projects/linux·blob master · refs

478 lines · 12.7 KiB

1// SPDX-License-Identifier: GPL-2.0-or-later
2/*
3 * Handle incoming frames
4 * Linux ethernet bridge
5 *
6 * Authors:
7 * Lennert Buytenhek <[email protected]>
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
26static int
27br_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
33static 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 */
76int 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(p)) {
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 (test_bit(BR_PORT_LOCKED_BIT, &p->flags)) {
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 (test_bit(BR_PORT_MAB_BIT, &p->flags))
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 (test_bit(BR_LEARNING_BIT, &p->flags))
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 = test_bit(BR_ISOLATED_BIT, &p->flags);
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 (ipv6_mod_enabled() &&
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;
180
181 msg = br_is_nd_neigh_msg(skb);
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 br->dev->flags & IFF_ALLMULTI) {
194 local_rcv = true;
195 DEV_STATS_INC(br->dev, multicast);
196 }
197 mcast_hit = true;
198 } else {
199 local_rcv = true;
200 DEV_STATS_INC(br->dev, multicast);
201 }
202 break;
203 case BR_PKT_UNICAST:
204 dst = br_fdb_find_rcu(br, eth_hdr(skb)->h_dest, vid);
205 if (unlikely(!dst && vid &&
206 br_opt_get(br, BROPT_FDB_LOCAL_VLAN_0))) {
207 dst = br_fdb_find_rcu(br, eth_hdr(skb)->h_dest, 0);
208 if (dst &&
209 (!test_bit(BR_FDB_LOCAL, &dst->flags) ||
210 test_bit(BR_FDB_ADDED_BY_USER, &dst->flags)))
211 dst = NULL;
212 }
213 break;
214 default:
215 break;
216 }
217
218 if (dst) {
219 unsigned long now = jiffies;
220
221 if (test_bit(BR_FDB_LOCAL, &dst->flags))
222 return br_pass_frame_up(skb, false);
223
224 if (now != READ_ONCE(dst->used))
225 WRITE_ONCE(dst->used, now);
226 br_forward(READ_ONCE(dst->dst), skb, local_rcv, false);
227 } else {
228 if (!mcast_hit)
229 br_flood(br, skb, pkt_type, local_rcv, false, vid);
230 else
231 br_multicast_flood(mdst, skb, brmctx, local_rcv, false);
232 }
233
234 if (local_rcv)
235 return br_pass_frame_up(skb, promisc);
236
237out:
238 return 0;
239drop:
240 kfree_skb_reason(skb, reason);
241 goto out;
242}
243EXPORT_SYMBOL_GPL(br_handle_frame_finish);
244
245static void __br_handle_local_finish(struct sk_buff *skb)
246{
247 struct net_bridge_port *p = br_port_get_rcu(skb->dev);
248 u16 vid = 0;
249
250 /* check if vlan is allowed, to avoid spoofing */
251 if (test_bit(BR_LEARNING_BIT, &p->flags) &&
252 nbp_state_should_learn(p) &&
253 !br_opt_get(p->br, BROPT_NO_LL_LEARN) &&
254 br_should_learn(p, skb, &vid))
255 br_fdb_update(p->br, p, eth_hdr(skb)->h_source, vid, 0);
256}
257
258/* note: already called with rcu_read_lock */
259static int br_handle_local_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
260{
261 __br_handle_local_finish(skb);
262
263 /* return 1 to signal the okfn() was called so it's ok to use the skb */
264 return 1;
265}
266
267static int nf_hook_bridge_pre(struct sk_buff *skb, struct sk_buff **pskb)
268{
269#ifdef CONFIG_NETFILTER_FAMILY_BRIDGE
270 struct nf_hook_entries *e = NULL;
271 struct nf_hook_state state;
272 unsigned int verdict, i;
273 struct net *net;
274 int ret;
275
276 net = dev_net(skb->dev);
277#ifdef CONFIG_JUMP_LABEL
278 if (!static_key_false(&nf_hooks_needed[NFPROTO_BRIDGE][NF_BR_PRE_ROUTING]))
279 goto frame_finish;
280#endif
281
282 e = rcu_dereference(net->nf.hooks_bridge[NF_BR_PRE_ROUTING]);
283 if (!e)
284 goto frame_finish;
285
286 nf_hook_state_init(&state, NF_BR_PRE_ROUTING,
287 NFPROTO_BRIDGE, skb->dev, NULL, NULL,
288 net, br_handle_frame_finish);
289
290 for (i = 0; i < e->num_hook_entries; i++) {
291 verdict = nf_hook_entry_hookfn(&e->hooks[i], skb, &state);
292 switch (verdict & NF_VERDICT_MASK) {
293 case NF_ACCEPT:
294 if (BR_INPUT_SKB_CB(skb)->br_netfilter_broute) {
295 *pskb = skb;
296 return RX_HANDLER_PASS;
297 }
298 break;
299 case NF_DROP:
300 kfree_skb(skb);
301 return RX_HANDLER_CONSUMED;
302 case NF_QUEUE:
303 ret = nf_queue(skb, &state, i, verdict);
304 if (ret == 1)
305 continue;
306 return RX_HANDLER_CONSUMED;
307 default: /* STOLEN */
308 return RX_HANDLER_CONSUMED;
309 }
310 }
311frame_finish:
312 net = dev_net(skb->dev);
313 br_handle_frame_finish(net, NULL, skb);
314#else
315 br_handle_frame_finish(dev_net(skb->dev), NULL, skb);
316#endif
317 return RX_HANDLER_CONSUMED;
318}
319
320#define BR_CFM_MRP_OPTS \
321 ((IS_ENABLED(CONFIG_BRIDGE_CFM) ? BIT(BROPT_CFM_ENABLED) : 0UL) | \
322 (IS_ENABLED(CONFIG_BRIDGE_MRP) ? BIT(BROPT_MRP_ENABLED) : 0UL))
323
324/* Return 0 if the frame was not processed otherwise 1
325 * note: already called with rcu_read_lock
326 */
327static int br_process_frame_type(struct net_bridge_port *p,
328 struct sk_buff *skb)
329{
330 struct net_bridge *br = p->br;
331
332 if (skb->protocol == htons(ETH_P_CFM) &&
333 br_opt_get(br, BROPT_CFM_ENABLED))
334 return br_cfm_frame_rx(p, skb);
335
336 if (skb->protocol == htons(ETH_P_MRP) &&
337 br_opt_get(br, BROPT_MRP_ENABLED))
338 return br_mrp_process(p, skb);
339
340 return 0;
341}
342
343/*
344 * Return NULL if skb is handled
345 * note: already called with rcu_read_lock
346 */
347static rx_handler_result_t br_handle_frame(struct sk_buff **pskb)
348{
349 enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED;
350 struct net_bridge_port *p;
351 struct sk_buff *skb = *pskb;
352 const unsigned char *dest = eth_hdr(skb)->h_dest;
353
354 if (unlikely(skb->pkt_type == PACKET_LOOPBACK))
355 return RX_HANDLER_PASS;
356
357 if (!is_valid_ether_addr(eth_hdr(skb)->h_source)) {
358 reason = SKB_DROP_REASON_MAC_INVALID_SOURCE;
359 goto drop;
360 }
361
362 skb = skb_share_check(skb, GFP_ATOMIC);
363 if (!skb)
364 return RX_HANDLER_CONSUMED;
365
366 memset(skb->cb, 0, sizeof(struct br_input_skb_cb));
367 br_tc_skb_miss_set(skb, false);
368
369 p = br_port_get_rcu(skb->dev);
370 if (test_bit(BR_VLAN_TUNNEL_BIT, &p->flags))
371 br_handle_ingress_vlan_tunnel(skb, p, nbp_vlan_group_rcu(p));
372
373 if (unlikely(is_link_local_ether_addr(dest))) {
374 u16 fwd_mask = p->br->group_fwd_mask_required;
375
376 /*
377 * See IEEE 802.1D Table 7-10 Reserved addresses
378 *
379 * Assignment Value
380 * Bridge Group Address 01-80-C2-00-00-00
381 * (MAC Control) 802.3 01-80-C2-00-00-01
382 * (Link Aggregation) 802.3 01-80-C2-00-00-02
383 * 802.1X PAE address 01-80-C2-00-00-03
384 *
385 * 802.1AB LLDP 01-80-C2-00-00-0E
386 *
387 * Others reserved for future standardization
388 */
389 fwd_mask |= p->group_fwd_mask;
390 switch (dest[5]) {
391 case 0x00: /* Bridge Group Address */
392 /* If STP is turned off,
393 then must forward to keep loop detection */
394 if (p->br->stp_enabled == BR_NO_STP ||
395 fwd_mask & (1u << dest[5]))
396 goto forward;
397 *pskb = skb;
398 __br_handle_local_finish(skb);
399 return RX_HANDLER_PASS;
400
401 case 0x01: /* IEEE MAC (Pause) */
402 reason = SKB_DROP_REASON_MAC_IEEE_MAC_CONTROL;
403 goto drop;
404
405 case 0x0E: /* 802.1AB LLDP */
406 fwd_mask |= p->br->group_fwd_mask;
407 if (fwd_mask & (1u << dest[5]))
408 goto forward;
409 *pskb = skb;
410 __br_handle_local_finish(skb);
411 return RX_HANDLER_PASS;
412
413 default:
414 /* Allow selective forwarding for most other protocols */
415 fwd_mask |= p->br->group_fwd_mask;
416 if (fwd_mask & (1u << dest[5]))
417 goto forward;
418 }
419
420 BR_INPUT_SKB_CB(skb)->promisc = false;
421
422 /* The else clause should be hit when nf_hook():
423 * - returns < 0 (drop/error)
424 * - returns = 0 (stolen/nf_queue)
425 * Thus return 1 from the okfn() to signal the skb is ok to pass
426 */
427 if (NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_IN,
428 dev_net(skb->dev), NULL, skb, skb->dev, NULL,
429 br_handle_local_finish) == 1) {
430 return RX_HANDLER_PASS;
431 } else {
432 return RX_HANDLER_CONSUMED;
433 }
434 }
435
436 if (unlikely((READ_ONCE(p->br->options) & BR_CFM_MRP_OPTS) &&
437 br_process_frame_type(p, skb)))
438 return RX_HANDLER_PASS;
439
440forward:
441 if (br_mst_is_enabled(p))
442 goto defer_stp_filtering;
443
444 switch (p->state) {
445 case BR_STATE_FORWARDING:
446 case BR_STATE_LEARNING:
447defer_stp_filtering:
448 if (ether_addr_equal(p->br->dev->dev_addr, dest))
449 skb->pkt_type = PACKET_HOST;
450
451 return nf_hook_bridge_pre(skb, pskb);
452 default:
453 reason = SKB_DROP_REASON_BRIDGE_INGRESS_STP_STATE;
454drop:
455 kfree_skb_reason(skb, reason);
456 }
457 return RX_HANDLER_CONSUMED;
458}
459
460/* This function has no purpose other than to appease the br_port_get_rcu/rtnl
461 * helpers which identify bridged ports according to the rx_handler installed
462 * on them (so there _needs_ to be a bridge rx_handler even if we don't need it
463 * to do anything useful). This bridge won't support traffic to/from the stack,
464 * but only hardware bridging. So return RX_HANDLER_PASS so we don't steal
465 * frames from the ETH_P_XDSA packet_type handler.
466 */
467static rx_handler_result_t br_handle_frame_dummy(struct sk_buff **pskb)
468{
469 return RX_HANDLER_PASS;
470}
471
472rx_handler_func_t *br_get_rx_handler(const struct net_device *dev)
473{
474 if (netdev_uses_dsa(dev))
475 return br_handle_frame_dummy;
476
477 return br_handle_frame;
478}