Annotation of src/sys/kern/uipc_socket.c, Revision 1.160.2.4
1.160.2.4! yamt 1: /* $NetBSD: uipc_socket.c,v 1.160.2.3 2009/09/16 13:38:01 yamt Exp $ */
1.64 thorpej 2:
3: /*-
1.160.2.2 yamt 4: * Copyright (c) 2002, 2007, 2008, 2009 The NetBSD Foundation, Inc.
1.64 thorpej 5: * All rights reserved.
6: *
7: * This code is derived from software contributed to The NetBSD Foundation
1.160.2.2 yamt 8: * by Jason R. Thorpe of Wasabi Systems, Inc, and by Andrew Doran.
1.64 thorpej 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: */
1.16 cgd 31:
1.1 cgd 32: /*
1.159 ad 33: * Copyright (c) 2004 The FreeBSD Foundation
34: * Copyright (c) 2004 Robert Watson
1.15 mycroft 35: * Copyright (c) 1982, 1986, 1988, 1990, 1993
36: * The Regents of the University of California. All rights reserved.
1.1 cgd 37: *
38: * Redistribution and use in source and binary forms, with or without
39: * modification, are permitted provided that the following conditions
40: * are met:
41: * 1. Redistributions of source code must retain the above copyright
42: * notice, this list of conditions and the following disclaimer.
43: * 2. Redistributions in binary form must reproduce the above copyright
44: * notice, this list of conditions and the following disclaimer in the
45: * documentation and/or other materials provided with the distribution.
1.85 agc 46: * 3. Neither the name of the University nor the names of its contributors
1.1 cgd 47: * may be used to endorse or promote products derived from this software
48: * without specific prior written permission.
49: *
50: * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
51: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
52: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
53: * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
54: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
55: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
56: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
57: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
58: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
59: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
60: * SUCH DAMAGE.
61: *
1.32 fvdl 62: * @(#)uipc_socket.c 8.6 (Berkeley) 5/2/95
1.1 cgd 63: */
1.59 lukem 64:
65: #include <sys/cdefs.h>
1.160.2.4! yamt 66: __KERNEL_RCSID(0, "$NetBSD: uipc_socket.c,v 1.160.2.3 2009/09/16 13:38:01 yamt Exp $");
1.64 thorpej 67:
1.160.2.2 yamt 68: #include "opt_compat_netbsd.h"
1.64 thorpej 69: #include "opt_sock_counters.h"
70: #include "opt_sosend_loan.h"
1.81 martin 71: #include "opt_mbuftrace.h"
1.84 ragge 72: #include "opt_somaxkva.h"
1.160.2.2 yamt 73: #include "opt_multiprocessor.h" /* XXX */
1.1 cgd 74:
1.9 mycroft 75: #include <sys/param.h>
76: #include <sys/systm.h>
77: #include <sys/proc.h>
78: #include <sys/file.h>
1.142 dyoung 79: #include <sys/filedesc.h>
1.160.2.2 yamt 80: #include <sys/kmem.h>
1.9 mycroft 81: #include <sys/mbuf.h>
82: #include <sys/domain.h>
83: #include <sys/kernel.h>
84: #include <sys/protosw.h>
85: #include <sys/socket.h>
86: #include <sys/socketvar.h>
1.21 christos 87: #include <sys/signalvar.h>
1.9 mycroft 88: #include <sys/resourcevar.h>
1.160.2.2 yamt 89: #include <sys/uidinfo.h>
1.72 jdolecek 90: #include <sys/event.h>
1.89 christos 91: #include <sys/poll.h>
1.118 elad 92: #include <sys/kauth.h>
1.136 ad 93: #include <sys/mutex.h>
94: #include <sys/condvar.h>
1.37 thorpej 95:
1.160.2.2 yamt 96: #ifdef COMPAT_50
97: #include <compat/sys/time.h>
98: #include <compat/sys/socket.h>
99: #endif
100:
1.64 thorpej 101: #include <uvm/uvm.h>
102:
1.77 thorpej 103: MALLOC_DEFINE(M_SOOPTS, "soopts", "socket options");
104: MALLOC_DEFINE(M_SONAME, "soname", "socket name");
1.37 thorpej 105:
1.142 dyoung 106: extern const struct fileops socketops;
107:
1.54 lukem 108: extern int somaxconn; /* patchable (XXX sysctl) */
109: int somaxconn = SOMAXCONN;
1.160 ad 110: kmutex_t *softnet_lock;
1.49 jonathan 111:
1.64 thorpej 112: #ifdef SOSEND_COUNTERS
113: #include <sys/device.h>
114:
1.113 thorpej 115: static struct evcnt sosend_loan_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
1.64 thorpej 116: NULL, "sosend", "loan big");
1.113 thorpej 117: static struct evcnt sosend_copy_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
1.64 thorpej 118: NULL, "sosend", "copy big");
1.113 thorpej 119: static struct evcnt sosend_copy_small = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
1.64 thorpej 120: NULL, "sosend", "copy small");
1.113 thorpej 121: static struct evcnt sosend_kvalimit = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
1.64 thorpej 122: NULL, "sosend", "kva limit");
123:
124: #define SOSEND_COUNTER_INCR(ev) (ev)->ev_count++
125:
1.101 matt 126: EVCNT_ATTACH_STATIC(sosend_loan_big);
127: EVCNT_ATTACH_STATIC(sosend_copy_big);
128: EVCNT_ATTACH_STATIC(sosend_copy_small);
129: EVCNT_ATTACH_STATIC(sosend_kvalimit);
1.64 thorpej 130: #else
131:
132: #define SOSEND_COUNTER_INCR(ev) /* nothing */
133:
134: #endif /* SOSEND_COUNTERS */
135:
1.119 yamt 136: static struct callback_entry sokva_reclaimerentry;
1.1 cgd 137:
1.160.2.2 yamt 138: #if defined(SOSEND_NO_LOAN) || defined(MULTIPROCESSOR)
1.121 yamt 139: int sock_loan_thresh = -1;
1.71 thorpej 140: #else
1.121 yamt 141: int sock_loan_thresh = 4096;
1.65 thorpej 142: #endif
1.64 thorpej 143:
1.136 ad 144: static kmutex_t so_pendfree_lock;
1.113 thorpej 145: static struct mbuf *so_pendfree;
1.64 thorpej 146:
1.84 ragge 147: #ifndef SOMAXKVA
148: #define SOMAXKVA (16 * 1024 * 1024)
149: #endif
150: int somaxkva = SOMAXKVA;
1.113 thorpej 151: static int socurkva;
1.136 ad 152: static kcondvar_t socurkva_cv;
1.64 thorpej 153:
1.160.2.4! yamt 154: static kauth_listener_t socket_listener;
! 155:
1.64 thorpej 156: #define SOCK_LOAN_CHUNK 65536
157:
1.117 yamt 158: static size_t sodopendfree(void);
159: static size_t sodopendfreel(void);
1.93 yamt 160:
1.160.2.2 yamt 161: static void sysctl_kern_somaxkva_setup(void);
162: static struct sysctllog *socket_sysctllog;
163:
1.113 thorpej 164: static vsize_t
1.129 yamt 165: sokvareserve(struct socket *so, vsize_t len)
1.80 yamt 166: {
1.98 christos 167: int error;
1.80 yamt 168:
1.136 ad 169: mutex_enter(&so_pendfree_lock);
1.80 yamt 170: while (socurkva + len > somaxkva) {
1.93 yamt 171: size_t freed;
172:
173: /*
174: * try to do pendfree.
175: */
176:
1.117 yamt 177: freed = sodopendfreel();
1.93 yamt 178:
179: /*
180: * if some kva was freed, try again.
181: */
182:
183: if (freed)
1.80 yamt 184: continue;
1.93 yamt 185:
1.80 yamt 186: SOSEND_COUNTER_INCR(&sosend_kvalimit);
1.136 ad 187: error = cv_wait_sig(&socurkva_cv, &so_pendfree_lock);
1.98 christos 188: if (error) {
189: len = 0;
190: break;
191: }
1.80 yamt 192: }
1.93 yamt 193: socurkva += len;
1.136 ad 194: mutex_exit(&so_pendfree_lock);
1.98 christos 195: return len;
1.95 yamt 196: }
197:
1.113 thorpej 198: static void
1.95 yamt 199: sokvaunreserve(vsize_t len)
200: {
201:
1.136 ad 202: mutex_enter(&so_pendfree_lock);
1.95 yamt 203: socurkva -= len;
1.136 ad 204: cv_broadcast(&socurkva_cv);
205: mutex_exit(&so_pendfree_lock);
1.95 yamt 206: }
207:
208: /*
209: * sokvaalloc: allocate kva for loan.
210: */
211:
212: vaddr_t
213: sokvaalloc(vsize_t len, struct socket *so)
214: {
215: vaddr_t lva;
216:
217: /*
218: * reserve kva.
219: */
220:
1.98 christos 221: if (sokvareserve(so, len) == 0)
222: return 0;
1.93 yamt 223:
224: /*
225: * allocate kva.
226: */
1.80 yamt 227:
1.109 yamt 228: lva = uvm_km_alloc(kernel_map, len, 0, UVM_KMF_VAONLY | UVM_KMF_WAITVA);
1.95 yamt 229: if (lva == 0) {
230: sokvaunreserve(len);
1.80 yamt 231: return (0);
1.95 yamt 232: }
1.80 yamt 233:
234: return lva;
235: }
236:
1.93 yamt 237: /*
238: * sokvafree: free kva for loan.
239: */
240:
1.80 yamt 241: void
242: sokvafree(vaddr_t sva, vsize_t len)
243: {
1.93 yamt 244:
245: /*
246: * free kva.
247: */
1.80 yamt 248:
1.109 yamt 249: uvm_km_free(kernel_map, sva, len, UVM_KMF_VAONLY);
1.93 yamt 250:
251: /*
252: * unreserve kva.
253: */
254:
1.95 yamt 255: sokvaunreserve(len);
1.80 yamt 256: }
257:
1.64 thorpej 258: static void
1.134 christos 259: sodoloanfree(struct vm_page **pgs, void *buf, size_t size)
1.64 thorpej 260: {
1.156 yamt 261: vaddr_t sva, eva;
1.64 thorpej 262: vsize_t len;
1.156 yamt 263: int npgs;
264:
265: KASSERT(pgs != NULL);
1.64 thorpej 266:
267: eva = round_page((vaddr_t) buf + size);
268: sva = trunc_page((vaddr_t) buf);
269: len = eva - sva;
270: npgs = len >> PAGE_SHIFT;
271:
272: pmap_kremove(sva, len);
273: pmap_update(pmap_kernel());
274: uvm_unloan(pgs, npgs, UVM_LOAN_TOPAGE);
1.80 yamt 275: sokvafree(sva, len);
1.64 thorpej 276: }
277:
278: static size_t
1.152 matt 279: sodopendfree(void)
1.64 thorpej 280: {
1.93 yamt 281: size_t rv;
1.64 thorpej 282:
1.160 ad 283: if (__predict_true(so_pendfree == NULL))
284: return 0;
285:
1.136 ad 286: mutex_enter(&so_pendfree_lock);
1.117 yamt 287: rv = sodopendfreel();
1.136 ad 288: mutex_exit(&so_pendfree_lock);
1.93 yamt 289:
290: return rv;
291: }
292:
293: /*
294: * sodopendfreel: free mbufs on "pendfree" list.
1.136 ad 295: * unlock and relock so_pendfree_lock when freeing mbufs.
1.93 yamt 296: *
1.136 ad 297: * => called with so_pendfree_lock held.
1.93 yamt 298: */
299:
300: static size_t
1.152 matt 301: sodopendfreel(void)
1.93 yamt 302: {
1.137 ad 303: struct mbuf *m, *next;
1.93 yamt 304: size_t rv = 0;
305:
1.136 ad 306: KASSERT(mutex_owned(&so_pendfree_lock));
1.64 thorpej 307:
1.137 ad 308: while (so_pendfree != NULL) {
1.64 thorpej 309: m = so_pendfree;
1.93 yamt 310: so_pendfree = NULL;
1.136 ad 311: mutex_exit(&so_pendfree_lock);
1.93 yamt 312:
313: for (; m != NULL; m = next) {
314: next = m->m_next;
1.156 yamt 315: KASSERT((~m->m_flags & (M_EXT|M_EXT_PAGES)) == 0);
316: KASSERT(m->m_ext.ext_refcnt == 0);
1.93 yamt 317:
318: rv += m->m_ext.ext_size;
1.156 yamt 319: sodoloanfree(m->m_ext.ext_pgs, m->m_ext.ext_buf,
1.93 yamt 320: m->m_ext.ext_size);
1.145 ad 321: pool_cache_put(mb_cache, m);
1.93 yamt 322: }
1.64 thorpej 323:
1.136 ad 324: mutex_enter(&so_pendfree_lock);
1.64 thorpej 325: }
326:
327: return (rv);
328: }
329:
1.80 yamt 330: void
1.134 christos 331: soloanfree(struct mbuf *m, void *buf, size_t size, void *arg)
1.64 thorpej 332: {
333:
1.156 yamt 334: KASSERT(m != NULL);
1.64 thorpej 335:
1.93 yamt 336: /*
337: * postpone freeing mbuf.
338: *
339: * we can't do it in interrupt context
340: * because we need to put kva back to kernel_map.
341: */
342:
1.136 ad 343: mutex_enter(&so_pendfree_lock);
1.92 yamt 344: m->m_next = so_pendfree;
345: so_pendfree = m;
1.136 ad 346: cv_broadcast(&socurkva_cv);
347: mutex_exit(&so_pendfree_lock);
1.64 thorpej 348: }
349:
350: static long
351: sosend_loan(struct socket *so, struct uio *uio, struct mbuf *m, long space)
352: {
353: struct iovec *iov = uio->uio_iov;
354: vaddr_t sva, eva;
355: vsize_t len;
1.156 yamt 356: vaddr_t lva;
357: int npgs, error;
358: vaddr_t va;
359: int i;
1.64 thorpej 360:
1.116 yamt 361: if (VMSPACE_IS_KERNEL_P(uio->uio_vmspace))
1.64 thorpej 362: return (0);
363:
364: if (iov->iov_len < (size_t) space)
365: space = iov->iov_len;
366: if (space > SOCK_LOAN_CHUNK)
367: space = SOCK_LOAN_CHUNK;
368:
369: eva = round_page((vaddr_t) iov->iov_base + space);
370: sva = trunc_page((vaddr_t) iov->iov_base);
371: len = eva - sva;
372: npgs = len >> PAGE_SHIFT;
373:
1.79 thorpej 374: KASSERT(npgs <= M_EXT_MAXPAGES);
375:
1.80 yamt 376: lva = sokvaalloc(len, so);
1.64 thorpej 377: if (lva == 0)
1.80 yamt 378: return 0;
1.64 thorpej 379:
1.116 yamt 380: error = uvm_loan(&uio->uio_vmspace->vm_map, sva, len,
1.79 thorpej 381: m->m_ext.ext_pgs, UVM_LOAN_TOPAGE);
1.64 thorpej 382: if (error) {
1.80 yamt 383: sokvafree(lva, len);
1.64 thorpej 384: return (0);
385: }
386:
387: for (i = 0, va = lva; i < npgs; i++, va += PAGE_SIZE)
1.79 thorpej 388: pmap_kenter_pa(va, VM_PAGE_TO_PHYS(m->m_ext.ext_pgs[i]),
1.160.2.4! yamt 389: VM_PROT_READ, 0);
1.64 thorpej 390: pmap_update(pmap_kernel());
391:
392: lva += (vaddr_t) iov->iov_base & PAGE_MASK;
393:
1.134 christos 394: MEXTADD(m, (void *) lva, space, M_MBUF, soloanfree, so);
1.79 thorpej 395: m->m_flags |= M_EXT_PAGES | M_EXT_ROMAP;
1.64 thorpej 396:
397: uio->uio_resid -= space;
398: /* uio_offset not updated, not set/used for write(2) */
1.134 christos 399: uio->uio_iov->iov_base = (char *)uio->uio_iov->iov_base + space;
1.64 thorpej 400: uio->uio_iov->iov_len -= space;
401: if (uio->uio_iov->iov_len == 0) {
402: uio->uio_iov++;
403: uio->uio_iovcnt--;
404: }
405:
406: return (space);
407: }
408:
1.119 yamt 409: static int
1.129 yamt 410: sokva_reclaim_callback(struct callback_entry *ce, void *obj, void *arg)
1.119 yamt 411: {
412:
413: KASSERT(ce == &sokva_reclaimerentry);
414: KASSERT(obj == NULL);
415:
416: sodopendfree();
417: if (!vm_map_starved_p(kernel_map)) {
418: return CALLBACK_CHAIN_ABORT;
419: }
420: return CALLBACK_CHAIN_CONTINUE;
421: }
422:
1.142 dyoung 423: struct mbuf *
1.147 dyoung 424: getsombuf(struct socket *so, int type)
1.142 dyoung 425: {
426: struct mbuf *m;
427:
1.147 dyoung 428: m = m_get(M_WAIT, type);
1.142 dyoung 429: MCLAIM(m, so->so_mowner);
430: return m;
431: }
432:
1.160.2.4! yamt 433: static int
! 434: socket_listener_cb(kauth_cred_t cred, kauth_action_t action, void *cookie,
! 435: void *arg0, void *arg1, void *arg2, void *arg3)
! 436: {
! 437: int result;
! 438: enum kauth_network_req req;
! 439:
! 440: result = KAUTH_RESULT_DEFER;
! 441: req = (enum kauth_network_req)arg0;
! 442:
! 443: if ((action != KAUTH_NETWORK_SOCKET) &&
! 444: (action != KAUTH_NETWORK_BIND))
! 445: return result;
! 446:
! 447: switch (req) {
! 448: case KAUTH_REQ_NETWORK_BIND_PORT:
! 449: result = KAUTH_RESULT_ALLOW;
! 450: break;
! 451:
! 452: case KAUTH_REQ_NETWORK_SOCKET_DROP: {
! 453: /* Normal users can only drop their own connections. */
! 454: struct socket *so = (struct socket *)arg1;
! 455:
! 456: if (proc_uidmatch(cred, so->so_cred))
! 457: result = KAUTH_RESULT_ALLOW;
! 458:
! 459: break;
! 460: }
! 461:
! 462: case KAUTH_REQ_NETWORK_SOCKET_OPEN:
! 463: /* We allow "raw" routing/bluetooth sockets to anyone. */
! 464: if ((u_long)arg1 == PF_ROUTE || (u_long)arg1 == PF_BLUETOOTH)
! 465: result = KAUTH_RESULT_ALLOW;
! 466: else {
! 467: /* Privileged, let secmodel handle this. */
! 468: if ((u_long)arg2 == SOCK_RAW)
! 469: break;
! 470: }
! 471:
! 472: result = KAUTH_RESULT_ALLOW;
! 473:
! 474: break;
! 475:
! 476: case KAUTH_REQ_NETWORK_SOCKET_CANSEE:
! 477: result = KAUTH_RESULT_ALLOW;
! 478:
! 479: break;
! 480:
! 481: default:
! 482: break;
! 483: }
! 484:
! 485: return result;
! 486: }
! 487:
1.119 yamt 488: void
489: soinit(void)
490: {
491:
1.160.2.2 yamt 492: sysctl_kern_somaxkva_setup();
493:
1.148 ad 494: mutex_init(&so_pendfree_lock, MUTEX_DEFAULT, IPL_VM);
1.160 ad 495: softnet_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE);
1.136 ad 496: cv_init(&socurkva_cv, "sokva");
1.160.2.2 yamt 497: soinit2();
1.136 ad 498:
1.119 yamt 499: /* Set the initial adjusted socket buffer size. */
500: if (sb_max_set(sb_max))
501: panic("bad initial sb_max value: %lu", sb_max);
502:
503: callback_register(&vm_map_to_kernel(kernel_map)->vmk_reclaim_callback,
504: &sokva_reclaimerentry, NULL, sokva_reclaim_callback);
1.160.2.4! yamt 505:
! 506: socket_listener = kauth_listen_scope(KAUTH_SCOPE_NETWORK,
! 507: socket_listener_cb, NULL);
1.119 yamt 508: }
509:
1.1 cgd 510: /*
511: * Socket operation routines.
512: * These routines are called by the routines in
513: * sys_socket.c or from a system process, and
514: * implement the semantics of socket operations by
515: * switching out to the protocol specific routines.
516: */
517: /*ARGSUSED*/
1.3 andrew 518: int
1.160 ad 519: socreate(int dom, struct socket **aso, int type, int proto, struct lwp *l,
520: struct socket *lockso)
1.1 cgd 521: {
1.99 matt 522: const struct protosw *prp;
1.54 lukem 523: struct socket *so;
1.115 yamt 524: uid_t uid;
1.160 ad 525: int error;
526: kmutex_t *lock;
1.1 cgd 527:
1.132 elad 528: error = kauth_authorize_network(l->l_cred, KAUTH_NETWORK_SOCKET,
529: KAUTH_REQ_NETWORK_SOCKET_OPEN, KAUTH_ARG(dom), KAUTH_ARG(type),
530: KAUTH_ARG(proto));
1.140 dyoung 531: if (error != 0)
532: return error;
1.127 elad 533:
1.1 cgd 534: if (proto)
535: prp = pffindproto(dom, proto, type);
536: else
537: prp = pffindtype(dom, type);
1.140 dyoung 538: if (prp == NULL) {
1.120 ginsbach 539: /* no support for domain */
540: if (pffinddomain(dom) == 0)
1.140 dyoung 541: return EAFNOSUPPORT;
1.120 ginsbach 542: /* no support for socket type */
543: if (proto == 0 && type != 0)
1.140 dyoung 544: return EPROTOTYPE;
545: return EPROTONOSUPPORT;
1.120 ginsbach 546: }
1.140 dyoung 547: if (prp->pr_usrreq == NULL)
548: return EPROTONOSUPPORT;
1.1 cgd 549: if (prp->pr_type != type)
1.140 dyoung 550: return EPROTOTYPE;
1.160 ad 551:
552: so = soget(true);
1.1 cgd 553: so->so_type = type;
554: so->so_proto = prp;
1.33 matt 555: so->so_send = sosend;
556: so->so_receive = soreceive;
1.78 matt 557: #ifdef MBUFTRACE
558: so->so_rcv.sb_mowner = &prp->pr_domain->dom_mowner;
559: so->so_snd.sb_mowner = &prp->pr_domain->dom_mowner;
560: so->so_mowner = &prp->pr_domain->dom_mowner;
561: #endif
1.138 rmind 562: uid = kauth_cred_geteuid(l->l_cred);
1.115 yamt 563: so->so_uidinfo = uid_find(uid);
1.160.2.2 yamt 564: so->so_cpid = l->l_proc->p_pid;
1.160 ad 565: if (lockso != NULL) {
566: /* Caller wants us to share a lock. */
567: lock = lockso->so_lock;
568: so->so_lock = lock;
569: mutex_obj_hold(lock);
570: mutex_enter(lock);
571: } else {
572: /* Lock assigned and taken during PRU_ATTACH. */
573: }
1.140 dyoung 574: error = (*prp->pr_usrreq)(so, PRU_ATTACH, NULL,
575: (struct mbuf *)(long)proto, NULL, l);
1.160 ad 576: KASSERT(solocked(so));
1.140 dyoung 577: if (error != 0) {
1.1 cgd 578: so->so_state |= SS_NOFDREF;
579: sofree(so);
1.140 dyoung 580: return error;
1.1 cgd 581: }
1.160.2.4! yamt 582: so->so_cred = kauth_cred_dup(l->l_cred);
1.160 ad 583: sounlock(so);
1.1 cgd 584: *aso = so;
1.140 dyoung 585: return 0;
1.1 cgd 586: }
587:
1.142 dyoung 588: /* On success, write file descriptor to fdout and return zero. On
589: * failure, return non-zero; *fdout will be undefined.
590: */
591: int
592: fsocreate(int domain, struct socket **sop, int type, int protocol,
593: struct lwp *l, int *fdout)
594: {
595: struct socket *so;
596: struct file *fp;
597: int fd, error;
598:
1.155 ad 599: if ((error = fd_allocfile(&fp, &fd)) != 0)
1.142 dyoung 600: return (error);
601: fp->f_flag = FREAD|FWRITE;
602: fp->f_type = DTYPE_SOCKET;
603: fp->f_ops = &socketops;
1.160 ad 604: error = socreate(domain, &so, type, protocol, l, NULL);
1.142 dyoung 605: if (error != 0) {
1.155 ad 606: fd_abort(curproc, fp, fd);
1.142 dyoung 607: } else {
608: if (sop != NULL)
609: *sop = so;
610: fp->f_data = so;
1.155 ad 611: fd_affix(curproc, fp, fd);
1.142 dyoung 612: *fdout = fd;
613: }
614: return error;
615: }
616:
1.3 andrew 617: int
1.160.2.3 yamt 618: sofamily(const struct socket *so)
619: {
620: const struct protosw *pr;
621: const struct domain *dom;
622:
623: if ((pr = so->so_proto) == NULL)
624: return AF_UNSPEC;
625: if ((dom = pr->pr_domain) == NULL)
626: return AF_UNSPEC;
627: return dom->dom_family;
628: }
629:
630: int
1.114 christos 631: sobind(struct socket *so, struct mbuf *nam, struct lwp *l)
1.1 cgd 632: {
1.160 ad 633: int error;
1.1 cgd 634:
1.160 ad 635: solock(so);
1.140 dyoung 636: error = (*so->so_proto->pr_usrreq)(so, PRU_BIND, NULL, nam, NULL, l);
1.160 ad 637: sounlock(so);
1.140 dyoung 638: return error;
1.1 cgd 639: }
640:
1.3 andrew 641: int
1.150 elad 642: solisten(struct socket *so, int backlog, struct lwp *l)
1.1 cgd 643: {
1.160 ad 644: int error;
1.1 cgd 645:
1.160 ad 646: solock(so);
1.158 ad 647: if ((so->so_state & (SS_ISCONNECTED | SS_ISCONNECTING |
1.160.2.1 yamt 648: SS_ISDISCONNECTING)) != 0) {
649: sounlock(so);
1.158 ad 650: return (EOPNOTSUPP);
1.160.2.1 yamt 651: }
1.140 dyoung 652: error = (*so->so_proto->pr_usrreq)(so, PRU_LISTEN, NULL,
1.150 elad 653: NULL, NULL, l);
1.140 dyoung 654: if (error != 0) {
1.160 ad 655: sounlock(so);
1.140 dyoung 656: return error;
1.1 cgd 657: }
1.63 matt 658: if (TAILQ_EMPTY(&so->so_q))
1.1 cgd 659: so->so_options |= SO_ACCEPTCONN;
660: if (backlog < 0)
661: backlog = 0;
1.49 jonathan 662: so->so_qlimit = min(backlog, somaxconn);
1.160 ad 663: sounlock(so);
1.140 dyoung 664: return 0;
1.1 cgd 665: }
666:
1.21 christos 667: void
1.54 lukem 668: sofree(struct socket *so)
1.1 cgd 669: {
1.160.2.1 yamt 670: u_int refs;
1.1 cgd 671:
1.160 ad 672: KASSERT(solocked(so));
673:
674: if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0) {
675: sounlock(so);
1.1 cgd 676: return;
1.160 ad 677: }
1.43 mycroft 678: if (so->so_head) {
679: /*
680: * We must not decommission a socket that's on the accept(2)
681: * queue. If we do, then accept(2) may hang after select(2)
682: * indicated that the listening socket was ready.
683: */
1.160 ad 684: if (!soqremque(so, 0)) {
685: sounlock(so);
1.43 mycroft 686: return;
1.160 ad 687: }
1.43 mycroft 688: }
1.98 christos 689: if (so->so_rcv.sb_hiwat)
1.110 christos 690: (void)chgsbsize(so->so_uidinfo, &so->so_rcv.sb_hiwat, 0,
1.98 christos 691: RLIM_INFINITY);
692: if (so->so_snd.sb_hiwat)
1.110 christos 693: (void)chgsbsize(so->so_uidinfo, &so->so_snd.sb_hiwat, 0,
1.98 christos 694: RLIM_INFINITY);
695: sbrelease(&so->so_snd, so);
1.160 ad 696: KASSERT(!cv_has_waiters(&so->so_cv));
697: KASSERT(!cv_has_waiters(&so->so_rcv.sb_cv));
698: KASSERT(!cv_has_waiters(&so->so_snd.sb_cv));
1.1 cgd 699: sorflush(so);
1.160.2.1 yamt 700: refs = so->so_aborting; /* XXX */
1.160.2.2 yamt 701: /* Remove acccept filter if one is present. */
702: if (so->so_accf != NULL)
703: (void)accept_filt_clear(so);
1.160 ad 704: sounlock(so);
1.160.2.1 yamt 705: if (refs == 0) /* XXX */
706: soput(so);
1.1 cgd 707: }
708:
709: /*
710: * Close a socket on last file table reference removal.
711: * Initiate disconnect if connected.
712: * Free socket when disconnect complete.
713: */
1.3 andrew 714: int
1.54 lukem 715: soclose(struct socket *so)
1.1 cgd 716: {
1.54 lukem 717: struct socket *so2;
1.160 ad 718: int error;
719: int error2;
1.1 cgd 720:
1.54 lukem 721: error = 0;
1.160 ad 722: solock(so);
1.1 cgd 723: if (so->so_options & SO_ACCEPTCONN) {
1.160.2.2 yamt 724: for (;;) {
1.160 ad 725: if ((so2 = TAILQ_FIRST(&so->so_q0)) != 0) {
726: KASSERT(solocked2(so, so2));
727: (void) soqremque(so2, 0);
728: /* soabort drops the lock. */
729: (void) soabort(so2);
730: solock(so);
731: continue;
732: }
733: if ((so2 = TAILQ_FIRST(&so->so_q)) != 0) {
734: KASSERT(solocked2(so, so2));
735: (void) soqremque(so2, 1);
736: /* soabort drops the lock. */
737: (void) soabort(so2);
738: solock(so);
739: continue;
740: }
1.160.2.2 yamt 741: break;
742: }
1.1 cgd 743: }
744: if (so->so_pcb == 0)
745: goto discard;
746: if (so->so_state & SS_ISCONNECTED) {
747: if ((so->so_state & SS_ISDISCONNECTING) == 0) {
748: error = sodisconnect(so);
749: if (error)
750: goto drop;
751: }
752: if (so->so_options & SO_LINGER) {
1.151 ad 753: if ((so->so_state & SS_ISDISCONNECTING) && so->so_nbio)
1.1 cgd 754: goto drop;
1.21 christos 755: while (so->so_state & SS_ISCONNECTED) {
1.160.2.2 yamt 756: error = sowait(so, true, so->so_linger * hz);
1.21 christos 757: if (error)
1.1 cgd 758: break;
1.21 christos 759: }
1.1 cgd 760: }
761: }
1.54 lukem 762: drop:
1.1 cgd 763: if (so->so_pcb) {
1.160 ad 764: error2 = (*so->so_proto->pr_usrreq)(so, PRU_DETACH,
1.140 dyoung 765: NULL, NULL, NULL, NULL);
1.1 cgd 766: if (error == 0)
767: error = error2;
768: }
1.54 lukem 769: discard:
1.1 cgd 770: if (so->so_state & SS_NOFDREF)
771: panic("soclose: NOFDREF");
1.160.2.4! yamt 772: kauth_cred_free(so->so_cred);
1.1 cgd 773: so->so_state |= SS_NOFDREF;
774: sofree(so);
775: return (error);
776: }
777:
778: /*
1.160 ad 779: * Must be called with the socket locked.. Will return with it unlocked.
1.1 cgd 780: */
1.3 andrew 781: int
1.54 lukem 782: soabort(struct socket *so)
1.1 cgd 783: {
1.160.2.1 yamt 784: u_int refs;
1.139 yamt 785: int error;
1.160 ad 786:
787: KASSERT(solocked(so));
788: KASSERT(so->so_head == NULL);
1.1 cgd 789:
1.160.2.1 yamt 790: so->so_aborting++; /* XXX */
1.140 dyoung 791: error = (*so->so_proto->pr_usrreq)(so, PRU_ABORT, NULL,
792: NULL, NULL, NULL);
1.160.2.1 yamt 793: refs = --so->so_aborting; /* XXX */
794: if (error || (refs == 0)) {
1.139 yamt 795: sofree(so);
1.160 ad 796: } else {
797: sounlock(so);
1.139 yamt 798: }
799: return error;
1.1 cgd 800: }
801:
1.3 andrew 802: int
1.54 lukem 803: soaccept(struct socket *so, struct mbuf *nam)
1.1 cgd 804: {
1.160 ad 805: int error;
806:
807: KASSERT(solocked(so));
1.1 cgd 808:
1.54 lukem 809: error = 0;
1.1 cgd 810: if ((so->so_state & SS_NOFDREF) == 0)
811: panic("soaccept: !NOFDREF");
812: so->so_state &= ~SS_NOFDREF;
1.55 thorpej 813: if ((so->so_state & SS_ISDISCONNECTED) == 0 ||
814: (so->so_proto->pr_flags & PR_ABRTACPTDIS) == 0)
1.41 mycroft 815: error = (*so->so_proto->pr_usrreq)(so, PRU_ACCEPT,
1.140 dyoung 816: NULL, nam, NULL, NULL);
1.41 mycroft 817: else
1.53 itojun 818: error = ECONNABORTED;
1.52 itojun 819:
1.1 cgd 820: return (error);
821: }
822:
1.3 andrew 823: int
1.114 christos 824: soconnect(struct socket *so, struct mbuf *nam, struct lwp *l)
1.1 cgd 825: {
1.160 ad 826: int error;
827:
828: KASSERT(solocked(so));
1.1 cgd 829:
830: if (so->so_options & SO_ACCEPTCONN)
831: return (EOPNOTSUPP);
832: /*
833: * If protocol is connection-based, can only connect once.
834: * Otherwise, if connected, try to disconnect first.
835: * This allows user to disconnect by connecting to, e.g.,
836: * a null address.
837: */
838: if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) &&
839: ((so->so_proto->pr_flags & PR_CONNREQUIRED) ||
840: (error = sodisconnect(so))))
841: error = EISCONN;
842: else
843: error = (*so->so_proto->pr_usrreq)(so, PRU_CONNECT,
1.140 dyoung 844: NULL, nam, NULL, l);
1.1 cgd 845: return (error);
846: }
847:
1.3 andrew 848: int
1.54 lukem 849: soconnect2(struct socket *so1, struct socket *so2)
1.1 cgd 850: {
1.160 ad 851: int error;
852:
853: KASSERT(solocked2(so1, so2));
1.1 cgd 854:
1.22 mycroft 855: error = (*so1->so_proto->pr_usrreq)(so1, PRU_CONNECT2,
1.140 dyoung 856: NULL, (struct mbuf *)so2, NULL, NULL);
1.1 cgd 857: return (error);
858: }
859:
1.3 andrew 860: int
1.54 lukem 861: sodisconnect(struct socket *so)
1.1 cgd 862: {
1.160 ad 863: int error;
864:
865: KASSERT(solocked(so));
1.1 cgd 866:
867: if ((so->so_state & SS_ISCONNECTED) == 0) {
868: error = ENOTCONN;
1.160 ad 869: } else if (so->so_state & SS_ISDISCONNECTING) {
1.1 cgd 870: error = EALREADY;
1.160 ad 871: } else {
872: error = (*so->so_proto->pr_usrreq)(so, PRU_DISCONNECT,
873: NULL, NULL, NULL, NULL);
1.1 cgd 874: }
1.117 yamt 875: sodopendfree();
1.1 cgd 876: return (error);
877: }
878:
1.15 mycroft 879: #define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK)
1.1 cgd 880: /*
881: * Send on a socket.
882: * If send must go all at once and message is larger than
883: * send buffering, then hard error.
884: * Lock against other senders.
885: * If must go all at once and not enough room now, then
886: * inform user that this would block and do nothing.
887: * Otherwise, if nonblocking, send as much as possible.
888: * The data to be sent is described by "uio" if nonzero,
889: * otherwise by the mbuf chain "top" (which must be null
890: * if uio is not). Data provided in mbuf chain must be small
891: * enough to send all at once.
892: *
893: * Returns nonzero on error, timeout or signal; callers
894: * must check for short counts if EINTR/ERESTART are returned.
895: * Data and control buffers are freed on return.
896: */
1.3 andrew 897: int
1.54 lukem 898: sosend(struct socket *so, struct mbuf *addr, struct uio *uio, struct mbuf *top,
1.114 christos 899: struct mbuf *control, int flags, struct lwp *l)
1.1 cgd 900: {
1.54 lukem 901: struct mbuf **mp, *m;
1.114 christos 902: struct proc *p;
1.58 jdolecek 903: long space, len, resid, clen, mlen;
904: int error, s, dontroute, atomic;
1.160.2.4! yamt 905: short wakeup_state = 0;
1.54 lukem 906:
1.114 christos 907: p = l->l_proc;
1.117 yamt 908: sodopendfree();
1.160 ad 909: clen = 0;
1.64 thorpej 910:
1.160 ad 911: /*
912: * solock() provides atomicity of access. splsoftnet() prevents
913: * protocol processing soft interrupts from interrupting us and
914: * blocking (expensive).
915: */
916: s = splsoftnet();
917: solock(so);
1.54 lukem 918: atomic = sosendallatonce(so) || top;
1.1 cgd 919: if (uio)
920: resid = uio->uio_resid;
921: else
922: resid = top->m_pkthdr.len;
1.7 cgd 923: /*
924: * In theory resid should be unsigned.
925: * However, space must be signed, as it might be less than 0
926: * if we over-committed, and we must use a signed comparison
927: * of space and resid. On the other hand, a negative resid
928: * causes us to loop sending 0-length segments to the protocol.
929: */
1.29 mycroft 930: if (resid < 0) {
931: error = EINVAL;
932: goto out;
933: }
1.1 cgd 934: dontroute =
935: (flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 &&
936: (so->so_proto->pr_flags & PR_ATOMIC);
1.160.2.2 yamt 937: l->l_ru.ru_msgsnd++;
1.1 cgd 938: if (control)
939: clen = control->m_len;
1.54 lukem 940: restart:
1.21 christos 941: if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0)
1.1 cgd 942: goto out;
943: do {
1.160 ad 944: if (so->so_state & SS_CANTSENDMORE) {
945: error = EPIPE;
946: goto release;
947: }
1.48 thorpej 948: if (so->so_error) {
949: error = so->so_error;
950: so->so_error = 0;
951: goto release;
952: }
1.1 cgd 953: if ((so->so_state & SS_ISCONNECTED) == 0) {
954: if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
955: if ((so->so_state & SS_ISCONFIRMING) == 0 &&
1.160 ad 956: !(resid == 0 && clen != 0)) {
957: error = ENOTCONN;
958: goto release;
959: }
960: } else if (addr == 0) {
961: error = EDESTADDRREQ;
962: goto release;
963: }
1.1 cgd 964: }
965: space = sbspace(&so->so_snd);
966: if (flags & MSG_OOB)
967: space += 1024;
1.21 christos 968: if ((atomic && resid > so->so_snd.sb_hiwat) ||
1.160 ad 969: clen > so->so_snd.sb_hiwat) {
970: error = EMSGSIZE;
971: goto release;
972: }
1.96 mycroft 973: if (space < resid + clen &&
1.1 cgd 974: (atomic || space < so->so_snd.sb_lowat || space < clen)) {
1.160 ad 975: if (so->so_nbio) {
976: error = EWOULDBLOCK;
977: goto release;
978: }
1.1 cgd 979: sbunlock(&so->so_snd);
1.160.2.4! yamt 980: if (wakeup_state & SS_RESTARTSYS) {
! 981: error = ERESTART;
! 982: goto out;
! 983: }
1.1 cgd 984: error = sbwait(&so->so_snd);
985: if (error)
986: goto out;
1.160.2.4! yamt 987: wakeup_state = so->so_state;
1.1 cgd 988: goto restart;
989: }
1.160.2.4! yamt 990: wakeup_state = 0;
1.1 cgd 991: mp = ⊤
992: space -= clen;
993: do {
1.45 tv 994: if (uio == NULL) {
995: /*
996: * Data is prepackaged in "top".
997: */
998: resid = 0;
999: if (flags & MSG_EOR)
1000: top->m_flags |= M_EOR;
1001: } else do {
1.160 ad 1002: sounlock(so);
1003: splx(s);
1.144 dyoung 1004: if (top == NULL) {
1.78 matt 1005: m = m_gethdr(M_WAIT, MT_DATA);
1.45 tv 1006: mlen = MHLEN;
1007: m->m_pkthdr.len = 0;
1.140 dyoung 1008: m->m_pkthdr.rcvif = NULL;
1.45 tv 1009: } else {
1.78 matt 1010: m = m_get(M_WAIT, MT_DATA);
1.45 tv 1011: mlen = MLEN;
1012: }
1.78 matt 1013: MCLAIM(m, so->so_snd.sb_mowner);
1.121 yamt 1014: if (sock_loan_thresh >= 0 &&
1015: uio->uio_iov->iov_len >= sock_loan_thresh &&
1016: space >= sock_loan_thresh &&
1.64 thorpej 1017: (len = sosend_loan(so, uio, m,
1018: space)) != 0) {
1019: SOSEND_COUNTER_INCR(&sosend_loan_big);
1020: space -= len;
1021: goto have_data;
1022: }
1.45 tv 1023: if (resid >= MINCLSIZE && space >= MCLBYTES) {
1.64 thorpej 1024: SOSEND_COUNTER_INCR(&sosend_copy_big);
1.78 matt 1025: m_clget(m, M_WAIT);
1.45 tv 1026: if ((m->m_flags & M_EXT) == 0)
1027: goto nopages;
1028: mlen = MCLBYTES;
1029: if (atomic && top == 0) {
1.58 jdolecek 1030: len = lmin(MCLBYTES - max_hdr,
1.54 lukem 1031: resid);
1.45 tv 1032: m->m_data += max_hdr;
1033: } else
1.58 jdolecek 1034: len = lmin(MCLBYTES, resid);
1.45 tv 1035: space -= len;
1036: } else {
1.64 thorpej 1037: nopages:
1038: SOSEND_COUNTER_INCR(&sosend_copy_small);
1.58 jdolecek 1039: len = lmin(lmin(mlen, resid), space);
1.45 tv 1040: space -= len;
1041: /*
1042: * For datagram protocols, leave room
1043: * for protocol headers in first mbuf.
1044: */
1045: if (atomic && top == 0 && len < mlen)
1046: MH_ALIGN(m, len);
1047: }
1.144 dyoung 1048: error = uiomove(mtod(m, void *), (int)len, uio);
1.64 thorpej 1049: have_data:
1.45 tv 1050: resid = uio->uio_resid;
1051: m->m_len = len;
1052: *mp = m;
1053: top->m_pkthdr.len += len;
1.160 ad 1054: s = splsoftnet();
1055: solock(so);
1.144 dyoung 1056: if (error != 0)
1.45 tv 1057: goto release;
1058: mp = &m->m_next;
1059: if (resid <= 0) {
1060: if (flags & MSG_EOR)
1061: top->m_flags |= M_EOR;
1062: break;
1063: }
1064: } while (space > 0 && atomic);
1.108 perry 1065:
1.160 ad 1066: if (so->so_state & SS_CANTSENDMORE) {
1067: error = EPIPE;
1068: goto release;
1069: }
1.45 tv 1070: if (dontroute)
1071: so->so_options |= SO_DONTROUTE;
1072: if (resid > 0)
1073: so->so_state |= SS_MORETOCOME;
1.46 sommerfe 1074: error = (*so->so_proto->pr_usrreq)(so,
1075: (flags & MSG_OOB) ? PRU_SENDOOB : PRU_SEND,
1.160 ad 1076: top, addr, control, curlwp);
1.45 tv 1077: if (dontroute)
1078: so->so_options &= ~SO_DONTROUTE;
1079: if (resid > 0)
1080: so->so_state &= ~SS_MORETOCOME;
1081: clen = 0;
1.144 dyoung 1082: control = NULL;
1083: top = NULL;
1.45 tv 1084: mp = ⊤
1.144 dyoung 1085: if (error != 0)
1.1 cgd 1086: goto release;
1087: } while (resid && space > 0);
1088: } while (resid);
1089:
1.54 lukem 1090: release:
1.1 cgd 1091: sbunlock(&so->so_snd);
1.54 lukem 1092: out:
1.160 ad 1093: sounlock(so);
1094: splx(s);
1.1 cgd 1095: if (top)
1096: m_freem(top);
1097: if (control)
1098: m_freem(control);
1099: return (error);
1100: }
1101:
1102: /*
1.159 ad 1103: * Following replacement or removal of the first mbuf on the first
1104: * mbuf chain of a socket buffer, push necessary state changes back
1105: * into the socket buffer so that other consumers see the values
1106: * consistently. 'nextrecord' is the callers locally stored value of
1107: * the original value of sb->sb_mb->m_nextpkt which must be restored
1108: * when the lead mbuf changes. NOTE: 'nextrecord' may be NULL.
1109: */
1110: static void
1111: sbsync(struct sockbuf *sb, struct mbuf *nextrecord)
1112: {
1113:
1.160 ad 1114: KASSERT(solocked(sb->sb_so));
1115:
1.159 ad 1116: /*
1117: * First, update for the new value of nextrecord. If necessary,
1118: * make it the first record.
1119: */
1120: if (sb->sb_mb != NULL)
1121: sb->sb_mb->m_nextpkt = nextrecord;
1122: else
1123: sb->sb_mb = nextrecord;
1124:
1125: /*
1126: * Now update any dependent socket buffer fields to reflect
1127: * the new state. This is an inline of SB_EMPTY_FIXUP, with
1128: * the addition of a second clause that takes care of the
1129: * case where sb_mb has been updated, but remains the last
1130: * record.
1131: */
1132: if (sb->sb_mb == NULL) {
1133: sb->sb_mbtail = NULL;
1134: sb->sb_lastrecord = NULL;
1135: } else if (sb->sb_mb->m_nextpkt == NULL)
1136: sb->sb_lastrecord = sb->sb_mb;
1137: }
1138:
1139: /*
1.1 cgd 1140: * Implement receive operations on a socket.
1141: * We depend on the way that records are added to the sockbuf
1142: * by sbappend*. In particular, each record (mbufs linked through m_next)
1143: * must begin with an address if the protocol so specifies,
1144: * followed by an optional mbuf or mbufs containing ancillary data,
1145: * and then zero or more mbufs of data.
1146: * In order to avoid blocking network interrupts for the entire time here,
1147: * we splx() while doing the actual copy to user space.
1148: * Although the sockbuf is locked, new data may still be appended,
1149: * and thus we must maintain consistency of the sockbuf during that time.
1150: *
1151: * The caller may receive the data as a single mbuf chain by supplying
1152: * an mbuf **mp0 for use in returning the chain. The uio is then used
1153: * only for the count in uio_resid.
1154: */
1.3 andrew 1155: int
1.54 lukem 1156: soreceive(struct socket *so, struct mbuf **paddr, struct uio *uio,
1157: struct mbuf **mp0, struct mbuf **controlp, int *flagsp)
1.1 cgd 1158: {
1.116 yamt 1159: struct lwp *l = curlwp;
1.160 ad 1160: struct mbuf *m, **mp, *mt;
1.146 dyoung 1161: int atomic, flags, len, error, s, offset, moff, type, orig_resid;
1.99 matt 1162: const struct protosw *pr;
1.54 lukem 1163: struct mbuf *nextrecord;
1.67 he 1164: int mbuf_removed = 0;
1.146 dyoung 1165: const struct domain *dom;
1.160.2.4! yamt 1166: short wakeup_state = 0;
1.64 thorpej 1167:
1.54 lukem 1168: pr = so->so_proto;
1.146 dyoung 1169: atomic = pr->pr_flags & PR_ATOMIC;
1170: dom = pr->pr_domain;
1.1 cgd 1171: mp = mp0;
1.54 lukem 1172: type = 0;
1173: orig_resid = uio->uio_resid;
1.102 jonathan 1174:
1.144 dyoung 1175: if (paddr != NULL)
1176: *paddr = NULL;
1177: if (controlp != NULL)
1178: *controlp = NULL;
1179: if (flagsp != NULL)
1.1 cgd 1180: flags = *flagsp &~ MSG_EOR;
1181: else
1182: flags = 0;
1.66 enami 1183:
1184: if ((flags & MSG_DONTWAIT) == 0)
1.117 yamt 1185: sodopendfree();
1.66 enami 1186:
1.1 cgd 1187: if (flags & MSG_OOB) {
1188: m = m_get(M_WAIT, MT_DATA);
1.160 ad 1189: solock(so);
1.17 cgd 1190: error = (*pr->pr_usrreq)(so, PRU_RCVOOB, m,
1.140 dyoung 1191: (struct mbuf *)(long)(flags & MSG_PEEK), NULL, l);
1.160 ad 1192: sounlock(so);
1.1 cgd 1193: if (error)
1194: goto bad;
1195: do {
1.134 christos 1196: error = uiomove(mtod(m, void *),
1.1 cgd 1197: (int) min(uio->uio_resid, m->m_len), uio);
1198: m = m_free(m);
1.144 dyoung 1199: } while (uio->uio_resid > 0 && error == 0 && m);
1.54 lukem 1200: bad:
1.144 dyoung 1201: if (m != NULL)
1.1 cgd 1202: m_freem(m);
1.144 dyoung 1203: return error;
1.1 cgd 1204: }
1.144 dyoung 1205: if (mp != NULL)
1.140 dyoung 1206: *mp = NULL;
1.160 ad 1207:
1208: /*
1209: * solock() provides atomicity of access. splsoftnet() prevents
1210: * protocol processing soft interrupts from interrupting us and
1211: * blocking (expensive).
1212: */
1213: s = splsoftnet();
1214: solock(so);
1.1 cgd 1215: if (so->so_state & SS_ISCONFIRMING && uio->uio_resid)
1.140 dyoung 1216: (*pr->pr_usrreq)(so, PRU_RCVD, NULL, NULL, NULL, l);
1.1 cgd 1217:
1.54 lukem 1218: restart:
1.160 ad 1219: if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0) {
1220: sounlock(so);
1221: splx(s);
1.144 dyoung 1222: return error;
1.160 ad 1223: }
1.1 cgd 1224:
1225: m = so->so_rcv.sb_mb;
1226: /*
1227: * If we have less data than requested, block awaiting more
1228: * (subject to any timeout) if:
1.15 mycroft 1229: * 1. the current count is less than the low water mark,
1.1 cgd 1230: * 2. MSG_WAITALL is set, and it is possible to do the entire
1.15 mycroft 1231: * receive operation at once if we block (resid <= hiwat), or
1232: * 3. MSG_DONTWAIT is not set.
1.1 cgd 1233: * If MSG_WAITALL is set but resid is larger than the receive buffer,
1234: * we have to do the receive in sections, and thus risk returning
1235: * a short count if a timeout or signal occurs after we start.
1236: */
1.144 dyoung 1237: if (m == NULL ||
1238: ((flags & MSG_DONTWAIT) == 0 &&
1239: so->so_rcv.sb_cc < uio->uio_resid &&
1240: (so->so_rcv.sb_cc < so->so_rcv.sb_lowat ||
1241: ((flags & MSG_WAITALL) &&
1242: uio->uio_resid <= so->so_rcv.sb_hiwat)) &&
1.146 dyoung 1243: m->m_nextpkt == NULL && !atomic)) {
1.1 cgd 1244: #ifdef DIAGNOSTIC
1.144 dyoung 1245: if (m == NULL && so->so_rcv.sb_cc)
1.1 cgd 1246: panic("receive 1");
1247: #endif
1248: if (so->so_error) {
1.144 dyoung 1249: if (m != NULL)
1.15 mycroft 1250: goto dontblock;
1.1 cgd 1251: error = so->so_error;
1252: if ((flags & MSG_PEEK) == 0)
1253: so->so_error = 0;
1254: goto release;
1255: }
1256: if (so->so_state & SS_CANTRCVMORE) {
1.144 dyoung 1257: if (m != NULL)
1.15 mycroft 1258: goto dontblock;
1.1 cgd 1259: else
1260: goto release;
1261: }
1.144 dyoung 1262: for (; m != NULL; m = m->m_next)
1.1 cgd 1263: if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) {
1264: m = so->so_rcv.sb_mb;
1265: goto dontblock;
1266: }
1267: if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
1268: (so->so_proto->pr_flags & PR_CONNREQUIRED)) {
1269: error = ENOTCONN;
1270: goto release;
1271: }
1272: if (uio->uio_resid == 0)
1273: goto release;
1.151 ad 1274: if (so->so_nbio || (flags & MSG_DONTWAIT)) {
1.1 cgd 1275: error = EWOULDBLOCK;
1276: goto release;
1277: }
1.69 thorpej 1278: SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 1");
1279: SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 1");
1.1 cgd 1280: sbunlock(&so->so_rcv);
1.160.2.4! yamt 1281: if (wakeup_state & SS_RESTARTSYS)
! 1282: error = ERESTART;
! 1283: else
! 1284: error = sbwait(&so->so_rcv);
1.160 ad 1285: if (error != 0) {
1286: sounlock(so);
1287: splx(s);
1.144 dyoung 1288: return error;
1.160 ad 1289: }
1.160.2.4! yamt 1290: wakeup_state = so->so_state;
1.1 cgd 1291: goto restart;
1292: }
1.54 lukem 1293: dontblock:
1.69 thorpej 1294: /*
1295: * On entry here, m points to the first record of the socket buffer.
1.159 ad 1296: * From this point onward, we maintain 'nextrecord' as a cache of the
1297: * pointer to the next record in the socket buffer. We must keep the
1298: * various socket buffer pointers and local stack versions of the
1299: * pointers in sync, pushing out modifications before dropping the
1.160 ad 1300: * socket lock, and re-reading them when picking it up.
1.159 ad 1301: *
1302: * Otherwise, we will race with the network stack appending new data
1303: * or records onto the socket buffer by using inconsistent/stale
1304: * versions of the field, possibly resulting in socket buffer
1305: * corruption.
1306: *
1307: * By holding the high-level sblock(), we prevent simultaneous
1308: * readers from pulling off the front of the socket buffer.
1.69 thorpej 1309: */
1.144 dyoung 1310: if (l != NULL)
1.157 ad 1311: l->l_ru.ru_msgrcv++;
1.69 thorpej 1312: KASSERT(m == so->so_rcv.sb_mb);
1313: SBLASTRECORDCHK(&so->so_rcv, "soreceive 1");
1314: SBLASTMBUFCHK(&so->so_rcv, "soreceive 1");
1.1 cgd 1315: nextrecord = m->m_nextpkt;
1316: if (pr->pr_flags & PR_ADDR) {
1317: #ifdef DIAGNOSTIC
1318: if (m->m_type != MT_SONAME)
1319: panic("receive 1a");
1320: #endif
1.3 andrew 1321: orig_resid = 0;
1.1 cgd 1322: if (flags & MSG_PEEK) {
1323: if (paddr)
1324: *paddr = m_copy(m, 0, m->m_len);
1325: m = m->m_next;
1326: } else {
1327: sbfree(&so->so_rcv, m);
1.67 he 1328: mbuf_removed = 1;
1.144 dyoung 1329: if (paddr != NULL) {
1.1 cgd 1330: *paddr = m;
1331: so->so_rcv.sb_mb = m->m_next;
1.144 dyoung 1332: m->m_next = NULL;
1.1 cgd 1333: m = so->so_rcv.sb_mb;
1334: } else {
1335: MFREE(m, so->so_rcv.sb_mb);
1336: m = so->so_rcv.sb_mb;
1337: }
1.159 ad 1338: sbsync(&so->so_rcv, nextrecord);
1.1 cgd 1339: }
1340: }
1.159 ad 1341:
1342: /*
1343: * Process one or more MT_CONTROL mbufs present before any data mbufs
1344: * in the first mbuf chain on the socket buffer. If MSG_PEEK, we
1345: * just copy the data; if !MSG_PEEK, we call into the protocol to
1346: * perform externalization (or freeing if controlp == NULL).
1347: */
1348: if (__predict_false(m != NULL && m->m_type == MT_CONTROL)) {
1349: struct mbuf *cm = NULL, *cmn;
1350: struct mbuf **cme = &cm;
1351:
1352: do {
1353: if (flags & MSG_PEEK) {
1354: if (controlp != NULL) {
1355: *controlp = m_copy(m, 0, m->m_len);
1356: controlp = &(*controlp)->m_next;
1357: }
1358: m = m->m_next;
1359: } else {
1360: sbfree(&so->so_rcv, m);
1.1 cgd 1361: so->so_rcv.sb_mb = m->m_next;
1.144 dyoung 1362: m->m_next = NULL;
1.159 ad 1363: *cme = m;
1364: cme = &(*cme)->m_next;
1.1 cgd 1365: m = so->so_rcv.sb_mb;
1.159 ad 1366: }
1367: } while (m != NULL && m->m_type == MT_CONTROL);
1368: if ((flags & MSG_PEEK) == 0)
1369: sbsync(&so->so_rcv, nextrecord);
1370: for (; cm != NULL; cm = cmn) {
1371: cmn = cm->m_next;
1372: cm->m_next = NULL;
1373: type = mtod(cm, struct cmsghdr *)->cmsg_type;
1374: if (controlp != NULL) {
1375: if (dom->dom_externalize != NULL &&
1376: type == SCM_RIGHTS) {
1.160 ad 1377: sounlock(so);
1.159 ad 1378: splx(s);
1379: error = (*dom->dom_externalize)(cm, l);
1380: s = splsoftnet();
1.160 ad 1381: solock(so);
1.159 ad 1382: }
1383: *controlp = cm;
1384: while (*controlp != NULL)
1385: controlp = &(*controlp)->m_next;
1.1 cgd 1386: } else {
1.106 itojun 1387: /*
1388: * Dispose of any SCM_RIGHTS message that went
1389: * through the read path rather than recv.
1390: */
1.159 ad 1391: if (dom->dom_dispose != NULL &&
1392: type == SCM_RIGHTS) {
1.160 ad 1393: sounlock(so);
1.159 ad 1394: (*dom->dom_dispose)(cm);
1.160 ad 1395: solock(so);
1.159 ad 1396: }
1397: m_freem(cm);
1.1 cgd 1398: }
1399: }
1.159 ad 1400: if (m != NULL)
1401: nextrecord = so->so_rcv.sb_mb->m_nextpkt;
1402: else
1403: nextrecord = so->so_rcv.sb_mb;
1404: orig_resid = 0;
1.1 cgd 1405: }
1.69 thorpej 1406:
1.159 ad 1407: /* If m is non-NULL, we have some data to read. */
1408: if (__predict_true(m != NULL)) {
1.1 cgd 1409: type = m->m_type;
1410: if (type == MT_OOBDATA)
1411: flags |= MSG_OOB;
1412: }
1.69 thorpej 1413: SBLASTRECORDCHK(&so->so_rcv, "soreceive 2");
1414: SBLASTMBUFCHK(&so->so_rcv, "soreceive 2");
1415:
1.1 cgd 1416: moff = 0;
1417: offset = 0;
1.144 dyoung 1418: while (m != NULL && uio->uio_resid > 0 && error == 0) {
1.1 cgd 1419: if (m->m_type == MT_OOBDATA) {
1420: if (type != MT_OOBDATA)
1421: break;
1422: } else if (type == MT_OOBDATA)
1423: break;
1424: #ifdef DIAGNOSTIC
1425: else if (m->m_type != MT_DATA && m->m_type != MT_HEADER)
1426: panic("receive 3");
1427: #endif
1428: so->so_state &= ~SS_RCVATMARK;
1.160.2.4! yamt 1429: wakeup_state = 0;
1.1 cgd 1430: len = uio->uio_resid;
1431: if (so->so_oobmark && len > so->so_oobmark - offset)
1432: len = so->so_oobmark - offset;
1433: if (len > m->m_len - moff)
1434: len = m->m_len - moff;
1435: /*
1436: * If mp is set, just pass back the mbufs.
1437: * Otherwise copy them out via the uio, then free.
1438: * Sockbuf must be consistent here (points to current mbuf,
1439: * it points to next record) when we drop priority;
1440: * we must note any additions to the sockbuf when we
1441: * block interrupts again.
1442: */
1.144 dyoung 1443: if (mp == NULL) {
1.69 thorpej 1444: SBLASTRECORDCHK(&so->so_rcv, "soreceive uiomove");
1445: SBLASTMBUFCHK(&so->so_rcv, "soreceive uiomove");
1.160 ad 1446: sounlock(so);
1.1 cgd 1447: splx(s);
1.134 christos 1448: error = uiomove(mtod(m, char *) + moff, (int)len, uio);
1.20 mycroft 1449: s = splsoftnet();
1.160 ad 1450: solock(so);
1.144 dyoung 1451: if (error != 0) {
1.67 he 1452: /*
1453: * If any part of the record has been removed
1454: * (such as the MT_SONAME mbuf, which will
1455: * happen when PR_ADDR, and thus also
1456: * PR_ATOMIC, is set), then drop the entire
1457: * record to maintain the atomicity of the
1458: * receive operation.
1459: *
1460: * This avoids a later panic("receive 1a")
1461: * when compiled with DIAGNOSTIC.
1462: */
1.146 dyoung 1463: if (m && mbuf_removed && atomic)
1.67 he 1464: (void) sbdroprecord(&so->so_rcv);
1465:
1.57 jdolecek 1466: goto release;
1.67 he 1467: }
1.1 cgd 1468: } else
1469: uio->uio_resid -= len;
1470: if (len == m->m_len - moff) {
1471: if (m->m_flags & M_EOR)
1472: flags |= MSG_EOR;
1473: if (flags & MSG_PEEK) {
1474: m = m->m_next;
1475: moff = 0;
1476: } else {
1477: nextrecord = m->m_nextpkt;
1478: sbfree(&so->so_rcv, m);
1479: if (mp) {
1480: *mp = m;
1481: mp = &m->m_next;
1482: so->so_rcv.sb_mb = m = m->m_next;
1.140 dyoung 1483: *mp = NULL;
1.1 cgd 1484: } else {
1485: MFREE(m, so->so_rcv.sb_mb);
1486: m = so->so_rcv.sb_mb;
1487: }
1.69 thorpej 1488: /*
1489: * If m != NULL, we also know that
1490: * so->so_rcv.sb_mb != NULL.
1491: */
1492: KASSERT(so->so_rcv.sb_mb == m);
1493: if (m) {
1.1 cgd 1494: m->m_nextpkt = nextrecord;
1.69 thorpej 1495: if (nextrecord == NULL)
1496: so->so_rcv.sb_lastrecord = m;
1497: } else {
1498: so->so_rcv.sb_mb = nextrecord;
1.70 thorpej 1499: SB_EMPTY_FIXUP(&so->so_rcv);
1.69 thorpej 1500: }
1501: SBLASTRECORDCHK(&so->so_rcv, "soreceive 3");
1502: SBLASTMBUFCHK(&so->so_rcv, "soreceive 3");
1.1 cgd 1503: }
1.144 dyoung 1504: } else if (flags & MSG_PEEK)
1505: moff += len;
1506: else {
1.160 ad 1507: if (mp != NULL) {
1508: mt = m_copym(m, 0, len, M_NOWAIT);
1509: if (__predict_false(mt == NULL)) {
1510: sounlock(so);
1511: mt = m_copym(m, 0, len, M_WAIT);
1512: solock(so);
1513: }
1514: *mp = mt;
1515: }
1.144 dyoung 1516: m->m_data += len;
1517: m->m_len -= len;
1518: so->so_rcv.sb_cc -= len;
1.1 cgd 1519: }
1520: if (so->so_oobmark) {
1521: if ((flags & MSG_PEEK) == 0) {
1522: so->so_oobmark -= len;
1523: if (so->so_oobmark == 0) {
1524: so->so_state |= SS_RCVATMARK;
1525: break;
1526: }
1.7 cgd 1527: } else {
1.1 cgd 1528: offset += len;
1.7 cgd 1529: if (offset == so->so_oobmark)
1530: break;
1531: }
1.1 cgd 1532: }
1533: if (flags & MSG_EOR)
1534: break;
1535: /*
1536: * If the MSG_WAITALL flag is set (for non-atomic socket),
1537: * we must not quit until "uio->uio_resid == 0" or an error
1538: * termination. If a signal/timeout occurs, return
1539: * with a short count but without error.
1540: * Keep sockbuf locked against other readers.
1541: */
1.144 dyoung 1542: while (flags & MSG_WAITALL && m == NULL && uio->uio_resid > 0 &&
1.3 andrew 1543: !sosendallatonce(so) && !nextrecord) {
1.1 cgd 1544: if (so->so_error || so->so_state & SS_CANTRCVMORE)
1545: break;
1.68 matt 1546: /*
1547: * If we are peeking and the socket receive buffer is
1548: * full, stop since we can't get more data to peek at.
1549: */
1550: if ((flags & MSG_PEEK) && sbspace(&so->so_rcv) <= 0)
1551: break;
1552: /*
1553: * If we've drained the socket buffer, tell the
1554: * protocol in case it needs to do something to
1555: * get it filled again.
1556: */
1557: if ((pr->pr_flags & PR_WANTRCVD) && so->so_pcb)
1558: (*pr->pr_usrreq)(so, PRU_RCVD,
1.140 dyoung 1559: NULL, (struct mbuf *)(long)flags, NULL, l);
1.69 thorpej 1560: SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 2");
1561: SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 2");
1.160.2.4! yamt 1562: if (wakeup_state & SS_RESTARTSYS)
! 1563: error = ERESTART;
! 1564: else
! 1565: error = sbwait(&so->so_rcv);
1.144 dyoung 1566: if (error != 0) {
1.1 cgd 1567: sbunlock(&so->so_rcv);
1.160 ad 1568: sounlock(so);
1.1 cgd 1569: splx(s);
1.144 dyoung 1570: return 0;
1.1 cgd 1571: }
1.21 christos 1572: if ((m = so->so_rcv.sb_mb) != NULL)
1.1 cgd 1573: nextrecord = m->m_nextpkt;
1.160.2.4! yamt 1574: wakeup_state = so->so_state;
1.1 cgd 1575: }
1576: }
1.3 andrew 1577:
1.146 dyoung 1578: if (m && atomic) {
1.3 andrew 1579: flags |= MSG_TRUNC;
1580: if ((flags & MSG_PEEK) == 0)
1581: (void) sbdroprecord(&so->so_rcv);
1582: }
1.1 cgd 1583: if ((flags & MSG_PEEK) == 0) {
1.144 dyoung 1584: if (m == NULL) {
1.69 thorpej 1585: /*
1.70 thorpej 1586: * First part is an inline SB_EMPTY_FIXUP(). Second
1.69 thorpej 1587: * part makes sure sb_lastrecord is up-to-date if
1588: * there is still data in the socket buffer.
1589: */
1.1 cgd 1590: so->so_rcv.sb_mb = nextrecord;
1.69 thorpej 1591: if (so->so_rcv.sb_mb == NULL) {
1592: so->so_rcv.sb_mbtail = NULL;
1593: so->so_rcv.sb_lastrecord = NULL;
1594: } else if (nextrecord->m_nextpkt == NULL)
1595: so->so_rcv.sb_lastrecord = nextrecord;
1596: }
1597: SBLASTRECORDCHK(&so->so_rcv, "soreceive 4");
1598: SBLASTMBUFCHK(&so->so_rcv, "soreceive 4");
1.1 cgd 1599: if (pr->pr_flags & PR_WANTRCVD && so->so_pcb)
1.140 dyoung 1600: (*pr->pr_usrreq)(so, PRU_RCVD, NULL,
1601: (struct mbuf *)(long)flags, NULL, l);
1.1 cgd 1602: }
1.3 andrew 1603: if (orig_resid == uio->uio_resid && orig_resid &&
1604: (flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) {
1605: sbunlock(&so->so_rcv);
1606: goto restart;
1607: }
1.108 perry 1608:
1.144 dyoung 1609: if (flagsp != NULL)
1.1 cgd 1610: *flagsp |= flags;
1.54 lukem 1611: release:
1.1 cgd 1612: sbunlock(&so->so_rcv);
1.160 ad 1613: sounlock(so);
1.1 cgd 1614: splx(s);
1.144 dyoung 1615: return error;
1.1 cgd 1616: }
1617:
1.14 mycroft 1618: int
1.54 lukem 1619: soshutdown(struct socket *so, int how)
1.1 cgd 1620: {
1.99 matt 1621: const struct protosw *pr;
1.160 ad 1622: int error;
1623:
1624: KASSERT(solocked(so));
1.34 kleink 1625:
1.54 lukem 1626: pr = so->so_proto;
1.34 kleink 1627: if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR))
1628: return (EINVAL);
1.1 cgd 1629:
1.160 ad 1630: if (how == SHUT_RD || how == SHUT_RDWR) {
1.1 cgd 1631: sorflush(so);
1.160 ad 1632: error = 0;
1633: }
1.34 kleink 1634: if (how == SHUT_WR || how == SHUT_RDWR)
1.160 ad 1635: error = (*pr->pr_usrreq)(so, PRU_SHUTDOWN, NULL,
1.140 dyoung 1636: NULL, NULL, NULL);
1.160 ad 1637:
1638: return error;
1.1 cgd 1639: }
1640:
1.160.2.4! yamt 1641: void
! 1642: sorestart(struct socket *so)
1.160.2.2 yamt 1643: {
1.160.2.4! yamt 1644: /*
! 1645: * An application has called close() on an fd on which another
! 1646: * of its threads has called a socket system call.
! 1647: * Mark this and wake everyone up, and code that would block again
! 1648: * instead returns ERESTART.
! 1649: * On system call re-entry the fd is validated and EBADF returned.
! 1650: * Any other fd will block again on the 2nd syscall.
! 1651: */
1.160.2.2 yamt 1652: solock(so);
1.160.2.4! yamt 1653: so->so_state |= SS_RESTARTSYS;
1.160.2.2 yamt 1654: cv_broadcast(&so->so_cv);
1.160.2.4! yamt 1655: cv_broadcast(&so->so_snd.sb_cv);
! 1656: cv_broadcast(&so->so_rcv.sb_cv);
1.160.2.2 yamt 1657: sounlock(so);
1658: }
1659:
1.14 mycroft 1660: void
1.54 lukem 1661: sorflush(struct socket *so)
1.1 cgd 1662: {
1.54 lukem 1663: struct sockbuf *sb, asb;
1.99 matt 1664: const struct protosw *pr;
1.160 ad 1665:
1666: KASSERT(solocked(so));
1.1 cgd 1667:
1.54 lukem 1668: sb = &so->so_rcv;
1669: pr = so->so_proto;
1.160 ad 1670: socantrcvmore(so);
1.1 cgd 1671: sb->sb_flags |= SB_NOINTR;
1.160 ad 1672: (void )sblock(sb, M_WAITOK);
1.1 cgd 1673: sbunlock(sb);
1674: asb = *sb;
1.86 wrstuden 1675: /*
1676: * Clear most of the sockbuf structure, but leave some of the
1677: * fields valid.
1678: */
1679: memset(&sb->sb_startzero, 0,
1680: sizeof(*sb) - offsetof(struct sockbuf, sb_startzero));
1.160 ad 1681: if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose) {
1682: sounlock(so);
1.1 cgd 1683: (*pr->pr_domain->dom_dispose)(asb.sb_mb);
1.160 ad 1684: solock(so);
1685: }
1.98 christos 1686: sbrelease(&asb, so);
1.1 cgd 1687: }
1688:
1.160.2.2 yamt 1689: /*
1690: * internal set SOL_SOCKET options
1691: */
1.142 dyoung 1692: static int
1.160.2.2 yamt 1693: sosetopt1(struct socket *so, const struct sockopt *sopt)
1.1 cgd 1694: {
1.160.2.2 yamt 1695: int error = EINVAL, optval, opt;
1696: struct linger l;
1697: struct timeval tv;
1.142 dyoung 1698:
1.160.2.2 yamt 1699: switch ((opt = sopt->sopt_name)) {
1.142 dyoung 1700:
1.160.2.2 yamt 1701: case SO_ACCEPTFILTER:
1702: error = accept_filt_setopt(so, sopt);
1703: KASSERT(solocked(so));
1.142 dyoung 1704: break;
1.1 cgd 1705:
1.160.2.2 yamt 1706: case SO_LINGER:
1707: error = sockopt_get(sopt, &l, sizeof(l));
1708: solock(so);
1709: if (error)
1710: break;
1711: if (l.l_linger < 0 || l.l_linger > USHRT_MAX ||
1712: l.l_linger > (INT_MAX / hz)) {
1713: error = EDOM;
1714: break;
1715: }
1716: so->so_linger = l.l_linger;
1717: if (l.l_onoff)
1718: so->so_options |= SO_LINGER;
1719: else
1720: so->so_options &= ~SO_LINGER;
1721: break;
1722:
1.142 dyoung 1723: case SO_DEBUG:
1724: case SO_KEEPALIVE:
1725: case SO_DONTROUTE:
1726: case SO_USELOOPBACK:
1727: case SO_BROADCAST:
1728: case SO_REUSEADDR:
1729: case SO_REUSEPORT:
1730: case SO_OOBINLINE:
1731: case SO_TIMESTAMP:
1.160.2.2 yamt 1732: #ifdef SO_OTIMESTAMP
1733: case SO_OTIMESTAMP:
1734: #endif
1735: error = sockopt_getint(sopt, &optval);
1736: solock(so);
1737: if (error)
1738: break;
1739: if (optval)
1740: so->so_options |= opt;
1.142 dyoung 1741: else
1.160.2.2 yamt 1742: so->so_options &= ~opt;
1.142 dyoung 1743: break;
1744:
1745: case SO_SNDBUF:
1746: case SO_RCVBUF:
1747: case SO_SNDLOWAT:
1748: case SO_RCVLOWAT:
1.160.2.2 yamt 1749: error = sockopt_getint(sopt, &optval);
1750: solock(so);
1751: if (error)
1752: break;
1.1 cgd 1753:
1.142 dyoung 1754: /*
1755: * Values < 1 make no sense for any of these
1756: * options, so disallow them.
1757: */
1.160.2.2 yamt 1758: if (optval < 1) {
1759: error = EINVAL;
1760: break;
1761: }
1.1 cgd 1762:
1.160.2.2 yamt 1763: switch (opt) {
1.1 cgd 1764: case SO_SNDBUF:
1.160.2.2 yamt 1765: if (sbreserve(&so->so_snd, (u_long)optval, so) == 0) {
1766: error = ENOBUFS;
1767: break;
1768: }
1769: so->so_snd.sb_flags &= ~SB_AUTOSIZE;
1770: break;
1771:
1.1 cgd 1772: case SO_RCVBUF:
1.160.2.2 yamt 1773: if (sbreserve(&so->so_rcv, (u_long)optval, so) == 0) {
1774: error = ENOBUFS;
1775: break;
1776: }
1777: so->so_rcv.sb_flags &= ~SB_AUTOSIZE;
1.142 dyoung 1778: break;
1779:
1780: /*
1781: * Make sure the low-water is never greater than
1782: * the high-water.
1783: */
1.1 cgd 1784: case SO_SNDLOWAT:
1.160.2.2 yamt 1785: if (optval > so->so_snd.sb_hiwat)
1786: optval = so->so_snd.sb_hiwat;
1787:
1788: so->so_snd.sb_lowat = optval;
1.142 dyoung 1789: break;
1.160.2.2 yamt 1790:
1.1 cgd 1791: case SO_RCVLOWAT:
1.160.2.2 yamt 1792: if (optval > so->so_rcv.sb_hiwat)
1793: optval = so->so_rcv.sb_hiwat;
1794:
1795: so->so_rcv.sb_lowat = optval;
1.142 dyoung 1796: break;
1797: }
1798: break;
1.28 thorpej 1799:
1.160.2.2 yamt 1800: #ifdef COMPAT_50
1801: case SO_OSNDTIMEO:
1802: case SO_ORCVTIMEO: {
1803: struct timeval50 otv;
1804: error = sockopt_get(sopt, &otv, sizeof(otv));
1805: if (error) {
1806: solock(so);
1807: break;
1808: }
1809: timeval50_to_timeval(&otv, &tv);
1810: opt = opt == SO_OSNDTIMEO ? SO_SNDTIMEO : SO_RCVTIMEO;
1811: error = 0;
1812: /*FALLTHROUGH*/
1813: }
1814: #endif /* COMPAT_50 */
1815:
1.142 dyoung 1816: case SO_SNDTIMEO:
1817: case SO_RCVTIMEO:
1.160.2.2 yamt 1818: if (error)
1819: error = sockopt_get(sopt, &tv, sizeof(tv));
1820: solock(so);
1821: if (error)
1822: break;
1.28 thorpej 1823:
1.160.2.2 yamt 1824: if (tv.tv_sec > (INT_MAX - tv.tv_usec / tick) / hz) {
1825: error = EDOM;
1826: break;
1827: }
1828:
1829: optval = tv.tv_sec * hz + tv.tv_usec / tick;
1830: if (optval == 0 && tv.tv_usec != 0)
1831: optval = 1;
1.28 thorpej 1832:
1.160.2.2 yamt 1833: switch (opt) {
1.142 dyoung 1834: case SO_SNDTIMEO:
1.160.2.2 yamt 1835: so->so_snd.sb_timeo = optval;
1.1 cgd 1836: break;
1837: case SO_RCVTIMEO:
1.160.2.2 yamt 1838: so->so_rcv.sb_timeo = optval;
1.142 dyoung 1839: break;
1840: }
1841: break;
1.1 cgd 1842:
1.142 dyoung 1843: default:
1.160.2.2 yamt 1844: solock(so);
1845: error = ENOPROTOOPT;
1846: break;
1.142 dyoung 1847: }
1.160.2.2 yamt 1848: KASSERT(solocked(so));
1849: return error;
1.142 dyoung 1850: }
1.1 cgd 1851:
1.142 dyoung 1852: int
1.160.2.2 yamt 1853: sosetopt(struct socket *so, struct sockopt *sopt)
1.142 dyoung 1854: {
1855: int error, prerr;
1.1 cgd 1856:
1.160.2.2 yamt 1857: if (sopt->sopt_level == SOL_SOCKET) {
1858: error = sosetopt1(so, sopt);
1859: KASSERT(solocked(so));
1860: } else {
1.142 dyoung 1861: error = ENOPROTOOPT;
1.160.2.2 yamt 1862: solock(so);
1863: }
1.1 cgd 1864:
1.142 dyoung 1865: if ((error == 0 || error == ENOPROTOOPT) &&
1866: so->so_proto != NULL && so->so_proto->pr_ctloutput != NULL) {
1867: /* give the protocol stack a shot */
1.160.2.2 yamt 1868: prerr = (*so->so_proto->pr_ctloutput)(PRCO_SETOPT, so, sopt);
1.142 dyoung 1869: if (prerr == 0)
1870: error = 0;
1871: else if (prerr != ENOPROTOOPT)
1872: error = prerr;
1.160.2.2 yamt 1873: }
1.160 ad 1874: sounlock(so);
1.142 dyoung 1875: return error;
1.1 cgd 1876: }
1877:
1.160.2.2 yamt 1878: /*
1879: * so_setsockopt() is a wrapper providing a sockopt structure for sosetopt()
1880: */
1881: int
1882: so_setsockopt(struct lwp *l, struct socket *so, int level, int name,
1883: const void *val, size_t valsize)
1884: {
1885: struct sockopt sopt;
1886: int error;
1887:
1888: KASSERT(valsize == 0 || val != NULL);
1889:
1890: sockopt_init(&sopt, level, name, valsize);
1891: sockopt_set(&sopt, val, valsize);
1892:
1893: error = sosetopt(so, &sopt);
1894:
1895: sockopt_destroy(&sopt);
1896:
1897: return error;
1898: }
1899:
1900: /*
1901: * internal get SOL_SOCKET options
1902: */
1903: static int
1904: sogetopt1(struct socket *so, struct sockopt *sopt)
1905: {
1906: int error, optval, opt;
1907: struct linger l;
1908: struct timeval tv;
1909:
1910: switch ((opt = sopt->sopt_name)) {
1911:
1912: case SO_ACCEPTFILTER:
1913: error = accept_filt_getopt(so, sopt);
1914: break;
1915:
1916: case SO_LINGER:
1917: l.l_onoff = (so->so_options & SO_LINGER) ? 1 : 0;
1918: l.l_linger = so->so_linger;
1919:
1920: error = sockopt_set(sopt, &l, sizeof(l));
1921: break;
1922:
1923: case SO_USELOOPBACK:
1924: case SO_DONTROUTE:
1925: case SO_DEBUG:
1926: case SO_KEEPALIVE:
1927: case SO_REUSEADDR:
1928: case SO_REUSEPORT:
1929: case SO_BROADCAST:
1930: case SO_OOBINLINE:
1931: case SO_TIMESTAMP:
1932: #ifdef SO_OTIMESTAMP
1933: case SO_OTIMESTAMP:
1934: #endif
1935: error = sockopt_setint(sopt, (so->so_options & opt) ? 1 : 0);
1936: break;
1937:
1938: case SO_TYPE:
1939: error = sockopt_setint(sopt, so->so_type);
1940: break;
1941:
1942: case SO_ERROR:
1943: error = sockopt_setint(sopt, so->so_error);
1944: so->so_error = 0;
1945: break;
1946:
1947: case SO_SNDBUF:
1948: error = sockopt_setint(sopt, so->so_snd.sb_hiwat);
1949: break;
1950:
1951: case SO_RCVBUF:
1952: error = sockopt_setint(sopt, so->so_rcv.sb_hiwat);
1953: break;
1954:
1955: case SO_SNDLOWAT:
1956: error = sockopt_setint(sopt, so->so_snd.sb_lowat);
1957: break;
1958:
1959: case SO_RCVLOWAT:
1960: error = sockopt_setint(sopt, so->so_rcv.sb_lowat);
1961: break;
1962:
1963: #ifdef COMPAT_50
1964: case SO_OSNDTIMEO:
1965: case SO_ORCVTIMEO: {
1966: struct timeval50 otv;
1967:
1968: optval = (opt == SO_OSNDTIMEO ?
1969: so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
1970:
1971: otv.tv_sec = optval / hz;
1972: otv.tv_usec = (optval % hz) * tick;
1973:
1974: error = sockopt_set(sopt, &otv, sizeof(otv));
1975: break;
1976: }
1977: #endif /* COMPAT_50 */
1978:
1979: case SO_SNDTIMEO:
1980: case SO_RCVTIMEO:
1981: optval = (opt == SO_SNDTIMEO ?
1982: so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
1983:
1984: tv.tv_sec = optval / hz;
1985: tv.tv_usec = (optval % hz) * tick;
1986:
1987: error = sockopt_set(sopt, &tv, sizeof(tv));
1988: break;
1989:
1990: case SO_OVERFLOWED:
1991: error = sockopt_setint(sopt, so->so_rcv.sb_overflowed);
1992: break;
1993:
1994: default:
1995: error = ENOPROTOOPT;
1996: break;
1997: }
1998:
1999: return (error);
2000: }
2001:
1.14 mycroft 2002: int
1.160.2.2 yamt 2003: sogetopt(struct socket *so, struct sockopt *sopt)
1.1 cgd 2004: {
1.160 ad 2005: int error;
1.1 cgd 2006:
1.160 ad 2007: solock(so);
1.160.2.2 yamt 2008: if (sopt->sopt_level != SOL_SOCKET) {
1.1 cgd 2009: if (so->so_proto && so->so_proto->pr_ctloutput) {
1.160 ad 2010: error = ((*so->so_proto->pr_ctloutput)
1.160.2.2 yamt 2011: (PRCO_GETOPT, so, sopt));
1.1 cgd 2012: } else
1.160 ad 2013: error = (ENOPROTOOPT);
1.1 cgd 2014: } else {
1.160.2.2 yamt 2015: error = sogetopt1(so, sopt);
2016: }
2017: sounlock(so);
2018: return (error);
2019: }
1.1 cgd 2020:
1.160.2.2 yamt 2021: /*
2022: * alloc sockopt data buffer buffer
2023: * - will be released at destroy
2024: */
2025: static int
2026: sockopt_alloc(struct sockopt *sopt, size_t len, km_flag_t kmflag)
2027: {
1.1 cgd 2028:
1.160.2.2 yamt 2029: KASSERT(sopt->sopt_size == 0);
1.1 cgd 2030:
1.160.2.2 yamt 2031: if (len > sizeof(sopt->sopt_buf)) {
2032: sopt->sopt_data = kmem_zalloc(len, kmflag);
2033: if (sopt->sopt_data == NULL)
2034: return ENOMEM;
2035: } else
2036: sopt->sopt_data = sopt->sopt_buf;
1.1 cgd 2037:
1.160.2.2 yamt 2038: sopt->sopt_size = len;
2039: return 0;
2040: }
1.1 cgd 2041:
1.160.2.2 yamt 2042: /*
2043: * initialise sockopt storage
2044: * - MAY sleep during allocation
2045: */
2046: void
2047: sockopt_init(struct sockopt *sopt, int level, int name, size_t size)
2048: {
1.1 cgd 2049:
1.160.2.2 yamt 2050: memset(sopt, 0, sizeof(*sopt));
1.1 cgd 2051:
1.160.2.2 yamt 2052: sopt->sopt_level = level;
2053: sopt->sopt_name = name;
2054: (void)sockopt_alloc(sopt, size, KM_SLEEP);
2055: }
1.1 cgd 2056:
1.160.2.2 yamt 2057: /*
2058: * destroy sockopt storage
2059: * - will release any held memory references
2060: */
2061: void
2062: sockopt_destroy(struct sockopt *sopt)
2063: {
1.1 cgd 2064:
1.160.2.2 yamt 2065: if (sopt->sopt_data != sopt->sopt_buf)
2066: kmem_free(sopt->sopt_data, sopt->sopt_size);
1.1 cgd 2067:
1.160.2.2 yamt 2068: memset(sopt, 0, sizeof(*sopt));
2069: }
1.1 cgd 2070:
1.160.2.2 yamt 2071: /*
2072: * set sockopt value
2073: * - value is copied into sockopt
2074: * - memory is allocated when necessary, will not sleep
2075: */
2076: int
2077: sockopt_set(struct sockopt *sopt, const void *buf, size_t len)
2078: {
2079: int error;
1.107 darrenr 2080:
1.160.2.2 yamt 2081: if (sopt->sopt_size == 0) {
2082: error = sockopt_alloc(sopt, len, KM_NOSLEEP);
2083: if (error)
2084: return error;
2085: }
2086:
2087: KASSERT(sopt->sopt_size == len);
2088: memcpy(sopt->sopt_data, buf, len);
2089: return 0;
2090: }
2091:
2092: /*
2093: * common case of set sockopt integer value
2094: */
2095: int
2096: sockopt_setint(struct sockopt *sopt, int val)
2097: {
2098:
2099: return sockopt_set(sopt, &val, sizeof(int));
2100: }
2101:
2102: /*
2103: * get sockopt value
2104: * - correct size must be given
2105: */
2106: int
2107: sockopt_get(const struct sockopt *sopt, void *buf, size_t len)
2108: {
2109:
2110: if (sopt->sopt_size != len)
2111: return EINVAL;
2112:
2113: memcpy(buf, sopt->sopt_data, len);
2114: return 0;
2115: }
2116:
2117: /*
2118: * common case of get sockopt integer value
2119: */
2120: int
2121: sockopt_getint(const struct sockopt *sopt, int *valp)
2122: {
2123:
2124: return sockopt_get(sopt, valp, sizeof(int));
2125: }
2126:
2127: /*
2128: * set sockopt value from mbuf
2129: * - ONLY for legacy code
2130: * - mbuf is released by sockopt
2131: * - will not sleep
2132: */
2133: int
2134: sockopt_setmbuf(struct sockopt *sopt, struct mbuf *m)
2135: {
2136: size_t len;
2137: int error;
2138:
2139: len = m_length(m);
2140:
2141: if (sopt->sopt_size == 0) {
2142: error = sockopt_alloc(sopt, len, KM_NOSLEEP);
2143: if (error)
2144: return error;
2145: }
2146:
2147: KASSERT(sopt->sopt_size == len);
2148: m_copydata(m, 0, len, sopt->sopt_data);
2149: m_freem(m);
2150:
2151: return 0;
2152: }
2153:
2154: /*
2155: * get sockopt value into mbuf
2156: * - ONLY for legacy code
2157: * - mbuf to be released by the caller
2158: * - will not sleep
2159: */
2160: struct mbuf *
2161: sockopt_getmbuf(const struct sockopt *sopt)
2162: {
2163: struct mbuf *m;
2164:
2165: if (sopt->sopt_size > MCLBYTES)
2166: return NULL;
2167:
2168: m = m_get(M_DONTWAIT, MT_SOOPTS);
2169: if (m == NULL)
2170: return NULL;
2171:
2172: if (sopt->sopt_size > MLEN) {
2173: MCLGET(m, M_DONTWAIT);
2174: if ((m->m_flags & M_EXT) == 0) {
2175: m_free(m);
2176: return NULL;
1.1 cgd 2177: }
2178: }
1.160 ad 2179:
1.160.2.2 yamt 2180: memcpy(mtod(m, void *), sopt->sopt_data, sopt->sopt_size);
2181: m->m_len = sopt->sopt_size;
2182:
2183: return m;
1.1 cgd 2184: }
2185:
1.14 mycroft 2186: void
1.54 lukem 2187: sohasoutofband(struct socket *so)
1.1 cgd 2188: {
1.153 rmind 2189:
1.90 christos 2190: fownsignal(so->so_pgid, SIGURG, POLL_PRI, POLLPRI|POLLRDBAND, so);
1.160.2.2 yamt 2191: selnotify(&so->so_rcv.sb_sel, POLLPRI | POLLRDBAND, NOTE_SUBMIT);
1.1 cgd 2192: }
1.72 jdolecek 2193:
2194: static void
2195: filt_sordetach(struct knote *kn)
2196: {
2197: struct socket *so;
2198:
1.155 ad 2199: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2200: solock(so);
1.73 christos 2201: SLIST_REMOVE(&so->so_rcv.sb_sel.sel_klist, kn, knote, kn_selnext);
2202: if (SLIST_EMPTY(&so->so_rcv.sb_sel.sel_klist))
1.72 jdolecek 2203: so->so_rcv.sb_flags &= ~SB_KNOTE;
1.160 ad 2204: sounlock(so);
1.72 jdolecek 2205: }
2206:
2207: /*ARGSUSED*/
2208: static int
1.129 yamt 2209: filt_soread(struct knote *kn, long hint)
1.72 jdolecek 2210: {
2211: struct socket *so;
1.160 ad 2212: int rv;
1.72 jdolecek 2213:
1.155 ad 2214: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2215: if (hint != NOTE_SUBMIT)
2216: solock(so);
1.72 jdolecek 2217: kn->kn_data = so->so_rcv.sb_cc;
2218: if (so->so_state & SS_CANTRCVMORE) {
1.108 perry 2219: kn->kn_flags |= EV_EOF;
1.72 jdolecek 2220: kn->kn_fflags = so->so_error;
1.160 ad 2221: rv = 1;
2222: } else if (so->so_error) /* temporary udp error */
2223: rv = 1;
2224: else if (kn->kn_sfflags & NOTE_LOWAT)
2225: rv = (kn->kn_data >= kn->kn_sdata);
2226: else
2227: rv = (kn->kn_data >= so->so_rcv.sb_lowat);
2228: if (hint != NOTE_SUBMIT)
2229: sounlock(so);
2230: return rv;
1.72 jdolecek 2231: }
2232:
2233: static void
2234: filt_sowdetach(struct knote *kn)
2235: {
2236: struct socket *so;
2237:
1.155 ad 2238: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2239: solock(so);
1.73 christos 2240: SLIST_REMOVE(&so->so_snd.sb_sel.sel_klist, kn, knote, kn_selnext);
2241: if (SLIST_EMPTY(&so->so_snd.sb_sel.sel_klist))
1.72 jdolecek 2242: so->so_snd.sb_flags &= ~SB_KNOTE;
1.160 ad 2243: sounlock(so);
1.72 jdolecek 2244: }
2245:
2246: /*ARGSUSED*/
2247: static int
1.129 yamt 2248: filt_sowrite(struct knote *kn, long hint)
1.72 jdolecek 2249: {
2250: struct socket *so;
1.160 ad 2251: int rv;
1.72 jdolecek 2252:
1.155 ad 2253: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2254: if (hint != NOTE_SUBMIT)
2255: solock(so);
1.72 jdolecek 2256: kn->kn_data = sbspace(&so->so_snd);
2257: if (so->so_state & SS_CANTSENDMORE) {
1.108 perry 2258: kn->kn_flags |= EV_EOF;
1.72 jdolecek 2259: kn->kn_fflags = so->so_error;
1.160 ad 2260: rv = 1;
2261: } else if (so->so_error) /* temporary udp error */
2262: rv = 1;
2263: else if (((so->so_state & SS_ISCONNECTED) == 0) &&
1.72 jdolecek 2264: (so->so_proto->pr_flags & PR_CONNREQUIRED))
1.160 ad 2265: rv = 0;
2266: else if (kn->kn_sfflags & NOTE_LOWAT)
2267: rv = (kn->kn_data >= kn->kn_sdata);
2268: else
2269: rv = (kn->kn_data >= so->so_snd.sb_lowat);
2270: if (hint != NOTE_SUBMIT)
2271: sounlock(so);
2272: return rv;
1.72 jdolecek 2273: }
2274:
2275: /*ARGSUSED*/
2276: static int
1.129 yamt 2277: filt_solisten(struct knote *kn, long hint)
1.72 jdolecek 2278: {
2279: struct socket *so;
1.160 ad 2280: int rv;
1.72 jdolecek 2281:
1.155 ad 2282: so = ((file_t *)kn->kn_obj)->f_data;
1.72 jdolecek 2283:
2284: /*
2285: * Set kn_data to number of incoming connections, not
2286: * counting partial (incomplete) connections.
1.108 perry 2287: */
1.160 ad 2288: if (hint != NOTE_SUBMIT)
2289: solock(so);
1.72 jdolecek 2290: kn->kn_data = so->so_qlen;
1.160 ad 2291: rv = (kn->kn_data > 0);
2292: if (hint != NOTE_SUBMIT)
2293: sounlock(so);
2294: return rv;
1.72 jdolecek 2295: }
2296:
2297: static const struct filterops solisten_filtops =
2298: { 1, NULL, filt_sordetach, filt_solisten };
2299: static const struct filterops soread_filtops =
2300: { 1, NULL, filt_sordetach, filt_soread };
2301: static const struct filterops sowrite_filtops =
2302: { 1, NULL, filt_sowdetach, filt_sowrite };
2303:
2304: int
1.129 yamt 2305: soo_kqfilter(struct file *fp, struct knote *kn)
1.72 jdolecek 2306: {
2307: struct socket *so;
2308: struct sockbuf *sb;
2309:
1.155 ad 2310: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2311: solock(so);
1.72 jdolecek 2312: switch (kn->kn_filter) {
2313: case EVFILT_READ:
2314: if (so->so_options & SO_ACCEPTCONN)
2315: kn->kn_fop = &solisten_filtops;
2316: else
2317: kn->kn_fop = &soread_filtops;
2318: sb = &so->so_rcv;
2319: break;
2320: case EVFILT_WRITE:
2321: kn->kn_fop = &sowrite_filtops;
2322: sb = &so->so_snd;
2323: break;
2324: default:
1.160 ad 2325: sounlock(so);
1.149 pooka 2326: return (EINVAL);
1.72 jdolecek 2327: }
1.73 christos 2328: SLIST_INSERT_HEAD(&sb->sb_sel.sel_klist, kn, kn_selnext);
1.72 jdolecek 2329: sb->sb_flags |= SB_KNOTE;
1.160 ad 2330: sounlock(so);
1.72 jdolecek 2331: return (0);
2332: }
2333:
1.154 ad 2334: static int
2335: sodopoll(struct socket *so, int events)
2336: {
2337: int revents;
2338:
2339: revents = 0;
2340:
2341: if (events & (POLLIN | POLLRDNORM))
2342: if (soreadable(so))
2343: revents |= events & (POLLIN | POLLRDNORM);
2344:
2345: if (events & (POLLOUT | POLLWRNORM))
2346: if (sowritable(so))
2347: revents |= events & (POLLOUT | POLLWRNORM);
2348:
2349: if (events & (POLLPRI | POLLRDBAND))
2350: if (so->so_oobmark || (so->so_state & SS_RCVATMARK))
2351: revents |= events & (POLLPRI | POLLRDBAND);
2352:
2353: return revents;
2354: }
2355:
2356: int
2357: sopoll(struct socket *so, int events)
2358: {
2359: int revents = 0;
2360:
1.160 ad 2361: #ifndef DIAGNOSTIC
2362: /*
2363: * Do a quick, unlocked check in expectation that the socket
2364: * will be ready for I/O. Don't do this check if DIAGNOSTIC,
2365: * as the solocked() assertions will fail.
2366: */
1.154 ad 2367: if ((revents = sodopoll(so, events)) != 0)
2368: return revents;
1.160 ad 2369: #endif
1.154 ad 2370:
1.160 ad 2371: solock(so);
1.154 ad 2372: if ((revents = sodopoll(so, events)) == 0) {
2373: if (events & (POLLIN | POLLPRI | POLLRDNORM | POLLRDBAND)) {
2374: selrecord(curlwp, &so->so_rcv.sb_sel);
1.160 ad 2375: so->so_rcv.sb_flags |= SB_NOTIFY;
1.154 ad 2376: }
2377:
2378: if (events & (POLLOUT | POLLWRNORM)) {
2379: selrecord(curlwp, &so->so_snd.sb_sel);
1.160 ad 2380: so->so_snd.sb_flags |= SB_NOTIFY;
1.154 ad 2381: }
2382: }
1.160 ad 2383: sounlock(so);
1.154 ad 2384:
2385: return revents;
2386: }
2387:
2388:
1.94 yamt 2389: #include <sys/sysctl.h>
2390:
2391: static int sysctl_kern_somaxkva(SYSCTLFN_PROTO);
2392:
2393: /*
2394: * sysctl helper routine for kern.somaxkva. ensures that the given
2395: * value is not too small.
2396: * (XXX should we maybe make sure it's not too large as well?)
2397: */
2398: static int
2399: sysctl_kern_somaxkva(SYSCTLFN_ARGS)
2400: {
2401: int error, new_somaxkva;
2402: struct sysctlnode node;
2403:
2404: new_somaxkva = somaxkva;
2405: node = *rnode;
2406: node.sysctl_data = &new_somaxkva;
2407: error = sysctl_lookup(SYSCTLFN_CALL(&node));
2408: if (error || newp == NULL)
2409: return (error);
2410:
2411: if (new_somaxkva < (16 * 1024 * 1024)) /* sanity */
2412: return (EINVAL);
2413:
1.136 ad 2414: mutex_enter(&so_pendfree_lock);
1.94 yamt 2415: somaxkva = new_somaxkva;
1.136 ad 2416: cv_broadcast(&socurkva_cv);
2417: mutex_exit(&so_pendfree_lock);
1.94 yamt 2418:
2419: return (error);
2420: }
2421:
1.160.2.2 yamt 2422: static void
2423: sysctl_kern_somaxkva_setup(void)
1.94 yamt 2424: {
2425:
1.160.2.2 yamt 2426: KASSERT(socket_sysctllog == NULL);
2427: sysctl_createv(&socket_sysctllog, 0, NULL, NULL,
1.97 atatat 2428: CTLFLAG_PERMANENT,
2429: CTLTYPE_NODE, "kern", NULL,
2430: NULL, 0, NULL, 0,
2431: CTL_KERN, CTL_EOL);
2432:
1.160.2.2 yamt 2433: sysctl_createv(&socket_sysctllog, 0, NULL, NULL,
1.97 atatat 2434: CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1.103 atatat 2435: CTLTYPE_INT, "somaxkva",
2436: SYSCTL_DESCR("Maximum amount of kernel memory to be "
2437: "used for socket buffers"),
1.94 yamt 2438: sysctl_kern_somaxkva, 0, NULL, 0,
2439: CTL_KERN, KERN_SOMAXKVA, CTL_EOL);
2440: }
CVSweb <webmaster@jp.NetBSD.org>