1 /* $NetBSD: uipc_socket.c,v 1.248 2016/06/10 13:27:15 ozaki-r Exp $ */
2
3 /*-
4 * Copyright (c) 2002, 2007, 2008, 2009 The NetBSD Foundation, Inc.
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, and by Andrew Doran.
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 */
31
32 /*
33 * Copyright (c) 2004 The FreeBSD Foundation
34 * Copyright (c) 2004 Robert Watson
35 * Copyright (c) 1982, 1986, 1988, 1990, 1993
36 * The Regents of the University of California. All rights reserved.
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.
46 * 3. Neither the name of the University nor the names of its contributors
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 *
62 * @(#)uipc_socket.c 8.6 (Berkeley) 5/2/95
63 */
64
65 /*
66 * Socket operation routines.
67 *
68 * These routines are called by the routines in sys_socket.c or from a
69 * system process, and implement the semantics of socket operations by
70 * switching out to the protocol specific routines.
71 */
72
73 #include <sys/cdefs.h>
74 __KERNEL_RCSID(0, "$NetBSD: uipc_socket.c,v 1.248 2016/06/10 13:27:15 ozaki-r Exp $");
75
76 #ifdef _KERNEL_OPT
77 #include "opt_compat_netbsd.h"
78 #include "opt_sock_counters.h"
79 #include "opt_sosend_loan.h"
80 #include "opt_mbuftrace.h"
81 #include "opt_somaxkva.h"
82 #include "opt_multiprocessor.h" /* XXX */
83 #include "opt_sctp.h"
84 #endif
85
86 #include <sys/param.h>
87 #include <sys/systm.h>
88 #include <sys/proc.h>
89 #include <sys/file.h>
90 #include <sys/filedesc.h>
91 #include <sys/kmem.h>
92 #include <sys/mbuf.h>
93 #include <sys/domain.h>
94 #include <sys/kernel.h>
95 #include <sys/protosw.h>
96 #include <sys/socket.h>
97 #include <sys/socketvar.h>
98 #include <sys/signalvar.h>
99 #include <sys/resourcevar.h>
100 #include <sys/uidinfo.h>
101 #include <sys/event.h>
102 #include <sys/poll.h>
103 #include <sys/kauth.h>
104 #include <sys/mutex.h>
105 #include <sys/condvar.h>
106 #include <sys/kthread.h>
107
108 #ifdef COMPAT_50
109 #include <compat/sys/time.h>
110 #include <compat/sys/socket.h>
111 #endif
112
113 #include <uvm/uvm_extern.h>
114 #include <uvm/uvm_loan.h>
115 #include <uvm/uvm_page.h>
116
117 MALLOC_DEFINE(M_SONAME, "soname", "socket name");
118
119 extern const struct fileops socketops;
120
121 extern int somaxconn; /* patchable (XXX sysctl) */
122 int somaxconn = SOMAXCONN;
123 kmutex_t *softnet_lock;
124
125 #ifdef SOSEND_COUNTERS
126 #include <sys/device.h>
127
128 static struct evcnt sosend_loan_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
129 NULL, "sosend", "loan big");
130 static struct evcnt sosend_copy_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
131 NULL, "sosend", "copy big");
132 static struct evcnt sosend_copy_small = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
133 NULL, "sosend", "copy small");
134 static struct evcnt sosend_kvalimit = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
135 NULL, "sosend", "kva limit");
136
137 #define SOSEND_COUNTER_INCR(ev) (ev)->ev_count++
138
139 EVCNT_ATTACH_STATIC(sosend_loan_big);
140 EVCNT_ATTACH_STATIC(sosend_copy_big);
141 EVCNT_ATTACH_STATIC(sosend_copy_small);
142 EVCNT_ATTACH_STATIC(sosend_kvalimit);
143 #else
144
145 #define SOSEND_COUNTER_INCR(ev) /* nothing */
146
147 #endif /* SOSEND_COUNTERS */
148
149 #if defined(SOSEND_NO_LOAN) || defined(MULTIPROCESSOR)
150 int sock_loan_thresh = -1;
151 #else
152 int sock_loan_thresh = 4096;
153 #endif
154
155 static kmutex_t so_pendfree_lock;
156 static struct mbuf *so_pendfree = NULL;
157
158 #ifndef SOMAXKVA
159 #define SOMAXKVA (16 * 1024 * 1024)
160 #endif
161 int somaxkva = SOMAXKVA;
162 static int socurkva;
163 static kcondvar_t socurkva_cv;
164
165 static kauth_listener_t socket_listener;
166
167 #define SOCK_LOAN_CHUNK 65536
168
169 static void sopendfree_thread(void *);
170 static kcondvar_t pendfree_thread_cv;
171 static lwp_t *sopendfree_lwp;
172
173 static void sysctl_kern_socket_setup(void);
174 static struct sysctllog *socket_sysctllog;
175
176 static vsize_t
sokvareserve(struct socket * so,vsize_t len)177 sokvareserve(struct socket *so, vsize_t len)
178 {
179 int error;
180
181 mutex_enter(&so_pendfree_lock);
182 while (socurkva + len > somaxkva) {
183 SOSEND_COUNTER_INCR(&sosend_kvalimit);
184 error = cv_wait_sig(&socurkva_cv, &so_pendfree_lock);
185 if (error) {
186 len = 0;
187 break;
188 }
189 }
190 socurkva += len;
191 mutex_exit(&so_pendfree_lock);
192 return len;
193 }
194
195 static void
sokvaunreserve(vsize_t len)196 sokvaunreserve(vsize_t len)
197 {
198
199 mutex_enter(&so_pendfree_lock);
200 socurkva -= len;
201 cv_broadcast(&socurkva_cv);
202 mutex_exit(&so_pendfree_lock);
203 }
204
205 /*
206 * sokvaalloc: allocate kva for loan.
207 */
208
209 vaddr_t
sokvaalloc(vaddr_t sva,vsize_t len,struct socket * so)210 sokvaalloc(vaddr_t sva, vsize_t len, struct socket *so)
211 {
212 vaddr_t lva;
213
214 /*
215 * reserve kva.
216 */
217
218 if (sokvareserve(so, len) == 0)
219 return 0;
220
221 /*
222 * allocate kva.
223 */
224
225 lva = uvm_km_alloc(kernel_map, len, atop(sva) & uvmexp.colormask,
226 UVM_KMF_COLORMATCH | UVM_KMF_VAONLY | UVM_KMF_WAITVA);
227 if (lva == 0) {
228 sokvaunreserve(len);
229 return (0);
230 }
231
232 return lva;
233 }
234
235 /*
236 * sokvafree: free kva for loan.
237 */
238
239 void
sokvafree(vaddr_t sva,vsize_t len)240 sokvafree(vaddr_t sva, vsize_t len)
241 {
242
243 /*
244 * free kva.
245 */
246
247 uvm_km_free(kernel_map, sva, len, UVM_KMF_VAONLY);
248
249 /*
250 * unreserve kva.
251 */
252
253 sokvaunreserve(len);
254 }
255
256 static void
sodoloanfree(struct vm_page ** pgs,void * buf,size_t size)257 sodoloanfree(struct vm_page **pgs, void *buf, size_t size)
258 {
259 vaddr_t sva, eva;
260 vsize_t len;
261 int npgs;
262
263 KASSERT(pgs != NULL);
264
265 eva = round_page((vaddr_t) buf + size);
266 sva = trunc_page((vaddr_t) buf);
267 len = eva - sva;
268 npgs = len >> PAGE_SHIFT;
269
270 pmap_kremove(sva, len);
271 pmap_update(pmap_kernel());
272 uvm_unloan(pgs, npgs, UVM_LOAN_TOPAGE);
273 sokvafree(sva, len);
274 }
275
276 /*
277 * sopendfree_thread: free mbufs on "pendfree" list.
278 * unlock and relock so_pendfree_lock when freeing mbufs.
279 */
280
281 static void
sopendfree_thread(void * v)282 sopendfree_thread(void *v)
283 {
284 struct mbuf *m, *next;
285 size_t rv;
286
287 mutex_enter(&so_pendfree_lock);
288
289 for (;;) {
290 rv = 0;
291 while (so_pendfree != NULL) {
292 m = so_pendfree;
293 so_pendfree = NULL;
294 mutex_exit(&so_pendfree_lock);
295
296 for (; m != NULL; m = next) {
297 next = m->m_next;
298 KASSERT((~m->m_flags & (M_EXT|M_EXT_PAGES)) == 0);
299 KASSERT(m->m_ext.ext_refcnt == 0);
300
301 rv += m->m_ext.ext_size;
302 sodoloanfree(m->m_ext.ext_pgs, m->m_ext.ext_buf,
303 m->m_ext.ext_size);
304 pool_cache_put(mb_cache, m);
305 }
306
307 mutex_enter(&so_pendfree_lock);
308 }
309 if (rv)
310 cv_broadcast(&socurkva_cv);
311 cv_wait(&pendfree_thread_cv, &so_pendfree_lock);
312 }
313 panic("sopendfree_thread");
314 /* NOTREACHED */
315 }
316
317 void
soloanfree(struct mbuf * m,void * buf,size_t size,void * arg)318 soloanfree(struct mbuf *m, void *buf, size_t size, void *arg)
319 {
320
321 KASSERT(m != NULL);
322
323 /*
324 * postpone freeing mbuf.
325 *
326 * we can't do it in interrupt context
327 * because we need to put kva back to kernel_map.
328 */
329
330 mutex_enter(&so_pendfree_lock);
331 m->m_next = so_pendfree;
332 so_pendfree = m;
333 cv_signal(&pendfree_thread_cv);
334 mutex_exit(&so_pendfree_lock);
335 }
336
337 static long
sosend_loan(struct socket * so,struct uio * uio,struct mbuf * m,long space)338 sosend_loan(struct socket *so, struct uio *uio, struct mbuf *m, long space)
339 {
340 struct iovec *iov = uio->uio_iov;
341 vaddr_t sva, eva;
342 vsize_t len;
343 vaddr_t lva;
344 int npgs, error;
345 vaddr_t va;
346 int i;
347
348 if (VMSPACE_IS_KERNEL_P(uio->uio_vmspace))
349 return (0);
350
351 if (iov->iov_len < (size_t) space)
352 space = iov->iov_len;
353 if (space > SOCK_LOAN_CHUNK)
354 space = SOCK_LOAN_CHUNK;
355
356 eva = round_page((vaddr_t) iov->iov_base + space);
357 sva = trunc_page((vaddr_t) iov->iov_base);
358 len = eva - sva;
359 npgs = len >> PAGE_SHIFT;
360
361 KASSERT(npgs <= M_EXT_MAXPAGES);
362
363 lva = sokvaalloc(sva, len, so);
364 if (lva == 0)
365 return 0;
366
367 error = uvm_loan(&uio->uio_vmspace->vm_map, sva, len,
368 m->m_ext.ext_pgs, UVM_LOAN_TOPAGE);
369 if (error) {
370 sokvafree(lva, len);
371 return (0);
372 }
373
374 for (i = 0, va = lva; i < npgs; i++, va += PAGE_SIZE)
375 pmap_kenter_pa(va, VM_PAGE_TO_PHYS(m->m_ext.ext_pgs[i]),
376 VM_PROT_READ, 0);
377 pmap_update(pmap_kernel());
378
379 lva += (vaddr_t) iov->iov_base & PAGE_MASK;
380
381 MEXTADD(m, (void *) lva, space, M_MBUF, soloanfree, so);
382 m->m_flags |= M_EXT_PAGES | M_EXT_ROMAP;
383
384 uio->uio_resid -= space;
385 /* uio_offset not updated, not set/used for write(2) */
386 uio->uio_iov->iov_base = (char *)uio->uio_iov->iov_base + space;
387 uio->uio_iov->iov_len -= space;
388 if (uio->uio_iov->iov_len == 0) {
389 uio->uio_iov++;
390 uio->uio_iovcnt--;
391 }
392
393 return (space);
394 }
395
396 struct mbuf *
getsombuf(struct socket * so,int type)397 getsombuf(struct socket *so, int type)
398 {
399 struct mbuf *m;
400
401 m = m_get(M_WAIT, type);
402 MCLAIM(m, so->so_mowner);
403 return m;
404 }
405
406 static int
socket_listener_cb(kauth_cred_t cred,kauth_action_t action,void * cookie,void * arg0,void * arg1,void * arg2,void * arg3)407 socket_listener_cb(kauth_cred_t cred, kauth_action_t action, void *cookie,
408 void *arg0, void *arg1, void *arg2, void *arg3)
409 {
410 int result;
411 enum kauth_network_req req;
412
413 result = KAUTH_RESULT_DEFER;
414 req = (enum kauth_network_req)(uintptr_t)arg0;
415
416 if ((action != KAUTH_NETWORK_SOCKET) &&
417 (action != KAUTH_NETWORK_BIND))
418 return result;
419
420 switch (req) {
421 case KAUTH_REQ_NETWORK_BIND_PORT:
422 result = KAUTH_RESULT_ALLOW;
423 break;
424
425 case KAUTH_REQ_NETWORK_SOCKET_DROP: {
426 /* Normal users can only drop their own connections. */
427 struct socket *so = (struct socket *)arg1;
428
429 if (so->so_cred && proc_uidmatch(cred, so->so_cred) == 0)
430 result = KAUTH_RESULT_ALLOW;
431
432 break;
433 }
434
435 case KAUTH_REQ_NETWORK_SOCKET_OPEN:
436 /* We allow "raw" routing/bluetooth sockets to anyone. */
437 if ((u_long)arg1 == PF_ROUTE || (u_long)arg1 == PF_OROUTE
438 || (u_long)arg1 == PF_BLUETOOTH) {
439 result = KAUTH_RESULT_ALLOW;
440 } else {
441 /* Privileged, let secmodel handle this. */
442 if ((u_long)arg2 == SOCK_RAW)
443 break;
444 }
445
446 result = KAUTH_RESULT_ALLOW;
447
448 break;
449
450 case KAUTH_REQ_NETWORK_SOCKET_CANSEE:
451 result = KAUTH_RESULT_ALLOW;
452
453 break;
454
455 default:
456 break;
457 }
458
459 return result;
460 }
461
462 void
soinit(void)463 soinit(void)
464 {
465
466 sysctl_kern_socket_setup();
467
468 mutex_init(&so_pendfree_lock, MUTEX_DEFAULT, IPL_VM);
469 softnet_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE);
470 cv_init(&socurkva_cv, "sokva");
471 cv_init(&pendfree_thread_cv, "sopendfr");
472 soinit2();
473
474 /* Set the initial adjusted socket buffer size. */
475 if (sb_max_set(sb_max))
476 panic("bad initial sb_max value: %lu", sb_max);
477
478 socket_listener = kauth_listen_scope(KAUTH_SCOPE_NETWORK,
479 socket_listener_cb, NULL);
480 }
481
482 void
soinit1(void)483 soinit1(void)
484 {
485 int error = kthread_create(PRI_NONE, KTHREAD_MPSAFE, NULL,
486 sopendfree_thread, NULL, &sopendfree_lwp, "sopendfree");
487 if (error)
488 panic("soinit1 %d", error);
489 }
490
491 /*
492 * socreate: create a new socket of the specified type and the protocol.
493 *
494 * => Caller may specify another socket for lock sharing (must not be held).
495 * => Returns the new socket without lock held.
496 */
497 int
socreate(int dom,struct socket ** aso,int type,int proto,struct lwp * l,struct socket * lockso)498 socreate(int dom, struct socket **aso, int type, int proto, struct lwp *l,
499 struct socket *lockso)
500 {
501 const struct protosw *prp;
502 struct socket *so;
503 uid_t uid;
504 int error;
505 kmutex_t *lock;
506
507 error = kauth_authorize_network(l->l_cred, KAUTH_NETWORK_SOCKET,
508 KAUTH_REQ_NETWORK_SOCKET_OPEN, KAUTH_ARG(dom), KAUTH_ARG(type),
509 KAUTH_ARG(proto));
510 if (error != 0)
511 return error;
512
513 if (proto)
514 prp = pffindproto(dom, proto, type);
515 else
516 prp = pffindtype(dom, type);
517 if (prp == NULL) {
518 /* no support for domain */
519 if (pffinddomain(dom) == 0)
520 return EAFNOSUPPORT;
521 /* no support for socket type */
522 if (proto == 0 && type != 0)
523 return EPROTOTYPE;
524 return EPROTONOSUPPORT;
525 }
526 if (prp->pr_usrreqs == NULL)
527 return EPROTONOSUPPORT;
528 if (prp->pr_type != type)
529 return EPROTOTYPE;
530
531 so = soget(true);
532 so->so_type = type;
533 so->so_proto = prp;
534 so->so_send = sosend;
535 so->so_receive = soreceive;
536 #ifdef MBUFTRACE
537 so->so_rcv.sb_mowner = &prp->pr_domain->dom_mowner;
538 so->so_snd.sb_mowner = &prp->pr_domain->dom_mowner;
539 so->so_mowner = &prp->pr_domain->dom_mowner;
540 #endif
541 uid = kauth_cred_geteuid(l->l_cred);
542 so->so_uidinfo = uid_find(uid);
543 so->so_cpid = l->l_proc->p_pid;
544
545 /*
546 * Lock assigned and taken during PCB attach, unless we share
547 * the lock with another socket, e.g. socketpair(2) case.
548 */
549 if (lockso) {
550 lock = lockso->so_lock;
551 so->so_lock = lock;
552 mutex_obj_hold(lock);
553 mutex_enter(lock);
554 }
555
556 /* Attach the PCB (returns with the socket lock held). */
557 error = (*prp->pr_usrreqs->pr_attach)(so, proto);
558 KASSERT(solocked(so));
559
560 if (error) {
561 KASSERT(so->so_pcb == NULL);
562 so->so_state |= SS_NOFDREF;
563 sofree(so);
564 return error;
565 }
566 so->so_cred = kauth_cred_dup(l->l_cred);
567 sounlock(so);
568
569 *aso = so;
570 return 0;
571 }
572
573 /*
574 * fsocreate: create a socket and a file descriptor associated with it.
575 *
576 * => On success, write file descriptor to fdout and return zero.
577 * => On failure, return non-zero; *fdout will be undefined.
578 */
579 int
fsocreate(int domain,struct socket ** sop,int type,int proto,int * fdout)580 fsocreate(int domain, struct socket **sop, int type, int proto, int *fdout)
581 {
582 lwp_t *l = curlwp;
583 int error, fd, flags;
584 struct socket *so;
585 struct file *fp;
586
587 if ((error = fd_allocfile(&fp, &fd)) != 0) {
588 return error;
589 }
590 flags = type & SOCK_FLAGS_MASK;
591 fd_set_exclose(l, fd, (flags & SOCK_CLOEXEC) != 0);
592 fp->f_flag = FREAD|FWRITE|((flags & SOCK_NONBLOCK) ? FNONBLOCK : 0)|
593 ((flags & SOCK_NOSIGPIPE) ? FNOSIGPIPE : 0);
594 fp->f_type = DTYPE_SOCKET;
595 fp->f_ops = &socketops;
596
597 type &= ~SOCK_FLAGS_MASK;
598 error = socreate(domain, &so, type, proto, l, NULL);
599 if (error) {
600 fd_abort(curproc, fp, fd);
601 return error;
602 }
603 if (flags & SOCK_NONBLOCK) {
604 so->so_state |= SS_NBIO;
605 }
606 fp->f_socket = so;
607 fd_affix(curproc, fp, fd);
608
609 if (sop != NULL) {
610 *sop = so;
611 }
612 *fdout = fd;
613 return error;
614 }
615
616 int
sofamily(const struct socket * so)617 sofamily(const struct socket *so)
618 {
619 const struct protosw *pr;
620 const struct domain *dom;
621
622 if ((pr = so->so_proto) == NULL)
623 return AF_UNSPEC;
624 if ((dom = pr->pr_domain) == NULL)
625 return AF_UNSPEC;
626 return dom->dom_family;
627 }
628
629 int
sobind(struct socket * so,struct sockaddr * nam,struct lwp * l)630 sobind(struct socket *so, struct sockaddr *nam, struct lwp *l)
631 {
632 int error;
633
634 solock(so);
635 if (nam->sa_family != so->so_proto->pr_domain->dom_family) {
636 sounlock(so);
637 return EAFNOSUPPORT;
638 }
639 error = (*so->so_proto->pr_usrreqs->pr_bind)(so, nam, l);
640 sounlock(so);
641 return error;
642 }
643
644 int
solisten(struct socket * so,int backlog,struct lwp * l)645 solisten(struct socket *so, int backlog, struct lwp *l)
646 {
647 int error;
648 short oldopt, oldqlimit;
649
650 solock(so);
651 if ((so->so_state & (SS_ISCONNECTED | SS_ISCONNECTING |
652 SS_ISDISCONNECTING)) != 0) {
653 sounlock(so);
654 return EINVAL;
655 }
656 oldopt = so->so_options;
657 oldqlimit = so->so_qlimit;
658 if (TAILQ_EMPTY(&so->so_q))
659 so->so_options |= SO_ACCEPTCONN;
660 if (backlog < 0)
661 backlog = 0;
662 so->so_qlimit = min(backlog, somaxconn);
663
664 error = (*so->so_proto->pr_usrreqs->pr_listen)(so, l);
665 if (error != 0) {
666 so->so_options = oldopt;
667 so->so_qlimit = oldqlimit;
668 sounlock(so);
669 return error;
670 }
671 sounlock(so);
672 return 0;
673 }
674
675 void
sofree(struct socket * so)676 sofree(struct socket *so)
677 {
678 u_int refs;
679
680 KASSERT(solocked(so));
681
682 if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0) {
683 sounlock(so);
684 return;
685 }
686 if (so->so_head) {
687 /*
688 * We must not decommission a socket that's on the accept(2)
689 * queue. If we do, then accept(2) may hang after select(2)
690 * indicated that the listening socket was ready.
691 */
692 if (!soqremque(so, 0)) {
693 sounlock(so);
694 return;
695 }
696 }
697 if (so->so_rcv.sb_hiwat)
698 (void)chgsbsize(so->so_uidinfo, &so->so_rcv.sb_hiwat, 0,
699 RLIM_INFINITY);
700 if (so->so_snd.sb_hiwat)
701 (void)chgsbsize(so->so_uidinfo, &so->so_snd.sb_hiwat, 0,
702 RLIM_INFINITY);
703 sbrelease(&so->so_snd, so);
704 KASSERT(!cv_has_waiters(&so->so_cv));
705 KASSERT(!cv_has_waiters(&so->so_rcv.sb_cv));
706 KASSERT(!cv_has_waiters(&so->so_snd.sb_cv));
707 sorflush(so);
708 refs = so->so_aborting; /* XXX */
709 /* Remove acccept filter if one is present. */
710 if (so->so_accf != NULL)
711 (void)accept_filt_clear(so);
712 sounlock(so);
713 if (refs == 0) /* XXX */
714 soput(so);
715 }
716
717 /*
718 * soclose: close a socket on last file table reference removal.
719 * Initiate disconnect if connected. Free socket when disconnect complete.
720 */
721 int
soclose(struct socket * so)722 soclose(struct socket *so)
723 {
724 struct socket *so2;
725 int error = 0;
726
727 solock(so);
728 if (so->so_options & SO_ACCEPTCONN) {
729 for (;;) {
730 if ((so2 = TAILQ_FIRST(&so->so_q0)) != 0) {
731 KASSERT(solocked2(so, so2));
732 (void) soqremque(so2, 0);
733 /* soabort drops the lock. */
734 (void) soabort(so2);
735 solock(so);
736 continue;
737 }
738 if ((so2 = TAILQ_FIRST(&so->so_q)) != 0) {
739 KASSERT(solocked2(so, so2));
740 (void) soqremque(so2, 1);
741 /* soabort drops the lock. */
742 (void) soabort(so2);
743 solock(so);
744 continue;
745 }
746 break;
747 }
748 }
749 if (so->so_pcb == NULL)
750 goto discard;
751 if (so->so_state & SS_ISCONNECTED) {
752 if ((so->so_state & SS_ISDISCONNECTING) == 0) {
753 error = sodisconnect(so);
754 if (error)
755 goto drop;
756 }
757 if (so->so_options & SO_LINGER) {
758 if ((so->so_state & (SS_ISDISCONNECTING|SS_NBIO)) ==
759 (SS_ISDISCONNECTING|SS_NBIO))
760 goto drop;
761 while (so->so_state & SS_ISCONNECTED) {
762 error = sowait(so, true, so->so_linger * hz);
763 if (error)
764 break;
765 }
766 }
767 }
768 drop:
769 if (so->so_pcb) {
770 KASSERT(solocked(so));
771 (*so->so_proto->pr_usrreqs->pr_detach)(so);
772 }
773 discard:
774 KASSERT((so->so_state & SS_NOFDREF) == 0);
775 kauth_cred_free(so->so_cred);
776 so->so_state |= SS_NOFDREF;
777 sofree(so);
778 return error;
779 }
780
781 /*
782 * Must be called with the socket locked.. Will return with it unlocked.
783 */
784 int
soabort(struct socket * so)785 soabort(struct socket *so)
786 {
787 u_int refs;
788 int error;
789
790 KASSERT(solocked(so));
791 KASSERT(so->so_head == NULL);
792
793 so->so_aborting++; /* XXX */
794 error = (*so->so_proto->pr_usrreqs->pr_abort)(so);
795 refs = --so->so_aborting; /* XXX */
796 if (error || (refs == 0)) {
797 sofree(so);
798 } else {
799 sounlock(so);
800 }
801 return error;
802 }
803
804 int
soaccept(struct socket * so,struct sockaddr * nam)805 soaccept(struct socket *so, struct sockaddr *nam)
806 {
807 int error;
808
809 KASSERT(solocked(so));
810 KASSERT((so->so_state & SS_NOFDREF) != 0);
811
812 so->so_state &= ~SS_NOFDREF;
813 if ((so->so_state & SS_ISDISCONNECTED) == 0 ||
814 (so->so_proto->pr_flags & PR_ABRTACPTDIS) == 0)
815 error = (*so->so_proto->pr_usrreqs->pr_accept)(so, nam);
816 else
817 error = ECONNABORTED;
818
819 return error;
820 }
821
822 int
soconnect(struct socket * so,struct sockaddr * nam,struct lwp * l)823 soconnect(struct socket *so, struct sockaddr *nam, struct lwp *l)
824 {
825 int error;
826
827 KASSERT(solocked(so));
828
829 if (so->so_options & SO_ACCEPTCONN)
830 return EOPNOTSUPP;
831 /*
832 * If protocol is connection-based, can only connect once.
833 * Otherwise, if connected, try to disconnect first.
834 * This allows user to disconnect by connecting to, e.g.,
835 * a null address.
836 */
837 if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) &&
838 ((so->so_proto->pr_flags & PR_CONNREQUIRED) ||
839 (error = sodisconnect(so)))) {
840 error = EISCONN;
841 } else {
842 if (nam->sa_family != so->so_proto->pr_domain->dom_family) {
843 return EAFNOSUPPORT;
844 }
845 error = (*so->so_proto->pr_usrreqs->pr_connect)(so, nam, l);
846 }
847
848 return error;
849 }
850
851 int
soconnect2(struct socket * so1,struct socket * so2)852 soconnect2(struct socket *so1, struct socket *so2)
853 {
854 KASSERT(solocked2(so1, so2));
855
856 return (*so1->so_proto->pr_usrreqs->pr_connect2)(so1, so2);
857 }
858
859 int
sodisconnect(struct socket * so)860 sodisconnect(struct socket *so)
861 {
862 int error;
863
864 KASSERT(solocked(so));
865
866 if ((so->so_state & SS_ISCONNECTED) == 0) {
867 error = ENOTCONN;
868 } else if (so->so_state & SS_ISDISCONNECTING) {
869 error = EALREADY;
870 } else {
871 error = (*so->so_proto->pr_usrreqs->pr_disconnect)(so);
872 }
873 return (error);
874 }
875
876 #define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK)
877 /*
878 * Send on a socket.
879 * If send must go all at once and message is larger than
880 * send buffering, then hard error.
881 * Lock against other senders.
882 * If must go all at once and not enough room now, then
883 * inform user that this would block and do nothing.
884 * Otherwise, if nonblocking, send as much as possible.
885 * The data to be sent is described by "uio" if nonzero,
886 * otherwise by the mbuf chain "top" (which must be null
887 * if uio is not). Data provided in mbuf chain must be small
888 * enough to send all at once.
889 *
890 * Returns nonzero on error, timeout or signal; callers
891 * must check for short counts if EINTR/ERESTART are returned.
892 * Data and control buffers are freed on return.
893 */
894 int
sosend(struct socket * so,struct sockaddr * addr,struct uio * uio,struct mbuf * top,struct mbuf * control,int flags,struct lwp * l)895 sosend(struct socket *so, struct sockaddr *addr, struct uio *uio,
896 struct mbuf *top, struct mbuf *control, int flags, struct lwp *l)
897 {
898 struct mbuf **mp, *m;
899 long space, len, resid, clen, mlen;
900 int error, s, dontroute, atomic;
901 short wakeup_state = 0;
902
903 clen = 0;
904
905 /*
906 * solock() provides atomicity of access. splsoftnet() prevents
907 * protocol processing soft interrupts from interrupting us and
908 * blocking (expensive).
909 */
910 s = splsoftnet();
911 solock(so);
912 atomic = sosendallatonce(so) || top;
913 if (uio)
914 resid = uio->uio_resid;
915 else
916 resid = top->m_pkthdr.len;
917 /*
918 * In theory resid should be unsigned.
919 * However, space must be signed, as it might be less than 0
920 * if we over-committed, and we must use a signed comparison
921 * of space and resid. On the other hand, a negative resid
922 * causes us to loop sending 0-length segments to the protocol.
923 */
924 if (resid < 0) {
925 error = EINVAL;
926 goto out;
927 }
928 dontroute =
929 (flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 &&
930 (so->so_proto->pr_flags & PR_ATOMIC);
931 l->l_ru.ru_msgsnd++;
932 if (control)
933 clen = control->m_len;
934 restart:
935 if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0)
936 goto out;
937 do {
938 if (so->so_state & SS_CANTSENDMORE) {
939 error = EPIPE;
940 goto release;
941 }
942 if (so->so_error) {
943 error = so->so_error;
944 so->so_error = 0;
945 goto release;
946 }
947 if ((so->so_state & SS_ISCONNECTED) == 0) {
948 if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
949 if (resid || clen == 0) {
950 error = ENOTCONN;
951 goto release;
952 }
953 } else if (addr == NULL) {
954 error = EDESTADDRREQ;
955 goto release;
956 }
957 }
958 space = sbspace(&so->so_snd);
959 if (flags & MSG_OOB)
960 space += 1024;
961 if ((atomic && resid > so->so_snd.sb_hiwat) ||
962 clen > so->so_snd.sb_hiwat) {
963 error = EMSGSIZE;
964 goto release;
965 }
966 if (space < resid + clen &&
967 (atomic || space < so->so_snd.sb_lowat || space < clen)) {
968 if ((so->so_state & SS_NBIO) || (flags & MSG_NBIO)) {
969 error = EWOULDBLOCK;
970 goto release;
971 }
972 sbunlock(&so->so_snd);
973 if (wakeup_state & SS_RESTARTSYS) {
974 error = ERESTART;
975 goto out;
976 }
977 error = sbwait(&so->so_snd);
978 if (error)
979 goto out;
980 wakeup_state = so->so_state;
981 goto restart;
982 }
983 wakeup_state = 0;
984 mp = ⊤
985 space -= clen;
986 do {
987 if (uio == NULL) {
988 /*
989 * Data is prepackaged in "top".
990 */
991 resid = 0;
992 if (flags & MSG_EOR)
993 top->m_flags |= M_EOR;
994 } else do {
995 sounlock(so);
996 splx(s);
997 if (top == NULL) {
998 m = m_gethdr(M_WAIT, MT_DATA);
999 mlen = MHLEN;
1000 m->m_pkthdr.len = 0;
1001 m_reset_rcvif(m);
1002 } else {
1003 m = m_get(M_WAIT, MT_DATA);
1004 mlen = MLEN;
1005 }
1006 MCLAIM(m, so->so_snd.sb_mowner);
1007 if (sock_loan_thresh >= 0 &&
1008 uio->uio_iov->iov_len >= sock_loan_thresh &&
1009 space >= sock_loan_thresh &&
1010 (len = sosend_loan(so, uio, m,
1011 space)) != 0) {
1012 SOSEND_COUNTER_INCR(&sosend_loan_big);
1013 space -= len;
1014 goto have_data;
1015 }
1016 if (resid >= MINCLSIZE && space >= MCLBYTES) {
1017 SOSEND_COUNTER_INCR(&sosend_copy_big);
1018 m_clget(m, M_DONTWAIT);
1019 if ((m->m_flags & M_EXT) == 0)
1020 goto nopages;
1021 mlen = MCLBYTES;
1022 if (atomic && top == 0) {
1023 len = lmin(MCLBYTES - max_hdr,
1024 resid);
1025 m->m_data += max_hdr;
1026 } else
1027 len = lmin(MCLBYTES, resid);
1028 space -= len;
1029 } else {
1030 nopages:
1031 SOSEND_COUNTER_INCR(&sosend_copy_small);
1032 len = lmin(lmin(mlen, resid), space);
1033 space -= len;
1034 /*
1035 * For datagram protocols, leave room
1036 * for protocol headers in first mbuf.
1037 */
1038 if (atomic && top == 0 && len < mlen)
1039 MH_ALIGN(m, len);
1040 }
1041 error = uiomove(mtod(m, void *), (int)len, uio);
1042 have_data:
1043 resid = uio->uio_resid;
1044 m->m_len = len;
1045 *mp = m;
1046 top->m_pkthdr.len += len;
1047 s = splsoftnet();
1048 solock(so);
1049 if (error != 0)
1050 goto release;
1051 mp = &m->m_next;
1052 if (resid <= 0) {
1053 if (flags & MSG_EOR)
1054 top->m_flags |= M_EOR;
1055 break;
1056 }
1057 } while (space > 0 && atomic);
1058
1059 if (so->so_state & SS_CANTSENDMORE) {
1060 error = EPIPE;
1061 goto release;
1062 }
1063 if (dontroute)
1064 so->so_options |= SO_DONTROUTE;
1065 if (resid > 0)
1066 so->so_state |= SS_MORETOCOME;
1067 if (flags & MSG_OOB) {
1068 error = (*so->so_proto->pr_usrreqs->pr_sendoob)(so,
1069 top, control);
1070 } else {
1071 error = (*so->so_proto->pr_usrreqs->pr_send)(so,
1072 top, addr, control, l);
1073 }
1074 if (dontroute)
1075 so->so_options &= ~SO_DONTROUTE;
1076 if (resid > 0)
1077 so->so_state &= ~SS_MORETOCOME;
1078 clen = 0;
1079 control = NULL;
1080 top = NULL;
1081 mp = ⊤
1082 if (error != 0)
1083 goto release;
1084 } while (resid && space > 0);
1085 } while (resid);
1086
1087 release:
1088 sbunlock(&so->so_snd);
1089 out:
1090 sounlock(so);
1091 splx(s);
1092 if (top)
1093 m_freem(top);
1094 if (control)
1095 m_freem(control);
1096 return (error);
1097 }
1098
1099 /*
1100 * Following replacement or removal of the first mbuf on the first
1101 * mbuf chain of a socket buffer, push necessary state changes back
1102 * into the socket buffer so that other consumers see the values
1103 * consistently. 'nextrecord' is the callers locally stored value of
1104 * the original value of sb->sb_mb->m_nextpkt which must be restored
1105 * when the lead mbuf changes. NOTE: 'nextrecord' may be NULL.
1106 */
1107 static void
sbsync(struct sockbuf * sb,struct mbuf * nextrecord)1108 sbsync(struct sockbuf *sb, struct mbuf *nextrecord)
1109 {
1110
1111 KASSERT(solocked(sb->sb_so));
1112
1113 /*
1114 * First, update for the new value of nextrecord. If necessary,
1115 * make it the first record.
1116 */
1117 if (sb->sb_mb != NULL)
1118 sb->sb_mb->m_nextpkt = nextrecord;
1119 else
1120 sb->sb_mb = nextrecord;
1121
1122 /*
1123 * Now update any dependent socket buffer fields to reflect
1124 * the new state. This is an inline of SB_EMPTY_FIXUP, with
1125 * the addition of a second clause that takes care of the
1126 * case where sb_mb has been updated, but remains the last
1127 * record.
1128 */
1129 if (sb->sb_mb == NULL) {
1130 sb->sb_mbtail = NULL;
1131 sb->sb_lastrecord = NULL;
1132 } else if (sb->sb_mb->m_nextpkt == NULL)
1133 sb->sb_lastrecord = sb->sb_mb;
1134 }
1135
1136 /*
1137 * Implement receive operations on a socket.
1138 * We depend on the way that records are added to the sockbuf
1139 * by sbappend*. In particular, each record (mbufs linked through m_next)
1140 * must begin with an address if the protocol so specifies,
1141 * followed by an optional mbuf or mbufs containing ancillary data,
1142 * and then zero or more mbufs of data.
1143 * In order to avoid blocking network interrupts for the entire time here,
1144 * we splx() while doing the actual copy to user space.
1145 * Although the sockbuf is locked, new data may still be appended,
1146 * and thus we must maintain consistency of the sockbuf during that time.
1147 *
1148 * The caller may receive the data as a single mbuf chain by supplying
1149 * an mbuf **mp0 for use in returning the chain. The uio is then used
1150 * only for the count in uio_resid.
1151 */
1152 int
soreceive(struct socket * so,struct mbuf ** paddr,struct uio * uio,struct mbuf ** mp0,struct mbuf ** controlp,int * flagsp)1153 soreceive(struct socket *so, struct mbuf **paddr, struct uio *uio,
1154 struct mbuf **mp0, struct mbuf **controlp, int *flagsp)
1155 {
1156 struct lwp *l = curlwp;
1157 struct mbuf *m, **mp, *mt;
1158 size_t len, offset, moff, orig_resid;
1159 int atomic, flags, error, s, type;
1160 const struct protosw *pr;
1161 struct mbuf *nextrecord;
1162 int mbuf_removed = 0;
1163 const struct domain *dom;
1164 short wakeup_state = 0;
1165
1166 pr = so->so_proto;
1167 atomic = pr->pr_flags & PR_ATOMIC;
1168 dom = pr->pr_domain;
1169 mp = mp0;
1170 type = 0;
1171 orig_resid = uio->uio_resid;
1172
1173 if (paddr != NULL)
1174 *paddr = NULL;
1175 if (controlp != NULL)
1176 *controlp = NULL;
1177 if (flagsp != NULL)
1178 flags = *flagsp &~ MSG_EOR;
1179 else
1180 flags = 0;
1181
1182 if (flags & MSG_OOB) {
1183 m = m_get(M_WAIT, MT_DATA);
1184 solock(so);
1185 error = (*pr->pr_usrreqs->pr_recvoob)(so, m, flags & MSG_PEEK);
1186 sounlock(so);
1187 if (error)
1188 goto bad;
1189 do {
1190 error = uiomove(mtod(m, void *),
1191 MIN(uio->uio_resid, m->m_len), uio);
1192 m = m_free(m);
1193 } while (uio->uio_resid > 0 && error == 0 && m);
1194 bad:
1195 if (m != NULL)
1196 m_freem(m);
1197 return error;
1198 }
1199 if (mp != NULL)
1200 *mp = NULL;
1201
1202 /*
1203 * solock() provides atomicity of access. splsoftnet() prevents
1204 * protocol processing soft interrupts from interrupting us and
1205 * blocking (expensive).
1206 */
1207 s = splsoftnet();
1208 solock(so);
1209 restart:
1210 if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0) {
1211 sounlock(so);
1212 splx(s);
1213 return error;
1214 }
1215
1216 m = so->so_rcv.sb_mb;
1217 /*
1218 * If we have less data than requested, block awaiting more
1219 * (subject to any timeout) if:
1220 * 1. the current count is less than the low water mark,
1221 * 2. MSG_WAITALL is set, and it is possible to do the entire
1222 * receive operation at once if we block (resid <= hiwat), or
1223 * 3. MSG_DONTWAIT is not set.
1224 * If MSG_WAITALL is set but resid is larger than the receive buffer,
1225 * we have to do the receive in sections, and thus risk returning
1226 * a short count if a timeout or signal occurs after we start.
1227 */
1228 if (m == NULL ||
1229 ((flags & MSG_DONTWAIT) == 0 &&
1230 so->so_rcv.sb_cc < uio->uio_resid &&
1231 (so->so_rcv.sb_cc < so->so_rcv.sb_lowat ||
1232 ((flags & MSG_WAITALL) &&
1233 uio->uio_resid <= so->so_rcv.sb_hiwat)) &&
1234 m->m_nextpkt == NULL && !atomic)) {
1235 #ifdef DIAGNOSTIC
1236 if (m == NULL && so->so_rcv.sb_cc)
1237 panic("receive 1");
1238 #endif
1239 if (so->so_error) {
1240 if (m != NULL)
1241 goto dontblock;
1242 error = so->so_error;
1243 if ((flags & MSG_PEEK) == 0)
1244 so->so_error = 0;
1245 goto release;
1246 }
1247 if (so->so_state & SS_CANTRCVMORE) {
1248 if (m != NULL)
1249 goto dontblock;
1250 else
1251 goto release;
1252 }
1253 for (; m != NULL; m = m->m_next)
1254 if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) {
1255 m = so->so_rcv.sb_mb;
1256 goto dontblock;
1257 }
1258 if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
1259 (so->so_proto->pr_flags & PR_CONNREQUIRED)) {
1260 error = ENOTCONN;
1261 goto release;
1262 }
1263 if (uio->uio_resid == 0)
1264 goto release;
1265 if ((so->so_state & SS_NBIO) ||
1266 (flags & (MSG_DONTWAIT|MSG_NBIO))) {
1267 error = EWOULDBLOCK;
1268 goto release;
1269 }
1270 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 1");
1271 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 1");
1272 sbunlock(&so->so_rcv);
1273 if (wakeup_state & SS_RESTARTSYS)
1274 error = ERESTART;
1275 else
1276 error = sbwait(&so->so_rcv);
1277 if (error != 0) {
1278 sounlock(so);
1279 splx(s);
1280 return error;
1281 }
1282 wakeup_state = so->so_state;
1283 goto restart;
1284 }
1285 dontblock:
1286 /*
1287 * On entry here, m points to the first record of the socket buffer.
1288 * From this point onward, we maintain 'nextrecord' as a cache of the
1289 * pointer to the next record in the socket buffer. We must keep the
1290 * various socket buffer pointers and local stack versions of the
1291 * pointers in sync, pushing out modifications before dropping the
1292 * socket lock, and re-reading them when picking it up.
1293 *
1294 * Otherwise, we will race with the network stack appending new data
1295 * or records onto the socket buffer by using inconsistent/stale
1296 * versions of the field, possibly resulting in socket buffer
1297 * corruption.
1298 *
1299 * By holding the high-level sblock(), we prevent simultaneous
1300 * readers from pulling off the front of the socket buffer.
1301 */
1302 if (l != NULL)
1303 l->l_ru.ru_msgrcv++;
1304 KASSERT(m == so->so_rcv.sb_mb);
1305 SBLASTRECORDCHK(&so->so_rcv, "soreceive 1");
1306 SBLASTMBUFCHK(&so->so_rcv, "soreceive 1");
1307 nextrecord = m->m_nextpkt;
1308 if (pr->pr_flags & PR_ADDR) {
1309 #ifdef DIAGNOSTIC
1310 if (m->m_type != MT_SONAME)
1311 panic("receive 1a");
1312 #endif
1313 orig_resid = 0;
1314 if (flags & MSG_PEEK) {
1315 if (paddr)
1316 *paddr = m_copy(m, 0, m->m_len);
1317 m = m->m_next;
1318 } else {
1319 sbfree(&so->so_rcv, m);
1320 mbuf_removed = 1;
1321 if (paddr != NULL) {
1322 *paddr = m;
1323 so->so_rcv.sb_mb = m->m_next;
1324 m->m_next = NULL;
1325 m = so->so_rcv.sb_mb;
1326 } else {
1327 MFREE(m, so->so_rcv.sb_mb);
1328 m = so->so_rcv.sb_mb;
1329 }
1330 sbsync(&so->so_rcv, nextrecord);
1331 }
1332 }
1333 if (pr->pr_flags & PR_ADDR_OPT) {
1334 /*
1335 * For SCTP we may be getting a
1336 * whole message OR a partial delivery.
1337 */
1338 if (m->m_type == MT_SONAME) {
1339 orig_resid = 0;
1340 if (flags & MSG_PEEK) {
1341 if (paddr)
1342 *paddr = m_copy(m, 0, m->m_len);
1343 m = m->m_next;
1344 } else {
1345 sbfree(&so->so_rcv, m);
1346 if (paddr) {
1347 *paddr = m;
1348 so->so_rcv.sb_mb = m->m_next;
1349 m->m_next = 0;
1350 m = so->so_rcv.sb_mb;
1351 } else {
1352 MFREE(m, so->so_rcv.sb_mb);
1353 m = so->so_rcv.sb_mb;
1354 }
1355 }
1356 }
1357 }
1358
1359 /*
1360 * Process one or more MT_CONTROL mbufs present before any data mbufs
1361 * in the first mbuf chain on the socket buffer. If MSG_PEEK, we
1362 * just copy the data; if !MSG_PEEK, we call into the protocol to
1363 * perform externalization (or freeing if controlp == NULL).
1364 */
1365 if (__predict_false(m != NULL && m->m_type == MT_CONTROL)) {
1366 struct mbuf *cm = NULL, *cmn;
1367 struct mbuf **cme = &cm;
1368
1369 do {
1370 if (flags & MSG_PEEK) {
1371 if (controlp != NULL) {
1372 *controlp = m_copy(m, 0, m->m_len);
1373 controlp = &(*controlp)->m_next;
1374 }
1375 m = m->m_next;
1376 } else {
1377 sbfree(&so->so_rcv, m);
1378 so->so_rcv.sb_mb = m->m_next;
1379 m->m_next = NULL;
1380 *cme = m;
1381 cme = &(*cme)->m_next;
1382 m = so->so_rcv.sb_mb;
1383 }
1384 } while (m != NULL && m->m_type == MT_CONTROL);
1385 if ((flags & MSG_PEEK) == 0)
1386 sbsync(&so->so_rcv, nextrecord);
1387 for (; cm != NULL; cm = cmn) {
1388 cmn = cm->m_next;
1389 cm->m_next = NULL;
1390 type = mtod(cm, struct cmsghdr *)->cmsg_type;
1391 if (controlp != NULL) {
1392 if (dom->dom_externalize != NULL &&
1393 type == SCM_RIGHTS) {
1394 sounlock(so);
1395 splx(s);
1396 error = (*dom->dom_externalize)(cm, l,
1397 (flags & MSG_CMSG_CLOEXEC) ?
1398 O_CLOEXEC : 0);
1399 s = splsoftnet();
1400 solock(so);
1401 }
1402 *controlp = cm;
1403 while (*controlp != NULL)
1404 controlp = &(*controlp)->m_next;
1405 } else {
1406 /*
1407 * Dispose of any SCM_RIGHTS message that went
1408 * through the read path rather than recv.
1409 */
1410 if (dom->dom_dispose != NULL &&
1411 type == SCM_RIGHTS) {
1412 sounlock(so);
1413 (*dom->dom_dispose)(cm);
1414 solock(so);
1415 }
1416 m_freem(cm);
1417 }
1418 }
1419 if (m != NULL)
1420 nextrecord = so->so_rcv.sb_mb->m_nextpkt;
1421 else
1422 nextrecord = so->so_rcv.sb_mb;
1423 orig_resid = 0;
1424 }
1425
1426 /* If m is non-NULL, we have some data to read. */
1427 if (__predict_true(m != NULL)) {
1428 type = m->m_type;
1429 if (type == MT_OOBDATA)
1430 flags |= MSG_OOB;
1431 }
1432 SBLASTRECORDCHK(&so->so_rcv, "soreceive 2");
1433 SBLASTMBUFCHK(&so->so_rcv, "soreceive 2");
1434
1435 moff = 0;
1436 offset = 0;
1437 while (m != NULL && uio->uio_resid > 0 && error == 0) {
1438 if (m->m_type == MT_OOBDATA) {
1439 if (type != MT_OOBDATA)
1440 break;
1441 } else if (type == MT_OOBDATA)
1442 break;
1443 #ifdef DIAGNOSTIC
1444 else if (m->m_type != MT_DATA && m->m_type != MT_HEADER)
1445 panic("receive 3");
1446 #endif
1447 so->so_state &= ~SS_RCVATMARK;
1448 wakeup_state = 0;
1449 len = uio->uio_resid;
1450 if (so->so_oobmark && len > so->so_oobmark - offset)
1451 len = so->so_oobmark - offset;
1452 if (len > m->m_len - moff)
1453 len = m->m_len - moff;
1454 /*
1455 * If mp is set, just pass back the mbufs.
1456 * Otherwise copy them out via the uio, then free.
1457 * Sockbuf must be consistent here (points to current mbuf,
1458 * it points to next record) when we drop priority;
1459 * we must note any additions to the sockbuf when we
1460 * block interrupts again.
1461 */
1462 if (mp == NULL) {
1463 SBLASTRECORDCHK(&so->so_rcv, "soreceive uiomove");
1464 SBLASTMBUFCHK(&so->so_rcv, "soreceive uiomove");
1465 sounlock(so);
1466 splx(s);
1467 error = uiomove(mtod(m, char *) + moff, len, uio);
1468 s = splsoftnet();
1469 solock(so);
1470 if (error != 0) {
1471 /*
1472 * If any part of the record has been removed
1473 * (such as the MT_SONAME mbuf, which will
1474 * happen when PR_ADDR, and thus also
1475 * PR_ATOMIC, is set), then drop the entire
1476 * record to maintain the atomicity of the
1477 * receive operation.
1478 *
1479 * This avoids a later panic("receive 1a")
1480 * when compiled with DIAGNOSTIC.
1481 */
1482 if (m && mbuf_removed && atomic)
1483 (void) sbdroprecord(&so->so_rcv);
1484
1485 goto release;
1486 }
1487 } else
1488 uio->uio_resid -= len;
1489 if (len == m->m_len - moff) {
1490 if (m->m_flags & M_EOR)
1491 flags |= MSG_EOR;
1492 #ifdef SCTP
1493 if (m->m_flags & M_NOTIFICATION)
1494 flags |= MSG_NOTIFICATION;
1495 #endif /* SCTP */
1496 if (flags & MSG_PEEK) {
1497 m = m->m_next;
1498 moff = 0;
1499 } else {
1500 nextrecord = m->m_nextpkt;
1501 sbfree(&so->so_rcv, m);
1502 if (mp) {
1503 *mp = m;
1504 mp = &m->m_next;
1505 so->so_rcv.sb_mb = m = m->m_next;
1506 *mp = NULL;
1507 } else {
1508 MFREE(m, so->so_rcv.sb_mb);
1509 m = so->so_rcv.sb_mb;
1510 }
1511 /*
1512 * If m != NULL, we also know that
1513 * so->so_rcv.sb_mb != NULL.
1514 */
1515 KASSERT(so->so_rcv.sb_mb == m);
1516 if (m) {
1517 m->m_nextpkt = nextrecord;
1518 if (nextrecord == NULL)
1519 so->so_rcv.sb_lastrecord = m;
1520 } else {
1521 so->so_rcv.sb_mb = nextrecord;
1522 SB_EMPTY_FIXUP(&so->so_rcv);
1523 }
1524 SBLASTRECORDCHK(&so->so_rcv, "soreceive 3");
1525 SBLASTMBUFCHK(&so->so_rcv, "soreceive 3");
1526 }
1527 } else if (flags & MSG_PEEK)
1528 moff += len;
1529 else {
1530 if (mp != NULL) {
1531 mt = m_copym(m, 0, len, M_NOWAIT);
1532 if (__predict_false(mt == NULL)) {
1533 sounlock(so);
1534 mt = m_copym(m, 0, len, M_WAIT);
1535 solock(so);
1536 }
1537 *mp = mt;
1538 }
1539 m->m_data += len;
1540 m->m_len -= len;
1541 so->so_rcv.sb_cc -= len;
1542 }
1543 if (so->so_oobmark) {
1544 if ((flags & MSG_PEEK) == 0) {
1545 so->so_oobmark -= len;
1546 if (so->so_oobmark == 0) {
1547 so->so_state |= SS_RCVATMARK;
1548 break;
1549 }
1550 } else {
1551 offset += len;
1552 if (offset == so->so_oobmark)
1553 break;
1554 }
1555 }
1556 if (flags & MSG_EOR)
1557 break;
1558 /*
1559 * If the MSG_WAITALL flag is set (for non-atomic socket),
1560 * we must not quit until "uio->uio_resid == 0" or an error
1561 * termination. If a signal/timeout occurs, return
1562 * with a short count but without error.
1563 * Keep sockbuf locked against other readers.
1564 */
1565 while (flags & MSG_WAITALL && m == NULL && uio->uio_resid > 0 &&
1566 !sosendallatonce(so) && !nextrecord) {
1567 if (so->so_error || so->so_state & SS_CANTRCVMORE)
1568 break;
1569 /*
1570 * If we are peeking and the socket receive buffer is
1571 * full, stop since we can't get more data to peek at.
1572 */
1573 if ((flags & MSG_PEEK) && sbspace(&so->so_rcv) <= 0)
1574 break;
1575 /*
1576 * If we've drained the socket buffer, tell the
1577 * protocol in case it needs to do something to
1578 * get it filled again.
1579 */
1580 if ((pr->pr_flags & PR_WANTRCVD) && so->so_pcb)
1581 (*pr->pr_usrreqs->pr_rcvd)(so, flags, l);
1582 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 2");
1583 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 2");
1584 if (wakeup_state & SS_RESTARTSYS)
1585 error = ERESTART;
1586 else
1587 error = sbwait(&so->so_rcv);
1588 if (error != 0) {
1589 sbunlock(&so->so_rcv);
1590 sounlock(so);
1591 splx(s);
1592 return 0;
1593 }
1594 if ((m = so->so_rcv.sb_mb) != NULL)
1595 nextrecord = m->m_nextpkt;
1596 wakeup_state = so->so_state;
1597 }
1598 }
1599
1600 if (m && atomic) {
1601 flags |= MSG_TRUNC;
1602 if ((flags & MSG_PEEK) == 0)
1603 (void) sbdroprecord(&so->so_rcv);
1604 }
1605 if ((flags & MSG_PEEK) == 0) {
1606 if (m == NULL) {
1607 /*
1608 * First part is an inline SB_EMPTY_FIXUP(). Second
1609 * part makes sure sb_lastrecord is up-to-date if
1610 * there is still data in the socket buffer.
1611 */
1612 so->so_rcv.sb_mb = nextrecord;
1613 if (so->so_rcv.sb_mb == NULL) {
1614 so->so_rcv.sb_mbtail = NULL;
1615 so->so_rcv.sb_lastrecord = NULL;
1616 } else if (nextrecord->m_nextpkt == NULL)
1617 so->so_rcv.sb_lastrecord = nextrecord;
1618 }
1619 SBLASTRECORDCHK(&so->so_rcv, "soreceive 4");
1620 SBLASTMBUFCHK(&so->so_rcv, "soreceive 4");
1621 if (pr->pr_flags & PR_WANTRCVD && so->so_pcb)
1622 (*pr->pr_usrreqs->pr_rcvd)(so, flags, l);
1623 }
1624 if (orig_resid == uio->uio_resid && orig_resid &&
1625 (flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) {
1626 sbunlock(&so->so_rcv);
1627 goto restart;
1628 }
1629
1630 if (flagsp != NULL)
1631 *flagsp |= flags;
1632 release:
1633 sbunlock(&so->so_rcv);
1634 sounlock(so);
1635 splx(s);
1636 return error;
1637 }
1638
1639 int
soshutdown(struct socket * so,int how)1640 soshutdown(struct socket *so, int how)
1641 {
1642 const struct protosw *pr;
1643 int error;
1644
1645 KASSERT(solocked(so));
1646
1647 pr = so->so_proto;
1648 if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR))
1649 return (EINVAL);
1650
1651 if (how == SHUT_RD || how == SHUT_RDWR) {
1652 sorflush(so);
1653 error = 0;
1654 }
1655 if (how == SHUT_WR || how == SHUT_RDWR)
1656 error = (*pr->pr_usrreqs->pr_shutdown)(so);
1657
1658 return error;
1659 }
1660
1661 void
sorestart(struct socket * so)1662 sorestart(struct socket *so)
1663 {
1664 /*
1665 * An application has called close() on an fd on which another
1666 * of its threads has called a socket system call.
1667 * Mark this and wake everyone up, and code that would block again
1668 * instead returns ERESTART.
1669 * On system call re-entry the fd is validated and EBADF returned.
1670 * Any other fd will block again on the 2nd syscall.
1671 */
1672 solock(so);
1673 so->so_state |= SS_RESTARTSYS;
1674 cv_broadcast(&so->so_cv);
1675 cv_broadcast(&so->so_snd.sb_cv);
1676 cv_broadcast(&so->so_rcv.sb_cv);
1677 sounlock(so);
1678 }
1679
1680 void
sorflush(struct socket * so)1681 sorflush(struct socket *so)
1682 {
1683 struct sockbuf *sb, asb;
1684 const struct protosw *pr;
1685
1686 KASSERT(solocked(so));
1687
1688 sb = &so->so_rcv;
1689 pr = so->so_proto;
1690 socantrcvmore(so);
1691 sb->sb_flags |= SB_NOINTR;
1692 (void )sblock(sb, M_WAITOK);
1693 sbunlock(sb);
1694 asb = *sb;
1695 /*
1696 * Clear most of the sockbuf structure, but leave some of the
1697 * fields valid.
1698 */
1699 memset(&sb->sb_startzero, 0,
1700 sizeof(*sb) - offsetof(struct sockbuf, sb_startzero));
1701 if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose) {
1702 sounlock(so);
1703 (*pr->pr_domain->dom_dispose)(asb.sb_mb);
1704 solock(so);
1705 }
1706 sbrelease(&asb, so);
1707 }
1708
1709 /*
1710 * internal set SOL_SOCKET options
1711 */
1712 static int
sosetopt1(struct socket * so,const struct sockopt * sopt)1713 sosetopt1(struct socket *so, const struct sockopt *sopt)
1714 {
1715 int error = EINVAL, opt;
1716 int optval = 0; /* XXX: gcc */
1717 struct linger l;
1718 struct timeval tv;
1719
1720 switch ((opt = sopt->sopt_name)) {
1721
1722 case SO_ACCEPTFILTER:
1723 error = accept_filt_setopt(so, sopt);
1724 KASSERT(solocked(so));
1725 break;
1726
1727 case SO_LINGER:
1728 error = sockopt_get(sopt, &l, sizeof(l));
1729 solock(so);
1730 if (error)
1731 break;
1732 if (l.l_linger < 0 || l.l_linger > USHRT_MAX ||
1733 l.l_linger > (INT_MAX / hz)) {
1734 error = EDOM;
1735 break;
1736 }
1737 so->so_linger = l.l_linger;
1738 if (l.l_onoff)
1739 so->so_options |= SO_LINGER;
1740 else
1741 so->so_options &= ~SO_LINGER;
1742 break;
1743
1744 case SO_DEBUG:
1745 case SO_KEEPALIVE:
1746 case SO_DONTROUTE:
1747 case SO_USELOOPBACK:
1748 case SO_BROADCAST:
1749 case SO_REUSEADDR:
1750 case SO_REUSEPORT:
1751 case SO_OOBINLINE:
1752 case SO_TIMESTAMP:
1753 case SO_NOSIGPIPE:
1754 #ifdef SO_OTIMESTAMP
1755 case SO_OTIMESTAMP:
1756 #endif
1757 error = sockopt_getint(sopt, &optval);
1758 solock(so);
1759 if (error)
1760 break;
1761 if (optval)
1762 so->so_options |= opt;
1763 else
1764 so->so_options &= ~opt;
1765 break;
1766
1767 case SO_SNDBUF:
1768 case SO_RCVBUF:
1769 case SO_SNDLOWAT:
1770 case SO_RCVLOWAT:
1771 error = sockopt_getint(sopt, &optval);
1772 solock(so);
1773 if (error)
1774 break;
1775
1776 /*
1777 * Values < 1 make no sense for any of these
1778 * options, so disallow them.
1779 */
1780 if (optval < 1) {
1781 error = EINVAL;
1782 break;
1783 }
1784
1785 switch (opt) {
1786 case SO_SNDBUF:
1787 if (sbreserve(&so->so_snd, (u_long)optval, so) == 0) {
1788 error = ENOBUFS;
1789 break;
1790 }
1791 so->so_snd.sb_flags &= ~SB_AUTOSIZE;
1792 break;
1793
1794 case SO_RCVBUF:
1795 if (sbreserve(&so->so_rcv, (u_long)optval, so) == 0) {
1796 error = ENOBUFS;
1797 break;
1798 }
1799 so->so_rcv.sb_flags &= ~SB_AUTOSIZE;
1800 break;
1801
1802 /*
1803 * Make sure the low-water is never greater than
1804 * the high-water.
1805 */
1806 case SO_SNDLOWAT:
1807 if (optval > so->so_snd.sb_hiwat)
1808 optval = so->so_snd.sb_hiwat;
1809
1810 so->so_snd.sb_lowat = optval;
1811 break;
1812
1813 case SO_RCVLOWAT:
1814 if (optval > so->so_rcv.sb_hiwat)
1815 optval = so->so_rcv.sb_hiwat;
1816
1817 so->so_rcv.sb_lowat = optval;
1818 break;
1819 }
1820 break;
1821
1822 #ifdef COMPAT_50
1823 case SO_OSNDTIMEO:
1824 case SO_ORCVTIMEO: {
1825 struct timeval50 otv;
1826 error = sockopt_get(sopt, &otv, sizeof(otv));
1827 if (error) {
1828 solock(so);
1829 break;
1830 }
1831 timeval50_to_timeval(&otv, &tv);
1832 opt = opt == SO_OSNDTIMEO ? SO_SNDTIMEO : SO_RCVTIMEO;
1833 error = 0;
1834 /*FALLTHROUGH*/
1835 }
1836 #endif /* COMPAT_50 */
1837
1838 case SO_SNDTIMEO:
1839 case SO_RCVTIMEO:
1840 if (error)
1841 error = sockopt_get(sopt, &tv, sizeof(tv));
1842 solock(so);
1843 if (error)
1844 break;
1845
1846 if (tv.tv_sec > (INT_MAX - tv.tv_usec / tick) / hz) {
1847 error = EDOM;
1848 break;
1849 }
1850
1851 optval = tv.tv_sec * hz + tv.tv_usec / tick;
1852 if (optval == 0 && tv.tv_usec != 0)
1853 optval = 1;
1854
1855 switch (opt) {
1856 case SO_SNDTIMEO:
1857 so->so_snd.sb_timeo = optval;
1858 break;
1859 case SO_RCVTIMEO:
1860 so->so_rcv.sb_timeo = optval;
1861 break;
1862 }
1863 break;
1864
1865 default:
1866 solock(so);
1867 error = ENOPROTOOPT;
1868 break;
1869 }
1870 KASSERT(solocked(so));
1871 return error;
1872 }
1873
1874 int
sosetopt(struct socket * so,struct sockopt * sopt)1875 sosetopt(struct socket *so, struct sockopt *sopt)
1876 {
1877 int error, prerr;
1878
1879 if (sopt->sopt_level == SOL_SOCKET) {
1880 error = sosetopt1(so, sopt);
1881 KASSERT(solocked(so));
1882 } else {
1883 error = ENOPROTOOPT;
1884 solock(so);
1885 }
1886
1887 if ((error == 0 || error == ENOPROTOOPT) &&
1888 so->so_proto != NULL && so->so_proto->pr_ctloutput != NULL) {
1889 /* give the protocol stack a shot */
1890 prerr = (*so->so_proto->pr_ctloutput)(PRCO_SETOPT, so, sopt);
1891 if (prerr == 0)
1892 error = 0;
1893 else if (prerr != ENOPROTOOPT)
1894 error = prerr;
1895 }
1896 sounlock(so);
1897 return error;
1898 }
1899
1900 /*
1901 * so_setsockopt() is a wrapper providing a sockopt structure for sosetopt()
1902 */
1903 int
so_setsockopt(struct lwp * l,struct socket * so,int level,int name,const void * val,size_t valsize)1904 so_setsockopt(struct lwp *l, struct socket *so, int level, int name,
1905 const void *val, size_t valsize)
1906 {
1907 struct sockopt sopt;
1908 int error;
1909
1910 KASSERT(valsize == 0 || val != NULL);
1911
1912 sockopt_init(&sopt, level, name, valsize);
1913 sockopt_set(&sopt, val, valsize);
1914
1915 error = sosetopt(so, &sopt);
1916
1917 sockopt_destroy(&sopt);
1918
1919 return error;
1920 }
1921
1922 /*
1923 * internal get SOL_SOCKET options
1924 */
1925 static int
sogetopt1(struct socket * so,struct sockopt * sopt)1926 sogetopt1(struct socket *so, struct sockopt *sopt)
1927 {
1928 int error, optval, opt;
1929 struct linger l;
1930 struct timeval tv;
1931
1932 switch ((opt = sopt->sopt_name)) {
1933
1934 case SO_ACCEPTFILTER:
1935 error = accept_filt_getopt(so, sopt);
1936 break;
1937
1938 case SO_LINGER:
1939 l.l_onoff = (so->so_options & SO_LINGER) ? 1 : 0;
1940 l.l_linger = so->so_linger;
1941
1942 error = sockopt_set(sopt, &l, sizeof(l));
1943 break;
1944
1945 case SO_USELOOPBACK:
1946 case SO_DONTROUTE:
1947 case SO_DEBUG:
1948 case SO_KEEPALIVE:
1949 case SO_REUSEADDR:
1950 case SO_REUSEPORT:
1951 case SO_BROADCAST:
1952 case SO_OOBINLINE:
1953 case SO_TIMESTAMP:
1954 case SO_NOSIGPIPE:
1955 #ifdef SO_OTIMESTAMP
1956 case SO_OTIMESTAMP:
1957 #endif
1958 case SO_ACCEPTCONN:
1959 error = sockopt_setint(sopt, (so->so_options & opt) ? 1 : 0);
1960 break;
1961
1962 case SO_TYPE:
1963 error = sockopt_setint(sopt, so->so_type);
1964 break;
1965
1966 case SO_ERROR:
1967 error = sockopt_setint(sopt, so->so_error);
1968 so->so_error = 0;
1969 break;
1970
1971 case SO_SNDBUF:
1972 error = sockopt_setint(sopt, so->so_snd.sb_hiwat);
1973 break;
1974
1975 case SO_RCVBUF:
1976 error = sockopt_setint(sopt, so->so_rcv.sb_hiwat);
1977 break;
1978
1979 case SO_SNDLOWAT:
1980 error = sockopt_setint(sopt, so->so_snd.sb_lowat);
1981 break;
1982
1983 case SO_RCVLOWAT:
1984 error = sockopt_setint(sopt, so->so_rcv.sb_lowat);
1985 break;
1986
1987 #ifdef COMPAT_50
1988 case SO_OSNDTIMEO:
1989 case SO_ORCVTIMEO: {
1990 struct timeval50 otv;
1991
1992 optval = (opt == SO_OSNDTIMEO ?
1993 so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
1994
1995 otv.tv_sec = optval / hz;
1996 otv.tv_usec = (optval % hz) * tick;
1997
1998 error = sockopt_set(sopt, &otv, sizeof(otv));
1999 break;
2000 }
2001 #endif /* COMPAT_50 */
2002
2003 case SO_SNDTIMEO:
2004 case SO_RCVTIMEO:
2005 optval = (opt == SO_SNDTIMEO ?
2006 so->so_snd.sb_timeo : so->so_rcv.sb_timeo);
2007
2008 tv.tv_sec = optval / hz;
2009 tv.tv_usec = (optval % hz) * tick;
2010
2011 error = sockopt_set(sopt, &tv, sizeof(tv));
2012 break;
2013
2014 case SO_OVERFLOWED:
2015 error = sockopt_setint(sopt, so->so_rcv.sb_overflowed);
2016 break;
2017
2018 default:
2019 error = ENOPROTOOPT;
2020 break;
2021 }
2022
2023 return (error);
2024 }
2025
2026 int
sogetopt(struct socket * so,struct sockopt * sopt)2027 sogetopt(struct socket *so, struct sockopt *sopt)
2028 {
2029 int error;
2030
2031 solock(so);
2032 if (sopt->sopt_level != SOL_SOCKET) {
2033 if (so->so_proto && so->so_proto->pr_ctloutput) {
2034 error = ((*so->so_proto->pr_ctloutput)
2035 (PRCO_GETOPT, so, sopt));
2036 } else
2037 error = (ENOPROTOOPT);
2038 } else {
2039 error = sogetopt1(so, sopt);
2040 }
2041 sounlock(so);
2042 return (error);
2043 }
2044
2045 /*
2046 * alloc sockopt data buffer buffer
2047 * - will be released at destroy
2048 */
2049 static int
sockopt_alloc(struct sockopt * sopt,size_t len,km_flag_t kmflag)2050 sockopt_alloc(struct sockopt *sopt, size_t len, km_flag_t kmflag)
2051 {
2052
2053 KASSERT(sopt->sopt_size == 0);
2054
2055 if (len > sizeof(sopt->sopt_buf)) {
2056 sopt->sopt_data = kmem_zalloc(len, kmflag);
2057 if (sopt->sopt_data == NULL)
2058 return ENOMEM;
2059 } else
2060 sopt->sopt_data = sopt->sopt_buf;
2061
2062 sopt->sopt_size = len;
2063 return 0;
2064 }
2065
2066 /*
2067 * initialise sockopt storage
2068 * - MAY sleep during allocation
2069 */
2070 void
sockopt_init(struct sockopt * sopt,int level,int name,size_t size)2071 sockopt_init(struct sockopt *sopt, int level, int name, size_t size)
2072 {
2073
2074 memset(sopt, 0, sizeof(*sopt));
2075
2076 sopt->sopt_level = level;
2077 sopt->sopt_name = name;
2078 (void)sockopt_alloc(sopt, size, KM_SLEEP);
2079 }
2080
2081 /*
2082 * destroy sockopt storage
2083 * - will release any held memory references
2084 */
2085 void
sockopt_destroy(struct sockopt * sopt)2086 sockopt_destroy(struct sockopt *sopt)
2087 {
2088
2089 if (sopt->sopt_data != sopt->sopt_buf)
2090 kmem_free(sopt->sopt_data, sopt->sopt_size);
2091
2092 memset(sopt, 0, sizeof(*sopt));
2093 }
2094
2095 /*
2096 * set sockopt value
2097 * - value is copied into sockopt
2098 * - memory is allocated when necessary, will not sleep
2099 */
2100 int
sockopt_set(struct sockopt * sopt,const void * buf,size_t len)2101 sockopt_set(struct sockopt *sopt, const void *buf, size_t len)
2102 {
2103 int error;
2104
2105 if (sopt->sopt_size == 0) {
2106 error = sockopt_alloc(sopt, len, KM_NOSLEEP);
2107 if (error)
2108 return error;
2109 }
2110
2111 KASSERT(sopt->sopt_size == len);
2112 memcpy(sopt->sopt_data, buf, len);
2113 return 0;
2114 }
2115
2116 /*
2117 * common case of set sockopt integer value
2118 */
2119 int
sockopt_setint(struct sockopt * sopt,int val)2120 sockopt_setint(struct sockopt *sopt, int val)
2121 {
2122
2123 return sockopt_set(sopt, &val, sizeof(int));
2124 }
2125
2126 /*
2127 * get sockopt value
2128 * - correct size must be given
2129 */
2130 int
sockopt_get(const struct sockopt * sopt,void * buf,size_t len)2131 sockopt_get(const struct sockopt *sopt, void *buf, size_t len)
2132 {
2133
2134 if (sopt->sopt_size != len)
2135 return EINVAL;
2136
2137 memcpy(buf, sopt->sopt_data, len);
2138 return 0;
2139 }
2140
2141 /*
2142 * common case of get sockopt integer value
2143 */
2144 int
sockopt_getint(const struct sockopt * sopt,int * valp)2145 sockopt_getint(const struct sockopt *sopt, int *valp)
2146 {
2147
2148 return sockopt_get(sopt, valp, sizeof(int));
2149 }
2150
2151 /*
2152 * set sockopt value from mbuf
2153 * - ONLY for legacy code
2154 * - mbuf is released by sockopt
2155 * - will not sleep
2156 */
2157 int
sockopt_setmbuf(struct sockopt * sopt,struct mbuf * m)2158 sockopt_setmbuf(struct sockopt *sopt, struct mbuf *m)
2159 {
2160 size_t len;
2161 int error;
2162
2163 len = m_length(m);
2164
2165 if (sopt->sopt_size == 0) {
2166 error = sockopt_alloc(sopt, len, KM_NOSLEEP);
2167 if (error)
2168 return error;
2169 }
2170
2171 KASSERT(sopt->sopt_size == len);
2172 m_copydata(m, 0, len, sopt->sopt_data);
2173 m_freem(m);
2174
2175 return 0;
2176 }
2177
2178 /*
2179 * get sockopt value into mbuf
2180 * - ONLY for legacy code
2181 * - mbuf to be released by the caller
2182 * - will not sleep
2183 */
2184 struct mbuf *
sockopt_getmbuf(const struct sockopt * sopt)2185 sockopt_getmbuf(const struct sockopt *sopt)
2186 {
2187 struct mbuf *m;
2188
2189 if (sopt->sopt_size > MCLBYTES)
2190 return NULL;
2191
2192 m = m_get(M_DONTWAIT, MT_SOOPTS);
2193 if (m == NULL)
2194 return NULL;
2195
2196 if (sopt->sopt_size > MLEN) {
2197 MCLGET(m, M_DONTWAIT);
2198 if ((m->m_flags & M_EXT) == 0) {
2199 m_free(m);
2200 return NULL;
2201 }
2202 }
2203
2204 memcpy(mtod(m, void *), sopt->sopt_data, sopt->sopt_size);
2205 m->m_len = sopt->sopt_size;
2206
2207 return m;
2208 }
2209
2210 void
sohasoutofband(struct socket * so)2211 sohasoutofband(struct socket *so)
2212 {
2213
2214 fownsignal(so->so_pgid, SIGURG, POLL_PRI, POLLPRI|POLLRDBAND, so);
2215 selnotify(&so->so_rcv.sb_sel, POLLPRI | POLLRDBAND, NOTE_SUBMIT);
2216 }
2217
2218 static void
filt_sordetach(struct knote * kn)2219 filt_sordetach(struct knote *kn)
2220 {
2221 struct socket *so;
2222
2223 so = ((file_t *)kn->kn_obj)->f_socket;
2224 solock(so);
2225 SLIST_REMOVE(&so->so_rcv.sb_sel.sel_klist, kn, knote, kn_selnext);
2226 if (SLIST_EMPTY(&so->so_rcv.sb_sel.sel_klist))
2227 so->so_rcv.sb_flags &= ~SB_KNOTE;
2228 sounlock(so);
2229 }
2230
2231 /*ARGSUSED*/
2232 static int
filt_soread(struct knote * kn,long hint)2233 filt_soread(struct knote *kn, long hint)
2234 {
2235 struct socket *so;
2236 int rv;
2237
2238 so = ((file_t *)kn->kn_obj)->f_socket;
2239 if (hint != NOTE_SUBMIT)
2240 solock(so);
2241 kn->kn_data = so->so_rcv.sb_cc;
2242 if (so->so_state & SS_CANTRCVMORE) {
2243 kn->kn_flags |= EV_EOF;
2244 kn->kn_fflags = so->so_error;
2245 rv = 1;
2246 } else if (so->so_error) /* temporary udp error */
2247 rv = 1;
2248 else if (kn->kn_sfflags & NOTE_LOWAT)
2249 rv = (kn->kn_data >= kn->kn_sdata);
2250 else
2251 rv = (kn->kn_data >= so->so_rcv.sb_lowat);
2252 if (hint != NOTE_SUBMIT)
2253 sounlock(so);
2254 return rv;
2255 }
2256
2257 static void
filt_sowdetach(struct knote * kn)2258 filt_sowdetach(struct knote *kn)
2259 {
2260 struct socket *so;
2261
2262 so = ((file_t *)kn->kn_obj)->f_socket;
2263 solock(so);
2264 SLIST_REMOVE(&so->so_snd.sb_sel.sel_klist, kn, knote, kn_selnext);
2265 if (SLIST_EMPTY(&so->so_snd.sb_sel.sel_klist))
2266 so->so_snd.sb_flags &= ~SB_KNOTE;
2267 sounlock(so);
2268 }
2269
2270 /*ARGSUSED*/
2271 static int
filt_sowrite(struct knote * kn,long hint)2272 filt_sowrite(struct knote *kn, long hint)
2273 {
2274 struct socket *so;
2275 int rv;
2276
2277 so = ((file_t *)kn->kn_obj)->f_socket;
2278 if (hint != NOTE_SUBMIT)
2279 solock(so);
2280 kn->kn_data = sbspace(&so->so_snd);
2281 if (so->so_state & SS_CANTSENDMORE) {
2282 kn->kn_flags |= EV_EOF;
2283 kn->kn_fflags = so->so_error;
2284 rv = 1;
2285 } else if (so->so_error) /* temporary udp error */
2286 rv = 1;
2287 else if (((so->so_state & SS_ISCONNECTED) == 0) &&
2288 (so->so_proto->pr_flags & PR_CONNREQUIRED))
2289 rv = 0;
2290 else if (kn->kn_sfflags & NOTE_LOWAT)
2291 rv = (kn->kn_data >= kn->kn_sdata);
2292 else
2293 rv = (kn->kn_data >= so->so_snd.sb_lowat);
2294 if (hint != NOTE_SUBMIT)
2295 sounlock(so);
2296 return rv;
2297 }
2298
2299 /*ARGSUSED*/
2300 static int
filt_solisten(struct knote * kn,long hint)2301 filt_solisten(struct knote *kn, long hint)
2302 {
2303 struct socket *so;
2304 int rv;
2305
2306 so = ((file_t *)kn->kn_obj)->f_socket;
2307
2308 /*
2309 * Set kn_data to number of incoming connections, not
2310 * counting partial (incomplete) connections.
2311 */
2312 if (hint != NOTE_SUBMIT)
2313 solock(so);
2314 kn->kn_data = so->so_qlen;
2315 rv = (kn->kn_data > 0);
2316 if (hint != NOTE_SUBMIT)
2317 sounlock(so);
2318 return rv;
2319 }
2320
2321 static const struct filterops solisten_filtops =
2322 { 1, NULL, filt_sordetach, filt_solisten };
2323 static const struct filterops soread_filtops =
2324 { 1, NULL, filt_sordetach, filt_soread };
2325 static const struct filterops sowrite_filtops =
2326 { 1, NULL, filt_sowdetach, filt_sowrite };
2327
2328 int
soo_kqfilter(struct file * fp,struct knote * kn)2329 soo_kqfilter(struct file *fp, struct knote *kn)
2330 {
2331 struct socket *so;
2332 struct sockbuf *sb;
2333
2334 so = ((file_t *)kn->kn_obj)->f_socket;
2335 solock(so);
2336 switch (kn->kn_filter) {
2337 case EVFILT_READ:
2338 if (so->so_options & SO_ACCEPTCONN)
2339 kn->kn_fop = &solisten_filtops;
2340 else
2341 kn->kn_fop = &soread_filtops;
2342 sb = &so->so_rcv;
2343 break;
2344 case EVFILT_WRITE:
2345 kn->kn_fop = &sowrite_filtops;
2346 sb = &so->so_snd;
2347 break;
2348 default:
2349 sounlock(so);
2350 return (EINVAL);
2351 }
2352 SLIST_INSERT_HEAD(&sb->sb_sel.sel_klist, kn, kn_selnext);
2353 sb->sb_flags |= SB_KNOTE;
2354 sounlock(so);
2355 return (0);
2356 }
2357
2358 static int
sodopoll(struct socket * so,int events)2359 sodopoll(struct socket *so, int events)
2360 {
2361 int revents;
2362
2363 revents = 0;
2364
2365 if (events & (POLLIN | POLLRDNORM))
2366 if (soreadable(so))
2367 revents |= events & (POLLIN | POLLRDNORM);
2368
2369 if (events & (POLLOUT | POLLWRNORM))
2370 if (sowritable(so))
2371 revents |= events & (POLLOUT | POLLWRNORM);
2372
2373 if (events & (POLLPRI | POLLRDBAND))
2374 if (so->so_oobmark || (so->so_state & SS_RCVATMARK))
2375 revents |= events & (POLLPRI | POLLRDBAND);
2376
2377 return revents;
2378 }
2379
2380 int
sopoll(struct socket * so,int events)2381 sopoll(struct socket *so, int events)
2382 {
2383 int revents = 0;
2384
2385 #ifndef DIAGNOSTIC
2386 /*
2387 * Do a quick, unlocked check in expectation that the socket
2388 * will be ready for I/O. Don't do this check if DIAGNOSTIC,
2389 * as the solocked() assertions will fail.
2390 */
2391 if ((revents = sodopoll(so, events)) != 0)
2392 return revents;
2393 #endif
2394
2395 solock(so);
2396 if ((revents = sodopoll(so, events)) == 0) {
2397 if (events & (POLLIN | POLLPRI | POLLRDNORM | POLLRDBAND)) {
2398 selrecord(curlwp, &so->so_rcv.sb_sel);
2399 so->so_rcv.sb_flags |= SB_NOTIFY;
2400 }
2401
2402 if (events & (POLLOUT | POLLWRNORM)) {
2403 selrecord(curlwp, &so->so_snd.sb_sel);
2404 so->so_snd.sb_flags |= SB_NOTIFY;
2405 }
2406 }
2407 sounlock(so);
2408
2409 return revents;
2410 }
2411
2412
2413 #include <sys/sysctl.h>
2414
2415 static int sysctl_kern_somaxkva(SYSCTLFN_PROTO);
2416 static int sysctl_kern_sbmax(SYSCTLFN_PROTO);
2417
2418 /*
2419 * sysctl helper routine for kern.somaxkva. ensures that the given
2420 * value is not too small.
2421 * (XXX should we maybe make sure it's not too large as well?)
2422 */
2423 static int
sysctl_kern_somaxkva(SYSCTLFN_ARGS)2424 sysctl_kern_somaxkva(SYSCTLFN_ARGS)
2425 {
2426 int error, new_somaxkva;
2427 struct sysctlnode node;
2428
2429 new_somaxkva = somaxkva;
2430 node = *rnode;
2431 node.sysctl_data = &new_somaxkva;
2432 error = sysctl_lookup(SYSCTLFN_CALL(&node));
2433 if (error || newp == NULL)
2434 return (error);
2435
2436 if (new_somaxkva < (16 * 1024 * 1024)) /* sanity */
2437 return (EINVAL);
2438
2439 mutex_enter(&so_pendfree_lock);
2440 somaxkva = new_somaxkva;
2441 cv_broadcast(&socurkva_cv);
2442 mutex_exit(&so_pendfree_lock);
2443
2444 return (error);
2445 }
2446
2447 /*
2448 * sysctl helper routine for kern.sbmax. Basically just ensures that
2449 * any new value is not too small.
2450 */
2451 static int
sysctl_kern_sbmax(SYSCTLFN_ARGS)2452 sysctl_kern_sbmax(SYSCTLFN_ARGS)
2453 {
2454 int error, new_sbmax;
2455 struct sysctlnode node;
2456
2457 new_sbmax = sb_max;
2458 node = *rnode;
2459 node.sysctl_data = &new_sbmax;
2460 error = sysctl_lookup(SYSCTLFN_CALL(&node));
2461 if (error || newp == NULL)
2462 return (error);
2463
2464 KERNEL_LOCK(1, NULL);
2465 error = sb_max_set(new_sbmax);
2466 KERNEL_UNLOCK_ONE(NULL);
2467
2468 return (error);
2469 }
2470
2471 static void
sysctl_kern_socket_setup(void)2472 sysctl_kern_socket_setup(void)
2473 {
2474
2475 KASSERT(socket_sysctllog == NULL);
2476
2477 sysctl_createv(&socket_sysctllog, 0, NULL, NULL,
2478 CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
2479 CTLTYPE_INT, "somaxkva",
2480 SYSCTL_DESCR("Maximum amount of kernel memory to be "
2481 "used for socket buffers"),
2482 sysctl_kern_somaxkva, 0, NULL, 0,
2483 CTL_KERN, KERN_SOMAXKVA, CTL_EOL);
2484
2485 sysctl_createv(&socket_sysctllog, 0, NULL, NULL,
2486 CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
2487 CTLTYPE_INT, "sbmax",
2488 SYSCTL_DESCR("Maximum socket buffer size"),
2489 sysctl_kern_sbmax, 0, NULL, 0,
2490 CTL_KERN, KERN_SBMAX, CTL_EOL);
2491 }
2492