addrconf.c 171.8 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
						  int plen,
						  const struct net_device *dev,
176 177
						  u32 flags, u32 noflags,
						  bool no_gw);
178

179
static void addrconf_dad_start(struct inet6_ifaddr *ifp);
180
static void addrconf_dad_work(struct work_struct *w);
181 182
static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id,
				   bool send_na);
183
static void addrconf_dad_run(struct inet6_dev *idev, bool restart);
184
static void addrconf_rs_timer(struct timer_list *t);
L
Linus Torvalds 已提交
185 186 187
static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);
static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);

188
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
189 190
				struct prefix_info *pinfo);

191
static struct ipv6_devconf ipv6_devconf __read_mostly = {
L
Linus Torvalds 已提交
192 193 194 195 196 197 198
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
	.force_mld_version	= 0,
199 200
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
201 202 203
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
204
	.rtr_solicit_max_interval = RTR_SOLICITATION_MAX_INTERVAL,
L
Linus Torvalds 已提交
205
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
206
	.use_tempaddr		= 0,
L
Linus Torvalds 已提交
207 208 209 210 211
	.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,
212
	.accept_ra_defrtr	= 1,
213
	.accept_ra_from_local	= 0,
214
	.accept_ra_min_hop_limit= 1,
215
	.accept_ra_pinfo	= 1,
216 217
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
218
	.rtr_probe_interval	= 60 * HZ,
219
#ifdef CONFIG_IPV6_ROUTE_INFO
220
	.accept_ra_rt_info_min_plen = 0,
221 222
	.accept_ra_rt_info_max_plen = 0,
#endif
223
#endif
224
	.proxy_ndp		= 0,
225
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
226
	.disable_ipv6		= 0,
227
	.accept_dad		= 0,
228
	.suppress_frag_ndisc	= 1,
229
	.accept_ra_mtu		= 1,
230 231
	.stable_secret		= {
		.initialized = false,
232 233
	},
	.use_oif_addrs_only	= 0,
234
	.ignore_routes_with_linkdown = 0,
235
	.keep_addr_on_down	= 0,
236
	.seg6_enabled		= 0,
237 238 239
#ifdef CONFIG_IPV6_SEG6_HMAC
	.seg6_require_hmac	= 0,
#endif
240
	.enhanced_dad           = 1,
241
	.addr_gen_mode		= IN6_ADDR_GEN_MODE_EUI64,
242
	.disable_policy		= 0,
L
Linus Torvalds 已提交
243 244
};

245
static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
L
Linus Torvalds 已提交
246 247 248 249 250 251
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
252 253 254
	.force_mld_version	= 0,
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
255 256 257
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
258
	.rtr_solicit_max_interval = RTR_SOLICITATION_MAX_INTERVAL,
L
Linus Torvalds 已提交
259 260 261 262 263 264 265
	.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,
266
	.accept_ra_defrtr	= 1,
267
	.accept_ra_from_local	= 0,
268
	.accept_ra_min_hop_limit= 1,
269
	.accept_ra_pinfo	= 1,
270 271
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
272
	.rtr_probe_interval	= 60 * HZ,
273
#ifdef CONFIG_IPV6_ROUTE_INFO
274
	.accept_ra_rt_info_min_plen = 0,
275 276
	.accept_ra_rt_info_max_plen = 0,
#endif
277
#endif
278
	.proxy_ndp		= 0,
279
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
280
	.disable_ipv6		= 0,
281
	.accept_dad		= 1,
282
	.suppress_frag_ndisc	= 1,
283
	.accept_ra_mtu		= 1,
284 285 286
	.stable_secret		= {
		.initialized = false,
	},
287
	.use_oif_addrs_only	= 0,
288
	.ignore_routes_with_linkdown = 0,
289
	.keep_addr_on_down	= 0,
290
	.seg6_enabled		= 0,
291 292 293
#ifdef CONFIG_IPV6_SEG6_HMAC
	.seg6_require_hmac	= 0,
#endif
294
	.enhanced_dad           = 1,
295
	.addr_gen_mode		= IN6_ADDR_GEN_MODE_EUI64,
296
	.disable_policy		= 0,
L
Linus Torvalds 已提交
297 298
};

299 300
/* 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)
301
{
302
	return netif_oper_up(dev) && !qdisc_tx_is_noop(dev);
303 304
}

305
static void addrconf_del_rs_timer(struct inet6_dev *idev)
L
Linus Torvalds 已提交
306
{
307 308 309 310
	if (del_timer(&idev->rs_timer))
		__in6_dev_put(idev);
}

311
static void addrconf_del_dad_work(struct inet6_ifaddr *ifp)
312
{
313
	if (cancel_delayed_work(&ifp->dad_work))
L
Linus Torvalds 已提交
314 315 316
		__in6_ifa_put(ifp);
}

317 318 319 320 321 322 323
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
324

325 326
static void addrconf_mod_dad_work(struct inet6_ifaddr *ifp,
				   unsigned long delay)
L
Linus Torvalds 已提交
327
{
328 329 330
	in6_ifa_hold(ifp);
	if (mod_delayed_work(addrconf_wq, &ifp->dad_work, delay))
		in6_ifa_put(ifp);
L
Linus Torvalds 已提交
331 332
}

333 334
static int snmp6_alloc_dev(struct inet6_dev *idev)
{
335 336
	int i;

W
WANG Cong 已提交
337 338
	idev->stats.ipv6 = alloc_percpu(struct ipstats_mib);
	if (!idev->stats.ipv6)
339
		goto err_ip;
340 341 342

	for_each_possible_cpu(i) {
		struct ipstats_mib *addrconf_stats;
W
WANG Cong 已提交
343
		addrconf_stats = per_cpu_ptr(idev->stats.ipv6, i);
344 345 346 347
		u64_stats_init(&addrconf_stats->syncp);
	}


348 349 350
	idev->stats.icmpv6dev = kzalloc(sizeof(struct icmpv6_mib_device),
					GFP_KERNEL);
	if (!idev->stats.icmpv6dev)
351
		goto err_icmp;
352 353 354
	idev->stats.icmpv6msgdev = kzalloc(sizeof(struct icmpv6msg_mib_device),
					   GFP_KERNEL);
	if (!idev->stats.icmpv6msgdev)
355
		goto err_icmpmsg;
356 357 358

	return 0;

359
err_icmpmsg:
360
	kfree(idev->stats.icmpv6dev);
361
err_icmp:
W
WANG Cong 已提交
362
	free_percpu(idev->stats.ipv6);
363
err_ip:
364
	return -ENOMEM;
365 366
}

367
static struct inet6_dev *ipv6_add_dev(struct net_device *dev)
L
Linus Torvalds 已提交
368 369
{
	struct inet6_dev *ndev;
370
	int err = -ENOMEM;
L
Linus Torvalds 已提交
371 372 373 374

	ASSERT_RTNL();

	if (dev->mtu < IPV6_MIN_MTU)
375
		return ERR_PTR(-EINVAL);
L
Linus Torvalds 已提交
376

377
	ndev = kzalloc(sizeof(struct inet6_dev), GFP_KERNEL);
378
	if (!ndev)
379
		return ERR_PTR(err);
380 381 382

	rwlock_init(&ndev->lock);
	ndev->dev = dev;
383
	INIT_LIST_HEAD(&ndev->addr_list);
384
	timer_setup(&ndev->rs_timer, addrconf_rs_timer, 0);
385
	memcpy(&ndev->cnf, dev_net(dev)->ipv6.devconf_dflt, sizeof(ndev->cnf));
386 387

	if (ndev->cnf.stable_secret.initialized)
388
		ndev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
389

390 391
	ndev->cnf.mtu6 = dev->mtu;
	ndev->nd_parms = neigh_parms_alloc(dev, &nd_tbl);
392
	if (!ndev->nd_parms) {
393
		kfree(ndev);
394
		return ERR_PTR(err);
395
	}
396 397
	if (ndev->cnf.forwarding)
		dev_disable_lro(dev);
398 399
	/* We refer to the device */
	dev_hold(dev);
L
Linus Torvalds 已提交
400

401
	if (snmp6_alloc_dev(ndev) < 0) {
402 403
		netdev_dbg(dev, "%s: cannot allocate memory for statistics\n",
			   __func__);
404
		neigh_parms_release(&nd_tbl, ndev->nd_parms);
R
Roy Li 已提交
405 406
		dev_put(dev);
		kfree(ndev);
407
		return ERR_PTR(err);
408
	}
L
Linus Torvalds 已提交
409

410
	if (snmp6_register_dev(ndev) < 0) {
411 412
		netdev_dbg(dev, "%s: cannot create /proc/net/dev_snmp6/%s\n",
			   __func__, dev->name);
413
		goto err_release;
414 415
	}

416
	/* One reference from device. */
417
	refcount_set(&ndev->refcnt, 1);
L
Linus Torvalds 已提交
418

419 420 421
	if (dev->flags & (IFF_NOARP | IFF_LOOPBACK))
		ndev->cnf.accept_dad = -1;

A
Amerigo Wang 已提交
422
#if IS_ENABLED(CONFIG_IPV6_SIT)
423
	if (dev->type == ARPHRD_SIT && (dev->priv_flags & IFF_ISATAP)) {
424
		pr_info("%s: Disabled Multicast RS\n", dev->name);
425 426 427 428
		ndev->cnf.rtr_solicits = 0;
	}
#endif

429
	INIT_LIST_HEAD(&ndev->tempaddr_list);
J
Jiri Bohac 已提交
430
	ndev->desync_factor = U32_MAX;
431 432
	if ((dev->flags&IFF_LOOPBACK) ||
	    dev->type == ARPHRD_TUNNEL ||
433
	    dev->type == ARPHRD_TUNNEL6 ||
434 435
	    dev->type == ARPHRD_SIT ||
	    dev->type == ARPHRD_NONE) {
436
		ndev->cnf.use_tempaddr = -1;
437 438
	} else
		ipv6_regen_rndid(ndev);
439

440
	ndev->token = in6addr_any;
L
Linus Torvalds 已提交
441

442
	if (netif_running(dev) && addrconf_link_ready(dev))
443 444
		ndev->if_flags |= IF_READY;

445 446
	ipv6_mc_init_dev(ndev);
	ndev->tstamp = jiffies;
447 448 449
	err = addrconf_sysctl_register(ndev);
	if (err) {
		ipv6_mc_destroy_dev(ndev);
450
		snmp6_unregister_dev(ndev);
451 452
		goto err_release;
	}
453
	/* protected by rtnl_lock */
454
	rcu_assign_pointer(dev->ip6_ptr, ndev);
455

456 457 458
	/* Join interface-local all-node multicast group */
	ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allnodes);

459
	/* Join all-node multicast group */
460
	ipv6_dev_mc_inc(dev, &in6addr_linklocal_allnodes);
461

462
	/* Join all-router multicast group if forwarding is set */
L
Li Wei 已提交
463
	if (ndev->cnf.forwarding && (dev->flags & IFF_MULTICAST))
464 465
		ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);

L
Linus Torvalds 已提交
466
	return ndev;
467 468 469 470 471 472

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
473 474
}

475
static struct inet6_dev *ipv6_find_idev(struct net_device *dev)
L
Linus Torvalds 已提交
476 477 478 479 480
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
481 482 483
	idev = __in6_dev_get(dev);
	if (!idev) {
		idev = ipv6_add_dev(dev);
484
		if (IS_ERR(idev))
L
Linus Torvalds 已提交
485 486
			return NULL;
	}
487

L
Linus Torvalds 已提交
488 489 490 491 492
	if (dev->flags&IFF_UP)
		ipv6_mc_up(idev);
	return idev;
}

493 494 495 496
static int inet6_netconf_msgsize_devconf(int type)
{
	int size =  NLMSG_ALIGN(sizeof(struct netconfmsg))
		    + nla_total_size(4);	/* NETCONFA_IFINDEX */
497
	bool all = false;
498

499 500 501 502
	if (type == NETCONFA_ALL)
		all = true;

	if (all || type == NETCONFA_FORWARDING)
503
		size += nla_total_size(4);
504
#ifdef CONFIG_IPV6_MROUTE
505
	if (all || type == NETCONFA_MC_FORWARDING)
506
		size += nla_total_size(4);
507
#endif
508
	if (all || type == NETCONFA_PROXY_NEIGH)
509
		size += nla_total_size(4);
510

511
	if (all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN)
512 513
		size += nla_total_size(4);

514 515 516 517 518 519 520 521 522 523
	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;
524
	bool all = false;
525 526 527

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct netconfmsg),
			flags);
528
	if (!nlh)
529 530
		return -EMSGSIZE;

531 532 533
	if (type == NETCONFA_ALL)
		all = true;

534 535 536 537 538 539
	ncm = nlmsg_data(nlh);
	ncm->ncm_family = AF_INET6;

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

540 541 542
	if (!devconf)
		goto out;

543
	if ((all || type == NETCONFA_FORWARDING) &&
544 545
	    nla_put_s32(skb, NETCONFA_FORWARDING, devconf->forwarding) < 0)
		goto nla_put_failure;
546
#ifdef CONFIG_IPV6_MROUTE
547
	if ((all || type == NETCONFA_MC_FORWARDING) &&
548 549 550
	    nla_put_s32(skb, NETCONFA_MC_FORWARDING,
			devconf->mc_forwarding) < 0)
		goto nla_put_failure;
551
#endif
552
	if ((all || type == NETCONFA_PROXY_NEIGH) &&
553 554 555
	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
		goto nla_put_failure;

556
	if ((all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN) &&
557 558 559 560
	    nla_put_s32(skb, NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
			devconf->ignore_routes_with_linkdown) < 0)
		goto nla_put_failure;

561
out:
562 563
	nlmsg_end(skb, nlh);
	return 0;
564 565 566 567 568 569

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

570 571
void inet6_netconf_notify_devconf(struct net *net, int event, int type,
				  int ifindex, struct ipv6_devconf *devconf)
572 573 574 575
{
	struct sk_buff *skb;
	int err = -ENOBUFS;

576
	skb = nlmsg_new(inet6_netconf_msgsize_devconf(type), GFP_KERNEL);
577
	if (!skb)
578 579 580
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf, 0, 0,
581
					 event, 0, type);
582 583 584 585 586 587
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
588
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_NETCONF, NULL, GFP_KERNEL);
589 590
	return;
errout:
591
	rtnl_set_sk_err(net, RTNLGRP_IPV6_NETCONF, err);
592 593
}

594 595 596
static const struct nla_policy devconf_ipv6_policy[NETCONFA_MAX+1] = {
	[NETCONFA_IFINDEX]	= { .len = sizeof(int) },
	[NETCONFA_FORWARDING]	= { .len = sizeof(int) },
597
	[NETCONFA_PROXY_NEIGH]	= { .len = sizeof(int) },
598
	[NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN]	= { .len = sizeof(int) },
599 600
};

601 602 603 604 605 606 607 608 609 610 611 612 613
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))
614 615 616
		return nlmsg_parse_deprecated(nlh, sizeof(struct netconfmsg),
					      tb, NETCONFA_MAX,
					      devconf_ipv6_policy, extack);
617

618 619 620
	err = nlmsg_parse_deprecated_strict(nlh, sizeof(struct netconfmsg),
					    tb, NETCONFA_MAX,
					    devconf_ipv6_policy, extack);
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639
	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;
}

640
static int inet6_netconf_get_devconf(struct sk_buff *in_skb,
641 642
				     struct nlmsghdr *nlh,
				     struct netlink_ext_ack *extack)
643 644 645
{
	struct net *net = sock_net(in_skb->sk);
	struct nlattr *tb[NETCONFA_MAX+1];
646 647
	struct inet6_dev *in6_dev = NULL;
	struct net_device *dev = NULL;
648 649 650 651 652
	struct sk_buff *skb;
	struct ipv6_devconf *devconf;
	int ifindex;
	int err;

653
	err = inet6_netconf_valid_get_req(in_skb, nlh, tb, extack);
654
	if (err < 0)
655
		return err;
656 657

	if (!tb[NETCONFA_IFINDEX])
658
		return -EINVAL;
659

660
	err = -EINVAL;
661 662 663 664 665 666 667 668 669
	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:
670
		dev = dev_get_by_index(net, ifindex);
671
		if (!dev)
672 673
			return -EINVAL;
		in6_dev = in6_dev_get(dev);
674
		if (!in6_dev)
675 676 677 678 679 680
			goto errout;
		devconf = &in6_dev->cnf;
		break;
	}

	err = -ENOBUFS;
681
	skb = nlmsg_new(inet6_netconf_msgsize_devconf(NETCONFA_ALL), GFP_KERNEL);
682
	if (!skb)
683 684 685 686 687
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf,
					 NETLINK_CB(in_skb).portid,
					 nlh->nlmsg_seq, RTM_NEWNETCONF, 0,
688
					 NETCONFA_ALL);
689 690 691 692 693 694 695 696
	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:
697 698 699 700
	if (in6_dev)
		in6_dev_put(in6_dev);
	if (dev)
		dev_put(dev);
701 702 703
	return err;
}

704 705 706
static int inet6_netconf_dump_devconf(struct sk_buff *skb,
				      struct netlink_callback *cb)
{
707
	const struct nlmsghdr *nlh = cb->nlh;
708 709 710 711 712 713 714
	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;

715 716 717 718 719 720 721 722 723 724 725 726 727 728 729
	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;
		}
	}

730 731 732 733 734 735 736
	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();
737 738
		cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^
			  net->dev_base_seq;
739 740 741 742 743 744 745 746 747 748
		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,
749
						       nlh->nlmsg_seq,
750 751
						       RTM_NEWNETCONF,
						       NLM_F_MULTI,
752
						       NETCONFA_ALL) < 0) {
753 754 755
				rcu_read_unlock();
				goto done;
			}
756
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
757 758 759 760 761 762 763 764 765
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,
766
					       nlh->nlmsg_seq,
767
					       RTM_NEWNETCONF, NLM_F_MULTI,
768
					       NETCONFA_ALL) < 0)
769 770 771 772 773 774 775 776
			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,
777
					       nlh->nlmsg_seq,
778
					       RTM_NEWNETCONF, NLM_F_MULTI,
779
					       NETCONFA_ALL) < 0)
780 781 782 783 784 785 786 787 788 789 790
			goto done;
		else
			h++;
	}
done:
	cb->args[0] = h;
	cb->args[1] = idx;

	return skb->len;
}

L
Linus Torvalds 已提交
791 792 793 794 795 796 797 798 799
#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;
800 801
	if (idev->cnf.forwarding)
		dev_disable_lro(dev);
A
Amerigo Wang 已提交
802
	if (dev->flags & IFF_MULTICAST) {
803
		if (idev->cnf.forwarding) {
804
			ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);
805 806 807
			ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_inc(dev, &in6addr_sitelocal_allrouters);
		} else {
808
			ipv6_dev_mc_dec(dev, &in6addr_linklocal_allrouters);
809 810 811
			ipv6_dev_mc_dec(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_dec(dev, &in6addr_sitelocal_allrouters);
		}
L
Linus Torvalds 已提交
812
	}
813 814

	list_for_each_entry(ifa, &idev->addr_list, if_list) {
M
Michal Wrobel 已提交
815 816
		if (ifa->flags&IFA_F_TENTATIVE)
			continue;
L
Linus Torvalds 已提交
817 818 819 820 821
		if (idev->cnf.forwarding)
			addrconf_join_anycast(ifa);
		else
			addrconf_leave_anycast(ifa);
	}
822 823
	inet6_netconf_notify_devconf(dev_net(dev), RTM_NEWNETCONF,
				     NETCONFA_FORWARDING,
824
				     dev->ifindex, &idev->cnf);
L
Linus Torvalds 已提交
825 826 827
}


828
static void addrconf_forward_change(struct net *net, __s32 newf)
L
Linus Torvalds 已提交
829 830 831 832
{
	struct net_device *dev;
	struct inet6_dev *idev;

833
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
834 835
		idev = __in6_dev_get(dev);
		if (idev) {
836 837
			int changed = (!idev->cnf.forwarding) ^ (!newf);
			idev->cnf.forwarding = newf;
L
Linus Torvalds 已提交
838 839 840 841 842
			if (changed)
				dev_forward_change(idev);
		}
	}
}
843

844
static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
845
{
846
	struct net *net;
847 848 849 850
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
851 852

	net = (struct net *)table->extra2;
853 854
	old = *p;
	*p = newf;
855

856
	if (p == &net->ipv6.devconf_dflt->forwarding) {
857
		if ((!newf) ^ (!old))
858 859
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
860 861
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
862 863
		rtnl_unlock();
		return 0;
E
Eric W. Biederman 已提交
864
	}
865

866
	if (p == &net->ipv6.devconf_all->forwarding) {
867 868
		int old_dflt = net->ipv6.devconf_dflt->forwarding;

869
		net->ipv6.devconf_dflt->forwarding = newf;
870
		if ((!newf) ^ (!old_dflt))
871 872
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
873 874 875
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);

876
		addrconf_forward_change(net, newf);
877
		if ((!newf) ^ (!old))
878 879
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
880 881
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
882
	} else if ((!newf) ^ (!old))
883
		dev_forward_change((struct inet6_dev *)table->extra1);
884
	rtnl_unlock();
885

886
	if (newf)
887
		rt6_purge_dflt_routers(net);
888
	return 1;
889
}
890 891 892 893 894 895 896 897 898 899 900 901 902 903

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),
904
							     RTM_NEWNETCONF,
905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926
							     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,
927
						     RTM_NEWNETCONF,
928 929 930 931 932 933 934 935 936 937 938 939
						     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,
940
						     RTM_NEWNETCONF,
941 942 943 944 945 946 947 948 949
						     NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
	}
	rtnl_unlock();

	return 1;
}

L
Linus Torvalds 已提交
950 951
#endif

952
/* Nobody refers to this ifaddr, destroy it */
L
Linus Torvalds 已提交
953 954
void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp)
{
955
	WARN_ON(!hlist_unhashed(&ifp->addr_lst));
956

L
Linus Torvalds 已提交
957
#ifdef NET_REFCNT_DEBUG
958
	pr_debug("%s\n", __func__);
L
Linus Torvalds 已提交
959 960 961 962
#endif

	in6_dev_put(ifp->idev);

963 964 965
	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
966

967
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
968
		pr_warn("Freeing alive inet6 address %p\n", ifp);
L
Linus Torvalds 已提交
969 970 971
		return;
	}

972
	kfree_rcu(ifp, rcu);
L
Linus Torvalds 已提交
973 974
}

B
Brian Haley 已提交
975 976 977
static void
ipv6_link_dev_addr(struct inet6_dev *idev, struct inet6_ifaddr *ifp)
{
978
	struct list_head *p;
979
	int ifp_scope = ipv6_addr_src_scope(&ifp->addr);
B
Brian Haley 已提交
980 981 982 983 984

	/*
	 * Each device address list is sorted in order of scope -
	 * global before linklocal.
	 */
985 986 987
	list_for_each(p, &idev->addr_list) {
		struct inet6_ifaddr *ifa
			= list_entry(p, struct inet6_ifaddr, if_list);
988
		if (ifp_scope >= ipv6_addr_src_scope(&ifa->addr))
B
Brian Haley 已提交
989 990 991
			break;
	}

992
	list_add_tail_rcu(&ifp->if_list, p);
B
Brian Haley 已提交
993 994
}

995
static u32 inet6_addr_hash(const struct net *net, const struct in6_addr *addr)
996
{
997 998 999
	u32 val = ipv6_addr_hash(addr) ^ net_hash_mix(net);

	return hash_32(val, IN6_ADDR_HSIZE_SHIFT);
1000 1001
}

1002
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
1003
			       struct net_device *dev, unsigned int hash)
1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017
{
	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;
}

1018 1019
static int ipv6_add_addr_hash(struct net_device *dev, struct inet6_ifaddr *ifa)
{
1020
	unsigned int hash = inet6_addr_hash(dev_net(dev), &ifa->addr);
1021 1022 1023 1024 1025
	int err = 0;

	spin_lock(&addrconf_hash_lock);

	/* Ignore adding duplicate addresses on an interface */
1026
	if (ipv6_chk_same_addr(dev_net(dev), &ifa->addr, dev, hash)) {
1027
		netdev_dbg(dev, "ipv6_add_addr: already assigned\n");
1028
		err = -EEXIST;
1029 1030
	} else {
		hlist_add_head_rcu(&ifa->addr_lst, &inet6_addr_lst[hash]);
1031 1032 1033 1034 1035 1036 1037
	}

	spin_unlock(&addrconf_hash_lock);

	return err;
}

L
Linus Torvalds 已提交
1038 1039 1040
/* On success it returns ifp with increased reference count */

static struct inet6_ifaddr *
1041
ipv6_add_addr(struct inet6_dev *idev, struct ifa6_config *cfg,
1042
	      bool can_block, struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
1043
{
1044
	gfp_t gfp_flags = can_block ? GFP_KERNEL : GFP_ATOMIC;
1045
	int addr_type = ipv6_addr_type(cfg->pfx);
1046
	struct net *net = dev_net(idev->dev);
L
Linus Torvalds 已提交
1047
	struct inet6_ifaddr *ifa = NULL;
1048
	struct fib6_info *f6i = NULL;
L
Linus Torvalds 已提交
1049
	int err = 0;
1050 1051 1052 1053

	if (addr_type == IPV6_ADDR_ANY ||
	    addr_type & IPV6_ADDR_MULTICAST ||
	    (!(idev->dev->flags & IFF_LOOPBACK) &&
1054
	     !netif_is_l3_master(idev->dev) &&
1055 1056
	     addr_type & IPV6_ADDR_LOOPBACK))
		return ERR_PTR(-EADDRNOTAVAIL);
L
Linus Torvalds 已提交
1057 1058 1059

	if (idev->dead) {
		err = -ENODEV;			/*XXX*/
1060
		goto out;
L
Linus Torvalds 已提交
1061 1062
	}

1063
	if (idev->cnf.disable_ipv6) {
1064
		err = -EACCES;
1065
		goto out;
1066 1067
	}

1068 1069 1070 1071 1072
	/* validator notifier needs to be blocking;
	 * do not call in atomic context
	 */
	if (can_block) {
		struct in6_validator_info i6vi = {
1073
			.i6vi_addr = *cfg->pfx,
1074
			.i6vi_dev = idev,
1075
			.extack = extack,
1076 1077 1078 1079 1080 1081 1082
		};

		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
1083

1084
	ifa = kzalloc(sizeof(*ifa), gfp_flags);
1085
	if (!ifa) {
L
Linus Torvalds 已提交
1086 1087 1088 1089
		err = -ENOBUFS;
		goto out;
	}

1090
	f6i = addrconf_f6i_alloc(net, idev, cfg->pfx, false, gfp_flags);
1091 1092 1093
	if (IS_ERR(f6i)) {
		err = PTR_ERR(f6i);
		f6i = NULL;
L
Linus Torvalds 已提交
1094 1095 1096
		goto out;
	}

1097 1098
	if (net->ipv6.devconf_all->disable_policy ||
	    idev->cnf.disable_policy)
1099
		f6i->dst_nopolicy = true;
1100

1101 1102
	neigh_parms_data_state_setall(idev->nd_parms);

1103 1104 1105
	ifa->addr = *cfg->pfx;
	if (cfg->peer_pfx)
		ifa->peer_addr = *cfg->peer_pfx;
L
Linus Torvalds 已提交
1106 1107

	spin_lock_init(&ifa->lock);
1108
	INIT_DELAYED_WORK(&ifa->dad_work, addrconf_dad_work);
1109
	INIT_HLIST_NODE(&ifa->addr_lst);
1110 1111
	ifa->scope = cfg->scope;
	ifa->prefix_len = cfg->plen;
1112
	ifa->rt_priority = cfg->rt_priority;
1113
	ifa->flags = cfg->ifa_flags;
1114
	/* No need to add the TENTATIVE flag for addresses with NODAD */
1115
	if (!(cfg->ifa_flags & IFA_F_NODAD))
1116
		ifa->flags |= IFA_F_TENTATIVE;
1117 1118
	ifa->valid_lft = cfg->valid_lft;
	ifa->prefered_lft = cfg->preferred_lft;
L
Linus Torvalds 已提交
1119
	ifa->cstamp = ifa->tstamp = jiffies;
1120
	ifa->tokenized = false;
L
Linus Torvalds 已提交
1121

1122
	ifa->rt = f6i;
1123

L
Linus Torvalds 已提交
1124
	ifa->idev = idev;
1125 1126
	in6_dev_hold(idev);

L
Linus Torvalds 已提交
1127
	/* For caller */
1128
	refcount_set(&ifa->refcnt, 1);
L
Linus Torvalds 已提交
1129

1130
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
1131

1132 1133 1134 1135 1136
	err = ipv6_add_addr_hash(idev->dev, ifa);
	if (err < 0) {
		rcu_read_unlock_bh();
		goto out;
	}
L
Linus Torvalds 已提交
1137 1138

	write_lock(&idev->lock);
1139

L
Linus Torvalds 已提交
1140
	/* Add to inet6_dev unicast addr list. */
B
Brian Haley 已提交
1141
	ipv6_link_dev_addr(idev, ifa);
L
Linus Torvalds 已提交
1142 1143

	if (ifa->flags&IFA_F_TEMPORARY) {
1144
		list_add(&ifa->tmp_list, &idev->tempaddr_list);
L
Linus Torvalds 已提交
1145 1146 1147 1148 1149
		in6_ifa_hold(ifa);
	}

	in6_ifa_hold(ifa);
	write_unlock(&idev->lock);
1150

1151
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1152

1153 1154 1155
	inet6addr_notifier_call_chain(NETDEV_UP, ifa);
out:
	if (unlikely(err < 0)) {
1156
		fib6_info_release(f6i);
1157

1158 1159 1160 1161 1162
		if (ifa) {
			if (ifa->idev)
				in6_dev_put(ifa->idev);
			kfree(ifa);
		}
L
Linus Torvalds 已提交
1163 1164 1165 1166 1167 1168
		ifa = ERR_PTR(err);
	}

	return ifa;
}

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
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;
1206 1207
		if (ifa->prefix_len != ifp->prefix_len ||
		    !ipv6_prefix_equal(&ifa->addr, &ifp->addr,
1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233
				       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)
{
1234
	struct fib6_info *f6i;
1235

1236 1237
	f6i = addrconf_get_prefix_route(&ifp->addr, ifp->prefix_len,
					ifp->idev->dev, 0, RTF_DEFAULT, true);
1238
	if (f6i) {
1239
		if (del_rt)
1240
			ip6_del_rt(dev_net(ifp->idev->dev), f6i);
1241
		else {
1242 1243 1244
			if (!(f6i->fib6_flags & RTF_EXPIRES))
				fib6_set_expires(f6i, expires);
			fib6_info_release(f6i);
1245 1246 1247 1248 1249
		}
	}
}


L
Linus Torvalds 已提交
1250 1251 1252 1253
/* This function wants to get referenced ifp and releases it before return */

static void ipv6_del_addr(struct inet6_ifaddr *ifp)
{
1254
	int state;
1255 1256
	enum cleanup_prefix_rt_t action = CLEANUP_PREFIX_RT_NOP;
	unsigned long expires;
L
Linus Torvalds 已提交
1257

1258 1259
	ASSERT_RTNL();

1260
	spin_lock_bh(&ifp->lock);
1261
	state = ifp->state;
1262
	ifp->state = INET6_IFADDR_STATE_DEAD;
1263
	spin_unlock_bh(&ifp->lock);
1264 1265 1266

	if (state == INET6_IFADDR_STATE_DEAD)
		goto out;
L
Linus Torvalds 已提交
1267

1268 1269 1270
	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
1271

1272
	write_lock_bh(&ifp->idev->lock);
1273

L
Linus Torvalds 已提交
1274
	if (ifp->flags&IFA_F_TEMPORARY) {
1275 1276 1277 1278
		list_del(&ifp->tmp_list);
		if (ifp->ifpub) {
			in6_ifa_put(ifp->ifpub);
			ifp->ifpub = NULL;
L
Linus Torvalds 已提交
1279
		}
1280
		__in6_ifa_put(ifp);
L
Linus Torvalds 已提交
1281 1282
	}

1283 1284
	if (ifp->flags & IFA_F_PERMANENT && !(ifp->flags & IFA_F_NOPREFIXROUTE))
		action = check_cleanup_prefix_route(ifp, &expires);
1285

1286
	list_del_rcu(&ifp->if_list);
1287 1288 1289
	__in6_ifa_put(ifp);

	write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
1290

1291
	addrconf_del_dad_work(ifp);
1292

L
Linus Torvalds 已提交
1293 1294
	ipv6_ifa_notify(RTM_DELADDR, ifp);

1295
	inet6addr_notifier_call_chain(NETDEV_DOWN, ifp);
L
Linus Torvalds 已提交
1296

1297 1298 1299
	if (action != CLEANUP_PREFIX_RT_NOP) {
		cleanup_prefix_route(ifp, expires,
			action == CLEANUP_PREFIX_RT_DEL);
L
Linus Torvalds 已提交
1300 1301
	}

1302 1303
	/* clean up prefsrc entries */
	rt6_remove_prefsrc(ifp);
1304
out:
L
Linus Torvalds 已提交
1305 1306 1307
	in6_ifa_put(ifp);
}

E
Eric Dumazet 已提交
1308 1309 1310
static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp,
				struct inet6_ifaddr *ift,
				bool block)
L
Linus Torvalds 已提交
1311 1312 1313
{
	struct inet6_dev *idev = ifp->idev;
	struct in6_addr addr, *tmpaddr;
1314
	unsigned long tmp_tstamp, age;
1315
	unsigned long regen_advance;
1316
	struct ifa6_config cfg;
L
Linus Torvalds 已提交
1317
	int ret = 0;
1318
	unsigned long now = jiffies;
J
Jiri Bohac 已提交
1319
	long max_desync_factor;
1320
	s32 cnf_temp_preferred_lft;
L
Linus Torvalds 已提交
1321

1322
	write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333
	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) {
1334
		write_unlock_bh(&idev->lock);
1335
		pr_info("%s: use_tempaddr is disabled\n", __func__);
L
Linus Torvalds 已提交
1336 1337 1338 1339 1340 1341 1342 1343
		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);
1344
		write_unlock_bh(&idev->lock);
1345 1346
		pr_warn("%s: regeneration time exceeded - disabled temporary address support\n",
			__func__);
L
Linus Torvalds 已提交
1347 1348 1349 1350 1351 1352
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}
	in6_ifa_hold(ifp);
	memcpy(addr.s6_addr, ifp->addr.s6_addr, 8);
1353
	ipv6_try_regen_rndid(idev, tmpaddr);
L
Linus Torvalds 已提交
1354
	memcpy(&addr.s6_addr[8], idev->rndid, 8);
1355
	age = (now - ifp->tstamp) / HZ;
J
Jiri Bohac 已提交
1356 1357 1358 1359 1360 1361 1362 1363

	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
	 */
1364
	cnf_temp_preferred_lft = READ_ONCE(idev->cnf.temp_prefered_lft);
J
Jiri Bohac 已提交
1365 1366
	max_desync_factor = min_t(__u32,
				  idev->cnf.max_desync_factor,
1367
				  cnf_temp_preferred_lft - regen_advance);
J
Jiri Bohac 已提交
1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378

	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;
		}
	}

1379
	memset(&cfg, 0, sizeof(cfg));
1380
	cfg.valid_lft = min_t(__u32, ifp->valid_lft,
1381
			      idev->cnf.temp_valid_lft + age);
1382 1383 1384 1385
	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
1386 1387 1388
	tmp_tstamp = ifp->tstamp;
	spin_unlock_bh(&ifp->lock);

1389
	write_unlock_bh(&idev->lock);
1390

1391 1392 1393 1394
	/* 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.
1395 1396
	 * Use age calculation as in addrconf_verify to avoid unnecessary
	 * temporary addresses being generated.
1397
	 */
1398
	age = (now - tmp_tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
1399
	if (cfg.preferred_lft <= regen_advance + age) {
1400 1401 1402 1403 1404 1405
		in6_ifa_put(ifp);
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}

1406
	cfg.ifa_flags = IFA_F_TEMPORARY;
1407 1408
	/* set in addrconf_prefix_rcv() */
	if (ifp->flags & IFA_F_OPTIMISTIC)
1409
		cfg.ifa_flags |= IFA_F_OPTIMISTIC;
1410

1411 1412
	cfg.pfx = &addr;
	cfg.scope = ipv6_addr_scope(cfg.pfx);
1413

1414
	ift = ipv6_add_addr(idev, &cfg, block, NULL);
1415
	if (IS_ERR(ift)) {
L
Linus Torvalds 已提交
1416 1417
		in6_ifa_put(ifp);
		in6_dev_put(idev);
1418
		pr_info("%s: retry temporary address regeneration\n", __func__);
L
Linus Torvalds 已提交
1419
		tmpaddr = &addr;
1420
		write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1421 1422 1423 1424 1425
		goto retry;
	}

	spin_lock_bh(&ift->lock);
	ift->ifpub = ifp;
1426
	ift->cstamp = now;
L
Linus Torvalds 已提交
1427 1428 1429
	ift->tstamp = tmp_tstamp;
	spin_unlock_bh(&ift->lock);

1430
	addrconf_dad_start(ift);
L
Linus Torvalds 已提交
1431 1432 1433 1434 1435 1436 1437
	in6_ifa_put(ift);
	in6_dev_put(idev);
out:
	return ret;
}

/*
1438
 *	Choose an appropriate source address (RFC3484)
L
Linus Torvalds 已提交
1439
 */
1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452
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,
1453 1454 1455
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	IPV6_SADDR_RULE_NOT_OPTIMISTIC,
#endif
1456 1457 1458
	IPV6_SADDR_RULE_MAX
};

1459
struct ipv6_saddr_score {
1460 1461 1462 1463 1464 1465
	int			rule;
	int			addr_type;
	struct inet6_ifaddr	*ifa;
	DECLARE_BITMAP(scorebits, IPV6_SADDR_RULE_MAX);
	int			scopedist;
	int			matchlen;
1466 1467
};

1468
struct ipv6_saddr_dst {
1469
	const struct in6_addr *addr;
1470 1471 1472
	int ifindex;
	int scope;
	int label;
1473
	unsigned int prefs;
1474
};
1475

D
Dave Jones 已提交
1476
static inline int ipv6_saddr_preferred(int type)
L
Linus Torvalds 已提交
1477
{
U
Ulrich Weber 已提交
1478
	if (type & (IPV6_ADDR_MAPPED|IPV6_ADDR_COMPATv4|IPV6_ADDR_LOOPBACK))
1479 1480
		return 1;
	return 0;
L
Linus Torvalds 已提交
1481 1482
}

1483 1484
static bool ipv6_use_optimistic_addr(struct net *net,
				     struct inet6_dev *idev)
1485 1486
{
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
1487 1488 1489 1490 1491 1492 1493 1494
	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;
1495 1496 1497 1498 1499
#else
	return false;
#endif
}

1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514
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
}

1515 1516
static int ipv6_get_saddr_eval(struct net *net,
			       struct ipv6_saddr_score *score,
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 1554 1555
			       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
1556
		 *  B-15 |    \        if scope is enough for destination.
1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574
		 *       |             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:
1575
	    {
1576
		/* Rule 3: Avoid deprecated and optimistic addresses */
1577 1578
		u8 avoid = IFA_F_DEPRECATED;

1579
		if (!ipv6_use_optimistic_addr(net, score->ifa->idev))
1580
			avoid |= IFA_F_OPTIMISTIC;
1581
		ret = ipv6_saddr_preferred(score->addr_type) ||
1582
		      !(score->ifa->flags & avoid);
1583
		break;
1584
	    }
1585 1586
#ifdef CONFIG_IPV6_MIP6
	case IPV6_SADDR_RULE_HOA:
1587
	    {
1588
		/* Rule 4: Prefer home address */
1589 1590
		int prefhome = !(dst->prefs & IPV6_PREFER_SRC_COA);
		ret = !(score->ifa->flags & IFA_F_HOMEADDRESS) ^ prefhome;
1591
		break;
1592
	    }
1593 1594 1595 1596 1597 1598 1599 1600
#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 */
1601 1602
		ret = ipv6_addr_label(net,
				      &score->ifa->addr, score->addr_type,
1603 1604 1605
				      score->ifa->idev->dev->ifindex) == dst->label;
		break;
	case IPV6_SADDR_RULE_PRIVACY:
1606
	    {
1607
		/* Rule 7: Prefer public address
L
Lucas De Marchi 已提交
1608
		 * Note: prefer temporary address if use_tempaddr >= 2
1609
		 */
1610 1611 1612 1613
		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;
1614
		break;
1615
	    }
1616 1617 1618 1619 1620 1621 1622 1623 1624
	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 */
1625 1626 1627 1628
		ret = ipv6_addr_diff(&score->ifa->addr, dst->addr);
		if (ret > score->ifa->prefix_len)
			ret = score->ifa->prefix_len;
		score->matchlen = ret;
1629
		break;
1630 1631 1632 1633 1634 1635 1636 1637
#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
1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648
	default:
		ret = 0;
	}

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

1649 1650 1651 1652 1653
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)
1654
{
1655
	struct ipv6_saddr_score *score = &scores[1 - hiscore_idx], *hiscore = &scores[hiscore_idx];
1656

1657
	list_for_each_entry_rcu(score->ifa, &idev->addr_list, if_list) {
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 1706 1707
		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);
1708
				hiscore_idx = 1 - hiscore_idx;
1709 1710 1711 1712 1713 1714 1715 1716 1717

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

				break;
			}
		}
	}
out:
1718
	return hiscore_idx;
1719 1720
}

1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742
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;
}

1743
int ipv6_dev_get_saddr(struct net *net, const struct net_device *dst_dev,
1744
		       const struct in6_addr *daddr, unsigned int prefs,
1745
		       struct in6_addr *saddr)
L
Linus Torvalds 已提交
1746
{
1747
	struct ipv6_saddr_score scores[2], *hiscore;
1748
	struct ipv6_saddr_dst dst;
1749
	struct inet6_dev *idev;
1750
	struct net_device *dev;
1751
	int dst_type;
1752
	bool use_oif_addr = false;
1753
	int hiscore_idx = 0;
1754
	int ret = 0;
L
Linus Torvalds 已提交
1755

1756 1757 1758 1759
	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);
1760
	dst.label = ipv6_addr_label(net, daddr, dst_type, dst.ifindex);
1761
	dst.prefs = prefs;
1762

1763 1764
	scores[hiscore_idx].rule = -1;
	scores[hiscore_idx].ifa = NULL;
L
Linus Torvalds 已提交
1765

1766
	rcu_read_lock();
L
Linus Torvalds 已提交
1767

1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778
	/* 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.)
1779 1780 1781 1782
	 *  - "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)
1783 1784
	 */
	if (dst_dev) {
1785
		idev = __in6_dev_get(dst_dev);
1786
		if ((dst_type & IPV6_ADDR_MULTICAST) ||
1787 1788
		    dst.scope <= IPV6_ADDR_SCOPE_LINKLOCAL ||
		    (idev && idev->cnf.use_oif_addrs_only)) {
1789 1790 1791
			use_oif_addr = true;
		}
	}
1792

1793
	if (use_oif_addr) {
1794
		if (idev)
1795
			hiscore_idx = __ipv6_dev_get_saddr(net, &dst, idev, scores, hiscore_idx);
1796
	} else {
1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815
		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;
		}

1816
		for_each_netdev_rcu(net, dev) {
1817 1818 1819 1820 1821
			/* only consider addresses on devices in the
			 * same L3 domain
			 */
			if (l3mdev_master_ifindex_rcu(dev) != master_idx)
				continue;
1822 1823
			idev = __in6_dev_get(dev);
			if (!idev)
1824
				continue;
1825
			hiscore_idx = __ipv6_dev_get_saddr(net, &dst, idev, scores, hiscore_idx);
L
Linus Torvalds 已提交
1826 1827
		}
	}
1828 1829

out:
1830
	hiscore = &scores[hiscore_idx];
1831
	if (!hiscore->ifa)
1832 1833 1834
		ret = -EADDRNOTAVAIL;
	else
		*saddr = hiscore->ifa->addr;
1835

1836 1837
	rcu_read_unlock();
	return ret;
L
Linus Torvalds 已提交
1838
}
1839
EXPORT_SYMBOL(ipv6_dev_get_saddr);
L
Linus Torvalds 已提交
1840

1841
int __ipv6_get_lladdr(struct inet6_dev *idev, struct in6_addr *addr,
1842
		      u32 banned_flags)
1843 1844 1845 1846
{
	struct inet6_ifaddr *ifp;
	int err = -EADDRNOTAVAIL;

1847 1848 1849
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
1850 1851 1852 1853 1854 1855 1856 1857 1858 1859
		if (ifp->scope == IFA_LINK &&
		    !(ifp->flags & banned_flags)) {
			*addr = ifp->addr;
			err = 0;
			break;
		}
	}
	return err;
}

1860
int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr,
1861
		    u32 banned_flags)
L
Linus Torvalds 已提交
1862 1863 1864 1865
{
	struct inet6_dev *idev;
	int err = -EADDRNOTAVAIL;

1866
	rcu_read_lock();
S
Stephen Hemminger 已提交
1867 1868
	idev = __in6_dev_get(dev);
	if (idev) {
L
Linus Torvalds 已提交
1869
		read_lock_bh(&idev->lock);
1870
		err = __ipv6_get_lladdr(idev, addr, banned_flags);
L
Linus Torvalds 已提交
1871 1872
		read_unlock_bh(&idev->lock);
	}
1873
	rcu_read_unlock();
L
Linus Torvalds 已提交
1874 1875 1876
	return err;
}

1877
static int ipv6_count_addresses(const struct inet6_dev *idev)
L
Linus Torvalds 已提交
1878
{
1879
	const struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
1880 1881
	int cnt = 0;

1882 1883
	rcu_read_lock();
	list_for_each_entry_rcu(ifp, &idev->addr_list, if_list)
L
Linus Torvalds 已提交
1884
		cnt++;
1885
	rcu_read_unlock();
L
Linus Torvalds 已提交
1886 1887 1888
	return cnt;
}

1889
int ipv6_chk_addr(struct net *net, const struct in6_addr *addr,
1890
		  const struct net_device *dev, int strict)
1891
{
1892 1893
	return ipv6_chk_addr_and_flags(net, addr, dev, !dev,
				       strict, IFA_F_TENTATIVE);
1894 1895 1896
}
EXPORT_SYMBOL(ipv6_chk_addr);

1897 1898 1899 1900 1901 1902 1903 1904 1905
/* 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)
 */
1906
int ipv6_chk_addr_and_flags(struct net *net, const struct in6_addr *addr,
1907 1908
			    const struct net_device *dev, bool skip_dev_check,
			    int strict, u32 banned_flags)
L
Linus Torvalds 已提交
1909
{
1910
	unsigned int hash = inet6_addr_hash(net, addr);
1911
	const struct net_device *l3mdev;
1912
	struct inet6_ifaddr *ifp;
1913
	u32 ifp_flags;
L
Linus Torvalds 已提交
1914

1915
	rcu_read_lock();
1916

1917
	l3mdev = l3mdev_master_dev_rcu(dev);
1918 1919 1920
	if (skip_dev_check)
		dev = NULL;

1921
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
1922
		if (!net_eq(dev_net(ifp->idev->dev), net))
1923
			continue;
1924 1925 1926 1927

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

1928 1929 1930 1931 1932 1933
		/* 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
1934
		if (ipv6_addr_equal(&ifp->addr, addr) &&
1935
		    !(ifp_flags&banned_flags) &&
1936
		    (!dev || ifp->idev->dev == dev ||
1937
		     !(ifp->scope&(IFA_LINK|IFA_HOST) || strict))) {
1938
			rcu_read_unlock();
1939
			return 1;
L
Linus Torvalds 已提交
1940 1941
		}
	}
1942

1943
	rcu_read_unlock();
1944
	return 0;
L
Linus Torvalds 已提交
1945
}
1946
EXPORT_SYMBOL(ipv6_chk_addr_and_flags);
1947

L
Linus Torvalds 已提交
1948

1949 1950 1951 1952 1953 1954
/* 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)
{
1955 1956
	const struct inet6_ifaddr *ifa;
	const struct inet6_dev *idev;
1957 1958 1959 1960 1961
	bool ret = false;

	rcu_read_lock();
	idev = __in6_dev_get(dev);
	if (idev) {
1962
		list_for_each_entry_rcu(ifa, &idev->addr_list, if_list) {
1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973
			ret = ipv6_prefix_equal(addr, &ifa->addr, prefix_len);
			if (ret)
				break;
		}
	}
	rcu_read_unlock();

	return ret;
}
EXPORT_SYMBOL(ipv6_chk_custom_prefix);

1974
int ipv6_chk_prefix(const struct in6_addr *addr, struct net_device *dev)
1975
{
1976 1977
	const struct inet6_ifaddr *ifa;
	const struct inet6_dev *idev;
1978 1979 1980 1981 1982 1983
	int	onlink;

	onlink = 0;
	rcu_read_lock();
	idev = __in6_dev_get(dev);
	if (idev) {
1984
		list_for_each_entry_rcu(ifa, &idev->addr_list, if_list) {
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
			onlink = ipv6_prefix_equal(addr, &ifa->addr,
						   ifa->prefix_len);
			if (onlink)
				break;
		}
	}
	rcu_read_unlock();
	return onlink;
}
EXPORT_SYMBOL(ipv6_chk_prefix);

1996
struct inet6_ifaddr *ipv6_get_ifaddr(struct net *net, const struct in6_addr *addr,
1997
				     struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1998
{
1999
	unsigned int hash = inet6_addr_hash(net, addr);
2000
	struct inet6_ifaddr *ifp, *result = NULL;
L
Linus Torvalds 已提交
2001

2002 2003
	rcu_read_lock();
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
2004
		if (!net_eq(dev_net(ifp->idev->dev), net))
2005
			continue;
L
Linus Torvalds 已提交
2006
		if (ipv6_addr_equal(&ifp->addr, addr)) {
2007
			if (!dev || ifp->idev->dev == dev ||
L
Linus Torvalds 已提交
2008
			    !(ifp->scope&(IFA_LINK|IFA_HOST) || strict)) {
2009
				result = ifp;
L
Linus Torvalds 已提交
2010 2011 2012 2013 2014
				in6_ifa_hold(ifp);
				break;
			}
		}
	}
2015
	rcu_read_unlock();
L
Linus Torvalds 已提交
2016

2017
	return result;
L
Linus Torvalds 已提交
2018 2019 2020 2021
}

/* Gets referenced address, destroys ifaddr */

B
Brian Haley 已提交
2022
static void addrconf_dad_stop(struct inet6_ifaddr *ifp, int dad_failed)
L
Linus Torvalds 已提交
2023
{
2024 2025 2026
	if (dad_failed)
		ifp->flags |= IFA_F_DADFAILED;

2027
	if (ifp->flags&IFA_F_TEMPORARY) {
L
Linus Torvalds 已提交
2028 2029 2030 2031 2032 2033
		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
2034
			ipv6_create_tempaddr(ifpub, ifp, true);
L
Linus Torvalds 已提交
2035 2036 2037 2038 2039
			in6_ifa_put(ifpub);
		} else {
			spin_unlock_bh(&ifp->lock);
		}
		ipv6_del_addr(ifp);
2040 2041 2042 2043
	} else if (ifp->flags&IFA_F_PERMANENT || !dad_failed) {
		spin_lock_bh(&ifp->lock);
		addrconf_del_dad_work(ifp);
		ifp->flags |= IFA_F_TENTATIVE;
2044 2045
		if (dad_failed)
			ifp->flags &= ~IFA_F_OPTIMISTIC;
2046 2047 2048 2049
		spin_unlock_bh(&ifp->lock);
		if (dad_failed)
			ipv6_ifa_notify(0, ifp);
		in6_ifa_put(ifp);
2050
	} else {
L
Linus Torvalds 已提交
2051
		ipv6_del_addr(ifp);
2052
	}
L
Linus Torvalds 已提交
2053 2054
}

H
Herbert Xu 已提交
2055 2056 2057 2058
static int addrconf_dad_end(struct inet6_ifaddr *ifp)
{
	int err = -ENOENT;

2059
	spin_lock_bh(&ifp->lock);
H
Herbert Xu 已提交
2060 2061 2062 2063
	if (ifp->state == INET6_IFADDR_STATE_DAD) {
		ifp->state = INET6_IFADDR_STATE_POSTDAD;
		err = 0;
	}
2064
	spin_unlock_bh(&ifp->lock);
H
Herbert Xu 已提交
2065 2066 2067 2068

	return err;
}

2069
void addrconf_dad_failure(struct sk_buff *skb, struct inet6_ifaddr *ifp)
2070
{
2071
	struct inet6_dev *idev = ifp->idev;
2072
	struct net *net = dev_net(ifp->idev->dev);
2073

2074 2075
	if (addrconf_dad_end(ifp)) {
		in6_ifa_put(ifp);
H
Herbert Xu 已提交
2076
		return;
2077
	}
H
Herbert Xu 已提交
2078

2079 2080
	net_info_ratelimited("%s: IPv6 duplicate address %pI6c used by %pM detected!\n",
			     ifp->idev->dev->name, &ifp->addr, eth_hdr(skb)->h_source);
2081

2082 2083 2084 2085 2086 2087
	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;
2088 2089 2090 2091 2092 2093 2094 2095
		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,
		};
2096

2097
		if (retries > net->ipv6.sysctl.idgen_retries) {
2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108
			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);
2109

2110 2111 2112 2113 2114 2115 2116 2117
		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);

2118
		ifp2 = ipv6_add_addr(idev, &cfg, false, NULL);
2119 2120 2121 2122 2123 2124 2125 2126
		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);

2127
		addrconf_mod_dad_work(ifp2, net->ipv6.sysctl.idgen_delay);
2128 2129 2130
		in6_ifa_put(ifp2);
lock_errdad:
		spin_lock_bh(&ifp->lock);
2131 2132
	}

2133
errdad:
2134 2135
	/* transition from _POSTDAD to _ERRDAD */
	ifp->state = INET6_IFADDR_STATE_ERRDAD;
2136
	spin_unlock_bh(&ifp->lock);
2137 2138

	addrconf_mod_dad_work(ifp, 0);
2139
	in6_ifa_put(ifp);
2140
}
L
Linus Torvalds 已提交
2141

2142 2143
/* Join to solicited addr multicast group.
 * caller must hold RTNL */
2144
void addrconf_join_solict(struct net_device *dev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2145 2146 2147 2148 2149 2150 2151 2152 2153 2154
{
	struct in6_addr maddr;

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

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

2155
/* caller must hold RTNL */
2156
void addrconf_leave_solict(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2157 2158 2159 2160 2161 2162 2163 2164 2165 2166
{
	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);
}

2167
/* caller must hold RTNL */
2168
static void addrconf_join_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
2169 2170
{
	struct in6_addr addr;
2171

2172
	if (ifp->prefix_len >= 127) /* RFC 6164 */
2173
		return;
L
Linus Torvalds 已提交
2174 2175 2176
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
W
WANG Cong 已提交
2177
	__ipv6_dev_ac_inc(ifp->idev, &addr);
L
Linus Torvalds 已提交
2178 2179
}

2180
/* caller must hold RTNL */
2181
static void addrconf_leave_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
2182 2183
{
	struct in6_addr addr;
2184

2185
	if (ifp->prefix_len >= 127) /* RFC 6164 */
2186
		return;
L
Linus Torvalds 已提交
2187 2188 2189 2190 2191 2192
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
	__ipv6_dev_ac_dec(ifp->idev, &addr);
}

2193
static int addrconf_ifid_6lowpan(u8 *eui, struct net_device *dev)
2194
{
2195 2196
	switch (dev->addr_len) {
	case ETH_ALEN:
2197 2198 2199 2200 2201
		memcpy(eui, dev->dev_addr, 3);
		eui[3] = 0xFF;
		eui[4] = 0xFE;
		memcpy(eui + 5, dev->dev_addr + 3, 3);
		break;
2202 2203 2204 2205 2206
	case EUI64_ADDR_LEN:
		memcpy(eui, dev->dev_addr, EUI64_ADDR_LEN);
		eui[0] ^= 2;
		break;
	default:
2207
		return -1;
2208 2209
	}

2210 2211 2212
	return 0;
}

2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226
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;
}

2227 2228 2229 2230 2231 2232
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);
2233
	eui[7] = *(u8 *)dev->dev_addr;
2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245
	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;
}

2246
static int __ipv6_isatap_ifid(u8 *eui, __be32 addr)
2247
{
2248 2249
	if (addr == 0)
		return -1;
2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269
	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;
}

2270 2271 2272 2273 2274
static int addrconf_ifid_gre(u8 *eui, struct net_device *dev)
{
	return __ipv6_isatap_ifid(eui, *(__be32 *)dev->dev_addr);
}

2275 2276 2277 2278 2279 2280 2281 2282 2283 2284
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
2285 2286 2287 2288 2289
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev)
{
	switch (dev->type) {
	case ARPHRD_ETHER:
	case ARPHRD_FDDI:
2290
		return addrconf_ifid_eui48(eui, dev);
L
Linus Torvalds 已提交
2291
	case ARPHRD_ARCNET:
2292
		return addrconf_ifid_arcnet(eui, dev);
L
Linus Torvalds 已提交
2293
	case ARPHRD_INFINIBAND:
2294
		return addrconf_ifid_infiniband(eui, dev);
F
Fred L. Templin 已提交
2295
	case ARPHRD_SIT:
2296
		return addrconf_ifid_sit(eui, dev);
2297
	case ARPHRD_IPGRE:
2298
	case ARPHRD_TUNNEL:
2299
		return addrconf_ifid_gre(eui, dev);
2300
	case ARPHRD_6LOWPAN:
2301
		return addrconf_ifid_6lowpan(eui, dev);
2302 2303
	case ARPHRD_IEEE1394:
		return addrconf_ifid_ieee1394(eui, dev);
2304
	case ARPHRD_TUNNEL6:
2305
	case ARPHRD_IP6GRE:
2306
	case ARPHRD_RAWIP:
2307
		return addrconf_ifid_ip6tnl(eui, dev);
L
Linus Torvalds 已提交
2308 2309 2310 2311 2312 2313 2314 2315 2316 2317
	}
	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);
2318 2319 2320
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
L
Linus Torvalds 已提交
2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331
		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) */
2332
static void ipv6_regen_rndid(struct inet6_dev *idev)
L
Linus Torvalds 已提交
2333 2334
{
regen:
2335
	get_random_bytes(idev->rndid, sizeof(idev->rndid));
L
Linus Torvalds 已提交
2336 2337 2338 2339 2340 2341 2342 2343
	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
2344
	 *  - ISATAP (RFC4214) 6.1
L
Linus Torvalds 已提交
2345 2346 2347 2348
	 *	00-00-5E-FE-xx-xx-xx-xx
	 *  - value 0
	 *  - XXX: already assigned to an address on the device
	 */
2349
	if (idev->rndid[0] == 0xfd &&
L
Linus Torvalds 已提交
2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360
	    (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;
	}
}

2361
static void  ipv6_try_regen_rndid(struct inet6_dev *idev, struct in6_addr *tmpaddr)
2362
{
L
Linus Torvalds 已提交
2363
	if (tmpaddr && memcmp(idev->rndid, &tmpaddr->s6_addr[8], 8) == 0)
2364
		ipv6_regen_rndid(idev);
L
Linus Torvalds 已提交
2365 2366 2367 2368 2369 2370 2371
}

/*
 *	Add prefix route.
 */

static void
2372 2373 2374
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
2375
{
2376
	struct fib6_config cfg = {
D
David Ahern 已提交
2377
		.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_PREFIX,
2378
		.fc_metric = metric ? : IP6_RT_PRIO_ADDRCONF,
2379 2380 2381 2382
		.fc_ifindex = dev->ifindex,
		.fc_expires = expires,
		.fc_dst_len = plen,
		.fc_flags = RTF_UP | flags,
2383
		.fc_nlinfo.nl_net = dev_net(dev),
2384
		.fc_protocol = RTPROT_KERNEL,
2385
		.fc_type = RTN_UNICAST,
2386
	};
L
Linus Torvalds 已提交
2387

A
Alexey Dobriyan 已提交
2388
	cfg.fc_dst = *pfx;
L
Linus Torvalds 已提交
2389 2390 2391 2392 2393

	/* 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
2394
#if IS_ENABLED(CONFIG_IPV6_SIT)
2395 2396
	if (dev->type == ARPHRD_SIT && (dev->flags & IFF_POINTOPOINT))
		cfg.fc_flags |= RTF_NONEXTHOP;
2397
#endif
L
Linus Torvalds 已提交
2398

2399
	ip6_route_add(&cfg, gfp_flags, NULL);
L
Linus Torvalds 已提交
2400 2401
}

2402

2403
static struct fib6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
2404 2405
						  int plen,
						  const struct net_device *dev,
2406 2407
						  u32 flags, u32 noflags,
						  bool no_gw)
2408 2409
{
	struct fib6_node *fn;
2410
	struct fib6_info *rt = NULL;
2411
	struct fib6_table *table;
D
David Ahern 已提交
2412
	u32 tb_id = l3mdev_fib_table(dev) ? : RT6_TABLE_PREFIX;
2413

D
David Ahern 已提交
2414
	table = fib6_get_table(dev_net(dev), tb_id);
2415
	if (!table)
2416 2417
		return NULL;

2418
	rcu_read_lock();
2419
	fn = fib6_locate(&table->tb6_root, pfx, plen, NULL, 0, true);
2420 2421
	if (!fn)
		goto out;
2422

2423
	for_each_fib6_node_rt_rcu(fn) {
D
David Ahern 已提交
2424
		if (rt->fib6_nh.fib_nh_dev->ifindex != dev->ifindex)
2425
			continue;
2426
		if (no_gw && rt->fib6_nh.fib_nh_gw_family)
2427
			continue;
2428
		if ((rt->fib6_flags & flags) != flags)
2429
			continue;
2430
		if ((rt->fib6_flags & noflags) != 0)
2431
			continue;
2432 2433
		if (!fib6_info_hold_safe(rt))
			continue;
2434 2435 2436
		break;
	}
out:
2437
	rcu_read_unlock();
2438 2439 2440 2441
	return rt;
}


L
Linus Torvalds 已提交
2442 2443 2444 2445
/* Create "default" multicast route to the interface */

static void addrconf_add_mroute(struct net_device *dev)
{
2446
	struct fib6_config cfg = {
D
David Ahern 已提交
2447
		.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_LOCAL,
2448 2449 2450 2451
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_dst_len = 8,
		.fc_flags = RTF_UP,
2452
		.fc_type = RTN_UNICAST,
2453
		.fc_nlinfo.nl_net = dev_net(dev),
2454 2455 2456 2457
	};

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

2458
	ip6_route_add(&cfg, GFP_KERNEL, NULL);
L
Linus Torvalds 已提交
2459 2460 2461 2462 2463 2464 2465 2466
}

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

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
2467 2468
	idev = ipv6_find_idev(dev);
	if (!idev)
2469 2470 2471 2472
		return ERR_PTR(-ENOBUFS);

	if (idev->cnf.disable_ipv6)
		return ERR_PTR(-EACCES);
L
Linus Torvalds 已提交
2473 2474

	/* Add default multicast route */
2475
	if (!(dev->flags & IFF_LOOPBACK) && !netif_is_l3_master(dev))
2476
		addrconf_add_mroute(dev);
L
Linus Torvalds 已提交
2477 2478 2479 2480

	return idev;
}

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 2506 2507 2508 2509 2510
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
2511
			       idev->desync_factor - age;
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 2537 2538 2539 2540 2541
		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
2542
		ipv6_create_tempaddr(ifp, NULL, false);
2543 2544 2545 2546 2547
	} else {
		read_unlock_bh(&idev->lock);
	}
}

2548 2549
static bool is_addr_mode_generate_stable(struct inet6_dev *idev)
{
2550 2551
	return idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_STABLE_PRIVACY ||
	       idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_RANDOM;
2552 2553
}

A
Alexander Aring 已提交
2554 2555 2556 2557 2558 2559
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)
2560 2561 2562 2563 2564 2565
{
	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;
2566 2567 2568 2569 2570 2571 2572 2573
		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,
		};
2574 2575

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
2576 2577
		if ((net->ipv6.devconf_all->optimistic_dad ||
		     in6_dev->cnf.optimistic_dad) &&
2578
		    !net->ipv6.devconf_all->forwarding && sllao)
2579
			cfg.ifa_flags |= IFA_F_OPTIMISTIC;
2580 2581 2582 2583 2584 2585 2586
#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)
2587
			ifp = ipv6_add_addr(in6_dev, &cfg, false, NULL);
2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612

		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;
2613
		if (!create && stored_lft) {
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 2648 2649 2650 2651 2652
			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
2653
EXPORT_SYMBOL_GPL(addrconf_prefix_rcv_add_addr);
2654

2655
void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
L
Linus Torvalds 已提交
2656 2657 2658 2659
{
	struct prefix_info *pinfo;
	__u32 valid_lft;
	__u32 prefered_lft;
2660
	int addr_type, err;
2661
	u32 addr_flags = 0;
L
Linus Torvalds 已提交
2662
	struct inet6_dev *in6_dev;
2663
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
2664 2665

	pinfo = (struct prefix_info *) opt;
2666

L
Linus Torvalds 已提交
2667
	if (len < sizeof(struct prefix_info)) {
2668
		netdev_dbg(dev, "addrconf: prefix option too short\n");
L
Linus Torvalds 已提交
2669 2670
		return;
	}
2671

L
Linus Torvalds 已提交
2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684
	/*
	 *	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) {
2685
		net_warn_ratelimited("addrconf: prefix option has invalid lifetime\n");
L
Linus Torvalds 已提交
2686 2687 2688 2689 2690
		return;
	}

	in6_dev = in6_dev_get(dev);

2691
	if (!in6_dev) {
2692 2693
		net_dbg_ratelimited("addrconf: device %s not configured\n",
				    dev->name);
L
Linus Torvalds 已提交
2694 2695 2696 2697 2698 2699 2700 2701 2702
		return;
	}

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

2703
	if (pinfo->onlink) {
2704
		struct fib6_info *rt;
2705 2706
		unsigned long rt_expires;

2707 2708 2709 2710 2711
		/* 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.
		 */
2712 2713 2714 2715
		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
2716

2717 2718
		if (addrconf_finite_timeout(rt_expires))
			rt_expires *= HZ;
L
Linus Torvalds 已提交
2719

2720 2721 2722 2723
		rt = addrconf_get_prefix_route(&pinfo->prefix,
					       pinfo->prefix_len,
					       dev,
					       RTF_ADDRCONF | RTF_PREFIX_RT,
2724
					       RTF_DEFAULT, true);
L
Linus Torvalds 已提交
2725

2726
		if (rt) {
2727 2728
			/* Autoconf prefix route */
			if (valid_lft == 0) {
2729
				ip6_del_rt(net, rt);
2730
				rt = NULL;
2731
			} else if (addrconf_finite_timeout(rt_expires)) {
2732
				/* not infinity */
2733
				fib6_set_expires(rt, jiffies + rt_expires);
2734
			} else {
2735
				fib6_clean_expires(rt);
L
Linus Torvalds 已提交
2736 2737
			}
		} else if (valid_lft) {
2738
			clock_t expires = 0;
2739 2740
			int flags = RTF_ADDRCONF | RTF_PREFIX_RT;
			if (addrconf_finite_timeout(rt_expires)) {
2741 2742 2743 2744
				/* not infinity */
				flags |= RTF_EXPIRES;
				expires = jiffies_to_clock_t(rt_expires);
			}
L
Linus Torvalds 已提交
2745
			addrconf_prefix_route(&pinfo->prefix, pinfo->prefix_len,
2746 2747
					      0, dev, expires, flags,
					      GFP_ATOMIC);
L
Linus Torvalds 已提交
2748
		}
2749
		fib6_info_release(rt);
L
Linus Torvalds 已提交
2750 2751 2752 2753 2754 2755
	}

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

	if (pinfo->autoconf && in6_dev->cnf.autoconf) {
		struct in6_addr addr;
2756
		bool tokenized = false, dev_addr_generated = false;
L
Linus Torvalds 已提交
2757 2758 2759

		if (pinfo->prefix_len == 64) {
			memcpy(&addr, &pinfo->prefix, 8);
2760 2761 2762 2763 2764 2765

			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);
2766
				tokenized = true;
2767
			} else if (is_addr_mode_generate_stable(in6_dev) &&
2768 2769
				   !ipv6_generate_stable_address(&addr, 0,
								 in6_dev)) {
2770
				addr_flags |= IFA_F_STABLE_PRIVACY;
2771
				goto ok;
2772 2773
			} else if (ipv6_generate_eui64(addr.s6_addr + 8, dev) &&
				   ipv6_inherit_eui64(addr.s6_addr + 8, in6_dev)) {
2774
				goto put;
2775 2776
			} else {
				dev_addr_generated = true;
L
Linus Torvalds 已提交
2777 2778 2779
			}
			goto ok;
		}
2780 2781
		net_dbg_ratelimited("IPv6 addrconf: prefix with wrong length %d\n",
				    pinfo->prefix_len);
2782
		goto put;
L
Linus Torvalds 已提交
2783 2784

ok:
2785 2786 2787 2788 2789 2790 2791
		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;
2792 2793 2794 2795 2796 2797 2798 2799 2800

		/* 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
2801 2802
	}
	inet6_prefix_notify(RTM_NEWPREFIX, in6_dev, pinfo);
2803
put:
L
Linus Torvalds 已提交
2804 2805 2806 2807 2808 2809 2810 2811
	in6_dev_put(in6_dev);
}

/*
 *	Set destination address.
 *	Special case for SIT interfaces where we create a new "virtual"
 *	device.
 */
2812
int addrconf_set_dstaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823
{
	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;

2824
	dev = __dev_get_by_index(net, ireq.ifr6_ifindex);
L
Linus Torvalds 已提交
2825 2826

	err = -ENODEV;
2827
	if (!dev)
L
Linus Torvalds 已提交
2828 2829
		goto err_exit;

A
Amerigo Wang 已提交
2830
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2831
	if (dev->type == ARPHRD_SIT) {
2832
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846
		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;
2847
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;
L
Linus Torvalds 已提交
2848

2849 2850 2851 2852 2853 2854 2855 2856
		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
2857 2858 2859

		if (err == 0) {
			err = -ENOBUFS;
2860 2861
			dev = __dev_get_by_name(net, p.name);
			if (!dev)
L
Linus Torvalds 已提交
2862
				goto err_exit;
2863
			err = dev_open(dev, NULL);
L
Linus Torvalds 已提交
2864 2865
		}
	}
2866
#endif
L
Linus Torvalds 已提交
2867 2868 2869 2870 2871 2872

err_exit:
	rtnl_unlock();
	return err;
}

2873 2874 2875 2876 2877 2878 2879 2880 2881
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)
2882
		ret = ipv6_sock_mc_join(sk, ifindex, addr);
2883
	else
2884
		ret = ipv6_sock_mc_drop(sk, ifindex, addr);
2885 2886 2887 2888 2889
	release_sock(sk);

	return ret;
}

L
Linus Torvalds 已提交
2890 2891 2892
/*
 *	Manual configuration of address on an interface
 */
2893
static int inet6_addr_add(struct net *net, int ifindex,
2894
			  struct ifa6_config *cfg,
2895
			  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
2896 2897 2898 2899
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2900 2901
	unsigned long timeout;
	clock_t expires;
2902
	u32 flags;
L
Linus Torvalds 已提交
2903 2904

	ASSERT_RTNL();
2905

2906
	if (cfg->plen > 128)
2907 2908
		return -EINVAL;

2909
	/* check the lifetime */
2910
	if (!cfg->valid_lft || cfg->preferred_lft > cfg->valid_lft)
2911 2912
		return -EINVAL;

2913
	if (cfg->ifa_flags & IFA_F_MANAGETEMPADDR && cfg->plen != 64)
2914 2915
		return -EINVAL;

2916 2917
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2918
		return -ENODEV;
2919

2920 2921 2922
	idev = addrconf_add_dev(dev);
	if (IS_ERR(idev))
		return PTR_ERR(idev);
L
Linus Torvalds 已提交
2923

2924
	if (cfg->ifa_flags & IFA_F_MCAUTOJOIN) {
2925
		int ret = ipv6_mc_config(net->ipv6.mc_autojoin_sk,
2926
					 true, cfg->pfx, ifindex);
2927 2928 2929 2930 2931

		if (ret < 0)
			return ret;
	}

2932
	cfg->scope = ipv6_addr_scope(cfg->pfx);
L
Linus Torvalds 已提交
2933

2934
	timeout = addrconf_timeout_fixup(cfg->valid_lft, HZ);
2935 2936
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
2937
		cfg->valid_lft = timeout;
2938
		flags = RTF_EXPIRES;
2939 2940 2941
	} else {
		expires = 0;
		flags = 0;
2942
		cfg->ifa_flags |= IFA_F_PERMANENT;
2943
	}
2944

2945
	timeout = addrconf_timeout_fixup(cfg->preferred_lft, HZ);
2946 2947
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
2948 2949
			cfg->ifa_flags |= IFA_F_DEPRECATED;
		cfg->preferred_lft = timeout;
2950
	}
2951

2952
	ifp = ipv6_add_addr(idev, cfg, true, extack);
L
Linus Torvalds 已提交
2953
	if (!IS_ERR(ifp)) {
2954
		if (!(cfg->ifa_flags & IFA_F_NOPREFIXROUTE)) {
2955 2956 2957
			addrconf_prefix_route(&ifp->addr, ifp->prefix_len,
					      ifp->rt_priority, dev, expires,
					      flags, GFP_KERNEL);
2958 2959
		}

2960 2961 2962 2963 2964
		/* 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);
2965 2966 2967 2968 2969
		/*
		 * Note that section 3.1 of RFC 4429 indicates
		 * that the Optimistic flag should not be set for
		 * manually configured addresses
		 */
2970
		addrconf_dad_start(ifp);
2971 2972 2973
		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
2974
		in6_ifa_put(ifp);
2975
		addrconf_verify_rtnl();
L
Linus Torvalds 已提交
2976
		return 0;
2977 2978 2979
	} 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
2980 2981 2982 2983 2984
	}

	return PTR_ERR(ifp);
}

2985 2986
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
2987 2988 2989 2990
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2991

2992 2993 2994
	if (plen > 128)
		return -EINVAL;

2995 2996
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2997 2998
		return -ENODEV;

2999
	idev = __in6_dev_get(dev);
3000
	if (!idev)
L
Linus Torvalds 已提交
3001 3002 3003
		return -ENXIO;

	read_lock_bh(&idev->lock);
3004
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
L
Linus Torvalds 已提交
3005 3006 3007 3008
		if (ifp->prefix_len == plen &&
		    ipv6_addr_equal(pfx, &ifp->addr)) {
			in6_ifa_hold(ifp);
			read_unlock_bh(&idev->lock);
3009

3010 3011 3012 3013
			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
3014
			ipv6_del_addr(ifp);
3015
			addrconf_verify_rtnl();
3016 3017 3018 3019
			if (ipv6_addr_is_multicast(pfx)) {
				ipv6_mc_config(net->ipv6.mc_autojoin_sk,
					       false, pfx, dev->ifindex);
			}
L
Linus Torvalds 已提交
3020 3021 3022 3023 3024 3025 3026 3027
			return 0;
		}
	}
	read_unlock_bh(&idev->lock);
	return -EADDRNOTAVAIL;
}


3028
int addrconf_add_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
3029
{
3030 3031 3032 3033 3034
	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
3035 3036
	struct in6_ifreq ireq;
	int err;
3037

3038
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
3039
		return -EPERM;
3040

L
Linus Torvalds 已提交
3041 3042 3043
	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		return -EFAULT;

3044 3045 3046
	cfg.pfx = &ireq.ifr6_addr;
	cfg.plen = ireq.ifr6_prefixlen;

L
Linus Torvalds 已提交
3047
	rtnl_lock();
3048
	err = inet6_addr_add(net, ireq.ifr6_ifindex, &cfg, NULL);
L
Linus Torvalds 已提交
3049 3050 3051 3052
	rtnl_unlock();
	return err;
}

3053
int addrconf_del_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
3054 3055 3056
{
	struct in6_ifreq ireq;
	int err;
3057

3058
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
3059 3060 3061 3062 3063 3064
		return -EPERM;

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

	rtnl_lock();
3065
	err = inet6_addr_del(net, ireq.ifr6_ifindex, 0, &ireq.ifr6_addr,
3066
			     ireq.ifr6_prefixlen);
L
Linus Torvalds 已提交
3067 3068 3069 3070
	rtnl_unlock();
	return err;
}

3071 3072 3073 3074
static void add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
		     int plen, int scope)
{
	struct inet6_ifaddr *ifp;
3075 3076 3077 3078 3079 3080 3081 3082
	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
	};
3083

3084
	ifp = ipv6_add_addr(idev, &cfg, true, NULL);
3085 3086 3087 3088
	if (!IS_ERR(ifp)) {
		spin_lock_bh(&ifp->lock);
		ifp->flags &= ~IFA_F_TENTATIVE;
		spin_unlock_bh(&ifp->lock);
3089
		rt_genid_bump_ipv6(dev_net(idev->dev));
3090 3091 3092 3093 3094
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
		in6_ifa_put(ifp);
	}
}

A
Amerigo Wang 已提交
3095
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3096 3097 3098 3099
static void sit_add_v4_addrs(struct inet6_dev *idev)
{
	struct in6_addr addr;
	struct net_device *dev;
3100
	struct net *net = dev_net(idev->dev);
3101
	int scope, plen;
3102
	u32 pflags = 0;
L
Linus Torvalds 已提交
3103 3104 3105 3106 3107 3108 3109 3110 3111

	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;
3112
		plen = 64;
L
Linus Torvalds 已提交
3113 3114
	} else {
		scope = IPV6_ADDR_COMPATv4;
3115
		plen = 96;
3116
		pflags |= RTF_NONEXTHOP;
L
Linus Torvalds 已提交
3117 3118 3119
	}

	if (addr.s6_addr32[3]) {
3120
		add_addr(idev, &addr, plen, scope);
3121
		addrconf_prefix_route(&addr, plen, 0, idev->dev, 0, pflags,
3122
				      GFP_KERNEL);
L
Linus Torvalds 已提交
3123 3124 3125
		return;
	}

3126
	for_each_netdev(net, dev) {
3127
		struct in_device *in_dev = __in_dev_get_rtnl(dev);
L
Linus Torvalds 已提交
3128
		if (in_dev && (dev->flags & IFF_UP)) {
3129
			struct in_ifaddr *ifa;
L
Linus Torvalds 已提交
3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144

			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;
				}

3145
				add_addr(idev, &addr, plen, flag);
3146
				addrconf_prefix_route(&addr, plen, 0, idev->dev,
3147
						      0, pflags, GFP_KERNEL);
L
Linus Torvalds 已提交
3148 3149
			}
		}
3150
	}
L
Linus Torvalds 已提交
3151
}
3152
#endif
L
Linus Torvalds 已提交
3153 3154 3155 3156 3157 3158 3159 3160 3161

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

	/* ::1 */

	ASSERT_RTNL();

3162
	idev = ipv6_find_idev(dev);
3163
	if (!idev) {
3164
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
3165 3166 3167
		return;
	}

3168
	add_addr(idev, &in6addr_loopback, 128, IFA_HOST);
L
Linus Torvalds 已提交
3169 3170
}

3171 3172
void addrconf_add_linklocal(struct inet6_dev *idev,
			    const struct in6_addr *addr, u32 flags)
L
Linus Torvalds 已提交
3173
{
3174 3175 3176 3177 3178 3179 3180 3181
	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
	};
3182
	struct inet6_ifaddr *ifp;
3183 3184

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
3185 3186
	if ((dev_net(idev->dev)->ipv6.devconf_all->optimistic_dad ||
	     idev->cnf.optimistic_dad) &&
3187
	    !dev_net(idev->dev)->ipv6.devconf_all->forwarding)
3188
		cfg.ifa_flags |= IFA_F_OPTIMISTIC;
3189
#endif
L
Linus Torvalds 已提交
3190

3191
	ifp = ipv6_add_addr(idev, &cfg, true, NULL);
L
Linus Torvalds 已提交
3192
	if (!IS_ERR(ifp)) {
3193
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, 0, idev->dev,
3194
				      0, 0, GFP_ATOMIC);
3195
		addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
3196 3197 3198
		in6_ifa_put(ifp);
	}
}
3199
EXPORT_SYMBOL_GPL(addrconf_add_linklocal);
L
Linus Torvalds 已提交
3200

3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228
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;
3229
			__be32 prefix[2];
3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254
			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);
3255 3256
	data.prefix[0] = address->s6_addr32[0];
	data.prefix[1] = address->s6_addr32[1];
3257 3258 3259 3260 3261 3262
	data.secret = secret;
	data.dad_count = dad_count;

	sha_transform(digest, data.__data, workspace);

	temp = *address;
3263 3264
	temp.s6_addr32[2] = (__force __be32)digest[0];
	temp.s6_addr32[3] = (__force __be32)digest[1];
3265 3266 3267 3268 3269

	spin_unlock_bh(&lock);

	if (ipv6_reserved_interfaceid(temp)) {
		dad_count++;
3270
		if (dad_count > dev_net(idev->dev)->ipv6.sysctl.idgen_retries)
3271 3272 3273 3274 3275 3276 3277 3278
			return -1;
		goto retry;
	}

	*address = temp;
	return 0;
}

3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289
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;
}

3290 3291
static void addrconf_addr_gen(struct inet6_dev *idev, bool prefix_route)
{
3292 3293
	struct in6_addr addr;

D
David Ahern 已提交
3294 3295 3296 3297
	/* no link local addresses on L3 master devices */
	if (netif_is_l3_master(idev->dev))
		return;

3298
	ipv6_addr_set(&addr, htonl(0xFE800000), 0, 0, 0);
3299

3300
	switch (idev->cnf.addr_gen_mode) {
3301 3302 3303 3304
	case IN6_ADDR_GEN_MODE_RANDOM:
		ipv6_gen_mode_random_init(idev);
		/* fallthrough */
	case IN6_ADDR_GEN_MODE_STABLE_PRIVACY:
3305
		if (!ipv6_generate_stable_address(&addr, 0, idev))
3306 3307
			addrconf_add_linklocal(idev, &addr,
					       IFA_F_STABLE_PRIVACY);
3308
		else if (prefix_route)
3309
			addrconf_prefix_route(&addr, 64, 0, idev->dev,
3310
					      0, 0, GFP_KERNEL);
3311 3312
		break;
	case IN6_ADDR_GEN_MODE_EUI64:
3313 3314 3315 3316 3317
		/* 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)
3318
			addrconf_add_linklocal(idev, &addr, 0);
3319
		else if (prefix_route)
3320
			addrconf_prefix_route(&addr, 64, 0, idev->dev,
3321
					      0, 0, GFP_KERNEL);
3322 3323 3324 3325 3326
		break;
	case IN6_ADDR_GEN_MODE_NONE:
	default:
		/* will not add any link local address */
		break;
3327 3328 3329
	}
}

L
Linus Torvalds 已提交
3330 3331
static void addrconf_dev_config(struct net_device *dev)
{
3332
	struct inet6_dev *idev;
L
Linus Torvalds 已提交
3333 3334 3335

	ASSERT_RTNL();

3336 3337 3338
	if ((dev->type != ARPHRD_ETHER) &&
	    (dev->type != ARPHRD_FDDI) &&
	    (dev->type != ARPHRD_ARCNET) &&
3339
	    (dev->type != ARPHRD_INFINIBAND) &&
3340
	    (dev->type != ARPHRD_IEEE1394) &&
3341
	    (dev->type != ARPHRD_TUNNEL6) &&
3342
	    (dev->type != ARPHRD_6LOWPAN) &&
3343 3344 3345
	    (dev->type != ARPHRD_IP6GRE) &&
	    (dev->type != ARPHRD_IPGRE) &&
	    (dev->type != ARPHRD_TUNNEL) &&
3346 3347
	    (dev->type != ARPHRD_NONE) &&
	    (dev->type != ARPHRD_RAWIP)) {
3348 3349 3350 3351
		/* Alas, we support only Ethernet autoconfiguration. */
		return;
	}

L
Linus Torvalds 已提交
3352
	idev = addrconf_add_dev(dev);
3353
	if (IS_ERR(idev))
L
Linus Torvalds 已提交
3354 3355
		return;

3356 3357
	/* this device type has no EUI support */
	if (dev->type == ARPHRD_NONE &&
3358 3359
	    idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_EUI64)
		idev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_RANDOM;
3360

3361
	addrconf_addr_gen(idev, false);
L
Linus Torvalds 已提交
3362 3363
}

A
Amerigo Wang 已提交
3364
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3365 3366 3367 3368 3369 3370
static void addrconf_sit_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3371 3372 3373
	/*
	 * Configure the tunnel with one of our IPv4
	 * addresses... we should configure all of
L
Linus Torvalds 已提交
3374 3375 3376
	 * our v4 addrs in the tunnel
	 */

3377
	idev = ipv6_find_idev(dev);
3378
	if (!idev) {
3379
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
3380 3381 3382
		return;
	}

F
Fred L. Templin 已提交
3383
	if (dev->priv_flags & IFF_ISATAP) {
3384
		addrconf_addr_gen(idev, false);
F
Fred L. Templin 已提交
3385 3386 3387
		return;
	}

L
Linus Torvalds 已提交
3388 3389
	sit_add_v4_addrs(idev);

3390
	if (dev->flags&IFF_POINTOPOINT)
L
Linus Torvalds 已提交
3391 3392
		addrconf_add_mroute(dev);
}
3393
#endif
L
Linus Torvalds 已提交
3394

A
Amerigo Wang 已提交
3395
#if IS_ENABLED(CONFIG_NET_IPGRE)
3396 3397 3398 3399 3400 3401
static void addrconf_gre_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3402
	idev = ipv6_find_idev(dev);
3403
	if (!idev) {
3404
		pr_debug("%s: add_dev failed\n", __func__);
3405 3406 3407
		return;
	}

3408
	addrconf_addr_gen(idev, true);
3409 3410
	if (dev->flags & IFF_POINTOPOINT)
		addrconf_add_mroute(dev);
3411 3412 3413
}
#endif

3414 3415
static int fixup_permanent_addr(struct net *net,
				struct inet6_dev *idev,
3416 3417
				struct inet6_ifaddr *ifp)
{
3418
	/* !fib6_node means the host route was removed from the
3419 3420 3421
	 * FIB, for example, if 'lo' device is taken down. In that
	 * case regenerate the host route.
	 */
3422
	if (!ifp->rt || !ifp->rt->fib6_node) {
3423
		struct fib6_info *f6i, *prev;
3424

3425 3426 3427 3428
		f6i = addrconf_f6i_alloc(net, idev, &ifp->addr, false,
					 GFP_ATOMIC);
		if (IS_ERR(f6i))
			return PTR_ERR(f6i);
3429

3430 3431 3432
		/* ifp->rt can be accessed outside of rtnl */
		spin_lock(&ifp->lock);
		prev = ifp->rt;
3433
		ifp->rt = f6i;
3434 3435
		spin_unlock(&ifp->lock);

3436
		fib6_info_release(prev);
3437 3438 3439 3440
	}

	if (!(ifp->flags & IFA_F_NOPREFIXROUTE)) {
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len,
3441 3442
				      ifp->rt_priority, idev->dev, 0, 0,
				      GFP_ATOMIC);
3443 3444
	}

3445 3446
	if (ifp->state == INET6_IFADDR_STATE_PREDAD)
		addrconf_dad_start(ifp);
3447 3448 3449 3450

	return 0;
}

3451
static void addrconf_permanent_addr(struct net *net, struct net_device *dev)
3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463
{
	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) &&
3464
		    fixup_permanent_addr(net, idev, ifp) < 0) {
3465
			write_unlock_bh(&idev->lock);
3466
			in6_ifa_hold(ifp);
3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477
			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);
}

3478
static int addrconf_notify(struct notifier_block *this, unsigned long event,
3479
			   void *ptr)
L
Linus Torvalds 已提交
3480
{
3481
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
3482
	struct netdev_notifier_change_info *change_info;
3483
	struct netdev_notifier_changeupper_info *info;
3484
	struct inet6_dev *idev = __in6_dev_get(dev);
3485
	struct net *net = dev_net(dev);
3486
	int run_pending = 0;
3487
	int err;
L
Linus Torvalds 已提交
3488

S
Stephen Hemminger 已提交
3489
	switch (event) {
3490
	case NETDEV_REGISTER:
3491
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
3492
			idev = ipv6_add_dev(dev);
3493 3494
			if (IS_ERR(idev))
				return notifier_from_errno(PTR_ERR(idev));
3495 3496
		}
		break;
S
Stephen Hemminger 已提交
3497

3498 3499 3500
	case NETDEV_CHANGEMTU:
		/* if MTU under IPV6_MIN_MTU stop IPv6 on this interface. */
		if (dev->mtu < IPV6_MIN_MTU) {
3501
			addrconf_ifdown(dev, dev != net->loopback_dev);
3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523
			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
3524
	case NETDEV_UP:
3525
	case NETDEV_CHANGE:
3526 3527 3528
		if (dev->flags & IFF_SLAVE)
			break;

3529 3530 3531
		if (idev && idev->cnf.disable_ipv6)
			break;

3532
		if (event == NETDEV_UP) {
3533
			/* restore routes for permanent addresses */
3534
			addrconf_permanent_addr(net, dev);
3535

3536
			if (!addrconf_link_ready(dev)) {
3537
				/* device is not ready yet. */
3538 3539
				pr_debug("ADDRCONF(NETDEV_UP): %s: link is not ready\n",
					 dev->name);
3540 3541
				break;
			}
3542

3543 3544 3545
			if (!idev && dev->mtu >= IPV6_MIN_MTU)
				idev = ipv6_add_dev(dev);

3546
			if (!IS_ERR_OR_NULL(idev)) {
3547
				idev->if_flags |= IF_READY;
3548 3549
				run_pending = 1;
			}
3550
		} else if (event == NETDEV_CHANGE) {
3551
			if (!addrconf_link_ready(dev)) {
3552
				/* device is still not ready. */
3553
				rt6_sync_down_dev(dev, event);
3554 3555 3556
				break;
			}

3557
			if (!IS_ERR_OR_NULL(idev)) {
3558 3559 3560 3561 3562 3563 3564
				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);
3565 3566 3567
					change_info = ptr;
					if (change_info->flags_changed & IFF_NOARP)
						addrconf_dad_run(idev, true);
3568
					rt6_sync_up(dev, RTNH_F_LINKDOWN);
3569
					break;
3570
				}
3571 3572 3573
				idev->if_flags |= IF_READY;
			}

3574 3575
			pr_info("ADDRCONF(NETDEV_CHANGE): %s: link becomes ready\n",
				dev->name);
3576

3577
			run_pending = 1;
3578 3579
		}

S
Stephen Hemminger 已提交
3580
		switch (dev->type) {
A
Amerigo Wang 已提交
3581
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3582 3583 3584
		case ARPHRD_SIT:
			addrconf_sit_config(dev);
			break;
3585
#endif
A
Amerigo Wang 已提交
3586
#if IS_ENABLED(CONFIG_NET_IPGRE)
3587 3588 3589
		case ARPHRD_IPGRE:
			addrconf_gre_config(dev);
			break;
3590
#endif
L
Linus Torvalds 已提交
3591 3592 3593 3594 3595 3596 3597
		case ARPHRD_LOOPBACK:
			init_loopback(dev);
			break;

		default:
			addrconf_dev_config(dev);
			break;
3598
		}
S
Stephen Hemminger 已提交
3599

3600
		if (!IS_ERR_OR_NULL(idev)) {
3601
			if (run_pending)
3602
				addrconf_dad_run(idev, false);
3603

3604 3605 3606
			/* Device has an address by now */
			rt6_sync_up(dev, RTNH_F_DEAD);

S
Stephen Hemminger 已提交
3607 3608 3609 3610
			/*
			 * 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
3611
			 */
S
Stephen Hemminger 已提交
3612 3613
			if (idev->cnf.mtu6 != dev->mtu &&
			    dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
3614 3615 3616 3617 3618
				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
3619 3620 3621 3622

			/*
			 * If the changed mtu during down is lower than
			 * IPV6_MIN_MTU stop IPv6 on this interface.
L
Linus Torvalds 已提交
3623 3624
			 */
			if (dev->mtu < IPV6_MIN_MTU)
3625
				addrconf_ifdown(dev, dev != net->loopback_dev);
L
Linus Torvalds 已提交
3626 3627 3628 3629 3630 3631 3632 3633 3634 3635
		}
		break;

	case NETDEV_DOWN:
	case NETDEV_UNREGISTER:
		/*
		 *	Remove all addresses from this interface.
		 */
		addrconf_ifdown(dev, event != NETDEV_DOWN);
		break;
3636

L
Linus Torvalds 已提交
3637 3638
	case NETDEV_CHANGENAME:
		if (idev) {
3639
			snmp6_unregister_dev(idev);
3640
			addrconf_sysctl_unregister(idev);
3641
			err = addrconf_sysctl_register(idev);
3642 3643
			if (err)
				return notifier_from_errno(err);
3644 3645 3646 3647 3648
			err = snmp6_register_dev(idev);
			if (err) {
				addrconf_sysctl_unregister(idev);
				return notifier_from_errno(err);
			}
3649
		}
L
Linus Torvalds 已提交
3650
		break;
S
Stephen Hemminger 已提交
3651

3652 3653
	case NETDEV_PRE_TYPE_CHANGE:
	case NETDEV_POST_TYPE_CHANGE:
3654 3655
		if (idev)
			addrconf_type_change(dev, event);
3656
		break;
3657 3658 3659 3660 3661 3662 3663 3664 3665

	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);
3666
	}
L
Linus Torvalds 已提交
3667 3668 3669 3670 3671 3672 3673 3674 3675

	return NOTIFY_OK;
}

/*
 *	addrconf module should be notified of a device going up
 */
static struct notifier_block ipv6_dev_notf = {
	.notifier_call = addrconf_notify,
3676
	.priority = ADDRCONF_NOTIFY_PRIORITY,
L
Linus Torvalds 已提交
3677 3678
};

3679
static void addrconf_type_change(struct net_device *dev, unsigned long event)
3680 3681 3682 3683 3684 3685
{
	struct inet6_dev *idev;
	ASSERT_RTNL();

	idev = __in6_dev_get(dev);

3686
	if (event == NETDEV_POST_TYPE_CHANGE)
3687
		ipv6_mc_remap(idev);
3688
	else if (event == NETDEV_PRE_TYPE_CHANGE)
3689 3690 3691
		ipv6_mc_unmap(idev);
}

3692 3693 3694 3695 3696 3697
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
3698 3699
static int addrconf_ifdown(struct net_device *dev, int how)
{
3700
	unsigned long event = how ? NETDEV_UNREGISTER : NETDEV_DOWN;
3701
	struct net *net = dev_net(dev);
3702
	struct inet6_dev *idev;
3703
	struct inet6_ifaddr *ifa, *tmp;
3704
	bool keep_addr = false;
3705
	int state, i;
L
Linus Torvalds 已提交
3706 3707 3708

	ASSERT_RTNL();

3709
	rt6_disable_ip(dev, event);
L
Linus Torvalds 已提交
3710 3711

	idev = __in6_dev_get(dev);
3712
	if (!idev)
L
Linus Torvalds 已提交
3713 3714
		return -ENODEV;

S
Stephen Hemminger 已提交
3715 3716 3717
	/*
	 * Step 1: remove reference to ipv6 device from parent device.
	 *	   Do not dev_put!
L
Linus Torvalds 已提交
3718
	 */
3719
	if (how) {
L
Linus Torvalds 已提交
3720
		idev->dead = 1;
3721 3722

		/* protected by rtnl_lock */
3723
		RCU_INIT_POINTER(dev->ip6_ptr, NULL);
L
Linus Torvalds 已提交
3724 3725 3726 3727 3728 3729

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

	}

3730 3731 3732
	/* combine the user config with event to determine if permanent
	 * addresses are to be removed from address hash table
	 */
3733 3734 3735 3736 3737 3738 3739 3740 3741
	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);
	}
3742

3743 3744 3745 3746 3747
	/* 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);
3748
restart:
3749
		hlist_for_each_entry_rcu(ifa, h, addr_lst) {
3750
			if (ifa->idev == idev) {
3751
				addrconf_del_dad_work(ifa);
3752 3753 3754 3755
				/* combined flag + permanent flag decide if
				 * address is retained on a down event
				 */
				if (!keep_addr ||
3756 3757
				    !(ifa->flags & IFA_F_PERMANENT) ||
				    addr_is_local(&ifa->addr)) {
3758 3759 3760
					hlist_del_init_rcu(&ifa->addr_lst);
					goto restart;
				}
3761 3762 3763 3764 3765
			}
		}
		spin_unlock_bh(&addrconf_hash_lock);
	}

L
Linus Torvalds 已提交
3766 3767
	write_lock_bh(&idev->lock);

3768 3769
	addrconf_del_rs_timer(idev);

S
Stephen Hemminger 已提交
3770
	/* Step 2: clear flags for stateless addrconf */
3771
	if (!how)
3772
		idev->if_flags &= ~(IF_RS_SENT|IF_RA_RCVD|IF_READY);
L
Linus Torvalds 已提交
3773

S
Stephen Hemminger 已提交
3774
	/* Step 3: clear tempaddr list */
3775 3776 3777 3778
	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
3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789
		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);
	}
3790

3791
	list_for_each_entry_safe(ifa, tmp, &idev->addr_list, if_list) {
3792
		struct fib6_info *rt = NULL;
3793
		bool keep;
3794

3795
		addrconf_del_dad_work(ifa);
3796

3797 3798 3799
		keep = keep_addr && (ifa->flags & IFA_F_PERMANENT) &&
			!addr_is_local(&ifa->addr);

3800
		write_unlock_bh(&idev->lock);
3801
		spin_lock_bh(&ifa->lock);
3802

3803
		if (keep) {
3804 3805
			/* set state to skip the notifier below */
			state = INET6_IFADDR_STATE_DEAD;
3806
			ifa->state = INET6_IFADDR_STATE_PREDAD;
3807 3808
			if (!(ifa->flags & IFA_F_NODAD))
				ifa->flags |= IFA_F_TENTATIVE;
3809 3810 3811

			rt = ifa->rt;
			ifa->rt = NULL;
3812 3813 3814 3815 3816
		} else {
			state = ifa->state;
			ifa->state = INET6_IFADDR_STATE_DEAD;
		}

3817
		spin_unlock_bh(&ifa->lock);
3818

3819
		if (rt)
3820
			ip6_del_rt(net, rt);
3821

3822 3823
		if (state != INET6_IFADDR_STATE_DEAD) {
			__ipv6_ifa_notify(RTM_DELADDR, ifa);
3824
			inet6addr_notifier_call_chain(NETDEV_DOWN, ifa);
3825 3826 3827 3828
		} else {
			if (idev->cnf.forwarding)
				addrconf_leave_anycast(ifa);
			addrconf_leave_solict(ifa->idev, &ifa->addr);
3829
		}
3830

3831
		write_lock_bh(&idev->lock);
3832 3833 3834 3835
		if (!keep) {
			list_del_rcu(&ifa->if_list);
			in6_ifa_put(ifa);
		}
3836
	}
3837

L
Linus Torvalds 已提交
3838 3839
	write_unlock_bh(&idev->lock);

3840 3841 3842
	/* Step 5: Discard anycast and multicast list */
	if (how) {
		ipv6_ac_destroy_dev(idev);
L
Linus Torvalds 已提交
3843
		ipv6_mc_destroy_dev(idev);
3844
	} else {
L
Linus Torvalds 已提交
3845
		ipv6_mc_down(idev);
3846
	}
L
Linus Torvalds 已提交
3847 3848

	idev->tstamp = jiffies;
3849

S
Stephen Hemminger 已提交
3850
	/* Last: Shot the device (if unregistered) */
3851
	if (how) {
3852
		addrconf_sysctl_unregister(idev);
L
Linus Torvalds 已提交
3853 3854 3855 3856 3857 3858 3859
		neigh_parms_release(&nd_tbl, idev->nd_parms);
		neigh_ifdown(&nd_tbl, dev);
		in6_dev_put(idev);
	}
	return 0;
}

3860
static void addrconf_rs_timer(struct timer_list *t)
L
Linus Torvalds 已提交
3861
{
3862
	struct inet6_dev *idev = from_timer(idev, t, rs_timer);
3863
	struct net_device *dev = idev->dev;
3864
	struct in6_addr lladdr;
L
Linus Torvalds 已提交
3865

3866
	write_lock(&idev->lock);
S
stephen hemminger 已提交
3867
	if (idev->dead || !(idev->if_flags & IF_READY))
L
Linus Torvalds 已提交
3868 3869
		goto out;

3870
	if (!ipv6_accept_ra(idev))
S
stephen hemminger 已提交
3871 3872 3873 3874
		goto out;

	/* Announcement received after solicitation was sent */
	if (idev->if_flags & IF_RA_RCVD)
L
Linus Torvalds 已提交
3875 3876
		goto out;

3877
	if (idev->rs_probes++ < idev->cnf.rtr_solicits || idev->cnf.rtr_solicits < 0) {
3878 3879 3880
		write_unlock(&idev->lock);
		if (!ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
			ndisc_send_rs(dev, &lladdr,
3881 3882
				      &in6addr_linklocal_allrouters);
		else
3883
			goto put;
L
Linus Torvalds 已提交
3884

3885
		write_lock(&idev->lock);
3886 3887
		idev->rs_interval = rfc3315_s14_backoff_update(
			idev->rs_interval, idev->cnf.rtr_solicit_max_interval);
3888 3889 3890 3891
		/* 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 :
3892
				      idev->rs_interval);
L
Linus Torvalds 已提交
3893 3894 3895 3896 3897
	} else {
		/*
		 * Note: we do not support deprecated "all on-link"
		 * assumption any longer.
		 */
3898
		pr_debug("%s: no IPv6 routers present\n", idev->dev->name);
L
Linus Torvalds 已提交
3899 3900 3901
	}

out:
3902
	write_unlock(&idev->lock);
3903
put:
3904
	in6_dev_put(idev);
L
Linus Torvalds 已提交
3905 3906 3907 3908 3909
}

/*
 *	Duplicate Address Detection
 */
3910 3911 3912 3913
static void addrconf_dad_kick(struct inet6_ifaddr *ifp)
{
	unsigned long rand_num;
	struct inet6_dev *idev = ifp->idev;
3914
	u64 nonce;
3915

3916 3917 3918
	if (ifp->flags & IFA_F_OPTIMISTIC)
		rand_num = 0;
	else
3919
		rand_num = prandom_u32() % (idev->cnf.rtr_solicit_delay ? : 1);
3920

3921 3922 3923 3924 3925 3926 3927 3928
	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;
3929
	ifp->dad_probes = idev->cnf.dad_transmits;
3930
	addrconf_mod_dad_work(ifp, rand_num);
3931 3932
}

3933
static void addrconf_dad_begin(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3934 3935 3936
{
	struct inet6_dev *idev = ifp->idev;
	struct net_device *dev = idev->dev;
3937
	bool bump_id, notify = false;
3938
	struct net *net;
L
Linus Torvalds 已提交
3939 3940 3941

	addrconf_join_solict(dev, &ifp->addr);

3942
	prandom_seed((__force u32) ifp->addr.s6_addr32[3]);
L
Linus Torvalds 已提交
3943 3944

	read_lock_bh(&idev->lock);
3945
	spin_lock(&ifp->lock);
3946
	if (ifp->state == INET6_IFADDR_STATE_DEAD)
L
Linus Torvalds 已提交
3947 3948
		goto out;

3949
	net = dev_net(dev);
L
Linus Torvalds 已提交
3950
	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
3951
	    (net->ipv6.devconf_all->accept_dad < 1 &&
3952
	     idev->cnf.accept_dad < 1) ||
3953 3954
	    !(ifp->flags&IFA_F_TENTATIVE) ||
	    ifp->flags & IFA_F_NODAD) {
3955 3956 3957 3958 3959
		bool send_na = false;

		if (ifp->flags & IFA_F_TENTATIVE &&
		    !(ifp->flags & IFA_F_OPTIMISTIC))
			send_na = true;
3960
		bump_id = ifp->flags & IFA_F_TENTATIVE;
B
Brian Haley 已提交
3961
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3962
		spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3963 3964
		read_unlock_bh(&idev->lock);

3965
		addrconf_dad_completed(ifp, bump_id, send_na);
L
Linus Torvalds 已提交
3966 3967 3968
		return;
	}

3969
	if (!(idev->if_flags & IF_READY)) {
3970
		spin_unlock(&ifp->lock);
3971
		read_unlock_bh(&idev->lock);
3972
		/*
3973
		 * If the device is not ready:
3974 3975 3976 3977
		 * - keep it tentative if it is a permanent address.
		 * - otherwise, kill it.
		 */
		in6_ifa_hold(ifp);
B
Brian Haley 已提交
3978
		addrconf_dad_stop(ifp, 0);
3979
		return;
3980
	}
3981 3982 3983 3984 3985

	/*
	 * Optimistic nodes can start receiving
	 * Frames right away
	 */
3986
	if (ifp->flags & IFA_F_OPTIMISTIC) {
3987 3988
		ip6_ins_rt(net, ifp->rt);
		if (ipv6_use_optimistic_addr(net, idev)) {
3989 3990 3991
			/* Because optimistic nodes can use this address,
			 * notify listeners. If DAD fails, RTM_DELADDR is sent.
			 */
3992
			notify = true;
3993 3994
		}
	}
3995

3996
	addrconf_dad_kick(ifp);
L
Linus Torvalds 已提交
3997
out:
3998
	spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3999
	read_unlock_bh(&idev->lock);
4000 4001
	if (notify)
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
L
Linus Torvalds 已提交
4002 4003
}

4004
static void addrconf_dad_start(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
4005
{
4006 4007
	bool begin_dad = false;

4008
	spin_lock_bh(&ifp->lock);
4009 4010 4011 4012
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
		ifp->state = INET6_IFADDR_STATE_PREDAD;
		begin_dad = true;
	}
4013
	spin_unlock_bh(&ifp->lock);
4014 4015 4016 4017 4018 4019 4020 4021 4022 4023

	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
4024
	struct inet6_dev *idev = ifp->idev;
4025
	bool bump_id, disable_ipv6 = false;
L
Linus Torvalds 已提交
4026 4027
	struct in6_addr mcaddr;

4028 4029 4030 4031 4032 4033 4034 4035
	enum {
		DAD_PROCESS,
		DAD_BEGIN,
		DAD_ABORT,
	} action = DAD_PROCESS;

	rtnl_lock();

4036
	spin_lock_bh(&ifp->lock);
4037 4038 4039 4040 4041 4042
	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;
4043

4044 4045 4046
		if ((dev_net(idev->dev)->ipv6.devconf_all->accept_dad > 1 ||
		     idev->cnf.accept_dad > 1) &&
		    !idev->cnf.disable_ipv6 &&
4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062
		    !(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;
			}
		}
4063
	}
4064
	spin_unlock_bh(&ifp->lock);
4065 4066 4067 4068 4069

	if (action == DAD_BEGIN) {
		addrconf_dad_begin(ifp);
		goto out;
	} else if (action == DAD_ABORT) {
4070
		in6_ifa_hold(ifp);
4071
		addrconf_dad_stop(ifp, 1);
4072 4073
		if (disable_ipv6)
			addrconf_ifdown(idev->dev, 0);
4074 4075 4076
		goto out;
	}

4077
	if (!ifp->dad_probes && addrconf_dad_end(ifp))
H
Herbert Xu 已提交
4078 4079
		goto out;

4080
	write_lock_bh(&idev->lock);
4081
	if (idev->dead || !(idev->if_flags & IF_READY)) {
4082
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
4083 4084
		goto out;
	}
4085 4086

	spin_lock(&ifp->lock);
4087 4088
	if (ifp->state == INET6_IFADDR_STATE_DEAD) {
		spin_unlock(&ifp->lock);
4089
		write_unlock_bh(&idev->lock);
4090 4091 4092
		goto out;
	}

4093
	if (ifp->dad_probes == 0) {
4094 4095
		bool send_na = false;

L
Linus Torvalds 已提交
4096 4097 4098 4099
		/*
		 * DAD was successful
		 */

4100 4101 4102
		if (ifp->flags & IFA_F_TENTATIVE &&
		    !(ifp->flags & IFA_F_OPTIMISTIC))
			send_na = true;
4103
		bump_id = ifp->flags & IFA_F_TENTATIVE;
B
Brian Haley 已提交
4104
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
4105
		spin_unlock(&ifp->lock);
4106
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
4107

4108
		addrconf_dad_completed(ifp, bump_id, send_na);
L
Linus Torvalds 已提交
4109 4110 4111 4112

		goto out;
	}

4113
	ifp->dad_probes--;
4114 4115
	addrconf_mod_dad_work(ifp,
			      NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME));
4116
	spin_unlock(&ifp->lock);
4117
	write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
4118 4119 4120

	/* send a neighbour solicitation for our addr */
	addrconf_addr_solict_mult(&ifp->addr, &mcaddr);
4121 4122
	ndisc_send_ns(ifp->idev->dev, &ifp->addr, &mcaddr, &in6addr_any,
		      ifp->dad_nonce);
L
Linus Torvalds 已提交
4123 4124
out:
	in6_ifa_put(ifp);
4125
	rtnl_unlock();
L
Linus Torvalds 已提交
4126 4127
}

4128 4129 4130 4131 4132 4133
/* 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;

4134 4135 4136
	list_for_each_entry_reverse(ifpiter, &idev->addr_list, if_list) {
		if (ifpiter->scope > IFA_LINK)
			break;
4137 4138 4139 4140 4141 4142 4143 4144 4145
		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;
}

4146 4147
static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id,
				   bool send_na)
L
Linus Torvalds 已提交
4148
{
S
Stephen Hemminger 已提交
4149
	struct net_device *dev = ifp->idev->dev;
4150
	struct in6_addr lladdr;
4151
	bool send_rs, send_mld;
4152

4153
	addrconf_del_dad_work(ifp);
L
Linus Torvalds 已提交
4154 4155 4156 4157 4158 4159 4160

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

	ipv6_ifa_notify(RTM_NEWADDR, ifp);

4161 4162
	/* If added prefix is link local and we are prepared to process
	   router advertisements, start sending router solicitations.
L
Linus Torvalds 已提交
4163 4164
	 */

4165
	read_lock_bh(&ifp->idev->lock);
4166
	send_mld = ifp->scope == IFA_LINK && ipv6_lonely_lladdr(ifp);
4167 4168
	send_rs = send_mld &&
		  ipv6_accept_ra(ifp->idev) &&
4169
		  ifp->idev->cnf.rtr_solicits != 0 &&
4170
		  (dev->flags&IFF_LOOPBACK) == 0;
4171 4172
	read_unlock_bh(&ifp->idev->lock);

4173 4174 4175 4176 4177 4178
	/* 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);

4179 4180 4181 4182 4183 4184 4185 4186 4187 4188
	/* 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);
	}

4189
	if (send_rs) {
L
Linus Torvalds 已提交
4190 4191 4192 4193 4194
		/*
		 *	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
		 */
4195
		if (ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
4196
			return;
4197
		ndisc_send_rs(dev, &lladdr, &in6addr_linklocal_allrouters);
L
Linus Torvalds 已提交
4198

4199 4200
		write_lock_bh(&ifp->idev->lock);
		spin_lock(&ifp->lock);
4201 4202
		ifp->idev->rs_interval = rfc3315_s14_backoff_init(
			ifp->idev->cnf.rtr_solicit_interval);
4203
		ifp->idev->rs_probes = 1;
L
Linus Torvalds 已提交
4204
		ifp->idev->if_flags |= IF_RS_SENT;
4205
		addrconf_mod_rs_timer(ifp->idev, ifp->idev->rs_interval);
4206 4207
		spin_unlock(&ifp->lock);
		write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
4208
	}
4209 4210 4211

	if (bump_id)
		rt_genid_bump_ipv6(dev_net(dev));
4212 4213 4214 4215 4216 4217

	/* 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
4218 4219
}

4220
static void addrconf_dad_run(struct inet6_dev *idev, bool restart)
S
Stephen Hemminger 已提交
4221
{
4222 4223 4224
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
4225
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
4226
		spin_lock(&ifp->lock);
4227 4228 4229 4230
		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
4231
			addrconf_dad_kick(ifp);
4232
		}
4233
		spin_unlock(&ifp->lock);
4234 4235 4236 4237
	}
	read_unlock_bh(&idev->lock);
}

L
Linus Torvalds 已提交
4238 4239
#ifdef CONFIG_PROC_FS
struct if6_iter_state {
4240
	struct seq_net_private p;
L
Linus Torvalds 已提交
4241
	int bucket;
4242
	int offset;
L
Linus Torvalds 已提交
4243 4244
};

4245
static struct inet6_ifaddr *if6_get_first(struct seq_file *seq, loff_t pos)
L
Linus Torvalds 已提交
4246 4247
{
	struct if6_iter_state *state = seq->private;
4248
	struct net *net = seq_file_net(seq);
4249
	struct inet6_ifaddr *ifa = NULL;
4250
	int p = 0;
L
Linus Torvalds 已提交
4251

4252 4253 4254 4255 4256 4257 4258
	/* initial bucket if pos is 0 */
	if (pos == 0) {
		state->bucket = 0;
		state->offset = 0;
	}

	for (; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
4259
		hlist_for_each_entry_rcu(ifa, &inet6_addr_lst[state->bucket],
4260
					 addr_lst) {
4261 4262
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
4263 4264 4265 4266 4267
			/* sync with offset */
			if (p < state->offset) {
				p++;
				continue;
			}
4268
			return ifa;
4269 4270 4271 4272 4273
		}

		/* prepare for next bucket */
		state->offset = 0;
		p = 0;
L
Linus Torvalds 已提交
4274
	}
4275
	return NULL;
L
Linus Torvalds 已提交
4276 4277
}

4278 4279
static struct inet6_ifaddr *if6_get_next(struct seq_file *seq,
					 struct inet6_ifaddr *ifa)
L
Linus Torvalds 已提交
4280 4281
{
	struct if6_iter_state *state = seq->private;
4282
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
4283

4284
	hlist_for_each_entry_continue_rcu(ifa, addr_lst) {
4285 4286
		if (!net_eq(dev_net(ifa->idev->dev), net))
			continue;
4287
		state->offset++;
4288
		return ifa;
4289
	}
4290

4291
	state->offset = 0;
4292
	while (++state->bucket < IN6_ADDR_HSIZE) {
4293
		hlist_for_each_entry_rcu(ifa,
4294
				     &inet6_addr_lst[state->bucket], addr_lst) {
4295 4296 4297
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
			return ifa;
4298
		}
L
Linus Torvalds 已提交
4299
	}
4300

4301
	return NULL;
L
Linus Torvalds 已提交
4302 4303 4304
}

static void *if6_seq_start(struct seq_file *seq, loff_t *pos)
4305
	__acquires(rcu)
L
Linus Torvalds 已提交
4306
{
4307
	rcu_read_lock();
4308
	return if6_get_first(seq, *pos);
L
Linus Torvalds 已提交
4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320
}

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)
4321
	__releases(rcu)
L
Linus Torvalds 已提交
4322
{
4323
	rcu_read_unlock();
L
Linus Torvalds 已提交
4324 4325 4326 4327 4328
}

static int if6_seq_show(struct seq_file *seq, void *v)
{
	struct inet6_ifaddr *ifp = (struct inet6_ifaddr *)v;
4329
	seq_printf(seq, "%pi6 %02x %02x %02x %02x %8s\n",
4330
		   &ifp->addr,
L
Linus Torvalds 已提交
4331 4332 4333
		   ifp->idev->dev->ifindex,
		   ifp->prefix_len,
		   ifp->scope,
4334
		   (u8) ifp->flags,
L
Linus Torvalds 已提交
4335 4336 4337 4338
		   ifp->idev->dev->name);
	return 0;
}

4339
static const struct seq_operations if6_seq_ops = {
L
Linus Torvalds 已提交
4340 4341 4342 4343 4344 4345
	.start	= if6_seq_start,
	.next	= if6_seq_next,
	.show	= if6_seq_show,
	.stop	= if6_seq_stop,
};

4346
static int __net_init if6_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
4347
{
4348 4349
	if (!proc_create_net("if_inet6", 0444, net->proc_net, &if6_seq_ops,
			sizeof(struct if6_iter_state)))
L
Linus Torvalds 已提交
4350 4351 4352 4353
		return -ENOMEM;
	return 0;
}

4354
static void __net_exit if6_proc_net_exit(struct net *net)
4355
{
4356
	remove_proc_entry("if_inet6", net->proc_net);
4357 4358 4359
}

static struct pernet_operations if6_proc_net_ops = {
4360 4361
	.init = if6_proc_net_init,
	.exit = if6_proc_net_exit,
4362 4363 4364 4365 4366 4367 4368
};

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

L
Linus Torvalds 已提交
4369 4370
void if6_proc_exit(void)
{
4371
	unregister_pernet_subsys(&if6_proc_net_ops);
L
Linus Torvalds 已提交
4372 4373 4374
}
#endif	/* CONFIG_PROC_FS */

A
Amerigo Wang 已提交
4375
#if IS_ENABLED(CONFIG_IPV6_MIP6)
4376
/* Check if address is a home address configured on any interface. */
4377
int ipv6_chk_home_addr(struct net *net, const struct in6_addr *addr)
4378
{
4379
	unsigned int hash = inet6_addr_hash(net, addr);
4380
	struct inet6_ifaddr *ifp = NULL;
4381
	int ret = 0;
4382

4383 4384
	rcu_read_lock();
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
4385
		if (!net_eq(dev_net(ifp->idev->dev), net))
4386
			continue;
4387
		if (ipv6_addr_equal(&ifp->addr, addr) &&
4388 4389 4390 4391 4392
		    (ifp->flags & IFA_F_HOMEADDRESS)) {
			ret = 1;
			break;
		}
	}
4393
	rcu_read_unlock();
4394 4395 4396 4397
	return ret;
}
#endif

L
Linus Torvalds 已提交
4398 4399 4400 4401
/*
 *	Periodic address status verification
 */

4402
static void addrconf_verify_rtnl(void)
L
Linus Torvalds 已提交
4403
{
4404
	unsigned long now, next, next_sec, next_sched;
L
Linus Torvalds 已提交
4405 4406 4407
	struct inet6_ifaddr *ifp;
	int i;

4408 4409
	ASSERT_RTNL();

4410
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
4411
	now = jiffies;
4412
	next = round_jiffies_up(now + ADDR_CHECK_FREQUENCY);
L
Linus Torvalds 已提交
4413

4414
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
4415

S
Stephen Hemminger 已提交
4416
	for (i = 0; i < IN6_ADDR_HSIZE; i++) {
L
Linus Torvalds 已提交
4417
restart:
4418
		hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[i], addr_lst) {
L
Linus Torvalds 已提交
4419 4420
			unsigned long age;

4421 4422 4423 4424 4425 4426
			/* 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
4427 4428 4429
				continue;

			spin_lock(&ifp->lock);
4430 4431
			/* We try to batch several events at once. */
			age = (now - ifp->tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
L
Linus Torvalds 已提交
4432

4433 4434
			if (ifp->valid_lft != INFINITY_LIFE_TIME &&
			    age >= ifp->valid_lft) {
L
Linus Torvalds 已提交
4435 4436 4437 4438
				spin_unlock(&ifp->lock);
				in6_ifa_hold(ifp);
				ipv6_del_addr(ifp);
				goto restart;
4439 4440 4441
			} else if (ifp->prefered_lft == INFINITY_LIFE_TIME) {
				spin_unlock(&ifp->lock);
				continue;
L
Linus Torvalds 已提交
4442
			} else if (age >= ifp->prefered_lft) {
4443
				/* jiffies - ifp->tstamp > age >= ifp->prefered_lft */
L
Linus Torvalds 已提交
4444 4445 4446 4447 4448 4449 4450
				int deprecate = 0;

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

4451 4452
				if ((ifp->valid_lft != INFINITY_LIFE_TIME) &&
				    (time_before(ifp->tstamp + ifp->valid_lft * HZ, next)))
L
Linus Torvalds 已提交
4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465
					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)) {
4466 4467
				unsigned long regen_advance = ifp->idev->cnf.regen_max_retry *
					ifp->idev->cnf.dad_transmits *
J
Jiri Pirko 已提交
4468
					NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME) / HZ;
4469

L
Linus Torvalds 已提交
4470 4471 4472 4473 4474 4475 4476 4477 4478
				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);
4479

4480 4481 4482
						spin_lock(&ifpub->lock);
						ifpub->regen_count = 0;
						spin_unlock(&ifpub->lock);
4483
						rcu_read_unlock_bh();
E
Eric Dumazet 已提交
4484
						ipv6_create_tempaddr(ifpub, ifp, true);
L
Linus Torvalds 已提交
4485 4486
						in6_ifa_put(ifpub);
						in6_ifa_put(ifp);
4487
						rcu_read_lock_bh();
L
Linus Torvalds 已提交
4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501
						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);
			}
		}
	}

4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512
	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;

4513 4514
	pr_debug("now = %lu, schedule = %lu, rounded schedule = %lu => %lu\n",
		 now, next, next_sec, next_sched);
4515
	mod_delayed_work(addrconf_wq, &addr_chk_work, next_sched - now);
4516
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
4517 4518
}

4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530
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);
}

4531 4532
static struct in6_addr *extract_addr(struct nlattr *addr, struct nlattr *local,
				     struct in6_addr **peer_pfx)
4533 4534 4535
{
	struct in6_addr *pfx = NULL;

4536 4537
	*peer_pfx = NULL;

4538 4539 4540 4541 4542
	if (addr)
		pfx = nla_data(addr);

	if (local) {
		if (pfx && nla_memcmp(local, pfx, sizeof(*pfx)))
4543 4544
			*peer_pfx = pfx;
		pfx = nla_data(local);
4545 4546 4547 4548 4549
	}

	return pfx;
}

4550
static const struct nla_policy ifa_ipv6_policy[IFA_MAX+1] = {
4551 4552 4553
	[IFA_ADDRESS]		= { .len = sizeof(struct in6_addr) },
	[IFA_LOCAL]		= { .len = sizeof(struct in6_addr) },
	[IFA_CACHEINFO]		= { .len = sizeof(struct ifa_cacheinfo) },
4554
	[IFA_FLAGS]		= { .len = sizeof(u32) },
4555
	[IFA_RT_PRIORITY]	= { .len = sizeof(u32) },
4556
	[IFA_TARGET_NETNSID]	= { .type = NLA_S32 },
4557 4558
};

L
Linus Torvalds 已提交
4559
static int
4560 4561
inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh,
		  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
4562
{
4563
	struct net *net = sock_net(skb->sk);
4564 4565
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4566
	struct in6_addr *pfx, *peer_pfx;
4567
	u32 ifa_flags;
4568
	int err;
L
Linus Torvalds 已提交
4569

4570 4571
	err = nlmsg_parse_deprecated(nlh, sizeof(*ifm), tb, IFA_MAX,
				     ifa_ipv6_policy, extack);
4572 4573 4574 4575
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
4576
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
4577
	if (!pfx)
L
Linus Torvalds 已提交
4578 4579
		return -EINVAL;

4580 4581 4582 4583 4584 4585 4586
	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
4587 4588
}

4589 4590 4591 4592
static int modify_prefix_route(struct inet6_ifaddr *ifp,
			       unsigned long expires, u32 flags)
{
	struct fib6_info *f6i;
4593
	u32 prio;
4594

4595 4596
	f6i = addrconf_get_prefix_route(&ifp->addr, ifp->prefix_len,
					ifp->idev->dev, 0, RTF_DEFAULT, true);
4597 4598 4599
	if (!f6i)
		return -ENOENT;

4600 4601 4602 4603 4604
	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);

4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620
		/* 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;
}

4621
static int inet6_addr_modify(struct inet6_ifaddr *ifp, struct ifa6_config *cfg)
4622
{
4623 4624
	u32 flags;
	clock_t expires;
4625
	unsigned long timeout;
4626
	bool was_managetempaddr;
4627
	bool had_prefixroute;
4628

4629 4630
	ASSERT_RTNL();

4631
	if (!cfg->valid_lft || cfg->preferred_lft > cfg->valid_lft)
4632 4633
		return -EINVAL;

4634
	if (cfg->ifa_flags & IFA_F_MANAGETEMPADDR &&
4635 4636 4637
	    (ifp->flags & IFA_F_TEMPORARY || ifp->prefix_len != 64))
		return -EINVAL;

4638
	if (!(ifp->flags & IFA_F_TENTATIVE) || ifp->flags & IFA_F_DADFAILED)
4639
		cfg->ifa_flags &= ~IFA_F_OPTIMISTIC;
4640

4641
	timeout = addrconf_timeout_fixup(cfg->valid_lft, HZ);
4642 4643
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
4644
		cfg->valid_lft = timeout;
4645
		flags = RTF_EXPIRES;
4646 4647 4648
	} else {
		expires = 0;
		flags = 0;
4649
		cfg->ifa_flags |= IFA_F_PERMANENT;
4650
	}
4651

4652
	timeout = addrconf_timeout_fixup(cfg->preferred_lft, HZ);
4653 4654
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
4655 4656
			cfg->ifa_flags |= IFA_F_DEPRECATED;
		cfg->preferred_lft = timeout;
4657
	}
4658 4659

	spin_lock_bh(&ifp->lock);
4660
	was_managetempaddr = ifp->flags & IFA_F_MANAGETEMPADDR;
4661 4662
	had_prefixroute = ifp->flags & IFA_F_PERMANENT &&
			  !(ifp->flags & IFA_F_NOPREFIXROUTE);
4663
	ifp->flags &= ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD |
4664 4665
			IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
			IFA_F_NOPREFIXROUTE);
4666
	ifp->flags |= cfg->ifa_flags;
4667
	ifp->tstamp = jiffies;
4668 4669
	ifp->valid_lft = cfg->valid_lft;
	ifp->prefered_lft = cfg->preferred_lft;
4670

4671 4672
	if (cfg->rt_priority && cfg->rt_priority != ifp->rt_priority)
		ifp->rt_priority = cfg->rt_priority;
4673 4674 4675 4676 4677

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

4678
	if (!(cfg->ifa_flags & IFA_F_NOPREFIXROUTE)) {
4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689
		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);
		}
4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701
	} 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);
		}
4702
	}
4703 4704

	if (was_managetempaddr || ifp->flags & IFA_F_MANAGETEMPADDR) {
4705 4706 4707 4708 4709 4710 4711 4712
		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);
4713 4714
	}

4715
	addrconf_verify_rtnl();
4716 4717 4718 4719

	return 0;
}

L
Linus Torvalds 已提交
4720
static int
4721 4722
inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh,
		  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
4723
{
4724
	struct net *net = sock_net(skb->sk);
4725 4726
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4727
	struct in6_addr *peer_pfx;
4728 4729
	struct inet6_ifaddr *ifa;
	struct net_device *dev;
4730
	struct inet6_dev *idev;
4731
	struct ifa6_config cfg;
4732
	int err;
L
Linus Torvalds 已提交
4733

4734 4735
	err = nlmsg_parse_deprecated(nlh, sizeof(*ifm), tb, IFA_MAX,
				     ifa_ipv6_policy, extack);
4736 4737 4738
	if (err < 0)
		return err;

4739 4740
	memset(&cfg, 0, sizeof(cfg));

4741
	ifm = nlmsg_data(nlh);
4742 4743
	cfg.pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
	if (!cfg.pfx)
L
Linus Torvalds 已提交
4744 4745
		return -EINVAL;

4746 4747
	cfg.peer_pfx = peer_pfx;
	cfg.plen = ifm->ifa_prefixlen;
4748 4749 4750
	if (tb[IFA_RT_PRIORITY])
		cfg.rt_priority = nla_get_u32(tb[IFA_RT_PRIORITY]);

4751 4752 4753
	cfg.valid_lft = INFINITY_LIFE_TIME;
	cfg.preferred_lft = INFINITY_LIFE_TIME;

4754
	if (tb[IFA_CACHEINFO]) {
4755
		struct ifa_cacheinfo *ci;
4756 4757

		ci = nla_data(tb[IFA_CACHEINFO]);
4758 4759
		cfg.valid_lft = ci->ifa_valid;
		cfg.preferred_lft = ci->ifa_prefered;
4760 4761
	}

4762
	dev =  __dev_get_by_index(net, ifm->ifa_index);
4763
	if (!dev)
4764 4765
		return -ENODEV;

4766 4767 4768 4769
	if (tb[IFA_FLAGS])
		cfg.ifa_flags = nla_get_u32(tb[IFA_FLAGS]);
	else
		cfg.ifa_flags = ifm->ifa_flags;
4770

4771
	/* We ignore other flags so far. */
4772 4773 4774
	cfg.ifa_flags &= IFA_F_NODAD | IFA_F_HOMEADDRESS |
			 IFA_F_MANAGETEMPADDR | IFA_F_NOPREFIXROUTE |
			 IFA_F_MCAUTOJOIN | IFA_F_OPTIMISTIC;
4775 4776

	idev = ipv6_find_idev(dev);
4777 4778
	if (!idev)
		return -ENOBUFS;
4779 4780

	if (!ipv6_allow_optimistic_dad(net, idev))
4781
		cfg.ifa_flags &= ~IFA_F_OPTIMISTIC;
4782

4783 4784
	if (cfg.ifa_flags & IFA_F_NODAD &&
	    cfg.ifa_flags & IFA_F_OPTIMISTIC) {
4785 4786 4787
		NL_SET_ERR_MSG(extack, "IFA_F_NODAD and IFA_F_OPTIMISTIC are mutually exclusive");
		return -EINVAL;
	}
4788

4789
	ifa = ipv6_get_ifaddr(net, cfg.pfx, dev, 1);
4790
	if (!ifa) {
4791 4792
		/*
		 * It would be best to check for !NLM_F_CREATE here but
4793
		 * userspace already relies on not having to provide this.
4794
		 */
4795
		return inet6_addr_add(net, ifm->ifa_index, &cfg, extack);
4796 4797
	}

4798 4799 4800 4801
	if (nlh->nlmsg_flags & NLM_F_EXCL ||
	    !(nlh->nlmsg_flags & NLM_F_REPLACE))
		err = -EEXIST;
	else
4802
		err = inet6_addr_modify(ifa, &cfg);
4803 4804 4805 4806

	in6_ifa_put(ifa);

	return err;
L
Linus Torvalds 已提交
4807 4808
}

4809
static void put_ifaddrmsg(struct nlmsghdr *nlh, u8 prefixlen, u32 flags,
4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821
			  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;
}

4822 4823 4824 4825 4826
static int put_cacheinfo(struct sk_buff *skb, unsigned long cstamp,
			 unsigned long tstamp, u32 preferred, u32 valid)
{
	struct ifa_cacheinfo ci;

4827 4828
	ci.cstamp = cstamp_delta(cstamp);
	ci.tstamp = cstamp_delta(tstamp);
4829 4830 4831 4832 4833 4834
	ci.ifa_prefered = preferred;
	ci.ifa_valid = valid;

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

4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846
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;
}

4847 4848
static inline int inet6_ifaddr_msgsize(void)
{
4849
	return NLMSG_ALIGN(sizeof(struct ifaddrmsg))
4850
	       + nla_total_size(16) /* IFA_LOCAL */
4851
	       + nla_total_size(16) /* IFA_ADDRESS */
4852
	       + nla_total_size(sizeof(struct ifa_cacheinfo))
4853 4854
	       + nla_total_size(4)  /* IFA_FLAGS */
	       + nla_total_size(4)  /* IFA_RT_PRIORITY */;
4855
}
4856

4857 4858 4859 4860 4861 4862
enum addr_type_t {
	UNICAST_ADDR,
	MULTICAST_ADDR,
	ANYCAST_ADDR,
};

C
Christian Brauner 已提交
4863 4864 4865 4866 4867 4868
struct inet6_fill_args {
	u32 portid;
	u32 seq;
	int event;
	unsigned int flags;
	int netnsid;
4869
	int ifindex;
4870
	enum addr_type_t type;
C
Christian Brauner 已提交
4871 4872
};

L
Linus Torvalds 已提交
4873
static int inet6_fill_ifaddr(struct sk_buff *skb, struct inet6_ifaddr *ifa,
C
Christian Brauner 已提交
4874
			     struct inet6_fill_args *args)
L
Linus Torvalds 已提交
4875 4876
{
	struct nlmsghdr  *nlh;
4877
	u32 preferred, valid;
L
Linus Torvalds 已提交
4878

C
Christian Brauner 已提交
4879 4880
	nlh = nlmsg_put(skb, args->portid, args->seq, args->event,
			sizeof(struct ifaddrmsg), args->flags);
4881
	if (!nlh)
4882
		return -EMSGSIZE;
4883

4884 4885 4886
	put_ifaddrmsg(nlh, ifa->prefix_len, ifa->flags, rt_scope(ifa->scope),
		      ifa->idev->dev->ifindex);

C
Christian Brauner 已提交
4887 4888
	if (args->netnsid >= 0 &&
	    nla_put_s32(skb, IFA_TARGET_NETNSID, args->netnsid))
4889 4890
		goto error;

4891 4892
	if (!((ifa->flags&IFA_F_PERMANENT) &&
	      (ifa->prefered_lft == INFINITY_LIFE_TIME))) {
4893 4894 4895
		preferred = ifa->prefered_lft;
		valid = ifa->valid_lft;
		if (preferred != INFINITY_LIFE_TIME) {
L
Linus Torvalds 已提交
4896
			long tval = (jiffies - ifa->tstamp)/HZ;
4897 4898 4899 4900
			if (preferred > tval)
				preferred -= tval;
			else
				preferred = 0;
4901 4902 4903 4904 4905 4906
			if (valid != INFINITY_LIFE_TIME) {
				if (valid > tval)
					valid -= tval;
				else
					valid = 0;
			}
L
Linus Torvalds 已提交
4907 4908
		}
	} else {
4909 4910 4911 4912
		preferred = INFINITY_LIFE_TIME;
		valid = INFINITY_LIFE_TIME;
	}

C
Cong Wang 已提交
4913
	if (!ipv6_addr_any(&ifa->peer_addr)) {
4914 4915
		if (nla_put_in6_addr(skb, IFA_LOCAL, &ifa->addr) < 0 ||
		    nla_put_in6_addr(skb, IFA_ADDRESS, &ifa->peer_addr) < 0)
4916 4917
			goto error;
	} else
4918
		if (nla_put_in6_addr(skb, IFA_ADDRESS, &ifa->addr) < 0)
4919 4920
			goto error;

4921 4922 4923 4924
	if (ifa->rt_priority &&
	    nla_put_u32(skb, IFA_RT_PRIORITY, ifa->rt_priority))
		goto error;

4925 4926
	if (put_cacheinfo(skb, ifa->cstamp, ifa->tstamp, preferred, valid) < 0)
		goto error;
L
Linus Torvalds 已提交
4927

4928 4929 4930
	if (nla_put_u32(skb, IFA_FLAGS, ifa->flags) < 0)
		goto error;

4931 4932
	nlmsg_end(skb, nlh);
	return 0;
4933 4934 4935 4936

error:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4937 4938 4939
}

static int inet6_fill_ifmcaddr(struct sk_buff *skb, struct ifmcaddr6 *ifmca,
C
Christian Brauner 已提交
4940
			       struct inet6_fill_args *args)
L
Linus Torvalds 已提交
4941 4942
{
	struct nlmsghdr  *nlh;
4943 4944 4945 4946 4947
	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
4948

C
Christian Brauner 已提交
4949 4950
	nlh = nlmsg_put(skb, args->portid, args->seq, args->event,
			sizeof(struct ifaddrmsg), args->flags);
4951
	if (!nlh)
4952
		return -EMSGSIZE;
4953

C
Christian Brauner 已提交
4954 4955
	if (args->netnsid >= 0 &&
	    nla_put_s32(skb, IFA_TARGET_NETNSID, args->netnsid))
4956 4957
		return -EMSGSIZE;

4958
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
4959
	if (nla_put_in6_addr(skb, IFA_MULTICAST, &ifmca->mca_addr) < 0 ||
4960
	    put_cacheinfo(skb, ifmca->mca_cstamp, ifmca->mca_tstamp,
4961 4962 4963 4964
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
4965

4966 4967
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4968 4969 4970
}

static int inet6_fill_ifacaddr(struct sk_buff *skb, struct ifacaddr6 *ifaca,
C
Christian Brauner 已提交
4971
			       struct inet6_fill_args *args)
L
Linus Torvalds 已提交
4972
{
D
David Ahern 已提交
4973 4974
	struct net_device *dev = fib6_info_nh_dev(ifaca->aca_rt);
	int ifindex = dev ? dev->ifindex : 1;
L
Linus Torvalds 已提交
4975
	struct nlmsghdr  *nlh;
4976 4977 4978 4979
	u8 scope = RT_SCOPE_UNIVERSE;

	if (ipv6_addr_scope(&ifaca->aca_addr) & IFA_SITE)
		scope = RT_SCOPE_SITE;
L
Linus Torvalds 已提交
4980

C
Christian Brauner 已提交
4981 4982
	nlh = nlmsg_put(skb, args->portid, args->seq, args->event,
			sizeof(struct ifaddrmsg), args->flags);
4983
	if (!nlh)
4984
		return -EMSGSIZE;
4985

C
Christian Brauner 已提交
4986 4987
	if (args->netnsid >= 0 &&
	    nla_put_s32(skb, IFA_TARGET_NETNSID, args->netnsid))
4988 4989
		return -EMSGSIZE;

4990
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
4991
	if (nla_put_in6_addr(skb, IFA_ANYCAST, &ifaca->aca_addr) < 0 ||
4992
	    put_cacheinfo(skb, ifaca->aca_cstamp, ifaca->aca_tstamp,
4993 4994 4995 4996
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
L
Linus Torvalds 已提交
4997

4998 4999
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
5000 5001
}

E
Eric Dumazet 已提交
5002 5003
/* called with rcu_read_lock() */
static int in6_dump_addrs(struct inet6_dev *idev, struct sk_buff *skb,
5004
			  struct netlink_callback *cb, int s_ip_idx,
5005
			  struct inet6_fill_args *fillargs)
E
Eric Dumazet 已提交
5006 5007 5008
{
	struct ifmcaddr6 *ifmca;
	struct ifacaddr6 *ifaca;
5009
	int ip_idx = 0;
E
Eric Dumazet 已提交
5010 5011 5012
	int err = 1;

	read_lock_bh(&idev->lock);
5013
	switch (fillargs->type) {
5014 5015
	case UNICAST_ADDR: {
		struct inet6_ifaddr *ifa;
5016
		fillargs->event = RTM_NEWADDR;
5017

E
Eric Dumazet 已提交
5018
		/* unicast address incl. temp addr */
5019
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
5020 5021
			if (ip_idx < s_ip_idx)
				goto next;
5022
			err = inet6_fill_ifaddr(skb, ifa, fillargs);
5023
			if (err < 0)
E
Eric Dumazet 已提交
5024
				break;
5025
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
5026 5027
next:
			ip_idx++;
E
Eric Dumazet 已提交
5028 5029
		}
		break;
5030
	}
E
Eric Dumazet 已提交
5031
	case MULTICAST_ADDR:
5032
		fillargs->event = RTM_GETMULTICAST;
C
Christian Brauner 已提交
5033

E
Eric Dumazet 已提交
5034 5035 5036 5037 5038
		/* multicast address */
		for (ifmca = idev->mc_list; ifmca;
		     ifmca = ifmca->next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
5039
			err = inet6_fill_ifmcaddr(skb, ifmca, fillargs);
5040
			if (err < 0)
E
Eric Dumazet 已提交
5041 5042 5043 5044
				break;
		}
		break;
	case ANYCAST_ADDR:
5045
		fillargs->event = RTM_GETANYCAST;
E
Eric Dumazet 已提交
5046 5047 5048 5049 5050
		/* anycast address */
		for (ifaca = idev->ac_list; ifaca;
		     ifaca = ifaca->aca_next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
5051
			err = inet6_fill_ifacaddr(skb, ifaca, fillargs);
5052
			if (err < 0)
E
Eric Dumazet 已提交
5053 5054 5055 5056 5057 5058 5059
				break;
		}
		break;
	default:
		break;
	}
	read_unlock_bh(&idev->lock);
5060
	cb->args[2] = ip_idx;
E
Eric Dumazet 已提交
5061 5062 5063
	return err;
}

5064 5065 5066
static int inet6_valid_dump_ifaddr_req(const struct nlmsghdr *nlh,
				       struct inet6_fill_args *fillargs,
				       struct net **tgt_net, struct sock *sk,
5067
				       struct netlink_callback *cb)
5068
{
5069
	struct netlink_ext_ack *extack = cb->extack;
5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083
	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;
	}
5084 5085 5086 5087 5088

	fillargs->ifindex = ifm->ifa_index;
	if (fillargs->ifindex) {
		cb->answer_flags |= NLM_F_DUMP_FILTERED;
		fillargs->flags |= NLM_F_DUMP_FILTERED;
5089 5090
	}

5091 5092
	err = nlmsg_parse_deprecated_strict(nlh, sizeof(*ifm), tb, IFA_MAX,
					    ifa_ipv6_policy, extack);
5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105
	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)) {
5106
				fillargs->netnsid = -1;
5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119
				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
5120 5121 5122
static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
			   enum addr_type_t type)
{
5123
	const struct nlmsghdr *nlh = cb->nlh;
5124 5125 5126 5127 5128 5129 5130
	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
5131
	struct net *net = sock_net(skb->sk);
5132
	struct net *tgt_net = net;
5133
	int idx, s_idx, s_ip_idx;
E
Eric Dumazet 已提交
5134
	int h, s_h;
L
Linus Torvalds 已提交
5135
	struct net_device *dev;
E
Eric Dumazet 已提交
5136 5137
	struct inet6_dev *idev;
	struct hlist_head *head;
5138
	int err = 0;
5139

E
Eric Dumazet 已提交
5140 5141
	s_h = cb->args[0];
	s_idx = idx = cb->args[1];
5142
	s_ip_idx = cb->args[2];
5143

5144 5145
	if (cb->strict_check) {
		err = inet6_valid_dump_ifaddr_req(nlh, &fillargs, &tgt_net,
5146
						  skb->sk, cb);
5147
		if (err < 0)
5148
			goto put_tgt_net;
5149

5150
		err = 0;
5151 5152
		if (fillargs.ifindex) {
			dev = __dev_get_by_index(tgt_net, fillargs.ifindex);
5153 5154 5155 5156
			if (!dev) {
				err = -ENODEV;
				goto put_tgt_net;
			}
5157 5158 5159 5160
			idev = __in6_dev_get(dev);
			if (idev) {
				err = in6_dump_addrs(idev, skb, cb, s_ip_idx,
						     &fillargs);
5161 5162
				if (err > 0)
					err = 0;
5163 5164 5165
			}
			goto put_tgt_net;
		}
5166 5167
	}

E
Eric Dumazet 已提交
5168
	rcu_read_lock();
5169
	cb->seq = atomic_read(&tgt_net->ipv6.dev_addr_genid) ^ tgt_net->dev_base_seq;
E
Eric Dumazet 已提交
5170 5171
	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
5172
		head = &tgt_net->dev_index_head[h];
5173
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
5174 5175
			if (idx < s_idx)
				goto cont;
5176
			if (h > s_h || idx > s_idx)
E
Eric Dumazet 已提交
5177
				s_ip_idx = 0;
S
Stephen Hemminger 已提交
5178 5179
			idev = __in6_dev_get(dev);
			if (!idev)
E
Eric Dumazet 已提交
5180 5181
				goto cont;

5182
			if (in6_dump_addrs(idev, skb, cb, s_ip_idx,
5183
					   &fillargs) < 0)
E
Eric Dumazet 已提交
5184
				goto done;
5185
cont:
E
Eric Dumazet 已提交
5186 5187
			idx++;
		}
L
Linus Torvalds 已提交
5188
	}
E
Eric Dumazet 已提交
5189 5190 5191 5192
done:
	rcu_read_unlock();
	cb->args[0] = h;
	cb->args[1] = idx;
5193
put_tgt_net:
5194
	if (fillargs.netnsid >= 0)
5195
		put_net(tgt_net);
E
Eric Dumazet 已提交
5196

5197
	return skb->len ? : err;
L
Linus Torvalds 已提交
5198 5199 5200 5201 5202
}

static int inet6_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = UNICAST_ADDR;
5203

L
Linus Torvalds 已提交
5204 5205 5206 5207 5208 5209
	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;
5210

L
Linus Torvalds 已提交
5211 5212 5213 5214 5215 5216 5217
	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;
5218

L
Linus Torvalds 已提交
5219 5220 5221
	return inet6_dump_addr(skb, cb, type);
}

5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241
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))
5242 5243
		return nlmsg_parse_deprecated(nlh, sizeof(*ifm), tb, IFA_MAX,
					      ifa_ipv6_policy, extack);
5244

5245 5246
	err = nlmsg_parse_deprecated_strict(nlh, sizeof(*ifm), tb, IFA_MAX,
					    ifa_ipv6_policy, extack);
5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267
	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;
}

5268 5269
static int inet6_rtm_getaddr(struct sk_buff *in_skb, struct nlmsghdr *nlh,
			     struct netlink_ext_ack *extack)
5270
{
5271
	struct net *net = sock_net(in_skb->sk);
C
Christian Brauner 已提交
5272 5273 5274 5275 5276 5277 5278
	struct inet6_fill_args fillargs = {
		.portid = NETLINK_CB(in_skb).portid,
		.seq = nlh->nlmsg_seq,
		.event = RTM_NEWADDR,
		.flags = 0,
		.netnsid = -1,
	};
5279
	struct net *tgt_net = net;
5280 5281
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
5282
	struct in6_addr *addr = NULL, *peer;
5283 5284 5285 5286 5287
	struct net_device *dev = NULL;
	struct inet6_ifaddr *ifa;
	struct sk_buff *skb;
	int err;

5288
	err = inet6_rtm_valid_getaddr_req(in_skb, nlh, tb, extack);
5289
	if (err < 0)
5290
		return err;
5291

5292
	if (tb[IFA_TARGET_NETNSID]) {
C
Christian Brauner 已提交
5293
		fillargs.netnsid = nla_get_s32(tb[IFA_TARGET_NETNSID]);
5294 5295

		tgt_net = rtnl_get_net_ns_capable(NETLINK_CB(in_skb).sk,
C
Christian Brauner 已提交
5296
						  fillargs.netnsid);
5297 5298 5299 5300
		if (IS_ERR(tgt_net))
			return PTR_ERR(tgt_net);
	}

5301
	addr = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer);
5302 5303
	if (!addr)
		return -EINVAL;
5304

5305
	ifm = nlmsg_data(nlh);
5306
	if (ifm->ifa_index)
5307
		dev = dev_get_by_index(tgt_net, ifm->ifa_index);
5308

5309
	ifa = ipv6_get_ifaddr(tgt_net, addr, dev, 1);
S
Stephen Hemminger 已提交
5310
	if (!ifa) {
5311 5312 5313
		err = -EADDRNOTAVAIL;
		goto errout;
	}
5314

S
Stephen Hemminger 已提交
5315 5316
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_KERNEL);
	if (!skb) {
5317
		err = -ENOBUFS;
5318
		goto errout_ifa;
5319 5320
	}

C
Christian Brauner 已提交
5321
	err = inet6_fill_ifaddr(skb, ifa, &fillargs);
5322 5323 5324 5325 5326 5327
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout_ifa;
	}
5328
	err = rtnl_unicast(skb, tgt_net, NETLINK_CB(in_skb).portid);
5329
errout_ifa:
5330
	in6_ifa_put(ifa);
5331
errout:
5332 5333
	if (dev)
		dev_put(dev);
C
Christian Brauner 已提交
5334
	if (fillargs.netnsid >= 0)
5335 5336
		put_net(tgt_net);

5337 5338 5339
	return err;
}

L
Linus Torvalds 已提交
5340 5341 5342
static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa)
{
	struct sk_buff *skb;
5343
	struct net *net = dev_net(ifa->idev->dev);
C
Christian Brauner 已提交
5344 5345 5346 5347 5348 5349 5350
	struct inet6_fill_args fillargs = {
		.portid = 0,
		.seq = 0,
		.event = event,
		.flags = 0,
		.netnsid = -1,
	};
5351
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
5352

5353
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_ATOMIC);
5354
	if (!skb)
5355 5356
		goto errout;

C
Christian Brauner 已提交
5357
	err = inet6_fill_ifaddr(skb, ifa, &fillargs);
5358 5359 5360 5361 5362 5363
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5364 5365
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFADDR, NULL, GFP_ATOMIC);
	return;
5366 5367
errout:
	if (err < 0)
5368
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFADDR, err);
L
Linus Torvalds 已提交
5369 5370
}

D
Dave Jones 已提交
5371
static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
L
Linus Torvalds 已提交
5372 5373
				__s32 *array, int bytes)
{
5374 5375
	BUG_ON(bytes < (DEVCONF_MAX * 4));

L
Linus Torvalds 已提交
5376 5377 5378 5379 5380 5381 5382 5383 5384
	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;
5385 5386
	array[DEVCONF_RTR_SOLICIT_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_interval);
5387 5388
	array[DEVCONF_RTR_SOLICIT_MAX_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_max_interval);
5389 5390
	array[DEVCONF_RTR_SOLICIT_DELAY] =
		jiffies_to_msecs(cnf->rtr_solicit_delay);
L
Linus Torvalds 已提交
5391
	array[DEVCONF_FORCE_MLD_VERSION] = cnf->force_mld_version;
5392 5393 5394 5395
	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
5396 5397 5398 5399 5400 5401
	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;
5402
	array[DEVCONF_ACCEPT_RA_DEFRTR] = cnf->accept_ra_defrtr;
5403
	array[DEVCONF_ACCEPT_RA_MIN_HOP_LIMIT] = cnf->accept_ra_min_hop_limit;
5404
	array[DEVCONF_ACCEPT_RA_PINFO] = cnf->accept_ra_pinfo;
5405 5406
#ifdef CONFIG_IPV6_ROUTER_PREF
	array[DEVCONF_ACCEPT_RA_RTR_PREF] = cnf->accept_ra_rtr_pref;
5407 5408
	array[DEVCONF_RTR_PROBE_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_probe_interval);
N
Neil Horman 已提交
5409
#ifdef CONFIG_IPV6_ROUTE_INFO
5410
	array[DEVCONF_ACCEPT_RA_RT_INFO_MIN_PLEN] = cnf->accept_ra_rt_info_min_plen;
5411 5412
	array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
#endif
5413
#endif
5414
	array[DEVCONF_PROXY_NDP] = cnf->proxy_ndp;
5415
	array[DEVCONF_ACCEPT_SOURCE_ROUTE] = cnf->accept_source_route;
5416 5417
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
5418
	array[DEVCONF_USE_OPTIMISTIC] = cnf->use_optimistic;
5419
#endif
5420 5421 5422
#ifdef CONFIG_IPV6_MROUTE
	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
#endif
5423
	array[DEVCONF_DISABLE_IPV6] = cnf->disable_ipv6;
5424
	array[DEVCONF_ACCEPT_DAD] = cnf->accept_dad;
5425
	array[DEVCONF_FORCE_TLLAO] = cnf->force_tllao;
5426
	array[DEVCONF_NDISC_NOTIFY] = cnf->ndisc_notify;
5427
	array[DEVCONF_SUPPRESS_FRAG_NDISC] = cnf->suppress_frag_ndisc;
5428
	array[DEVCONF_ACCEPT_RA_FROM_LOCAL] = cnf->accept_ra_from_local;
5429
	array[DEVCONF_ACCEPT_RA_MTU] = cnf->accept_ra_mtu;
5430
	array[DEVCONF_IGNORE_ROUTES_WITH_LINKDOWN] = cnf->ignore_routes_with_linkdown;
5431
	/* we omit DEVCONF_STABLE_SECRET for now */
5432
	array[DEVCONF_USE_OIF_ADDRS_ONLY] = cnf->use_oif_addrs_only;
5433
	array[DEVCONF_DROP_UNICAST_IN_L2_MULTICAST] = cnf->drop_unicast_in_l2_multicast;
5434
	array[DEVCONF_DROP_UNSOLICITED_NA] = cnf->drop_unsolicited_na;
5435
	array[DEVCONF_KEEP_ADDR_ON_DOWN] = cnf->keep_addr_on_down;
5436
	array[DEVCONF_SEG6_ENABLED] = cnf->seg6_enabled;
5437 5438 5439
#ifdef CONFIG_IPV6_SEG6_HMAC
	array[DEVCONF_SEG6_REQUIRE_HMAC] = cnf->seg6_require_hmac;
#endif
5440
	array[DEVCONF_ENHANCED_DAD] = cnf->enhanced_dad;
5441
	array[DEVCONF_ADDR_GEN_MODE] = cnf->addr_gen_mode;
5442
	array[DEVCONF_DISABLE_POLICY] = cnf->disable_policy;
5443
	array[DEVCONF_NDISC_TCLASS] = cnf->ndisc_tclass;
L
Linus Torvalds 已提交
5444 5445
}

T
Thomas Graf 已提交
5446 5447 5448 5449 5450 5451
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 */
5452
	     + nla_total_size(ICMP6_MIB_MAX * 8) /* IFLA_INET6_ICMP6STATS */
5453 5454 5455
	     + 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
5456 5457
}

5458 5459 5460 5461 5462 5463 5464
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 */
5465
	       + nla_total_size(1) /* IFLA_OPERSTATE */
T
Thomas Graf 已提交
5466
	       + nla_total_size(inet6_ifla6_size()); /* IFLA_PROTINFO */
5467
}
5468

5469
static inline void __snmp6_fill_statsdev(u64 *stats, atomic_long_t *mib,
5470
					int bytes)
5471 5472
{
	int i;
5473
	int pad = bytes - sizeof(u64) * ICMP6_MIB_MAX;
5474 5475 5476
	BUG_ON(pad < 0);

	/* Use put_unaligned() because stats may not be aligned for u64. */
5477 5478
	put_unaligned(ICMP6_MIB_MAX, &stats[0]);
	for (i = 1; i < ICMP6_MIB_MAX; i++)
5479
		put_unaligned(atomic_long_read(&mib[i]), &stats[i]);
5480

5481
	memset(&stats[ICMP6_MIB_MAX], 0, pad);
5482 5483
}

W
WANG Cong 已提交
5484
static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
5485
					int bytes, size_t syncpoff)
5486
{
5487 5488 5489 5490
	int i, c;
	u64 buff[IPSTATS_MIB_MAX];
	int pad = bytes - sizeof(u64) * IPSTATS_MIB_MAX;

5491 5492
	BUG_ON(pad < 0);

5493 5494
	memset(buff, 0, sizeof(buff));
	buff[0] = IPSTATS_MIB_MAX;
5495

5496 5497 5498 5499 5500 5501 5502
	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);
5503 5504
}

5505 5506 5507
static void snmp6_fill_stats(u64 *stats, struct inet6_dev *idev, int attrtype,
			     int bytes)
{
S
Stephen Hemminger 已提交
5508
	switch (attrtype) {
5509
	case IFLA_INET6_STATS:
5510 5511
		__snmp6_fill_stats64(stats, idev->stats.ipv6, bytes,
				     offsetof(struct ipstats_mib, syncp));
5512 5513
		break;
	case IFLA_INET6_ICMP6STATS:
5514
		__snmp6_fill_statsdev(stats, idev->stats.icmpv6dev->mibs, bytes);
5515 5516 5517 5518
		break;
	}
}

5519 5520
static int inet6_fill_ifla6_attrs(struct sk_buff *skb, struct inet6_dev *idev,
				  u32 ext_filter_mask)
T
Thomas Graf 已提交
5521 5522 5523 5524
{
	struct nlattr *nla;
	struct ifla_cacheinfo ci;

D
David S. Miller 已提交
5525 5526
	if (nla_put_u32(skb, IFLA_INET6_FLAGS, idev->if_flags))
		goto nla_put_failure;
T
Thomas Graf 已提交
5527
	ci.max_reasm_len = IPV6_MAXPLEN;
5528 5529
	ci.tstamp = cstamp_delta(idev->tstamp);
	ci.reachable_time = jiffies_to_msecs(idev->nd_parms->reachable_time);
J
Jiri Pirko 已提交
5530
	ci.retrans_time = jiffies_to_msecs(NEIGH_VAR(idev->nd_parms, RETRANS_TIME));
D
David S. Miller 已提交
5531 5532
	if (nla_put(skb, IFLA_INET6_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
T
Thomas Graf 已提交
5533
	nla = nla_reserve(skb, IFLA_INET6_CONF, DEVCONF_MAX * sizeof(s32));
5534
	if (!nla)
T
Thomas Graf 已提交
5535 5536 5537 5538 5539
		goto nla_put_failure;
	ipv6_store_devconf(&idev->cnf, nla_data(nla), nla_len(nla));

	/* XXX - MC not implemented */

5540 5541 5542
	if (ext_filter_mask & RTEXT_FILTER_SKIP_STATS)
		return 0;

T
Thomas Graf 已提交
5543
	nla = nla_reserve(skb, IFLA_INET6_STATS, IPSTATS_MIB_MAX * sizeof(u64));
5544
	if (!nla)
T
Thomas Graf 已提交
5545 5546 5547 5548
		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));
5549
	if (!nla)
T
Thomas Graf 已提交
5550 5551 5552
		goto nla_put_failure;
	snmp6_fill_stats(nla_data(nla), idev, IFLA_INET6_ICMP6STATS, nla_len(nla));

5553
	nla = nla_reserve(skb, IFLA_INET6_TOKEN, sizeof(struct in6_addr));
5554
	if (!nla)
5555
		goto nla_put_failure;
5556

5557
	if (nla_put_u8(skb, IFLA_INET6_ADDR_GEN_MODE, idev->cnf.addr_gen_mode))
5558 5559
		goto nla_put_failure;

5560 5561 5562 5563
	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
5564 5565 5566 5567 5568 5569
	return 0;

nla_put_failure:
	return -EMSGSIZE;
}

5570 5571
static size_t inet6_get_link_af_size(const struct net_device *dev,
				     u32 ext_filter_mask)
T
Thomas Graf 已提交
5572 5573 5574 5575 5576 5577 5578
{
	if (!__in6_dev_get(dev))
		return 0;

	return inet6_ifla6_size();
}

5579 5580
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
5581 5582 5583 5584 5585 5586
{
	struct inet6_dev *idev = __in6_dev_get(dev);

	if (!idev)
		return -ENODATA;

5587
	if (inet6_fill_ifla6_attrs(skb, idev, ext_filter_mask) < 0)
T
Thomas Graf 已提交
5588 5589 5590 5591 5592
		return -EMSGSIZE;

	return 0;
}

5593 5594 5595 5596
static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token)
{
	struct inet6_ifaddr *ifp;
	struct net_device *dev = idev->dev;
5597
	bool clear_token, update_rs = false;
5598
	struct in6_addr ll_addr;
5599

5600 5601
	ASSERT_RTNL();

5602
	if (!token)
5603 5604 5605 5606 5607
		return -EINVAL;
	if (dev->flags & (IFF_LOOPBACK | IFF_NOARP))
		return -EINVAL;
	if (!ipv6_accept_ra(idev))
		return -EINVAL;
5608
	if (idev->cnf.rtr_solicits == 0)
5609 5610 5611 5612 5613 5614 5615 5616 5617
		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);

5618 5619 5620 5621
	clear_token = ipv6_addr_any(token);
	if (clear_token)
		goto update_lft;

5622 5623 5624
	if (!idev->dead && (idev->if_flags & IF_READY) &&
	    !ipv6_get_lladdr(dev, &ll_addr, IFA_F_TENTATIVE |
			     IFA_F_OPTIMISTIC)) {
5625 5626 5627 5628 5629 5630
		/* 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;
	}
5631

5632
update_lft:
5633
	write_lock_bh(&idev->lock);
5634

5635
	if (update_rs) {
5636
		idev->if_flags |= IF_RS_SENT;
5637 5638
		idev->rs_interval = rfc3315_s14_backoff_init(
			idev->cnf.rtr_solicit_interval);
5639
		idev->rs_probes = 1;
5640
		addrconf_mod_rs_timer(idev, idev->rs_interval);
5641
	}
5642 5643 5644 5645

	/* Well, that's kinda nasty ... */
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
		spin_lock(&ifp->lock);
5646
		if (ifp->tokenized) {
5647 5648 5649 5650 5651 5652 5653
			ifp->valid_lft = 0;
			ifp->prefered_lft = 0;
		}
		spin_unlock(&ifp->lock);
	}

	write_unlock_bh(&idev->lock);
5654
	inet6_ifinfo_notify(RTM_NEWLINK, idev);
5655
	addrconf_verify_rtnl();
5656 5657 5658
	return 0;
}

5659 5660 5661 5662 5663
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) },
};

5664 5665 5666 5667 5668 5669 5670 5671 5672 5673 5674 5675 5676 5677 5678 5679 5680 5681 5682 5683
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;
}

5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696 5697 5698 5699 5700 5701 5702 5703 5704 5705 5706 5707 5708 5709 5710 5711 5712 5713 5714 5715 5716
static int inet6_validate_link_af(const struct net_device *dev,
				  const struct nlattr *nla)
{
	struct nlattr *tb[IFLA_INET6_MAX + 1];
	struct inet6_dev *idev = NULL;
	int err;

	if (dev) {
		idev = __in6_dev_get(dev);
		if (!idev)
			return -EAFNOSUPPORT;
	}

	err = nla_parse_nested_deprecated(tb, IFLA_INET6_MAX, nla,
					  inet6_af_policy, NULL);
	if (err)
		return err;

	if (!tb[IFLA_INET6_TOKEN] && !tb[IFLA_INET6_ADDR_GEN_MODE])
		return -EINVAL;

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

		if (check_addr_gen_mode(mode) < 0)
			return -EINVAL;
		if (dev && check_stable_privacy(idev, dev_net(dev), mode) < 0)
			return -EINVAL;
	}

	return 0;
}

5717 5718 5719 5720
static int inet6_set_link_af(struct net_device *dev, const struct nlattr *nla)
{
	struct inet6_dev *idev = __in6_dev_get(dev);
	struct nlattr *tb[IFLA_INET6_MAX + 1];
5721
	int err;
5722

5723
	if (nla_parse_nested_deprecated(tb, IFLA_INET6_MAX, nla, NULL, NULL) < 0)
5724 5725
		BUG();

5726
	if (tb[IFLA_INET6_TOKEN]) {
5727
		err = inet6_set_iftoken(idev, nla_data(tb[IFLA_INET6_TOKEN]));
5728 5729 5730 5731 5732 5733 5734
		if (err)
			return err;
	}

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

5735
		idev->cnf.addr_gen_mode = mode;
5736
	}
5737

5738
	return 0;
5739 5740
}

5741
static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
5742
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
5743
{
5744 5745 5746 5747 5748
	struct net_device *dev = idev->dev;
	struct ifinfomsg *hdr;
	struct nlmsghdr *nlh;
	void *protoinfo;

5749
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*hdr), flags);
5750
	if (!nlh)
5751
		return -EMSGSIZE;
5752 5753 5754 5755 5756 5757 5758 5759 5760

	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
5761 5762 5763 5764
	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) ||
5765
	    (dev->ifindex != dev_get_iflink(dev) &&
5766 5767 5768
	     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
5769
		goto nla_put_failure;
5770
	protoinfo = nla_nest_start_noflag(skb, IFLA_PROTINFO);
5771
	if (!protoinfo)
5772
		goto nla_put_failure;
L
Linus Torvalds 已提交
5773

5774
	if (inet6_fill_ifla6_attrs(skb, idev, 0) < 0)
5775
		goto nla_put_failure;
L
Linus Torvalds 已提交
5776

5777
	nla_nest_end(skb, protoinfo);
5778 5779
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
5780

5781
nla_put_failure:
5782 5783
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
5784 5785
}

5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796 5797 5798 5799 5800 5801 5802 5803 5804 5805 5806 5807 5808 5809 5810
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
5811 5812
static int inet6_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
{
5813
	struct net *net = sock_net(skb->sk);
E
Eric Dumazet 已提交
5814
	int h, s_h;
5815
	int idx = 0, s_idx;
L
Linus Torvalds 已提交
5816 5817
	struct net_device *dev;
	struct inet6_dev *idev;
E
Eric Dumazet 已提交
5818
	struct hlist_head *head;
L
Linus Torvalds 已提交
5819

5820 5821 5822 5823 5824 5825 5826 5827 5828 5829
	/* 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
5830 5831 5832 5833 5834 5835 5836
	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];
5837
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
5838 5839 5840 5841 5842 5843
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;
			if (inet6_fill_ifinfo(skb, idev,
5844
					      NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
5845
					      cb->nlh->nlmsg_seq,
5846
					      RTM_NEWLINK, NLM_F_MULTI) < 0)
E
Eric Dumazet 已提交
5847
				goto out;
5848
cont:
E
Eric Dumazet 已提交
5849 5850
			idx++;
		}
L
Linus Torvalds 已提交
5851
	}
E
Eric Dumazet 已提交
5852 5853 5854 5855
out:
	rcu_read_unlock();
	cb->args[1] = idx;
	cb->args[0] = h;
L
Linus Torvalds 已提交
5856 5857 5858 5859 5860 5861 5862

	return skb->len;
}

void inet6_ifinfo_notify(int event, struct inet6_dev *idev)
{
	struct sk_buff *skb;
5863
	struct net *net = dev_net(idev->dev);
5864
	int err = -ENOBUFS;
5865

5866
	skb = nlmsg_new(inet6_if_nlmsg_size(), GFP_ATOMIC);
5867
	if (!skb)
5868 5869 5870
		goto errout;

	err = inet6_fill_ifinfo(skb, idev, 0, 0, event, 0);
5871 5872 5873 5874 5875 5876
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_if_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5877
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFINFO, NULL, GFP_ATOMIC);
5878
	return;
5879 5880
errout:
	if (err < 0)
5881
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFINFO, err);
L
Linus Torvalds 已提交
5882 5883
}

5884 5885 5886 5887 5888 5889
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));
}
5890

L
Linus Torvalds 已提交
5891
static int inet6_fill_prefix(struct sk_buff *skb, struct inet6_dev *idev,
5892
			     struct prefix_info *pinfo, u32 portid, u32 seq,
5893
			     int event, unsigned int flags)
L
Linus Torvalds 已提交
5894
{
5895 5896
	struct prefixmsg *pmsg;
	struct nlmsghdr *nlh;
L
Linus Torvalds 已提交
5897 5898
	struct prefix_cacheinfo	ci;

5899
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*pmsg), flags);
5900
	if (!nlh)
5901
		return -EMSGSIZE;
5902 5903

	pmsg = nlmsg_data(nlh);
L
Linus Torvalds 已提交
5904
	pmsg->prefix_family = AF_INET6;
5905 5906
	pmsg->prefix_pad1 = 0;
	pmsg->prefix_pad2 = 0;
L
Linus Torvalds 已提交
5907 5908 5909
	pmsg->prefix_ifindex = idev->dev->ifindex;
	pmsg->prefix_len = pinfo->prefix_len;
	pmsg->prefix_type = pinfo->type;
5910
	pmsg->prefix_pad3 = 0;
L
Linus Torvalds 已提交
5911 5912 5913 5914 5915 5916
	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
5917 5918
	if (nla_put(skb, PREFIX_ADDRESS, sizeof(pinfo->prefix), &pinfo->prefix))
		goto nla_put_failure;
L
Linus Torvalds 已提交
5919 5920
	ci.preferred_time = ntohl(pinfo->prefered);
	ci.valid_time = ntohl(pinfo->valid);
D
David S. Miller 已提交
5921 5922
	if (nla_put(skb, PREFIX_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
5923 5924
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
5925

5926
nla_put_failure:
5927 5928
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
5929 5930
}

5931
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
5932 5933 5934
			 struct prefix_info *pinfo)
{
	struct sk_buff *skb;
5935
	struct net *net = dev_net(idev->dev);
5936
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
5937

5938
	skb = nlmsg_new(inet6_prefix_nlmsg_size(), GFP_ATOMIC);
5939
	if (!skb)
5940 5941 5942
		goto errout;

	err = inet6_fill_prefix(skb, idev, pinfo, 0, 0, event, 0);
5943 5944 5945 5946 5947 5948
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_prefix_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5949 5950
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_PREFIX, NULL, GFP_ATOMIC);
	return;
5951 5952
errout:
	if (err < 0)
5953
		rtnl_set_sk_err(net, RTNLGRP_IPV6_PREFIX, err);
L
Linus Torvalds 已提交
5954 5955 5956 5957
}

static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
5958 5959
	struct net *net = dev_net(ifp->idev->dev);

5960 5961 5962
	if (event)
		ASSERT_RTNL();

L
Linus Torvalds 已提交
5963 5964 5965 5966
	inet6_ifa_notify(event ? : RTM_NEWADDR, ifp);

	switch (event) {
	case RTM_NEWADDR:
5967 5968 5969 5970 5971 5972
		/*
		 * 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
		 */
5973
		if (!rcu_access_pointer(ifp->rt->fib6_node))
5974
			ip6_ins_rt(net, ifp->rt);
L
Linus Torvalds 已提交
5975 5976
		if (ifp->idev->cnf.forwarding)
			addrconf_join_anycast(ifp);
C
Cong Wang 已提交
5977
		if (!ipv6_addr_any(&ifp->peer_addr))
5978
			addrconf_prefix_route(&ifp->peer_addr, 128, 0,
5979
					      ifp->idev->dev, 0, 0,
5980
					      GFP_ATOMIC);
L
Linus Torvalds 已提交
5981 5982 5983 5984 5985
		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
5986
		if (!ipv6_addr_any(&ifp->peer_addr)) {
5987
			struct fib6_info *rt;
5988

5989
			rt = addrconf_get_prefix_route(&ifp->peer_addr, 128,
5990 5991
						       ifp->idev->dev, 0, 0,
						       false);
M
Martin KaFai Lau 已提交
5992
			if (rt)
5993
				ip6_del_rt(net, rt);
5994
		}
5995
		if (ifp->rt) {
5996 5997
			ip6_del_rt(net, ifp->rt);
			ifp->rt = NULL;
5998
		}
H
Hannes Frederic Sowa 已提交
5999
		rt_genid_bump_ipv6(net);
L
Linus Torvalds 已提交
6000 6001
		break;
	}
6002
	atomic_inc(&net->ipv6.dev_addr_genid);
L
Linus Torvalds 已提交
6003 6004 6005 6006
}

static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
6007
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
6008 6009
	if (likely(ifp->idev->dead == 0))
		__ipv6_ifa_notify(event, ifp);
6010
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
6011 6012 6013 6014 6015
}

#ifdef CONFIG_SYSCTL

static
6016
int addrconf_sysctl_forward(struct ctl_table *ctl, int write,
L
Linus Torvalds 已提交
6017 6018 6019 6020
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
6021
	loff_t pos = *ppos;
6022
	struct ctl_table lctl;
L
Linus Torvalds 已提交
6023 6024
	int ret;

6025 6026 6027 6028 6029 6030 6031 6032
	/*
	 * 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
6033

6034
	if (write)
6035
		ret = addrconf_fixup_forwarding(ctl, valp, val);
E
Eric W. Biederman 已提交
6036 6037
	if (ret)
		*ppos = pos;
6038
	return ret;
L
Linus Torvalds 已提交
6039 6040
}

6041 6042 6043 6044 6045 6046 6047 6048 6049 6050 6051 6052 6053 6054 6055
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);
}

6056 6057
static void dev_disable_change(struct inet6_dev *idev)
{
6058 6059
	struct netdev_notifier_info info;

6060 6061 6062
	if (!idev || !idev->dev)
		return;

6063
	netdev_notifier_info_init(&info, idev->dev);
6064
	if (idev->cnf.disable_ipv6)
6065
		addrconf_notify(NULL, NETDEV_DOWN, &info);
6066
	else
6067
		addrconf_notify(NULL, NETDEV_UP, &info);
6068 6069 6070 6071 6072 6073 6074
}

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

6075
	for_each_netdev(net, dev) {
6076 6077 6078 6079 6080 6081 6082 6083 6084 6085
		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);
		}
	}
}

6086
static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
6087 6088
{
	struct net *net;
6089 6090 6091 6092
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
6093 6094

	net = (struct net *)table->extra2;
6095 6096
	old = *p;
	*p = newf;
6097

6098 6099
	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
		rtnl_unlock();
6100
		return 0;
E
Eric W. Biederman 已提交
6101
	}
6102 6103 6104 6105

	if (p == &net->ipv6.devconf_all->disable_ipv6) {
		net->ipv6.devconf_dflt->disable_ipv6 = newf;
		addrconf_disable_change(net, newf);
6106
	} else if ((!newf) ^ (!old))
6107 6108 6109 6110 6111 6112 6113
		dev_disable_change((struct inet6_dev *)table->extra1);

	rtnl_unlock();
	return 0;
}

static
6114
int addrconf_sysctl_disable(struct ctl_table *ctl, int write,
6115 6116 6117 6118
			    void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
6119
	loff_t pos = *ppos;
6120
	struct ctl_table lctl;
6121 6122
	int ret;

6123 6124 6125 6126 6127 6128 6129 6130
	/*
	 * 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);
6131 6132 6133

	if (write)
		ret = addrconf_disable_ipv6(ctl, valp, val);
E
Eric W. Biederman 已提交
6134 6135
	if (ret)
		*ppos = pos;
6136 6137 6138
	return ret;
}

6139 6140 6141 6142 6143 6144 6145 6146 6147 6148 6149 6150 6151 6152 6153 6154 6155 6156 6157
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)
6158 6159
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
6160 6161 6162
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
		else if (valp == &net->ipv6.devconf_all->proxy_ndp)
6163 6164
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
6165 6166 6167 6168 6169
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
		else {
			struct inet6_dev *idev = ctl->extra1;

6170 6171
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
6172 6173 6174 6175 6176 6177 6178 6179 6180
						     idev->dev->ifindex,
						     &idev->cnf);
		}
		rtnl_unlock();
	}

	return ret;
}

6181 6182 6183 6184 6185
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;
6186
	u32 new_val;
6187 6188
	struct inet6_dev *idev = (struct inet6_dev *)ctl->extra1;
	struct net *net = (struct net *)ctl->extra2;
6189 6190 6191 6192 6193
	struct ctl_table tmp = {
		.data = &new_val,
		.maxlen = sizeof(new_val),
		.mode = ctl->mode,
	};
6194

6195 6196 6197
	if (!rtnl_trylock())
		return restart_syscall();

6198
	new_val = *((u32 *)ctl->data);
6199

6200 6201 6202
	ret = proc_douintvec(&tmp, write, buffer, lenp, ppos);
	if (ret != 0)
		goto out;
6203

6204
	if (write) {
6205 6206 6207 6208
		if (check_addr_gen_mode(new_val) < 0) {
			ret = -EINVAL;
			goto out;
		}
6209

6210
		if (idev) {
6211 6212 6213 6214
			if (check_stable_privacy(idev, net, new_val) < 0) {
				ret = -EINVAL;
				goto out;
			}
6215 6216 6217 6218 6219

			if (idev->cnf.addr_gen_mode != new_val) {
				idev->cnf.addr_gen_mode = new_val;
				addrconf_dev_config(idev->dev);
			}
6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231
		} 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);
				}
			}
6232
		}
6233 6234

		*((u32 *)ctl->data) = new_val;
6235 6236
	}

6237 6238 6239
out:
	rtnl_unlock();

6240 6241 6242
	return ret;
}

6243 6244 6245 6246 6247 6248 6249 6250
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;
6251
	struct net *net = ctl->extra2;
6252 6253
	struct ipv6_stable_secret *secret = ctl->data;

6254 6255 6256
	if (&net->ipv6.devconf_all->stable_secret == ctl->data)
		return -EIO;

6257 6258 6259 6260 6261 6262 6263 6264 6265 6266 6267
	lctl.maxlen = IPV6_MAX_STRLEN;
	lctl.data = str;

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

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

6268 6269 6270 6271
	err = snprintf(str, sizeof(str), "%pI6", &secret->secret);
	if (err >= sizeof(str)) {
		err = -EIO;
		goto out;
6272 6273 6274 6275 6276 6277 6278 6279 6280 6281 6282 6283 6284 6285
	}

	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;

6286 6287 6288 6289 6290 6291 6292
	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) {
6293
				idev->cnf.addr_gen_mode =
6294 6295 6296 6297 6298 6299
					IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
			}
		}
	} else {
		struct inet6_dev *idev = ctl->extra1;

6300
		idev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
6301 6302
	}

6303 6304 6305 6306 6307
out:
	rtnl_unlock();

	return err;
}
6308

6309 6310 6311 6312 6313 6314 6315 6316 6317 6318 6319 6320 6321 6322 6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334 6335 6336
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;
}

6337 6338 6339 6340 6341 6342 6343 6344 6345 6346 6347 6348 6349 6350 6351 6352 6353 6354 6355 6356
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) {
6357
			struct fib6_info *rt = ifa->rt;
6358 6359
			int cpu;

6360
			rcu_read_lock();
6361
			ifa->rt->dst_nopolicy = val ? true : false;
6362 6363 6364 6365 6366 6367 6368 6369
			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);
				}
			}
6370
			rcu_read_unlock();
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 6416 6417 6418 6419 6420 6421 6422 6423 6424 6425 6426 6427 6428 6429 6430 6431 6432 6433 6434
		}
		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;
}

6435
static int minus_one = -1;
6436
static const int zero = 0;
6437 6438 6439
static const int one = 1;
static const int two_five_five = 255;

6440
static const struct ctl_table addrconf_sysctl[] = {
6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451 6452
	{
		.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,
6453 6454 6455
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= (void *)&one,
		.extra2		= (void *)&two_five_five,
6456 6457 6458 6459 6460 6461 6462 6463 6464 6465 6466 6467 6468 6469 6470 6471 6472 6473 6474 6475 6476 6477 6478 6479 6480 6481 6482 6483 6484 6485 6486 6487 6488 6489 6490 6491 6492 6493 6494 6495 6496
	},
	{
		.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,
6497 6498
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &minus_one,
6499 6500 6501 6502 6503 6504 6505 6506
	},
	{
		.procname	= "router_solicitation_interval",
		.data		= &ipv6_devconf.rtr_solicit_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
6507 6508 6509 6510 6511 6512 6513
	{
		.procname	= "router_solicitation_max_interval",
		.data		= &ipv6_devconf.rtr_solicit_max_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
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 6588 6589 6590 6591 6592 6593 6594 6595 6596 6597 6598 6599 6600 6601 6602 6603 6604 6605 6606
	{
		.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,
	},
6607
#ifdef CONFIG_IPV6_ROUTER_PREF
6608 6609 6610 6611 6612 6613 6614 6615 6616 6617 6618 6619 6620 6621
	{
		.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
6622
#ifdef CONFIG_IPV6_ROUTE_INFO
6623 6624 6625 6626 6627 6628 6629
	{
		.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,
	},
6630 6631 6632 6633 6634 6635 6636
	{
		.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,
	},
6637
#endif
6638
#endif
6639 6640 6641 6642 6643 6644 6645 6646 6647 6648 6649 6650 6651 6652
	{
		.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,
	},
6653
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
6654 6655 6656 6657 6658 6659 6660 6661 6662 6663 6664 6665 6666 6667
	{
		.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,
	},
6668 6669
#endif
#ifdef CONFIG_IPV6_MROUTE
6670 6671 6672 6673 6674 6675 6676
	{
		.procname	= "mc_forwarding",
		.data		= &ipv6_devconf.mc_forwarding,
		.maxlen		= sizeof(int),
		.mode		= 0444,
		.proc_handler	= proc_dointvec,
	},
6677
#endif
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 6751 6752 6753 6754 6755 6756 6757 6758 6759 6760 6761 6762 6763 6764 6765 6766 6767 6768 6769
	{
		.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,

	},
6770 6771 6772 6773 6774 6775 6776
	{
		.procname	= "seg6_enabled",
		.data		= &ipv6_devconf.seg6_enabled,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
6777 6778 6779 6780 6781 6782 6783 6784 6785
#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
6786 6787 6788 6789 6790 6791 6792
	{
		.procname       = "enhanced_dad",
		.data           = &ipv6_devconf.enhanced_dad,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = proc_dointvec,
	},
6793 6794 6795 6796 6797 6798 6799
	{
		.procname		= "addr_gen_mode",
		.data			= &ipv6_devconf.addr_gen_mode,
		.maxlen			= sizeof(int),
		.mode			= 0644,
		.proc_handler	= addrconf_sysctl_addr_gen_mode,
	},
6800 6801 6802 6803 6804 6805 6806
	{
		.procname       = "disable_policy",
		.data           = &ipv6_devconf.disable_policy,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = addrconf_sysctl_disable_policy,
	},
6807 6808 6809 6810 6811 6812 6813 6814 6815
	{
		.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,
	},
6816 6817 6818
	{
		/* sentinel */
	}
L
Linus Torvalds 已提交
6819 6820
};

6821
static int __addrconf_sysctl_register(struct net *net, char *dev_name,
6822
		struct inet6_dev *idev, struct ipv6_devconf *p)
L
Linus Torvalds 已提交
6823
{
6824
	int i, ifindex;
6825
	struct ctl_table *table;
6826
	char path[sizeof("net/ipv6/conf/") + IFNAMSIZ];
6827

6828 6829
	table = kmemdup(addrconf_sysctl, sizeof(addrconf_sysctl), GFP_KERNEL);
	if (!table)
6830 6831
		goto out;

6832 6833
	for (i = 0; table[i].data; i++) {
		table[i].data += (char *)p - (char *)&ipv6_devconf;
6834 6835 6836 6837 6838 6839 6840 6841
		/* 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
6842 6843
	}

6844
	snprintf(path, sizeof(path), "net/ipv6/conf/%s", dev_name);
L
Linus Torvalds 已提交
6845

6846 6847
	p->sysctl_header = register_net_sysctl(net, path, table);
	if (!p->sysctl_header)
6848
		goto free;
6849

6850 6851 6852 6853 6854 6855
	if (!strcmp(dev_name, "all"))
		ifindex = NETCONFA_IFINDEX_ALL;
	else if (!strcmp(dev_name, "default"))
		ifindex = NETCONFA_IFINDEX_DEFAULT;
	else
		ifindex = idev->dev->ifindex;
6856 6857
	inet6_netconf_notify_devconf(net, RTM_NEWNETCONF, NETCONFA_ALL,
				     ifindex, p);
6858
	return 0;
L
Linus Torvalds 已提交
6859

6860
free:
6861
	kfree(table);
6862
out:
6863
	return -ENOBUFS;
L
Linus Torvalds 已提交
6864 6865
}

6866 6867
static void __addrconf_sysctl_unregister(struct net *net,
					 struct ipv6_devconf *p, int ifindex)
6868
{
6869
	struct ctl_table *table;
6870

6871
	if (!p->sysctl_header)
6872 6873
		return;

6874 6875 6876 6877
	table = p->sysctl_header->ctl_table_arg;
	unregister_net_sysctl_table(p->sysctl_header);
	p->sysctl_header = NULL;
	kfree(table);
6878 6879

	inet6_netconf_notify_devconf(net, RTM_DELNETCONF, 0, ifindex, NULL);
6880 6881
}

6882
static int addrconf_sysctl_register(struct inet6_dev *idev)
6883
{
6884 6885 6886 6887 6888 6889 6890 6891 6892 6893 6894 6895 6896 6897 6898
	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;
6899 6900
}

6901
static void addrconf_sysctl_unregister(struct inet6_dev *idev)
L
Linus Torvalds 已提交
6902
{
6903 6904
	__addrconf_sysctl_unregister(dev_net(idev->dev), &idev->cnf,
				     idev->dev->ifindex);
6905
	neigh_sysctl_unregister(idev->nd_parms);
L
Linus Torvalds 已提交
6906 6907 6908 6909 6910
}


#endif

6911
static int __net_init addrconf_init_net(struct net *net)
6912
{
6913
	int err = -ENOMEM;
6914 6915
	struct ipv6_devconf *all, *dflt;

6916
	all = kmemdup(&ipv6_devconf, sizeof(ipv6_devconf), GFP_KERNEL);
6917
	if (!all)
6918
		goto err_alloc_all;
6919

6920
	dflt = kmemdup(&ipv6_devconf_dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
6921
	if (!dflt)
6922
		goto err_alloc_dflt;
6923

6924 6925
	if (IS_ENABLED(CONFIG_SYSCTL) &&
	    sysctl_devconf_inherit_init_net == 1 && !net_eq(net, &init_net)) {
6926 6927 6928 6929
		memcpy(all, init_net.ipv6.devconf_all, sizeof(ipv6_devconf));
		memcpy(dflt, init_net.ipv6.devconf_dflt, sizeof(ipv6_devconf_dflt));
	}

6930 6931 6932
	/* these will be inherited by all namespaces */
	dflt->autoconf = ipv6_defaults.autoconf;
	dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
6933

6934 6935 6936
	dflt->stable_secret.initialized = false;
	all->stable_secret.initialized = false;

6937 6938 6939 6940
	net->ipv6.devconf_all = all;
	net->ipv6.devconf_dflt = dflt;

#ifdef CONFIG_SYSCTL
6941
	err = __addrconf_sysctl_register(net, "all", NULL, all);
6942 6943 6944
	if (err < 0)
		goto err_reg_all;

6945
	err = __addrconf_sysctl_register(net, "default", NULL, dflt);
6946 6947 6948 6949 6950 6951 6952
	if (err < 0)
		goto err_reg_dflt;
#endif
	return 0;

#ifdef CONFIG_SYSCTL
err_reg_dflt:
6953
	__addrconf_sysctl_unregister(net, all, NETCONFA_IFINDEX_ALL);
6954 6955 6956 6957 6958 6959 6960 6961 6962
err_reg_all:
	kfree(dflt);
#endif
err_alloc_dflt:
	kfree(all);
err_alloc_all:
	return err;
}

6963
static void __net_exit addrconf_exit_net(struct net *net)
6964 6965
{
#ifdef CONFIG_SYSCTL
6966 6967 6968 6969
	__addrconf_sysctl_unregister(net, net->ipv6.devconf_dflt,
				     NETCONFA_IFINDEX_DEFAULT);
	__addrconf_sysctl_unregister(net, net->ipv6.devconf_all,
				     NETCONFA_IFINDEX_ALL);
6970
#endif
Z
zhuyj 已提交
6971 6972
	kfree(net->ipv6.devconf_dflt);
	kfree(net->ipv6.devconf_all);
6973 6974 6975 6976 6977 6978 6979
}

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

6980
static struct rtnl_af_ops inet6_ops __read_mostly = {
T
Thomas Graf 已提交
6981 6982 6983
	.family		  = AF_INET6,
	.fill_link_af	  = inet6_fill_link_af,
	.get_link_af_size = inet6_get_link_af_size,
6984
	.validate_link_af = inet6_validate_link_af,
6985
	.set_link_af	  = inet6_set_link_af,
T
Thomas Graf 已提交
6986 6987
};

L
Linus Torvalds 已提交
6988 6989 6990 6991 6992 6993
/*
 *	Init / cleanup code
 */

int __init addrconf_init(void)
{
6994
	struct inet6_dev *idev;
6995
	int i, err;
6996

S
Stephen Hemminger 已提交
6997 6998
	err = ipv6_addr_label_init();
	if (err < 0) {
6999 7000
		pr_crit("%s: cannot initialize default policy table: %d\n",
			__func__, err);
7001
		goto out;
7002
	}
L
Linus Torvalds 已提交
7003

7004 7005 7006
	err = register_pernet_subsys(&addrconf_ops);
	if (err < 0)
		goto out_addrlabel;
7007

7008 7009 7010 7011 7012 7013
	addrconf_wq = create_workqueue("ipv6_addrconf");
	if (!addrconf_wq) {
		err = -ENOMEM;
		goto out_nowq;
	}

L
Linus Torvalds 已提交
7014 7015 7016 7017 7018 7019 7020 7021 7022 7023 7024 7025 7026 7027 7028 7029 7030 7031 7032
	/* 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();
7033
	idev = ipv6_add_dev(init_net.loopback_dev);
L
Linus Torvalds 已提交
7034
	rtnl_unlock();
7035 7036
	if (IS_ERR(idev)) {
		err = PTR_ERR(idev);
7037
		goto errlo;
7038
	}
L
Linus Torvalds 已提交
7039

7040 7041
	ip6_route_init_special_entries();

7042 7043 7044
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		INIT_HLIST_HEAD(&inet6_addr_lst[i]);

L
Linus Torvalds 已提交
7045 7046
	register_netdevice_notifier(&ipv6_dev_notf);

7047
	addrconf_verify();
7048

7049
	rtnl_af_register(&inet6_ops);
T
Thomas Graf 已提交
7050

7051 7052
	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETLINK,
				   NULL, inet6_dump_ifinfo, 0);
7053 7054 7055
	if (err < 0)
		goto errout;

7056 7057 7058 7059 7060 7061 7062 7063 7064 7065 7066 7067 7068 7069 7070 7071 7072 7073 7074 7075 7076 7077 7078 7079 7080 7081 7082
	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;
7083 7084 7085
	err = ipv6_addr_label_rtnl_register();
	if (err < 0)
		goto errout;
7086

L
Linus Torvalds 已提交
7087
	return 0;
7088
errout:
7089
	rtnl_unregister_all(PF_INET6);
T
Thomas Graf 已提交
7090
	rtnl_af_unregister(&inet6_ops);
7091
	unregister_netdevice_notifier(&ipv6_dev_notf);
7092
errlo:
7093 7094
	destroy_workqueue(addrconf_wq);
out_nowq:
7095
	unregister_pernet_subsys(&addrconf_ops);
7096 7097 7098
out_addrlabel:
	ipv6_addr_label_cleanup();
out:
7099
	return err;
L
Linus Torvalds 已提交
7100 7101
}

7102
void addrconf_cleanup(void)
L
Linus Torvalds 已提交
7103
{
7104
	struct net_device *dev;
L
Linus Torvalds 已提交
7105 7106 7107
	int i;

	unregister_netdevice_notifier(&ipv6_dev_notf);
7108
	unregister_pernet_subsys(&addrconf_ops);
7109
	ipv6_addr_label_cleanup();
L
Linus Torvalds 已提交
7110

7111
	rtnl_af_unregister(&inet6_ops);
L
Linus Torvalds 已提交
7112

7113
	rtnl_lock();
T
Thomas Graf 已提交
7114

7115 7116 7117 7118 7119 7120 7121 7122
	/* 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
7123 7124 7125
	/*
	 *	Check hash table.
	 */
7126
	spin_lock_bh(&addrconf_hash_lock);
7127 7128
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		WARN_ON(!hlist_empty(&inet6_addr_lst[i]));
7129
	spin_unlock_bh(&addrconf_hash_lock);
7130
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
7131
	rtnl_unlock();
7132 7133

	destroy_workqueue(addrconf_wq);
L
Linus Torvalds 已提交
7134
}