ipmr.c 43.2 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3
/*
 *	IP multicast routing support for mrouted 3.6/3.8
 *
4
 *		(c) 1995 Alan Cox, <alan@lxorguk.ukuu.org.uk>
L
Linus Torvalds 已提交
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
 *	  Linux Consultancy and Custom Driver Development
 *
 *	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.
 *
 *	Fixes:
 *	Michael Chastain	:	Incorrect size of copying.
 *	Alan Cox		:	Added the cache manager code
 *	Alan Cox		:	Fixed the clone/copy bug and device race.
 *	Mike McLagan		:	Routing by source
 *	Malcolm Beattie		:	Buffer handling fixes.
 *	Alexey Kuznetsov	:	Double buffer free and other fixes.
 *	SVR Anand		:	Fixed several multicast bugs and problems.
 *	Alexey Kuznetsov	:	Status, optimisations and more.
 *	Brad Parker		:	Better behaviour on mrouted upcall
 *					overflow.
 *      Carlos Picoto           :       PIMv1 Support
 *	Pavlin Ivanov Radoslavov:	PIMv2 Registers must checksum only PIM header
 *					Relax this requrement to work with older peers.
 *
 */

#include <asm/system.h>
#include <asm/uaccess.h>
#include <linux/types.h>
32
#include <linux/capability.h>
L
Linus Torvalds 已提交
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
#include <linux/errno.h>
#include <linux/timer.h>
#include <linux/mm.h>
#include <linux/kernel.h>
#include <linux/fcntl.h>
#include <linux/stat.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/inetdevice.h>
#include <linux/igmp.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/mroute.h>
#include <linux/init.h>
49
#include <linux/if_ether.h>
50
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
51 52 53
#include <net/ip.h>
#include <net/protocol.h>
#include <linux/skbuff.h>
54
#include <net/route.h>
L
Linus Torvalds 已提交
55 56 57 58 59 60 61 62 63
#include <net/sock.h>
#include <net/icmp.h>
#include <net/udp.h>
#include <net/raw.h>
#include <linux/notifier.h>
#include <linux/if_arp.h>
#include <linux/netfilter_ipv4.h>
#include <net/ipip.h>
#include <net/checksum.h>
64
#include <net/netlink.h>
L
Linus Torvalds 已提交
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106

#if defined(CONFIG_IP_PIMSM_V1) || defined(CONFIG_IP_PIMSM_V2)
#define CONFIG_IP_PIMSM	1
#endif

static struct sock *mroute_socket;


/* Big lock, protecting vif table, mrt cache and mroute socket state.
   Note that the changes are semaphored via rtnl_lock.
 */

static DEFINE_RWLOCK(mrt_lock);

/*
 *	Multicast router control variables
 */

static struct vif_device vif_table[MAXVIFS];		/* Devices 		*/
static int maxvif;

#define VIF_EXISTS(idx) (vif_table[idx].dev != NULL)

static int mroute_do_assert;				/* Set in PIM assert	*/
static int mroute_do_pim;

static struct mfc_cache *mfc_cache_array[MFC_LINES];	/* Forwarding cache	*/

static struct mfc_cache *mfc_unres_queue;		/* Queue of unresolved entries */
static atomic_t cache_resolve_queue_len;		/* Size of unresolved	*/

/* Special spinlock for queue of unresolved entries */
static DEFINE_SPINLOCK(mfc_unres_lock);

/* We return to original Alan's scheme. Hash table of resolved
   entries is changed only in process context and protected
   with weak lock mrt_lock. Queue of unresolved entries is protected
   with strong spinlock mfc_unres_lock.

   In this case data path is free of exclusive locks at all.
 */

107
static struct kmem_cache *mrt_cachep __read_mostly;
L
Linus Torvalds 已提交
108 109 110 111 112 113 114 115 116 117 118 119 120

static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local);
static int ipmr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert);
static int ipmr_fill_mroute(struct sk_buff *skb, struct mfc_cache *c, struct rtmsg *rtm);

#ifdef CONFIG_IP_PIMSM_V2
static struct net_protocol pim_protocol;
#endif

static struct timer_list ipmr_expire_timer;

/* Service routines creating virtual interfaces: DVMRP tunnels and PIMREG */

121 122 123 124 125 126
static void ipmr_del_tunnel(struct net_device *dev, struct vifctl *v)
{
	dev_close(dev);

	dev = __dev_get_by_name(&init_net, "tunl0");
	if (dev) {
127
		const struct net_device_ops *ops = dev->netdev_ops;
128 129 130 131 132 133 134 135 136 137 138 139
		struct ifreq ifr;
		struct ip_tunnel_parm p;

		memset(&p, 0, sizeof(p));
		p.iph.daddr = v->vifc_rmt_addr.s_addr;
		p.iph.saddr = v->vifc_lcl_addr.s_addr;
		p.iph.version = 4;
		p.iph.ihl = 5;
		p.iph.protocol = IPPROTO_IPIP;
		sprintf(p.name, "dvmrp%d", v->vifc_vifi);
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;

140 141 142 143 144 145 146
		if (ops->ndo_do_ioctl) {
			mm_segment_t oldfs = get_fs();

			set_fs(KERNEL_DS);
			ops->ndo_do_ioctl(dev, &ifr, SIOCDELTUNNEL);
			set_fs(oldfs);
		}
147 148 149
	}
}

L
Linus Torvalds 已提交
150 151 152 153 154
static
struct net_device *ipmr_new_tunnel(struct vifctl *v)
{
	struct net_device  *dev;

155
	dev = __dev_get_by_name(&init_net, "tunl0");
L
Linus Torvalds 已提交
156 157

	if (dev) {
158
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
159 160 161 162 163 164 165 166 167 168 169 170
		int err;
		struct ifreq ifr;
		struct ip_tunnel_parm p;
		struct in_device  *in_dev;

		memset(&p, 0, sizeof(p));
		p.iph.daddr = v->vifc_rmt_addr.s_addr;
		p.iph.saddr = v->vifc_lcl_addr.s_addr;
		p.iph.version = 4;
		p.iph.ihl = 5;
		p.iph.protocol = IPPROTO_IPIP;
		sprintf(p.name, "dvmrp%d", v->vifc_vifi);
S
Stephen Hemminger 已提交
171
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;
L
Linus Torvalds 已提交
172

173 174 175 176 177 178 179 180
		if (ops->ndo_do_ioctl) {
			mm_segment_t oldfs = get_fs();

			set_fs(KERNEL_DS);
			err = ops->ndo_do_ioctl(dev, &ifr, SIOCADDTUNNEL);
			set_fs(oldfs);
		} else
			err = -EOPNOTSUPP;
L
Linus Torvalds 已提交
181 182 183

		dev = NULL;

184
		if (err == 0 && (dev = __dev_get_by_name(&init_net, p.name)) != NULL) {
L
Linus Torvalds 已提交
185 186
			dev->flags |= IFF_MULTICAST;

187
			in_dev = __in_dev_get_rtnl(dev);
188
			if (in_dev == NULL)
L
Linus Torvalds 已提交
189
				goto failure;
190 191 192

			ipv4_devconf_setall(in_dev);
			IPV4_DEVCONF(in_dev->cnf, RP_FILTER) = 0;
L
Linus Torvalds 已提交
193 194 195

			if (dev_open(dev))
				goto failure;
196
			dev_hold(dev);
L
Linus Torvalds 已提交
197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216
		}
	}
	return dev;

failure:
	/* allow the register to be completed before unregistering. */
	rtnl_unlock();
	rtnl_lock();

	unregister_netdevice(dev);
	return NULL;
}

#ifdef CONFIG_IP_PIMSM

static int reg_vif_num = -1;

static int reg_vif_xmit(struct sk_buff *skb, struct net_device *dev)
{
	read_lock(&mrt_lock);
217 218
	dev->stats.tx_bytes += skb->len;
	dev->stats.tx_packets++;
L
Linus Torvalds 已提交
219 220 221 222 223 224
	ipmr_cache_report(skb, reg_vif_num, IGMPMSG_WHOLEPKT);
	read_unlock(&mrt_lock);
	kfree_skb(skb);
	return 0;
}

225 226 227 228
static const struct net_device_ops reg_vif_netdev_ops = {
	.ndo_start_xmit	= reg_vif_xmit,
};

L
Linus Torvalds 已提交
229 230 231
static void reg_vif_setup(struct net_device *dev)
{
	dev->type		= ARPHRD_PIMREG;
232
	dev->mtu		= ETH_DATA_LEN - sizeof(struct iphdr) - 8;
L
Linus Torvalds 已提交
233
	dev->flags		= IFF_NOARP;
234
	dev->netdev_ops		= &reg_vif_netdev_ops,
L
Linus Torvalds 已提交
235 236 237 238 239 240 241 242
	dev->destructor		= free_netdev;
}

static struct net_device *ipmr_reg_vif(void)
{
	struct net_device *dev;
	struct in_device *in_dev;

243
	dev = alloc_netdev(0, "pimreg", reg_vif_setup);
L
Linus Torvalds 已提交
244 245 246 247 248 249 250 251 252 253

	if (dev == NULL)
		return NULL;

	if (register_netdevice(dev)) {
		free_netdev(dev);
		return NULL;
	}
	dev->iflink = 0;

254 255 256
	rcu_read_lock();
	if ((in_dev = __in_dev_get_rcu(dev)) == NULL) {
		rcu_read_unlock();
L
Linus Torvalds 已提交
257
		goto failure;
258
	}
L
Linus Torvalds 已提交
259

260 261 262
	ipv4_devconf_setall(in_dev);
	IPV4_DEVCONF(in_dev->cnf, RP_FILTER) = 0;
	rcu_read_unlock();
L
Linus Torvalds 已提交
263 264 265 266

	if (dev_open(dev))
		goto failure;

267 268
	dev_hold(dev);

L
Linus Torvalds 已提交
269 270 271 272 273 274 275 276 277 278 279 280 281 282
	return dev;

failure:
	/* allow the register to be completed before unregistering. */
	rtnl_unlock();
	rtnl_lock();

	unregister_netdevice(dev);
	return NULL;
}
#endif

/*
 *	Delete a VIF entry
283
 *	@notify: Set to 1, if the caller is a notifier_call
L
Linus Torvalds 已提交
284
 */
285

286
static int vif_delete(int vifi, int notify)
L
Linus Torvalds 已提交
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
{
	struct vif_device *v;
	struct net_device *dev;
	struct in_device *in_dev;

	if (vifi < 0 || vifi >= maxvif)
		return -EADDRNOTAVAIL;

	v = &vif_table[vifi];

	write_lock_bh(&mrt_lock);
	dev = v->dev;
	v->dev = NULL;

	if (!dev) {
		write_unlock_bh(&mrt_lock);
		return -EADDRNOTAVAIL;
	}

#ifdef CONFIG_IP_PIMSM
	if (vifi == reg_vif_num)
		reg_vif_num = -1;
#endif

	if (vifi+1 == maxvif) {
		int tmp;
		for (tmp=vifi-1; tmp>=0; tmp--) {
			if (VIF_EXISTS(tmp))
				break;
		}
		maxvif = tmp+1;
	}

	write_unlock_bh(&mrt_lock);

	dev_set_allmulti(dev, -1);

324
	if ((in_dev = __in_dev_get_rtnl(dev)) != NULL) {
325
		IPV4_DEVCONF(in_dev->cnf, MC_FORWARDING)--;
L
Linus Torvalds 已提交
326 327 328
		ip_rt_multicast_event(in_dev);
	}

329
	if (v->flags&(VIFF_TUNNEL|VIFF_REGISTER) && !notify)
L
Linus Torvalds 已提交
330 331 332 333 334 335 336 337 338 339 340 341 342
		unregister_netdevice(dev);

	dev_put(dev);
	return 0;
}

/* Destroy an unresolved cache entry, killing queued skbs
   and reporting error to netlink readers.
 */

static void ipmr_destroy_unres(struct mfc_cache *c)
{
	struct sk_buff *skb;
343
	struct nlmsgerr *e;
L
Linus Torvalds 已提交
344 345 346

	atomic_dec(&cache_resolve_queue_len);

J
Jianjun Kong 已提交
347
	while ((skb = skb_dequeue(&c->mfc_un.unres.unresolved))) {
348
		if (ip_hdr(skb)->version == 0) {
L
Linus Torvalds 已提交
349 350 351 352
			struct nlmsghdr *nlh = (struct nlmsghdr *)skb_pull(skb, sizeof(struct iphdr));
			nlh->nlmsg_type = NLMSG_ERROR;
			nlh->nlmsg_len = NLMSG_LENGTH(sizeof(struct nlmsgerr));
			skb_trim(skb, nlh->nlmsg_len);
353 354 355
			e = NLMSG_DATA(nlh);
			e->error = -ETIMEDOUT;
			memset(&e->msg, 0, sizeof(e->msg));
356

357
			rtnl_unicast(skb, &init_net, NETLINK_CB(skb).pid);
L
Linus Torvalds 已提交
358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408
		} else
			kfree_skb(skb);
	}

	kmem_cache_free(mrt_cachep, c);
}


/* Single timer process for all the unresolved queue. */

static void ipmr_expire_process(unsigned long dummy)
{
	unsigned long now;
	unsigned long expires;
	struct mfc_cache *c, **cp;

	if (!spin_trylock(&mfc_unres_lock)) {
		mod_timer(&ipmr_expire_timer, jiffies+HZ/10);
		return;
	}

	if (atomic_read(&cache_resolve_queue_len) == 0)
		goto out;

	now = jiffies;
	expires = 10*HZ;
	cp = &mfc_unres_queue;

	while ((c=*cp) != NULL) {
		if (time_after(c->mfc_un.unres.expires, now)) {
			unsigned long interval = c->mfc_un.unres.expires - now;
			if (interval < expires)
				expires = interval;
			cp = &c->next;
			continue;
		}

		*cp = c->next;

		ipmr_destroy_unres(c);
	}

	if (atomic_read(&cache_resolve_queue_len))
		mod_timer(&ipmr_expire_timer, jiffies + expires);

out:
	spin_unlock(&mfc_unres_lock);
}

/* Fill oifs list. It is called under write locked mrt_lock. */

409
static void ipmr_update_thresholds(struct mfc_cache *cache, unsigned char *ttls)
L
Linus Torvalds 已提交
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433
{
	int vifi;

	cache->mfc_un.res.minvif = MAXVIFS;
	cache->mfc_un.res.maxvif = 0;
	memset(cache->mfc_un.res.ttls, 255, MAXVIFS);

	for (vifi=0; vifi<maxvif; vifi++) {
		if (VIF_EXISTS(vifi) && ttls[vifi] && ttls[vifi] < 255) {
			cache->mfc_un.res.ttls[vifi] = ttls[vifi];
			if (cache->mfc_un.res.minvif > vifi)
				cache->mfc_un.res.minvif = vifi;
			if (cache->mfc_un.res.maxvif <= vifi)
				cache->mfc_un.res.maxvif = vifi + 1;
		}
	}
}

static int vif_add(struct vifctl *vifc, int mrtsock)
{
	int vifi = vifc->vifc_vifi;
	struct vif_device *v = &vif_table[vifi];
	struct net_device *dev;
	struct in_device *in_dev;
434
	int err;
L
Linus Torvalds 已提交
435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451

	/* Is vif busy ? */
	if (VIF_EXISTS(vifi))
		return -EADDRINUSE;

	switch (vifc->vifc_flags) {
#ifdef CONFIG_IP_PIMSM
	case VIFF_REGISTER:
		/*
		 * Special Purpose VIF in PIM
		 * All the packets will be sent to the daemon
		 */
		if (reg_vif_num >= 0)
			return -EADDRINUSE;
		dev = ipmr_reg_vif();
		if (!dev)
			return -ENOBUFS;
452 453 454
		err = dev_set_allmulti(dev, 1);
		if (err) {
			unregister_netdevice(dev);
455
			dev_put(dev);
456 457
			return err;
		}
L
Linus Torvalds 已提交
458 459
		break;
#endif
460
	case VIFF_TUNNEL:
L
Linus Torvalds 已提交
461 462 463
		dev = ipmr_new_tunnel(vifc);
		if (!dev)
			return -ENOBUFS;
464 465 466
		err = dev_set_allmulti(dev, 1);
		if (err) {
			ipmr_del_tunnel(dev, vifc);
467
			dev_put(dev);
468 469
			return err;
		}
L
Linus Torvalds 已提交
470 471
		break;
	case 0:
472
		dev = ip_dev_find(&init_net, vifc->vifc_lcl_addr.s_addr);
L
Linus Torvalds 已提交
473 474
		if (!dev)
			return -EADDRNOTAVAIL;
475
		err = dev_set_allmulti(dev, 1);
476 477
		if (err) {
			dev_put(dev);
478
			return err;
479
		}
L
Linus Torvalds 已提交
480 481 482 483 484
		break;
	default:
		return -EINVAL;
	}

485
	if ((in_dev = __in_dev_get_rtnl(dev)) == NULL)
L
Linus Torvalds 已提交
486
		return -EADDRNOTAVAIL;
487
	IPV4_DEVCONF(in_dev->cnf, MC_FORWARDING)++;
L
Linus Torvalds 已提交
488 489 490 491 492
	ip_rt_multicast_event(in_dev);

	/*
	 *	Fill in the VIF structures
	 */
J
Jianjun Kong 已提交
493 494 495 496
	v->rate_limit = vifc->vifc_rate_limit;
	v->local = vifc->vifc_lcl_addr.s_addr;
	v->remote = vifc->vifc_rmt_addr.s_addr;
	v->flags = vifc->vifc_flags;
L
Linus Torvalds 已提交
497 498
	if (!mrtsock)
		v->flags |= VIFF_STATIC;
J
Jianjun Kong 已提交
499
	v->threshold = vifc->vifc_threshold;
L
Linus Torvalds 已提交
500 501 502 503 504 505 506 507 508 509
	v->bytes_in = 0;
	v->bytes_out = 0;
	v->pkt_in = 0;
	v->pkt_out = 0;
	v->link = dev->ifindex;
	if (v->flags&(VIFF_TUNNEL|VIFF_REGISTER))
		v->link = dev->iflink;

	/* And finish update writing critical data */
	write_lock_bh(&mrt_lock);
J
Jianjun Kong 已提交
510
	v->dev = dev;
L
Linus Torvalds 已提交
511 512 513 514 515 516 517 518 519 520
#ifdef CONFIG_IP_PIMSM
	if (v->flags&VIFF_REGISTER)
		reg_vif_num = vifi;
#endif
	if (vifi+1 > maxvif)
		maxvif = vifi+1;
	write_unlock_bh(&mrt_lock);
	return 0;
}

A
Al Viro 已提交
521
static struct mfc_cache *ipmr_cache_find(__be32 origin, __be32 mcastgrp)
L
Linus Torvalds 已提交
522
{
J
Jianjun Kong 已提交
523
	int line = MFC_HASH(mcastgrp, origin);
L
Linus Torvalds 已提交
524 525 526 527 528 529 530 531 532 533 534 535 536 537
	struct mfc_cache *c;

	for (c=mfc_cache_array[line]; c; c = c->next) {
		if (c->mfc_origin==origin && c->mfc_mcastgrp==mcastgrp)
			break;
	}
	return c;
}

/*
 *	Allocate a multicast cache entry
 */
static struct mfc_cache *ipmr_cache_alloc(void)
{
J
Jianjun Kong 已提交
538 539
	struct mfc_cache *c = kmem_cache_zalloc(mrt_cachep, GFP_KERNEL);
	if (c == NULL)
L
Linus Torvalds 已提交
540 541 542 543 544 545 546
		return NULL;
	c->mfc_un.res.minvif = MAXVIFS;
	return c;
}

static struct mfc_cache *ipmr_cache_alloc_unres(void)
{
J
Jianjun Kong 已提交
547 548
	struct mfc_cache *c = kmem_cache_zalloc(mrt_cachep, GFP_ATOMIC);
	if (c == NULL)
L
Linus Torvalds 已提交
549 550 551 552 553 554 555 556 557
		return NULL;
	skb_queue_head_init(&c->mfc_un.unres.unresolved);
	c->mfc_un.unres.expires = jiffies + 10*HZ;
	return c;
}

/*
 *	A cache entry has gone into a resolved state from queued
 */
558

L
Linus Torvalds 已提交
559 560 561
static void ipmr_cache_resolve(struct mfc_cache *uc, struct mfc_cache *c)
{
	struct sk_buff *skb;
562
	struct nlmsgerr *e;
L
Linus Torvalds 已提交
563 564 565 566 567

	/*
	 *	Play the pending entries through our router
	 */

J
Jianjun Kong 已提交
568
	while ((skb = __skb_dequeue(&uc->mfc_un.unres.unresolved))) {
569
		if (ip_hdr(skb)->version == 0) {
L
Linus Torvalds 已提交
570 571 572
			struct nlmsghdr *nlh = (struct nlmsghdr *)skb_pull(skb, sizeof(struct iphdr));

			if (ipmr_fill_mroute(skb, c, NLMSG_DATA(nlh)) > 0) {
573 574
				nlh->nlmsg_len = (skb_tail_pointer(skb) -
						  (u8 *)nlh);
L
Linus Torvalds 已提交
575 576 577 578
			} else {
				nlh->nlmsg_type = NLMSG_ERROR;
				nlh->nlmsg_len = NLMSG_LENGTH(sizeof(struct nlmsgerr));
				skb_trim(skb, nlh->nlmsg_len);
579 580 581
				e = NLMSG_DATA(nlh);
				e->error = -EMSGSIZE;
				memset(&e->msg, 0, sizeof(e->msg));
L
Linus Torvalds 已提交
582
			}
583

584
			rtnl_unicast(skb, &init_net, NETLINK_CB(skb).pid);
L
Linus Torvalds 已提交
585 586 587 588 589 590 591 592 593 594 595
		} else
			ip_mr_forward(skb, c, 0);
	}
}

/*
 *	Bounce a cache query up to mrouted. We could use netlink for this but mrouted
 *	expects the following bizarre scheme.
 *
 *	Called under mrt_lock.
 */
596

L
Linus Torvalds 已提交
597 598 599
static int ipmr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert)
{
	struct sk_buff *skb;
600
	const int ihl = ip_hdrlen(pkt);
L
Linus Torvalds 已提交
601 602 603 604 605 606 607 608 609 610 611
	struct igmphdr *igmp;
	struct igmpmsg *msg;
	int ret;

#ifdef CONFIG_IP_PIMSM
	if (assert == IGMPMSG_WHOLEPKT)
		skb = skb_realloc_headroom(pkt, sizeof(struct iphdr));
	else
#endif
		skb = alloc_skb(128, GFP_ATOMIC);

S
Stephen Hemminger 已提交
612
	if (!skb)
L
Linus Torvalds 已提交
613 614 615 616 617 618 619 620 621
		return -ENOBUFS;

#ifdef CONFIG_IP_PIMSM
	if (assert == IGMPMSG_WHOLEPKT) {
		/* Ugly, but we have no choice with this interface.
		   Duplicate old header, fix ihl, length etc.
		   And all this only to mangle msg->im_msgtype and
		   to set msg->im_mbz to "mbz" :-)
		 */
622 623
		skb_push(skb, sizeof(struct iphdr));
		skb_reset_network_header(skb);
624
		skb_reset_transport_header(skb);
625
		msg = (struct igmpmsg *)skb_network_header(skb);
626
		memcpy(msg, skb_network_header(pkt), sizeof(struct iphdr));
L
Linus Torvalds 已提交
627 628
		msg->im_msgtype = IGMPMSG_WHOLEPKT;
		msg->im_mbz = 0;
629
		msg->im_vif = reg_vif_num;
630 631 632
		ip_hdr(skb)->ihl = sizeof(struct iphdr) >> 2;
		ip_hdr(skb)->tot_len = htons(ntohs(ip_hdr(pkt)->tot_len) +
					     sizeof(struct iphdr));
633
	} else
L
Linus Torvalds 已提交
634
#endif
635 636
	{

L
Linus Torvalds 已提交
637 638 639 640
	/*
	 *	Copy the IP header
	 */

641
	skb->network_header = skb->tail;
642
	skb_put(skb, ihl);
643
	skb_copy_to_linear_data(skb, pkt->data, ihl);
644 645
	ip_hdr(skb)->protocol = 0;			/* Flag to the kernel this is a route add */
	msg = (struct igmpmsg *)skb_network_header(skb);
L
Linus Torvalds 已提交
646 647 648 649 650 651 652
	msg->im_vif = vifi;
	skb->dst = dst_clone(pkt->dst);

	/*
	 *	Add our header
	 */

J
Jianjun Kong 已提交
653
	igmp=(struct igmphdr *)skb_put(skb, sizeof(struct igmphdr));
L
Linus Torvalds 已提交
654 655 656
	igmp->type	=
	msg->im_msgtype = assert;
	igmp->code 	=	0;
657
	ip_hdr(skb)->tot_len = htons(skb->len);			/* Fix the length */
658
	skb->transport_header = skb->network_header;
659
	}
L
Linus Torvalds 已提交
660 661 662 663 664 665 666 667 668

	if (mroute_socket == NULL) {
		kfree_skb(skb);
		return -EINVAL;
	}

	/*
	 *	Deliver to mrouted
	 */
J
Jianjun Kong 已提交
669
	if ((ret = sock_queue_rcv_skb(mroute_socket, skb))<0) {
L
Linus Torvalds 已提交
670 671 672 673 674 675 676 677 678 679 680
		if (net_ratelimit())
			printk(KERN_WARNING "mroute: pending queue full, dropping entries.\n");
		kfree_skb(skb);
	}

	return ret;
}

/*
 *	Queue a packet for resolution. It gets locked cache entry!
 */
681

L
Linus Torvalds 已提交
682 683 684 685 686
static int
ipmr_cache_unresolved(vifi_t vifi, struct sk_buff *skb)
{
	int err;
	struct mfc_cache *c;
687
	const struct iphdr *iph = ip_hdr(skb);
L
Linus Torvalds 已提交
688 689 690

	spin_lock_bh(&mfc_unres_lock);
	for (c=mfc_unres_queue; c; c=c->next) {
691 692
		if (c->mfc_mcastgrp == iph->daddr &&
		    c->mfc_origin == iph->saddr)
L
Linus Torvalds 已提交
693 694 695 696 697 698 699 700
			break;
	}

	if (c == NULL) {
		/*
		 *	Create a new entry if allowable
		 */

J
Jianjun Kong 已提交
701
		if (atomic_read(&cache_resolve_queue_len) >= 10 ||
L
Linus Torvalds 已提交
702 703 704 705 706 707 708 709 710 711
		    (c=ipmr_cache_alloc_unres())==NULL) {
			spin_unlock_bh(&mfc_unres_lock);

			kfree_skb(skb);
			return -ENOBUFS;
		}

		/*
		 *	Fill in the new cache entry
		 */
712 713 714
		c->mfc_parent	= -1;
		c->mfc_origin	= iph->saddr;
		c->mfc_mcastgrp	= iph->daddr;
L
Linus Torvalds 已提交
715 716 717 718 719

		/*
		 *	Reflect first query at mrouted.
		 */
		if ((err = ipmr_cache_report(skb, vifi, IGMPMSG_NOCACHE))<0) {
720
			/* If the report failed throw the cache entry
L
Linus Torvalds 已提交
721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743
			   out - Brad Parker
			 */
			spin_unlock_bh(&mfc_unres_lock);

			kmem_cache_free(mrt_cachep, c);
			kfree_skb(skb);
			return err;
		}

		atomic_inc(&cache_resolve_queue_len);
		c->next = mfc_unres_queue;
		mfc_unres_queue = c;

		mod_timer(&ipmr_expire_timer, c->mfc_un.unres.expires);
	}

	/*
	 *	See if we can append the packet
	 */
	if (c->mfc_un.unres.unresolved.qlen>3) {
		kfree_skb(skb);
		err = -ENOBUFS;
	} else {
J
Jianjun Kong 已提交
744
		skb_queue_tail(&c->mfc_un.unres.unresolved, skb);
L
Linus Torvalds 已提交
745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760
		err = 0;
	}

	spin_unlock_bh(&mfc_unres_lock);
	return err;
}

/*
 *	MFC cache manipulation by user space mroute daemon
 */

static int ipmr_mfc_delete(struct mfcctl *mfc)
{
	int line;
	struct mfc_cache *c, **cp;

J
Jianjun Kong 已提交
761
	line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
L
Linus Torvalds 已提交
762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781

	for (cp=&mfc_cache_array[line]; (c=*cp) != NULL; cp = &c->next) {
		if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
		    c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr) {
			write_lock_bh(&mrt_lock);
			*cp = c->next;
			write_unlock_bh(&mrt_lock);

			kmem_cache_free(mrt_cachep, c);
			return 0;
		}
	}
	return -ENOENT;
}

static int ipmr_mfc_add(struct mfcctl *mfc, int mrtsock)
{
	int line;
	struct mfc_cache *uc, *c, **cp;

J
Jianjun Kong 已提交
782
	line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
L
Linus Torvalds 已提交
783 784 785 786 787 788 789 790 791 792

	for (cp=&mfc_cache_array[line]; (c=*cp) != NULL; cp = &c->next) {
		if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
		    c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr)
			break;
	}

	if (c != NULL) {
		write_lock_bh(&mrt_lock);
		c->mfc_parent = mfc->mfcc_parent;
793
		ipmr_update_thresholds(c, mfc->mfcc_ttls);
L
Linus Torvalds 已提交
794 795 796 797 798 799
		if (!mrtsock)
			c->mfc_flags |= MFC_STATIC;
		write_unlock_bh(&mrt_lock);
		return 0;
	}

800
	if (!ipv4_is_multicast(mfc->mfcc_mcastgrp.s_addr))
L
Linus Torvalds 已提交
801 802
		return -EINVAL;

J
Jianjun Kong 已提交
803 804
	c = ipmr_cache_alloc();
	if (c == NULL)
L
Linus Torvalds 已提交
805 806
		return -ENOMEM;

J
Jianjun Kong 已提交
807 808 809
	c->mfc_origin = mfc->mfcc_origin.s_addr;
	c->mfc_mcastgrp = mfc->mfcc_mcastgrp.s_addr;
	c->mfc_parent = mfc->mfcc_parent;
810
	ipmr_update_thresholds(c, mfc->mfcc_ttls);
L
Linus Torvalds 已提交
811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845
	if (!mrtsock)
		c->mfc_flags |= MFC_STATIC;

	write_lock_bh(&mrt_lock);
	c->next = mfc_cache_array[line];
	mfc_cache_array[line] = c;
	write_unlock_bh(&mrt_lock);

	/*
	 *	Check to see if we resolved a queued list. If so we
	 *	need to send on the frames and tidy up.
	 */
	spin_lock_bh(&mfc_unres_lock);
	for (cp = &mfc_unres_queue; (uc=*cp) != NULL;
	     cp = &uc->next) {
		if (uc->mfc_origin == c->mfc_origin &&
		    uc->mfc_mcastgrp == c->mfc_mcastgrp) {
			*cp = uc->next;
			if (atomic_dec_and_test(&cache_resolve_queue_len))
				del_timer(&ipmr_expire_timer);
			break;
		}
	}
	spin_unlock_bh(&mfc_unres_lock);

	if (uc) {
		ipmr_cache_resolve(uc, c);
		kmem_cache_free(mrt_cachep, uc);
	}
	return 0;
}

/*
 *	Close the multicast socket, and clear the vif tables etc
 */
846

L
Linus Torvalds 已提交
847 848 849
static void mroute_clean_tables(struct sock *sk)
{
	int i;
850

L
Linus Torvalds 已提交
851 852 853
	/*
	 *	Shut down all active vif entries
	 */
S
Stephen Hemminger 已提交
854
	for (i=0; i<maxvif; i++) {
L
Linus Torvalds 已提交
855
		if (!(vif_table[i].flags&VIFF_STATIC))
856
			vif_delete(i, 0);
L
Linus Torvalds 已提交
857 858 859 860 861
	}

	/*
	 *	Wipe the cache
	 */
J
Jianjun Kong 已提交
862
	for (i=0; i<MFC_LINES; i++) {
L
Linus Torvalds 已提交
863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899
		struct mfc_cache *c, **cp;

		cp = &mfc_cache_array[i];
		while ((c = *cp) != NULL) {
			if (c->mfc_flags&MFC_STATIC) {
				cp = &c->next;
				continue;
			}
			write_lock_bh(&mrt_lock);
			*cp = c->next;
			write_unlock_bh(&mrt_lock);

			kmem_cache_free(mrt_cachep, c);
		}
	}

	if (atomic_read(&cache_resolve_queue_len) != 0) {
		struct mfc_cache *c;

		spin_lock_bh(&mfc_unres_lock);
		while (mfc_unres_queue != NULL) {
			c = mfc_unres_queue;
			mfc_unres_queue = c->next;
			spin_unlock_bh(&mfc_unres_lock);

			ipmr_destroy_unres(c);

			spin_lock_bh(&mfc_unres_lock);
		}
		spin_unlock_bh(&mfc_unres_lock);
	}
}

static void mrtsock_destruct(struct sock *sk)
{
	rtnl_lock();
	if (sk == mroute_socket) {
900
		IPV4_DEVCONF_ALL(sock_net(sk), MC_FORWARDING)--;
L
Linus Torvalds 已提交
901 902

		write_lock_bh(&mrt_lock);
J
Jianjun Kong 已提交
903
		mroute_socket = NULL;
L
Linus Torvalds 已提交
904 905 906 907 908 909 910 911 912 913 914 915 916
		write_unlock_bh(&mrt_lock);

		mroute_clean_tables(sk);
	}
	rtnl_unlock();
}

/*
 *	Socket options and virtual interface manipulation. The whole
 *	virtual interface system is a complete heap, but unfortunately
 *	that's how BSD mrouted happens to think. Maybe one day with a proper
 *	MOSPF/PIM router set up we can clean this up.
 */
917

J
Jianjun Kong 已提交
918
int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, int optlen)
L
Linus Torvalds 已提交
919 920 921 922
{
	int ret;
	struct vifctl vif;
	struct mfcctl mfc;
923

S
Stephen Hemminger 已提交
924 925
	if (optname != MRT_INIT) {
		if (sk != mroute_socket && !capable(CAP_NET_ADMIN))
L
Linus Torvalds 已提交
926 927 928
			return -EACCES;
	}

S
Stephen Hemminger 已提交
929 930 931 932 933
	switch (optname) {
	case MRT_INIT:
		if (sk->sk_type != SOCK_RAW ||
		    inet_sk(sk)->num != IPPROTO_IGMP)
			return -EOPNOTSUPP;
J
Jianjun Kong 已提交
934
		if (optlen != sizeof(int))
S
Stephen Hemminger 已提交
935
			return -ENOPROTOOPT;
L
Linus Torvalds 已提交
936

S
Stephen Hemminger 已提交
937 938
		rtnl_lock();
		if (mroute_socket) {
L
Linus Torvalds 已提交
939
			rtnl_unlock();
S
Stephen Hemminger 已提交
940 941 942 943 944 945
			return -EADDRINUSE;
		}

		ret = ip_ra_control(sk, 1, mrtsock_destruct);
		if (ret == 0) {
			write_lock_bh(&mrt_lock);
J
Jianjun Kong 已提交
946
			mroute_socket = sk;
S
Stephen Hemminger 已提交
947 948
			write_unlock_bh(&mrt_lock);

949
			IPV4_DEVCONF_ALL(sock_net(sk), MC_FORWARDING)++;
S
Stephen Hemminger 已提交
950 951 952 953
		}
		rtnl_unlock();
		return ret;
	case MRT_DONE:
J
Jianjun Kong 已提交
954
		if (sk != mroute_socket)
S
Stephen Hemminger 已提交
955 956 957 958
			return -EACCES;
		return ip_ra_control(sk, 0, NULL);
	case MRT_ADD_VIF:
	case MRT_DEL_VIF:
J
Jianjun Kong 已提交
959
		if (optlen != sizeof(vif))
S
Stephen Hemminger 已提交
960
			return -EINVAL;
J
Jianjun Kong 已提交
961
		if (copy_from_user(&vif, optval, sizeof(vif)))
S
Stephen Hemminger 已提交
962 963 964 965
			return -EFAULT;
		if (vif.vifc_vifi >= MAXVIFS)
			return -ENFILE;
		rtnl_lock();
J
Jianjun Kong 已提交
966
		if (optname == MRT_ADD_VIF) {
S
Stephen Hemminger 已提交
967 968
			ret = vif_add(&vif, sk==mroute_socket);
		} else {
969
			ret = vif_delete(vif.vifc_vifi, 0);
S
Stephen Hemminger 已提交
970 971 972
		}
		rtnl_unlock();
		return ret;
L
Linus Torvalds 已提交
973 974 975 976 977

		/*
		 *	Manipulate the forwarding caches. These live
		 *	in a sort of kernel/user symbiosis.
		 */
S
Stephen Hemminger 已提交
978 979
	case MRT_ADD_MFC:
	case MRT_DEL_MFC:
J
Jianjun Kong 已提交
980
		if (optlen != sizeof(mfc))
S
Stephen Hemminger 已提交
981
			return -EINVAL;
J
Jianjun Kong 已提交
982
		if (copy_from_user(&mfc, optval, sizeof(mfc)))
S
Stephen Hemminger 已提交
983 984
			return -EFAULT;
		rtnl_lock();
J
Jianjun Kong 已提交
985
		if (optname == MRT_DEL_MFC)
S
Stephen Hemminger 已提交
986 987 988 989 990
			ret = ipmr_mfc_delete(&mfc);
		else
			ret = ipmr_mfc_add(&mfc, sk==mroute_socket);
		rtnl_unlock();
		return ret;
L
Linus Torvalds 已提交
991 992 993
		/*
		 *	Control PIM assert.
		 */
S
Stephen Hemminger 已提交
994 995 996 997 998 999 1000 1001
	case MRT_ASSERT:
	{
		int v;
		if (get_user(v,(int __user *)optval))
			return -EFAULT;
		mroute_do_assert=(v)?1:0;
		return 0;
	}
L
Linus Torvalds 已提交
1002
#ifdef CONFIG_IP_PIMSM
S
Stephen Hemminger 已提交
1003 1004
	case MRT_PIM:
	{
S
Stephen Hemminger 已提交
1005 1006
		int v;

S
Stephen Hemminger 已提交
1007 1008
		if (get_user(v,(int __user *)optval))
			return -EFAULT;
S
Stephen Hemminger 已提交
1009 1010
		v = (v) ? 1 : 0;

S
Stephen Hemminger 已提交
1011 1012 1013 1014 1015
		rtnl_lock();
		ret = 0;
		if (v != mroute_do_pim) {
			mroute_do_pim = v;
			mroute_do_assert = v;
L
Linus Torvalds 已提交
1016
#ifdef CONFIG_IP_PIMSM_V2
S
Stephen Hemminger 已提交
1017 1018 1019 1020 1021 1022 1023 1024
			if (mroute_do_pim)
				ret = inet_add_protocol(&pim_protocol,
							IPPROTO_PIM);
			else
				ret = inet_del_protocol(&pim_protocol,
							IPPROTO_PIM);
			if (ret < 0)
				ret = -EAGAIN;
L
Linus Torvalds 已提交
1025 1026
#endif
		}
S
Stephen Hemminger 已提交
1027 1028 1029
		rtnl_unlock();
		return ret;
	}
L
Linus Torvalds 已提交
1030
#endif
S
Stephen Hemminger 已提交
1031 1032 1033 1034 1035 1036
	/*
	 *	Spurious command, or MRT_VERSION which you cannot
	 *	set.
	 */
	default:
		return -ENOPROTOOPT;
L
Linus Torvalds 已提交
1037 1038 1039 1040 1041 1042
	}
}

/*
 *	Getsock opt support for the multicast routing system.
 */
1043

J
Jianjun Kong 已提交
1044
int ip_mroute_getsockopt(struct sock *sk, int optname, char __user *optval, int __user *optlen)
L
Linus Torvalds 已提交
1045 1046 1047 1048
{
	int olr;
	int val;

J
Jianjun Kong 已提交
1049
	if (optname != MRT_VERSION &&
L
Linus Torvalds 已提交
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
#ifdef CONFIG_IP_PIMSM
	   optname!=MRT_PIM &&
#endif
	   optname!=MRT_ASSERT)
		return -ENOPROTOOPT;

	if (get_user(olr, optlen))
		return -EFAULT;

	olr = min_t(unsigned int, olr, sizeof(int));
	if (olr < 0)
		return -EINVAL;
1062

J
Jianjun Kong 已提交
1063
	if (put_user(olr, optlen))
L
Linus Torvalds 已提交
1064
		return -EFAULT;
J
Jianjun Kong 已提交
1065 1066
	if (optname == MRT_VERSION)
		val = 0x0305;
L
Linus Torvalds 已提交
1067
#ifdef CONFIG_IP_PIMSM
J
Jianjun Kong 已提交
1068 1069
	else if (optname == MRT_PIM)
		val = mroute_do_pim;
L
Linus Torvalds 已提交
1070 1071
#endif
	else
J
Jianjun Kong 已提交
1072 1073
		val = mroute_do_assert;
	if (copy_to_user(optval, &val, olr))
L
Linus Torvalds 已提交
1074 1075 1076 1077 1078 1079 1080
		return -EFAULT;
	return 0;
}

/*
 *	The IP multicast ioctl support routines.
 */
1081

L
Linus Torvalds 已提交
1082 1083 1084 1085 1086 1087
int ipmr_ioctl(struct sock *sk, int cmd, void __user *arg)
{
	struct sioc_sg_req sr;
	struct sioc_vif_req vr;
	struct vif_device *vif;
	struct mfc_cache *c;
1088

S
Stephen Hemminger 已提交
1089 1090
	switch (cmd) {
	case SIOCGETVIFCNT:
J
Jianjun Kong 已提交
1091
		if (copy_from_user(&vr, arg, sizeof(vr)))
S
Stephen Hemminger 已提交
1092
			return -EFAULT;
J
Jianjun Kong 已提交
1093
		if (vr.vifi >= maxvif)
S
Stephen Hemminger 已提交
1094 1095 1096 1097
			return -EINVAL;
		read_lock(&mrt_lock);
		vif=&vif_table[vr.vifi];
		if (VIF_EXISTS(vr.vifi))	{
J
Jianjun Kong 已提交
1098 1099 1100 1101
			vr.icount = vif->pkt_in;
			vr.ocount = vif->pkt_out;
			vr.ibytes = vif->bytes_in;
			vr.obytes = vif->bytes_out;
L
Linus Torvalds 已提交
1102 1103
			read_unlock(&mrt_lock);

J
Jianjun Kong 已提交
1104
			if (copy_to_user(arg, &vr, sizeof(vr)))
S
Stephen Hemminger 已提交
1105 1106 1107 1108 1109 1110
				return -EFAULT;
			return 0;
		}
		read_unlock(&mrt_lock);
		return -EADDRNOTAVAIL;
	case SIOCGETSGCNT:
J
Jianjun Kong 已提交
1111
		if (copy_from_user(&sr, arg, sizeof(sr)))
S
Stephen Hemminger 已提交
1112 1113 1114 1115 1116 1117 1118 1119
			return -EFAULT;

		read_lock(&mrt_lock);
		c = ipmr_cache_find(sr.src.s_addr, sr.grp.s_addr);
		if (c) {
			sr.pktcnt = c->mfc_un.res.pkt;
			sr.bytecnt = c->mfc_un.res.bytes;
			sr.wrong_if = c->mfc_un.res.wrong_if;
L
Linus Torvalds 已提交
1120
			read_unlock(&mrt_lock);
S
Stephen Hemminger 已提交
1121

J
Jianjun Kong 已提交
1122
			if (copy_to_user(arg, &sr, sizeof(sr)))
S
Stephen Hemminger 已提交
1123 1124 1125 1126 1127 1128 1129
				return -EFAULT;
			return 0;
		}
		read_unlock(&mrt_lock);
		return -EADDRNOTAVAIL;
	default:
		return -ENOIOCTLCMD;
L
Linus Torvalds 已提交
1130 1131 1132 1133 1134 1135
	}
}


static int ipmr_device_event(struct notifier_block *this, unsigned long event, void *ptr)
{
1136
	struct net_device *dev = ptr;
L
Linus Torvalds 已提交
1137 1138
	struct vif_device *v;
	int ct;
1139

1140
	if (!net_eq(dev_net(dev), &init_net))
1141 1142
		return NOTIFY_DONE;

L
Linus Torvalds 已提交
1143 1144 1145
	if (event != NETDEV_UNREGISTER)
		return NOTIFY_DONE;
	v=&vif_table[0];
J
Jianjun Kong 已提交
1146 1147
	for (ct=0; ct<maxvif; ct++,v++) {
		if (v->dev == dev)
1148
			vif_delete(ct, 1);
L
Linus Torvalds 已提交
1149 1150 1151 1152 1153
	}
	return NOTIFY_DONE;
}


J
Jianjun Kong 已提交
1154
static struct notifier_block ip_mr_notifier = {
L
Linus Torvalds 已提交
1155 1156 1157 1158 1159 1160 1161 1162
	.notifier_call = ipmr_device_event,
};

/*
 * 	Encapsulate a packet by attaching a valid IPIP header to it.
 *	This avoids tunnel drivers and other mess and gives us the speed so
 *	important for multicast video.
 */
1163

A
Al Viro 已提交
1164
static void ip_encap(struct sk_buff *skb, __be32 saddr, __be32 daddr)
L
Linus Torvalds 已提交
1165
{
1166
	struct iphdr *iph;
1167
	struct iphdr *old_iph = ip_hdr(skb);
1168 1169

	skb_push(skb, sizeof(struct iphdr));
1170
	skb->transport_header = skb->network_header;
1171
	skb_reset_network_header(skb);
1172
	iph = ip_hdr(skb);
L
Linus Torvalds 已提交
1173 1174

	iph->version	= 	4;
1175 1176
	iph->tos	=	old_iph->tos;
	iph->ttl	=	old_iph->ttl;
L
Linus Torvalds 已提交
1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193
	iph->frag_off	=	0;
	iph->daddr	=	daddr;
	iph->saddr	=	saddr;
	iph->protocol	=	IPPROTO_IPIP;
	iph->ihl	=	5;
	iph->tot_len	=	htons(skb->len);
	ip_select_ident(iph, skb->dst, NULL);
	ip_send_check(iph);

	memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
	nf_reset(skb);
}

static inline int ipmr_forward_finish(struct sk_buff *skb)
{
	struct ip_options * opt	= &(IPCB(skb)->opt);

1194
	IP_INC_STATS_BH(dev_net(skb->dst->dev), IPSTATS_MIB_OUTFORWDATAGRAMS);
L
Linus Torvalds 已提交
1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207

	if (unlikely(opt->optlen))
		ip_forward_options(skb);

	return dst_output(skb);
}

/*
 *	Processing handlers for ipmr_forward
 */

static void ipmr_queue_xmit(struct sk_buff *skb, struct mfc_cache *c, int vifi)
{
1208
	const struct iphdr *iph = ip_hdr(skb);
L
Linus Torvalds 已提交
1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219
	struct vif_device *vif = &vif_table[vifi];
	struct net_device *dev;
	struct rtable *rt;
	int    encap = 0;

	if (vif->dev == NULL)
		goto out_free;

#ifdef CONFIG_IP_PIMSM
	if (vif->flags & VIFF_REGISTER) {
		vif->pkt_out++;
J
Jianjun Kong 已提交
1220
		vif->bytes_out += skb->len;
1221 1222
		vif->dev->stats.tx_bytes += skb->len;
		vif->dev->stats.tx_packets++;
L
Linus Torvalds 已提交
1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235
		ipmr_cache_report(skb, vifi, IGMPMSG_WHOLEPKT);
		kfree_skb(skb);
		return;
	}
#endif

	if (vif->flags&VIFF_TUNNEL) {
		struct flowi fl = { .oif = vif->link,
				    .nl_u = { .ip4_u =
					      { .daddr = vif->remote,
						.saddr = vif->local,
						.tos = RT_TOS(iph->tos) } },
				    .proto = IPPROTO_IPIP };
1236
		if (ip_route_output_key(&init_net, &rt, &fl))
L
Linus Torvalds 已提交
1237 1238 1239 1240 1241 1242 1243 1244
			goto out_free;
		encap = sizeof(struct iphdr);
	} else {
		struct flowi fl = { .oif = vif->link,
				    .nl_u = { .ip4_u =
					      { .daddr = iph->daddr,
						.tos = RT_TOS(iph->tos) } },
				    .proto = IPPROTO_IPIP };
1245
		if (ip_route_output_key(&init_net, &rt, &fl))
L
Linus Torvalds 已提交
1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256
			goto out_free;
	}

	dev = rt->u.dst.dev;

	if (skb->len+encap > dst_mtu(&rt->u.dst) && (ntohs(iph->frag_off) & IP_DF)) {
		/* Do not fragment multicasts. Alas, IPv4 does not
		   allow to send ICMP, so that packets will disappear
		   to blackhole.
		 */

1257
		IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
1258 1259 1260 1261 1262 1263 1264
		ip_rt_put(rt);
		goto out_free;
	}

	encap += LL_RESERVED_SPACE(dev) + rt->u.dst.header_len;

	if (skb_cow(skb, encap)) {
1265
		ip_rt_put(rt);
L
Linus Torvalds 已提交
1266 1267 1268 1269
		goto out_free;
	}

	vif->pkt_out++;
J
Jianjun Kong 已提交
1270
	vif->bytes_out += skb->len;
L
Linus Torvalds 已提交
1271 1272 1273

	dst_release(skb->dst);
	skb->dst = &rt->u.dst;
1274
	ip_decrease_ttl(ip_hdr(skb));
L
Linus Torvalds 已提交
1275 1276 1277 1278 1279 1280

	/* FIXME: forward and output firewalls used to be called here.
	 * What do we do with netfilter? -- RR */
	if (vif->flags & VIFF_TUNNEL) {
		ip_encap(skb, vif->local, vif->remote);
		/* FIXME: extra output firewall step used to be here. --RR */
1281 1282
		vif->dev->stats.tx_packets++;
		vif->dev->stats.tx_bytes += skb->len;
L
Linus Torvalds 已提交
1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297
	}

	IPCB(skb)->flags |= IPSKB_FORWARDED;

	/*
	 * RFC1584 teaches, that DVMRP/PIM router must deliver packets locally
	 * not only before forwarding, but after forwarding on all output
	 * interfaces. It is clear, if mrouter runs a multicasting
	 * program, it should receive packets not depending to what interface
	 * program is joined.
	 * If we will not make it, the program will have to join on all
	 * interfaces. On the other hand, multihoming host (or router, but
	 * not mrouter) cannot join to more than one interface - it will
	 * result in receiving multiple packets.
	 */
1298
	NF_HOOK(PF_INET, NF_INET_FORWARD, skb, skb->dev, dev,
L
Linus Torvalds 已提交
1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333
		ipmr_forward_finish);
	return;

out_free:
	kfree_skb(skb);
	return;
}

static int ipmr_find_vif(struct net_device *dev)
{
	int ct;
	for (ct=maxvif-1; ct>=0; ct--) {
		if (vif_table[ct].dev == dev)
			break;
	}
	return ct;
}

/* "local" means that we should preserve one skb (for local delivery) */

static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local)
{
	int psend = -1;
	int vif, ct;

	vif = cache->mfc_parent;
	cache->mfc_un.res.pkt++;
	cache->mfc_un.res.bytes += skb->len;

	/*
	 * Wrong interface: drop packet and (maybe) send PIM assert.
	 */
	if (vif_table[vif].dev != skb->dev) {
		int true_vifi;

1334
		if (skb->rtable->fl.iif == 0) {
L
Linus Torvalds 已提交
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358
			/* It is our own packet, looped back.
			   Very complicated situation...

			   The best workaround until routing daemons will be
			   fixed is not to redistribute packet, if it was
			   send through wrong interface. It means, that
			   multicast applications WILL NOT work for
			   (S,G), which have default multicast route pointing
			   to wrong oif. In any case, it is not a good
			   idea to use multicasting applications on router.
			 */
			goto dont_forward;
		}

		cache->mfc_un.res.wrong_if++;
		true_vifi = ipmr_find_vif(skb->dev);

		if (true_vifi >= 0 && mroute_do_assert &&
		    /* pimsm uses asserts, when switching from RPT to SPT,
		       so that we cannot check that packet arrived on an oif.
		       It is bad, but otherwise we would need to move pretty
		       large chunk of pimd to kernel. Ough... --ANK
		     */
		    (mroute_do_pim || cache->mfc_un.res.ttls[true_vifi] < 255) &&
1359
		    time_after(jiffies,
L
Linus Torvalds 已提交
1360 1361 1362 1363 1364 1365 1366 1367
			       cache->mfc_un.res.last_assert + MFC_ASSERT_THRESH)) {
			cache->mfc_un.res.last_assert = jiffies;
			ipmr_cache_report(skb, true_vifi, IGMPMSG_WRONGVIF);
		}
		goto dont_forward;
	}

	vif_table[vif].pkt_in++;
J
Jianjun Kong 已提交
1368
	vif_table[vif].bytes_in += skb->len;
L
Linus Torvalds 已提交
1369 1370 1371 1372 1373

	/*
	 *	Forward the frame
	 */
	for (ct = cache->mfc_un.res.maxvif-1; ct >= cache->mfc_un.res.minvif; ct--) {
1374
		if (ip_hdr(skb)->ttl > cache->mfc_un.res.ttls[ct]) {
L
Linus Torvalds 已提交
1375 1376 1377 1378 1379
			if (psend != -1) {
				struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
				if (skb2)
					ipmr_queue_xmit(skb2, cache, psend);
			}
J
Jianjun Kong 已提交
1380
			psend = ct;
L
Linus Torvalds 已提交
1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407
		}
	}
	if (psend != -1) {
		if (local) {
			struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
			if (skb2)
				ipmr_queue_xmit(skb2, cache, psend);
		} else {
			ipmr_queue_xmit(skb, cache, psend);
			return 0;
		}
	}

dont_forward:
	if (!local)
		kfree_skb(skb);
	return 0;
}


/*
 *	Multicast packets for forwarding arrive here
 */

int ip_mr_input(struct sk_buff *skb)
{
	struct mfc_cache *cache;
1408
	int local = skb->rtable->rt_flags&RTCF_LOCAL;
L
Linus Torvalds 已提交
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419

	/* Packet is looped back after forward, it should not be
	   forwarded second time, but still can be delivered locally.
	 */
	if (IPCB(skb)->flags&IPSKB_FORWARDED)
		goto dont_forward;

	if (!local) {
		    if (IPCB(skb)->opt.router_alert) {
			    if (ip_call_ra_chain(skb))
				    return 0;
1420
		    } else if (ip_hdr(skb)->protocol == IPPROTO_IGMP){
L
Linus Torvalds 已提交
1421 1422 1423 1424 1425 1426 1427 1428
			    /* IGMPv1 (and broken IGMPv2 implementations sort of
			       Cisco IOS <= 11.2(8)) do not put router alert
			       option to IGMP packets destined to routable
			       groups. It is very bad, because it means
			       that we can forward NO IGMP messages.
			     */
			    read_lock(&mrt_lock);
			    if (mroute_socket) {
1429
				    nf_reset(skb);
L
Linus Torvalds 已提交
1430 1431 1432 1433 1434 1435 1436 1437 1438
				    raw_rcv(mroute_socket, skb);
				    read_unlock(&mrt_lock);
				    return 0;
			    }
			    read_unlock(&mrt_lock);
		    }
	}

	read_lock(&mrt_lock);
1439
	cache = ipmr_cache_find(ip_hdr(skb)->saddr, ip_hdr(skb)->daddr);
L
Linus Torvalds 已提交
1440 1441 1442 1443

	/*
	 *	No usable cache entry
	 */
J
Jianjun Kong 已提交
1444
	if (cache == NULL) {
L
Linus Torvalds 已提交
1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484
		int vif;

		if (local) {
			struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
			ip_local_deliver(skb);
			if (skb2 == NULL) {
				read_unlock(&mrt_lock);
				return -ENOBUFS;
			}
			skb = skb2;
		}

		vif = ipmr_find_vif(skb->dev);
		if (vif >= 0) {
			int err = ipmr_cache_unresolved(vif, skb);
			read_unlock(&mrt_lock);

			return err;
		}
		read_unlock(&mrt_lock);
		kfree_skb(skb);
		return -ENODEV;
	}

	ip_mr_forward(skb, cache, local);

	read_unlock(&mrt_lock);

	if (local)
		return ip_local_deliver(skb);

	return 0;

dont_forward:
	if (local)
		return ip_local_deliver(skb);
	kfree_skb(skb);
	return 0;
}

I
Ilpo Järvinen 已提交
1485 1486
#ifdef CONFIG_IP_PIMSM
static int __pim_rcv(struct sk_buff *skb, unsigned int pimlen)
L
Linus Torvalds 已提交
1487
{
I
Ilpo Järvinen 已提交
1488 1489
	struct net_device *reg_dev = NULL;
	struct iphdr *encap;
L
Linus Torvalds 已提交
1490

I
Ilpo Järvinen 已提交
1491
	encap = (struct iphdr *)(skb_transport_header(skb) + pimlen);
L
Linus Torvalds 已提交
1492 1493 1494 1495 1496 1497
	/*
	   Check that:
	   a. packet is really destinted to a multicast group
	   b. packet is not a NULL-REGISTER
	   c. packet is not truncated
	 */
1498
	if (!ipv4_is_multicast(encap->daddr) ||
L
Linus Torvalds 已提交
1499
	    encap->tot_len == 0 ||
I
Ilpo Järvinen 已提交
1500 1501
	    ntohs(encap->tot_len) + pimlen > skb->len)
		return 1;
L
Linus Torvalds 已提交
1502 1503 1504 1505 1506 1507 1508 1509

	read_lock(&mrt_lock);
	if (reg_vif_num >= 0)
		reg_dev = vif_table[reg_vif_num].dev;
	if (reg_dev)
		dev_hold(reg_dev);
	read_unlock(&mrt_lock);

1510
	if (reg_dev == NULL)
I
Ilpo Järvinen 已提交
1511
		return 1;
L
Linus Torvalds 已提交
1512

1513
	skb->mac_header = skb->network_header;
L
Linus Torvalds 已提交
1514
	skb_pull(skb, (u8*)encap - skb->data);
1515
	skb_reset_network_header(skb);
L
Linus Torvalds 已提交
1516 1517 1518 1519 1520 1521
	skb->dev = reg_dev;
	skb->protocol = htons(ETH_P_IP);
	skb->ip_summed = 0;
	skb->pkt_type = PACKET_HOST;
	dst_release(skb->dst);
	skb->dst = NULL;
1522 1523
	reg_dev->stats.rx_bytes += skb->len;
	reg_dev->stats.rx_packets++;
L
Linus Torvalds 已提交
1524 1525 1526
	nf_reset(skb);
	netif_rx(skb);
	dev_put(reg_dev);
I
Ilpo Järvinen 已提交
1527

L
Linus Torvalds 已提交
1528
	return 0;
I
Ilpo Järvinen 已提交
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553
}
#endif

#ifdef CONFIG_IP_PIMSM_V1
/*
 * Handle IGMP messages of PIMv1
 */

int pim_rcv_v1(struct sk_buff * skb)
{
	struct igmphdr *pim;

	if (!pskb_may_pull(skb, sizeof(*pim) + sizeof(struct iphdr)))
		goto drop;

	pim = igmp_hdr(skb);

	if (!mroute_do_pim ||
	    pim->group != PIM_V1_VERSION || pim->code != PIM_V1_REGISTER)
		goto drop;

	if (__pim_rcv(skb, sizeof(*pim))) {
drop:
		kfree_skb(skb);
	}
L
Linus Torvalds 已提交
1554 1555 1556 1557 1558 1559 1560 1561 1562
	return 0;
}
#endif

#ifdef CONFIG_IP_PIMSM_V2
static int pim_rcv(struct sk_buff * skb)
{
	struct pimreghdr *pim;

I
Ilpo Järvinen 已提交
1563
	if (!pskb_may_pull(skb, sizeof(*pim) + sizeof(struct iphdr)))
L
Linus Torvalds 已提交
1564 1565
		goto drop;

1566
	pim = (struct pimreghdr *)skb_transport_header(skb);
1567
	if (pim->type != ((PIM_VERSION<<4)|(PIM_REGISTER)) ||
L
Linus Torvalds 已提交
1568
	    (pim->flags&PIM_NULL_REGISTER) ||
1569
	    (ip_compute_csum((void *)pim, sizeof(*pim)) != 0 &&
1570
	     csum_fold(skb_checksum(skb, 0, skb->len, 0))))
L
Linus Torvalds 已提交
1571 1572
		goto drop;

I
Ilpo Järvinen 已提交
1573 1574 1575 1576
	if (__pim_rcv(skb, sizeof(*pim))) {
drop:
		kfree_skb(skb);
	}
L
Linus Torvalds 已提交
1577 1578 1579 1580 1581 1582 1583 1584 1585 1586
	return 0;
}
#endif

static int
ipmr_fill_mroute(struct sk_buff *skb, struct mfc_cache *c, struct rtmsg *rtm)
{
	int ct;
	struct rtnexthop *nhp;
	struct net_device *dev = vif_table[c->mfc_parent].dev;
1587
	u8 *b = skb_tail_pointer(skb);
L
Linus Torvalds 已提交
1588 1589 1590 1591 1592
	struct rtattr *mp_head;

	if (dev)
		RTA_PUT(skb, RTA_IIF, 4, &dev->ifindex);

J
Jianjun Kong 已提交
1593
	mp_head = (struct rtattr *)skb_put(skb, RTA_LENGTH(0));
L
Linus Torvalds 已提交
1594 1595 1596 1597 1598

	for (ct = c->mfc_un.res.minvif; ct < c->mfc_un.res.maxvif; ct++) {
		if (c->mfc_un.res.ttls[ct] < 255) {
			if (skb_tailroom(skb) < RTA_ALIGN(RTA_ALIGN(sizeof(*nhp)) + 4))
				goto rtattr_failure;
J
Jianjun Kong 已提交
1599
			nhp = (struct rtnexthop *)skb_put(skb, RTA_ALIGN(sizeof(*nhp)));
L
Linus Torvalds 已提交
1600 1601 1602 1603 1604 1605 1606
			nhp->rtnh_flags = 0;
			nhp->rtnh_hops = c->mfc_un.res.ttls[ct];
			nhp->rtnh_ifindex = vif_table[ct].dev->ifindex;
			nhp->rtnh_len = sizeof(*nhp);
		}
	}
	mp_head->rta_type = RTA_MULTIPATH;
1607
	mp_head->rta_len = skb_tail_pointer(skb) - (u8 *)mp_head;
L
Linus Torvalds 已提交
1608 1609 1610 1611
	rtm->rtm_type = RTN_MULTICAST;
	return 1;

rtattr_failure:
1612
	nlmsg_trim(skb, b);
L
Linus Torvalds 已提交
1613 1614 1615 1616 1617 1618 1619
	return -EMSGSIZE;
}

int ipmr_get_route(struct sk_buff *skb, struct rtmsg *rtm, int nowait)
{
	int err;
	struct mfc_cache *cache;
1620
	struct rtable *rt = skb->rtable;
L
Linus Torvalds 已提交
1621 1622 1623 1624

	read_lock(&mrt_lock);
	cache = ipmr_cache_find(rt->rt_src, rt->rt_dst);

J
Jianjun Kong 已提交
1625
	if (cache == NULL) {
1626
		struct sk_buff *skb2;
1627
		struct iphdr *iph;
L
Linus Torvalds 已提交
1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640
		struct net_device *dev;
		int vif;

		if (nowait) {
			read_unlock(&mrt_lock);
			return -EAGAIN;
		}

		dev = skb->dev;
		if (dev == NULL || (vif = ipmr_find_vif(dev)) < 0) {
			read_unlock(&mrt_lock);
			return -ENODEV;
		}
1641 1642 1643 1644 1645 1646
		skb2 = skb_clone(skb, GFP_ATOMIC);
		if (!skb2) {
			read_unlock(&mrt_lock);
			return -ENOMEM;
		}

1647 1648
		skb_push(skb2, sizeof(struct iphdr));
		skb_reset_network_header(skb2);
1649 1650 1651 1652 1653
		iph = ip_hdr(skb2);
		iph->ihl = sizeof(struct iphdr) >> 2;
		iph->saddr = rt->rt_src;
		iph->daddr = rt->rt_dst;
		iph->version = 0;
1654
		err = ipmr_cache_unresolved(vif, skb2);
L
Linus Torvalds 已提交
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665
		read_unlock(&mrt_lock);
		return err;
	}

	if (!nowait && (rtm->rtm_flags&RTM_F_NOTIFY))
		cache->mfc_flags |= MFC_NOTIFY;
	err = ipmr_fill_mroute(skb, cache, rtm);
	read_unlock(&mrt_lock);
	return err;
}

1666
#ifdef CONFIG_PROC_FS
L
Linus Torvalds 已提交
1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677
/*
 *	The /proc interfaces to multicast routing /proc/ip_mr_cache /proc/ip_mr_vif
 */
struct ipmr_vif_iter {
	int ct;
};

static struct vif_device *ipmr_vif_seq_idx(struct ipmr_vif_iter *iter,
					   loff_t pos)
{
	for (iter->ct = 0; iter->ct < maxvif; ++iter->ct) {
S
Stephen Hemminger 已提交
1678
		if (!VIF_EXISTS(iter->ct))
L
Linus Torvalds 已提交
1679
			continue;
1680
		if (pos-- == 0)
L
Linus Torvalds 已提交
1681 1682 1683 1684 1685 1686
			return &vif_table[iter->ct];
	}
	return NULL;
}

static void *ipmr_vif_seq_start(struct seq_file *seq, loff_t *pos)
S
Stephen Hemminger 已提交
1687
	__acquires(mrt_lock)
L
Linus Torvalds 已提交
1688 1689
{
	read_lock(&mrt_lock);
1690
	return *pos ? ipmr_vif_seq_idx(seq->private, *pos - 1)
L
Linus Torvalds 已提交
1691 1692 1693 1694 1695 1696 1697 1698 1699 1700
		: SEQ_START_TOKEN;
}

static void *ipmr_vif_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	struct ipmr_vif_iter *iter = seq->private;

	++*pos;
	if (v == SEQ_START_TOKEN)
		return ipmr_vif_seq_idx(iter, 0);
1701

L
Linus Torvalds 已提交
1702
	while (++iter->ct < maxvif) {
S
Stephen Hemminger 已提交
1703
		if (!VIF_EXISTS(iter->ct))
L
Linus Torvalds 已提交
1704 1705 1706 1707 1708 1709 1710
			continue;
		return &vif_table[iter->ct];
	}
	return NULL;
}

static void ipmr_vif_seq_stop(struct seq_file *seq, void *v)
S
Stephen Hemminger 已提交
1711
	__releases(mrt_lock)
L
Linus Torvalds 已提交
1712 1713 1714 1715 1716 1717 1718
{
	read_unlock(&mrt_lock);
}

static int ipmr_vif_seq_show(struct seq_file *seq, void *v)
{
	if (v == SEQ_START_TOKEN) {
1719
		seq_puts(seq,
L
Linus Torvalds 已提交
1720 1721 1722 1723 1724 1725 1726 1727
			 "Interface      BytesIn  PktsIn  BytesOut PktsOut Flags Local    Remote\n");
	} else {
		const struct vif_device *vif = v;
		const char *name =  vif->dev ? vif->dev->name : "none";

		seq_printf(seq,
			   "%2Zd %-10s %8ld %7ld  %8ld %7ld %05X %08X %08X\n",
			   vif - vif_table,
1728
			   name, vif->bytes_in, vif->pkt_in,
L
Linus Torvalds 已提交
1729 1730 1731 1732 1733 1734
			   vif->bytes_out, vif->pkt_out,
			   vif->flags, vif->local, vif->remote);
	}
	return 0;
}

1735
static const struct seq_operations ipmr_vif_seq_ops = {
L
Linus Torvalds 已提交
1736 1737 1738 1739 1740 1741 1742 1743
	.start = ipmr_vif_seq_start,
	.next  = ipmr_vif_seq_next,
	.stop  = ipmr_vif_seq_stop,
	.show  = ipmr_vif_seq_show,
};

static int ipmr_vif_open(struct inode *inode, struct file *file)
{
1744 1745
	return seq_open_private(file, &ipmr_vif_seq_ops,
			sizeof(struct ipmr_vif_iter));
L
Linus Torvalds 已提交
1746 1747
}

1748
static const struct file_operations ipmr_vif_fops = {
L
Linus Torvalds 已提交
1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767
	.owner	 = THIS_MODULE,
	.open    = ipmr_vif_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release_private,
};

struct ipmr_mfc_iter {
	struct mfc_cache **cache;
	int ct;
};


static struct mfc_cache *ipmr_mfc_seq_idx(struct ipmr_mfc_iter *it, loff_t pos)
{
	struct mfc_cache *mfc;

	it->cache = mfc_cache_array;
	read_lock(&mrt_lock);
1768
	for (it->ct = 0; it->ct < MFC_LINES; it->ct++)
S
Stephen Hemminger 已提交
1769
		for (mfc = mfc_cache_array[it->ct]; mfc; mfc = mfc->next)
1770
			if (pos-- == 0)
L
Linus Torvalds 已提交
1771 1772 1773 1774 1775
				return mfc;
	read_unlock(&mrt_lock);

	it->cache = &mfc_unres_queue;
	spin_lock_bh(&mfc_unres_lock);
S
Stephen Hemminger 已提交
1776
	for (mfc = mfc_unres_queue; mfc; mfc = mfc->next)
L
Linus Torvalds 已提交
1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790
		if (pos-- == 0)
			return mfc;
	spin_unlock_bh(&mfc_unres_lock);

	it->cache = NULL;
	return NULL;
}


static void *ipmr_mfc_seq_start(struct seq_file *seq, loff_t *pos)
{
	struct ipmr_mfc_iter *it = seq->private;
	it->cache = NULL;
	it->ct = 0;
1791
	return *pos ? ipmr_mfc_seq_idx(seq->private, *pos - 1)
L
Linus Torvalds 已提交
1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806
		: SEQ_START_TOKEN;
}

static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	struct mfc_cache *mfc = v;
	struct ipmr_mfc_iter *it = seq->private;

	++*pos;

	if (v == SEQ_START_TOKEN)
		return ipmr_mfc_seq_idx(seq->private, 0);

	if (mfc->next)
		return mfc->next;
1807 1808

	if (it->cache == &mfc_unres_queue)
L
Linus Torvalds 已提交
1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822
		goto end_of_list;

	BUG_ON(it->cache != mfc_cache_array);

	while (++it->ct < MFC_LINES) {
		mfc = mfc_cache_array[it->ct];
		if (mfc)
			return mfc;
	}

	/* exhausted cache_array, show unresolved */
	read_unlock(&mrt_lock);
	it->cache = &mfc_unres_queue;
	it->ct = 0;
1823

L
Linus Torvalds 已提交
1824 1825
	spin_lock_bh(&mfc_unres_lock);
	mfc = mfc_unres_queue;
1826
	if (mfc)
L
Linus Torvalds 已提交
1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850
		return mfc;

 end_of_list:
	spin_unlock_bh(&mfc_unres_lock);
	it->cache = NULL;

	return NULL;
}

static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
{
	struct ipmr_mfc_iter *it = seq->private;

	if (it->cache == &mfc_unres_queue)
		spin_unlock_bh(&mfc_unres_lock);
	else if (it->cache == mfc_cache_array)
		read_unlock(&mrt_lock);
}

static int ipmr_mfc_seq_show(struct seq_file *seq, void *v)
{
	int n;

	if (v == SEQ_START_TOKEN) {
1851
		seq_puts(seq,
L
Linus Torvalds 已提交
1852 1853 1854 1855
		 "Group    Origin   Iif     Pkts    Bytes    Wrong Oifs\n");
	} else {
		const struct mfc_cache *mfc = v;
		const struct ipmr_mfc_iter *it = seq->private;
1856

1857
		seq_printf(seq, "%08lX %08lX %-3hd",
L
Linus Torvalds 已提交
1858 1859
			   (unsigned long) mfc->mfc_mcastgrp,
			   (unsigned long) mfc->mfc_origin,
1860
			   mfc->mfc_parent);
L
Linus Torvalds 已提交
1861 1862

		if (it->cache != &mfc_unres_queue) {
1863 1864 1865 1866
			seq_printf(seq, " %8lu %8lu %8lu",
				   mfc->mfc_un.res.pkt,
				   mfc->mfc_un.res.bytes,
				   mfc->mfc_un.res.wrong_if);
S
Stephen Hemminger 已提交
1867 1868 1869
			for (n = mfc->mfc_un.res.minvif;
			     n < mfc->mfc_un.res.maxvif; n++ ) {
				if (VIF_EXISTS(n)
L
Linus Torvalds 已提交
1870
				   && mfc->mfc_un.res.ttls[n] < 255)
1871 1872
				seq_printf(seq,
					   " %2d:%-3d",
L
Linus Torvalds 已提交
1873 1874
					   n, mfc->mfc_un.res.ttls[n]);
			}
1875 1876 1877 1878 1879
		} else {
			/* unresolved mfc_caches don't contain
			 * pkt, bytes and wrong_if values
			 */
			seq_printf(seq, " %8lu %8lu %8lu", 0ul, 0ul, 0ul);
L
Linus Torvalds 已提交
1880 1881 1882 1883 1884 1885
		}
		seq_putc(seq, '\n');
	}
	return 0;
}

1886
static const struct seq_operations ipmr_mfc_seq_ops = {
L
Linus Torvalds 已提交
1887 1888 1889 1890 1891 1892 1893 1894
	.start = ipmr_mfc_seq_start,
	.next  = ipmr_mfc_seq_next,
	.stop  = ipmr_mfc_seq_stop,
	.show  = ipmr_mfc_seq_show,
};

static int ipmr_mfc_open(struct inode *inode, struct file *file)
{
1895 1896
	return seq_open_private(file, &ipmr_mfc_seq_ops,
			sizeof(struct ipmr_mfc_iter));
L
Linus Torvalds 已提交
1897 1898
}

1899
static const struct file_operations ipmr_mfc_fops = {
L
Linus Torvalds 已提交
1900 1901 1902 1903 1904 1905
	.owner	 = THIS_MODULE,
	.open    = ipmr_mfc_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release_private,
};
1906
#endif
L
Linus Torvalds 已提交
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917

#ifdef CONFIG_IP_PIMSM_V2
static struct net_protocol pim_protocol = {
	.handler	=	pim_rcv,
};
#endif


/*
 *	Setup for IP multicast routing
 */
1918

W
Wang Chen 已提交
1919
int __init ip_mr_init(void)
L
Linus Torvalds 已提交
1920
{
W
Wang Chen 已提交
1921 1922
	int err;

L
Linus Torvalds 已提交
1923 1924
	mrt_cachep = kmem_cache_create("ip_mrt_cache",
				       sizeof(struct mfc_cache),
A
Alexey Dobriyan 已提交
1925
				       0, SLAB_HWCACHE_ALIGN|SLAB_PANIC,
1926
				       NULL);
W
Wang Chen 已提交
1927 1928 1929
	if (!mrt_cachep)
		return -ENOMEM;

1930
	setup_timer(&ipmr_expire_timer, ipmr_expire_process, 0);
W
Wang Chen 已提交
1931 1932 1933
	err = register_netdevice_notifier(&ip_mr_notifier);
	if (err)
		goto reg_notif_fail;
1934
#ifdef CONFIG_PROC_FS
W
Wang Chen 已提交
1935 1936 1937 1938 1939
	err = -ENOMEM;
	if (!proc_net_fops_create(&init_net, "ip_mr_vif", 0, &ipmr_vif_fops))
		goto proc_vif_fail;
	if (!proc_net_fops_create(&init_net, "ip_mr_cache", 0, &ipmr_mfc_fops))
		goto proc_cache_fail;
1940
#endif
W
Wang Chen 已提交
1941 1942 1943 1944
	return 0;
#ifdef CONFIG_PROC_FS
proc_cache_fail:
	proc_net_remove(&init_net, "ip_mr_vif");
B
Benjamin Thery 已提交
1945 1946
proc_vif_fail:
	unregister_netdevice_notifier(&ip_mr_notifier);
W
Wang Chen 已提交
1947
#endif
B
Benjamin Thery 已提交
1948 1949 1950
reg_notif_fail:
	del_timer(&ipmr_expire_timer);
	kmem_cache_destroy(mrt_cachep);
W
Wang Chen 已提交
1951
	return err;
L
Linus Torvalds 已提交
1952
}