addrconf.c 128.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5
/*
 *	IPv6 Address [auto]configuration
 *	Linux INET6 implementation
 *
 *	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 25 26 27 28 29 30 31 32 33 34 35
 *	Alexey Kuznetsov	<kuznet@ms2.inr.ac.ru>
 *
 *	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:
 *
 *	Janos Farkas			:	delete timer on ifdown
 *	<chexum@bankinf.banki.hu>
 *	Andi Kleen			:	kill double kfree on module
 *						unload.
 *	Maciej W. Rozycki		:	FDDI support
 *	sekiya@USAGI			:	Don't send too many RS
 *						packets.
 *	yoshfuji@USAGI			:       Fixed interval between DAD
 *						packets.
 *	YOSHIFUJI Hideaki @USAGI	:	improved accuracy of
 *						address validation timer.
 *	YOSHIFUJI Hideaki @USAGI	:	Privacy Extensions (RFC3041)
 *						support.
 *	Yuji SEKIYA @USAGI		:	Don't assign a same IPv6
 *						address on a same interface.
 *	YOSHIFUJI Hideaki @USAGI	:	ARCnet support
 *	YOSHIFUJI Hideaki @USAGI	:	convert /proc/net/if_inet6 to
 *						seq_file.
36 37 38
 *	YOSHIFUJI Hideaki @USAGI	:	improved source address
 *						selection; consider scope,
 *						status etc.
L
Linus Torvalds 已提交
39 40
 */

41 42
#define pr_fmt(fmt) "IPv6: " fmt

L
Linus Torvalds 已提交
43 44
#include <linux/errno.h>
#include <linux/types.h>
45
#include <linux/kernel.h>
L
Linus Torvalds 已提交
46 47 48 49 50
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/in6.h>
#include <linux/netdevice.h>
51
#include <linux/if_addr.h>
L
Linus Torvalds 已提交
52 53 54 55 56 57
#include <linux/if_arp.h>
#include <linux/if_arcnet.h>
#include <linux/if_infiniband.h>
#include <linux/route.h>
#include <linux/inetdevice.h>
#include <linux/init.h>
58
#include <linux/slab.h>
L
Linus Torvalds 已提交
59 60 61
#ifdef CONFIG_SYSCTL
#include <linux/sysctl.h>
#endif
62
#include <linux/capability.h>
L
Linus Torvalds 已提交
63 64
#include <linux/delay.h>
#include <linux/notifier.h>
65
#include <linux/string.h>
E
Eric Dumazet 已提交
66
#include <linux/hash.h>
L
Linus Torvalds 已提交
67

68
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
69 70 71
#include <net/sock.h>
#include <net/snmp.h>

72
#include <net/af_ieee802154.h>
73
#include <net/firewire.h>
L
Linus Torvalds 已提交
74 75 76 77 78 79 80
#include <net/ipv6.h>
#include <net/protocol.h>
#include <net/ndisc.h>
#include <net/ip6_route.h>
#include <net/addrconf.h>
#include <net/tcp.h>
#include <net/ip.h>
81
#include <net/netlink.h>
82
#include <net/pkt_sched.h>
L
Linus Torvalds 已提交
83 84
#include <linux/if_tunnel.h>
#include <linux/rtnetlink.h>
85
#include <linux/netconf.h>
L
Linus Torvalds 已提交
86
#include <linux/random.h>
S
Stephen Hemminger 已提交
87
#include <linux/uaccess.h>
88
#include <asm/unaligned.h>
L
Linus Torvalds 已提交
89 90 91

#include <linux/proc_fs.h>
#include <linux/seq_file.h>
92
#include <linux/export.h>
L
Linus Torvalds 已提交
93 94 95 96 97

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

#if ACONF_DEBUG >= 3
98
#define ADBG(fmt, ...) printk(fmt, ##__VA_ARGS__)
L
Linus Torvalds 已提交
99
#else
100
#define ADBG(fmt, ...) do { if (0) printk(fmt, ##__VA_ARGS__); } while (0)
L
Linus Torvalds 已提交
101 102 103
#endif

#define	INFINITY_LIFE_TIME	0xFFFFFFFF
104 105 106 107 108

static inline u32 cstamp_delta(unsigned long cstamp)
{
	return (cstamp - INITIAL_JIFFIES) * 100UL / HZ;
}
L
Linus Torvalds 已提交
109 110

#ifdef CONFIG_SYSCTL
111
static void addrconf_sysctl_register(struct inet6_dev *idev);
112 113 114 115 116 117 118 119 120
static void addrconf_sysctl_unregister(struct inet6_dev *idev);
#else
static inline void addrconf_sysctl_register(struct inet6_dev *idev)
{
}

static inline void addrconf_sysctl_unregister(struct inet6_dev *idev)
{
}
L
Linus Torvalds 已提交
121 122
#endif

S
Sorin Dumitru 已提交
123 124
static void __ipv6_regen_rndid(struct inet6_dev *idev);
static void __ipv6_try_regen_rndid(struct inet6_dev *idev, struct in6_addr *tmpaddr);
L
Linus Torvalds 已提交
125 126
static void ipv6_regen_rndid(unsigned long data);

127
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev);
L
Linus Torvalds 已提交
128 129 130 131 132
static int ipv6_count_addresses(struct inet6_dev *idev);

/*
 *	Configured unicast address hash table
 */
133
static struct hlist_head inet6_addr_lst[IN6_ADDR_HSIZE];
134
static DEFINE_SPINLOCK(addrconf_hash_lock);
L
Linus Torvalds 已提交
135 136 137

static void addrconf_verify(unsigned long);

138
static DEFINE_TIMER(addr_chk_timer, addrconf_verify, 0, 0);
L
Linus Torvalds 已提交
139 140 141 142 143
static DEFINE_SPINLOCK(addrconf_verify_lock);

static void addrconf_join_anycast(struct inet6_ifaddr *ifp);
static void addrconf_leave_anycast(struct inet6_ifaddr *ifp);

144 145
static void addrconf_type_change(struct net_device *dev,
				 unsigned long event);
L
Linus Torvalds 已提交
146 147
static int addrconf_ifdown(struct net_device *dev, int how);

148 149 150 151 152
static struct rt6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags);

153
static void addrconf_dad_start(struct inet6_ifaddr *ifp);
L
Linus Torvalds 已提交
154 155
static void addrconf_dad_timer(unsigned long data);
static void addrconf_dad_completed(struct inet6_ifaddr *ifp);
156
static void addrconf_dad_run(struct inet6_dev *idev);
L
Linus Torvalds 已提交
157 158 159 160
static void addrconf_rs_timer(unsigned long data);
static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);
static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);

161
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
162
				struct prefix_info *pinfo);
163 164
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev);
L
Linus Torvalds 已提交
165

166
static struct ipv6_devconf ipv6_devconf __read_mostly = {
L
Linus Torvalds 已提交
167 168 169 170 171 172 173
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
	.force_mld_version	= 0,
174 175
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
176 177 178 179 180 181 182 183 184 185
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
	.use_tempaddr 		= 0,
	.temp_valid_lft		= TEMP_VALID_LIFETIME,
	.temp_prefered_lft	= TEMP_PREFERRED_LIFETIME,
	.regen_max_retry	= REGEN_MAX_RETRY,
	.max_desync_factor	= MAX_DESYNC_FACTOR,
	.max_addresses		= IPV6_MAX_ADDRESSES,
186
	.accept_ra_defrtr	= 1,
187
	.accept_ra_pinfo	= 1,
188 189
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
190
	.rtr_probe_interval	= 60 * HZ,
191 192 193
#ifdef CONFIG_IPV6_ROUTE_INFO
	.accept_ra_rt_info_max_plen = 0,
#endif
194
#endif
195
	.proxy_ndp		= 0,
196
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
197
	.disable_ipv6		= 0,
198
	.accept_dad		= 1,
199
	.suppress_frag_ndisc	= 1,
L
Linus Torvalds 已提交
200 201
};

202
static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
L
Linus Torvalds 已提交
203 204 205 206 207 208
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
209 210 211
	.force_mld_version	= 0,
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
212 213 214 215 216 217 218 219 220 221
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
	.use_tempaddr		= 0,
	.temp_valid_lft		= TEMP_VALID_LIFETIME,
	.temp_prefered_lft	= TEMP_PREFERRED_LIFETIME,
	.regen_max_retry	= REGEN_MAX_RETRY,
	.max_desync_factor	= MAX_DESYNC_FACTOR,
	.max_addresses		= IPV6_MAX_ADDRESSES,
222
	.accept_ra_defrtr	= 1,
223
	.accept_ra_pinfo	= 1,
224 225
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
226
	.rtr_probe_interval	= 60 * HZ,
227 228 229
#ifdef CONFIG_IPV6_ROUTE_INFO
	.accept_ra_rt_info_max_plen = 0,
#endif
230
#endif
231
	.proxy_ndp		= 0,
232
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
233
	.disable_ipv6		= 0,
234
	.accept_dad		= 1,
235
	.suppress_frag_ndisc	= 1,
L
Linus Torvalds 已提交
236 237
};

238
/* Check if a valid qdisc is available */
239
static inline bool addrconf_qdisc_ok(const struct net_device *dev)
240
{
241
	return !qdisc_tx_is_noop(dev);
242 243
}

244
static void addrconf_del_rs_timer(struct inet6_dev *idev)
L
Linus Torvalds 已提交
245
{
246 247 248 249 250 251 252
	if (del_timer(&idev->rs_timer))
		__in6_dev_put(idev);
}

static void addrconf_del_dad_timer(struct inet6_ifaddr *ifp)
{
	if (del_timer(&ifp->dad_timer))
L
Linus Torvalds 已提交
253 254 255
		__in6_ifa_put(ifp);
}

256 257 258 259 260 261 262
static void addrconf_mod_rs_timer(struct inet6_dev *idev,
				  unsigned long when)
{
	if (!timer_pending(&idev->rs_timer))
		in6_dev_hold(idev);
	mod_timer(&idev->rs_timer, jiffies + when);
}
L
Linus Torvalds 已提交
263

264 265
static void addrconf_mod_dad_timer(struct inet6_ifaddr *ifp,
				   unsigned long when)
L
Linus Torvalds 已提交
266
{
267
	if (!timer_pending(&ifp->dad_timer))
L
Linus Torvalds 已提交
268
		in6_ifa_hold(ifp);
269
	mod_timer(&ifp->dad_timer, jiffies + when);
L
Linus Torvalds 已提交
270 271
}

272 273
static int snmp6_alloc_dev(struct inet6_dev *idev)
{
274 275
	int i;

276
	if (snmp_mib_init((void __percpu **)idev->stats.ipv6,
277 278
			  sizeof(struct ipstats_mib),
			  __alignof__(struct ipstats_mib)) < 0)
279
		goto err_ip;
280 281 282 283 284 285 286 287 288 289 290 291

	for_each_possible_cpu(i) {
		struct ipstats_mib *addrconf_stats;
		addrconf_stats = per_cpu_ptr(idev->stats.ipv6[0], i);
		u64_stats_init(&addrconf_stats->syncp);
#if SNMP_ARRAY_SZ == 2
		addrconf_stats = per_cpu_ptr(idev->stats.ipv6[1], i);
		u64_stats_init(&addrconf_stats->syncp);
#endif
	}


292 293 294
	idev->stats.icmpv6dev = kzalloc(sizeof(struct icmpv6_mib_device),
					GFP_KERNEL);
	if (!idev->stats.icmpv6dev)
295
		goto err_icmp;
296 297 298
	idev->stats.icmpv6msgdev = kzalloc(sizeof(struct icmpv6msg_mib_device),
					   GFP_KERNEL);
	if (!idev->stats.icmpv6msgdev)
299
		goto err_icmpmsg;
300 301 302

	return 0;

303
err_icmpmsg:
304
	kfree(idev->stats.icmpv6dev);
305
err_icmp:
306
	snmp_mib_free((void __percpu **)idev->stats.ipv6);
307
err_ip:
308
	return -ENOMEM;
309 310
}

311
static struct inet6_dev *ipv6_add_dev(struct net_device *dev)
L
Linus Torvalds 已提交
312 313 314 315 316 317 318 319
{
	struct inet6_dev *ndev;

	ASSERT_RTNL();

	if (dev->mtu < IPV6_MIN_MTU)
		return NULL;

320
	ndev = kzalloc(sizeof(struct inet6_dev), GFP_KERNEL);
321

322 323
	if (ndev == NULL)
		return NULL;
324 325 326

	rwlock_init(&ndev->lock);
	ndev->dev = dev;
327
	INIT_LIST_HEAD(&ndev->addr_list);
328 329
	setup_timer(&ndev->rs_timer, addrconf_rs_timer,
		    (unsigned long)ndev);
330
	memcpy(&ndev->cnf, dev_net(dev)->ipv6.devconf_dflt, sizeof(ndev->cnf));
331 332 333 334 335 336 337
	ndev->cnf.mtu6 = dev->mtu;
	ndev->cnf.sysctl = NULL;
	ndev->nd_parms = neigh_parms_alloc(dev, &nd_tbl);
	if (ndev->nd_parms == NULL) {
		kfree(ndev);
		return NULL;
	}
338 339
	if (ndev->cnf.forwarding)
		dev_disable_lro(dev);
340 341
	/* We refer to the device */
	dev_hold(dev);
L
Linus Torvalds 已提交
342

343
	if (snmp6_alloc_dev(ndev) < 0) {
344
		ADBG(KERN_WARNING
345
			"%s: cannot allocate memory for statistics; dev=%s.\n",
346
			__func__, dev->name);
347
		neigh_parms_release(&nd_tbl, ndev->nd_parms);
R
Roy Li 已提交
348 349
		dev_put(dev);
		kfree(ndev);
350 351
		return NULL;
	}
L
Linus Torvalds 已提交
352

353
	if (snmp6_register_dev(ndev) < 0) {
354
		ADBG(KERN_WARNING
355
			"%s: cannot create /proc/net/dev_snmp6/%s\n",
356
			__func__, dev->name);
357 358 359 360 361 362 363 364 365 366
		neigh_parms_release(&nd_tbl, ndev->nd_parms);
		ndev->dead = 1;
		in6_dev_finish_destroy(ndev);
		return NULL;
	}

	/* One reference from device.  We must do this before
	 * we invoke __ipv6_regen_rndid().
	 */
	in6_dev_hold(ndev);
L
Linus Torvalds 已提交
367

368 369 370
	if (dev->flags & (IFF_NOARP | IFF_LOOPBACK))
		ndev->cnf.accept_dad = -1;

A
Amerigo Wang 已提交
371
#if IS_ENABLED(CONFIG_IPV6_SIT)
372
	if (dev->type == ARPHRD_SIT && (dev->priv_flags & IFF_ISATAP)) {
373
		pr_info("%s: Disabled Multicast RS\n", dev->name);
374 375 376 377
		ndev->cnf.rtr_solicits = 0;
	}
#endif

378
	INIT_LIST_HEAD(&ndev->tempaddr_list);
379
	setup_timer(&ndev->regen_timer, ipv6_regen_rndid, (unsigned long)ndev);
380 381
	if ((dev->flags&IFF_LOOPBACK) ||
	    dev->type == ARPHRD_TUNNEL ||
382
	    dev->type == ARPHRD_TUNNEL6 ||
383 384
	    dev->type == ARPHRD_SIT ||
	    dev->type == ARPHRD_NONE) {
385 386 387 388 389
		ndev->cnf.use_tempaddr = -1;
	} else {
		in6_dev_hold(ndev);
		ipv6_regen_rndid((unsigned long) ndev);
	}
390

391
	ndev->token = in6addr_any;
L
Linus Torvalds 已提交
392

393
	if (netif_running(dev) && addrconf_qdisc_ok(dev))
394 395
		ndev->if_flags |= IF_READY;

396 397
	ipv6_mc_init_dev(ndev);
	ndev->tstamp = jiffies;
398
	addrconf_sysctl_register(ndev);
399
	/* protected by rtnl_lock */
400
	rcu_assign_pointer(dev->ip6_ptr, ndev);
401

402 403 404
	/* Join interface-local all-node multicast group */
	ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allnodes);

405
	/* Join all-node multicast group */
406
	ipv6_dev_mc_inc(dev, &in6addr_linklocal_allnodes);
407

408
	/* Join all-router multicast group if forwarding is set */
L
Li Wei 已提交
409
	if (ndev->cnf.forwarding && (dev->flags & IFF_MULTICAST))
410 411
		ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);

L
Linus Torvalds 已提交
412 413 414
	return ndev;
}

415
static struct inet6_dev *ipv6_find_idev(struct net_device *dev)
L
Linus Torvalds 已提交
416 417 418 419 420
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
421 422 423 424
	idev = __in6_dev_get(dev);
	if (!idev) {
		idev = ipv6_add_dev(dev);
		if (!idev)
L
Linus Torvalds 已提交
425 426
			return NULL;
	}
427

L
Linus Torvalds 已提交
428 429 430 431 432
	if (dev->flags&IFF_UP)
		ipv6_mc_up(idev);
	return idev;
}

433 434 435 436 437
static int inet6_netconf_msgsize_devconf(int type)
{
	int size =  NLMSG_ALIGN(sizeof(struct netconfmsg))
		    + nla_total_size(4);	/* NETCONFA_IFINDEX */

438 439
	/* type -1 is used for ALL */
	if (type == -1 || type == NETCONFA_FORWARDING)
440
		size += nla_total_size(4);
441
#ifdef CONFIG_IPV6_MROUTE
442 443
	if (type == -1 || type == NETCONFA_MC_FORWARDING)
		size += nla_total_size(4);
444
#endif
445 446
	if (type == -1 || type == NETCONFA_PROXY_NEIGH)
		size += nla_total_size(4);
447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469

	return size;
}

static int inet6_netconf_fill_devconf(struct sk_buff *skb, int ifindex,
				      struct ipv6_devconf *devconf, u32 portid,
				      u32 seq, int event, unsigned int flags,
				      int type)
{
	struct nlmsghdr  *nlh;
	struct netconfmsg *ncm;

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct netconfmsg),
			flags);
	if (nlh == NULL)
		return -EMSGSIZE;

	ncm = nlmsg_data(nlh);
	ncm->ncm_family = AF_INET6;

	if (nla_put_s32(skb, NETCONFA_IFINDEX, ifindex) < 0)
		goto nla_put_failure;

470 471
	/* type -1 is used for ALL */
	if ((type == -1 || type == NETCONFA_FORWARDING) &&
472 473
	    nla_put_s32(skb, NETCONFA_FORWARDING, devconf->forwarding) < 0)
		goto nla_put_failure;
474
#ifdef CONFIG_IPV6_MROUTE
475 476 477 478
	if ((type == -1 || type == NETCONFA_MC_FORWARDING) &&
	    nla_put_s32(skb, NETCONFA_MC_FORWARDING,
			devconf->mc_forwarding) < 0)
		goto nla_put_failure;
479
#endif
480 481 482 483
	if ((type == -1 || type == NETCONFA_PROXY_NEIGH) &&
	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
		goto nla_put_failure;

484 485 486 487 488 489 490
	return nlmsg_end(skb, nlh);

nla_put_failure:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
}

491 492
void inet6_netconf_notify_devconf(struct net *net, int type, int ifindex,
				  struct ipv6_devconf *devconf)
493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511
{
	struct sk_buff *skb;
	int err = -ENOBUFS;

	skb = nlmsg_new(inet6_netconf_msgsize_devconf(type), GFP_ATOMIC);
	if (skb == NULL)
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf, 0, 0,
					 RTM_NEWNETCONF, 0, type);
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_NETCONF, NULL, GFP_ATOMIC);
	return;
errout:
512
	rtnl_set_sk_err(net, RTNLGRP_IPV6_NETCONF, err);
513 514
}

515 516 517
static const struct nla_policy devconf_ipv6_policy[NETCONFA_MAX+1] = {
	[NETCONFA_IFINDEX]	= { .len = sizeof(int) },
	[NETCONFA_FORWARDING]	= { .len = sizeof(int) },
518
	[NETCONFA_PROXY_NEIGH]	= { .len = sizeof(int) },
519 520 521
};

static int inet6_netconf_get_devconf(struct sk_buff *in_skb,
522
				     struct nlmsghdr *nlh)
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581
{
	struct net *net = sock_net(in_skb->sk);
	struct nlattr *tb[NETCONFA_MAX+1];
	struct netconfmsg *ncm;
	struct sk_buff *skb;
	struct ipv6_devconf *devconf;
	struct inet6_dev *in6_dev;
	struct net_device *dev;
	int ifindex;
	int err;

	err = nlmsg_parse(nlh, sizeof(*ncm), tb, NETCONFA_MAX,
			  devconf_ipv6_policy);
	if (err < 0)
		goto errout;

	err = EINVAL;
	if (!tb[NETCONFA_IFINDEX])
		goto errout;

	ifindex = nla_get_s32(tb[NETCONFA_IFINDEX]);
	switch (ifindex) {
	case NETCONFA_IFINDEX_ALL:
		devconf = net->ipv6.devconf_all;
		break;
	case NETCONFA_IFINDEX_DEFAULT:
		devconf = net->ipv6.devconf_dflt;
		break;
	default:
		dev = __dev_get_by_index(net, ifindex);
		if (dev == NULL)
			goto errout;
		in6_dev = __in6_dev_get(dev);
		if (in6_dev == NULL)
			goto errout;
		devconf = &in6_dev->cnf;
		break;
	}

	err = -ENOBUFS;
	skb = nlmsg_new(inet6_netconf_msgsize_devconf(-1), GFP_ATOMIC);
	if (skb == NULL)
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf,
					 NETLINK_CB(in_skb).portid,
					 nlh->nlmsg_seq, RTM_NEWNETCONF, 0,
					 -1);
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
errout:
	return err;
}

582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598
static int inet6_netconf_dump_devconf(struct sk_buff *skb,
				      struct netlink_callback *cb)
{
	struct net *net = sock_net(skb->sk);
	int h, s_h;
	int idx, s_idx;
	struct net_device *dev;
	struct inet6_dev *idev;
	struct hlist_head *head;

	s_h = cb->args[0];
	s_idx = idx = cb->args[1];

	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
		rcu_read_lock();
599 600
		cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^
			  net->dev_base_seq;
601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;

			if (inet6_netconf_fill_devconf(skb, dev->ifindex,
						       &idev->cnf,
						       NETLINK_CB(cb->skb).portid,
						       cb->nlh->nlmsg_seq,
						       RTM_NEWNETCONF,
						       NLM_F_MULTI,
						       -1) <= 0) {
				rcu_read_unlock();
				goto done;
			}
618
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652
cont:
			idx++;
		}
		rcu_read_unlock();
	}
	if (h == NETDEV_HASHENTRIES) {
		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_ALL,
					       net->ipv6.devconf_all,
					       NETLINK_CB(cb->skb).portid,
					       cb->nlh->nlmsg_seq,
					       RTM_NEWNETCONF, NLM_F_MULTI,
					       -1) <= 0)
			goto done;
		else
			h++;
	}
	if (h == NETDEV_HASHENTRIES + 1) {
		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_DEFAULT,
					       net->ipv6.devconf_dflt,
					       NETLINK_CB(cb->skb).portid,
					       cb->nlh->nlmsg_seq,
					       RTM_NEWNETCONF, NLM_F_MULTI,
					       -1) <= 0)
			goto done;
		else
			h++;
	}
done:
	cb->args[0] = h;
	cb->args[1] = idx;

	return skb->len;
}

L
Linus Torvalds 已提交
653 654 655 656 657 658 659 660 661
#ifdef CONFIG_SYSCTL
static void dev_forward_change(struct inet6_dev *idev)
{
	struct net_device *dev;
	struct inet6_ifaddr *ifa;

	if (!idev)
		return;
	dev = idev->dev;
662 663
	if (idev->cnf.forwarding)
		dev_disable_lro(dev);
A
Amerigo Wang 已提交
664
	if (dev->flags & IFF_MULTICAST) {
665
		if (idev->cnf.forwarding) {
666
			ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);
667 668 669
			ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_inc(dev, &in6addr_sitelocal_allrouters);
		} else {
670
			ipv6_dev_mc_dec(dev, &in6addr_linklocal_allrouters);
671 672 673
			ipv6_dev_mc_dec(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_dec(dev, &in6addr_sitelocal_allrouters);
		}
L
Linus Torvalds 已提交
674
	}
675 676

	list_for_each_entry(ifa, &idev->addr_list, if_list) {
M
Michal Wrobel 已提交
677 678
		if (ifa->flags&IFA_F_TENTATIVE)
			continue;
L
Linus Torvalds 已提交
679 680 681 682 683
		if (idev->cnf.forwarding)
			addrconf_join_anycast(ifa);
		else
			addrconf_leave_anycast(ifa);
	}
684 685
	inet6_netconf_notify_devconf(dev_net(dev), NETCONFA_FORWARDING,
				     dev->ifindex, &idev->cnf);
L
Linus Torvalds 已提交
686 687 688
}


689
static void addrconf_forward_change(struct net *net, __s32 newf)
L
Linus Torvalds 已提交
690 691 692 693
{
	struct net_device *dev;
	struct inet6_dev *idev;

694
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
695 696
		idev = __in6_dev_get(dev);
		if (idev) {
697 698
			int changed = (!idev->cnf.forwarding) ^ (!newf);
			idev->cnf.forwarding = newf;
L
Linus Torvalds 已提交
699 700 701 702 703
			if (changed)
				dev_forward_change(idev);
		}
	}
}
704

705
static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
706
{
707
	struct net *net;
708 709 710 711
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
712 713

	net = (struct net *)table->extra2;
714 715
	old = *p;
	*p = newf;
716

717
	if (p == &net->ipv6.devconf_dflt->forwarding) {
718 719 720 721
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net, NETCONFA_FORWARDING,
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
722 723
		rtnl_unlock();
		return 0;
E
Eric W. Biederman 已提交
724
	}
725

726 727 728
	if (p == &net->ipv6.devconf_all->forwarding) {
		net->ipv6.devconf_dflt->forwarding = newf;
		addrconf_forward_change(net, newf);
729 730 731 732
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net, NETCONFA_FORWARDING,
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
733
	} else if ((!newf) ^ (!old))
734
		dev_forward_change((struct inet6_dev *)table->extra1);
735
	rtnl_unlock();
736

737
	if (newf)
738
		rt6_purge_dflt_routers(net);
739
	return 1;
740
}
L
Linus Torvalds 已提交
741 742
#endif

743
/* Nobody refers to this ifaddr, destroy it */
L
Linus Torvalds 已提交
744 745
void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp)
{
746
	WARN_ON(!hlist_unhashed(&ifp->addr_lst));
747

L
Linus Torvalds 已提交
748
#ifdef NET_REFCNT_DEBUG
749
	pr_debug("%s\n", __func__);
L
Linus Torvalds 已提交
750 751 752 753
#endif

	in6_dev_put(ifp->idev);

754
	if (del_timer(&ifp->dad_timer))
S
Stephen Hemminger 已提交
755
		pr_notice("Timer is still running, when freeing ifa=%p\n", ifp);
L
Linus Torvalds 已提交
756

757
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
758
		pr_warn("Freeing alive inet6 address %p\n", ifp);
L
Linus Torvalds 已提交
759 760
		return;
	}
A
Amerigo Wang 已提交
761
	ip6_rt_put(ifp->rt);
L
Linus Torvalds 已提交
762

763
	kfree_rcu(ifp, rcu);
L
Linus Torvalds 已提交
764 765
}

B
Brian Haley 已提交
766 767 768
static void
ipv6_link_dev_addr(struct inet6_dev *idev, struct inet6_ifaddr *ifp)
{
769
	struct list_head *p;
770
	int ifp_scope = ipv6_addr_src_scope(&ifp->addr);
B
Brian Haley 已提交
771 772 773 774 775

	/*
	 * Each device address list is sorted in order of scope -
	 * global before linklocal.
	 */
776 777 778
	list_for_each(p, &idev->addr_list) {
		struct inet6_ifaddr *ifa
			= list_entry(p, struct inet6_ifaddr, if_list);
779
		if (ifp_scope >= ipv6_addr_src_scope(&ifa->addr))
B
Brian Haley 已提交
780 781 782
			break;
	}

783
	list_add_tail(&ifp->if_list, p);
B
Brian Haley 已提交
784 785
}

E
Eric Dumazet 已提交
786
static u32 inet6_addr_hash(const struct in6_addr *addr)
787
{
E
Eric Dumazet 已提交
788
	return hash_32(ipv6_addr_hash(addr), IN6_ADDR_HSIZE_SHIFT);
789 790
}

L
Linus Torvalds 已提交
791 792 793
/* On success it returns ifp with increased reference count */

static struct inet6_ifaddr *
794 795
ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
	      const struct in6_addr *peer_addr, int pfxlen,
796
	      int scope, u32 flags, u32 valid_lft, u32 prefered_lft)
L
Linus Torvalds 已提交
797 798 799
{
	struct inet6_ifaddr *ifa = NULL;
	struct rt6_info *rt;
800
	unsigned int hash;
L
Linus Torvalds 已提交
801
	int err = 0;
802 803 804 805 806 807 808
	int addr_type = ipv6_addr_type(addr);

	if (addr_type == IPV6_ADDR_ANY ||
	    addr_type & IPV6_ADDR_MULTICAST ||
	    (!(idev->dev->flags & IFF_LOOPBACK) &&
	     addr_type & IPV6_ADDR_LOOPBACK))
		return ERR_PTR(-EADDRNOTAVAIL);
L
Linus Torvalds 已提交
809

810
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
811 812 813 814 815
	if (idev->dead) {
		err = -ENODEV;			/*XXX*/
		goto out2;
	}

816
	if (idev->cnf.disable_ipv6) {
817 818 819 820
		err = -EACCES;
		goto out2;
	}

821
	spin_lock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
822 823

	/* Ignore adding duplicate addresses on an interface */
824
	if (ipv6_chk_same_addr(dev_net(idev->dev), addr, idev->dev)) {
825
		ADBG("ipv6_add_addr: already assigned\n");
L
Linus Torvalds 已提交
826 827 828 829
		err = -EEXIST;
		goto out;
	}

830
	ifa = kzalloc(sizeof(struct inet6_ifaddr), GFP_ATOMIC);
L
Linus Torvalds 已提交
831 832

	if (ifa == NULL) {
833
		ADBG("ipv6_add_addr: malloc failed\n");
L
Linus Torvalds 已提交
834 835 836 837
		err = -ENOBUFS;
		goto out;
	}

838
	rt = addrconf_dst_alloc(idev, addr, false);
L
Linus Torvalds 已提交
839 840 841 842 843
	if (IS_ERR(rt)) {
		err = PTR_ERR(rt);
		goto out;
	}

844 845
	neigh_parms_data_state_setall(idev->nd_parms);

A
Alexey Dobriyan 已提交
846
	ifa->addr = *addr;
847 848
	if (peer_addr)
		ifa->peer_addr = *peer_addr;
L
Linus Torvalds 已提交
849 850

	spin_lock_init(&ifa->lock);
851
	spin_lock_init(&ifa->state_lock);
852 853
	setup_timer(&ifa->dad_timer, addrconf_dad_timer,
		    (unsigned long)ifa);
854
	INIT_HLIST_NODE(&ifa->addr_lst);
L
Linus Torvalds 已提交
855 856 857
	ifa->scope = scope;
	ifa->prefix_len = pfxlen;
	ifa->flags = flags | IFA_F_TENTATIVE;
858 859
	ifa->valid_lft = valid_lft;
	ifa->prefered_lft = prefered_lft;
L
Linus Torvalds 已提交
860
	ifa->cstamp = ifa->tstamp = jiffies;
861
	ifa->tokenized = false;
L
Linus Torvalds 已提交
862

863 864
	ifa->rt = rt;

L
Linus Torvalds 已提交
865 866 867 868 869 870
	ifa->idev = idev;
	in6_dev_hold(idev);
	/* For caller */
	in6_ifa_hold(ifa);

	/* Add to big hash table */
E
Eric Dumazet 已提交
871
	hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
872

873 874
	hlist_add_head_rcu(&ifa->addr_lst, &inet6_addr_lst[hash]);
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
875 876 877

	write_lock(&idev->lock);
	/* Add to inet6_dev unicast addr list. */
B
Brian Haley 已提交
878
	ipv6_link_dev_addr(idev, ifa);
L
Linus Torvalds 已提交
879 880

	if (ifa->flags&IFA_F_TEMPORARY) {
881
		list_add(&ifa->tmp_list, &idev->tempaddr_list);
L
Linus Torvalds 已提交
882 883 884 885 886 887
		in6_ifa_hold(ifa);
	}

	in6_ifa_hold(ifa);
	write_unlock(&idev->lock);
out2:
888
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
889

890
	if (likely(err == 0))
891
		inet6addr_notifier_call_chain(NETDEV_UP, ifa);
L
Linus Torvalds 已提交
892 893 894 895 896 897 898
	else {
		kfree(ifa);
		ifa = ERR_PTR(err);
	}

	return ifa;
out:
899
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
900 901 902
	goto out2;
}

903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984
enum cleanup_prefix_rt_t {
	CLEANUP_PREFIX_RT_NOP,    /* no cleanup action for prefix route */
	CLEANUP_PREFIX_RT_DEL,    /* delete the prefix route */
	CLEANUP_PREFIX_RT_EXPIRE, /* update the lifetime of the prefix route */
};

/*
 * Check, whether the prefix for ifp would still need a prefix route
 * after deleting ifp. The function returns one of the CLEANUP_PREFIX_RT_*
 * constants.
 *
 * 1) we don't purge prefix if address was not permanent.
 *    prefix is managed by its own lifetime.
 * 2) we also don't purge, if the address was IFA_F_NOPREFIXROUTE.
 * 3) if there are no addresses, delete prefix.
 * 4) if there are still other permanent address(es),
 *    corresponding prefix is still permanent.
 * 5) if there are still other addresses with IFA_F_NOPREFIXROUTE,
 *    don't purge the prefix, assume user space is managing it.
 * 6) otherwise, update prefix lifetime to the
 *    longest valid lifetime among the corresponding
 *    addresses on the device.
 *    Note: subsequent RA will update lifetime.
 **/
static enum cleanup_prefix_rt_t
check_cleanup_prefix_route(struct inet6_ifaddr *ifp, unsigned long *expires)
{
	struct inet6_ifaddr *ifa;
	struct inet6_dev *idev = ifp->idev;
	unsigned long lifetime;
	enum cleanup_prefix_rt_t action = CLEANUP_PREFIX_RT_DEL;

	*expires = jiffies;

	list_for_each_entry(ifa, &idev->addr_list, if_list) {
		if (ifa == ifp)
			continue;
		if (!ipv6_prefix_equal(&ifa->addr, &ifp->addr,
				       ifp->prefix_len))
			continue;
		if (ifa->flags & (IFA_F_PERMANENT | IFA_F_NOPREFIXROUTE))
			return CLEANUP_PREFIX_RT_NOP;

		action = CLEANUP_PREFIX_RT_EXPIRE;

		spin_lock(&ifa->lock);

		lifetime = addrconf_timeout_fixup(ifa->valid_lft, HZ);
		/*
		 * Note: Because this address is
		 * not permanent, lifetime <
		 * LONG_MAX / HZ here.
		 */
		if (time_before(*expires, ifa->tstamp + lifetime * HZ))
			*expires = ifa->tstamp + lifetime * HZ;
		spin_unlock(&ifa->lock);
	}

	return action;
}

static void
cleanup_prefix_route(struct inet6_ifaddr *ifp, unsigned long expires, bool del_rt)
{
	struct rt6_info *rt;

	rt = addrconf_get_prefix_route(&ifp->addr,
				       ifp->prefix_len,
				       ifp->idev->dev,
				       0, RTF_GATEWAY | RTF_DEFAULT);
	if (rt) {
		if (del_rt)
			ip6_del_rt(rt);
		else {
			if (!(rt->rt6i_flags & RTF_EXPIRES))
				rt6_set_expires(rt, expires);
			ip6_rt_put(rt);
		}
	}
}


L
Linus Torvalds 已提交
985 986 987 988
/* This function wants to get referenced ifp and releases it before return */

static void ipv6_del_addr(struct inet6_ifaddr *ifp)
{
989
	int state;
990 991
	enum cleanup_prefix_rt_t action = CLEANUP_PREFIX_RT_NOP;
	unsigned long expires;
L
Linus Torvalds 已提交
992

993 994
	spin_lock_bh(&ifp->state_lock);
	state = ifp->state;
995
	ifp->state = INET6_IFADDR_STATE_DEAD;
996 997 998 999
	spin_unlock_bh(&ifp->state_lock);

	if (state == INET6_IFADDR_STATE_DEAD)
		goto out;
L
Linus Torvalds 已提交
1000

1001 1002 1003
	spin_lock_bh(&addrconf_hash_lock);
	hlist_del_init_rcu(&ifp->addr_lst);
	spin_unlock_bh(&addrconf_hash_lock);
L
Linus Torvalds 已提交
1004

1005
	write_lock_bh(&ifp->idev->lock);
1006

L
Linus Torvalds 已提交
1007
	if (ifp->flags&IFA_F_TEMPORARY) {
1008 1009 1010 1011
		list_del(&ifp->tmp_list);
		if (ifp->ifpub) {
			in6_ifa_put(ifp->ifpub);
			ifp->ifpub = NULL;
L
Linus Torvalds 已提交
1012
		}
1013
		__in6_ifa_put(ifp);
L
Linus Torvalds 已提交
1014 1015
	}

1016 1017
	if (ifp->flags & IFA_F_PERMANENT && !(ifp->flags & IFA_F_NOPREFIXROUTE))
		action = check_cleanup_prefix_route(ifp, &expires);
1018

1019 1020 1021 1022
	list_del_init(&ifp->if_list);
	__in6_ifa_put(ifp);

	write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
1023

1024
	addrconf_del_dad_timer(ifp);
1025

L
Linus Torvalds 已提交
1026 1027
	ipv6_ifa_notify(RTM_DELADDR, ifp);

1028
	inet6addr_notifier_call_chain(NETDEV_DOWN, ifp);
L
Linus Torvalds 已提交
1029

1030 1031 1032
	if (action != CLEANUP_PREFIX_RT_NOP) {
		cleanup_prefix_route(ifp, expires,
			action == CLEANUP_PREFIX_RT_DEL);
L
Linus Torvalds 已提交
1033 1034
	}

1035 1036
	/* clean up prefsrc entries */
	rt6_remove_prefsrc(ifp);
1037
out:
L
Linus Torvalds 已提交
1038 1039 1040 1041 1042 1043 1044
	in6_ifa_put(ifp);
}

static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp, struct inet6_ifaddr *ift)
{
	struct inet6_dev *idev = ifp->idev;
	struct in6_addr addr, *tmpaddr;
1045
	unsigned long tmp_prefered_lft, tmp_valid_lft, tmp_tstamp, age;
1046
	unsigned long regen_advance;
L
Linus Torvalds 已提交
1047 1048
	int tmp_plen;
	int ret = 0;
1049
	u32 addr_flags;
1050
	unsigned long now = jiffies;
L
Linus Torvalds 已提交
1051

1052
	write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063
	if (ift) {
		spin_lock_bh(&ift->lock);
		memcpy(&addr.s6_addr[8], &ift->addr.s6_addr[8], 8);
		spin_unlock_bh(&ift->lock);
		tmpaddr = &addr;
	} else {
		tmpaddr = NULL;
	}
retry:
	in6_dev_hold(idev);
	if (idev->cnf.use_tempaddr <= 0) {
1064
		write_unlock_bh(&idev->lock);
1065
		pr_info("%s: use_tempaddr is disabled\n", __func__);
L
Linus Torvalds 已提交
1066 1067 1068 1069 1070 1071 1072 1073
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}
	spin_lock_bh(&ifp->lock);
	if (ifp->regen_count++ >= idev->cnf.regen_max_retry) {
		idev->cnf.use_tempaddr = -1;	/*XXX*/
		spin_unlock_bh(&ifp->lock);
1074
		write_unlock_bh(&idev->lock);
1075 1076
		pr_warn("%s: regeneration time exceeded - disabled temporary address support\n",
			__func__);
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081 1082
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}
	in6_ifa_hold(ifp);
	memcpy(addr.s6_addr, ifp->addr.s6_addr, 8);
S
Sorin Dumitru 已提交
1083
	__ipv6_try_regen_rndid(idev, tmpaddr);
L
Linus Torvalds 已提交
1084
	memcpy(&addr.s6_addr[8], idev->rndid, 8);
1085
	age = (now - ifp->tstamp) / HZ;
L
Linus Torvalds 已提交
1086 1087
	tmp_valid_lft = min_t(__u32,
			      ifp->valid_lft,
1088
			      idev->cnf.temp_valid_lft + age);
1089 1090
	tmp_prefered_lft = min_t(__u32,
				 ifp->prefered_lft,
1091
				 idev->cnf.temp_prefered_lft + age -
1092
				 idev->cnf.max_desync_factor);
L
Linus Torvalds 已提交
1093 1094 1095 1096
	tmp_plen = ifp->prefix_len;
	tmp_tstamp = ifp->tstamp;
	spin_unlock_bh(&ifp->lock);

1097
	regen_advance = idev->cnf.regen_max_retry *
1098
	                idev->cnf.dad_transmits *
J
Jiri Pirko 已提交
1099
	                NEIGH_VAR(idev->nd_parms, RETRANS_TIME) / HZ;
1100
	write_unlock_bh(&idev->lock);
1101

1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113
	/* A temporary address is created only if this calculated Preferred
	 * Lifetime is greater than REGEN_ADVANCE time units.  In particular,
	 * an implementation must not create a temporary address with a zero
	 * Preferred Lifetime.
	 */
	if (tmp_prefered_lft <= regen_advance) {
		in6_ifa_put(ifp);
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}

1114 1115 1116 1117 1118
	addr_flags = IFA_F_TEMPORARY;
	/* set in addrconf_prefix_rcv() */
	if (ifp->flags & IFA_F_OPTIMISTIC)
		addr_flags |= IFA_F_OPTIMISTIC;

1119 1120 1121 1122
	ift = ipv6_add_addr(idev, &addr, NULL, tmp_plen,
			    ipv6_addr_scope(&addr), addr_flags,
			    tmp_valid_lft, tmp_prefered_lft);
	if (IS_ERR(ift)) {
L
Linus Torvalds 已提交
1123 1124
		in6_ifa_put(ifp);
		in6_dev_put(idev);
1125
		pr_info("%s: retry temporary address regeneration\n", __func__);
L
Linus Torvalds 已提交
1126
		tmpaddr = &addr;
1127
		write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1128 1129 1130 1131 1132
		goto retry;
	}

	spin_lock_bh(&ift->lock);
	ift->ifpub = ifp;
1133
	ift->cstamp = now;
L
Linus Torvalds 已提交
1134 1135 1136
	ift->tstamp = tmp_tstamp;
	spin_unlock_bh(&ift->lock);

1137
	addrconf_dad_start(ift);
L
Linus Torvalds 已提交
1138 1139 1140 1141 1142 1143 1144
	in6_ifa_put(ift);
	in6_dev_put(idev);
out:
	return ret;
}

/*
1145
 *	Choose an appropriate source address (RFC3484)
L
Linus Torvalds 已提交
1146
 */
1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162
enum {
	IPV6_SADDR_RULE_INIT = 0,
	IPV6_SADDR_RULE_LOCAL,
	IPV6_SADDR_RULE_SCOPE,
	IPV6_SADDR_RULE_PREFERRED,
#ifdef CONFIG_IPV6_MIP6
	IPV6_SADDR_RULE_HOA,
#endif
	IPV6_SADDR_RULE_OIF,
	IPV6_SADDR_RULE_LABEL,
	IPV6_SADDR_RULE_PRIVACY,
	IPV6_SADDR_RULE_ORCHID,
	IPV6_SADDR_RULE_PREFIX,
	IPV6_SADDR_RULE_MAX
};

1163
struct ipv6_saddr_score {
1164 1165 1166 1167 1168 1169
	int			rule;
	int			addr_type;
	struct inet6_ifaddr	*ifa;
	DECLARE_BITMAP(scorebits, IPV6_SADDR_RULE_MAX);
	int			scopedist;
	int			matchlen;
1170 1171
};

1172
struct ipv6_saddr_dst {
1173
	const struct in6_addr *addr;
1174 1175 1176
	int ifindex;
	int scope;
	int label;
1177
	unsigned int prefs;
1178
};
1179

D
Dave Jones 已提交
1180
static inline int ipv6_saddr_preferred(int type)
L
Linus Torvalds 已提交
1181
{
U
Ulrich Weber 已提交
1182
	if (type & (IPV6_ADDR_MAPPED|IPV6_ADDR_COMPATv4|IPV6_ADDR_LOOPBACK))
1183 1184
		return 1;
	return 0;
L
Linus Torvalds 已提交
1185 1186
}

1187 1188
static int ipv6_get_saddr_eval(struct net *net,
			       struct ipv6_saddr_score *score,
1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227
			       struct ipv6_saddr_dst *dst,
			       int i)
{
	int ret;

	if (i <= score->rule) {
		switch (i) {
		case IPV6_SADDR_RULE_SCOPE:
			ret = score->scopedist;
			break;
		case IPV6_SADDR_RULE_PREFIX:
			ret = score->matchlen;
			break;
		default:
			ret = !!test_bit(i, score->scorebits);
		}
		goto out;
	}

	switch (i) {
	case IPV6_SADDR_RULE_INIT:
		/* Rule 0: remember if hiscore is not ready yet */
		ret = !!score->ifa;
		break;
	case IPV6_SADDR_RULE_LOCAL:
		/* Rule 1: Prefer same address */
		ret = ipv6_addr_equal(&score->ifa->addr, dst->addr);
		break;
	case IPV6_SADDR_RULE_SCOPE:
		/* Rule 2: Prefer appropriate scope
		 *
		 *      ret
		 *       ^
		 *    -1 |  d 15
		 *    ---+--+-+---> scope
		 *       |
		 *       |             d is scope of the destination.
		 *  B-d  |  \
		 *       |   \      <- smaller scope is better if
1228
		 *  B-15 |    \        if scope is enough for destination.
1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252
		 *       |             ret = B - scope (-1 <= scope >= d <= 15).
		 * d-C-1 | /
		 *       |/         <- greater is better
		 *   -C  /             if scope is not enough for destination.
		 *      /|             ret = scope - C (-1 <= d < scope <= 15).
		 *
		 * d - C - 1 < B -15 (for all -1 <= d <= 15).
		 * C > d + 14 - B >= 15 + 14 - B = 29 - B.
		 * Assume B = 0 and we get C > 29.
		 */
		ret = __ipv6_addr_src_scope(score->addr_type);
		if (ret >= dst->scope)
			ret = -ret;
		else
			ret -= 128;	/* 30 is enough */
		score->scopedist = ret;
		break;
	case IPV6_SADDR_RULE_PREFERRED:
		/* Rule 3: Avoid deprecated and optimistic addresses */
		ret = ipv6_saddr_preferred(score->addr_type) ||
		      !(score->ifa->flags & (IFA_F_DEPRECATED|IFA_F_OPTIMISTIC));
		break;
#ifdef CONFIG_IPV6_MIP6
	case IPV6_SADDR_RULE_HOA:
1253
	    {
1254
		/* Rule 4: Prefer home address */
1255 1256
		int prefhome = !(dst->prefs & IPV6_PREFER_SRC_COA);
		ret = !(score->ifa->flags & IFA_F_HOMEADDRESS) ^ prefhome;
1257
		break;
1258
	    }
1259 1260 1261 1262 1263 1264 1265 1266
#endif
	case IPV6_SADDR_RULE_OIF:
		/* Rule 5: Prefer outgoing interface */
		ret = (!dst->ifindex ||
		       dst->ifindex == score->ifa->idev->dev->ifindex);
		break;
	case IPV6_SADDR_RULE_LABEL:
		/* Rule 6: Prefer matching label */
1267 1268
		ret = ipv6_addr_label(net,
				      &score->ifa->addr, score->addr_type,
1269 1270 1271
				      score->ifa->idev->dev->ifindex) == dst->label;
		break;
	case IPV6_SADDR_RULE_PRIVACY:
1272
	    {
1273
		/* Rule 7: Prefer public address
L
Lucas De Marchi 已提交
1274
		 * Note: prefer temporary address if use_tempaddr >= 2
1275
		 */
1276 1277 1278 1279
		int preftmp = dst->prefs & (IPV6_PREFER_SRC_PUBLIC|IPV6_PREFER_SRC_TMP) ?
				!!(dst->prefs & IPV6_PREFER_SRC_TMP) :
				score->ifa->idev->cnf.use_tempaddr >= 2;
		ret = (!(score->ifa->flags & IFA_F_TEMPORARY)) ^ preftmp;
1280
		break;
1281
	    }
1282 1283 1284 1285 1286 1287 1288 1289 1290
	case IPV6_SADDR_RULE_ORCHID:
		/* Rule 8-: Prefer ORCHID vs ORCHID or
		 *	    non-ORCHID vs non-ORCHID
		 */
		ret = !(ipv6_addr_orchid(&score->ifa->addr) ^
			ipv6_addr_orchid(dst->addr));
		break;
	case IPV6_SADDR_RULE_PREFIX:
		/* Rule 8: Use longest matching prefix */
1291 1292 1293 1294
		ret = ipv6_addr_diff(&score->ifa->addr, dst->addr);
		if (ret > score->ifa->prefix_len)
			ret = score->ifa->prefix_len;
		score->matchlen = ret;
1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306
		break;
	default:
		ret = 0;
	}

	if (ret)
		__set_bit(i, score->scorebits);
	score->rule = i;
out:
	return ret;
}

1307
int ipv6_dev_get_saddr(struct net *net, const struct net_device *dst_dev,
1308
		       const struct in6_addr *daddr, unsigned int prefs,
1309
		       struct in6_addr *saddr)
L
Linus Torvalds 已提交
1310
{
1311 1312 1313
	struct ipv6_saddr_score scores[2],
				*score = &scores[0], *hiscore = &scores[1];
	struct ipv6_saddr_dst dst;
1314
	struct net_device *dev;
1315
	int dst_type;
L
Linus Torvalds 已提交
1316

1317 1318 1319 1320
	dst_type = __ipv6_addr_type(daddr);
	dst.addr = daddr;
	dst.ifindex = dst_dev ? dst_dev->ifindex : 0;
	dst.scope = __ipv6_addr_src_scope(dst_type);
1321
	dst.label = ipv6_addr_label(net, daddr, dst_type, dst.ifindex);
1322
	dst.prefs = prefs;
1323 1324 1325

	hiscore->rule = -1;
	hiscore->ifa = NULL;
L
Linus Torvalds 已提交
1326

1327
	rcu_read_lock();
L
Linus Torvalds 已提交
1328

1329
	for_each_netdev_rcu(net, dev) {
1330 1331
		struct inet6_dev *idev;

1332
		/* Candidate Source Address (section 4)
1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
		 *  - multicast and link-local destination address,
		 *    the set of candidate source address MUST only
		 *    include addresses assigned to interfaces
		 *    belonging to the same link as the outgoing
		 *    interface.
		 * (- For site-local destination addresses, the
		 *    set of candidate source addresses MUST only
		 *    include addresses assigned to interfaces
		 *    belonging to the same site as the outgoing
		 *    interface.)
		 */
1344 1345 1346
		if (((dst_type & IPV6_ADDR_MULTICAST) ||
		     dst.scope <= IPV6_ADDR_SCOPE_LINKLOCAL) &&
		    dst.ifindex && dev->ifindex != dst.ifindex)
1347
			continue;
L
Linus Torvalds 已提交
1348 1349

		idev = __in6_dev_get(dev);
1350 1351 1352 1353
		if (!idev)
			continue;

		read_lock_bh(&idev->lock);
1354
		list_for_each_entry(score->ifa, &idev->addr_list, if_list) {
1355
			int i;
1356

1357
			/*
1358 1359 1360
			 * - Tentative Address (RFC2462 section 5.4)
			 *  - A tentative address is not considered
			 *    "assigned to an interface" in the traditional
1361
			 *    sense, unless it is also flagged as optimistic.
1362
			 * - Candidate Source Address (section 4)
1363 1364 1365 1366
			 *  - In any case, anycast addresses, multicast
			 *    addresses, and the unspecified address MUST
			 *    NOT be included in a candidate set.
			 */
1367 1368
			if ((score->ifa->flags & IFA_F_TENTATIVE) &&
			    (!(score->ifa->flags & IFA_F_OPTIMISTIC)))
1369
				continue;
1370 1371 1372 1373 1374

			score->addr_type = __ipv6_addr_type(&score->ifa->addr);

			if (unlikely(score->addr_type == IPV6_ADDR_ANY ||
				     score->addr_type & IPV6_ADDR_MULTICAST)) {
1375
				LIMIT_NETDEBUG(KERN_DEBUG
J
Joe Perches 已提交
1376
					       "ADDRCONF: unspecified / multicast address "
1377 1378 1379 1380 1381
					       "assigned as unicast address on %s",
					       dev->name);
				continue;
			}

1382 1383 1384 1385 1386 1387
			score->rule = -1;
			bitmap_zero(score->scorebits, IPV6_SADDR_RULE_MAX);

			for (i = 0; i < IPV6_SADDR_RULE_MAX; i++) {
				int minihiscore, miniscore;

1388 1389
				minihiscore = ipv6_get_saddr_eval(net, hiscore, &dst, i);
				miniscore = ipv6_get_saddr_eval(net, score, &dst, i);
1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410

				if (minihiscore > miniscore) {
					if (i == IPV6_SADDR_RULE_SCOPE &&
					    score->scopedist > 0) {
						/*
						 * special case:
						 * each remaining entry
						 * has too small (not enough)
						 * scope, because ifa entries
						 * are sorted by their scope
						 * values.
						 */
						goto try_nextdev;
					}
					break;
				} else if (minihiscore < miniscore) {
					if (hiscore->ifa)
						in6_ifa_put(hiscore->ifa);

					in6_ifa_hold(score->ifa);

1411
					swap(hiscore, score);
1412 1413 1414 1415 1416

					/* restore our iterator */
					score->ifa = hiscore->ifa;

					break;
L
Linus Torvalds 已提交
1417
				}
1418
			}
L
Linus Torvalds 已提交
1419
		}
1420
try_nextdev:
1421
		read_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
1422
	}
1423
	rcu_read_unlock();
L
Linus Torvalds 已提交
1424

1425
	if (!hiscore->ifa)
1426
		return -EADDRNOTAVAIL;
1427

A
Alexey Dobriyan 已提交
1428
	*saddr = hiscore->ifa->addr;
1429
	in6_ifa_put(hiscore->ifa);
1430
	return 0;
L
Linus Torvalds 已提交
1431
}
1432
EXPORT_SYMBOL(ipv6_dev_get_saddr);
L
Linus Torvalds 已提交
1433

1434
int __ipv6_get_lladdr(struct inet6_dev *idev, struct in6_addr *addr,
1435
		      u32 banned_flags)
1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450
{
	struct inet6_ifaddr *ifp;
	int err = -EADDRNOTAVAIL;

	list_for_each_entry(ifp, &idev->addr_list, if_list) {
		if (ifp->scope == IFA_LINK &&
		    !(ifp->flags & banned_flags)) {
			*addr = ifp->addr;
			err = 0;
			break;
		}
	}
	return err;
}

1451
int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr,
1452
		    u32 banned_flags)
L
Linus Torvalds 已提交
1453 1454 1455 1456
{
	struct inet6_dev *idev;
	int err = -EADDRNOTAVAIL;

1457
	rcu_read_lock();
S
Stephen Hemminger 已提交
1458 1459
	idev = __in6_dev_get(dev);
	if (idev) {
L
Linus Torvalds 已提交
1460
		read_lock_bh(&idev->lock);
1461
		err = __ipv6_get_lladdr(idev, addr, banned_flags);
L
Linus Torvalds 已提交
1462 1463
		read_unlock_bh(&idev->lock);
	}
1464
	rcu_read_unlock();
L
Linus Torvalds 已提交
1465 1466 1467 1468 1469 1470 1471 1472 1473
	return err;
}

static int ipv6_count_addresses(struct inet6_dev *idev)
{
	int cnt = 0;
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
1474
	list_for_each_entry(ifp, &idev->addr_list, if_list)
L
Linus Torvalds 已提交
1475 1476 1477 1478 1479
		cnt++;
	read_unlock_bh(&idev->lock);
	return cnt;
}

1480
int ipv6_chk_addr(struct net *net, const struct in6_addr *addr,
1481
		  const struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1482
{
1483
	struct inet6_ifaddr *ifp;
E
Eric Dumazet 已提交
1484
	unsigned int hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
1485

1486
	rcu_read_lock_bh();
1487
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
1488
		if (!net_eq(dev_net(ifp->idev->dev), net))
1489
			continue;
L
Linus Torvalds 已提交
1490
		if (ipv6_addr_equal(&ifp->addr, addr) &&
1491 1492 1493 1494 1495
		    !(ifp->flags&IFA_F_TENTATIVE) &&
		    (dev == NULL || ifp->idev->dev == dev ||
		     !(ifp->scope&(IFA_LINK|IFA_HOST) || strict))) {
			rcu_read_unlock_bh();
			return 1;
L
Linus Torvalds 已提交
1496 1497
		}
	}
1498

1499 1500
	rcu_read_unlock_bh();
	return 0;
L
Linus Torvalds 已提交
1501
}
1502 1503
EXPORT_SYMBOL(ipv6_chk_addr);

1504 1505
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev)
L
Linus Torvalds 已提交
1506
{
E
Eric Dumazet 已提交
1507
	unsigned int hash = inet6_addr_hash(addr);
1508
	struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
1509

1510
	hlist_for_each_entry(ifp, &inet6_addr_lst[hash], addr_lst) {
1511
		if (!net_eq(dev_net(ifp->idev->dev), net))
1512
			continue;
L
Linus Torvalds 已提交
1513 1514
		if (ipv6_addr_equal(&ifp->addr, addr)) {
			if (dev == NULL || ifp->idev->dev == dev)
1515
				return true;
L
Linus Torvalds 已提交
1516 1517
		}
	}
1518
	return false;
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 1542 1543 1544 1545 1546 1547
/* Compares an address/prefix_len with addresses on device @dev.
 * If one is found it returns true.
 */
bool ipv6_chk_custom_prefix(const struct in6_addr *addr,
	const unsigned int prefix_len, struct net_device *dev)
{
	struct inet6_dev *idev;
	struct inet6_ifaddr *ifa;
	bool ret = false;

	rcu_read_lock();
	idev = __in6_dev_get(dev);
	if (idev) {
		read_lock_bh(&idev->lock);
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
			ret = ipv6_prefix_equal(addr, &ifa->addr, prefix_len);
			if (ret)
				break;
		}
		read_unlock_bh(&idev->lock);
	}
	rcu_read_unlock();

	return ret;
}
EXPORT_SYMBOL(ipv6_chk_custom_prefix);

1548
int ipv6_chk_prefix(const struct in6_addr *addr, struct net_device *dev)
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
{
	struct inet6_dev *idev;
	struct inet6_ifaddr *ifa;
	int	onlink;

	onlink = 0;
	rcu_read_lock();
	idev = __in6_dev_get(dev);
	if (idev) {
		read_lock_bh(&idev->lock);
1559
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571
			onlink = ipv6_prefix_equal(addr, &ifa->addr,
						   ifa->prefix_len);
			if (onlink)
				break;
		}
		read_unlock_bh(&idev->lock);
	}
	rcu_read_unlock();
	return onlink;
}
EXPORT_SYMBOL(ipv6_chk_prefix);

1572
struct inet6_ifaddr *ipv6_get_ifaddr(struct net *net, const struct in6_addr *addr,
1573
				     struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1574
{
1575
	struct inet6_ifaddr *ifp, *result = NULL;
E
Eric Dumazet 已提交
1576
	unsigned int hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
1577

1578
	rcu_read_lock_bh();
1579
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
1580
		if (!net_eq(dev_net(ifp->idev->dev), net))
1581
			continue;
L
Linus Torvalds 已提交
1582 1583 1584
		if (ipv6_addr_equal(&ifp->addr, addr)) {
			if (dev == NULL || ifp->idev->dev == dev ||
			    !(ifp->scope&(IFA_LINK|IFA_HOST) || strict)) {
1585
				result = ifp;
L
Linus Torvalds 已提交
1586 1587 1588 1589 1590
				in6_ifa_hold(ifp);
				break;
			}
		}
	}
1591
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1592

1593
	return result;
L
Linus Torvalds 已提交
1594 1595 1596 1597
}

/* Gets referenced address, destroys ifaddr */

B
Brian Haley 已提交
1598
static void addrconf_dad_stop(struct inet6_ifaddr *ifp, int dad_failed)
L
Linus Torvalds 已提交
1599 1600 1601
{
	if (ifp->flags&IFA_F_PERMANENT) {
		spin_lock_bh(&ifp->lock);
1602
		addrconf_del_dad_timer(ifp);
L
Linus Torvalds 已提交
1603
		ifp->flags |= IFA_F_TENTATIVE;
B
Brian Haley 已提交
1604 1605
		if (dad_failed)
			ifp->flags |= IFA_F_DADFAILED;
L
Linus Torvalds 已提交
1606
		spin_unlock_bh(&ifp->lock);
1607 1608
		if (dad_failed)
			ipv6_ifa_notify(0, ifp);
L
Linus Torvalds 已提交
1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626
		in6_ifa_put(ifp);
	} else if (ifp->flags&IFA_F_TEMPORARY) {
		struct inet6_ifaddr *ifpub;
		spin_lock_bh(&ifp->lock);
		ifpub = ifp->ifpub;
		if (ifpub) {
			in6_ifa_hold(ifpub);
			spin_unlock_bh(&ifp->lock);
			ipv6_create_tempaddr(ifpub, ifp);
			in6_ifa_put(ifpub);
		} else {
			spin_unlock_bh(&ifp->lock);
		}
		ipv6_del_addr(ifp);
	} else
		ipv6_del_addr(ifp);
}

H
Herbert Xu 已提交
1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640
static int addrconf_dad_end(struct inet6_ifaddr *ifp)
{
	int err = -ENOENT;

	spin_lock(&ifp->state_lock);
	if (ifp->state == INET6_IFADDR_STATE_DAD) {
		ifp->state = INET6_IFADDR_STATE_POSTDAD;
		err = 0;
	}
	spin_unlock(&ifp->state_lock);

	return err;
}

1641 1642
void addrconf_dad_failure(struct inet6_ifaddr *ifp)
{
1643
	struct inet6_dev *idev = ifp->idev;
1644

1645 1646
	if (addrconf_dad_end(ifp)) {
		in6_ifa_put(ifp);
H
Herbert Xu 已提交
1647
		return;
1648
	}
H
Herbert Xu 已提交
1649

1650 1651
	net_info_ratelimited("%s: IPv6 duplicate address %pI6c detected!\n",
			     ifp->idev->dev->name, &ifp->addr);
1652

1653 1654 1655 1656 1657 1658 1659 1660 1661 1662
	if (idev->cnf.accept_dad > 1 && !idev->cnf.disable_ipv6) {
		struct in6_addr addr;

		addr.s6_addr32[0] = htonl(0xfe800000);
		addr.s6_addr32[1] = 0;

		if (!ipv6_generate_eui64(addr.s6_addr + 8, idev->dev) &&
		    ipv6_addr_equal(&ifp->addr, &addr)) {
			/* DAD failed for link-local based on MAC address */
			idev->cnf.disable_ipv6 = 1;
1663

1664
			pr_info("%s: IPv6 being disabled!\n",
1665
				ifp->idev->dev->name);
1666 1667 1668
		}
	}

B
Brian Haley 已提交
1669
	addrconf_dad_stop(ifp, 1);
1670
}
L
Linus Torvalds 已提交
1671 1672 1673

/* Join to solicited addr multicast group. */

1674
void addrconf_join_solict(struct net_device *dev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
1675 1676 1677 1678 1679 1680 1681 1682 1683 1684
{
	struct in6_addr maddr;

	if (dev->flags&(IFF_LOOPBACK|IFF_NOARP))
		return;

	addrconf_addr_solict_mult(addr, &maddr);
	ipv6_dev_mc_inc(dev, &maddr);
}

1685
void addrconf_leave_solict(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
1686 1687 1688 1689 1690 1691 1692 1693 1694 1695
{
	struct in6_addr maddr;

	if (idev->dev->flags&(IFF_LOOPBACK|IFF_NOARP))
		return;

	addrconf_addr_solict_mult(addr, &maddr);
	__ipv6_dev_mc_dec(idev, &maddr);
}

1696
static void addrconf_join_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
1697 1698
{
	struct in6_addr addr;
1699
	if (ifp->prefix_len >= 127) /* RFC 6164 */
1700
		return;
L
Linus Torvalds 已提交
1701 1702 1703 1704 1705 1706
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
	ipv6_dev_ac_inc(ifp->idev->dev, &addr);
}

1707
static void addrconf_leave_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
1708 1709
{
	struct in6_addr addr;
1710
	if (ifp->prefix_len >= 127) /* RFC 6164 */
1711
		return;
L
Linus Torvalds 已提交
1712 1713 1714 1715 1716 1717
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
	__ipv6_dev_ac_dec(ifp->idev, &addr);
}

1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748
static int addrconf_ifid_eui48(u8 *eui, struct net_device *dev)
{
	if (dev->addr_len != ETH_ALEN)
		return -1;
	memcpy(eui, dev->dev_addr, 3);
	memcpy(eui + 5, dev->dev_addr + 3, 3);

	/*
	 * The zSeries OSA network cards can be shared among various
	 * OS instances, but the OSA cards have only one MAC address.
	 * This leads to duplicate address conflicts in conjunction
	 * with IPv6 if more than one instance uses the same card.
	 *
	 * The driver for these cards can deliver a unique 16-bit
	 * identifier for each instance sharing the same card.  It is
	 * placed instead of 0xFFFE in the interface identifier.  The
	 * "u" bit of the interface identifier is not inverted in this
	 * case.  Hence the resulting interface identifier has local
	 * scope according to RFC2373.
	 */
	if (dev->dev_id) {
		eui[3] = (dev->dev_id >> 8) & 0xFF;
		eui[4] = dev->dev_id & 0xFF;
	} else {
		eui[3] = 0xFF;
		eui[4] = 0xFE;
		eui[0] ^= 2;
	}
	return 0;
}

1749 1750 1751 1752 1753
static int addrconf_ifid_eui64(u8 *eui, struct net_device *dev)
{
	if (dev->addr_len != IEEE802154_ADDR_LEN)
		return -1;
	memcpy(eui, dev->dev_addr, 8);
1754
	eui[0] ^= 2;
1755 1756 1757
	return 0;
}

1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771
static int addrconf_ifid_ieee1394(u8 *eui, struct net_device *dev)
{
	union fwnet_hwaddr *ha;

	if (dev->addr_len != FWNET_ALEN)
		return -1;

	ha = (union fwnet_hwaddr *)dev->dev_addr;

	memcpy(eui, &ha->uc.uniq_id, sizeof(ha->uc.uniq_id));
	eui[0] ^= 2;
	return 0;
}

1772 1773 1774 1775 1776 1777
static int addrconf_ifid_arcnet(u8 *eui, struct net_device *dev)
{
	/* XXX: inherit EUI-64 from other interface -- yoshfuji */
	if (dev->addr_len != ARCNET_ALEN)
		return -1;
	memset(eui, 0, 7);
1778
	eui[7] = *(u8 *)dev->dev_addr;
1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790
	return 0;
}

static int addrconf_ifid_infiniband(u8 *eui, struct net_device *dev)
{
	if (dev->addr_len != INFINIBAND_ALEN)
		return -1;
	memcpy(eui, dev->dev_addr + 12, 8);
	eui[0] |= 2;
	return 0;
}

1791
static int __ipv6_isatap_ifid(u8 *eui, __be32 addr)
1792
{
1793 1794
	if (addr == 0)
		return -1;
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814
	eui[0] = (ipv4_is_zeronet(addr) || ipv4_is_private_10(addr) ||
		  ipv4_is_loopback(addr) || ipv4_is_linklocal_169(addr) ||
		  ipv4_is_private_172(addr) || ipv4_is_test_192(addr) ||
		  ipv4_is_anycast_6to4(addr) || ipv4_is_private_192(addr) ||
		  ipv4_is_test_198(addr) || ipv4_is_multicast(addr) ||
		  ipv4_is_lbcast(addr)) ? 0x00 : 0x02;
	eui[1] = 0;
	eui[2] = 0x5E;
	eui[3] = 0xFE;
	memcpy(eui + 4, &addr, 4);
	return 0;
}

static int addrconf_ifid_sit(u8 *eui, struct net_device *dev)
{
	if (dev->priv_flags & IFF_ISATAP)
		return __ipv6_isatap_ifid(eui, *(__be32 *)dev->dev_addr);
	return -1;
}

1815 1816 1817 1818 1819
static int addrconf_ifid_gre(u8 *eui, struct net_device *dev)
{
	return __ipv6_isatap_ifid(eui, *(__be32 *)dev->dev_addr);
}

1820 1821 1822 1823 1824 1825 1826 1827 1828 1829
static int addrconf_ifid_ip6tnl(u8 *eui, struct net_device *dev)
{
	memcpy(eui, dev->perm_addr, 3);
	memcpy(eui + 5, dev->perm_addr + 3, 3);
	eui[3] = 0xFF;
	eui[4] = 0xFE;
	eui[0] ^= 2;
	return 0;
}

L
Linus Torvalds 已提交
1830 1831 1832 1833 1834
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev)
{
	switch (dev->type) {
	case ARPHRD_ETHER:
	case ARPHRD_FDDI:
1835
		return addrconf_ifid_eui48(eui, dev);
L
Linus Torvalds 已提交
1836
	case ARPHRD_ARCNET:
1837
		return addrconf_ifid_arcnet(eui, dev);
L
Linus Torvalds 已提交
1838
	case ARPHRD_INFINIBAND:
1839
		return addrconf_ifid_infiniband(eui, dev);
F
Fred L. Templin 已提交
1840
	case ARPHRD_SIT:
1841
		return addrconf_ifid_sit(eui, dev);
1842 1843
	case ARPHRD_IPGRE:
		return addrconf_ifid_gre(eui, dev);
1844
	case ARPHRD_6LOWPAN:
1845 1846
	case ARPHRD_IEEE802154:
		return addrconf_ifid_eui64(eui, dev);
1847 1848
	case ARPHRD_IEEE1394:
		return addrconf_ifid_ieee1394(eui, dev);
1849 1850
	case ARPHRD_TUNNEL6:
		return addrconf_ifid_ip6tnl(eui, dev);
L
Linus Torvalds 已提交
1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
	}
	return -1;
}

static int ipv6_inherit_eui64(u8 *eui, struct inet6_dev *idev)
{
	int err = -1;
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
1861
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
L
Linus Torvalds 已提交
1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872
		if (ifp->scope == IFA_LINK && !(ifp->flags&IFA_F_TENTATIVE)) {
			memcpy(eui, ifp->addr.s6_addr+8, 8);
			err = 0;
			break;
		}
	}
	read_unlock_bh(&idev->lock);
	return err;
}

/* (re)generation of randomized interface identifier (RFC 3041 3.2, 3.5) */
S
Sorin Dumitru 已提交
1873
static void __ipv6_regen_rndid(struct inet6_dev *idev)
L
Linus Torvalds 已提交
1874 1875
{
regen:
1876
	get_random_bytes(idev->rndid, sizeof(idev->rndid));
L
Linus Torvalds 已提交
1877 1878 1879 1880 1881 1882 1883 1884
	idev->rndid[0] &= ~0x02;

	/*
	 * <draft-ietf-ipngwg-temp-addresses-v2-00.txt>:
	 * check if generated address is not inappropriate
	 *
	 *  - Reserved subnet anycast (RFC 2526)
	 *	11111101 11....11 1xxxxxxx
F
Fred L. Templin 已提交
1885
	 *  - ISATAP (RFC4214) 6.1
L
Linus Torvalds 已提交
1886 1887 1888 1889
	 *	00-00-5E-FE-xx-xx-xx-xx
	 *  - value 0
	 *  - XXX: already assigned to an address on the device
	 */
1890
	if (idev->rndid[0] == 0xfd &&
L
Linus Torvalds 已提交
1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906
	    (idev->rndid[1]&idev->rndid[2]&idev->rndid[3]&idev->rndid[4]&idev->rndid[5]&idev->rndid[6]) == 0xff &&
	    (idev->rndid[7]&0x80))
		goto regen;
	if ((idev->rndid[0]|idev->rndid[1]) == 0) {
		if (idev->rndid[2] == 0x5e && idev->rndid[3] == 0xfe)
			goto regen;
		if ((idev->rndid[2]|idev->rndid[3]|idev->rndid[4]|idev->rndid[5]|idev->rndid[6]|idev->rndid[7]) == 0x00)
			goto regen;
	}
}

static void ipv6_regen_rndid(unsigned long data)
{
	struct inet6_dev *idev = (struct inet6_dev *) data;
	unsigned long expires;

1907
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
1908 1909 1910 1911 1912
	write_lock_bh(&idev->lock);

	if (idev->dead)
		goto out;

S
Sorin Dumitru 已提交
1913
	__ipv6_regen_rndid(idev);
1914

L
Linus Torvalds 已提交
1915
	expires = jiffies +
1916
		idev->cnf.temp_prefered_lft * HZ -
J
Jiri Pirko 已提交
1917 1918
		idev->cnf.regen_max_retry * idev->cnf.dad_transmits *
		NEIGH_VAR(idev->nd_parms, RETRANS_TIME) -
1919
		idev->cnf.max_desync_factor * HZ;
L
Linus Torvalds 已提交
1920
	if (time_before(expires, jiffies)) {
1921 1922
		pr_warn("%s: too short regeneration interval; timer disabled for %s\n",
			__func__, idev->dev->name);
L
Linus Torvalds 已提交
1923 1924 1925 1926 1927 1928 1929 1930
		goto out;
	}

	if (!mod_timer(&idev->regen_timer, expires))
		in6_dev_hold(idev);

out:
	write_unlock_bh(&idev->lock);
1931
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1932 1933 1934
	in6_dev_put(idev);
}

S
Sorin Dumitru 已提交
1935
static void  __ipv6_try_regen_rndid(struct inet6_dev *idev, struct in6_addr *tmpaddr)
1936
{
L
Linus Torvalds 已提交
1937
	if (tmpaddr && memcmp(idev->rndid, &tmpaddr->s6_addr[8], 8) == 0)
S
Sorin Dumitru 已提交
1938
		__ipv6_regen_rndid(idev);
L
Linus Torvalds 已提交
1939 1940 1941 1942 1943 1944 1945 1946
}

/*
 *	Add prefix route.
 */

static void
addrconf_prefix_route(struct in6_addr *pfx, int plen, struct net_device *dev,
J
Jamal Hadi Salim 已提交
1947
		      unsigned long expires, u32 flags)
L
Linus Torvalds 已提交
1948
{
1949 1950 1951 1952 1953 1954 1955
	struct fib6_config cfg = {
		.fc_table = RT6_TABLE_PREFIX,
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_expires = expires,
		.fc_dst_len = plen,
		.fc_flags = RTF_UP | flags,
1956
		.fc_nlinfo.nl_net = dev_net(dev),
1957
		.fc_protocol = RTPROT_KERNEL,
1958
	};
L
Linus Torvalds 已提交
1959

A
Alexey Dobriyan 已提交
1960
	cfg.fc_dst = *pfx;
L
Linus Torvalds 已提交
1961 1962 1963 1964 1965

	/* Prevent useless cloning on PtP SIT.
	   This thing is done here expecting that the whole
	   class of non-broadcast devices need not cloning.
	 */
A
Amerigo Wang 已提交
1966
#if IS_ENABLED(CONFIG_IPV6_SIT)
1967 1968
	if (dev->type == ARPHRD_SIT && (dev->flags & IFF_POINTOPOINT))
		cfg.fc_flags |= RTF_NONEXTHOP;
1969
#endif
L
Linus Torvalds 已提交
1970

1971
	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
1972 1973
}

1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987

static struct rt6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags)
{
	struct fib6_node *fn;
	struct rt6_info *rt = NULL;
	struct fib6_table *table;

	table = fib6_get_table(dev_net(dev), RT6_TABLE_PREFIX);
	if (table == NULL)
		return NULL;

1988
	read_lock_bh(&table->tb6_lock);
1989 1990 1991 1992
	fn = fib6_locate(&table->tb6_root, pfx, plen, NULL, 0);
	if (!fn)
		goto out;
	for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) {
1993
		if (rt->dst.dev->ifindex != dev->ifindex)
1994 1995 1996
			continue;
		if ((rt->rt6i_flags & flags) != flags)
			continue;
1997
		if ((rt->rt6i_flags & noflags) != 0)
1998 1999 2000 2001 2002
			continue;
		dst_hold(&rt->dst);
		break;
	}
out:
2003
	read_unlock_bh(&table->tb6_lock);
2004 2005 2006 2007
	return rt;
}


L
Linus Torvalds 已提交
2008 2009 2010 2011
/* Create "default" multicast route to the interface */

static void addrconf_add_mroute(struct net_device *dev)
{
2012 2013 2014 2015 2016 2017
	struct fib6_config cfg = {
		.fc_table = RT6_TABLE_LOCAL,
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_dst_len = 8,
		.fc_flags = RTF_UP,
2018
		.fc_nlinfo.nl_net = dev_net(dev),
2019 2020 2021 2022 2023
	};

	ipv6_addr_set(&cfg.fc_dst, htonl(0xFF000000), 0, 0, 0);

	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2024 2025 2026 2027 2028 2029 2030 2031
}

static struct inet6_dev *addrconf_add_dev(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
2032 2033
	idev = ipv6_find_idev(dev);
	if (!idev)
2034 2035 2036 2037
		return ERR_PTR(-ENOBUFS);

	if (idev->cnf.disable_ipv6)
		return ERR_PTR(-EACCES);
L
Linus Torvalds 已提交
2038 2039

	/* Add default multicast route */
2040 2041
	if (!(dev->flags & IFF_LOOPBACK))
		addrconf_add_mroute(dev);
L
Linus Torvalds 已提交
2042 2043 2044 2045

	return idev;
}

2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112
static void manage_tempaddrs(struct inet6_dev *idev,
			     struct inet6_ifaddr *ifp,
			     __u32 valid_lft, __u32 prefered_lft,
			     bool create, unsigned long now)
{
	u32 flags;
	struct inet6_ifaddr *ift;

	read_lock_bh(&idev->lock);
	/* update all temporary addresses in the list */
	list_for_each_entry(ift, &idev->tempaddr_list, tmp_list) {
		int age, max_valid, max_prefered;

		if (ifp != ift->ifpub)
			continue;

		/* RFC 4941 section 3.3:
		 * If a received option will extend the lifetime of a public
		 * address, the lifetimes of temporary addresses should
		 * be extended, subject to the overall constraint that no
		 * temporary addresses should ever remain "valid" or "preferred"
		 * for a time longer than (TEMP_VALID_LIFETIME) or
		 * (TEMP_PREFERRED_LIFETIME - DESYNC_FACTOR), respectively.
		 */
		age = (now - ift->cstamp) / HZ;
		max_valid = idev->cnf.temp_valid_lft - age;
		if (max_valid < 0)
			max_valid = 0;

		max_prefered = idev->cnf.temp_prefered_lft -
			       idev->cnf.max_desync_factor - age;
		if (max_prefered < 0)
			max_prefered = 0;

		if (valid_lft > max_valid)
			valid_lft = max_valid;

		if (prefered_lft > max_prefered)
			prefered_lft = max_prefered;

		spin_lock(&ift->lock);
		flags = ift->flags;
		ift->valid_lft = valid_lft;
		ift->prefered_lft = prefered_lft;
		ift->tstamp = now;
		if (prefered_lft > 0)
			ift->flags &= ~IFA_F_DEPRECATED;

		spin_unlock(&ift->lock);
		if (!(flags&IFA_F_TENTATIVE))
			ipv6_ifa_notify(0, ift);
	}

	if ((create || list_empty(&idev->tempaddr_list)) &&
	    idev->cnf.use_tempaddr > 0) {
		/* When a new public address is created as described
		 * in [ADDRCONF], also create a new temporary address.
		 * Also create a temporary address if it's enabled but
		 * no temporary address currently exists.
		 */
		read_unlock_bh(&idev->lock);
		ipv6_create_tempaddr(ifp, NULL);
	} else {
		read_unlock_bh(&idev->lock);
	}
}

2113
void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
L
Linus Torvalds 已提交
2114 2115 2116 2117 2118 2119
{
	struct prefix_info *pinfo;
	__u32 valid_lft;
	__u32 prefered_lft;
	int addr_type;
	struct inet6_dev *in6_dev;
2120
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
2121 2122

	pinfo = (struct prefix_info *) opt;
2123

L
Linus Torvalds 已提交
2124
	if (len < sizeof(struct prefix_info)) {
2125
		ADBG("addrconf: prefix option too short\n");
L
Linus Torvalds 已提交
2126 2127
		return;
	}
2128

L
Linus Torvalds 已提交
2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141
	/*
	 *	Validation checks ([ADDRCONF], page 19)
	 */

	addr_type = ipv6_addr_type(&pinfo->prefix);

	if (addr_type & (IPV6_ADDR_MULTICAST|IPV6_ADDR_LINKLOCAL))
		return;

	valid_lft = ntohl(pinfo->valid);
	prefered_lft = ntohl(pinfo->prefered);

	if (prefered_lft > valid_lft) {
2142
		net_warn_ratelimited("addrconf: prefix option has invalid lifetime\n");
L
Linus Torvalds 已提交
2143 2144 2145 2146 2147 2148
		return;
	}

	in6_dev = in6_dev_get(dev);

	if (in6_dev == NULL) {
2149 2150
		net_dbg_ratelimited("addrconf: device %s not configured\n",
				    dev->name);
L
Linus Torvalds 已提交
2151 2152 2153 2154 2155 2156 2157 2158 2159
		return;
	}

	/*
	 *	Two things going on here:
	 *	1) Add routes for on-link prefixes
	 *	2) Configure prefixes with the auto flag set
	 */

2160 2161 2162 2163
	if (pinfo->onlink) {
		struct rt6_info *rt;
		unsigned long rt_expires;

2164 2165 2166 2167 2168
		/* Avoid arithmetic overflow. Really, we could
		 * save rt_expires in seconds, likely valid_lft,
		 * but it would require division in fib gc, that it
		 * not good.
		 */
2169 2170 2171 2172
		if (HZ > USER_HZ)
			rt_expires = addrconf_timeout_fixup(valid_lft, HZ);
		else
			rt_expires = addrconf_timeout_fixup(valid_lft, USER_HZ);
Y
YOSHIFUJI Hideaki 已提交
2173

2174 2175
		if (addrconf_finite_timeout(rt_expires))
			rt_expires *= HZ;
L
Linus Torvalds 已提交
2176

2177 2178 2179 2180 2181
		rt = addrconf_get_prefix_route(&pinfo->prefix,
					       pinfo->prefix_len,
					       dev,
					       RTF_ADDRCONF | RTF_PREFIX_RT,
					       RTF_GATEWAY | RTF_DEFAULT);
L
Linus Torvalds 已提交
2182

2183
		if (rt) {
2184 2185 2186 2187
			/* Autoconf prefix route */
			if (valid_lft == 0) {
				ip6_del_rt(rt);
				rt = NULL;
2188
			} else if (addrconf_finite_timeout(rt_expires)) {
2189
				/* not infinity */
2190
				rt6_set_expires(rt, jiffies + rt_expires);
2191
			} else {
2192
				rt6_clean_expires(rt);
L
Linus Torvalds 已提交
2193 2194
			}
		} else if (valid_lft) {
2195
			clock_t expires = 0;
2196 2197
			int flags = RTF_ADDRCONF | RTF_PREFIX_RT;
			if (addrconf_finite_timeout(rt_expires)) {
2198 2199 2200 2201
				/* not infinity */
				flags |= RTF_EXPIRES;
				expires = jiffies_to_clock_t(rt_expires);
			}
L
Linus Torvalds 已提交
2202
			addrconf_prefix_route(&pinfo->prefix, pinfo->prefix_len,
2203
					      dev, expires, flags);
L
Linus Torvalds 已提交
2204
		}
A
Amerigo Wang 已提交
2205
		ip6_rt_put(rt);
L
Linus Torvalds 已提交
2206 2207 2208 2209 2210
	}

	/* Try to figure out our local address for this prefix */

	if (pinfo->autoconf && in6_dev->cnf.autoconf) {
2211
		struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
2212 2213
		struct in6_addr addr;
		int create = 0, update_lft = 0;
2214
		bool tokenized = false;
L
Linus Torvalds 已提交
2215 2216 2217

		if (pinfo->prefix_len == 64) {
			memcpy(&addr, &pinfo->prefix, 8);
2218 2219 2220 2221 2222 2223

			if (!ipv6_addr_any(&in6_dev->token)) {
				read_lock_bh(&in6_dev->lock);
				memcpy(addr.s6_addr + 8,
				       in6_dev->token.s6_addr + 8, 8);
				read_unlock_bh(&in6_dev->lock);
2224
				tokenized = true;
2225 2226
			} else if (ipv6_generate_eui64(addr.s6_addr + 8, dev) &&
				   ipv6_inherit_eui64(addr.s6_addr + 8, in6_dev)) {
L
Linus Torvalds 已提交
2227 2228 2229 2230 2231
				in6_dev_put(in6_dev);
				return;
			}
			goto ok;
		}
2232 2233
		net_dbg_ratelimited("IPv6 addrconf: prefix with wrong length %d\n",
				    pinfo->prefix_len);
L
Linus Torvalds 已提交
2234 2235 2236 2237 2238
		in6_dev_put(in6_dev);
		return;

ok:

2239
		ifp = ipv6_get_ifaddr(net, &addr, dev, 1);
L
Linus Torvalds 已提交
2240 2241 2242

		if (ifp == NULL && valid_lft) {
			int max_addresses = in6_dev->cnf.max_addresses;
2243 2244 2245 2246
			u32 addr_flags = 0;

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
			if (in6_dev->cnf.optimistic_dad &&
2247
			    !net->ipv6.devconf_all->forwarding && sllao)
2248 2249
				addr_flags = IFA_F_OPTIMISTIC;
#endif
L
Linus Torvalds 已提交
2250 2251 2252 2253 2254 2255

			/* Do not allow to create too much of autoconfigured
			 * addresses; this would be too easy way to crash kernel.
			 */
			if (!max_addresses ||
			    ipv6_count_addresses(in6_dev) < max_addresses)
2256 2257
				ifp = ipv6_add_addr(in6_dev, &addr, NULL,
						    pinfo->prefix_len,
2258
						    addr_type&IPV6_ADDR_SCOPE_MASK,
2259 2260
						    addr_flags, valid_lft,
						    prefered_lft);
L
Linus Torvalds 已提交
2261

2262
			if (IS_ERR_OR_NULL(ifp)) {
L
Linus Torvalds 已提交
2263 2264 2265 2266
				in6_dev_put(in6_dev);
				return;
			}

2267
			ifp->flags |= IFA_F_MANAGETEMPADDR;
2268 2269
			update_lft = 0;
			create = 1;
L
Linus Torvalds 已提交
2270
			ifp->cstamp = jiffies;
2271
			ifp->tokenized = tokenized;
2272
			addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
2273 2274 2275
		}

		if (ifp) {
2276
			u32 flags;
L
Linus Torvalds 已提交
2277 2278 2279 2280 2281 2282 2283 2284 2285 2286
			unsigned long now;
			u32 stored_lft;

			/* update lifetime (RFC2462 5.5.3 e) */
			spin_lock(&ifp->lock);
			now = jiffies;
			if (ifp->valid_lft > (now - ifp->tstamp) / HZ)
				stored_lft = ifp->valid_lft - (now - ifp->tstamp) / HZ;
			else
				stored_lft = 0;
2287
			if (!update_lft && !create && stored_lft) {
2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302
				const u32 minimum_lft = min(
					stored_lft, (u32)MIN_VALID_LIFETIME);
				valid_lft = max(valid_lft, minimum_lft);

				/* RFC4862 Section 5.5.3e:
				 * "Note that the preferred lifetime of the
				 *  corresponding address is always reset to
				 *  the Preferred Lifetime in the received
				 *  Prefix Information option, regardless of
				 *  whether the valid lifetime is also reset or
				 *  ignored."
				 *
				 * So we should always update prefered_lft here.
				 */
				update_lft = 1;
L
Linus Torvalds 已提交
2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317
			}

			if (update_lft) {
				ifp->valid_lft = valid_lft;
				ifp->prefered_lft = prefered_lft;
				ifp->tstamp = now;
				flags = ifp->flags;
				ifp->flags &= ~IFA_F_DEPRECATED;
				spin_unlock(&ifp->lock);

				if (!(flags&IFA_F_TENTATIVE))
					ipv6_ifa_notify(0, ifp);
			} else
				spin_unlock(&ifp->lock);

2318 2319
			manage_tempaddrs(in6_dev, ifp, valid_lft, prefered_lft,
					 create, now);
2320

L
Linus Torvalds 已提交
2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333
			in6_ifa_put(ifp);
			addrconf_verify(0);
		}
	}
	inet6_prefix_notify(RTM_NEWPREFIX, in6_dev, pinfo);
	in6_dev_put(in6_dev);
}

/*
 *	Set destination address.
 *	Special case for SIT interfaces where we create a new "virtual"
 *	device.
 */
2334
int addrconf_set_dstaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345
{
	struct in6_ifreq ireq;
	struct net_device *dev;
	int err = -EINVAL;

	rtnl_lock();

	err = -EFAULT;
	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		goto err_exit;

2346
	dev = __dev_get_by_index(net, ireq.ifr6_ifindex);
L
Linus Torvalds 已提交
2347 2348 2349 2350 2351

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

A
Amerigo Wang 已提交
2352
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2353
	if (dev->type == ARPHRD_SIT) {
2354
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368
		struct ifreq ifr;
		struct ip_tunnel_parm p;

		err = -EADDRNOTAVAIL;
		if (!(ipv6_addr_type(&ireq.ifr6_addr) & IPV6_ADDR_COMPATv4))
			goto err_exit;

		memset(&p, 0, sizeof(p));
		p.iph.daddr = ireq.ifr6_addr.s6_addr32[3];
		p.iph.saddr = 0;
		p.iph.version = 4;
		p.iph.ihl = 5;
		p.iph.protocol = IPPROTO_IPV6;
		p.iph.ttl = 64;
2369
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;
L
Linus Torvalds 已提交
2370

2371 2372 2373 2374 2375 2376 2377 2378
		if (ops->ndo_do_ioctl) {
			mm_segment_t oldfs = get_fs();

			set_fs(KERNEL_DS);
			err = ops->ndo_do_ioctl(dev, &ifr, SIOCADDTUNNEL);
			set_fs(oldfs);
		} else
			err = -EOPNOTSUPP;
L
Linus Torvalds 已提交
2379 2380 2381

		if (err == 0) {
			err = -ENOBUFS;
2382 2383
			dev = __dev_get_by_name(net, p.name);
			if (!dev)
L
Linus Torvalds 已提交
2384 2385 2386 2387
				goto err_exit;
			err = dev_open(dev);
		}
	}
2388
#endif
L
Linus Torvalds 已提交
2389 2390 2391 2392 2393 2394 2395 2396 2397

err_exit:
	rtnl_unlock();
	return err;
}

/*
 *	Manual configuration of address on an interface
 */
2398 2399
static int inet6_addr_add(struct net *net, int ifindex,
			  const struct in6_addr *pfx,
2400
			  const struct in6_addr *peer_pfx,
2401 2402
			  unsigned int plen, __u32 ifa_flags,
			  __u32 prefered_lft, __u32 valid_lft)
L
Linus Torvalds 已提交
2403 2404 2405 2406 2407
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
	int scope;
2408 2409
	u32 flags;
	clock_t expires;
2410
	unsigned long timeout;
L
Linus Torvalds 已提交
2411 2412

	ASSERT_RTNL();
2413

2414 2415 2416
	if (plen > 128)
		return -EINVAL;

2417 2418 2419 2420
	/* check the lifetime */
	if (!valid_lft || prefered_lft > valid_lft)
		return -EINVAL;

2421 2422 2423
	if (ifa_flags & IFA_F_MANAGETEMPADDR && plen != 64)
		return -EINVAL;

2424 2425
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2426
		return -ENODEV;
2427

2428 2429 2430
	idev = addrconf_add_dev(dev);
	if (IS_ERR(idev))
		return PTR_ERR(idev);
L
Linus Torvalds 已提交
2431 2432 2433

	scope = ipv6_addr_scope(pfx);

2434 2435 2436 2437
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
2438
		flags = RTF_EXPIRES;
2439 2440 2441 2442
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
2443
	}
2444

2445 2446 2447 2448 2449 2450
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
2451

2452 2453
	ifp = ipv6_add_addr(idev, pfx, peer_pfx, plen, scope, ifa_flags,
			    valid_lft, prefered_lft);
2454

L
Linus Torvalds 已提交
2455
	if (!IS_ERR(ifp)) {
2456 2457 2458 2459 2460
		if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
			addrconf_prefix_route(&ifp->addr, ifp->prefix_len, dev,
					      expires, flags);
		}

2461 2462 2463 2464 2465
		/*
		 * Note that section 3.1 of RFC 4429 indicates
		 * that the Optimistic flag should not be set for
		 * manually configured addresses
		 */
2466
		addrconf_dad_start(ifp);
2467 2468 2469
		if (ifa_flags & IFA_F_MANAGETEMPADDR)
			manage_tempaddrs(idev, ifp, valid_lft, prefered_lft,
					 true, jiffies);
L
Linus Torvalds 已提交
2470
		in6_ifa_put(ifp);
2471
		addrconf_verify(0);
L
Linus Torvalds 已提交
2472 2473 2474 2475 2476 2477
		return 0;
	}

	return PTR_ERR(ifp);
}

2478
static int inet6_addr_del(struct net *net, int ifindex, const struct in6_addr *pfx,
2479
			  unsigned int plen)
L
Linus Torvalds 已提交
2480 2481 2482 2483
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2484

2485 2486 2487
	if (plen > 128)
		return -EINVAL;

2488 2489
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2490 2491
		return -ENODEV;

2492
	if ((idev = __in6_dev_get(dev)) == NULL)
L
Linus Torvalds 已提交
2493 2494 2495
		return -ENXIO;

	read_lock_bh(&idev->lock);
2496
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
L
Linus Torvalds 已提交
2497 2498 2499 2500
		if (ifp->prefix_len == plen &&
		    ipv6_addr_equal(pfx, &ifp->addr)) {
			in6_ifa_hold(ifp);
			read_unlock_bh(&idev->lock);
2501

L
Linus Torvalds 已提交
2502 2503 2504 2505 2506 2507 2508 2509 2510
			ipv6_del_addr(ifp);
			return 0;
		}
	}
	read_unlock_bh(&idev->lock);
	return -EADDRNOTAVAIL;
}


2511
int addrconf_add_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2512 2513 2514
{
	struct in6_ifreq ireq;
	int err;
2515

2516
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2517
		return -EPERM;
2518

L
Linus Torvalds 已提交
2519 2520 2521 2522
	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		return -EFAULT;

	rtnl_lock();
2523
	err = inet6_addr_add(net, ireq.ifr6_ifindex, &ireq.ifr6_addr, NULL,
2524 2525
			     ireq.ifr6_prefixlen, IFA_F_PERMANENT,
			     INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
2526 2527 2528 2529
	rtnl_unlock();
	return err;
}

2530
int addrconf_del_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2531 2532 2533
{
	struct in6_ifreq ireq;
	int err;
2534

2535
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2536 2537 2538 2539 2540 2541
		return -EPERM;

	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		return -EFAULT;

	rtnl_lock();
2542 2543
	err = inet6_addr_del(net, ireq.ifr6_ifindex, &ireq.ifr6_addr,
			     ireq.ifr6_prefixlen);
L
Linus Torvalds 已提交
2544 2545 2546 2547
	rtnl_unlock();
	return err;
}

2548 2549 2550 2551 2552
static void add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
		     int plen, int scope)
{
	struct inet6_ifaddr *ifp;

2553
	ifp = ipv6_add_addr(idev, addr, NULL, plen,
2554 2555
			    scope, IFA_F_PERMANENT,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
2556 2557 2558 2559 2560 2561 2562 2563 2564
	if (!IS_ERR(ifp)) {
		spin_lock_bh(&ifp->lock);
		ifp->flags &= ~IFA_F_TENTATIVE;
		spin_unlock_bh(&ifp->lock);
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
		in6_ifa_put(ifp);
	}
}

A
Amerigo Wang 已提交
2565
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2566 2567 2568 2569
static void sit_add_v4_addrs(struct inet6_dev *idev)
{
	struct in6_addr addr;
	struct net_device *dev;
2570
	struct net *net = dev_net(idev->dev);
2571
	int scope, plen;
2572
	u32 pflags = 0;
L
Linus Torvalds 已提交
2573 2574 2575 2576 2577 2578 2579 2580 2581

	ASSERT_RTNL();

	memset(&addr, 0, sizeof(struct in6_addr));
	memcpy(&addr.s6_addr32[3], idev->dev->dev_addr, 4);

	if (idev->dev->flags&IFF_POINTOPOINT) {
		addr.s6_addr32[0] = htonl(0xfe800000);
		scope = IFA_LINK;
2582
		plen = 64;
L
Linus Torvalds 已提交
2583 2584
	} else {
		scope = IPV6_ADDR_COMPATv4;
2585
		plen = 96;
2586
		pflags |= RTF_NONEXTHOP;
L
Linus Torvalds 已提交
2587 2588 2589
	}

	if (addr.s6_addr32[3]) {
2590
		add_addr(idev, &addr, plen, scope);
2591
		addrconf_prefix_route(&addr, plen, idev->dev, 0, pflags);
L
Linus Torvalds 已提交
2592 2593 2594
		return;
	}

2595
	for_each_netdev(net, dev) {
2596
		struct in_device *in_dev = __in_dev_get_rtnl(dev);
L
Linus Torvalds 已提交
2597
		if (in_dev && (dev->flags & IFF_UP)) {
2598
			struct in_ifaddr *ifa;
L
Linus Torvalds 已提交
2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613

			int flag = scope;

			for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) {

				addr.s6_addr32[3] = ifa->ifa_local;

				if (ifa->ifa_scope == RT_SCOPE_LINK)
					continue;
				if (ifa->ifa_scope >= RT_SCOPE_HOST) {
					if (idev->dev->flags&IFF_POINTOPOINT)
						continue;
					flag |= IFA_HOST;
				}

2614
				add_addr(idev, &addr, plen, flag);
2615 2616
				addrconf_prefix_route(&addr, plen, idev->dev, 0,
						      pflags);
L
Linus Torvalds 已提交
2617 2618
			}
		}
2619
	}
L
Linus Torvalds 已提交
2620
}
2621
#endif
L
Linus Torvalds 已提交
2622 2623 2624 2625

static void init_loopback(struct net_device *dev)
{
	struct inet6_dev  *idev;
2626 2627 2628
	struct net_device *sp_dev;
	struct inet6_ifaddr *sp_ifa;
	struct rt6_info *sp_rt;
L
Linus Torvalds 已提交
2629 2630 2631 2632 2633

	/* ::1 */

	ASSERT_RTNL();

2634
	if ((idev = ipv6_find_idev(dev)) == NULL) {
2635
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
2636 2637 2638
		return;
	}

2639
	add_addr(idev, &in6addr_loopback, 128, IFA_HOST);
2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655

	/* Add routes to other interface's IPv6 addresses */
	for_each_netdev(dev_net(dev), sp_dev) {
		if (!strcmp(sp_dev->name, dev->name))
			continue;

		idev = __in6_dev_get(sp_dev);
		if (!idev)
			continue;

		read_lock_bh(&idev->lock);
		list_for_each_entry(sp_ifa, &idev->addr_list, if_list) {

			if (sp_ifa->flags & (IFA_F_DADFAILED | IFA_F_TENTATIVE))
				continue;

2656 2657 2658
			if (sp_ifa->rt)
				continue;

2659
			sp_rt = addrconf_dst_alloc(idev, &sp_ifa->addr, false);
2660 2661

			/* Failure cases are ignored */
2662 2663
			if (!IS_ERR(sp_rt)) {
				sp_ifa->rt = sp_rt;
2664
				ip6_ins_rt(sp_rt);
2665
			}
2666 2667 2668
		}
		read_unlock_bh(&idev->lock);
	}
L
Linus Torvalds 已提交
2669 2670
}

2671
static void addrconf_add_linklocal(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2672
{
2673
	struct inet6_ifaddr *ifp;
2674 2675 2676 2677
	u32 addr_flags = IFA_F_PERMANENT;

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	if (idev->cnf.optimistic_dad &&
2678
	    !dev_net(idev->dev)->ipv6.devconf_all->forwarding)
2679 2680
		addr_flags |= IFA_F_OPTIMISTIC;
#endif
L
Linus Torvalds 已提交
2681

2682

2683 2684
	ifp = ipv6_add_addr(idev, addr, NULL, 64, IFA_LINK, addr_flags,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
2685
	if (!IS_ERR(ifp)) {
2686
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, idev->dev, 0, 0);
2687
		addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
2688 2689 2690 2691 2692 2693 2694
		in6_ifa_put(ifp);
	}
}

static void addrconf_dev_config(struct net_device *dev)
{
	struct in6_addr addr;
2695
	struct inet6_dev *idev;
L
Linus Torvalds 已提交
2696 2697 2698

	ASSERT_RTNL();

2699 2700 2701
	if ((dev->type != ARPHRD_ETHER) &&
	    (dev->type != ARPHRD_FDDI) &&
	    (dev->type != ARPHRD_ARCNET) &&
2702
	    (dev->type != ARPHRD_INFINIBAND) &&
2703
	    (dev->type != ARPHRD_IEEE802154) &&
2704
	    (dev->type != ARPHRD_IEEE1394) &&
2705 2706
	    (dev->type != ARPHRD_TUNNEL6) &&
	    (dev->type != ARPHRD_6LOWPAN)) {
2707 2708 2709 2710
		/* Alas, we support only Ethernet autoconfiguration. */
		return;
	}

L
Linus Torvalds 已提交
2711
	idev = addrconf_add_dev(dev);
2712
	if (IS_ERR(idev))
L
Linus Torvalds 已提交
2713 2714 2715 2716 2717 2718 2719 2720 2721
		return;

	memset(&addr, 0, sizeof(struct in6_addr));
	addr.s6_addr32[0] = htonl(0xFE800000);

	if (ipv6_generate_eui64(addr.s6_addr + 8, dev) == 0)
		addrconf_add_linklocal(idev, &addr);
}

A
Amerigo Wang 已提交
2722
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2723 2724 2725 2726 2727 2728
static void addrconf_sit_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

2729 2730 2731
	/*
	 * Configure the tunnel with one of our IPv4
	 * addresses... we should configure all of
L
Linus Torvalds 已提交
2732 2733 2734
	 * our v4 addrs in the tunnel
	 */

2735
	if ((idev = ipv6_find_idev(dev)) == NULL) {
2736
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
2737 2738 2739
		return;
	}

F
Fred L. Templin 已提交
2740 2741 2742 2743 2744 2745 2746 2747 2748
	if (dev->priv_flags & IFF_ISATAP) {
		struct in6_addr addr;

		ipv6_addr_set(&addr,  htonl(0xFE800000), 0, 0, 0);
		if (!ipv6_generate_eui64(addr.s6_addr + 8, dev))
			addrconf_add_linklocal(idev, &addr);
		return;
	}

L
Linus Torvalds 已提交
2749 2750
	sit_add_v4_addrs(idev);

2751
	if (dev->flags&IFF_POINTOPOINT)
L
Linus Torvalds 已提交
2752 2753
		addrconf_add_mroute(dev);
}
2754
#endif
L
Linus Torvalds 已提交
2755

A
Amerigo Wang 已提交
2756
#if IS_ENABLED(CONFIG_NET_IPGRE)
2757 2758 2759 2760 2761 2762 2763
static void addrconf_gre_config(struct net_device *dev)
{
	struct inet6_dev *idev;
	struct in6_addr addr;

	ASSERT_RTNL();

2764
	if ((idev = ipv6_find_idev(dev)) == NULL) {
2765
		pr_debug("%s: add_dev failed\n", __func__);
2766 2767 2768 2769 2770 2771 2772 2773 2774
		return;
	}

	ipv6_addr_set(&addr,  htonl(0xFE800000), 0, 0, 0);
	if (!ipv6_generate_eui64(addr.s6_addr + 8, dev))
		addrconf_add_linklocal(idev, &addr);
}
#endif

L
Linus Torvalds 已提交
2775 2776 2777 2778 2779
static inline int
ipv6_inherit_linklocal(struct inet6_dev *idev, struct net_device *link_dev)
{
	struct in6_addr lladdr;

2780
	if (!ipv6_get_lladdr(link_dev, &lladdr, IFA_F_TENTATIVE)) {
L
Linus Torvalds 已提交
2781 2782 2783 2784 2785 2786
		addrconf_add_linklocal(idev, &lladdr);
		return 0;
	}
	return -1;
}

2787
static int addrconf_notify(struct notifier_block *this, unsigned long event,
2788
			   void *ptr)
L
Linus Torvalds 已提交
2789
{
2790
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
2791
	struct inet6_dev *idev = __in6_dev_get(dev);
2792
	int run_pending = 0;
2793
	int err;
L
Linus Torvalds 已提交
2794

S
Stephen Hemminger 已提交
2795
	switch (event) {
2796
	case NETDEV_REGISTER:
2797
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
2798 2799
			idev = ipv6_add_dev(dev);
			if (!idev)
2800
				return notifier_from_errno(-ENOMEM);
2801 2802
		}
		break;
S
Stephen Hemminger 已提交
2803

L
Linus Torvalds 已提交
2804
	case NETDEV_UP:
2805
	case NETDEV_CHANGE:
2806 2807 2808
		if (dev->flags & IFF_SLAVE)
			break;

2809
		if (event == NETDEV_UP) {
2810
			if (!addrconf_qdisc_ok(dev)) {
2811
				/* device is not ready yet. */
2812
				pr_info("ADDRCONF(NETDEV_UP): %s: link is not ready\n",
2813 2814 2815
					dev->name);
				break;
			}
2816

2817 2818 2819
			if (!idev && dev->mtu >= IPV6_MIN_MTU)
				idev = ipv6_add_dev(dev);

2820
			if (idev) {
2821
				idev->if_flags |= IF_READY;
2822 2823
				run_pending = 1;
			}
2824
		} else {
2825
			if (!addrconf_qdisc_ok(dev)) {
2826 2827 2828 2829 2830
				/* device is still not ready. */
				break;
			}

			if (idev) {
S
Stephen Hemminger 已提交
2831
				if (idev->if_flags & IF_READY)
2832 2833 2834 2835 2836
					/* device is already configured. */
					break;
				idev->if_flags |= IF_READY;
			}

2837 2838
			pr_info("ADDRCONF(NETDEV_CHANGE): %s: link becomes ready\n",
				dev->name);
2839

2840
			run_pending = 1;
2841 2842
		}

S
Stephen Hemminger 已提交
2843
		switch (dev->type) {
A
Amerigo Wang 已提交
2844
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2845 2846 2847
		case ARPHRD_SIT:
			addrconf_sit_config(dev);
			break;
2848
#endif
A
Amerigo Wang 已提交
2849
#if IS_ENABLED(CONFIG_NET_IPGRE)
2850 2851 2852
		case ARPHRD_IPGRE:
			addrconf_gre_config(dev);
			break;
2853
#endif
L
Linus Torvalds 已提交
2854 2855 2856 2857 2858 2859 2860
		case ARPHRD_LOOPBACK:
			init_loopback(dev);
			break;

		default:
			addrconf_dev_config(dev);
			break;
2861
		}
S
Stephen Hemminger 已提交
2862

L
Linus Torvalds 已提交
2863
		if (idev) {
2864 2865 2866
			if (run_pending)
				addrconf_dad_run(idev);

S
Stephen Hemminger 已提交
2867 2868 2869 2870
			/*
			 * If the MTU changed during the interface down,
			 * when the interface up, the changed MTU must be
			 * reflected in the idev as well as routers.
L
Linus Torvalds 已提交
2871
			 */
S
Stephen Hemminger 已提交
2872 2873
			if (idev->cnf.mtu6 != dev->mtu &&
			    dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
2874 2875 2876 2877 2878
				rt6_mtu_change(dev, dev->mtu);
				idev->cnf.mtu6 = dev->mtu;
			}
			idev->tstamp = jiffies;
			inet6_ifinfo_notify(RTM_NEWLINK, idev);
S
Stephen Hemminger 已提交
2879 2880 2881 2882

			/*
			 * If the changed mtu during down is lower than
			 * IPV6_MIN_MTU stop IPv6 on this interface.
L
Linus Torvalds 已提交
2883 2884
			 */
			if (dev->mtu < IPV6_MIN_MTU)
S
Stephen Hemminger 已提交
2885
				addrconf_ifdown(dev, 1);
L
Linus Torvalds 已提交
2886 2887 2888 2889
		}
		break;

	case NETDEV_CHANGEMTU:
2890
		if (idev && dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
2891 2892 2893 2894 2895
			rt6_mtu_change(dev, dev->mtu);
			idev->cnf.mtu6 = dev->mtu;
			break;
		}

2896 2897 2898 2899 2900 2901
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
			idev = ipv6_add_dev(dev);
			if (idev)
				break;
		}

S
Stephen Hemminger 已提交
2902
		/*
2903
		 * if MTU under IPV6_MIN_MTU.
S
Stephen Hemminger 已提交
2904 2905
		 * Stop IPv6 on this interface.
		 */
L
Linus Torvalds 已提交
2906 2907 2908 2909 2910 2911 2912 2913

	case NETDEV_DOWN:
	case NETDEV_UNREGISTER:
		/*
		 *	Remove all addresses from this interface.
		 */
		addrconf_ifdown(dev, event != NETDEV_DOWN);
		break;
2914

L
Linus Torvalds 已提交
2915 2916
	case NETDEV_CHANGENAME:
		if (idev) {
2917
			snmp6_unregister_dev(idev);
2918
			addrconf_sysctl_unregister(idev);
2919
			addrconf_sysctl_register(idev);
2920 2921 2922
			err = snmp6_register_dev(idev);
			if (err)
				return notifier_from_errno(err);
2923
		}
L
Linus Torvalds 已提交
2924
		break;
S
Stephen Hemminger 已提交
2925

2926 2927 2928
	case NETDEV_PRE_TYPE_CHANGE:
	case NETDEV_POST_TYPE_CHANGE:
		addrconf_type_change(dev, event);
2929
		break;
2930
	}
L
Linus Torvalds 已提交
2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941

	return NOTIFY_OK;
}

/*
 *	addrconf module should be notified of a device going up
 */
static struct notifier_block ipv6_dev_notf = {
	.notifier_call = addrconf_notify,
};

2942
static void addrconf_type_change(struct net_device *dev, unsigned long event)
2943 2944 2945 2946 2947 2948
{
	struct inet6_dev *idev;
	ASSERT_RTNL();

	idev = __in6_dev_get(dev);

2949
	if (event == NETDEV_POST_TYPE_CHANGE)
2950
		ipv6_mc_remap(idev);
2951
	else if (event == NETDEV_PRE_TYPE_CHANGE)
2952 2953 2954
		ipv6_mc_unmap(idev);
}

L
Linus Torvalds 已提交
2955 2956
static int addrconf_ifdown(struct net_device *dev, int how)
{
2957
	struct net *net = dev_net(dev);
2958 2959
	struct inet6_dev *idev;
	struct inet6_ifaddr *ifa;
2960
	int state, i;
L
Linus Torvalds 已提交
2961 2962 2963

	ASSERT_RTNL();

2964 2965
	rt6_ifdown(net, dev);
	neigh_ifdown(&nd_tbl, dev);
L
Linus Torvalds 已提交
2966 2967 2968 2969 2970

	idev = __in6_dev_get(dev);
	if (idev == NULL)
		return -ENODEV;

S
Stephen Hemminger 已提交
2971 2972 2973
	/*
	 * Step 1: remove reference to ipv6 device from parent device.
	 *	   Do not dev_put!
L
Linus Torvalds 已提交
2974
	 */
2975
	if (how) {
L
Linus Torvalds 已提交
2976
		idev->dead = 1;
2977 2978

		/* protected by rtnl_lock */
2979
		RCU_INIT_POINTER(dev->ip6_ptr, NULL);
L
Linus Torvalds 已提交
2980 2981 2982 2983 2984 2985

		/* Step 1.5: remove snmp6 entry */
		snmp6_unregister_dev(idev);

	}

2986 2987 2988 2989 2990 2991
	/* Step 2: clear hash table */
	for (i = 0; i < IN6_ADDR_HSIZE; i++) {
		struct hlist_head *h = &inet6_addr_lst[i];

		spin_lock_bh(&addrconf_hash_lock);
	restart:
2992
		hlist_for_each_entry_rcu(ifa, h, addr_lst) {
2993 2994
			if (ifa->idev == idev) {
				hlist_del_init_rcu(&ifa->addr_lst);
2995
				addrconf_del_dad_timer(ifa);
2996 2997 2998 2999 3000 3001
				goto restart;
			}
		}
		spin_unlock_bh(&addrconf_hash_lock);
	}

L
Linus Torvalds 已提交
3002 3003
	write_lock_bh(&idev->lock);

3004 3005
	addrconf_del_rs_timer(idev);

S
Stephen Hemminger 已提交
3006
	/* Step 2: clear flags for stateless addrconf */
3007
	if (!how)
3008
		idev->if_flags &= ~(IF_RS_SENT|IF_RA_RCVD|IF_READY);
L
Linus Torvalds 已提交
3009

3010
	if (how && del_timer(&idev->regen_timer))
L
Linus Torvalds 已提交
3011 3012
		in6_dev_put(idev);

S
Stephen Hemminger 已提交
3013
	/* Step 3: clear tempaddr list */
3014 3015 3016 3017
	while (!list_empty(&idev->tempaddr_list)) {
		ifa = list_first_entry(&idev->tempaddr_list,
				       struct inet6_ifaddr, tmp_list);
		list_del(&ifa->tmp_list);
L
Linus Torvalds 已提交
3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028
		write_unlock_bh(&idev->lock);
		spin_lock_bh(&ifa->lock);

		if (ifa->ifpub) {
			in6_ifa_put(ifa->ifpub);
			ifa->ifpub = NULL;
		}
		spin_unlock_bh(&ifa->lock);
		in6_ifa_put(ifa);
		write_lock_bh(&idev->lock);
	}
3029

3030 3031 3032
	while (!list_empty(&idev->addr_list)) {
		ifa = list_first_entry(&idev->addr_list,
				       struct inet6_ifaddr, if_list);
3033
		addrconf_del_dad_timer(ifa);
3034

3035
		list_del(&ifa->if_list);
3036

3037
		write_unlock_bh(&idev->lock);
3038

3039 3040 3041 3042
		spin_lock_bh(&ifa->state_lock);
		state = ifa->state;
		ifa->state = INET6_IFADDR_STATE_DEAD;
		spin_unlock_bh(&ifa->state_lock);
3043

3044 3045
		if (state != INET6_IFADDR_STATE_DEAD) {
			__ipv6_ifa_notify(RTM_DELADDR, ifa);
3046
			inet6addr_notifier_call_chain(NETDEV_DOWN, ifa);
3047
		}
3048
		in6_ifa_put(ifa);
3049

3050 3051
		write_lock_bh(&idev->lock);
	}
3052

L
Linus Torvalds 已提交
3053 3054 3055
	write_unlock_bh(&idev->lock);

	/* Step 5: Discard multicast list */
3056
	if (how)
L
Linus Torvalds 已提交
3057 3058 3059 3060 3061
		ipv6_mc_destroy_dev(idev);
	else
		ipv6_mc_down(idev);

	idev->tstamp = jiffies;
3062

S
Stephen Hemminger 已提交
3063
	/* Last: Shot the device (if unregistered) */
3064
	if (how) {
3065
		addrconf_sysctl_unregister(idev);
L
Linus Torvalds 已提交
3066 3067 3068 3069 3070 3071 3072 3073 3074
		neigh_parms_release(&nd_tbl, idev->nd_parms);
		neigh_ifdown(&nd_tbl, dev);
		in6_dev_put(idev);
	}
	return 0;
}

static void addrconf_rs_timer(unsigned long data)
{
3075
	struct inet6_dev *idev = (struct inet6_dev *)data;
3076
	struct net_device *dev = idev->dev;
3077
	struct in6_addr lladdr;
L
Linus Torvalds 已提交
3078

3079
	write_lock(&idev->lock);
S
stephen hemminger 已提交
3080
	if (idev->dead || !(idev->if_flags & IF_READY))
L
Linus Torvalds 已提交
3081 3082
		goto out;

3083
	if (!ipv6_accept_ra(idev))
S
stephen hemminger 已提交
3084 3085 3086 3087
		goto out;

	/* Announcement received after solicitation was sent */
	if (idev->if_flags & IF_RA_RCVD)
L
Linus Torvalds 已提交
3088 3089
		goto out;

3090
	if (idev->rs_probes++ < idev->cnf.rtr_solicits) {
3091 3092 3093
		write_unlock(&idev->lock);
		if (!ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
			ndisc_send_rs(dev, &lladdr,
3094 3095
				      &in6addr_linklocal_allrouters);
		else
3096
			goto put;
L
Linus Torvalds 已提交
3097

3098
		write_lock(&idev->lock);
3099 3100 3101 3102 3103
		/* The wait after the last probe can be shorter */
		addrconf_mod_rs_timer(idev, (idev->rs_probes ==
					     idev->cnf.rtr_solicits) ?
				      idev->cnf.rtr_solicit_delay :
				      idev->cnf.rtr_solicit_interval);
L
Linus Torvalds 已提交
3104 3105 3106 3107 3108
	} else {
		/*
		 * Note: we do not support deprecated "all on-link"
		 * assumption any longer.
		 */
3109
		pr_debug("%s: no IPv6 routers present\n", idev->dev->name);
L
Linus Torvalds 已提交
3110 3111 3112
	}

out:
3113
	write_unlock(&idev->lock);
3114
put:
3115
	in6_dev_put(idev);
L
Linus Torvalds 已提交
3116 3117 3118 3119 3120
}

/*
 *	Duplicate Address Detection
 */
3121 3122 3123 3124 3125
static void addrconf_dad_kick(struct inet6_ifaddr *ifp)
{
	unsigned long rand_num;
	struct inet6_dev *idev = ifp->idev;

3126 3127 3128
	if (ifp->flags & IFA_F_OPTIMISTIC)
		rand_num = 0;
	else
3129
		rand_num = prandom_u32() % (idev->cnf.rtr_solicit_delay ? : 1);
3130

3131 3132
	ifp->dad_probes = idev->cnf.dad_transmits;
	addrconf_mod_dad_timer(ifp, rand_num);
3133 3134
}

3135
static void addrconf_dad_start(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3136 3137 3138 3139 3140 3141
{
	struct inet6_dev *idev = ifp->idev;
	struct net_device *dev = idev->dev;

	addrconf_join_solict(dev, &ifp->addr);

3142
	prandom_seed((__force u32) ifp->addr.s6_addr32[3]);
L
Linus Torvalds 已提交
3143 3144

	read_lock_bh(&idev->lock);
3145
	spin_lock(&ifp->lock);
3146
	if (ifp->state == INET6_IFADDR_STATE_DEAD)
L
Linus Torvalds 已提交
3147 3148 3149
		goto out;

	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
3150
	    idev->cnf.accept_dad < 1 ||
3151 3152
	    !(ifp->flags&IFA_F_TENTATIVE) ||
	    ifp->flags & IFA_F_NODAD) {
B
Brian Haley 已提交
3153
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3154
		spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3155 3156 3157 3158 3159 3160
		read_unlock_bh(&idev->lock);

		addrconf_dad_completed(ifp);
		return;
	}

3161
	if (!(idev->if_flags & IF_READY)) {
3162
		spin_unlock(&ifp->lock);
3163
		read_unlock_bh(&idev->lock);
3164
		/*
3165
		 * If the device is not ready:
3166 3167 3168 3169
		 * - keep it tentative if it is a permanent address.
		 * - otherwise, kill it.
		 */
		in6_ifa_hold(ifp);
B
Brian Haley 已提交
3170
		addrconf_dad_stop(ifp, 0);
3171
		return;
3172
	}
3173 3174 3175 3176 3177

	/*
	 * Optimistic nodes can start receiving
	 * Frames right away
	 */
S
Stephen Hemminger 已提交
3178
	if (ifp->flags & IFA_F_OPTIMISTIC)
3179 3180
		ip6_ins_rt(ifp->rt);

3181
	addrconf_dad_kick(ifp);
L
Linus Torvalds 已提交
3182
out:
3183
	spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3184 3185 3186 3187 3188 3189 3190 3191 3192
	read_unlock_bh(&idev->lock);
}

static void addrconf_dad_timer(unsigned long data)
{
	struct inet6_ifaddr *ifp = (struct inet6_ifaddr *) data;
	struct inet6_dev *idev = ifp->idev;
	struct in6_addr mcaddr;

3193
	if (!ifp->dad_probes && addrconf_dad_end(ifp))
H
Herbert Xu 已提交
3194 3195
		goto out;

3196
	write_lock(&idev->lock);
3197
	if (idev->dead || !(idev->if_flags & IF_READY)) {
3198
		write_unlock(&idev->lock);
L
Linus Torvalds 已提交
3199 3200
		goto out;
	}
3201 3202

	spin_lock(&ifp->lock);
3203 3204
	if (ifp->state == INET6_IFADDR_STATE_DEAD) {
		spin_unlock(&ifp->lock);
3205
		write_unlock(&idev->lock);
3206 3207 3208
		goto out;
	}

3209
	if (ifp->dad_probes == 0) {
L
Linus Torvalds 已提交
3210 3211 3212 3213
		/*
		 * DAD was successful
		 */

B
Brian Haley 已提交
3214
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3215
		spin_unlock(&ifp->lock);
3216
		write_unlock(&idev->lock);
L
Linus Torvalds 已提交
3217 3218 3219 3220 3221 3222

		addrconf_dad_completed(ifp);

		goto out;
	}

3223
	ifp->dad_probes--;
J
Jiri Pirko 已提交
3224 3225
	addrconf_mod_dad_timer(ifp,
			       NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME));
3226
	spin_unlock(&ifp->lock);
3227
	write_unlock(&idev->lock);
L
Linus Torvalds 已提交
3228 3229 3230

	/* send a neighbour solicitation for our addr */
	addrconf_addr_solict_mult(&ifp->addr, &mcaddr);
3231
	ndisc_send_ns(ifp->idev->dev, NULL, &ifp->addr, &mcaddr, &in6addr_any);
L
Linus Torvalds 已提交
3232 3233 3234 3235 3236 3237
out:
	in6_ifa_put(ifp);
}

static void addrconf_dad_completed(struct inet6_ifaddr *ifp)
{
S
Stephen Hemminger 已提交
3238
	struct net_device *dev = ifp->idev->dev;
3239
	struct in6_addr lladdr;
3240
	bool send_rs, send_mld;
3241 3242

	addrconf_del_dad_timer(ifp);
L
Linus Torvalds 已提交
3243 3244 3245 3246 3247 3248 3249

	/*
	 *	Configure the address for reception. Now it is valid.
	 */

	ipv6_ifa_notify(RTM_NEWADDR, ifp);

3250 3251
	/* If added prefix is link local and we are prepared to process
	   router advertisements, start sending router solicitations.
L
Linus Torvalds 已提交
3252 3253
	 */

3254 3255
	read_lock_bh(&ifp->idev->lock);
	spin_lock(&ifp->lock);
3256 3257 3258 3259
	send_mld = ipv6_addr_type(&ifp->addr) & IPV6_ADDR_LINKLOCAL &&
		   ifp->idev->valid_ll_addr_cnt == 1;
	send_rs = send_mld &&
		  ipv6_accept_ra(ifp->idev) &&
3260
		  ifp->idev->cnf.rtr_solicits > 0 &&
3261
		  (dev->flags&IFF_LOOPBACK) == 0;
3262 3263 3264
	spin_unlock(&ifp->lock);
	read_unlock_bh(&ifp->idev->lock);

3265 3266 3267 3268 3269 3270
	/* While dad is in progress mld report's source address is in6_addrany.
	 * Resend with proper ll now.
	 */
	if (send_mld)
		ipv6_mc_dad_complete(ifp->idev);

3271
	if (send_rs) {
L
Linus Torvalds 已提交
3272 3273 3274 3275 3276
		/*
		 *	If a host as already performed a random delay
		 *	[...] as part of DAD [...] there is no need
		 *	to delay again before sending the first RS
		 */
3277
		if (ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
3278
			return;
3279
		ndisc_send_rs(dev, &lladdr, &in6addr_linklocal_allrouters);
L
Linus Torvalds 已提交
3280

3281 3282 3283
		write_lock_bh(&ifp->idev->lock);
		spin_lock(&ifp->lock);
		ifp->idev->rs_probes = 1;
L
Linus Torvalds 已提交
3284
		ifp->idev->if_flags |= IF_RS_SENT;
3285 3286 3287 3288
		addrconf_mod_rs_timer(ifp->idev,
				      ifp->idev->cnf.rtr_solicit_interval);
		spin_unlock(&ifp->lock);
		write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
3289 3290 3291
	}
}

S
Stephen Hemminger 已提交
3292 3293
static void addrconf_dad_run(struct inet6_dev *idev)
{
3294 3295 3296
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
3297
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
3298
		spin_lock(&ifp->lock);
H
Herbert Xu 已提交
3299 3300 3301
		if (ifp->flags & IFA_F_TENTATIVE &&
		    ifp->state == INET6_IFADDR_STATE_DAD)
			addrconf_dad_kick(ifp);
3302
		spin_unlock(&ifp->lock);
3303 3304 3305 3306
	}
	read_unlock_bh(&idev->lock);
}

L
Linus Torvalds 已提交
3307 3308
#ifdef CONFIG_PROC_FS
struct if6_iter_state {
3309
	struct seq_net_private p;
L
Linus Torvalds 已提交
3310
	int bucket;
3311
	int offset;
L
Linus Torvalds 已提交
3312 3313
};

3314
static struct inet6_ifaddr *if6_get_first(struct seq_file *seq, loff_t pos)
L
Linus Torvalds 已提交
3315 3316 3317
{
	struct inet6_ifaddr *ifa = NULL;
	struct if6_iter_state *state = seq->private;
3318
	struct net *net = seq_file_net(seq);
3319
	int p = 0;
L
Linus Torvalds 已提交
3320

3321 3322 3323 3324 3325 3326 3327
	/* initial bucket if pos is 0 */
	if (pos == 0) {
		state->bucket = 0;
		state->offset = 0;
	}

	for (; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
3328
		hlist_for_each_entry_rcu_bh(ifa, &inet6_addr_lst[state->bucket],
3329
					 addr_lst) {
3330 3331
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
3332 3333 3334 3335 3336 3337
			/* sync with offset */
			if (p < state->offset) {
				p++;
				continue;
			}
			state->offset++;
3338
			return ifa;
3339 3340 3341 3342 3343
		}

		/* prepare for next bucket */
		state->offset = 0;
		p = 0;
L
Linus Torvalds 已提交
3344
	}
3345
	return NULL;
L
Linus Torvalds 已提交
3346 3347
}

3348 3349
static struct inet6_ifaddr *if6_get_next(struct seq_file *seq,
					 struct inet6_ifaddr *ifa)
L
Linus Torvalds 已提交
3350 3351
{
	struct if6_iter_state *state = seq->private;
3352
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
3353

3354
	hlist_for_each_entry_continue_rcu_bh(ifa, addr_lst) {
3355 3356
		if (!net_eq(dev_net(ifa->idev->dev), net))
			continue;
3357
		state->offset++;
3358
		return ifa;
3359
	}
3360

3361
	while (++state->bucket < IN6_ADDR_HSIZE) {
3362
		state->offset = 0;
3363
		hlist_for_each_entry_rcu_bh(ifa,
3364
				     &inet6_addr_lst[state->bucket], addr_lst) {
3365 3366
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
3367
			state->offset++;
3368
			return ifa;
3369
		}
L
Linus Torvalds 已提交
3370
	}
3371

3372
	return NULL;
L
Linus Torvalds 已提交
3373 3374 3375
}

static void *if6_seq_start(struct seq_file *seq, loff_t *pos)
E
Eric Dumazet 已提交
3376
	__acquires(rcu_bh)
L
Linus Torvalds 已提交
3377
{
3378
	rcu_read_lock_bh();
3379
	return if6_get_first(seq, *pos);
L
Linus Torvalds 已提交
3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391
}

static void *if6_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	struct inet6_ifaddr *ifa;

	ifa = if6_get_next(seq, v);
	++*pos;
	return ifa;
}

static void if6_seq_stop(struct seq_file *seq, void *v)
E
Eric Dumazet 已提交
3392
	__releases(rcu_bh)
L
Linus Torvalds 已提交
3393
{
3394
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3395 3396 3397 3398 3399
}

static int if6_seq_show(struct seq_file *seq, void *v)
{
	struct inet6_ifaddr *ifp = (struct inet6_ifaddr *)v;
3400
	seq_printf(seq, "%pi6 %02x %02x %02x %02x %8s\n",
3401
		   &ifp->addr,
L
Linus Torvalds 已提交
3402 3403 3404
		   ifp->idev->dev->ifindex,
		   ifp->prefix_len,
		   ifp->scope,
3405
		   (u8) ifp->flags,
L
Linus Torvalds 已提交
3406 3407 3408 3409
		   ifp->idev->dev->name);
	return 0;
}

3410
static const struct seq_operations if6_seq_ops = {
L
Linus Torvalds 已提交
3411 3412 3413 3414 3415 3416 3417 3418
	.start	= if6_seq_start,
	.next	= if6_seq_next,
	.show	= if6_seq_show,
	.stop	= if6_seq_stop,
};

static int if6_seq_open(struct inode *inode, struct file *file)
{
3419 3420
	return seq_open_net(inode, file, &if6_seq_ops,
			    sizeof(struct if6_iter_state));
L
Linus Torvalds 已提交
3421 3422
}

3423
static const struct file_operations if6_fops = {
L
Linus Torvalds 已提交
3424 3425 3426 3427
	.owner		= THIS_MODULE,
	.open		= if6_seq_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
3428
	.release	= seq_release_net,
L
Linus Torvalds 已提交
3429 3430
};

3431
static int __net_init if6_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
3432
{
3433
	if (!proc_create("if_inet6", S_IRUGO, net->proc_net, &if6_fops))
L
Linus Torvalds 已提交
3434 3435 3436 3437
		return -ENOMEM;
	return 0;
}

3438
static void __net_exit if6_proc_net_exit(struct net *net)
3439
{
3440
	remove_proc_entry("if_inet6", net->proc_net);
3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452
}

static struct pernet_operations if6_proc_net_ops = {
       .init = if6_proc_net_init,
       .exit = if6_proc_net_exit,
};

int __init if6_proc_init(void)
{
	return register_pernet_subsys(&if6_proc_net_ops);
}

L
Linus Torvalds 已提交
3453 3454
void if6_proc_exit(void)
{
3455
	unregister_pernet_subsys(&if6_proc_net_ops);
L
Linus Torvalds 已提交
3456 3457 3458
}
#endif	/* CONFIG_PROC_FS */

A
Amerigo Wang 已提交
3459
#if IS_ENABLED(CONFIG_IPV6_MIP6)
3460
/* Check if address is a home address configured on any interface. */
3461
int ipv6_chk_home_addr(struct net *net, const struct in6_addr *addr)
3462 3463
{
	int ret = 0;
3464
	struct inet6_ifaddr *ifp = NULL;
E
Eric Dumazet 已提交
3465
	unsigned int hash = inet6_addr_hash(addr);
3466

3467
	rcu_read_lock_bh();
3468
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
3469
		if (!net_eq(dev_net(ifp->idev->dev), net))
3470
			continue;
3471
		if (ipv6_addr_equal(&ifp->addr, addr) &&
3472 3473 3474 3475 3476
		    (ifp->flags & IFA_F_HOMEADDRESS)) {
			ret = 1;
			break;
		}
	}
3477
	rcu_read_unlock_bh();
3478 3479 3480 3481
	return ret;
}
#endif

L
Linus Torvalds 已提交
3482 3483 3484 3485 3486 3487
/*
 *	Periodic address status verification
 */

static void addrconf_verify(unsigned long foo)
{
3488
	unsigned long now, next, next_sec, next_sched;
L
Linus Torvalds 已提交
3489 3490 3491
	struct inet6_ifaddr *ifp;
	int i;

3492 3493
	rcu_read_lock_bh();
	spin_lock(&addrconf_verify_lock);
L
Linus Torvalds 已提交
3494
	now = jiffies;
3495
	next = round_jiffies_up(now + ADDR_CHECK_FREQUENCY);
L
Linus Torvalds 已提交
3496 3497 3498

	del_timer(&addr_chk_timer);

S
Stephen Hemminger 已提交
3499
	for (i = 0; i < IN6_ADDR_HSIZE; i++) {
L
Linus Torvalds 已提交
3500
restart:
3501
		hlist_for_each_entry_rcu_bh(ifp,
3502
					 &inet6_addr_lst[i], addr_lst) {
L
Linus Torvalds 已提交
3503 3504
			unsigned long age;

3505 3506 3507 3508 3509 3510
			/* When setting preferred_lft to a value not zero or
			 * infinity, while valid_lft is infinity
			 * IFA_F_PERMANENT has a non-infinity life time.
			 */
			if ((ifp->flags & IFA_F_PERMANENT) &&
			    (ifp->prefered_lft == INFINITY_LIFE_TIME))
L
Linus Torvalds 已提交
3511 3512 3513
				continue;

			spin_lock(&ifp->lock);
3514 3515
			/* We try to batch several events at once. */
			age = (now - ifp->tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
L
Linus Torvalds 已提交
3516

3517 3518
			if (ifp->valid_lft != INFINITY_LIFE_TIME &&
			    age >= ifp->valid_lft) {
L
Linus Torvalds 已提交
3519 3520 3521 3522
				spin_unlock(&ifp->lock);
				in6_ifa_hold(ifp);
				ipv6_del_addr(ifp);
				goto restart;
3523 3524 3525
			} else if (ifp->prefered_lft == INFINITY_LIFE_TIME) {
				spin_unlock(&ifp->lock);
				continue;
L
Linus Torvalds 已提交
3526
			} else if (age >= ifp->prefered_lft) {
3527
				/* jiffies - ifp->tstamp > age >= ifp->prefered_lft */
L
Linus Torvalds 已提交
3528 3529 3530 3531 3532 3533 3534
				int deprecate = 0;

				if (!(ifp->flags&IFA_F_DEPRECATED)) {
					deprecate = 1;
					ifp->flags |= IFA_F_DEPRECATED;
				}

3535 3536
				if ((ifp->valid_lft != INFINITY_LIFE_TIME) &&
				    (time_before(ifp->tstamp + ifp->valid_lft * HZ, next)))
L
Linus Torvalds 已提交
3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549
					next = ifp->tstamp + ifp->valid_lft * HZ;

				spin_unlock(&ifp->lock);

				if (deprecate) {
					in6_ifa_hold(ifp);

					ipv6_ifa_notify(0, ifp);
					in6_ifa_put(ifp);
					goto restart;
				}
			} else if ((ifp->flags&IFA_F_TEMPORARY) &&
				   !(ifp->flags&IFA_F_TENTATIVE)) {
3550 3551
				unsigned long regen_advance = ifp->idev->cnf.regen_max_retry *
					ifp->idev->cnf.dad_transmits *
J
Jiri Pirko 已提交
3552
					NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME) / HZ;
3553

L
Linus Torvalds 已提交
3554 3555 3556 3557 3558 3559 3560 3561 3562
				if (age >= ifp->prefered_lft - regen_advance) {
					struct inet6_ifaddr *ifpub = ifp->ifpub;
					if (time_before(ifp->tstamp + ifp->prefered_lft * HZ, next))
						next = ifp->tstamp + ifp->prefered_lft * HZ;
					if (!ifp->regen_count && ifpub) {
						ifp->regen_count++;
						in6_ifa_hold(ifp);
						in6_ifa_hold(ifpub);
						spin_unlock(&ifp->lock);
3563

3564 3565 3566
						spin_lock(&ifpub->lock);
						ifpub->regen_count = 0;
						spin_unlock(&ifpub->lock);
L
Linus Torvalds 已提交
3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583
						ipv6_create_tempaddr(ifpub, ifp);
						in6_ifa_put(ifpub);
						in6_ifa_put(ifp);
						goto restart;
					}
				} else if (time_before(ifp->tstamp + ifp->prefered_lft * HZ - regen_advance * HZ, next))
					next = ifp->tstamp + ifp->prefered_lft * HZ - regen_advance * HZ;
				spin_unlock(&ifp->lock);
			} else {
				/* ifp->prefered_lft <= ifp->valid_lft */
				if (time_before(ifp->tstamp + ifp->prefered_lft * HZ, next))
					next = ifp->tstamp + ifp->prefered_lft * HZ;
				spin_unlock(&ifp->lock);
			}
		}
	}

3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594
	next_sec = round_jiffies_up(next);
	next_sched = next;

	/* If rounded timeout is accurate enough, accept it. */
	if (time_before(next_sec, next + ADDRCONF_TIMER_FUZZ))
		next_sched = next_sec;

	/* And minimum interval is ADDRCONF_TIMER_FUZZ_MAX. */
	if (time_before(next_sched, jiffies + ADDRCONF_TIMER_FUZZ_MAX))
		next_sched = jiffies + ADDRCONF_TIMER_FUZZ_MAX;

3595 3596
	ADBG(KERN_DEBUG "now = %lu, schedule = %lu, rounded schedule = %lu => %lu\n",
	      now, next, next_sec, next_sched);
3597 3598

	addr_chk_timer.expires = next_sched;
L
Linus Torvalds 已提交
3599
	add_timer(&addr_chk_timer);
3600 3601
	spin_unlock(&addrconf_verify_lock);
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3602 3603
}

3604 3605
static struct in6_addr *extract_addr(struct nlattr *addr, struct nlattr *local,
				     struct in6_addr **peer_pfx)
3606 3607 3608
{
	struct in6_addr *pfx = NULL;

3609 3610
	*peer_pfx = NULL;

3611 3612 3613 3614 3615
	if (addr)
		pfx = nla_data(addr);

	if (local) {
		if (pfx && nla_memcmp(local, pfx, sizeof(*pfx)))
3616 3617
			*peer_pfx = pfx;
		pfx = nla_data(local);
3618 3619 3620 3621 3622
	}

	return pfx;
}

3623
static const struct nla_policy ifa_ipv6_policy[IFA_MAX+1] = {
3624 3625 3626
	[IFA_ADDRESS]		= { .len = sizeof(struct in6_addr) },
	[IFA_LOCAL]		= { .len = sizeof(struct in6_addr) },
	[IFA_CACHEINFO]		= { .len = sizeof(struct ifa_cacheinfo) },
3627
	[IFA_FLAGS]		= { .len = sizeof(u32) },
3628 3629
};

L
Linus Torvalds 已提交
3630
static int
3631
inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
3632
{
3633
	struct net *net = sock_net(skb->sk);
3634 3635
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
3636
	struct in6_addr *pfx, *peer_pfx;
3637
	int err;
L
Linus Torvalds 已提交
3638

3639 3640 3641 3642 3643
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
3644
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
L
Linus Torvalds 已提交
3645 3646 3647
	if (pfx == NULL)
		return -EINVAL;

3648
	return inet6_addr_del(net, ifm->ifa_index, pfx, ifm->ifa_prefixlen);
L
Linus Torvalds 已提交
3649 3650
}

3651
static int inet6_addr_modify(struct inet6_ifaddr *ifp, u32 ifa_flags,
3652
			     u32 prefered_lft, u32 valid_lft)
3653
{
3654 3655
	u32 flags;
	clock_t expires;
3656
	unsigned long timeout;
3657
	bool was_managetempaddr;
3658
	bool had_prefixroute;
3659

3660 3661 3662
	if (!valid_lft || (prefered_lft > valid_lft))
		return -EINVAL;

3663 3664 3665 3666
	if (ifa_flags & IFA_F_MANAGETEMPADDR &&
	    (ifp->flags & IFA_F_TEMPORARY || ifp->prefix_len != 64))
		return -EINVAL;

3667 3668 3669 3670
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
3671
		flags = RTF_EXPIRES;
3672 3673 3674 3675
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
3676
	}
3677

3678 3679 3680 3681 3682 3683
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
3684 3685

	spin_lock_bh(&ifp->lock);
3686
	was_managetempaddr = ifp->flags & IFA_F_MANAGETEMPADDR;
3687 3688
	had_prefixroute = ifp->flags & IFA_F_PERMANENT &&
			  !(ifp->flags & IFA_F_NOPREFIXROUTE);
3689
	ifp->flags &= ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD |
3690 3691
			IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
			IFA_F_NOPREFIXROUTE);
3692
	ifp->flags |= ifa_flags;
3693 3694 3695 3696 3697 3698 3699 3700
	ifp->tstamp = jiffies;
	ifp->valid_lft = valid_lft;
	ifp->prefered_lft = prefered_lft;

	spin_unlock_bh(&ifp->lock);
	if (!(ifp->flags&IFA_F_TENTATIVE))
		ipv6_ifa_notify(0, ifp);

3701 3702 3703
	if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, ifp->idev->dev,
				      expires, flags);
3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715
	} else if (had_prefixroute) {
		enum cleanup_prefix_rt_t action;
		unsigned long rt_expires;

		write_lock_bh(&ifp->idev->lock);
		action = check_cleanup_prefix_route(ifp, &rt_expires);
		write_unlock_bh(&ifp->idev->lock);

		if (action != CLEANUP_PREFIX_RT_NOP) {
			cleanup_prefix_route(ifp, rt_expires,
				action == CLEANUP_PREFIX_RT_DEL);
		}
3716
	}
3717 3718 3719 3720 3721 3722 3723 3724

	if (was_managetempaddr || ifp->flags & IFA_F_MANAGETEMPADDR) {
		if (was_managetempaddr && !(ifp->flags & IFA_F_MANAGETEMPADDR))
			valid_lft = prefered_lft = 0;
		manage_tempaddrs(ifp->idev, ifp, valid_lft, prefered_lft,
				 !was_managetempaddr, jiffies);
	}

3725 3726 3727 3728 3729
	addrconf_verify(0);

	return 0;
}

L
Linus Torvalds 已提交
3730
static int
3731
inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
3732
{
3733
	struct net *net = sock_net(skb->sk);
3734 3735
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
3736
	struct in6_addr *pfx, *peer_pfx;
3737 3738
	struct inet6_ifaddr *ifa;
	struct net_device *dev;
3739
	u32 valid_lft = INFINITY_LIFE_TIME, preferred_lft = INFINITY_LIFE_TIME;
3740
	u32 ifa_flags;
3741
	int err;
L
Linus Torvalds 已提交
3742

3743 3744 3745 3746 3747
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
3748
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
L
Linus Torvalds 已提交
3749 3750 3751
	if (pfx == NULL)
		return -EINVAL;

3752
	if (tb[IFA_CACHEINFO]) {
3753
		struct ifa_cacheinfo *ci;
3754 3755

		ci = nla_data(tb[IFA_CACHEINFO]);
3756
		valid_lft = ci->ifa_valid;
3757 3758 3759 3760
		preferred_lft = ci->ifa_prefered;
	} else {
		preferred_lft = INFINITY_LIFE_TIME;
		valid_lft = INFINITY_LIFE_TIME;
3761 3762
	}

3763
	dev =  __dev_get_by_index(net, ifm->ifa_index);
3764 3765 3766
	if (dev == NULL)
		return -ENODEV;

3767 3768
	ifa_flags = tb[IFA_FLAGS] ? nla_get_u32(tb[IFA_FLAGS]) : ifm->ifa_flags;

3769
	/* We ignore other flags so far. */
3770 3771
	ifa_flags &= IFA_F_NODAD | IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
		     IFA_F_NOPREFIXROUTE;
3772

3773
	ifa = ipv6_get_ifaddr(net, pfx, dev, 1);
3774 3775 3776
	if (ifa == NULL) {
		/*
		 * It would be best to check for !NLM_F_CREATE here but
3777
		 * userspace already relies on not having to provide this.
3778
		 */
3779
		return inet6_addr_add(net, ifm->ifa_index, pfx, peer_pfx,
3780 3781
				      ifm->ifa_prefixlen, ifa_flags,
				      preferred_lft, valid_lft);
3782 3783
	}

3784 3785 3786 3787
	if (nlh->nlmsg_flags & NLM_F_EXCL ||
	    !(nlh->nlmsg_flags & NLM_F_REPLACE))
		err = -EEXIST;
	else
3788
		err = inet6_addr_modify(ifa, ifa_flags, preferred_lft, valid_lft);
3789 3790 3791 3792

	in6_ifa_put(ifa);

	return err;
L
Linus Torvalds 已提交
3793 3794
}

3795
static void put_ifaddrmsg(struct nlmsghdr *nlh, u8 prefixlen, u32 flags,
3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807
			  u8 scope, int ifindex)
{
	struct ifaddrmsg *ifm;

	ifm = nlmsg_data(nlh);
	ifm->ifa_family = AF_INET6;
	ifm->ifa_prefixlen = prefixlen;
	ifm->ifa_flags = flags;
	ifm->ifa_scope = scope;
	ifm->ifa_index = ifindex;
}

3808 3809 3810 3811 3812
static int put_cacheinfo(struct sk_buff *skb, unsigned long cstamp,
			 unsigned long tstamp, u32 preferred, u32 valid)
{
	struct ifa_cacheinfo ci;

3813 3814
	ci.cstamp = cstamp_delta(cstamp);
	ci.tstamp = cstamp_delta(tstamp);
3815 3816 3817 3818 3819 3820
	ci.ifa_prefered = preferred;
	ci.ifa_valid = valid;

	return nla_put(skb, IFA_CACHEINFO, sizeof(ci), &ci);
}

3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832
static inline int rt_scope(int ifa_scope)
{
	if (ifa_scope & IFA_HOST)
		return RT_SCOPE_HOST;
	else if (ifa_scope & IFA_LINK)
		return RT_SCOPE_LINK;
	else if (ifa_scope & IFA_SITE)
		return RT_SCOPE_SITE;
	else
		return RT_SCOPE_UNIVERSE;
}

3833 3834
static inline int inet6_ifaddr_msgsize(void)
{
3835
	return NLMSG_ALIGN(sizeof(struct ifaddrmsg))
3836
	       + nla_total_size(16) /* IFA_LOCAL */
3837
	       + nla_total_size(16) /* IFA_ADDRESS */
3838 3839
	       + nla_total_size(sizeof(struct ifa_cacheinfo))
	       + nla_total_size(4)  /* IFA_FLAGS */;
3840
}
3841

L
Linus Torvalds 已提交
3842
static int inet6_fill_ifaddr(struct sk_buff *skb, struct inet6_ifaddr *ifa,
3843
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
3844 3845
{
	struct nlmsghdr  *nlh;
3846
	u32 preferred, valid;
L
Linus Torvalds 已提交
3847

3848
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
3849
	if (nlh == NULL)
3850
		return -EMSGSIZE;
3851

3852 3853 3854
	put_ifaddrmsg(nlh, ifa->prefix_len, ifa->flags, rt_scope(ifa->scope),
		      ifa->idev->dev->ifindex);

3855 3856
	if (!((ifa->flags&IFA_F_PERMANENT) &&
	      (ifa->prefered_lft == INFINITY_LIFE_TIME))) {
3857 3858 3859
		preferred = ifa->prefered_lft;
		valid = ifa->valid_lft;
		if (preferred != INFINITY_LIFE_TIME) {
L
Linus Torvalds 已提交
3860
			long tval = (jiffies - ifa->tstamp)/HZ;
3861 3862 3863 3864
			if (preferred > tval)
				preferred -= tval;
			else
				preferred = 0;
3865 3866 3867 3868 3869 3870
			if (valid != INFINITY_LIFE_TIME) {
				if (valid > tval)
					valid -= tval;
				else
					valid = 0;
			}
L
Linus Torvalds 已提交
3871 3872
		}
	} else {
3873 3874 3875 3876
		preferred = INFINITY_LIFE_TIME;
		valid = INFINITY_LIFE_TIME;
	}

C
Cong Wang 已提交
3877
	if (!ipv6_addr_any(&ifa->peer_addr)) {
3878 3879 3880 3881 3882 3883 3884 3885 3886
		if (nla_put(skb, IFA_LOCAL, 16, &ifa->addr) < 0 ||
		    nla_put(skb, IFA_ADDRESS, 16, &ifa->peer_addr) < 0)
			goto error;
	} else
		if (nla_put(skb, IFA_ADDRESS, 16, &ifa->addr) < 0)
			goto error;

	if (put_cacheinfo(skb, ifa->cstamp, ifa->tstamp, preferred, valid) < 0)
		goto error;
L
Linus Torvalds 已提交
3887

3888 3889 3890
	if (nla_put_u32(skb, IFA_FLAGS, ifa->flags) < 0)
		goto error;

3891
	return nlmsg_end(skb, nlh);
3892 3893 3894 3895

error:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
3896 3897 3898
}

static int inet6_fill_ifmcaddr(struct sk_buff *skb, struct ifmcaddr6 *ifmca,
3899
				u32 portid, u32 seq, int event, u16 flags)
L
Linus Torvalds 已提交
3900 3901
{
	struct nlmsghdr  *nlh;
3902 3903 3904 3905 3906
	u8 scope = RT_SCOPE_UNIVERSE;
	int ifindex = ifmca->idev->dev->ifindex;

	if (ipv6_addr_scope(&ifmca->mca_addr) & IFA_SITE)
		scope = RT_SCOPE_SITE;
L
Linus Torvalds 已提交
3907

3908
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
3909
	if (nlh == NULL)
3910
		return -EMSGSIZE;
3911

3912 3913 3914
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
	if (nla_put(skb, IFA_MULTICAST, 16, &ifmca->mca_addr) < 0 ||
	    put_cacheinfo(skb, ifmca->mca_cstamp, ifmca->mca_tstamp,
3915 3916 3917 3918
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
3919

3920
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
3921 3922 3923
}

static int inet6_fill_ifacaddr(struct sk_buff *skb, struct ifacaddr6 *ifaca,
3924
				u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
3925 3926
{
	struct nlmsghdr  *nlh;
3927 3928 3929 3930 3931
	u8 scope = RT_SCOPE_UNIVERSE;
	int ifindex = ifaca->aca_idev->dev->ifindex;

	if (ipv6_addr_scope(&ifaca->aca_addr) & IFA_SITE)
		scope = RT_SCOPE_SITE;
L
Linus Torvalds 已提交
3932

3933
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
3934
	if (nlh == NULL)
3935
		return -EMSGSIZE;
3936

3937 3938 3939
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
	if (nla_put(skb, IFA_ANYCAST, 16, &ifaca->aca_addr) < 0 ||
	    put_cacheinfo(skb, ifaca->aca_cstamp, ifaca->aca_tstamp,
3940 3941 3942 3943
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
L
Linus Torvalds 已提交
3944

3945
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
3946 3947
}

S
Stephen Hemminger 已提交
3948
enum addr_type_t {
L
Linus Torvalds 已提交
3949 3950 3951 3952 3953
	UNICAST_ADDR,
	MULTICAST_ADDR,
	ANYCAST_ADDR,
};

E
Eric Dumazet 已提交
3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965
/* called with rcu_read_lock() */
static int in6_dump_addrs(struct inet6_dev *idev, struct sk_buff *skb,
			  struct netlink_callback *cb, enum addr_type_t type,
			  int s_ip_idx, int *p_ip_idx)
{
	struct ifmcaddr6 *ifmca;
	struct ifacaddr6 *ifaca;
	int err = 1;
	int ip_idx = *p_ip_idx;

	read_lock_bh(&idev->lock);
	switch (type) {
3966 3967 3968
	case UNICAST_ADDR: {
		struct inet6_ifaddr *ifa;

E
Eric Dumazet 已提交
3969
		/* unicast address incl. temp addr */
3970 3971
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
			if (++ip_idx < s_ip_idx)
E
Eric Dumazet 已提交
3972 3973
				continue;
			err = inet6_fill_ifaddr(skb, ifa,
3974
						NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
3975 3976 3977 3978 3979
						cb->nlh->nlmsg_seq,
						RTM_NEWADDR,
						NLM_F_MULTI);
			if (err <= 0)
				break;
3980
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
E
Eric Dumazet 已提交
3981 3982
		}
		break;
3983
	}
E
Eric Dumazet 已提交
3984 3985 3986 3987 3988 3989 3990
	case MULTICAST_ADDR:
		/* multicast address */
		for (ifmca = idev->mc_list; ifmca;
		     ifmca = ifmca->next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
			err = inet6_fill_ifmcaddr(skb, ifmca,
3991
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005
						  cb->nlh->nlmsg_seq,
						  RTM_GETMULTICAST,
						  NLM_F_MULTI);
			if (err <= 0)
				break;
		}
		break;
	case ANYCAST_ADDR:
		/* anycast address */
		for (ifaca = idev->ac_list; ifaca;
		     ifaca = ifaca->aca_next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
			err = inet6_fill_ifacaddr(skb, ifaca,
4006
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021
						  cb->nlh->nlmsg_seq,
						  RTM_GETANYCAST,
						  NLM_F_MULTI);
			if (err <= 0)
				break;
		}
		break;
	default:
		break;
	}
	read_unlock_bh(&idev->lock);
	*p_ip_idx = ip_idx;
	return err;
}

L
Linus Torvalds 已提交
4022 4023 4024
static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
			   enum addr_type_t type)
{
E
Eric Dumazet 已提交
4025 4026
	struct net *net = sock_net(skb->sk);
	int h, s_h;
L
Linus Torvalds 已提交
4027 4028 4029
	int idx, ip_idx;
	int s_idx, s_ip_idx;
	struct net_device *dev;
E
Eric Dumazet 已提交
4030 4031
	struct inet6_dev *idev;
	struct hlist_head *head;
4032

E
Eric Dumazet 已提交
4033 4034 4035
	s_h = cb->args[0];
	s_idx = idx = cb->args[1];
	s_ip_idx = ip_idx = cb->args[2];
4036

E
Eric Dumazet 已提交
4037
	rcu_read_lock();
4038
	cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^ net->dev_base_seq;
E
Eric Dumazet 已提交
4039 4040 4041
	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
4042
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
4043 4044
			if (idx < s_idx)
				goto cont;
4045
			if (h > s_h || idx > s_idx)
E
Eric Dumazet 已提交
4046 4047
				s_ip_idx = 0;
			ip_idx = 0;
S
Stephen Hemminger 已提交
4048 4049
			idev = __in6_dev_get(dev);
			if (!idev)
E
Eric Dumazet 已提交
4050 4051 4052 4053 4054
				goto cont;

			if (in6_dump_addrs(idev, skb, cb, type,
					   s_ip_idx, &ip_idx) <= 0)
				goto done;
4055
cont:
E
Eric Dumazet 已提交
4056 4057
			idx++;
		}
L
Linus Torvalds 已提交
4058
	}
E
Eric Dumazet 已提交
4059 4060 4061 4062 4063 4064
done:
	rcu_read_unlock();
	cb->args[0] = h;
	cb->args[1] = idx;
	cb->args[2] = ip_idx;

L
Linus Torvalds 已提交
4065 4066 4067 4068 4069 4070
	return skb->len;
}

static int inet6_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = UNICAST_ADDR;
4071

L
Linus Torvalds 已提交
4072 4073 4074 4075 4076 4077
	return inet6_dump_addr(skb, cb, type);
}

static int inet6_dump_ifmcaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = MULTICAST_ADDR;
4078

L
Linus Torvalds 已提交
4079 4080 4081 4082 4083 4084 4085
	return inet6_dump_addr(skb, cb, type);
}


static int inet6_dump_ifacaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = ANYCAST_ADDR;
4086

L
Linus Torvalds 已提交
4087 4088 4089
	return inet6_dump_addr(skb, cb, type);
}

4090
static int inet6_rtm_getaddr(struct sk_buff *in_skb, struct nlmsghdr *nlh)
4091
{
4092
	struct net *net = sock_net(in_skb->sk);
4093 4094
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4095
	struct in6_addr *addr = NULL, *peer;
4096 4097 4098 4099 4100
	struct net_device *dev = NULL;
	struct inet6_ifaddr *ifa;
	struct sk_buff *skb;
	int err;

4101 4102 4103 4104
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		goto errout;

4105
	addr = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer);
4106 4107 4108
	if (addr == NULL) {
		err = -EINVAL;
		goto errout;
4109 4110
	}

4111
	ifm = nlmsg_data(nlh);
4112
	if (ifm->ifa_index)
4113
		dev = __dev_get_by_index(net, ifm->ifa_index);
4114

S
Stephen Hemminger 已提交
4115 4116
	ifa = ipv6_get_ifaddr(net, addr, dev, 1);
	if (!ifa) {
4117 4118 4119
		err = -EADDRNOTAVAIL;
		goto errout;
	}
4120

S
Stephen Hemminger 已提交
4121 4122
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_KERNEL);
	if (!skb) {
4123
		err = -ENOBUFS;
4124
		goto errout_ifa;
4125 4126
	}

4127
	err = inet6_fill_ifaddr(skb, ifa, NETLINK_CB(in_skb).portid,
4128
				nlh->nlmsg_seq, RTM_NEWADDR, 0);
4129 4130 4131 4132 4133 4134
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout_ifa;
	}
4135
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
4136
errout_ifa:
4137
	in6_ifa_put(ifa);
4138
errout:
4139 4140 4141
	return err;
}

L
Linus Torvalds 已提交
4142 4143 4144
static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa)
{
	struct sk_buff *skb;
4145
	struct net *net = dev_net(ifa->idev->dev);
4146
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
4147

4148
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_ATOMIC);
4149 4150 4151 4152
	if (skb == NULL)
		goto errout;

	err = inet6_fill_ifaddr(skb, ifa, 0, 0, event, 0);
4153 4154 4155 4156 4157 4158
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4159 4160
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFADDR, NULL, GFP_ATOMIC);
	return;
4161 4162
errout:
	if (err < 0)
4163
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFADDR, err);
L
Linus Torvalds 已提交
4164 4165
}

D
Dave Jones 已提交
4166
static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
L
Linus Torvalds 已提交
4167 4168
				__s32 *array, int bytes)
{
4169 4170
	BUG_ON(bytes < (DEVCONF_MAX * 4));

L
Linus Torvalds 已提交
4171 4172 4173 4174 4175 4176 4177 4178 4179
	memset(array, 0, bytes);
	array[DEVCONF_FORWARDING] = cnf->forwarding;
	array[DEVCONF_HOPLIMIT] = cnf->hop_limit;
	array[DEVCONF_MTU6] = cnf->mtu6;
	array[DEVCONF_ACCEPT_RA] = cnf->accept_ra;
	array[DEVCONF_ACCEPT_REDIRECTS] = cnf->accept_redirects;
	array[DEVCONF_AUTOCONF] = cnf->autoconf;
	array[DEVCONF_DAD_TRANSMITS] = cnf->dad_transmits;
	array[DEVCONF_RTR_SOLICITS] = cnf->rtr_solicits;
4180 4181 4182 4183
	array[DEVCONF_RTR_SOLICIT_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_interval);
	array[DEVCONF_RTR_SOLICIT_DELAY] =
		jiffies_to_msecs(cnf->rtr_solicit_delay);
L
Linus Torvalds 已提交
4184
	array[DEVCONF_FORCE_MLD_VERSION] = cnf->force_mld_version;
4185 4186 4187 4188
	array[DEVCONF_MLDV1_UNSOLICITED_REPORT_INTERVAL] =
		jiffies_to_msecs(cnf->mldv1_unsolicited_report_interval);
	array[DEVCONF_MLDV2_UNSOLICITED_REPORT_INTERVAL] =
		jiffies_to_msecs(cnf->mldv2_unsolicited_report_interval);
L
Linus Torvalds 已提交
4189 4190 4191 4192 4193 4194
	array[DEVCONF_USE_TEMPADDR] = cnf->use_tempaddr;
	array[DEVCONF_TEMP_VALID_LFT] = cnf->temp_valid_lft;
	array[DEVCONF_TEMP_PREFERED_LFT] = cnf->temp_prefered_lft;
	array[DEVCONF_REGEN_MAX_RETRY] = cnf->regen_max_retry;
	array[DEVCONF_MAX_DESYNC_FACTOR] = cnf->max_desync_factor;
	array[DEVCONF_MAX_ADDRESSES] = cnf->max_addresses;
4195
	array[DEVCONF_ACCEPT_RA_DEFRTR] = cnf->accept_ra_defrtr;
4196
	array[DEVCONF_ACCEPT_RA_PINFO] = cnf->accept_ra_pinfo;
4197 4198
#ifdef CONFIG_IPV6_ROUTER_PREF
	array[DEVCONF_ACCEPT_RA_RTR_PREF] = cnf->accept_ra_rtr_pref;
4199 4200
	array[DEVCONF_RTR_PROBE_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_probe_interval);
N
Neil Horman 已提交
4201
#ifdef CONFIG_IPV6_ROUTE_INFO
4202 4203
	array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
#endif
4204
#endif
4205
	array[DEVCONF_PROXY_NDP] = cnf->proxy_ndp;
4206
	array[DEVCONF_ACCEPT_SOURCE_ROUTE] = cnf->accept_source_route;
4207 4208 4209
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
#endif
4210 4211 4212
#ifdef CONFIG_IPV6_MROUTE
	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
#endif
4213
	array[DEVCONF_DISABLE_IPV6] = cnf->disable_ipv6;
4214
	array[DEVCONF_ACCEPT_DAD] = cnf->accept_dad;
4215
	array[DEVCONF_FORCE_TLLAO] = cnf->force_tllao;
4216
	array[DEVCONF_NDISC_NOTIFY] = cnf->ndisc_notify;
4217
	array[DEVCONF_SUPPRESS_FRAG_NDISC] = cnf->suppress_frag_ndisc;
L
Linus Torvalds 已提交
4218 4219
}

T
Thomas Graf 已提交
4220 4221 4222 4223 4224 4225
static inline size_t inet6_ifla6_size(void)
{
	return nla_total_size(4) /* IFLA_INET6_FLAGS */
	     + nla_total_size(sizeof(struct ifla_cacheinfo))
	     + nla_total_size(DEVCONF_MAX * 4) /* IFLA_INET6_CONF */
	     + nla_total_size(IPSTATS_MIB_MAX * 8) /* IFLA_INET6_STATS */
4226 4227
	     + nla_total_size(ICMP6_MIB_MAX * 8) /* IFLA_INET6_ICMP6STATS */
	     + nla_total_size(sizeof(struct in6_addr)); /* IFLA_INET6_TOKEN */
T
Thomas Graf 已提交
4228 4229
}

4230 4231 4232 4233 4234 4235 4236
static inline size_t inet6_if_nlmsg_size(void)
{
	return NLMSG_ALIGN(sizeof(struct ifinfomsg))
	       + nla_total_size(IFNAMSIZ) /* IFLA_IFNAME */
	       + nla_total_size(MAX_ADDR_LEN) /* IFLA_ADDRESS */
	       + nla_total_size(4) /* IFLA_MTU */
	       + nla_total_size(4) /* IFLA_LINK */
T
Thomas Graf 已提交
4237
	       + nla_total_size(inet6_ifla6_size()); /* IFLA_PROTINFO */
4238
}
4239

4240
static inline void __snmp6_fill_statsdev(u64 *stats, atomic_long_t *mib,
4241
				      int items, int bytes)
4242 4243 4244 4245 4246 4247 4248 4249
{
	int i;
	int pad = bytes - sizeof(u64) * items;
	BUG_ON(pad < 0);

	/* Use put_unaligned() because stats may not be aligned for u64. */
	put_unaligned(items, &stats[0]);
	for (i = 1; i < items; i++)
4250
		put_unaligned(atomic_long_read(&mib[i]), &stats[i]);
4251 4252 4253 4254

	memset(&stats[items], 0, pad);
}

4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269
static inline void __snmp6_fill_stats64(u64 *stats, void __percpu **mib,
				      int items, int bytes, size_t syncpoff)
{
	int i;
	int pad = bytes - sizeof(u64) * items;
	BUG_ON(pad < 0);

	/* Use put_unaligned() because stats may not be aligned for u64. */
	put_unaligned(items, &stats[0]);
	for (i = 1; i < items; i++)
		put_unaligned(snmp_fold_field64(mib, i, syncpoff), &stats[i]);

	memset(&stats[items], 0, pad);
}

4270 4271 4272
static void snmp6_fill_stats(u64 *stats, struct inet6_dev *idev, int attrtype,
			     int bytes)
{
S
Stephen Hemminger 已提交
4273
	switch (attrtype) {
4274
	case IFLA_INET6_STATS:
4275 4276
		__snmp6_fill_stats64(stats, (void __percpu **)idev->stats.ipv6,
				     IPSTATS_MIB_MAX, bytes, offsetof(struct ipstats_mib, syncp));
4277 4278
		break;
	case IFLA_INET6_ICMP6STATS:
4279
		__snmp6_fill_statsdev(stats, idev->stats.icmpv6dev->mibs, ICMP6_MIB_MAX, bytes);
4280 4281 4282 4283
		break;
	}
}

T
Thomas Graf 已提交
4284 4285 4286 4287 4288
static int inet6_fill_ifla6_attrs(struct sk_buff *skb, struct inet6_dev *idev)
{
	struct nlattr *nla;
	struct ifla_cacheinfo ci;

D
David S. Miller 已提交
4289 4290
	if (nla_put_u32(skb, IFLA_INET6_FLAGS, idev->if_flags))
		goto nla_put_failure;
T
Thomas Graf 已提交
4291
	ci.max_reasm_len = IPV6_MAXPLEN;
4292 4293
	ci.tstamp = cstamp_delta(idev->tstamp);
	ci.reachable_time = jiffies_to_msecs(idev->nd_parms->reachable_time);
J
Jiri Pirko 已提交
4294
	ci.retrans_time = jiffies_to_msecs(NEIGH_VAR(idev->nd_parms, RETRANS_TIME));
D
David S. Miller 已提交
4295 4296
	if (nla_put(skb, IFLA_INET6_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
T
Thomas Graf 已提交
4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313
	nla = nla_reserve(skb, IFLA_INET6_CONF, DEVCONF_MAX * sizeof(s32));
	if (nla == NULL)
		goto nla_put_failure;
	ipv6_store_devconf(&idev->cnf, nla_data(nla), nla_len(nla));

	/* XXX - MC not implemented */

	nla = nla_reserve(skb, IFLA_INET6_STATS, IPSTATS_MIB_MAX * sizeof(u64));
	if (nla == NULL)
		goto nla_put_failure;
	snmp6_fill_stats(nla_data(nla), idev, IFLA_INET6_STATS, nla_len(nla));

	nla = nla_reserve(skb, IFLA_INET6_ICMP6STATS, ICMP6_MIB_MAX * sizeof(u64));
	if (nla == NULL)
		goto nla_put_failure;
	snmp6_fill_stats(nla_data(nla), idev, IFLA_INET6_ICMP6STATS, nla_len(nla));

4314 4315 4316 4317 4318 4319 4320
	nla = nla_reserve(skb, IFLA_INET6_TOKEN, sizeof(struct in6_addr));
	if (nla == NULL)
		goto nla_put_failure;
	read_lock_bh(&idev->lock);
	memcpy(nla_data(nla), idev->token.s6_addr, nla_len(nla));
	read_unlock_bh(&idev->lock);

T
Thomas Graf 已提交
4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347
	return 0;

nla_put_failure:
	return -EMSGSIZE;
}

static size_t inet6_get_link_af_size(const struct net_device *dev)
{
	if (!__in6_dev_get(dev))
		return 0;

	return inet6_ifla6_size();
}

static int inet6_fill_link_af(struct sk_buff *skb, const struct net_device *dev)
{
	struct inet6_dev *idev = __in6_dev_get(dev);

	if (!idev)
		return -ENODATA;

	if (inet6_fill_ifla6_attrs(skb, idev) < 0)
		return -EMSGSIZE;

	return 0;
}

4348 4349 4350 4351
static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token)
{
	struct inet6_ifaddr *ifp;
	struct net_device *dev = idev->dev;
4352
	bool update_rs = false;
4353
	struct in6_addr ll_addr;
4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372

	if (token == NULL)
		return -EINVAL;
	if (ipv6_addr_any(token))
		return -EINVAL;
	if (dev->flags & (IFF_LOOPBACK | IFF_NOARP))
		return -EINVAL;
	if (!ipv6_accept_ra(idev))
		return -EINVAL;
	if (idev->cnf.rtr_solicits <= 0)
		return -EINVAL;

	write_lock_bh(&idev->lock);

	BUILD_BUG_ON(sizeof(token->s6_addr) != 16);
	memcpy(idev->token.s6_addr + 8, token->s6_addr + 8, 8);

	write_unlock_bh(&idev->lock);

4373 4374 4375
	if (!idev->dead && (idev->if_flags & IF_READY) &&
	    !ipv6_get_lladdr(dev, &ll_addr, IFA_F_TENTATIVE |
			     IFA_F_OPTIMISTIC)) {
4376 4377 4378 4379 4380 4381 4382

		/* If we're not ready, then normal ifup will take care
		 * of this. Otherwise, we need to request our rs here.
		 */
		ndisc_send_rs(dev, &ll_addr, &in6addr_linklocal_allrouters);
		update_rs = true;
	}
4383 4384

	write_lock_bh(&idev->lock);
4385

4386
	if (update_rs) {
4387
		idev->if_flags |= IF_RS_SENT;
4388 4389 4390
		idev->rs_probes = 1;
		addrconf_mod_rs_timer(idev, idev->cnf.rtr_solicit_interval);
	}
4391 4392 4393 4394

	/* Well, that's kinda nasty ... */
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
		spin_lock(&ifp->lock);
4395
		if (ifp->tokenized) {
4396 4397 4398 4399 4400 4401 4402
			ifp->valid_lft = 0;
			ifp->prefered_lft = 0;
		}
		spin_unlock(&ifp->lock);
	}

	write_unlock_bh(&idev->lock);
4403
	addrconf_verify(0);
4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424
	return 0;
}

static int inet6_set_link_af(struct net_device *dev, const struct nlattr *nla)
{
	int err = -EINVAL;
	struct inet6_dev *idev = __in6_dev_get(dev);
	struct nlattr *tb[IFLA_INET6_MAX + 1];

	if (!idev)
		return -EAFNOSUPPORT;

	if (nla_parse_nested(tb, IFLA_INET6_MAX, nla, NULL) < 0)
		BUG();

	if (tb[IFLA_INET6_TOKEN])
		err = inet6_set_iftoken(idev, nla_data(tb[IFLA_INET6_TOKEN]));

	return err;
}

4425
static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
4426
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4427
{
4428 4429 4430 4431 4432
	struct net_device *dev = idev->dev;
	struct ifinfomsg *hdr;
	struct nlmsghdr *nlh;
	void *protoinfo;

4433
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*hdr), flags);
4434
	if (nlh == NULL)
4435
		return -EMSGSIZE;
4436 4437 4438 4439 4440 4441 4442 4443 4444

	hdr = nlmsg_data(nlh);
	hdr->ifi_family = AF_INET6;
	hdr->__ifi_pad = 0;
	hdr->ifi_type = dev->type;
	hdr->ifi_index = dev->ifindex;
	hdr->ifi_flags = dev_get_flags(dev);
	hdr->ifi_change = 0;

D
David S. Miller 已提交
4445 4446 4447 4448 4449 4450 4451
	if (nla_put_string(skb, IFLA_IFNAME, dev->name) ||
	    (dev->addr_len &&
	     nla_put(skb, IFLA_ADDRESS, dev->addr_len, dev->dev_addr)) ||
	    nla_put_u32(skb, IFLA_MTU, dev->mtu) ||
	    (dev->ifindex != dev->iflink &&
	     nla_put_u32(skb, IFLA_LINK, dev->iflink)))
		goto nla_put_failure;
4452 4453 4454
	protoinfo = nla_nest_start(skb, IFLA_PROTINFO);
	if (protoinfo == NULL)
		goto nla_put_failure;
L
Linus Torvalds 已提交
4455

T
Thomas Graf 已提交
4456
	if (inet6_fill_ifla6_attrs(skb, idev) < 0)
4457
		goto nla_put_failure;
L
Linus Torvalds 已提交
4458

4459 4460
	nla_nest_end(skb, protoinfo);
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
4461

4462
nla_put_failure:
4463 4464
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4465 4466 4467 4468
}

static int inet6_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
{
4469
	struct net *net = sock_net(skb->sk);
E
Eric Dumazet 已提交
4470
	int h, s_h;
4471
	int idx = 0, s_idx;
L
Linus Torvalds 已提交
4472 4473
	struct net_device *dev;
	struct inet6_dev *idev;
E
Eric Dumazet 已提交
4474
	struct hlist_head *head;
L
Linus Torvalds 已提交
4475

E
Eric Dumazet 已提交
4476 4477 4478 4479 4480 4481 4482
	s_h = cb->args[0];
	s_idx = cb->args[1];

	rcu_read_lock();
	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
4483
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
4484 4485 4486 4487 4488 4489
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;
			if (inet6_fill_ifinfo(skb, idev,
4490
					      NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4491 4492 4493
					      cb->nlh->nlmsg_seq,
					      RTM_NEWLINK, NLM_F_MULTI) <= 0)
				goto out;
4494
cont:
E
Eric Dumazet 已提交
4495 4496
			idx++;
		}
L
Linus Torvalds 已提交
4497
	}
E
Eric Dumazet 已提交
4498 4499 4500 4501
out:
	rcu_read_unlock();
	cb->args[1] = idx;
	cb->args[0] = h;
L
Linus Torvalds 已提交
4502 4503 4504 4505 4506 4507 4508

	return skb->len;
}

void inet6_ifinfo_notify(int event, struct inet6_dev *idev)
{
	struct sk_buff *skb;
4509
	struct net *net = dev_net(idev->dev);
4510
	int err = -ENOBUFS;
4511

4512
	skb = nlmsg_new(inet6_if_nlmsg_size(), GFP_ATOMIC);
4513 4514 4515 4516
	if (skb == NULL)
		goto errout;

	err = inet6_fill_ifinfo(skb, idev, 0, 0, event, 0);
4517 4518 4519 4520 4521 4522
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_if_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4523
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFINFO, NULL, GFP_ATOMIC);
4524
	return;
4525 4526
errout:
	if (err < 0)
4527
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFINFO, err);
L
Linus Torvalds 已提交
4528 4529
}

4530 4531 4532 4533 4534 4535
static inline size_t inet6_prefix_nlmsg_size(void)
{
	return NLMSG_ALIGN(sizeof(struct prefixmsg))
	       + nla_total_size(sizeof(struct in6_addr))
	       + nla_total_size(sizeof(struct prefix_cacheinfo));
}
4536

L
Linus Torvalds 已提交
4537
static int inet6_fill_prefix(struct sk_buff *skb, struct inet6_dev *idev,
4538
			     struct prefix_info *pinfo, u32 portid, u32 seq,
4539
			     int event, unsigned int flags)
L
Linus Torvalds 已提交
4540
{
4541 4542
	struct prefixmsg *pmsg;
	struct nlmsghdr *nlh;
L
Linus Torvalds 已提交
4543 4544
	struct prefix_cacheinfo	ci;

4545
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*pmsg), flags);
4546
	if (nlh == NULL)
4547
		return -EMSGSIZE;
4548 4549

	pmsg = nlmsg_data(nlh);
L
Linus Torvalds 已提交
4550
	pmsg->prefix_family = AF_INET6;
4551 4552
	pmsg->prefix_pad1 = 0;
	pmsg->prefix_pad2 = 0;
L
Linus Torvalds 已提交
4553 4554 4555
	pmsg->prefix_ifindex = idev->dev->ifindex;
	pmsg->prefix_len = pinfo->prefix_len;
	pmsg->prefix_type = pinfo->type;
4556
	pmsg->prefix_pad3 = 0;
L
Linus Torvalds 已提交
4557 4558 4559 4560 4561 4562
	pmsg->prefix_flags = 0;
	if (pinfo->onlink)
		pmsg->prefix_flags |= IF_PREFIX_ONLINK;
	if (pinfo->autoconf)
		pmsg->prefix_flags |= IF_PREFIX_AUTOCONF;

D
David S. Miller 已提交
4563 4564
	if (nla_put(skb, PREFIX_ADDRESS, sizeof(pinfo->prefix), &pinfo->prefix))
		goto nla_put_failure;
L
Linus Torvalds 已提交
4565 4566
	ci.preferred_time = ntohl(pinfo->prefered);
	ci.valid_time = ntohl(pinfo->valid);
D
David S. Miller 已提交
4567 4568
	if (nla_put(skb, PREFIX_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
4569
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
4570

4571
nla_put_failure:
4572 4573
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4574 4575
}

4576
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
4577 4578 4579
			 struct prefix_info *pinfo)
{
	struct sk_buff *skb;
4580
	struct net *net = dev_net(idev->dev);
4581
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
4582

4583
	skb = nlmsg_new(inet6_prefix_nlmsg_size(), GFP_ATOMIC);
4584 4585 4586 4587
	if (skb == NULL)
		goto errout;

	err = inet6_fill_prefix(skb, idev, pinfo, 0, 0, event, 0);
4588 4589 4590 4591 4592 4593
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_prefix_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4594 4595
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_PREFIX, NULL, GFP_ATOMIC);
	return;
4596 4597
errout:
	if (err < 0)
4598
		rtnl_set_sk_err(net, RTNLGRP_IPV6_PREFIX, err);
L
Linus Torvalds 已提交
4599 4600
}

4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613
static void update_valid_ll_addr_cnt(struct inet6_ifaddr *ifp, int count)
{
	write_lock_bh(&ifp->idev->lock);
	spin_lock(&ifp->lock);
	if (((ifp->flags & (IFA_F_PERMANENT|IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|
			    IFA_F_DADFAILED)) == IFA_F_PERMANENT) &&
	    (ipv6_addr_type(&ifp->addr) & IPV6_ADDR_LINKLOCAL))
		ifp->idev->valid_ll_addr_cnt += count;
	WARN_ON(ifp->idev->valid_ll_addr_cnt < 0);
	spin_unlock(&ifp->lock);
	write_unlock_bh(&ifp->idev->lock);
}

L
Linus Torvalds 已提交
4614 4615
static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
4616 4617
	struct net *net = dev_net(ifp->idev->dev);

L
Linus Torvalds 已提交
4618 4619 4620 4621
	inet6_ifa_notify(event ? : RTM_NEWADDR, ifp);

	switch (event) {
	case RTM_NEWADDR:
4622 4623
		update_valid_ll_addr_cnt(ifp, 1);

4624 4625 4626 4627 4628 4629 4630 4631
		/*
		 * If the address was optimistic
		 * we inserted the route at the start of
		 * our DAD process, so we don't need
		 * to do it again
		 */
		if (!(ifp->rt->rt6i_node))
			ip6_ins_rt(ifp->rt);
L
Linus Torvalds 已提交
4632 4633
		if (ifp->idev->cnf.forwarding)
			addrconf_join_anycast(ifp);
C
Cong Wang 已提交
4634
		if (!ipv6_addr_any(&ifp->peer_addr))
4635 4636
			addrconf_prefix_route(&ifp->peer_addr, 128,
					      ifp->idev->dev, 0, 0);
L
Linus Torvalds 已提交
4637 4638
		break;
	case RTM_DELADDR:
4639 4640
		update_valid_ll_addr_cnt(ifp, -1);

L
Linus Torvalds 已提交
4641 4642 4643
		if (ifp->idev->cnf.forwarding)
			addrconf_leave_anycast(ifp);
		addrconf_leave_solict(ifp->idev, &ifp->addr);
C
Cong Wang 已提交
4644
		if (!ipv6_addr_any(&ifp->peer_addr)) {
4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655
			struct rt6_info *rt;
			struct net_device *dev = ifp->idev->dev;

			rt = rt6_lookup(dev_net(dev), &ifp->peer_addr, NULL,
					dev->ifindex, 1);
			if (rt) {
				dst_hold(&rt->dst);
				if (ip6_del_rt(rt))
					dst_free(&rt->dst);
			}
		}
4656
		dst_hold(&ifp->rt->dst);
4657

4658
		if (ip6_del_rt(ifp->rt))
4659
			dst_free(&ifp->rt->dst);
L
Linus Torvalds 已提交
4660 4661
		break;
	}
4662
	atomic_inc(&net->ipv6.dev_addr_genid);
4663
	rt_genid_bump_ipv6(net);
L
Linus Torvalds 已提交
4664 4665 4666 4667
}

static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
4668
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
4669 4670
	if (likely(ifp->idev->dead == 0))
		__ipv6_ifa_notify(event, ifp);
4671
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
4672 4673 4674 4675 4676
}

#ifdef CONFIG_SYSCTL

static
4677
int addrconf_sysctl_forward(struct ctl_table *ctl, int write,
L
Linus Torvalds 已提交
4678 4679 4680 4681
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
4682
	loff_t pos = *ppos;
4683
	struct ctl_table lctl;
L
Linus Torvalds 已提交
4684 4685
	int ret;

4686 4687 4688 4689 4690 4691 4692 4693
	/*
	 * ctl->data points to idev->cnf.forwarding, we should
	 * not modify it until we get the rtnl lock.
	 */
	lctl = *ctl;
	lctl.data = &val;

	ret = proc_dointvec(&lctl, write, buffer, lenp, ppos);
L
Linus Torvalds 已提交
4694

4695
	if (write)
4696
		ret = addrconf_fixup_forwarding(ctl, valp, val);
E
Eric W. Biederman 已提交
4697 4698
	if (ret)
		*ppos = pos;
4699
	return ret;
L
Linus Torvalds 已提交
4700 4701
}

4702 4703
static void dev_disable_change(struct inet6_dev *idev)
{
4704 4705
	struct netdev_notifier_info info;

4706 4707 4708
	if (!idev || !idev->dev)
		return;

4709
	netdev_notifier_info_init(&info, idev->dev);
4710
	if (idev->cnf.disable_ipv6)
4711
		addrconf_notify(NULL, NETDEV_DOWN, &info);
4712
	else
4713
		addrconf_notify(NULL, NETDEV_UP, &info);
4714 4715 4716 4717 4718 4719 4720
}

static void addrconf_disable_change(struct net *net, __s32 newf)
{
	struct net_device *dev;
	struct inet6_dev *idev;

4721 4722
	rcu_read_lock();
	for_each_netdev_rcu(net, dev) {
4723 4724 4725 4726 4727 4728 4729 4730
		idev = __in6_dev_get(dev);
		if (idev) {
			int changed = (!idev->cnf.disable_ipv6) ^ (!newf);
			idev->cnf.disable_ipv6 = newf;
			if (changed)
				dev_disable_change(idev);
		}
	}
4731
	rcu_read_unlock();
4732 4733
}

4734
static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
4735 4736
{
	struct net *net;
4737 4738 4739 4740
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
4741 4742

	net = (struct net *)table->extra2;
4743 4744
	old = *p;
	*p = newf;
4745

4746 4747
	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
		rtnl_unlock();
4748
		return 0;
E
Eric W. Biederman 已提交
4749
	}
4750 4751 4752 4753

	if (p == &net->ipv6.devconf_all->disable_ipv6) {
		net->ipv6.devconf_dflt->disable_ipv6 = newf;
		addrconf_disable_change(net, newf);
4754
	} else if ((!newf) ^ (!old))
4755 4756 4757 4758 4759 4760 4761
		dev_disable_change((struct inet6_dev *)table->extra1);

	rtnl_unlock();
	return 0;
}

static
4762
int addrconf_sysctl_disable(struct ctl_table *ctl, int write,
4763 4764 4765 4766
			    void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
4767
	loff_t pos = *ppos;
4768
	struct ctl_table lctl;
4769 4770
	int ret;

4771 4772 4773 4774 4775 4776 4777 4778
	/*
	 * ctl->data points to idev->cnf.disable_ipv6, we should
	 * not modify it until we get the rtnl lock.
	 */
	lctl = *ctl;
	lctl.data = &val;

	ret = proc_dointvec(&lctl, write, buffer, lenp, ppos);
4779 4780 4781

	if (write)
		ret = addrconf_disable_ipv6(ctl, valp, val);
E
Eric W. Biederman 已提交
4782 4783
	if (ret)
		*ppos = pos;
4784 4785 4786
	return ret;
}

4787 4788 4789 4790 4791 4792 4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826
static
int addrconf_sysctl_proxy_ndp(struct ctl_table *ctl, int write,
			      void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int ret;
	int old, new;

	old = *valp;
	ret = proc_dointvec(ctl, write, buffer, lenp, ppos);
	new = *valp;

	if (write && old != new) {
		struct net *net = ctl->extra2;

		if (!rtnl_trylock())
			return restart_syscall();

		if (valp == &net->ipv6.devconf_dflt->proxy_ndp)
			inet6_netconf_notify_devconf(net, NETCONFA_PROXY_NEIGH,
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
		else if (valp == &net->ipv6.devconf_all->proxy_ndp)
			inet6_netconf_notify_devconf(net, NETCONFA_PROXY_NEIGH,
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
		else {
			struct inet6_dev *idev = ctl->extra1;

			inet6_netconf_notify_devconf(net, NETCONFA_PROXY_NEIGH,
						     idev->dev->ifindex,
						     &idev->cnf);
		}
		rtnl_unlock();
	}

	return ret;
}


L
Linus Torvalds 已提交
4827 4828 4829
static struct addrconf_sysctl_table
{
	struct ctl_table_header *sysctl_header;
4830
	struct ctl_table addrconf_vars[DEVCONF_MAX+1];
4831
} addrconf_sysctl __read_mostly = {
L
Linus Torvalds 已提交
4832 4833
	.sysctl_header = NULL,
	.addrconf_vars = {
4834
		{
S
Stephen Hemminger 已提交
4835 4836 4837 4838 4839
			.procname	= "forwarding",
			.data		= &ipv6_devconf.forwarding,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= addrconf_sysctl_forward,
L
Linus Torvalds 已提交
4840 4841
		},
		{
S
Stephen Hemminger 已提交
4842 4843 4844 4845 4846
			.procname	= "hop_limit",
			.data		= &ipv6_devconf.hop_limit,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4847 4848
		},
		{
S
Stephen Hemminger 已提交
4849 4850 4851 4852 4853
			.procname	= "mtu",
			.data		= &ipv6_devconf.mtu6,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4854 4855
		},
		{
S
Stephen Hemminger 已提交
4856 4857 4858 4859 4860
			.procname	= "accept_ra",
			.data		= &ipv6_devconf.accept_ra,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4861 4862
		},
		{
S
Stephen Hemminger 已提交
4863 4864 4865 4866 4867
			.procname	= "accept_redirects",
			.data		= &ipv6_devconf.accept_redirects,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4868 4869
		},
		{
S
Stephen Hemminger 已提交
4870 4871 4872 4873 4874
			.procname	= "autoconf",
			.data		= &ipv6_devconf.autoconf,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4875 4876
		},
		{
S
Stephen Hemminger 已提交
4877 4878 4879 4880 4881
			.procname	= "dad_transmits",
			.data		= &ipv6_devconf.dad_transmits,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4882 4883
		},
		{
S
Stephen Hemminger 已提交
4884 4885 4886 4887 4888
			.procname	= "router_solicitations",
			.data		= &ipv6_devconf.rtr_solicits,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4889 4890
		},
		{
S
Stephen Hemminger 已提交
4891 4892 4893 4894 4895
			.procname	= "router_solicitation_interval",
			.data		= &ipv6_devconf.rtr_solicit_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
4896 4897
		},
		{
S
Stephen Hemminger 已提交
4898 4899 4900 4901 4902
			.procname	= "router_solicitation_delay",
			.data		= &ipv6_devconf.rtr_solicit_delay,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
4903 4904
		},
		{
S
Stephen Hemminger 已提交
4905 4906 4907 4908 4909
			.procname	= "force_mld_version",
			.data		= &ipv6_devconf.force_mld_version,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4910
		},
4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926
		{
			.procname	= "mldv1_unsolicited_report_interval",
			.data		=
				&ipv6_devconf.mldv1_unsolicited_report_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_ms_jiffies,
		},
		{
			.procname	= "mldv2_unsolicited_report_interval",
			.data		=
				&ipv6_devconf.mldv2_unsolicited_report_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_ms_jiffies,
		},
L
Linus Torvalds 已提交
4927
		{
S
Stephen Hemminger 已提交
4928 4929 4930 4931 4932
			.procname	= "use_tempaddr",
			.data		= &ipv6_devconf.use_tempaddr,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4933 4934
		},
		{
S
Stephen Hemminger 已提交
4935 4936 4937 4938 4939
			.procname	= "temp_valid_lft",
			.data		= &ipv6_devconf.temp_valid_lft,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4940 4941
		},
		{
S
Stephen Hemminger 已提交
4942 4943 4944 4945 4946
			.procname	= "temp_prefered_lft",
			.data		= &ipv6_devconf.temp_prefered_lft,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4947 4948
		},
		{
S
Stephen Hemminger 已提交
4949 4950 4951 4952 4953
			.procname	= "regen_max_retry",
			.data		= &ipv6_devconf.regen_max_retry,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4954 4955
		},
		{
S
Stephen Hemminger 已提交
4956 4957 4958 4959 4960
			.procname	= "max_desync_factor",
			.data		= &ipv6_devconf.max_desync_factor,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4961 4962
		},
		{
S
Stephen Hemminger 已提交
4963 4964 4965 4966 4967
			.procname	= "max_addresses",
			.data		= &ipv6_devconf.max_addresses,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
4968
		},
4969
		{
S
Stephen Hemminger 已提交
4970 4971 4972 4973 4974
			.procname	= "accept_ra_defrtr",
			.data		= &ipv6_devconf.accept_ra_defrtr,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
4975
		},
4976
		{
S
Stephen Hemminger 已提交
4977 4978 4979 4980 4981
			.procname	= "accept_ra_pinfo",
			.data		= &ipv6_devconf.accept_ra_pinfo,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
4982
		},
4983 4984
#ifdef CONFIG_IPV6_ROUTER_PREF
		{
S
Stephen Hemminger 已提交
4985 4986 4987 4988 4989
			.procname	= "accept_ra_rtr_pref",
			.data		= &ipv6_devconf.accept_ra_rtr_pref,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
4990
		},
4991
		{
S
Stephen Hemminger 已提交
4992 4993 4994 4995 4996
			.procname	= "router_probe_interval",
			.data		= &ipv6_devconf.rtr_probe_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_jiffies,
4997
		},
N
Neil Horman 已提交
4998
#ifdef CONFIG_IPV6_ROUTE_INFO
4999
		{
S
Stephen Hemminger 已提交
5000 5001 5002 5003 5004
			.procname	= "accept_ra_rt_info_max_plen",
			.data		= &ipv6_devconf.accept_ra_rt_info_max_plen,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5005 5006
		},
#endif
5007
#endif
5008
		{
S
Stephen Hemminger 已提交
5009 5010 5011 5012
			.procname	= "proxy_ndp",
			.data		= &ipv6_devconf.proxy_ndp,
			.maxlen		= sizeof(int),
			.mode		= 0644,
5013
			.proc_handler	= addrconf_sysctl_proxy_ndp,
5014
		},
5015
		{
S
Stephen Hemminger 已提交
5016 5017 5018 5019 5020
			.procname	= "accept_source_route",
			.data		= &ipv6_devconf.accept_source_route,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5021
		},
5022 5023
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
		{
S
Stephen Hemminger 已提交
5024 5025 5026 5027 5028
			.procname       = "optimistic_dad",
			.data           = &ipv6_devconf.optimistic_dad,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec,
5029 5030

		},
5031 5032 5033
#endif
#ifdef CONFIG_IPV6_MROUTE
		{
S
Stephen Hemminger 已提交
5034 5035 5036 5037 5038
			.procname	= "mc_forwarding",
			.data		= &ipv6_devconf.mc_forwarding,
			.maxlen		= sizeof(int),
			.mode		= 0444,
			.proc_handler	= proc_dointvec,
5039
		},
5040
#endif
5041
		{
S
Stephen Hemminger 已提交
5042 5043 5044 5045 5046
			.procname	= "disable_ipv6",
			.data		= &ipv6_devconf.disable_ipv6,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= addrconf_sysctl_disable,
5047
		},
5048
		{
S
Stephen Hemminger 已提交
5049 5050 5051 5052 5053
			.procname	= "accept_dad",
			.data		= &ipv6_devconf.accept_dad,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5054
		},
5055 5056 5057 5058 5059 5060 5061
		{
			.procname       = "force_tllao",
			.data           = &ipv6_devconf.force_tllao,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec
		},
5062 5063 5064 5065 5066 5067 5068
		{
			.procname       = "ndisc_notify",
			.data           = &ipv6_devconf.ndisc_notify,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec
		},
5069 5070 5071 5072 5073 5074 5075
		{
			.procname	= "suppress_frag_ndisc",
			.data		= &ipv6_devconf.suppress_frag_ndisc,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec
		},
L
Linus Torvalds 已提交
5076
		{
5077
			/* sentinel */
L
Linus Torvalds 已提交
5078 5079 5080 5081
		}
	},
};

5082
static int __addrconf_sysctl_register(struct net *net, char *dev_name,
5083
		struct inet6_dev *idev, struct ipv6_devconf *p)
L
Linus Torvalds 已提交
5084 5085 5086
{
	int i;
	struct addrconf_sysctl_table *t;
5087
	char path[sizeof("net/ipv6/conf/") + IFNAMSIZ];
5088

A
Arnaldo Carvalho de Melo 已提交
5089
	t = kmemdup(&addrconf_sysctl, sizeof(*t), GFP_KERNEL);
L
Linus Torvalds 已提交
5090
	if (t == NULL)
5091 5092
		goto out;

S
Stephen Hemminger 已提交
5093
	for (i = 0; t->addrconf_vars[i].data; i++) {
S
Stephen Hemminger 已提交
5094
		t->addrconf_vars[i].data += (char *)p - (char *)&ipv6_devconf;
L
Linus Torvalds 已提交
5095
		t->addrconf_vars[i].extra1 = idev; /* embedded; no ref */
5096
		t->addrconf_vars[i].extra2 = net;
L
Linus Torvalds 已提交
5097 5098
	}

5099
	snprintf(path, sizeof(path), "net/ipv6/conf/%s", dev_name);
L
Linus Torvalds 已提交
5100

5101
	t->sysctl_header = register_net_sysctl(net, path, t->addrconf_vars);
L
Linus Torvalds 已提交
5102
	if (t->sysctl_header == NULL)
5103
		goto free;
5104 5105

	p->sysctl = t;
5106
	return 0;
L
Linus Torvalds 已提交
5107

5108
free:
L
Linus Torvalds 已提交
5109
	kfree(t);
5110
out:
5111
	return -ENOBUFS;
L
Linus Torvalds 已提交
5112 5113
}

5114 5115 5116 5117 5118 5119 5120 5121 5122
static void __addrconf_sysctl_unregister(struct ipv6_devconf *p)
{
	struct addrconf_sysctl_table *t;

	if (p->sysctl == NULL)
		return;

	t = p->sysctl;
	p->sysctl = NULL;
5123
	unregister_net_sysctl_table(t->sysctl_header);
5124 5125 5126
	kfree(t);
}

5127 5128
static void addrconf_sysctl_register(struct inet6_dev *idev)
{
5129
	neigh_sysctl_register(idev->dev, idev->nd_parms,
5130
			      &ndisc_ifinfo_sysctl_change);
5131
	__addrconf_sysctl_register(dev_net(idev->dev), idev->dev->name,
5132
					idev, &idev->cnf);
5133 5134
}

5135
static void addrconf_sysctl_unregister(struct inet6_dev *idev)
L
Linus Torvalds 已提交
5136
{
5137 5138
	__addrconf_sysctl_unregister(&idev->cnf);
	neigh_sysctl_unregister(idev->nd_parms);
L
Linus Torvalds 已提交
5139 5140 5141 5142 5143
}


#endif

5144
static int __net_init addrconf_init_net(struct net *net)
5145
{
5146
	int err = -ENOMEM;
5147 5148
	struct ipv6_devconf *all, *dflt;

5149 5150 5151
	all = kmemdup(&ipv6_devconf, sizeof(ipv6_devconf), GFP_KERNEL);
	if (all == NULL)
		goto err_alloc_all;
5152

5153 5154 5155
	dflt = kmemdup(&ipv6_devconf_dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
	if (dflt == NULL)
		goto err_alloc_dflt;
5156

5157 5158 5159
	/* these will be inherited by all namespaces */
	dflt->autoconf = ipv6_defaults.autoconf;
	dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
5160 5161 5162 5163 5164

	net->ipv6.devconf_all = all;
	net->ipv6.devconf_dflt = dflt;

#ifdef CONFIG_SYSCTL
5165
	err = __addrconf_sysctl_register(net, "all", NULL, all);
5166 5167 5168
	if (err < 0)
		goto err_reg_all;

5169
	err = __addrconf_sysctl_register(net, "default", NULL, dflt);
5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186
	if (err < 0)
		goto err_reg_dflt;
#endif
	return 0;

#ifdef CONFIG_SYSCTL
err_reg_dflt:
	__addrconf_sysctl_unregister(all);
err_reg_all:
	kfree(dflt);
#endif
err_alloc_dflt:
	kfree(all);
err_alloc_all:
	return err;
}

5187
static void __net_exit addrconf_exit_net(struct net *net)
5188 5189 5190 5191 5192
{
#ifdef CONFIG_SYSCTL
	__addrconf_sysctl_unregister(net->ipv6.devconf_dflt);
	__addrconf_sysctl_unregister(net->ipv6.devconf_all);
#endif
O
Octavian Purdila 已提交
5193
	if (!net_eq(net, &init_net)) {
5194 5195 5196 5197 5198 5199 5200 5201 5202 5203
		kfree(net->ipv6.devconf_dflt);
		kfree(net->ipv6.devconf_all);
	}
}

static struct pernet_operations addrconf_ops = {
	.init = addrconf_init_net,
	.exit = addrconf_exit_net,
};

T
Thomas Graf 已提交
5204 5205 5206 5207
static struct rtnl_af_ops inet6_ops = {
	.family		  = AF_INET6,
	.fill_link_af	  = inet6_fill_link_af,
	.get_link_af_size = inet6_get_link_af_size,
5208
	.set_link_af	  = inet6_set_link_af,
T
Thomas Graf 已提交
5209 5210
};

L
Linus Torvalds 已提交
5211 5212 5213 5214 5215 5216
/*
 *	Init / cleanup code
 */

int __init addrconf_init(void)
{
5217
	int i, err;
5218

S
Stephen Hemminger 已提交
5219 5220
	err = ipv6_addr_label_init();
	if (err < 0) {
5221 5222
		pr_crit("%s: cannot initialize default policy table: %d\n",
			__func__, err);
5223
		goto out;
5224
	}
L
Linus Torvalds 已提交
5225

5226 5227 5228
	err = register_pernet_subsys(&addrconf_ops);
	if (err < 0)
		goto out_addrlabel;
5229

L
Linus Torvalds 已提交
5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248
	/* The addrconf netdev notifier requires that loopback_dev
	 * has it's ipv6 private information allocated and setup
	 * before it can bring up and give link-local addresses
	 * to other devices which are up.
	 *
	 * Unfortunately, loopback_dev is not necessarily the first
	 * entry in the global dev_base list of net devices.  In fact,
	 * it is likely to be the very last entry on that list.
	 * So this causes the notifier registry below to try and
	 * give link-local addresses to all devices besides loopback_dev
	 * first, then loopback_dev, which cases all the non-loopback_dev
	 * devices to fail to get a link-local address.
	 *
	 * So, as a temporary fix, allocate the ipv6 structure for
	 * loopback_dev first by hand.
	 * Longer term, all of the dependencies ipv6 has upon the loopback
	 * device and it being up should be removed.
	 */
	rtnl_lock();
5249
	if (!ipv6_add_dev(init_net.loopback_dev))
L
Linus Torvalds 已提交
5250 5251 5252
		err = -ENOMEM;
	rtnl_unlock();
	if (err)
5253
		goto errlo;
L
Linus Torvalds 已提交
5254

5255 5256 5257
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		INIT_HLIST_HEAD(&inet6_addr_lst[i]);

L
Linus Torvalds 已提交
5258 5259 5260
	register_netdevice_notifier(&ipv6_dev_notf);

	addrconf_verify(0);
5261

5262
	rtnl_af_register(&inet6_ops);
T
Thomas Graf 已提交
5263

5264 5265
	err = __rtnl_register(PF_INET6, RTM_GETLINK, NULL, inet6_dump_ifinfo,
			      NULL);
5266 5267 5268 5269
	if (err < 0)
		goto errout;

	/* Only the first call to __rtnl_register can fail */
5270 5271 5272 5273 5274 5275 5276 5277
	__rtnl_register(PF_INET6, RTM_NEWADDR, inet6_rtm_newaddr, NULL, NULL);
	__rtnl_register(PF_INET6, RTM_DELADDR, inet6_rtm_deladdr, NULL, NULL);
	__rtnl_register(PF_INET6, RTM_GETADDR, inet6_rtm_getaddr,
			inet6_dump_ifaddr, NULL);
	__rtnl_register(PF_INET6, RTM_GETMULTICAST, NULL,
			inet6_dump_ifmcaddr, NULL);
	__rtnl_register(PF_INET6, RTM_GETANYCAST, NULL,
			inet6_dump_ifacaddr, NULL);
5278
	__rtnl_register(PF_INET6, RTM_GETNETCONF, inet6_netconf_get_devconf,
5279
			inet6_netconf_dump_devconf, NULL);
5280

5281 5282
	ipv6_addr_label_rtnl_register();

L
Linus Torvalds 已提交
5283
	return 0;
5284
errout:
T
Thomas Graf 已提交
5285
	rtnl_af_unregister(&inet6_ops);
5286
	unregister_netdevice_notifier(&ipv6_dev_notf);
5287 5288
errlo:
	unregister_pernet_subsys(&addrconf_ops);
5289 5290 5291
out_addrlabel:
	ipv6_addr_label_cleanup();
out:
5292
	return err;
L
Linus Torvalds 已提交
5293 5294
}

5295
void addrconf_cleanup(void)
L
Linus Torvalds 已提交
5296
{
5297
	struct net_device *dev;
L
Linus Torvalds 已提交
5298 5299 5300
	int i;

	unregister_netdevice_notifier(&ipv6_dev_notf);
5301
	unregister_pernet_subsys(&addrconf_ops);
5302
	ipv6_addr_label_cleanup();
L
Linus Torvalds 已提交
5303 5304 5305

	rtnl_lock();

T
Thomas Graf 已提交
5306 5307
	__rtnl_af_unregister(&inet6_ops);

5308 5309 5310 5311 5312 5313 5314 5315
	/* clean dev list */
	for_each_netdev(&init_net, dev) {
		if (__in6_dev_get(dev) == NULL)
			continue;
		addrconf_ifdown(dev, 1);
	}
	addrconf_ifdown(init_net.loopback_dev, 2);

L
Linus Torvalds 已提交
5316 5317 5318
	/*
	 *	Check hash table.
	 */
5319
	spin_lock_bh(&addrconf_hash_lock);
5320 5321
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		WARN_ON(!hlist_empty(&inet6_addr_lst[i]));
5322
	spin_unlock_bh(&addrconf_hash_lock);
L
Linus Torvalds 已提交
5323 5324 5325 5326

	del_timer(&addr_chk_timer);
	rtnl_unlock();
}