xref: /netbsd-src/sys/net/if.c (revision a5847cc334d9a7029f6352b847e9e8d71a0f9e0c)
1 /*	$NetBSD: if.c,v 1.256 2011/10/28 20:11:58 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 1.256 2011/10/28 20:11:58 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@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 	kmem_free(il, sizeof(*il));
1958 }
1959 
1960 /*
1961  * Return interface configuration
1962  * of system.  List may be used
1963  * in later ioctl's (above) to get
1964  * other information.
1965  *
1966  * Each record is a struct ifreq.  Before the addition of
1967  * sockaddr_storage, the API rule was that sockaddr flavors that did
1968  * not fit would extend beyond the struct ifreq, with the next struct
1969  * ifreq starting sa_len beyond the struct sockaddr.  Because the
1970  * union in struct ifreq includes struct sockaddr_storage, every kind
1971  * of sockaddr must fit.  Thus, there are no longer any overlength
1972  * records.
1973  *
1974  * Records are added to the user buffer if they fit, and ifc_len is
1975  * adjusted to the length that was written.  Thus, the user is only
1976  * assured of getting the complete list if ifc_len on return is at
1977  * least sizeof(struct ifreq) less than it was on entry.
1978  *
1979  * If the user buffer pointer is NULL, this routine copies no data and
1980  * returns the amount of space that would be needed.
1981  *
1982  * Invariants:
1983  * ifrp points to the next part of the user's buffer to be used.  If
1984  * ifrp != NULL, space holds the number of bytes remaining that we may
1985  * write at ifrp.  Otherwise, space holds the number of bytes that
1986  * would have been written had there been adequate space.
1987  */
1988 /*ARGSUSED*/
1989 int
1990 ifconf(u_long cmd, void *data)
1991 {
1992 	struct ifconf *ifc = (struct ifconf *)data;
1993 	struct ifnet *ifp;
1994 	struct ifaddr *ifa;
1995 	struct ifreq ifr, *ifrp;
1996 	int space, error = 0;
1997 	const int sz = (int)sizeof(struct ifreq);
1998 
1999 	if ((ifrp = ifc->ifc_req) == NULL)
2000 		space = 0;
2001 	else
2002 		space = ifc->ifc_len;
2003 	IFNET_FOREACH(ifp) {
2004 		(void)strncpy(ifr.ifr_name, ifp->if_xname,
2005 		    sizeof(ifr.ifr_name));
2006 		if (ifr.ifr_name[sizeof(ifr.ifr_name) - 1] != '\0')
2007 			return ENAMETOOLONG;
2008 		if (IFADDR_EMPTY(ifp)) {
2009 			/* Interface with no addresses - send zero sockaddr. */
2010 			memset(&ifr.ifr_addr, 0, sizeof(ifr.ifr_addr));
2011 			if (ifrp == NULL) {
2012 				space += sz;
2013 				continue;
2014 			}
2015 			if (space >= sz) {
2016 				error = copyout(&ifr, ifrp, sz);
2017 				if (error != 0)
2018 					return error;
2019 				ifrp++;
2020 				space -= sz;
2021 			}
2022 		}
2023 
2024 		IFADDR_FOREACH(ifa, ifp) {
2025 			struct sockaddr *sa = ifa->ifa_addr;
2026 			/* all sockaddrs must fit in sockaddr_storage */
2027 			KASSERT(sa->sa_len <= sizeof(ifr.ifr_ifru));
2028 
2029 			if (ifrp == NULL) {
2030 				space += sz;
2031 				continue;
2032 			}
2033 			memcpy(&ifr.ifr_space, sa, sa->sa_len);
2034 			if (space >= sz) {
2035 				error = copyout(&ifr, ifrp, sz);
2036 				if (error != 0)
2037 					return (error);
2038 				ifrp++; space -= sz;
2039 			}
2040 		}
2041 	}
2042 	if (ifrp != NULL) {
2043 		KASSERT(0 <= space && space <= ifc->ifc_len);
2044 		ifc->ifc_len -= space;
2045 	} else {
2046 		KASSERT(space >= 0);
2047 		ifc->ifc_len = space;
2048 	}
2049 	return (0);
2050 }
2051 
2052 int
2053 ifreq_setaddr(u_long cmd, struct ifreq *ifr, const struct sockaddr *sa)
2054 {
2055 	uint8_t len;
2056 #ifdef COMPAT_OIFREQ
2057 	struct ifreq ifrb;
2058 	struct oifreq *oifr = NULL;
2059 	u_long ocmd = cmd;
2060 	cmd = compat_cvtcmd(cmd);
2061 	if (cmd != ocmd) {
2062 		oifr = (struct oifreq *)(void *)ifr;
2063 		ifr = &ifrb;
2064 		ifreqo2n(oifr, ifr);
2065 		len = sizeof(oifr->ifr_addr);
2066 	} else
2067 #endif
2068 		len = sizeof(ifr->ifr_ifru.ifru_space);
2069 
2070 	if (len < sa->sa_len)
2071 		return EFBIG;
2072 
2073 	memset(&ifr->ifr_addr, 0, len);
2074 	sockaddr_copy(&ifr->ifr_addr, len, sa);
2075 
2076 #ifdef COMPAT_OIFREQ
2077 	if (cmd != ocmd)
2078 		ifreqn2o(oifr, ifr);
2079 #endif
2080 	return 0;
2081 }
2082 
2083 /*
2084  * Queue message on interface, and start output if interface
2085  * not yet active.
2086  */
2087 int
2088 ifq_enqueue(struct ifnet *ifp, struct mbuf *m
2089     ALTQ_COMMA ALTQ_DECL(struct altq_pktattr *pktattr))
2090 {
2091 	int len = m->m_pkthdr.len;
2092 	int mflags = m->m_flags;
2093 	int s = splnet();
2094 	int error;
2095 
2096 	IFQ_ENQUEUE(&ifp->if_snd, m, pktattr, error);
2097 	if (error != 0)
2098 		goto out;
2099 	ifp->if_obytes += len;
2100 	if (mflags & M_MCAST)
2101 		ifp->if_omcasts++;
2102 	if ((ifp->if_flags & IFF_OACTIVE) == 0)
2103 		(*ifp->if_start)(ifp);
2104 out:
2105 	splx(s);
2106 	return error;
2107 }
2108 
2109 /*
2110  * Queue message on interface, possibly using a second fast queue
2111  */
2112 int
2113 ifq_enqueue2(struct ifnet *ifp, struct ifqueue *ifq, struct mbuf *m
2114     ALTQ_COMMA ALTQ_DECL(struct altq_pktattr *pktattr))
2115 {
2116 	int error = 0;
2117 
2118 	if (ifq != NULL
2119 #ifdef ALTQ
2120 	    && ALTQ_IS_ENABLED(&ifp->if_snd) == 0
2121 #endif
2122 	    ) {
2123 		if (IF_QFULL(ifq)) {
2124 			IF_DROP(&ifp->if_snd);
2125 			m_freem(m);
2126 			if (error == 0)
2127 				error = ENOBUFS;
2128 		} else
2129 			IF_ENQUEUE(ifq, m);
2130 	} else
2131 		IFQ_ENQUEUE(&ifp->if_snd, m, pktattr, error);
2132 	if (error != 0) {
2133 		++ifp->if_oerrors;
2134 		return error;
2135 	}
2136 	return 0;
2137 }
2138 
2139 int
2140 if_addr_init(ifnet_t *ifp, struct ifaddr *ifa, const bool src)
2141 {
2142 	int rc;
2143 
2144 	if (ifp->if_initaddr != NULL)
2145 		rc = (*ifp->if_initaddr)(ifp, ifa, src);
2146 	else if (src ||
2147 	         (rc = (*ifp->if_ioctl)(ifp, SIOCSIFDSTADDR, ifa)) == ENOTTY)
2148 		rc = (*ifp->if_ioctl)(ifp, SIOCINITIFADDR, ifa);
2149 
2150 	return rc;
2151 }
2152 
2153 int
2154 if_flags_set(ifnet_t *ifp, const short flags)
2155 {
2156 	int rc;
2157 
2158 	if (ifp->if_setflags != NULL)
2159 		rc = (*ifp->if_setflags)(ifp, flags);
2160 	else {
2161 		short cantflags;
2162 		struct ifreq ifr;
2163 
2164 		memset(&ifr, 0, sizeof(ifr));
2165 
2166 		cantflags = (ifp->if_flags ^ flags) & IFF_CANTCHANGE;
2167 		if (cantflags != 0)
2168 			ifp->if_flags ^= cantflags;
2169 
2170 		ifr.ifr_flags = flags & ~IFF_CANTCHANGE;
2171 		rc = (*ifp->if_ioctl)(ifp, SIOCSIFFLAGS, &ifr);
2172 
2173 		if (rc != 0 && cantflags != 0)
2174 			ifp->if_flags ^= cantflags;
2175 	}
2176 
2177 	return rc;
2178 }
2179 
2180 int
2181 if_mcast_op(ifnet_t *ifp, const unsigned long cmd, const struct sockaddr *sa)
2182 {
2183 	int rc;
2184 	struct ifreq ifr;
2185 
2186 	if (ifp->if_mcastop != NULL)
2187 		rc = (*ifp->if_mcastop)(ifp, cmd, sa);
2188 	else {
2189 		ifreq_setaddr(cmd, &ifr, sa);
2190 		rc = (*ifp->if_ioctl)(ifp, cmd, &ifr);
2191 	}
2192 
2193 	return rc;
2194 }
2195 
2196 static void
2197 sysctl_sndq_setup(struct sysctllog **clog, const char *ifname,
2198     struct ifaltq *ifq)
2199 {
2200 	const struct sysctlnode *cnode, *rnode;
2201 
2202 	if (sysctl_createv(clog, 0, NULL, &rnode,
2203 		       CTLFLAG_PERMANENT,
2204 		       CTLTYPE_NODE, "net", NULL,
2205 		       NULL, 0, NULL, 0,
2206 		       CTL_NET, CTL_EOL) != 0)
2207 		goto bad;
2208 
2209 	if (sysctl_createv(clog, 0, &rnode, &rnode,
2210 		       CTLFLAG_PERMANENT,
2211 		       CTLTYPE_NODE, "interfaces",
2212 		       SYSCTL_DESCR("Per-interface controls"),
2213 		       NULL, 0, NULL, 0,
2214 		       CTL_CREATE, CTL_EOL) != 0)
2215 		goto bad;
2216 
2217 	if (sysctl_createv(clog, 0, &rnode, &rnode,
2218 		       CTLFLAG_PERMANENT,
2219 		       CTLTYPE_NODE, ifname,
2220 		       SYSCTL_DESCR("Interface controls"),
2221 		       NULL, 0, NULL, 0,
2222 		       CTL_CREATE, CTL_EOL) != 0)
2223 		goto bad;
2224 
2225 	if (sysctl_createv(clog, 0, &rnode, &rnode,
2226 		       CTLFLAG_PERMANENT,
2227 		       CTLTYPE_NODE, "sndq",
2228 		       SYSCTL_DESCR("Interface output queue controls"),
2229 		       NULL, 0, NULL, 0,
2230 		       CTL_CREATE, CTL_EOL) != 0)
2231 		goto bad;
2232 
2233 	if (sysctl_createv(clog, 0, &rnode, &cnode,
2234 		       CTLFLAG_PERMANENT,
2235 		       CTLTYPE_INT, "len",
2236 		       SYSCTL_DESCR("Current output queue length"),
2237 		       NULL, 0, &ifq->ifq_len, 0,
2238 		       CTL_CREATE, CTL_EOL) != 0)
2239 		goto bad;
2240 
2241 	if (sysctl_createv(clog, 0, &rnode, &cnode,
2242 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
2243 		       CTLTYPE_INT, "maxlen",
2244 		       SYSCTL_DESCR("Maximum allowed output queue length"),
2245 		       NULL, 0, &ifq->ifq_maxlen, 0,
2246 		       CTL_CREATE, CTL_EOL) != 0)
2247 		goto bad;
2248 
2249 	if (sysctl_createv(clog, 0, &rnode, &cnode,
2250 		       CTLFLAG_PERMANENT,
2251 		       CTLTYPE_INT, "drops",
2252 		       SYSCTL_DESCR("Packets dropped due to full output queue"),
2253 		       NULL, 0, &ifq->ifq_drops, 0,
2254 		       CTL_CREATE, CTL_EOL) != 0)
2255 		goto bad;
2256 
2257 	return;
2258 bad:
2259 	printf("%s: could not attach sysctl nodes\n", ifname);
2260 	return;
2261 }
2262 
2263 #if defined(INET) || defined(INET6)
2264 static void
2265 sysctl_net_ifq_setup(struct sysctllog **clog,
2266 		     int pf, const char *pfname,
2267 		     int ipn, const char *ipname,
2268 		     int qid, struct ifqueue *ifq)
2269 {
2270 
2271 	sysctl_createv(clog, 0, NULL, NULL,
2272 		       CTLFLAG_PERMANENT,
2273 		       CTLTYPE_NODE, "net", NULL,
2274 		       NULL, 0, NULL, 0,
2275 		       CTL_NET, CTL_EOL);
2276 	sysctl_createv(clog, 0, NULL, NULL,
2277 		       CTLFLAG_PERMANENT,
2278 		       CTLTYPE_NODE, pfname, NULL,
2279 		       NULL, 0, NULL, 0,
2280 		       CTL_NET, pf, CTL_EOL);
2281 	sysctl_createv(clog, 0, NULL, NULL,
2282 		       CTLFLAG_PERMANENT,
2283 		       CTLTYPE_NODE, ipname, NULL,
2284 		       NULL, 0, NULL, 0,
2285 		       CTL_NET, pf, ipn, CTL_EOL);
2286 	sysctl_createv(clog, 0, NULL, NULL,
2287 		       CTLFLAG_PERMANENT,
2288 		       CTLTYPE_NODE, "ifq",
2289 		       SYSCTL_DESCR("Protocol input queue controls"),
2290 		       NULL, 0, NULL, 0,
2291 		       CTL_NET, pf, ipn, qid, CTL_EOL);
2292 
2293 	sysctl_createv(clog, 0, NULL, NULL,
2294 		       CTLFLAG_PERMANENT,
2295 		       CTLTYPE_INT, "len",
2296 		       SYSCTL_DESCR("Current input queue length"),
2297 		       NULL, 0, &ifq->ifq_len, 0,
2298 		       CTL_NET, pf, ipn, qid, IFQCTL_LEN, CTL_EOL);
2299 	sysctl_createv(clog, 0, NULL, NULL,
2300 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
2301 		       CTLTYPE_INT, "maxlen",
2302 		       SYSCTL_DESCR("Maximum allowed input queue length"),
2303 		       NULL, 0, &ifq->ifq_maxlen, 0,
2304 		       CTL_NET, pf, ipn, qid, IFQCTL_MAXLEN, CTL_EOL);
2305 #ifdef notyet
2306 	sysctl_createv(clog, 0, NULL, NULL,
2307 		       CTLFLAG_PERMANENT,
2308 		       CTLTYPE_INT, "peak",
2309 		       SYSCTL_DESCR("Highest input queue length"),
2310 		       NULL, 0, &ifq->ifq_peak, 0,
2311 		       CTL_NET, pf, ipn, qid, IFQCTL_PEAK, CTL_EOL);
2312 #endif
2313 	sysctl_createv(clog, 0, NULL, NULL,
2314 		       CTLFLAG_PERMANENT,
2315 		       CTLTYPE_INT, "drops",
2316 		       SYSCTL_DESCR("Packets dropped due to full input queue"),
2317 		       NULL, 0, &ifq->ifq_drops, 0,
2318 		       CTL_NET, pf, ipn, qid, IFQCTL_DROPS, CTL_EOL);
2319 }
2320 #endif /* INET || INET6 */
2321