xref: /freebsd/sys/net/if_bridge.c (revision 6ff6d951ade3f3379932df7f878ef3ea272cfc59)
1 /*	$NetBSD: if_bridge.c,v 1.31 2005/06/01 19:45:34 jdc Exp $	*/
2 
3 /*
4  * Copyright 2001 Wasabi Systems, Inc.
5  * All rights reserved.
6  *
7  * Written by Jason R. Thorpe for Wasabi Systems, Inc.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  * 1. Redistributions of source code must retain the above copyright
13  *    notice, this list of conditions and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  * 3. All advertising materials mentioning features or use of this software
18  *    must display the following acknowledgement:
19  *	This product includes software developed for the NetBSD Project by
20  *	Wasabi Systems, Inc.
21  * 4. The name of Wasabi Systems, Inc. may not be used to endorse
22  *    or promote products derived from this software without specific prior
23  *    written permission.
24  *
25  * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND
26  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
27  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
28  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL WASABI SYSTEMS, INC
29  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
30  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
31  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
32  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
33  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
34  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
35  * POSSIBILITY OF SUCH DAMAGE.
36  */
37 
38 /*
39  * Copyright (c) 1999, 2000 Jason L. Wright (jason@thought.net)
40  * All rights reserved.
41  *
42  * Redistribution and use in source and binary forms, with or without
43  * modification, are permitted provided that the following conditions
44  * are met:
45  * 1. Redistributions of source code must retain the above copyright
46  *    notice, this list of conditions and the following disclaimer.
47  * 2. Redistributions in binary form must reproduce the above copyright
48  *    notice, this list of conditions and the following disclaimer in the
49  *    documentation and/or other materials provided with the distribution.
50  *
51  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
52  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
53  * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
54  * DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT,
55  * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
56  * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
57  * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
58  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
59  * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
60  * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
61  * POSSIBILITY OF SUCH DAMAGE.
62  *
63  * OpenBSD: if_bridge.c,v 1.60 2001/06/15 03:38:33 itojun Exp
64  */
65 
66 /*
67  * Network interface bridge support.
68  *
69  * TODO:
70  *
71  *	- Currently only supports Ethernet-like interfaces (Ethernet,
72  *	  802.11, VLANs on Ethernet, etc.)  Figure out a nice way
73  *	  to bridge other types of interfaces (FDDI-FDDI, and maybe
74  *	  consider heterogenous bridges).
75  */
76 
77 #include <sys/cdefs.h>
78 __FBSDID("$FreeBSD$");
79 
80 #include "opt_inet.h"
81 #include "opt_inet6.h"
82 #include "opt_carp.h"
83 
84 #include <sys/param.h>
85 #include <sys/mbuf.h>
86 #include <sys/malloc.h>
87 #include <sys/protosw.h>
88 #include <sys/systm.h>
89 #include <sys/time.h>
90 #include <sys/socket.h> /* for net/if.h */
91 #include <sys/sockio.h>
92 #include <sys/ctype.h>  /* string functions */
93 #include <sys/kernel.h>
94 #include <sys/random.h>
95 #include <sys/syslog.h>
96 #include <sys/sysctl.h>
97 #include <vm/uma.h>
98 #include <sys/module.h>
99 #include <sys/priv.h>
100 #include <sys/proc.h>
101 #include <sys/lock.h>
102 #include <sys/mutex.h>
103 
104 #include <net/bpf.h>
105 #include <net/if.h>
106 #include <net/if_clone.h>
107 #include <net/if_dl.h>
108 #include <net/if_types.h>
109 #include <net/if_var.h>
110 #include <net/pfil.h>
111 
112 #include <netinet/in.h> /* for struct arpcom */
113 #include <netinet/in_systm.h>
114 #include <netinet/in_var.h>
115 #include <netinet/ip.h>
116 #include <netinet/ip_var.h>
117 #ifdef INET6
118 #include <netinet/ip6.h>
119 #include <netinet6/ip6_var.h>
120 #endif
121 #ifdef DEV_CARP
122 #include <netinet/ip_carp.h>
123 #endif
124 #include <machine/in_cksum.h>
125 #include <netinet/if_ether.h> /* for struct arpcom */
126 #include <net/bridgestp.h>
127 #include <net/if_bridgevar.h>
128 #include <net/if_llc.h>
129 #include <net/if_vlan_var.h>
130 
131 #include <net/route.h>
132 #include <netinet/ip_fw.h>
133 #include <netinet/ip_dummynet.h>
134 
135 /*
136  * Size of the route hash table.  Must be a power of two.
137  */
138 #ifndef BRIDGE_RTHASH_SIZE
139 #define	BRIDGE_RTHASH_SIZE		1024
140 #endif
141 
142 #define	BRIDGE_RTHASH_MASK		(BRIDGE_RTHASH_SIZE - 1)
143 
144 /*
145  * Maximum number of addresses to cache.
146  */
147 #ifndef BRIDGE_RTABLE_MAX
148 #define	BRIDGE_RTABLE_MAX		100
149 #endif
150 
151 /*
152  * Timeout (in seconds) for entries learned dynamically.
153  */
154 #ifndef BRIDGE_RTABLE_TIMEOUT
155 #define	BRIDGE_RTABLE_TIMEOUT		(20 * 60)	/* same as ARP */
156 #endif
157 
158 /*
159  * Number of seconds between walks of the route list.
160  */
161 #ifndef BRIDGE_RTABLE_PRUNE_PERIOD
162 #define	BRIDGE_RTABLE_PRUNE_PERIOD	(5 * 60)
163 #endif
164 
165 /*
166  * List of capabilities to mask on the member interface.
167  */
168 #define	BRIDGE_IFCAPS_MASK		IFCAP_TXCSUM
169 
170 /*
171  * Bridge interface list entry.
172  */
173 struct bridge_iflist {
174 	LIST_ENTRY(bridge_iflist) bif_next;
175 	struct ifnet		*bif_ifp;	/* member if */
176 	struct bstp_port	bif_stp;	/* STP state */
177 	uint32_t		bif_flags;	/* member if flags */
178 	int			bif_mutecap;	/* member muted caps */
179 	uint32_t		bif_addrmax;	/* max # of addresses */
180 	uint32_t		bif_addrcnt;	/* cur. # of addresses */
181 	uint32_t		bif_addrexceeded;/* # of address violations */
182 };
183 
184 /*
185  * Bridge route node.
186  */
187 struct bridge_rtnode {
188 	LIST_ENTRY(bridge_rtnode) brt_hash;	/* hash table linkage */
189 	LIST_ENTRY(bridge_rtnode) brt_list;	/* list linkage */
190 	struct bridge_iflist	*brt_dst;	/* destination if */
191 	unsigned long		brt_expire;	/* expiration time */
192 	uint8_t			brt_flags;	/* address flags */
193 	uint8_t			brt_addr[ETHER_ADDR_LEN];
194 	uint16_t		brt_vlan;	/* vlan id */
195 };
196 #define	brt_ifp			brt_dst->bif_ifp
197 
198 /*
199  * Software state for each bridge.
200  */
201 struct bridge_softc {
202 	struct ifnet		*sc_ifp;	/* make this an interface */
203 	LIST_ENTRY(bridge_softc) sc_list;
204 	struct mtx		sc_mtx;
205 	struct cv		sc_cv;
206 	uint32_t		sc_brtmax;	/* max # of addresses */
207 	uint32_t		sc_brtcnt;	/* cur. # of addresses */
208 	uint32_t		sc_brttimeout;	/* rt timeout in seconds */
209 	struct callout		sc_brcallout;	/* bridge callout */
210 	uint32_t		sc_iflist_ref;	/* refcount for sc_iflist */
211 	uint32_t		sc_iflist_xcnt;	/* refcount for sc_iflist */
212 	LIST_HEAD(, bridge_iflist) sc_iflist;	/* member interface list */
213 	LIST_HEAD(, bridge_rtnode) *sc_rthash;	/* our forwarding table */
214 	LIST_HEAD(, bridge_rtnode) sc_rtlist;	/* list version of above */
215 	uint32_t		sc_rthash_key;	/* key for hash */
216 	LIST_HEAD(, bridge_iflist) sc_spanlist;	/* span ports list */
217 	struct bstp_state	sc_stp;		/* STP state */
218 	uint32_t		sc_brtexceeded;	/* # of cache drops */
219 };
220 
221 static struct mtx 	bridge_list_mtx;
222 eventhandler_tag	bridge_detach_cookie = NULL;
223 
224 int	bridge_rtable_prune_period = BRIDGE_RTABLE_PRUNE_PERIOD;
225 
226 uma_zone_t bridge_rtnode_zone;
227 
228 static int	bridge_clone_create(struct if_clone *, int, caddr_t);
229 static void	bridge_clone_destroy(struct ifnet *);
230 
231 static int	bridge_ioctl(struct ifnet *, u_long, caddr_t);
232 static void	bridge_mutecaps(struct bridge_iflist *, int);
233 static void	bridge_ifdetach(void *arg __unused, struct ifnet *);
234 static void	bridge_init(void *);
235 static void	bridge_dummynet(struct mbuf *, struct ifnet *);
236 static void	bridge_stop(struct ifnet *, int);
237 static void	bridge_start(struct ifnet *);
238 static struct mbuf *bridge_input(struct ifnet *, struct mbuf *);
239 static int	bridge_output(struct ifnet *, struct mbuf *, struct sockaddr *,
240 		    struct rtentry *);
241 static void	bridge_enqueue(struct bridge_softc *, struct ifnet *,
242 		    struct mbuf *);
243 static void	bridge_rtdelete(struct bridge_softc *, struct ifnet *ifp, int);
244 
245 static void	bridge_forward(struct bridge_softc *, struct bridge_iflist *,
246 		    struct mbuf *m);
247 
248 static void	bridge_timer(void *);
249 
250 static void	bridge_broadcast(struct bridge_softc *, struct ifnet *,
251 		    struct mbuf *, int);
252 static void	bridge_span(struct bridge_softc *, struct mbuf *);
253 
254 static int	bridge_rtupdate(struct bridge_softc *, const uint8_t *,
255 		    uint16_t, struct bridge_iflist *, int, uint8_t);
256 static struct ifnet *bridge_rtlookup(struct bridge_softc *, const uint8_t *,
257 		    uint16_t);
258 static void	bridge_rttrim(struct bridge_softc *);
259 static void	bridge_rtage(struct bridge_softc *);
260 static void	bridge_rtflush(struct bridge_softc *, int);
261 static int	bridge_rtdaddr(struct bridge_softc *, const uint8_t *,
262 		    uint16_t);
263 
264 static int	bridge_rtable_init(struct bridge_softc *);
265 static void	bridge_rtable_fini(struct bridge_softc *);
266 
267 static int	bridge_rtnode_addr_cmp(const uint8_t *, const uint8_t *);
268 static struct bridge_rtnode *bridge_rtnode_lookup(struct bridge_softc *,
269 		    const uint8_t *, uint16_t);
270 static int	bridge_rtnode_insert(struct bridge_softc *,
271 		    struct bridge_rtnode *);
272 static void	bridge_rtnode_destroy(struct bridge_softc *,
273 		    struct bridge_rtnode *);
274 static void	bridge_rtable_expire(struct ifnet *, int);
275 static void	bridge_state_change(struct ifnet *, int);
276 
277 static struct bridge_iflist *bridge_lookup_member(struct bridge_softc *,
278 		    const char *name);
279 static struct bridge_iflist *bridge_lookup_member_if(struct bridge_softc *,
280 		    struct ifnet *ifp);
281 static void	bridge_delete_member(struct bridge_softc *,
282 		    struct bridge_iflist *, int);
283 static void	bridge_delete_span(struct bridge_softc *,
284 		    struct bridge_iflist *);
285 
286 static int	bridge_ioctl_add(struct bridge_softc *, void *);
287 static int	bridge_ioctl_del(struct bridge_softc *, void *);
288 static int	bridge_ioctl_gifflags(struct bridge_softc *, void *);
289 static int	bridge_ioctl_sifflags(struct bridge_softc *, void *);
290 static int	bridge_ioctl_scache(struct bridge_softc *, void *);
291 static int	bridge_ioctl_gcache(struct bridge_softc *, void *);
292 static int	bridge_ioctl_gifs(struct bridge_softc *, void *);
293 static int	bridge_ioctl_rts(struct bridge_softc *, void *);
294 static int	bridge_ioctl_saddr(struct bridge_softc *, void *);
295 static int	bridge_ioctl_sto(struct bridge_softc *, void *);
296 static int	bridge_ioctl_gto(struct bridge_softc *, void *);
297 static int	bridge_ioctl_daddr(struct bridge_softc *, void *);
298 static int	bridge_ioctl_flush(struct bridge_softc *, void *);
299 static int	bridge_ioctl_gpri(struct bridge_softc *, void *);
300 static int	bridge_ioctl_spri(struct bridge_softc *, void *);
301 static int	bridge_ioctl_ght(struct bridge_softc *, void *);
302 static int	bridge_ioctl_sht(struct bridge_softc *, void *);
303 static int	bridge_ioctl_gfd(struct bridge_softc *, void *);
304 static int	bridge_ioctl_sfd(struct bridge_softc *, void *);
305 static int	bridge_ioctl_gma(struct bridge_softc *, void *);
306 static int	bridge_ioctl_sma(struct bridge_softc *, void *);
307 static int	bridge_ioctl_sifprio(struct bridge_softc *, void *);
308 static int	bridge_ioctl_sifcost(struct bridge_softc *, void *);
309 static int	bridge_ioctl_sifmaxaddr(struct bridge_softc *, void *);
310 static int	bridge_ioctl_addspan(struct bridge_softc *, void *);
311 static int	bridge_ioctl_delspan(struct bridge_softc *, void *);
312 static int	bridge_ioctl_gbparam(struct bridge_softc *, void *);
313 static int	bridge_ioctl_grte(struct bridge_softc *, void *);
314 static int	bridge_ioctl_gifsstp(struct bridge_softc *, void *);
315 static int	bridge_ioctl_sproto(struct bridge_softc *, void *);
316 static int	bridge_ioctl_stxhc(struct bridge_softc *, void *);
317 static int	bridge_pfil(struct mbuf **, struct ifnet *, struct ifnet *,
318 		    int);
319 static int	bridge_ip_checkbasic(struct mbuf **mp);
320 #ifdef INET6
321 static int	bridge_ip6_checkbasic(struct mbuf **mp);
322 #endif /* INET6 */
323 static int	bridge_fragment(struct ifnet *, struct mbuf *,
324 		    struct ether_header *, int, struct llc *);
325 
326 /* The default bridge vlan is 1 (IEEE 802.1Q-2003 Table 9-2) */
327 #define	VLANTAGOF(_m)	\
328     (_m->m_flags & M_VLANTAG) ? EVL_VLANOFTAG(_m->m_pkthdr.ether_vtag) : 1
329 
330 static struct bstp_cb_ops bridge_ops = {
331 	.bcb_state = bridge_state_change,
332 	.bcb_rtage = bridge_rtable_expire
333 };
334 
335 SYSCTL_DECL(_net_link);
336 SYSCTL_NODE(_net_link, IFT_BRIDGE, bridge, CTLFLAG_RW, 0, "Bridge");
337 
338 static int pfil_onlyip = 1; /* only pass IP[46] packets when pfil is enabled */
339 static int pfil_bridge = 1; /* run pfil hooks on the bridge interface */
340 static int pfil_member = 1; /* run pfil hooks on the member interface */
341 static int pfil_ipfw = 0;   /* layer2 filter with ipfw */
342 static int pfil_ipfw_arp = 0;   /* layer2 filter with ipfw */
343 static int pfil_local_phys = 0; /* run pfil hooks on the physical interface for
344                                    locally destined packets */
345 static int log_stp   = 0;   /* log STP state changes */
346 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_onlyip, CTLFLAG_RW,
347     &pfil_onlyip, 0, "Only pass IP packets when pfil is enabled");
348 SYSCTL_INT(_net_link_bridge, OID_AUTO, ipfw_arp, CTLFLAG_RW,
349     &pfil_ipfw_arp, 0, "Filter ARP packets through IPFW layer2");
350 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_bridge, CTLFLAG_RW,
351     &pfil_bridge, 0, "Packet filter on the bridge interface");
352 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_member, CTLFLAG_RW,
353     &pfil_member, 0, "Packet filter on the member interface");
354 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_local_phys, CTLFLAG_RW,
355     &pfil_local_phys, 0,
356     "Packet filter on the physical interface for locally destined packets");
357 SYSCTL_INT(_net_link_bridge, OID_AUTO, log_stp, CTLFLAG_RW,
358     &log_stp, 0, "Log STP state changes");
359 
360 struct bridge_control {
361 	int	(*bc_func)(struct bridge_softc *, void *);
362 	int	bc_argsize;
363 	int	bc_flags;
364 };
365 
366 #define	BC_F_COPYIN		0x01	/* copy arguments in */
367 #define	BC_F_COPYOUT		0x02	/* copy arguments out */
368 #define	BC_F_SUSER		0x04	/* do super-user check */
369 
370 const struct bridge_control bridge_control_table[] = {
371 	{ bridge_ioctl_add,		sizeof(struct ifbreq),
372 	  BC_F_COPYIN|BC_F_SUSER },
373 	{ bridge_ioctl_del,		sizeof(struct ifbreq),
374 	  BC_F_COPYIN|BC_F_SUSER },
375 
376 	{ bridge_ioctl_gifflags,	sizeof(struct ifbreq),
377 	  BC_F_COPYIN|BC_F_COPYOUT },
378 	{ bridge_ioctl_sifflags,	sizeof(struct ifbreq),
379 	  BC_F_COPYIN|BC_F_SUSER },
380 
381 	{ bridge_ioctl_scache,		sizeof(struct ifbrparam),
382 	  BC_F_COPYIN|BC_F_SUSER },
383 	{ bridge_ioctl_gcache,		sizeof(struct ifbrparam),
384 	  BC_F_COPYOUT },
385 
386 	{ bridge_ioctl_gifs,		sizeof(struct ifbifconf),
387 	  BC_F_COPYIN|BC_F_COPYOUT },
388 	{ bridge_ioctl_rts,		sizeof(struct ifbaconf),
389 	  BC_F_COPYIN|BC_F_COPYOUT },
390 
391 	{ bridge_ioctl_saddr,		sizeof(struct ifbareq),
392 	  BC_F_COPYIN|BC_F_SUSER },
393 
394 	{ bridge_ioctl_sto,		sizeof(struct ifbrparam),
395 	  BC_F_COPYIN|BC_F_SUSER },
396 	{ bridge_ioctl_gto,		sizeof(struct ifbrparam),
397 	  BC_F_COPYOUT },
398 
399 	{ bridge_ioctl_daddr,		sizeof(struct ifbareq),
400 	  BC_F_COPYIN|BC_F_SUSER },
401 
402 	{ bridge_ioctl_flush,		sizeof(struct ifbreq),
403 	  BC_F_COPYIN|BC_F_SUSER },
404 
405 	{ bridge_ioctl_gpri,		sizeof(struct ifbrparam),
406 	  BC_F_COPYOUT },
407 	{ bridge_ioctl_spri,		sizeof(struct ifbrparam),
408 	  BC_F_COPYIN|BC_F_SUSER },
409 
410 	{ bridge_ioctl_ght,		sizeof(struct ifbrparam),
411 	  BC_F_COPYOUT },
412 	{ bridge_ioctl_sht,		sizeof(struct ifbrparam),
413 	  BC_F_COPYIN|BC_F_SUSER },
414 
415 	{ bridge_ioctl_gfd,		sizeof(struct ifbrparam),
416 	  BC_F_COPYOUT },
417 	{ bridge_ioctl_sfd,		sizeof(struct ifbrparam),
418 	  BC_F_COPYIN|BC_F_SUSER },
419 
420 	{ bridge_ioctl_gma,		sizeof(struct ifbrparam),
421 	  BC_F_COPYOUT },
422 	{ bridge_ioctl_sma,		sizeof(struct ifbrparam),
423 	  BC_F_COPYIN|BC_F_SUSER },
424 
425 	{ bridge_ioctl_sifprio,		sizeof(struct ifbreq),
426 	  BC_F_COPYIN|BC_F_SUSER },
427 
428 	{ bridge_ioctl_sifcost,		sizeof(struct ifbreq),
429 	  BC_F_COPYIN|BC_F_SUSER },
430 
431 	{ bridge_ioctl_addspan,		sizeof(struct ifbreq),
432 	  BC_F_COPYIN|BC_F_SUSER },
433 	{ bridge_ioctl_delspan,		sizeof(struct ifbreq),
434 	  BC_F_COPYIN|BC_F_SUSER },
435 
436 	{ bridge_ioctl_gbparam,		sizeof(struct ifbropreq),
437 	  BC_F_COPYOUT },
438 
439 	{ bridge_ioctl_grte,		sizeof(struct ifbrparam),
440 	  BC_F_COPYOUT },
441 
442 	{ bridge_ioctl_gifsstp,		sizeof(struct ifbpstpconf),
443 	  BC_F_COPYIN|BC_F_COPYOUT },
444 
445 	{ bridge_ioctl_sproto,		sizeof(struct ifbrparam),
446 	  BC_F_COPYIN|BC_F_SUSER },
447 
448 	{ bridge_ioctl_stxhc,		sizeof(struct ifbrparam),
449 	  BC_F_COPYIN|BC_F_SUSER },
450 
451 	{ bridge_ioctl_sifmaxaddr,	sizeof(struct ifbreq),
452 	  BC_F_COPYIN|BC_F_SUSER },
453 
454 };
455 const int bridge_control_table_size =
456     sizeof(bridge_control_table) / sizeof(bridge_control_table[0]);
457 
458 LIST_HEAD(, bridge_softc) bridge_list;
459 
460 IFC_SIMPLE_DECLARE(bridge, 0);
461 
462 static int
463 bridge_modevent(module_t mod, int type, void *data)
464 {
465 
466 	switch (type) {
467 	case MOD_LOAD:
468 		mtx_init(&bridge_list_mtx, "if_bridge list", NULL, MTX_DEF);
469 		if_clone_attach(&bridge_cloner);
470 		bridge_rtnode_zone = uma_zcreate("bridge_rtnode",
471 		    sizeof(struct bridge_rtnode), NULL, NULL, NULL, NULL,
472 		    UMA_ALIGN_PTR, 0);
473 		LIST_INIT(&bridge_list);
474 		bridge_input_p = bridge_input;
475 		bridge_output_p = bridge_output;
476 		bridge_dn_p = bridge_dummynet;
477 		bridge_detach_cookie = EVENTHANDLER_REGISTER(
478 		    ifnet_departure_event, bridge_ifdetach, NULL,
479 		    EVENTHANDLER_PRI_ANY);
480 		break;
481 	case MOD_UNLOAD:
482 		EVENTHANDLER_DEREGISTER(ifnet_departure_event,
483 		    bridge_detach_cookie);
484 		if_clone_detach(&bridge_cloner);
485 		uma_zdestroy(bridge_rtnode_zone);
486 		bridge_input_p = NULL;
487 		bridge_output_p = NULL;
488 		bridge_dn_p = NULL;
489 		mtx_destroy(&bridge_list_mtx);
490 		break;
491 	default:
492 		return (EOPNOTSUPP);
493 	}
494 	return (0);
495 }
496 
497 static moduledata_t bridge_mod = {
498 	"if_bridge",
499 	bridge_modevent,
500 	0
501 };
502 
503 DECLARE_MODULE(if_bridge, bridge_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
504 MODULE_DEPEND(if_bridge, bridgestp, 1, 1, 1);
505 
506 /*
507  * handler for net.link.bridge.pfil_ipfw
508  */
509 static int
510 sysctl_pfil_ipfw(SYSCTL_HANDLER_ARGS)
511 {
512 	int enable = pfil_ipfw;
513 	int error;
514 
515 	error = sysctl_handle_int(oidp, &enable, 0, req);
516 	enable = (enable) ? 1 : 0;
517 
518 	if (enable != pfil_ipfw) {
519 		pfil_ipfw = enable;
520 
521 		/*
522 		 * Disable pfil so that ipfw doesnt run twice, if the user
523 		 * really wants both then they can re-enable pfil_bridge and/or
524 		 * pfil_member. Also allow non-ip packets as ipfw can filter by
525 		 * layer2 type.
526 		 */
527 		if (pfil_ipfw) {
528 			pfil_onlyip = 0;
529 			pfil_bridge = 0;
530 			pfil_member = 0;
531 		}
532 	}
533 
534 	return (error);
535 }
536 SYSCTL_PROC(_net_link_bridge, OID_AUTO, ipfw, CTLTYPE_INT|CTLFLAG_RW,
537 	    &pfil_ipfw, 0, &sysctl_pfil_ipfw, "I", "Layer2 filter with IPFW");
538 
539 /*
540  * bridge_clone_create:
541  *
542  *	Create a new bridge instance.
543  */
544 static int
545 bridge_clone_create(struct if_clone *ifc, int unit, caddr_t params)
546 {
547 	struct bridge_softc *sc, *sc2;
548 	struct ifnet *bifp, *ifp;
549 	u_char eaddr[6];
550 	int retry;
551 
552 	sc = malloc(sizeof(*sc), M_DEVBUF, M_WAITOK|M_ZERO);
553 	ifp = sc->sc_ifp = if_alloc(IFT_ETHER);
554 	if (ifp == NULL) {
555 		free(sc, M_DEVBUF);
556 		return (ENOSPC);
557 	}
558 
559 	BRIDGE_LOCK_INIT(sc);
560 	sc->sc_brtmax = BRIDGE_RTABLE_MAX;
561 	sc->sc_brttimeout = BRIDGE_RTABLE_TIMEOUT;
562 
563 	/* Initialize our routing table. */
564 	bridge_rtable_init(sc);
565 
566 	callout_init_mtx(&sc->sc_brcallout, &sc->sc_mtx, 0);
567 
568 	LIST_INIT(&sc->sc_iflist);
569 	LIST_INIT(&sc->sc_spanlist);
570 
571 	ifp->if_softc = sc;
572 	if_initname(ifp, ifc->ifc_name, unit);
573 	ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
574 	ifp->if_ioctl = bridge_ioctl;
575 	ifp->if_start = bridge_start;
576 	ifp->if_init = bridge_init;
577 	ifp->if_type = IFT_BRIDGE;
578 	IFQ_SET_MAXLEN(&ifp->if_snd, ifqmaxlen);
579 	ifp->if_snd.ifq_drv_maxlen = ifqmaxlen;
580 	IFQ_SET_READY(&ifp->if_snd);
581 
582 	/*
583 	 * Generate a random ethernet address with a locally administered
584 	 * address.
585 	 *
586 	 * Since we are using random ethernet addresses for the bridge, it is
587 	 * possible that we might have address collisions, so make sure that
588 	 * this hardware address isn't already in use on another bridge.
589 	 */
590 	for (retry = 1; retry != 0;) {
591 		arc4rand(eaddr, ETHER_ADDR_LEN, 1);
592 		eaddr[0] &= ~1;		/* clear multicast bit */
593 		eaddr[0] |= 2;		/* set the LAA bit */
594 		retry = 0;
595 		mtx_lock(&bridge_list_mtx);
596 		LIST_FOREACH(sc2, &bridge_list, sc_list) {
597 			bifp = sc2->sc_ifp;
598 			if (memcmp(eaddr, IF_LLADDR(bifp), ETHER_ADDR_LEN) == 0)
599 				retry = 1;
600 		}
601 		mtx_unlock(&bridge_list_mtx);
602 	}
603 
604 	bstp_attach(&sc->sc_stp, &bridge_ops);
605 	ether_ifattach(ifp, eaddr);
606 	/* Now undo some of the damage... */
607 	ifp->if_baudrate = 0;
608 	ifp->if_type = IFT_BRIDGE;
609 
610 	mtx_lock(&bridge_list_mtx);
611 	LIST_INSERT_HEAD(&bridge_list, sc, sc_list);
612 	mtx_unlock(&bridge_list_mtx);
613 
614 	return (0);
615 }
616 
617 /*
618  * bridge_clone_destroy:
619  *
620  *	Destroy a bridge instance.
621  */
622 static void
623 bridge_clone_destroy(struct ifnet *ifp)
624 {
625 	struct bridge_softc *sc = ifp->if_softc;
626 	struct bridge_iflist *bif;
627 
628 	BRIDGE_LOCK(sc);
629 
630 	bridge_stop(ifp, 1);
631 	ifp->if_flags &= ~IFF_UP;
632 
633 	while ((bif = LIST_FIRST(&sc->sc_iflist)) != NULL)
634 		bridge_delete_member(sc, bif, 0);
635 
636 	while ((bif = LIST_FIRST(&sc->sc_spanlist)) != NULL) {
637 		bridge_delete_span(sc, bif);
638 	}
639 
640 	BRIDGE_UNLOCK(sc);
641 
642 	callout_drain(&sc->sc_brcallout);
643 
644 	mtx_lock(&bridge_list_mtx);
645 	LIST_REMOVE(sc, sc_list);
646 	mtx_unlock(&bridge_list_mtx);
647 
648 	bstp_detach(&sc->sc_stp);
649 	ether_ifdetach(ifp);
650 	if_free_type(ifp, IFT_ETHER);
651 
652 	/* Tear down the routing table. */
653 	bridge_rtable_fini(sc);
654 
655 	BRIDGE_LOCK_DESTROY(sc);
656 	free(sc, M_DEVBUF);
657 }
658 
659 /*
660  * bridge_ioctl:
661  *
662  *	Handle a control request from the operator.
663  */
664 static int
665 bridge_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data)
666 {
667 	struct bridge_softc *sc = ifp->if_softc;
668 	struct thread *td = curthread;
669 	union {
670 		struct ifbreq ifbreq;
671 		struct ifbifconf ifbifconf;
672 		struct ifbareq ifbareq;
673 		struct ifbaconf ifbaconf;
674 		struct ifbrparam ifbrparam;
675 		struct ifbropreq ifbropreq;
676 	} args;
677 	struct ifdrv *ifd = (struct ifdrv *) data;
678 	const struct bridge_control *bc;
679 	int error = 0;
680 
681 	switch (cmd) {
682 
683 	case SIOCADDMULTI:
684 	case SIOCDELMULTI:
685 		break;
686 
687 	case SIOCGDRVSPEC:
688 	case SIOCSDRVSPEC:
689 		if (ifd->ifd_cmd >= bridge_control_table_size) {
690 			error = EINVAL;
691 			break;
692 		}
693 		bc = &bridge_control_table[ifd->ifd_cmd];
694 
695 		if (cmd == SIOCGDRVSPEC &&
696 		    (bc->bc_flags & BC_F_COPYOUT) == 0) {
697 			error = EINVAL;
698 			break;
699 		}
700 		else if (cmd == SIOCSDRVSPEC &&
701 		    (bc->bc_flags & BC_F_COPYOUT) != 0) {
702 			error = EINVAL;
703 			break;
704 		}
705 
706 		if (bc->bc_flags & BC_F_SUSER) {
707 			error = priv_check(td, PRIV_NET_BRIDGE);
708 			if (error)
709 				break;
710 		}
711 
712 		if (ifd->ifd_len != bc->bc_argsize ||
713 		    ifd->ifd_len > sizeof(args)) {
714 			error = EINVAL;
715 			break;
716 		}
717 
718 		bzero(&args, sizeof(args));
719 		if (bc->bc_flags & BC_F_COPYIN) {
720 			error = copyin(ifd->ifd_data, &args, ifd->ifd_len);
721 			if (error)
722 				break;
723 		}
724 
725 		BRIDGE_LOCK(sc);
726 		error = (*bc->bc_func)(sc, &args);
727 		BRIDGE_UNLOCK(sc);
728 		if (error)
729 			break;
730 
731 		if (bc->bc_flags & BC_F_COPYOUT)
732 			error = copyout(&args, ifd->ifd_data, ifd->ifd_len);
733 
734 		break;
735 
736 	case SIOCSIFFLAGS:
737 		if (!(ifp->if_flags & IFF_UP) &&
738 		    (ifp->if_drv_flags & IFF_DRV_RUNNING)) {
739 			/*
740 			 * If interface is marked down and it is running,
741 			 * then stop and disable it.
742 			 */
743 			BRIDGE_LOCK(sc);
744 			bridge_stop(ifp, 1);
745 			BRIDGE_UNLOCK(sc);
746 		} else if ((ifp->if_flags & IFF_UP) &&
747 		    !(ifp->if_drv_flags & IFF_DRV_RUNNING)) {
748 			/*
749 			 * If interface is marked up and it is stopped, then
750 			 * start it.
751 			 */
752 			(*ifp->if_init)(sc);
753 		}
754 		break;
755 
756 	case SIOCSIFMTU:
757 		/* Do not allow the MTU to be changed on the bridge */
758 		error = EINVAL;
759 		break;
760 
761 	default:
762 		/*
763 		 * drop the lock as ether_ioctl() will call bridge_start() and
764 		 * cause the lock to be recursed.
765 		 */
766 		error = ether_ioctl(ifp, cmd, data);
767 		break;
768 	}
769 
770 	return (error);
771 }
772 
773 /*
774  * bridge_mutecaps:
775  *
776  *	Clear or restore unwanted capabilities on the member interface
777  */
778 static void
779 bridge_mutecaps(struct bridge_iflist *bif, int mute)
780 {
781 	struct ifnet *ifp = bif->bif_ifp;
782 	struct ifreq ifr;
783 	int error;
784 
785 	if (ifp->if_ioctl == NULL)
786 		return;
787 
788 	bzero(&ifr, sizeof(ifr));
789 	ifr.ifr_reqcap = ifp->if_capenable;
790 
791 	if (mute) {
792 		/* mask off and save capabilities */
793 		bif->bif_mutecap = ifr.ifr_reqcap & BRIDGE_IFCAPS_MASK;
794 		if (bif->bif_mutecap != 0)
795 			ifr.ifr_reqcap &= ~BRIDGE_IFCAPS_MASK;
796 	} else
797 		/* restore muted capabilities */
798 		ifr.ifr_reqcap |= bif->bif_mutecap;
799 
800 
801 	if (bif->bif_mutecap != 0) {
802 		IFF_LOCKGIANT(ifp);
803 		error = (*ifp->if_ioctl)(ifp, SIOCSIFCAP, (caddr_t)&ifr);
804 		IFF_UNLOCKGIANT(ifp);
805 	}
806 }
807 
808 /*
809  * bridge_lookup_member:
810  *
811  *	Lookup a bridge member interface.
812  */
813 static struct bridge_iflist *
814 bridge_lookup_member(struct bridge_softc *sc, const char *name)
815 {
816 	struct bridge_iflist *bif;
817 	struct ifnet *ifp;
818 
819 	BRIDGE_LOCK_ASSERT(sc);
820 
821 	LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
822 		ifp = bif->bif_ifp;
823 		if (strcmp(ifp->if_xname, name) == 0)
824 			return (bif);
825 	}
826 
827 	return (NULL);
828 }
829 
830 /*
831  * bridge_lookup_member_if:
832  *
833  *	Lookup a bridge member interface by ifnet*.
834  */
835 static struct bridge_iflist *
836 bridge_lookup_member_if(struct bridge_softc *sc, struct ifnet *member_ifp)
837 {
838 	struct bridge_iflist *bif;
839 
840 	BRIDGE_LOCK_ASSERT(sc);
841 
842 	LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
843 		if (bif->bif_ifp == member_ifp)
844 			return (bif);
845 	}
846 
847 	return (NULL);
848 }
849 
850 /*
851  * bridge_delete_member:
852  *
853  *	Delete the specified member interface.
854  */
855 static void
856 bridge_delete_member(struct bridge_softc *sc, struct bridge_iflist *bif,
857     int gone)
858 {
859 	struct ifnet *ifs = bif->bif_ifp;
860 
861 	BRIDGE_LOCK_ASSERT(sc);
862 
863 	if (!gone) {
864 		switch (ifs->if_type) {
865 		case IFT_ETHER:
866 		case IFT_L2VLAN:
867 			/*
868 			 * Take the interface out of promiscuous mode.
869 			 */
870 			(void) ifpromisc(ifs, 0);
871 			bridge_mutecaps(bif, 0);
872 			break;
873 
874 		case IFT_GIF:
875 			break;
876 
877 		default:
878 #ifdef DIAGNOSTIC
879 			panic("bridge_delete_member: impossible");
880 #endif
881 			break;
882 		}
883 	}
884 
885 	if (bif->bif_flags & IFBIF_STP)
886 		bstp_disable(&bif->bif_stp);
887 
888 	ifs->if_bridge = NULL;
889 	BRIDGE_XLOCK(sc);
890 	LIST_REMOVE(bif, bif_next);
891 	BRIDGE_XDROP(sc);
892 
893 	bridge_rtdelete(sc, ifs, IFBF_FLUSHALL);
894 	KASSERT(bif->bif_addrcnt == 0,
895 	    ("%s: %d bridge routes referenced", __func__, bif->bif_addrcnt));
896 
897 	BRIDGE_UNLOCK(sc);
898 	bstp_destroy(&bif->bif_stp);	/* prepare to free */
899 	BRIDGE_LOCK(sc);
900 	free(bif, M_DEVBUF);
901 }
902 
903 /*
904  * bridge_delete_span:
905  *
906  *	Delete the specified span interface.
907  */
908 static void
909 bridge_delete_span(struct bridge_softc *sc, struct bridge_iflist *bif)
910 {
911 	BRIDGE_LOCK_ASSERT(sc);
912 
913 	KASSERT(bif->bif_ifp->if_bridge == NULL,
914 	    ("%s: not a span interface", __func__));
915 
916 	LIST_REMOVE(bif, bif_next);
917 	free(bif, M_DEVBUF);
918 }
919 
920 static int
921 bridge_ioctl_add(struct bridge_softc *sc, void *arg)
922 {
923 	struct ifbreq *req = arg;
924 	struct bridge_iflist *bif = NULL;
925 	struct ifnet *ifs;
926 	int error = 0;
927 
928 	ifs = ifunit(req->ifbr_ifsname);
929 	if (ifs == NULL)
930 		return (ENOENT);
931 
932 	/* If it's in the span list, it can't be a member. */
933 	LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
934 		if (ifs == bif->bif_ifp)
935 			return (EBUSY);
936 
937 	/* Allow the first Ethernet member to define the MTU */
938 	if (ifs->if_type != IFT_GIF) {
939 		if (LIST_EMPTY(&sc->sc_iflist))
940 			sc->sc_ifp->if_mtu = ifs->if_mtu;
941 		else if (sc->sc_ifp->if_mtu != ifs->if_mtu) {
942 			if_printf(sc->sc_ifp, "invalid MTU for %s\n",
943 			    ifs->if_xname);
944 			return (EINVAL);
945 		}
946 	}
947 
948 	if (ifs->if_bridge == sc)
949 		return (EEXIST);
950 
951 	if (ifs->if_bridge != NULL)
952 		return (EBUSY);
953 
954 	bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO);
955 	if (bif == NULL)
956 		return (ENOMEM);
957 
958 	bif->bif_ifp = ifs;
959 	bif->bif_flags = IFBIF_LEARNING | IFBIF_DISCOVER;
960 
961 	switch (ifs->if_type) {
962 	case IFT_ETHER:
963 	case IFT_L2VLAN:
964 		/*
965 		 * Place the interface into promiscuous mode.
966 		 */
967 		error = ifpromisc(ifs, 1);
968 		if (error)
969 			goto out;
970 
971 		bridge_mutecaps(bif, 1);
972 		break;
973 
974 	case IFT_GIF:
975 		break;
976 
977 	default:
978 		error = EINVAL;
979 		goto out;
980 	}
981 
982 	ifs->if_bridge = sc;
983 	bstp_create(&sc->sc_stp, &bif->bif_stp, bif->bif_ifp);
984 	/*
985 	 * XXX: XLOCK HERE!?!
986 	 *
987 	 * NOTE: insert_***HEAD*** should be safe for the traversals.
988 	 */
989 	LIST_INSERT_HEAD(&sc->sc_iflist, bif, bif_next);
990 
991 out:
992 	if (error) {
993 		if (bif != NULL)
994 			free(bif, M_DEVBUF);
995 	}
996 	return (error);
997 }
998 
999 static int
1000 bridge_ioctl_del(struct bridge_softc *sc, void *arg)
1001 {
1002 	struct ifbreq *req = arg;
1003 	struct bridge_iflist *bif;
1004 
1005 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1006 	if (bif == NULL)
1007 		return (ENOENT);
1008 
1009 	bridge_delete_member(sc, bif, 0);
1010 
1011 	return (0);
1012 }
1013 
1014 static int
1015 bridge_ioctl_gifflags(struct bridge_softc *sc, void *arg)
1016 {
1017 	struct ifbreq *req = arg;
1018 	struct bridge_iflist *bif;
1019 	struct bstp_port *bp;
1020 
1021 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1022 	if (bif == NULL)
1023 		return (ENOENT);
1024 
1025 	bp = &bif->bif_stp;
1026 	req->ifbr_ifsflags = bif->bif_flags;
1027 	req->ifbr_state = bp->bp_state;
1028 	req->ifbr_priority = bp->bp_priority;
1029 	req->ifbr_path_cost = bp->bp_path_cost;
1030 	req->ifbr_portno = bif->bif_ifp->if_index & 0xfff;
1031 	req->ifbr_proto = bp->bp_protover;
1032 	req->ifbr_role = bp->bp_role;
1033 	req->ifbr_stpflags = bp->bp_flags;
1034 	req->ifbr_addrcnt = bif->bif_addrcnt;
1035 	req->ifbr_addrmax = bif->bif_addrmax;
1036 	req->ifbr_addrexceeded = bif->bif_addrexceeded;
1037 
1038 	/* Copy STP state options as flags */
1039 	if (bp->bp_operedge)
1040 		req->ifbr_ifsflags |= IFBIF_BSTP_EDGE;
1041 	if (bp->bp_flags & BSTP_PORT_AUTOEDGE)
1042 		req->ifbr_ifsflags |= IFBIF_BSTP_AUTOEDGE;
1043 	if (bp->bp_ptp_link)
1044 		req->ifbr_ifsflags |= IFBIF_BSTP_PTP;
1045 	if (bp->bp_flags & BSTP_PORT_AUTOPTP)
1046 		req->ifbr_ifsflags |= IFBIF_BSTP_AUTOPTP;
1047 	if (bp->bp_flags & BSTP_PORT_ADMEDGE)
1048 		req->ifbr_ifsflags |= IFBIF_BSTP_ADMEDGE;
1049 	if (bp->bp_flags & BSTP_PORT_ADMCOST)
1050 		req->ifbr_ifsflags |= IFBIF_BSTP_ADMCOST;
1051 	return (0);
1052 }
1053 
1054 static int
1055 bridge_ioctl_sifflags(struct bridge_softc *sc, void *arg)
1056 {
1057 	struct ifbreq *req = arg;
1058 	struct bridge_iflist *bif;
1059 	struct bstp_port *bp;
1060 	int error;
1061 
1062 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1063 	if (bif == NULL)
1064 		return (ENOENT);
1065 	bp = &bif->bif_stp;
1066 
1067 	if (req->ifbr_ifsflags & IFBIF_SPAN)
1068 		/* SPAN is readonly */
1069 		return (EINVAL);
1070 
1071 	if (req->ifbr_ifsflags & IFBIF_STP) {
1072 		if ((bif->bif_flags & IFBIF_STP) == 0) {
1073 			error = bstp_enable(&bif->bif_stp);
1074 			if (error)
1075 				return (error);
1076 		}
1077 	} else {
1078 		if ((bif->bif_flags & IFBIF_STP) != 0)
1079 			bstp_disable(&bif->bif_stp);
1080 	}
1081 
1082 	/* Pass on STP flags */
1083 	bstp_set_edge(bp, req->ifbr_ifsflags & IFBIF_BSTP_EDGE ? 1 : 0);
1084 	bstp_set_autoedge(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOEDGE ? 1 : 0);
1085 	bstp_set_ptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_PTP ? 1 : 0);
1086 	bstp_set_autoptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOPTP ? 1 : 0);
1087 
1088 	/* Save the bits relating to the bridge */
1089 	bif->bif_flags = req->ifbr_ifsflags & IFBIFMASK;
1090 
1091 	return (0);
1092 }
1093 
1094 static int
1095 bridge_ioctl_scache(struct bridge_softc *sc, void *arg)
1096 {
1097 	struct ifbrparam *param = arg;
1098 
1099 	sc->sc_brtmax = param->ifbrp_csize;
1100 	bridge_rttrim(sc);
1101 
1102 	return (0);
1103 }
1104 
1105 static int
1106 bridge_ioctl_gcache(struct bridge_softc *sc, void *arg)
1107 {
1108 	struct ifbrparam *param = arg;
1109 
1110 	param->ifbrp_csize = sc->sc_brtmax;
1111 
1112 	return (0);
1113 }
1114 
1115 static int
1116 bridge_ioctl_gifs(struct bridge_softc *sc, void *arg)
1117 {
1118 	struct ifbifconf *bifc = arg;
1119 	struct bridge_iflist *bif;
1120 	struct ifbreq breq;
1121 	char *buf, *outbuf;
1122 	int count, buflen, len, error = 0;
1123 
1124 	count = 0;
1125 	LIST_FOREACH(bif, &sc->sc_iflist, bif_next)
1126 		count++;
1127 	LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1128 		count++;
1129 
1130 	buflen = sizeof(breq) * count;
1131 	if (bifc->ifbic_len == 0) {
1132 		bifc->ifbic_len = buflen;
1133 		return (0);
1134 	}
1135 	BRIDGE_UNLOCK(sc);
1136 	outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO);
1137 	BRIDGE_LOCK(sc);
1138 
1139 	count = 0;
1140 	buf = outbuf;
1141 	len = min(bifc->ifbic_len, buflen);
1142 	bzero(&breq, sizeof(breq));
1143 	LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1144 		if (len < sizeof(breq))
1145 			break;
1146 
1147 		strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname,
1148 		    sizeof(breq.ifbr_ifsname));
1149 		/* Fill in the ifbreq structure */
1150 		error = bridge_ioctl_gifflags(sc, &breq);
1151 		if (error)
1152 			break;
1153 		memcpy(buf, &breq, sizeof(breq));
1154 		count++;
1155 		buf += sizeof(breq);
1156 		len -= sizeof(breq);
1157 	}
1158 	LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) {
1159 		if (len < sizeof(breq))
1160 			break;
1161 
1162 		strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname,
1163 		    sizeof(breq.ifbr_ifsname));
1164 		breq.ifbr_ifsflags = bif->bif_flags;
1165 		breq.ifbr_portno = bif->bif_ifp->if_index & 0xfff;
1166 		memcpy(buf, &breq, sizeof(breq));
1167 		count++;
1168 		buf += sizeof(breq);
1169 		len -= sizeof(breq);
1170 	}
1171 
1172 	BRIDGE_UNLOCK(sc);
1173 	bifc->ifbic_len = sizeof(breq) * count;
1174 	error = copyout(outbuf, bifc->ifbic_req, bifc->ifbic_len);
1175 	BRIDGE_LOCK(sc);
1176 	free(outbuf, M_TEMP);
1177 	return (error);
1178 }
1179 
1180 static int
1181 bridge_ioctl_rts(struct bridge_softc *sc, void *arg)
1182 {
1183 	struct ifbaconf *bac = arg;
1184 	struct bridge_rtnode *brt;
1185 	struct ifbareq bareq;
1186 	char *buf, *outbuf;
1187 	int count, buflen, len, error = 0;
1188 
1189 	if (bac->ifbac_len == 0)
1190 		return (0);
1191 
1192 	count = 0;
1193 	LIST_FOREACH(brt, &sc->sc_rtlist, brt_list)
1194 		count++;
1195 	buflen = sizeof(bareq) * count;
1196 
1197 	BRIDGE_UNLOCK(sc);
1198 	outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO);
1199 	BRIDGE_LOCK(sc);
1200 
1201 	count = 0;
1202 	buf = outbuf;
1203 	len = min(bac->ifbac_len, buflen);
1204 	bzero(&bareq, sizeof(bareq));
1205 	LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) {
1206 		if (len < sizeof(bareq))
1207 			goto out;
1208 		strlcpy(bareq.ifba_ifsname, brt->brt_ifp->if_xname,
1209 		    sizeof(bareq.ifba_ifsname));
1210 		memcpy(bareq.ifba_dst, brt->brt_addr, sizeof(brt->brt_addr));
1211 		bareq.ifba_vlan = brt->brt_vlan;
1212 		if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC &&
1213 				time_uptime < brt->brt_expire)
1214 			bareq.ifba_expire = brt->brt_expire - time_uptime;
1215 		else
1216 			bareq.ifba_expire = 0;
1217 		bareq.ifba_flags = brt->brt_flags;
1218 
1219 		memcpy(buf, &bareq, sizeof(bareq));
1220 		count++;
1221 		buf += sizeof(bareq);
1222 		len -= sizeof(bareq);
1223 	}
1224 out:
1225 	BRIDGE_UNLOCK(sc);
1226 	bac->ifbac_len = sizeof(bareq) * count;
1227 	error = copyout(outbuf, bac->ifbac_req, bac->ifbac_len);
1228 	BRIDGE_LOCK(sc);
1229 	free(outbuf, M_TEMP);
1230 	return (error);
1231 }
1232 
1233 static int
1234 bridge_ioctl_saddr(struct bridge_softc *sc, void *arg)
1235 {
1236 	struct ifbareq *req = arg;
1237 	struct bridge_iflist *bif;
1238 	int error;
1239 
1240 	bif = bridge_lookup_member(sc, req->ifba_ifsname);
1241 	if (bif == NULL)
1242 		return (ENOENT);
1243 
1244 	error = bridge_rtupdate(sc, req->ifba_dst, req->ifba_vlan, bif, 1,
1245 	    req->ifba_flags);
1246 
1247 	return (error);
1248 }
1249 
1250 static int
1251 bridge_ioctl_sto(struct bridge_softc *sc, void *arg)
1252 {
1253 	struct ifbrparam *param = arg;
1254 
1255 	sc->sc_brttimeout = param->ifbrp_ctime;
1256 	return (0);
1257 }
1258 
1259 static int
1260 bridge_ioctl_gto(struct bridge_softc *sc, void *arg)
1261 {
1262 	struct ifbrparam *param = arg;
1263 
1264 	param->ifbrp_ctime = sc->sc_brttimeout;
1265 	return (0);
1266 }
1267 
1268 static int
1269 bridge_ioctl_daddr(struct bridge_softc *sc, void *arg)
1270 {
1271 	struct ifbareq *req = arg;
1272 
1273 	return (bridge_rtdaddr(sc, req->ifba_dst, req->ifba_vlan));
1274 }
1275 
1276 static int
1277 bridge_ioctl_flush(struct bridge_softc *sc, void *arg)
1278 {
1279 	struct ifbreq *req = arg;
1280 
1281 	bridge_rtflush(sc, req->ifbr_ifsflags);
1282 	return (0);
1283 }
1284 
1285 static int
1286 bridge_ioctl_gpri(struct bridge_softc *sc, void *arg)
1287 {
1288 	struct ifbrparam *param = arg;
1289 	struct bstp_state *bs = &sc->sc_stp;
1290 
1291 	param->ifbrp_prio = bs->bs_bridge_priority;
1292 	return (0);
1293 }
1294 
1295 static int
1296 bridge_ioctl_spri(struct bridge_softc *sc, void *arg)
1297 {
1298 	struct ifbrparam *param = arg;
1299 
1300 	return (bstp_set_priority(&sc->sc_stp, param->ifbrp_prio));
1301 }
1302 
1303 static int
1304 bridge_ioctl_ght(struct bridge_softc *sc, void *arg)
1305 {
1306 	struct ifbrparam *param = arg;
1307 	struct bstp_state *bs = &sc->sc_stp;
1308 
1309 	param->ifbrp_hellotime = bs->bs_bridge_htime >> 8;
1310 	return (0);
1311 }
1312 
1313 static int
1314 bridge_ioctl_sht(struct bridge_softc *sc, void *arg)
1315 {
1316 	struct ifbrparam *param = arg;
1317 
1318 	return (bstp_set_htime(&sc->sc_stp, param->ifbrp_hellotime));
1319 }
1320 
1321 static int
1322 bridge_ioctl_gfd(struct bridge_softc *sc, void *arg)
1323 {
1324 	struct ifbrparam *param = arg;
1325 	struct bstp_state *bs = &sc->sc_stp;
1326 
1327 	param->ifbrp_fwddelay = bs->bs_bridge_fdelay >> 8;
1328 	return (0);
1329 }
1330 
1331 static int
1332 bridge_ioctl_sfd(struct bridge_softc *sc, void *arg)
1333 {
1334 	struct ifbrparam *param = arg;
1335 
1336 	return (bstp_set_fdelay(&sc->sc_stp, param->ifbrp_fwddelay));
1337 }
1338 
1339 static int
1340 bridge_ioctl_gma(struct bridge_softc *sc, void *arg)
1341 {
1342 	struct ifbrparam *param = arg;
1343 	struct bstp_state *bs = &sc->sc_stp;
1344 
1345 	param->ifbrp_maxage = bs->bs_bridge_max_age >> 8;
1346 	return (0);
1347 }
1348 
1349 static int
1350 bridge_ioctl_sma(struct bridge_softc *sc, void *arg)
1351 {
1352 	struct ifbrparam *param = arg;
1353 
1354 	return (bstp_set_maxage(&sc->sc_stp, param->ifbrp_maxage));
1355 }
1356 
1357 static int
1358 bridge_ioctl_sifprio(struct bridge_softc *sc, void *arg)
1359 {
1360 	struct ifbreq *req = arg;
1361 	struct bridge_iflist *bif;
1362 
1363 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1364 	if (bif == NULL)
1365 		return (ENOENT);
1366 
1367 	return (bstp_set_port_priority(&bif->bif_stp, req->ifbr_priority));
1368 }
1369 
1370 static int
1371 bridge_ioctl_sifcost(struct bridge_softc *sc, void *arg)
1372 {
1373 	struct ifbreq *req = arg;
1374 	struct bridge_iflist *bif;
1375 
1376 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1377 	if (bif == NULL)
1378 		return (ENOENT);
1379 
1380 	return (bstp_set_path_cost(&bif->bif_stp, req->ifbr_path_cost));
1381 }
1382 
1383 static int
1384 bridge_ioctl_sifmaxaddr(struct bridge_softc *sc, void *arg)
1385 {
1386 	struct ifbreq *req = arg;
1387 	struct bridge_iflist *bif;
1388 
1389 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1390 	if (bif == NULL)
1391 		return (ENOENT);
1392 
1393 	bif->bif_addrmax = req->ifbr_addrmax;
1394 	return (0);
1395 }
1396 
1397 static int
1398 bridge_ioctl_addspan(struct bridge_softc *sc, void *arg)
1399 {
1400 	struct ifbreq *req = arg;
1401 	struct bridge_iflist *bif = NULL;
1402 	struct ifnet *ifs;
1403 
1404 	ifs = ifunit(req->ifbr_ifsname);
1405 	if (ifs == NULL)
1406 		return (ENOENT);
1407 
1408 	LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1409 		if (ifs == bif->bif_ifp)
1410 			return (EBUSY);
1411 
1412 	if (ifs->if_bridge != NULL)
1413 		return (EBUSY);
1414 
1415 	switch (ifs->if_type) {
1416 		case IFT_ETHER:
1417 		case IFT_GIF:
1418 		case IFT_L2VLAN:
1419 			break;
1420 		default:
1421 			return (EINVAL);
1422 	}
1423 
1424 	bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO);
1425 	if (bif == NULL)
1426 		return (ENOMEM);
1427 
1428 	bif->bif_ifp = ifs;
1429 	bif->bif_flags = IFBIF_SPAN;
1430 
1431 	LIST_INSERT_HEAD(&sc->sc_spanlist, bif, bif_next);
1432 
1433 	return (0);
1434 }
1435 
1436 static int
1437 bridge_ioctl_delspan(struct bridge_softc *sc, void *arg)
1438 {
1439 	struct ifbreq *req = arg;
1440 	struct bridge_iflist *bif;
1441 	struct ifnet *ifs;
1442 
1443 	ifs = ifunit(req->ifbr_ifsname);
1444 	if (ifs == NULL)
1445 		return (ENOENT);
1446 
1447 	LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1448 		if (ifs == bif->bif_ifp)
1449 			break;
1450 
1451 	if (bif == NULL)
1452 		return (ENOENT);
1453 
1454 	bridge_delete_span(sc, bif);
1455 
1456 	return (0);
1457 }
1458 
1459 static int
1460 bridge_ioctl_gbparam(struct bridge_softc *sc, void *arg)
1461 {
1462 	struct ifbropreq *req = arg;
1463 	struct bstp_state *bs = &sc->sc_stp;
1464 	struct bstp_port *root_port;
1465 
1466 	req->ifbop_maxage = bs->bs_bridge_max_age >> 8;
1467 	req->ifbop_hellotime = bs->bs_bridge_htime >> 8;
1468 	req->ifbop_fwddelay = bs->bs_bridge_fdelay >> 8;
1469 
1470 	root_port = bs->bs_root_port;
1471 	if (root_port == NULL)
1472 		req->ifbop_root_port = 0;
1473 	else
1474 		req->ifbop_root_port = root_port->bp_ifp->if_index;
1475 
1476 	req->ifbop_holdcount = bs->bs_txholdcount;
1477 	req->ifbop_priority = bs->bs_bridge_priority;
1478 	req->ifbop_protocol = bs->bs_protover;
1479 	req->ifbop_root_path_cost = bs->bs_root_pv.pv_cost;
1480 	req->ifbop_bridgeid = bs->bs_bridge_pv.pv_dbridge_id;
1481 	req->ifbop_designated_root = bs->bs_root_pv.pv_root_id;
1482 	req->ifbop_designated_bridge = bs->bs_root_pv.pv_dbridge_id;
1483 	req->ifbop_last_tc_time.tv_sec = bs->bs_last_tc_time.tv_sec;
1484 	req->ifbop_last_tc_time.tv_usec = bs->bs_last_tc_time.tv_usec;
1485 
1486 	return (0);
1487 }
1488 
1489 static int
1490 bridge_ioctl_grte(struct bridge_softc *sc, void *arg)
1491 {
1492 	struct ifbrparam *param = arg;
1493 
1494 	param->ifbrp_cexceeded = sc->sc_brtexceeded;
1495 	return (0);
1496 }
1497 
1498 static int
1499 bridge_ioctl_gifsstp(struct bridge_softc *sc, void *arg)
1500 {
1501 	struct ifbpstpconf *bifstp = arg;
1502 	struct bridge_iflist *bif;
1503 	struct bstp_port *bp;
1504 	struct ifbpstpreq bpreq;
1505 	char *buf, *outbuf;
1506 	int count, buflen, len, error = 0;
1507 
1508 	count = 0;
1509 	LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1510 		if ((bif->bif_flags & IFBIF_STP) != 0)
1511 			count++;
1512 	}
1513 
1514 	buflen = sizeof(bpreq) * count;
1515 	if (bifstp->ifbpstp_len == 0) {
1516 		bifstp->ifbpstp_len = buflen;
1517 		return (0);
1518 	}
1519 
1520 	BRIDGE_UNLOCK(sc);
1521 	outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO);
1522 	BRIDGE_LOCK(sc);
1523 
1524 	count = 0;
1525 	buf = outbuf;
1526 	len = min(bifstp->ifbpstp_len, buflen);
1527 	bzero(&bpreq, sizeof(bpreq));
1528 	LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1529 		if (len < sizeof(bpreq))
1530 			break;
1531 
1532 		if ((bif->bif_flags & IFBIF_STP) == 0)
1533 			continue;
1534 
1535 		bp = &bif->bif_stp;
1536 		bpreq.ifbp_portno = bif->bif_ifp->if_index & 0xfff;
1537 		bpreq.ifbp_fwd_trans = bp->bp_forward_transitions;
1538 		bpreq.ifbp_design_cost = bp->bp_desg_pv.pv_cost;
1539 		bpreq.ifbp_design_port = bp->bp_desg_pv.pv_port_id;
1540 		bpreq.ifbp_design_bridge = bp->bp_desg_pv.pv_dbridge_id;
1541 		bpreq.ifbp_design_root = bp->bp_desg_pv.pv_root_id;
1542 
1543 		memcpy(buf, &bpreq, sizeof(bpreq));
1544 		count++;
1545 		buf += sizeof(bpreq);
1546 		len -= sizeof(bpreq);
1547 	}
1548 
1549 	BRIDGE_UNLOCK(sc);
1550 	bifstp->ifbpstp_len = sizeof(bpreq) * count;
1551 	error = copyout(outbuf, bifstp->ifbpstp_req, bifstp->ifbpstp_len);
1552 	BRIDGE_LOCK(sc);
1553 	free(outbuf, M_TEMP);
1554 	return (error);
1555 }
1556 
1557 static int
1558 bridge_ioctl_sproto(struct bridge_softc *sc, void *arg)
1559 {
1560 	struct ifbrparam *param = arg;
1561 
1562 	return (bstp_set_protocol(&sc->sc_stp, param->ifbrp_proto));
1563 }
1564 
1565 static int
1566 bridge_ioctl_stxhc(struct bridge_softc *sc, void *arg)
1567 {
1568 	struct ifbrparam *param = arg;
1569 
1570 	return (bstp_set_holdcount(&sc->sc_stp, param->ifbrp_txhc));
1571 }
1572 
1573 /*
1574  * bridge_ifdetach:
1575  *
1576  *	Detach an interface from a bridge.  Called when a member
1577  *	interface is detaching.
1578  */
1579 static void
1580 bridge_ifdetach(void *arg __unused, struct ifnet *ifp)
1581 {
1582 	struct bridge_softc *sc = ifp->if_bridge;
1583 	struct bridge_iflist *bif;
1584 
1585 	/* Check if the interface is a bridge member */
1586 	if (sc != NULL) {
1587 		BRIDGE_LOCK(sc);
1588 
1589 		bif = bridge_lookup_member_if(sc, ifp);
1590 		if (bif != NULL)
1591 			bridge_delete_member(sc, bif, 1);
1592 
1593 		BRIDGE_UNLOCK(sc);
1594 		return;
1595 	}
1596 
1597 	/* Check if the interface is a span port */
1598 	mtx_lock(&bridge_list_mtx);
1599 	LIST_FOREACH(sc, &bridge_list, sc_list) {
1600 		BRIDGE_LOCK(sc);
1601 		LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1602 			if (ifp == bif->bif_ifp) {
1603 				bridge_delete_span(sc, bif);
1604 				break;
1605 			}
1606 
1607 		BRIDGE_UNLOCK(sc);
1608 	}
1609 	mtx_unlock(&bridge_list_mtx);
1610 }
1611 
1612 /*
1613  * bridge_init:
1614  *
1615  *	Initialize a bridge interface.
1616  */
1617 static void
1618 bridge_init(void *xsc)
1619 {
1620 	struct bridge_softc *sc = (struct bridge_softc *)xsc;
1621 	struct ifnet *ifp = sc->sc_ifp;
1622 
1623 	if (ifp->if_drv_flags & IFF_DRV_RUNNING)
1624 		return;
1625 
1626 	BRIDGE_LOCK(sc);
1627 	callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz,
1628 	    bridge_timer, sc);
1629 
1630 	ifp->if_drv_flags |= IFF_DRV_RUNNING;
1631 	bstp_init(&sc->sc_stp);		/* Initialize Spanning Tree */
1632 
1633 	BRIDGE_UNLOCK(sc);
1634 }
1635 
1636 /*
1637  * bridge_stop:
1638  *
1639  *	Stop the bridge interface.
1640  */
1641 static void
1642 bridge_stop(struct ifnet *ifp, int disable)
1643 {
1644 	struct bridge_softc *sc = ifp->if_softc;
1645 
1646 	BRIDGE_LOCK_ASSERT(sc);
1647 
1648 	if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
1649 		return;
1650 
1651 	callout_stop(&sc->sc_brcallout);
1652 	bstp_stop(&sc->sc_stp);
1653 
1654 	bridge_rtflush(sc, IFBF_FLUSHDYN);
1655 
1656 	ifp->if_drv_flags &= ~IFF_DRV_RUNNING;
1657 }
1658 
1659 /*
1660  * bridge_enqueue:
1661  *
1662  *	Enqueue a packet on a bridge member interface.
1663  *
1664  */
1665 static void
1666 bridge_enqueue(struct bridge_softc *sc, struct ifnet *dst_ifp, struct mbuf *m)
1667 {
1668 	int len, err = 0;
1669 	short mflags;
1670 	struct mbuf *m0;
1671 
1672 	len = m->m_pkthdr.len;
1673 	mflags = m->m_flags;
1674 
1675 	/* We may be sending a fragment so traverse the mbuf */
1676 	for (; m; m = m0) {
1677 		m0 = m->m_nextpkt;
1678 		m->m_nextpkt = NULL;
1679 
1680 		/*
1681 		 * If underlying interface can not do VLAN tag insertion itself
1682 		 * then attach a packet tag that holds it.
1683 		 */
1684 		if ((m->m_flags & M_VLANTAG) &&
1685 		    (dst_ifp->if_capenable & IFCAP_VLAN_HWTAGGING) == 0) {
1686 			m = ether_vlanencap(m, m->m_pkthdr.ether_vtag);
1687 			if (m == NULL) {
1688 				if_printf(dst_ifp,
1689 				    "unable to prepend VLAN header\n");
1690 				dst_ifp->if_oerrors++;
1691 				continue;
1692 			}
1693 			m->m_flags &= ~M_VLANTAG;
1694 		}
1695 
1696 		if (err == 0)
1697 			IFQ_ENQUEUE(&dst_ifp->if_snd, m, err);
1698 	}
1699 
1700 	if (err == 0) {
1701 
1702 		sc->sc_ifp->if_opackets++;
1703 		sc->sc_ifp->if_obytes += len;
1704 
1705 		dst_ifp->if_obytes += len;
1706 
1707 		if (mflags & M_MCAST) {
1708 			sc->sc_ifp->if_omcasts++;
1709 			dst_ifp->if_omcasts++;
1710 		}
1711 	}
1712 
1713 	if ((dst_ifp->if_drv_flags & IFF_DRV_OACTIVE) == 0)
1714 		(*dst_ifp->if_start)(dst_ifp);
1715 }
1716 
1717 /*
1718  * bridge_dummynet:
1719  *
1720  * 	Receive a queued packet from dummynet and pass it on to the output
1721  * 	interface.
1722  *
1723  *	The mbuf has the Ethernet header already attached.
1724  */
1725 static void
1726 bridge_dummynet(struct mbuf *m, struct ifnet *ifp)
1727 {
1728 	struct bridge_softc *sc;
1729 
1730 	sc = ifp->if_bridge;
1731 
1732 	/*
1733 	 * The packet didnt originate from a member interface. This should only
1734 	 * ever happen if a member interface is removed while packets are
1735 	 * queued for it.
1736 	 */
1737 	if (sc == NULL) {
1738 		m_freem(m);
1739 		return;
1740 	}
1741 
1742 	if (PFIL_HOOKED(&inet_pfil_hook)
1743 #ifdef INET6
1744 	    || PFIL_HOOKED(&inet6_pfil_hook)
1745 #endif
1746 	    ) {
1747 		if (bridge_pfil(&m, sc->sc_ifp, ifp, PFIL_OUT) != 0)
1748 			return;
1749 		if (m == NULL)
1750 			return;
1751 	}
1752 
1753 	bridge_enqueue(sc, ifp, m);
1754 }
1755 
1756 /*
1757  * bridge_output:
1758  *
1759  *	Send output from a bridge member interface.  This
1760  *	performs the bridging function for locally originated
1761  *	packets.
1762  *
1763  *	The mbuf has the Ethernet header already attached.  We must
1764  *	enqueue or free the mbuf before returning.
1765  */
1766 static int
1767 bridge_output(struct ifnet *ifp, struct mbuf *m, struct sockaddr *sa,
1768     struct rtentry *rt)
1769 {
1770 	struct ether_header *eh;
1771 	struct ifnet *dst_if;
1772 	struct bridge_softc *sc;
1773 	uint16_t vlan;
1774 
1775 	if (m->m_len < ETHER_HDR_LEN) {
1776 		m = m_pullup(m, ETHER_HDR_LEN);
1777 		if (m == NULL)
1778 			return (0);
1779 	}
1780 
1781 	eh = mtod(m, struct ether_header *);
1782 	sc = ifp->if_bridge;
1783 	vlan = VLANTAGOF(m);
1784 
1785 	BRIDGE_LOCK(sc);
1786 
1787 	/*
1788 	 * If bridge is down, but the original output interface is up,
1789 	 * go ahead and send out that interface.  Otherwise, the packet
1790 	 * is dropped below.
1791 	 */
1792 	if ((sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) {
1793 		dst_if = ifp;
1794 		goto sendunicast;
1795 	}
1796 
1797 	/*
1798 	 * If the packet is a multicast, or we don't know a better way to
1799 	 * get there, send to all interfaces.
1800 	 */
1801 	if (ETHER_IS_MULTICAST(eh->ether_dhost))
1802 		dst_if = NULL;
1803 	else
1804 		dst_if = bridge_rtlookup(sc, eh->ether_dhost, vlan);
1805 	if (dst_if == NULL) {
1806 		struct bridge_iflist *bif;
1807 		struct mbuf *mc;
1808 		int error = 0, used = 0;
1809 
1810 		bridge_span(sc, m);
1811 
1812 		BRIDGE_LOCK2REF(sc, error);
1813 		if (error) {
1814 			m_freem(m);
1815 			return (0);
1816 		}
1817 
1818 		LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1819 			dst_if = bif->bif_ifp;
1820 
1821 			if (dst_if->if_type == IFT_GIF)
1822 				continue;
1823 			if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
1824 				continue;
1825 
1826 			/*
1827 			 * If this is not the original output interface,
1828 			 * and the interface is participating in spanning
1829 			 * tree, make sure the port is in a state that
1830 			 * allows forwarding.
1831 			 */
1832 			if (dst_if != ifp && (bif->bif_flags & IFBIF_STP) &&
1833 			    bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
1834 				continue;
1835 
1836 			if (LIST_NEXT(bif, bif_next) == NULL) {
1837 				used = 1;
1838 				mc = m;
1839 			} else {
1840 				mc = m_copypacket(m, M_DONTWAIT);
1841 				if (mc == NULL) {
1842 					sc->sc_ifp->if_oerrors++;
1843 					continue;
1844 				}
1845 			}
1846 
1847 			bridge_enqueue(sc, dst_if, mc);
1848 		}
1849 		if (used == 0)
1850 			m_freem(m);
1851 		BRIDGE_UNREF(sc);
1852 		return (0);
1853 	}
1854 
1855 sendunicast:
1856 	/*
1857 	 * XXX Spanning tree consideration here?
1858 	 */
1859 
1860 	bridge_span(sc, m);
1861 	if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) {
1862 		m_freem(m);
1863 		BRIDGE_UNLOCK(sc);
1864 		return (0);
1865 	}
1866 
1867 	BRIDGE_UNLOCK(sc);
1868 	bridge_enqueue(sc, dst_if, m);
1869 	return (0);
1870 }
1871 
1872 /*
1873  * bridge_start:
1874  *
1875  *	Start output on a bridge.
1876  *
1877  */
1878 static void
1879 bridge_start(struct ifnet *ifp)
1880 {
1881 	struct bridge_softc *sc;
1882 	struct mbuf *m;
1883 	struct ether_header *eh;
1884 	struct ifnet *dst_if;
1885 
1886 	sc = ifp->if_softc;
1887 
1888 	ifp->if_drv_flags |= IFF_DRV_OACTIVE;
1889 	for (;;) {
1890 		IFQ_DEQUEUE(&ifp->if_snd, m);
1891 		if (m == 0)
1892 			break;
1893 		ETHER_BPF_MTAP(ifp, m);
1894 
1895 		eh = mtod(m, struct ether_header *);
1896 		dst_if = NULL;
1897 
1898 		BRIDGE_LOCK(sc);
1899 		if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) {
1900 			dst_if = bridge_rtlookup(sc, eh->ether_dhost, 1);
1901 		}
1902 
1903 		if (dst_if == NULL)
1904 			bridge_broadcast(sc, ifp, m, 0);
1905 		else {
1906 			BRIDGE_UNLOCK(sc);
1907 			bridge_enqueue(sc, dst_if, m);
1908 		}
1909 	}
1910 	ifp->if_drv_flags &= ~IFF_DRV_OACTIVE;
1911 }
1912 
1913 /*
1914  * bridge_forward:
1915  *
1916  *	The forwarding function of the bridge.
1917  *
1918  *	NOTE: Releases the lock on return.
1919  */
1920 static void
1921 bridge_forward(struct bridge_softc *sc, struct bridge_iflist *sbif,
1922     struct mbuf *m)
1923 {
1924 	struct bridge_iflist *dbif;
1925 	struct ifnet *src_if, *dst_if, *ifp;
1926 	struct ether_header *eh;
1927 	uint16_t vlan;
1928 	uint8_t *dst;
1929 	int error;
1930 
1931 	src_if = m->m_pkthdr.rcvif;
1932 	ifp = sc->sc_ifp;
1933 
1934 	ifp->if_ipackets++;
1935 	ifp->if_ibytes += m->m_pkthdr.len;
1936 	vlan = VLANTAGOF(m);
1937 
1938 	if ((sbif->bif_flags & IFBIF_STP) &&
1939 	    sbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
1940 		goto drop;
1941 
1942 	eh = mtod(m, struct ether_header *);
1943 	dst = eh->ether_dhost;
1944 
1945 	/* If the interface is learning, record the address. */
1946 	if (sbif->bif_flags & IFBIF_LEARNING) {
1947 		error = bridge_rtupdate(sc, eh->ether_shost, vlan,
1948 		    sbif, 0, IFBAF_DYNAMIC);
1949 		/*
1950 		 * If the interface has addresses limits then deny any source
1951 		 * that is not in the cache.
1952 		 */
1953 		if (error && sbif->bif_addrmax)
1954 			goto drop;
1955 	}
1956 
1957 	if ((sbif->bif_flags & IFBIF_STP) != 0 &&
1958 	    sbif->bif_stp.bp_state == BSTP_IFSTATE_LEARNING)
1959 		goto drop;
1960 
1961 	/*
1962 	 * At this point, the port either doesn't participate
1963 	 * in spanning tree or it is in the forwarding state.
1964 	 */
1965 
1966 	/*
1967 	 * If the packet is unicast, destined for someone on
1968 	 * "this" side of the bridge, drop it.
1969 	 */
1970 	if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) {
1971 		dst_if = bridge_rtlookup(sc, dst, vlan);
1972 		if (src_if == dst_if)
1973 			goto drop;
1974 	} else {
1975 		/*
1976 		 * Check if its a reserved multicast address, any address
1977 		 * listed in 802.1D section 7.12.6 may not be forwarded by the
1978 		 * bridge.
1979 		 * This is currently 01-80-C2-00-00-00 to 01-80-C2-00-00-0F
1980 		 */
1981 		if (dst[0] == 0x01 && dst[1] == 0x80 &&
1982 		    dst[2] == 0xc2 && dst[3] == 0x00 &&
1983 		    dst[4] == 0x00 && dst[5] <= 0x0f)
1984 			goto drop;
1985 
1986 		/* ...forward it to all interfaces. */
1987 		ifp->if_imcasts++;
1988 		dst_if = NULL;
1989 	}
1990 
1991 	/*
1992 	 * If we have a destination interface which is a member of our bridge,
1993 	 * OR this is a unicast packet, push it through the bpf(4) machinery.
1994 	 * For broadcast or multicast packets, don't bother because it will
1995 	 * be reinjected into ether_input. We do this before we pass the packets
1996 	 * through the pfil(9) framework, as it is possible that pfil(9) will
1997 	 * drop the packet, or possibly modify it, making it difficult to debug
1998 	 * firewall issues on the bridge.
1999 	 */
2000 	if (dst_if != NULL || (m->m_flags & (M_BCAST | M_MCAST)) == 0)
2001 		ETHER_BPF_MTAP(ifp, m);
2002 
2003 	/* run the packet filter */
2004 	if (PFIL_HOOKED(&inet_pfil_hook)
2005 #ifdef INET6
2006 	    || PFIL_HOOKED(&inet6_pfil_hook)
2007 #endif
2008 	    ) {
2009 		BRIDGE_UNLOCK(sc);
2010 		if (bridge_pfil(&m, ifp, src_if, PFIL_IN) != 0)
2011 			return;
2012 		if (m == NULL)
2013 			return;
2014 		BRIDGE_LOCK(sc);
2015 	}
2016 
2017 	if (dst_if == NULL) {
2018 		bridge_broadcast(sc, src_if, m, 1);
2019 		return;
2020 	}
2021 
2022 	/*
2023 	 * At this point, we're dealing with a unicast frame
2024 	 * going to a different interface.
2025 	 */
2026 	if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2027 		goto drop;
2028 
2029 	dbif = bridge_lookup_member_if(sc, dst_if);
2030 	if (dbif == NULL)
2031 		/* Not a member of the bridge (anymore?) */
2032 		goto drop;
2033 
2034 	/* Private segments can not talk to each other */
2035 	if (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE)
2036 		goto drop;
2037 
2038 	if ((dbif->bif_flags & IFBIF_STP) &&
2039 	    dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2040 		goto drop;
2041 
2042 	BRIDGE_UNLOCK(sc);
2043 
2044 	if (PFIL_HOOKED(&inet_pfil_hook)
2045 #ifdef INET6
2046 	    || PFIL_HOOKED(&inet6_pfil_hook)
2047 #endif
2048 	    ) {
2049 		if (bridge_pfil(&m, ifp, dst_if, PFIL_OUT) != 0)
2050 			return;
2051 		if (m == NULL)
2052 			return;
2053 	}
2054 
2055 	bridge_enqueue(sc, dst_if, m);
2056 	return;
2057 
2058 drop:
2059 	BRIDGE_UNLOCK(sc);
2060 	m_freem(m);
2061 }
2062 
2063 /*
2064  * bridge_input:
2065  *
2066  *	Receive input from a member interface.  Queue the packet for
2067  *	bridging if it is not for us.
2068  */
2069 static struct mbuf *
2070 bridge_input(struct ifnet *ifp, struct mbuf *m)
2071 {
2072 	struct bridge_softc *sc = ifp->if_bridge;
2073 	struct bridge_iflist *bif, *bif2;
2074 	struct ifnet *bifp;
2075 	struct ether_header *eh;
2076 	struct mbuf *mc, *mc2;
2077 	uint16_t vlan;
2078 	int error;
2079 
2080 	if ((sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
2081 		return (m);
2082 
2083 	bifp = sc->sc_ifp;
2084 	vlan = VLANTAGOF(m);
2085 
2086 	/*
2087 	 * Implement support for bridge monitoring. If this flag has been
2088 	 * set on this interface, discard the packet once we push it through
2089 	 * the bpf(4) machinery, but before we do, increment the byte and
2090 	 * packet counters associated with this interface.
2091 	 */
2092 	if ((bifp->if_flags & IFF_MONITOR) != 0) {
2093 		m->m_pkthdr.rcvif  = bifp;
2094 		ETHER_BPF_MTAP(bifp, m);
2095 		bifp->if_ipackets++;
2096 		bifp->if_ibytes += m->m_pkthdr.len;
2097 		m_freem(m);
2098 		return (NULL);
2099 	}
2100 	BRIDGE_LOCK(sc);
2101 	bif = bridge_lookup_member_if(sc, ifp);
2102 	if (bif == NULL) {
2103 		BRIDGE_UNLOCK(sc);
2104 		return (m);
2105 	}
2106 
2107 	eh = mtod(m, struct ether_header *);
2108 
2109 	bridge_span(sc, m);
2110 
2111 	if (m->m_flags & (M_BCAST|M_MCAST)) {
2112 		/* Tap off 802.1D packets; they do not get forwarded. */
2113 		if (memcmp(eh->ether_dhost, bstp_etheraddr,
2114 		    ETHER_ADDR_LEN) == 0) {
2115 			m = bstp_input(&bif->bif_stp, ifp, m);
2116 			if (m == NULL) {
2117 				BRIDGE_UNLOCK(sc);
2118 				return (NULL);
2119 			}
2120 		}
2121 
2122 		if ((bif->bif_flags & IFBIF_STP) &&
2123 		    bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) {
2124 			BRIDGE_UNLOCK(sc);
2125 			return (m);
2126 		}
2127 
2128 		/*
2129 		 * Make a deep copy of the packet and enqueue the copy
2130 		 * for bridge processing; return the original packet for
2131 		 * local processing.
2132 		 */
2133 		mc = m_dup(m, M_DONTWAIT);
2134 		if (mc == NULL) {
2135 			BRIDGE_UNLOCK(sc);
2136 			return (m);
2137 		}
2138 
2139 		/* Perform the bridge forwarding function with the copy. */
2140 		bridge_forward(sc, bif, mc);
2141 
2142 		/*
2143 		 * Reinject the mbuf as arriving on the bridge so we have a
2144 		 * chance at claiming multicast packets. We can not loop back
2145 		 * here from ether_input as a bridge is never a member of a
2146 		 * bridge.
2147 		 */
2148 		KASSERT(bifp->if_bridge == NULL,
2149 		    ("loop created in bridge_input"));
2150 		mc2 = m_dup(m, M_DONTWAIT);
2151 		if (mc2 != NULL) {
2152 			/* Keep the layer3 header aligned */
2153 			int i = min(mc2->m_pkthdr.len, max_protohdr);
2154 			mc2 = m_copyup(mc2, i, ETHER_ALIGN);
2155 		}
2156 		if (mc2 != NULL) {
2157 			mc2->m_pkthdr.rcvif = bifp;
2158 			(*bifp->if_input)(bifp, mc2);
2159 		}
2160 
2161 		/* Return the original packet for local processing. */
2162 		return (m);
2163 	}
2164 
2165 	if ((bif->bif_flags & IFBIF_STP) &&
2166 	    bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) {
2167 		BRIDGE_UNLOCK(sc);
2168 		return (m);
2169 	}
2170 
2171 #ifdef DEV_CARP
2172 #   define OR_CARP_CHECK_WE_ARE_DST(iface) \
2173 	|| ((iface)->if_carp \
2174 	    && carp_forus((iface)->if_carp, eh->ether_dhost))
2175 #   define OR_CARP_CHECK_WE_ARE_SRC(iface) \
2176 	|| ((iface)->if_carp \
2177 	    && carp_forus((iface)->if_carp, eh->ether_shost))
2178 #else
2179 #   define OR_CARP_CHECK_WE_ARE_DST(iface)
2180 #   define OR_CARP_CHECK_WE_ARE_SRC(iface)
2181 #endif
2182 
2183 #ifdef INET6
2184 #   define OR_PFIL_HOOKED_INET6 \
2185 	|| PFIL_HOOKED(&inet6_pfil_hook)
2186 #else
2187 #   define OR_PFIL_HOOKED_INET6
2188 #endif
2189 
2190 #define GRAB_OUR_PACKETS(iface) \
2191 	if ((iface)->if_type == IFT_GIF) \
2192 		continue; \
2193 	/* It is destined for us. */ \
2194 	if (memcmp(IF_LLADDR((iface)), eh->ether_dhost,  ETHER_ADDR_LEN) == 0 \
2195 	    OR_CARP_CHECK_WE_ARE_DST((iface))				\
2196 	    ) {								\
2197 		if ((iface)->if_type == IFT_BRIDGE) {			\
2198 			ETHER_BPF_MTAP(iface, m);			\
2199 			iface->if_ipackets++;				\
2200 			/* Filter on the physical interface. */		\
2201 			if (pfil_local_phys &&				\
2202 			    (PFIL_HOOKED(&inet_pfil_hook)		\
2203 			     OR_PFIL_HOOKED_INET6)) {			\
2204 				if (bridge_pfil(&m, NULL, ifp,		\
2205 				    PFIL_IN) != 0 || m == NULL) {	\
2206 					BRIDGE_UNLOCK(sc);		\
2207 					return (NULL);			\
2208 				}					\
2209 			}						\
2210 		}							\
2211 		if (bif->bif_flags & IFBIF_LEARNING) {			\
2212 			error = bridge_rtupdate(sc, eh->ether_shost,	\
2213 			    vlan, bif, 0, IFBAF_DYNAMIC);		\
2214 			if (error && bif->bif_addrmax) {		\
2215 				BRIDGE_UNLOCK(sc);			\
2216 				m_freem(m);				\
2217 				return (NULL);				\
2218 			}						\
2219 		}							\
2220 		m->m_pkthdr.rcvif = iface;				\
2221 		BRIDGE_UNLOCK(sc);					\
2222 		return (m);						\
2223 	}								\
2224 									\
2225 	/* We just received a packet that we sent out. */		\
2226 	if (memcmp(IF_LLADDR((iface)), eh->ether_shost, ETHER_ADDR_LEN) == 0 \
2227 	    OR_CARP_CHECK_WE_ARE_SRC((iface))			\
2228 	    ) {								\
2229 		BRIDGE_UNLOCK(sc);					\
2230 		m_freem(m);						\
2231 		return (NULL);						\
2232 	}
2233 
2234 	/*
2235 	 * Unicast.  Make sure it's not for the bridge.
2236 	 */
2237 	do { GRAB_OUR_PACKETS(bifp) } while (0);
2238 
2239 	/*
2240 	 * Give a chance for ifp at first priority. This will help when	the
2241 	 * packet comes through the interface like VLAN's with the same MACs
2242 	 * on several interfaces from the same bridge. This also will save
2243 	 * some CPU cycles in case the destination interface and the input
2244 	 * interface (eq ifp) are the same.
2245 	 */
2246 	do { GRAB_OUR_PACKETS(ifp) } while (0);
2247 
2248 	/* Now check the all bridge members. */
2249 	LIST_FOREACH(bif2, &sc->sc_iflist, bif_next) {
2250 		GRAB_OUR_PACKETS(bif2->bif_ifp)
2251 	}
2252 
2253 #undef OR_CARP_CHECK_WE_ARE_DST
2254 #undef OR_CARP_CHECK_WE_ARE_SRC
2255 #undef OR_PFIL_HOOKED_INET6
2256 #undef GRAB_OUR_PACKETS
2257 
2258 	/* Perform the bridge forwarding function. */
2259 	bridge_forward(sc, bif, m);
2260 
2261 	return (NULL);
2262 }
2263 
2264 /*
2265  * bridge_broadcast:
2266  *
2267  *	Send a frame to all interfaces that are members of
2268  *	the bridge, except for the one on which the packet
2269  *	arrived.
2270  *
2271  *	NOTE: Releases the lock on return.
2272  */
2273 static void
2274 bridge_broadcast(struct bridge_softc *sc, struct ifnet *src_if,
2275     struct mbuf *m, int runfilt)
2276 {
2277 	struct bridge_iflist *dbif, *sbif;
2278 	struct mbuf *mc;
2279 	struct ifnet *dst_if;
2280 	int error = 0, used = 0, i;
2281 
2282 	sbif = bridge_lookup_member_if(sc, src_if);
2283 
2284 	BRIDGE_LOCK2REF(sc, error);
2285 	if (error) {
2286 		m_freem(m);
2287 		return;
2288 	}
2289 
2290 	/* Filter on the bridge interface before broadcasting */
2291 	if (runfilt && (PFIL_HOOKED(&inet_pfil_hook)
2292 #ifdef INET6
2293 	    || PFIL_HOOKED(&inet6_pfil_hook)
2294 #endif
2295 	    )) {
2296 		if (bridge_pfil(&m, sc->sc_ifp, NULL, PFIL_OUT) != 0)
2297 			goto out;
2298 		if (m == NULL)
2299 			goto out;
2300 	}
2301 
2302 	LIST_FOREACH(dbif, &sc->sc_iflist, bif_next) {
2303 		dst_if = dbif->bif_ifp;
2304 		if (dst_if == src_if)
2305 			continue;
2306 
2307 		/* Private segments can not talk to each other */
2308 		if (sbif && (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE))
2309 			continue;
2310 
2311 		if ((dbif->bif_flags & IFBIF_STP) &&
2312 		    dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2313 			continue;
2314 
2315 		if ((dbif->bif_flags & IFBIF_DISCOVER) == 0 &&
2316 		    (m->m_flags & (M_BCAST|M_MCAST)) == 0)
2317 			continue;
2318 
2319 		if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2320 			continue;
2321 
2322 		if (LIST_NEXT(dbif, bif_next) == NULL) {
2323 			mc = m;
2324 			used = 1;
2325 		} else {
2326 			mc = m_dup(m, M_DONTWAIT);
2327 			if (mc == NULL) {
2328 				sc->sc_ifp->if_oerrors++;
2329 				continue;
2330 			}
2331 		}
2332 
2333 		/*
2334 		 * Filter on the output interface. Pass a NULL bridge interface
2335 		 * pointer so we do not redundantly filter on the bridge for
2336 		 * each interface we broadcast on.
2337 		 */
2338 		if (runfilt && (PFIL_HOOKED(&inet_pfil_hook)
2339 #ifdef INET6
2340 		    || PFIL_HOOKED(&inet6_pfil_hook)
2341 #endif
2342 		    )) {
2343 			if (used == 0) {
2344 				/* Keep the layer3 header aligned */
2345 				i = min(mc->m_pkthdr.len, max_protohdr);
2346 				mc = m_copyup(mc, i, ETHER_ALIGN);
2347 				if (mc == NULL) {
2348 					sc->sc_ifp->if_oerrors++;
2349 					continue;
2350 				}
2351 			}
2352 			if (bridge_pfil(&mc, NULL, dst_if, PFIL_OUT) != 0)
2353 				continue;
2354 			if (mc == NULL)
2355 				continue;
2356 		}
2357 
2358 		bridge_enqueue(sc, dst_if, mc);
2359 	}
2360 	if (used == 0)
2361 		m_freem(m);
2362 
2363 out:
2364 	BRIDGE_UNREF(sc);
2365 }
2366 
2367 /*
2368  * bridge_span:
2369  *
2370  *	Duplicate a packet out one or more interfaces that are in span mode,
2371  *	the original mbuf is unmodified.
2372  */
2373 static void
2374 bridge_span(struct bridge_softc *sc, struct mbuf *m)
2375 {
2376 	struct bridge_iflist *bif;
2377 	struct ifnet *dst_if;
2378 	struct mbuf *mc;
2379 
2380 	if (LIST_EMPTY(&sc->sc_spanlist))
2381 		return;
2382 
2383 	LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) {
2384 		dst_if = bif->bif_ifp;
2385 
2386 		if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2387 			continue;
2388 
2389 		mc = m_copypacket(m, M_DONTWAIT);
2390 		if (mc == NULL) {
2391 			sc->sc_ifp->if_oerrors++;
2392 			continue;
2393 		}
2394 
2395 		bridge_enqueue(sc, dst_if, mc);
2396 	}
2397 }
2398 
2399 /*
2400  * bridge_rtupdate:
2401  *
2402  *	Add a bridge routing entry.
2403  */
2404 static int
2405 bridge_rtupdate(struct bridge_softc *sc, const uint8_t *dst, uint16_t vlan,
2406     struct bridge_iflist *bif, int setflags, uint8_t flags)
2407 {
2408 	struct bridge_rtnode *brt;
2409 	int error;
2410 
2411 	BRIDGE_LOCK_ASSERT(sc);
2412 
2413 	/* Check the source address is valid and not multicast. */
2414 	if (ETHER_IS_MULTICAST(dst) ||
2415 	    (dst[0] == 0 && dst[1] == 0 && dst[2] == 0 &&
2416 	     dst[3] == 0 && dst[4] == 0 && dst[5] == 0) != 0)
2417 		return (EINVAL);
2418 
2419 	/* 802.1p frames map to vlan 1 */
2420 	if (vlan == 0)
2421 		vlan = 1;
2422 
2423 	/*
2424 	 * A route for this destination might already exist.  If so,
2425 	 * update it, otherwise create a new one.
2426 	 */
2427 	if ((brt = bridge_rtnode_lookup(sc, dst, vlan)) == NULL) {
2428 		if (sc->sc_brtcnt >= sc->sc_brtmax) {
2429 			sc->sc_brtexceeded++;
2430 			return (ENOSPC);
2431 		}
2432 		/* Check per interface address limits (if enabled) */
2433 		if (bif->bif_addrmax && bif->bif_addrcnt >= bif->bif_addrmax) {
2434 			bif->bif_addrexceeded++;
2435 			return (ENOSPC);
2436 		}
2437 
2438 		/*
2439 		 * Allocate a new bridge forwarding node, and
2440 		 * initialize the expiration time and Ethernet
2441 		 * address.
2442 		 */
2443 		brt = uma_zalloc(bridge_rtnode_zone, M_NOWAIT | M_ZERO);
2444 		if (brt == NULL)
2445 			return (ENOMEM);
2446 
2447 		if (bif->bif_flags & IFBIF_STICKY)
2448 			brt->brt_flags = IFBAF_STICKY;
2449 		else
2450 			brt->brt_flags = IFBAF_DYNAMIC;
2451 
2452 		memcpy(brt->brt_addr, dst, ETHER_ADDR_LEN);
2453 		brt->brt_vlan = vlan;
2454 
2455 		if ((error = bridge_rtnode_insert(sc, brt)) != 0) {
2456 			uma_zfree(bridge_rtnode_zone, brt);
2457 			return (error);
2458 		}
2459 		brt->brt_dst = bif;
2460 		bif->bif_addrcnt++;
2461 	}
2462 
2463 	if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC &&
2464 	    brt->brt_dst != bif) {
2465 		brt->brt_dst->bif_addrcnt--;
2466 		brt->brt_dst = bif;
2467 		brt->brt_dst->bif_addrcnt++;
2468 	}
2469 
2470 	if ((flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2471 		brt->brt_expire = time_uptime + sc->sc_brttimeout;
2472 	if (setflags)
2473 		brt->brt_flags = flags;
2474 
2475 	return (0);
2476 }
2477 
2478 /*
2479  * bridge_rtlookup:
2480  *
2481  *	Lookup the destination interface for an address.
2482  */
2483 static struct ifnet *
2484 bridge_rtlookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2485 {
2486 	struct bridge_rtnode *brt;
2487 
2488 	BRIDGE_LOCK_ASSERT(sc);
2489 
2490 	if ((brt = bridge_rtnode_lookup(sc, addr, vlan)) == NULL)
2491 		return (NULL);
2492 
2493 	return (brt->brt_ifp);
2494 }
2495 
2496 /*
2497  * bridge_rttrim:
2498  *
2499  *	Trim the routine table so that we have a number
2500  *	of routing entries less than or equal to the
2501  *	maximum number.
2502  */
2503 static void
2504 bridge_rttrim(struct bridge_softc *sc)
2505 {
2506 	struct bridge_rtnode *brt, *nbrt;
2507 
2508 	BRIDGE_LOCK_ASSERT(sc);
2509 
2510 	/* Make sure we actually need to do this. */
2511 	if (sc->sc_brtcnt <= sc->sc_brtmax)
2512 		return;
2513 
2514 	/* Force an aging cycle; this might trim enough addresses. */
2515 	bridge_rtage(sc);
2516 	if (sc->sc_brtcnt <= sc->sc_brtmax)
2517 		return;
2518 
2519 	LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2520 		if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) {
2521 			bridge_rtnode_destroy(sc, brt);
2522 			if (sc->sc_brtcnt <= sc->sc_brtmax)
2523 				return;
2524 		}
2525 	}
2526 }
2527 
2528 /*
2529  * bridge_timer:
2530  *
2531  *	Aging timer for the bridge.
2532  */
2533 static void
2534 bridge_timer(void *arg)
2535 {
2536 	struct bridge_softc *sc = arg;
2537 
2538 	BRIDGE_LOCK_ASSERT(sc);
2539 
2540 	bridge_rtage(sc);
2541 
2542 	if (sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING)
2543 		callout_reset(&sc->sc_brcallout,
2544 		    bridge_rtable_prune_period * hz, bridge_timer, sc);
2545 }
2546 
2547 /*
2548  * bridge_rtage:
2549  *
2550  *	Perform an aging cycle.
2551  */
2552 static void
2553 bridge_rtage(struct bridge_softc *sc)
2554 {
2555 	struct bridge_rtnode *brt, *nbrt;
2556 
2557 	BRIDGE_LOCK_ASSERT(sc);
2558 
2559 	LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2560 		if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) {
2561 			if (time_uptime >= brt->brt_expire)
2562 				bridge_rtnode_destroy(sc, brt);
2563 		}
2564 	}
2565 }
2566 
2567 /*
2568  * bridge_rtflush:
2569  *
2570  *	Remove all dynamic addresses from the bridge.
2571  */
2572 static void
2573 bridge_rtflush(struct bridge_softc *sc, int full)
2574 {
2575 	struct bridge_rtnode *brt, *nbrt;
2576 
2577 	BRIDGE_LOCK_ASSERT(sc);
2578 
2579 	LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2580 		if (full || (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2581 			bridge_rtnode_destroy(sc, brt);
2582 	}
2583 }
2584 
2585 /*
2586  * bridge_rtdaddr:
2587  *
2588  *	Remove an address from the table.
2589  */
2590 static int
2591 bridge_rtdaddr(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2592 {
2593 	struct bridge_rtnode *brt;
2594 	int found = 0;
2595 
2596 	BRIDGE_LOCK_ASSERT(sc);
2597 
2598 	/*
2599 	 * If vlan is zero then we want to delete for all vlans so the lookup
2600 	 * may return more than one.
2601 	 */
2602 	while ((brt = bridge_rtnode_lookup(sc, addr, vlan)) != NULL) {
2603 		bridge_rtnode_destroy(sc, brt);
2604 		found = 1;
2605 	}
2606 
2607 	return (found ? 0 : ENOENT);
2608 }
2609 
2610 /*
2611  * bridge_rtdelete:
2612  *
2613  *	Delete routes to a speicifc member interface.
2614  */
2615 static void
2616 bridge_rtdelete(struct bridge_softc *sc, struct ifnet *ifp, int full)
2617 {
2618 	struct bridge_rtnode *brt, *nbrt;
2619 
2620 	BRIDGE_LOCK_ASSERT(sc);
2621 
2622 	LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2623 		if (brt->brt_ifp == ifp && (full ||
2624 			    (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC))
2625 			bridge_rtnode_destroy(sc, brt);
2626 	}
2627 }
2628 
2629 /*
2630  * bridge_rtable_init:
2631  *
2632  *	Initialize the route table for this bridge.
2633  */
2634 static int
2635 bridge_rtable_init(struct bridge_softc *sc)
2636 {
2637 	int i;
2638 
2639 	sc->sc_rthash = malloc(sizeof(*sc->sc_rthash) * BRIDGE_RTHASH_SIZE,
2640 	    M_DEVBUF, M_NOWAIT);
2641 	if (sc->sc_rthash == NULL)
2642 		return (ENOMEM);
2643 
2644 	for (i = 0; i < BRIDGE_RTHASH_SIZE; i++)
2645 		LIST_INIT(&sc->sc_rthash[i]);
2646 
2647 	sc->sc_rthash_key = arc4random();
2648 
2649 	LIST_INIT(&sc->sc_rtlist);
2650 
2651 	return (0);
2652 }
2653 
2654 /*
2655  * bridge_rtable_fini:
2656  *
2657  *	Deconstruct the route table for this bridge.
2658  */
2659 static void
2660 bridge_rtable_fini(struct bridge_softc *sc)
2661 {
2662 
2663 	KASSERT(sc->sc_brtcnt == 0,
2664 	    ("%s: %d bridge routes referenced", __func__, sc->sc_brtcnt));
2665 	free(sc->sc_rthash, M_DEVBUF);
2666 }
2667 
2668 /*
2669  * The following hash function is adapted from "Hash Functions" by Bob Jenkins
2670  * ("Algorithm Alley", Dr. Dobbs Journal, September 1997).
2671  */
2672 #define	mix(a, b, c)							\
2673 do {									\
2674 	a -= b; a -= c; a ^= (c >> 13);					\
2675 	b -= c; b -= a; b ^= (a << 8);					\
2676 	c -= a; c -= b; c ^= (b >> 13);					\
2677 	a -= b; a -= c; a ^= (c >> 12);					\
2678 	b -= c; b -= a; b ^= (a << 16);					\
2679 	c -= a; c -= b; c ^= (b >> 5);					\
2680 	a -= b; a -= c; a ^= (c >> 3);					\
2681 	b -= c; b -= a; b ^= (a << 10);					\
2682 	c -= a; c -= b; c ^= (b >> 15);					\
2683 } while (/*CONSTCOND*/0)
2684 
2685 static __inline uint32_t
2686 bridge_rthash(struct bridge_softc *sc, const uint8_t *addr)
2687 {
2688 	uint32_t a = 0x9e3779b9, b = 0x9e3779b9, c = sc->sc_rthash_key;
2689 
2690 	b += addr[5] << 8;
2691 	b += addr[4];
2692 	a += addr[3] << 24;
2693 	a += addr[2] << 16;
2694 	a += addr[1] << 8;
2695 	a += addr[0];
2696 
2697 	mix(a, b, c);
2698 
2699 	return (c & BRIDGE_RTHASH_MASK);
2700 }
2701 
2702 #undef mix
2703 
2704 static int
2705 bridge_rtnode_addr_cmp(const uint8_t *a, const uint8_t *b)
2706 {
2707 	int i, d;
2708 
2709 	for (i = 0, d = 0; i < ETHER_ADDR_LEN && d == 0; i++) {
2710 		d = ((int)a[i]) - ((int)b[i]);
2711 	}
2712 
2713 	return (d);
2714 }
2715 
2716 /*
2717  * bridge_rtnode_lookup:
2718  *
2719  *	Look up a bridge route node for the specified destination. Compare the
2720  *	vlan id or if zero then just return the first match.
2721  */
2722 static struct bridge_rtnode *
2723 bridge_rtnode_lookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2724 {
2725 	struct bridge_rtnode *brt;
2726 	uint32_t hash;
2727 	int dir;
2728 
2729 	BRIDGE_LOCK_ASSERT(sc);
2730 
2731 	hash = bridge_rthash(sc, addr);
2732 	LIST_FOREACH(brt, &sc->sc_rthash[hash], brt_hash) {
2733 		dir = bridge_rtnode_addr_cmp(addr, brt->brt_addr);
2734 		if (dir == 0 && (brt->brt_vlan == vlan || vlan == 0))
2735 			return (brt);
2736 		if (dir > 0)
2737 			return (NULL);
2738 	}
2739 
2740 	return (NULL);
2741 }
2742 
2743 /*
2744  * bridge_rtnode_insert:
2745  *
2746  *	Insert the specified bridge node into the route table.  We
2747  *	assume the entry is not already in the table.
2748  */
2749 static int
2750 bridge_rtnode_insert(struct bridge_softc *sc, struct bridge_rtnode *brt)
2751 {
2752 	struct bridge_rtnode *lbrt;
2753 	uint32_t hash;
2754 	int dir;
2755 
2756 	BRIDGE_LOCK_ASSERT(sc);
2757 
2758 	hash = bridge_rthash(sc, brt->brt_addr);
2759 
2760 	lbrt = LIST_FIRST(&sc->sc_rthash[hash]);
2761 	if (lbrt == NULL) {
2762 		LIST_INSERT_HEAD(&sc->sc_rthash[hash], brt, brt_hash);
2763 		goto out;
2764 	}
2765 
2766 	do {
2767 		dir = bridge_rtnode_addr_cmp(brt->brt_addr, lbrt->brt_addr);
2768 		if (dir == 0 && brt->brt_vlan == lbrt->brt_vlan)
2769 			return (EEXIST);
2770 		if (dir > 0) {
2771 			LIST_INSERT_BEFORE(lbrt, brt, brt_hash);
2772 			goto out;
2773 		}
2774 		if (LIST_NEXT(lbrt, brt_hash) == NULL) {
2775 			LIST_INSERT_AFTER(lbrt, brt, brt_hash);
2776 			goto out;
2777 		}
2778 		lbrt = LIST_NEXT(lbrt, brt_hash);
2779 	} while (lbrt != NULL);
2780 
2781 #ifdef DIAGNOSTIC
2782 	panic("bridge_rtnode_insert: impossible");
2783 #endif
2784 
2785 out:
2786 	LIST_INSERT_HEAD(&sc->sc_rtlist, brt, brt_list);
2787 	sc->sc_brtcnt++;
2788 
2789 	return (0);
2790 }
2791 
2792 /*
2793  * bridge_rtnode_destroy:
2794  *
2795  *	Destroy a bridge rtnode.
2796  */
2797 static void
2798 bridge_rtnode_destroy(struct bridge_softc *sc, struct bridge_rtnode *brt)
2799 {
2800 	BRIDGE_LOCK_ASSERT(sc);
2801 
2802 	LIST_REMOVE(brt, brt_hash);
2803 
2804 	LIST_REMOVE(brt, brt_list);
2805 	sc->sc_brtcnt--;
2806 	brt->brt_dst->bif_addrcnt--;
2807 	uma_zfree(bridge_rtnode_zone, brt);
2808 }
2809 
2810 /*
2811  * bridge_rtable_expire:
2812  *
2813  *	Set the expiry time for all routes on an interface.
2814  */
2815 static void
2816 bridge_rtable_expire(struct ifnet *ifp, int age)
2817 {
2818 	struct bridge_softc *sc = ifp->if_bridge;
2819 	struct bridge_rtnode *brt;
2820 
2821 	BRIDGE_LOCK(sc);
2822 
2823 	/*
2824 	 * If the age is zero then flush, otherwise set all the expiry times to
2825 	 * age for the interface
2826 	 */
2827 	if (age == 0)
2828 		bridge_rtdelete(sc, ifp, IFBF_FLUSHDYN);
2829 	else {
2830 		LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) {
2831 			/* Cap the expiry time to 'age' */
2832 			if (brt->brt_ifp == ifp &&
2833 			    brt->brt_expire > time_uptime + age &&
2834 			    (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2835 				brt->brt_expire = time_uptime + age;
2836 		}
2837 	}
2838 	BRIDGE_UNLOCK(sc);
2839 }
2840 
2841 /*
2842  * bridge_state_change:
2843  *
2844  *	Callback from the bridgestp code when a port changes states.
2845  */
2846 static void
2847 bridge_state_change(struct ifnet *ifp, int state)
2848 {
2849 	struct bridge_softc *sc = ifp->if_bridge;
2850 	static const char *stpstates[] = {
2851 		"disabled",
2852 		"listening",
2853 		"learning",
2854 		"forwarding",
2855 		"blocking",
2856 		"discarding"
2857 	};
2858 
2859 	if (log_stp)
2860 		log(LOG_NOTICE, "%s: state changed to %s on %s\n",
2861 		    sc->sc_ifp->if_xname, stpstates[state], ifp->if_xname);
2862 }
2863 
2864 /*
2865  * Send bridge packets through pfil if they are one of the types pfil can deal
2866  * with, or if they are ARP or REVARP.  (pfil will pass ARP and REVARP without
2867  * question.) If *bifp or *ifp are NULL then packet filtering is skipped for
2868  * that interface.
2869  */
2870 static int
2871 bridge_pfil(struct mbuf **mp, struct ifnet *bifp, struct ifnet *ifp, int dir)
2872 {
2873 	int snap, error, i, hlen;
2874 	struct ether_header *eh1, eh2;
2875 	struct ip_fw_args args;
2876 	struct ip *ip;
2877 	struct llc llc1;
2878 	u_int16_t ether_type;
2879 
2880 	snap = 0;
2881 	error = -1;	/* Default error if not error == 0 */
2882 
2883 #if 0
2884 	/* we may return with the IP fields swapped, ensure its not shared */
2885 	KASSERT(M_WRITABLE(*mp), ("%s: modifying a shared mbuf", __func__));
2886 #endif
2887 
2888 	if (pfil_bridge == 0 && pfil_member == 0 && pfil_ipfw == 0)
2889 		return (0); /* filtering is disabled */
2890 
2891 	i = min((*mp)->m_pkthdr.len, max_protohdr);
2892 	if ((*mp)->m_len < i) {
2893 	    *mp = m_pullup(*mp, i);
2894 	    if (*mp == NULL) {
2895 		printf("%s: m_pullup failed\n", __func__);
2896 		return (-1);
2897 	    }
2898 	}
2899 
2900 	eh1 = mtod(*mp, struct ether_header *);
2901 	ether_type = ntohs(eh1->ether_type);
2902 
2903 	/*
2904 	 * Check for SNAP/LLC.
2905 	 */
2906 	if (ether_type < ETHERMTU) {
2907 		struct llc *llc2 = (struct llc *)(eh1 + 1);
2908 
2909 		if ((*mp)->m_len >= ETHER_HDR_LEN + 8 &&
2910 		    llc2->llc_dsap == LLC_SNAP_LSAP &&
2911 		    llc2->llc_ssap == LLC_SNAP_LSAP &&
2912 		    llc2->llc_control == LLC_UI) {
2913 			ether_type = htons(llc2->llc_un.type_snap.ether_type);
2914 			snap = 1;
2915 		}
2916 	}
2917 
2918 	/*
2919 	 * If we're trying to filter bridge traffic, don't look at anything
2920 	 * other than IP and ARP traffic.  If the filter doesn't understand
2921 	 * IPv6, don't allow IPv6 through the bridge either.  This is lame
2922 	 * since if we really wanted, say, an AppleTalk filter, we are hosed,
2923 	 * but of course we don't have an AppleTalk filter to begin with.
2924 	 * (Note that since pfil doesn't understand ARP it will pass *ALL*
2925 	 * ARP traffic.)
2926 	 */
2927 	switch (ether_type) {
2928 		case ETHERTYPE_ARP:
2929 		case ETHERTYPE_REVARP:
2930 			if (pfil_ipfw_arp == 0)
2931 				return (0); /* Automatically pass */
2932 			break;
2933 
2934 		case ETHERTYPE_IP:
2935 #ifdef INET6
2936 		case ETHERTYPE_IPV6:
2937 #endif /* INET6 */
2938 			break;
2939 		default:
2940 			/*
2941 			 * Check to see if the user wants to pass non-ip
2942 			 * packets, these will not be checked by pfil(9) and
2943 			 * passed unconditionally so the default is to drop.
2944 			 */
2945 			if (pfil_onlyip)
2946 				goto bad;
2947 	}
2948 
2949 	/* Strip off the Ethernet header and keep a copy. */
2950 	m_copydata(*mp, 0, ETHER_HDR_LEN, (caddr_t) &eh2);
2951 	m_adj(*mp, ETHER_HDR_LEN);
2952 
2953 	/* Strip off snap header, if present */
2954 	if (snap) {
2955 		m_copydata(*mp, 0, sizeof(struct llc), (caddr_t) &llc1);
2956 		m_adj(*mp, sizeof(struct llc));
2957 	}
2958 
2959 	/*
2960 	 * Check the IP header for alignment and errors
2961 	 */
2962 	if (dir == PFIL_IN) {
2963 		switch (ether_type) {
2964 			case ETHERTYPE_IP:
2965 				error = bridge_ip_checkbasic(mp);
2966 				break;
2967 #ifdef INET6
2968 			case ETHERTYPE_IPV6:
2969 				error = bridge_ip6_checkbasic(mp);
2970 				break;
2971 #endif /* INET6 */
2972 			default:
2973 				error = 0;
2974 		}
2975 		if (error)
2976 			goto bad;
2977 	}
2978 
2979 	if (IPFW_LOADED && pfil_ipfw != 0 && dir == PFIL_OUT && ifp != NULL) {
2980 		error = -1;
2981 		args.rule = ip_dn_claim_rule(*mp);
2982 		if (args.rule != NULL && fw_one_pass)
2983 			goto ipfwpass; /* packet already partially processed */
2984 
2985 		args.m = *mp;
2986 		args.oif = ifp;
2987 		args.next_hop = NULL;
2988 		args.eh = &eh2;
2989 		args.inp = NULL;	/* used by ipfw uid/gid/jail rules */
2990 		i = ip_fw_chk_ptr(&args);
2991 		*mp = args.m;
2992 
2993 		if (*mp == NULL)
2994 			return (error);
2995 
2996 		if (DUMMYNET_LOADED && (i == IP_FW_DUMMYNET)) {
2997 
2998 			/* put the Ethernet header back on */
2999 			M_PREPEND(*mp, ETHER_HDR_LEN, M_DONTWAIT);
3000 			if (*mp == NULL)
3001 				return (error);
3002 			bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN);
3003 
3004 			/*
3005 			 * Pass the pkt to dummynet, which consumes it. The
3006 			 * packet will return to us via bridge_dummynet().
3007 			 */
3008 			args.oif = ifp;
3009 			ip_dn_io_ptr(mp, DN_TO_IFB_FWD, &args);
3010 			return (error);
3011 		}
3012 
3013 		if (i != IP_FW_PASS) /* drop */
3014 			goto bad;
3015 	}
3016 
3017 ipfwpass:
3018 	error = 0;
3019 
3020 	/*
3021 	 * Run the packet through pfil
3022 	 */
3023 	switch (ether_type) {
3024 	case ETHERTYPE_IP:
3025 		/*
3026 		 * before calling the firewall, swap fields the same as
3027 		 * IP does. here we assume the header is contiguous
3028 		 */
3029 		ip = mtod(*mp, struct ip *);
3030 
3031 		ip->ip_len = ntohs(ip->ip_len);
3032 		ip->ip_off = ntohs(ip->ip_off);
3033 
3034 		/*
3035 		 * Run pfil on the member interface and the bridge, both can
3036 		 * be skipped by clearing pfil_member or pfil_bridge.
3037 		 *
3038 		 * Keep the order:
3039 		 *   in_if -> bridge_if -> out_if
3040 		 */
3041 		if (pfil_bridge && dir == PFIL_OUT && bifp != NULL)
3042 			error = pfil_run_hooks(&inet_pfil_hook, mp, bifp,
3043 					dir, NULL);
3044 
3045 		if (*mp == NULL || error != 0) /* filter may consume */
3046 			break;
3047 
3048 		if (pfil_member && ifp != NULL)
3049 			error = pfil_run_hooks(&inet_pfil_hook, mp, ifp,
3050 					dir, NULL);
3051 
3052 		if (*mp == NULL || error != 0) /* filter may consume */
3053 			break;
3054 
3055 		if (pfil_bridge && dir == PFIL_IN && bifp != NULL)
3056 			error = pfil_run_hooks(&inet_pfil_hook, mp, bifp,
3057 					dir, NULL);
3058 
3059 		if (*mp == NULL || error != 0) /* filter may consume */
3060 			break;
3061 
3062 		/* check if we need to fragment the packet */
3063 		if (pfil_member && ifp != NULL && dir == PFIL_OUT) {
3064 			i = (*mp)->m_pkthdr.len;
3065 			if (i > ifp->if_mtu) {
3066 				error = bridge_fragment(ifp, *mp, &eh2, snap,
3067 					    &llc1);
3068 				return (error);
3069 			}
3070 		}
3071 
3072 		/* Recalculate the ip checksum and restore byte ordering */
3073 		ip = mtod(*mp, struct ip *);
3074 		hlen = ip->ip_hl << 2;
3075 		if (hlen < sizeof(struct ip))
3076 			goto bad;
3077 		if (hlen > (*mp)->m_len) {
3078 			if ((*mp = m_pullup(*mp, hlen)) == 0)
3079 				goto bad;
3080 			ip = mtod(*mp, struct ip *);
3081 			if (ip == NULL)
3082 				goto bad;
3083 		}
3084 		ip->ip_len = htons(ip->ip_len);
3085 		ip->ip_off = htons(ip->ip_off);
3086 		ip->ip_sum = 0;
3087 		if (hlen == sizeof(struct ip))
3088 			ip->ip_sum = in_cksum_hdr(ip);
3089 		else
3090 			ip->ip_sum = in_cksum(*mp, hlen);
3091 
3092 		break;
3093 #ifdef INET6
3094 	case ETHERTYPE_IPV6:
3095 		if (pfil_bridge && dir == PFIL_OUT && bifp != NULL)
3096 			error = pfil_run_hooks(&inet6_pfil_hook, mp, bifp,
3097 					dir, NULL);
3098 
3099 		if (*mp == NULL || error != 0) /* filter may consume */
3100 			break;
3101 
3102 		if (pfil_member && ifp != NULL)
3103 			error = pfil_run_hooks(&inet6_pfil_hook, mp, ifp,
3104 					dir, NULL);
3105 
3106 		if (*mp == NULL || error != 0) /* filter may consume */
3107 			break;
3108 
3109 		if (pfil_bridge && dir == PFIL_IN && bifp != NULL)
3110 			error = pfil_run_hooks(&inet6_pfil_hook, mp, bifp,
3111 					dir, NULL);
3112 		break;
3113 #endif
3114 	default:
3115 		error = 0;
3116 		break;
3117 	}
3118 
3119 	if (*mp == NULL)
3120 		return (error);
3121 	if (error != 0)
3122 		goto bad;
3123 
3124 	error = -1;
3125 
3126 	/*
3127 	 * Finally, put everything back the way it was and return
3128 	 */
3129 	if (snap) {
3130 		M_PREPEND(*mp, sizeof(struct llc), M_DONTWAIT);
3131 		if (*mp == NULL)
3132 			return (error);
3133 		bcopy(&llc1, mtod(*mp, caddr_t), sizeof(struct llc));
3134 	}
3135 
3136 	M_PREPEND(*mp, ETHER_HDR_LEN, M_DONTWAIT);
3137 	if (*mp == NULL)
3138 		return (error);
3139 	bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN);
3140 
3141 	return (0);
3142 
3143 bad:
3144 	m_freem(*mp);
3145 	*mp = NULL;
3146 	return (error);
3147 }
3148 
3149 /*
3150  * Perform basic checks on header size since
3151  * pfil assumes ip_input has already processed
3152  * it for it.  Cut-and-pasted from ip_input.c.
3153  * Given how simple the IPv6 version is,
3154  * does the IPv4 version really need to be
3155  * this complicated?
3156  *
3157  * XXX Should we update ipstat here, or not?
3158  * XXX Right now we update ipstat but not
3159  * XXX csum_counter.
3160  */
3161 static int
3162 bridge_ip_checkbasic(struct mbuf **mp)
3163 {
3164 	struct mbuf *m = *mp;
3165 	struct ip *ip;
3166 	int len, hlen;
3167 	u_short sum;
3168 
3169 	if (*mp == NULL)
3170 		return (-1);
3171 
3172 	if (IP_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) {
3173 		if ((m = m_copyup(m, sizeof(struct ip),
3174 			(max_linkhdr + 3) & ~3)) == NULL) {
3175 			/* XXXJRT new stat, please */
3176 			ipstat.ips_toosmall++;
3177 			goto bad;
3178 		}
3179 	} else if (__predict_false(m->m_len < sizeof (struct ip))) {
3180 		if ((m = m_pullup(m, sizeof (struct ip))) == NULL) {
3181 			ipstat.ips_toosmall++;
3182 			goto bad;
3183 		}
3184 	}
3185 	ip = mtod(m, struct ip *);
3186 	if (ip == NULL) goto bad;
3187 
3188 	if (ip->ip_v != IPVERSION) {
3189 		ipstat.ips_badvers++;
3190 		goto bad;
3191 	}
3192 	hlen = ip->ip_hl << 2;
3193 	if (hlen < sizeof(struct ip)) { /* minimum header length */
3194 		ipstat.ips_badhlen++;
3195 		goto bad;
3196 	}
3197 	if (hlen > m->m_len) {
3198 		if ((m = m_pullup(m, hlen)) == 0) {
3199 			ipstat.ips_badhlen++;
3200 			goto bad;
3201 		}
3202 		ip = mtod(m, struct ip *);
3203 		if (ip == NULL) goto bad;
3204 	}
3205 
3206 	if (m->m_pkthdr.csum_flags & CSUM_IP_CHECKED) {
3207 		sum = !(m->m_pkthdr.csum_flags & CSUM_IP_VALID);
3208 	} else {
3209 		if (hlen == sizeof(struct ip)) {
3210 			sum = in_cksum_hdr(ip);
3211 		} else {
3212 			sum = in_cksum(m, hlen);
3213 		}
3214 	}
3215 	if (sum) {
3216 		ipstat.ips_badsum++;
3217 		goto bad;
3218 	}
3219 
3220 	/* Retrieve the packet length. */
3221 	len = ntohs(ip->ip_len);
3222 
3223 	/*
3224 	 * Check for additional length bogosity
3225 	 */
3226 	if (len < hlen) {
3227 		ipstat.ips_badlen++;
3228 		goto bad;
3229 	}
3230 
3231 	/*
3232 	 * Check that the amount of data in the buffers
3233 	 * is as at least much as the IP header would have us expect.
3234 	 * Drop packet if shorter than we expect.
3235 	 */
3236 	if (m->m_pkthdr.len < len) {
3237 		ipstat.ips_tooshort++;
3238 		goto bad;
3239 	}
3240 
3241 	/* Checks out, proceed */
3242 	*mp = m;
3243 	return (0);
3244 
3245 bad:
3246 	*mp = m;
3247 	return (-1);
3248 }
3249 
3250 #ifdef INET6
3251 /*
3252  * Same as above, but for IPv6.
3253  * Cut-and-pasted from ip6_input.c.
3254  * XXX Should we update ip6stat, or not?
3255  */
3256 static int
3257 bridge_ip6_checkbasic(struct mbuf **mp)
3258 {
3259 	struct mbuf *m = *mp;
3260 	struct ip6_hdr *ip6;
3261 
3262 	/*
3263 	 * If the IPv6 header is not aligned, slurp it up into a new
3264 	 * mbuf with space for link headers, in the event we forward
3265 	 * it.  Otherwise, if it is aligned, make sure the entire base
3266 	 * IPv6 header is in the first mbuf of the chain.
3267 	 */
3268 	if (IP6_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) {
3269 		struct ifnet *inifp = m->m_pkthdr.rcvif;
3270 		if ((m = m_copyup(m, sizeof(struct ip6_hdr),
3271 			    (max_linkhdr + 3) & ~3)) == NULL) {
3272 			/* XXXJRT new stat, please */
3273 			ip6stat.ip6s_toosmall++;
3274 			in6_ifstat_inc(inifp, ifs6_in_hdrerr);
3275 			goto bad;
3276 		}
3277 	} else if (__predict_false(m->m_len < sizeof(struct ip6_hdr))) {
3278 		struct ifnet *inifp = m->m_pkthdr.rcvif;
3279 		if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) {
3280 			ip6stat.ip6s_toosmall++;
3281 			in6_ifstat_inc(inifp, ifs6_in_hdrerr);
3282 			goto bad;
3283 		}
3284 	}
3285 
3286 	ip6 = mtod(m, struct ip6_hdr *);
3287 
3288 	if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) {
3289 		ip6stat.ip6s_badvers++;
3290 		in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr);
3291 		goto bad;
3292 	}
3293 
3294 	/* Checks out, proceed */
3295 	*mp = m;
3296 	return (0);
3297 
3298 bad:
3299 	*mp = m;
3300 	return (-1);
3301 }
3302 #endif /* INET6 */
3303 
3304 /*
3305  * bridge_fragment:
3306  *
3307  *	Return a fragmented mbuf chain.
3308  */
3309 static int
3310 bridge_fragment(struct ifnet *ifp, struct mbuf *m, struct ether_header *eh,
3311     int snap, struct llc *llc)
3312 {
3313 	struct mbuf *m0;
3314 	struct ip *ip;
3315 	int error = -1;
3316 
3317 	if (m->m_len < sizeof(struct ip) &&
3318 	    (m = m_pullup(m, sizeof(struct ip))) == NULL)
3319 		goto out;
3320 	ip = mtod(m, struct ip *);
3321 
3322 	error = ip_fragment(ip, &m, ifp->if_mtu, ifp->if_hwassist,
3323 		    CSUM_DELAY_IP);
3324 	if (error)
3325 		goto out;
3326 
3327 	/* walk the chain and re-add the Ethernet header */
3328 	for (m0 = m; m0; m0 = m0->m_nextpkt) {
3329 		if (error == 0) {
3330 			if (snap) {
3331 				M_PREPEND(m0, sizeof(struct llc), M_DONTWAIT);
3332 				if (m0 == NULL) {
3333 					error = ENOBUFS;
3334 					continue;
3335 				}
3336 				bcopy(llc, mtod(m0, caddr_t),
3337 				    sizeof(struct llc));
3338 			}
3339 			M_PREPEND(m0, ETHER_HDR_LEN, M_DONTWAIT);
3340 			if (m0 == NULL) {
3341 				error = ENOBUFS;
3342 				continue;
3343 			}
3344 			bcopy(eh, mtod(m0, caddr_t), ETHER_HDR_LEN);
3345 		} else
3346 			m_freem(m);
3347 	}
3348 
3349 	if (error == 0)
3350 		ipstat.ips_fragmented++;
3351 
3352 	return (error);
3353 
3354 out:
3355 	if (m != NULL)
3356 		m_freem(m);
3357 	return (error);
3358 }
3359