2 * Copyright (c) 1982, 1986, 1990, 1993
3 * The Regents of the University of California. All rights reserved.
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 * 4. Neither the name of the University nor the names of its contributors
14 * may be used to endorse or promote products derived from this software
15 * without specific prior written permission.
17 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * @(#)socketvar.h 8.3 (Berkeley) 2/19/95
34 #ifndef _SYS_SOCKETVAR_H_
35 #define _SYS_SOCKETVAR_H_
37 #include <sys/queue.h> /* for TAILQ macros */
38 #include <sys/selinfo.h> /* for struct selinfo */
39 #include <sys/_lock.h>
40 #include <sys/_mutex.h>
42 #include <sys/sockbuf.h>
43 #include <sys/sockstate.h>
45 #include <sys/sockopt.h>
49 * Kernel structure per socket.
50 * Contains send and receive buffer queues,
51 * handle on protocol and pointer to protocol
52 * private data and error information.
54 typedef u_quad_t so_gen_t;
57 * Locking key to struct socket:
58 * (a) constant after allocation, no locking required.
59 * (b) locked by SOCK_LOCK(so).
60 * (c) locked by SOCKBUF_LOCK(&so->so_rcv).
61 * (d) locked by SOCKBUF_LOCK(&so->so_snd).
62 * (e) locked by ACCEPT_LOCK().
63 * (f) not locked since integer reads/writes are atomic.
64 * (g) used only as a sleep/wakeup address, no value.
65 * (h) locked by global mutex so_global_mtx.
68 int so_count; /* (b) reference count */
69 short so_type; /* (a) generic type, see socket.h */
70 short so_options; /* from socket call, see socket.h */
71 short so_linger; /* time to linger while closing */
72 short so_state; /* (b) internal state flags SS_* */
73 int so_qstate; /* (e) internal state flags SQ_* */
74 void *so_pcb; /* protocol control block */
75 struct protosw *so_proto; /* (a) protocol handle */
77 * Variables for connection queuing.
78 * Socket where accepts occur is so_head in all subsidiary sockets.
79 * If so_head is 0, socket is not related to an accept.
80 * For head socket so_incomp queues partially completed connections,
81 * while so_comp is a queue of connections ready to be accepted.
82 * If a connection is aborted and it has so_head set, then
83 * it has to be pulled out of either so_incomp or so_comp.
84 * We allow connections to queue up based on current queue lengths
85 * and limit on number of queued connections for this socket.
87 struct socket *so_head; /* (e) back pointer to listen socket */
88 TAILQ_HEAD(, socket) so_incomp; /* (e) queue of partial unaccepted connections */
89 TAILQ_HEAD(, socket) so_comp; /* (e) queue of complete unaccepted connections */
90 TAILQ_ENTRY(socket) so_list; /* (e) list of unaccepted connections */
91 u_short so_qlen; /* (e) number of unaccepted connections */
92 u_short so_incqlen; /* (e) number of unaccepted incomplete
94 u_short so_qlimit; /* (e) max number queued connections */
95 short so_timeo; /* (g) connection timeout */
96 u_short so_error; /* (f) error affecting connection */
97 struct sigio *so_sigio; /* [sg] information for async I/O or
98 out of band data (SIGURG) */
99 u_long so_oobmark; /* (c) chars to oob mark */
100 TAILQ_HEAD(, aiocblist) so_aiojobq; /* AIO ops waiting on socket */
101 struct sockbuf so_rcv, so_snd;
103 void (*so_upcall)(struct socket *, void *, int);
105 struct ucred *so_cred; /* (a) user credentials */
106 struct label *so_label; /* (b) MAC label for socket */
107 struct label *so_peerlabel; /* (b) cached MAC label for peer */
108 /* NB: generation count must not be first. */
109 so_gen_t so_gencnt; /* (h) generation count */
110 void *so_emuldata; /* (b) private data for emulators */
112 struct accept_filter *so_accept_filter;
113 void *so_accept_filter_arg; /* saved filter args */
114 char *so_accept_filter_str; /* saved user args */
116 int so_fibnum; /* routing domain for this socket */
120 * Global accept mutex to serialize access to accept queues and
121 * fields associated with multiple sockets. This allows us to
122 * avoid defining a lock order between listen and accept sockets
123 * until such time as it proves to be a good idea.
125 extern struct mtx accept_mtx;
126 #define ACCEPT_LOCK_ASSERT() mtx_assert(&accept_mtx, MA_OWNED)
127 #define ACCEPT_UNLOCK_ASSERT() mtx_assert(&accept_mtx, MA_NOTOWNED)
128 #define ACCEPT_LOCK() mtx_lock(&accept_mtx)
129 #define ACCEPT_UNLOCK() mtx_unlock(&accept_mtx)
132 * Per-socket mutex: we reuse the receive socket buffer mutex for space
133 * efficiency. This decision should probably be revisited as we optimize
134 * locking for the socket code.
136 #define SOCK_MTX(_so) SOCKBUF_MTX(&(_so)->so_rcv)
137 #define SOCK_LOCK(_so) SOCKBUF_LOCK(&(_so)->so_rcv)
138 #define SOCK_OWNED(_so) SOCKBUF_OWNED(&(_so)->so_rcv)
139 #define SOCK_UNLOCK(_so) SOCKBUF_UNLOCK(&(_so)->so_rcv)
140 #define SOCK_LOCK_ASSERT(_so) SOCKBUF_LOCK_ASSERT(&(_so)->so_rcv)
143 * Socket state bits stored in so_qstate.
145 #define SQ_INCOMP 0x0800 /* unaccepted, incomplete connection */
146 #define SQ_COMP 0x1000 /* unaccepted, complete connection */
149 * Externalized form of struct socket used by the sysctl(3) interface.
152 size_t xso_len; /* length of this structure */
153 struct socket *xso_so; /* makes a convenient handle sometimes */
158 caddr_t so_pcb; /* another convenient handle */
168 struct xsockbuf so_rcv, so_snd;
169 uid_t so_uid; /* XXX */
175 * Macros for sockets and socket buffering.
181 #define SBL_WAIT 0x00000001 /* Wait if not immediately available. */
182 #define SBL_NOINTR 0x00000002 /* Force non-interruptible sleep. */
183 #define SBL_VALID (SBL_WAIT | SBL_NOINTR)
186 * Do we need to notify the other side when I/O is possible?
188 #define sb_notify(sb) (((sb)->sb_flags & (SB_WAIT | SB_SEL | SB_ASYNC | \
189 SB_UPCALL | SB_AIO | SB_KNOTE)) != 0)
191 /* do we have to send all at once on a socket? */
192 #define sosendallatonce(so) \
193 ((so)->so_proto->pr_flags & PR_ATOMIC)
195 /* can we read something from so? */
196 #define soreadable(so) \
197 ((so)->so_rcv.sb_cc >= (so)->so_rcv.sb_lowat || \
198 ((so)->so_rcv.sb_state & SBS_CANTRCVMORE) || \
199 !TAILQ_EMPTY(&(so)->so_comp) || (so)->so_error)
201 /* can we write something to so? */
202 #define sowriteable(so) \
203 ((sbspace(&(so)->so_snd) >= (so)->so_snd.sb_lowat && \
204 (((so)->so_state&SS_ISCONNECTED) || \
205 ((so)->so_proto->pr_flags&PR_CONNREQUIRED)==0)) || \
206 ((so)->so_snd.sb_state & SBS_CANTSENDMORE) || \
210 * soref()/sorele() ref-count the socket structure. Note that you must
211 * still explicitly close the socket, but the last ref count will free
214 #define soref(so) do { \
215 SOCK_LOCK_ASSERT(so); \
219 #define sorele(so) do { \
220 ACCEPT_LOCK_ASSERT(); \
221 SOCK_LOCK_ASSERT(so); \
222 if ((so)->so_count <= 0) \
224 if (--(so)->so_count == 0) \
232 #define sotryfree(so) do { \
233 ACCEPT_LOCK_ASSERT(); \
234 SOCK_LOCK_ASSERT(so); \
235 if ((so)->so_count == 0) \
244 * In sorwakeup() and sowwakeup(), acquire the socket buffer lock to
245 * avoid a non-atomic test-and-wakeup. However, sowakeup is
246 * responsible for releasing the lock if it is called. We unlock only
247 * if we don't call into sowakeup. If any code is introduced that
248 * directly invokes the underlying sowakeup() primitives, it must
249 * maintain the same semantics.
251 #define sorwakeup_locked(so) do { \
252 SOCKBUF_LOCK_ASSERT(&(so)->so_rcv); \
253 if (sb_notify(&(so)->so_rcv)) \
254 sowakeup((so), &(so)->so_rcv); \
256 SOCKBUF_UNLOCK(&(so)->so_rcv); \
259 #define sorwakeup(so) do { \
260 SOCKBUF_LOCK(&(so)->so_rcv); \
261 sorwakeup_locked(so); \
264 #define sowwakeup_locked(so) do { \
265 SOCKBUF_LOCK_ASSERT(&(so)->so_snd); \
266 if (sb_notify(&(so)->so_snd)) \
267 sowakeup((so), &(so)->so_snd); \
269 SOCKBUF_UNLOCK(&(so)->so_snd); \
272 #define sowwakeup(so) do { \
273 SOCKBUF_LOCK(&(so)->so_snd); \
274 sowwakeup_locked(so); \
277 struct accept_filter {
279 void (*accf_callback)
280 (struct socket *so, void *arg, int waitflag);
281 void * (*accf_create)
282 (struct socket *so, char *arg);
285 SLIST_ENTRY(accept_filter) accf_next;
288 #ifdef MALLOC_DECLARE
289 MALLOC_DECLARE(M_ACCF);
290 MALLOC_DECLARE(M_PCB);
291 MALLOC_DECLARE(M_SONAME);
294 extern int maxsockets;
295 extern u_long sb_max;
296 extern struct uma_zone *socket_zone;
297 extern so_gen_t so_gencnt;
305 * From uipc_socket and friends
307 int sockargs(struct mbuf **mp, caddr_t buf, int buflen, int type);
308 int getsockaddr(struct sockaddr **namp, caddr_t uaddr, size_t len);
309 void soabort(struct socket *so);
310 int soaccept(struct socket *so, struct sockaddr **nam);
311 int socheckuid(struct socket *so, uid_t uid);
312 int sobind(struct socket *so, struct sockaddr *nam, struct thread *td);
313 int soclose(struct socket *so);
314 int soconnect(struct socket *so, struct sockaddr *nam, struct thread *td);
315 int soconnect2(struct socket *so1, struct socket *so2);
316 int socow_setup(struct mbuf *m0, struct uio *uio);
317 int socreate(int dom, struct socket **aso, int type, int proto,
318 struct ucred *cred, struct thread *td);
319 int sodisconnect(struct socket *so);
320 struct sockaddr *sodupsockaddr(const struct sockaddr *sa, int mflags);
321 void sofree(struct socket *so);
322 void sohasoutofband(struct socket *so);
323 int solisten(struct socket *so, int backlog, struct thread *td);
324 void solisten_proto(struct socket *so, int backlog);
325 int solisten_proto_check(struct socket *so);
327 sonewconn(struct socket *head, int connstatus);
330 int sopoll(struct socket *so, int events, struct ucred *active_cred,
332 int sopoll_generic(struct socket *so, int events,
333 struct ucred *active_cred, struct thread *td);
334 int soreceive(struct socket *so, struct sockaddr **paddr, struct uio *uio,
335 struct mbuf **mp0, struct mbuf **controlp, int *flagsp);
336 int soreceive_dgram(struct socket *so, struct sockaddr **paddr,
337 struct uio *uio, struct mbuf **mp0, struct mbuf **controlp,
339 int soreceive_generic(struct socket *so, struct sockaddr **paddr,
340 struct uio *uio, struct mbuf **mp0, struct mbuf **controlp,
342 int soreserve(struct socket *so, u_long sndcc, u_long rcvcc);
343 void sorflush(struct socket *so);
344 int sosend(struct socket *so, struct sockaddr *addr, struct uio *uio,
345 struct mbuf *top, struct mbuf *control, int flags,
347 int sosend_dgram(struct socket *so, struct sockaddr *addr,
348 struct uio *uio, struct mbuf *top, struct mbuf *control,
349 int flags, struct thread *td);
350 int sosend_generic(struct socket *so, struct sockaddr *addr,
351 struct uio *uio, struct mbuf *top, struct mbuf *control,
352 int flags, struct thread *td);
353 int soshutdown(struct socket *so, int how);
354 void sotoxsocket(struct socket *so, struct xsocket *xso);
355 void sowakeup(struct socket *so, struct sockbuf *sb);
358 * Accept filter functions (duh).
360 int accept_filt_add(struct accept_filter *filt);
361 int accept_filt_del(char *name);
362 struct accept_filter *accept_filt_get(char *name);
363 #ifdef ACCEPT_FILTER_MOD
365 SYSCTL_DECL(_net_inet_accf);
367 int accept_filt_generic_mod_event(module_t mod, int event, void *data);
372 #endif /* !_SYS_SOCKETVAR_H_ */