Home | History | Annotate | Line # | Download | only in net
if.c revision 1.108
      1 /*	$NetBSD: if.c,v 1.108 2002/05/30 05:06:28 itojun Exp $	*/
      2 
      3 /*-
      4  * Copyright (c) 1999, 2000, 2001 The NetBSD Foundation, Inc.
      5  * All rights reserved.
      6  *
      7  * This code is derived from software contributed to The NetBSD Foundation
      8  * by William Studnemund 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  * 3. All advertising materials mentioning features or use of this software
     19  *    must display the following acknowledgement:
     20  *	This product includes software developed by the NetBSD
     21  *	Foundation, Inc. and its contributors.
     22  * 4. Neither the name of The NetBSD Foundation nor the names of its
     23  *    contributors may be used to endorse or promote products derived
     24  *    from this software without specific prior written permission.
     25  *
     26  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
     27  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
     28  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
     29  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
     30  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
     31  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
     32  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
     33  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
     34  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
     35  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
     36  * POSSIBILITY OF SUCH DAMAGE.
     37  */
     38 
     39 /*
     40  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
     41  * All rights reserved.
     42  *
     43  * Redistribution and use in source and binary forms, with or without
     44  * modification, are permitted provided that the following conditions
     45  * are met:
     46  * 1. Redistributions of source code must retain the above copyright
     47  *    notice, this list of conditions and the following disclaimer.
     48  * 2. Redistributions in binary form must reproduce the above copyright
     49  *    notice, this list of conditions and the following disclaimer in the
     50  *    documentation and/or other materials provided with the distribution.
     51  * 3. Neither the name of the project nor the names of its contributors
     52  *    may be used to endorse or promote products derived from this software
     53  *    without specific prior written permission.
     54  *
     55  * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
     56  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     57  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     58  * ARE DISCLAIMED.  IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
     59  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     60  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     61  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     62  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     63  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     64  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     65  * SUCH DAMAGE.
     66  */
     67 
     68 /*
     69  * Copyright (c) 1980, 1986, 1993
     70  *	The Regents of the University of California.  All rights reserved.
     71  *
     72  * Redistribution and use in source and binary forms, with or without
     73  * modification, are permitted provided that the following conditions
     74  * are met:
     75  * 1. Redistributions of source code must retain the above copyright
     76  *    notice, this list of conditions and the following disclaimer.
     77  * 2. Redistributions in binary form must reproduce the above copyright
     78  *    notice, this list of conditions and the following disclaimer in the
     79  *    documentation and/or other materials provided with the distribution.
     80  * 3. All advertising materials mentioning features or use of this software
     81  *    must display the following acknowledgement:
     82  *	This product includes software developed by the University of
     83  *	California, Berkeley and its contributors.
     84  * 4. Neither the name of the University nor the names of its contributors
     85  *    may be used to endorse or promote products derived from this software
     86  *    without specific prior written permission.
     87  *
     88  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     89  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     90  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     91  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     92  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     93  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     94  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     95  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     96  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     97  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     98  * SUCH DAMAGE.
     99  *
    100  *	@(#)if.c	8.5 (Berkeley) 1/9/95
    101  */
    102 
    103 #include <sys/cdefs.h>
    104 __KERNEL_RCSID(0, "$NetBSD: if.c,v 1.108 2002/05/30 05:06:28 itojun Exp $");
    105 
    106 #include "opt_inet.h"
    107 
    108 #include "opt_compat_linux.h"
    109 #include "opt_compat_svr4.h"
    110 #include "opt_compat_43.h"
    111 #include "opt_atalk.h"
    112 #include "opt_pfil_hooks.h"
    113 
    114 #include <sys/param.h>
    115 #include <sys/mbuf.h>
    116 #include <sys/systm.h>
    117 #include <sys/callout.h>
    118 #include <sys/proc.h>
    119 #include <sys/socket.h>
    120 #include <sys/socketvar.h>
    121 #include <sys/domain.h>
    122 #include <sys/protosw.h>
    123 #include <sys/kernel.h>
    124 #include <sys/ioctl.h>
    125 
    126 #include <net/if.h>
    127 #include <net/if_dl.h>
    128 #include <net/if_ether.h>
    129 #include <net/if_ieee80211.h>
    130 #include <net/if_types.h>
    131 #include <net/radix.h>
    132 #include <net/route.h>
    133 #include <net/netisr.h>
    134 #ifdef NETATALK
    135 #include <netatalk/at_extern.h>
    136 #include <netatalk/at.h>
    137 #endif
    138 
    139 #ifdef INET6
    140 #include <netinet/in.h>
    141 #include <netinet6/in6_var.h>
    142 #include <netinet6/nd6.h>
    143 #endif
    144 
    145 int	ifqmaxlen = IFQ_MAXLEN;
    146 struct	callout if_slowtimo_ch;
    147 
    148 int netisr;			/* scheduling bits for network */
    149 
    150 int if_rt_walktree __P((struct radix_node *, void *));
    151 
    152 struct if_clone *if_clone_lookup __P((const char *, int *));
    153 int if_clone_list __P((struct if_clonereq *));
    154 
    155 LIST_HEAD(, if_clone) if_cloners = LIST_HEAD_INITIALIZER(if_cloners);
    156 int if_cloners_count;
    157 
    158 #if defined(INET) || defined(INET6) || defined(NETATALK) || defined(NS) || \
    159     defined(ISO) || defined(CCITT) || defined(NATM)
    160 static void if_detach_queues __P((struct ifnet *, struct ifqueue *));
    161 #endif
    162 
    163 /*
    164  * Network interface utility routines.
    165  *
    166  * Routines with ifa_ifwith* names take sockaddr *'s as
    167  * parameters.
    168  */
    169 void
    170 ifinit()
    171 {
    172 
    173 	callout_init(&if_slowtimo_ch);
    174 	if_slowtimo(NULL);
    175 }
    176 
    177 /*
    178  * Null routines used while an interface is going away.  These routines
    179  * just return an error.
    180  */
    181 
    182 int
    183 if_nulloutput(ifp, m, so, rt)
    184 	struct ifnet *ifp;
    185 	struct mbuf *m;
    186 	struct sockaddr *so;
    187 	struct rtentry *rt;
    188 {
    189 
    190 	return (ENXIO);
    191 }
    192 
    193 void
    194 if_nullinput(ifp, m)
    195 	struct ifnet *ifp;
    196 	struct mbuf *m;
    197 {
    198 
    199 	/* Nothing. */
    200 }
    201 
    202 void
    203 if_nullstart(ifp)
    204 	struct ifnet *ifp;
    205 {
    206 
    207 	/* Nothing. */
    208 }
    209 
    210 int
    211 if_nullioctl(ifp, cmd, data)
    212 	struct ifnet *ifp;
    213 	u_long cmd;
    214 	caddr_t data;
    215 {
    216 
    217 	return (ENXIO);
    218 }
    219 
    220 int
    221 if_nullinit(ifp)
    222 	struct ifnet *ifp;
    223 {
    224 
    225 	return (ENXIO);
    226 }
    227 
    228 void
    229 if_nullstop(ifp, disable)
    230 	struct ifnet *ifp;
    231 	int disable;
    232 {
    233 
    234 	/* Nothing. */
    235 }
    236 
    237 void
    238 if_nullwatchdog(ifp)
    239 	struct ifnet *ifp;
    240 {
    241 
    242 	/* Nothing. */
    243 }
    244 
    245 void
    246 if_nulldrain(ifp)
    247 	struct ifnet *ifp;
    248 {
    249 
    250 	/* Nothing. */
    251 }
    252 
    253 int if_index = 1;
    254 struct ifnet_head ifnet;
    255 struct ifaddr **ifnet_addrs = NULL;
    256 struct ifnet **ifindex2ifnet = NULL;
    257 
    258 /*
    259  * Allocate the link level name for the specified interface.  This
    260  * is an attachment helper.  It must be called after ifp->if_addrlen
    261  * is initialized, which may not be the case when if_attach() is
    262  * called.
    263  */
    264 void
    265 if_alloc_sadl(struct ifnet *ifp)
    266 {
    267 	unsigned socksize, ifasize;
    268 	int namelen, masklen;
    269 	struct sockaddr_dl *sdl;
    270 	struct ifaddr *ifa;
    271 
    272 	/*
    273 	 * If the interface already has a link name, release it
    274 	 * now.  This is useful for interfaces that can change
    275 	 * link types, and thus switch link names often.
    276 	 */
    277 	if (ifp->if_sadl != NULL)
    278 		if_free_sadl(ifp);
    279 
    280 	namelen = strlen(ifp->if_xname);
    281 	masklen = offsetof(struct sockaddr_dl, sdl_data[0]) + namelen;
    282 	socksize = masklen + ifp->if_addrlen;
    283 #define ROUNDUP(a) (1 + (((a) - 1) | (sizeof(long) - 1)))
    284 	if (socksize < sizeof(*sdl))
    285 		socksize = sizeof(*sdl);
    286 	socksize = ROUNDUP(socksize);
    287 	ifasize = sizeof(*ifa) + 2 * socksize;
    288 	ifa = (struct ifaddr *)malloc(ifasize, M_IFADDR, M_WAITOK);
    289 	memset((caddr_t)ifa, 0, ifasize);
    290 	sdl = (struct sockaddr_dl *)(ifa + 1);
    291 	sdl->sdl_len = socksize;
    292 	sdl->sdl_family = AF_LINK;
    293 	bcopy(ifp->if_xname, sdl->sdl_data, namelen);
    294 	sdl->sdl_nlen = namelen;
    295 	sdl->sdl_alen = ifp->if_addrlen;
    296 	sdl->sdl_index = ifp->if_index;
    297 	sdl->sdl_type = ifp->if_type;
    298 	ifnet_addrs[ifp->if_index] = ifa;
    299 	IFAREF(ifa);
    300 	ifa->ifa_ifp = ifp;
    301 	ifa->ifa_rtrequest = link_rtrequest;
    302 	TAILQ_INSERT_HEAD(&ifp->if_addrlist, ifa, ifa_list);
    303 	IFAREF(ifa);
    304 	ifa->ifa_addr = (struct sockaddr *)sdl;
    305 	ifp->if_sadl = sdl;
    306 	sdl = (struct sockaddr_dl *)(socksize + (caddr_t)sdl);
    307 	ifa->ifa_netmask = (struct sockaddr *)sdl;
    308 	sdl->sdl_len = masklen;
    309 	while (namelen != 0)
    310 		sdl->sdl_data[--namelen] = 0xff;
    311 }
    312 
    313 /*
    314  * Free the link level name for the specified interface.  This is
    315  * a detach helper.  This is called from if_detach() or from
    316  * link layer type specific detach functions.
    317  */
    318 void
    319 if_free_sadl(struct ifnet *ifp)
    320 {
    321 	struct ifaddr *ifa;
    322 	int s;
    323 
    324 	ifa = ifnet_addrs[ifp->if_index];
    325 	if (ifa == NULL) {
    326 		KASSERT(ifp->if_sadl == NULL);
    327 		return;
    328 	}
    329 
    330 	KASSERT(ifp->if_sadl != NULL);
    331 
    332 	s = splnet();
    333 	rtinit(ifa, RTM_DELETE, 0);
    334 	TAILQ_REMOVE(&ifp->if_addrlist, ifa, ifa_list);
    335 	IFAFREE(ifa);
    336 
    337 	ifp->if_sadl = NULL;
    338 
    339 	ifnet_addrs[ifp->if_index] = NULL;
    340 	IFAFREE(ifa);
    341 	splx(s);
    342 }
    343 
    344 /*
    345  * Attach an interface to the
    346  * list of "active" interfaces.
    347  */
    348 void
    349 if_attach(ifp)
    350 	struct ifnet *ifp;
    351 {
    352 	static size_t if_indexlim = 0;
    353 	int indexlim = 0;
    354 
    355 	if (if_indexlim == 0) {
    356 		TAILQ_INIT(&ifnet);
    357 		if_indexlim = 8;
    358 	}
    359 	TAILQ_INIT(&ifp->if_addrlist);
    360 	TAILQ_INSERT_TAIL(&ifnet, ifp, if_list);
    361 	ifp->if_index = if_index;
    362 	if (ifindex2ifnet == 0)
    363 		if_index++;
    364 	else
    365 		while (ifindex2ifnet[ifp->if_index] != NULL) {
    366 			++if_index;
    367 			if (if_index == 0)
    368 				if_index = 1;
    369 			/*
    370 			 * If we hit USHRT_MAX, we skip back to 0 since
    371 			 * there are a number of places where the value
    372 			 * of if_index or if_index itself is compared
    373 			 * to to or stored in an unsigned short.  By
    374 			 * jumping back, we won't botch those assignments
    375 			 * or comparisons.
    376 			 */
    377 			else if (if_index == USHRT_MAX) {
    378 				/*
    379 				 * However, if we have to jump back to
    380 				 * zero *twice* without finding an empty
    381 				 * slot in ifindex2ifnet[], then there
    382 				 * there are too many (>65535) interfaces.
    383 				 */
    384 				if (indexlim++)
    385 					panic("too many interfaces");
    386 				else
    387 					if_index = 1;
    388 			}
    389 			ifp->if_index = if_index;
    390 		}
    391 
    392 	/*
    393 	 * We have some arrays that should be indexed by if_index.
    394 	 * since if_index will grow dynamically, they should grow too.
    395 	 *	struct ifadd **ifnet_addrs
    396 	 *	struct ifnet **ifindex2ifnet
    397 	 */
    398 	if (ifnet_addrs == 0 || ifindex2ifnet == 0 ||
    399 	    ifp->if_index >= if_indexlim) {
    400 		size_t n;
    401 		caddr_t q;
    402 
    403 		while (ifp->if_index >= if_indexlim)
    404 			if_indexlim <<= 1;
    405 
    406 		/* grow ifnet_addrs */
    407 		n = if_indexlim * sizeof(struct ifaddr *);
    408 		q = (caddr_t)malloc(n, M_IFADDR, M_WAITOK);
    409 		memset(q, 0, n);
    410 		if (ifnet_addrs) {
    411 			bcopy((caddr_t)ifnet_addrs, q, n/2);
    412 			free((caddr_t)ifnet_addrs, M_IFADDR);
    413 		}
    414 		ifnet_addrs = (struct ifaddr **)q;
    415 
    416 		/* grow ifindex2ifnet */
    417 		n = if_indexlim * sizeof(struct ifnet *);
    418 		q = (caddr_t)malloc(n, M_IFADDR, M_WAITOK);
    419 		memset(q, 0, n);
    420 		if (ifindex2ifnet) {
    421 			bcopy((caddr_t)ifindex2ifnet, q, n/2);
    422 			free((caddr_t)ifindex2ifnet, M_IFADDR);
    423 		}
    424 		ifindex2ifnet = (struct ifnet **)q;
    425 	}
    426 
    427 	ifindex2ifnet[ifp->if_index] = ifp;
    428 
    429 	/*
    430 	 * Link level name is allocated later by a separate call to
    431 	 * if_alloc_sadl().
    432 	 */
    433 
    434 	if (ifp->if_snd.ifq_maxlen == 0)
    435 		ifp->if_snd.ifq_maxlen = ifqmaxlen;
    436 	ifp->if_broadcastaddr = 0; /* reliably crash if used uninitialized */
    437 
    438 	ifp->if_link_state = LINK_STATE_UNKNOWN;
    439 
    440 	ifp->if_capenable = 0;
    441 	ifp->if_csum_flags_tx = 0;
    442 	ifp->if_csum_flags_rx = 0;
    443 
    444 #ifdef ALTQ
    445 	ifp->if_snd.altq_type = 0;
    446 	ifp->if_snd.altq_disc = NULL;
    447 	ifp->if_snd.altq_flags &= ALTQF_CANTCHANGE;
    448 	ifp->if_snd.altq_tbr  = NULL;
    449 	ifp->if_snd.altq_ifp  = ifp;
    450 #endif
    451 
    452 #ifdef PFIL_HOOKS
    453 	ifp->if_pfil.ph_type = PFIL_TYPE_IFNET;
    454 	ifp->if_pfil.ph_ifnet = ifp;
    455 	if (pfil_head_register(&ifp->if_pfil) != 0)
    456 		printf("%s: WARNING: unable to register pfil hook\n",
    457 		    ifp->if_xname);
    458 #endif
    459 
    460 	if (domains)
    461 		if_attachdomain1(ifp);
    462 
    463 	/* Announce the interface. */
    464 	rt_ifannouncemsg(ifp, IFAN_ARRIVAL);
    465 }
    466 
    467 void
    468 if_attachdomain()
    469 {
    470 	struct ifnet *ifp;
    471 
    472 	for (ifp = TAILQ_FIRST(&ifnet); ifp; ifp = TAILQ_NEXT(ifp, if_list))
    473 		if_attachdomain1(ifp);
    474 }
    475 
    476 void
    477 if_attachdomain1(ifp)
    478 	struct ifnet *ifp;
    479 {
    480 	struct domain *dp;
    481 
    482 	/* address family dependent data region */
    483 	memset(ifp->if_afdata, 0, sizeof(ifp->if_afdata));
    484 	for (dp = domains; dp; dp = dp->dom_next) {
    485 		if (dp->dom_ifattach)
    486 			ifp->if_afdata[dp->dom_family] =
    487 			    (*dp->dom_ifattach)(ifp);
    488 	}
    489 }
    490 
    491 /*
    492  * Deactivate an interface.  This points all of the procedure
    493  * handles at error stubs.  May be called from interrupt context.
    494  */
    495 void
    496 if_deactivate(ifp)
    497 	struct ifnet *ifp;
    498 {
    499 	int s;
    500 
    501 	s = splnet();
    502 
    503 	ifp->if_output	 = if_nulloutput;
    504 	ifp->if_input	 = if_nullinput;
    505 	ifp->if_start	 = if_nullstart;
    506 	ifp->if_ioctl	 = if_nullioctl;
    507 	ifp->if_init	 = if_nullinit;
    508 	ifp->if_stop	 = if_nullstop;
    509 	ifp->if_watchdog = if_nullwatchdog;
    510 	ifp->if_drain	 = if_nulldrain;
    511 
    512 	/* No more packets may be enqueued. */
    513 	ifp->if_snd.ifq_maxlen = 0;
    514 
    515 	splx(s);
    516 }
    517 
    518 /*
    519  * Detach an interface from the list of "active" interfaces,
    520  * freeing any resources as we go along.
    521  *
    522  * NOTE: This routine must be called with a valid thread context,
    523  * as it may block.
    524  */
    525 void
    526 if_detach(ifp)
    527 	struct ifnet *ifp;
    528 {
    529 	struct socket so;
    530 	struct ifaddr *ifa;
    531 #ifdef IFAREF_DEBUG
    532 	struct ifaddr *last_ifa = NULL;
    533 #endif
    534 	struct domain *dp;
    535 	struct protosw *pr;
    536 	struct radix_node_head *rnh;
    537 	int s, i, family, purged;
    538 
    539 	/*
    540 	 * XXX It's kind of lame that we have to have the
    541 	 * XXX socket structure...
    542 	 */
    543 	memset(&so, 0, sizeof(so));
    544 
    545 	s = splnet();
    546 
    547 	/*
    548 	 * Do an if_down() to give protocols a chance to do something.
    549 	 */
    550 	if_down(ifp);
    551 
    552 #ifdef ALTQ
    553 	if (ALTQ_IS_ENABLED(&ifp->if_snd))
    554 		altq_disable(&ifp->if_snd);
    555 	if (ALTQ_IS_ATTACHED(&ifp->if_snd))
    556 		altq_detach(&ifp->if_snd);
    557 #endif
    558 
    559 #ifdef PFIL_HOOKS
    560 	(void) pfil_head_unregister(&ifp->if_pfil);
    561 #endif
    562 
    563 	if_free_sadl(ifp);
    564 
    565 	/*
    566 	 * Rip all the addresses off the interface.  This should make
    567 	 * all of the routes go away.
    568 	 */
    569 	while ((ifa = TAILQ_FIRST(&ifp->if_addrlist)) != NULL) {
    570 		family = ifa->ifa_addr->sa_family;
    571 #ifdef IFAREF_DEBUG
    572 		printf("if_detach: ifaddr %p, family %d, refcnt %d\n",
    573 		    ifa, family, ifa->ifa_refcnt);
    574 		if (last_ifa != NULL && ifa == last_ifa)
    575 			panic("if_detach: loop detected");
    576 		last_ifa = ifa;
    577 #endif
    578 		if (family == AF_LINK) {
    579 			/*
    580 			 * XXX This case may now be obsolete by
    581 			 * XXX the call to if_free_sadl().
    582 			 */
    583 			rtinit(ifa, RTM_DELETE, 0);
    584 			TAILQ_REMOVE(&ifp->if_addrlist, ifa, ifa_list);
    585 			IFAFREE(ifa);
    586 		} else {
    587 			dp = pffinddomain(family);
    588 #ifdef DIAGNOSTIC
    589 			if (dp == NULL)
    590 				panic("if_detach: no domain for AF %d\n",
    591 				    family);
    592 #endif
    593 			purged = 0;
    594 			for (pr = dp->dom_protosw;
    595 			     pr < dp->dom_protoswNPROTOSW; pr++) {
    596 				so.so_proto = pr;
    597 				if (pr->pr_usrreq != NULL) {
    598 					(void) (*pr->pr_usrreq)(&so,
    599 					    PRU_PURGEIF, NULL, NULL,
    600 					    (struct mbuf *) ifp, curproc);
    601 					purged = 1;
    602 				}
    603 			}
    604 			if (purged == 0) {
    605 				/*
    606 				 * XXX What's really the best thing to do
    607 				 * XXX here?  --thorpej (at) netbsd.org
    608 				 */
    609 				printf("if_detach: WARNING: AF %d not purged\n",
    610 				    family);
    611 			}
    612 		}
    613 	}
    614 
    615 	/* Walk the routing table looking for straglers. */
    616 	for (i = 0; i <= AF_MAX; i++) {
    617 		if ((rnh = rt_tables[i]) != NULL)
    618 			(void) (*rnh->rnh_walktree)(rnh, if_rt_walktree, ifp);
    619 	}
    620 
    621 	for (dp = domains; dp; dp = dp->dom_next) {
    622 		if (dp->dom_ifdetach && ifp->if_afdata[dp->dom_family])
    623 			(*dp->dom_ifdetach)(ifp,
    624 			    ifp->if_afdata[dp->dom_family]);
    625 	}
    626 
    627 	/* Announce that the interface is gone. */
    628 	rt_ifannouncemsg(ifp, IFAN_DEPARTURE);
    629 
    630 	ifindex2ifnet[ifp->if_index] = NULL;
    631 
    632 	TAILQ_REMOVE(&ifnet, ifp, if_list);
    633 
    634 	/*
    635 	 * remove packets came from ifp, from software interrupt queues.
    636 	 * net/netisr_dispatch.h is not usable, as some of them use
    637 	 * strange queue names.
    638 	 */
    639 #define IF_DETACH_QUEUES(x) \
    640 do { \
    641 	extern struct ifqueue x; \
    642 	if_detach_queues(ifp, & x); \
    643 } while (0)
    644 #ifdef INET
    645 #if NARP > 0
    646 	IF_DETACH_QUEUES(arpintrq);
    647 #endif
    648 	IF_DETACH_QUEUES(ipintrq);
    649 #endif
    650 #ifdef INET6
    651 	IF_DETACH_QUEUES(ip6intrq);
    652 #endif
    653 #ifdef NETATALK
    654 	IF_DETACH_QUEUES(atintrq1);
    655 	IF_DETACH_QUEUES(atintrq2);
    656 #endif
    657 #ifdef NS
    658 	IF_DETACH_QUEUES(nsintrq);
    659 #endif
    660 #ifdef ISO
    661 	IF_DETACH_QUEUES(clnlintrq);
    662 #endif
    663 #ifdef CCITT
    664 	IF_DETACH_QUEUES(llcintrq);
    665 	IF_DETACH_QUEUES(hdintrq);
    666 #endif
    667 #ifdef NATM
    668 	IF_DETACH_QUEUES(natmintrq);
    669 #endif
    670 #undef IF_DETACH_QUEUES
    671 
    672 	splx(s);
    673 }
    674 
    675 #if defined(INET) || defined(INET6) || defined(NETATALK) || defined(NS) || \
    676     defined(ISO) || defined(CCITT) || defined(NATM)
    677 static void
    678 if_detach_queues(ifp, q)
    679 	struct ifnet *ifp;
    680 	struct ifqueue *q;
    681 {
    682 	struct mbuf *m, *prev, *next;
    683 
    684 	prev = NULL;
    685 	for (m = q->ifq_head; m; m = next) {
    686 		next = m->m_nextpkt;
    687 #ifdef DIAGNOSTIC
    688 		if ((m->m_flags & M_PKTHDR) == 0) {
    689 			prev = m;
    690 			continue;
    691 		}
    692 #endif
    693 		if (m->m_pkthdr.rcvif != ifp) {
    694 			prev = m;
    695 			continue;
    696 		}
    697 
    698 		if (prev)
    699 			prev->m_nextpkt = m->m_nextpkt;
    700 		else
    701 			q->ifq_head = m->m_nextpkt;
    702 		if (q->ifq_tail == m)
    703 			q->ifq_tail = prev;
    704 		q->ifq_len--;
    705 
    706 		m->m_nextpkt = NULL;
    707 		m_freem(m);
    708 		IF_DROP(q);
    709 	}
    710 }
    711 #endif /* defined(INET) || ... */
    712 
    713 /*
    714  * Callback for a radix tree walk to delete all references to an
    715  * ifnet.
    716  */
    717 int
    718 if_rt_walktree(rn, v)
    719 	struct radix_node *rn;
    720 	void *v;
    721 {
    722 	struct ifnet *ifp = (struct ifnet *)v;
    723 	struct rtentry *rt = (struct rtentry *)rn;
    724 	int error;
    725 
    726 	if (rt->rt_ifp == ifp) {
    727 		/* Delete the entry. */
    728 		error = rtrequest(RTM_DELETE, rt_key(rt), rt->rt_gateway,
    729 		    rt_mask(rt), rt->rt_flags, NULL);
    730 		if (error)
    731 			printf("%s: warning: unable to delete rtentry @ %p, "
    732 			    "error = %d\n", ifp->if_xname, rt, error);
    733 	}
    734 	return (0);
    735 }
    736 
    737 /*
    738  * Create a clone network interface.
    739  */
    740 int
    741 if_clone_create(name)
    742 	const char *name;
    743 {
    744 	struct if_clone *ifc;
    745 	int unit;
    746 
    747 	ifc = if_clone_lookup(name, &unit);
    748 	if (ifc == NULL)
    749 		return (EINVAL);
    750 
    751 	if (ifunit(name) != NULL)
    752 		return (EEXIST);
    753 
    754 	return ((*ifc->ifc_create)(ifc, unit));
    755 }
    756 
    757 /*
    758  * Destroy a clone network interface.
    759  */
    760 int
    761 if_clone_destroy(name)
    762 	const char *name;
    763 {
    764 	struct if_clone *ifc;
    765 	struct ifnet *ifp;
    766 
    767 	ifc = if_clone_lookup(name, NULL);
    768 	if (ifc == NULL)
    769 		return (EINVAL);
    770 
    771 	ifp = ifunit(name);
    772 	if (ifp == NULL)
    773 		return (ENXIO);
    774 
    775 	if (ifc->ifc_destroy == NULL)
    776 		return (EOPNOTSUPP);
    777 
    778 	(*ifc->ifc_destroy)(ifp);
    779 	return (0);
    780 }
    781 
    782 /*
    783  * Look up a network interface cloner.
    784  */
    785 struct if_clone *
    786 if_clone_lookup(name, unitp)
    787 	const char *name;
    788 	int *unitp;
    789 {
    790 	struct if_clone *ifc;
    791 	const char *cp;
    792 	int i;
    793 
    794 	for (ifc = LIST_FIRST(&if_cloners); ifc != NULL;) {
    795 		for (cp = name, i = 0; i < ifc->ifc_namelen; i++, cp++) {
    796 			if (ifc->ifc_name[i] != *cp)
    797 				goto next_ifc;
    798 		}
    799 		goto found_name;
    800  next_ifc:
    801 		ifc = LIST_NEXT(ifc, ifc_list);
    802 	}
    803 
    804 	/* No match. */
    805 	return (NULL);
    806 
    807  found_name:
    808 	for (i = 0; *cp != '\0'; cp++) {
    809 		if (*cp < '0' || *cp > '9') {
    810 			/* Bogus unit number. */
    811 			return (NULL);
    812 		}
    813 		i = (i * 10) + (*cp - '0');
    814 	}
    815 
    816 	if (unitp != NULL)
    817 		*unitp = i;
    818 	return (ifc);
    819 }
    820 
    821 /*
    822  * Register a network interface cloner.
    823  */
    824 void
    825 if_clone_attach(ifc)
    826 	struct if_clone *ifc;
    827 {
    828 
    829 	LIST_INSERT_HEAD(&if_cloners, ifc, ifc_list);
    830 	if_cloners_count++;
    831 }
    832 
    833 /*
    834  * Unregister a network interface cloner.
    835  */
    836 void
    837 if_clone_detach(ifc)
    838 	struct if_clone *ifc;
    839 {
    840 
    841 	LIST_REMOVE(ifc, ifc_list);
    842 	if_cloners_count--;
    843 }
    844 
    845 /*
    846  * Provide list of interface cloners to userspace.
    847  */
    848 int
    849 if_clone_list(ifcr)
    850 	struct if_clonereq *ifcr;
    851 {
    852 	char outbuf[IFNAMSIZ], *dst;
    853 	struct if_clone *ifc;
    854 	int count, error = 0;
    855 
    856 	ifcr->ifcr_total = if_cloners_count;
    857 	if ((dst = ifcr->ifcr_buffer) == NULL) {
    858 		/* Just asking how many there are. */
    859 		return (0);
    860 	}
    861 
    862 	if (ifcr->ifcr_count < 0)
    863 		return (EINVAL);
    864 
    865 	count = (if_cloners_count < ifcr->ifcr_count) ?
    866 	    if_cloners_count : ifcr->ifcr_count;
    867 
    868 	for (ifc = LIST_FIRST(&if_cloners); ifc != NULL && count != 0;
    869 	     ifc = LIST_NEXT(ifc, ifc_list), count--, dst += IFNAMSIZ) {
    870 		strncpy(outbuf, ifc->ifc_name, IFNAMSIZ);
    871 		outbuf[IFNAMSIZ - 1] = '\0';	/* sanity */
    872 		error = copyout(outbuf, dst, IFNAMSIZ);
    873 		if (error)
    874 			break;
    875 	}
    876 
    877 	return (error);
    878 }
    879 
    880 /*
    881  * Locate an interface based on a complete address.
    882  */
    883 /*ARGSUSED*/
    884 struct ifaddr *
    885 ifa_ifwithaddr(addr)
    886 	struct sockaddr *addr;
    887 {
    888 	struct ifnet *ifp;
    889 	struct ifaddr *ifa;
    890 
    891 #define	equal(a1, a2) \
    892   (bcmp((caddr_t)(a1), (caddr_t)(a2), ((struct sockaddr *)(a1))->sa_len) == 0)
    893 
    894 	for (ifp = TAILQ_FIRST(&ifnet); ifp != NULL;
    895 	     ifp = TAILQ_NEXT(ifp, if_list)) {
    896 		if (ifp->if_output == if_nulloutput)
    897 			continue;
    898 		for (ifa = TAILQ_FIRST(&ifp->if_addrlist); ifa != NULL;
    899 		     ifa = TAILQ_NEXT(ifa, ifa_list)) {
    900 			if (ifa->ifa_addr->sa_family != addr->sa_family)
    901 				continue;
    902 			if (equal(addr, ifa->ifa_addr))
    903 				return (ifa);
    904 			if ((ifp->if_flags & IFF_BROADCAST) &&
    905 			    ifa->ifa_broadaddr &&
    906 			    /* IP6 doesn't have broadcast */
    907 			    ifa->ifa_broadaddr->sa_len != 0 &&
    908 			    equal(ifa->ifa_broadaddr, addr))
    909 				return (ifa);
    910 		}
    911 	}
    912 	return (NULL);
    913 }
    914 
    915 /*
    916  * Locate the point to point interface with a given destination address.
    917  */
    918 /*ARGSUSED*/
    919 struct ifaddr *
    920 ifa_ifwithdstaddr(addr)
    921 	struct sockaddr *addr;
    922 {
    923 	struct ifnet *ifp;
    924 	struct ifaddr *ifa;
    925 
    926 	for (ifp = TAILQ_FIRST(&ifnet); ifp != NULL;
    927 	     ifp = TAILQ_NEXT(ifp, if_list)) {
    928 		if (ifp->if_output == if_nulloutput)
    929 			continue;
    930 		if (ifp->if_flags & IFF_POINTOPOINT) {
    931 			for (ifa = TAILQ_FIRST(&ifp->if_addrlist); ifa != NULL;
    932 			     ifa = TAILQ_NEXT(ifa, ifa_list)) {
    933 				if (ifa->ifa_addr->sa_family !=
    934 				      addr->sa_family ||
    935 				    ifa->ifa_dstaddr == NULL)
    936 					continue;
    937 				if (equal(addr, ifa->ifa_dstaddr))
    938 					return (ifa);
    939 			}
    940 		}
    941 	}
    942 	return (NULL);
    943 }
    944 
    945 /*
    946  * Find an interface on a specific network.  If many, choice
    947  * is most specific found.
    948  */
    949 struct ifaddr *
    950 ifa_ifwithnet(addr)
    951 	struct sockaddr *addr;
    952 {
    953 	struct ifnet *ifp;
    954 	struct ifaddr *ifa;
    955 	struct sockaddr_dl *sdl;
    956 	struct ifaddr *ifa_maybe = 0;
    957 	u_int af = addr->sa_family;
    958 	char *addr_data = addr->sa_data, *cplim;
    959 
    960 	if (af == AF_LINK) {
    961 		sdl = (struct sockaddr_dl *)addr;
    962 		if (sdl->sdl_index && sdl->sdl_index <= if_index &&
    963 		    ifindex2ifnet[sdl->sdl_index]->if_output != if_nulloutput)
    964 			return (ifnet_addrs[sdl->sdl_index]);
    965 	}
    966 #ifdef NETATALK
    967 	if (af == AF_APPLETALK) {
    968 		struct sockaddr_at *sat, *sat2;
    969 		sat = (struct sockaddr_at *)addr;
    970 		for (ifp = TAILQ_FIRST(&ifnet); ifp != NULL;
    971 		     ifp = TAILQ_NEXT(ifp, if_list)) {
    972 			if (ifp->if_output == if_nulloutput)
    973 				continue;
    974 			ifa = at_ifawithnet((struct sockaddr_at *)addr, ifp);
    975 			if (ifa == NULL)
    976 				continue;
    977 			sat2 = (struct sockaddr_at *)ifa->ifa_addr;
    978 			if (sat2->sat_addr.s_net == sat->sat_addr.s_net)
    979 				return (ifa); /* exact match */
    980 			if (ifa_maybe == NULL) {
    981 				/* else keep the if with the rigth range */
    982 				ifa_maybe = ifa;
    983 			}
    984 		}
    985 		return (ifa_maybe);
    986 	}
    987 #endif
    988 	for (ifp = TAILQ_FIRST(&ifnet); ifp != NULL;
    989 	     ifp = TAILQ_NEXT(ifp, if_list)) {
    990 		if (ifp->if_output == if_nulloutput)
    991 			continue;
    992 		for (ifa = TAILQ_FIRST(&ifp->if_addrlist); ifa != NULL;
    993 		     ifa = TAILQ_NEXT(ifa, ifa_list)) {
    994 			char *cp, *cp2, *cp3;
    995 
    996 			if (ifa->ifa_addr->sa_family != af ||
    997 			    ifa->ifa_netmask == 0)
    998  next:				continue;
    999 			cp = addr_data;
   1000 			cp2 = ifa->ifa_addr->sa_data;
   1001 			cp3 = ifa->ifa_netmask->sa_data;
   1002 			cplim = (char *)ifa->ifa_netmask +
   1003 			    ifa->ifa_netmask->sa_len;
   1004 			while (cp3 < cplim) {
   1005 				if ((*cp++ ^ *cp2++) & *cp3++) {
   1006 					/* want to continue for() loop */
   1007 					goto next;
   1008 				}
   1009 			}
   1010 			if (ifa_maybe == 0 ||
   1011 			    rn_refines((caddr_t)ifa->ifa_netmask,
   1012 			    (caddr_t)ifa_maybe->ifa_netmask))
   1013 				ifa_maybe = ifa;
   1014 		}
   1015 	}
   1016 	return (ifa_maybe);
   1017 }
   1018 
   1019 /*
   1020  * Find the interface of the addresss.
   1021  */
   1022 struct ifaddr *
   1023 ifa_ifwithladdr(addr)
   1024 	struct sockaddr *addr;
   1025 {
   1026 	struct ifaddr *ia;
   1027 
   1028 	if ((ia = ifa_ifwithaddr(addr)) || (ia = ifa_ifwithdstaddr(addr)) ||
   1029 	    (ia = ifa_ifwithnet(addr)))
   1030 		return (ia);
   1031 	return (NULL);
   1032 }
   1033 
   1034 /*
   1035  * Find an interface using a specific address family
   1036  */
   1037 struct ifaddr *
   1038 ifa_ifwithaf(af)
   1039 	int af;
   1040 {
   1041 	struct ifnet *ifp;
   1042 	struct ifaddr *ifa;
   1043 
   1044 	for (ifp = TAILQ_FIRST(&ifnet); ifp != NULL;
   1045 	     ifp = TAILQ_NEXT(ifp, if_list)) {
   1046 		if (ifp->if_output == if_nulloutput)
   1047 			continue;
   1048 		for (ifa = TAILQ_FIRST(&ifp->if_addrlist); ifa != NULL;
   1049 		     ifa = TAILQ_NEXT(ifa, ifa_list)) {
   1050 			if (ifa->ifa_addr->sa_family == af)
   1051 				return (ifa);
   1052 		}
   1053 	}
   1054 	return (NULL);
   1055 }
   1056 
   1057 /*
   1058  * Find an interface address specific to an interface best matching
   1059  * a given address.
   1060  */
   1061 struct ifaddr *
   1062 ifaof_ifpforaddr(addr, ifp)
   1063 	struct sockaddr *addr;
   1064 	struct ifnet *ifp;
   1065 {
   1066 	struct ifaddr *ifa;
   1067 	char *cp, *cp2, *cp3;
   1068 	char *cplim;
   1069 	struct ifaddr *ifa_maybe = 0;
   1070 	u_int af = addr->sa_family;
   1071 
   1072 	if (ifp->if_output == if_nulloutput)
   1073 		return (NULL);
   1074 
   1075 	if (af >= AF_MAX)
   1076 		return (NULL);
   1077 
   1078 	for (ifa = TAILQ_FIRST(&ifp->if_addrlist); ifa != NULL;
   1079 	     ifa = TAILQ_NEXT(ifa, ifa_list)) {
   1080 		if (ifa->ifa_addr->sa_family != af)
   1081 			continue;
   1082 		ifa_maybe = ifa;
   1083 		if (ifa->ifa_netmask == 0) {
   1084 			if (equal(addr, ifa->ifa_addr) ||
   1085 			    (ifa->ifa_dstaddr &&
   1086 			     equal(addr, ifa->ifa_dstaddr)))
   1087 				return (ifa);
   1088 			continue;
   1089 		}
   1090 		cp = addr->sa_data;
   1091 		cp2 = ifa->ifa_addr->sa_data;
   1092 		cp3 = ifa->ifa_netmask->sa_data;
   1093 		cplim = ifa->ifa_netmask->sa_len + (char *)ifa->ifa_netmask;
   1094 		for (; cp3 < cplim; cp3++) {
   1095 			if ((*cp++ ^ *cp2++) & *cp3)
   1096 				break;
   1097 		}
   1098 		if (cp3 == cplim)
   1099 			return (ifa);
   1100 	}
   1101 	return (ifa_maybe);
   1102 }
   1103 
   1104 /*
   1105  * Default action when installing a route with a Link Level gateway.
   1106  * Lookup an appropriate real ifa to point to.
   1107  * This should be moved to /sys/net/link.c eventually.
   1108  */
   1109 void
   1110 link_rtrequest(cmd, rt, info)
   1111 	int cmd;
   1112 	struct rtentry *rt;
   1113 	struct rt_addrinfo *info;
   1114 {
   1115 	struct ifaddr *ifa;
   1116 	struct sockaddr *dst;
   1117 	struct ifnet *ifp;
   1118 
   1119 	if (cmd != RTM_ADD || ((ifa = rt->rt_ifa) == 0) ||
   1120 	    ((ifp = ifa->ifa_ifp) == 0) || ((dst = rt_key(rt)) == 0))
   1121 		return;
   1122 	if ((ifa = ifaof_ifpforaddr(dst, ifp)) != NULL) {
   1123 		IFAFREE(rt->rt_ifa);
   1124 		rt->rt_ifa = ifa;
   1125 		IFAREF(ifa);
   1126 		if (ifa->ifa_rtrequest && ifa->ifa_rtrequest != link_rtrequest)
   1127 			ifa->ifa_rtrequest(cmd, rt, info);
   1128 	}
   1129 }
   1130 
   1131 /*
   1132  * Mark an interface down and notify protocols of
   1133  * the transition.
   1134  * NOTE: must be called at splsoftnet or equivalent.
   1135  */
   1136 void
   1137 if_down(ifp)
   1138 	struct ifnet *ifp;
   1139 {
   1140 	struct ifaddr *ifa;
   1141 
   1142 	ifp->if_flags &= ~IFF_UP;
   1143 	microtime(&ifp->if_lastchange);
   1144 	for (ifa = TAILQ_FIRST(&ifp->if_addrlist); ifa != NULL;
   1145 	     ifa = TAILQ_NEXT(ifa, ifa_list))
   1146 		pfctlinput(PRC_IFDOWN, ifa->ifa_addr);
   1147 	IFQ_PURGE(&ifp->if_snd);
   1148 	rt_ifmsg(ifp);
   1149 }
   1150 
   1151 /*
   1152  * Mark an interface up and notify protocols of
   1153  * the transition.
   1154  * NOTE: must be called at splsoftnet or equivalent.
   1155  */
   1156 void
   1157 if_up(ifp)
   1158 	struct ifnet *ifp;
   1159 {
   1160 #ifdef notyet
   1161 	struct ifaddr *ifa;
   1162 #endif
   1163 
   1164 	ifp->if_flags |= IFF_UP;
   1165 	microtime(&ifp->if_lastchange);
   1166 #ifdef notyet
   1167 	/* this has no effect on IP, and will kill all ISO connections XXX */
   1168 	for (ifa = TAILQ_FIRST(&ifp->if_addrlist); ifa != NULL;
   1169 	     ifa = TAILQ_NEXT(ifa, ifa_list))
   1170 		pfctlinput(PRC_IFUP, ifa->ifa_addr);
   1171 #endif
   1172 	rt_ifmsg(ifp);
   1173 #ifdef INET6
   1174 	in6_if_up(ifp);
   1175 #endif
   1176 }
   1177 
   1178 /*
   1179  * Handle interface watchdog timer routines.  Called
   1180  * from softclock, we decrement timers (if set) and
   1181  * call the appropriate interface routine on expiration.
   1182  */
   1183 void
   1184 if_slowtimo(arg)
   1185 	void *arg;
   1186 {
   1187 	struct ifnet *ifp;
   1188 	int s = splnet();
   1189 
   1190 	for (ifp = TAILQ_FIRST(&ifnet); ifp != NULL;
   1191 	     ifp = TAILQ_NEXT(ifp, if_list)) {
   1192 		if (ifp->if_timer == 0 || --ifp->if_timer)
   1193 			continue;
   1194 		if (ifp->if_watchdog)
   1195 			(*ifp->if_watchdog)(ifp);
   1196 	}
   1197 	splx(s);
   1198 	callout_reset(&if_slowtimo_ch, hz / IFNET_SLOWHZ,
   1199 	    if_slowtimo, NULL);
   1200 }
   1201 
   1202 /*
   1203  * Set/clear promiscuous mode on interface ifp based on the truth value
   1204  * of pswitch.  The calls are reference counted so that only the first
   1205  * "on" request actually has an effect, as does the final "off" request.
   1206  * Results are undefined if the "off" and "on" requests are not matched.
   1207  */
   1208 int
   1209 ifpromisc(ifp, pswitch)
   1210 	struct ifnet *ifp;
   1211 	int pswitch;
   1212 {
   1213 	int pcount, ret;
   1214 	short flags;
   1215 	struct ifreq ifr;
   1216 
   1217 	pcount = ifp->if_pcount;
   1218 	flags = ifp->if_flags;
   1219 	if (pswitch) {
   1220 		/*
   1221 		 * Allow the device to be "placed" into promiscuous
   1222 		 * mode even if it is not configured up.  It will
   1223 		 * consult IFF_PROMISC when it is is brought up.
   1224 		 */
   1225 		if (ifp->if_pcount++ != 0)
   1226 			return (0);
   1227 		ifp->if_flags |= IFF_PROMISC;
   1228 		if ((ifp->if_flags & IFF_UP) == 0)
   1229 			return (0);
   1230 	} else {
   1231 		if (--ifp->if_pcount > 0)
   1232 			return (0);
   1233 		ifp->if_flags &= ~IFF_PROMISC;
   1234 		/*
   1235 		 * If the device is not configured up, we should not need to
   1236 		 * turn off promiscuous mode (device should have turned it
   1237 		 * off when interface went down; and will look at IFF_PROMISC
   1238 		 * again next time interface comes up).
   1239 		 */
   1240 		if ((ifp->if_flags & IFF_UP) == 0)
   1241 			return (0);
   1242 	}
   1243 	memset(&ifr, 0, sizeof(ifr));
   1244 	ifr.ifr_flags = ifp->if_flags;
   1245 	ret = (*ifp->if_ioctl)(ifp, SIOCSIFFLAGS, (caddr_t) &ifr);
   1246 	/* Restore interface state if not successful. */
   1247 	if (ret != 0) {
   1248 		ifp->if_pcount = pcount;
   1249 		ifp->if_flags = flags;
   1250 	}
   1251 	return (ret);
   1252 }
   1253 
   1254 /*
   1255  * Map interface name to
   1256  * interface structure pointer.
   1257  */
   1258 struct ifnet *
   1259 ifunit(name)
   1260 	const char *name;
   1261 {
   1262 	struct ifnet *ifp;
   1263 	const char *cp = name;
   1264 	u_int unit = 0;
   1265 	u_int i;
   1266 
   1267 	/*
   1268 	 * If the entire name is a number, treat it as an ifindex.
   1269 	 */
   1270 	for (i = 0; i < IFNAMSIZ && *cp >= '0' && *cp <= '9'; i++, cp++) {
   1271 		unit = unit * 10 + (*cp - '0');
   1272 	}
   1273 
   1274 	/*
   1275 	 * If the number took all of the name, then it's a valid ifindex.
   1276 	 */
   1277 	if (i == IFNAMSIZ || (cp != name && *cp == '\0')) {
   1278 		if (unit >= if_index)
   1279 			return (NULL);
   1280 		ifp = ifindex2ifnet[unit];
   1281 		if (ifp == NULL || ifp->if_output == if_nulloutput)
   1282 			return (NULL);
   1283 		return (ifp);
   1284 	}
   1285 
   1286 	for (ifp = TAILQ_FIRST(&ifnet); ifp != NULL;
   1287 	     ifp = TAILQ_NEXT(ifp, if_list)) {
   1288 		if (ifp->if_output == if_nulloutput)
   1289 			continue;
   1290 	 	if (strcmp(ifp->if_xname, name) == 0)
   1291 			return (ifp);
   1292 	}
   1293 	return (NULL);
   1294 }
   1295 
   1296 /*
   1297  * Interface ioctls.
   1298  */
   1299 int
   1300 ifioctl(so, cmd, data, p)
   1301 	struct socket *so;
   1302 	u_long cmd;
   1303 	caddr_t data;
   1304 	struct proc *p;
   1305 {
   1306 	struct ifnet *ifp;
   1307 	struct ifreq *ifr;
   1308 	struct ifcapreq *ifcr;
   1309 	struct ifdatareq *ifdr;
   1310 	int s, error = 0;
   1311 	short oif_flags;
   1312 
   1313 	switch (cmd) {
   1314 
   1315 	case SIOCGIFCONF:
   1316 	case OSIOCGIFCONF:
   1317 		return (ifconf(cmd, data));
   1318 	}
   1319 	ifr = (struct ifreq *)data;
   1320 	ifcr = (struct ifcapreq *)data;
   1321 	ifdr = (struct ifdatareq *)data;
   1322 
   1323 	switch (cmd) {
   1324 	case SIOCIFCREATE:
   1325 	case SIOCIFDESTROY:
   1326 		if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
   1327 			return (error);
   1328 		return ((cmd == SIOCIFCREATE) ?
   1329 			if_clone_create(ifr->ifr_name) :
   1330 			if_clone_destroy(ifr->ifr_name));
   1331 
   1332 	case SIOCIFGCLONERS:
   1333 		return (if_clone_list((struct if_clonereq *)data));
   1334 	}
   1335 
   1336 	ifp = ifunit(ifr->ifr_name);
   1337 	if (ifp == 0)
   1338 		return (ENXIO);
   1339 	oif_flags = ifp->if_flags;
   1340 	switch (cmd) {
   1341 
   1342 	case SIOCGIFFLAGS:
   1343 		ifr->ifr_flags = ifp->if_flags;
   1344 		break;
   1345 
   1346 	case SIOCGIFMETRIC:
   1347 		ifr->ifr_metric = ifp->if_metric;
   1348 		break;
   1349 
   1350 	case SIOCGIFMTU:
   1351 		ifr->ifr_mtu = ifp->if_mtu;
   1352 		break;
   1353 
   1354 	case SIOCGIFDLT:
   1355 		ifr->ifr_dlt = ifp->if_dlt;
   1356 		break;
   1357 
   1358 	case SIOCSIFFLAGS:
   1359 		if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
   1360 			return (error);
   1361 		if (ifp->if_flags & IFF_UP && (ifr->ifr_flags & IFF_UP) == 0) {
   1362 			s = splnet();
   1363 			if_down(ifp);
   1364 			splx(s);
   1365 		}
   1366 		if (ifr->ifr_flags & IFF_UP && (ifp->if_flags & IFF_UP) == 0) {
   1367 			s = splnet();
   1368 			if_up(ifp);
   1369 			splx(s);
   1370 		}
   1371 		ifp->if_flags = (ifp->if_flags & IFF_CANTCHANGE) |
   1372 			(ifr->ifr_flags &~ IFF_CANTCHANGE);
   1373 		if (ifp->if_ioctl)
   1374 			(void) (*ifp->if_ioctl)(ifp, cmd, data);
   1375 		break;
   1376 
   1377 	case SIOCGIFCAP:
   1378 		ifcr->ifcr_capabilities = ifp->if_capabilities;
   1379 		ifcr->ifcr_capenable = ifp->if_capenable;
   1380 		break;
   1381 
   1382 	case SIOCSIFCAP:
   1383 		if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
   1384 			return (error);
   1385 		if ((ifcr->ifcr_capenable & ~ifp->if_capabilities) != 0)
   1386 			return (EINVAL);
   1387 		if (ifp->if_ioctl == NULL)
   1388 			return (EOPNOTSUPP);
   1389 
   1390 		/* Must prevent race with packet reception here. */
   1391 		s = splnet();
   1392 		if (ifcr->ifcr_capenable != ifp->if_capenable) {
   1393 			struct ifreq ifrq;
   1394 
   1395 			ifrq.ifr_flags = ifp->if_flags;
   1396 			ifp->if_capenable = ifcr->ifcr_capenable;
   1397 
   1398 			/* Pre-compute the checksum flags mask. */
   1399 			ifp->if_csum_flags_tx = 0;
   1400 			ifp->if_csum_flags_rx = 0;
   1401 			if (ifp->if_capenable & IFCAP_CSUM_IPv4) {
   1402 				ifp->if_csum_flags_tx |= M_CSUM_IPv4;
   1403 				ifp->if_csum_flags_rx |= M_CSUM_IPv4;
   1404 			}
   1405 
   1406 			if (ifp->if_capenable & IFCAP_CSUM_TCPv4) {
   1407 				ifp->if_csum_flags_tx |= M_CSUM_TCPv4;
   1408 				ifp->if_csum_flags_rx |= M_CSUM_TCPv4;
   1409 			} else if (ifp->if_capenable & IFCAP_CSUM_TCPv4_Rx)
   1410 				ifp->if_csum_flags_rx |= M_CSUM_TCPv4;
   1411 
   1412 			if (ifp->if_capenable & IFCAP_CSUM_UDPv4) {
   1413 				ifp->if_csum_flags_tx |= M_CSUM_UDPv4;
   1414 				ifp->if_csum_flags_rx |= M_CSUM_UDPv4;
   1415 			} else if (ifp->if_capenable & IFCAP_CSUM_UDPv4_Rx)
   1416 				ifp->if_csum_flags_rx |= M_CSUM_UDPv4;
   1417 
   1418 			if (ifp->if_capenable & IFCAP_CSUM_TCPv6) {
   1419 				ifp->if_csum_flags_tx |= M_CSUM_TCPv6;
   1420 				ifp->if_csum_flags_rx |= M_CSUM_TCPv6;
   1421 			}
   1422 
   1423 			if (ifp->if_capenable & IFCAP_CSUM_UDPv6) {
   1424 				ifp->if_csum_flags_tx |= M_CSUM_UDPv6;
   1425 				ifp->if_csum_flags_rx |= M_CSUM_UDPv6;
   1426 			}
   1427 
   1428 			/*
   1429 			 * Only kick the interface if it's up.  If it's
   1430 			 * not up now, it will notice the cap enables
   1431 			 * when it is brought up later.
   1432 			 */
   1433 			if (ifp->if_flags & IFF_UP)
   1434 				(void) (*ifp->if_ioctl)(ifp, SIOCSIFFLAGS,
   1435 				    (caddr_t) &ifrq);
   1436 		}
   1437 		splx(s);
   1438 		break;
   1439 
   1440 	case SIOCSIFMETRIC:
   1441 		if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
   1442 			return (error);
   1443 		ifp->if_metric = ifr->ifr_metric;
   1444 		break;
   1445 
   1446 	case SIOCGIFDATA:
   1447 		ifdr->ifdr_data = ifp->if_data;
   1448 		break;
   1449 
   1450 	case SIOCZIFDATA:
   1451 		if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
   1452 			return (error);
   1453 		ifdr->ifdr_data = ifp->if_data;
   1454 		/*
   1455 		 * Assumes that the volatile counters that can be
   1456 		 * zero'ed are at the end of if_data.
   1457 		 */
   1458 		memset(&ifp->if_data.ifi_ipackets, 0, sizeof(ifp->if_data) -
   1459 		    offsetof(struct if_data, ifi_ipackets));
   1460 		break;
   1461 
   1462 	case SIOCSIFMTU:
   1463 	{
   1464 		u_long oldmtu = ifp->if_mtu;
   1465 
   1466 		error = suser(p->p_ucred, &p->p_acflag);
   1467 		if (error)
   1468 			return (error);
   1469 		if (ifp->if_ioctl == NULL)
   1470 			return (EOPNOTSUPP);
   1471 		error = (*ifp->if_ioctl)(ifp, cmd, data);
   1472 
   1473 		/*
   1474 		 * If the link MTU changed, do network layer specific procedure.
   1475 		 */
   1476 		if (ifp->if_mtu != oldmtu) {
   1477 #ifdef INET6
   1478 			nd6_setmtu(ifp);
   1479 #endif
   1480 		}
   1481 		break;
   1482 	}
   1483 	case SIOCSIFPHYADDR:
   1484 	case SIOCDIFPHYADDR:
   1485 #ifdef INET6
   1486 	case SIOCSIFPHYADDR_IN6:
   1487 #endif
   1488 	case SIOCSLIFPHYADDR:
   1489 	case SIOCADDMULTI:
   1490 	case SIOCDELMULTI:
   1491 	case SIOCSIFMEDIA:
   1492 		if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
   1493 			return (error);
   1494 		/* FALLTHROUGH */
   1495 	case SIOCGIFPSRCADDR:
   1496 	case SIOCGIFPDSTADDR:
   1497 	case SIOCGLIFPHYADDR:
   1498 	case SIOCGIFMEDIA:
   1499 		if (ifp->if_ioctl == 0)
   1500 			return (EOPNOTSUPP);
   1501 		error = (*ifp->if_ioctl)(ifp, cmd, data);
   1502 		break;
   1503 
   1504 	case SIOCSDRVSPEC:
   1505 	case SIOCS80211NWID:
   1506 	case SIOCS80211NWKEY:
   1507 	case SIOCS80211POWER:
   1508 		/* XXX:  need to pass proc pointer through to driver... */
   1509 		if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
   1510 			return (error);
   1511 	/* FALLTHROUGH */
   1512 	default:
   1513 		if (so->so_proto == 0)
   1514 			return (EOPNOTSUPP);
   1515 #if !defined(COMPAT_43) && !defined(COMPAT_LINUX) && !defined(COMPAT_SVR4) && !defined(LKM)
   1516 		error = ((*so->so_proto->pr_usrreq)(so, PRU_CONTROL,
   1517 		    (struct mbuf *)cmd, (struct mbuf *)data,
   1518 		    (struct mbuf *)ifp, p));
   1519 #else
   1520 	    {
   1521 		int ocmd = cmd;
   1522 
   1523 		switch (cmd) {
   1524 
   1525 		case SIOCSIFADDR:
   1526 		case SIOCSIFDSTADDR:
   1527 		case SIOCSIFBRDADDR:
   1528 		case SIOCSIFNETMASK:
   1529 #if BYTE_ORDER != BIG_ENDIAN
   1530 			if (ifr->ifr_addr.sa_family == 0 &&
   1531 			    ifr->ifr_addr.sa_len < 16) {
   1532 				ifr->ifr_addr.sa_family = ifr->ifr_addr.sa_len;
   1533 				ifr->ifr_addr.sa_len = 16;
   1534 			}
   1535 #else
   1536 			if (ifr->ifr_addr.sa_len == 0)
   1537 				ifr->ifr_addr.sa_len = 16;
   1538 #endif
   1539 			break;
   1540 
   1541 		case OSIOCGIFADDR:
   1542 			cmd = SIOCGIFADDR;
   1543 			break;
   1544 
   1545 		case OSIOCGIFDSTADDR:
   1546 			cmd = SIOCGIFDSTADDR;
   1547 			break;
   1548 
   1549 		case OSIOCGIFBRDADDR:
   1550 			cmd = SIOCGIFBRDADDR;
   1551 			break;
   1552 
   1553 		case OSIOCGIFNETMASK:
   1554 			cmd = SIOCGIFNETMASK;
   1555 		}
   1556 
   1557 		error = ((*so->so_proto->pr_usrreq)(so, PRU_CONTROL,
   1558 		    (struct mbuf *)cmd, (struct mbuf *)data,
   1559 		    (struct mbuf *)ifp, p));
   1560 
   1561 		switch (ocmd) {
   1562 		case OSIOCGIFADDR:
   1563 		case OSIOCGIFDSTADDR:
   1564 		case OSIOCGIFBRDADDR:
   1565 		case OSIOCGIFNETMASK:
   1566 			*(u_int16_t *)&ifr->ifr_addr = ifr->ifr_addr.sa_family;
   1567 		}
   1568 	    }
   1569 #endif /* COMPAT_43 */
   1570 		break;
   1571 	}
   1572 
   1573 	if (((oif_flags ^ ifp->if_flags) & IFF_UP) != 0) {
   1574 #ifdef INET6
   1575 		if ((ifp->if_flags & IFF_UP) != 0) {
   1576 			s = splnet();
   1577 			in6_if_up(ifp);
   1578 			splx(s);
   1579 		}
   1580 #endif
   1581 	}
   1582 
   1583 	return (error);
   1584 }
   1585 
   1586 /*
   1587  * Return interface configuration
   1588  * of system.  List may be used
   1589  * in later ioctl's (above) to get
   1590  * other information.
   1591  */
   1592 /*ARGSUSED*/
   1593 int
   1594 ifconf(cmd, data)
   1595 	u_long cmd;
   1596 	caddr_t data;
   1597 {
   1598 	struct ifconf *ifc = (struct ifconf *)data;
   1599 	struct ifnet *ifp;
   1600 	struct ifaddr *ifa;
   1601 	struct ifreq ifr, *ifrp;
   1602 	int space = ifc->ifc_len, error = 0;
   1603 
   1604 	ifrp = ifc->ifc_req;
   1605 	TAILQ_FOREACH(ifp, &ifnet, if_list) {
   1606 		bcopy(ifp->if_xname, ifr.ifr_name, IFNAMSIZ);
   1607 		if ((ifa = TAILQ_FIRST(&ifp->if_addrlist)) == 0) {
   1608 			memset((caddr_t)&ifr.ifr_addr, 0, sizeof(ifr.ifr_addr));
   1609 			if (space >= (int)sizeof (ifr)) {
   1610 				error = copyout((caddr_t)&ifr, (caddr_t)ifrp,
   1611 						sizeof(ifr));
   1612 				if (error)
   1613 					break;
   1614 			}
   1615 			space -= sizeof (ifr), ifrp++;
   1616 		} else
   1617 		    for (; ifa != 0; ifa = TAILQ_NEXT(ifa, ifa_list)) {
   1618 			struct sockaddr *sa = ifa->ifa_addr;
   1619 #if defined(COMPAT_43) || defined(COMPAT_LINUX) || defined(COMPAT_SVR4)
   1620 			if (cmd == OSIOCGIFCONF) {
   1621 				struct osockaddr *osa =
   1622 					 (struct osockaddr *)&ifr.ifr_addr;
   1623 				ifr.ifr_addr = *sa;
   1624 				osa->sa_family = sa->sa_family;
   1625 				if (space >= (int)sizeof (ifr)) {
   1626 					error = copyout((caddr_t)&ifr,
   1627 							(caddr_t)ifrp,
   1628 							sizeof (ifr));
   1629 					ifrp++;
   1630 				}
   1631 			} else
   1632 #endif
   1633 			if (sa->sa_len <= sizeof(*sa)) {
   1634 				ifr.ifr_addr = *sa;
   1635 				if (space >= (int)sizeof (ifr)) {
   1636 					error = copyout((caddr_t)&ifr,
   1637 							(caddr_t)ifrp,
   1638 							sizeof (ifr));
   1639 					ifrp++;
   1640 				}
   1641 			} else {
   1642 				space -= sa->sa_len - sizeof(*sa);
   1643 				if (space >= (int)sizeof (ifr)) {
   1644 					error = copyout((caddr_t)&ifr,
   1645 							(caddr_t)ifrp,
   1646 							sizeof (ifr.ifr_name));
   1647 					if (error == 0) {
   1648 						error = copyout((caddr_t)sa,
   1649 						  (caddr_t)&ifrp->ifr_addr,
   1650 						  sa->sa_len);
   1651 					}
   1652 					ifrp = (struct ifreq *)
   1653 						(sa->sa_len +
   1654 						 (caddr_t)&ifrp->ifr_addr);
   1655 				}
   1656 			}
   1657 			if (error)
   1658 				break;
   1659 			space -= sizeof (ifr);
   1660 		}
   1661 	}
   1662 	ifc->ifc_len -= space;
   1663 	return (error);
   1664 }
   1665