Annotation of src/sys/kern/uipc_socket.c, Revision 1.199
1.199 ! elad 1: /* $NetBSD: uipc_socket.c,v 1.198 2009/12/29 04:23:43 elad Exp $ */
1.64 thorpej 2:
3: /*-
1.188 ad 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
1.188 ad 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>
1.199 ! elad 66: __KERNEL_RCSID(0, "$NetBSD: uipc_socket.c,v 1.198 2009/12/29 04:23:43 elad Exp $");
1.64 thorpej 67:
1.179 christos 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"
1.167 ad 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>
1.173 plunky 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>
1.174 pooka 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:
1.179 christos 96: #ifdef COMPAT_50
97: #include <compat/sys/time.h>
1.184 christos 98: #include <compat/sys/socket.h>
1.179 christos 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:
1.167 ad 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:
1.191 elad 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:
1.178 pooka 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]),
1.194 cegger 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:
1.191 elad 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:
1.193 elad 443: if ((action != KAUTH_NETWORK_SOCKET) &&
444: (action != KAUTH_NETWORK_BIND))
1.191 elad 445: return result;
446:
447: switch (req) {
1.193 elad 448: case KAUTH_REQ_NETWORK_BIND_PORT:
449: result = KAUTH_RESULT_ALLOW;
450: break;
451:
1.191 elad 452: case KAUTH_REQ_NETWORK_SOCKET_DROP: {
453: /* Normal users can only drop their own connections. */
454: struct socket *so = (struct socket *)arg1;
455:
1.199 ! elad 456: if (proc_uidmatch(cred, so->so_cred))
1.191 elad 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:
1.192 elad 476: case KAUTH_REQ_NETWORK_SOCKET_CANSEE:
477: result = KAUTH_RESULT_ALLOW;
478:
479: break;
480:
1.191 elad 481: default:
482: break;
483: }
484:
485: return result;
486: }
487:
1.119 yamt 488: void
489: soinit(void)
490: {
491:
1.178 pooka 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");
1.166 ad 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);
1.191 elad 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);
1.168 yamt 564: so->so_egid = kauth_cred_getegid(l->l_cred);
565: so->so_cpid = l->l_proc->p_pid;
1.160 ad 566: if (lockso != NULL) {
567: /* Caller wants us to share a lock. */
568: lock = lockso->so_lock;
569: so->so_lock = lock;
570: mutex_obj_hold(lock);
571: mutex_enter(lock);
572: } else {
573: /* Lock assigned and taken during PRU_ATTACH. */
574: }
1.140 dyoung 575: error = (*prp->pr_usrreq)(so, PRU_ATTACH, NULL,
576: (struct mbuf *)(long)proto, NULL, l);
1.160 ad 577: KASSERT(solocked(so));
1.140 dyoung 578: if (error != 0) {
1.1 cgd 579: so->so_state |= SS_NOFDREF;
580: sofree(so);
1.140 dyoung 581: return error;
1.1 cgd 582: }
1.198 elad 583: so->so_cred = kauth_cred_dup(l->l_cred);
1.160 ad 584: sounlock(so);
1.1 cgd 585: *aso = so;
1.140 dyoung 586: return 0;
1.1 cgd 587: }
588:
1.142 dyoung 589: /* On success, write file descriptor to fdout and return zero. On
590: * failure, return non-zero; *fdout will be undefined.
591: */
592: int
593: fsocreate(int domain, struct socket **sop, int type, int protocol,
594: struct lwp *l, int *fdout)
595: {
596: struct socket *so;
597: struct file *fp;
598: int fd, error;
599:
1.155 ad 600: if ((error = fd_allocfile(&fp, &fd)) != 0)
1.142 dyoung 601: return (error);
602: fp->f_flag = FREAD|FWRITE;
603: fp->f_type = DTYPE_SOCKET;
604: fp->f_ops = &socketops;
1.160 ad 605: error = socreate(domain, &so, type, protocol, l, NULL);
1.142 dyoung 606: if (error != 0) {
1.155 ad 607: fd_abort(curproc, fp, fd);
1.142 dyoung 608: } else {
609: if (sop != NULL)
610: *sop = so;
611: fp->f_data = so;
1.155 ad 612: fd_affix(curproc, fp, fd);
1.142 dyoung 613: *fdout = fd;
614: }
615: return error;
616: }
617:
1.3 andrew 618: int
1.190 dyoung 619: sofamily(const struct socket *so)
620: {
621: const struct protosw *pr;
622: const struct domain *dom;
623:
624: if ((pr = so->so_proto) == NULL)
625: return AF_UNSPEC;
626: if ((dom = pr->pr_domain) == NULL)
627: return AF_UNSPEC;
628: return dom->dom_family;
629: }
630:
631: int
1.114 christos 632: sobind(struct socket *so, struct mbuf *nam, struct lwp *l)
1.1 cgd 633: {
1.160 ad 634: int error;
1.1 cgd 635:
1.160 ad 636: solock(so);
1.140 dyoung 637: error = (*so->so_proto->pr_usrreq)(so, PRU_BIND, NULL, nam, NULL, l);
1.160 ad 638: sounlock(so);
1.140 dyoung 639: return error;
1.1 cgd 640: }
641:
1.3 andrew 642: int
1.150 elad 643: solisten(struct socket *so, int backlog, struct lwp *l)
1.1 cgd 644: {
1.160 ad 645: int error;
1.1 cgd 646:
1.160 ad 647: solock(so);
1.158 ad 648: if ((so->so_state & (SS_ISCONNECTED | SS_ISCONNECTING |
1.163 ad 649: SS_ISDISCONNECTING)) != 0) {
650: sounlock(so);
1.158 ad 651: return (EOPNOTSUPP);
1.163 ad 652: }
1.140 dyoung 653: error = (*so->so_proto->pr_usrreq)(so, PRU_LISTEN, NULL,
1.150 elad 654: NULL, NULL, l);
1.140 dyoung 655: if (error != 0) {
1.160 ad 656: sounlock(so);
1.140 dyoung 657: return error;
1.1 cgd 658: }
1.63 matt 659: if (TAILQ_EMPTY(&so->so_q))
1.1 cgd 660: so->so_options |= SO_ACCEPTCONN;
661: if (backlog < 0)
662: backlog = 0;
1.49 jonathan 663: so->so_qlimit = min(backlog, somaxconn);
1.160 ad 664: sounlock(so);
1.140 dyoung 665: return 0;
1.1 cgd 666: }
667:
1.21 christos 668: void
1.54 lukem 669: sofree(struct socket *so)
1.1 cgd 670: {
1.161 ad 671: u_int refs;
1.1 cgd 672:
1.160 ad 673: KASSERT(solocked(so));
674:
675: if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0) {
676: sounlock(so);
1.1 cgd 677: return;
1.160 ad 678: }
1.43 mycroft 679: if (so->so_head) {
680: /*
681: * We must not decommission a socket that's on the accept(2)
682: * queue. If we do, then accept(2) may hang after select(2)
683: * indicated that the listening socket was ready.
684: */
1.160 ad 685: if (!soqremque(so, 0)) {
686: sounlock(so);
1.43 mycroft 687: return;
1.160 ad 688: }
1.43 mycroft 689: }
1.98 christos 690: if (so->so_rcv.sb_hiwat)
1.110 christos 691: (void)chgsbsize(so->so_uidinfo, &so->so_rcv.sb_hiwat, 0,
1.98 christos 692: RLIM_INFINITY);
693: if (so->so_snd.sb_hiwat)
1.110 christos 694: (void)chgsbsize(so->so_uidinfo, &so->so_snd.sb_hiwat, 0,
1.98 christos 695: RLIM_INFINITY);
696: sbrelease(&so->so_snd, so);
1.160 ad 697: KASSERT(!cv_has_waiters(&so->so_cv));
698: KASSERT(!cv_has_waiters(&so->so_rcv.sb_cv));
699: KASSERT(!cv_has_waiters(&so->so_snd.sb_cv));
1.1 cgd 700: sorflush(so);
1.161 ad 701: refs = so->so_aborting; /* XXX */
1.177 ad 702: /* Remove acccept filter if one is present. */
1.170 tls 703: if (so->so_accf != NULL)
1.177 ad 704: (void)accept_filt_clear(so);
1.160 ad 705: sounlock(so);
1.161 ad 706: if (refs == 0) /* XXX */
707: soput(so);
1.1 cgd 708: }
709:
710: /*
711: * Close a socket on last file table reference removal.
712: * Initiate disconnect if connected.
713: * Free socket when disconnect complete.
714: */
1.3 andrew 715: int
1.54 lukem 716: soclose(struct socket *so)
1.1 cgd 717: {
1.54 lukem 718: struct socket *so2;
1.160 ad 719: int error;
720: int error2;
1.1 cgd 721:
1.54 lukem 722: error = 0;
1.160 ad 723: solock(so);
1.1 cgd 724: if (so->so_options & SO_ACCEPTCONN) {
1.172 ad 725: for (;;) {
726: if ((so2 = TAILQ_FIRST(&so->so_q0)) != 0) {
1.160 ad 727: KASSERT(solocked2(so, so2));
728: (void) soqremque(so2, 0);
729: /* soabort drops the lock. */
730: (void) soabort(so2);
731: solock(so);
1.172 ad 732: continue;
1.160 ad 733: }
1.172 ad 734: if ((so2 = TAILQ_FIRST(&so->so_q)) != 0) {
1.160 ad 735: KASSERT(solocked2(so, so2));
736: (void) soqremque(so2, 1);
737: /* soabort drops the lock. */
738: (void) soabort(so2);
739: solock(so);
1.172 ad 740: continue;
1.160 ad 741: }
1.172 ad 742: break;
743: }
1.1 cgd 744: }
745: if (so->so_pcb == 0)
746: goto discard;
747: if (so->so_state & SS_ISCONNECTED) {
748: if ((so->so_state & SS_ISDISCONNECTING) == 0) {
749: error = sodisconnect(so);
750: if (error)
751: goto drop;
752: }
753: if (so->so_options & SO_LINGER) {
1.151 ad 754: if ((so->so_state & SS_ISDISCONNECTING) && so->so_nbio)
1.1 cgd 755: goto drop;
1.21 christos 756: while (so->so_state & SS_ISCONNECTED) {
1.185 yamt 757: error = sowait(so, true, so->so_linger * hz);
1.21 christos 758: if (error)
1.1 cgd 759: break;
1.21 christos 760: }
1.1 cgd 761: }
762: }
1.54 lukem 763: drop:
1.1 cgd 764: if (so->so_pcb) {
1.160 ad 765: error2 = (*so->so_proto->pr_usrreq)(so, PRU_DETACH,
1.140 dyoung 766: NULL, NULL, NULL, NULL);
1.1 cgd 767: if (error == 0)
768: error = error2;
769: }
1.54 lukem 770: discard:
1.1 cgd 771: if (so->so_state & SS_NOFDREF)
772: panic("soclose: NOFDREF");
1.198 elad 773: kauth_cred_free(so->so_cred);
1.1 cgd 774: so->so_state |= SS_NOFDREF;
775: sofree(so);
776: return (error);
777: }
778:
779: /*
1.160 ad 780: * Must be called with the socket locked.. Will return with it unlocked.
1.1 cgd 781: */
1.3 andrew 782: int
1.54 lukem 783: soabort(struct socket *so)
1.1 cgd 784: {
1.161 ad 785: u_int refs;
1.139 yamt 786: int error;
1.160 ad 787:
788: KASSERT(solocked(so));
789: KASSERT(so->so_head == NULL);
1.1 cgd 790:
1.161 ad 791: so->so_aborting++; /* XXX */
1.140 dyoung 792: error = (*so->so_proto->pr_usrreq)(so, PRU_ABORT, NULL,
793: NULL, NULL, NULL);
1.161 ad 794: refs = --so->so_aborting; /* XXX */
1.164 drochner 795: if (error || (refs == 0)) {
1.139 yamt 796: sofree(so);
1.160 ad 797: } else {
798: sounlock(so);
1.139 yamt 799: }
800: return error;
1.1 cgd 801: }
802:
1.3 andrew 803: int
1.54 lukem 804: soaccept(struct socket *so, struct mbuf *nam)
1.1 cgd 805: {
1.160 ad 806: int error;
807:
808: KASSERT(solocked(so));
1.1 cgd 809:
1.54 lukem 810: error = 0;
1.1 cgd 811: if ((so->so_state & SS_NOFDREF) == 0)
812: panic("soaccept: !NOFDREF");
813: so->so_state &= ~SS_NOFDREF;
1.55 thorpej 814: if ((so->so_state & SS_ISDISCONNECTED) == 0 ||
815: (so->so_proto->pr_flags & PR_ABRTACPTDIS) == 0)
1.41 mycroft 816: error = (*so->so_proto->pr_usrreq)(so, PRU_ACCEPT,
1.140 dyoung 817: NULL, nam, NULL, NULL);
1.41 mycroft 818: else
1.53 itojun 819: error = ECONNABORTED;
1.52 itojun 820:
1.1 cgd 821: return (error);
822: }
823:
1.3 andrew 824: int
1.114 christos 825: soconnect(struct socket *so, struct mbuf *nam, struct lwp *l)
1.1 cgd 826: {
1.160 ad 827: int error;
828:
829: KASSERT(solocked(so));
1.1 cgd 830:
831: if (so->so_options & SO_ACCEPTCONN)
832: return (EOPNOTSUPP);
833: /*
834: * If protocol is connection-based, can only connect once.
835: * Otherwise, if connected, try to disconnect first.
836: * This allows user to disconnect by connecting to, e.g.,
837: * a null address.
838: */
839: if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) &&
840: ((so->so_proto->pr_flags & PR_CONNREQUIRED) ||
841: (error = sodisconnect(so))))
842: error = EISCONN;
843: else
844: error = (*so->so_proto->pr_usrreq)(so, PRU_CONNECT,
1.140 dyoung 845: NULL, nam, NULL, l);
1.1 cgd 846: return (error);
847: }
848:
1.3 andrew 849: int
1.54 lukem 850: soconnect2(struct socket *so1, struct socket *so2)
1.1 cgd 851: {
1.160 ad 852: int error;
853:
854: KASSERT(solocked2(so1, so2));
1.1 cgd 855:
1.22 mycroft 856: error = (*so1->so_proto->pr_usrreq)(so1, PRU_CONNECT2,
1.140 dyoung 857: NULL, (struct mbuf *)so2, NULL, NULL);
1.1 cgd 858: return (error);
859: }
860:
1.3 andrew 861: int
1.54 lukem 862: sodisconnect(struct socket *so)
1.1 cgd 863: {
1.160 ad 864: int error;
865:
866: KASSERT(solocked(so));
1.1 cgd 867:
868: if ((so->so_state & SS_ISCONNECTED) == 0) {
869: error = ENOTCONN;
1.160 ad 870: } else if (so->so_state & SS_ISDISCONNECTING) {
1.1 cgd 871: error = EALREADY;
1.160 ad 872: } else {
873: error = (*so->so_proto->pr_usrreq)(so, PRU_DISCONNECT,
874: NULL, NULL, NULL, NULL);
1.1 cgd 875: }
1.117 yamt 876: sodopendfree();
1.1 cgd 877: return (error);
878: }
879:
1.15 mycroft 880: #define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK)
1.1 cgd 881: /*
882: * Send on a socket.
883: * If send must go all at once and message is larger than
884: * send buffering, then hard error.
885: * Lock against other senders.
886: * If must go all at once and not enough room now, then
887: * inform user that this would block and do nothing.
888: * Otherwise, if nonblocking, send as much as possible.
889: * The data to be sent is described by "uio" if nonzero,
890: * otherwise by the mbuf chain "top" (which must be null
891: * if uio is not). Data provided in mbuf chain must be small
892: * enough to send all at once.
893: *
894: * Returns nonzero on error, timeout or signal; callers
895: * must check for short counts if EINTR/ERESTART are returned.
896: * Data and control buffers are freed on return.
897: */
1.3 andrew 898: int
1.54 lukem 899: sosend(struct socket *so, struct mbuf *addr, struct uio *uio, struct mbuf *top,
1.114 christos 900: struct mbuf *control, int flags, struct lwp *l)
1.1 cgd 901: {
1.54 lukem 902: struct mbuf **mp, *m;
1.114 christos 903: struct proc *p;
1.58 jdolecek 904: long space, len, resid, clen, mlen;
905: int error, s, dontroute, atomic;
1.196 dsl 906: short wakeup_state = 0;
1.54 lukem 907:
1.114 christos 908: p = l->l_proc;
1.117 yamt 909: sodopendfree();
1.160 ad 910: clen = 0;
1.64 thorpej 911:
1.160 ad 912: /*
913: * solock() provides atomicity of access. splsoftnet() prevents
914: * protocol processing soft interrupts from interrupting us and
915: * blocking (expensive).
916: */
917: s = splsoftnet();
918: solock(so);
1.54 lukem 919: atomic = sosendallatonce(so) || top;
1.1 cgd 920: if (uio)
921: resid = uio->uio_resid;
922: else
923: resid = top->m_pkthdr.len;
1.7 cgd 924: /*
925: * In theory resid should be unsigned.
926: * However, space must be signed, as it might be less than 0
927: * if we over-committed, and we must use a signed comparison
928: * of space and resid. On the other hand, a negative resid
929: * causes us to loop sending 0-length segments to the protocol.
930: */
1.29 mycroft 931: if (resid < 0) {
932: error = EINVAL;
933: goto out;
934: }
1.1 cgd 935: dontroute =
936: (flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 &&
937: (so->so_proto->pr_flags & PR_ATOMIC);
1.165 christos 938: l->l_ru.ru_msgsnd++;
1.1 cgd 939: if (control)
940: clen = control->m_len;
1.54 lukem 941: restart:
1.21 christos 942: if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0)
1.1 cgd 943: goto out;
944: do {
1.160 ad 945: if (so->so_state & SS_CANTSENDMORE) {
946: error = EPIPE;
947: goto release;
948: }
1.48 thorpej 949: if (so->so_error) {
950: error = so->so_error;
951: so->so_error = 0;
952: goto release;
953: }
1.1 cgd 954: if ((so->so_state & SS_ISCONNECTED) == 0) {
955: if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
956: if ((so->so_state & SS_ISCONFIRMING) == 0 &&
1.160 ad 957: !(resid == 0 && clen != 0)) {
958: error = ENOTCONN;
959: goto release;
960: }
961: } else if (addr == 0) {
962: error = EDESTADDRREQ;
963: goto release;
964: }
1.1 cgd 965: }
966: space = sbspace(&so->so_snd);
967: if (flags & MSG_OOB)
968: space += 1024;
1.21 christos 969: if ((atomic && resid > so->so_snd.sb_hiwat) ||
1.160 ad 970: clen > so->so_snd.sb_hiwat) {
971: error = EMSGSIZE;
972: goto release;
973: }
1.96 mycroft 974: if (space < resid + clen &&
1.1 cgd 975: (atomic || space < so->so_snd.sb_lowat || space < clen)) {
1.160 ad 976: if (so->so_nbio) {
977: error = EWOULDBLOCK;
978: goto release;
979: }
1.1 cgd 980: sbunlock(&so->so_snd);
1.196 dsl 981: if (wakeup_state & SS_RESTARTSYS) {
982: error = ERESTART;
983: goto out;
984: }
1.1 cgd 985: error = sbwait(&so->so_snd);
986: if (error)
987: goto out;
1.196 dsl 988: wakeup_state = so->so_state;
1.1 cgd 989: goto restart;
990: }
1.196 dsl 991: wakeup_state = 0;
1.1 cgd 992: mp = ⊤
993: space -= clen;
994: do {
1.45 tv 995: if (uio == NULL) {
996: /*
997: * Data is prepackaged in "top".
998: */
999: resid = 0;
1000: if (flags & MSG_EOR)
1001: top->m_flags |= M_EOR;
1002: } else do {
1.160 ad 1003: sounlock(so);
1004: splx(s);
1.144 dyoung 1005: if (top == NULL) {
1.78 matt 1006: m = m_gethdr(M_WAIT, MT_DATA);
1.45 tv 1007: mlen = MHLEN;
1008: m->m_pkthdr.len = 0;
1.140 dyoung 1009: m->m_pkthdr.rcvif = NULL;
1.45 tv 1010: } else {
1.78 matt 1011: m = m_get(M_WAIT, MT_DATA);
1.45 tv 1012: mlen = MLEN;
1013: }
1.78 matt 1014: MCLAIM(m, so->so_snd.sb_mowner);
1.121 yamt 1015: if (sock_loan_thresh >= 0 &&
1016: uio->uio_iov->iov_len >= sock_loan_thresh &&
1017: space >= sock_loan_thresh &&
1.64 thorpej 1018: (len = sosend_loan(so, uio, m,
1019: space)) != 0) {
1020: SOSEND_COUNTER_INCR(&sosend_loan_big);
1021: space -= len;
1022: goto have_data;
1023: }
1.45 tv 1024: if (resid >= MINCLSIZE && space >= MCLBYTES) {
1.64 thorpej 1025: SOSEND_COUNTER_INCR(&sosend_copy_big);
1.78 matt 1026: m_clget(m, M_WAIT);
1.45 tv 1027: if ((m->m_flags & M_EXT) == 0)
1028: goto nopages;
1029: mlen = MCLBYTES;
1030: if (atomic && top == 0) {
1.58 jdolecek 1031: len = lmin(MCLBYTES - max_hdr,
1.54 lukem 1032: resid);
1.45 tv 1033: m->m_data += max_hdr;
1034: } else
1.58 jdolecek 1035: len = lmin(MCLBYTES, resid);
1.45 tv 1036: space -= len;
1037: } else {
1.64 thorpej 1038: nopages:
1039: SOSEND_COUNTER_INCR(&sosend_copy_small);
1.58 jdolecek 1040: len = lmin(lmin(mlen, resid), space);
1.45 tv 1041: space -= len;
1042: /*
1043: * For datagram protocols, leave room
1044: * for protocol headers in first mbuf.
1045: */
1046: if (atomic && top == 0 && len < mlen)
1047: MH_ALIGN(m, len);
1048: }
1.144 dyoung 1049: error = uiomove(mtod(m, void *), (int)len, uio);
1.64 thorpej 1050: have_data:
1.45 tv 1051: resid = uio->uio_resid;
1052: m->m_len = len;
1053: *mp = m;
1054: top->m_pkthdr.len += len;
1.160 ad 1055: s = splsoftnet();
1056: solock(so);
1.144 dyoung 1057: if (error != 0)
1.45 tv 1058: goto release;
1059: mp = &m->m_next;
1060: if (resid <= 0) {
1061: if (flags & MSG_EOR)
1062: top->m_flags |= M_EOR;
1063: break;
1064: }
1065: } while (space > 0 && atomic);
1.108 perry 1066:
1.160 ad 1067: if (so->so_state & SS_CANTSENDMORE) {
1068: error = EPIPE;
1069: goto release;
1070: }
1.45 tv 1071: if (dontroute)
1072: so->so_options |= SO_DONTROUTE;
1073: if (resid > 0)
1074: so->so_state |= SS_MORETOCOME;
1.46 sommerfe 1075: error = (*so->so_proto->pr_usrreq)(so,
1076: (flags & MSG_OOB) ? PRU_SENDOOB : PRU_SEND,
1.160 ad 1077: top, addr, control, curlwp);
1.45 tv 1078: if (dontroute)
1079: so->so_options &= ~SO_DONTROUTE;
1080: if (resid > 0)
1081: so->so_state &= ~SS_MORETOCOME;
1082: clen = 0;
1.144 dyoung 1083: control = NULL;
1084: top = NULL;
1.45 tv 1085: mp = ⊤
1.144 dyoung 1086: if (error != 0)
1.1 cgd 1087: goto release;
1088: } while (resid && space > 0);
1089: } while (resid);
1090:
1.54 lukem 1091: release:
1.1 cgd 1092: sbunlock(&so->so_snd);
1.54 lukem 1093: out:
1.160 ad 1094: sounlock(so);
1095: splx(s);
1.1 cgd 1096: if (top)
1097: m_freem(top);
1098: if (control)
1099: m_freem(control);
1100: return (error);
1101: }
1102:
1103: /*
1.159 ad 1104: * Following replacement or removal of the first mbuf on the first
1105: * mbuf chain of a socket buffer, push necessary state changes back
1106: * into the socket buffer so that other consumers see the values
1107: * consistently. 'nextrecord' is the callers locally stored value of
1108: * the original value of sb->sb_mb->m_nextpkt which must be restored
1109: * when the lead mbuf changes. NOTE: 'nextrecord' may be NULL.
1110: */
1111: static void
1112: sbsync(struct sockbuf *sb, struct mbuf *nextrecord)
1113: {
1114:
1.160 ad 1115: KASSERT(solocked(sb->sb_so));
1116:
1.159 ad 1117: /*
1118: * First, update for the new value of nextrecord. If necessary,
1119: * make it the first record.
1120: */
1121: if (sb->sb_mb != NULL)
1122: sb->sb_mb->m_nextpkt = nextrecord;
1123: else
1124: sb->sb_mb = nextrecord;
1125:
1126: /*
1127: * Now update any dependent socket buffer fields to reflect
1128: * the new state. This is an inline of SB_EMPTY_FIXUP, with
1129: * the addition of a second clause that takes care of the
1130: * case where sb_mb has been updated, but remains the last
1131: * record.
1132: */
1133: if (sb->sb_mb == NULL) {
1134: sb->sb_mbtail = NULL;
1135: sb->sb_lastrecord = NULL;
1136: } else if (sb->sb_mb->m_nextpkt == NULL)
1137: sb->sb_lastrecord = sb->sb_mb;
1138: }
1139:
1140: /*
1.1 cgd 1141: * Implement receive operations on a socket.
1142: * We depend on the way that records are added to the sockbuf
1143: * by sbappend*. In particular, each record (mbufs linked through m_next)
1144: * must begin with an address if the protocol so specifies,
1145: * followed by an optional mbuf or mbufs containing ancillary data,
1146: * and then zero or more mbufs of data.
1147: * In order to avoid blocking network interrupts for the entire time here,
1148: * we splx() while doing the actual copy to user space.
1149: * Although the sockbuf is locked, new data may still be appended,
1150: * and thus we must maintain consistency of the sockbuf during that time.
1151: *
1152: * The caller may receive the data as a single mbuf chain by supplying
1153: * an mbuf **mp0 for use in returning the chain. The uio is then used
1154: * only for the count in uio_resid.
1155: */
1.3 andrew 1156: int
1.54 lukem 1157: soreceive(struct socket *so, struct mbuf **paddr, struct uio *uio,
1158: struct mbuf **mp0, struct mbuf **controlp, int *flagsp)
1.1 cgd 1159: {
1.116 yamt 1160: struct lwp *l = curlwp;
1.160 ad 1161: struct mbuf *m, **mp, *mt;
1.146 dyoung 1162: int atomic, flags, len, error, s, offset, moff, type, orig_resid;
1.99 matt 1163: const struct protosw *pr;
1.54 lukem 1164: struct mbuf *nextrecord;
1.67 he 1165: int mbuf_removed = 0;
1.146 dyoung 1166: const struct domain *dom;
1.196 dsl 1167: short wakeup_state = 0;
1.64 thorpej 1168:
1.54 lukem 1169: pr = so->so_proto;
1.146 dyoung 1170: atomic = pr->pr_flags & PR_ATOMIC;
1171: dom = pr->pr_domain;
1.1 cgd 1172: mp = mp0;
1.54 lukem 1173: type = 0;
1174: orig_resid = uio->uio_resid;
1.102 jonathan 1175:
1.144 dyoung 1176: if (paddr != NULL)
1177: *paddr = NULL;
1178: if (controlp != NULL)
1179: *controlp = NULL;
1180: if (flagsp != NULL)
1.1 cgd 1181: flags = *flagsp &~ MSG_EOR;
1182: else
1183: flags = 0;
1.66 enami 1184:
1185: if ((flags & MSG_DONTWAIT) == 0)
1.117 yamt 1186: sodopendfree();
1.66 enami 1187:
1.1 cgd 1188: if (flags & MSG_OOB) {
1189: m = m_get(M_WAIT, MT_DATA);
1.160 ad 1190: solock(so);
1.17 cgd 1191: error = (*pr->pr_usrreq)(so, PRU_RCVOOB, m,
1.140 dyoung 1192: (struct mbuf *)(long)(flags & MSG_PEEK), NULL, l);
1.160 ad 1193: sounlock(so);
1.1 cgd 1194: if (error)
1195: goto bad;
1196: do {
1.134 christos 1197: error = uiomove(mtod(m, void *),
1.1 cgd 1198: (int) min(uio->uio_resid, m->m_len), uio);
1199: m = m_free(m);
1.144 dyoung 1200: } while (uio->uio_resid > 0 && error == 0 && m);
1.54 lukem 1201: bad:
1.144 dyoung 1202: if (m != NULL)
1.1 cgd 1203: m_freem(m);
1.144 dyoung 1204: return error;
1.1 cgd 1205: }
1.144 dyoung 1206: if (mp != NULL)
1.140 dyoung 1207: *mp = NULL;
1.160 ad 1208:
1209: /*
1210: * solock() provides atomicity of access. splsoftnet() prevents
1211: * protocol processing soft interrupts from interrupting us and
1212: * blocking (expensive).
1213: */
1214: s = splsoftnet();
1215: solock(so);
1.1 cgd 1216: if (so->so_state & SS_ISCONFIRMING && uio->uio_resid)
1.140 dyoung 1217: (*pr->pr_usrreq)(so, PRU_RCVD, NULL, NULL, NULL, l);
1.1 cgd 1218:
1.54 lukem 1219: restart:
1.160 ad 1220: if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0) {
1221: sounlock(so);
1222: splx(s);
1.144 dyoung 1223: return error;
1.160 ad 1224: }
1.1 cgd 1225:
1226: m = so->so_rcv.sb_mb;
1227: /*
1228: * If we have less data than requested, block awaiting more
1229: * (subject to any timeout) if:
1.15 mycroft 1230: * 1. the current count is less than the low water mark,
1.1 cgd 1231: * 2. MSG_WAITALL is set, and it is possible to do the entire
1.15 mycroft 1232: * receive operation at once if we block (resid <= hiwat), or
1233: * 3. MSG_DONTWAIT is not set.
1.1 cgd 1234: * If MSG_WAITALL is set but resid is larger than the receive buffer,
1235: * we have to do the receive in sections, and thus risk returning
1236: * a short count if a timeout or signal occurs after we start.
1237: */
1.144 dyoung 1238: if (m == NULL ||
1239: ((flags & MSG_DONTWAIT) == 0 &&
1240: so->so_rcv.sb_cc < uio->uio_resid &&
1241: (so->so_rcv.sb_cc < so->so_rcv.sb_lowat ||
1242: ((flags & MSG_WAITALL) &&
1243: uio->uio_resid <= so->so_rcv.sb_hiwat)) &&
1.146 dyoung 1244: m->m_nextpkt == NULL && !atomic)) {
1.1 cgd 1245: #ifdef DIAGNOSTIC
1.144 dyoung 1246: if (m == NULL && so->so_rcv.sb_cc)
1.1 cgd 1247: panic("receive 1");
1248: #endif
1249: if (so->so_error) {
1.144 dyoung 1250: if (m != NULL)
1.15 mycroft 1251: goto dontblock;
1.1 cgd 1252: error = so->so_error;
1253: if ((flags & MSG_PEEK) == 0)
1254: so->so_error = 0;
1255: goto release;
1256: }
1257: if (so->so_state & SS_CANTRCVMORE) {
1.144 dyoung 1258: if (m != NULL)
1.15 mycroft 1259: goto dontblock;
1.1 cgd 1260: else
1261: goto release;
1262: }
1.144 dyoung 1263: for (; m != NULL; m = m->m_next)
1.1 cgd 1264: if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) {
1265: m = so->so_rcv.sb_mb;
1266: goto dontblock;
1267: }
1268: if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
1269: (so->so_proto->pr_flags & PR_CONNREQUIRED)) {
1270: error = ENOTCONN;
1271: goto release;
1272: }
1273: if (uio->uio_resid == 0)
1274: goto release;
1.151 ad 1275: if (so->so_nbio || (flags & MSG_DONTWAIT)) {
1.1 cgd 1276: error = EWOULDBLOCK;
1277: goto release;
1278: }
1.69 thorpej 1279: SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 1");
1280: SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 1");
1.1 cgd 1281: sbunlock(&so->so_rcv);
1.196 dsl 1282: if (wakeup_state & SS_RESTARTSYS)
1283: error = ERESTART;
1284: else
1285: error = sbwait(&so->so_rcv);
1.160 ad 1286: if (error != 0) {
1287: sounlock(so);
1288: splx(s);
1.144 dyoung 1289: return error;
1.160 ad 1290: }
1.196 dsl 1291: wakeup_state = so->so_state;
1.1 cgd 1292: goto restart;
1293: }
1.54 lukem 1294: dontblock:
1.69 thorpej 1295: /*
1296: * On entry here, m points to the first record of the socket buffer.
1.159 ad 1297: * From this point onward, we maintain 'nextrecord' as a cache of the
1298: * pointer to the next record in the socket buffer. We must keep the
1299: * various socket buffer pointers and local stack versions of the
1300: * pointers in sync, pushing out modifications before dropping the
1.160 ad 1301: * socket lock, and re-reading them when picking it up.
1.159 ad 1302: *
1303: * Otherwise, we will race with the network stack appending new data
1304: * or records onto the socket buffer by using inconsistent/stale
1305: * versions of the field, possibly resulting in socket buffer
1306: * corruption.
1307: *
1308: * By holding the high-level sblock(), we prevent simultaneous
1309: * readers from pulling off the front of the socket buffer.
1.69 thorpej 1310: */
1.144 dyoung 1311: if (l != NULL)
1.157 ad 1312: l->l_ru.ru_msgrcv++;
1.69 thorpej 1313: KASSERT(m == so->so_rcv.sb_mb);
1314: SBLASTRECORDCHK(&so->so_rcv, "soreceive 1");
1315: SBLASTMBUFCHK(&so->so_rcv, "soreceive 1");
1.1 cgd 1316: nextrecord = m->m_nextpkt;
1317: if (pr->pr_flags & PR_ADDR) {
1318: #ifdef DIAGNOSTIC
1319: if (m->m_type != MT_SONAME)
1320: panic("receive 1a");
1321: #endif
1.3 andrew 1322: orig_resid = 0;
1.1 cgd 1323: if (flags & MSG_PEEK) {
1324: if (paddr)
1325: *paddr = m_copy(m, 0, m->m_len);
1326: m = m->m_next;
1327: } else {
1328: sbfree(&so->so_rcv, m);
1.67 he 1329: mbuf_removed = 1;
1.144 dyoung 1330: if (paddr != NULL) {
1.1 cgd 1331: *paddr = m;
1332: so->so_rcv.sb_mb = m->m_next;
1.144 dyoung 1333: m->m_next = NULL;
1.1 cgd 1334: m = so->so_rcv.sb_mb;
1335: } else {
1336: MFREE(m, so->so_rcv.sb_mb);
1337: m = so->so_rcv.sb_mb;
1338: }
1.159 ad 1339: sbsync(&so->so_rcv, nextrecord);
1.1 cgd 1340: }
1341: }
1.159 ad 1342:
1343: /*
1344: * Process one or more MT_CONTROL mbufs present before any data mbufs
1345: * in the first mbuf chain on the socket buffer. If MSG_PEEK, we
1346: * just copy the data; if !MSG_PEEK, we call into the protocol to
1347: * perform externalization (or freeing if controlp == NULL).
1348: */
1349: if (__predict_false(m != NULL && m->m_type == MT_CONTROL)) {
1350: struct mbuf *cm = NULL, *cmn;
1351: struct mbuf **cme = &cm;
1352:
1353: do {
1354: if (flags & MSG_PEEK) {
1355: if (controlp != NULL) {
1356: *controlp = m_copy(m, 0, m->m_len);
1357: controlp = &(*controlp)->m_next;
1358: }
1359: m = m->m_next;
1360: } else {
1361: sbfree(&so->so_rcv, m);
1.1 cgd 1362: so->so_rcv.sb_mb = m->m_next;
1.144 dyoung 1363: m->m_next = NULL;
1.159 ad 1364: *cme = m;
1365: cme = &(*cme)->m_next;
1.1 cgd 1366: m = so->so_rcv.sb_mb;
1.159 ad 1367: }
1368: } while (m != NULL && m->m_type == MT_CONTROL);
1369: if ((flags & MSG_PEEK) == 0)
1370: sbsync(&so->so_rcv, nextrecord);
1371: for (; cm != NULL; cm = cmn) {
1372: cmn = cm->m_next;
1373: cm->m_next = NULL;
1374: type = mtod(cm, struct cmsghdr *)->cmsg_type;
1375: if (controlp != NULL) {
1376: if (dom->dom_externalize != NULL &&
1377: type == SCM_RIGHTS) {
1.160 ad 1378: sounlock(so);
1.159 ad 1379: splx(s);
1380: error = (*dom->dom_externalize)(cm, l);
1381: s = splsoftnet();
1.160 ad 1382: solock(so);
1.159 ad 1383: }
1384: *controlp = cm;
1385: while (*controlp != NULL)
1386: controlp = &(*controlp)->m_next;
1.1 cgd 1387: } else {
1.106 itojun 1388: /*
1389: * Dispose of any SCM_RIGHTS message that went
1390: * through the read path rather than recv.
1391: */
1.159 ad 1392: if (dom->dom_dispose != NULL &&
1393: type == SCM_RIGHTS) {
1.160 ad 1394: sounlock(so);
1.159 ad 1395: (*dom->dom_dispose)(cm);
1.160 ad 1396: solock(so);
1.159 ad 1397: }
1398: m_freem(cm);
1.1 cgd 1399: }
1400: }
1.159 ad 1401: if (m != NULL)
1402: nextrecord = so->so_rcv.sb_mb->m_nextpkt;
1403: else
1404: nextrecord = so->so_rcv.sb_mb;
1405: orig_resid = 0;
1.1 cgd 1406: }
1.69 thorpej 1407:
1.159 ad 1408: /* If m is non-NULL, we have some data to read. */
1409: if (__predict_true(m != NULL)) {
1.1 cgd 1410: type = m->m_type;
1411: if (type == MT_OOBDATA)
1412: flags |= MSG_OOB;
1413: }
1.69 thorpej 1414: SBLASTRECORDCHK(&so->so_rcv, "soreceive 2");
1415: SBLASTMBUFCHK(&so->so_rcv, "soreceive 2");
1416:
1.1 cgd 1417: moff = 0;
1418: offset = 0;
1.144 dyoung 1419: while (m != NULL && uio->uio_resid > 0 && error == 0) {
1.1 cgd 1420: if (m->m_type == MT_OOBDATA) {
1421: if (type != MT_OOBDATA)
1422: break;
1423: } else if (type == MT_OOBDATA)
1424: break;
1425: #ifdef DIAGNOSTIC
1426: else if (m->m_type != MT_DATA && m->m_type != MT_HEADER)
1427: panic("receive 3");
1428: #endif
1429: so->so_state &= ~SS_RCVATMARK;
1.196 dsl 1430: wakeup_state = 0;
1.1 cgd 1431: len = uio->uio_resid;
1432: if (so->so_oobmark && len > so->so_oobmark - offset)
1433: len = so->so_oobmark - offset;
1434: if (len > m->m_len - moff)
1435: len = m->m_len - moff;
1436: /*
1437: * If mp is set, just pass back the mbufs.
1438: * Otherwise copy them out via the uio, then free.
1439: * Sockbuf must be consistent here (points to current mbuf,
1440: * it points to next record) when we drop priority;
1441: * we must note any additions to the sockbuf when we
1442: * block interrupts again.
1443: */
1.144 dyoung 1444: if (mp == NULL) {
1.69 thorpej 1445: SBLASTRECORDCHK(&so->so_rcv, "soreceive uiomove");
1446: SBLASTMBUFCHK(&so->so_rcv, "soreceive uiomove");
1.160 ad 1447: sounlock(so);
1.1 cgd 1448: splx(s);
1.134 christos 1449: error = uiomove(mtod(m, char *) + moff, (int)len, uio);
1.20 mycroft 1450: s = splsoftnet();
1.160 ad 1451: solock(so);
1.144 dyoung 1452: if (error != 0) {
1.67 he 1453: /*
1454: * If any part of the record has been removed
1455: * (such as the MT_SONAME mbuf, which will
1456: * happen when PR_ADDR, and thus also
1457: * PR_ATOMIC, is set), then drop the entire
1458: * record to maintain the atomicity of the
1459: * receive operation.
1460: *
1461: * This avoids a later panic("receive 1a")
1462: * when compiled with DIAGNOSTIC.
1463: */
1.146 dyoung 1464: if (m && mbuf_removed && atomic)
1.67 he 1465: (void) sbdroprecord(&so->so_rcv);
1466:
1.57 jdolecek 1467: goto release;
1.67 he 1468: }
1.1 cgd 1469: } else
1470: uio->uio_resid -= len;
1471: if (len == m->m_len - moff) {
1472: if (m->m_flags & M_EOR)
1473: flags |= MSG_EOR;
1474: if (flags & MSG_PEEK) {
1475: m = m->m_next;
1476: moff = 0;
1477: } else {
1478: nextrecord = m->m_nextpkt;
1479: sbfree(&so->so_rcv, m);
1480: if (mp) {
1481: *mp = m;
1482: mp = &m->m_next;
1483: so->so_rcv.sb_mb = m = m->m_next;
1.140 dyoung 1484: *mp = NULL;
1.1 cgd 1485: } else {
1486: MFREE(m, so->so_rcv.sb_mb);
1487: m = so->so_rcv.sb_mb;
1488: }
1.69 thorpej 1489: /*
1490: * If m != NULL, we also know that
1491: * so->so_rcv.sb_mb != NULL.
1492: */
1493: KASSERT(so->so_rcv.sb_mb == m);
1494: if (m) {
1.1 cgd 1495: m->m_nextpkt = nextrecord;
1.69 thorpej 1496: if (nextrecord == NULL)
1497: so->so_rcv.sb_lastrecord = m;
1498: } else {
1499: so->so_rcv.sb_mb = nextrecord;
1.70 thorpej 1500: SB_EMPTY_FIXUP(&so->so_rcv);
1.69 thorpej 1501: }
1502: SBLASTRECORDCHK(&so->so_rcv, "soreceive 3");
1503: SBLASTMBUFCHK(&so->so_rcv, "soreceive 3");
1.1 cgd 1504: }
1.144 dyoung 1505: } else if (flags & MSG_PEEK)
1506: moff += len;
1507: else {
1.160 ad 1508: if (mp != NULL) {
1509: mt = m_copym(m, 0, len, M_NOWAIT);
1510: if (__predict_false(mt == NULL)) {
1511: sounlock(so);
1512: mt = m_copym(m, 0, len, M_WAIT);
1513: solock(so);
1514: }
1515: *mp = mt;
1516: }
1.144 dyoung 1517: m->m_data += len;
1518: m->m_len -= len;
1519: so->so_rcv.sb_cc -= len;
1.1 cgd 1520: }
1521: if (so->so_oobmark) {
1522: if ((flags & MSG_PEEK) == 0) {
1523: so->so_oobmark -= len;
1524: if (so->so_oobmark == 0) {
1525: so->so_state |= SS_RCVATMARK;
1526: break;
1527: }
1.7 cgd 1528: } else {
1.1 cgd 1529: offset += len;
1.7 cgd 1530: if (offset == so->so_oobmark)
1531: break;
1532: }
1.1 cgd 1533: }
1534: if (flags & MSG_EOR)
1535: break;
1536: /*
1537: * If the MSG_WAITALL flag is set (for non-atomic socket),
1538: * we must not quit until "uio->uio_resid == 0" or an error
1539: * termination. If a signal/timeout occurs, return
1540: * with a short count but without error.
1541: * Keep sockbuf locked against other readers.
1542: */
1.144 dyoung 1543: while (flags & MSG_WAITALL && m == NULL && uio->uio_resid > 0 &&
1.3 andrew 1544: !sosendallatonce(so) && !nextrecord) {
1.1 cgd 1545: if (so->so_error || so->so_state & SS_CANTRCVMORE)
1546: break;
1.68 matt 1547: /*
1548: * If we are peeking and the socket receive buffer is
1549: * full, stop since we can't get more data to peek at.
1550: */
1551: if ((flags & MSG_PEEK) && sbspace(&so->so_rcv) <= 0)
1552: break;
1553: /*
1554: * If we've drained the socket buffer, tell the
1555: * protocol in case it needs to do something to
1556: * get it filled again.
1557: */
1558: if ((pr->pr_flags & PR_WANTRCVD) && so->so_pcb)
1559: (*pr->pr_usrreq)(so, PRU_RCVD,
1.140 dyoung 1560: NULL, (struct mbuf *)(long)flags, NULL, l);
1.69 thorpej 1561: SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 2");
1562: SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 2");
1.196 dsl 1563: if (wakeup_state & SS_RESTARTSYS)
1564: error = ERESTART;
1565: else
1566: error = sbwait(&so->so_rcv);
1.144 dyoung 1567: if (error != 0) {
1.1 cgd 1568: sbunlock(&so->so_rcv);
1.160 ad 1569: sounlock(so);
1.1 cgd 1570: splx(s);
1.144 dyoung 1571: return 0;
1.1 cgd 1572: }
1.21 christos 1573: if ((m = so->so_rcv.sb_mb) != NULL)
1.1 cgd 1574: nextrecord = m->m_nextpkt;
1.196 dsl 1575: wakeup_state = so->so_state;
1.1 cgd 1576: }
1577: }
1.3 andrew 1578:
1.146 dyoung 1579: if (m && atomic) {
1.3 andrew 1580: flags |= MSG_TRUNC;
1581: if ((flags & MSG_PEEK) == 0)
1582: (void) sbdroprecord(&so->so_rcv);
1583: }
1.1 cgd 1584: if ((flags & MSG_PEEK) == 0) {
1.144 dyoung 1585: if (m == NULL) {
1.69 thorpej 1586: /*
1.70 thorpej 1587: * First part is an inline SB_EMPTY_FIXUP(). Second
1.69 thorpej 1588: * part makes sure sb_lastrecord is up-to-date if
1589: * there is still data in the socket buffer.
1590: */
1.1 cgd 1591: so->so_rcv.sb_mb = nextrecord;
1.69 thorpej 1592: if (so->so_rcv.sb_mb == NULL) {
1593: so->so_rcv.sb_mbtail = NULL;
1594: so->so_rcv.sb_lastrecord = NULL;
1595: } else if (nextrecord->m_nextpkt == NULL)
1596: so->so_rcv.sb_lastrecord = nextrecord;
1597: }
1598: SBLASTRECORDCHK(&so->so_rcv, "soreceive 4");
1599: SBLASTMBUFCHK(&so->so_rcv, "soreceive 4");
1.1 cgd 1600: if (pr->pr_flags & PR_WANTRCVD && so->so_pcb)
1.140 dyoung 1601: (*pr->pr_usrreq)(so, PRU_RCVD, NULL,
1602: (struct mbuf *)(long)flags, NULL, l);
1.1 cgd 1603: }
1.3 andrew 1604: if (orig_resid == uio->uio_resid && orig_resid &&
1605: (flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) {
1606: sbunlock(&so->so_rcv);
1607: goto restart;
1608: }
1.108 perry 1609:
1.144 dyoung 1610: if (flagsp != NULL)
1.1 cgd 1611: *flagsp |= flags;
1.54 lukem 1612: release:
1.1 cgd 1613: sbunlock(&so->so_rcv);
1.160 ad 1614: sounlock(so);
1.1 cgd 1615: splx(s);
1.144 dyoung 1616: return error;
1.1 cgd 1617: }
1618:
1.14 mycroft 1619: int
1.54 lukem 1620: soshutdown(struct socket *so, int how)
1.1 cgd 1621: {
1.99 matt 1622: const struct protosw *pr;
1.160 ad 1623: int error;
1624:
1625: KASSERT(solocked(so));
1.34 kleink 1626:
1.54 lukem 1627: pr = so->so_proto;
1.34 kleink 1628: if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR))
1629: return (EINVAL);
1.1 cgd 1630:
1.160 ad 1631: if (how == SHUT_RD || how == SHUT_RDWR) {
1.1 cgd 1632: sorflush(so);
1.160 ad 1633: error = 0;
1634: }
1.34 kleink 1635: if (how == SHUT_WR || how == SHUT_RDWR)
1.160 ad 1636: error = (*pr->pr_usrreq)(so, PRU_SHUTDOWN, NULL,
1.140 dyoung 1637: NULL, NULL, NULL);
1.160 ad 1638:
1639: return error;
1.1 cgd 1640: }
1641:
1.195 dsl 1642: void
1.196 dsl 1643: sorestart(struct socket *so)
1.188 ad 1644: {
1.196 dsl 1645: /*
1646: * An application has called close() on an fd on which another
1647: * of its threads has called a socket system call.
1648: * Mark this and wake everyone up, and code that would block again
1649: * instead returns ERESTART.
1650: * On system call re-entry the fd is validated and EBADF returned.
1651: * Any other fd will block again on the 2nd syscall.
1652: */
1.188 ad 1653: solock(so);
1.196 dsl 1654: so->so_state |= SS_RESTARTSYS;
1.188 ad 1655: cv_broadcast(&so->so_cv);
1.196 dsl 1656: cv_broadcast(&so->so_snd.sb_cv);
1657: cv_broadcast(&so->so_rcv.sb_cv);
1.188 ad 1658: sounlock(so);
1659: }
1660:
1.14 mycroft 1661: void
1.54 lukem 1662: sorflush(struct socket *so)
1.1 cgd 1663: {
1.54 lukem 1664: struct sockbuf *sb, asb;
1.99 matt 1665: const struct protosw *pr;
1.160 ad 1666:
1667: KASSERT(solocked(so));
1.1 cgd 1668:
1.54 lukem 1669: sb = &so->so_rcv;
1670: pr = so->so_proto;
1.160 ad 1671: socantrcvmore(so);
1.1 cgd 1672: sb->sb_flags |= SB_NOINTR;
1.160 ad 1673: (void )sblock(sb, M_WAITOK);
1.1 cgd 1674: sbunlock(sb);
1675: asb = *sb;
1.86 wrstuden 1676: /*
1677: * Clear most of the sockbuf structure, but leave some of the
1678: * fields valid.
1679: */
1680: memset(&sb->sb_startzero, 0,
1681: sizeof(*sb) - offsetof(struct sockbuf, sb_startzero));
1.160 ad 1682: if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose) {
1683: sounlock(so);
1.1 cgd 1684: (*pr->pr_domain->dom_dispose)(asb.sb_mb);
1.160 ad 1685: solock(so);
1686: }
1.98 christos 1687: sbrelease(&asb, so);
1.1 cgd 1688: }
1689:
1.171 plunky 1690: /*
1691: * internal set SOL_SOCKET options
1692: */
1.142 dyoung 1693: static int
1.171 plunky 1694: sosetopt1(struct socket *so, const struct sockopt *sopt)
1.1 cgd 1695: {
1.182 christos 1696: int error = EINVAL, optval, opt;
1.171 plunky 1697: struct linger l;
1698: struct timeval tv;
1.142 dyoung 1699:
1.179 christos 1700: switch ((opt = sopt->sopt_name)) {
1.142 dyoung 1701:
1.170 tls 1702: case SO_ACCEPTFILTER:
1.177 ad 1703: error = accept_filt_setopt(so, sopt);
1704: KASSERT(solocked(so));
1.170 tls 1705: break;
1706:
1.171 plunky 1707: case SO_LINGER:
1708: error = sockopt_get(sopt, &l, sizeof(l));
1.177 ad 1709: solock(so);
1.171 plunky 1710: if (error)
1.177 ad 1711: break;
1.171 plunky 1712: if (l.l_linger < 0 || l.l_linger > USHRT_MAX ||
1.177 ad 1713: l.l_linger > (INT_MAX / hz)) {
1714: error = EDOM;
1715: break;
1716: }
1.171 plunky 1717: so->so_linger = l.l_linger;
1718: if (l.l_onoff)
1719: so->so_options |= SO_LINGER;
1720: else
1721: so->so_options &= ~SO_LINGER;
1.177 ad 1722: break;
1.1 cgd 1723:
1.142 dyoung 1724: case SO_DEBUG:
1725: case SO_KEEPALIVE:
1726: case SO_DONTROUTE:
1727: case SO_USELOOPBACK:
1728: case SO_BROADCAST:
1729: case SO_REUSEADDR:
1730: case SO_REUSEPORT:
1731: case SO_OOBINLINE:
1732: case SO_TIMESTAMP:
1.184 christos 1733: #ifdef SO_OTIMESTAMP
1734: case SO_OTIMESTAMP:
1735: #endif
1.171 plunky 1736: error = sockopt_getint(sopt, &optval);
1.177 ad 1737: solock(so);
1.171 plunky 1738: if (error)
1.177 ad 1739: break;
1.171 plunky 1740: if (optval)
1.179 christos 1741: so->so_options |= opt;
1.142 dyoung 1742: else
1.179 christos 1743: so->so_options &= ~opt;
1.142 dyoung 1744: break;
1745:
1746: case SO_SNDBUF:
1747: case SO_RCVBUF:
1748: case SO_SNDLOWAT:
1749: case SO_RCVLOWAT:
1.171 plunky 1750: error = sockopt_getint(sopt, &optval);
1.177 ad 1751: solock(so);
1.171 plunky 1752: if (error)
1.177 ad 1753: break;
1.1 cgd 1754:
1.142 dyoung 1755: /*
1756: * Values < 1 make no sense for any of these
1757: * options, so disallow them.
1758: */
1.177 ad 1759: if (optval < 1) {
1760: error = EINVAL;
1761: break;
1762: }
1.1 cgd 1763:
1.179 christos 1764: switch (opt) {
1.171 plunky 1765: case SO_SNDBUF:
1.177 ad 1766: if (sbreserve(&so->so_snd, (u_long)optval, so) == 0) {
1767: error = ENOBUFS;
1768: break;
1769: }
1.171 plunky 1770: so->so_snd.sb_flags &= ~SB_AUTOSIZE;
1771: break;
1.1 cgd 1772:
1773: case SO_RCVBUF:
1.177 ad 1774: if (sbreserve(&so->so_rcv, (u_long)optval, so) == 0) {
1775: error = ENOBUFS;
1776: break;
1777: }
1.171 plunky 1778: so->so_rcv.sb_flags &= ~SB_AUTOSIZE;
1.142 dyoung 1779: break;
1780:
1781: /*
1782: * Make sure the low-water is never greater than
1783: * the high-water.
1784: */
1.1 cgd 1785: case SO_SNDLOWAT:
1.171 plunky 1786: if (optval > so->so_snd.sb_hiwat)
1787: optval = so->so_snd.sb_hiwat;
1788:
1789: so->so_snd.sb_lowat = optval;
1.142 dyoung 1790: break;
1.171 plunky 1791:
1.1 cgd 1792: case SO_RCVLOWAT:
1.171 plunky 1793: if (optval > so->so_rcv.sb_hiwat)
1794: optval = so->so_rcv.sb_hiwat;
1795:
1796: so->so_rcv.sb_lowat = optval;
1.142 dyoung 1797: break;
1798: }
1799: break;
1.28 thorpej 1800:
1.179 christos 1801: #ifdef COMPAT_50
1802: case SO_OSNDTIMEO:
1803: case SO_ORCVTIMEO: {
1804: struct timeval50 otv;
1805: error = sockopt_get(sopt, &otv, sizeof(otv));
1.186 pooka 1806: if (error) {
1807: solock(so);
1.183 christos 1808: break;
1.186 pooka 1809: }
1.179 christos 1810: timeval50_to_timeval(&otv, &tv);
1811: opt = opt == SO_OSNDTIMEO ? SO_SNDTIMEO : SO_RCVTIMEO;
1.182 christos 1812: error = 0;
1.179 christos 1813: /*FALLTHROUGH*/
1814: }
1815: #endif /* COMPAT_50 */
1816:
1.142 dyoung 1817: case SO_SNDTIMEO:
1818: case SO_RCVTIMEO:
1.182 christos 1819: if (error)
1.179 christos 1820: error = sockopt_get(sopt, &tv, sizeof(tv));
1.177 ad 1821: solock(so);
1.171 plunky 1822: if (error)
1.177 ad 1823: break;
1.171 plunky 1824:
1.177 ad 1825: if (tv.tv_sec > (INT_MAX - tv.tv_usec / tick) / hz) {
1826: error = EDOM;
1827: break;
1828: }
1.28 thorpej 1829:
1.171 plunky 1830: optval = tv.tv_sec * hz + tv.tv_usec / tick;
1831: if (optval == 0 && tv.tv_usec != 0)
1832: optval = 1;
1.28 thorpej 1833:
1.179 christos 1834: switch (opt) {
1.142 dyoung 1835: case SO_SNDTIMEO:
1.171 plunky 1836: so->so_snd.sb_timeo = optval;
1.1 cgd 1837: break;
1838: case SO_RCVTIMEO:
1.171 plunky 1839: so->so_rcv.sb_timeo = optval;
1.142 dyoung 1840: break;
1841: }
1842: break;
1.1 cgd 1843:
1.142 dyoung 1844: default:
1.177 ad 1845: solock(so);
1846: error = ENOPROTOOPT;
1847: break;
1.142 dyoung 1848: }
1.177 ad 1849: KASSERT(solocked(so));
1850: return error;
1.142 dyoung 1851: }
1.1 cgd 1852:
1.142 dyoung 1853: int
1.171 plunky 1854: sosetopt(struct socket *so, struct sockopt *sopt)
1.142 dyoung 1855: {
1856: int error, prerr;
1.1 cgd 1857:
1.177 ad 1858: if (sopt->sopt_level == SOL_SOCKET) {
1.171 plunky 1859: error = sosetopt1(so, sopt);
1.177 ad 1860: KASSERT(solocked(so));
1861: } else {
1.142 dyoung 1862: error = ENOPROTOOPT;
1.177 ad 1863: solock(so);
1864: }
1.1 cgd 1865:
1.142 dyoung 1866: if ((error == 0 || error == ENOPROTOOPT) &&
1867: so->so_proto != NULL && so->so_proto->pr_ctloutput != NULL) {
1868: /* give the protocol stack a shot */
1.171 plunky 1869: prerr = (*so->so_proto->pr_ctloutput)(PRCO_SETOPT, so, sopt);
1.142 dyoung 1870: if (prerr == 0)
1871: error = 0;
1872: else if (prerr != ENOPROTOOPT)
1873: error = prerr;
1.171 plunky 1874: }
1.160 ad 1875: sounlock(so);
1.142 dyoung 1876: return error;
1.1 cgd 1877: }
1878:
1.171 plunky 1879: /*
1880: * so_setsockopt() is a wrapper providing a sockopt structure for sosetopt()
1881: */
1882: int
1883: so_setsockopt(struct lwp *l, struct socket *so, int level, int name,
1884: const void *val, size_t valsize)
1885: {
1886: struct sockopt sopt;
1887: int error;
1888:
1889: KASSERT(valsize == 0 || val != NULL);
1890:
1891: sockopt_init(&sopt, level, name, valsize);
1892: sockopt_set(&sopt, val, valsize);
1893:
1894: error = sosetopt(so, &sopt);
1895:
1896: sockopt_destroy(&sopt);
1897:
1898: return error;
1899: }
1900:
1901: /*
1902: * internal get SOL_SOCKET options
1903: */
1904: static int
1905: sogetopt1(struct socket *so, struct sockopt *sopt)
1906: {
1.179 christos 1907: int error, optval, opt;
1.171 plunky 1908: struct linger l;
1909: struct timeval tv;
1910:
1.179 christos 1911: switch ((opt = sopt->sopt_name)) {
1.171 plunky 1912:
1913: case SO_ACCEPTFILTER:
1.177 ad 1914: error = accept_filt_getopt(so, sopt);
1.171 plunky 1915: break;
1916:
1917: case SO_LINGER:
1918: l.l_onoff = (so->so_options & SO_LINGER) ? 1 : 0;
1919: l.l_linger = so->so_linger;
1920:
1921: error = sockopt_set(sopt, &l, sizeof(l));
1922: break;
1923:
1924: case SO_USELOOPBACK:
1925: case SO_DONTROUTE:
1926: case SO_DEBUG:
1927: case SO_KEEPALIVE:
1928: case SO_REUSEADDR:
1929: case SO_REUSEPORT:
1930: case SO_BROADCAST:
1931: case SO_OOBINLINE:
1932: case SO_TIMESTAMP:
1.184 christos 1933: #ifdef SO_OTIMESTAMP
1934: case SO_OTIMESTAMP:
1935: #endif
1.179 christos 1936: error = sockopt_setint(sopt, (so->so_options & opt) ? 1 : 0);
1.171 plunky 1937: break;
1938:
1939: case SO_TYPE:
1940: error = sockopt_setint(sopt, so->so_type);
1941: break;
1942:
1943: case SO_ERROR:
1944: error = sockopt_setint(sopt, so->so_error);
1945: so->so_error = 0;
1946: break;
1947:
1948: case SO_SNDBUF:
1949: error = sockopt_setint(sopt, so->so_snd.sb_hiwat);
1950: break;
1951:
1952: case SO_RCVBUF:
1953: error = sockopt_setint(sopt, so->so_rcv.sb_hiwat);
1954: break;
1955:
1956: case SO_SNDLOWAT:
1957: error = sockopt_setint(sopt, so->so_snd.sb_lowat);
1958: break;
1959:
1960: case SO_RCVLOWAT:
1961: error = sockopt_setint(sopt, so->so_rcv.sb_lowat);
1962: break;
1963:
1.179 christos 1964: #ifdef COMPAT_50
1965: case SO_OSNDTIMEO:
1966: case SO_ORCVTIMEO: {
1967: struct timeval50 otv;
1968:
1969: optval = (opt == SO_OSNDTIMEO ?
1970: so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
1971:
1972: otv.tv_sec = optval / hz;
1973: otv.tv_usec = (optval % hz) * tick;
1974:
1975: error = sockopt_set(sopt, &otv, sizeof(otv));
1976: break;
1977: }
1978: #endif /* COMPAT_50 */
1979:
1.171 plunky 1980: case SO_SNDTIMEO:
1981: case SO_RCVTIMEO:
1.179 christos 1982: optval = (opt == SO_SNDTIMEO ?
1.171 plunky 1983: so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
1984:
1985: tv.tv_sec = optval / hz;
1986: tv.tv_usec = (optval % hz) * tick;
1987:
1988: error = sockopt_set(sopt, &tv, sizeof(tv));
1989: break;
1990:
1991: case SO_OVERFLOWED:
1992: error = sockopt_setint(sopt, so->so_rcv.sb_overflowed);
1993: break;
1994:
1995: default:
1996: error = ENOPROTOOPT;
1997: break;
1998: }
1999:
2000: return (error);
2001: }
2002:
1.14 mycroft 2003: int
1.171 plunky 2004: sogetopt(struct socket *so, struct sockopt *sopt)
1.1 cgd 2005: {
1.160 ad 2006: int error;
1.1 cgd 2007:
1.160 ad 2008: solock(so);
1.171 plunky 2009: if (sopt->sopt_level != SOL_SOCKET) {
1.1 cgd 2010: if (so->so_proto && so->so_proto->pr_ctloutput) {
1.160 ad 2011: error = ((*so->so_proto->pr_ctloutput)
1.171 plunky 2012: (PRCO_GETOPT, so, sopt));
1.1 cgd 2013: } else
1.160 ad 2014: error = (ENOPROTOOPT);
1.1 cgd 2015: } else {
1.171 plunky 2016: error = sogetopt1(so, sopt);
2017: }
2018: sounlock(so);
2019: return (error);
2020: }
2021:
2022: /*
2023: * alloc sockopt data buffer buffer
2024: * - will be released at destroy
2025: */
1.176 plunky 2026: static int
2027: sockopt_alloc(struct sockopt *sopt, size_t len, km_flag_t kmflag)
1.171 plunky 2028: {
2029:
2030: KASSERT(sopt->sopt_size == 0);
2031:
1.176 plunky 2032: if (len > sizeof(sopt->sopt_buf)) {
2033: sopt->sopt_data = kmem_zalloc(len, kmflag);
2034: if (sopt->sopt_data == NULL)
2035: return ENOMEM;
2036: } else
1.171 plunky 2037: sopt->sopt_data = sopt->sopt_buf;
2038:
2039: sopt->sopt_size = len;
1.176 plunky 2040: return 0;
1.171 plunky 2041: }
2042:
2043: /*
2044: * initialise sockopt storage
1.176 plunky 2045: * - MAY sleep during allocation
1.171 plunky 2046: */
2047: void
2048: sockopt_init(struct sockopt *sopt, int level, int name, size_t size)
2049: {
1.1 cgd 2050:
1.171 plunky 2051: memset(sopt, 0, sizeof(*sopt));
1.1 cgd 2052:
1.171 plunky 2053: sopt->sopt_level = level;
2054: sopt->sopt_name = name;
1.176 plunky 2055: (void)sockopt_alloc(sopt, size, KM_SLEEP);
1.171 plunky 2056: }
2057:
2058: /*
2059: * destroy sockopt storage
2060: * - will release any held memory references
2061: */
2062: void
2063: sockopt_destroy(struct sockopt *sopt)
2064: {
2065:
2066: if (sopt->sopt_data != sopt->sopt_buf)
1.173 plunky 2067: kmem_free(sopt->sopt_data, sopt->sopt_size);
1.171 plunky 2068:
2069: memset(sopt, 0, sizeof(*sopt));
2070: }
2071:
2072: /*
2073: * set sockopt value
2074: * - value is copied into sockopt
1.176 plunky 2075: * - memory is allocated when necessary, will not sleep
1.171 plunky 2076: */
2077: int
2078: sockopt_set(struct sockopt *sopt, const void *buf, size_t len)
2079: {
1.176 plunky 2080: int error;
1.171 plunky 2081:
1.176 plunky 2082: if (sopt->sopt_size == 0) {
2083: error = sockopt_alloc(sopt, len, KM_NOSLEEP);
2084: if (error)
2085: return error;
2086: }
1.171 plunky 2087:
2088: KASSERT(sopt->sopt_size == len);
2089: memcpy(sopt->sopt_data, buf, len);
2090: return 0;
2091: }
2092:
2093: /*
2094: * common case of set sockopt integer value
2095: */
2096: int
2097: sockopt_setint(struct sockopt *sopt, int val)
2098: {
2099:
2100: return sockopt_set(sopt, &val, sizeof(int));
2101: }
2102:
2103: /*
2104: * get sockopt value
2105: * - correct size must be given
2106: */
2107: int
2108: sockopt_get(const struct sockopt *sopt, void *buf, size_t len)
2109: {
1.170 tls 2110:
1.171 plunky 2111: if (sopt->sopt_size != len)
2112: return EINVAL;
1.1 cgd 2113:
1.171 plunky 2114: memcpy(buf, sopt->sopt_data, len);
2115: return 0;
2116: }
1.1 cgd 2117:
1.171 plunky 2118: /*
2119: * common case of get sockopt integer value
2120: */
2121: int
2122: sockopt_getint(const struct sockopt *sopt, int *valp)
2123: {
1.1 cgd 2124:
1.171 plunky 2125: return sockopt_get(sopt, valp, sizeof(int));
2126: }
1.1 cgd 2127:
1.171 plunky 2128: /*
2129: * set sockopt value from mbuf
2130: * - ONLY for legacy code
2131: * - mbuf is released by sockopt
1.176 plunky 2132: * - will not sleep
1.171 plunky 2133: */
2134: int
2135: sockopt_setmbuf(struct sockopt *sopt, struct mbuf *m)
2136: {
2137: size_t len;
1.176 plunky 2138: int error;
1.1 cgd 2139:
1.171 plunky 2140: len = m_length(m);
1.1 cgd 2141:
1.176 plunky 2142: if (sopt->sopt_size == 0) {
2143: error = sockopt_alloc(sopt, len, KM_NOSLEEP);
2144: if (error)
2145: return error;
2146: }
1.1 cgd 2147:
1.171 plunky 2148: KASSERT(sopt->sopt_size == len);
2149: m_copydata(m, 0, len, sopt->sopt_data);
2150: m_freem(m);
1.1 cgd 2151:
1.171 plunky 2152: return 0;
2153: }
1.1 cgd 2154:
1.171 plunky 2155: /*
2156: * get sockopt value into mbuf
2157: * - ONLY for legacy code
2158: * - mbuf to be released by the caller
1.176 plunky 2159: * - will not sleep
1.171 plunky 2160: */
2161: struct mbuf *
2162: sockopt_getmbuf(const struct sockopt *sopt)
2163: {
2164: struct mbuf *m;
1.107 darrenr 2165:
1.176 plunky 2166: if (sopt->sopt_size > MCLBYTES)
2167: return NULL;
2168:
2169: m = m_get(M_DONTWAIT, MT_SOOPTS);
1.171 plunky 2170: if (m == NULL)
2171: return NULL;
2172:
1.176 plunky 2173: if (sopt->sopt_size > MLEN) {
2174: MCLGET(m, M_DONTWAIT);
2175: if ((m->m_flags & M_EXT) == 0) {
2176: m_free(m);
2177: return NULL;
2178: }
1.1 cgd 2179: }
1.176 plunky 2180:
2181: memcpy(mtod(m, void *), sopt->sopt_data, sopt->sopt_size);
2182: m->m_len = sopt->sopt_size;
1.160 ad 2183:
1.171 plunky 2184: return m;
1.1 cgd 2185: }
2186:
1.14 mycroft 2187: void
1.54 lukem 2188: sohasoutofband(struct socket *so)
1.1 cgd 2189: {
1.153 rmind 2190:
1.90 christos 2191: fownsignal(so->so_pgid, SIGURG, POLL_PRI, POLLPRI|POLLRDBAND, so);
1.189 ad 2192: selnotify(&so->so_rcv.sb_sel, POLLPRI | POLLRDBAND, NOTE_SUBMIT);
1.1 cgd 2193: }
1.72 jdolecek 2194:
2195: static void
2196: filt_sordetach(struct knote *kn)
2197: {
2198: struct socket *so;
2199:
1.155 ad 2200: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2201: solock(so);
1.73 christos 2202: SLIST_REMOVE(&so->so_rcv.sb_sel.sel_klist, kn, knote, kn_selnext);
2203: if (SLIST_EMPTY(&so->so_rcv.sb_sel.sel_klist))
1.72 jdolecek 2204: so->so_rcv.sb_flags &= ~SB_KNOTE;
1.160 ad 2205: sounlock(so);
1.72 jdolecek 2206: }
2207:
2208: /*ARGSUSED*/
2209: static int
1.129 yamt 2210: filt_soread(struct knote *kn, long hint)
1.72 jdolecek 2211: {
2212: struct socket *so;
1.160 ad 2213: int rv;
1.72 jdolecek 2214:
1.155 ad 2215: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2216: if (hint != NOTE_SUBMIT)
2217: solock(so);
1.72 jdolecek 2218: kn->kn_data = so->so_rcv.sb_cc;
2219: if (so->so_state & SS_CANTRCVMORE) {
1.108 perry 2220: kn->kn_flags |= EV_EOF;
1.72 jdolecek 2221: kn->kn_fflags = so->so_error;
1.160 ad 2222: rv = 1;
2223: } else if (so->so_error) /* temporary udp error */
2224: rv = 1;
2225: else if (kn->kn_sfflags & NOTE_LOWAT)
2226: rv = (kn->kn_data >= kn->kn_sdata);
2227: else
2228: rv = (kn->kn_data >= so->so_rcv.sb_lowat);
2229: if (hint != NOTE_SUBMIT)
2230: sounlock(so);
2231: return rv;
1.72 jdolecek 2232: }
2233:
2234: static void
2235: filt_sowdetach(struct knote *kn)
2236: {
2237: struct socket *so;
2238:
1.155 ad 2239: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2240: solock(so);
1.73 christos 2241: SLIST_REMOVE(&so->so_snd.sb_sel.sel_klist, kn, knote, kn_selnext);
2242: if (SLIST_EMPTY(&so->so_snd.sb_sel.sel_klist))
1.72 jdolecek 2243: so->so_snd.sb_flags &= ~SB_KNOTE;
1.160 ad 2244: sounlock(so);
1.72 jdolecek 2245: }
2246:
2247: /*ARGSUSED*/
2248: static int
1.129 yamt 2249: filt_sowrite(struct knote *kn, long hint)
1.72 jdolecek 2250: {
2251: struct socket *so;
1.160 ad 2252: int rv;
1.72 jdolecek 2253:
1.155 ad 2254: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2255: if (hint != NOTE_SUBMIT)
2256: solock(so);
1.72 jdolecek 2257: kn->kn_data = sbspace(&so->so_snd);
2258: if (so->so_state & SS_CANTSENDMORE) {
1.108 perry 2259: kn->kn_flags |= EV_EOF;
1.72 jdolecek 2260: kn->kn_fflags = so->so_error;
1.160 ad 2261: rv = 1;
2262: } else if (so->so_error) /* temporary udp error */
2263: rv = 1;
2264: else if (((so->so_state & SS_ISCONNECTED) == 0) &&
1.72 jdolecek 2265: (so->so_proto->pr_flags & PR_CONNREQUIRED))
1.160 ad 2266: rv = 0;
2267: else if (kn->kn_sfflags & NOTE_LOWAT)
2268: rv = (kn->kn_data >= kn->kn_sdata);
2269: else
2270: rv = (kn->kn_data >= so->so_snd.sb_lowat);
2271: if (hint != NOTE_SUBMIT)
2272: sounlock(so);
2273: return rv;
1.72 jdolecek 2274: }
2275:
2276: /*ARGSUSED*/
2277: static int
1.129 yamt 2278: filt_solisten(struct knote *kn, long hint)
1.72 jdolecek 2279: {
2280: struct socket *so;
1.160 ad 2281: int rv;
1.72 jdolecek 2282:
1.155 ad 2283: so = ((file_t *)kn->kn_obj)->f_data;
1.72 jdolecek 2284:
2285: /*
2286: * Set kn_data to number of incoming connections, not
2287: * counting partial (incomplete) connections.
1.108 perry 2288: */
1.160 ad 2289: if (hint != NOTE_SUBMIT)
2290: solock(so);
1.72 jdolecek 2291: kn->kn_data = so->so_qlen;
1.160 ad 2292: rv = (kn->kn_data > 0);
2293: if (hint != NOTE_SUBMIT)
2294: sounlock(so);
2295: return rv;
1.72 jdolecek 2296: }
2297:
2298: static const struct filterops solisten_filtops =
2299: { 1, NULL, filt_sordetach, filt_solisten };
2300: static const struct filterops soread_filtops =
2301: { 1, NULL, filt_sordetach, filt_soread };
2302: static const struct filterops sowrite_filtops =
2303: { 1, NULL, filt_sowdetach, filt_sowrite };
2304:
2305: int
1.129 yamt 2306: soo_kqfilter(struct file *fp, struct knote *kn)
1.72 jdolecek 2307: {
2308: struct socket *so;
2309: struct sockbuf *sb;
2310:
1.155 ad 2311: so = ((file_t *)kn->kn_obj)->f_data;
1.160 ad 2312: solock(so);
1.72 jdolecek 2313: switch (kn->kn_filter) {
2314: case EVFILT_READ:
2315: if (so->so_options & SO_ACCEPTCONN)
2316: kn->kn_fop = &solisten_filtops;
2317: else
2318: kn->kn_fop = &soread_filtops;
2319: sb = &so->so_rcv;
2320: break;
2321: case EVFILT_WRITE:
2322: kn->kn_fop = &sowrite_filtops;
2323: sb = &so->so_snd;
2324: break;
2325: default:
1.160 ad 2326: sounlock(so);
1.149 pooka 2327: return (EINVAL);
1.72 jdolecek 2328: }
1.73 christos 2329: SLIST_INSERT_HEAD(&sb->sb_sel.sel_klist, kn, kn_selnext);
1.72 jdolecek 2330: sb->sb_flags |= SB_KNOTE;
1.160 ad 2331: sounlock(so);
1.72 jdolecek 2332: return (0);
2333: }
2334:
1.154 ad 2335: static int
2336: sodopoll(struct socket *so, int events)
2337: {
2338: int revents;
2339:
2340: revents = 0;
2341:
2342: if (events & (POLLIN | POLLRDNORM))
2343: if (soreadable(so))
2344: revents |= events & (POLLIN | POLLRDNORM);
2345:
2346: if (events & (POLLOUT | POLLWRNORM))
2347: if (sowritable(so))
2348: revents |= events & (POLLOUT | POLLWRNORM);
2349:
2350: if (events & (POLLPRI | POLLRDBAND))
2351: if (so->so_oobmark || (so->so_state & SS_RCVATMARK))
2352: revents |= events & (POLLPRI | POLLRDBAND);
2353:
2354: return revents;
2355: }
2356:
2357: int
2358: sopoll(struct socket *so, int events)
2359: {
2360: int revents = 0;
2361:
1.160 ad 2362: #ifndef DIAGNOSTIC
2363: /*
2364: * Do a quick, unlocked check in expectation that the socket
2365: * will be ready for I/O. Don't do this check if DIAGNOSTIC,
2366: * as the solocked() assertions will fail.
2367: */
1.154 ad 2368: if ((revents = sodopoll(so, events)) != 0)
2369: return revents;
1.160 ad 2370: #endif
1.154 ad 2371:
1.160 ad 2372: solock(so);
1.154 ad 2373: if ((revents = sodopoll(so, events)) == 0) {
2374: if (events & (POLLIN | POLLPRI | POLLRDNORM | POLLRDBAND)) {
2375: selrecord(curlwp, &so->so_rcv.sb_sel);
1.160 ad 2376: so->so_rcv.sb_flags |= SB_NOTIFY;
1.154 ad 2377: }
2378:
2379: if (events & (POLLOUT | POLLWRNORM)) {
2380: selrecord(curlwp, &so->so_snd.sb_sel);
1.160 ad 2381: so->so_snd.sb_flags |= SB_NOTIFY;
1.154 ad 2382: }
2383: }
1.160 ad 2384: sounlock(so);
1.154 ad 2385:
2386: return revents;
2387: }
2388:
2389:
1.94 yamt 2390: #include <sys/sysctl.h>
2391:
2392: static int sysctl_kern_somaxkva(SYSCTLFN_PROTO);
2393:
2394: /*
2395: * sysctl helper routine for kern.somaxkva. ensures that the given
2396: * value is not too small.
2397: * (XXX should we maybe make sure it's not too large as well?)
2398: */
2399: static int
2400: sysctl_kern_somaxkva(SYSCTLFN_ARGS)
2401: {
2402: int error, new_somaxkva;
2403: struct sysctlnode node;
2404:
2405: new_somaxkva = somaxkva;
2406: node = *rnode;
2407: node.sysctl_data = &new_somaxkva;
2408: error = sysctl_lookup(SYSCTLFN_CALL(&node));
2409: if (error || newp == NULL)
2410: return (error);
2411:
2412: if (new_somaxkva < (16 * 1024 * 1024)) /* sanity */
2413: return (EINVAL);
2414:
1.136 ad 2415: mutex_enter(&so_pendfree_lock);
1.94 yamt 2416: somaxkva = new_somaxkva;
1.136 ad 2417: cv_broadcast(&socurkva_cv);
2418: mutex_exit(&so_pendfree_lock);
1.94 yamt 2419:
2420: return (error);
2421: }
2422:
1.178 pooka 2423: static void
1.187 cegger 2424: sysctl_kern_somaxkva_setup(void)
1.94 yamt 2425: {
2426:
1.178 pooka 2427: KASSERT(socket_sysctllog == NULL);
2428: sysctl_createv(&socket_sysctllog, 0, NULL, NULL,
1.97 atatat 2429: CTLFLAG_PERMANENT,
2430: CTLTYPE_NODE, "kern", NULL,
2431: NULL, 0, NULL, 0,
2432: CTL_KERN, CTL_EOL);
2433:
1.178 pooka 2434: sysctl_createv(&socket_sysctllog, 0, NULL, NULL,
1.97 atatat 2435: CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1.103 atatat 2436: CTLTYPE_INT, "somaxkva",
2437: SYSCTL_DESCR("Maximum amount of kernel memory to be "
2438: "used for socket buffers"),
1.94 yamt 2439: sysctl_kern_somaxkva, 0, NULL, 0,
2440: CTL_KERN, KERN_SOMAXKVA, CTL_EOL);
2441: }
CVSweb <webmaster@jp.NetBSD.org>
![[BACK]](/icons/back.gif){kind=icon}
Return to uipc_socket.c CVS log ![[TXT]](/icons/text.gif){kind=icon}
![[DIR]](/icons/dir.gif){kind=icon}
Up to [cvs.NetBSD.org] / src / sys / kern