| 1 | /*	$NetBSD: socketvar.h,v 1.171 2025/04/07 21:02:19 andvar Exp $	*/ |
| 2 | |
| 3 | /*- |
| 4 | * Copyright (c) 2008, 2009 The NetBSD Foundation, Inc. |
| 5 | * All rights reserved. |
| 6 | * |
| 7 | * This code is derived from software contributed to The NetBSD Foundation |
| 8 | * by Andrew Doran. |
| 9 | * |
| 10 | * Redistribution and use in source and binary forms, with or without |
| 11 | * modification, are permitted provided that the following conditions |
| 12 | * are met: |
| 13 | * 1. Redistributions of source code must retain the above copyright |
| 14 | * notice, this list of conditions and the following disclaimer. |
| 15 | * 2. Redistributions in binary form must reproduce the above copyright |
| 16 | * notice, this list of conditions and the following disclaimer in the |
| 17 | * documentation and/or other materials provided with the distribution. |
| 18 | * |
| 19 | * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS |
| 20 | * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED |
| 21 | * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR |
| 22 | * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS |
| 23 | * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR |
| 24 | * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF |
| 25 | * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS |
| 26 | * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN |
| 27 | * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) |
| 28 | * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE |
| 29 | * POSSIBILITY OF SUCH DAMAGE. |
| 30 | */ |
| 31 | |
| 32 | /*- |
| 33 | * Copyright (c) 1982, 1986, 1990, 1993 |
| 34 | *	The Regents of the University of California. All rights reserved. |
| 35 | * |
| 36 | * Redistribution and use in source and binary forms, with or without |
| 37 | * modification, are permitted provided that the following conditions |
| 38 | * are met: |
| 39 | * 1. Redistributions of source code must retain the above copyright |
| 40 | * notice, this list of conditions and the following disclaimer. |
| 41 | * 2. Redistributions in binary form must reproduce the above copyright |
| 42 | * notice, this list of conditions and the following disclaimer in the |
| 43 | * documentation and/or other materials provided with the distribution. |
| 44 | * 3. Neither the name of the University nor the names of its contributors |
| 45 | * may be used to endorse or promote products derived from this software |
| 46 | * without specific prior written permission. |
| 47 | * |
| 48 | * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND |
| 49 | * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
| 50 | * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE |
| 51 | * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE |
| 52 | * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL |
| 53 | * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS |
| 54 | * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) |
| 55 | * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT |
| 56 | * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY |
| 57 | * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF |
| 58 | * SUCH DAMAGE. |
| 59 | * |
| 60 | *	@(#)socketvar.h	8.3 (Berkeley) 2/19/95 |
| 61 | */ |
| 62 | |
| 63 | #ifndef _SYS_SOCKETVAR_H_ |
| 64 | #define	_SYS_SOCKETVAR_H_ |
| 65 | |
| 66 | #include <sys/select.h> |
| 67 | #include <sys/selinfo.h>		/* for struct selinfo */ |
| 68 | #include <sys/queue.h> |
| 69 | #include <sys/mutex.h> |
| 70 | #include <sys/condvar.h> |
| 71 | |
| 72 | #if !defined(_KERNEL) |
| 73 | struct uio; |
| 74 | struct lwp; |
| 75 | struct uidinfo; |
| 76 | #else |
| 77 | #include <sys/atomic.h> |
| 78 | #include <sys/uidinfo.h> |
| 79 | #endif |
| 80 | |
| 81 | TAILQ_HEAD(soqhead, socket); |
| 82 | |
| 83 | /* |
| 84 | * Variables for socket buffering. |
| 85 | */ |
| 86 | struct sockbuf { |
| 87 | 	struct selinfo sb_sel;		/* process selecting read/write */ |
| 88 | 	struct mowner *sb_mowner;	/* who owns data for this sockbuf */ |
| 89 | 	struct socket *sb_so;		/* back pointer to socket */ |
| 90 | 	kcondvar_t sb_cv;		/* notifier */ |
| 91 | 	/* When re-zeroing this struct, we zero from sb_startzero to the end */ |
| 92 | #define	sb_startzero	sb_cc |
| 93 | 	u_long	sb_cc;			/* actual chars in buffer */ |
| 94 | 	u_long	sb_hiwat;		/* max actual char count */ |
| 95 | 	u_long	sb_mbcnt;		/* chars of mbufs used */ |
| 96 | 	u_long	sb_mbmax;		/* max chars of mbufs to use */ |
| 97 | 	u_long	sb_lowat;		/* low water mark */ |
| 98 | 	struct mbuf *sb_mb;		/* the mbuf chain */ |
| 99 | 	struct mbuf *sb_mbtail;		/* the last mbuf in the chain */ |
| 100 | 	struct mbuf *sb_lastrecord;	/* first mbuf of last record in |
| 101 | 					 socket buffer */ |
| 102 | 	int	sb_flags;		/* flags, see below */ |
| 103 | 	int	sb_timeo;		/* timeout for read/write */ |
| 104 | 	u_long	sb_overflowed;		/* # of drops due to full buffer */ |
| 105 | }; |
| 106 | |
| 107 | #ifndef SB_MAX |
| 108 | #define	SB_MAX		(256*1024)	/* default for max chars in sockbuf */ |
| 109 | #endif |
| 110 | |
| 111 | #define	SB_LOCK		0x01		/* lock on data queue */ |
| 112 | #define	SB_NOTIFY	0x04		/* someone is waiting for data/space */ |
| 113 | #define	SB_ASYNC	0x10		/* ASYNC I/O, need signals */ |
| 114 | #define	SB_UPCALL	0x20		/* someone wants an upcall */ |
| 115 | #define	SB_NOINTR	0x40		/* operations not interruptible */ |
| 116 | #define	SB_KNOTE	0x100		/* kernel note attached */ |
| 117 | #define	SB_AUTOSIZE	0x800		/* automatically size socket buffer */ |
| 118 | |
| 119 | /* |
| 120 | * Kernel structure per socket. |
| 121 | * Contains send and receive buffer queues, |
| 122 | * handle on protocol and pointer to protocol |
| 123 | * private data and error information. |
| 124 | */ |
| 125 | struct so_accf { |
| 126 | 	struct accept_filter	*so_accept_filter; |
| 127 | 	void	*so_accept_filter_arg;	/* saved filter args */ |
| 128 | 	char	*so_accept_filter_str;	/* saved user args */ |
| 129 | }; |
| 130 | |
| 131 | struct sockaddr; |
| 132 | |
| 133 | struct socket { |
| 134 | 	kmutex_t * volatile so_lock;	/* pointer to lock on structure */ |
| 135 | 	kcondvar_t	so_cv;		/* notifier */ |
| 136 | 	short		so_type;	/* generic type, see socket.h */ |
| 137 | 	short		so_options;	/* from socket call, see socket.h */ |
| 138 | 	u_short		so_linger;	/* time to linger while closing */ |
| 139 | 	short		so_state;	/* internal state flags SS_*, below */ |
| 140 | 	int		so_unused;	/* used to be so_nbio */ |
| 141 | 	void		*so_pcb;	/* protocol control block */ |
| 142 | 	const struct protosw *so_proto;	/* protocol handle */ |
| 143 | /* |
| 144 | * Variables for connection queueing. |
| 145 | * Socket where accepts occur is so_head in all subsidiary sockets. |
| 146 | * If so_head is 0, socket is not related to an accept. |
| 147 | * For head socket so_q0 queues partially completed connections, |
| 148 | * while so_q is a queue of connections ready to be accepted. |
| 149 | * If a connection is aborted and it has so_head set, then |
| 150 | * it has to be pulled out of either so_q0 or so_q. |
| 151 | * We allow connections to queue up based on current queue lengths |
| 152 | * and limit on number of queued connections for this socket. |
| 153 | */ |
| 154 | 	struct socket	*so_head;	/* back pointer to accept socket */ |
| 155 | 	struct soqhead	*so_onq;	/* queue (q or q0) that we're on */ |
| 156 | 	struct soqhead	so_q0;		/* queue of partial connections */ |
| 157 | 	struct soqhead	so_q;		/* queue of incoming connections */ |
| 158 | 	TAILQ_ENTRY(socket) so_qe;	/* our queue entry (q or q0) */ |
| 159 | 	short		so_q0len;	/* partials on so_q0 */ |
| 160 | 	short		so_qlen;	/* number of connections on so_q */ |
| 161 | 	short		so_qlimit;	/* max number queued connections */ |
| 162 | 	short		so_timeo;	/* connection timeout */ |
| 163 | 	u_short		so_error;	/* error affecting connection */ |
| 164 | 	u_short		so_rerror;	/* error affecting receiving */ |
| 165 | 	u_short		so_aborting;	/* references from soabort() */ |
| 166 | 	pid_t		so_pgid;	/* pgid for signals */ |
| 167 | 	u_long		so_oobmark;	/* chars to oob mark */ |
| 168 | 	struct sockbuf	so_snd;		/* send buffer */ |
| 169 | 	struct sockbuf	so_rcv;		/* receive buffer */ |
| 170 | |
| 171 | 	void		*so_internal;	/* Space for svr4 stream data */ |
| 172 | 	void		(*so_upcall) (struct socket *, void *, int, int); |
| 173 | 	void *		so_upcallarg;	/* Arg for above */ |
| 174 | 	int		(*so_send) (struct socket *, struct sockaddr *, |
| 175 | 					struct uio *, struct mbuf *, |
| 176 | 					struct mbuf *, int, struct lwp *); |
| 177 | 	int		(*so_receive) (struct socket *, |
| 178 | 					struct mbuf **, |
| 179 | 					struct uio *, struct mbuf **, |
| 180 | 					struct mbuf **, int *); |
| 181 | 	struct mowner	*so_mowner;	/* who owns mbufs for this socket */ |
| 182 | 	struct uidinfo	*so_uidinfo;	/* who opened the socket */ |
| 183 | 	gid_t		so_egid;	/* creator effective gid */ |
| 184 | 	pid_t		so_cpid;	/* creator pid */ |
| 185 | 	struct so_accf	*so_accf; |
| 186 | 	kauth_cred_t	so_cred;	/* socket credentials */ |
| 187 | }; |
| 188 | |
| 189 | /* |
| 190 | * Socket state bits. |
| 191 | */ |
| 192 | #define	SS_NOFDREF		0x001	/* no file table ref any more */ |
| 193 | #define	SS_ISCONNECTED		0x002	/* socket connected to a peer */ |
| 194 | #define	SS_ISCONNECTING		0x004	/* in process of connecting to peer */ |
| 195 | #define	SS_ISDISCONNECTING	0x008	/* in process of disconnecting */ |
| 196 | #define	SS_CANTSENDMORE		0x010	/* can't send more data to peer */ |
| 197 | #define	SS_CANTRCVMORE		0x020	/* can't receive more data from peer */ |
| 198 | #define	SS_RCVATMARK		0x040	/* at mark on input */ |
| 199 | #define	SS_ISABORTING		0x080	/* aborting fd references - close() */ |
| 200 | #define	SS_RESTARTSYS		0x100	/* restart blocked system calls */ |
| 201 | #define	SS_POLLRDBAND		0x200	/* poll should return POLLRDBAND */ |
| 202 | #define	SS_MORETOCOME		0x400	/* |
| 203 | 					 * hint from sosend to lower layer; |
| 204 | 					 * more data coming |
| 205 | 					 */ |
| 206 | #define	SS_ISDISCONNECTED	0x800	/* socket disconnected from peer */ |
| 207 | #define	SS_ISAPIPE 		0x1000	/* socket is implementing a pipe */ |
| 208 | #define	SS_NBIO			0x2000	/* socket is in non blocking I/O */ |
| 209 | |
| 210 | #ifdef _KERNEL |
| 211 | |
| 212 | struct accept_filter { |
| 213 | 	char	accf_name[16]; |
| 214 | 	void	(*accf_callback) |
| 215 | 		(struct socket *, void *, int, int); |
| 216 | 	void *	(*accf_create) |
| 217 | 		(struct socket *, char *); |
| 218 | 	void	(*accf_destroy) |
| 219 | 		(struct socket *); |
| 220 | 	LIST_ENTRY(accept_filter) accf_next; |
| 221 | 	u_int	accf_refcnt; |
| 222 | }; |
| 223 | |
| 224 | struct sockopt { |
| 225 | 	int		sopt_level;		/* option level */ |
| 226 | 	int		sopt_name;		/* option name */ |
| 227 | 	size_t		sopt_size;		/* data length */ |
| 228 | 	size_t		sopt_retsize;		/* returned data length */ |
| 229 | 	void *		sopt_data;		/* data pointer */ |
| 230 | 	uint8_t		sopt_buf[sizeof(int)];	/* internal storage */ |
| 231 | }; |
| 232 | |
| 233 | #define	SB_EMPTY_FIXUP(sb)						\ |
| 234 | do {									\ |
| 235 | 	KASSERT(solocked((sb)->sb_so));					\ |
| 236 | 	if ((sb)->sb_mb == NULL) {					\ |
| 237 | 		(sb)->sb_mbtail = NULL;					\ |
| 238 | 		(sb)->sb_lastrecord = NULL;				\ |
| 239 | 	}								\ |
| 240 | } while (0) |
| 241 | |
| 242 | extern u_long		sb_max; |
| 243 | extern int		somaxkva; |
| 244 | extern int		sock_loan_thresh; |
| 245 | extern kmutex_t		*softnet_lock; |
| 246 | |
| 247 | struct mbuf; |
| 248 | struct lwp; |
| 249 | struct msghdr; |
| 250 | struct stat; |
| 251 | struct knote; |
| 252 | struct sockaddr_big; |
| 253 | enum uio_seg; |
| 254 | |
| 255 | /* 0x400 is SO_OTIMESTAMP */ |
| 256 | #define SOOPT_TIMESTAMP(o) ((o) & (SO_TIMESTAMP | 0x400)) |
| 257 | |
| 258 | /* |
| 259 | * File operations on sockets. |
| 260 | */ |
| 261 | int	soo_read(file_t *, off_t *, struct uio *, kauth_cred_t, int); |
| 262 | int	soo_write(file_t *, off_t *, struct uio *, kauth_cred_t, int); |
| 263 | int	soo_fcntl(file_t *, u_int cmd, void *); |
| 264 | int	soo_ioctl(file_t *, u_long cmd, void *); |
| 265 | int	soo_poll(file_t *, int); |
| 266 | int	soo_kqfilter(file_t *, struct knote *); |
| 267 | int 	soo_close(file_t *); |
| 268 | int	soo_stat(file_t *, struct stat *); |
| 269 | void	soo_restart(file_t *); |
| 270 | void	sbappend(struct sockbuf *, struct mbuf *); |
| 271 | void	sbappendstream(struct sockbuf *, struct mbuf *); |
| 272 | int	sbappendaddr(struct sockbuf *, const struct sockaddr *, struct mbuf *, |
| 273 | 	 struct mbuf *); |
| 274 | int	sbappendaddrchain(struct sockbuf *, const struct sockaddr *, |
| 275 | 	 struct mbuf *, int); |
| 276 | int	sbappendcontrol(struct sockbuf *, struct mbuf *, struct mbuf *); |
| 277 | void	sbappendrecord(struct sockbuf *, struct mbuf *); |
| 278 | void	sbcheck(struct sockbuf *); |
| 279 | void	sbcompress(struct sockbuf *, struct mbuf *, struct mbuf *); |
| 280 | struct mbuf * |
| 281 | 	sbcreatecontrol(void *, int, int, int); |
| 282 | struct mbuf * |
| 283 | 	sbcreatecontrol1(void **, int, int, int, int); |
| 284 | struct mbuf ** |
| 285 | 	sbsavetimestamp(int, struct mbuf **); |
| 286 | void	sbdrop(struct sockbuf *, int); |
| 287 | void	sbdroprecord(struct sockbuf *); |
| 288 | void	sbflush(struct sockbuf *); |
| 289 | void	sbinsertoob(struct sockbuf *, struct mbuf *); |
| 290 | void	sbrelease(struct sockbuf *, struct socket *); |
| 291 | int	sbreserve(struct sockbuf *, u_long, struct socket *); |
| 292 | int	sbwait(struct sockbuf *); |
| 293 | int	sb_max_set(u_long); |
| 294 | void	soinit(void); |
| 295 | void	soinit1(void); |
| 296 | void	soinit2(void); |
| 297 | int	soabort(struct socket *); |
| 298 | int	soaccept(struct socket *, struct sockaddr *); |
| 299 | int	sofamily(const struct socket *); |
| 300 | int	sobind(struct socket *, struct sockaddr *, struct lwp *); |
| 301 | void	socantrcvmore(struct socket *); |
| 302 | void	socantsendmore(struct socket *); |
| 303 | void	soroverflow(struct socket *); |
| 304 | int	soclose(struct socket *); |
| 305 | int	soconnect(struct socket *, struct sockaddr *, struct lwp *); |
| 306 | int	soconnect2(struct socket *, struct socket *); |
| 307 | int	socreate(int, struct socket **, int, int, struct lwp *, |
| 308 | 		 struct socket *); |
| 309 | int	fsocreate(int, struct socket **, int, int, int *, file_t **, |
| 310 | 		struct socket *); |
| 311 | int	sodisconnect(struct socket *); |
| 312 | void	sofree(struct socket *); |
| 313 | int	sogetopt(struct socket *, struct sockopt *); |
| 314 | void	sohasoutofband(struct socket *); |
| 315 | void	soisconnected(struct socket *); |
| 316 | void	soisconnecting(struct socket *); |
| 317 | void	soisdisconnected(struct socket *); |
| 318 | void	soisdisconnecting(struct socket *); |
| 319 | int	solisten(struct socket *, int, struct lwp *); |
| 320 | struct socket * |
| 321 | 	sonewconn(struct socket *, bool); |
| 322 | void	soqinsque(struct socket *, struct socket *, int); |
| 323 | bool	soqremque(struct socket *, int); |
| 324 | int	soreceive(struct socket *, struct mbuf **, struct uio *, |
| 325 | 	 struct mbuf **, struct mbuf **, int *); |
| 326 | int	soreserve(struct socket *, u_long, u_long); |
| 327 | void	sorflush(struct socket *); |
| 328 | int	sosend(struct socket *, struct sockaddr *, struct uio *, |
| 329 | 	 struct mbuf *, struct mbuf *, int, struct lwp *); |
| 330 | int	sosetopt(struct socket *, struct sockopt *); |
| 331 | int	so_setsockopt(struct lwp *, struct socket *, int, int, const void *, size_t); |
| 332 | int	soshutdown(struct socket *, int); |
| 333 | void	sorestart(struct socket *); |
| 334 | void	sowakeup(struct socket *, struct sockbuf *, int); |
| 335 | int	sockargs(struct mbuf **, const void *, size_t, enum uio_seg, int); |
| 336 | int	sopoll(struct socket *, int); |
| 337 | struct	socket *soget(bool); |
| 338 | void	soput(struct socket *); |
| 339 | bool	solocked(const struct socket *); |
| 340 | bool	solocked2(const struct socket *, const struct socket *); |
| 341 | int	sblock(struct sockbuf *, int); |
| 342 | void	sbunlock(struct sockbuf *); |
| 343 | int	sowait(struct socket *, bool, int); |
| 344 | void	solockretry(struct socket *, kmutex_t *); |
| 345 | void	sosetlock(struct socket *); |
| 346 | void	solockreset(struct socket *, kmutex_t *); |
| 347 | |
| 348 | void	sockopt_init(struct sockopt *, int, int, size_t); |
| 349 | void	sockopt_destroy(struct sockopt *); |
| 350 | int	sockopt_set(struct sockopt *, const void *, size_t); |
| 351 | int	sockopt_setint(struct sockopt *, int); |
| 352 | int	sockopt_get(const struct sockopt *, void *, size_t); |
| 353 | int	sockopt_getint(const struct sockopt *, int *); |
| 354 | int	sockopt_setmbuf(struct sockopt *, struct mbuf *); |
| 355 | struct mbuf *sockopt_getmbuf(const struct sockopt *); |
| 356 | |
| 357 | int	copyout_sockname(struct sockaddr *, unsigned int *, int, struct mbuf *); |
| 358 | int	copyout_sockname_sb(struct sockaddr *, unsigned int *, |
| 359 | int , struct sockaddr_big *); |
| 360 | int	copyout_msg_control(struct lwp *, struct msghdr *, struct mbuf *); |
| 361 | void	free_control_mbuf(struct lwp *, struct mbuf *, struct mbuf *); |
| 362 | |
| 363 | int	do_sys_getpeername(int, struct sockaddr *); |
| 364 | int	do_sys_getsockname(int, struct sockaddr *); |
| 365 | |
| 366 | int	do_sys_sendmsg(struct lwp *, int, struct msghdr *, int, register_t *); |
| 367 | int	do_sys_sendmsg_so(struct lwp *, int, struct socket *, file_t *, |
| 368 | 	 struct msghdr *, int, register_t *); |
| 369 | |
| 370 | int	do_sys_recvmsg(struct lwp *, int, struct msghdr *, |
| 371 | 	 struct mbuf **, struct mbuf **, register_t *); |
| 372 | int	do_sys_recvmsg_so(struct lwp *, int, struct socket *, |
| 373 | 	 struct msghdr *mp, struct mbuf **, struct mbuf **, register_t *); |
| 374 | |
| 375 | int	do_sys_bind(struct lwp *, int, struct sockaddr *); |
| 376 | int	do_sys_connect(struct lwp *, int, struct sockaddr *); |
| 377 | int	do_sys_accept(struct lwp *, int, struct sockaddr *, register_t *, |
| 378 | 	 const sigset_t *, int, int); |
| 379 | |
| 380 | int	do_sys_peeloff(struct socket *, void *); |
| 381 | /* |
| 382 | * Inline functions for sockets and socket buffering. |
| 383 | */ |
| 384 | |
| 385 | #include <sys/protosw.h> |
| 386 | #include <sys/mbuf.h> |
| 387 | |
| 388 | /* |
| 389 | * Do we need to notify the other side when I/O is possible? |
| 390 | */ |
| 391 | static __inline int |
| 392 | sb_notify(struct sockbuf *sb) |
| 393 | { |
| 394 | |
| 395 | 	KASSERT(solocked(sb->sb_so)); |
| 396 | |
| 397 | 	return sb->sb_flags & (SB_NOTIFY | SB_ASYNC | SB_UPCALL | SB_KNOTE); |
| 398 | } |
| 399 | |
| 400 | /* |
| 401 | * How much space is there in a socket buffer (so->so_snd or so->so_rcv)? |
| 402 | * Since the fields are unsigned, detect overflow and return 0. |
| 403 | */ |
| 404 | static __inline u_long |
| 405 | sbspace(const struct sockbuf *sb) |
| 406 | { |
| 407 | |
| 408 | 	KASSERT(solocked(sb->sb_so)); |
| 409 | 	if (sb->sb_hiwat <= sb->sb_cc || sb->sb_mbmax <= sb->sb_mbcnt) |
| 410 | 		return 0; |
| 411 | 	return lmin(sb->sb_hiwat - sb->sb_cc, sb->sb_mbmax - sb->sb_mbcnt); |
| 412 | } |
| 413 | |
| 414 | static __inline u_long |
| 415 | sbspace_oob(const struct sockbuf *sb) |
| 416 | { |
| 417 | 	u_long hiwat = sb->sb_hiwat; |
| 418 | |
| 419 | 	if (hiwat < ULONG_MAX - 1024) |
| 420 | 		hiwat += 1024; |
| 421 | |
| 422 | 	KASSERT(solocked(sb->sb_so)); |
| 423 | |
| 424 | 	if (hiwat <= sb->sb_cc || sb->sb_mbmax <= sb->sb_mbcnt) |
| 425 | 		return 0; |
| 426 | 	return lmin(hiwat - sb->sb_cc, sb->sb_mbmax - sb->sb_mbcnt); |
| 427 | } |
| 428 | |
| 429 | /* |
| 430 | * How much socket buffer space has been used? |
| 431 | */ |
| 432 | static __inline u_long |
| 433 | sbused(const struct sockbuf *sb) |
| 434 | { |
| 435 | |
| 436 | 	KASSERT(solocked(sb->sb_so)); |
| 437 | 	return sb->sb_cc; |
| 438 | } |
| 439 | |
| 440 | /* do we have to send all at once on a socket? */ |
| 441 | static __inline int |
| 442 | sosendallatonce(const struct socket *so) |
| 443 | { |
| 444 | |
| 445 | 	return so->so_proto->pr_flags & PR_ATOMIC; |
| 446 | } |
| 447 | |
| 448 | /* can we read something from so? */ |
| 449 | static __inline int |
| 450 | soreadable(const struct socket *so) |
| 451 | { |
| 452 | |
| 453 | 	KASSERT(solocked(so)); |
| 454 | |
| 455 | 	return so->so_rcv.sb_cc >= so->so_rcv.sb_lowat || |
| 456 | 	 (so->so_state & SS_CANTRCVMORE) != 0 || |
| 457 | 	 so->so_qlen != 0 || so->so_error != 0 || so->so_rerror != 0; |
| 458 | } |
| 459 | |
| 460 | /* can we write something to so? */ |
| 461 | static __inline int |
| 462 | sowritable(const struct socket *so) |
| 463 | { |
| 464 | |
| 465 | 	KASSERT(solocked(so)); |
| 466 | |
| 467 | 	return (sbspace(&so->so_snd) >= so->so_snd.sb_lowat && |
| 468 | 	 ((so->so_state & SS_ISCONNECTED) != 0 || |
| 469 | 	 (so->so_proto->pr_flags & PR_CONNREQUIRED) == 0)) || |
| 470 | 	 (so->so_state & SS_CANTSENDMORE) != 0 || |
| 471 | 	 so->so_error != 0; |
| 472 | } |
| 473 | |
| 474 | /* adjust counters in sb reflecting allocation of m */ |
| 475 | static __inline void |
| 476 | sballoc(struct sockbuf *sb, struct mbuf *m) |
| 477 | { |
| 478 | |
| 479 | 	KASSERT(solocked(sb->sb_so)); |
| 480 | |
| 481 | 	sb->sb_cc += m->m_len; |
| 482 | 	sb->sb_mbcnt += MSIZE; |
| 483 | 	if (m->m_flags & M_EXT) |
| 484 | 		sb->sb_mbcnt += m->m_ext.ext_size; |
| 485 | } |
| 486 | |
| 487 | /* adjust counters in sb reflecting freeing of m */ |
| 488 | static __inline void |
| 489 | sbfree(struct sockbuf *sb, struct mbuf *m) |
| 490 | { |
| 491 | |
| 492 | 	KASSERT(solocked(sb->sb_so)); |
| 493 | |
| 494 | 	sb->sb_cc -= m->m_len; |
| 495 | 	sb->sb_mbcnt -= MSIZE; |
| 496 | 	if (m->m_flags & M_EXT) |
| 497 | 		sb->sb_mbcnt -= m->m_ext.ext_size; |
| 498 | } |
| 499 | |
| 500 | static __inline void |
| 501 | sorwakeup(struct socket *so) |
| 502 | { |
| 503 | |
| 504 | 	KASSERT(solocked(so)); |
| 505 | |
| 506 | 	if (sb_notify(&so->so_rcv)) |
| 507 | 		sowakeup(so, &so->so_rcv, POLL_IN); |
| 508 | } |
| 509 | |
| 510 | static __inline void |
| 511 | sowwakeup(struct socket *so) |
| 512 | { |
| 513 | |
| 514 | 	KASSERT(solocked(so)); |
| 515 | |
| 516 | 	if (sb_notify(&so->so_snd)) |
| 517 | 		sowakeup(so, &so->so_snd, POLL_OUT); |
| 518 | } |
| 519 | |
| 520 | static __inline void |
| 521 | solock(struct socket *so) |
| 522 | { |
| 523 | 	kmutex_t *lock; |
| 524 | |
| 525 | 	lock = atomic_load_consume(&so->so_lock); |
| 526 | 	mutex_enter(lock); |
| 527 | 	if (__predict_false(lock != atomic_load_relaxed(&so->so_lock))) |
| 528 | 		solockretry(so, lock); |
| 529 | } |
| 530 | |
| 531 | static __inline void |
| 532 | sounlock(struct socket *so) |
| 533 | { |
| 534 | |
| 535 | 	mutex_exit(so->so_lock); |
| 536 | } |
| 537 | |
| 538 | #ifdef SOCKBUF_DEBUG |
| 539 | /* |
| 540 | * SBLASTRECORDCHK: check sb->sb_lastrecord is maintained correctly. |
| 541 | * SBLASTMBUFCHK: check sb->sb_mbtail is maintained correctly. |
| 542 | * |
| 543 | * => panic if the socket buffer is inconsistent. |
| 544 | * => 'where' is used for a panic message. |
| 545 | */ |
| 546 | void	sblastrecordchk(struct sockbuf *, const char *); |
| 547 | #define	SBLASTRECORDCHK(sb, where)	sblastrecordchk((sb), (where)) |
| 548 | |
| 549 | void	sblastmbufchk(struct sockbuf *, const char *); |
| 550 | #define	SBLASTMBUFCHK(sb, where)	sblastmbufchk((sb), (where)) |
| 551 | #define	SBCHECK(sb)			sbcheck(sb) |
| 552 | #else |
| 553 | #define	SBLASTRECORDCHK(sb, where)	/* nothing */ |
| 554 | #define	SBLASTMBUFCHK(sb, where)	/* nothing */ |
| 555 | #define	SBCHECK(sb)			/* nothing */ |
| 556 | #endif /* SOCKBUF_DEBUG */ |
| 557 | |
| 558 | /* sosend loan */ |
| 559 | vaddr_t	sokvaalloc(vaddr_t, vsize_t, struct socket *); |
| 560 | void	sokvafree(vaddr_t, vsize_t); |
| 561 | void	soloanfree(struct mbuf *, void *, size_t, void *); |
| 562 | |
| 563 | /* |
| 564 | * Values for socket-buffer-append priority argument to sbappendaddrchain(). |
| 565 | * The following flags are reserved for future implementation: |
| 566 | * |
| 567 | * SB_PRIO_NONE: honour normal socket-buffer limits. |
| 568 | * |
| 569 | * SB_PRIO_ONESHOT_OVERFLOW: if the socket has any space, |
| 570 | *	deliver the entire chain. Intended for large requests |
| 571 | * that should be delivered in their entirety, or not at all. |
| 572 | * |
| 573 | * SB_PRIO_OVERDRAFT: allow a small (2*MLEN) overflow, over and |
| 574 | *	above normal socket limits. Intended messages indicating |
| 575 | * buffer overflow in earlier normal/lower-priority messages . |
| 576 | * |
| 577 | * SB_PRIO_BESTEFFORT: Ignore limits entirely. Intended only for |
| 578 | * 	kernel-generated messages to specially-marked sockets which |
| 579 | *	require "reliable" delivery, nd where the source socket/protocol |
| 580 | *	message generator enforce some hard limit (but possibly well |
| 581 | *	above kern.sbmax). It is entirely up to the in-kernel source to |
| 582 | *	avoid complete mbuf exhaustion or DoS scenarios. |
| 583 | */ |
| 584 | #define SB_PRIO_NONE 	 	0 |
| 585 | #define SB_PRIO_ONESHOT_OVERFLOW 1 |
| 586 | #define SB_PRIO_OVERDRAFT	2 |
| 587 | #define SB_PRIO_BESTEFFORT	3 |
| 588 | |
| 589 | /* |
| 590 | * Accept filter functions (duh). |
| 591 | */ |
| 592 | int	accept_filt_getopt(struct socket *, struct sockopt *); |
| 593 | int	accept_filt_setopt(struct socket *, const struct sockopt *); |
| 594 | int	accept_filt_clear(struct socket *); |
| 595 | int	accept_filt_add(struct accept_filter *); |
| 596 | int	accept_filt_del(struct accept_filter *); |
| 597 | struct	accept_filter *accept_filt_get(char *); |
| 598 | #ifdef ACCEPT_FILTER_MOD |
| 599 | #ifdef SYSCTL_DECL |
| 600 | SYSCTL_DECL(_net_inet_accf); |
| 601 | #endif |
| 602 | void	accept_filter_init(void); |
| 603 | #endif |
| 604 | #ifdef DDB |
| 605 | int sofindproc(struct socket *so, int all, void (*pr)(const char *, ...)); |
| 606 | void socket_print(const char *modif, void (*pr)(const char *, ...)); |
| 607 | #endif |
| 608 | |
| 609 | #endif /* _KERNEL */ |
| 610 | |
| 611 | #endif /* !_SYS_SOCKETVAR_H_ */ |