route.c 86.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5
/*
 *	Linux INET6 implementation
 *	FIB front-end.
 *
 *	Authors:
6
 *	Pedro Roque		<roque@di.fc.ul.pt>
L
Linus Torvalds 已提交
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
 *
 *	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.
 */

/*	Changes:
 *
 *	YOSHIFUJI Hideaki @USAGI
 *		reworked default router selection.
 *		- respect outgoing interface
 *		- select from (probably) reachable routers (i.e.
 *		routers in REACHABLE, STALE, DELAY or PROBE states).
 *		- always select the same router if it is (probably)
 *		reachable.  otherwise, round-robin the list.
23 24
 *	Ville Nuorvala
 *		Fixed routing subtrees.
L
Linus Torvalds 已提交
25 26
 */

27 28
#define pr_fmt(fmt) "IPv6: " fmt

29
#include <linux/capability.h>
L
Linus Torvalds 已提交
30
#include <linux/errno.h>
31
#include <linux/export.h>
L
Linus Torvalds 已提交
32 33 34 35 36 37 38 39
#include <linux/types.h>
#include <linux/times.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/route.h>
#include <linux/netdevice.h>
#include <linux/in6.h>
40
#include <linux/mroute6.h>
L
Linus Torvalds 已提交
41 42 43 44
#include <linux/init.h>
#include <linux/if_arp.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
45
#include <linux/nsproxy.h>
46
#include <linux/slab.h>
47
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
48 49 50 51 52 53 54 55 56
#include <net/snmp.h>
#include <net/ipv6.h>
#include <net/ip6_fib.h>
#include <net/ip6_route.h>
#include <net/ndisc.h>
#include <net/addrconf.h>
#include <net/tcp.h>
#include <linux/rtnetlink.h>
#include <net/dst.h>
57
#include <net/dst_metadata.h>
L
Linus Torvalds 已提交
58
#include <net/xfrm.h>
59
#include <net/netevent.h>
60
#include <net/netlink.h>
61
#include <net/nexthop.h>
62
#include <net/lwtunnel.h>
63
#include <net/ip_tunnels.h>
L
Linus Torvalds 已提交
64 65 66 67 68 69 70

#include <asm/uaccess.h>

#ifdef CONFIG_SYSCTL
#include <linux/sysctl.h>
#endif

71
enum rt6_nud_state {
J
Jiri Benc 已提交
72 73 74
	RT6_NUD_FAIL_HARD = -3,
	RT6_NUD_FAIL_PROBE = -2,
	RT6_NUD_FAIL_DO_RR = -1,
75 76 77
	RT6_NUD_SUCCEED = 1
};

M
Martin KaFai Lau 已提交
78
static void ip6_rt_copy_init(struct rt6_info *rt, struct rt6_info *ort);
L
Linus Torvalds 已提交
79
static struct dst_entry	*ip6_dst_check(struct dst_entry *dst, u32 cookie);
80
static unsigned int	 ip6_default_advmss(const struct dst_entry *dst);
81
static unsigned int	 ip6_mtu(const struct dst_entry *dst);
L
Linus Torvalds 已提交
82 83 84 85
static struct dst_entry *ip6_negative_advice(struct dst_entry *);
static void		ip6_dst_destroy(struct dst_entry *);
static void		ip6_dst_ifdown(struct dst_entry *,
				       struct net_device *dev, int how);
86
static int		 ip6_dst_gc(struct dst_ops *ops);
L
Linus Torvalds 已提交
87 88

static int		ip6_pkt_discard(struct sk_buff *skb);
89
static int		ip6_pkt_discard_out(struct sock *sk, struct sk_buff *skb);
90
static int		ip6_pkt_prohibit(struct sk_buff *skb);
91
static int		ip6_pkt_prohibit_out(struct sock *sk, struct sk_buff *skb);
L
Linus Torvalds 已提交
92
static void		ip6_link_failure(struct sk_buff *skb);
93 94 95 96
static void		ip6_rt_update_pmtu(struct dst_entry *dst, struct sock *sk,
					   struct sk_buff *skb, u32 mtu);
static void		rt6_do_redirect(struct dst_entry *dst, struct sock *sk,
					struct sk_buff *skb);
97
static void		rt6_dst_from_metrics_check(struct rt6_info *rt);
98
static int rt6_score_route(struct rt6_info *rt, int oif, int strict);
L
Linus Torvalds 已提交
99

100
#ifdef CONFIG_IPV6_ROUTE_INFO
101
static struct rt6_info *rt6_add_route_info(struct net *net,
102 103
					   const struct in6_addr *prefix, int prefixlen,
					   const struct in6_addr *gwaddr, int ifindex,
104
					   unsigned int pref);
105
static struct rt6_info *rt6_get_route_info(struct net *net,
106 107
					   const struct in6_addr *prefix, int prefixlen,
					   const struct in6_addr *gwaddr, int ifindex);
108 109
#endif

110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
struct uncached_list {
	spinlock_t		lock;
	struct list_head	head;
};

static DEFINE_PER_CPU_ALIGNED(struct uncached_list, rt6_uncached_list);

static void rt6_uncached_list_add(struct rt6_info *rt)
{
	struct uncached_list *ul = raw_cpu_ptr(&rt6_uncached_list);

	rt->dst.flags |= DST_NOCACHE;
	rt->rt6i_uncached_list = ul;

	spin_lock_bh(&ul->lock);
	list_add_tail(&rt->rt6i_uncached, &ul->head);
	spin_unlock_bh(&ul->lock);
}

static void rt6_uncached_list_del(struct rt6_info *rt)
{
	if (!list_empty(&rt->rt6i_uncached)) {
		struct uncached_list *ul = rt->rt6i_uncached_list;

		spin_lock_bh(&ul->lock);
		list_del(&rt->rt6i_uncached);
		spin_unlock_bh(&ul->lock);
	}
}

static void rt6_uncached_list_flush_dev(struct net *net, struct net_device *dev)
{
	struct net_device *loopback_dev = net->loopback_dev;
	int cpu;

	for_each_possible_cpu(cpu) {
		struct uncached_list *ul = per_cpu_ptr(&rt6_uncached_list, cpu);
		struct rt6_info *rt;

		spin_lock_bh(&ul->lock);
		list_for_each_entry(rt, &ul->head, rt6i_uncached) {
			struct inet6_dev *rt_idev = rt->rt6i_idev;
			struct net_device *rt_dev = rt->dst.dev;

			if (rt_idev && (rt_idev->dev == dev || !dev) &&
			    rt_idev->dev != loopback_dev) {
				rt->rt6i_idev = in6_dev_get(loopback_dev);
				in6_dev_put(rt_idev);
			}

			if (rt_dev && (rt_dev == dev || !dev) &&
			    rt_dev != loopback_dev) {
				rt->dst.dev = loopback_dev;
				dev_hold(rt->dst.dev);
				dev_put(rt_dev);
			}
		}
		spin_unlock_bh(&ul->lock);
	}
}

M
Martin KaFai Lau 已提交
171 172 173 174 175
static u32 *rt6_pcpu_cow_metrics(struct rt6_info *rt)
{
	return dst_metrics_write_ptr(rt->dst.from);
}

176 177
static u32 *ipv6_cow_metrics(struct dst_entry *dst, unsigned long old)
{
178
	struct rt6_info *rt = (struct rt6_info *)dst;
179

M
Martin KaFai Lau 已提交
180 181 182
	if (rt->rt6i_flags & RTF_PCPU)
		return rt6_pcpu_cow_metrics(rt);
	else if (rt->rt6i_flags & RTF_CACHE)
183 184
		return NULL;
	else
185
		return dst_cow_metrics_generic(dst, old);
186 187
}

188 189 190
static inline const void *choose_neigh_daddr(struct rt6_info *rt,
					     struct sk_buff *skb,
					     const void *daddr)
191 192 193
{
	struct in6_addr *p = &rt->rt6i_gateway;

D
David S. Miller 已提交
194
	if (!ipv6_addr_any(p))
195
		return (const void *) p;
196 197
	else if (skb)
		return &ipv6_hdr(skb)->daddr;
198 199 200
	return daddr;
}

201 202 203
static struct neighbour *ip6_neigh_lookup(const struct dst_entry *dst,
					  struct sk_buff *skb,
					  const void *daddr)
204
{
205 206 207
	struct rt6_info *rt = (struct rt6_info *) dst;
	struct neighbour *n;

208
	daddr = choose_neigh_daddr(rt, skb, daddr);
209
	n = __ipv6_neigh_lookup(dst->dev, daddr);
210 211 212 213 214
	if (n)
		return n;
	return neigh_create(&nd_tbl, daddr, dst->dev);
}

215
static struct dst_ops ip6_dst_ops_template = {
L
Linus Torvalds 已提交
216 217 218 219
	.family			=	AF_INET6,
	.gc			=	ip6_dst_gc,
	.gc_thresh		=	1024,
	.check			=	ip6_dst_check,
220
	.default_advmss		=	ip6_default_advmss,
221
	.mtu			=	ip6_mtu,
222
	.cow_metrics		=	ipv6_cow_metrics,
L
Linus Torvalds 已提交
223 224 225 226 227
	.destroy		=	ip6_dst_destroy,
	.ifdown			=	ip6_dst_ifdown,
	.negative_advice	=	ip6_negative_advice,
	.link_failure		=	ip6_link_failure,
	.update_pmtu		=	ip6_rt_update_pmtu,
228
	.redirect		=	rt6_do_redirect,
229
	.local_out		=	__ip6_local_out,
230
	.neigh_lookup		=	ip6_neigh_lookup,
L
Linus Torvalds 已提交
231 232
};

233
static unsigned int ip6_blackhole_mtu(const struct dst_entry *dst)
234
{
235 236 237
	unsigned int mtu = dst_metric_raw(dst, RTAX_MTU);

	return mtu ? : dst->dev->mtu;
238 239
}

240 241
static void ip6_rt_blackhole_update_pmtu(struct dst_entry *dst, struct sock *sk,
					 struct sk_buff *skb, u32 mtu)
242 243 244
{
}

245 246
static void ip6_rt_blackhole_redirect(struct dst_entry *dst, struct sock *sk,
				      struct sk_buff *skb)
247 248 249
{
}

250 251 252 253 254 255
static u32 *ip6_rt_blackhole_cow_metrics(struct dst_entry *dst,
					 unsigned long old)
{
	return NULL;
}

256 257 258 259
static struct dst_ops ip6_dst_blackhole_ops = {
	.family			=	AF_INET6,
	.destroy		=	ip6_dst_destroy,
	.check			=	ip6_dst_check,
260
	.mtu			=	ip6_blackhole_mtu,
261
	.default_advmss		=	ip6_default_advmss,
262
	.update_pmtu		=	ip6_rt_blackhole_update_pmtu,
263
	.redirect		=	ip6_rt_blackhole_redirect,
264
	.cow_metrics		=	ip6_rt_blackhole_cow_metrics,
265
	.neigh_lookup		=	ip6_neigh_lookup,
266 267
};

268
static const u32 ip6_template_metrics[RTAX_MAX] = {
L
Li RongQing 已提交
269
	[RTAX_HOPLIMIT - 1] = 0,
270 271
};

272
static const struct rt6_info ip6_null_entry_template = {
273 274 275
	.dst = {
		.__refcnt	= ATOMIC_INIT(1),
		.__use		= 1,
276
		.obsolete	= DST_OBSOLETE_FORCE_CHK,
277 278 279
		.error		= -ENETUNREACH,
		.input		= ip6_pkt_discard,
		.output		= ip6_pkt_discard_out,
L
Linus Torvalds 已提交
280 281
	},
	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
282
	.rt6i_protocol  = RTPROT_KERNEL,
L
Linus Torvalds 已提交
283 284 285 286
	.rt6i_metric	= ~(u32) 0,
	.rt6i_ref	= ATOMIC_INIT(1),
};

T
Thomas Graf 已提交
287 288
#ifdef CONFIG_IPV6_MULTIPLE_TABLES

289
static const struct rt6_info ip6_prohibit_entry_template = {
290 291 292
	.dst = {
		.__refcnt	= ATOMIC_INIT(1),
		.__use		= 1,
293
		.obsolete	= DST_OBSOLETE_FORCE_CHK,
294 295 296
		.error		= -EACCES,
		.input		= ip6_pkt_prohibit,
		.output		= ip6_pkt_prohibit_out,
T
Thomas Graf 已提交
297 298
	},
	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
299
	.rt6i_protocol  = RTPROT_KERNEL,
T
Thomas Graf 已提交
300 301 302 303
	.rt6i_metric	= ~(u32) 0,
	.rt6i_ref	= ATOMIC_INIT(1),
};

304
static const struct rt6_info ip6_blk_hole_entry_template = {
305 306 307
	.dst = {
		.__refcnt	= ATOMIC_INIT(1),
		.__use		= 1,
308
		.obsolete	= DST_OBSOLETE_FORCE_CHK,
309 310
		.error		= -EINVAL,
		.input		= dst_discard,
311
		.output		= dst_discard_sk,
T
Thomas Graf 已提交
312 313
	},
	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
314
	.rt6i_protocol  = RTPROT_KERNEL,
T
Thomas Graf 已提交
315 316 317 318 319 320
	.rt6i_metric	= ~(u32) 0,
	.rt6i_ref	= ATOMIC_INIT(1),
};

#endif

L
Linus Torvalds 已提交
321
/* allocate dst with ip6_dst_ops */
M
Martin KaFai Lau 已提交
322 323
static struct rt6_info *__ip6_dst_alloc(struct net *net,
					struct net_device *dev,
324
					int flags)
L
Linus Torvalds 已提交
325
{
326
	struct rt6_info *rt = dst_alloc(&net->ipv6.ip6_dst_ops, dev,
327
					0, DST_OBSOLETE_FORCE_CHK, flags);
328

329
	if (rt) {
330 331 332
		struct dst_entry *dst = &rt->dst;

		memset(dst + 1, 0, sizeof(*rt) - sizeof(*dst));
333
		INIT_LIST_HEAD(&rt->rt6i_siblings);
334
		INIT_LIST_HEAD(&rt->rt6i_uncached);
335
	}
336
	return rt;
L
Linus Torvalds 已提交
337 338
}

M
Martin KaFai Lau 已提交
339 340
static struct rt6_info *ip6_dst_alloc(struct net *net,
				      struct net_device *dev,
341
				      int flags)
M
Martin KaFai Lau 已提交
342
{
343
	struct rt6_info *rt = __ip6_dst_alloc(net, dev, flags);
M
Martin KaFai Lau 已提交
344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365

	if (rt) {
		rt->rt6i_pcpu = alloc_percpu_gfp(struct rt6_info *, GFP_ATOMIC);
		if (rt->rt6i_pcpu) {
			int cpu;

			for_each_possible_cpu(cpu) {
				struct rt6_info **p;

				p = per_cpu_ptr(rt->rt6i_pcpu, cpu);
				/* no one shares rt */
				*p =  NULL;
			}
		} else {
			dst_destroy((struct dst_entry *)rt);
			return NULL;
		}
	}

	return rt;
}

L
Linus Torvalds 已提交
366 367 368
static void ip6_dst_destroy(struct dst_entry *dst)
{
	struct rt6_info *rt = (struct rt6_info *)dst;
369
	struct dst_entry *from = dst->from;
370
	struct inet6_dev *idev;
L
Linus Torvalds 已提交
371

372
	dst_destroy_metrics_generic(dst);
373
	free_percpu(rt->rt6i_pcpu);
374 375 376
	rt6_uncached_list_del(rt);

	idev = rt->rt6i_idev;
377
	if (idev) {
L
Linus Torvalds 已提交
378 379
		rt->rt6i_idev = NULL;
		in6_dev_put(idev);
380
	}
381

382 383
	dst->from = NULL;
	dst_release(from);
384 385
}

L
Linus Torvalds 已提交
386 387 388 389 390
static void ip6_dst_ifdown(struct dst_entry *dst, struct net_device *dev,
			   int how)
{
	struct rt6_info *rt = (struct rt6_info *)dst;
	struct inet6_dev *idev = rt->rt6i_idev;
391
	struct net_device *loopback_dev =
392
		dev_net(dev)->loopback_dev;
L
Linus Torvalds 已提交
393

394 395 396 397 398 399 400 401 402
	if (dev != loopback_dev) {
		if (idev && idev->dev == dev) {
			struct inet6_dev *loopback_idev =
				in6_dev_get(loopback_dev);
			if (loopback_idev) {
				rt->rt6i_idev = loopback_idev;
				in6_dev_put(idev);
			}
		}
L
Linus Torvalds 已提交
403 404 405
	}
}

406
static bool rt6_check_expired(const struct rt6_info *rt)
L
Linus Torvalds 已提交
407
{
408 409
	if (rt->rt6i_flags & RTF_EXPIRES) {
		if (time_after(jiffies, rt->dst.expires))
410
			return true;
411
	} else if (rt->dst.from) {
412
		return rt6_check_expired((struct rt6_info *) rt->dst.from);
413
	}
414
	return false;
L
Linus Torvalds 已提交
415 416
}

417 418 419 420 421 422 423 424 425
/* Multipath route selection:
 *   Hash based function using packet header and flowlabel.
 * Adapted from fib_info_hashfn()
 */
static int rt6_info_hash_nhsfn(unsigned int candidate_count,
			       const struct flowi6 *fl6)
{
	unsigned int val = fl6->flowi6_proto;

426 427
	val ^= ipv6_addr_hash(&fl6->daddr);
	val ^= ipv6_addr_hash(&fl6->saddr);
428 429 430 431 432 433

	/* Work only if this not encapsulated */
	switch (fl6->flowi6_proto) {
	case IPPROTO_UDP:
	case IPPROTO_TCP:
	case IPPROTO_SCTP:
434 435
		val ^= (__force u16)fl6->fl6_sport;
		val ^= (__force u16)fl6->fl6_dport;
436 437 438
		break;

	case IPPROTO_ICMPV6:
439 440
		val ^= (__force u16)fl6->fl6_icmp_type;
		val ^= (__force u16)fl6->fl6_icmp_code;
441 442 443
		break;
	}
	/* RFC6438 recommands to use flowlabel */
444
	val ^= (__force u32)fl6->flowlabel;
445 446 447 448 449 450 451

	/* Perhaps, we need to tune, this function? */
	val = val ^ (val >> 7) ^ (val >> 12);
	return val % candidate_count;
}

static struct rt6_info *rt6_multipath_select(struct rt6_info *match,
452 453
					     struct flowi6 *fl6, int oif,
					     int strict)
454 455 456 457 458 459 460 461 462 463 464 465 466
{
	struct rt6_info *sibling, *next_sibling;
	int route_choosen;

	route_choosen = rt6_info_hash_nhsfn(match->rt6i_nsiblings + 1, fl6);
	/* Don't change the route, if route_choosen == 0
	 * (siblings does not include ourself)
	 */
	if (route_choosen)
		list_for_each_entry_safe(sibling, next_sibling,
				&match->rt6i_siblings, rt6i_siblings) {
			route_choosen--;
			if (route_choosen == 0) {
467 468
				if (rt6_score_route(sibling, oif, strict) < 0)
					break;
469 470 471 472 473 474 475
				match = sibling;
				break;
			}
		}
	return match;
}

L
Linus Torvalds 已提交
476
/*
T
Thomas Graf 已提交
477
 *	Route lookup. Any table->tb6_lock is implied.
L
Linus Torvalds 已提交
478 479
 */

480 481
static inline struct rt6_info *rt6_device_match(struct net *net,
						    struct rt6_info *rt,
482
						    const struct in6_addr *saddr,
L
Linus Torvalds 已提交
483
						    int oif,
484
						    int flags)
L
Linus Torvalds 已提交
485 486 487 488
{
	struct rt6_info *local = NULL;
	struct rt6_info *sprt;

489 490 491
	if (!oif && ipv6_addr_any(saddr))
		goto out;

492
	for (sprt = rt; sprt; sprt = sprt->dst.rt6_next) {
493
		struct net_device *dev = sprt->dst.dev;
494 495

		if (oif) {
L
Linus Torvalds 已提交
496 497 498
			if (dev->ifindex == oif)
				return sprt;
			if (dev->flags & IFF_LOOPBACK) {
499
				if (!sprt->rt6i_idev ||
L
Linus Torvalds 已提交
500
				    sprt->rt6i_idev->dev->ifindex != oif) {
501
					if (flags & RT6_LOOKUP_F_IFACE && oif)
L
Linus Torvalds 已提交
502
						continue;
503
					if (local && (!oif ||
L
Linus Torvalds 已提交
504 505 506 507 508
						      local->rt6i_idev->dev->ifindex == oif))
						continue;
				}
				local = sprt;
			}
509 510 511 512
		} else {
			if (ipv6_chk_addr(net, saddr, dev,
					  flags & RT6_LOOKUP_F_IFACE))
				return sprt;
L
Linus Torvalds 已提交
513
		}
514
	}
L
Linus Torvalds 已提交
515

516
	if (oif) {
L
Linus Torvalds 已提交
517 518 519
		if (local)
			return local;

520
		if (flags & RT6_LOOKUP_F_IFACE)
521
			return net->ipv6.ip6_null_entry;
L
Linus Torvalds 已提交
522
	}
523
out:
L
Linus Torvalds 已提交
524 525 526
	return rt;
}

527
#ifdef CONFIG_IPV6_ROUTER_PREF
528 529 530 531 532 533 534 535 536 537 538 539 540
struct __rt6_probe_work {
	struct work_struct work;
	struct in6_addr target;
	struct net_device *dev;
};

static void rt6_probe_deferred(struct work_struct *w)
{
	struct in6_addr mcaddr;
	struct __rt6_probe_work *work =
		container_of(w, struct __rt6_probe_work, work);

	addrconf_addr_solict_mult(&work->target, &mcaddr);
541
	ndisc_send_ns(work->dev, NULL, &work->target, &mcaddr, NULL, NULL);
542
	dev_put(work->dev);
543
	kfree(work);
544 545
}

546 547
static void rt6_probe(struct rt6_info *rt)
{
548
	struct __rt6_probe_work *work;
549
	struct neighbour *neigh;
550 551 552 553 554 555 556 557
	/*
	 * Okay, this does not seem to be appropriate
	 * for now, however, we need to check if it
	 * is really so; aka Router Reachability Probing.
	 *
	 * Router Reachability Probe MUST be rate-limited
	 * to no more than one per minute.
	 */
558
	if (!rt || !(rt->rt6i_flags & RTF_GATEWAY))
559
		return;
560 561 562
	rcu_read_lock_bh();
	neigh = __ipv6_neigh_lookup_noref(rt->dst.dev, &rt->rt6i_gateway);
	if (neigh) {
563 564 565
		if (neigh->nud_state & NUD_VALID)
			goto out;

566
		work = NULL;
567
		write_lock(&neigh->lock);
568 569 570 571 572 573 574
		if (!(neigh->nud_state & NUD_VALID) &&
		    time_after(jiffies,
			       neigh->updated +
			       rt->rt6i_idev->cnf.rtr_probe_interval)) {
			work = kmalloc(sizeof(*work), GFP_ATOMIC);
			if (work)
				__neigh_set_probe_once(neigh);
575
		}
576
		write_unlock(&neigh->lock);
577 578
	} else {
		work = kmalloc(sizeof(*work), GFP_ATOMIC);
579
	}
580 581 582 583 584 585 586 587 588

	if (work) {
		INIT_WORK(&work->work, rt6_probe_deferred);
		work->target = rt->rt6i_gateway;
		dev_hold(rt->dst.dev);
		work->dev = rt->dst.dev;
		schedule_work(&work->work);
	}

589
out:
590
	rcu_read_unlock_bh();
591 592 593 594 595 596 597
}
#else
static inline void rt6_probe(struct rt6_info *rt)
{
}
#endif

L
Linus Torvalds 已提交
598
/*
599
 * Default Router Selection (RFC 2461 6.3.6)
L
Linus Torvalds 已提交
600
 */
D
Dave Jones 已提交
601
static inline int rt6_check_dev(struct rt6_info *rt, int oif)
602
{
603
	struct net_device *dev = rt->dst.dev;
604
	if (!oif || dev->ifindex == oif)
605
		return 2;
606 607 608 609
	if ((dev->flags & IFF_LOOPBACK) &&
	    rt->rt6i_idev && rt->rt6i_idev->dev->ifindex == oif)
		return 1;
	return 0;
610
}
L
Linus Torvalds 已提交
611

612
static inline enum rt6_nud_state rt6_check_neigh(struct rt6_info *rt)
L
Linus Torvalds 已提交
613
{
614
	struct neighbour *neigh;
615
	enum rt6_nud_state ret = RT6_NUD_FAIL_HARD;
616

617 618
	if (rt->rt6i_flags & RTF_NONEXTHOP ||
	    !(rt->rt6i_flags & RTF_GATEWAY))
619
		return RT6_NUD_SUCCEED;
620 621 622 623 624

	rcu_read_lock_bh();
	neigh = __ipv6_neigh_lookup_noref(rt->dst.dev, &rt->rt6i_gateway);
	if (neigh) {
		read_lock(&neigh->lock);
625
		if (neigh->nud_state & NUD_VALID)
626
			ret = RT6_NUD_SUCCEED;
627
#ifdef CONFIG_IPV6_ROUTER_PREF
628
		else if (!(neigh->nud_state & NUD_FAILED))
629
			ret = RT6_NUD_SUCCEED;
J
Jiri Benc 已提交
630 631
		else
			ret = RT6_NUD_FAIL_PROBE;
632
#endif
633
		read_unlock(&neigh->lock);
634 635
	} else {
		ret = IS_ENABLED(CONFIG_IPV6_ROUTER_PREF) ?
J
Jiri Benc 已提交
636
		      RT6_NUD_SUCCEED : RT6_NUD_FAIL_DO_RR;
637
	}
638 639
	rcu_read_unlock_bh();

640
	return ret;
L
Linus Torvalds 已提交
641 642
}

643 644
static int rt6_score_route(struct rt6_info *rt, int oif,
			   int strict)
L
Linus Torvalds 已提交
645
{
646
	int m;
647

648
	m = rt6_check_dev(rt, oif);
649
	if (!m && (strict & RT6_LOOKUP_F_IFACE))
650
		return RT6_NUD_FAIL_HARD;
651 652 653
#ifdef CONFIG_IPV6_ROUTER_PREF
	m |= IPV6_DECODE_PREF(IPV6_EXTRACT_PREF(rt->rt6i_flags)) << 2;
#endif
654 655 656 657 658
	if (strict & RT6_LOOKUP_F_REACHABLE) {
		int n = rt6_check_neigh(rt);
		if (n < 0)
			return n;
	}
659 660 661
	return m;
}

662
static struct rt6_info *find_match(struct rt6_info *rt, int oif, int strict,
663 664
				   int *mpri, struct rt6_info *match,
				   bool *do_rr)
665
{
666
	int m;
667
	bool match_do_rr = false;
668 669 670 671 672 673
	struct inet6_dev *idev = rt->rt6i_idev;
	struct net_device *dev = rt->dst.dev;

	if (dev && !netif_carrier_ok(dev) &&
	    idev->cnf.ignore_routes_with_linkdown)
		goto out;
674 675 676 677 678

	if (rt6_check_expired(rt))
		goto out;

	m = rt6_score_route(rt, oif, strict);
J
Jiri Benc 已提交
679
	if (m == RT6_NUD_FAIL_DO_RR) {
680 681
		match_do_rr = true;
		m = 0; /* lowest valid score */
J
Jiri Benc 已提交
682
	} else if (m == RT6_NUD_FAIL_HARD) {
683
		goto out;
684 685 686 687
	}

	if (strict & RT6_LOOKUP_F_REACHABLE)
		rt6_probe(rt);
688

J
Jiri Benc 已提交
689
	/* note that m can be RT6_NUD_FAIL_PROBE at this point */
690
	if (m > *mpri) {
691
		*do_rr = match_do_rr;
692 693 694 695 696 697 698 699 700
		*mpri = m;
		match = rt;
	}
out:
	return match;
}

static struct rt6_info *find_rr_leaf(struct fib6_node *fn,
				     struct rt6_info *rr_head,
701 702
				     u32 metric, int oif, int strict,
				     bool *do_rr)
703
{
704
	struct rt6_info *rt, *match, *cont;
705
	int mpri = -1;
L
Linus Torvalds 已提交
706

707
	match = NULL;
708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723
	cont = NULL;
	for (rt = rr_head; rt; rt = rt->dst.rt6_next) {
		if (rt->rt6i_metric != metric) {
			cont = rt;
			break;
		}

		match = find_match(rt, oif, strict, &mpri, match, do_rr);
	}

	for (rt = fn->leaf; rt && rt != rr_head; rt = rt->dst.rt6_next) {
		if (rt->rt6i_metric != metric) {
			cont = rt;
			break;
		}

724
		match = find_match(rt, oif, strict, &mpri, match, do_rr);
725 726 727 728 729 730
	}

	if (match || !cont)
		return match;

	for (rt = cont; rt; rt = rt->dst.rt6_next)
731
		match = find_match(rt, oif, strict, &mpri, match, do_rr);
L
Linus Torvalds 已提交
732

733 734
	return match;
}
L
Linus Torvalds 已提交
735

736 737 738
static struct rt6_info *rt6_select(struct fib6_node *fn, int oif, int strict)
{
	struct rt6_info *match, *rt0;
739
	struct net *net;
740
	bool do_rr = false;
L
Linus Torvalds 已提交
741

742 743 744
	rt0 = fn->rr_ptr;
	if (!rt0)
		fn->rr_ptr = rt0 = fn->leaf;
L
Linus Torvalds 已提交
745

746 747
	match = find_rr_leaf(fn, rt0, rt0->rt6i_metric, oif, strict,
			     &do_rr);
L
Linus Torvalds 已提交
748

749
	if (do_rr) {
750
		struct rt6_info *next = rt0->dst.rt6_next;
751

752
		/* no entries matched; do round-robin */
753 754 755 756 757
		if (!next || next->rt6i_metric != rt0->rt6i_metric)
			next = fn->leaf;

		if (next != rt0)
			fn->rr_ptr = next;
L
Linus Torvalds 已提交
758 759
	}

760
	net = dev_net(rt0->dst.dev);
E
Eric Dumazet 已提交
761
	return match ? match : net->ipv6.ip6_null_entry;
L
Linus Torvalds 已提交
762 763
}

764 765 766 767 768
static bool rt6_is_gw_or_nonexthop(const struct rt6_info *rt)
{
	return (rt->rt6i_flags & (RTF_NONEXTHOP | RTF_GATEWAY));
}

769 770
#ifdef CONFIG_IPV6_ROUTE_INFO
int rt6_route_rcv(struct net_device *dev, u8 *opt, int len,
771
		  const struct in6_addr *gwaddr)
772
{
773
	struct net *net = dev_net(dev);
774 775 776
	struct route_info *rinfo = (struct route_info *) opt;
	struct in6_addr prefix_buf, *prefix;
	unsigned int pref;
777
	unsigned long lifetime;
778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800
	struct rt6_info *rt;

	if (len < sizeof(struct route_info)) {
		return -EINVAL;
	}

	/* Sanity check for prefix_len and length */
	if (rinfo->length > 3) {
		return -EINVAL;
	} else if (rinfo->prefix_len > 128) {
		return -EINVAL;
	} else if (rinfo->prefix_len > 64) {
		if (rinfo->length < 2) {
			return -EINVAL;
		}
	} else if (rinfo->prefix_len > 0) {
		if (rinfo->length < 1) {
			return -EINVAL;
		}
	}

	pref = rinfo->route_pref;
	if (pref == ICMPV6_ROUTER_PREF_INVALID)
801
		return -EINVAL;
802

803
	lifetime = addrconf_timeout_fixup(ntohl(rinfo->lifetime), HZ);
804 805 806 807 808 809 810 811 812 813 814

	if (rinfo->length == 3)
		prefix = (struct in6_addr *)rinfo->prefix;
	else {
		/* this function is safe */
		ipv6_addr_prefix(&prefix_buf,
				 (struct in6_addr *)rinfo->prefix,
				 rinfo->prefix_len);
		prefix = &prefix_buf;
	}

815 816 817 818 819
	if (rinfo->prefix_len == 0)
		rt = rt6_get_dflt_router(gwaddr, dev);
	else
		rt = rt6_get_route_info(net, prefix, rinfo->prefix_len,
					gwaddr, dev->ifindex);
820 821

	if (rt && !lifetime) {
822
		ip6_del_rt(rt);
823 824 825 826
		rt = NULL;
	}

	if (!rt && lifetime)
827
		rt = rt6_add_route_info(net, prefix, rinfo->prefix_len, gwaddr, dev->ifindex,
828 829 830 831 832 833
					pref);
	else if (rt)
		rt->rt6i_flags = RTF_ROUTEINFO |
				 (rt->rt6i_flags & ~RTF_PREF_MASK) | RTF_PREF(pref);

	if (rt) {
834 835 836 837 838
		if (!addrconf_finite_timeout(lifetime))
			rt6_clean_expires(rt);
		else
			rt6_set_expires(rt, jiffies + HZ * lifetime);

A
Amerigo Wang 已提交
839
		ip6_rt_put(rt);
840 841 842 843 844
	}
	return 0;
}
#endif

M
Martin KaFai Lau 已提交
845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860
static struct fib6_node* fib6_backtrack(struct fib6_node *fn,
					struct in6_addr *saddr)
{
	struct fib6_node *pn;
	while (1) {
		if (fn->fn_flags & RTN_TL_ROOT)
			return NULL;
		pn = fn->parent;
		if (FIB6_SUBTREE(pn) && FIB6_SUBTREE(pn) != fn)
			fn = fib6_lookup(FIB6_SUBTREE(pn), NULL, saddr);
		else
			fn = pn;
		if (fn->fn_flags & RTN_RTINFO)
			return fn;
	}
}
T
Thomas Graf 已提交
861

862 863
static struct rt6_info *ip6_pol_route_lookup(struct net *net,
					     struct fib6_table *table,
864
					     struct flowi6 *fl6, int flags)
L
Linus Torvalds 已提交
865 866 867 868
{
	struct fib6_node *fn;
	struct rt6_info *rt;

T
Thomas Graf 已提交
869
	read_lock_bh(&table->tb6_lock);
870
	fn = fib6_lookup(&table->tb6_root, &fl6->daddr, &fl6->saddr);
T
Thomas Graf 已提交
871 872
restart:
	rt = fn->leaf;
873
	rt = rt6_device_match(net, rt, &fl6->saddr, fl6->flowi6_oif, flags);
874
	if (rt->rt6i_nsiblings && fl6->flowi6_oif == 0)
875
		rt = rt6_multipath_select(rt, fl6, fl6->flowi6_oif, flags);
M
Martin KaFai Lau 已提交
876 877 878 879 880
	if (rt == net->ipv6.ip6_null_entry) {
		fn = fib6_backtrack(fn, &fl6->saddr);
		if (fn)
			goto restart;
	}
881
	dst_use(&rt->dst, jiffies);
T
Thomas Graf 已提交
882 883 884 885 886
	read_unlock_bh(&table->tb6_lock);
	return rt;

}

887
struct dst_entry *ip6_route_lookup(struct net *net, struct flowi6 *fl6,
F
Florian Westphal 已提交
888 889 890 891 892 893
				    int flags)
{
	return fib6_rule_lookup(net, fl6, flags, ip6_pol_route_lookup);
}
EXPORT_SYMBOL_GPL(ip6_route_lookup);

894 895
struct rt6_info *rt6_lookup(struct net *net, const struct in6_addr *daddr,
			    const struct in6_addr *saddr, int oif, int strict)
T
Thomas Graf 已提交
896
{
897 898 899
	struct flowi6 fl6 = {
		.flowi6_oif = oif,
		.daddr = *daddr,
T
Thomas Graf 已提交
900 901
	};
	struct dst_entry *dst;
902
	int flags = strict ? RT6_LOOKUP_F_IFACE : 0;
T
Thomas Graf 已提交
903

904
	if (saddr) {
905
		memcpy(&fl6.saddr, saddr, sizeof(*saddr));
906 907 908
		flags |= RT6_LOOKUP_F_HAS_SADDR;
	}

909
	dst = fib6_rule_lookup(net, &fl6, flags, ip6_pol_route_lookup);
T
Thomas Graf 已提交
910 911 912 913 914
	if (dst->error == 0)
		return (struct rt6_info *) dst;

	dst_release(dst);

L
Linus Torvalds 已提交
915 916
	return NULL;
}
917 918
EXPORT_SYMBOL(rt6_lookup);

T
Thomas Graf 已提交
919
/* ip6_ins_rt is called with FREE table->tb6_lock.
L
Linus Torvalds 已提交
920 921 922 923 924
   It takes new route entry, the addition fails by any reason the
   route is freed. In any case, if caller does not hold it, it may
   be destroyed.
 */

925
static int __ip6_ins_rt(struct rt6_info *rt, struct nl_info *info,
926
			struct mx6_config *mxc)
L
Linus Torvalds 已提交
927 928
{
	int err;
T
Thomas Graf 已提交
929
	struct fib6_table *table;
L
Linus Torvalds 已提交
930

T
Thomas Graf 已提交
931 932
	table = rt->rt6i_table;
	write_lock_bh(&table->tb6_lock);
933
	err = fib6_add(&table->tb6_root, rt, info, mxc);
T
Thomas Graf 已提交
934
	write_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
935 936 937 938

	return err;
}

939 940
int ip6_ins_rt(struct rt6_info *rt)
{
941 942 943 944
	struct nl_info info = {	.nl_net = dev_net(rt->dst.dev), };
	struct mx6_config mxc = { .mx = NULL, };

	return __ip6_ins_rt(rt, &info, &mxc);
945 946
}

947 948 949
static struct rt6_info *ip6_rt_cache_alloc(struct rt6_info *ort,
					   const struct in6_addr *daddr,
					   const struct in6_addr *saddr)
L
Linus Torvalds 已提交
950 951 952 953 954 955 956
{
	struct rt6_info *rt;

	/*
	 *	Clone the route.
	 */

M
Martin KaFai Lau 已提交
957
	if (ort->rt6i_flags & (RTF_CACHE | RTF_PCPU))
M
Martin KaFai Lau 已提交
958
		ort = (struct rt6_info *)ort->dst.from;
L
Linus Torvalds 已提交
959

960
	rt = __ip6_dst_alloc(dev_net(ort->dst.dev), ort->dst.dev, 0);
M
Martin KaFai Lau 已提交
961 962 963 964 965 966 967 968 969 970

	if (!rt)
		return NULL;

	ip6_rt_copy_init(rt, ort);
	rt->rt6i_flags |= RTF_CACHE;
	rt->rt6i_metric = 0;
	rt->dst.flags |= DST_HOST;
	rt->rt6i_dst.addr = *daddr;
	rt->rt6i_dst.plen = 128;
L
Linus Torvalds 已提交
971

M
Martin KaFai Lau 已提交
972 973 974 975
	if (!rt6_is_gw_or_nonexthop(ort)) {
		if (ort->rt6i_dst.plen != 128 &&
		    ipv6_addr_equal(&ort->rt6i_dst.addr, daddr))
			rt->rt6i_flags |= RTF_ANYCAST;
L
Linus Torvalds 已提交
976
#ifdef CONFIG_IPV6_SUBTREES
M
Martin KaFai Lau 已提交
977 978 979
		if (rt->rt6i_src.plen && saddr) {
			rt->rt6i_src.addr = *saddr;
			rt->rt6i_src.plen = 128;
980
		}
M
Martin KaFai Lau 已提交
981
#endif
982
	}
L
Linus Torvalds 已提交
983

984 985
	return rt;
}
L
Linus Torvalds 已提交
986

M
Martin KaFai Lau 已提交
987 988 989 990 991
static struct rt6_info *ip6_rt_pcpu_alloc(struct rt6_info *rt)
{
	struct rt6_info *pcpu_rt;

	pcpu_rt = __ip6_dst_alloc(dev_net(rt->dst.dev),
992
				  rt->dst.dev, rt->dst.flags);
M
Martin KaFai Lau 已提交
993 994 995 996 997 998 999 1000 1001 1002 1003 1004

	if (!pcpu_rt)
		return NULL;
	ip6_rt_copy_init(pcpu_rt, rt);
	pcpu_rt->rt6i_protocol = rt->rt6i_protocol;
	pcpu_rt->rt6i_flags |= RTF_PCPU;
	return pcpu_rt;
}

/* It should be called with read_lock_bh(&tb6_lock) acquired */
static struct rt6_info *rt6_get_pcpu_route(struct rt6_info *rt)
{
1005
	struct rt6_info *pcpu_rt, **p;
M
Martin KaFai Lau 已提交
1006 1007 1008 1009

	p = this_cpu_ptr(rt->rt6i_pcpu);
	pcpu_rt = *p;

1010 1011 1012 1013 1014 1015 1016 1017 1018
	if (pcpu_rt) {
		dst_hold(&pcpu_rt->dst);
		rt6_dst_from_metrics_check(pcpu_rt);
	}
	return pcpu_rt;
}

static struct rt6_info *rt6_make_pcpu_route(struct rt6_info *rt)
{
1019
	struct fib6_table *table = rt->rt6i_table;
1020
	struct rt6_info *pcpu_rt, *prev, **p;
M
Martin KaFai Lau 已提交
1021 1022 1023 1024 1025

	pcpu_rt = ip6_rt_pcpu_alloc(rt);
	if (!pcpu_rt) {
		struct net *net = dev_net(rt->dst.dev);

1026 1027
		dst_hold(&net->ipv6.ip6_null_entry->dst);
		return net->ipv6.ip6_null_entry;
M
Martin KaFai Lau 已提交
1028 1029
	}

1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045
	read_lock_bh(&table->tb6_lock);
	if (rt->rt6i_pcpu) {
		p = this_cpu_ptr(rt->rt6i_pcpu);
		prev = cmpxchg(p, NULL, pcpu_rt);
		if (prev) {
			/* If someone did it before us, return prev instead */
			dst_destroy(&pcpu_rt->dst);
			pcpu_rt = prev;
		}
	} else {
		/* rt has been removed from the fib6 tree
		 * before we have a chance to acquire the read_lock.
		 * In this case, don't brother to create a pcpu rt
		 * since rt is going away anyway.  The next
		 * dst_check() will trigger a re-lookup.
		 */
M
Martin KaFai Lau 已提交
1046
		dst_destroy(&pcpu_rt->dst);
1047
		pcpu_rt = rt;
M
Martin KaFai Lau 已提交
1048 1049 1050
	}
	dst_hold(&pcpu_rt->dst);
	rt6_dst_from_metrics_check(pcpu_rt);
1051
	read_unlock_bh(&table->tb6_lock);
M
Martin KaFai Lau 已提交
1052 1053 1054
	return pcpu_rt;
}

1055
static struct rt6_info *ip6_pol_route(struct net *net, struct fib6_table *table, int oif,
1056
				      struct flowi6 *fl6, int flags)
L
Linus Torvalds 已提交
1057
{
1058
	struct fib6_node *fn, *saved_fn;
1059
	struct rt6_info *rt;
T
Thomas Graf 已提交
1060
	int strict = 0;
L
Linus Torvalds 已提交
1061

1062
	strict |= flags & RT6_LOOKUP_F_IFACE;
1063 1064
	if (net->ipv6.devconf_all->forwarding == 0)
		strict |= RT6_LOOKUP_F_REACHABLE;
L
Linus Torvalds 已提交
1065

T
Thomas Graf 已提交
1066
	read_lock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1067

1068
	fn = fib6_lookup(&table->tb6_root, &fl6->daddr, &fl6->saddr);
1069
	saved_fn = fn;
L
Linus Torvalds 已提交
1070

M
Martin KaFai Lau 已提交
1071
redo_rt6_select:
1072
	rt = rt6_select(fn, oif, strict);
1073
	if (rt->rt6i_nsiblings)
1074
		rt = rt6_multipath_select(rt, fl6, oif, strict);
M
Martin KaFai Lau 已提交
1075 1076 1077 1078
	if (rt == net->ipv6.ip6_null_entry) {
		fn = fib6_backtrack(fn, &fl6->saddr);
		if (fn)
			goto redo_rt6_select;
1079 1080 1081 1082 1083 1084
		else if (strict & RT6_LOOKUP_F_REACHABLE) {
			/* also consider unreachable route */
			strict &= ~RT6_LOOKUP_F_REACHABLE;
			fn = saved_fn;
			goto redo_rt6_select;
		}
M
Martin KaFai Lau 已提交
1085 1086
	}

1087

1088
	if (rt == net->ipv6.ip6_null_entry || (rt->rt6i_flags & RTF_CACHE)) {
M
Martin KaFai Lau 已提交
1089 1090 1091 1092 1093
		dst_use(&rt->dst, jiffies);
		read_unlock_bh(&table->tb6_lock);

		rt6_dst_from_metrics_check(rt);
		return rt;
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103
	} else if (unlikely((fl6->flowi6_flags & FLOWI_FLAG_KNOWN_NH) &&
			    !(rt->rt6i_flags & RTF_GATEWAY))) {
		/* Create a RTF_CACHE clone which will not be
		 * owned by the fib6 tree.  It is for the special case where
		 * the daddr in the skb during the neighbor look-up is different
		 * from the fl6->daddr used to look-up route here.
		 */

		struct rt6_info *uncached_rt;

M
Martin KaFai Lau 已提交
1104 1105 1106
		dst_use(&rt->dst, jiffies);
		read_unlock_bh(&table->tb6_lock);

1107 1108
		uncached_rt = ip6_rt_cache_alloc(rt, &fl6->daddr, NULL);
		dst_release(&rt->dst);
T
Thomas Graf 已提交
1109

1110
		if (uncached_rt)
1111
			rt6_uncached_list_add(uncached_rt);
1112 1113
		else
			uncached_rt = net->ipv6.ip6_null_entry;
M
Martin KaFai Lau 已提交
1114

1115 1116 1117
		dst_hold(&uncached_rt->dst);
		return uncached_rt;

M
Martin KaFai Lau 已提交
1118 1119 1120 1121 1122 1123 1124 1125 1126
	} else {
		/* Get a percpu copy */

		struct rt6_info *pcpu_rt;

		rt->dst.lastuse = jiffies;
		rt->dst.__use++;
		pcpu_rt = rt6_get_pcpu_route(rt);

1127 1128 1129 1130 1131 1132 1133 1134 1135
		if (pcpu_rt) {
			read_unlock_bh(&table->tb6_lock);
		} else {
			/* We have to do the read_unlock first
			 * because rt6_make_pcpu_route() may trigger
			 * ip6_dst_gc() which will take the write_lock.
			 */
			dst_hold(&rt->dst);
			read_unlock_bh(&table->tb6_lock);
1136
			pcpu_rt = rt6_make_pcpu_route(rt);
1137 1138
			dst_release(&rt->dst);
		}
M
Martin KaFai Lau 已提交
1139 1140

		return pcpu_rt;
1141

M
Martin KaFai Lau 已提交
1142
	}
L
Linus Torvalds 已提交
1143 1144
}

1145
static struct rt6_info *ip6_pol_route_input(struct net *net, struct fib6_table *table,
1146
					    struct flowi6 *fl6, int flags)
1147
{
1148
	return ip6_pol_route(net, table, fl6->flowi6_iif, fl6, flags);
1149 1150
}

1151 1152 1153 1154 1155 1156 1157 1158 1159 1160
static struct dst_entry *ip6_route_input_lookup(struct net *net,
						struct net_device *dev,
						struct flowi6 *fl6, int flags)
{
	if (rt6_need_strict(&fl6->daddr) && dev->type != ARPHRD_PIMREG)
		flags |= RT6_LOOKUP_F_IFACE;

	return fib6_rule_lookup(net, fl6, flags, ip6_pol_route_input);
}

T
Thomas Graf 已提交
1161 1162
void ip6_route_input(struct sk_buff *skb)
{
1163
	const struct ipv6hdr *iph = ipv6_hdr(skb);
1164
	struct net *net = dev_net(skb->dev);
1165
	int flags = RT6_LOOKUP_F_HAS_SADDR;
1166
	struct ip_tunnel_info *tun_info;
1167 1168 1169 1170
	struct flowi6 fl6 = {
		.flowi6_iif = skb->dev->ifindex,
		.daddr = iph->daddr,
		.saddr = iph->saddr,
1171
		.flowlabel = ip6_flowinfo(iph),
1172 1173
		.flowi6_mark = skb->mark,
		.flowi6_proto = iph->nexthdr,
T
Thomas Graf 已提交
1174
	};
1175

1176
	tun_info = skb_tunnel_info(skb);
1177
	if (tun_info && !(tun_info->mode & IP_TUNNEL_INFO_TX))
1178
		fl6.flowi6_tun_key.tun_id = tun_info->key.tun_id;
1179
	skb_dst_drop(skb);
1180
	skb_dst_set(skb, ip6_route_input_lookup(net, skb->dev, &fl6, flags));
T
Thomas Graf 已提交
1181 1182
}

1183
static struct rt6_info *ip6_pol_route_output(struct net *net, struct fib6_table *table,
1184
					     struct flowi6 *fl6, int flags)
L
Linus Torvalds 已提交
1185
{
1186
	return ip6_pol_route(net, table, fl6->flowi6_oif, fl6, flags);
T
Thomas Graf 已提交
1187 1188
}

1189
struct dst_entry *ip6_route_output(struct net *net, const struct sock *sk,
1190
				    struct flowi6 *fl6)
T
Thomas Graf 已提交
1191 1192 1193
{
	int flags = 0;

1194
	fl6->flowi6_iif = LOOPBACK_IFINDEX;
1195

1196
	if ((sk && sk->sk_bound_dev_if) || rt6_need_strict(&fl6->daddr))
1197
		flags |= RT6_LOOKUP_F_IFACE;
T
Thomas Graf 已提交
1198

1199
	if (!ipv6_addr_any(&fl6->saddr))
1200
		flags |= RT6_LOOKUP_F_HAS_SADDR;
1201 1202
	else if (sk)
		flags |= rt6_srcprefs2flags(inet6_sk(sk)->srcprefs);
1203

1204
	return fib6_rule_lookup(net, fl6, flags, ip6_pol_route_output);
L
Linus Torvalds 已提交
1205
}
1206
EXPORT_SYMBOL(ip6_route_output);
L
Linus Torvalds 已提交
1207

1208
struct dst_entry *ip6_blackhole_route(struct net *net, struct dst_entry *dst_orig)
1209
{
1210
	struct rt6_info *rt, *ort = (struct rt6_info *) dst_orig;
1211 1212
	struct dst_entry *new = NULL;

1213
	rt = dst_alloc(&ip6_dst_blackhole_ops, ort->dst.dev, 1, DST_OBSOLETE_NONE, 0);
1214
	if (rt) {
1215
		new = &rt->dst;
1216

1217 1218
		memset(new + 1, 0, sizeof(*rt) - sizeof(*new));

1219
		new->__use = 1;
1220
		new->input = dst_discard;
1221
		new->output = dst_discard_sk;
1222

E
Eric Dumazet 已提交
1223 1224 1225 1226
		if (dst_metrics_read_only(&ort->dst))
			new->_metrics = ort->dst._metrics;
		else
			dst_copy_metrics(new, &ort->dst);
1227 1228 1229 1230
		rt->rt6i_idev = ort->rt6i_idev;
		if (rt->rt6i_idev)
			in6_dev_hold(rt->rt6i_idev);

A
Alexey Dobriyan 已提交
1231
		rt->rt6i_gateway = ort->rt6i_gateway;
1232
		rt->rt6i_flags = ort->rt6i_flags;
1233 1234 1235 1236 1237 1238 1239 1240 1241 1242
		rt->rt6i_metric = 0;

		memcpy(&rt->rt6i_dst, &ort->rt6i_dst, sizeof(struct rt6key));
#ifdef CONFIG_IPV6_SUBTREES
		memcpy(&rt->rt6i_src, &ort->rt6i_src, sizeof(struct rt6key));
#endif

		dst_free(new);
	}

1243 1244
	dst_release(dst_orig);
	return new ? new : ERR_PTR(-ENOMEM);
1245 1246
}

L
Linus Torvalds 已提交
1247 1248 1249 1250
/*
 *	Destination cache support functions
 */

1251 1252 1253 1254 1255 1256 1257
static void rt6_dst_from_metrics_check(struct rt6_info *rt)
{
	if (rt->dst.from &&
	    dst_metrics_ptr(&rt->dst) != dst_metrics_ptr(rt->dst.from))
		dst_init_metrics(&rt->dst, dst_metrics_ptr(rt->dst.from), true);
}

1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277
static struct dst_entry *rt6_check(struct rt6_info *rt, u32 cookie)
{
	if (!rt->rt6i_node || (rt->rt6i_node->fn_sernum != cookie))
		return NULL;

	if (rt6_check_expired(rt))
		return NULL;

	return &rt->dst;
}

static struct dst_entry *rt6_dst_from_check(struct rt6_info *rt, u32 cookie)
{
	if (rt->dst.obsolete == DST_OBSOLETE_FORCE_CHK &&
	    rt6_check((struct rt6_info *)(rt->dst.from), cookie))
		return &rt->dst;
	else
		return NULL;
}

L
Linus Torvalds 已提交
1278 1279 1280 1281 1282 1283
static struct dst_entry *ip6_dst_check(struct dst_entry *dst, u32 cookie)
{
	struct rt6_info *rt;

	rt = (struct rt6_info *) dst;

1284 1285 1286 1287
	/* All IPV6 dsts are created with ->obsolete set to the value
	 * DST_OBSOLETE_FORCE_CHK which forces validation calls down
	 * into this function always.
	 */
1288

1289 1290
	rt6_dst_from_metrics_check(rt);

M
Martin KaFai Lau 已提交
1291
	if ((rt->rt6i_flags & RTF_PCPU) || unlikely(dst->flags & DST_NOCACHE))
1292 1293 1294
		return rt6_dst_from_check(rt, cookie);
	else
		return rt6_check(rt, cookie);
L
Linus Torvalds 已提交
1295 1296 1297 1298 1299 1300 1301
}

static struct dst_entry *ip6_negative_advice(struct dst_entry *dst)
{
	struct rt6_info *rt = (struct rt6_info *) dst;

	if (rt) {
1302 1303 1304 1305 1306 1307
		if (rt->rt6i_flags & RTF_CACHE) {
			if (rt6_check_expired(rt)) {
				ip6_del_rt(rt);
				dst = NULL;
			}
		} else {
L
Linus Torvalds 已提交
1308
			dst_release(dst);
1309 1310
			dst = NULL;
		}
L
Linus Torvalds 已提交
1311
	}
1312
	return dst;
L
Linus Torvalds 已提交
1313 1314 1315 1316 1317 1318
}

static void ip6_link_failure(struct sk_buff *skb)
{
	struct rt6_info *rt;

1319
	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_ADDR_UNREACH, 0);
L
Linus Torvalds 已提交
1320

E
Eric Dumazet 已提交
1321
	rt = (struct rt6_info *) skb_dst(skb);
L
Linus Torvalds 已提交
1322
	if (rt) {
1323 1324 1325 1326 1327
		if (rt->rt6i_flags & RTF_CACHE) {
			dst_hold(&rt->dst);
			if (ip6_del_rt(rt))
				dst_free(&rt->dst);
		} else if (rt->rt6i_node && (rt->rt6i_flags & RTF_DEFAULT)) {
L
Linus Torvalds 已提交
1328
			rt->rt6i_node->fn_sernum = -1;
1329
		}
L
Linus Torvalds 已提交
1330 1331 1332
	}
}

1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
static void rt6_do_update_pmtu(struct rt6_info *rt, u32 mtu)
{
	struct net *net = dev_net(rt->dst.dev);

	rt->rt6i_flags |= RTF_MODIFIED;
	rt->rt6i_pmtu = mtu;
	rt6_update_expires(rt, net->ipv6.sysctl.ip6_rt_mtu_expires);
}

static void __ip6_rt_update_pmtu(struct dst_entry *dst, const struct sock *sk,
				 const struct ipv6hdr *iph, u32 mtu)
L
Linus Torvalds 已提交
1344
{
1345
	struct rt6_info *rt6 = (struct rt6_info *)dst;
L
Linus Torvalds 已提交
1346

1347 1348
	if (rt6->rt6i_flags & RTF_LOCAL)
		return;
1349

1350 1351 1352 1353
	dst_confirm(dst);
	mtu = max_t(u32, mtu, IPV6_MIN_MTU);
	if (mtu >= dst_mtu(dst))
		return;
1354

1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380
	if (rt6->rt6i_flags & RTF_CACHE) {
		rt6_do_update_pmtu(rt6, mtu);
	} else {
		const struct in6_addr *daddr, *saddr;
		struct rt6_info *nrt6;

		if (iph) {
			daddr = &iph->daddr;
			saddr = &iph->saddr;
		} else if (sk) {
			daddr = &sk->sk_v6_daddr;
			saddr = &inet6_sk(sk)->saddr;
		} else {
			return;
		}
		nrt6 = ip6_rt_cache_alloc(rt6, daddr, saddr);
		if (nrt6) {
			rt6_do_update_pmtu(nrt6, mtu);

			/* ip6_ins_rt(nrt6) will bump the
			 * rt6->rt6i_node->fn_sernum
			 * which will fail the next rt6_check() and
			 * invalidate the sk->sk_dst_cache.
			 */
			ip6_ins_rt(nrt6);
		}
L
Linus Torvalds 已提交
1381 1382 1383
	}
}

1384 1385 1386 1387 1388 1389
static void ip6_rt_update_pmtu(struct dst_entry *dst, struct sock *sk,
			       struct sk_buff *skb, u32 mtu)
{
	__ip6_rt_update_pmtu(dst, sk, skb ? ipv6_hdr(skb) : NULL, mtu);
}

1390 1391
void ip6_update_pmtu(struct sk_buff *skb, struct net *net, __be32 mtu,
		     int oif, u32 mark)
1392 1393 1394 1395 1396 1397 1398
{
	const struct ipv6hdr *iph = (struct ipv6hdr *) skb->data;
	struct dst_entry *dst;
	struct flowi6 fl6;

	memset(&fl6, 0, sizeof(fl6));
	fl6.flowi6_oif = oif;
1399
	fl6.flowi6_mark = mark ? mark : IP6_REPLY_MARK(net, skb->mark);
1400 1401
	fl6.daddr = iph->daddr;
	fl6.saddr = iph->saddr;
1402
	fl6.flowlabel = ip6_flowinfo(iph);
1403 1404 1405

	dst = ip6_route_output(net, NULL, &fl6);
	if (!dst->error)
1406
		__ip6_rt_update_pmtu(dst, NULL, iph, ntohl(mtu));
1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417
	dst_release(dst);
}
EXPORT_SYMBOL_GPL(ip6_update_pmtu);

void ip6_sk_update_pmtu(struct sk_buff *skb, struct sock *sk, __be32 mtu)
{
	ip6_update_pmtu(skb, sock_net(sk), mtu,
			sk->sk_bound_dev_if, sk->sk_mark);
}
EXPORT_SYMBOL_GPL(ip6_sk_update_pmtu);

1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463
/* Handle redirects */
struct ip6rd_flowi {
	struct flowi6 fl6;
	struct in6_addr gateway;
};

static struct rt6_info *__ip6_route_redirect(struct net *net,
					     struct fib6_table *table,
					     struct flowi6 *fl6,
					     int flags)
{
	struct ip6rd_flowi *rdfl = (struct ip6rd_flowi *)fl6;
	struct rt6_info *rt;
	struct fib6_node *fn;

	/* Get the "current" route for this destination and
	 * check if the redirect has come from approriate router.
	 *
	 * RFC 4861 specifies that redirects should only be
	 * accepted if they come from the nexthop to the target.
	 * Due to the way the routes are chosen, this notion
	 * is a bit fuzzy and one might need to check all possible
	 * routes.
	 */

	read_lock_bh(&table->tb6_lock);
	fn = fib6_lookup(&table->tb6_root, &fl6->daddr, &fl6->saddr);
restart:
	for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) {
		if (rt6_check_expired(rt))
			continue;
		if (rt->dst.error)
			break;
		if (!(rt->rt6i_flags & RTF_GATEWAY))
			continue;
		if (fl6->flowi6_oif != rt->dst.dev->ifindex)
			continue;
		if (!ipv6_addr_equal(&rdfl->gateway, &rt->rt6i_gateway))
			continue;
		break;
	}

	if (!rt)
		rt = net->ipv6.ip6_null_entry;
	else if (rt->dst.error) {
		rt = net->ipv6.ip6_null_entry;
1464 1465 1466 1467
		goto out;
	}

	if (rt == net->ipv6.ip6_null_entry) {
M
Martin KaFai Lau 已提交
1468 1469 1470
		fn = fib6_backtrack(fn, &fl6->saddr);
		if (fn)
			goto restart;
1471
	}
M
Martin KaFai Lau 已提交
1472

1473
out:
1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494
	dst_hold(&rt->dst);

	read_unlock_bh(&table->tb6_lock);

	return rt;
};

static struct dst_entry *ip6_route_redirect(struct net *net,
					const struct flowi6 *fl6,
					const struct in6_addr *gateway)
{
	int flags = RT6_LOOKUP_F_HAS_SADDR;
	struct ip6rd_flowi rdfl;

	rdfl.fl6 = *fl6;
	rdfl.gateway = *gateway;

	return fib6_rule_lookup(net, &rdfl.fl6,
				flags, __ip6_route_redirect);
}

1495 1496 1497 1498 1499 1500 1501
void ip6_redirect(struct sk_buff *skb, struct net *net, int oif, u32 mark)
{
	const struct ipv6hdr *iph = (struct ipv6hdr *) skb->data;
	struct dst_entry *dst;
	struct flowi6 fl6;

	memset(&fl6, 0, sizeof(fl6));
1502
	fl6.flowi6_iif = LOOPBACK_IFINDEX;
1503 1504 1505 1506
	fl6.flowi6_oif = oif;
	fl6.flowi6_mark = mark;
	fl6.daddr = iph->daddr;
	fl6.saddr = iph->saddr;
1507
	fl6.flowlabel = ip6_flowinfo(iph);
1508

1509 1510
	dst = ip6_route_redirect(net, &fl6, &ipv6_hdr(skb)->saddr);
	rt6_do_redirect(dst, NULL, skb);
1511 1512 1513 1514
	dst_release(dst);
}
EXPORT_SYMBOL_GPL(ip6_redirect);

1515 1516 1517 1518 1519 1520 1521 1522 1523
void ip6_redirect_no_header(struct sk_buff *skb, struct net *net, int oif,
			    u32 mark)
{
	const struct ipv6hdr *iph = ipv6_hdr(skb);
	const struct rd_msg *msg = (struct rd_msg *)icmp6_hdr(skb);
	struct dst_entry *dst;
	struct flowi6 fl6;

	memset(&fl6, 0, sizeof(fl6));
1524
	fl6.flowi6_iif = LOOPBACK_IFINDEX;
1525 1526 1527 1528 1529
	fl6.flowi6_oif = oif;
	fl6.flowi6_mark = mark;
	fl6.daddr = msg->dest;
	fl6.saddr = iph->daddr;

1530 1531
	dst = ip6_route_redirect(net, &fl6, &iph->saddr);
	rt6_do_redirect(dst, NULL, skb);
1532 1533 1534
	dst_release(dst);
}

1535 1536 1537 1538 1539 1540
void ip6_sk_redirect(struct sk_buff *skb, struct sock *sk)
{
	ip6_redirect(skb, sock_net(sk), sk->sk_bound_dev_if, sk->sk_mark);
}
EXPORT_SYMBOL_GPL(ip6_sk_redirect);

1541
static unsigned int ip6_default_advmss(const struct dst_entry *dst)
L
Linus Torvalds 已提交
1542
{
1543 1544 1545 1546
	struct net_device *dev = dst->dev;
	unsigned int mtu = dst_mtu(dst);
	struct net *net = dev_net(dev);

L
Linus Torvalds 已提交
1547 1548
	mtu -= sizeof(struct ipv6hdr) + sizeof(struct tcphdr);

1549 1550
	if (mtu < net->ipv6.sysctl.ip6_rt_min_advmss)
		mtu = net->ipv6.sysctl.ip6_rt_min_advmss;
L
Linus Torvalds 已提交
1551 1552

	/*
1553 1554 1555
	 * Maximal non-jumbo IPv6 payload is IPV6_MAXPLEN and
	 * corresponding MSS is IPV6_MAXPLEN - tcp_header_size.
	 * IPV6_MAXPLEN is also valid and means: "any MSS,
L
Linus Torvalds 已提交
1556 1557 1558 1559 1560 1561 1562
	 * rely only on pmtu discovery"
	 */
	if (mtu > IPV6_MAXPLEN - sizeof(struct tcphdr))
		mtu = IPV6_MAXPLEN;
	return mtu;
}

1563
static unsigned int ip6_mtu(const struct dst_entry *dst)
1564
{
1565 1566
	const struct rt6_info *rt = (const struct rt6_info *)dst;
	unsigned int mtu = rt->rt6i_pmtu;
1567
	struct inet6_dev *idev;
1568

1569 1570 1571 1572
	if (mtu)
		goto out;

	mtu = dst_metric_raw(dst, RTAX_MTU);
1573
	if (mtu)
E
Eric Dumazet 已提交
1574
		goto out;
1575 1576

	mtu = IPV6_MIN_MTU;
1577 1578 1579 1580 1581 1582 1583

	rcu_read_lock();
	idev = __in6_dev_get(dst->dev);
	if (idev)
		mtu = idev->cnf.mtu6;
	rcu_read_unlock();

E
Eric Dumazet 已提交
1584 1585
out:
	return min_t(unsigned int, mtu, IP6_MAX_MTU);
1586 1587
}

1588 1589
static struct dst_entry *icmp6_dst_gc_list;
static DEFINE_SPINLOCK(icmp6_dst_lock);
1590

1591
struct dst_entry *icmp6_dst_alloc(struct net_device *dev,
1592
				  struct flowi6 *fl6)
L
Linus Torvalds 已提交
1593
{
1594
	struct dst_entry *dst;
L
Linus Torvalds 已提交
1595 1596
	struct rt6_info *rt;
	struct inet6_dev *idev = in6_dev_get(dev);
1597
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
1598

1599
	if (unlikely(!idev))
E
Eric Dumazet 已提交
1600
		return ERR_PTR(-ENODEV);
L
Linus Torvalds 已提交
1601

1602
	rt = ip6_dst_alloc(net, dev, 0);
1603
	if (unlikely(!rt)) {
L
Linus Torvalds 已提交
1604
		in6_dev_put(idev);
1605
		dst = ERR_PTR(-ENOMEM);
L
Linus Torvalds 已提交
1606 1607 1608
		goto out;
	}

1609 1610
	rt->dst.flags |= DST_HOST;
	rt->dst.output  = ip6_output;
1611
	atomic_set(&rt->dst.__refcnt, 1);
1612
	rt->rt6i_gateway  = fl6->daddr;
1613
	rt->rt6i_dst.addr = fl6->daddr;
1614 1615
	rt->rt6i_dst.plen = 128;
	rt->rt6i_idev     = idev;
L
Li RongQing 已提交
1616
	dst_metric_set(&rt->dst, RTAX_HOPLIMIT, 0);
L
Linus Torvalds 已提交
1617

1618
	spin_lock_bh(&icmp6_dst_lock);
1619 1620
	rt->dst.next = icmp6_dst_gc_list;
	icmp6_dst_gc_list = &rt->dst;
1621
	spin_unlock_bh(&icmp6_dst_lock);
L
Linus Torvalds 已提交
1622

1623
	fib6_force_start_gc(net);
L
Linus Torvalds 已提交
1624

1625 1626
	dst = xfrm_lookup(net, &rt->dst, flowi6_to_flowi(fl6), NULL, 0);

L
Linus Torvalds 已提交
1627
out:
1628
	return dst;
L
Linus Torvalds 已提交
1629 1630
}

1631
int icmp6_dst_gc(void)
L
Linus Torvalds 已提交
1632
{
1633
	struct dst_entry *dst, **pprev;
1634
	int more = 0;
L
Linus Torvalds 已提交
1635

1636 1637
	spin_lock_bh(&icmp6_dst_lock);
	pprev = &icmp6_dst_gc_list;
1638

L
Linus Torvalds 已提交
1639 1640 1641 1642 1643 1644
	while ((dst = *pprev) != NULL) {
		if (!atomic_read(&dst->__refcnt)) {
			*pprev = dst->next;
			dst_free(dst);
		} else {
			pprev = &dst->next;
1645
			++more;
L
Linus Torvalds 已提交
1646 1647 1648
		}
	}

1649
	spin_unlock_bh(&icmp6_dst_lock);
1650

1651
	return more;
L
Linus Torvalds 已提交
1652 1653
}

1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672
static void icmp6_clean_all(int (*func)(struct rt6_info *rt, void *arg),
			    void *arg)
{
	struct dst_entry *dst, **pprev;

	spin_lock_bh(&icmp6_dst_lock);
	pprev = &icmp6_dst_gc_list;
	while ((dst = *pprev) != NULL) {
		struct rt6_info *rt = (struct rt6_info *) dst;
		if (func(rt, arg)) {
			*pprev = dst->next;
			dst_free(dst);
		} else {
			pprev = &dst->next;
		}
	}
	spin_unlock_bh(&icmp6_dst_lock);
}

1673
static int ip6_dst_gc(struct dst_ops *ops)
L
Linus Torvalds 已提交
1674
{
1675
	struct net *net = container_of(ops, struct net, ipv6.ip6_dst_ops);
1676 1677 1678 1679 1680
	int rt_min_interval = net->ipv6.sysctl.ip6_rt_gc_min_interval;
	int rt_max_size = net->ipv6.sysctl.ip6_rt_max_size;
	int rt_elasticity = net->ipv6.sysctl.ip6_rt_gc_elasticity;
	int rt_gc_timeout = net->ipv6.sysctl.ip6_rt_gc_timeout;
	unsigned long rt_last_gc = net->ipv6.ip6_rt_last_gc;
1681
	int entries;
1682

1683
	entries = dst_entries_get_fast(ops);
1684
	if (time_after(rt_last_gc + rt_min_interval, jiffies) &&
1685
	    entries <= rt_max_size)
L
Linus Torvalds 已提交
1686 1687
		goto out;

1688
	net->ipv6.ip6_rt_gc_expire++;
1689
	fib6_run_gc(net->ipv6.ip6_rt_gc_expire, net, true);
1690 1691
	entries = dst_entries_get_slow(ops);
	if (entries < ops->gc_thresh)
1692
		net->ipv6.ip6_rt_gc_expire = rt_gc_timeout>>1;
L
Linus Torvalds 已提交
1693
out:
1694
	net->ipv6.ip6_rt_gc_expire -= net->ipv6.ip6_rt_gc_expire>>rt_elasticity;
1695
	return entries > rt_max_size;
L
Linus Torvalds 已提交
1696 1697
}

1698 1699 1700 1701 1702 1703 1704
static int ip6_convert_metrics(struct mx6_config *mxc,
			       const struct fib6_config *cfg)
{
	struct nlattr *nla;
	int remaining;
	u32 *mp;

1705
	if (!cfg->fc_mx)
1706 1707 1708 1709 1710 1711 1712 1713 1714 1715
		return 0;

	mp = kzalloc(sizeof(u32) * RTAX_MAX, GFP_KERNEL);
	if (unlikely(!mp))
		return -ENOMEM;

	nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) {
		int type = nla_type(nla);

		if (type) {
1716 1717
			u32 val;

1718 1719
			if (unlikely(type > RTAX_MAX))
				goto err;
1720 1721 1722 1723 1724 1725 1726 1727 1728 1729
			if (type == RTAX_CC_ALGO) {
				char tmp[TCP_CA_NAME_MAX];

				nla_strlcpy(tmp, nla, sizeof(tmp));
				val = tcp_ca_get_key_by_name(tmp);
				if (val == TCP_CA_UNSPEC)
					goto err;
			} else {
				val = nla_get_u32(nla);
			}
1730

1731
			mp[type - 1] = val;
1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742
			__set_bit(type - 1, mxc->mx_valid);
		}
	}

	mxc->mx = mp;

	return 0;
 err:
	kfree(mp);
	return -EINVAL;
}
L
Linus Torvalds 已提交
1743

1744
int ip6_route_add(struct fib6_config *cfg)
L
Linus Torvalds 已提交
1745 1746
{
	int err;
1747
	struct net *net = cfg->fc_nlinfo.nl_net;
L
Linus Torvalds 已提交
1748 1749 1750
	struct rt6_info *rt = NULL;
	struct net_device *dev = NULL;
	struct inet6_dev *idev = NULL;
T
Thomas Graf 已提交
1751
	struct fib6_table *table;
1752
	struct mx6_config mxc = { .mx = NULL, };
L
Linus Torvalds 已提交
1753 1754
	int addr_type;

1755
	if (cfg->fc_dst_len > 128 || cfg->fc_src_len > 128)
L
Linus Torvalds 已提交
1756 1757
		return -EINVAL;
#ifndef CONFIG_IPV6_SUBTREES
1758
	if (cfg->fc_src_len)
L
Linus Torvalds 已提交
1759 1760
		return -EINVAL;
#endif
1761
	if (cfg->fc_ifindex) {
L
Linus Torvalds 已提交
1762
		err = -ENODEV;
1763
		dev = dev_get_by_index(net, cfg->fc_ifindex);
L
Linus Torvalds 已提交
1764 1765 1766 1767 1768 1769 1770
		if (!dev)
			goto out;
		idev = in6_dev_get(dev);
		if (!idev)
			goto out;
	}

1771 1772
	if (cfg->fc_metric == 0)
		cfg->fc_metric = IP6_RT_PRIO_USER;
L
Linus Torvalds 已提交
1773

1774
	err = -ENOBUFS;
1775 1776
	if (cfg->fc_nlinfo.nlh &&
	    !(cfg->fc_nlinfo.nlh->nlmsg_flags & NLM_F_CREATE)) {
1777
		table = fib6_get_table(net, cfg->fc_table);
1778
		if (!table) {
1779
			pr_warn("NLM_F_CREATE should be specified when creating new route\n");
1780 1781 1782 1783 1784
			table = fib6_new_table(net, cfg->fc_table);
		}
	} else {
		table = fib6_new_table(net, cfg->fc_table);
	}
1785 1786

	if (!table)
T
Thomas Graf 已提交
1787 1788
		goto out;

1789 1790
	rt = ip6_dst_alloc(net, NULL,
			   (cfg->fc_flags & RTF_ADDRCONF) ? 0 : DST_NOCOUNT);
L
Linus Torvalds 已提交
1791

1792
	if (!rt) {
L
Linus Torvalds 已提交
1793 1794 1795 1796
		err = -ENOMEM;
		goto out;
	}

1797 1798 1799 1800 1801
	if (cfg->fc_flags & RTF_EXPIRES)
		rt6_set_expires(rt, jiffies +
				clock_t_to_jiffies(cfg->fc_expires));
	else
		rt6_clean_expires(rt);
L
Linus Torvalds 已提交
1802

1803 1804 1805 1806 1807
	if (cfg->fc_protocol == RTPROT_UNSPEC)
		cfg->fc_protocol = RTPROT_BOOT;
	rt->rt6i_protocol = cfg->fc_protocol;

	addr_type = ipv6_addr_type(&cfg->fc_dst);
L
Linus Torvalds 已提交
1808 1809

	if (addr_type & IPV6_ADDR_MULTICAST)
1810
		rt->dst.input = ip6_mc_input;
1811 1812
	else if (cfg->fc_flags & RTF_LOCAL)
		rt->dst.input = ip6_input;
L
Linus Torvalds 已提交
1813
	else
1814
		rt->dst.input = ip6_forward;
L
Linus Torvalds 已提交
1815

1816
	rt->dst.output = ip6_output;
L
Linus Torvalds 已提交
1817

1818 1819 1820 1821
	if (cfg->fc_encap) {
		struct lwtunnel_state *lwtstate;

		err = lwtunnel_build_state(dev, cfg->fc_encap_type,
1822 1823
					   cfg->fc_encap, AF_INET6, cfg,
					   &lwtstate);
1824 1825
		if (err)
			goto out;
1826 1827 1828 1829
		rt->dst.lwtstate = lwtstate_get(lwtstate);
		if (lwtunnel_output_redirect(rt->dst.lwtstate)) {
			rt->dst.lwtstate->orig_output = rt->dst.output;
			rt->dst.output = lwtunnel_output;
1830
		}
1831 1832 1833
		if (lwtunnel_input_redirect(rt->dst.lwtstate)) {
			rt->dst.lwtstate->orig_input = rt->dst.input;
			rt->dst.input = lwtunnel_input;
1834
		}
1835 1836
	}

1837 1838
	ipv6_addr_prefix(&rt->rt6i_dst.addr, &cfg->fc_dst, cfg->fc_dst_len);
	rt->rt6i_dst.plen = cfg->fc_dst_len;
1839
	if (rt->rt6i_dst.plen == 128)
1840 1841
		rt->dst.flags |= DST_HOST;

L
Linus Torvalds 已提交
1842
#ifdef CONFIG_IPV6_SUBTREES
1843 1844
	ipv6_addr_prefix(&rt->rt6i_src.addr, &cfg->fc_src, cfg->fc_src_len);
	rt->rt6i_src.plen = cfg->fc_src_len;
L
Linus Torvalds 已提交
1845 1846
#endif

1847
	rt->rt6i_metric = cfg->fc_metric;
L
Linus Torvalds 已提交
1848 1849 1850 1851

	/* We cannot add true routes via loopback here,
	   they would result in kernel looping; promote them to reject routes
	 */
1852
	if ((cfg->fc_flags & RTF_REJECT) ||
1853 1854 1855
	    (dev && (dev->flags & IFF_LOOPBACK) &&
	     !(addr_type & IPV6_ADDR_LOOPBACK) &&
	     !(cfg->fc_flags & RTF_LOCAL))) {
L
Linus Torvalds 已提交
1856
		/* hold loopback dev/idev if we haven't done so. */
1857
		if (dev != net->loopback_dev) {
L
Linus Torvalds 已提交
1858 1859 1860 1861
			if (dev) {
				dev_put(dev);
				in6_dev_put(idev);
			}
1862
			dev = net->loopback_dev;
L
Linus Torvalds 已提交
1863 1864 1865 1866 1867 1868 1869 1870
			dev_hold(dev);
			idev = in6_dev_get(dev);
			if (!idev) {
				err = -ENODEV;
				goto out;
			}
		}
		rt->rt6i_flags = RTF_REJECT|RTF_NONEXTHOP;
1871 1872 1873
		switch (cfg->fc_type) {
		case RTN_BLACKHOLE:
			rt->dst.error = -EINVAL;
1874
			rt->dst.output = dst_discard_sk;
1875
			rt->dst.input = dst_discard;
1876 1877 1878
			break;
		case RTN_PROHIBIT:
			rt->dst.error = -EACCES;
1879 1880
			rt->dst.output = ip6_pkt_prohibit_out;
			rt->dst.input = ip6_pkt_prohibit;
1881
			break;
1882
		case RTN_THROW:
1883
		default:
1884 1885 1886 1887
			rt->dst.error = (cfg->fc_type == RTN_THROW) ? -EAGAIN
					: -ENETUNREACH;
			rt->dst.output = ip6_pkt_discard_out;
			rt->dst.input = ip6_pkt_discard;
1888 1889
			break;
		}
L
Linus Torvalds 已提交
1890 1891 1892
		goto install_route;
	}

1893
	if (cfg->fc_flags & RTF_GATEWAY) {
1894
		const struct in6_addr *gw_addr;
L
Linus Torvalds 已提交
1895 1896
		int gwa_type;

1897
		gw_addr = &cfg->fc_gateway;
1898
		gwa_type = ipv6_addr_type(gw_addr);
1899 1900 1901 1902 1903 1904 1905

		/* if gw_addr is local we will fail to detect this in case
		 * address is still TENTATIVE (DAD in progress). rt6_lookup()
		 * will return already-added prefix route via interface that
		 * prefix route was assigned to, which might be non-loopback.
		 */
		err = -EINVAL;
1906 1907 1908
		if (ipv6_chk_addr_and_flags(net, gw_addr,
					    gwa_type & IPV6_ADDR_LINKLOCAL ?
					    dev : NULL, 0, 0))
1909 1910
			goto out;

A
Alexey Dobriyan 已提交
1911
		rt->rt6i_gateway = *gw_addr;
L
Linus Torvalds 已提交
1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922

		if (gwa_type != (IPV6_ADDR_LINKLOCAL|IPV6_ADDR_UNICAST)) {
			struct rt6_info *grt;

			/* IPv6 strictly inhibits using not link-local
			   addresses as nexthop address.
			   Otherwise, router will not able to send redirects.
			   It is very good, but in some (rare!) circumstances
			   (SIT, PtP, NBMA NOARP links) it is handy to allow
			   some exceptions. --ANK
			 */
1923
			if (!(gwa_type & IPV6_ADDR_UNICAST))
L
Linus Torvalds 已提交
1924 1925
				goto out;

1926
			grt = rt6_lookup(net, gw_addr, NULL, cfg->fc_ifindex, 1);
L
Linus Torvalds 已提交
1927 1928

			err = -EHOSTUNREACH;
1929
			if (!grt)
L
Linus Torvalds 已提交
1930 1931
				goto out;
			if (dev) {
1932
				if (dev != grt->dst.dev) {
A
Amerigo Wang 已提交
1933
					ip6_rt_put(grt);
L
Linus Torvalds 已提交
1934 1935 1936
					goto out;
				}
			} else {
1937
				dev = grt->dst.dev;
L
Linus Torvalds 已提交
1938 1939 1940 1941
				idev = grt->rt6i_idev;
				dev_hold(dev);
				in6_dev_hold(grt->rt6i_idev);
			}
1942
			if (!(grt->rt6i_flags & RTF_GATEWAY))
L
Linus Torvalds 已提交
1943
				err = 0;
A
Amerigo Wang 已提交
1944
			ip6_rt_put(grt);
L
Linus Torvalds 已提交
1945 1946 1947 1948 1949

			if (err)
				goto out;
		}
		err = -EINVAL;
1950
		if (!dev || (dev->flags & IFF_LOOPBACK))
L
Linus Torvalds 已提交
1951 1952 1953 1954
			goto out;
	}

	err = -ENODEV;
1955
	if (!dev)
L
Linus Torvalds 已提交
1956 1957
		goto out;

1958 1959 1960 1961 1962
	if (!ipv6_addr_any(&cfg->fc_prefsrc)) {
		if (!ipv6_chk_addr(net, &cfg->fc_prefsrc, dev, 0)) {
			err = -EINVAL;
			goto out;
		}
A
Alexey Dobriyan 已提交
1963
		rt->rt6i_prefsrc.addr = cfg->fc_prefsrc;
1964 1965 1966 1967
		rt->rt6i_prefsrc.plen = 128;
	} else
		rt->rt6i_prefsrc.plen = 0;

1968
	rt->rt6i_flags = cfg->fc_flags;
L
Linus Torvalds 已提交
1969 1970

install_route:
1971
	rt->dst.dev = dev;
L
Linus Torvalds 已提交
1972
	rt->rt6i_idev = idev;
T
Thomas Graf 已提交
1973
	rt->rt6i_table = table;
1974

1975
	cfg->fc_nlinfo.nl_net = dev_net(dev);
1976

1977 1978 1979
	err = ip6_convert_metrics(&mxc, cfg);
	if (err)
		goto out;
L
Linus Torvalds 已提交
1980

1981 1982 1983 1984
	err = __ip6_ins_rt(rt, &cfg->fc_nlinfo, &mxc);

	kfree(mxc.mx);
	return err;
L
Linus Torvalds 已提交
1985 1986 1987 1988 1989 1990
out:
	if (dev)
		dev_put(dev);
	if (idev)
		in6_dev_put(idev);
	if (rt)
1991
		dst_free(&rt->dst);
L
Linus Torvalds 已提交
1992 1993 1994
	return err;
}

1995
static int __ip6_del_rt(struct rt6_info *rt, struct nl_info *info)
L
Linus Torvalds 已提交
1996 1997
{
	int err;
T
Thomas Graf 已提交
1998
	struct fib6_table *table;
1999
	struct net *net = dev_net(rt->dst.dev);
L
Linus Torvalds 已提交
2000

2001 2002 2003 2004
	if (rt == net->ipv6.ip6_null_entry) {
		err = -ENOENT;
		goto out;
	}
2005

T
Thomas Graf 已提交
2006 2007
	table = rt->rt6i_table;
	write_lock_bh(&table->tb6_lock);
2008
	err = fib6_del(rt, info);
T
Thomas Graf 已提交
2009
	write_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
2010

2011
out:
A
Amerigo Wang 已提交
2012
	ip6_rt_put(rt);
L
Linus Torvalds 已提交
2013 2014 2015
	return err;
}

2016 2017
int ip6_del_rt(struct rt6_info *rt)
{
2018
	struct nl_info info = {
2019
		.nl_net = dev_net(rt->dst.dev),
2020
	};
2021
	return __ip6_del_rt(rt, &info);
2022 2023
}

2024
static int ip6_route_del(struct fib6_config *cfg)
L
Linus Torvalds 已提交
2025
{
T
Thomas Graf 已提交
2026
	struct fib6_table *table;
L
Linus Torvalds 已提交
2027 2028 2029 2030
	struct fib6_node *fn;
	struct rt6_info *rt;
	int err = -ESRCH;

2031
	table = fib6_get_table(cfg->fc_nlinfo.nl_net, cfg->fc_table);
2032
	if (!table)
T
Thomas Graf 已提交
2033 2034 2035
		return err;

	read_lock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
2036

T
Thomas Graf 已提交
2037
	fn = fib6_locate(&table->tb6_root,
2038 2039
			 &cfg->fc_dst, cfg->fc_dst_len,
			 &cfg->fc_src, cfg->fc_src_len);
2040

L
Linus Torvalds 已提交
2041
	if (fn) {
2042
		for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) {
2043 2044 2045
			if ((rt->rt6i_flags & RTF_CACHE) &&
			    !(cfg->fc_flags & RTF_CACHE))
				continue;
2046
			if (cfg->fc_ifindex &&
2047 2048
			    (!rt->dst.dev ||
			     rt->dst.dev->ifindex != cfg->fc_ifindex))
L
Linus Torvalds 已提交
2049
				continue;
2050 2051
			if (cfg->fc_flags & RTF_GATEWAY &&
			    !ipv6_addr_equal(&cfg->fc_gateway, &rt->rt6i_gateway))
L
Linus Torvalds 已提交
2052
				continue;
2053
			if (cfg->fc_metric && cfg->fc_metric != rt->rt6i_metric)
L
Linus Torvalds 已提交
2054
				continue;
2055
			dst_hold(&rt->dst);
T
Thomas Graf 已提交
2056
			read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
2057

2058
			return __ip6_del_rt(rt, &cfg->fc_nlinfo);
L
Linus Torvalds 已提交
2059 2060
		}
	}
T
Thomas Graf 已提交
2061
	read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
2062 2063 2064 2065

	return err;
}

2066
static void rt6_do_redirect(struct dst_entry *dst, struct sock *sk, struct sk_buff *skb)
2067
{
2068
	struct net *net = dev_net(skb->dev);
2069
	struct netevent_redirect netevent;
2070 2071 2072 2073
	struct rt6_info *rt, *nrt = NULL;
	struct ndisc_options ndopts;
	struct inet6_dev *in6_dev;
	struct neighbour *neigh;
2074
	struct rd_msg *msg;
2075 2076
	int optlen, on_link;
	u8 *lladdr;
2077

2078
	optlen = skb_tail_pointer(skb) - skb_transport_header(skb);
2079
	optlen -= sizeof(*msg);
2080 2081

	if (optlen < 0) {
2082
		net_dbg_ratelimited("rt6_do_redirect: packet too short\n");
2083 2084 2085
		return;
	}

2086
	msg = (struct rd_msg *)icmp6_hdr(skb);
2087

2088
	if (ipv6_addr_is_multicast(&msg->dest)) {
2089
		net_dbg_ratelimited("rt6_do_redirect: destination address is multicast\n");
2090 2091 2092
		return;
	}

2093
	on_link = 0;
2094
	if (ipv6_addr_equal(&msg->dest, &msg->target)) {
2095
		on_link = 1;
2096
	} else if (ipv6_addr_type(&msg->target) !=
2097
		   (IPV6_ADDR_UNICAST|IPV6_ADDR_LINKLOCAL)) {
2098
		net_dbg_ratelimited("rt6_do_redirect: target address is not link-local unicast\n");
2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112
		return;
	}

	in6_dev = __in6_dev_get(skb->dev);
	if (!in6_dev)
		return;
	if (in6_dev->cnf.forwarding || !in6_dev->cnf.accept_redirects)
		return;

	/* RFC2461 8.1:
	 *	The IP source address of the Redirect MUST be the same as the current
	 *	first-hop router for the specified ICMP Destination Address.
	 */

2113
	if (!ndisc_parse_options(msg->opt, optlen, &ndopts)) {
2114 2115 2116
		net_dbg_ratelimited("rt6_redirect: invalid ND options\n");
		return;
	}
2117 2118

	lladdr = NULL;
2119 2120 2121 2122 2123 2124 2125 2126 2127
	if (ndopts.nd_opts_tgt_lladdr) {
		lladdr = ndisc_opt_addr_data(ndopts.nd_opts_tgt_lladdr,
					     skb->dev);
		if (!lladdr) {
			net_dbg_ratelimited("rt6_redirect: invalid link-layer address length\n");
			return;
		}
	}

2128 2129 2130
	rt = (struct rt6_info *) dst;
	if (rt == net->ipv6.ip6_null_entry) {
		net_dbg_ratelimited("rt6_redirect: source isn't a valid nexthop for redirect target\n");
2131
		return;
2132
	}
2133

2134 2135 2136 2137 2138
	/* Redirect received -> path was valid.
	 * Look, redirects are sent only in response to data packets,
	 * so that this nexthop apparently is reachable. --ANK
	 */
	dst_confirm(&rt->dst);
2139

2140
	neigh = __neigh_lookup(&nd_tbl, &msg->target, skb->dev, 1);
2141 2142
	if (!neigh)
		return;
2143

L
Linus Torvalds 已提交
2144 2145 2146 2147
	/*
	 *	We have finally decided to accept it.
	 */

2148
	neigh_update(neigh, lladdr, NUD_STALE,
L
Linus Torvalds 已提交
2149 2150 2151 2152 2153 2154
		     NEIGH_UPDATE_F_WEAK_OVERRIDE|
		     NEIGH_UPDATE_F_OVERRIDE|
		     (on_link ? 0 : (NEIGH_UPDATE_F_OVERRIDE_ISROUTER|
				     NEIGH_UPDATE_F_ISROUTER))
		     );

M
Martin KaFai Lau 已提交
2155
	nrt = ip6_rt_cache_alloc(rt, &msg->dest, NULL);
2156
	if (!nrt)
L
Linus Torvalds 已提交
2157 2158 2159 2160 2161 2162
		goto out;

	nrt->rt6i_flags = RTF_GATEWAY|RTF_UP|RTF_DYNAMIC|RTF_CACHE;
	if (on_link)
		nrt->rt6i_flags &= ~RTF_GATEWAY;

A
Alexey Dobriyan 已提交
2163
	nrt->rt6i_gateway = *(struct in6_addr *)neigh->primary_key;
L
Linus Torvalds 已提交
2164

2165
	if (ip6_ins_rt(nrt))
L
Linus Torvalds 已提交
2166 2167
		goto out;

2168 2169
	netevent.old = &rt->dst;
	netevent.new = &nrt->dst;
2170
	netevent.daddr = &msg->dest;
2171
	netevent.neigh = neigh;
2172 2173
	call_netevent_notifiers(NETEVENT_REDIRECT, &netevent);

2174
	if (rt->rt6i_flags & RTF_CACHE) {
2175
		rt = (struct rt6_info *) dst_clone(&rt->dst);
2176
		ip6_del_rt(rt);
L
Linus Torvalds 已提交
2177 2178 2179
	}

out:
2180
	neigh_release(neigh);
2181 2182
}

L
Linus Torvalds 已提交
2183 2184 2185 2186
/*
 *	Misc support functions
 */

2187 2188 2189 2190 2191 2192 2193 2194 2195 2196
static void rt6_set_from(struct rt6_info *rt, struct rt6_info *from)
{
	BUG_ON(from->dst.from);

	rt->rt6i_flags &= ~RTF_EXPIRES;
	dst_hold(&from->dst);
	rt->dst.from = &from->dst;
	dst_init_metrics(&rt->dst, dst_metrics_ptr(&from->dst), true);
}

M
Martin KaFai Lau 已提交
2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210
static void ip6_rt_copy_init(struct rt6_info *rt, struct rt6_info *ort)
{
	rt->dst.input = ort->dst.input;
	rt->dst.output = ort->dst.output;
	rt->rt6i_dst = ort->rt6i_dst;
	rt->dst.error = ort->dst.error;
	rt->rt6i_idev = ort->rt6i_idev;
	if (rt->rt6i_idev)
		in6_dev_hold(rt->rt6i_idev);
	rt->dst.lastuse = jiffies;
	rt->rt6i_gateway = ort->rt6i_gateway;
	rt->rt6i_flags = ort->rt6i_flags;
	rt6_set_from(rt, ort);
	rt->rt6i_metric = ort->rt6i_metric;
L
Linus Torvalds 已提交
2211
#ifdef CONFIG_IPV6_SUBTREES
M
Martin KaFai Lau 已提交
2212
	rt->rt6i_src = ort->rt6i_src;
L
Linus Torvalds 已提交
2213
#endif
M
Martin KaFai Lau 已提交
2214 2215
	rt->rt6i_prefsrc = ort->rt6i_prefsrc;
	rt->rt6i_table = ort->rt6i_table;
2216
	rt->dst.lwtstate = lwtstate_get(ort->dst.lwtstate);
L
Linus Torvalds 已提交
2217 2218
}

2219
#ifdef CONFIG_IPV6_ROUTE_INFO
2220
static struct rt6_info *rt6_get_route_info(struct net *net,
2221 2222
					   const struct in6_addr *prefix, int prefixlen,
					   const struct in6_addr *gwaddr, int ifindex)
2223 2224 2225
{
	struct fib6_node *fn;
	struct rt6_info *rt = NULL;
T
Thomas Graf 已提交
2226 2227
	struct fib6_table *table;

2228
	table = fib6_get_table(net, RT6_TABLE_INFO);
2229
	if (!table)
T
Thomas Graf 已提交
2230
		return NULL;
2231

2232
	read_lock_bh(&table->tb6_lock);
2233
	fn = fib6_locate(&table->tb6_root, prefix, prefixlen, NULL, 0);
2234 2235 2236
	if (!fn)
		goto out;

2237
	for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) {
2238
		if (rt->dst.dev->ifindex != ifindex)
2239 2240 2241 2242 2243
			continue;
		if ((rt->rt6i_flags & (RTF_ROUTEINFO|RTF_GATEWAY)) != (RTF_ROUTEINFO|RTF_GATEWAY))
			continue;
		if (!ipv6_addr_equal(&rt->rt6i_gateway, gwaddr))
			continue;
2244
		dst_hold(&rt->dst);
2245 2246 2247
		break;
	}
out:
2248
	read_unlock_bh(&table->tb6_lock);
2249 2250 2251
	return rt;
}

2252
static struct rt6_info *rt6_add_route_info(struct net *net,
2253 2254
					   const struct in6_addr *prefix, int prefixlen,
					   const struct in6_addr *gwaddr, int ifindex,
2255
					   unsigned int pref)
2256
{
2257 2258
	struct fib6_config cfg = {
		.fc_table	= RT6_TABLE_INFO,
2259
		.fc_metric	= IP6_RT_PRIO_USER,
2260 2261 2262 2263
		.fc_ifindex	= ifindex,
		.fc_dst_len	= prefixlen,
		.fc_flags	= RTF_GATEWAY | RTF_ADDRCONF | RTF_ROUTEINFO |
				  RTF_UP | RTF_PREF(pref),
2264
		.fc_nlinfo.portid = 0,
2265 2266
		.fc_nlinfo.nlh = NULL,
		.fc_nlinfo.nl_net = net,
2267 2268
	};

A
Alexey Dobriyan 已提交
2269 2270
	cfg.fc_dst = *prefix;
	cfg.fc_gateway = *gwaddr;
2271

2272 2273
	/* We should treat it as a default route if prefix length is 0. */
	if (!prefixlen)
2274
		cfg.fc_flags |= RTF_DEFAULT;
2275

2276
	ip6_route_add(&cfg);
2277

2278
	return rt6_get_route_info(net, prefix, prefixlen, gwaddr, ifindex);
2279 2280 2281
}
#endif

2282
struct rt6_info *rt6_get_dflt_router(const struct in6_addr *addr, struct net_device *dev)
2283
{
L
Linus Torvalds 已提交
2284
	struct rt6_info *rt;
T
Thomas Graf 已提交
2285
	struct fib6_table *table;
L
Linus Torvalds 已提交
2286

2287
	table = fib6_get_table(dev_net(dev), RT6_TABLE_DFLT);
2288
	if (!table)
T
Thomas Graf 已提交
2289
		return NULL;
L
Linus Torvalds 已提交
2290

2291
	read_lock_bh(&table->tb6_lock);
2292
	for (rt = table->tb6_root.leaf; rt; rt = rt->dst.rt6_next) {
2293
		if (dev == rt->dst.dev &&
2294
		    ((rt->rt6i_flags & (RTF_ADDRCONF | RTF_DEFAULT)) == (RTF_ADDRCONF | RTF_DEFAULT)) &&
L
Linus Torvalds 已提交
2295 2296 2297 2298
		    ipv6_addr_equal(&rt->rt6i_gateway, addr))
			break;
	}
	if (rt)
2299
		dst_hold(&rt->dst);
2300
	read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
2301 2302 2303
	return rt;
}

2304
struct rt6_info *rt6_add_dflt_router(const struct in6_addr *gwaddr,
2305 2306
				     struct net_device *dev,
				     unsigned int pref)
L
Linus Torvalds 已提交
2307
{
2308 2309
	struct fib6_config cfg = {
		.fc_table	= RT6_TABLE_DFLT,
2310
		.fc_metric	= IP6_RT_PRIO_USER,
2311 2312 2313
		.fc_ifindex	= dev->ifindex,
		.fc_flags	= RTF_GATEWAY | RTF_ADDRCONF | RTF_DEFAULT |
				  RTF_UP | RTF_EXPIRES | RTF_PREF(pref),
2314
		.fc_nlinfo.portid = 0,
2315
		.fc_nlinfo.nlh = NULL,
2316
		.fc_nlinfo.nl_net = dev_net(dev),
2317
	};
L
Linus Torvalds 已提交
2318

A
Alexey Dobriyan 已提交
2319
	cfg.fc_gateway = *gwaddr;
L
Linus Torvalds 已提交
2320

2321
	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2322 2323 2324 2325

	return rt6_get_dflt_router(gwaddr, dev);
}

2326
void rt6_purge_dflt_routers(struct net *net)
L
Linus Torvalds 已提交
2327 2328
{
	struct rt6_info *rt;
T
Thomas Graf 已提交
2329 2330 2331
	struct fib6_table *table;

	/* NOTE: Keep consistent with rt6_get_dflt_router */
2332
	table = fib6_get_table(net, RT6_TABLE_DFLT);
2333
	if (!table)
T
Thomas Graf 已提交
2334
		return;
L
Linus Torvalds 已提交
2335 2336

restart:
T
Thomas Graf 已提交
2337
	read_lock_bh(&table->tb6_lock);
2338
	for (rt = table->tb6_root.leaf; rt; rt = rt->dst.rt6_next) {
2339 2340
		if (rt->rt6i_flags & (RTF_DEFAULT | RTF_ADDRCONF) &&
		    (!rt->rt6i_idev || rt->rt6i_idev->cnf.accept_ra != 2)) {
2341
			dst_hold(&rt->dst);
T
Thomas Graf 已提交
2342
			read_unlock_bh(&table->tb6_lock);
2343
			ip6_del_rt(rt);
L
Linus Torvalds 已提交
2344 2345 2346
			goto restart;
		}
	}
T
Thomas Graf 已提交
2347
	read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
2348 2349
}

2350 2351
static void rtmsg_to_fib6_config(struct net *net,
				 struct in6_rtmsg *rtmsg,
2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363
				 struct fib6_config *cfg)
{
	memset(cfg, 0, sizeof(*cfg));

	cfg->fc_table = RT6_TABLE_MAIN;
	cfg->fc_ifindex = rtmsg->rtmsg_ifindex;
	cfg->fc_metric = rtmsg->rtmsg_metric;
	cfg->fc_expires = rtmsg->rtmsg_info;
	cfg->fc_dst_len = rtmsg->rtmsg_dst_len;
	cfg->fc_src_len = rtmsg->rtmsg_src_len;
	cfg->fc_flags = rtmsg->rtmsg_flags;

2364
	cfg->fc_nlinfo.nl_net = net;
2365

A
Alexey Dobriyan 已提交
2366 2367 2368
	cfg->fc_dst = rtmsg->rtmsg_dst;
	cfg->fc_src = rtmsg->rtmsg_src;
	cfg->fc_gateway = rtmsg->rtmsg_gateway;
2369 2370
}

2371
int ipv6_route_ioctl(struct net *net, unsigned int cmd, void __user *arg)
L
Linus Torvalds 已提交
2372
{
2373
	struct fib6_config cfg;
L
Linus Torvalds 已提交
2374 2375 2376
	struct in6_rtmsg rtmsg;
	int err;

2377
	switch (cmd) {
L
Linus Torvalds 已提交
2378 2379
	case SIOCADDRT:		/* Add a route */
	case SIOCDELRT:		/* Delete a route */
2380
		if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2381 2382 2383 2384 2385
			return -EPERM;
		err = copy_from_user(&rtmsg, arg,
				     sizeof(struct in6_rtmsg));
		if (err)
			return -EFAULT;
2386

2387
		rtmsg_to_fib6_config(net, &rtmsg, &cfg);
2388

L
Linus Torvalds 已提交
2389 2390 2391
		rtnl_lock();
		switch (cmd) {
		case SIOCADDRT:
2392
			err = ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2393 2394
			break;
		case SIOCDELRT:
2395
			err = ip6_route_del(&cfg);
L
Linus Torvalds 已提交
2396 2397 2398 2399 2400 2401 2402
			break;
		default:
			err = -EINVAL;
		}
		rtnl_unlock();

		return err;
2403
	}
L
Linus Torvalds 已提交
2404 2405 2406 2407 2408 2409 2410 2411

	return -EINVAL;
}

/*
 *	Drop the packet on the floor
 */

2412
static int ip6_pkt_drop(struct sk_buff *skb, u8 code, int ipstats_mib_noroutes)
L
Linus Torvalds 已提交
2413
{
2414
	int type;
E
Eric Dumazet 已提交
2415
	struct dst_entry *dst = skb_dst(skb);
2416 2417
	switch (ipstats_mib_noroutes) {
	case IPSTATS_MIB_INNOROUTES:
2418
		type = ipv6_addr_type(&ipv6_hdr(skb)->daddr);
U
Ulrich Weber 已提交
2419
		if (type == IPV6_ADDR_ANY) {
2420 2421
			IP6_INC_STATS(dev_net(dst->dev), ip6_dst_idev(dst),
				      IPSTATS_MIB_INADDRERRORS);
2422 2423 2424 2425
			break;
		}
		/* FALLTHROUGH */
	case IPSTATS_MIB_OUTNOROUTES:
2426 2427
		IP6_INC_STATS(dev_net(dst->dev), ip6_dst_idev(dst),
			      ipstats_mib_noroutes);
2428 2429
		break;
	}
2430
	icmpv6_send(skb, ICMPV6_DEST_UNREACH, code, 0);
L
Linus Torvalds 已提交
2431 2432 2433 2434
	kfree_skb(skb);
	return 0;
}

2435 2436
static int ip6_pkt_discard(struct sk_buff *skb)
{
2437
	return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_INNOROUTES);
2438 2439
}

2440
static int ip6_pkt_discard_out(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
2441
{
E
Eric Dumazet 已提交
2442
	skb->dev = skb_dst(skb)->dev;
2443
	return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
2444 2445
}

2446 2447
static int ip6_pkt_prohibit(struct sk_buff *skb)
{
2448
	return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_INNOROUTES);
2449 2450
}

2451
static int ip6_pkt_prohibit_out(struct sock *sk, struct sk_buff *skb)
2452
{
E
Eric Dumazet 已提交
2453
	skb->dev = skb_dst(skb)->dev;
2454
	return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_OUTNOROUTES);
2455 2456
}

L
Linus Torvalds 已提交
2457 2458 2459 2460 2461 2462
/*
 *	Allocate a dst for local (unicast / anycast) address.
 */

struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
				    const struct in6_addr *addr,
2463
				    bool anycast)
L
Linus Torvalds 已提交
2464
{
2465
	struct net *net = dev_net(idev->dev);
2466
	struct rt6_info *rt = ip6_dst_alloc(net, net->loopback_dev,
2467
					    DST_NOCOUNT);
2468
	if (!rt)
L
Linus Torvalds 已提交
2469 2470 2471 2472
		return ERR_PTR(-ENOMEM);

	in6_dev_hold(idev);

2473
	rt->dst.flags |= DST_HOST;
2474 2475
	rt->dst.input = ip6_input;
	rt->dst.output = ip6_output;
L
Linus Torvalds 已提交
2476 2477 2478
	rt->rt6i_idev = idev;

	rt->rt6i_flags = RTF_UP | RTF_NONEXTHOP;
2479 2480 2481
	if (anycast)
		rt->rt6i_flags |= RTF_ANYCAST;
	else
L
Linus Torvalds 已提交
2482 2483
		rt->rt6i_flags |= RTF_LOCAL;

2484
	rt->rt6i_gateway  = *addr;
A
Alexey Dobriyan 已提交
2485
	rt->rt6i_dst.addr = *addr;
L
Linus Torvalds 已提交
2486
	rt->rt6i_dst.plen = 128;
2487
	rt->rt6i_table = fib6_get_table(net, RT6_TABLE_LOCAL);
L
Linus Torvalds 已提交
2488

2489
	atomic_set(&rt->dst.__refcnt, 1);
L
Linus Torvalds 已提交
2490 2491 2492 2493

	return rt;
}

2494 2495
int ip6_route_get_saddr(struct net *net,
			struct rt6_info *rt,
2496
			const struct in6_addr *daddr,
2497 2498 2499
			unsigned int prefs,
			struct in6_addr *saddr)
{
2500 2501
	struct inet6_dev *idev =
		rt ? ip6_dst_idev((struct dst_entry *)rt) : NULL;
2502
	int err = 0;
2503
	if (rt && rt->rt6i_prefsrc.plen)
A
Alexey Dobriyan 已提交
2504
		*saddr = rt->rt6i_prefsrc.addr;
2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523
	else
		err = ipv6_dev_get_saddr(net, idev ? idev->dev : NULL,
					 daddr, prefs, saddr);
	return err;
}

/* remove deleted ip from prefsrc entries */
struct arg_dev_net_ip {
	struct net_device *dev;
	struct net *net;
	struct in6_addr *addr;
};

static int fib6_remove_prefsrc(struct rt6_info *rt, void *arg)
{
	struct net_device *dev = ((struct arg_dev_net_ip *)arg)->dev;
	struct net *net = ((struct arg_dev_net_ip *)arg)->net;
	struct in6_addr *addr = ((struct arg_dev_net_ip *)arg)->addr;

2524
	if (((void *)rt->dst.dev == dev || !dev) &&
2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540
	    rt != net->ipv6.ip6_null_entry &&
	    ipv6_addr_equal(addr, &rt->rt6i_prefsrc.addr)) {
		/* remove prefsrc entry */
		rt->rt6i_prefsrc.plen = 0;
	}
	return 0;
}

void rt6_remove_prefsrc(struct inet6_ifaddr *ifp)
{
	struct net *net = dev_net(ifp->idev->dev);
	struct arg_dev_net_ip adni = {
		.dev = ifp->idev->dev,
		.net = net,
		.addr = &ifp->addr,
	};
2541
	fib6_clean_all(net, fib6_remove_prefsrc, &adni);
2542 2543
}

2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564
#define RTF_RA_ROUTER		(RTF_ADDRCONF | RTF_DEFAULT | RTF_GATEWAY)
#define RTF_CACHE_GATEWAY	(RTF_GATEWAY | RTF_CACHE)

/* Remove routers and update dst entries when gateway turn into host. */
static int fib6_clean_tohost(struct rt6_info *rt, void *arg)
{
	struct in6_addr *gateway = (struct in6_addr *)arg;

	if ((((rt->rt6i_flags & RTF_RA_ROUTER) == RTF_RA_ROUTER) ||
	     ((rt->rt6i_flags & RTF_CACHE_GATEWAY) == RTF_CACHE_GATEWAY)) &&
	     ipv6_addr_equal(gateway, &rt->rt6i_gateway)) {
		return -1;
	}
	return 0;
}

void rt6_clean_tohost(struct net *net, struct in6_addr *gateway)
{
	fib6_clean_all(net, fib6_clean_tohost, gateway);
}

2565 2566 2567 2568 2569
struct arg_dev_net {
	struct net_device *dev;
	struct net *net;
};

L
Linus Torvalds 已提交
2570 2571
static int fib6_ifdown(struct rt6_info *rt, void *arg)
{
S
stephen hemminger 已提交
2572 2573
	const struct arg_dev_net *adn = arg;
	const struct net_device *dev = adn->dev;
2574

2575
	if ((rt->dst.dev == dev || !dev) &&
2576
	    rt != adn->net->ipv6.ip6_null_entry)
L
Linus Torvalds 已提交
2577
		return -1;
2578

L
Linus Torvalds 已提交
2579 2580 2581
	return 0;
}

2582
void rt6_ifdown(struct net *net, struct net_device *dev)
L
Linus Torvalds 已提交
2583
{
2584 2585 2586 2587 2588
	struct arg_dev_net adn = {
		.dev = dev,
		.net = net,
	};

2589
	fib6_clean_all(net, fib6_ifdown, &adn);
2590
	icmp6_clean_all(fib6_ifdown, &adn);
2591
	rt6_uncached_list_flush_dev(net, dev);
L
Linus Torvalds 已提交
2592 2593
}

2594
struct rt6_mtu_change_arg {
L
Linus Torvalds 已提交
2595
	struct net_device *dev;
2596
	unsigned int mtu;
L
Linus Torvalds 已提交
2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610
};

static int rt6_mtu_change_route(struct rt6_info *rt, void *p_arg)
{
	struct rt6_mtu_change_arg *arg = (struct rt6_mtu_change_arg *) p_arg;
	struct inet6_dev *idev;

	/* In IPv6 pmtu discovery is not optional,
	   so that RTAX_MTU lock cannot disable it.
	   We still use this lock to block changes
	   caused by addrconf/ndisc.
	*/

	idev = __in6_dev_get(arg->dev);
2611
	if (!idev)
L
Linus Torvalds 已提交
2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627
		return 0;

	/* For administrative MTU increase, there is no way to discover
	   IPv6 PMTU increase, so PMTU increase should be updated here.
	   Since RFC 1981 doesn't include administrative MTU increase
	   update PMTU increase is a MUST. (i.e. jumbo frame)
	 */
	/*
	   If new MTU is less than route PMTU, this new MTU will be the
	   lowest MTU in the path, update the route PMTU to reflect PMTU
	   decreases; if new MTU is greater than route PMTU, and the
	   old MTU is the lowest MTU in the path, update the route PMTU
	   to reflect the increase. In this case if the other nodes' MTU
	   also have the lowest MTU, TOO BIG MESSAGE will be lead to
	   PMTU discouvery.
	 */
2628
	if (rt->dst.dev == arg->dev &&
2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642
	    !dst_metric_locked(&rt->dst, RTAX_MTU)) {
		if (rt->rt6i_flags & RTF_CACHE) {
			/* For RTF_CACHE with rt6i_pmtu == 0
			 * (i.e. a redirected route),
			 * the metrics of its rt->dst.from has already
			 * been updated.
			 */
			if (rt->rt6i_pmtu && rt->rt6i_pmtu > arg->mtu)
				rt->rt6i_pmtu = arg->mtu;
		} else if (dst_mtu(&rt->dst) >= arg->mtu ||
			   (dst_mtu(&rt->dst) < arg->mtu &&
			    dst_mtu(&rt->dst) == idev->cnf.mtu6)) {
			dst_metric_set(&rt->dst, RTAX_MTU, arg->mtu);
		}
2643
	}
L
Linus Torvalds 已提交
2644 2645 2646
	return 0;
}

2647
void rt6_mtu_change(struct net_device *dev, unsigned int mtu)
L
Linus Torvalds 已提交
2648
{
T
Thomas Graf 已提交
2649 2650 2651 2652
	struct rt6_mtu_change_arg arg = {
		.dev = dev,
		.mtu = mtu,
	};
L
Linus Torvalds 已提交
2653

2654
	fib6_clean_all(dev_net(dev), rt6_mtu_change_route, &arg);
L
Linus Torvalds 已提交
2655 2656
}

2657
static const struct nla_policy rtm_ipv6_policy[RTA_MAX+1] = {
2658
	[RTA_GATEWAY]           = { .len = sizeof(struct in6_addr) },
2659
	[RTA_OIF]               = { .type = NLA_U32 },
2660
	[RTA_IIF]		= { .type = NLA_U32 },
2661 2662
	[RTA_PRIORITY]          = { .type = NLA_U32 },
	[RTA_METRICS]           = { .type = NLA_NESTED },
2663
	[RTA_MULTIPATH]		= { .len = sizeof(struct rtnexthop) },
2664
	[RTA_PREF]              = { .type = NLA_U8 },
2665 2666
	[RTA_ENCAP_TYPE]	= { .type = NLA_U16 },
	[RTA_ENCAP]		= { .type = NLA_NESTED },
2667 2668 2669 2670
};

static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh,
			      struct fib6_config *cfg)
L
Linus Torvalds 已提交
2671
{
2672 2673
	struct rtmsg *rtm;
	struct nlattr *tb[RTA_MAX+1];
2674
	unsigned int pref;
2675
	int err;
L
Linus Torvalds 已提交
2676

2677 2678 2679
	err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv6_policy);
	if (err < 0)
		goto errout;
L
Linus Torvalds 已提交
2680

2681 2682 2683 2684 2685 2686 2687 2688 2689
	err = -EINVAL;
	rtm = nlmsg_data(nlh);
	memset(cfg, 0, sizeof(*cfg));

	cfg->fc_table = rtm->rtm_table;
	cfg->fc_dst_len = rtm->rtm_dst_len;
	cfg->fc_src_len = rtm->rtm_src_len;
	cfg->fc_flags = RTF_UP;
	cfg->fc_protocol = rtm->rtm_protocol;
2690
	cfg->fc_type = rtm->rtm_type;
2691

2692 2693
	if (rtm->rtm_type == RTN_UNREACHABLE ||
	    rtm->rtm_type == RTN_BLACKHOLE ||
2694 2695
	    rtm->rtm_type == RTN_PROHIBIT ||
	    rtm->rtm_type == RTN_THROW)
2696 2697
		cfg->fc_flags |= RTF_REJECT;

2698 2699 2700
	if (rtm->rtm_type == RTN_LOCAL)
		cfg->fc_flags |= RTF_LOCAL;

2701 2702 2703
	if (rtm->rtm_flags & RTM_F_CLONED)
		cfg->fc_flags |= RTF_CACHE;

2704
	cfg->fc_nlinfo.portid = NETLINK_CB(skb).portid;
2705
	cfg->fc_nlinfo.nlh = nlh;
2706
	cfg->fc_nlinfo.nl_net = sock_net(skb->sk);
2707 2708

	if (tb[RTA_GATEWAY]) {
2709
		cfg->fc_gateway = nla_get_in6_addr(tb[RTA_GATEWAY]);
2710
		cfg->fc_flags |= RTF_GATEWAY;
L
Linus Torvalds 已提交
2711
	}
2712 2713 2714 2715 2716 2717 2718 2719

	if (tb[RTA_DST]) {
		int plen = (rtm->rtm_dst_len + 7) >> 3;

		if (nla_len(tb[RTA_DST]) < plen)
			goto errout;

		nla_memcpy(&cfg->fc_dst, tb[RTA_DST], plen);
L
Linus Torvalds 已提交
2720
	}
2721 2722 2723 2724 2725 2726 2727 2728

	if (tb[RTA_SRC]) {
		int plen = (rtm->rtm_src_len + 7) >> 3;

		if (nla_len(tb[RTA_SRC]) < plen)
			goto errout;

		nla_memcpy(&cfg->fc_src, tb[RTA_SRC], plen);
L
Linus Torvalds 已提交
2729
	}
2730

2731
	if (tb[RTA_PREFSRC])
2732
		cfg->fc_prefsrc = nla_get_in6_addr(tb[RTA_PREFSRC]);
2733

2734 2735 2736 2737 2738 2739 2740 2741 2742
	if (tb[RTA_OIF])
		cfg->fc_ifindex = nla_get_u32(tb[RTA_OIF]);

	if (tb[RTA_PRIORITY])
		cfg->fc_metric = nla_get_u32(tb[RTA_PRIORITY]);

	if (tb[RTA_METRICS]) {
		cfg->fc_mx = nla_data(tb[RTA_METRICS]);
		cfg->fc_mx_len = nla_len(tb[RTA_METRICS]);
L
Linus Torvalds 已提交
2743
	}
2744 2745 2746 2747

	if (tb[RTA_TABLE])
		cfg->fc_table = nla_get_u32(tb[RTA_TABLE]);

2748 2749 2750 2751 2752
	if (tb[RTA_MULTIPATH]) {
		cfg->fc_mp = nla_data(tb[RTA_MULTIPATH]);
		cfg->fc_mp_len = nla_len(tb[RTA_MULTIPATH]);
	}

2753 2754 2755 2756 2757 2758 2759 2760
	if (tb[RTA_PREF]) {
		pref = nla_get_u8(tb[RTA_PREF]);
		if (pref != ICMPV6_ROUTER_PREF_LOW &&
		    pref != ICMPV6_ROUTER_PREF_HIGH)
			pref = ICMPV6_ROUTER_PREF_MEDIUM;
		cfg->fc_flags |= RTF_PREF(pref);
	}

2761 2762 2763 2764 2765 2766
	if (tb[RTA_ENCAP])
		cfg->fc_encap = tb[RTA_ENCAP];

	if (tb[RTA_ENCAP_TYPE])
		cfg->fc_encap_type = nla_get_u16(tb[RTA_ENCAP_TYPE]);

2767 2768 2769
	err = 0;
errout:
	return err;
L
Linus Torvalds 已提交
2770 2771
}

2772 2773 2774 2775 2776 2777 2778 2779
static int ip6_route_multipath(struct fib6_config *cfg, int add)
{
	struct fib6_config r_cfg;
	struct rtnexthop *rtnh;
	int remaining;
	int attrlen;
	int err = 0, last_err = 0;

2780
	remaining = cfg->fc_mp_len;
2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795
beginning:
	rtnh = (struct rtnexthop *)cfg->fc_mp;

	/* Parse a Multipath Entry */
	while (rtnh_ok(rtnh, remaining)) {
		memcpy(&r_cfg, cfg, sizeof(*cfg));
		if (rtnh->rtnh_ifindex)
			r_cfg.fc_ifindex = rtnh->rtnh_ifindex;

		attrlen = rtnh_attrlen(rtnh);
		if (attrlen > 0) {
			struct nlattr *nla, *attrs = rtnh_attrs(rtnh);

			nla = nla_find(attrs, attrlen, RTA_GATEWAY);
			if (nla) {
2796
				r_cfg.fc_gateway = nla_get_in6_addr(nla);
2797 2798
				r_cfg.fc_flags |= RTF_GATEWAY;
			}
2799 2800 2801 2802
			r_cfg.fc_encap = nla_find(attrs, attrlen, RTA_ENCAP);
			nla = nla_find(attrs, attrlen, RTA_ENCAP_TYPE);
			if (nla)
				r_cfg.fc_encap_type = nla_get_u16(nla);
2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815
		}
		err = add ? ip6_route_add(&r_cfg) : ip6_route_del(&r_cfg);
		if (err) {
			last_err = err;
			/* If we are trying to remove a route, do not stop the
			 * loop when ip6_route_del() fails (because next hop is
			 * already gone), we should try to remove all next hops.
			 */
			if (add) {
				/* If add fails, we should try to delete all
				 * next hops that have been already added.
				 */
				add = 0;
2816
				remaining = cfg->fc_mp_len - remaining;
2817 2818 2819
				goto beginning;
			}
		}
2820
		/* Because each route is added like a single route we remove
2821 2822 2823 2824 2825
		 * these flags after the first nexthop: if there is a collision,
		 * we have already failed to add the first nexthop:
		 * fib6_add_rt2node() has rejected it; when replacing, old
		 * nexthops have been replaced by first new, the rest should
		 * be added to it.
2826
		 */
2827 2828
		cfg->fc_nlinfo.nlh->nlmsg_flags &= ~(NLM_F_EXCL |
						     NLM_F_REPLACE);
2829 2830 2831 2832 2833 2834
		rtnh = rtnh_next(rtnh, &remaining);
	}

	return last_err;
}

2835
static int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
2836
{
2837 2838
	struct fib6_config cfg;
	int err;
L
Linus Torvalds 已提交
2839

2840 2841 2842 2843
	err = rtm_to_fib6_config(skb, nlh, &cfg);
	if (err < 0)
		return err;

2844 2845 2846 2847
	if (cfg.fc_mp)
		return ip6_route_multipath(&cfg, 0);
	else
		return ip6_route_del(&cfg);
L
Linus Torvalds 已提交
2848 2849
}

2850
static int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
2851
{
2852 2853
	struct fib6_config cfg;
	int err;
L
Linus Torvalds 已提交
2854

2855 2856 2857 2858
	err = rtm_to_fib6_config(skb, nlh, &cfg);
	if (err < 0)
		return err;

2859 2860 2861 2862
	if (cfg.fc_mp)
		return ip6_route_multipath(&cfg, 1);
	else
		return ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2863 2864
}

2865
static inline size_t rt6_nlmsg_size(struct rt6_info *rt)
2866 2867 2868 2869 2870 2871 2872 2873 2874 2875
{
	return NLMSG_ALIGN(sizeof(struct rtmsg))
	       + nla_total_size(16) /* RTA_SRC */
	       + nla_total_size(16) /* RTA_DST */
	       + nla_total_size(16) /* RTA_GATEWAY */
	       + nla_total_size(16) /* RTA_PREFSRC */
	       + nla_total_size(4) /* RTA_TABLE */
	       + nla_total_size(4) /* RTA_IIF */
	       + nla_total_size(4) /* RTA_OIF */
	       + nla_total_size(4) /* RTA_PRIORITY */
2876
	       + RTAX_MAX * nla_total_size(4) /* RTA_METRICS */
2877
	       + nla_total_size(sizeof(struct rta_cacheinfo))
2878
	       + nla_total_size(TCP_CA_NAME_MAX) /* RTAX_CC_ALGO */
2879
	       + nla_total_size(1) /* RTA_PREF */
2880
	       + lwtunnel_get_encap_size(rt->dst.lwtstate);
2881 2882
}

2883 2884
static int rt6_fill_node(struct net *net,
			 struct sk_buff *skb, struct rt6_info *rt,
2885
			 struct in6_addr *dst, struct in6_addr *src,
2886
			 int iif, int type, u32 portid, u32 seq,
2887
			 int prefix, int nowait, unsigned int flags)
L
Linus Torvalds 已提交
2888
{
2889
	u32 metrics[RTAX_MAX];
L
Linus Torvalds 已提交
2890
	struct rtmsg *rtm;
2891
	struct nlmsghdr *nlh;
2892
	long expires;
2893
	u32 table;
L
Linus Torvalds 已提交
2894 2895 2896 2897 2898 2899 2900 2901

	if (prefix) {	/* user wants prefix routes only */
		if (!(rt->rt6i_flags & RTF_PREFIX_RT)) {
			/* success since this is not a prefix route */
			return 1;
		}
	}

2902
	nlh = nlmsg_put(skb, portid, seq, type, sizeof(*rtm), flags);
2903
	if (!nlh)
2904
		return -EMSGSIZE;
2905 2906

	rtm = nlmsg_data(nlh);
L
Linus Torvalds 已提交
2907 2908 2909 2910
	rtm->rtm_family = AF_INET6;
	rtm->rtm_dst_len = rt->rt6i_dst.plen;
	rtm->rtm_src_len = rt->rt6i_src.plen;
	rtm->rtm_tos = 0;
T
Thomas Graf 已提交
2911
	if (rt->rt6i_table)
2912
		table = rt->rt6i_table->tb6_id;
T
Thomas Graf 已提交
2913
	else
2914 2915
		table = RT6_TABLE_UNSPEC;
	rtm->rtm_table = table;
D
David S. Miller 已提交
2916 2917
	if (nla_put_u32(skb, RTA_TABLE, table))
		goto nla_put_failure;
2918 2919 2920 2921 2922 2923 2924 2925
	if (rt->rt6i_flags & RTF_REJECT) {
		switch (rt->dst.error) {
		case -EINVAL:
			rtm->rtm_type = RTN_BLACKHOLE;
			break;
		case -EACCES:
			rtm->rtm_type = RTN_PROHIBIT;
			break;
2926 2927 2928
		case -EAGAIN:
			rtm->rtm_type = RTN_THROW;
			break;
2929 2930 2931 2932 2933
		default:
			rtm->rtm_type = RTN_UNREACHABLE;
			break;
		}
	}
2934
	else if (rt->rt6i_flags & RTF_LOCAL)
2935
		rtm->rtm_type = RTN_LOCAL;
2936
	else if (rt->dst.dev && (rt->dst.dev->flags & IFF_LOOPBACK))
L
Linus Torvalds 已提交
2937 2938 2939 2940
		rtm->rtm_type = RTN_LOCAL;
	else
		rtm->rtm_type = RTN_UNICAST;
	rtm->rtm_flags = 0;
2941
	if (!netif_carrier_ok(rt->dst.dev)) {
2942
		rtm->rtm_flags |= RTNH_F_LINKDOWN;
2943 2944 2945
		if (rt->rt6i_idev->cnf.ignore_routes_with_linkdown)
			rtm->rtm_flags |= RTNH_F_DEAD;
	}
L
Linus Torvalds 已提交
2946 2947
	rtm->rtm_scope = RT_SCOPE_UNIVERSE;
	rtm->rtm_protocol = rt->rt6i_protocol;
2948
	if (rt->rt6i_flags & RTF_DYNAMIC)
L
Linus Torvalds 已提交
2949
		rtm->rtm_protocol = RTPROT_REDIRECT;
2950 2951 2952 2953 2954 2955
	else if (rt->rt6i_flags & RTF_ADDRCONF) {
		if (rt->rt6i_flags & (RTF_DEFAULT | RTF_ROUTEINFO))
			rtm->rtm_protocol = RTPROT_RA;
		else
			rtm->rtm_protocol = RTPROT_KERNEL;
	}
L
Linus Torvalds 已提交
2956

2957
	if (rt->rt6i_flags & RTF_CACHE)
L
Linus Torvalds 已提交
2958 2959 2960
		rtm->rtm_flags |= RTM_F_CLONED;

	if (dst) {
2961
		if (nla_put_in6_addr(skb, RTA_DST, dst))
D
David S. Miller 已提交
2962
			goto nla_put_failure;
2963
		rtm->rtm_dst_len = 128;
L
Linus Torvalds 已提交
2964
	} else if (rtm->rtm_dst_len)
2965
		if (nla_put_in6_addr(skb, RTA_DST, &rt->rt6i_dst.addr))
D
David S. Miller 已提交
2966
			goto nla_put_failure;
L
Linus Torvalds 已提交
2967 2968
#ifdef CONFIG_IPV6_SUBTREES
	if (src) {
2969
		if (nla_put_in6_addr(skb, RTA_SRC, src))
D
David S. Miller 已提交
2970
			goto nla_put_failure;
2971
		rtm->rtm_src_len = 128;
D
David S. Miller 已提交
2972
	} else if (rtm->rtm_src_len &&
2973
		   nla_put_in6_addr(skb, RTA_SRC, &rt->rt6i_src.addr))
D
David S. Miller 已提交
2974
		goto nla_put_failure;
L
Linus Torvalds 已提交
2975
#endif
2976 2977 2978
	if (iif) {
#ifdef CONFIG_IPV6_MROUTE
		if (ipv6_addr_is_multicast(&rt->rt6i_dst.addr)) {
2979
			int err = ip6mr_get_route(net, skb, rtm, nowait);
2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991
			if (err <= 0) {
				if (!nowait) {
					if (err == 0)
						return 0;
					goto nla_put_failure;
				} else {
					if (err == -EMSGSIZE)
						goto nla_put_failure;
				}
			}
		} else
#endif
D
David S. Miller 已提交
2992 2993
			if (nla_put_u32(skb, RTA_IIF, iif))
				goto nla_put_failure;
2994
	} else if (dst) {
L
Linus Torvalds 已提交
2995
		struct in6_addr saddr_buf;
D
David S. Miller 已提交
2996
		if (ip6_route_get_saddr(net, rt, dst, 0, &saddr_buf) == 0 &&
2997
		    nla_put_in6_addr(skb, RTA_PREFSRC, &saddr_buf))
D
David S. Miller 已提交
2998
			goto nla_put_failure;
L
Linus Torvalds 已提交
2999
	}
3000

3001 3002
	if (rt->rt6i_prefsrc.plen) {
		struct in6_addr saddr_buf;
A
Alexey Dobriyan 已提交
3003
		saddr_buf = rt->rt6i_prefsrc.addr;
3004
		if (nla_put_in6_addr(skb, RTA_PREFSRC, &saddr_buf))
D
David S. Miller 已提交
3005
			goto nla_put_failure;
3006 3007
	}

3008 3009 3010 3011
	memcpy(metrics, dst_metrics_ptr(&rt->dst), sizeof(metrics));
	if (rt->rt6i_pmtu)
		metrics[RTAX_MTU - 1] = rt->rt6i_pmtu;
	if (rtnetlink_put_metrics(skb, metrics) < 0)
3012 3013
		goto nla_put_failure;

3014
	if (rt->rt6i_flags & RTF_GATEWAY) {
3015
		if (nla_put_in6_addr(skb, RTA_GATEWAY, &rt->rt6i_gateway) < 0)
3016 3017
			goto nla_put_failure;
	}
3018

D
David S. Miller 已提交
3019 3020 3021 3022 3023
	if (rt->dst.dev &&
	    nla_put_u32(skb, RTA_OIF, rt->dst.dev->ifindex))
		goto nla_put_failure;
	if (nla_put_u32(skb, RTA_PRIORITY, rt->rt6i_metric))
		goto nla_put_failure;
3024 3025

	expires = (rt->rt6i_flags & RTF_EXPIRES) ? rt->dst.expires - jiffies : 0;
3026

3027
	if (rtnl_put_cacheinfo(skb, &rt->dst, 0, expires, rt->dst.error) < 0)
3028
		goto nla_put_failure;
3029

3030 3031 3032
	if (nla_put_u8(skb, RTA_PREF, IPV6_EXTRACT_PREF(rt->rt6i_flags)))
		goto nla_put_failure;

3033
	lwtunnel_fill_encap(skb, rt->dst.lwtstate);
3034

3035 3036
	nlmsg_end(skb, nlh);
	return 0;
3037 3038

nla_put_failure:
3039 3040
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
3041 3042
}

3043
int rt6_dump_route(struct rt6_info *rt, void *p_arg)
L
Linus Torvalds 已提交
3044 3045 3046 3047
{
	struct rt6_rtnl_dump_arg *arg = (struct rt6_rtnl_dump_arg *) p_arg;
	int prefix;

3048 3049
	if (nlmsg_len(arg->cb->nlh) >= sizeof(struct rtmsg)) {
		struct rtmsg *rtm = nlmsg_data(arg->cb->nlh);
L
Linus Torvalds 已提交
3050 3051 3052 3053
		prefix = (rtm->rtm_flags & RTM_F_PREFIX) != 0;
	} else
		prefix = 0;

3054 3055
	return rt6_fill_node(arg->net,
		     arg->skb, rt, NULL, NULL, 0, RTM_NEWROUTE,
3056
		     NETLINK_CB(arg->cb->skb).portid, arg->cb->nlh->nlmsg_seq,
3057
		     prefix, 0, NLM_F_MULTI);
L
Linus Torvalds 已提交
3058 3059
}

3060
static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
3061
{
3062
	struct net *net = sock_net(in_skb->sk);
3063 3064
	struct nlattr *tb[RTA_MAX+1];
	struct rt6_info *rt;
L
Linus Torvalds 已提交
3065
	struct sk_buff *skb;
3066
	struct rtmsg *rtm;
3067
	struct flowi6 fl6;
3068
	int err, iif = 0, oif = 0;
L
Linus Torvalds 已提交
3069

3070 3071 3072
	err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv6_policy);
	if (err < 0)
		goto errout;
L
Linus Torvalds 已提交
3073

3074
	err = -EINVAL;
3075
	memset(&fl6, 0, sizeof(fl6));
L
Linus Torvalds 已提交
3076

3077 3078 3079 3080
	if (tb[RTA_SRC]) {
		if (nla_len(tb[RTA_SRC]) < sizeof(struct in6_addr))
			goto errout;

A
Alexey Dobriyan 已提交
3081
		fl6.saddr = *(struct in6_addr *)nla_data(tb[RTA_SRC]);
3082 3083 3084 3085 3086 3087
	}

	if (tb[RTA_DST]) {
		if (nla_len(tb[RTA_DST]) < sizeof(struct in6_addr))
			goto errout;

A
Alexey Dobriyan 已提交
3088
		fl6.daddr = *(struct in6_addr *)nla_data(tb[RTA_DST]);
3089 3090 3091 3092 3093 3094
	}

	if (tb[RTA_IIF])
		iif = nla_get_u32(tb[RTA_IIF]);

	if (tb[RTA_OIF])
3095
		oif = nla_get_u32(tb[RTA_OIF]);
L
Linus Torvalds 已提交
3096

3097 3098 3099
	if (tb[RTA_MARK])
		fl6.flowi6_mark = nla_get_u32(tb[RTA_MARK]);

L
Linus Torvalds 已提交
3100 3101
	if (iif) {
		struct net_device *dev;
3102 3103
		int flags = 0;

3104
		dev = __dev_get_by_index(net, iif);
L
Linus Torvalds 已提交
3105 3106
		if (!dev) {
			err = -ENODEV;
3107
			goto errout;
L
Linus Torvalds 已提交
3108
		}
3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120

		fl6.flowi6_iif = iif;

		if (!ipv6_addr_any(&fl6.saddr))
			flags |= RT6_LOOKUP_F_HAS_SADDR;

		rt = (struct rt6_info *)ip6_route_input_lookup(net, dev, &fl6,
							       flags);
	} else {
		fl6.flowi6_oif = oif;

		rt = (struct rt6_info *)ip6_route_output(net, NULL, &fl6);
L
Linus Torvalds 已提交
3121 3122
	}

3123
	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
3124
	if (!skb) {
A
Amerigo Wang 已提交
3125
		ip6_rt_put(rt);
3126 3127 3128
		err = -ENOBUFS;
		goto errout;
	}
L
Linus Torvalds 已提交
3129

3130 3131 3132
	/* Reserve room for dummy headers, this skb can pass
	   through good chunk of routing engine.
	 */
3133
	skb_reset_mac_header(skb);
3134
	skb_reserve(skb, MAX_HEADER + sizeof(struct ipv6hdr));
L
Linus Torvalds 已提交
3135

3136
	skb_dst_set(skb, &rt->dst);
L
Linus Torvalds 已提交
3137

3138
	err = rt6_fill_node(net, skb, rt, &fl6.daddr, &fl6.saddr, iif,
3139
			    RTM_NEWROUTE, NETLINK_CB(in_skb).portid,
3140
			    nlh->nlmsg_seq, 0, 0, 0);
L
Linus Torvalds 已提交
3141
	if (err < 0) {
3142 3143
		kfree_skb(skb);
		goto errout;
L
Linus Torvalds 已提交
3144 3145
	}

3146
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
3147
errout:
L
Linus Torvalds 已提交
3148 3149 3150
	return err;
}

3151
void inet6_rt_notify(int event, struct rt6_info *rt, struct nl_info *info)
L
Linus Torvalds 已提交
3152 3153
{
	struct sk_buff *skb;
3154
	struct net *net = info->nl_net;
3155 3156 3157 3158
	u32 seq;
	int err;

	err = -ENOBUFS;
3159
	seq = info->nlh ? info->nlh->nlmsg_seq : 0;
3160

3161
	skb = nlmsg_new(rt6_nlmsg_size(rt), gfp_any());
3162
	if (!skb)
3163 3164
		goto errout;

3165
	err = rt6_fill_node(net, skb, rt, NULL, NULL, 0,
3166
				event, info->portid, seq, 0, 0, 0);
3167 3168 3169 3170 3171 3172
	if (err < 0) {
		/* -EMSGSIZE implies BUG in rt6_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
3173
	rtnl_notify(skb, net, info->portid, RTNLGRP_IPV6_ROUTE,
3174 3175
		    info->nlh, gfp_any());
	return;
3176 3177
errout:
	if (err < 0)
3178
		rtnl_set_sk_err(net, RTNLGRP_IPV6_ROUTE, err);
L
Linus Torvalds 已提交
3179 3180
}

3181
static int ip6_route_dev_notify(struct notifier_block *this,
3182
				unsigned long event, void *ptr)
3183
{
3184
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
3185
	struct net *net = dev_net(dev);
3186 3187

	if (event == NETDEV_REGISTER && (dev->flags & IFF_LOOPBACK)) {
3188
		net->ipv6.ip6_null_entry->dst.dev = dev;
3189 3190
		net->ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(dev);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
3191
		net->ipv6.ip6_prohibit_entry->dst.dev = dev;
3192
		net->ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(dev);
3193
		net->ipv6.ip6_blk_hole_entry->dst.dev = dev;
3194 3195 3196 3197 3198 3199 3200
		net->ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(dev);
#endif
	}

	return NOTIFY_OK;
}

L
Linus Torvalds 已提交
3201 3202 3203 3204 3205 3206
/*
 *	/proc
 */

#ifdef CONFIG_PROC_FS

3207 3208 3209 3210 3211
static const struct file_operations ipv6_route_proc_fops = {
	.owner		= THIS_MODULE,
	.open		= ipv6_route_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
3212
	.release	= seq_release_net,
3213 3214
};

L
Linus Torvalds 已提交
3215 3216
static int rt6_stats_seq_show(struct seq_file *seq, void *v)
{
3217
	struct net *net = (struct net *)seq->private;
L
Linus Torvalds 已提交
3218
	seq_printf(seq, "%04x %04x %04x %04x %04x %04x %04x\n",
3219 3220 3221 3222 3223
		   net->ipv6.rt6_stats->fib_nodes,
		   net->ipv6.rt6_stats->fib_route_nodes,
		   net->ipv6.rt6_stats->fib_rt_alloc,
		   net->ipv6.rt6_stats->fib_rt_entries,
		   net->ipv6.rt6_stats->fib_rt_cache,
3224
		   dst_entries_get_slow(&net->ipv6.ip6_dst_ops),
3225
		   net->ipv6.rt6_stats->fib_discarded_routes);
L
Linus Torvalds 已提交
3226 3227 3228 3229 3230 3231

	return 0;
}

static int rt6_stats_seq_open(struct inode *inode, struct file *file)
{
3232
	return single_open_net(inode, file, rt6_stats_seq_show);
3233 3234
}

3235
static const struct file_operations rt6_stats_seq_fops = {
L
Linus Torvalds 已提交
3236 3237 3238 3239
	.owner	 = THIS_MODULE,
	.open	 = rt6_stats_seq_open,
	.read	 = seq_read,
	.llseek	 = seq_lseek,
3240
	.release = single_release_net,
L
Linus Torvalds 已提交
3241 3242 3243 3244 3245 3246
};
#endif	/* CONFIG_PROC_FS */

#ifdef CONFIG_SYSCTL

static
3247
int ipv6_sysctl_rtcache_flush(struct ctl_table *ctl, int write,
L
Linus Torvalds 已提交
3248 3249
			      void __user *buffer, size_t *lenp, loff_t *ppos)
{
3250 3251 3252
	struct net *net;
	int delay;
	if (!write)
L
Linus Torvalds 已提交
3253
		return -EINVAL;
3254 3255 3256 3257

	net = (struct net *)ctl->extra1;
	delay = net->ipv6.sysctl.flush_delay;
	proc_dointvec(ctl, write, buffer, lenp, ppos);
3258
	fib6_run_gc(delay <= 0 ? 0 : (unsigned long)delay, net, delay > 0);
3259
	return 0;
L
Linus Torvalds 已提交
3260 3261
}

3262
struct ctl_table ipv6_route_table_template[] = {
3263
	{
L
Linus Torvalds 已提交
3264
		.procname	=	"flush",
3265
		.data		=	&init_net.ipv6.sysctl.flush_delay,
L
Linus Torvalds 已提交
3266
		.maxlen		=	sizeof(int),
3267
		.mode		=	0200,
A
Alexey Dobriyan 已提交
3268
		.proc_handler	=	ipv6_sysctl_rtcache_flush
L
Linus Torvalds 已提交
3269 3270 3271
	},
	{
		.procname	=	"gc_thresh",
3272
		.data		=	&ip6_dst_ops_template.gc_thresh,
L
Linus Torvalds 已提交
3273 3274
		.maxlen		=	sizeof(int),
		.mode		=	0644,
A
Alexey Dobriyan 已提交
3275
		.proc_handler	=	proc_dointvec,
L
Linus Torvalds 已提交
3276 3277 3278
	},
	{
		.procname	=	"max_size",
3279
		.data		=	&init_net.ipv6.sysctl.ip6_rt_max_size,
L
Linus Torvalds 已提交
3280 3281
		.maxlen		=	sizeof(int),
		.mode		=	0644,
A
Alexey Dobriyan 已提交
3282
		.proc_handler	=	proc_dointvec,
L
Linus Torvalds 已提交
3283 3284 3285
	},
	{
		.procname	=	"gc_min_interval",
3286
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_min_interval,
L
Linus Torvalds 已提交
3287 3288
		.maxlen		=	sizeof(int),
		.mode		=	0644,
A
Alexey Dobriyan 已提交
3289
		.proc_handler	=	proc_dointvec_jiffies,
L
Linus Torvalds 已提交
3290 3291 3292
	},
	{
		.procname	=	"gc_timeout",
3293
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_timeout,
L
Linus Torvalds 已提交
3294 3295
		.maxlen		=	sizeof(int),
		.mode		=	0644,
A
Alexey Dobriyan 已提交
3296
		.proc_handler	=	proc_dointvec_jiffies,
L
Linus Torvalds 已提交
3297 3298 3299
	},
	{
		.procname	=	"gc_interval",
3300
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_interval,
L
Linus Torvalds 已提交
3301 3302
		.maxlen		=	sizeof(int),
		.mode		=	0644,
A
Alexey Dobriyan 已提交
3303
		.proc_handler	=	proc_dointvec_jiffies,
L
Linus Torvalds 已提交
3304 3305 3306
	},
	{
		.procname	=	"gc_elasticity",
3307
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_elasticity,
L
Linus Torvalds 已提交
3308 3309
		.maxlen		=	sizeof(int),
		.mode		=	0644,
3310
		.proc_handler	=	proc_dointvec,
L
Linus Torvalds 已提交
3311 3312 3313
	},
	{
		.procname	=	"mtu_expires",
3314
		.data		=	&init_net.ipv6.sysctl.ip6_rt_mtu_expires,
L
Linus Torvalds 已提交
3315 3316
		.maxlen		=	sizeof(int),
		.mode		=	0644,
A
Alexey Dobriyan 已提交
3317
		.proc_handler	=	proc_dointvec_jiffies,
L
Linus Torvalds 已提交
3318 3319 3320
	},
	{
		.procname	=	"min_adv_mss",
3321
		.data		=	&init_net.ipv6.sysctl.ip6_rt_min_advmss,
L
Linus Torvalds 已提交
3322 3323
		.maxlen		=	sizeof(int),
		.mode		=	0644,
3324
		.proc_handler	=	proc_dointvec,
L
Linus Torvalds 已提交
3325 3326 3327
	},
	{
		.procname	=	"gc_min_interval_ms",
3328
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_min_interval,
L
Linus Torvalds 已提交
3329 3330
		.maxlen		=	sizeof(int),
		.mode		=	0644,
A
Alexey Dobriyan 已提交
3331
		.proc_handler	=	proc_dointvec_ms_jiffies,
L
Linus Torvalds 已提交
3332
	},
3333
	{ }
L
Linus Torvalds 已提交
3334 3335
};

3336
struct ctl_table * __net_init ipv6_route_sysctl_init(struct net *net)
3337 3338 3339 3340 3341 3342
{
	struct ctl_table *table;

	table = kmemdup(ipv6_route_table_template,
			sizeof(ipv6_route_table_template),
			GFP_KERNEL);
3343 3344 3345

	if (table) {
		table[0].data = &net->ipv6.sysctl.flush_delay;
3346
		table[0].extra1 = net;
3347
		table[1].data = &net->ipv6.ip6_dst_ops.gc_thresh;
3348 3349 3350 3351 3352 3353 3354
		table[2].data = &net->ipv6.sysctl.ip6_rt_max_size;
		table[3].data = &net->ipv6.sysctl.ip6_rt_gc_min_interval;
		table[4].data = &net->ipv6.sysctl.ip6_rt_gc_timeout;
		table[5].data = &net->ipv6.sysctl.ip6_rt_gc_interval;
		table[6].data = &net->ipv6.sysctl.ip6_rt_gc_elasticity;
		table[7].data = &net->ipv6.sysctl.ip6_rt_mtu_expires;
		table[8].data = &net->ipv6.sysctl.ip6_rt_min_advmss;
3355
		table[9].data = &net->ipv6.sysctl.ip6_rt_gc_min_interval;
3356 3357 3358 3359

		/* Don't export sysctls to unprivileged users */
		if (net->user_ns != &init_user_ns)
			table[0].procname = NULL;
3360 3361
	}

3362 3363
	return table;
}
L
Linus Torvalds 已提交
3364 3365
#endif

3366
static int __net_init ip6_route_net_init(struct net *net)
3367
{
3368
	int ret = -ENOMEM;
3369

3370 3371
	memcpy(&net->ipv6.ip6_dst_ops, &ip6_dst_ops_template,
	       sizeof(net->ipv6.ip6_dst_ops));
3372

3373 3374 3375
	if (dst_entries_init(&net->ipv6.ip6_dst_ops) < 0)
		goto out_ip6_dst_ops;

3376 3377 3378 3379
	net->ipv6.ip6_null_entry = kmemdup(&ip6_null_entry_template,
					   sizeof(*net->ipv6.ip6_null_entry),
					   GFP_KERNEL);
	if (!net->ipv6.ip6_null_entry)
3380
		goto out_ip6_dst_entries;
3381
	net->ipv6.ip6_null_entry->dst.path =
3382
		(struct dst_entry *)net->ipv6.ip6_null_entry;
3383
	net->ipv6.ip6_null_entry->dst.ops = &net->ipv6.ip6_dst_ops;
3384 3385
	dst_init_metrics(&net->ipv6.ip6_null_entry->dst,
			 ip6_template_metrics, true);
3386 3387 3388 3389 3390

#ifdef CONFIG_IPV6_MULTIPLE_TABLES
	net->ipv6.ip6_prohibit_entry = kmemdup(&ip6_prohibit_entry_template,
					       sizeof(*net->ipv6.ip6_prohibit_entry),
					       GFP_KERNEL);
3391 3392
	if (!net->ipv6.ip6_prohibit_entry)
		goto out_ip6_null_entry;
3393
	net->ipv6.ip6_prohibit_entry->dst.path =
3394
		(struct dst_entry *)net->ipv6.ip6_prohibit_entry;
3395
	net->ipv6.ip6_prohibit_entry->dst.ops = &net->ipv6.ip6_dst_ops;
3396 3397
	dst_init_metrics(&net->ipv6.ip6_prohibit_entry->dst,
			 ip6_template_metrics, true);
3398 3399 3400 3401

	net->ipv6.ip6_blk_hole_entry = kmemdup(&ip6_blk_hole_entry_template,
					       sizeof(*net->ipv6.ip6_blk_hole_entry),
					       GFP_KERNEL);
3402 3403
	if (!net->ipv6.ip6_blk_hole_entry)
		goto out_ip6_prohibit_entry;
3404
	net->ipv6.ip6_blk_hole_entry->dst.path =
3405
		(struct dst_entry *)net->ipv6.ip6_blk_hole_entry;
3406
	net->ipv6.ip6_blk_hole_entry->dst.ops = &net->ipv6.ip6_dst_ops;
3407 3408
	dst_init_metrics(&net->ipv6.ip6_blk_hole_entry->dst,
			 ip6_template_metrics, true);
3409 3410
#endif

3411 3412 3413 3414 3415 3416 3417 3418 3419
	net->ipv6.sysctl.flush_delay = 0;
	net->ipv6.sysctl.ip6_rt_max_size = 4096;
	net->ipv6.sysctl.ip6_rt_gc_min_interval = HZ / 2;
	net->ipv6.sysctl.ip6_rt_gc_timeout = 60*HZ;
	net->ipv6.sysctl.ip6_rt_gc_interval = 30*HZ;
	net->ipv6.sysctl.ip6_rt_gc_elasticity = 9;
	net->ipv6.sysctl.ip6_rt_mtu_expires = 10*60*HZ;
	net->ipv6.sysctl.ip6_rt_min_advmss = IPV6_MIN_MTU - 20 - 40;

3420 3421
	net->ipv6.ip6_rt_gc_expire = 30*HZ;

3422 3423 3424
	ret = 0;
out:
	return ret;
3425

3426 3427 3428 3429 3430 3431
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
out_ip6_prohibit_entry:
	kfree(net->ipv6.ip6_prohibit_entry);
out_ip6_null_entry:
	kfree(net->ipv6.ip6_null_entry);
#endif
3432 3433
out_ip6_dst_entries:
	dst_entries_destroy(&net->ipv6.ip6_dst_ops);
3434 3435
out_ip6_dst_ops:
	goto out;
3436 3437
}

3438
static void __net_exit ip6_route_net_exit(struct net *net)
3439
{
3440 3441 3442 3443 3444
	kfree(net->ipv6.ip6_null_entry);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
	kfree(net->ipv6.ip6_prohibit_entry);
	kfree(net->ipv6.ip6_blk_hole_entry);
#endif
3445
	dst_entries_destroy(&net->ipv6.ip6_dst_ops);
3446 3447
}

3448 3449 3450
static int __net_init ip6_route_net_init_late(struct net *net)
{
#ifdef CONFIG_PROC_FS
3451 3452
	proc_create("ipv6_route", 0, net->proc_net, &ipv6_route_proc_fops);
	proc_create("rt6_stats", S_IRUGO, net->proc_net, &rt6_stats_seq_fops);
3453 3454 3455 3456 3457 3458 3459
#endif
	return 0;
}

static void __net_exit ip6_route_net_exit_late(struct net *net)
{
#ifdef CONFIG_PROC_FS
3460 3461
	remove_proc_entry("ipv6_route", net->proc_net);
	remove_proc_entry("rt6_stats", net->proc_net);
3462 3463 3464
#endif
}

3465 3466 3467 3468 3469
static struct pernet_operations ip6_route_net_ops = {
	.init = ip6_route_net_init,
	.exit = ip6_route_net_exit,
};

3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485
static int __net_init ipv6_inetpeer_init(struct net *net)
{
	struct inet_peer_base *bp = kmalloc(sizeof(*bp), GFP_KERNEL);

	if (!bp)
		return -ENOMEM;
	inet_peer_base_init(bp);
	net->ipv6.peers = bp;
	return 0;
}

static void __net_exit ipv6_inetpeer_exit(struct net *net)
{
	struct inet_peer_base *bp = net->ipv6.peers;

	net->ipv6.peers = NULL;
3486
	inetpeer_invalidate_tree(bp);
3487 3488 3489
	kfree(bp);
}

3490
static struct pernet_operations ipv6_inetpeer_ops = {
3491 3492 3493 3494
	.init	=	ipv6_inetpeer_init,
	.exit	=	ipv6_inetpeer_exit,
};

3495 3496 3497 3498 3499
static struct pernet_operations ip6_route_net_late_ops = {
	.init = ip6_route_net_init_late,
	.exit = ip6_route_net_exit_late,
};

3500 3501 3502 3503 3504
static struct notifier_block ip6_route_dev_notifier = {
	.notifier_call = ip6_route_dev_notify,
	.priority = 0,
};

3505
int __init ip6_route_init(void)
L
Linus Torvalds 已提交
3506
{
3507
	int ret;
3508
	int cpu;
3509

3510 3511
	ret = -ENOMEM;
	ip6_dst_ops_template.kmem_cachep =
A
Alexey Dobriyan 已提交
3512
		kmem_cache_create("ip6_dst_cache", sizeof(struct rt6_info), 0,
3513
				  SLAB_HWCACHE_ALIGN, NULL);
3514
	if (!ip6_dst_ops_template.kmem_cachep)
3515
		goto out;
3516

3517
	ret = dst_entries_init(&ip6_dst_blackhole_ops);
3518
	if (ret)
3519 3520
		goto out_kmem_cache;

3521 3522
	ret = register_pernet_subsys(&ipv6_inetpeer_ops);
	if (ret)
3523
		goto out_dst_entries;
3524

3525 3526 3527
	ret = register_pernet_subsys(&ip6_route_net_ops);
	if (ret)
		goto out_register_inetpeer;
3528

3529 3530
	ip6_dst_blackhole_ops.kmem_cachep = ip6_dst_ops_template.kmem_cachep;

3531 3532 3533
	/* Registering of the loopback is done before this portion of code,
	 * the loopback reference in rt6_info will not be taken, do it
	 * manually for init_net */
3534
	init_net.ipv6.ip6_null_entry->dst.dev = init_net.loopback_dev;
3535 3536
	init_net.ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
  #ifdef CONFIG_IPV6_MULTIPLE_TABLES
3537
	init_net.ipv6.ip6_prohibit_entry->dst.dev = init_net.loopback_dev;
3538
	init_net.ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
3539
	init_net.ipv6.ip6_blk_hole_entry->dst.dev = init_net.loopback_dev;
3540 3541
	init_net.ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
  #endif
3542
	ret = fib6_init();
3543
	if (ret)
3544
		goto out_register_subsys;
3545 3546 3547

	ret = xfrm6_init();
	if (ret)
3548
		goto out_fib6_init;
3549

3550 3551 3552
	ret = fib6_rules_init();
	if (ret)
		goto xfrm6_init;
3553

3554 3555 3556 3557
	ret = register_pernet_subsys(&ip6_route_net_late_ops);
	if (ret)
		goto fib6_rules_init;

3558
	ret = -ENOBUFS;
3559 3560 3561
	if (__rtnl_register(PF_INET6, RTM_NEWROUTE, inet6_rtm_newroute, NULL, NULL) ||
	    __rtnl_register(PF_INET6, RTM_DELROUTE, inet6_rtm_delroute, NULL, NULL) ||
	    __rtnl_register(PF_INET6, RTM_GETROUTE, inet6_rtm_getroute, NULL, NULL))
3562
		goto out_register_late_subsys;
3563

3564
	ret = register_netdevice_notifier(&ip6_route_dev_notifier);
3565
	if (ret)
3566
		goto out_register_late_subsys;
3567

3568 3569 3570 3571 3572 3573 3574
	for_each_possible_cpu(cpu) {
		struct uncached_list *ul = per_cpu_ptr(&rt6_uncached_list, cpu);

		INIT_LIST_HEAD(&ul->head);
		spin_lock_init(&ul->lock);
	}

3575 3576 3577
out:
	return ret;

3578 3579
out_register_late_subsys:
	unregister_pernet_subsys(&ip6_route_net_late_ops);
3580 3581 3582 3583
fib6_rules_init:
	fib6_rules_cleanup();
xfrm6_init:
	xfrm6_fini();
3584 3585
out_fib6_init:
	fib6_gc_cleanup();
3586 3587
out_register_subsys:
	unregister_pernet_subsys(&ip6_route_net_ops);
3588 3589
out_register_inetpeer:
	unregister_pernet_subsys(&ipv6_inetpeer_ops);
3590 3591
out_dst_entries:
	dst_entries_destroy(&ip6_dst_blackhole_ops);
3592
out_kmem_cache:
3593
	kmem_cache_destroy(ip6_dst_ops_template.kmem_cachep);
3594
	goto out;
L
Linus Torvalds 已提交
3595 3596 3597 3598
}

void ip6_route_cleanup(void)
{
3599
	unregister_netdevice_notifier(&ip6_route_dev_notifier);
3600
	unregister_pernet_subsys(&ip6_route_net_late_ops);
T
Thomas Graf 已提交
3601
	fib6_rules_cleanup();
L
Linus Torvalds 已提交
3602 3603
	xfrm6_fini();
	fib6_gc_cleanup();
3604
	unregister_pernet_subsys(&ipv6_inetpeer_ops);
3605
	unregister_pernet_subsys(&ip6_route_net_ops);
3606
	dst_entries_destroy(&ip6_dst_blackhole_ops);
3607
	kmem_cache_destroy(ip6_dst_ops_template.kmem_cachep);
L
Linus Torvalds 已提交
3608
}