addrconf.c 140.5 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
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
49
#include <linux/inet.h>
L
Linus Torvalds 已提交
50 51
#include <linux/in6.h>
#include <linux/netdevice.h>
52
#include <linux/if_addr.h>
L
Linus Torvalds 已提交
53 54 55 56 57 58
#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>
59
#include <linux/slab.h>
L
Linus Torvalds 已提交
60 61 62
#ifdef CONFIG_SYSCTL
#include <linux/sysctl.h>
#endif
63
#include <linux/capability.h>
L
Linus Torvalds 已提交
64 65
#include <linux/delay.h>
#include <linux/notifier.h>
66
#include <linux/string.h>
E
Eric Dumazet 已提交
67
#include <linux/hash.h>
L
Linus Torvalds 已提交
68

69
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
70 71 72
#include <net/sock.h>
#include <net/snmp.h>

73
#include <net/af_ieee802154.h>
74
#include <net/firewire.h>
L
Linus Torvalds 已提交
75 76 77 78 79 80 81
#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>
82
#include <net/netlink.h>
83
#include <net/pkt_sched.h>
L
Linus Torvalds 已提交
84 85
#include <linux/if_tunnel.h>
#include <linux/rtnetlink.h>
86
#include <linux/netconf.h>
L
Linus Torvalds 已提交
87
#include <linux/random.h>
S
Stephen Hemminger 已提交
88
#include <linux/uaccess.h>
89
#include <asm/unaligned.h>
L
Linus Torvalds 已提交
90 91 92

#include <linux/proc_fs.h>
#include <linux/seq_file.h>
93
#include <linux/export.h>
L
Linus Torvalds 已提交
94 95 96 97 98

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

#if ACONF_DEBUG >= 3
99
#define ADBG(fmt, ...) printk(fmt, ##__VA_ARGS__)
L
Linus Torvalds 已提交
100
#else
101
#define ADBG(fmt, ...) do { if (0) printk(fmt, ##__VA_ARGS__); } while (0)
L
Linus Torvalds 已提交
102 103 104
#endif

#define	INFINITY_LIFE_TIME	0xFFFFFFFF
105

106 107 108
#define IPV6_MAX_STRLEN \
	sizeof("ffff:ffff:ffff:ffff:ffff:ffff:255.255.255.255")

109 110 111 112
static inline u32 cstamp_delta(unsigned long cstamp)
{
	return (cstamp - INITIAL_JIFFIES) * 100UL / HZ;
}
L
Linus Torvalds 已提交
113 114

#ifdef CONFIG_SYSCTL
115
static int addrconf_sysctl_register(struct inet6_dev *idev);
116 117
static void addrconf_sysctl_unregister(struct inet6_dev *idev);
#else
118
static inline int addrconf_sysctl_register(struct inet6_dev *idev)
119
{
120
	return 0;
121 122 123 124 125
}

static inline void addrconf_sysctl_unregister(struct inet6_dev *idev)
{
}
L
Linus Torvalds 已提交
126 127
#endif

S
Sorin Dumitru 已提交
128 129
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
130 131
static void ipv6_regen_rndid(unsigned long data);

132
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev);
L
Linus Torvalds 已提交
133
static int ipv6_count_addresses(struct inet6_dev *idev);
134 135 136
static int ipv6_generate_stable_address(struct in6_addr *addr,
					u8 dad_count,
					const struct inet6_dev *idev);
L
Linus Torvalds 已提交
137 138 139 140

/*
 *	Configured unicast address hash table
 */
141
static struct hlist_head inet6_addr_lst[IN6_ADDR_HSIZE];
142
static DEFINE_SPINLOCK(addrconf_hash_lock);
L
Linus Torvalds 已提交
143

144 145 146
static void addrconf_verify(void);
static void addrconf_verify_rtnl(void);
static void addrconf_verify_work(struct work_struct *);
L
Linus Torvalds 已提交
147

148 149
static struct workqueue_struct *addrconf_wq;
static DECLARE_DELAYED_WORK(addr_chk_work, addrconf_verify_work);
L
Linus Torvalds 已提交
150 151 152 153

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

154 155
static void addrconf_type_change(struct net_device *dev,
				 unsigned long event);
L
Linus Torvalds 已提交
156 157
static int addrconf_ifdown(struct net_device *dev, int how);

158 159 160 161 162
static struct rt6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags);

163
static void addrconf_dad_start(struct inet6_ifaddr *ifp);
164
static void addrconf_dad_work(struct work_struct *w);
L
Linus Torvalds 已提交
165
static void addrconf_dad_completed(struct inet6_ifaddr *ifp);
166
static void addrconf_dad_run(struct inet6_dev *idev);
L
Linus Torvalds 已提交
167 168 169 170
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);

171
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
172
				struct prefix_info *pinfo);
173 174
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev);
L
Linus Torvalds 已提交
175

176
static struct ipv6_devconf ipv6_devconf __read_mostly = {
L
Linus Torvalds 已提交
177 178 179 180 181 182 183
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
	.force_mld_version	= 0,
184 185
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
186 187 188 189
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
190
	.use_tempaddr		= 0,
L
Linus Torvalds 已提交
191 192 193 194 195
	.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,
196
	.accept_ra_defrtr	= 1,
197
	.accept_ra_from_local	= 0,
198
	.accept_ra_pinfo	= 1,
199 200
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
201
	.rtr_probe_interval	= 60 * HZ,
202 203 204
#ifdef CONFIG_IPV6_ROUTE_INFO
	.accept_ra_rt_info_max_plen = 0,
#endif
205
#endif
206
	.proxy_ndp		= 0,
207
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
208
	.disable_ipv6		= 0,
209
	.accept_dad		= 1,
210
	.suppress_frag_ndisc	= 1,
211
	.accept_ra_mtu		= 1,
212 213 214
	.stable_secret		= {
		.initialized = false,
	}
L
Linus Torvalds 已提交
215 216
};

217
static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
L
Linus Torvalds 已提交
218 219 220 221 222 223
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
224 225 226
	.force_mld_version	= 0,
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
227 228 229 230 231 232 233 234 235 236
	.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,
237
	.accept_ra_defrtr	= 1,
238
	.accept_ra_from_local	= 0,
239
	.accept_ra_pinfo	= 1,
240 241
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
242
	.rtr_probe_interval	= 60 * HZ,
243 244 245
#ifdef CONFIG_IPV6_ROUTE_INFO
	.accept_ra_rt_info_max_plen = 0,
#endif
246
#endif
247
	.proxy_ndp		= 0,
248
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
249
	.disable_ipv6		= 0,
250
	.accept_dad		= 1,
251
	.suppress_frag_ndisc	= 1,
252
	.accept_ra_mtu		= 1,
253 254 255
	.stable_secret		= {
		.initialized = false,
	},
L
Linus Torvalds 已提交
256 257
};

258
/* Check if a valid qdisc is available */
259
static inline bool addrconf_qdisc_ok(const struct net_device *dev)
260
{
261
	return !qdisc_tx_is_noop(dev);
262 263
}

264
static void addrconf_del_rs_timer(struct inet6_dev *idev)
L
Linus Torvalds 已提交
265
{
266 267 268 269
	if (del_timer(&idev->rs_timer))
		__in6_dev_put(idev);
}

270
static void addrconf_del_dad_work(struct inet6_ifaddr *ifp)
271
{
272
	if (cancel_delayed_work(&ifp->dad_work))
L
Linus Torvalds 已提交
273 274 275
		__in6_ifa_put(ifp);
}

276 277 278 279 280 281 282
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
283

284 285
static void addrconf_mod_dad_work(struct inet6_ifaddr *ifp,
				   unsigned long delay)
L
Linus Torvalds 已提交
286
{
287
	if (!delayed_work_pending(&ifp->dad_work))
L
Linus Torvalds 已提交
288
		in6_ifa_hold(ifp);
289
	mod_delayed_work(addrconf_wq, &ifp->dad_work, delay);
L
Linus Torvalds 已提交
290 291
}

292 293
static int snmp6_alloc_dev(struct inet6_dev *idev)
{
294 295
	int i;

W
WANG Cong 已提交
296 297
	idev->stats.ipv6 = alloc_percpu(struct ipstats_mib);
	if (!idev->stats.ipv6)
298
		goto err_ip;
299 300 301

	for_each_possible_cpu(i) {
		struct ipstats_mib *addrconf_stats;
W
WANG Cong 已提交
302
		addrconf_stats = per_cpu_ptr(idev->stats.ipv6, i);
303 304 305 306
		u64_stats_init(&addrconf_stats->syncp);
	}


307 308 309
	idev->stats.icmpv6dev = kzalloc(sizeof(struct icmpv6_mib_device),
					GFP_KERNEL);
	if (!idev->stats.icmpv6dev)
310
		goto err_icmp;
311 312 313
	idev->stats.icmpv6msgdev = kzalloc(sizeof(struct icmpv6msg_mib_device),
					   GFP_KERNEL);
	if (!idev->stats.icmpv6msgdev)
314
		goto err_icmpmsg;
315 316 317

	return 0;

318
err_icmpmsg:
319
	kfree(idev->stats.icmpv6dev);
320
err_icmp:
W
WANG Cong 已提交
321
	free_percpu(idev->stats.ipv6);
322
err_ip:
323
	return -ENOMEM;
324 325
}

326
static struct inet6_dev *ipv6_add_dev(struct net_device *dev)
L
Linus Torvalds 已提交
327 328
{
	struct inet6_dev *ndev;
329
	int err = -ENOMEM;
L
Linus Torvalds 已提交
330 331 332 333

	ASSERT_RTNL();

	if (dev->mtu < IPV6_MIN_MTU)
334
		return ERR_PTR(-EINVAL);
L
Linus Torvalds 已提交
335

336 337
	ndev = kzalloc(sizeof(struct inet6_dev), GFP_KERNEL);
	if (ndev == NULL)
338
		return ERR_PTR(err);
339 340 341

	rwlock_init(&ndev->lock);
	ndev->dev = dev;
342
	INIT_LIST_HEAD(&ndev->addr_list);
343 344
	setup_timer(&ndev->rs_timer, addrconf_rs_timer,
		    (unsigned long)ndev);
345
	memcpy(&ndev->cnf, dev_net(dev)->ipv6.devconf_dflt, sizeof(ndev->cnf));
346 347 348 349 350
	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);
351
		return ERR_PTR(err);
352
	}
353 354
	if (ndev->cnf.forwarding)
		dev_disable_lro(dev);
355 356
	/* We refer to the device */
	dev_hold(dev);
L
Linus Torvalds 已提交
357

358
	if (snmp6_alloc_dev(ndev) < 0) {
359
		ADBG(KERN_WARNING
360
			"%s: cannot allocate memory for statistics; dev=%s.\n",
361
			__func__, dev->name);
362
		neigh_parms_release(&nd_tbl, ndev->nd_parms);
R
Roy Li 已提交
363 364
		dev_put(dev);
		kfree(ndev);
365
		return ERR_PTR(err);
366
	}
L
Linus Torvalds 已提交
367

368
	if (snmp6_register_dev(ndev) < 0) {
369
		ADBG(KERN_WARNING
370
			"%s: cannot create /proc/net/dev_snmp6/%s\n",
371
			__func__, dev->name);
372
		goto err_release;
373 374 375 376 377 378
	}

	/* One reference from device.  We must do this before
	 * we invoke __ipv6_regen_rndid().
	 */
	in6_dev_hold(ndev);
L
Linus Torvalds 已提交
379

380 381 382
	if (dev->flags & (IFF_NOARP | IFF_LOOPBACK))
		ndev->cnf.accept_dad = -1;

A
Amerigo Wang 已提交
383
#if IS_ENABLED(CONFIG_IPV6_SIT)
384
	if (dev->type == ARPHRD_SIT && (dev->priv_flags & IFF_ISATAP)) {
385
		pr_info("%s: Disabled Multicast RS\n", dev->name);
386 387 388 389
		ndev->cnf.rtr_solicits = 0;
	}
#endif

390
	INIT_LIST_HEAD(&ndev->tempaddr_list);
391
	setup_timer(&ndev->regen_timer, ipv6_regen_rndid, (unsigned long)ndev);
392 393
	if ((dev->flags&IFF_LOOPBACK) ||
	    dev->type == ARPHRD_TUNNEL ||
394
	    dev->type == ARPHRD_TUNNEL6 ||
395 396
	    dev->type == ARPHRD_SIT ||
	    dev->type == ARPHRD_NONE) {
397 398 399 400 401
		ndev->cnf.use_tempaddr = -1;
	} else {
		in6_dev_hold(ndev);
		ipv6_regen_rndid((unsigned long) ndev);
	}
402

403
	ndev->token = in6addr_any;
L
Linus Torvalds 已提交
404

405
	if (netif_running(dev) && addrconf_qdisc_ok(dev))
406 407
		ndev->if_flags |= IF_READY;

408 409
	ipv6_mc_init_dev(ndev);
	ndev->tstamp = jiffies;
410 411 412 413 414 415
	err = addrconf_sysctl_register(ndev);
	if (err) {
		ipv6_mc_destroy_dev(ndev);
		del_timer(&ndev->regen_timer);
		goto err_release;
	}
416
	/* protected by rtnl_lock */
417
	rcu_assign_pointer(dev->ip6_ptr, ndev);
418

419 420 421
	/* Join interface-local all-node multicast group */
	ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allnodes);

422
	/* Join all-node multicast group */
423
	ipv6_dev_mc_inc(dev, &in6addr_linklocal_allnodes);
424

425
	/* Join all-router multicast group if forwarding is set */
L
Li Wei 已提交
426
	if (ndev->cnf.forwarding && (dev->flags & IFF_MULTICAST))
427 428
		ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);

L
Linus Torvalds 已提交
429
	return ndev;
430 431 432 433 434 435

err_release:
	neigh_parms_release(&nd_tbl, ndev->nd_parms);
	ndev->dead = 1;
	in6_dev_finish_destroy(ndev);
	return ERR_PTR(err);
L
Linus Torvalds 已提交
436 437
}

438
static struct inet6_dev *ipv6_find_idev(struct net_device *dev)
L
Linus Torvalds 已提交
439 440 441 442 443
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
444 445 446
	idev = __in6_dev_get(dev);
	if (!idev) {
		idev = ipv6_add_dev(dev);
447
		if (IS_ERR(idev))
L
Linus Torvalds 已提交
448 449
			return NULL;
	}
450

L
Linus Torvalds 已提交
451 452 453 454 455
	if (dev->flags&IFF_UP)
		ipv6_mc_up(idev);
	return idev;
}

456 457 458 459 460
static int inet6_netconf_msgsize_devconf(int type)
{
	int size =  NLMSG_ALIGN(sizeof(struct netconfmsg))
		    + nla_total_size(4);	/* NETCONFA_IFINDEX */

461 462
	/* type -1 is used for ALL */
	if (type == -1 || type == NETCONFA_FORWARDING)
463
		size += nla_total_size(4);
464
#ifdef CONFIG_IPV6_MROUTE
465 466
	if (type == -1 || type == NETCONFA_MC_FORWARDING)
		size += nla_total_size(4);
467
#endif
468 469
	if (type == -1 || type == NETCONFA_PROXY_NEIGH)
		size += nla_total_size(4);
470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492

	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;

493 494
	/* type -1 is used for ALL */
	if ((type == -1 || type == NETCONFA_FORWARDING) &&
495 496
	    nla_put_s32(skb, NETCONFA_FORWARDING, devconf->forwarding) < 0)
		goto nla_put_failure;
497
#ifdef CONFIG_IPV6_MROUTE
498 499 500 501
	if ((type == -1 || type == NETCONFA_MC_FORWARDING) &&
	    nla_put_s32(skb, NETCONFA_MC_FORWARDING,
			devconf->mc_forwarding) < 0)
		goto nla_put_failure;
502
#endif
503 504 505 506
	if ((type == -1 || type == NETCONFA_PROXY_NEIGH) &&
	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
		goto nla_put_failure;

507 508
	nlmsg_end(skb, nlh);
	return 0;
509 510 511 512 513 514

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

515 516
void inet6_netconf_notify_devconf(struct net *net, int type, int ifindex,
				  struct ipv6_devconf *devconf)
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535
{
	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:
536
	rtnl_set_sk_err(net, RTNLGRP_IPV6_NETCONF, err);
537 538
}

539 540 541
static const struct nla_policy devconf_ipv6_policy[NETCONFA_MAX+1] = {
	[NETCONFA_IFINDEX]	= { .len = sizeof(int) },
	[NETCONFA_FORWARDING]	= { .len = sizeof(int) },
542
	[NETCONFA_PROXY_NEIGH]	= { .len = sizeof(int) },
543 544 545
};

static int inet6_netconf_get_devconf(struct sk_buff *in_skb,
546
				     struct nlmsghdr *nlh)
547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605
{
	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;
}

606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622
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();
623 624
		cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^
			  net->dev_base_seq;
625 626 627 628 629 630 631 632 633 634 635 636 637
		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,
638
						       -1) < 0) {
639 640 641
				rcu_read_unlock();
				goto done;
			}
642
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
643 644 645 646 647 648 649 650 651 652 653
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,
654
					       -1) < 0)
655 656 657 658 659 660 661 662 663 664
			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,
665
					       -1) < 0)
666 667 668 669 670 671 672 673 674 675 676
			goto done;
		else
			h++;
	}
done:
	cb->args[0] = h;
	cb->args[1] = idx;

	return skb->len;
}

L
Linus Torvalds 已提交
677 678 679 680 681 682 683 684 685
#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;
686 687
	if (idev->cnf.forwarding)
		dev_disable_lro(dev);
A
Amerigo Wang 已提交
688
	if (dev->flags & IFF_MULTICAST) {
689
		if (idev->cnf.forwarding) {
690
			ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);
691 692 693
			ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_inc(dev, &in6addr_sitelocal_allrouters);
		} else {
694
			ipv6_dev_mc_dec(dev, &in6addr_linklocal_allrouters);
695 696 697
			ipv6_dev_mc_dec(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_dec(dev, &in6addr_sitelocal_allrouters);
		}
L
Linus Torvalds 已提交
698
	}
699 700

	list_for_each_entry(ifa, &idev->addr_list, if_list) {
M
Michal Wrobel 已提交
701 702
		if (ifa->flags&IFA_F_TENTATIVE)
			continue;
L
Linus Torvalds 已提交
703 704 705 706 707
		if (idev->cnf.forwarding)
			addrconf_join_anycast(ifa);
		else
			addrconf_leave_anycast(ifa);
	}
708 709
	inet6_netconf_notify_devconf(dev_net(dev), NETCONFA_FORWARDING,
				     dev->ifindex, &idev->cnf);
L
Linus Torvalds 已提交
710 711 712
}


713
static void addrconf_forward_change(struct net *net, __s32 newf)
L
Linus Torvalds 已提交
714 715 716 717
{
	struct net_device *dev;
	struct inet6_dev *idev;

718
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
719 720
		idev = __in6_dev_get(dev);
		if (idev) {
721 722
			int changed = (!idev->cnf.forwarding) ^ (!newf);
			idev->cnf.forwarding = newf;
L
Linus Torvalds 已提交
723 724 725 726 727
			if (changed)
				dev_forward_change(idev);
		}
	}
}
728

729
static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
730
{
731
	struct net *net;
732 733 734 735
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
736 737

	net = (struct net *)table->extra2;
738 739
	old = *p;
	*p = newf;
740

741
	if (p == &net->ipv6.devconf_dflt->forwarding) {
742 743 744 745
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net, NETCONFA_FORWARDING,
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
746 747
		rtnl_unlock();
		return 0;
E
Eric W. Biederman 已提交
748
	}
749

750 751 752
	if (p == &net->ipv6.devconf_all->forwarding) {
		net->ipv6.devconf_dflt->forwarding = newf;
		addrconf_forward_change(net, newf);
753 754 755 756
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net, NETCONFA_FORWARDING,
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
757
	} else if ((!newf) ^ (!old))
758
		dev_forward_change((struct inet6_dev *)table->extra1);
759
	rtnl_unlock();
760

761
	if (newf)
762
		rt6_purge_dflt_routers(net);
763
	return 1;
764
}
L
Linus Torvalds 已提交
765 766
#endif

767
/* Nobody refers to this ifaddr, destroy it */
L
Linus Torvalds 已提交
768 769
void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp)
{
770
	WARN_ON(!hlist_unhashed(&ifp->addr_lst));
771

L
Linus Torvalds 已提交
772
#ifdef NET_REFCNT_DEBUG
773
	pr_debug("%s\n", __func__);
L
Linus Torvalds 已提交
774 775 776 777
#endif

	in6_dev_put(ifp->idev);

778 779 780
	if (cancel_delayed_work(&ifp->dad_work))
		pr_notice("delayed DAD work was pending while freeing ifa=%p\n",
			  ifp);
L
Linus Torvalds 已提交
781

782
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
783
		pr_warn("Freeing alive inet6 address %p\n", ifp);
L
Linus Torvalds 已提交
784 785
		return;
	}
A
Amerigo Wang 已提交
786
	ip6_rt_put(ifp->rt);
L
Linus Torvalds 已提交
787

788
	kfree_rcu(ifp, rcu);
L
Linus Torvalds 已提交
789 790
}

B
Brian Haley 已提交
791 792 793
static void
ipv6_link_dev_addr(struct inet6_dev *idev, struct inet6_ifaddr *ifp)
{
794
	struct list_head *p;
795
	int ifp_scope = ipv6_addr_src_scope(&ifp->addr);
B
Brian Haley 已提交
796 797 798 799 800

	/*
	 * Each device address list is sorted in order of scope -
	 * global before linklocal.
	 */
801 802 803
	list_for_each(p, &idev->addr_list) {
		struct inet6_ifaddr *ifa
			= list_entry(p, struct inet6_ifaddr, if_list);
804
		if (ifp_scope >= ipv6_addr_src_scope(&ifa->addr))
B
Brian Haley 已提交
805 806 807
			break;
	}

808
	list_add_tail(&ifp->if_list, p);
B
Brian Haley 已提交
809 810
}

E
Eric Dumazet 已提交
811
static u32 inet6_addr_hash(const struct in6_addr *addr)
812
{
E
Eric Dumazet 已提交
813
	return hash_32(ipv6_addr_hash(addr), IN6_ADDR_HSIZE_SHIFT);
814 815
}

L
Linus Torvalds 已提交
816 817 818
/* On success it returns ifp with increased reference count */

static struct inet6_ifaddr *
819 820
ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
	      const struct in6_addr *peer_addr, int pfxlen,
821
	      int scope, u32 flags, u32 valid_lft, u32 prefered_lft)
L
Linus Torvalds 已提交
822 823 824
{
	struct inet6_ifaddr *ifa = NULL;
	struct rt6_info *rt;
825
	unsigned int hash;
L
Linus Torvalds 已提交
826
	int err = 0;
827 828 829 830 831 832 833
	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
834

835
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
836 837 838 839 840
	if (idev->dead) {
		err = -ENODEV;			/*XXX*/
		goto out2;
	}

841
	if (idev->cnf.disable_ipv6) {
842 843 844 845
		err = -EACCES;
		goto out2;
	}

846
	spin_lock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
847 848

	/* Ignore adding duplicate addresses on an interface */
849
	if (ipv6_chk_same_addr(dev_net(idev->dev), addr, idev->dev)) {
850
		ADBG("ipv6_add_addr: already assigned\n");
L
Linus Torvalds 已提交
851 852 853 854
		err = -EEXIST;
		goto out;
	}

855
	ifa = kzalloc(sizeof(struct inet6_ifaddr), GFP_ATOMIC);
L
Linus Torvalds 已提交
856 857

	if (ifa == NULL) {
858
		ADBG("ipv6_add_addr: malloc failed\n");
L
Linus Torvalds 已提交
859 860 861 862
		err = -ENOBUFS;
		goto out;
	}

863
	rt = addrconf_dst_alloc(idev, addr, false);
L
Linus Torvalds 已提交
864 865 866 867 868
	if (IS_ERR(rt)) {
		err = PTR_ERR(rt);
		goto out;
	}

869 870
	neigh_parms_data_state_setall(idev->nd_parms);

A
Alexey Dobriyan 已提交
871
	ifa->addr = *addr;
872 873
	if (peer_addr)
		ifa->peer_addr = *peer_addr;
L
Linus Torvalds 已提交
874 875

	spin_lock_init(&ifa->lock);
876
	INIT_DELAYED_WORK(&ifa->dad_work, addrconf_dad_work);
877
	INIT_HLIST_NODE(&ifa->addr_lst);
L
Linus Torvalds 已提交
878 879 880
	ifa->scope = scope;
	ifa->prefix_len = pfxlen;
	ifa->flags = flags | IFA_F_TENTATIVE;
881 882
	ifa->valid_lft = valid_lft;
	ifa->prefered_lft = prefered_lft;
L
Linus Torvalds 已提交
883
	ifa->cstamp = ifa->tstamp = jiffies;
884
	ifa->tokenized = false;
L
Linus Torvalds 已提交
885

886 887
	ifa->rt = rt;

L
Linus Torvalds 已提交
888 889 890 891 892 893
	ifa->idev = idev;
	in6_dev_hold(idev);
	/* For caller */
	in6_ifa_hold(ifa);

	/* Add to big hash table */
E
Eric Dumazet 已提交
894
	hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
895

896 897
	hlist_add_head_rcu(&ifa->addr_lst, &inet6_addr_lst[hash]);
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
898 899 900

	write_lock(&idev->lock);
	/* Add to inet6_dev unicast addr list. */
B
Brian Haley 已提交
901
	ipv6_link_dev_addr(idev, ifa);
L
Linus Torvalds 已提交
902 903

	if (ifa->flags&IFA_F_TEMPORARY) {
904
		list_add(&ifa->tmp_list, &idev->tempaddr_list);
L
Linus Torvalds 已提交
905 906 907 908 909 910
		in6_ifa_hold(ifa);
	}

	in6_ifa_hold(ifa);
	write_unlock(&idev->lock);
out2:
911
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
912

913
	if (likely(err == 0))
914
		inet6addr_notifier_call_chain(NETDEV_UP, ifa);
L
Linus Torvalds 已提交
915 916 917 918 919 920 921
	else {
		kfree(ifa);
		ifa = ERR_PTR(err);
	}

	return ifa;
out:
922
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
923 924 925
	goto out2;
}

926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
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
1008 1009 1010 1011
/* This function wants to get referenced ifp and releases it before return */

static void ipv6_del_addr(struct inet6_ifaddr *ifp)
{
1012
	int state;
1013 1014
	enum cleanup_prefix_rt_t action = CLEANUP_PREFIX_RT_NOP;
	unsigned long expires;
L
Linus Torvalds 已提交
1015

1016 1017
	ASSERT_RTNL();

1018
	spin_lock_bh(&ifp->lock);
1019
	state = ifp->state;
1020
	ifp->state = INET6_IFADDR_STATE_DEAD;
1021
	spin_unlock_bh(&ifp->lock);
1022 1023 1024

	if (state == INET6_IFADDR_STATE_DEAD)
		goto out;
L
Linus Torvalds 已提交
1025

1026 1027 1028
	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
1029

1030
	write_lock_bh(&ifp->idev->lock);
1031

L
Linus Torvalds 已提交
1032
	if (ifp->flags&IFA_F_TEMPORARY) {
1033 1034 1035 1036
		list_del(&ifp->tmp_list);
		if (ifp->ifpub) {
			in6_ifa_put(ifp->ifpub);
			ifp->ifpub = NULL;
L
Linus Torvalds 已提交
1037
		}
1038
		__in6_ifa_put(ifp);
L
Linus Torvalds 已提交
1039 1040
	}

1041 1042
	if (ifp->flags & IFA_F_PERMANENT && !(ifp->flags & IFA_F_NOPREFIXROUTE))
		action = check_cleanup_prefix_route(ifp, &expires);
1043

1044 1045 1046 1047
	list_del_init(&ifp->if_list);
	__in6_ifa_put(ifp);

	write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
1048

1049
	addrconf_del_dad_work(ifp);
1050

L
Linus Torvalds 已提交
1051 1052
	ipv6_ifa_notify(RTM_DELADDR, ifp);

1053
	inet6addr_notifier_call_chain(NETDEV_DOWN, ifp);
L
Linus Torvalds 已提交
1054

1055 1056 1057
	if (action != CLEANUP_PREFIX_RT_NOP) {
		cleanup_prefix_route(ifp, expires,
			action == CLEANUP_PREFIX_RT_DEL);
L
Linus Torvalds 已提交
1058 1059
	}

1060 1061
	/* clean up prefsrc entries */
	rt6_remove_prefsrc(ifp);
1062
out:
L
Linus Torvalds 已提交
1063 1064 1065 1066 1067 1068 1069
	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;
1070
	unsigned long tmp_prefered_lft, tmp_valid_lft, tmp_tstamp, age;
1071
	unsigned long regen_advance;
L
Linus Torvalds 已提交
1072 1073
	int tmp_plen;
	int ret = 0;
1074
	u32 addr_flags;
1075
	unsigned long now = jiffies;
L
Linus Torvalds 已提交
1076

1077
	write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088
	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) {
1089
		write_unlock_bh(&idev->lock);
1090
		pr_info("%s: use_tempaddr is disabled\n", __func__);
L
Linus Torvalds 已提交
1091 1092 1093 1094 1095 1096 1097 1098
		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);
1099
		write_unlock_bh(&idev->lock);
1100 1101
		pr_warn("%s: regeneration time exceeded - disabled temporary address support\n",
			__func__);
L
Linus Torvalds 已提交
1102 1103 1104 1105 1106 1107
		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
1108
	__ipv6_try_regen_rndid(idev, tmpaddr);
L
Linus Torvalds 已提交
1109
	memcpy(&addr.s6_addr[8], idev->rndid, 8);
1110
	age = (now - ifp->tstamp) / HZ;
L
Linus Torvalds 已提交
1111 1112
	tmp_valid_lft = min_t(__u32,
			      ifp->valid_lft,
1113
			      idev->cnf.temp_valid_lft + age);
1114 1115
	tmp_prefered_lft = min_t(__u32,
				 ifp->prefered_lft,
1116
				 idev->cnf.temp_prefered_lft + age -
1117
				 idev->cnf.max_desync_factor);
L
Linus Torvalds 已提交
1118 1119 1120 1121
	tmp_plen = ifp->prefix_len;
	tmp_tstamp = ifp->tstamp;
	spin_unlock_bh(&ifp->lock);

1122
	regen_advance = idev->cnf.regen_max_retry *
1123 1124
			idev->cnf.dad_transmits *
			NEIGH_VAR(idev->nd_parms, RETRANS_TIME) / HZ;
1125
	write_unlock_bh(&idev->lock);
1126

1127 1128 1129 1130
	/* 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.
1131 1132
	 * Use age calculation as in addrconf_verify to avoid unnecessary
	 * temporary addresses being generated.
1133
	 */
1134 1135
	age = (now - tmp_tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
	if (tmp_prefered_lft <= regen_advance + age) {
1136 1137 1138 1139 1140 1141
		in6_ifa_put(ifp);
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}

1142 1143 1144 1145 1146
	addr_flags = IFA_F_TEMPORARY;
	/* set in addrconf_prefix_rcv() */
	if (ifp->flags & IFA_F_OPTIMISTIC)
		addr_flags |= IFA_F_OPTIMISTIC;

1147 1148 1149 1150
	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
1151 1152
		in6_ifa_put(ifp);
		in6_dev_put(idev);
1153
		pr_info("%s: retry temporary address regeneration\n", __func__);
L
Linus Torvalds 已提交
1154
		tmpaddr = &addr;
1155
		write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1156 1157 1158 1159 1160
		goto retry;
	}

	spin_lock_bh(&ift->lock);
	ift->ifpub = ifp;
1161
	ift->cstamp = now;
L
Linus Torvalds 已提交
1162 1163 1164
	ift->tstamp = tmp_tstamp;
	spin_unlock_bh(&ift->lock);

1165
	addrconf_dad_start(ift);
L
Linus Torvalds 已提交
1166 1167 1168 1169 1170 1171 1172
	in6_ifa_put(ift);
	in6_dev_put(idev);
out:
	return ret;
}

/*
1173
 *	Choose an appropriate source address (RFC3484)
L
Linus Torvalds 已提交
1174
 */
1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
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,
1188 1189 1190
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	IPV6_SADDR_RULE_NOT_OPTIMISTIC,
#endif
1191 1192 1193
	IPV6_SADDR_RULE_MAX
};

1194
struct ipv6_saddr_score {
1195 1196 1197 1198 1199 1200
	int			rule;
	int			addr_type;
	struct inet6_ifaddr	*ifa;
	DECLARE_BITMAP(scorebits, IPV6_SADDR_RULE_MAX);
	int			scopedist;
	int			matchlen;
1201 1202
};

1203
struct ipv6_saddr_dst {
1204
	const struct in6_addr *addr;
1205 1206 1207
	int ifindex;
	int scope;
	int label;
1208
	unsigned int prefs;
1209
};
1210

D
Dave Jones 已提交
1211
static inline int ipv6_saddr_preferred(int type)
L
Linus Torvalds 已提交
1212
{
U
Ulrich Weber 已提交
1213
	if (type & (IPV6_ADDR_MAPPED|IPV6_ADDR_COMPATv4|IPV6_ADDR_LOOPBACK))
1214 1215
		return 1;
	return 0;
L
Linus Torvalds 已提交
1216 1217
}

1218 1219 1220 1221 1222 1223 1224 1225 1226
static inline bool ipv6_use_optimistic_addr(struct inet6_dev *idev)
{
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	return idev && idev->cnf.optimistic_dad && idev->cnf.use_optimistic;
#else
	return false;
#endif
}

1227 1228
static int ipv6_get_saddr_eval(struct net *net,
			       struct ipv6_saddr_score *score,
1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267
			       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
1268
		 *  B-15 |    \        if scope is enough for destination.
1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286
		 *       |             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:
1287
	    {
1288
		/* Rule 3: Avoid deprecated and optimistic addresses */
1289 1290 1291 1292
		u8 avoid = IFA_F_DEPRECATED;

		if (!ipv6_use_optimistic_addr(score->ifa->idev))
			avoid |= IFA_F_OPTIMISTIC;
1293
		ret = ipv6_saddr_preferred(score->addr_type) ||
1294
		      !(score->ifa->flags & avoid);
1295
		break;
1296
	    }
1297 1298
#ifdef CONFIG_IPV6_MIP6
	case IPV6_SADDR_RULE_HOA:
1299
	    {
1300
		/* Rule 4: Prefer home address */
1301 1302
		int prefhome = !(dst->prefs & IPV6_PREFER_SRC_COA);
		ret = !(score->ifa->flags & IFA_F_HOMEADDRESS) ^ prefhome;
1303
		break;
1304
	    }
1305 1306 1307 1308 1309 1310 1311 1312
#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 */
1313 1314
		ret = ipv6_addr_label(net,
				      &score->ifa->addr, score->addr_type,
1315 1316 1317
				      score->ifa->idev->dev->ifindex) == dst->label;
		break;
	case IPV6_SADDR_RULE_PRIVACY:
1318
	    {
1319
		/* Rule 7: Prefer public address
L
Lucas De Marchi 已提交
1320
		 * Note: prefer temporary address if use_tempaddr >= 2
1321
		 */
1322 1323 1324 1325
		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;
1326
		break;
1327
	    }
1328 1329 1330 1331 1332 1333 1334 1335 1336
	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 */
1337 1338 1339 1340
		ret = ipv6_addr_diff(&score->ifa->addr, dst->addr);
		if (ret > score->ifa->prefix_len)
			ret = score->ifa->prefix_len;
		score->matchlen = ret;
1341
		break;
1342 1343 1344 1345 1346 1347 1348 1349
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	case IPV6_SADDR_RULE_NOT_OPTIMISTIC:
		/* Optimistic addresses still have lower precedence than other
		 * preferred addresses.
		 */
		ret = !(score->ifa->flags & IFA_F_OPTIMISTIC);
		break;
#endif
1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360
	default:
		ret = 0;
	}

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

1361
int ipv6_dev_get_saddr(struct net *net, const struct net_device *dst_dev,
1362
		       const struct in6_addr *daddr, unsigned int prefs,
1363
		       struct in6_addr *saddr)
L
Linus Torvalds 已提交
1364
{
1365 1366 1367
	struct ipv6_saddr_score scores[2],
				*score = &scores[0], *hiscore = &scores[1];
	struct ipv6_saddr_dst dst;
1368
	struct net_device *dev;
1369
	int dst_type;
L
Linus Torvalds 已提交
1370

1371 1372 1373 1374
	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);
1375
	dst.label = ipv6_addr_label(net, daddr, dst_type, dst.ifindex);
1376
	dst.prefs = prefs;
1377 1378 1379

	hiscore->rule = -1;
	hiscore->ifa = NULL;
L
Linus Torvalds 已提交
1380

1381
	rcu_read_lock();
L
Linus Torvalds 已提交
1382

1383
	for_each_netdev_rcu(net, dev) {
1384 1385
		struct inet6_dev *idev;

1386
		/* Candidate Source Address (section 4)
1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397
		 *  - 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.)
		 */
1398 1399 1400
		if (((dst_type & IPV6_ADDR_MULTICAST) ||
		     dst.scope <= IPV6_ADDR_SCOPE_LINKLOCAL) &&
		    dst.ifindex && dev->ifindex != dst.ifindex)
1401
			continue;
L
Linus Torvalds 已提交
1402 1403

		idev = __in6_dev_get(dev);
1404 1405 1406 1407
		if (!idev)
			continue;

		read_lock_bh(&idev->lock);
1408
		list_for_each_entry(score->ifa, &idev->addr_list, if_list) {
1409
			int i;
1410

1411
			/*
1412 1413 1414
			 * - Tentative Address (RFC2462 section 5.4)
			 *  - A tentative address is not considered
			 *    "assigned to an interface" in the traditional
1415
			 *    sense, unless it is also flagged as optimistic.
1416
			 * - Candidate Source Address (section 4)
1417 1418 1419 1420
			 *  - In any case, anycast addresses, multicast
			 *    addresses, and the unspecified address MUST
			 *    NOT be included in a candidate set.
			 */
1421 1422
			if ((score->ifa->flags & IFA_F_TENTATIVE) &&
			    (!(score->ifa->flags & IFA_F_OPTIMISTIC)))
1423
				continue;
1424 1425 1426 1427 1428

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

			if (unlikely(score->addr_type == IPV6_ADDR_ANY ||
				     score->addr_type & IPV6_ADDR_MULTICAST)) {
1429 1430
				net_dbg_ratelimited("ADDRCONF: unspecified / multicast address assigned as unicast address on %s",
						    dev->name);
1431 1432 1433
				continue;
			}

1434 1435 1436 1437 1438 1439
			score->rule = -1;
			bitmap_zero(score->scorebits, IPV6_SADDR_RULE_MAX);

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

1440 1441
				minihiscore = ipv6_get_saddr_eval(net, hiscore, &dst, i);
				miniscore = ipv6_get_saddr_eval(net, score, &dst, i);
1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462

				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);

1463
					swap(hiscore, score);
1464 1465 1466 1467 1468

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

					break;
L
Linus Torvalds 已提交
1469
				}
1470
			}
L
Linus Torvalds 已提交
1471
		}
1472
try_nextdev:
1473
		read_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
1474
	}
1475
	rcu_read_unlock();
L
Linus Torvalds 已提交
1476

1477
	if (!hiscore->ifa)
1478
		return -EADDRNOTAVAIL;
1479

A
Alexey Dobriyan 已提交
1480
	*saddr = hiscore->ifa->addr;
1481
	in6_ifa_put(hiscore->ifa);
1482
	return 0;
L
Linus Torvalds 已提交
1483
}
1484
EXPORT_SYMBOL(ipv6_dev_get_saddr);
L
Linus Torvalds 已提交
1485

1486
int __ipv6_get_lladdr(struct inet6_dev *idev, struct in6_addr *addr,
1487
		      u32 banned_flags)
1488 1489 1490 1491
{
	struct inet6_ifaddr *ifp;
	int err = -EADDRNOTAVAIL;

1492 1493 1494
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
1495 1496 1497 1498 1499 1500 1501 1502 1503 1504
		if (ifp->scope == IFA_LINK &&
		    !(ifp->flags & banned_flags)) {
			*addr = ifp->addr;
			err = 0;
			break;
		}
	}
	return err;
}

1505
int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr,
1506
		    u32 banned_flags)
L
Linus Torvalds 已提交
1507 1508 1509 1510
{
	struct inet6_dev *idev;
	int err = -EADDRNOTAVAIL;

1511
	rcu_read_lock();
S
Stephen Hemminger 已提交
1512 1513
	idev = __in6_dev_get(dev);
	if (idev) {
L
Linus Torvalds 已提交
1514
		read_lock_bh(&idev->lock);
1515
		err = __ipv6_get_lladdr(idev, addr, banned_flags);
L
Linus Torvalds 已提交
1516 1517
		read_unlock_bh(&idev->lock);
	}
1518
	rcu_read_unlock();
L
Linus Torvalds 已提交
1519 1520 1521 1522 1523 1524 1525 1526 1527
	return err;
}

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

	read_lock_bh(&idev->lock);
1528
	list_for_each_entry(ifp, &idev->addr_list, if_list)
L
Linus Torvalds 已提交
1529 1530 1531 1532 1533
		cnt++;
	read_unlock_bh(&idev->lock);
	return cnt;
}

1534
int ipv6_chk_addr(struct net *net, const struct in6_addr *addr,
1535
		  const struct net_device *dev, int strict)
1536 1537 1538 1539 1540 1541 1542 1543
{
	return ipv6_chk_addr_and_flags(net, addr, dev, strict, IFA_F_TENTATIVE);
}
EXPORT_SYMBOL(ipv6_chk_addr);

int ipv6_chk_addr_and_flags(struct net *net, const struct in6_addr *addr,
			    const struct net_device *dev, int strict,
			    u32 banned_flags)
L
Linus Torvalds 已提交
1544
{
1545
	struct inet6_ifaddr *ifp;
E
Eric Dumazet 已提交
1546
	unsigned int hash = inet6_addr_hash(addr);
1547
	u32 ifp_flags;
L
Linus Torvalds 已提交
1548

1549
	rcu_read_lock_bh();
1550
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
1551
		if (!net_eq(dev_net(ifp->idev->dev), net))
1552
			continue;
1553 1554 1555 1556 1557 1558
		/* Decouple optimistic from tentative for evaluation here.
		 * Ban optimistic addresses explicitly, when required.
		 */
		ifp_flags = (ifp->flags&IFA_F_OPTIMISTIC)
			    ? (ifp->flags&~IFA_F_TENTATIVE)
			    : ifp->flags;
L
Linus Torvalds 已提交
1559
		if (ipv6_addr_equal(&ifp->addr, addr) &&
1560
		    !(ifp_flags&banned_flags) &&
1561 1562 1563 1564
		    (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
1565 1566
		}
	}
1567

1568 1569
	rcu_read_unlock_bh();
	return 0;
L
Linus Torvalds 已提交
1570
}
1571
EXPORT_SYMBOL(ipv6_chk_addr_and_flags);
1572

1573 1574
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev)
L
Linus Torvalds 已提交
1575
{
E
Eric Dumazet 已提交
1576
	unsigned int hash = inet6_addr_hash(addr);
1577
	struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
1578

1579
	hlist_for_each_entry(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
		if (ipv6_addr_equal(&ifp->addr, addr)) {
			if (dev == NULL || ifp->idev->dev == dev)
1584
				return true;
L
Linus Torvalds 已提交
1585 1586
		}
	}
1587
	return false;
L
Linus Torvalds 已提交
1588 1589
}

1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
/* 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);

1617
int ipv6_chk_prefix(const struct in6_addr *addr, struct net_device *dev)
1618 1619 1620 1621 1622 1623 1624 1625 1626 1627
{
	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);
1628
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640
			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);

1641
struct inet6_ifaddr *ipv6_get_ifaddr(struct net *net, const struct in6_addr *addr,
1642
				     struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1643
{
1644
	struct inet6_ifaddr *ifp, *result = NULL;
E
Eric Dumazet 已提交
1645
	unsigned int hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
1646

1647
	rcu_read_lock_bh();
1648
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
1649
		if (!net_eq(dev_net(ifp->idev->dev), net))
1650
			continue;
L
Linus Torvalds 已提交
1651 1652 1653
		if (ipv6_addr_equal(&ifp->addr, addr)) {
			if (dev == NULL || ifp->idev->dev == dev ||
			    !(ifp->scope&(IFA_LINK|IFA_HOST) || strict)) {
1654
				result = ifp;
L
Linus Torvalds 已提交
1655 1656 1657 1658 1659
				in6_ifa_hold(ifp);
				break;
			}
		}
	}
1660
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1661

1662
	return result;
L
Linus Torvalds 已提交
1663 1664 1665 1666
}

/* Gets referenced address, destroys ifaddr */

B
Brian Haley 已提交
1667
static void addrconf_dad_stop(struct inet6_ifaddr *ifp, int dad_failed)
L
Linus Torvalds 已提交
1668 1669 1670
{
	if (ifp->flags&IFA_F_PERMANENT) {
		spin_lock_bh(&ifp->lock);
1671
		addrconf_del_dad_work(ifp);
L
Linus Torvalds 已提交
1672
		ifp->flags |= IFA_F_TENTATIVE;
B
Brian Haley 已提交
1673 1674
		if (dad_failed)
			ifp->flags |= IFA_F_DADFAILED;
L
Linus Torvalds 已提交
1675
		spin_unlock_bh(&ifp->lock);
1676 1677
		if (dad_failed)
			ipv6_ifa_notify(0, ifp);
L
Linus Torvalds 已提交
1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691
		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);
1692
	} else {
L
Linus Torvalds 已提交
1693
		ipv6_del_addr(ifp);
1694
	}
L
Linus Torvalds 已提交
1695 1696
}

H
Herbert Xu 已提交
1697 1698 1699 1700
static int addrconf_dad_end(struct inet6_ifaddr *ifp)
{
	int err = -ENOENT;

1701
	spin_lock_bh(&ifp->lock);
H
Herbert Xu 已提交
1702 1703 1704 1705
	if (ifp->state == INET6_IFADDR_STATE_DAD) {
		ifp->state = INET6_IFADDR_STATE_POSTDAD;
		err = 0;
	}
1706
	spin_unlock_bh(&ifp->lock);
H
Herbert Xu 已提交
1707 1708 1709 1710

	return err;
}

1711 1712
void addrconf_dad_failure(struct inet6_ifaddr *ifp)
{
1713
	struct in6_addr addr;
1714
	struct inet6_dev *idev = ifp->idev;
1715
	struct net *net = dev_net(ifp->idev->dev);
1716

1717 1718
	if (addrconf_dad_end(ifp)) {
		in6_ifa_put(ifp);
H
Herbert Xu 已提交
1719
		return;
1720
	}
H
Herbert Xu 已提交
1721

1722 1723
	net_info_ratelimited("%s: IPv6 duplicate address %pI6c detected!\n",
			     ifp->idev->dev->name, &ifp->addr);
1724

1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735
	spin_lock_bh(&ifp->lock);

	if (ifp->flags & IFA_F_STABLE_PRIVACY) {
		int scope = ifp->scope;
		u32 flags = ifp->flags;
		struct in6_addr new_addr;
		struct inet6_ifaddr *ifp2;
		u32 valid_lft, preferred_lft;
		int pfxlen = ifp->prefix_len;
		int retries = ifp->stable_privacy_retry + 1;

1736
		if (retries > net->ipv6.sysctl.idgen_retries) {
1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750
			net_info_ratelimited("%s: privacy stable address generation failed because of DAD conflicts!\n",
					     ifp->idev->dev->name);
			goto errdad;
		}

		new_addr = ifp->addr;
		if (ipv6_generate_stable_address(&new_addr, retries,
						 idev))
			goto errdad;

		valid_lft = ifp->valid_lft;
		preferred_lft = ifp->prefered_lft;

		spin_unlock_bh(&ifp->lock);
1751

1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770
		if (idev->cnf.max_addresses &&
		    ipv6_count_addresses(idev) >=
		    idev->cnf.max_addresses)
			goto lock_errdad;

		net_info_ratelimited("%s: generating new stable privacy address because of DAD conflict\n",
				     ifp->idev->dev->name);

		ifp2 = ipv6_add_addr(idev, &new_addr, NULL, pfxlen,
				     scope, flags, valid_lft,
				     preferred_lft);
		if (IS_ERR(ifp2))
			goto lock_errdad;

		spin_lock_bh(&ifp2->lock);
		ifp2->stable_privacy_retry = retries;
		ifp2->state = INET6_IFADDR_STATE_PREDAD;
		spin_unlock_bh(&ifp2->lock);

1771
		addrconf_mod_dad_work(ifp2, net->ipv6.sysctl.idgen_delay);
1772 1773 1774 1775
		in6_ifa_put(ifp2);
lock_errdad:
		spin_lock_bh(&ifp->lock);
	} else if (idev->cnf.accept_dad > 1 && !idev->cnf.disable_ipv6) {
1776 1777 1778 1779 1780 1781 1782
		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;
1783

1784
			pr_info("%s: IPv6 being disabled!\n",
1785
				ifp->idev->dev->name);
1786 1787 1788
		}
	}

1789
errdad:
1790 1791
	/* transition from _POSTDAD to _ERRDAD */
	ifp->state = INET6_IFADDR_STATE_ERRDAD;
1792
	spin_unlock_bh(&ifp->lock);
1793 1794

	addrconf_mod_dad_work(ifp, 0);
1795
}
L
Linus Torvalds 已提交
1796

1797 1798
/* Join to solicited addr multicast group.
 * caller must hold RTNL */
1799
void addrconf_join_solict(struct net_device *dev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809
{
	struct in6_addr maddr;

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

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

1810
/* caller must hold RTNL */
1811
void addrconf_leave_solict(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
1812 1813 1814 1815 1816 1817 1818 1819 1820 1821
{
	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);
}

1822
/* caller must hold RTNL */
1823
static void addrconf_join_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
1824 1825
{
	struct in6_addr addr;
1826

1827
	if (ifp->prefix_len >= 127) /* RFC 6164 */
1828
		return;
L
Linus Torvalds 已提交
1829 1830 1831
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
W
WANG Cong 已提交
1832
	__ipv6_dev_ac_inc(ifp->idev, &addr);
L
Linus Torvalds 已提交
1833 1834
}

1835
/* caller must hold RTNL */
1836
static void addrconf_leave_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
1837 1838
{
	struct in6_addr addr;
1839

1840
	if (ifp->prefix_len >= 127) /* RFC 6164 */
1841
		return;
L
Linus Torvalds 已提交
1842 1843 1844 1845 1846 1847
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
	__ipv6_dev_ac_dec(ifp->idev, &addr);
}

1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878
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;
}

1879 1880 1881 1882 1883
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);
1884
	eui[0] ^= 2;
1885 1886 1887
	return 0;
}

1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901
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;
}

1902 1903 1904 1905 1906 1907
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);
1908
	eui[7] = *(u8 *)dev->dev_addr;
1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920
	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;
}

1921
static int __ipv6_isatap_ifid(u8 *eui, __be32 addr)
1922
{
1923 1924
	if (addr == 0)
		return -1;
1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944
	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;
}

1945 1946 1947 1948 1949
static int addrconf_ifid_gre(u8 *eui, struct net_device *dev)
{
	return __ipv6_isatap_ifid(eui, *(__be32 *)dev->dev_addr);
}

1950 1951 1952 1953 1954 1955 1956 1957 1958 1959
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
1960 1961 1962 1963 1964
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev)
{
	switch (dev->type) {
	case ARPHRD_ETHER:
	case ARPHRD_FDDI:
1965
		return addrconf_ifid_eui48(eui, dev);
L
Linus Torvalds 已提交
1966
	case ARPHRD_ARCNET:
1967
		return addrconf_ifid_arcnet(eui, dev);
L
Linus Torvalds 已提交
1968
	case ARPHRD_INFINIBAND:
1969
		return addrconf_ifid_infiniband(eui, dev);
F
Fred L. Templin 已提交
1970
	case ARPHRD_SIT:
1971
		return addrconf_ifid_sit(eui, dev);
1972 1973
	case ARPHRD_IPGRE:
		return addrconf_ifid_gre(eui, dev);
1974
	case ARPHRD_6LOWPAN:
1975 1976
	case ARPHRD_IEEE802154:
		return addrconf_ifid_eui64(eui, dev);
1977 1978
	case ARPHRD_IEEE1394:
		return addrconf_ifid_ieee1394(eui, dev);
1979 1980
	case ARPHRD_TUNNEL6:
		return addrconf_ifid_ip6tnl(eui, dev);
L
Linus Torvalds 已提交
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
	}
	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);
1991 1992 1993
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
L
Linus Torvalds 已提交
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
		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
2005
static void __ipv6_regen_rndid(struct inet6_dev *idev)
L
Linus Torvalds 已提交
2006 2007
{
regen:
2008
	get_random_bytes(idev->rndid, sizeof(idev->rndid));
L
Linus Torvalds 已提交
2009 2010 2011 2012 2013 2014 2015 2016
	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
2017
	 *  - ISATAP (RFC4214) 6.1
L
Linus Torvalds 已提交
2018 2019 2020 2021
	 *	00-00-5E-FE-xx-xx-xx-xx
	 *  - value 0
	 *  - XXX: already assigned to an address on the device
	 */
2022
	if (idev->rndid[0] == 0xfd &&
L
Linus Torvalds 已提交
2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038
	    (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;

2039
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
2040 2041 2042 2043 2044
	write_lock_bh(&idev->lock);

	if (idev->dead)
		goto out;

S
Sorin Dumitru 已提交
2045
	__ipv6_regen_rndid(idev);
2046

L
Linus Torvalds 已提交
2047
	expires = jiffies +
2048
		idev->cnf.temp_prefered_lft * HZ -
J
Jiri Pirko 已提交
2049 2050
		idev->cnf.regen_max_retry * idev->cnf.dad_transmits *
		NEIGH_VAR(idev->nd_parms, RETRANS_TIME) -
2051
		idev->cnf.max_desync_factor * HZ;
L
Linus Torvalds 已提交
2052
	if (time_before(expires, jiffies)) {
2053 2054
		pr_warn("%s: too short regeneration interval; timer disabled for %s\n",
			__func__, idev->dev->name);
L
Linus Torvalds 已提交
2055 2056 2057 2058 2059 2060 2061 2062
		goto out;
	}

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

out:
	write_unlock_bh(&idev->lock);
2063
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
2064 2065 2066
	in6_dev_put(idev);
}

S
Sorin Dumitru 已提交
2067
static void  __ipv6_try_regen_rndid(struct inet6_dev *idev, struct in6_addr *tmpaddr)
2068
{
L
Linus Torvalds 已提交
2069
	if (tmpaddr && memcmp(idev->rndid, &tmpaddr->s6_addr[8], 8) == 0)
S
Sorin Dumitru 已提交
2070
		__ipv6_regen_rndid(idev);
L
Linus Torvalds 已提交
2071 2072 2073 2074 2075 2076 2077 2078
}

/*
 *	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
2079
		      unsigned long expires, u32 flags)
L
Linus Torvalds 已提交
2080
{
2081 2082 2083 2084 2085 2086 2087
	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,
2088
		.fc_nlinfo.nl_net = dev_net(dev),
2089
		.fc_protocol = RTPROT_KERNEL,
2090
	};
L
Linus Torvalds 已提交
2091

A
Alexey Dobriyan 已提交
2092
	cfg.fc_dst = *pfx;
L
Linus Torvalds 已提交
2093 2094 2095 2096 2097

	/* 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
2098
#if IS_ENABLED(CONFIG_IPV6_SIT)
2099 2100
	if (dev->type == ARPHRD_SIT && (dev->flags & IFF_POINTOPOINT))
		cfg.fc_flags |= RTF_NONEXTHOP;
2101
#endif
L
Linus Torvalds 已提交
2102

2103
	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2104 2105
}

2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119

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;

2120
	read_lock_bh(&table->tb6_lock);
2121 2122 2123 2124
	fn = fib6_locate(&table->tb6_root, pfx, plen, NULL, 0);
	if (!fn)
		goto out;
	for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) {
2125
		if (rt->dst.dev->ifindex != dev->ifindex)
2126 2127 2128
			continue;
		if ((rt->rt6i_flags & flags) != flags)
			continue;
2129
		if ((rt->rt6i_flags & noflags) != 0)
2130 2131 2132 2133 2134
			continue;
		dst_hold(&rt->dst);
		break;
	}
out:
2135
	read_unlock_bh(&table->tb6_lock);
2136 2137 2138 2139
	return rt;
}


L
Linus Torvalds 已提交
2140 2141 2142 2143
/* Create "default" multicast route to the interface */

static void addrconf_add_mroute(struct net_device *dev)
{
2144 2145 2146 2147 2148 2149
	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,
2150
		.fc_nlinfo.nl_net = dev_net(dev),
2151 2152 2153 2154 2155
	};

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

	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2156 2157 2158 2159 2160 2161 2162 2163
}

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

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
2164 2165
	idev = ipv6_find_idev(dev);
	if (!idev)
2166 2167 2168 2169
		return ERR_PTR(-ENOBUFS);

	if (idev->cnf.disable_ipv6)
		return ERR_PTR(-EACCES);
L
Linus Torvalds 已提交
2170 2171

	/* Add default multicast route */
2172 2173
	if (!(dev->flags & IFF_LOOPBACK))
		addrconf_add_mroute(dev);
L
Linus Torvalds 已提交
2174 2175 2176 2177

	return idev;
}

2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244
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);
	}
}

2245
void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
L
Linus Torvalds 已提交
2246 2247 2248 2249 2250
{
	struct prefix_info *pinfo;
	__u32 valid_lft;
	__u32 prefered_lft;
	int addr_type;
2251
	u32 addr_flags = 0;
L
Linus Torvalds 已提交
2252
	struct inet6_dev *in6_dev;
2253
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
2254 2255

	pinfo = (struct prefix_info *) opt;
2256

L
Linus Torvalds 已提交
2257
	if (len < sizeof(struct prefix_info)) {
2258
		ADBG("addrconf: prefix option too short\n");
L
Linus Torvalds 已提交
2259 2260
		return;
	}
2261

L
Linus Torvalds 已提交
2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274
	/*
	 *	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) {
2275
		net_warn_ratelimited("addrconf: prefix option has invalid lifetime\n");
L
Linus Torvalds 已提交
2276 2277 2278 2279 2280 2281
		return;
	}

	in6_dev = in6_dev_get(dev);

	if (in6_dev == NULL) {
2282 2283
		net_dbg_ratelimited("addrconf: device %s not configured\n",
				    dev->name);
L
Linus Torvalds 已提交
2284 2285 2286 2287 2288 2289 2290 2291 2292
		return;
	}

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

2293 2294 2295 2296
	if (pinfo->onlink) {
		struct rt6_info *rt;
		unsigned long rt_expires;

2297 2298 2299 2300 2301
		/* 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.
		 */
2302 2303 2304 2305
		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
2306

2307 2308
		if (addrconf_finite_timeout(rt_expires))
			rt_expires *= HZ;
L
Linus Torvalds 已提交
2309

2310 2311 2312 2313 2314
		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
2315

2316
		if (rt) {
2317 2318 2319 2320
			/* Autoconf prefix route */
			if (valid_lft == 0) {
				ip6_del_rt(rt);
				rt = NULL;
2321
			} else if (addrconf_finite_timeout(rt_expires)) {
2322
				/* not infinity */
2323
				rt6_set_expires(rt, jiffies + rt_expires);
2324
			} else {
2325
				rt6_clean_expires(rt);
L
Linus Torvalds 已提交
2326 2327
			}
		} else if (valid_lft) {
2328
			clock_t expires = 0;
2329 2330
			int flags = RTF_ADDRCONF | RTF_PREFIX_RT;
			if (addrconf_finite_timeout(rt_expires)) {
2331 2332 2333 2334
				/* not infinity */
				flags |= RTF_EXPIRES;
				expires = jiffies_to_clock_t(rt_expires);
			}
L
Linus Torvalds 已提交
2335
			addrconf_prefix_route(&pinfo->prefix, pinfo->prefix_len,
2336
					      dev, expires, flags);
L
Linus Torvalds 已提交
2337
		}
A
Amerigo Wang 已提交
2338
		ip6_rt_put(rt);
L
Linus Torvalds 已提交
2339 2340 2341 2342 2343
	}

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

	if (pinfo->autoconf && in6_dev->cnf.autoconf) {
2344
		struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
2345 2346
		struct in6_addr addr;
		int create = 0, update_lft = 0;
2347
		bool tokenized = false;
L
Linus Torvalds 已提交
2348 2349 2350

		if (pinfo->prefix_len == 64) {
			memcpy(&addr, &pinfo->prefix, 8);
2351 2352 2353 2354 2355 2356

			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);
2357
				tokenized = true;
2358 2359 2360 2361
			} else if (in6_dev->addr_gen_mode ==
				   IN6_ADDR_GEN_MODE_STABLE_PRIVACY &&
				   !ipv6_generate_stable_address(&addr, 0,
								 in6_dev)) {
2362
				addr_flags |= IFA_F_STABLE_PRIVACY;
2363
				goto ok;
2364 2365
			} 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
2366 2367 2368 2369 2370
				in6_dev_put(in6_dev);
				return;
			}
			goto ok;
		}
2371 2372
		net_dbg_ratelimited("IPv6 addrconf: prefix with wrong length %d\n",
				    pinfo->prefix_len);
L
Linus Torvalds 已提交
2373 2374 2375 2376 2377
		in6_dev_put(in6_dev);
		return;

ok:

2378
		ifp = ipv6_get_ifaddr(net, &addr, dev, 1);
L
Linus Torvalds 已提交
2379 2380 2381

		if (ifp == NULL && valid_lft) {
			int max_addresses = in6_dev->cnf.max_addresses;
2382 2383 2384

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
			if (in6_dev->cnf.optimistic_dad &&
2385
			    !net->ipv6.devconf_all->forwarding && sllao)
2386 2387
				addr_flags = IFA_F_OPTIMISTIC;
#endif
L
Linus Torvalds 已提交
2388 2389 2390 2391 2392 2393

			/* 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)
2394 2395
				ifp = ipv6_add_addr(in6_dev, &addr, NULL,
						    pinfo->prefix_len,
2396
						    addr_type&IPV6_ADDR_SCOPE_MASK,
2397 2398
						    addr_flags, valid_lft,
						    prefered_lft);
L
Linus Torvalds 已提交
2399

2400
			if (IS_ERR_OR_NULL(ifp)) {
L
Linus Torvalds 已提交
2401 2402 2403 2404
				in6_dev_put(in6_dev);
				return;
			}

2405 2406
			update_lft = 0;
			create = 1;
2407 2408
			spin_lock_bh(&ifp->lock);
			ifp->flags |= IFA_F_MANAGETEMPADDR;
L
Linus Torvalds 已提交
2409
			ifp->cstamp = jiffies;
2410
			ifp->tokenized = tokenized;
2411
			spin_unlock_bh(&ifp->lock);
2412
			addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
2413 2414 2415
		}

		if (ifp) {
2416
			u32 flags;
L
Linus Torvalds 已提交
2417 2418 2419 2420
			unsigned long now;
			u32 stored_lft;

			/* update lifetime (RFC2462 5.5.3 e) */
2421
			spin_lock_bh(&ifp->lock);
L
Linus Torvalds 已提交
2422 2423 2424 2425 2426
			now = jiffies;
			if (ifp->valid_lft > (now - ifp->tstamp) / HZ)
				stored_lft = ifp->valid_lft - (now - ifp->tstamp) / HZ;
			else
				stored_lft = 0;
2427
			if (!update_lft && !create && stored_lft) {
2428 2429
				const u32 minimum_lft = min_t(u32,
					stored_lft, MIN_VALID_LIFETIME);
2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442
				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
2443 2444 2445 2446 2447 2448 2449 2450
			}

			if (update_lft) {
				ifp->valid_lft = valid_lft;
				ifp->prefered_lft = prefered_lft;
				ifp->tstamp = now;
				flags = ifp->flags;
				ifp->flags &= ~IFA_F_DEPRECATED;
2451
				spin_unlock_bh(&ifp->lock);
L
Linus Torvalds 已提交
2452 2453 2454 2455

				if (!(flags&IFA_F_TENTATIVE))
					ipv6_ifa_notify(0, ifp);
			} else
2456
				spin_unlock_bh(&ifp->lock);
L
Linus Torvalds 已提交
2457

2458 2459
			manage_tempaddrs(in6_dev, ifp, valid_lft, prefered_lft,
					 create, now);
2460

L
Linus Torvalds 已提交
2461
			in6_ifa_put(ifp);
2462
			addrconf_verify();
L
Linus Torvalds 已提交
2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473
		}
	}
	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.
 */
2474
int addrconf_set_dstaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485
{
	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;

2486
	dev = __dev_get_by_index(net, ireq.ifr6_ifindex);
L
Linus Torvalds 已提交
2487 2488 2489 2490 2491

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

A
Amerigo Wang 已提交
2492
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2493
	if (dev->type == ARPHRD_SIT) {
2494
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508
		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;
2509
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;
L
Linus Torvalds 已提交
2510

2511 2512 2513 2514 2515 2516 2517 2518
		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
2519 2520 2521

		if (err == 0) {
			err = -ENOBUFS;
2522 2523
			dev = __dev_get_by_name(net, p.name);
			if (!dev)
L
Linus Torvalds 已提交
2524 2525 2526 2527
				goto err_exit;
			err = dev_open(dev);
		}
	}
2528
#endif
L
Linus Torvalds 已提交
2529 2530 2531 2532 2533 2534

err_exit:
	rtnl_unlock();
	return err;
}

2535 2536 2537 2538 2539 2540 2541 2542 2543
static int ipv6_mc_config(struct sock *sk, bool join,
			  const struct in6_addr *addr, int ifindex)
{
	int ret;

	ASSERT_RTNL();

	lock_sock(sk);
	if (join)
2544
		ret = ipv6_sock_mc_join(sk, ifindex, addr);
2545
	else
2546
		ret = ipv6_sock_mc_drop(sk, ifindex, addr);
2547 2548 2549 2550 2551
	release_sock(sk);

	return ret;
}

L
Linus Torvalds 已提交
2552 2553 2554
/*
 *	Manual configuration of address on an interface
 */
2555 2556
static int inet6_addr_add(struct net *net, int ifindex,
			  const struct in6_addr *pfx,
2557
			  const struct in6_addr *peer_pfx,
2558 2559
			  unsigned int plen, __u32 ifa_flags,
			  __u32 prefered_lft, __u32 valid_lft)
L
Linus Torvalds 已提交
2560 2561 2562 2563
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2564 2565
	unsigned long timeout;
	clock_t expires;
L
Linus Torvalds 已提交
2566
	int scope;
2567
	u32 flags;
L
Linus Torvalds 已提交
2568 2569

	ASSERT_RTNL();
2570

2571 2572 2573
	if (plen > 128)
		return -EINVAL;

2574 2575 2576 2577
	/* check the lifetime */
	if (!valid_lft || prefered_lft > valid_lft)
		return -EINVAL;

2578 2579 2580
	if (ifa_flags & IFA_F_MANAGETEMPADDR && plen != 64)
		return -EINVAL;

2581 2582
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2583
		return -ENODEV;
2584

2585 2586 2587
	idev = addrconf_add_dev(dev);
	if (IS_ERR(idev))
		return PTR_ERR(idev);
L
Linus Torvalds 已提交
2588

2589 2590 2591 2592 2593 2594 2595 2596
	if (ifa_flags & IFA_F_MCAUTOJOIN) {
		int ret = ipv6_mc_config(net->ipv6.mc_autojoin_sk,
					 true, pfx, ifindex);

		if (ret < 0)
			return ret;
	}

L
Linus Torvalds 已提交
2597 2598
	scope = ipv6_addr_scope(pfx);

2599 2600 2601 2602
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
2603
		flags = RTF_EXPIRES;
2604 2605 2606 2607
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
2608
	}
2609

2610 2611 2612 2613 2614 2615
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
2616

2617 2618
	ifp = ipv6_add_addr(idev, pfx, peer_pfx, plen, scope, ifa_flags,
			    valid_lft, prefered_lft);
2619

L
Linus Torvalds 已提交
2620
	if (!IS_ERR(ifp)) {
2621 2622 2623 2624 2625
		if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
			addrconf_prefix_route(&ifp->addr, ifp->prefix_len, dev,
					      expires, flags);
		}

2626 2627 2628 2629 2630
		/*
		 * Note that section 3.1 of RFC 4429 indicates
		 * that the Optimistic flag should not be set for
		 * manually configured addresses
		 */
2631
		addrconf_dad_start(ifp);
2632 2633 2634
		if (ifa_flags & IFA_F_MANAGETEMPADDR)
			manage_tempaddrs(idev, ifp, valid_lft, prefered_lft,
					 true, jiffies);
L
Linus Torvalds 已提交
2635
		in6_ifa_put(ifp);
2636
		addrconf_verify_rtnl();
L
Linus Torvalds 已提交
2637
		return 0;
2638 2639 2640
	} else if (ifa_flags & IFA_F_MCAUTOJOIN) {
		ipv6_mc_config(net->ipv6.mc_autojoin_sk,
			       false, pfx, ifindex);
L
Linus Torvalds 已提交
2641 2642 2643 2644 2645
	}

	return PTR_ERR(ifp);
}

2646 2647
static int inet6_addr_del(struct net *net, int ifindex, u32 ifa_flags,
			  const struct in6_addr *pfx, unsigned int plen)
L
Linus Torvalds 已提交
2648 2649 2650 2651
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2652

2653 2654 2655
	if (plen > 128)
		return -EINVAL;

2656 2657
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2658 2659
		return -ENODEV;

2660 2661
	idev = __in6_dev_get(dev);
	if (idev == NULL)
L
Linus Torvalds 已提交
2662 2663 2664
		return -ENXIO;

	read_lock_bh(&idev->lock);
2665
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
L
Linus Torvalds 已提交
2666 2667 2668 2669
		if (ifp->prefix_len == plen &&
		    ipv6_addr_equal(pfx, &ifp->addr)) {
			in6_ifa_hold(ifp);
			read_unlock_bh(&idev->lock);
2670

2671 2672 2673 2674
			if (!(ifp->flags & IFA_F_TEMPORARY) &&
			    (ifa_flags & IFA_F_MANAGETEMPADDR))
				manage_tempaddrs(idev, ifp, 0, 0, false,
						 jiffies);
L
Linus Torvalds 已提交
2675
			ipv6_del_addr(ifp);
2676
			addrconf_verify_rtnl();
2677 2678 2679 2680
			if (ipv6_addr_is_multicast(pfx)) {
				ipv6_mc_config(net->ipv6.mc_autojoin_sk,
					       false, pfx, dev->ifindex);
			}
L
Linus Torvalds 已提交
2681 2682 2683 2684 2685 2686 2687 2688
			return 0;
		}
	}
	read_unlock_bh(&idev->lock);
	return -EADDRNOTAVAIL;
}


2689
int addrconf_add_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2690 2691 2692
{
	struct in6_ifreq ireq;
	int err;
2693

2694
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2695
		return -EPERM;
2696

L
Linus Torvalds 已提交
2697 2698 2699 2700
	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		return -EFAULT;

	rtnl_lock();
2701
	err = inet6_addr_add(net, ireq.ifr6_ifindex, &ireq.ifr6_addr, NULL,
2702 2703
			     ireq.ifr6_prefixlen, IFA_F_PERMANENT,
			     INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
2704 2705 2706 2707
	rtnl_unlock();
	return err;
}

2708
int addrconf_del_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2709 2710 2711
{
	struct in6_ifreq ireq;
	int err;
2712

2713
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2714 2715 2716 2717 2718 2719
		return -EPERM;

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

	rtnl_lock();
2720
	err = inet6_addr_del(net, ireq.ifr6_ifindex, 0, &ireq.ifr6_addr,
2721
			     ireq.ifr6_prefixlen);
L
Linus Torvalds 已提交
2722 2723 2724 2725
	rtnl_unlock();
	return err;
}

2726 2727 2728 2729 2730
static void add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
		     int plen, int scope)
{
	struct inet6_ifaddr *ifp;

2731
	ifp = ipv6_add_addr(idev, addr, NULL, plen,
2732 2733
			    scope, IFA_F_PERMANENT,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
2734 2735 2736 2737 2738 2739 2740 2741 2742
	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
2743
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2744 2745 2746 2747
static void sit_add_v4_addrs(struct inet6_dev *idev)
{
	struct in6_addr addr;
	struct net_device *dev;
2748
	struct net *net = dev_net(idev->dev);
2749
	int scope, plen;
2750
	u32 pflags = 0;
L
Linus Torvalds 已提交
2751 2752 2753 2754 2755 2756 2757 2758 2759

	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;
2760
		plen = 64;
L
Linus Torvalds 已提交
2761 2762
	} else {
		scope = IPV6_ADDR_COMPATv4;
2763
		plen = 96;
2764
		pflags |= RTF_NONEXTHOP;
L
Linus Torvalds 已提交
2765 2766 2767
	}

	if (addr.s6_addr32[3]) {
2768
		add_addr(idev, &addr, plen, scope);
2769
		addrconf_prefix_route(&addr, plen, idev->dev, 0, pflags);
L
Linus Torvalds 已提交
2770 2771 2772
		return;
	}

2773
	for_each_netdev(net, dev) {
2774
		struct in_device *in_dev = __in_dev_get_rtnl(dev);
L
Linus Torvalds 已提交
2775
		if (in_dev && (dev->flags & IFF_UP)) {
2776
			struct in_ifaddr *ifa;
L
Linus Torvalds 已提交
2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791

			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;
				}

2792
				add_addr(idev, &addr, plen, flag);
2793 2794
				addrconf_prefix_route(&addr, plen, idev->dev, 0,
						      pflags);
L
Linus Torvalds 已提交
2795 2796
			}
		}
2797
	}
L
Linus Torvalds 已提交
2798
}
2799
#endif
L
Linus Torvalds 已提交
2800 2801 2802 2803

static void init_loopback(struct net_device *dev)
{
	struct inet6_dev  *idev;
2804 2805 2806
	struct net_device *sp_dev;
	struct inet6_ifaddr *sp_ifa;
	struct rt6_info *sp_rt;
L
Linus Torvalds 已提交
2807 2808 2809 2810 2811

	/* ::1 */

	ASSERT_RTNL();

2812 2813
	idev = ipv6_find_idev(dev);
	if (idev == NULL) {
2814
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
2815 2816 2817
		return;
	}

2818
	add_addr(idev, &in6addr_loopback, 128, IFA_HOST);
2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834

	/* 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;

2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846
			if (sp_ifa->rt) {
				/* This dst has been added to garbage list when
				 * lo device down, release this obsolete dst and
				 * reallocate a new router for ifa.
				 */
				if (sp_ifa->rt->dst.obsolete > 0) {
					ip6_rt_put(sp_ifa->rt);
					sp_ifa->rt = NULL;
				} else {
					continue;
				}
			}
2847

2848
			sp_rt = addrconf_dst_alloc(idev, &sp_ifa->addr, false);
2849 2850

			/* Failure cases are ignored */
2851 2852
			if (!IS_ERR(sp_rt)) {
				sp_ifa->rt = sp_rt;
2853
				ip6_ins_rt(sp_rt);
2854
			}
2855 2856 2857
		}
		read_unlock_bh(&idev->lock);
	}
L
Linus Torvalds 已提交
2858 2859
}

2860 2861
static void addrconf_add_linklocal(struct inet6_dev *idev,
				   const struct in6_addr *addr, u32 flags)
L
Linus Torvalds 已提交
2862
{
2863
	struct inet6_ifaddr *ifp;
2864
	u32 addr_flags = flags | IFA_F_PERMANENT;
2865 2866 2867

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	if (idev->cnf.optimistic_dad &&
2868
	    !dev_net(idev->dev)->ipv6.devconf_all->forwarding)
2869 2870
		addr_flags |= IFA_F_OPTIMISTIC;
#endif
L
Linus Torvalds 已提交
2871

2872 2873
	ifp = ipv6_add_addr(idev, addr, NULL, 64, IFA_LINK, addr_flags,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
2874
	if (!IS_ERR(ifp)) {
2875
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, idev->dev, 0, 0);
2876
		addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
2877 2878 2879 2880
		in6_ifa_put(ifp);
	}
}

2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949
static bool ipv6_reserved_interfaceid(struct in6_addr address)
{
	if ((address.s6_addr32[2] | address.s6_addr32[3]) == 0)
		return true;

	if (address.s6_addr32[2] == htonl(0x02005eff) &&
	    ((address.s6_addr32[3] & htonl(0xfe000000)) == htonl(0xfe000000)))
		return true;

	if (address.s6_addr32[2] == htonl(0xfdffffff) &&
	    ((address.s6_addr32[3] & htonl(0xffffff80)) == htonl(0xffffff80)))
		return true;

	return false;
}

static int ipv6_generate_stable_address(struct in6_addr *address,
					u8 dad_count,
					const struct inet6_dev *idev)
{
	static DEFINE_SPINLOCK(lock);
	static __u32 digest[SHA_DIGEST_WORDS];
	static __u32 workspace[SHA_WORKSPACE_WORDS];

	static union {
		char __data[SHA_MESSAGE_BYTES];
		struct {
			struct in6_addr secret;
			__be64 prefix;
			unsigned char hwaddr[MAX_ADDR_LEN];
			u8 dad_count;
		} __packed;
	} data;

	struct in6_addr secret;
	struct in6_addr temp;
	struct net *net = dev_net(idev->dev);

	BUILD_BUG_ON(sizeof(data.__data) != sizeof(data));

	if (idev->cnf.stable_secret.initialized)
		secret = idev->cnf.stable_secret.secret;
	else if (net->ipv6.devconf_dflt->stable_secret.initialized)
		secret = net->ipv6.devconf_dflt->stable_secret.secret;
	else
		return -1;

retry:
	spin_lock_bh(&lock);

	sha_init(digest);
	memset(&data, 0, sizeof(data));
	memset(workspace, 0, sizeof(workspace));
	memcpy(data.hwaddr, idev->dev->perm_addr, idev->dev->addr_len);
	data.prefix = ((__be64)address->s6_addr32[0] << 32) |
		       (__be64)address->s6_addr32[1];
	data.secret = secret;
	data.dad_count = dad_count;

	sha_transform(digest, data.__data, workspace);

	temp = *address;
	temp.s6_addr32[2] = digest[0];
	temp.s6_addr32[3] = digest[1];

	spin_unlock_bh(&lock);

	if (ipv6_reserved_interfaceid(temp)) {
		dad_count++;
2950
		if (dad_count > dev_net(idev->dev)->ipv6.sysctl.idgen_retries)
2951 2952 2953 2954 2955 2956 2957 2958
			return -1;
		goto retry;
	}

	*address = temp;
	return 0;
}

2959 2960
static void addrconf_addr_gen(struct inet6_dev *idev, bool prefix_route)
{
2961 2962 2963
	struct in6_addr addr;

	ipv6_addr_set(&addr, htonl(0xFE800000), 0, 0, 0);
2964

2965 2966
	if (idev->addr_gen_mode == IN6_ADDR_GEN_MODE_STABLE_PRIVACY) {
		if (!ipv6_generate_stable_address(&addr, 0, idev))
2967 2968
			addrconf_add_linklocal(idev, &addr,
					       IFA_F_STABLE_PRIVACY);
2969 2970 2971
		else if (prefix_route)
			addrconf_prefix_route(&addr, 64, idev->dev, 0, 0);
	} else if (idev->addr_gen_mode == IN6_ADDR_GEN_MODE_EUI64) {
2972 2973 2974 2975 2976
		/* addrconf_add_linklocal also adds a prefix_route and we
		 * only need to care about prefix routes if ipv6_generate_eui64
		 * couldn't generate one.
		 */
		if (ipv6_generate_eui64(addr.s6_addr + 8, idev->dev) == 0)
2977
			addrconf_add_linklocal(idev, &addr, 0);
2978 2979 2980 2981 2982
		else if (prefix_route)
			addrconf_prefix_route(&addr, 64, idev->dev, 0, 0);
	}
}

L
Linus Torvalds 已提交
2983 2984
static void addrconf_dev_config(struct net_device *dev)
{
2985
	struct inet6_dev *idev;
L
Linus Torvalds 已提交
2986 2987 2988

	ASSERT_RTNL();

2989 2990 2991
	if ((dev->type != ARPHRD_ETHER) &&
	    (dev->type != ARPHRD_FDDI) &&
	    (dev->type != ARPHRD_ARCNET) &&
2992
	    (dev->type != ARPHRD_INFINIBAND) &&
2993
	    (dev->type != ARPHRD_IEEE802154) &&
2994
	    (dev->type != ARPHRD_IEEE1394) &&
2995 2996
	    (dev->type != ARPHRD_TUNNEL6) &&
	    (dev->type != ARPHRD_6LOWPAN)) {
2997 2998 2999 3000
		/* Alas, we support only Ethernet autoconfiguration. */
		return;
	}

L
Linus Torvalds 已提交
3001
	idev = addrconf_add_dev(dev);
3002
	if (IS_ERR(idev))
L
Linus Torvalds 已提交
3003 3004
		return;

3005
	addrconf_addr_gen(idev, false);
L
Linus Torvalds 已提交
3006 3007
}

A
Amerigo Wang 已提交
3008
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3009 3010 3011 3012 3013 3014
static void addrconf_sit_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3015 3016 3017
	/*
	 * Configure the tunnel with one of our IPv4
	 * addresses... we should configure all of
L
Linus Torvalds 已提交
3018 3019 3020
	 * our v4 addrs in the tunnel
	 */

3021 3022
	idev = ipv6_find_idev(dev);
	if (idev == NULL) {
3023
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
3024 3025 3026
		return;
	}

F
Fred L. Templin 已提交
3027
	if (dev->priv_flags & IFF_ISATAP) {
3028
		addrconf_addr_gen(idev, false);
F
Fred L. Templin 已提交
3029 3030 3031
		return;
	}

L
Linus Torvalds 已提交
3032 3033
	sit_add_v4_addrs(idev);

3034
	if (dev->flags&IFF_POINTOPOINT)
L
Linus Torvalds 已提交
3035 3036
		addrconf_add_mroute(dev);
}
3037
#endif
L
Linus Torvalds 已提交
3038

A
Amerigo Wang 已提交
3039
#if IS_ENABLED(CONFIG_NET_IPGRE)
3040 3041 3042 3043 3044 3045
static void addrconf_gre_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3046 3047
	idev = ipv6_find_idev(dev);
	if (idev == NULL) {
3048
		pr_debug("%s: add_dev failed\n", __func__);
3049 3050 3051
		return;
	}

3052
	addrconf_addr_gen(idev, true);
3053 3054 3055
}
#endif

3056
static int addrconf_notify(struct notifier_block *this, unsigned long event,
3057
			   void *ptr)
L
Linus Torvalds 已提交
3058
{
3059
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
3060
	struct inet6_dev *idev = __in6_dev_get(dev);
3061
	int run_pending = 0;
3062
	int err;
L
Linus Torvalds 已提交
3063

S
Stephen Hemminger 已提交
3064
	switch (event) {
3065
	case NETDEV_REGISTER:
3066
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
3067
			idev = ipv6_add_dev(dev);
3068 3069
			if (IS_ERR(idev))
				return notifier_from_errno(PTR_ERR(idev));
3070 3071
		}
		break;
S
Stephen Hemminger 已提交
3072

L
Linus Torvalds 已提交
3073
	case NETDEV_UP:
3074
	case NETDEV_CHANGE:
3075 3076 3077
		if (dev->flags & IFF_SLAVE)
			break;

3078 3079 3080
		if (idev && idev->cnf.disable_ipv6)
			break;

3081
		if (event == NETDEV_UP) {
3082
			if (!addrconf_qdisc_ok(dev)) {
3083
				/* device is not ready yet. */
3084
				pr_info("ADDRCONF(NETDEV_UP): %s: link is not ready\n",
3085 3086 3087
					dev->name);
				break;
			}
3088

3089 3090 3091
			if (!idev && dev->mtu >= IPV6_MIN_MTU)
				idev = ipv6_add_dev(dev);

3092
			if (!IS_ERR_OR_NULL(idev)) {
3093
				idev->if_flags |= IF_READY;
3094 3095
				run_pending = 1;
			}
3096
		} else {
3097
			if (!addrconf_qdisc_ok(dev)) {
3098 3099 3100 3101 3102
				/* device is still not ready. */
				break;
			}

			if (idev) {
S
Stephen Hemminger 已提交
3103
				if (idev->if_flags & IF_READY)
3104 3105 3106 3107 3108
					/* device is already configured. */
					break;
				idev->if_flags |= IF_READY;
			}

3109 3110
			pr_info("ADDRCONF(NETDEV_CHANGE): %s: link becomes ready\n",
				dev->name);
3111

3112
			run_pending = 1;
3113 3114
		}

S
Stephen Hemminger 已提交
3115
		switch (dev->type) {
A
Amerigo Wang 已提交
3116
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3117 3118 3119
		case ARPHRD_SIT:
			addrconf_sit_config(dev);
			break;
3120
#endif
A
Amerigo Wang 已提交
3121
#if IS_ENABLED(CONFIG_NET_IPGRE)
3122 3123 3124
		case ARPHRD_IPGRE:
			addrconf_gre_config(dev);
			break;
3125
#endif
L
Linus Torvalds 已提交
3126 3127 3128 3129 3130 3131 3132
		case ARPHRD_LOOPBACK:
			init_loopback(dev);
			break;

		default:
			addrconf_dev_config(dev);
			break;
3133
		}
S
Stephen Hemminger 已提交
3134

3135
		if (!IS_ERR_OR_NULL(idev)) {
3136 3137 3138
			if (run_pending)
				addrconf_dad_run(idev);

S
Stephen Hemminger 已提交
3139 3140 3141 3142
			/*
			 * 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
3143
			 */
S
Stephen Hemminger 已提交
3144 3145
			if (idev->cnf.mtu6 != dev->mtu &&
			    dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
3146 3147 3148 3149 3150
				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
3151 3152 3153 3154

			/*
			 * If the changed mtu during down is lower than
			 * IPV6_MIN_MTU stop IPv6 on this interface.
L
Linus Torvalds 已提交
3155 3156
			 */
			if (dev->mtu < IPV6_MIN_MTU)
S
Stephen Hemminger 已提交
3157
				addrconf_ifdown(dev, 1);
L
Linus Torvalds 已提交
3158 3159 3160 3161
		}
		break;

	case NETDEV_CHANGEMTU:
3162
		if (idev && dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
3163 3164 3165 3166 3167
			rt6_mtu_change(dev, dev->mtu);
			idev->cnf.mtu6 = dev->mtu;
			break;
		}

3168 3169
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
			idev = ipv6_add_dev(dev);
3170
			if (!IS_ERR(idev))
3171 3172 3173
				break;
		}

S
Stephen Hemminger 已提交
3174
		/*
3175
		 * if MTU under IPV6_MIN_MTU.
S
Stephen Hemminger 已提交
3176 3177
		 * Stop IPv6 on this interface.
		 */
L
Linus Torvalds 已提交
3178 3179 3180 3181 3182 3183 3184 3185

	case NETDEV_DOWN:
	case NETDEV_UNREGISTER:
		/*
		 *	Remove all addresses from this interface.
		 */
		addrconf_ifdown(dev, event != NETDEV_DOWN);
		break;
3186

L
Linus Torvalds 已提交
3187 3188
	case NETDEV_CHANGENAME:
		if (idev) {
3189
			snmp6_unregister_dev(idev);
3190
			addrconf_sysctl_unregister(idev);
3191
			err = addrconf_sysctl_register(idev);
3192 3193
			if (err)
				return notifier_from_errno(err);
3194 3195 3196 3197 3198
			err = snmp6_register_dev(idev);
			if (err) {
				addrconf_sysctl_unregister(idev);
				return notifier_from_errno(err);
			}
3199
		}
L
Linus Torvalds 已提交
3200
		break;
S
Stephen Hemminger 已提交
3201

3202 3203 3204
	case NETDEV_PRE_TYPE_CHANGE:
	case NETDEV_POST_TYPE_CHANGE:
		addrconf_type_change(dev, event);
3205
		break;
3206
	}
L
Linus Torvalds 已提交
3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217

	return NOTIFY_OK;
}

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

3218
static void addrconf_type_change(struct net_device *dev, unsigned long event)
3219 3220 3221 3222 3223 3224
{
	struct inet6_dev *idev;
	ASSERT_RTNL();

	idev = __in6_dev_get(dev);

3225
	if (event == NETDEV_POST_TYPE_CHANGE)
3226
		ipv6_mc_remap(idev);
3227
	else if (event == NETDEV_PRE_TYPE_CHANGE)
3228 3229 3230
		ipv6_mc_unmap(idev);
}

L
Linus Torvalds 已提交
3231 3232
static int addrconf_ifdown(struct net_device *dev, int how)
{
3233
	struct net *net = dev_net(dev);
3234 3235
	struct inet6_dev *idev;
	struct inet6_ifaddr *ifa;
3236
	int state, i;
L
Linus Torvalds 已提交
3237 3238 3239

	ASSERT_RTNL();

3240 3241
	rt6_ifdown(net, dev);
	neigh_ifdown(&nd_tbl, dev);
L
Linus Torvalds 已提交
3242 3243 3244 3245 3246

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

S
Stephen Hemminger 已提交
3247 3248 3249
	/*
	 * Step 1: remove reference to ipv6 device from parent device.
	 *	   Do not dev_put!
L
Linus Torvalds 已提交
3250
	 */
3251
	if (how) {
L
Linus Torvalds 已提交
3252
		idev->dead = 1;
3253 3254

		/* protected by rtnl_lock */
3255
		RCU_INIT_POINTER(dev->ip6_ptr, NULL);
L
Linus Torvalds 已提交
3256 3257 3258 3259 3260 3261

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

	}

3262 3263 3264 3265 3266
	/* 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);
3267
restart:
3268
		hlist_for_each_entry_rcu(ifa, h, addr_lst) {
3269 3270
			if (ifa->idev == idev) {
				hlist_del_init_rcu(&ifa->addr_lst);
3271
				addrconf_del_dad_work(ifa);
3272 3273 3274 3275 3276 3277
				goto restart;
			}
		}
		spin_unlock_bh(&addrconf_hash_lock);
	}

L
Linus Torvalds 已提交
3278 3279
	write_lock_bh(&idev->lock);

3280 3281
	addrconf_del_rs_timer(idev);

S
Stephen Hemminger 已提交
3282
	/* Step 2: clear flags for stateless addrconf */
3283
	if (!how)
3284
		idev->if_flags &= ~(IF_RS_SENT|IF_RA_RCVD|IF_READY);
L
Linus Torvalds 已提交
3285

3286
	if (how && del_timer(&idev->regen_timer))
L
Linus Torvalds 已提交
3287 3288
		in6_dev_put(idev);

S
Stephen Hemminger 已提交
3289
	/* Step 3: clear tempaddr list */
3290 3291 3292 3293
	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
3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304
		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);
	}
3305

3306 3307 3308
	while (!list_empty(&idev->addr_list)) {
		ifa = list_first_entry(&idev->addr_list,
				       struct inet6_ifaddr, if_list);
3309
		addrconf_del_dad_work(ifa);
3310

3311
		list_del(&ifa->if_list);
3312

3313
		write_unlock_bh(&idev->lock);
3314

3315
		spin_lock_bh(&ifa->lock);
3316 3317
		state = ifa->state;
		ifa->state = INET6_IFADDR_STATE_DEAD;
3318
		spin_unlock_bh(&ifa->lock);
3319

3320 3321
		if (state != INET6_IFADDR_STATE_DEAD) {
			__ipv6_ifa_notify(RTM_DELADDR, ifa);
3322
			inet6addr_notifier_call_chain(NETDEV_DOWN, ifa);
3323
		}
3324
		in6_ifa_put(ifa);
3325

3326 3327
		write_lock_bh(&idev->lock);
	}
3328

L
Linus Torvalds 已提交
3329 3330
	write_unlock_bh(&idev->lock);

3331 3332 3333
	/* Step 5: Discard anycast and multicast list */
	if (how) {
		ipv6_ac_destroy_dev(idev);
L
Linus Torvalds 已提交
3334
		ipv6_mc_destroy_dev(idev);
3335
	} else {
L
Linus Torvalds 已提交
3336
		ipv6_mc_down(idev);
3337
	}
L
Linus Torvalds 已提交
3338 3339

	idev->tstamp = jiffies;
3340

S
Stephen Hemminger 已提交
3341
	/* Last: Shot the device (if unregistered) */
3342
	if (how) {
3343
		addrconf_sysctl_unregister(idev);
L
Linus Torvalds 已提交
3344 3345 3346 3347 3348 3349 3350 3351 3352
		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)
{
3353
	struct inet6_dev *idev = (struct inet6_dev *)data;
3354
	struct net_device *dev = idev->dev;
3355
	struct in6_addr lladdr;
L
Linus Torvalds 已提交
3356

3357
	write_lock(&idev->lock);
S
stephen hemminger 已提交
3358
	if (idev->dead || !(idev->if_flags & IF_READY))
L
Linus Torvalds 已提交
3359 3360
		goto out;

3361
	if (!ipv6_accept_ra(idev))
S
stephen hemminger 已提交
3362 3363 3364 3365
		goto out;

	/* Announcement received after solicitation was sent */
	if (idev->if_flags & IF_RA_RCVD)
L
Linus Torvalds 已提交
3366 3367
		goto out;

3368
	if (idev->rs_probes++ < idev->cnf.rtr_solicits) {
3369 3370 3371
		write_unlock(&idev->lock);
		if (!ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
			ndisc_send_rs(dev, &lladdr,
3372 3373
				      &in6addr_linklocal_allrouters);
		else
3374
			goto put;
L
Linus Torvalds 已提交
3375

3376
		write_lock(&idev->lock);
3377 3378 3379 3380 3381
		/* 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
3382 3383 3384 3385 3386
	} else {
		/*
		 * Note: we do not support deprecated "all on-link"
		 * assumption any longer.
		 */
3387
		pr_debug("%s: no IPv6 routers present\n", idev->dev->name);
L
Linus Torvalds 已提交
3388 3389 3390
	}

out:
3391
	write_unlock(&idev->lock);
3392
put:
3393
	in6_dev_put(idev);
L
Linus Torvalds 已提交
3394 3395 3396 3397 3398
}

/*
 *	Duplicate Address Detection
 */
3399 3400 3401 3402 3403
static void addrconf_dad_kick(struct inet6_ifaddr *ifp)
{
	unsigned long rand_num;
	struct inet6_dev *idev = ifp->idev;

3404 3405 3406
	if (ifp->flags & IFA_F_OPTIMISTIC)
		rand_num = 0;
	else
3407
		rand_num = prandom_u32() % (idev->cnf.rtr_solicit_delay ? : 1);
3408

3409
	ifp->dad_probes = idev->cnf.dad_transmits;
3410
	addrconf_mod_dad_work(ifp, rand_num);
3411 3412
}

3413
static void addrconf_dad_begin(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3414 3415 3416 3417 3418 3419
{
	struct inet6_dev *idev = ifp->idev;
	struct net_device *dev = idev->dev;

	addrconf_join_solict(dev, &ifp->addr);

3420
	prandom_seed((__force u32) ifp->addr.s6_addr32[3]);
L
Linus Torvalds 已提交
3421 3422

	read_lock_bh(&idev->lock);
3423
	spin_lock(&ifp->lock);
3424
	if (ifp->state == INET6_IFADDR_STATE_DEAD)
L
Linus Torvalds 已提交
3425 3426 3427
		goto out;

	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
3428
	    idev->cnf.accept_dad < 1 ||
3429 3430
	    !(ifp->flags&IFA_F_TENTATIVE) ||
	    ifp->flags & IFA_F_NODAD) {
B
Brian Haley 已提交
3431
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3432
		spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3433 3434 3435 3436 3437 3438
		read_unlock_bh(&idev->lock);

		addrconf_dad_completed(ifp);
		return;
	}

3439
	if (!(idev->if_flags & IF_READY)) {
3440
		spin_unlock(&ifp->lock);
3441
		read_unlock_bh(&idev->lock);
3442
		/*
3443
		 * If the device is not ready:
3444 3445 3446 3447
		 * - keep it tentative if it is a permanent address.
		 * - otherwise, kill it.
		 */
		in6_ifa_hold(ifp);
B
Brian Haley 已提交
3448
		addrconf_dad_stop(ifp, 0);
3449
		return;
3450
	}
3451 3452 3453 3454 3455

	/*
	 * Optimistic nodes can start receiving
	 * Frames right away
	 */
3456
	if (ifp->flags & IFA_F_OPTIMISTIC) {
3457
		ip6_ins_rt(ifp->rt);
3458 3459 3460 3461 3462 3463 3464
		if (ipv6_use_optimistic_addr(idev)) {
			/* Because optimistic nodes can use this address,
			 * notify listeners. If DAD fails, RTM_DELADDR is sent.
			 */
			ipv6_ifa_notify(RTM_NEWADDR, ifp);
		}
	}
3465

3466
	addrconf_dad_kick(ifp);
L
Linus Torvalds 已提交
3467
out:
3468
	spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3469 3470 3471
	read_unlock_bh(&idev->lock);
}

3472
static void addrconf_dad_start(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3473
{
3474 3475
	bool begin_dad = false;

3476
	spin_lock_bh(&ifp->lock);
3477 3478 3479 3480
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
		ifp->state = INET6_IFADDR_STATE_PREDAD;
		begin_dad = true;
	}
3481
	spin_unlock_bh(&ifp->lock);
3482 3483 3484 3485 3486 3487 3488 3489 3490 3491

	if (begin_dad)
		addrconf_mod_dad_work(ifp, 0);
}

static void addrconf_dad_work(struct work_struct *w)
{
	struct inet6_ifaddr *ifp = container_of(to_delayed_work(w),
						struct inet6_ifaddr,
						dad_work);
L
Linus Torvalds 已提交
3492 3493 3494
	struct inet6_dev *idev = ifp->idev;
	struct in6_addr mcaddr;

3495 3496 3497 3498 3499 3500 3501 3502
	enum {
		DAD_PROCESS,
		DAD_BEGIN,
		DAD_ABORT,
	} action = DAD_PROCESS;

	rtnl_lock();

3503
	spin_lock_bh(&ifp->lock);
3504 3505 3506 3507 3508 3509 3510
	if (ifp->state == INET6_IFADDR_STATE_PREDAD) {
		action = DAD_BEGIN;
		ifp->state = INET6_IFADDR_STATE_DAD;
	} else if (ifp->state == INET6_IFADDR_STATE_ERRDAD) {
		action = DAD_ABORT;
		ifp->state = INET6_IFADDR_STATE_POSTDAD;
	}
3511
	spin_unlock_bh(&ifp->lock);
3512 3513 3514 3515 3516 3517 3518 3519 3520

	if (action == DAD_BEGIN) {
		addrconf_dad_begin(ifp);
		goto out;
	} else if (action == DAD_ABORT) {
		addrconf_dad_stop(ifp, 1);
		goto out;
	}

3521
	if (!ifp->dad_probes && addrconf_dad_end(ifp))
H
Herbert Xu 已提交
3522 3523
		goto out;

3524
	write_lock_bh(&idev->lock);
3525
	if (idev->dead || !(idev->if_flags & IF_READY)) {
3526
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3527 3528
		goto out;
	}
3529 3530

	spin_lock(&ifp->lock);
3531 3532
	if (ifp->state == INET6_IFADDR_STATE_DEAD) {
		spin_unlock(&ifp->lock);
3533
		write_unlock_bh(&idev->lock);
3534 3535 3536
		goto out;
	}

3537
	if (ifp->dad_probes == 0) {
L
Linus Torvalds 已提交
3538 3539 3540 3541
		/*
		 * DAD was successful
		 */

B
Brian Haley 已提交
3542
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3543
		spin_unlock(&ifp->lock);
3544
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3545 3546 3547 3548 3549 3550

		addrconf_dad_completed(ifp);

		goto out;
	}

3551
	ifp->dad_probes--;
3552 3553
	addrconf_mod_dad_work(ifp,
			      NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME));
3554
	spin_unlock(&ifp->lock);
3555
	write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3556 3557 3558

	/* send a neighbour solicitation for our addr */
	addrconf_addr_solict_mult(&ifp->addr, &mcaddr);
3559
	ndisc_send_ns(ifp->idev->dev, NULL, &ifp->addr, &mcaddr, &in6addr_any);
L
Linus Torvalds 已提交
3560 3561
out:
	in6_ifa_put(ifp);
3562
	rtnl_unlock();
L
Linus Torvalds 已提交
3563 3564
}

3565 3566 3567 3568 3569 3570
/* ifp->idev must be at least read locked */
static bool ipv6_lonely_lladdr(struct inet6_ifaddr *ifp)
{
	struct inet6_ifaddr *ifpiter;
	struct inet6_dev *idev = ifp->idev;

3571 3572 3573
	list_for_each_entry_reverse(ifpiter, &idev->addr_list, if_list) {
		if (ifpiter->scope > IFA_LINK)
			break;
3574 3575 3576 3577 3578 3579 3580 3581 3582
		if (ifp != ifpiter && ifpiter->scope == IFA_LINK &&
		    (ifpiter->flags & (IFA_F_PERMANENT|IFA_F_TENTATIVE|
				       IFA_F_OPTIMISTIC|IFA_F_DADFAILED)) ==
		    IFA_F_PERMANENT)
			return false;
	}
	return true;
}

L
Linus Torvalds 已提交
3583 3584
static void addrconf_dad_completed(struct inet6_ifaddr *ifp)
{
S
Stephen Hemminger 已提交
3585
	struct net_device *dev = ifp->idev->dev;
3586
	struct in6_addr lladdr;
3587
	bool send_rs, send_mld;
3588

3589
	addrconf_del_dad_work(ifp);
L
Linus Torvalds 已提交
3590 3591 3592 3593 3594 3595 3596

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

	ipv6_ifa_notify(RTM_NEWADDR, ifp);

3597 3598
	/* If added prefix is link local and we are prepared to process
	   router advertisements, start sending router solicitations.
L
Linus Torvalds 已提交
3599 3600
	 */

3601
	read_lock_bh(&ifp->idev->lock);
3602
	send_mld = ifp->scope == IFA_LINK && ipv6_lonely_lladdr(ifp);
3603 3604
	send_rs = send_mld &&
		  ipv6_accept_ra(ifp->idev) &&
3605
		  ifp->idev->cnf.rtr_solicits > 0 &&
3606
		  (dev->flags&IFF_LOOPBACK) == 0;
3607 3608
	read_unlock_bh(&ifp->idev->lock);

3609 3610 3611 3612 3613 3614
	/* 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);

3615
	if (send_rs) {
L
Linus Torvalds 已提交
3616 3617 3618 3619 3620
		/*
		 *	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
		 */
3621
		if (ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
3622
			return;
3623
		ndisc_send_rs(dev, &lladdr, &in6addr_linklocal_allrouters);
L
Linus Torvalds 已提交
3624

3625 3626 3627
		write_lock_bh(&ifp->idev->lock);
		spin_lock(&ifp->lock);
		ifp->idev->rs_probes = 1;
L
Linus Torvalds 已提交
3628
		ifp->idev->if_flags |= IF_RS_SENT;
3629 3630 3631 3632
		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
3633 3634 3635
	}
}

S
Stephen Hemminger 已提交
3636 3637
static void addrconf_dad_run(struct inet6_dev *idev)
{
3638 3639 3640
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
3641
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
3642
		spin_lock(&ifp->lock);
H
Herbert Xu 已提交
3643 3644 3645
		if (ifp->flags & IFA_F_TENTATIVE &&
		    ifp->state == INET6_IFADDR_STATE_DAD)
			addrconf_dad_kick(ifp);
3646
		spin_unlock(&ifp->lock);
3647 3648 3649 3650
	}
	read_unlock_bh(&idev->lock);
}

L
Linus Torvalds 已提交
3651 3652
#ifdef CONFIG_PROC_FS
struct if6_iter_state {
3653
	struct seq_net_private p;
L
Linus Torvalds 已提交
3654
	int bucket;
3655
	int offset;
L
Linus Torvalds 已提交
3656 3657
};

3658
static struct inet6_ifaddr *if6_get_first(struct seq_file *seq, loff_t pos)
L
Linus Torvalds 已提交
3659 3660 3661
{
	struct inet6_ifaddr *ifa = NULL;
	struct if6_iter_state *state = seq->private;
3662
	struct net *net = seq_file_net(seq);
3663
	int p = 0;
L
Linus Torvalds 已提交
3664

3665 3666 3667 3668 3669 3670 3671
	/* initial bucket if pos is 0 */
	if (pos == 0) {
		state->bucket = 0;
		state->offset = 0;
	}

	for (; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
3672
		hlist_for_each_entry_rcu_bh(ifa, &inet6_addr_lst[state->bucket],
3673
					 addr_lst) {
3674 3675
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
3676 3677 3678 3679 3680 3681
			/* sync with offset */
			if (p < state->offset) {
				p++;
				continue;
			}
			state->offset++;
3682
			return ifa;
3683 3684 3685 3686 3687
		}

		/* prepare for next bucket */
		state->offset = 0;
		p = 0;
L
Linus Torvalds 已提交
3688
	}
3689
	return NULL;
L
Linus Torvalds 已提交
3690 3691
}

3692 3693
static struct inet6_ifaddr *if6_get_next(struct seq_file *seq,
					 struct inet6_ifaddr *ifa)
L
Linus Torvalds 已提交
3694 3695
{
	struct if6_iter_state *state = seq->private;
3696
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
3697

3698
	hlist_for_each_entry_continue_rcu_bh(ifa, addr_lst) {
3699 3700
		if (!net_eq(dev_net(ifa->idev->dev), net))
			continue;
3701
		state->offset++;
3702
		return ifa;
3703
	}
3704

3705
	while (++state->bucket < IN6_ADDR_HSIZE) {
3706
		state->offset = 0;
3707
		hlist_for_each_entry_rcu_bh(ifa,
3708
				     &inet6_addr_lst[state->bucket], addr_lst) {
3709 3710
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
3711
			state->offset++;
3712
			return ifa;
3713
		}
L
Linus Torvalds 已提交
3714
	}
3715

3716
	return NULL;
L
Linus Torvalds 已提交
3717 3718 3719
}

static void *if6_seq_start(struct seq_file *seq, loff_t *pos)
E
Eric Dumazet 已提交
3720
	__acquires(rcu_bh)
L
Linus Torvalds 已提交
3721
{
3722
	rcu_read_lock_bh();
3723
	return if6_get_first(seq, *pos);
L
Linus Torvalds 已提交
3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735
}

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
3736
	__releases(rcu_bh)
L
Linus Torvalds 已提交
3737
{
3738
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3739 3740 3741 3742 3743
}

static int if6_seq_show(struct seq_file *seq, void *v)
{
	struct inet6_ifaddr *ifp = (struct inet6_ifaddr *)v;
3744
	seq_printf(seq, "%pi6 %02x %02x %02x %02x %8s\n",
3745
		   &ifp->addr,
L
Linus Torvalds 已提交
3746 3747 3748
		   ifp->idev->dev->ifindex,
		   ifp->prefix_len,
		   ifp->scope,
3749
		   (u8) ifp->flags,
L
Linus Torvalds 已提交
3750 3751 3752 3753
		   ifp->idev->dev->name);
	return 0;
}

3754
static const struct seq_operations if6_seq_ops = {
L
Linus Torvalds 已提交
3755 3756 3757 3758 3759 3760 3761 3762
	.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)
{
3763 3764
	return seq_open_net(inode, file, &if6_seq_ops,
			    sizeof(struct if6_iter_state));
L
Linus Torvalds 已提交
3765 3766
}

3767
static const struct file_operations if6_fops = {
L
Linus Torvalds 已提交
3768 3769 3770 3771
	.owner		= THIS_MODULE,
	.open		= if6_seq_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
3772
	.release	= seq_release_net,
L
Linus Torvalds 已提交
3773 3774
};

3775
static int __net_init if6_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
3776
{
3777
	if (!proc_create("if_inet6", S_IRUGO, net->proc_net, &if6_fops))
L
Linus Torvalds 已提交
3778 3779 3780 3781
		return -ENOMEM;
	return 0;
}

3782
static void __net_exit if6_proc_net_exit(struct net *net)
3783
{
3784
	remove_proc_entry("if_inet6", net->proc_net);
3785 3786 3787
}

static struct pernet_operations if6_proc_net_ops = {
3788 3789
	.init = if6_proc_net_init,
	.exit = if6_proc_net_exit,
3790 3791 3792 3793 3794 3795 3796
};

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

L
Linus Torvalds 已提交
3797 3798
void if6_proc_exit(void)
{
3799
	unregister_pernet_subsys(&if6_proc_net_ops);
L
Linus Torvalds 已提交
3800 3801 3802
}
#endif	/* CONFIG_PROC_FS */

A
Amerigo Wang 已提交
3803
#if IS_ENABLED(CONFIG_IPV6_MIP6)
3804
/* Check if address is a home address configured on any interface. */
3805
int ipv6_chk_home_addr(struct net *net, const struct in6_addr *addr)
3806 3807
{
	int ret = 0;
3808
	struct inet6_ifaddr *ifp = NULL;
E
Eric Dumazet 已提交
3809
	unsigned int hash = inet6_addr_hash(addr);
3810

3811
	rcu_read_lock_bh();
3812
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
3813
		if (!net_eq(dev_net(ifp->idev->dev), net))
3814
			continue;
3815
		if (ipv6_addr_equal(&ifp->addr, addr) &&
3816 3817 3818 3819 3820
		    (ifp->flags & IFA_F_HOMEADDRESS)) {
			ret = 1;
			break;
		}
	}
3821
	rcu_read_unlock_bh();
3822 3823 3824 3825
	return ret;
}
#endif

L
Linus Torvalds 已提交
3826 3827 3828 3829
/*
 *	Periodic address status verification
 */

3830
static void addrconf_verify_rtnl(void)
L
Linus Torvalds 已提交
3831
{
3832
	unsigned long now, next, next_sec, next_sched;
L
Linus Torvalds 已提交
3833 3834 3835
	struct inet6_ifaddr *ifp;
	int i;

3836 3837
	ASSERT_RTNL();

3838
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
3839
	now = jiffies;
3840
	next = round_jiffies_up(now + ADDR_CHECK_FREQUENCY);
L
Linus Torvalds 已提交
3841

3842
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
3843

S
Stephen Hemminger 已提交
3844
	for (i = 0; i < IN6_ADDR_HSIZE; i++) {
L
Linus Torvalds 已提交
3845
restart:
3846
		hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[i], addr_lst) {
L
Linus Torvalds 已提交
3847 3848
			unsigned long age;

3849 3850 3851 3852 3853 3854
			/* 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
3855 3856 3857
				continue;

			spin_lock(&ifp->lock);
3858 3859
			/* We try to batch several events at once. */
			age = (now - ifp->tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
L
Linus Torvalds 已提交
3860

3861 3862
			if (ifp->valid_lft != INFINITY_LIFE_TIME &&
			    age >= ifp->valid_lft) {
L
Linus Torvalds 已提交
3863 3864 3865 3866
				spin_unlock(&ifp->lock);
				in6_ifa_hold(ifp);
				ipv6_del_addr(ifp);
				goto restart;
3867 3868 3869
			} else if (ifp->prefered_lft == INFINITY_LIFE_TIME) {
				spin_unlock(&ifp->lock);
				continue;
L
Linus Torvalds 已提交
3870
			} else if (age >= ifp->prefered_lft) {
3871
				/* jiffies - ifp->tstamp > age >= ifp->prefered_lft */
L
Linus Torvalds 已提交
3872 3873 3874 3875 3876 3877 3878
				int deprecate = 0;

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

3879 3880
				if ((ifp->valid_lft != INFINITY_LIFE_TIME) &&
				    (time_before(ifp->tstamp + ifp->valid_lft * HZ, next)))
L
Linus Torvalds 已提交
3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893
					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)) {
3894 3895
				unsigned long regen_advance = ifp->idev->cnf.regen_max_retry *
					ifp->idev->cnf.dad_transmits *
J
Jiri Pirko 已提交
3896
					NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME) / HZ;
3897

L
Linus Torvalds 已提交
3898 3899 3900 3901 3902 3903 3904 3905 3906
				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);
3907

3908 3909 3910
						spin_lock(&ifpub->lock);
						ifpub->regen_count = 0;
						spin_unlock(&ifpub->lock);
L
Linus Torvalds 已提交
3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927
						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);
			}
		}
	}

3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938
	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;

3939 3940
	ADBG(KERN_DEBUG "now = %lu, schedule = %lu, rounded schedule = %lu => %lu\n",
	      now, next, next_sec, next_sched);
3941
	mod_delayed_work(addrconf_wq, &addr_chk_work, next_sched - now);
3942
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3943 3944
}

3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956
static void addrconf_verify_work(struct work_struct *w)
{
	rtnl_lock();
	addrconf_verify_rtnl();
	rtnl_unlock();
}

static void addrconf_verify(void)
{
	mod_delayed_work(addrconf_wq, &addr_chk_work, 0);
}

3957 3958
static struct in6_addr *extract_addr(struct nlattr *addr, struct nlattr *local,
				     struct in6_addr **peer_pfx)
3959 3960 3961
{
	struct in6_addr *pfx = NULL;

3962 3963
	*peer_pfx = NULL;

3964 3965 3966 3967 3968
	if (addr)
		pfx = nla_data(addr);

	if (local) {
		if (pfx && nla_memcmp(local, pfx, sizeof(*pfx)))
3969 3970
			*peer_pfx = pfx;
		pfx = nla_data(local);
3971 3972 3973 3974 3975
	}

	return pfx;
}

3976
static const struct nla_policy ifa_ipv6_policy[IFA_MAX+1] = {
3977 3978 3979
	[IFA_ADDRESS]		= { .len = sizeof(struct in6_addr) },
	[IFA_LOCAL]		= { .len = sizeof(struct in6_addr) },
	[IFA_CACHEINFO]		= { .len = sizeof(struct ifa_cacheinfo) },
3980
	[IFA_FLAGS]		= { .len = sizeof(u32) },
3981 3982
};

L
Linus Torvalds 已提交
3983
static int
3984
inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
3985
{
3986
	struct net *net = sock_net(skb->sk);
3987 3988
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
3989
	struct in6_addr *pfx, *peer_pfx;
3990
	u32 ifa_flags;
3991
	int err;
L
Linus Torvalds 已提交
3992

3993 3994 3995 3996 3997
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
3998
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
L
Linus Torvalds 已提交
3999 4000 4001
	if (pfx == NULL)
		return -EINVAL;

4002 4003 4004 4005 4006 4007 4008
	ifa_flags = tb[IFA_FLAGS] ? nla_get_u32(tb[IFA_FLAGS]) : ifm->ifa_flags;

	/* We ignore other flags so far. */
	ifa_flags &= IFA_F_MANAGETEMPADDR;

	return inet6_addr_del(net, ifm->ifa_index, ifa_flags, pfx,
			      ifm->ifa_prefixlen);
L
Linus Torvalds 已提交
4009 4010
}

4011
static int inet6_addr_modify(struct inet6_ifaddr *ifp, u32 ifa_flags,
4012
			     u32 prefered_lft, u32 valid_lft)
4013
{
4014 4015
	u32 flags;
	clock_t expires;
4016
	unsigned long timeout;
4017
	bool was_managetempaddr;
4018
	bool had_prefixroute;
4019

4020 4021
	ASSERT_RTNL();

4022 4023 4024
	if (!valid_lft || (prefered_lft > valid_lft))
		return -EINVAL;

4025 4026 4027 4028
	if (ifa_flags & IFA_F_MANAGETEMPADDR &&
	    (ifp->flags & IFA_F_TEMPORARY || ifp->prefix_len != 64))
		return -EINVAL;

4029 4030 4031 4032
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
4033
		flags = RTF_EXPIRES;
4034 4035 4036 4037
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
4038
	}
4039

4040 4041 4042 4043 4044 4045
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
4046 4047

	spin_lock_bh(&ifp->lock);
4048
	was_managetempaddr = ifp->flags & IFA_F_MANAGETEMPADDR;
4049 4050
	had_prefixroute = ifp->flags & IFA_F_PERMANENT &&
			  !(ifp->flags & IFA_F_NOPREFIXROUTE);
4051
	ifp->flags &= ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD |
4052 4053
			IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
			IFA_F_NOPREFIXROUTE);
4054
	ifp->flags |= ifa_flags;
4055 4056 4057 4058 4059 4060 4061 4062
	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);

4063 4064 4065
	if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, ifp->idev->dev,
				      expires, flags);
4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077
	} 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);
		}
4078
	}
4079 4080 4081 4082 4083 4084 4085 4086

	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);
	}

4087
	addrconf_verify_rtnl();
4088 4089 4090 4091

	return 0;
}

L
Linus Torvalds 已提交
4092
static int
4093
inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
4094
{
4095
	struct net *net = sock_net(skb->sk);
4096 4097
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4098
	struct in6_addr *pfx, *peer_pfx;
4099 4100
	struct inet6_ifaddr *ifa;
	struct net_device *dev;
4101
	u32 valid_lft = INFINITY_LIFE_TIME, preferred_lft = INFINITY_LIFE_TIME;
4102
	u32 ifa_flags;
4103
	int err;
L
Linus Torvalds 已提交
4104

4105 4106 4107 4108 4109
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
4110
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
L
Linus Torvalds 已提交
4111 4112 4113
	if (pfx == NULL)
		return -EINVAL;

4114
	if (tb[IFA_CACHEINFO]) {
4115
		struct ifa_cacheinfo *ci;
4116 4117

		ci = nla_data(tb[IFA_CACHEINFO]);
4118
		valid_lft = ci->ifa_valid;
4119 4120 4121 4122
		preferred_lft = ci->ifa_prefered;
	} else {
		preferred_lft = INFINITY_LIFE_TIME;
		valid_lft = INFINITY_LIFE_TIME;
4123 4124
	}

4125
	dev =  __dev_get_by_index(net, ifm->ifa_index);
4126 4127 4128
	if (dev == NULL)
		return -ENODEV;

4129 4130
	ifa_flags = tb[IFA_FLAGS] ? nla_get_u32(tb[IFA_FLAGS]) : ifm->ifa_flags;

4131
	/* We ignore other flags so far. */
4132
	ifa_flags &= IFA_F_NODAD | IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
4133
		     IFA_F_NOPREFIXROUTE | IFA_F_MCAUTOJOIN;
4134

4135
	ifa = ipv6_get_ifaddr(net, pfx, dev, 1);
4136 4137 4138
	if (ifa == NULL) {
		/*
		 * It would be best to check for !NLM_F_CREATE here but
4139
		 * userspace already relies on not having to provide this.
4140
		 */
4141
		return inet6_addr_add(net, ifm->ifa_index, pfx, peer_pfx,
4142 4143
				      ifm->ifa_prefixlen, ifa_flags,
				      preferred_lft, valid_lft);
4144 4145
	}

4146 4147 4148 4149
	if (nlh->nlmsg_flags & NLM_F_EXCL ||
	    !(nlh->nlmsg_flags & NLM_F_REPLACE))
		err = -EEXIST;
	else
4150
		err = inet6_addr_modify(ifa, ifa_flags, preferred_lft, valid_lft);
4151 4152 4153 4154

	in6_ifa_put(ifa);

	return err;
L
Linus Torvalds 已提交
4155 4156
}

4157
static void put_ifaddrmsg(struct nlmsghdr *nlh, u8 prefixlen, u32 flags,
4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169
			  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;
}

4170 4171 4172 4173 4174
static int put_cacheinfo(struct sk_buff *skb, unsigned long cstamp,
			 unsigned long tstamp, u32 preferred, u32 valid)
{
	struct ifa_cacheinfo ci;

4175 4176
	ci.cstamp = cstamp_delta(cstamp);
	ci.tstamp = cstamp_delta(tstamp);
4177 4178 4179 4180 4181 4182
	ci.ifa_prefered = preferred;
	ci.ifa_valid = valid;

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

4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194
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;
}

4195 4196
static inline int inet6_ifaddr_msgsize(void)
{
4197
	return NLMSG_ALIGN(sizeof(struct ifaddrmsg))
4198
	       + nla_total_size(16) /* IFA_LOCAL */
4199
	       + nla_total_size(16) /* IFA_ADDRESS */
4200 4201
	       + nla_total_size(sizeof(struct ifa_cacheinfo))
	       + nla_total_size(4)  /* IFA_FLAGS */;
4202
}
4203

L
Linus Torvalds 已提交
4204
static int inet6_fill_ifaddr(struct sk_buff *skb, struct inet6_ifaddr *ifa,
4205
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4206 4207
{
	struct nlmsghdr  *nlh;
4208
	u32 preferred, valid;
L
Linus Torvalds 已提交
4209

4210
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4211
	if (nlh == NULL)
4212
		return -EMSGSIZE;
4213

4214 4215 4216
	put_ifaddrmsg(nlh, ifa->prefix_len, ifa->flags, rt_scope(ifa->scope),
		      ifa->idev->dev->ifindex);

4217 4218
	if (!((ifa->flags&IFA_F_PERMANENT) &&
	      (ifa->prefered_lft == INFINITY_LIFE_TIME))) {
4219 4220 4221
		preferred = ifa->prefered_lft;
		valid = ifa->valid_lft;
		if (preferred != INFINITY_LIFE_TIME) {
L
Linus Torvalds 已提交
4222
			long tval = (jiffies - ifa->tstamp)/HZ;
4223 4224 4225 4226
			if (preferred > tval)
				preferred -= tval;
			else
				preferred = 0;
4227 4228 4229 4230 4231 4232
			if (valid != INFINITY_LIFE_TIME) {
				if (valid > tval)
					valid -= tval;
				else
					valid = 0;
			}
L
Linus Torvalds 已提交
4233 4234
		}
	} else {
4235 4236 4237 4238
		preferred = INFINITY_LIFE_TIME;
		valid = INFINITY_LIFE_TIME;
	}

C
Cong Wang 已提交
4239
	if (!ipv6_addr_any(&ifa->peer_addr)) {
4240 4241 4242 4243 4244 4245 4246 4247 4248
		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
4249

4250 4251 4252
	if (nla_put_u32(skb, IFA_FLAGS, ifa->flags) < 0)
		goto error;

4253 4254
	nlmsg_end(skb, nlh);
	return 0;
4255 4256 4257 4258

error:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4259 4260 4261
}

static int inet6_fill_ifmcaddr(struct sk_buff *skb, struct ifmcaddr6 *ifmca,
4262
				u32 portid, u32 seq, int event, u16 flags)
L
Linus Torvalds 已提交
4263 4264
{
	struct nlmsghdr  *nlh;
4265 4266 4267 4268 4269
	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
4270

4271
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4272
	if (nlh == NULL)
4273
		return -EMSGSIZE;
4274

4275 4276 4277
	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,
4278 4279 4280 4281
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
4282

4283 4284
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4285 4286 4287
}

static int inet6_fill_ifacaddr(struct sk_buff *skb, struct ifacaddr6 *ifaca,
4288
				u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4289 4290
{
	struct nlmsghdr  *nlh;
4291 4292 4293 4294 4295
	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
4296

4297
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4298
	if (nlh == NULL)
4299
		return -EMSGSIZE;
4300

4301 4302 4303
	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,
4304 4305 4306 4307
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
L
Linus Torvalds 已提交
4308

4309 4310
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4311 4312
}

S
Stephen Hemminger 已提交
4313
enum addr_type_t {
L
Linus Torvalds 已提交
4314 4315 4316 4317 4318
	UNICAST_ADDR,
	MULTICAST_ADDR,
	ANYCAST_ADDR,
};

E
Eric Dumazet 已提交
4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330
/* 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) {
4331 4332 4333
	case UNICAST_ADDR: {
		struct inet6_ifaddr *ifa;

E
Eric Dumazet 已提交
4334
		/* unicast address incl. temp addr */
4335 4336
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
			if (++ip_idx < s_ip_idx)
E
Eric Dumazet 已提交
4337 4338
				continue;
			err = inet6_fill_ifaddr(skb, ifa,
4339
						NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4340 4341 4342
						cb->nlh->nlmsg_seq,
						RTM_NEWADDR,
						NLM_F_MULTI);
4343
			if (err < 0)
E
Eric Dumazet 已提交
4344
				break;
4345
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
E
Eric Dumazet 已提交
4346 4347
		}
		break;
4348
	}
E
Eric Dumazet 已提交
4349 4350 4351 4352 4353 4354 4355
	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,
4356
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4357 4358 4359
						  cb->nlh->nlmsg_seq,
						  RTM_GETMULTICAST,
						  NLM_F_MULTI);
4360
			if (err < 0)
E
Eric Dumazet 已提交
4361 4362 4363 4364 4365 4366 4367 4368 4369 4370
				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,
4371
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4372 4373 4374
						  cb->nlh->nlmsg_seq,
						  RTM_GETANYCAST,
						  NLM_F_MULTI);
4375
			if (err < 0)
E
Eric Dumazet 已提交
4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386
				break;
		}
		break;
	default:
		break;
	}
	read_unlock_bh(&idev->lock);
	*p_ip_idx = ip_idx;
	return err;
}

L
Linus Torvalds 已提交
4387 4388 4389
static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
			   enum addr_type_t type)
{
E
Eric Dumazet 已提交
4390 4391
	struct net *net = sock_net(skb->sk);
	int h, s_h;
L
Linus Torvalds 已提交
4392 4393 4394
	int idx, ip_idx;
	int s_idx, s_ip_idx;
	struct net_device *dev;
E
Eric Dumazet 已提交
4395 4396
	struct inet6_dev *idev;
	struct hlist_head *head;
4397

E
Eric Dumazet 已提交
4398 4399 4400
	s_h = cb->args[0];
	s_idx = idx = cb->args[1];
	s_ip_idx = ip_idx = cb->args[2];
4401

E
Eric Dumazet 已提交
4402
	rcu_read_lock();
4403
	cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^ net->dev_base_seq;
E
Eric Dumazet 已提交
4404 4405 4406
	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
4407
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
4408 4409
			if (idx < s_idx)
				goto cont;
4410
			if (h > s_h || idx > s_idx)
E
Eric Dumazet 已提交
4411 4412
				s_ip_idx = 0;
			ip_idx = 0;
S
Stephen Hemminger 已提交
4413 4414
			idev = __in6_dev_get(dev);
			if (!idev)
E
Eric Dumazet 已提交
4415 4416 4417
				goto cont;

			if (in6_dump_addrs(idev, skb, cb, type,
4418
					   s_ip_idx, &ip_idx) < 0)
E
Eric Dumazet 已提交
4419
				goto done;
4420
cont:
E
Eric Dumazet 已提交
4421 4422
			idx++;
		}
L
Linus Torvalds 已提交
4423
	}
E
Eric Dumazet 已提交
4424 4425 4426 4427 4428 4429
done:
	rcu_read_unlock();
	cb->args[0] = h;
	cb->args[1] = idx;
	cb->args[2] = ip_idx;

L
Linus Torvalds 已提交
4430 4431 4432 4433 4434 4435
	return skb->len;
}

static int inet6_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = UNICAST_ADDR;
4436

L
Linus Torvalds 已提交
4437 4438 4439 4440 4441 4442
	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;
4443

L
Linus Torvalds 已提交
4444 4445 4446 4447 4448 4449 4450
	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;
4451

L
Linus Torvalds 已提交
4452 4453 4454
	return inet6_dump_addr(skb, cb, type);
}

4455
static int inet6_rtm_getaddr(struct sk_buff *in_skb, struct nlmsghdr *nlh)
4456
{
4457
	struct net *net = sock_net(in_skb->sk);
4458 4459
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4460
	struct in6_addr *addr = NULL, *peer;
4461 4462 4463 4464 4465
	struct net_device *dev = NULL;
	struct inet6_ifaddr *ifa;
	struct sk_buff *skb;
	int err;

4466 4467 4468 4469
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		goto errout;

4470
	addr = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer);
4471 4472 4473
	if (addr == NULL) {
		err = -EINVAL;
		goto errout;
4474 4475
	}

4476
	ifm = nlmsg_data(nlh);
4477
	if (ifm->ifa_index)
4478
		dev = __dev_get_by_index(net, ifm->ifa_index);
4479

S
Stephen Hemminger 已提交
4480 4481
	ifa = ipv6_get_ifaddr(net, addr, dev, 1);
	if (!ifa) {
4482 4483 4484
		err = -EADDRNOTAVAIL;
		goto errout;
	}
4485

S
Stephen Hemminger 已提交
4486 4487
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_KERNEL);
	if (!skb) {
4488
		err = -ENOBUFS;
4489
		goto errout_ifa;
4490 4491
	}

4492
	err = inet6_fill_ifaddr(skb, ifa, NETLINK_CB(in_skb).portid,
4493
				nlh->nlmsg_seq, RTM_NEWADDR, 0);
4494 4495 4496 4497 4498 4499
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout_ifa;
	}
4500
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
4501
errout_ifa:
4502
	in6_ifa_put(ifa);
4503
errout:
4504 4505 4506
	return err;
}

L
Linus Torvalds 已提交
4507 4508 4509
static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa)
{
	struct sk_buff *skb;
4510
	struct net *net = dev_net(ifa->idev->dev);
4511
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
4512

4513
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_ATOMIC);
4514 4515 4516 4517
	if (skb == NULL)
		goto errout;

	err = inet6_fill_ifaddr(skb, ifa, 0, 0, event, 0);
4518 4519 4520 4521 4522 4523
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4524 4525
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFADDR, NULL, GFP_ATOMIC);
	return;
4526 4527
errout:
	if (err < 0)
4528
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFADDR, err);
L
Linus Torvalds 已提交
4529 4530
}

D
Dave Jones 已提交
4531
static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
L
Linus Torvalds 已提交
4532 4533
				__s32 *array, int bytes)
{
4534 4535
	BUG_ON(bytes < (DEVCONF_MAX * 4));

L
Linus Torvalds 已提交
4536 4537 4538 4539 4540 4541 4542 4543 4544
	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;
4545 4546 4547 4548
	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
4549
	array[DEVCONF_FORCE_MLD_VERSION] = cnf->force_mld_version;
4550 4551 4552 4553
	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
4554 4555 4556 4557 4558 4559
	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;
4560
	array[DEVCONF_ACCEPT_RA_DEFRTR] = cnf->accept_ra_defrtr;
4561
	array[DEVCONF_ACCEPT_RA_PINFO] = cnf->accept_ra_pinfo;
4562 4563
#ifdef CONFIG_IPV6_ROUTER_PREF
	array[DEVCONF_ACCEPT_RA_RTR_PREF] = cnf->accept_ra_rtr_pref;
4564 4565
	array[DEVCONF_RTR_PROBE_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_probe_interval);
N
Neil Horman 已提交
4566
#ifdef CONFIG_IPV6_ROUTE_INFO
4567 4568
	array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
#endif
4569
#endif
4570
	array[DEVCONF_PROXY_NDP] = cnf->proxy_ndp;
4571
	array[DEVCONF_ACCEPT_SOURCE_ROUTE] = cnf->accept_source_route;
4572 4573
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
4574
	array[DEVCONF_USE_OPTIMISTIC] = cnf->use_optimistic;
4575
#endif
4576 4577 4578
#ifdef CONFIG_IPV6_MROUTE
	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
#endif
4579
	array[DEVCONF_DISABLE_IPV6] = cnf->disable_ipv6;
4580
	array[DEVCONF_ACCEPT_DAD] = cnf->accept_dad;
4581
	array[DEVCONF_FORCE_TLLAO] = cnf->force_tllao;
4582
	array[DEVCONF_NDISC_NOTIFY] = cnf->ndisc_notify;
4583
	array[DEVCONF_SUPPRESS_FRAG_NDISC] = cnf->suppress_frag_ndisc;
4584
	array[DEVCONF_ACCEPT_RA_FROM_LOCAL] = cnf->accept_ra_from_local;
4585
	array[DEVCONF_ACCEPT_RA_MTU] = cnf->accept_ra_mtu;
4586
	/* we omit DEVCONF_STABLE_SECRET for now */
L
Linus Torvalds 已提交
4587 4588
}

T
Thomas Graf 已提交
4589 4590 4591 4592 4593 4594
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 */
4595 4596
	     + 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
4597 4598
}

4599 4600 4601 4602 4603 4604 4605
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
4606
	       + nla_total_size(inet6_ifla6_size()); /* IFLA_PROTINFO */
4607
}
4608

4609
static inline void __snmp6_fill_statsdev(u64 *stats, atomic_long_t *mib,
4610
				      int items, int bytes)
4611 4612 4613 4614 4615 4616 4617 4618
{
	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++)
4619
		put_unaligned(atomic_long_read(&mib[i]), &stats[i]);
4620 4621 4622 4623

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

W
WANG Cong 已提交
4624
static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638
				      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);
}

4639 4640 4641
static void snmp6_fill_stats(u64 *stats, struct inet6_dev *idev, int attrtype,
			     int bytes)
{
S
Stephen Hemminger 已提交
4642
	switch (attrtype) {
4643
	case IFLA_INET6_STATS:
W
WANG Cong 已提交
4644
		__snmp6_fill_stats64(stats, idev->stats.ipv6,
4645
				     IPSTATS_MIB_MAX, bytes, offsetof(struct ipstats_mib, syncp));
4646 4647
		break;
	case IFLA_INET6_ICMP6STATS:
4648
		__snmp6_fill_statsdev(stats, idev->stats.icmpv6dev->mibs, ICMP6_MIB_MAX, bytes);
4649 4650 4651 4652
		break;
	}
}

T
Thomas Graf 已提交
4653 4654 4655 4656 4657
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
4658 4659
	if (nla_put_u32(skb, IFLA_INET6_FLAGS, idev->if_flags))
		goto nla_put_failure;
T
Thomas Graf 已提交
4660
	ci.max_reasm_len = IPV6_MAXPLEN;
4661 4662
	ci.tstamp = cstamp_delta(idev->tstamp);
	ci.reachable_time = jiffies_to_msecs(idev->nd_parms->reachable_time);
J
Jiri Pirko 已提交
4663
	ci.retrans_time = jiffies_to_msecs(NEIGH_VAR(idev->nd_parms, RETRANS_TIME));
D
David S. Miller 已提交
4664 4665
	if (nla_put(skb, IFLA_INET6_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
T
Thomas Graf 已提交
4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682
	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));

4683 4684 4685
	nla = nla_reserve(skb, IFLA_INET6_TOKEN, sizeof(struct in6_addr));
	if (nla == NULL)
		goto nla_put_failure;
4686 4687 4688 4689

	if (nla_put_u8(skb, IFLA_INET6_ADDR_GEN_MODE, idev->addr_gen_mode))
		goto nla_put_failure;

4690 4691 4692 4693
	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
4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720
	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;
}

4721 4722 4723 4724
static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token)
{
	struct inet6_ifaddr *ifp;
	struct net_device *dev = idev->dev;
4725
	bool update_rs = false;
4726
	struct in6_addr ll_addr;
4727

4728 4729
	ASSERT_RTNL();

4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747
	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);

4748 4749 4750
	if (!idev->dead && (idev->if_flags & IF_READY) &&
	    !ipv6_get_lladdr(dev, &ll_addr, IFA_F_TENTATIVE |
			     IFA_F_OPTIMISTIC)) {
4751 4752 4753 4754 4755 4756 4757

		/* 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;
	}
4758 4759

	write_lock_bh(&idev->lock);
4760

4761
	if (update_rs) {
4762
		idev->if_flags |= IF_RS_SENT;
4763 4764 4765
		idev->rs_probes = 1;
		addrconf_mod_rs_timer(idev, idev->cnf.rtr_solicit_interval);
	}
4766 4767 4768 4769

	/* Well, that's kinda nasty ... */
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
		spin_lock(&ifp->lock);
4770
		if (ifp->tokenized) {
4771 4772 4773 4774 4775 4776 4777
			ifp->valid_lft = 0;
			ifp->prefered_lft = 0;
		}
		spin_unlock(&ifp->lock);
	}

	write_unlock_bh(&idev->lock);
4778
	inet6_ifinfo_notify(RTM_NEWLINK, idev);
4779
	addrconf_verify_rtnl();
4780 4781 4782
	return 0;
}

4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793 4794 4795 4796 4797 4798
static const struct nla_policy inet6_af_policy[IFLA_INET6_MAX + 1] = {
	[IFLA_INET6_ADDR_GEN_MODE]	= { .type = NLA_U8 },
	[IFLA_INET6_TOKEN]		= { .len = sizeof(struct in6_addr) },
};

static int inet6_validate_link_af(const struct net_device *dev,
				  const struct nlattr *nla)
{
	struct nlattr *tb[IFLA_INET6_MAX + 1];

	if (dev && !__in6_dev_get(dev))
		return -EAFNOSUPPORT;

	return nla_parse_nested(tb, IFLA_INET6_MAX, nla, inet6_af_policy);
}

4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810
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();

4811
	if (tb[IFLA_INET6_TOKEN]) {
4812
		err = inet6_set_iftoken(idev, nla_data(tb[IFLA_INET6_TOKEN]));
4813 4814 4815 4816 4817 4818 4819 4820
		if (err)
			return err;
	}

	if (tb[IFLA_INET6_ADDR_GEN_MODE]) {
		u8 mode = nla_get_u8(tb[IFLA_INET6_ADDR_GEN_MODE]);

		if (mode != IN6_ADDR_GEN_MODE_EUI64 &&
4821 4822 4823 4824 4825 4826 4827
		    mode != IN6_ADDR_GEN_MODE_NONE &&
		    mode != IN6_ADDR_GEN_MODE_STABLE_PRIVACY)
			return -EINVAL;

		if (mode == IN6_ADDR_GEN_MODE_STABLE_PRIVACY &&
		    !idev->cnf.stable_secret.initialized &&
		    !dev_net(dev)->ipv6.devconf_dflt->stable_secret.initialized)
4828
			return -EINVAL;
4829

4830 4831 4832
		idev->addr_gen_mode = mode;
		err = 0;
	}
4833 4834 4835 4836

	return err;
}

4837
static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
4838
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4839
{
4840 4841 4842 4843 4844
	struct net_device *dev = idev->dev;
	struct ifinfomsg *hdr;
	struct nlmsghdr *nlh;
	void *protoinfo;

4845
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*hdr), flags);
4846
	if (nlh == NULL)
4847
		return -EMSGSIZE;
4848 4849 4850 4851 4852 4853 4854 4855 4856

	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
4857 4858 4859 4860 4861 4862 4863
	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;
4864 4865 4866
	protoinfo = nla_nest_start(skb, IFLA_PROTINFO);
	if (protoinfo == NULL)
		goto nla_put_failure;
L
Linus Torvalds 已提交
4867

T
Thomas Graf 已提交
4868
	if (inet6_fill_ifla6_attrs(skb, idev) < 0)
4869
		goto nla_put_failure;
L
Linus Torvalds 已提交
4870

4871
	nla_nest_end(skb, protoinfo);
4872 4873
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4874

4875
nla_put_failure:
4876 4877
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4878 4879 4880 4881
}

static int inet6_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
{
4882
	struct net *net = sock_net(skb->sk);
E
Eric Dumazet 已提交
4883
	int h, s_h;
4884
	int idx = 0, s_idx;
L
Linus Torvalds 已提交
4885 4886
	struct net_device *dev;
	struct inet6_dev *idev;
E
Eric Dumazet 已提交
4887
	struct hlist_head *head;
L
Linus Torvalds 已提交
4888

E
Eric Dumazet 已提交
4889 4890 4891 4892 4893 4894 4895
	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];
4896
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
4897 4898 4899 4900 4901 4902
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;
			if (inet6_fill_ifinfo(skb, idev,
4903
					      NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4904
					      cb->nlh->nlmsg_seq,
4905
					      RTM_NEWLINK, NLM_F_MULTI) < 0)
E
Eric Dumazet 已提交
4906
				goto out;
4907
cont:
E
Eric Dumazet 已提交
4908 4909
			idx++;
		}
L
Linus Torvalds 已提交
4910
	}
E
Eric Dumazet 已提交
4911 4912 4913 4914
out:
	rcu_read_unlock();
	cb->args[1] = idx;
	cb->args[0] = h;
L
Linus Torvalds 已提交
4915 4916 4917 4918 4919 4920 4921

	return skb->len;
}

void inet6_ifinfo_notify(int event, struct inet6_dev *idev)
{
	struct sk_buff *skb;
4922
	struct net *net = dev_net(idev->dev);
4923
	int err = -ENOBUFS;
4924

4925
	skb = nlmsg_new(inet6_if_nlmsg_size(), GFP_ATOMIC);
4926 4927 4928 4929
	if (skb == NULL)
		goto errout;

	err = inet6_fill_ifinfo(skb, idev, 0, 0, event, 0);
4930 4931 4932 4933 4934 4935
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_if_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4936
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFINFO, NULL, GFP_ATOMIC);
4937
	return;
4938 4939
errout:
	if (err < 0)
4940
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFINFO, err);
L
Linus Torvalds 已提交
4941 4942
}

4943 4944 4945 4946 4947 4948
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));
}
4949

L
Linus Torvalds 已提交
4950
static int inet6_fill_prefix(struct sk_buff *skb, struct inet6_dev *idev,
4951
			     struct prefix_info *pinfo, u32 portid, u32 seq,
4952
			     int event, unsigned int flags)
L
Linus Torvalds 已提交
4953
{
4954 4955
	struct prefixmsg *pmsg;
	struct nlmsghdr *nlh;
L
Linus Torvalds 已提交
4956 4957
	struct prefix_cacheinfo	ci;

4958
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*pmsg), flags);
4959
	if (nlh == NULL)
4960
		return -EMSGSIZE;
4961 4962

	pmsg = nlmsg_data(nlh);
L
Linus Torvalds 已提交
4963
	pmsg->prefix_family = AF_INET6;
4964 4965
	pmsg->prefix_pad1 = 0;
	pmsg->prefix_pad2 = 0;
L
Linus Torvalds 已提交
4966 4967 4968
	pmsg->prefix_ifindex = idev->dev->ifindex;
	pmsg->prefix_len = pinfo->prefix_len;
	pmsg->prefix_type = pinfo->type;
4969
	pmsg->prefix_pad3 = 0;
L
Linus Torvalds 已提交
4970 4971 4972 4973 4974 4975
	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
4976 4977
	if (nla_put(skb, PREFIX_ADDRESS, sizeof(pinfo->prefix), &pinfo->prefix))
		goto nla_put_failure;
L
Linus Torvalds 已提交
4978 4979
	ci.preferred_time = ntohl(pinfo->prefered);
	ci.valid_time = ntohl(pinfo->valid);
D
David S. Miller 已提交
4980 4981
	if (nla_put(skb, PREFIX_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
4982 4983
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4984

4985
nla_put_failure:
4986 4987
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4988 4989
}

4990
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
4991 4992 4993
			 struct prefix_info *pinfo)
{
	struct sk_buff *skb;
4994
	struct net *net = dev_net(idev->dev);
4995
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
4996

4997
	skb = nlmsg_new(inet6_prefix_nlmsg_size(), GFP_ATOMIC);
4998 4999 5000 5001
	if (skb == NULL)
		goto errout;

	err = inet6_fill_prefix(skb, idev, pinfo, 0, 0, event, 0);
5002 5003 5004 5005 5006 5007
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_prefix_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5008 5009
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_PREFIX, NULL, GFP_ATOMIC);
	return;
5010 5011
errout:
	if (err < 0)
5012
		rtnl_set_sk_err(net, RTNLGRP_IPV6_PREFIX, err);
L
Linus Torvalds 已提交
5013 5014 5015 5016
}

static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
5017 5018
	struct net *net = dev_net(ifp->idev->dev);

5019 5020 5021
	if (event)
		ASSERT_RTNL();

L
Linus Torvalds 已提交
5022 5023 5024 5025
	inet6_ifa_notify(event ? : RTM_NEWADDR, ifp);

	switch (event) {
	case RTM_NEWADDR:
5026 5027 5028 5029 5030 5031 5032 5033
		/*
		 * 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
5034 5035
		if (ifp->idev->cnf.forwarding)
			addrconf_join_anycast(ifp);
C
Cong Wang 已提交
5036
		if (!ipv6_addr_any(&ifp->peer_addr))
5037 5038
			addrconf_prefix_route(&ifp->peer_addr, 128,
					      ifp->idev->dev, 0, 0);
L
Linus Torvalds 已提交
5039 5040 5041 5042 5043
		break;
	case RTM_DELADDR:
		if (ifp->idev->cnf.forwarding)
			addrconf_leave_anycast(ifp);
		addrconf_leave_solict(ifp->idev, &ifp->addr);
C
Cong Wang 已提交
5044
		if (!ipv6_addr_any(&ifp->peer_addr)) {
5045 5046
			struct rt6_info *rt;

5047 5048
			rt = addrconf_get_prefix_route(&ifp->peer_addr, 128,
						       ifp->idev->dev, 0, 0);
5049 5050
			if (rt && ip6_del_rt(rt))
				dst_free(&rt->dst);
5051
		}
5052
		dst_hold(&ifp->rt->dst);
5053

5054
		if (ip6_del_rt(ifp->rt))
5055
			dst_free(&ifp->rt->dst);
H
Hannes Frederic Sowa 已提交
5056 5057

		rt_genid_bump_ipv6(net);
L
Linus Torvalds 已提交
5058 5059
		break;
	}
5060
	atomic_inc(&net->ipv6.dev_addr_genid);
L
Linus Torvalds 已提交
5061 5062 5063 5064
}

static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
5065
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
5066 5067
	if (likely(ifp->idev->dead == 0))
		__ipv6_ifa_notify(event, ifp);
5068
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
5069 5070 5071 5072 5073
}

#ifdef CONFIG_SYSCTL

static
5074
int addrconf_sysctl_forward(struct ctl_table *ctl, int write,
L
Linus Torvalds 已提交
5075 5076 5077 5078
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
5079
	loff_t pos = *ppos;
5080
	struct ctl_table lctl;
L
Linus Torvalds 已提交
5081 5082
	int ret;

5083 5084 5085 5086 5087 5088 5089 5090
	/*
	 * 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
5091

5092
	if (write)
5093
		ret = addrconf_fixup_forwarding(ctl, valp, val);
E
Eric W. Biederman 已提交
5094 5095
	if (ret)
		*ppos = pos;
5096
	return ret;
L
Linus Torvalds 已提交
5097 5098
}

5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113
static
int addrconf_sysctl_mtu(struct ctl_table *ctl, int write,
			void __user *buffer, size_t *lenp, loff_t *ppos)
{
	struct inet6_dev *idev = ctl->extra1;
	int min_mtu = IPV6_MIN_MTU;
	struct ctl_table lctl;

	lctl = *ctl;
	lctl.extra1 = &min_mtu;
	lctl.extra2 = idev ? &idev->dev->mtu : NULL;

	return proc_dointvec_minmax(&lctl, write, buffer, lenp, ppos);
}

5114 5115
static void dev_disable_change(struct inet6_dev *idev)
{
5116 5117
	struct netdev_notifier_info info;

5118 5119 5120
	if (!idev || !idev->dev)
		return;

5121
	netdev_notifier_info_init(&info, idev->dev);
5122
	if (idev->cnf.disable_ipv6)
5123
		addrconf_notify(NULL, NETDEV_DOWN, &info);
5124
	else
5125
		addrconf_notify(NULL, NETDEV_UP, &info);
5126 5127 5128 5129 5130 5131 5132
}

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

5133 5134
	rcu_read_lock();
	for_each_netdev_rcu(net, dev) {
5135 5136 5137 5138 5139 5140 5141 5142
		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);
		}
	}
5143
	rcu_read_unlock();
5144 5145
}

5146
static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
5147 5148
{
	struct net *net;
5149 5150 5151 5152
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
5153 5154

	net = (struct net *)table->extra2;
5155 5156
	old = *p;
	*p = newf;
5157

5158 5159
	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
		rtnl_unlock();
5160
		return 0;
E
Eric W. Biederman 已提交
5161
	}
5162 5163 5164 5165

	if (p == &net->ipv6.devconf_all->disable_ipv6) {
		net->ipv6.devconf_dflt->disable_ipv6 = newf;
		addrconf_disable_change(net, newf);
5166
	} else if ((!newf) ^ (!old))
5167 5168 5169 5170 5171 5172 5173
		dev_disable_change((struct inet6_dev *)table->extra1);

	rtnl_unlock();
	return 0;
}

static
5174
int addrconf_sysctl_disable(struct ctl_table *ctl, int write,
5175 5176 5177 5178
			    void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
5179
	loff_t pos = *ppos;
5180
	struct ctl_table lctl;
5181 5182
	int ret;

5183 5184 5185 5186 5187 5188 5189 5190
	/*
	 * 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);
5191 5192 5193

	if (write)
		ret = addrconf_disable_ipv6(ctl, valp, val);
E
Eric W. Biederman 已提交
5194 5195
	if (ret)
		*ppos = pos;
5196 5197 5198
	return ret;
}

5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237
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;
}

5238 5239 5240 5241 5242 5243 5244 5245
static int addrconf_sysctl_stable_secret(struct ctl_table *ctl, int write,
					 void __user *buffer, size_t *lenp,
					 loff_t *ppos)
{
	int err;
	struct in6_addr addr;
	char str[IPV6_MAX_STRLEN];
	struct ctl_table lctl = *ctl;
5246
	struct net *net = ctl->extra2;
5247 5248
	struct ipv6_stable_secret *secret = ctl->data;

5249 5250 5251
	if (&net->ipv6.devconf_all->stable_secret == ctl->data)
		return -EIO;

5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283
	lctl.maxlen = IPV6_MAX_STRLEN;
	lctl.data = str;

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

	if (!write && !secret->initialized) {
		err = -EIO;
		goto out;
	}

	if (!write) {
		err = snprintf(str, sizeof(str), "%pI6",
			       &secret->secret);
		if (err >= sizeof(str)) {
			err = -EIO;
			goto out;
		}
	}

	err = proc_dostring(&lctl, write, buffer, lenp, ppos);
	if (err || !write)
		goto out;

	if (in6_pton(str, -1, addr.in6_u.u6_addr8, -1, NULL) != 1) {
		err = -EIO;
		goto out;
	}

	secret->initialized = true;
	secret->secret = addr;

5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300
	if (&net->ipv6.devconf_dflt->stable_secret == ctl->data) {
		struct net_device *dev;

		for_each_netdev(net, dev) {
			struct inet6_dev *idev = __in6_dev_get(dev);

			if (idev) {
				idev->addr_gen_mode =
					IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
			}
		}
	} else {
		struct inet6_dev *idev = ctl->extra1;

		idev->addr_gen_mode = IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
	}

5301 5302 5303 5304 5305
out:
	rtnl_unlock();

	return err;
}
5306

L
Linus Torvalds 已提交
5307 5308 5309
static struct addrconf_sysctl_table
{
	struct ctl_table_header *sysctl_header;
5310
	struct ctl_table addrconf_vars[DEVCONF_MAX+1];
5311
} addrconf_sysctl __read_mostly = {
L
Linus Torvalds 已提交
5312 5313
	.sysctl_header = NULL,
	.addrconf_vars = {
5314
		{
S
Stephen Hemminger 已提交
5315 5316 5317 5318 5319
			.procname	= "forwarding",
			.data		= &ipv6_devconf.forwarding,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= addrconf_sysctl_forward,
L
Linus Torvalds 已提交
5320 5321
		},
		{
S
Stephen Hemminger 已提交
5322 5323 5324 5325 5326
			.procname	= "hop_limit",
			.data		= &ipv6_devconf.hop_limit,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5327 5328
		},
		{
S
Stephen Hemminger 已提交
5329 5330 5331 5332
			.procname	= "mtu",
			.data		= &ipv6_devconf.mtu6,
			.maxlen		= sizeof(int),
			.mode		= 0644,
5333
			.proc_handler	= addrconf_sysctl_mtu,
L
Linus Torvalds 已提交
5334 5335
		},
		{
S
Stephen Hemminger 已提交
5336 5337 5338 5339 5340
			.procname	= "accept_ra",
			.data		= &ipv6_devconf.accept_ra,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5341 5342
		},
		{
S
Stephen Hemminger 已提交
5343 5344 5345 5346 5347
			.procname	= "accept_redirects",
			.data		= &ipv6_devconf.accept_redirects,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5348 5349
		},
		{
S
Stephen Hemminger 已提交
5350 5351 5352 5353 5354
			.procname	= "autoconf",
			.data		= &ipv6_devconf.autoconf,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5355 5356
		},
		{
S
Stephen Hemminger 已提交
5357 5358 5359 5360 5361
			.procname	= "dad_transmits",
			.data		= &ipv6_devconf.dad_transmits,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5362 5363
		},
		{
S
Stephen Hemminger 已提交
5364 5365 5366 5367 5368
			.procname	= "router_solicitations",
			.data		= &ipv6_devconf.rtr_solicits,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5369 5370
		},
		{
S
Stephen Hemminger 已提交
5371 5372 5373 5374 5375
			.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
5376 5377
		},
		{
S
Stephen Hemminger 已提交
5378 5379 5380 5381 5382
			.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
5383 5384
		},
		{
S
Stephen Hemminger 已提交
5385 5386 5387 5388 5389
			.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
5390
		},
5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406
		{
			.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
5407
		{
S
Stephen Hemminger 已提交
5408 5409 5410 5411 5412
			.procname	= "use_tempaddr",
			.data		= &ipv6_devconf.use_tempaddr,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5413 5414
		},
		{
S
Stephen Hemminger 已提交
5415 5416 5417 5418 5419
			.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
5420 5421
		},
		{
S
Stephen Hemminger 已提交
5422 5423 5424 5425 5426
			.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
5427 5428
		},
		{
S
Stephen Hemminger 已提交
5429 5430 5431 5432 5433
			.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
5434 5435
		},
		{
S
Stephen Hemminger 已提交
5436 5437 5438 5439 5440
			.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
5441 5442
		},
		{
S
Stephen Hemminger 已提交
5443 5444 5445 5446 5447
			.procname	= "max_addresses",
			.data		= &ipv6_devconf.max_addresses,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5448
		},
5449
		{
S
Stephen Hemminger 已提交
5450 5451 5452 5453 5454
			.procname	= "accept_ra_defrtr",
			.data		= &ipv6_devconf.accept_ra_defrtr,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5455
		},
5456
		{
S
Stephen Hemminger 已提交
5457 5458 5459 5460 5461
			.procname	= "accept_ra_pinfo",
			.data		= &ipv6_devconf.accept_ra_pinfo,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5462
		},
5463 5464
#ifdef CONFIG_IPV6_ROUTER_PREF
		{
S
Stephen Hemminger 已提交
5465 5466 5467 5468 5469
			.procname	= "accept_ra_rtr_pref",
			.data		= &ipv6_devconf.accept_ra_rtr_pref,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5470
		},
5471
		{
S
Stephen Hemminger 已提交
5472 5473 5474 5475 5476
			.procname	= "router_probe_interval",
			.data		= &ipv6_devconf.rtr_probe_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_jiffies,
5477
		},
N
Neil Horman 已提交
5478
#ifdef CONFIG_IPV6_ROUTE_INFO
5479
		{
S
Stephen Hemminger 已提交
5480 5481 5482 5483 5484
			.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,
5485 5486
		},
#endif
5487
#endif
5488
		{
S
Stephen Hemminger 已提交
5489 5490 5491 5492
			.procname	= "proxy_ndp",
			.data		= &ipv6_devconf.proxy_ndp,
			.maxlen		= sizeof(int),
			.mode		= 0644,
5493
			.proc_handler	= addrconf_sysctl_proxy_ndp,
5494
		},
5495
		{
S
Stephen Hemminger 已提交
5496 5497 5498 5499 5500
			.procname	= "accept_source_route",
			.data		= &ipv6_devconf.accept_source_route,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5501
		},
5502 5503
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
		{
S
Stephen Hemminger 已提交
5504 5505 5506 5507 5508
			.procname       = "optimistic_dad",
			.data           = &ipv6_devconf.optimistic_dad,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec,
5509 5510

		},
5511 5512 5513 5514 5515 5516 5517 5518
		{
			.procname       = "use_optimistic",
			.data           = &ipv6_devconf.use_optimistic,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec,

		},
5519 5520 5521
#endif
#ifdef CONFIG_IPV6_MROUTE
		{
S
Stephen Hemminger 已提交
5522 5523 5524 5525 5526
			.procname	= "mc_forwarding",
			.data		= &ipv6_devconf.mc_forwarding,
			.maxlen		= sizeof(int),
			.mode		= 0444,
			.proc_handler	= proc_dointvec,
5527
		},
5528
#endif
5529
		{
S
Stephen Hemminger 已提交
5530 5531 5532 5533 5534
			.procname	= "disable_ipv6",
			.data		= &ipv6_devconf.disable_ipv6,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= addrconf_sysctl_disable,
5535
		},
5536
		{
S
Stephen Hemminger 已提交
5537 5538 5539 5540 5541
			.procname	= "accept_dad",
			.data		= &ipv6_devconf.accept_dad,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5542
		},
5543 5544 5545 5546 5547 5548 5549
		{
			.procname       = "force_tllao",
			.data           = &ipv6_devconf.force_tllao,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec
		},
5550 5551 5552 5553 5554 5555 5556
		{
			.procname       = "ndisc_notify",
			.data           = &ipv6_devconf.ndisc_notify,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec
		},
5557 5558 5559 5560 5561 5562 5563
		{
			.procname	= "suppress_frag_ndisc",
			.data		= &ipv6_devconf.suppress_frag_ndisc,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec
		},
5564 5565 5566 5567 5568 5569 5570
		{
			.procname	= "accept_ra_from_local",
			.data		= &ipv6_devconf.accept_ra_from_local,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
		},
5571 5572 5573 5574 5575 5576 5577
		{
			.procname	= "accept_ra_mtu",
			.data		= &ipv6_devconf.accept_ra_mtu,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
		},
5578 5579 5580 5581 5582 5583 5584
		{
			.procname	= "stable_secret",
			.data		= &ipv6_devconf.stable_secret,
			.maxlen		= IPV6_MAX_STRLEN,
			.mode		= 0600,
			.proc_handler	= addrconf_sysctl_stable_secret,
		},
L
Linus Torvalds 已提交
5585
		{
5586
			/* sentinel */
L
Linus Torvalds 已提交
5587 5588 5589 5590
		}
	},
};

5591
static int __addrconf_sysctl_register(struct net *net, char *dev_name,
5592
		struct inet6_dev *idev, struct ipv6_devconf *p)
L
Linus Torvalds 已提交
5593 5594 5595
{
	int i;
	struct addrconf_sysctl_table *t;
5596
	char path[sizeof("net/ipv6/conf/") + IFNAMSIZ];
5597

A
Arnaldo Carvalho de Melo 已提交
5598
	t = kmemdup(&addrconf_sysctl, sizeof(*t), GFP_KERNEL);
L
Linus Torvalds 已提交
5599
	if (t == NULL)
5600 5601
		goto out;

S
Stephen Hemminger 已提交
5602
	for (i = 0; t->addrconf_vars[i].data; i++) {
S
Stephen Hemminger 已提交
5603
		t->addrconf_vars[i].data += (char *)p - (char *)&ipv6_devconf;
L
Linus Torvalds 已提交
5604
		t->addrconf_vars[i].extra1 = idev; /* embedded; no ref */
5605
		t->addrconf_vars[i].extra2 = net;
L
Linus Torvalds 已提交
5606 5607
	}

5608
	snprintf(path, sizeof(path), "net/ipv6/conf/%s", dev_name);
L
Linus Torvalds 已提交
5609

5610
	t->sysctl_header = register_net_sysctl(net, path, t->addrconf_vars);
L
Linus Torvalds 已提交
5611
	if (t->sysctl_header == NULL)
5612
		goto free;
5613 5614

	p->sysctl = t;
5615
	return 0;
L
Linus Torvalds 已提交
5616

5617
free:
L
Linus Torvalds 已提交
5618
	kfree(t);
5619
out:
5620
	return -ENOBUFS;
L
Linus Torvalds 已提交
5621 5622
}

5623 5624 5625 5626 5627 5628 5629 5630 5631
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;
5632
	unregister_net_sysctl_table(t->sysctl_header);
5633 5634 5635
	kfree(t);
}

5636
static int addrconf_sysctl_register(struct inet6_dev *idev)
5637
{
5638 5639 5640 5641 5642 5643 5644 5645 5646 5647 5648 5649 5650 5651 5652
	int err;

	if (!sysctl_dev_name_is_allowed(idev->dev->name))
		return -EINVAL;

	err = neigh_sysctl_register(idev->dev, idev->nd_parms,
				    &ndisc_ifinfo_sysctl_change);
	if (err)
		return err;
	err = __addrconf_sysctl_register(dev_net(idev->dev), idev->dev->name,
					 idev, &idev->cnf);
	if (err)
		neigh_sysctl_unregister(idev->nd_parms);

	return err;
5653 5654
}

5655
static void addrconf_sysctl_unregister(struct inet6_dev *idev)
L
Linus Torvalds 已提交
5656
{
5657 5658
	__addrconf_sysctl_unregister(&idev->cnf);
	neigh_sysctl_unregister(idev->nd_parms);
L
Linus Torvalds 已提交
5659 5660 5661 5662 5663
}


#endif

5664
static int __net_init addrconf_init_net(struct net *net)
5665
{
5666
	int err = -ENOMEM;
5667 5668
	struct ipv6_devconf *all, *dflt;

5669 5670 5671
	all = kmemdup(&ipv6_devconf, sizeof(ipv6_devconf), GFP_KERNEL);
	if (all == NULL)
		goto err_alloc_all;
5672

5673 5674 5675
	dflt = kmemdup(&ipv6_devconf_dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
	if (dflt == NULL)
		goto err_alloc_dflt;
5676

5677 5678 5679
	/* these will be inherited by all namespaces */
	dflt->autoconf = ipv6_defaults.autoconf;
	dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
5680

5681 5682 5683
	dflt->stable_secret.initialized = false;
	all->stable_secret.initialized = false;

5684 5685 5686 5687
	net->ipv6.devconf_all = all;
	net->ipv6.devconf_dflt = dflt;

#ifdef CONFIG_SYSCTL
5688
	err = __addrconf_sysctl_register(net, "all", NULL, all);
5689 5690 5691
	if (err < 0)
		goto err_reg_all;

5692
	err = __addrconf_sysctl_register(net, "default", NULL, dflt);
5693 5694 5695 5696 5697 5698 5699 5700 5701 5702 5703 5704 5705 5706 5707 5708 5709
	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;
}

5710
static void __net_exit addrconf_exit_net(struct net *net)
5711 5712 5713 5714 5715
{
#ifdef CONFIG_SYSCTL
	__addrconf_sysctl_unregister(net->ipv6.devconf_dflt);
	__addrconf_sysctl_unregister(net->ipv6.devconf_all);
#endif
Z
zhuyj 已提交
5716 5717
	kfree(net->ipv6.devconf_dflt);
	kfree(net->ipv6.devconf_all);
5718 5719 5720 5721 5722 5723 5724
}

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

5725
static struct rtnl_af_ops inet6_ops __read_mostly = {
T
Thomas Graf 已提交
5726 5727 5728
	.family		  = AF_INET6,
	.fill_link_af	  = inet6_fill_link_af,
	.get_link_af_size = inet6_get_link_af_size,
5729
	.validate_link_af = inet6_validate_link_af,
5730
	.set_link_af	  = inet6_set_link_af,
T
Thomas Graf 已提交
5731 5732
};

L
Linus Torvalds 已提交
5733 5734 5735 5736 5737 5738
/*
 *	Init / cleanup code
 */

int __init addrconf_init(void)
{
5739
	struct inet6_dev *idev;
5740
	int i, err;
5741

S
Stephen Hemminger 已提交
5742 5743
	err = ipv6_addr_label_init();
	if (err < 0) {
5744 5745
		pr_crit("%s: cannot initialize default policy table: %d\n",
			__func__, err);
5746
		goto out;
5747
	}
L
Linus Torvalds 已提交
5748

5749 5750 5751
	err = register_pernet_subsys(&addrconf_ops);
	if (err < 0)
		goto out_addrlabel;
5752

5753 5754 5755 5756 5757 5758
	addrconf_wq = create_workqueue("ipv6_addrconf");
	if (!addrconf_wq) {
		err = -ENOMEM;
		goto out_nowq;
	}

L
Linus Torvalds 已提交
5759 5760 5761 5762 5763 5764 5765 5766 5767 5768 5769 5770 5771 5772 5773 5774 5775 5776 5777
	/* 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();
5778
	idev = ipv6_add_dev(init_net.loopback_dev);
L
Linus Torvalds 已提交
5779
	rtnl_unlock();
5780 5781
	if (IS_ERR(idev)) {
		err = PTR_ERR(idev);
5782
		goto errlo;
5783
	}
L
Linus Torvalds 已提交
5784

5785 5786 5787
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		INIT_HLIST_HEAD(&inet6_addr_lst[i]);

L
Linus Torvalds 已提交
5788 5789
	register_netdevice_notifier(&ipv6_dev_notf);

5790
	addrconf_verify();
5791

5792
	rtnl_af_register(&inet6_ops);
T
Thomas Graf 已提交
5793

5794 5795
	err = __rtnl_register(PF_INET6, RTM_GETLINK, NULL, inet6_dump_ifinfo,
			      NULL);
5796 5797 5798 5799
	if (err < 0)
		goto errout;

	/* Only the first call to __rtnl_register can fail */
5800 5801 5802 5803 5804 5805 5806 5807
	__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);
5808
	__rtnl_register(PF_INET6, RTM_GETNETCONF, inet6_netconf_get_devconf,
5809
			inet6_netconf_dump_devconf, NULL);
5810

5811 5812
	ipv6_addr_label_rtnl_register();

L
Linus Torvalds 已提交
5813
	return 0;
5814
errout:
T
Thomas Graf 已提交
5815
	rtnl_af_unregister(&inet6_ops);
5816
	unregister_netdevice_notifier(&ipv6_dev_notf);
5817
errlo:
5818 5819
	destroy_workqueue(addrconf_wq);
out_nowq:
5820
	unregister_pernet_subsys(&addrconf_ops);
5821 5822 5823
out_addrlabel:
	ipv6_addr_label_cleanup();
out:
5824
	return err;
L
Linus Torvalds 已提交
5825 5826
}

5827
void addrconf_cleanup(void)
L
Linus Torvalds 已提交
5828
{
5829
	struct net_device *dev;
L
Linus Torvalds 已提交
5830 5831 5832
	int i;

	unregister_netdevice_notifier(&ipv6_dev_notf);
5833
	unregister_pernet_subsys(&addrconf_ops);
5834
	ipv6_addr_label_cleanup();
L
Linus Torvalds 已提交
5835 5836 5837

	rtnl_lock();

T
Thomas Graf 已提交
5838 5839
	__rtnl_af_unregister(&inet6_ops);

5840 5841 5842 5843 5844 5845 5846 5847
	/* 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
5848 5849 5850
	/*
	 *	Check hash table.
	 */
5851
	spin_lock_bh(&addrconf_hash_lock);
5852 5853
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		WARN_ON(!hlist_empty(&inet6_addr_lst[i]));
5854
	spin_unlock_bh(&addrconf_hash_lock);
5855
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
5856
	rtnl_unlock();
5857 5858

	destroy_workqueue(addrconf_wq);
L
Linus Torvalds 已提交
5859
}