br_multicast.c 58.9 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 17 18
#include <linux/if_ether.h>
#include <linux/igmp.h>
#include <linux/jhash.h>
#include <linux/kernel.h>
19
#include <linux/log2.h>
20 21 22 23 24 25 26
#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>
27
#include <linux/inetdevice.h>
28
#include <linux/mroute.h>
29
#include <net/ip.h>
E
Eric Dumazet 已提交
30
#if IS_ENABLED(CONFIG_IPV6)
31 32
#include <net/ipv6.h>
#include <net/mld.h>
33
#include <net/ip6_checksum.h>
34
#include <net/addrconf.h>
35
#endif
36 37 38

#include "br_private.h"

39
static void br_multicast_start_querier(struct net_bridge *br,
40
				       struct bridge_mcast_own_query *query);
41 42
static void br_multicast_add_router(struct net_bridge *br,
				    struct net_bridge_port *port);
43 44 45 46 47 48 49 50 51 52
static void br_ip4_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
					 __be32 group,
					 __u16 vid);
#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,
					 __u16 vid);
#endif
C
Cong Wang 已提交
53
unsigned int br_mdb_rehash_seq;
54

55 56 57 58
static inline int br_ip_equal(const struct br_ip *a, const struct br_ip *b)
{
	if (a->proto != b->proto)
		return 0;
59 60
	if (a->vid != b->vid)
		return 0;
61 62 63
	switch (a->proto) {
	case htons(ETH_P_IP):
		return a->u.ip4 == b->u.ip4;
E
Eric Dumazet 已提交
64
#if IS_ENABLED(CONFIG_IPV6)
65 66 67
	case htons(ETH_P_IPV6):
		return ipv6_addr_equal(&a->u.ip6, &b->u.ip6);
#endif
68 69 70 71
	}
	return 0;
}

72 73
static inline int __br_ip4_hash(struct net_bridge_mdb_htable *mdb, __be32 ip,
				__u16 vid)
74
{
75
	return jhash_2words((__force u32)ip, vid, mdb->secret) & (mdb->max - 1);
76 77
}

E
Eric Dumazet 已提交
78
#if IS_ENABLED(CONFIG_IPV6)
79
static inline int __br_ip6_hash(struct net_bridge_mdb_htable *mdb,
80 81
				const struct in6_addr *ip,
				__u16 vid)
82
{
83 84
	return jhash_2words(ipv6_addr_hash(ip), vid,
			    mdb->secret) & (mdb->max - 1);
85 86 87
}
#endif

88 89 90 91 92
static inline int br_ip_hash(struct net_bridge_mdb_htable *mdb,
			     struct br_ip *ip)
{
	switch (ip->proto) {
	case htons(ETH_P_IP):
93
		return __br_ip4_hash(mdb, ip->u.ip4, ip->vid);
E
Eric Dumazet 已提交
94
#if IS_ENABLED(CONFIG_IPV6)
95
	case htons(ETH_P_IPV6):
96
		return __br_ip6_hash(mdb, &ip->u.ip6, ip->vid);
97
#endif
98 99
	}
	return 0;
100 101 102
}

static struct net_bridge_mdb_entry *__br_mdb_ip_get(
103
	struct net_bridge_mdb_htable *mdb, struct br_ip *dst, int hash)
104 105 106
{
	struct net_bridge_mdb_entry *mp;

107
	hlist_for_each_entry_rcu(mp, &mdb->mhash[hash], hlist[mdb->ver]) {
108
		if (br_ip_equal(&mp->addr, dst))
109 110 111 112 113 114
			return mp;
	}

	return NULL;
}

115 116
struct net_bridge_mdb_entry *br_mdb_ip_get(struct net_bridge_mdb_htable *mdb,
					   struct br_ip *dst)
117 118 119 120 121 122 123
{
	if (!mdb)
		return NULL;

	return __br_mdb_ip_get(mdb, dst, br_ip_hash(mdb, dst));
}

124
static struct net_bridge_mdb_entry *br_mdb_ip4_get(
125
	struct net_bridge_mdb_htable *mdb, __be32 dst, __u16 vid)
126
{
127 128 129 130
	struct br_ip br_dst;

	br_dst.u.ip4 = dst;
	br_dst.proto = htons(ETH_P_IP);
131
	br_dst.vid = vid;
132

133
	return br_mdb_ip_get(mdb, &br_dst);
134 135
}

E
Eric Dumazet 已提交
136
#if IS_ENABLED(CONFIG_IPV6)
137
static struct net_bridge_mdb_entry *br_mdb_ip6_get(
138 139
	struct net_bridge_mdb_htable *mdb, const struct in6_addr *dst,
	__u16 vid)
140 141
{
	struct br_ip br_dst;
142

A
Alexey Dobriyan 已提交
143
	br_dst.u.ip6 = *dst;
144
	br_dst.proto = htons(ETH_P_IPV6);
145
	br_dst.vid = vid;
146

147
	return br_mdb_ip_get(mdb, &br_dst);
148 149 150
}
#endif

151
struct net_bridge_mdb_entry *br_mdb_get(struct net_bridge *br,
152
					struct sk_buff *skb, u16 vid)
153
{
154
	struct net_bridge_mdb_htable *mdb = rcu_dereference(br->mdb);
155 156
	struct br_ip ip;

157
	if (br->multicast_disabled)
158 159
		return NULL;

160
	if (BR_INPUT_SKB_CB(skb)->igmp)
161 162
		return NULL;

163
	ip.proto = skb->protocol;
164
	ip.vid = vid;
165

166 167
	switch (skb->protocol) {
	case htons(ETH_P_IP):
168 169
		ip.u.ip4 = ip_hdr(skb)->daddr;
		break;
E
Eric Dumazet 已提交
170
#if IS_ENABLED(CONFIG_IPV6)
171
	case htons(ETH_P_IPV6):
A
Alexey Dobriyan 已提交
172
		ip.u.ip6 = ipv6_hdr(skb)->daddr;
173 174
		break;
#endif
175 176
	default:
		return NULL;
177 178
	}

179
	return br_mdb_ip_get(mdb, &ip);
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202
}

static void br_mdb_free(struct rcu_head *head)
{
	struct net_bridge_mdb_htable *mdb =
		container_of(head, struct net_bridge_mdb_htable, rcu);
	struct net_bridge_mdb_htable *old = mdb->old;

	mdb->old = NULL;
	kfree(old->mhash);
	kfree(old);
}

static int br_mdb_copy(struct net_bridge_mdb_htable *new,
		       struct net_bridge_mdb_htable *old,
		       int elasticity)
{
	struct net_bridge_mdb_entry *mp;
	int maxlen;
	int len;
	int i;

	for (i = 0; i < old->max; i++)
203
		hlist_for_each_entry(mp, &old->mhash[i], hlist[old->ver])
204
			hlist_add_head(&mp->hlist[new->ver],
205
				       &new->mhash[br_ip_hash(new, &mp->addr)]);
206 207 208 209 210 211 212

	if (!elasticity)
		return 0;

	maxlen = 0;
	for (i = 0; i < new->max; i++) {
		len = 0;
213
		hlist_for_each_entry(mp, &new->mhash[i], hlist[new->ver])
214 215 216 217 218 219 220 221
			len++;
		if (len > maxlen)
			maxlen = len;
	}

	return maxlen > elasticity ? -EINVAL : 0;
}

222
void br_multicast_free_pg(struct rcu_head *head)
223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247
{
	struct net_bridge_port_group *p =
		container_of(head, struct net_bridge_port_group, rcu);

	kfree(p);
}

static void br_multicast_free_group(struct rcu_head *head)
{
	struct net_bridge_mdb_entry *mp =
		container_of(head, struct net_bridge_mdb_entry, rcu);

	kfree(mp);
}

static void br_multicast_group_expired(unsigned long data)
{
	struct net_bridge_mdb_entry *mp = (void *)data;
	struct net_bridge *br = mp->br;
	struct net_bridge_mdb_htable *mdb;

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

248
	mp->mglist = false;
249 250 251 252

	if (mp->ports)
		goto out;

253 254
	mdb = mlock_dereference(br->mdb, br);

255 256 257 258 259 260 261 262 263 264 265 266
	hlist_del_rcu(&mp->hlist[mdb->ver]);
	mdb->size--;

	call_rcu_bh(&mp->rcu, br_multicast_free_group);

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

static void br_multicast_del_pg(struct net_bridge *br,
				struct net_bridge_port_group *pg)
{
267
	struct net_bridge_mdb_htable *mdb;
268 269
	struct net_bridge_mdb_entry *mp;
	struct net_bridge_port_group *p;
270 271 272
	struct net_bridge_port_group __rcu **pp;

	mdb = mlock_dereference(br->mdb, br);
273

274
	mp = br_mdb_ip_get(mdb, &pg->addr);
275 276 277
	if (WARN_ON(!mp))
		return;

278 279 280
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
281 282 283
		if (p != pg)
			continue;

284
		rcu_assign_pointer(*pp, p->next);
285 286
		hlist_del_init(&p->mglist);
		del_timer(&p->timer);
287 288
		br_mdb_notify(br->dev, p->port, &pg->addr, RTM_DELMDB,
			      p->flags);
289 290
		call_rcu_bh(&p->rcu, br_multicast_free_pg);

291
		if (!mp->ports && !mp->mglist &&
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307
		    netif_running(br->dev))
			mod_timer(&mp->timer, jiffies);

		return;
	}

	WARN_ON(1);
}

static void br_multicast_port_group_expired(unsigned long data)
{
	struct net_bridge_port_group *pg = (void *)data;
	struct net_bridge *br = pg->port->br;

	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) || timer_pending(&pg->timer) ||
308
	    hlist_unhashed(&pg->mglist) || pg->flags & MDB_PG_FLAGS_PERMANENT)
309 310 311 312 313 314 315 316
		goto out;

	br_multicast_del_pg(br, pg);

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

317
static int br_mdb_rehash(struct net_bridge_mdb_htable __rcu **mdbp, int max,
318 319
			 int elasticity)
{
320
	struct net_bridge_mdb_htable *old = rcu_dereference_protected(*mdbp, 1);
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354
	struct net_bridge_mdb_htable *mdb;
	int err;

	mdb = kmalloc(sizeof(*mdb), GFP_ATOMIC);
	if (!mdb)
		return -ENOMEM;

	mdb->max = max;
	mdb->old = old;

	mdb->mhash = kzalloc(max * sizeof(*mdb->mhash), GFP_ATOMIC);
	if (!mdb->mhash) {
		kfree(mdb);
		return -ENOMEM;
	}

	mdb->size = old ? old->size : 0;
	mdb->ver = old ? old->ver ^ 1 : 0;

	if (!old || elasticity)
		get_random_bytes(&mdb->secret, sizeof(mdb->secret));
	else
		mdb->secret = old->secret;

	if (!old)
		goto out;

	err = br_mdb_copy(mdb, old, elasticity);
	if (err) {
		kfree(mdb->mhash);
		kfree(mdb);
		return err;
	}

C
Cong Wang 已提交
355
	br_mdb_rehash_seq++;
356 357 358 359 360 361 362 363
	call_rcu_bh(&mdb->rcu, br_mdb_free);

out:
	rcu_assign_pointer(*mdbp, mdb);

	return 0;
}

364
static struct sk_buff *br_ip4_multicast_alloc_query(struct net_bridge *br,
365 366
						    __be32 group,
						    u8 *igmp_type)
367
{
368 369
	struct igmpv3_query *ihv3;
	size_t igmp_hdr_size;
370 371 372 373 374
	struct sk_buff *skb;
	struct igmphdr *ih;
	struct ethhdr *eth;
	struct iphdr *iph;

375 376 377
	igmp_hdr_size = sizeof(*ih);
	if (br->multicast_igmp_version == 3)
		igmp_hdr_size = sizeof(*ihv3);
378
	skb = netdev_alloc_skb_ip_align(br->dev, sizeof(*eth) + sizeof(*iph) +
379
						 igmp_hdr_size + 4);
380 381 382 383 384 385 386 387
	if (!skb)
		goto out;

	skb->protocol = htons(ETH_P_IP);

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

388
	ether_addr_copy(eth->h_source, br->dev->dev_addr);
389 390 391 392 393 394 395 396 397 398 399 400 401 402 403
	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;
404
	iph->tot_len = htons(sizeof(*iph) + igmp_hdr_size + 4);
405 406 407 408
	iph->id = 0;
	iph->frag_off = htons(IP_DF);
	iph->ttl = 1;
	iph->protocol = IPPROTO_IGMP;
409 410
	iph->saddr = br->multicast_query_use_ifaddr ?
		     inet_select_addr(br->dev, 0, RT_SCOPE_LINK) : 0;
411 412 413 414 415 416 417 418 419
	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);
420
	*igmp_type = IGMP_HOST_MEMBERSHIP_QUERY;
421

422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450
	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);
451 452 453 454 455 456
	__skb_pull(skb, sizeof(*eth));

out:
	return skb;
}

E
Eric Dumazet 已提交
457
#if IS_ENABLED(CONFIG_IPV6)
458
static struct sk_buff *br_ip6_multicast_alloc_query(struct net_bridge *br,
459 460
						    const struct in6_addr *grp,
						    u8 *igmp_type)
461
{
462 463
	struct mld2_query *mld2q;
	unsigned long interval;
464 465
	struct ipv6hdr *ip6h;
	struct mld_msg *mldq;
466 467
	size_t mld_hdr_size;
	struct sk_buff *skb;
468 469 470
	struct ethhdr *eth;
	u8 *hopopt;

471 472 473
	mld_hdr_size = sizeof(*mldq);
	if (br->multicast_mld_version == 2)
		mld_hdr_size = sizeof(*mld2q);
474
	skb = netdev_alloc_skb_ip_align(br->dev, sizeof(*eth) + sizeof(*ip6h) +
475
						 8 + mld_hdr_size);
476 477 478 479 480 481 482 483 484
	if (!skb)
		goto out;

	skb->protocol = htons(ETH_P_IPV6);

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

485
	ether_addr_copy(eth->h_source, br->dev->dev_addr);
486 487 488 489 490 491 492 493
	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);
494
	ip6h->payload_len = htons(8 + mld_hdr_size);
495 496
	ip6h->nexthdr = IPPROTO_HOPOPTS;
	ip6h->hop_limit = 1;
497
	ipv6_addr_set(&ip6h->daddr, htonl(0xff020000), 0, 0, htonl(1));
498 499 500
	if (ipv6_dev_get_saddr(dev_net(br->dev), br->dev, &ip6h->daddr, 0,
			       &ip6h->saddr)) {
		kfree_skb(skb);
501
		br->has_ipv6_addr = 0;
502 503
		return NULL;
	}
504 505

	br->has_ipv6_addr = 1;
506
	ipv6_eth_mc_map(&ip6h->daddr, eth->h_dest);
507 508 509 510 511 512 513 514

	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;
515 516
	hopopt[6] = IPV6_TLV_PAD1;		/* Pad1 */
	hopopt[7] = IPV6_TLV_PAD1;		/* Pad1 */
517 518 519 520 521

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

	/* ICMPv6 */
	skb_set_transport_header(skb, skb->len);
522
	interval = ipv6_addr_any(grp) ?
523 524
			br->multicast_query_response_interval :
			br->multicast_last_member_interval;
525
	*igmp_type = ICMPV6_MGM_QUERY;
526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542
	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);
543
		mld2q->mld2q_mrc = htons((u16)jiffies_to_msecs(interval));
544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562
		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);
563 564 565 566 567 568 569 570

	__skb_pull(skb, sizeof(*eth));

out:
	return skb;
}
#endif

571
static struct sk_buff *br_multicast_alloc_query(struct net_bridge *br,
572 573
						struct br_ip *addr,
						u8 *igmp_type)
574 575 576
{
	switch (addr->proto) {
	case htons(ETH_P_IP):
577
		return br_ip4_multicast_alloc_query(br, addr->u.ip4, igmp_type);
E
Eric Dumazet 已提交
578
#if IS_ENABLED(CONFIG_IPV6)
579
	case htons(ETH_P_IPV6):
580 581
		return br_ip6_multicast_alloc_query(br, &addr->u.ip6,
						    igmp_type);
582
#endif
583 584 585 586
	}
	return NULL;
}

587
static struct net_bridge_mdb_entry *br_multicast_get_group(
588 589
	struct net_bridge *br, struct net_bridge_port *port,
	struct br_ip *group, int hash)
590
{
591
	struct net_bridge_mdb_htable *mdb;
592
	struct net_bridge_mdb_entry *mp;
593 594
	unsigned int count = 0;
	unsigned int max;
595 596 597
	int elasticity;
	int err;

598
	mdb = rcu_dereference_protected(br->mdb, 1);
599
	hlist_for_each_entry(mp, &mdb->mhash[hash], hlist[mdb->ver]) {
600
		count++;
601
		if (unlikely(br_ip_equal(group, &mp->addr)))
602 603 604 605 606 607 608 609
			return mp;
	}

	elasticity = 0;
	max = mdb->max;

	if (unlikely(count > br->hash_elasticity && count)) {
		if (net_ratelimit())
610 611 612
			br_info(br, "Multicast hash table "
				"chain limit reached: %s\n",
				port ? port->dev->name : br->dev->name);
613 614 615 616 617 618

		elasticity = br->hash_elasticity;
	}

	if (mdb->size >= max) {
		max *= 2;
619 620 621 622 623
		if (unlikely(max > br->hash_max)) {
			br_warn(br, "Multicast hash table maximum of %d "
				"reached, disabling snooping: %s\n",
				br->hash_max,
				port ? port->dev->name : br->dev->name);
624 625 626 627 628 629 630 631 632 633
			err = -E2BIG;
disable:
			br->multicast_disabled = 1;
			goto err;
		}
	}

	if (max > mdb->max || elasticity) {
		if (mdb->old) {
			if (net_ratelimit())
634 635 636
				br_info(br, "Multicast hash table "
					"on fire: %s\n",
					port ? port->dev->name : br->dev->name);
637 638 639 640 641 642
			err = -EEXIST;
			goto err;
		}

		err = br_mdb_rehash(&br->mdb, max, elasticity);
		if (err) {
643 644 645 646
			br_warn(br, "Cannot rehash multicast "
				"hash table, disabling snooping: %s, %d, %d\n",
				port ? port->dev->name : br->dev->name,
				mdb->size, err);
647 648 649 650 651 652 653 654 655 656 657 658 659 660
			goto disable;
		}

		err = -EAGAIN;
		goto err;
	}

	return NULL;

err:
	mp = ERR_PTR(err);
	return mp;
}

661
struct net_bridge_mdb_entry *br_multicast_new_group(struct net_bridge *br,
662 663
						    struct net_bridge_port *p,
						    struct br_ip *group)
664
{
665
	struct net_bridge_mdb_htable *mdb;
666 667
	struct net_bridge_mdb_entry *mp;
	int hash;
668
	int err;
669

670
	mdb = rcu_dereference_protected(br->mdb, 1);
671
	if (!mdb) {
672 673 674
		err = br_mdb_rehash(&br->mdb, BR_HASH_SIZE, 0);
		if (err)
			return ERR_PTR(err);
675 676 677 678
		goto rehash;
	}

	hash = br_ip_hash(mdb, group);
679
	mp = br_multicast_get_group(br, p, group, hash);
680 681 682 683 684 685
	switch (PTR_ERR(mp)) {
	case 0:
		break;

	case -EAGAIN:
rehash:
686
		mdb = rcu_dereference_protected(br->mdb, 1);
687 688 689 690 691 692 693 694 695
		hash = br_ip_hash(mdb, group);
		break;

	default:
		goto out;
	}

	mp = kzalloc(sizeof(*mp), GFP_ATOMIC);
	if (unlikely(!mp))
696
		return ERR_PTR(-ENOMEM);
697 698

	mp->br = br;
699
	mp->addr = *group;
700 701 702
	setup_timer(&mp->timer, br_multicast_group_expired,
		    (unsigned long)mp);

703 704 705 706 707 708 709
	hlist_add_head_rcu(&mp->hlist[mdb->ver], &mdb->mhash[hash]);
	mdb->size++;

out:
	return mp;
}

710 711 712
struct net_bridge_port_group *br_multicast_new_port_group(
			struct net_bridge_port *port,
			struct br_ip *group,
713
			struct net_bridge_port_group __rcu *next,
714
			unsigned char flags)
715 716 717 718 719 720 721 722 723
{
	struct net_bridge_port_group *p;

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

	p->addr = *group;
	p->port = port;
724
	p->flags = flags;
725
	rcu_assign_pointer(p->next, next);
726 727 728 729 730 731
	hlist_add_head(&p->mglist, &port->mglist);
	setup_timer(&p->timer, br_multicast_port_group_expired,
		    (unsigned long)p);
	return p;
}

732
static int br_multicast_add_group(struct net_bridge *br,
733 734
				  struct net_bridge_port *port,
				  struct br_ip *group)
735
{
736
	struct net_bridge_port_group __rcu **pp;
737 738
	struct net_bridge_port_group *p;
	struct net_bridge_mdb_entry *mp;
739
	unsigned long now = jiffies;
740 741 742 743 744 745 746 747 748
	int err;

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

	mp = br_multicast_new_group(br, port, group);
	err = PTR_ERR(mp);
749
	if (IS_ERR(mp))
750 751 752
		goto err;

	if (!port) {
753
		mp->mglist = true;
754
		mod_timer(&mp->timer, now + br->multicast_membership_interval);
755 756 757
		goto out;
	}

758 759 760
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
761
		if (p->port == port)
762
			goto found;
763 764 765 766
		if ((unsigned long)p->port < (unsigned long)port)
			break;
	}

767
	p = br_multicast_new_port_group(port, group, *pp, 0);
768 769 770
	if (unlikely(!p))
		goto err;
	rcu_assign_pointer(*pp, p);
771
	br_mdb_notify(br->dev, port, group, RTM_NEWMDB, 0);
772

773 774
found:
	mod_timer(&p->timer, now + br->multicast_membership_interval);
775 776 777 778 779 780 781 782
out:
	err = 0;

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

783 784
static int br_ip4_multicast_add_group(struct net_bridge *br,
				      struct net_bridge_port *port,
785 786
				      __be32 group,
				      __u16 vid)
787 788 789 790 791 792 793 794
{
	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);
795
	br_group.vid = vid;
796 797 798 799

	return br_multicast_add_group(br, port, &br_group);
}

E
Eric Dumazet 已提交
800
#if IS_ENABLED(CONFIG_IPV6)
801 802
static int br_ip6_multicast_add_group(struct net_bridge *br,
				      struct net_bridge_port *port,
803 804
				      const struct in6_addr *group,
				      __u16 vid)
805 806 807
{
	struct br_ip br_group;

808
	if (ipv6_addr_is_ll_all_nodes(group))
809 810
		return 0;

A
Alexey Dobriyan 已提交
811
	br_group.u.ip6 = *group;
812
	br_group.proto = htons(ETH_P_IPV6);
813
	br_group.vid = vid;
814 815 816 817 818

	return br_multicast_add_group(br, port, &br_group);
}
#endif

819 820 821 822 823 824
static void br_multicast_router_expired(unsigned long data)
{
	struct net_bridge_port *port = (void *)data;
	struct net_bridge *br = port->br;

	spin_lock(&br->multicast_lock);
825 826
	if (port->multicast_router == MDB_RTR_TYPE_DISABLED ||
	    port->multicast_router == MDB_RTR_TYPE_PERM ||
827 828 829 830 831
	    timer_pending(&port->multicast_router_timer) ||
	    hlist_unhashed(&port->rlist))
		goto out;

	hlist_del_init_rcu(&port->rlist);
832
	br_rtr_notify(br->dev, port, RTM_DELMDB);
833 834 835
	/* Don't allow timer refresh if the router expired */
	if (port->multicast_router == MDB_RTR_TYPE_TEMP)
		port->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
836 837 838 839 840 841 842 843 844

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

static void br_multicast_local_router_expired(unsigned long data)
{
}

845
static void br_multicast_querier_expired(struct net_bridge *br,
846
					 struct bridge_mcast_own_query *query)
847 848 849 850 851
{
	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) || br->multicast_disabled)
		goto out;

852
	br_multicast_start_querier(br, query);
853 854 855 856 857

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

858 859 860 861
static void br_ip4_multicast_querier_expired(unsigned long data)
{
	struct net_bridge *br = (void *)data;

862
	br_multicast_querier_expired(br, &br->ip4_own_query);
863 864 865 866 867 868 869
}

#if IS_ENABLED(CONFIG_IPV6)
static void br_ip6_multicast_querier_expired(unsigned long data)
{
	struct net_bridge *br = (void *)data;

870
	br_multicast_querier_expired(br, &br->ip6_own_query);
871 872 873
}
#endif

874 875 876 877 878 879 880 881 882 883 884 885
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
}

886 887 888
static void __br_multicast_send_query(struct net_bridge *br,
				      struct net_bridge_port *port,
				      struct br_ip *ip)
889 890
{
	struct sk_buff *skb;
891
	u8 igmp_type;
892

893
	skb = br_multicast_alloc_query(br, ip, &igmp_type);
894
	if (!skb)
895
		return;
896 897 898

	if (port) {
		skb->dev = port->dev;
899
		br_multicast_count(br, port, skb, igmp_type,
900
				   BR_MCAST_DIR_TX);
901 902
		NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_OUT,
			dev_net(port->dev), NULL, skb, NULL, skb->dev,
903
			br_dev_queue_push_xmit);
904 905
	} else {
		br_multicast_select_own_querier(br, ip, skb);
906
		br_multicast_count(br, port, skb, igmp_type,
907
				   BR_MCAST_DIR_RX);
908
		netif_rx(skb);
909
	}
910 911 912
}

static void br_multicast_send_query(struct net_bridge *br,
913
				    struct net_bridge_port *port,
914
				    struct bridge_mcast_own_query *own_query)
915
{
916
	struct bridge_mcast_other_query *other_query = NULL;
917 918
	struct br_ip br_group;
	unsigned long time;
919 920

	if (!netif_running(br->dev) || br->multicast_disabled ||
921
	    !br->multicast_querier)
922 923
		return;

924 925
	memset(&br_group.u, 0, sizeof(br_group.u));

926 927 928
	if (port ? (own_query == &port->ip4_own_query) :
		   (own_query == &br->ip4_own_query)) {
		other_query = &br->ip4_other_query;
929
		br_group.proto = htons(ETH_P_IP);
E
Eric Dumazet 已提交
930
#if IS_ENABLED(CONFIG_IPV6)
931
	} else {
932
		other_query = &br->ip6_other_query;
933
		br_group.proto = htons(ETH_P_IPV6);
934
#endif
935 936
	}

937
	if (!other_query || timer_pending(&other_query->timer))
938 939 940
		return;

	__br_multicast_send_query(br, port, &br_group);
941 942

	time = jiffies;
943
	time += own_query->startup_sent < br->multicast_startup_query_count ?
944 945
		br->multicast_startup_query_interval :
		br->multicast_query_interval;
946
	mod_timer(&own_query->timer, time);
947 948
}

949 950 951
static void
br_multicast_port_query_expired(struct net_bridge_port *port,
				struct bridge_mcast_own_query *query)
952 953 954 955
{
	struct net_bridge *br = port->br;

	spin_lock(&br->multicast_lock);
956 957
	if (port->state == BR_STATE_DISABLED ||
	    port->state == BR_STATE_BLOCKING)
958 959
		goto out;

960 961
	if (query->startup_sent < br->multicast_startup_query_count)
		query->startup_sent++;
962

963
	br_multicast_send_query(port->br, port, query);
964 965 966 967 968

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

969 970 971 972
static void br_ip4_multicast_port_query_expired(unsigned long data)
{
	struct net_bridge_port *port = (void *)data;

973
	br_multicast_port_query_expired(port, &port->ip4_own_query);
974 975 976 977 978 979 980
}

#if IS_ENABLED(CONFIG_IPV6)
static void br_ip6_multicast_port_query_expired(unsigned long data)
{
	struct net_bridge_port *port = (void *)data;

981
	br_multicast_port_query_expired(port, &port->ip6_own_query);
982 983 984
}
#endif

985
int br_multicast_add_port(struct net_bridge_port *port)
986
{
987
	port->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
988 989 990

	setup_timer(&port->multicast_router_timer, br_multicast_router_expired,
		    (unsigned long)port);
991 992
	setup_timer(&port->ip4_own_query.timer,
		    br_ip4_multicast_port_query_expired, (unsigned long)port);
993
#if IS_ENABLED(CONFIG_IPV6)
994 995
	setup_timer(&port->ip6_own_query.timer,
		    br_ip6_multicast_port_query_expired, (unsigned long)port);
996
#endif
997 998 999 1000 1001
	port->mcast_stats = netdev_alloc_pcpu_stats(struct bridge_mcast_stats);
	if (!port->mcast_stats)
		return -ENOMEM;

	return 0;
1002 1003 1004 1005
}

void br_multicast_del_port(struct net_bridge_port *port)
{
1006 1007 1008 1009 1010 1011 1012 1013 1014
	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);
1015
	del_timer_sync(&port->multicast_router_timer);
1016
	free_percpu(port->mcast_stats);
1017 1018
}

1019
static void br_multicast_enable(struct bridge_mcast_own_query *query)
1020
{
1021
	query->startup_sent = 0;
1022

1023 1024 1025
	if (try_to_del_timer_sync(&query->timer) >= 0 ||
	    del_timer(&query->timer))
		mod_timer(&query->timer, jiffies);
1026 1027
}

1028
static void __br_multicast_enable_port(struct net_bridge_port *port)
1029 1030 1031 1032
{
	struct net_bridge *br = port->br;

	if (br->multicast_disabled || !netif_running(br->dev))
1033
		return;
1034

1035
	br_multicast_enable(&port->ip4_own_query);
1036
#if IS_ENABLED(CONFIG_IPV6)
1037
	br_multicast_enable(&port->ip6_own_query);
1038
#endif
1039 1040
	if (port->multicast_router == MDB_RTR_TYPE_PERM &&
	    hlist_unhashed(&port->rlist))
1041
		br_multicast_add_router(br, port);
1042
}
1043

1044 1045 1046 1047 1048 1049
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);
1050 1051 1052 1053 1054 1055 1056
	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;
1057
	struct hlist_node *n;
1058 1059

	spin_lock(&br->multicast_lock);
1060
	hlist_for_each_entry_safe(pg, n, &port->mglist, mglist)
1061
		if (!(pg->flags & MDB_PG_FLAGS_PERMANENT))
1062
			br_multicast_del_pg(br, pg);
1063

1064
	if (!hlist_unhashed(&port->rlist)) {
1065
		hlist_del_init_rcu(&port->rlist);
1066
		br_rtr_notify(br->dev, port, RTM_DELMDB);
1067 1068 1069
		/* Don't allow timer refresh if disabling */
		if (port->multicast_router == MDB_RTR_TYPE_TEMP)
			port->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
1070
	}
1071
	del_timer(&port->multicast_router_timer);
1072
	del_timer(&port->ip4_own_query.timer);
1073
#if IS_ENABLED(CONFIG_IPV6)
1074
	del_timer(&port->ip6_own_query.timer);
1075
#endif
1076 1077 1078
	spin_unlock(&br->multicast_lock);
}

1079 1080
static int br_ip4_multicast_igmp3_report(struct net_bridge *br,
					 struct net_bridge_port *port,
1081 1082
					 struct sk_buff *skb,
					 u16 vid)
1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094
{
	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);
1095
	len = skb_transport_offset(skb) + sizeof(*ih);
1096 1097 1098 1099 1100 1101

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

H
Herbert Xu 已提交
1102
		grec = (void *)(skb->data + len - sizeof(*grec));
1103 1104 1105
		group = grec->grec_mca;
		type = grec->grec_type;

E
Eric Dumazet 已提交
1106
		len += ntohs(grec->grec_nsrcs) * 4;
1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123
		if (!pskb_may_pull(skb, len))
			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;
		}

1124 1125 1126 1127 1128 1129 1130 1131 1132
		if ((type == IGMPV3_CHANGE_TO_INCLUDE ||
		     type == IGMPV3_MODE_IS_INCLUDE) &&
		    ntohs(grec->grec_nsrcs) == 0) {
			br_ip4_multicast_leave_group(br, port, group, vid);
		} else {
			err = br_ip4_multicast_add_group(br, port, group, vid);
			if (err)
				break;
		}
1133 1134 1135 1136 1137
	}

	return err;
}

E
Eric Dumazet 已提交
1138
#if IS_ENABLED(CONFIG_IPV6)
1139 1140
static int br_ip6_multicast_mld2_report(struct net_bridge *br,
					struct net_bridge_port *port,
1141 1142
					struct sk_buff *skb,
					u16 vid)
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155
{
	struct icmp6hdr *icmp6h;
	struct mld2_grec *grec;
	int i;
	int len;
	int num;
	int err = 0;

	if (!pskb_may_pull(skb, sizeof(*icmp6h)))
		return -EINVAL;

	icmp6h = icmp6_hdr(skb);
	num = ntohs(icmp6h->icmp6_dataun.un_data16[1]);
1156
	len = skb_transport_offset(skb) + sizeof(*icmp6h);
1157 1158 1159 1160 1161 1162

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

		nsrcs = skb_header_pointer(skb,
					   len + offsetof(struct mld2_grec,
1163
							  grec_nsrcs),
1164 1165 1166 1167 1168 1169
					   sizeof(_nsrcs), &_nsrcs);
		if (!nsrcs)
			return -EINVAL;

		if (!pskb_may_pull(skb,
				   len + sizeof(*grec) +
1170
				   sizeof(struct in6_addr) * ntohs(*nsrcs)))
1171 1172 1173
			return -EINVAL;

		grec = (struct mld2_grec *)(skb->data + len);
1174 1175
		len += sizeof(*grec) +
		       sizeof(struct in6_addr) * ntohs(*nsrcs);
1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190

		/* 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;
		}

1191 1192 1193 1194 1195 1196 1197 1198
		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,
						     vid);
		} else {
			err = br_ip6_multicast_add_group(br, port,
							 &grec->grec_mca, vid);
1199
			if (err)
1200 1201
				break;
		}
1202 1203 1204 1205 1206 1207
	}

	return err;
}
#endif

1208
static bool br_ip4_multicast_select_querier(struct net_bridge *br,
1209
					    struct net_bridge_port *port,
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226
					    __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;

1227 1228 1229
	/* update protected by general multicast_lock by caller */
	rcu_assign_pointer(br->ip4_querier.port, port);

1230 1231 1232 1233 1234
	return true;
}

#if IS_ENABLED(CONFIG_IPV6)
static bool br_ip6_multicast_select_querier(struct net_bridge *br,
1235
					    struct net_bridge_port *port,
1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249
					    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;

1250 1251 1252
	/* update protected by general multicast_lock by caller */
	rcu_assign_pointer(br->ip6_querier.port, port);

1253 1254 1255 1256 1257
	return true;
}
#endif

static bool br_multicast_select_querier(struct net_bridge *br,
1258
					struct net_bridge_port *port,
1259 1260 1261 1262
					struct br_ip *saddr)
{
	switch (saddr->proto) {
	case htons(ETH_P_IP):
1263
		return br_ip4_multicast_select_querier(br, port, saddr->u.ip4);
1264 1265
#if IS_ENABLED(CONFIG_IPV6)
	case htons(ETH_P_IPV6):
1266
		return br_ip6_multicast_select_querier(br, port, &saddr->u.ip6);
1267 1268 1269 1270 1271 1272
#endif
	}

	return false;
}

1273
static void
1274 1275 1276
br_multicast_update_query_timer(struct net_bridge *br,
				struct bridge_mcast_other_query *query,
				unsigned long max_delay)
1277
{
1278 1279
	if (!timer_pending(&query->timer))
		query->delay_time = jiffies + max_delay;
1280

1281
	mod_timer(&query->timer, jiffies + br->multicast_querier_interval);
1282 1283
}

1284
/*
C
Cong Wang 已提交
1285
 * Add port to router_list
1286 1287 1288
 *  list is maintained ordered by pointer value
 *  and locked by br->multicast_lock and RCU
 */
1289 1290 1291
static void br_multicast_add_router(struct net_bridge *br,
				    struct net_bridge_port *port)
{
1292
	struct net_bridge_port *p;
1293
	struct hlist_node *slot = NULL;
1294

1295 1296 1297
	if (!hlist_unhashed(&port->rlist))
		return;

1298
	hlist_for_each_entry(p, &br->router_list, rlist) {
1299 1300
		if ((unsigned long) port >= (unsigned long) p)
			break;
1301
		slot = &p->rlist;
1302 1303
	}

1304
	if (slot)
1305
		hlist_add_behind_rcu(&port->rlist, slot);
1306 1307
	else
		hlist_add_head_rcu(&port->rlist, &br->router_list);
1308
	br_rtr_notify(br->dev, port, RTM_NEWMDB);
1309 1310
}

1311 1312 1313 1314 1315 1316
static void br_multicast_mark_router(struct net_bridge *br,
				     struct net_bridge_port *port)
{
	unsigned long now = jiffies;

	if (!port) {
1317
		if (br->multicast_router == MDB_RTR_TYPE_TEMP_QUERY)
1318 1319 1320 1321 1322
			mod_timer(&br->multicast_router_timer,
				  now + br->multicast_querier_interval);
		return;
	}

1323 1324
	if (port->multicast_router == MDB_RTR_TYPE_DISABLED ||
	    port->multicast_router == MDB_RTR_TYPE_PERM)
1325 1326
		return;

1327
	br_multicast_add_router(br, port);
1328 1329 1330 1331 1332 1333 1334

	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,
1335
					struct bridge_mcast_other_query *query,
1336
					struct br_ip *saddr,
1337
					unsigned long max_delay)
1338
{
1339
	if (!br_multicast_select_querier(br, port, saddr))
1340 1341
		return;

1342
	br_multicast_update_query_timer(br, query, max_delay);
1343 1344 1345
	br_multicast_mark_router(br, port);
}

1346 1347
static int br_ip4_multicast_query(struct net_bridge *br,
				  struct net_bridge_port *port,
1348 1349
				  struct sk_buff *skb,
				  u16 vid)
1350
{
1351
	const struct iphdr *iph = ip_hdr(skb);
1352 1353 1354 1355
	struct igmphdr *ih = igmp_hdr(skb);
	struct net_bridge_mdb_entry *mp;
	struct igmpv3_query *ih3;
	struct net_bridge_port_group *p;
1356
	struct net_bridge_port_group __rcu **pp;
1357
	struct br_ip saddr;
1358 1359
	unsigned long max_delay;
	unsigned long now = jiffies;
1360
	unsigned int offset = skb_transport_offset(skb);
1361
	__be32 group;
1362
	int err = 0;
1363 1364 1365 1366 1367 1368 1369 1370

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

	group = ih->group;

1371
	if (skb->len == offset + sizeof(*ih)) {
1372 1373 1374 1375 1376 1377
		max_delay = ih->code * (HZ / IGMP_TIMER_SCALE);

		if (!max_delay) {
			max_delay = 10 * HZ;
			group = 0;
		}
1378
	} else if (skb->len >= offset + sizeof(*ih3)) {
1379 1380
		ih3 = igmpv3_query_hdr(skb);
		if (ih3->nsrcs)
1381
			goto out;
1382

1383 1384
		max_delay = ih3->code ?
			    IGMPV3_MRC(ih3->code) * (HZ / IGMP_TIMER_SCALE) : 1;
1385
	} else {
1386 1387 1388
		goto out;
	}

1389 1390 1391
	if (!group) {
		saddr.proto = htons(ETH_P_IP);
		saddr.u.ip4 = iph->saddr;
1392

1393 1394
		br_multicast_query_received(br, port, &br->ip4_other_query,
					    &saddr, max_delay);
1395
		goto out;
1396
	}
1397

1398
	mp = br_mdb_ip4_get(mlock_dereference(br->mdb, br), group, vid);
1399 1400 1401 1402 1403
	if (!mp)
		goto out;

	max_delay *= br->multicast_last_member_count;

1404
	if (mp->mglist &&
1405 1406 1407 1408 1409
	    (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);

1410 1411 1412
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
1413 1414 1415
		if (timer_pending(&p->timer) ?
		    time_after(p->timer.expires, now + max_delay) :
		    try_to_del_timer_sync(&p->timer) >= 0)
1416
			mod_timer(&p->timer, now + max_delay);
1417 1418 1419 1420
	}

out:
	spin_unlock(&br->multicast_lock);
1421
	return err;
1422 1423
}

E
Eric Dumazet 已提交
1424
#if IS_ENABLED(CONFIG_IPV6)
1425 1426
static int br_ip6_multicast_query(struct net_bridge *br,
				  struct net_bridge_port *port,
1427 1428
				  struct sk_buff *skb,
				  u16 vid)
1429
{
1430
	const struct ipv6hdr *ip6h = ipv6_hdr(skb);
1431
	struct mld_msg *mld;
1432 1433
	struct net_bridge_mdb_entry *mp;
	struct mld2_query *mld2q;
1434 1435
	struct net_bridge_port_group *p;
	struct net_bridge_port_group __rcu **pp;
1436
	struct br_ip saddr;
1437 1438
	unsigned long max_delay;
	unsigned long now = jiffies;
1439
	unsigned int offset = skb_transport_offset(skb);
1440
	const struct in6_addr *group = NULL;
1441
	bool is_general_query;
1442 1443 1444 1445 1446 1447 1448
	int err = 0;

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

1449 1450
	if (skb->len == offset + sizeof(*mld)) {
		if (!pskb_may_pull(skb, offset + sizeof(*mld))) {
1451 1452 1453 1454
			err = -EINVAL;
			goto out;
		}
		mld = (struct mld_msg *) icmp6_hdr(skb);
L
Li RongQing 已提交
1455
		max_delay = msecs_to_jiffies(ntohs(mld->mld_maxdelay));
1456 1457
		if (max_delay)
			group = &mld->mld_mca;
1458
	} else {
1459
		if (!pskb_may_pull(skb, offset + sizeof(*mld2q))) {
1460 1461 1462 1463 1464 1465
			err = -EINVAL;
			goto out;
		}
		mld2q = (struct mld2_query *)icmp6_hdr(skb);
		if (!mld2q->mld2q_nsrcs)
			group = &mld2q->mld2q_mca;
1466 1467

		max_delay = max(msecs_to_jiffies(mldv2_mrc(mld2q)), 1UL);
1468 1469
	}

1470 1471
	is_general_query = group && ipv6_addr_any(group);

1472 1473 1474
	if (is_general_query) {
		saddr.proto = htons(ETH_P_IPV6);
		saddr.u.ip6 = ip6h->saddr;
1475

1476 1477
		br_multicast_query_received(br, port, &br->ip6_other_query,
					    &saddr, max_delay);
1478
		goto out;
1479 1480
	} else if (!group) {
		goto out;
1481
	}
1482

1483
	mp = br_mdb_ip6_get(mlock_dereference(br->mdb, br), group, vid);
1484 1485 1486 1487
	if (!mp)
		goto out;

	max_delay *= br->multicast_last_member_count;
1488
	if (mp->mglist &&
1489 1490 1491 1492 1493
	    (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);

1494 1495 1496
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
1497 1498 1499
		if (timer_pending(&p->timer) ?
		    time_after(p->timer.expires, now + max_delay) :
		    try_to_del_timer_sync(&p->timer) >= 0)
1500
			mod_timer(&p->timer, now + max_delay);
1501 1502 1503 1504 1505 1506 1507 1508
	}

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

1509 1510 1511 1512 1513 1514
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,
			 struct bridge_mcast_own_query *own_query)
1515 1516 1517 1518 1519 1520 1521 1522 1523
{
	struct net_bridge_mdb_htable *mdb;
	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) ||
1524
	    (port && port->state == BR_STATE_DISABLED))
1525 1526
		goto out;

1527
	mdb = mlock_dereference(br->mdb, br);
1528 1529 1530 1531
	mp = br_mdb_ip_get(mdb, group);
	if (!mp)
		goto out;

1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544
	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) {
			if (p->port != port)
				continue;

			rcu_assign_pointer(*pp, p->next);
			hlist_del_init(&p->mglist);
			del_timer(&p->timer);
			call_rcu_bh(&p->rcu, br_multicast_free_pg);
1545 1546
			br_mdb_notify(br->dev, port, group, RTM_DELMDB,
				      p->flags);
1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557

			if (!mp->ports && !mp->mglist &&
			    netif_running(br->dev))
				mod_timer(&mp->timer, jiffies);
		}
		goto out;
	}

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

1558
	if (br->multicast_querier) {
1559 1560 1561 1562
		__br_multicast_send_query(br, port, &mp->addr);

		time = jiffies + br->multicast_last_member_count *
				 br->multicast_last_member_interval;
1563

1564
		mod_timer(&own_query->timer, time);
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582

		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;
		}
	}

1583 1584 1585 1586 1587
	now = jiffies;
	time = now + br->multicast_last_member_count *
		     br->multicast_last_member_interval;

	if (!port) {
1588
		if (mp->mglist &&
1589 1590 1591 1592 1593
		    (timer_pending(&mp->timer) ?
		     time_after(mp->timer.expires, time) :
		     try_to_del_timer_sync(&mp->timer) >= 0)) {
			mod_timer(&mp->timer, time);
		}
1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611

		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;
1612 1613 1614 1615 1616
	}
out:
	spin_unlock(&br->multicast_lock);
}

1617 1618
static void br_ip4_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
1619 1620
					 __be32 group,
					 __u16 vid)
1621 1622
{
	struct br_ip br_group;
1623
	struct bridge_mcast_own_query *own_query;
1624 1625 1626 1627

	if (ipv4_is_local_multicast(group))
		return;

1628 1629
	own_query = port ? &port->ip4_own_query : &br->ip4_own_query;

1630 1631
	br_group.u.ip4 = group;
	br_group.proto = htons(ETH_P_IP);
1632
	br_group.vid = vid;
1633

1634 1635
	br_multicast_leave_group(br, port, &br_group, &br->ip4_other_query,
				 own_query);
1636 1637
}

E
Eric Dumazet 已提交
1638
#if IS_ENABLED(CONFIG_IPV6)
1639 1640
static void br_ip6_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
1641 1642
					 const struct in6_addr *group,
					 __u16 vid)
1643 1644
{
	struct br_ip br_group;
1645
	struct bridge_mcast_own_query *own_query;
1646

1647
	if (ipv6_addr_is_ll_all_nodes(group))
1648 1649
		return;

1650 1651
	own_query = port ? &port->ip6_own_query : &br->ip6_own_query;

A
Alexey Dobriyan 已提交
1652
	br_group.u.ip6 = *group;
1653
	br_group.proto = htons(ETH_P_IPV6);
1654
	br_group.vid = vid;
1655

1656 1657
	br_multicast_leave_group(br, port, &br_group, &br->ip6_other_query,
				 own_query);
1658 1659
}
#endif
1660

1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693
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;

	if (!br->multicast_stats_enabled)
		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);
}

1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708
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);
}

1709 1710
static int br_multicast_ipv4_rcv(struct net_bridge *br,
				 struct net_bridge_port *port,
1711 1712
				 struct sk_buff *skb,
				 u16 vid)
1713
{
1714
	struct sk_buff *skb_trimmed = NULL;
1715 1716 1717
	struct igmphdr *ih;
	int err;

1718
	err = ip_mc_check_igmp(skb, &skb_trimmed);
1719

1720
	if (err == -ENOMSG) {
1721
		if (!ipv4_is_local_multicast(ip_hdr(skb)->daddr)) {
1722
			BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1723 1724 1725 1726
		} else if (pim_ipv4_all_pim_routers(ip_hdr(skb)->daddr)) {
			if (ip_hdr(skb)->protocol == IPPROTO_PIM)
				br_multicast_pim(br, port, skb);
		}
1727
		return 0;
1728
	} else if (err < 0) {
1729
		br_multicast_err_count(br, port, skb->protocol);
1730
		return err;
1731
	}
1732

1733
	ih = igmp_hdr(skb);
1734
	BR_INPUT_SKB_CB(skb)->igmp = ih->type;
1735 1736 1737 1738

	switch (ih->type) {
	case IGMP_HOST_MEMBERSHIP_REPORT:
	case IGMPV2_HOST_MEMBERSHIP_REPORT:
1739
		BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1740
		err = br_ip4_multicast_add_group(br, port, ih->group, vid);
1741 1742
		break;
	case IGMPV3_HOST_MEMBERSHIP_REPORT:
1743
		err = br_ip4_multicast_igmp3_report(br, port, skb_trimmed, vid);
1744 1745
		break;
	case IGMP_HOST_MEMBERSHIP_QUERY:
1746
		err = br_ip4_multicast_query(br, port, skb_trimmed, vid);
1747 1748
		break;
	case IGMP_HOST_LEAVE_MESSAGE:
1749
		br_ip4_multicast_leave_group(br, port, ih->group, vid);
1750 1751 1752
		break;
	}

1753
	if (skb_trimmed && skb_trimmed != skb)
1754 1755
		kfree_skb(skb_trimmed);

1756
	br_multicast_count(br, port, skb, BR_INPUT_SKB_CB(skb)->igmp,
1757 1758
			   BR_MCAST_DIR_RX);

1759 1760 1761
	return err;
}

E
Eric Dumazet 已提交
1762
#if IS_ENABLED(CONFIG_IPV6)
1763 1764
static int br_multicast_ipv6_rcv(struct net_bridge *br,
				 struct net_bridge_port *port,
1765 1766
				 struct sk_buff *skb,
				 u16 vid)
1767
{
1768 1769
	struct sk_buff *skb_trimmed = NULL;
	struct mld_msg *mld;
1770 1771
	int err;

1772
	err = ipv6_mc_check_mld(skb, &skb_trimmed);
1773

1774 1775 1776
	if (err == -ENOMSG) {
		if (!ipv6_addr_is_ll_all_nodes(&ipv6_hdr(skb)->daddr))
			BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1777
		return 0;
1778
	} else if (err < 0) {
1779
		br_multicast_err_count(br, port, skb->protocol);
1780
		return err;
1781 1782
	}

1783
	mld = (struct mld_msg *)skb_transport_header(skb);
1784
	BR_INPUT_SKB_CB(skb)->igmp = mld->mld_type;
1785

1786
	switch (mld->mld_type) {
1787
	case ICMPV6_MGM_REPORT:
1788
		BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1789
		err = br_ip6_multicast_add_group(br, port, &mld->mld_mca, vid);
1790 1791
		break;
	case ICMPV6_MLD2_REPORT:
1792
		err = br_ip6_multicast_mld2_report(br, port, skb_trimmed, vid);
1793 1794
		break;
	case ICMPV6_MGM_QUERY:
1795
		err = br_ip6_multicast_query(br, port, skb_trimmed, vid);
1796 1797
		break;
	case ICMPV6_MGM_REDUCTION:
1798
		br_ip6_multicast_leave_group(br, port, &mld->mld_mca, vid);
1799
		break;
1800 1801
	}

1802
	if (skb_trimmed && skb_trimmed != skb)
1803 1804
		kfree_skb(skb_trimmed);

1805
	br_multicast_count(br, port, skb, BR_INPUT_SKB_CB(skb)->igmp,
1806 1807
			   BR_MCAST_DIR_RX);

1808 1809 1810 1811
	return err;
}
#endif

1812
int br_multicast_rcv(struct net_bridge *br, struct net_bridge_port *port,
1813
		     struct sk_buff *skb, u16 vid)
1814
{
1815 1816
	int ret = 0;

1817 1818 1819
	BR_INPUT_SKB_CB(skb)->igmp = 0;
	BR_INPUT_SKB_CB(skb)->mrouters_only = 0;

1820 1821 1822 1823 1824
	if (br->multicast_disabled)
		return 0;

	switch (skb->protocol) {
	case htons(ETH_P_IP):
1825 1826
		ret = br_multicast_ipv4_rcv(br, port, skb, vid);
		break;
E
Eric Dumazet 已提交
1827
#if IS_ENABLED(CONFIG_IPV6)
1828
	case htons(ETH_P_IPV6):
1829 1830
		ret = br_multicast_ipv6_rcv(br, port, skb, vid);
		break;
1831
#endif
1832 1833
	}

1834
	return ret;
1835 1836
}

1837
static void br_multicast_query_expired(struct net_bridge *br,
1838 1839
				       struct bridge_mcast_own_query *query,
				       struct bridge_mcast_querier *querier)
1840 1841 1842 1843 1844
{
	spin_lock(&br->multicast_lock);
	if (query->startup_sent < br->multicast_startup_query_count)
		query->startup_sent++;

1845
	RCU_INIT_POINTER(querier->port, NULL);
1846 1847 1848 1849 1850
	br_multicast_send_query(br, NULL, query);
	spin_unlock(&br->multicast_lock);
}

static void br_ip4_multicast_query_expired(unsigned long data)
1851 1852 1853
{
	struct net_bridge *br = (void *)data;

1854
	br_multicast_query_expired(br, &br->ip4_own_query, &br->ip4_querier);
1855
}
1856

1857 1858 1859 1860
#if IS_ENABLED(CONFIG_IPV6)
static void br_ip6_multicast_query_expired(unsigned long data)
{
	struct net_bridge *br = (void *)data;
1861

1862
	br_multicast_query_expired(br, &br->ip6_own_query, &br->ip6_querier);
1863
}
1864
#endif
1865 1866 1867 1868 1869 1870

void br_multicast_init(struct net_bridge *br)
{
	br->hash_elasticity = 4;
	br->hash_max = 512;

1871
	br->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
1872
	br->multicast_querier = 0;
1873
	br->multicast_query_use_ifaddr = 0;
1874 1875 1876 1877 1878 1879 1880 1881 1882 1883
	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;

1884
	br->ip4_other_query.delay_time = 0;
1885
	br->ip4_querier.port = NULL;
1886
	br->multicast_igmp_version = 2;
1887
#if IS_ENABLED(CONFIG_IPV6)
1888
	br->multicast_mld_version = 1;
1889
	br->ip6_other_query.delay_time = 0;
1890
	br->ip6_querier.port = NULL;
1891
#endif
1892
	br->has_ipv6_addr = 1;
1893

1894 1895 1896
	spin_lock_init(&br->multicast_lock);
	setup_timer(&br->multicast_router_timer,
		    br_multicast_local_router_expired, 0);
1897 1898 1899
	setup_timer(&br->ip4_other_query.timer,
		    br_ip4_multicast_querier_expired, (unsigned long)br);
	setup_timer(&br->ip4_own_query.timer, br_ip4_multicast_query_expired,
1900
		    (unsigned long)br);
1901
#if IS_ENABLED(CONFIG_IPV6)
1902 1903 1904
	setup_timer(&br->ip6_other_query.timer,
		    br_ip6_multicast_querier_expired, (unsigned long)br);
	setup_timer(&br->ip6_own_query.timer, br_ip6_multicast_query_expired,
1905 1906
		    (unsigned long)br);
#endif
1907 1908
}

1909
static void __br_multicast_open(struct net_bridge *br,
1910
				struct bridge_mcast_own_query *query)
1911
{
1912
	query->startup_sent = 0;
1913 1914 1915 1916

	if (br->multicast_disabled)
		return;

1917 1918 1919 1920 1921
	mod_timer(&query->timer, jiffies);
}

void br_multicast_open(struct net_bridge *br)
{
1922
	__br_multicast_open(br, &br->ip4_own_query);
1923
#if IS_ENABLED(CONFIG_IPV6)
1924
	__br_multicast_open(br, &br->ip6_own_query);
1925
#endif
1926 1927 1928 1929 1930
}

void br_multicast_stop(struct net_bridge *br)
{
	del_timer_sync(&br->multicast_router_timer);
1931 1932
	del_timer_sync(&br->ip4_other_query.timer);
	del_timer_sync(&br->ip4_own_query.timer);
1933
#if IS_ENABLED(CONFIG_IPV6)
1934 1935
	del_timer_sync(&br->ip6_other_query.timer);
	del_timer_sync(&br->ip6_own_query.timer);
1936
#endif
1937 1938 1939 1940 1941 1942 1943 1944 1945
}

void br_multicast_dev_del(struct net_bridge *br)
{
	struct net_bridge_mdb_htable *mdb;
	struct net_bridge_mdb_entry *mp;
	struct hlist_node *n;
	u32 ver;
	int i;
1946 1947

	spin_lock_bh(&br->multicast_lock);
1948
	mdb = mlock_dereference(br->mdb, br);
1949 1950 1951 1952 1953 1954 1955
	if (!mdb)
		goto out;

	br->mdb = NULL;

	ver = mdb->ver;
	for (i = 0; i < mdb->max; i++) {
1956
		hlist_for_each_entry_safe(mp, n, &mdb->mhash[i],
1957 1958 1959 1960 1961 1962 1963 1964
					  hlist[ver]) {
			del_timer(&mp->timer);
			call_rcu_bh(&mp->rcu, br_multicast_free_group);
		}
	}

	if (mdb->old) {
		spin_unlock_bh(&br->multicast_lock);
1965
		rcu_barrier_bh();
1966 1967 1968 1969 1970 1971 1972 1973 1974
		spin_lock_bh(&br->multicast_lock);
		WARN_ON(mdb->old);
	}

	mdb->old = mdb;
	call_rcu_bh(&mdb->rcu, br_mdb_free);

out:
	spin_unlock_bh(&br->multicast_lock);
1975 1976

	free_percpu(br->mcast_stats);
1977
}
1978 1979 1980

int br_multicast_set_router(struct net_bridge *br, unsigned long val)
{
1981
	int err = -EINVAL;
1982 1983 1984 1985

	spin_lock_bh(&br->multicast_lock);

	switch (val) {
1986 1987
	case MDB_RTR_TYPE_DISABLED:
	case MDB_RTR_TYPE_PERM:
1988 1989
		del_timer(&br->multicast_router_timer);
		/* fall through */
1990
	case MDB_RTR_TYPE_TEMP_QUERY:
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
		br->multicast_router = val;
		err = 0;
		break;
	}

	spin_unlock_bh(&br->multicast_lock);

	return err;
}

2001 2002 2003 2004 2005 2006 2007 2008
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);
}

2009 2010 2011
int br_multicast_set_port_router(struct net_bridge_port *p, unsigned long val)
{
	struct net_bridge *br = p->br;
2012
	unsigned long now = jiffies;
2013
	int err = -EINVAL;
2014 2015

	spin_lock(&br->multicast_lock);
2016
	if (p->multicast_router == val) {
2017 2018 2019 2020
		/* 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);
2021 2022 2023
		err = 0;
		goto unlock;
	}
2024
	switch (val) {
2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035
	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;
2036 2037 2038
		del_timer(&p->multicast_router_timer);
		br_multicast_add_router(br, p);
		break;
2039 2040 2041 2042
	case MDB_RTR_TYPE_TEMP:
		p->multicast_router = MDB_RTR_TYPE_TEMP;
		br_multicast_mark_router(br, p);
		break;
2043 2044
	default:
		goto unlock;
2045
	}
2046 2047
	err = 0;
unlock:
2048 2049 2050 2051
	spin_unlock(&br->multicast_lock);

	return err;
}
2052

2053
static void br_multicast_start_querier(struct net_bridge *br,
2054
				       struct bridge_mcast_own_query *query)
2055 2056
{
	struct net_bridge_port *port;
2057

2058
	__br_multicast_open(br, query);
2059 2060 2061 2062 2063 2064

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

2065 2066
		if (query == &br->ip4_own_query)
			br_multicast_enable(&port->ip4_own_query);
2067 2068
#if IS_ENABLED(CONFIG_IPV6)
		else
2069
			br_multicast_enable(&port->ip6_own_query);
2070
#endif
2071 2072 2073 2074 2075
	}
}

int br_multicast_toggle(struct net_bridge *br, unsigned long val)
{
2076
	struct net_bridge_mdb_htable *mdb;
2077 2078
	struct net_bridge_port *port;
	int err = 0;
2079

2080
	spin_lock_bh(&br->multicast_lock);
2081 2082 2083 2084 2085 2086 2087
	if (br->multicast_disabled == !val)
		goto unlock;

	br->multicast_disabled = !val;
	if (br->multicast_disabled)
		goto unlock;

2088 2089 2090
	if (!netif_running(br->dev))
		goto unlock;

2091 2092 2093
	mdb = mlock_dereference(br->mdb, br);
	if (mdb) {
		if (mdb->old) {
2094 2095 2096 2097 2098 2099
			err = -EEXIST;
rollback:
			br->multicast_disabled = !!val;
			goto unlock;
		}

2100
		err = br_mdb_rehash(&br->mdb, mdb->max,
2101 2102 2103 2104 2105
				    br->hash_elasticity);
		if (err)
			goto rollback;
	}

2106 2107 2108
	br_multicast_open(br);
	list_for_each_entry(port, &br->port_list, list)
		__br_multicast_enable_port(port);
2109 2110

unlock:
2111
	spin_unlock_bh(&br->multicast_lock);
2112 2113 2114

	return err;
}
2115

2116 2117
int br_multicast_set_querier(struct net_bridge *br, unsigned long val)
{
2118 2119
	unsigned long max_delay;

2120 2121 2122 2123 2124 2125 2126
	val = !!val;

	spin_lock_bh(&br->multicast_lock);
	if (br->multicast_querier == val)
		goto unlock;

	br->multicast_querier = val;
2127 2128 2129 2130 2131
	if (!val)
		goto unlock;

	max_delay = br->multicast_query_response_interval;

2132 2133
	if (!timer_pending(&br->ip4_other_query.timer))
		br->ip4_other_query.delay_time = jiffies + max_delay;
2134

2135
	br_multicast_start_querier(br, &br->ip4_own_query);
2136 2137

#if IS_ENABLED(CONFIG_IPV6)
2138 2139
	if (!timer_pending(&br->ip6_other_query.timer))
		br->ip6_other_query.delay_time = jiffies + max_delay;
2140

2141
	br_multicast_start_querier(br, &br->ip6_own_query);
2142
#endif
2143 2144 2145 2146 2147 2148 2149

unlock:
	spin_unlock_bh(&br->multicast_lock);

	return 0;
}

2150 2151
int br_multicast_set_hash_max(struct net_bridge *br, unsigned long val)
{
2152
	int err = -EINVAL;
2153
	u32 old;
2154
	struct net_bridge_mdb_htable *mdb;
2155

2156
	spin_lock_bh(&br->multicast_lock);
2157 2158
	if (!is_power_of_2(val))
		goto unlock;
2159 2160 2161

	mdb = mlock_dereference(br->mdb, br);
	if (mdb && val < mdb->size)
2162 2163 2164 2165 2166 2167 2168
		goto unlock;

	err = 0;

	old = br->hash_max;
	br->hash_max = val;

2169 2170
	if (mdb) {
		if (mdb->old) {
2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183
			err = -EEXIST;
rollback:
			br->hash_max = old;
			goto unlock;
		}

		err = br_mdb_rehash(&br->mdb, br->hash_max,
				    br->hash_elasticity);
		if (err)
			goto rollback;
	}

unlock:
2184
	spin_unlock_bh(&br->multicast_lock);
2185 2186 2187

	return err;
}
2188

2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206
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;
}

2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226
#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

2227 2228 2229 2230 2231 2232 2233 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 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282
/**
 * 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);
2283

2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320
/**
 * 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);

2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351
/**
 * 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;
2352
#if IS_ENABLED(CONFIG_IPV6)
2353 2354 2355 2356 2357
	case ETH_P_IPV6:
		if (!timer_pending(&br->ip6_other_query.timer) ||
		    rcu_dereference(br->ip6_querier.port) == port)
			goto unlock;
		break;
2358
#endif
2359 2360 2361 2362 2363 2364 2365 2366 2367 2368
	default:
		goto unlock;
	}

	ret = true;
unlock:
	rcu_read_unlock();
	return ret;
}
EXPORT_SYMBOL_GPL(br_multicast_has_querier_adjacent);
2369 2370

static void br_mcast_stats_add(struct bridge_mcast_stats __percpu *stats,
2371
			       const struct sk_buff *skb, u8 type, u8 dir)
2372 2373
{
	struct bridge_mcast_stats *pstats = this_cpu_ptr(stats);
2374 2375
	__be16 proto = skb->protocol;
	unsigned int t_len;
2376 2377 2378 2379

	u64_stats_update_begin(&pstats->syncp);
	switch (proto) {
	case htons(ETH_P_IP):
2380
		t_len = ntohs(ip_hdr(skb)->tot_len) - ip_hdrlen(skb);
2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391
		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:
2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406
			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]++;
			}
2407 2408 2409 2410 2411 2412 2413 2414
			break;
		case IGMP_HOST_LEAVE_MESSAGE:
			pstats->mstats.igmp_leaves[dir]++;
			break;
		}
		break;
#if IS_ENABLED(CONFIG_IPV6)
	case htons(ETH_P_IPV6):
2415 2416 2417
		t_len = ntohs(ipv6_hdr(skb)->payload_len) +
			sizeof(struct ipv6hdr);
		t_len -= skb_network_header_len(skb);
2418 2419 2420 2421 2422 2423 2424 2425
		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:
2426 2427 2428 2429
			if (t_len != sizeof(struct mld_msg))
				pstats->mstats.mld_v2queries[dir]++;
			else
				pstats->mstats.mld_v1queries[dir]++;
2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441
			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,
2442
			const struct sk_buff *skb, u8 type, u8 dir)
2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456
{
	struct bridge_mcast_stats __percpu *stats;

	/* if multicast_disabled is true then igmp type can't be set */
	if (!type || !br->multicast_stats_enabled)
		return;

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

2457
	br_mcast_stats_add(stats, skb, type, dir);
2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501
}

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;
}

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));

2502 2503 2504
		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);
2505 2506 2507 2508 2509 2510
		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;

2511 2512
		mcast_stats_add_dir(tdst.mld_v1queries, temp.mld_v1queries);
		mcast_stats_add_dir(tdst.mld_v2queries, temp.mld_v2queries);
2513 2514 2515 2516 2517 2518 2519
		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));
}