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