Meg's Mirrors

zig

1	/*-
2	* SPDX-License-Identifier: BSD-3-Clause
3	*
4	* Copyright (c) 1986, 1989, 1991, 1993
5	* The Regents of the University of California. All rights reserved.
6	* (c) UNIX System Laboratories, Inc.
7	* All or some portions of this file are derived from material licensed
8	* to the University of California by American Telephone and Telegraph
9	* Co. or Unix System Laboratories, Inc. and are reproduced herein with
10	* the permission of UNIX System Laboratories, Inc.
11	*
12	* Redistribution and use in source and binary forms, with or without
13	* modification, are permitted provided that the following conditions
14	* are met:
15	* 1. Redistributions of source code must retain the above copyright
16	* notice, this list of conditions and the following disclaimer.
17	* 2. Redistributions in binary form must reproduce the above copyright
18	* notice, this list of conditions and the following disclaimer in the
19	* documentation and/or other materials provided with the distribution.
20	* 3. Neither the name of the University nor the names of its contributors
21	* may be used to endorse or promote products derived from this software
22	* without specific prior written permission.
23	*
24	* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
25	* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26	* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27	* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
28	* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29	* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
30	* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31	* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32	* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
33	* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34	* SUCH DAMAGE.
35	*/
36
37	#ifndef _SYS_PROC_H_
38	#define _SYS_PROC_H_
39
40	#include <sys/callout.h> /* For struct callout. */
41	#include <sys/event.h> /* For struct klist. */
42	#ifdef _KERNEL
43	#include <sys/_eventhandler.h>
44	#endif
45	#include <sys/_exterr.h>
46	#include <sys/condvar.h>
47	#ifndef _KERNEL
48	#include <sys/filedesc.h>
49	#endif
50	#include <sys/queue.h>
51	#include <sys/_lock.h>
52	#include <sys/lock_profile.h>
53	#include <sys/_mutex.h>
54	#include <sys/osd.h>
55	#include <sys/priority.h>
56	#include <sys/rtprio.h> /* XXX. */
57	#include <sys/resource.h>
58	#include <sys/sigio.h>
59	#include <sys/signal.h>
60	#include <sys/signalvar.h>
61	#ifndef _KERNEL
62	#include <sys/time.h> /* For structs itimerval, timeval. */
63	#else
64	#include <sys/pcpu.h>
65	#include <sys/systm.h>
66	#endif
67	#include <sys/ucontext.h>
68	#include <sys/ucred.h>
69	#include <sys/types.h>
70	#include <sys/_domainset.h>
71
72	#include <machine/proc.h> /* Machine-dependent proc substruct. */
73	#ifdef _KERNEL
74	#include <machine/cpu.h>
75	#endif
76
77	/*
78	* One structure allocated per session.
79	*
80	* List of locks
81	* (m) locked by s_mtx mtx
82	* (e) locked by proctree_lock sx
83	* (c) const until freeing
84	*/
85	struct session {
86	u_int s_count; /* Ref cnt; pgrps in session - atomic. */
87	struct proc *s_leader; /* (m + e) Session leader. */
88	struct vnode *s_ttyvp; /* (m) Vnode of controlling tty. */
89	struct cdev_priv *s_ttydp; /* (m) Device of controlling tty. */
90	struct tty *s_ttyp; /* (e) Controlling tty. */
91	pid_t s_sid; /* (c) Session ID. */
92	/* (m) Setlogin() name: */
93	char s_login[roundup(MAXLOGNAME, sizeof(long))];
94	struct mtx s_mtx; /* Mutex to protect members. */
95	};
96
97	/*
98	* One structure allocated per process group.
99	*
100	* List of locks
101	* (m) locked by pg_mtx mtx
102	* (e) locked by proctree_lock sx
103	* (c) const until freeing
104	*/
105	struct pgrp {
106	LIST_ENTRY(pgrp) pg_hash; /* (e) Hash chain. */
107	LIST_HEAD(, proc) pg_members; /* (m + e) Pointer to pgrp members. */
108	struct session *pg_session; /* (c) Pointer to session. */
109	struct sigiolst pg_sigiolst; /* (m) List of sigio sources. */
110	pid_t pg_id; /* (c) Process group id. */
111	struct mtx pg_mtx; /* Mutex to protect members */
112	int pg_flags; /* (m) PGRP_ flags */
113	struct sx pg_killsx; /* Mutual exclusion between group member
114	* fork() and killpg() */
115	};
116
117	#define PGRP_ORPHANED 0x00000001 /* Group is orphaned */
118
119	/*
120	* pargs, used to hold a copy of the command line, if it had a sane length.
121	*/
122	struct pargs {
123	u_int ar_ref; /* Reference count. */
124	u_int ar_length; /* Length. */
125	u_char ar_args[1]; /* Arguments. */
126	};
127
128	/*-
129	* Description of a process.
130	*
131	* This structure contains the information needed to manage a thread of
132	* control, known in UN*X as a process; it has references to substructures
133	* containing descriptions of things that the process uses, but may share
134	* with related processes. The process structure and the substructures
135	* are always addressable except for those marked "(CPU)" below,
136	* which might be addressable only on a processor on which the process
137	* is running.
138	*
139	* Below is a key of locks used to protect each member of struct proc. The
140	* lock is indicated by a reference to a specific character in parens in the
141	* associated comment.
142	* * - not yet protected
143	* a - only touched by curproc or parent during fork/wait
144	* b - created at fork, never changes
145	* (exception aiods switch vmspaces, but they are also
146	* marked 'P_SYSTEM' so hopefully it will be left alone)
147	* c - locked by proc mtx
148	* d - locked by allproc_lock lock
149	* e - locked by proctree_lock lock
150	* f - session mtx
151	* g - process group mtx
152	* h - callout_lock mtx
153	* i - by curproc or the master session mtx
154	* j - locked by proc slock
155	* k - only accessed by curthread
156	* k*- only accessed by curthread and from an interrupt
157	* kx- only accessed by curthread and by debugger
158	* l - the attaching proc or attaching proc parent
159	* n - not locked, lazy
160	* o - ktrace lock
161	* q - td_contested lock
162	* r - p_peers lock
163	* s - see sleepq_switch(), sleeping_on_old_rtc(), and sleep(9)
164	* t - thread lock
165	* u - process stat lock
166	* w - process timer lock
167	* x - created at fork, only changes during single threading in exec
168	* y - created at first aio, doesn't change until exit or exec at which
169	* point we are single-threaded and only curthread changes it
170	*
171	* If the locking key specifies two identifiers (for example, p_pptr) then
172	* either lock is sufficient for read access, but both locks must be held
173	* for write access.
174	*/
175	struct cpuset;
176	struct filecaps;
177	struct filemon;
178	struct kaioinfo;
179	struct kaudit_record;
180	struct kcov_info;
181	struct kdtrace_proc;
182	struct kdtrace_thread;
183	struct kmsan_td;
184	struct kq_timer_cb_data;
185	struct mqueue_notifier;
186	struct p_sched;
187	struct proc;
188	struct procdesc;
189	struct racct;
190	struct sbuf;
191	struct sleepqueue;
192	struct socket;
193	struct td_sched;
194	struct thread;
195	struct trapframe;
196	struct turnstile;
197	struct vm_map;
198	struct vm_map_entry;
199	struct epoch_tracker;
200
201	struct syscall_args {
202	u_int code;
203	u_int original_code;
204	struct sysent *callp;
205	register_t args[8];
206	};
207
208	/*
209	* XXX: Does this belong in resource.h or resourcevar.h instead?
210	* Resource usage extension. The times in rusage structs in the kernel are
211	* never up to date. The actual times are kept as runtimes and tick counts
212	* (with control info in the "previous" times), and are converted when
213	* userland asks for rusage info. Backwards compatibility prevents putting
214	* this directly in the user-visible rusage struct.
215	*
216	* Locking for p_rux: (cu) means (u) for p_rux and (c) for p_crux.
217	* Locking for td_rux: (t) for all fields.
218	*/
219	struct rusage_ext {
220	uint64_t rux_runtime; /* (cu) Real time. */
221	uint64_t rux_uticks; /* (cu) Statclock hits in user mode. */
222	uint64_t rux_sticks; /* (cu) Statclock hits in sys mode. */
223	uint64_t rux_iticks; /* (cu) Statclock hits in intr mode. */
224	uint64_t rux_uu; /* (c) Previous user time in usec. */
225	uint64_t rux_su; /* (c) Previous sys time in usec. */
226	uint64_t rux_tu; /* (c) Previous total time in usec. */
227	};
228
229	/*
230	* Kernel runnable context (thread).
231	* This is what is put to sleep and reactivated.
232	* Thread context. Processes may have multiple threads.
233	*/
234	struct thread {
235	struct mtx *volatile td_lock; /* replaces sched lock */
236	struct proc *td_proc; /* (*) Associated process. */
237	TAILQ_ENTRY(thread) td_plist; /* (*) All threads in this proc. */
238	TAILQ_ENTRY(thread) td_runq; /* (t) Run queue. */
239	union {
240	TAILQ_ENTRY(thread) td_slpq; /* (t) Sleep queue. */
241	struct thread *td_zombie; /* Zombie list linkage */
242	};
243	TAILQ_ENTRY(thread) td_lockq; /* (t) Lock queue. */
244	LIST_ENTRY(thread) td_hash; /* (d) Hash chain. */
245	struct cpuset *td_cpuset; /* (t) CPU affinity mask. */
246	struct domainset_ref td_domain; /* (a) NUMA policy */
247	struct seltd *td_sel; /* Select queue/channel. */
248	struct sleepqueue *td_sleepqueue; /* (k) Associated sleep queue. */
249	struct turnstile *td_turnstile; /* (k) Associated turnstile. */
250	void *td_pad1; /* Available */
251	struct umtx_q *td_umtxq; /* (c?) Link for when we're blocked. */
252	lwpid_t td_tid; /* (b) Thread ID. */
253	sigqueue_t td_sigqueue; /* (c) Sigs arrived, not delivered. */
254	#define td_siglist td_sigqueue.sq_signals
255	u_char td_lend_user_pri; /* (t) Lend user pri. */
256	u_char td_allocdomain; /* (b) NUMA domain backing this struct thread. */
257	u_char td_base_ithread_pri; /* (t) Base ithread pri */
258	struct kmsan_td *td_kmsan; /* (k) KMSAN state */
259
260	/* Cleared during fork1(), thread_create(), or kthread_add(). */
261	#define td_startzero td_flags
262	int td_flags; /* (t) TDF_* flags. */
263	int td_ast; /* (t) TDA_* indicators */
264	int td_inhibitors; /* (t) Why can not run. */
265	int td_pflags; /* (k) Private thread (TDP_*) flags. */
266	int td_pflags2; /* (k) Private thread (TDP2_*) flags. */
267	int td_dupfd; /* (k) Ret value from fdopen. XXX */
268	int td_sqqueue; /* (t) Sleepqueue queue blocked on. */
269	const void *td_wchan; /* (t) Sleep address. */
270	const char *td_wmesg; /* (t) Reason for sleep. */
271	volatile u_char td_owepreempt; /* (k*) Preempt on last critical_exit */
272	u_char td_tsqueue; /* (t) Turnstile queue blocked on. */
273	u_char _td_pad0[2]; /* Available. */
274	int td_locks; /* (k) Debug: count of non-spin locks */
275	int td_rw_rlocks; /* (k) Count of rwlock read locks. */
276	int td_sx_slocks; /* (k) Count of sx shared locks. */
277	int td_lk_slocks; /* (k) Count of lockmgr shared locks. */
278	struct lock_object *td_wantedlock; /* (k) Lock we are contending on */
279	struct turnstile *td_blocked; /* (t) Lock thread is blocked on. */
280	const char *td_lockname; /* (t) Name of lock blocked on. */
281	LIST_HEAD(, turnstile) td_contested; /* (q) Contested locks. */
282	struct lock_list_entry *td_sleeplocks; /* (k) Held sleep locks. */
283	int td_intr_nesting_level; /* (k) Interrupt recursion. */
284	int td_pinned; /* (k) Temporary cpu pin count. */
285	struct ucred *td_realucred; /* (k) Reference to credentials. */
286	struct ucred *td_ucred; /* (k) Used credentials, temporarily switchable. */
287	struct plimit *td_limit; /* (k) Resource limits. */
288	int td_slptick; /* (t) Time at sleep. */
289	int td_blktick; /* (t) Time spent blocked. */
290	int td_swvoltick; /* (t) Time at last SW_VOL switch. */
291	int td_swinvoltick; /* (t) Time at last SW_INVOL switch. */
292	u_int td_cow; /* (*) Number of copy-on-write faults */
293	struct rusage td_ru; /* (t) rusage information. */
294	struct rusage_ext td_rux; /* (t) Internal rusage information. */
295	uint64_t td_incruntime; /* (t) Cpu ticks to transfer to proc. */
296	uint64_t td_runtime; /* (t) How many cpu ticks we've run. */
297	u_int td_pticks; /* (t) Statclock hits for profiling */
298	u_int td_sticks; /* (t) Statclock hits in system mode. */
299	u_int td_iticks; /* (t) Statclock hits in intr mode. */
300	u_int td_uticks; /* (t) Statclock hits in user mode. */
301	int td_intrval; /* (t) Return value for sleepq. */
302	sigset_t td_oldsigmask; /* (k) Saved mask from pre sigpause. */
303	volatile u_int td_generation; /* (k) For detection of preemption */
304	stack_t td_sigstk; /* (k) Stack ptr and on-stack flag. */
305	int td_xsig; /* (c) Signal for ptrace */
306	u_long td_profil_addr; /* (k) Temporary addr until AST. */
307	u_int td_profil_ticks; /* (k) Temporary ticks until AST. */
308	char td_name[MAXCOMLEN + 1]; /* (*) Thread name. */
309	struct file *td_fpop; /* (k) file referencing cdev under op */
310	int td_dbgflags; /* (c) Userland debugger flags */
311	siginfo_t td_si; /* (c) For debugger or core file */
312	int td_ng_outbound; /* (k) Thread entered ng from above. */
313	struct osd td_osd; /* (k) Object specific data. */
314	struct vm_map_entry *td_map_def_user; /* (k) Deferred entries. */
315	pid_t td_dbg_forked; /* (c) Child pid for debugger. */
316	u_int td_no_sleeping; /* (k) Sleeping disabled count. */
317	struct vnode *td_vp_reserved;/* (k) Preallocated vnode. */
318	void *td_su; /* (k) FFS SU private */
319	sbintime_t td_sleeptimo; /* (t) Sleep timeout. */
320	int td_rtcgen; /* (s) rtc_generation of abs. sleep */
321	int td_errno; /* (k) Error from last syscall. */
322	size_t td_vslock_sz; /* (k) amount of vslock-ed space */
323	struct kcov_info *td_kcov_info; /* (*) Kernel code coverage data */
324	long td_ucredref; /* (k) references on td_realucred */
325	struct kexterr td_kexterr;
326	#define td_endzero td_sigmask
327
328	/* Copied during fork1(), thread_create(), or kthread_add(). */
329	#define td_startcopy td_endzero
330	sigset_t td_sigmask; /* (c) Current signal mask. */
331	u_char td_rqindex; /* (t) Run queue index. */
332	u_char td_base_pri; /* (t) Thread base kernel priority. */
333	u_char td_priority; /* (t) Thread active priority. */
334	u_char td_pri_class; /* (t) Scheduling class. */
335	u_char td_user_pri; /* (t) User pri from estcpu and nice. */
336	u_char td_base_user_pri; /* (t) Base user pri */
337	uintptr_t td_rb_list; /* (k) Robust list head. */
338	uintptr_t td_rbp_list; /* (k) Robust priv list head. */
339	uintptr_t td_rb_inact; /* (k) Current in-action mutex loc. */
340	struct syscall_args td_sa; /* (kx) Syscall parameters. Copied on
341	fork for child tracing. */
342	void *td_sigblock_ptr; /* (k) uptr for fast sigblock. */
343	uint32_t td_sigblock_val; /* (k) fast sigblock value read at
344	td_sigblock_ptr on kern entry */
345	void *td_exterr_ptr;
346	#define td_endcopy td_pcb
347
348	/*
349	* Fields that must be manually set in fork1(), thread_create(), kthread_add(),
350	* or already have been set in the allocator, constructor, etc.
351	*/
352	struct pcb *td_pcb; /* (k) Kernel VA of pcb and kstack. */
353	enum td_states {
354	TDS_INACTIVE = 0x0,
355	TDS_INHIBITED,
356	TDS_CAN_RUN,
357	TDS_RUNQ,
358	TDS_RUNNING
359	} td_state; /* (t) thread state */
360	/* Note: td_state must be accessed using TD_{GET,SET}_STATE(). */
361	union {
362	syscallarg_t tdu_retval[2];
363	off_t tdu_off;
364	} td_uretoff; /* (k) Syscall aux returns. */
365	#define td_retval td_uretoff.tdu_retval
366	u_int td_cowgen; /* (k) Generation of COW pointers. */
367	/* LP64 hole */
368	struct callout td_slpcallout; /* (h) Callout for sleep. */
369	struct trapframe *td_frame; /* (k) */
370	vm_offset_t td_kstack; /* (a) Kernel VA of kstack. */
371	u_short td_kstack_pages; /* (a) Size of the kstack. */
372	u_short td_kstack_domain; /* (a) Domain backing kstack KVA. */
373	volatile u_int td_critnest; /* (k*) Critical section nest level. */
374	struct mdthread td_md; /* (k) Any machine-dependent fields. */
375	struct kaudit_record *td_ar; /* (k) Active audit record, if any. */
376	struct lpohead td_lprof[2]; /* (a) lock profiling objects. */
377	struct kdtrace_thread *td_dtrace; /* (*) DTrace-specific data. */
378	struct vnet *td_vnet; /* (k) Effective vnet. */
379	const char *td_vnet_lpush; /* (k) Debugging vnet push / pop. */
380	struct trapframe *td_intr_frame;/* (k) Frame of the current irq */
381	struct proc *td_rfppwait_p; /* (k) The vforked child */
382	struct vm_page **td_ma; /* (k) uio pages held */
383	int td_ma_cnt; /* (k) size of *td_ma */
384	/* LP64 hole */
385	void *td_emuldata; /* Emulator state data */
386	int td_lastcpu; /* (t) Last cpu we were on. */
387	int td_oncpu; /* (t) Which cpu we are on. */
388	void *td_lkpi_task; /* LinuxKPI task struct pointer */
389	int td_pmcpend;
390	void *td_remotereq; /* (c) dbg remote request. */
391	off_t td_ktr_io_lim; /* (k) limit for ktrace file size */
392	#ifdef EPOCH_TRACE
393	SLIST_HEAD(, epoch_tracker) td_epochs;
394	#endif
395	};
396
397	struct thread0_storage {
398	struct thread t0st_thread;
399	uint64_t t0st_sched[10];
400	};
401
402	struct mtx *thread_lock_block(struct thread *);
403	void thread_lock_block_wait(struct thread *);
404	void thread_lock_set(struct thread *, struct mtx *);
405	void thread_lock_unblock(struct thread *, struct mtx *);
406	#define THREAD_LOCK_ASSERT(td, type) \
407	mtx_assert((td)->td_lock, (type))
408
409	#define THREAD_LOCK_BLOCKED_ASSERT(td, type) \
410	do { \
411	struct mtx *__m = (td)->td_lock; \
412	if (__m != &blocked_lock) \
413	mtx_assert(__m, (type)); \
414	} while (0)
415
416	#ifdef INVARIANTS
417	#define THREAD_LOCKPTR_ASSERT(td, lock) \
418	do { \
419	struct mtx *__m; \
420	__m = (td)->td_lock; \
421	KASSERT(__m == (lock), \
422	("Thread %p lock %p does not match %p", td, __m, (lock))); \
423	} while (0)
424
425	#define THREAD_LOCKPTR_BLOCKED_ASSERT(td, lock) \
426	do { \
427	struct mtx *__m; \
428	__m = (td)->td_lock; \
429	KASSERT(__m == (lock) || __m == &blocked_lock, \
430	("Thread %p lock %p does not match %p", td, __m, (lock))); \
431	} while (0)
432
433	#define TD_LOCKS_INC(td) ((td)->td_locks++)
434	#define TD_LOCKS_DEC(td) do { \
435	KASSERT(SCHEDULER_STOPPED() || (td)->td_locks > 0, \
436	("Thread %p owns no locks", (td))); \
437	(td)->td_locks--; \
438	} while (0)
439	#else
440	#define THREAD_LOCKPTR_ASSERT(td, lock)
441	#define THREAD_LOCKPTR_BLOCKED_ASSERT(td, lock)
442
443	#define TD_LOCKS_INC(td)
444	#define TD_LOCKS_DEC(td)
445	#endif
446
447	/*
448	* Flags kept in td_flags:
449	* To change these you MUST have the scheduler lock.
450	*/
451	#define TDF_BORROWING 0x00000001 /* Thread is borrowing pri from another. */
452	#define TDF_INPANIC 0x00000002 /* Caused a panic, let it drive crashdump. */
453	#define TDF_INMEM 0x00000004 /* Thread's stack is in memory. */
454	#define TDF_SINTR 0x00000008 /* Sleep is interruptible. */
455	#define TDF_TIMEOUT 0x00000010 /* Timing out during sleep. */
456	#define TDF_IDLETD 0x00000020 /* This is a per-CPU idle thread. */
457	#define TDF_UNUSED11 0x00000040 /* Available */
458	#define TDF_SIGWAIT 0x00000080 /* Ignore ignored signals */
459	#define TDF_KTH_SUSP 0x00000100 /* kthread is suspended */
460	#define TDF_ALLPROCSUSP 0x00000200 /* suspended by SINGLE_ALLPROC */
461	#define TDF_BOUNDARY 0x00000400 /* Thread suspended at user boundary */
462	#define TDF_UNUSED1 0x00000800 /* Available */
463	#define TDF_UNUSED2 0x00001000 /* Available */
464	#define TDF_SBDRY 0x00002000 /* Stop only on usermode boundary. */
465	#define TDF_UPIBLOCKED 0x00004000 /* Thread blocked on user PI mutex. */
466	#define TDF_UNUSED3 0x00008000 /* Available */
467	#define TDF_UNUSED4 0x00010000 /* Available */
468	#define TDF_UNUSED5 0x00020000 /* Available */
469	#define TDF_NOLOAD 0x00040000 /* Ignore during load avg calculations. */
470	#define TDF_SERESTART 0x00080000 /* ERESTART on stop attempts. */
471	#define TDF_THRWAKEUP 0x00100000 /* Libthr thread must not suspend itself. */
472	#define TDF_SEINTR 0x00200000 /* EINTR on stop attempts. */
473	#define TDF_UNUSED12 0x00400000 /* Available */
474	#define TDF_UNUSED6 0x00800000 /* Available */
475	#define TDF_SCHED0 0x01000000 /* Reserved for scheduler private use */
476	#define TDF_SCHED1 0x02000000 /* Reserved for scheduler private use */
477	#define TDF_SCHED2 0x04000000 /* Reserved for scheduler private use */
478	#define TDF_SCHED3 0x08000000 /* Reserved for scheduler private use */
479	#define TDF_UNUSED7 0x10000000 /* Available */
480	#define TDF_UNUSED8 0x20000000 /* Available */
481	#define TDF_UNUSED9 0x40000000 /* Available */
482	#define TDF_UNUSED10 0x80000000 /* Available */
483
484	enum {
485	TDA_AST = 0, /* Special: call all non-flagged AST handlers */
486	TDA_OWEUPC,
487	TDA_HWPMC,
488	TDA_VFORK,
489	TDA_ALRM,
490	TDA_PROF,
491	TDA_MAC,
492	TDA_SCHED,
493	TDA_UFS,
494	TDA_GEOM,
495	TDA_KQUEUE,
496	TDA_RACCT,
497	TDA_MOD1, /* For third party use, before signals are */
498	TDA_MOD2, /* processed .. */
499	TDA_PSELECT, /* For discarding temporary signal mask */
500	TDA_SIG,
501	TDA_KTRACE,
502	TDA_SUSPEND,
503	TDA_SIGSUSPEND,
504	TDA_MOD3, /* .. and after */
505	TDA_MOD4,
506	TDA_MAX,
507	};
508	#define TDAI(tda) (1U << (tda))
509	#define td_ast_pending(td, tda) ((td->td_ast & TDAI(tda)) != 0)
510
511	/* Userland debug flags */
512	#define TDB_SUSPEND 0x00000001 /* Thread is suspended by debugger */
513	#define TDB_XSIG 0x00000002 /* Thread is exchanging signal under trace */
514	#define TDB_USERWR 0x00000004 /* Debugger modified memory or registers */
515	#define TDB_SCE 0x00000008 /* Thread performs syscall enter */
516	#define TDB_SCX 0x00000010 /* Thread performs syscall exit */
517	#define TDB_EXEC 0x00000020 /* TDB_SCX from exec(2) family */
518	#define TDB_FORK 0x00000040 /* TDB_SCX from fork(2) that created new
519	process */
520	#define TDB_STOPATFORK 0x00000080 /* Stop at the return from fork (child
521	only) */
522	#define TDB_CHILD 0x00000100 /* New child indicator for ptrace() */
523	#define TDB_BORN 0x00000200 /* New LWP indicator for ptrace() */
524	#define TDB_EXIT 0x00000400 /* Exiting LWP indicator for ptrace() */
525	#define TDB_VFORK 0x00000800 /* vfork indicator for ptrace() */
526	#define TDB_FSTP 0x00001000 /* The thread is PT_ATTACH leader */
527	#define TDB_STEP 0x00002000 /* (x86) PSL_T set for PT_STEP */
528	#define TDB_SSWITCH 0x00004000 /* Suspended in ptracestop */
529	#define TDB_BOUNDARY 0x00008000 /* ptracestop() at boundary */
530	#define TDB_COREDUMPREQ 0x00010000 /* Coredump request */
531	#define TDB_SCREMOTEREQ 0x00020000 /* Remote syscall request */
532
533	/*
534	* "Private" flags kept in td_pflags:
535	* These are only written by curthread and thus need no locking.
536	*/
537	#define TDP_OLDMASK 0x00000001 /* Need to restore mask after suspend. */
538	#define TDP_INKTR 0x00000002 /* Thread is currently in KTR code. */
539	#define TDP_INKTRACE 0x00000004 /* Thread is currently in KTRACE code. */
540	#define TDP_BUFNEED 0x00000008 /* Do not recurse into the buf flush */
541	#define TDP_COWINPROGRESS 0x00000010 /* Snapshot copy-on-write in progress. */
542	#define TDP_ALTSTACK 0x00000020 /* Have alternate signal stack. */
543	#define TDP_DEADLKTREAT 0x00000040 /* Lock acquisition - deadlock treatment. */
544	#define TDP_NOFAULTING 0x00000080 /* Do not handle page faults. */
545	#define TDP_SIGFASTBLOCK 0x00000100 /* Fast sigblock active */
546	#define TDP_OWEUPC 0x00000200 /* Call addupc() at next AST. */
547	#define TDP_ITHREAD 0x00000400 /* Thread is an interrupt thread. */
548	#define TDP_SYNCIO 0x00000800 /* Local override, disable async i/o. */
549	#define TDP_SCHED1 0x00001000 /* Reserved for scheduler private use */
550	#define TDP_SCHED2 0x00002000 /* Reserved for scheduler private use */
551	#define TDP_SCHED3 0x00004000 /* Reserved for scheduler private use */
552	#define TDP_SCHED4 0x00008000 /* Reserved for scheduler private use */
553	#define TDP_GEOM 0x00010000 /* Settle GEOM before finishing syscall */
554	#define TDP_SOFTDEP 0x00020000 /* Stuck processing softdep worklist */
555	#define TDP_NORUNNINGBUF 0x00040000 /* Ignore runningbufspace check */
556	#define TDP_WAKEUP 0x00080000 /* Don't sleep in umtx cond_wait */
557	#define TDP_INBDFLUSH 0x00100000 /* Already in BO_BDFLUSH, do not recurse */
558	#define TDP_KTHREAD 0x00200000 /* This is an official kernel thread */
559	#define TDP_CALLCHAIN 0x00400000 /* Capture thread's callchain */
560	#define TDP_IGNSUSP 0x00800000 /* Permission to ignore the MNTK_SUSPEND* */
561	#define TDP_AUDITREC 0x01000000 /* Audit record pending on thread */
562	#define TDP_RFPPWAIT 0x02000000 /* Handle RFPPWAIT on syscall exit */
563	#define TDP_RESETSPUR 0x04000000 /* Reset spurious page fault history. */
564	#define TDP_NERRNO 0x08000000 /* Last errno is already in td_errno */
565	#define TDP_UIOHELD 0x10000000 /* Current uio has pages held in td_ma */
566	#define TDP_EFIRT 0x20000000 /* In firmware (EFI RT) call */
567	#define TDP_EXECVMSPC 0x40000000 /* Execve destroyed old vmspace */
568	#define TDP_SIGFASTPENDING 0x80000000 /* Pending signal due to sigfastblock */
569
570	#define TDP2_SBPAGES 0x00000001 /* Owns sbusy on some pages */
571	#define TDP2_COMPAT32RB 0x00000002 /* compat32 ABI for robust lists */
572	#define TDP2_ACCT 0x00000004 /* Doing accounting */
573	#define TDP2_SAN_QUIET 0x00000008 /* Disable warnings from K(A|M)SAN */
574	#define TDP2_EXTERR 0x00000010 /* Kernel reported ext error */
575	#define TDP2_UEXTERR 0x00000020 /* User set ext error reporting ptr */
576
577	/*
578	* Reasons that the current thread can not be run yet.
579	* More than one may apply.
580	*/
581	#define TDI_SUSPENDED 0x0001 /* On suspension queue. */
582	#define TDI_SLEEPING 0x0002 /* Actually asleep! (tricky). */
583	#define TDI_LOCK 0x0008 /* Stopped on a lock. */
584	#define TDI_IWAIT 0x0010 /* Awaiting interrupt. */
585
586	#define TD_IS_SLEEPING(td) ((td)->td_inhibitors & TDI_SLEEPING)
587	#define TD_ON_SLEEPQ(td) ((td)->td_wchan != NULL)
588	#define TD_IS_SUSPENDED(td) ((td)->td_inhibitors & TDI_SUSPENDED)
589	#define TD_ON_LOCK(td) ((td)->td_inhibitors & TDI_LOCK)
590	#define TD_AWAITING_INTR(td) ((td)->td_inhibitors & TDI_IWAIT)
591	#ifdef _KERNEL
592	#define TD_GET_STATE(td) atomic_load_int(&(td)->td_state)
593	#else
594	#define TD_GET_STATE(td) ((td)->td_state)
595	#endif
596	#define TD_IS_RUNNING(td) (TD_GET_STATE(td) == TDS_RUNNING)
597	#define TD_ON_RUNQ(td) (TD_GET_STATE(td) == TDS_RUNQ)
598	#define TD_CAN_RUN(td) (TD_GET_STATE(td) == TDS_CAN_RUN)
599	#define TD_IS_INHIBITED(td) (TD_GET_STATE(td) == TDS_INHIBITED)
600	#define TD_ON_UPILOCK(td) ((td)->td_flags & TDF_UPIBLOCKED)
601	#define TD_IS_IDLETHREAD(td) ((td)->td_flags & TDF_IDLETD)
602
603	#define TD_CAN_ABORT(td) (TD_ON_SLEEPQ((td)) && \
604	((td)->td_flags & TDF_SINTR) != 0)
605
606	#define KTDSTATE(td) \
607	(((td)->td_inhibitors & TDI_SLEEPING) != 0 ? "sleep" : \
608	((td)->td_inhibitors & TDI_SUSPENDED) != 0 ? "suspended" : \
609	((td)->td_inhibitors & TDI_LOCK) != 0 ? "blocked" : \
610	((td)->td_inhibitors & TDI_IWAIT) != 0 ? "iwait" : "yielding")
611
612	#define TD_SET_INHIB(td, inhib) do { \
613	TD_SET_STATE(td, TDS_INHIBITED); \
614	(td)->td_inhibitors |= (inhib); \
615	} while (0)
616
617	#define TD_CLR_INHIB(td, inhib) do { \
618	if (((td)->td_inhibitors & (inhib)) && \
619	(((td)->td_inhibitors &= ~(inhib)) == 0)) \
620	TD_SET_STATE(td, TDS_CAN_RUN); \
621	} while (0)
622
623	#define TD_SET_SLEEPING(td) TD_SET_INHIB((td), TDI_SLEEPING)
624	#define TD_SET_LOCK(td) TD_SET_INHIB((td), TDI_LOCK)
625	#define TD_SET_SUSPENDED(td) TD_SET_INHIB((td), TDI_SUSPENDED)
626	#define TD_SET_IWAIT(td) TD_SET_INHIB((td), TDI_IWAIT)
627	#define TD_SET_EXITING(td) TD_SET_INHIB((td), TDI_EXITING)
628
629	#define TD_CLR_SLEEPING(td) TD_CLR_INHIB((td), TDI_SLEEPING)
630	#define TD_CLR_LOCK(td) TD_CLR_INHIB((td), TDI_LOCK)
631	#define TD_CLR_SUSPENDED(td) TD_CLR_INHIB((td), TDI_SUSPENDED)
632	#define TD_CLR_IWAIT(td) TD_CLR_INHIB((td), TDI_IWAIT)
633
634	#ifdef _KERNEL
635	#define TD_SET_STATE(td, state) atomic_store_int(&(td)->td_state, state)
636	#else
637	#define TD_SET_STATE(td, state) (td)->td_state = state
638	#endif
639	#define TD_SET_RUNNING(td) TD_SET_STATE(td, TDS_RUNNING)
640	#define TD_SET_RUNQ(td) TD_SET_STATE(td, TDS_RUNQ)
641	#define TD_SET_CAN_RUN(td) TD_SET_STATE(td, TDS_CAN_RUN)
642
643
644	#define TD_SBDRY_INTR(td) \
645	(((td)->td_flags & (TDF_SEINTR | TDF_SERESTART)) != 0)
646	#define TD_SBDRY_ERRNO(td) \
647	(((td)->td_flags & TDF_SEINTR) != 0 ? EINTR : ERESTART)
648
649	/*
650	* Process structure.
651	*/
652	struct proc {
653	LIST_ENTRY(proc) p_list; /* (d) List of all processes. */
654	TAILQ_HEAD(, thread) p_threads; /* (c) all threads. */
655	struct mtx p_slock; /* process spin lock */
656	struct ucred *p_ucred; /* (c) Process owner's identity. */
657	struct filedesc *p_fd; /* (b) Open files. */
658	struct filedesc_to_leader *p_fdtol; /* (b) Tracking node */
659	struct pwddesc *p_pd; /* (b) Cwd, chroot, jail, umask */
660	struct pstats *p_stats; /* (b) Accounting/statistics (CPU). */
661	struct plimit *p_limit; /* (c) Resource limits. */
662	struct callout p_limco; /* (c) Limit callout handle */
663	struct sigacts *p_sigacts; /* (x) Signal actions, state (CPU). */
664
665	int p_flag; /* (c) P_* flags. */
666	int p_flag2; /* (c) P2_* flags. */
667	enum p_states {
668	PRS_NEW = 0, /* In creation */
669	PRS_NORMAL, /* threads can be run. */
670	PRS_ZOMBIE
671	} p_state; /* (j/c) Process status. */
672	pid_t p_pid; /* (b) Process identifier. */
673	LIST_ENTRY(proc) p_hash; /* (d) Hash chain. */
674	LIST_ENTRY(proc) p_pglist; /* (g + e) List of processes in pgrp. */
675	struct proc *p_pptr; /* (c + e) Pointer to parent process. */
676	LIST_ENTRY(proc) p_sibling; /* (e) List of sibling processes. */
677	LIST_HEAD(, proc) p_children; /* (e) Pointer to list of children. */
678	struct proc *p_reaper; /* (e) My reaper. */
679	LIST_HEAD(, proc) p_reaplist; /* (e) List of my descendants
680	(if I am reaper). */
681	LIST_ENTRY(proc) p_reapsibling; /* (e) List of siblings - descendants of
682	the same reaper. */
683	struct mtx p_mtx; /* (n) Lock for this struct. */
684	struct mtx p_statmtx; /* Lock for the stats */
685	struct mtx p_itimmtx; /* Lock for the virt/prof timers */
686	struct mtx p_profmtx; /* Lock for the profiling */
687	struct ksiginfo *p_ksi; /* Locked by parent proc lock */
688	sigqueue_t p_sigqueue; /* (c) Sigs not delivered to a td. */
689	#define p_siglist p_sigqueue.sq_signals
690	pid_t p_oppid; /* (c + e) Real parent pid. */
691
692	/* The following fields are all zeroed upon creation in fork. */
693	#define p_startzero p_vmspace
694	struct vmspace *p_vmspace; /* (b) Address space. */
695	u_int p_swtick; /* (c) Tick when swapped in or out. */
696	u_int p_cowgen; /* (c) Generation of COW pointers. */
697	struct itimerval p_realtimer; /* (c) Alarm timer. */
698	struct rusage p_ru; /* (a) Exit information. */
699	struct rusage_ext p_rux; /* (cu) Internal resource usage. */
700	struct rusage_ext p_crux; /* (c) Internal child resource usage. */
701	int p_profthreads; /* (c) Num threads in addupc_task. */
702	volatile int p_exitthreads; /* (j) Number of threads exiting */
703	int p_traceflag; /* (o) Kernel trace points. */
704	struct ktr_io_params *p_ktrioparms; /* (c + o) Params for ktrace. */
705	struct vnode *p_textvp; /* (b) Vnode of executable. */
706	struct vnode *p_textdvp; /* (b) Dir containing textvp. */
707	char *p_binname; /* (b) Binary hardlink name. */
708	u_int p_lock; /* (c) Prevent exit. */
709	struct sigiolst p_sigiolst; /* (c) List of sigio sources. */
710	int p_sigparent; /* (c) Signal to parent on exit. */
711	int p_sig; /* (n) For core dump/debugger XXX. */
712	u_int p_ptevents; /* (c + e) ptrace() event mask. */
713	struct kaioinfo *p_aioinfo; /* (y) ASYNC I/O info. */
714	struct thread *p_singlethread;/* (c + j) If single threading this is it */
715	int p_suspcount; /* (j) Num threads in suspended mode. */
716	struct thread *p_xthread; /* (c) Trap thread */
717	int p_boundary_count;/* (j) Num threads at user boundary */
718	int p_pendingcnt; /* (c) how many signals are pending */
719	struct itimers *p_itimers; /* (c) POSIX interval timers. */
720	struct procdesc *p_procdesc; /* (e) Process descriptor, if any. */
721	u_int p_treeflag; /* (e) P_TREE flags */
722	int p_pendingexits; /* (c) Count of pending thread exits. */
723	struct filemon *p_filemon; /* (c) filemon-specific data. */
724	int p_pdeathsig; /* (c) Signal from parent on exit. */
725	/* End area that is zeroed on creation. */
726	#define p_endzero p_magic
727
728	/* The following fields are all copied upon creation in fork. */
729	#define p_startcopy p_endzero
730	u_int p_magic; /* (b) Magic number. */
731	int p_osrel; /* (x) osreldate for the
732	binary (from ELF note, if any) */
733	uint32_t p_fctl0; /* (x) ABI feature control, ELF note */
734	char p_comm[MAXCOMLEN + 1]; /* (x) Process name. */
735	struct sysentvec *p_sysent; /* (b) Syscall dispatch info. */
736	struct pargs *p_args; /* (c) Process arguments. */
737	rlim_t p_cpulimit; /* (c) Current CPU limit in seconds. */
738	signed char p_nice; /* (c) Process "nice" value. */
739	int p_fibnum; /* in this routing domain XXX MRT */
740	pid_t p_reapsubtree; /* (e) Pid of the direct child of the
741	reaper which spawned
742	our subtree. */
743	uint64_t p_elf_flags; /* (x) ELF flags */
744	void *p_elf_brandinfo; /* (x) Elf_Brandinfo, NULL for
745	non ELF binaries. */
746	sbintime_t p_umtx_min_timeout;
747	/* End area that is copied on creation. */
748	#define p_endcopy p_xexit
749
750	u_int p_xexit; /* (c) Exit code. */
751	u_int p_xsig; /* (c) Stop/kill sig. */
752	struct pgrp *p_pgrp; /* (c + e) Pointer to process group. */
753	struct knlist *p_klist; /* (c) Knotes attached to this proc. */
754	int p_numthreads; /* (c) Number of threads. */
755	struct mdproc p_md; /* Any machine-dependent fields. */
756	struct callout p_itcallout; /* (h + c) Interval timer callout. */
757	u_short p_acflag; /* (c) Accounting flags. */
758	struct proc *p_peers; /* (r) */
759	struct proc *p_leader; /* (b) */
760	void *p_emuldata; /* (c) Emulator state data. */
761	struct label *p_label; /* (*) Proc (not subject) MAC label. */
762	STAILQ_HEAD(, ktr_request) p_ktr; /* (o) KTR event queue. */
763	LIST_HEAD(, mqueue_notifier) p_mqnotifier; /* (c) mqueue notifiers.*/
764	struct kdtrace_proc *p_dtrace; /* (*) DTrace-specific data. */
765	struct cv p_pwait; /* (*) wait cv for exit/exec. */
766	struct racct *p_racct; /* (b) Resource accounting. */
767	int p_throttled; /* (c) Flag for racct pcpu throttling */
768	/*
769	* An orphan is the child that has been re-parented to the
770	* debugger as a result of attaching to it. Need to keep
771	* track of them for parent to be able to collect the exit
772	* status of what used to be children.
773	*/
774	LIST_ENTRY(proc) p_orphan; /* (e) List of orphan processes. */
775	LIST_HEAD(, proc) p_orphans; /* (e) Pointer to list of orphans. */
776
777	TAILQ_HEAD(, kq_timer_cb_data) p_kqtim_stop; /* (c) */
778	LIST_ENTRY(proc) p_jaillist; /* (d) Jail process linkage. */
779	};
780
781	#define p_session p_pgrp->pg_session
782	#define p_pgid p_pgrp->pg_id
783
784	#define NOCPU (-1) /* For when we aren't on a CPU. */
785	#define NOCPU_OLD (255)
786	#define MAXCPU_OLD (254)
787
788	#define PROC_SLOCK(p) mtx_lock_spin(&(p)->p_slock)
789	#define PROC_SUNLOCK(p) mtx_unlock_spin(&(p)->p_slock)
790	#define PROC_SLOCK_ASSERT(p, type) mtx_assert(&(p)->p_slock, (type))
791
792	#define PROC_STATLOCK(p) mtx_lock_spin(&(p)->p_statmtx)
793	#define PROC_STATUNLOCK(p) mtx_unlock_spin(&(p)->p_statmtx)
794	#define PROC_STATLOCK_ASSERT(p, type) mtx_assert(&(p)->p_statmtx, (type))
795
796	#define PROC_ITIMLOCK(p) mtx_lock_spin(&(p)->p_itimmtx)
797	#define PROC_ITIMUNLOCK(p) mtx_unlock_spin(&(p)->p_itimmtx)
798	#define PROC_ITIMLOCK_ASSERT(p, type) mtx_assert(&(p)->p_itimmtx, (type))
799
800	#define PROC_PROFLOCK(p) mtx_lock_spin(&(p)->p_profmtx)
801	#define PROC_PROFUNLOCK(p) mtx_unlock_spin(&(p)->p_profmtx)
802	#define PROC_PROFLOCK_ASSERT(p, type) mtx_assert(&(p)->p_profmtx, (type))
803
804	/* These flags are kept in p_flag. */
805	#define P_ADVLOCK 0x00000001 /* Process may hold a POSIX advisory
806	lock. */
807	#define P_CONTROLT 0x00000002 /* Has a controlling terminal. */
808	#define P_KPROC 0x00000004 /* Kernel process. */
809	#define P_IDLEPROC 0x00000008 /* Container for system idle threads. */
810	#define P_PPWAIT 0x00000010 /* Parent is waiting for child to
811	exec/exit. */
812	#define P_PROFIL 0x00000020 /* Has started profiling. */
813	#define P_STOPPROF 0x00000040 /* Has thread requesting to stop
814	profiling. */
815	#define P_HADTHREADS 0x00000080 /* Has had threads (no cleanup
816	shortcuts) */
817	#define P_SUGID 0x00000100 /* Had set id privileges since last
818	exec. */
819	#define P_SYSTEM 0x00000200 /* System proc: no sigs or stats. */
820	#define P_SINGLE_EXIT 0x00000400 /* Threads suspending should exit,
821	not wait. */
822	#define P_TRACED 0x00000800 /* Debugged process being traced. */
823	#define P_WAITED 0x00001000 /* Someone is waiting for us. */
824	#define P_WEXIT 0x00002000 /* Working on exiting. */
825	#define P_EXEC 0x00004000 /* Process called exec. */
826	#define P_WKILLED 0x00008000 /* Killed, go to kernel/user boundary
827	ASAP. */
828	#define P_CONTINUED 0x00010000 /* Proc has continued from a stopped
829	state. */
830	#define P_STOPPED_SIG 0x00020000 /* Stopped due to SIGSTOP/SIGTSTP. */
831	#define P_STOPPED_TRACE 0x00040000 /* Stopped because of tracing. */
832	#define P_STOPPED_SINGLE 0x00080000 /* Only 1 thread can continue (not to
833	user). */
834	#define P_PROTECTED 0x00100000 /* Do not kill on memory overcommit. */
835	#define P_SIGEVENT 0x00200000 /* Process pending signals changed. */
836	#define P_SINGLE_BOUNDARY 0x00400000 /* Threads should suspend at user
837	boundary. */
838	#define P_HWPMC 0x00800000 /* Process is using HWPMCs */
839	#define P_JAILED 0x01000000 /* Process is in jail. */
840	#define P_TOTAL_STOP 0x02000000 /* Stopped in stop_all_proc. */
841	#define P_INEXEC 0x04000000 /* Process is in execve(). */
842	#define P_STATCHILD 0x08000000 /* Child process stopped or exited. */
843	#define P_INMEM 0x10000000 /* Loaded into memory, always set. */
844	#define P_UNUSED1 0x20000000 /* --available-- */
845	#define P_UNUSED2 0x40000000 /* --available-- */
846	#define P_PPTRACE 0x80000000 /* PT_TRACEME by vforked child. */
847
848	#define P_STOPPED (P_STOPPED_SIG|P_STOPPED_SINGLE|P_STOPPED_TRACE)
849	#define P_SHOULDSTOP(p) ((p)->p_flag & P_STOPPED)
850	#define P_KILLED(p) ((p)->p_flag & P_WKILLED)
851
852	/* These flags are kept in p_flag2. */
853	#define P2_INHERIT_PROTECTED 0x00000001 /* New children get
854	P_PROTECTED. */
855	#define P2_NOTRACE 0x00000002 /* No ptrace(2) attach or
856	coredumps. */
857	#define P2_NOTRACE_EXEC 0x00000004 /* Keep P2_NOPTRACE on
858	exec(2). */
859	#define P2_AST_SU 0x00000008 /* Handles SU ast for
860	kthreads. */
861	#define P2_PTRACE_FSTP 0x00000010 /* SIGSTOP from PT_ATTACH not
862	yet handled. */
863	#define P2_TRAPCAP 0x00000020 /* SIGTRAP on ENOTCAPABLE */
864	#define P2_ASLR_ENABLE 0x00000040 /* Force enable ASLR. */
865	#define P2_ASLR_DISABLE 0x00000080 /* Force disable ASLR. */
866	#define P2_ASLR_IGNSTART 0x00000100 /* Enable ASLR to consume sbrk
867	area. */
868	#define P2_PROTMAX_ENABLE 0x00000200 /* Force enable implied
869	PROT_MAX. */
870	#define P2_PROTMAX_DISABLE 0x00000400 /* Force disable implied
871	PROT_MAX. */
872	#define P2_STKGAP_DISABLE 0x00000800 /* Disable stack gap for
873	MAP_STACK */
874	#define P2_STKGAP_DISABLE_EXEC 0x00001000 /* Stack gap disabled
875	after exec */
876	#define P2_ITSTOPPED 0x00002000 /* itimers stopped */
877	#define P2_PTRACEREQ 0x00004000 /* Active ptrace req */
878	#define P2_NO_NEW_PRIVS 0x00008000 /* Ignore setuid */
879	#define P2_WXORX_DISABLE 0x00010000 /* WX mappings enabled */
880	#define P2_WXORX_ENABLE_EXEC 0x00020000 /* WXORX enabled after exec */
881	#define P2_WEXIT 0x00040000 /* exit just started, no
882	external thread_single() is
883	permitted */
884	#define P2_REAPKILLED 0x00080000 /* REAP_KILL pass touched me */
885	#define P2_MEMBAR_PRIVE 0x00100000 /* membar private expedited
886	registered */
887	#define P2_MEMBAR_PRIVE_SYNCORE 0x00200000 /* membar private expedited
888	sync core registered */
889	#define P2_MEMBAR_GLOBE 0x00400000 /* membar global expedited
890	registered */
891
892	#define P2_LOGSIGEXIT_ENABLE 0x00800000 /* Disable logging on sigexit */
893	#define P2_LOGSIGEXIT_CTL 0x01000000 /* Override kern.logsigexit */
894
895	#define P2_HWT 0x02000000 /* Process is using HWT. */
896
897	/* Flags protected by proctree_lock, kept in p_treeflags. */
898	#define P_TREE_ORPHANED 0x00000001 /* Reparented, on orphan list */
899	#define P_TREE_FIRST_ORPHAN 0x00000002 /* First element of orphan
900	list */
901	#define P_TREE_REAPER 0x00000004 /* Reaper of subtree */
902	#define P_TREE_GRPEXITED 0x00000008 /* exit1() done with job ctl */
903
904	/*
905	* These were process status values (p_stat), now they are only used in
906	* legacy conversion code.
907	*/
908	#define SIDL 1 /* Process being created by fork. */
909	#define SRUN 2 /* Currently runnable. */
910	#define SSLEEP 3 /* Sleeping on an address. */
911	#define SSTOP 4 /* Process debugging or suspension. */
912	#define SZOMB 5 /* Awaiting collection by parent. */
913	#define SWAIT 6 /* Waiting for interrupt. */
914	#define SLOCK 7 /* Blocked on a lock. */
915
916	#define P_MAGIC 0xbeefface
917
918	#ifdef _KERNEL
919
920	/* Types and flags for mi_switch(9). */
921	#define SW_TYPE_MASK 0xff /* First 8 bits are switch type */
922	#define SWT_OWEPREEMPT 1 /* Switching due to owepreempt. */
923	#define SWT_TURNSTILE 2 /* Turnstile contention. */
924	#define SWT_SLEEPQ 3 /* Sleepq wait. */
925	#define SWT_RELINQUISH 4 /* yield call. */
926	#define SWT_NEEDRESCHED 5 /* NEEDRESCHED was set. */
927	#define SWT_IDLE 6 /* Switching from the idle thread. */
928	#define SWT_IWAIT 7 /* Waiting for interrupts. */
929	#define SWT_SUSPEND 8 /* Thread suspended. */
930	#define SWT_REMOTEPREEMPT 9 /* Remote processor preempted. */
931	#define SWT_REMOTEWAKEIDLE 10 /* Remote processor preempted idle. */
932	#define SWT_BIND 11 /* Thread bound to a new CPU. */
933	#define SWT_COUNT 12 /* Number of switch types. */
934	/* Flags */
935	#define SW_VOL 0x0100 /* Voluntary switch. */
936	#define SW_INVOL 0x0200 /* Involuntary switch. */
937	#define SW_PREEMPT 0x0400 /* The invol switch is a preemption */
938
939	/* How values for thread_single(). */
940	#define SINGLE_NO_EXIT 0
941	#define SINGLE_EXIT 1
942	#define SINGLE_BOUNDARY 2
943	#define SINGLE_ALLPROC 3
944
945	#define FOREACH_PROC_IN_SYSTEM(p) \
946	LIST_FOREACH((p), &allproc, p_list)
947	#define FOREACH_THREAD_IN_PROC(p, td) \
948	TAILQ_FOREACH((td), &(p)->p_threads, td_plist)
949
950	#define FIRST_THREAD_IN_PROC(p) TAILQ_FIRST(&(p)->p_threads)
951
952	/*
953	* We use process IDs <= pid_max <= PID_MAX; PID_MAX + 1 must also fit
954	* in a pid_t, as it is used to represent "no process group".
955	*/
956	#define PID_MAX 99999
957	#define NO_PID (PID_MAX + 1)
958	#define THREAD0_TID NO_PID
959	extern pid_t pid_max;
960
961	#define SESS_LEADER(p) ((p)->p_session->s_leader == (p))
962
963	/* Lock and unlock a process. */
964	#define PROC_LOCK(p) mtx_lock(&(p)->p_mtx)
965	#define PROC_TRYLOCK(p) mtx_trylock(&(p)->p_mtx)
966	#define PROC_UNLOCK(p) mtx_unlock(&(p)->p_mtx)
967	#define PROC_LOCKED(p) mtx_owned(&(p)->p_mtx)
968	#define PROC_WAIT_UNLOCKED(p) mtx_wait_unlocked(&(p)->p_mtx)
969	#define PROC_LOCK_ASSERT(p, type) mtx_assert(&(p)->p_mtx, (type))
970
971	/* Lock and unlock a process group. */
972	#define PGRP_LOCK(pg) mtx_lock(&(pg)->pg_mtx)
973	#define PGRP_UNLOCK(pg) mtx_unlock(&(pg)->pg_mtx)
974	#define PGRP_LOCKED(pg) mtx_owned(&(pg)->pg_mtx)
975	#define PGRP_LOCK_ASSERT(pg, type) mtx_assert(&(pg)->pg_mtx, (type))
976
977	#define PGRP_LOCK_PGSIGNAL(pg) do { \
978	if ((pg) != NULL) \
979	PGRP_LOCK(pg); \
980	} while (0)
981	#define PGRP_UNLOCK_PGSIGNAL(pg) do { \
982	if ((pg) != NULL) \
983	PGRP_UNLOCK(pg); \
984	} while (0)
985
986	/* Lock and unlock a session. */
987	#define SESS_LOCK(s) mtx_lock(&(s)->s_mtx)
988	#define SESS_UNLOCK(s) mtx_unlock(&(s)->s_mtx)
989	#define SESS_LOCKED(s) mtx_owned(&(s)->s_mtx)
990	#define SESS_LOCK_ASSERT(s, type) mtx_assert(&(s)->s_mtx, (type))
991
992	/*
993	* A non-zero p_lock prevents the process from exiting; it will sleep in exit1()
994	* until the count reaches zero.
995	*
996	* PHOLD() asserts that the process (except the current process) is
997	* not exiting and increments p_lock.
998	* _PHOLD() is same as PHOLD(), it takes the process locked.
999	*/
1000	#define PHOLD(p) do { \
1001	PROC_LOCK(p); \
1002	_PHOLD(p); \
1003	PROC_UNLOCK(p); \
1004	} while (0)
1005	#define _PHOLD(p) do { \
1006	PROC_LOCK_ASSERT((p), MA_OWNED); \
1007	KASSERT(!((p)->p_flag & P_WEXIT) || (p) == curproc, \
1008	("PHOLD of exiting process %p", p)); \
1009	(p)->p_lock++; \
1010	} while (0)
1011	#define PROC_ASSERT_HELD(p) do { \
1012	KASSERT((p)->p_lock > 0, ("process %p not held", p)); \
1013	} while (0)
1014
1015	#define PRELE(p) do { \
1016	PROC_LOCK((p)); \
1017	_PRELE((p)); \
1018	PROC_UNLOCK((p)); \
1019	} while (0)
1020	#define _PRELE(p) do { \
1021	PROC_LOCK_ASSERT((p), MA_OWNED); \
1022	PROC_ASSERT_HELD(p); \
1023	(--(p)->p_lock); \
1024	if (((p)->p_flag & P_WEXIT) && (p)->p_lock == 0) \
1025	wakeup(&(p)->p_lock); \
1026	} while (0)
1027	#define PROC_ASSERT_NOT_HELD(p) do { \
1028	KASSERT((p)->p_lock == 0, ("process %p held", p)); \
1029	} while (0)
1030
1031	#define PROC_UPDATE_COW(p) do { \
1032	struct proc *_p = (p); \
1033	PROC_LOCK_ASSERT((_p), MA_OWNED); \
1034	atomic_store_int(&_p->p_cowgen, _p->p_cowgen + 1); \
1035	} while (0)
1036
1037	#define PROC_COW_CHANGECOUNT(td, p) ({ \
1038	struct thread *_td = (td); \
1039	struct proc *_p = (p); \
1040	MPASS(_td == curthread); \
1041	PROC_LOCK_ASSERT(_p, MA_OWNED); \
1042	_p->p_cowgen - _td->td_cowgen; \
1043	})
1044
1045	/* Control whether or not it is safe for curthread to sleep. */
1046	#define THREAD_NO_SLEEPING() do { \
1047	curthread->td_no_sleeping++; \
1048	MPASS(curthread->td_no_sleeping > 0); \
1049	} while (0)
1050
1051	#define THREAD_SLEEPING_OK() do { \
1052	MPASS(curthread->td_no_sleeping > 0); \
1053	curthread->td_no_sleeping--; \
1054	} while (0)
1055
1056	#define THREAD_CAN_SLEEP() ((curthread)->td_no_sleeping == 0)
1057
1058	#define THREAD_CONTENDS_ON_LOCK(lo) do { \
1059	MPASS(curthread->td_wantedlock == NULL); \
1060	curthread->td_wantedlock = lo; \
1061	} while (0)
1062
1063	#define THREAD_CONTENTION_DONE(lo) do { \
1064	MPASS(curthread->td_wantedlock == lo); \
1065	curthread->td_wantedlock = NULL; \
1066	} while (0)
1067
1068	#define PIDHASH(pid) (&pidhashtbl[(pid) & pidhash])
1069	#define PIDHASHLOCK(pid) (&pidhashtbl_lock[((pid) & pidhashlock)])
1070	extern LIST_HEAD(pidhashhead, proc) *pidhashtbl;
1071	extern struct sx *pidhashtbl_lock;
1072	extern u_long pidhash;
1073	extern u_long pidhashlock;
1074
1075	#define PGRPHASH(pgid) (&pgrphashtbl[(pgid) & pgrphash])
1076	extern LIST_HEAD(pgrphashhead, pgrp) *pgrphashtbl;
1077	extern u_long pgrphash;
1078
1079	extern struct sx allproc_lock;
1080	extern int allproc_gen;
1081	extern struct sx proctree_lock;
1082	extern struct mtx ppeers_lock;
1083	extern struct mtx procid_lock;
1084	extern struct proc proc0; /* Process slot for swapper. */
1085	extern struct thread0_storage thread0_st; /* Primary thread in proc0. */
1086	#define thread0 (thread0_st.t0st_thread)
1087	extern struct vmspace vmspace0; /* VM space for proc0. */
1088	extern int hogticks; /* Limit on kernel cpu hogs. */
1089	extern int lastpid;
1090	extern int nprocs, maxproc; /* Current and max number of procs. */
1091	extern int maxprocperuid; /* Max procs per uid. */
1092	extern u_long ps_arg_cache_limit;
1093
1094	LIST_HEAD(proclist, proc);
1095	TAILQ_HEAD(procqueue, proc);
1096	TAILQ_HEAD(threadqueue, thread);
1097	extern struct proclist allproc; /* List of all processes. */
1098	extern struct proc *initproc, *pageproc; /* Process slots for init, pager. */
1099
1100	extern struct uma_zone *proc_zone;
1101	extern struct uma_zone *pgrp_zone;
1102
1103	struct proc *pfind(pid_t); /* Find process by id. */
1104	struct proc *pfind_any(pid_t); /* Find (zombie) process by id. */
1105	struct proc *pfind_any_locked(pid_t pid); /* Find process by id, locked. */
1106	struct pgrp *pgfind(pid_t); /* Find process group by id. */
1107	void pidhash_slockall(void); /* Shared lock all pid hash lists. */
1108	void pidhash_sunlockall(void); /* Shared unlock all pid hash lists. */
1109
1110	struct fork_req {
1111	int fr_flags;
1112	int fr_pages;
1113	int *fr_pidp;
1114	struct proc **fr_procp;
1115	int *fr_pd_fd;
1116	int fr_pd_flags;
1117	struct filecaps *fr_pd_fcaps;
1118	int fr_flags2;
1119	#define FR2_DROPSIG_CAUGHT 0x00000001 /* Drop caught non-DFL signals */
1120	#define FR2_SHARE_PATHS 0x00000002 /* Invert sense of RFFDG for paths */
1121	#define FR2_KPROC 0x00000004 /* Create a kernel process */
1122	};
1123
1124	/*
1125	* pget() flags.
1126	*/
1127	#define PGET_HOLD 0x00001 /* Hold the process. */
1128	#define PGET_CANSEE 0x00002 /* Check against p_cansee(). */
1129	#define PGET_CANDEBUG 0x00004 /* Check against p_candebug(). */
1130	#define PGET_ISCURRENT 0x00008 /* Check that the found process is current. */
1131	#define PGET_NOTWEXIT 0x00010 /* Check that the process is not in P_WEXIT. */
1132	#define PGET_NOTINEXEC 0x00020 /* Check that the process is not in P_INEXEC. */
1133	#define PGET_NOTID 0x00040 /* Do not assume tid if pid > PID_MAX. */
1134
1135	#define PGET_WANTREAD (PGET_HOLD | PGET_CANDEBUG | PGET_NOTWEXIT)
1136
1137	int pget(pid_t pid, int flags, struct proc **pp);
1138
1139	/* ast_register() flags */
1140	#define ASTR_ASTF_REQUIRED 0x0001 /* td_ast TDAI(TDA_X) flag set is
1141	required for call */
1142	#define ASTR_TDP 0x0002 /* td_pflags flag set is required */
1143	#define ASTR_KCLEAR 0x0004 /* call me on ast_kclear() */
1144	#define ASTR_UNCOND 0x0008 /* call me always */
1145
1146	void ast(struct trapframe *framep);
1147	void ast_kclear(struct thread *td);
1148	void ast_register(int ast, int ast_flags, int tdp,
1149	void (*f)(struct thread *td, int asts));
1150	void ast_deregister(int tda);
1151	void ast_sched_locked(struct thread *td, int tda);
1152	void ast_sched_mask(struct thread *td, int ast);
1153	void ast_sched(struct thread *td, int tda);
1154	void ast_unsched_locked(struct thread *td, int tda);
1155
1156	struct thread *choosethread(void);
1157	int cr_bsd_visible(struct ucred *u1, struct ucred *u2);
1158	int cr_cansee(struct ucred *u1, struct ucred *u2);
1159	int cr_canseesocket(struct ucred *cred, struct socket *so);
1160	int cr_cansignal(struct ucred *cred, struct proc *proc, int signum);
1161	int enterpgrp(struct proc *p, pid_t pgid, struct pgrp *pgrp,
1162	struct session *sess);
1163	int enterthispgrp(struct proc *p, struct pgrp *pgrp);
1164	int fork1(struct thread *, struct fork_req *);
1165	void fork_exit(void (*)(void *, struct trapframe *), void *,
1166	struct trapframe *);
1167	void fork_return(struct thread *, struct trapframe *);
1168	int inferior(struct proc *p);
1169	void itimer_proc_continue(struct proc *p);
1170	void kqtimer_proc_continue(struct proc *p);
1171	void kern_proc_vmmap_resident(struct vm_map *map, struct vm_map_entry *entry,
1172	int *resident_count, bool *super);
1173	void kern_yield(int);
1174	void killjobc(void);
1175	int leavepgrp(struct proc *p);
1176	int maybe_preempt(struct thread *td);
1177	void maybe_yield(void);
1178	void mi_switch(int flags);
1179	int p_candebug(struct thread *td, struct proc *p);
1180	int p_cansee(struct thread *td, struct proc *p);
1181	int p_cansched(struct thread *td, struct proc *p);
1182	int p_cansignal(struct thread *td, struct proc *p, int signum);
1183	int p_canwait(struct thread *td, struct proc *p);
1184	struct pargs *pargs_alloc(int len);
1185	void pargs_drop(struct pargs *pa);
1186	void pargs_hold(struct pargs *pa);
1187	int pgrp_calc_jobc(struct pgrp *pgrp);
1188	void proc_add_orphan(struct proc *child, struct proc *parent);
1189	int proc_get_binpath(struct proc *p, char *binname, char **fullpath,
1190	char **freepath);
1191	int proc_getargv(struct thread *td, struct proc *p, struct sbuf *sb);
1192	int proc_getauxv(struct thread *td, struct proc *p, struct sbuf *sb);
1193	int proc_getenvv(struct thread *td, struct proc *p, struct sbuf *sb);
1194	void procinit(void);
1195	int proc_iterate(int (*cb)(struct proc *, void *), void *cbarg);
1196	void proc_linkup0(struct proc *p, struct thread *td);
1197	void proc_linkup(struct proc *p, struct thread *td);
1198	struct proc *proc_realparent(struct proc *child);
1199	void proc_reap(struct thread *td, struct proc *p, int *status, int options);
1200	void proc_reparent(struct proc *child, struct proc *newparent, bool set_oppid);
1201	void proc_set_p2_wexit(struct proc *p);
1202	void proc_set_traced(struct proc *p, bool stop);
1203	void proc_wkilled(struct proc *p);
1204	struct pstats *pstats_alloc(void);
1205	void pstats_fork(struct pstats *src, struct pstats *dst);
1206	void pstats_free(struct pstats *ps);
1207	void proc_clear_orphan(struct proc *p);
1208	void reaper_abandon_children(struct proc *p, bool exiting);
1209	int securelevel_ge(struct ucred *cr, int level);
1210	int securelevel_gt(struct ucred *cr, int level);
1211	void sess_hold(struct session *);
1212	void sess_release(struct session *);
1213	void setrunnable(struct thread *, int);
1214	void setsugid(struct proc *p);
1215	bool should_yield(void);
1216	int sigonstack(size_t sp);
1217	void stopevent(struct proc *, u_int, u_int);
1218	struct thread *tdfind(lwpid_t, pid_t);
1219	void threadinit(void);
1220	void tidhash_add(struct thread *);
1221	void tidhash_remove(struct thread *);
1222	void cpu_idle(int);
1223	int cpu_idle_wakeup(int);
1224	extern void (*cpu_idle_hook)(sbintime_t); /* Hook to machdep CPU idler. */
1225	void cpu_switch(struct thread *, struct thread *, struct mtx *);
1226	void cpu_sync_core(void);
1227	void cpu_throw(struct thread *, struct thread *) __dead2;
1228	void cpu_update_pcb(struct thread *);
1229	bool curproc_sigkilled(void);
1230	void userret(struct thread *, struct trapframe *);
1231
1232	void cpu_exit(struct thread *);
1233	void exit1(struct thread *, int, int) __dead2;
1234	void cpu_copy_thread(struct thread *td, struct thread *td0);
1235	bool cpu_exec_vmspace_reuse(struct proc *p, struct vm_map *map);
1236	int cpu_fetch_syscall_args(struct thread *td);
1237	void cpu_fork(struct thread *, struct proc *, struct thread *, int);
1238	void cpu_fork_kthread_handler(struct thread *, void (*)(void *), void *);
1239	int cpu_procctl(struct thread *td, int idtype, id_t id, int com,
1240	void *data);
1241	void cpu_set_syscall_retval(struct thread *, int);
1242	int cpu_set_upcall(struct thread *, void (*)(void *), void *,
1243	stack_t *);
1244	int cpu_set_user_tls(struct thread *, void *tls_base, int flags);
1245	void cpu_thread_alloc(struct thread *);
1246	void cpu_thread_clean(struct thread *);
1247	void cpu_thread_exit(struct thread *);
1248	void cpu_thread_free(struct thread *);
1249	struct thread *thread_alloc(int pages);
1250	int thread_check_susp(struct thread *td, bool sleep);
1251	void thread_cow_get_proc(struct thread *newtd, struct proc *p);
1252	void thread_cow_get(struct thread *newtd, struct thread *td);
1253	void thread_cow_free(struct thread *td);
1254	void thread_cow_update(struct thread *td);
1255	void thread_cow_synced(struct thread *td);
1256	int thread_create(struct thread *td, struct rtprio *rtp,
1257	int (*initialize_thread)(struct thread *, void *), void *thunk);
1258	void thread_exit(void) __dead2;
1259	void thread_free(struct thread *td);
1260	void thread_link(struct thread *td, struct proc *p);
1261	void thread_reap_barrier(void);
1262	int thread_recycle(struct thread *, int pages);
1263	int thread_single(struct proc *p, int how);
1264	void thread_single_end(struct proc *p, int how);
1265	void thread_stash(struct thread *td);
1266	void thread_stopped(struct proc *p);
1267	void childproc_stopped(struct proc *child, int reason);
1268	void childproc_continued(struct proc *child);
1269	void childproc_exited(struct proc *child);
1270	void thread_run_flash(struct thread *td);
1271	int thread_suspend_check(int how);
1272	bool thread_suspend_check_needed(void);
1273	void thread_suspend_switch(struct thread *, struct proc *p);
1274	void thread_suspend_one(struct thread *td);
1275	void thread_unlink(struct thread *td);
1276	void thread_unsuspend(struct proc *p);
1277	void thread_wait(struct proc *p);
1278
1279	bool stop_all_proc_block(void);
1280	void stop_all_proc_unblock(void);
1281	void stop_all_proc(void);
1282	void resume_all_proc(void);
1283
1284	static __inline int
1285	curthread_pflags_set(int flags)
1286	{
1287	struct thread *td;
1288	int save;
1289
1290	td = curthread;
1291	save = ~flags | (td->td_pflags & flags);
1292	td->td_pflags |= flags;
1293	return (save);
1294	}
1295
1296	static __inline void
1297	curthread_pflags_restore(int save)
1298	{
1299
1300	curthread->td_pflags &= save;
1301	}
1302
1303	static __inline int
1304	curthread_pflags2_set(int flags)
1305	{
1306	struct thread *td;
1307	int save;
1308
1309	td = curthread;
1310	save = ~flags | (td->td_pflags2 & flags);
1311	td->td_pflags2 |= flags;
1312	return (save);
1313	}
1314
1315	static __inline void
1316	curthread_pflags2_restore(int save)
1317	{
1318
1319	curthread->td_pflags2 &= save;
1320	}
1321
1322	static __inline __pure2 struct td_sched *
1323	td_get_sched(struct thread *td)
1324	{
1325
1326	return ((struct td_sched *)&td[1]);
1327	}
1328
1329	#define PROC_ID_PID 0
1330	#define PROC_ID_GROUP 1
1331	#define PROC_ID_SESSION 2
1332	#define PROC_ID_REAP 3
1333
1334	void proc_id_set(int type, pid_t id);
1335	void proc_id_set_cond(int type, pid_t id);
1336	void proc_id_clear(int type, pid_t id);
1337
1338	EVENTHANDLER_LIST_DECLARE(process_ctor);
1339	EVENTHANDLER_LIST_DECLARE(process_dtor);
1340	EVENTHANDLER_LIST_DECLARE(process_init);
1341	EVENTHANDLER_LIST_DECLARE(process_fini);
1342	EVENTHANDLER_LIST_DECLARE(process_exit);
1343	EVENTHANDLER_LIST_DECLARE(process_fork);
1344	EVENTHANDLER_LIST_DECLARE(process_exec);
1345
1346	EVENTHANDLER_LIST_DECLARE(thread_ctor);
1347	EVENTHANDLER_LIST_DECLARE(thread_dtor);
1348	EVENTHANDLER_LIST_DECLARE(thread_init);
1349
1350	#endif /* _KERNEL */
1351
1352	#endif /* !_SYS_PROC_H_ */