if.c revision 1.257 1 /* $NetBSD: if.c,v 1.257 2011/11/16 06:09:37 dyoung Exp $ */
2
3 /*-
4 * Copyright (c) 1999, 2000, 2001, 2008 The NetBSD Foundation, Inc.
5 * All rights reserved.
6 *
7 * This code is derived from software contributed to The NetBSD Foundation
8 * by William Studenmund and Jason R. Thorpe.
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) 1995, 1996, 1997, and 1998 WIDE Project.
34 * All rights reserved.
35 *
36 * Redistribution and use in source and binary forms, with or without
37 * modification, are permitted provided that the following conditions
38 * are met:
39 * 1. Redistributions of source code must retain the above copyright
40 * notice, this list of conditions and the following disclaimer.
41 * 2. Redistributions in binary form must reproduce the above copyright
42 * notice, this list of conditions and the following disclaimer in the
43 * documentation and/or other materials provided with the distribution.
44 * 3. Neither the name of the project nor the names of its contributors
45 * may be used to endorse or promote products derived from this software
46 * without specific prior written permission.
47 *
48 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
49 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
50 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
51 * ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
52 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
53 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
54 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
55 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
56 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
57 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
58 * SUCH DAMAGE.
59 */
60
61 /*
62 * Copyright (c) 1980, 1986, 1993
63 * The Regents of the University of California. All rights reserved.
64 *
65 * Redistribution and use in source and binary forms, with or without
66 * modification, are permitted provided that the following conditions
67 * are met:
68 * 1. Redistributions of source code must retain the above copyright
69 * notice, this list of conditions and the following disclaimer.
70 * 2. Redistributions in binary form must reproduce the above copyright
71 * notice, this list of conditions and the following disclaimer in the
72 * documentation and/or other materials provided with the distribution.
73 * 3. Neither the name of the University nor the names of its contributors
74 * may be used to endorse or promote products derived from this software
75 * without specific prior written permission.
76 *
77 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
78 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
79 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
80 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
81 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
82 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
83 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
84 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
85 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
86 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
87 * SUCH DAMAGE.
88 *
89 * @(#)if.c 8.5 (Berkeley) 1/9/95
90 */
91
92 #include <sys/cdefs.h>
93 __KERNEL_RCSID(0, "$NetBSD: if.c,v 1.257 2011/11/16 06:09:37 dyoung Exp $");
94
95 #include "opt_inet.h"
96
97 #include "opt_atalk.h"
98 #include "opt_natm.h"
99 #include "opt_pfil_hooks.h"
100
101 #include <sys/param.h>
102 #include <sys/mbuf.h>
103 #include <sys/systm.h>
104 #include <sys/callout.h>
105 #include <sys/proc.h>
106 #include <sys/socket.h>
107 #include <sys/socketvar.h>
108 #include <sys/domain.h>
109 #include <sys/protosw.h>
110 #include <sys/kernel.h>
111 #include <sys/ioctl.h>
112 #include <sys/sysctl.h>
113 #include <sys/syslog.h>
114 #include <sys/kauth.h>
115 #include <sys/kmem.h>
116
117 #include <net/if.h>
118 #include <net/if_dl.h>
119 #include <net/if_ether.h>
120 #include <net/if_media.h>
121 #include <net80211/ieee80211.h>
122 #include <net80211/ieee80211_ioctl.h>
123 #include <net/if_types.h>
124 #include <net/radix.h>
125 #include <net/route.h>
126 #include <net/netisr.h>
127 #ifdef NETATALK
128 #include <netatalk/at_extern.h>
129 #include <netatalk/at.h>
130 #endif
131 #include <net/pfil.h>
132
133 #ifdef INET6
134 #include <netinet/in.h>
135 #include <netinet6/in6_var.h>
136 #include <netinet6/nd6.h>
137 #endif
138
139 #include "carp.h"
140 #if NCARP > 0
141 #include <netinet/ip_carp.h>
142 #endif
143
144 #include <compat/sys/sockio.h>
145 #include <compat/sys/socket.h>
146
147 MALLOC_DEFINE(M_IFADDR, "ifaddr", "interface address");
148 MALLOC_DEFINE(M_IFMADDR, "ether_multi", "link-level multicast address");
149
150 int ifqmaxlen = IFQ_MAXLEN;
151 callout_t if_slowtimo_ch;
152
153 int netisr; /* scheduling bits for network */
154
155 static int if_rt_walktree(struct rtentry *, void *);
156
157 static struct if_clone *if_clone_lookup(const char *, int *);
158 static int if_clone_list(struct if_clonereq *);
159
160 static LIST_HEAD(, if_clone) if_cloners = LIST_HEAD_INITIALIZER(if_cloners);
161 static int if_cloners_count;
162
163 static uint64_t index_gen;
164 static kmutex_t index_gen_mtx;
165
166 #ifdef PFIL_HOOKS
167 struct pfil_head if_pfil; /* packet filtering hook for interfaces */
168 #endif
169
170 static kauth_listener_t if_listener;
171
172 static int ifioctl_attach(struct ifnet *);
173 static void ifioctl_detach(struct ifnet *);
174 static void ifnet_lock_enter(struct ifnet_lock *);
175 static void ifnet_lock_exit(struct ifnet_lock *);
176 static void if_detach_queues(struct ifnet *, struct ifqueue *);
177 static void sysctl_sndq_setup(struct sysctllog **, const char *,
178 struct ifaltq *);
179
180 #if defined(INET) || defined(INET6)
181 static void sysctl_net_ifq_setup(struct sysctllog **, int, const char *,
182 int, const char *, int, struct ifqueue *);
183 #endif
184
185 static int
186 if_listener_cb(kauth_cred_t cred, kauth_action_t action, void *cookie,
187 void *arg0, void *arg1, void *arg2, void *arg3)
188 {
189 int result;
190 enum kauth_network_req req;
191
192 result = KAUTH_RESULT_DEFER;
193 req = (enum kauth_network_req)arg1;
194
195 if (action != KAUTH_NETWORK_INTERFACE)
196 return result;
197
198 if ((req == KAUTH_REQ_NETWORK_INTERFACE_GET) ||
199 (req == KAUTH_REQ_NETWORK_INTERFACE_SET))
200 result = KAUTH_RESULT_ALLOW;
201
202 return result;
203 }
204
205 /*
206 * Network interface utility routines.
207 *
208 * Routines with ifa_ifwith* names take sockaddr *'s as
209 * parameters.
210 */
211 void
212 ifinit(void)
213 {
214 #ifdef INET
215 {extern struct ifqueue ipintrq;
216 sysctl_net_ifq_setup(NULL, PF_INET, "inet", IPPROTO_IP, "ip",
217 IPCTL_IFQ, &ipintrq);}
218 #endif /* INET */
219 #ifdef INET6
220 {extern struct ifqueue ip6intrq;
221 sysctl_net_ifq_setup(NULL, PF_INET6, "inet6", IPPROTO_IPV6, "ip6",
222 IPV6CTL_IFQ, &ip6intrq);}
223 #endif /* INET6 */
224
225 callout_init(&if_slowtimo_ch, 0);
226 if_slowtimo(NULL);
227
228 if_listener = kauth_listen_scope(KAUTH_SCOPE_NETWORK,
229 if_listener_cb, NULL);
230 }
231
232 /*
233 * XXX Initialization before configure().
234 * XXX hack to get pfil_add_hook working in autoconf.
235 */
236 void
237 ifinit1(void)
238 {
239
240 mutex_init(&index_gen_mtx, MUTEX_DEFAULT, IPL_NONE);
241
242 #ifdef PFIL_HOOKS
243 if_pfil.ph_type = PFIL_TYPE_IFNET;
244 if_pfil.ph_ifnet = NULL;
245 if (pfil_head_register(&if_pfil) != 0)
246 printf("WARNING: unable to register pfil hook\n");
247 #endif
248 }
249
250 struct ifnet *
251 if_alloc(u_char type)
252 {
253 return malloc(sizeof(struct ifnet), M_DEVBUF, M_WAITOK|M_ZERO);
254 }
255
256 void
257 if_free(struct ifnet *ifp)
258 {
259 free(ifp, M_DEVBUF);
260 }
261
262 void
263 if_initname(struct ifnet *ifp, const char *name, int unit)
264 {
265 (void)snprintf(ifp->if_xname, sizeof(ifp->if_xname),
266 "%s%d", name, unit);
267 }
268
269 /*
270 * Null routines used while an interface is going away. These routines
271 * just return an error.
272 */
273
274 int
275 if_nulloutput(struct ifnet *ifp, struct mbuf *m,
276 const struct sockaddr *so, struct rtentry *rt)
277 {
278
279 return ENXIO;
280 }
281
282 void
283 if_nullinput(struct ifnet *ifp, struct mbuf *m)
284 {
285
286 /* Nothing. */
287 }
288
289 void
290 if_nullstart(struct ifnet *ifp)
291 {
292
293 /* Nothing. */
294 }
295
296 int
297 if_nullioctl(struct ifnet *ifp, u_long cmd, void *data)
298 {
299
300 /* Wake ifioctl_detach(), who may wait for all threads to
301 * quit the critical section.
302 */
303 cv_signal(&ifp->if_ioctl_lock->il_emptied);
304 return ENXIO;
305 }
306
307 int
308 if_nullinit(struct ifnet *ifp)
309 {
310
311 return ENXIO;
312 }
313
314 void
315 if_nullstop(struct ifnet *ifp, int disable)
316 {
317
318 /* Nothing. */
319 }
320
321 void
322 if_nullwatchdog(struct ifnet *ifp)
323 {
324
325 /* Nothing. */
326 }
327
328 void
329 if_nulldrain(struct ifnet *ifp)
330 {
331
332 /* Nothing. */
333 }
334
335 static u_int if_index = 1;
336 struct ifnet_head ifnet;
337 size_t if_indexlim = 0;
338 struct ifaddr **ifnet_addrs = NULL;
339 struct ifnet **ifindex2ifnet = NULL;
340 struct ifnet *lo0ifp;
341
342 void
343 if_set_sadl(struct ifnet *ifp, const void *lla, u_char addrlen, bool factory)
344 {
345 struct ifaddr *ifa;
346 struct sockaddr_dl *sdl;
347
348 ifp->if_addrlen = addrlen;
349 if_alloc_sadl(ifp);
350 ifa = ifp->if_dl;
351 sdl = satosdl(ifa->ifa_addr);
352
353 (void)sockaddr_dl_setaddr(sdl, sdl->sdl_len, lla, ifp->if_addrlen);
354 if (factory) {
355 ifp->if_hwdl = ifp->if_dl;
356 IFAREF(ifp->if_hwdl);
357 }
358 /* TBD routing socket */
359 }
360
361 struct ifaddr *
362 if_dl_create(const struct ifnet *ifp, const struct sockaddr_dl **sdlp)
363 {
364 unsigned socksize, ifasize;
365 int addrlen, namelen;
366 struct sockaddr_dl *mask, *sdl;
367 struct ifaddr *ifa;
368
369 namelen = strlen(ifp->if_xname);
370 addrlen = ifp->if_addrlen;
371 socksize = roundup(sockaddr_dl_measure(namelen, addrlen), sizeof(long));
372 ifasize = sizeof(*ifa) + 2 * socksize;
373 ifa = (struct ifaddr *)malloc(ifasize, M_IFADDR, M_WAITOK|M_ZERO);
374
375 sdl = (struct sockaddr_dl *)(ifa + 1);
376 mask = (struct sockaddr_dl *)(socksize + (char *)sdl);
377
378 sockaddr_dl_init(sdl, socksize, ifp->if_index, ifp->if_type,
379 ifp->if_xname, namelen, NULL, addrlen);
380 mask->sdl_len = sockaddr_dl_measure(namelen, 0);
381 memset(&mask->sdl_data[0], 0xff, namelen);
382 ifa->ifa_rtrequest = link_rtrequest;
383 ifa->ifa_addr = (struct sockaddr *)sdl;
384 ifa->ifa_netmask = (struct sockaddr *)mask;
385
386 *sdlp = sdl;
387
388 return ifa;
389 }
390
391 static void
392 if_sadl_setrefs(struct ifnet *ifp, struct ifaddr *ifa)
393 {
394 const struct sockaddr_dl *sdl;
395 ifnet_addrs[ifp->if_index] = ifa;
396 IFAREF(ifa);
397 ifp->if_dl = ifa;
398 IFAREF(ifa);
399 sdl = satosdl(ifa->ifa_addr);
400 ifp->if_sadl = sdl;
401 }
402
403 /*
404 * Allocate the link level name for the specified interface. This
405 * is an attachment helper. It must be called after ifp->if_addrlen
406 * is initialized, which may not be the case when if_attach() is
407 * called.
408 */
409 void
410 if_alloc_sadl(struct ifnet *ifp)
411 {
412 struct ifaddr *ifa;
413 const struct sockaddr_dl *sdl;
414
415 /*
416 * If the interface already has a link name, release it
417 * now. This is useful for interfaces that can change
418 * link types, and thus switch link names often.
419 */
420 if (ifp->if_sadl != NULL)
421 if_free_sadl(ifp);
422
423 ifa = if_dl_create(ifp, &sdl);
424
425 ifa_insert(ifp, ifa);
426 if_sadl_setrefs(ifp, ifa);
427 }
428
429 static void
430 if_deactivate_sadl(struct ifnet *ifp)
431 {
432 struct ifaddr *ifa;
433
434 KASSERT(ifp->if_dl != NULL);
435
436 ifa = ifp->if_dl;
437
438 ifp->if_sadl = NULL;
439
440 ifnet_addrs[ifp->if_index] = NULL;
441 IFAFREE(ifa);
442 ifp->if_dl = NULL;
443 IFAFREE(ifa);
444 }
445
446 void
447 if_activate_sadl(struct ifnet *ifp, struct ifaddr *ifa,
448 const struct sockaddr_dl *sdl)
449 {
450 int s;
451
452 s = splnet();
453
454 if_deactivate_sadl(ifp);
455
456 if_sadl_setrefs(ifp, ifa);
457 IFADDR_FOREACH(ifa, ifp)
458 rtinit(ifa, RTM_LLINFO_UPD, 0);
459 splx(s);
460 }
461
462 /*
463 * Free the link level name for the specified interface. This is
464 * a detach helper. This is called from if_detach() or from
465 * link layer type specific detach functions.
466 */
467 void
468 if_free_sadl(struct ifnet *ifp)
469 {
470 struct ifaddr *ifa;
471 int s;
472
473 ifa = ifnet_addrs[ifp->if_index];
474 if (ifa == NULL) {
475 KASSERT(ifp->if_sadl == NULL);
476 KASSERT(ifp->if_dl == NULL);
477 return;
478 }
479
480 KASSERT(ifp->if_sadl != NULL);
481 KASSERT(ifp->if_dl != NULL);
482
483 s = splnet();
484 rtinit(ifa, RTM_DELETE, 0);
485 ifa_remove(ifp, ifa);
486 if_deactivate_sadl(ifp);
487 if (ifp->if_hwdl == ifa) {
488 IFAFREE(ifa);
489 ifp->if_hwdl = NULL;
490 }
491 splx(s);
492 }
493
494 /*
495 * Attach an interface to the
496 * list of "active" interfaces.
497 */
498 void
499 if_attach(struct ifnet *ifp)
500 {
501 int indexlim = 0;
502
503 if (if_indexlim == 0) {
504 TAILQ_INIT(&ifnet);
505 if_indexlim = 8;
506 }
507 TAILQ_INIT(&ifp->if_addrlist);
508 TAILQ_INSERT_TAIL(&ifnet, ifp, if_list);
509
510 if (ifioctl_attach(ifp) != 0)
511 panic("%s: ifioctl_attach() failed", __func__);
512
513 mutex_enter(&index_gen_mtx);
514 ifp->if_index_gen = index_gen++;
515 mutex_exit(&index_gen_mtx);
516
517 ifp->if_index = if_index;
518 if (ifindex2ifnet == NULL)
519 if_index++;
520 else
521 while (ifp->if_index < if_indexlim &&
522 ifindex2ifnet[ifp->if_index] != NULL) {
523 ++if_index;
524 if (if_index == 0)
525 if_index = 1;
526 /*
527 * If we hit USHRT_MAX, we skip back to 0 since
528 * there are a number of places where the value
529 * of if_index or if_index itself is compared
530 * to or stored in an unsigned short. By
531 * jumping back, we won't botch those assignments
532 * or comparisons.
533 */
534 else if (if_index == USHRT_MAX) {
535 /*
536 * However, if we have to jump back to
537 * zero *twice* without finding an empty
538 * slot in ifindex2ifnet[], then there
539 * there are too many (>65535) interfaces.
540 */
541 if (indexlim++)
542 panic("too many interfaces");
543 else
544 if_index = 1;
545 }
546 ifp->if_index = if_index;
547 }
548
549 /*
550 * We have some arrays that should be indexed by if_index.
551 * since if_index will grow dynamically, they should grow too.
552 * struct ifadd **ifnet_addrs
553 * struct ifnet **ifindex2ifnet
554 */
555 if (ifnet_addrs == NULL || ifindex2ifnet == NULL ||
556 ifp->if_index >= if_indexlim) {
557 size_t m, n, oldlim;
558 void *q;
559
560 oldlim = if_indexlim;
561 while (ifp->if_index >= if_indexlim)
562 if_indexlim <<= 1;
563
564 /* grow ifnet_addrs */
565 m = oldlim * sizeof(struct ifaddr *);
566 n = if_indexlim * sizeof(struct ifaddr *);
567 q = malloc(n, M_IFADDR, M_WAITOK|M_ZERO);
568 if (ifnet_addrs != NULL) {
569 memcpy(q, ifnet_addrs, m);
570 free(ifnet_addrs, M_IFADDR);
571 }
572 ifnet_addrs = (struct ifaddr **)q;
573
574 /* grow ifindex2ifnet */
575 m = oldlim * sizeof(struct ifnet *);
576 n = if_indexlim * sizeof(struct ifnet *);
577 q = malloc(n, M_IFADDR, M_WAITOK|M_ZERO);
578 if (ifindex2ifnet != NULL) {
579 memcpy(q, ifindex2ifnet, m);
580 free(ifindex2ifnet, M_IFADDR);
581 }
582 ifindex2ifnet = (struct ifnet **)q;
583 }
584
585 ifindex2ifnet[ifp->if_index] = ifp;
586
587 /*
588 * Link level name is allocated later by a separate call to
589 * if_alloc_sadl().
590 */
591
592 if (ifp->if_snd.ifq_maxlen == 0)
593 ifp->if_snd.ifq_maxlen = ifqmaxlen;
594
595 sysctl_sndq_setup(&ifp->if_sysctl_log, ifp->if_xname, &ifp->if_snd);
596
597 ifp->if_broadcastaddr = 0; /* reliably crash if used uninitialized */
598
599 ifp->if_link_state = LINK_STATE_UNKNOWN;
600
601 ifp->if_capenable = 0;
602 ifp->if_csum_flags_tx = 0;
603 ifp->if_csum_flags_rx = 0;
604
605 #ifdef ALTQ
606 ifp->if_snd.altq_type = 0;
607 ifp->if_snd.altq_disc = NULL;
608 ifp->if_snd.altq_flags &= ALTQF_CANTCHANGE;
609 ifp->if_snd.altq_tbr = NULL;
610 ifp->if_snd.altq_ifp = ifp;
611 #endif
612
613 #ifdef PFIL_HOOKS
614 ifp->if_pfil.ph_type = PFIL_TYPE_IFNET;
615 ifp->if_pfil.ph_ifnet = ifp;
616 if (pfil_head_register(&ifp->if_pfil) != 0)
617 printf("%s: WARNING: unable to register pfil hook\n",
618 ifp->if_xname);
619 (void)pfil_run_hooks(&if_pfil,
620 (struct mbuf **)PFIL_IFNET_ATTACH, ifp, PFIL_IFNET);
621 #endif
622
623 if (!STAILQ_EMPTY(&domains))
624 if_attachdomain1(ifp);
625
626 /* Announce the interface. */
627 rt_ifannouncemsg(ifp, IFAN_ARRIVAL);
628 }
629
630 void
631 if_attachdomain(void)
632 {
633 struct ifnet *ifp;
634 int s;
635
636 s = splnet();
637 IFNET_FOREACH(ifp)
638 if_attachdomain1(ifp);
639 splx(s);
640 }
641
642 void
643 if_attachdomain1(struct ifnet *ifp)
644 {
645 struct domain *dp;
646 int s;
647
648 s = splnet();
649
650 /* address family dependent data region */
651 memset(ifp->if_afdata, 0, sizeof(ifp->if_afdata));
652 DOMAIN_FOREACH(dp) {
653 if (dp->dom_ifattach != NULL)
654 ifp->if_afdata[dp->dom_family] =
655 (*dp->dom_ifattach)(ifp);
656 }
657
658 splx(s);
659 }
660
661 /*
662 * Deactivate an interface. This points all of the procedure
663 * handles at error stubs. May be called from interrupt context.
664 */
665 void
666 if_deactivate(struct ifnet *ifp)
667 {
668 int s;
669
670 s = splnet();
671
672 ifp->if_output = if_nulloutput;
673 ifp->if_input = if_nullinput;
674 ifp->if_start = if_nullstart;
675 ifp->if_ioctl = if_nullioctl;
676 ifp->if_init = if_nullinit;
677 ifp->if_stop = if_nullstop;
678 ifp->if_watchdog = if_nullwatchdog;
679 ifp->if_drain = if_nulldrain;
680
681 /* No more packets may be enqueued. */
682 ifp->if_snd.ifq_maxlen = 0;
683
684 splx(s);
685 }
686
687 void
688 if_purgeaddrs(struct ifnet *ifp, int family, void (*purgeaddr)(struct ifaddr *))
689 {
690 struct ifaddr *ifa, *nifa;
691
692 for (ifa = IFADDR_FIRST(ifp); ifa != NULL; ifa = nifa) {
693 nifa = IFADDR_NEXT(ifa);
694 if (ifa->ifa_addr->sa_family != family)
695 continue;
696 (*purgeaddr)(ifa);
697 }
698 }
699
700 /*
701 * Detach an interface from the list of "active" interfaces,
702 * freeing any resources as we go along.
703 *
704 * NOTE: This routine must be called with a valid thread context,
705 * as it may block.
706 */
707 void
708 if_detach(struct ifnet *ifp)
709 {
710 struct socket so;
711 struct ifaddr *ifa;
712 #ifdef IFAREF_DEBUG
713 struct ifaddr *last_ifa = NULL;
714 #endif
715 struct domain *dp;
716 const struct protosw *pr;
717 int s, i, family, purged;
718
719 /*
720 * XXX It's kind of lame that we have to have the
721 * XXX socket structure...
722 */
723 memset(&so, 0, sizeof(so));
724
725 s = splnet();
726
727 /*
728 * Do an if_down() to give protocols a chance to do something.
729 */
730 if_down(ifp);
731
732 #ifdef ALTQ
733 if (ALTQ_IS_ENABLED(&ifp->if_snd))
734 altq_disable(&ifp->if_snd);
735 if (ALTQ_IS_ATTACHED(&ifp->if_snd))
736 altq_detach(&ifp->if_snd);
737 #endif
738
739 sysctl_teardown(&ifp->if_sysctl_log);
740
741 #if NCARP > 0
742 /* Remove the interface from any carp group it is a part of. */
743 if (ifp->if_carp != NULL && ifp->if_type != IFT_CARP)
744 carp_ifdetach(ifp);
745 #endif
746
747 /*
748 * Rip all the addresses off the interface. This should make
749 * all of the routes go away.
750 *
751 * pr_usrreq calls can remove an arbitrary number of ifaddrs
752 * from the list, including our "cursor", ifa. For safety,
753 * and to honor the TAILQ abstraction, I just restart the
754 * loop after each removal. Note that the loop will exit
755 * when all of the remaining ifaddrs belong to the AF_LINK
756 * family. I am counting on the historical fact that at
757 * least one pr_usrreq in each address domain removes at
758 * least one ifaddr.
759 */
760 again:
761 IFADDR_FOREACH(ifa, ifp) {
762 family = ifa->ifa_addr->sa_family;
763 #ifdef IFAREF_DEBUG
764 printf("if_detach: ifaddr %p, family %d, refcnt %d\n",
765 ifa, family, ifa->ifa_refcnt);
766 if (last_ifa != NULL && ifa == last_ifa)
767 panic("if_detach: loop detected");
768 last_ifa = ifa;
769 #endif
770 if (family == AF_LINK)
771 continue;
772 dp = pffinddomain(family);
773 #ifdef DIAGNOSTIC
774 if (dp == NULL)
775 panic("if_detach: no domain for AF %d",
776 family);
777 #endif
778 /*
779 * XXX These PURGEIF calls are redundant with the
780 * purge-all-families calls below, but are left in for
781 * now both to make a smaller change, and to avoid
782 * unplanned interactions with clearing of
783 * ifp->if_addrlist.
784 */
785 purged = 0;
786 for (pr = dp->dom_protosw;
787 pr < dp->dom_protoswNPROTOSW; pr++) {
788 so.so_proto = pr;
789 if (pr->pr_usrreq != NULL) {
790 (void) (*pr->pr_usrreq)(&so,
791 PRU_PURGEIF, NULL, NULL,
792 (struct mbuf *) ifp, curlwp);
793 purged = 1;
794 }
795 }
796 if (purged == 0) {
797 /*
798 * XXX What's really the best thing to do
799 * XXX here? --thorpej (at) NetBSD.org
800 */
801 printf("if_detach: WARNING: AF %d not purged\n",
802 family);
803 ifa_remove(ifp, ifa);
804 }
805 goto again;
806 }
807
808 if_free_sadl(ifp);
809
810 /* Walk the routing table looking for stragglers. */
811 for (i = 0; i <= AF_MAX; i++) {
812 while (rt_walktree(i, if_rt_walktree, ifp) == ERESTART)
813 ;
814 }
815
816 DOMAIN_FOREACH(dp) {
817 if (dp->dom_ifdetach != NULL && ifp->if_afdata[dp->dom_family])
818 (*dp->dom_ifdetach)(ifp,
819 ifp->if_afdata[dp->dom_family]);
820
821 /*
822 * One would expect multicast memberships (INET and
823 * INET6) on UDP sockets to be purged by the PURGEIF
824 * calls above, but if all addresses were removed from
825 * the interface prior to destruction, the calls will
826 * not be made (e.g. ppp, for which pppd(8) generally
827 * removes addresses before destroying the interface).
828 * Because there is no invariant that multicast
829 * memberships only exist for interfaces with IPv4
830 * addresses, we must call PURGEIF regardless of
831 * addresses. (Protocols which might store ifnet
832 * pointers are marked with PR_PURGEIF.)
833 */
834 for (pr = dp->dom_protosw; pr < dp->dom_protoswNPROTOSW; pr++) {
835 so.so_proto = pr;
836 if (pr->pr_usrreq != NULL && pr->pr_flags & PR_PURGEIF)
837 (void)(*pr->pr_usrreq)(&so, PRU_PURGEIF, NULL,
838 NULL, (struct mbuf *)ifp, curlwp);
839 }
840 }
841
842 #ifdef PFIL_HOOKS
843 (void)pfil_run_hooks(&if_pfil,
844 (struct mbuf **)PFIL_IFNET_DETACH, ifp, PFIL_IFNET);
845 (void)pfil_head_unregister(&ifp->if_pfil);
846 #endif
847
848 /* Announce that the interface is gone. */
849 rt_ifannouncemsg(ifp, IFAN_DEPARTURE);
850
851 ifindex2ifnet[ifp->if_index] = NULL;
852
853 TAILQ_REMOVE(&ifnet, ifp, if_list);
854
855 ifioctl_detach(ifp);
856
857 /*
858 * remove packets that came from ifp, from software interrupt queues.
859 */
860 DOMAIN_FOREACH(dp) {
861 for (i = 0; i < __arraycount(dp->dom_ifqueues); i++) {
862 if (dp->dom_ifqueues[i] == NULL)
863 break;
864 if_detach_queues(ifp, dp->dom_ifqueues[i]);
865 }
866 }
867
868 splx(s);
869 }
870
871 static void
872 if_detach_queues(struct ifnet *ifp, struct ifqueue *q)
873 {
874 struct mbuf *m, *prev, *next;
875
876 prev = NULL;
877 for (m = q->ifq_head; m != NULL; m = next) {
878 next = m->m_nextpkt;
879 #ifdef DIAGNOSTIC
880 if ((m->m_flags & M_PKTHDR) == 0) {
881 prev = m;
882 continue;
883 }
884 #endif
885 if (m->m_pkthdr.rcvif != ifp) {
886 prev = m;
887 continue;
888 }
889
890 if (prev != NULL)
891 prev->m_nextpkt = m->m_nextpkt;
892 else
893 q->ifq_head = m->m_nextpkt;
894 if (q->ifq_tail == m)
895 q->ifq_tail = prev;
896 q->ifq_len--;
897
898 m->m_nextpkt = NULL;
899 m_freem(m);
900 IF_DROP(q);
901 }
902 }
903
904 /*
905 * Callback for a radix tree walk to delete all references to an
906 * ifnet.
907 */
908 static int
909 if_rt_walktree(struct rtentry *rt, void *v)
910 {
911 struct ifnet *ifp = (struct ifnet *)v;
912 int error;
913
914 if (rt->rt_ifp != ifp)
915 return 0;
916
917 /* Delete the entry. */
918 ++rt->rt_refcnt;
919 error = rtrequest(RTM_DELETE, rt_getkey(rt), rt->rt_gateway,
920 rt_mask(rt), rt->rt_flags, NULL);
921 KASSERT((rt->rt_flags & RTF_UP) == 0);
922 rt->rt_ifp = NULL;
923 RTFREE(rt);
924 if (error != 0)
925 printf("%s: warning: unable to delete rtentry @ %p, "
926 "error = %d\n", ifp->if_xname, rt, error);
927 return ERESTART;
928 }
929
930 /*
931 * Create a clone network interface.
932 */
933 int
934 if_clone_create(const char *name)
935 {
936 struct if_clone *ifc;
937 int unit;
938
939 ifc = if_clone_lookup(name, &unit);
940 if (ifc == NULL)
941 return EINVAL;
942
943 if (ifunit(name) != NULL)
944 return EEXIST;
945
946 return (*ifc->ifc_create)(ifc, unit);
947 }
948
949 /*
950 * Destroy a clone network interface.
951 */
952 int
953 if_clone_destroy(const char *name)
954 {
955 struct if_clone *ifc;
956 struct ifnet *ifp;
957
958 ifc = if_clone_lookup(name, NULL);
959 if (ifc == NULL)
960 return EINVAL;
961
962 ifp = ifunit(name);
963 if (ifp == NULL)
964 return ENXIO;
965
966 if (ifc->ifc_destroy == NULL)
967 return EOPNOTSUPP;
968
969 return (*ifc->ifc_destroy)(ifp);
970 }
971
972 /*
973 * Look up a network interface cloner.
974 */
975 static struct if_clone *
976 if_clone_lookup(const char *name, int *unitp)
977 {
978 struct if_clone *ifc;
979 const char *cp;
980 int unit;
981
982 /* separate interface name from unit */
983 for (cp = name;
984 cp - name < IFNAMSIZ && *cp && (*cp < '0' || *cp > '9');
985 cp++)
986 continue;
987
988 if (cp == name || cp - name == IFNAMSIZ || !*cp)
989 return NULL; /* No name or unit number */
990
991 LIST_FOREACH(ifc, &if_cloners, ifc_list) {
992 if (strlen(ifc->ifc_name) == cp - name &&
993 strncmp(name, ifc->ifc_name, cp - name) == 0)
994 break;
995 }
996
997 if (ifc == NULL)
998 return NULL;
999
1000 unit = 0;
1001 while (cp - name < IFNAMSIZ && *cp) {
1002 if (*cp < '0' || *cp > '9' || unit >= INT_MAX / 10) {
1003 /* Bogus unit number. */
1004 return NULL;
1005 }
1006 unit = (unit * 10) + (*cp++ - '0');
1007 }
1008
1009 if (unitp != NULL)
1010 *unitp = unit;
1011 return ifc;
1012 }
1013
1014 /*
1015 * Register a network interface cloner.
1016 */
1017 void
1018 if_clone_attach(struct if_clone *ifc)
1019 {
1020
1021 LIST_INSERT_HEAD(&if_cloners, ifc, ifc_list);
1022 if_cloners_count++;
1023 }
1024
1025 /*
1026 * Unregister a network interface cloner.
1027 */
1028 void
1029 if_clone_detach(struct if_clone *ifc)
1030 {
1031
1032 LIST_REMOVE(ifc, ifc_list);
1033 if_cloners_count--;
1034 }
1035
1036 /*
1037 * Provide list of interface cloners to userspace.
1038 */
1039 static int
1040 if_clone_list(struct if_clonereq *ifcr)
1041 {
1042 char outbuf[IFNAMSIZ], *dst;
1043 struct if_clone *ifc;
1044 int count, error = 0;
1045
1046 ifcr->ifcr_total = if_cloners_count;
1047 if ((dst = ifcr->ifcr_buffer) == NULL) {
1048 /* Just asking how many there are. */
1049 return 0;
1050 }
1051
1052 if (ifcr->ifcr_count < 0)
1053 return EINVAL;
1054
1055 count = (if_cloners_count < ifcr->ifcr_count) ?
1056 if_cloners_count : ifcr->ifcr_count;
1057
1058 for (ifc = LIST_FIRST(&if_cloners); ifc != NULL && count != 0;
1059 ifc = LIST_NEXT(ifc, ifc_list), count--, dst += IFNAMSIZ) {
1060 (void)strncpy(outbuf, ifc->ifc_name, sizeof(outbuf));
1061 if (outbuf[sizeof(outbuf) - 1] != '\0')
1062 return ENAMETOOLONG;
1063 error = copyout(outbuf, dst, sizeof(outbuf));
1064 if (error != 0)
1065 break;
1066 }
1067
1068 return error;
1069 }
1070
1071 void
1072 ifa_insert(struct ifnet *ifp, struct ifaddr *ifa)
1073 {
1074 ifa->ifa_ifp = ifp;
1075 TAILQ_INSERT_TAIL(&ifp->if_addrlist, ifa, ifa_list);
1076 IFAREF(ifa);
1077 }
1078
1079 void
1080 ifa_remove(struct ifnet *ifp, struct ifaddr *ifa)
1081 {
1082 KASSERT(ifa->ifa_ifp == ifp);
1083 TAILQ_REMOVE(&ifp->if_addrlist, ifa, ifa_list);
1084 IFAFREE(ifa);
1085 }
1086
1087 static inline int
1088 equal(const struct sockaddr *sa1, const struct sockaddr *sa2)
1089 {
1090 return sockaddr_cmp(sa1, sa2) == 0;
1091 }
1092
1093 /*
1094 * Locate an interface based on a complete address.
1095 */
1096 /*ARGSUSED*/
1097 struct ifaddr *
1098 ifa_ifwithaddr(const struct sockaddr *addr)
1099 {
1100 struct ifnet *ifp;
1101 struct ifaddr *ifa;
1102
1103 IFNET_FOREACH(ifp) {
1104 if (ifp->if_output == if_nulloutput)
1105 continue;
1106 IFADDR_FOREACH(ifa, ifp) {
1107 if (ifa->ifa_addr->sa_family != addr->sa_family)
1108 continue;
1109 if (equal(addr, ifa->ifa_addr))
1110 return ifa;
1111 if ((ifp->if_flags & IFF_BROADCAST) &&
1112 ifa->ifa_broadaddr &&
1113 /* IP6 doesn't have broadcast */
1114 ifa->ifa_broadaddr->sa_len != 0 &&
1115 equal(ifa->ifa_broadaddr, addr))
1116 return ifa;
1117 }
1118 }
1119 return NULL;
1120 }
1121
1122 /*
1123 * Locate the point to point interface with a given destination address.
1124 */
1125 /*ARGSUSED*/
1126 struct ifaddr *
1127 ifa_ifwithdstaddr(const struct sockaddr *addr)
1128 {
1129 struct ifnet *ifp;
1130 struct ifaddr *ifa;
1131
1132 IFNET_FOREACH(ifp) {
1133 if (ifp->if_output == if_nulloutput)
1134 continue;
1135 if ((ifp->if_flags & IFF_POINTOPOINT) == 0)
1136 continue;
1137 IFADDR_FOREACH(ifa, ifp) {
1138 if (ifa->ifa_addr->sa_family != addr->sa_family ||
1139 ifa->ifa_dstaddr == NULL)
1140 continue;
1141 if (equal(addr, ifa->ifa_dstaddr))
1142 return ifa;
1143 }
1144 }
1145 return NULL;
1146 }
1147
1148 /*
1149 * Find an interface on a specific network. If many, choice
1150 * is most specific found.
1151 */
1152 struct ifaddr *
1153 ifa_ifwithnet(const struct sockaddr *addr)
1154 {
1155 struct ifnet *ifp;
1156 struct ifaddr *ifa;
1157 const struct sockaddr_dl *sdl;
1158 struct ifaddr *ifa_maybe = 0;
1159 u_int af = addr->sa_family;
1160 const char *addr_data = addr->sa_data, *cplim;
1161
1162 if (af == AF_LINK) {
1163 sdl = satocsdl(addr);
1164 if (sdl->sdl_index && sdl->sdl_index < if_indexlim &&
1165 ifindex2ifnet[sdl->sdl_index] &&
1166 ifindex2ifnet[sdl->sdl_index]->if_output != if_nulloutput)
1167 return ifnet_addrs[sdl->sdl_index];
1168 }
1169 #ifdef NETATALK
1170 if (af == AF_APPLETALK) {
1171 const struct sockaddr_at *sat, *sat2;
1172 sat = (const struct sockaddr_at *)addr;
1173 IFNET_FOREACH(ifp) {
1174 if (ifp->if_output == if_nulloutput)
1175 continue;
1176 ifa = at_ifawithnet((const struct sockaddr_at *)addr, ifp);
1177 if (ifa == NULL)
1178 continue;
1179 sat2 = (struct sockaddr_at *)ifa->ifa_addr;
1180 if (sat2->sat_addr.s_net == sat->sat_addr.s_net)
1181 return ifa; /* exact match */
1182 if (ifa_maybe == NULL) {
1183 /* else keep the if with the right range */
1184 ifa_maybe = ifa;
1185 }
1186 }
1187 return ifa_maybe;
1188 }
1189 #endif
1190 IFNET_FOREACH(ifp) {
1191 if (ifp->if_output == if_nulloutput)
1192 continue;
1193 IFADDR_FOREACH(ifa, ifp) {
1194 const char *cp, *cp2, *cp3;
1195
1196 if (ifa->ifa_addr->sa_family != af ||
1197 ifa->ifa_netmask == NULL)
1198 next: continue;
1199 cp = addr_data;
1200 cp2 = ifa->ifa_addr->sa_data;
1201 cp3 = ifa->ifa_netmask->sa_data;
1202 cplim = (const char *)ifa->ifa_netmask +
1203 ifa->ifa_netmask->sa_len;
1204 while (cp3 < cplim) {
1205 if ((*cp++ ^ *cp2++) & *cp3++) {
1206 /* want to continue for() loop */
1207 goto next;
1208 }
1209 }
1210 if (ifa_maybe == NULL ||
1211 rn_refines((void *)ifa->ifa_netmask,
1212 (void *)ifa_maybe->ifa_netmask))
1213 ifa_maybe = ifa;
1214 }
1215 }
1216 return ifa_maybe;
1217 }
1218
1219 /*
1220 * Find the interface of the addresss.
1221 */
1222 struct ifaddr *
1223 ifa_ifwithladdr(const struct sockaddr *addr)
1224 {
1225 struct ifaddr *ia;
1226
1227 if ((ia = ifa_ifwithaddr(addr)) || (ia = ifa_ifwithdstaddr(addr)) ||
1228 (ia = ifa_ifwithnet(addr)))
1229 return ia;
1230 return NULL;
1231 }
1232
1233 /*
1234 * Find an interface using a specific address family
1235 */
1236 struct ifaddr *
1237 ifa_ifwithaf(int af)
1238 {
1239 struct ifnet *ifp;
1240 struct ifaddr *ifa;
1241
1242 IFNET_FOREACH(ifp) {
1243 if (ifp->if_output == if_nulloutput)
1244 continue;
1245 IFADDR_FOREACH(ifa, ifp) {
1246 if (ifa->ifa_addr->sa_family == af)
1247 return ifa;
1248 }
1249 }
1250 return NULL;
1251 }
1252
1253 /*
1254 * Find an interface address specific to an interface best matching
1255 * a given address.
1256 */
1257 struct ifaddr *
1258 ifaof_ifpforaddr(const struct sockaddr *addr, struct ifnet *ifp)
1259 {
1260 struct ifaddr *ifa;
1261 const char *cp, *cp2, *cp3;
1262 const char *cplim;
1263 struct ifaddr *ifa_maybe = 0;
1264 u_int af = addr->sa_family;
1265
1266 if (ifp->if_output == if_nulloutput)
1267 return NULL;
1268
1269 if (af >= AF_MAX)
1270 return NULL;
1271
1272 IFADDR_FOREACH(ifa, ifp) {
1273 if (ifa->ifa_addr->sa_family != af)
1274 continue;
1275 ifa_maybe = ifa;
1276 if (ifa->ifa_netmask == NULL) {
1277 if (equal(addr, ifa->ifa_addr) ||
1278 (ifa->ifa_dstaddr &&
1279 equal(addr, ifa->ifa_dstaddr)))
1280 return ifa;
1281 continue;
1282 }
1283 cp = addr->sa_data;
1284 cp2 = ifa->ifa_addr->sa_data;
1285 cp3 = ifa->ifa_netmask->sa_data;
1286 cplim = ifa->ifa_netmask->sa_len + (char *)ifa->ifa_netmask;
1287 for (; cp3 < cplim; cp3++) {
1288 if ((*cp++ ^ *cp2++) & *cp3)
1289 break;
1290 }
1291 if (cp3 == cplim)
1292 return ifa;
1293 }
1294 return ifa_maybe;
1295 }
1296
1297 /*
1298 * Default action when installing a route with a Link Level gateway.
1299 * Lookup an appropriate real ifa to point to.
1300 * This should be moved to /sys/net/link.c eventually.
1301 */
1302 void
1303 link_rtrequest(int cmd, struct rtentry *rt, const struct rt_addrinfo *info)
1304 {
1305 struct ifaddr *ifa;
1306 const struct sockaddr *dst;
1307 struct ifnet *ifp;
1308
1309 if (cmd != RTM_ADD || (ifa = rt->rt_ifa) == NULL ||
1310 (ifp = ifa->ifa_ifp) == NULL || (dst = rt_getkey(rt)) == NULL)
1311 return;
1312 if ((ifa = ifaof_ifpforaddr(dst, ifp)) != NULL) {
1313 rt_replace_ifa(rt, ifa);
1314 if (ifa->ifa_rtrequest && ifa->ifa_rtrequest != link_rtrequest)
1315 ifa->ifa_rtrequest(cmd, rt, info);
1316 }
1317 }
1318
1319 /*
1320 * Handle a change in the interface link state.
1321 */
1322 void
1323 if_link_state_change(struct ifnet *ifp, int link_state)
1324 {
1325 if (ifp->if_link_state == link_state)
1326 return;
1327 ifp->if_link_state = link_state;
1328 /* Notify that the link state has changed. */
1329 rt_ifmsg(ifp);
1330 #if NCARP > 0
1331 if (ifp->if_carp)
1332 carp_carpdev_state(ifp);
1333 #endif
1334 }
1335
1336 /*
1337 * Mark an interface down and notify protocols of
1338 * the transition.
1339 * NOTE: must be called at splsoftnet or equivalent.
1340 */
1341 void
1342 if_down(struct ifnet *ifp)
1343 {
1344 struct ifaddr *ifa;
1345
1346 ifp->if_flags &= ~IFF_UP;
1347 nanotime(&ifp->if_lastchange);
1348 IFADDR_FOREACH(ifa, ifp)
1349 pfctlinput(PRC_IFDOWN, ifa->ifa_addr);
1350 IFQ_PURGE(&ifp->if_snd);
1351 #if NCARP > 0
1352 if (ifp->if_carp)
1353 carp_carpdev_state(ifp);
1354 #endif
1355 rt_ifmsg(ifp);
1356 }
1357
1358 /*
1359 * Mark an interface up and notify protocols of
1360 * the transition.
1361 * NOTE: must be called at splsoftnet or equivalent.
1362 */
1363 void
1364 if_up(struct ifnet *ifp)
1365 {
1366 #ifdef notyet
1367 struct ifaddr *ifa;
1368 #endif
1369
1370 ifp->if_flags |= IFF_UP;
1371 nanotime(&ifp->if_lastchange);
1372 #ifdef notyet
1373 /* this has no effect on IP, and will kill all ISO connections XXX */
1374 IFADDR_FOREACH(ifa, ifp)
1375 pfctlinput(PRC_IFUP, ifa->ifa_addr);
1376 #endif
1377 #if NCARP > 0
1378 if (ifp->if_carp)
1379 carp_carpdev_state(ifp);
1380 #endif
1381 rt_ifmsg(ifp);
1382 #ifdef INET6
1383 in6_if_up(ifp);
1384 #endif
1385 }
1386
1387 /*
1388 * Handle interface watchdog timer routines. Called
1389 * from softclock, we decrement timers (if set) and
1390 * call the appropriate interface routine on expiration.
1391 */
1392 void
1393 if_slowtimo(void *arg)
1394 {
1395 struct ifnet *ifp;
1396 int s = splnet();
1397
1398 IFNET_FOREACH(ifp) {
1399 if (ifp->if_timer == 0 || --ifp->if_timer)
1400 continue;
1401 if (ifp->if_watchdog != NULL)
1402 (*ifp->if_watchdog)(ifp);
1403 }
1404 splx(s);
1405 callout_reset(&if_slowtimo_ch, hz / IFNET_SLOWHZ, if_slowtimo, NULL);
1406 }
1407
1408 /*
1409 * Set/clear promiscuous mode on interface ifp based on the truth value
1410 * of pswitch. The calls are reference counted so that only the first
1411 * "on" request actually has an effect, as does the final "off" request.
1412 * Results are undefined if the "off" and "on" requests are not matched.
1413 */
1414 int
1415 ifpromisc(struct ifnet *ifp, int pswitch)
1416 {
1417 int pcount, ret;
1418 short flags, nflags;
1419
1420 pcount = ifp->if_pcount;
1421 flags = ifp->if_flags;
1422 if (pswitch) {
1423 /*
1424 * Allow the device to be "placed" into promiscuous
1425 * mode even if it is not configured up. It will
1426 * consult IFF_PROMISC when it is brought up.
1427 */
1428 if (ifp->if_pcount++ != 0)
1429 return 0;
1430 nflags = ifp->if_flags | IFF_PROMISC;
1431 if ((nflags & IFF_UP) == 0)
1432 return 0;
1433 } else {
1434 if (--ifp->if_pcount > 0)
1435 return 0;
1436 nflags = ifp->if_flags & ~IFF_PROMISC;
1437 /*
1438 * If the device is not configured up, we should not need to
1439 * turn off promiscuous mode (device should have turned it
1440 * off when interface went down; and will look at IFF_PROMISC
1441 * again next time interface comes up).
1442 */
1443 if ((nflags & IFF_UP) == 0)
1444 return 0;
1445 }
1446 ret = if_flags_set(ifp, nflags);
1447 /* Restore interface state if not successful. */
1448 if (ret != 0) {
1449 ifp->if_pcount = pcount;
1450 }
1451 return ret;
1452 }
1453
1454 /*
1455 * Map interface name to
1456 * interface structure pointer.
1457 */
1458 struct ifnet *
1459 ifunit(const char *name)
1460 {
1461 struct ifnet *ifp;
1462 const char *cp = name;
1463 u_int unit = 0;
1464 u_int i;
1465
1466 /*
1467 * If the entire name is a number, treat it as an ifindex.
1468 */
1469 for (i = 0; i < IFNAMSIZ && *cp >= '0' && *cp <= '9'; i++, cp++) {
1470 unit = unit * 10 + (*cp - '0');
1471 }
1472
1473 /*
1474 * If the number took all of the name, then it's a valid ifindex.
1475 */
1476 if (i == IFNAMSIZ || (cp != name && *cp == '\0')) {
1477 if (unit >= if_indexlim)
1478 return NULL;
1479 ifp = ifindex2ifnet[unit];
1480 if (ifp == NULL || ifp->if_output == if_nulloutput)
1481 return NULL;
1482 return ifp;
1483 }
1484
1485 IFNET_FOREACH(ifp) {
1486 if (ifp->if_output == if_nulloutput)
1487 continue;
1488 if (strcmp(ifp->if_xname, name) == 0)
1489 return ifp;
1490 }
1491 return NULL;
1492 }
1493
1494 ifnet_t *
1495 if_byindex(u_int idx)
1496 {
1497
1498 return (idx < if_indexlim) ? ifindex2ifnet[idx] : NULL;
1499 }
1500
1501 /* common */
1502 int
1503 ifioctl_common(struct ifnet *ifp, u_long cmd, void *data)
1504 {
1505 int s;
1506 struct ifreq *ifr;
1507 struct ifcapreq *ifcr;
1508 struct ifdatareq *ifdr;
1509
1510 switch (cmd) {
1511 case SIOCSIFCAP:
1512 ifcr = data;
1513 if ((ifcr->ifcr_capenable & ~ifp->if_capabilities) != 0)
1514 return EINVAL;
1515
1516 if (ifcr->ifcr_capenable == ifp->if_capenable)
1517 return 0;
1518
1519 ifp->if_capenable = ifcr->ifcr_capenable;
1520
1521 /* Pre-compute the checksum flags mask. */
1522 ifp->if_csum_flags_tx = 0;
1523 ifp->if_csum_flags_rx = 0;
1524 if (ifp->if_capenable & IFCAP_CSUM_IPv4_Tx) {
1525 ifp->if_csum_flags_tx |= M_CSUM_IPv4;
1526 }
1527 if (ifp->if_capenable & IFCAP_CSUM_IPv4_Rx) {
1528 ifp->if_csum_flags_rx |= M_CSUM_IPv4;
1529 }
1530
1531 if (ifp->if_capenable & IFCAP_CSUM_TCPv4_Tx) {
1532 ifp->if_csum_flags_tx |= M_CSUM_TCPv4;
1533 }
1534 if (ifp->if_capenable & IFCAP_CSUM_TCPv4_Rx) {
1535 ifp->if_csum_flags_rx |= M_CSUM_TCPv4;
1536 }
1537
1538 if (ifp->if_capenable & IFCAP_CSUM_UDPv4_Tx) {
1539 ifp->if_csum_flags_tx |= M_CSUM_UDPv4;
1540 }
1541 if (ifp->if_capenable & IFCAP_CSUM_UDPv4_Rx) {
1542 ifp->if_csum_flags_rx |= M_CSUM_UDPv4;
1543 }
1544
1545 if (ifp->if_capenable & IFCAP_CSUM_TCPv6_Tx) {
1546 ifp->if_csum_flags_tx |= M_CSUM_TCPv6;
1547 }
1548 if (ifp->if_capenable & IFCAP_CSUM_TCPv6_Rx) {
1549 ifp->if_csum_flags_rx |= M_CSUM_TCPv6;
1550 }
1551
1552 if (ifp->if_capenable & IFCAP_CSUM_UDPv6_Tx) {
1553 ifp->if_csum_flags_tx |= M_CSUM_UDPv6;
1554 }
1555 if (ifp->if_capenable & IFCAP_CSUM_UDPv6_Rx) {
1556 ifp->if_csum_flags_rx |= M_CSUM_UDPv6;
1557 }
1558 if (ifp->if_flags & IFF_UP)
1559 return ENETRESET;
1560 return 0;
1561 case SIOCSIFFLAGS:
1562 ifr = data;
1563 if (ifp->if_flags & IFF_UP && (ifr->ifr_flags & IFF_UP) == 0) {
1564 s = splnet();
1565 if_down(ifp);
1566 splx(s);
1567 }
1568 if (ifr->ifr_flags & IFF_UP && (ifp->if_flags & IFF_UP) == 0) {
1569 s = splnet();
1570 if_up(ifp);
1571 splx(s);
1572 }
1573 ifp->if_flags = (ifp->if_flags & IFF_CANTCHANGE) |
1574 (ifr->ifr_flags &~ IFF_CANTCHANGE);
1575 break;
1576 case SIOCGIFFLAGS:
1577 ifr = data;
1578 ifr->ifr_flags = ifp->if_flags;
1579 break;
1580
1581 case SIOCGIFMETRIC:
1582 ifr = data;
1583 ifr->ifr_metric = ifp->if_metric;
1584 break;
1585
1586 case SIOCGIFMTU:
1587 ifr = data;
1588 ifr->ifr_mtu = ifp->if_mtu;
1589 break;
1590
1591 case SIOCGIFDLT:
1592 ifr = data;
1593 ifr->ifr_dlt = ifp->if_dlt;
1594 break;
1595
1596 case SIOCGIFCAP:
1597 ifcr = data;
1598 ifcr->ifcr_capabilities = ifp->if_capabilities;
1599 ifcr->ifcr_capenable = ifp->if_capenable;
1600 break;
1601
1602 case SIOCSIFMETRIC:
1603 ifr = data;
1604 ifp->if_metric = ifr->ifr_metric;
1605 break;
1606
1607 case SIOCGIFDATA:
1608 ifdr = data;
1609 ifdr->ifdr_data = ifp->if_data;
1610 break;
1611
1612 case SIOCZIFDATA:
1613 ifdr = data;
1614 ifdr->ifdr_data = ifp->if_data;
1615 /*
1616 * Assumes that the volatile counters that can be
1617 * zero'ed are at the end of if_data.
1618 */
1619 memset(&ifp->if_data.ifi_ipackets, 0, sizeof(ifp->if_data) -
1620 offsetof(struct if_data, ifi_ipackets));
1621 break;
1622 case SIOCSIFMTU:
1623 ifr = data;
1624 if (ifp->if_mtu == ifr->ifr_mtu)
1625 break;
1626 ifp->if_mtu = ifr->ifr_mtu;
1627 /*
1628 * If the link MTU changed, do network layer specific procedure.
1629 */
1630 #ifdef INET6
1631 nd6_setmtu(ifp);
1632 #endif
1633 return ENETRESET;
1634 default:
1635 return ENOTTY;
1636 }
1637 return 0;
1638 }
1639
1640 int
1641 ifaddrpref_ioctl(struct socket *so, u_long cmd, void *data, struct ifnet *ifp,
1642 lwp_t *l)
1643 {
1644 struct if_addrprefreq *ifap = (struct if_addrprefreq *)data;
1645 struct ifaddr *ifa;
1646 const struct sockaddr *any, *sa;
1647 union {
1648 struct sockaddr sa;
1649 struct sockaddr_storage ss;
1650 } u, v;
1651
1652 switch (cmd) {
1653 case SIOCSIFADDRPREF:
1654 if (kauth_authorize_network(l->l_cred, KAUTH_NETWORK_INTERFACE,
1655 KAUTH_REQ_NETWORK_INTERFACE_SETPRIV, ifp, (void *)cmd,
1656 NULL) != 0)
1657 return EPERM;
1658 case SIOCGIFADDRPREF:
1659 break;
1660 default:
1661 return EOPNOTSUPP;
1662 }
1663
1664 /* sanity checks */
1665 if (data == NULL || ifp == NULL) {
1666 panic("invalid argument to %s", __func__);
1667 /*NOTREACHED*/
1668 }
1669
1670 /* address must be specified on ADD and DELETE */
1671 sa = sstocsa(&ifap->ifap_addr);
1672 if (sa->sa_family != sofamily(so))
1673 return EINVAL;
1674 if ((any = sockaddr_any(sa)) == NULL || sa->sa_len != any->sa_len)
1675 return EINVAL;
1676
1677 sockaddr_externalize(&v.sa, sizeof(v.ss), sa);
1678
1679 IFADDR_FOREACH(ifa, ifp) {
1680 if (ifa->ifa_addr->sa_family != sa->sa_family)
1681 continue;
1682 sockaddr_externalize(&u.sa, sizeof(u.ss), ifa->ifa_addr);
1683 if (sockaddr_cmp(&u.sa, &v.sa) == 0)
1684 break;
1685 }
1686 if (ifa == NULL)
1687 return EADDRNOTAVAIL;
1688
1689 switch (cmd) {
1690 case SIOCSIFADDRPREF:
1691 ifa->ifa_preference = ifap->ifap_preference;
1692 return 0;
1693 case SIOCGIFADDRPREF:
1694 /* fill in the if_laddrreq structure */
1695 (void)sockaddr_copy(sstosa(&ifap->ifap_addr),
1696 sizeof(ifap->ifap_addr), ifa->ifa_addr);
1697 ifap->ifap_preference = ifa->ifa_preference;
1698 return 0;
1699 default:
1700 return EOPNOTSUPP;
1701 }
1702 }
1703
1704 static void
1705 ifnet_lock_enter(struct ifnet_lock *il)
1706 {
1707 uint64_t *nenter;
1708
1709 /* Before trying to acquire the mutex, increase the count of threads
1710 * who have entered or who wait to enter the critical section.
1711 * Avoid one costly locked memory transaction by keeping a count for
1712 * each CPU.
1713 */
1714 nenter = percpu_getref(il->il_nenter);
1715 (*nenter)++;
1716 percpu_putref(il->il_nenter);
1717 mutex_enter(&il->il_lock);
1718 }
1719
1720 static void
1721 ifnet_lock_exit(struct ifnet_lock *il)
1722 {
1723 /* Increase the count of threads who have exited the critical
1724 * section. Increase while we still hold the lock.
1725 */
1726 il->il_nexit++;
1727 mutex_exit(&il->il_lock);
1728 }
1729
1730 /*
1731 * Interface ioctls.
1732 */
1733 int
1734 ifioctl(struct socket *so, u_long cmd, void *data, struct lwp *l)
1735 {
1736 struct ifnet *ifp;
1737 struct ifreq *ifr;
1738 int error = 0;
1739 #if defined(COMPAT_OSOCK) || defined(COMPAT_OIFREQ)
1740 u_long ocmd = cmd;
1741 #endif
1742 short oif_flags;
1743 #ifdef COMPAT_OIFREQ
1744 struct ifreq ifrb;
1745 struct oifreq *oifr = NULL;
1746 #endif
1747
1748 switch (cmd) {
1749 #ifdef COMPAT_OIFREQ
1750 case OSIOCGIFCONF:
1751 case OOSIOCGIFCONF:
1752 return compat_ifconf(cmd, data);
1753 #endif
1754 #ifdef COMPAT_OIFDATA
1755 case OSIOCGIFDATA:
1756 case OSIOCZIFDATA:
1757 return compat_ifdatareq(l, cmd, data);
1758 #endif
1759 case SIOCGIFCONF:
1760 return ifconf(cmd, data);
1761 case SIOCINITIFADDR:
1762 return EPERM;
1763 }
1764
1765 #ifdef COMPAT_OIFREQ
1766 cmd = compat_cvtcmd(cmd);
1767 if (cmd != ocmd) {
1768 oifr = data;
1769 data = ifr = &ifrb;
1770 ifreqo2n(oifr, ifr);
1771 } else
1772 #endif
1773 ifr = data;
1774
1775 ifp = ifunit(ifr->ifr_name);
1776
1777 switch (cmd) {
1778 case SIOCIFCREATE:
1779 case SIOCIFDESTROY:
1780 if (l != NULL) {
1781 error = kauth_authorize_network(l->l_cred,
1782 KAUTH_NETWORK_INTERFACE,
1783 KAUTH_REQ_NETWORK_INTERFACE_SETPRIV, ifp,
1784 (void *)cmd, NULL);
1785 if (error != 0)
1786 return error;
1787 }
1788 return (cmd == SIOCIFCREATE) ?
1789 if_clone_create(ifr->ifr_name) :
1790 if_clone_destroy(ifr->ifr_name);
1791
1792 case SIOCIFGCLONERS:
1793 return if_clone_list((struct if_clonereq *)data);
1794 }
1795
1796 if (ifp == NULL)
1797 return ENXIO;
1798
1799 switch (cmd) {
1800 case SIOCALIFADDR:
1801 case SIOCDLIFADDR:
1802 case SIOCSIFADDRPREF:
1803 case SIOCSIFFLAGS:
1804 case SIOCSIFCAP:
1805 case SIOCSIFMETRIC:
1806 case SIOCZIFDATA:
1807 case SIOCSIFMTU:
1808 case SIOCSIFPHYADDR:
1809 case SIOCDIFPHYADDR:
1810 #ifdef INET6
1811 case SIOCSIFPHYADDR_IN6:
1812 #endif
1813 case SIOCSLIFPHYADDR:
1814 case SIOCADDMULTI:
1815 case SIOCDELMULTI:
1816 case SIOCSIFMEDIA:
1817 case SIOCSDRVSPEC:
1818 case SIOCG80211:
1819 case SIOCS80211:
1820 case SIOCS80211NWID:
1821 case SIOCS80211NWKEY:
1822 case SIOCS80211POWER:
1823 case SIOCS80211BSSID:
1824 case SIOCS80211CHANNEL:
1825 case SIOCSLINKSTR:
1826 if (l != NULL) {
1827 error = kauth_authorize_network(l->l_cred,
1828 KAUTH_NETWORK_INTERFACE,
1829 KAUTH_REQ_NETWORK_INTERFACE_SETPRIV, ifp,
1830 (void *)cmd, NULL);
1831 if (error != 0)
1832 return error;
1833 }
1834 }
1835
1836 oif_flags = ifp->if_flags;
1837
1838 ifnet_lock_enter(ifp->if_ioctl_lock);
1839 error = (*ifp->if_ioctl)(ifp, cmd, data);
1840 if (error != ENOTTY)
1841 ;
1842 else if (so->so_proto == NULL)
1843 error = EOPNOTSUPP;
1844 else {
1845 #ifdef COMPAT_OSOCK
1846 error = compat_ifioctl(so, ocmd, cmd, data, l);
1847 #else
1848 error = (*so->so_proto->pr_usrreq)(so, PRU_CONTROL,
1849 (struct mbuf *)cmd, (struct mbuf *)data,
1850 (struct mbuf *)ifp, l);
1851 #endif
1852 }
1853
1854 if (((oif_flags ^ ifp->if_flags) & IFF_UP) != 0) {
1855 #ifdef INET6
1856 if ((ifp->if_flags & IFF_UP) != 0) {
1857 int s = splnet();
1858 in6_if_up(ifp);
1859 splx(s);
1860 }
1861 #endif
1862 }
1863 #ifdef COMPAT_OIFREQ
1864 if (cmd != ocmd)
1865 ifreqn2o(oifr, ifr);
1866 #endif
1867
1868 ifnet_lock_exit(ifp->if_ioctl_lock);
1869 return error;
1870 }
1871
1872 /* This callback adds to the sum in `arg' the number of
1873 * threads on `ci' who have entered or who wait to enter the
1874 * critical section.
1875 */
1876 static void
1877 ifnet_lock_sum(void *p, void *arg, struct cpu_info *ci)
1878 {
1879 uint64_t *sum = arg, *nenter = p;
1880
1881 *sum += *nenter;
1882 }
1883
1884 /* Return the number of threads who have entered or who wait
1885 * to enter the critical section on all CPUs.
1886 */
1887 static uint64_t
1888 ifnet_lock_entrances(struct ifnet_lock *il)
1889 {
1890 uint64_t sum = 0;
1891
1892 percpu_foreach(il->il_nenter, ifnet_lock_sum, &sum);
1893
1894 return sum;
1895 }
1896
1897 static int
1898 ifioctl_attach(struct ifnet *ifp)
1899 {
1900 struct ifnet_lock *il;
1901
1902 /* If the driver has not supplied its own if_ioctl, then
1903 * supply the default.
1904 */
1905 if (ifp->if_ioctl == NULL)
1906 ifp->if_ioctl = ifioctl_common;
1907
1908 /* Create an ifnet_lock for synchronizing ifioctls. */
1909 if ((il = kmem_zalloc(sizeof(*il), KM_SLEEP)) == NULL)
1910 return ENOMEM;
1911
1912 il->il_nenter = percpu_alloc(sizeof(uint64_t));
1913 if (il->il_nenter == NULL) {
1914 kmem_free(il, sizeof(*il));
1915 return ENOMEM;
1916 }
1917
1918 mutex_init(&il->il_lock, MUTEX_DEFAULT, IPL_NONE);
1919 cv_init(&il->il_emptied, ifp->if_xname);
1920
1921 ifp->if_ioctl_lock = il;
1922
1923 return 0;
1924 }
1925
1926 /*
1927 * This must not be called until after `ifp' has been withdrawn from the
1928 * ifnet tables so that ifioctl() cannot get a handle on it by calling
1929 * ifunit().
1930 */
1931 static void
1932 ifioctl_detach(struct ifnet *ifp)
1933 {
1934 struct ifnet_lock *il;
1935
1936 il = ifp->if_ioctl_lock;
1937 mutex_enter(&il->il_lock);
1938 /* Install if_nullioctl to make sure that any thread that
1939 * subsequently enters the critical section will quit it
1940 * immediately and signal the condition variable that we
1941 * wait on, below.
1942 */
1943 ifp->if_ioctl = if_nullioctl;
1944 /* Sleep while threads are still in the critical section or
1945 * wait to enter it.
1946 */
1947 while (ifnet_lock_entrances(il) != il->il_nexit)
1948 cv_wait(&il->il_emptied, &il->il_lock);
1949 /* At this point, we are the only thread still in the critical
1950 * section, and no new thread can get a handle on the ifioctl
1951 * lock, so it is safe to free its memory.
1952 */
1953 mutex_exit(&il->il_lock);
1954 ifp->if_ioctl_lock = NULL;
1955 percpu_free(il->il_nenter, sizeof(uint64_t));
1956 il->il_nenter = NULL;
1957 mutex_destroy(&il->il_lock);
1958 kmem_free(il, sizeof(*il));
1959 }
1960
1961 /*
1962 * Return interface configuration
1963 * of system. List may be used
1964 * in later ioctl's (above) to get
1965 * other information.
1966 *
1967 * Each record is a struct ifreq. Before the addition of
1968 * sockaddr_storage, the API rule was that sockaddr flavors that did
1969 * not fit would extend beyond the struct ifreq, with the next struct
1970 * ifreq starting sa_len beyond the struct sockaddr. Because the
1971 * union in struct ifreq includes struct sockaddr_storage, every kind
1972 * of sockaddr must fit. Thus, there are no longer any overlength
1973 * records.
1974 *
1975 * Records are added to the user buffer if they fit, and ifc_len is
1976 * adjusted to the length that was written. Thus, the user is only
1977 * assured of getting the complete list if ifc_len on return is at
1978 * least sizeof(struct ifreq) less than it was on entry.
1979 *
1980 * If the user buffer pointer is NULL, this routine copies no data and
1981 * returns the amount of space that would be needed.
1982 *
1983 * Invariants:
1984 * ifrp points to the next part of the user's buffer to be used. If
1985 * ifrp != NULL, space holds the number of bytes remaining that we may
1986 * write at ifrp. Otherwise, space holds the number of bytes that
1987 * would have been written had there been adequate space.
1988 */
1989 /*ARGSUSED*/
1990 int
1991 ifconf(u_long cmd, void *data)
1992 {
1993 struct ifconf *ifc = (struct ifconf *)data;
1994 struct ifnet *ifp;
1995 struct ifaddr *ifa;
1996 struct ifreq ifr, *ifrp;
1997 int space, error = 0;
1998 const int sz = (int)sizeof(struct ifreq);
1999
2000 if ((ifrp = ifc->ifc_req) == NULL)
2001 space = 0;
2002 else
2003 space = ifc->ifc_len;
2004 IFNET_FOREACH(ifp) {
2005 (void)strncpy(ifr.ifr_name, ifp->if_xname,
2006 sizeof(ifr.ifr_name));
2007 if (ifr.ifr_name[sizeof(ifr.ifr_name) - 1] != '\0')
2008 return ENAMETOOLONG;
2009 if (IFADDR_EMPTY(ifp)) {
2010 /* Interface with no addresses - send zero sockaddr. */
2011 memset(&ifr.ifr_addr, 0, sizeof(ifr.ifr_addr));
2012 if (ifrp == NULL) {
2013 space += sz;
2014 continue;
2015 }
2016 if (space >= sz) {
2017 error = copyout(&ifr, ifrp, sz);
2018 if (error != 0)
2019 return error;
2020 ifrp++;
2021 space -= sz;
2022 }
2023 }
2024
2025 IFADDR_FOREACH(ifa, ifp) {
2026 struct sockaddr *sa = ifa->ifa_addr;
2027 /* all sockaddrs must fit in sockaddr_storage */
2028 KASSERT(sa->sa_len <= sizeof(ifr.ifr_ifru));
2029
2030 if (ifrp == NULL) {
2031 space += sz;
2032 continue;
2033 }
2034 memcpy(&ifr.ifr_space, sa, sa->sa_len);
2035 if (space >= sz) {
2036 error = copyout(&ifr, ifrp, sz);
2037 if (error != 0)
2038 return (error);
2039 ifrp++; space -= sz;
2040 }
2041 }
2042 }
2043 if (ifrp != NULL) {
2044 KASSERT(0 <= space && space <= ifc->ifc_len);
2045 ifc->ifc_len -= space;
2046 } else {
2047 KASSERT(space >= 0);
2048 ifc->ifc_len = space;
2049 }
2050 return (0);
2051 }
2052
2053 int
2054 ifreq_setaddr(u_long cmd, struct ifreq *ifr, const struct sockaddr *sa)
2055 {
2056 uint8_t len;
2057 #ifdef COMPAT_OIFREQ
2058 struct ifreq ifrb;
2059 struct oifreq *oifr = NULL;
2060 u_long ocmd = cmd;
2061 cmd = compat_cvtcmd(cmd);
2062 if (cmd != ocmd) {
2063 oifr = (struct oifreq *)(void *)ifr;
2064 ifr = &ifrb;
2065 ifreqo2n(oifr, ifr);
2066 len = sizeof(oifr->ifr_addr);
2067 } else
2068 #endif
2069 len = sizeof(ifr->ifr_ifru.ifru_space);
2070
2071 if (len < sa->sa_len)
2072 return EFBIG;
2073
2074 memset(&ifr->ifr_addr, 0, len);
2075 sockaddr_copy(&ifr->ifr_addr, len, sa);
2076
2077 #ifdef COMPAT_OIFREQ
2078 if (cmd != ocmd)
2079 ifreqn2o(oifr, ifr);
2080 #endif
2081 return 0;
2082 }
2083
2084 /*
2085 * Queue message on interface, and start output if interface
2086 * not yet active.
2087 */
2088 int
2089 ifq_enqueue(struct ifnet *ifp, struct mbuf *m
2090 ALTQ_COMMA ALTQ_DECL(struct altq_pktattr *pktattr))
2091 {
2092 int len = m->m_pkthdr.len;
2093 int mflags = m->m_flags;
2094 int s = splnet();
2095 int error;
2096
2097 IFQ_ENQUEUE(&ifp->if_snd, m, pktattr, error);
2098 if (error != 0)
2099 goto out;
2100 ifp->if_obytes += len;
2101 if (mflags & M_MCAST)
2102 ifp->if_omcasts++;
2103 if ((ifp->if_flags & IFF_OACTIVE) == 0)
2104 (*ifp->if_start)(ifp);
2105 out:
2106 splx(s);
2107 return error;
2108 }
2109
2110 /*
2111 * Queue message on interface, possibly using a second fast queue
2112 */
2113 int
2114 ifq_enqueue2(struct ifnet *ifp, struct ifqueue *ifq, struct mbuf *m
2115 ALTQ_COMMA ALTQ_DECL(struct altq_pktattr *pktattr))
2116 {
2117 int error = 0;
2118
2119 if (ifq != NULL
2120 #ifdef ALTQ
2121 && ALTQ_IS_ENABLED(&ifp->if_snd) == 0
2122 #endif
2123 ) {
2124 if (IF_QFULL(ifq)) {
2125 IF_DROP(&ifp->if_snd);
2126 m_freem(m);
2127 if (error == 0)
2128 error = ENOBUFS;
2129 } else
2130 IF_ENQUEUE(ifq, m);
2131 } else
2132 IFQ_ENQUEUE(&ifp->if_snd, m, pktattr, error);
2133 if (error != 0) {
2134 ++ifp->if_oerrors;
2135 return error;
2136 }
2137 return 0;
2138 }
2139
2140 int
2141 if_addr_init(ifnet_t *ifp, struct ifaddr *ifa, const bool src)
2142 {
2143 int rc;
2144
2145 if (ifp->if_initaddr != NULL)
2146 rc = (*ifp->if_initaddr)(ifp, ifa, src);
2147 else if (src ||
2148 (rc = (*ifp->if_ioctl)(ifp, SIOCSIFDSTADDR, ifa)) == ENOTTY)
2149 rc = (*ifp->if_ioctl)(ifp, SIOCINITIFADDR, ifa);
2150
2151 return rc;
2152 }
2153
2154 int
2155 if_flags_set(ifnet_t *ifp, const short flags)
2156 {
2157 int rc;
2158
2159 if (ifp->if_setflags != NULL)
2160 rc = (*ifp->if_setflags)(ifp, flags);
2161 else {
2162 short cantflags;
2163 struct ifreq ifr;
2164
2165 memset(&ifr, 0, sizeof(ifr));
2166
2167 cantflags = (ifp->if_flags ^ flags) & IFF_CANTCHANGE;
2168 if (cantflags != 0)
2169 ifp->if_flags ^= cantflags;
2170
2171 ifr.ifr_flags = flags & ~IFF_CANTCHANGE;
2172 rc = (*ifp->if_ioctl)(ifp, SIOCSIFFLAGS, &ifr);
2173
2174 if (rc != 0 && cantflags != 0)
2175 ifp->if_flags ^= cantflags;
2176 }
2177
2178 return rc;
2179 }
2180
2181 int
2182 if_mcast_op(ifnet_t *ifp, const unsigned long cmd, const struct sockaddr *sa)
2183 {
2184 int rc;
2185 struct ifreq ifr;
2186
2187 if (ifp->if_mcastop != NULL)
2188 rc = (*ifp->if_mcastop)(ifp, cmd, sa);
2189 else {
2190 ifreq_setaddr(cmd, &ifr, sa);
2191 rc = (*ifp->if_ioctl)(ifp, cmd, &ifr);
2192 }
2193
2194 return rc;
2195 }
2196
2197 static void
2198 sysctl_sndq_setup(struct sysctllog **clog, const char *ifname,
2199 struct ifaltq *ifq)
2200 {
2201 const struct sysctlnode *cnode, *rnode;
2202
2203 if (sysctl_createv(clog, 0, NULL, &rnode,
2204 CTLFLAG_PERMANENT,
2205 CTLTYPE_NODE, "net", NULL,
2206 NULL, 0, NULL, 0,
2207 CTL_NET, CTL_EOL) != 0)
2208 goto bad;
2209
2210 if (sysctl_createv(clog, 0, &rnode, &rnode,
2211 CTLFLAG_PERMANENT,
2212 CTLTYPE_NODE, "interfaces",
2213 SYSCTL_DESCR("Per-interface controls"),
2214 NULL, 0, NULL, 0,
2215 CTL_CREATE, CTL_EOL) != 0)
2216 goto bad;
2217
2218 if (sysctl_createv(clog, 0, &rnode, &rnode,
2219 CTLFLAG_PERMANENT,
2220 CTLTYPE_NODE, ifname,
2221 SYSCTL_DESCR("Interface controls"),
2222 NULL, 0, NULL, 0,
2223 CTL_CREATE, CTL_EOL) != 0)
2224 goto bad;
2225
2226 if (sysctl_createv(clog, 0, &rnode, &rnode,
2227 CTLFLAG_PERMANENT,
2228 CTLTYPE_NODE, "sndq",
2229 SYSCTL_DESCR("Interface output queue controls"),
2230 NULL, 0, NULL, 0,
2231 CTL_CREATE, CTL_EOL) != 0)
2232 goto bad;
2233
2234 if (sysctl_createv(clog, 0, &rnode, &cnode,
2235 CTLFLAG_PERMANENT,
2236 CTLTYPE_INT, "len",
2237 SYSCTL_DESCR("Current output queue length"),
2238 NULL, 0, &ifq->ifq_len, 0,
2239 CTL_CREATE, CTL_EOL) != 0)
2240 goto bad;
2241
2242 if (sysctl_createv(clog, 0, &rnode, &cnode,
2243 CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
2244 CTLTYPE_INT, "maxlen",
2245 SYSCTL_DESCR("Maximum allowed output queue length"),
2246 NULL, 0, &ifq->ifq_maxlen, 0,
2247 CTL_CREATE, CTL_EOL) != 0)
2248 goto bad;
2249
2250 if (sysctl_createv(clog, 0, &rnode, &cnode,
2251 CTLFLAG_PERMANENT,
2252 CTLTYPE_INT, "drops",
2253 SYSCTL_DESCR("Packets dropped due to full output queue"),
2254 NULL, 0, &ifq->ifq_drops, 0,
2255 CTL_CREATE, CTL_EOL) != 0)
2256 goto bad;
2257
2258 return;
2259 bad:
2260 printf("%s: could not attach sysctl nodes\n", ifname);
2261 return;
2262 }
2263
2264 #if defined(INET) || defined(INET6)
2265 static void
2266 sysctl_net_ifq_setup(struct sysctllog **clog,
2267 int pf, const char *pfname,
2268 int ipn, const char *ipname,
2269 int qid, struct ifqueue *ifq)
2270 {
2271
2272 sysctl_createv(clog, 0, NULL, NULL,
2273 CTLFLAG_PERMANENT,
2274 CTLTYPE_NODE, "net", NULL,
2275 NULL, 0, NULL, 0,
2276 CTL_NET, CTL_EOL);
2277 sysctl_createv(clog, 0, NULL, NULL,
2278 CTLFLAG_PERMANENT,
2279 CTLTYPE_NODE, pfname, NULL,
2280 NULL, 0, NULL, 0,
2281 CTL_NET, pf, CTL_EOL);
2282 sysctl_createv(clog, 0, NULL, NULL,
2283 CTLFLAG_PERMANENT,
2284 CTLTYPE_NODE, ipname, NULL,
2285 NULL, 0, NULL, 0,
2286 CTL_NET, pf, ipn, CTL_EOL);
2287 sysctl_createv(clog, 0, NULL, NULL,
2288 CTLFLAG_PERMANENT,
2289 CTLTYPE_NODE, "ifq",
2290 SYSCTL_DESCR("Protocol input queue controls"),
2291 NULL, 0, NULL, 0,
2292 CTL_NET, pf, ipn, qid, CTL_EOL);
2293
2294 sysctl_createv(clog, 0, NULL, NULL,
2295 CTLFLAG_PERMANENT,
2296 CTLTYPE_INT, "len",
2297 SYSCTL_DESCR("Current input queue length"),
2298 NULL, 0, &ifq->ifq_len, 0,
2299 CTL_NET, pf, ipn, qid, IFQCTL_LEN, CTL_EOL);
2300 sysctl_createv(clog, 0, NULL, NULL,
2301 CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
2302 CTLTYPE_INT, "maxlen",
2303 SYSCTL_DESCR("Maximum allowed input queue length"),
2304 NULL, 0, &ifq->ifq_maxlen, 0,
2305 CTL_NET, pf, ipn, qid, IFQCTL_MAXLEN, CTL_EOL);
2306 #ifdef notyet
2307 sysctl_createv(clog, 0, NULL, NULL,
2308 CTLFLAG_PERMANENT,
2309 CTLTYPE_INT, "peak",
2310 SYSCTL_DESCR("Highest input queue length"),
2311 NULL, 0, &ifq->ifq_peak, 0,
2312 CTL_NET, pf, ipn, qid, IFQCTL_PEAK, CTL_EOL);
2313 #endif
2314 sysctl_createv(clog, 0, NULL, NULL,
2315 CTLFLAG_PERMANENT,
2316 CTLTYPE_INT, "drops",
2317 SYSCTL_DESCR("Packets dropped due to full input queue"),
2318 NULL, 0, &ifq->ifq_drops, 0,
2319 CTL_NET, pf, ipn, qid, IFQCTL_DROPS, CTL_EOL);
2320 }
2321 #endif /* INET || INET6 */
2322