br_multicast.c 57.7 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)
461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479
{
	struct sk_buff *skb;
	struct ipv6hdr *ip6h;
	struct mld_msg *mldq;
	struct ethhdr *eth;
	u8 *hopopt;
	unsigned long interval;

	skb = netdev_alloc_skb_ip_align(br->dev, sizeof(*eth) + sizeof(*ip6h) +
						 8 + sizeof(*mldq));
	if (!skb)
		goto out;

	skb->protocol = htons(ETH_P_IPV6);

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

480
	ether_addr_copy(eth->h_source, br->dev->dev_addr);
481 482 483 484 485 486 487 488
	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);
489
	ip6h->payload_len = htons(8 + sizeof(*mldq));
490 491
	ip6h->nexthdr = IPPROTO_HOPOPTS;
	ip6h->hop_limit = 1;
492
	ipv6_addr_set(&ip6h->daddr, htonl(0xff020000), 0, 0, htonl(1));
493 494 495
	if (ipv6_dev_get_saddr(dev_net(br->dev), br->dev, &ip6h->daddr, 0,
			       &ip6h->saddr)) {
		kfree_skb(skb);
496
		br->has_ipv6_addr = 0;
497 498
		return NULL;
	}
499 500

	br->has_ipv6_addr = 1;
501
	ipv6_eth_mc_map(&ip6h->daddr, eth->h_dest);
502 503 504 505 506 507 508 509

	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;
510 511
	hopopt[6] = IPV6_TLV_PAD1;		/* Pad1 */
	hopopt[7] = IPV6_TLV_PAD1;		/* Pad1 */
512 513 514 515 516 517 518

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

	/* ICMPv6 */
	skb_set_transport_header(skb, skb->len);
	mldq = (struct mld_msg *) icmp6_hdr(skb);

519
	interval = ipv6_addr_any(grp) ?
520 521
			br->multicast_query_response_interval :
			br->multicast_last_member_interval;
522

523
	*igmp_type = ICMPV6_MGM_QUERY;
524 525 526 527 528
	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;
529
	mldq->mld_mca = *grp;
530 531 532 533 534 535 536 537 538 539 540 541 542 543 544

	/* checksum */
	mldq->mld_cksum = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
					  sizeof(*mldq), IPPROTO_ICMPV6,
					  csum_partial(mldq,
						       sizeof(*mldq), 0));
	skb_put(skb, sizeof(*mldq));

	__skb_pull(skb, sizeof(*eth));

out:
	return skb;
}
#endif

545
static struct sk_buff *br_multicast_alloc_query(struct net_bridge *br,
546 547
						struct br_ip *addr,
						u8 *igmp_type)
548 549 550
{
	switch (addr->proto) {
	case htons(ETH_P_IP):
551
		return br_ip4_multicast_alloc_query(br, addr->u.ip4, igmp_type);
E
Eric Dumazet 已提交
552
#if IS_ENABLED(CONFIG_IPV6)
553
	case htons(ETH_P_IPV6):
554 555
		return br_ip6_multicast_alloc_query(br, &addr->u.ip6,
						    igmp_type);
556
#endif
557 558 559 560
	}
	return NULL;
}

561
static struct net_bridge_mdb_entry *br_multicast_get_group(
562 563
	struct net_bridge *br, struct net_bridge_port *port,
	struct br_ip *group, int hash)
564
{
565
	struct net_bridge_mdb_htable *mdb;
566
	struct net_bridge_mdb_entry *mp;
567 568
	unsigned int count = 0;
	unsigned int max;
569 570 571
	int elasticity;
	int err;

572
	mdb = rcu_dereference_protected(br->mdb, 1);
573
	hlist_for_each_entry(mp, &mdb->mhash[hash], hlist[mdb->ver]) {
574
		count++;
575
		if (unlikely(br_ip_equal(group, &mp->addr)))
576 577 578 579 580 581 582 583
			return mp;
	}

	elasticity = 0;
	max = mdb->max;

	if (unlikely(count > br->hash_elasticity && count)) {
		if (net_ratelimit())
584 585 586
			br_info(br, "Multicast hash table "
				"chain limit reached: %s\n",
				port ? port->dev->name : br->dev->name);
587 588 589 590 591 592

		elasticity = br->hash_elasticity;
	}

	if (mdb->size >= max) {
		max *= 2;
593 594 595 596 597
		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);
598 599 600 601 602 603 604 605 606 607
			err = -E2BIG;
disable:
			br->multicast_disabled = 1;
			goto err;
		}
	}

	if (max > mdb->max || elasticity) {
		if (mdb->old) {
			if (net_ratelimit())
608 609 610
				br_info(br, "Multicast hash table "
					"on fire: %s\n",
					port ? port->dev->name : br->dev->name);
611 612 613 614 615 616
			err = -EEXIST;
			goto err;
		}

		err = br_mdb_rehash(&br->mdb, max, elasticity);
		if (err) {
617 618 619 620
			br_warn(br, "Cannot rehash multicast "
				"hash table, disabling snooping: %s, %d, %d\n",
				port ? port->dev->name : br->dev->name,
				mdb->size, err);
621 622 623 624 625 626 627 628 629 630 631 632 633 634
			goto disable;
		}

		err = -EAGAIN;
		goto err;
	}

	return NULL;

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

635
struct net_bridge_mdb_entry *br_multicast_new_group(struct net_bridge *br,
636 637
						    struct net_bridge_port *p,
						    struct br_ip *group)
638
{
639
	struct net_bridge_mdb_htable *mdb;
640 641
	struct net_bridge_mdb_entry *mp;
	int hash;
642
	int err;
643

644
	mdb = rcu_dereference_protected(br->mdb, 1);
645
	if (!mdb) {
646 647 648
		err = br_mdb_rehash(&br->mdb, BR_HASH_SIZE, 0);
		if (err)
			return ERR_PTR(err);
649 650 651 652
		goto rehash;
	}

	hash = br_ip_hash(mdb, group);
653
	mp = br_multicast_get_group(br, p, group, hash);
654 655 656 657 658 659
	switch (PTR_ERR(mp)) {
	case 0:
		break;

	case -EAGAIN:
rehash:
660
		mdb = rcu_dereference_protected(br->mdb, 1);
661 662 663 664 665 666 667 668 669
		hash = br_ip_hash(mdb, group);
		break;

	default:
		goto out;
	}

	mp = kzalloc(sizeof(*mp), GFP_ATOMIC);
	if (unlikely(!mp))
670
		return ERR_PTR(-ENOMEM);
671 672

	mp->br = br;
673
	mp->addr = *group;
674 675 676
	setup_timer(&mp->timer, br_multicast_group_expired,
		    (unsigned long)mp);

677 678 679 680 681 682 683
	hlist_add_head_rcu(&mp->hlist[mdb->ver], &mdb->mhash[hash]);
	mdb->size++;

out:
	return mp;
}

684 685 686
struct net_bridge_port_group *br_multicast_new_port_group(
			struct net_bridge_port *port,
			struct br_ip *group,
687
			struct net_bridge_port_group __rcu *next,
688
			unsigned char flags)
689 690 691 692 693 694 695 696 697
{
	struct net_bridge_port_group *p;

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

	p->addr = *group;
	p->port = port;
698
	p->flags = flags;
699
	rcu_assign_pointer(p->next, next);
700 701 702 703 704 705
	hlist_add_head(&p->mglist, &port->mglist);
	setup_timer(&p->timer, br_multicast_port_group_expired,
		    (unsigned long)p);
	return p;
}

706
static int br_multicast_add_group(struct net_bridge *br,
707 708
				  struct net_bridge_port *port,
				  struct br_ip *group)
709
{
710
	struct net_bridge_port_group __rcu **pp;
711 712
	struct net_bridge_port_group *p;
	struct net_bridge_mdb_entry *mp;
713
	unsigned long now = jiffies;
714 715 716 717 718 719 720 721 722
	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);
723
	if (IS_ERR(mp))
724 725 726
		goto err;

	if (!port) {
727
		mp->mglist = true;
728
		mod_timer(&mp->timer, now + br->multicast_membership_interval);
729 730 731
		goto out;
	}

732 733 734
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
735
		if (p->port == port)
736
			goto found;
737 738 739 740
		if ((unsigned long)p->port < (unsigned long)port)
			break;
	}

741
	p = br_multicast_new_port_group(port, group, *pp, 0);
742 743 744
	if (unlikely(!p))
		goto err;
	rcu_assign_pointer(*pp, p);
745
	br_mdb_notify(br->dev, port, group, RTM_NEWMDB, 0);
746

747 748
found:
	mod_timer(&p->timer, now + br->multicast_membership_interval);
749 750 751 752 753 754 755 756
out:
	err = 0;

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

757 758
static int br_ip4_multicast_add_group(struct net_bridge *br,
				      struct net_bridge_port *port,
759 760
				      __be32 group,
				      __u16 vid)
761 762 763 764 765 766 767 768
{
	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);
769
	br_group.vid = vid;
770 771 772 773

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

E
Eric Dumazet 已提交
774
#if IS_ENABLED(CONFIG_IPV6)
775 776
static int br_ip6_multicast_add_group(struct net_bridge *br,
				      struct net_bridge_port *port,
777 778
				      const struct in6_addr *group,
				      __u16 vid)
779 780 781
{
	struct br_ip br_group;

782
	if (ipv6_addr_is_ll_all_nodes(group))
783 784
		return 0;

A
Alexey Dobriyan 已提交
785
	br_group.u.ip6 = *group;
786
	br_group.proto = htons(ETH_P_IPV6);
787
	br_group.vid = vid;
788 789 790 791 792

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

793 794 795 796 797 798
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);
799 800
	if (port->multicast_router == MDB_RTR_TYPE_DISABLED ||
	    port->multicast_router == MDB_RTR_TYPE_PERM ||
801 802 803 804 805
	    timer_pending(&port->multicast_router_timer) ||
	    hlist_unhashed(&port->rlist))
		goto out;

	hlist_del_init_rcu(&port->rlist);
806
	br_rtr_notify(br->dev, port, RTM_DELMDB);
807 808 809
	/* 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;
810 811 812 813 814 815 816 817 818

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

static void br_multicast_local_router_expired(unsigned long data)
{
}

819
static void br_multicast_querier_expired(struct net_bridge *br,
820
					 struct bridge_mcast_own_query *query)
821 822 823 824 825
{
	spin_lock(&br->multicast_lock);
	if (!netif_running(br->dev) || br->multicast_disabled)
		goto out;

826
	br_multicast_start_querier(br, query);
827 828 829 830 831

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

832 833 834 835
static void br_ip4_multicast_querier_expired(unsigned long data)
{
	struct net_bridge *br = (void *)data;

836
	br_multicast_querier_expired(br, &br->ip4_own_query);
837 838 839 840 841 842 843
}

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

844
	br_multicast_querier_expired(br, &br->ip6_own_query);
845 846 847
}
#endif

848 849 850 851 852 853 854 855 856 857 858 859
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
}

860 861 862
static void __br_multicast_send_query(struct net_bridge *br,
				      struct net_bridge_port *port,
				      struct br_ip *ip)
863 864
{
	struct sk_buff *skb;
865
	u8 igmp_type;
866

867
	skb = br_multicast_alloc_query(br, ip, &igmp_type);
868
	if (!skb)
869
		return;
870 871 872

	if (port) {
		skb->dev = port->dev;
873
		br_multicast_count(br, port, skb, igmp_type,
874
				   BR_MCAST_DIR_TX);
875 876
		NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_OUT,
			dev_net(port->dev), NULL, skb, NULL, skb->dev,
877
			br_dev_queue_push_xmit);
878 879
	} else {
		br_multicast_select_own_querier(br, ip, skb);
880
		br_multicast_count(br, port, skb, igmp_type,
881
				   BR_MCAST_DIR_RX);
882
		netif_rx(skb);
883
	}
884 885 886
}

static void br_multicast_send_query(struct net_bridge *br,
887
				    struct net_bridge_port *port,
888
				    struct bridge_mcast_own_query *own_query)
889
{
890
	struct bridge_mcast_other_query *other_query = NULL;
891 892
	struct br_ip br_group;
	unsigned long time;
893 894

	if (!netif_running(br->dev) || br->multicast_disabled ||
895
	    !br->multicast_querier)
896 897
		return;

898 899
	memset(&br_group.u, 0, sizeof(br_group.u));

900 901 902
	if (port ? (own_query == &port->ip4_own_query) :
		   (own_query == &br->ip4_own_query)) {
		other_query = &br->ip4_other_query;
903
		br_group.proto = htons(ETH_P_IP);
E
Eric Dumazet 已提交
904
#if IS_ENABLED(CONFIG_IPV6)
905
	} else {
906
		other_query = &br->ip6_other_query;
907
		br_group.proto = htons(ETH_P_IPV6);
908
#endif
909 910
	}

911
	if (!other_query || timer_pending(&other_query->timer))
912 913 914
		return;

	__br_multicast_send_query(br, port, &br_group);
915 916

	time = jiffies;
917
	time += own_query->startup_sent < br->multicast_startup_query_count ?
918 919
		br->multicast_startup_query_interval :
		br->multicast_query_interval;
920
	mod_timer(&own_query->timer, time);
921 922
}

923 924 925
static void
br_multicast_port_query_expired(struct net_bridge_port *port,
				struct bridge_mcast_own_query *query)
926 927 928 929
{
	struct net_bridge *br = port->br;

	spin_lock(&br->multicast_lock);
930 931
	if (port->state == BR_STATE_DISABLED ||
	    port->state == BR_STATE_BLOCKING)
932 933
		goto out;

934 935
	if (query->startup_sent < br->multicast_startup_query_count)
		query->startup_sent++;
936

937
	br_multicast_send_query(port->br, port, query);
938 939 940 941 942

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

943 944 945 946
static void br_ip4_multicast_port_query_expired(unsigned long data)
{
	struct net_bridge_port *port = (void *)data;

947
	br_multicast_port_query_expired(port, &port->ip4_own_query);
948 949 950 951 952 953 954
}

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

955
	br_multicast_port_query_expired(port, &port->ip6_own_query);
956 957 958
}
#endif

959
int br_multicast_add_port(struct net_bridge_port *port)
960
{
961
	port->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
962 963 964

	setup_timer(&port->multicast_router_timer, br_multicast_router_expired,
		    (unsigned long)port);
965 966
	setup_timer(&port->ip4_own_query.timer,
		    br_ip4_multicast_port_query_expired, (unsigned long)port);
967
#if IS_ENABLED(CONFIG_IPV6)
968 969
	setup_timer(&port->ip6_own_query.timer,
		    br_ip6_multicast_port_query_expired, (unsigned long)port);
970
#endif
971 972 973 974 975
	port->mcast_stats = netdev_alloc_pcpu_stats(struct bridge_mcast_stats);
	if (!port->mcast_stats)
		return -ENOMEM;

	return 0;
976 977 978 979
}

void br_multicast_del_port(struct net_bridge_port *port)
{
980 981 982 983 984 985 986 987 988
	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);
989
	del_timer_sync(&port->multicast_router_timer);
990
	free_percpu(port->mcast_stats);
991 992
}

993
static void br_multicast_enable(struct bridge_mcast_own_query *query)
994
{
995
	query->startup_sent = 0;
996

997 998 999
	if (try_to_del_timer_sync(&query->timer) >= 0 ||
	    del_timer(&query->timer))
		mod_timer(&query->timer, jiffies);
1000 1001
}

1002
static void __br_multicast_enable_port(struct net_bridge_port *port)
1003 1004 1005 1006
{
	struct net_bridge *br = port->br;

	if (br->multicast_disabled || !netif_running(br->dev))
1007
		return;
1008

1009
	br_multicast_enable(&port->ip4_own_query);
1010
#if IS_ENABLED(CONFIG_IPV6)
1011
	br_multicast_enable(&port->ip6_own_query);
1012
#endif
1013 1014
	if (port->multicast_router == MDB_RTR_TYPE_PERM &&
	    hlist_unhashed(&port->rlist))
1015
		br_multicast_add_router(br, port);
1016
}
1017

1018 1019 1020 1021 1022 1023
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);
1024 1025 1026 1027 1028 1029 1030
	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;
1031
	struct hlist_node *n;
1032 1033

	spin_lock(&br->multicast_lock);
1034
	hlist_for_each_entry_safe(pg, n, &port->mglist, mglist)
1035
		if (!(pg->flags & MDB_PG_FLAGS_PERMANENT))
1036
			br_multicast_del_pg(br, pg);
1037

1038
	if (!hlist_unhashed(&port->rlist)) {
1039
		hlist_del_init_rcu(&port->rlist);
1040
		br_rtr_notify(br->dev, port, RTM_DELMDB);
1041 1042 1043
		/* Don't allow timer refresh if disabling */
		if (port->multicast_router == MDB_RTR_TYPE_TEMP)
			port->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
1044
	}
1045
	del_timer(&port->multicast_router_timer);
1046
	del_timer(&port->ip4_own_query.timer);
1047
#if IS_ENABLED(CONFIG_IPV6)
1048
	del_timer(&port->ip6_own_query.timer);
1049
#endif
1050 1051 1052
	spin_unlock(&br->multicast_lock);
}

1053 1054
static int br_ip4_multicast_igmp3_report(struct net_bridge *br,
					 struct net_bridge_port *port,
1055 1056
					 struct sk_buff *skb,
					 u16 vid)
1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068
{
	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);
1069
	len = skb_transport_offset(skb) + sizeof(*ih);
1070 1071 1072 1073 1074 1075

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

H
Herbert Xu 已提交
1076
		grec = (void *)(skb->data + len - sizeof(*grec));
1077 1078 1079
		group = grec->grec_mca;
		type = grec->grec_type;

E
Eric Dumazet 已提交
1080
		len += ntohs(grec->grec_nsrcs) * 4;
1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
		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;
		}

1098 1099 1100 1101 1102 1103 1104 1105 1106
		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;
		}
1107 1108 1109 1110 1111
	}

	return err;
}

E
Eric Dumazet 已提交
1112
#if IS_ENABLED(CONFIG_IPV6)
1113 1114
static int br_ip6_multicast_mld2_report(struct net_bridge *br,
					struct net_bridge_port *port,
1115 1116
					struct sk_buff *skb,
					u16 vid)
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129
{
	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]);
1130
	len = skb_transport_offset(skb) + sizeof(*icmp6h);
1131 1132 1133 1134 1135 1136

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

		nsrcs = skb_header_pointer(skb,
					   len + offsetof(struct mld2_grec,
1137
							  grec_nsrcs),
1138 1139 1140 1141 1142 1143
					   sizeof(_nsrcs), &_nsrcs);
		if (!nsrcs)
			return -EINVAL;

		if (!pskb_may_pull(skb,
				   len + sizeof(*grec) +
1144
				   sizeof(struct in6_addr) * ntohs(*nsrcs)))
1145 1146 1147
			return -EINVAL;

		grec = (struct mld2_grec *)(skb->data + len);
1148 1149
		len += sizeof(*grec) +
		       sizeof(struct in6_addr) * ntohs(*nsrcs);
1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164

		/* 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;
		}

1165 1166 1167 1168 1169 1170 1171 1172
		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);
1173
			if (err)
1174 1175
				break;
		}
1176 1177 1178 1179 1180 1181
	}

	return err;
}
#endif

1182
static bool br_ip4_multicast_select_querier(struct net_bridge *br,
1183
					    struct net_bridge_port *port,
1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200
					    __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;

1201 1202 1203
	/* update protected by general multicast_lock by caller */
	rcu_assign_pointer(br->ip4_querier.port, port);

1204 1205 1206 1207 1208
	return true;
}

#if IS_ENABLED(CONFIG_IPV6)
static bool br_ip6_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
					    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;

1224 1225 1226
	/* update protected by general multicast_lock by caller */
	rcu_assign_pointer(br->ip6_querier.port, port);

1227 1228 1229 1230 1231
	return true;
}
#endif

static bool br_multicast_select_querier(struct net_bridge *br,
1232
					struct net_bridge_port *port,
1233 1234 1235 1236
					struct br_ip *saddr)
{
	switch (saddr->proto) {
	case htons(ETH_P_IP):
1237
		return br_ip4_multicast_select_querier(br, port, saddr->u.ip4);
1238 1239
#if IS_ENABLED(CONFIG_IPV6)
	case htons(ETH_P_IPV6):
1240
		return br_ip6_multicast_select_querier(br, port, &saddr->u.ip6);
1241 1242 1243 1244 1245 1246
#endif
	}

	return false;
}

1247
static void
1248 1249 1250
br_multicast_update_query_timer(struct net_bridge *br,
				struct bridge_mcast_other_query *query,
				unsigned long max_delay)
1251
{
1252 1253
	if (!timer_pending(&query->timer))
		query->delay_time = jiffies + max_delay;
1254

1255
	mod_timer(&query->timer, jiffies + br->multicast_querier_interval);
1256 1257
}

1258
/*
C
Cong Wang 已提交
1259
 * Add port to router_list
1260 1261 1262
 *  list is maintained ordered by pointer value
 *  and locked by br->multicast_lock and RCU
 */
1263 1264 1265
static void br_multicast_add_router(struct net_bridge *br,
				    struct net_bridge_port *port)
{
1266
	struct net_bridge_port *p;
1267
	struct hlist_node *slot = NULL;
1268

1269 1270 1271
	if (!hlist_unhashed(&port->rlist))
		return;

1272
	hlist_for_each_entry(p, &br->router_list, rlist) {
1273 1274
		if ((unsigned long) port >= (unsigned long) p)
			break;
1275
		slot = &p->rlist;
1276 1277
	}

1278
	if (slot)
1279
		hlist_add_behind_rcu(&port->rlist, slot);
1280 1281
	else
		hlist_add_head_rcu(&port->rlist, &br->router_list);
1282
	br_rtr_notify(br->dev, port, RTM_NEWMDB);
1283 1284
}

1285 1286 1287 1288 1289 1290
static void br_multicast_mark_router(struct net_bridge *br,
				     struct net_bridge_port *port)
{
	unsigned long now = jiffies;

	if (!port) {
1291
		if (br->multicast_router == MDB_RTR_TYPE_TEMP_QUERY)
1292 1293 1294 1295 1296
			mod_timer(&br->multicast_router_timer,
				  now + br->multicast_querier_interval);
		return;
	}

1297 1298
	if (port->multicast_router == MDB_RTR_TYPE_DISABLED ||
	    port->multicast_router == MDB_RTR_TYPE_PERM)
1299 1300
		return;

1301
	br_multicast_add_router(br, port);
1302 1303 1304 1305 1306 1307 1308

	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,
1309
					struct bridge_mcast_other_query *query,
1310
					struct br_ip *saddr,
1311
					unsigned long max_delay)
1312
{
1313
	if (!br_multicast_select_querier(br, port, saddr))
1314 1315
		return;

1316
	br_multicast_update_query_timer(br, query, max_delay);
1317 1318 1319
	br_multicast_mark_router(br, port);
}

1320 1321
static int br_ip4_multicast_query(struct net_bridge *br,
				  struct net_bridge_port *port,
1322 1323
				  struct sk_buff *skb,
				  u16 vid)
1324
{
1325
	const struct iphdr *iph = ip_hdr(skb);
1326 1327 1328 1329
	struct igmphdr *ih = igmp_hdr(skb);
	struct net_bridge_mdb_entry *mp;
	struct igmpv3_query *ih3;
	struct net_bridge_port_group *p;
1330
	struct net_bridge_port_group __rcu **pp;
1331
	struct br_ip saddr;
1332 1333
	unsigned long max_delay;
	unsigned long now = jiffies;
1334
	unsigned int offset = skb_transport_offset(skb);
1335
	__be32 group;
1336
	int err = 0;
1337 1338 1339 1340 1341 1342 1343 1344

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

	group = ih->group;

1345
	if (skb->len == offset + sizeof(*ih)) {
1346 1347 1348 1349 1350 1351
		max_delay = ih->code * (HZ / IGMP_TIMER_SCALE);

		if (!max_delay) {
			max_delay = 10 * HZ;
			group = 0;
		}
1352
	} else if (skb->len >= offset + sizeof(*ih3)) {
1353 1354
		ih3 = igmpv3_query_hdr(skb);
		if (ih3->nsrcs)
1355
			goto out;
1356

1357 1358
		max_delay = ih3->code ?
			    IGMPV3_MRC(ih3->code) * (HZ / IGMP_TIMER_SCALE) : 1;
1359
	} else {
1360 1361 1362
		goto out;
	}

1363 1364 1365
	if (!group) {
		saddr.proto = htons(ETH_P_IP);
		saddr.u.ip4 = iph->saddr;
1366

1367 1368
		br_multicast_query_received(br, port, &br->ip4_other_query,
					    &saddr, max_delay);
1369
		goto out;
1370
	}
1371

1372
	mp = br_mdb_ip4_get(mlock_dereference(br->mdb, br), group, vid);
1373 1374 1375 1376 1377
	if (!mp)
		goto out;

	max_delay *= br->multicast_last_member_count;

1378
	if (mp->mglist &&
1379 1380 1381 1382 1383
	    (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);

1384 1385 1386
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
1387 1388 1389
		if (timer_pending(&p->timer) ?
		    time_after(p->timer.expires, now + max_delay) :
		    try_to_del_timer_sync(&p->timer) >= 0)
1390
			mod_timer(&p->timer, now + max_delay);
1391 1392 1393 1394
	}

out:
	spin_unlock(&br->multicast_lock);
1395
	return err;
1396 1397
}

E
Eric Dumazet 已提交
1398
#if IS_ENABLED(CONFIG_IPV6)
1399 1400
static int br_ip6_multicast_query(struct net_bridge *br,
				  struct net_bridge_port *port,
1401 1402
				  struct sk_buff *skb,
				  u16 vid)
1403
{
1404
	const struct ipv6hdr *ip6h = ipv6_hdr(skb);
1405
	struct mld_msg *mld;
1406 1407
	struct net_bridge_mdb_entry *mp;
	struct mld2_query *mld2q;
1408 1409
	struct net_bridge_port_group *p;
	struct net_bridge_port_group __rcu **pp;
1410
	struct br_ip saddr;
1411 1412
	unsigned long max_delay;
	unsigned long now = jiffies;
1413
	unsigned int offset = skb_transport_offset(skb);
1414
	const struct in6_addr *group = NULL;
1415
	bool is_general_query;
1416 1417 1418 1419 1420 1421 1422
	int err = 0;

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

1423 1424
	if (skb->len == offset + sizeof(*mld)) {
		if (!pskb_may_pull(skb, offset + sizeof(*mld))) {
1425 1426 1427 1428
			err = -EINVAL;
			goto out;
		}
		mld = (struct mld_msg *) icmp6_hdr(skb);
L
Li RongQing 已提交
1429
		max_delay = msecs_to_jiffies(ntohs(mld->mld_maxdelay));
1430 1431
		if (max_delay)
			group = &mld->mld_mca;
1432
	} else {
1433
		if (!pskb_may_pull(skb, offset + sizeof(*mld2q))) {
1434 1435 1436 1437 1438 1439
			err = -EINVAL;
			goto out;
		}
		mld2q = (struct mld2_query *)icmp6_hdr(skb);
		if (!mld2q->mld2q_nsrcs)
			group = &mld2q->mld2q_mca;
1440 1441

		max_delay = max(msecs_to_jiffies(mldv2_mrc(mld2q)), 1UL);
1442 1443
	}

1444 1445
	is_general_query = group && ipv6_addr_any(group);

1446 1447 1448
	if (is_general_query) {
		saddr.proto = htons(ETH_P_IPV6);
		saddr.u.ip6 = ip6h->saddr;
1449

1450 1451
		br_multicast_query_received(br, port, &br->ip6_other_query,
					    &saddr, max_delay);
1452
		goto out;
1453 1454
	} else if (!group) {
		goto out;
1455
	}
1456

1457
	mp = br_mdb_ip6_get(mlock_dereference(br->mdb, br), group, vid);
1458 1459 1460 1461
	if (!mp)
		goto out;

	max_delay *= br->multicast_last_member_count;
1462
	if (mp->mglist &&
1463 1464 1465 1466 1467
	    (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);

1468 1469 1470
	for (pp = &mp->ports;
	     (p = mlock_dereference(*pp, br)) != NULL;
	     pp = &p->next) {
1471 1472 1473
		if (timer_pending(&p->timer) ?
		    time_after(p->timer.expires, now + max_delay) :
		    try_to_del_timer_sync(&p->timer) >= 0)
1474
			mod_timer(&p->timer, now + max_delay);
1475 1476 1477 1478 1479 1480 1481 1482
	}

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

1483 1484 1485 1486 1487 1488
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)
1489 1490 1491 1492 1493 1494 1495 1496 1497
{
	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) ||
1498
	    (port && port->state == BR_STATE_DISABLED))
1499 1500
		goto out;

1501
	mdb = mlock_dereference(br->mdb, br);
1502 1503 1504 1505
	mp = br_mdb_ip_get(mdb, group);
	if (!mp)
		goto out;

1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518
	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);
1519 1520
			br_mdb_notify(br->dev, port, group, RTM_DELMDB,
				      p->flags);
1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531

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

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

1532
	if (br->multicast_querier) {
1533 1534 1535 1536
		__br_multicast_send_query(br, port, &mp->addr);

		time = jiffies + br->multicast_last_member_count *
				 br->multicast_last_member_interval;
1537

1538
		mod_timer(&own_query->timer, time);
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556

		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;
		}
	}

1557 1558 1559 1560 1561
	now = jiffies;
	time = now + br->multicast_last_member_count *
		     br->multicast_last_member_interval;

	if (!port) {
1562
		if (mp->mglist &&
1563 1564 1565 1566 1567
		    (timer_pending(&mp->timer) ?
		     time_after(mp->timer.expires, time) :
		     try_to_del_timer_sync(&mp->timer) >= 0)) {
			mod_timer(&mp->timer, time);
		}
1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585

		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;
1586 1587 1588 1589 1590
	}
out:
	spin_unlock(&br->multicast_lock);
}

1591 1592
static void br_ip4_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
1593 1594
					 __be32 group,
					 __u16 vid)
1595 1596
{
	struct br_ip br_group;
1597
	struct bridge_mcast_own_query *own_query;
1598 1599 1600 1601

	if (ipv4_is_local_multicast(group))
		return;

1602 1603
	own_query = port ? &port->ip4_own_query : &br->ip4_own_query;

1604 1605
	br_group.u.ip4 = group;
	br_group.proto = htons(ETH_P_IP);
1606
	br_group.vid = vid;
1607

1608 1609
	br_multicast_leave_group(br, port, &br_group, &br->ip4_other_query,
				 own_query);
1610 1611
}

E
Eric Dumazet 已提交
1612
#if IS_ENABLED(CONFIG_IPV6)
1613 1614
static void br_ip6_multicast_leave_group(struct net_bridge *br,
					 struct net_bridge_port *port,
1615 1616
					 const struct in6_addr *group,
					 __u16 vid)
1617 1618
{
	struct br_ip br_group;
1619
	struct bridge_mcast_own_query *own_query;
1620

1621
	if (ipv6_addr_is_ll_all_nodes(group))
1622 1623
		return;

1624 1625
	own_query = port ? &port->ip6_own_query : &br->ip6_own_query;

A
Alexey Dobriyan 已提交
1626
	br_group.u.ip6 = *group;
1627
	br_group.proto = htons(ETH_P_IPV6);
1628
	br_group.vid = vid;
1629

1630 1631
	br_multicast_leave_group(br, port, &br_group, &br->ip6_other_query,
				 own_query);
1632 1633
}
#endif
1634

1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667
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);
}

1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682
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);
}

1683 1684
static int br_multicast_ipv4_rcv(struct net_bridge *br,
				 struct net_bridge_port *port,
1685 1686
				 struct sk_buff *skb,
				 u16 vid)
1687
{
1688
	struct sk_buff *skb_trimmed = NULL;
1689 1690 1691
	struct igmphdr *ih;
	int err;

1692
	err = ip_mc_check_igmp(skb, &skb_trimmed);
1693

1694
	if (err == -ENOMSG) {
1695
		if (!ipv4_is_local_multicast(ip_hdr(skb)->daddr)) {
1696
			BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1697 1698 1699 1700
		} else if (pim_ipv4_all_pim_routers(ip_hdr(skb)->daddr)) {
			if (ip_hdr(skb)->protocol == IPPROTO_PIM)
				br_multicast_pim(br, port, skb);
		}
1701
		return 0;
1702
	} else if (err < 0) {
1703
		br_multicast_err_count(br, port, skb->protocol);
1704
		return err;
1705
	}
1706

1707
	ih = igmp_hdr(skb);
1708
	BR_INPUT_SKB_CB(skb)->igmp = ih->type;
1709 1710 1711 1712

	switch (ih->type) {
	case IGMP_HOST_MEMBERSHIP_REPORT:
	case IGMPV2_HOST_MEMBERSHIP_REPORT:
1713
		BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1714
		err = br_ip4_multicast_add_group(br, port, ih->group, vid);
1715 1716
		break;
	case IGMPV3_HOST_MEMBERSHIP_REPORT:
1717
		err = br_ip4_multicast_igmp3_report(br, port, skb_trimmed, vid);
1718 1719
		break;
	case IGMP_HOST_MEMBERSHIP_QUERY:
1720
		err = br_ip4_multicast_query(br, port, skb_trimmed, vid);
1721 1722
		break;
	case IGMP_HOST_LEAVE_MESSAGE:
1723
		br_ip4_multicast_leave_group(br, port, ih->group, vid);
1724 1725 1726
		break;
	}

1727
	if (skb_trimmed && skb_trimmed != skb)
1728 1729
		kfree_skb(skb_trimmed);

1730
	br_multicast_count(br, port, skb, BR_INPUT_SKB_CB(skb)->igmp,
1731 1732
			   BR_MCAST_DIR_RX);

1733 1734 1735
	return err;
}

E
Eric Dumazet 已提交
1736
#if IS_ENABLED(CONFIG_IPV6)
1737 1738
static int br_multicast_ipv6_rcv(struct net_bridge *br,
				 struct net_bridge_port *port,
1739 1740
				 struct sk_buff *skb,
				 u16 vid)
1741
{
1742 1743
	struct sk_buff *skb_trimmed = NULL;
	struct mld_msg *mld;
1744 1745
	int err;

1746
	err = ipv6_mc_check_mld(skb, &skb_trimmed);
1747

1748 1749 1750
	if (err == -ENOMSG) {
		if (!ipv6_addr_is_ll_all_nodes(&ipv6_hdr(skb)->daddr))
			BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1751
		return 0;
1752
	} else if (err < 0) {
1753
		br_multicast_err_count(br, port, skb->protocol);
1754
		return err;
1755 1756
	}

1757
	mld = (struct mld_msg *)skb_transport_header(skb);
1758
	BR_INPUT_SKB_CB(skb)->igmp = mld->mld_type;
1759

1760
	switch (mld->mld_type) {
1761
	case ICMPV6_MGM_REPORT:
1762
		BR_INPUT_SKB_CB(skb)->mrouters_only = 1;
1763
		err = br_ip6_multicast_add_group(br, port, &mld->mld_mca, vid);
1764 1765
		break;
	case ICMPV6_MLD2_REPORT:
1766
		err = br_ip6_multicast_mld2_report(br, port, skb_trimmed, vid);
1767 1768
		break;
	case ICMPV6_MGM_QUERY:
1769
		err = br_ip6_multicast_query(br, port, skb_trimmed, vid);
1770 1771
		break;
	case ICMPV6_MGM_REDUCTION:
1772
		br_ip6_multicast_leave_group(br, port, &mld->mld_mca, vid);
1773
		break;
1774 1775
	}

1776
	if (skb_trimmed && skb_trimmed != skb)
1777 1778
		kfree_skb(skb_trimmed);

1779
	br_multicast_count(br, port, skb, BR_INPUT_SKB_CB(skb)->igmp,
1780 1781
			   BR_MCAST_DIR_RX);

1782 1783 1784 1785
	return err;
}
#endif

1786
int br_multicast_rcv(struct net_bridge *br, struct net_bridge_port *port,
1787
		     struct sk_buff *skb, u16 vid)
1788
{
1789 1790
	int ret = 0;

1791 1792 1793
	BR_INPUT_SKB_CB(skb)->igmp = 0;
	BR_INPUT_SKB_CB(skb)->mrouters_only = 0;

1794 1795 1796 1797 1798
	if (br->multicast_disabled)
		return 0;

	switch (skb->protocol) {
	case htons(ETH_P_IP):
1799 1800
		ret = br_multicast_ipv4_rcv(br, port, skb, vid);
		break;
E
Eric Dumazet 已提交
1801
#if IS_ENABLED(CONFIG_IPV6)
1802
	case htons(ETH_P_IPV6):
1803 1804
		ret = br_multicast_ipv6_rcv(br, port, skb, vid);
		break;
1805
#endif
1806 1807
	}

1808
	return ret;
1809 1810
}

1811
static void br_multicast_query_expired(struct net_bridge *br,
1812 1813
				       struct bridge_mcast_own_query *query,
				       struct bridge_mcast_querier *querier)
1814 1815 1816 1817 1818
{
	spin_lock(&br->multicast_lock);
	if (query->startup_sent < br->multicast_startup_query_count)
		query->startup_sent++;

1819
	RCU_INIT_POINTER(querier->port, NULL);
1820 1821 1822 1823 1824
	br_multicast_send_query(br, NULL, query);
	spin_unlock(&br->multicast_lock);
}

static void br_ip4_multicast_query_expired(unsigned long data)
1825 1826 1827
{
	struct net_bridge *br = (void *)data;

1828
	br_multicast_query_expired(br, &br->ip4_own_query, &br->ip4_querier);
1829
}
1830

1831 1832 1833 1834
#if IS_ENABLED(CONFIG_IPV6)
static void br_ip6_multicast_query_expired(unsigned long data)
{
	struct net_bridge *br = (void *)data;
1835

1836
	br_multicast_query_expired(br, &br->ip6_own_query, &br->ip6_querier);
1837
}
1838
#endif
1839 1840 1841 1842 1843 1844

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

1845
	br->multicast_igmp_version = 2;
1846
	br->multicast_router = MDB_RTR_TYPE_TEMP_QUERY;
1847
	br->multicast_querier = 0;
1848
	br->multicast_query_use_ifaddr = 0;
1849 1850 1851 1852 1853 1854 1855 1856 1857 1858
	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;

1859
	br->ip4_other_query.delay_time = 0;
1860
	br->ip4_querier.port = NULL;
1861
#if IS_ENABLED(CONFIG_IPV6)
1862
	br->ip6_other_query.delay_time = 0;
1863
	br->ip6_querier.port = NULL;
1864
#endif
1865
	br->has_ipv6_addr = 1;
1866

1867 1868 1869
	spin_lock_init(&br->multicast_lock);
	setup_timer(&br->multicast_router_timer,
		    br_multicast_local_router_expired, 0);
1870 1871 1872
	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,
1873
		    (unsigned long)br);
1874
#if IS_ENABLED(CONFIG_IPV6)
1875 1876 1877
	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,
1878 1879
		    (unsigned long)br);
#endif
1880 1881
}

1882
static void __br_multicast_open(struct net_bridge *br,
1883
				struct bridge_mcast_own_query *query)
1884
{
1885
	query->startup_sent = 0;
1886 1887 1888 1889

	if (br->multicast_disabled)
		return;

1890 1891 1892 1893 1894
	mod_timer(&query->timer, jiffies);
}

void br_multicast_open(struct net_bridge *br)
{
1895
	__br_multicast_open(br, &br->ip4_own_query);
1896
#if IS_ENABLED(CONFIG_IPV6)
1897
	__br_multicast_open(br, &br->ip6_own_query);
1898
#endif
1899 1900 1901 1902 1903
}

void br_multicast_stop(struct net_bridge *br)
{
	del_timer_sync(&br->multicast_router_timer);
1904 1905
	del_timer_sync(&br->ip4_other_query.timer);
	del_timer_sync(&br->ip4_own_query.timer);
1906
#if IS_ENABLED(CONFIG_IPV6)
1907 1908
	del_timer_sync(&br->ip6_other_query.timer);
	del_timer_sync(&br->ip6_own_query.timer);
1909
#endif
1910 1911 1912 1913 1914 1915 1916 1917 1918
}

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;
1919 1920

	spin_lock_bh(&br->multicast_lock);
1921
	mdb = mlock_dereference(br->mdb, br);
1922 1923 1924 1925 1926 1927 1928
	if (!mdb)
		goto out;

	br->mdb = NULL;

	ver = mdb->ver;
	for (i = 0; i < mdb->max; i++) {
1929
		hlist_for_each_entry_safe(mp, n, &mdb->mhash[i],
1930 1931 1932 1933 1934 1935 1936 1937
					  hlist[ver]) {
			del_timer(&mp->timer);
			call_rcu_bh(&mp->rcu, br_multicast_free_group);
		}
	}

	if (mdb->old) {
		spin_unlock_bh(&br->multicast_lock);
1938
		rcu_barrier_bh();
1939 1940 1941 1942 1943 1944 1945 1946 1947
		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);
1948 1949

	free_percpu(br->mcast_stats);
1950
}
1951 1952 1953

int br_multicast_set_router(struct net_bridge *br, unsigned long val)
{
1954
	int err = -EINVAL;
1955 1956 1957 1958

	spin_lock_bh(&br->multicast_lock);

	switch (val) {
1959 1960
	case MDB_RTR_TYPE_DISABLED:
	case MDB_RTR_TYPE_PERM:
1961 1962
		del_timer(&br->multicast_router_timer);
		/* fall through */
1963
	case MDB_RTR_TYPE_TEMP_QUERY:
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973
		br->multicast_router = val;
		err = 0;
		break;
	}

	spin_unlock_bh(&br->multicast_lock);

	return err;
}

1974 1975 1976 1977 1978 1979 1980 1981
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);
}

1982 1983 1984
int br_multicast_set_port_router(struct net_bridge_port *p, unsigned long val)
{
	struct net_bridge *br = p->br;
1985
	unsigned long now = jiffies;
1986
	int err = -EINVAL;
1987 1988

	spin_lock(&br->multicast_lock);
1989
	if (p->multicast_router == val) {
1990 1991 1992 1993
		/* 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);
1994 1995 1996
		err = 0;
		goto unlock;
	}
1997
	switch (val) {
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
	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;
2009 2010 2011
		del_timer(&p->multicast_router_timer);
		br_multicast_add_router(br, p);
		break;
2012 2013 2014 2015
	case MDB_RTR_TYPE_TEMP:
		p->multicast_router = MDB_RTR_TYPE_TEMP;
		br_multicast_mark_router(br, p);
		break;
2016 2017
	default:
		goto unlock;
2018
	}
2019 2020
	err = 0;
unlock:
2021 2022 2023 2024
	spin_unlock(&br->multicast_lock);

	return err;
}
2025

2026
static void br_multicast_start_querier(struct net_bridge *br,
2027
				       struct bridge_mcast_own_query *query)
2028 2029
{
	struct net_bridge_port *port;
2030

2031
	__br_multicast_open(br, query);
2032 2033 2034 2035 2036 2037

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

2038 2039
		if (query == &br->ip4_own_query)
			br_multicast_enable(&port->ip4_own_query);
2040 2041
#if IS_ENABLED(CONFIG_IPV6)
		else
2042
			br_multicast_enable(&port->ip6_own_query);
2043
#endif
2044 2045 2046 2047 2048
	}
}

int br_multicast_toggle(struct net_bridge *br, unsigned long val)
{
2049
	struct net_bridge_mdb_htable *mdb;
2050 2051
	struct net_bridge_port *port;
	int err = 0;
2052

2053
	spin_lock_bh(&br->multicast_lock);
2054 2055 2056 2057 2058 2059 2060
	if (br->multicast_disabled == !val)
		goto unlock;

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

2061 2062 2063
	if (!netif_running(br->dev))
		goto unlock;

2064 2065 2066
	mdb = mlock_dereference(br->mdb, br);
	if (mdb) {
		if (mdb->old) {
2067 2068 2069 2070 2071 2072
			err = -EEXIST;
rollback:
			br->multicast_disabled = !!val;
			goto unlock;
		}

2073
		err = br_mdb_rehash(&br->mdb, mdb->max,
2074 2075 2076 2077 2078
				    br->hash_elasticity);
		if (err)
			goto rollback;
	}

2079 2080 2081
	br_multicast_open(br);
	list_for_each_entry(port, &br->port_list, list)
		__br_multicast_enable_port(port);
2082 2083

unlock:
2084
	spin_unlock_bh(&br->multicast_lock);
2085 2086 2087

	return err;
}
2088

2089 2090
int br_multicast_set_querier(struct net_bridge *br, unsigned long val)
{
2091 2092
	unsigned long max_delay;

2093 2094 2095 2096 2097 2098 2099
	val = !!val;

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

	br->multicast_querier = val;
2100 2101 2102 2103 2104
	if (!val)
		goto unlock;

	max_delay = br->multicast_query_response_interval;

2105 2106
	if (!timer_pending(&br->ip4_other_query.timer))
		br->ip4_other_query.delay_time = jiffies + max_delay;
2107

2108
	br_multicast_start_querier(br, &br->ip4_own_query);
2109 2110

#if IS_ENABLED(CONFIG_IPV6)
2111 2112
	if (!timer_pending(&br->ip6_other_query.timer))
		br->ip6_other_query.delay_time = jiffies + max_delay;
2113

2114
	br_multicast_start_querier(br, &br->ip6_own_query);
2115
#endif
2116 2117 2118 2119 2120 2121 2122

unlock:
	spin_unlock_bh(&br->multicast_lock);

	return 0;
}

2123 2124
int br_multicast_set_hash_max(struct net_bridge *br, unsigned long val)
{
2125
	int err = -EINVAL;
2126
	u32 old;
2127
	struct net_bridge_mdb_htable *mdb;
2128

2129
	spin_lock_bh(&br->multicast_lock);
2130 2131
	if (!is_power_of_2(val))
		goto unlock;
2132 2133 2134

	mdb = mlock_dereference(br->mdb, br);
	if (mdb && val < mdb->size)
2135 2136 2137 2138 2139 2140 2141
		goto unlock;

	err = 0;

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

2142 2143
	if (mdb) {
		if (mdb->old) {
2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156
			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:
2157
	spin_unlock_bh(&br->multicast_lock);
2158 2159 2160

	return err;
}
2161

2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179
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;
}

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 2234 2235
/**
 * 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);
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
/**
 * 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);

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 2303 2304
/**
 * 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;
2305
#if IS_ENABLED(CONFIG_IPV6)
2306 2307 2308 2309 2310
	case ETH_P_IPV6:
		if (!timer_pending(&br->ip6_other_query.timer) ||
		    rcu_dereference(br->ip6_querier.port) == port)
			goto unlock;
		break;
2311
#endif
2312 2313 2314 2315 2316 2317 2318 2319 2320 2321
	default:
		goto unlock;
	}

	ret = true;
unlock:
	rcu_read_unlock();
	return ret;
}
EXPORT_SYMBOL_GPL(br_multicast_has_querier_adjacent);
2322 2323

static void br_mcast_stats_add(struct bridge_mcast_stats __percpu *stats,
2324
			       const struct sk_buff *skb, u8 type, u8 dir)
2325 2326
{
	struct bridge_mcast_stats *pstats = this_cpu_ptr(stats);
2327 2328
	__be16 proto = skb->protocol;
	unsigned int t_len;
2329 2330 2331 2332

	u64_stats_update_begin(&pstats->syncp);
	switch (proto) {
	case htons(ETH_P_IP):
2333
		t_len = ntohs(ip_hdr(skb)->tot_len) - ip_hdrlen(skb);
2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344
		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:
2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359
			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]++;
			}
2360 2361 2362 2363 2364 2365 2366 2367
			break;
		case IGMP_HOST_LEAVE_MESSAGE:
			pstats->mstats.igmp_leaves[dir]++;
			break;
		}
		break;
#if IS_ENABLED(CONFIG_IPV6)
	case htons(ETH_P_IPV6):
2368 2369 2370
		t_len = ntohs(ipv6_hdr(skb)->payload_len) +
			sizeof(struct ipv6hdr);
		t_len -= skb_network_header_len(skb);
2371 2372 2373 2374 2375 2376 2377 2378
		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:
2379 2380 2381 2382
			if (t_len != sizeof(struct mld_msg))
				pstats->mstats.mld_v2queries[dir]++;
			else
				pstats->mstats.mld_v1queries[dir]++;
2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394
			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,
2395
			const struct sk_buff *skb, u8 type, u8 dir)
2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409
{
	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;

2410
	br_mcast_stats_add(stats, skb, type, dir);
2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 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
}

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));

2455 2456 2457
		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);
2458 2459 2460 2461 2462 2463
		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;

2464 2465
		mcast_stats_add_dir(tdst.mld_v1queries, temp.mld_v1queries);
		mcast_stats_add_dir(tdst.mld_v2queries, temp.mld_v2queries);
2466 2467 2468 2469 2470 2471 2472
		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));
}