Annotation of src/sys/kern/uipc_socket.c, Revision 1.141.2.1
1.141.2.1! matt 1: /* $NetBSD: uipc_socket.c,v 1.144 2007/10/05 02:40:25 dyoung Exp $ */
1.64 thorpej 2:
3: /*-
1.137 ad 4: * Copyright (c) 2002, 2007 The NetBSD Foundation, Inc.
1.64 thorpej 5: * All rights reserved.
6: *
7: * This code is derived from software contributed to The NetBSD Foundation
8: * by Jason R. Thorpe of Wasabi Systems, Inc.
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: * 3. All advertising materials mentioning features or use of this software
19: * must display the following acknowledgement:
20: * This product includes software developed by the NetBSD
21: * Foundation, Inc. and its contributors.
22: * 4. Neither the name of The NetBSD Foundation nor the names of its
23: * contributors may be used to endorse or promote products derived
24: * from this software without specific prior written permission.
25: *
26: * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
27: * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
28: * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
29: * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
30: * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31: * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32: * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33: * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34: * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35: * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36: * POSSIBILITY OF SUCH DAMAGE.
37: */
1.16 cgd 38:
1.1 cgd 39: /*
1.15 mycroft 40: * Copyright (c) 1982, 1986, 1988, 1990, 1993
41: * The Regents of the University of California. All rights reserved.
1.1 cgd 42: *
43: * Redistribution and use in source and binary forms, with or without
44: * modification, are permitted provided that the following conditions
45: * are met:
46: * 1. Redistributions of source code must retain the above copyright
47: * notice, this list of conditions and the following disclaimer.
48: * 2. Redistributions in binary form must reproduce the above copyright
49: * notice, this list of conditions and the following disclaimer in the
50: * documentation and/or other materials provided with the distribution.
1.85 agc 51: * 3. Neither the name of the University nor the names of its contributors
1.1 cgd 52: * may be used to endorse or promote products derived from this software
53: * without specific prior written permission.
54: *
55: * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
56: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
57: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
58: * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
59: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
60: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
61: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
62: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
63: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
64: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
65: * SUCH DAMAGE.
66: *
1.32 fvdl 67: * @(#)uipc_socket.c 8.6 (Berkeley) 5/2/95
1.1 cgd 68: */
1.59 lukem 69:
70: #include <sys/cdefs.h>
1.141.2.1! matt 71: __KERNEL_RCSID(0, "$NetBSD: uipc_socket.c,v 1.144 2007/10/05 02:40:25 dyoung Exp $");
1.64 thorpej 72:
73: #include "opt_sock_counters.h"
74: #include "opt_sosend_loan.h"
1.81 martin 75: #include "opt_mbuftrace.h"
1.84 ragge 76: #include "opt_somaxkva.h"
1.1 cgd 77:
1.9 mycroft 78: #include <sys/param.h>
79: #include <sys/systm.h>
80: #include <sys/proc.h>
81: #include <sys/file.h>
1.141.2.1! matt 82: #include <sys/filedesc.h>
1.9 mycroft 83: #include <sys/malloc.h>
84: #include <sys/mbuf.h>
85: #include <sys/domain.h>
86: #include <sys/kernel.h>
87: #include <sys/protosw.h>
88: #include <sys/socket.h>
89: #include <sys/socketvar.h>
1.21 christos 90: #include <sys/signalvar.h>
1.9 mycroft 91: #include <sys/resourcevar.h>
1.37 thorpej 92: #include <sys/pool.h>
1.72 jdolecek 93: #include <sys/event.h>
1.89 christos 94: #include <sys/poll.h>
1.118 elad 95: #include <sys/kauth.h>
1.136 ad 96: #include <sys/mutex.h>
97: #include <sys/condvar.h>
1.37 thorpej 98:
1.64 thorpej 99: #include <uvm/uvm.h>
100:
1.135 ad 101: POOL_INIT(socket_pool, sizeof(struct socket), 0, 0, 0, "sockpl", NULL,
102: IPL_SOFTNET);
1.77 thorpej 103:
104: MALLOC_DEFINE(M_SOOPTS, "soopts", "socket options");
105: MALLOC_DEFINE(M_SONAME, "soname", "socket name");
1.37 thorpej 106:
1.141.2.1! matt 107: extern const struct fileops socketops;
! 108:
1.54 lukem 109: extern int somaxconn; /* patchable (XXX sysctl) */
110: int somaxconn = SOMAXCONN;
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:
1.71 thorpej 138: #ifdef SOSEND_NO_LOAN
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:
154: #define SOCK_LOAN_CHUNK 65536
155:
1.117 yamt 156: static size_t sodopendfree(void);
157: static size_t sodopendfreel(void);
1.93 yamt 158:
1.113 thorpej 159: static vsize_t
1.129 yamt 160: sokvareserve(struct socket *so, vsize_t len)
1.80 yamt 161: {
1.98 christos 162: int error;
1.80 yamt 163:
1.136 ad 164: mutex_enter(&so_pendfree_lock);
1.80 yamt 165: while (socurkva + len > somaxkva) {
1.93 yamt 166: size_t freed;
167:
168: /*
169: * try to do pendfree.
170: */
171:
1.117 yamt 172: freed = sodopendfreel();
1.93 yamt 173:
174: /*
175: * if some kva was freed, try again.
176: */
177:
178: if (freed)
1.80 yamt 179: continue;
1.93 yamt 180:
1.80 yamt 181: SOSEND_COUNTER_INCR(&sosend_kvalimit);
1.136 ad 182: error = cv_wait_sig(&socurkva_cv, &so_pendfree_lock);
1.98 christos 183: if (error) {
184: len = 0;
185: break;
186: }
1.80 yamt 187: }
1.93 yamt 188: socurkva += len;
1.136 ad 189: mutex_exit(&so_pendfree_lock);
1.98 christos 190: return len;
1.95 yamt 191: }
192:
1.113 thorpej 193: static void
1.95 yamt 194: sokvaunreserve(vsize_t len)
195: {
196:
1.136 ad 197: mutex_enter(&so_pendfree_lock);
1.95 yamt 198: socurkva -= len;
1.136 ad 199: cv_broadcast(&socurkva_cv);
200: mutex_exit(&so_pendfree_lock);
1.95 yamt 201: }
202:
203: /*
204: * sokvaalloc: allocate kva for loan.
205: */
206:
207: vaddr_t
208: sokvaalloc(vsize_t len, struct socket *so)
209: {
210: vaddr_t lva;
211:
212: /*
213: * reserve kva.
214: */
215:
1.98 christos 216: if (sokvareserve(so, len) == 0)
217: return 0;
1.93 yamt 218:
219: /*
220: * allocate kva.
221: */
1.80 yamt 222:
1.109 yamt 223: lva = uvm_km_alloc(kernel_map, len, 0, UVM_KMF_VAONLY | UVM_KMF_WAITVA);
1.95 yamt 224: if (lva == 0) {
225: sokvaunreserve(len);
1.80 yamt 226: return (0);
1.95 yamt 227: }
1.80 yamt 228:
229: return lva;
230: }
231:
1.93 yamt 232: /*
233: * sokvafree: free kva for loan.
234: */
235:
1.80 yamt 236: void
237: sokvafree(vaddr_t sva, vsize_t len)
238: {
1.93 yamt 239:
240: /*
241: * free kva.
242: */
1.80 yamt 243:
1.109 yamt 244: uvm_km_free(kernel_map, sva, len, UVM_KMF_VAONLY);
1.93 yamt 245:
246: /*
247: * unreserve kva.
248: */
249:
1.95 yamt 250: sokvaunreserve(len);
1.80 yamt 251: }
252:
1.64 thorpej 253: static void
1.134 christos 254: sodoloanfree(struct vm_page **pgs, void *buf, size_t size)
1.64 thorpej 255: {
256: vaddr_t va, sva, eva;
257: vsize_t len;
258: paddr_t pa;
259: int i, npgs;
260:
261: eva = round_page((vaddr_t) buf + size);
262: sva = trunc_page((vaddr_t) buf);
263: len = eva - sva;
264: npgs = len >> PAGE_SHIFT;
265:
1.79 thorpej 266: if (__predict_false(pgs == NULL)) {
267: pgs = alloca(npgs * sizeof(*pgs));
1.64 thorpej 268:
1.79 thorpej 269: for (i = 0, va = sva; va < eva; i++, va += PAGE_SIZE) {
1.133 thorpej 270: if (pmap_extract(pmap_kernel(), va, &pa) == false)
1.79 thorpej 271: panic("sodoloanfree: va 0x%lx not mapped", va);
272: pgs[i] = PHYS_TO_VM_PAGE(pa);
273: }
1.64 thorpej 274: }
275:
276: pmap_kremove(sva, len);
277: pmap_update(pmap_kernel());
278: uvm_unloan(pgs, npgs, UVM_LOAN_TOPAGE);
1.80 yamt 279: sokvafree(sva, len);
1.64 thorpej 280: }
281:
282: static size_t
1.117 yamt 283: sodopendfree()
1.64 thorpej 284: {
1.93 yamt 285: size_t rv;
1.64 thorpej 286:
1.136 ad 287: mutex_enter(&so_pendfree_lock);
1.117 yamt 288: rv = sodopendfreel();
1.136 ad 289: mutex_exit(&so_pendfree_lock);
1.93 yamt 290:
291: return rv;
292: }
293:
294: /*
295: * sodopendfreel: free mbufs on "pendfree" list.
1.136 ad 296: * unlock and relock so_pendfree_lock when freeing mbufs.
1.93 yamt 297: *
1.136 ad 298: * => called with so_pendfree_lock held.
1.93 yamt 299: */
300:
301: static size_t
1.117 yamt 302: sodopendfreel()
1.93 yamt 303: {
1.137 ad 304: struct mbuf *m, *next;
1.93 yamt 305: size_t rv = 0;
1.137 ad 306: int s;
1.93 yamt 307:
1.136 ad 308: KASSERT(mutex_owned(&so_pendfree_lock));
1.64 thorpej 309:
1.137 ad 310: while (so_pendfree != NULL) {
1.64 thorpej 311: m = so_pendfree;
1.93 yamt 312: so_pendfree = NULL;
1.136 ad 313: mutex_exit(&so_pendfree_lock);
1.93 yamt 314:
315: for (; m != NULL; m = next) {
316: next = m->m_next;
317:
318: rv += m->m_ext.ext_size;
319: sodoloanfree((m->m_flags & M_EXT_PAGES) ?
320: m->m_ext.ext_pgs : NULL, m->m_ext.ext_buf,
321: m->m_ext.ext_size);
1.137 ad 322: s = splvm();
1.93 yamt 323: pool_cache_put(&mbpool_cache, m);
1.137 ad 324: splx(s);
1.93 yamt 325: }
1.64 thorpej 326:
1.136 ad 327: mutex_enter(&so_pendfree_lock);
1.64 thorpej 328: }
329:
330: return (rv);
331: }
332:
1.80 yamt 333: void
1.134 christos 334: soloanfree(struct mbuf *m, void *buf, size_t size, void *arg)
1.64 thorpej 335: {
336:
337: if (m == NULL) {
1.93 yamt 338:
339: /*
340: * called from MEXTREMOVE.
341: */
342:
1.79 thorpej 343: sodoloanfree(NULL, buf, size);
1.64 thorpej 344: return;
345: }
346:
1.93 yamt 347: /*
348: * postpone freeing mbuf.
349: *
350: * we can't do it in interrupt context
351: * because we need to put kva back to kernel_map.
352: */
353:
1.136 ad 354: mutex_enter(&so_pendfree_lock);
1.92 yamt 355: m->m_next = so_pendfree;
356: so_pendfree = m;
1.136 ad 357: cv_broadcast(&socurkva_cv);
358: mutex_exit(&so_pendfree_lock);
1.64 thorpej 359: }
360:
361: static long
362: sosend_loan(struct socket *so, struct uio *uio, struct mbuf *m, long space)
363: {
364: struct iovec *iov = uio->uio_iov;
365: vaddr_t sva, eva;
366: vsize_t len;
367: vaddr_t lva, va;
1.80 yamt 368: int npgs, i, error;
1.64 thorpej 369:
1.116 yamt 370: if (VMSPACE_IS_KERNEL_P(uio->uio_vmspace))
1.64 thorpej 371: return (0);
372:
373: if (iov->iov_len < (size_t) space)
374: space = iov->iov_len;
375: if (space > SOCK_LOAN_CHUNK)
376: space = SOCK_LOAN_CHUNK;
377:
378: eva = round_page((vaddr_t) iov->iov_base + space);
379: sva = trunc_page((vaddr_t) iov->iov_base);
380: len = eva - sva;
381: npgs = len >> PAGE_SHIFT;
382:
1.79 thorpej 383: /* XXX KDASSERT */
384: KASSERT(npgs <= M_EXT_MAXPAGES);
385:
1.80 yamt 386: lva = sokvaalloc(len, so);
1.64 thorpej 387: if (lva == 0)
1.80 yamt 388: return 0;
1.64 thorpej 389:
1.116 yamt 390: error = uvm_loan(&uio->uio_vmspace->vm_map, sva, len,
1.79 thorpej 391: m->m_ext.ext_pgs, UVM_LOAN_TOPAGE);
1.64 thorpej 392: if (error) {
1.80 yamt 393: sokvafree(lva, len);
1.64 thorpej 394: return (0);
395: }
396:
397: for (i = 0, va = lva; i < npgs; i++, va += PAGE_SIZE)
1.79 thorpej 398: pmap_kenter_pa(va, VM_PAGE_TO_PHYS(m->m_ext.ext_pgs[i]),
399: VM_PROT_READ);
1.64 thorpej 400: pmap_update(pmap_kernel());
401:
402: lva += (vaddr_t) iov->iov_base & PAGE_MASK;
403:
1.134 christos 404: MEXTADD(m, (void *) lva, space, M_MBUF, soloanfree, so);
1.79 thorpej 405: m->m_flags |= M_EXT_PAGES | M_EXT_ROMAP;
1.64 thorpej 406:
407: uio->uio_resid -= space;
408: /* uio_offset not updated, not set/used for write(2) */
1.134 christos 409: uio->uio_iov->iov_base = (char *)uio->uio_iov->iov_base + space;
1.64 thorpej 410: uio->uio_iov->iov_len -= space;
411: if (uio->uio_iov->iov_len == 0) {
412: uio->uio_iov++;
413: uio->uio_iovcnt--;
414: }
415:
416: return (space);
417: }
418:
1.119 yamt 419: static int
1.129 yamt 420: sokva_reclaim_callback(struct callback_entry *ce, void *obj, void *arg)
1.119 yamt 421: {
422:
423: KASSERT(ce == &sokva_reclaimerentry);
424: KASSERT(obj == NULL);
425:
426: sodopendfree();
427: if (!vm_map_starved_p(kernel_map)) {
428: return CALLBACK_CHAIN_ABORT;
429: }
430: return CALLBACK_CHAIN_CONTINUE;
431: }
432:
1.141.2.1! matt 433: struct mbuf *
! 434: getsombuf(struct socket *so)
! 435: {
! 436: struct mbuf *m;
! 437:
! 438: m = m_get(M_WAIT, MT_SONAME);
! 439: MCLAIM(m, so->so_mowner);
! 440: return m;
! 441: }
! 442:
! 443: struct mbuf *
! 444: m_intopt(struct socket *so, int val)
! 445: {
! 446: struct mbuf *m;
! 447:
! 448: m = getsombuf(so);
! 449: m->m_len = sizeof(int);
! 450: *mtod(m, int *) = val;
! 451: return m;
! 452: }
! 453:
1.119 yamt 454: void
455: soinit(void)
456: {
457:
1.136 ad 458: mutex_init(&so_pendfree_lock, MUTEX_DRIVER, IPL_VM);
459: cv_init(&socurkva_cv, "sokva");
460:
1.119 yamt 461: /* Set the initial adjusted socket buffer size. */
462: if (sb_max_set(sb_max))
463: panic("bad initial sb_max value: %lu", sb_max);
464:
465: callback_register(&vm_map_to_kernel(kernel_map)->vmk_reclaim_callback,
466: &sokva_reclaimerentry, NULL, sokva_reclaim_callback);
467: }
468:
1.1 cgd 469: /*
470: * Socket operation routines.
471: * These routines are called by the routines in
472: * sys_socket.c or from a system process, and
473: * implement the semantics of socket operations by
474: * switching out to the protocol specific routines.
475: */
476: /*ARGSUSED*/
1.3 andrew 477: int
1.114 christos 478: socreate(int dom, struct socket **aso, int type, int proto, struct lwp *l)
1.1 cgd 479: {
1.99 matt 480: const struct protosw *prp;
1.54 lukem 481: struct socket *so;
1.115 yamt 482: uid_t uid;
1.54 lukem 483: int error, s;
1.1 cgd 484:
1.132 elad 485: error = kauth_authorize_network(l->l_cred, KAUTH_NETWORK_SOCKET,
486: KAUTH_REQ_NETWORK_SOCKET_OPEN, KAUTH_ARG(dom), KAUTH_ARG(type),
487: KAUTH_ARG(proto));
1.140 dyoung 488: if (error != 0)
489: return error;
1.127 elad 490:
1.1 cgd 491: if (proto)
492: prp = pffindproto(dom, proto, type);
493: else
494: prp = pffindtype(dom, type);
1.140 dyoung 495: if (prp == NULL) {
1.120 ginsbach 496: /* no support for domain */
497: if (pffinddomain(dom) == 0)
1.140 dyoung 498: return EAFNOSUPPORT;
1.120 ginsbach 499: /* no support for socket type */
500: if (proto == 0 && type != 0)
1.140 dyoung 501: return EPROTOTYPE;
502: return EPROTONOSUPPORT;
1.120 ginsbach 503: }
1.140 dyoung 504: if (prp->pr_usrreq == NULL)
505: return EPROTONOSUPPORT;
1.1 cgd 506: if (prp->pr_type != type)
1.140 dyoung 507: return EPROTOTYPE;
1.39 matt 508: s = splsoftnet();
1.37 thorpej 509: so = pool_get(&socket_pool, PR_WAITOK);
1.140 dyoung 510: memset(so, 0, sizeof(*so));
1.31 thorpej 511: TAILQ_INIT(&so->so_q0);
512: TAILQ_INIT(&so->so_q);
1.1 cgd 513: so->so_type = type;
514: so->so_proto = prp;
1.33 matt 515: so->so_send = sosend;
516: so->so_receive = soreceive;
1.78 matt 517: #ifdef MBUFTRACE
518: so->so_rcv.sb_mowner = &prp->pr_domain->dom_mowner;
519: so->so_snd.sb_mowner = &prp->pr_domain->dom_mowner;
520: so->so_mowner = &prp->pr_domain->dom_mowner;
521: #endif
1.141.2.1! matt 522: selinit(&so->so_rcv.sb_sel);
! 523: selinit(&so->so_snd.sb_sel);
1.138 rmind 524: uid = kauth_cred_geteuid(l->l_cred);
1.115 yamt 525: so->so_uidinfo = uid_find(uid);
1.140 dyoung 526: error = (*prp->pr_usrreq)(so, PRU_ATTACH, NULL,
527: (struct mbuf *)(long)proto, NULL, l);
528: if (error != 0) {
1.1 cgd 529: so->so_state |= SS_NOFDREF;
530: sofree(so);
1.39 matt 531: splx(s);
1.140 dyoung 532: return error;
1.1 cgd 533: }
1.39 matt 534: splx(s);
1.1 cgd 535: *aso = so;
1.140 dyoung 536: return 0;
1.1 cgd 537: }
538:
1.141.2.1! matt 539: /* On success, write file descriptor to fdout and return zero. On
! 540: * failure, return non-zero; *fdout will be undefined.
! 541: */
! 542: int
! 543: fsocreate(int domain, struct socket **sop, int type, int protocol,
! 544: struct lwp *l, int *fdout)
! 545: {
! 546: struct filedesc *fdp;
! 547: struct socket *so;
! 548: struct file *fp;
! 549: int fd, error;
! 550:
! 551: fdp = l->l_proc->p_fd;
! 552: /* falloc() will use the desciptor for us */
! 553: if ((error = falloc(l, &fp, &fd)) != 0)
! 554: return (error);
! 555: fp->f_flag = FREAD|FWRITE;
! 556: fp->f_type = DTYPE_SOCKET;
! 557: fp->f_ops = &socketops;
! 558: error = socreate(domain, &so, type, protocol, l);
! 559: if (error != 0) {
! 560: FILE_UNUSE(fp, l);
! 561: fdremove(fdp, fd);
! 562: ffree(fp);
! 563: } else {
! 564: if (sop != NULL)
! 565: *sop = so;
! 566: fp->f_data = so;
! 567: FILE_SET_MATURE(fp);
! 568: FILE_UNUSE(fp, l);
! 569: *fdout = fd;
! 570: }
! 571: return error;
! 572: }
! 573:
1.3 andrew 574: int
1.114 christos 575: sobind(struct socket *so, struct mbuf *nam, struct lwp *l)
1.1 cgd 576: {
1.54 lukem 577: int s, error;
1.1 cgd 578:
1.54 lukem 579: s = splsoftnet();
1.140 dyoung 580: error = (*so->so_proto->pr_usrreq)(so, PRU_BIND, NULL, nam, NULL, l);
1.1 cgd 581: splx(s);
1.140 dyoung 582: return error;
1.1 cgd 583: }
584:
1.3 andrew 585: int
1.54 lukem 586: solisten(struct socket *so, int backlog)
1.1 cgd 587: {
1.54 lukem 588: int s, error;
1.1 cgd 589:
1.54 lukem 590: s = splsoftnet();
1.140 dyoung 591: error = (*so->so_proto->pr_usrreq)(so, PRU_LISTEN, NULL,
592: NULL, NULL, NULL);
593: if (error != 0) {
1.1 cgd 594: splx(s);
1.140 dyoung 595: return error;
1.1 cgd 596: }
1.63 matt 597: if (TAILQ_EMPTY(&so->so_q))
1.1 cgd 598: so->so_options |= SO_ACCEPTCONN;
599: if (backlog < 0)
600: backlog = 0;
1.49 jonathan 601: so->so_qlimit = min(backlog, somaxconn);
1.1 cgd 602: splx(s);
1.140 dyoung 603: return 0;
1.1 cgd 604: }
605:
1.21 christos 606: void
1.54 lukem 607: sofree(struct socket *so)
1.1 cgd 608: {
609:
1.43 mycroft 610: if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0)
1.1 cgd 611: return;
1.43 mycroft 612: if (so->so_head) {
613: /*
614: * We must not decommission a socket that's on the accept(2)
615: * queue. If we do, then accept(2) may hang after select(2)
616: * indicated that the listening socket was ready.
617: */
618: if (!soqremque(so, 0))
619: return;
620: }
1.98 christos 621: if (so->so_rcv.sb_hiwat)
1.110 christos 622: (void)chgsbsize(so->so_uidinfo, &so->so_rcv.sb_hiwat, 0,
1.98 christos 623: RLIM_INFINITY);
624: if (so->so_snd.sb_hiwat)
1.110 christos 625: (void)chgsbsize(so->so_uidinfo, &so->so_snd.sb_hiwat, 0,
1.98 christos 626: RLIM_INFINITY);
627: sbrelease(&so->so_snd, so);
1.1 cgd 628: sorflush(so);
1.141.2.1! matt 629: seldestroy(&so->so_rcv.sb_sel);
! 630: seldestroy(&so->so_snd.sb_sel);
1.37 thorpej 631: pool_put(&socket_pool, so);
1.1 cgd 632: }
633:
634: /*
635: * Close a socket on last file table reference removal.
636: * Initiate disconnect if connected.
637: * Free socket when disconnect complete.
638: */
1.3 andrew 639: int
1.54 lukem 640: soclose(struct socket *so)
1.1 cgd 641: {
1.54 lukem 642: struct socket *so2;
643: int s, error;
1.1 cgd 644:
1.54 lukem 645: error = 0;
646: s = splsoftnet(); /* conservative */
1.1 cgd 647: if (so->so_options & SO_ACCEPTCONN) {
1.63 matt 648: while ((so2 = TAILQ_FIRST(&so->so_q0)) != 0) {
1.42 mycroft 649: (void) soqremque(so2, 0);
1.41 mycroft 650: (void) soabort(so2);
651: }
1.63 matt 652: while ((so2 = TAILQ_FIRST(&so->so_q)) != 0) {
1.42 mycroft 653: (void) soqremque(so2, 1);
1.41 mycroft 654: (void) soabort(so2);
655: }
1.1 cgd 656: }
657: if (so->so_pcb == 0)
658: goto discard;
659: if (so->so_state & SS_ISCONNECTED) {
660: if ((so->so_state & SS_ISDISCONNECTING) == 0) {
661: error = sodisconnect(so);
662: if (error)
663: goto drop;
664: }
665: if (so->so_options & SO_LINGER) {
666: if ((so->so_state & SS_ISDISCONNECTING) &&
667: (so->so_state & SS_NBIO))
668: goto drop;
1.21 christos 669: while (so->so_state & SS_ISCONNECTED) {
1.134 christos 670: error = tsleep((void *)&so->so_timeo,
1.21 christos 671: PSOCK | PCATCH, netcls,
1.30 thorpej 672: so->so_linger * hz);
1.21 christos 673: if (error)
1.1 cgd 674: break;
1.21 christos 675: }
1.1 cgd 676: }
677: }
1.54 lukem 678: drop:
1.1 cgd 679: if (so->so_pcb) {
1.22 mycroft 680: int error2 = (*so->so_proto->pr_usrreq)(so, PRU_DETACH,
1.140 dyoung 681: NULL, NULL, NULL, NULL);
1.1 cgd 682: if (error == 0)
683: error = error2;
684: }
1.54 lukem 685: discard:
1.1 cgd 686: if (so->so_state & SS_NOFDREF)
687: panic("soclose: NOFDREF");
688: so->so_state |= SS_NOFDREF;
689: sofree(so);
690: splx(s);
691: return (error);
692: }
693:
694: /*
1.20 mycroft 695: * Must be called at splsoftnet...
1.1 cgd 696: */
1.3 andrew 697: int
1.54 lukem 698: soabort(struct socket *so)
1.1 cgd 699: {
1.139 yamt 700: int error;
1.1 cgd 701:
1.139 yamt 702: KASSERT(so->so_head == NULL);
1.140 dyoung 703: error = (*so->so_proto->pr_usrreq)(so, PRU_ABORT, NULL,
704: NULL, NULL, NULL);
1.139 yamt 705: if (error) {
706: sofree(so);
707: }
708: return error;
1.1 cgd 709: }
710:
1.3 andrew 711: int
1.54 lukem 712: soaccept(struct socket *so, struct mbuf *nam)
1.1 cgd 713: {
1.54 lukem 714: int s, error;
1.1 cgd 715:
1.54 lukem 716: error = 0;
717: s = splsoftnet();
1.1 cgd 718: if ((so->so_state & SS_NOFDREF) == 0)
719: panic("soaccept: !NOFDREF");
720: so->so_state &= ~SS_NOFDREF;
1.55 thorpej 721: if ((so->so_state & SS_ISDISCONNECTED) == 0 ||
722: (so->so_proto->pr_flags & PR_ABRTACPTDIS) == 0)
1.41 mycroft 723: error = (*so->so_proto->pr_usrreq)(so, PRU_ACCEPT,
1.140 dyoung 724: NULL, nam, NULL, NULL);
1.41 mycroft 725: else
1.53 itojun 726: error = ECONNABORTED;
1.52 itojun 727:
1.1 cgd 728: splx(s);
729: return (error);
730: }
731:
1.3 andrew 732: int
1.114 christos 733: soconnect(struct socket *so, struct mbuf *nam, struct lwp *l)
1.1 cgd 734: {
1.54 lukem 735: int s, error;
1.1 cgd 736:
737: if (so->so_options & SO_ACCEPTCONN)
738: return (EOPNOTSUPP);
1.20 mycroft 739: s = splsoftnet();
1.1 cgd 740: /*
741: * If protocol is connection-based, can only connect once.
742: * Otherwise, if connected, try to disconnect first.
743: * This allows user to disconnect by connecting to, e.g.,
744: * a null address.
745: */
746: if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) &&
747: ((so->so_proto->pr_flags & PR_CONNREQUIRED) ||
748: (error = sodisconnect(so))))
749: error = EISCONN;
750: else
751: error = (*so->so_proto->pr_usrreq)(so, PRU_CONNECT,
1.140 dyoung 752: NULL, nam, NULL, l);
1.1 cgd 753: splx(s);
754: return (error);
755: }
756:
1.3 andrew 757: int
1.54 lukem 758: soconnect2(struct socket *so1, struct socket *so2)
1.1 cgd 759: {
1.54 lukem 760: int s, error;
1.1 cgd 761:
1.54 lukem 762: s = splsoftnet();
1.22 mycroft 763: error = (*so1->so_proto->pr_usrreq)(so1, PRU_CONNECT2,
1.140 dyoung 764: NULL, (struct mbuf *)so2, NULL, NULL);
1.1 cgd 765: splx(s);
766: return (error);
767: }
768:
1.3 andrew 769: int
1.54 lukem 770: sodisconnect(struct socket *so)
1.1 cgd 771: {
1.54 lukem 772: int s, error;
1.1 cgd 773:
1.54 lukem 774: s = splsoftnet();
1.1 cgd 775: if ((so->so_state & SS_ISCONNECTED) == 0) {
776: error = ENOTCONN;
777: goto bad;
778: }
779: if (so->so_state & SS_ISDISCONNECTING) {
780: error = EALREADY;
781: goto bad;
782: }
1.22 mycroft 783: error = (*so->so_proto->pr_usrreq)(so, PRU_DISCONNECT,
1.140 dyoung 784: NULL, NULL, NULL, NULL);
1.54 lukem 785: bad:
1.1 cgd 786: splx(s);
1.117 yamt 787: sodopendfree();
1.1 cgd 788: return (error);
789: }
790:
1.15 mycroft 791: #define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK)
1.1 cgd 792: /*
793: * Send on a socket.
794: * If send must go all at once and message is larger than
795: * send buffering, then hard error.
796: * Lock against other senders.
797: * If must go all at once and not enough room now, then
798: * inform user that this would block and do nothing.
799: * Otherwise, if nonblocking, send as much as possible.
800: * The data to be sent is described by "uio" if nonzero,
801: * otherwise by the mbuf chain "top" (which must be null
802: * if uio is not). Data provided in mbuf chain must be small
803: * enough to send all at once.
804: *
805: * Returns nonzero on error, timeout or signal; callers
806: * must check for short counts if EINTR/ERESTART are returned.
807: * Data and control buffers are freed on return.
808: */
1.3 andrew 809: int
1.54 lukem 810: sosend(struct socket *so, struct mbuf *addr, struct uio *uio, struct mbuf *top,
1.114 christos 811: struct mbuf *control, int flags, struct lwp *l)
1.1 cgd 812: {
1.54 lukem 813: struct mbuf **mp, *m;
1.114 christos 814: struct proc *p;
1.58 jdolecek 815: long space, len, resid, clen, mlen;
816: int error, s, dontroute, atomic;
1.54 lukem 817:
1.114 christos 818: p = l->l_proc;
1.117 yamt 819: sodopendfree();
1.64 thorpej 820:
1.54 lukem 821: clen = 0;
822: atomic = sosendallatonce(so) || top;
1.1 cgd 823: if (uio)
824: resid = uio->uio_resid;
825: else
826: resid = top->m_pkthdr.len;
1.7 cgd 827: /*
828: * In theory resid should be unsigned.
829: * However, space must be signed, as it might be less than 0
830: * if we over-committed, and we must use a signed comparison
831: * of space and resid. On the other hand, a negative resid
832: * causes us to loop sending 0-length segments to the protocol.
833: */
1.29 mycroft 834: if (resid < 0) {
835: error = EINVAL;
836: goto out;
837: }
1.1 cgd 838: dontroute =
839: (flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 &&
840: (so->so_proto->pr_flags & PR_ATOMIC);
1.102 jonathan 841: if (p)
842: p->p_stats->p_ru.ru_msgsnd++;
1.1 cgd 843: if (control)
844: clen = control->m_len;
845: #define snderr(errno) { error = errno; splx(s); goto release; }
846:
1.54 lukem 847: restart:
1.21 christos 848: if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0)
1.1 cgd 849: goto out;
850: do {
1.20 mycroft 851: s = splsoftnet();
1.1 cgd 852: if (so->so_state & SS_CANTSENDMORE)
853: snderr(EPIPE);
1.48 thorpej 854: if (so->so_error) {
855: error = so->so_error;
856: so->so_error = 0;
857: splx(s);
858: goto release;
859: }
1.1 cgd 860: if ((so->so_state & SS_ISCONNECTED) == 0) {
861: if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
862: if ((so->so_state & SS_ISCONFIRMING) == 0 &&
863: !(resid == 0 && clen != 0))
864: snderr(ENOTCONN);
865: } else if (addr == 0)
866: snderr(EDESTADDRREQ);
867: }
868: space = sbspace(&so->so_snd);
869: if (flags & MSG_OOB)
870: space += 1024;
1.21 christos 871: if ((atomic && resid > so->so_snd.sb_hiwat) ||
1.11 mycroft 872: clen > so->so_snd.sb_hiwat)
873: snderr(EMSGSIZE);
1.96 mycroft 874: if (space < resid + clen &&
1.1 cgd 875: (atomic || space < so->so_snd.sb_lowat || space < clen)) {
876: if (so->so_state & SS_NBIO)
877: snderr(EWOULDBLOCK);
878: sbunlock(&so->so_snd);
879: error = sbwait(&so->so_snd);
880: splx(s);
881: if (error)
882: goto out;
883: goto restart;
884: }
885: splx(s);
886: mp = ⊤
887: space -= clen;
888: do {
1.45 tv 889: if (uio == NULL) {
890: /*
891: * Data is prepackaged in "top".
892: */
893: resid = 0;
894: if (flags & MSG_EOR)
895: top->m_flags |= M_EOR;
896: } else do {
1.141.2.1! matt 897: if (top == NULL) {
1.78 matt 898: m = m_gethdr(M_WAIT, MT_DATA);
1.45 tv 899: mlen = MHLEN;
900: m->m_pkthdr.len = 0;
1.140 dyoung 901: m->m_pkthdr.rcvif = NULL;
1.45 tv 902: } else {
1.78 matt 903: m = m_get(M_WAIT, MT_DATA);
1.45 tv 904: mlen = MLEN;
905: }
1.78 matt 906: MCLAIM(m, so->so_snd.sb_mowner);
1.121 yamt 907: if (sock_loan_thresh >= 0 &&
908: uio->uio_iov->iov_len >= sock_loan_thresh &&
909: space >= sock_loan_thresh &&
1.64 thorpej 910: (len = sosend_loan(so, uio, m,
911: space)) != 0) {
912: SOSEND_COUNTER_INCR(&sosend_loan_big);
913: space -= len;
914: goto have_data;
915: }
1.45 tv 916: if (resid >= MINCLSIZE && space >= MCLBYTES) {
1.64 thorpej 917: SOSEND_COUNTER_INCR(&sosend_copy_big);
1.78 matt 918: m_clget(m, M_WAIT);
1.45 tv 919: if ((m->m_flags & M_EXT) == 0)
920: goto nopages;
921: mlen = MCLBYTES;
922: if (atomic && top == 0) {
1.58 jdolecek 923: len = lmin(MCLBYTES - max_hdr,
1.54 lukem 924: resid);
1.45 tv 925: m->m_data += max_hdr;
926: } else
1.58 jdolecek 927: len = lmin(MCLBYTES, resid);
1.45 tv 928: space -= len;
929: } else {
1.64 thorpej 930: nopages:
931: SOSEND_COUNTER_INCR(&sosend_copy_small);
1.58 jdolecek 932: len = lmin(lmin(mlen, resid), space);
1.45 tv 933: space -= len;
934: /*
935: * For datagram protocols, leave room
936: * for protocol headers in first mbuf.
937: */
938: if (atomic && top == 0 && len < mlen)
939: MH_ALIGN(m, len);
940: }
1.141.2.1! matt 941: error = uiomove(mtod(m, void *), (int)len, uio);
1.64 thorpej 942: have_data:
1.45 tv 943: resid = uio->uio_resid;
944: m->m_len = len;
945: *mp = m;
946: top->m_pkthdr.len += len;
1.141.2.1! matt 947: if (error != 0)
1.45 tv 948: goto release;
949: mp = &m->m_next;
950: if (resid <= 0) {
951: if (flags & MSG_EOR)
952: top->m_flags |= M_EOR;
953: break;
954: }
955: } while (space > 0 && atomic);
1.108 perry 956:
1.46 sommerfe 957: s = splsoftnet();
958:
959: if (so->so_state & SS_CANTSENDMORE)
960: snderr(EPIPE);
1.45 tv 961:
962: if (dontroute)
963: so->so_options |= SO_DONTROUTE;
964: if (resid > 0)
965: so->so_state |= SS_MORETOCOME;
1.46 sommerfe 966: error = (*so->so_proto->pr_usrreq)(so,
967: (flags & MSG_OOB) ? PRU_SENDOOB : PRU_SEND,
1.114 christos 968: top, addr, control, curlwp); /* XXX */
1.45 tv 969: if (dontroute)
970: so->so_options &= ~SO_DONTROUTE;
971: if (resid > 0)
972: so->so_state &= ~SS_MORETOCOME;
1.46 sommerfe 973: splx(s);
974:
1.45 tv 975: clen = 0;
1.141.2.1! matt 976: control = NULL;
! 977: top = NULL;
1.45 tv 978: mp = ⊤
1.141.2.1! matt 979: if (error != 0)
1.1 cgd 980: goto release;
981: } while (resid && space > 0);
982: } while (resid);
983:
1.54 lukem 984: release:
1.1 cgd 985: sbunlock(&so->so_snd);
1.54 lukem 986: out:
1.1 cgd 987: if (top)
988: m_freem(top);
989: if (control)
990: m_freem(control);
991: return (error);
992: }
993:
994: /*
995: * Implement receive operations on a socket.
996: * We depend on the way that records are added to the sockbuf
997: * by sbappend*. In particular, each record (mbufs linked through m_next)
998: * must begin with an address if the protocol so specifies,
999: * followed by an optional mbuf or mbufs containing ancillary data,
1000: * and then zero or more mbufs of data.
1001: * In order to avoid blocking network interrupts for the entire time here,
1002: * we splx() while doing the actual copy to user space.
1003: * Although the sockbuf is locked, new data may still be appended,
1004: * and thus we must maintain consistency of the sockbuf during that time.
1005: *
1006: * The caller may receive the data as a single mbuf chain by supplying
1007: * an mbuf **mp0 for use in returning the chain. The uio is then used
1008: * only for the count in uio_resid.
1009: */
1.3 andrew 1010: int
1.54 lukem 1011: soreceive(struct socket *so, struct mbuf **paddr, struct uio *uio,
1012: struct mbuf **mp0, struct mbuf **controlp, int *flagsp)
1.1 cgd 1013: {
1.116 yamt 1014: struct lwp *l = curlwp;
1.54 lukem 1015: struct mbuf *m, **mp;
1016: int flags, len, error, s, offset, moff, type, orig_resid;
1.99 matt 1017: const struct protosw *pr;
1.54 lukem 1018: struct mbuf *nextrecord;
1.67 he 1019: int mbuf_removed = 0;
1.64 thorpej 1020:
1.54 lukem 1021: pr = so->so_proto;
1.1 cgd 1022: mp = mp0;
1.54 lukem 1023: type = 0;
1024: orig_resid = uio->uio_resid;
1.102 jonathan 1025:
1.141.2.1! matt 1026: if (paddr != NULL)
! 1027: *paddr = NULL;
! 1028: if (controlp != NULL)
! 1029: *controlp = NULL;
! 1030: if (flagsp != NULL)
1.1 cgd 1031: flags = *flagsp &~ MSG_EOR;
1032: else
1033: flags = 0;
1.66 enami 1034:
1035: if ((flags & MSG_DONTWAIT) == 0)
1.117 yamt 1036: sodopendfree();
1.66 enami 1037:
1.1 cgd 1038: if (flags & MSG_OOB) {
1039: m = m_get(M_WAIT, MT_DATA);
1.17 cgd 1040: error = (*pr->pr_usrreq)(so, PRU_RCVOOB, m,
1.140 dyoung 1041: (struct mbuf *)(long)(flags & MSG_PEEK), NULL, l);
1.1 cgd 1042: if (error)
1043: goto bad;
1044: do {
1.134 christos 1045: error = uiomove(mtod(m, void *),
1.1 cgd 1046: (int) min(uio->uio_resid, m->m_len), uio);
1047: m = m_free(m);
1.141.2.1! matt 1048: } while (uio->uio_resid > 0 && error == 0 && m);
1.54 lukem 1049: bad:
1.141.2.1! matt 1050: if (m != NULL)
1.1 cgd 1051: m_freem(m);
1.141.2.1! matt 1052: return error;
1.1 cgd 1053: }
1.141.2.1! matt 1054: if (mp != NULL)
1.140 dyoung 1055: *mp = NULL;
1.1 cgd 1056: if (so->so_state & SS_ISCONFIRMING && uio->uio_resid)
1.140 dyoung 1057: (*pr->pr_usrreq)(so, PRU_RCVD, NULL, NULL, NULL, l);
1.1 cgd 1058:
1.54 lukem 1059: restart:
1.21 christos 1060: if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0)
1.141.2.1! matt 1061: return error;
1.20 mycroft 1062: s = splsoftnet();
1.1 cgd 1063:
1064: m = so->so_rcv.sb_mb;
1065: /*
1066: * If we have less data than requested, block awaiting more
1067: * (subject to any timeout) if:
1.15 mycroft 1068: * 1. the current count is less than the low water mark,
1.1 cgd 1069: * 2. MSG_WAITALL is set, and it is possible to do the entire
1.15 mycroft 1070: * receive operation at once if we block (resid <= hiwat), or
1071: * 3. MSG_DONTWAIT is not set.
1.1 cgd 1072: * If MSG_WAITALL is set but resid is larger than the receive buffer,
1073: * we have to do the receive in sections, and thus risk returning
1074: * a short count if a timeout or signal occurs after we start.
1075: */
1.141.2.1! matt 1076: if (m == NULL ||
! 1077: ((flags & MSG_DONTWAIT) == 0 &&
! 1078: so->so_rcv.sb_cc < uio->uio_resid &&
! 1079: (so->so_rcv.sb_cc < so->so_rcv.sb_lowat ||
! 1080: ((flags & MSG_WAITALL) &&
! 1081: uio->uio_resid <= so->so_rcv.sb_hiwat)) &&
! 1082: m->m_nextpkt == NULL &&
! 1083: (pr->pr_flags & PR_ATOMIC) == 0)) {
1.1 cgd 1084: #ifdef DIAGNOSTIC
1.141.2.1! matt 1085: if (m == NULL && so->so_rcv.sb_cc)
1.1 cgd 1086: panic("receive 1");
1087: #endif
1088: if (so->so_error) {
1.141.2.1! matt 1089: if (m != NULL)
1.15 mycroft 1090: goto dontblock;
1.1 cgd 1091: error = so->so_error;
1092: if ((flags & MSG_PEEK) == 0)
1093: so->so_error = 0;
1094: goto release;
1095: }
1096: if (so->so_state & SS_CANTRCVMORE) {
1.141.2.1! matt 1097: if (m != NULL)
1.15 mycroft 1098: goto dontblock;
1.1 cgd 1099: else
1100: goto release;
1101: }
1.141.2.1! matt 1102: for (; m != NULL; m = m->m_next)
1.1 cgd 1103: if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) {
1104: m = so->so_rcv.sb_mb;
1105: goto dontblock;
1106: }
1107: if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
1108: (so->so_proto->pr_flags & PR_CONNREQUIRED)) {
1109: error = ENOTCONN;
1110: goto release;
1111: }
1112: if (uio->uio_resid == 0)
1113: goto release;
1.15 mycroft 1114: if ((so->so_state & SS_NBIO) || (flags & MSG_DONTWAIT)) {
1.1 cgd 1115: error = EWOULDBLOCK;
1116: goto release;
1117: }
1.69 thorpej 1118: SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 1");
1119: SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 1");
1.1 cgd 1120: sbunlock(&so->so_rcv);
1121: error = sbwait(&so->so_rcv);
1122: splx(s);
1.141.2.1! matt 1123: if (error != 0)
! 1124: return error;
1.1 cgd 1125: goto restart;
1126: }
1.54 lukem 1127: dontblock:
1.69 thorpej 1128: /*
1129: * On entry here, m points to the first record of the socket buffer.
1130: * While we process the initial mbufs containing address and control
1131: * info, we save a copy of m->m_nextpkt into nextrecord.
1132: */
1.141.2.1! matt 1133: if (l != NULL)
1.114 christos 1134: l->l_proc->p_stats->p_ru.ru_msgrcv++;
1.69 thorpej 1135: KASSERT(m == so->so_rcv.sb_mb);
1136: SBLASTRECORDCHK(&so->so_rcv, "soreceive 1");
1137: SBLASTMBUFCHK(&so->so_rcv, "soreceive 1");
1.1 cgd 1138: nextrecord = m->m_nextpkt;
1139: if (pr->pr_flags & PR_ADDR) {
1140: #ifdef DIAGNOSTIC
1141: if (m->m_type != MT_SONAME)
1142: panic("receive 1a");
1143: #endif
1.3 andrew 1144: orig_resid = 0;
1.1 cgd 1145: if (flags & MSG_PEEK) {
1146: if (paddr)
1147: *paddr = m_copy(m, 0, m->m_len);
1148: m = m->m_next;
1149: } else {
1150: sbfree(&so->so_rcv, m);
1.67 he 1151: mbuf_removed = 1;
1.141.2.1! matt 1152: if (paddr != NULL) {
1.1 cgd 1153: *paddr = m;
1154: so->so_rcv.sb_mb = m->m_next;
1.141.2.1! matt 1155: m->m_next = NULL;
1.1 cgd 1156: m = so->so_rcv.sb_mb;
1157: } else {
1158: MFREE(m, so->so_rcv.sb_mb);
1159: m = so->so_rcv.sb_mb;
1160: }
1161: }
1162: }
1.141.2.1! matt 1163: while (m != NULL && m->m_type == MT_CONTROL && error == 0) {
1.1 cgd 1164: if (flags & MSG_PEEK) {
1.141.2.1! matt 1165: if (controlp != NULL)
1.1 cgd 1166: *controlp = m_copy(m, 0, m->m_len);
1167: m = m->m_next;
1168: } else {
1169: sbfree(&so->so_rcv, m);
1.67 he 1170: mbuf_removed = 1;
1.141.2.1! matt 1171: if (controlp != NULL) {
1.102 jonathan 1172: struct domain *dom = pr->pr_domain;
1.114 christos 1173: if (dom->dom_externalize && l &&
1.1 cgd 1174: mtod(m, struct cmsghdr *)->cmsg_type ==
1175: SCM_RIGHTS)
1.114 christos 1176: error = (*dom->dom_externalize)(m, l);
1.1 cgd 1177: *controlp = m;
1178: so->so_rcv.sb_mb = m->m_next;
1.141.2.1! matt 1179: m->m_next = NULL;
1.1 cgd 1180: m = so->so_rcv.sb_mb;
1181: } else {
1.106 itojun 1182: /*
1183: * Dispose of any SCM_RIGHTS message that went
1184: * through the read path rather than recv.
1185: */
1186: if (pr->pr_domain->dom_dispose &&
1187: mtod(m, struct cmsghdr *)->cmsg_type == SCM_RIGHTS)
1188: (*pr->pr_domain->dom_dispose)(m);
1.1 cgd 1189: MFREE(m, so->so_rcv.sb_mb);
1190: m = so->so_rcv.sb_mb;
1191: }
1192: }
1.141.2.1! matt 1193: if (controlp != NULL) {
1.3 andrew 1194: orig_resid = 0;
1.1 cgd 1195: controlp = &(*controlp)->m_next;
1.3 andrew 1196: }
1.1 cgd 1197: }
1.69 thorpej 1198:
1199: /*
1200: * If m is non-NULL, we have some data to read. From now on,
1201: * make sure to keep sb_lastrecord consistent when working on
1202: * the last packet on the chain (nextrecord == NULL) and we
1203: * change m->m_nextpkt.
1204: */
1.141.2.1! matt 1205: if (m != NULL) {
1.69 thorpej 1206: if ((flags & MSG_PEEK) == 0) {
1.1 cgd 1207: m->m_nextpkt = nextrecord;
1.69 thorpej 1208: /*
1209: * If nextrecord == NULL (this is a single chain),
1210: * then sb_lastrecord may not be valid here if m
1211: * was changed earlier.
1212: */
1213: if (nextrecord == NULL) {
1214: KASSERT(so->so_rcv.sb_mb == m);
1215: so->so_rcv.sb_lastrecord = m;
1216: }
1217: }
1.1 cgd 1218: type = m->m_type;
1219: if (type == MT_OOBDATA)
1220: flags |= MSG_OOB;
1.69 thorpej 1221: } else {
1222: if ((flags & MSG_PEEK) == 0) {
1223: KASSERT(so->so_rcv.sb_mb == m);
1224: so->so_rcv.sb_mb = nextrecord;
1.70 thorpej 1225: SB_EMPTY_FIXUP(&so->so_rcv);
1.69 thorpej 1226: }
1.1 cgd 1227: }
1.69 thorpej 1228: SBLASTRECORDCHK(&so->so_rcv, "soreceive 2");
1229: SBLASTMBUFCHK(&so->so_rcv, "soreceive 2");
1230:
1.1 cgd 1231: moff = 0;
1232: offset = 0;
1.141.2.1! matt 1233: while (m != NULL && uio->uio_resid > 0 && error == 0) {
1.1 cgd 1234: if (m->m_type == MT_OOBDATA) {
1235: if (type != MT_OOBDATA)
1236: break;
1237: } else if (type == MT_OOBDATA)
1238: break;
1239: #ifdef DIAGNOSTIC
1240: else if (m->m_type != MT_DATA && m->m_type != MT_HEADER)
1241: panic("receive 3");
1242: #endif
1243: so->so_state &= ~SS_RCVATMARK;
1244: len = uio->uio_resid;
1245: if (so->so_oobmark && len > so->so_oobmark - offset)
1246: len = so->so_oobmark - offset;
1247: if (len > m->m_len - moff)
1248: len = m->m_len - moff;
1249: /*
1250: * If mp is set, just pass back the mbufs.
1251: * Otherwise copy them out via the uio, then free.
1252: * Sockbuf must be consistent here (points to current mbuf,
1253: * it points to next record) when we drop priority;
1254: * we must note any additions to the sockbuf when we
1255: * block interrupts again.
1256: */
1.141.2.1! matt 1257: if (mp == NULL) {
1.69 thorpej 1258: SBLASTRECORDCHK(&so->so_rcv, "soreceive uiomove");
1259: SBLASTMBUFCHK(&so->so_rcv, "soreceive uiomove");
1.1 cgd 1260: splx(s);
1.134 christos 1261: error = uiomove(mtod(m, char *) + moff, (int)len, uio);
1.20 mycroft 1262: s = splsoftnet();
1.141.2.1! matt 1263: if (error != 0) {
1.67 he 1264: /*
1265: * If any part of the record has been removed
1266: * (such as the MT_SONAME mbuf, which will
1267: * happen when PR_ADDR, and thus also
1268: * PR_ATOMIC, is set), then drop the entire
1269: * record to maintain the atomicity of the
1270: * receive operation.
1271: *
1272: * This avoids a later panic("receive 1a")
1273: * when compiled with DIAGNOSTIC.
1274: */
1275: if (m && mbuf_removed
1276: && (pr->pr_flags & PR_ATOMIC))
1277: (void) sbdroprecord(&so->so_rcv);
1278:
1.57 jdolecek 1279: goto release;
1.67 he 1280: }
1.1 cgd 1281: } else
1282: uio->uio_resid -= len;
1283: if (len == m->m_len - moff) {
1284: if (m->m_flags & M_EOR)
1285: flags |= MSG_EOR;
1286: if (flags & MSG_PEEK) {
1287: m = m->m_next;
1288: moff = 0;
1289: } else {
1290: nextrecord = m->m_nextpkt;
1291: sbfree(&so->so_rcv, m);
1292: if (mp) {
1293: *mp = m;
1294: mp = &m->m_next;
1295: so->so_rcv.sb_mb = m = m->m_next;
1.140 dyoung 1296: *mp = NULL;
1.1 cgd 1297: } else {
1298: MFREE(m, so->so_rcv.sb_mb);
1299: m = so->so_rcv.sb_mb;
1300: }
1.69 thorpej 1301: /*
1302: * If m != NULL, we also know that
1303: * so->so_rcv.sb_mb != NULL.
1304: */
1305: KASSERT(so->so_rcv.sb_mb == m);
1306: if (m) {
1.1 cgd 1307: m->m_nextpkt = nextrecord;
1.69 thorpej 1308: if (nextrecord == NULL)
1309: so->so_rcv.sb_lastrecord = m;
1310: } else {
1311: so->so_rcv.sb_mb = nextrecord;
1.70 thorpej 1312: SB_EMPTY_FIXUP(&so->so_rcv);
1.69 thorpej 1313: }
1314: SBLASTRECORDCHK(&so->so_rcv, "soreceive 3");
1315: SBLASTMBUFCHK(&so->so_rcv, "soreceive 3");
1.1 cgd 1316: }
1.141.2.1! matt 1317: } else if (flags & MSG_PEEK)
! 1318: moff += len;
! 1319: else {
! 1320: if (mp != NULL)
! 1321: *mp = m_copym(m, 0, len, M_WAIT);
! 1322: m->m_data += len;
! 1323: m->m_len -= len;
! 1324: so->so_rcv.sb_cc -= len;
1.1 cgd 1325: }
1326: if (so->so_oobmark) {
1327: if ((flags & MSG_PEEK) == 0) {
1328: so->so_oobmark -= len;
1329: if (so->so_oobmark == 0) {
1330: so->so_state |= SS_RCVATMARK;
1331: break;
1332: }
1.7 cgd 1333: } else {
1.1 cgd 1334: offset += len;
1.7 cgd 1335: if (offset == so->so_oobmark)
1336: break;
1337: }
1.1 cgd 1338: }
1339: if (flags & MSG_EOR)
1340: break;
1341: /*
1342: * If the MSG_WAITALL flag is set (for non-atomic socket),
1343: * we must not quit until "uio->uio_resid == 0" or an error
1344: * termination. If a signal/timeout occurs, return
1345: * with a short count but without error.
1346: * Keep sockbuf locked against other readers.
1347: */
1.141.2.1! matt 1348: while (flags & MSG_WAITALL && m == NULL && uio->uio_resid > 0 &&
1.3 andrew 1349: !sosendallatonce(so) && !nextrecord) {
1.1 cgd 1350: if (so->so_error || so->so_state & SS_CANTRCVMORE)
1351: break;
1.68 matt 1352: /*
1353: * If we are peeking and the socket receive buffer is
1354: * full, stop since we can't get more data to peek at.
1355: */
1356: if ((flags & MSG_PEEK) && sbspace(&so->so_rcv) <= 0)
1357: break;
1358: /*
1359: * If we've drained the socket buffer, tell the
1360: * protocol in case it needs to do something to
1361: * get it filled again.
1362: */
1363: if ((pr->pr_flags & PR_WANTRCVD) && so->so_pcb)
1364: (*pr->pr_usrreq)(so, PRU_RCVD,
1.140 dyoung 1365: NULL, (struct mbuf *)(long)flags, NULL, l);
1.69 thorpej 1366: SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 2");
1367: SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 2");
1.1 cgd 1368: error = sbwait(&so->so_rcv);
1.141.2.1! matt 1369: if (error != 0) {
1.1 cgd 1370: sbunlock(&so->so_rcv);
1371: splx(s);
1.141.2.1! matt 1372: return 0;
1.1 cgd 1373: }
1.21 christos 1374: if ((m = so->so_rcv.sb_mb) != NULL)
1.1 cgd 1375: nextrecord = m->m_nextpkt;
1376: }
1377: }
1.3 andrew 1378:
1379: if (m && pr->pr_flags & PR_ATOMIC) {
1380: flags |= MSG_TRUNC;
1381: if ((flags & MSG_PEEK) == 0)
1382: (void) sbdroprecord(&so->so_rcv);
1383: }
1.1 cgd 1384: if ((flags & MSG_PEEK) == 0) {
1.141.2.1! matt 1385: if (m == NULL) {
1.69 thorpej 1386: /*
1.70 thorpej 1387: * First part is an inline SB_EMPTY_FIXUP(). Second
1.69 thorpej 1388: * part makes sure sb_lastrecord is up-to-date if
1389: * there is still data in the socket buffer.
1390: */
1.1 cgd 1391: so->so_rcv.sb_mb = nextrecord;
1.69 thorpej 1392: if (so->so_rcv.sb_mb == NULL) {
1393: so->so_rcv.sb_mbtail = NULL;
1394: so->so_rcv.sb_lastrecord = NULL;
1395: } else if (nextrecord->m_nextpkt == NULL)
1396: so->so_rcv.sb_lastrecord = nextrecord;
1397: }
1398: SBLASTRECORDCHK(&so->so_rcv, "soreceive 4");
1399: SBLASTMBUFCHK(&so->so_rcv, "soreceive 4");
1.1 cgd 1400: if (pr->pr_flags & PR_WANTRCVD && so->so_pcb)
1.140 dyoung 1401: (*pr->pr_usrreq)(so, PRU_RCVD, NULL,
1402: (struct mbuf *)(long)flags, NULL, l);
1.1 cgd 1403: }
1.3 andrew 1404: if (orig_resid == uio->uio_resid && orig_resid &&
1405: (flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) {
1406: sbunlock(&so->so_rcv);
1407: splx(s);
1408: goto restart;
1409: }
1.108 perry 1410:
1.141.2.1! matt 1411: if (flagsp != NULL)
1.1 cgd 1412: *flagsp |= flags;
1.54 lukem 1413: release:
1.1 cgd 1414: sbunlock(&so->so_rcv);
1415: splx(s);
1.141.2.1! matt 1416: return error;
1.1 cgd 1417: }
1418:
1.14 mycroft 1419: int
1.54 lukem 1420: soshutdown(struct socket *so, int how)
1.1 cgd 1421: {
1.99 matt 1422: const struct protosw *pr;
1.34 kleink 1423:
1.54 lukem 1424: pr = so->so_proto;
1.34 kleink 1425: if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR))
1426: return (EINVAL);
1.1 cgd 1427:
1.34 kleink 1428: if (how == SHUT_RD || how == SHUT_RDWR)
1.1 cgd 1429: sorflush(so);
1.34 kleink 1430: if (how == SHUT_WR || how == SHUT_RDWR)
1.140 dyoung 1431: return (*pr->pr_usrreq)(so, PRU_SHUTDOWN, NULL,
1432: NULL, NULL, NULL);
1.141.2.1! matt 1433: return 0;
1.1 cgd 1434: }
1435:
1.14 mycroft 1436: void
1.54 lukem 1437: sorflush(struct socket *so)
1.1 cgd 1438: {
1.54 lukem 1439: struct sockbuf *sb, asb;
1.99 matt 1440: const struct protosw *pr;
1.54 lukem 1441: int s;
1.1 cgd 1442:
1.54 lukem 1443: sb = &so->so_rcv;
1444: pr = so->so_proto;
1.1 cgd 1445: sb->sb_flags |= SB_NOINTR;
1.15 mycroft 1446: (void) sblock(sb, M_WAITOK);
1.56 thorpej 1447: s = splnet();
1.1 cgd 1448: socantrcvmore(so);
1449: sbunlock(sb);
1450: asb = *sb;
1.86 wrstuden 1451: /*
1452: * Clear most of the sockbuf structure, but leave some of the
1453: * fields valid.
1454: */
1455: memset(&sb->sb_startzero, 0,
1456: sizeof(*sb) - offsetof(struct sockbuf, sb_startzero));
1.1 cgd 1457: splx(s);
1458: if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose)
1459: (*pr->pr_domain->dom_dispose)(asb.sb_mb);
1.98 christos 1460: sbrelease(&asb, so);
1.1 cgd 1461: }
1462:
1.141.2.1! matt 1463: static int
! 1464: sosetopt1(struct socket *so, int level, int optname, struct mbuf *m)
1.1 cgd 1465: {
1.141.2.1! matt 1466: int optval, val;
1.130 christos 1467: struct linger *l;
1.141 yamt 1468: struct sockbuf *sb;
1.141.2.1! matt 1469: struct timeval *tv;
1.1 cgd 1470:
1.141.2.1! matt 1471: switch (optname) {
1.1 cgd 1472:
1.141.2.1! matt 1473: case SO_LINGER:
! 1474: if (m == NULL || m->m_len != sizeof(struct linger))
! 1475: return EINVAL;
! 1476: l = mtod(m, struct linger *);
! 1477: if (l->l_linger < 0 || l->l_linger > USHRT_MAX ||
! 1478: l->l_linger > (INT_MAX / hz))
! 1479: return EDOM;
! 1480: so->so_linger = l->l_linger;
! 1481: if (l->l_onoff)
! 1482: so->so_options |= SO_LINGER;
! 1483: else
! 1484: so->so_options &= ~SO_LINGER;
! 1485: break;
1.1 cgd 1486:
1.141.2.1! matt 1487: case SO_DEBUG:
! 1488: case SO_KEEPALIVE:
! 1489: case SO_DONTROUTE:
! 1490: case SO_USELOOPBACK:
! 1491: case SO_BROADCAST:
! 1492: case SO_REUSEADDR:
! 1493: case SO_REUSEPORT:
! 1494: case SO_OOBINLINE:
! 1495: case SO_TIMESTAMP:
! 1496: if (m == NULL || m->m_len < sizeof(int))
! 1497: return EINVAL;
! 1498: if (*mtod(m, int *))
! 1499: so->so_options |= optname;
! 1500: else
! 1501: so->so_options &= ~optname;
! 1502: break;
! 1503:
! 1504: case SO_SNDBUF:
! 1505: case SO_RCVBUF:
! 1506: case SO_SNDLOWAT:
! 1507: case SO_RCVLOWAT:
! 1508: if (m == NULL || m->m_len < sizeof(int))
! 1509: return EINVAL;
! 1510:
! 1511: /*
! 1512: * Values < 1 make no sense for any of these
! 1513: * options, so disallow them.
! 1514: */
! 1515: optval = *mtod(m, int *);
! 1516: if (optval < 1)
! 1517: return EINVAL;
! 1518:
! 1519: switch (optname) {
1.1 cgd 1520:
1521: case SO_SNDBUF:
1522: case SO_RCVBUF:
1.141.2.1! matt 1523: sb = (optname == SO_SNDBUF) ?
! 1524: &so->so_snd : &so->so_rcv;
! 1525: if (sbreserve(sb, (u_long)optval, so) == 0)
! 1526: return ENOBUFS;
! 1527: sb->sb_flags &= ~SB_AUTOSIZE;
! 1528: break;
! 1529:
! 1530: /*
! 1531: * Make sure the low-water is never greater than
! 1532: * the high-water.
! 1533: */
1.1 cgd 1534: case SO_SNDLOWAT:
1.141.2.1! matt 1535: so->so_snd.sb_lowat =
! 1536: (optval > so->so_snd.sb_hiwat) ?
! 1537: so->so_snd.sb_hiwat : optval;
! 1538: break;
1.1 cgd 1539: case SO_RCVLOWAT:
1.141.2.1! matt 1540: so->so_rcv.sb_lowat =
! 1541: (optval > so->so_rcv.sb_hiwat) ?
! 1542: so->so_rcv.sb_hiwat : optval;
! 1543: break;
! 1544: }
! 1545: break;
1.28 thorpej 1546:
1.141.2.1! matt 1547: case SO_SNDTIMEO:
! 1548: case SO_RCVTIMEO:
! 1549: if (m == NULL || m->m_len < sizeof(*tv))
! 1550: return EINVAL;
! 1551: tv = mtod(m, struct timeval *);
! 1552: if (tv->tv_sec > (INT_MAX - tv->tv_usec / tick) / hz)
! 1553: return EDOM;
! 1554: val = tv->tv_sec * hz + tv->tv_usec / tick;
! 1555: if (val == 0 && tv->tv_usec != 0)
! 1556: val = 1;
1.1 cgd 1557:
1.141.2.1! matt 1558: switch (optname) {
1.1 cgd 1559:
1560: case SO_SNDTIMEO:
1.141.2.1! matt 1561: so->so_snd.sb_timeo = val;
! 1562: break;
1.1 cgd 1563: case SO_RCVTIMEO:
1.141.2.1! matt 1564: so->so_rcv.sb_timeo = val;
! 1565: break;
! 1566: }
! 1567: break;
1.1 cgd 1568:
1.141.2.1! matt 1569: default:
! 1570: return ENOPROTOOPT;
! 1571: }
! 1572: return 0;
! 1573: }
1.1 cgd 1574:
1.141.2.1! matt 1575: int
! 1576: sosetopt(struct socket *so, int level, int optname, struct mbuf *m)
! 1577: {
! 1578: int error, prerr;
1.1 cgd 1579:
1.141.2.1! matt 1580: if (level == SOL_SOCKET)
! 1581: error = sosetopt1(so, level, optname, m);
! 1582: else
! 1583: error = ENOPROTOOPT;
1.1 cgd 1584:
1.141.2.1! matt 1585: if ((error == 0 || error == ENOPROTOOPT) &&
! 1586: so->so_proto != NULL && so->so_proto->pr_ctloutput != NULL) {
! 1587: /* give the protocol stack a shot */
! 1588: prerr = (*so->so_proto->pr_ctloutput)(PRCO_SETOPT, so, level,
! 1589: optname, &m);
! 1590: if (prerr == 0)
! 1591: error = 0;
! 1592: else if (prerr != ENOPROTOOPT)
! 1593: error = prerr;
! 1594: } else if (m != NULL)
! 1595: (void)m_free(m);
! 1596: return error;
1.1 cgd 1597: }
1598:
1.14 mycroft 1599: int
1.54 lukem 1600: sogetopt(struct socket *so, int level, int optname, struct mbuf **mp)
1.1 cgd 1601: {
1.54 lukem 1602: struct mbuf *m;
1.1 cgd 1603:
1604: if (level != SOL_SOCKET) {
1605: if (so->so_proto && so->so_proto->pr_ctloutput) {
1606: return ((*so->so_proto->pr_ctloutput)
1607: (PRCO_GETOPT, so, level, optname, mp));
1608: } else
1609: return (ENOPROTOOPT);
1610: } else {
1611: m = m_get(M_WAIT, MT_SOOPTS);
1.36 perry 1612: m->m_len = sizeof(int);
1.1 cgd 1613:
1614: switch (optname) {
1615:
1616: case SO_LINGER:
1.36 perry 1617: m->m_len = sizeof(struct linger);
1.1 cgd 1618: mtod(m, struct linger *)->l_onoff =
1.131 christos 1619: (so->so_options & SO_LINGER) ? 1 : 0;
1.1 cgd 1620: mtod(m, struct linger *)->l_linger = so->so_linger;
1621: break;
1622:
1623: case SO_USELOOPBACK:
1624: case SO_DONTROUTE:
1625: case SO_DEBUG:
1626: case SO_KEEPALIVE:
1627: case SO_REUSEADDR:
1.15 mycroft 1628: case SO_REUSEPORT:
1.1 cgd 1629: case SO_BROADCAST:
1630: case SO_OOBINLINE:
1.26 thorpej 1631: case SO_TIMESTAMP:
1.131 christos 1632: *mtod(m, int *) = (so->so_options & optname) ? 1 : 0;
1.1 cgd 1633: break;
1634:
1635: case SO_TYPE:
1636: *mtod(m, int *) = so->so_type;
1637: break;
1638:
1639: case SO_ERROR:
1640: *mtod(m, int *) = so->so_error;
1641: so->so_error = 0;
1642: break;
1643:
1644: case SO_SNDBUF:
1645: *mtod(m, int *) = so->so_snd.sb_hiwat;
1646: break;
1647:
1648: case SO_RCVBUF:
1649: *mtod(m, int *) = so->so_rcv.sb_hiwat;
1650: break;
1651:
1652: case SO_SNDLOWAT:
1653: *mtod(m, int *) = so->so_snd.sb_lowat;
1654: break;
1655:
1656: case SO_RCVLOWAT:
1657: *mtod(m, int *) = so->so_rcv.sb_lowat;
1658: break;
1659:
1660: case SO_SNDTIMEO:
1661: case SO_RCVTIMEO:
1662: {
1663: int val = (optname == SO_SNDTIMEO ?
1664: so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
1665:
1666: m->m_len = sizeof(struct timeval);
1667: mtod(m, struct timeval *)->tv_sec = val / hz;
1668: mtod(m, struct timeval *)->tv_usec =
1.27 kleink 1669: (val % hz) * tick;
1.1 cgd 1670: break;
1671: }
1672:
1.107 darrenr 1673: case SO_OVERFLOWED:
1674: *mtod(m, int *) = so->so_rcv.sb_overflowed;
1675: break;
1676:
1.1 cgd 1677: default:
1678: (void)m_free(m);
1679: return (ENOPROTOOPT);
1680: }
1681: *mp = m;
1682: return (0);
1683: }
1684: }
1685:
1.14 mycroft 1686: void
1.54 lukem 1687: sohasoutofband(struct socket *so)
1.1 cgd 1688: {
1.90 christos 1689: fownsignal(so->so_pgid, SIGURG, POLL_PRI, POLLPRI|POLLRDBAND, so);
1.2 cgd 1690: selwakeup(&so->so_rcv.sb_sel);
1.1 cgd 1691: }
1.72 jdolecek 1692:
1693: static void
1694: filt_sordetach(struct knote *kn)
1695: {
1696: struct socket *so;
1697:
1698: so = (struct socket *)kn->kn_fp->f_data;
1.73 christos 1699: SLIST_REMOVE(&so->so_rcv.sb_sel.sel_klist, kn, knote, kn_selnext);
1700: if (SLIST_EMPTY(&so->so_rcv.sb_sel.sel_klist))
1.72 jdolecek 1701: so->so_rcv.sb_flags &= ~SB_KNOTE;
1702: }
1703:
1704: /*ARGSUSED*/
1705: static int
1.129 yamt 1706: filt_soread(struct knote *kn, long hint)
1.72 jdolecek 1707: {
1708: struct socket *so;
1709:
1710: so = (struct socket *)kn->kn_fp->f_data;
1711: kn->kn_data = so->so_rcv.sb_cc;
1712: if (so->so_state & SS_CANTRCVMORE) {
1.108 perry 1713: kn->kn_flags |= EV_EOF;
1.72 jdolecek 1714: kn->kn_fflags = so->so_error;
1715: return (1);
1716: }
1717: if (so->so_error) /* temporary udp error */
1718: return (1);
1719: if (kn->kn_sfflags & NOTE_LOWAT)
1720: return (kn->kn_data >= kn->kn_sdata);
1721: return (kn->kn_data >= so->so_rcv.sb_lowat);
1722: }
1723:
1724: static void
1725: filt_sowdetach(struct knote *kn)
1726: {
1727: struct socket *so;
1728:
1729: so = (struct socket *)kn->kn_fp->f_data;
1.73 christos 1730: SLIST_REMOVE(&so->so_snd.sb_sel.sel_klist, kn, knote, kn_selnext);
1731: if (SLIST_EMPTY(&so->so_snd.sb_sel.sel_klist))
1.72 jdolecek 1732: so->so_snd.sb_flags &= ~SB_KNOTE;
1733: }
1734:
1735: /*ARGSUSED*/
1736: static int
1.129 yamt 1737: filt_sowrite(struct knote *kn, long hint)
1.72 jdolecek 1738: {
1739: struct socket *so;
1740:
1741: so = (struct socket *)kn->kn_fp->f_data;
1742: kn->kn_data = sbspace(&so->so_snd);
1743: if (so->so_state & SS_CANTSENDMORE) {
1.108 perry 1744: kn->kn_flags |= EV_EOF;
1.72 jdolecek 1745: kn->kn_fflags = so->so_error;
1746: return (1);
1747: }
1748: if (so->so_error) /* temporary udp error */
1749: return (1);
1750: if (((so->so_state & SS_ISCONNECTED) == 0) &&
1751: (so->so_proto->pr_flags & PR_CONNREQUIRED))
1752: return (0);
1753: if (kn->kn_sfflags & NOTE_LOWAT)
1754: return (kn->kn_data >= kn->kn_sdata);
1755: return (kn->kn_data >= so->so_snd.sb_lowat);
1756: }
1757:
1758: /*ARGSUSED*/
1759: static int
1.129 yamt 1760: filt_solisten(struct knote *kn, long hint)
1.72 jdolecek 1761: {
1762: struct socket *so;
1763:
1764: so = (struct socket *)kn->kn_fp->f_data;
1765:
1766: /*
1767: * Set kn_data to number of incoming connections, not
1768: * counting partial (incomplete) connections.
1.108 perry 1769: */
1.72 jdolecek 1770: kn->kn_data = so->so_qlen;
1771: return (kn->kn_data > 0);
1772: }
1773:
1774: static const struct filterops solisten_filtops =
1775: { 1, NULL, filt_sordetach, filt_solisten };
1776: static const struct filterops soread_filtops =
1777: { 1, NULL, filt_sordetach, filt_soread };
1778: static const struct filterops sowrite_filtops =
1779: { 1, NULL, filt_sowdetach, filt_sowrite };
1780:
1781: int
1.129 yamt 1782: soo_kqfilter(struct file *fp, struct knote *kn)
1.72 jdolecek 1783: {
1784: struct socket *so;
1785: struct sockbuf *sb;
1786:
1787: so = (struct socket *)kn->kn_fp->f_data;
1788: switch (kn->kn_filter) {
1789: case EVFILT_READ:
1790: if (so->so_options & SO_ACCEPTCONN)
1791: kn->kn_fop = &solisten_filtops;
1792: else
1793: kn->kn_fop = &soread_filtops;
1794: sb = &so->so_rcv;
1795: break;
1796: case EVFILT_WRITE:
1797: kn->kn_fop = &sowrite_filtops;
1798: sb = &so->so_snd;
1799: break;
1800: default:
1801: return (1);
1802: }
1.73 christos 1803: SLIST_INSERT_HEAD(&sb->sb_sel.sel_klist, kn, kn_selnext);
1.72 jdolecek 1804: sb->sb_flags |= SB_KNOTE;
1805: return (0);
1806: }
1807:
1.94 yamt 1808: #include <sys/sysctl.h>
1809:
1810: static int sysctl_kern_somaxkva(SYSCTLFN_PROTO);
1811:
1812: /*
1813: * sysctl helper routine for kern.somaxkva. ensures that the given
1814: * value is not too small.
1815: * (XXX should we maybe make sure it's not too large as well?)
1816: */
1817: static int
1818: sysctl_kern_somaxkva(SYSCTLFN_ARGS)
1819: {
1820: int error, new_somaxkva;
1821: struct sysctlnode node;
1822:
1823: new_somaxkva = somaxkva;
1824: node = *rnode;
1825: node.sysctl_data = &new_somaxkva;
1826: error = sysctl_lookup(SYSCTLFN_CALL(&node));
1827: if (error || newp == NULL)
1828: return (error);
1829:
1830: if (new_somaxkva < (16 * 1024 * 1024)) /* sanity */
1831: return (EINVAL);
1832:
1.136 ad 1833: mutex_enter(&so_pendfree_lock);
1.94 yamt 1834: somaxkva = new_somaxkva;
1.136 ad 1835: cv_broadcast(&socurkva_cv);
1836: mutex_exit(&so_pendfree_lock);
1.94 yamt 1837:
1838: return (error);
1839: }
1840:
1841: SYSCTL_SETUP(sysctl_kern_somaxkva_setup, "sysctl kern.somaxkva setup")
1842: {
1843:
1.97 atatat 1844: sysctl_createv(clog, 0, NULL, NULL,
1845: CTLFLAG_PERMANENT,
1846: CTLTYPE_NODE, "kern", NULL,
1847: NULL, 0, NULL, 0,
1848: CTL_KERN, CTL_EOL);
1849:
1850: sysctl_createv(clog, 0, NULL, NULL,
1851: CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1.103 atatat 1852: CTLTYPE_INT, "somaxkva",
1853: SYSCTL_DESCR("Maximum amount of kernel memory to be "
1854: "used for socket buffers"),
1.94 yamt 1855: sysctl_kern_somaxkva, 0, NULL, 0,
1856: CTL_KERN, KERN_SOMAXKVA, CTL_EOL);
1857: }
CVSweb <webmaster@jp.NetBSD.org>