addrconf.c 171.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5
/*
 *	IPv6 Address [auto]configuration
 *	Linux INET6 implementation
 *
 *	Authors:
6
 *	Pedro Roque		<roque@di.fc.ul.pt>
L
Linus Torvalds 已提交
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
 *	Alexey Kuznetsov	<kuznet@ms2.inr.ac.ru>
 *
 *	This program is free software; you can redistribute it and/or
 *      modify it under the terms of the GNU General Public License
 *      as published by the Free Software Foundation; either version
 *      2 of the License, or (at your option) any later version.
 */

/*
 *	Changes:
 *
 *	Janos Farkas			:	delete timer on ifdown
 *	<chexum@bankinf.banki.hu>
 *	Andi Kleen			:	kill double kfree on module
 *						unload.
 *	Maciej W. Rozycki		:	FDDI support
 *	sekiya@USAGI			:	Don't send too many RS
 *						packets.
 *	yoshfuji@USAGI			:       Fixed interval between DAD
 *						packets.
 *	YOSHIFUJI Hideaki @USAGI	:	improved accuracy of
 *						address validation timer.
 *	YOSHIFUJI Hideaki @USAGI	:	Privacy Extensions (RFC3041)
 *						support.
 *	Yuji SEKIYA @USAGI		:	Don't assign a same IPv6
 *						address on a same interface.
 *	YOSHIFUJI Hideaki @USAGI	:	ARCnet support
 *	YOSHIFUJI Hideaki @USAGI	:	convert /proc/net/if_inet6 to
 *						seq_file.
36 37 38
 *	YOSHIFUJI Hideaki @USAGI	:	improved source address
 *						selection; consider scope,
 *						status etc.
L
Linus Torvalds 已提交
39 40
 */

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

L
Linus Torvalds 已提交
43 44
#include <linux/errno.h>
#include <linux/types.h>
45
#include <linux/kernel.h>
46
#include <linux/sched/signal.h>
L
Linus Torvalds 已提交
47 48 49
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
50
#include <linux/inet.h>
L
Linus Torvalds 已提交
51 52
#include <linux/in6.h>
#include <linux/netdevice.h>
53
#include <linux/if_addr.h>
L
Linus Torvalds 已提交
54 55 56 57 58 59
#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>
60
#include <linux/slab.h>
L
Linus Torvalds 已提交
61 62 63
#ifdef CONFIG_SYSCTL
#include <linux/sysctl.h>
#endif
64
#include <linux/capability.h>
L
Linus Torvalds 已提交
65 66
#include <linux/delay.h>
#include <linux/notifier.h>
67
#include <linux/string.h>
E
Eric Dumazet 已提交
68
#include <linux/hash.h>
L
Linus Torvalds 已提交
69

70
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
71 72 73
#include <net/sock.h>
#include <net/snmp.h>

74
#include <net/6lowpan.h>
75
#include <net/firewire.h>
L
Linus Torvalds 已提交
76 77 78 79 80 81 82
#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>
83
#include <net/netlink.h>
84
#include <net/pkt_sched.h>
D
David Ahern 已提交
85
#include <net/l3mdev.h>
L
Linus Torvalds 已提交
86 87
#include <linux/if_tunnel.h>
#include <linux/rtnetlink.h>
88
#include <linux/netconf.h>
L
Linus Torvalds 已提交
89
#include <linux/random.h>
S
Stephen Hemminger 已提交
90
#include <linux/uaccess.h>
91
#include <asm/unaligned.h>
L
Linus Torvalds 已提交
92 93 94

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

#define	INFINITY_LIFE_TIME	0xFFFFFFFF
98

99 100 101
#define IPV6_MAX_STRLEN \
	sizeof("ffff:ffff:ffff:ffff:ffff:ffff:255.255.255.255")

102 103 104 105
static inline u32 cstamp_delta(unsigned long cstamp)
{
	return (cstamp - INITIAL_JIFFIES) * 100UL / HZ;
}
L
Linus Torvalds 已提交
106

107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
static inline s32 rfc3315_s14_backoff_init(s32 irt)
{
	/* multiply 'initial retransmission time' by 0.9 .. 1.1 */
	u64 tmp = (900000 + prandom_u32() % 200001) * (u64)irt;
	do_div(tmp, 1000000);
	return (s32)tmp;
}

static inline s32 rfc3315_s14_backoff_update(s32 rt, s32 mrt)
{
	/* multiply 'retransmission timeout' by 1.9 .. 2.1 */
	u64 tmp = (1900000 + prandom_u32() % 200001) * (u64)rt;
	do_div(tmp, 1000000);
	if ((s32)tmp > mrt) {
		/* multiply 'maximum retransmission time' by 0.9 .. 1.1 */
		tmp = (900000 + prandom_u32() % 200001) * (u64)mrt;
		do_div(tmp, 1000000);
	}
	return (s32)tmp;
}

L
Linus Torvalds 已提交
128
#ifdef CONFIG_SYSCTL
129
static int addrconf_sysctl_register(struct inet6_dev *idev);
130 131
static void addrconf_sysctl_unregister(struct inet6_dev *idev);
#else
132
static inline int addrconf_sysctl_register(struct inet6_dev *idev)
133
{
134
	return 0;
135 136 137 138 139
}

static inline void addrconf_sysctl_unregister(struct inet6_dev *idev)
{
}
L
Linus Torvalds 已提交
140 141
#endif

142 143
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
144

145
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev);
146
static int ipv6_count_addresses(const struct inet6_dev *idev);
147 148 149
static int ipv6_generate_stable_address(struct in6_addr *addr,
					u8 dad_count,
					const struct inet6_dev *idev);
L
Linus Torvalds 已提交
150

151 152
#define IN6_ADDR_HSIZE_SHIFT	8
#define IN6_ADDR_HSIZE		(1 << IN6_ADDR_HSIZE_SHIFT)
L
Linus Torvalds 已提交
153 154 155
/*
 *	Configured unicast address hash table
 */
156
static struct hlist_head inet6_addr_lst[IN6_ADDR_HSIZE];
157
static DEFINE_SPINLOCK(addrconf_hash_lock);
L
Linus Torvalds 已提交
158

159 160 161
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
162

163 164
static struct workqueue_struct *addrconf_wq;
static DECLARE_DELAYED_WORK(addr_chk_work, addrconf_verify_work);
L
Linus Torvalds 已提交
165 166 167 168

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

169 170
static void addrconf_type_change(struct net_device *dev,
				 unsigned long event);
L
Linus Torvalds 已提交
171 172
static int addrconf_ifdown(struct net_device *dev, int how);

173
static struct fib6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
174 175 176 177
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags);

178
static void addrconf_dad_start(struct inet6_ifaddr *ifp);
179
static void addrconf_dad_work(struct work_struct *w);
180 181
static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id,
				   bool send_na);
182
static void addrconf_dad_run(struct inet6_dev *idev, bool restart);
183
static void addrconf_rs_timer(struct timer_list *t);
L
Linus Torvalds 已提交
184 185 186
static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);
static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);

187
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
188 189
				struct prefix_info *pinfo);

190
static struct ipv6_devconf ipv6_devconf __read_mostly = {
L
Linus Torvalds 已提交
191 192 193 194 195 196 197
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
	.force_mld_version	= 0,
198 199
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
200 201 202
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
203
	.rtr_solicit_max_interval = RTR_SOLICITATION_MAX_INTERVAL,
L
Linus Torvalds 已提交
204
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
205
	.use_tempaddr		= 0,
L
Linus Torvalds 已提交
206 207 208 209 210
	.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,
211
	.accept_ra_defrtr	= 1,
212
	.accept_ra_from_local	= 0,
213
	.accept_ra_min_hop_limit= 1,
214
	.accept_ra_pinfo	= 1,
215 216
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
217
	.rtr_probe_interval	= 60 * HZ,
218
#ifdef CONFIG_IPV6_ROUTE_INFO
219
	.accept_ra_rt_info_min_plen = 0,
220 221
	.accept_ra_rt_info_max_plen = 0,
#endif
222
#endif
223
	.proxy_ndp		= 0,
224
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
225
	.disable_ipv6		= 0,
226
	.accept_dad		= 0,
227
	.suppress_frag_ndisc	= 1,
228
	.accept_ra_mtu		= 1,
229 230
	.stable_secret		= {
		.initialized = false,
231 232
	},
	.use_oif_addrs_only	= 0,
233
	.ignore_routes_with_linkdown = 0,
234
	.keep_addr_on_down	= 0,
235
	.seg6_enabled		= 0,
236 237 238
#ifdef CONFIG_IPV6_SEG6_HMAC
	.seg6_require_hmac	= 0,
#endif
239
	.enhanced_dad           = 1,
240
	.addr_gen_mode		= IN6_ADDR_GEN_MODE_EUI64,
241
	.disable_policy		= 0,
L
Linus Torvalds 已提交
242 243
};

244
static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
L
Linus Torvalds 已提交
245 246 247 248 249 250
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
251 252 253
	.force_mld_version	= 0,
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
254 255 256
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
257
	.rtr_solicit_max_interval = RTR_SOLICITATION_MAX_INTERVAL,
L
Linus Torvalds 已提交
258 259 260 261 262 263 264
	.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,
265
	.accept_ra_defrtr	= 1,
266
	.accept_ra_from_local	= 0,
267
	.accept_ra_min_hop_limit= 1,
268
	.accept_ra_pinfo	= 1,
269 270
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
271
	.rtr_probe_interval	= 60 * HZ,
272
#ifdef CONFIG_IPV6_ROUTE_INFO
273
	.accept_ra_rt_info_min_plen = 0,
274 275
	.accept_ra_rt_info_max_plen = 0,
#endif
276
#endif
277
	.proxy_ndp		= 0,
278
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
279
	.disable_ipv6		= 0,
280
	.accept_dad		= 1,
281
	.suppress_frag_ndisc	= 1,
282
	.accept_ra_mtu		= 1,
283 284 285
	.stable_secret		= {
		.initialized = false,
	},
286
	.use_oif_addrs_only	= 0,
287
	.ignore_routes_with_linkdown = 0,
288
	.keep_addr_on_down	= 0,
289
	.seg6_enabled		= 0,
290 291 292
#ifdef CONFIG_IPV6_SEG6_HMAC
	.seg6_require_hmac	= 0,
#endif
293
	.enhanced_dad           = 1,
294
	.addr_gen_mode		= IN6_ADDR_GEN_MODE_EUI64,
295
	.disable_policy		= 0,
L
Linus Torvalds 已提交
296 297
};

298 299
/* Check if link is ready: is it up and is a valid qdisc available */
static inline bool addrconf_link_ready(const struct net_device *dev)
300
{
301
	return netif_oper_up(dev) && !qdisc_tx_is_noop(dev);
302 303
}

304
static void addrconf_del_rs_timer(struct inet6_dev *idev)
L
Linus Torvalds 已提交
305
{
306 307 308 309
	if (del_timer(&idev->rs_timer))
		__in6_dev_put(idev);
}

310
static void addrconf_del_dad_work(struct inet6_ifaddr *ifp)
311
{
312
	if (cancel_delayed_work(&ifp->dad_work))
L
Linus Torvalds 已提交
313 314 315
		__in6_ifa_put(ifp);
}

316 317 318 319 320 321 322
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
323

324 325
static void addrconf_mod_dad_work(struct inet6_ifaddr *ifp,
				   unsigned long delay)
L
Linus Torvalds 已提交
326
{
327 328 329
	in6_ifa_hold(ifp);
	if (mod_delayed_work(addrconf_wq, &ifp->dad_work, delay))
		in6_ifa_put(ifp);
L
Linus Torvalds 已提交
330 331
}

332 333
static int snmp6_alloc_dev(struct inet6_dev *idev)
{
334 335
	int i;

W
WANG Cong 已提交
336 337
	idev->stats.ipv6 = alloc_percpu(struct ipstats_mib);
	if (!idev->stats.ipv6)
338
		goto err_ip;
339 340 341

	for_each_possible_cpu(i) {
		struct ipstats_mib *addrconf_stats;
W
WANG Cong 已提交
342
		addrconf_stats = per_cpu_ptr(idev->stats.ipv6, i);
343 344 345 346
		u64_stats_init(&addrconf_stats->syncp);
	}


347 348 349
	idev->stats.icmpv6dev = kzalloc(sizeof(struct icmpv6_mib_device),
					GFP_KERNEL);
	if (!idev->stats.icmpv6dev)
350
		goto err_icmp;
351 352 353
	idev->stats.icmpv6msgdev = kzalloc(sizeof(struct icmpv6msg_mib_device),
					   GFP_KERNEL);
	if (!idev->stats.icmpv6msgdev)
354
		goto err_icmpmsg;
355 356 357

	return 0;

358
err_icmpmsg:
359
	kfree(idev->stats.icmpv6dev);
360
err_icmp:
W
WANG Cong 已提交
361
	free_percpu(idev->stats.ipv6);
362
err_ip:
363
	return -ENOMEM;
364 365
}

366
static struct inet6_dev *ipv6_add_dev(struct net_device *dev)
L
Linus Torvalds 已提交
367 368
{
	struct inet6_dev *ndev;
369
	int err = -ENOMEM;
L
Linus Torvalds 已提交
370 371 372 373

	ASSERT_RTNL();

	if (dev->mtu < IPV6_MIN_MTU)
374
		return ERR_PTR(-EINVAL);
L
Linus Torvalds 已提交
375

376
	ndev = kzalloc(sizeof(struct inet6_dev), GFP_KERNEL);
377
	if (!ndev)
378
		return ERR_PTR(err);
379 380 381

	rwlock_init(&ndev->lock);
	ndev->dev = dev;
382
	INIT_LIST_HEAD(&ndev->addr_list);
383
	timer_setup(&ndev->rs_timer, addrconf_rs_timer, 0);
384
	memcpy(&ndev->cnf, dev_net(dev)->ipv6.devconf_dflt, sizeof(ndev->cnf));
385 386

	if (ndev->cnf.stable_secret.initialized)
387
		ndev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
388

389 390
	ndev->cnf.mtu6 = dev->mtu;
	ndev->nd_parms = neigh_parms_alloc(dev, &nd_tbl);
391
	if (!ndev->nd_parms) {
392
		kfree(ndev);
393
		return ERR_PTR(err);
394
	}
395 396
	if (ndev->cnf.forwarding)
		dev_disable_lro(dev);
397 398
	/* We refer to the device */
	dev_hold(dev);
L
Linus Torvalds 已提交
399

400
	if (snmp6_alloc_dev(ndev) < 0) {
401 402
		netdev_dbg(dev, "%s: cannot allocate memory for statistics\n",
			   __func__);
403
		neigh_parms_release(&nd_tbl, ndev->nd_parms);
R
Roy Li 已提交
404 405
		dev_put(dev);
		kfree(ndev);
406
		return ERR_PTR(err);
407
	}
L
Linus Torvalds 已提交
408

409
	if (snmp6_register_dev(ndev) < 0) {
410 411
		netdev_dbg(dev, "%s: cannot create /proc/net/dev_snmp6/%s\n",
			   __func__, dev->name);
412
		goto err_release;
413 414
	}

415
	/* One reference from device. */
416
	refcount_set(&ndev->refcnt, 1);
L
Linus Torvalds 已提交
417

418 419 420
	if (dev->flags & (IFF_NOARP | IFF_LOOPBACK))
		ndev->cnf.accept_dad = -1;

A
Amerigo Wang 已提交
421
#if IS_ENABLED(CONFIG_IPV6_SIT)
422
	if (dev->type == ARPHRD_SIT && (dev->priv_flags & IFF_ISATAP)) {
423
		pr_info("%s: Disabled Multicast RS\n", dev->name);
424 425 426 427
		ndev->cnf.rtr_solicits = 0;
	}
#endif

428
	INIT_LIST_HEAD(&ndev->tempaddr_list);
J
Jiri Bohac 已提交
429
	ndev->desync_factor = U32_MAX;
430 431
	if ((dev->flags&IFF_LOOPBACK) ||
	    dev->type == ARPHRD_TUNNEL ||
432
	    dev->type == ARPHRD_TUNNEL6 ||
433 434
	    dev->type == ARPHRD_SIT ||
	    dev->type == ARPHRD_NONE) {
435
		ndev->cnf.use_tempaddr = -1;
436 437
	} else
		ipv6_regen_rndid(ndev);
438

439
	ndev->token = in6addr_any;
L
Linus Torvalds 已提交
440

441
	if (netif_running(dev) && addrconf_link_ready(dev))
442 443
		ndev->if_flags |= IF_READY;

444 445
	ipv6_mc_init_dev(ndev);
	ndev->tstamp = jiffies;
446 447 448
	err = addrconf_sysctl_register(ndev);
	if (err) {
		ipv6_mc_destroy_dev(ndev);
449
		snmp6_unregister_dev(ndev);
450 451
		goto err_release;
	}
452
	/* protected by rtnl_lock */
453
	rcu_assign_pointer(dev->ip6_ptr, ndev);
454

455 456 457
	/* Join interface-local all-node multicast group */
	ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allnodes);

458
	/* Join all-node multicast group */
459
	ipv6_dev_mc_inc(dev, &in6addr_linklocal_allnodes);
460

461
	/* Join all-router multicast group if forwarding is set */
L
Li Wei 已提交
462
	if (ndev->cnf.forwarding && (dev->flags & IFF_MULTICAST))
463 464
		ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);

L
Linus Torvalds 已提交
465
	return ndev;
466 467 468 469 470 471

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
472 473
}

474
static struct inet6_dev *ipv6_find_idev(struct net_device *dev)
L
Linus Torvalds 已提交
475 476 477 478 479
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
480 481 482
	idev = __in6_dev_get(dev);
	if (!idev) {
		idev = ipv6_add_dev(dev);
483
		if (IS_ERR(idev))
L
Linus Torvalds 已提交
484 485
			return NULL;
	}
486

L
Linus Torvalds 已提交
487 488 489 490 491
	if (dev->flags&IFF_UP)
		ipv6_mc_up(idev);
	return idev;
}

492 493 494 495
static int inet6_netconf_msgsize_devconf(int type)
{
	int size =  NLMSG_ALIGN(sizeof(struct netconfmsg))
		    + nla_total_size(4);	/* NETCONFA_IFINDEX */
496
	bool all = false;
497

498 499 500 501
	if (type == NETCONFA_ALL)
		all = true;

	if (all || type == NETCONFA_FORWARDING)
502
		size += nla_total_size(4);
503
#ifdef CONFIG_IPV6_MROUTE
504
	if (all || type == NETCONFA_MC_FORWARDING)
505
		size += nla_total_size(4);
506
#endif
507
	if (all || type == NETCONFA_PROXY_NEIGH)
508
		size += nla_total_size(4);
509

510
	if (all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN)
511 512
		size += nla_total_size(4);

513 514 515 516 517 518 519 520 521 522
	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;
523
	bool all = false;
524 525 526

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct netconfmsg),
			flags);
527
	if (!nlh)
528 529
		return -EMSGSIZE;

530 531 532
	if (type == NETCONFA_ALL)
		all = true;

533 534 535 536 537 538
	ncm = nlmsg_data(nlh);
	ncm->ncm_family = AF_INET6;

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

539 540 541
	if (!devconf)
		goto out;

542
	if ((all || type == NETCONFA_FORWARDING) &&
543 544
	    nla_put_s32(skb, NETCONFA_FORWARDING, devconf->forwarding) < 0)
		goto nla_put_failure;
545
#ifdef CONFIG_IPV6_MROUTE
546
	if ((all || type == NETCONFA_MC_FORWARDING) &&
547 548 549
	    nla_put_s32(skb, NETCONFA_MC_FORWARDING,
			devconf->mc_forwarding) < 0)
		goto nla_put_failure;
550
#endif
551
	if ((all || type == NETCONFA_PROXY_NEIGH) &&
552 553 554
	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
		goto nla_put_failure;

555
	if ((all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN) &&
556 557 558 559
	    nla_put_s32(skb, NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
			devconf->ignore_routes_with_linkdown) < 0)
		goto nla_put_failure;

560
out:
561 562
	nlmsg_end(skb, nlh);
	return 0;
563 564 565 566 567 568

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

569 570
void inet6_netconf_notify_devconf(struct net *net, int event, int type,
				  int ifindex, struct ipv6_devconf *devconf)
571 572 573 574
{
	struct sk_buff *skb;
	int err = -ENOBUFS;

575
	skb = nlmsg_new(inet6_netconf_msgsize_devconf(type), GFP_KERNEL);
576
	if (!skb)
577 578 579
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf, 0, 0,
580
					 event, 0, type);
581 582 583 584 585 586
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
587
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_NETCONF, NULL, GFP_KERNEL);
588 589
	return;
errout:
590
	rtnl_set_sk_err(net, RTNLGRP_IPV6_NETCONF, err);
591 592
}

593 594 595
static const struct nla_policy devconf_ipv6_policy[NETCONFA_MAX+1] = {
	[NETCONFA_IFINDEX]	= { .len = sizeof(int) },
	[NETCONFA_FORWARDING]	= { .len = sizeof(int) },
596
	[NETCONFA_PROXY_NEIGH]	= { .len = sizeof(int) },
597
	[NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN]	= { .len = sizeof(int) },
598 599
};

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636
static int inet6_netconf_valid_get_req(struct sk_buff *skb,
				       const struct nlmsghdr *nlh,
				       struct nlattr **tb,
				       struct netlink_ext_ack *extack)
{
	int i, err;

	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(struct netconfmsg))) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid header for netconf get request");
		return -EINVAL;
	}

	if (!netlink_strict_get_check(skb))
		return nlmsg_parse(nlh, sizeof(struct netconfmsg), tb,
				   NETCONFA_MAX, devconf_ipv6_policy, extack);

	err = nlmsg_parse_strict(nlh, sizeof(struct netconfmsg), tb,
				 NETCONFA_MAX, devconf_ipv6_policy, extack);
	if (err)
		return err;

	for (i = 0; i <= NETCONFA_MAX; i++) {
		if (!tb[i])
			continue;

		switch (i) {
		case NETCONFA_IFINDEX:
			break;
		default:
			NL_SET_ERR_MSG_MOD(extack, "Unsupported attribute in netconf get request");
			return -EINVAL;
		}
	}

	return 0;
}

637
static int inet6_netconf_get_devconf(struct sk_buff *in_skb,
638 639
				     struct nlmsghdr *nlh,
				     struct netlink_ext_ack *extack)
640 641 642
{
	struct net *net = sock_net(in_skb->sk);
	struct nlattr *tb[NETCONFA_MAX+1];
643 644
	struct inet6_dev *in6_dev = NULL;
	struct net_device *dev = NULL;
645 646 647 648 649
	struct sk_buff *skb;
	struct ipv6_devconf *devconf;
	int ifindex;
	int err;

650
	err = inet6_netconf_valid_get_req(in_skb, nlh, tb, extack);
651
	if (err < 0)
652
		return err;
653 654

	if (!tb[NETCONFA_IFINDEX])
655
		return -EINVAL;
656

657
	err = -EINVAL;
658 659 660 661 662 663 664 665 666
	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:
667
		dev = dev_get_by_index(net, ifindex);
668
		if (!dev)
669 670
			return -EINVAL;
		in6_dev = in6_dev_get(dev);
671
		if (!in6_dev)
672 673 674 675 676 677
			goto errout;
		devconf = &in6_dev->cnf;
		break;
	}

	err = -ENOBUFS;
678
	skb = nlmsg_new(inet6_netconf_msgsize_devconf(NETCONFA_ALL), GFP_KERNEL);
679
	if (!skb)
680 681 682 683 684
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf,
					 NETLINK_CB(in_skb).portid,
					 nlh->nlmsg_seq, RTM_NEWNETCONF, 0,
685
					 NETCONFA_ALL);
686 687 688 689 690 691 692 693
	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:
694 695 696 697
	if (in6_dev)
		in6_dev_put(in6_dev);
	if (dev)
		dev_put(dev);
698 699 700
	return err;
}

701 702 703
static int inet6_netconf_dump_devconf(struct sk_buff *skb,
				      struct netlink_callback *cb)
{
704
	const struct nlmsghdr *nlh = cb->nlh;
705 706 707 708 709 710 711
	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;

712 713 714 715 716 717 718 719 720 721 722 723 724 725 726
	if (cb->strict_check) {
		struct netlink_ext_ack *extack = cb->extack;
		struct netconfmsg *ncm;

		if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*ncm))) {
			NL_SET_ERR_MSG_MOD(extack, "Invalid header for netconf dump request");
			return -EINVAL;
		}

		if (nlmsg_attrlen(nlh, sizeof(*ncm))) {
			NL_SET_ERR_MSG_MOD(extack, "Invalid data after header in netconf dump request");
			return -EINVAL;
		}
	}

727 728 729 730 731 732 733
	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();
734 735
		cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^
			  net->dev_base_seq;
736 737 738 739 740 741 742 743 744 745
		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,
746
						       nlh->nlmsg_seq,
747 748
						       RTM_NEWNETCONF,
						       NLM_F_MULTI,
749
						       NETCONFA_ALL) < 0) {
750 751 752
				rcu_read_unlock();
				goto done;
			}
753
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
754 755 756 757 758 759 760 761 762
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,
763
					       nlh->nlmsg_seq,
764
					       RTM_NEWNETCONF, NLM_F_MULTI,
765
					       NETCONFA_ALL) < 0)
766 767 768 769 770 771 772 773
			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,
774
					       nlh->nlmsg_seq,
775
					       RTM_NEWNETCONF, NLM_F_MULTI,
776
					       NETCONFA_ALL) < 0)
777 778 779 780 781 782 783 784 785 786 787
			goto done;
		else
			h++;
	}
done:
	cb->args[0] = h;
	cb->args[1] = idx;

	return skb->len;
}

L
Linus Torvalds 已提交
788 789 790 791 792 793 794 795 796
#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;
797 798
	if (idev->cnf.forwarding)
		dev_disable_lro(dev);
A
Amerigo Wang 已提交
799
	if (dev->flags & IFF_MULTICAST) {
800
		if (idev->cnf.forwarding) {
801
			ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);
802 803 804
			ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_inc(dev, &in6addr_sitelocal_allrouters);
		} else {
805
			ipv6_dev_mc_dec(dev, &in6addr_linklocal_allrouters);
806 807 808
			ipv6_dev_mc_dec(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_dec(dev, &in6addr_sitelocal_allrouters);
		}
L
Linus Torvalds 已提交
809
	}
810 811

	list_for_each_entry(ifa, &idev->addr_list, if_list) {
M
Michal Wrobel 已提交
812 813
		if (ifa->flags&IFA_F_TENTATIVE)
			continue;
L
Linus Torvalds 已提交
814 815 816 817 818
		if (idev->cnf.forwarding)
			addrconf_join_anycast(ifa);
		else
			addrconf_leave_anycast(ifa);
	}
819 820
	inet6_netconf_notify_devconf(dev_net(dev), RTM_NEWNETCONF,
				     NETCONFA_FORWARDING,
821
				     dev->ifindex, &idev->cnf);
L
Linus Torvalds 已提交
822 823 824
}


825
static void addrconf_forward_change(struct net *net, __s32 newf)
L
Linus Torvalds 已提交
826 827 828 829
{
	struct net_device *dev;
	struct inet6_dev *idev;

830
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
831 832
		idev = __in6_dev_get(dev);
		if (idev) {
833 834
			int changed = (!idev->cnf.forwarding) ^ (!newf);
			idev->cnf.forwarding = newf;
L
Linus Torvalds 已提交
835 836 837 838 839
			if (changed)
				dev_forward_change(idev);
		}
	}
}
840

841
static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
842
{
843
	struct net *net;
844 845 846 847
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
848 849

	net = (struct net *)table->extra2;
850 851
	old = *p;
	*p = newf;
852

853
	if (p == &net->ipv6.devconf_dflt->forwarding) {
854
		if ((!newf) ^ (!old))
855 856
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
857 858
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
859 860
		rtnl_unlock();
		return 0;
E
Eric W. Biederman 已提交
861
	}
862

863
	if (p == &net->ipv6.devconf_all->forwarding) {
864 865
		int old_dflt = net->ipv6.devconf_dflt->forwarding;

866
		net->ipv6.devconf_dflt->forwarding = newf;
867
		if ((!newf) ^ (!old_dflt))
868 869
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
870 871 872
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);

873
		addrconf_forward_change(net, newf);
874
		if ((!newf) ^ (!old))
875 876
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
877 878
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
879
	} else if ((!newf) ^ (!old))
880
		dev_forward_change((struct inet6_dev *)table->extra1);
881
	rtnl_unlock();
882

883
	if (newf)
884
		rt6_purge_dflt_routers(net);
885
	return 1;
886
}
887 888 889 890 891 892 893 894 895 896 897 898 899 900

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

	for_each_netdev(net, dev) {
		idev = __in6_dev_get(dev);
		if (idev) {
			int changed = (!idev->cnf.ignore_routes_with_linkdown) ^ (!newf);

			idev->cnf.ignore_routes_with_linkdown = newf;
			if (changed)
				inet6_netconf_notify_devconf(dev_net(dev),
901
							     RTM_NEWNETCONF,
902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923
							     NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
							     dev->ifindex,
							     &idev->cnf);
		}
	}
}

static int addrconf_fixup_linkdown(struct ctl_table *table, int *p, int newf)
{
	struct net *net;
	int old;

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

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

	if (p == &net->ipv6.devconf_dflt->ignore_routes_with_linkdown) {
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net,
924
						     RTM_NEWNETCONF,
925 926 927 928 929 930 931 932 933 934 935 936
						     NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
		rtnl_unlock();
		return 0;
	}

	if (p == &net->ipv6.devconf_all->ignore_routes_with_linkdown) {
		net->ipv6.devconf_dflt->ignore_routes_with_linkdown = newf;
		addrconf_linkdown_change(net, newf);
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net,
937
						     RTM_NEWNETCONF,
938 939 940 941 942 943 944 945 946
						     NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
	}
	rtnl_unlock();

	return 1;
}

L
Linus Torvalds 已提交
947 948
#endif

949
/* Nobody refers to this ifaddr, destroy it */
L
Linus Torvalds 已提交
950 951
void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp)
{
952
	WARN_ON(!hlist_unhashed(&ifp->addr_lst));
953

L
Linus Torvalds 已提交
954
#ifdef NET_REFCNT_DEBUG
955
	pr_debug("%s\n", __func__);
L
Linus Torvalds 已提交
956 957 958 959
#endif

	in6_dev_put(ifp->idev);

960 961 962
	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
963

964
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
965
		pr_warn("Freeing alive inet6 address %p\n", ifp);
L
Linus Torvalds 已提交
966 967 968
		return;
	}

969
	kfree_rcu(ifp, rcu);
L
Linus Torvalds 已提交
970 971
}

B
Brian Haley 已提交
972 973 974
static void
ipv6_link_dev_addr(struct inet6_dev *idev, struct inet6_ifaddr *ifp)
{
975
	struct list_head *p;
976
	int ifp_scope = ipv6_addr_src_scope(&ifp->addr);
B
Brian Haley 已提交
977 978 979 980 981

	/*
	 * Each device address list is sorted in order of scope -
	 * global before linklocal.
	 */
982 983 984
	list_for_each(p, &idev->addr_list) {
		struct inet6_ifaddr *ifa
			= list_entry(p, struct inet6_ifaddr, if_list);
985
		if (ifp_scope >= ipv6_addr_src_scope(&ifa->addr))
B
Brian Haley 已提交
986 987 988
			break;
	}

989
	list_add_tail_rcu(&ifp->if_list, p);
B
Brian Haley 已提交
990 991
}

992
static u32 inet6_addr_hash(const struct net *net, const struct in6_addr *addr)
993
{
994 995 996
	u32 val = ipv6_addr_hash(addr) ^ net_hash_mix(net);

	return hash_32(val, IN6_ADDR_HSIZE_SHIFT);
997 998
}

999
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
1000
			       struct net_device *dev, unsigned int hash)
1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014
{
	struct inet6_ifaddr *ifp;

	hlist_for_each_entry(ifp, &inet6_addr_lst[hash], addr_lst) {
		if (!net_eq(dev_net(ifp->idev->dev), net))
			continue;
		if (ipv6_addr_equal(&ifp->addr, addr)) {
			if (!dev || ifp->idev->dev == dev)
				return true;
		}
	}
	return false;
}

1015 1016
static int ipv6_add_addr_hash(struct net_device *dev, struct inet6_ifaddr *ifa)
{
1017
	unsigned int hash = inet6_addr_hash(dev_net(dev), &ifa->addr);
1018 1019 1020 1021 1022
	int err = 0;

	spin_lock(&addrconf_hash_lock);

	/* Ignore adding duplicate addresses on an interface */
1023
	if (ipv6_chk_same_addr(dev_net(dev), &ifa->addr, dev, hash)) {
1024
		netdev_dbg(dev, "ipv6_add_addr: already assigned\n");
1025
		err = -EEXIST;
1026 1027
	} else {
		hlist_add_head_rcu(&ifa->addr_lst, &inet6_addr_lst[hash]);
1028 1029 1030 1031 1032 1033 1034
	}

	spin_unlock(&addrconf_hash_lock);

	return err;
}

L
Linus Torvalds 已提交
1035 1036 1037
/* On success it returns ifp with increased reference count */

static struct inet6_ifaddr *
1038
ipv6_add_addr(struct inet6_dev *idev, struct ifa6_config *cfg,
1039
	      bool can_block, struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
1040
{
1041
	gfp_t gfp_flags = can_block ? GFP_KERNEL : GFP_ATOMIC;
1042
	int addr_type = ipv6_addr_type(cfg->pfx);
1043
	struct net *net = dev_net(idev->dev);
L
Linus Torvalds 已提交
1044
	struct inet6_ifaddr *ifa = NULL;
1045
	struct fib6_info *f6i = NULL;
L
Linus Torvalds 已提交
1046
	int err = 0;
1047 1048 1049 1050

	if (addr_type == IPV6_ADDR_ANY ||
	    addr_type & IPV6_ADDR_MULTICAST ||
	    (!(idev->dev->flags & IFF_LOOPBACK) &&
1051
	     !netif_is_l3_master(idev->dev) &&
1052 1053
	     addr_type & IPV6_ADDR_LOOPBACK))
		return ERR_PTR(-EADDRNOTAVAIL);
L
Linus Torvalds 已提交
1054 1055 1056

	if (idev->dead) {
		err = -ENODEV;			/*XXX*/
1057
		goto out;
L
Linus Torvalds 已提交
1058 1059
	}

1060
	if (idev->cnf.disable_ipv6) {
1061
		err = -EACCES;
1062
		goto out;
1063 1064
	}

1065 1066 1067 1068 1069
	/* validator notifier needs to be blocking;
	 * do not call in atomic context
	 */
	if (can_block) {
		struct in6_validator_info i6vi = {
1070
			.i6vi_addr = *cfg->pfx,
1071
			.i6vi_dev = idev,
1072
			.extack = extack,
1073 1074 1075 1076 1077 1078 1079
		};

		err = inet6addr_validator_notifier_call_chain(NETDEV_UP, &i6vi);
		err = notifier_to_errno(err);
		if (err < 0)
			goto out;
	}
L
Linus Torvalds 已提交
1080

1081
	ifa = kzalloc(sizeof(*ifa), gfp_flags);
1082
	if (!ifa) {
L
Linus Torvalds 已提交
1083 1084 1085 1086
		err = -ENOBUFS;
		goto out;
	}

1087
	f6i = addrconf_f6i_alloc(net, idev, cfg->pfx, false, gfp_flags);
1088 1089 1090
	if (IS_ERR(f6i)) {
		err = PTR_ERR(f6i);
		f6i = NULL;
L
Linus Torvalds 已提交
1091 1092 1093
		goto out;
	}

1094 1095
	if (net->ipv6.devconf_all->disable_policy ||
	    idev->cnf.disable_policy)
1096
		f6i->dst_nopolicy = true;
1097

1098 1099
	neigh_parms_data_state_setall(idev->nd_parms);

1100 1101 1102
	ifa->addr = *cfg->pfx;
	if (cfg->peer_pfx)
		ifa->peer_addr = *cfg->peer_pfx;
L
Linus Torvalds 已提交
1103 1104

	spin_lock_init(&ifa->lock);
1105
	INIT_DELAYED_WORK(&ifa->dad_work, addrconf_dad_work);
1106
	INIT_HLIST_NODE(&ifa->addr_lst);
1107 1108
	ifa->scope = cfg->scope;
	ifa->prefix_len = cfg->plen;
1109
	ifa->rt_priority = cfg->rt_priority;
1110
	ifa->flags = cfg->ifa_flags;
1111
	/* No need to add the TENTATIVE flag for addresses with NODAD */
1112
	if (!(cfg->ifa_flags & IFA_F_NODAD))
1113
		ifa->flags |= IFA_F_TENTATIVE;
1114 1115
	ifa->valid_lft = cfg->valid_lft;
	ifa->prefered_lft = cfg->preferred_lft;
L
Linus Torvalds 已提交
1116
	ifa->cstamp = ifa->tstamp = jiffies;
1117
	ifa->tokenized = false;
L
Linus Torvalds 已提交
1118

1119
	ifa->rt = f6i;
1120

L
Linus Torvalds 已提交
1121
	ifa->idev = idev;
1122 1123
	in6_dev_hold(idev);

L
Linus Torvalds 已提交
1124
	/* For caller */
1125
	refcount_set(&ifa->refcnt, 1);
L
Linus Torvalds 已提交
1126

1127
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
1128

1129 1130 1131 1132 1133
	err = ipv6_add_addr_hash(idev->dev, ifa);
	if (err < 0) {
		rcu_read_unlock_bh();
		goto out;
	}
L
Linus Torvalds 已提交
1134 1135

	write_lock(&idev->lock);
1136

L
Linus Torvalds 已提交
1137
	/* Add to inet6_dev unicast addr list. */
B
Brian Haley 已提交
1138
	ipv6_link_dev_addr(idev, ifa);
L
Linus Torvalds 已提交
1139 1140

	if (ifa->flags&IFA_F_TEMPORARY) {
1141
		list_add(&ifa->tmp_list, &idev->tempaddr_list);
L
Linus Torvalds 已提交
1142 1143 1144 1145 1146
		in6_ifa_hold(ifa);
	}

	in6_ifa_hold(ifa);
	write_unlock(&idev->lock);
1147

1148
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1149

1150 1151 1152
	inet6addr_notifier_call_chain(NETDEV_UP, ifa);
out:
	if (unlikely(err < 0)) {
1153
		fib6_info_release(f6i);
1154

1155 1156 1157 1158 1159
		if (ifa) {
			if (ifa->idev)
				in6_dev_put(ifa->idev);
			kfree(ifa);
		}
L
Linus Torvalds 已提交
1160 1161 1162 1163 1164 1165
		ifa = ERR_PTR(err);
	}

	return ifa;
}

1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229
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)
{
1230
	struct fib6_info *f6i;
1231

1232
	f6i = addrconf_get_prefix_route(&ifp->addr,
1233 1234 1235
				       ifp->prefix_len,
				       ifp->idev->dev,
				       0, RTF_GATEWAY | RTF_DEFAULT);
1236
	if (f6i) {
1237
		if (del_rt)
1238
			ip6_del_rt(dev_net(ifp->idev->dev), f6i);
1239
		else {
1240 1241 1242
			if (!(f6i->fib6_flags & RTF_EXPIRES))
				fib6_set_expires(f6i, expires);
			fib6_info_release(f6i);
1243 1244 1245 1246 1247
		}
	}
}


L
Linus Torvalds 已提交
1248 1249 1250 1251
/* This function wants to get referenced ifp and releases it before return */

static void ipv6_del_addr(struct inet6_ifaddr *ifp)
{
1252
	int state;
1253 1254
	enum cleanup_prefix_rt_t action = CLEANUP_PREFIX_RT_NOP;
	unsigned long expires;
L
Linus Torvalds 已提交
1255

1256 1257
	ASSERT_RTNL();

1258
	spin_lock_bh(&ifp->lock);
1259
	state = ifp->state;
1260
	ifp->state = INET6_IFADDR_STATE_DEAD;
1261
	spin_unlock_bh(&ifp->lock);
1262 1263 1264

	if (state == INET6_IFADDR_STATE_DEAD)
		goto out;
L
Linus Torvalds 已提交
1265

1266 1267 1268
	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
1269

1270
	write_lock_bh(&ifp->idev->lock);
1271

L
Linus Torvalds 已提交
1272
	if (ifp->flags&IFA_F_TEMPORARY) {
1273 1274 1275 1276
		list_del(&ifp->tmp_list);
		if (ifp->ifpub) {
			in6_ifa_put(ifp->ifpub);
			ifp->ifpub = NULL;
L
Linus Torvalds 已提交
1277
		}
1278
		__in6_ifa_put(ifp);
L
Linus Torvalds 已提交
1279 1280
	}

1281 1282
	if (ifp->flags & IFA_F_PERMANENT && !(ifp->flags & IFA_F_NOPREFIXROUTE))
		action = check_cleanup_prefix_route(ifp, &expires);
1283

1284
	list_del_rcu(&ifp->if_list);
1285 1286 1287
	__in6_ifa_put(ifp);

	write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
1288

1289
	addrconf_del_dad_work(ifp);
1290

L
Linus Torvalds 已提交
1291 1292
	ipv6_ifa_notify(RTM_DELADDR, ifp);

1293
	inet6addr_notifier_call_chain(NETDEV_DOWN, ifp);
L
Linus Torvalds 已提交
1294

1295 1296 1297
	if (action != CLEANUP_PREFIX_RT_NOP) {
		cleanup_prefix_route(ifp, expires,
			action == CLEANUP_PREFIX_RT_DEL);
L
Linus Torvalds 已提交
1298 1299
	}

1300 1301
	/* clean up prefsrc entries */
	rt6_remove_prefsrc(ifp);
1302
out:
L
Linus Torvalds 已提交
1303 1304 1305
	in6_ifa_put(ifp);
}

E
Eric Dumazet 已提交
1306 1307 1308
static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp,
				struct inet6_ifaddr *ift,
				bool block)
L
Linus Torvalds 已提交
1309 1310 1311
{
	struct inet6_dev *idev = ifp->idev;
	struct in6_addr addr, *tmpaddr;
1312
	unsigned long tmp_tstamp, age;
1313
	unsigned long regen_advance;
1314
	struct ifa6_config cfg;
L
Linus Torvalds 已提交
1315
	int ret = 0;
1316
	unsigned long now = jiffies;
J
Jiri Bohac 已提交
1317
	long max_desync_factor;
1318
	s32 cnf_temp_preferred_lft;
L
Linus Torvalds 已提交
1319

1320
	write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331
	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) {
1332
		write_unlock_bh(&idev->lock);
1333
		pr_info("%s: use_tempaddr is disabled\n", __func__);
L
Linus Torvalds 已提交
1334 1335 1336 1337 1338 1339 1340 1341
		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);
1342
		write_unlock_bh(&idev->lock);
1343 1344
		pr_warn("%s: regeneration time exceeded - disabled temporary address support\n",
			__func__);
L
Linus Torvalds 已提交
1345 1346 1347 1348 1349 1350
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}
	in6_ifa_hold(ifp);
	memcpy(addr.s6_addr, ifp->addr.s6_addr, 8);
1351
	ipv6_try_regen_rndid(idev, tmpaddr);
L
Linus Torvalds 已提交
1352
	memcpy(&addr.s6_addr[8], idev->rndid, 8);
1353
	age = (now - ifp->tstamp) / HZ;
J
Jiri Bohac 已提交
1354 1355 1356 1357 1358 1359 1360 1361

	regen_advance = idev->cnf.regen_max_retry *
			idev->cnf.dad_transmits *
			NEIGH_VAR(idev->nd_parms, RETRANS_TIME) / HZ;

	/* recalculate max_desync_factor each time and update
	 * idev->desync_factor if it's larger
	 */
1362
	cnf_temp_preferred_lft = READ_ONCE(idev->cnf.temp_prefered_lft);
J
Jiri Bohac 已提交
1363 1364
	max_desync_factor = min_t(__u32,
				  idev->cnf.max_desync_factor,
1365
				  cnf_temp_preferred_lft - regen_advance);
J
Jiri Bohac 已提交
1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376

	if (unlikely(idev->desync_factor > max_desync_factor)) {
		if (max_desync_factor > 0) {
			get_random_bytes(&idev->desync_factor,
					 sizeof(idev->desync_factor));
			idev->desync_factor %= max_desync_factor;
		} else {
			idev->desync_factor = 0;
		}
	}

1377
	memset(&cfg, 0, sizeof(cfg));
1378
	cfg.valid_lft = min_t(__u32, ifp->valid_lft,
1379
			      idev->cnf.temp_valid_lft + age);
1380 1381 1382 1383
	cfg.preferred_lft = cnf_temp_preferred_lft + age - idev->desync_factor;
	cfg.preferred_lft = min_t(__u32, ifp->prefered_lft, cfg.preferred_lft);

	cfg.plen = ifp->prefix_len;
L
Linus Torvalds 已提交
1384 1385 1386
	tmp_tstamp = ifp->tstamp;
	spin_unlock_bh(&ifp->lock);

1387
	write_unlock_bh(&idev->lock);
1388

1389 1390 1391 1392
	/* 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.
1393 1394
	 * Use age calculation as in addrconf_verify to avoid unnecessary
	 * temporary addresses being generated.
1395
	 */
1396
	age = (now - tmp_tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
1397
	if (cfg.preferred_lft <= regen_advance + age) {
1398 1399 1400 1401 1402 1403
		in6_ifa_put(ifp);
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}

1404
	cfg.ifa_flags = IFA_F_TEMPORARY;
1405 1406
	/* set in addrconf_prefix_rcv() */
	if (ifp->flags & IFA_F_OPTIMISTIC)
1407
		cfg.ifa_flags |= IFA_F_OPTIMISTIC;
1408

1409 1410
	cfg.pfx = &addr;
	cfg.scope = ipv6_addr_scope(cfg.pfx);
1411

1412
	ift = ipv6_add_addr(idev, &cfg, block, NULL);
1413
	if (IS_ERR(ift)) {
L
Linus Torvalds 已提交
1414 1415
		in6_ifa_put(ifp);
		in6_dev_put(idev);
1416
		pr_info("%s: retry temporary address regeneration\n", __func__);
L
Linus Torvalds 已提交
1417
		tmpaddr = &addr;
1418
		write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1419 1420 1421 1422 1423
		goto retry;
	}

	spin_lock_bh(&ift->lock);
	ift->ifpub = ifp;
1424
	ift->cstamp = now;
L
Linus Torvalds 已提交
1425 1426 1427
	ift->tstamp = tmp_tstamp;
	spin_unlock_bh(&ift->lock);

1428
	addrconf_dad_start(ift);
L
Linus Torvalds 已提交
1429 1430 1431 1432 1433 1434 1435
	in6_ifa_put(ift);
	in6_dev_put(idev);
out:
	return ret;
}

/*
1436
 *	Choose an appropriate source address (RFC3484)
L
Linus Torvalds 已提交
1437
 */
1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450
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,
1451 1452 1453
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	IPV6_SADDR_RULE_NOT_OPTIMISTIC,
#endif
1454 1455 1456
	IPV6_SADDR_RULE_MAX
};

1457
struct ipv6_saddr_score {
1458 1459 1460 1461 1462 1463
	int			rule;
	int			addr_type;
	struct inet6_ifaddr	*ifa;
	DECLARE_BITMAP(scorebits, IPV6_SADDR_RULE_MAX);
	int			scopedist;
	int			matchlen;
1464 1465
};

1466
struct ipv6_saddr_dst {
1467
	const struct in6_addr *addr;
1468 1469 1470
	int ifindex;
	int scope;
	int label;
1471
	unsigned int prefs;
1472
};
1473

D
Dave Jones 已提交
1474
static inline int ipv6_saddr_preferred(int type)
L
Linus Torvalds 已提交
1475
{
U
Ulrich Weber 已提交
1476
	if (type & (IPV6_ADDR_MAPPED|IPV6_ADDR_COMPATv4|IPV6_ADDR_LOOPBACK))
1477 1478
		return 1;
	return 0;
L
Linus Torvalds 已提交
1479 1480
}

1481 1482
static bool ipv6_use_optimistic_addr(struct net *net,
				     struct inet6_dev *idev)
1483 1484
{
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
1485 1486 1487 1488 1489 1490 1491 1492
	if (!idev)
		return false;
	if (!net->ipv6.devconf_all->optimistic_dad && !idev->cnf.optimistic_dad)
		return false;
	if (!net->ipv6.devconf_all->use_optimistic && !idev->cnf.use_optimistic)
		return false;

	return true;
1493 1494 1495 1496 1497
#else
	return false;
#endif
}

1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512
static bool ipv6_allow_optimistic_dad(struct net *net,
				      struct inet6_dev *idev)
{
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	if (!idev)
		return false;
	if (!net->ipv6.devconf_all->optimistic_dad && !idev->cnf.optimistic_dad)
		return false;

	return true;
#else
	return false;
#endif
}

1513 1514
static int ipv6_get_saddr_eval(struct net *net,
			       struct ipv6_saddr_score *score,
1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553
			       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
1554
		 *  B-15 |    \        if scope is enough for destination.
1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572
		 *       |             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:
1573
	    {
1574
		/* Rule 3: Avoid deprecated and optimistic addresses */
1575 1576
		u8 avoid = IFA_F_DEPRECATED;

1577
		if (!ipv6_use_optimistic_addr(net, score->ifa->idev))
1578
			avoid |= IFA_F_OPTIMISTIC;
1579
		ret = ipv6_saddr_preferred(score->addr_type) ||
1580
		      !(score->ifa->flags & avoid);
1581
		break;
1582
	    }
1583 1584
#ifdef CONFIG_IPV6_MIP6
	case IPV6_SADDR_RULE_HOA:
1585
	    {
1586
		/* Rule 4: Prefer home address */
1587 1588
		int prefhome = !(dst->prefs & IPV6_PREFER_SRC_COA);
		ret = !(score->ifa->flags & IFA_F_HOMEADDRESS) ^ prefhome;
1589
		break;
1590
	    }
1591 1592 1593 1594 1595 1596 1597 1598
#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 */
1599 1600
		ret = ipv6_addr_label(net,
				      &score->ifa->addr, score->addr_type,
1601 1602 1603
				      score->ifa->idev->dev->ifindex) == dst->label;
		break;
	case IPV6_SADDR_RULE_PRIVACY:
1604
	    {
1605
		/* Rule 7: Prefer public address
L
Lucas De Marchi 已提交
1606
		 * Note: prefer temporary address if use_tempaddr >= 2
1607
		 */
1608 1609 1610 1611
		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;
1612
		break;
1613
	    }
1614 1615 1616 1617 1618 1619 1620 1621 1622
	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 */
1623 1624 1625 1626
		ret = ipv6_addr_diff(&score->ifa->addr, dst->addr);
		if (ret > score->ifa->prefix_len)
			ret = score->ifa->prefix_len;
		score->matchlen = ret;
1627
		break;
1628 1629 1630 1631 1632 1633 1634 1635
#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
1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646
	default:
		ret = 0;
	}

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

1647 1648 1649 1650 1651
static int __ipv6_dev_get_saddr(struct net *net,
				struct ipv6_saddr_dst *dst,
				struct inet6_dev *idev,
				struct ipv6_saddr_score *scores,
				int hiscore_idx)
1652
{
1653
	struct ipv6_saddr_score *score = &scores[1 - hiscore_idx], *hiscore = &scores[hiscore_idx];
1654

1655
	list_for_each_entry_rcu(score->ifa, &idev->addr_list, if_list) {
1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705
		int i;

		/*
		 * - Tentative Address (RFC2462 section 5.4)
		 *  - A tentative address is not considered
		 *    "assigned to an interface" in the traditional
		 *    sense, unless it is also flagged as optimistic.
		 * - Candidate Source Address (section 4)
		 *  - In any case, anycast addresses, multicast
		 *    addresses, and the unspecified address MUST
		 *    NOT be included in a candidate set.
		 */
		if ((score->ifa->flags & IFA_F_TENTATIVE) &&
		    (!(score->ifa->flags & IFA_F_OPTIMISTIC)))
			continue;

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

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

		score->rule = -1;
		bitmap_zero(score->scorebits, IPV6_SADDR_RULE_MAX);

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

			minihiscore = ipv6_get_saddr_eval(net, hiscore, dst, i);
			miniscore = ipv6_get_saddr_eval(net, score, dst, i);

			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 out;
				}
				break;
			} else if (minihiscore < miniscore) {
				swap(hiscore, score);
1706
				hiscore_idx = 1 - hiscore_idx;
1707 1708 1709 1710 1711 1712 1713 1714 1715

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

				break;
			}
		}
	}
out:
1716
	return hiscore_idx;
1717 1718
}

1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740
static int ipv6_get_saddr_master(struct net *net,
				 const struct net_device *dst_dev,
				 const struct net_device *master,
				 struct ipv6_saddr_dst *dst,
				 struct ipv6_saddr_score *scores,
				 int hiscore_idx)
{
	struct inet6_dev *idev;

	idev = __in6_dev_get(dst_dev);
	if (idev)
		hiscore_idx = __ipv6_dev_get_saddr(net, dst, idev,
						   scores, hiscore_idx);

	idev = __in6_dev_get(master);
	if (idev)
		hiscore_idx = __ipv6_dev_get_saddr(net, dst, idev,
						   scores, hiscore_idx);

	return hiscore_idx;
}

1741
int ipv6_dev_get_saddr(struct net *net, const struct net_device *dst_dev,
1742
		       const struct in6_addr *daddr, unsigned int prefs,
1743
		       struct in6_addr *saddr)
L
Linus Torvalds 已提交
1744
{
1745
	struct ipv6_saddr_score scores[2], *hiscore;
1746
	struct ipv6_saddr_dst dst;
1747
	struct inet6_dev *idev;
1748
	struct net_device *dev;
1749
	int dst_type;
1750
	bool use_oif_addr = false;
1751
	int hiscore_idx = 0;
1752
	int ret = 0;
L
Linus Torvalds 已提交
1753

1754 1755 1756 1757
	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);
1758
	dst.label = ipv6_addr_label(net, daddr, dst_type, dst.ifindex);
1759
	dst.prefs = prefs;
1760

1761 1762
	scores[hiscore_idx].rule = -1;
	scores[hiscore_idx].ifa = NULL;
L
Linus Torvalds 已提交
1763

1764
	rcu_read_lock();
L
Linus Torvalds 已提交
1765

1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776
	/* Candidate Source Address (section 4)
	 *  - 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.)
1777 1778 1779 1780
	 *  - "It is RECOMMENDED that the candidate source addresses
	 *    be the set of unicast addresses assigned to the
	 *    interface that will be used to send to the destination
	 *    (the 'outgoing' interface)." (RFC 6724)
1781 1782
	 */
	if (dst_dev) {
1783
		idev = __in6_dev_get(dst_dev);
1784
		if ((dst_type & IPV6_ADDR_MULTICAST) ||
1785 1786
		    dst.scope <= IPV6_ADDR_SCOPE_LINKLOCAL ||
		    (idev && idev->cnf.use_oif_addrs_only)) {
1787 1788 1789
			use_oif_addr = true;
		}
	}
1790

1791
	if (use_oif_addr) {
1792
		if (idev)
1793
			hiscore_idx = __ipv6_dev_get_saddr(net, &dst, idev, scores, hiscore_idx);
1794
	} else {
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813
		const struct net_device *master;
		int master_idx = 0;

		/* if dst_dev exists and is enslaved to an L3 device, then
		 * prefer addresses from dst_dev and then the master over
		 * any other enslaved devices in the L3 domain.
		 */
		master = l3mdev_master_dev_rcu(dst_dev);
		if (master) {
			master_idx = master->ifindex;

			hiscore_idx = ipv6_get_saddr_master(net, dst_dev,
							    master, &dst,
							    scores, hiscore_idx);

			if (scores[hiscore_idx].ifa)
				goto out;
		}

1814
		for_each_netdev_rcu(net, dev) {
1815 1816 1817 1818 1819
			/* only consider addresses on devices in the
			 * same L3 domain
			 */
			if (l3mdev_master_ifindex_rcu(dev) != master_idx)
				continue;
1820 1821
			idev = __in6_dev_get(dev);
			if (!idev)
1822
				continue;
1823
			hiscore_idx = __ipv6_dev_get_saddr(net, &dst, idev, scores, hiscore_idx);
L
Linus Torvalds 已提交
1824 1825
		}
	}
1826 1827

out:
1828
	hiscore = &scores[hiscore_idx];
1829
	if (!hiscore->ifa)
1830 1831 1832
		ret = -EADDRNOTAVAIL;
	else
		*saddr = hiscore->ifa->addr;
1833

1834 1835
	rcu_read_unlock();
	return ret;
L
Linus Torvalds 已提交
1836
}
1837
EXPORT_SYMBOL(ipv6_dev_get_saddr);
L
Linus Torvalds 已提交
1838

1839
int __ipv6_get_lladdr(struct inet6_dev *idev, struct in6_addr *addr,
1840
		      u32 banned_flags)
1841 1842 1843 1844
{
	struct inet6_ifaddr *ifp;
	int err = -EADDRNOTAVAIL;

1845 1846 1847
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
1848 1849 1850 1851 1852 1853 1854 1855 1856 1857
		if (ifp->scope == IFA_LINK &&
		    !(ifp->flags & banned_flags)) {
			*addr = ifp->addr;
			err = 0;
			break;
		}
	}
	return err;
}

1858
int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr,
1859
		    u32 banned_flags)
L
Linus Torvalds 已提交
1860 1861 1862 1863
{
	struct inet6_dev *idev;
	int err = -EADDRNOTAVAIL;

1864
	rcu_read_lock();
S
Stephen Hemminger 已提交
1865 1866
	idev = __in6_dev_get(dev);
	if (idev) {
L
Linus Torvalds 已提交
1867
		read_lock_bh(&idev->lock);
1868
		err = __ipv6_get_lladdr(idev, addr, banned_flags);
L
Linus Torvalds 已提交
1869 1870
		read_unlock_bh(&idev->lock);
	}
1871
	rcu_read_unlock();
L
Linus Torvalds 已提交
1872 1873 1874
	return err;
}

1875
static int ipv6_count_addresses(const struct inet6_dev *idev)
L
Linus Torvalds 已提交
1876
{
1877
	const struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
1878 1879
	int cnt = 0;

1880 1881
	rcu_read_lock();
	list_for_each_entry_rcu(ifp, &idev->addr_list, if_list)
L
Linus Torvalds 已提交
1882
		cnt++;
1883
	rcu_read_unlock();
L
Linus Torvalds 已提交
1884 1885 1886
	return cnt;
}

1887
int ipv6_chk_addr(struct net *net, const struct in6_addr *addr,
1888
		  const struct net_device *dev, int strict)
1889
{
1890 1891
	return ipv6_chk_addr_and_flags(net, addr, dev, !dev,
				       strict, IFA_F_TENTATIVE);
1892 1893 1894
}
EXPORT_SYMBOL(ipv6_chk_addr);

1895 1896 1897 1898 1899 1900 1901 1902 1903
/* device argument is used to find the L3 domain of interest. If
 * skip_dev_check is set, then the ifp device is not checked against
 * the passed in dev argument. So the 2 cases for addresses checks are:
 *   1. does the address exist in the L3 domain that dev is part of
 *      (skip_dev_check = true), or
 *
 *   2. does the address exist on the specific device
 *      (skip_dev_check = false)
 */
1904
int ipv6_chk_addr_and_flags(struct net *net, const struct in6_addr *addr,
1905 1906
			    const struct net_device *dev, bool skip_dev_check,
			    int strict, u32 banned_flags)
L
Linus Torvalds 已提交
1907
{
1908
	unsigned int hash = inet6_addr_hash(net, addr);
1909
	const struct net_device *l3mdev;
1910
	struct inet6_ifaddr *ifp;
1911
	u32 ifp_flags;
L
Linus Torvalds 已提交
1912

1913
	rcu_read_lock();
1914

1915
	l3mdev = l3mdev_master_dev_rcu(dev);
1916 1917 1918
	if (skip_dev_check)
		dev = NULL;

1919
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
1920
		if (!net_eq(dev_net(ifp->idev->dev), net))
1921
			continue;
1922 1923 1924 1925

		if (l3mdev_master_dev_rcu(ifp->idev->dev) != l3mdev)
			continue;

1926 1927 1928 1929 1930 1931
		/* 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
1932
		if (ipv6_addr_equal(&ifp->addr, addr) &&
1933
		    !(ifp_flags&banned_flags) &&
1934
		    (!dev || ifp->idev->dev == dev ||
1935
		     !(ifp->scope&(IFA_LINK|IFA_HOST) || strict))) {
1936
			rcu_read_unlock();
1937
			return 1;
L
Linus Torvalds 已提交
1938 1939
		}
	}
1940

1941
	rcu_read_unlock();
1942
	return 0;
L
Linus Torvalds 已提交
1943
}
1944
EXPORT_SYMBOL(ipv6_chk_addr_and_flags);
1945

L
Linus Torvalds 已提交
1946

1947 1948 1949 1950 1951 1952
/* 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)
{
1953 1954
	const struct inet6_ifaddr *ifa;
	const struct inet6_dev *idev;
1955 1956 1957 1958 1959
	bool ret = false;

	rcu_read_lock();
	idev = __in6_dev_get(dev);
	if (idev) {
1960
		list_for_each_entry_rcu(ifa, &idev->addr_list, if_list) {
1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971
			ret = ipv6_prefix_equal(addr, &ifa->addr, prefix_len);
			if (ret)
				break;
		}
	}
	rcu_read_unlock();

	return ret;
}
EXPORT_SYMBOL(ipv6_chk_custom_prefix);

1972
int ipv6_chk_prefix(const struct in6_addr *addr, struct net_device *dev)
1973
{
1974 1975
	const struct inet6_ifaddr *ifa;
	const struct inet6_dev *idev;
1976 1977 1978 1979 1980 1981
	int	onlink;

	onlink = 0;
	rcu_read_lock();
	idev = __in6_dev_get(dev);
	if (idev) {
1982
		list_for_each_entry_rcu(ifa, &idev->addr_list, if_list) {
1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
			onlink = ipv6_prefix_equal(addr, &ifa->addr,
						   ifa->prefix_len);
			if (onlink)
				break;
		}
	}
	rcu_read_unlock();
	return onlink;
}
EXPORT_SYMBOL(ipv6_chk_prefix);

1994
struct inet6_ifaddr *ipv6_get_ifaddr(struct net *net, const struct in6_addr *addr,
1995
				     struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1996
{
1997
	unsigned int hash = inet6_addr_hash(net, addr);
1998
	struct inet6_ifaddr *ifp, *result = NULL;
L
Linus Torvalds 已提交
1999

2000 2001
	rcu_read_lock();
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
2002
		if (!net_eq(dev_net(ifp->idev->dev), net))
2003
			continue;
L
Linus Torvalds 已提交
2004
		if (ipv6_addr_equal(&ifp->addr, addr)) {
2005
			if (!dev || ifp->idev->dev == dev ||
L
Linus Torvalds 已提交
2006
			    !(ifp->scope&(IFA_LINK|IFA_HOST) || strict)) {
2007
				result = ifp;
L
Linus Torvalds 已提交
2008 2009 2010 2011 2012
				in6_ifa_hold(ifp);
				break;
			}
		}
	}
2013
	rcu_read_unlock();
L
Linus Torvalds 已提交
2014

2015
	return result;
L
Linus Torvalds 已提交
2016 2017 2018 2019
}

/* Gets referenced address, destroys ifaddr */

B
Brian Haley 已提交
2020
static void addrconf_dad_stop(struct inet6_ifaddr *ifp, int dad_failed)
L
Linus Torvalds 已提交
2021
{
2022 2023 2024
	if (dad_failed)
		ifp->flags |= IFA_F_DADFAILED;

2025
	if (ifp->flags&IFA_F_TEMPORARY) {
L
Linus Torvalds 已提交
2026 2027 2028 2029 2030 2031
		struct inet6_ifaddr *ifpub;
		spin_lock_bh(&ifp->lock);
		ifpub = ifp->ifpub;
		if (ifpub) {
			in6_ifa_hold(ifpub);
			spin_unlock_bh(&ifp->lock);
E
Eric Dumazet 已提交
2032
			ipv6_create_tempaddr(ifpub, ifp, true);
L
Linus Torvalds 已提交
2033 2034 2035 2036 2037
			in6_ifa_put(ifpub);
		} else {
			spin_unlock_bh(&ifp->lock);
		}
		ipv6_del_addr(ifp);
2038 2039 2040 2041
	} else if (ifp->flags&IFA_F_PERMANENT || !dad_failed) {
		spin_lock_bh(&ifp->lock);
		addrconf_del_dad_work(ifp);
		ifp->flags |= IFA_F_TENTATIVE;
2042 2043
		if (dad_failed)
			ifp->flags &= ~IFA_F_OPTIMISTIC;
2044 2045 2046 2047
		spin_unlock_bh(&ifp->lock);
		if (dad_failed)
			ipv6_ifa_notify(0, ifp);
		in6_ifa_put(ifp);
2048
	} else {
L
Linus Torvalds 已提交
2049
		ipv6_del_addr(ifp);
2050
	}
L
Linus Torvalds 已提交
2051 2052
}

H
Herbert Xu 已提交
2053 2054 2055 2056
static int addrconf_dad_end(struct inet6_ifaddr *ifp)
{
	int err = -ENOENT;

2057
	spin_lock_bh(&ifp->lock);
H
Herbert Xu 已提交
2058 2059 2060 2061
	if (ifp->state == INET6_IFADDR_STATE_DAD) {
		ifp->state = INET6_IFADDR_STATE_POSTDAD;
		err = 0;
	}
2062
	spin_unlock_bh(&ifp->lock);
H
Herbert Xu 已提交
2063 2064 2065 2066

	return err;
}

2067
void addrconf_dad_failure(struct sk_buff *skb, struct inet6_ifaddr *ifp)
2068
{
2069
	struct inet6_dev *idev = ifp->idev;
2070
	struct net *net = dev_net(ifp->idev->dev);
2071

2072 2073
	if (addrconf_dad_end(ifp)) {
		in6_ifa_put(ifp);
H
Herbert Xu 已提交
2074
		return;
2075
	}
H
Herbert Xu 已提交
2076

2077 2078
	net_info_ratelimited("%s: IPv6 duplicate address %pI6c used by %pM detected!\n",
			     ifp->idev->dev->name, &ifp->addr, eth_hdr(skb)->h_source);
2079

2080 2081 2082 2083 2084 2085
	spin_lock_bh(&ifp->lock);

	if (ifp->flags & IFA_F_STABLE_PRIVACY) {
		struct in6_addr new_addr;
		struct inet6_ifaddr *ifp2;
		int retries = ifp->stable_privacy_retry + 1;
2086 2087 2088 2089 2090 2091 2092 2093
		struct ifa6_config cfg = {
			.pfx = &new_addr,
			.plen = ifp->prefix_len,
			.ifa_flags = ifp->flags,
			.valid_lft = ifp->valid_lft,
			.preferred_lft = ifp->prefered_lft,
			.scope = ifp->scope,
		};
2094

2095
		if (retries > net->ipv6.sysctl.idgen_retries) {
2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106
			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;

		spin_unlock_bh(&ifp->lock);
2107

2108 2109 2110 2111 2112 2113 2114 2115
		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);

2116
		ifp2 = ipv6_add_addr(idev, &cfg, false, NULL);
2117 2118 2119 2120 2121 2122 2123 2124
		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);

2125
		addrconf_mod_dad_work(ifp2, net->ipv6.sysctl.idgen_delay);
2126 2127 2128
		in6_ifa_put(ifp2);
lock_errdad:
		spin_lock_bh(&ifp->lock);
2129 2130
	}

2131
errdad:
2132 2133
	/* transition from _POSTDAD to _ERRDAD */
	ifp->state = INET6_IFADDR_STATE_ERRDAD;
2134
	spin_unlock_bh(&ifp->lock);
2135 2136

	addrconf_mod_dad_work(ifp, 0);
2137
	in6_ifa_put(ifp);
2138
}
L
Linus Torvalds 已提交
2139

2140 2141
/* Join to solicited addr multicast group.
 * caller must hold RTNL */
2142
void addrconf_join_solict(struct net_device *dev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2143 2144 2145 2146 2147 2148 2149 2150 2151 2152
{
	struct in6_addr maddr;

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

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

2153
/* caller must hold RTNL */
2154
void addrconf_leave_solict(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2155 2156 2157 2158 2159 2160 2161 2162 2163 2164
{
	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);
}

2165
/* caller must hold RTNL */
2166
static void addrconf_join_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
2167 2168
{
	struct in6_addr addr;
2169

2170
	if (ifp->prefix_len >= 127) /* RFC 6164 */
2171
		return;
L
Linus Torvalds 已提交
2172 2173 2174
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
W
WANG Cong 已提交
2175
	__ipv6_dev_ac_inc(ifp->idev, &addr);
L
Linus Torvalds 已提交
2176 2177
}

2178
/* caller must hold RTNL */
2179
static void addrconf_leave_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
2180 2181
{
	struct in6_addr addr;
2182

2183
	if (ifp->prefix_len >= 127) /* RFC 6164 */
2184
		return;
L
Linus Torvalds 已提交
2185 2186 2187 2188 2189 2190
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
	__ipv6_dev_ac_dec(ifp->idev, &addr);
}

2191
static int addrconf_ifid_6lowpan(u8 *eui, struct net_device *dev)
2192
{
2193 2194
	switch (dev->addr_len) {
	case ETH_ALEN:
2195 2196 2197 2198 2199
		memcpy(eui, dev->dev_addr, 3);
		eui[3] = 0xFF;
		eui[4] = 0xFE;
		memcpy(eui + 5, dev->dev_addr + 3, 3);
		break;
2200 2201 2202 2203 2204
	case EUI64_ADDR_LEN:
		memcpy(eui, dev->dev_addr, EUI64_ADDR_LEN);
		eui[0] ^= 2;
		break;
	default:
2205
		return -1;
2206 2207
	}

2208 2209 2210
	return 0;
}

2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224
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;
}

2225 2226 2227 2228 2229 2230
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);
2231
	eui[7] = *(u8 *)dev->dev_addr;
2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243
	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;
}

2244
static int __ipv6_isatap_ifid(u8 *eui, __be32 addr)
2245
{
2246 2247
	if (addr == 0)
		return -1;
2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267
	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;
}

2268 2269 2270 2271 2272
static int addrconf_ifid_gre(u8 *eui, struct net_device *dev)
{
	return __ipv6_isatap_ifid(eui, *(__be32 *)dev->dev_addr);
}

2273 2274 2275 2276 2277 2278 2279 2280 2281 2282
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
2283 2284 2285 2286 2287
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev)
{
	switch (dev->type) {
	case ARPHRD_ETHER:
	case ARPHRD_FDDI:
2288
		return addrconf_ifid_eui48(eui, dev);
L
Linus Torvalds 已提交
2289
	case ARPHRD_ARCNET:
2290
		return addrconf_ifid_arcnet(eui, dev);
L
Linus Torvalds 已提交
2291
	case ARPHRD_INFINIBAND:
2292
		return addrconf_ifid_infiniband(eui, dev);
F
Fred L. Templin 已提交
2293
	case ARPHRD_SIT:
2294
		return addrconf_ifid_sit(eui, dev);
2295
	case ARPHRD_IPGRE:
2296
	case ARPHRD_TUNNEL:
2297
		return addrconf_ifid_gre(eui, dev);
2298
	case ARPHRD_6LOWPAN:
2299
		return addrconf_ifid_6lowpan(eui, dev);
2300 2301
	case ARPHRD_IEEE1394:
		return addrconf_ifid_ieee1394(eui, dev);
2302
	case ARPHRD_TUNNEL6:
2303
	case ARPHRD_IP6GRE:
2304
	case ARPHRD_RAWIP:
2305
		return addrconf_ifid_ip6tnl(eui, dev);
L
Linus Torvalds 已提交
2306 2307 2308 2309 2310 2311 2312 2313 2314 2315
	}
	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);
2316 2317 2318
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
L
Linus Torvalds 已提交
2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329
		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) */
2330
static void ipv6_regen_rndid(struct inet6_dev *idev)
L
Linus Torvalds 已提交
2331 2332
{
regen:
2333
	get_random_bytes(idev->rndid, sizeof(idev->rndid));
L
Linus Torvalds 已提交
2334 2335 2336 2337 2338 2339 2340 2341
	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
2342
	 *  - ISATAP (RFC4214) 6.1
L
Linus Torvalds 已提交
2343 2344 2345 2346
	 *	00-00-5E-FE-xx-xx-xx-xx
	 *  - value 0
	 *  - XXX: already assigned to an address on the device
	 */
2347
	if (idev->rndid[0] == 0xfd &&
L
Linus Torvalds 已提交
2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358
	    (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;
	}
}

2359
static void  ipv6_try_regen_rndid(struct inet6_dev *idev, struct in6_addr *tmpaddr)
2360
{
L
Linus Torvalds 已提交
2361
	if (tmpaddr && memcmp(idev->rndid, &tmpaddr->s6_addr[8], 8) == 0)
2362
		ipv6_regen_rndid(idev);
L
Linus Torvalds 已提交
2363 2364 2365 2366 2367 2368 2369
}

/*
 *	Add prefix route.
 */

static void
2370 2371 2372
addrconf_prefix_route(struct in6_addr *pfx, int plen, u32 metric,
		      struct net_device *dev, unsigned long expires,
		      u32 flags, gfp_t gfp_flags)
L
Linus Torvalds 已提交
2373
{
2374
	struct fib6_config cfg = {
D
David Ahern 已提交
2375
		.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_PREFIX,
2376
		.fc_metric = metric ? : IP6_RT_PRIO_ADDRCONF,
2377 2378 2379 2380
		.fc_ifindex = dev->ifindex,
		.fc_expires = expires,
		.fc_dst_len = plen,
		.fc_flags = RTF_UP | flags,
2381
		.fc_nlinfo.nl_net = dev_net(dev),
2382
		.fc_protocol = RTPROT_KERNEL,
2383
		.fc_type = RTN_UNICAST,
2384
	};
L
Linus Torvalds 已提交
2385

A
Alexey Dobriyan 已提交
2386
	cfg.fc_dst = *pfx;
L
Linus Torvalds 已提交
2387 2388 2389 2390 2391

	/* 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
2392
#if IS_ENABLED(CONFIG_IPV6_SIT)
2393 2394
	if (dev->type == ARPHRD_SIT && (dev->flags & IFF_POINTOPOINT))
		cfg.fc_flags |= RTF_NONEXTHOP;
2395
#endif
L
Linus Torvalds 已提交
2396

2397
	ip6_route_add(&cfg, gfp_flags, NULL);
L
Linus Torvalds 已提交
2398 2399
}

2400

2401
static struct fib6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
2402 2403 2404 2405 2406
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags)
{
	struct fib6_node *fn;
2407
	struct fib6_info *rt = NULL;
2408
	struct fib6_table *table;
D
David Ahern 已提交
2409
	u32 tb_id = l3mdev_fib_table(dev) ? : RT6_TABLE_PREFIX;
2410

D
David Ahern 已提交
2411
	table = fib6_get_table(dev_net(dev), tb_id);
2412
	if (!table)
2413 2414
		return NULL;

2415
	rcu_read_lock();
2416
	fn = fib6_locate(&table->tb6_root, pfx, plen, NULL, 0, true);
2417 2418
	if (!fn)
		goto out;
2419

2420
	for_each_fib6_node_rt_rcu(fn) {
2421
		if (rt->fib6_nh.nh_dev->ifindex != dev->ifindex)
2422
			continue;
2423
		if ((rt->fib6_flags & flags) != flags)
2424
			continue;
2425
		if ((rt->fib6_flags & noflags) != 0)
2426
			continue;
2427 2428
		if (!fib6_info_hold_safe(rt))
			continue;
2429 2430 2431
		break;
	}
out:
2432
	rcu_read_unlock();
2433 2434 2435 2436
	return rt;
}


L
Linus Torvalds 已提交
2437 2438 2439 2440
/* Create "default" multicast route to the interface */

static void addrconf_add_mroute(struct net_device *dev)
{
2441
	struct fib6_config cfg = {
D
David Ahern 已提交
2442
		.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_LOCAL,
2443 2444 2445 2446
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_dst_len = 8,
		.fc_flags = RTF_UP,
2447
		.fc_type = RTN_UNICAST,
2448
		.fc_nlinfo.nl_net = dev_net(dev),
2449 2450 2451 2452
	};

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

2453
	ip6_route_add(&cfg, GFP_KERNEL, NULL);
L
Linus Torvalds 已提交
2454 2455 2456 2457 2458 2459 2460 2461
}

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

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
2462 2463
	idev = ipv6_find_idev(dev);
	if (!idev)
2464 2465 2466 2467
		return ERR_PTR(-ENOBUFS);

	if (idev->cnf.disable_ipv6)
		return ERR_PTR(-EACCES);
L
Linus Torvalds 已提交
2468 2469

	/* Add default multicast route */
2470
	if (!(dev->flags & IFF_LOOPBACK) && !netif_is_l3_master(dev))
2471
		addrconf_add_mroute(dev);
L
Linus Torvalds 已提交
2472 2473 2474 2475

	return idev;
}

2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505
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 -
J
Jiri Bohac 已提交
2506
			       idev->desync_factor - age;
2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536
		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);
E
Eric Dumazet 已提交
2537
		ipv6_create_tempaddr(ifp, NULL, false);
2538 2539 2540 2541 2542
	} else {
		read_unlock_bh(&idev->lock);
	}
}

2543 2544
static bool is_addr_mode_generate_stable(struct inet6_dev *idev)
{
2545 2546
	return idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_STABLE_PRIVACY ||
	       idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_RANDOM;
2547 2548
}

A
Alexander Aring 已提交
2549 2550 2551 2552 2553 2554
int addrconf_prefix_rcv_add_addr(struct net *net, struct net_device *dev,
				 const struct prefix_info *pinfo,
				 struct inet6_dev *in6_dev,
				 const struct in6_addr *addr, int addr_type,
				 u32 addr_flags, bool sllao, bool tokenized,
				 __u32 valid_lft, u32 prefered_lft)
2555 2556 2557 2558 2559 2560
{
	struct inet6_ifaddr *ifp = ipv6_get_ifaddr(net, addr, dev, 1);
	int create = 0, update_lft = 0;

	if (!ifp && valid_lft) {
		int max_addresses = in6_dev->cnf.max_addresses;
2561 2562 2563 2564 2565 2566 2567 2568
		struct ifa6_config cfg = {
			.pfx = addr,
			.plen = pinfo->prefix_len,
			.ifa_flags = addr_flags,
			.valid_lft = valid_lft,
			.preferred_lft = prefered_lft,
			.scope = addr_type & IPV6_ADDR_SCOPE_MASK,
		};
2569 2570

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
2571 2572
		if ((net->ipv6.devconf_all->optimistic_dad ||
		     in6_dev->cnf.optimistic_dad) &&
2573
		    !net->ipv6.devconf_all->forwarding && sllao)
2574
			cfg.ifa_flags |= IFA_F_OPTIMISTIC;
2575 2576 2577 2578 2579 2580 2581
#endif

		/* 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)
2582
			ifp = ipv6_add_addr(in6_dev, &cfg, false, NULL);
2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607

		if (IS_ERR_OR_NULL(ifp))
			return -1;

		create = 1;
		spin_lock_bh(&ifp->lock);
		ifp->flags |= IFA_F_MANAGETEMPADDR;
		ifp->cstamp = jiffies;
		ifp->tokenized = tokenized;
		spin_unlock_bh(&ifp->lock);
		addrconf_dad_start(ifp);
	}

	if (ifp) {
		u32 flags;
		unsigned long now;
		u32 stored_lft;

		/* update lifetime (RFC2462 5.5.3 e) */
		spin_lock_bh(&ifp->lock);
		now = jiffies;
		if (ifp->valid_lft > (now - ifp->tstamp) / HZ)
			stored_lft = ifp->valid_lft - (now - ifp->tstamp) / HZ;
		else
			stored_lft = 0;
2608
		if (!create && stored_lft) {
2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647
			const u32 minimum_lft = min_t(u32,
				stored_lft, MIN_VALID_LIFETIME);
			valid_lft = max(valid_lft, minimum_lft);

			/* RFC4862 Section 5.5.3e:
			 * "Note that the preferred lifetime of the
			 *  corresponding address is always reset to
			 *  the Preferred Lifetime in the received
			 *  Prefix Information option, regardless of
			 *  whether the valid lifetime is also reset or
			 *  ignored."
			 *
			 * So we should always update prefered_lft here.
			 */
			update_lft = 1;
		}

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

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

		manage_tempaddrs(in6_dev, ifp, valid_lft, prefered_lft,
				 create, now);

		in6_ifa_put(ifp);
		addrconf_verify();
	}

	return 0;
}
A
Alexander Aring 已提交
2648
EXPORT_SYMBOL_GPL(addrconf_prefix_rcv_add_addr);
2649

2650
void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
L
Linus Torvalds 已提交
2651 2652 2653 2654
{
	struct prefix_info *pinfo;
	__u32 valid_lft;
	__u32 prefered_lft;
2655
	int addr_type, err;
2656
	u32 addr_flags = 0;
L
Linus Torvalds 已提交
2657
	struct inet6_dev *in6_dev;
2658
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
2659 2660

	pinfo = (struct prefix_info *) opt;
2661

L
Linus Torvalds 已提交
2662
	if (len < sizeof(struct prefix_info)) {
2663
		netdev_dbg(dev, "addrconf: prefix option too short\n");
L
Linus Torvalds 已提交
2664 2665
		return;
	}
2666

L
Linus Torvalds 已提交
2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679
	/*
	 *	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) {
2680
		net_warn_ratelimited("addrconf: prefix option has invalid lifetime\n");
L
Linus Torvalds 已提交
2681 2682 2683 2684 2685
		return;
	}

	in6_dev = in6_dev_get(dev);

2686
	if (!in6_dev) {
2687 2688
		net_dbg_ratelimited("addrconf: device %s not configured\n",
				    dev->name);
L
Linus Torvalds 已提交
2689 2690 2691 2692 2693 2694 2695 2696 2697
		return;
	}

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

2698
	if (pinfo->onlink) {
2699
		struct fib6_info *rt;
2700 2701
		unsigned long rt_expires;

2702 2703 2704 2705 2706
		/* 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.
		 */
2707 2708 2709 2710
		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
2711

2712 2713
		if (addrconf_finite_timeout(rt_expires))
			rt_expires *= HZ;
L
Linus Torvalds 已提交
2714

2715 2716 2717 2718 2719
		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
2720

2721
		if (rt) {
2722 2723
			/* Autoconf prefix route */
			if (valid_lft == 0) {
2724
				ip6_del_rt(net, rt);
2725
				rt = NULL;
2726
			} else if (addrconf_finite_timeout(rt_expires)) {
2727
				/* not infinity */
2728
				fib6_set_expires(rt, jiffies + rt_expires);
2729
			} else {
2730
				fib6_clean_expires(rt);
L
Linus Torvalds 已提交
2731 2732
			}
		} else if (valid_lft) {
2733
			clock_t expires = 0;
2734 2735
			int flags = RTF_ADDRCONF | RTF_PREFIX_RT;
			if (addrconf_finite_timeout(rt_expires)) {
2736 2737 2738 2739
				/* not infinity */
				flags |= RTF_EXPIRES;
				expires = jiffies_to_clock_t(rt_expires);
			}
L
Linus Torvalds 已提交
2740
			addrconf_prefix_route(&pinfo->prefix, pinfo->prefix_len,
2741 2742
					      0, dev, expires, flags,
					      GFP_ATOMIC);
L
Linus Torvalds 已提交
2743
		}
2744
		fib6_info_release(rt);
L
Linus Torvalds 已提交
2745 2746 2747 2748 2749 2750
	}

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

	if (pinfo->autoconf && in6_dev->cnf.autoconf) {
		struct in6_addr addr;
2751
		bool tokenized = false, dev_addr_generated = false;
L
Linus Torvalds 已提交
2752 2753 2754

		if (pinfo->prefix_len == 64) {
			memcpy(&addr, &pinfo->prefix, 8);
2755 2756 2757 2758 2759 2760

			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);
2761
				tokenized = true;
2762
			} else if (is_addr_mode_generate_stable(in6_dev) &&
2763 2764
				   !ipv6_generate_stable_address(&addr, 0,
								 in6_dev)) {
2765
				addr_flags |= IFA_F_STABLE_PRIVACY;
2766
				goto ok;
2767 2768
			} else if (ipv6_generate_eui64(addr.s6_addr + 8, dev) &&
				   ipv6_inherit_eui64(addr.s6_addr + 8, in6_dev)) {
2769
				goto put;
2770 2771
			} else {
				dev_addr_generated = true;
L
Linus Torvalds 已提交
2772 2773 2774
			}
			goto ok;
		}
2775 2776
		net_dbg_ratelimited("IPv6 addrconf: prefix with wrong length %d\n",
				    pinfo->prefix_len);
2777
		goto put;
L
Linus Torvalds 已提交
2778 2779

ok:
2780 2781 2782 2783 2784 2785 2786
		err = addrconf_prefix_rcv_add_addr(net, dev, pinfo, in6_dev,
						   &addr, addr_type,
						   addr_flags, sllao,
						   tokenized, valid_lft,
						   prefered_lft);
		if (err)
			goto put;
2787 2788 2789 2790 2791 2792 2793 2794 2795

		/* Ignore error case here because previous prefix add addr was
		 * successful which will be notified.
		 */
		ndisc_ops_prefix_rcv_add_addr(net, dev, pinfo, in6_dev, &addr,
					      addr_type, addr_flags, sllao,
					      tokenized, valid_lft,
					      prefered_lft,
					      dev_addr_generated);
L
Linus Torvalds 已提交
2796 2797
	}
	inet6_prefix_notify(RTM_NEWPREFIX, in6_dev, pinfo);
2798
put:
L
Linus Torvalds 已提交
2799 2800 2801 2802 2803 2804 2805 2806
	in6_dev_put(in6_dev);
}

/*
 *	Set destination address.
 *	Special case for SIT interfaces where we create a new "virtual"
 *	device.
 */
2807
int addrconf_set_dstaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818
{
	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;

2819
	dev = __dev_get_by_index(net, ireq.ifr6_ifindex);
L
Linus Torvalds 已提交
2820 2821

	err = -ENODEV;
2822
	if (!dev)
L
Linus Torvalds 已提交
2823 2824
		goto err_exit;

A
Amerigo Wang 已提交
2825
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2826
	if (dev->type == ARPHRD_SIT) {
2827
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841
		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;
2842
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;
L
Linus Torvalds 已提交
2843

2844 2845 2846 2847 2848 2849 2850 2851
		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
2852 2853 2854

		if (err == 0) {
			err = -ENOBUFS;
2855 2856
			dev = __dev_get_by_name(net, p.name);
			if (!dev)
L
Linus Torvalds 已提交
2857
				goto err_exit;
2858
			err = dev_open(dev, NULL);
L
Linus Torvalds 已提交
2859 2860
		}
	}
2861
#endif
L
Linus Torvalds 已提交
2862 2863 2864 2865 2866 2867

err_exit:
	rtnl_unlock();
	return err;
}

2868 2869 2870 2871 2872 2873 2874 2875 2876
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)
2877
		ret = ipv6_sock_mc_join(sk, ifindex, addr);
2878
	else
2879
		ret = ipv6_sock_mc_drop(sk, ifindex, addr);
2880 2881 2882 2883 2884
	release_sock(sk);

	return ret;
}

L
Linus Torvalds 已提交
2885 2886 2887
/*
 *	Manual configuration of address on an interface
 */
2888
static int inet6_addr_add(struct net *net, int ifindex,
2889
			  struct ifa6_config *cfg,
2890
			  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
2891 2892 2893 2894
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2895 2896
	unsigned long timeout;
	clock_t expires;
2897
	u32 flags;
L
Linus Torvalds 已提交
2898 2899

	ASSERT_RTNL();
2900

2901
	if (cfg->plen > 128)
2902 2903
		return -EINVAL;

2904
	/* check the lifetime */
2905
	if (!cfg->valid_lft || cfg->preferred_lft > cfg->valid_lft)
2906 2907
		return -EINVAL;

2908
	if (cfg->ifa_flags & IFA_F_MANAGETEMPADDR && cfg->plen != 64)
2909 2910
		return -EINVAL;

2911 2912
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2913
		return -ENODEV;
2914

2915 2916 2917
	idev = addrconf_add_dev(dev);
	if (IS_ERR(idev))
		return PTR_ERR(idev);
L
Linus Torvalds 已提交
2918

2919
	if (cfg->ifa_flags & IFA_F_MCAUTOJOIN) {
2920
		int ret = ipv6_mc_config(net->ipv6.mc_autojoin_sk,
2921
					 true, cfg->pfx, ifindex);
2922 2923 2924 2925 2926

		if (ret < 0)
			return ret;
	}

2927
	cfg->scope = ipv6_addr_scope(cfg->pfx);
L
Linus Torvalds 已提交
2928

2929
	timeout = addrconf_timeout_fixup(cfg->valid_lft, HZ);
2930 2931
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
2932
		cfg->valid_lft = timeout;
2933
		flags = RTF_EXPIRES;
2934 2935 2936
	} else {
		expires = 0;
		flags = 0;
2937
		cfg->ifa_flags |= IFA_F_PERMANENT;
2938
	}
2939

2940
	timeout = addrconf_timeout_fixup(cfg->preferred_lft, HZ);
2941 2942
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
2943 2944
			cfg->ifa_flags |= IFA_F_DEPRECATED;
		cfg->preferred_lft = timeout;
2945
	}
2946

2947
	ifp = ipv6_add_addr(idev, cfg, true, extack);
L
Linus Torvalds 已提交
2948
	if (!IS_ERR(ifp)) {
2949
		if (!(cfg->ifa_flags & IFA_F_NOPREFIXROUTE)) {
2950 2951 2952
			addrconf_prefix_route(&ifp->addr, ifp->prefix_len,
					      ifp->rt_priority, dev, expires,
					      flags, GFP_KERNEL);
2953 2954
		}

2955 2956 2957 2958 2959
		/* Send a netlink notification if DAD is enabled and
		 * optimistic flag is not set
		 */
		if (!(ifp->flags & (IFA_F_OPTIMISTIC | IFA_F_NODAD)))
			ipv6_ifa_notify(0, ifp);
2960 2961 2962 2963 2964
		/*
		 * Note that section 3.1 of RFC 4429 indicates
		 * that the Optimistic flag should not be set for
		 * manually configured addresses
		 */
2965
		addrconf_dad_start(ifp);
2966 2967 2968
		if (cfg->ifa_flags & IFA_F_MANAGETEMPADDR)
			manage_tempaddrs(idev, ifp, cfg->valid_lft,
					 cfg->preferred_lft, true, jiffies);
L
Linus Torvalds 已提交
2969
		in6_ifa_put(ifp);
2970
		addrconf_verify_rtnl();
L
Linus Torvalds 已提交
2971
		return 0;
2972 2973 2974
	} else if (cfg->ifa_flags & IFA_F_MCAUTOJOIN) {
		ipv6_mc_config(net->ipv6.mc_autojoin_sk, false,
			       cfg->pfx, ifindex);
L
Linus Torvalds 已提交
2975 2976 2977 2978 2979
	}

	return PTR_ERR(ifp);
}

2980 2981
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
2982 2983 2984 2985
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2986

2987 2988 2989
	if (plen > 128)
		return -EINVAL;

2990 2991
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2992 2993
		return -ENODEV;

2994
	idev = __in6_dev_get(dev);
2995
	if (!idev)
L
Linus Torvalds 已提交
2996 2997 2998
		return -ENXIO;

	read_lock_bh(&idev->lock);
2999
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
L
Linus Torvalds 已提交
3000 3001 3002 3003
		if (ifp->prefix_len == plen &&
		    ipv6_addr_equal(pfx, &ifp->addr)) {
			in6_ifa_hold(ifp);
			read_unlock_bh(&idev->lock);
3004

3005 3006 3007 3008
			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
3009
			ipv6_del_addr(ifp);
3010
			addrconf_verify_rtnl();
3011 3012 3013 3014
			if (ipv6_addr_is_multicast(pfx)) {
				ipv6_mc_config(net->ipv6.mc_autojoin_sk,
					       false, pfx, dev->ifindex);
			}
L
Linus Torvalds 已提交
3015 3016 3017 3018 3019 3020 3021 3022
			return 0;
		}
	}
	read_unlock_bh(&idev->lock);
	return -EADDRNOTAVAIL;
}


3023
int addrconf_add_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
3024
{
3025 3026 3027 3028 3029
	struct ifa6_config cfg = {
		.ifa_flags = IFA_F_PERMANENT,
		.preferred_lft = INFINITY_LIFE_TIME,
		.valid_lft = INFINITY_LIFE_TIME,
	};
L
Linus Torvalds 已提交
3030 3031
	struct in6_ifreq ireq;
	int err;
3032

3033
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
3034
		return -EPERM;
3035

L
Linus Torvalds 已提交
3036 3037 3038
	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		return -EFAULT;

3039 3040 3041
	cfg.pfx = &ireq.ifr6_addr;
	cfg.plen = ireq.ifr6_prefixlen;

L
Linus Torvalds 已提交
3042
	rtnl_lock();
3043
	err = inet6_addr_add(net, ireq.ifr6_ifindex, &cfg, NULL);
L
Linus Torvalds 已提交
3044 3045 3046 3047
	rtnl_unlock();
	return err;
}

3048
int addrconf_del_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
3049 3050 3051
{
	struct in6_ifreq ireq;
	int err;
3052

3053
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
3054 3055 3056 3057 3058 3059
		return -EPERM;

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

	rtnl_lock();
3060
	err = inet6_addr_del(net, ireq.ifr6_ifindex, 0, &ireq.ifr6_addr,
3061
			     ireq.ifr6_prefixlen);
L
Linus Torvalds 已提交
3062 3063 3064 3065
	rtnl_unlock();
	return err;
}

3066 3067 3068 3069
static void add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
		     int plen, int scope)
{
	struct inet6_ifaddr *ifp;
3070 3071 3072 3073 3074 3075 3076 3077
	struct ifa6_config cfg = {
		.pfx = addr,
		.plen = plen,
		.ifa_flags = IFA_F_PERMANENT,
		.valid_lft = INFINITY_LIFE_TIME,
		.preferred_lft = INFINITY_LIFE_TIME,
		.scope = scope
	};
3078

3079
	ifp = ipv6_add_addr(idev, &cfg, true, NULL);
3080 3081 3082 3083
	if (!IS_ERR(ifp)) {
		spin_lock_bh(&ifp->lock);
		ifp->flags &= ~IFA_F_TENTATIVE;
		spin_unlock_bh(&ifp->lock);
3084
		rt_genid_bump_ipv6(dev_net(idev->dev));
3085 3086 3087 3088 3089
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
		in6_ifa_put(ifp);
	}
}

A
Amerigo Wang 已提交
3090
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3091 3092 3093 3094
static void sit_add_v4_addrs(struct inet6_dev *idev)
{
	struct in6_addr addr;
	struct net_device *dev;
3095
	struct net *net = dev_net(idev->dev);
3096
	int scope, plen;
3097
	u32 pflags = 0;
L
Linus Torvalds 已提交
3098 3099 3100 3101 3102 3103 3104 3105 3106

	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;
3107
		plen = 64;
L
Linus Torvalds 已提交
3108 3109
	} else {
		scope = IPV6_ADDR_COMPATv4;
3110
		plen = 96;
3111
		pflags |= RTF_NONEXTHOP;
L
Linus Torvalds 已提交
3112 3113 3114
	}

	if (addr.s6_addr32[3]) {
3115
		add_addr(idev, &addr, plen, scope);
3116
		addrconf_prefix_route(&addr, plen, 0, idev->dev, 0, pflags,
3117
				      GFP_KERNEL);
L
Linus Torvalds 已提交
3118 3119 3120
		return;
	}

3121
	for_each_netdev(net, dev) {
3122
		struct in_device *in_dev = __in_dev_get_rtnl(dev);
L
Linus Torvalds 已提交
3123
		if (in_dev && (dev->flags & IFF_UP)) {
3124
			struct in_ifaddr *ifa;
L
Linus Torvalds 已提交
3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139

			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;
				}

3140
				add_addr(idev, &addr, plen, flag);
3141
				addrconf_prefix_route(&addr, plen, 0, idev->dev,
3142
						      0, pflags, GFP_KERNEL);
L
Linus Torvalds 已提交
3143 3144
			}
		}
3145
	}
L
Linus Torvalds 已提交
3146
}
3147
#endif
L
Linus Torvalds 已提交
3148 3149 3150 3151 3152 3153 3154 3155 3156

static void init_loopback(struct net_device *dev)
{
	struct inet6_dev  *idev;

	/* ::1 */

	ASSERT_RTNL();

3157
	idev = ipv6_find_idev(dev);
3158
	if (!idev) {
3159
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
3160 3161 3162
		return;
	}

3163
	add_addr(idev, &in6addr_loopback, 128, IFA_HOST);
L
Linus Torvalds 已提交
3164 3165
}

3166 3167
void addrconf_add_linklocal(struct inet6_dev *idev,
			    const struct in6_addr *addr, u32 flags)
L
Linus Torvalds 已提交
3168
{
3169 3170 3171 3172 3173 3174 3175 3176
	struct ifa6_config cfg = {
		.pfx = addr,
		.plen = 64,
		.ifa_flags = flags | IFA_F_PERMANENT,
		.valid_lft = INFINITY_LIFE_TIME,
		.preferred_lft = INFINITY_LIFE_TIME,
		.scope = IFA_LINK
	};
3177
	struct inet6_ifaddr *ifp;
3178 3179

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
3180 3181
	if ((dev_net(idev->dev)->ipv6.devconf_all->optimistic_dad ||
	     idev->cnf.optimistic_dad) &&
3182
	    !dev_net(idev->dev)->ipv6.devconf_all->forwarding)
3183
		cfg.ifa_flags |= IFA_F_OPTIMISTIC;
3184
#endif
L
Linus Torvalds 已提交
3185

3186
	ifp = ipv6_add_addr(idev, &cfg, true, NULL);
L
Linus Torvalds 已提交
3187
	if (!IS_ERR(ifp)) {
3188
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, 0, idev->dev,
3189
				      0, 0, GFP_ATOMIC);
3190
		addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
3191 3192 3193
		in6_ifa_put(ifp);
	}
}
3194
EXPORT_SYMBOL_GPL(addrconf_add_linklocal);
L
Linus Torvalds 已提交
3195

3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223
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;
3224
			__be32 prefix[2];
3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249
			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);
3250 3251
	data.prefix[0] = address->s6_addr32[0];
	data.prefix[1] = address->s6_addr32[1];
3252 3253 3254 3255 3256 3257
	data.secret = secret;
	data.dad_count = dad_count;

	sha_transform(digest, data.__data, workspace);

	temp = *address;
3258 3259
	temp.s6_addr32[2] = (__force __be32)digest[0];
	temp.s6_addr32[3] = (__force __be32)digest[1];
3260 3261 3262 3263 3264

	spin_unlock_bh(&lock);

	if (ipv6_reserved_interfaceid(temp)) {
		dad_count++;
3265
		if (dad_count > dev_net(idev->dev)->ipv6.sysctl.idgen_retries)
3266 3267 3268 3269 3270 3271 3272 3273
			return -1;
		goto retry;
	}

	*address = temp;
	return 0;
}

3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284
static void ipv6_gen_mode_random_init(struct inet6_dev *idev)
{
	struct ipv6_stable_secret *s = &idev->cnf.stable_secret;

	if (s->initialized)
		return;
	s = &idev->cnf.stable_secret;
	get_random_bytes(&s->secret, sizeof(s->secret));
	s->initialized = true;
}

3285 3286
static void addrconf_addr_gen(struct inet6_dev *idev, bool prefix_route)
{
3287 3288
	struct in6_addr addr;

D
David Ahern 已提交
3289 3290 3291 3292
	/* no link local addresses on L3 master devices */
	if (netif_is_l3_master(idev->dev))
		return;

3293
	ipv6_addr_set(&addr, htonl(0xFE800000), 0, 0, 0);
3294

3295
	switch (idev->cnf.addr_gen_mode) {
3296 3297 3298 3299
	case IN6_ADDR_GEN_MODE_RANDOM:
		ipv6_gen_mode_random_init(idev);
		/* fallthrough */
	case IN6_ADDR_GEN_MODE_STABLE_PRIVACY:
3300
		if (!ipv6_generate_stable_address(&addr, 0, idev))
3301 3302
			addrconf_add_linklocal(idev, &addr,
					       IFA_F_STABLE_PRIVACY);
3303
		else if (prefix_route)
3304
			addrconf_prefix_route(&addr, 64, 0, idev->dev,
3305
					      0, 0, GFP_KERNEL);
3306 3307
		break;
	case IN6_ADDR_GEN_MODE_EUI64:
3308 3309 3310 3311 3312
		/* 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)
3313
			addrconf_add_linklocal(idev, &addr, 0);
3314
		else if (prefix_route)
3315
			addrconf_prefix_route(&addr, 64, 0, idev->dev,
3316
					      0, 0, GFP_KERNEL);
3317 3318 3319 3320 3321
		break;
	case IN6_ADDR_GEN_MODE_NONE:
	default:
		/* will not add any link local address */
		break;
3322 3323 3324
	}
}

L
Linus Torvalds 已提交
3325 3326
static void addrconf_dev_config(struct net_device *dev)
{
3327
	struct inet6_dev *idev;
L
Linus Torvalds 已提交
3328 3329 3330

	ASSERT_RTNL();

3331 3332 3333
	if ((dev->type != ARPHRD_ETHER) &&
	    (dev->type != ARPHRD_FDDI) &&
	    (dev->type != ARPHRD_ARCNET) &&
3334
	    (dev->type != ARPHRD_INFINIBAND) &&
3335
	    (dev->type != ARPHRD_IEEE1394) &&
3336
	    (dev->type != ARPHRD_TUNNEL6) &&
3337
	    (dev->type != ARPHRD_6LOWPAN) &&
3338 3339 3340
	    (dev->type != ARPHRD_IP6GRE) &&
	    (dev->type != ARPHRD_IPGRE) &&
	    (dev->type != ARPHRD_TUNNEL) &&
3341 3342
	    (dev->type != ARPHRD_NONE) &&
	    (dev->type != ARPHRD_RAWIP)) {
3343 3344 3345 3346
		/* Alas, we support only Ethernet autoconfiguration. */
		return;
	}

L
Linus Torvalds 已提交
3347
	idev = addrconf_add_dev(dev);
3348
	if (IS_ERR(idev))
L
Linus Torvalds 已提交
3349 3350
		return;

3351 3352
	/* this device type has no EUI support */
	if (dev->type == ARPHRD_NONE &&
3353 3354
	    idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_EUI64)
		idev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_RANDOM;
3355

3356
	addrconf_addr_gen(idev, false);
L
Linus Torvalds 已提交
3357 3358
}

A
Amerigo Wang 已提交
3359
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3360 3361 3362 3363 3364 3365
static void addrconf_sit_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3366 3367 3368
	/*
	 * Configure the tunnel with one of our IPv4
	 * addresses... we should configure all of
L
Linus Torvalds 已提交
3369 3370 3371
	 * our v4 addrs in the tunnel
	 */

3372
	idev = ipv6_find_idev(dev);
3373
	if (!idev) {
3374
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
3375 3376 3377
		return;
	}

F
Fred L. Templin 已提交
3378
	if (dev->priv_flags & IFF_ISATAP) {
3379
		addrconf_addr_gen(idev, false);
F
Fred L. Templin 已提交
3380 3381 3382
		return;
	}

L
Linus Torvalds 已提交
3383 3384
	sit_add_v4_addrs(idev);

3385
	if (dev->flags&IFF_POINTOPOINT)
L
Linus Torvalds 已提交
3386 3387
		addrconf_add_mroute(dev);
}
3388
#endif
L
Linus Torvalds 已提交
3389

A
Amerigo Wang 已提交
3390
#if IS_ENABLED(CONFIG_NET_IPGRE)
3391 3392 3393 3394 3395 3396
static void addrconf_gre_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3397
	idev = ipv6_find_idev(dev);
3398
	if (!idev) {
3399
		pr_debug("%s: add_dev failed\n", __func__);
3400 3401 3402
		return;
	}

3403
	addrconf_addr_gen(idev, true);
3404 3405
	if (dev->flags & IFF_POINTOPOINT)
		addrconf_add_mroute(dev);
3406 3407 3408
}
#endif

3409 3410
static int fixup_permanent_addr(struct net *net,
				struct inet6_dev *idev,
3411 3412
				struct inet6_ifaddr *ifp)
{
3413
	/* !fib6_node means the host route was removed from the
3414 3415 3416
	 * FIB, for example, if 'lo' device is taken down. In that
	 * case regenerate the host route.
	 */
3417
	if (!ifp->rt || !ifp->rt->fib6_node) {
3418
		struct fib6_info *f6i, *prev;
3419

3420 3421 3422 3423
		f6i = addrconf_f6i_alloc(net, idev, &ifp->addr, false,
					 GFP_ATOMIC);
		if (IS_ERR(f6i))
			return PTR_ERR(f6i);
3424

3425 3426 3427
		/* ifp->rt can be accessed outside of rtnl */
		spin_lock(&ifp->lock);
		prev = ifp->rt;
3428
		ifp->rt = f6i;
3429 3430
		spin_unlock(&ifp->lock);

3431
		fib6_info_release(prev);
3432 3433 3434 3435
	}

	if (!(ifp->flags & IFA_F_NOPREFIXROUTE)) {
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len,
3436 3437
				      ifp->rt_priority, idev->dev, 0, 0,
				      GFP_ATOMIC);
3438 3439
	}

3440 3441
	if (ifp->state == INET6_IFADDR_STATE_PREDAD)
		addrconf_dad_start(ifp);
3442 3443 3444 3445

	return 0;
}

3446
static void addrconf_permanent_addr(struct net *net, struct net_device *dev)
3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458
{
	struct inet6_ifaddr *ifp, *tmp;
	struct inet6_dev *idev;

	idev = __in6_dev_get(dev);
	if (!idev)
		return;

	write_lock_bh(&idev->lock);

	list_for_each_entry_safe(ifp, tmp, &idev->addr_list, if_list) {
		if ((ifp->flags & IFA_F_PERMANENT) &&
3459
		    fixup_permanent_addr(net, idev, ifp) < 0) {
3460
			write_unlock_bh(&idev->lock);
3461
			in6_ifa_hold(ifp);
3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472
			ipv6_del_addr(ifp);
			write_lock_bh(&idev->lock);

			net_info_ratelimited("%s: Failed to add prefix route for address %pI6c; dropping\n",
					     idev->dev->name, &ifp->addr);
		}
	}

	write_unlock_bh(&idev->lock);
}

3473
static int addrconf_notify(struct notifier_block *this, unsigned long event,
3474
			   void *ptr)
L
Linus Torvalds 已提交
3475
{
3476
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
3477
	struct netdev_notifier_change_info *change_info;
3478
	struct netdev_notifier_changeupper_info *info;
3479
	struct inet6_dev *idev = __in6_dev_get(dev);
3480
	struct net *net = dev_net(dev);
3481
	int run_pending = 0;
3482
	int err;
L
Linus Torvalds 已提交
3483

S
Stephen Hemminger 已提交
3484
	switch (event) {
3485
	case NETDEV_REGISTER:
3486
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
3487
			idev = ipv6_add_dev(dev);
3488 3489
			if (IS_ERR(idev))
				return notifier_from_errno(PTR_ERR(idev));
3490 3491
		}
		break;
S
Stephen Hemminger 已提交
3492

3493 3494 3495
	case NETDEV_CHANGEMTU:
		/* if MTU under IPV6_MIN_MTU stop IPv6 on this interface. */
		if (dev->mtu < IPV6_MIN_MTU) {
3496
			addrconf_ifdown(dev, dev != net->loopback_dev);
3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518
			break;
		}

		if (idev) {
			rt6_mtu_change(dev, dev->mtu);
			idev->cnf.mtu6 = dev->mtu;
			break;
		}

		/* allocate new idev */
		idev = ipv6_add_dev(dev);
		if (IS_ERR(idev))
			break;

		/* device is still not ready */
		if (!(idev->if_flags & IF_READY))
			break;

		run_pending = 1;

		/* fall through */

L
Linus Torvalds 已提交
3519
	case NETDEV_UP:
3520
	case NETDEV_CHANGE:
3521 3522 3523
		if (dev->flags & IFF_SLAVE)
			break;

3524 3525 3526
		if (idev && idev->cnf.disable_ipv6)
			break;

3527
		if (event == NETDEV_UP) {
3528
			/* restore routes for permanent addresses */
3529
			addrconf_permanent_addr(net, dev);
3530

3531
			if (!addrconf_link_ready(dev)) {
3532
				/* device is not ready yet. */
3533
				pr_info("ADDRCONF(NETDEV_UP): %s: link is not ready\n",
3534 3535 3536
					dev->name);
				break;
			}
3537

3538 3539 3540
			if (!idev && dev->mtu >= IPV6_MIN_MTU)
				idev = ipv6_add_dev(dev);

3541
			if (!IS_ERR_OR_NULL(idev)) {
3542
				idev->if_flags |= IF_READY;
3543 3544
				run_pending = 1;
			}
3545
		} else if (event == NETDEV_CHANGE) {
3546
			if (!addrconf_link_ready(dev)) {
3547
				/* device is still not ready. */
3548
				rt6_sync_down_dev(dev, event);
3549 3550 3551
				break;
			}

3552
			if (!IS_ERR_OR_NULL(idev)) {
3553 3554 3555 3556 3557 3558 3559
				if (idev->if_flags & IF_READY) {
					/* device is already configured -
					 * but resend MLD reports, we might
					 * have roamed and need to update
					 * multicast snooping switches
					 */
					ipv6_mc_up(idev);
3560 3561 3562
					change_info = ptr;
					if (change_info->flags_changed & IFF_NOARP)
						addrconf_dad_run(idev, true);
3563
					rt6_sync_up(dev, RTNH_F_LINKDOWN);
3564
					break;
3565
				}
3566 3567 3568
				idev->if_flags |= IF_READY;
			}

3569 3570
			pr_info("ADDRCONF(NETDEV_CHANGE): %s: link becomes ready\n",
				dev->name);
3571

3572
			run_pending = 1;
3573 3574
		}

S
Stephen Hemminger 已提交
3575
		switch (dev->type) {
A
Amerigo Wang 已提交
3576
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3577 3578 3579
		case ARPHRD_SIT:
			addrconf_sit_config(dev);
			break;
3580
#endif
A
Amerigo Wang 已提交
3581
#if IS_ENABLED(CONFIG_NET_IPGRE)
3582 3583 3584
		case ARPHRD_IPGRE:
			addrconf_gre_config(dev);
			break;
3585
#endif
L
Linus Torvalds 已提交
3586 3587 3588 3589 3590 3591 3592
		case ARPHRD_LOOPBACK:
			init_loopback(dev);
			break;

		default:
			addrconf_dev_config(dev);
			break;
3593
		}
S
Stephen Hemminger 已提交
3594

3595
		if (!IS_ERR_OR_NULL(idev)) {
3596
			if (run_pending)
3597
				addrconf_dad_run(idev, false);
3598

3599 3600 3601
			/* Device has an address by now */
			rt6_sync_up(dev, RTNH_F_DEAD);

S
Stephen Hemminger 已提交
3602 3603 3604 3605
			/*
			 * 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
3606
			 */
S
Stephen Hemminger 已提交
3607 3608
			if (idev->cnf.mtu6 != dev->mtu &&
			    dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
3609 3610 3611 3612 3613
				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
3614 3615 3616 3617

			/*
			 * If the changed mtu during down is lower than
			 * IPV6_MIN_MTU stop IPv6 on this interface.
L
Linus Torvalds 已提交
3618 3619
			 */
			if (dev->mtu < IPV6_MIN_MTU)
3620
				addrconf_ifdown(dev, dev != net->loopback_dev);
L
Linus Torvalds 已提交
3621 3622 3623 3624 3625 3626 3627 3628 3629 3630
		}
		break;

	case NETDEV_DOWN:
	case NETDEV_UNREGISTER:
		/*
		 *	Remove all addresses from this interface.
		 */
		addrconf_ifdown(dev, event != NETDEV_DOWN);
		break;
3631

L
Linus Torvalds 已提交
3632 3633
	case NETDEV_CHANGENAME:
		if (idev) {
3634
			snmp6_unregister_dev(idev);
3635
			addrconf_sysctl_unregister(idev);
3636
			err = addrconf_sysctl_register(idev);
3637 3638
			if (err)
				return notifier_from_errno(err);
3639 3640 3641 3642 3643
			err = snmp6_register_dev(idev);
			if (err) {
				addrconf_sysctl_unregister(idev);
				return notifier_from_errno(err);
			}
3644
		}
L
Linus Torvalds 已提交
3645
		break;
S
Stephen Hemminger 已提交
3646

3647 3648
	case NETDEV_PRE_TYPE_CHANGE:
	case NETDEV_POST_TYPE_CHANGE:
3649 3650
		if (idev)
			addrconf_type_change(dev, event);
3651
		break;
3652 3653 3654 3655 3656 3657 3658 3659 3660

	case NETDEV_CHANGEUPPER:
		info = ptr;

		/* flush all routes if dev is linked to or unlinked from
		 * an L3 master device (e.g., VRF)
		 */
		if (info->upper_dev && netif_is_l3_master(info->upper_dev))
			addrconf_ifdown(dev, 0);
3661
	}
L
Linus Torvalds 已提交
3662 3663 3664 3665 3666 3667 3668 3669 3670

	return NOTIFY_OK;
}

/*
 *	addrconf module should be notified of a device going up
 */
static struct notifier_block ipv6_dev_notf = {
	.notifier_call = addrconf_notify,
3671
	.priority = ADDRCONF_NOTIFY_PRIORITY,
L
Linus Torvalds 已提交
3672 3673
};

3674
static void addrconf_type_change(struct net_device *dev, unsigned long event)
3675 3676 3677 3678 3679 3680
{
	struct inet6_dev *idev;
	ASSERT_RTNL();

	idev = __in6_dev_get(dev);

3681
	if (event == NETDEV_POST_TYPE_CHANGE)
3682
		ipv6_mc_remap(idev);
3683
	else if (event == NETDEV_PRE_TYPE_CHANGE)
3684 3685 3686
		ipv6_mc_unmap(idev);
}

3687 3688 3689 3690 3691 3692
static bool addr_is_local(const struct in6_addr *addr)
{
	return ipv6_addr_type(addr) &
		(IPV6_ADDR_LINKLOCAL | IPV6_ADDR_LOOPBACK);
}

L
Linus Torvalds 已提交
3693 3694
static int addrconf_ifdown(struct net_device *dev, int how)
{
3695
	unsigned long event = how ? NETDEV_UNREGISTER : NETDEV_DOWN;
3696
	struct net *net = dev_net(dev);
3697
	struct inet6_dev *idev;
3698
	struct inet6_ifaddr *ifa, *tmp;
3699
	bool keep_addr = false;
3700
	int state, i;
L
Linus Torvalds 已提交
3701 3702 3703

	ASSERT_RTNL();

3704
	rt6_disable_ip(dev, event);
L
Linus Torvalds 已提交
3705 3706

	idev = __in6_dev_get(dev);
3707
	if (!idev)
L
Linus Torvalds 已提交
3708 3709
		return -ENODEV;

S
Stephen Hemminger 已提交
3710 3711 3712
	/*
	 * Step 1: remove reference to ipv6 device from parent device.
	 *	   Do not dev_put!
L
Linus Torvalds 已提交
3713
	 */
3714
	if (how) {
L
Linus Torvalds 已提交
3715
		idev->dead = 1;
3716 3717

		/* protected by rtnl_lock */
3718
		RCU_INIT_POINTER(dev->ip6_ptr, NULL);
L
Linus Torvalds 已提交
3719 3720 3721 3722 3723 3724

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

	}

3725 3726 3727
	/* combine the user config with event to determine if permanent
	 * addresses are to be removed from address hash table
	 */
3728 3729 3730 3731 3732 3733 3734 3735 3736
	if (!how && !idev->cnf.disable_ipv6) {
		/* aggregate the system setting and interface setting */
		int _keep_addr = net->ipv6.devconf_all->keep_addr_on_down;

		if (!_keep_addr)
			_keep_addr = idev->cnf.keep_addr_on_down;

		keep_addr = (_keep_addr > 0);
	}
3737

3738 3739 3740 3741 3742
	/* 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);
3743
restart:
3744
		hlist_for_each_entry_rcu(ifa, h, addr_lst) {
3745
			if (ifa->idev == idev) {
3746
				addrconf_del_dad_work(ifa);
3747 3748 3749 3750
				/* combined flag + permanent flag decide if
				 * address is retained on a down event
				 */
				if (!keep_addr ||
3751 3752
				    !(ifa->flags & IFA_F_PERMANENT) ||
				    addr_is_local(&ifa->addr)) {
3753 3754 3755
					hlist_del_init_rcu(&ifa->addr_lst);
					goto restart;
				}
3756 3757 3758 3759 3760
			}
		}
		spin_unlock_bh(&addrconf_hash_lock);
	}

L
Linus Torvalds 已提交
3761 3762
	write_lock_bh(&idev->lock);

3763 3764
	addrconf_del_rs_timer(idev);

S
Stephen Hemminger 已提交
3765
	/* Step 2: clear flags for stateless addrconf */
3766
	if (!how)
3767
		idev->if_flags &= ~(IF_RS_SENT|IF_RA_RCVD|IF_READY);
L
Linus Torvalds 已提交
3768

S
Stephen Hemminger 已提交
3769
	/* Step 3: clear tempaddr list */
3770 3771 3772 3773
	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
3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784
		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);
	}
3785

3786
	list_for_each_entry_safe(ifa, tmp, &idev->addr_list, if_list) {
3787
		struct fib6_info *rt = NULL;
3788
		bool keep;
3789

3790
		addrconf_del_dad_work(ifa);
3791

3792 3793 3794
		keep = keep_addr && (ifa->flags & IFA_F_PERMANENT) &&
			!addr_is_local(&ifa->addr);

3795
		write_unlock_bh(&idev->lock);
3796
		spin_lock_bh(&ifa->lock);
3797

3798
		if (keep) {
3799 3800
			/* set state to skip the notifier below */
			state = INET6_IFADDR_STATE_DEAD;
3801
			ifa->state = INET6_IFADDR_STATE_PREDAD;
3802 3803
			if (!(ifa->flags & IFA_F_NODAD))
				ifa->flags |= IFA_F_TENTATIVE;
3804 3805 3806

			rt = ifa->rt;
			ifa->rt = NULL;
3807 3808 3809 3810 3811
		} else {
			state = ifa->state;
			ifa->state = INET6_IFADDR_STATE_DEAD;
		}

3812
		spin_unlock_bh(&ifa->lock);
3813

3814
		if (rt)
3815
			ip6_del_rt(net, rt);
3816

3817 3818
		if (state != INET6_IFADDR_STATE_DEAD) {
			__ipv6_ifa_notify(RTM_DELADDR, ifa);
3819
			inet6addr_notifier_call_chain(NETDEV_DOWN, ifa);
3820 3821 3822 3823
		} else {
			if (idev->cnf.forwarding)
				addrconf_leave_anycast(ifa);
			addrconf_leave_solict(ifa->idev, &ifa->addr);
3824
		}
3825

3826
		write_lock_bh(&idev->lock);
3827 3828 3829 3830
		if (!keep) {
			list_del_rcu(&ifa->if_list);
			in6_ifa_put(ifa);
		}
3831
	}
3832

L
Linus Torvalds 已提交
3833 3834
	write_unlock_bh(&idev->lock);

3835 3836 3837
	/* Step 5: Discard anycast and multicast list */
	if (how) {
		ipv6_ac_destroy_dev(idev);
L
Linus Torvalds 已提交
3838
		ipv6_mc_destroy_dev(idev);
3839
	} else {
L
Linus Torvalds 已提交
3840
		ipv6_mc_down(idev);
3841
	}
L
Linus Torvalds 已提交
3842 3843

	idev->tstamp = jiffies;
3844

S
Stephen Hemminger 已提交
3845
	/* Last: Shot the device (if unregistered) */
3846
	if (how) {
3847
		addrconf_sysctl_unregister(idev);
L
Linus Torvalds 已提交
3848 3849 3850 3851 3852 3853 3854
		neigh_parms_release(&nd_tbl, idev->nd_parms);
		neigh_ifdown(&nd_tbl, dev);
		in6_dev_put(idev);
	}
	return 0;
}

3855
static void addrconf_rs_timer(struct timer_list *t)
L
Linus Torvalds 已提交
3856
{
3857
	struct inet6_dev *idev = from_timer(idev, t, rs_timer);
3858
	struct net_device *dev = idev->dev;
3859
	struct in6_addr lladdr;
L
Linus Torvalds 已提交
3860

3861
	write_lock(&idev->lock);
S
stephen hemminger 已提交
3862
	if (idev->dead || !(idev->if_flags & IF_READY))
L
Linus Torvalds 已提交
3863 3864
		goto out;

3865
	if (!ipv6_accept_ra(idev))
S
stephen hemminger 已提交
3866 3867 3868 3869
		goto out;

	/* Announcement received after solicitation was sent */
	if (idev->if_flags & IF_RA_RCVD)
L
Linus Torvalds 已提交
3870 3871
		goto out;

3872
	if (idev->rs_probes++ < idev->cnf.rtr_solicits || idev->cnf.rtr_solicits < 0) {
3873 3874 3875
		write_unlock(&idev->lock);
		if (!ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
			ndisc_send_rs(dev, &lladdr,
3876 3877
				      &in6addr_linklocal_allrouters);
		else
3878
			goto put;
L
Linus Torvalds 已提交
3879

3880
		write_lock(&idev->lock);
3881 3882
		idev->rs_interval = rfc3315_s14_backoff_update(
			idev->rs_interval, idev->cnf.rtr_solicit_max_interval);
3883 3884 3885 3886
		/* 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 :
3887
				      idev->rs_interval);
L
Linus Torvalds 已提交
3888 3889 3890 3891 3892
	} else {
		/*
		 * Note: we do not support deprecated "all on-link"
		 * assumption any longer.
		 */
3893
		pr_debug("%s: no IPv6 routers present\n", idev->dev->name);
L
Linus Torvalds 已提交
3894 3895 3896
	}

out:
3897
	write_unlock(&idev->lock);
3898
put:
3899
	in6_dev_put(idev);
L
Linus Torvalds 已提交
3900 3901 3902 3903 3904
}

/*
 *	Duplicate Address Detection
 */
3905 3906 3907 3908
static void addrconf_dad_kick(struct inet6_ifaddr *ifp)
{
	unsigned long rand_num;
	struct inet6_dev *idev = ifp->idev;
3909
	u64 nonce;
3910

3911 3912 3913
	if (ifp->flags & IFA_F_OPTIMISTIC)
		rand_num = 0;
	else
3914
		rand_num = prandom_u32() % (idev->cnf.rtr_solicit_delay ? : 1);
3915

3916 3917 3918 3919 3920 3921 3922 3923
	nonce = 0;
	if (idev->cnf.enhanced_dad ||
	    dev_net(idev->dev)->ipv6.devconf_all->enhanced_dad) {
		do
			get_random_bytes(&nonce, 6);
		while (nonce == 0);
	}
	ifp->dad_nonce = nonce;
3924
	ifp->dad_probes = idev->cnf.dad_transmits;
3925
	addrconf_mod_dad_work(ifp, rand_num);
3926 3927
}

3928
static void addrconf_dad_begin(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3929 3930 3931
{
	struct inet6_dev *idev = ifp->idev;
	struct net_device *dev = idev->dev;
3932
	bool bump_id, notify = false;
3933
	struct net *net;
L
Linus Torvalds 已提交
3934 3935 3936

	addrconf_join_solict(dev, &ifp->addr);

3937
	prandom_seed((__force u32) ifp->addr.s6_addr32[3]);
L
Linus Torvalds 已提交
3938 3939

	read_lock_bh(&idev->lock);
3940
	spin_lock(&ifp->lock);
3941
	if (ifp->state == INET6_IFADDR_STATE_DEAD)
L
Linus Torvalds 已提交
3942 3943
		goto out;

3944
	net = dev_net(dev);
L
Linus Torvalds 已提交
3945
	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
3946
	    (net->ipv6.devconf_all->accept_dad < 1 &&
3947
	     idev->cnf.accept_dad < 1) ||
3948 3949
	    !(ifp->flags&IFA_F_TENTATIVE) ||
	    ifp->flags & IFA_F_NODAD) {
3950 3951 3952 3953 3954
		bool send_na = false;

		if (ifp->flags & IFA_F_TENTATIVE &&
		    !(ifp->flags & IFA_F_OPTIMISTIC))
			send_na = true;
3955
		bump_id = ifp->flags & IFA_F_TENTATIVE;
B
Brian Haley 已提交
3956
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3957
		spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3958 3959
		read_unlock_bh(&idev->lock);

3960
		addrconf_dad_completed(ifp, bump_id, send_na);
L
Linus Torvalds 已提交
3961 3962 3963
		return;
	}

3964
	if (!(idev->if_flags & IF_READY)) {
3965
		spin_unlock(&ifp->lock);
3966
		read_unlock_bh(&idev->lock);
3967
		/*
3968
		 * If the device is not ready:
3969 3970 3971 3972
		 * - keep it tentative if it is a permanent address.
		 * - otherwise, kill it.
		 */
		in6_ifa_hold(ifp);
B
Brian Haley 已提交
3973
		addrconf_dad_stop(ifp, 0);
3974
		return;
3975
	}
3976 3977 3978 3979 3980

	/*
	 * Optimistic nodes can start receiving
	 * Frames right away
	 */
3981
	if (ifp->flags & IFA_F_OPTIMISTIC) {
3982 3983
		ip6_ins_rt(net, ifp->rt);
		if (ipv6_use_optimistic_addr(net, idev)) {
3984 3985 3986
			/* Because optimistic nodes can use this address,
			 * notify listeners. If DAD fails, RTM_DELADDR is sent.
			 */
3987
			notify = true;
3988 3989
		}
	}
3990

3991
	addrconf_dad_kick(ifp);
L
Linus Torvalds 已提交
3992
out:
3993
	spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3994
	read_unlock_bh(&idev->lock);
3995 3996
	if (notify)
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
L
Linus Torvalds 已提交
3997 3998
}

3999
static void addrconf_dad_start(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
4000
{
4001 4002
	bool begin_dad = false;

4003
	spin_lock_bh(&ifp->lock);
4004 4005 4006 4007
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
		ifp->state = INET6_IFADDR_STATE_PREDAD;
		begin_dad = true;
	}
4008
	spin_unlock_bh(&ifp->lock);
4009 4010 4011 4012 4013 4014 4015 4016 4017 4018

	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
4019
	struct inet6_dev *idev = ifp->idev;
4020
	bool bump_id, disable_ipv6 = false;
L
Linus Torvalds 已提交
4021 4022
	struct in6_addr mcaddr;

4023 4024 4025 4026 4027 4028 4029 4030
	enum {
		DAD_PROCESS,
		DAD_BEGIN,
		DAD_ABORT,
	} action = DAD_PROCESS;

	rtnl_lock();

4031
	spin_lock_bh(&ifp->lock);
4032 4033 4034 4035 4036 4037
	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;
4038

4039 4040 4041
		if ((dev_net(idev->dev)->ipv6.devconf_all->accept_dad > 1 ||
		     idev->cnf.accept_dad > 1) &&
		    !idev->cnf.disable_ipv6 &&
4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057
		    !(ifp->flags & IFA_F_STABLE_PRIVACY)) {
			struct in6_addr addr;

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

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

				pr_info("%s: IPv6 being disabled!\n",
					ifp->idev->dev->name);
				disable_ipv6 = true;
			}
		}
4058
	}
4059
	spin_unlock_bh(&ifp->lock);
4060 4061 4062 4063 4064

	if (action == DAD_BEGIN) {
		addrconf_dad_begin(ifp);
		goto out;
	} else if (action == DAD_ABORT) {
4065
		in6_ifa_hold(ifp);
4066
		addrconf_dad_stop(ifp, 1);
4067 4068
		if (disable_ipv6)
			addrconf_ifdown(idev->dev, 0);
4069 4070 4071
		goto out;
	}

4072
	if (!ifp->dad_probes && addrconf_dad_end(ifp))
H
Herbert Xu 已提交
4073 4074
		goto out;

4075
	write_lock_bh(&idev->lock);
4076
	if (idev->dead || !(idev->if_flags & IF_READY)) {
4077
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
4078 4079
		goto out;
	}
4080 4081

	spin_lock(&ifp->lock);
4082 4083
	if (ifp->state == INET6_IFADDR_STATE_DEAD) {
		spin_unlock(&ifp->lock);
4084
		write_unlock_bh(&idev->lock);
4085 4086 4087
		goto out;
	}

4088
	if (ifp->dad_probes == 0) {
4089 4090
		bool send_na = false;

L
Linus Torvalds 已提交
4091 4092 4093 4094
		/*
		 * DAD was successful
		 */

4095 4096 4097
		if (ifp->flags & IFA_F_TENTATIVE &&
		    !(ifp->flags & IFA_F_OPTIMISTIC))
			send_na = true;
4098
		bump_id = ifp->flags & IFA_F_TENTATIVE;
B
Brian Haley 已提交
4099
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
4100
		spin_unlock(&ifp->lock);
4101
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
4102

4103
		addrconf_dad_completed(ifp, bump_id, send_na);
L
Linus Torvalds 已提交
4104 4105 4106 4107

		goto out;
	}

4108
	ifp->dad_probes--;
4109 4110
	addrconf_mod_dad_work(ifp,
			      NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME));
4111
	spin_unlock(&ifp->lock);
4112
	write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
4113 4114 4115

	/* send a neighbour solicitation for our addr */
	addrconf_addr_solict_mult(&ifp->addr, &mcaddr);
4116 4117
	ndisc_send_ns(ifp->idev->dev, &ifp->addr, &mcaddr, &in6addr_any,
		      ifp->dad_nonce);
L
Linus Torvalds 已提交
4118 4119
out:
	in6_ifa_put(ifp);
4120
	rtnl_unlock();
L
Linus Torvalds 已提交
4121 4122
}

4123 4124 4125 4126 4127 4128
/* 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;

4129 4130 4131
	list_for_each_entry_reverse(ifpiter, &idev->addr_list, if_list) {
		if (ifpiter->scope > IFA_LINK)
			break;
4132 4133 4134 4135 4136 4137 4138 4139 4140
		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;
}

4141 4142
static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id,
				   bool send_na)
L
Linus Torvalds 已提交
4143
{
S
Stephen Hemminger 已提交
4144
	struct net_device *dev = ifp->idev->dev;
4145
	struct in6_addr lladdr;
4146
	bool send_rs, send_mld;
4147

4148
	addrconf_del_dad_work(ifp);
L
Linus Torvalds 已提交
4149 4150 4151 4152 4153 4154 4155

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

	ipv6_ifa_notify(RTM_NEWADDR, ifp);

4156 4157
	/* If added prefix is link local and we are prepared to process
	   router advertisements, start sending router solicitations.
L
Linus Torvalds 已提交
4158 4159
	 */

4160
	read_lock_bh(&ifp->idev->lock);
4161
	send_mld = ifp->scope == IFA_LINK && ipv6_lonely_lladdr(ifp);
4162 4163
	send_rs = send_mld &&
		  ipv6_accept_ra(ifp->idev) &&
4164
		  ifp->idev->cnf.rtr_solicits != 0 &&
4165
		  (dev->flags&IFF_LOOPBACK) == 0;
4166 4167
	read_unlock_bh(&ifp->idev->lock);

4168 4169 4170 4171 4172 4173
	/* 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);

4174 4175 4176 4177 4178 4179 4180 4181 4182 4183
	/* send unsolicited NA if enabled */
	if (send_na &&
	    (ifp->idev->cnf.ndisc_notify ||
	     dev_net(dev)->ipv6.devconf_all->ndisc_notify)) {
		ndisc_send_na(dev, &in6addr_linklocal_allnodes, &ifp->addr,
			      /*router=*/ !!ifp->idev->cnf.forwarding,
			      /*solicited=*/ false, /*override=*/ true,
			      /*inc_opt=*/ true);
	}

4184
	if (send_rs) {
L
Linus Torvalds 已提交
4185 4186 4187 4188 4189
		/*
		 *	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
		 */
4190
		if (ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
4191
			return;
4192
		ndisc_send_rs(dev, &lladdr, &in6addr_linklocal_allrouters);
L
Linus Torvalds 已提交
4193

4194 4195
		write_lock_bh(&ifp->idev->lock);
		spin_lock(&ifp->lock);
4196 4197
		ifp->idev->rs_interval = rfc3315_s14_backoff_init(
			ifp->idev->cnf.rtr_solicit_interval);
4198
		ifp->idev->rs_probes = 1;
L
Linus Torvalds 已提交
4199
		ifp->idev->if_flags |= IF_RS_SENT;
4200
		addrconf_mod_rs_timer(ifp->idev, ifp->idev->rs_interval);
4201 4202
		spin_unlock(&ifp->lock);
		write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
4203
	}
4204 4205 4206

	if (bump_id)
		rt_genid_bump_ipv6(dev_net(dev));
4207 4208 4209 4210 4211 4212

	/* Make sure that a new temporary address will be created
	 * before this temporary address becomes deprecated.
	 */
	if (ifp->flags & IFA_F_TEMPORARY)
		addrconf_verify_rtnl();
L
Linus Torvalds 已提交
4213 4214
}

4215
static void addrconf_dad_run(struct inet6_dev *idev, bool restart)
S
Stephen Hemminger 已提交
4216
{
4217 4218 4219
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
4220
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
4221
		spin_lock(&ifp->lock);
4222 4223 4224 4225
		if ((ifp->flags & IFA_F_TENTATIVE &&
		     ifp->state == INET6_IFADDR_STATE_DAD) || restart) {
			if (restart)
				ifp->state = INET6_IFADDR_STATE_PREDAD;
H
Herbert Xu 已提交
4226
			addrconf_dad_kick(ifp);
4227
		}
4228
		spin_unlock(&ifp->lock);
4229 4230 4231 4232
	}
	read_unlock_bh(&idev->lock);
}

L
Linus Torvalds 已提交
4233 4234
#ifdef CONFIG_PROC_FS
struct if6_iter_state {
4235
	struct seq_net_private p;
L
Linus Torvalds 已提交
4236
	int bucket;
4237
	int offset;
L
Linus Torvalds 已提交
4238 4239
};

4240
static struct inet6_ifaddr *if6_get_first(struct seq_file *seq, loff_t pos)
L
Linus Torvalds 已提交
4241 4242
{
	struct if6_iter_state *state = seq->private;
4243
	struct net *net = seq_file_net(seq);
4244
	struct inet6_ifaddr *ifa = NULL;
4245
	int p = 0;
L
Linus Torvalds 已提交
4246

4247 4248 4249 4250 4251 4252 4253
	/* initial bucket if pos is 0 */
	if (pos == 0) {
		state->bucket = 0;
		state->offset = 0;
	}

	for (; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
4254
		hlist_for_each_entry_rcu(ifa, &inet6_addr_lst[state->bucket],
4255
					 addr_lst) {
4256 4257
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
4258 4259 4260 4261 4262
			/* sync with offset */
			if (p < state->offset) {
				p++;
				continue;
			}
4263
			return ifa;
4264 4265 4266 4267 4268
		}

		/* prepare for next bucket */
		state->offset = 0;
		p = 0;
L
Linus Torvalds 已提交
4269
	}
4270
	return NULL;
L
Linus Torvalds 已提交
4271 4272
}

4273 4274
static struct inet6_ifaddr *if6_get_next(struct seq_file *seq,
					 struct inet6_ifaddr *ifa)
L
Linus Torvalds 已提交
4275 4276
{
	struct if6_iter_state *state = seq->private;
4277
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
4278

4279
	hlist_for_each_entry_continue_rcu(ifa, addr_lst) {
4280 4281
		if (!net_eq(dev_net(ifa->idev->dev), net))
			continue;
4282
		state->offset++;
4283
		return ifa;
4284
	}
4285

4286
	state->offset = 0;
4287
	while (++state->bucket < IN6_ADDR_HSIZE) {
4288
		hlist_for_each_entry_rcu(ifa,
4289
				     &inet6_addr_lst[state->bucket], addr_lst) {
4290 4291 4292
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
			return ifa;
4293
		}
L
Linus Torvalds 已提交
4294
	}
4295

4296
	return NULL;
L
Linus Torvalds 已提交
4297 4298 4299
}

static void *if6_seq_start(struct seq_file *seq, loff_t *pos)
4300
	__acquires(rcu)
L
Linus Torvalds 已提交
4301
{
4302
	rcu_read_lock();
4303
	return if6_get_first(seq, *pos);
L
Linus Torvalds 已提交
4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315
}

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)
4316
	__releases(rcu)
L
Linus Torvalds 已提交
4317
{
4318
	rcu_read_unlock();
L
Linus Torvalds 已提交
4319 4320 4321 4322 4323
}

static int if6_seq_show(struct seq_file *seq, void *v)
{
	struct inet6_ifaddr *ifp = (struct inet6_ifaddr *)v;
4324
	seq_printf(seq, "%pi6 %02x %02x %02x %02x %8s\n",
4325
		   &ifp->addr,
L
Linus Torvalds 已提交
4326 4327 4328
		   ifp->idev->dev->ifindex,
		   ifp->prefix_len,
		   ifp->scope,
4329
		   (u8) ifp->flags,
L
Linus Torvalds 已提交
4330 4331 4332 4333
		   ifp->idev->dev->name);
	return 0;
}

4334
static const struct seq_operations if6_seq_ops = {
L
Linus Torvalds 已提交
4335 4336 4337 4338 4339 4340
	.start	= if6_seq_start,
	.next	= if6_seq_next,
	.show	= if6_seq_show,
	.stop	= if6_seq_stop,
};

4341
static int __net_init if6_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
4342
{
4343 4344
	if (!proc_create_net("if_inet6", 0444, net->proc_net, &if6_seq_ops,
			sizeof(struct if6_iter_state)))
L
Linus Torvalds 已提交
4345 4346 4347 4348
		return -ENOMEM;
	return 0;
}

4349
static void __net_exit if6_proc_net_exit(struct net *net)
4350
{
4351
	remove_proc_entry("if_inet6", net->proc_net);
4352 4353 4354
}

static struct pernet_operations if6_proc_net_ops = {
4355 4356
	.init = if6_proc_net_init,
	.exit = if6_proc_net_exit,
4357 4358 4359 4360 4361 4362 4363
};

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

L
Linus Torvalds 已提交
4364 4365
void if6_proc_exit(void)
{
4366
	unregister_pernet_subsys(&if6_proc_net_ops);
L
Linus Torvalds 已提交
4367 4368 4369
}
#endif	/* CONFIG_PROC_FS */

A
Amerigo Wang 已提交
4370
#if IS_ENABLED(CONFIG_IPV6_MIP6)
4371
/* Check if address is a home address configured on any interface. */
4372
int ipv6_chk_home_addr(struct net *net, const struct in6_addr *addr)
4373
{
4374
	unsigned int hash = inet6_addr_hash(net, addr);
4375
	struct inet6_ifaddr *ifp = NULL;
4376
	int ret = 0;
4377

4378 4379
	rcu_read_lock();
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
4380
		if (!net_eq(dev_net(ifp->idev->dev), net))
4381
			continue;
4382
		if (ipv6_addr_equal(&ifp->addr, addr) &&
4383 4384 4385 4386 4387
		    (ifp->flags & IFA_F_HOMEADDRESS)) {
			ret = 1;
			break;
		}
	}
4388
	rcu_read_unlock();
4389 4390 4391 4392
	return ret;
}
#endif

L
Linus Torvalds 已提交
4393 4394 4395 4396
/*
 *	Periodic address status verification
 */

4397
static void addrconf_verify_rtnl(void)
L
Linus Torvalds 已提交
4398
{
4399
	unsigned long now, next, next_sec, next_sched;
L
Linus Torvalds 已提交
4400 4401 4402
	struct inet6_ifaddr *ifp;
	int i;

4403 4404
	ASSERT_RTNL();

4405
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
4406
	now = jiffies;
4407
	next = round_jiffies_up(now + ADDR_CHECK_FREQUENCY);
L
Linus Torvalds 已提交
4408

4409
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
4410

S
Stephen Hemminger 已提交
4411
	for (i = 0; i < IN6_ADDR_HSIZE; i++) {
L
Linus Torvalds 已提交
4412
restart:
4413
		hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[i], addr_lst) {
L
Linus Torvalds 已提交
4414 4415
			unsigned long age;

4416 4417 4418 4419 4420 4421
			/* 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
4422 4423 4424
				continue;

			spin_lock(&ifp->lock);
4425 4426
			/* We try to batch several events at once. */
			age = (now - ifp->tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
L
Linus Torvalds 已提交
4427

4428 4429
			if (ifp->valid_lft != INFINITY_LIFE_TIME &&
			    age >= ifp->valid_lft) {
L
Linus Torvalds 已提交
4430 4431 4432 4433
				spin_unlock(&ifp->lock);
				in6_ifa_hold(ifp);
				ipv6_del_addr(ifp);
				goto restart;
4434 4435 4436
			} else if (ifp->prefered_lft == INFINITY_LIFE_TIME) {
				spin_unlock(&ifp->lock);
				continue;
L
Linus Torvalds 已提交
4437
			} else if (age >= ifp->prefered_lft) {
4438
				/* jiffies - ifp->tstamp > age >= ifp->prefered_lft */
L
Linus Torvalds 已提交
4439 4440 4441 4442 4443 4444 4445
				int deprecate = 0;

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

4446 4447
				if ((ifp->valid_lft != INFINITY_LIFE_TIME) &&
				    (time_before(ifp->tstamp + ifp->valid_lft * HZ, next)))
L
Linus Torvalds 已提交
4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460
					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)) {
4461 4462
				unsigned long regen_advance = ifp->idev->cnf.regen_max_retry *
					ifp->idev->cnf.dad_transmits *
J
Jiri Pirko 已提交
4463
					NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME) / HZ;
4464

L
Linus Torvalds 已提交
4465 4466 4467 4468 4469 4470 4471 4472 4473
				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);
4474

4475 4476 4477
						spin_lock(&ifpub->lock);
						ifpub->regen_count = 0;
						spin_unlock(&ifpub->lock);
4478
						rcu_read_unlock_bh();
E
Eric Dumazet 已提交
4479
						ipv6_create_tempaddr(ifpub, ifp, true);
L
Linus Torvalds 已提交
4480 4481
						in6_ifa_put(ifpub);
						in6_ifa_put(ifp);
4482
						rcu_read_lock_bh();
L
Linus Torvalds 已提交
4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496
						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);
			}
		}
	}

4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507
	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;

4508 4509
	pr_debug("now = %lu, schedule = %lu, rounded schedule = %lu => %lu\n",
		 now, next, next_sec, next_sched);
4510
	mod_delayed_work(addrconf_wq, &addr_chk_work, next_sched - now);
4511
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
4512 4513
}

4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525
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);
}

4526 4527
static struct in6_addr *extract_addr(struct nlattr *addr, struct nlattr *local,
				     struct in6_addr **peer_pfx)
4528 4529 4530
{
	struct in6_addr *pfx = NULL;

4531 4532
	*peer_pfx = NULL;

4533 4534 4535 4536 4537
	if (addr)
		pfx = nla_data(addr);

	if (local) {
		if (pfx && nla_memcmp(local, pfx, sizeof(*pfx)))
4538 4539
			*peer_pfx = pfx;
		pfx = nla_data(local);
4540 4541 4542 4543 4544
	}

	return pfx;
}

4545
static const struct nla_policy ifa_ipv6_policy[IFA_MAX+1] = {
4546 4547 4548
	[IFA_ADDRESS]		= { .len = sizeof(struct in6_addr) },
	[IFA_LOCAL]		= { .len = sizeof(struct in6_addr) },
	[IFA_CACHEINFO]		= { .len = sizeof(struct ifa_cacheinfo) },
4549
	[IFA_FLAGS]		= { .len = sizeof(u32) },
4550
	[IFA_RT_PRIORITY]	= { .len = sizeof(u32) },
4551
	[IFA_TARGET_NETNSID]	= { .type = NLA_S32 },
4552 4553
};

L
Linus Torvalds 已提交
4554
static int
4555 4556
inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh,
		  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
4557
{
4558
	struct net *net = sock_net(skb->sk);
4559 4560
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4561
	struct in6_addr *pfx, *peer_pfx;
4562
	u32 ifa_flags;
4563
	int err;
L
Linus Torvalds 已提交
4564

4565
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy,
4566
			  extack);
4567 4568 4569 4570
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
4571
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
4572
	if (!pfx)
L
Linus Torvalds 已提交
4573 4574
		return -EINVAL;

4575 4576 4577 4578 4579 4580 4581
	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
4582 4583
}

4584 4585 4586 4587
static int modify_prefix_route(struct inet6_ifaddr *ifp,
			       unsigned long expires, u32 flags)
{
	struct fib6_info *f6i;
4588
	u32 prio;
4589 4590 4591 4592 4593 4594 4595 4596

	f6i = addrconf_get_prefix_route(&ifp->addr,
					ifp->prefix_len,
					ifp->idev->dev,
					0, RTF_GATEWAY | RTF_DEFAULT);
	if (!f6i)
		return -ENOENT;

4597 4598 4599 4600 4601
	prio = ifp->rt_priority ? : IP6_RT_PRIO_ADDRCONF;
	if (f6i->fib6_metric != prio) {
		/* delete old one */
		ip6_del_rt(dev_net(ifp->idev->dev), f6i);

4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617
		/* add new one */
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len,
				      ifp->rt_priority, ifp->idev->dev,
				      expires, flags, GFP_KERNEL);
	} else {
		if (!expires)
			fib6_clean_expires(f6i);
		else
			fib6_set_expires(f6i, expires);

		fib6_info_release(f6i);
	}

	return 0;
}

4618
static int inet6_addr_modify(struct inet6_ifaddr *ifp, struct ifa6_config *cfg)
4619
{
4620 4621
	u32 flags;
	clock_t expires;
4622
	unsigned long timeout;
4623
	bool was_managetempaddr;
4624
	bool had_prefixroute;
4625

4626 4627
	ASSERT_RTNL();

4628
	if (!cfg->valid_lft || cfg->preferred_lft > cfg->valid_lft)
4629 4630
		return -EINVAL;

4631
	if (cfg->ifa_flags & IFA_F_MANAGETEMPADDR &&
4632 4633 4634
	    (ifp->flags & IFA_F_TEMPORARY || ifp->prefix_len != 64))
		return -EINVAL;

4635
	if (!(ifp->flags & IFA_F_TENTATIVE) || ifp->flags & IFA_F_DADFAILED)
4636
		cfg->ifa_flags &= ~IFA_F_OPTIMISTIC;
4637

4638
	timeout = addrconf_timeout_fixup(cfg->valid_lft, HZ);
4639 4640
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
4641
		cfg->valid_lft = timeout;
4642
		flags = RTF_EXPIRES;
4643 4644 4645
	} else {
		expires = 0;
		flags = 0;
4646
		cfg->ifa_flags |= IFA_F_PERMANENT;
4647
	}
4648

4649
	timeout = addrconf_timeout_fixup(cfg->preferred_lft, HZ);
4650 4651
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
4652 4653
			cfg->ifa_flags |= IFA_F_DEPRECATED;
		cfg->preferred_lft = timeout;
4654
	}
4655 4656

	spin_lock_bh(&ifp->lock);
4657
	was_managetempaddr = ifp->flags & IFA_F_MANAGETEMPADDR;
4658 4659
	had_prefixroute = ifp->flags & IFA_F_PERMANENT &&
			  !(ifp->flags & IFA_F_NOPREFIXROUTE);
4660
	ifp->flags &= ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD |
4661 4662
			IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
			IFA_F_NOPREFIXROUTE);
4663
	ifp->flags |= cfg->ifa_flags;
4664
	ifp->tstamp = jiffies;
4665 4666
	ifp->valid_lft = cfg->valid_lft;
	ifp->prefered_lft = cfg->preferred_lft;
4667

4668 4669
	if (cfg->rt_priority && cfg->rt_priority != ifp->rt_priority)
		ifp->rt_priority = cfg->rt_priority;
4670 4671 4672 4673 4674

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

4675
	if (!(cfg->ifa_flags & IFA_F_NOPREFIXROUTE)) {
4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686
		int rc = -ENOENT;

		if (had_prefixroute)
			rc = modify_prefix_route(ifp, expires, flags);

		/* prefix route could have been deleted; if so restore it */
		if (rc == -ENOENT) {
			addrconf_prefix_route(&ifp->addr, ifp->prefix_len,
					      ifp->rt_priority, ifp->idev->dev,
					      expires, flags, GFP_KERNEL);
		}
4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698
	} 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);
		}
4699
	}
4700 4701

	if (was_managetempaddr || ifp->flags & IFA_F_MANAGETEMPADDR) {
4702 4703 4704 4705 4706 4707 4708 4709
		if (was_managetempaddr &&
		    !(ifp->flags & IFA_F_MANAGETEMPADDR)) {
			cfg->valid_lft = 0;
			cfg->preferred_lft = 0;
		}
		manage_tempaddrs(ifp->idev, ifp, cfg->valid_lft,
				 cfg->preferred_lft, !was_managetempaddr,
				 jiffies);
4710 4711
	}

4712
	addrconf_verify_rtnl();
4713 4714 4715 4716

	return 0;
}

L
Linus Torvalds 已提交
4717
static int
4718 4719
inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh,
		  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
4720
{
4721
	struct net *net = sock_net(skb->sk);
4722 4723
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4724
	struct in6_addr *peer_pfx;
4725 4726
	struct inet6_ifaddr *ifa;
	struct net_device *dev;
4727
	struct inet6_dev *idev;
4728
	struct ifa6_config cfg;
4729
	int err;
L
Linus Torvalds 已提交
4730

4731
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy,
4732
			  extack);
4733 4734 4735
	if (err < 0)
		return err;

4736 4737
	memset(&cfg, 0, sizeof(cfg));

4738
	ifm = nlmsg_data(nlh);
4739 4740
	cfg.pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
	if (!cfg.pfx)
L
Linus Torvalds 已提交
4741 4742
		return -EINVAL;

4743 4744
	cfg.peer_pfx = peer_pfx;
	cfg.plen = ifm->ifa_prefixlen;
4745 4746 4747
	if (tb[IFA_RT_PRIORITY])
		cfg.rt_priority = nla_get_u32(tb[IFA_RT_PRIORITY]);

4748 4749 4750
	cfg.valid_lft = INFINITY_LIFE_TIME;
	cfg.preferred_lft = INFINITY_LIFE_TIME;

4751
	if (tb[IFA_CACHEINFO]) {
4752
		struct ifa_cacheinfo *ci;
4753 4754

		ci = nla_data(tb[IFA_CACHEINFO]);
4755 4756
		cfg.valid_lft = ci->ifa_valid;
		cfg.preferred_lft = ci->ifa_prefered;
4757 4758
	}

4759
	dev =  __dev_get_by_index(net, ifm->ifa_index);
4760
	if (!dev)
4761 4762
		return -ENODEV;

4763 4764 4765 4766
	if (tb[IFA_FLAGS])
		cfg.ifa_flags = nla_get_u32(tb[IFA_FLAGS]);
	else
		cfg.ifa_flags = ifm->ifa_flags;
4767

4768
	/* We ignore other flags so far. */
4769 4770 4771
	cfg.ifa_flags &= IFA_F_NODAD | IFA_F_HOMEADDRESS |
			 IFA_F_MANAGETEMPADDR | IFA_F_NOPREFIXROUTE |
			 IFA_F_MCAUTOJOIN | IFA_F_OPTIMISTIC;
4772 4773

	idev = ipv6_find_idev(dev);
4774 4775
	if (!idev)
		return -ENOBUFS;
4776 4777

	if (!ipv6_allow_optimistic_dad(net, idev))
4778
		cfg.ifa_flags &= ~IFA_F_OPTIMISTIC;
4779

4780 4781
	if (cfg.ifa_flags & IFA_F_NODAD &&
	    cfg.ifa_flags & IFA_F_OPTIMISTIC) {
4782 4783 4784
		NL_SET_ERR_MSG(extack, "IFA_F_NODAD and IFA_F_OPTIMISTIC are mutually exclusive");
		return -EINVAL;
	}
4785

4786
	ifa = ipv6_get_ifaddr(net, cfg.pfx, dev, 1);
4787
	if (!ifa) {
4788 4789
		/*
		 * It would be best to check for !NLM_F_CREATE here but
4790
		 * userspace already relies on not having to provide this.
4791
		 */
4792
		return inet6_addr_add(net, ifm->ifa_index, &cfg, extack);
4793 4794
	}

4795 4796 4797 4798
	if (nlh->nlmsg_flags & NLM_F_EXCL ||
	    !(nlh->nlmsg_flags & NLM_F_REPLACE))
		err = -EEXIST;
	else
4799
		err = inet6_addr_modify(ifa, &cfg);
4800 4801 4802 4803

	in6_ifa_put(ifa);

	return err;
L
Linus Torvalds 已提交
4804 4805
}

4806
static void put_ifaddrmsg(struct nlmsghdr *nlh, u8 prefixlen, u32 flags,
4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818
			  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;
}

4819 4820 4821 4822 4823
static int put_cacheinfo(struct sk_buff *skb, unsigned long cstamp,
			 unsigned long tstamp, u32 preferred, u32 valid)
{
	struct ifa_cacheinfo ci;

4824 4825
	ci.cstamp = cstamp_delta(cstamp);
	ci.tstamp = cstamp_delta(tstamp);
4826 4827 4828 4829 4830 4831
	ci.ifa_prefered = preferred;
	ci.ifa_valid = valid;

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

4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843
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;
}

4844 4845
static inline int inet6_ifaddr_msgsize(void)
{
4846
	return NLMSG_ALIGN(sizeof(struct ifaddrmsg))
4847
	       + nla_total_size(16) /* IFA_LOCAL */
4848
	       + nla_total_size(16) /* IFA_ADDRESS */
4849
	       + nla_total_size(sizeof(struct ifa_cacheinfo))
4850 4851
	       + nla_total_size(4)  /* IFA_FLAGS */
	       + nla_total_size(4)  /* IFA_RT_PRIORITY */;
4852
}
4853

4854 4855 4856 4857 4858 4859
enum addr_type_t {
	UNICAST_ADDR,
	MULTICAST_ADDR,
	ANYCAST_ADDR,
};

C
Christian Brauner 已提交
4860 4861 4862 4863 4864 4865
struct inet6_fill_args {
	u32 portid;
	u32 seq;
	int event;
	unsigned int flags;
	int netnsid;
4866
	int ifindex;
4867
	enum addr_type_t type;
C
Christian Brauner 已提交
4868 4869
};

L
Linus Torvalds 已提交
4870
static int inet6_fill_ifaddr(struct sk_buff *skb, struct inet6_ifaddr *ifa,
C
Christian Brauner 已提交
4871
			     struct inet6_fill_args *args)
L
Linus Torvalds 已提交
4872 4873
{
	struct nlmsghdr  *nlh;
4874
	u32 preferred, valid;
L
Linus Torvalds 已提交
4875

C
Christian Brauner 已提交
4876 4877
	nlh = nlmsg_put(skb, args->portid, args->seq, args->event,
			sizeof(struct ifaddrmsg), args->flags);
4878
	if (!nlh)
4879
		return -EMSGSIZE;
4880

4881 4882 4883
	put_ifaddrmsg(nlh, ifa->prefix_len, ifa->flags, rt_scope(ifa->scope),
		      ifa->idev->dev->ifindex);

C
Christian Brauner 已提交
4884 4885
	if (args->netnsid >= 0 &&
	    nla_put_s32(skb, IFA_TARGET_NETNSID, args->netnsid))
4886 4887
		goto error;

4888 4889
	if (!((ifa->flags&IFA_F_PERMANENT) &&
	      (ifa->prefered_lft == INFINITY_LIFE_TIME))) {
4890 4891 4892
		preferred = ifa->prefered_lft;
		valid = ifa->valid_lft;
		if (preferred != INFINITY_LIFE_TIME) {
L
Linus Torvalds 已提交
4893
			long tval = (jiffies - ifa->tstamp)/HZ;
4894 4895 4896 4897
			if (preferred > tval)
				preferred -= tval;
			else
				preferred = 0;
4898 4899 4900 4901 4902 4903
			if (valid != INFINITY_LIFE_TIME) {
				if (valid > tval)
					valid -= tval;
				else
					valid = 0;
			}
L
Linus Torvalds 已提交
4904 4905
		}
	} else {
4906 4907 4908 4909
		preferred = INFINITY_LIFE_TIME;
		valid = INFINITY_LIFE_TIME;
	}

C
Cong Wang 已提交
4910
	if (!ipv6_addr_any(&ifa->peer_addr)) {
4911 4912
		if (nla_put_in6_addr(skb, IFA_LOCAL, &ifa->addr) < 0 ||
		    nla_put_in6_addr(skb, IFA_ADDRESS, &ifa->peer_addr) < 0)
4913 4914
			goto error;
	} else
4915
		if (nla_put_in6_addr(skb, IFA_ADDRESS, &ifa->addr) < 0)
4916 4917
			goto error;

4918 4919 4920 4921
	if (ifa->rt_priority &&
	    nla_put_u32(skb, IFA_RT_PRIORITY, ifa->rt_priority))
		goto error;

4922 4923
	if (put_cacheinfo(skb, ifa->cstamp, ifa->tstamp, preferred, valid) < 0)
		goto error;
L
Linus Torvalds 已提交
4924

4925 4926 4927
	if (nla_put_u32(skb, IFA_FLAGS, ifa->flags) < 0)
		goto error;

4928 4929
	nlmsg_end(skb, nlh);
	return 0;
4930 4931 4932 4933

error:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4934 4935 4936
}

static int inet6_fill_ifmcaddr(struct sk_buff *skb, struct ifmcaddr6 *ifmca,
C
Christian Brauner 已提交
4937
			       struct inet6_fill_args *args)
L
Linus Torvalds 已提交
4938 4939
{
	struct nlmsghdr  *nlh;
4940 4941 4942 4943 4944
	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
4945

C
Christian Brauner 已提交
4946 4947
	nlh = nlmsg_put(skb, args->portid, args->seq, args->event,
			sizeof(struct ifaddrmsg), args->flags);
4948
	if (!nlh)
4949
		return -EMSGSIZE;
4950

C
Christian Brauner 已提交
4951 4952
	if (args->netnsid >= 0 &&
	    nla_put_s32(skb, IFA_TARGET_NETNSID, args->netnsid))
4953 4954
		return -EMSGSIZE;

4955
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
4956
	if (nla_put_in6_addr(skb, IFA_MULTICAST, &ifmca->mca_addr) < 0 ||
4957
	    put_cacheinfo(skb, ifmca->mca_cstamp, ifmca->mca_tstamp,
4958 4959 4960 4961
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
4962

4963 4964
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4965 4966 4967
}

static int inet6_fill_ifacaddr(struct sk_buff *skb, struct ifacaddr6 *ifaca,
C
Christian Brauner 已提交
4968
			       struct inet6_fill_args *args)
L
Linus Torvalds 已提交
4969
{
D
David Ahern 已提交
4970 4971
	struct net_device *dev = fib6_info_nh_dev(ifaca->aca_rt);
	int ifindex = dev ? dev->ifindex : 1;
L
Linus Torvalds 已提交
4972
	struct nlmsghdr  *nlh;
4973 4974 4975 4976
	u8 scope = RT_SCOPE_UNIVERSE;

	if (ipv6_addr_scope(&ifaca->aca_addr) & IFA_SITE)
		scope = RT_SCOPE_SITE;
L
Linus Torvalds 已提交
4977

C
Christian Brauner 已提交
4978 4979
	nlh = nlmsg_put(skb, args->portid, args->seq, args->event,
			sizeof(struct ifaddrmsg), args->flags);
4980
	if (!nlh)
4981
		return -EMSGSIZE;
4982

C
Christian Brauner 已提交
4983 4984
	if (args->netnsid >= 0 &&
	    nla_put_s32(skb, IFA_TARGET_NETNSID, args->netnsid))
4985 4986
		return -EMSGSIZE;

4987
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
4988
	if (nla_put_in6_addr(skb, IFA_ANYCAST, &ifaca->aca_addr) < 0 ||
4989
	    put_cacheinfo(skb, ifaca->aca_cstamp, ifaca->aca_tstamp,
4990 4991 4992 4993
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
L
Linus Torvalds 已提交
4994

4995 4996
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4997 4998
}

E
Eric Dumazet 已提交
4999 5000
/* called with rcu_read_lock() */
static int in6_dump_addrs(struct inet6_dev *idev, struct sk_buff *skb,
5001
			  struct netlink_callback *cb, int s_ip_idx,
5002
			  struct inet6_fill_args *fillargs)
E
Eric Dumazet 已提交
5003 5004 5005
{
	struct ifmcaddr6 *ifmca;
	struct ifacaddr6 *ifaca;
5006
	int ip_idx = 0;
E
Eric Dumazet 已提交
5007 5008 5009
	int err = 1;

	read_lock_bh(&idev->lock);
5010
	switch (fillargs->type) {
5011 5012
	case UNICAST_ADDR: {
		struct inet6_ifaddr *ifa;
5013
		fillargs->event = RTM_NEWADDR;
5014

E
Eric Dumazet 已提交
5015
		/* unicast address incl. temp addr */
5016
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
5017 5018
			if (ip_idx < s_ip_idx)
				goto next;
5019
			err = inet6_fill_ifaddr(skb, ifa, fillargs);
5020
			if (err < 0)
E
Eric Dumazet 已提交
5021
				break;
5022
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
5023 5024
next:
			ip_idx++;
E
Eric Dumazet 已提交
5025 5026
		}
		break;
5027
	}
E
Eric Dumazet 已提交
5028
	case MULTICAST_ADDR:
5029
		fillargs->event = RTM_GETMULTICAST;
C
Christian Brauner 已提交
5030

E
Eric Dumazet 已提交
5031 5032 5033 5034 5035
		/* multicast address */
		for (ifmca = idev->mc_list; ifmca;
		     ifmca = ifmca->next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
5036
			err = inet6_fill_ifmcaddr(skb, ifmca, fillargs);
5037
			if (err < 0)
E
Eric Dumazet 已提交
5038 5039 5040 5041
				break;
		}
		break;
	case ANYCAST_ADDR:
5042
		fillargs->event = RTM_GETANYCAST;
E
Eric Dumazet 已提交
5043 5044 5045 5046 5047
		/* anycast address */
		for (ifaca = idev->ac_list; ifaca;
		     ifaca = ifaca->aca_next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
5048
			err = inet6_fill_ifacaddr(skb, ifaca, fillargs);
5049
			if (err < 0)
E
Eric Dumazet 已提交
5050 5051 5052 5053 5054 5055 5056
				break;
		}
		break;
	default:
		break;
	}
	read_unlock_bh(&idev->lock);
5057
	cb->args[2] = ip_idx;
E
Eric Dumazet 已提交
5058 5059 5060
	return err;
}

5061 5062 5063
static int inet6_valid_dump_ifaddr_req(const struct nlmsghdr *nlh,
				       struct inet6_fill_args *fillargs,
				       struct net **tgt_net, struct sock *sk,
5064
				       struct netlink_callback *cb)
5065
{
5066
	struct netlink_ext_ack *extack = cb->extack;
5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080
	struct nlattr *tb[IFA_MAX+1];
	struct ifaddrmsg *ifm;
	int err, i;

	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*ifm))) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid header for address dump request");
		return -EINVAL;
	}

	ifm = nlmsg_data(nlh);
	if (ifm->ifa_prefixlen || ifm->ifa_flags || ifm->ifa_scope) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid values in header for address dump request");
		return -EINVAL;
	}
5081 5082 5083 5084 5085

	fillargs->ifindex = ifm->ifa_index;
	if (fillargs->ifindex) {
		cb->answer_flags |= NLM_F_DUMP_FILTERED;
		fillargs->flags |= NLM_F_DUMP_FILTERED;
5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102
	}

	err = nlmsg_parse_strict(nlh, sizeof(*ifm), tb, IFA_MAX,
				 ifa_ipv6_policy, extack);
	if (err < 0)
		return err;

	for (i = 0; i <= IFA_MAX; ++i) {
		if (!tb[i])
			continue;

		if (i == IFA_TARGET_NETNSID) {
			struct net *net;

			fillargs->netnsid = nla_get_s32(tb[i]);
			net = rtnl_get_net_ns_capable(sk, fillargs->netnsid);
			if (IS_ERR(net)) {
5103
				fillargs->netnsid = -1;
5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116
				NL_SET_ERR_MSG_MOD(extack, "Invalid target network namespace id");
				return PTR_ERR(net);
			}
			*tgt_net = net;
		} else {
			NL_SET_ERR_MSG_MOD(extack, "Unsupported attribute in dump request");
			return -EINVAL;
		}
	}

	return 0;
}

L
Linus Torvalds 已提交
5117 5118 5119
static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
			   enum addr_type_t type)
{
5120
	const struct nlmsghdr *nlh = cb->nlh;
5121 5122 5123 5124 5125 5126 5127
	struct inet6_fill_args fillargs = {
		.portid = NETLINK_CB(cb->skb).portid,
		.seq = cb->nlh->nlmsg_seq,
		.flags = NLM_F_MULTI,
		.netnsid = -1,
		.type = type,
	};
E
Eric Dumazet 已提交
5128
	struct net *net = sock_net(skb->sk);
5129
	struct net *tgt_net = net;
5130
	int idx, s_idx, s_ip_idx;
E
Eric Dumazet 已提交
5131
	int h, s_h;
L
Linus Torvalds 已提交
5132
	struct net_device *dev;
E
Eric Dumazet 已提交
5133 5134
	struct inet6_dev *idev;
	struct hlist_head *head;
5135
	int err = 0;
5136

E
Eric Dumazet 已提交
5137 5138
	s_h = cb->args[0];
	s_idx = idx = cb->args[1];
5139
	s_ip_idx = cb->args[2];
5140

5141 5142
	if (cb->strict_check) {
		err = inet6_valid_dump_ifaddr_req(nlh, &fillargs, &tgt_net,
5143
						  skb->sk, cb);
5144
		if (err < 0)
5145
			goto put_tgt_net;
5146

5147
		err = 0;
5148 5149
		if (fillargs.ifindex) {
			dev = __dev_get_by_index(tgt_net, fillargs.ifindex);
5150 5151 5152 5153
			if (!dev) {
				err = -ENODEV;
				goto put_tgt_net;
			}
5154 5155 5156 5157 5158 5159 5160
			idev = __in6_dev_get(dev);
			if (idev) {
				err = in6_dump_addrs(idev, skb, cb, s_ip_idx,
						     &fillargs);
			}
			goto put_tgt_net;
		}
5161 5162
	}

E
Eric Dumazet 已提交
5163
	rcu_read_lock();
5164
	cb->seq = atomic_read(&tgt_net->ipv6.dev_addr_genid) ^ tgt_net->dev_base_seq;
E
Eric Dumazet 已提交
5165 5166
	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
5167
		head = &tgt_net->dev_index_head[h];
5168
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
5169 5170
			if (idx < s_idx)
				goto cont;
5171
			if (h > s_h || idx > s_idx)
E
Eric Dumazet 已提交
5172
				s_ip_idx = 0;
S
Stephen Hemminger 已提交
5173 5174
			idev = __in6_dev_get(dev);
			if (!idev)
E
Eric Dumazet 已提交
5175 5176
				goto cont;

5177
			if (in6_dump_addrs(idev, skb, cb, s_ip_idx,
5178
					   &fillargs) < 0)
E
Eric Dumazet 已提交
5179
				goto done;
5180
cont:
E
Eric Dumazet 已提交
5181 5182
			idx++;
		}
L
Linus Torvalds 已提交
5183
	}
E
Eric Dumazet 已提交
5184 5185 5186 5187
done:
	rcu_read_unlock();
	cb->args[0] = h;
	cb->args[1] = idx;
5188
put_tgt_net:
5189
	if (fillargs.netnsid >= 0)
5190
		put_net(tgt_net);
E
Eric Dumazet 已提交
5191

5192
	return skb->len ? : err;
L
Linus Torvalds 已提交
5193 5194 5195 5196 5197
}

static int inet6_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = UNICAST_ADDR;
5198

L
Linus Torvalds 已提交
5199 5200 5201 5202 5203 5204
	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;
5205

L
Linus Torvalds 已提交
5206 5207 5208 5209 5210 5211 5212
	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;
5213

L
Linus Torvalds 已提交
5214 5215 5216
	return inet6_dump_addr(skb, cb, type);
}

5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262
static int inet6_rtm_valid_getaddr_req(struct sk_buff *skb,
				       const struct nlmsghdr *nlh,
				       struct nlattr **tb,
				       struct netlink_ext_ack *extack)
{
	struct ifaddrmsg *ifm;
	int i, err;

	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*ifm))) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid header for get address request");
		return -EINVAL;
	}

	ifm = nlmsg_data(nlh);
	if (ifm->ifa_prefixlen || ifm->ifa_flags || ifm->ifa_scope) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid values in header for get address request");
		return -EINVAL;
	}

	if (!netlink_strict_get_check(skb))
		return nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX,
				   ifa_ipv6_policy, extack);

	err = nlmsg_parse_strict(nlh, sizeof(*ifm), tb, IFA_MAX,
				 ifa_ipv6_policy, extack);
	if (err)
		return err;

	for (i = 0; i <= IFA_MAX; i++) {
		if (!tb[i])
			continue;

		switch (i) {
		case IFA_TARGET_NETNSID:
		case IFA_ADDRESS:
		case IFA_LOCAL:
			break;
		default:
			NL_SET_ERR_MSG_MOD(extack, "Unsupported attribute in get address request");
			return -EINVAL;
		}
	}

	return 0;
}

5263 5264
static int inet6_rtm_getaddr(struct sk_buff *in_skb, struct nlmsghdr *nlh,
			     struct netlink_ext_ack *extack)
5265
{
5266
	struct net *net = sock_net(in_skb->sk);
C
Christian Brauner 已提交
5267 5268 5269 5270 5271 5272 5273
	struct inet6_fill_args fillargs = {
		.portid = NETLINK_CB(in_skb).portid,
		.seq = nlh->nlmsg_seq,
		.event = RTM_NEWADDR,
		.flags = 0,
		.netnsid = -1,
	};
5274
	struct net *tgt_net = net;
5275 5276
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
5277
	struct in6_addr *addr = NULL, *peer;
5278 5279 5280 5281 5282
	struct net_device *dev = NULL;
	struct inet6_ifaddr *ifa;
	struct sk_buff *skb;
	int err;

5283
	err = inet6_rtm_valid_getaddr_req(in_skb, nlh, tb, extack);
5284
	if (err < 0)
5285
		return err;
5286

5287
	if (tb[IFA_TARGET_NETNSID]) {
C
Christian Brauner 已提交
5288
		fillargs.netnsid = nla_get_s32(tb[IFA_TARGET_NETNSID]);
5289 5290

		tgt_net = rtnl_get_net_ns_capable(NETLINK_CB(in_skb).sk,
C
Christian Brauner 已提交
5291
						  fillargs.netnsid);
5292 5293 5294 5295
		if (IS_ERR(tgt_net))
			return PTR_ERR(tgt_net);
	}

5296
	addr = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer);
5297 5298
	if (!addr)
		return -EINVAL;
5299

5300
	ifm = nlmsg_data(nlh);
5301
	if (ifm->ifa_index)
5302
		dev = dev_get_by_index(tgt_net, ifm->ifa_index);
5303

5304
	ifa = ipv6_get_ifaddr(tgt_net, addr, dev, 1);
S
Stephen Hemminger 已提交
5305
	if (!ifa) {
5306 5307 5308
		err = -EADDRNOTAVAIL;
		goto errout;
	}
5309

S
Stephen Hemminger 已提交
5310 5311
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_KERNEL);
	if (!skb) {
5312
		err = -ENOBUFS;
5313
		goto errout_ifa;
5314 5315
	}

C
Christian Brauner 已提交
5316
	err = inet6_fill_ifaddr(skb, ifa, &fillargs);
5317 5318 5319 5320 5321 5322
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout_ifa;
	}
5323
	err = rtnl_unicast(skb, tgt_net, NETLINK_CB(in_skb).portid);
5324
errout_ifa:
5325
	in6_ifa_put(ifa);
5326
errout:
5327 5328
	if (dev)
		dev_put(dev);
C
Christian Brauner 已提交
5329
	if (fillargs.netnsid >= 0)
5330 5331
		put_net(tgt_net);

5332 5333 5334
	return err;
}

L
Linus Torvalds 已提交
5335 5336 5337
static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa)
{
	struct sk_buff *skb;
5338
	struct net *net = dev_net(ifa->idev->dev);
C
Christian Brauner 已提交
5339 5340 5341 5342 5343 5344 5345
	struct inet6_fill_args fillargs = {
		.portid = 0,
		.seq = 0,
		.event = event,
		.flags = 0,
		.netnsid = -1,
	};
5346
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
5347

5348
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_ATOMIC);
5349
	if (!skb)
5350 5351
		goto errout;

C
Christian Brauner 已提交
5352
	err = inet6_fill_ifaddr(skb, ifa, &fillargs);
5353 5354 5355 5356 5357 5358
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5359 5360
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFADDR, NULL, GFP_ATOMIC);
	return;
5361 5362
errout:
	if (err < 0)
5363
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFADDR, err);
L
Linus Torvalds 已提交
5364 5365
}

D
Dave Jones 已提交
5366
static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
L
Linus Torvalds 已提交
5367 5368
				__s32 *array, int bytes)
{
5369 5370
	BUG_ON(bytes < (DEVCONF_MAX * 4));

L
Linus Torvalds 已提交
5371 5372 5373 5374 5375 5376 5377 5378 5379
	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;
5380 5381
	array[DEVCONF_RTR_SOLICIT_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_interval);
5382 5383
	array[DEVCONF_RTR_SOLICIT_MAX_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_max_interval);
5384 5385
	array[DEVCONF_RTR_SOLICIT_DELAY] =
		jiffies_to_msecs(cnf->rtr_solicit_delay);
L
Linus Torvalds 已提交
5386
	array[DEVCONF_FORCE_MLD_VERSION] = cnf->force_mld_version;
5387 5388 5389 5390
	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
5391 5392 5393 5394 5395 5396
	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;
5397
	array[DEVCONF_ACCEPT_RA_DEFRTR] = cnf->accept_ra_defrtr;
5398
	array[DEVCONF_ACCEPT_RA_MIN_HOP_LIMIT] = cnf->accept_ra_min_hop_limit;
5399
	array[DEVCONF_ACCEPT_RA_PINFO] = cnf->accept_ra_pinfo;
5400 5401
#ifdef CONFIG_IPV6_ROUTER_PREF
	array[DEVCONF_ACCEPT_RA_RTR_PREF] = cnf->accept_ra_rtr_pref;
5402 5403
	array[DEVCONF_RTR_PROBE_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_probe_interval);
N
Neil Horman 已提交
5404
#ifdef CONFIG_IPV6_ROUTE_INFO
5405
	array[DEVCONF_ACCEPT_RA_RT_INFO_MIN_PLEN] = cnf->accept_ra_rt_info_min_plen;
5406 5407
	array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
#endif
5408
#endif
5409
	array[DEVCONF_PROXY_NDP] = cnf->proxy_ndp;
5410
	array[DEVCONF_ACCEPT_SOURCE_ROUTE] = cnf->accept_source_route;
5411 5412
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
5413
	array[DEVCONF_USE_OPTIMISTIC] = cnf->use_optimistic;
5414
#endif
5415 5416 5417
#ifdef CONFIG_IPV6_MROUTE
	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
#endif
5418
	array[DEVCONF_DISABLE_IPV6] = cnf->disable_ipv6;
5419
	array[DEVCONF_ACCEPT_DAD] = cnf->accept_dad;
5420
	array[DEVCONF_FORCE_TLLAO] = cnf->force_tllao;
5421
	array[DEVCONF_NDISC_NOTIFY] = cnf->ndisc_notify;
5422
	array[DEVCONF_SUPPRESS_FRAG_NDISC] = cnf->suppress_frag_ndisc;
5423
	array[DEVCONF_ACCEPT_RA_FROM_LOCAL] = cnf->accept_ra_from_local;
5424
	array[DEVCONF_ACCEPT_RA_MTU] = cnf->accept_ra_mtu;
5425
	array[DEVCONF_IGNORE_ROUTES_WITH_LINKDOWN] = cnf->ignore_routes_with_linkdown;
5426
	/* we omit DEVCONF_STABLE_SECRET for now */
5427
	array[DEVCONF_USE_OIF_ADDRS_ONLY] = cnf->use_oif_addrs_only;
5428
	array[DEVCONF_DROP_UNICAST_IN_L2_MULTICAST] = cnf->drop_unicast_in_l2_multicast;
5429
	array[DEVCONF_DROP_UNSOLICITED_NA] = cnf->drop_unsolicited_na;
5430
	array[DEVCONF_KEEP_ADDR_ON_DOWN] = cnf->keep_addr_on_down;
5431
	array[DEVCONF_SEG6_ENABLED] = cnf->seg6_enabled;
5432 5433 5434
#ifdef CONFIG_IPV6_SEG6_HMAC
	array[DEVCONF_SEG6_REQUIRE_HMAC] = cnf->seg6_require_hmac;
#endif
5435
	array[DEVCONF_ENHANCED_DAD] = cnf->enhanced_dad;
5436
	array[DEVCONF_ADDR_GEN_MODE] = cnf->addr_gen_mode;
5437
	array[DEVCONF_DISABLE_POLICY] = cnf->disable_policy;
5438
	array[DEVCONF_NDISC_TCLASS] = cnf->ndisc_tclass;
L
Linus Torvalds 已提交
5439 5440
}

T
Thomas Graf 已提交
5441 5442 5443 5444 5445 5446
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 */
5447
	     + nla_total_size(ICMP6_MIB_MAX * 8) /* IFLA_INET6_ICMP6STATS */
5448 5449 5450
	     + nla_total_size(sizeof(struct in6_addr)) /* IFLA_INET6_TOKEN */
	     + nla_total_size(1) /* IFLA_INET6_ADDR_GEN_MODE */
	     + 0;
T
Thomas Graf 已提交
5451 5452
}

5453 5454 5455 5456 5457 5458 5459
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 */
5460
	       + nla_total_size(1) /* IFLA_OPERSTATE */
T
Thomas Graf 已提交
5461
	       + nla_total_size(inet6_ifla6_size()); /* IFLA_PROTINFO */
5462
}
5463

5464
static inline void __snmp6_fill_statsdev(u64 *stats, atomic_long_t *mib,
5465
					int bytes)
5466 5467
{
	int i;
5468
	int pad = bytes - sizeof(u64) * ICMP6_MIB_MAX;
5469 5470 5471
	BUG_ON(pad < 0);

	/* Use put_unaligned() because stats may not be aligned for u64. */
5472 5473
	put_unaligned(ICMP6_MIB_MAX, &stats[0]);
	for (i = 1; i < ICMP6_MIB_MAX; i++)
5474
		put_unaligned(atomic_long_read(&mib[i]), &stats[i]);
5475

5476
	memset(&stats[ICMP6_MIB_MAX], 0, pad);
5477 5478
}

W
WANG Cong 已提交
5479
static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
5480
					int bytes, size_t syncpoff)
5481
{
5482 5483 5484 5485
	int i, c;
	u64 buff[IPSTATS_MIB_MAX];
	int pad = bytes - sizeof(u64) * IPSTATS_MIB_MAX;

5486 5487
	BUG_ON(pad < 0);

5488 5489
	memset(buff, 0, sizeof(buff));
	buff[0] = IPSTATS_MIB_MAX;
5490

5491 5492 5493 5494 5495 5496 5497
	for_each_possible_cpu(c) {
		for (i = 1; i < IPSTATS_MIB_MAX; i++)
			buff[i] += snmp_get_cpu_field64(mib, c, i, syncpoff);
	}

	memcpy(stats, buff, IPSTATS_MIB_MAX * sizeof(u64));
	memset(&stats[IPSTATS_MIB_MAX], 0, pad);
5498 5499
}

5500 5501 5502
static void snmp6_fill_stats(u64 *stats, struct inet6_dev *idev, int attrtype,
			     int bytes)
{
S
Stephen Hemminger 已提交
5503
	switch (attrtype) {
5504
	case IFLA_INET6_STATS:
5505 5506
		__snmp6_fill_stats64(stats, idev->stats.ipv6, bytes,
				     offsetof(struct ipstats_mib, syncp));
5507 5508
		break;
	case IFLA_INET6_ICMP6STATS:
5509
		__snmp6_fill_statsdev(stats, idev->stats.icmpv6dev->mibs, bytes);
5510 5511 5512 5513
		break;
	}
}

5514 5515
static int inet6_fill_ifla6_attrs(struct sk_buff *skb, struct inet6_dev *idev,
				  u32 ext_filter_mask)
T
Thomas Graf 已提交
5516 5517 5518 5519
{
	struct nlattr *nla;
	struct ifla_cacheinfo ci;

D
David S. Miller 已提交
5520 5521
	if (nla_put_u32(skb, IFLA_INET6_FLAGS, idev->if_flags))
		goto nla_put_failure;
T
Thomas Graf 已提交
5522
	ci.max_reasm_len = IPV6_MAXPLEN;
5523 5524
	ci.tstamp = cstamp_delta(idev->tstamp);
	ci.reachable_time = jiffies_to_msecs(idev->nd_parms->reachable_time);
J
Jiri Pirko 已提交
5525
	ci.retrans_time = jiffies_to_msecs(NEIGH_VAR(idev->nd_parms, RETRANS_TIME));
D
David S. Miller 已提交
5526 5527
	if (nla_put(skb, IFLA_INET6_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
T
Thomas Graf 已提交
5528
	nla = nla_reserve(skb, IFLA_INET6_CONF, DEVCONF_MAX * sizeof(s32));
5529
	if (!nla)
T
Thomas Graf 已提交
5530 5531 5532 5533 5534
		goto nla_put_failure;
	ipv6_store_devconf(&idev->cnf, nla_data(nla), nla_len(nla));

	/* XXX - MC not implemented */

5535 5536 5537
	if (ext_filter_mask & RTEXT_FILTER_SKIP_STATS)
		return 0;

T
Thomas Graf 已提交
5538
	nla = nla_reserve(skb, IFLA_INET6_STATS, IPSTATS_MIB_MAX * sizeof(u64));
5539
	if (!nla)
T
Thomas Graf 已提交
5540 5541 5542 5543
		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));
5544
	if (!nla)
T
Thomas Graf 已提交
5545 5546 5547
		goto nla_put_failure;
	snmp6_fill_stats(nla_data(nla), idev, IFLA_INET6_ICMP6STATS, nla_len(nla));

5548
	nla = nla_reserve(skb, IFLA_INET6_TOKEN, sizeof(struct in6_addr));
5549
	if (!nla)
5550
		goto nla_put_failure;
5551

5552
	if (nla_put_u8(skb, IFLA_INET6_ADDR_GEN_MODE, idev->cnf.addr_gen_mode))
5553 5554
		goto nla_put_failure;

5555 5556 5557 5558
	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
5559 5560 5561 5562 5563 5564
	return 0;

nla_put_failure:
	return -EMSGSIZE;
}

5565 5566
static size_t inet6_get_link_af_size(const struct net_device *dev,
				     u32 ext_filter_mask)
T
Thomas Graf 已提交
5567 5568 5569 5570 5571 5572 5573
{
	if (!__in6_dev_get(dev))
		return 0;

	return inet6_ifla6_size();
}

5574 5575
static int inet6_fill_link_af(struct sk_buff *skb, const struct net_device *dev,
			      u32 ext_filter_mask)
T
Thomas Graf 已提交
5576 5577 5578 5579 5580 5581
{
	struct inet6_dev *idev = __in6_dev_get(dev);

	if (!idev)
		return -ENODATA;

5582
	if (inet6_fill_ifla6_attrs(skb, idev, ext_filter_mask) < 0)
T
Thomas Graf 已提交
5583 5584 5585 5586 5587
		return -EMSGSIZE;

	return 0;
}

5588 5589 5590 5591
static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token)
{
	struct inet6_ifaddr *ifp;
	struct net_device *dev = idev->dev;
5592
	bool clear_token, update_rs = false;
5593
	struct in6_addr ll_addr;
5594

5595 5596
	ASSERT_RTNL();

5597
	if (!token)
5598 5599 5600 5601 5602
		return -EINVAL;
	if (dev->flags & (IFF_LOOPBACK | IFF_NOARP))
		return -EINVAL;
	if (!ipv6_accept_ra(idev))
		return -EINVAL;
5603
	if (idev->cnf.rtr_solicits == 0)
5604 5605 5606 5607 5608 5609 5610 5611 5612
		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);

5613 5614 5615 5616
	clear_token = ipv6_addr_any(token);
	if (clear_token)
		goto update_lft;

5617 5618 5619
	if (!idev->dead && (idev->if_flags & IF_READY) &&
	    !ipv6_get_lladdr(dev, &ll_addr, IFA_F_TENTATIVE |
			     IFA_F_OPTIMISTIC)) {
5620 5621 5622 5623 5624 5625
		/* 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;
	}
5626

5627
update_lft:
5628
	write_lock_bh(&idev->lock);
5629

5630
	if (update_rs) {
5631
		idev->if_flags |= IF_RS_SENT;
5632 5633
		idev->rs_interval = rfc3315_s14_backoff_init(
			idev->cnf.rtr_solicit_interval);
5634
		idev->rs_probes = 1;
5635
		addrconf_mod_rs_timer(idev, idev->rs_interval);
5636
	}
5637 5638 5639 5640

	/* Well, that's kinda nasty ... */
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
		spin_lock(&ifp->lock);
5641
		if (ifp->tokenized) {
5642 5643 5644 5645 5646 5647 5648
			ifp->valid_lft = 0;
			ifp->prefered_lft = 0;
		}
		spin_unlock(&ifp->lock);
	}

	write_unlock_bh(&idev->lock);
5649
	inet6_ifinfo_notify(RTM_NEWLINK, idev);
5650
	addrconf_verify_rtnl();
5651 5652 5653
	return 0;
}

5654 5655 5656 5657 5658 5659 5660 5661 5662 5663 5664 5665 5666
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;

5667 5668
	return nla_parse_nested(tb, IFLA_INET6_MAX, nla, inet6_af_policy,
				NULL);
5669 5670
}

5671 5672 5673 5674 5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690
static int check_addr_gen_mode(int mode)
{
	if (mode != IN6_ADDR_GEN_MODE_EUI64 &&
	    mode != IN6_ADDR_GEN_MODE_NONE &&
	    mode != IN6_ADDR_GEN_MODE_STABLE_PRIVACY &&
	    mode != IN6_ADDR_GEN_MODE_RANDOM)
		return -EINVAL;
	return 1;
}

static int check_stable_privacy(struct inet6_dev *idev, struct net *net,
				int mode)
{
	if (mode == IN6_ADDR_GEN_MODE_STABLE_PRIVACY &&
	    !idev->cnf.stable_secret.initialized &&
	    !net->ipv6.devconf_dflt->stable_secret.initialized)
		return -EINVAL;
	return 1;
}

5691 5692 5693 5694 5695 5696 5697 5698 5699
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;

5700
	if (nla_parse_nested(tb, IFLA_INET6_MAX, nla, NULL, NULL) < 0)
5701 5702
		BUG();

5703
	if (tb[IFLA_INET6_TOKEN]) {
5704
		err = inet6_set_iftoken(idev, nla_data(tb[IFLA_INET6_TOKEN]));
5705 5706 5707 5708 5709 5710 5711
		if (err)
			return err;
	}

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

5712 5713
		if (check_addr_gen_mode(mode) < 0 ||
		    check_stable_privacy(idev, dev_net(dev), mode) < 0)
5714
			return -EINVAL;
5715

5716
		idev->cnf.addr_gen_mode = mode;
5717 5718
		err = 0;
	}
5719 5720 5721 5722

	return err;
}

5723
static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
5724
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
5725
{
5726 5727 5728 5729 5730
	struct net_device *dev = idev->dev;
	struct ifinfomsg *hdr;
	struct nlmsghdr *nlh;
	void *protoinfo;

5731
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*hdr), flags);
5732
	if (!nlh)
5733
		return -EMSGSIZE;
5734 5735 5736 5737 5738 5739 5740 5741 5742

	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
5743 5744 5745 5746
	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) ||
5747
	    (dev->ifindex != dev_get_iflink(dev) &&
5748 5749 5750
	     nla_put_u32(skb, IFLA_LINK, dev_get_iflink(dev))) ||
	    nla_put_u8(skb, IFLA_OPERSTATE,
		       netif_running(dev) ? dev->operstate : IF_OPER_DOWN))
D
David S. Miller 已提交
5751
		goto nla_put_failure;
5752
	protoinfo = nla_nest_start(skb, IFLA_PROTINFO);
5753
	if (!protoinfo)
5754
		goto nla_put_failure;
L
Linus Torvalds 已提交
5755

5756
	if (inet6_fill_ifla6_attrs(skb, idev, 0) < 0)
5757
		goto nla_put_failure;
L
Linus Torvalds 已提交
5758

5759
	nla_nest_end(skb, protoinfo);
5760 5761
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
5762

5763
nla_put_failure:
5764 5765
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
5766 5767
}

5768 5769 5770 5771 5772 5773 5774 5775 5776 5777 5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792
static int inet6_valid_dump_ifinfo(const struct nlmsghdr *nlh,
				   struct netlink_ext_ack *extack)
{
	struct ifinfomsg *ifm;

	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*ifm))) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid header for link dump request");
		return -EINVAL;
	}

	if (nlmsg_attrlen(nlh, sizeof(*ifm))) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid data after header");
		return -EINVAL;
	}

	ifm = nlmsg_data(nlh);
	if (ifm->__ifi_pad || ifm->ifi_type || ifm->ifi_flags ||
	    ifm->ifi_change || ifm->ifi_index) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid values in header for dump request");
		return -EINVAL;
	}

	return 0;
}

L
Linus Torvalds 已提交
5793 5794
static int inet6_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
{
5795
	struct net *net = sock_net(skb->sk);
E
Eric Dumazet 已提交
5796
	int h, s_h;
5797
	int idx = 0, s_idx;
L
Linus Torvalds 已提交
5798 5799
	struct net_device *dev;
	struct inet6_dev *idev;
E
Eric Dumazet 已提交
5800
	struct hlist_head *head;
L
Linus Torvalds 已提交
5801

5802 5803 5804 5805 5806 5807 5808 5809 5810 5811
	/* only requests using strict checking can pass data to
	 * influence the dump
	 */
	if (cb->strict_check) {
		int err = inet6_valid_dump_ifinfo(cb->nlh, cb->extack);

		if (err < 0)
			return err;
	}

E
Eric Dumazet 已提交
5812 5813 5814 5815 5816 5817 5818
	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];
5819
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
5820 5821 5822 5823 5824 5825
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;
			if (inet6_fill_ifinfo(skb, idev,
5826
					      NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
5827
					      cb->nlh->nlmsg_seq,
5828
					      RTM_NEWLINK, NLM_F_MULTI) < 0)
E
Eric Dumazet 已提交
5829
				goto out;
5830
cont:
E
Eric Dumazet 已提交
5831 5832
			idx++;
		}
L
Linus Torvalds 已提交
5833
	}
E
Eric Dumazet 已提交
5834 5835 5836 5837
out:
	rcu_read_unlock();
	cb->args[1] = idx;
	cb->args[0] = h;
L
Linus Torvalds 已提交
5838 5839 5840 5841 5842 5843 5844

	return skb->len;
}

void inet6_ifinfo_notify(int event, struct inet6_dev *idev)
{
	struct sk_buff *skb;
5845
	struct net *net = dev_net(idev->dev);
5846
	int err = -ENOBUFS;
5847

5848
	skb = nlmsg_new(inet6_if_nlmsg_size(), GFP_ATOMIC);
5849
	if (!skb)
5850 5851 5852
		goto errout;

	err = inet6_fill_ifinfo(skb, idev, 0, 0, event, 0);
5853 5854 5855 5856 5857 5858
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_if_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5859
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFINFO, NULL, GFP_ATOMIC);
5860
	return;
5861 5862
errout:
	if (err < 0)
5863
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFINFO, err);
L
Linus Torvalds 已提交
5864 5865
}

5866 5867 5868 5869 5870 5871
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));
}
5872

L
Linus Torvalds 已提交
5873
static int inet6_fill_prefix(struct sk_buff *skb, struct inet6_dev *idev,
5874
			     struct prefix_info *pinfo, u32 portid, u32 seq,
5875
			     int event, unsigned int flags)
L
Linus Torvalds 已提交
5876
{
5877 5878
	struct prefixmsg *pmsg;
	struct nlmsghdr *nlh;
L
Linus Torvalds 已提交
5879 5880
	struct prefix_cacheinfo	ci;

5881
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*pmsg), flags);
5882
	if (!nlh)
5883
		return -EMSGSIZE;
5884 5885

	pmsg = nlmsg_data(nlh);
L
Linus Torvalds 已提交
5886
	pmsg->prefix_family = AF_INET6;
5887 5888
	pmsg->prefix_pad1 = 0;
	pmsg->prefix_pad2 = 0;
L
Linus Torvalds 已提交
5889 5890 5891
	pmsg->prefix_ifindex = idev->dev->ifindex;
	pmsg->prefix_len = pinfo->prefix_len;
	pmsg->prefix_type = pinfo->type;
5892
	pmsg->prefix_pad3 = 0;
L
Linus Torvalds 已提交
5893 5894 5895 5896 5897 5898
	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
5899 5900
	if (nla_put(skb, PREFIX_ADDRESS, sizeof(pinfo->prefix), &pinfo->prefix))
		goto nla_put_failure;
L
Linus Torvalds 已提交
5901 5902
	ci.preferred_time = ntohl(pinfo->prefered);
	ci.valid_time = ntohl(pinfo->valid);
D
David S. Miller 已提交
5903 5904
	if (nla_put(skb, PREFIX_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
5905 5906
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
5907

5908
nla_put_failure:
5909 5910
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
5911 5912
}

5913
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
5914 5915 5916
			 struct prefix_info *pinfo)
{
	struct sk_buff *skb;
5917
	struct net *net = dev_net(idev->dev);
5918
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
5919

5920
	skb = nlmsg_new(inet6_prefix_nlmsg_size(), GFP_ATOMIC);
5921
	if (!skb)
5922 5923 5924
		goto errout;

	err = inet6_fill_prefix(skb, idev, pinfo, 0, 0, event, 0);
5925 5926 5927 5928 5929 5930
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_prefix_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5931 5932
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_PREFIX, NULL, GFP_ATOMIC);
	return;
5933 5934
errout:
	if (err < 0)
5935
		rtnl_set_sk_err(net, RTNLGRP_IPV6_PREFIX, err);
L
Linus Torvalds 已提交
5936 5937 5938 5939
}

static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
5940 5941
	struct net *net = dev_net(ifp->idev->dev);

5942 5943 5944
	if (event)
		ASSERT_RTNL();

L
Linus Torvalds 已提交
5945 5946 5947 5948
	inet6_ifa_notify(event ? : RTM_NEWADDR, ifp);

	switch (event) {
	case RTM_NEWADDR:
5949 5950 5951 5952 5953 5954
		/*
		 * 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
		 */
5955
		if (!rcu_access_pointer(ifp->rt->fib6_node))
5956
			ip6_ins_rt(net, ifp->rt);
L
Linus Torvalds 已提交
5957 5958
		if (ifp->idev->cnf.forwarding)
			addrconf_join_anycast(ifp);
C
Cong Wang 已提交
5959
		if (!ipv6_addr_any(&ifp->peer_addr))
5960
			addrconf_prefix_route(&ifp->peer_addr, 128, 0,
5961
					      ifp->idev->dev, 0, 0,
5962
					      GFP_ATOMIC);
L
Linus Torvalds 已提交
5963 5964 5965 5966 5967
		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
5968
		if (!ipv6_addr_any(&ifp->peer_addr)) {
5969
			struct fib6_info *rt;
5970

5971 5972
			rt = addrconf_get_prefix_route(&ifp->peer_addr, 128,
						       ifp->idev->dev, 0, 0);
M
Martin KaFai Lau 已提交
5973
			if (rt)
5974
				ip6_del_rt(net, rt);
5975
		}
5976
		if (ifp->rt) {
5977 5978
			ip6_del_rt(net, ifp->rt);
			ifp->rt = NULL;
5979
		}
H
Hannes Frederic Sowa 已提交
5980
		rt_genid_bump_ipv6(net);
L
Linus Torvalds 已提交
5981 5982
		break;
	}
5983
	atomic_inc(&net->ipv6.dev_addr_genid);
L
Linus Torvalds 已提交
5984 5985 5986 5987
}

static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
5988
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
5989 5990
	if (likely(ifp->idev->dead == 0))
		__ipv6_ifa_notify(event, ifp);
5991
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
5992 5993 5994 5995 5996
}

#ifdef CONFIG_SYSCTL

static
5997
int addrconf_sysctl_forward(struct ctl_table *ctl, int write,
L
Linus Torvalds 已提交
5998 5999 6000 6001
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
6002
	loff_t pos = *ppos;
6003
	struct ctl_table lctl;
L
Linus Torvalds 已提交
6004 6005
	int ret;

6006 6007 6008 6009 6010 6011 6012 6013
	/*
	 * 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
6014

6015
	if (write)
6016
		ret = addrconf_fixup_forwarding(ctl, valp, val);
E
Eric W. Biederman 已提交
6017 6018
	if (ret)
		*ppos = pos;
6019
	return ret;
L
Linus Torvalds 已提交
6020 6021
}

6022 6023 6024 6025 6026 6027 6028 6029 6030 6031 6032 6033 6034 6035 6036
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);
}

6037 6038
static void dev_disable_change(struct inet6_dev *idev)
{
6039 6040
	struct netdev_notifier_info info;

6041 6042 6043
	if (!idev || !idev->dev)
		return;

6044
	netdev_notifier_info_init(&info, idev->dev);
6045
	if (idev->cnf.disable_ipv6)
6046
		addrconf_notify(NULL, NETDEV_DOWN, &info);
6047
	else
6048
		addrconf_notify(NULL, NETDEV_UP, &info);
6049 6050 6051 6052 6053 6054 6055
}

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

6056
	for_each_netdev(net, dev) {
6057 6058 6059 6060 6061 6062 6063 6064 6065 6066
		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);
		}
	}
}

6067
static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
6068 6069
{
	struct net *net;
6070 6071 6072 6073
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
6074 6075

	net = (struct net *)table->extra2;
6076 6077
	old = *p;
	*p = newf;
6078

6079 6080
	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
		rtnl_unlock();
6081
		return 0;
E
Eric W. Biederman 已提交
6082
	}
6083 6084 6085 6086

	if (p == &net->ipv6.devconf_all->disable_ipv6) {
		net->ipv6.devconf_dflt->disable_ipv6 = newf;
		addrconf_disable_change(net, newf);
6087
	} else if ((!newf) ^ (!old))
6088 6089 6090 6091 6092 6093 6094
		dev_disable_change((struct inet6_dev *)table->extra1);

	rtnl_unlock();
	return 0;
}

static
6095
int addrconf_sysctl_disable(struct ctl_table *ctl, int write,
6096 6097 6098 6099
			    void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
6100
	loff_t pos = *ppos;
6101
	struct ctl_table lctl;
6102 6103
	int ret;

6104 6105 6106 6107 6108 6109 6110 6111
	/*
	 * 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);
6112 6113 6114

	if (write)
		ret = addrconf_disable_ipv6(ctl, valp, val);
E
Eric W. Biederman 已提交
6115 6116
	if (ret)
		*ppos = pos;
6117 6118 6119
	return ret;
}

6120 6121 6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138
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)
6139 6140
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
6141 6142 6143
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
		else if (valp == &net->ipv6.devconf_all->proxy_ndp)
6144 6145
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
6146 6147 6148 6149 6150
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
		else {
			struct inet6_dev *idev = ctl->extra1;

6151 6152
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
6153 6154 6155 6156 6157 6158 6159 6160 6161
						     idev->dev->ifindex,
						     &idev->cnf);
		}
		rtnl_unlock();
	}

	return ret;
}

6162 6163 6164 6165 6166
static int addrconf_sysctl_addr_gen_mode(struct ctl_table *ctl, int write,
					 void __user *buffer, size_t *lenp,
					 loff_t *ppos)
{
	int ret = 0;
6167
	u32 new_val;
6168 6169
	struct inet6_dev *idev = (struct inet6_dev *)ctl->extra1;
	struct net *net = (struct net *)ctl->extra2;
6170 6171 6172 6173 6174
	struct ctl_table tmp = {
		.data = &new_val,
		.maxlen = sizeof(new_val),
		.mode = ctl->mode,
	};
6175

6176 6177 6178
	if (!rtnl_trylock())
		return restart_syscall();

6179
	new_val = *((u32 *)ctl->data);
6180

6181 6182 6183
	ret = proc_douintvec(&tmp, write, buffer, lenp, ppos);
	if (ret != 0)
		goto out;
6184

6185
	if (write) {
6186 6187 6188 6189
		if (check_addr_gen_mode(new_val) < 0) {
			ret = -EINVAL;
			goto out;
		}
6190

6191
		if (idev) {
6192 6193 6194 6195
			if (check_stable_privacy(idev, net, new_val) < 0) {
				ret = -EINVAL;
				goto out;
			}
6196 6197 6198 6199 6200

			if (idev->cnf.addr_gen_mode != new_val) {
				idev->cnf.addr_gen_mode = new_val;
				addrconf_dev_config(idev->dev);
			}
6201 6202 6203 6204 6205 6206 6207 6208 6209 6210 6211 6212
		} else if (&net->ipv6.devconf_all->addr_gen_mode == ctl->data) {
			struct net_device *dev;

			net->ipv6.devconf_dflt->addr_gen_mode = new_val;
			for_each_netdev(net, dev) {
				idev = __in6_dev_get(dev);
				if (idev &&
				    idev->cnf.addr_gen_mode != new_val) {
					idev->cnf.addr_gen_mode = new_val;
					addrconf_dev_config(idev->dev);
				}
			}
6213
		}
6214 6215

		*((u32 *)ctl->data) = new_val;
6216 6217
	}

6218 6219 6220
out:
	rtnl_unlock();

6221 6222 6223
	return ret;
}

6224 6225 6226 6227 6228 6229 6230 6231
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;
6232
	struct net *net = ctl->extra2;
6233 6234
	struct ipv6_stable_secret *secret = ctl->data;

6235 6236 6237
	if (&net->ipv6.devconf_all->stable_secret == ctl->data)
		return -EIO;

6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248
	lctl.maxlen = IPV6_MAX_STRLEN;
	lctl.data = str;

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

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

6249 6250 6251 6252
	err = snprintf(str, sizeof(str), "%pI6", &secret->secret);
	if (err >= sizeof(str)) {
		err = -EIO;
		goto out;
6253 6254 6255 6256 6257 6258 6259 6260 6261 6262 6263 6264 6265 6266
	}

	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;

6267 6268 6269 6270 6271 6272 6273
	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) {
6274
				idev->cnf.addr_gen_mode =
6275 6276 6277 6278 6279 6280
					IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
			}
		}
	} else {
		struct inet6_dev *idev = ctl->extra1;

6281
		idev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
6282 6283
	}

6284 6285 6286 6287 6288
out:
	rtnl_unlock();

	return err;
}
6289

6290 6291 6292 6293 6294 6295 6296 6297 6298 6299 6300 6301 6302 6303 6304 6305 6306 6307 6308 6309 6310 6311 6312 6313 6314 6315 6316 6317
static
int addrconf_sysctl_ignore_routes_with_linkdown(struct ctl_table *ctl,
						int write,
						void __user *buffer,
						size_t *lenp,
						loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
	loff_t pos = *ppos;
	struct ctl_table lctl;
	int ret;

	/* ctl->data points to idev->cnf.ignore_routes_when_linkdown
	 * we should not modify it until we get the rtnl lock.
	 */
	lctl = *ctl;
	lctl.data = &val;

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

	if (write)
		ret = addrconf_fixup_linkdown(ctl, valp, val);
	if (ret)
		*ppos = pos;
	return ret;
}

6318 6319 6320 6321 6322 6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334 6335 6336 6337
static
void addrconf_set_nopolicy(struct rt6_info *rt, int action)
{
	if (rt) {
		if (action)
			rt->dst.flags |= DST_NOPOLICY;
		else
			rt->dst.flags &= ~DST_NOPOLICY;
	}
}

static
void addrconf_disable_policy_idev(struct inet6_dev *idev, int val)
{
	struct inet6_ifaddr *ifa;

	read_lock_bh(&idev->lock);
	list_for_each_entry(ifa, &idev->addr_list, if_list) {
		spin_lock(&ifa->lock);
		if (ifa->rt) {
6338
			struct fib6_info *rt = ifa->rt;
6339 6340
			int cpu;

6341
			rcu_read_lock();
6342
			ifa->rt->dst_nopolicy = val ? true : false;
6343 6344 6345 6346 6347 6348 6349 6350
			if (rt->rt6i_pcpu) {
				for_each_possible_cpu(cpu) {
					struct rt6_info **rtp;

					rtp = per_cpu_ptr(rt->rt6i_pcpu, cpu);
					addrconf_set_nopolicy(*rtp, val);
				}
			}
6351
			rcu_read_unlock();
6352 6353 6354 6355 6356 6357 6358 6359 6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373 6374 6375 6376 6377 6378 6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389 6390 6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401 6402 6403 6404 6405 6406 6407 6408 6409 6410 6411 6412 6413 6414 6415
		}
		spin_unlock(&ifa->lock);
	}
	read_unlock_bh(&idev->lock);
}

static
int addrconf_disable_policy(struct ctl_table *ctl, int *valp, int val)
{
	struct inet6_dev *idev;
	struct net *net;

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

	*valp = val;

	net = (struct net *)ctl->extra2;
	if (valp == &net->ipv6.devconf_dflt->disable_policy) {
		rtnl_unlock();
		return 0;
	}

	if (valp == &net->ipv6.devconf_all->disable_policy)  {
		struct net_device *dev;

		for_each_netdev(net, dev) {
			idev = __in6_dev_get(dev);
			if (idev)
				addrconf_disable_policy_idev(idev, val);
		}
	} else {
		idev = (struct inet6_dev *)ctl->extra1;
		addrconf_disable_policy_idev(idev, val);
	}

	rtnl_unlock();
	return 0;
}

static
int addrconf_sysctl_disable_policy(struct ctl_table *ctl, int write,
				   void __user *buffer, size_t *lenp,
				   loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
	loff_t pos = *ppos;
	struct ctl_table lctl;
	int ret;

	lctl = *ctl;
	lctl.data = &val;
	ret = proc_dointvec(&lctl, write, buffer, lenp, ppos);

	if (write && (*valp != val))
		ret = addrconf_disable_policy(ctl, valp, val);

	if (ret)
		*ppos = pos;

	return ret;
}

6416
static int minus_one = -1;
6417
static const int zero = 0;
6418 6419 6420
static const int one = 1;
static const int two_five_five = 255;

6421
static const struct ctl_table addrconf_sysctl[] = {
6422 6423 6424 6425 6426 6427 6428 6429 6430 6431 6432 6433
	{
		.procname	= "forwarding",
		.data		= &ipv6_devconf.forwarding,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= addrconf_sysctl_forward,
	},
	{
		.procname	= "hop_limit",
		.data		= &ipv6_devconf.hop_limit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
6434 6435 6436
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= (void *)&one,
		.extra2		= (void *)&two_five_five,
6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460 6461 6462 6463 6464 6465 6466 6467 6468 6469 6470 6471 6472 6473 6474 6475 6476 6477
	},
	{
		.procname	= "mtu",
		.data		= &ipv6_devconf.mtu6,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= addrconf_sysctl_mtu,
	},
	{
		.procname	= "accept_ra",
		.data		= &ipv6_devconf.accept_ra,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "accept_redirects",
		.data		= &ipv6_devconf.accept_redirects,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "autoconf",
		.data		= &ipv6_devconf.autoconf,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "dad_transmits",
		.data		= &ipv6_devconf.dad_transmits,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "router_solicitations",
		.data		= &ipv6_devconf.rtr_solicits,
		.maxlen		= sizeof(int),
		.mode		= 0644,
6478 6479
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &minus_one,
6480 6481 6482 6483 6484 6485 6486 6487
	},
	{
		.procname	= "router_solicitation_interval",
		.data		= &ipv6_devconf.rtr_solicit_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
6488 6489 6490 6491 6492 6493 6494
	{
		.procname	= "router_solicitation_max_interval",
		.data		= &ipv6_devconf.rtr_solicit_max_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
6495 6496 6497 6498 6499 6500 6501 6502 6503 6504 6505 6506 6507 6508 6509 6510 6511 6512 6513 6514 6515 6516 6517 6518 6519 6520 6521 6522 6523 6524 6525 6526 6527 6528 6529 6530 6531 6532 6533 6534 6535 6536 6537 6538 6539 6540 6541 6542 6543 6544 6545 6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556 6557 6558 6559 6560 6561 6562 6563 6564 6565 6566 6567 6568 6569 6570 6571 6572 6573 6574 6575 6576 6577 6578 6579 6580 6581 6582 6583 6584 6585 6586 6587
	{
		.procname	= "router_solicitation_delay",
		.data		= &ipv6_devconf.rtr_solicit_delay,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
	{
		.procname	= "force_mld_version",
		.data		= &ipv6_devconf.force_mld_version,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.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,
	},
	{
		.procname	= "use_tempaddr",
		.data		= &ipv6_devconf.use_tempaddr,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "temp_valid_lft",
		.data		= &ipv6_devconf.temp_valid_lft,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "temp_prefered_lft",
		.data		= &ipv6_devconf.temp_prefered_lft,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "regen_max_retry",
		.data		= &ipv6_devconf.regen_max_retry,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "max_desync_factor",
		.data		= &ipv6_devconf.max_desync_factor,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "max_addresses",
		.data		= &ipv6_devconf.max_addresses,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "accept_ra_defrtr",
		.data		= &ipv6_devconf.accept_ra_defrtr,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "accept_ra_min_hop_limit",
		.data		= &ipv6_devconf.accept_ra_min_hop_limit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "accept_ra_pinfo",
		.data		= &ipv6_devconf.accept_ra_pinfo,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
6588
#ifdef CONFIG_IPV6_ROUTER_PREF
6589 6590 6591 6592 6593 6594 6595 6596 6597 6598 6599 6600 6601 6602
	{
		.procname	= "accept_ra_rtr_pref",
		.data		= &ipv6_devconf.accept_ra_rtr_pref,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "router_probe_interval",
		.data		= &ipv6_devconf.rtr_probe_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
N
Neil Horman 已提交
6603
#ifdef CONFIG_IPV6_ROUTE_INFO
6604 6605 6606 6607 6608 6609 6610
	{
		.procname	= "accept_ra_rt_info_min_plen",
		.data		= &ipv6_devconf.accept_ra_rt_info_min_plen,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
6611 6612 6613 6614 6615 6616 6617
	{
		.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,
	},
6618
#endif
6619
#endif
6620 6621 6622 6623 6624 6625 6626 6627 6628 6629 6630 6631 6632 6633
	{
		.procname	= "proxy_ndp",
		.data		= &ipv6_devconf.proxy_ndp,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= addrconf_sysctl_proxy_ndp,
	},
	{
		.procname	= "accept_source_route",
		.data		= &ipv6_devconf.accept_source_route,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
6634
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
6635 6636 6637 6638 6639 6640 6641 6642 6643 6644 6645 6646 6647 6648
	{
		.procname	= "optimistic_dad",
		.data		= &ipv6_devconf.optimistic_dad,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler   = proc_dointvec,
	},
	{
		.procname	= "use_optimistic",
		.data		= &ipv6_devconf.use_optimistic,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
6649 6650
#endif
#ifdef CONFIG_IPV6_MROUTE
6651 6652 6653 6654 6655 6656 6657
	{
		.procname	= "mc_forwarding",
		.data		= &ipv6_devconf.mc_forwarding,
		.maxlen		= sizeof(int),
		.mode		= 0444,
		.proc_handler	= proc_dointvec,
	},
6658
#endif
6659 6660 6661 6662 6663 6664 6665 6666 6667 6668 6669 6670 6671 6672 6673 6674 6675 6676 6677 6678 6679 6680 6681 6682 6683 6684 6685 6686 6687 6688 6689 6690 6691 6692 6693 6694 6695 6696 6697 6698 6699 6700 6701 6702 6703 6704 6705 6706 6707 6708 6709 6710 6711 6712 6713 6714 6715 6716 6717 6718 6719 6720 6721 6722 6723 6724 6725 6726 6727 6728 6729 6730 6731 6732 6733 6734 6735 6736 6737 6738 6739 6740 6741 6742 6743 6744 6745 6746 6747 6748 6749 6750
	{
		.procname	= "disable_ipv6",
		.data		= &ipv6_devconf.disable_ipv6,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= addrconf_sysctl_disable,
	},
	{
		.procname	= "accept_dad",
		.data		= &ipv6_devconf.accept_dad,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "force_tllao",
		.data		= &ipv6_devconf.force_tllao,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "ndisc_notify",
		.data		= &ipv6_devconf.ndisc_notify,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "suppress_frag_ndisc",
		.data		= &ipv6_devconf.suppress_frag_ndisc,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "accept_ra_from_local",
		.data		= &ipv6_devconf.accept_ra_from_local,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "accept_ra_mtu",
		.data		= &ipv6_devconf.accept_ra_mtu,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "stable_secret",
		.data		= &ipv6_devconf.stable_secret,
		.maxlen		= IPV6_MAX_STRLEN,
		.mode		= 0600,
		.proc_handler	= addrconf_sysctl_stable_secret,
	},
	{
		.procname	= "use_oif_addrs_only",
		.data		= &ipv6_devconf.use_oif_addrs_only,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "ignore_routes_with_linkdown",
		.data		= &ipv6_devconf.ignore_routes_with_linkdown,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= addrconf_sysctl_ignore_routes_with_linkdown,
	},
	{
		.procname	= "drop_unicast_in_l2_multicast",
		.data		= &ipv6_devconf.drop_unicast_in_l2_multicast,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "drop_unsolicited_na",
		.data		= &ipv6_devconf.drop_unsolicited_na,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "keep_addr_on_down",
		.data		= &ipv6_devconf.keep_addr_on_down,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,

	},
6751 6752 6753 6754 6755 6756 6757
	{
		.procname	= "seg6_enabled",
		.data		= &ipv6_devconf.seg6_enabled,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
6758 6759 6760 6761 6762 6763 6764 6765 6766
#ifdef CONFIG_IPV6_SEG6_HMAC
	{
		.procname	= "seg6_require_hmac",
		.data		= &ipv6_devconf.seg6_require_hmac,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
#endif
6767 6768 6769 6770 6771 6772 6773
	{
		.procname       = "enhanced_dad",
		.data           = &ipv6_devconf.enhanced_dad,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = proc_dointvec,
	},
6774 6775 6776 6777 6778 6779 6780
	{
		.procname		= "addr_gen_mode",
		.data			= &ipv6_devconf.addr_gen_mode,
		.maxlen			= sizeof(int),
		.mode			= 0644,
		.proc_handler	= addrconf_sysctl_addr_gen_mode,
	},
6781 6782 6783 6784 6785 6786 6787
	{
		.procname       = "disable_policy",
		.data           = &ipv6_devconf.disable_policy,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = addrconf_sysctl_disable_policy,
	},
6788 6789 6790 6791 6792 6793 6794 6795 6796
	{
		.procname	= "ndisc_tclass",
		.data		= &ipv6_devconf.ndisc_tclass,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= (void *)&zero,
		.extra2		= (void *)&two_five_five,
	},
6797 6798 6799
	{
		/* sentinel */
	}
L
Linus Torvalds 已提交
6800 6801
};

6802
static int __addrconf_sysctl_register(struct net *net, char *dev_name,
6803
		struct inet6_dev *idev, struct ipv6_devconf *p)
L
Linus Torvalds 已提交
6804
{
6805
	int i, ifindex;
6806
	struct ctl_table *table;
6807
	char path[sizeof("net/ipv6/conf/") + IFNAMSIZ];
6808

6809 6810
	table = kmemdup(addrconf_sysctl, sizeof(addrconf_sysctl), GFP_KERNEL);
	if (!table)
6811 6812
		goto out;

6813 6814
	for (i = 0; table[i].data; i++) {
		table[i].data += (char *)p - (char *)&ipv6_devconf;
6815 6816 6817 6818 6819 6820 6821 6822
		/* If one of these is already set, then it is not safe to
		 * overwrite either of them: this makes proc_dointvec_minmax
		 * usable.
		 */
		if (!table[i].extra1 && !table[i].extra2) {
			table[i].extra1 = idev; /* embedded; no ref */
			table[i].extra2 = net;
		}
L
Linus Torvalds 已提交
6823 6824
	}

6825
	snprintf(path, sizeof(path), "net/ipv6/conf/%s", dev_name);
L
Linus Torvalds 已提交
6826

6827 6828
	p->sysctl_header = register_net_sysctl(net, path, table);
	if (!p->sysctl_header)
6829
		goto free;
6830

6831 6832 6833 6834 6835 6836
	if (!strcmp(dev_name, "all"))
		ifindex = NETCONFA_IFINDEX_ALL;
	else if (!strcmp(dev_name, "default"))
		ifindex = NETCONFA_IFINDEX_DEFAULT;
	else
		ifindex = idev->dev->ifindex;
6837 6838
	inet6_netconf_notify_devconf(net, RTM_NEWNETCONF, NETCONFA_ALL,
				     ifindex, p);
6839
	return 0;
L
Linus Torvalds 已提交
6840

6841
free:
6842
	kfree(table);
6843
out:
6844
	return -ENOBUFS;
L
Linus Torvalds 已提交
6845 6846
}

6847 6848
static void __addrconf_sysctl_unregister(struct net *net,
					 struct ipv6_devconf *p, int ifindex)
6849
{
6850
	struct ctl_table *table;
6851

6852
	if (!p->sysctl_header)
6853 6854
		return;

6855 6856 6857 6858
	table = p->sysctl_header->ctl_table_arg;
	unregister_net_sysctl_table(p->sysctl_header);
	p->sysctl_header = NULL;
	kfree(table);
6859 6860

	inet6_netconf_notify_devconf(net, RTM_DELNETCONF, 0, ifindex, NULL);
6861 6862
}

6863
static int addrconf_sysctl_register(struct inet6_dev *idev)
6864
{
6865 6866 6867 6868 6869 6870 6871 6872 6873 6874 6875 6876 6877 6878 6879
	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;
6880 6881
}

6882
static void addrconf_sysctl_unregister(struct inet6_dev *idev)
L
Linus Torvalds 已提交
6883
{
6884 6885
	__addrconf_sysctl_unregister(dev_net(idev->dev), &idev->cnf,
				     idev->dev->ifindex);
6886
	neigh_sysctl_unregister(idev->nd_parms);
L
Linus Torvalds 已提交
6887 6888 6889 6890 6891
}


#endif

6892
static int __net_init addrconf_init_net(struct net *net)
6893
{
6894
	int err = -ENOMEM;
6895 6896
	struct ipv6_devconf *all, *dflt;

6897
	all = kmemdup(&ipv6_devconf, sizeof(ipv6_devconf), GFP_KERNEL);
6898
	if (!all)
6899
		goto err_alloc_all;
6900

6901
	dflt = kmemdup(&ipv6_devconf_dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
6902
	if (!dflt)
6903
		goto err_alloc_dflt;
6904

6905 6906 6907
	/* these will be inherited by all namespaces */
	dflt->autoconf = ipv6_defaults.autoconf;
	dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
6908

6909 6910 6911
	dflt->stable_secret.initialized = false;
	all->stable_secret.initialized = false;

6912 6913 6914 6915
	net->ipv6.devconf_all = all;
	net->ipv6.devconf_dflt = dflt;

#ifdef CONFIG_SYSCTL
6916
	err = __addrconf_sysctl_register(net, "all", NULL, all);
6917 6918 6919
	if (err < 0)
		goto err_reg_all;

6920
	err = __addrconf_sysctl_register(net, "default", NULL, dflt);
6921 6922 6923 6924 6925 6926 6927
	if (err < 0)
		goto err_reg_dflt;
#endif
	return 0;

#ifdef CONFIG_SYSCTL
err_reg_dflt:
6928
	__addrconf_sysctl_unregister(net, all, NETCONFA_IFINDEX_ALL);
6929 6930 6931 6932 6933 6934 6935 6936 6937
err_reg_all:
	kfree(dflt);
#endif
err_alloc_dflt:
	kfree(all);
err_alloc_all:
	return err;
}

6938
static void __net_exit addrconf_exit_net(struct net *net)
6939 6940
{
#ifdef CONFIG_SYSCTL
6941 6942 6943 6944
	__addrconf_sysctl_unregister(net, net->ipv6.devconf_dflt,
				     NETCONFA_IFINDEX_DEFAULT);
	__addrconf_sysctl_unregister(net, net->ipv6.devconf_all,
				     NETCONFA_IFINDEX_ALL);
6945
#endif
Z
zhuyj 已提交
6946 6947
	kfree(net->ipv6.devconf_dflt);
	kfree(net->ipv6.devconf_all);
6948 6949 6950 6951 6952 6953 6954
}

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

6955
static struct rtnl_af_ops inet6_ops __read_mostly = {
T
Thomas Graf 已提交
6956 6957 6958
	.family		  = AF_INET6,
	.fill_link_af	  = inet6_fill_link_af,
	.get_link_af_size = inet6_get_link_af_size,
6959
	.validate_link_af = inet6_validate_link_af,
6960
	.set_link_af	  = inet6_set_link_af,
T
Thomas Graf 已提交
6961 6962
};

L
Linus Torvalds 已提交
6963 6964 6965 6966 6967 6968
/*
 *	Init / cleanup code
 */

int __init addrconf_init(void)
{
6969
	struct inet6_dev *idev;
6970
	int i, err;
6971

S
Stephen Hemminger 已提交
6972 6973
	err = ipv6_addr_label_init();
	if (err < 0) {
6974 6975
		pr_crit("%s: cannot initialize default policy table: %d\n",
			__func__, err);
6976
		goto out;
6977
	}
L
Linus Torvalds 已提交
6978

6979 6980 6981
	err = register_pernet_subsys(&addrconf_ops);
	if (err < 0)
		goto out_addrlabel;
6982

6983 6984 6985 6986 6987 6988
	addrconf_wq = create_workqueue("ipv6_addrconf");
	if (!addrconf_wq) {
		err = -ENOMEM;
		goto out_nowq;
	}

L
Linus Torvalds 已提交
6989 6990 6991 6992 6993 6994 6995 6996 6997 6998 6999 7000 7001 7002 7003 7004 7005 7006 7007
	/* 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();
7008
	idev = ipv6_add_dev(init_net.loopback_dev);
L
Linus Torvalds 已提交
7009
	rtnl_unlock();
7010 7011
	if (IS_ERR(idev)) {
		err = PTR_ERR(idev);
7012
		goto errlo;
7013
	}
L
Linus Torvalds 已提交
7014

7015 7016
	ip6_route_init_special_entries();

7017 7018 7019
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		INIT_HLIST_HEAD(&inet6_addr_lst[i]);

L
Linus Torvalds 已提交
7020 7021
	register_netdevice_notifier(&ipv6_dev_notf);

7022
	addrconf_verify();
7023

7024
	rtnl_af_register(&inet6_ops);
T
Thomas Graf 已提交
7025

7026 7027
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETLINK,
				   NULL, inet6_dump_ifinfo, 0);
7028 7029 7030
	if (err < 0)
		goto errout;

7031 7032 7033 7034 7035 7036 7037 7038 7039 7040 7041 7042 7043 7044 7045 7046 7047 7048 7049 7050 7051 7052 7053 7054 7055 7056 7057
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_NEWADDR,
				   inet6_rtm_newaddr, NULL, 0);
	if (err < 0)
		goto errout;
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_DELADDR,
				   inet6_rtm_deladdr, NULL, 0);
	if (err < 0)
		goto errout;
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETADDR,
				   inet6_rtm_getaddr, inet6_dump_ifaddr,
				   RTNL_FLAG_DOIT_UNLOCKED);
	if (err < 0)
		goto errout;
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETMULTICAST,
				   NULL, inet6_dump_ifmcaddr, 0);
	if (err < 0)
		goto errout;
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETANYCAST,
				   NULL, inet6_dump_ifacaddr, 0);
	if (err < 0)
		goto errout;
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETNETCONF,
				   inet6_netconf_get_devconf,
				   inet6_netconf_dump_devconf,
				   RTNL_FLAG_DOIT_UNLOCKED);
	if (err < 0)
		goto errout;
7058 7059 7060
	err = ipv6_addr_label_rtnl_register();
	if (err < 0)
		goto errout;
7061

L
Linus Torvalds 已提交
7062
	return 0;
7063
errout:
7064
	rtnl_unregister_all(PF_INET6);
T
Thomas Graf 已提交
7065
	rtnl_af_unregister(&inet6_ops);
7066
	unregister_netdevice_notifier(&ipv6_dev_notf);
7067
errlo:
7068 7069
	destroy_workqueue(addrconf_wq);
out_nowq:
7070
	unregister_pernet_subsys(&addrconf_ops);
7071 7072 7073
out_addrlabel:
	ipv6_addr_label_cleanup();
out:
7074
	return err;
L
Linus Torvalds 已提交
7075 7076
}

7077
void addrconf_cleanup(void)
L
Linus Torvalds 已提交
7078
{
7079
	struct net_device *dev;
L
Linus Torvalds 已提交
7080 7081 7082
	int i;

	unregister_netdevice_notifier(&ipv6_dev_notf);
7083
	unregister_pernet_subsys(&addrconf_ops);
7084
	ipv6_addr_label_cleanup();
L
Linus Torvalds 已提交
7085

7086
	rtnl_af_unregister(&inet6_ops);
L
Linus Torvalds 已提交
7087

7088
	rtnl_lock();
T
Thomas Graf 已提交
7089

7090 7091 7092 7093 7094 7095 7096 7097
	/* 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
7098 7099 7100
	/*
	 *	Check hash table.
	 */
7101
	spin_lock_bh(&addrconf_hash_lock);
7102 7103
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		WARN_ON(!hlist_empty(&inet6_addr_lst[i]));
7104
	spin_unlock_bh(&addrconf_hash_lock);
7105
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
7106
	rtnl_unlock();
7107 7108

	destroy_workqueue(addrconf_wq);
L
Linus Torvalds 已提交
7109
}