1 // SPDX-License-Identifier: GPL-2.0-or-later 2 /* 3 * Handle incoming frames 4 * Linux ethernet bridge 5 * 6 * Authors: 7 * Lennert Buytenhek <buytenh@gnu.org> 8 */ 9 10 #include <linux/slab.h> 11 #include <linux/kernel.h> 12 #include <linux/netdevice.h> 13 #include <linux/etherdevice.h> 14 #include <linux/netfilter_bridge.h> 15 #ifdef CONFIG_NETFILTER_FAMILY_BRIDGE 16 #include <net/netfilter/nf_queue.h> 17 #endif 18 #include <linux/neighbour.h> 19 #include <net/arp.h> 20 #include <net/dsa.h> 21 #include <linux/export.h> 22 #include <linux/rculist.h> 23 #include "br_private.h" 24 #include "br_private_tunnel.h" 25 26 static int 27 br_netif_receive_skb(struct net *net, struct sock *sk, struct sk_buff *skb) 28 { 29 br_drop_fake_rtable(skb); 30 return netif_receive_skb(skb); 31 } 32 33 static int br_pass_frame_up(struct sk_buff *skb, bool promisc) 34 { 35 struct net_device *indev, *brdev = BR_INPUT_SKB_CB(skb)->brdev; 36 struct net_bridge *br = netdev_priv(brdev); 37 struct net_bridge_vlan_group *vg; 38 39 dev_sw_netstats_rx_add(brdev, skb->len); 40 41 vg = br_vlan_group_rcu(br); 42 43 /* Reset the offload_fwd_mark because there could be a stacked 44 * bridge above, and it should not think this bridge it doing 45 * that bridge's work forwarding out its ports. 46 */ 47 br_switchdev_frame_unmark(skb); 48 49 /* Bridge is just like any other port. Make sure the 50 * packet is allowed except in promisc mode when someone 51 * may be running packet capture. 52 */ 53 if (!(brdev->flags & IFF_PROMISC) && 54 !br_allowed_egress(vg, skb)) { 55 kfree_skb(skb); 56 return NET_RX_DROP; 57 } 58 59 indev = skb->dev; 60 skb->dev = brdev; 61 skb = br_handle_vlan(br, NULL, vg, skb); 62 if (!skb) 63 return NET_RX_DROP; 64 /* update the multicast stats if the packet is IGMP/MLD */ 65 br_multicast_count(br, NULL, skb, br_multicast_igmp_type(skb), 66 BR_MCAST_DIR_TX); 67 68 BR_INPUT_SKB_CB(skb)->promisc = promisc; 69 70 return NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_IN, 71 dev_net(indev), NULL, skb, indev, NULL, 72 br_netif_receive_skb); 73 } 74 75 /* note: already called with rcu_read_lock */ 76 int br_handle_frame_finish(struct net *net, struct sock *sk, struct sk_buff *skb) 77 { 78 enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; 79 struct net_bridge_port *p = br_port_get_rcu(skb->dev); 80 enum br_pkt_type pkt_type = BR_PKT_UNICAST; 81 struct net_bridge_fdb_entry *dst = NULL; 82 struct net_bridge_mcast_port *pmctx; 83 struct net_bridge_mdb_entry *mdst; 84 bool local_rcv, mcast_hit = false; 85 struct net_bridge_mcast *brmctx; 86 struct net_bridge_vlan *vlan; 87 struct net_bridge *br; 88 bool promisc; 89 u16 vid = 0; 90 u8 state; 91 92 if (!p) 93 goto drop; 94 95 br = p->br; 96 97 if (br_mst_is_enabled(br)) { 98 state = BR_STATE_FORWARDING; 99 } else { 100 if (p->state == BR_STATE_DISABLED) { 101 reason = SKB_DROP_REASON_BRIDGE_INGRESS_STP_STATE; 102 goto drop; 103 } 104 105 state = p->state; 106 } 107 108 brmctx = &p->br->multicast_ctx; 109 pmctx = &p->multicast_ctx; 110 if (!br_allowed_ingress(p->br, nbp_vlan_group_rcu(p), skb, &vid, 111 &state, &vlan)) 112 goto out; 113 114 if (p->flags & BR_PORT_LOCKED) { 115 struct net_bridge_fdb_entry *fdb_src = 116 br_fdb_find_rcu(br, eth_hdr(skb)->h_source, vid); 117 118 if (!fdb_src) { 119 /* FDB miss. Create locked FDB entry if MAB is enabled 120 * and drop the packet. 121 */ 122 if (p->flags & BR_PORT_MAB) 123 br_fdb_update(br, p, eth_hdr(skb)->h_source, 124 vid, BIT(BR_FDB_LOCKED)); 125 goto drop; 126 } else if (READ_ONCE(fdb_src->dst) != p || 127 test_bit(BR_FDB_LOCAL, &fdb_src->flags)) { 128 /* FDB mismatch. Drop the packet without roaming. */ 129 goto drop; 130 } else if (test_bit(BR_FDB_LOCKED, &fdb_src->flags)) { 131 /* FDB match, but entry is locked. Refresh it and drop 132 * the packet. 133 */ 134 br_fdb_update(br, p, eth_hdr(skb)->h_source, vid, 135 BIT(BR_FDB_LOCKED)); 136 goto drop; 137 } 138 } 139 140 nbp_switchdev_frame_mark(p, skb); 141 142 /* insert into forwarding database after filtering to avoid spoofing */ 143 if (p->flags & BR_LEARNING) 144 br_fdb_update(br, p, eth_hdr(skb)->h_source, vid, 0); 145 146 promisc = !!(br->dev->flags & IFF_PROMISC); 147 local_rcv = promisc; 148 149 if (is_multicast_ether_addr(eth_hdr(skb)->h_dest)) { 150 /* by definition the broadcast is also a multicast address */ 151 if (is_broadcast_ether_addr(eth_hdr(skb)->h_dest)) { 152 pkt_type = BR_PKT_BROADCAST; 153 local_rcv = true; 154 } else { 155 pkt_type = BR_PKT_MULTICAST; 156 if (br_multicast_rcv(&brmctx, &pmctx, vlan, skb, vid)) 157 goto drop; 158 } 159 } 160 161 if (state == BR_STATE_LEARNING) { 162 reason = SKB_DROP_REASON_BRIDGE_INGRESS_STP_STATE; 163 goto drop; 164 } 165 166 BR_INPUT_SKB_CB(skb)->brdev = br->dev; 167 BR_INPUT_SKB_CB(skb)->src_port_isolated = !!(p->flags & BR_ISOLATED); 168 169 if (IS_ENABLED(CONFIG_INET) && 170 (skb->protocol == htons(ETH_P_ARP) || 171 skb->protocol == htons(ETH_P_RARP))) { 172 br_do_proxy_suppress_arp(skb, br, vid, p); 173 } else if (IS_ENABLED(CONFIG_IPV6) && 174 skb->protocol == htons(ETH_P_IPV6) && 175 br_opt_get(br, BROPT_NEIGH_SUPPRESS_ENABLED) && 176 pskb_may_pull(skb, sizeof(struct ipv6hdr) + 177 sizeof(struct nd_msg)) && 178 ipv6_hdr(skb)->nexthdr == IPPROTO_ICMPV6) { 179 struct nd_msg *msg, _msg; 180 181 msg = br_is_nd_neigh_msg(skb, &_msg); 182 if (msg) 183 br_do_suppress_nd(skb, br, vid, p, msg); 184 } 185 186 switch (pkt_type) { 187 case BR_PKT_MULTICAST: 188 mdst = br_mdb_entry_skb_get(brmctx, skb, vid); 189 if ((mdst || BR_INPUT_SKB_CB_MROUTERS_ONLY(skb)) && 190 br_multicast_querier_exists(brmctx, eth_hdr(skb), mdst)) { 191 if ((mdst && mdst->host_joined) || 192 br_multicast_is_router(brmctx, skb)) { 193 local_rcv = true; 194 DEV_STATS_INC(br->dev, multicast); 195 } 196 mcast_hit = true; 197 } else { 198 local_rcv = true; 199 DEV_STATS_INC(br->dev, multicast); 200 } 201 break; 202 case BR_PKT_UNICAST: 203 dst = br_fdb_find_rcu(br, eth_hdr(skb)->h_dest, vid); 204 break; 205 default: 206 break; 207 } 208 209 if (dst) { 210 unsigned long now = jiffies; 211 212 if (test_bit(BR_FDB_LOCAL, &dst->flags)) 213 return br_pass_frame_up(skb, false); 214 215 if (now != dst->used) 216 dst->used = now; 217 br_forward(dst->dst, skb, local_rcv, false); 218 } else { 219 if (!mcast_hit) 220 br_flood(br, skb, pkt_type, local_rcv, false, vid); 221 else 222 br_multicast_flood(mdst, skb, brmctx, local_rcv, false); 223 } 224 225 if (local_rcv) 226 return br_pass_frame_up(skb, promisc); 227 228 out: 229 return 0; 230 drop: 231 kfree_skb_reason(skb, reason); 232 goto out; 233 } 234 EXPORT_SYMBOL_GPL(br_handle_frame_finish); 235 236 static void __br_handle_local_finish(struct sk_buff *skb) 237 { 238 struct net_bridge_port *p = br_port_get_rcu(skb->dev); 239 u16 vid = 0; 240 241 /* check if vlan is allowed, to avoid spoofing */ 242 if ((p->flags & BR_LEARNING) && 243 nbp_state_should_learn(p) && 244 !br_opt_get(p->br, BROPT_NO_LL_LEARN) && 245 br_should_learn(p, skb, &vid)) 246 br_fdb_update(p->br, p, eth_hdr(skb)->h_source, vid, 0); 247 } 248 249 /* note: already called with rcu_read_lock */ 250 static int br_handle_local_finish(struct net *net, struct sock *sk, struct sk_buff *skb) 251 { 252 __br_handle_local_finish(skb); 253 254 /* return 1 to signal the okfn() was called so it's ok to use the skb */ 255 return 1; 256 } 257 258 static int nf_hook_bridge_pre(struct sk_buff *skb, struct sk_buff **pskb) 259 { 260 #ifdef CONFIG_NETFILTER_FAMILY_BRIDGE 261 struct nf_hook_entries *e = NULL; 262 struct nf_hook_state state; 263 unsigned int verdict, i; 264 struct net *net; 265 int ret; 266 267 net = dev_net(skb->dev); 268 #ifdef HAVE_JUMP_LABEL 269 if (!static_key_false(&nf_hooks_needed[NFPROTO_BRIDGE][NF_BR_PRE_ROUTING])) 270 goto frame_finish; 271 #endif 272 273 e = rcu_dereference(net->nf.hooks_bridge[NF_BR_PRE_ROUTING]); 274 if (!e) 275 goto frame_finish; 276 277 nf_hook_state_init(&state, NF_BR_PRE_ROUTING, 278 NFPROTO_BRIDGE, skb->dev, NULL, NULL, 279 net, br_handle_frame_finish); 280 281 for (i = 0; i < e->num_hook_entries; i++) { 282 verdict = nf_hook_entry_hookfn(&e->hooks[i], skb, &state); 283 switch (verdict & NF_VERDICT_MASK) { 284 case NF_ACCEPT: 285 if (BR_INPUT_SKB_CB(skb)->br_netfilter_broute) { 286 *pskb = skb; 287 return RX_HANDLER_PASS; 288 } 289 break; 290 case NF_DROP: 291 kfree_skb(skb); 292 return RX_HANDLER_CONSUMED; 293 case NF_QUEUE: 294 ret = nf_queue(skb, &state, i, verdict); 295 if (ret == 1) 296 continue; 297 return RX_HANDLER_CONSUMED; 298 default: /* STOLEN */ 299 return RX_HANDLER_CONSUMED; 300 } 301 } 302 frame_finish: 303 net = dev_net(skb->dev); 304 br_handle_frame_finish(net, NULL, skb); 305 #else 306 br_handle_frame_finish(dev_net(skb->dev), NULL, skb); 307 #endif 308 return RX_HANDLER_CONSUMED; 309 } 310 311 /* Return 0 if the frame was not processed otherwise 1 312 * note: already called with rcu_read_lock 313 */ 314 static int br_process_frame_type(struct net_bridge_port *p, 315 struct sk_buff *skb) 316 { 317 struct br_frame_type *tmp; 318 319 hlist_for_each_entry_rcu(tmp, &p->br->frame_type_list, list) 320 if (unlikely(tmp->type == skb->protocol)) 321 return tmp->frame_handler(p, skb); 322 323 return 0; 324 } 325 326 /* 327 * Return NULL if skb is handled 328 * note: already called with rcu_read_lock 329 */ 330 static rx_handler_result_t br_handle_frame(struct sk_buff **pskb) 331 { 332 enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; 333 struct net_bridge_port *p; 334 struct sk_buff *skb = *pskb; 335 const unsigned char *dest = eth_hdr(skb)->h_dest; 336 337 if (unlikely(skb->pkt_type == PACKET_LOOPBACK)) 338 return RX_HANDLER_PASS; 339 340 if (!is_valid_ether_addr(eth_hdr(skb)->h_source)) { 341 reason = SKB_DROP_REASON_MAC_INVALID_SOURCE; 342 goto drop; 343 } 344 345 skb = skb_share_check(skb, GFP_ATOMIC); 346 if (!skb) 347 return RX_HANDLER_CONSUMED; 348 349 memset(skb->cb, 0, sizeof(struct br_input_skb_cb)); 350 br_tc_skb_miss_set(skb, false); 351 352 p = br_port_get_rcu(skb->dev); 353 if (p->flags & BR_VLAN_TUNNEL) 354 br_handle_ingress_vlan_tunnel(skb, p, nbp_vlan_group_rcu(p)); 355 356 if (unlikely(is_link_local_ether_addr(dest))) { 357 u16 fwd_mask = p->br->group_fwd_mask_required; 358 359 /* 360 * See IEEE 802.1D Table 7-10 Reserved addresses 361 * 362 * Assignment Value 363 * Bridge Group Address 01-80-C2-00-00-00 364 * (MAC Control) 802.3 01-80-C2-00-00-01 365 * (Link Aggregation) 802.3 01-80-C2-00-00-02 366 * 802.1X PAE address 01-80-C2-00-00-03 367 * 368 * 802.1AB LLDP 01-80-C2-00-00-0E 369 * 370 * Others reserved for future standardization 371 */ 372 fwd_mask |= p->group_fwd_mask; 373 switch (dest[5]) { 374 case 0x00: /* Bridge Group Address */ 375 /* If STP is turned off, 376 then must forward to keep loop detection */ 377 if (p->br->stp_enabled == BR_NO_STP || 378 fwd_mask & (1u << dest[5])) 379 goto forward; 380 *pskb = skb; 381 __br_handle_local_finish(skb); 382 return RX_HANDLER_PASS; 383 384 case 0x01: /* IEEE MAC (Pause) */ 385 reason = SKB_DROP_REASON_MAC_IEEE_MAC_CONTROL; 386 goto drop; 387 388 case 0x0E: /* 802.1AB LLDP */ 389 fwd_mask |= p->br->group_fwd_mask; 390 if (fwd_mask & (1u << dest[5])) 391 goto forward; 392 *pskb = skb; 393 __br_handle_local_finish(skb); 394 return RX_HANDLER_PASS; 395 396 default: 397 /* Allow selective forwarding for most other protocols */ 398 fwd_mask |= p->br->group_fwd_mask; 399 if (fwd_mask & (1u << dest[5])) 400 goto forward; 401 } 402 403 BR_INPUT_SKB_CB(skb)->promisc = false; 404 405 /* The else clause should be hit when nf_hook(): 406 * - returns < 0 (drop/error) 407 * - returns = 0 (stolen/nf_queue) 408 * Thus return 1 from the okfn() to signal the skb is ok to pass 409 */ 410 if (NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_IN, 411 dev_net(skb->dev), NULL, skb, skb->dev, NULL, 412 br_handle_local_finish) == 1) { 413 return RX_HANDLER_PASS; 414 } else { 415 return RX_HANDLER_CONSUMED; 416 } 417 } 418 419 if (unlikely(br_process_frame_type(p, skb))) 420 return RX_HANDLER_PASS; 421 422 forward: 423 if (br_mst_is_enabled(p->br)) 424 goto defer_stp_filtering; 425 426 switch (p->state) { 427 case BR_STATE_FORWARDING: 428 case BR_STATE_LEARNING: 429 defer_stp_filtering: 430 if (ether_addr_equal(p->br->dev->dev_addr, dest)) 431 skb->pkt_type = PACKET_HOST; 432 433 return nf_hook_bridge_pre(skb, pskb); 434 default: 435 reason = SKB_DROP_REASON_BRIDGE_INGRESS_STP_STATE; 436 drop: 437 kfree_skb_reason(skb, reason); 438 } 439 return RX_HANDLER_CONSUMED; 440 } 441 442 /* This function has no purpose other than to appease the br_port_get_rcu/rtnl 443 * helpers which identify bridged ports according to the rx_handler installed 444 * on them (so there _needs_ to be a bridge rx_handler even if we don't need it 445 * to do anything useful). This bridge won't support traffic to/from the stack, 446 * but only hardware bridging. So return RX_HANDLER_PASS so we don't steal 447 * frames from the ETH_P_XDSA packet_type handler. 448 */ 449 static rx_handler_result_t br_handle_frame_dummy(struct sk_buff **pskb) 450 { 451 return RX_HANDLER_PASS; 452 } 453 454 rx_handler_func_t *br_get_rx_handler(const struct net_device *dev) 455 { 456 if (netdev_uses_dsa(dev)) 457 return br_handle_frame_dummy; 458 459 return br_handle_frame; 460 } 461 462 void br_add_frame(struct net_bridge *br, struct br_frame_type *ft) 463 { 464 hlist_add_head_rcu(&ft->list, &br->frame_type_list); 465 } 466 467 void br_del_frame(struct net_bridge *br, struct br_frame_type *ft) 468 { 469 struct br_frame_type *tmp; 470 471 hlist_for_each_entry(tmp, &br->frame_type_list, list) 472 if (ft == tmp) { 473 hlist_del_rcu(&ft->list); 474 return; 475 } 476 } 477