socketvar.h revision 95959
1/*-
2 * Copyright (c) 1982, 1986, 1990, 1993
3 *	The Regents of the University of California.  All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
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 * 3. All advertising materials mentioning features or use of this software
14 *    must display the following acknowledgement:
15 *	This product includes software developed by the University of
16 *	California, Berkeley and its contributors.
17 * 4. Neither the name of the University nor the names of its contributors
18 *    may be used to endorse or promote products derived from this software
19 *    without specific prior written permission.
20 *
21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31 * SUCH DAMAGE.
32 *
33 *	@(#)socketvar.h	8.3 (Berkeley) 2/19/95
34 * $FreeBSD: head/sys/sys/socketvar.h 95959 2002-05-02 22:03:19Z alfred $
35 */
36
37#ifndef _SYS_SOCKETVAR_H_
38#define _SYS_SOCKETVAR_H_
39
40#include <sys/queue.h>			/* for TAILQ macros */
41#include <sys/selinfo.h>		/* for struct selinfo */
42
43/*
44 * Kernel structure per socket.
45 * Contains send and receive buffer queues,
46 * handle on protocol and pointer to protocol
47 * private data and error information.
48 */
49typedef	u_quad_t so_gen_t;
50
51/*
52 * List of locks:
53 * (c)	const, inited in either socreate() or sonewconn()
54 * (m)	sb_mtx mutex
55 * (mr)	so_rcv.sb_mtx mutex
56 * (sg)	sigio_lock sx
57 * (sh)	sohead_lock sx
58 *
59 * Lock of so_rcv.sb_mtx can duplicate, provided that sohead_lock
60 * is exclusively locked.
61 *
62 * Brackets mean that this data is not protected yet.
63 */
64struct socket {
65	int	so_count;		/* reference count */
66	short	so_type;		/* generic type, see socket.h */
67	short	so_options;		/* from socket call, see socket.h */
68	short	so_linger;		/* time to linger while closing */
69	short	so_state;		/* internal state flags SS_*, below */
70	caddr_t	so_pcb;			/* protocol control block */
71	struct	protosw *so_proto;	/* protocol handle */
72/*
73 * Variables for connection queuing.
74 * Socket where accepts occur is so_head in all subsidiary sockets.
75 * If so_head is 0, socket is not related to an accept.
76 * For head socket so_incomp queues partially completed connections,
77 * while so_comp is a queue of connections ready to be accepted.
78 * If a connection is aborted and it has so_head set, then
79 * it has to be pulled out of either so_incomp or so_comp.
80 * We allow connections to queue up based on current queue lengths
81 * and limit on number of queued connections for this socket.
82 */
83	struct	socket *so_head;	/* back pointer to accept socket */
84	TAILQ_HEAD(, socket) so_incomp;	/* queue of partial unaccepted connections */
85	TAILQ_HEAD(, socket) so_comp;	/* queue of complete unaccepted connections */
86	TAILQ_ENTRY(socket) so_list;	/* list of unaccepted connections */
87	short	so_qlen;		/* number of unaccepted connections */
88	short	so_incqlen;		/* number of unaccepted incomplete
89					   connections */
90	short	so_qlimit;		/* max number queued connections */
91	short	so_timeo;		/* connection timeout */
92	u_short	so_error;		/* error affecting connection */
93	struct	sigio *so_sigio;	/* [sg] information for async I/O or
94					   out of band data (SIGURG) */
95	u_long	so_oobmark;		/* chars to oob mark */
96	TAILQ_HEAD(, aiocblist) so_aiojobq; /* AIO ops waiting on socket */
97/*
98 * Variables for socket buffering.
99 */
100	struct sockbuf {
101		u_long	sb_cc;		/* actual chars in buffer */
102		u_long	sb_hiwat;	/* max actual char count */
103		u_long	sb_mbcnt;	/* chars of mbufs used */
104		u_long	sb_mbmax;	/* max chars of mbufs to use */
105		long	sb_lowat;	/* low water mark */
106		struct	mbuf *sb_mb;	/* the mbuf chain */
107		struct	selinfo sb_sel;	/* process selecting read/write */
108		short	sb_flags;	/* flags, see below */
109		short	sb_timeo;	/* timeout for read/write */
110	} so_rcv, so_snd;
111#define	SB_MAX		(256*1024)	/* default for max chars in sockbuf */
112#define	SB_LOCK		0x01		/* lock on data queue */
113#define	SB_WANT		0x02		/* someone is waiting to lock */
114#define	SB_WAIT		0x04		/* someone is waiting for data/space */
115#define	SB_SEL		0x08		/* someone is selecting */
116#define	SB_ASYNC	0x10		/* ASYNC I/O, need signals */
117#define	SB_UPCALL	0x20		/* someone wants an upcall */
118#define	SB_NOINTR	0x40		/* operations not interruptible */
119#define SB_AIO		0x80		/* AIO operations queued */
120#define SB_KNOTE	0x100		/* kernel note attached */
121
122	void	(*so_upcall)(struct socket *, void *, int);
123	void	*so_upcallarg;
124	struct	ucred *so_cred;		/* user credentials */
125	/* NB: generation count must not be first; easiest to make it last. */
126	so_gen_t so_gencnt;		/* generation count */
127	void	*so_emuldata;		/* private data for emulators */
128 	struct so_accf {
129		struct	accept_filter *so_accept_filter;
130		void	*so_accept_filter_arg;	/* saved filter args */
131		char	*so_accept_filter_str;	/* saved user args */
132	} *so_accf;
133};
134
135/*
136 * Socket state bits.
137 */
138#define	SS_NOFDREF		0x0001	/* no file table ref any more */
139#define	SS_ISCONNECTED		0x0002	/* socket connected to a peer */
140#define	SS_ISCONNECTING		0x0004	/* in process of connecting to peer */
141#define	SS_ISDISCONNECTING	0x0008	/* in process of disconnecting */
142#define	SS_CANTSENDMORE		0x0010	/* can't send more data to peer */
143#define	SS_CANTRCVMORE		0x0020	/* can't receive more data from peer */
144#define	SS_RCVATMARK		0x0040	/* at mark on input */
145
146#define	SS_NBIO			0x0100	/* non-blocking ops */
147#define	SS_ASYNC		0x0200	/* async i/o notify */
148#define	SS_ISCONFIRMING		0x0400	/* deciding to accept connection req */
149
150#define	SS_INCOMP		0x0800	/* unaccepted, incomplete connection */
151#define	SS_COMP			0x1000	/* unaccepted, complete connection */
152#define	SS_ISDISCONNECTED	0x2000	/* socket disconnected from peer */
153
154/*
155 * Externalized form of struct socket used by the sysctl(3) interface.
156 */
157struct xsocket {
158	size_t	xso_len;	/* length of this structure */
159	struct	socket *xso_so;	/* makes a convenient handle sometimes */
160	short	so_type;
161	short	so_options;
162	short	so_linger;
163	short	so_state;
164	caddr_t	so_pcb;		/* another convenient handle */
165	int	xso_protocol;
166	int	xso_family;
167	short	so_qlen;
168	short	so_incqlen;
169	short	so_qlimit;
170	short	so_timeo;
171	u_short	so_error;
172	pid_t	so_pgid;
173	u_long	so_oobmark;
174	struct xsockbuf {
175		u_long	sb_cc;
176		u_long	sb_hiwat;
177		u_long	sb_mbcnt;
178		u_long	sb_mbmax;
179		long	sb_lowat;
180		short	sb_flags;
181		short	sb_timeo;
182	} so_rcv, so_snd;
183	uid_t	so_uid;		/* XXX */
184};
185
186/*
187 * Macros for sockets and socket buffering.
188 */
189
190/*
191 * Do we need to notify the other side when I/O is possible?
192 */
193#define	sb_notify(sb)	(((sb)->sb_flags & (SB_WAIT | SB_SEL | SB_ASYNC | \
194    SB_UPCALL | SB_AIO | SB_KNOTE)) != 0)
195
196/*
197 * How much space is there in a socket buffer (so->so_snd or so->so_rcv)?
198 * This is problematical if the fields are unsigned, as the space might
199 * still be negative (cc > hiwat or mbcnt > mbmax).  Should detect
200 * overflow and return 0.  Should use "lmin" but it doesn't exist now.
201 */
202#define	sbspace(sb) \
203    ((long) imin((int)((sb)->sb_hiwat - (sb)->sb_cc), \
204	 (int)((sb)->sb_mbmax - (sb)->sb_mbcnt)))
205
206/* do we have to send all at once on a socket? */
207#define	sosendallatonce(so) \
208    ((so)->so_proto->pr_flags & PR_ATOMIC)
209
210/* can we read something from so? */
211#define	soreadable(so) \
212    ((so)->so_rcv.sb_cc >= (so)->so_rcv.sb_lowat || \
213	((so)->so_state & SS_CANTRCVMORE) || \
214	!TAILQ_EMPTY(&(so)->so_comp) || (so)->so_error)
215
216/* can we write something to so? */
217#define	sowriteable(so) \
218    ((sbspace(&(so)->so_snd) >= (so)->so_snd.sb_lowat && \
219	(((so)->so_state&SS_ISCONNECTED) || \
220	  ((so)->so_proto->pr_flags&PR_CONNREQUIRED)==0)) || \
221     ((so)->so_state & SS_CANTSENDMORE) || \
222     (so)->so_error)
223
224/* adjust counters in sb reflecting allocation of m */
225#define	sballoc(sb, m) { \
226	(sb)->sb_cc += (m)->m_len; \
227	(sb)->sb_mbcnt += MSIZE; \
228	if ((m)->m_flags & M_EXT) \
229		(sb)->sb_mbcnt += (m)->m_ext.ext_size; \
230}
231
232/* adjust counters in sb reflecting freeing of m */
233#define	sbfree(sb, m) { \
234	(sb)->sb_cc -= (m)->m_len; \
235	(sb)->sb_mbcnt -= MSIZE; \
236	if ((m)->m_flags & M_EXT) \
237		(sb)->sb_mbcnt -= (m)->m_ext.ext_size; \
238}
239
240/*
241 * Set lock on sockbuf sb; sleep if lock is already held.
242 * Unless SB_NOINTR is set on sockbuf, sleep is interruptible.
243 * Returns error without lock if sleep is interrupted.
244 */
245#define sblock(sb, wf) ((sb)->sb_flags & SB_LOCK ? \
246		(((wf) == M_WAITOK) ? sb_lock(sb) : EWOULDBLOCK) : \
247		((sb)->sb_flags |= SB_LOCK), 0)
248
249/* release lock on sockbuf sb */
250#define	sbunlock(sb) { \
251	(sb)->sb_flags &= ~SB_LOCK; \
252	if ((sb)->sb_flags & SB_WANT) { \
253		(sb)->sb_flags &= ~SB_WANT; \
254		wakeup((caddr_t)&(sb)->sb_flags); \
255	} \
256}
257
258/*
259 * soref()/sorele() ref-count the socket structure.  Note that you must
260 * still explicitly close the socket, but the last ref count will free
261 * the structure.
262 */
263#define soref(so)	do {			\
264				++(so)->so_count; \
265			} while (0)
266
267#define sorele(so)	do {				\
268				if ((so)->so_count <= 0)	\
269					panic("sorele");\
270				if (--(so)->so_count == 0)\
271					sofree(so);	\
272			} while (0)
273
274#define sotryfree(so)	do {				\
275				if ((so)->so_count == 0)	\
276					sofree(so);	\
277			} while(0)
278
279#define	sorwakeup_locked(so)	do {					\
280					if (sb_notify(&(so)->so_rcv))	\
281						sowakeup((so), &(so)->so_rcv); \
282				} while (0)
283
284#define	sorwakeup(so)		do {					\
285					sorwakeup_locked(so);		\
286				} while (0)
287
288#define	sowwakeup_locked(so)	do {					\
289					if (sb_notify(&(so)->so_snd))	\
290						sowakeup((so), &(so)->so_snd); \
291				} while (0)
292
293#define	sowwakeup(so)		do {					\
294					sowwakeup_locked(so);		\
295				} while (0)
296
297#ifdef _KERNEL
298
299/*
300 * Argument structure for sosetopt et seq.  This is in the KERNEL
301 * section because it will never be visible to user code.
302 */
303enum sopt_dir { SOPT_GET, SOPT_SET };
304struct sockopt {
305	enum	sopt_dir sopt_dir; /* is this a get or a set? */
306	int	sopt_level;	/* second arg of [gs]etsockopt */
307	int	sopt_name;	/* third arg of [gs]etsockopt */
308	void   *sopt_val;	/* fourth arg of [gs]etsockopt */
309	size_t	sopt_valsize;	/* (almost) fifth arg of [gs]etsockopt */
310	struct	thread *sopt_td; /* calling thread or null if kernel */
311};
312
313struct sf_buf {
314	SLIST_ENTRY(sf_buf) free_list;	/* list of free buffer slots */
315	struct		vm_page *m;	/* currently mapped page */
316	vm_offset_t	kva;		/* va of mapping */
317};
318
319struct accept_filter {
320	char	accf_name[16];
321	void	(*accf_callback)
322		(struct socket *so, void *arg, int waitflag);
323	void *	(*accf_create)
324		(struct socket *so, char *arg);
325	void	(*accf_destroy)
326		(struct socket *so);
327	SLIST_ENTRY(accept_filter) accf_next;
328};
329
330#ifdef MALLOC_DECLARE
331MALLOC_DECLARE(M_ACCF);
332MALLOC_DECLARE(M_PCB);
333MALLOC_DECLARE(M_SONAME);
334#endif
335
336extern int	maxsockets;
337extern u_long	sb_max;
338extern struct uma_zone *socket_zone;
339extern so_gen_t so_gencnt;
340
341struct file;
342struct filedesc;
343struct knote;
344struct mbuf;
345struct sockaddr;
346struct stat;
347struct ucred;
348struct uio;
349
350/*
351 * File operations on sockets.
352 */
353int	soo_read(struct file *fp, struct uio *uio, struct ucred *cred,
354	    int flags, struct thread *td);
355int	soo_write(struct file *fp, struct uio *uio, struct ucred *cred,
356	    int flags, struct thread *td);
357int	soo_close(struct file *fp, struct thread *td);
358int	soo_ioctl(struct file *fp, u_long cmd, caddr_t data,
359	    struct thread *td);
360int	soo_poll(struct file *fp, int events, struct ucred *cred,
361	    struct thread *td);
362int	soo_stat(struct file *fp, struct stat *ub, struct thread *td);
363int	sokqfilter(struct file *fp, struct knote *kn);
364
365/*
366 * From uipc_socket and friends
367 */
368struct	sockaddr *dup_sockaddr(struct sockaddr *sa, int canwait);
369int	sockargs(struct mbuf **mp, caddr_t buf, int buflen, int type);
370int	getsockaddr(struct sockaddr **namp, caddr_t uaddr, size_t len);
371void	sbappend(struct sockbuf *sb, struct mbuf *m);
372int	sbappendaddr(struct sockbuf *sb, struct sockaddr *asa,
373	    struct mbuf *m0, struct mbuf *control);
374int	sbappendcontrol(struct sockbuf *sb, struct mbuf *m0,
375	    struct mbuf *control);
376void	sbappendrecord(struct sockbuf *sb, struct mbuf *m0);
377void	sbcheck(struct sockbuf *sb);
378void	sbcompress(struct sockbuf *sb, struct mbuf *m, struct mbuf *n);
379struct mbuf *
380	sbcreatecontrol(caddr_t p, int size, int type, int level);
381void	sbdrop(struct sockbuf *sb, int len);
382void	sbdroprecord(struct sockbuf *sb);
383void	sbflush(struct sockbuf *sb);
384void	sbinsertoob(struct sockbuf *sb, struct mbuf *m0);
385void	sbrelease(struct sockbuf *sb, struct socket *so);
386int	sbreserve(struct sockbuf *sb, u_long cc, struct socket *so,
387	    struct thread *td);
388void	sbtoxsockbuf(struct sockbuf *sb, struct xsockbuf *xsb);
389int	sbwait(struct sockbuf *sb);
390int	sb_lock(struct sockbuf *sb);
391int	soabort(struct socket *so);
392int	soaccept(struct socket *so, struct sockaddr **nam);
393int	socheckuid(struct socket *so, uid_t uid);
394int	socheckproc(struct socket *so, struct proc *p);
395struct	socket *soalloc(int waitok);
396int	sobind(struct socket *so, struct sockaddr *nam, struct thread *td);
397void	socantrcvmore(struct socket *so);
398void	socantsendmore(struct socket *so);
399int	soclose(struct socket *so);
400int	soconnect(struct socket *so, struct sockaddr *nam, struct thread *td);
401int	soconnect2(struct socket *so1, struct socket *so2);
402int	socreate(int dom, struct socket **aso, int type, int proto,
403	    struct ucred *cred, struct thread *td);
404int	sodisconnect(struct socket *so);
405void	soisconnected_locked(struct socket *so);
406void	sofree(struct socket *so);
407int	sogetopt(struct socket *so, struct sockopt *sopt);
408void	sohasoutofband(struct socket *so);
409void	soisconnected(struct socket *so);
410void	soisconnecting(struct socket *so);
411void	soisdisconnected(struct socket *so);
412void	soisdisconnected_locked(struct socket *so);
413void	soisdisconnecting(struct socket *so);
414int	solisten(struct socket *so, int backlog, struct thread *td);
415struct socket *
416	sonewconn(struct socket *head, int connstatus);
417int	sooptcopyin(struct sockopt *sopt, void *buf, size_t len, size_t minlen);
418int	sooptcopyout(struct sockopt *sopt, void *buf, size_t len);
419
420/* XXX; prepare mbuf for (__FreeBSD__ < 3) routines. */
421int	soopt_getm(struct sockopt *sopt, struct mbuf **mp);
422int	soopt_mcopyin(struct sockopt *sopt, struct mbuf *m);
423int	soopt_mcopyout(struct sockopt *sopt, struct mbuf *m);
424
425int	sopoll(struct socket *so, int events, struct ucred *cred,
426	    struct thread *td);
427int	soreceive(struct socket *so, struct sockaddr **paddr, struct uio *uio,
428	    struct mbuf **mp0, struct mbuf **controlp, int *flagsp);
429int	soreserve(struct socket *so, u_long sndcc, u_long rcvcc);
430void	sorflush(struct socket *so);
431int	sosend(struct socket *so, struct sockaddr *addr, struct uio *uio,
432	    struct mbuf *top, struct mbuf *control, int flags,
433	    struct thread *td);
434int	sosetopt(struct socket *so, struct sockopt *sopt);
435int	soshutdown(struct socket *so, int how);
436void	sotoxsocket(struct socket *so, struct xsocket *xso);
437void	sowakeup(struct socket *so, struct sockbuf *sb);
438
439/*
440 * Accept filter functions (duh).
441 */
442int	accept_filt_add(struct accept_filter *filt);
443int	accept_filt_del(char *name);
444struct	accept_filter *accept_filt_get(char *name);
445#ifdef ACCEPT_FILTER_MOD
446#ifdef SYSCTL_DECL
447SYSCTL_DECL(_net_inet_accf);
448#endif
449int	accept_filt_generic_mod_event(module_t mod, int event, void *data);
450#endif
451
452#endif /* _KERNEL */
453
454#endif /* !_SYS_SOCKETVAR_H_ */
455