route.c 66.2 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 23 24
 *
 *	$Id: route.c,v 1.56 2001/10/31 21:55:55 davem Exp $
 *
 *	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.
25 26
 *	Ville Nuorvala
 *		Fixed routing subtrees.
L
Linus Torvalds 已提交
27 28
 */

29
#include <linux/capability.h>
L
Linus Torvalds 已提交
30 31 32 33 34 35 36 37 38
#include <linux/errno.h>
#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>
39
#include <linux/mroute6.h>
L
Linus Torvalds 已提交
40 41 42 43
#include <linux/init.h>
#include <linux/if_arp.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
44
#include <linux/nsproxy.h>
45
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
46 47 48 49 50 51 52 53 54 55
#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>
#include <net/xfrm.h>
56
#include <net/netevent.h>
57
#include <net/netlink.h>
L
Linus Torvalds 已提交
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75

#include <asm/uaccess.h>

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

/* Set to 3 to get tracing. */
#define RT6_DEBUG 2

#if RT6_DEBUG >= 3
#define RDBG(x) printk x
#define RT6_TRACE(x...) printk(KERN_DEBUG x)
#else
#define RDBG(x)
#define RT6_TRACE(x...) do { ; } while (0)
#endif

76
#define CLONE_OFFLINK_ROUTE 0
L
Linus Torvalds 已提交
77 78 79 80 81 82 83

static struct rt6_info * ip6_rt_copy(struct rt6_info *ort);
static struct dst_entry	*ip6_dst_check(struct dst_entry *dst, u32 cookie);
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);
84
static int		 ip6_dst_gc(struct dst_ops *ops);
L
Linus Torvalds 已提交
85 86 87 88 89 90

static int		ip6_pkt_discard(struct sk_buff *skb);
static int		ip6_pkt_discard_out(struct sk_buff *skb);
static void		ip6_link_failure(struct sk_buff *skb);
static void		ip6_rt_update_pmtu(struct dst_entry *dst, u32 mtu);

91
#ifdef CONFIG_IPV6_ROUTE_INFO
92 93
static struct rt6_info *rt6_add_route_info(struct net *net,
					   struct in6_addr *prefix, int prefixlen,
94 95
					   struct in6_addr *gwaddr, int ifindex,
					   unsigned pref);
96 97
static struct rt6_info *rt6_get_route_info(struct net *net,
					   struct in6_addr *prefix, int prefixlen,
98 99 100
					   struct in6_addr *gwaddr, int ifindex);
#endif

101
static struct dst_ops ip6_dst_ops_template = {
L
Linus Torvalds 已提交
102 103 104 105 106 107 108 109 110 111
	.family			=	AF_INET6,
	.protocol		=	__constant_htons(ETH_P_IPV6),
	.gc			=	ip6_dst_gc,
	.gc_thresh		=	1024,
	.check			=	ip6_dst_check,
	.destroy		=	ip6_dst_destroy,
	.ifdown			=	ip6_dst_ifdown,
	.negative_advice	=	ip6_negative_advice,
	.link_failure		=	ip6_link_failure,
	.update_pmtu		=	ip6_rt_update_pmtu,
112
	.local_out		=	__ip6_local_out,
L
Linus Torvalds 已提交
113
	.entry_size		=	sizeof(struct rt6_info),
114
	.entries		=	ATOMIC_INIT(0),
L
Linus Torvalds 已提交
115 116
};

117 118 119 120 121 122 123 124 125 126 127
static void ip6_rt_blackhole_update_pmtu(struct dst_entry *dst, u32 mtu)
{
}

static struct dst_ops ip6_dst_blackhole_ops = {
	.family			=	AF_INET6,
	.protocol		=	__constant_htons(ETH_P_IPV6),
	.destroy		=	ip6_dst_destroy,
	.check			=	ip6_dst_check,
	.update_pmtu		=	ip6_rt_blackhole_update_pmtu,
	.entry_size		=	sizeof(struct rt6_info),
128
	.entries		=	ATOMIC_INIT(0),
129 130
};

131
static struct rt6_info ip6_null_entry_template = {
L
Linus Torvalds 已提交
132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
	.u = {
		.dst = {
			.__refcnt	= ATOMIC_INIT(1),
			.__use		= 1,
			.obsolete	= -1,
			.error		= -ENETUNREACH,
			.metrics	= { [RTAX_HOPLIMIT - 1] = 255, },
			.input		= ip6_pkt_discard,
			.output		= ip6_pkt_discard_out,
		}
	},
	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
	.rt6i_metric	= ~(u32) 0,
	.rt6i_ref	= ATOMIC_INIT(1),
};

T
Thomas Graf 已提交
148 149
#ifdef CONFIG_IPV6_MULTIPLE_TABLES

150 151 152
static int ip6_pkt_prohibit(struct sk_buff *skb);
static int ip6_pkt_prohibit_out(struct sk_buff *skb);

153
static struct rt6_info ip6_prohibit_entry_template = {
T
Thomas Graf 已提交
154 155 156 157 158 159 160
	.u = {
		.dst = {
			.__refcnt	= ATOMIC_INIT(1),
			.__use		= 1,
			.obsolete	= -1,
			.error		= -EACCES,
			.metrics	= { [RTAX_HOPLIMIT - 1] = 255, },
161 162
			.input		= ip6_pkt_prohibit,
			.output		= ip6_pkt_prohibit_out,
T
Thomas Graf 已提交
163 164 165 166 167 168 169
		}
	},
	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
	.rt6i_metric	= ~(u32) 0,
	.rt6i_ref	= ATOMIC_INIT(1),
};

170
static struct rt6_info ip6_blk_hole_entry_template = {
T
Thomas Graf 已提交
171 172 173 174 175 176 177
	.u = {
		.dst = {
			.__refcnt	= ATOMIC_INIT(1),
			.__use		= 1,
			.obsolete	= -1,
			.error		= -EINVAL,
			.metrics	= { [RTAX_HOPLIMIT - 1] = 255, },
178 179
			.input		= dst_discard,
			.output		= dst_discard,
T
Thomas Graf 已提交
180 181 182 183 184 185 186 187 188
		}
	},
	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
	.rt6i_metric	= ~(u32) 0,
	.rt6i_ref	= ATOMIC_INIT(1),
};

#endif

L
Linus Torvalds 已提交
189
/* allocate dst with ip6_dst_ops */
190
static inline struct rt6_info *ip6_dst_alloc(struct dst_ops *ops)
L
Linus Torvalds 已提交
191
{
192
	return (struct rt6_info *)dst_alloc(ops);
L
Linus Torvalds 已提交
193 194 195 196 197 198 199 200 201 202
}

static void ip6_dst_destroy(struct dst_entry *dst)
{
	struct rt6_info *rt = (struct rt6_info *)dst;
	struct inet6_dev *idev = rt->rt6i_idev;

	if (idev != NULL) {
		rt->rt6i_idev = NULL;
		in6_dev_put(idev);
203
	}
L
Linus Torvalds 已提交
204 205 206 207 208 209 210
}

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;
211
	struct net_device *loopback_dev =
212
		dev_net(dev)->loopback_dev;
L
Linus Torvalds 已提交
213

214 215 216
	if (dev != loopback_dev && idev != NULL && idev->dev == dev) {
		struct inet6_dev *loopback_idev =
			in6_dev_get(loopback_dev);
L
Linus Torvalds 已提交
217 218 219 220 221 222 223 224 225 226 227 228 229
		if (loopback_idev != NULL) {
			rt->rt6i_idev = loopback_idev;
			in6_dev_put(idev);
		}
	}
}

static __inline__ int rt6_check_expired(const struct rt6_info *rt)
{
	return (rt->rt6i_flags & RTF_EXPIRES &&
		time_after(jiffies, rt->rt6i_expires));
}

T
Thomas Graf 已提交
230 231 232 233 234 235
static inline int rt6_need_strict(struct in6_addr *daddr)
{
	return (ipv6_addr_type(daddr) &
		(IPV6_ADDR_MULTICAST | IPV6_ADDR_LINKLOCAL));
}

L
Linus Torvalds 已提交
236
/*
T
Thomas Graf 已提交
237
 *	Route lookup. Any table->tb6_lock is implied.
L
Linus Torvalds 已提交
238 239
 */

240 241
static inline struct rt6_info *rt6_device_match(struct net *net,
						    struct rt6_info *rt,
L
Linus Torvalds 已提交
242 243 244 245 246 247 248
						    int oif,
						    int strict)
{
	struct rt6_info *local = NULL;
	struct rt6_info *sprt;

	if (oif) {
249
		for (sprt = rt; sprt; sprt = sprt->u.dst.rt6_next) {
L
Linus Torvalds 已提交
250 251 252 253 254 255 256 257
			struct net_device *dev = sprt->rt6i_dev;
			if (dev->ifindex == oif)
				return sprt;
			if (dev->flags & IFF_LOOPBACK) {
				if (sprt->rt6i_idev == NULL ||
				    sprt->rt6i_idev->dev->ifindex != oif) {
					if (strict && oif)
						continue;
258
					if (local && (!oif ||
L
Linus Torvalds 已提交
259 260 261 262 263 264 265 266 267 268 269
						      local->rt6i_idev->dev->ifindex == oif))
						continue;
				}
				local = sprt;
			}
		}

		if (local)
			return local;

		if (strict)
270
			return net->ipv6.ip6_null_entry;
L
Linus Torvalds 已提交
271 272 273 274
	}
	return rt;
}

275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290
#ifdef CONFIG_IPV6_ROUTER_PREF
static void rt6_probe(struct rt6_info *rt)
{
	struct neighbour *neigh = rt ? rt->rt6i_nexthop : NULL;
	/*
	 * 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.
	 */
	if (!neigh || (neigh->nud_state & NUD_VALID))
		return;
	read_lock_bh(&neigh->lock);
	if (!(neigh->nud_state & NUD_VALID) &&
291
	    time_after(jiffies, neigh->updated + rt->rt6i_idev->cnf.rtr_probe_interval)) {
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
		struct in6_addr mcaddr;
		struct in6_addr *target;

		neigh->updated = jiffies;
		read_unlock_bh(&neigh->lock);

		target = (struct in6_addr *)&neigh->primary_key;
		addrconf_addr_solict_mult(target, &mcaddr);
		ndisc_send_ns(rt->rt6i_dev, NULL, target, &mcaddr, NULL);
	} else
		read_unlock_bh(&neigh->lock);
}
#else
static inline void rt6_probe(struct rt6_info *rt)
{
	return;
}
#endif

L
Linus Torvalds 已提交
311
/*
312
 * Default Router Selection (RFC 2461 6.3.6)
L
Linus Torvalds 已提交
313
 */
D
Dave Jones 已提交
314
static inline int rt6_check_dev(struct rt6_info *rt, int oif)
315 316
{
	struct net_device *dev = rt->rt6i_dev;
317
	if (!oif || dev->ifindex == oif)
318
		return 2;
319 320 321 322
	if ((dev->flags & IFF_LOOPBACK) &&
	    rt->rt6i_idev && rt->rt6i_idev->dev->ifindex == oif)
		return 1;
	return 0;
323
}
L
Linus Torvalds 已提交
324

D
Dave Jones 已提交
325
static inline int rt6_check_neigh(struct rt6_info *rt)
L
Linus Torvalds 已提交
326
{
327
	struct neighbour *neigh = rt->rt6i_nexthop;
328
	int m;
329 330 331 332
	if (rt->rt6i_flags & RTF_NONEXTHOP ||
	    !(rt->rt6i_flags & RTF_GATEWAY))
		m = 1;
	else if (neigh) {
333 334
		read_lock_bh(&neigh->lock);
		if (neigh->nud_state & NUD_VALID)
335
			m = 2;
336 337 338 339 340
#ifdef CONFIG_IPV6_ROUTER_PREF
		else if (neigh->nud_state & NUD_FAILED)
			m = 0;
#endif
		else
341
			m = 1;
342
		read_unlock_bh(&neigh->lock);
343 344
	} else
		m = 0;
345
	return m;
L
Linus Torvalds 已提交
346 347
}

348 349
static int rt6_score_route(struct rt6_info *rt, int oif,
			   int strict)
L
Linus Torvalds 已提交
350
{
351
	int m, n;
352

353
	m = rt6_check_dev(rt, oif);
354
	if (!m && (strict & RT6_LOOKUP_F_IFACE))
355
		return -1;
356 357 358
#ifdef CONFIG_IPV6_ROUTER_PREF
	m |= IPV6_DECODE_PREF(IPV6_EXTRACT_PREF(rt->rt6i_flags)) << 2;
#endif
359
	n = rt6_check_neigh(rt);
360
	if (!n && (strict & RT6_LOOKUP_F_REACHABLE))
361 362 363 364
		return -1;
	return m;
}

365 366
static struct rt6_info *find_match(struct rt6_info *rt, int oif, int strict,
				   int *mpri, struct rt6_info *match)
367
{
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
	int m;

	if (rt6_check_expired(rt))
		goto out;

	m = rt6_score_route(rt, oif, strict);
	if (m < 0)
		goto out;

	if (m > *mpri) {
		if (strict & RT6_LOOKUP_F_REACHABLE)
			rt6_probe(match);
		*mpri = m;
		match = rt;
	} else if (strict & RT6_LOOKUP_F_REACHABLE) {
		rt6_probe(rt);
	}

out:
	return match;
}

static struct rt6_info *find_rr_leaf(struct fib6_node *fn,
				     struct rt6_info *rr_head,
				     u32 metric, int oif, int strict)
{
	struct rt6_info *rt, *match;
395
	int mpri = -1;
L
Linus Torvalds 已提交
396

397 398 399 400 401 402 403
	match = NULL;
	for (rt = rr_head; rt && rt->rt6i_metric == metric;
	     rt = rt->u.dst.rt6_next)
		match = find_match(rt, oif, strict, &mpri, match);
	for (rt = fn->leaf; rt && rt != rr_head && rt->rt6i_metric == metric;
	     rt = rt->u.dst.rt6_next)
		match = find_match(rt, oif, strict, &mpri, match);
L
Linus Torvalds 已提交
404

405 406
	return match;
}
L
Linus Torvalds 已提交
407

408 409 410
static struct rt6_info *rt6_select(struct fib6_node *fn, int oif, int strict)
{
	struct rt6_info *match, *rt0;
411
	struct net *net;
L
Linus Torvalds 已提交
412

413
	RT6_TRACE("%s(fn->leaf=%p, oif=%d)\n",
414
		  __func__, fn->leaf, oif);
415

416 417 418
	rt0 = fn->rr_ptr;
	if (!rt0)
		fn->rr_ptr = rt0 = fn->leaf;
L
Linus Torvalds 已提交
419

420
	match = find_rr_leaf(fn, rt0, rt0->rt6i_metric, oif, strict);
L
Linus Torvalds 已提交
421

422
	if (!match &&
423 424 425
	    (strict & RT6_LOOKUP_F_REACHABLE)) {
		struct rt6_info *next = rt0->u.dst.rt6_next;

426
		/* no entries matched; do round-robin */
427 428 429 430 431
		if (!next || next->rt6i_metric != rt0->rt6i_metric)
			next = fn->leaf;

		if (next != rt0)
			fn->rr_ptr = next;
L
Linus Torvalds 已提交
432 433
	}

434
	RT6_TRACE("%s() => %p\n",
435
		  __func__, match);
L
Linus Torvalds 已提交
436

437
	net = dev_net(rt0->rt6i_dev);
438
	return (match ? match : net->ipv6.ip6_null_entry);
L
Linus Torvalds 已提交
439 440
}

441 442 443 444
#ifdef CONFIG_IPV6_ROUTE_INFO
int rt6_route_rcv(struct net_device *dev, u8 *opt, int len,
		  struct in6_addr *gwaddr)
{
445
	struct net *net = dev_net(dev);
446 447 448
	struct route_info *rinfo = (struct route_info *) opt;
	struct in6_addr prefix_buf, *prefix;
	unsigned int pref;
449
	unsigned long lifetime;
450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474
	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)
		pref = ICMPV6_ROUTER_PREF_MEDIUM;

475
	lifetime = addrconf_timeout_fixup(ntohl(rinfo->lifetime), HZ);
476 477 478 479 480 481 482 483 484 485 486

	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;
	}

487 488
	rt = rt6_get_route_info(net, prefix, rinfo->prefix_len, gwaddr,
				dev->ifindex);
489 490

	if (rt && !lifetime) {
491
		ip6_del_rt(rt);
492 493 494 495
		rt = NULL;
	}

	if (!rt && lifetime)
496
		rt = rt6_add_route_info(net, prefix, rinfo->prefix_len, gwaddr, dev->ifindex,
497 498 499 500 501 502
					pref);
	else if (rt)
		rt->rt6i_flags = RTF_ROUTEINFO |
				 (rt->rt6i_flags & ~RTF_PREF_MASK) | RTF_PREF(pref);

	if (rt) {
503
		if (!addrconf_finite_timeout(lifetime)) {
504 505 506 507 508 509 510 511 512 513 514
			rt->rt6i_flags &= ~RTF_EXPIRES;
		} else {
			rt->rt6i_expires = jiffies + HZ * lifetime;
			rt->rt6i_flags |= RTF_EXPIRES;
		}
		dst_release(&rt->u.dst);
	}
	return 0;
}
#endif

515
#define BACKTRACK(__net, saddr)			\
516
do { \
517
	if (rt == __net->ipv6.ip6_null_entry) {	\
518
		struct fib6_node *pn; \
V
Ville Nuorvala 已提交
519
		while (1) { \
520 521 522 523
			if (fn->fn_flags & RTN_TL_ROOT) \
				goto out; \
			pn = fn->parent; \
			if (FIB6_SUBTREE(pn) && FIB6_SUBTREE(pn) != fn) \
524
				fn = fib6_lookup(FIB6_SUBTREE(pn), NULL, saddr); \
525 526 527 528
			else \
				fn = pn; \
			if (fn->fn_flags & RTN_RTINFO) \
				goto restart; \
T
Thomas Graf 已提交
529 530
		} \
	} \
531
} while(0)
T
Thomas Graf 已提交
532

533 534
static struct rt6_info *ip6_pol_route_lookup(struct net *net,
					     struct fib6_table *table,
T
Thomas Graf 已提交
535
					     struct flowi *fl, int flags)
L
Linus Torvalds 已提交
536 537 538 539
{
	struct fib6_node *fn;
	struct rt6_info *rt;

T
Thomas Graf 已提交
540 541 542 543
	read_lock_bh(&table->tb6_lock);
	fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src);
restart:
	rt = fn->leaf;
544 545
	rt = rt6_device_match(net, rt, fl->oif, flags);
	BACKTRACK(net, &fl->fl6_src);
T
Thomas Graf 已提交
546
out:
547
	dst_use(&rt->u.dst, jiffies);
T
Thomas Graf 已提交
548 549 550 551 552
	read_unlock_bh(&table->tb6_lock);
	return rt;

}

553 554
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
555 556 557 558 559 560 561 562 563 564
{
	struct flowi fl = {
		.oif = oif,
		.nl_u = {
			.ip6_u = {
				.daddr = *daddr,
			},
		},
	};
	struct dst_entry *dst;
565
	int flags = strict ? RT6_LOOKUP_F_IFACE : 0;
T
Thomas Graf 已提交
566

567 568 569 570 571
	if (saddr) {
		memcpy(&fl.fl6_src, saddr, sizeof(*saddr));
		flags |= RT6_LOOKUP_F_HAS_SADDR;
	}

572
	dst = fib6_rule_lookup(net, &fl, flags, ip6_pol_route_lookup);
T
Thomas Graf 已提交
573 574 575 576 577
	if (dst->error == 0)
		return (struct rt6_info *) dst;

	dst_release(dst);

L
Linus Torvalds 已提交
578 579 580
	return NULL;
}

581 582
EXPORT_SYMBOL(rt6_lookup);

T
Thomas Graf 已提交
583
/* ip6_ins_rt is called with FREE table->tb6_lock.
L
Linus Torvalds 已提交
584 585 586 587 588
   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.
 */

589
static int __ip6_ins_rt(struct rt6_info *rt, struct nl_info *info)
L
Linus Torvalds 已提交
590 591
{
	int err;
T
Thomas Graf 已提交
592
	struct fib6_table *table;
L
Linus Torvalds 已提交
593

T
Thomas Graf 已提交
594 595
	table = rt->rt6i_table;
	write_lock_bh(&table->tb6_lock);
596
	err = fib6_add(&table->tb6_root, rt, info);
T
Thomas Graf 已提交
597
	write_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
598 599 600 601

	return err;
}

602 603
int ip6_ins_rt(struct rt6_info *rt)
{
604
	struct nl_info info = {
605
		.nl_net = dev_net(rt->rt6i_dev),
606
	};
607
	return __ip6_ins_rt(rt, &info);
608 609
}

610 611
static struct rt6_info *rt6_alloc_cow(struct rt6_info *ort, struct in6_addr *daddr,
				      struct in6_addr *saddr)
L
Linus Torvalds 已提交
612 613 614 615 616 617 618 619 620 621
{
	struct rt6_info *rt;

	/*
	 *	Clone the route.
	 */

	rt = ip6_rt_copy(ort);

	if (rt) {
622 623 624 625
		if (!(rt->rt6i_flags&RTF_GATEWAY)) {
			if (rt->rt6i_dst.plen != 128 &&
			    ipv6_addr_equal(&rt->rt6i_dst.addr, daddr))
				rt->rt6i_flags |= RTF_ANYCAST;
L
Linus Torvalds 已提交
626
			ipv6_addr_copy(&rt->rt6i_gateway, daddr);
627
		}
L
Linus Torvalds 已提交
628

629
		ipv6_addr_copy(&rt->rt6i_dst.addr, daddr);
L
Linus Torvalds 已提交
630 631 632 633 634 635 636 637 638 639 640 641 642
		rt->rt6i_dst.plen = 128;
		rt->rt6i_flags |= RTF_CACHE;
		rt->u.dst.flags |= DST_HOST;

#ifdef CONFIG_IPV6_SUBTREES
		if (rt->rt6i_src.plen && saddr) {
			ipv6_addr_copy(&rt->rt6i_src.addr, saddr);
			rt->rt6i_src.plen = 128;
		}
#endif

		rt->rt6i_nexthop = ndisc_get_neigh(rt->rt6i_dev, &rt->rt6i_gateway);

643
	}
L
Linus Torvalds 已提交
644

645 646
	return rt;
}
L
Linus Torvalds 已提交
647

648 649 650 651 652 653 654 655 656 657 658 659 660
static struct rt6_info *rt6_alloc_clone(struct rt6_info *ort, struct in6_addr *daddr)
{
	struct rt6_info *rt = ip6_rt_copy(ort);
	if (rt) {
		ipv6_addr_copy(&rt->rt6i_dst.addr, daddr);
		rt->rt6i_dst.plen = 128;
		rt->rt6i_flags |= RTF_CACHE;
		rt->u.dst.flags |= DST_HOST;
		rt->rt6i_nexthop = neigh_clone(ort->rt6i_nexthop);
	}
	return rt;
}

661 662
static struct rt6_info *ip6_pol_route(struct net *net, struct fib6_table *table, int oif,
				      struct flowi *fl, int flags)
L
Linus Torvalds 已提交
663 664
{
	struct fib6_node *fn;
665
	struct rt6_info *rt, *nrt;
T
Thomas Graf 已提交
666
	int strict = 0;
L
Linus Torvalds 已提交
667
	int attempts = 3;
668
	int err;
669
	int reachable = ipv6_devconf.forwarding ? 0 : RT6_LOOKUP_F_REACHABLE;
L
Linus Torvalds 已提交
670

671
	strict |= flags & RT6_LOOKUP_F_IFACE;
L
Linus Torvalds 已提交
672 673

relookup:
T
Thomas Graf 已提交
674
	read_lock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
675

676
restart_2:
T
Thomas Graf 已提交
677
	fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src);
L
Linus Torvalds 已提交
678 679

restart:
680
	rt = rt6_select(fn, oif, strict | reachable);
681 682 683

	BACKTRACK(net, &fl->fl6_src);
	if (rt == net->ipv6.ip6_null_entry ||
684
	    rt->rt6i_flags & RTF_CACHE)
685
		goto out;
L
Linus Torvalds 已提交
686

687
	dst_hold(&rt->u.dst);
T
Thomas Graf 已提交
688
	read_unlock_bh(&table->tb6_lock);
689

690
	if (!rt->rt6i_nexthop && !(rt->rt6i_flags & RTF_NONEXTHOP))
T
Thomas Graf 已提交
691
		nrt = rt6_alloc_cow(rt, &fl->fl6_dst, &fl->fl6_src);
692 693
	else {
#if CLONE_OFFLINK_ROUTE
T
Thomas Graf 已提交
694
		nrt = rt6_alloc_clone(rt, &fl->fl6_dst);
695 696 697 698
#else
		goto out2;
#endif
	}
699

700
	dst_release(&rt->u.dst);
701
	rt = nrt ? : net->ipv6.ip6_null_entry;
L
Linus Torvalds 已提交
702

703 704
	dst_hold(&rt->u.dst);
	if (nrt) {
705
		err = ip6_ins_rt(nrt);
706
		if (!err)
L
Linus Torvalds 已提交
707 708 709
			goto out2;
	}

710 711 712 713
	if (--attempts <= 0)
		goto out2;

	/*
T
Thomas Graf 已提交
714
	 * Race condition! In the gap, when table->tb6_lock was
715 716 717 718 719 720
	 * released someone could insert this route.  Relookup.
	 */
	dst_release(&rt->u.dst);
	goto relookup;

out:
721 722 723 724
	if (reachable) {
		reachable = 0;
		goto restart_2;
	}
725
	dst_hold(&rt->u.dst);
T
Thomas Graf 已提交
726
	read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
727 728 729
out2:
	rt->u.dst.lastuse = jiffies;
	rt->u.dst.__use++;
T
Thomas Graf 已提交
730 731

	return rt;
L
Linus Torvalds 已提交
732 733
}

734
static struct rt6_info *ip6_pol_route_input(struct net *net, struct fib6_table *table,
735 736
					    struct flowi *fl, int flags)
{
737
	return ip6_pol_route(net, table, fl->iif, fl, flags);
738 739
}

T
Thomas Graf 已提交
740 741
void ip6_route_input(struct sk_buff *skb)
{
742
	struct ipv6hdr *iph = ipv6_hdr(skb);
743
	struct net *net = dev_net(skb->dev);
744
	int flags = RT6_LOOKUP_F_HAS_SADDR;
T
Thomas Graf 已提交
745 746 747 748 749 750
	struct flowi fl = {
		.iif = skb->dev->ifindex,
		.nl_u = {
			.ip6_u = {
				.daddr = iph->daddr,
				.saddr = iph->saddr,
A
Al Viro 已提交
751
				.flowlabel = (* (__be32 *) iph)&IPV6_FLOWINFO_MASK,
T
Thomas Graf 已提交
752 753
			},
		},
754
		.mark = skb->mark,
T
Thomas Graf 已提交
755 756
		.proto = iph->nexthdr,
	};
757 758 759

	if (rt6_need_strict(&iph->daddr))
		flags |= RT6_LOOKUP_F_IFACE;
T
Thomas Graf 已提交
760

761
	skb->dst = fib6_rule_lookup(net, &fl, flags, ip6_pol_route_input);
T
Thomas Graf 已提交
762 763
}

764
static struct rt6_info *ip6_pol_route_output(struct net *net, struct fib6_table *table,
T
Thomas Graf 已提交
765
					     struct flowi *fl, int flags)
L
Linus Torvalds 已提交
766
{
767
	return ip6_pol_route(net, table, fl->oif, fl, flags);
T
Thomas Graf 已提交
768 769
}

770 771
struct dst_entry * ip6_route_output(struct net *net, struct sock *sk,
				    struct flowi *fl)
T
Thomas Graf 已提交
772 773 774 775
{
	int flags = 0;

	if (rt6_need_strict(&fl->fl6_dst))
776
		flags |= RT6_LOOKUP_F_IFACE;
T
Thomas Graf 已提交
777

778 779
	if (!ipv6_addr_any(&fl->fl6_src))
		flags |= RT6_LOOKUP_F_HAS_SADDR;
780 781 782 783 784 785 786 787 788
	else if (sk) {
		unsigned int prefs = inet6_sk(sk)->srcprefs;
		if (prefs & IPV6_PREFER_SRC_TMP)
			flags |= RT6_LOOKUP_F_SRCPREF_TMP;
		if (prefs & IPV6_PREFER_SRC_PUBLIC)
			flags |= RT6_LOOKUP_F_SRCPREF_PUBLIC;
		if (prefs & IPV6_PREFER_SRC_COA)
			flags |= RT6_LOOKUP_F_SRCPREF_COA;
	}
789

790
	return fib6_rule_lookup(net, fl, flags, ip6_pol_route_output);
L
Linus Torvalds 已提交
791 792
}

793
EXPORT_SYMBOL(ip6_route_output);
L
Linus Torvalds 已提交
794

795 796 797 798 799 800 801 802 803 804 805 806
int ip6_dst_blackhole(struct sock *sk, struct dst_entry **dstp, struct flowi *fl)
{
	struct rt6_info *ort = (struct rt6_info *) *dstp;
	struct rt6_info *rt = (struct rt6_info *)
		dst_alloc(&ip6_dst_blackhole_ops);
	struct dst_entry *new = NULL;

	if (rt) {
		new = &rt->u.dst;

		atomic_set(&new->__refcnt, 1);
		new->__use = 1;
807 808
		new->input = dst_discard;
		new->output = dst_discard;
809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836

		memcpy(new->metrics, ort->u.dst.metrics, RTAX_MAX*sizeof(u32));
		new->dev = ort->u.dst.dev;
		if (new->dev)
			dev_hold(new->dev);
		rt->rt6i_idev = ort->rt6i_idev;
		if (rt->rt6i_idev)
			in6_dev_hold(rt->rt6i_idev);
		rt->rt6i_expires = 0;

		ipv6_addr_copy(&rt->rt6i_gateway, &ort->rt6i_gateway);
		rt->rt6i_flags = ort->rt6i_flags & ~RTF_EXPIRES;
		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);
	}

	dst_release(*dstp);
	*dstp = new;
	return (new ? 0 : -ENOMEM);
}
EXPORT_SYMBOL_GPL(ip6_dst_blackhole);

L
Linus Torvalds 已提交
837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858
/*
 *	Destination cache support functions
 */

static struct dst_entry *ip6_dst_check(struct dst_entry *dst, u32 cookie)
{
	struct rt6_info *rt;

	rt = (struct rt6_info *) dst;

	if (rt && rt->rt6i_node && (rt->rt6i_node->fn_sernum == cookie))
		return dst;

	return NULL;
}

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

	if (rt) {
		if (rt->rt6i_flags & RTF_CACHE)
859
			ip6_del_rt(rt);
L
Linus Torvalds 已提交
860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892
		else
			dst_release(dst);
	}
	return NULL;
}

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

	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_ADDR_UNREACH, 0, skb->dev);

	rt = (struct rt6_info *) skb->dst;
	if (rt) {
		if (rt->rt6i_flags&RTF_CACHE) {
			dst_set_expires(&rt->u.dst, 0);
			rt->rt6i_flags |= RTF_EXPIRES;
		} else if (rt->rt6i_node && (rt->rt6i_flags & RTF_DEFAULT))
			rt->rt6i_node->fn_sernum = -1;
	}
}

static void ip6_rt_update_pmtu(struct dst_entry *dst, u32 mtu)
{
	struct rt6_info *rt6 = (struct rt6_info*)dst;

	if (mtu < dst_mtu(dst) && rt6->rt6i_dst.plen == 128) {
		rt6->rt6i_flags |= RTF_MODIFIED;
		if (mtu < IPV6_MIN_MTU) {
			mtu = IPV6_MIN_MTU;
			dst->metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
		}
		dst->metrics[RTAX_MTU-1] = mtu;
893
		call_netevent_notifiers(NETEVENT_PMTU_UPDATE, dst);
L
Linus Torvalds 已提交
894 895 896 897 898
	}
}

static int ipv6_get_mtu(struct net_device *dev);

899
static inline unsigned int ipv6_advmss(struct net *net, unsigned int mtu)
L
Linus Torvalds 已提交
900 901 902
{
	mtu -= sizeof(struct ipv6hdr) + sizeof(struct tcphdr);

903 904
	if (mtu < net->ipv6.sysctl.ip6_rt_min_advmss)
		mtu = net->ipv6.sysctl.ip6_rt_min_advmss;
L
Linus Torvalds 已提交
905 906

	/*
907 908 909
	 * 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
910 911 912 913 914 915 916
	 * rely only on pmtu discovery"
	 */
	if (mtu > IPV6_MAXPLEN - sizeof(struct tcphdr))
		mtu = IPV6_MAXPLEN;
	return mtu;
}

917 918
static struct dst_entry *icmp6_dst_gc_list;
static DEFINE_SPINLOCK(icmp6_dst_lock);
919

920
struct dst_entry *icmp6_dst_alloc(struct net_device *dev,
L
Linus Torvalds 已提交
921
				  struct neighbour *neigh,
922
				  const struct in6_addr *addr)
L
Linus Torvalds 已提交
923 924 925
{
	struct rt6_info *rt;
	struct inet6_dev *idev = in6_dev_get(dev);
926
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
927 928 929 930

	if (unlikely(idev == NULL))
		return NULL;

931
	rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops);
L
Linus Torvalds 已提交
932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948
	if (unlikely(rt == NULL)) {
		in6_dev_put(idev);
		goto out;
	}

	dev_hold(dev);
	if (neigh)
		neigh_hold(neigh);
	else
		neigh = ndisc_get_neigh(dev, addr);

	rt->rt6i_dev	  = dev;
	rt->rt6i_idev     = idev;
	rt->rt6i_nexthop  = neigh;
	atomic_set(&rt->u.dst.__refcnt, 1);
	rt->u.dst.metrics[RTAX_HOPLIMIT-1] = 255;
	rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(rt->rt6i_dev);
949
	rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->u.dst));
950
	rt->u.dst.output  = ip6_output;
L
Linus Torvalds 已提交
951 952

#if 0	/* there's no chance to use these for ndisc */
953 954
	rt->u.dst.flags   = ipv6_addr_type(addr) & IPV6_ADDR_UNICAST
				? DST_HOST
L
Linus Torvalds 已提交
955 956 957 958 959
				: 0;
	ipv6_addr_copy(&rt->rt6i_dst.addr, addr);
	rt->rt6i_dst.plen = 128;
#endif

960 961 962 963
	spin_lock_bh(&icmp6_dst_lock);
	rt->u.dst.next = icmp6_dst_gc_list;
	icmp6_dst_gc_list = &rt->u.dst;
	spin_unlock_bh(&icmp6_dst_lock);
L
Linus Torvalds 已提交
964

965
	fib6_force_start_gc(net);
L
Linus Torvalds 已提交
966 967

out:
968
	return &rt->u.dst;
L
Linus Torvalds 已提交
969 970
}

971
int icmp6_dst_gc(int *more)
L
Linus Torvalds 已提交
972 973 974 975 976
{
	struct dst_entry *dst, *next, **pprev;
	int freed;

	next = NULL;
977
	freed = 0;
978

979 980
	spin_lock_bh(&icmp6_dst_lock);
	pprev = &icmp6_dst_gc_list;
981

L
Linus Torvalds 已提交
982 983 984 985 986 987 988 989 990 991 992
	while ((dst = *pprev) != NULL) {
		if (!atomic_read(&dst->__refcnt)) {
			*pprev = dst->next;
			dst_free(dst);
			freed++;
		} else {
			pprev = &dst->next;
			(*more)++;
		}
	}

993
	spin_unlock_bh(&icmp6_dst_lock);
994

L
Linus Torvalds 已提交
995 996 997
	return freed;
}

998
static int ip6_dst_gc(struct dst_ops *ops)
L
Linus Torvalds 已提交
999 1000
{
	unsigned long now = jiffies;
1001 1002 1003 1004 1005 1006 1007 1008 1009
	struct net *net = ops->dst_net;
	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;

	if (time_after(rt_last_gc + rt_min_interval, now) &&
	    atomic_read(&ops->entries) <= rt_max_size)
L
Linus Torvalds 已提交
1010 1011
		goto out;

1012 1013 1014
	net->ipv6.ip6_rt_gc_expire++;
	fib6_run_gc(net->ipv6.ip6_rt_gc_expire, net);
	net->ipv6.ip6_rt_last_gc = now;
1015 1016
	if (atomic_read(&ops->entries) < ops->gc_thresh)
		net->ipv6.ip6_rt_gc_expire = rt_gc_timeout>>1;
L
Linus Torvalds 已提交
1017
out:
1018 1019
	net->ipv6.ip6_rt_gc_expire -= net->ipv6.ip6_rt_gc_expire>>rt_elasticity;
	return (atomic_read(&ops->entries) > rt_max_size);
L
Linus Torvalds 已提交
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
}

/* Clean host part of a prefix. Not necessary in radix tree,
   but results in cleaner routing tables.

   Remove it only when all the things will work!
 */

static int ipv6_get_mtu(struct net_device *dev)
{
	int mtu = IPV6_MIN_MTU;
	struct inet6_dev *idev;

	idev = in6_dev_get(dev);
	if (idev) {
		mtu = idev->cnf.mtu6;
		in6_dev_put(idev);
	}
	return mtu;
}

1041
int ip6_dst_hoplimit(struct dst_entry *dst)
L
Linus Torvalds 已提交
1042
{
1043 1044 1045 1046 1047 1048 1049 1050 1051
	int hoplimit = dst_metric(dst, RTAX_HOPLIMIT);
	if (hoplimit < 0) {
		struct net_device *dev = dst->dev;
		struct inet6_dev *idev = in6_dev_get(dev);
		if (idev) {
			hoplimit = idev->cnf.hop_limit;
			in6_dev_put(idev);
		} else
			hoplimit = ipv6_devconf.hop_limit;
L
Linus Torvalds 已提交
1052 1053 1054 1055 1056 1057 1058 1059
	}
	return hoplimit;
}

/*
 *
 */

1060
int ip6_route_add(struct fib6_config *cfg)
L
Linus Torvalds 已提交
1061 1062
{
	int err;
1063
	struct net *net = cfg->fc_nlinfo.nl_net;
L
Linus Torvalds 已提交
1064 1065 1066
	struct rt6_info *rt = NULL;
	struct net_device *dev = NULL;
	struct inet6_dev *idev = NULL;
T
Thomas Graf 已提交
1067
	struct fib6_table *table;
L
Linus Torvalds 已提交
1068 1069
	int addr_type;

1070
	if (cfg->fc_dst_len > 128 || cfg->fc_src_len > 128)
L
Linus Torvalds 已提交
1071 1072
		return -EINVAL;
#ifndef CONFIG_IPV6_SUBTREES
1073
	if (cfg->fc_src_len)
L
Linus Torvalds 已提交
1074 1075
		return -EINVAL;
#endif
1076
	if (cfg->fc_ifindex) {
L
Linus Torvalds 已提交
1077
		err = -ENODEV;
1078
		dev = dev_get_by_index(net, cfg->fc_ifindex);
L
Linus Torvalds 已提交
1079 1080 1081 1082 1083 1084 1085
		if (!dev)
			goto out;
		idev = in6_dev_get(dev);
		if (!idev)
			goto out;
	}

1086 1087
	if (cfg->fc_metric == 0)
		cfg->fc_metric = IP6_RT_PRIO_USER;
L
Linus Torvalds 已提交
1088

1089
	table = fib6_new_table(net, cfg->fc_table);
T
Thomas Graf 已提交
1090 1091 1092 1093 1094
	if (table == NULL) {
		err = -ENOBUFS;
		goto out;
	}

1095
	rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops);
L
Linus Torvalds 已提交
1096 1097 1098 1099 1100 1101 1102

	if (rt == NULL) {
		err = -ENOMEM;
		goto out;
	}

	rt->u.dst.obsolete = -1;
1103 1104 1105
	rt->rt6i_expires = (cfg->fc_flags & RTF_EXPIRES) ?
				jiffies + clock_t_to_jiffies(cfg->fc_expires) :
				0;
L
Linus Torvalds 已提交
1106

1107 1108 1109 1110 1111
	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
1112 1113 1114 1115 1116 1117 1118 1119

	if (addr_type & IPV6_ADDR_MULTICAST)
		rt->u.dst.input = ip6_mc_input;
	else
		rt->u.dst.input = ip6_forward;

	rt->u.dst.output = ip6_output;

1120 1121
	ipv6_addr_prefix(&rt->rt6i_dst.addr, &cfg->fc_dst, cfg->fc_dst_len);
	rt->rt6i_dst.plen = cfg->fc_dst_len;
L
Linus Torvalds 已提交
1122 1123 1124 1125
	if (rt->rt6i_dst.plen == 128)
	       rt->u.dst.flags = DST_HOST;

#ifdef CONFIG_IPV6_SUBTREES
1126 1127
	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
1128 1129
#endif

1130
	rt->rt6i_metric = cfg->fc_metric;
L
Linus Torvalds 已提交
1131 1132 1133 1134

	/* We cannot add true routes via loopback here,
	   they would result in kernel looping; promote them to reject routes
	 */
1135
	if ((cfg->fc_flags & RTF_REJECT) ||
L
Linus Torvalds 已提交
1136 1137
	    (dev && (dev->flags&IFF_LOOPBACK) && !(addr_type&IPV6_ADDR_LOOPBACK))) {
		/* hold loopback dev/idev if we haven't done so. */
1138
		if (dev != net->loopback_dev) {
L
Linus Torvalds 已提交
1139 1140 1141 1142
			if (dev) {
				dev_put(dev);
				in6_dev_put(idev);
			}
1143
			dev = net->loopback_dev;
L
Linus Torvalds 已提交
1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157
			dev_hold(dev);
			idev = in6_dev_get(dev);
			if (!idev) {
				err = -ENODEV;
				goto out;
			}
		}
		rt->u.dst.output = ip6_pkt_discard_out;
		rt->u.dst.input = ip6_pkt_discard;
		rt->u.dst.error = -ENETUNREACH;
		rt->rt6i_flags = RTF_REJECT|RTF_NONEXTHOP;
		goto install_route;
	}

1158
	if (cfg->fc_flags & RTF_GATEWAY) {
L
Linus Torvalds 已提交
1159 1160 1161
		struct in6_addr *gw_addr;
		int gwa_type;

1162 1163
		gw_addr = &cfg->fc_gateway;
		ipv6_addr_copy(&rt->rt6i_gateway, gw_addr);
L
Linus Torvalds 已提交
1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179
		gwa_type = ipv6_addr_type(gw_addr);

		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
			 */
			err = -EINVAL;
			if (!(gwa_type&IPV6_ADDR_UNICAST))
				goto out;

1180
			grt = rt6_lookup(net, gw_addr, NULL, cfg->fc_ifindex, 1);
L
Linus Torvalds 已提交
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211

			err = -EHOSTUNREACH;
			if (grt == NULL)
				goto out;
			if (dev) {
				if (dev != grt->rt6i_dev) {
					dst_release(&grt->u.dst);
					goto out;
				}
			} else {
				dev = grt->rt6i_dev;
				idev = grt->rt6i_idev;
				dev_hold(dev);
				in6_dev_hold(grt->rt6i_idev);
			}
			if (!(grt->rt6i_flags&RTF_GATEWAY))
				err = 0;
			dst_release(&grt->u.dst);

			if (err)
				goto out;
		}
		err = -EINVAL;
		if (dev == NULL || (dev->flags&IFF_LOOPBACK))
			goto out;
	}

	err = -ENODEV;
	if (dev == NULL)
		goto out;

1212
	if (cfg->fc_flags & (RTF_GATEWAY | RTF_NONEXTHOP)) {
L
Linus Torvalds 已提交
1213 1214 1215 1216 1217 1218 1219 1220
		rt->rt6i_nexthop = __neigh_lookup_errno(&nd_tbl, &rt->rt6i_gateway, dev);
		if (IS_ERR(rt->rt6i_nexthop)) {
			err = PTR_ERR(rt->rt6i_nexthop);
			rt->rt6i_nexthop = NULL;
			goto out;
		}
	}

1221
	rt->rt6i_flags = cfg->fc_flags;
L
Linus Torvalds 已提交
1222 1223

install_route:
1224 1225 1226 1227 1228
	if (cfg->fc_mx) {
		struct nlattr *nla;
		int remaining;

		nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) {
1229
			int type = nla_type(nla);
1230 1231 1232

			if (type) {
				if (type > RTAX_MAX) {
L
Linus Torvalds 已提交
1233 1234 1235
					err = -EINVAL;
					goto out;
				}
1236 1237

				rt->u.dst.metrics[type - 1] = nla_get_u32(nla);
L
Linus Torvalds 已提交
1238 1239 1240 1241
			}
		}
	}

1242
	if (dst_metric(&rt->u.dst, RTAX_HOPLIMIT) == 0)
L
Linus Torvalds 已提交
1243
		rt->u.dst.metrics[RTAX_HOPLIMIT-1] = -1;
1244
	if (!dst_metric(&rt->u.dst, RTAX_MTU))
L
Linus Torvalds 已提交
1245
		rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(dev);
1246
	if (!dst_metric(&rt->u.dst, RTAX_ADVMSS))
1247
		rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->u.dst));
L
Linus Torvalds 已提交
1248 1249
	rt->u.dst.dev = dev;
	rt->rt6i_idev = idev;
T
Thomas Graf 已提交
1250
	rt->rt6i_table = table;
1251

1252
	cfg->fc_nlinfo.nl_net = dev_net(dev);
1253

1254
	return __ip6_ins_rt(rt, &cfg->fc_nlinfo);
L
Linus Torvalds 已提交
1255 1256 1257 1258 1259 1260 1261

out:
	if (dev)
		dev_put(dev);
	if (idev)
		in6_dev_put(idev);
	if (rt)
1262
		dst_free(&rt->u.dst);
L
Linus Torvalds 已提交
1263 1264 1265
	return err;
}

1266
static int __ip6_del_rt(struct rt6_info *rt, struct nl_info *info)
L
Linus Torvalds 已提交
1267 1268
{
	int err;
T
Thomas Graf 已提交
1269
	struct fib6_table *table;
1270
	struct net *net = dev_net(rt->rt6i_dev);
L
Linus Torvalds 已提交
1271

1272
	if (rt == net->ipv6.ip6_null_entry)
1273 1274
		return -ENOENT;

T
Thomas Graf 已提交
1275 1276
	table = rt->rt6i_table;
	write_lock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1277

1278
	err = fib6_del(rt, info);
L
Linus Torvalds 已提交
1279 1280
	dst_release(&rt->u.dst);

T
Thomas Graf 已提交
1281
	write_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1282 1283 1284 1285

	return err;
}

1286 1287
int ip6_del_rt(struct rt6_info *rt)
{
1288
	struct nl_info info = {
1289
		.nl_net = dev_net(rt->rt6i_dev),
1290
	};
1291
	return __ip6_del_rt(rt, &info);
1292 1293
}

1294
static int ip6_route_del(struct fib6_config *cfg)
L
Linus Torvalds 已提交
1295
{
T
Thomas Graf 已提交
1296
	struct fib6_table *table;
L
Linus Torvalds 已提交
1297 1298 1299 1300
	struct fib6_node *fn;
	struct rt6_info *rt;
	int err = -ESRCH;

1301
	table = fib6_get_table(cfg->fc_nlinfo.nl_net, cfg->fc_table);
T
Thomas Graf 已提交
1302 1303 1304 1305
	if (table == NULL)
		return err;

	read_lock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1306

T
Thomas Graf 已提交
1307
	fn = fib6_locate(&table->tb6_root,
1308 1309
			 &cfg->fc_dst, cfg->fc_dst_len,
			 &cfg->fc_src, cfg->fc_src_len);
1310

L
Linus Torvalds 已提交
1311
	if (fn) {
1312
		for (rt = fn->leaf; rt; rt = rt->u.dst.rt6_next) {
1313
			if (cfg->fc_ifindex &&
L
Linus Torvalds 已提交
1314
			    (rt->rt6i_dev == NULL ||
1315
			     rt->rt6i_dev->ifindex != cfg->fc_ifindex))
L
Linus Torvalds 已提交
1316
				continue;
1317 1318
			if (cfg->fc_flags & RTF_GATEWAY &&
			    !ipv6_addr_equal(&cfg->fc_gateway, &rt->rt6i_gateway))
L
Linus Torvalds 已提交
1319
				continue;
1320
			if (cfg->fc_metric && cfg->fc_metric != rt->rt6i_metric)
L
Linus Torvalds 已提交
1321 1322
				continue;
			dst_hold(&rt->u.dst);
T
Thomas Graf 已提交
1323
			read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1324

1325
			return __ip6_del_rt(rt, &cfg->fc_nlinfo);
L
Linus Torvalds 已提交
1326 1327
		}
	}
T
Thomas Graf 已提交
1328
	read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1329 1330 1331 1332 1333 1334 1335

	return err;
}

/*
 *	Handle redirects
 */
1336 1337 1338 1339 1340
struct ip6rd_flowi {
	struct flowi fl;
	struct in6_addr gateway;
};

1341 1342
static struct rt6_info *__ip6_route_redirect(struct net *net,
					     struct fib6_table *table,
1343 1344
					     struct flowi *fl,
					     int flags)
L
Linus Torvalds 已提交
1345
{
1346 1347
	struct ip6rd_flowi *rdfl = (struct ip6rd_flowi *)fl;
	struct rt6_info *rt;
1348
	struct fib6_node *fn;
T
Thomas Graf 已提交
1349

L
Linus Torvalds 已提交
1350
	/*
1351 1352 1353 1354 1355 1356 1357 1358
	 * Get the "current" route for this destination and
	 * check if the redirect has come from approriate router.
	 *
	 * RFC 2461 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.
L
Linus Torvalds 已提交
1359 1360
	 */

T
Thomas Graf 已提交
1361
	read_lock_bh(&table->tb6_lock);
1362
	fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src);
1363
restart:
1364
	for (rt = fn->leaf; rt; rt = rt->u.dst.rt6_next) {
1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
		/*
		 * Current route is on-link; redirect is always invalid.
		 *
		 * Seems, previous statement is not true. It could
		 * be node, which looks for us as on-link (f.e. proxy ndisc)
		 * But then router serving it might decide, that we should
		 * know truth 8)8) --ANK (980726).
		 */
		if (rt6_check_expired(rt))
			continue;
		if (!(rt->rt6i_flags & RTF_GATEWAY))
			continue;
1377
		if (fl->oif != rt->rt6i_dev->ifindex)
1378
			continue;
1379
		if (!ipv6_addr_equal(&rdfl->gateway, &rt->rt6i_gateway))
1380 1381 1382
			continue;
		break;
	}
1383

1384
	if (!rt)
1385 1386
		rt = net->ipv6.ip6_null_entry;
	BACKTRACK(net, &fl->fl6_src);
1387
out:
1388 1389
	dst_hold(&rt->u.dst);

T
Thomas Graf 已提交
1390
	read_unlock_bh(&table->tb6_lock);
1391

1392 1393 1394 1395 1396 1397 1398 1399
	return rt;
};

static struct rt6_info *ip6_route_redirect(struct in6_addr *dest,
					   struct in6_addr *src,
					   struct in6_addr *gateway,
					   struct net_device *dev)
{
1400
	int flags = RT6_LOOKUP_F_HAS_SADDR;
1401
	struct net *net = dev_net(dev);
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413
	struct ip6rd_flowi rdfl = {
		.fl = {
			.oif = dev->ifindex,
			.nl_u = {
				.ip6_u = {
					.daddr = *dest,
					.saddr = *src,
				},
			},
		},
		.gateway = *gateway,
	};
1414 1415 1416

	if (rt6_need_strict(dest))
		flags |= RT6_LOOKUP_F_IFACE;
1417

1418
	return (struct rt6_info *)fib6_rule_lookup(net, (struct flowi *)&rdfl,
1419
						   flags, __ip6_route_redirect);
1420 1421 1422 1423 1424 1425 1426 1427
}

void rt6_redirect(struct in6_addr *dest, struct in6_addr *src,
		  struct in6_addr *saddr,
		  struct neighbour *neigh, u8 *lladdr, int on_link)
{
	struct rt6_info *rt, *nrt = NULL;
	struct netevent_redirect netevent;
1428
	struct net *net = dev_net(neigh->dev);
1429 1430 1431

	rt = ip6_route_redirect(dest, src, saddr, neigh->dev);

1432
	if (rt == net->ipv6.ip6_null_entry) {
L
Linus Torvalds 已提交
1433 1434 1435
		if (net_ratelimit())
			printk(KERN_DEBUG "rt6_redirect: source isn't a valid nexthop "
			       "for redirect target\n");
1436
		goto out;
L
Linus Torvalds 已提交
1437 1438 1439 1440 1441 1442
	}

	/*
	 *	We have finally decided to accept it.
	 */

1443
	neigh_update(neigh, lladdr, NUD_STALE,
L
Linus Torvalds 已提交
1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476
		     NEIGH_UPDATE_F_WEAK_OVERRIDE|
		     NEIGH_UPDATE_F_OVERRIDE|
		     (on_link ? 0 : (NEIGH_UPDATE_F_OVERRIDE_ISROUTER|
				     NEIGH_UPDATE_F_ISROUTER))
		     );

	/*
	 * 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->u.dst);

	/* Duplicate redirect: silently ignore. */
	if (neigh == rt->u.dst.neighbour)
		goto out;

	nrt = ip6_rt_copy(rt);
	if (nrt == NULL)
		goto out;

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

	ipv6_addr_copy(&nrt->rt6i_dst.addr, dest);
	nrt->rt6i_dst.plen = 128;
	nrt->u.dst.flags |= DST_HOST;

	ipv6_addr_copy(&nrt->rt6i_gateway, (struct in6_addr*)neigh->primary_key);
	nrt->rt6i_nexthop = neigh_clone(neigh);
	/* Reset pmtu, it may be better */
	nrt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(neigh->dev);
1477
	nrt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(dev_net(neigh->dev),
1478
							dst_mtu(&nrt->u.dst));
L
Linus Torvalds 已提交
1479

1480
	if (ip6_ins_rt(nrt))
L
Linus Torvalds 已提交
1481 1482
		goto out;

1483 1484 1485 1486
	netevent.old = &rt->u.dst;
	netevent.new = &nrt->u.dst;
	call_netevent_notifiers(NETEVENT_REDIRECT, &netevent);

L
Linus Torvalds 已提交
1487
	if (rt->rt6i_flags&RTF_CACHE) {
1488
		ip6_del_rt(rt);
L
Linus Torvalds 已提交
1489 1490 1491 1492
		return;
	}

out:
1493
	dst_release(&rt->u.dst);
L
Linus Torvalds 已提交
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505
	return;
}

/*
 *	Handle ICMP "packet too big" messages
 *	i.e. Path MTU discovery
 */

void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
			struct net_device *dev, u32 pmtu)
{
	struct rt6_info *rt, *nrt;
1506
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
1507 1508
	int allfrag = 0;

1509
	rt = rt6_lookup(net, daddr, saddr, dev->ifindex, 0);
L
Linus Torvalds 已提交
1510 1511 1512 1513 1514 1515 1516 1517
	if (rt == NULL)
		return;

	if (pmtu >= dst_mtu(&rt->u.dst))
		goto out;

	if (pmtu < IPV6_MIN_MTU) {
		/*
1518
		 * According to RFC2460, PMTU is set to the IPv6 Minimum Link
L
Linus Torvalds 已提交
1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541
		 * MTU (1280) and a fragment header should always be included
		 * after a node receiving Too Big message reporting PMTU is
		 * less than the IPv6 Minimum Link MTU.
		 */
		pmtu = IPV6_MIN_MTU;
		allfrag = 1;
	}

	/* New mtu received -> path was valid.
	   They are sent only in response to data packets,
	   so that this nexthop apparently is reachable. --ANK
	 */
	dst_confirm(&rt->u.dst);

	/* Host route. If it is static, it would be better
	   not to override it, but add new one, so that
	   when cache entry will expire old pmtu
	   would return automatically.
	 */
	if (rt->rt6i_flags & RTF_CACHE) {
		rt->u.dst.metrics[RTAX_MTU-1] = pmtu;
		if (allfrag)
			rt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
1542
		dst_set_expires(&rt->u.dst, net->ipv6.sysctl.ip6_rt_mtu_expires);
L
Linus Torvalds 已提交
1543 1544 1545 1546 1547 1548 1549 1550 1551
		rt->rt6i_flags |= RTF_MODIFIED|RTF_EXPIRES;
		goto out;
	}

	/* Network route.
	   Two cases are possible:
	   1. It is connected route. Action: COW
	   2. It is gatewayed route or NONEXTHOP route. Action: clone it.
	 */
1552
	if (!rt->rt6i_nexthop && !(rt->rt6i_flags & RTF_NONEXTHOP))
1553
		nrt = rt6_alloc_cow(rt, daddr, saddr);
1554 1555
	else
		nrt = rt6_alloc_clone(rt, daddr);
1556

1557
	if (nrt) {
1558 1559 1560 1561 1562 1563 1564 1565 1566 1567
		nrt->u.dst.metrics[RTAX_MTU-1] = pmtu;
		if (allfrag)
			nrt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;

		/* According to RFC 1981, detecting PMTU increase shouldn't be
		 * happened within 5 mins, the recommended timer is 10 mins.
		 * Here this route expiration time is set to ip6_rt_mtu_expires
		 * which is 10 mins. After 10 mins the decreased pmtu is expired
		 * and detecting PMTU increase will be automatically happened.
		 */
1568
		dst_set_expires(&nrt->u.dst, net->ipv6.sysctl.ip6_rt_mtu_expires);
1569 1570
		nrt->rt6i_flags |= RTF_DYNAMIC|RTF_EXPIRES;

1571
		ip6_ins_rt(nrt);
L
Linus Torvalds 已提交
1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582
	}
out:
	dst_release(&rt->u.dst);
}

/*
 *	Misc support functions
 */

static struct rt6_info * ip6_rt_copy(struct rt6_info *ort)
{
1583
	struct net *net = dev_net(ort->rt6i_dev);
1584
	struct rt6_info *rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops);
L
Linus Torvalds 已提交
1585 1586 1587 1588 1589 1590

	if (rt) {
		rt->u.dst.input = ort->u.dst.input;
		rt->u.dst.output = ort->u.dst.output;

		memcpy(rt->u.dst.metrics, ort->u.dst.metrics, RTAX_MAX*sizeof(u32));
1591
		rt->u.dst.error = ort->u.dst.error;
L
Linus Torvalds 已提交
1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608
		rt->u.dst.dev = ort->u.dst.dev;
		if (rt->u.dst.dev)
			dev_hold(rt->u.dst.dev);
		rt->rt6i_idev = ort->rt6i_idev;
		if (rt->rt6i_idev)
			in6_dev_hold(rt->rt6i_idev);
		rt->u.dst.lastuse = jiffies;
		rt->rt6i_expires = 0;

		ipv6_addr_copy(&rt->rt6i_gateway, &ort->rt6i_gateway);
		rt->rt6i_flags = ort->rt6i_flags & ~RTF_EXPIRES;
		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
T
Thomas Graf 已提交
1609
		rt->rt6i_table = ort->rt6i_table;
L
Linus Torvalds 已提交
1610 1611 1612 1613
	}
	return rt;
}

1614
#ifdef CONFIG_IPV6_ROUTE_INFO
1615 1616
static struct rt6_info *rt6_get_route_info(struct net *net,
					   struct in6_addr *prefix, int prefixlen,
1617 1618 1619 1620
					   struct in6_addr *gwaddr, int ifindex)
{
	struct fib6_node *fn;
	struct rt6_info *rt = NULL;
T
Thomas Graf 已提交
1621 1622
	struct fib6_table *table;

1623
	table = fib6_get_table(net, RT6_TABLE_INFO);
T
Thomas Graf 已提交
1624 1625
	if (table == NULL)
		return NULL;
1626

T
Thomas Graf 已提交
1627 1628
	write_lock_bh(&table->tb6_lock);
	fn = fib6_locate(&table->tb6_root, prefix ,prefixlen, NULL, 0);
1629 1630 1631
	if (!fn)
		goto out;

1632
	for (rt = fn->leaf; rt; rt = rt->u.dst.rt6_next) {
1633 1634 1635 1636 1637 1638 1639 1640 1641 1642
		if (rt->rt6i_dev->ifindex != ifindex)
			continue;
		if ((rt->rt6i_flags & (RTF_ROUTEINFO|RTF_GATEWAY)) != (RTF_ROUTEINFO|RTF_GATEWAY))
			continue;
		if (!ipv6_addr_equal(&rt->rt6i_gateway, gwaddr))
			continue;
		dst_hold(&rt->u.dst);
		break;
	}
out:
T
Thomas Graf 已提交
1643
	write_unlock_bh(&table->tb6_lock);
1644 1645 1646
	return rt;
}

1647 1648
static struct rt6_info *rt6_add_route_info(struct net *net,
					   struct in6_addr *prefix, int prefixlen,
1649 1650 1651
					   struct in6_addr *gwaddr, int ifindex,
					   unsigned pref)
{
1652 1653
	struct fib6_config cfg = {
		.fc_table	= RT6_TABLE_INFO,
1654
		.fc_metric	= IP6_RT_PRIO_USER,
1655 1656 1657 1658
		.fc_ifindex	= ifindex,
		.fc_dst_len	= prefixlen,
		.fc_flags	= RTF_GATEWAY | RTF_ADDRCONF | RTF_ROUTEINFO |
				  RTF_UP | RTF_PREF(pref),
1659 1660 1661
		.fc_nlinfo.pid = 0,
		.fc_nlinfo.nlh = NULL,
		.fc_nlinfo.nl_net = net,
1662 1663 1664 1665
	};

	ipv6_addr_copy(&cfg.fc_dst, prefix);
	ipv6_addr_copy(&cfg.fc_gateway, gwaddr);
1666

1667 1668
	/* We should treat it as a default route if prefix length is 0. */
	if (!prefixlen)
1669
		cfg.fc_flags |= RTF_DEFAULT;
1670

1671
	ip6_route_add(&cfg);
1672

1673
	return rt6_get_route_info(net, prefix, prefixlen, gwaddr, ifindex);
1674 1675 1676
}
#endif

L
Linus Torvalds 已提交
1677
struct rt6_info *rt6_get_dflt_router(struct in6_addr *addr, struct net_device *dev)
1678
{
L
Linus Torvalds 已提交
1679
	struct rt6_info *rt;
T
Thomas Graf 已提交
1680
	struct fib6_table *table;
L
Linus Torvalds 已提交
1681

1682
	table = fib6_get_table(dev_net(dev), RT6_TABLE_DFLT);
T
Thomas Graf 已提交
1683 1684
	if (table == NULL)
		return NULL;
L
Linus Torvalds 已提交
1685

T
Thomas Graf 已提交
1686
	write_lock_bh(&table->tb6_lock);
1687
	for (rt = table->tb6_root.leaf; rt; rt=rt->u.dst.rt6_next) {
L
Linus Torvalds 已提交
1688
		if (dev == rt->rt6i_dev &&
1689
		    ((rt->rt6i_flags & (RTF_ADDRCONF | RTF_DEFAULT)) == (RTF_ADDRCONF | RTF_DEFAULT)) &&
L
Linus Torvalds 已提交
1690 1691 1692 1693 1694
		    ipv6_addr_equal(&rt->rt6i_gateway, addr))
			break;
	}
	if (rt)
		dst_hold(&rt->u.dst);
T
Thomas Graf 已提交
1695
	write_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1696 1697 1698 1699
	return rt;
}

struct rt6_info *rt6_add_dflt_router(struct in6_addr *gwaddr,
1700 1701
				     struct net_device *dev,
				     unsigned int pref)
L
Linus Torvalds 已提交
1702
{
1703 1704
	struct fib6_config cfg = {
		.fc_table	= RT6_TABLE_DFLT,
1705
		.fc_metric	= IP6_RT_PRIO_USER,
1706 1707 1708
		.fc_ifindex	= dev->ifindex,
		.fc_flags	= RTF_GATEWAY | RTF_ADDRCONF | RTF_DEFAULT |
				  RTF_UP | RTF_EXPIRES | RTF_PREF(pref),
1709 1710
		.fc_nlinfo.pid = 0,
		.fc_nlinfo.nlh = NULL,
1711
		.fc_nlinfo.nl_net = dev_net(dev),
1712
	};
L
Linus Torvalds 已提交
1713

1714
	ipv6_addr_copy(&cfg.fc_gateway, gwaddr);
L
Linus Torvalds 已提交
1715

1716
	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
1717 1718 1719 1720

	return rt6_get_dflt_router(gwaddr, dev);
}

1721
void rt6_purge_dflt_routers(struct net *net)
L
Linus Torvalds 已提交
1722 1723
{
	struct rt6_info *rt;
T
Thomas Graf 已提交
1724 1725 1726
	struct fib6_table *table;

	/* NOTE: Keep consistent with rt6_get_dflt_router */
1727
	table = fib6_get_table(net, RT6_TABLE_DFLT);
T
Thomas Graf 已提交
1728 1729
	if (table == NULL)
		return;
L
Linus Torvalds 已提交
1730 1731

restart:
T
Thomas Graf 已提交
1732
	read_lock_bh(&table->tb6_lock);
1733
	for (rt = table->tb6_root.leaf; rt; rt = rt->u.dst.rt6_next) {
L
Linus Torvalds 已提交
1734 1735
		if (rt->rt6i_flags & (RTF_DEFAULT | RTF_ADDRCONF)) {
			dst_hold(&rt->u.dst);
T
Thomas Graf 已提交
1736
			read_unlock_bh(&table->tb6_lock);
1737
			ip6_del_rt(rt);
L
Linus Torvalds 已提交
1738 1739 1740
			goto restart;
		}
	}
T
Thomas Graf 已提交
1741
	read_unlock_bh(&table->tb6_lock);
L
Linus Torvalds 已提交
1742 1743
}

1744 1745
static void rtmsg_to_fib6_config(struct net *net,
				 struct in6_rtmsg *rtmsg,
1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757
				 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;

1758
	cfg->fc_nlinfo.nl_net = net;
1759

1760 1761 1762 1763 1764
	ipv6_addr_copy(&cfg->fc_dst, &rtmsg->rtmsg_dst);
	ipv6_addr_copy(&cfg->fc_src, &rtmsg->rtmsg_src);
	ipv6_addr_copy(&cfg->fc_gateway, &rtmsg->rtmsg_gateway);
}

1765
int ipv6_route_ioctl(struct net *net, unsigned int cmd, void __user *arg)
L
Linus Torvalds 已提交
1766
{
1767
	struct fib6_config cfg;
L
Linus Torvalds 已提交
1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779
	struct in6_rtmsg rtmsg;
	int err;

	switch(cmd) {
	case SIOCADDRT:		/* Add a route */
	case SIOCDELRT:		/* Delete a route */
		if (!capable(CAP_NET_ADMIN))
			return -EPERM;
		err = copy_from_user(&rtmsg, arg,
				     sizeof(struct in6_rtmsg));
		if (err)
			return -EFAULT;
1780

1781
		rtmsg_to_fib6_config(net, &rtmsg, &cfg);
1782

L
Linus Torvalds 已提交
1783 1784 1785
		rtnl_lock();
		switch (cmd) {
		case SIOCADDRT:
1786
			err = ip6_route_add(&cfg);
L
Linus Torvalds 已提交
1787 1788
			break;
		case SIOCDELRT:
1789
			err = ip6_route_del(&cfg);
L
Linus Torvalds 已提交
1790 1791 1792 1793 1794 1795 1796
			break;
		default:
			err = -EINVAL;
		}
		rtnl_unlock();

		return err;
1797
	}
L
Linus Torvalds 已提交
1798 1799 1800 1801 1802 1803 1804 1805

	return -EINVAL;
}

/*
 *	Drop the packet on the floor
 */

I
Ilpo Järvinen 已提交
1806
static int ip6_pkt_drop(struct sk_buff *skb, int code, int ipstats_mib_noroutes)
L
Linus Torvalds 已提交
1807
{
1808 1809 1810
	int type;
	switch (ipstats_mib_noroutes) {
	case IPSTATS_MIB_INNOROUTES:
1811
		type = ipv6_addr_type(&ipv6_hdr(skb)->daddr);
1812 1813 1814 1815 1816 1817 1818 1819 1820
		if (type == IPV6_ADDR_ANY || type == IPV6_ADDR_RESERVED) {
			IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_INADDRERRORS);
			break;
		}
		/* FALLTHROUGH */
	case IPSTATS_MIB_OUTNOROUTES:
		IP6_INC_STATS(ip6_dst_idev(skb->dst), ipstats_mib_noroutes);
		break;
	}
1821
	icmpv6_send(skb, ICMPV6_DEST_UNREACH, code, 0, skb->dev);
L
Linus Torvalds 已提交
1822 1823 1824 1825
	kfree_skb(skb);
	return 0;
}

1826 1827
static int ip6_pkt_discard(struct sk_buff *skb)
{
1828
	return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_INNOROUTES);
1829 1830
}

1831
static int ip6_pkt_discard_out(struct sk_buff *skb)
L
Linus Torvalds 已提交
1832 1833
{
	skb->dev = skb->dst->dev;
1834
	return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
1835 1836
}

1837 1838
#ifdef CONFIG_IPV6_MULTIPLE_TABLES

1839 1840
static int ip6_pkt_prohibit(struct sk_buff *skb)
{
1841
	return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_INNOROUTES);
1842 1843 1844 1845 1846
}

static int ip6_pkt_prohibit_out(struct sk_buff *skb)
{
	skb->dev = skb->dst->dev;
1847
	return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_OUTNOROUTES);
1848 1849
}

1850 1851
#endif

L
Linus Torvalds 已提交
1852 1853 1854 1855 1856 1857 1858 1859
/*
 *	Allocate a dst for local (unicast / anycast) address.
 */

struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
				    const struct in6_addr *addr,
				    int anycast)
{
1860
	struct net *net = dev_net(idev->dev);
1861
	struct rt6_info *rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops);
L
Linus Torvalds 已提交
1862 1863 1864 1865

	if (rt == NULL)
		return ERR_PTR(-ENOMEM);

1866
	dev_hold(net->loopback_dev);
L
Linus Torvalds 已提交
1867 1868 1869 1870 1871
	in6_dev_hold(idev);

	rt->u.dst.flags = DST_HOST;
	rt->u.dst.input = ip6_input;
	rt->u.dst.output = ip6_output;
1872
	rt->rt6i_dev = net->loopback_dev;
L
Linus Torvalds 已提交
1873 1874
	rt->rt6i_idev = idev;
	rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(rt->rt6i_dev);
1875
	rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->u.dst));
L
Linus Torvalds 已提交
1876 1877 1878 1879
	rt->u.dst.metrics[RTAX_HOPLIMIT-1] = -1;
	rt->u.dst.obsolete = -1;

	rt->rt6i_flags = RTF_UP | RTF_NONEXTHOP;
1880 1881 1882
	if (anycast)
		rt->rt6i_flags |= RTF_ANYCAST;
	else
L
Linus Torvalds 已提交
1883 1884 1885
		rt->rt6i_flags |= RTF_LOCAL;
	rt->rt6i_nexthop = ndisc_get_neigh(rt->rt6i_dev, &rt->rt6i_gateway);
	if (rt->rt6i_nexthop == NULL) {
1886
		dst_free(&rt->u.dst);
L
Linus Torvalds 已提交
1887 1888 1889 1890 1891
		return ERR_PTR(-ENOMEM);
	}

	ipv6_addr_copy(&rt->rt6i_dst.addr, addr);
	rt->rt6i_dst.plen = 128;
1892
	rt->rt6i_table = fib6_get_table(net, RT6_TABLE_LOCAL);
L
Linus Torvalds 已提交
1893 1894 1895 1896 1897 1898

	atomic_set(&rt->u.dst.__refcnt, 1);

	return rt;
}

1899 1900 1901 1902 1903
struct arg_dev_net {
	struct net_device *dev;
	struct net *net;
};

L
Linus Torvalds 已提交
1904 1905
static int fib6_ifdown(struct rt6_info *rt, void *arg)
{
1906 1907 1908 1909 1910
	struct net_device *dev = ((struct arg_dev_net *)arg)->dev;
	struct net *net = ((struct arg_dev_net *)arg)->net;

	if (((void *)rt->rt6i_dev == dev || dev == NULL) &&
	    rt != net->ipv6.ip6_null_entry) {
L
Linus Torvalds 已提交
1911 1912 1913 1914 1915 1916
		RT6_TRACE("deleted by ifdown %p\n", rt);
		return -1;
	}
	return 0;
}

1917
void rt6_ifdown(struct net *net, struct net_device *dev)
L
Linus Torvalds 已提交
1918
{
1919 1920 1921 1922 1923 1924
	struct arg_dev_net adn = {
		.dev = dev,
		.net = net,
	};

	fib6_clean_all(net, fib6_ifdown, 0, &adn);
L
Linus Torvalds 已提交
1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936
}

struct rt6_mtu_change_arg
{
	struct net_device *dev;
	unsigned mtu;
};

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;
1937
	struct net *net = dev_net(arg->dev);
L
Linus Torvalds 已提交
1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964

	/* 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);
	if (idev == NULL)
		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.
	 */
	if (rt->rt6i_dev == arg->dev &&
	    !dst_metric_locked(&rt->u.dst, RTAX_MTU) &&
1965
	    (dst_mtu(&rt->u.dst) >= arg->mtu ||
1966
	     (dst_mtu(&rt->u.dst) < arg->mtu &&
1967
	      dst_mtu(&rt->u.dst) == idev->cnf.mtu6))) {
L
Linus Torvalds 已提交
1968
		rt->u.dst.metrics[RTAX_MTU-1] = arg->mtu;
1969
		rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, arg->mtu);
1970
	}
L
Linus Torvalds 已提交
1971 1972 1973 1974 1975
	return 0;
}

void rt6_mtu_change(struct net_device *dev, unsigned mtu)
{
T
Thomas Graf 已提交
1976 1977 1978 1979
	struct rt6_mtu_change_arg arg = {
		.dev = dev,
		.mtu = mtu,
	};
L
Linus Torvalds 已提交
1980

1981
	fib6_clean_all(dev_net(dev), rt6_mtu_change_route, 0, &arg);
L
Linus Torvalds 已提交
1982 1983
}

1984
static const struct nla_policy rtm_ipv6_policy[RTA_MAX+1] = {
1985
	[RTA_GATEWAY]           = { .len = sizeof(struct in6_addr) },
1986
	[RTA_OIF]               = { .type = NLA_U32 },
1987
	[RTA_IIF]		= { .type = NLA_U32 },
1988 1989 1990 1991 1992 1993
	[RTA_PRIORITY]          = { .type = NLA_U32 },
	[RTA_METRICS]           = { .type = NLA_NESTED },
};

static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh,
			      struct fib6_config *cfg)
L
Linus Torvalds 已提交
1994
{
1995 1996 1997
	struct rtmsg *rtm;
	struct nlattr *tb[RTA_MAX+1];
	int err;
L
Linus Torvalds 已提交
1998

1999 2000 2001
	err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv6_policy);
	if (err < 0)
		goto errout;
L
Linus Torvalds 已提交
2002

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
	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;

	if (rtm->rtm_type == RTN_UNREACHABLE)
		cfg->fc_flags |= RTF_REJECT;

	cfg->fc_nlinfo.pid = NETLINK_CB(skb).pid;
	cfg->fc_nlinfo.nlh = nlh;
2018
	cfg->fc_nlinfo.nl_net = sock_net(skb->sk);
2019 2020 2021 2022

	if (tb[RTA_GATEWAY]) {
		nla_memcpy(&cfg->fc_gateway, tb[RTA_GATEWAY], 16);
		cfg->fc_flags |= RTF_GATEWAY;
L
Linus Torvalds 已提交
2023
	}
2024 2025 2026 2027 2028 2029 2030 2031

	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
2032
	}
2033 2034 2035 2036 2037 2038 2039 2040

	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
2041
	}
2042 2043 2044 2045 2046 2047 2048 2049 2050 2051

	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
2052
	}
2053 2054 2055 2056 2057 2058 2059

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

	err = 0;
errout:
	return err;
L
Linus Torvalds 已提交
2060 2061
}

2062
static int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
L
Linus Torvalds 已提交
2063
{
2064 2065
	struct fib6_config cfg;
	int err;
L
Linus Torvalds 已提交
2066

2067 2068 2069 2070 2071
	err = rtm_to_fib6_config(skb, nlh, &cfg);
	if (err < 0)
		return err;

	return ip6_route_del(&cfg);
L
Linus Torvalds 已提交
2072 2073
}

2074
static int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
L
Linus Torvalds 已提交
2075
{
2076 2077
	struct fib6_config cfg;
	int err;
L
Linus Torvalds 已提交
2078

2079 2080 2081 2082 2083
	err = rtm_to_fib6_config(skb, nlh, &cfg);
	if (err < 0)
		return err;

	return ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2084 2085
}

2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096
static inline size_t rt6_nlmsg_size(void)
{
	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 */
2097
	       + RTAX_MAX * nla_total_size(4) /* RTA_METRICS */
2098 2099 2100
	       + nla_total_size(sizeof(struct rta_cacheinfo));
}

L
Linus Torvalds 已提交
2101
static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
2102 2103
			 struct in6_addr *dst, struct in6_addr *src,
			 int iif, int type, u32 pid, u32 seq,
2104
			 int prefix, int nowait, unsigned int flags)
L
Linus Torvalds 已提交
2105 2106
{
	struct rtmsg *rtm;
2107
	struct nlmsghdr *nlh;
2108
	long expires;
2109
	u32 table;
L
Linus Torvalds 已提交
2110 2111 2112 2113 2114 2115 2116 2117

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

2118 2119
	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*rtm), flags);
	if (nlh == NULL)
2120
		return -EMSGSIZE;
2121 2122

	rtm = nlmsg_data(nlh);
L
Linus Torvalds 已提交
2123 2124 2125 2126
	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
2127
	if (rt->rt6i_table)
2128
		table = rt->rt6i_table->tb6_id;
T
Thomas Graf 已提交
2129
	else
2130 2131
		table = RT6_TABLE_UNSPEC;
	rtm->rtm_table = table;
2132
	NLA_PUT_U32(skb, RTA_TABLE, table);
L
Linus Torvalds 已提交
2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152
	if (rt->rt6i_flags&RTF_REJECT)
		rtm->rtm_type = RTN_UNREACHABLE;
	else if (rt->rt6i_dev && (rt->rt6i_dev->flags&IFF_LOOPBACK))
		rtm->rtm_type = RTN_LOCAL;
	else
		rtm->rtm_type = RTN_UNICAST;
	rtm->rtm_flags = 0;
	rtm->rtm_scope = RT_SCOPE_UNIVERSE;
	rtm->rtm_protocol = rt->rt6i_protocol;
	if (rt->rt6i_flags&RTF_DYNAMIC)
		rtm->rtm_protocol = RTPROT_REDIRECT;
	else if (rt->rt6i_flags & RTF_ADDRCONF)
		rtm->rtm_protocol = RTPROT_KERNEL;
	else if (rt->rt6i_flags&RTF_DEFAULT)
		rtm->rtm_protocol = RTPROT_RA;

	if (rt->rt6i_flags&RTF_CACHE)
		rtm->rtm_flags |= RTM_F_CLONED;

	if (dst) {
2153
		NLA_PUT(skb, RTA_DST, 16, dst);
2154
		rtm->rtm_dst_len = 128;
L
Linus Torvalds 已提交
2155
	} else if (rtm->rtm_dst_len)
2156
		NLA_PUT(skb, RTA_DST, 16, &rt->rt6i_dst.addr);
L
Linus Torvalds 已提交
2157 2158
#ifdef CONFIG_IPV6_SUBTREES
	if (src) {
2159
		NLA_PUT(skb, RTA_SRC, 16, src);
2160
		rtm->rtm_src_len = 128;
L
Linus Torvalds 已提交
2161
	} else if (rtm->rtm_src_len)
2162
		NLA_PUT(skb, RTA_SRC, 16, &rt->rt6i_src.addr);
L
Linus Torvalds 已提交
2163
#endif
2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181
	if (iif) {
#ifdef CONFIG_IPV6_MROUTE
		if (ipv6_addr_is_multicast(&rt->rt6i_dst.addr)) {
			int err = ip6mr_get_route(skb, rtm, nowait);
			if (err <= 0) {
				if (!nowait) {
					if (err == 0)
						return 0;
					goto nla_put_failure;
				} else {
					if (err == -EMSGSIZE)
						goto nla_put_failure;
				}
			}
		} else
#endif
			NLA_PUT_U32(skb, RTA_IIF, iif);
	} else if (dst) {
L
Linus Torvalds 已提交
2182
		struct in6_addr saddr_buf;
2183
		if (ipv6_dev_get_saddr(ip6_dst_idev(&rt->u.dst)->dev,
2184
				       dst, 0, &saddr_buf) == 0)
2185
			NLA_PUT(skb, RTA_PREFSRC, 16, &saddr_buf);
L
Linus Torvalds 已提交
2186
	}
2187

L
Linus Torvalds 已提交
2188
	if (rtnetlink_put_metrics(skb, rt->u.dst.metrics) < 0)
2189 2190
		goto nla_put_failure;

L
Linus Torvalds 已提交
2191
	if (rt->u.dst.neighbour)
2192 2193
		NLA_PUT(skb, RTA_GATEWAY, 16, &rt->u.dst.neighbour->primary_key);

L
Linus Torvalds 已提交
2194
	if (rt->u.dst.dev)
2195 2196 2197
		NLA_PUT_U32(skb, RTA_OIF, rt->rt6i_dev->ifindex);

	NLA_PUT_U32(skb, RTA_PRIORITY, rt->rt6i_metric);
2198

2199 2200 2201 2202 2203 2204
	if (!(rt->rt6i_flags & RTF_EXPIRES))
		expires = 0;
	else if (rt->rt6i_expires - jiffies < INT_MAX)
		expires = rt->rt6i_expires - jiffies;
	else
		expires = INT_MAX;
2205

2206 2207 2208
	if (rtnl_put_cacheinfo(skb, &rt->u.dst, 0, 0, 0,
			       expires, rt->u.dst.error) < 0)
		goto nla_put_failure;
2209 2210 2211 2212

	return nlmsg_end(skb, nlh);

nla_put_failure:
2213 2214
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
2215 2216
}

2217
int rt6_dump_route(struct rt6_info *rt, void *p_arg)
L
Linus Torvalds 已提交
2218 2219 2220 2221
{
	struct rt6_rtnl_dump_arg *arg = (struct rt6_rtnl_dump_arg *) p_arg;
	int prefix;

2222 2223
	if (nlmsg_len(arg->cb->nlh) >= sizeof(struct rtmsg)) {
		struct rtmsg *rtm = nlmsg_data(arg->cb->nlh);
L
Linus Torvalds 已提交
2224 2225 2226 2227 2228 2229
		prefix = (rtm->rtm_flags & RTM_F_PREFIX) != 0;
	} else
		prefix = 0;

	return rt6_fill_node(arg->skb, rt, NULL, NULL, 0, RTM_NEWROUTE,
		     NETLINK_CB(arg->cb->skb).pid, arg->cb->nlh->nlmsg_seq,
2230
		     prefix, 0, NLM_F_MULTI);
L
Linus Torvalds 已提交
2231 2232
}

2233
static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
L
Linus Torvalds 已提交
2234
{
2235
	struct net *net = sock_net(in_skb->sk);
2236 2237
	struct nlattr *tb[RTA_MAX+1];
	struct rt6_info *rt;
L
Linus Torvalds 已提交
2238
	struct sk_buff *skb;
2239
	struct rtmsg *rtm;
L
Linus Torvalds 已提交
2240
	struct flowi fl;
2241
	int err, iif = 0;
L
Linus Torvalds 已提交
2242

2243 2244 2245
	err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv6_policy);
	if (err < 0)
		goto errout;
L
Linus Torvalds 已提交
2246

2247
	err = -EINVAL;
L
Linus Torvalds 已提交
2248 2249
	memset(&fl, 0, sizeof(fl));

2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268
	if (tb[RTA_SRC]) {
		if (nla_len(tb[RTA_SRC]) < sizeof(struct in6_addr))
			goto errout;

		ipv6_addr_copy(&fl.fl6_src, nla_data(tb[RTA_SRC]));
	}

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

		ipv6_addr_copy(&fl.fl6_dst, nla_data(tb[RTA_DST]));
	}

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

	if (tb[RTA_OIF])
		fl.oif = nla_get_u32(tb[RTA_OIF]);
L
Linus Torvalds 已提交
2269 2270 2271

	if (iif) {
		struct net_device *dev;
2272
		dev = __dev_get_by_index(net, iif);
L
Linus Torvalds 已提交
2273 2274
		if (!dev) {
			err = -ENODEV;
2275
			goto errout;
L
Linus Torvalds 已提交
2276 2277 2278
		}
	}

2279 2280 2281 2282 2283
	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (skb == NULL) {
		err = -ENOBUFS;
		goto errout;
	}
L
Linus Torvalds 已提交
2284

2285 2286 2287
	/* Reserve room for dummy headers, this skb can pass
	   through good chunk of routing engine.
	 */
2288
	skb_reset_mac_header(skb);
2289
	skb_reserve(skb, MAX_HEADER + sizeof(struct ipv6hdr));
L
Linus Torvalds 已提交
2290

2291
	rt = (struct rt6_info*) ip6_route_output(net, NULL, &fl);
L
Linus Torvalds 已提交
2292 2293
	skb->dst = &rt->u.dst;

2294
	err = rt6_fill_node(skb, rt, &fl.fl6_dst, &fl.fl6_src, iif,
L
Linus Torvalds 已提交
2295
			    RTM_NEWROUTE, NETLINK_CB(in_skb).pid,
2296
			    nlh->nlmsg_seq, 0, 0, 0);
L
Linus Torvalds 已提交
2297
	if (err < 0) {
2298 2299
		kfree_skb(skb);
		goto errout;
L
Linus Torvalds 已提交
2300 2301
	}

2302
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).pid);
2303
errout:
L
Linus Torvalds 已提交
2304 2305 2306
	return err;
}

2307
void inet6_rt_notify(int event, struct rt6_info *rt, struct nl_info *info)
L
Linus Torvalds 已提交
2308 2309
{
	struct sk_buff *skb;
2310
	struct net *net = info->nl_net;
2311 2312 2313 2314 2315
	u32 seq;
	int err;

	err = -ENOBUFS;
	seq = info->nlh != NULL ? info->nlh->nlmsg_seq : 0;
2316

2317
	skb = nlmsg_new(rt6_nlmsg_size(), gfp_any());
2318 2319 2320
	if (skb == NULL)
		goto errout;

2321
	err = rt6_fill_node(skb, rt, NULL, NULL, 0,
2322
				event, info->pid, seq, 0, 0, 0);
2323 2324 2325 2326 2327 2328
	if (err < 0) {
		/* -EMSGSIZE implies BUG in rt6_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
2329 2330
	err = rtnl_notify(skb, net, info->pid, RTNLGRP_IPV6_ROUTE,
			  info->nlh, gfp_any());
2331 2332
errout:
	if (err < 0)
2333
		rtnl_set_sk_err(net, RTNLGRP_IPV6_ROUTE, err);
L
Linus Torvalds 已提交
2334 2335
}

2336 2337 2338 2339
static int ip6_route_dev_notify(struct notifier_block *this,
				unsigned long event, void *data)
{
	struct net_device *dev = (struct net_device *)data;
2340
	struct net *net = dev_net(dev);
2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355

	if (event == NETDEV_REGISTER && (dev->flags & IFF_LOOPBACK)) {
		net->ipv6.ip6_null_entry->u.dst.dev = dev;
		net->ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(dev);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
		net->ipv6.ip6_prohibit_entry->u.dst.dev = dev;
		net->ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(dev);
		net->ipv6.ip6_blk_hole_entry->u.dst.dev = dev;
		net->ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(dev);
#endif
	}

	return NOTIFY_OK;
}

L
Linus Torvalds 已提交
2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374
/*
 *	/proc
 */

#ifdef CONFIG_PROC_FS

#define RT6_INFO_LEN (32 + 4 + 32 + 4 + 32 + 40 + 5 + 1)

struct rt6_proc_arg
{
	char *buffer;
	int offset;
	int length;
	int skip;
	int len;
};

static int rt6_info_route(struct rt6_info *rt, void *p_arg)
{
2375
	struct seq_file *m = p_arg;
L
Linus Torvalds 已提交
2376

2377 2378
	seq_printf(m, NIP6_SEQFMT " %02x ", NIP6(rt->rt6i_dst.addr),
		   rt->rt6i_dst.plen);
L
Linus Torvalds 已提交
2379 2380

#ifdef CONFIG_IPV6_SUBTREES
2381 2382
	seq_printf(m, NIP6_SEQFMT " %02x ", NIP6(rt->rt6i_src.addr),
		   rt->rt6i_src.plen);
L
Linus Torvalds 已提交
2383
#else
2384
	seq_puts(m, "00000000000000000000000000000000 00 ");
L
Linus Torvalds 已提交
2385 2386 2387
#endif

	if (rt->rt6i_nexthop) {
2388 2389
		seq_printf(m, NIP6_SEQFMT,
			   NIP6(*((struct in6_addr *)rt->rt6i_nexthop->primary_key)));
L
Linus Torvalds 已提交
2390
	} else {
2391
		seq_puts(m, "00000000000000000000000000000000");
L
Linus Torvalds 已提交
2392
	}
2393 2394 2395 2396
	seq_printf(m, " %08x %08x %08x %08x %8s\n",
		   rt->rt6i_metric, atomic_read(&rt->u.dst.__refcnt),
		   rt->u.dst.__use, rt->rt6i_flags,
		   rt->rt6i_dev ? rt->rt6i_dev->name : "");
L
Linus Torvalds 已提交
2397 2398 2399
	return 0;
}

2400
static int ipv6_route_show(struct seq_file *m, void *v)
L
Linus Torvalds 已提交
2401
{
2402 2403
	struct net *net = (struct net *)m->private;
	fib6_clean_all(net, rt6_info_route, 0, m);
2404 2405
	return 0;
}
L
Linus Torvalds 已提交
2406

2407 2408
static int ipv6_route_open(struct inode *inode, struct file *file)
{
2409
	int err;
2410 2411 2412
	struct net *net = get_proc_net(inode);
	if (!net)
		return -ENXIO;
2413 2414 2415 2416 2417 2418 2419 2420

	err = single_open(file, ipv6_route_show, net);
	if (err < 0) {
		put_net(net);
		return err;
	}

	return 0;
2421 2422 2423 2424 2425 2426 2427 2428
}

static int ipv6_route_release(struct inode *inode, struct file *file)
{
	struct seq_file *seq = file->private_data;
	struct net *net = seq->private;
	put_net(net);
	return single_release(inode, file);
L
Linus Torvalds 已提交
2429 2430
}

2431 2432 2433 2434 2435
static const struct file_operations ipv6_route_proc_fops = {
	.owner		= THIS_MODULE,
	.open		= ipv6_route_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
2436
	.release	= ipv6_route_release,
2437 2438
};

L
Linus Torvalds 已提交
2439 2440
static int rt6_stats_seq_show(struct seq_file *seq, void *v)
{
2441
	struct net *net = (struct net *)seq->private;
L
Linus Torvalds 已提交
2442
	seq_printf(seq, "%04x %04x %04x %04x %04x %04x %04x\n",
2443 2444 2445 2446 2447
		   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,
2448
		   atomic_read(&net->ipv6.ip6_dst_ops->entries),
2449
		   net->ipv6.rt6_stats->fib_discarded_routes);
L
Linus Torvalds 已提交
2450 2451 2452 2453 2454 2455

	return 0;
}

static int rt6_stats_seq_open(struct inode *inode, struct file *file)
{
2456
	int err;
2457
	struct net *net = get_proc_net(inode);
2458 2459 2460 2461 2462 2463 2464 2465 2466 2467
	if (!net)
		return -ENXIO;

	err = single_open(file, rt6_stats_seq_show, net);
	if (err < 0) {
		put_net(net);
		return err;
	}

	return 0;
2468 2469 2470 2471 2472 2473 2474 2475
}

static int rt6_stats_seq_release(struct inode *inode, struct file *file)
{
	struct seq_file *seq = file->private_data;
	struct net *net = (struct net *)seq->private;
	put_net(net);
	return single_release(inode, file);
L
Linus Torvalds 已提交
2476 2477
}

2478
static const struct file_operations rt6_stats_seq_fops = {
L
Linus Torvalds 已提交
2479 2480 2481 2482
	.owner	 = THIS_MODULE,
	.open	 = rt6_stats_seq_open,
	.read	 = seq_read,
	.llseek	 = seq_lseek,
2483
	.release = rt6_stats_seq_release,
L
Linus Torvalds 已提交
2484 2485 2486 2487 2488 2489 2490 2491 2492
};
#endif	/* CONFIG_PROC_FS */

#ifdef CONFIG_SYSCTL

static
int ipv6_sysctl_rtcache_flush(ctl_table *ctl, int write, struct file * filp,
			      void __user *buffer, size_t *lenp, loff_t *ppos)
{
2493 2494
	struct net *net = current->nsproxy->net_ns;
	int delay = net->ipv6.sysctl.flush_delay;
L
Linus Torvalds 已提交
2495 2496
	if (write) {
		proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
2497
		fib6_run_gc(delay <= 0 ? ~0UL : (unsigned long)delay, net);
L
Linus Torvalds 已提交
2498 2499 2500 2501 2502
		return 0;
	} else
		return -EINVAL;
}

2503
ctl_table ipv6_route_table_template[] = {
2504
	{
L
Linus Torvalds 已提交
2505
		.procname	=	"flush",
2506
		.data		=	&init_net.ipv6.sysctl.flush_delay,
L
Linus Torvalds 已提交
2507
		.maxlen		=	sizeof(int),
2508
		.mode		=	0200,
2509
		.proc_handler	=	&ipv6_sysctl_rtcache_flush
L
Linus Torvalds 已提交
2510 2511 2512 2513
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_GC_THRESH,
		.procname	=	"gc_thresh",
2514
		.data		=	&ip6_dst_ops_template.gc_thresh,
L
Linus Torvalds 已提交
2515 2516
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2517
		.proc_handler	=	&proc_dointvec,
L
Linus Torvalds 已提交
2518 2519 2520 2521
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_MAX_SIZE,
		.procname	=	"max_size",
2522
		.data		=	&init_net.ipv6.sysctl.ip6_rt_max_size,
L
Linus Torvalds 已提交
2523 2524
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2525
		.proc_handler	=	&proc_dointvec,
L
Linus Torvalds 已提交
2526 2527 2528 2529
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_GC_MIN_INTERVAL,
		.procname	=	"gc_min_interval",
2530
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_min_interval,
L
Linus Torvalds 已提交
2531 2532
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2533
		.proc_handler	=	&proc_dointvec_jiffies,
L
Linus Torvalds 已提交
2534 2535 2536 2537 2538
		.strategy	=	&sysctl_jiffies,
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_GC_TIMEOUT,
		.procname	=	"gc_timeout",
2539
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_timeout,
L
Linus Torvalds 已提交
2540 2541
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2542
		.proc_handler	=	&proc_dointvec_jiffies,
L
Linus Torvalds 已提交
2543 2544 2545 2546 2547
		.strategy	=	&sysctl_jiffies,
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_GC_INTERVAL,
		.procname	=	"gc_interval",
2548
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_interval,
L
Linus Torvalds 已提交
2549 2550
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2551
		.proc_handler	=	&proc_dointvec_jiffies,
L
Linus Torvalds 已提交
2552 2553 2554 2555 2556
		.strategy	=	&sysctl_jiffies,
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_GC_ELASTICITY,
		.procname	=	"gc_elasticity",
2557
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_elasticity,
L
Linus Torvalds 已提交
2558 2559
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2560
		.proc_handler	=	&proc_dointvec_jiffies,
L
Linus Torvalds 已提交
2561 2562 2563 2564 2565
		.strategy	=	&sysctl_jiffies,
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_MTU_EXPIRES,
		.procname	=	"mtu_expires",
2566
		.data		=	&init_net.ipv6.sysctl.ip6_rt_mtu_expires,
L
Linus Torvalds 已提交
2567 2568
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2569
		.proc_handler	=	&proc_dointvec_jiffies,
L
Linus Torvalds 已提交
2570 2571 2572 2573 2574
		.strategy	=	&sysctl_jiffies,
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_MIN_ADVMSS,
		.procname	=	"min_adv_mss",
2575
		.data		=	&init_net.ipv6.sysctl.ip6_rt_min_advmss,
L
Linus Torvalds 已提交
2576 2577
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2578
		.proc_handler	=	&proc_dointvec_jiffies,
L
Linus Torvalds 已提交
2579 2580 2581 2582 2583
		.strategy	=	&sysctl_jiffies,
	},
	{
		.ctl_name	=	NET_IPV6_ROUTE_GC_MIN_INTERVAL_MS,
		.procname	=	"gc_min_interval_ms",
2584
		.data		=	&init_net.ipv6.sysctl.ip6_rt_gc_min_interval,
L
Linus Torvalds 已提交
2585 2586
		.maxlen		=	sizeof(int),
		.mode		=	0644,
2587
		.proc_handler	=	&proc_dointvec_ms_jiffies,
L
Linus Torvalds 已提交
2588 2589 2590 2591 2592
		.strategy	=	&sysctl_ms_jiffies,
	},
	{ .ctl_name = 0 }
};

2593 2594 2595 2596 2597 2598 2599
struct ctl_table *ipv6_route_sysctl_init(struct net *net)
{
	struct ctl_table *table;

	table = kmemdup(ipv6_route_table_template,
			sizeof(ipv6_route_table_template),
			GFP_KERNEL);
2600 2601 2602

	if (table) {
		table[0].data = &net->ipv6.sysctl.flush_delay;
2603
		table[1].data = &net->ipv6.ip6_dst_ops->gc_thresh;
2604 2605 2606 2607 2608 2609 2610 2611 2612
		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;
	}

2613 2614
	return table;
}
L
Linus Torvalds 已提交
2615 2616
#endif

2617 2618
static int ip6_route_net_init(struct net *net)
{
2619
	int ret = -ENOMEM;
2620

2621 2622 2623 2624 2625
	net->ipv6.ip6_dst_ops = kmemdup(&ip6_dst_ops_template,
					sizeof(*net->ipv6.ip6_dst_ops),
					GFP_KERNEL);
	if (!net->ipv6.ip6_dst_ops)
		goto out;
2626
	net->ipv6.ip6_dst_ops->dst_net = hold_net(net);
2627

2628 2629 2630 2631
	net->ipv6.ip6_null_entry = kmemdup(&ip6_null_entry_template,
					   sizeof(*net->ipv6.ip6_null_entry),
					   GFP_KERNEL);
	if (!net->ipv6.ip6_null_entry)
2632
		goto out_ip6_dst_ops;
2633 2634
	net->ipv6.ip6_null_entry->u.dst.path =
		(struct dst_entry *)net->ipv6.ip6_null_entry;
2635
	net->ipv6.ip6_null_entry->u.dst.ops = net->ipv6.ip6_dst_ops;
2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646

#ifdef CONFIG_IPV6_MULTIPLE_TABLES
	net->ipv6.ip6_prohibit_entry = kmemdup(&ip6_prohibit_entry_template,
					       sizeof(*net->ipv6.ip6_prohibit_entry),
					       GFP_KERNEL);
	if (!net->ipv6.ip6_prohibit_entry) {
		kfree(net->ipv6.ip6_null_entry);
		goto out;
	}
	net->ipv6.ip6_prohibit_entry->u.dst.path =
		(struct dst_entry *)net->ipv6.ip6_prohibit_entry;
2647
	net->ipv6.ip6_prohibit_entry->u.dst.ops = net->ipv6.ip6_dst_ops;
2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658

	net->ipv6.ip6_blk_hole_entry = kmemdup(&ip6_blk_hole_entry_template,
					       sizeof(*net->ipv6.ip6_blk_hole_entry),
					       GFP_KERNEL);
	if (!net->ipv6.ip6_blk_hole_entry) {
		kfree(net->ipv6.ip6_null_entry);
		kfree(net->ipv6.ip6_prohibit_entry);
		goto out;
	}
	net->ipv6.ip6_blk_hole_entry->u.dst.path =
		(struct dst_entry *)net->ipv6.ip6_blk_hole_entry;
2659
	net->ipv6.ip6_blk_hole_entry->u.dst.ops = net->ipv6.ip6_dst_ops;
2660 2661
#endif

2662 2663 2664 2665
#ifdef CONFIG_PROC_FS
	proc_net_fops_create(net, "ipv6_route", 0, &ipv6_route_proc_fops);
	proc_net_fops_create(net, "rt6_stats", S_IRUGO, &rt6_stats_seq_fops);
#endif
2666 2667
	net->ipv6.ip6_rt_gc_expire = 30*HZ;

2668 2669 2670
	ret = 0;
out:
	return ret;
2671 2672

out_ip6_dst_ops:
2673
	release_net(net->ipv6.ip6_dst_ops->dst_net);
2674 2675
	kfree(net->ipv6.ip6_dst_ops);
	goto out;
2676 2677 2678 2679 2680 2681 2682 2683
}

static void ip6_route_net_exit(struct net *net)
{
#ifdef CONFIG_PROC_FS
	proc_net_remove(net, "ipv6_route");
	proc_net_remove(net, "rt6_stats");
#endif
2684 2685 2686 2687 2688
	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
2689
	release_net(net->ipv6.ip6_dst_ops->dst_net);
2690
	kfree(net->ipv6.ip6_dst_ops);
2691 2692 2693 2694 2695 2696 2697
}

static struct pernet_operations ip6_route_net_ops = {
	.init = ip6_route_net_init,
	.exit = ip6_route_net_exit,
};

2698 2699 2700 2701 2702
static struct notifier_block ip6_route_dev_notifier = {
	.notifier_call = ip6_route_dev_notify,
	.priority = 0,
};

2703
int __init ip6_route_init(void)
L
Linus Torvalds 已提交
2704
{
2705 2706
	int ret;

2707 2708
	ret = -ENOMEM;
	ip6_dst_ops_template.kmem_cachep =
A
Alexey Dobriyan 已提交
2709
		kmem_cache_create("ip6_dst_cache", sizeof(struct rt6_info), 0,
2710
				  SLAB_HWCACHE_ALIGN, NULL);
2711
	if (!ip6_dst_ops_template.kmem_cachep)
2712
		goto out;;
2713

2714 2715
	ret = register_pernet_subsys(&ip6_route_net_ops);
	if (ret)
2716 2717
		goto out_kmem_cache;

2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728
	/* 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 */
	init_net.ipv6.ip6_null_entry->u.dst.dev = init_net.loopback_dev;
	init_net.ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
  #ifdef CONFIG_IPV6_MULTIPLE_TABLES
	init_net.ipv6.ip6_prohibit_entry->u.dst.dev = init_net.loopback_dev;
	init_net.ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
	init_net.ipv6.ip6_blk_hole_entry->u.dst.dev = init_net.loopback_dev;
	init_net.ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
  #endif
2729 2730
	ret = fib6_init();
	if (ret)
2731
		goto out_register_subsys;
2732 2733 2734

	ret = xfrm6_init();
	if (ret)
2735
		goto out_fib6_init;
2736

2737 2738 2739
	ret = fib6_rules_init();
	if (ret)
		goto xfrm6_init;
2740

2741 2742 2743 2744 2745
	ret = -ENOBUFS;
	if (__rtnl_register(PF_INET6, RTM_NEWROUTE, inet6_rtm_newroute, NULL) ||
	    __rtnl_register(PF_INET6, RTM_DELROUTE, inet6_rtm_delroute, NULL) ||
	    __rtnl_register(PF_INET6, RTM_GETROUTE, inet6_rtm_getroute, NULL))
		goto fib6_rules_init;
2746

2747
	ret = register_netdevice_notifier(&ip6_route_dev_notifier);
2748 2749
	if (ret)
		goto fib6_rules_init;
2750

2751 2752 2753 2754 2755 2756 2757 2758 2759
out:
	return ret;

fib6_rules_init:
	fib6_rules_cleanup();
xfrm6_init:
	xfrm6_fini();
out_fib6_init:
	fib6_gc_cleanup();
2760 2761
out_register_subsys:
	unregister_pernet_subsys(&ip6_route_net_ops);
2762
out_kmem_cache:
2763
	kmem_cache_destroy(ip6_dst_ops_template.kmem_cachep);
2764
	goto out;
L
Linus Torvalds 已提交
2765 2766 2767 2768
}

void ip6_route_cleanup(void)
{
2769
	unregister_netdevice_notifier(&ip6_route_dev_notifier);
T
Thomas Graf 已提交
2770
	fib6_rules_cleanup();
L
Linus Torvalds 已提交
2771 2772
	xfrm6_fini();
	fib6_gc_cleanup();
2773
	unregister_pernet_subsys(&ip6_route_net_ops);
2774
	kmem_cache_destroy(ip6_dst_ops_template.kmem_cachep);
L
Linus Torvalds 已提交
2775
}