FreeBSD/Linux Kernel Cross Reference
sys/net/if_bridge.c
1 /* $NetBSD: if_bridge.c,v 1.31 2005/06/01 19:45:34 jdc Exp $ */
2
3 /*
4 * Copyright 2001 Wasabi Systems, Inc.
5 * All rights reserved.
6 *
7 * Written by Jason R. Thorpe for Wasabi Systems, Inc.
8 *
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
11 * are met:
12 * 1. Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 * notice, this list of conditions and the following disclaimer in the
16 * documentation and/or other materials provided with the distribution.
17 * 3. All advertising materials mentioning features or use of this software
18 * must display the following acknowledgement:
19 * This product includes software developed for the NetBSD Project by
20 * Wasabi Systems, Inc.
21 * 4. The name of Wasabi Systems, Inc. may not be used to endorse
22 * or promote products derived from this software without specific prior
23 * written permission.
24 *
25 * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND
26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
27 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
28 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WASABI SYSTEMS, INC
29 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
30 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
31 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
32 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
33 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
34 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
35 * POSSIBILITY OF SUCH DAMAGE.
36 */
37
38 /*
39 * Copyright (c) 1999, 2000 Jason L. Wright (jason@thought.net)
40 * All rights reserved.
41 *
42 * Redistribution and use in source and binary forms, with or without
43 * modification, are permitted provided that the following conditions
44 * are met:
45 * 1. Redistributions of source code must retain the above copyright
46 * notice, this list of conditions and the following disclaimer.
47 * 2. Redistributions in binary form must reproduce the above copyright
48 * notice, this list of conditions and the following disclaimer in the
49 * documentation and/or other materials provided with the distribution.
50 *
51 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
52 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
53 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
54 * DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT,
55 * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
56 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
57 * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
58 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
59 * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
60 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
61 * POSSIBILITY OF SUCH DAMAGE.
62 *
63 * OpenBSD: if_bridge.c,v 1.60 2001/06/15 03:38:33 itojun Exp
64 */
65
66 /*
67 * Network interface bridge support.
68 *
69 * TODO:
70 *
71 * - Currently only supports Ethernet-like interfaces (Ethernet,
72 * 802.11, VLANs on Ethernet, etc.) Figure out a nice way
73 * to bridge other types of interfaces (FDDI-FDDI, and maybe
74 * consider heterogenous bridges).
75 */
76
77 #include <sys/cdefs.h>
78 __FBSDID("$FreeBSD$");
79
80 #include "opt_inet.h"
81 #include "opt_inet6.h"
82
83 #include <sys/param.h>
84 #include <sys/mbuf.h>
85 #include <sys/malloc.h>
86 #include <sys/protosw.h>
87 #include <sys/systm.h>
88 #include <sys/jail.h>
89 #include <sys/time.h>
90 #include <sys/socket.h> /* for net/if.h */
91 #include <sys/sockio.h>
92 #include <sys/ctype.h> /* string functions */
93 #include <sys/kernel.h>
94 #include <sys/random.h>
95 #include <sys/syslog.h>
96 #include <sys/sysctl.h>
97 #include <vm/uma.h>
98 #include <sys/module.h>
99 #include <sys/priv.h>
100 #include <sys/proc.h>
101 #include <sys/lock.h>
102 #include <sys/mutex.h>
103 #include <sys/rwlock.h>
104
105 #include <net/bpf.h>
106 #include <net/if.h>
107 #include <net/if_clone.h>
108 #include <net/if_dl.h>
109 #include <net/if_types.h>
110 #include <net/if_var.h>
111 #include <net/pfil.h>
112 #include <net/vnet.h>
113
114 #include <netinet/in.h> /* for struct arpcom */
115 #include <netinet/in_systm.h>
116 #include <netinet/in_var.h>
117 #include <netinet/ip.h>
118 #include <netinet/ip_var.h>
119 #ifdef INET6
120 #include <netinet/ip6.h>
121 #include <netinet6/ip6_var.h>
122 #include <netinet6/in6_ifattach.h>
123 #endif
124 #if defined(INET) || defined(INET6)
125 #include <netinet/ip_carp.h>
126 #endif
127 #include <machine/in_cksum.h>
128 #include <netinet/if_ether.h> /* for struct arpcom */
129 #include <net/bridgestp.h>
130 #include <net/if_bridgevar.h>
131 #include <net/if_llc.h>
132 #include <net/if_vlan_var.h>
133
134 #include <net/route.h>
135 #include <netinet/ip_fw.h>
136 #include <netpfil/ipfw/ip_fw_private.h>
137
138 /*
139 * Size of the route hash table. Must be a power of two.
140 */
141 #ifndef BRIDGE_RTHASH_SIZE
142 #define BRIDGE_RTHASH_SIZE 1024
143 #endif
144
145 #define BRIDGE_RTHASH_MASK (BRIDGE_RTHASH_SIZE - 1)
146
147 /*
148 * Default maximum number of addresses to cache.
149 */
150 #ifndef BRIDGE_RTABLE_MAX
151 #define BRIDGE_RTABLE_MAX 2000
152 #endif
153
154 /*
155 * Timeout (in seconds) for entries learned dynamically.
156 */
157 #ifndef BRIDGE_RTABLE_TIMEOUT
158 #define BRIDGE_RTABLE_TIMEOUT (20 * 60) /* same as ARP */
159 #endif
160
161 /*
162 * Number of seconds between walks of the route list.
163 */
164 #ifndef BRIDGE_RTABLE_PRUNE_PERIOD
165 #define BRIDGE_RTABLE_PRUNE_PERIOD (5 * 60)
166 #endif
167
168 /*
169 * List of capabilities to possibly mask on the member interface.
170 */
171 #define BRIDGE_IFCAPS_MASK (IFCAP_TOE|IFCAP_TSO|IFCAP_TXCSUM)
172
173 /*
174 * List of capabilities to strip
175 */
176 #define BRIDGE_IFCAPS_STRIP IFCAP_LRO
177
178 /*
179 * Bridge interface list entry.
180 */
181 struct bridge_iflist {
182 LIST_ENTRY(bridge_iflist) bif_next;
183 struct ifnet *bif_ifp; /* member if */
184 struct bstp_port bif_stp; /* STP state */
185 uint32_t bif_flags; /* member if flags */
186 int bif_savedcaps; /* saved capabilities */
187 uint32_t bif_addrmax; /* max # of addresses */
188 uint32_t bif_addrcnt; /* cur. # of addresses */
189 uint32_t bif_addrexceeded;/* # of address violations */
190 };
191
192 /*
193 * Bridge route node.
194 */
195 struct bridge_rtnode {
196 LIST_ENTRY(bridge_rtnode) brt_hash; /* hash table linkage */
197 LIST_ENTRY(bridge_rtnode) brt_list; /* list linkage */
198 struct bridge_iflist *brt_dst; /* destination if */
199 unsigned long brt_expire; /* expiration time */
200 uint8_t brt_flags; /* address flags */
201 uint8_t brt_addr[ETHER_ADDR_LEN];
202 uint16_t brt_vlan; /* vlan id */
203 };
204 #define brt_ifp brt_dst->bif_ifp
205
206 /*
207 * Software state for each bridge.
208 */
209 struct bridge_softc {
210 struct ifnet *sc_ifp; /* make this an interface */
211 LIST_ENTRY(bridge_softc) sc_list;
212 struct mtx sc_mtx;
213 struct cv sc_cv;
214 uint32_t sc_brtmax; /* max # of addresses */
215 uint32_t sc_brtcnt; /* cur. # of addresses */
216 uint32_t sc_brttimeout; /* rt timeout in seconds */
217 struct callout sc_brcallout; /* bridge callout */
218 uint32_t sc_iflist_ref; /* refcount for sc_iflist */
219 uint32_t sc_iflist_xcnt; /* refcount for sc_iflist */
220 LIST_HEAD(, bridge_iflist) sc_iflist; /* member interface list */
221 LIST_HEAD(, bridge_rtnode) *sc_rthash; /* our forwarding table */
222 LIST_HEAD(, bridge_rtnode) sc_rtlist; /* list version of above */
223 uint32_t sc_rthash_key; /* key for hash */
224 LIST_HEAD(, bridge_iflist) sc_spanlist; /* span ports list */
225 struct bstp_state sc_stp; /* STP state */
226 uint32_t sc_brtexceeded; /* # of cache drops */
227 struct ifnet *sc_ifaddr; /* member mac copied from */
228 u_char sc_defaddr[6]; /* Default MAC address */
229 };
230
231 static struct mtx bridge_list_mtx;
232 eventhandler_tag bridge_detach_cookie = NULL;
233
234 int bridge_rtable_prune_period = BRIDGE_RTABLE_PRUNE_PERIOD;
235
236 uma_zone_t bridge_rtnode_zone;
237
238 static int bridge_clone_create(struct if_clone *, int, caddr_t);
239 static void bridge_clone_destroy(struct ifnet *);
240
241 static int bridge_ioctl(struct ifnet *, u_long, caddr_t);
242 static void bridge_mutecaps(struct bridge_softc *);
243 static void bridge_set_ifcap(struct bridge_softc *, struct bridge_iflist *,
244 int);
245 static void bridge_ifdetach(void *arg __unused, struct ifnet *);
246 static void bridge_init(void *);
247 static void bridge_dummynet(struct mbuf *, struct ifnet *);
248 static void bridge_stop(struct ifnet *, int);
249 static void bridge_start(struct ifnet *);
250 static struct mbuf *bridge_input(struct ifnet *, struct mbuf *);
251 static int bridge_output(struct ifnet *, struct mbuf *, struct sockaddr *,
252 struct rtentry *);
253 static void bridge_enqueue(struct bridge_softc *, struct ifnet *,
254 struct mbuf *);
255 static void bridge_rtdelete(struct bridge_softc *, struct ifnet *ifp, int);
256
257 static void bridge_forward(struct bridge_softc *, struct bridge_iflist *,
258 struct mbuf *m);
259
260 static void bridge_timer(void *);
261
262 static void bridge_broadcast(struct bridge_softc *, struct ifnet *,
263 struct mbuf *, int);
264 static void bridge_span(struct bridge_softc *, struct mbuf *);
265
266 static int bridge_rtupdate(struct bridge_softc *, const uint8_t *,
267 uint16_t, struct bridge_iflist *, int, uint8_t);
268 static struct ifnet *bridge_rtlookup(struct bridge_softc *, const uint8_t *,
269 uint16_t);
270 static void bridge_rttrim(struct bridge_softc *);
271 static void bridge_rtage(struct bridge_softc *);
272 static void bridge_rtflush(struct bridge_softc *, int);
273 static int bridge_rtdaddr(struct bridge_softc *, const uint8_t *,
274 uint16_t);
275
276 static int bridge_rtable_init(struct bridge_softc *);
277 static void bridge_rtable_fini(struct bridge_softc *);
278
279 static int bridge_rtnode_addr_cmp(const uint8_t *, const uint8_t *);
280 static struct bridge_rtnode *bridge_rtnode_lookup(struct bridge_softc *,
281 const uint8_t *, uint16_t);
282 static int bridge_rtnode_insert(struct bridge_softc *,
283 struct bridge_rtnode *);
284 static void bridge_rtnode_destroy(struct bridge_softc *,
285 struct bridge_rtnode *);
286 static void bridge_rtable_expire(struct ifnet *, int);
287 static void bridge_state_change(struct ifnet *, int);
288
289 static struct bridge_iflist *bridge_lookup_member(struct bridge_softc *,
290 const char *name);
291 static struct bridge_iflist *bridge_lookup_member_if(struct bridge_softc *,
292 struct ifnet *ifp);
293 static void bridge_delete_member(struct bridge_softc *,
294 struct bridge_iflist *, int);
295 static void bridge_delete_span(struct bridge_softc *,
296 struct bridge_iflist *);
297
298 static int bridge_ioctl_add(struct bridge_softc *, void *);
299 static int bridge_ioctl_del(struct bridge_softc *, void *);
300 static int bridge_ioctl_gifflags(struct bridge_softc *, void *);
301 static int bridge_ioctl_sifflags(struct bridge_softc *, void *);
302 static int bridge_ioctl_scache(struct bridge_softc *, void *);
303 static int bridge_ioctl_gcache(struct bridge_softc *, void *);
304 static int bridge_ioctl_gifs(struct bridge_softc *, void *);
305 static int bridge_ioctl_rts(struct bridge_softc *, void *);
306 static int bridge_ioctl_saddr(struct bridge_softc *, void *);
307 static int bridge_ioctl_sto(struct bridge_softc *, void *);
308 static int bridge_ioctl_gto(struct bridge_softc *, void *);
309 static int bridge_ioctl_daddr(struct bridge_softc *, void *);
310 static int bridge_ioctl_flush(struct bridge_softc *, void *);
311 static int bridge_ioctl_gpri(struct bridge_softc *, void *);
312 static int bridge_ioctl_spri(struct bridge_softc *, void *);
313 static int bridge_ioctl_ght(struct bridge_softc *, void *);
314 static int bridge_ioctl_sht(struct bridge_softc *, void *);
315 static int bridge_ioctl_gfd(struct bridge_softc *, void *);
316 static int bridge_ioctl_sfd(struct bridge_softc *, void *);
317 static int bridge_ioctl_gma(struct bridge_softc *, void *);
318 static int bridge_ioctl_sma(struct bridge_softc *, void *);
319 static int bridge_ioctl_sifprio(struct bridge_softc *, void *);
320 static int bridge_ioctl_sifcost(struct bridge_softc *, void *);
321 static int bridge_ioctl_sifmaxaddr(struct bridge_softc *, void *);
322 static int bridge_ioctl_addspan(struct bridge_softc *, void *);
323 static int bridge_ioctl_delspan(struct bridge_softc *, void *);
324 static int bridge_ioctl_gbparam(struct bridge_softc *, void *);
325 static int bridge_ioctl_grte(struct bridge_softc *, void *);
326 static int bridge_ioctl_gifsstp(struct bridge_softc *, void *);
327 static int bridge_ioctl_sproto(struct bridge_softc *, void *);
328 static int bridge_ioctl_stxhc(struct bridge_softc *, void *);
329 static int bridge_pfil(struct mbuf **, struct ifnet *, struct ifnet *,
330 int);
331 static int bridge_ip_checkbasic(struct mbuf **mp);
332 #ifdef INET6
333 static int bridge_ip6_checkbasic(struct mbuf **mp);
334 #endif /* INET6 */
335 static int bridge_fragment(struct ifnet *, struct mbuf *,
336 struct ether_header *, int, struct llc *);
337 static void bridge_linkstate(struct ifnet *ifp);
338 static void bridge_linkcheck(struct bridge_softc *sc);
339
340 extern void (*bridge_linkstate_p)(struct ifnet *ifp);
341
342 /* The default bridge vlan is 1 (IEEE 802.1Q-2003 Table 9-2) */
343 #define VLANTAGOF(_m) \
344 (_m->m_flags & M_VLANTAG) ? EVL_VLANOFTAG(_m->m_pkthdr.ether_vtag) : 1
345
346 static struct bstp_cb_ops bridge_ops = {
347 .bcb_state = bridge_state_change,
348 .bcb_rtage = bridge_rtable_expire
349 };
350
351 SYSCTL_DECL(_net_link);
352 static SYSCTL_NODE(_net_link, IFT_BRIDGE, bridge, CTLFLAG_RW, 0, "Bridge");
353
354 static int pfil_onlyip = 1; /* only pass IP[46] packets when pfil is enabled */
355 static int pfil_bridge = 1; /* run pfil hooks on the bridge interface */
356 static int pfil_member = 1; /* run pfil hooks on the member interface */
357 static int pfil_ipfw = 0; /* layer2 filter with ipfw */
358 static int pfil_ipfw_arp = 0; /* layer2 filter with ipfw */
359 static int pfil_local_phys = 0; /* run pfil hooks on the physical interface for
360 locally destined packets */
361 static int log_stp = 0; /* log STP state changes */
362 static int bridge_inherit_mac = 0; /* share MAC with first bridge member */
363 TUNABLE_INT("net.link.bridge.pfil_onlyip", &pfil_onlyip);
364 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_onlyip, CTLFLAG_RW,
365 &pfil_onlyip, 0, "Only pass IP packets when pfil is enabled");
366 TUNABLE_INT("net.link.bridge.ipfw_arp", &pfil_ipfw_arp);
367 SYSCTL_INT(_net_link_bridge, OID_AUTO, ipfw_arp, CTLFLAG_RW,
368 &pfil_ipfw_arp, 0, "Filter ARP packets through IPFW layer2");
369 TUNABLE_INT("net.link.bridge.pfil_bridge", &pfil_bridge);
370 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_bridge, CTLFLAG_RW,
371 &pfil_bridge, 0, "Packet filter on the bridge interface");
372 TUNABLE_INT("net.link.bridge.pfil_member", &pfil_member);
373 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_member, CTLFLAG_RW,
374 &pfil_member, 0, "Packet filter on the member interface");
375 TUNABLE_INT("net.link.bridge.pfil_local_phys", &pfil_local_phys);
376 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_local_phys, CTLFLAG_RW,
377 &pfil_local_phys, 0,
378 "Packet filter on the physical interface for locally destined packets");
379 TUNABLE_INT("net.link.bridge.log_stp", &log_stp);
380 SYSCTL_INT(_net_link_bridge, OID_AUTO, log_stp, CTLFLAG_RW,
381 &log_stp, 0, "Log STP state changes");
382 TUNABLE_INT("net.link.bridge.inherit_mac", &bridge_inherit_mac);
383 SYSCTL_INT(_net_link_bridge, OID_AUTO, inherit_mac, CTLFLAG_RW,
384 &bridge_inherit_mac, 0,
385 "Inherit MAC address from the first bridge member");
386
387 static VNET_DEFINE(int, allow_llz_overlap) = 0;
388 #define V_allow_llz_overlap VNET(allow_llz_overlap)
389 SYSCTL_VNET_INT(_net_link_bridge, OID_AUTO, allow_llz_overlap, CTLFLAG_RW,
390 &VNET_NAME(allow_llz_overlap), 0, "Allow overlap of link-local scope "
391 "zones of a bridge interface and the member interfaces");
392
393 struct bridge_control {
394 int (*bc_func)(struct bridge_softc *, void *);
395 int bc_argsize;
396 int bc_flags;
397 };
398
399 #define BC_F_COPYIN 0x01 /* copy arguments in */
400 #define BC_F_COPYOUT 0x02 /* copy arguments out */
401 #define BC_F_SUSER 0x04 /* do super-user check */
402
403 const struct bridge_control bridge_control_table[] = {
404 { bridge_ioctl_add, sizeof(struct ifbreq),
405 BC_F_COPYIN|BC_F_SUSER },
406 { bridge_ioctl_del, sizeof(struct ifbreq),
407 BC_F_COPYIN|BC_F_SUSER },
408
409 { bridge_ioctl_gifflags, sizeof(struct ifbreq),
410 BC_F_COPYIN|BC_F_COPYOUT },
411 { bridge_ioctl_sifflags, sizeof(struct ifbreq),
412 BC_F_COPYIN|BC_F_SUSER },
413
414 { bridge_ioctl_scache, sizeof(struct ifbrparam),
415 BC_F_COPYIN|BC_F_SUSER },
416 { bridge_ioctl_gcache, sizeof(struct ifbrparam),
417 BC_F_COPYOUT },
418
419 { bridge_ioctl_gifs, sizeof(struct ifbifconf),
420 BC_F_COPYIN|BC_F_COPYOUT },
421 { bridge_ioctl_rts, sizeof(struct ifbaconf),
422 BC_F_COPYIN|BC_F_COPYOUT },
423
424 { bridge_ioctl_saddr, sizeof(struct ifbareq),
425 BC_F_COPYIN|BC_F_SUSER },
426
427 { bridge_ioctl_sto, sizeof(struct ifbrparam),
428 BC_F_COPYIN|BC_F_SUSER },
429 { bridge_ioctl_gto, sizeof(struct ifbrparam),
430 BC_F_COPYOUT },
431
432 { bridge_ioctl_daddr, sizeof(struct ifbareq),
433 BC_F_COPYIN|BC_F_SUSER },
434
435 { bridge_ioctl_flush, sizeof(struct ifbreq),
436 BC_F_COPYIN|BC_F_SUSER },
437
438 { bridge_ioctl_gpri, sizeof(struct ifbrparam),
439 BC_F_COPYOUT },
440 { bridge_ioctl_spri, sizeof(struct ifbrparam),
441 BC_F_COPYIN|BC_F_SUSER },
442
443 { bridge_ioctl_ght, sizeof(struct ifbrparam),
444 BC_F_COPYOUT },
445 { bridge_ioctl_sht, sizeof(struct ifbrparam),
446 BC_F_COPYIN|BC_F_SUSER },
447
448 { bridge_ioctl_gfd, sizeof(struct ifbrparam),
449 BC_F_COPYOUT },
450 { bridge_ioctl_sfd, sizeof(struct ifbrparam),
451 BC_F_COPYIN|BC_F_SUSER },
452
453 { bridge_ioctl_gma, sizeof(struct ifbrparam),
454 BC_F_COPYOUT },
455 { bridge_ioctl_sma, sizeof(struct ifbrparam),
456 BC_F_COPYIN|BC_F_SUSER },
457
458 { bridge_ioctl_sifprio, sizeof(struct ifbreq),
459 BC_F_COPYIN|BC_F_SUSER },
460
461 { bridge_ioctl_sifcost, sizeof(struct ifbreq),
462 BC_F_COPYIN|BC_F_SUSER },
463
464 { bridge_ioctl_addspan, sizeof(struct ifbreq),
465 BC_F_COPYIN|BC_F_SUSER },
466 { bridge_ioctl_delspan, sizeof(struct ifbreq),
467 BC_F_COPYIN|BC_F_SUSER },
468
469 { bridge_ioctl_gbparam, sizeof(struct ifbropreq),
470 BC_F_COPYOUT },
471
472 { bridge_ioctl_grte, sizeof(struct ifbrparam),
473 BC_F_COPYOUT },
474
475 { bridge_ioctl_gifsstp, sizeof(struct ifbpstpconf),
476 BC_F_COPYIN|BC_F_COPYOUT },
477
478 { bridge_ioctl_sproto, sizeof(struct ifbrparam),
479 BC_F_COPYIN|BC_F_SUSER },
480
481 { bridge_ioctl_stxhc, sizeof(struct ifbrparam),
482 BC_F_COPYIN|BC_F_SUSER },
483
484 { bridge_ioctl_sifmaxaddr, sizeof(struct ifbreq),
485 BC_F_COPYIN|BC_F_SUSER },
486
487 };
488 const int bridge_control_table_size =
489 sizeof(bridge_control_table) / sizeof(bridge_control_table[0]);
490
491 LIST_HEAD(, bridge_softc) bridge_list;
492
493 IFC_SIMPLE_DECLARE(bridge, 0);
494
495 static int
496 bridge_modevent(module_t mod, int type, void *data)
497 {
498
499 switch (type) {
500 case MOD_LOAD:
501 mtx_init(&bridge_list_mtx, "if_bridge list", NULL, MTX_DEF);
502 if_clone_attach(&bridge_cloner);
503 bridge_rtnode_zone = uma_zcreate("bridge_rtnode",
504 sizeof(struct bridge_rtnode), NULL, NULL, NULL, NULL,
505 UMA_ALIGN_PTR, 0);
506 LIST_INIT(&bridge_list);
507 bridge_input_p = bridge_input;
508 bridge_output_p = bridge_output;
509 bridge_dn_p = bridge_dummynet;
510 bridge_linkstate_p = bridge_linkstate;
511 bridge_detach_cookie = EVENTHANDLER_REGISTER(
512 ifnet_departure_event, bridge_ifdetach, NULL,
513 EVENTHANDLER_PRI_ANY);
514 break;
515 case MOD_UNLOAD:
516 EVENTHANDLER_DEREGISTER(ifnet_departure_event,
517 bridge_detach_cookie);
518 if_clone_detach(&bridge_cloner);
519 uma_zdestroy(bridge_rtnode_zone);
520 bridge_input_p = NULL;
521 bridge_output_p = NULL;
522 bridge_dn_p = NULL;
523 bridge_linkstate_p = NULL;
524 mtx_destroy(&bridge_list_mtx);
525 break;
526 default:
527 return (EOPNOTSUPP);
528 }
529 return (0);
530 }
531
532 static moduledata_t bridge_mod = {
533 "if_bridge",
534 bridge_modevent,
535 0
536 };
537
538 DECLARE_MODULE(if_bridge, bridge_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
539 MODULE_DEPEND(if_bridge, bridgestp, 1, 1, 1);
540
541 /*
542 * handler for net.link.bridge.pfil_ipfw
543 */
544 static int
545 sysctl_pfil_ipfw(SYSCTL_HANDLER_ARGS)
546 {
547 int enable = pfil_ipfw;
548 int error;
549
550 error = sysctl_handle_int(oidp, &enable, 0, req);
551 enable = (enable) ? 1 : 0;
552
553 if (enable != pfil_ipfw) {
554 pfil_ipfw = enable;
555
556 /*
557 * Disable pfil so that ipfw doesnt run twice, if the user
558 * really wants both then they can re-enable pfil_bridge and/or
559 * pfil_member. Also allow non-ip packets as ipfw can filter by
560 * layer2 type.
561 */
562 if (pfil_ipfw) {
563 pfil_onlyip = 0;
564 pfil_bridge = 0;
565 pfil_member = 0;
566 }
567 }
568
569 return (error);
570 }
571 SYSCTL_PROC(_net_link_bridge, OID_AUTO, ipfw, CTLTYPE_INT|CTLFLAG_RW,
572 &pfil_ipfw, 0, &sysctl_pfil_ipfw, "I", "Layer2 filter with IPFW");
573
574 /*
575 * bridge_clone_create:
576 *
577 * Create a new bridge instance.
578 */
579 static int
580 bridge_clone_create(struct if_clone *ifc, int unit, caddr_t params)
581 {
582 struct bridge_softc *sc, *sc2;
583 struct ifnet *bifp, *ifp;
584 int fb, retry;
585 unsigned long hostid;
586
587 sc = malloc(sizeof(*sc), M_DEVBUF, M_WAITOK|M_ZERO);
588 ifp = sc->sc_ifp = if_alloc(IFT_ETHER);
589 if (ifp == NULL) {
590 free(sc, M_DEVBUF);
591 return (ENOSPC);
592 }
593
594 BRIDGE_LOCK_INIT(sc);
595 sc->sc_brtmax = BRIDGE_RTABLE_MAX;
596 sc->sc_brttimeout = BRIDGE_RTABLE_TIMEOUT;
597
598 /* Initialize our routing table. */
599 bridge_rtable_init(sc);
600
601 callout_init_mtx(&sc->sc_brcallout, &sc->sc_mtx, 0);
602
603 LIST_INIT(&sc->sc_iflist);
604 LIST_INIT(&sc->sc_spanlist);
605
606 ifp->if_softc = sc;
607 if_initname(ifp, ifc->ifc_name, unit);
608 ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
609 ifp->if_ioctl = bridge_ioctl;
610 ifp->if_start = bridge_start;
611 ifp->if_init = bridge_init;
612 ifp->if_type = IFT_BRIDGE;
613 IFQ_SET_MAXLEN(&ifp->if_snd, ifqmaxlen);
614 ifp->if_snd.ifq_drv_maxlen = ifqmaxlen;
615 IFQ_SET_READY(&ifp->if_snd);
616
617 /*
618 * Generate an ethernet address with a locally administered address.
619 *
620 * Since we are using random ethernet addresses for the bridge, it is
621 * possible that we might have address collisions, so make sure that
622 * this hardware address isn't already in use on another bridge.
623 * The first try uses the hostid and falls back to arc4rand().
624 */
625 fb = 0;
626 getcredhostid(curthread->td_ucred, &hostid);
627 for (retry = 1; retry != 0;) {
628 if (fb || hostid == 0) {
629 arc4rand(sc->sc_defaddr, ETHER_ADDR_LEN, 1);
630 sc->sc_defaddr[0] &= ~1;/* clear multicast bit */
631 sc->sc_defaddr[0] |= 2; /* set the LAA bit */
632 } else {
633 sc->sc_defaddr[0] = 0x2;
634 sc->sc_defaddr[1] = (hostid >> 24) & 0xff;
635 sc->sc_defaddr[2] = (hostid >> 16) & 0xff;
636 sc->sc_defaddr[3] = (hostid >> 8 ) & 0xff;
637 sc->sc_defaddr[4] = hostid & 0xff;
638 sc->sc_defaddr[5] = ifp->if_dunit & 0xff;
639 }
640
641 fb = 1;
642 retry = 0;
643 mtx_lock(&bridge_list_mtx);
644 LIST_FOREACH(sc2, &bridge_list, sc_list) {
645 bifp = sc2->sc_ifp;
646 if (memcmp(sc->sc_defaddr,
647 IF_LLADDR(bifp), ETHER_ADDR_LEN) == 0)
648 retry = 1;
649 }
650 mtx_unlock(&bridge_list_mtx);
651 }
652
653 bstp_attach(&sc->sc_stp, &bridge_ops);
654 ether_ifattach(ifp, sc->sc_defaddr);
655 /* Now undo some of the damage... */
656 ifp->if_baudrate = 0;
657 ifp->if_type = IFT_BRIDGE;
658
659 mtx_lock(&bridge_list_mtx);
660 LIST_INSERT_HEAD(&bridge_list, sc, sc_list);
661 mtx_unlock(&bridge_list_mtx);
662
663 return (0);
664 }
665
666 /*
667 * bridge_clone_destroy:
668 *
669 * Destroy a bridge instance.
670 */
671 static void
672 bridge_clone_destroy(struct ifnet *ifp)
673 {
674 struct bridge_softc *sc = ifp->if_softc;
675 struct bridge_iflist *bif;
676
677 BRIDGE_LOCK(sc);
678
679 bridge_stop(ifp, 1);
680 ifp->if_flags &= ~IFF_UP;
681
682 while ((bif = LIST_FIRST(&sc->sc_iflist)) != NULL)
683 bridge_delete_member(sc, bif, 0);
684
685 while ((bif = LIST_FIRST(&sc->sc_spanlist)) != NULL) {
686 bridge_delete_span(sc, bif);
687 }
688
689 BRIDGE_UNLOCK(sc);
690
691 callout_drain(&sc->sc_brcallout);
692
693 mtx_lock(&bridge_list_mtx);
694 LIST_REMOVE(sc, sc_list);
695 mtx_unlock(&bridge_list_mtx);
696
697 bstp_detach(&sc->sc_stp);
698 ether_ifdetach(ifp);
699 if_free_type(ifp, IFT_ETHER);
700
701 /* Tear down the routing table. */
702 bridge_rtable_fini(sc);
703
704 BRIDGE_LOCK_DESTROY(sc);
705 free(sc, M_DEVBUF);
706 }
707
708 /*
709 * bridge_ioctl:
710 *
711 * Handle a control request from the operator.
712 */
713 static int
714 bridge_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data)
715 {
716 struct bridge_softc *sc = ifp->if_softc;
717 struct ifreq *ifr = (struct ifreq *)data;
718 struct bridge_iflist *bif;
719 struct thread *td = curthread;
720 union {
721 struct ifbreq ifbreq;
722 struct ifbifconf ifbifconf;
723 struct ifbareq ifbareq;
724 struct ifbaconf ifbaconf;
725 struct ifbrparam ifbrparam;
726 struct ifbropreq ifbropreq;
727 } args;
728 struct ifdrv *ifd = (struct ifdrv *) data;
729 const struct bridge_control *bc;
730 int error = 0;
731
732 switch (cmd) {
733
734 case SIOCADDMULTI:
735 case SIOCDELMULTI:
736 break;
737
738 case SIOCGDRVSPEC:
739 case SIOCSDRVSPEC:
740 if (ifd->ifd_cmd >= bridge_control_table_size) {
741 error = EINVAL;
742 break;
743 }
744 bc = &bridge_control_table[ifd->ifd_cmd];
745
746 if (cmd == SIOCGDRVSPEC &&
747 (bc->bc_flags & BC_F_COPYOUT) == 0) {
748 error = EINVAL;
749 break;
750 }
751 else if (cmd == SIOCSDRVSPEC &&
752 (bc->bc_flags & BC_F_COPYOUT) != 0) {
753 error = EINVAL;
754 break;
755 }
756
757 if (bc->bc_flags & BC_F_SUSER) {
758 error = priv_check(td, PRIV_NET_BRIDGE);
759 if (error)
760 break;
761 }
762
763 if (ifd->ifd_len != bc->bc_argsize ||
764 ifd->ifd_len > sizeof(args)) {
765 error = EINVAL;
766 break;
767 }
768
769 bzero(&args, sizeof(args));
770 if (bc->bc_flags & BC_F_COPYIN) {
771 error = copyin(ifd->ifd_data, &args, ifd->ifd_len);
772 if (error)
773 break;
774 }
775
776 BRIDGE_LOCK(sc);
777 error = (*bc->bc_func)(sc, &args);
778 BRIDGE_UNLOCK(sc);
779 if (error)
780 break;
781
782 if (bc->bc_flags & BC_F_COPYOUT)
783 error = copyout(&args, ifd->ifd_data, ifd->ifd_len);
784
785 break;
786
787 case SIOCSIFFLAGS:
788 if (!(ifp->if_flags & IFF_UP) &&
789 (ifp->if_drv_flags & IFF_DRV_RUNNING)) {
790 /*
791 * If interface is marked down and it is running,
792 * then stop and disable it.
793 */
794 BRIDGE_LOCK(sc);
795 bridge_stop(ifp, 1);
796 BRIDGE_UNLOCK(sc);
797 } else if ((ifp->if_flags & IFF_UP) &&
798 !(ifp->if_drv_flags & IFF_DRV_RUNNING)) {
799 /*
800 * If interface is marked up and it is stopped, then
801 * start it.
802 */
803 (*ifp->if_init)(sc);
804 }
805 break;
806
807 case SIOCSIFMTU:
808 if (ifr->ifr_mtu < 576) {
809 error = EINVAL;
810 break;
811 }
812 if (LIST_EMPTY(&sc->sc_iflist)) {
813 sc->sc_ifp->if_mtu = ifr->ifr_mtu;
814 break;
815 }
816 BRIDGE_LOCK(sc);
817 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
818 if (bif->bif_ifp->if_mtu != ifr->ifr_mtu) {
819 log(LOG_NOTICE, "%s: invalid MTU: %lu(%s)"
820 " != %d\n", sc->sc_ifp->if_xname,
821 bif->bif_ifp->if_mtu,
822 bif->bif_ifp->if_xname, ifr->ifr_mtu);
823 error = EINVAL;
824 break;
825 }
826 }
827 if (!error)
828 sc->sc_ifp->if_mtu = ifr->ifr_mtu;
829 BRIDGE_UNLOCK(sc);
830 break;
831 default:
832 /*
833 * drop the lock as ether_ioctl() will call bridge_start() and
834 * cause the lock to be recursed.
835 */
836 error = ether_ioctl(ifp, cmd, data);
837 break;
838 }
839
840 return (error);
841 }
842
843 /*
844 * bridge_mutecaps:
845 *
846 * Clear or restore unwanted capabilities on the member interface
847 */
848 static void
849 bridge_mutecaps(struct bridge_softc *sc)
850 {
851 struct bridge_iflist *bif;
852 int enabled, mask;
853
854 /* Initial bitmask of capabilities to test */
855 mask = BRIDGE_IFCAPS_MASK;
856
857 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
858 /* Every member must support it or its disabled */
859 mask &= bif->bif_savedcaps;
860 }
861
862 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
863 enabled = bif->bif_ifp->if_capenable;
864 enabled &= ~BRIDGE_IFCAPS_STRIP;
865 /* strip off mask bits and enable them again if allowed */
866 enabled &= ~BRIDGE_IFCAPS_MASK;
867 enabled |= mask;
868 bridge_set_ifcap(sc, bif, enabled);
869 }
870
871 }
872
873 static void
874 bridge_set_ifcap(struct bridge_softc *sc, struct bridge_iflist *bif, int set)
875 {
876 struct ifnet *ifp = bif->bif_ifp;
877 struct ifreq ifr;
878 int error;
879
880 bzero(&ifr, sizeof(ifr));
881 ifr.ifr_reqcap = set;
882
883 if (ifp->if_capenable != set) {
884 error = (*ifp->if_ioctl)(ifp, SIOCSIFCAP, (caddr_t)&ifr);
885 if (error)
886 if_printf(sc->sc_ifp,
887 "error setting interface capabilities on %s\n",
888 ifp->if_xname);
889 }
890 }
891
892 /*
893 * bridge_lookup_member:
894 *
895 * Lookup a bridge member interface.
896 */
897 static struct bridge_iflist *
898 bridge_lookup_member(struct bridge_softc *sc, const char *name)
899 {
900 struct bridge_iflist *bif;
901 struct ifnet *ifp;
902
903 BRIDGE_LOCK_ASSERT(sc);
904
905 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
906 ifp = bif->bif_ifp;
907 if (strcmp(ifp->if_xname, name) == 0)
908 return (bif);
909 }
910
911 return (NULL);
912 }
913
914 /*
915 * bridge_lookup_member_if:
916 *
917 * Lookup a bridge member interface by ifnet*.
918 */
919 static struct bridge_iflist *
920 bridge_lookup_member_if(struct bridge_softc *sc, struct ifnet *member_ifp)
921 {
922 struct bridge_iflist *bif;
923
924 BRIDGE_LOCK_ASSERT(sc);
925
926 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
927 if (bif->bif_ifp == member_ifp)
928 return (bif);
929 }
930
931 return (NULL);
932 }
933
934 /*
935 * bridge_delete_member:
936 *
937 * Delete the specified member interface.
938 */
939 static void
940 bridge_delete_member(struct bridge_softc *sc, struct bridge_iflist *bif,
941 int gone)
942 {
943 struct ifnet *ifs = bif->bif_ifp;
944 struct ifnet *fif = NULL;
945
946 BRIDGE_LOCK_ASSERT(sc);
947
948 if (bif->bif_flags & IFBIF_STP)
949 bstp_disable(&bif->bif_stp);
950
951 ifs->if_bridge = NULL;
952 BRIDGE_XLOCK(sc);
953 LIST_REMOVE(bif, bif_next);
954 BRIDGE_XDROP(sc);
955
956 /*
957 * If removing the interface that gave the bridge its mac address, set
958 * the mac address of the bridge to the address of the next member, or
959 * to its default address if no members are left.
960 */
961 if (bridge_inherit_mac && sc->sc_ifaddr == ifs) {
962 if (LIST_EMPTY(&sc->sc_iflist)) {
963 bcopy(sc->sc_defaddr,
964 IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN);
965 sc->sc_ifaddr = NULL;
966 } else {
967 fif = LIST_FIRST(&sc->sc_iflist)->bif_ifp;
968 bcopy(IF_LLADDR(fif),
969 IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN);
970 sc->sc_ifaddr = fif;
971 }
972 EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp);
973 }
974
975 bridge_linkcheck(sc);
976 bridge_mutecaps(sc); /* recalcuate now this interface is removed */
977 bridge_rtdelete(sc, ifs, IFBF_FLUSHALL);
978 KASSERT(bif->bif_addrcnt == 0,
979 ("%s: %d bridge routes referenced", __func__, bif->bif_addrcnt));
980
981 BRIDGE_UNLOCK(sc);
982 if (!gone) {
983 switch (ifs->if_type) {
984 case IFT_ETHER:
985 case IFT_L2VLAN:
986 /*
987 * Take the interface out of promiscuous mode.
988 */
989 (void) ifpromisc(ifs, 0);
990 break;
991
992 case IFT_GIF:
993 break;
994
995 default:
996 #ifdef DIAGNOSTIC
997 panic("bridge_delete_member: impossible");
998 #endif
999 break;
1000 }
1001 /* reneable any interface capabilities */
1002 bridge_set_ifcap(sc, bif, bif->bif_savedcaps);
1003 }
1004 bstp_destroy(&bif->bif_stp); /* prepare to free */
1005 BRIDGE_LOCK(sc);
1006 free(bif, M_DEVBUF);
1007 }
1008
1009 /*
1010 * bridge_delete_span:
1011 *
1012 * Delete the specified span interface.
1013 */
1014 static void
1015 bridge_delete_span(struct bridge_softc *sc, struct bridge_iflist *bif)
1016 {
1017 BRIDGE_LOCK_ASSERT(sc);
1018
1019 KASSERT(bif->bif_ifp->if_bridge == NULL,
1020 ("%s: not a span interface", __func__));
1021
1022 LIST_REMOVE(bif, bif_next);
1023 free(bif, M_DEVBUF);
1024 }
1025
1026 static int
1027 bridge_ioctl_add(struct bridge_softc *sc, void *arg)
1028 {
1029 struct ifbreq *req = arg;
1030 struct bridge_iflist *bif = NULL;
1031 struct ifnet *ifs;
1032 int error = 0;
1033
1034 ifs = ifunit(req->ifbr_ifsname);
1035 if (ifs == NULL)
1036 return (ENOENT);
1037 if (ifs->if_ioctl == NULL) /* must be supported */
1038 return (EINVAL);
1039
1040 /* If it's in the span list, it can't be a member. */
1041 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1042 if (ifs == bif->bif_ifp)
1043 return (EBUSY);
1044
1045 if (ifs->if_bridge == sc)
1046 return (EEXIST);
1047
1048 if (ifs->if_bridge != NULL)
1049 return (EBUSY);
1050
1051 switch (ifs->if_type) {
1052 case IFT_ETHER:
1053 case IFT_L2VLAN:
1054 case IFT_GIF:
1055 /* permitted interface types */
1056 break;
1057 default:
1058 return (EINVAL);
1059 }
1060
1061 #ifdef INET6
1062 /*
1063 * Two valid inet6 addresses with link-local scope must not be
1064 * on the parent interface and the member interfaces at the
1065 * same time. This restriction is needed to prevent violation
1066 * of link-local scope zone. Attempts to add a member
1067 * interface which has inet6 addresses when the parent has
1068 * inet6 triggers removal of all inet6 addresses on the member
1069 * interface.
1070 */
1071
1072 /* Check if the parent interface has a link-local scope addr. */
1073 if (V_allow_llz_overlap == 0 &&
1074 in6ifa_llaonifp(sc->sc_ifp) != NULL) {
1075 /*
1076 * If any, remove all inet6 addresses from the member
1077 * interfaces.
1078 */
1079 BRIDGE_XLOCK(sc);
1080 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1081 if (in6ifa_llaonifp(bif->bif_ifp)) {
1082 BRIDGE_UNLOCK(sc);
1083 in6_ifdetach(bif->bif_ifp);
1084 BRIDGE_LOCK(sc);
1085 if_printf(sc->sc_ifp,
1086 "IPv6 addresses on %s have been removed "
1087 "before adding it as a member to prevent "
1088 "IPv6 address scope violation.\n",
1089 bif->bif_ifp->if_xname);
1090 }
1091 }
1092 BRIDGE_XDROP(sc);
1093 if (in6ifa_llaonifp(ifs)) {
1094 BRIDGE_UNLOCK(sc);
1095 in6_ifdetach(ifs);
1096 BRIDGE_LOCK(sc);
1097 if_printf(sc->sc_ifp,
1098 "IPv6 addresses on %s have been removed "
1099 "before adding it as a member to prevent "
1100 "IPv6 address scope violation.\n",
1101 ifs->if_xname);
1102 }
1103 }
1104 #endif
1105 /* Allow the first Ethernet member to define the MTU */
1106 if (LIST_EMPTY(&sc->sc_iflist))
1107 sc->sc_ifp->if_mtu = ifs->if_mtu;
1108 else if (sc->sc_ifp->if_mtu != ifs->if_mtu) {
1109 if_printf(sc->sc_ifp, "invalid MTU: %lu(%s) != %lu\n",
1110 ifs->if_mtu, ifs->if_xname, sc->sc_ifp->if_mtu);
1111 return (EINVAL);
1112 }
1113
1114 bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO);
1115 if (bif == NULL)
1116 return (ENOMEM);
1117
1118 bif->bif_ifp = ifs;
1119 bif->bif_flags = IFBIF_LEARNING | IFBIF_DISCOVER;
1120 bif->bif_savedcaps = ifs->if_capenable;
1121
1122 /*
1123 * Assign the interface's MAC address to the bridge if it's the first
1124 * member and the MAC address of the bridge has not been changed from
1125 * the default randomly generated one.
1126 */
1127 if (bridge_inherit_mac && LIST_EMPTY(&sc->sc_iflist) &&
1128 !memcmp(IF_LLADDR(sc->sc_ifp), sc->sc_defaddr, ETHER_ADDR_LEN)) {
1129 bcopy(IF_LLADDR(ifs), IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN);
1130 sc->sc_ifaddr = ifs;
1131 EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp);
1132 }
1133
1134 ifs->if_bridge = sc;
1135 bstp_create(&sc->sc_stp, &bif->bif_stp, bif->bif_ifp);
1136 /*
1137 * XXX: XLOCK HERE!?!
1138 *
1139 * NOTE: insert_***HEAD*** should be safe for the traversals.
1140 */
1141 LIST_INSERT_HEAD(&sc->sc_iflist, bif, bif_next);
1142
1143 /* Set interface capabilities to the intersection set of all members */
1144 bridge_mutecaps(sc);
1145 bridge_linkcheck(sc);
1146
1147 /* Place the interface into promiscuous mode */
1148 switch (ifs->if_type) {
1149 case IFT_ETHER:
1150 case IFT_L2VLAN:
1151 BRIDGE_UNLOCK(sc);
1152 error = ifpromisc(ifs, 1);
1153 BRIDGE_LOCK(sc);
1154 break;
1155 }
1156
1157 if (error) {
1158 bridge_delete_member(sc, bif, 0);
1159 free(bif, M_DEVBUF);
1160 }
1161 return (error);
1162 }
1163
1164 static int
1165 bridge_ioctl_del(struct bridge_softc *sc, void *arg)
1166 {
1167 struct ifbreq *req = arg;
1168 struct bridge_iflist *bif;
1169
1170 bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1171 if (bif == NULL)
1172 return (ENOENT);
1173
1174 bridge_delete_member(sc, bif, 0);
1175
1176 return (0);
1177 }
1178
1179 static int
1180 bridge_ioctl_gifflags(struct bridge_softc *sc, void *arg)
1181 {
1182 struct ifbreq *req = arg;
1183 struct bridge_iflist *bif;
1184 struct bstp_port *bp;
1185
1186 bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1187 if (bif == NULL)
1188 return (ENOENT);
1189
1190 bp = &bif->bif_stp;
1191 req->ifbr_ifsflags = bif->bif_flags;
1192 req->ifbr_state = bp->bp_state;
1193 req->ifbr_priority = bp->bp_priority;
1194 req->ifbr_path_cost = bp->bp_path_cost;
1195 req->ifbr_portno = bif->bif_ifp->if_index & 0xfff;
1196 req->ifbr_proto = bp->bp_protover;
1197 req->ifbr_role = bp->bp_role;
1198 req->ifbr_stpflags = bp->bp_flags;
1199 req->ifbr_addrcnt = bif->bif_addrcnt;
1200 req->ifbr_addrmax = bif->bif_addrmax;
1201 req->ifbr_addrexceeded = bif->bif_addrexceeded;
1202
1203 /* Copy STP state options as flags */
1204 if (bp->bp_operedge)
1205 req->ifbr_ifsflags |= IFBIF_BSTP_EDGE;
1206 if (bp->bp_flags & BSTP_PORT_AUTOEDGE)
1207 req->ifbr_ifsflags |= IFBIF_BSTP_AUTOEDGE;
1208 if (bp->bp_ptp_link)
1209 req->ifbr_ifsflags |= IFBIF_BSTP_PTP;
1210 if (bp->bp_flags & BSTP_PORT_AUTOPTP)
1211 req->ifbr_ifsflags |= IFBIF_BSTP_AUTOPTP;
1212 if (bp->bp_flags & BSTP_PORT_ADMEDGE)
1213 req->ifbr_ifsflags |= IFBIF_BSTP_ADMEDGE;
1214 if (bp->bp_flags & BSTP_PORT_ADMCOST)
1215 req->ifbr_ifsflags |= IFBIF_BSTP_ADMCOST;
1216 return (0);
1217 }
1218
1219 static int
1220 bridge_ioctl_sifflags(struct bridge_softc *sc, void *arg)
1221 {
1222 struct ifbreq *req = arg;
1223 struct bridge_iflist *bif;
1224 struct bstp_port *bp;
1225 int error;
1226
1227 bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1228 if (bif == NULL)
1229 return (ENOENT);
1230 bp = &bif->bif_stp;
1231
1232 if (req->ifbr_ifsflags & IFBIF_SPAN)
1233 /* SPAN is readonly */
1234 return (EINVAL);
1235
1236 if (req->ifbr_ifsflags & IFBIF_STP) {
1237 if ((bif->bif_flags & IFBIF_STP) == 0) {
1238 error = bstp_enable(&bif->bif_stp);
1239 if (error)
1240 return (error);
1241 }
1242 } else {
1243 if ((bif->bif_flags & IFBIF_STP) != 0)
1244 bstp_disable(&bif->bif_stp);
1245 }
1246
1247 /* Pass on STP flags */
1248 bstp_set_edge(bp, req->ifbr_ifsflags & IFBIF_BSTP_EDGE ? 1 : 0);
1249 bstp_set_autoedge(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOEDGE ? 1 : 0);
1250 bstp_set_ptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_PTP ? 1 : 0);
1251 bstp_set_autoptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOPTP ? 1 : 0);
1252
1253 /* Save the bits relating to the bridge */
1254 bif->bif_flags = req->ifbr_ifsflags & IFBIFMASK;
1255
1256 return (0);
1257 }
1258
1259 static int
1260 bridge_ioctl_scache(struct bridge_softc *sc, void *arg)
1261 {
1262 struct ifbrparam *param = arg;
1263
1264 sc->sc_brtmax = param->ifbrp_csize;
1265 bridge_rttrim(sc);
1266
1267 return (0);
1268 }
1269
1270 static int
1271 bridge_ioctl_gcache(struct bridge_softc *sc, void *arg)
1272 {
1273 struct ifbrparam *param = arg;
1274
1275 param->ifbrp_csize = sc->sc_brtmax;
1276
1277 return (0);
1278 }
1279
1280 static int
1281 bridge_ioctl_gifs(struct bridge_softc *sc, void *arg)
1282 {
1283 struct ifbifconf *bifc = arg;
1284 struct bridge_iflist *bif;
1285 struct ifbreq breq;
1286 char *buf, *outbuf;
1287 int count, buflen, len, error = 0;
1288
1289 count = 0;
1290 LIST_FOREACH(bif, &sc->sc_iflist, bif_next)
1291 count++;
1292 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1293 count++;
1294
1295 buflen = sizeof(breq) * count;
1296 if (bifc->ifbic_len == 0) {
1297 bifc->ifbic_len = buflen;
1298 return (0);
1299 }
1300 BRIDGE_UNLOCK(sc);
1301 outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO);
1302 BRIDGE_LOCK(sc);
1303
1304 count = 0;
1305 buf = outbuf;
1306 len = min(bifc->ifbic_len, buflen);
1307 bzero(&breq, sizeof(breq));
1308 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1309 if (len < sizeof(breq))
1310 break;
1311
1312 strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname,
1313 sizeof(breq.ifbr_ifsname));
1314 /* Fill in the ifbreq structure */
1315 error = bridge_ioctl_gifflags(sc, &breq);
1316 if (error)
1317 break;
1318 memcpy(buf, &breq, sizeof(breq));
1319 count++;
1320 buf += sizeof(breq);
1321 len -= sizeof(breq);
1322 }
1323 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) {
1324 if (len < sizeof(breq))
1325 break;
1326
1327 strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname,
1328 sizeof(breq.ifbr_ifsname));
1329 breq.ifbr_ifsflags = bif->bif_flags;
1330 breq.ifbr_portno = bif->bif_ifp->if_index & 0xfff;
1331 memcpy(buf, &breq, sizeof(breq));
1332 count++;
1333 buf += sizeof(breq);
1334 len -= sizeof(breq);
1335 }
1336
1337 BRIDGE_UNLOCK(sc);
1338 bifc->ifbic_len = sizeof(breq) * count;
1339 error = copyout(outbuf, bifc->ifbic_req, bifc->ifbic_len);
1340 BRIDGE_LOCK(sc);
1341 free(outbuf, M_TEMP);
1342 return (error);
1343 }
1344
1345 static int
1346 bridge_ioctl_rts(struct bridge_softc *sc, void *arg)
1347 {
1348 struct ifbaconf *bac = arg;
1349 struct bridge_rtnode *brt;
1350 struct ifbareq bareq;
1351 char *buf, *outbuf;
1352 int count, buflen, len, error = 0;
1353
1354 if (bac->ifbac_len == 0)
1355 return (0);
1356
1357 count = 0;
1358 LIST_FOREACH(brt, &sc->sc_rtlist, brt_list)
1359 count++;
1360 buflen = sizeof(bareq) * count;
1361
1362 BRIDGE_UNLOCK(sc);
1363 outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO);
1364 BRIDGE_LOCK(sc);
1365
1366 count = 0;
1367 buf = outbuf;
1368 len = min(bac->ifbac_len, buflen);
1369 bzero(&bareq, sizeof(bareq));
1370 LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) {
1371 if (len < sizeof(bareq))
1372 goto out;
1373 strlcpy(bareq.ifba_ifsname, brt->brt_ifp->if_xname,
1374 sizeof(bareq.ifba_ifsname));
1375 memcpy(bareq.ifba_dst, brt->brt_addr, sizeof(brt->brt_addr));
1376 bareq.ifba_vlan = brt->brt_vlan;
1377 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC &&
1378 time_uptime < brt->brt_expire)
1379 bareq.ifba_expire = brt->brt_expire - time_uptime;
1380 else
1381 bareq.ifba_expire = 0;
1382 bareq.ifba_flags = brt->brt_flags;
1383
1384 memcpy(buf, &bareq, sizeof(bareq));
1385 count++;
1386 buf += sizeof(bareq);
1387 len -= sizeof(bareq);
1388 }
1389 out:
1390 BRIDGE_UNLOCK(sc);
1391 bac->ifbac_len = sizeof(bareq) * count;
1392 error = copyout(outbuf, bac->ifbac_req, bac->ifbac_len);
1393 BRIDGE_LOCK(sc);
1394 free(outbuf, M_TEMP);
1395 return (error);
1396 }
1397
1398 static int
1399 bridge_ioctl_saddr(struct bridge_softc *sc, void *arg)
1400 {
1401 struct ifbareq *req = arg;
1402 struct bridge_iflist *bif;
1403 int error;
1404
1405 bif = bridge_lookup_member(sc, req->ifba_ifsname);
1406 if (bif == NULL)
1407 return (ENOENT);
1408
1409 error = bridge_rtupdate(sc, req->ifba_dst, req->ifba_vlan, bif, 1,
1410 req->ifba_flags);
1411
1412 return (error);
1413 }
1414
1415 static int
1416 bridge_ioctl_sto(struct bridge_softc *sc, void *arg)
1417 {
1418 struct ifbrparam *param = arg;
1419
1420 sc->sc_brttimeout = param->ifbrp_ctime;
1421 return (0);
1422 }
1423
1424 static int
1425 bridge_ioctl_gto(struct bridge_softc *sc, void *arg)
1426 {
1427 struct ifbrparam *param = arg;
1428
1429 param->ifbrp_ctime = sc->sc_brttimeout;
1430 return (0);
1431 }
1432
1433 static int
1434 bridge_ioctl_daddr(struct bridge_softc *sc, void *arg)
1435 {
1436 struct ifbareq *req = arg;
1437
1438 return (bridge_rtdaddr(sc, req->ifba_dst, req->ifba_vlan));
1439 }
1440
1441 static int
1442 bridge_ioctl_flush(struct bridge_softc *sc, void *arg)
1443 {
1444 struct ifbreq *req = arg;
1445
1446 bridge_rtflush(sc, req->ifbr_ifsflags);
1447 return (0);
1448 }
1449
1450 static int
1451 bridge_ioctl_gpri(struct bridge_softc *sc, void *arg)
1452 {
1453 struct ifbrparam *param = arg;
1454 struct bstp_state *bs = &sc->sc_stp;
1455
1456 param->ifbrp_prio = bs->bs_bridge_priority;
1457 return (0);
1458 }
1459
1460 static int
1461 bridge_ioctl_spri(struct bridge_softc *sc, void *arg)
1462 {
1463 struct ifbrparam *param = arg;
1464
1465 return (bstp_set_priority(&sc->sc_stp, param->ifbrp_prio));
1466 }
1467
1468 static int
1469 bridge_ioctl_ght(struct bridge_softc *sc, void *arg)
1470 {
1471 struct ifbrparam *param = arg;
1472 struct bstp_state *bs = &sc->sc_stp;
1473
1474 param->ifbrp_hellotime = bs->bs_bridge_htime >> 8;
1475 return (0);
1476 }
1477
1478 static int
1479 bridge_ioctl_sht(struct bridge_softc *sc, void *arg)
1480 {
1481 struct ifbrparam *param = arg;
1482
1483 return (bstp_set_htime(&sc->sc_stp, param->ifbrp_hellotime));
1484 }
1485
1486 static int
1487 bridge_ioctl_gfd(struct bridge_softc *sc, void *arg)
1488 {
1489 struct ifbrparam *param = arg;
1490 struct bstp_state *bs = &sc->sc_stp;
1491
1492 param->ifbrp_fwddelay = bs->bs_bridge_fdelay >> 8;
1493 return (0);
1494 }
1495
1496 static int
1497 bridge_ioctl_sfd(struct bridge_softc *sc, void *arg)
1498 {
1499 struct ifbrparam *param = arg;
1500
1501 return (bstp_set_fdelay(&sc->sc_stp, param->ifbrp_fwddelay));
1502 }
1503
1504 static int
1505 bridge_ioctl_gma(struct bridge_softc *sc, void *arg)
1506 {
1507 struct ifbrparam *param = arg;
1508 struct bstp_state *bs = &sc->sc_stp;
1509
1510 param->ifbrp_maxage = bs->bs_bridge_max_age >> 8;
1511 return (0);
1512 }
1513
1514 static int
1515 bridge_ioctl_sma(struct bridge_softc *sc, void *arg)
1516 {
1517 struct ifbrparam *param = arg;
1518
1519 return (bstp_set_maxage(&sc->sc_stp, param->ifbrp_maxage));
1520 }
1521
1522 static int
1523 bridge_ioctl_sifprio(struct bridge_softc *sc, void *arg)
1524 {
1525 struct ifbreq *req = arg;
1526 struct bridge_iflist *bif;
1527
1528 bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1529 if (bif == NULL)
1530 return (ENOENT);
1531
1532 return (bstp_set_port_priority(&bif->bif_stp, req->ifbr_priority));
1533 }
1534
1535 static int
1536 bridge_ioctl_sifcost(struct bridge_softc *sc, void *arg)
1537 {
1538 struct ifbreq *req = arg;
1539 struct bridge_iflist *bif;
1540
1541 bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1542 if (bif == NULL)
1543 return (ENOENT);
1544
1545 return (bstp_set_path_cost(&bif->bif_stp, req->ifbr_path_cost));
1546 }
1547
1548 static int
1549 bridge_ioctl_sifmaxaddr(struct bridge_softc *sc, void *arg)
1550 {
1551 struct ifbreq *req = arg;
1552 struct bridge_iflist *bif;
1553
1554 bif = bridge_lookup_member(sc, req->ifbr_ifsname);
1555 if (bif == NULL)
1556 return (ENOENT);
1557
1558 bif->bif_addrmax = req->ifbr_addrmax;
1559 return (0);
1560 }
1561
1562 static int
1563 bridge_ioctl_addspan(struct bridge_softc *sc, void *arg)
1564 {
1565 struct ifbreq *req = arg;
1566 struct bridge_iflist *bif = NULL;
1567 struct ifnet *ifs;
1568
1569 ifs = ifunit(req->ifbr_ifsname);
1570 if (ifs == NULL)
1571 return (ENOENT);
1572
1573 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1574 if (ifs == bif->bif_ifp)
1575 return (EBUSY);
1576
1577 if (ifs->if_bridge != NULL)
1578 return (EBUSY);
1579
1580 switch (ifs->if_type) {
1581 case IFT_ETHER:
1582 case IFT_GIF:
1583 case IFT_L2VLAN:
1584 break;
1585 default:
1586 return (EINVAL);
1587 }
1588
1589 bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO);
1590 if (bif == NULL)
1591 return (ENOMEM);
1592
1593 bif->bif_ifp = ifs;
1594 bif->bif_flags = IFBIF_SPAN;
1595
1596 LIST_INSERT_HEAD(&sc->sc_spanlist, bif, bif_next);
1597
1598 return (0);
1599 }
1600
1601 static int
1602 bridge_ioctl_delspan(struct bridge_softc *sc, void *arg)
1603 {
1604 struct ifbreq *req = arg;
1605 struct bridge_iflist *bif;
1606 struct ifnet *ifs;
1607
1608 ifs = ifunit(req->ifbr_ifsname);
1609 if (ifs == NULL)
1610 return (ENOENT);
1611
1612 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1613 if (ifs == bif->bif_ifp)
1614 break;
1615
1616 if (bif == NULL)
1617 return (ENOENT);
1618
1619 bridge_delete_span(sc, bif);
1620
1621 return (0);
1622 }
1623
1624 static int
1625 bridge_ioctl_gbparam(struct bridge_softc *sc, void *arg)
1626 {
1627 struct ifbropreq *req = arg;
1628 struct bstp_state *bs = &sc->sc_stp;
1629 struct bstp_port *root_port;
1630
1631 req->ifbop_maxage = bs->bs_bridge_max_age >> 8;
1632 req->ifbop_hellotime = bs->bs_bridge_htime >> 8;
1633 req->ifbop_fwddelay = bs->bs_bridge_fdelay >> 8;
1634
1635 root_port = bs->bs_root_port;
1636 if (root_port == NULL)
1637 req->ifbop_root_port = 0;
1638 else
1639 req->ifbop_root_port = root_port->bp_ifp->if_index;
1640
1641 req->ifbop_holdcount = bs->bs_txholdcount;
1642 req->ifbop_priority = bs->bs_bridge_priority;
1643 req->ifbop_protocol = bs->bs_protover;
1644 req->ifbop_root_path_cost = bs->bs_root_pv.pv_cost;
1645 req->ifbop_bridgeid = bs->bs_bridge_pv.pv_dbridge_id;
1646 req->ifbop_designated_root = bs->bs_root_pv.pv_root_id;
1647 req->ifbop_designated_bridge = bs->bs_root_pv.pv_dbridge_id;
1648 req->ifbop_last_tc_time.tv_sec = bs->bs_last_tc_time.tv_sec;
1649 req->ifbop_last_tc_time.tv_usec = bs->bs_last_tc_time.tv_usec;
1650
1651 return (0);
1652 }
1653
1654 static int
1655 bridge_ioctl_grte(struct bridge_softc *sc, void *arg)
1656 {
1657 struct ifbrparam *param = arg;
1658
1659 param->ifbrp_cexceeded = sc->sc_brtexceeded;
1660 return (0);
1661 }
1662
1663 static int
1664 bridge_ioctl_gifsstp(struct bridge_softc *sc, void *arg)
1665 {
1666 struct ifbpstpconf *bifstp = arg;
1667 struct bridge_iflist *bif;
1668 struct bstp_port *bp;
1669 struct ifbpstpreq bpreq;
1670 char *buf, *outbuf;
1671 int count, buflen, len, error = 0;
1672
1673 count = 0;
1674 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1675 if ((bif->bif_flags & IFBIF_STP) != 0)
1676 count++;
1677 }
1678
1679 buflen = sizeof(bpreq) * count;
1680 if (bifstp->ifbpstp_len == 0) {
1681 bifstp->ifbpstp_len = buflen;
1682 return (0);
1683 }
1684
1685 BRIDGE_UNLOCK(sc);
1686 outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO);
1687 BRIDGE_LOCK(sc);
1688
1689 count = 0;
1690 buf = outbuf;
1691 len = min(bifstp->ifbpstp_len, buflen);
1692 bzero(&bpreq, sizeof(bpreq));
1693 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1694 if (len < sizeof(bpreq))
1695 break;
1696
1697 if ((bif->bif_flags & IFBIF_STP) == 0)
1698 continue;
1699
1700 bp = &bif->bif_stp;
1701 bpreq.ifbp_portno = bif->bif_ifp->if_index & 0xfff;
1702 bpreq.ifbp_fwd_trans = bp->bp_forward_transitions;
1703 bpreq.ifbp_design_cost = bp->bp_desg_pv.pv_cost;
1704 bpreq.ifbp_design_port = bp->bp_desg_pv.pv_port_id;
1705 bpreq.ifbp_design_bridge = bp->bp_desg_pv.pv_dbridge_id;
1706 bpreq.ifbp_design_root = bp->bp_desg_pv.pv_root_id;
1707
1708 memcpy(buf, &bpreq, sizeof(bpreq));
1709 count++;
1710 buf += sizeof(bpreq);
1711 len -= sizeof(bpreq);
1712 }
1713
1714 BRIDGE_UNLOCK(sc);
1715 bifstp->ifbpstp_len = sizeof(bpreq) * count;
1716 error = copyout(outbuf, bifstp->ifbpstp_req, bifstp->ifbpstp_len);
1717 BRIDGE_LOCK(sc);
1718 free(outbuf, M_TEMP);
1719 return (error);
1720 }
1721
1722 static int
1723 bridge_ioctl_sproto(struct bridge_softc *sc, void *arg)
1724 {
1725 struct ifbrparam *param = arg;
1726
1727 return (bstp_set_protocol(&sc->sc_stp, param->ifbrp_proto));
1728 }
1729
1730 static int
1731 bridge_ioctl_stxhc(struct bridge_softc *sc, void *arg)
1732 {
1733 struct ifbrparam *param = arg;
1734
1735 return (bstp_set_holdcount(&sc->sc_stp, param->ifbrp_txhc));
1736 }
1737
1738 /*
1739 * bridge_ifdetach:
1740 *
1741 * Detach an interface from a bridge. Called when a member
1742 * interface is detaching.
1743 */
1744 static void
1745 bridge_ifdetach(void *arg __unused, struct ifnet *ifp)
1746 {
1747 struct bridge_softc *sc = ifp->if_bridge;
1748 struct bridge_iflist *bif;
1749
1750 if (ifp->if_flags & IFF_RENAMING)
1751 return;
1752
1753 /* Check if the interface is a bridge member */
1754 if (sc != NULL) {
1755 BRIDGE_LOCK(sc);
1756
1757 bif = bridge_lookup_member_if(sc, ifp);
1758 if (bif != NULL)
1759 bridge_delete_member(sc, bif, 1);
1760
1761 BRIDGE_UNLOCK(sc);
1762 return;
1763 }
1764
1765 /* Check if the interface is a span port */
1766 mtx_lock(&bridge_list_mtx);
1767 LIST_FOREACH(sc, &bridge_list, sc_list) {
1768 BRIDGE_LOCK(sc);
1769 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next)
1770 if (ifp == bif->bif_ifp) {
1771 bridge_delete_span(sc, bif);
1772 break;
1773 }
1774
1775 BRIDGE_UNLOCK(sc);
1776 }
1777 mtx_unlock(&bridge_list_mtx);
1778 }
1779
1780 /*
1781 * bridge_init:
1782 *
1783 * Initialize a bridge interface.
1784 */
1785 static void
1786 bridge_init(void *xsc)
1787 {
1788 struct bridge_softc *sc = (struct bridge_softc *)xsc;
1789 struct ifnet *ifp = sc->sc_ifp;
1790
1791 if (ifp->if_drv_flags & IFF_DRV_RUNNING)
1792 return;
1793
1794 BRIDGE_LOCK(sc);
1795 callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz,
1796 bridge_timer, sc);
1797
1798 ifp->if_drv_flags |= IFF_DRV_RUNNING;
1799 bstp_init(&sc->sc_stp); /* Initialize Spanning Tree */
1800
1801 BRIDGE_UNLOCK(sc);
1802 }
1803
1804 /*
1805 * bridge_stop:
1806 *
1807 * Stop the bridge interface.
1808 */
1809 static void
1810 bridge_stop(struct ifnet *ifp, int disable)
1811 {
1812 struct bridge_softc *sc = ifp->if_softc;
1813
1814 BRIDGE_LOCK_ASSERT(sc);
1815
1816 if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
1817 return;
1818
1819 callout_stop(&sc->sc_brcallout);
1820 bstp_stop(&sc->sc_stp);
1821
1822 bridge_rtflush(sc, IFBF_FLUSHDYN);
1823
1824 ifp->if_drv_flags &= ~IFF_DRV_RUNNING;
1825 }
1826
1827 /*
1828 * bridge_enqueue:
1829 *
1830 * Enqueue a packet on a bridge member interface.
1831 *
1832 */
1833 static void
1834 bridge_enqueue(struct bridge_softc *sc, struct ifnet *dst_ifp, struct mbuf *m)
1835 {
1836 int len, err = 0;
1837 short mflags;
1838 struct mbuf *m0;
1839
1840 len = m->m_pkthdr.len;
1841 mflags = m->m_flags;
1842
1843 /* We may be sending a fragment so traverse the mbuf */
1844 for (; m; m = m0) {
1845 m0 = m->m_nextpkt;
1846 m->m_nextpkt = NULL;
1847
1848 /*
1849 * If underlying interface can not do VLAN tag insertion itself
1850 * then attach a packet tag that holds it.
1851 */
1852 if ((m->m_flags & M_VLANTAG) &&
1853 (dst_ifp->if_capenable & IFCAP_VLAN_HWTAGGING) == 0) {
1854 m = ether_vlanencap(m, m->m_pkthdr.ether_vtag);
1855 if (m == NULL) {
1856 if_printf(dst_ifp,
1857 "unable to prepend VLAN header\n");
1858 dst_ifp->if_oerrors++;
1859 continue;
1860 }
1861 m->m_flags &= ~M_VLANTAG;
1862 }
1863
1864 if ((err = dst_ifp->if_transmit(dst_ifp, m))) {
1865 m_freem(m0);
1866 break;
1867 }
1868 }
1869
1870 if (err == 0) {
1871 sc->sc_ifp->if_opackets++;
1872 sc->sc_ifp->if_obytes += len;
1873 if (mflags & M_MCAST)
1874 sc->sc_ifp->if_omcasts++;
1875 }
1876 }
1877
1878 /*
1879 * bridge_dummynet:
1880 *
1881 * Receive a queued packet from dummynet and pass it on to the output
1882 * interface.
1883 *
1884 * The mbuf has the Ethernet header already attached.
1885 */
1886 static void
1887 bridge_dummynet(struct mbuf *m, struct ifnet *ifp)
1888 {
1889 struct bridge_softc *sc;
1890
1891 sc = ifp->if_bridge;
1892
1893 /*
1894 * The packet didnt originate from a member interface. This should only
1895 * ever happen if a member interface is removed while packets are
1896 * queued for it.
1897 */
1898 if (sc == NULL) {
1899 m_freem(m);
1900 return;
1901 }
1902
1903 if (PFIL_HOOKED(&V_inet_pfil_hook)
1904 #ifdef INET6
1905 || PFIL_HOOKED(&V_inet6_pfil_hook)
1906 #endif
1907 ) {
1908 if (bridge_pfil(&m, sc->sc_ifp, ifp, PFIL_OUT) != 0)
1909 return;
1910 if (m == NULL)
1911 return;
1912 }
1913
1914 bridge_enqueue(sc, ifp, m);
1915 }
1916
1917 /*
1918 * bridge_output:
1919 *
1920 * Send output from a bridge member interface. This
1921 * performs the bridging function for locally originated
1922 * packets.
1923 *
1924 * The mbuf has the Ethernet header already attached. We must
1925 * enqueue or free the mbuf before returning.
1926 */
1927 static int
1928 bridge_output(struct ifnet *ifp, struct mbuf *m, struct sockaddr *sa,
1929 struct rtentry *rt)
1930 {
1931 struct ether_header *eh;
1932 struct ifnet *dst_if;
1933 struct bridge_softc *sc;
1934 uint16_t vlan;
1935
1936 if (m->m_len < ETHER_HDR_LEN) {
1937 m = m_pullup(m, ETHER_HDR_LEN);
1938 if (m == NULL)
1939 return (0);
1940 }
1941
1942 eh = mtod(m, struct ether_header *);
1943 sc = ifp->if_bridge;
1944 vlan = VLANTAGOF(m);
1945
1946 BRIDGE_LOCK(sc);
1947
1948 /*
1949 * If bridge is down, but the original output interface is up,
1950 * go ahead and send out that interface. Otherwise, the packet
1951 * is dropped below.
1952 */
1953 if ((sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) {
1954 dst_if = ifp;
1955 goto sendunicast;
1956 }
1957
1958 /*
1959 * If the packet is a multicast, or we don't know a better way to
1960 * get there, send to all interfaces.
1961 */
1962 if (ETHER_IS_MULTICAST(eh->ether_dhost))
1963 dst_if = NULL;
1964 else
1965 dst_if = bridge_rtlookup(sc, eh->ether_dhost, vlan);
1966 if (dst_if == NULL) {
1967 struct bridge_iflist *bif;
1968 struct mbuf *mc;
1969 int error = 0, used = 0;
1970
1971 bridge_span(sc, m);
1972
1973 BRIDGE_LOCK2REF(sc, error);
1974 if (error) {
1975 m_freem(m);
1976 return (0);
1977 }
1978
1979 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
1980 dst_if = bif->bif_ifp;
1981
1982 if (dst_if->if_type == IFT_GIF)
1983 continue;
1984 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
1985 continue;
1986
1987 /*
1988 * If this is not the original output interface,
1989 * and the interface is participating in spanning
1990 * tree, make sure the port is in a state that
1991 * allows forwarding.
1992 */
1993 if (dst_if != ifp && (bif->bif_flags & IFBIF_STP) &&
1994 bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
1995 continue;
1996
1997 if (LIST_NEXT(bif, bif_next) == NULL) {
1998 used = 1;
1999 mc = m;
2000 } else {
2001 mc = m_copypacket(m, M_DONTWAIT);
2002 if (mc == NULL) {
2003 sc->sc_ifp->if_oerrors++;
2004 continue;
2005 }
2006 }
2007
2008 bridge_enqueue(sc, dst_if, mc);
2009 }
2010 if (used == 0)
2011 m_freem(m);
2012 BRIDGE_UNREF(sc);
2013 return (0);
2014 }
2015
2016 sendunicast:
2017 /*
2018 * XXX Spanning tree consideration here?
2019 */
2020
2021 bridge_span(sc, m);
2022 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) {
2023 m_freem(m);
2024 BRIDGE_UNLOCK(sc);
2025 return (0);
2026 }
2027
2028 BRIDGE_UNLOCK(sc);
2029 bridge_enqueue(sc, dst_if, m);
2030 return (0);
2031 }
2032
2033 /*
2034 * bridge_start:
2035 *
2036 * Start output on a bridge.
2037 *
2038 */
2039 static void
2040 bridge_start(struct ifnet *ifp)
2041 {
2042 struct bridge_softc *sc;
2043 struct mbuf *m;
2044 struct ether_header *eh;
2045 struct ifnet *dst_if;
2046
2047 sc = ifp->if_softc;
2048
2049 ifp->if_drv_flags |= IFF_DRV_OACTIVE;
2050 for (;;) {
2051 IFQ_DEQUEUE(&ifp->if_snd, m);
2052 if (m == 0)
2053 break;
2054 ETHER_BPF_MTAP(ifp, m);
2055
2056 eh = mtod(m, struct ether_header *);
2057 dst_if = NULL;
2058
2059 BRIDGE_LOCK(sc);
2060 if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) {
2061 dst_if = bridge_rtlookup(sc, eh->ether_dhost, 1);
2062 }
2063
2064 if (dst_if == NULL)
2065 bridge_broadcast(sc, ifp, m, 0);
2066 else {
2067 BRIDGE_UNLOCK(sc);
2068 bridge_enqueue(sc, dst_if, m);
2069 }
2070 }
2071 ifp->if_drv_flags &= ~IFF_DRV_OACTIVE;
2072 }
2073
2074 /*
2075 * bridge_forward:
2076 *
2077 * The forwarding function of the bridge.
2078 *
2079 * NOTE: Releases the lock on return.
2080 */
2081 static void
2082 bridge_forward(struct bridge_softc *sc, struct bridge_iflist *sbif,
2083 struct mbuf *m)
2084 {
2085 struct bridge_iflist *dbif;
2086 struct ifnet *src_if, *dst_if, *ifp;
2087 struct ether_header *eh;
2088 uint16_t vlan;
2089 uint8_t *dst;
2090 int error;
2091
2092 src_if = m->m_pkthdr.rcvif;
2093 ifp = sc->sc_ifp;
2094
2095 ifp->if_ipackets++;
2096 ifp->if_ibytes += m->m_pkthdr.len;
2097 vlan = VLANTAGOF(m);
2098
2099 if ((sbif->bif_flags & IFBIF_STP) &&
2100 sbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2101 goto drop;
2102
2103 eh = mtod(m, struct ether_header *);
2104 dst = eh->ether_dhost;
2105
2106 /* If the interface is learning, record the address. */
2107 if (sbif->bif_flags & IFBIF_LEARNING) {
2108 error = bridge_rtupdate(sc, eh->ether_shost, vlan,
2109 sbif, 0, IFBAF_DYNAMIC);
2110 /*
2111 * If the interface has addresses limits then deny any source
2112 * that is not in the cache.
2113 */
2114 if (error && sbif->bif_addrmax)
2115 goto drop;
2116 }
2117
2118 if ((sbif->bif_flags & IFBIF_STP) != 0 &&
2119 sbif->bif_stp.bp_state == BSTP_IFSTATE_LEARNING)
2120 goto drop;
2121
2122 /*
2123 * At this point, the port either doesn't participate
2124 * in spanning tree or it is in the forwarding state.
2125 */
2126
2127 /*
2128 * If the packet is unicast, destined for someone on
2129 * "this" side of the bridge, drop it.
2130 */
2131 if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) {
2132 dst_if = bridge_rtlookup(sc, dst, vlan);
2133 if (src_if == dst_if)
2134 goto drop;
2135 } else {
2136 /*
2137 * Check if its a reserved multicast address, any address
2138 * listed in 802.1D section 7.12.6 may not be forwarded by the
2139 * bridge.
2140 * This is currently 01-80-C2-00-00-00 to 01-80-C2-00-00-0F
2141 */
2142 if (dst[0] == 0x01 && dst[1] == 0x80 &&
2143 dst[2] == 0xc2 && dst[3] == 0x00 &&
2144 dst[4] == 0x00 && dst[5] <= 0x0f)
2145 goto drop;
2146
2147 /* ...forward it to all interfaces. */
2148 ifp->if_imcasts++;
2149 dst_if = NULL;
2150 }
2151
2152 /*
2153 * If we have a destination interface which is a member of our bridge,
2154 * OR this is a unicast packet, push it through the bpf(4) machinery.
2155 * For broadcast or multicast packets, don't bother because it will
2156 * be reinjected into ether_input. We do this before we pass the packets
2157 * through the pfil(9) framework, as it is possible that pfil(9) will
2158 * drop the packet, or possibly modify it, making it difficult to debug
2159 * firewall issues on the bridge.
2160 */
2161 if (dst_if != NULL || (m->m_flags & (M_BCAST | M_MCAST)) == 0)
2162 ETHER_BPF_MTAP(ifp, m);
2163
2164 /* run the packet filter */
2165 if (PFIL_HOOKED(&V_inet_pfil_hook)
2166 #ifdef INET6
2167 || PFIL_HOOKED(&V_inet6_pfil_hook)
2168 #endif
2169 ) {
2170 BRIDGE_UNLOCK(sc);
2171 if (bridge_pfil(&m, ifp, src_if, PFIL_IN) != 0)
2172 return;
2173 if (m == NULL)
2174 return;
2175 BRIDGE_LOCK(sc);
2176 }
2177
2178 if (dst_if == NULL) {
2179 bridge_broadcast(sc, src_if, m, 1);
2180 return;
2181 }
2182
2183 /*
2184 * At this point, we're dealing with a unicast frame
2185 * going to a different interface.
2186 */
2187 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2188 goto drop;
2189
2190 dbif = bridge_lookup_member_if(sc, dst_if);
2191 if (dbif == NULL)
2192 /* Not a member of the bridge (anymore?) */
2193 goto drop;
2194
2195 /* Private segments can not talk to each other */
2196 if (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE)
2197 goto drop;
2198
2199 if ((dbif->bif_flags & IFBIF_STP) &&
2200 dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2201 goto drop;
2202
2203 BRIDGE_UNLOCK(sc);
2204
2205 if (PFIL_HOOKED(&V_inet_pfil_hook)
2206 #ifdef INET6
2207 || PFIL_HOOKED(&V_inet6_pfil_hook)
2208 #endif
2209 ) {
2210 if (bridge_pfil(&m, ifp, dst_if, PFIL_OUT) != 0)
2211 return;
2212 if (m == NULL)
2213 return;
2214 }
2215
2216 bridge_enqueue(sc, dst_if, m);
2217 return;
2218
2219 drop:
2220 BRIDGE_UNLOCK(sc);
2221 m_freem(m);
2222 }
2223
2224 /*
2225 * bridge_input:
2226 *
2227 * Receive input from a member interface. Queue the packet for
2228 * bridging if it is not for us.
2229 */
2230 static struct mbuf *
2231 bridge_input(struct ifnet *ifp, struct mbuf *m)
2232 {
2233 struct bridge_softc *sc = ifp->if_bridge;
2234 struct bridge_iflist *bif, *bif2;
2235 struct ifnet *bifp;
2236 struct ether_header *eh;
2237 struct mbuf *mc, *mc2;
2238 uint16_t vlan;
2239 int error;
2240
2241 if ((sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
2242 return (m);
2243
2244 bifp = sc->sc_ifp;
2245 vlan = VLANTAGOF(m);
2246
2247 /*
2248 * Implement support for bridge monitoring. If this flag has been
2249 * set on this interface, discard the packet once we push it through
2250 * the bpf(4) machinery, but before we do, increment the byte and
2251 * packet counters associated with this interface.
2252 */
2253 if ((bifp->if_flags & IFF_MONITOR) != 0) {
2254 m->m_pkthdr.rcvif = bifp;
2255 ETHER_BPF_MTAP(bifp, m);
2256 bifp->if_ipackets++;
2257 bifp->if_ibytes += m->m_pkthdr.len;
2258 m_freem(m);
2259 return (NULL);
2260 }
2261 BRIDGE_LOCK(sc);
2262 bif = bridge_lookup_member_if(sc, ifp);
2263 if (bif == NULL) {
2264 BRIDGE_UNLOCK(sc);
2265 return (m);
2266 }
2267
2268 eh = mtod(m, struct ether_header *);
2269
2270 bridge_span(sc, m);
2271
2272 if (m->m_flags & (M_BCAST|M_MCAST)) {
2273 /* Tap off 802.1D packets; they do not get forwarded. */
2274 if (memcmp(eh->ether_dhost, bstp_etheraddr,
2275 ETHER_ADDR_LEN) == 0) {
2276 bstp_input(&bif->bif_stp, ifp, m); /* consumes mbuf */
2277 BRIDGE_UNLOCK(sc);
2278 return (NULL);
2279 }
2280
2281 if ((bif->bif_flags & IFBIF_STP) &&
2282 bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) {
2283 BRIDGE_UNLOCK(sc);
2284 return (m);
2285 }
2286
2287 /*
2288 * Make a deep copy of the packet and enqueue the copy
2289 * for bridge processing; return the original packet for
2290 * local processing.
2291 */
2292 mc = m_dup(m, M_DONTWAIT);
2293 if (mc == NULL) {
2294 BRIDGE_UNLOCK(sc);
2295 return (m);
2296 }
2297
2298 /* Perform the bridge forwarding function with the copy. */
2299 bridge_forward(sc, bif, mc);
2300
2301 /*
2302 * Reinject the mbuf as arriving on the bridge so we have a
2303 * chance at claiming multicast packets. We can not loop back
2304 * here from ether_input as a bridge is never a member of a
2305 * bridge.
2306 */
2307 KASSERT(bifp->if_bridge == NULL,
2308 ("loop created in bridge_input"));
2309 mc2 = m_dup(m, M_DONTWAIT);
2310 if (mc2 != NULL) {
2311 /* Keep the layer3 header aligned */
2312 int i = min(mc2->m_pkthdr.len, max_protohdr);
2313 mc2 = m_copyup(mc2, i, ETHER_ALIGN);
2314 }
2315 if (mc2 != NULL) {
2316 mc2->m_pkthdr.rcvif = bifp;
2317 (*bifp->if_input)(bifp, mc2);
2318 }
2319
2320 /* Return the original packet for local processing. */
2321 return (m);
2322 }
2323
2324 if ((bif->bif_flags & IFBIF_STP) &&
2325 bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) {
2326 BRIDGE_UNLOCK(sc);
2327 return (m);
2328 }
2329
2330 #if (defined(INET) || defined(INET6))
2331 # define OR_CARP_CHECK_WE_ARE_DST(iface) \
2332 || ((iface)->if_carp \
2333 && (*carp_forus_p)((iface), eh->ether_dhost))
2334 # define OR_CARP_CHECK_WE_ARE_SRC(iface) \
2335 || ((iface)->if_carp \
2336 && (*carp_forus_p)((iface), eh->ether_shost))
2337 #else
2338 # define OR_CARP_CHECK_WE_ARE_DST(iface)
2339 # define OR_CARP_CHECK_WE_ARE_SRC(iface)
2340 #endif
2341
2342 #ifdef INET6
2343 # define OR_PFIL_HOOKED_INET6 \
2344 || PFIL_HOOKED(&V_inet6_pfil_hook)
2345 #else
2346 # define OR_PFIL_HOOKED_INET6
2347 #endif
2348
2349 #define GRAB_OUR_PACKETS(iface) \
2350 if ((iface)->if_type == IFT_GIF) \
2351 continue; \
2352 /* It is destined for us. */ \
2353 if (memcmp(IF_LLADDR((iface)), eh->ether_dhost, ETHER_ADDR_LEN) == 0 \
2354 OR_CARP_CHECK_WE_ARE_DST((iface)) \
2355 ) { \
2356 if ((iface)->if_type == IFT_BRIDGE) { \
2357 ETHER_BPF_MTAP(iface, m); \
2358 iface->if_ipackets++; \
2359 /* Filter on the physical interface. */ \
2360 if (pfil_local_phys && \
2361 (PFIL_HOOKED(&V_inet_pfil_hook) \
2362 OR_PFIL_HOOKED_INET6)) { \
2363 if (bridge_pfil(&m, NULL, ifp, \
2364 PFIL_IN) != 0 || m == NULL) { \
2365 BRIDGE_UNLOCK(sc); \
2366 return (NULL); \
2367 } \
2368 eh = mtod(m, struct ether_header *); \
2369 } \
2370 } \
2371 if (bif->bif_flags & IFBIF_LEARNING) { \
2372 error = bridge_rtupdate(sc, eh->ether_shost, \
2373 vlan, bif, 0, IFBAF_DYNAMIC); \
2374 if (error && bif->bif_addrmax) { \
2375 BRIDGE_UNLOCK(sc); \
2376 m_freem(m); \
2377 return (NULL); \
2378 } \
2379 } \
2380 m->m_pkthdr.rcvif = iface; \
2381 BRIDGE_UNLOCK(sc); \
2382 return (m); \
2383 } \
2384 \
2385 /* We just received a packet that we sent out. */ \
2386 if (memcmp(IF_LLADDR((iface)), eh->ether_shost, ETHER_ADDR_LEN) == 0 \
2387 OR_CARP_CHECK_WE_ARE_SRC((iface)) \
2388 ) { \
2389 BRIDGE_UNLOCK(sc); \
2390 m_freem(m); \
2391 return (NULL); \
2392 }
2393
2394 /*
2395 * Unicast. Make sure it's not for the bridge.
2396 */
2397 do { GRAB_OUR_PACKETS(bifp) } while (0);
2398
2399 /*
2400 * Give a chance for ifp at first priority. This will help when the
2401 * packet comes through the interface like VLAN's with the same MACs
2402 * on several interfaces from the same bridge. This also will save
2403 * some CPU cycles in case the destination interface and the input
2404 * interface (eq ifp) are the same.
2405 */
2406 do { GRAB_OUR_PACKETS(ifp) } while (0);
2407
2408 /* Now check the all bridge members. */
2409 LIST_FOREACH(bif2, &sc->sc_iflist, bif_next) {
2410 GRAB_OUR_PACKETS(bif2->bif_ifp)
2411 }
2412
2413 #undef OR_CARP_CHECK_WE_ARE_DST
2414 #undef OR_CARP_CHECK_WE_ARE_SRC
2415 #undef OR_PFIL_HOOKED_INET6
2416 #undef GRAB_OUR_PACKETS
2417
2418 /* Perform the bridge forwarding function. */
2419 bridge_forward(sc, bif, m);
2420
2421 return (NULL);
2422 }
2423
2424 /*
2425 * bridge_broadcast:
2426 *
2427 * Send a frame to all interfaces that are members of
2428 * the bridge, except for the one on which the packet
2429 * arrived.
2430 *
2431 * NOTE: Releases the lock on return.
2432 */
2433 static void
2434 bridge_broadcast(struct bridge_softc *sc, struct ifnet *src_if,
2435 struct mbuf *m, int runfilt)
2436 {
2437 struct bridge_iflist *dbif, *sbif;
2438 struct mbuf *mc;
2439 struct ifnet *dst_if;
2440 int error = 0, used = 0, i;
2441
2442 sbif = bridge_lookup_member_if(sc, src_if);
2443
2444 BRIDGE_LOCK2REF(sc, error);
2445 if (error) {
2446 m_freem(m);
2447 return;
2448 }
2449
2450 /* Filter on the bridge interface before broadcasting */
2451 if (runfilt && (PFIL_HOOKED(&V_inet_pfil_hook)
2452 #ifdef INET6
2453 || PFIL_HOOKED(&V_inet6_pfil_hook)
2454 #endif
2455 )) {
2456 if (bridge_pfil(&m, sc->sc_ifp, NULL, PFIL_OUT) != 0)
2457 goto out;
2458 if (m == NULL)
2459 goto out;
2460 }
2461
2462 LIST_FOREACH(dbif, &sc->sc_iflist, bif_next) {
2463 dst_if = dbif->bif_ifp;
2464 if (dst_if == src_if)
2465 continue;
2466
2467 /* Private segments can not talk to each other */
2468 if (sbif && (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE))
2469 continue;
2470
2471 if ((dbif->bif_flags & IFBIF_STP) &&
2472 dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING)
2473 continue;
2474
2475 if ((dbif->bif_flags & IFBIF_DISCOVER) == 0 &&
2476 (m->m_flags & (M_BCAST|M_MCAST)) == 0)
2477 continue;
2478
2479 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2480 continue;
2481
2482 if (LIST_NEXT(dbif, bif_next) == NULL) {
2483 mc = m;
2484 used = 1;
2485 } else {
2486 mc = m_dup(m, M_DONTWAIT);
2487 if (mc == NULL) {
2488 sc->sc_ifp->if_oerrors++;
2489 continue;
2490 }
2491 }
2492
2493 /*
2494 * Filter on the output interface. Pass a NULL bridge interface
2495 * pointer so we do not redundantly filter on the bridge for
2496 * each interface we broadcast on.
2497 */
2498 if (runfilt && (PFIL_HOOKED(&V_inet_pfil_hook)
2499 #ifdef INET6
2500 || PFIL_HOOKED(&V_inet6_pfil_hook)
2501 #endif
2502 )) {
2503 if (used == 0) {
2504 /* Keep the layer3 header aligned */
2505 i = min(mc->m_pkthdr.len, max_protohdr);
2506 mc = m_copyup(mc, i, ETHER_ALIGN);
2507 if (mc == NULL) {
2508 sc->sc_ifp->if_oerrors++;
2509 continue;
2510 }
2511 }
2512 if (bridge_pfil(&mc, NULL, dst_if, PFIL_OUT) != 0)
2513 continue;
2514 if (mc == NULL)
2515 continue;
2516 }
2517
2518 bridge_enqueue(sc, dst_if, mc);
2519 }
2520 if (used == 0)
2521 m_freem(m);
2522
2523 out:
2524 BRIDGE_UNREF(sc);
2525 }
2526
2527 /*
2528 * bridge_span:
2529 *
2530 * Duplicate a packet out one or more interfaces that are in span mode,
2531 * the original mbuf is unmodified.
2532 */
2533 static void
2534 bridge_span(struct bridge_softc *sc, struct mbuf *m)
2535 {
2536 struct bridge_iflist *bif;
2537 struct ifnet *dst_if;
2538 struct mbuf *mc;
2539
2540 if (LIST_EMPTY(&sc->sc_spanlist))
2541 return;
2542
2543 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) {
2544 dst_if = bif->bif_ifp;
2545
2546 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0)
2547 continue;
2548
2549 mc = m_copypacket(m, M_DONTWAIT);
2550 if (mc == NULL) {
2551 sc->sc_ifp->if_oerrors++;
2552 continue;
2553 }
2554
2555 bridge_enqueue(sc, dst_if, mc);
2556 }
2557 }
2558
2559 /*
2560 * bridge_rtupdate:
2561 *
2562 * Add a bridge routing entry.
2563 */
2564 static int
2565 bridge_rtupdate(struct bridge_softc *sc, const uint8_t *dst, uint16_t vlan,
2566 struct bridge_iflist *bif, int setflags, uint8_t flags)
2567 {
2568 struct bridge_rtnode *brt;
2569 int error;
2570
2571 BRIDGE_LOCK_ASSERT(sc);
2572
2573 /* Check the source address is valid and not multicast. */
2574 if (ETHER_IS_MULTICAST(dst) ||
2575 (dst[0] == 0 && dst[1] == 0 && dst[2] == 0 &&
2576 dst[3] == 0 && dst[4] == 0 && dst[5] == 0) != 0)
2577 return (EINVAL);
2578
2579 /* 802.1p frames map to vlan 1 */
2580 if (vlan == 0)
2581 vlan = 1;
2582
2583 /*
2584 * A route for this destination might already exist. If so,
2585 * update it, otherwise create a new one.
2586 */
2587 if ((brt = bridge_rtnode_lookup(sc, dst, vlan)) == NULL) {
2588 if (sc->sc_brtcnt >= sc->sc_brtmax) {
2589 sc->sc_brtexceeded++;
2590 return (ENOSPC);
2591 }
2592 /* Check per interface address limits (if enabled) */
2593 if (bif->bif_addrmax && bif->bif_addrcnt >= bif->bif_addrmax) {
2594 bif->bif_addrexceeded++;
2595 return (ENOSPC);
2596 }
2597
2598 /*
2599 * Allocate a new bridge forwarding node, and
2600 * initialize the expiration time and Ethernet
2601 * address.
2602 */
2603 brt = uma_zalloc(bridge_rtnode_zone, M_NOWAIT | M_ZERO);
2604 if (brt == NULL)
2605 return (ENOMEM);
2606
2607 if (bif->bif_flags & IFBIF_STICKY)
2608 brt->brt_flags = IFBAF_STICKY;
2609 else
2610 brt->brt_flags = IFBAF_DYNAMIC;
2611
2612 memcpy(brt->brt_addr, dst, ETHER_ADDR_LEN);
2613 brt->brt_vlan = vlan;
2614
2615 if ((error = bridge_rtnode_insert(sc, brt)) != 0) {
2616 uma_zfree(bridge_rtnode_zone, brt);
2617 return (error);
2618 }
2619 brt->brt_dst = bif;
2620 bif->bif_addrcnt++;
2621 }
2622
2623 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC &&
2624 brt->brt_dst != bif) {
2625 brt->brt_dst->bif_addrcnt--;
2626 brt->brt_dst = bif;
2627 brt->brt_dst->bif_addrcnt++;
2628 }
2629
2630 if ((flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2631 brt->brt_expire = time_uptime + sc->sc_brttimeout;
2632 if (setflags)
2633 brt->brt_flags = flags;
2634
2635 return (0);
2636 }
2637
2638 /*
2639 * bridge_rtlookup:
2640 *
2641 * Lookup the destination interface for an address.
2642 */
2643 static struct ifnet *
2644 bridge_rtlookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2645 {
2646 struct bridge_rtnode *brt;
2647
2648 BRIDGE_LOCK_ASSERT(sc);
2649
2650 if ((brt = bridge_rtnode_lookup(sc, addr, vlan)) == NULL)
2651 return (NULL);
2652
2653 return (brt->brt_ifp);
2654 }
2655
2656 /*
2657 * bridge_rttrim:
2658 *
2659 * Trim the routine table so that we have a number
2660 * of routing entries less than or equal to the
2661 * maximum number.
2662 */
2663 static void
2664 bridge_rttrim(struct bridge_softc *sc)
2665 {
2666 struct bridge_rtnode *brt, *nbrt;
2667
2668 BRIDGE_LOCK_ASSERT(sc);
2669
2670 /* Make sure we actually need to do this. */
2671 if (sc->sc_brtcnt <= sc->sc_brtmax)
2672 return;
2673
2674 /* Force an aging cycle; this might trim enough addresses. */
2675 bridge_rtage(sc);
2676 if (sc->sc_brtcnt <= sc->sc_brtmax)
2677 return;
2678
2679 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2680 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) {
2681 bridge_rtnode_destroy(sc, brt);
2682 if (sc->sc_brtcnt <= sc->sc_brtmax)
2683 return;
2684 }
2685 }
2686 }
2687
2688 /*
2689 * bridge_timer:
2690 *
2691 * Aging timer for the bridge.
2692 */
2693 static void
2694 bridge_timer(void *arg)
2695 {
2696 struct bridge_softc *sc = arg;
2697
2698 BRIDGE_LOCK_ASSERT(sc);
2699
2700 bridge_rtage(sc);
2701
2702 if (sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING)
2703 callout_reset(&sc->sc_brcallout,
2704 bridge_rtable_prune_period * hz, bridge_timer, sc);
2705 }
2706
2707 /*
2708 * bridge_rtage:
2709 *
2710 * Perform an aging cycle.
2711 */
2712 static void
2713 bridge_rtage(struct bridge_softc *sc)
2714 {
2715 struct bridge_rtnode *brt, *nbrt;
2716
2717 BRIDGE_LOCK_ASSERT(sc);
2718
2719 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2720 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) {
2721 if (time_uptime >= brt->brt_expire)
2722 bridge_rtnode_destroy(sc, brt);
2723 }
2724 }
2725 }
2726
2727 /*
2728 * bridge_rtflush:
2729 *
2730 * Remove all dynamic addresses from the bridge.
2731 */
2732 static void
2733 bridge_rtflush(struct bridge_softc *sc, int full)
2734 {
2735 struct bridge_rtnode *brt, *nbrt;
2736
2737 BRIDGE_LOCK_ASSERT(sc);
2738
2739 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2740 if (full || (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2741 bridge_rtnode_destroy(sc, brt);
2742 }
2743 }
2744
2745 /*
2746 * bridge_rtdaddr:
2747 *
2748 * Remove an address from the table.
2749 */
2750 static int
2751 bridge_rtdaddr(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2752 {
2753 struct bridge_rtnode *brt;
2754 int found = 0;
2755
2756 BRIDGE_LOCK_ASSERT(sc);
2757
2758 /*
2759 * If vlan is zero then we want to delete for all vlans so the lookup
2760 * may return more than one.
2761 */
2762 while ((brt = bridge_rtnode_lookup(sc, addr, vlan)) != NULL) {
2763 bridge_rtnode_destroy(sc, brt);
2764 found = 1;
2765 }
2766
2767 return (found ? 0 : ENOENT);
2768 }
2769
2770 /*
2771 * bridge_rtdelete:
2772 *
2773 * Delete routes to a speicifc member interface.
2774 */
2775 static void
2776 bridge_rtdelete(struct bridge_softc *sc, struct ifnet *ifp, int full)
2777 {
2778 struct bridge_rtnode *brt, *nbrt;
2779
2780 BRIDGE_LOCK_ASSERT(sc);
2781
2782 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) {
2783 if (brt->brt_ifp == ifp && (full ||
2784 (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC))
2785 bridge_rtnode_destroy(sc, brt);
2786 }
2787 }
2788
2789 /*
2790 * bridge_rtable_init:
2791 *
2792 * Initialize the route table for this bridge.
2793 */
2794 static int
2795 bridge_rtable_init(struct bridge_softc *sc)
2796 {
2797 int i;
2798
2799 sc->sc_rthash = malloc(sizeof(*sc->sc_rthash) * BRIDGE_RTHASH_SIZE,
2800 M_DEVBUF, M_NOWAIT);
2801 if (sc->sc_rthash == NULL)
2802 return (ENOMEM);
2803
2804 for (i = 0; i < BRIDGE_RTHASH_SIZE; i++)
2805 LIST_INIT(&sc->sc_rthash[i]);
2806
2807 sc->sc_rthash_key = arc4random();
2808
2809 LIST_INIT(&sc->sc_rtlist);
2810
2811 return (0);
2812 }
2813
2814 /*
2815 * bridge_rtable_fini:
2816 *
2817 * Deconstruct the route table for this bridge.
2818 */
2819 static void
2820 bridge_rtable_fini(struct bridge_softc *sc)
2821 {
2822
2823 KASSERT(sc->sc_brtcnt == 0,
2824 ("%s: %d bridge routes referenced", __func__, sc->sc_brtcnt));
2825 free(sc->sc_rthash, M_DEVBUF);
2826 }
2827
2828 /*
2829 * The following hash function is adapted from "Hash Functions" by Bob Jenkins
2830 * ("Algorithm Alley", Dr. Dobbs Journal, September 1997).
2831 */
2832 #define mix(a, b, c) \
2833 do { \
2834 a -= b; a -= c; a ^= (c >> 13); \
2835 b -= c; b -= a; b ^= (a << 8); \
2836 c -= a; c -= b; c ^= (b >> 13); \
2837 a -= b; a -= c; a ^= (c >> 12); \
2838 b -= c; b -= a; b ^= (a << 16); \
2839 c -= a; c -= b; c ^= (b >> 5); \
2840 a -= b; a -= c; a ^= (c >> 3); \
2841 b -= c; b -= a; b ^= (a << 10); \
2842 c -= a; c -= b; c ^= (b >> 15); \
2843 } while (/*CONSTCOND*/0)
2844
2845 static __inline uint32_t
2846 bridge_rthash(struct bridge_softc *sc, const uint8_t *addr)
2847 {
2848 uint32_t a = 0x9e3779b9, b = 0x9e3779b9, c = sc->sc_rthash_key;
2849
2850 b += addr[5] << 8;
2851 b += addr[4];
2852 a += addr[3] << 24;
2853 a += addr[2] << 16;
2854 a += addr[1] << 8;
2855 a += addr[0];
2856
2857 mix(a, b, c);
2858
2859 return (c & BRIDGE_RTHASH_MASK);
2860 }
2861
2862 #undef mix
2863
2864 static int
2865 bridge_rtnode_addr_cmp(const uint8_t *a, const uint8_t *b)
2866 {
2867 int i, d;
2868
2869 for (i = 0, d = 0; i < ETHER_ADDR_LEN && d == 0; i++) {
2870 d = ((int)a[i]) - ((int)b[i]);
2871 }
2872
2873 return (d);
2874 }
2875
2876 /*
2877 * bridge_rtnode_lookup:
2878 *
2879 * Look up a bridge route node for the specified destination. Compare the
2880 * vlan id or if zero then just return the first match.
2881 */
2882 static struct bridge_rtnode *
2883 bridge_rtnode_lookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan)
2884 {
2885 struct bridge_rtnode *brt;
2886 uint32_t hash;
2887 int dir;
2888
2889 BRIDGE_LOCK_ASSERT(sc);
2890
2891 hash = bridge_rthash(sc, addr);
2892 LIST_FOREACH(brt, &sc->sc_rthash[hash], brt_hash) {
2893 dir = bridge_rtnode_addr_cmp(addr, brt->brt_addr);
2894 if (dir == 0 && (brt->brt_vlan == vlan || vlan == 0))
2895 return (brt);
2896 if (dir > 0)
2897 return (NULL);
2898 }
2899
2900 return (NULL);
2901 }
2902
2903 /*
2904 * bridge_rtnode_insert:
2905 *
2906 * Insert the specified bridge node into the route table. We
2907 * assume the entry is not already in the table.
2908 */
2909 static int
2910 bridge_rtnode_insert(struct bridge_softc *sc, struct bridge_rtnode *brt)
2911 {
2912 struct bridge_rtnode *lbrt;
2913 uint32_t hash;
2914 int dir;
2915
2916 BRIDGE_LOCK_ASSERT(sc);
2917
2918 hash = bridge_rthash(sc, brt->brt_addr);
2919
2920 lbrt = LIST_FIRST(&sc->sc_rthash[hash]);
2921 if (lbrt == NULL) {
2922 LIST_INSERT_HEAD(&sc->sc_rthash[hash], brt, brt_hash);
2923 goto out;
2924 }
2925
2926 do {
2927 dir = bridge_rtnode_addr_cmp(brt->brt_addr, lbrt->brt_addr);
2928 if (dir == 0 && brt->brt_vlan == lbrt->brt_vlan)
2929 return (EEXIST);
2930 if (dir > 0) {
2931 LIST_INSERT_BEFORE(lbrt, brt, brt_hash);
2932 goto out;
2933 }
2934 if (LIST_NEXT(lbrt, brt_hash) == NULL) {
2935 LIST_INSERT_AFTER(lbrt, brt, brt_hash);
2936 goto out;
2937 }
2938 lbrt = LIST_NEXT(lbrt, brt_hash);
2939 } while (lbrt != NULL);
2940
2941 #ifdef DIAGNOSTIC
2942 panic("bridge_rtnode_insert: impossible");
2943 #endif
2944
2945 out:
2946 LIST_INSERT_HEAD(&sc->sc_rtlist, brt, brt_list);
2947 sc->sc_brtcnt++;
2948
2949 return (0);
2950 }
2951
2952 /*
2953 * bridge_rtnode_destroy:
2954 *
2955 * Destroy a bridge rtnode.
2956 */
2957 static void
2958 bridge_rtnode_destroy(struct bridge_softc *sc, struct bridge_rtnode *brt)
2959 {
2960 BRIDGE_LOCK_ASSERT(sc);
2961
2962 LIST_REMOVE(brt, brt_hash);
2963
2964 LIST_REMOVE(brt, brt_list);
2965 sc->sc_brtcnt--;
2966 brt->brt_dst->bif_addrcnt--;
2967 uma_zfree(bridge_rtnode_zone, brt);
2968 }
2969
2970 /*
2971 * bridge_rtable_expire:
2972 *
2973 * Set the expiry time for all routes on an interface.
2974 */
2975 static void
2976 bridge_rtable_expire(struct ifnet *ifp, int age)
2977 {
2978 struct bridge_softc *sc = ifp->if_bridge;
2979 struct bridge_rtnode *brt;
2980
2981 BRIDGE_LOCK(sc);
2982
2983 /*
2984 * If the age is zero then flush, otherwise set all the expiry times to
2985 * age for the interface
2986 */
2987 if (age == 0)
2988 bridge_rtdelete(sc, ifp, IFBF_FLUSHDYN);
2989 else {
2990 LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) {
2991 /* Cap the expiry time to 'age' */
2992 if (brt->brt_ifp == ifp &&
2993 brt->brt_expire > time_uptime + age &&
2994 (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)
2995 brt->brt_expire = time_uptime + age;
2996 }
2997 }
2998 BRIDGE_UNLOCK(sc);
2999 }
3000
3001 /*
3002 * bridge_state_change:
3003 *
3004 * Callback from the bridgestp code when a port changes states.
3005 */
3006 static void
3007 bridge_state_change(struct ifnet *ifp, int state)
3008 {
3009 struct bridge_softc *sc = ifp->if_bridge;
3010 static const char *stpstates[] = {
3011 "disabled",
3012 "listening",
3013 "learning",
3014 "forwarding",
3015 "blocking",
3016 "discarding"
3017 };
3018
3019 if (log_stp)
3020 log(LOG_NOTICE, "%s: state changed to %s on %s\n",
3021 sc->sc_ifp->if_xname, stpstates[state], ifp->if_xname);
3022 }
3023
3024 /*
3025 * Send bridge packets through pfil if they are one of the types pfil can deal
3026 * with, or if they are ARP or REVARP. (pfil will pass ARP and REVARP without
3027 * question.) If *bifp or *ifp are NULL then packet filtering is skipped for
3028 * that interface.
3029 */
3030 static int
3031 bridge_pfil(struct mbuf **mp, struct ifnet *bifp, struct ifnet *ifp, int dir)
3032 {
3033 int snap, error, i, hlen;
3034 struct ether_header *eh1, eh2;
3035 struct ip_fw_args args;
3036 struct ip *ip;
3037 struct llc llc1;
3038 u_int16_t ether_type;
3039
3040 snap = 0;
3041 error = -1; /* Default error if not error == 0 */
3042
3043 #if 0
3044 /* we may return with the IP fields swapped, ensure its not shared */
3045 KASSERT(M_WRITABLE(*mp), ("%s: modifying a shared mbuf", __func__));
3046 #endif
3047
3048 if (pfil_bridge == 0 && pfil_member == 0 && pfil_ipfw == 0)
3049 return (0); /* filtering is disabled */
3050
3051 i = min((*mp)->m_pkthdr.len, max_protohdr);
3052 if ((*mp)->m_len < i) {
3053 *mp = m_pullup(*mp, i);
3054 if (*mp == NULL) {
3055 printf("%s: m_pullup failed\n", __func__);
3056 return (-1);
3057 }
3058 }
3059
3060 eh1 = mtod(*mp, struct ether_header *);
3061 ether_type = ntohs(eh1->ether_type);
3062
3063 /*
3064 * Check for SNAP/LLC.
3065 */
3066 if (ether_type < ETHERMTU) {
3067 struct llc *llc2 = (struct llc *)(eh1 + 1);
3068
3069 if ((*mp)->m_len >= ETHER_HDR_LEN + 8 &&
3070 llc2->llc_dsap == LLC_SNAP_LSAP &&
3071 llc2->llc_ssap == LLC_SNAP_LSAP &&
3072 llc2->llc_control == LLC_UI) {
3073 ether_type = htons(llc2->llc_un.type_snap.ether_type);
3074 snap = 1;
3075 }
3076 }
3077
3078 /*
3079 * If we're trying to filter bridge traffic, don't look at anything
3080 * other than IP and ARP traffic. If the filter doesn't understand
3081 * IPv6, don't allow IPv6 through the bridge either. This is lame
3082 * since if we really wanted, say, an AppleTalk filter, we are hosed,
3083 * but of course we don't have an AppleTalk filter to begin with.
3084 * (Note that since pfil doesn't understand ARP it will pass *ALL*
3085 * ARP traffic.)
3086 */
3087 switch (ether_type) {
3088 case ETHERTYPE_ARP:
3089 case ETHERTYPE_REVARP:
3090 if (pfil_ipfw_arp == 0)
3091 return (0); /* Automatically pass */
3092 break;
3093
3094 case ETHERTYPE_IP:
3095 #ifdef INET6
3096 case ETHERTYPE_IPV6:
3097 #endif /* INET6 */
3098 break;
3099 default:
3100 /*
3101 * Check to see if the user wants to pass non-ip
3102 * packets, these will not be checked by pfil(9) and
3103 * passed unconditionally so the default is to drop.
3104 */
3105 if (pfil_onlyip)
3106 goto bad;
3107 }
3108
3109 /* Strip off the Ethernet header and keep a copy. */
3110 m_copydata(*mp, 0, ETHER_HDR_LEN, (caddr_t) &eh2);
3111 m_adj(*mp, ETHER_HDR_LEN);
3112
3113 /* Strip off snap header, if present */
3114 if (snap) {
3115 m_copydata(*mp, 0, sizeof(struct llc), (caddr_t) &llc1);
3116 m_adj(*mp, sizeof(struct llc));
3117 }
3118
3119 /*
3120 * Check the IP header for alignment and errors
3121 */
3122 if (dir == PFIL_IN) {
3123 switch (ether_type) {
3124 case ETHERTYPE_IP:
3125 error = bridge_ip_checkbasic(mp);
3126 break;
3127 #ifdef INET6
3128 case ETHERTYPE_IPV6:
3129 error = bridge_ip6_checkbasic(mp);
3130 break;
3131 #endif /* INET6 */
3132 default:
3133 error = 0;
3134 }
3135 if (error)
3136 goto bad;
3137 }
3138
3139 /* XXX this section is also in if_ethersubr.c */
3140 // XXX PFIL_OUT or DIR_OUT ?
3141 if (V_ip_fw_chk_ptr && pfil_ipfw != 0 &&
3142 dir == PFIL_OUT && ifp != NULL) {
3143 struct m_tag *mtag;
3144
3145 error = -1;
3146 /* fetch the start point from existing tags, if any */
3147 mtag = m_tag_locate(*mp, MTAG_IPFW_RULE, 0, NULL);
3148 if (mtag == NULL) {
3149 args.rule.slot = 0;
3150 } else {
3151 struct ipfw_rule_ref *r;
3152
3153 /* XXX can we free the tag after use ? */
3154 mtag->m_tag_id = PACKET_TAG_NONE;
3155 r = (struct ipfw_rule_ref *)(mtag + 1);
3156 /* packet already partially processed ? */
3157 if (r->info & IPFW_ONEPASS)
3158 goto ipfwpass;
3159 args.rule = *r;
3160 }
3161
3162 args.m = *mp;
3163 args.oif = ifp;
3164 args.next_hop = NULL;
3165 args.next_hop6 = NULL;
3166 args.eh = &eh2;
3167 args.inp = NULL; /* used by ipfw uid/gid/jail rules */
3168 i = V_ip_fw_chk_ptr(&args);
3169 *mp = args.m;
3170
3171 if (*mp == NULL)
3172 return (error);
3173
3174 if (ip_dn_io_ptr && (i == IP_FW_DUMMYNET)) {
3175
3176 /* put the Ethernet header back on */
3177 M_PREPEND(*mp, ETHER_HDR_LEN, M_DONTWAIT);
3178 if (*mp == NULL)
3179 return (error);
3180 bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN);
3181
3182 /*
3183 * Pass the pkt to dummynet, which consumes it. The
3184 * packet will return to us via bridge_dummynet().
3185 */
3186 args.oif = ifp;
3187 ip_dn_io_ptr(mp, DIR_FWD | PROTO_IFB, &args);
3188 return (error);
3189 }
3190
3191 if (i != IP_FW_PASS) /* drop */
3192 goto bad;
3193 }
3194
3195 ipfwpass:
3196 error = 0;
3197
3198 /*
3199 * Run the packet through pfil
3200 */
3201 switch (ether_type) {
3202 case ETHERTYPE_IP:
3203 /*
3204 * before calling the firewall, swap fields the same as
3205 * IP does. here we assume the header is contiguous
3206 */
3207 ip = mtod(*mp, struct ip *);
3208
3209 ip->ip_len = ntohs(ip->ip_len);
3210 ip->ip_off = ntohs(ip->ip_off);
3211
3212 /*
3213 * Run pfil on the member interface and the bridge, both can
3214 * be skipped by clearing pfil_member or pfil_bridge.
3215 *
3216 * Keep the order:
3217 * in_if -> bridge_if -> out_if
3218 */
3219 if (pfil_bridge && dir == PFIL_OUT && bifp != NULL)
3220 error = pfil_run_hooks(&V_inet_pfil_hook, mp, bifp,
3221 dir, NULL);
3222
3223 if (*mp == NULL || error != 0) /* filter may consume */
3224 break;
3225
3226 if (pfil_member && ifp != NULL)
3227 error = pfil_run_hooks(&V_inet_pfil_hook, mp, ifp,
3228 dir, NULL);
3229
3230 if (*mp == NULL || error != 0) /* filter may consume */
3231 break;
3232
3233 if (pfil_bridge && dir == PFIL_IN && bifp != NULL)
3234 error = pfil_run_hooks(&V_inet_pfil_hook, mp, bifp,
3235 dir, NULL);
3236
3237 if (*mp == NULL || error != 0) /* filter may consume */
3238 break;
3239
3240 /* check if we need to fragment the packet */
3241 if (pfil_member && ifp != NULL && dir == PFIL_OUT) {
3242 i = (*mp)->m_pkthdr.len;
3243 if (i > ifp->if_mtu) {
3244 error = bridge_fragment(ifp, *mp, &eh2, snap,
3245 &llc1);
3246 return (error);
3247 }
3248 }
3249
3250 /* Recalculate the ip checksum and restore byte ordering */
3251 ip = mtod(*mp, struct ip *);
3252 hlen = ip->ip_hl << 2;
3253 if (hlen < sizeof(struct ip))
3254 goto bad;
3255 if (hlen > (*mp)->m_len) {
3256 if ((*mp = m_pullup(*mp, hlen)) == 0)
3257 goto bad;
3258 ip = mtod(*mp, struct ip *);
3259 if (ip == NULL)
3260 goto bad;
3261 }
3262 ip->ip_len = htons(ip->ip_len);
3263 ip->ip_off = htons(ip->ip_off);
3264 ip->ip_sum = 0;
3265 if (hlen == sizeof(struct ip))
3266 ip->ip_sum = in_cksum_hdr(ip);
3267 else
3268 ip->ip_sum = in_cksum(*mp, hlen);
3269
3270 break;
3271 #ifdef INET6
3272 case ETHERTYPE_IPV6:
3273 if (pfil_bridge && dir == PFIL_OUT && bifp != NULL)
3274 error = pfil_run_hooks(&V_inet6_pfil_hook, mp, bifp,
3275 dir, NULL);
3276
3277 if (*mp == NULL || error != 0) /* filter may consume */
3278 break;
3279
3280 if (pfil_member && ifp != NULL)
3281 error = pfil_run_hooks(&V_inet6_pfil_hook, mp, ifp,
3282 dir, NULL);
3283
3284 if (*mp == NULL || error != 0) /* filter may consume */
3285 break;
3286
3287 if (pfil_bridge && dir == PFIL_IN && bifp != NULL)
3288 error = pfil_run_hooks(&V_inet6_pfil_hook, mp, bifp,
3289 dir, NULL);
3290 break;
3291 #endif
3292 default:
3293 error = 0;
3294 break;
3295 }
3296
3297 if (*mp == NULL)
3298 return (error);
3299 if (error != 0)
3300 goto bad;
3301
3302 error = -1;
3303
3304 /*
3305 * Finally, put everything back the way it was and return
3306 */
3307 if (snap) {
3308 M_PREPEND(*mp, sizeof(struct llc), M_DONTWAIT);
3309 if (*mp == NULL)
3310 return (error);
3311 bcopy(&llc1, mtod(*mp, caddr_t), sizeof(struct llc));
3312 }
3313
3314 M_PREPEND(*mp, ETHER_HDR_LEN, M_DONTWAIT);
3315 if (*mp == NULL)
3316 return (error);
3317 bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN);
3318
3319 return (0);
3320
3321 bad:
3322 m_freem(*mp);
3323 *mp = NULL;
3324 return (error);
3325 }
3326
3327 /*
3328 * Perform basic checks on header size since
3329 * pfil assumes ip_input has already processed
3330 * it for it. Cut-and-pasted from ip_input.c.
3331 * Given how simple the IPv6 version is,
3332 * does the IPv4 version really need to be
3333 * this complicated?
3334 *
3335 * XXX Should we update ipstat here, or not?
3336 * XXX Right now we update ipstat but not
3337 * XXX csum_counter.
3338 */
3339 static int
3340 bridge_ip_checkbasic(struct mbuf **mp)
3341 {
3342 struct mbuf *m = *mp;
3343 struct ip *ip;
3344 int len, hlen;
3345 u_short sum;
3346
3347 if (*mp == NULL)
3348 return (-1);
3349
3350 if (IP_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) {
3351 if ((m = m_copyup(m, sizeof(struct ip),
3352 (max_linkhdr + 3) & ~3)) == NULL) {
3353 /* XXXJRT new stat, please */
3354 KMOD_IPSTAT_INC(ips_toosmall);
3355 goto bad;
3356 }
3357 } else if (__predict_false(m->m_len < sizeof (struct ip))) {
3358 if ((m = m_pullup(m, sizeof (struct ip))) == NULL) {
3359 KMOD_IPSTAT_INC(ips_toosmall);
3360 goto bad;
3361 }
3362 }
3363 ip = mtod(m, struct ip *);
3364 if (ip == NULL) goto bad;
3365
3366 if (ip->ip_v != IPVERSION) {
3367 KMOD_IPSTAT_INC(ips_badvers);
3368 goto bad;
3369 }
3370 hlen = ip->ip_hl << 2;
3371 if (hlen < sizeof(struct ip)) { /* minimum header length */
3372 KMOD_IPSTAT_INC(ips_badhlen);
3373 goto bad;
3374 }
3375 if (hlen > m->m_len) {
3376 if ((m = m_pullup(m, hlen)) == 0) {
3377 KMOD_IPSTAT_INC(ips_badhlen);
3378 goto bad;
3379 }
3380 ip = mtod(m, struct ip *);
3381 if (ip == NULL) goto bad;
3382 }
3383
3384 if (m->m_pkthdr.csum_flags & CSUM_IP_CHECKED) {
3385 sum = !(m->m_pkthdr.csum_flags & CSUM_IP_VALID);
3386 } else {
3387 if (hlen == sizeof(struct ip)) {
3388 sum = in_cksum_hdr(ip);
3389 } else {
3390 sum = in_cksum(m, hlen);
3391 }
3392 }
3393 if (sum) {
3394 KMOD_IPSTAT_INC(ips_badsum);
3395 goto bad;
3396 }
3397
3398 /* Retrieve the packet length. */
3399 len = ntohs(ip->ip_len);
3400
3401 /*
3402 * Check for additional length bogosity
3403 */
3404 if (len < hlen) {
3405 KMOD_IPSTAT_INC(ips_badlen);
3406 goto bad;
3407 }
3408
3409 /*
3410 * Check that the amount of data in the buffers
3411 * is as at least much as the IP header would have us expect.
3412 * Drop packet if shorter than we expect.
3413 */
3414 if (m->m_pkthdr.len < len) {
3415 KMOD_IPSTAT_INC(ips_tooshort);
3416 goto bad;
3417 }
3418
3419 /* Checks out, proceed */
3420 *mp = m;
3421 return (0);
3422
3423 bad:
3424 *mp = m;
3425 return (-1);
3426 }
3427
3428 #ifdef INET6
3429 /*
3430 * Same as above, but for IPv6.
3431 * Cut-and-pasted from ip6_input.c.
3432 * XXX Should we update ip6stat, or not?
3433 */
3434 static int
3435 bridge_ip6_checkbasic(struct mbuf **mp)
3436 {
3437 struct mbuf *m = *mp;
3438 struct ip6_hdr *ip6;
3439
3440 /*
3441 * If the IPv6 header is not aligned, slurp it up into a new
3442 * mbuf with space for link headers, in the event we forward
3443 * it. Otherwise, if it is aligned, make sure the entire base
3444 * IPv6 header is in the first mbuf of the chain.
3445 */
3446 if (IP6_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) {
3447 struct ifnet *inifp = m->m_pkthdr.rcvif;
3448 if ((m = m_copyup(m, sizeof(struct ip6_hdr),
3449 (max_linkhdr + 3) & ~3)) == NULL) {
3450 /* XXXJRT new stat, please */
3451 IP6STAT_INC(ip6s_toosmall);
3452 in6_ifstat_inc(inifp, ifs6_in_hdrerr);
3453 goto bad;
3454 }
3455 } else if (__predict_false(m->m_len < sizeof(struct ip6_hdr))) {
3456 struct ifnet *inifp = m->m_pkthdr.rcvif;
3457 if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) {
3458 IP6STAT_INC(ip6s_toosmall);
3459 in6_ifstat_inc(inifp, ifs6_in_hdrerr);
3460 goto bad;
3461 }
3462 }
3463
3464 ip6 = mtod(m, struct ip6_hdr *);
3465
3466 if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) {
3467 IP6STAT_INC(ip6s_badvers);
3468 in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr);
3469 goto bad;
3470 }
3471
3472 /* Checks out, proceed */
3473 *mp = m;
3474 return (0);
3475
3476 bad:
3477 *mp = m;
3478 return (-1);
3479 }
3480 #endif /* INET6 */
3481
3482 /*
3483 * bridge_fragment:
3484 *
3485 * Return a fragmented mbuf chain.
3486 */
3487 static int
3488 bridge_fragment(struct ifnet *ifp, struct mbuf *m, struct ether_header *eh,
3489 int snap, struct llc *llc)
3490 {
3491 struct mbuf *m0;
3492 struct ip *ip;
3493 int error = -1;
3494
3495 if (m->m_len < sizeof(struct ip) &&
3496 (m = m_pullup(m, sizeof(struct ip))) == NULL)
3497 goto out;
3498 ip = mtod(m, struct ip *);
3499
3500 error = ip_fragment(ip, &m, ifp->if_mtu, ifp->if_hwassist,
3501 CSUM_DELAY_IP);
3502 if (error)
3503 goto out;
3504
3505 /* walk the chain and re-add the Ethernet header */
3506 for (m0 = m; m0; m0 = m0->m_nextpkt) {
3507 if (error == 0) {
3508 if (snap) {
3509 M_PREPEND(m0, sizeof(struct llc), M_DONTWAIT);
3510 if (m0 == NULL) {
3511 error = ENOBUFS;
3512 continue;
3513 }
3514 bcopy(llc, mtod(m0, caddr_t),
3515 sizeof(struct llc));
3516 }
3517 M_PREPEND(m0, ETHER_HDR_LEN, M_DONTWAIT);
3518 if (m0 == NULL) {
3519 error = ENOBUFS;
3520 continue;
3521 }
3522 bcopy(eh, mtod(m0, caddr_t), ETHER_HDR_LEN);
3523 } else
3524 m_freem(m);
3525 }
3526
3527 if (error == 0)
3528 KMOD_IPSTAT_INC(ips_fragmented);
3529
3530 return (error);
3531
3532 out:
3533 if (m != NULL)
3534 m_freem(m);
3535 return (error);
3536 }
3537
3538 static void
3539 bridge_linkstate(struct ifnet *ifp)
3540 {
3541 struct bridge_softc *sc = ifp->if_bridge;
3542 struct bridge_iflist *bif;
3543
3544 BRIDGE_LOCK(sc);
3545 bif = bridge_lookup_member_if(sc, ifp);
3546 if (bif == NULL) {
3547 BRIDGE_UNLOCK(sc);
3548 return;
3549 }
3550 bridge_linkcheck(sc);
3551 BRIDGE_UNLOCK(sc);
3552
3553 bstp_linkstate(&bif->bif_stp);
3554 }
3555
3556 static void
3557 bridge_linkcheck(struct bridge_softc *sc)
3558 {
3559 struct bridge_iflist *bif;
3560 int new_link, hasls;
3561
3562 BRIDGE_LOCK_ASSERT(sc);
3563 new_link = LINK_STATE_DOWN;
3564 hasls = 0;
3565 /* Our link is considered up if at least one of our ports is active */
3566 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) {
3567 if (bif->bif_ifp->if_capabilities & IFCAP_LINKSTATE)
3568 hasls++;
3569 if (bif->bif_ifp->if_link_state == LINK_STATE_UP) {
3570 new_link = LINK_STATE_UP;
3571 break;
3572 }
3573 }
3574 if (!LIST_EMPTY(&sc->sc_iflist) && !hasls) {
3575 /* If no interfaces support link-state then we default to up */
3576 new_link = LINK_STATE_UP;
3577 }
3578 if_link_state_change(sc->sc_ifp, new_link);
3579 }
Cache object: 8988afff82470eed0dc232723472bb3f
|