1 /* $NetBSD: if_bridge.c,v 1.31 2005/06/01 19:45:34 jdc Exp $ */ 2 3 /*- 4 * SPDX-License-Identifier: BSD-4-Clause 5 * 6 * Copyright 2001 Wasabi Systems, Inc. 7 * All rights reserved. 8 * 9 * Written by Jason R. Thorpe for Wasabi Systems, Inc. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 3. All advertising materials mentioning features or use of this software 20 * must display the following acknowledgement: 21 * This product includes software developed for the NetBSD Project by 22 * Wasabi Systems, Inc. 23 * 4. The name of Wasabi Systems, Inc. may not be used to endorse 24 * or promote products derived from this software without specific prior 25 * written permission. 26 * 27 * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND 28 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 29 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 30 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WASABI SYSTEMS, INC 31 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 32 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 33 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 34 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 35 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 36 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 37 * POSSIBILITY OF SUCH DAMAGE. 38 */ 39 40 /* 41 * Copyright (c) 1999, 2000 Jason L. Wright (jason@thought.net) 42 * All rights reserved. 43 * 44 * Redistribution and use in source and binary forms, with or without 45 * modification, are permitted provided that the following conditions 46 * are met: 47 * 1. Redistributions of source code must retain the above copyright 48 * notice, this list of conditions and the following disclaimer. 49 * 2. Redistributions in binary form must reproduce the above copyright 50 * notice, this list of conditions and the following disclaimer in the 51 * documentation and/or other materials provided with the distribution. 52 * 53 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR 54 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 55 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 56 * DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, 57 * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 58 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR 59 * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 60 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 61 * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN 62 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 63 * POSSIBILITY OF SUCH DAMAGE. 64 * 65 * OpenBSD: if_bridge.c,v 1.60 2001/06/15 03:38:33 itojun Exp 66 */ 67 68 /* 69 * Network interface bridge support. 70 * 71 * TODO: 72 * 73 * - Currently only supports Ethernet-like interfaces (Ethernet, 74 * 802.11, VLANs on Ethernet, etc.) Figure out a nice way 75 * to bridge other types of interfaces (maybe consider 76 * heterogeneous bridges). 77 */ 78 79 #include <sys/cdefs.h> 80 __FBSDID("$FreeBSD$"); 81 82 #include "opt_inet.h" 83 #include "opt_inet6.h" 84 85 #include <sys/param.h> 86 #include <sys/eventhandler.h> 87 #include <sys/mbuf.h> 88 #include <sys/malloc.h> 89 #include <sys/protosw.h> 90 #include <sys/systm.h> 91 #include <sys/jail.h> 92 #include <sys/time.h> 93 #include <sys/socket.h> /* for net/if.h */ 94 #include <sys/sockio.h> 95 #include <sys/ctype.h> /* string functions */ 96 #include <sys/kernel.h> 97 #include <sys/random.h> 98 #include <sys/syslog.h> 99 #include <sys/sysctl.h> 100 #include <vm/uma.h> 101 #include <sys/module.h> 102 #include <sys/priv.h> 103 #include <sys/proc.h> 104 #include <sys/lock.h> 105 #include <sys/mutex.h> 106 107 #include <net/bpf.h> 108 #include <net/if.h> 109 #include <net/if_clone.h> 110 #include <net/if_dl.h> 111 #include <net/if_types.h> 112 #include <net/if_var.h> 113 #include <net/pfil.h> 114 #include <net/vnet.h> 115 116 #include <netinet/in.h> 117 #include <netinet/in_systm.h> 118 #include <netinet/in_var.h> 119 #include <netinet/ip.h> 120 #include <netinet/ip_var.h> 121 #ifdef INET6 122 #include <netinet/ip6.h> 123 #include <netinet6/ip6_var.h> 124 #include <netinet6/in6_ifattach.h> 125 #endif 126 #if defined(INET) || defined(INET6) 127 #include <netinet/ip_carp.h> 128 #endif 129 #include <machine/in_cksum.h> 130 #include <netinet/if_ether.h> 131 #include <net/bridgestp.h> 132 #include <net/if_bridgevar.h> 133 #include <net/if_llc.h> 134 #include <net/if_vlan_var.h> 135 136 #include <net/route.h> 137 138 #ifdef INET6 139 /* 140 * XXX: declare here to avoid to include many inet6 related files.. 141 * should be more generalized? 142 */ 143 extern void nd6_setmtu(struct ifnet *); 144 #endif 145 146 /* 147 * Size of the route hash table. Must be a power of two. 148 */ 149 #ifndef BRIDGE_RTHASH_SIZE 150 #define BRIDGE_RTHASH_SIZE 1024 151 #endif 152 153 #define BRIDGE_RTHASH_MASK (BRIDGE_RTHASH_SIZE - 1) 154 155 /* 156 * Default maximum number of addresses to cache. 157 */ 158 #ifndef BRIDGE_RTABLE_MAX 159 #define BRIDGE_RTABLE_MAX 2000 160 #endif 161 162 /* 163 * Timeout (in seconds) for entries learned dynamically. 164 */ 165 #ifndef BRIDGE_RTABLE_TIMEOUT 166 #define BRIDGE_RTABLE_TIMEOUT (20 * 60) /* same as ARP */ 167 #endif 168 169 /* 170 * Number of seconds between walks of the route list. 171 */ 172 #ifndef BRIDGE_RTABLE_PRUNE_PERIOD 173 #define BRIDGE_RTABLE_PRUNE_PERIOD (5 * 60) 174 #endif 175 176 /* 177 * List of capabilities to possibly mask on the member interface. 178 */ 179 #define BRIDGE_IFCAPS_MASK (IFCAP_TOE|IFCAP_TSO|IFCAP_TXCSUM|\ 180 IFCAP_TXCSUM_IPV6) 181 182 /* 183 * List of capabilities to strip 184 */ 185 #define BRIDGE_IFCAPS_STRIP IFCAP_LRO 186 187 /* 188 * Bridge interface list entry. 189 */ 190 struct bridge_iflist { 191 LIST_ENTRY(bridge_iflist) bif_next; 192 struct ifnet *bif_ifp; /* member if */ 193 struct bstp_port bif_stp; /* STP state */ 194 uint32_t bif_flags; /* member if flags */ 195 int bif_savedcaps; /* saved capabilities */ 196 uint32_t bif_addrmax; /* max # of addresses */ 197 uint32_t bif_addrcnt; /* cur. # of addresses */ 198 uint32_t bif_addrexceeded;/* # of address violations */ 199 }; 200 201 /* 202 * Bridge route node. 203 */ 204 struct bridge_rtnode { 205 LIST_ENTRY(bridge_rtnode) brt_hash; /* hash table linkage */ 206 LIST_ENTRY(bridge_rtnode) brt_list; /* list linkage */ 207 struct bridge_iflist *brt_dst; /* destination if */ 208 unsigned long brt_expire; /* expiration time */ 209 uint8_t brt_flags; /* address flags */ 210 uint8_t brt_addr[ETHER_ADDR_LEN]; 211 uint16_t brt_vlan; /* vlan id */ 212 }; 213 #define brt_ifp brt_dst->bif_ifp 214 215 /* 216 * Software state for each bridge. 217 */ 218 struct bridge_softc { 219 struct ifnet *sc_ifp; /* make this an interface */ 220 LIST_ENTRY(bridge_softc) sc_list; 221 struct mtx sc_mtx; 222 struct cv sc_cv; 223 uint32_t sc_brtmax; /* max # of addresses */ 224 uint32_t sc_brtcnt; /* cur. # of addresses */ 225 uint32_t sc_brttimeout; /* rt timeout in seconds */ 226 struct callout sc_brcallout; /* bridge callout */ 227 uint32_t sc_iflist_ref; /* refcount for sc_iflist */ 228 uint32_t sc_iflist_xcnt; /* refcount for sc_iflist */ 229 LIST_HEAD(, bridge_iflist) sc_iflist; /* member interface list */ 230 LIST_HEAD(, bridge_rtnode) *sc_rthash; /* our forwarding table */ 231 LIST_HEAD(, bridge_rtnode) sc_rtlist; /* list version of above */ 232 uint32_t sc_rthash_key; /* key for hash */ 233 LIST_HEAD(, bridge_iflist) sc_spanlist; /* span ports list */ 234 struct bstp_state sc_stp; /* STP state */ 235 uint32_t sc_brtexceeded; /* # of cache drops */ 236 struct ifnet *sc_ifaddr; /* member mac copied from */ 237 struct ether_addr sc_defaddr; /* Default MAC address */ 238 }; 239 240 VNET_DEFINE_STATIC(struct mtx, bridge_list_mtx); 241 #define V_bridge_list_mtx VNET(bridge_list_mtx) 242 static eventhandler_tag bridge_detach_cookie; 243 244 int bridge_rtable_prune_period = BRIDGE_RTABLE_PRUNE_PERIOD; 245 246 VNET_DEFINE_STATIC(uma_zone_t, bridge_rtnode_zone); 247 #define V_bridge_rtnode_zone VNET(bridge_rtnode_zone) 248 249 static int bridge_clone_create(struct if_clone *, int, caddr_t); 250 static void bridge_clone_destroy(struct ifnet *); 251 252 static int bridge_ioctl(struct ifnet *, u_long, caddr_t); 253 static void bridge_mutecaps(struct bridge_softc *); 254 static void bridge_set_ifcap(struct bridge_softc *, struct bridge_iflist *, 255 int); 256 static void bridge_ifdetach(void *arg __unused, struct ifnet *); 257 static void bridge_init(void *); 258 static void bridge_dummynet(struct mbuf *, struct ifnet *); 259 static void bridge_stop(struct ifnet *, int); 260 static int bridge_transmit(struct ifnet *, struct mbuf *); 261 static void bridge_qflush(struct ifnet *); 262 static struct mbuf *bridge_input(struct ifnet *, struct mbuf *); 263 static int bridge_output(struct ifnet *, struct mbuf *, struct sockaddr *, 264 struct rtentry *); 265 static int bridge_enqueue(struct bridge_softc *, struct ifnet *, 266 struct mbuf *); 267 static void bridge_rtdelete(struct bridge_softc *, struct ifnet *ifp, int); 268 269 static void bridge_forward(struct bridge_softc *, struct bridge_iflist *, 270 struct mbuf *m); 271 272 static void bridge_timer(void *); 273 274 static void bridge_broadcast(struct bridge_softc *, struct ifnet *, 275 struct mbuf *, int); 276 static void bridge_span(struct bridge_softc *, struct mbuf *); 277 278 static int bridge_rtupdate(struct bridge_softc *, const uint8_t *, 279 uint16_t, struct bridge_iflist *, int, uint8_t); 280 static struct ifnet *bridge_rtlookup(struct bridge_softc *, const uint8_t *, 281 uint16_t); 282 static void bridge_rttrim(struct bridge_softc *); 283 static void bridge_rtage(struct bridge_softc *); 284 static void bridge_rtflush(struct bridge_softc *, int); 285 static int bridge_rtdaddr(struct bridge_softc *, const uint8_t *, 286 uint16_t); 287 288 static void bridge_rtable_init(struct bridge_softc *); 289 static void bridge_rtable_fini(struct bridge_softc *); 290 291 static int bridge_rtnode_addr_cmp(const uint8_t *, const uint8_t *); 292 static struct bridge_rtnode *bridge_rtnode_lookup(struct bridge_softc *, 293 const uint8_t *, uint16_t); 294 static int bridge_rtnode_insert(struct bridge_softc *, 295 struct bridge_rtnode *); 296 static void bridge_rtnode_destroy(struct bridge_softc *, 297 struct bridge_rtnode *); 298 static void bridge_rtable_expire(struct ifnet *, int); 299 static void bridge_state_change(struct ifnet *, int); 300 301 static struct bridge_iflist *bridge_lookup_member(struct bridge_softc *, 302 const char *name); 303 static struct bridge_iflist *bridge_lookup_member_if(struct bridge_softc *, 304 struct ifnet *ifp); 305 static void bridge_delete_member(struct bridge_softc *, 306 struct bridge_iflist *, int); 307 static void bridge_delete_span(struct bridge_softc *, 308 struct bridge_iflist *); 309 310 static int bridge_ioctl_add(struct bridge_softc *, void *); 311 static int bridge_ioctl_del(struct bridge_softc *, void *); 312 static int bridge_ioctl_gifflags(struct bridge_softc *, void *); 313 static int bridge_ioctl_sifflags(struct bridge_softc *, void *); 314 static int bridge_ioctl_scache(struct bridge_softc *, void *); 315 static int bridge_ioctl_gcache(struct bridge_softc *, void *); 316 static int bridge_ioctl_gifs(struct bridge_softc *, void *); 317 static int bridge_ioctl_rts(struct bridge_softc *, void *); 318 static int bridge_ioctl_saddr(struct bridge_softc *, void *); 319 static int bridge_ioctl_sto(struct bridge_softc *, void *); 320 static int bridge_ioctl_gto(struct bridge_softc *, void *); 321 static int bridge_ioctl_daddr(struct bridge_softc *, void *); 322 static int bridge_ioctl_flush(struct bridge_softc *, void *); 323 static int bridge_ioctl_gpri(struct bridge_softc *, void *); 324 static int bridge_ioctl_spri(struct bridge_softc *, void *); 325 static int bridge_ioctl_ght(struct bridge_softc *, void *); 326 static int bridge_ioctl_sht(struct bridge_softc *, void *); 327 static int bridge_ioctl_gfd(struct bridge_softc *, void *); 328 static int bridge_ioctl_sfd(struct bridge_softc *, void *); 329 static int bridge_ioctl_gma(struct bridge_softc *, void *); 330 static int bridge_ioctl_sma(struct bridge_softc *, void *); 331 static int bridge_ioctl_sifprio(struct bridge_softc *, void *); 332 static int bridge_ioctl_sifcost(struct bridge_softc *, void *); 333 static int bridge_ioctl_sifmaxaddr(struct bridge_softc *, void *); 334 static int bridge_ioctl_addspan(struct bridge_softc *, void *); 335 static int bridge_ioctl_delspan(struct bridge_softc *, void *); 336 static int bridge_ioctl_gbparam(struct bridge_softc *, void *); 337 static int bridge_ioctl_grte(struct bridge_softc *, void *); 338 static int bridge_ioctl_gifsstp(struct bridge_softc *, void *); 339 static int bridge_ioctl_sproto(struct bridge_softc *, void *); 340 static int bridge_ioctl_stxhc(struct bridge_softc *, void *); 341 static int bridge_pfil(struct mbuf **, struct ifnet *, struct ifnet *, 342 int); 343 static int bridge_ip_checkbasic(struct mbuf **mp); 344 #ifdef INET6 345 static int bridge_ip6_checkbasic(struct mbuf **mp); 346 #endif /* INET6 */ 347 static int bridge_fragment(struct ifnet *, struct mbuf **mp, 348 struct ether_header *, int, struct llc *); 349 static void bridge_linkstate(struct ifnet *ifp); 350 static void bridge_linkcheck(struct bridge_softc *sc); 351 352 353 /* The default bridge vlan is 1 (IEEE 802.1Q-2003 Table 9-2) */ 354 #define VLANTAGOF(_m) \ 355 (_m->m_flags & M_VLANTAG) ? EVL_VLANOFTAG(_m->m_pkthdr.ether_vtag) : 1 356 357 static struct bstp_cb_ops bridge_ops = { 358 .bcb_state = bridge_state_change, 359 .bcb_rtage = bridge_rtable_expire 360 }; 361 362 SYSCTL_DECL(_net_link); 363 static SYSCTL_NODE(_net_link, IFT_BRIDGE, bridge, CTLFLAG_RW, 0, "Bridge"); 364 365 /* only pass IP[46] packets when pfil is enabled */ 366 VNET_DEFINE_STATIC(int, pfil_onlyip) = 1; 367 #define V_pfil_onlyip VNET(pfil_onlyip) 368 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_onlyip, 369 CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_onlyip), 0, 370 "Only pass IP packets when pfil is enabled"); 371 372 /* run pfil hooks on the bridge interface */ 373 VNET_DEFINE_STATIC(int, pfil_bridge) = 1; 374 #define V_pfil_bridge VNET(pfil_bridge) 375 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_bridge, 376 CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_bridge), 0, 377 "Packet filter on the bridge interface"); 378 379 /* layer2 filter with ipfw */ 380 VNET_DEFINE_STATIC(int, pfil_ipfw); 381 #define V_pfil_ipfw VNET(pfil_ipfw) 382 383 /* layer2 ARP filter with ipfw */ 384 VNET_DEFINE_STATIC(int, pfil_ipfw_arp); 385 #define V_pfil_ipfw_arp VNET(pfil_ipfw_arp) 386 SYSCTL_INT(_net_link_bridge, OID_AUTO, ipfw_arp, 387 CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_ipfw_arp), 0, 388 "Filter ARP packets through IPFW layer2"); 389 390 /* run pfil hooks on the member interface */ 391 VNET_DEFINE_STATIC(int, pfil_member) = 1; 392 #define V_pfil_member VNET(pfil_member) 393 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_member, 394 CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_member), 0, 395 "Packet filter on the member interface"); 396 397 /* run pfil hooks on the physical interface for locally destined packets */ 398 VNET_DEFINE_STATIC(int, pfil_local_phys); 399 #define V_pfil_local_phys VNET(pfil_local_phys) 400 SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_local_phys, 401 CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_local_phys), 0, 402 "Packet filter on the physical interface for locally destined packets"); 403 404 /* log STP state changes */ 405 VNET_DEFINE_STATIC(int, log_stp); 406 #define V_log_stp VNET(log_stp) 407 SYSCTL_INT(_net_link_bridge, OID_AUTO, log_stp, 408 CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(log_stp), 0, 409 "Log STP state changes"); 410 411 /* share MAC with first bridge member */ 412 VNET_DEFINE_STATIC(int, bridge_inherit_mac); 413 #define V_bridge_inherit_mac VNET(bridge_inherit_mac) 414 SYSCTL_INT(_net_link_bridge, OID_AUTO, inherit_mac, 415 CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(bridge_inherit_mac), 0, 416 "Inherit MAC address from the first bridge member"); 417 418 VNET_DEFINE_STATIC(int, allow_llz_overlap) = 0; 419 #define V_allow_llz_overlap VNET(allow_llz_overlap) 420 SYSCTL_INT(_net_link_bridge, OID_AUTO, allow_llz_overlap, 421 CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(allow_llz_overlap), 0, 422 "Allow overlap of link-local scope " 423 "zones of a bridge interface and the member interfaces"); 424 425 struct bridge_control { 426 int (*bc_func)(struct bridge_softc *, void *); 427 int bc_argsize; 428 int bc_flags; 429 }; 430 431 #define BC_F_COPYIN 0x01 /* copy arguments in */ 432 #define BC_F_COPYOUT 0x02 /* copy arguments out */ 433 #define BC_F_SUSER 0x04 /* do super-user check */ 434 435 const struct bridge_control bridge_control_table[] = { 436 { bridge_ioctl_add, sizeof(struct ifbreq), 437 BC_F_COPYIN|BC_F_SUSER }, 438 { bridge_ioctl_del, sizeof(struct ifbreq), 439 BC_F_COPYIN|BC_F_SUSER }, 440 441 { bridge_ioctl_gifflags, sizeof(struct ifbreq), 442 BC_F_COPYIN|BC_F_COPYOUT }, 443 { bridge_ioctl_sifflags, sizeof(struct ifbreq), 444 BC_F_COPYIN|BC_F_SUSER }, 445 446 { bridge_ioctl_scache, sizeof(struct ifbrparam), 447 BC_F_COPYIN|BC_F_SUSER }, 448 { bridge_ioctl_gcache, sizeof(struct ifbrparam), 449 BC_F_COPYOUT }, 450 451 { bridge_ioctl_gifs, sizeof(struct ifbifconf), 452 BC_F_COPYIN|BC_F_COPYOUT }, 453 { bridge_ioctl_rts, sizeof(struct ifbaconf), 454 BC_F_COPYIN|BC_F_COPYOUT }, 455 456 { bridge_ioctl_saddr, sizeof(struct ifbareq), 457 BC_F_COPYIN|BC_F_SUSER }, 458 459 { bridge_ioctl_sto, sizeof(struct ifbrparam), 460 BC_F_COPYIN|BC_F_SUSER }, 461 { bridge_ioctl_gto, sizeof(struct ifbrparam), 462 BC_F_COPYOUT }, 463 464 { bridge_ioctl_daddr, sizeof(struct ifbareq), 465 BC_F_COPYIN|BC_F_SUSER }, 466 467 { bridge_ioctl_flush, sizeof(struct ifbreq), 468 BC_F_COPYIN|BC_F_SUSER }, 469 470 { bridge_ioctl_gpri, sizeof(struct ifbrparam), 471 BC_F_COPYOUT }, 472 { bridge_ioctl_spri, sizeof(struct ifbrparam), 473 BC_F_COPYIN|BC_F_SUSER }, 474 475 { bridge_ioctl_ght, sizeof(struct ifbrparam), 476 BC_F_COPYOUT }, 477 { bridge_ioctl_sht, sizeof(struct ifbrparam), 478 BC_F_COPYIN|BC_F_SUSER }, 479 480 { bridge_ioctl_gfd, sizeof(struct ifbrparam), 481 BC_F_COPYOUT }, 482 { bridge_ioctl_sfd, sizeof(struct ifbrparam), 483 BC_F_COPYIN|BC_F_SUSER }, 484 485 { bridge_ioctl_gma, sizeof(struct ifbrparam), 486 BC_F_COPYOUT }, 487 { bridge_ioctl_sma, sizeof(struct ifbrparam), 488 BC_F_COPYIN|BC_F_SUSER }, 489 490 { bridge_ioctl_sifprio, sizeof(struct ifbreq), 491 BC_F_COPYIN|BC_F_SUSER }, 492 493 { bridge_ioctl_sifcost, sizeof(struct ifbreq), 494 BC_F_COPYIN|BC_F_SUSER }, 495 496 { bridge_ioctl_addspan, sizeof(struct ifbreq), 497 BC_F_COPYIN|BC_F_SUSER }, 498 { bridge_ioctl_delspan, sizeof(struct ifbreq), 499 BC_F_COPYIN|BC_F_SUSER }, 500 501 { bridge_ioctl_gbparam, sizeof(struct ifbropreq), 502 BC_F_COPYOUT }, 503 504 { bridge_ioctl_grte, sizeof(struct ifbrparam), 505 BC_F_COPYOUT }, 506 507 { bridge_ioctl_gifsstp, sizeof(struct ifbpstpconf), 508 BC_F_COPYIN|BC_F_COPYOUT }, 509 510 { bridge_ioctl_sproto, sizeof(struct ifbrparam), 511 BC_F_COPYIN|BC_F_SUSER }, 512 513 { bridge_ioctl_stxhc, sizeof(struct ifbrparam), 514 BC_F_COPYIN|BC_F_SUSER }, 515 516 { bridge_ioctl_sifmaxaddr, sizeof(struct ifbreq), 517 BC_F_COPYIN|BC_F_SUSER }, 518 519 }; 520 const int bridge_control_table_size = nitems(bridge_control_table); 521 522 VNET_DEFINE_STATIC(LIST_HEAD(, bridge_softc), bridge_list); 523 #define V_bridge_list VNET(bridge_list) 524 #define BRIDGE_LIST_LOCK_INIT(x) mtx_init(&V_bridge_list_mtx, \ 525 "if_bridge list", NULL, MTX_DEF) 526 #define BRIDGE_LIST_LOCK_DESTROY(x) mtx_destroy(&V_bridge_list_mtx) 527 #define BRIDGE_LIST_LOCK(x) mtx_lock(&V_bridge_list_mtx) 528 #define BRIDGE_LIST_UNLOCK(x) mtx_unlock(&V_bridge_list_mtx) 529 530 VNET_DEFINE_STATIC(struct if_clone *, bridge_cloner); 531 #define V_bridge_cloner VNET(bridge_cloner) 532 533 static const char bridge_name[] = "bridge"; 534 535 static void 536 vnet_bridge_init(const void *unused __unused) 537 { 538 539 V_bridge_rtnode_zone = uma_zcreate("bridge_rtnode", 540 sizeof(struct bridge_rtnode), NULL, NULL, NULL, NULL, 541 UMA_ALIGN_PTR, 0); 542 BRIDGE_LIST_LOCK_INIT(); 543 LIST_INIT(&V_bridge_list); 544 V_bridge_cloner = if_clone_simple(bridge_name, 545 bridge_clone_create, bridge_clone_destroy, 0); 546 } 547 VNET_SYSINIT(vnet_bridge_init, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY, 548 vnet_bridge_init, NULL); 549 550 static void 551 vnet_bridge_uninit(const void *unused __unused) 552 { 553 554 if_clone_detach(V_bridge_cloner); 555 V_bridge_cloner = NULL; 556 BRIDGE_LIST_LOCK_DESTROY(); 557 uma_zdestroy(V_bridge_rtnode_zone); 558 } 559 VNET_SYSUNINIT(vnet_bridge_uninit, SI_SUB_PSEUDO, SI_ORDER_ANY, 560 vnet_bridge_uninit, NULL); 561 562 static int 563 bridge_modevent(module_t mod, int type, void *data) 564 { 565 566 switch (type) { 567 case MOD_LOAD: 568 bridge_dn_p = bridge_dummynet; 569 bridge_detach_cookie = EVENTHANDLER_REGISTER( 570 ifnet_departure_event, bridge_ifdetach, NULL, 571 EVENTHANDLER_PRI_ANY); 572 break; 573 case MOD_UNLOAD: 574 EVENTHANDLER_DEREGISTER(ifnet_departure_event, 575 bridge_detach_cookie); 576 bridge_dn_p = NULL; 577 break; 578 default: 579 return (EOPNOTSUPP); 580 } 581 return (0); 582 } 583 584 static moduledata_t bridge_mod = { 585 "if_bridge", 586 bridge_modevent, 587 0 588 }; 589 590 DECLARE_MODULE(if_bridge, bridge_mod, SI_SUB_PSEUDO, SI_ORDER_ANY); 591 MODULE_VERSION(if_bridge, 1); 592 MODULE_DEPEND(if_bridge, bridgestp, 1, 1, 1); 593 594 /* 595 * handler for net.link.bridge.ipfw 596 */ 597 static int 598 sysctl_pfil_ipfw(SYSCTL_HANDLER_ARGS) 599 { 600 int enable = V_pfil_ipfw; 601 int error; 602 603 error = sysctl_handle_int(oidp, &enable, 0, req); 604 enable &= 1; 605 606 if (enable != V_pfil_ipfw) { 607 V_pfil_ipfw = enable; 608 609 /* 610 * Disable pfil so that ipfw doesnt run twice, if the user 611 * really wants both then they can re-enable pfil_bridge and/or 612 * pfil_member. Also allow non-ip packets as ipfw can filter by 613 * layer2 type. 614 */ 615 if (V_pfil_ipfw) { 616 V_pfil_onlyip = 0; 617 V_pfil_bridge = 0; 618 V_pfil_member = 0; 619 } 620 } 621 622 return (error); 623 } 624 SYSCTL_PROC(_net_link_bridge, OID_AUTO, ipfw, 625 CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_VNET, 626 &VNET_NAME(pfil_ipfw), 0, &sysctl_pfil_ipfw, "I", 627 "Layer2 filter with IPFW"); 628 629 /* 630 * bridge_clone_create: 631 * 632 * Create a new bridge instance. 633 */ 634 static int 635 bridge_clone_create(struct if_clone *ifc, int unit, caddr_t params) 636 { 637 struct bridge_softc *sc, *sc2; 638 struct ifnet *bifp, *ifp; 639 int fb, retry; 640 unsigned long hostid; 641 642 sc = malloc(sizeof(*sc), M_DEVBUF, M_WAITOK|M_ZERO); 643 ifp = sc->sc_ifp = if_alloc(IFT_ETHER); 644 if (ifp == NULL) { 645 free(sc, M_DEVBUF); 646 return (ENOSPC); 647 } 648 649 BRIDGE_LOCK_INIT(sc); 650 sc->sc_brtmax = BRIDGE_RTABLE_MAX; 651 sc->sc_brttimeout = BRIDGE_RTABLE_TIMEOUT; 652 653 /* Initialize our routing table. */ 654 bridge_rtable_init(sc); 655 656 callout_init_mtx(&sc->sc_brcallout, &sc->sc_mtx, 0); 657 658 LIST_INIT(&sc->sc_iflist); 659 LIST_INIT(&sc->sc_spanlist); 660 661 ifp->if_softc = sc; 662 if_initname(ifp, bridge_name, unit); 663 ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST; 664 ifp->if_ioctl = bridge_ioctl; 665 ifp->if_transmit = bridge_transmit; 666 ifp->if_qflush = bridge_qflush; 667 ifp->if_init = bridge_init; 668 ifp->if_type = IFT_BRIDGE; 669 670 /* 671 * Generate an ethernet address with a locally administered address. 672 * 673 * Since we are using random ethernet addresses for the bridge, it is 674 * possible that we might have address collisions, so make sure that 675 * this hardware address isn't already in use on another bridge. 676 * The first try uses the hostid and falls back to arc4rand(). 677 */ 678 fb = 0; 679 getcredhostid(curthread->td_ucred, &hostid); 680 do { 681 if (fb || hostid == 0) { 682 ether_gen_addr(ifp, &sc->sc_defaddr); 683 } else { 684 sc->sc_defaddr.octet[0] = 0x2; 685 sc->sc_defaddr.octet[1] = (hostid >> 24) & 0xff; 686 sc->sc_defaddr.octet[2] = (hostid >> 16) & 0xff; 687 sc->sc_defaddr.octet[3] = (hostid >> 8 ) & 0xff; 688 sc->sc_defaddr.octet[4] = hostid & 0xff; 689 sc->sc_defaddr.octet[5] = ifp->if_dunit & 0xff; 690 } 691 692 fb = 1; 693 retry = 0; 694 BRIDGE_LIST_LOCK(); 695 LIST_FOREACH(sc2, &V_bridge_list, sc_list) { 696 bifp = sc2->sc_ifp; 697 if (memcmp(sc->sc_defaddr.octet, 698 IF_LLADDR(bifp), ETHER_ADDR_LEN) == 0) { 699 retry = 1; 700 break; 701 } 702 } 703 BRIDGE_LIST_UNLOCK(); 704 } while (retry == 1); 705 706 bstp_attach(&sc->sc_stp, &bridge_ops); 707 ether_ifattach(ifp, sc->sc_defaddr.octet); 708 /* Now undo some of the damage... */ 709 ifp->if_baudrate = 0; 710 ifp->if_type = IFT_BRIDGE; 711 712 BRIDGE_LIST_LOCK(); 713 LIST_INSERT_HEAD(&V_bridge_list, sc, sc_list); 714 BRIDGE_LIST_UNLOCK(); 715 716 return (0); 717 } 718 719 /* 720 * bridge_clone_destroy: 721 * 722 * Destroy a bridge instance. 723 */ 724 static void 725 bridge_clone_destroy(struct ifnet *ifp) 726 { 727 struct bridge_softc *sc = ifp->if_softc; 728 struct bridge_iflist *bif; 729 730 BRIDGE_LOCK(sc); 731 732 bridge_stop(ifp, 1); 733 ifp->if_flags &= ~IFF_UP; 734 735 while ((bif = LIST_FIRST(&sc->sc_iflist)) != NULL) 736 bridge_delete_member(sc, bif, 0); 737 738 while ((bif = LIST_FIRST(&sc->sc_spanlist)) != NULL) { 739 bridge_delete_span(sc, bif); 740 } 741 742 /* Tear down the routing table. */ 743 bridge_rtable_fini(sc); 744 745 BRIDGE_UNLOCK(sc); 746 747 callout_drain(&sc->sc_brcallout); 748 749 BRIDGE_LIST_LOCK(); 750 LIST_REMOVE(sc, sc_list); 751 BRIDGE_LIST_UNLOCK(); 752 753 bstp_detach(&sc->sc_stp); 754 ether_ifdetach(ifp); 755 if_free(ifp); 756 757 BRIDGE_LOCK_DESTROY(sc); 758 free(sc, M_DEVBUF); 759 } 760 761 /* 762 * bridge_ioctl: 763 * 764 * Handle a control request from the operator. 765 */ 766 static int 767 bridge_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data) 768 { 769 struct bridge_softc *sc = ifp->if_softc; 770 struct ifreq *ifr = (struct ifreq *)data; 771 struct bridge_iflist *bif; 772 struct thread *td = curthread; 773 union { 774 struct ifbreq ifbreq; 775 struct ifbifconf ifbifconf; 776 struct ifbareq ifbareq; 777 struct ifbaconf ifbaconf; 778 struct ifbrparam ifbrparam; 779 struct ifbropreq ifbropreq; 780 } args; 781 struct ifdrv *ifd = (struct ifdrv *) data; 782 const struct bridge_control *bc; 783 int error = 0, oldmtu; 784 785 switch (cmd) { 786 787 case SIOCADDMULTI: 788 case SIOCDELMULTI: 789 break; 790 791 case SIOCGDRVSPEC: 792 case SIOCSDRVSPEC: 793 if (ifd->ifd_cmd >= bridge_control_table_size) { 794 error = EINVAL; 795 break; 796 } 797 bc = &bridge_control_table[ifd->ifd_cmd]; 798 799 if (cmd == SIOCGDRVSPEC && 800 (bc->bc_flags & BC_F_COPYOUT) == 0) { 801 error = EINVAL; 802 break; 803 } 804 else if (cmd == SIOCSDRVSPEC && 805 (bc->bc_flags & BC_F_COPYOUT) != 0) { 806 error = EINVAL; 807 break; 808 } 809 810 if (bc->bc_flags & BC_F_SUSER) { 811 error = priv_check(td, PRIV_NET_BRIDGE); 812 if (error) 813 break; 814 } 815 816 if (ifd->ifd_len != bc->bc_argsize || 817 ifd->ifd_len > sizeof(args)) { 818 error = EINVAL; 819 break; 820 } 821 822 bzero(&args, sizeof(args)); 823 if (bc->bc_flags & BC_F_COPYIN) { 824 error = copyin(ifd->ifd_data, &args, ifd->ifd_len); 825 if (error) 826 break; 827 } 828 829 oldmtu = ifp->if_mtu; 830 BRIDGE_LOCK(sc); 831 error = (*bc->bc_func)(sc, &args); 832 BRIDGE_UNLOCK(sc); 833 if (error) 834 break; 835 836 /* 837 * Bridge MTU may change during addition of the first port. 838 * If it did, do network layer specific procedure. 839 */ 840 if (ifp->if_mtu != oldmtu) { 841 #ifdef INET6 842 nd6_setmtu(ifp); 843 #endif 844 rt_updatemtu(ifp); 845 } 846 847 if (bc->bc_flags & BC_F_COPYOUT) 848 error = copyout(&args, ifd->ifd_data, ifd->ifd_len); 849 850 break; 851 852 case SIOCSIFFLAGS: 853 if (!(ifp->if_flags & IFF_UP) && 854 (ifp->if_drv_flags & IFF_DRV_RUNNING)) { 855 /* 856 * If interface is marked down and it is running, 857 * then stop and disable it. 858 */ 859 BRIDGE_LOCK(sc); 860 bridge_stop(ifp, 1); 861 BRIDGE_UNLOCK(sc); 862 } else if ((ifp->if_flags & IFF_UP) && 863 !(ifp->if_drv_flags & IFF_DRV_RUNNING)) { 864 /* 865 * If interface is marked up and it is stopped, then 866 * start it. 867 */ 868 (*ifp->if_init)(sc); 869 } 870 break; 871 872 case SIOCSIFMTU: 873 if (ifr->ifr_mtu < 576) { 874 error = EINVAL; 875 break; 876 } 877 if (LIST_EMPTY(&sc->sc_iflist)) { 878 sc->sc_ifp->if_mtu = ifr->ifr_mtu; 879 break; 880 } 881 BRIDGE_LOCK(sc); 882 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 883 if (bif->bif_ifp->if_mtu != ifr->ifr_mtu) { 884 log(LOG_NOTICE, "%s: invalid MTU: %u(%s)" 885 " != %d\n", sc->sc_ifp->if_xname, 886 bif->bif_ifp->if_mtu, 887 bif->bif_ifp->if_xname, ifr->ifr_mtu); 888 error = EINVAL; 889 break; 890 } 891 } 892 if (!error) 893 sc->sc_ifp->if_mtu = ifr->ifr_mtu; 894 BRIDGE_UNLOCK(sc); 895 break; 896 default: 897 /* 898 * drop the lock as ether_ioctl() will call bridge_start() and 899 * cause the lock to be recursed. 900 */ 901 error = ether_ioctl(ifp, cmd, data); 902 break; 903 } 904 905 return (error); 906 } 907 908 /* 909 * bridge_mutecaps: 910 * 911 * Clear or restore unwanted capabilities on the member interface 912 */ 913 static void 914 bridge_mutecaps(struct bridge_softc *sc) 915 { 916 struct bridge_iflist *bif; 917 int enabled, mask; 918 919 /* Initial bitmask of capabilities to test */ 920 mask = BRIDGE_IFCAPS_MASK; 921 922 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 923 /* Every member must support it or its disabled */ 924 mask &= bif->bif_savedcaps; 925 } 926 927 BRIDGE_XLOCK(sc); 928 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 929 enabled = bif->bif_ifp->if_capenable; 930 enabled &= ~BRIDGE_IFCAPS_STRIP; 931 /* strip off mask bits and enable them again if allowed */ 932 enabled &= ~BRIDGE_IFCAPS_MASK; 933 enabled |= mask; 934 BRIDGE_UNLOCK(sc); 935 bridge_set_ifcap(sc, bif, enabled); 936 BRIDGE_LOCK(sc); 937 } 938 BRIDGE_XDROP(sc); 939 940 } 941 942 static void 943 bridge_set_ifcap(struct bridge_softc *sc, struct bridge_iflist *bif, int set) 944 { 945 struct ifnet *ifp = bif->bif_ifp; 946 struct ifreq ifr; 947 int error, mask, stuck; 948 949 BRIDGE_UNLOCK_ASSERT(sc); 950 951 bzero(&ifr, sizeof(ifr)); 952 ifr.ifr_reqcap = set; 953 954 if (ifp->if_capenable != set) { 955 error = (*ifp->if_ioctl)(ifp, SIOCSIFCAP, (caddr_t)&ifr); 956 if (error) 957 if_printf(sc->sc_ifp, 958 "error setting capabilities on %s: %d\n", 959 ifp->if_xname, error); 960 mask = BRIDGE_IFCAPS_MASK | BRIDGE_IFCAPS_STRIP; 961 stuck = ifp->if_capenable & mask & ~set; 962 if (stuck != 0) 963 if_printf(sc->sc_ifp, 964 "can't disable some capabilities on %s: 0x%x\n", 965 ifp->if_xname, stuck); 966 } 967 } 968 969 /* 970 * bridge_lookup_member: 971 * 972 * Lookup a bridge member interface. 973 */ 974 static struct bridge_iflist * 975 bridge_lookup_member(struct bridge_softc *sc, const char *name) 976 { 977 struct bridge_iflist *bif; 978 struct ifnet *ifp; 979 980 BRIDGE_LOCK_ASSERT(sc); 981 982 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 983 ifp = bif->bif_ifp; 984 if (strcmp(ifp->if_xname, name) == 0) 985 return (bif); 986 } 987 988 return (NULL); 989 } 990 991 /* 992 * bridge_lookup_member_if: 993 * 994 * Lookup a bridge member interface by ifnet*. 995 */ 996 static struct bridge_iflist * 997 bridge_lookup_member_if(struct bridge_softc *sc, struct ifnet *member_ifp) 998 { 999 struct bridge_iflist *bif; 1000 1001 BRIDGE_LOCK_ASSERT(sc); 1002 1003 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 1004 if (bif->bif_ifp == member_ifp) 1005 return (bif); 1006 } 1007 1008 return (NULL); 1009 } 1010 1011 /* 1012 * bridge_delete_member: 1013 * 1014 * Delete the specified member interface. 1015 */ 1016 static void 1017 bridge_delete_member(struct bridge_softc *sc, struct bridge_iflist *bif, 1018 int gone) 1019 { 1020 struct ifnet *ifs = bif->bif_ifp; 1021 struct ifnet *fif = NULL; 1022 1023 BRIDGE_LOCK_ASSERT(sc); 1024 1025 if (bif->bif_flags & IFBIF_STP) 1026 bstp_disable(&bif->bif_stp); 1027 1028 ifs->if_bridge = NULL; 1029 BRIDGE_XLOCK(sc); 1030 LIST_REMOVE(bif, bif_next); 1031 BRIDGE_XDROP(sc); 1032 1033 /* 1034 * If removing the interface that gave the bridge its mac address, set 1035 * the mac address of the bridge to the address of the next member, or 1036 * to its default address if no members are left. 1037 */ 1038 if (V_bridge_inherit_mac && sc->sc_ifaddr == ifs) { 1039 if (LIST_EMPTY(&sc->sc_iflist)) { 1040 bcopy(&sc->sc_defaddr, 1041 IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN); 1042 sc->sc_ifaddr = NULL; 1043 } else { 1044 fif = LIST_FIRST(&sc->sc_iflist)->bif_ifp; 1045 bcopy(IF_LLADDR(fif), 1046 IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN); 1047 sc->sc_ifaddr = fif; 1048 } 1049 EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp); 1050 } 1051 1052 bridge_linkcheck(sc); 1053 bridge_mutecaps(sc); /* recalcuate now this interface is removed */ 1054 bridge_rtdelete(sc, ifs, IFBF_FLUSHALL); 1055 KASSERT(bif->bif_addrcnt == 0, 1056 ("%s: %d bridge routes referenced", __func__, bif->bif_addrcnt)); 1057 1058 ifs->if_bridge_output = NULL; 1059 ifs->if_bridge_input = NULL; 1060 ifs->if_bridge_linkstate = NULL; 1061 BRIDGE_UNLOCK(sc); 1062 if (!gone) { 1063 switch (ifs->if_type) { 1064 case IFT_ETHER: 1065 case IFT_L2VLAN: 1066 /* 1067 * Take the interface out of promiscuous mode, but only 1068 * if it was promiscuous in the first place. It might 1069 * not be if we're in the bridge_ioctl_add() error path. 1070 */ 1071 if (ifs->if_flags & IFF_PROMISC) 1072 (void) ifpromisc(ifs, 0); 1073 break; 1074 1075 case IFT_GIF: 1076 break; 1077 1078 default: 1079 #ifdef DIAGNOSTIC 1080 panic("bridge_delete_member: impossible"); 1081 #endif 1082 break; 1083 } 1084 /* reneable any interface capabilities */ 1085 bridge_set_ifcap(sc, bif, bif->bif_savedcaps); 1086 } 1087 bstp_destroy(&bif->bif_stp); /* prepare to free */ 1088 BRIDGE_LOCK(sc); 1089 free(bif, M_DEVBUF); 1090 } 1091 1092 /* 1093 * bridge_delete_span: 1094 * 1095 * Delete the specified span interface. 1096 */ 1097 static void 1098 bridge_delete_span(struct bridge_softc *sc, struct bridge_iflist *bif) 1099 { 1100 BRIDGE_LOCK_ASSERT(sc); 1101 1102 KASSERT(bif->bif_ifp->if_bridge == NULL, 1103 ("%s: not a span interface", __func__)); 1104 1105 LIST_REMOVE(bif, bif_next); 1106 free(bif, M_DEVBUF); 1107 } 1108 1109 static int 1110 bridge_ioctl_add(struct bridge_softc *sc, void *arg) 1111 { 1112 struct ifbreq *req = arg; 1113 struct bridge_iflist *bif = NULL; 1114 struct ifnet *ifs; 1115 int error = 0; 1116 1117 ifs = ifunit(req->ifbr_ifsname); 1118 if (ifs == NULL) 1119 return (ENOENT); 1120 if (ifs->if_ioctl == NULL) /* must be supported */ 1121 return (EINVAL); 1122 1123 /* If it's in the span list, it can't be a member. */ 1124 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) 1125 if (ifs == bif->bif_ifp) 1126 return (EBUSY); 1127 1128 if (ifs->if_bridge == sc) 1129 return (EEXIST); 1130 1131 if (ifs->if_bridge != NULL) 1132 return (EBUSY); 1133 1134 switch (ifs->if_type) { 1135 case IFT_ETHER: 1136 case IFT_L2VLAN: 1137 case IFT_GIF: 1138 /* permitted interface types */ 1139 break; 1140 default: 1141 return (EINVAL); 1142 } 1143 1144 #ifdef INET6 1145 /* 1146 * Two valid inet6 addresses with link-local scope must not be 1147 * on the parent interface and the member interfaces at the 1148 * same time. This restriction is needed to prevent violation 1149 * of link-local scope zone. Attempts to add a member 1150 * interface which has inet6 addresses when the parent has 1151 * inet6 triggers removal of all inet6 addresses on the member 1152 * interface. 1153 */ 1154 1155 /* Check if the parent interface has a link-local scope addr. */ 1156 if (V_allow_llz_overlap == 0 && 1157 in6ifa_llaonifp(sc->sc_ifp) != NULL) { 1158 /* 1159 * If any, remove all inet6 addresses from the member 1160 * interfaces. 1161 */ 1162 BRIDGE_XLOCK(sc); 1163 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 1164 if (in6ifa_llaonifp(bif->bif_ifp)) { 1165 BRIDGE_UNLOCK(sc); 1166 in6_ifdetach(bif->bif_ifp); 1167 BRIDGE_LOCK(sc); 1168 if_printf(sc->sc_ifp, 1169 "IPv6 addresses on %s have been removed " 1170 "before adding it as a member to prevent " 1171 "IPv6 address scope violation.\n", 1172 bif->bif_ifp->if_xname); 1173 } 1174 } 1175 BRIDGE_XDROP(sc); 1176 if (in6ifa_llaonifp(ifs)) { 1177 BRIDGE_UNLOCK(sc); 1178 in6_ifdetach(ifs); 1179 BRIDGE_LOCK(sc); 1180 if_printf(sc->sc_ifp, 1181 "IPv6 addresses on %s have been removed " 1182 "before adding it as a member to prevent " 1183 "IPv6 address scope violation.\n", 1184 ifs->if_xname); 1185 } 1186 } 1187 #endif 1188 /* Allow the first Ethernet member to define the MTU */ 1189 if (LIST_EMPTY(&sc->sc_iflist)) 1190 sc->sc_ifp->if_mtu = ifs->if_mtu; 1191 else if (sc->sc_ifp->if_mtu != ifs->if_mtu) { 1192 if_printf(sc->sc_ifp, "invalid MTU: %u(%s) != %u\n", 1193 ifs->if_mtu, ifs->if_xname, sc->sc_ifp->if_mtu); 1194 return (EINVAL); 1195 } 1196 1197 bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO); 1198 if (bif == NULL) 1199 return (ENOMEM); 1200 1201 bif->bif_ifp = ifs; 1202 bif->bif_flags = IFBIF_LEARNING | IFBIF_DISCOVER; 1203 bif->bif_savedcaps = ifs->if_capenable; 1204 1205 /* 1206 * Assign the interface's MAC address to the bridge if it's the first 1207 * member and the MAC address of the bridge has not been changed from 1208 * the default randomly generated one. 1209 */ 1210 if (V_bridge_inherit_mac && LIST_EMPTY(&sc->sc_iflist) && 1211 !memcmp(IF_LLADDR(sc->sc_ifp), sc->sc_defaddr.octet, ETHER_ADDR_LEN)) { 1212 bcopy(IF_LLADDR(ifs), IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN); 1213 sc->sc_ifaddr = ifs; 1214 EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp); 1215 } 1216 1217 ifs->if_bridge = sc; 1218 ifs->if_bridge_output = bridge_output; 1219 ifs->if_bridge_input = bridge_input; 1220 ifs->if_bridge_linkstate = bridge_linkstate; 1221 bstp_create(&sc->sc_stp, &bif->bif_stp, bif->bif_ifp); 1222 /* 1223 * XXX: XLOCK HERE!?! 1224 * 1225 * NOTE: insert_***HEAD*** should be safe for the traversals. 1226 */ 1227 LIST_INSERT_HEAD(&sc->sc_iflist, bif, bif_next); 1228 1229 /* Set interface capabilities to the intersection set of all members */ 1230 bridge_mutecaps(sc); 1231 bridge_linkcheck(sc); 1232 1233 /* Place the interface into promiscuous mode */ 1234 switch (ifs->if_type) { 1235 case IFT_ETHER: 1236 case IFT_L2VLAN: 1237 BRIDGE_UNLOCK(sc); 1238 error = ifpromisc(ifs, 1); 1239 BRIDGE_LOCK(sc); 1240 break; 1241 } 1242 1243 if (error) 1244 bridge_delete_member(sc, bif, 0); 1245 return (error); 1246 } 1247 1248 static int 1249 bridge_ioctl_del(struct bridge_softc *sc, void *arg) 1250 { 1251 struct ifbreq *req = arg; 1252 struct bridge_iflist *bif; 1253 1254 bif = bridge_lookup_member(sc, req->ifbr_ifsname); 1255 if (bif == NULL) 1256 return (ENOENT); 1257 1258 bridge_delete_member(sc, bif, 0); 1259 1260 return (0); 1261 } 1262 1263 static int 1264 bridge_ioctl_gifflags(struct bridge_softc *sc, void *arg) 1265 { 1266 struct ifbreq *req = arg; 1267 struct bridge_iflist *bif; 1268 struct bstp_port *bp; 1269 1270 bif = bridge_lookup_member(sc, req->ifbr_ifsname); 1271 if (bif == NULL) 1272 return (ENOENT); 1273 1274 bp = &bif->bif_stp; 1275 req->ifbr_ifsflags = bif->bif_flags; 1276 req->ifbr_state = bp->bp_state; 1277 req->ifbr_priority = bp->bp_priority; 1278 req->ifbr_path_cost = bp->bp_path_cost; 1279 req->ifbr_portno = bif->bif_ifp->if_index & 0xfff; 1280 req->ifbr_proto = bp->bp_protover; 1281 req->ifbr_role = bp->bp_role; 1282 req->ifbr_stpflags = bp->bp_flags; 1283 req->ifbr_addrcnt = bif->bif_addrcnt; 1284 req->ifbr_addrmax = bif->bif_addrmax; 1285 req->ifbr_addrexceeded = bif->bif_addrexceeded; 1286 1287 /* Copy STP state options as flags */ 1288 if (bp->bp_operedge) 1289 req->ifbr_ifsflags |= IFBIF_BSTP_EDGE; 1290 if (bp->bp_flags & BSTP_PORT_AUTOEDGE) 1291 req->ifbr_ifsflags |= IFBIF_BSTP_AUTOEDGE; 1292 if (bp->bp_ptp_link) 1293 req->ifbr_ifsflags |= IFBIF_BSTP_PTP; 1294 if (bp->bp_flags & BSTP_PORT_AUTOPTP) 1295 req->ifbr_ifsflags |= IFBIF_BSTP_AUTOPTP; 1296 if (bp->bp_flags & BSTP_PORT_ADMEDGE) 1297 req->ifbr_ifsflags |= IFBIF_BSTP_ADMEDGE; 1298 if (bp->bp_flags & BSTP_PORT_ADMCOST) 1299 req->ifbr_ifsflags |= IFBIF_BSTP_ADMCOST; 1300 return (0); 1301 } 1302 1303 static int 1304 bridge_ioctl_sifflags(struct bridge_softc *sc, void *arg) 1305 { 1306 struct ifbreq *req = arg; 1307 struct bridge_iflist *bif; 1308 struct bstp_port *bp; 1309 int error; 1310 1311 bif = bridge_lookup_member(sc, req->ifbr_ifsname); 1312 if (bif == NULL) 1313 return (ENOENT); 1314 bp = &bif->bif_stp; 1315 1316 if (req->ifbr_ifsflags & IFBIF_SPAN) 1317 /* SPAN is readonly */ 1318 return (EINVAL); 1319 1320 if (req->ifbr_ifsflags & IFBIF_STP) { 1321 if ((bif->bif_flags & IFBIF_STP) == 0) { 1322 error = bstp_enable(&bif->bif_stp); 1323 if (error) 1324 return (error); 1325 } 1326 } else { 1327 if ((bif->bif_flags & IFBIF_STP) != 0) 1328 bstp_disable(&bif->bif_stp); 1329 } 1330 1331 /* Pass on STP flags */ 1332 bstp_set_edge(bp, req->ifbr_ifsflags & IFBIF_BSTP_EDGE ? 1 : 0); 1333 bstp_set_autoedge(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOEDGE ? 1 : 0); 1334 bstp_set_ptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_PTP ? 1 : 0); 1335 bstp_set_autoptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOPTP ? 1 : 0); 1336 1337 /* Save the bits relating to the bridge */ 1338 bif->bif_flags = req->ifbr_ifsflags & IFBIFMASK; 1339 1340 return (0); 1341 } 1342 1343 static int 1344 bridge_ioctl_scache(struct bridge_softc *sc, void *arg) 1345 { 1346 struct ifbrparam *param = arg; 1347 1348 sc->sc_brtmax = param->ifbrp_csize; 1349 bridge_rttrim(sc); 1350 1351 return (0); 1352 } 1353 1354 static int 1355 bridge_ioctl_gcache(struct bridge_softc *sc, void *arg) 1356 { 1357 struct ifbrparam *param = arg; 1358 1359 param->ifbrp_csize = sc->sc_brtmax; 1360 1361 return (0); 1362 } 1363 1364 static int 1365 bridge_ioctl_gifs(struct bridge_softc *sc, void *arg) 1366 { 1367 struct ifbifconf *bifc = arg; 1368 struct bridge_iflist *bif; 1369 struct ifbreq breq; 1370 char *buf, *outbuf; 1371 int count, buflen, len, error = 0; 1372 1373 count = 0; 1374 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) 1375 count++; 1376 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) 1377 count++; 1378 1379 buflen = sizeof(breq) * count; 1380 if (bifc->ifbic_len == 0) { 1381 bifc->ifbic_len = buflen; 1382 return (0); 1383 } 1384 BRIDGE_UNLOCK(sc); 1385 outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO); 1386 BRIDGE_LOCK(sc); 1387 1388 count = 0; 1389 buf = outbuf; 1390 len = min(bifc->ifbic_len, buflen); 1391 bzero(&breq, sizeof(breq)); 1392 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 1393 if (len < sizeof(breq)) 1394 break; 1395 1396 strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname, 1397 sizeof(breq.ifbr_ifsname)); 1398 /* Fill in the ifbreq structure */ 1399 error = bridge_ioctl_gifflags(sc, &breq); 1400 if (error) 1401 break; 1402 memcpy(buf, &breq, sizeof(breq)); 1403 count++; 1404 buf += sizeof(breq); 1405 len -= sizeof(breq); 1406 } 1407 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) { 1408 if (len < sizeof(breq)) 1409 break; 1410 1411 strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname, 1412 sizeof(breq.ifbr_ifsname)); 1413 breq.ifbr_ifsflags = bif->bif_flags; 1414 breq.ifbr_portno = bif->bif_ifp->if_index & 0xfff; 1415 memcpy(buf, &breq, sizeof(breq)); 1416 count++; 1417 buf += sizeof(breq); 1418 len -= sizeof(breq); 1419 } 1420 1421 BRIDGE_UNLOCK(sc); 1422 bifc->ifbic_len = sizeof(breq) * count; 1423 error = copyout(outbuf, bifc->ifbic_req, bifc->ifbic_len); 1424 BRIDGE_LOCK(sc); 1425 free(outbuf, M_TEMP); 1426 return (error); 1427 } 1428 1429 static int 1430 bridge_ioctl_rts(struct bridge_softc *sc, void *arg) 1431 { 1432 struct ifbaconf *bac = arg; 1433 struct bridge_rtnode *brt; 1434 struct ifbareq bareq; 1435 char *buf, *outbuf; 1436 int count, buflen, len, error = 0; 1437 1438 if (bac->ifbac_len == 0) 1439 return (0); 1440 1441 count = 0; 1442 LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) 1443 count++; 1444 buflen = sizeof(bareq) * count; 1445 1446 BRIDGE_UNLOCK(sc); 1447 outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO); 1448 BRIDGE_LOCK(sc); 1449 1450 count = 0; 1451 buf = outbuf; 1452 len = min(bac->ifbac_len, buflen); 1453 bzero(&bareq, sizeof(bareq)); 1454 LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) { 1455 if (len < sizeof(bareq)) 1456 goto out; 1457 strlcpy(bareq.ifba_ifsname, brt->brt_ifp->if_xname, 1458 sizeof(bareq.ifba_ifsname)); 1459 memcpy(bareq.ifba_dst, brt->brt_addr, sizeof(brt->brt_addr)); 1460 bareq.ifba_vlan = brt->brt_vlan; 1461 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC && 1462 time_uptime < brt->brt_expire) 1463 bareq.ifba_expire = brt->brt_expire - time_uptime; 1464 else 1465 bareq.ifba_expire = 0; 1466 bareq.ifba_flags = brt->brt_flags; 1467 1468 memcpy(buf, &bareq, sizeof(bareq)); 1469 count++; 1470 buf += sizeof(bareq); 1471 len -= sizeof(bareq); 1472 } 1473 out: 1474 BRIDGE_UNLOCK(sc); 1475 bac->ifbac_len = sizeof(bareq) * count; 1476 error = copyout(outbuf, bac->ifbac_req, bac->ifbac_len); 1477 BRIDGE_LOCK(sc); 1478 free(outbuf, M_TEMP); 1479 return (error); 1480 } 1481 1482 static int 1483 bridge_ioctl_saddr(struct bridge_softc *sc, void *arg) 1484 { 1485 struct ifbareq *req = arg; 1486 struct bridge_iflist *bif; 1487 int error; 1488 1489 bif = bridge_lookup_member(sc, req->ifba_ifsname); 1490 if (bif == NULL) 1491 return (ENOENT); 1492 1493 error = bridge_rtupdate(sc, req->ifba_dst, req->ifba_vlan, bif, 1, 1494 req->ifba_flags); 1495 1496 return (error); 1497 } 1498 1499 static int 1500 bridge_ioctl_sto(struct bridge_softc *sc, void *arg) 1501 { 1502 struct ifbrparam *param = arg; 1503 1504 sc->sc_brttimeout = param->ifbrp_ctime; 1505 return (0); 1506 } 1507 1508 static int 1509 bridge_ioctl_gto(struct bridge_softc *sc, void *arg) 1510 { 1511 struct ifbrparam *param = arg; 1512 1513 param->ifbrp_ctime = sc->sc_brttimeout; 1514 return (0); 1515 } 1516 1517 static int 1518 bridge_ioctl_daddr(struct bridge_softc *sc, void *arg) 1519 { 1520 struct ifbareq *req = arg; 1521 1522 return (bridge_rtdaddr(sc, req->ifba_dst, req->ifba_vlan)); 1523 } 1524 1525 static int 1526 bridge_ioctl_flush(struct bridge_softc *sc, void *arg) 1527 { 1528 struct ifbreq *req = arg; 1529 1530 bridge_rtflush(sc, req->ifbr_ifsflags); 1531 return (0); 1532 } 1533 1534 static int 1535 bridge_ioctl_gpri(struct bridge_softc *sc, void *arg) 1536 { 1537 struct ifbrparam *param = arg; 1538 struct bstp_state *bs = &sc->sc_stp; 1539 1540 param->ifbrp_prio = bs->bs_bridge_priority; 1541 return (0); 1542 } 1543 1544 static int 1545 bridge_ioctl_spri(struct bridge_softc *sc, void *arg) 1546 { 1547 struct ifbrparam *param = arg; 1548 1549 return (bstp_set_priority(&sc->sc_stp, param->ifbrp_prio)); 1550 } 1551 1552 static int 1553 bridge_ioctl_ght(struct bridge_softc *sc, void *arg) 1554 { 1555 struct ifbrparam *param = arg; 1556 struct bstp_state *bs = &sc->sc_stp; 1557 1558 param->ifbrp_hellotime = bs->bs_bridge_htime >> 8; 1559 return (0); 1560 } 1561 1562 static int 1563 bridge_ioctl_sht(struct bridge_softc *sc, void *arg) 1564 { 1565 struct ifbrparam *param = arg; 1566 1567 return (bstp_set_htime(&sc->sc_stp, param->ifbrp_hellotime)); 1568 } 1569 1570 static int 1571 bridge_ioctl_gfd(struct bridge_softc *sc, void *arg) 1572 { 1573 struct ifbrparam *param = arg; 1574 struct bstp_state *bs = &sc->sc_stp; 1575 1576 param->ifbrp_fwddelay = bs->bs_bridge_fdelay >> 8; 1577 return (0); 1578 } 1579 1580 static int 1581 bridge_ioctl_sfd(struct bridge_softc *sc, void *arg) 1582 { 1583 struct ifbrparam *param = arg; 1584 1585 return (bstp_set_fdelay(&sc->sc_stp, param->ifbrp_fwddelay)); 1586 } 1587 1588 static int 1589 bridge_ioctl_gma(struct bridge_softc *sc, void *arg) 1590 { 1591 struct ifbrparam *param = arg; 1592 struct bstp_state *bs = &sc->sc_stp; 1593 1594 param->ifbrp_maxage = bs->bs_bridge_max_age >> 8; 1595 return (0); 1596 } 1597 1598 static int 1599 bridge_ioctl_sma(struct bridge_softc *sc, void *arg) 1600 { 1601 struct ifbrparam *param = arg; 1602 1603 return (bstp_set_maxage(&sc->sc_stp, param->ifbrp_maxage)); 1604 } 1605 1606 static int 1607 bridge_ioctl_sifprio(struct bridge_softc *sc, void *arg) 1608 { 1609 struct ifbreq *req = arg; 1610 struct bridge_iflist *bif; 1611 1612 bif = bridge_lookup_member(sc, req->ifbr_ifsname); 1613 if (bif == NULL) 1614 return (ENOENT); 1615 1616 return (bstp_set_port_priority(&bif->bif_stp, req->ifbr_priority)); 1617 } 1618 1619 static int 1620 bridge_ioctl_sifcost(struct bridge_softc *sc, void *arg) 1621 { 1622 struct ifbreq *req = arg; 1623 struct bridge_iflist *bif; 1624 1625 bif = bridge_lookup_member(sc, req->ifbr_ifsname); 1626 if (bif == NULL) 1627 return (ENOENT); 1628 1629 return (bstp_set_path_cost(&bif->bif_stp, req->ifbr_path_cost)); 1630 } 1631 1632 static int 1633 bridge_ioctl_sifmaxaddr(struct bridge_softc *sc, void *arg) 1634 { 1635 struct ifbreq *req = arg; 1636 struct bridge_iflist *bif; 1637 1638 bif = bridge_lookup_member(sc, req->ifbr_ifsname); 1639 if (bif == NULL) 1640 return (ENOENT); 1641 1642 bif->bif_addrmax = req->ifbr_addrmax; 1643 return (0); 1644 } 1645 1646 static int 1647 bridge_ioctl_addspan(struct bridge_softc *sc, void *arg) 1648 { 1649 struct ifbreq *req = arg; 1650 struct bridge_iflist *bif = NULL; 1651 struct ifnet *ifs; 1652 1653 ifs = ifunit(req->ifbr_ifsname); 1654 if (ifs == NULL) 1655 return (ENOENT); 1656 1657 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) 1658 if (ifs == bif->bif_ifp) 1659 return (EBUSY); 1660 1661 if (ifs->if_bridge != NULL) 1662 return (EBUSY); 1663 1664 switch (ifs->if_type) { 1665 case IFT_ETHER: 1666 case IFT_GIF: 1667 case IFT_L2VLAN: 1668 break; 1669 default: 1670 return (EINVAL); 1671 } 1672 1673 bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO); 1674 if (bif == NULL) 1675 return (ENOMEM); 1676 1677 bif->bif_ifp = ifs; 1678 bif->bif_flags = IFBIF_SPAN; 1679 1680 LIST_INSERT_HEAD(&sc->sc_spanlist, bif, bif_next); 1681 1682 return (0); 1683 } 1684 1685 static int 1686 bridge_ioctl_delspan(struct bridge_softc *sc, void *arg) 1687 { 1688 struct ifbreq *req = arg; 1689 struct bridge_iflist *bif; 1690 struct ifnet *ifs; 1691 1692 ifs = ifunit(req->ifbr_ifsname); 1693 if (ifs == NULL) 1694 return (ENOENT); 1695 1696 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) 1697 if (ifs == bif->bif_ifp) 1698 break; 1699 1700 if (bif == NULL) 1701 return (ENOENT); 1702 1703 bridge_delete_span(sc, bif); 1704 1705 return (0); 1706 } 1707 1708 static int 1709 bridge_ioctl_gbparam(struct bridge_softc *sc, void *arg) 1710 { 1711 struct ifbropreq *req = arg; 1712 struct bstp_state *bs = &sc->sc_stp; 1713 struct bstp_port *root_port; 1714 1715 req->ifbop_maxage = bs->bs_bridge_max_age >> 8; 1716 req->ifbop_hellotime = bs->bs_bridge_htime >> 8; 1717 req->ifbop_fwddelay = bs->bs_bridge_fdelay >> 8; 1718 1719 root_port = bs->bs_root_port; 1720 if (root_port == NULL) 1721 req->ifbop_root_port = 0; 1722 else 1723 req->ifbop_root_port = root_port->bp_ifp->if_index; 1724 1725 req->ifbop_holdcount = bs->bs_txholdcount; 1726 req->ifbop_priority = bs->bs_bridge_priority; 1727 req->ifbop_protocol = bs->bs_protover; 1728 req->ifbop_root_path_cost = bs->bs_root_pv.pv_cost; 1729 req->ifbop_bridgeid = bs->bs_bridge_pv.pv_dbridge_id; 1730 req->ifbop_designated_root = bs->bs_root_pv.pv_root_id; 1731 req->ifbop_designated_bridge = bs->bs_root_pv.pv_dbridge_id; 1732 req->ifbop_last_tc_time.tv_sec = bs->bs_last_tc_time.tv_sec; 1733 req->ifbop_last_tc_time.tv_usec = bs->bs_last_tc_time.tv_usec; 1734 1735 return (0); 1736 } 1737 1738 static int 1739 bridge_ioctl_grte(struct bridge_softc *sc, void *arg) 1740 { 1741 struct ifbrparam *param = arg; 1742 1743 param->ifbrp_cexceeded = sc->sc_brtexceeded; 1744 return (0); 1745 } 1746 1747 static int 1748 bridge_ioctl_gifsstp(struct bridge_softc *sc, void *arg) 1749 { 1750 struct ifbpstpconf *bifstp = arg; 1751 struct bridge_iflist *bif; 1752 struct bstp_port *bp; 1753 struct ifbpstpreq bpreq; 1754 char *buf, *outbuf; 1755 int count, buflen, len, error = 0; 1756 1757 count = 0; 1758 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 1759 if ((bif->bif_flags & IFBIF_STP) != 0) 1760 count++; 1761 } 1762 1763 buflen = sizeof(bpreq) * count; 1764 if (bifstp->ifbpstp_len == 0) { 1765 bifstp->ifbpstp_len = buflen; 1766 return (0); 1767 } 1768 1769 BRIDGE_UNLOCK(sc); 1770 outbuf = malloc(buflen, M_TEMP, M_WAITOK | M_ZERO); 1771 BRIDGE_LOCK(sc); 1772 1773 count = 0; 1774 buf = outbuf; 1775 len = min(bifstp->ifbpstp_len, buflen); 1776 bzero(&bpreq, sizeof(bpreq)); 1777 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 1778 if (len < sizeof(bpreq)) 1779 break; 1780 1781 if ((bif->bif_flags & IFBIF_STP) == 0) 1782 continue; 1783 1784 bp = &bif->bif_stp; 1785 bpreq.ifbp_portno = bif->bif_ifp->if_index & 0xfff; 1786 bpreq.ifbp_fwd_trans = bp->bp_forward_transitions; 1787 bpreq.ifbp_design_cost = bp->bp_desg_pv.pv_cost; 1788 bpreq.ifbp_design_port = bp->bp_desg_pv.pv_port_id; 1789 bpreq.ifbp_design_bridge = bp->bp_desg_pv.pv_dbridge_id; 1790 bpreq.ifbp_design_root = bp->bp_desg_pv.pv_root_id; 1791 1792 memcpy(buf, &bpreq, sizeof(bpreq)); 1793 count++; 1794 buf += sizeof(bpreq); 1795 len -= sizeof(bpreq); 1796 } 1797 1798 BRIDGE_UNLOCK(sc); 1799 bifstp->ifbpstp_len = sizeof(bpreq) * count; 1800 error = copyout(outbuf, bifstp->ifbpstp_req, bifstp->ifbpstp_len); 1801 BRIDGE_LOCK(sc); 1802 free(outbuf, M_TEMP); 1803 return (error); 1804 } 1805 1806 static int 1807 bridge_ioctl_sproto(struct bridge_softc *sc, void *arg) 1808 { 1809 struct ifbrparam *param = arg; 1810 1811 return (bstp_set_protocol(&sc->sc_stp, param->ifbrp_proto)); 1812 } 1813 1814 static int 1815 bridge_ioctl_stxhc(struct bridge_softc *sc, void *arg) 1816 { 1817 struct ifbrparam *param = arg; 1818 1819 return (bstp_set_holdcount(&sc->sc_stp, param->ifbrp_txhc)); 1820 } 1821 1822 /* 1823 * bridge_ifdetach: 1824 * 1825 * Detach an interface from a bridge. Called when a member 1826 * interface is detaching. 1827 */ 1828 static void 1829 bridge_ifdetach(void *arg __unused, struct ifnet *ifp) 1830 { 1831 struct bridge_softc *sc = ifp->if_bridge; 1832 struct bridge_iflist *bif; 1833 1834 if (ifp->if_flags & IFF_RENAMING) 1835 return; 1836 if (V_bridge_cloner == NULL) { 1837 /* 1838 * This detach handler can be called after 1839 * vnet_bridge_uninit(). Just return in that case. 1840 */ 1841 return; 1842 } 1843 /* Check if the interface is a bridge member */ 1844 if (sc != NULL) { 1845 BRIDGE_LOCK(sc); 1846 1847 bif = bridge_lookup_member_if(sc, ifp); 1848 if (bif != NULL) 1849 bridge_delete_member(sc, bif, 1); 1850 1851 BRIDGE_UNLOCK(sc); 1852 return; 1853 } 1854 1855 /* Check if the interface is a span port */ 1856 BRIDGE_LIST_LOCK(); 1857 LIST_FOREACH(sc, &V_bridge_list, sc_list) { 1858 BRIDGE_LOCK(sc); 1859 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) 1860 if (ifp == bif->bif_ifp) { 1861 bridge_delete_span(sc, bif); 1862 break; 1863 } 1864 1865 BRIDGE_UNLOCK(sc); 1866 } 1867 BRIDGE_LIST_UNLOCK(); 1868 } 1869 1870 /* 1871 * bridge_init: 1872 * 1873 * Initialize a bridge interface. 1874 */ 1875 static void 1876 bridge_init(void *xsc) 1877 { 1878 struct bridge_softc *sc = (struct bridge_softc *)xsc; 1879 struct ifnet *ifp = sc->sc_ifp; 1880 1881 if (ifp->if_drv_flags & IFF_DRV_RUNNING) 1882 return; 1883 1884 BRIDGE_LOCK(sc); 1885 callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz, 1886 bridge_timer, sc); 1887 1888 ifp->if_drv_flags |= IFF_DRV_RUNNING; 1889 bstp_init(&sc->sc_stp); /* Initialize Spanning Tree */ 1890 1891 BRIDGE_UNLOCK(sc); 1892 } 1893 1894 /* 1895 * bridge_stop: 1896 * 1897 * Stop the bridge interface. 1898 */ 1899 static void 1900 bridge_stop(struct ifnet *ifp, int disable) 1901 { 1902 struct bridge_softc *sc = ifp->if_softc; 1903 1904 BRIDGE_LOCK_ASSERT(sc); 1905 1906 if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) 1907 return; 1908 1909 callout_stop(&sc->sc_brcallout); 1910 bstp_stop(&sc->sc_stp); 1911 1912 bridge_rtflush(sc, IFBF_FLUSHDYN); 1913 1914 ifp->if_drv_flags &= ~IFF_DRV_RUNNING; 1915 } 1916 1917 /* 1918 * bridge_enqueue: 1919 * 1920 * Enqueue a packet on a bridge member interface. 1921 * 1922 */ 1923 static int 1924 bridge_enqueue(struct bridge_softc *sc, struct ifnet *dst_ifp, struct mbuf *m) 1925 { 1926 int len, err = 0; 1927 short mflags; 1928 struct mbuf *m0; 1929 1930 /* We may be sending a fragment so traverse the mbuf */ 1931 for (; m; m = m0) { 1932 m0 = m->m_nextpkt; 1933 m->m_nextpkt = NULL; 1934 len = m->m_pkthdr.len; 1935 mflags = m->m_flags; 1936 1937 /* 1938 * If underlying interface can not do VLAN tag insertion itself 1939 * then attach a packet tag that holds it. 1940 */ 1941 if ((m->m_flags & M_VLANTAG) && 1942 (dst_ifp->if_capenable & IFCAP_VLAN_HWTAGGING) == 0) { 1943 m = ether_vlanencap(m, m->m_pkthdr.ether_vtag); 1944 if (m == NULL) { 1945 if_printf(dst_ifp, 1946 "unable to prepend VLAN header\n"); 1947 if_inc_counter(dst_ifp, IFCOUNTER_OERRORS, 1); 1948 continue; 1949 } 1950 m->m_flags &= ~M_VLANTAG; 1951 } 1952 1953 M_ASSERTPKTHDR(m); /* We shouldn't transmit mbuf without pkthdr */ 1954 if ((err = dst_ifp->if_transmit(dst_ifp, m))) { 1955 m_freem(m0); 1956 if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1); 1957 break; 1958 } 1959 1960 if_inc_counter(sc->sc_ifp, IFCOUNTER_OPACKETS, 1); 1961 if_inc_counter(sc->sc_ifp, IFCOUNTER_OBYTES, len); 1962 if (mflags & M_MCAST) 1963 if_inc_counter(sc->sc_ifp, IFCOUNTER_OMCASTS, 1); 1964 } 1965 1966 return (err); 1967 } 1968 1969 /* 1970 * bridge_dummynet: 1971 * 1972 * Receive a queued packet from dummynet and pass it on to the output 1973 * interface. 1974 * 1975 * The mbuf has the Ethernet header already attached. 1976 */ 1977 static void 1978 bridge_dummynet(struct mbuf *m, struct ifnet *ifp) 1979 { 1980 struct bridge_softc *sc; 1981 1982 sc = ifp->if_bridge; 1983 1984 /* 1985 * The packet didnt originate from a member interface. This should only 1986 * ever happen if a member interface is removed while packets are 1987 * queued for it. 1988 */ 1989 if (sc == NULL) { 1990 m_freem(m); 1991 return; 1992 } 1993 1994 if (PFIL_HOOKED_OUT(V_inet_pfil_head) 1995 #ifdef INET6 1996 || PFIL_HOOKED_OUT(V_inet6_pfil_head) 1997 #endif 1998 ) { 1999 if (bridge_pfil(&m, sc->sc_ifp, ifp, PFIL_OUT) != 0) 2000 return; 2001 if (m == NULL) 2002 return; 2003 } 2004 2005 bridge_enqueue(sc, ifp, m); 2006 } 2007 2008 /* 2009 * bridge_output: 2010 * 2011 * Send output from a bridge member interface. This 2012 * performs the bridging function for locally originated 2013 * packets. 2014 * 2015 * The mbuf has the Ethernet header already attached. We must 2016 * enqueue or free the mbuf before returning. 2017 */ 2018 static int 2019 bridge_output(struct ifnet *ifp, struct mbuf *m, struct sockaddr *sa, 2020 struct rtentry *rt) 2021 { 2022 struct ether_header *eh; 2023 struct ifnet *bifp, *dst_if; 2024 struct bridge_softc *sc; 2025 uint16_t vlan; 2026 2027 if (m->m_len < ETHER_HDR_LEN) { 2028 m = m_pullup(m, ETHER_HDR_LEN); 2029 if (m == NULL) 2030 return (0); 2031 } 2032 2033 eh = mtod(m, struct ether_header *); 2034 sc = ifp->if_bridge; 2035 vlan = VLANTAGOF(m); 2036 2037 BRIDGE_LOCK(sc); 2038 bifp = sc->sc_ifp; 2039 2040 /* 2041 * If bridge is down, but the original output interface is up, 2042 * go ahead and send out that interface. Otherwise, the packet 2043 * is dropped below. 2044 */ 2045 if ((bifp->if_drv_flags & IFF_DRV_RUNNING) == 0) { 2046 dst_if = ifp; 2047 goto sendunicast; 2048 } 2049 2050 /* 2051 * If the packet is a multicast, or we don't know a better way to 2052 * get there, send to all interfaces. 2053 */ 2054 if (ETHER_IS_MULTICAST(eh->ether_dhost)) 2055 dst_if = NULL; 2056 else 2057 dst_if = bridge_rtlookup(sc, eh->ether_dhost, vlan); 2058 /* Tap any traffic not passing back out the originating interface */ 2059 if (dst_if != ifp) 2060 ETHER_BPF_MTAP(bifp, m); 2061 if (dst_if == NULL) { 2062 struct bridge_iflist *bif; 2063 struct mbuf *mc; 2064 int error = 0, used = 0; 2065 2066 bridge_span(sc, m); 2067 2068 BRIDGE_LOCK2REF(sc, error); 2069 if (error) { 2070 m_freem(m); 2071 return (0); 2072 } 2073 2074 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 2075 dst_if = bif->bif_ifp; 2076 2077 if (dst_if->if_type == IFT_GIF) 2078 continue; 2079 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) 2080 continue; 2081 2082 /* 2083 * If this is not the original output interface, 2084 * and the interface is participating in spanning 2085 * tree, make sure the port is in a state that 2086 * allows forwarding. 2087 */ 2088 if (dst_if != ifp && (bif->bif_flags & IFBIF_STP) && 2089 bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) 2090 continue; 2091 2092 if (LIST_NEXT(bif, bif_next) == NULL) { 2093 used = 1; 2094 mc = m; 2095 } else { 2096 mc = m_copypacket(m, M_NOWAIT); 2097 if (mc == NULL) { 2098 if_inc_counter(bifp, IFCOUNTER_OERRORS, 1); 2099 continue; 2100 } 2101 } 2102 2103 bridge_enqueue(sc, dst_if, mc); 2104 } 2105 if (used == 0) 2106 m_freem(m); 2107 BRIDGE_UNREF(sc); 2108 return (0); 2109 } 2110 2111 sendunicast: 2112 /* 2113 * XXX Spanning tree consideration here? 2114 */ 2115 2116 bridge_span(sc, m); 2117 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) { 2118 m_freem(m); 2119 BRIDGE_UNLOCK(sc); 2120 return (0); 2121 } 2122 2123 BRIDGE_UNLOCK(sc); 2124 bridge_enqueue(sc, dst_if, m); 2125 return (0); 2126 } 2127 2128 /* 2129 * bridge_transmit: 2130 * 2131 * Do output on a bridge. 2132 * 2133 */ 2134 static int 2135 bridge_transmit(struct ifnet *ifp, struct mbuf *m) 2136 { 2137 struct bridge_softc *sc; 2138 struct ether_header *eh; 2139 struct ifnet *dst_if; 2140 int error = 0; 2141 2142 sc = ifp->if_softc; 2143 2144 ETHER_BPF_MTAP(ifp, m); 2145 2146 eh = mtod(m, struct ether_header *); 2147 2148 BRIDGE_LOCK(sc); 2149 if (((m->m_flags & (M_BCAST|M_MCAST)) == 0) && 2150 (dst_if = bridge_rtlookup(sc, eh->ether_dhost, 1)) != NULL) { 2151 BRIDGE_UNLOCK(sc); 2152 error = bridge_enqueue(sc, dst_if, m); 2153 } else 2154 bridge_broadcast(sc, ifp, m, 0); 2155 2156 return (error); 2157 } 2158 2159 /* 2160 * The ifp->if_qflush entry point for if_bridge(4) is no-op. 2161 */ 2162 static void 2163 bridge_qflush(struct ifnet *ifp __unused) 2164 { 2165 } 2166 2167 /* 2168 * bridge_forward: 2169 * 2170 * The forwarding function of the bridge. 2171 * 2172 * NOTE: Releases the lock on return. 2173 */ 2174 static void 2175 bridge_forward(struct bridge_softc *sc, struct bridge_iflist *sbif, 2176 struct mbuf *m) 2177 { 2178 struct bridge_iflist *dbif; 2179 struct ifnet *src_if, *dst_if, *ifp; 2180 struct ether_header *eh; 2181 uint16_t vlan; 2182 uint8_t *dst; 2183 int error; 2184 2185 src_if = m->m_pkthdr.rcvif; 2186 ifp = sc->sc_ifp; 2187 2188 if_inc_counter(ifp, IFCOUNTER_IPACKETS, 1); 2189 if_inc_counter(ifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); 2190 vlan = VLANTAGOF(m); 2191 2192 if ((sbif->bif_flags & IFBIF_STP) && 2193 sbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) 2194 goto drop; 2195 2196 eh = mtod(m, struct ether_header *); 2197 dst = eh->ether_dhost; 2198 2199 /* If the interface is learning, record the address. */ 2200 if (sbif->bif_flags & IFBIF_LEARNING) { 2201 error = bridge_rtupdate(sc, eh->ether_shost, vlan, 2202 sbif, 0, IFBAF_DYNAMIC); 2203 /* 2204 * If the interface has addresses limits then deny any source 2205 * that is not in the cache. 2206 */ 2207 if (error && sbif->bif_addrmax) 2208 goto drop; 2209 } 2210 2211 if ((sbif->bif_flags & IFBIF_STP) != 0 && 2212 sbif->bif_stp.bp_state == BSTP_IFSTATE_LEARNING) 2213 goto drop; 2214 2215 /* 2216 * At this point, the port either doesn't participate 2217 * in spanning tree or it is in the forwarding state. 2218 */ 2219 2220 /* 2221 * If the packet is unicast, destined for someone on 2222 * "this" side of the bridge, drop it. 2223 */ 2224 if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) { 2225 dst_if = bridge_rtlookup(sc, dst, vlan); 2226 if (src_if == dst_if) 2227 goto drop; 2228 } else { 2229 /* 2230 * Check if its a reserved multicast address, any address 2231 * listed in 802.1D section 7.12.6 may not be forwarded by the 2232 * bridge. 2233 * This is currently 01-80-C2-00-00-00 to 01-80-C2-00-00-0F 2234 */ 2235 if (dst[0] == 0x01 && dst[1] == 0x80 && 2236 dst[2] == 0xc2 && dst[3] == 0x00 && 2237 dst[4] == 0x00 && dst[5] <= 0x0f) 2238 goto drop; 2239 2240 /* ...forward it to all interfaces. */ 2241 if_inc_counter(ifp, IFCOUNTER_IMCASTS, 1); 2242 dst_if = NULL; 2243 } 2244 2245 /* 2246 * If we have a destination interface which is a member of our bridge, 2247 * OR this is a unicast packet, push it through the bpf(4) machinery. 2248 * For broadcast or multicast packets, don't bother because it will 2249 * be reinjected into ether_input. We do this before we pass the packets 2250 * through the pfil(9) framework, as it is possible that pfil(9) will 2251 * drop the packet, or possibly modify it, making it difficult to debug 2252 * firewall issues on the bridge. 2253 */ 2254 if (dst_if != NULL || (m->m_flags & (M_BCAST | M_MCAST)) == 0) 2255 ETHER_BPF_MTAP(ifp, m); 2256 2257 /* run the packet filter */ 2258 if (PFIL_HOOKED_IN(V_inet_pfil_head) 2259 #ifdef INET6 2260 || PFIL_HOOKED_IN(V_inet6_pfil_head) 2261 #endif 2262 ) { 2263 BRIDGE_UNLOCK(sc); 2264 if (bridge_pfil(&m, ifp, src_if, PFIL_IN) != 0) 2265 return; 2266 if (m == NULL) 2267 return; 2268 BRIDGE_LOCK(sc); 2269 } 2270 2271 if (dst_if == NULL) { 2272 bridge_broadcast(sc, src_if, m, 1); 2273 return; 2274 } 2275 2276 /* 2277 * At this point, we're dealing with a unicast frame 2278 * going to a different interface. 2279 */ 2280 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) 2281 goto drop; 2282 2283 dbif = bridge_lookup_member_if(sc, dst_if); 2284 if (dbif == NULL) 2285 /* Not a member of the bridge (anymore?) */ 2286 goto drop; 2287 2288 /* Private segments can not talk to each other */ 2289 if (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE) 2290 goto drop; 2291 2292 if ((dbif->bif_flags & IFBIF_STP) && 2293 dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) 2294 goto drop; 2295 2296 BRIDGE_UNLOCK(sc); 2297 2298 if (PFIL_HOOKED_OUT(V_inet_pfil_head) 2299 #ifdef INET6 2300 || PFIL_HOOKED_OUT(V_inet6_pfil_head) 2301 #endif 2302 ) { 2303 if (bridge_pfil(&m, ifp, dst_if, PFIL_OUT) != 0) 2304 return; 2305 if (m == NULL) 2306 return; 2307 } 2308 2309 bridge_enqueue(sc, dst_if, m); 2310 return; 2311 2312 drop: 2313 BRIDGE_UNLOCK(sc); 2314 m_freem(m); 2315 } 2316 2317 /* 2318 * bridge_input: 2319 * 2320 * Receive input from a member interface. Queue the packet for 2321 * bridging if it is not for us. 2322 */ 2323 static struct mbuf * 2324 bridge_input(struct ifnet *ifp, struct mbuf *m) 2325 { 2326 struct bridge_softc *sc = ifp->if_bridge; 2327 struct bridge_iflist *bif, *bif2; 2328 struct ifnet *bifp; 2329 struct ether_header *eh; 2330 struct mbuf *mc, *mc2; 2331 uint16_t vlan; 2332 int error; 2333 2334 if ((sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) 2335 return (m); 2336 2337 bifp = sc->sc_ifp; 2338 vlan = VLANTAGOF(m); 2339 2340 /* 2341 * Implement support for bridge monitoring. If this flag has been 2342 * set on this interface, discard the packet once we push it through 2343 * the bpf(4) machinery, but before we do, increment the byte and 2344 * packet counters associated with this interface. 2345 */ 2346 if ((bifp->if_flags & IFF_MONITOR) != 0) { 2347 m->m_pkthdr.rcvif = bifp; 2348 ETHER_BPF_MTAP(bifp, m); 2349 if_inc_counter(bifp, IFCOUNTER_IPACKETS, 1); 2350 if_inc_counter(bifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); 2351 m_freem(m); 2352 return (NULL); 2353 } 2354 BRIDGE_LOCK(sc); 2355 bif = bridge_lookup_member_if(sc, ifp); 2356 if (bif == NULL) { 2357 BRIDGE_UNLOCK(sc); 2358 return (m); 2359 } 2360 2361 eh = mtod(m, struct ether_header *); 2362 2363 bridge_span(sc, m); 2364 2365 if (m->m_flags & (M_BCAST|M_MCAST)) { 2366 /* Tap off 802.1D packets; they do not get forwarded. */ 2367 if (memcmp(eh->ether_dhost, bstp_etheraddr, 2368 ETHER_ADDR_LEN) == 0) { 2369 bstp_input(&bif->bif_stp, ifp, m); /* consumes mbuf */ 2370 BRIDGE_UNLOCK(sc); 2371 return (NULL); 2372 } 2373 2374 if ((bif->bif_flags & IFBIF_STP) && 2375 bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) { 2376 BRIDGE_UNLOCK(sc); 2377 return (m); 2378 } 2379 2380 /* 2381 * Make a deep copy of the packet and enqueue the copy 2382 * for bridge processing; return the original packet for 2383 * local processing. 2384 */ 2385 mc = m_dup(m, M_NOWAIT); 2386 if (mc == NULL) { 2387 BRIDGE_UNLOCK(sc); 2388 return (m); 2389 } 2390 2391 /* Perform the bridge forwarding function with the copy. */ 2392 bridge_forward(sc, bif, mc); 2393 2394 /* 2395 * Reinject the mbuf as arriving on the bridge so we have a 2396 * chance at claiming multicast packets. We can not loop back 2397 * here from ether_input as a bridge is never a member of a 2398 * bridge. 2399 */ 2400 KASSERT(bifp->if_bridge == NULL, 2401 ("loop created in bridge_input")); 2402 mc2 = m_dup(m, M_NOWAIT); 2403 if (mc2 != NULL) { 2404 /* Keep the layer3 header aligned */ 2405 int i = min(mc2->m_pkthdr.len, max_protohdr); 2406 mc2 = m_copyup(mc2, i, ETHER_ALIGN); 2407 } 2408 if (mc2 != NULL) { 2409 mc2->m_pkthdr.rcvif = bifp; 2410 (*bifp->if_input)(bifp, mc2); 2411 } 2412 2413 /* Return the original packet for local processing. */ 2414 return (m); 2415 } 2416 2417 if ((bif->bif_flags & IFBIF_STP) && 2418 bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) { 2419 BRIDGE_UNLOCK(sc); 2420 return (m); 2421 } 2422 2423 #if (defined(INET) || defined(INET6)) 2424 # define OR_CARP_CHECK_WE_ARE_DST(iface) \ 2425 || ((iface)->if_carp \ 2426 && (*carp_forus_p)((iface), eh->ether_dhost)) 2427 # define OR_CARP_CHECK_WE_ARE_SRC(iface) \ 2428 || ((iface)->if_carp \ 2429 && (*carp_forus_p)((iface), eh->ether_shost)) 2430 #else 2431 # define OR_CARP_CHECK_WE_ARE_DST(iface) 2432 # define OR_CARP_CHECK_WE_ARE_SRC(iface) 2433 #endif 2434 2435 #ifdef INET6 2436 # define OR_PFIL_HOOKED_INET6 \ 2437 || PFIL_HOOKED_IN(V_inet6_pfil_head) 2438 #else 2439 # define OR_PFIL_HOOKED_INET6 2440 #endif 2441 2442 #define GRAB_OUR_PACKETS(iface) \ 2443 if ((iface)->if_type == IFT_GIF) \ 2444 continue; \ 2445 /* It is destined for us. */ \ 2446 if (memcmp(IF_LLADDR((iface)), eh->ether_dhost, ETHER_ADDR_LEN) == 0 \ 2447 OR_CARP_CHECK_WE_ARE_DST((iface)) \ 2448 ) { \ 2449 if (bif->bif_flags & IFBIF_LEARNING) { \ 2450 error = bridge_rtupdate(sc, eh->ether_shost, \ 2451 vlan, bif, 0, IFBAF_DYNAMIC); \ 2452 if (error && bif->bif_addrmax) { \ 2453 BRIDGE_UNLOCK(sc); \ 2454 m_freem(m); \ 2455 return (NULL); \ 2456 } \ 2457 } \ 2458 m->m_pkthdr.rcvif = iface; \ 2459 if ((iface) == ifp) { \ 2460 /* Skip bridge processing... src == dest */ \ 2461 BRIDGE_UNLOCK(sc); \ 2462 return (m); \ 2463 } \ 2464 /* It's passing over or to the bridge, locally. */ \ 2465 ETHER_BPF_MTAP(bifp, m); \ 2466 if_inc_counter(bifp, IFCOUNTER_IPACKETS, 1); \ 2467 if_inc_counter(bifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); \ 2468 /* Filter on the physical interface. */ \ 2469 if (V_pfil_local_phys && (PFIL_HOOKED_IN(V_inet_pfil_head) \ 2470 OR_PFIL_HOOKED_INET6)) { \ 2471 if (bridge_pfil(&m, NULL, ifp, \ 2472 PFIL_IN) != 0 || m == NULL) { \ 2473 BRIDGE_UNLOCK(sc); \ 2474 return (NULL); \ 2475 } \ 2476 } \ 2477 if ((iface) != bifp) \ 2478 ETHER_BPF_MTAP(iface, m); \ 2479 BRIDGE_UNLOCK(sc); \ 2480 return (m); \ 2481 } \ 2482 \ 2483 /* We just received a packet that we sent out. */ \ 2484 if (memcmp(IF_LLADDR((iface)), eh->ether_shost, ETHER_ADDR_LEN) == 0 \ 2485 OR_CARP_CHECK_WE_ARE_SRC((iface)) \ 2486 ) { \ 2487 BRIDGE_UNLOCK(sc); \ 2488 m_freem(m); \ 2489 return (NULL); \ 2490 } 2491 2492 /* 2493 * Unicast. Make sure it's not for the bridge. 2494 */ 2495 do { GRAB_OUR_PACKETS(bifp) } while (0); 2496 2497 /* 2498 * Give a chance for ifp at first priority. This will help when the 2499 * packet comes through the interface like VLAN's with the same MACs 2500 * on several interfaces from the same bridge. This also will save 2501 * some CPU cycles in case the destination interface and the input 2502 * interface (eq ifp) are the same. 2503 */ 2504 do { GRAB_OUR_PACKETS(ifp) } while (0); 2505 2506 /* Now check the all bridge members. */ 2507 LIST_FOREACH(bif2, &sc->sc_iflist, bif_next) { 2508 GRAB_OUR_PACKETS(bif2->bif_ifp) 2509 } 2510 2511 #undef OR_CARP_CHECK_WE_ARE_DST 2512 #undef OR_CARP_CHECK_WE_ARE_SRC 2513 #undef OR_PFIL_HOOKED_INET6 2514 #undef GRAB_OUR_PACKETS 2515 2516 /* Perform the bridge forwarding function. */ 2517 bridge_forward(sc, bif, m); 2518 2519 return (NULL); 2520 } 2521 2522 /* 2523 * bridge_broadcast: 2524 * 2525 * Send a frame to all interfaces that are members of 2526 * the bridge, except for the one on which the packet 2527 * arrived. 2528 * 2529 * NOTE: Releases the lock on return. 2530 */ 2531 static void 2532 bridge_broadcast(struct bridge_softc *sc, struct ifnet *src_if, 2533 struct mbuf *m, int runfilt) 2534 { 2535 struct bridge_iflist *dbif, *sbif; 2536 struct mbuf *mc; 2537 struct ifnet *dst_if; 2538 int error = 0, used = 0, i; 2539 2540 sbif = bridge_lookup_member_if(sc, src_if); 2541 2542 BRIDGE_LOCK2REF(sc, error); 2543 if (error) { 2544 m_freem(m); 2545 return; 2546 } 2547 2548 /* Filter on the bridge interface before broadcasting */ 2549 if (runfilt && (PFIL_HOOKED_OUT(V_inet_pfil_head) 2550 #ifdef INET6 2551 || PFIL_HOOKED_OUT(V_inet6_pfil_head) 2552 #endif 2553 )) { 2554 if (bridge_pfil(&m, sc->sc_ifp, NULL, PFIL_OUT) != 0) 2555 goto out; 2556 if (m == NULL) 2557 goto out; 2558 } 2559 2560 LIST_FOREACH(dbif, &sc->sc_iflist, bif_next) { 2561 dst_if = dbif->bif_ifp; 2562 if (dst_if == src_if) 2563 continue; 2564 2565 /* Private segments can not talk to each other */ 2566 if (sbif && (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE)) 2567 continue; 2568 2569 if ((dbif->bif_flags & IFBIF_STP) && 2570 dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) 2571 continue; 2572 2573 if ((dbif->bif_flags & IFBIF_DISCOVER) == 0 && 2574 (m->m_flags & (M_BCAST|M_MCAST)) == 0) 2575 continue; 2576 2577 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) 2578 continue; 2579 2580 if (LIST_NEXT(dbif, bif_next) == NULL) { 2581 mc = m; 2582 used = 1; 2583 } else { 2584 mc = m_dup(m, M_NOWAIT); 2585 if (mc == NULL) { 2586 if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1); 2587 continue; 2588 } 2589 } 2590 2591 /* 2592 * Filter on the output interface. Pass a NULL bridge interface 2593 * pointer so we do not redundantly filter on the bridge for 2594 * each interface we broadcast on. 2595 */ 2596 if (runfilt && (PFIL_HOOKED_OUT(V_inet_pfil_head) 2597 #ifdef INET6 2598 || PFIL_HOOKED_OUT(V_inet6_pfil_head) 2599 #endif 2600 )) { 2601 if (used == 0) { 2602 /* Keep the layer3 header aligned */ 2603 i = min(mc->m_pkthdr.len, max_protohdr); 2604 mc = m_copyup(mc, i, ETHER_ALIGN); 2605 if (mc == NULL) { 2606 if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1); 2607 continue; 2608 } 2609 } 2610 if (bridge_pfil(&mc, NULL, dst_if, PFIL_OUT) != 0) 2611 continue; 2612 if (mc == NULL) 2613 continue; 2614 } 2615 2616 bridge_enqueue(sc, dst_if, mc); 2617 } 2618 if (used == 0) 2619 m_freem(m); 2620 2621 out: 2622 BRIDGE_UNREF(sc); 2623 } 2624 2625 /* 2626 * bridge_span: 2627 * 2628 * Duplicate a packet out one or more interfaces that are in span mode, 2629 * the original mbuf is unmodified. 2630 */ 2631 static void 2632 bridge_span(struct bridge_softc *sc, struct mbuf *m) 2633 { 2634 struct bridge_iflist *bif; 2635 struct ifnet *dst_if; 2636 struct mbuf *mc; 2637 2638 if (LIST_EMPTY(&sc->sc_spanlist)) 2639 return; 2640 2641 LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) { 2642 dst_if = bif->bif_ifp; 2643 2644 if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) 2645 continue; 2646 2647 mc = m_copypacket(m, M_NOWAIT); 2648 if (mc == NULL) { 2649 if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1); 2650 continue; 2651 } 2652 2653 bridge_enqueue(sc, dst_if, mc); 2654 } 2655 } 2656 2657 /* 2658 * bridge_rtupdate: 2659 * 2660 * Add a bridge routing entry. 2661 */ 2662 static int 2663 bridge_rtupdate(struct bridge_softc *sc, const uint8_t *dst, uint16_t vlan, 2664 struct bridge_iflist *bif, int setflags, uint8_t flags) 2665 { 2666 struct bridge_rtnode *brt; 2667 int error; 2668 2669 BRIDGE_LOCK_ASSERT(sc); 2670 2671 /* Check the source address is valid and not multicast. */ 2672 if (ETHER_IS_MULTICAST(dst) || 2673 (dst[0] == 0 && dst[1] == 0 && dst[2] == 0 && 2674 dst[3] == 0 && dst[4] == 0 && dst[5] == 0) != 0) 2675 return (EINVAL); 2676 2677 /* 802.1p frames map to vlan 1 */ 2678 if (vlan == 0) 2679 vlan = 1; 2680 2681 /* 2682 * A route for this destination might already exist. If so, 2683 * update it, otherwise create a new one. 2684 */ 2685 if ((brt = bridge_rtnode_lookup(sc, dst, vlan)) == NULL) { 2686 if (sc->sc_brtcnt >= sc->sc_brtmax) { 2687 sc->sc_brtexceeded++; 2688 return (ENOSPC); 2689 } 2690 /* Check per interface address limits (if enabled) */ 2691 if (bif->bif_addrmax && bif->bif_addrcnt >= bif->bif_addrmax) { 2692 bif->bif_addrexceeded++; 2693 return (ENOSPC); 2694 } 2695 2696 /* 2697 * Allocate a new bridge forwarding node, and 2698 * initialize the expiration time and Ethernet 2699 * address. 2700 */ 2701 brt = uma_zalloc(V_bridge_rtnode_zone, M_NOWAIT | M_ZERO); 2702 if (brt == NULL) 2703 return (ENOMEM); 2704 2705 if (bif->bif_flags & IFBIF_STICKY) 2706 brt->brt_flags = IFBAF_STICKY; 2707 else 2708 brt->brt_flags = IFBAF_DYNAMIC; 2709 2710 memcpy(brt->brt_addr, dst, ETHER_ADDR_LEN); 2711 brt->brt_vlan = vlan; 2712 2713 if ((error = bridge_rtnode_insert(sc, brt)) != 0) { 2714 uma_zfree(V_bridge_rtnode_zone, brt); 2715 return (error); 2716 } 2717 brt->brt_dst = bif; 2718 bif->bif_addrcnt++; 2719 } 2720 2721 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC && 2722 brt->brt_dst != bif) { 2723 brt->brt_dst->bif_addrcnt--; 2724 brt->brt_dst = bif; 2725 brt->brt_dst->bif_addrcnt++; 2726 } 2727 2728 if ((flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) 2729 brt->brt_expire = time_uptime + sc->sc_brttimeout; 2730 if (setflags) 2731 brt->brt_flags = flags; 2732 2733 return (0); 2734 } 2735 2736 /* 2737 * bridge_rtlookup: 2738 * 2739 * Lookup the destination interface for an address. 2740 */ 2741 static struct ifnet * 2742 bridge_rtlookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan) 2743 { 2744 struct bridge_rtnode *brt; 2745 2746 BRIDGE_LOCK_ASSERT(sc); 2747 2748 if ((brt = bridge_rtnode_lookup(sc, addr, vlan)) == NULL) 2749 return (NULL); 2750 2751 return (brt->brt_ifp); 2752 } 2753 2754 /* 2755 * bridge_rttrim: 2756 * 2757 * Trim the routine table so that we have a number 2758 * of routing entries less than or equal to the 2759 * maximum number. 2760 */ 2761 static void 2762 bridge_rttrim(struct bridge_softc *sc) 2763 { 2764 struct bridge_rtnode *brt, *nbrt; 2765 2766 BRIDGE_LOCK_ASSERT(sc); 2767 2768 /* Make sure we actually need to do this. */ 2769 if (sc->sc_brtcnt <= sc->sc_brtmax) 2770 return; 2771 2772 /* Force an aging cycle; this might trim enough addresses. */ 2773 bridge_rtage(sc); 2774 if (sc->sc_brtcnt <= sc->sc_brtmax) 2775 return; 2776 2777 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { 2778 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) { 2779 bridge_rtnode_destroy(sc, brt); 2780 if (sc->sc_brtcnt <= sc->sc_brtmax) 2781 return; 2782 } 2783 } 2784 } 2785 2786 /* 2787 * bridge_timer: 2788 * 2789 * Aging timer for the bridge. 2790 */ 2791 static void 2792 bridge_timer(void *arg) 2793 { 2794 struct bridge_softc *sc = arg; 2795 2796 BRIDGE_LOCK_ASSERT(sc); 2797 2798 /* Destruction of rtnodes requires a proper vnet context */ 2799 CURVNET_SET(sc->sc_ifp->if_vnet); 2800 bridge_rtage(sc); 2801 2802 if (sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) 2803 callout_reset(&sc->sc_brcallout, 2804 bridge_rtable_prune_period * hz, bridge_timer, sc); 2805 CURVNET_RESTORE(); 2806 } 2807 2808 /* 2809 * bridge_rtage: 2810 * 2811 * Perform an aging cycle. 2812 */ 2813 static void 2814 bridge_rtage(struct bridge_softc *sc) 2815 { 2816 struct bridge_rtnode *brt, *nbrt; 2817 2818 BRIDGE_LOCK_ASSERT(sc); 2819 2820 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { 2821 if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) { 2822 if (time_uptime >= brt->brt_expire) 2823 bridge_rtnode_destroy(sc, brt); 2824 } 2825 } 2826 } 2827 2828 /* 2829 * bridge_rtflush: 2830 * 2831 * Remove all dynamic addresses from the bridge. 2832 */ 2833 static void 2834 bridge_rtflush(struct bridge_softc *sc, int full) 2835 { 2836 struct bridge_rtnode *brt, *nbrt; 2837 2838 BRIDGE_LOCK_ASSERT(sc); 2839 2840 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { 2841 if (full || (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) 2842 bridge_rtnode_destroy(sc, brt); 2843 } 2844 } 2845 2846 /* 2847 * bridge_rtdaddr: 2848 * 2849 * Remove an address from the table. 2850 */ 2851 static int 2852 bridge_rtdaddr(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan) 2853 { 2854 struct bridge_rtnode *brt; 2855 int found = 0; 2856 2857 BRIDGE_LOCK_ASSERT(sc); 2858 2859 /* 2860 * If vlan is zero then we want to delete for all vlans so the lookup 2861 * may return more than one. 2862 */ 2863 while ((brt = bridge_rtnode_lookup(sc, addr, vlan)) != NULL) { 2864 bridge_rtnode_destroy(sc, brt); 2865 found = 1; 2866 } 2867 2868 return (found ? 0 : ENOENT); 2869 } 2870 2871 /* 2872 * bridge_rtdelete: 2873 * 2874 * Delete routes to a speicifc member interface. 2875 */ 2876 static void 2877 bridge_rtdelete(struct bridge_softc *sc, struct ifnet *ifp, int full) 2878 { 2879 struct bridge_rtnode *brt, *nbrt; 2880 2881 BRIDGE_LOCK_ASSERT(sc); 2882 2883 LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { 2884 if (brt->brt_ifp == ifp && (full || 2885 (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)) 2886 bridge_rtnode_destroy(sc, brt); 2887 } 2888 } 2889 2890 /* 2891 * bridge_rtable_init: 2892 * 2893 * Initialize the route table for this bridge. 2894 */ 2895 static void 2896 bridge_rtable_init(struct bridge_softc *sc) 2897 { 2898 int i; 2899 2900 sc->sc_rthash = malloc(sizeof(*sc->sc_rthash) * BRIDGE_RTHASH_SIZE, 2901 M_DEVBUF, M_WAITOK); 2902 2903 for (i = 0; i < BRIDGE_RTHASH_SIZE; i++) 2904 LIST_INIT(&sc->sc_rthash[i]); 2905 2906 sc->sc_rthash_key = arc4random(); 2907 LIST_INIT(&sc->sc_rtlist); 2908 } 2909 2910 /* 2911 * bridge_rtable_fini: 2912 * 2913 * Deconstruct the route table for this bridge. 2914 */ 2915 static void 2916 bridge_rtable_fini(struct bridge_softc *sc) 2917 { 2918 2919 KASSERT(sc->sc_brtcnt == 0, 2920 ("%s: %d bridge routes referenced", __func__, sc->sc_brtcnt)); 2921 free(sc->sc_rthash, M_DEVBUF); 2922 } 2923 2924 /* 2925 * The following hash function is adapted from "Hash Functions" by Bob Jenkins 2926 * ("Algorithm Alley", Dr. Dobbs Journal, September 1997). 2927 */ 2928 #define mix(a, b, c) \ 2929 do { \ 2930 a -= b; a -= c; a ^= (c >> 13); \ 2931 b -= c; b -= a; b ^= (a << 8); \ 2932 c -= a; c -= b; c ^= (b >> 13); \ 2933 a -= b; a -= c; a ^= (c >> 12); \ 2934 b -= c; b -= a; b ^= (a << 16); \ 2935 c -= a; c -= b; c ^= (b >> 5); \ 2936 a -= b; a -= c; a ^= (c >> 3); \ 2937 b -= c; b -= a; b ^= (a << 10); \ 2938 c -= a; c -= b; c ^= (b >> 15); \ 2939 } while (/*CONSTCOND*/0) 2940 2941 static __inline uint32_t 2942 bridge_rthash(struct bridge_softc *sc, const uint8_t *addr) 2943 { 2944 uint32_t a = 0x9e3779b9, b = 0x9e3779b9, c = sc->sc_rthash_key; 2945 2946 b += addr[5] << 8; 2947 b += addr[4]; 2948 a += addr[3] << 24; 2949 a += addr[2] << 16; 2950 a += addr[1] << 8; 2951 a += addr[0]; 2952 2953 mix(a, b, c); 2954 2955 return (c & BRIDGE_RTHASH_MASK); 2956 } 2957 2958 #undef mix 2959 2960 static int 2961 bridge_rtnode_addr_cmp(const uint8_t *a, const uint8_t *b) 2962 { 2963 int i, d; 2964 2965 for (i = 0, d = 0; i < ETHER_ADDR_LEN && d == 0; i++) { 2966 d = ((int)a[i]) - ((int)b[i]); 2967 } 2968 2969 return (d); 2970 } 2971 2972 /* 2973 * bridge_rtnode_lookup: 2974 * 2975 * Look up a bridge route node for the specified destination. Compare the 2976 * vlan id or if zero then just return the first match. 2977 */ 2978 static struct bridge_rtnode * 2979 bridge_rtnode_lookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan) 2980 { 2981 struct bridge_rtnode *brt; 2982 uint32_t hash; 2983 int dir; 2984 2985 BRIDGE_LOCK_ASSERT(sc); 2986 2987 hash = bridge_rthash(sc, addr); 2988 LIST_FOREACH(brt, &sc->sc_rthash[hash], brt_hash) { 2989 dir = bridge_rtnode_addr_cmp(addr, brt->brt_addr); 2990 if (dir == 0 && (brt->brt_vlan == vlan || vlan == 0)) 2991 return (brt); 2992 if (dir > 0) 2993 return (NULL); 2994 } 2995 2996 return (NULL); 2997 } 2998 2999 /* 3000 * bridge_rtnode_insert: 3001 * 3002 * Insert the specified bridge node into the route table. We 3003 * assume the entry is not already in the table. 3004 */ 3005 static int 3006 bridge_rtnode_insert(struct bridge_softc *sc, struct bridge_rtnode *brt) 3007 { 3008 struct bridge_rtnode *lbrt; 3009 uint32_t hash; 3010 int dir; 3011 3012 BRIDGE_LOCK_ASSERT(sc); 3013 3014 hash = bridge_rthash(sc, brt->brt_addr); 3015 3016 lbrt = LIST_FIRST(&sc->sc_rthash[hash]); 3017 if (lbrt == NULL) { 3018 LIST_INSERT_HEAD(&sc->sc_rthash[hash], brt, brt_hash); 3019 goto out; 3020 } 3021 3022 do { 3023 dir = bridge_rtnode_addr_cmp(brt->brt_addr, lbrt->brt_addr); 3024 if (dir == 0 && brt->brt_vlan == lbrt->brt_vlan) 3025 return (EEXIST); 3026 if (dir > 0) { 3027 LIST_INSERT_BEFORE(lbrt, brt, brt_hash); 3028 goto out; 3029 } 3030 if (LIST_NEXT(lbrt, brt_hash) == NULL) { 3031 LIST_INSERT_AFTER(lbrt, brt, brt_hash); 3032 goto out; 3033 } 3034 lbrt = LIST_NEXT(lbrt, brt_hash); 3035 } while (lbrt != NULL); 3036 3037 #ifdef DIAGNOSTIC 3038 panic("bridge_rtnode_insert: impossible"); 3039 #endif 3040 3041 out: 3042 LIST_INSERT_HEAD(&sc->sc_rtlist, brt, brt_list); 3043 sc->sc_brtcnt++; 3044 3045 return (0); 3046 } 3047 3048 /* 3049 * bridge_rtnode_destroy: 3050 * 3051 * Destroy a bridge rtnode. 3052 */ 3053 static void 3054 bridge_rtnode_destroy(struct bridge_softc *sc, struct bridge_rtnode *brt) 3055 { 3056 BRIDGE_LOCK_ASSERT(sc); 3057 3058 LIST_REMOVE(brt, brt_hash); 3059 3060 LIST_REMOVE(brt, brt_list); 3061 sc->sc_brtcnt--; 3062 brt->brt_dst->bif_addrcnt--; 3063 uma_zfree(V_bridge_rtnode_zone, brt); 3064 } 3065 3066 /* 3067 * bridge_rtable_expire: 3068 * 3069 * Set the expiry time for all routes on an interface. 3070 */ 3071 static void 3072 bridge_rtable_expire(struct ifnet *ifp, int age) 3073 { 3074 struct bridge_softc *sc = ifp->if_bridge; 3075 struct bridge_rtnode *brt; 3076 3077 CURVNET_SET(ifp->if_vnet); 3078 BRIDGE_LOCK(sc); 3079 3080 /* 3081 * If the age is zero then flush, otherwise set all the expiry times to 3082 * age for the interface 3083 */ 3084 if (age == 0) 3085 bridge_rtdelete(sc, ifp, IFBF_FLUSHDYN); 3086 else { 3087 LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) { 3088 /* Cap the expiry time to 'age' */ 3089 if (brt->brt_ifp == ifp && 3090 brt->brt_expire > time_uptime + age && 3091 (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) 3092 brt->brt_expire = time_uptime + age; 3093 } 3094 } 3095 BRIDGE_UNLOCK(sc); 3096 CURVNET_RESTORE(); 3097 } 3098 3099 /* 3100 * bridge_state_change: 3101 * 3102 * Callback from the bridgestp code when a port changes states. 3103 */ 3104 static void 3105 bridge_state_change(struct ifnet *ifp, int state) 3106 { 3107 struct bridge_softc *sc = ifp->if_bridge; 3108 static const char *stpstates[] = { 3109 "disabled", 3110 "listening", 3111 "learning", 3112 "forwarding", 3113 "blocking", 3114 "discarding" 3115 }; 3116 3117 CURVNET_SET(ifp->if_vnet); 3118 if (V_log_stp) 3119 log(LOG_NOTICE, "%s: state changed to %s on %s\n", 3120 sc->sc_ifp->if_xname, stpstates[state], ifp->if_xname); 3121 CURVNET_RESTORE(); 3122 } 3123 3124 /* 3125 * Send bridge packets through pfil if they are one of the types pfil can deal 3126 * with, or if they are ARP or REVARP. (pfil will pass ARP and REVARP without 3127 * question.) If *bifp or *ifp are NULL then packet filtering is skipped for 3128 * that interface. 3129 */ 3130 static int 3131 bridge_pfil(struct mbuf **mp, struct ifnet *bifp, struct ifnet *ifp, int dir) 3132 { 3133 int snap, error, i, hlen; 3134 struct ether_header *eh1, eh2; 3135 struct ip *ip; 3136 struct llc llc1; 3137 u_int16_t ether_type; 3138 pfil_return_t rv; 3139 3140 snap = 0; 3141 error = -1; /* Default error if not error == 0 */ 3142 3143 #if 0 3144 /* we may return with the IP fields swapped, ensure its not shared */ 3145 KASSERT(M_WRITABLE(*mp), ("%s: modifying a shared mbuf", __func__)); 3146 #endif 3147 3148 if (V_pfil_bridge == 0 && V_pfil_member == 0 && V_pfil_ipfw == 0) 3149 return (0); /* filtering is disabled */ 3150 3151 i = min((*mp)->m_pkthdr.len, max_protohdr); 3152 if ((*mp)->m_len < i) { 3153 *mp = m_pullup(*mp, i); 3154 if (*mp == NULL) { 3155 printf("%s: m_pullup failed\n", __func__); 3156 return (-1); 3157 } 3158 } 3159 3160 eh1 = mtod(*mp, struct ether_header *); 3161 ether_type = ntohs(eh1->ether_type); 3162 3163 /* 3164 * Check for SNAP/LLC. 3165 */ 3166 if (ether_type < ETHERMTU) { 3167 struct llc *llc2 = (struct llc *)(eh1 + 1); 3168 3169 if ((*mp)->m_len >= ETHER_HDR_LEN + 8 && 3170 llc2->llc_dsap == LLC_SNAP_LSAP && 3171 llc2->llc_ssap == LLC_SNAP_LSAP && 3172 llc2->llc_control == LLC_UI) { 3173 ether_type = htons(llc2->llc_un.type_snap.ether_type); 3174 snap = 1; 3175 } 3176 } 3177 3178 /* 3179 * If we're trying to filter bridge traffic, don't look at anything 3180 * other than IP and ARP traffic. If the filter doesn't understand 3181 * IPv6, don't allow IPv6 through the bridge either. This is lame 3182 * since if we really wanted, say, an AppleTalk filter, we are hosed, 3183 * but of course we don't have an AppleTalk filter to begin with. 3184 * (Note that since pfil doesn't understand ARP it will pass *ALL* 3185 * ARP traffic.) 3186 */ 3187 switch (ether_type) { 3188 case ETHERTYPE_ARP: 3189 case ETHERTYPE_REVARP: 3190 if (V_pfil_ipfw_arp == 0) 3191 return (0); /* Automatically pass */ 3192 break; 3193 3194 case ETHERTYPE_IP: 3195 #ifdef INET6 3196 case ETHERTYPE_IPV6: 3197 #endif /* INET6 */ 3198 break; 3199 default: 3200 /* 3201 * Check to see if the user wants to pass non-ip 3202 * packets, these will not be checked by pfil(9) and 3203 * passed unconditionally so the default is to drop. 3204 */ 3205 if (V_pfil_onlyip) 3206 goto bad; 3207 } 3208 3209 /* Run the packet through pfil before stripping link headers */ 3210 if (PFIL_HOOKED_OUT(V_link_pfil_head) && V_pfil_ipfw != 0 && 3211 dir == PFIL_OUT && ifp != NULL) { 3212 switch (pfil_run_hooks(V_link_pfil_head, mp, ifp, dir, NULL)) { 3213 case PFIL_DROPPED: 3214 return (EACCES); 3215 case PFIL_CONSUMED: 3216 return (0); 3217 } 3218 } 3219 3220 /* Strip off the Ethernet header and keep a copy. */ 3221 m_copydata(*mp, 0, ETHER_HDR_LEN, (caddr_t) &eh2); 3222 m_adj(*mp, ETHER_HDR_LEN); 3223 3224 /* Strip off snap header, if present */ 3225 if (snap) { 3226 m_copydata(*mp, 0, sizeof(struct llc), (caddr_t) &llc1); 3227 m_adj(*mp, sizeof(struct llc)); 3228 } 3229 3230 /* 3231 * Check the IP header for alignment and errors 3232 */ 3233 if (dir == PFIL_IN) { 3234 switch (ether_type) { 3235 case ETHERTYPE_IP: 3236 error = bridge_ip_checkbasic(mp); 3237 break; 3238 #ifdef INET6 3239 case ETHERTYPE_IPV6: 3240 error = bridge_ip6_checkbasic(mp); 3241 break; 3242 #endif /* INET6 */ 3243 default: 3244 error = 0; 3245 } 3246 if (error) 3247 goto bad; 3248 } 3249 3250 error = 0; 3251 3252 /* 3253 * Run the packet through pfil 3254 */ 3255 rv = PFIL_PASS; 3256 switch (ether_type) { 3257 case ETHERTYPE_IP: 3258 /* 3259 * Run pfil on the member interface and the bridge, both can 3260 * be skipped by clearing pfil_member or pfil_bridge. 3261 * 3262 * Keep the order: 3263 * in_if -> bridge_if -> out_if 3264 */ 3265 if (V_pfil_bridge && dir == PFIL_OUT && bifp != NULL && (rv = 3266 pfil_run_hooks(V_inet_pfil_head, mp, bifp, dir, NULL)) != 3267 PFIL_PASS) 3268 break; 3269 3270 if (V_pfil_member && ifp != NULL && (rv = 3271 pfil_run_hooks(V_inet_pfil_head, mp, ifp, dir, NULL)) != 3272 PFIL_PASS) 3273 break; 3274 3275 if (V_pfil_bridge && dir == PFIL_IN && bifp != NULL && (rv = 3276 pfil_run_hooks(V_inet_pfil_head, mp, bifp, dir, NULL)) != 3277 PFIL_PASS) 3278 break; 3279 3280 /* check if we need to fragment the packet */ 3281 /* bridge_fragment generates a mbuf chain of packets */ 3282 /* that already include eth headers */ 3283 if (V_pfil_member && ifp != NULL && dir == PFIL_OUT) { 3284 i = (*mp)->m_pkthdr.len; 3285 if (i > ifp->if_mtu) { 3286 error = bridge_fragment(ifp, mp, &eh2, snap, 3287 &llc1); 3288 return (error); 3289 } 3290 } 3291 3292 /* Recalculate the ip checksum. */ 3293 ip = mtod(*mp, struct ip *); 3294 hlen = ip->ip_hl << 2; 3295 if (hlen < sizeof(struct ip)) 3296 goto bad; 3297 if (hlen > (*mp)->m_len) { 3298 if ((*mp = m_pullup(*mp, hlen)) == NULL) 3299 goto bad; 3300 ip = mtod(*mp, struct ip *); 3301 if (ip == NULL) 3302 goto bad; 3303 } 3304 ip->ip_sum = 0; 3305 if (hlen == sizeof(struct ip)) 3306 ip->ip_sum = in_cksum_hdr(ip); 3307 else 3308 ip->ip_sum = in_cksum(*mp, hlen); 3309 3310 break; 3311 #ifdef INET6 3312 case ETHERTYPE_IPV6: 3313 if (V_pfil_bridge && dir == PFIL_OUT && bifp != NULL && (rv = 3314 pfil_run_hooks(V_inet6_pfil_head, mp, bifp, dir, NULL)) != 3315 PFIL_PASS) 3316 break; 3317 3318 if (V_pfil_member && ifp != NULL && (rv = 3319 pfil_run_hooks(V_inet6_pfil_head, mp, ifp, dir, NULL)) != 3320 PFIL_PASS) 3321 break; 3322 3323 if (V_pfil_bridge && dir == PFIL_IN && bifp != NULL && (rv = 3324 pfil_run_hooks(V_inet6_pfil_head, mp, bifp, dir, NULL)) != 3325 PFIL_PASS) 3326 break; 3327 break; 3328 #endif 3329 } 3330 3331 switch (rv) { 3332 case PFIL_CONSUMED: 3333 return (0); 3334 case PFIL_DROPPED: 3335 return (EACCES); 3336 default: 3337 break; 3338 } 3339 3340 error = -1; 3341 3342 /* 3343 * Finally, put everything back the way it was and return 3344 */ 3345 if (snap) { 3346 M_PREPEND(*mp, sizeof(struct llc), M_NOWAIT); 3347 if (*mp == NULL) 3348 return (error); 3349 bcopy(&llc1, mtod(*mp, caddr_t), sizeof(struct llc)); 3350 } 3351 3352 M_PREPEND(*mp, ETHER_HDR_LEN, M_NOWAIT); 3353 if (*mp == NULL) 3354 return (error); 3355 bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN); 3356 3357 return (0); 3358 3359 bad: 3360 m_freem(*mp); 3361 *mp = NULL; 3362 return (error); 3363 } 3364 3365 /* 3366 * Perform basic checks on header size since 3367 * pfil assumes ip_input has already processed 3368 * it for it. Cut-and-pasted from ip_input.c. 3369 * Given how simple the IPv6 version is, 3370 * does the IPv4 version really need to be 3371 * this complicated? 3372 * 3373 * XXX Should we update ipstat here, or not? 3374 * XXX Right now we update ipstat but not 3375 * XXX csum_counter. 3376 */ 3377 static int 3378 bridge_ip_checkbasic(struct mbuf **mp) 3379 { 3380 struct mbuf *m = *mp; 3381 struct ip *ip; 3382 int len, hlen; 3383 u_short sum; 3384 3385 if (*mp == NULL) 3386 return (-1); 3387 3388 if (IP_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) { 3389 if ((m = m_copyup(m, sizeof(struct ip), 3390 (max_linkhdr + 3) & ~3)) == NULL) { 3391 /* XXXJRT new stat, please */ 3392 KMOD_IPSTAT_INC(ips_toosmall); 3393 goto bad; 3394 } 3395 } else if (__predict_false(m->m_len < sizeof (struct ip))) { 3396 if ((m = m_pullup(m, sizeof (struct ip))) == NULL) { 3397 KMOD_IPSTAT_INC(ips_toosmall); 3398 goto bad; 3399 } 3400 } 3401 ip = mtod(m, struct ip *); 3402 if (ip == NULL) goto bad; 3403 3404 if (ip->ip_v != IPVERSION) { 3405 KMOD_IPSTAT_INC(ips_badvers); 3406 goto bad; 3407 } 3408 hlen = ip->ip_hl << 2; 3409 if (hlen < sizeof(struct ip)) { /* minimum header length */ 3410 KMOD_IPSTAT_INC(ips_badhlen); 3411 goto bad; 3412 } 3413 if (hlen > m->m_len) { 3414 if ((m = m_pullup(m, hlen)) == NULL) { 3415 KMOD_IPSTAT_INC(ips_badhlen); 3416 goto bad; 3417 } 3418 ip = mtod(m, struct ip *); 3419 if (ip == NULL) goto bad; 3420 } 3421 3422 if (m->m_pkthdr.csum_flags & CSUM_IP_CHECKED) { 3423 sum = !(m->m_pkthdr.csum_flags & CSUM_IP_VALID); 3424 } else { 3425 if (hlen == sizeof(struct ip)) { 3426 sum = in_cksum_hdr(ip); 3427 } else { 3428 sum = in_cksum(m, hlen); 3429 } 3430 } 3431 if (sum) { 3432 KMOD_IPSTAT_INC(ips_badsum); 3433 goto bad; 3434 } 3435 3436 /* Retrieve the packet length. */ 3437 len = ntohs(ip->ip_len); 3438 3439 /* 3440 * Check for additional length bogosity 3441 */ 3442 if (len < hlen) { 3443 KMOD_IPSTAT_INC(ips_badlen); 3444 goto bad; 3445 } 3446 3447 /* 3448 * Check that the amount of data in the buffers 3449 * is as at least much as the IP header would have us expect. 3450 * Drop packet if shorter than we expect. 3451 */ 3452 if (m->m_pkthdr.len < len) { 3453 KMOD_IPSTAT_INC(ips_tooshort); 3454 goto bad; 3455 } 3456 3457 /* Checks out, proceed */ 3458 *mp = m; 3459 return (0); 3460 3461 bad: 3462 *mp = m; 3463 return (-1); 3464 } 3465 3466 #ifdef INET6 3467 /* 3468 * Same as above, but for IPv6. 3469 * Cut-and-pasted from ip6_input.c. 3470 * XXX Should we update ip6stat, or not? 3471 */ 3472 static int 3473 bridge_ip6_checkbasic(struct mbuf **mp) 3474 { 3475 struct mbuf *m = *mp; 3476 struct ip6_hdr *ip6; 3477 3478 /* 3479 * If the IPv6 header is not aligned, slurp it up into a new 3480 * mbuf with space for link headers, in the event we forward 3481 * it. Otherwise, if it is aligned, make sure the entire base 3482 * IPv6 header is in the first mbuf of the chain. 3483 */ 3484 if (IP6_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) { 3485 struct ifnet *inifp = m->m_pkthdr.rcvif; 3486 if ((m = m_copyup(m, sizeof(struct ip6_hdr), 3487 (max_linkhdr + 3) & ~3)) == NULL) { 3488 /* XXXJRT new stat, please */ 3489 IP6STAT_INC(ip6s_toosmall); 3490 in6_ifstat_inc(inifp, ifs6_in_hdrerr); 3491 goto bad; 3492 } 3493 } else if (__predict_false(m->m_len < sizeof(struct ip6_hdr))) { 3494 struct ifnet *inifp = m->m_pkthdr.rcvif; 3495 if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) { 3496 IP6STAT_INC(ip6s_toosmall); 3497 in6_ifstat_inc(inifp, ifs6_in_hdrerr); 3498 goto bad; 3499 } 3500 } 3501 3502 ip6 = mtod(m, struct ip6_hdr *); 3503 3504 if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) { 3505 IP6STAT_INC(ip6s_badvers); 3506 in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr); 3507 goto bad; 3508 } 3509 3510 /* Checks out, proceed */ 3511 *mp = m; 3512 return (0); 3513 3514 bad: 3515 *mp = m; 3516 return (-1); 3517 } 3518 #endif /* INET6 */ 3519 3520 /* 3521 * bridge_fragment: 3522 * 3523 * Fragment mbuf chain in multiple packets and prepend ethernet header. 3524 */ 3525 static int 3526 bridge_fragment(struct ifnet *ifp, struct mbuf **mp, struct ether_header *eh, 3527 int snap, struct llc *llc) 3528 { 3529 struct mbuf *m = *mp, *nextpkt = NULL, *mprev = NULL, *mcur = NULL; 3530 struct ip *ip; 3531 int error = -1; 3532 3533 if (m->m_len < sizeof(struct ip) && 3534 (m = m_pullup(m, sizeof(struct ip))) == NULL) 3535 goto dropit; 3536 ip = mtod(m, struct ip *); 3537 3538 m->m_pkthdr.csum_flags |= CSUM_IP; 3539 error = ip_fragment(ip, &m, ifp->if_mtu, ifp->if_hwassist); 3540 if (error) 3541 goto dropit; 3542 3543 /* 3544 * Walk the chain and re-add the Ethernet header for 3545 * each mbuf packet. 3546 */ 3547 for (mcur = m; mcur; mcur = mcur->m_nextpkt) { 3548 nextpkt = mcur->m_nextpkt; 3549 mcur->m_nextpkt = NULL; 3550 if (snap) { 3551 M_PREPEND(mcur, sizeof(struct llc), M_NOWAIT); 3552 if (mcur == NULL) { 3553 error = ENOBUFS; 3554 if (mprev != NULL) 3555 mprev->m_nextpkt = nextpkt; 3556 goto dropit; 3557 } 3558 bcopy(llc, mtod(mcur, caddr_t),sizeof(struct llc)); 3559 } 3560 3561 M_PREPEND(mcur, ETHER_HDR_LEN, M_NOWAIT); 3562 if (mcur == NULL) { 3563 error = ENOBUFS; 3564 if (mprev != NULL) 3565 mprev->m_nextpkt = nextpkt; 3566 goto dropit; 3567 } 3568 bcopy(eh, mtod(mcur, caddr_t), ETHER_HDR_LEN); 3569 3570 /* 3571 * The previous two M_PREPEND could have inserted one or two 3572 * mbufs in front so we have to update the previous packet's 3573 * m_nextpkt. 3574 */ 3575 mcur->m_nextpkt = nextpkt; 3576 if (mprev != NULL) 3577 mprev->m_nextpkt = mcur; 3578 else { 3579 /* The first mbuf in the original chain needs to be 3580 * updated. */ 3581 *mp = mcur; 3582 } 3583 mprev = mcur; 3584 } 3585 3586 KMOD_IPSTAT_INC(ips_fragmented); 3587 return (error); 3588 3589 dropit: 3590 for (mcur = *mp; mcur; mcur = m) { /* droping the full packet chain */ 3591 m = mcur->m_nextpkt; 3592 m_freem(mcur); 3593 } 3594 return (error); 3595 } 3596 3597 static void 3598 bridge_linkstate(struct ifnet *ifp) 3599 { 3600 struct bridge_softc *sc = ifp->if_bridge; 3601 struct bridge_iflist *bif; 3602 3603 BRIDGE_LOCK(sc); 3604 bif = bridge_lookup_member_if(sc, ifp); 3605 if (bif == NULL) { 3606 BRIDGE_UNLOCK(sc); 3607 return; 3608 } 3609 bridge_linkcheck(sc); 3610 BRIDGE_UNLOCK(sc); 3611 3612 bstp_linkstate(&bif->bif_stp); 3613 } 3614 3615 static void 3616 bridge_linkcheck(struct bridge_softc *sc) 3617 { 3618 struct bridge_iflist *bif; 3619 int new_link, hasls; 3620 3621 BRIDGE_LOCK_ASSERT(sc); 3622 new_link = LINK_STATE_DOWN; 3623 hasls = 0; 3624 /* Our link is considered up if at least one of our ports is active */ 3625 LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { 3626 if (bif->bif_ifp->if_capabilities & IFCAP_LINKSTATE) 3627 hasls++; 3628 if (bif->bif_ifp->if_link_state == LINK_STATE_UP) { 3629 new_link = LINK_STATE_UP; 3630 break; 3631 } 3632 } 3633 if (!LIST_EMPTY(&sc->sc_iflist) && !hasls) { 3634 /* If no interfaces support link-state then we default to up */ 3635 new_link = LINK_STATE_UP; 3636 } 3637 if_link_state_change(sc->sc_ifp, new_link); 3638 } 3639