xref: /freebsd/sys/net/if_bridge.c (revision e2d997d1cbb9b4d651bd35e7d16d316ecd0981ad)
1 /*	$NetBSD: if_bridge.c,v 1.31 2005/06/01 19:45:34 jdc Exp $	*/
2 
3 /*-
4  * SPDX-License-Identifier: BSD-4-Clause
5  *
6  * Copyright 2001 Wasabi Systems, Inc.
7  * All rights reserved.
8  *
9  * Written by Jason R. Thorpe for Wasabi Systems, Inc.
10  *
11  * Redistribution and use in source and binary forms, with or without
12  * modification, are permitted provided that the following conditions
13  * are met:
14  * 1. Redistributions of source code must retain the above copyright
15  *    notice, this list of conditions and the following disclaimer.
16  * 2. Redistributions in binary form must reproduce the above copyright
17  *    notice, this list of conditions and the following disclaimer in the
18  *    documentation and/or other materials provided with the distribution.
19  * 3. All advertising materials mentioning features or use of this software
20  *    must display the following acknowledgement:
21  *	This product includes software developed for the NetBSD Project by
22  *	Wasabi Systems, Inc.
23  * 4. The name of Wasabi Systems, Inc. may not be used to endorse
24  *    or promote products derived from this software without specific prior
25  *    written permission.
26  *
27  * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND
28  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
29  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
30  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL WASABI SYSTEMS, INC
31  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37  * POSSIBILITY OF SUCH DAMAGE.
38  */
39 
40 /*
41  * Copyright (c) 1999, 2000 Jason L. Wright (jason@thought.net)
42  * All rights reserved.
43  *
44  * Redistribution and use in source and binary forms, with or without
45  * modification, are permitted provided that the following conditions
46  * are met:
47  * 1. Redistributions of source code must retain the above copyright
48  *    notice, this list of conditions and the following disclaimer.
49  * 2. Redistributions in binary form must reproduce the above copyright
50  *    notice, this list of conditions and the following disclaimer in the
51  *    documentation and/or other materials provided with the distribution.
52  *
53  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
54  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
55  * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
56  * DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT,
57  * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
58  * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
59  * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
60  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
61  * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
62  * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
63  * POSSIBILITY OF SUCH DAMAGE.
64  *
65  * OpenBSD: if_bridge.c,v 1.60 2001/06/15 03:38:33 itojun Exp
66  */
67 
68 /*
69  * Network interface bridge support.
70  *
71  * TODO:
72  *
73  *	- Currently only supports Ethernet-like interfaces (Ethernet,
74  *	  802.11, VLANs on Ethernet, etc.)  Figure out a nice way
75  *	  to bridge other types of interfaces (maybe consider
76  *	  heterogeneous bridges).
77  */
78 
79 #include <sys/cdefs.h>
80 __FBSDID("$FreeBSD$");
81 
82 #include "opt_inet.h"
83 #include "opt_inet6.h"
84 
85 #include <sys/param.h>
86 #include <sys/eventhandler.h>
87 #include <sys/mbuf.h>
88 #include <sys/malloc.h>
89 #include <sys/protosw.h>
90 #include <sys/systm.h>
91 #include <sys/jail.h>
92 #include <sys/time.h>
93 #include <sys/socket.h> /* for net/if.h */
94 #include <sys/sockio.h>
95 #include <sys/ctype.h>  /* string functions */
96 #include <sys/kernel.h>
97 #include <sys/random.h>
98 #include <sys/syslog.h>
99 #include <sys/sysctl.h>
100 #include <vm/uma.h>
101 #include <sys/module.h>
102 #include <sys/priv.h>
103 #include <sys/proc.h>
104 #include <sys/lock.h>
105 #include <sys/mutex.h>
106 
107 #include <net/bpf.h>
108 #include <net/if.h>
109 #include <net/if_clone.h>
110 #include <net/if_dl.h>
111 #include <net/if_types.h>
112 #include <net/if_var.h>
113 #include <net/if_private.h>
114 #include <net/pfil.h>
115 #include <net/vnet.h>
116 
117 #include <netinet/in.h>
118 #include <netinet/in_systm.h>
119 #include <netinet/in_var.h>
120 #include <netinet/ip.h>
121 #include <netinet/ip_var.h>
122 #ifdef INET6
123 #include <netinet/ip6.h>
124 #include <netinet6/ip6_var.h>
125 #include <netinet6/in6_ifattach.h>
126 #endif
127 #if defined(INET) || defined(INET6)
128 #include <netinet/ip_carp.h>
129 #endif
130 #include <machine/in_cksum.h>
131 #include <netinet/if_ether.h>
132 #include <net/bridgestp.h>
133 #include <net/if_bridgevar.h>
134 #include <net/if_llc.h>
135 #include <net/if_vlan_var.h>
136 
137 #include <net/route.h>
138 
139 /*
140  * Size of the route hash table.  Must be a power of two.
141  */
142 #ifndef BRIDGE_RTHASH_SIZE
143 #define	BRIDGE_RTHASH_SIZE		1024
144 #endif
145 
146 #define	BRIDGE_RTHASH_MASK		(BRIDGE_RTHASH_SIZE - 1)
147 
148 /*
149  * Default maximum number of addresses to cache.
150  */
151 #ifndef BRIDGE_RTABLE_MAX
152 #define	BRIDGE_RTABLE_MAX		2000
153 #endif
154 
155 /*
156  * Timeout (in seconds) for entries learned dynamically.
157  */
158 #ifndef BRIDGE_RTABLE_TIMEOUT
159 #define	BRIDGE_RTABLE_TIMEOUT		(20 * 60)	/* same as ARP */
160 #endif
161 
162 /*
163  * Number of seconds between walks of the route list.
164  */
165 #ifndef BRIDGE_RTABLE_PRUNE_PERIOD
166 #define	BRIDGE_RTABLE_PRUNE_PERIOD	(5 * 60)
167 #endif
168 
169 /*
170  * List of capabilities to possibly mask on the member interface.
171  */
172 #define	BRIDGE_IFCAPS_MASK		(IFCAP_TOE|IFCAP_TSO|IFCAP_TXCSUM|\
173 					 IFCAP_TXCSUM_IPV6)
174 
175 /*
176  * List of capabilities to strip
177  */
178 #define	BRIDGE_IFCAPS_STRIP		IFCAP_LRO
179 
180 /*
181  * Bridge locking
182  *
183  * The bridge relies heavily on the epoch(9) system to protect its data
184  * structures. This means we can safely use CK_LISTs while in NET_EPOCH, but we
185  * must ensure there is only one writer at a time.
186  *
187  * That is: for read accesses we only need to be in NET_EPOCH, but for write
188  * accesses we must hold:
189  *
190  *  - BRIDGE_RT_LOCK, for any change to bridge_rtnodes
191  *  - BRIDGE_LOCK, for any other change
192  *
193  * The BRIDGE_LOCK is a sleepable lock, because it is held across ioctl()
194  * calls to bridge member interfaces and these ioctl()s can sleep.
195  * The BRIDGE_RT_LOCK is a non-sleepable mutex, because it is sometimes
196  * required while we're in NET_EPOCH and then we're not allowed to sleep.
197  */
198 #define BRIDGE_LOCK_INIT(_sc)		do {			\
199 	sx_init(&(_sc)->sc_sx, "if_bridge");			\
200 	mtx_init(&(_sc)->sc_rt_mtx, "if_bridge rt", NULL, MTX_DEF);	\
201 } while (0)
202 #define BRIDGE_LOCK_DESTROY(_sc)	do {	\
203 	sx_destroy(&(_sc)->sc_sx);		\
204 	mtx_destroy(&(_sc)->sc_rt_mtx);		\
205 } while (0)
206 #define BRIDGE_LOCK(_sc)		sx_xlock(&(_sc)->sc_sx)
207 #define BRIDGE_UNLOCK(_sc)		sx_xunlock(&(_sc)->sc_sx)
208 #define BRIDGE_LOCK_ASSERT(_sc)		sx_assert(&(_sc)->sc_sx, SX_XLOCKED)
209 #define BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(_sc)	\
210 	    MPASS(in_epoch(net_epoch_preempt) || sx_xlocked(&(_sc)->sc_sx))
211 #define BRIDGE_UNLOCK_ASSERT(_sc)	sx_assert(&(_sc)->sc_sx, SX_UNLOCKED)
212 #define BRIDGE_RT_LOCK(_sc)		mtx_lock(&(_sc)->sc_rt_mtx)
213 #define BRIDGE_RT_UNLOCK(_sc)		mtx_unlock(&(_sc)->sc_rt_mtx)
214 #define BRIDGE_RT_LOCK_ASSERT(_sc)	mtx_assert(&(_sc)->sc_rt_mtx, MA_OWNED)
215 #define BRIDGE_RT_LOCK_OR_NET_EPOCH_ASSERT(_sc)	\
216 	    MPASS(in_epoch(net_epoch_preempt) || mtx_owned(&(_sc)->sc_rt_mtx))
217 
218 /*
219  * Bridge interface list entry.
220  */
221 struct bridge_iflist {
222 	CK_LIST_ENTRY(bridge_iflist) bif_next;
223 	struct ifnet		*bif_ifp;	/* member if */
224 	struct bstp_port	bif_stp;	/* STP state */
225 	uint32_t		bif_flags;	/* member if flags */
226 	int			bif_savedcaps;	/* saved capabilities */
227 	uint32_t		bif_addrmax;	/* max # of addresses */
228 	uint32_t		bif_addrcnt;	/* cur. # of addresses */
229 	uint32_t		bif_addrexceeded;/* # of address violations */
230 	struct epoch_context	bif_epoch_ctx;
231 };
232 
233 /*
234  * Bridge route node.
235  */
236 struct bridge_rtnode {
237 	CK_LIST_ENTRY(bridge_rtnode) brt_hash;	/* hash table linkage */
238 	CK_LIST_ENTRY(bridge_rtnode) brt_list;	/* list linkage */
239 	struct bridge_iflist	*brt_dst;	/* destination if */
240 	unsigned long		brt_expire;	/* expiration time */
241 	uint8_t			brt_flags;	/* address flags */
242 	uint8_t			brt_addr[ETHER_ADDR_LEN];
243 	uint16_t		brt_vlan;	/* vlan id */
244 	struct	vnet		*brt_vnet;
245 	struct	epoch_context	brt_epoch_ctx;
246 };
247 #define	brt_ifp			brt_dst->bif_ifp
248 
249 /*
250  * Software state for each bridge.
251  */
252 struct bridge_softc {
253 	struct ifnet		*sc_ifp;	/* make this an interface */
254 	LIST_ENTRY(bridge_softc) sc_list;
255 	struct sx		sc_sx;
256 	struct mtx		sc_rt_mtx;
257 	uint32_t		sc_brtmax;	/* max # of addresses */
258 	uint32_t		sc_brtcnt;	/* cur. # of addresses */
259 	uint32_t		sc_brttimeout;	/* rt timeout in seconds */
260 	struct callout		sc_brcallout;	/* bridge callout */
261 	CK_LIST_HEAD(, bridge_iflist) sc_iflist;	/* member interface list */
262 	CK_LIST_HEAD(, bridge_rtnode) *sc_rthash;	/* our forwarding table */
263 	CK_LIST_HEAD(, bridge_rtnode) sc_rtlist;	/* list version of above */
264 	uint32_t		sc_rthash_key;	/* key for hash */
265 	CK_LIST_HEAD(, bridge_iflist) sc_spanlist;	/* span ports list */
266 	struct bstp_state	sc_stp;		/* STP state */
267 	uint32_t		sc_brtexceeded;	/* # of cache drops */
268 	struct ifnet		*sc_ifaddr;	/* member mac copied from */
269 	struct ether_addr	sc_defaddr;	/* Default MAC address */
270 	struct epoch_context	sc_epoch_ctx;
271 };
272 
273 VNET_DEFINE_STATIC(struct sx, bridge_list_sx);
274 #define	V_bridge_list_sx	VNET(bridge_list_sx)
275 static eventhandler_tag bridge_detach_cookie;
276 
277 int	bridge_rtable_prune_period = BRIDGE_RTABLE_PRUNE_PERIOD;
278 
279 VNET_DEFINE_STATIC(uma_zone_t, bridge_rtnode_zone);
280 #define	V_bridge_rtnode_zone	VNET(bridge_rtnode_zone)
281 
282 static int	bridge_clone_create(struct if_clone *, char *, size_t,
283 		    struct ifc_data *, struct ifnet **);
284 static int	bridge_clone_destroy(struct if_clone *, struct ifnet *, uint32_t);
285 
286 static int	bridge_ioctl(struct ifnet *, u_long, caddr_t);
287 static void	bridge_mutecaps(struct bridge_softc *);
288 static void	bridge_set_ifcap(struct bridge_softc *, struct bridge_iflist *,
289 		    int);
290 static void	bridge_ifdetach(void *arg __unused, struct ifnet *);
291 static void	bridge_init(void *);
292 static void	bridge_dummynet(struct mbuf *, struct ifnet *);
293 static void	bridge_stop(struct ifnet *, int);
294 static int	bridge_transmit(struct ifnet *, struct mbuf *);
295 #ifdef ALTQ
296 static void	bridge_altq_start(if_t);
297 static int	bridge_altq_transmit(if_t, struct mbuf *);
298 #endif
299 static void	bridge_qflush(struct ifnet *);
300 static struct mbuf *bridge_input(struct ifnet *, struct mbuf *);
301 static int	bridge_output(struct ifnet *, struct mbuf *, struct sockaddr *,
302 		    struct rtentry *);
303 static int	bridge_enqueue(struct bridge_softc *, struct ifnet *,
304 		    struct mbuf *);
305 static void	bridge_rtdelete(struct bridge_softc *, struct ifnet *ifp, int);
306 
307 static void	bridge_forward(struct bridge_softc *, struct bridge_iflist *,
308 		    struct mbuf *m);
309 
310 static void	bridge_timer(void *);
311 
312 static void	bridge_broadcast(struct bridge_softc *, struct ifnet *,
313 		    struct mbuf *, int);
314 static void	bridge_span(struct bridge_softc *, struct mbuf *);
315 
316 static int	bridge_rtupdate(struct bridge_softc *, const uint8_t *,
317 		    uint16_t, struct bridge_iflist *, int, uint8_t);
318 static struct ifnet *bridge_rtlookup(struct bridge_softc *, const uint8_t *,
319 		    uint16_t);
320 static void	bridge_rttrim(struct bridge_softc *);
321 static void	bridge_rtage(struct bridge_softc *);
322 static void	bridge_rtflush(struct bridge_softc *, int);
323 static int	bridge_rtdaddr(struct bridge_softc *, const uint8_t *,
324 		    uint16_t);
325 
326 static void	bridge_rtable_init(struct bridge_softc *);
327 static void	bridge_rtable_fini(struct bridge_softc *);
328 
329 static int	bridge_rtnode_addr_cmp(const uint8_t *, const uint8_t *);
330 static struct bridge_rtnode *bridge_rtnode_lookup(struct bridge_softc *,
331 		    const uint8_t *, uint16_t);
332 static int	bridge_rtnode_insert(struct bridge_softc *,
333 		    struct bridge_rtnode *);
334 static void	bridge_rtnode_destroy(struct bridge_softc *,
335 		    struct bridge_rtnode *);
336 static void	bridge_rtable_expire(struct ifnet *, int);
337 static void	bridge_state_change(struct ifnet *, int);
338 
339 static struct bridge_iflist *bridge_lookup_member(struct bridge_softc *,
340 		    const char *name);
341 static struct bridge_iflist *bridge_lookup_member_if(struct bridge_softc *,
342 		    struct ifnet *ifp);
343 static void	bridge_delete_member(struct bridge_softc *,
344 		    struct bridge_iflist *, int);
345 static void	bridge_delete_span(struct bridge_softc *,
346 		    struct bridge_iflist *);
347 
348 static int	bridge_ioctl_add(struct bridge_softc *, void *);
349 static int	bridge_ioctl_del(struct bridge_softc *, void *);
350 static int	bridge_ioctl_gifflags(struct bridge_softc *, void *);
351 static int	bridge_ioctl_sifflags(struct bridge_softc *, void *);
352 static int	bridge_ioctl_scache(struct bridge_softc *, void *);
353 static int	bridge_ioctl_gcache(struct bridge_softc *, void *);
354 static int	bridge_ioctl_gifs(struct bridge_softc *, void *);
355 static int	bridge_ioctl_rts(struct bridge_softc *, void *);
356 static int	bridge_ioctl_saddr(struct bridge_softc *, void *);
357 static int	bridge_ioctl_sto(struct bridge_softc *, void *);
358 static int	bridge_ioctl_gto(struct bridge_softc *, void *);
359 static int	bridge_ioctl_daddr(struct bridge_softc *, void *);
360 static int	bridge_ioctl_flush(struct bridge_softc *, void *);
361 static int	bridge_ioctl_gpri(struct bridge_softc *, void *);
362 static int	bridge_ioctl_spri(struct bridge_softc *, void *);
363 static int	bridge_ioctl_ght(struct bridge_softc *, void *);
364 static int	bridge_ioctl_sht(struct bridge_softc *, void *);
365 static int	bridge_ioctl_gfd(struct bridge_softc *, void *);
366 static int	bridge_ioctl_sfd(struct bridge_softc *, void *);
367 static int	bridge_ioctl_gma(struct bridge_softc *, void *);
368 static int	bridge_ioctl_sma(struct bridge_softc *, void *);
369 static int	bridge_ioctl_sifprio(struct bridge_softc *, void *);
370 static int	bridge_ioctl_sifcost(struct bridge_softc *, void *);
371 static int	bridge_ioctl_sifmaxaddr(struct bridge_softc *, void *);
372 static int	bridge_ioctl_addspan(struct bridge_softc *, void *);
373 static int	bridge_ioctl_delspan(struct bridge_softc *, void *);
374 static int	bridge_ioctl_gbparam(struct bridge_softc *, void *);
375 static int	bridge_ioctl_grte(struct bridge_softc *, void *);
376 static int	bridge_ioctl_gifsstp(struct bridge_softc *, void *);
377 static int	bridge_ioctl_sproto(struct bridge_softc *, void *);
378 static int	bridge_ioctl_stxhc(struct bridge_softc *, void *);
379 static int	bridge_pfil(struct mbuf **, struct ifnet *, struct ifnet *,
380 		    int);
381 static int	bridge_ip_checkbasic(struct mbuf **mp);
382 #ifdef INET6
383 static int	bridge_ip6_checkbasic(struct mbuf **mp);
384 #endif /* INET6 */
385 static int	bridge_fragment(struct ifnet *, struct mbuf **mp,
386 		    struct ether_header *, int, struct llc *);
387 static void	bridge_linkstate(struct ifnet *ifp);
388 static void	bridge_linkcheck(struct bridge_softc *sc);
389 
390 /* The default bridge vlan is 1 (IEEE 802.1Q-2003 Table 9-2) */
391 #define	VLANTAGOF(_m)	\
392     (_m->m_flags & M_VLANTAG) ? EVL_VLANOFTAG(_m->m_pkthdr.ether_vtag) : 1
393 
394 static struct bstp_cb_ops bridge_ops = {
395 	.bcb_state = bridge_state_change,
396 	.bcb_rtage = bridge_rtable_expire
397 };
398 
399 SYSCTL_DECL(_net_link);
400 static SYSCTL_NODE(_net_link, IFT_BRIDGE, bridge, CTLFLAG_RW | CTLFLAG_MPSAFE, 0,
401     "Bridge");
402 
403 /* only pass IP[46] packets when pfil is enabled */
404 VNET_DEFINE_STATIC(int, pfil_onlyip) = 1;
405 #define	V_pfil_onlyip	VNET(pfil_onlyip)
406 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_onlyip,
407     CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_onlyip), 0,
408     "Only pass IP packets when pfil is enabled");
409 
410 /* run pfil hooks on the bridge interface */
411 VNET_DEFINE_STATIC(int, pfil_bridge) = 0;
412 #define	V_pfil_bridge	VNET(pfil_bridge)
413 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_bridge,
414     CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_bridge), 0,
415     "Packet filter on the bridge interface");
416 
417 /* layer2 filter with ipfw */
418 VNET_DEFINE_STATIC(int, pfil_ipfw);
419 #define	V_pfil_ipfw	VNET(pfil_ipfw)
420 
421 /* layer2 ARP filter with ipfw */
422 VNET_DEFINE_STATIC(int, pfil_ipfw_arp);
423 #define	V_pfil_ipfw_arp	VNET(pfil_ipfw_arp)
424 SYSCTL_INT(_net_link_bridge, OID_AUTO, ipfw_arp,
425     CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_ipfw_arp), 0,
426     "Filter ARP packets through IPFW layer2");
427 
428 /* run pfil hooks on the member interface */
429 VNET_DEFINE_STATIC(int, pfil_member) = 0;
430 #define	V_pfil_member	VNET(pfil_member)
431 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_member,
432     CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_member), 0,
433     "Packet filter on the member interface");
434 
435 /* run pfil hooks on the physical interface for locally destined packets */
436 VNET_DEFINE_STATIC(int, pfil_local_phys);
437 #define	V_pfil_local_phys	VNET(pfil_local_phys)
438 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_local_phys,
439     CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_local_phys), 0,
440     "Packet filter on the physical interface for locally destined packets");
441 
442 /* log STP state changes */
443 VNET_DEFINE_STATIC(int, log_stp);
444 #define	V_log_stp	VNET(log_stp)
445 SYSCTL_INT(_net_link_bridge, OID_AUTO, log_stp,
446     CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(log_stp), 0,
447     "Log STP state changes");
448 
449 /* share MAC with first bridge member */
450 VNET_DEFINE_STATIC(int, bridge_inherit_mac);
451 #define	V_bridge_inherit_mac	VNET(bridge_inherit_mac)
452 SYSCTL_INT(_net_link_bridge, OID_AUTO, inherit_mac,
453     CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(bridge_inherit_mac), 0,
454     "Inherit MAC address from the first bridge member");
455 
456 VNET_DEFINE_STATIC(int, allow_llz_overlap) = 0;
457 #define	V_allow_llz_overlap	VNET(allow_llz_overlap)
458 SYSCTL_INT(_net_link_bridge, OID_AUTO, allow_llz_overlap,
459     CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(allow_llz_overlap), 0,
460     "Allow overlap of link-local scope "
461     "zones of a bridge interface and the member interfaces");
462 
463 struct bridge_control {
464 	int	(*bc_func)(struct bridge_softc *, void *);
465 	int	bc_argsize;
466 	int	bc_flags;
467 };
468 
469 #define	BC_F_COPYIN		0x01	/* copy arguments in */
470 #define	BC_F_COPYOUT		0x02	/* copy arguments out */
471 #define	BC_F_SUSER		0x04	/* do super-user check */
472 
473 static const struct bridge_control bridge_control_table[] = {
474 	{ bridge_ioctl_add,		sizeof(struct ifbreq),
475 	  BC_F_COPYIN|BC_F_SUSER },
476 	{ bridge_ioctl_del,		sizeof(struct ifbreq),
477 	  BC_F_COPYIN|BC_F_SUSER },
478 
479 	{ bridge_ioctl_gifflags,	sizeof(struct ifbreq),
480 	  BC_F_COPYIN|BC_F_COPYOUT },
481 	{ bridge_ioctl_sifflags,	sizeof(struct ifbreq),
482 	  BC_F_COPYIN|BC_F_SUSER },
483 
484 	{ bridge_ioctl_scache,		sizeof(struct ifbrparam),
485 	  BC_F_COPYIN|BC_F_SUSER },
486 	{ bridge_ioctl_gcache,		sizeof(struct ifbrparam),
487 	  BC_F_COPYOUT },
488 
489 	{ bridge_ioctl_gifs,		sizeof(struct ifbifconf),
490 	  BC_F_COPYIN|BC_F_COPYOUT },
491 	{ bridge_ioctl_rts,		sizeof(struct ifbaconf),
492 	  BC_F_COPYIN|BC_F_COPYOUT },
493 
494 	{ bridge_ioctl_saddr,		sizeof(struct ifbareq),
495 	  BC_F_COPYIN|BC_F_SUSER },
496 
497 	{ bridge_ioctl_sto,		sizeof(struct ifbrparam),
498 	  BC_F_COPYIN|BC_F_SUSER },
499 	{ bridge_ioctl_gto,		sizeof(struct ifbrparam),
500 	  BC_F_COPYOUT },
501 
502 	{ bridge_ioctl_daddr,		sizeof(struct ifbareq),
503 	  BC_F_COPYIN|BC_F_SUSER },
504 
505 	{ bridge_ioctl_flush,		sizeof(struct ifbreq),
506 	  BC_F_COPYIN|BC_F_SUSER },
507 
508 	{ bridge_ioctl_gpri,		sizeof(struct ifbrparam),
509 	  BC_F_COPYOUT },
510 	{ bridge_ioctl_spri,		sizeof(struct ifbrparam),
511 	  BC_F_COPYIN|BC_F_SUSER },
512 
513 	{ bridge_ioctl_ght,		sizeof(struct ifbrparam),
514 	  BC_F_COPYOUT },
515 	{ bridge_ioctl_sht,		sizeof(struct ifbrparam),
516 	  BC_F_COPYIN|BC_F_SUSER },
517 
518 	{ bridge_ioctl_gfd,		sizeof(struct ifbrparam),
519 	  BC_F_COPYOUT },
520 	{ bridge_ioctl_sfd,		sizeof(struct ifbrparam),
521 	  BC_F_COPYIN|BC_F_SUSER },
522 
523 	{ bridge_ioctl_gma,		sizeof(struct ifbrparam),
524 	  BC_F_COPYOUT },
525 	{ bridge_ioctl_sma,		sizeof(struct ifbrparam),
526 	  BC_F_COPYIN|BC_F_SUSER },
527 
528 	{ bridge_ioctl_sifprio,		sizeof(struct ifbreq),
529 	  BC_F_COPYIN|BC_F_SUSER },
530 
531 	{ bridge_ioctl_sifcost,		sizeof(struct ifbreq),
532 	  BC_F_COPYIN|BC_F_SUSER },
533 
534 	{ bridge_ioctl_addspan,		sizeof(struct ifbreq),
535 	  BC_F_COPYIN|BC_F_SUSER },
536 	{ bridge_ioctl_delspan,		sizeof(struct ifbreq),
537 	  BC_F_COPYIN|BC_F_SUSER },
538 
539 	{ bridge_ioctl_gbparam,		sizeof(struct ifbropreq),
540 	  BC_F_COPYOUT },
541 
542 	{ bridge_ioctl_grte,		sizeof(struct ifbrparam),
543 	  BC_F_COPYOUT },
544 
545 	{ bridge_ioctl_gifsstp,		sizeof(struct ifbpstpconf),
546 	  BC_F_COPYIN|BC_F_COPYOUT },
547 
548 	{ bridge_ioctl_sproto,		sizeof(struct ifbrparam),
549 	  BC_F_COPYIN|BC_F_SUSER },
550 
551 	{ bridge_ioctl_stxhc,		sizeof(struct ifbrparam),
552 	  BC_F_COPYIN|BC_F_SUSER },
553 
554 	{ bridge_ioctl_sifmaxaddr,	sizeof(struct ifbreq),
555 	  BC_F_COPYIN|BC_F_SUSER },
556 
557 };
558 static const int bridge_control_table_size = nitems(bridge_control_table);
559 
560 VNET_DEFINE_STATIC(LIST_HEAD(, bridge_softc), bridge_list);
561 #define	V_bridge_list	VNET(bridge_list)
562 #define	BRIDGE_LIST_LOCK_INIT(x)	sx_init(&V_bridge_list_sx,	\
563 					    "if_bridge list")
564 #define	BRIDGE_LIST_LOCK_DESTROY(x)	sx_destroy(&V_bridge_list_sx)
565 #define	BRIDGE_LIST_LOCK(x)		sx_xlock(&V_bridge_list_sx)
566 #define	BRIDGE_LIST_UNLOCK(x)		sx_xunlock(&V_bridge_list_sx)
567 
568 VNET_DEFINE_STATIC(struct if_clone *, bridge_cloner);
569 #define	V_bridge_cloner	VNET(bridge_cloner)
570 
571 static const char bridge_name[] = "bridge";
572 
573 static void
574 vnet_bridge_init(const void *unused __unused)
575 {
576 
577 	V_bridge_rtnode_zone = uma_zcreate("bridge_rtnode",
578 	    sizeof(struct bridge_rtnode), NULL, NULL, NULL, NULL,
579 	    UMA_ALIGN_PTR, 0);
580 	BRIDGE_LIST_LOCK_INIT();
581 	LIST_INIT(&V_bridge_list);
582 
583 	struct if_clone_addreq req = {
584 		.create_f = bridge_clone_create,
585 		.destroy_f = bridge_clone_destroy,
586 		.flags = IFC_F_AUTOUNIT,
587 	};
588 	V_bridge_cloner = ifc_attach_cloner(bridge_name, &req);
589 }
590 VNET_SYSINIT(vnet_bridge_init, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY,
591     vnet_bridge_init, NULL);
592 
593 static void
594 vnet_bridge_uninit(const void *unused __unused)
595 {
596 
597 	ifc_detach_cloner(V_bridge_cloner);
598 	V_bridge_cloner = NULL;
599 	BRIDGE_LIST_LOCK_DESTROY();
600 
601 	/* Callbacks may use the UMA zone. */
602 	NET_EPOCH_DRAIN_CALLBACKS();
603 
604 	uma_zdestroy(V_bridge_rtnode_zone);
605 }
606 VNET_SYSUNINIT(vnet_bridge_uninit, SI_SUB_PSEUDO, SI_ORDER_ANY,
607     vnet_bridge_uninit, NULL);
608 
609 static int
610 bridge_modevent(module_t mod, int type, void *data)
611 {
612 
613 	switch (type) {
614 	case MOD_LOAD:
615 		bridge_dn_p = bridge_dummynet;
616 		bridge_detach_cookie = EVENTHANDLER_REGISTER(
617 		    ifnet_departure_event, bridge_ifdetach, NULL,
618 		    EVENTHANDLER_PRI_ANY);
619 		break;
620 	case MOD_UNLOAD:
621 		EVENTHANDLER_DEREGISTER(ifnet_departure_event,
622 		    bridge_detach_cookie);
623 		bridge_dn_p = NULL;
624 		break;
625 	default:
626 		return (EOPNOTSUPP);
627 	}
628 	return (0);
629 }
630 
631 static moduledata_t bridge_mod = {
632 	"if_bridge",
633 	bridge_modevent,
634 	0
635 };
636 
637 DECLARE_MODULE(if_bridge, bridge_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
638 MODULE_VERSION(if_bridge, 1);
639 MODULE_DEPEND(if_bridge, bridgestp, 1, 1, 1);
640 
641 /*
642  * handler for net.link.bridge.ipfw
643  */
644 static int
645 sysctl_pfil_ipfw(SYSCTL_HANDLER_ARGS)
646 {
647 	int enable = V_pfil_ipfw;
648 	int error;
649 
650 	error = sysctl_handle_int(oidp, &enable, 0, req);
651 	enable &= 1;
652 
653 	if (enable != V_pfil_ipfw) {
654 		V_pfil_ipfw = enable;
655 
656 		/*
657 		 * Disable pfil so that ipfw doesnt run twice, if the user
658 		 * really wants both then they can re-enable pfil_bridge and/or
659 		 * pfil_member. Also allow non-ip packets as ipfw can filter by
660 		 * layer2 type.
661 		 */
662 		if (V_pfil_ipfw) {
663 			V_pfil_onlyip = 0;
664 			V_pfil_bridge = 0;
665 			V_pfil_member = 0;
666 		}
667 	}
668 
669 	return (error);
670 }
671 SYSCTL_PROC(_net_link_bridge, OID_AUTO, ipfw,
672     CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_VNET | CTLFLAG_NEEDGIANT,
673     &VNET_NAME(pfil_ipfw), 0, &sysctl_pfil_ipfw, "I",
674     "Layer2 filter with IPFW");
675 
676 #ifdef VIMAGE
677 static void
678 bridge_reassign(struct ifnet *ifp, struct vnet *newvnet, char *arg)
679 {
680 	struct bridge_softc *sc = ifp->if_softc;
681 	struct bridge_iflist *bif;
682 
683 	BRIDGE_LOCK(sc);
684 
685 	while ((bif = CK_LIST_FIRST(&sc->sc_iflist)) != NULL)
686 		bridge_delete_member(sc, bif, 0);
687 
688 	while ((bif = CK_LIST_FIRST(&sc->sc_spanlist)) != NULL) {
689 		bridge_delete_span(sc, bif);
690 	}
691 
692 	BRIDGE_UNLOCK(sc);
693 
694 	ether_reassign(ifp, newvnet, arg);
695 }
696 #endif
697 
698 /*
699  * bridge_clone_create:
700  *
701  *	Create a new bridge instance.
702  */
703 static int
704 bridge_clone_create(struct if_clone *ifc, char *name, size_t len,
705     struct ifc_data *ifd, struct ifnet **ifpp)
706 {
707 	struct bridge_softc *sc;
708 	struct ifnet *ifp;
709 
710 	sc = malloc(sizeof(*sc), M_DEVBUF, M_WAITOK|M_ZERO);
711 	ifp = sc->sc_ifp = if_alloc(IFT_ETHER);
712 	if (ifp == NULL) {
713 		free(sc, M_DEVBUF);
714 		return (ENOSPC);
715 	}
716 
717 	BRIDGE_LOCK_INIT(sc);
718 	sc->sc_brtmax = BRIDGE_RTABLE_MAX;
719 	sc->sc_brttimeout = BRIDGE_RTABLE_TIMEOUT;
720 
721 	/* Initialize our routing table. */
722 	bridge_rtable_init(sc);
723 
724 	callout_init_mtx(&sc->sc_brcallout, &sc->sc_rt_mtx, 0);
725 
726 	CK_LIST_INIT(&sc->sc_iflist);
727 	CK_LIST_INIT(&sc->sc_spanlist);
728 
729 	ifp->if_softc = sc;
730 	if_initname(ifp, bridge_name, ifd->unit);
731 	ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
732 	ifp->if_ioctl = bridge_ioctl;
733 #ifdef ALTQ
734 	ifp->if_start = bridge_altq_start;
735 	ifp->if_transmit = bridge_altq_transmit;
736 	IFQ_SET_MAXLEN(&ifp->if_snd, ifqmaxlen);
737 	ifp->if_snd.ifq_drv_maxlen = 0;
738 	IFQ_SET_READY(&ifp->if_snd);
739 #else
740 	ifp->if_transmit = bridge_transmit;
741 #endif
742 	ifp->if_qflush = bridge_qflush;
743 	ifp->if_init = bridge_init;
744 	ifp->if_type = IFT_BRIDGE;
745 
746 	ether_gen_addr(ifp, &sc->sc_defaddr);
747 
748 	bstp_attach(&sc->sc_stp, &bridge_ops);
749 	ether_ifattach(ifp, sc->sc_defaddr.octet);
750 	/* Now undo some of the damage... */
751 	ifp->if_baudrate = 0;
752 	ifp->if_type = IFT_BRIDGE;
753 #ifdef VIMAGE
754 	ifp->if_reassign = bridge_reassign;
755 #endif
756 
757 	BRIDGE_LIST_LOCK();
758 	LIST_INSERT_HEAD(&V_bridge_list, sc, sc_list);
759 	BRIDGE_LIST_UNLOCK();
760 	*ifpp = ifp;
761 
762 	return (0);
763 }
764 
765 static void
766 bridge_clone_destroy_cb(struct epoch_context *ctx)
767 {
768 	struct bridge_softc *sc;
769 
770 	sc = __containerof(ctx, struct bridge_softc, sc_epoch_ctx);
771 
772 	BRIDGE_LOCK_DESTROY(sc);
773 	free(sc, M_DEVBUF);
774 }
775 
776 /*
777  * bridge_clone_destroy:
778  *
779  *	Destroy a bridge instance.
780  */
781 static int
782 bridge_clone_destroy(struct if_clone *ifc, struct ifnet *ifp, uint32_t flags)
783 {
784 	struct bridge_softc *sc = ifp->if_softc;
785 	struct bridge_iflist *bif;
786 	struct epoch_tracker et;
787 
788 	BRIDGE_LOCK(sc);
789 
790 	bridge_stop(ifp, 1);
791 	ifp->if_flags &= ~IFF_UP;
792 
793 	while ((bif = CK_LIST_FIRST(&sc->sc_iflist)) != NULL)
794 		bridge_delete_member(sc, bif, 0);
795 
796 	while ((bif = CK_LIST_FIRST(&sc->sc_spanlist)) != NULL) {
797 		bridge_delete_span(sc, bif);
798 	}
799 
800 	/* Tear down the routing table. */
801 	bridge_rtable_fini(sc);
802 
803 	BRIDGE_UNLOCK(sc);
804 
805 	NET_EPOCH_ENTER(et);
806 
807 	callout_drain(&sc->sc_brcallout);
808 
809 	BRIDGE_LIST_LOCK();
810 	LIST_REMOVE(sc, sc_list);
811 	BRIDGE_LIST_UNLOCK();
812 
813 	bstp_detach(&sc->sc_stp);
814 #ifdef ALTQ
815 	IFQ_PURGE(&ifp->if_snd);
816 #endif
817 	NET_EPOCH_EXIT(et);
818 
819 	ether_ifdetach(ifp);
820 	if_free(ifp);
821 
822 	NET_EPOCH_CALL(bridge_clone_destroy_cb, &sc->sc_epoch_ctx);
823 
824 	return (0);
825 }
826 
827 /*
828  * bridge_ioctl:
829  *
830  *	Handle a control request from the operator.
831  */
832 static int
833 bridge_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data)
834 {
835 	struct bridge_softc *sc = ifp->if_softc;
836 	struct ifreq *ifr = (struct ifreq *)data;
837 	struct bridge_iflist *bif;
838 	struct thread *td = curthread;
839 	union {
840 		struct ifbreq ifbreq;
841 		struct ifbifconf ifbifconf;
842 		struct ifbareq ifbareq;
843 		struct ifbaconf ifbaconf;
844 		struct ifbrparam ifbrparam;
845 		struct ifbropreq ifbropreq;
846 	} args;
847 	struct ifdrv *ifd = (struct ifdrv *) data;
848 	const struct bridge_control *bc;
849 	int error = 0, oldmtu;
850 
851 	BRIDGE_LOCK(sc);
852 
853 	switch (cmd) {
854 	case SIOCADDMULTI:
855 	case SIOCDELMULTI:
856 		break;
857 
858 	case SIOCGDRVSPEC:
859 	case SIOCSDRVSPEC:
860 		if (ifd->ifd_cmd >= bridge_control_table_size) {
861 			error = EINVAL;
862 			break;
863 		}
864 		bc = &bridge_control_table[ifd->ifd_cmd];
865 
866 		if (cmd == SIOCGDRVSPEC &&
867 		    (bc->bc_flags & BC_F_COPYOUT) == 0) {
868 			error = EINVAL;
869 			break;
870 		}
871 		else if (cmd == SIOCSDRVSPEC &&
872 		    (bc->bc_flags & BC_F_COPYOUT) != 0) {
873 			error = EINVAL;
874 			break;
875 		}
876 
877 		if (bc->bc_flags & BC_F_SUSER) {
878 			error = priv_check(td, PRIV_NET_BRIDGE);
879 			if (error)
880 				break;
881 		}
882 
883 		if (ifd->ifd_len != bc->bc_argsize ||
884 		    ifd->ifd_len > sizeof(args)) {
885 			error = EINVAL;
886 			break;
887 		}
888 
889 		bzero(&args, sizeof(args));
890 		if (bc->bc_flags & BC_F_COPYIN) {
891 			error = copyin(ifd->ifd_data, &args, ifd->ifd_len);
892 			if (error)
893 				break;
894 		}
895 
896 		oldmtu = ifp->if_mtu;
897 		error = (*bc->bc_func)(sc, &args);
898 		if (error)
899 			break;
900 
901 		/*
902 		 * Bridge MTU may change during addition of the first port.
903 		 * If it did, do network layer specific procedure.
904 		 */
905 		if (ifp->if_mtu != oldmtu)
906 			if_notifymtu(ifp);
907 
908 		if (bc->bc_flags & BC_F_COPYOUT)
909 			error = copyout(&args, ifd->ifd_data, ifd->ifd_len);
910 
911 		break;
912 
913 	case SIOCSIFFLAGS:
914 		if (!(ifp->if_flags & IFF_UP) &&
915 		    (ifp->if_drv_flags & IFF_DRV_RUNNING)) {
916 			/*
917 			 * If interface is marked down and it is running,
918 			 * then stop and disable it.
919 			 */
920 			bridge_stop(ifp, 1);
921 		} else if ((ifp->if_flags & IFF_UP) &&
922 		    !(ifp->if_drv_flags & IFF_DRV_RUNNING)) {
923 			/*
924 			 * If interface is marked up and it is stopped, then
925 			 * start it.
926 			 */
927 			BRIDGE_UNLOCK(sc);
928 			(*ifp->if_init)(sc);
929 			BRIDGE_LOCK(sc);
930 		}
931 		break;
932 
933 	case SIOCSIFMTU:
934 		oldmtu = sc->sc_ifp->if_mtu;
935 
936 		if (ifr->ifr_mtu < 576) {
937 			error = EINVAL;
938 			break;
939 		}
940 		if (CK_LIST_EMPTY(&sc->sc_iflist)) {
941 			sc->sc_ifp->if_mtu = ifr->ifr_mtu;
942 			break;
943 		}
944 		CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
945 			error = (*bif->bif_ifp->if_ioctl)(bif->bif_ifp,
946 			    SIOCSIFMTU, (caddr_t)ifr);
947 			if (error != 0) {
948 				log(LOG_NOTICE, "%s: invalid MTU: %u for"
949 				    " member %s\n", sc->sc_ifp->if_xname,
950 				    ifr->ifr_mtu,
951 				    bif->bif_ifp->if_xname);
952 				error = EINVAL;
953 				break;
954 			}
955 		}
956 		if (error) {
957 			/* Restore the previous MTU on all member interfaces. */
958 			ifr->ifr_mtu = oldmtu;
959 			CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
960 				(*bif->bif_ifp->if_ioctl)(bif->bif_ifp,
961 				    SIOCSIFMTU, (caddr_t)ifr);
962 			}
963 		} else {
964 			sc->sc_ifp->if_mtu = ifr->ifr_mtu;
965 		}
966 		break;
967 	default:
968 		/*
969 		 * drop the lock as ether_ioctl() will call bridge_start() and
970 		 * cause the lock to be recursed.
971 		 */
972 		BRIDGE_UNLOCK(sc);
973 		error = ether_ioctl(ifp, cmd, data);
974 		BRIDGE_LOCK(sc);
975 		break;
976 	}
977 
978 	BRIDGE_UNLOCK(sc);
979 
980 	return (error);
981 }
982 
983 /*
984  * bridge_mutecaps:
985  *
986  *	Clear or restore unwanted capabilities on the member interface
987  */
988 static void
989 bridge_mutecaps(struct bridge_softc *sc)
990 {
991 	struct bridge_iflist *bif;
992 	int enabled, mask;
993 
994 	BRIDGE_LOCK_ASSERT(sc);
995 
996 	/* Initial bitmask of capabilities to test */
997 	mask = BRIDGE_IFCAPS_MASK;
998 
999 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1000 		/* Every member must support it or its disabled */
1001 		mask &= bif->bif_savedcaps;
1002 	}
1003 
1004 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1005 		enabled = bif->bif_ifp->if_capenable;
1006 		enabled &= ~BRIDGE_IFCAPS_STRIP;
1007 		/* strip off mask bits and enable them again if allowed */
1008 		enabled &= ~BRIDGE_IFCAPS_MASK;
1009 		enabled |= mask;
1010 		bridge_set_ifcap(sc, bif, enabled);
1011 	}
1012 }
1013 
1014 static void
1015 bridge_set_ifcap(struct bridge_softc *sc, struct bridge_iflist *bif, int set)
1016 {
1017 	struct ifnet *ifp = bif->bif_ifp;
1018 	struct ifreq ifr;
1019 	int error, mask, stuck;
1020 
1021 	bzero(&ifr, sizeof(ifr));
1022 	ifr.ifr_reqcap = set;
1023 
1024 	if (ifp->if_capenable != set) {
1025 		error = (*ifp->if_ioctl)(ifp, SIOCSIFCAP, (caddr_t)&ifr);
1026 		if (error)
1027 			if_printf(sc->sc_ifp,
1028 			    "error setting capabilities on %s: %d\n",
1029 			    ifp->if_xname, error);
1030 		mask = BRIDGE_IFCAPS_MASK | BRIDGE_IFCAPS_STRIP;
1031 		stuck = ifp->if_capenable & mask & ~set;
1032 		if (stuck != 0)
1033 			if_printf(sc->sc_ifp,
1034 			    "can't disable some capabilities on %s: 0x%x\n",
1035 			    ifp->if_xname, stuck);
1036 	}
1037 }
1038 
1039 /*
1040  * bridge_lookup_member:
1041  *
1042  *	Lookup a bridge member interface.
1043  */
1044 static struct bridge_iflist *
1045 bridge_lookup_member(struct bridge_softc *sc, const char *name)
1046 {
1047 	struct bridge_iflist *bif;
1048 	struct ifnet *ifp;
1049 
1050 	BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc);
1051 
1052 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1053 		ifp = bif->bif_ifp;
1054 		if (strcmp(ifp->if_xname, name) == 0)
1055 			return (bif);
1056 	}
1057 
1058 	return (NULL);
1059 }
1060 
1061 /*
1062  * bridge_lookup_member_if:
1063  *
1064  *	Lookup a bridge member interface by ifnet*.
1065  */
1066 static struct bridge_iflist *
1067 bridge_lookup_member_if(struct bridge_softc *sc, struct ifnet *member_ifp)
1068 {
1069 	struct bridge_iflist *bif;
1070 
1071 	BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc);
1072 
1073 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1074 		if (bif->bif_ifp == member_ifp)
1075 			return (bif);
1076 	}
1077 
1078 	return (NULL);
1079 }
1080 
1081 static void
1082 bridge_delete_member_cb(struct epoch_context *ctx)
1083 {
1084 	struct bridge_iflist *bif;
1085 
1086 	bif = __containerof(ctx, struct bridge_iflist, bif_epoch_ctx);
1087 
1088 	free(bif, M_DEVBUF);
1089 }
1090 
1091 /*
1092  * bridge_delete_member:
1093  *
1094  *	Delete the specified member interface.
1095  */
1096 static void
1097 bridge_delete_member(struct bridge_softc *sc, struct bridge_iflist *bif,
1098     int gone)
1099 {
1100 	struct ifnet *ifs = bif->bif_ifp;
1101 	struct ifnet *fif = NULL;
1102 	struct bridge_iflist *bifl;
1103 
1104 	BRIDGE_LOCK_ASSERT(sc);
1105 
1106 	if (bif->bif_flags & IFBIF_STP)
1107 		bstp_disable(&bif->bif_stp);
1108 
1109 	ifs->if_bridge = NULL;
1110 	CK_LIST_REMOVE(bif, bif_next);
1111 
1112 	/*
1113 	 * If removing the interface that gave the bridge its mac address, set
1114 	 * the mac address of the bridge to the address of the next member, or
1115 	 * to its default address if no members are left.
1116 	 */
1117 	if (V_bridge_inherit_mac && sc->sc_ifaddr == ifs) {
1118 		if (CK_LIST_EMPTY(&sc->sc_iflist)) {
1119 			bcopy(&sc->sc_defaddr,
1120 			    IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN);
1121 			sc->sc_ifaddr = NULL;
1122 		} else {
1123 			bifl = CK_LIST_FIRST(&sc->sc_iflist);
1124 			fif = bifl->bif_ifp;
1125 			bcopy(IF_LLADDR(fif),
1126 			    IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN);
1127 			sc->sc_ifaddr = fif;
1128 		}
1129 		EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp);
1130 	}
1131 
1132 	bridge_linkcheck(sc);
1133 	bridge_mutecaps(sc);	/* recalcuate now this interface is removed */
1134 	BRIDGE_RT_LOCK(sc);
1135 	bridge_rtdelete(sc, ifs, IFBF_FLUSHALL);
1136 	BRIDGE_RT_UNLOCK(sc);
1137 	KASSERT(bif->bif_addrcnt == 0,
1138 	    ("%s: %d bridge routes referenced", __func__, bif->bif_addrcnt));
1139 
1140 	ifs->if_bridge_output = NULL;
1141 	ifs->if_bridge_input = NULL;
1142 	ifs->if_bridge_linkstate = NULL;
1143 	if (!gone) {
1144 		switch (ifs->if_type) {
1145 		case IFT_ETHER:
1146 		case IFT_L2VLAN:
1147 			/*
1148 			 * Take the interface out of promiscuous mode, but only
1149 			 * if it was promiscuous in the first place. It might
1150 			 * not be if we're in the bridge_ioctl_add() error path.
1151 			 */
1152 			if (ifs->if_flags & IFF_PROMISC)
1153 				(void) ifpromisc(ifs, 0);
1154 			break;
1155 
1156 		case IFT_GIF:
1157 			break;
1158 
1159 		default:
1160 #ifdef DIAGNOSTIC
1161 			panic("bridge_delete_member: impossible");
1162 #endif
1163 			break;
1164 		}
1165 		/* reneable any interface capabilities */
1166 		bridge_set_ifcap(sc, bif, bif->bif_savedcaps);
1167 	}
1168 	bstp_destroy(&bif->bif_stp);	/* prepare to free */
1169 
1170 	NET_EPOCH_CALL(bridge_delete_member_cb, &bif->bif_epoch_ctx);
1171 }
1172 
1173 /*
1174  * bridge_delete_span:
1175  *
1176  *	Delete the specified span interface.
1177  */
1178 static void
1179 bridge_delete_span(struct bridge_softc *sc, struct bridge_iflist *bif)
1180 {
1181 	BRIDGE_LOCK_ASSERT(sc);
1182 
1183 	KASSERT(bif->bif_ifp->if_bridge == NULL,
1184 	    ("%s: not a span interface", __func__));
1185 
1186 	CK_LIST_REMOVE(bif, bif_next);
1187 
1188 	NET_EPOCH_CALL(bridge_delete_member_cb, &bif->bif_epoch_ctx);
1189 }
1190 
1191 static int
1192 bridge_ioctl_add(struct bridge_softc *sc, void *arg)
1193 {
1194 	struct ifbreq *req = arg;
1195 	struct bridge_iflist *bif = NULL;
1196 	struct ifnet *ifs;
1197 	int error = 0;
1198 
1199 	ifs = ifunit(req->ifbr_ifsname);
1200 	if (ifs == NULL)
1201 		return (ENOENT);
1202 	if (ifs->if_ioctl == NULL)	/* must be supported */
1203 		return (EINVAL);
1204 
1205 	/* If it's in the span list, it can't be a member. */
1206 	CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1207 		if (ifs == bif->bif_ifp)
1208 			return (EBUSY);
1209 
1210 	if (ifs->if_bridge == sc)
1211 		return (EEXIST);
1212 
1213 	if (ifs->if_bridge != NULL)
1214 		return (EBUSY);
1215 
1216 	switch (ifs->if_type) {
1217 	case IFT_ETHER:
1218 	case IFT_L2VLAN:
1219 	case IFT_GIF:
1220 		/* permitted interface types */
1221 		break;
1222 	default:
1223 		return (EINVAL);
1224 	}
1225 
1226 #ifdef INET6
1227 	/*
1228 	 * Two valid inet6 addresses with link-local scope must not be
1229 	 * on the parent interface and the member interfaces at the
1230 	 * same time.  This restriction is needed to prevent violation
1231 	 * of link-local scope zone.  Attempts to add a member
1232 	 * interface which has inet6 addresses when the parent has
1233 	 * inet6 triggers removal of all inet6 addresses on the member
1234 	 * interface.
1235 	 */
1236 
1237 	/* Check if the parent interface has a link-local scope addr. */
1238 	if (V_allow_llz_overlap == 0 &&
1239 	    in6ifa_llaonifp(sc->sc_ifp) != NULL) {
1240 		/*
1241 		 * If any, remove all inet6 addresses from the member
1242 		 * interfaces.
1243 		 */
1244 		CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1245  			if (in6ifa_llaonifp(bif->bif_ifp)) {
1246 				in6_ifdetach(bif->bif_ifp);
1247 				if_printf(sc->sc_ifp,
1248 				    "IPv6 addresses on %s have been removed "
1249 				    "before adding it as a member to prevent "
1250 				    "IPv6 address scope violation.\n",
1251 				    bif->bif_ifp->if_xname);
1252 			}
1253 		}
1254 		if (in6ifa_llaonifp(ifs)) {
1255 			in6_ifdetach(ifs);
1256 			if_printf(sc->sc_ifp,
1257 			    "IPv6 addresses on %s have been removed "
1258 			    "before adding it as a member to prevent "
1259 			    "IPv6 address scope violation.\n",
1260 			    ifs->if_xname);
1261 		}
1262 	}
1263 #endif
1264 	/* Allow the first Ethernet member to define the MTU */
1265 	if (CK_LIST_EMPTY(&sc->sc_iflist))
1266 		sc->sc_ifp->if_mtu = ifs->if_mtu;
1267 	else if (sc->sc_ifp->if_mtu != ifs->if_mtu) {
1268 		struct ifreq ifr;
1269 
1270 		snprintf(ifr.ifr_name, sizeof(ifr.ifr_name), "%s",
1271 		    ifs->if_xname);
1272 		ifr.ifr_mtu = sc->sc_ifp->if_mtu;
1273 
1274 		error = (*ifs->if_ioctl)(ifs,
1275 		    SIOCSIFMTU, (caddr_t)&ifr);
1276 		if (error != 0) {
1277 			log(LOG_NOTICE, "%s: invalid MTU: %u for"
1278 			    " new member %s\n", sc->sc_ifp->if_xname,
1279 			    ifr.ifr_mtu,
1280 			    ifs->if_xname);
1281 			return (EINVAL);
1282 		}
1283 	}
1284 
1285 	bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO);
1286 	if (bif == NULL)
1287 		return (ENOMEM);
1288 
1289 	bif->bif_ifp = ifs;
1290 	bif->bif_flags = IFBIF_LEARNING | IFBIF_DISCOVER;
1291 	bif->bif_savedcaps = ifs->if_capenable;
1292 
1293 	/*
1294 	 * Assign the interface's MAC address to the bridge if it's the first
1295 	 * member and the MAC address of the bridge has not been changed from
1296 	 * the default randomly generated one.
1297 	 */
1298 	if (V_bridge_inherit_mac && CK_LIST_EMPTY(&sc->sc_iflist) &&
1299 	    !memcmp(IF_LLADDR(sc->sc_ifp), sc->sc_defaddr.octet, ETHER_ADDR_LEN)) {
1300 		bcopy(IF_LLADDR(ifs), IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN);
1301 		sc->sc_ifaddr = ifs;
1302 		EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp);
1303 	}
1304 
1305 	ifs->if_bridge = sc;
1306 	ifs->if_bridge_output = bridge_output;
1307 	ifs->if_bridge_input = bridge_input;
1308 	ifs->if_bridge_linkstate = bridge_linkstate;
1309 	bstp_create(&sc->sc_stp, &bif->bif_stp, bif->bif_ifp);
1310 	/*
1311 	 * XXX: XLOCK HERE!?!
1312 	 *
1313 	 * NOTE: insert_***HEAD*** should be safe for the traversals.
1314 	 */
1315 	CK_LIST_INSERT_HEAD(&sc->sc_iflist, bif, bif_next);
1316 
1317 	/* Set interface capabilities to the intersection set of all members */
1318 	bridge_mutecaps(sc);
1319 	bridge_linkcheck(sc);
1320 
1321 	/* Place the interface into promiscuous mode */
1322 	switch (ifs->if_type) {
1323 		case IFT_ETHER:
1324 		case IFT_L2VLAN:
1325 			error = ifpromisc(ifs, 1);
1326 			break;
1327 	}
1328 
1329 	if (error)
1330 		bridge_delete_member(sc, bif, 0);
1331 	return (error);
1332 }
1333 
1334 static int
1335 bridge_ioctl_del(struct bridge_softc *sc, void *arg)
1336 {
1337 	struct ifbreq *req = arg;
1338 	struct bridge_iflist *bif;
1339 
1340 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1341 	if (bif == NULL)
1342 		return (ENOENT);
1343 
1344 	bridge_delete_member(sc, bif, 0);
1345 
1346 	return (0);
1347 }
1348 
1349 static int
1350 bridge_ioctl_gifflags(struct bridge_softc *sc, void *arg)
1351 {
1352 	struct ifbreq *req = arg;
1353 	struct bridge_iflist *bif;
1354 	struct bstp_port *bp;
1355 
1356 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1357 	if (bif == NULL)
1358 		return (ENOENT);
1359 
1360 	bp = &bif->bif_stp;
1361 	req->ifbr_ifsflags = bif->bif_flags;
1362 	req->ifbr_state = bp->bp_state;
1363 	req->ifbr_priority = bp->bp_priority;
1364 	req->ifbr_path_cost = bp->bp_path_cost;
1365 	req->ifbr_portno = bif->bif_ifp->if_index & 0xfff;
1366 	req->ifbr_proto = bp->bp_protover;
1367 	req->ifbr_role = bp->bp_role;
1368 	req->ifbr_stpflags = bp->bp_flags;
1369 	req->ifbr_addrcnt = bif->bif_addrcnt;
1370 	req->ifbr_addrmax = bif->bif_addrmax;
1371 	req->ifbr_addrexceeded = bif->bif_addrexceeded;
1372 
1373 	/* Copy STP state options as flags */
1374 	if (bp->bp_operedge)
1375 		req->ifbr_ifsflags |= IFBIF_BSTP_EDGE;
1376 	if (bp->bp_flags & BSTP_PORT_AUTOEDGE)
1377 		req->ifbr_ifsflags |= IFBIF_BSTP_AUTOEDGE;
1378 	if (bp->bp_ptp_link)
1379 		req->ifbr_ifsflags |= IFBIF_BSTP_PTP;
1380 	if (bp->bp_flags & BSTP_PORT_AUTOPTP)
1381 		req->ifbr_ifsflags |= IFBIF_BSTP_AUTOPTP;
1382 	if (bp->bp_flags & BSTP_PORT_ADMEDGE)
1383 		req->ifbr_ifsflags |= IFBIF_BSTP_ADMEDGE;
1384 	if (bp->bp_flags & BSTP_PORT_ADMCOST)
1385 		req->ifbr_ifsflags |= IFBIF_BSTP_ADMCOST;
1386 	return (0);
1387 }
1388 
1389 static int
1390 bridge_ioctl_sifflags(struct bridge_softc *sc, void *arg)
1391 {
1392 	struct epoch_tracker et;
1393 	struct ifbreq *req = arg;
1394 	struct bridge_iflist *bif;
1395 	struct bstp_port *bp;
1396 	int error;
1397 
1398 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1399 	if (bif == NULL)
1400 		return (ENOENT);
1401 	bp = &bif->bif_stp;
1402 
1403 	if (req->ifbr_ifsflags & IFBIF_SPAN)
1404 		/* SPAN is readonly */
1405 		return (EINVAL);
1406 
1407 	NET_EPOCH_ENTER(et);
1408 
1409 	if (req->ifbr_ifsflags & IFBIF_STP) {
1410 		if ((bif->bif_flags & IFBIF_STP) == 0) {
1411 			error = bstp_enable(&bif->bif_stp);
1412 			if (error) {
1413 				NET_EPOCH_EXIT(et);
1414 				return (error);
1415 			}
1416 		}
1417 	} else {
1418 		if ((bif->bif_flags & IFBIF_STP) != 0)
1419 			bstp_disable(&bif->bif_stp);
1420 	}
1421 
1422 	/* Pass on STP flags */
1423 	bstp_set_edge(bp, req->ifbr_ifsflags & IFBIF_BSTP_EDGE ? 1 : 0);
1424 	bstp_set_autoedge(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOEDGE ? 1 : 0);
1425 	bstp_set_ptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_PTP ? 1 : 0);
1426 	bstp_set_autoptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOPTP ? 1 : 0);
1427 
1428 	/* Save the bits relating to the bridge */
1429 	bif->bif_flags = req->ifbr_ifsflags & IFBIFMASK;
1430 
1431 	NET_EPOCH_EXIT(et);
1432 
1433 	return (0);
1434 }
1435 
1436 static int
1437 bridge_ioctl_scache(struct bridge_softc *sc, void *arg)
1438 {
1439 	struct ifbrparam *param = arg;
1440 
1441 	sc->sc_brtmax = param->ifbrp_csize;
1442 	bridge_rttrim(sc);
1443 
1444 	return (0);
1445 }
1446 
1447 static int
1448 bridge_ioctl_gcache(struct bridge_softc *sc, void *arg)
1449 {
1450 	struct ifbrparam *param = arg;
1451 
1452 	param->ifbrp_csize = sc->sc_brtmax;
1453 
1454 	return (0);
1455 }
1456 
1457 static int
1458 bridge_ioctl_gifs(struct bridge_softc *sc, void *arg)
1459 {
1460 	struct ifbifconf *bifc = arg;
1461 	struct bridge_iflist *bif;
1462 	struct ifbreq breq;
1463 	char *buf, *outbuf;
1464 	int count, buflen, len, error = 0;
1465 
1466 	count = 0;
1467 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next)
1468 		count++;
1469 	CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1470 		count++;
1471 
1472 	buflen = sizeof(breq) * count;
1473 	if (bifc->ifbic_len == 0) {
1474 		bifc->ifbic_len = buflen;
1475 		return (0);
1476 	}
1477 	outbuf = malloc(buflen, M_TEMP, M_NOWAIT | M_ZERO);
1478 	if (outbuf == NULL)
1479 		return (ENOMEM);
1480 
1481 	count = 0;
1482 	buf = outbuf;
1483 	len = min(bifc->ifbic_len, buflen);
1484 	bzero(&breq, sizeof(breq));
1485 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1486 		if (len < sizeof(breq))
1487 			break;
1488 
1489 		strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname,
1490 		    sizeof(breq.ifbr_ifsname));
1491 		/* Fill in the ifbreq structure */
1492 		error = bridge_ioctl_gifflags(sc, &breq);
1493 		if (error)
1494 			break;
1495 		memcpy(buf, &breq, sizeof(breq));
1496 		count++;
1497 		buf += sizeof(breq);
1498 		len -= sizeof(breq);
1499 	}
1500 	CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) {
1501 		if (len < sizeof(breq))
1502 			break;
1503 
1504 		strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname,
1505 		    sizeof(breq.ifbr_ifsname));
1506 		breq.ifbr_ifsflags = bif->bif_flags;
1507 		breq.ifbr_portno = bif->bif_ifp->if_index & 0xfff;
1508 		memcpy(buf, &breq, sizeof(breq));
1509 		count++;
1510 		buf += sizeof(breq);
1511 		len -= sizeof(breq);
1512 	}
1513 
1514 	bifc->ifbic_len = sizeof(breq) * count;
1515 	error = copyout(outbuf, bifc->ifbic_req, bifc->ifbic_len);
1516 	free(outbuf, M_TEMP);
1517 	return (error);
1518 }
1519 
1520 static int
1521 bridge_ioctl_rts(struct bridge_softc *sc, void *arg)
1522 {
1523 	struct ifbaconf *bac = arg;
1524 	struct bridge_rtnode *brt;
1525 	struct ifbareq bareq;
1526 	char *buf, *outbuf;
1527 	int count, buflen, len, error = 0;
1528 
1529 	if (bac->ifbac_len == 0)
1530 		return (0);
1531 
1532 	count = 0;
1533 	CK_LIST_FOREACH(brt, &sc->sc_rtlist, brt_list)
1534 		count++;
1535 	buflen = sizeof(bareq) * count;
1536 
1537 	outbuf = malloc(buflen, M_TEMP, M_NOWAIT | M_ZERO);
1538 	if (outbuf == NULL)
1539 		return (ENOMEM);
1540 
1541 	count = 0;
1542 	buf = outbuf;
1543 	len = min(bac->ifbac_len, buflen);
1544 	bzero(&bareq, sizeof(bareq));
1545 	CK_LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) {
1546 		if (len < sizeof(bareq))
1547 			goto out;
1548 		strlcpy(bareq.ifba_ifsname, brt->brt_ifp->if_xname,
1549 		    sizeof(bareq.ifba_ifsname));
1550 		memcpy(bareq.ifba_dst, brt->brt_addr, sizeof(brt->brt_addr));
1551 		bareq.ifba_vlan = brt->brt_vlan;
1552 		if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC &&
1553 				time_uptime < brt->brt_expire)
1554 			bareq.ifba_expire = brt->brt_expire - time_uptime;
1555 		else
1556 			bareq.ifba_expire = 0;
1557 		bareq.ifba_flags = brt->brt_flags;
1558 
1559 		memcpy(buf, &bareq, sizeof(bareq));
1560 		count++;
1561 		buf += sizeof(bareq);
1562 		len -= sizeof(bareq);
1563 	}
1564 out:
1565 	bac->ifbac_len = sizeof(bareq) * count;
1566 	error = copyout(outbuf, bac->ifbac_req, bac->ifbac_len);
1567 	free(outbuf, M_TEMP);
1568 	return (error);
1569 }
1570 
1571 static int
1572 bridge_ioctl_saddr(struct bridge_softc *sc, void *arg)
1573 {
1574 	struct ifbareq *req = arg;
1575 	struct bridge_iflist *bif;
1576 	struct epoch_tracker et;
1577 	int error;
1578 
1579 	NET_EPOCH_ENTER(et);
1580 	bif = bridge_lookup_member(sc, req->ifba_ifsname);
1581 	if (bif == NULL) {
1582 		NET_EPOCH_EXIT(et);
1583 		return (ENOENT);
1584 	}
1585 
1586 	/* bridge_rtupdate() may acquire the lock. */
1587 	error = bridge_rtupdate(sc, req->ifba_dst, req->ifba_vlan, bif, 1,
1588 	    req->ifba_flags);
1589 	NET_EPOCH_EXIT(et);
1590 
1591 	return (error);
1592 }
1593 
1594 static int
1595 bridge_ioctl_sto(struct bridge_softc *sc, void *arg)
1596 {
1597 	struct ifbrparam *param = arg;
1598 
1599 	sc->sc_brttimeout = param->ifbrp_ctime;
1600 	return (0);
1601 }
1602 
1603 static int
1604 bridge_ioctl_gto(struct bridge_softc *sc, void *arg)
1605 {
1606 	struct ifbrparam *param = arg;
1607 
1608 	param->ifbrp_ctime = sc->sc_brttimeout;
1609 	return (0);
1610 }
1611 
1612 static int
1613 bridge_ioctl_daddr(struct bridge_softc *sc, void *arg)
1614 {
1615 	struct ifbareq *req = arg;
1616 
1617 	return (bridge_rtdaddr(sc, req->ifba_dst, req->ifba_vlan));
1618 }
1619 
1620 static int
1621 bridge_ioctl_flush(struct bridge_softc *sc, void *arg)
1622 {
1623 	struct ifbreq *req = arg;
1624 
1625 	BRIDGE_RT_LOCK(sc);
1626 	bridge_rtflush(sc, req->ifbr_ifsflags);
1627 	BRIDGE_RT_UNLOCK(sc);
1628 
1629 	return (0);
1630 }
1631 
1632 static int
1633 bridge_ioctl_gpri(struct bridge_softc *sc, void *arg)
1634 {
1635 	struct ifbrparam *param = arg;
1636 	struct bstp_state *bs = &sc->sc_stp;
1637 
1638 	param->ifbrp_prio = bs->bs_bridge_priority;
1639 	return (0);
1640 }
1641 
1642 static int
1643 bridge_ioctl_spri(struct bridge_softc *sc, void *arg)
1644 {
1645 	struct ifbrparam *param = arg;
1646 
1647 	return (bstp_set_priority(&sc->sc_stp, param->ifbrp_prio));
1648 }
1649 
1650 static int
1651 bridge_ioctl_ght(struct bridge_softc *sc, void *arg)
1652 {
1653 	struct ifbrparam *param = arg;
1654 	struct bstp_state *bs = &sc->sc_stp;
1655 
1656 	param->ifbrp_hellotime = bs->bs_bridge_htime >> 8;
1657 	return (0);
1658 }
1659 
1660 static int
1661 bridge_ioctl_sht(struct bridge_softc *sc, void *arg)
1662 {
1663 	struct ifbrparam *param = arg;
1664 
1665 	return (bstp_set_htime(&sc->sc_stp, param->ifbrp_hellotime));
1666 }
1667 
1668 static int
1669 bridge_ioctl_gfd(struct bridge_softc *sc, void *arg)
1670 {
1671 	struct ifbrparam *param = arg;
1672 	struct bstp_state *bs = &sc->sc_stp;
1673 
1674 	param->ifbrp_fwddelay = bs->bs_bridge_fdelay >> 8;
1675 	return (0);
1676 }
1677 
1678 static int
1679 bridge_ioctl_sfd(struct bridge_softc *sc, void *arg)
1680 {
1681 	struct ifbrparam *param = arg;
1682 
1683 	return (bstp_set_fdelay(&sc->sc_stp, param->ifbrp_fwddelay));
1684 }
1685 
1686 static int
1687 bridge_ioctl_gma(struct bridge_softc *sc, void *arg)
1688 {
1689 	struct ifbrparam *param = arg;
1690 	struct bstp_state *bs = &sc->sc_stp;
1691 
1692 	param->ifbrp_maxage = bs->bs_bridge_max_age >> 8;
1693 	return (0);
1694 }
1695 
1696 static int
1697 bridge_ioctl_sma(struct bridge_softc *sc, void *arg)
1698 {
1699 	struct ifbrparam *param = arg;
1700 
1701 	return (bstp_set_maxage(&sc->sc_stp, param->ifbrp_maxage));
1702 }
1703 
1704 static int
1705 bridge_ioctl_sifprio(struct bridge_softc *sc, void *arg)
1706 {
1707 	struct ifbreq *req = arg;
1708 	struct bridge_iflist *bif;
1709 
1710 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1711 	if (bif == NULL)
1712 		return (ENOENT);
1713 
1714 	return (bstp_set_port_priority(&bif->bif_stp, req->ifbr_priority));
1715 }
1716 
1717 static int
1718 bridge_ioctl_sifcost(struct bridge_softc *sc, void *arg)
1719 {
1720 	struct ifbreq *req = arg;
1721 	struct bridge_iflist *bif;
1722 
1723 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1724 	if (bif == NULL)
1725 		return (ENOENT);
1726 
1727 	return (bstp_set_path_cost(&bif->bif_stp, req->ifbr_path_cost));
1728 }
1729 
1730 static int
1731 bridge_ioctl_sifmaxaddr(struct bridge_softc *sc, void *arg)
1732 {
1733 	struct ifbreq *req = arg;
1734 	struct bridge_iflist *bif;
1735 
1736 	bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1737 	if (bif == NULL)
1738 		return (ENOENT);
1739 
1740 	bif->bif_addrmax = req->ifbr_addrmax;
1741 	return (0);
1742 }
1743 
1744 static int
1745 bridge_ioctl_addspan(struct bridge_softc *sc, void *arg)
1746 {
1747 	struct ifbreq *req = arg;
1748 	struct bridge_iflist *bif = NULL;
1749 	struct ifnet *ifs;
1750 
1751 	ifs = ifunit(req->ifbr_ifsname);
1752 	if (ifs == NULL)
1753 		return (ENOENT);
1754 
1755 	CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1756 		if (ifs == bif->bif_ifp)
1757 			return (EBUSY);
1758 
1759 	if (ifs->if_bridge != NULL)
1760 		return (EBUSY);
1761 
1762 	switch (ifs->if_type) {
1763 		case IFT_ETHER:
1764 		case IFT_GIF:
1765 		case IFT_L2VLAN:
1766 			break;
1767 		default:
1768 			return (EINVAL);
1769 	}
1770 
1771 	bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO);
1772 	if (bif == NULL)
1773 		return (ENOMEM);
1774 
1775 	bif->bif_ifp = ifs;
1776 	bif->bif_flags = IFBIF_SPAN;
1777 
1778 	CK_LIST_INSERT_HEAD(&sc->sc_spanlist, bif, bif_next);
1779 
1780 	return (0);
1781 }
1782 
1783 static int
1784 bridge_ioctl_delspan(struct bridge_softc *sc, void *arg)
1785 {
1786 	struct ifbreq *req = arg;
1787 	struct bridge_iflist *bif;
1788 	struct ifnet *ifs;
1789 
1790 	ifs = ifunit(req->ifbr_ifsname);
1791 	if (ifs == NULL)
1792 		return (ENOENT);
1793 
1794 	CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1795 		if (ifs == bif->bif_ifp)
1796 			break;
1797 
1798 	if (bif == NULL)
1799 		return (ENOENT);
1800 
1801 	bridge_delete_span(sc, bif);
1802 
1803 	return (0);
1804 }
1805 
1806 static int
1807 bridge_ioctl_gbparam(struct bridge_softc *sc, void *arg)
1808 {
1809 	struct ifbropreq *req = arg;
1810 	struct bstp_state *bs = &sc->sc_stp;
1811 	struct bstp_port *root_port;
1812 
1813 	req->ifbop_maxage = bs->bs_bridge_max_age >> 8;
1814 	req->ifbop_hellotime = bs->bs_bridge_htime >> 8;
1815 	req->ifbop_fwddelay = bs->bs_bridge_fdelay >> 8;
1816 
1817 	root_port = bs->bs_root_port;
1818 	if (root_port == NULL)
1819 		req->ifbop_root_port = 0;
1820 	else
1821 		req->ifbop_root_port = root_port->bp_ifp->if_index;
1822 
1823 	req->ifbop_holdcount = bs->bs_txholdcount;
1824 	req->ifbop_priority = bs->bs_bridge_priority;
1825 	req->ifbop_protocol = bs->bs_protover;
1826 	req->ifbop_root_path_cost = bs->bs_root_pv.pv_cost;
1827 	req->ifbop_bridgeid = bs->bs_bridge_pv.pv_dbridge_id;
1828 	req->ifbop_designated_root = bs->bs_root_pv.pv_root_id;
1829 	req->ifbop_designated_bridge = bs->bs_root_pv.pv_dbridge_id;
1830 	req->ifbop_last_tc_time.tv_sec = bs->bs_last_tc_time.tv_sec;
1831 	req->ifbop_last_tc_time.tv_usec = bs->bs_last_tc_time.tv_usec;
1832 
1833 	return (0);
1834 }
1835 
1836 static int
1837 bridge_ioctl_grte(struct bridge_softc *sc, void *arg)
1838 {
1839 	struct ifbrparam *param = arg;
1840 
1841 	param->ifbrp_cexceeded = sc->sc_brtexceeded;
1842 	return (0);
1843 }
1844 
1845 static int
1846 bridge_ioctl_gifsstp(struct bridge_softc *sc, void *arg)
1847 {
1848 	struct ifbpstpconf *bifstp = arg;
1849 	struct bridge_iflist *bif;
1850 	struct bstp_port *bp;
1851 	struct ifbpstpreq bpreq;
1852 	char *buf, *outbuf;
1853 	int count, buflen, len, error = 0;
1854 
1855 	count = 0;
1856 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1857 		if ((bif->bif_flags & IFBIF_STP) != 0)
1858 			count++;
1859 	}
1860 
1861 	buflen = sizeof(bpreq) * count;
1862 	if (bifstp->ifbpstp_len == 0) {
1863 		bifstp->ifbpstp_len = buflen;
1864 		return (0);
1865 	}
1866 
1867 	outbuf = malloc(buflen, M_TEMP, M_NOWAIT | M_ZERO);
1868 	if (outbuf == NULL)
1869 		return (ENOMEM);
1870 
1871 	count = 0;
1872 	buf = outbuf;
1873 	len = min(bifstp->ifbpstp_len, buflen);
1874 	bzero(&bpreq, sizeof(bpreq));
1875 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1876 		if (len < sizeof(bpreq))
1877 			break;
1878 
1879 		if ((bif->bif_flags & IFBIF_STP) == 0)
1880 			continue;
1881 
1882 		bp = &bif->bif_stp;
1883 		bpreq.ifbp_portno = bif->bif_ifp->if_index & 0xfff;
1884 		bpreq.ifbp_fwd_trans = bp->bp_forward_transitions;
1885 		bpreq.ifbp_design_cost = bp->bp_desg_pv.pv_cost;
1886 		bpreq.ifbp_design_port = bp->bp_desg_pv.pv_port_id;
1887 		bpreq.ifbp_design_bridge = bp->bp_desg_pv.pv_dbridge_id;
1888 		bpreq.ifbp_design_root = bp->bp_desg_pv.pv_root_id;
1889 
1890 		memcpy(buf, &bpreq, sizeof(bpreq));
1891 		count++;
1892 		buf += sizeof(bpreq);
1893 		len -= sizeof(bpreq);
1894 	}
1895 
1896 	bifstp->ifbpstp_len = sizeof(bpreq) * count;
1897 	error = copyout(outbuf, bifstp->ifbpstp_req, bifstp->ifbpstp_len);
1898 	free(outbuf, M_TEMP);
1899 	return (error);
1900 }
1901 
1902 static int
1903 bridge_ioctl_sproto(struct bridge_softc *sc, void *arg)
1904 {
1905 	struct ifbrparam *param = arg;
1906 
1907 	return (bstp_set_protocol(&sc->sc_stp, param->ifbrp_proto));
1908 }
1909 
1910 static int
1911 bridge_ioctl_stxhc(struct bridge_softc *sc, void *arg)
1912 {
1913 	struct ifbrparam *param = arg;
1914 
1915 	return (bstp_set_holdcount(&sc->sc_stp, param->ifbrp_txhc));
1916 }
1917 
1918 /*
1919  * bridge_ifdetach:
1920  *
1921  *	Detach an interface from a bridge.  Called when a member
1922  *	interface is detaching.
1923  */
1924 static void
1925 bridge_ifdetach(void *arg __unused, struct ifnet *ifp)
1926 {
1927 	struct bridge_softc *sc = ifp->if_bridge;
1928 	struct bridge_iflist *bif;
1929 
1930 	if (ifp->if_flags & IFF_RENAMING)
1931 		return;
1932 	if (V_bridge_cloner == NULL) {
1933 		/*
1934 		 * This detach handler can be called after
1935 		 * vnet_bridge_uninit().  Just return in that case.
1936 		 */
1937 		return;
1938 	}
1939 	/* Check if the interface is a bridge member */
1940 	if (sc != NULL) {
1941 		BRIDGE_LOCK(sc);
1942 
1943 		bif = bridge_lookup_member_if(sc, ifp);
1944 		if (bif != NULL)
1945 			bridge_delete_member(sc, bif, 1);
1946 
1947 		BRIDGE_UNLOCK(sc);
1948 		return;
1949 	}
1950 
1951 	/* Check if the interface is a span port */
1952 	BRIDGE_LIST_LOCK();
1953 	LIST_FOREACH(sc, &V_bridge_list, sc_list) {
1954 		BRIDGE_LOCK(sc);
1955 		CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1956 			if (ifp == bif->bif_ifp) {
1957 				bridge_delete_span(sc, bif);
1958 				break;
1959 			}
1960 
1961 		BRIDGE_UNLOCK(sc);
1962 	}
1963 	BRIDGE_LIST_UNLOCK();
1964 }
1965 
1966 /*
1967  * bridge_init:
1968  *
1969  *	Initialize a bridge interface.
1970  */
1971 static void
1972 bridge_init(void *xsc)
1973 {
1974 	struct bridge_softc *sc = (struct bridge_softc *)xsc;
1975 	struct ifnet *ifp = sc->sc_ifp;
1976 
1977 	if (ifp->if_drv_flags & IFF_DRV_RUNNING)
1978 		return;
1979 
1980 	BRIDGE_LOCK(sc);
1981 	callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz,
1982 	    bridge_timer, sc);
1983 
1984 	ifp->if_drv_flags |= IFF_DRV_RUNNING;
1985 	bstp_init(&sc->sc_stp);		/* Initialize Spanning Tree */
1986 
1987 	BRIDGE_UNLOCK(sc);
1988 }
1989 
1990 /*
1991  * bridge_stop:
1992  *
1993  *	Stop the bridge interface.
1994  */
1995 static void
1996 bridge_stop(struct ifnet *ifp, int disable)
1997 {
1998 	struct bridge_softc *sc = ifp->if_softc;
1999 
2000 	BRIDGE_LOCK_ASSERT(sc);
2001 
2002 	if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
2003 		return;
2004 
2005 	BRIDGE_RT_LOCK(sc);
2006 	callout_stop(&sc->sc_brcallout);
2007 
2008 	bstp_stop(&sc->sc_stp);
2009 
2010 	bridge_rtflush(sc, IFBF_FLUSHDYN);
2011 	BRIDGE_RT_UNLOCK(sc);
2012 
2013 	ifp->if_drv_flags &= ~IFF_DRV_RUNNING;
2014 }
2015 
2016 /*
2017  * bridge_enqueue:
2018  *
2019  *	Enqueue a packet on a bridge member interface.
2020  *
2021  */
2022 static int
2023 bridge_enqueue(struct bridge_softc *sc, struct ifnet *dst_ifp, struct mbuf *m)
2024 {
2025 	int len, err = 0;
2026 	short mflags;
2027 	struct mbuf *m0;
2028 
2029 	/* We may be sending a fragment so traverse the mbuf */
2030 	for (; m; m = m0) {
2031 		m0 = m->m_nextpkt;
2032 		m->m_nextpkt = NULL;
2033 		len = m->m_pkthdr.len;
2034 		mflags = m->m_flags;
2035 
2036 		/*
2037 		 * If underlying interface can not do VLAN tag insertion itself
2038 		 * then attach a packet tag that holds it.
2039 		 */
2040 		if ((m->m_flags & M_VLANTAG) &&
2041 		    (dst_ifp->if_capenable & IFCAP_VLAN_HWTAGGING) == 0) {
2042 			m = ether_vlanencap(m, m->m_pkthdr.ether_vtag);
2043 			if (m == NULL) {
2044 				if_printf(dst_ifp,
2045 				    "unable to prepend VLAN header\n");
2046 				if_inc_counter(dst_ifp, IFCOUNTER_OERRORS, 1);
2047 				continue;
2048 			}
2049 			m->m_flags &= ~M_VLANTAG;
2050 		}
2051 
2052 		M_ASSERTPKTHDR(m); /* We shouldn't transmit mbuf without pkthdr */
2053 		if ((err = dst_ifp->if_transmit(dst_ifp, m))) {
2054 			int n;
2055 
2056 			for (m = m0, n = 1; m != NULL; m = m0, n++) {
2057 				m0 = m->m_nextpkt;
2058 				m_freem(m);
2059 			}
2060 			if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, n);
2061 			break;
2062 		}
2063 
2064 		if_inc_counter(sc->sc_ifp, IFCOUNTER_OPACKETS, 1);
2065 		if_inc_counter(sc->sc_ifp, IFCOUNTER_OBYTES, len);
2066 		if (mflags & M_MCAST)
2067 			if_inc_counter(sc->sc_ifp, IFCOUNTER_OMCASTS, 1);
2068 	}
2069 
2070 	return (err);
2071 }
2072 
2073 /*
2074  * bridge_dummynet:
2075  *
2076  * 	Receive a queued packet from dummynet and pass it on to the output
2077  * 	interface.
2078  *
2079  *	The mbuf has the Ethernet header already attached.
2080  */
2081 static void
2082 bridge_dummynet(struct mbuf *m, struct ifnet *ifp)
2083 {
2084 	struct bridge_softc *sc;
2085 
2086 	sc = ifp->if_bridge;
2087 
2088 	/*
2089 	 * The packet didnt originate from a member interface. This should only
2090 	 * ever happen if a member interface is removed while packets are
2091 	 * queued for it.
2092 	 */
2093 	if (sc == NULL) {
2094 		m_freem(m);
2095 		return;
2096 	}
2097 
2098 	if (PFIL_HOOKED_OUT(V_inet_pfil_head)
2099 #ifdef INET6
2100 	    || PFIL_HOOKED_OUT(V_inet6_pfil_head)
2101 #endif
2102 	    ) {
2103 		if (bridge_pfil(&m, sc->sc_ifp, ifp, PFIL_OUT) != 0)
2104 			return;
2105 		if (m == NULL)
2106 			return;
2107 	}
2108 
2109 	bridge_enqueue(sc, ifp, m);
2110 }
2111 
2112 /*
2113  * bridge_output:
2114  *
2115  *	Send output from a bridge member interface.  This
2116  *	performs the bridging function for locally originated
2117  *	packets.
2118  *
2119  *	The mbuf has the Ethernet header already attached.  We must
2120  *	enqueue or free the mbuf before returning.
2121  */
2122 static int
2123 bridge_output(struct ifnet *ifp, struct mbuf *m, struct sockaddr *sa,
2124     struct rtentry *rt)
2125 {
2126 	struct ether_header *eh;
2127 	struct ifnet *bifp, *dst_if;
2128 	struct bridge_softc *sc;
2129 	uint16_t vlan;
2130 
2131 	NET_EPOCH_ASSERT();
2132 
2133 	if (m->m_len < ETHER_HDR_LEN) {
2134 		m = m_pullup(m, ETHER_HDR_LEN);
2135 		if (m == NULL)
2136 			return (0);
2137 	}
2138 
2139 	eh = mtod(m, struct ether_header *);
2140 	sc = ifp->if_bridge;
2141 	vlan = VLANTAGOF(m);
2142 
2143 	bifp = sc->sc_ifp;
2144 
2145 	/*
2146 	 * If bridge is down, but the original output interface is up,
2147 	 * go ahead and send out that interface.  Otherwise, the packet
2148 	 * is dropped below.
2149 	 */
2150 	if ((bifp->if_drv_flags & IFF_DRV_RUNNING) == 0) {
2151 		dst_if = ifp;
2152 		goto sendunicast;
2153 	}
2154 
2155 	/*
2156 	 * If the packet is a multicast, or we don't know a better way to
2157 	 * get there, send to all interfaces.
2158 	 */
2159 	if (ETHER_IS_MULTICAST(eh->ether_dhost))
2160 		dst_if = NULL;
2161 	else
2162 		dst_if = bridge_rtlookup(sc, eh->ether_dhost, vlan);
2163 	/* Tap any traffic not passing back out the originating interface */
2164 	if (dst_if != ifp)
2165 		ETHER_BPF_MTAP(bifp, m);
2166 	if (dst_if == NULL) {
2167 		struct bridge_iflist *bif;
2168 		struct mbuf *mc;
2169 		int used = 0;
2170 
2171 		bridge_span(sc, m);
2172 
2173 		CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
2174 			dst_if = bif->bif_ifp;
2175 
2176 			if (dst_if->if_type == IFT_GIF)
2177 				continue;
2178 			if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2179 				continue;
2180 
2181 			/*
2182 			 * If this is not the original output interface,
2183 			 * and the interface is participating in spanning
2184 			 * tree, make sure the port is in a state that
2185 			 * allows forwarding.
2186 			 */
2187 			if (dst_if != ifp && (bif->bif_flags & IFBIF_STP) &&
2188 			    bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2189 				continue;
2190 
2191 			if (CK_LIST_NEXT(bif, bif_next) == NULL) {
2192 				used = 1;
2193 				mc = m;
2194 			} else {
2195 				mc = m_dup(m, M_NOWAIT);
2196 				if (mc == NULL) {
2197 					if_inc_counter(bifp, IFCOUNTER_OERRORS, 1);
2198 					continue;
2199 				}
2200 			}
2201 
2202 			bridge_enqueue(sc, dst_if, mc);
2203 		}
2204 		if (used == 0)
2205 			m_freem(m);
2206 		return (0);
2207 	}
2208 
2209 sendunicast:
2210 	/*
2211 	 * XXX Spanning tree consideration here?
2212 	 */
2213 
2214 	bridge_span(sc, m);
2215 	if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) {
2216 		m_freem(m);
2217 		return (0);
2218 	}
2219 
2220 	bridge_enqueue(sc, dst_if, m);
2221 	return (0);
2222 }
2223 
2224 /*
2225  * bridge_transmit:
2226  *
2227  *	Do output on a bridge.
2228  *
2229  */
2230 static int
2231 bridge_transmit(struct ifnet *ifp, struct mbuf *m)
2232 {
2233 	struct bridge_softc *sc;
2234 	struct ether_header *eh;
2235 	struct ifnet *dst_if;
2236 	int error = 0;
2237 
2238 	sc = ifp->if_softc;
2239 
2240 	ETHER_BPF_MTAP(ifp, m);
2241 
2242 	eh = mtod(m, struct ether_header *);
2243 
2244 	if (((m->m_flags & (M_BCAST|M_MCAST)) == 0) &&
2245 	    (dst_if = bridge_rtlookup(sc, eh->ether_dhost, 1)) != NULL) {
2246 		error = bridge_enqueue(sc, dst_if, m);
2247 	} else
2248 		bridge_broadcast(sc, ifp, m, 0);
2249 
2250 	return (error);
2251 }
2252 
2253 #ifdef ALTQ
2254 static void
2255 bridge_altq_start(if_t ifp)
2256 {
2257 	struct ifaltq *ifq = &ifp->if_snd;
2258 	struct mbuf *m;
2259 
2260 	IFQ_LOCK(ifq);
2261 	IFQ_DEQUEUE_NOLOCK(ifq, m);
2262 	while (m != NULL) {
2263 		bridge_transmit(ifp, m);
2264 		IFQ_DEQUEUE_NOLOCK(ifq, m);
2265 	}
2266 	IFQ_UNLOCK(ifq);
2267 }
2268 
2269 static int
2270 bridge_altq_transmit(if_t ifp, struct mbuf *m)
2271 {
2272 	int err;
2273 
2274 	if (ALTQ_IS_ENABLED(&ifp->if_snd)) {
2275 		IFQ_ENQUEUE(&ifp->if_snd, m, err);
2276 		if (err == 0)
2277 			bridge_altq_start(ifp);
2278 	} else
2279 		err = bridge_transmit(ifp, m);
2280 
2281 	return (err);
2282 }
2283 #endif	/* ALTQ */
2284 
2285 /*
2286  * The ifp->if_qflush entry point for if_bridge(4) is no-op.
2287  */
2288 static void
2289 bridge_qflush(struct ifnet *ifp __unused)
2290 {
2291 }
2292 
2293 /*
2294  * bridge_forward:
2295  *
2296  *	The forwarding function of the bridge.
2297  *
2298  *	NOTE: Releases the lock on return.
2299  */
2300 static void
2301 bridge_forward(struct bridge_softc *sc, struct bridge_iflist *sbif,
2302     struct mbuf *m)
2303 {
2304 	struct bridge_iflist *dbif;
2305 	struct ifnet *src_if, *dst_if, *ifp;
2306 	struct ether_header *eh;
2307 	uint16_t vlan;
2308 	uint8_t *dst;
2309 	int error;
2310 
2311 	NET_EPOCH_ASSERT();
2312 
2313 	src_if = m->m_pkthdr.rcvif;
2314 	ifp = sc->sc_ifp;
2315 
2316 	if_inc_counter(ifp, IFCOUNTER_IPACKETS, 1);
2317 	if_inc_counter(ifp, IFCOUNTER_IBYTES, m->m_pkthdr.len);
2318 	vlan = VLANTAGOF(m);
2319 
2320 	if ((sbif->bif_flags & IFBIF_STP) &&
2321 	    sbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2322 		goto drop;
2323 
2324 	eh = mtod(m, struct ether_header *);
2325 	dst = eh->ether_dhost;
2326 
2327 	/* If the interface is learning, record the address. */
2328 	if (sbif->bif_flags & IFBIF_LEARNING) {
2329 		error = bridge_rtupdate(sc, eh->ether_shost, vlan,
2330 		    sbif, 0, IFBAF_DYNAMIC);
2331 		/*
2332 		 * If the interface has addresses limits then deny any source
2333 		 * that is not in the cache.
2334 		 */
2335 		if (error && sbif->bif_addrmax)
2336 			goto drop;
2337 	}
2338 
2339 	if ((sbif->bif_flags & IFBIF_STP) != 0 &&
2340 	    sbif->bif_stp.bp_state == BSTP_IFSTATE_LEARNING)
2341 		goto drop;
2342 
2343 	/*
2344 	 * At this point, the port either doesn't participate
2345 	 * in spanning tree or it is in the forwarding state.
2346 	 */
2347 
2348 	/*
2349 	 * If the packet is unicast, destined for someone on
2350 	 * "this" side of the bridge, drop it.
2351 	 */
2352 	if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) {
2353 		dst_if = bridge_rtlookup(sc, dst, vlan);
2354 		if (src_if == dst_if)
2355 			goto drop;
2356 	} else {
2357 		/*
2358 		 * Check if its a reserved multicast address, any address
2359 		 * listed in 802.1D section 7.12.6 may not be forwarded by the
2360 		 * bridge.
2361 		 * This is currently 01-80-C2-00-00-00 to 01-80-C2-00-00-0F
2362 		 */
2363 		if (dst[0] == 0x01 && dst[1] == 0x80 &&
2364 		    dst[2] == 0xc2 && dst[3] == 0x00 &&
2365 		    dst[4] == 0x00 && dst[5] <= 0x0f)
2366 			goto drop;
2367 
2368 		/* ...forward it to all interfaces. */
2369 		if_inc_counter(ifp, IFCOUNTER_IMCASTS, 1);
2370 		dst_if = NULL;
2371 	}
2372 
2373 	/*
2374 	 * If we have a destination interface which is a member of our bridge,
2375 	 * OR this is a unicast packet, push it through the bpf(4) machinery.
2376 	 * For broadcast or multicast packets, don't bother because it will
2377 	 * be reinjected into ether_input. We do this before we pass the packets
2378 	 * through the pfil(9) framework, as it is possible that pfil(9) will
2379 	 * drop the packet, or possibly modify it, making it difficult to debug
2380 	 * firewall issues on the bridge.
2381 	 */
2382 	if (dst_if != NULL || (m->m_flags & (M_BCAST | M_MCAST)) == 0)
2383 		ETHER_BPF_MTAP(ifp, m);
2384 
2385 	/* run the packet filter */
2386 	if (PFIL_HOOKED_IN(V_inet_pfil_head)
2387 #ifdef INET6
2388 	    || PFIL_HOOKED_IN(V_inet6_pfil_head)
2389 #endif
2390 	    ) {
2391 		if (bridge_pfil(&m, ifp, src_if, PFIL_IN) != 0)
2392 			return;
2393 		if (m == NULL)
2394 			return;
2395 	}
2396 
2397 	if (dst_if == NULL) {
2398 		bridge_broadcast(sc, src_if, m, 1);
2399 		return;
2400 	}
2401 
2402 	/*
2403 	 * At this point, we're dealing with a unicast frame
2404 	 * going to a different interface.
2405 	 */
2406 	if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2407 		goto drop;
2408 
2409 	dbif = bridge_lookup_member_if(sc, dst_if);
2410 	if (dbif == NULL)
2411 		/* Not a member of the bridge (anymore?) */
2412 		goto drop;
2413 
2414 	/* Private segments can not talk to each other */
2415 	if (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE)
2416 		goto drop;
2417 
2418 	if ((dbif->bif_flags & IFBIF_STP) &&
2419 	    dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2420 		goto drop;
2421 
2422 	if (PFIL_HOOKED_OUT(V_inet_pfil_head)
2423 #ifdef INET6
2424 	    || PFIL_HOOKED_OUT(V_inet6_pfil_head)
2425 #endif
2426 	    ) {
2427 		if (bridge_pfil(&m, ifp, dst_if, PFIL_OUT) != 0)
2428 			return;
2429 		if (m == NULL)
2430 			return;
2431 	}
2432 
2433 	bridge_enqueue(sc, dst_if, m);
2434 	return;
2435 
2436 drop:
2437 	m_freem(m);
2438 }
2439 
2440 /*
2441  * bridge_input:
2442  *
2443  *	Receive input from a member interface.  Queue the packet for
2444  *	bridging if it is not for us.
2445  */
2446 static struct mbuf *
2447 bridge_input(struct ifnet *ifp, struct mbuf *m)
2448 {
2449 	struct bridge_softc *sc = ifp->if_bridge;
2450 	struct bridge_iflist *bif, *bif2;
2451 	struct ifnet *bifp;
2452 	struct ether_header *eh;
2453 	struct mbuf *mc, *mc2;
2454 	uint16_t vlan;
2455 	int error;
2456 
2457 	NET_EPOCH_ASSERT();
2458 
2459 	if ((sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
2460 		return (m);
2461 
2462 	bifp = sc->sc_ifp;
2463 	vlan = VLANTAGOF(m);
2464 
2465 	/*
2466 	 * Implement support for bridge monitoring. If this flag has been
2467 	 * set on this interface, discard the packet once we push it through
2468 	 * the bpf(4) machinery, but before we do, increment the byte and
2469 	 * packet counters associated with this interface.
2470 	 */
2471 	if ((bifp->if_flags & IFF_MONITOR) != 0) {
2472 		m->m_pkthdr.rcvif  = bifp;
2473 		ETHER_BPF_MTAP(bifp, m);
2474 		if_inc_counter(bifp, IFCOUNTER_IPACKETS, 1);
2475 		if_inc_counter(bifp, IFCOUNTER_IBYTES, m->m_pkthdr.len);
2476 		m_freem(m);
2477 		return (NULL);
2478 	}
2479 	bif = bridge_lookup_member_if(sc, ifp);
2480 	if (bif == NULL) {
2481 		return (m);
2482 	}
2483 
2484 	eh = mtod(m, struct ether_header *);
2485 
2486 	bridge_span(sc, m);
2487 
2488 	if (m->m_flags & (M_BCAST|M_MCAST)) {
2489 		/* Tap off 802.1D packets; they do not get forwarded. */
2490 		if (memcmp(eh->ether_dhost, bstp_etheraddr,
2491 		    ETHER_ADDR_LEN) == 0) {
2492 			bstp_input(&bif->bif_stp, ifp, m); /* consumes mbuf */
2493 			return (NULL);
2494 		}
2495 
2496 		if ((bif->bif_flags & IFBIF_STP) &&
2497 		    bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) {
2498 			return (m);
2499 		}
2500 
2501 		/*
2502 		 * Make a deep copy of the packet and enqueue the copy
2503 		 * for bridge processing; return the original packet for
2504 		 * local processing.
2505 		 */
2506 		mc = m_dup(m, M_NOWAIT);
2507 		if (mc == NULL) {
2508 			return (m);
2509 		}
2510 
2511 		/* Perform the bridge forwarding function with the copy. */
2512 		bridge_forward(sc, bif, mc);
2513 
2514 		/*
2515 		 * Reinject the mbuf as arriving on the bridge so we have a
2516 		 * chance at claiming multicast packets. We can not loop back
2517 		 * here from ether_input as a bridge is never a member of a
2518 		 * bridge.
2519 		 */
2520 		KASSERT(bifp->if_bridge == NULL,
2521 		    ("loop created in bridge_input"));
2522 		mc2 = m_dup(m, M_NOWAIT);
2523 		if (mc2 != NULL) {
2524 			/* Keep the layer3 header aligned */
2525 			int i = min(mc2->m_pkthdr.len, max_protohdr);
2526 			mc2 = m_copyup(mc2, i, ETHER_ALIGN);
2527 		}
2528 		if (mc2 != NULL) {
2529 			mc2->m_pkthdr.rcvif = bifp;
2530 			(*bifp->if_input)(bifp, mc2);
2531 		}
2532 
2533 		/* Return the original packet for local processing. */
2534 		return (m);
2535 	}
2536 
2537 	if ((bif->bif_flags & IFBIF_STP) &&
2538 	    bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) {
2539 		return (m);
2540 	}
2541 
2542 #if defined(INET) || defined(INET6)
2543 #define	CARP_CHECK_WE_ARE_DST(iface) \
2544 	((iface)->if_carp && (*carp_forus_p)((iface), eh->ether_dhost))
2545 #define	CARP_CHECK_WE_ARE_SRC(iface) \
2546 	((iface)->if_carp && (*carp_forus_p)((iface), eh->ether_shost))
2547 #else
2548 #define	CARP_CHECK_WE_ARE_DST(iface)	false
2549 #define	CARP_CHECK_WE_ARE_SRC(iface)	false
2550 #endif
2551 
2552 #ifdef INET6
2553 #define	PFIL_HOOKED_INET6	PFIL_HOOKED_IN(V_inet6_pfil_head)
2554 #else
2555 #define	PFIL_HOOKED_INET6	false
2556 #endif
2557 
2558 #define GRAB_OUR_PACKETS(iface)						\
2559 	if ((iface)->if_type == IFT_GIF)				\
2560 		continue;						\
2561 	/* It is destined for us. */					\
2562 	if (memcmp(IF_LLADDR(iface), eh->ether_dhost, ETHER_ADDR_LEN) == 0 || \
2563 	    CARP_CHECK_WE_ARE_DST(iface)) {				\
2564 		if (bif->bif_flags & IFBIF_LEARNING) {			\
2565 			error = bridge_rtupdate(sc, eh->ether_shost,	\
2566 			    vlan, bif, 0, IFBAF_DYNAMIC);		\
2567 			if (error && bif->bif_addrmax) {		\
2568 				m_freem(m);				\
2569 				return (NULL);				\
2570 			}						\
2571 		}							\
2572 		m->m_pkthdr.rcvif = iface;				\
2573 		if ((iface) == ifp) {					\
2574 			/* Skip bridge processing... src == dest */	\
2575 			return (m);					\
2576 		}							\
2577 		/* It's passing over or to the bridge, locally. */	\
2578 		ETHER_BPF_MTAP(bifp, m);				\
2579 		if_inc_counter(bifp, IFCOUNTER_IPACKETS, 1);		\
2580 		if_inc_counter(bifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); \
2581 		/* Filter on the physical interface. */			\
2582 		if (V_pfil_local_phys && (PFIL_HOOKED_IN(V_inet_pfil_head) || \
2583 		    PFIL_HOOKED_INET6)) {				\
2584 			if (bridge_pfil(&m, NULL, ifp,			\
2585 			    PFIL_IN) != 0 || m == NULL) {		\
2586 				return (NULL);				\
2587 			}						\
2588 		}							\
2589 		if ((iface) != bifp)					\
2590 			ETHER_BPF_MTAP(iface, m);			\
2591 		return (m);						\
2592 	}								\
2593 									\
2594 	/* We just received a packet that we sent out. */		\
2595 	if (memcmp(IF_LLADDR(iface), eh->ether_shost, ETHER_ADDR_LEN) == 0 || \
2596 	    CARP_CHECK_WE_ARE_SRC(iface)) {				\
2597 		m_freem(m);						\
2598 		return (NULL);						\
2599 	}
2600 
2601 	/*
2602 	 * Unicast.  Make sure it's not for the bridge.
2603 	 */
2604 	do { GRAB_OUR_PACKETS(bifp) } while (0);
2605 
2606 	/*
2607 	 * Give a chance for ifp at first priority. This will help when	the
2608 	 * packet comes through the interface like VLAN's with the same MACs
2609 	 * on several interfaces from the same bridge. This also will save
2610 	 * some CPU cycles in case the destination interface and the input
2611 	 * interface (eq ifp) are the same.
2612 	 */
2613 	do { GRAB_OUR_PACKETS(ifp) } while (0);
2614 
2615 	/* Now check the all bridge members. */
2616 	CK_LIST_FOREACH(bif2, &sc->sc_iflist, bif_next) {
2617 		GRAB_OUR_PACKETS(bif2->bif_ifp)
2618 	}
2619 
2620 #undef CARP_CHECK_WE_ARE_DST
2621 #undef CARP_CHECK_WE_ARE_SRC
2622 #undef PFIL_HOOKED_INET6
2623 #undef GRAB_OUR_PACKETS
2624 
2625 	/* Perform the bridge forwarding function. */
2626 	bridge_forward(sc, bif, m);
2627 
2628 	return (NULL);
2629 }
2630 
2631 /*
2632  * bridge_broadcast:
2633  *
2634  *	Send a frame to all interfaces that are members of
2635  *	the bridge, except for the one on which the packet
2636  *	arrived.
2637  *
2638  *	NOTE: Releases the lock on return.
2639  */
2640 static void
2641 bridge_broadcast(struct bridge_softc *sc, struct ifnet *src_if,
2642     struct mbuf *m, int runfilt)
2643 {
2644 	struct bridge_iflist *dbif, *sbif;
2645 	struct mbuf *mc;
2646 	struct ifnet *dst_if;
2647 	int used = 0, i;
2648 
2649 	NET_EPOCH_ASSERT();
2650 
2651 	sbif = bridge_lookup_member_if(sc, src_if);
2652 
2653 	/* Filter on the bridge interface before broadcasting */
2654 	if (runfilt && (PFIL_HOOKED_OUT(V_inet_pfil_head)
2655 #ifdef INET6
2656 	    || PFIL_HOOKED_OUT(V_inet6_pfil_head)
2657 #endif
2658 	    )) {
2659 		if (bridge_pfil(&m, sc->sc_ifp, NULL, PFIL_OUT) != 0)
2660 			return;
2661 		if (m == NULL)
2662 			return;
2663 	}
2664 
2665 	CK_LIST_FOREACH(dbif, &sc->sc_iflist, bif_next) {
2666 		dst_if = dbif->bif_ifp;
2667 		if (dst_if == src_if)
2668 			continue;
2669 
2670 		/* Private segments can not talk to each other */
2671 		if (sbif && (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE))
2672 			continue;
2673 
2674 		if ((dbif->bif_flags & IFBIF_STP) &&
2675 		    dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2676 			continue;
2677 
2678 		if ((dbif->bif_flags & IFBIF_DISCOVER) == 0 &&
2679 		    (m->m_flags & (M_BCAST|M_MCAST)) == 0)
2680 			continue;
2681 
2682 		if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2683 			continue;
2684 
2685 		if (CK_LIST_NEXT(dbif, bif_next) == NULL) {
2686 			mc = m;
2687 			used = 1;
2688 		} else {
2689 			mc = m_dup(m, M_NOWAIT);
2690 			if (mc == NULL) {
2691 				if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1);
2692 				continue;
2693 			}
2694 		}
2695 
2696 		/*
2697 		 * Filter on the output interface. Pass a NULL bridge interface
2698 		 * pointer so we do not redundantly filter on the bridge for
2699 		 * each interface we broadcast on.
2700 		 */
2701 		if (runfilt && (PFIL_HOOKED_OUT(V_inet_pfil_head)
2702 #ifdef INET6
2703 		    || PFIL_HOOKED_OUT(V_inet6_pfil_head)
2704 #endif
2705 		    )) {
2706 			if (used == 0) {
2707 				/* Keep the layer3 header aligned */
2708 				i = min(mc->m_pkthdr.len, max_protohdr);
2709 				mc = m_copyup(mc, i, ETHER_ALIGN);
2710 				if (mc == NULL) {
2711 					if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1);
2712 					continue;
2713 				}
2714 			}
2715 			if (bridge_pfil(&mc, NULL, dst_if, PFIL_OUT) != 0)
2716 				continue;
2717 			if (mc == NULL)
2718 				continue;
2719 		}
2720 
2721 		bridge_enqueue(sc, dst_if, mc);
2722 	}
2723 	if (used == 0)
2724 		m_freem(m);
2725 }
2726 
2727 /*
2728  * bridge_span:
2729  *
2730  *	Duplicate a packet out one or more interfaces that are in span mode,
2731  *	the original mbuf is unmodified.
2732  */
2733 static void
2734 bridge_span(struct bridge_softc *sc, struct mbuf *m)
2735 {
2736 	struct bridge_iflist *bif;
2737 	struct ifnet *dst_if;
2738 	struct mbuf *mc;
2739 
2740 	NET_EPOCH_ASSERT();
2741 
2742 	if (CK_LIST_EMPTY(&sc->sc_spanlist))
2743 		return;
2744 
2745 	CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) {
2746 		dst_if = bif->bif_ifp;
2747 
2748 		if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2749 			continue;
2750 
2751 		mc = m_dup(m, M_NOWAIT);
2752 		if (mc == NULL) {
2753 			if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1);
2754 			continue;
2755 		}
2756 
2757 		bridge_enqueue(sc, dst_if, mc);
2758 	}
2759 }
2760 
2761 /*
2762  * bridge_rtupdate:
2763  *
2764  *	Add a bridge routing entry.
2765  */
2766 static int
2767 bridge_rtupdate(struct bridge_softc *sc, const uint8_t *dst, uint16_t vlan,
2768     struct bridge_iflist *bif, int setflags, uint8_t flags)
2769 {
2770 	struct bridge_rtnode *brt;
2771 	int error;
2772 
2773 	BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc);
2774 
2775 	/* Check the source address is valid and not multicast. */
2776 	if (ETHER_IS_MULTICAST(dst) ||
2777 	    (dst[0] == 0 && dst[1] == 0 && dst[2] == 0 &&
2778 	     dst[3] == 0 && dst[4] == 0 && dst[5] == 0) != 0)
2779 		return (EINVAL);
2780 
2781 	/* 802.1p frames map to vlan 1 */
2782 	if (vlan == 0)
2783 		vlan = 1;
2784 
2785 	/*
2786 	 * A route for this destination might already exist.  If so,
2787 	 * update it, otherwise create a new one.
2788 	 */
2789 	if ((brt = bridge_rtnode_lookup(sc, dst, vlan)) == NULL) {
2790 		BRIDGE_RT_LOCK(sc);
2791 
2792 		/* Check again, now that we have the lock. There could have
2793 		 * been a race and we only want to insert this once. */
2794 		if ((brt = bridge_rtnode_lookup(sc, dst, vlan)) != NULL) {
2795 			BRIDGE_RT_UNLOCK(sc);
2796 			return (0);
2797 		}
2798 
2799 		if (sc->sc_brtcnt >= sc->sc_brtmax) {
2800 			sc->sc_brtexceeded++;
2801 			BRIDGE_RT_UNLOCK(sc);
2802 			return (ENOSPC);
2803 		}
2804 		/* Check per interface address limits (if enabled) */
2805 		if (bif->bif_addrmax && bif->bif_addrcnt >= bif->bif_addrmax) {
2806 			bif->bif_addrexceeded++;
2807 			BRIDGE_RT_UNLOCK(sc);
2808 			return (ENOSPC);
2809 		}
2810 
2811 		/*
2812 		 * Allocate a new bridge forwarding node, and
2813 		 * initialize the expiration time and Ethernet
2814 		 * address.
2815 		 */
2816 		brt = uma_zalloc(V_bridge_rtnode_zone, M_NOWAIT | M_ZERO);
2817 		if (brt == NULL) {
2818 			BRIDGE_RT_UNLOCK(sc);
2819 			return (ENOMEM);
2820 		}
2821 		brt->brt_vnet = curvnet;
2822 
2823 		if (bif->bif_flags & IFBIF_STICKY)
2824 			brt->brt_flags = IFBAF_STICKY;
2825 		else
2826 			brt->brt_flags = IFBAF_DYNAMIC;
2827 
2828 		memcpy(brt->brt_addr, dst, ETHER_ADDR_LEN);
2829 		brt->brt_vlan = vlan;
2830 
2831 		if ((error = bridge_rtnode_insert(sc, brt)) != 0) {
2832 			uma_zfree(V_bridge_rtnode_zone, brt);
2833 			BRIDGE_RT_UNLOCK(sc);
2834 			return (error);
2835 		}
2836 		brt->brt_dst = bif;
2837 		bif->bif_addrcnt++;
2838 
2839 		BRIDGE_RT_UNLOCK(sc);
2840 	}
2841 
2842 	if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC &&
2843 	    brt->brt_dst != bif) {
2844 		BRIDGE_RT_LOCK(sc);
2845 		brt->brt_dst->bif_addrcnt--;
2846 		brt->brt_dst = bif;
2847 		brt->brt_dst->bif_addrcnt++;
2848 		BRIDGE_RT_UNLOCK(sc);
2849 	}
2850 
2851 	if ((flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2852 		brt->brt_expire = time_uptime + sc->sc_brttimeout;
2853 	if (setflags)
2854 		brt->brt_flags = flags;
2855 
2856 	return (0);
2857 }
2858 
2859 /*
2860  * bridge_rtlookup:
2861  *
2862  *	Lookup the destination interface for an address.
2863  */
2864 static struct ifnet *
2865 bridge_rtlookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2866 {
2867 	struct bridge_rtnode *brt;
2868 
2869 	NET_EPOCH_ASSERT();
2870 
2871 	if ((brt = bridge_rtnode_lookup(sc, addr, vlan)) == NULL)
2872 		return (NULL);
2873 
2874 	return (brt->brt_ifp);
2875 }
2876 
2877 /*
2878  * bridge_rttrim:
2879  *
2880  *	Trim the routine table so that we have a number
2881  *	of routing entries less than or equal to the
2882  *	maximum number.
2883  */
2884 static void
2885 bridge_rttrim(struct bridge_softc *sc)
2886 {
2887 	struct bridge_rtnode *brt, *nbrt;
2888 
2889 	NET_EPOCH_ASSERT();
2890 	BRIDGE_RT_LOCK_ASSERT(sc);
2891 
2892 	/* Make sure we actually need to do this. */
2893 	if (sc->sc_brtcnt <= sc->sc_brtmax)
2894 		return;
2895 
2896 	/* Force an aging cycle; this might trim enough addresses. */
2897 	bridge_rtage(sc);
2898 	if (sc->sc_brtcnt <= sc->sc_brtmax)
2899 		return;
2900 
2901 	CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2902 		if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) {
2903 			bridge_rtnode_destroy(sc, brt);
2904 			if (sc->sc_brtcnt <= sc->sc_brtmax)
2905 				return;
2906 		}
2907 	}
2908 }
2909 
2910 /*
2911  * bridge_timer:
2912  *
2913  *	Aging timer for the bridge.
2914  */
2915 static void
2916 bridge_timer(void *arg)
2917 {
2918 	struct bridge_softc *sc = arg;
2919 
2920 	BRIDGE_RT_LOCK_ASSERT(sc);
2921 
2922 	/* Destruction of rtnodes requires a proper vnet context */
2923 	CURVNET_SET(sc->sc_ifp->if_vnet);
2924 	bridge_rtage(sc);
2925 
2926 	if (sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING)
2927 		callout_reset(&sc->sc_brcallout,
2928 		    bridge_rtable_prune_period * hz, bridge_timer, sc);
2929 	CURVNET_RESTORE();
2930 }
2931 
2932 /*
2933  * bridge_rtage:
2934  *
2935  *	Perform an aging cycle.
2936  */
2937 static void
2938 bridge_rtage(struct bridge_softc *sc)
2939 {
2940 	struct bridge_rtnode *brt, *nbrt;
2941 
2942 	BRIDGE_RT_LOCK_ASSERT(sc);
2943 
2944 	CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2945 		if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) {
2946 			if (time_uptime >= brt->brt_expire)
2947 				bridge_rtnode_destroy(sc, brt);
2948 		}
2949 	}
2950 }
2951 
2952 /*
2953  * bridge_rtflush:
2954  *
2955  *	Remove all dynamic addresses from the bridge.
2956  */
2957 static void
2958 bridge_rtflush(struct bridge_softc *sc, int full)
2959 {
2960 	struct bridge_rtnode *brt, *nbrt;
2961 
2962 	BRIDGE_RT_LOCK_ASSERT(sc);
2963 
2964 	CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2965 		if (full || (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2966 			bridge_rtnode_destroy(sc, brt);
2967 	}
2968 }
2969 
2970 /*
2971  * bridge_rtdaddr:
2972  *
2973  *	Remove an address from the table.
2974  */
2975 static int
2976 bridge_rtdaddr(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2977 {
2978 	struct bridge_rtnode *brt;
2979 	int found = 0;
2980 
2981 	BRIDGE_RT_LOCK(sc);
2982 
2983 	/*
2984 	 * If vlan is zero then we want to delete for all vlans so the lookup
2985 	 * may return more than one.
2986 	 */
2987 	while ((brt = bridge_rtnode_lookup(sc, addr, vlan)) != NULL) {
2988 		bridge_rtnode_destroy(sc, brt);
2989 		found = 1;
2990 	}
2991 
2992 	BRIDGE_RT_UNLOCK(sc);
2993 
2994 	return (found ? 0 : ENOENT);
2995 }
2996 
2997 /*
2998  * bridge_rtdelete:
2999  *
3000  *	Delete routes to a speicifc member interface.
3001  */
3002 static void
3003 bridge_rtdelete(struct bridge_softc *sc, struct ifnet *ifp, int full)
3004 {
3005 	struct bridge_rtnode *brt, *nbrt;
3006 
3007 	BRIDGE_RT_LOCK_ASSERT(sc);
3008 
3009 	CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
3010 		if (brt->brt_ifp == ifp && (full ||
3011 			    (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC))
3012 			bridge_rtnode_destroy(sc, brt);
3013 	}
3014 }
3015 
3016 /*
3017  * bridge_rtable_init:
3018  *
3019  *	Initialize the route table for this bridge.
3020  */
3021 static void
3022 bridge_rtable_init(struct bridge_softc *sc)
3023 {
3024 	int i;
3025 
3026 	sc->sc_rthash = malloc(sizeof(*sc->sc_rthash) * BRIDGE_RTHASH_SIZE,
3027 	    M_DEVBUF, M_WAITOK);
3028 
3029 	for (i = 0; i < BRIDGE_RTHASH_SIZE; i++)
3030 		CK_LIST_INIT(&sc->sc_rthash[i]);
3031 
3032 	sc->sc_rthash_key = arc4random();
3033 	CK_LIST_INIT(&sc->sc_rtlist);
3034 }
3035 
3036 /*
3037  * bridge_rtable_fini:
3038  *
3039  *	Deconstruct the route table for this bridge.
3040  */
3041 static void
3042 bridge_rtable_fini(struct bridge_softc *sc)
3043 {
3044 
3045 	KASSERT(sc->sc_brtcnt == 0,
3046 	    ("%s: %d bridge routes referenced", __func__, sc->sc_brtcnt));
3047 	free(sc->sc_rthash, M_DEVBUF);
3048 }
3049 
3050 /*
3051  * The following hash function is adapted from "Hash Functions" by Bob Jenkins
3052  * ("Algorithm Alley", Dr. Dobbs Journal, September 1997).
3053  */
3054 #define	mix(a, b, c)							\
3055 do {									\
3056 	a -= b; a -= c; a ^= (c >> 13);					\
3057 	b -= c; b -= a; b ^= (a << 8);					\
3058 	c -= a; c -= b; c ^= (b >> 13);					\
3059 	a -= b; a -= c; a ^= (c >> 12);					\
3060 	b -= c; b -= a; b ^= (a << 16);					\
3061 	c -= a; c -= b; c ^= (b >> 5);					\
3062 	a -= b; a -= c; a ^= (c >> 3);					\
3063 	b -= c; b -= a; b ^= (a << 10);					\
3064 	c -= a; c -= b; c ^= (b >> 15);					\
3065 } while (/*CONSTCOND*/0)
3066 
3067 static __inline uint32_t
3068 bridge_rthash(struct bridge_softc *sc, const uint8_t *addr)
3069 {
3070 	uint32_t a = 0x9e3779b9, b = 0x9e3779b9, c = sc->sc_rthash_key;
3071 
3072 	b += addr[5] << 8;
3073 	b += addr[4];
3074 	a += addr[3] << 24;
3075 	a += addr[2] << 16;
3076 	a += addr[1] << 8;
3077 	a += addr[0];
3078 
3079 	mix(a, b, c);
3080 
3081 	return (c & BRIDGE_RTHASH_MASK);
3082 }
3083 
3084 #undef mix
3085 
3086 static int
3087 bridge_rtnode_addr_cmp(const uint8_t *a, const uint8_t *b)
3088 {
3089 	int i, d;
3090 
3091 	for (i = 0, d = 0; i < ETHER_ADDR_LEN && d == 0; i++) {
3092 		d = ((int)a[i]) - ((int)b[i]);
3093 	}
3094 
3095 	return (d);
3096 }
3097 
3098 /*
3099  * bridge_rtnode_lookup:
3100  *
3101  *	Look up a bridge route node for the specified destination. Compare the
3102  *	vlan id or if zero then just return the first match.
3103  */
3104 static struct bridge_rtnode *
3105 bridge_rtnode_lookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
3106 {
3107 	struct bridge_rtnode *brt;
3108 	uint32_t hash;
3109 	int dir;
3110 
3111 	BRIDGE_RT_LOCK_OR_NET_EPOCH_ASSERT(sc);
3112 
3113 	hash = bridge_rthash(sc, addr);
3114 	CK_LIST_FOREACH(brt, &sc->sc_rthash[hash], brt_hash) {
3115 		dir = bridge_rtnode_addr_cmp(addr, brt->brt_addr);
3116 		if (dir == 0 && (brt->brt_vlan == vlan || vlan == 0))
3117 			return (brt);
3118 		if (dir > 0)
3119 			return (NULL);
3120 	}
3121 
3122 	return (NULL);
3123 }
3124 
3125 /*
3126  * bridge_rtnode_insert:
3127  *
3128  *	Insert the specified bridge node into the route table.  We
3129  *	assume the entry is not already in the table.
3130  */
3131 static int
3132 bridge_rtnode_insert(struct bridge_softc *sc, struct bridge_rtnode *brt)
3133 {
3134 	struct bridge_rtnode *lbrt;
3135 	uint32_t hash;
3136 	int dir;
3137 
3138 	BRIDGE_RT_LOCK_ASSERT(sc);
3139 
3140 	hash = bridge_rthash(sc, brt->brt_addr);
3141 
3142 	lbrt = CK_LIST_FIRST(&sc->sc_rthash[hash]);
3143 	if (lbrt == NULL) {
3144 		CK_LIST_INSERT_HEAD(&sc->sc_rthash[hash], brt, brt_hash);
3145 		goto out;
3146 	}
3147 
3148 	do {
3149 		dir = bridge_rtnode_addr_cmp(brt->brt_addr, lbrt->brt_addr);
3150 		if (dir == 0 && brt->brt_vlan == lbrt->brt_vlan)
3151 			return (EEXIST);
3152 		if (dir > 0) {
3153 			CK_LIST_INSERT_BEFORE(lbrt, brt, brt_hash);
3154 			goto out;
3155 		}
3156 		if (CK_LIST_NEXT(lbrt, brt_hash) == NULL) {
3157 			CK_LIST_INSERT_AFTER(lbrt, brt, brt_hash);
3158 			goto out;
3159 		}
3160 		lbrt = CK_LIST_NEXT(lbrt, brt_hash);
3161 	} while (lbrt != NULL);
3162 
3163 #ifdef DIAGNOSTIC
3164 	panic("bridge_rtnode_insert: impossible");
3165 #endif
3166 
3167 out:
3168 	CK_LIST_INSERT_HEAD(&sc->sc_rtlist, brt, brt_list);
3169 	sc->sc_brtcnt++;
3170 
3171 	return (0);
3172 }
3173 
3174 static void
3175 bridge_rtnode_destroy_cb(struct epoch_context *ctx)
3176 {
3177 	struct bridge_rtnode *brt;
3178 
3179 	brt = __containerof(ctx, struct bridge_rtnode, brt_epoch_ctx);
3180 
3181 	CURVNET_SET(brt->brt_vnet);
3182 	uma_zfree(V_bridge_rtnode_zone, brt);
3183 	CURVNET_RESTORE();
3184 }
3185 
3186 /*
3187  * bridge_rtnode_destroy:
3188  *
3189  *	Destroy a bridge rtnode.
3190  */
3191 static void
3192 bridge_rtnode_destroy(struct bridge_softc *sc, struct bridge_rtnode *brt)
3193 {
3194 	BRIDGE_RT_LOCK_ASSERT(sc);
3195 
3196 	CK_LIST_REMOVE(brt, brt_hash);
3197 
3198 	CK_LIST_REMOVE(brt, brt_list);
3199 	sc->sc_brtcnt--;
3200 	brt->brt_dst->bif_addrcnt--;
3201 
3202 	NET_EPOCH_CALL(bridge_rtnode_destroy_cb, &brt->brt_epoch_ctx);
3203 }
3204 
3205 /*
3206  * bridge_rtable_expire:
3207  *
3208  *	Set the expiry time for all routes on an interface.
3209  */
3210 static void
3211 bridge_rtable_expire(struct ifnet *ifp, int age)
3212 {
3213 	struct bridge_softc *sc = ifp->if_bridge;
3214 	struct bridge_rtnode *brt;
3215 
3216 	CURVNET_SET(ifp->if_vnet);
3217 	BRIDGE_RT_LOCK(sc);
3218 
3219 	/*
3220 	 * If the age is zero then flush, otherwise set all the expiry times to
3221 	 * age for the interface
3222 	 */
3223 	if (age == 0)
3224 		bridge_rtdelete(sc, ifp, IFBF_FLUSHDYN);
3225 	else {
3226 		CK_LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) {
3227 			/* Cap the expiry time to 'age' */
3228 			if (brt->brt_ifp == ifp &&
3229 			    brt->brt_expire > time_uptime + age &&
3230 			    (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
3231 				brt->brt_expire = time_uptime + age;
3232 		}
3233 	}
3234 	BRIDGE_RT_UNLOCK(sc);
3235 	CURVNET_RESTORE();
3236 }
3237 
3238 /*
3239  * bridge_state_change:
3240  *
3241  *	Callback from the bridgestp code when a port changes states.
3242  */
3243 static void
3244 bridge_state_change(struct ifnet *ifp, int state)
3245 {
3246 	struct bridge_softc *sc = ifp->if_bridge;
3247 	static const char *stpstates[] = {
3248 		"disabled",
3249 		"listening",
3250 		"learning",
3251 		"forwarding",
3252 		"blocking",
3253 		"discarding"
3254 	};
3255 
3256 	CURVNET_SET(ifp->if_vnet);
3257 	if (V_log_stp)
3258 		log(LOG_NOTICE, "%s: state changed to %s on %s\n",
3259 		    sc->sc_ifp->if_xname, stpstates[state], ifp->if_xname);
3260 	CURVNET_RESTORE();
3261 }
3262 
3263 /*
3264  * Send bridge packets through pfil if they are one of the types pfil can deal
3265  * with, or if they are ARP or REVARP.  (pfil will pass ARP and REVARP without
3266  * question.) If *bifp or *ifp are NULL then packet filtering is skipped for
3267  * that interface.
3268  */
3269 static int
3270 bridge_pfil(struct mbuf **mp, struct ifnet *bifp, struct ifnet *ifp, int dir)
3271 {
3272 	int snap, error, i, hlen;
3273 	struct ether_header *eh1, eh2;
3274 	struct ip *ip;
3275 	struct llc llc1;
3276 	u_int16_t ether_type;
3277 	pfil_return_t rv;
3278 
3279 	snap = 0;
3280 	error = -1;	/* Default error if not error == 0 */
3281 
3282 #if 0
3283 	/* we may return with the IP fields swapped, ensure its not shared */
3284 	KASSERT(M_WRITABLE(*mp), ("%s: modifying a shared mbuf", __func__));
3285 #endif
3286 
3287 	if (V_pfil_bridge == 0 && V_pfil_member == 0 && V_pfil_ipfw == 0)
3288 		return (0); /* filtering is disabled */
3289 
3290 	i = min((*mp)->m_pkthdr.len, max_protohdr);
3291 	if ((*mp)->m_len < i) {
3292 	    *mp = m_pullup(*mp, i);
3293 	    if (*mp == NULL) {
3294 		printf("%s: m_pullup failed\n", __func__);
3295 		return (-1);
3296 	    }
3297 	}
3298 
3299 	eh1 = mtod(*mp, struct ether_header *);
3300 	ether_type = ntohs(eh1->ether_type);
3301 
3302 	/*
3303 	 * Check for SNAP/LLC.
3304 	 */
3305 	if (ether_type < ETHERMTU) {
3306 		struct llc *llc2 = (struct llc *)(eh1 + 1);
3307 
3308 		if ((*mp)->m_len >= ETHER_HDR_LEN + 8 &&
3309 		    llc2->llc_dsap == LLC_SNAP_LSAP &&
3310 		    llc2->llc_ssap == LLC_SNAP_LSAP &&
3311 		    llc2->llc_control == LLC_UI) {
3312 			ether_type = htons(llc2->llc_un.type_snap.ether_type);
3313 			snap = 1;
3314 		}
3315 	}
3316 
3317 	/*
3318 	 * If we're trying to filter bridge traffic, don't look at anything
3319 	 * other than IP and ARP traffic.  If the filter doesn't understand
3320 	 * IPv6, don't allow IPv6 through the bridge either.  This is lame
3321 	 * since if we really wanted, say, an AppleTalk filter, we are hosed,
3322 	 * but of course we don't have an AppleTalk filter to begin with.
3323 	 * (Note that since pfil doesn't understand ARP it will pass *ALL*
3324 	 * ARP traffic.)
3325 	 */
3326 	switch (ether_type) {
3327 		case ETHERTYPE_ARP:
3328 		case ETHERTYPE_REVARP:
3329 			if (V_pfil_ipfw_arp == 0)
3330 				return (0); /* Automatically pass */
3331 			break;
3332 
3333 		case ETHERTYPE_IP:
3334 #ifdef INET6
3335 		case ETHERTYPE_IPV6:
3336 #endif /* INET6 */
3337 			break;
3338 		default:
3339 			/*
3340 			 * Check to see if the user wants to pass non-ip
3341 			 * packets, these will not be checked by pfil(9) and
3342 			 * passed unconditionally so the default is to drop.
3343 			 */
3344 			if (V_pfil_onlyip)
3345 				goto bad;
3346 	}
3347 
3348 	/* Run the packet through pfil before stripping link headers */
3349 	if (PFIL_HOOKED_OUT(V_link_pfil_head) && V_pfil_ipfw != 0 &&
3350 	    dir == PFIL_OUT && ifp != NULL) {
3351 		switch (pfil_mbuf_out(V_link_pfil_head, mp, ifp, NULL)) {
3352 		case PFIL_DROPPED:
3353 			return (EACCES);
3354 		case PFIL_CONSUMED:
3355 			return (0);
3356 		}
3357 	}
3358 
3359 	/* Strip off the Ethernet header and keep a copy. */
3360 	m_copydata(*mp, 0, ETHER_HDR_LEN, (caddr_t) &eh2);
3361 	m_adj(*mp, ETHER_HDR_LEN);
3362 
3363 	/* Strip off snap header, if present */
3364 	if (snap) {
3365 		m_copydata(*mp, 0, sizeof(struct llc), (caddr_t) &llc1);
3366 		m_adj(*mp, sizeof(struct llc));
3367 	}
3368 
3369 	/*
3370 	 * Check the IP header for alignment and errors
3371 	 */
3372 	if (dir == PFIL_IN) {
3373 		switch (ether_type) {
3374 			case ETHERTYPE_IP:
3375 				error = bridge_ip_checkbasic(mp);
3376 				break;
3377 #ifdef INET6
3378 			case ETHERTYPE_IPV6:
3379 				error = bridge_ip6_checkbasic(mp);
3380 				break;
3381 #endif /* INET6 */
3382 			default:
3383 				error = 0;
3384 		}
3385 		if (error)
3386 			goto bad;
3387 	}
3388 
3389 	error = 0;
3390 
3391 	/*
3392 	 * Run the packet through pfil
3393 	 */
3394 	rv = PFIL_PASS;
3395 	switch (ether_type) {
3396 	case ETHERTYPE_IP:
3397 		/*
3398 		 * Run pfil on the member interface and the bridge, both can
3399 		 * be skipped by clearing pfil_member or pfil_bridge.
3400 		 *
3401 		 * Keep the order:
3402 		 *   in_if -> bridge_if -> out_if
3403 		 */
3404 		if (V_pfil_bridge && dir == PFIL_OUT && bifp != NULL && (rv =
3405 		    pfil_mbuf_out(V_inet_pfil_head, mp, bifp, NULL)) !=
3406 		    PFIL_PASS)
3407 			break;
3408 
3409 		if (V_pfil_member && ifp != NULL) {
3410 			rv = (dir == PFIL_OUT) ?
3411 			    pfil_mbuf_out(V_inet_pfil_head, mp, ifp, NULL) :
3412 			    pfil_mbuf_in(V_inet_pfil_head, mp, ifp, NULL);
3413 			if (rv != PFIL_PASS)
3414 				break;
3415 		}
3416 
3417 		if (V_pfil_bridge && dir == PFIL_IN && bifp != NULL && (rv =
3418 		    pfil_mbuf_in(V_inet_pfil_head, mp, bifp, NULL)) !=
3419 		    PFIL_PASS)
3420 			break;
3421 
3422 		/* check if we need to fragment the packet */
3423 		/* bridge_fragment generates a mbuf chain of packets */
3424 		/* that already include eth headers */
3425 		if (V_pfil_member && ifp != NULL && dir == PFIL_OUT) {
3426 			i = (*mp)->m_pkthdr.len;
3427 			if (i > ifp->if_mtu) {
3428 				error = bridge_fragment(ifp, mp, &eh2, snap,
3429 					    &llc1);
3430 				return (error);
3431 			}
3432 		}
3433 
3434 		/* Recalculate the ip checksum. */
3435 		ip = mtod(*mp, struct ip *);
3436 		hlen = ip->ip_hl << 2;
3437 		if (hlen < sizeof(struct ip))
3438 			goto bad;
3439 		if (hlen > (*mp)->m_len) {
3440 			if ((*mp = m_pullup(*mp, hlen)) == NULL)
3441 				goto bad;
3442 			ip = mtod(*mp, struct ip *);
3443 			if (ip == NULL)
3444 				goto bad;
3445 		}
3446 		ip->ip_sum = 0;
3447 		if (hlen == sizeof(struct ip))
3448 			ip->ip_sum = in_cksum_hdr(ip);
3449 		else
3450 			ip->ip_sum = in_cksum(*mp, hlen);
3451 
3452 		break;
3453 #ifdef INET6
3454 	case ETHERTYPE_IPV6:
3455 		if (V_pfil_bridge && dir == PFIL_OUT && bifp != NULL && (rv =
3456 		    pfil_mbuf_out(V_inet6_pfil_head, mp, bifp, NULL)) !=
3457 		    PFIL_PASS)
3458 			break;
3459 
3460 		if (V_pfil_member && ifp != NULL) {
3461 			rv = (dir == PFIL_OUT) ?
3462 			    pfil_mbuf_out(V_inet6_pfil_head, mp, ifp, NULL) :
3463 			    pfil_mbuf_in(V_inet6_pfil_head, mp, ifp, NULL);
3464 			if (rv != PFIL_PASS)
3465 				break;
3466 		}
3467 
3468 		if (V_pfil_bridge && dir == PFIL_IN && bifp != NULL && (rv =
3469 		    pfil_mbuf_in(V_inet6_pfil_head, mp, bifp, NULL)) !=
3470 		    PFIL_PASS)
3471 			break;
3472 		break;
3473 #endif
3474 	}
3475 
3476 	switch (rv) {
3477 	case PFIL_CONSUMED:
3478 		return (0);
3479 	case PFIL_DROPPED:
3480 		return (EACCES);
3481 	default:
3482 		break;
3483 	}
3484 
3485 	error = -1;
3486 
3487 	/*
3488 	 * Finally, put everything back the way it was and return
3489 	 */
3490 	if (snap) {
3491 		M_PREPEND(*mp, sizeof(struct llc), M_NOWAIT);
3492 		if (*mp == NULL)
3493 			return (error);
3494 		bcopy(&llc1, mtod(*mp, caddr_t), sizeof(struct llc));
3495 	}
3496 
3497 	M_PREPEND(*mp, ETHER_HDR_LEN, M_NOWAIT);
3498 	if (*mp == NULL)
3499 		return (error);
3500 	bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN);
3501 
3502 	return (0);
3503 
3504 bad:
3505 	m_freem(*mp);
3506 	*mp = NULL;
3507 	return (error);
3508 }
3509 
3510 /*
3511  * Perform basic checks on header size since
3512  * pfil assumes ip_input has already processed
3513  * it for it.  Cut-and-pasted from ip_input.c.
3514  * Given how simple the IPv6 version is,
3515  * does the IPv4 version really need to be
3516  * this complicated?
3517  *
3518  * XXX Should we update ipstat here, or not?
3519  * XXX Right now we update ipstat but not
3520  * XXX csum_counter.
3521  */
3522 static int
3523 bridge_ip_checkbasic(struct mbuf **mp)
3524 {
3525 	struct mbuf *m = *mp;
3526 	struct ip *ip;
3527 	int len, hlen;
3528 	u_short sum;
3529 
3530 	if (*mp == NULL)
3531 		return (-1);
3532 
3533 	if (IP_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) {
3534 		if ((m = m_copyup(m, sizeof(struct ip),
3535 			(max_linkhdr + 3) & ~3)) == NULL) {
3536 			/* XXXJRT new stat, please */
3537 			KMOD_IPSTAT_INC(ips_toosmall);
3538 			goto bad;
3539 		}
3540 	} else if (__predict_false(m->m_len < sizeof (struct ip))) {
3541 		if ((m = m_pullup(m, sizeof (struct ip))) == NULL) {
3542 			KMOD_IPSTAT_INC(ips_toosmall);
3543 			goto bad;
3544 		}
3545 	}
3546 	ip = mtod(m, struct ip *);
3547 	if (ip == NULL) goto bad;
3548 
3549 	if (ip->ip_v != IPVERSION) {
3550 		KMOD_IPSTAT_INC(ips_badvers);
3551 		goto bad;
3552 	}
3553 	hlen = ip->ip_hl << 2;
3554 	if (hlen < sizeof(struct ip)) { /* minimum header length */
3555 		KMOD_IPSTAT_INC(ips_badhlen);
3556 		goto bad;
3557 	}
3558 	if (hlen > m->m_len) {
3559 		if ((m = m_pullup(m, hlen)) == NULL) {
3560 			KMOD_IPSTAT_INC(ips_badhlen);
3561 			goto bad;
3562 		}
3563 		ip = mtod(m, struct ip *);
3564 		if (ip == NULL) goto bad;
3565 	}
3566 
3567 	if (m->m_pkthdr.csum_flags & CSUM_IP_CHECKED) {
3568 		sum = !(m->m_pkthdr.csum_flags & CSUM_IP_VALID);
3569 	} else {
3570 		if (hlen == sizeof(struct ip)) {
3571 			sum = in_cksum_hdr(ip);
3572 		} else {
3573 			sum = in_cksum(m, hlen);
3574 		}
3575 	}
3576 	if (sum) {
3577 		KMOD_IPSTAT_INC(ips_badsum);
3578 		goto bad;
3579 	}
3580 
3581 	/* Retrieve the packet length. */
3582 	len = ntohs(ip->ip_len);
3583 
3584 	/*
3585 	 * Check for additional length bogosity
3586 	 */
3587 	if (len < hlen) {
3588 		KMOD_IPSTAT_INC(ips_badlen);
3589 		goto bad;
3590 	}
3591 
3592 	/*
3593 	 * Check that the amount of data in the buffers
3594 	 * is as at least much as the IP header would have us expect.
3595 	 * Drop packet if shorter than we expect.
3596 	 */
3597 	if (m->m_pkthdr.len < len) {
3598 		KMOD_IPSTAT_INC(ips_tooshort);
3599 		goto bad;
3600 	}
3601 
3602 	/* Checks out, proceed */
3603 	*mp = m;
3604 	return (0);
3605 
3606 bad:
3607 	*mp = m;
3608 	return (-1);
3609 }
3610 
3611 #ifdef INET6
3612 /*
3613  * Same as above, but for IPv6.
3614  * Cut-and-pasted from ip6_input.c.
3615  * XXX Should we update ip6stat, or not?
3616  */
3617 static int
3618 bridge_ip6_checkbasic(struct mbuf **mp)
3619 {
3620 	struct mbuf *m = *mp;
3621 	struct ip6_hdr *ip6;
3622 
3623 	/*
3624 	 * If the IPv6 header is not aligned, slurp it up into a new
3625 	 * mbuf with space for link headers, in the event we forward
3626 	 * it.  Otherwise, if it is aligned, make sure the entire base
3627 	 * IPv6 header is in the first mbuf of the chain.
3628 	 */
3629 	if (IP6_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) {
3630 		struct ifnet *inifp = m->m_pkthdr.rcvif;
3631 		if ((m = m_copyup(m, sizeof(struct ip6_hdr),
3632 			    (max_linkhdr + 3) & ~3)) == NULL) {
3633 			/* XXXJRT new stat, please */
3634 			IP6STAT_INC(ip6s_toosmall);
3635 			in6_ifstat_inc(inifp, ifs6_in_hdrerr);
3636 			goto bad;
3637 		}
3638 	} else if (__predict_false(m->m_len < sizeof(struct ip6_hdr))) {
3639 		struct ifnet *inifp = m->m_pkthdr.rcvif;
3640 		if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) {
3641 			IP6STAT_INC(ip6s_toosmall);
3642 			in6_ifstat_inc(inifp, ifs6_in_hdrerr);
3643 			goto bad;
3644 		}
3645 	}
3646 
3647 	ip6 = mtod(m, struct ip6_hdr *);
3648 
3649 	if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) {
3650 		IP6STAT_INC(ip6s_badvers);
3651 		in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr);
3652 		goto bad;
3653 	}
3654 
3655 	/* Checks out, proceed */
3656 	*mp = m;
3657 	return (0);
3658 
3659 bad:
3660 	*mp = m;
3661 	return (-1);
3662 }
3663 #endif /* INET6 */
3664 
3665 /*
3666  * bridge_fragment:
3667  *
3668  *	Fragment mbuf chain in multiple packets and prepend ethernet header.
3669  */
3670 static int
3671 bridge_fragment(struct ifnet *ifp, struct mbuf **mp, struct ether_header *eh,
3672     int snap, struct llc *llc)
3673 {
3674 	struct mbuf *m = *mp, *nextpkt = NULL, *mprev = NULL, *mcur = NULL;
3675 	struct ip *ip;
3676 	int error = -1;
3677 
3678 	if (m->m_len < sizeof(struct ip) &&
3679 	    (m = m_pullup(m, sizeof(struct ip))) == NULL)
3680 		goto dropit;
3681 	ip = mtod(m, struct ip *);
3682 
3683 	m->m_pkthdr.csum_flags |= CSUM_IP;
3684 	error = ip_fragment(ip, &m, ifp->if_mtu, ifp->if_hwassist);
3685 	if (error)
3686 		goto dropit;
3687 
3688 	/*
3689 	 * Walk the chain and re-add the Ethernet header for
3690 	 * each mbuf packet.
3691 	 */
3692 	for (mcur = m; mcur; mcur = mcur->m_nextpkt) {
3693 		nextpkt = mcur->m_nextpkt;
3694 		mcur->m_nextpkt = NULL;
3695 		if (snap) {
3696 			M_PREPEND(mcur, sizeof(struct llc), M_NOWAIT);
3697 			if (mcur == NULL) {
3698 				error = ENOBUFS;
3699 				if (mprev != NULL)
3700 					mprev->m_nextpkt = nextpkt;
3701 				goto dropit;
3702 			}
3703 			bcopy(llc, mtod(mcur, caddr_t),sizeof(struct llc));
3704 		}
3705 
3706 		M_PREPEND(mcur, ETHER_HDR_LEN, M_NOWAIT);
3707 		if (mcur == NULL) {
3708 			error = ENOBUFS;
3709 			if (mprev != NULL)
3710 				mprev->m_nextpkt = nextpkt;
3711 			goto dropit;
3712 		}
3713 		bcopy(eh, mtod(mcur, caddr_t), ETHER_HDR_LEN);
3714 
3715 		/*
3716 		 * The previous two M_PREPEND could have inserted one or two
3717 		 * mbufs in front so we have to update the previous packet's
3718 		 * m_nextpkt.
3719 		 */
3720 		mcur->m_nextpkt = nextpkt;
3721 		if (mprev != NULL)
3722 			mprev->m_nextpkt = mcur;
3723 		else {
3724 			/* The first mbuf in the original chain needs to be
3725 			 * updated. */
3726 			*mp = mcur;
3727 		}
3728 		mprev = mcur;
3729 	}
3730 
3731 	KMOD_IPSTAT_INC(ips_fragmented);
3732 	return (error);
3733 
3734 dropit:
3735 	for (mcur = *mp; mcur; mcur = m) { /* droping the full packet chain */
3736 		m = mcur->m_nextpkt;
3737 		m_freem(mcur);
3738 	}
3739 	return (error);
3740 }
3741 
3742 static void
3743 bridge_linkstate(struct ifnet *ifp)
3744 {
3745 	struct bridge_softc *sc = ifp->if_bridge;
3746 	struct bridge_iflist *bif;
3747 	struct epoch_tracker et;
3748 
3749 	NET_EPOCH_ENTER(et);
3750 
3751 	bif = bridge_lookup_member_if(sc, ifp);
3752 	if (bif == NULL) {
3753 		NET_EPOCH_EXIT(et);
3754 		return;
3755 	}
3756 	bridge_linkcheck(sc);
3757 
3758 	bstp_linkstate(&bif->bif_stp);
3759 
3760 	NET_EPOCH_EXIT(et);
3761 }
3762 
3763 static void
3764 bridge_linkcheck(struct bridge_softc *sc)
3765 {
3766 	struct bridge_iflist *bif;
3767 	int new_link, hasls;
3768 
3769 	BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc);
3770 
3771 	new_link = LINK_STATE_DOWN;
3772 	hasls = 0;
3773 	/* Our link is considered up if at least one of our ports is active */
3774 	CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
3775 		if (bif->bif_ifp->if_capabilities & IFCAP_LINKSTATE)
3776 			hasls++;
3777 		if (bif->bif_ifp->if_link_state == LINK_STATE_UP) {
3778 			new_link = LINK_STATE_UP;
3779 			break;
3780 		}
3781 	}
3782 	if (!CK_LIST_EMPTY(&sc->sc_iflist) && !hasls) {
3783 		/* If no interfaces support link-state then we default to up */
3784 		new_link = LINK_STATE_UP;
3785 	}
3786 	if_link_state_change(sc->sc_ifp, new_link);
3787 }
3788