br_multicast.c 59.9 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13
/*
 * Bridge multicast support.
 *
 * Copyright (c) 2010 Herbert Xu <herbert@gondor.apana.org.au>
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License as published by the Free
 * Software Foundation; either version 2 of the License, or (at your option)
 * any later version.
 *
 */

#include <linux/err.h>
14
#include <linux/export.h>
15 16
#include <linux/if_ether.h>
#include <linux/igmp.h>
17
#include <linux/in.h>
18 19
#include <linux/jhash.h>
#include <linux/kernel.h>
20
#include <linux/log2.h>
21 22 23 24 25 26 27
#include <linux/netdevice.h>
#include <linux/netfilter_bridge.h>
#include <linux/random.h>
#include <linux/rculist.h>
#include <linux/skbuff.h>
#include <linux/slab.h>
#include <linux/timer.h>
28
#include <linux/inetdevice.h>
29
#include <linux/mroute.h>
30
#include <net/ip.h>
31
#include <net/switchdev.h>
E
Eric Dumazet 已提交
32
#if IS_ENABLED(CONFIG_IPV6)
33
#include <linux/icmpv6.h>
34 35
#include <net/ipv6.h>
#include <net/mld.h>
36
#include <net/ip6_checksum.h>
37
#include <net/addrconf.h>
38
#endif
39 40 41

#include "br_private.h"

42 43 44 45 46 47 48 49
static const struct rhashtable_params br_mdb_rht_params = {
	.head_offset = offsetof(struct net_bridge_mdb_entry, rhnode),
	.key_offset = offsetof(struct net_bridge_mdb_entry, addr),
	.key_len = sizeof(struct br_ip),
	.automatic_shrinking = true,
	.locks_mul = 1,
};

50
static void br_multicast_start_querier(struct net_bridge *br,
51
				       struct bridge_mcast_own_query *query);
52 53
static void br_multicast_add_router(struct net_bridge *br,
				    struct net_bridge_port *port);
54 55 56
static void br_ip4_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
					 __be32 group,
F
Felix Fietkau 已提交
57 58 59
					 __u16 vid,
					 const unsigned char *src);

60
static void __del_port_router(struct net_bridge_port *p);
61 62 63 64
#if IS_ENABLED(CONFIG_IPV6)
static void br_ip6_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
					 const struct in6_addr *group,
F
Felix Fietkau 已提交
65
					 __u16 vid, const unsigned char *src);
66
#endif
67

68 69 70 71
static inline int br_ip_equal(const struct br_ip *a, const struct br_ip *b)
{
	if (a->proto != b->proto)
		return 0;
72 73
	if (a->vid != b->vid)
		return 0;
74 75 76
	switch (a->proto) {
	case htons(ETH_P_IP):
		return a->u.ip4 == b->u.ip4;
E
Eric Dumazet 已提交
77
#if IS_ENABLED(CONFIG_IPV6)
78 79 80
	case htons(ETH_P_IPV6):
		return ipv6_addr_equal(&a->u.ip6, &b->u.ip6);
#endif
81 82 83 84
	}
	return 0;
}

85 86
static struct net_bridge_mdb_entry *br_mdb_ip_get_rcu(struct net_bridge *br,
						      struct br_ip *dst)
87
{
88
	return rhashtable_lookup(&br->mdb_hash_tbl, dst, br_mdb_rht_params);
89 90
}

91 92
struct net_bridge_mdb_entry *br_mdb_ip_get(struct net_bridge *br,
					   struct br_ip *dst)
93
{
94
	struct net_bridge_mdb_entry *ent;
95

96
	lockdep_assert_held_once(&br->multicast_lock);
97

98 99 100
	rcu_read_lock();
	ent = rhashtable_lookup(&br->mdb_hash_tbl, dst, br_mdb_rht_params);
	rcu_read_unlock();
101

102
	return ent;
103 104
}

105 106
static struct net_bridge_mdb_entry *br_mdb_ip4_get(struct net_bridge *br,
						   __be32 dst, __u16 vid)
107
{
108 109
	struct br_ip br_dst;

110
	memset(&br_dst, 0, sizeof(br_dst));
111 112
	br_dst.u.ip4 = dst;
	br_dst.proto = htons(ETH_P_IP);
113
	br_dst.vid = vid;
114

115
	return br_mdb_ip_get(br, &br_dst);
116 117
}

E
Eric Dumazet 已提交
118
#if IS_ENABLED(CONFIG_IPV6)
119 120 121
static struct net_bridge_mdb_entry *br_mdb_ip6_get(struct net_bridge *br,
						   const struct in6_addr *dst,
						   __u16 vid)
122 123
{
	struct br_ip br_dst;
124

125
	memset(&br_dst, 0, sizeof(br_dst));
A
Alexey Dobriyan 已提交
126
	br_dst.u.ip6 = *dst;
127
	br_dst.proto = htons(ETH_P_IPV6);
128
	br_dst.vid = vid;
129

130
	return br_mdb_ip_get(br, &br_dst);
131 132 133
}
#endif

134
struct net_bridge_mdb_entry *br_mdb_get(struct net_bridge *br,
135
					struct sk_buff *skb, u16 vid)
136
{
137 138
	struct br_ip ip;

139
	if (!br_opt_get(br, BROPT_MULTICAST_ENABLED))
140 141
		return NULL;

142
	if (BR_INPUT_SKB_CB(skb)->igmp)
143 144
		return NULL;

145
	memset(&ip, 0, sizeof(ip));
146
	ip.proto = skb->protocol;
147
	ip.vid = vid;
148

149 150
	switch (skb->protocol) {
	case htons(ETH_P_IP):
151 152
		ip.u.ip4 = ip_hdr(skb)->daddr;
		break;
E
Eric Dumazet 已提交
153
#if IS_ENABLED(CONFIG_IPV6)
154
	case htons(ETH_P_IPV6):
A
Alexey Dobriyan 已提交
155
		ip.u.ip6 = ipv6_hdr(skb)->daddr;
156 157
		break;
#endif
158 159
	default:
		return NULL;
160 161
	}

162
	return br_mdb_ip_get_rcu(br, &ip);
163 164
}

165
static void br_multicast_group_expired(struct timer_list *t)
166
{
167
	struct net_bridge_mdb_entry *mp = from_timer(mp, t, timer);
168 169 170 171 172 173
	struct net_bridge *br = mp->br;

	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) || timer_pending(&mp->timer))
		goto out;

174
	mp->host_joined = false;
175
	br_mdb_notify(br->dev, NULL, &mp->addr, RTM_DELMDB, 0);
176 177 178 179

	if (mp->ports)
		goto out;

180 181 182
	rhashtable_remove_fast(&br->mdb_hash_tbl, &mp->rhnode,
			       br_mdb_rht_params);
	hlist_del_rcu(&mp->mdb_node);
183

184
	kfree_rcu(mp, rcu);
185 186 187 188 189 190 191 192 193 194

out:
	spin_unlock(&br->multicast_lock);
}

static void br_multicast_del_pg(struct net_bridge *br,
				struct net_bridge_port_group *pg)
{
	struct net_bridge_mdb_entry *mp;
	struct net_bridge_port_group *p;
195 196
	struct net_bridge_port_group __rcu **pp;

197
	mp = br_mdb_ip_get(br, &pg->addr);
198 199 200
	if (WARN_ON(!mp))
		return;

201 202 203
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
204 205 206
		if (p != pg)
			continue;

207
		rcu_assign_pointer(*pp, p->next);
208 209
		hlist_del_init(&p->mglist);
		del_timer(&p->timer);
210 211
		br_mdb_notify(br->dev, p->port, &pg->addr, RTM_DELMDB,
			      p->flags);
212
		kfree_rcu(p, rcu);
213

214
		if (!mp->ports && !mp->host_joined &&
215 216 217 218 219 220 221 222 223
		    netif_running(br->dev))
			mod_timer(&mp->timer, jiffies);

		return;
	}

	WARN_ON(1);
}

224
static void br_multicast_port_group_expired(struct timer_list *t)
225
{
226
	struct net_bridge_port_group *pg = from_timer(pg, t, timer);
227 228 229 230
	struct net_bridge *br = pg->port->br;

	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) || timer_pending(&pg->timer) ||
231
	    hlist_unhashed(&pg->mglist) || pg->flags & MDB_PG_FLAGS_PERMANENT)
232 233 234 235 236 237 238 239
		goto out;

	br_multicast_del_pg(br, pg);

out:
	spin_unlock(&br->multicast_lock);
}

240
static struct sk_buff *br_ip4_multicast_alloc_query(struct net_bridge *br,
241 242
						    __be32 group,
						    u8 *igmp_type)
243
{
244 245
	struct igmpv3_query *ihv3;
	size_t igmp_hdr_size;
246 247 248 249 250
	struct sk_buff *skb;
	struct igmphdr *ih;
	struct ethhdr *eth;
	struct iphdr *iph;

251 252 253
	igmp_hdr_size = sizeof(*ih);
	if (br->multicast_igmp_version == 3)
		igmp_hdr_size = sizeof(*ihv3);
254
	skb = netdev_alloc_skb_ip_align(br->dev, sizeof(*eth) + sizeof(*iph) +
255
						 igmp_hdr_size + 4);
256 257 258 259 260 261 262 263
	if (!skb)
		goto out;

	skb->protocol = htons(ETH_P_IP);

	skb_reset_mac_header(skb);
	eth = eth_hdr(skb);

264
	ether_addr_copy(eth->h_source, br->dev->dev_addr);
265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
	eth->h_dest[0] = 1;
	eth->h_dest[1] = 0;
	eth->h_dest[2] = 0x5e;
	eth->h_dest[3] = 0;
	eth->h_dest[4] = 0;
	eth->h_dest[5] = 1;
	eth->h_proto = htons(ETH_P_IP);
	skb_put(skb, sizeof(*eth));

	skb_set_network_header(skb, skb->len);
	iph = ip_hdr(skb);

	iph->version = 4;
	iph->ihl = 6;
	iph->tos = 0xc0;
280
	iph->tot_len = htons(sizeof(*iph) + igmp_hdr_size + 4);
281 282 283 284
	iph->id = 0;
	iph->frag_off = htons(IP_DF);
	iph->ttl = 1;
	iph->protocol = IPPROTO_IGMP;
285
	iph->saddr = br_opt_get(br, BROPT_MULTICAST_QUERY_USE_IFADDR) ?
286
		     inet_select_addr(br->dev, 0, RT_SCOPE_LINK) : 0;
287 288 289 290 291 292 293 294 295
	iph->daddr = htonl(INADDR_ALLHOSTS_GROUP);
	((u8 *)&iph[1])[0] = IPOPT_RA;
	((u8 *)&iph[1])[1] = 4;
	((u8 *)&iph[1])[2] = 0;
	((u8 *)&iph[1])[3] = 0;
	ip_send_check(iph);
	skb_put(skb, 24);

	skb_set_transport_header(skb, skb->len);
296
	*igmp_type = IGMP_HOST_MEMBERSHIP_QUERY;
297

298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326
	switch (br->multicast_igmp_version) {
	case 2:
		ih = igmp_hdr(skb);
		ih->type = IGMP_HOST_MEMBERSHIP_QUERY;
		ih->code = (group ? br->multicast_last_member_interval :
				    br->multicast_query_response_interval) /
			   (HZ / IGMP_TIMER_SCALE);
		ih->group = group;
		ih->csum = 0;
		ih->csum = ip_compute_csum((void *)ih, sizeof(*ih));
		break;
	case 3:
		ihv3 = igmpv3_query_hdr(skb);
		ihv3->type = IGMP_HOST_MEMBERSHIP_QUERY;
		ihv3->code = (group ? br->multicast_last_member_interval :
				      br->multicast_query_response_interval) /
			     (HZ / IGMP_TIMER_SCALE);
		ihv3->group = group;
		ihv3->qqic = br->multicast_query_interval / HZ;
		ihv3->nsrcs = 0;
		ihv3->resv = 0;
		ihv3->suppress = 0;
		ihv3->qrv = 2;
		ihv3->csum = 0;
		ihv3->csum = ip_compute_csum((void *)ihv3, sizeof(*ihv3));
		break;
	}

	skb_put(skb, igmp_hdr_size);
327 328 329 330 331 332
	__skb_pull(skb, sizeof(*eth));

out:
	return skb;
}

E
Eric Dumazet 已提交
333
#if IS_ENABLED(CONFIG_IPV6)
334
static struct sk_buff *br_ip6_multicast_alloc_query(struct net_bridge *br,
335 336
						    const struct in6_addr *grp,
						    u8 *igmp_type)
337
{
338 339
	struct mld2_query *mld2q;
	unsigned long interval;
340 341
	struct ipv6hdr *ip6h;
	struct mld_msg *mldq;
342 343
	size_t mld_hdr_size;
	struct sk_buff *skb;
344 345 346
	struct ethhdr *eth;
	u8 *hopopt;

347 348 349
	mld_hdr_size = sizeof(*mldq);
	if (br->multicast_mld_version == 2)
		mld_hdr_size = sizeof(*mld2q);
350
	skb = netdev_alloc_skb_ip_align(br->dev, sizeof(*eth) + sizeof(*ip6h) +
351
						 8 + mld_hdr_size);
352 353 354 355 356 357 358 359 360
	if (!skb)
		goto out;

	skb->protocol = htons(ETH_P_IPV6);

	/* Ethernet header */
	skb_reset_mac_header(skb);
	eth = eth_hdr(skb);

361
	ether_addr_copy(eth->h_source, br->dev->dev_addr);
362 363 364 365 366 367 368 369
	eth->h_proto = htons(ETH_P_IPV6);
	skb_put(skb, sizeof(*eth));

	/* IPv6 header + HbH option */
	skb_set_network_header(skb, skb->len);
	ip6h = ipv6_hdr(skb);

	*(__force __be32 *)ip6h = htonl(0x60000000);
370
	ip6h->payload_len = htons(8 + mld_hdr_size);
371 372
	ip6h->nexthdr = IPPROTO_HOPOPTS;
	ip6h->hop_limit = 1;
373
	ipv6_addr_set(&ip6h->daddr, htonl(0xff020000), 0, 0, htonl(1));
374 375 376
	if (ipv6_dev_get_saddr(dev_net(br->dev), br->dev, &ip6h->daddr, 0,
			       &ip6h->saddr)) {
		kfree_skb(skb);
377
		br_opt_toggle(br, BROPT_HAS_IPV6_ADDR, false);
378 379
		return NULL;
	}
380

381
	br_opt_toggle(br, BROPT_HAS_IPV6_ADDR, true);
382
	ipv6_eth_mc_map(&ip6h->daddr, eth->h_dest);
383 384 385 386 387 388 389 390

	hopopt = (u8 *)(ip6h + 1);
	hopopt[0] = IPPROTO_ICMPV6;		/* next hdr */
	hopopt[1] = 0;				/* length of HbH */
	hopopt[2] = IPV6_TLV_ROUTERALERT;	/* Router Alert */
	hopopt[3] = 2;				/* Length of RA Option */
	hopopt[4] = 0;				/* Type = 0x0000 (MLD) */
	hopopt[5] = 0;
391 392
	hopopt[6] = IPV6_TLV_PAD1;		/* Pad1 */
	hopopt[7] = IPV6_TLV_PAD1;		/* Pad1 */
393 394 395 396 397

	skb_put(skb, sizeof(*ip6h) + 8);

	/* ICMPv6 */
	skb_set_transport_header(skb, skb->len);
398
	interval = ipv6_addr_any(grp) ?
399 400
			br->multicast_query_response_interval :
			br->multicast_last_member_interval;
401
	*igmp_type = ICMPV6_MGM_QUERY;
402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418
	switch (br->multicast_mld_version) {
	case 1:
		mldq = (struct mld_msg *)icmp6_hdr(skb);
		mldq->mld_type = ICMPV6_MGM_QUERY;
		mldq->mld_code = 0;
		mldq->mld_cksum = 0;
		mldq->mld_maxdelay = htons((u16)jiffies_to_msecs(interval));
		mldq->mld_reserved = 0;
		mldq->mld_mca = *grp;
		mldq->mld_cksum = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
						  sizeof(*mldq), IPPROTO_ICMPV6,
						  csum_partial(mldq,
							       sizeof(*mldq),
							       0));
		break;
	case 2:
		mld2q = (struct mld2_query *)icmp6_hdr(skb);
419
		mld2q->mld2q_mrc = htons((u16)jiffies_to_msecs(interval));
420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438
		mld2q->mld2q_type = ICMPV6_MGM_QUERY;
		mld2q->mld2q_code = 0;
		mld2q->mld2q_cksum = 0;
		mld2q->mld2q_resv1 = 0;
		mld2q->mld2q_resv2 = 0;
		mld2q->mld2q_suppress = 0;
		mld2q->mld2q_qrv = 2;
		mld2q->mld2q_nsrcs = 0;
		mld2q->mld2q_qqic = br->multicast_query_interval / HZ;
		mld2q->mld2q_mca = *grp;
		mld2q->mld2q_cksum = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
						     sizeof(*mld2q),
						     IPPROTO_ICMPV6,
						     csum_partial(mld2q,
								  sizeof(*mld2q),
								  0));
		break;
	}
	skb_put(skb, mld_hdr_size);
439 440 441 442 443 444 445 446

	__skb_pull(skb, sizeof(*eth));

out:
	return skb;
}
#endif

447
static struct sk_buff *br_multicast_alloc_query(struct net_bridge *br,
448 449
						struct br_ip *addr,
						u8 *igmp_type)
450 451 452
{
	switch (addr->proto) {
	case htons(ETH_P_IP):
453
		return br_ip4_multicast_alloc_query(br, addr->u.ip4, igmp_type);
E
Eric Dumazet 已提交
454
#if IS_ENABLED(CONFIG_IPV6)
455
	case htons(ETH_P_IPV6):
456 457
		return br_ip6_multicast_alloc_query(br, &addr->u.ip6,
						    igmp_type);
458
#endif
459 460 461 462
	}
	return NULL;
}

463
struct net_bridge_mdb_entry *br_multicast_new_group(struct net_bridge *br,
464
						    struct br_ip *group)
465 466
{
	struct net_bridge_mdb_entry *mp;
467
	int err;
468

469 470 471
	mp = br_mdb_ip_get(br, group);
	if (mp)
		return mp;
472

473 474 475
	if (atomic_read(&br->mdb_hash_tbl.nelems) >= br->hash_max) {
		br_opt_toggle(br, BROPT_MULTICAST_ENABLED, false);
		return ERR_PTR(-E2BIG);
476 477 478 479
	}

	mp = kzalloc(sizeof(*mp), GFP_ATOMIC);
	if (unlikely(!mp))
480
		return ERR_PTR(-ENOMEM);
481 482

	mp->br = br;
483
	mp->addr = *group;
484
	timer_setup(&mp->timer, br_multicast_group_expired, 0);
485 486 487 488 489 490 491 492
	err = rhashtable_lookup_insert_fast(&br->mdb_hash_tbl, &mp->rhnode,
					    br_mdb_rht_params);
	if (err) {
		kfree(mp);
		mp = ERR_PTR(err);
	} else {
		hlist_add_head_rcu(&mp->mdb_node, &br->mdb_list);
	}
493

494 495 496
	return mp;
}

497 498 499
struct net_bridge_port_group *br_multicast_new_port_group(
			struct net_bridge_port *port,
			struct br_ip *group,
500
			struct net_bridge_port_group __rcu *next,
F
Felix Fietkau 已提交
501 502
			unsigned char flags,
			const unsigned char *src)
503 504 505 506 507 508 509 510 511
{
	struct net_bridge_port_group *p;

	p = kzalloc(sizeof(*p), GFP_ATOMIC);
	if (unlikely(!p))
		return NULL;

	p->addr = *group;
	p->port = port;
512
	p->flags = flags;
513
	rcu_assign_pointer(p->next, next);
514
	hlist_add_head(&p->mglist, &port->mglist);
515
	timer_setup(&p->timer, br_multicast_port_group_expired, 0);
F
Felix Fietkau 已提交
516 517 518 519 520 521

	if (src)
		memcpy(p->eth_addr, src, ETH_ALEN);
	else
		memset(p->eth_addr, 0xff, ETH_ALEN);

522 523 524
	return p;
}

F
Felix Fietkau 已提交
525 526 527 528 529 530 531 532 533 534 535 536 537
static bool br_port_group_equal(struct net_bridge_port_group *p,
				struct net_bridge_port *port,
				const unsigned char *src)
{
	if (p->port != port)
		return false;

	if (!(port->flags & BR_MULTICAST_TO_UNICAST))
		return true;

	return ether_addr_equal(src, p->eth_addr);
}

538
static int br_multicast_add_group(struct net_bridge *br,
539
				  struct net_bridge_port *port,
F
Felix Fietkau 已提交
540 541
				  struct br_ip *group,
				  const unsigned char *src)
542
{
543
	struct net_bridge_port_group __rcu **pp;
544 545
	struct net_bridge_port_group *p;
	struct net_bridge_mdb_entry *mp;
546
	unsigned long now = jiffies;
547 548 549 550 551 552 553
	int err;

	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) ||
	    (port && port->state == BR_STATE_DISABLED))
		goto out;

554
	mp = br_multicast_new_group(br, group);
555
	err = PTR_ERR(mp);
556
	if (IS_ERR(mp))
557 558 559
		goto err;

	if (!port) {
560 561 562 563
		if (!mp->host_joined) {
			mp->host_joined = true;
			br_mdb_notify(br->dev, NULL, &mp->addr, RTM_NEWMDB, 0);
		}
564
		mod_timer(&mp->timer, now + br->multicast_membership_interval);
565 566 567
		goto out;
	}

568 569 570
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
F
Felix Fietkau 已提交
571
		if (br_port_group_equal(p, port, src))
572
			goto found;
573 574 575 576
		if ((unsigned long)p->port < (unsigned long)port)
			break;
	}

F
Felix Fietkau 已提交
577
	p = br_multicast_new_port_group(port, group, *pp, 0, src);
578 579 580
	if (unlikely(!p))
		goto err;
	rcu_assign_pointer(*pp, p);
581
	br_mdb_notify(br->dev, port, group, RTM_NEWMDB, 0);
582

583 584
found:
	mod_timer(&p->timer, now + br->multicast_membership_interval);
585 586 587 588 589 590 591 592
out:
	err = 0;

err:
	spin_unlock(&br->multicast_lock);
	return err;
}

593 594
static int br_ip4_multicast_add_group(struct net_bridge *br,
				      struct net_bridge_port *port,
595
				      __be32 group,
F
Felix Fietkau 已提交
596 597
				      __u16 vid,
				      const unsigned char *src)
598 599 600 601 602 603 604 605
{
	struct br_ip br_group;

	if (ipv4_is_local_multicast(group))
		return 0;

	br_group.u.ip4 = group;
	br_group.proto = htons(ETH_P_IP);
606
	br_group.vid = vid;
607

F
Felix Fietkau 已提交
608
	return br_multicast_add_group(br, port, &br_group, src);
609 610
}

E
Eric Dumazet 已提交
611
#if IS_ENABLED(CONFIG_IPV6)
612 613
static int br_ip6_multicast_add_group(struct net_bridge *br,
				      struct net_bridge_port *port,
614
				      const struct in6_addr *group,
F
Felix Fietkau 已提交
615 616
				      __u16 vid,
				      const unsigned char *src)
617 618 619
{
	struct br_ip br_group;

620
	if (ipv6_addr_is_ll_all_nodes(group))
621 622
		return 0;

623
	memset(&br_group, 0, sizeof(br_group));
A
Alexey Dobriyan 已提交
624
	br_group.u.ip6 = *group;
625
	br_group.proto = htons(ETH_P_IPV6);
626
	br_group.vid = vid;
627

F
Felix Fietkau 已提交
628
	return br_multicast_add_group(br, port, &br_group, src);
629 630 631
}
#endif

632
static void br_multicast_router_expired(struct timer_list *t)
633
{
634 635
	struct net_bridge_port *port =
			from_timer(port, t, multicast_router_timer);
636 637 638
	struct net_bridge *br = port->br;

	spin_lock(&br->multicast_lock);
639 640
	if (port->multicast_router == MDB_RTR_TYPE_DISABLED ||
	    port->multicast_router == MDB_RTR_TYPE_PERM ||
641
	    timer_pending(&port->multicast_router_timer))
642 643
		goto out;

644
	__del_port_router(port);
645 646 647 648
out:
	spin_unlock(&br->multicast_lock);
}

649 650 651 652 653 654 655 656 657 658 659 660 661
static void br_mc_router_state_change(struct net_bridge *p,
				      bool is_mc_router)
{
	struct switchdev_attr attr = {
		.orig_dev = p->dev,
		.id = SWITCHDEV_ATTR_ID_BRIDGE_MROUTER,
		.flags = SWITCHDEV_F_DEFER,
		.u.mrouter = is_mc_router,
	};

	switchdev_port_attr_set(p->dev, &attr);
}

662
static void br_multicast_local_router_expired(struct timer_list *t)
663
{
664
	struct net_bridge *br = from_timer(br, t, multicast_router_timer);
665 666 667 668 669 670 671 672 673 674

	spin_lock(&br->multicast_lock);
	if (br->multicast_router == MDB_RTR_TYPE_DISABLED ||
	    br->multicast_router == MDB_RTR_TYPE_PERM ||
	    timer_pending(&br->multicast_router_timer))
		goto out;

	br_mc_router_state_change(br, false);
out:
	spin_unlock(&br->multicast_lock);
675 676
}

677
static void br_multicast_querier_expired(struct net_bridge *br,
678
					 struct bridge_mcast_own_query *query)
679 680
{
	spin_lock(&br->multicast_lock);
681
	if (!netif_running(br->dev) || !br_opt_get(br, BROPT_MULTICAST_ENABLED))
682 683
		goto out;

684
	br_multicast_start_querier(br, query);
685 686 687 688 689

out:
	spin_unlock(&br->multicast_lock);
}

690
static void br_ip4_multicast_querier_expired(struct timer_list *t)
691
{
692
	struct net_bridge *br = from_timer(br, t, ip4_other_query.timer);
693

694
	br_multicast_querier_expired(br, &br->ip4_own_query);
695 696 697
}

#if IS_ENABLED(CONFIG_IPV6)
698
static void br_ip6_multicast_querier_expired(struct timer_list *t)
699
{
700
	struct net_bridge *br = from_timer(br, t, ip6_other_query.timer);
701

702
	br_multicast_querier_expired(br, &br->ip6_own_query);
703 704 705
}
#endif

706 707 708 709 710 711 712 713 714 715 716 717
static void br_multicast_select_own_querier(struct net_bridge *br,
					    struct br_ip *ip,
					    struct sk_buff *skb)
{
	if (ip->proto == htons(ETH_P_IP))
		br->ip4_querier.addr.u.ip4 = ip_hdr(skb)->saddr;
#if IS_ENABLED(CONFIG_IPV6)
	else
		br->ip6_querier.addr.u.ip6 = ipv6_hdr(skb)->saddr;
#endif
}

718 719 720
static void __br_multicast_send_query(struct net_bridge *br,
				      struct net_bridge_port *port,
				      struct br_ip *ip)
721 722
{
	struct sk_buff *skb;
723
	u8 igmp_type;
724

725
	skb = br_multicast_alloc_query(br, ip, &igmp_type);
726
	if (!skb)
727
		return;
728 729 730

	if (port) {
		skb->dev = port->dev;
731
		br_multicast_count(br, port, skb, igmp_type,
732
				   BR_MCAST_DIR_TX);
733 734
		NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_OUT,
			dev_net(port->dev), NULL, skb, NULL, skb->dev,
735
			br_dev_queue_push_xmit);
736 737
	} else {
		br_multicast_select_own_querier(br, ip, skb);
738
		br_multicast_count(br, port, skb, igmp_type,
739
				   BR_MCAST_DIR_RX);
740
		netif_rx(skb);
741
	}
742 743 744
}

static void br_multicast_send_query(struct net_bridge *br,
745
				    struct net_bridge_port *port,
746
				    struct bridge_mcast_own_query *own_query)
747
{
748
	struct bridge_mcast_other_query *other_query = NULL;
749 750
	struct br_ip br_group;
	unsigned long time;
751

752 753
	if (!netif_running(br->dev) ||
	    !br_opt_get(br, BROPT_MULTICAST_ENABLED) ||
754
	    !br_opt_get(br, BROPT_MULTICAST_QUERIER))
755 756
		return;

757 758
	memset(&br_group.u, 0, sizeof(br_group.u));

759 760 761
	if (port ? (own_query == &port->ip4_own_query) :
		   (own_query == &br->ip4_own_query)) {
		other_query = &br->ip4_other_query;
762
		br_group.proto = htons(ETH_P_IP);
E
Eric Dumazet 已提交
763
#if IS_ENABLED(CONFIG_IPV6)
764
	} else {
765
		other_query = &br->ip6_other_query;
766
		br_group.proto = htons(ETH_P_IPV6);
767
#endif
768 769
	}

770
	if (!other_query || timer_pending(&other_query->timer))
771 772 773
		return;

	__br_multicast_send_query(br, port, &br_group);
774 775

	time = jiffies;
776
	time += own_query->startup_sent < br->multicast_startup_query_count ?
777 778
		br->multicast_startup_query_interval :
		br->multicast_query_interval;
779
	mod_timer(&own_query->timer, time);
780 781
}

782 783 784
static void
br_multicast_port_query_expired(struct net_bridge_port *port,
				struct bridge_mcast_own_query *query)
785 786 787 788
{
	struct net_bridge *br = port->br;

	spin_lock(&br->multicast_lock);
789 790
	if (port->state == BR_STATE_DISABLED ||
	    port->state == BR_STATE_BLOCKING)
791 792
		goto out;

793 794
	if (query->startup_sent < br->multicast_startup_query_count)
		query->startup_sent++;
795

796
	br_multicast_send_query(port->br, port, query);
797 798 799 800 801

out:
	spin_unlock(&br->multicast_lock);
}

802
static void br_ip4_multicast_port_query_expired(struct timer_list *t)
803
{
804
	struct net_bridge_port *port = from_timer(port, t, ip4_own_query.timer);
805

806
	br_multicast_port_query_expired(port, &port->ip4_own_query);
807 808 809
}

#if IS_ENABLED(CONFIG_IPV6)
810
static void br_ip6_multicast_port_query_expired(struct timer_list *t)
811
{
812
	struct net_bridge_port *port = from_timer(port, t, ip6_own_query.timer);
813

814
	br_multicast_port_query_expired(port, &port->ip6_own_query);
815 816 817
}
#endif

818 819 820 821 822 823
static void br_mc_disabled_update(struct net_device *dev, bool value)
{
	struct switchdev_attr attr = {
		.orig_dev = dev,
		.id = SWITCHDEV_ATTR_ID_BRIDGE_MC_DISABLED,
		.flags = SWITCHDEV_F_DEFER,
824
		.u.mc_disabled = !value,
825 826 827 828 829
	};

	switchdev_port_attr_set(dev, &attr);
}

830
int br_multicast_add_port(struct net_bridge_port *port)
831
{
832
	port->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
833

834 835 836 837
	timer_setup(&port->multicast_router_timer,
		    br_multicast_router_expired, 0);
	timer_setup(&port->ip4_own_query.timer,
		    br_ip4_multicast_port_query_expired, 0);
838
#if IS_ENABLED(CONFIG_IPV6)
839 840
	timer_setup(&port->ip6_own_query.timer,
		    br_ip6_multicast_port_query_expired, 0);
841
#endif
842 843
	br_mc_disabled_update(port->dev,
			      br_opt_get(port->br, BROPT_MULTICAST_ENABLED));
844

845 846 847 848 849
	port->mcast_stats = netdev_alloc_pcpu_stats(struct bridge_mcast_stats);
	if (!port->mcast_stats)
		return -ENOMEM;

	return 0;
850 851 852 853
}

void br_multicast_del_port(struct net_bridge_port *port)
{
854 855 856 857 858 859 860 861 862
	struct net_bridge *br = port->br;
	struct net_bridge_port_group *pg;
	struct hlist_node *n;

	/* Take care of the remaining groups, only perm ones should be left */
	spin_lock_bh(&br->multicast_lock);
	hlist_for_each_entry_safe(pg, n, &port->mglist, mglist)
		br_multicast_del_pg(br, pg);
	spin_unlock_bh(&br->multicast_lock);
863
	del_timer_sync(&port->multicast_router_timer);
864
	free_percpu(port->mcast_stats);
865 866
}

867
static void br_multicast_enable(struct bridge_mcast_own_query *query)
868
{
869
	query->startup_sent = 0;
870

871 872 873
	if (try_to_del_timer_sync(&query->timer) >= 0 ||
	    del_timer(&query->timer))
		mod_timer(&query->timer, jiffies);
874 875
}

876
static void __br_multicast_enable_port(struct net_bridge_port *port)
877 878 879
{
	struct net_bridge *br = port->br;

880
	if (!br_opt_get(br, BROPT_MULTICAST_ENABLED) || !netif_running(br->dev))
881
		return;
882

883
	br_multicast_enable(&port->ip4_own_query);
884
#if IS_ENABLED(CONFIG_IPV6)
885
	br_multicast_enable(&port->ip6_own_query);
886
#endif
887 888
	if (port->multicast_router == MDB_RTR_TYPE_PERM &&
	    hlist_unhashed(&port->rlist))
889
		br_multicast_add_router(br, port);
890
}
891

892 893 894 895 896 897
void br_multicast_enable_port(struct net_bridge_port *port)
{
	struct net_bridge *br = port->br;

	spin_lock(&br->multicast_lock);
	__br_multicast_enable_port(port);
898 899 900 901 902 903 904
	spin_unlock(&br->multicast_lock);
}

void br_multicast_disable_port(struct net_bridge_port *port)
{
	struct net_bridge *br = port->br;
	struct net_bridge_port_group *pg;
905
	struct hlist_node *n;
906 907

	spin_lock(&br->multicast_lock);
908
	hlist_for_each_entry_safe(pg, n, &port->mglist, mglist)
909
		if (!(pg->flags & MDB_PG_FLAGS_PERMANENT))
910
			br_multicast_del_pg(br, pg);
911

912 913
	__del_port_router(port);

914
	del_timer(&port->multicast_router_timer);
915
	del_timer(&port->ip4_own_query.timer);
916
#if IS_ENABLED(CONFIG_IPV6)
917
	del_timer(&port->ip6_own_query.timer);
918
#endif
919 920 921
	spin_unlock(&br->multicast_lock);
}

922 923
static int br_ip4_multicast_igmp3_report(struct net_bridge *br,
					 struct net_bridge_port *port,
924 925
					 struct sk_buff *skb,
					 u16 vid)
926
{
F
Felix Fietkau 已提交
927
	const unsigned char *src;
928 929 930 931 932 933 934 935 936 937 938
	struct igmpv3_report *ih;
	struct igmpv3_grec *grec;
	int i;
	int len;
	int num;
	int type;
	int err = 0;
	__be32 group;

	ih = igmpv3_report_hdr(skb);
	num = ntohs(ih->ngrec);
939
	len = skb_transport_offset(skb) + sizeof(*ih);
940 941 942

	for (i = 0; i < num; i++) {
		len += sizeof(*grec);
943
		if (!ip_mc_may_pull(skb, len))
944 945
			return -EINVAL;

H
Herbert Xu 已提交
946
		grec = (void *)(skb->data + len - sizeof(*grec));
947 948 949
		group = grec->grec_mca;
		type = grec->grec_type;

E
Eric Dumazet 已提交
950
		len += ntohs(grec->grec_nsrcs) * 4;
951
		if (!ip_mc_may_pull(skb, len))
952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967
			return -EINVAL;

		/* We treat this as an IGMPv2 report for now. */
		switch (type) {
		case IGMPV3_MODE_IS_INCLUDE:
		case IGMPV3_MODE_IS_EXCLUDE:
		case IGMPV3_CHANGE_TO_INCLUDE:
		case IGMPV3_CHANGE_TO_EXCLUDE:
		case IGMPV3_ALLOW_NEW_SOURCES:
		case IGMPV3_BLOCK_OLD_SOURCES:
			break;

		default:
			continue;
		}

F
Felix Fietkau 已提交
968
		src = eth_hdr(skb)->h_source;
969 970 971
		if ((type == IGMPV3_CHANGE_TO_INCLUDE ||
		     type == IGMPV3_MODE_IS_INCLUDE) &&
		    ntohs(grec->grec_nsrcs) == 0) {
F
Felix Fietkau 已提交
972
			br_ip4_multicast_leave_group(br, port, group, vid, src);
973
		} else {
F
Felix Fietkau 已提交
974 975
			err = br_ip4_multicast_add_group(br, port, group, vid,
							 src);
976 977 978
			if (err)
				break;
		}
979 980 981 982 983
	}

	return err;
}

E
Eric Dumazet 已提交
984
#if IS_ENABLED(CONFIG_IPV6)
985 986
static int br_ip6_multicast_mld2_report(struct net_bridge *br,
					struct net_bridge_port *port,
987 988
					struct sk_buff *skb,
					u16 vid)
989
{
990
	unsigned int nsrcs_offset;
F
Felix Fietkau 已提交
991
	const unsigned char *src;
992 993
	struct icmp6hdr *icmp6h;
	struct mld2_grec *grec;
994
	unsigned int grec_len;
995 996 997 998 999
	int i;
	int len;
	int num;
	int err = 0;

1000
	if (!ipv6_mc_may_pull(skb, sizeof(*icmp6h)))
1001 1002 1003 1004
		return -EINVAL;

	icmp6h = icmp6_hdr(skb);
	num = ntohs(icmp6h->icmp6_dataun.un_data16[1]);
1005
	len = skb_transport_offset(skb) + sizeof(*icmp6h);
1006 1007 1008 1009

	for (i = 0; i < num; i++) {
		__be16 *nsrcs, _nsrcs;

1010 1011 1012 1013 1014 1015 1016
		nsrcs_offset = len + offsetof(struct mld2_grec, grec_nsrcs);

		if (skb_transport_offset(skb) + ipv6_transport_len(skb) <
		    nsrcs_offset + sizeof(_nsrcs))
			return -EINVAL;

		nsrcs = skb_header_pointer(skb, nsrcs_offset,
1017 1018 1019 1020
					   sizeof(_nsrcs), &_nsrcs);
		if (!nsrcs)
			return -EINVAL;

1021 1022 1023 1024
		grec_len = sizeof(*grec) +
			   sizeof(struct in6_addr) * ntohs(*nsrcs);

		if (!ipv6_mc_may_pull(skb, len + grec_len))
1025 1026 1027
			return -EINVAL;

		grec = (struct mld2_grec *)(skb->data + len);
1028
		len += grec_len;
1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043

		/* We treat these as MLDv1 reports for now. */
		switch (grec->grec_type) {
		case MLD2_MODE_IS_INCLUDE:
		case MLD2_MODE_IS_EXCLUDE:
		case MLD2_CHANGE_TO_INCLUDE:
		case MLD2_CHANGE_TO_EXCLUDE:
		case MLD2_ALLOW_NEW_SOURCES:
		case MLD2_BLOCK_OLD_SOURCES:
			break;

		default:
			continue;
		}

F
Felix Fietkau 已提交
1044
		src = eth_hdr(skb)->h_source;
1045 1046 1047 1048
		if ((grec->grec_type == MLD2_CHANGE_TO_INCLUDE ||
		     grec->grec_type == MLD2_MODE_IS_INCLUDE) &&
		    ntohs(*nsrcs) == 0) {
			br_ip6_multicast_leave_group(br, port, &grec->grec_mca,
F
Felix Fietkau 已提交
1049
						     vid, src);
1050 1051
		} else {
			err = br_ip6_multicast_add_group(br, port,
F
Felix Fietkau 已提交
1052 1053
							 &grec->grec_mca, vid,
							 src);
1054
			if (err)
1055 1056
				break;
		}
1057 1058 1059 1060 1061 1062
	}

	return err;
}
#endif

1063
static bool br_ip4_multicast_select_querier(struct net_bridge *br,
1064
					    struct net_bridge_port *port,
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081
					    __be32 saddr)
{
	if (!timer_pending(&br->ip4_own_query.timer) &&
	    !timer_pending(&br->ip4_other_query.timer))
		goto update;

	if (!br->ip4_querier.addr.u.ip4)
		goto update;

	if (ntohl(saddr) <= ntohl(br->ip4_querier.addr.u.ip4))
		goto update;

	return false;

update:
	br->ip4_querier.addr.u.ip4 = saddr;

1082 1083 1084
	/* update protected by general multicast_lock by caller */
	rcu_assign_pointer(br->ip4_querier.port, port);

1085 1086 1087 1088 1089
	return true;
}

#if IS_ENABLED(CONFIG_IPV6)
static bool br_ip6_multicast_select_querier(struct net_bridge *br,
1090
					    struct net_bridge_port *port,
1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104
					    struct in6_addr *saddr)
{
	if (!timer_pending(&br->ip6_own_query.timer) &&
	    !timer_pending(&br->ip6_other_query.timer))
		goto update;

	if (ipv6_addr_cmp(saddr, &br->ip6_querier.addr.u.ip6) <= 0)
		goto update;

	return false;

update:
	br->ip6_querier.addr.u.ip6 = *saddr;

1105 1106 1107
	/* update protected by general multicast_lock by caller */
	rcu_assign_pointer(br->ip6_querier.port, port);

1108 1109 1110 1111 1112
	return true;
}
#endif

static bool br_multicast_select_querier(struct net_bridge *br,
1113
					struct net_bridge_port *port,
1114 1115 1116 1117
					struct br_ip *saddr)
{
	switch (saddr->proto) {
	case htons(ETH_P_IP):
1118
		return br_ip4_multicast_select_querier(br, port, saddr->u.ip4);
1119 1120
#if IS_ENABLED(CONFIG_IPV6)
	case htons(ETH_P_IPV6):
1121
		return br_ip6_multicast_select_querier(br, port, &saddr->u.ip6);
1122 1123 1124 1125 1126 1127
#endif
	}

	return false;
}

1128
static void
1129 1130 1131
br_multicast_update_query_timer(struct net_bridge *br,
				struct bridge_mcast_other_query *query,
				unsigned long max_delay)
1132
{
1133 1134
	if (!timer_pending(&query->timer))
		query->delay_time = jiffies + max_delay;
1135

1136
	mod_timer(&query->timer, jiffies + br->multicast_querier_interval);
1137 1138
}

1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
static void br_port_mc_router_state_change(struct net_bridge_port *p,
					   bool is_mc_router)
{
	struct switchdev_attr attr = {
		.orig_dev = p->dev,
		.id = SWITCHDEV_ATTR_ID_PORT_MROUTER,
		.flags = SWITCHDEV_F_DEFER,
		.u.mrouter = is_mc_router,
	};

	switchdev_port_attr_set(p->dev, &attr);
}

1152
/*
C
Cong Wang 已提交
1153
 * Add port to router_list
1154 1155 1156
 *  list is maintained ordered by pointer value
 *  and locked by br->multicast_lock and RCU
 */
1157 1158 1159
static void br_multicast_add_router(struct net_bridge *br,
				    struct net_bridge_port *port)
{
1160
	struct net_bridge_port *p;
1161
	struct hlist_node *slot = NULL;
1162

1163 1164 1165
	if (!hlist_unhashed(&port->rlist))
		return;

1166
	hlist_for_each_entry(p, &br->router_list, rlist) {
1167 1168
		if ((unsigned long) port >= (unsigned long) p)
			break;
1169
		slot = &p->rlist;
1170 1171
	}

1172
	if (slot)
1173
		hlist_add_behind_rcu(&port->rlist, slot);
1174 1175
	else
		hlist_add_head_rcu(&port->rlist, &br->router_list);
1176
	br_rtr_notify(br->dev, port, RTM_NEWMDB);
1177
	br_port_mc_router_state_change(port, true);
1178 1179
}

1180 1181 1182 1183 1184 1185
static void br_multicast_mark_router(struct net_bridge *br,
				     struct net_bridge_port *port)
{
	unsigned long now = jiffies;

	if (!port) {
1186 1187 1188
		if (br->multicast_router == MDB_RTR_TYPE_TEMP_QUERY) {
			if (!timer_pending(&br->multicast_router_timer))
				br_mc_router_state_change(br, true);
1189 1190
			mod_timer(&br->multicast_router_timer,
				  now + br->multicast_querier_interval);
1191
		}
1192 1193 1194
		return;
	}

1195 1196
	if (port->multicast_router == MDB_RTR_TYPE_DISABLED ||
	    port->multicast_router == MDB_RTR_TYPE_PERM)
1197 1198
		return;

1199
	br_multicast_add_router(br, port);
1200 1201 1202 1203 1204 1205 1206

	mod_timer(&port->multicast_router_timer,
		  now + br->multicast_querier_interval);
}

static void br_multicast_query_received(struct net_bridge *br,
					struct net_bridge_port *port,
1207
					struct bridge_mcast_other_query *query,
1208
					struct br_ip *saddr,
1209
					unsigned long max_delay)
1210
{
1211
	if (!br_multicast_select_querier(br, port, saddr))
1212 1213
		return;

1214
	br_multicast_update_query_timer(br, query, max_delay);
1215 1216 1217 1218 1219 1220

	/* Based on RFC4541, section 2.1.1 IGMP Forwarding Rules,
	 * the arrival port for IGMP Queries where the source address
	 * is 0.0.0.0 should not be added to router port list.
	 */
	if ((saddr->proto == htons(ETH_P_IP) && saddr->u.ip4) ||
1221
	    saddr->proto == htons(ETH_P_IPV6))
1222
		br_multicast_mark_router(br, port);
1223 1224
}

1225 1226 1227 1228
static void br_ip4_multicast_query(struct net_bridge *br,
				   struct net_bridge_port *port,
				   struct sk_buff *skb,
				   u16 vid)
1229
{
1230
	unsigned int transport_len = ip_transport_len(skb);
1231
	const struct iphdr *iph = ip_hdr(skb);
1232 1233 1234 1235
	struct igmphdr *ih = igmp_hdr(skb);
	struct net_bridge_mdb_entry *mp;
	struct igmpv3_query *ih3;
	struct net_bridge_port_group *p;
1236
	struct net_bridge_port_group __rcu **pp;
1237
	struct br_ip saddr;
1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248
	unsigned long max_delay;
	unsigned long now = jiffies;
	__be32 group;

	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) ||
	    (port && port->state == BR_STATE_DISABLED))
		goto out;

	group = ih->group;

1249
	if (transport_len == sizeof(*ih)) {
1250 1251 1252 1253 1254 1255
		max_delay = ih->code * (HZ / IGMP_TIMER_SCALE);

		if (!max_delay) {
			max_delay = 10 * HZ;
			group = 0;
		}
1256
	} else if (transport_len >= sizeof(*ih3)) {
1257 1258
		ih3 = igmpv3_query_hdr(skb);
		if (ih3->nsrcs)
1259
			goto out;
1260

1261 1262
		max_delay = ih3->code ?
			    IGMPV3_MRC(ih3->code) * (HZ / IGMP_TIMER_SCALE) : 1;
1263
	} else {
1264 1265 1266
		goto out;
	}

1267 1268 1269
	if (!group) {
		saddr.proto = htons(ETH_P_IP);
		saddr.u.ip4 = iph->saddr;
1270

1271 1272
		br_multicast_query_received(br, port, &br->ip4_other_query,
					    &saddr, max_delay);
1273
		goto out;
1274
	}
1275

1276
	mp = br_mdb_ip4_get(br, group, vid);
1277 1278 1279 1280 1281
	if (!mp)
		goto out;

	max_delay *= br->multicast_last_member_count;

1282
	if (mp->host_joined &&
1283 1284 1285 1286 1287
	    (timer_pending(&mp->timer) ?
	     time_after(mp->timer.expires, now + max_delay) :
	     try_to_del_timer_sync(&mp->timer) >= 0))
		mod_timer(&mp->timer, now + max_delay);

1288 1289 1290
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
1291 1292 1293
		if (timer_pending(&p->timer) ?
		    time_after(p->timer.expires, now + max_delay) :
		    try_to_del_timer_sync(&p->timer) >= 0)
1294
			mod_timer(&p->timer, now + max_delay);
1295 1296 1297 1298 1299 1300
	}

out:
	spin_unlock(&br->multicast_lock);
}

E
Eric Dumazet 已提交
1301
#if IS_ENABLED(CONFIG_IPV6)
1302 1303
static int br_ip6_multicast_query(struct net_bridge *br,
				  struct net_bridge_port *port,
1304 1305
				  struct sk_buff *skb,
				  u16 vid)
1306
{
1307
	unsigned int transport_len = ipv6_transport_len(skb);
1308
	const struct ipv6hdr *ip6h = ipv6_hdr(skb);
1309
	struct mld_msg *mld;
1310 1311
	struct net_bridge_mdb_entry *mp;
	struct mld2_query *mld2q;
1312 1313
	struct net_bridge_port_group *p;
	struct net_bridge_port_group __rcu **pp;
1314
	struct br_ip saddr;
1315 1316
	unsigned long max_delay;
	unsigned long now = jiffies;
1317
	unsigned int offset = skb_transport_offset(skb);
1318
	const struct in6_addr *group = NULL;
1319
	bool is_general_query;
1320 1321 1322 1323 1324 1325 1326
	int err = 0;

	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) ||
	    (port && port->state == BR_STATE_DISABLED))
		goto out;

1327
	if (transport_len == sizeof(*mld)) {
1328
		if (!pskb_may_pull(skb, offset + sizeof(*mld))) {
1329 1330 1331 1332
			err = -EINVAL;
			goto out;
		}
		mld = (struct mld_msg *) icmp6_hdr(skb);
L
Li RongQing 已提交
1333
		max_delay = msecs_to_jiffies(ntohs(mld->mld_maxdelay));
1334 1335
		if (max_delay)
			group = &mld->mld_mca;
1336
	} else {
1337
		if (!pskb_may_pull(skb, offset + sizeof(*mld2q))) {
1338 1339 1340 1341 1342 1343
			err = -EINVAL;
			goto out;
		}
		mld2q = (struct mld2_query *)icmp6_hdr(skb);
		if (!mld2q->mld2q_nsrcs)
			group = &mld2q->mld2q_mca;
1344 1345

		max_delay = max(msecs_to_jiffies(mldv2_mrc(mld2q)), 1UL);
1346 1347
	}

1348 1349
	is_general_query = group && ipv6_addr_any(group);

1350 1351 1352
	if (is_general_query) {
		saddr.proto = htons(ETH_P_IPV6);
		saddr.u.ip6 = ip6h->saddr;
1353

1354 1355
		br_multicast_query_received(br, port, &br->ip6_other_query,
					    &saddr, max_delay);
1356
		goto out;
1357 1358
	} else if (!group) {
		goto out;
1359
	}
1360

1361
	mp = br_mdb_ip6_get(br, group, vid);
1362 1363 1364 1365
	if (!mp)
		goto out;

	max_delay *= br->multicast_last_member_count;
1366
	if (mp->host_joined &&
1367 1368 1369 1370 1371
	    (timer_pending(&mp->timer) ?
	     time_after(mp->timer.expires, now + max_delay) :
	     try_to_del_timer_sync(&mp->timer) >= 0))
		mod_timer(&mp->timer, now + max_delay);

1372 1373 1374
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
1375 1376 1377
		if (timer_pending(&p->timer) ?
		    time_after(p->timer.expires, now + max_delay) :
		    try_to_del_timer_sync(&p->timer) >= 0)
1378
			mod_timer(&p->timer, now + max_delay);
1379 1380 1381 1382 1383 1384 1385 1386
	}

out:
	spin_unlock(&br->multicast_lock);
	return err;
}
#endif

1387 1388 1389 1390 1391
static void
br_multicast_leave_group(struct net_bridge *br,
			 struct net_bridge_port *port,
			 struct br_ip *group,
			 struct bridge_mcast_other_query *other_query,
F
Felix Fietkau 已提交
1392 1393
			 struct bridge_mcast_own_query *own_query,
			 const unsigned char *src)
1394 1395 1396 1397 1398 1399 1400 1401
{
	struct net_bridge_mdb_entry *mp;
	struct net_bridge_port_group *p;
	unsigned long now;
	unsigned long time;

	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) ||
1402
	    (port && port->state == BR_STATE_DISABLED))
1403 1404
		goto out;

1405
	mp = br_mdb_ip_get(br, group);
1406 1407 1408
	if (!mp)
		goto out;

1409 1410 1411 1412 1413 1414
	if (port && (port->flags & BR_MULTICAST_FAST_LEAVE)) {
		struct net_bridge_port_group __rcu **pp;

		for (pp = &mp->ports;
		     (p = mlock_dereference(*pp, br)) != NULL;
		     pp = &p->next) {
F
Felix Fietkau 已提交
1415
			if (!br_port_group_equal(p, port, src))
1416 1417 1418 1419 1420
				continue;

			rcu_assign_pointer(*pp, p->next);
			hlist_del_init(&p->mglist);
			del_timer(&p->timer);
1421
			kfree_rcu(p, rcu);
1422 1423
			br_mdb_notify(br->dev, port, group, RTM_DELMDB,
				      p->flags);
1424

1425
			if (!mp->ports && !mp->host_joined &&
1426 1427 1428 1429 1430 1431 1432 1433 1434
			    netif_running(br->dev))
				mod_timer(&mp->timer, jiffies);
		}
		goto out;
	}

	if (timer_pending(&other_query->timer))
		goto out;

1435
	if (br_opt_get(br, BROPT_MULTICAST_QUERIER)) {
1436 1437 1438 1439
		__br_multicast_send_query(br, port, &mp->addr);

		time = jiffies + br->multicast_last_member_count *
				 br->multicast_last_member_interval;
1440

1441
		mod_timer(&own_query->timer, time);
1442 1443 1444 1445

		for (p = mlock_dereference(mp->ports, br);
		     p != NULL;
		     p = mlock_dereference(p->next, br)) {
F
Felix Fietkau 已提交
1446
			if (!br_port_group_equal(p, port, src))
1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459
				continue;

			if (!hlist_unhashed(&p->mglist) &&
			    (timer_pending(&p->timer) ?
			     time_after(p->timer.expires, time) :
			     try_to_del_timer_sync(&p->timer) >= 0)) {
				mod_timer(&p->timer, time);
			}

			break;
		}
	}

1460 1461 1462 1463 1464
	now = jiffies;
	time = now + br->multicast_last_member_count *
		     br->multicast_last_member_interval;

	if (!port) {
1465
		if (mp->host_joined &&
1466 1467 1468 1469 1470
		    (timer_pending(&mp->timer) ?
		     time_after(mp->timer.expires, time) :
		     try_to_del_timer_sync(&mp->timer) >= 0)) {
			mod_timer(&mp->timer, time);
		}
1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488

		goto out;
	}

	for (p = mlock_dereference(mp->ports, br);
	     p != NULL;
	     p = mlock_dereference(p->next, br)) {
		if (p->port != port)
			continue;

		if (!hlist_unhashed(&p->mglist) &&
		    (timer_pending(&p->timer) ?
		     time_after(p->timer.expires, time) :
		     try_to_del_timer_sync(&p->timer) >= 0)) {
			mod_timer(&p->timer, time);
		}

		break;
1489 1490 1491 1492 1493
	}
out:
	spin_unlock(&br->multicast_lock);
}

1494 1495
static void br_ip4_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
1496
					 __be32 group,
F
Felix Fietkau 已提交
1497 1498
					 __u16 vid,
					 const unsigned char *src)
1499 1500
{
	struct br_ip br_group;
1501
	struct bridge_mcast_own_query *own_query;
1502 1503 1504 1505

	if (ipv4_is_local_multicast(group))
		return;

1506 1507
	own_query = port ? &port->ip4_own_query : &br->ip4_own_query;

1508 1509
	br_group.u.ip4 = group;
	br_group.proto = htons(ETH_P_IP);
1510
	br_group.vid = vid;
1511

1512
	br_multicast_leave_group(br, port, &br_group, &br->ip4_other_query,
F
Felix Fietkau 已提交
1513
				 own_query, src);
1514 1515
}

E
Eric Dumazet 已提交
1516
#if IS_ENABLED(CONFIG_IPV6)
1517 1518
static void br_ip6_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
1519
					 const struct in6_addr *group,
F
Felix Fietkau 已提交
1520 1521
					 __u16 vid,
					 const unsigned char *src)
1522 1523
{
	struct br_ip br_group;
1524
	struct bridge_mcast_own_query *own_query;
1525

1526
	if (ipv6_addr_is_ll_all_nodes(group))
1527 1528
		return;

1529 1530
	own_query = port ? &port->ip6_own_query : &br->ip6_own_query;

A
Alexey Dobriyan 已提交
1531
	br_group.u.ip6 = *group;
1532
	br_group.proto = htons(ETH_P_IPV6);
1533
	br_group.vid = vid;
1534

1535
	br_multicast_leave_group(br, port, &br_group, &br->ip6_other_query,
F
Felix Fietkau 已提交
1536
				 own_query, src);
1537 1538
}
#endif
1539

1540 1541 1542 1543 1544 1545 1546
static void br_multicast_err_count(const struct net_bridge *br,
				   const struct net_bridge_port *p,
				   __be16 proto)
{
	struct bridge_mcast_stats __percpu *stats;
	struct bridge_mcast_stats *pstats;

1547
	if (!br_opt_get(br, BROPT_MULTICAST_STATS_ENABLED))
1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572
		return;

	if (p)
		stats = p->mcast_stats;
	else
		stats = br->mcast_stats;
	if (WARN_ON(!stats))
		return;

	pstats = this_cpu_ptr(stats);

	u64_stats_update_begin(&pstats->syncp);
	switch (proto) {
	case htons(ETH_P_IP):
		pstats->mstats.igmp_parse_errors++;
		break;
#if IS_ENABLED(CONFIG_IPV6)
	case htons(ETH_P_IPV6):
		pstats->mstats.mld_parse_errors++;
		break;
#endif
	}
	u64_stats_update_end(&pstats->syncp);
}

1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587
static void br_multicast_pim(struct net_bridge *br,
			     struct net_bridge_port *port,
			     const struct sk_buff *skb)
{
	unsigned int offset = skb_transport_offset(skb);
	struct pimhdr *pimhdr, _pimhdr;

	pimhdr = skb_header_pointer(skb, offset, sizeof(_pimhdr), &_pimhdr);
	if (!pimhdr || pim_hdr_version(pimhdr) != PIM_VERSION ||
	    pim_hdr_type(pimhdr) != PIM_TYPE_HELLO)
		return;

	br_multicast_mark_router(br, port);
}

1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600
static int br_ip4_multicast_mrd_rcv(struct net_bridge *br,
				    struct net_bridge_port *port,
				    struct sk_buff *skb)
{
	if (ip_hdr(skb)->protocol != IPPROTO_IGMP ||
	    igmp_hdr(skb)->type != IGMP_MRDISC_ADV)
		return -ENOMSG;

	br_multicast_mark_router(br, port);

	return 0;
}

1601 1602
static int br_multicast_ipv4_rcv(struct net_bridge *br,
				 struct net_bridge_port *port,
1603 1604
				 struct sk_buff *skb,
				 u16 vid)
1605
{
F
Felix Fietkau 已提交
1606
	const unsigned char *src;
1607 1608 1609
	struct igmphdr *ih;
	int err;

1610
	err = ip_mc_check_igmp(skb);
1611

1612
	if (err == -ENOMSG) {
1613
		if (!ipv4_is_local_multicast(ip_hdr(skb)->daddr)) {
1614
			BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1615 1616 1617
		} else if (pim_ipv4_all_pim_routers(ip_hdr(skb)->daddr)) {
			if (ip_hdr(skb)->protocol == IPPROTO_PIM)
				br_multicast_pim(br, port, skb);
1618 1619 1620 1621 1622 1623 1624
		} else if (ipv4_is_all_snoopers(ip_hdr(skb)->daddr)) {
			err = br_ip4_multicast_mrd_rcv(br, port, skb);

			if (err < 0 && err != -ENOMSG) {
				br_multicast_err_count(br, port, skb->protocol);
				return err;
			}
1625
		}
1626

1627
		return 0;
1628
	} else if (err < 0) {
1629
		br_multicast_err_count(br, port, skb->protocol);
1630
		return err;
1631
	}
1632

1633
	ih = igmp_hdr(skb);
F
Felix Fietkau 已提交
1634
	src = eth_hdr(skb)->h_source;
1635
	BR_INPUT_SKB_CB(skb)->igmp = ih->type;
1636 1637 1638 1639

	switch (ih->type) {
	case IGMP_HOST_MEMBERSHIP_REPORT:
	case IGMPV2_HOST_MEMBERSHIP_REPORT:
1640
		BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
F
Felix Fietkau 已提交
1641
		err = br_ip4_multicast_add_group(br, port, ih->group, vid, src);
1642 1643
		break;
	case IGMPV3_HOST_MEMBERSHIP_REPORT:
1644
		err = br_ip4_multicast_igmp3_report(br, port, skb, vid);
1645 1646
		break;
	case IGMP_HOST_MEMBERSHIP_QUERY:
1647
		br_ip4_multicast_query(br, port, skb, vid);
1648 1649
		break;
	case IGMP_HOST_LEAVE_MESSAGE:
F
Felix Fietkau 已提交
1650
		br_ip4_multicast_leave_group(br, port, ih->group, vid, src);
1651 1652 1653
		break;
	}

1654
	br_multicast_count(br, port, skb, BR_INPUT_SKB_CB(skb)->igmp,
1655 1656
			   BR_MCAST_DIR_RX);

1657 1658 1659
	return err;
}

E
Eric Dumazet 已提交
1660
#if IS_ENABLED(CONFIG_IPV6)
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681
static int br_ip6_multicast_mrd_rcv(struct net_bridge *br,
				    struct net_bridge_port *port,
				    struct sk_buff *skb)
{
	int ret;

	if (ipv6_hdr(skb)->nexthdr != IPPROTO_ICMPV6)
		return -ENOMSG;

	ret = ipv6_mc_check_icmpv6(skb);
	if (ret < 0)
		return ret;

	if (icmp6_hdr(skb)->icmp6_type != ICMPV6_MRDISC_ADV)
		return -ENOMSG;

	br_multicast_mark_router(br, port);

	return 0;
}

1682 1683
static int br_multicast_ipv6_rcv(struct net_bridge *br,
				 struct net_bridge_port *port,
1684 1685
				 struct sk_buff *skb,
				 u16 vid)
1686
{
F
Felix Fietkau 已提交
1687
	const unsigned char *src;
1688
	struct mld_msg *mld;
1689 1690
	int err;

1691
	err = ipv6_mc_check_mld(skb);
1692

1693 1694 1695
	if (err == -ENOMSG) {
		if (!ipv6_addr_is_ll_all_nodes(&ipv6_hdr(skb)->daddr))
			BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1696 1697 1698 1699 1700 1701 1702 1703 1704 1705

		if (ipv6_addr_is_all_snoopers(&ipv6_hdr(skb)->daddr)) {
			err = br_ip6_multicast_mrd_rcv(br, port, skb);

			if (err < 0 && err != -ENOMSG) {
				br_multicast_err_count(br, port, skb->protocol);
				return err;
			}
		}

1706
		return 0;
1707
	} else if (err < 0) {
1708
		br_multicast_err_count(br, port, skb->protocol);
1709
		return err;
1710 1711
	}

1712
	mld = (struct mld_msg *)skb_transport_header(skb);
1713
	BR_INPUT_SKB_CB(skb)->igmp = mld->mld_type;
1714

1715
	switch (mld->mld_type) {
1716
	case ICMPV6_MGM_REPORT:
F
Felix Fietkau 已提交
1717
		src = eth_hdr(skb)->h_source;
1718
		BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
F
Felix Fietkau 已提交
1719 1720
		err = br_ip6_multicast_add_group(br, port, &mld->mld_mca, vid,
						 src);
1721 1722
		break;
	case ICMPV6_MLD2_REPORT:
1723
		err = br_ip6_multicast_mld2_report(br, port, skb, vid);
1724 1725
		break;
	case ICMPV6_MGM_QUERY:
1726
		err = br_ip6_multicast_query(br, port, skb, vid);
1727 1728
		break;
	case ICMPV6_MGM_REDUCTION:
F
Felix Fietkau 已提交
1729 1730
		src = eth_hdr(skb)->h_source;
		br_ip6_multicast_leave_group(br, port, &mld->mld_mca, vid, src);
1731
		break;
1732 1733
	}

1734
	br_multicast_count(br, port, skb, BR_INPUT_SKB_CB(skb)->igmp,
1735 1736
			   BR_MCAST_DIR_RX);

1737 1738 1739 1740
	return err;
}
#endif

1741
int br_multicast_rcv(struct net_bridge *br, struct net_bridge_port *port,
1742
		     struct sk_buff *skb, u16 vid)
1743
{
1744 1745
	int ret = 0;

1746 1747 1748
	BR_INPUT_SKB_CB(skb)->igmp = 0;
	BR_INPUT_SKB_CB(skb)->mrouters_only = 0;

1749
	if (!br_opt_get(br, BROPT_MULTICAST_ENABLED))
1750 1751 1752 1753
		return 0;

	switch (skb->protocol) {
	case htons(ETH_P_IP):
1754 1755
		ret = br_multicast_ipv4_rcv(br, port, skb, vid);
		break;
E
Eric Dumazet 已提交
1756
#if IS_ENABLED(CONFIG_IPV6)
1757
	case htons(ETH_P_IPV6):
1758 1759
		ret = br_multicast_ipv6_rcv(br, port, skb, vid);
		break;
1760
#endif
1761 1762
	}

1763
	return ret;
1764 1765
}

1766
static void br_multicast_query_expired(struct net_bridge *br,
1767 1768
				       struct bridge_mcast_own_query *query,
				       struct bridge_mcast_querier *querier)
1769 1770 1771 1772 1773
{
	spin_lock(&br->multicast_lock);
	if (query->startup_sent < br->multicast_startup_query_count)
		query->startup_sent++;

1774
	RCU_INIT_POINTER(querier->port, NULL);
1775 1776 1777 1778
	br_multicast_send_query(br, NULL, query);
	spin_unlock(&br->multicast_lock);
}

1779
static void br_ip4_multicast_query_expired(struct timer_list *t)
1780
{
1781
	struct net_bridge *br = from_timer(br, t, ip4_own_query.timer);
1782

1783
	br_multicast_query_expired(br, &br->ip4_own_query, &br->ip4_querier);
1784
}
1785

1786
#if IS_ENABLED(CONFIG_IPV6)
1787
static void br_ip6_multicast_query_expired(struct timer_list *t)
1788
{
1789
	struct net_bridge *br = from_timer(br, t, ip6_own_query.timer);
1790

1791
	br_multicast_query_expired(br, &br->ip6_own_query, &br->ip6_querier);
1792
}
1793
#endif
1794 1795 1796

void br_multicast_init(struct net_bridge *br)
{
1797
	br->hash_max = BR_MULTICAST_DEFAULT_HASH_MAX;
1798

1799
	br->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809
	br->multicast_last_member_count = 2;
	br->multicast_startup_query_count = 2;

	br->multicast_last_member_interval = HZ;
	br->multicast_query_response_interval = 10 * HZ;
	br->multicast_startup_query_interval = 125 * HZ / 4;
	br->multicast_query_interval = 125 * HZ;
	br->multicast_querier_interval = 255 * HZ;
	br->multicast_membership_interval = 260 * HZ;

1810
	br->ip4_other_query.delay_time = 0;
1811
	br->ip4_querier.port = NULL;
1812
	br->multicast_igmp_version = 2;
1813
#if IS_ENABLED(CONFIG_IPV6)
1814
	br->multicast_mld_version = 1;
1815
	br->ip6_other_query.delay_time = 0;
1816
	br->ip6_querier.port = NULL;
1817
#endif
1818
	br_opt_toggle(br, BROPT_MULTICAST_ENABLED, true);
1819
	br_opt_toggle(br, BROPT_HAS_IPV6_ADDR, true);
1820

1821
	spin_lock_init(&br->multicast_lock);
1822 1823 1824 1825 1826 1827
	timer_setup(&br->multicast_router_timer,
		    br_multicast_local_router_expired, 0);
	timer_setup(&br->ip4_other_query.timer,
		    br_ip4_multicast_querier_expired, 0);
	timer_setup(&br->ip4_own_query.timer,
		    br_ip4_multicast_query_expired, 0);
1828
#if IS_ENABLED(CONFIG_IPV6)
1829 1830 1831 1832
	timer_setup(&br->ip6_other_query.timer,
		    br_ip6_multicast_querier_expired, 0);
	timer_setup(&br->ip6_own_query.timer,
		    br_ip6_multicast_query_expired, 0);
1833
#endif
1834
	INIT_HLIST_HEAD(&br->mdb_list);
1835 1836
}

1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898
static void br_ip4_multicast_join_snoopers(struct net_bridge *br)
{
	struct in_device *in_dev = in_dev_get(br->dev);

	if (!in_dev)
		return;

	ip_mc_inc_group(in_dev, htonl(INADDR_ALLSNOOPERS_GROUP));
	in_dev_put(in_dev);
}

#if IS_ENABLED(CONFIG_IPV6)
static void br_ip6_multicast_join_snoopers(struct net_bridge *br)
{
	struct in6_addr addr;

	ipv6_addr_set(&addr, htonl(0xff020000), 0, 0, htonl(0x6a));
	ipv6_dev_mc_inc(br->dev, &addr);
}
#else
static inline void br_ip6_multicast_join_snoopers(struct net_bridge *br)
{
}
#endif

static void br_multicast_join_snoopers(struct net_bridge *br)
{
	br_ip4_multicast_join_snoopers(br);
	br_ip6_multicast_join_snoopers(br);
}

static void br_ip4_multicast_leave_snoopers(struct net_bridge *br)
{
	struct in_device *in_dev = in_dev_get(br->dev);

	if (WARN_ON(!in_dev))
		return;

	ip_mc_dec_group(in_dev, htonl(INADDR_ALLSNOOPERS_GROUP));
	in_dev_put(in_dev);
}

#if IS_ENABLED(CONFIG_IPV6)
static void br_ip6_multicast_leave_snoopers(struct net_bridge *br)
{
	struct in6_addr addr;

	ipv6_addr_set(&addr, htonl(0xff020000), 0, 0, htonl(0x6a));
	ipv6_dev_mc_dec(br->dev, &addr);
}
#else
static inline void br_ip6_multicast_leave_snoopers(struct net_bridge *br)
{
}
#endif

static void br_multicast_leave_snoopers(struct net_bridge *br)
{
	br_ip4_multicast_leave_snoopers(br);
	br_ip6_multicast_leave_snoopers(br);
}

1899
static void __br_multicast_open(struct net_bridge *br,
1900
				struct bridge_mcast_own_query *query)
1901
{
1902
	query->startup_sent = 0;
1903

1904
	if (!br_opt_get(br, BROPT_MULTICAST_ENABLED))
1905 1906
		return;

1907 1908 1909 1910 1911
	mod_timer(&query->timer, jiffies);
}

void br_multicast_open(struct net_bridge *br)
{
1912 1913 1914
	if (br_opt_get(br, BROPT_MULTICAST_ENABLED))
		br_multicast_join_snoopers(br);

1915
	__br_multicast_open(br, &br->ip4_own_query);
1916
#if IS_ENABLED(CONFIG_IPV6)
1917
	__br_multicast_open(br, &br->ip6_own_query);
1918
#endif
1919 1920 1921 1922 1923
}

void br_multicast_stop(struct net_bridge *br)
{
	del_timer_sync(&br->multicast_router_timer);
1924 1925
	del_timer_sync(&br->ip4_other_query.timer);
	del_timer_sync(&br->ip4_own_query.timer);
1926
#if IS_ENABLED(CONFIG_IPV6)
1927 1928
	del_timer_sync(&br->ip6_other_query.timer);
	del_timer_sync(&br->ip6_own_query.timer);
1929
#endif
1930 1931 1932

	if (br_opt_get(br, BROPT_MULTICAST_ENABLED))
		br_multicast_leave_snoopers(br);
1933 1934 1935 1936 1937
}

void br_multicast_dev_del(struct net_bridge *br)
{
	struct net_bridge_mdb_entry *mp;
1938
	struct hlist_node *tmp;
1939 1940

	spin_lock_bh(&br->multicast_lock);
1941 1942 1943 1944 1945
	hlist_for_each_entry_safe(mp, tmp, &br->mdb_list, mdb_node) {
		del_timer(&mp->timer);
		rhashtable_remove_fast(&br->mdb_hash_tbl, &mp->rhnode,
				       br_mdb_rht_params);
		hlist_del_rcu(&mp->mdb_node);
1946
		kfree_rcu(mp, rcu);
1947 1948
	}
	spin_unlock_bh(&br->multicast_lock);
1949

1950
	rcu_barrier();
1951
}
1952 1953 1954

int br_multicast_set_router(struct net_bridge *br, unsigned long val)
{
1955
	int err = -EINVAL;
1956 1957 1958 1959

	spin_lock_bh(&br->multicast_lock);

	switch (val) {
1960 1961
	case MDB_RTR_TYPE_DISABLED:
	case MDB_RTR_TYPE_PERM:
1962
		br_mc_router_state_change(br, val == MDB_RTR_TYPE_PERM);
1963
		del_timer(&br->multicast_router_timer);
1964 1965 1966
		br->multicast_router = val;
		err = 0;
		break;
1967
	case MDB_RTR_TYPE_TEMP_QUERY:
1968 1969
		if (br->multicast_router != MDB_RTR_TYPE_TEMP_QUERY)
			br_mc_router_state_change(br, false);
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979
		br->multicast_router = val;
		err = 0;
		break;
	}

	spin_unlock_bh(&br->multicast_lock);

	return err;
}

1980 1981 1982 1983 1984 1985
static void __del_port_router(struct net_bridge_port *p)
{
	if (hlist_unhashed(&p->rlist))
		return;
	hlist_del_init_rcu(&p->rlist);
	br_rtr_notify(p->br->dev, p, RTM_DELMDB);
1986
	br_port_mc_router_state_change(p, false);
1987 1988 1989 1990

	/* don't allow timer refresh */
	if (p->multicast_router == MDB_RTR_TYPE_TEMP)
		p->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
1991 1992
}

1993 1994 1995
int br_multicast_set_port_router(struct net_bridge_port *p, unsigned long val)
{
	struct net_bridge *br = p->br;
1996
	unsigned long now = jiffies;
1997
	int err = -EINVAL;
1998 1999

	spin_lock(&br->multicast_lock);
2000
	if (p->multicast_router == val) {
2001 2002 2003 2004
		/* Refresh the temp router port timer */
		if (p->multicast_router == MDB_RTR_TYPE_TEMP)
			mod_timer(&p->multicast_router_timer,
				  now + br->multicast_querier_interval);
2005 2006 2007
		err = 0;
		goto unlock;
	}
2008
	switch (val) {
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
	case MDB_RTR_TYPE_DISABLED:
		p->multicast_router = MDB_RTR_TYPE_DISABLED;
		__del_port_router(p);
		del_timer(&p->multicast_router_timer);
		break;
	case MDB_RTR_TYPE_TEMP_QUERY:
		p->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
		__del_port_router(p);
		break;
	case MDB_RTR_TYPE_PERM:
		p->multicast_router = MDB_RTR_TYPE_PERM;
2020 2021 2022
		del_timer(&p->multicast_router_timer);
		br_multicast_add_router(br, p);
		break;
2023 2024 2025 2026
	case MDB_RTR_TYPE_TEMP:
		p->multicast_router = MDB_RTR_TYPE_TEMP;
		br_multicast_mark_router(br, p);
		break;
2027 2028
	default:
		goto unlock;
2029
	}
2030 2031
	err = 0;
unlock:
2032 2033 2034 2035
	spin_unlock(&br->multicast_lock);

	return err;
}
2036

2037
static void br_multicast_start_querier(struct net_bridge *br,
2038
				       struct bridge_mcast_own_query *query)
2039 2040
{
	struct net_bridge_port *port;
2041

2042
	__br_multicast_open(br, query);
2043 2044 2045 2046 2047 2048

	list_for_each_entry(port, &br->port_list, list) {
		if (port->state == BR_STATE_DISABLED ||
		    port->state == BR_STATE_BLOCKING)
			continue;

2049 2050
		if (query == &br->ip4_own_query)
			br_multicast_enable(&port->ip4_own_query);
2051 2052
#if IS_ENABLED(CONFIG_IPV6)
		else
2053
			br_multicast_enable(&port->ip6_own_query);
2054
#endif
2055 2056 2057 2058 2059
	}
}

int br_multicast_toggle(struct net_bridge *br, unsigned long val)
{
2060
	struct net_bridge_port *port;
2061

2062
	spin_lock_bh(&br->multicast_lock);
2063
	if (!!br_opt_get(br, BROPT_MULTICAST_ENABLED) == !!val)
2064 2065
		goto unlock;

2066 2067
	br_mc_disabled_update(br->dev, val);
	br_opt_toggle(br, BROPT_MULTICAST_ENABLED, !!val);
2068 2069
	if (!br_opt_get(br, BROPT_MULTICAST_ENABLED)) {
		br_multicast_leave_snoopers(br);
2070
		goto unlock;
2071
	}
2072

2073 2074 2075
	if (!netif_running(br->dev))
		goto unlock;

2076 2077 2078
	br_multicast_open(br);
	list_for_each_entry(port, &br->port_list, list)
		__br_multicast_enable_port(port);
2079 2080

unlock:
2081
	spin_unlock_bh(&br->multicast_lock);
2082

2083
	return 0;
2084
}
2085

2086 2087 2088 2089
bool br_multicast_enabled(const struct net_device *dev)
{
	struct net_bridge *br = netdev_priv(dev);

2090
	return !!br_opt_get(br, BROPT_MULTICAST_ENABLED);
2091 2092 2093
}
EXPORT_SYMBOL_GPL(br_multicast_enabled);

2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105
bool br_multicast_router(const struct net_device *dev)
{
	struct net_bridge *br = netdev_priv(dev);
	bool is_router;

	spin_lock_bh(&br->multicast_lock);
	is_router = br_multicast_is_router(br);
	spin_unlock_bh(&br->multicast_lock);
	return is_router;
}
EXPORT_SYMBOL_GPL(br_multicast_router);

2106 2107
int br_multicast_set_querier(struct net_bridge *br, unsigned long val)
{
2108 2109
	unsigned long max_delay;

2110 2111 2112
	val = !!val;

	spin_lock_bh(&br->multicast_lock);
2113
	if (br_opt_get(br, BROPT_MULTICAST_QUERIER) == val)
2114 2115
		goto unlock;

2116
	br_opt_toggle(br, BROPT_MULTICAST_QUERIER, !!val);
2117 2118 2119 2120 2121
	if (!val)
		goto unlock;

	max_delay = br->multicast_query_response_interval;

2122 2123
	if (!timer_pending(&br->ip4_other_query.timer))
		br->ip4_other_query.delay_time = jiffies + max_delay;
2124

2125
	br_multicast_start_querier(br, &br->ip4_own_query);
2126 2127

#if IS_ENABLED(CONFIG_IPV6)
2128 2129
	if (!timer_pending(&br->ip6_other_query.timer))
		br->ip6_other_query.delay_time = jiffies + max_delay;
2130

2131
	br_multicast_start_querier(br, &br->ip6_own_query);
2132
#endif
2133 2134 2135 2136 2137 2138 2139

unlock:
	spin_unlock_bh(&br->multicast_lock);

	return 0;
}

2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157
int br_multicast_set_igmp_version(struct net_bridge *br, unsigned long val)
{
	/* Currently we support only version 2 and 3 */
	switch (val) {
	case 2:
	case 3:
		break;
	default:
		return -EINVAL;
	}

	spin_lock_bh(&br->multicast_lock);
	br->multicast_igmp_version = val;
	spin_unlock_bh(&br->multicast_lock);

	return 0;
}

2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177
#if IS_ENABLED(CONFIG_IPV6)
int br_multicast_set_mld_version(struct net_bridge *br, unsigned long val)
{
	/* Currently we support version 1 and 2 */
	switch (val) {
	case 1:
	case 2:
		break;
	default:
		return -EINVAL;
	}

	spin_lock_bh(&br->multicast_lock);
	br->multicast_mld_version = val;
	spin_unlock_bh(&br->multicast_lock);

	return 0;
}
#endif

2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233
/**
 * br_multicast_list_adjacent - Returns snooped multicast addresses
 * @dev:	The bridge port adjacent to which to retrieve addresses
 * @br_ip_list:	The list to store found, snooped multicast IP addresses in
 *
 * Creates a list of IP addresses (struct br_ip_list) sensed by the multicast
 * snooping feature on all bridge ports of dev's bridge device, excluding
 * the addresses from dev itself.
 *
 * Returns the number of items added to br_ip_list.
 *
 * Notes:
 * - br_ip_list needs to be initialized by caller
 * - br_ip_list might contain duplicates in the end
 *   (needs to be taken care of by caller)
 * - br_ip_list needs to be freed by caller
 */
int br_multicast_list_adjacent(struct net_device *dev,
			       struct list_head *br_ip_list)
{
	struct net_bridge *br;
	struct net_bridge_port *port;
	struct net_bridge_port_group *group;
	struct br_ip_list *entry;
	int count = 0;

	rcu_read_lock();
	if (!br_ip_list || !br_port_exists(dev))
		goto unlock;

	port = br_port_get_rcu(dev);
	if (!port || !port->br)
		goto unlock;

	br = port->br;

	list_for_each_entry_rcu(port, &br->port_list, list) {
		if (!port->dev || port->dev == dev)
			continue;

		hlist_for_each_entry_rcu(group, &port->mglist, mglist) {
			entry = kmalloc(sizeof(*entry), GFP_ATOMIC);
			if (!entry)
				goto unlock;

			entry->addr = group->addr;
			list_add(&entry->list, br_ip_list);
			count++;
		}
	}

unlock:
	rcu_read_unlock();
	return count;
}
EXPORT_SYMBOL_GPL(br_multicast_list_adjacent);
2234

2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271
/**
 * br_multicast_has_querier_anywhere - Checks for a querier on a bridge
 * @dev: The bridge port providing the bridge on which to check for a querier
 * @proto: The protocol family to check for: IGMP -> ETH_P_IP, MLD -> ETH_P_IPV6
 *
 * Checks whether the given interface has a bridge on top and if so returns
 * true if a valid querier exists anywhere on the bridged link layer.
 * Otherwise returns false.
 */
bool br_multicast_has_querier_anywhere(struct net_device *dev, int proto)
{
	struct net_bridge *br;
	struct net_bridge_port *port;
	struct ethhdr eth;
	bool ret = false;

	rcu_read_lock();
	if (!br_port_exists(dev))
		goto unlock;

	port = br_port_get_rcu(dev);
	if (!port || !port->br)
		goto unlock;

	br = port->br;

	memset(&eth, 0, sizeof(eth));
	eth.h_proto = htons(proto);

	ret = br_multicast_querier_exists(br, &eth);

unlock:
	rcu_read_unlock();
	return ret;
}
EXPORT_SYMBOL_GPL(br_multicast_has_querier_anywhere);

2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302
/**
 * br_multicast_has_querier_adjacent - Checks for a querier behind a bridge port
 * @dev: The bridge port adjacent to which to check for a querier
 * @proto: The protocol family to check for: IGMP -> ETH_P_IP, MLD -> ETH_P_IPV6
 *
 * Checks whether the given interface has a bridge on top and if so returns
 * true if a selected querier is behind one of the other ports of this
 * bridge. Otherwise returns false.
 */
bool br_multicast_has_querier_adjacent(struct net_device *dev, int proto)
{
	struct net_bridge *br;
	struct net_bridge_port *port;
	bool ret = false;

	rcu_read_lock();
	if (!br_port_exists(dev))
		goto unlock;

	port = br_port_get_rcu(dev);
	if (!port || !port->br)
		goto unlock;

	br = port->br;

	switch (proto) {
	case ETH_P_IP:
		if (!timer_pending(&br->ip4_other_query.timer) ||
		    rcu_dereference(br->ip4_querier.port) == port)
			goto unlock;
		break;
2303
#if IS_ENABLED(CONFIG_IPV6)
2304 2305 2306 2307 2308
	case ETH_P_IPV6:
		if (!timer_pending(&br->ip6_other_query.timer) ||
		    rcu_dereference(br->ip6_querier.port) == port)
			goto unlock;
		break;
2309
#endif
2310 2311 2312 2313 2314 2315 2316 2317 2318 2319
	default:
		goto unlock;
	}

	ret = true;
unlock:
	rcu_read_unlock();
	return ret;
}
EXPORT_SYMBOL_GPL(br_multicast_has_querier_adjacent);
2320 2321

static void br_mcast_stats_add(struct bridge_mcast_stats __percpu *stats,
2322
			       const struct sk_buff *skb, u8 type, u8 dir)
2323 2324
{
	struct bridge_mcast_stats *pstats = this_cpu_ptr(stats);
2325 2326
	__be16 proto = skb->protocol;
	unsigned int t_len;
2327 2328 2329 2330

	u64_stats_update_begin(&pstats->syncp);
	switch (proto) {
	case htons(ETH_P_IP):
2331
		t_len = ntohs(ip_hdr(skb)->tot_len) - ip_hdrlen(skb);
2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342
		switch (type) {
		case IGMP_HOST_MEMBERSHIP_REPORT:
			pstats->mstats.igmp_v1reports[dir]++;
			break;
		case IGMPV2_HOST_MEMBERSHIP_REPORT:
			pstats->mstats.igmp_v2reports[dir]++;
			break;
		case IGMPV3_HOST_MEMBERSHIP_REPORT:
			pstats->mstats.igmp_v3reports[dir]++;
			break;
		case IGMP_HOST_MEMBERSHIP_QUERY:
2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357
			if (t_len != sizeof(struct igmphdr)) {
				pstats->mstats.igmp_v3queries[dir]++;
			} else {
				unsigned int offset = skb_transport_offset(skb);
				struct igmphdr *ih, _ihdr;

				ih = skb_header_pointer(skb, offset,
							sizeof(_ihdr), &_ihdr);
				if (!ih)
					break;
				if (!ih->code)
					pstats->mstats.igmp_v1queries[dir]++;
				else
					pstats->mstats.igmp_v2queries[dir]++;
			}
2358 2359 2360 2361 2362 2363 2364 2365
			break;
		case IGMP_HOST_LEAVE_MESSAGE:
			pstats->mstats.igmp_leaves[dir]++;
			break;
		}
		break;
#if IS_ENABLED(CONFIG_IPV6)
	case htons(ETH_P_IPV6):
2366 2367 2368
		t_len = ntohs(ipv6_hdr(skb)->payload_len) +
			sizeof(struct ipv6hdr);
		t_len -= skb_network_header_len(skb);
2369 2370 2371 2372 2373 2374 2375 2376
		switch (type) {
		case ICMPV6_MGM_REPORT:
			pstats->mstats.mld_v1reports[dir]++;
			break;
		case ICMPV6_MLD2_REPORT:
			pstats->mstats.mld_v2reports[dir]++;
			break;
		case ICMPV6_MGM_QUERY:
2377 2378 2379 2380
			if (t_len != sizeof(struct mld_msg))
				pstats->mstats.mld_v2queries[dir]++;
			else
				pstats->mstats.mld_v1queries[dir]++;
2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392
			break;
		case ICMPV6_MGM_REDUCTION:
			pstats->mstats.mld_leaves[dir]++;
			break;
		}
		break;
#endif /* CONFIG_IPV6 */
	}
	u64_stats_update_end(&pstats->syncp);
}

void br_multicast_count(struct net_bridge *br, const struct net_bridge_port *p,
2393
			const struct sk_buff *skb, u8 type, u8 dir)
2394 2395 2396 2397
{
	struct bridge_mcast_stats __percpu *stats;

	/* if multicast_disabled is true then igmp type can't be set */
2398
	if (!type || !br_opt_get(br, BROPT_MULTICAST_STATS_ENABLED))
2399 2400 2401 2402 2403 2404 2405 2406 2407
		return;

	if (p)
		stats = p->mcast_stats;
	else
		stats = br->mcast_stats;
	if (WARN_ON(!stats))
		return;

2408
	br_mcast_stats_add(stats, skb, type, dir);
2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419
}

int br_multicast_init_stats(struct net_bridge *br)
{
	br->mcast_stats = netdev_alloc_pcpu_stats(struct bridge_mcast_stats);
	if (!br->mcast_stats)
		return -ENOMEM;

	return 0;
}

2420 2421 2422 2423 2424
void br_multicast_uninit_stats(struct net_bridge *br)
{
	free_percpu(br->mcast_stats);
}

2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457
static void mcast_stats_add_dir(u64 *dst, u64 *src)
{
	dst[BR_MCAST_DIR_RX] += src[BR_MCAST_DIR_RX];
	dst[BR_MCAST_DIR_TX] += src[BR_MCAST_DIR_TX];
}

void br_multicast_get_stats(const struct net_bridge *br,
			    const struct net_bridge_port *p,
			    struct br_mcast_stats *dest)
{
	struct bridge_mcast_stats __percpu *stats;
	struct br_mcast_stats tdst;
	int i;

	memset(dest, 0, sizeof(*dest));
	if (p)
		stats = p->mcast_stats;
	else
		stats = br->mcast_stats;
	if (WARN_ON(!stats))
		return;

	memset(&tdst, 0, sizeof(tdst));
	for_each_possible_cpu(i) {
		struct bridge_mcast_stats *cpu_stats = per_cpu_ptr(stats, i);
		struct br_mcast_stats temp;
		unsigned int start;

		do {
			start = u64_stats_fetch_begin_irq(&cpu_stats->syncp);
			memcpy(&temp, &cpu_stats->mstats, sizeof(temp));
		} while (u64_stats_fetch_retry_irq(&cpu_stats->syncp, start));

2458 2459 2460
		mcast_stats_add_dir(tdst.igmp_v1queries, temp.igmp_v1queries);
		mcast_stats_add_dir(tdst.igmp_v2queries, temp.igmp_v2queries);
		mcast_stats_add_dir(tdst.igmp_v3queries, temp.igmp_v3queries);
2461 2462 2463 2464 2465 2466
		mcast_stats_add_dir(tdst.igmp_leaves, temp.igmp_leaves);
		mcast_stats_add_dir(tdst.igmp_v1reports, temp.igmp_v1reports);
		mcast_stats_add_dir(tdst.igmp_v2reports, temp.igmp_v2reports);
		mcast_stats_add_dir(tdst.igmp_v3reports, temp.igmp_v3reports);
		tdst.igmp_parse_errors += temp.igmp_parse_errors;

2467 2468
		mcast_stats_add_dir(tdst.mld_v1queries, temp.mld_v1queries);
		mcast_stats_add_dir(tdst.mld_v2queries, temp.mld_v2queries);
2469 2470 2471 2472 2473 2474 2475
		mcast_stats_add_dir(tdst.mld_leaves, temp.mld_leaves);
		mcast_stats_add_dir(tdst.mld_v1reports, temp.mld_v1reports);
		mcast_stats_add_dir(tdst.mld_v2reports, temp.mld_v2reports);
		tdst.mld_parse_errors += temp.mld_parse_errors;
	}
	memcpy(dest, &tdst, sizeof(*dest));
}
2476 2477 2478 2479 2480 2481 2482 2483 2484 2485

int br_mdb_hash_init(struct net_bridge *br)
{
	return rhashtable_init(&br->mdb_hash_tbl, &br_mdb_rht_params);
}

void br_mdb_hash_fini(struct net_bridge *br)
{
	rhashtable_destroy(&br->mdb_hash_tbl);
}