addrconf.c 159.4 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 98 99 100

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

#if ACONF_DEBUG >= 3
101
#define ADBG(fmt, ...) printk(fmt, ##__VA_ARGS__)
L
Linus Torvalds 已提交
102
#else
103
#define ADBG(fmt, ...) do { if (0) printk(fmt, ##__VA_ARGS__); } while (0)
L
Linus Torvalds 已提交
104 105 106
#endif

#define	INFINITY_LIFE_TIME	0xFFFFFFFF
107

108 109 110
#define IPV6_MAX_STRLEN \
	sizeof("ffff:ffff:ffff:ffff:ffff:ffff:255.255.255.255")

111 112 113 114
static inline u32 cstamp_delta(unsigned long cstamp)
{
	return (cstamp - INITIAL_JIFFIES) * 100UL / HZ;
}
L
Linus Torvalds 已提交
115

116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136
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
137
#ifdef CONFIG_SYSCTL
138
static int addrconf_sysctl_register(struct inet6_dev *idev);
139 140
static void addrconf_sysctl_unregister(struct inet6_dev *idev);
#else
141
static inline int addrconf_sysctl_register(struct inet6_dev *idev)
142
{
143
	return 0;
144 145 146 147 148
}

static inline void addrconf_sysctl_unregister(struct inet6_dev *idev)
{
}
L
Linus Torvalds 已提交
149 150
#endif

151 152
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
153

154
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev);
L
Linus Torvalds 已提交
155
static int ipv6_count_addresses(struct inet6_dev *idev);
156 157 158
static int ipv6_generate_stable_address(struct in6_addr *addr,
					u8 dad_count,
					const struct inet6_dev *idev);
L
Linus Torvalds 已提交
159 160 161 162

/*
 *	Configured unicast address hash table
 */
163
static struct hlist_head inet6_addr_lst[IN6_ADDR_HSIZE];
164
static DEFINE_SPINLOCK(addrconf_hash_lock);
L
Linus Torvalds 已提交
165

166 167 168
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
169

170 171
static struct workqueue_struct *addrconf_wq;
static DECLARE_DELAYED_WORK(addr_chk_work, addrconf_verify_work);
L
Linus Torvalds 已提交
172 173 174 175

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

176 177
static void addrconf_type_change(struct net_device *dev,
				 unsigned long event);
L
Linus Torvalds 已提交
178 179
static int addrconf_ifdown(struct net_device *dev, int how);

180 181 182 183 184
static struct rt6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags);

185
static void addrconf_dad_start(struct inet6_ifaddr *ifp);
186
static void addrconf_dad_work(struct work_struct *w);
187
static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id);
188
static void addrconf_dad_run(struct inet6_dev *idev);
L
Linus Torvalds 已提交
189 190 191 192
static void addrconf_rs_timer(unsigned long data);
static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);
static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifa);

193
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
194
				struct prefix_info *pinfo);
195 196
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev);
L
Linus Torvalds 已提交
197

198
static struct ipv6_devconf ipv6_devconf __read_mostly = {
L
Linus Torvalds 已提交
199 200 201 202 203 204 205
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
	.force_mld_version	= 0,
206 207
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
208 209 210
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
211
	.rtr_solicit_max_interval = RTR_SOLICITATION_MAX_INTERVAL,
L
Linus Torvalds 已提交
212
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
213
	.use_tempaddr		= 0,
L
Linus Torvalds 已提交
214 215 216 217 218
	.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,
219
	.accept_ra_defrtr	= 1,
220
	.accept_ra_from_local	= 0,
221
	.accept_ra_min_hop_limit= 1,
222
	.accept_ra_pinfo	= 1,
223 224
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
225
	.rtr_probe_interval	= 60 * HZ,
226
#ifdef CONFIG_IPV6_ROUTE_INFO
227
	.accept_ra_rt_info_min_plen = 0,
228 229
	.accept_ra_rt_info_max_plen = 0,
#endif
230
#endif
231
	.proxy_ndp		= 0,
232
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
233
	.disable_ipv6		= 0,
234
	.accept_dad		= 1,
235
	.suppress_frag_ndisc	= 1,
236
	.accept_ra_mtu		= 1,
237 238
	.stable_secret		= {
		.initialized = false,
239 240
	},
	.use_oif_addrs_only	= 0,
241
	.ignore_routes_with_linkdown = 0,
242
	.keep_addr_on_down	= 0,
243
	.seg6_enabled		= 0,
244 245 246
#ifdef CONFIG_IPV6_SEG6_HMAC
	.seg6_require_hmac	= 0,
#endif
247
	.enhanced_dad           = 1,
248
	.addr_gen_mode		= IN6_ADDR_GEN_MODE_EUI64,
249
	.disable_policy		= 0,
L
Linus Torvalds 已提交
250 251
};

252
static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
L
Linus Torvalds 已提交
253 254 255 256 257 258
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
259 260 261
	.force_mld_version	= 0,
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
262 263 264
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
265
	.rtr_solicit_max_interval = RTR_SOLICITATION_MAX_INTERVAL,
L
Linus Torvalds 已提交
266 267 268 269 270 271 272
	.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,
273
	.accept_ra_defrtr	= 1,
274
	.accept_ra_from_local	= 0,
275
	.accept_ra_min_hop_limit= 1,
276
	.accept_ra_pinfo	= 1,
277 278
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
279
	.rtr_probe_interval	= 60 * HZ,
280
#ifdef CONFIG_IPV6_ROUTE_INFO
281
	.accept_ra_rt_info_min_plen = 0,
282 283
	.accept_ra_rt_info_max_plen = 0,
#endif
284
#endif
285
	.proxy_ndp		= 0,
286
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
287
	.disable_ipv6		= 0,
288
	.accept_dad		= 1,
289
	.suppress_frag_ndisc	= 1,
290
	.accept_ra_mtu		= 1,
291 292 293
	.stable_secret		= {
		.initialized = false,
	},
294
	.use_oif_addrs_only	= 0,
295
	.ignore_routes_with_linkdown = 0,
296
	.keep_addr_on_down	= 0,
297
	.seg6_enabled		= 0,
298 299 300
#ifdef CONFIG_IPV6_SEG6_HMAC
	.seg6_require_hmac	= 0,
#endif
301
	.enhanced_dad           = 1,
302
	.addr_gen_mode		= IN6_ADDR_GEN_MODE_EUI64,
303
	.disable_policy		= 0,
L
Linus Torvalds 已提交
304 305
};

306
/* Check if a valid qdisc is available */
307
static inline bool addrconf_qdisc_ok(const struct net_device *dev)
308
{
309
	return !qdisc_tx_is_noop(dev);
310 311
}

312
static void addrconf_del_rs_timer(struct inet6_dev *idev)
L
Linus Torvalds 已提交
313
{
314 315 316 317
	if (del_timer(&idev->rs_timer))
		__in6_dev_put(idev);
}

318
static void addrconf_del_dad_work(struct inet6_ifaddr *ifp)
319
{
320
	if (cancel_delayed_work(&ifp->dad_work))
L
Linus Torvalds 已提交
321 322 323
		__in6_ifa_put(ifp);
}

324 325 326 327 328 329 330
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
331

332 333
static void addrconf_mod_dad_work(struct inet6_ifaddr *ifp,
				   unsigned long delay)
L
Linus Torvalds 已提交
334
{
335 336 337
	in6_ifa_hold(ifp);
	if (mod_delayed_work(addrconf_wq, &ifp->dad_work, delay))
		in6_ifa_put(ifp);
L
Linus Torvalds 已提交
338 339
}

340 341
static int snmp6_alloc_dev(struct inet6_dev *idev)
{
342 343
	int i;

W
WANG Cong 已提交
344 345
	idev->stats.ipv6 = alloc_percpu(struct ipstats_mib);
	if (!idev->stats.ipv6)
346
		goto err_ip;
347 348 349

	for_each_possible_cpu(i) {
		struct ipstats_mib *addrconf_stats;
W
WANG Cong 已提交
350
		addrconf_stats = per_cpu_ptr(idev->stats.ipv6, i);
351 352 353 354
		u64_stats_init(&addrconf_stats->syncp);
	}


355 356 357
	idev->stats.icmpv6dev = kzalloc(sizeof(struct icmpv6_mib_device),
					GFP_KERNEL);
	if (!idev->stats.icmpv6dev)
358
		goto err_icmp;
359 360 361
	idev->stats.icmpv6msgdev = kzalloc(sizeof(struct icmpv6msg_mib_device),
					   GFP_KERNEL);
	if (!idev->stats.icmpv6msgdev)
362
		goto err_icmpmsg;
363 364 365

	return 0;

366
err_icmpmsg:
367
	kfree(idev->stats.icmpv6dev);
368
err_icmp:
W
WANG Cong 已提交
369
	free_percpu(idev->stats.ipv6);
370
err_ip:
371
	return -ENOMEM;
372 373
}

374
static struct inet6_dev *ipv6_add_dev(struct net_device *dev)
L
Linus Torvalds 已提交
375 376
{
	struct inet6_dev *ndev;
377
	int err = -ENOMEM;
L
Linus Torvalds 已提交
378 379 380 381

	ASSERT_RTNL();

	if (dev->mtu < IPV6_MIN_MTU)
382
		return ERR_PTR(-EINVAL);
L
Linus Torvalds 已提交
383

384
	ndev = kzalloc(sizeof(struct inet6_dev), GFP_KERNEL);
385
	if (!ndev)
386
		return ERR_PTR(err);
387 388 389

	rwlock_init(&ndev->lock);
	ndev->dev = dev;
390
	INIT_LIST_HEAD(&ndev->addr_list);
391 392
	setup_timer(&ndev->rs_timer, addrconf_rs_timer,
		    (unsigned long)ndev);
393
	memcpy(&ndev->cnf, dev_net(dev)->ipv6.devconf_dflt, sizeof(ndev->cnf));
394 395

	if (ndev->cnf.stable_secret.initialized)
396
		ndev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
397
	else
398
		ndev->cnf.addr_gen_mode = ipv6_devconf_dflt.addr_gen_mode;
399

400 401
	ndev->cnf.mtu6 = dev->mtu;
	ndev->nd_parms = neigh_parms_alloc(dev, &nd_tbl);
402
	if (!ndev->nd_parms) {
403
		kfree(ndev);
404
		return ERR_PTR(err);
405
	}
406 407
	if (ndev->cnf.forwarding)
		dev_disable_lro(dev);
408 409
	/* We refer to the device */
	dev_hold(dev);
L
Linus Torvalds 已提交
410

411
	if (snmp6_alloc_dev(ndev) < 0) {
412
		ADBG(KERN_WARNING
413
			"%s: cannot allocate memory for statistics; dev=%s.\n",
414
			__func__, dev->name);
415
		neigh_parms_release(&nd_tbl, ndev->nd_parms);
R
Roy Li 已提交
416 417
		dev_put(dev);
		kfree(ndev);
418
		return ERR_PTR(err);
419
	}
L
Linus Torvalds 已提交
420

421
	if (snmp6_register_dev(ndev) < 0) {
422
		ADBG(KERN_WARNING
423
			"%s: cannot create /proc/net/dev_snmp6/%s\n",
424
			__func__, dev->name);
425
		goto err_release;
426 427
	}

428
	/* One reference from device. */
429
	refcount_set(&ndev->refcnt, 1);
L
Linus Torvalds 已提交
430

431 432 433
	if (dev->flags & (IFF_NOARP | IFF_LOOPBACK))
		ndev->cnf.accept_dad = -1;

A
Amerigo Wang 已提交
434
#if IS_ENABLED(CONFIG_IPV6_SIT)
435
	if (dev->type == ARPHRD_SIT && (dev->priv_flags & IFF_ISATAP)) {
436
		pr_info("%s: Disabled Multicast RS\n", dev->name);
437 438 439 440
		ndev->cnf.rtr_solicits = 0;
	}
#endif

441
	INIT_LIST_HEAD(&ndev->tempaddr_list);
J
Jiri Bohac 已提交
442
	ndev->desync_factor = U32_MAX;
443 444
	if ((dev->flags&IFF_LOOPBACK) ||
	    dev->type == ARPHRD_TUNNEL ||
445
	    dev->type == ARPHRD_TUNNEL6 ||
446 447
	    dev->type == ARPHRD_SIT ||
	    dev->type == ARPHRD_NONE) {
448
		ndev->cnf.use_tempaddr = -1;
449 450
	} else
		ipv6_regen_rndid(ndev);
451

452
	ndev->token = in6addr_any;
L
Linus Torvalds 已提交
453

454
	if (netif_running(dev) && addrconf_qdisc_ok(dev))
455 456
		ndev->if_flags |= IF_READY;

457 458
	ipv6_mc_init_dev(ndev);
	ndev->tstamp = jiffies;
459 460 461
	err = addrconf_sysctl_register(ndev);
	if (err) {
		ipv6_mc_destroy_dev(ndev);
462
		snmp6_unregister_dev(ndev);
463 464
		goto err_release;
	}
465
	/* protected by rtnl_lock */
466
	rcu_assign_pointer(dev->ip6_ptr, ndev);
467

468 469 470
	/* Join interface-local all-node multicast group */
	ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allnodes);

471
	/* Join all-node multicast group */
472
	ipv6_dev_mc_inc(dev, &in6addr_linklocal_allnodes);
473

474
	/* Join all-router multicast group if forwarding is set */
L
Li Wei 已提交
475
	if (ndev->cnf.forwarding && (dev->flags & IFF_MULTICAST))
476 477
		ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);

L
Linus Torvalds 已提交
478
	return ndev;
479 480 481 482 483 484

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
485 486
}

487
static struct inet6_dev *ipv6_find_idev(struct net_device *dev)
L
Linus Torvalds 已提交
488 489 490 491 492
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
493 494 495
	idev = __in6_dev_get(dev);
	if (!idev) {
		idev = ipv6_add_dev(dev);
496
		if (IS_ERR(idev))
L
Linus Torvalds 已提交
497 498
			return NULL;
	}
499

L
Linus Torvalds 已提交
500 501 502 503 504
	if (dev->flags&IFF_UP)
		ipv6_mc_up(idev);
	return idev;
}

505 506 507 508
static int inet6_netconf_msgsize_devconf(int type)
{
	int size =  NLMSG_ALIGN(sizeof(struct netconfmsg))
		    + nla_total_size(4);	/* NETCONFA_IFINDEX */
509
	bool all = false;
510

511 512 513 514
	if (type == NETCONFA_ALL)
		all = true;

	if (all || type == NETCONFA_FORWARDING)
515
		size += nla_total_size(4);
516
#ifdef CONFIG_IPV6_MROUTE
517
	if (all || type == NETCONFA_MC_FORWARDING)
518
		size += nla_total_size(4);
519
#endif
520
	if (all || type == NETCONFA_PROXY_NEIGH)
521
		size += nla_total_size(4);
522

523
	if (all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN)
524 525
		size += nla_total_size(4);

526 527 528 529 530 531 532 533 534 535
	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;
536
	bool all = false;
537 538 539

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct netconfmsg),
			flags);
540
	if (!nlh)
541 542
		return -EMSGSIZE;

543 544 545
	if (type == NETCONFA_ALL)
		all = true;

546 547 548 549 550 551
	ncm = nlmsg_data(nlh);
	ncm->ncm_family = AF_INET6;

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

552 553 554
	if (!devconf)
		goto out;

555
	if ((all || type == NETCONFA_FORWARDING) &&
556 557
	    nla_put_s32(skb, NETCONFA_FORWARDING, devconf->forwarding) < 0)
		goto nla_put_failure;
558
#ifdef CONFIG_IPV6_MROUTE
559
	if ((all || type == NETCONFA_MC_FORWARDING) &&
560 561 562
	    nla_put_s32(skb, NETCONFA_MC_FORWARDING,
			devconf->mc_forwarding) < 0)
		goto nla_put_failure;
563
#endif
564
	if ((all || type == NETCONFA_PROXY_NEIGH) &&
565 566 567
	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
		goto nla_put_failure;

568
	if ((all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN) &&
569 570 571 572
	    nla_put_s32(skb, NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
			devconf->ignore_routes_with_linkdown) < 0)
		goto nla_put_failure;

573
out:
574 575
	nlmsg_end(skb, nlh);
	return 0;
576 577 578 579 580 581

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

582 583
void inet6_netconf_notify_devconf(struct net *net, int event, int type,
				  int ifindex, struct ipv6_devconf *devconf)
584 585 586 587
{
	struct sk_buff *skb;
	int err = -ENOBUFS;

588
	skb = nlmsg_new(inet6_netconf_msgsize_devconf(type), GFP_KERNEL);
589
	if (!skb)
590 591 592
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf, 0, 0,
593
					 event, 0, type);
594 595 596 597 598 599
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
600
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_NETCONF, NULL, GFP_KERNEL);
601 602
	return;
errout:
603
	rtnl_set_sk_err(net, RTNLGRP_IPV6_NETCONF, err);
604 605
}

606 607 608
static const struct nla_policy devconf_ipv6_policy[NETCONFA_MAX+1] = {
	[NETCONFA_IFINDEX]	= { .len = sizeof(int) },
	[NETCONFA_FORWARDING]	= { .len = sizeof(int) },
609
	[NETCONFA_PROXY_NEIGH]	= { .len = sizeof(int) },
610
	[NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN]	= { .len = sizeof(int) },
611 612 613
};

static int inet6_netconf_get_devconf(struct sk_buff *in_skb,
614 615
				     struct nlmsghdr *nlh,
				     struct netlink_ext_ack *extack)
616 617 618 619 620 621 622 623 624 625 626 627
{
	struct net *net = sock_net(in_skb->sk);
	struct nlattr *tb[NETCONFA_MAX+1];
	struct netconfmsg *ncm;
	struct sk_buff *skb;
	struct ipv6_devconf *devconf;
	struct inet6_dev *in6_dev;
	struct net_device *dev;
	int ifindex;
	int err;

	err = nlmsg_parse(nlh, sizeof(*ncm), tb, NETCONFA_MAX,
628
			  devconf_ipv6_policy, extack);
629 630 631
	if (err < 0)
		goto errout;

632
	err = -EINVAL;
633 634 635 636 637 638 639 640 641 642 643 644 645
	if (!tb[NETCONFA_IFINDEX])
		goto errout;

	ifindex = nla_get_s32(tb[NETCONFA_IFINDEX]);
	switch (ifindex) {
	case NETCONFA_IFINDEX_ALL:
		devconf = net->ipv6.devconf_all;
		break;
	case NETCONFA_IFINDEX_DEFAULT:
		devconf = net->ipv6.devconf_dflt;
		break;
	default:
		dev = __dev_get_by_index(net, ifindex);
646
		if (!dev)
647 648
			goto errout;
		in6_dev = __in6_dev_get(dev);
649
		if (!in6_dev)
650 651 652 653 654 655
			goto errout;
		devconf = &in6_dev->cnf;
		break;
	}

	err = -ENOBUFS;
656
	skb = nlmsg_new(inet6_netconf_msgsize_devconf(NETCONFA_ALL), GFP_ATOMIC);
657
	if (!skb)
658 659 660 661 662
		goto errout;

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf,
					 NETLINK_CB(in_skb).portid,
					 nlh->nlmsg_seq, RTM_NEWNETCONF, 0,
663
					 NETCONFA_ALL);
664 665 666 667 668 669 670 671 672 673 674
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
errout:
	return err;
}

675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691
static int inet6_netconf_dump_devconf(struct sk_buff *skb,
				      struct netlink_callback *cb)
{
	struct net *net = sock_net(skb->sk);
	int h, s_h;
	int idx, s_idx;
	struct net_device *dev;
	struct inet6_dev *idev;
	struct hlist_head *head;

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

	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
		rcu_read_lock();
692 693
		cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^
			  net->dev_base_seq;
694 695 696 697 698 699 700 701 702 703 704 705 706
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;

			if (inet6_netconf_fill_devconf(skb, dev->ifindex,
						       &idev->cnf,
						       NETLINK_CB(cb->skb).portid,
						       cb->nlh->nlmsg_seq,
						       RTM_NEWNETCONF,
						       NLM_F_MULTI,
707
						       NETCONFA_ALL) < 0) {
708 709 710
				rcu_read_unlock();
				goto done;
			}
711
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
712 713 714 715 716 717 718 719 720 721 722
cont:
			idx++;
		}
		rcu_read_unlock();
	}
	if (h == NETDEV_HASHENTRIES) {
		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_ALL,
					       net->ipv6.devconf_all,
					       NETLINK_CB(cb->skb).portid,
					       cb->nlh->nlmsg_seq,
					       RTM_NEWNETCONF, NLM_F_MULTI,
723
					       NETCONFA_ALL) < 0)
724 725 726 727 728 729 730 731 732 733
			goto done;
		else
			h++;
	}
	if (h == NETDEV_HASHENTRIES + 1) {
		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_DEFAULT,
					       net->ipv6.devconf_dflt,
					       NETLINK_CB(cb->skb).portid,
					       cb->nlh->nlmsg_seq,
					       RTM_NEWNETCONF, NLM_F_MULTI,
734
					       NETCONFA_ALL) < 0)
735 736 737 738 739 740 741 742 743 744 745
			goto done;
		else
			h++;
	}
done:
	cb->args[0] = h;
	cb->args[1] = idx;

	return skb->len;
}

L
Linus Torvalds 已提交
746 747 748 749 750 751 752 753 754
#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;
755 756
	if (idev->cnf.forwarding)
		dev_disable_lro(dev);
A
Amerigo Wang 已提交
757
	if (dev->flags & IFF_MULTICAST) {
758
		if (idev->cnf.forwarding) {
759
			ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);
760 761 762
			ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_inc(dev, &in6addr_sitelocal_allrouters);
		} else {
763
			ipv6_dev_mc_dec(dev, &in6addr_linklocal_allrouters);
764 765 766
			ipv6_dev_mc_dec(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_dec(dev, &in6addr_sitelocal_allrouters);
		}
L
Linus Torvalds 已提交
767
	}
768 769

	list_for_each_entry(ifa, &idev->addr_list, if_list) {
M
Michal Wrobel 已提交
770 771
		if (ifa->flags&IFA_F_TENTATIVE)
			continue;
L
Linus Torvalds 已提交
772 773 774 775 776
		if (idev->cnf.forwarding)
			addrconf_join_anycast(ifa);
		else
			addrconf_leave_anycast(ifa);
	}
777 778
	inet6_netconf_notify_devconf(dev_net(dev), RTM_NEWNETCONF,
				     NETCONFA_FORWARDING,
779
				     dev->ifindex, &idev->cnf);
L
Linus Torvalds 已提交
780 781 782
}


783
static void addrconf_forward_change(struct net *net, __s32 newf)
L
Linus Torvalds 已提交
784 785 786 787
{
	struct net_device *dev;
	struct inet6_dev *idev;

788
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
789 790
		idev = __in6_dev_get(dev);
		if (idev) {
791 792
			int changed = (!idev->cnf.forwarding) ^ (!newf);
			idev->cnf.forwarding = newf;
L
Linus Torvalds 已提交
793 794 795 796 797
			if (changed)
				dev_forward_change(idev);
		}
	}
}
798

799
static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
800
{
801
	struct net *net;
802 803 804 805
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
806 807

	net = (struct net *)table->extra2;
808 809
	old = *p;
	*p = newf;
810

811
	if (p == &net->ipv6.devconf_dflt->forwarding) {
812
		if ((!newf) ^ (!old))
813 814
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
815 816
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
817 818
		rtnl_unlock();
		return 0;
E
Eric W. Biederman 已提交
819
	}
820

821
	if (p == &net->ipv6.devconf_all->forwarding) {
822 823
		int old_dflt = net->ipv6.devconf_dflt->forwarding;

824
		net->ipv6.devconf_dflt->forwarding = newf;
825
		if ((!newf) ^ (!old_dflt))
826 827
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
828 829 830
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);

831
		addrconf_forward_change(net, newf);
832
		if ((!newf) ^ (!old))
833 834
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_FORWARDING,
835 836
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
837
	} else if ((!newf) ^ (!old))
838
		dev_forward_change((struct inet6_dev *)table->extra1);
839
	rtnl_unlock();
840

841
	if (newf)
842
		rt6_purge_dflt_routers(net);
843
	return 1;
844
}
845 846 847 848 849 850 851 852 853 854 855 856 857 858

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),
859
							     RTM_NEWNETCONF,
860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881
							     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,
882
						     RTM_NEWNETCONF,
883 884 885 886 887 888 889 890 891 892 893 894
						     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,
895
						     RTM_NEWNETCONF,
896 897 898 899 900 901 902 903 904
						     NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
	}
	rtnl_unlock();

	return 1;
}

L
Linus Torvalds 已提交
905 906
#endif

907
/* Nobody refers to this ifaddr, destroy it */
L
Linus Torvalds 已提交
908 909
void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp)
{
910
	WARN_ON(!hlist_unhashed(&ifp->addr_lst));
911

L
Linus Torvalds 已提交
912
#ifdef NET_REFCNT_DEBUG
913
	pr_debug("%s\n", __func__);
L
Linus Torvalds 已提交
914 915 916 917
#endif

	in6_dev_put(ifp->idev);

918 919 920
	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
921

922
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
923
		pr_warn("Freeing alive inet6 address %p\n", ifp);
L
Linus Torvalds 已提交
924 925
		return;
	}
A
Amerigo Wang 已提交
926
	ip6_rt_put(ifp->rt);
L
Linus Torvalds 已提交
927

928
	kfree_rcu(ifp, rcu);
L
Linus Torvalds 已提交
929 930
}

B
Brian Haley 已提交
931 932 933
static void
ipv6_link_dev_addr(struct inet6_dev *idev, struct inet6_ifaddr *ifp)
{
934
	struct list_head *p;
935
	int ifp_scope = ipv6_addr_src_scope(&ifp->addr);
B
Brian Haley 已提交
936 937 938 939 940

	/*
	 * Each device address list is sorted in order of scope -
	 * global before linklocal.
	 */
941 942 943
	list_for_each(p, &idev->addr_list) {
		struct inet6_ifaddr *ifa
			= list_entry(p, struct inet6_ifaddr, if_list);
944
		if (ifp_scope >= ipv6_addr_src_scope(&ifa->addr))
B
Brian Haley 已提交
945 946 947
			break;
	}

948
	list_add_tail(&ifp->if_list, p);
B
Brian Haley 已提交
949 950
}

E
Eric Dumazet 已提交
951
static u32 inet6_addr_hash(const struct in6_addr *addr)
952
{
E
Eric Dumazet 已提交
953
	return hash_32(ipv6_addr_hash(addr), IN6_ADDR_HSIZE_SHIFT);
954 955
}

L
Linus Torvalds 已提交
956 957 958
/* On success it returns ifp with increased reference count */

static struct inet6_ifaddr *
959 960
ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
	      const struct in6_addr *peer_addr, int pfxlen,
961
	      int scope, u32 flags, u32 valid_lft, u32 prefered_lft)
L
Linus Torvalds 已提交
962
{
963
	struct net *net = dev_net(idev->dev);
L
Linus Torvalds 已提交
964 965
	struct inet6_ifaddr *ifa = NULL;
	struct rt6_info *rt;
966
	struct in6_validator_info i6vi;
967
	unsigned int hash;
L
Linus Torvalds 已提交
968
	int err = 0;
969 970 971 972 973 974 975
	int addr_type = ipv6_addr_type(addr);

	if (addr_type == IPV6_ADDR_ANY ||
	    addr_type & IPV6_ADDR_MULTICAST ||
	    (!(idev->dev->flags & IFF_LOOPBACK) &&
	     addr_type & IPV6_ADDR_LOOPBACK))
		return ERR_PTR(-EADDRNOTAVAIL);
L
Linus Torvalds 已提交
976

977
	rcu_read_lock_bh();
978 979 980

	in6_dev_hold(idev);

L
Linus Torvalds 已提交
981 982 983 984 985
	if (idev->dead) {
		err = -ENODEV;			/*XXX*/
		goto out2;
	}

986
	if (idev->cnf.disable_ipv6) {
987 988 989 990
		err = -EACCES;
		goto out2;
	}

991 992 993 994 995 996 997 998 999 1000 1001
	i6vi.i6vi_addr = *addr;
	i6vi.i6vi_dev = idev;
	rcu_read_unlock_bh();

	err = inet6addr_validator_notifier_call_chain(NETDEV_UP, &i6vi);

	rcu_read_lock_bh();
	err = notifier_to_errno(err);
	if (err)
		goto out2;

1002
	spin_lock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
1003 1004

	/* Ignore adding duplicate addresses on an interface */
1005
	if (ipv6_chk_same_addr(dev_net(idev->dev), addr, idev->dev)) {
1006
		ADBG("ipv6_add_addr: already assigned\n");
L
Linus Torvalds 已提交
1007 1008 1009 1010
		err = -EEXIST;
		goto out;
	}

1011
	ifa = kzalloc(sizeof(struct inet6_ifaddr), GFP_ATOMIC);
L
Linus Torvalds 已提交
1012

1013
	if (!ifa) {
1014
		ADBG("ipv6_add_addr: malloc failed\n");
L
Linus Torvalds 已提交
1015 1016 1017 1018
		err = -ENOBUFS;
		goto out;
	}

1019
	rt = addrconf_dst_alloc(idev, addr, false);
L
Linus Torvalds 已提交
1020 1021 1022 1023 1024
	if (IS_ERR(rt)) {
		err = PTR_ERR(rt);
		goto out;
	}

1025 1026 1027 1028
	if (net->ipv6.devconf_all->disable_policy ||
	    idev->cnf.disable_policy)
		rt->dst.flags |= DST_NOPOLICY;

1029 1030
	neigh_parms_data_state_setall(idev->nd_parms);

A
Alexey Dobriyan 已提交
1031
	ifa->addr = *addr;
1032 1033
	if (peer_addr)
		ifa->peer_addr = *peer_addr;
L
Linus Torvalds 已提交
1034 1035

	spin_lock_init(&ifa->lock);
1036
	INIT_DELAYED_WORK(&ifa->dad_work, addrconf_dad_work);
1037
	INIT_HLIST_NODE(&ifa->addr_lst);
L
Linus Torvalds 已提交
1038 1039
	ifa->scope = scope;
	ifa->prefix_len = pfxlen;
1040 1041 1042 1043
	ifa->flags = flags;
	/* No need to add the TENTATIVE flag for addresses with NODAD */
	if (!(flags & IFA_F_NODAD))
		ifa->flags |= IFA_F_TENTATIVE;
1044 1045
	ifa->valid_lft = valid_lft;
	ifa->prefered_lft = prefered_lft;
L
Linus Torvalds 已提交
1046
	ifa->cstamp = ifa->tstamp = jiffies;
1047
	ifa->tokenized = false;
L
Linus Torvalds 已提交
1048

1049 1050
	ifa->rt = rt;

L
Linus Torvalds 已提交
1051 1052
	ifa->idev = idev;
	/* For caller */
1053
	refcount_set(&ifa->refcnt, 1);
L
Linus Torvalds 已提交
1054 1055

	/* Add to big hash table */
E
Eric Dumazet 已提交
1056
	hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
1057

1058 1059
	hlist_add_head_rcu(&ifa->addr_lst, &inet6_addr_lst[hash]);
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
1060 1061 1062

	write_lock(&idev->lock);
	/* Add to inet6_dev unicast addr list. */
B
Brian Haley 已提交
1063
	ipv6_link_dev_addr(idev, ifa);
L
Linus Torvalds 已提交
1064 1065

	if (ifa->flags&IFA_F_TEMPORARY) {
1066
		list_add(&ifa->tmp_list, &idev->tempaddr_list);
L
Linus Torvalds 已提交
1067 1068 1069 1070 1071 1072
		in6_ifa_hold(ifa);
	}

	in6_ifa_hold(ifa);
	write_unlock(&idev->lock);
out2:
1073
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1074

1075
	if (likely(err == 0))
1076
		inet6addr_notifier_call_chain(NETDEV_UP, ifa);
L
Linus Torvalds 已提交
1077 1078
	else {
		kfree(ifa);
1079
		in6_dev_put(idev);
L
Linus Torvalds 已提交
1080 1081 1082 1083 1084
		ifa = ERR_PTR(err);
	}

	return ifa;
out:
1085
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
1086 1087 1088
	goto out2;
}

1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170
enum cleanup_prefix_rt_t {
	CLEANUP_PREFIX_RT_NOP,    /* no cleanup action for prefix route */
	CLEANUP_PREFIX_RT_DEL,    /* delete the prefix route */
	CLEANUP_PREFIX_RT_EXPIRE, /* update the lifetime of the prefix route */
};

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

	*expires = jiffies;

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

		action = CLEANUP_PREFIX_RT_EXPIRE;

		spin_lock(&ifa->lock);

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

	return action;
}

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

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


L
Linus Torvalds 已提交
1171 1172 1173 1174
/* This function wants to get referenced ifp and releases it before return */

static void ipv6_del_addr(struct inet6_ifaddr *ifp)
{
1175
	int state;
1176 1177
	enum cleanup_prefix_rt_t action = CLEANUP_PREFIX_RT_NOP;
	unsigned long expires;
L
Linus Torvalds 已提交
1178

1179 1180
	ASSERT_RTNL();

1181
	spin_lock_bh(&ifp->lock);
1182
	state = ifp->state;
1183
	ifp->state = INET6_IFADDR_STATE_DEAD;
1184
	spin_unlock_bh(&ifp->lock);
1185 1186 1187

	if (state == INET6_IFADDR_STATE_DEAD)
		goto out;
L
Linus Torvalds 已提交
1188

1189 1190 1191
	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
1192

1193
	write_lock_bh(&ifp->idev->lock);
1194

L
Linus Torvalds 已提交
1195
	if (ifp->flags&IFA_F_TEMPORARY) {
1196 1197 1198 1199
		list_del(&ifp->tmp_list);
		if (ifp->ifpub) {
			in6_ifa_put(ifp->ifpub);
			ifp->ifpub = NULL;
L
Linus Torvalds 已提交
1200
		}
1201
		__in6_ifa_put(ifp);
L
Linus Torvalds 已提交
1202 1203
	}

1204 1205
	if (ifp->flags & IFA_F_PERMANENT && !(ifp->flags & IFA_F_NOPREFIXROUTE))
		action = check_cleanup_prefix_route(ifp, &expires);
1206

1207 1208 1209 1210
	list_del_init(&ifp->if_list);
	__in6_ifa_put(ifp);

	write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
1211

1212
	addrconf_del_dad_work(ifp);
1213

L
Linus Torvalds 已提交
1214 1215
	ipv6_ifa_notify(RTM_DELADDR, ifp);

1216
	inet6addr_notifier_call_chain(NETDEV_DOWN, ifp);
L
Linus Torvalds 已提交
1217

1218 1219 1220
	if (action != CLEANUP_PREFIX_RT_NOP) {
		cleanup_prefix_route(ifp, expires,
			action == CLEANUP_PREFIX_RT_DEL);
L
Linus Torvalds 已提交
1221 1222
	}

1223 1224
	/* clean up prefsrc entries */
	rt6_remove_prefsrc(ifp);
1225
out:
L
Linus Torvalds 已提交
1226 1227 1228 1229 1230 1231 1232
	in6_ifa_put(ifp);
}

static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp, struct inet6_ifaddr *ift)
{
	struct inet6_dev *idev = ifp->idev;
	struct in6_addr addr, *tmpaddr;
1233
	unsigned long tmp_prefered_lft, tmp_valid_lft, tmp_tstamp, age;
1234
	unsigned long regen_advance;
L
Linus Torvalds 已提交
1235 1236
	int tmp_plen;
	int ret = 0;
1237
	u32 addr_flags;
1238
	unsigned long now = jiffies;
J
Jiri Bohac 已提交
1239
	long max_desync_factor;
1240
	s32 cnf_temp_preferred_lft;
L
Linus Torvalds 已提交
1241

1242
	write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253
	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) {
1254
		write_unlock_bh(&idev->lock);
1255
		pr_info("%s: use_tempaddr is disabled\n", __func__);
L
Linus Torvalds 已提交
1256 1257 1258 1259 1260 1261 1262 1263
		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);
1264
		write_unlock_bh(&idev->lock);
1265 1266
		pr_warn("%s: regeneration time exceeded - disabled temporary address support\n",
			__func__);
L
Linus Torvalds 已提交
1267 1268 1269 1270 1271 1272
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}
	in6_ifa_hold(ifp);
	memcpy(addr.s6_addr, ifp->addr.s6_addr, 8);
1273
	ipv6_try_regen_rndid(idev, tmpaddr);
L
Linus Torvalds 已提交
1274
	memcpy(&addr.s6_addr[8], idev->rndid, 8);
1275
	age = (now - ifp->tstamp) / HZ;
J
Jiri Bohac 已提交
1276 1277 1278 1279 1280 1281 1282 1283

	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
	 */
1284
	cnf_temp_preferred_lft = READ_ONCE(idev->cnf.temp_prefered_lft);
J
Jiri Bohac 已提交
1285 1286
	max_desync_factor = min_t(__u32,
				  idev->cnf.max_desync_factor,
1287
				  cnf_temp_preferred_lft - regen_advance);
J
Jiri Bohac 已提交
1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298

	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;
		}
	}

L
Linus Torvalds 已提交
1299 1300
	tmp_valid_lft = min_t(__u32,
			      ifp->valid_lft,
1301
			      idev->cnf.temp_valid_lft + age);
1302
	tmp_prefered_lft = cnf_temp_preferred_lft + age -
J
Jiri Bohac 已提交
1303 1304
			    idev->desync_factor;
	tmp_prefered_lft = min_t(__u32, ifp->prefered_lft, tmp_prefered_lft);
L
Linus Torvalds 已提交
1305 1306 1307 1308
	tmp_plen = ifp->prefix_len;
	tmp_tstamp = ifp->tstamp;
	spin_unlock_bh(&ifp->lock);

1309
	write_unlock_bh(&idev->lock);
1310

1311 1312 1313 1314
	/* 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.
1315 1316
	 * Use age calculation as in addrconf_verify to avoid unnecessary
	 * temporary addresses being generated.
1317
	 */
1318 1319
	age = (now - tmp_tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
	if (tmp_prefered_lft <= regen_advance + age) {
1320 1321 1322 1323 1324 1325
		in6_ifa_put(ifp);
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}

1326 1327 1328 1329 1330
	addr_flags = IFA_F_TEMPORARY;
	/* set in addrconf_prefix_rcv() */
	if (ifp->flags & IFA_F_OPTIMISTIC)
		addr_flags |= IFA_F_OPTIMISTIC;

1331 1332 1333 1334
	ift = ipv6_add_addr(idev, &addr, NULL, tmp_plen,
			    ipv6_addr_scope(&addr), addr_flags,
			    tmp_valid_lft, tmp_prefered_lft);
	if (IS_ERR(ift)) {
L
Linus Torvalds 已提交
1335 1336
		in6_ifa_put(ifp);
		in6_dev_put(idev);
1337
		pr_info("%s: retry temporary address regeneration\n", __func__);
L
Linus Torvalds 已提交
1338
		tmpaddr = &addr;
1339
		write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1340 1341 1342 1343 1344
		goto retry;
	}

	spin_lock_bh(&ift->lock);
	ift->ifpub = ifp;
1345
	ift->cstamp = now;
L
Linus Torvalds 已提交
1346 1347 1348
	ift->tstamp = tmp_tstamp;
	spin_unlock_bh(&ift->lock);

1349
	addrconf_dad_start(ift);
L
Linus Torvalds 已提交
1350 1351 1352 1353 1354 1355 1356
	in6_ifa_put(ift);
	in6_dev_put(idev);
out:
	return ret;
}

/*
1357
 *	Choose an appropriate source address (RFC3484)
L
Linus Torvalds 已提交
1358
 */
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371
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,
1372 1373 1374
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	IPV6_SADDR_RULE_NOT_OPTIMISTIC,
#endif
1375 1376 1377
	IPV6_SADDR_RULE_MAX
};

1378
struct ipv6_saddr_score {
1379 1380 1381 1382 1383 1384
	int			rule;
	int			addr_type;
	struct inet6_ifaddr	*ifa;
	DECLARE_BITMAP(scorebits, IPV6_SADDR_RULE_MAX);
	int			scopedist;
	int			matchlen;
1385 1386
};

1387
struct ipv6_saddr_dst {
1388
	const struct in6_addr *addr;
1389 1390 1391
	int ifindex;
	int scope;
	int label;
1392
	unsigned int prefs;
1393
};
1394

D
Dave Jones 已提交
1395
static inline int ipv6_saddr_preferred(int type)
L
Linus Torvalds 已提交
1396
{
U
Ulrich Weber 已提交
1397
	if (type & (IPV6_ADDR_MAPPED|IPV6_ADDR_COMPATv4|IPV6_ADDR_LOOPBACK))
1398 1399
		return 1;
	return 0;
L
Linus Torvalds 已提交
1400 1401
}

1402 1403 1404 1405 1406 1407 1408 1409 1410
static inline bool ipv6_use_optimistic_addr(struct inet6_dev *idev)
{
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	return idev && idev->cnf.optimistic_dad && idev->cnf.use_optimistic;
#else
	return false;
#endif
}

1411 1412
static int ipv6_get_saddr_eval(struct net *net,
			       struct ipv6_saddr_score *score,
1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451
			       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
1452
		 *  B-15 |    \        if scope is enough for destination.
1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470
		 *       |             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:
1471
	    {
1472
		/* Rule 3: Avoid deprecated and optimistic addresses */
1473 1474 1475 1476
		u8 avoid = IFA_F_DEPRECATED;

		if (!ipv6_use_optimistic_addr(score->ifa->idev))
			avoid |= IFA_F_OPTIMISTIC;
1477
		ret = ipv6_saddr_preferred(score->addr_type) ||
1478
		      !(score->ifa->flags & avoid);
1479
		break;
1480
	    }
1481 1482
#ifdef CONFIG_IPV6_MIP6
	case IPV6_SADDR_RULE_HOA:
1483
	    {
1484
		/* Rule 4: Prefer home address */
1485 1486
		int prefhome = !(dst->prefs & IPV6_PREFER_SRC_COA);
		ret = !(score->ifa->flags & IFA_F_HOMEADDRESS) ^ prefhome;
1487
		break;
1488
	    }
1489 1490 1491 1492 1493 1494 1495 1496
#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 */
1497 1498
		ret = ipv6_addr_label(net,
				      &score->ifa->addr, score->addr_type,
1499 1500 1501
				      score->ifa->idev->dev->ifindex) == dst->label;
		break;
	case IPV6_SADDR_RULE_PRIVACY:
1502
	    {
1503
		/* Rule 7: Prefer public address
L
Lucas De Marchi 已提交
1504
		 * Note: prefer temporary address if use_tempaddr >= 2
1505
		 */
1506 1507 1508 1509
		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;
1510
		break;
1511
	    }
1512 1513 1514 1515 1516 1517 1518 1519 1520
	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 */
1521 1522 1523 1524
		ret = ipv6_addr_diff(&score->ifa->addr, dst->addr);
		if (ret > score->ifa->prefix_len)
			ret = score->ifa->prefix_len;
		score->matchlen = ret;
1525
		break;
1526 1527 1528 1529 1530 1531 1532 1533
#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
1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544
	default:
		ret = 0;
	}

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

1545 1546 1547 1548 1549
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)
1550
{
1551
	struct ipv6_saddr_score *score = &scores[1 - hiscore_idx], *hiscore = &scores[hiscore_idx];
1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609

	read_lock_bh(&idev->lock);
	list_for_each_entry(score->ifa, &idev->addr_list, if_list) {
		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) {
				if (hiscore->ifa)
					in6_ifa_put(hiscore->ifa);

				in6_ifa_hold(score->ifa);

				swap(hiscore, score);
1610
				hiscore_idx = 1 - hiscore_idx;
1611 1612 1613 1614 1615 1616 1617 1618 1619 1620

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

				break;
			}
		}
	}
out:
	read_unlock_bh(&idev->lock);
1621
	return hiscore_idx;
1622 1623
}

1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645
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;
}

1646
int ipv6_dev_get_saddr(struct net *net, const struct net_device *dst_dev,
1647
		       const struct in6_addr *daddr, unsigned int prefs,
1648
		       struct in6_addr *saddr)
L
Linus Torvalds 已提交
1649
{
1650
	struct ipv6_saddr_score scores[2], *hiscore;
1651
	struct ipv6_saddr_dst dst;
1652
	struct inet6_dev *idev;
1653
	struct net_device *dev;
1654
	int dst_type;
1655
	bool use_oif_addr = false;
1656
	int hiscore_idx = 0;
L
Linus Torvalds 已提交
1657

1658 1659 1660 1661
	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);
1662
	dst.label = ipv6_addr_label(net, daddr, dst_type, dst.ifindex);
1663
	dst.prefs = prefs;
1664

1665 1666
	scores[hiscore_idx].rule = -1;
	scores[hiscore_idx].ifa = NULL;
L
Linus Torvalds 已提交
1667

1668
	rcu_read_lock();
L
Linus Torvalds 已提交
1669

1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680
	/* 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.)
1681 1682 1683 1684
	 *  - "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)
1685 1686
	 */
	if (dst_dev) {
1687
		idev = __in6_dev_get(dst_dev);
1688
		if ((dst_type & IPV6_ADDR_MULTICAST) ||
1689 1690
		    dst.scope <= IPV6_ADDR_SCOPE_LINKLOCAL ||
		    (idev && idev->cnf.use_oif_addrs_only)) {
1691 1692 1693
			use_oif_addr = true;
		}
	}
1694

1695
	if (use_oif_addr) {
1696
		if (idev)
1697
			hiscore_idx = __ipv6_dev_get_saddr(net, &dst, idev, scores, hiscore_idx);
1698
	} else {
1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717
		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;
		}

1718
		for_each_netdev_rcu(net, dev) {
1719 1720 1721 1722 1723
			/* only consider addresses on devices in the
			 * same L3 domain
			 */
			if (l3mdev_master_ifindex_rcu(dev) != master_idx)
				continue;
1724 1725
			idev = __in6_dev_get(dev);
			if (!idev)
1726
				continue;
1727
			hiscore_idx = __ipv6_dev_get_saddr(net, &dst, idev, scores, hiscore_idx);
L
Linus Torvalds 已提交
1728 1729
		}
	}
1730 1731

out:
1732
	rcu_read_unlock();
L
Linus Torvalds 已提交
1733

1734
	hiscore = &scores[hiscore_idx];
1735
	if (!hiscore->ifa)
1736
		return -EADDRNOTAVAIL;
1737

A
Alexey Dobriyan 已提交
1738
	*saddr = hiscore->ifa->addr;
1739
	in6_ifa_put(hiscore->ifa);
1740
	return 0;
L
Linus Torvalds 已提交
1741
}
1742
EXPORT_SYMBOL(ipv6_dev_get_saddr);
L
Linus Torvalds 已提交
1743

1744
int __ipv6_get_lladdr(struct inet6_dev *idev, struct in6_addr *addr,
1745
		      u32 banned_flags)
1746 1747 1748 1749
{
	struct inet6_ifaddr *ifp;
	int err = -EADDRNOTAVAIL;

1750 1751 1752
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762
		if (ifp->scope == IFA_LINK &&
		    !(ifp->flags & banned_flags)) {
			*addr = ifp->addr;
			err = 0;
			break;
		}
	}
	return err;
}

1763
int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr,
1764
		    u32 banned_flags)
L
Linus Torvalds 已提交
1765 1766 1767 1768
{
	struct inet6_dev *idev;
	int err = -EADDRNOTAVAIL;

1769
	rcu_read_lock();
S
Stephen Hemminger 已提交
1770 1771
	idev = __in6_dev_get(dev);
	if (idev) {
L
Linus Torvalds 已提交
1772
		read_lock_bh(&idev->lock);
1773
		err = __ipv6_get_lladdr(idev, addr, banned_flags);
L
Linus Torvalds 已提交
1774 1775
		read_unlock_bh(&idev->lock);
	}
1776
	rcu_read_unlock();
L
Linus Torvalds 已提交
1777 1778 1779 1780 1781 1782 1783 1784 1785
	return err;
}

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

	read_lock_bh(&idev->lock);
1786
	list_for_each_entry(ifp, &idev->addr_list, if_list)
L
Linus Torvalds 已提交
1787 1788 1789 1790 1791
		cnt++;
	read_unlock_bh(&idev->lock);
	return cnt;
}

1792
int ipv6_chk_addr(struct net *net, const struct in6_addr *addr,
1793
		  const struct net_device *dev, int strict)
1794 1795 1796 1797 1798 1799 1800 1801
{
	return ipv6_chk_addr_and_flags(net, addr, dev, strict, IFA_F_TENTATIVE);
}
EXPORT_SYMBOL(ipv6_chk_addr);

int ipv6_chk_addr_and_flags(struct net *net, const struct in6_addr *addr,
			    const struct net_device *dev, int strict,
			    u32 banned_flags)
L
Linus Torvalds 已提交
1802
{
1803
	struct inet6_ifaddr *ifp;
E
Eric Dumazet 已提交
1804
	unsigned int hash = inet6_addr_hash(addr);
1805
	u32 ifp_flags;
L
Linus Torvalds 已提交
1806

1807
	rcu_read_lock_bh();
1808
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
1809
		if (!net_eq(dev_net(ifp->idev->dev), net))
1810
			continue;
1811 1812 1813 1814 1815 1816
		/* 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
1817
		if (ipv6_addr_equal(&ifp->addr, addr) &&
1818
		    !(ifp_flags&banned_flags) &&
1819
		    (!dev || ifp->idev->dev == dev ||
1820 1821 1822
		     !(ifp->scope&(IFA_LINK|IFA_HOST) || strict))) {
			rcu_read_unlock_bh();
			return 1;
L
Linus Torvalds 已提交
1823 1824
		}
	}
1825

1826 1827
	rcu_read_unlock_bh();
	return 0;
L
Linus Torvalds 已提交
1828
}
1829
EXPORT_SYMBOL(ipv6_chk_addr_and_flags);
1830

1831 1832
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev)
L
Linus Torvalds 已提交
1833
{
E
Eric Dumazet 已提交
1834
	unsigned int hash = inet6_addr_hash(addr);
1835
	struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
1836

1837
	hlist_for_each_entry(ifp, &inet6_addr_lst[hash], addr_lst) {
1838
		if (!net_eq(dev_net(ifp->idev->dev), net))
1839
			continue;
L
Linus Torvalds 已提交
1840
		if (ipv6_addr_equal(&ifp->addr, addr)) {
1841
			if (!dev || ifp->idev->dev == dev)
1842
				return true;
L
Linus Torvalds 已提交
1843 1844
		}
	}
1845
	return false;
L
Linus Torvalds 已提交
1846 1847
}

1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874
/* Compares an address/prefix_len with addresses on device @dev.
 * If one is found it returns true.
 */
bool ipv6_chk_custom_prefix(const struct in6_addr *addr,
	const unsigned int prefix_len, struct net_device *dev)
{
	struct inet6_dev *idev;
	struct inet6_ifaddr *ifa;
	bool ret = false;

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

	return ret;
}
EXPORT_SYMBOL(ipv6_chk_custom_prefix);

1875
int ipv6_chk_prefix(const struct in6_addr *addr, struct net_device *dev)
1876 1877 1878 1879 1880 1881 1882 1883 1884 1885
{
	struct inet6_dev *idev;
	struct inet6_ifaddr *ifa;
	int	onlink;

	onlink = 0;
	rcu_read_lock();
	idev = __in6_dev_get(dev);
	if (idev) {
		read_lock_bh(&idev->lock);
1886
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898
			onlink = ipv6_prefix_equal(addr, &ifa->addr,
						   ifa->prefix_len);
			if (onlink)
				break;
		}
		read_unlock_bh(&idev->lock);
	}
	rcu_read_unlock();
	return onlink;
}
EXPORT_SYMBOL(ipv6_chk_prefix);

1899
struct inet6_ifaddr *ipv6_get_ifaddr(struct net *net, const struct in6_addr *addr,
1900
				     struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1901
{
1902
	struct inet6_ifaddr *ifp, *result = NULL;
E
Eric Dumazet 已提交
1903
	unsigned int hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
1904

1905
	rcu_read_lock_bh();
1906
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
1907
		if (!net_eq(dev_net(ifp->idev->dev), net))
1908
			continue;
L
Linus Torvalds 已提交
1909
		if (ipv6_addr_equal(&ifp->addr, addr)) {
1910
			if (!dev || ifp->idev->dev == dev ||
L
Linus Torvalds 已提交
1911
			    !(ifp->scope&(IFA_LINK|IFA_HOST) || strict)) {
1912
				result = ifp;
L
Linus Torvalds 已提交
1913 1914 1915 1916 1917
				in6_ifa_hold(ifp);
				break;
			}
		}
	}
1918
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1919

1920
	return result;
L
Linus Torvalds 已提交
1921 1922 1923 1924
}

/* Gets referenced address, destroys ifaddr */

B
Brian Haley 已提交
1925
static void addrconf_dad_stop(struct inet6_ifaddr *ifp, int dad_failed)
L
Linus Torvalds 已提交
1926
{
1927 1928 1929
	if (dad_failed)
		ifp->flags |= IFA_F_DADFAILED;

1930
	if (ifp->flags&IFA_F_TEMPORARY) {
L
Linus Torvalds 已提交
1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942
		struct inet6_ifaddr *ifpub;
		spin_lock_bh(&ifp->lock);
		ifpub = ifp->ifpub;
		if (ifpub) {
			in6_ifa_hold(ifpub);
			spin_unlock_bh(&ifp->lock);
			ipv6_create_tempaddr(ifpub, ifp);
			in6_ifa_put(ifpub);
		} else {
			spin_unlock_bh(&ifp->lock);
		}
		ipv6_del_addr(ifp);
1943 1944 1945 1946 1947 1948 1949 1950
	} else if (ifp->flags&IFA_F_PERMANENT || !dad_failed) {
		spin_lock_bh(&ifp->lock);
		addrconf_del_dad_work(ifp);
		ifp->flags |= IFA_F_TENTATIVE;
		spin_unlock_bh(&ifp->lock);
		if (dad_failed)
			ipv6_ifa_notify(0, ifp);
		in6_ifa_put(ifp);
1951
	} else {
L
Linus Torvalds 已提交
1952
		ipv6_del_addr(ifp);
1953
	}
L
Linus Torvalds 已提交
1954 1955
}

H
Herbert Xu 已提交
1956 1957 1958 1959
static int addrconf_dad_end(struct inet6_ifaddr *ifp)
{
	int err = -ENOENT;

1960
	spin_lock_bh(&ifp->lock);
H
Herbert Xu 已提交
1961 1962 1963 1964
	if (ifp->state == INET6_IFADDR_STATE_DAD) {
		ifp->state = INET6_IFADDR_STATE_POSTDAD;
		err = 0;
	}
1965
	spin_unlock_bh(&ifp->lock);
H
Herbert Xu 已提交
1966 1967 1968 1969

	return err;
}

1970 1971
void addrconf_dad_failure(struct inet6_ifaddr *ifp)
{
1972
	struct inet6_dev *idev = ifp->idev;
1973
	struct net *net = dev_net(ifp->idev->dev);
1974

1975 1976
	if (addrconf_dad_end(ifp)) {
		in6_ifa_put(ifp);
H
Herbert Xu 已提交
1977
		return;
1978
	}
H
Herbert Xu 已提交
1979

1980 1981
	net_info_ratelimited("%s: IPv6 duplicate address %pI6c detected!\n",
			     ifp->idev->dev->name, &ifp->addr);
1982

1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
	spin_lock_bh(&ifp->lock);

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

1994
		if (retries > net->ipv6.sysctl.idgen_retries) {
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
			net_info_ratelimited("%s: privacy stable address generation failed because of DAD conflicts!\n",
					     ifp->idev->dev->name);
			goto errdad;
		}

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

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

		spin_unlock_bh(&ifp->lock);
2009

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028
		if (idev->cnf.max_addresses &&
		    ipv6_count_addresses(idev) >=
		    idev->cnf.max_addresses)
			goto lock_errdad;

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

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

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

2029
		addrconf_mod_dad_work(ifp2, net->ipv6.sysctl.idgen_delay);
2030 2031 2032
		in6_ifa_put(ifp2);
lock_errdad:
		spin_lock_bh(&ifp->lock);
2033 2034
	}

2035
errdad:
2036 2037
	/* transition from _POSTDAD to _ERRDAD */
	ifp->state = INET6_IFADDR_STATE_ERRDAD;
2038
	spin_unlock_bh(&ifp->lock);
2039 2040

	addrconf_mod_dad_work(ifp, 0);
2041
	in6_ifa_put(ifp);
2042
}
L
Linus Torvalds 已提交
2043

2044 2045
/* Join to solicited addr multicast group.
 * caller must hold RTNL */
2046
void addrconf_join_solict(struct net_device *dev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2047 2048 2049 2050 2051 2052 2053 2054 2055 2056
{
	struct in6_addr maddr;

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

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

2057
/* caller must hold RTNL */
2058
void addrconf_leave_solict(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2059 2060 2061 2062 2063 2064 2065 2066 2067 2068
{
	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);
}

2069
/* caller must hold RTNL */
2070
static void addrconf_join_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
2071 2072
{
	struct in6_addr addr;
2073

2074
	if (ifp->prefix_len >= 127) /* RFC 6164 */
2075
		return;
L
Linus Torvalds 已提交
2076 2077 2078
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
W
WANG Cong 已提交
2079
	__ipv6_dev_ac_inc(ifp->idev, &addr);
L
Linus Torvalds 已提交
2080 2081
}

2082
/* caller must hold RTNL */
2083
static void addrconf_leave_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
2084 2085
{
	struct in6_addr addr;
2086

2087
	if (ifp->prefix_len >= 127) /* RFC 6164 */
2088
		return;
L
Linus Torvalds 已提交
2089 2090 2091 2092 2093 2094
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
	__ipv6_dev_ac_dec(ifp->idev, &addr);
}

2095
static int addrconf_ifid_6lowpan(u8 *eui, struct net_device *dev)
2096
{
2097 2098
	switch (dev->addr_len) {
	case ETH_ALEN:
2099 2100 2101 2102 2103
		memcpy(eui, dev->dev_addr, 3);
		eui[3] = 0xFF;
		eui[4] = 0xFE;
		memcpy(eui + 5, dev->dev_addr + 3, 3);
		break;
2104 2105 2106 2107 2108
	case EUI64_ADDR_LEN:
		memcpy(eui, dev->dev_addr, EUI64_ADDR_LEN);
		eui[0] ^= 2;
		break;
	default:
2109
		return -1;
2110 2111
	}

2112 2113 2114
	return 0;
}

2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128
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;
}

2129 2130 2131 2132 2133 2134
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);
2135
	eui[7] = *(u8 *)dev->dev_addr;
2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147
	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;
}

2148
static int __ipv6_isatap_ifid(u8 *eui, __be32 addr)
2149
{
2150 2151
	if (addr == 0)
		return -1;
2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171
	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;
}

2172 2173 2174 2175 2176
static int addrconf_ifid_gre(u8 *eui, struct net_device *dev)
{
	return __ipv6_isatap_ifid(eui, *(__be32 *)dev->dev_addr);
}

2177 2178 2179 2180 2181 2182 2183 2184 2185 2186
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
2187 2188 2189 2190 2191
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev)
{
	switch (dev->type) {
	case ARPHRD_ETHER:
	case ARPHRD_FDDI:
2192
		return addrconf_ifid_eui48(eui, dev);
L
Linus Torvalds 已提交
2193
	case ARPHRD_ARCNET:
2194
		return addrconf_ifid_arcnet(eui, dev);
L
Linus Torvalds 已提交
2195
	case ARPHRD_INFINIBAND:
2196
		return addrconf_ifid_infiniband(eui, dev);
F
Fred L. Templin 已提交
2197
	case ARPHRD_SIT:
2198
		return addrconf_ifid_sit(eui, dev);
2199
	case ARPHRD_IPGRE:
2200
	case ARPHRD_TUNNEL:
2201
		return addrconf_ifid_gre(eui, dev);
2202
	case ARPHRD_6LOWPAN:
2203
		return addrconf_ifid_6lowpan(eui, dev);
2204 2205
	case ARPHRD_IEEE1394:
		return addrconf_ifid_ieee1394(eui, dev);
2206
	case ARPHRD_TUNNEL6:
2207
	case ARPHRD_IP6GRE:
2208
		return addrconf_ifid_ip6tnl(eui, dev);
L
Linus Torvalds 已提交
2209 2210 2211 2212 2213 2214 2215 2216 2217 2218
	}
	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);
2219 2220 2221
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
L
Linus Torvalds 已提交
2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232
		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) */
2233
static void ipv6_regen_rndid(struct inet6_dev *idev)
L
Linus Torvalds 已提交
2234 2235
{
regen:
2236
	get_random_bytes(idev->rndid, sizeof(idev->rndid));
L
Linus Torvalds 已提交
2237 2238 2239 2240 2241 2242 2243 2244
	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
2245
	 *  - ISATAP (RFC4214) 6.1
L
Linus Torvalds 已提交
2246 2247 2248 2249
	 *	00-00-5E-FE-xx-xx-xx-xx
	 *  - value 0
	 *  - XXX: already assigned to an address on the device
	 */
2250
	if (idev->rndid[0] == 0xfd &&
L
Linus Torvalds 已提交
2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261
	    (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;
	}
}

2262
static void  ipv6_try_regen_rndid(struct inet6_dev *idev, struct in6_addr *tmpaddr)
2263
{
L
Linus Torvalds 已提交
2264
	if (tmpaddr && memcmp(idev->rndid, &tmpaddr->s6_addr[8], 8) == 0)
2265
		ipv6_regen_rndid(idev);
L
Linus Torvalds 已提交
2266 2267 2268 2269 2270 2271 2272 2273
}

/*
 *	Add prefix route.
 */

static void
addrconf_prefix_route(struct in6_addr *pfx, int plen, struct net_device *dev,
J
Jamal Hadi Salim 已提交
2274
		      unsigned long expires, u32 flags)
L
Linus Torvalds 已提交
2275
{
2276
	struct fib6_config cfg = {
D
David Ahern 已提交
2277
		.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_PREFIX,
2278 2279 2280 2281 2282
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_expires = expires,
		.fc_dst_len = plen,
		.fc_flags = RTF_UP | flags,
2283
		.fc_nlinfo.nl_net = dev_net(dev),
2284
		.fc_protocol = RTPROT_KERNEL,
2285
	};
L
Linus Torvalds 已提交
2286

A
Alexey Dobriyan 已提交
2287
	cfg.fc_dst = *pfx;
L
Linus Torvalds 已提交
2288 2289 2290 2291 2292

	/* 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
2293
#if IS_ENABLED(CONFIG_IPV6_SIT)
2294 2295
	if (dev->type == ARPHRD_SIT && (dev->flags & IFF_POINTOPOINT))
		cfg.fc_flags |= RTF_NONEXTHOP;
2296
#endif
L
Linus Torvalds 已提交
2297

2298
	ip6_route_add(&cfg, NULL);
L
Linus Torvalds 已提交
2299 2300
}

2301 2302 2303 2304 2305 2306 2307 2308 2309

static struct rt6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags)
{
	struct fib6_node *fn;
	struct rt6_info *rt = NULL;
	struct fib6_table *table;
D
David Ahern 已提交
2310
	u32 tb_id = l3mdev_fib_table(dev) ? : RT6_TABLE_PREFIX;
2311

D
David Ahern 已提交
2312
	table = fib6_get_table(dev_net(dev), tb_id);
2313
	if (!table)
2314 2315
		return NULL;

2316
	read_lock_bh(&table->tb6_lock);
2317 2318 2319
	fn = fib6_locate(&table->tb6_root, pfx, plen, NULL, 0);
	if (!fn)
		goto out;
2320 2321

	noflags |= RTF_CACHE;
2322
	for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) {
2323
		if (rt->dst.dev->ifindex != dev->ifindex)
2324 2325 2326
			continue;
		if ((rt->rt6i_flags & flags) != flags)
			continue;
2327
		if ((rt->rt6i_flags & noflags) != 0)
2328 2329 2330 2331 2332
			continue;
		dst_hold(&rt->dst);
		break;
	}
out:
2333
	read_unlock_bh(&table->tb6_lock);
2334 2335 2336 2337
	return rt;
}


L
Linus Torvalds 已提交
2338 2339 2340 2341
/* Create "default" multicast route to the interface */

static void addrconf_add_mroute(struct net_device *dev)
{
2342
	struct fib6_config cfg = {
D
David Ahern 已提交
2343
		.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_LOCAL,
2344 2345 2346 2347
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_dst_len = 8,
		.fc_flags = RTF_UP,
2348
		.fc_nlinfo.nl_net = dev_net(dev),
2349 2350 2351 2352
	};

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

2353
	ip6_route_add(&cfg, NULL);
L
Linus Torvalds 已提交
2354 2355 2356 2357 2358 2359 2360 2361
}

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

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
2362 2363
	idev = ipv6_find_idev(dev);
	if (!idev)
2364 2365 2366 2367
		return ERR_PTR(-ENOBUFS);

	if (idev->cnf.disable_ipv6)
		return ERR_PTR(-EACCES);
L
Linus Torvalds 已提交
2368 2369

	/* Add default multicast route */
2370
	if (!(dev->flags & IFF_LOOPBACK) && !netif_is_l3_master(dev))
2371
		addrconf_add_mroute(dev);
L
Linus Torvalds 已提交
2372 2373 2374 2375

	return idev;
}

2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405
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
2406
			       idev->desync_factor - age;
2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442
		if (max_prefered < 0)
			max_prefered = 0;

		if (valid_lft > max_valid)
			valid_lft = max_valid;

		if (prefered_lft > max_prefered)
			prefered_lft = max_prefered;

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

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

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

2443 2444
static bool is_addr_mode_generate_stable(struct inet6_dev *idev)
{
2445 2446
	return idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_STABLE_PRIVACY ||
	       idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_RANDOM;
2447 2448
}

A
Alexander Aring 已提交
2449 2450 2451 2452 2453 2454
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)
2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543
{
	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;

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
		if (in6_dev->cnf.optimistic_dad &&
		    !net->ipv6.devconf_all->forwarding && sllao)
			addr_flags |= IFA_F_OPTIMISTIC;
#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)
			ifp = ipv6_add_addr(in6_dev, addr, NULL,
					    pinfo->prefix_len,
					    addr_type&IPV6_ADDR_SCOPE_MASK,
					    addr_flags, valid_lft,
					    prefered_lft);

		if (IS_ERR_OR_NULL(ifp))
			return -1;

		update_lft = 0;
		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;
		if (!update_lft && !create && stored_lft) {
			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
2544
EXPORT_SYMBOL_GPL(addrconf_prefix_rcv_add_addr);
2545

2546
void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
L
Linus Torvalds 已提交
2547 2548 2549 2550
{
	struct prefix_info *pinfo;
	__u32 valid_lft;
	__u32 prefered_lft;
2551
	int addr_type, err;
2552
	u32 addr_flags = 0;
L
Linus Torvalds 已提交
2553
	struct inet6_dev *in6_dev;
2554
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
2555 2556

	pinfo = (struct prefix_info *) opt;
2557

L
Linus Torvalds 已提交
2558
	if (len < sizeof(struct prefix_info)) {
2559
		ADBG("addrconf: prefix option too short\n");
L
Linus Torvalds 已提交
2560 2561
		return;
	}
2562

L
Linus Torvalds 已提交
2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575
	/*
	 *	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) {
2576
		net_warn_ratelimited("addrconf: prefix option has invalid lifetime\n");
L
Linus Torvalds 已提交
2577 2578 2579 2580 2581
		return;
	}

	in6_dev = in6_dev_get(dev);

2582
	if (!in6_dev) {
2583 2584
		net_dbg_ratelimited("addrconf: device %s not configured\n",
				    dev->name);
L
Linus Torvalds 已提交
2585 2586 2587 2588 2589 2590 2591 2592 2593
		return;
	}

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

2594 2595 2596 2597
	if (pinfo->onlink) {
		struct rt6_info *rt;
		unsigned long rt_expires;

2598 2599 2600 2601 2602
		/* 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.
		 */
2603 2604 2605 2606
		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
2607

2608 2609
		if (addrconf_finite_timeout(rt_expires))
			rt_expires *= HZ;
L
Linus Torvalds 已提交
2610

2611 2612 2613 2614 2615
		rt = addrconf_get_prefix_route(&pinfo->prefix,
					       pinfo->prefix_len,
					       dev,
					       RTF_ADDRCONF | RTF_PREFIX_RT,
					       RTF_GATEWAY | RTF_DEFAULT);
L
Linus Torvalds 已提交
2616

2617
		if (rt) {
2618 2619 2620 2621
			/* Autoconf prefix route */
			if (valid_lft == 0) {
				ip6_del_rt(rt);
				rt = NULL;
2622
			} else if (addrconf_finite_timeout(rt_expires)) {
2623
				/* not infinity */
2624
				rt6_set_expires(rt, jiffies + rt_expires);
2625
			} else {
2626
				rt6_clean_expires(rt);
L
Linus Torvalds 已提交
2627 2628
			}
		} else if (valid_lft) {
2629
			clock_t expires = 0;
2630 2631
			int flags = RTF_ADDRCONF | RTF_PREFIX_RT;
			if (addrconf_finite_timeout(rt_expires)) {
2632 2633 2634 2635
				/* not infinity */
				flags |= RTF_EXPIRES;
				expires = jiffies_to_clock_t(rt_expires);
			}
L
Linus Torvalds 已提交
2636
			addrconf_prefix_route(&pinfo->prefix, pinfo->prefix_len,
2637
					      dev, expires, flags);
L
Linus Torvalds 已提交
2638
		}
A
Amerigo Wang 已提交
2639
		ip6_rt_put(rt);
L
Linus Torvalds 已提交
2640 2641 2642 2643 2644 2645
	}

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

	if (pinfo->autoconf && in6_dev->cnf.autoconf) {
		struct in6_addr addr;
2646
		bool tokenized = false, dev_addr_generated = false;
L
Linus Torvalds 已提交
2647 2648 2649

		if (pinfo->prefix_len == 64) {
			memcpy(&addr, &pinfo->prefix, 8);
2650 2651 2652 2653 2654 2655

			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);
2656
				tokenized = true;
2657
			} else if (is_addr_mode_generate_stable(in6_dev) &&
2658 2659
				   !ipv6_generate_stable_address(&addr, 0,
								 in6_dev)) {
2660
				addr_flags |= IFA_F_STABLE_PRIVACY;
2661
				goto ok;
2662 2663
			} else if (ipv6_generate_eui64(addr.s6_addr + 8, dev) &&
				   ipv6_inherit_eui64(addr.s6_addr + 8, in6_dev)) {
2664
				goto put;
2665 2666
			} else {
				dev_addr_generated = true;
L
Linus Torvalds 已提交
2667 2668 2669
			}
			goto ok;
		}
2670 2671
		net_dbg_ratelimited("IPv6 addrconf: prefix with wrong length %d\n",
				    pinfo->prefix_len);
2672
		goto put;
L
Linus Torvalds 已提交
2673 2674

ok:
2675 2676 2677 2678 2679 2680 2681
		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;
2682 2683 2684 2685 2686 2687 2688 2689 2690

		/* 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
2691 2692
	}
	inet6_prefix_notify(RTM_NEWPREFIX, in6_dev, pinfo);
2693
put:
L
Linus Torvalds 已提交
2694 2695 2696 2697 2698 2699 2700 2701
	in6_dev_put(in6_dev);
}

/*
 *	Set destination address.
 *	Special case for SIT interfaces where we create a new "virtual"
 *	device.
 */
2702
int addrconf_set_dstaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713
{
	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;

2714
	dev = __dev_get_by_index(net, ireq.ifr6_ifindex);
L
Linus Torvalds 已提交
2715 2716

	err = -ENODEV;
2717
	if (!dev)
L
Linus Torvalds 已提交
2718 2719
		goto err_exit;

A
Amerigo Wang 已提交
2720
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2721
	if (dev->type == ARPHRD_SIT) {
2722
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736
		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;
2737
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;
L
Linus Torvalds 已提交
2738

2739 2740 2741 2742 2743 2744 2745 2746
		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
2747 2748 2749

		if (err == 0) {
			err = -ENOBUFS;
2750 2751
			dev = __dev_get_by_name(net, p.name);
			if (!dev)
L
Linus Torvalds 已提交
2752 2753 2754 2755
				goto err_exit;
			err = dev_open(dev);
		}
	}
2756
#endif
L
Linus Torvalds 已提交
2757 2758 2759 2760 2761 2762

err_exit:
	rtnl_unlock();
	return err;
}

2763 2764 2765 2766 2767 2768 2769 2770 2771
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)
2772
		ret = ipv6_sock_mc_join(sk, ifindex, addr);
2773
	else
2774
		ret = ipv6_sock_mc_drop(sk, ifindex, addr);
2775 2776 2777 2778 2779
	release_sock(sk);

	return ret;
}

L
Linus Torvalds 已提交
2780 2781 2782
/*
 *	Manual configuration of address on an interface
 */
2783 2784
static int inet6_addr_add(struct net *net, int ifindex,
			  const struct in6_addr *pfx,
2785
			  const struct in6_addr *peer_pfx,
2786 2787
			  unsigned int plen, __u32 ifa_flags,
			  __u32 prefered_lft, __u32 valid_lft)
L
Linus Torvalds 已提交
2788 2789 2790 2791
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2792 2793
	unsigned long timeout;
	clock_t expires;
L
Linus Torvalds 已提交
2794
	int scope;
2795
	u32 flags;
L
Linus Torvalds 已提交
2796 2797

	ASSERT_RTNL();
2798

2799 2800 2801
	if (plen > 128)
		return -EINVAL;

2802 2803 2804 2805
	/* check the lifetime */
	if (!valid_lft || prefered_lft > valid_lft)
		return -EINVAL;

2806 2807 2808
	if (ifa_flags & IFA_F_MANAGETEMPADDR && plen != 64)
		return -EINVAL;

2809 2810
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2811
		return -ENODEV;
2812

2813 2814 2815
	idev = addrconf_add_dev(dev);
	if (IS_ERR(idev))
		return PTR_ERR(idev);
L
Linus Torvalds 已提交
2816

2817 2818 2819 2820 2821 2822 2823 2824
	if (ifa_flags & IFA_F_MCAUTOJOIN) {
		int ret = ipv6_mc_config(net->ipv6.mc_autojoin_sk,
					 true, pfx, ifindex);

		if (ret < 0)
			return ret;
	}

L
Linus Torvalds 已提交
2825 2826
	scope = ipv6_addr_scope(pfx);

2827 2828 2829 2830
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
2831
		flags = RTF_EXPIRES;
2832 2833 2834 2835
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
2836
	}
2837

2838 2839 2840 2841 2842 2843
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
2844

2845 2846
	ifp = ipv6_add_addr(idev, pfx, peer_pfx, plen, scope, ifa_flags,
			    valid_lft, prefered_lft);
2847

L
Linus Torvalds 已提交
2848
	if (!IS_ERR(ifp)) {
2849 2850 2851 2852 2853
		if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
			addrconf_prefix_route(&ifp->addr, ifp->prefix_len, dev,
					      expires, flags);
		}

2854 2855 2856 2857 2858
		/*
		 * Note that section 3.1 of RFC 4429 indicates
		 * that the Optimistic flag should not be set for
		 * manually configured addresses
		 */
2859
		addrconf_dad_start(ifp);
2860 2861 2862
		if (ifa_flags & IFA_F_MANAGETEMPADDR)
			manage_tempaddrs(idev, ifp, valid_lft, prefered_lft,
					 true, jiffies);
L
Linus Torvalds 已提交
2863
		in6_ifa_put(ifp);
2864
		addrconf_verify_rtnl();
L
Linus Torvalds 已提交
2865
		return 0;
2866 2867 2868
	} else if (ifa_flags & IFA_F_MCAUTOJOIN) {
		ipv6_mc_config(net->ipv6.mc_autojoin_sk,
			       false, pfx, ifindex);
L
Linus Torvalds 已提交
2869 2870 2871 2872 2873
	}

	return PTR_ERR(ifp);
}

2874 2875
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
2876 2877 2878 2879
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2880

2881 2882 2883
	if (plen > 128)
		return -EINVAL;

2884 2885
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2886 2887
		return -ENODEV;

2888
	idev = __in6_dev_get(dev);
2889
	if (!idev)
L
Linus Torvalds 已提交
2890 2891 2892
		return -ENXIO;

	read_lock_bh(&idev->lock);
2893
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
L
Linus Torvalds 已提交
2894 2895 2896 2897
		if (ifp->prefix_len == plen &&
		    ipv6_addr_equal(pfx, &ifp->addr)) {
			in6_ifa_hold(ifp);
			read_unlock_bh(&idev->lock);
2898

2899 2900 2901 2902
			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
2903
			ipv6_del_addr(ifp);
2904
			addrconf_verify_rtnl();
2905 2906 2907 2908
			if (ipv6_addr_is_multicast(pfx)) {
				ipv6_mc_config(net->ipv6.mc_autojoin_sk,
					       false, pfx, dev->ifindex);
			}
L
Linus Torvalds 已提交
2909 2910 2911 2912 2913 2914 2915 2916
			return 0;
		}
	}
	read_unlock_bh(&idev->lock);
	return -EADDRNOTAVAIL;
}


2917
int addrconf_add_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2918 2919 2920
{
	struct in6_ifreq ireq;
	int err;
2921

2922
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2923
		return -EPERM;
2924

L
Linus Torvalds 已提交
2925 2926 2927 2928
	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		return -EFAULT;

	rtnl_lock();
2929
	err = inet6_addr_add(net, ireq.ifr6_ifindex, &ireq.ifr6_addr, NULL,
2930 2931
			     ireq.ifr6_prefixlen, IFA_F_PERMANENT,
			     INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
2932 2933 2934 2935
	rtnl_unlock();
	return err;
}

2936
int addrconf_del_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2937 2938 2939
{
	struct in6_ifreq ireq;
	int err;
2940

2941
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2942 2943 2944 2945 2946 2947
		return -EPERM;

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

	rtnl_lock();
2948
	err = inet6_addr_del(net, ireq.ifr6_ifindex, 0, &ireq.ifr6_addr,
2949
			     ireq.ifr6_prefixlen);
L
Linus Torvalds 已提交
2950 2951 2952 2953
	rtnl_unlock();
	return err;
}

2954 2955 2956 2957 2958
static void add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
		     int plen, int scope)
{
	struct inet6_ifaddr *ifp;

2959
	ifp = ipv6_add_addr(idev, addr, NULL, plen,
2960 2961
			    scope, IFA_F_PERMANENT,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
2962 2963 2964 2965
	if (!IS_ERR(ifp)) {
		spin_lock_bh(&ifp->lock);
		ifp->flags &= ~IFA_F_TENTATIVE;
		spin_unlock_bh(&ifp->lock);
2966
		rt_genid_bump_ipv6(dev_net(idev->dev));
2967 2968 2969 2970 2971
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
		in6_ifa_put(ifp);
	}
}

A
Amerigo Wang 已提交
2972
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2973 2974 2975 2976
static void sit_add_v4_addrs(struct inet6_dev *idev)
{
	struct in6_addr addr;
	struct net_device *dev;
2977
	struct net *net = dev_net(idev->dev);
2978
	int scope, plen;
2979
	u32 pflags = 0;
L
Linus Torvalds 已提交
2980 2981 2982 2983 2984 2985 2986 2987 2988

	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;
2989
		plen = 64;
L
Linus Torvalds 已提交
2990 2991
	} else {
		scope = IPV6_ADDR_COMPATv4;
2992
		plen = 96;
2993
		pflags |= RTF_NONEXTHOP;
L
Linus Torvalds 已提交
2994 2995 2996
	}

	if (addr.s6_addr32[3]) {
2997
		add_addr(idev, &addr, plen, scope);
2998
		addrconf_prefix_route(&addr, plen, idev->dev, 0, pflags);
L
Linus Torvalds 已提交
2999 3000 3001
		return;
	}

3002
	for_each_netdev(net, dev) {
3003
		struct in_device *in_dev = __in_dev_get_rtnl(dev);
L
Linus Torvalds 已提交
3004
		if (in_dev && (dev->flags & IFF_UP)) {
3005
			struct in_ifaddr *ifa;
L
Linus Torvalds 已提交
3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020

			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;
				}

3021
				add_addr(idev, &addr, plen, flag);
3022 3023
				addrconf_prefix_route(&addr, plen, idev->dev, 0,
						      pflags);
L
Linus Torvalds 已提交
3024 3025
			}
		}
3026
	}
L
Linus Torvalds 已提交
3027
}
3028
#endif
L
Linus Torvalds 已提交
3029 3030 3031 3032 3033 3034 3035 3036 3037

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

	/* ::1 */

	ASSERT_RTNL();

3038
	idev = ipv6_find_idev(dev);
3039
	if (!idev) {
3040
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
3041 3042 3043
		return;
	}

3044
	add_addr(idev, &in6addr_loopback, 128, IFA_HOST);
L
Linus Torvalds 已提交
3045 3046
}

3047 3048
void addrconf_add_linklocal(struct inet6_dev *idev,
			    const struct in6_addr *addr, u32 flags)
L
Linus Torvalds 已提交
3049
{
3050
	struct inet6_ifaddr *ifp;
3051
	u32 addr_flags = flags | IFA_F_PERMANENT;
3052 3053 3054

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	if (idev->cnf.optimistic_dad &&
3055
	    !dev_net(idev->dev)->ipv6.devconf_all->forwarding)
3056 3057
		addr_flags |= IFA_F_OPTIMISTIC;
#endif
L
Linus Torvalds 已提交
3058

3059 3060
	ifp = ipv6_add_addr(idev, addr, NULL, 64, IFA_LINK, addr_flags,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
3061
	if (!IS_ERR(ifp)) {
3062
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, idev->dev, 0, 0);
3063
		addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
3064 3065 3066
		in6_ifa_put(ifp);
	}
}
3067
EXPORT_SYMBOL_GPL(addrconf_add_linklocal);
L
Linus Torvalds 已提交
3068

3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096
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;
3097
			__be32 prefix[2];
3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122
			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);
3123 3124
	data.prefix[0] = address->s6_addr32[0];
	data.prefix[1] = address->s6_addr32[1];
3125 3126 3127 3128 3129 3130
	data.secret = secret;
	data.dad_count = dad_count;

	sha_transform(digest, data.__data, workspace);

	temp = *address;
3131 3132
	temp.s6_addr32[2] = (__force __be32)digest[0];
	temp.s6_addr32[3] = (__force __be32)digest[1];
3133 3134 3135 3136 3137

	spin_unlock_bh(&lock);

	if (ipv6_reserved_interfaceid(temp)) {
		dad_count++;
3138
		if (dad_count > dev_net(idev->dev)->ipv6.sysctl.idgen_retries)
3139 3140 3141 3142 3143 3144 3145 3146
			return -1;
		goto retry;
	}

	*address = temp;
	return 0;
}

3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157
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;
}

3158 3159
static void addrconf_addr_gen(struct inet6_dev *idev, bool prefix_route)
{
3160 3161
	struct in6_addr addr;

D
David Ahern 已提交
3162 3163 3164 3165
	/* no link local addresses on L3 master devices */
	if (netif_is_l3_master(idev->dev))
		return;

3166
	ipv6_addr_set(&addr, htonl(0xFE800000), 0, 0, 0);
3167

3168
	switch (idev->cnf.addr_gen_mode) {
3169 3170 3171 3172
	case IN6_ADDR_GEN_MODE_RANDOM:
		ipv6_gen_mode_random_init(idev);
		/* fallthrough */
	case IN6_ADDR_GEN_MODE_STABLE_PRIVACY:
3173
		if (!ipv6_generate_stable_address(&addr, 0, idev))
3174 3175
			addrconf_add_linklocal(idev, &addr,
					       IFA_F_STABLE_PRIVACY);
3176 3177
		else if (prefix_route)
			addrconf_prefix_route(&addr, 64, idev->dev, 0, 0);
3178 3179
		break;
	case IN6_ADDR_GEN_MODE_EUI64:
3180 3181 3182 3183 3184
		/* 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)
3185
			addrconf_add_linklocal(idev, &addr, 0);
3186 3187
		else if (prefix_route)
			addrconf_prefix_route(&addr, 64, idev->dev, 0, 0);
3188 3189 3190 3191 3192
		break;
	case IN6_ADDR_GEN_MODE_NONE:
	default:
		/* will not add any link local address */
		break;
3193 3194 3195
	}
}

L
Linus Torvalds 已提交
3196 3197
static void addrconf_dev_config(struct net_device *dev)
{
3198
	struct inet6_dev *idev;
L
Linus Torvalds 已提交
3199 3200 3201

	ASSERT_RTNL();

3202 3203 3204
	if ((dev->type != ARPHRD_ETHER) &&
	    (dev->type != ARPHRD_FDDI) &&
	    (dev->type != ARPHRD_ARCNET) &&
3205
	    (dev->type != ARPHRD_INFINIBAND) &&
3206
	    (dev->type != ARPHRD_IEEE1394) &&
3207
	    (dev->type != ARPHRD_TUNNEL6) &&
3208
	    (dev->type != ARPHRD_6LOWPAN) &&
3209 3210 3211
	    (dev->type != ARPHRD_IP6GRE) &&
	    (dev->type != ARPHRD_IPGRE) &&
	    (dev->type != ARPHRD_TUNNEL) &&
3212
	    (dev->type != ARPHRD_NONE)) {
3213 3214 3215 3216
		/* Alas, we support only Ethernet autoconfiguration. */
		return;
	}

L
Linus Torvalds 已提交
3217
	idev = addrconf_add_dev(dev);
3218
	if (IS_ERR(idev))
L
Linus Torvalds 已提交
3219 3220
		return;

3221 3222
	/* this device type has no EUI support */
	if (dev->type == ARPHRD_NONE &&
3223 3224
	    idev->cnf.addr_gen_mode == IN6_ADDR_GEN_MODE_EUI64)
		idev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_RANDOM;
3225

3226
	addrconf_addr_gen(idev, false);
L
Linus Torvalds 已提交
3227 3228
}

A
Amerigo Wang 已提交
3229
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3230 3231 3232 3233 3234 3235
static void addrconf_sit_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3236 3237 3238
	/*
	 * Configure the tunnel with one of our IPv4
	 * addresses... we should configure all of
L
Linus Torvalds 已提交
3239 3240 3241
	 * our v4 addrs in the tunnel
	 */

3242
	idev = ipv6_find_idev(dev);
3243
	if (!idev) {
3244
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
3245 3246 3247
		return;
	}

F
Fred L. Templin 已提交
3248
	if (dev->priv_flags & IFF_ISATAP) {
3249
		addrconf_addr_gen(idev, false);
F
Fred L. Templin 已提交
3250 3251 3252
		return;
	}

L
Linus Torvalds 已提交
3253 3254
	sit_add_v4_addrs(idev);

3255
	if (dev->flags&IFF_POINTOPOINT)
L
Linus Torvalds 已提交
3256 3257
		addrconf_add_mroute(dev);
}
3258
#endif
L
Linus Torvalds 已提交
3259

A
Amerigo Wang 已提交
3260
#if IS_ENABLED(CONFIG_NET_IPGRE)
3261 3262 3263 3264 3265 3266
static void addrconf_gre_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

3267
	idev = ipv6_find_idev(dev);
3268
	if (!idev) {
3269
		pr_debug("%s: add_dev failed\n", __func__);
3270 3271 3272
		return;
	}

3273
	addrconf_addr_gen(idev, true);
3274 3275
	if (dev->flags & IFF_POINTOPOINT)
		addrconf_add_mroute(dev);
3276 3277 3278
}
#endif

3279 3280 3281
static int fixup_permanent_addr(struct inet6_dev *idev,
				struct inet6_ifaddr *ifp)
{
3282
	/* !rt6i_node means the host route was removed from the
3283 3284 3285
	 * FIB, for example, if 'lo' device is taken down. In that
	 * case regenerate the host route.
	 */
3286
	if (!ifp->rt || !ifp->rt->rt6i_node) {
3287
		struct rt6_info *rt, *prev;
3288 3289 3290 3291 3292

		rt = addrconf_dst_alloc(idev, &ifp->addr, false);
		if (unlikely(IS_ERR(rt)))
			return PTR_ERR(rt);

3293 3294 3295
		/* ifp->rt can be accessed outside of rtnl */
		spin_lock(&ifp->lock);
		prev = ifp->rt;
3296
		ifp->rt = rt;
3297 3298 3299
		spin_unlock(&ifp->lock);

		ip6_rt_put(prev);
3300 3301 3302 3303 3304 3305 3306
	}

	if (!(ifp->flags & IFA_F_NOPREFIXROUTE)) {
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len,
				      idev->dev, 0, 0);
	}

3307 3308
	if (ifp->state == INET6_IFADDR_STATE_PREDAD)
		addrconf_dad_start(ifp);
3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338

	return 0;
}

static void addrconf_permanent_addr(struct net_device *dev)
{
	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) &&
		    fixup_permanent_addr(idev, ifp) < 0) {
			write_unlock_bh(&idev->lock);
			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);
}

3339
static int addrconf_notify(struct notifier_block *this, unsigned long event,
3340
			   void *ptr)
L
Linus Torvalds 已提交
3341
{
3342
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
3343
	struct netdev_notifier_changeupper_info *info;
3344
	struct inet6_dev *idev = __in6_dev_get(dev);
3345
	struct net *net = dev_net(dev);
3346
	int run_pending = 0;
3347
	int err;
L
Linus Torvalds 已提交
3348

S
Stephen Hemminger 已提交
3349
	switch (event) {
3350
	case NETDEV_REGISTER:
3351
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
3352
			idev = ipv6_add_dev(dev);
3353 3354
			if (IS_ERR(idev))
				return notifier_from_errno(PTR_ERR(idev));
3355 3356
		}
		break;
S
Stephen Hemminger 已提交
3357

3358 3359 3360
	case NETDEV_CHANGEMTU:
		/* if MTU under IPV6_MIN_MTU stop IPv6 on this interface. */
		if (dev->mtu < IPV6_MIN_MTU) {
3361
			addrconf_ifdown(dev, dev != net->loopback_dev);
3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383
			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
3384
	case NETDEV_UP:
3385
	case NETDEV_CHANGE:
3386 3387 3388
		if (dev->flags & IFF_SLAVE)
			break;

3389 3390 3391
		if (idev && idev->cnf.disable_ipv6)
			break;

3392
		if (event == NETDEV_UP) {
3393 3394 3395
			/* restore routes for permanent addresses */
			addrconf_permanent_addr(dev);

3396
			if (!addrconf_qdisc_ok(dev)) {
3397
				/* device is not ready yet. */
3398
				pr_info("ADDRCONF(NETDEV_UP): %s: link is not ready\n",
3399 3400 3401
					dev->name);
				break;
			}
3402

3403 3404 3405
			if (!idev && dev->mtu >= IPV6_MIN_MTU)
				idev = ipv6_add_dev(dev);

3406
			if (!IS_ERR_OR_NULL(idev)) {
3407
				idev->if_flags |= IF_READY;
3408 3409
				run_pending = 1;
			}
3410
		} else if (event == NETDEV_CHANGE) {
3411
			if (!addrconf_qdisc_ok(dev)) {
3412 3413 3414 3415 3416
				/* device is still not ready. */
				break;
			}

			if (idev) {
3417 3418 3419 3420 3421 3422 3423
				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);
3424
					break;
3425
				}
3426 3427 3428
				idev->if_flags |= IF_READY;
			}

3429 3430
			pr_info("ADDRCONF(NETDEV_CHANGE): %s: link becomes ready\n",
				dev->name);
3431

3432
			run_pending = 1;
3433 3434
		}

S
Stephen Hemminger 已提交
3435
		switch (dev->type) {
A
Amerigo Wang 已提交
3436
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
3437 3438 3439
		case ARPHRD_SIT:
			addrconf_sit_config(dev);
			break;
3440
#endif
A
Amerigo Wang 已提交
3441
#if IS_ENABLED(CONFIG_NET_IPGRE)
3442 3443 3444
		case ARPHRD_IPGRE:
			addrconf_gre_config(dev);
			break;
3445
#endif
L
Linus Torvalds 已提交
3446 3447 3448 3449 3450 3451 3452
		case ARPHRD_LOOPBACK:
			init_loopback(dev);
			break;

		default:
			addrconf_dev_config(dev);
			break;
3453
		}
S
Stephen Hemminger 已提交
3454

3455
		if (!IS_ERR_OR_NULL(idev)) {
3456 3457 3458
			if (run_pending)
				addrconf_dad_run(idev);

S
Stephen Hemminger 已提交
3459 3460 3461 3462
			/*
			 * 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
3463
			 */
S
Stephen Hemminger 已提交
3464 3465
			if (idev->cnf.mtu6 != dev->mtu &&
			    dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
3466 3467 3468 3469 3470
				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
3471 3472 3473 3474

			/*
			 * If the changed mtu during down is lower than
			 * IPV6_MIN_MTU stop IPv6 on this interface.
L
Linus Torvalds 已提交
3475 3476
			 */
			if (dev->mtu < IPV6_MIN_MTU)
3477
				addrconf_ifdown(dev, dev != net->loopback_dev);
L
Linus Torvalds 已提交
3478 3479 3480 3481 3482 3483 3484 3485 3486 3487
		}
		break;

	case NETDEV_DOWN:
	case NETDEV_UNREGISTER:
		/*
		 *	Remove all addresses from this interface.
		 */
		addrconf_ifdown(dev, event != NETDEV_DOWN);
		break;
3488

L
Linus Torvalds 已提交
3489 3490
	case NETDEV_CHANGENAME:
		if (idev) {
3491
			snmp6_unregister_dev(idev);
3492
			addrconf_sysctl_unregister(idev);
3493
			err = addrconf_sysctl_register(idev);
3494 3495
			if (err)
				return notifier_from_errno(err);
3496 3497 3498 3499 3500
			err = snmp6_register_dev(idev);
			if (err) {
				addrconf_sysctl_unregister(idev);
				return notifier_from_errno(err);
			}
3501
		}
L
Linus Torvalds 已提交
3502
		break;
S
Stephen Hemminger 已提交
3503

3504 3505
	case NETDEV_PRE_TYPE_CHANGE:
	case NETDEV_POST_TYPE_CHANGE:
3506 3507
		if (idev)
			addrconf_type_change(dev, event);
3508
		break;
3509 3510 3511 3512 3513 3514 3515 3516 3517

	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);
3518
	}
L
Linus Torvalds 已提交
3519 3520 3521 3522 3523 3524 3525 3526 3527

	return NOTIFY_OK;
}

/*
 *	addrconf module should be notified of a device going up
 */
static struct notifier_block ipv6_dev_notf = {
	.notifier_call = addrconf_notify,
3528
	.priority = ADDRCONF_NOTIFY_PRIORITY,
L
Linus Torvalds 已提交
3529 3530
};

3531
static void addrconf_type_change(struct net_device *dev, unsigned long event)
3532 3533 3534 3535 3536 3537
{
	struct inet6_dev *idev;
	ASSERT_RTNL();

	idev = __in6_dev_get(dev);

3538
	if (event == NETDEV_POST_TYPE_CHANGE)
3539
		ipv6_mc_remap(idev);
3540
	else if (event == NETDEV_PRE_TYPE_CHANGE)
3541 3542 3543
		ipv6_mc_unmap(idev);
}

3544 3545 3546 3547 3548 3549
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
3550 3551
static int addrconf_ifdown(struct net_device *dev, int how)
{
3552
	struct net *net = dev_net(dev);
3553
	struct inet6_dev *idev;
3554 3555 3556 3557
	struct inet6_ifaddr *ifa, *tmp;
	struct list_head del_list;
	int _keep_addr;
	bool keep_addr;
3558
	int state, i;
L
Linus Torvalds 已提交
3559 3560 3561

	ASSERT_RTNL();

3562 3563
	rt6_ifdown(net, dev);
	neigh_ifdown(&nd_tbl, dev);
L
Linus Torvalds 已提交
3564 3565

	idev = __in6_dev_get(dev);
3566
	if (!idev)
L
Linus Torvalds 已提交
3567 3568
		return -ENODEV;

S
Stephen Hemminger 已提交
3569 3570 3571
	/*
	 * Step 1: remove reference to ipv6 device from parent device.
	 *	   Do not dev_put!
L
Linus Torvalds 已提交
3572
	 */
3573
	if (how) {
L
Linus Torvalds 已提交
3574
		idev->dead = 1;
3575 3576

		/* protected by rtnl_lock */
3577
		RCU_INIT_POINTER(dev->ip6_ptr, NULL);
L
Linus Torvalds 已提交
3578 3579 3580 3581 3582 3583

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

	}

3584 3585 3586 3587 3588 3589 3590 3591
	/* aggregate the system setting and interface setting */
	_keep_addr = net->ipv6.devconf_all->keep_addr_on_down;
	if (!_keep_addr)
		_keep_addr = idev->cnf.keep_addr_on_down;

	/* combine the user config with event to determine if permanent
	 * addresses are to be removed from address hash table
	 */
3592
	keep_addr = !(how || _keep_addr <= 0 || idev->cnf.disable_ipv6);
3593

3594 3595 3596 3597 3598
	/* 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);
3599
restart:
3600
		hlist_for_each_entry_rcu(ifa, h, addr_lst) {
3601
			if (ifa->idev == idev) {
3602
				addrconf_del_dad_work(ifa);
3603 3604 3605 3606
				/* combined flag + permanent flag decide if
				 * address is retained on a down event
				 */
				if (!keep_addr ||
3607 3608
				    !(ifa->flags & IFA_F_PERMANENT) ||
				    addr_is_local(&ifa->addr)) {
3609 3610 3611
					hlist_del_init_rcu(&ifa->addr_lst);
					goto restart;
				}
3612 3613 3614 3615 3616
			}
		}
		spin_unlock_bh(&addrconf_hash_lock);
	}

L
Linus Torvalds 已提交
3617 3618
	write_lock_bh(&idev->lock);

3619 3620
	addrconf_del_rs_timer(idev);

S
Stephen Hemminger 已提交
3621
	/* Step 2: clear flags for stateless addrconf */
3622
	if (!how)
3623
		idev->if_flags &= ~(IF_RS_SENT|IF_RA_RCVD|IF_READY);
L
Linus Torvalds 已提交
3624

S
Stephen Hemminger 已提交
3625
	/* Step 3: clear tempaddr list */
3626 3627 3628 3629
	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
3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640
		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);
	}
3641

3642 3643 3644
	/* re-combine the user config with event to determine if permanent
	 * addresses are to be removed from the interface list
	 */
3645
	keep_addr = (!how && _keep_addr > 0 && !idev->cnf.disable_ipv6);
3646

3647 3648
	INIT_LIST_HEAD(&del_list);
	list_for_each_entry_safe(ifa, tmp, &idev->addr_list, if_list) {
3649
		struct rt6_info *rt = NULL;
3650
		bool keep;
3651

3652
		addrconf_del_dad_work(ifa);
3653

3654 3655 3656 3657 3658
		keep = keep_addr && (ifa->flags & IFA_F_PERMANENT) &&
			!addr_is_local(&ifa->addr);
		if (!keep)
			list_move(&ifa->if_list, &del_list);

3659
		write_unlock_bh(&idev->lock);
3660
		spin_lock_bh(&ifa->lock);
3661

3662
		if (keep) {
3663 3664
			/* set state to skip the notifier below */
			state = INET6_IFADDR_STATE_DEAD;
3665
			ifa->state = INET6_IFADDR_STATE_PREDAD;
3666 3667
			if (!(ifa->flags & IFA_F_NODAD))
				ifa->flags |= IFA_F_TENTATIVE;
3668 3669 3670

			rt = ifa->rt;
			ifa->rt = NULL;
3671 3672 3673 3674 3675
		} else {
			state = ifa->state;
			ifa->state = INET6_IFADDR_STATE_DEAD;
		}

3676
		spin_unlock_bh(&ifa->lock);
3677

3678 3679 3680
		if (rt)
			ip6_del_rt(rt);

3681 3682
		if (state != INET6_IFADDR_STATE_DEAD) {
			__ipv6_ifa_notify(RTM_DELADDR, ifa);
3683
			inet6addr_notifier_call_chain(NETDEV_DOWN, ifa);
3684 3685 3686 3687
		} else {
			if (idev->cnf.forwarding)
				addrconf_leave_anycast(ifa);
			addrconf_leave_solict(ifa->idev, &ifa->addr);
3688
		}
3689

3690 3691
		write_lock_bh(&idev->lock);
	}
3692

L
Linus Torvalds 已提交
3693 3694
	write_unlock_bh(&idev->lock);

3695 3696 3697 3698 3699 3700 3701 3702 3703
	/* now clean up addresses to be removed */
	while (!list_empty(&del_list)) {
		ifa = list_first_entry(&del_list,
				       struct inet6_ifaddr, if_list);
		list_del(&ifa->if_list);

		in6_ifa_put(ifa);
	}

3704 3705 3706
	/* Step 5: Discard anycast and multicast list */
	if (how) {
		ipv6_ac_destroy_dev(idev);
L
Linus Torvalds 已提交
3707
		ipv6_mc_destroy_dev(idev);
3708
	} else {
L
Linus Torvalds 已提交
3709
		ipv6_mc_down(idev);
3710
	}
L
Linus Torvalds 已提交
3711 3712

	idev->tstamp = jiffies;
3713

S
Stephen Hemminger 已提交
3714
	/* Last: Shot the device (if unregistered) */
3715
	if (how) {
3716
		addrconf_sysctl_unregister(idev);
L
Linus Torvalds 已提交
3717 3718 3719 3720 3721 3722 3723 3724 3725
		neigh_parms_release(&nd_tbl, idev->nd_parms);
		neigh_ifdown(&nd_tbl, dev);
		in6_dev_put(idev);
	}
	return 0;
}

static void addrconf_rs_timer(unsigned long data)
{
3726
	struct inet6_dev *idev = (struct inet6_dev *)data;
3727
	struct net_device *dev = idev->dev;
3728
	struct in6_addr lladdr;
L
Linus Torvalds 已提交
3729

3730
	write_lock(&idev->lock);
S
stephen hemminger 已提交
3731
	if (idev->dead || !(idev->if_flags & IF_READY))
L
Linus Torvalds 已提交
3732 3733
		goto out;

3734
	if (!ipv6_accept_ra(idev))
S
stephen hemminger 已提交
3735 3736 3737 3738
		goto out;

	/* Announcement received after solicitation was sent */
	if (idev->if_flags & IF_RA_RCVD)
L
Linus Torvalds 已提交
3739 3740
		goto out;

3741
	if (idev->rs_probes++ < idev->cnf.rtr_solicits || idev->cnf.rtr_solicits < 0) {
3742 3743 3744
		write_unlock(&idev->lock);
		if (!ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
			ndisc_send_rs(dev, &lladdr,
3745 3746
				      &in6addr_linklocal_allrouters);
		else
3747
			goto put;
L
Linus Torvalds 已提交
3748

3749
		write_lock(&idev->lock);
3750 3751
		idev->rs_interval = rfc3315_s14_backoff_update(
			idev->rs_interval, idev->cnf.rtr_solicit_max_interval);
3752 3753 3754 3755
		/* 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 :
3756
				      idev->rs_interval);
L
Linus Torvalds 已提交
3757 3758 3759 3760 3761
	} else {
		/*
		 * Note: we do not support deprecated "all on-link"
		 * assumption any longer.
		 */
3762
		pr_debug("%s: no IPv6 routers present\n", idev->dev->name);
L
Linus Torvalds 已提交
3763 3764 3765
	}

out:
3766
	write_unlock(&idev->lock);
3767
put:
3768
	in6_dev_put(idev);
L
Linus Torvalds 已提交
3769 3770 3771 3772 3773
}

/*
 *	Duplicate Address Detection
 */
3774 3775 3776 3777
static void addrconf_dad_kick(struct inet6_ifaddr *ifp)
{
	unsigned long rand_num;
	struct inet6_dev *idev = ifp->idev;
3778
	u64 nonce;
3779

3780 3781 3782
	if (ifp->flags & IFA_F_OPTIMISTIC)
		rand_num = 0;
	else
3783
		rand_num = prandom_u32() % (idev->cnf.rtr_solicit_delay ? : 1);
3784

3785 3786 3787 3788 3789 3790 3791 3792
	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;
3793
	ifp->dad_probes = idev->cnf.dad_transmits;
3794
	addrconf_mod_dad_work(ifp, rand_num);
3795 3796
}

3797
static void addrconf_dad_begin(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3798 3799 3800
{
	struct inet6_dev *idev = ifp->idev;
	struct net_device *dev = idev->dev;
3801
	bool bump_id, notify = false;
L
Linus Torvalds 已提交
3802 3803 3804

	addrconf_join_solict(dev, &ifp->addr);

3805
	prandom_seed((__force u32) ifp->addr.s6_addr32[3]);
L
Linus Torvalds 已提交
3806 3807

	read_lock_bh(&idev->lock);
3808
	spin_lock(&ifp->lock);
3809
	if (ifp->state == INET6_IFADDR_STATE_DEAD)
L
Linus Torvalds 已提交
3810 3811 3812
		goto out;

	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
3813
	    idev->cnf.accept_dad < 1 ||
3814 3815
	    !(ifp->flags&IFA_F_TENTATIVE) ||
	    ifp->flags & IFA_F_NODAD) {
3816
		bump_id = ifp->flags & IFA_F_TENTATIVE;
B
Brian Haley 已提交
3817
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3818
		spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3819 3820
		read_unlock_bh(&idev->lock);

3821
		addrconf_dad_completed(ifp, bump_id);
L
Linus Torvalds 已提交
3822 3823 3824
		return;
	}

3825
	if (!(idev->if_flags & IF_READY)) {
3826
		spin_unlock(&ifp->lock);
3827
		read_unlock_bh(&idev->lock);
3828
		/*
3829
		 * If the device is not ready:
3830 3831 3832 3833
		 * - keep it tentative if it is a permanent address.
		 * - otherwise, kill it.
		 */
		in6_ifa_hold(ifp);
B
Brian Haley 已提交
3834
		addrconf_dad_stop(ifp, 0);
3835
		return;
3836
	}
3837 3838 3839 3840 3841

	/*
	 * Optimistic nodes can start receiving
	 * Frames right away
	 */
3842
	if (ifp->flags & IFA_F_OPTIMISTIC) {
3843
		ip6_ins_rt(ifp->rt);
3844 3845 3846 3847
		if (ipv6_use_optimistic_addr(idev)) {
			/* Because optimistic nodes can use this address,
			 * notify listeners. If DAD fails, RTM_DELADDR is sent.
			 */
3848
			notify = true;
3849 3850
		}
	}
3851

3852
	addrconf_dad_kick(ifp);
L
Linus Torvalds 已提交
3853
out:
3854
	spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3855
	read_unlock_bh(&idev->lock);
3856 3857
	if (notify)
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
L
Linus Torvalds 已提交
3858 3859
}

3860
static void addrconf_dad_start(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3861
{
3862 3863
	bool begin_dad = false;

3864
	spin_lock_bh(&ifp->lock);
3865 3866 3867 3868
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
		ifp->state = INET6_IFADDR_STATE_PREDAD;
		begin_dad = true;
	}
3869
	spin_unlock_bh(&ifp->lock);
3870 3871 3872 3873 3874 3875 3876 3877 3878 3879

	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
3880
	struct inet6_dev *idev = ifp->idev;
3881
	bool bump_id, disable_ipv6 = false;
L
Linus Torvalds 已提交
3882 3883
	struct in6_addr mcaddr;

3884 3885 3886 3887 3888 3889 3890 3891
	enum {
		DAD_PROCESS,
		DAD_BEGIN,
		DAD_ABORT,
	} action = DAD_PROCESS;

	rtnl_lock();

3892
	spin_lock_bh(&ifp->lock);
3893 3894 3895 3896 3897 3898
	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;
3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916

		if (idev->cnf.accept_dad > 1 && !idev->cnf.disable_ipv6 &&
		    !(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;
			}
		}
3917
	}
3918
	spin_unlock_bh(&ifp->lock);
3919 3920 3921 3922 3923

	if (action == DAD_BEGIN) {
		addrconf_dad_begin(ifp);
		goto out;
	} else if (action == DAD_ABORT) {
3924
		in6_ifa_hold(ifp);
3925
		addrconf_dad_stop(ifp, 1);
3926 3927
		if (disable_ipv6)
			addrconf_ifdown(idev->dev, 0);
3928 3929 3930
		goto out;
	}

3931
	if (!ifp->dad_probes && addrconf_dad_end(ifp))
H
Herbert Xu 已提交
3932 3933
		goto out;

3934
	write_lock_bh(&idev->lock);
3935
	if (idev->dead || !(idev->if_flags & IF_READY)) {
3936
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3937 3938
		goto out;
	}
3939 3940

	spin_lock(&ifp->lock);
3941 3942
	if (ifp->state == INET6_IFADDR_STATE_DEAD) {
		spin_unlock(&ifp->lock);
3943
		write_unlock_bh(&idev->lock);
3944 3945 3946
		goto out;
	}

3947
	if (ifp->dad_probes == 0) {
L
Linus Torvalds 已提交
3948 3949 3950 3951
		/*
		 * DAD was successful
		 */

3952
		bump_id = ifp->flags & IFA_F_TENTATIVE;
B
Brian Haley 已提交
3953
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3954
		spin_unlock(&ifp->lock);
3955
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3956

3957
		addrconf_dad_completed(ifp, bump_id);
L
Linus Torvalds 已提交
3958 3959 3960 3961

		goto out;
	}

3962
	ifp->dad_probes--;
3963 3964
	addrconf_mod_dad_work(ifp,
			      NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME));
3965
	spin_unlock(&ifp->lock);
3966
	write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3967 3968 3969

	/* send a neighbour solicitation for our addr */
	addrconf_addr_solict_mult(&ifp->addr, &mcaddr);
3970 3971
	ndisc_send_ns(ifp->idev->dev, &ifp->addr, &mcaddr, &in6addr_any,
		      ifp->dad_nonce);
L
Linus Torvalds 已提交
3972 3973
out:
	in6_ifa_put(ifp);
3974
	rtnl_unlock();
L
Linus Torvalds 已提交
3975 3976
}

3977 3978 3979 3980 3981 3982
/* 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;

3983 3984 3985
	list_for_each_entry_reverse(ifpiter, &idev->addr_list, if_list) {
		if (ifpiter->scope > IFA_LINK)
			break;
3986 3987 3988 3989 3990 3991 3992 3993 3994
		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;
}

3995
static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id)
L
Linus Torvalds 已提交
3996
{
S
Stephen Hemminger 已提交
3997
	struct net_device *dev = ifp->idev->dev;
3998
	struct in6_addr lladdr;
3999
	bool send_rs, send_mld;
4000

4001
	addrconf_del_dad_work(ifp);
L
Linus Torvalds 已提交
4002 4003 4004 4005 4006 4007 4008

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

	ipv6_ifa_notify(RTM_NEWADDR, ifp);

4009 4010
	/* If added prefix is link local and we are prepared to process
	   router advertisements, start sending router solicitations.
L
Linus Torvalds 已提交
4011 4012
	 */

4013
	read_lock_bh(&ifp->idev->lock);
4014
	send_mld = ifp->scope == IFA_LINK && ipv6_lonely_lladdr(ifp);
4015 4016
	send_rs = send_mld &&
		  ipv6_accept_ra(ifp->idev) &&
4017
		  ifp->idev->cnf.rtr_solicits != 0 &&
4018
		  (dev->flags&IFF_LOOPBACK) == 0;
4019 4020
	read_unlock_bh(&ifp->idev->lock);

4021 4022 4023 4024 4025 4026
	/* 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);

4027
	if (send_rs) {
L
Linus Torvalds 已提交
4028 4029 4030 4031 4032
		/*
		 *	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
		 */
4033
		if (ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
4034
			return;
4035
		ndisc_send_rs(dev, &lladdr, &in6addr_linklocal_allrouters);
L
Linus Torvalds 已提交
4036

4037 4038
		write_lock_bh(&ifp->idev->lock);
		spin_lock(&ifp->lock);
4039 4040
		ifp->idev->rs_interval = rfc3315_s14_backoff_init(
			ifp->idev->cnf.rtr_solicit_interval);
4041
		ifp->idev->rs_probes = 1;
L
Linus Torvalds 已提交
4042
		ifp->idev->if_flags |= IF_RS_SENT;
4043
		addrconf_mod_rs_timer(ifp->idev, ifp->idev->rs_interval);
4044 4045
		spin_unlock(&ifp->lock);
		write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
4046
	}
4047 4048 4049

	if (bump_id)
		rt_genid_bump_ipv6(dev_net(dev));
4050 4051 4052 4053 4054 4055

	/* 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
4056 4057
}

S
Stephen Hemminger 已提交
4058 4059
static void addrconf_dad_run(struct inet6_dev *idev)
{
4060 4061 4062
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
4063
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
4064
		spin_lock(&ifp->lock);
H
Herbert Xu 已提交
4065 4066 4067
		if (ifp->flags & IFA_F_TENTATIVE &&
		    ifp->state == INET6_IFADDR_STATE_DAD)
			addrconf_dad_kick(ifp);
4068
		spin_unlock(&ifp->lock);
4069 4070 4071 4072
	}
	read_unlock_bh(&idev->lock);
}

L
Linus Torvalds 已提交
4073 4074
#ifdef CONFIG_PROC_FS
struct if6_iter_state {
4075
	struct seq_net_private p;
L
Linus Torvalds 已提交
4076
	int bucket;
4077
	int offset;
L
Linus Torvalds 已提交
4078 4079
};

4080
static struct inet6_ifaddr *if6_get_first(struct seq_file *seq, loff_t pos)
L
Linus Torvalds 已提交
4081 4082 4083
{
	struct inet6_ifaddr *ifa = NULL;
	struct if6_iter_state *state = seq->private;
4084
	struct net *net = seq_file_net(seq);
4085
	int p = 0;
L
Linus Torvalds 已提交
4086

4087 4088 4089 4090 4091 4092 4093
	/* initial bucket if pos is 0 */
	if (pos == 0) {
		state->bucket = 0;
		state->offset = 0;
	}

	for (; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
4094
		hlist_for_each_entry_rcu_bh(ifa, &inet6_addr_lst[state->bucket],
4095
					 addr_lst) {
4096 4097
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
4098 4099 4100 4101 4102 4103
			/* sync with offset */
			if (p < state->offset) {
				p++;
				continue;
			}
			state->offset++;
4104
			return ifa;
4105 4106 4107 4108 4109
		}

		/* prepare for next bucket */
		state->offset = 0;
		p = 0;
L
Linus Torvalds 已提交
4110
	}
4111
	return NULL;
L
Linus Torvalds 已提交
4112 4113
}

4114 4115
static struct inet6_ifaddr *if6_get_next(struct seq_file *seq,
					 struct inet6_ifaddr *ifa)
L
Linus Torvalds 已提交
4116 4117
{
	struct if6_iter_state *state = seq->private;
4118
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
4119

4120
	hlist_for_each_entry_continue_rcu_bh(ifa, addr_lst) {
4121 4122
		if (!net_eq(dev_net(ifa->idev->dev), net))
			continue;
4123
		state->offset++;
4124
		return ifa;
4125
	}
4126

4127
	while (++state->bucket < IN6_ADDR_HSIZE) {
4128
		state->offset = 0;
4129
		hlist_for_each_entry_rcu_bh(ifa,
4130
				     &inet6_addr_lst[state->bucket], addr_lst) {
4131 4132
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
4133
			state->offset++;
4134
			return ifa;
4135
		}
L
Linus Torvalds 已提交
4136
	}
4137

4138
	return NULL;
L
Linus Torvalds 已提交
4139 4140 4141
}

static void *if6_seq_start(struct seq_file *seq, loff_t *pos)
E
Eric Dumazet 已提交
4142
	__acquires(rcu_bh)
L
Linus Torvalds 已提交
4143
{
4144
	rcu_read_lock_bh();
4145
	return if6_get_first(seq, *pos);
L
Linus Torvalds 已提交
4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157
}

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

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

static void if6_seq_stop(struct seq_file *seq, void *v)
E
Eric Dumazet 已提交
4158
	__releases(rcu_bh)
L
Linus Torvalds 已提交
4159
{
4160
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
4161 4162 4163 4164 4165
}

static int if6_seq_show(struct seq_file *seq, void *v)
{
	struct inet6_ifaddr *ifp = (struct inet6_ifaddr *)v;
4166
	seq_printf(seq, "%pi6 %02x %02x %02x %02x %8s\n",
4167
		   &ifp->addr,
L
Linus Torvalds 已提交
4168 4169 4170
		   ifp->idev->dev->ifindex,
		   ifp->prefix_len,
		   ifp->scope,
4171
		   (u8) ifp->flags,
L
Linus Torvalds 已提交
4172 4173 4174 4175
		   ifp->idev->dev->name);
	return 0;
}

4176
static const struct seq_operations if6_seq_ops = {
L
Linus Torvalds 已提交
4177 4178 4179 4180 4181 4182 4183 4184
	.start	= if6_seq_start,
	.next	= if6_seq_next,
	.show	= if6_seq_show,
	.stop	= if6_seq_stop,
};

static int if6_seq_open(struct inode *inode, struct file *file)
{
4185 4186
	return seq_open_net(inode, file, &if6_seq_ops,
			    sizeof(struct if6_iter_state));
L
Linus Torvalds 已提交
4187 4188
}

4189
static const struct file_operations if6_fops = {
L
Linus Torvalds 已提交
4190 4191 4192 4193
	.owner		= THIS_MODULE,
	.open		= if6_seq_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
4194
	.release	= seq_release_net,
L
Linus Torvalds 已提交
4195 4196
};

4197
static int __net_init if6_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
4198
{
4199
	if (!proc_create("if_inet6", S_IRUGO, net->proc_net, &if6_fops))
L
Linus Torvalds 已提交
4200 4201 4202 4203
		return -ENOMEM;
	return 0;
}

4204
static void __net_exit if6_proc_net_exit(struct net *net)
4205
{
4206
	remove_proc_entry("if_inet6", net->proc_net);
4207 4208 4209
}

static struct pernet_operations if6_proc_net_ops = {
4210 4211
	.init = if6_proc_net_init,
	.exit = if6_proc_net_exit,
4212 4213 4214 4215 4216 4217 4218
};

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

L
Linus Torvalds 已提交
4219 4220
void if6_proc_exit(void)
{
4221
	unregister_pernet_subsys(&if6_proc_net_ops);
L
Linus Torvalds 已提交
4222 4223 4224
}
#endif	/* CONFIG_PROC_FS */

A
Amerigo Wang 已提交
4225
#if IS_ENABLED(CONFIG_IPV6_MIP6)
4226
/* Check if address is a home address configured on any interface. */
4227
int ipv6_chk_home_addr(struct net *net, const struct in6_addr *addr)
4228 4229
{
	int ret = 0;
4230
	struct inet6_ifaddr *ifp = NULL;
E
Eric Dumazet 已提交
4231
	unsigned int hash = inet6_addr_hash(addr);
4232

4233
	rcu_read_lock_bh();
4234
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
4235
		if (!net_eq(dev_net(ifp->idev->dev), net))
4236
			continue;
4237
		if (ipv6_addr_equal(&ifp->addr, addr) &&
4238 4239 4240 4241 4242
		    (ifp->flags & IFA_F_HOMEADDRESS)) {
			ret = 1;
			break;
		}
	}
4243
	rcu_read_unlock_bh();
4244 4245 4246 4247
	return ret;
}
#endif

L
Linus Torvalds 已提交
4248 4249 4250 4251
/*
 *	Periodic address status verification
 */

4252
static void addrconf_verify_rtnl(void)
L
Linus Torvalds 已提交
4253
{
4254
	unsigned long now, next, next_sec, next_sched;
L
Linus Torvalds 已提交
4255 4256 4257
	struct inet6_ifaddr *ifp;
	int i;

4258 4259
	ASSERT_RTNL();

4260
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
4261
	now = jiffies;
4262
	next = round_jiffies_up(now + ADDR_CHECK_FREQUENCY);
L
Linus Torvalds 已提交
4263

4264
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
4265

S
Stephen Hemminger 已提交
4266
	for (i = 0; i < IN6_ADDR_HSIZE; i++) {
L
Linus Torvalds 已提交
4267
restart:
4268
		hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[i], addr_lst) {
L
Linus Torvalds 已提交
4269 4270
			unsigned long age;

4271 4272 4273 4274 4275 4276
			/* 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
4277 4278 4279
				continue;

			spin_lock(&ifp->lock);
4280 4281
			/* We try to batch several events at once. */
			age = (now - ifp->tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
L
Linus Torvalds 已提交
4282

4283 4284
			if (ifp->valid_lft != INFINITY_LIFE_TIME &&
			    age >= ifp->valid_lft) {
L
Linus Torvalds 已提交
4285 4286 4287 4288
				spin_unlock(&ifp->lock);
				in6_ifa_hold(ifp);
				ipv6_del_addr(ifp);
				goto restart;
4289 4290 4291
			} else if (ifp->prefered_lft == INFINITY_LIFE_TIME) {
				spin_unlock(&ifp->lock);
				continue;
L
Linus Torvalds 已提交
4292
			} else if (age >= ifp->prefered_lft) {
4293
				/* jiffies - ifp->tstamp > age >= ifp->prefered_lft */
L
Linus Torvalds 已提交
4294 4295 4296 4297 4298 4299 4300
				int deprecate = 0;

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

4301 4302
				if ((ifp->valid_lft != INFINITY_LIFE_TIME) &&
				    (time_before(ifp->tstamp + ifp->valid_lft * HZ, next)))
L
Linus Torvalds 已提交
4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315
					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)) {
4316 4317
				unsigned long regen_advance = ifp->idev->cnf.regen_max_retry *
					ifp->idev->cnf.dad_transmits *
J
Jiri Pirko 已提交
4318
					NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME) / HZ;
4319

L
Linus Torvalds 已提交
4320 4321 4322 4323 4324 4325 4326 4327 4328
				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);
4329

4330 4331 4332
						spin_lock(&ifpub->lock);
						ifpub->regen_count = 0;
						spin_unlock(&ifpub->lock);
L
Linus Torvalds 已提交
4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349
						ipv6_create_tempaddr(ifpub, ifp);
						in6_ifa_put(ifpub);
						in6_ifa_put(ifp);
						goto restart;
					}
				} else if (time_before(ifp->tstamp + ifp->prefered_lft * HZ - regen_advance * HZ, next))
					next = ifp->tstamp + ifp->prefered_lft * HZ - regen_advance * HZ;
				spin_unlock(&ifp->lock);
			} else {
				/* ifp->prefered_lft <= ifp->valid_lft */
				if (time_before(ifp->tstamp + ifp->prefered_lft * HZ, next))
					next = ifp->tstamp + ifp->prefered_lft * HZ;
				spin_unlock(&ifp->lock);
			}
		}
	}

4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360
	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;

4361 4362
	ADBG(KERN_DEBUG "now = %lu, schedule = %lu, rounded schedule = %lu => %lu\n",
	      now, next, next_sec, next_sched);
4363
	mod_delayed_work(addrconf_wq, &addr_chk_work, next_sched - now);
4364
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
4365 4366
}

4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378
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);
}

4379 4380
static struct in6_addr *extract_addr(struct nlattr *addr, struct nlattr *local,
				     struct in6_addr **peer_pfx)
4381 4382 4383
{
	struct in6_addr *pfx = NULL;

4384 4385
	*peer_pfx = NULL;

4386 4387 4388 4389 4390
	if (addr)
		pfx = nla_data(addr);

	if (local) {
		if (pfx && nla_memcmp(local, pfx, sizeof(*pfx)))
4391 4392
			*peer_pfx = pfx;
		pfx = nla_data(local);
4393 4394 4395 4396 4397
	}

	return pfx;
}

4398
static const struct nla_policy ifa_ipv6_policy[IFA_MAX+1] = {
4399 4400 4401
	[IFA_ADDRESS]		= { .len = sizeof(struct in6_addr) },
	[IFA_LOCAL]		= { .len = sizeof(struct in6_addr) },
	[IFA_CACHEINFO]		= { .len = sizeof(struct ifa_cacheinfo) },
4402
	[IFA_FLAGS]		= { .len = sizeof(u32) },
4403 4404
};

L
Linus Torvalds 已提交
4405
static int
4406 4407
inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh,
		  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
4408
{
4409
	struct net *net = sock_net(skb->sk);
4410 4411
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4412
	struct in6_addr *pfx, *peer_pfx;
4413
	u32 ifa_flags;
4414
	int err;
L
Linus Torvalds 已提交
4415

4416
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy,
4417
			  extack);
4418 4419 4420 4421
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
4422
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
4423
	if (!pfx)
L
Linus Torvalds 已提交
4424 4425
		return -EINVAL;

4426 4427 4428 4429 4430 4431 4432
	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
4433 4434
}

4435
static int inet6_addr_modify(struct inet6_ifaddr *ifp, u32 ifa_flags,
4436
			     u32 prefered_lft, u32 valid_lft)
4437
{
4438 4439
	u32 flags;
	clock_t expires;
4440
	unsigned long timeout;
4441
	bool was_managetempaddr;
4442
	bool had_prefixroute;
4443

4444 4445
	ASSERT_RTNL();

4446 4447 4448
	if (!valid_lft || (prefered_lft > valid_lft))
		return -EINVAL;

4449 4450 4451 4452
	if (ifa_flags & IFA_F_MANAGETEMPADDR &&
	    (ifp->flags & IFA_F_TEMPORARY || ifp->prefix_len != 64))
		return -EINVAL;

4453 4454 4455 4456
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
4457
		flags = RTF_EXPIRES;
4458 4459 4460 4461
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
4462
	}
4463

4464 4465 4466 4467 4468 4469
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
4470 4471

	spin_lock_bh(&ifp->lock);
4472
	was_managetempaddr = ifp->flags & IFA_F_MANAGETEMPADDR;
4473 4474
	had_prefixroute = ifp->flags & IFA_F_PERMANENT &&
			  !(ifp->flags & IFA_F_NOPREFIXROUTE);
4475
	ifp->flags &= ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD |
4476 4477
			IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
			IFA_F_NOPREFIXROUTE);
4478
	ifp->flags |= ifa_flags;
4479 4480 4481 4482 4483 4484 4485 4486
	ifp->tstamp = jiffies;
	ifp->valid_lft = valid_lft;
	ifp->prefered_lft = prefered_lft;

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

4487 4488 4489
	if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, ifp->idev->dev,
				      expires, flags);
4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501
	} 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);
		}
4502
	}
4503 4504 4505 4506 4507 4508 4509 4510

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

4511
	addrconf_verify_rtnl();
4512 4513 4514 4515

	return 0;
}

L
Linus Torvalds 已提交
4516
static int
4517 4518
inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh,
		  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
4519
{
4520
	struct net *net = sock_net(skb->sk);
4521 4522
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4523
	struct in6_addr *pfx, *peer_pfx;
4524 4525
	struct inet6_ifaddr *ifa;
	struct net_device *dev;
4526
	u32 valid_lft = INFINITY_LIFE_TIME, preferred_lft = INFINITY_LIFE_TIME;
4527
	u32 ifa_flags;
4528
	int err;
L
Linus Torvalds 已提交
4529

4530
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy,
4531
			  extack);
4532 4533 4534 4535
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
4536
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
4537
	if (!pfx)
L
Linus Torvalds 已提交
4538 4539
		return -EINVAL;

4540
	if (tb[IFA_CACHEINFO]) {
4541
		struct ifa_cacheinfo *ci;
4542 4543

		ci = nla_data(tb[IFA_CACHEINFO]);
4544
		valid_lft = ci->ifa_valid;
4545 4546 4547 4548
		preferred_lft = ci->ifa_prefered;
	} else {
		preferred_lft = INFINITY_LIFE_TIME;
		valid_lft = INFINITY_LIFE_TIME;
4549 4550
	}

4551
	dev =  __dev_get_by_index(net, ifm->ifa_index);
4552
	if (!dev)
4553 4554
		return -ENODEV;

4555 4556
	ifa_flags = tb[IFA_FLAGS] ? nla_get_u32(tb[IFA_FLAGS]) : ifm->ifa_flags;

4557
	/* We ignore other flags so far. */
4558
	ifa_flags &= IFA_F_NODAD | IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
4559
		     IFA_F_NOPREFIXROUTE | IFA_F_MCAUTOJOIN;
4560

4561
	ifa = ipv6_get_ifaddr(net, pfx, dev, 1);
4562
	if (!ifa) {
4563 4564
		/*
		 * It would be best to check for !NLM_F_CREATE here but
4565
		 * userspace already relies on not having to provide this.
4566
		 */
4567
		return inet6_addr_add(net, ifm->ifa_index, pfx, peer_pfx,
4568 4569
				      ifm->ifa_prefixlen, ifa_flags,
				      preferred_lft, valid_lft);
4570 4571
	}

4572 4573 4574 4575
	if (nlh->nlmsg_flags & NLM_F_EXCL ||
	    !(nlh->nlmsg_flags & NLM_F_REPLACE))
		err = -EEXIST;
	else
4576
		err = inet6_addr_modify(ifa, ifa_flags, preferred_lft, valid_lft);
4577 4578 4579 4580

	in6_ifa_put(ifa);

	return err;
L
Linus Torvalds 已提交
4581 4582
}

4583
static void put_ifaddrmsg(struct nlmsghdr *nlh, u8 prefixlen, u32 flags,
4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595
			  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;
}

4596 4597 4598 4599 4600
static int put_cacheinfo(struct sk_buff *skb, unsigned long cstamp,
			 unsigned long tstamp, u32 preferred, u32 valid)
{
	struct ifa_cacheinfo ci;

4601 4602
	ci.cstamp = cstamp_delta(cstamp);
	ci.tstamp = cstamp_delta(tstamp);
4603 4604 4605 4606 4607 4608
	ci.ifa_prefered = preferred;
	ci.ifa_valid = valid;

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

4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620
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;
}

4621 4622
static inline int inet6_ifaddr_msgsize(void)
{
4623
	return NLMSG_ALIGN(sizeof(struct ifaddrmsg))
4624
	       + nla_total_size(16) /* IFA_LOCAL */
4625
	       + nla_total_size(16) /* IFA_ADDRESS */
4626 4627
	       + nla_total_size(sizeof(struct ifa_cacheinfo))
	       + nla_total_size(4)  /* IFA_FLAGS */;
4628
}
4629

L
Linus Torvalds 已提交
4630
static int inet6_fill_ifaddr(struct sk_buff *skb, struct inet6_ifaddr *ifa,
4631
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4632 4633
{
	struct nlmsghdr  *nlh;
4634
	u32 preferred, valid;
L
Linus Torvalds 已提交
4635

4636
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4637
	if (!nlh)
4638
		return -EMSGSIZE;
4639

4640 4641 4642
	put_ifaddrmsg(nlh, ifa->prefix_len, ifa->flags, rt_scope(ifa->scope),
		      ifa->idev->dev->ifindex);

4643 4644
	if (!((ifa->flags&IFA_F_PERMANENT) &&
	      (ifa->prefered_lft == INFINITY_LIFE_TIME))) {
4645 4646 4647
		preferred = ifa->prefered_lft;
		valid = ifa->valid_lft;
		if (preferred != INFINITY_LIFE_TIME) {
L
Linus Torvalds 已提交
4648
			long tval = (jiffies - ifa->tstamp)/HZ;
4649 4650 4651 4652
			if (preferred > tval)
				preferred -= tval;
			else
				preferred = 0;
4653 4654 4655 4656 4657 4658
			if (valid != INFINITY_LIFE_TIME) {
				if (valid > tval)
					valid -= tval;
				else
					valid = 0;
			}
L
Linus Torvalds 已提交
4659 4660
		}
	} else {
4661 4662 4663 4664
		preferred = INFINITY_LIFE_TIME;
		valid = INFINITY_LIFE_TIME;
	}

C
Cong Wang 已提交
4665
	if (!ipv6_addr_any(&ifa->peer_addr)) {
4666 4667
		if (nla_put_in6_addr(skb, IFA_LOCAL, &ifa->addr) < 0 ||
		    nla_put_in6_addr(skb, IFA_ADDRESS, &ifa->peer_addr) < 0)
4668 4669
			goto error;
	} else
4670
		if (nla_put_in6_addr(skb, IFA_ADDRESS, &ifa->addr) < 0)
4671 4672 4673 4674
			goto error;

	if (put_cacheinfo(skb, ifa->cstamp, ifa->tstamp, preferred, valid) < 0)
		goto error;
L
Linus Torvalds 已提交
4675

4676 4677 4678
	if (nla_put_u32(skb, IFA_FLAGS, ifa->flags) < 0)
		goto error;

4679 4680
	nlmsg_end(skb, nlh);
	return 0;
4681 4682 4683 4684

error:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4685 4686 4687
}

static int inet6_fill_ifmcaddr(struct sk_buff *skb, struct ifmcaddr6 *ifmca,
4688
				u32 portid, u32 seq, int event, u16 flags)
L
Linus Torvalds 已提交
4689 4690
{
	struct nlmsghdr  *nlh;
4691 4692 4693 4694 4695
	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
4696

4697
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4698
	if (!nlh)
4699
		return -EMSGSIZE;
4700

4701
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
4702
	if (nla_put_in6_addr(skb, IFA_MULTICAST, &ifmca->mca_addr) < 0 ||
4703
	    put_cacheinfo(skb, ifmca->mca_cstamp, ifmca->mca_tstamp,
4704 4705 4706 4707
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
4708

4709 4710
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4711 4712 4713
}

static int inet6_fill_ifacaddr(struct sk_buff *skb, struct ifacaddr6 *ifaca,
4714
				u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4715 4716
{
	struct nlmsghdr  *nlh;
4717 4718 4719 4720 4721
	u8 scope = RT_SCOPE_UNIVERSE;
	int ifindex = ifaca->aca_idev->dev->ifindex;

	if (ipv6_addr_scope(&ifaca->aca_addr) & IFA_SITE)
		scope = RT_SCOPE_SITE;
L
Linus Torvalds 已提交
4722

4723
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4724
	if (!nlh)
4725
		return -EMSGSIZE;
4726

4727
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
4728
	if (nla_put_in6_addr(skb, IFA_ANYCAST, &ifaca->aca_addr) < 0 ||
4729
	    put_cacheinfo(skb, ifaca->aca_cstamp, ifaca->aca_tstamp,
4730 4731 4732 4733
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
L
Linus Torvalds 已提交
4734

4735 4736
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
4737 4738
}

S
Stephen Hemminger 已提交
4739
enum addr_type_t {
L
Linus Torvalds 已提交
4740 4741 4742 4743 4744
	UNICAST_ADDR,
	MULTICAST_ADDR,
	ANYCAST_ADDR,
};

E
Eric Dumazet 已提交
4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755 4756
/* called with rcu_read_lock() */
static int in6_dump_addrs(struct inet6_dev *idev, struct sk_buff *skb,
			  struct netlink_callback *cb, enum addr_type_t type,
			  int s_ip_idx, int *p_ip_idx)
{
	struct ifmcaddr6 *ifmca;
	struct ifacaddr6 *ifaca;
	int err = 1;
	int ip_idx = *p_ip_idx;

	read_lock_bh(&idev->lock);
	switch (type) {
4757 4758 4759
	case UNICAST_ADDR: {
		struct inet6_ifaddr *ifa;

E
Eric Dumazet 已提交
4760
		/* unicast address incl. temp addr */
4761 4762
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
			if (++ip_idx < s_ip_idx)
E
Eric Dumazet 已提交
4763 4764
				continue;
			err = inet6_fill_ifaddr(skb, ifa,
4765
						NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4766 4767 4768
						cb->nlh->nlmsg_seq,
						RTM_NEWADDR,
						NLM_F_MULTI);
4769
			if (err < 0)
E
Eric Dumazet 已提交
4770
				break;
4771
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
E
Eric Dumazet 已提交
4772 4773
		}
		break;
4774
	}
E
Eric Dumazet 已提交
4775 4776 4777 4778 4779 4780 4781
	case MULTICAST_ADDR:
		/* multicast address */
		for (ifmca = idev->mc_list; ifmca;
		     ifmca = ifmca->next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
			err = inet6_fill_ifmcaddr(skb, ifmca,
4782
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4783 4784 4785
						  cb->nlh->nlmsg_seq,
						  RTM_GETMULTICAST,
						  NLM_F_MULTI);
4786
			if (err < 0)
E
Eric Dumazet 已提交
4787 4788 4789 4790 4791 4792 4793 4794 4795 4796
				break;
		}
		break;
	case ANYCAST_ADDR:
		/* anycast address */
		for (ifaca = idev->ac_list; ifaca;
		     ifaca = ifaca->aca_next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
			err = inet6_fill_ifacaddr(skb, ifaca,
4797
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4798 4799 4800
						  cb->nlh->nlmsg_seq,
						  RTM_GETANYCAST,
						  NLM_F_MULTI);
4801
			if (err < 0)
E
Eric Dumazet 已提交
4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812
				break;
		}
		break;
	default:
		break;
	}
	read_unlock_bh(&idev->lock);
	*p_ip_idx = ip_idx;
	return err;
}

L
Linus Torvalds 已提交
4813 4814 4815
static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
			   enum addr_type_t type)
{
E
Eric Dumazet 已提交
4816 4817
	struct net *net = sock_net(skb->sk);
	int h, s_h;
L
Linus Torvalds 已提交
4818 4819 4820
	int idx, ip_idx;
	int s_idx, s_ip_idx;
	struct net_device *dev;
E
Eric Dumazet 已提交
4821 4822
	struct inet6_dev *idev;
	struct hlist_head *head;
4823

E
Eric Dumazet 已提交
4824 4825 4826
	s_h = cb->args[0];
	s_idx = idx = cb->args[1];
	s_ip_idx = ip_idx = cb->args[2];
4827

E
Eric Dumazet 已提交
4828
	rcu_read_lock();
4829
	cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^ net->dev_base_seq;
E
Eric Dumazet 已提交
4830 4831 4832
	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
4833
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
4834 4835
			if (idx < s_idx)
				goto cont;
4836
			if (h > s_h || idx > s_idx)
E
Eric Dumazet 已提交
4837 4838
				s_ip_idx = 0;
			ip_idx = 0;
S
Stephen Hemminger 已提交
4839 4840
			idev = __in6_dev_get(dev);
			if (!idev)
E
Eric Dumazet 已提交
4841 4842 4843
				goto cont;

			if (in6_dump_addrs(idev, skb, cb, type,
4844
					   s_ip_idx, &ip_idx) < 0)
E
Eric Dumazet 已提交
4845
				goto done;
4846
cont:
E
Eric Dumazet 已提交
4847 4848
			idx++;
		}
L
Linus Torvalds 已提交
4849
	}
E
Eric Dumazet 已提交
4850 4851 4852 4853 4854 4855
done:
	rcu_read_unlock();
	cb->args[0] = h;
	cb->args[1] = idx;
	cb->args[2] = ip_idx;

L
Linus Torvalds 已提交
4856 4857 4858 4859 4860 4861
	return skb->len;
}

static int inet6_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = UNICAST_ADDR;
4862

L
Linus Torvalds 已提交
4863 4864 4865 4866 4867 4868
	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;
4869

L
Linus Torvalds 已提交
4870 4871 4872 4873 4874 4875 4876
	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;
4877

L
Linus Torvalds 已提交
4878 4879 4880
	return inet6_dump_addr(skb, cb, type);
}

4881 4882
static int inet6_rtm_getaddr(struct sk_buff *in_skb, struct nlmsghdr *nlh,
			     struct netlink_ext_ack *extack)
4883
{
4884
	struct net *net = sock_net(in_skb->sk);
4885 4886
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4887
	struct in6_addr *addr = NULL, *peer;
4888 4889 4890 4891 4892
	struct net_device *dev = NULL;
	struct inet6_ifaddr *ifa;
	struct sk_buff *skb;
	int err;

4893
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy,
4894
			  extack);
4895 4896 4897
	if (err < 0)
		goto errout;

4898
	addr = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer);
4899
	if (!addr) {
4900 4901
		err = -EINVAL;
		goto errout;
4902 4903
	}

4904
	ifm = nlmsg_data(nlh);
4905
	if (ifm->ifa_index)
4906
		dev = __dev_get_by_index(net, ifm->ifa_index);
4907

S
Stephen Hemminger 已提交
4908 4909
	ifa = ipv6_get_ifaddr(net, addr, dev, 1);
	if (!ifa) {
4910 4911 4912
		err = -EADDRNOTAVAIL;
		goto errout;
	}
4913

S
Stephen Hemminger 已提交
4914 4915
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_KERNEL);
	if (!skb) {
4916
		err = -ENOBUFS;
4917
		goto errout_ifa;
4918 4919
	}

4920
	err = inet6_fill_ifaddr(skb, ifa, NETLINK_CB(in_skb).portid,
4921
				nlh->nlmsg_seq, RTM_NEWADDR, 0);
4922 4923 4924 4925 4926 4927
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout_ifa;
	}
4928
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
4929
errout_ifa:
4930
	in6_ifa_put(ifa);
4931
errout:
4932 4933 4934
	return err;
}

L
Linus Torvalds 已提交
4935 4936 4937
static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa)
{
	struct sk_buff *skb;
4938
	struct net *net = dev_net(ifa->idev->dev);
4939
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
4940

4941 4942 4943 4944 4945 4946 4947
	/* Don't send DELADDR notification for TENTATIVE address,
	 * since NEWADDR notification is sent only after removing
	 * TENTATIVE flag.
	 */
	if (ifa->flags & IFA_F_TENTATIVE && event == RTM_DELADDR)
		return;

4948
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_ATOMIC);
4949
	if (!skb)
4950 4951 4952
		goto errout;

	err = inet6_fill_ifaddr(skb, ifa, 0, 0, event, 0);
4953 4954 4955 4956 4957 4958
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4959 4960
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFADDR, NULL, GFP_ATOMIC);
	return;
4961 4962
errout:
	if (err < 0)
4963
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFADDR, err);
L
Linus Torvalds 已提交
4964 4965
}

D
Dave Jones 已提交
4966
static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
L
Linus Torvalds 已提交
4967 4968
				__s32 *array, int bytes)
{
4969 4970
	BUG_ON(bytes < (DEVCONF_MAX * 4));

L
Linus Torvalds 已提交
4971 4972 4973 4974 4975 4976 4977 4978 4979
	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;
4980 4981
	array[DEVCONF_RTR_SOLICIT_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_interval);
4982 4983
	array[DEVCONF_RTR_SOLICIT_MAX_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_max_interval);
4984 4985
	array[DEVCONF_RTR_SOLICIT_DELAY] =
		jiffies_to_msecs(cnf->rtr_solicit_delay);
L
Linus Torvalds 已提交
4986
	array[DEVCONF_FORCE_MLD_VERSION] = cnf->force_mld_version;
4987 4988 4989 4990
	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
4991 4992 4993 4994 4995 4996
	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;
4997
	array[DEVCONF_ACCEPT_RA_DEFRTR] = cnf->accept_ra_defrtr;
4998
	array[DEVCONF_ACCEPT_RA_MIN_HOP_LIMIT] = cnf->accept_ra_min_hop_limit;
4999
	array[DEVCONF_ACCEPT_RA_PINFO] = cnf->accept_ra_pinfo;
5000 5001
#ifdef CONFIG_IPV6_ROUTER_PREF
	array[DEVCONF_ACCEPT_RA_RTR_PREF] = cnf->accept_ra_rtr_pref;
5002 5003
	array[DEVCONF_RTR_PROBE_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_probe_interval);
N
Neil Horman 已提交
5004
#ifdef CONFIG_IPV6_ROUTE_INFO
5005
	array[DEVCONF_ACCEPT_RA_RT_INFO_MIN_PLEN] = cnf->accept_ra_rt_info_min_plen;
5006 5007
	array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
#endif
5008
#endif
5009
	array[DEVCONF_PROXY_NDP] = cnf->proxy_ndp;
5010
	array[DEVCONF_ACCEPT_SOURCE_ROUTE] = cnf->accept_source_route;
5011 5012
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
5013
	array[DEVCONF_USE_OPTIMISTIC] = cnf->use_optimistic;
5014
#endif
5015 5016 5017
#ifdef CONFIG_IPV6_MROUTE
	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
#endif
5018
	array[DEVCONF_DISABLE_IPV6] = cnf->disable_ipv6;
5019
	array[DEVCONF_ACCEPT_DAD] = cnf->accept_dad;
5020
	array[DEVCONF_FORCE_TLLAO] = cnf->force_tllao;
5021
	array[DEVCONF_NDISC_NOTIFY] = cnf->ndisc_notify;
5022
	array[DEVCONF_SUPPRESS_FRAG_NDISC] = cnf->suppress_frag_ndisc;
5023
	array[DEVCONF_ACCEPT_RA_FROM_LOCAL] = cnf->accept_ra_from_local;
5024
	array[DEVCONF_ACCEPT_RA_MTU] = cnf->accept_ra_mtu;
5025
	array[DEVCONF_IGNORE_ROUTES_WITH_LINKDOWN] = cnf->ignore_routes_with_linkdown;
5026
	/* we omit DEVCONF_STABLE_SECRET for now */
5027
	array[DEVCONF_USE_OIF_ADDRS_ONLY] = cnf->use_oif_addrs_only;
5028
	array[DEVCONF_DROP_UNICAST_IN_L2_MULTICAST] = cnf->drop_unicast_in_l2_multicast;
5029
	array[DEVCONF_DROP_UNSOLICITED_NA] = cnf->drop_unsolicited_na;
5030
	array[DEVCONF_KEEP_ADDR_ON_DOWN] = cnf->keep_addr_on_down;
5031
	array[DEVCONF_SEG6_ENABLED] = cnf->seg6_enabled;
5032 5033 5034
#ifdef CONFIG_IPV6_SEG6_HMAC
	array[DEVCONF_SEG6_REQUIRE_HMAC] = cnf->seg6_require_hmac;
#endif
5035
	array[DEVCONF_ENHANCED_DAD] = cnf->enhanced_dad;
5036
	array[DEVCONF_ADDR_GEN_MODE] = cnf->addr_gen_mode;
5037
	array[DEVCONF_DISABLE_POLICY] = cnf->disable_policy;
L
Linus Torvalds 已提交
5038 5039
}

T
Thomas Graf 已提交
5040 5041 5042 5043 5044 5045
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 */
5046 5047
	     + nla_total_size(ICMP6_MIB_MAX * 8) /* IFLA_INET6_ICMP6STATS */
	     + nla_total_size(sizeof(struct in6_addr)); /* IFLA_INET6_TOKEN */
T
Thomas Graf 已提交
5048 5049
}

5050 5051 5052 5053 5054 5055 5056
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 */
5057
	       + nla_total_size(1) /* IFLA_OPERSTATE */
T
Thomas Graf 已提交
5058
	       + nla_total_size(inet6_ifla6_size()); /* IFLA_PROTINFO */
5059
}
5060

5061
static inline void __snmp6_fill_statsdev(u64 *stats, atomic_long_t *mib,
5062
					int bytes)
5063 5064
{
	int i;
5065
	int pad = bytes - sizeof(u64) * ICMP6_MIB_MAX;
5066 5067 5068
	BUG_ON(pad < 0);

	/* Use put_unaligned() because stats may not be aligned for u64. */
5069 5070
	put_unaligned(ICMP6_MIB_MAX, &stats[0]);
	for (i = 1; i < ICMP6_MIB_MAX; i++)
5071
		put_unaligned(atomic_long_read(&mib[i]), &stats[i]);
5072

5073
	memset(&stats[ICMP6_MIB_MAX], 0, pad);
5074 5075
}

W
WANG Cong 已提交
5076
static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
5077
					int bytes, size_t syncpoff)
5078
{
5079 5080 5081 5082
	int i, c;
	u64 buff[IPSTATS_MIB_MAX];
	int pad = bytes - sizeof(u64) * IPSTATS_MIB_MAX;

5083 5084
	BUG_ON(pad < 0);

5085 5086
	memset(buff, 0, sizeof(buff));
	buff[0] = IPSTATS_MIB_MAX;
5087

5088 5089 5090 5091 5092 5093 5094
	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);
5095 5096
}

5097 5098 5099
static void snmp6_fill_stats(u64 *stats, struct inet6_dev *idev, int attrtype,
			     int bytes)
{
S
Stephen Hemminger 已提交
5100
	switch (attrtype) {
5101
	case IFLA_INET6_STATS:
5102 5103
		__snmp6_fill_stats64(stats, idev->stats.ipv6, bytes,
				     offsetof(struct ipstats_mib, syncp));
5104 5105
		break;
	case IFLA_INET6_ICMP6STATS:
5106
		__snmp6_fill_statsdev(stats, idev->stats.icmpv6dev->mibs, bytes);
5107 5108 5109 5110
		break;
	}
}

5111 5112
static int inet6_fill_ifla6_attrs(struct sk_buff *skb, struct inet6_dev *idev,
				  u32 ext_filter_mask)
T
Thomas Graf 已提交
5113 5114 5115 5116
{
	struct nlattr *nla;
	struct ifla_cacheinfo ci;

D
David S. Miller 已提交
5117 5118
	if (nla_put_u32(skb, IFLA_INET6_FLAGS, idev->if_flags))
		goto nla_put_failure;
T
Thomas Graf 已提交
5119
	ci.max_reasm_len = IPV6_MAXPLEN;
5120 5121
	ci.tstamp = cstamp_delta(idev->tstamp);
	ci.reachable_time = jiffies_to_msecs(idev->nd_parms->reachable_time);
J
Jiri Pirko 已提交
5122
	ci.retrans_time = jiffies_to_msecs(NEIGH_VAR(idev->nd_parms, RETRANS_TIME));
D
David S. Miller 已提交
5123 5124
	if (nla_put(skb, IFLA_INET6_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
T
Thomas Graf 已提交
5125
	nla = nla_reserve(skb, IFLA_INET6_CONF, DEVCONF_MAX * sizeof(s32));
5126
	if (!nla)
T
Thomas Graf 已提交
5127 5128 5129 5130 5131
		goto nla_put_failure;
	ipv6_store_devconf(&idev->cnf, nla_data(nla), nla_len(nla));

	/* XXX - MC not implemented */

5132 5133 5134
	if (ext_filter_mask & RTEXT_FILTER_SKIP_STATS)
		return 0;

T
Thomas Graf 已提交
5135
	nla = nla_reserve(skb, IFLA_INET6_STATS, IPSTATS_MIB_MAX * sizeof(u64));
5136
	if (!nla)
T
Thomas Graf 已提交
5137 5138 5139 5140
		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));
5141
	if (!nla)
T
Thomas Graf 已提交
5142 5143 5144
		goto nla_put_failure;
	snmp6_fill_stats(nla_data(nla), idev, IFLA_INET6_ICMP6STATS, nla_len(nla));

5145
	nla = nla_reserve(skb, IFLA_INET6_TOKEN, sizeof(struct in6_addr));
5146
	if (!nla)
5147
		goto nla_put_failure;
5148

5149
	if (nla_put_u8(skb, IFLA_INET6_ADDR_GEN_MODE, idev->cnf.addr_gen_mode))
5150 5151
		goto nla_put_failure;

5152 5153 5154 5155
	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
5156 5157 5158 5159 5160 5161
	return 0;

nla_put_failure:
	return -EMSGSIZE;
}

5162 5163
static size_t inet6_get_link_af_size(const struct net_device *dev,
				     u32 ext_filter_mask)
T
Thomas Graf 已提交
5164 5165 5166 5167 5168 5169 5170
{
	if (!__in6_dev_get(dev))
		return 0;

	return inet6_ifla6_size();
}

5171 5172
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
5173 5174 5175 5176 5177 5178
{
	struct inet6_dev *idev = __in6_dev_get(dev);

	if (!idev)
		return -ENODATA;

5179
	if (inet6_fill_ifla6_attrs(skb, idev, ext_filter_mask) < 0)
T
Thomas Graf 已提交
5180 5181 5182 5183 5184
		return -EMSGSIZE;

	return 0;
}

5185 5186 5187 5188
static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token)
{
	struct inet6_ifaddr *ifp;
	struct net_device *dev = idev->dev;
5189
	bool clear_token, update_rs = false;
5190
	struct in6_addr ll_addr;
5191

5192 5193
	ASSERT_RTNL();

5194
	if (!token)
5195 5196 5197 5198 5199
		return -EINVAL;
	if (dev->flags & (IFF_LOOPBACK | IFF_NOARP))
		return -EINVAL;
	if (!ipv6_accept_ra(idev))
		return -EINVAL;
5200
	if (idev->cnf.rtr_solicits == 0)
5201 5202 5203 5204 5205 5206 5207 5208 5209
		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);

5210 5211 5212 5213
	clear_token = ipv6_addr_any(token);
	if (clear_token)
		goto update_lft;

5214 5215 5216
	if (!idev->dead && (idev->if_flags & IF_READY) &&
	    !ipv6_get_lladdr(dev, &ll_addr, IFA_F_TENTATIVE |
			     IFA_F_OPTIMISTIC)) {
5217 5218 5219 5220 5221 5222
		/* 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;
	}
5223

5224
update_lft:
5225
	write_lock_bh(&idev->lock);
5226

5227
	if (update_rs) {
5228
		idev->if_flags |= IF_RS_SENT;
5229 5230
		idev->rs_interval = rfc3315_s14_backoff_init(
			idev->cnf.rtr_solicit_interval);
5231
		idev->rs_probes = 1;
5232
		addrconf_mod_rs_timer(idev, idev->rs_interval);
5233
	}
5234 5235 5236 5237

	/* Well, that's kinda nasty ... */
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
		spin_lock(&ifp->lock);
5238
		if (ifp->tokenized) {
5239 5240 5241 5242 5243 5244 5245
			ifp->valid_lft = 0;
			ifp->prefered_lft = 0;
		}
		spin_unlock(&ifp->lock);
	}

	write_unlock_bh(&idev->lock);
5246
	inet6_ifinfo_notify(RTM_NEWLINK, idev);
5247
	addrconf_verify_rtnl();
5248 5249 5250
	return 0;
}

5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263
static const struct nla_policy inet6_af_policy[IFLA_INET6_MAX + 1] = {
	[IFLA_INET6_ADDR_GEN_MODE]	= { .type = NLA_U8 },
	[IFLA_INET6_TOKEN]		= { .len = sizeof(struct in6_addr) },
};

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

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

5264 5265
	return nla_parse_nested(tb, IFLA_INET6_MAX, nla, inet6_af_policy,
				NULL);
5266 5267
}

5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287
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;
}

5288 5289 5290 5291 5292 5293 5294 5295 5296
static int inet6_set_link_af(struct net_device *dev, const struct nlattr *nla)
{
	int err = -EINVAL;
	struct inet6_dev *idev = __in6_dev_get(dev);
	struct nlattr *tb[IFLA_INET6_MAX + 1];

	if (!idev)
		return -EAFNOSUPPORT;

5297
	if (nla_parse_nested(tb, IFLA_INET6_MAX, nla, NULL, NULL) < 0)
5298 5299
		BUG();

5300
	if (tb[IFLA_INET6_TOKEN]) {
5301
		err = inet6_set_iftoken(idev, nla_data(tb[IFLA_INET6_TOKEN]));
5302 5303 5304 5305 5306 5307 5308
		if (err)
			return err;
	}

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

5309 5310
		if (check_addr_gen_mode(mode) < 0 ||
		    check_stable_privacy(idev, dev_net(dev), mode) < 0)
5311
			return -EINVAL;
5312

5313
		idev->cnf.addr_gen_mode = mode;
5314 5315
		err = 0;
	}
5316 5317 5318 5319

	return err;
}

5320
static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
5321
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
5322
{
5323 5324 5325 5326 5327
	struct net_device *dev = idev->dev;
	struct ifinfomsg *hdr;
	struct nlmsghdr *nlh;
	void *protoinfo;

5328
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*hdr), flags);
5329
	if (!nlh)
5330
		return -EMSGSIZE;
5331 5332 5333 5334 5335 5336 5337 5338 5339

	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
5340 5341 5342 5343
	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) ||
5344
	    (dev->ifindex != dev_get_iflink(dev) &&
5345 5346 5347
	     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
5348
		goto nla_put_failure;
5349
	protoinfo = nla_nest_start(skb, IFLA_PROTINFO);
5350
	if (!protoinfo)
5351
		goto nla_put_failure;
L
Linus Torvalds 已提交
5352

5353
	if (inet6_fill_ifla6_attrs(skb, idev, 0) < 0)
5354
		goto nla_put_failure;
L
Linus Torvalds 已提交
5355

5356
	nla_nest_end(skb, protoinfo);
5357 5358
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
5359

5360
nla_put_failure:
5361 5362
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
5363 5364 5365 5366
}

static int inet6_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
{
5367
	struct net *net = sock_net(skb->sk);
E
Eric Dumazet 已提交
5368
	int h, s_h;
5369
	int idx = 0, s_idx;
L
Linus Torvalds 已提交
5370 5371
	struct net_device *dev;
	struct inet6_dev *idev;
E
Eric Dumazet 已提交
5372
	struct hlist_head *head;
L
Linus Torvalds 已提交
5373

E
Eric Dumazet 已提交
5374 5375 5376 5377 5378 5379 5380
	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];
5381
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
5382 5383 5384 5385 5386 5387
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;
			if (inet6_fill_ifinfo(skb, idev,
5388
					      NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
5389
					      cb->nlh->nlmsg_seq,
5390
					      RTM_NEWLINK, NLM_F_MULTI) < 0)
E
Eric Dumazet 已提交
5391
				goto out;
5392
cont:
E
Eric Dumazet 已提交
5393 5394
			idx++;
		}
L
Linus Torvalds 已提交
5395
	}
E
Eric Dumazet 已提交
5396 5397 5398 5399
out:
	rcu_read_unlock();
	cb->args[1] = idx;
	cb->args[0] = h;
L
Linus Torvalds 已提交
5400 5401 5402 5403 5404 5405 5406

	return skb->len;
}

void inet6_ifinfo_notify(int event, struct inet6_dev *idev)
{
	struct sk_buff *skb;
5407
	struct net *net = dev_net(idev->dev);
5408
	int err = -ENOBUFS;
5409

5410
	skb = nlmsg_new(inet6_if_nlmsg_size(), GFP_ATOMIC);
5411
	if (!skb)
5412 5413 5414
		goto errout;

	err = inet6_fill_ifinfo(skb, idev, 0, 0, event, 0);
5415 5416 5417 5418 5419 5420
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_if_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5421
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFINFO, NULL, GFP_ATOMIC);
5422
	return;
5423 5424
errout:
	if (err < 0)
5425
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFINFO, err);
L
Linus Torvalds 已提交
5426 5427
}

5428 5429 5430 5431 5432 5433
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));
}
5434

L
Linus Torvalds 已提交
5435
static int inet6_fill_prefix(struct sk_buff *skb, struct inet6_dev *idev,
5436
			     struct prefix_info *pinfo, u32 portid, u32 seq,
5437
			     int event, unsigned int flags)
L
Linus Torvalds 已提交
5438
{
5439 5440
	struct prefixmsg *pmsg;
	struct nlmsghdr *nlh;
L
Linus Torvalds 已提交
5441 5442
	struct prefix_cacheinfo	ci;

5443
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*pmsg), flags);
5444
	if (!nlh)
5445
		return -EMSGSIZE;
5446 5447

	pmsg = nlmsg_data(nlh);
L
Linus Torvalds 已提交
5448
	pmsg->prefix_family = AF_INET6;
5449 5450
	pmsg->prefix_pad1 = 0;
	pmsg->prefix_pad2 = 0;
L
Linus Torvalds 已提交
5451 5452 5453
	pmsg->prefix_ifindex = idev->dev->ifindex;
	pmsg->prefix_len = pinfo->prefix_len;
	pmsg->prefix_type = pinfo->type;
5454
	pmsg->prefix_pad3 = 0;
L
Linus Torvalds 已提交
5455 5456 5457 5458 5459 5460
	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
5461 5462
	if (nla_put(skb, PREFIX_ADDRESS, sizeof(pinfo->prefix), &pinfo->prefix))
		goto nla_put_failure;
L
Linus Torvalds 已提交
5463 5464
	ci.preferred_time = ntohl(pinfo->prefered);
	ci.valid_time = ntohl(pinfo->valid);
D
David S. Miller 已提交
5465 5466
	if (nla_put(skb, PREFIX_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
5467 5468
	nlmsg_end(skb, nlh);
	return 0;
L
Linus Torvalds 已提交
5469

5470
nla_put_failure:
5471 5472
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
5473 5474
}

5475
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
5476 5477 5478
			 struct prefix_info *pinfo)
{
	struct sk_buff *skb;
5479
	struct net *net = dev_net(idev->dev);
5480
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
5481

5482
	skb = nlmsg_new(inet6_prefix_nlmsg_size(), GFP_ATOMIC);
5483
	if (!skb)
5484 5485 5486
		goto errout;

	err = inet6_fill_prefix(skb, idev, pinfo, 0, 0, event, 0);
5487 5488 5489 5490 5491 5492
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_prefix_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
5493 5494
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_PREFIX, NULL, GFP_ATOMIC);
	return;
5495 5496
errout:
	if (err < 0)
5497
		rtnl_set_sk_err(net, RTNLGRP_IPV6_PREFIX, err);
L
Linus Torvalds 已提交
5498 5499 5500 5501
}

static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
5502 5503
	struct net *net = dev_net(ifp->idev->dev);

5504 5505 5506
	if (event)
		ASSERT_RTNL();

L
Linus Torvalds 已提交
5507 5508 5509 5510
	inet6_ifa_notify(event ? : RTM_NEWADDR, ifp);

	switch (event) {
	case RTM_NEWADDR:
5511 5512 5513 5514 5515 5516 5517 5518
		/*
		 * If the address was optimistic
		 * we inserted the route at the start of
		 * our DAD process, so we don't need
		 * to do it again
		 */
		if (!(ifp->rt->rt6i_node))
			ip6_ins_rt(ifp->rt);
L
Linus Torvalds 已提交
5519 5520
		if (ifp->idev->cnf.forwarding)
			addrconf_join_anycast(ifp);
C
Cong Wang 已提交
5521
		if (!ipv6_addr_any(&ifp->peer_addr))
5522 5523
			addrconf_prefix_route(&ifp->peer_addr, 128,
					      ifp->idev->dev, 0, 0);
L
Linus Torvalds 已提交
5524 5525 5526 5527 5528
		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
5529
		if (!ipv6_addr_any(&ifp->peer_addr)) {
5530 5531
			struct rt6_info *rt;

5532 5533
			rt = addrconf_get_prefix_route(&ifp->peer_addr, 128,
						       ifp->idev->dev, 0, 0);
M
Martin KaFai Lau 已提交
5534 5535
			if (rt)
				ip6_del_rt(rt);
5536
		}
5537
		if (ifp->rt) {
W
Wei Wang 已提交
5538 5539
			if (dst_hold_safe(&ifp->rt->dst))
				ip6_del_rt(ifp->rt);
5540
		}
H
Hannes Frederic Sowa 已提交
5541
		rt_genid_bump_ipv6(net);
L
Linus Torvalds 已提交
5542 5543
		break;
	}
5544
	atomic_inc(&net->ipv6.dev_addr_genid);
L
Linus Torvalds 已提交
5545 5546 5547 5548
}

static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
5549
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
5550 5551
	if (likely(ifp->idev->dead == 0))
		__ipv6_ifa_notify(event, ifp);
5552
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
5553 5554 5555 5556 5557
}

#ifdef CONFIG_SYSCTL

static
5558
int addrconf_sysctl_forward(struct ctl_table *ctl, int write,
L
Linus Torvalds 已提交
5559 5560 5561 5562
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
5563
	loff_t pos = *ppos;
5564
	struct ctl_table lctl;
L
Linus Torvalds 已提交
5565 5566
	int ret;

5567 5568 5569 5570 5571 5572 5573 5574
	/*
	 * 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
5575

5576
	if (write)
5577
		ret = addrconf_fixup_forwarding(ctl, valp, val);
E
Eric W. Biederman 已提交
5578 5579
	if (ret)
		*ppos = pos;
5580
	return ret;
L
Linus Torvalds 已提交
5581 5582
}

5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596 5597
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);
}

5598 5599
static void dev_disable_change(struct inet6_dev *idev)
{
5600 5601
	struct netdev_notifier_info info;

5602 5603 5604
	if (!idev || !idev->dev)
		return;

5605
	netdev_notifier_info_init(&info, idev->dev);
5606
	if (idev->cnf.disable_ipv6)
5607
		addrconf_notify(NULL, NETDEV_DOWN, &info);
5608
	else
5609
		addrconf_notify(NULL, NETDEV_UP, &info);
5610 5611 5612 5613 5614 5615 5616
}

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

5617
	for_each_netdev(net, dev) {
5618 5619 5620 5621 5622 5623 5624 5625 5626 5627
		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);
		}
	}
}

5628
static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
5629 5630
{
	struct net *net;
5631 5632 5633 5634
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
5635 5636

	net = (struct net *)table->extra2;
5637 5638
	old = *p;
	*p = newf;
5639

5640 5641
	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
		rtnl_unlock();
5642
		return 0;
E
Eric W. Biederman 已提交
5643
	}
5644 5645 5646 5647

	if (p == &net->ipv6.devconf_all->disable_ipv6) {
		net->ipv6.devconf_dflt->disable_ipv6 = newf;
		addrconf_disable_change(net, newf);
5648
	} else if ((!newf) ^ (!old))
5649 5650 5651 5652 5653 5654 5655
		dev_disable_change((struct inet6_dev *)table->extra1);

	rtnl_unlock();
	return 0;
}

static
5656
int addrconf_sysctl_disable(struct ctl_table *ctl, int write,
5657 5658 5659 5660
			    void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
5661
	loff_t pos = *ppos;
5662
	struct ctl_table lctl;
5663 5664
	int ret;

5665 5666 5667 5668 5669 5670 5671 5672
	/*
	 * 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);
5673 5674 5675

	if (write)
		ret = addrconf_disable_ipv6(ctl, valp, val);
E
Eric W. Biederman 已提交
5676 5677
	if (ret)
		*ppos = pos;
5678 5679 5680
	return ret;
}

5681 5682 5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696 5697 5698 5699
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)
5700 5701
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
5702 5703 5704
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
		else if (valp == &net->ipv6.devconf_all->proxy_ndp)
5705 5706
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
5707 5708 5709 5710 5711
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
		else {
			struct inet6_dev *idev = ctl->extra1;

5712 5713
			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
						     NETCONFA_PROXY_NEIGH,
5714 5715 5716 5717 5718 5719 5720 5721 5722
						     idev->dev->ifindex,
						     &idev->cnf);
		}
		rtnl_unlock();
	}

	return ret;
}

5723 5724 5725 5726 5727 5728 5729 5730 5731
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;
	int new_val;
	struct inet6_dev *idev = (struct inet6_dev *)ctl->extra1;
	struct net *net = (struct net *)ctl->extra2;

5732 5733 5734
	if (!rtnl_trylock())
		return restart_syscall();

5735 5736 5737 5738 5739
	ret = proc_dointvec(ctl, write, buffer, lenp, ppos);

	if (write) {
		new_val = *((int *)ctl->data);

5740 5741 5742 5743
		if (check_addr_gen_mode(new_val) < 0) {
			ret = -EINVAL;
			goto out;
		}
5744 5745 5746 5747 5748 5749 5750 5751

		/* request for default */
		if (&net->ipv6.devconf_dflt->addr_gen_mode == ctl->data) {
			ipv6_devconf_dflt.addr_gen_mode = new_val;

		/* request for individual net device */
		} else {
			if (!idev)
5752
				goto out;
5753

5754 5755 5756 5757
			if (check_stable_privacy(idev, net, new_val) < 0) {
				ret = -EINVAL;
				goto out;
			}
5758 5759 5760 5761 5762 5763 5764 5765

			if (idev->cnf.addr_gen_mode != new_val) {
				idev->cnf.addr_gen_mode = new_val;
				addrconf_dev_config(idev->dev);
			}
		}
	}

5766 5767 5768
out:
	rtnl_unlock();

5769 5770 5771
	return ret;
}

5772 5773 5774 5775 5776 5777 5778 5779
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;
5780
	struct net *net = ctl->extra2;
5781 5782
	struct ipv6_stable_secret *secret = ctl->data;

5783 5784 5785
	if (&net->ipv6.devconf_all->stable_secret == ctl->data)
		return -EIO;

5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796
	lctl.maxlen = IPV6_MAX_STRLEN;
	lctl.data = str;

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

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

5797 5798 5799 5800
	err = snprintf(str, sizeof(str), "%pI6", &secret->secret);
	if (err >= sizeof(str)) {
		err = -EIO;
		goto out;
5801 5802 5803 5804 5805 5806 5807 5808 5809 5810 5811 5812 5813 5814
	}

	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;

5815 5816 5817 5818 5819 5820 5821
	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) {
5822
				idev->cnf.addr_gen_mode =
5823 5824 5825 5826 5827 5828
					IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
			}
		}
	} else {
		struct inet6_dev *idev = ctl->extra1;

5829
		idev->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_STABLE_PRIVACY;
5830 5831
	}

5832 5833 5834 5835 5836
out:
	rtnl_unlock();

	return err;
}
5837

5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849 5850 5851 5852 5853 5854 5855 5856 5857 5858 5859 5860 5861 5862 5863 5864 5865
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;
}

5866 5867 5868 5869 5870 5871 5872 5873 5874 5875 5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891 5892 5893 5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916 5917 5918 5919 5920 5921 5922 5923 5924 5925 5926 5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943 5944 5945 5946 5947 5948 5949 5950 5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964
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) {
			struct rt6_info *rt = ifa->rt;
			struct fib6_table *table = rt->rt6i_table;
			int cpu;

			read_lock(&table->tb6_lock);
			addrconf_set_nopolicy(ifa->rt, val);
			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);
				}
			}
			read_unlock(&table->tb6_lock);
		}
		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;
}

5965
static int minus_one = -1;
5966 5967 5968
static const int one = 1;
static const int two_five_five = 255;

5969
static const struct ctl_table addrconf_sysctl[] = {
5970 5971 5972 5973 5974 5975 5976 5977 5978 5979 5980 5981
	{
		.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,
5982 5983 5984
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= (void *)&one,
		.extra2		= (void *)&two_five_five,
5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000 6001 6002 6003 6004 6005 6006 6007 6008 6009 6010 6011 6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022 6023 6024 6025
	},
	{
		.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,
6026 6027
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &minus_one,
6028 6029 6030 6031 6032 6033 6034 6035
	},
	{
		.procname	= "router_solicitation_interval",
		.data		= &ipv6_devconf.rtr_solicit_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
6036 6037 6038 6039 6040 6041 6042
	{
		.procname	= "router_solicitation_max_interval",
		.data		= &ipv6_devconf.rtr_solicit_max_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
6043 6044 6045 6046 6047 6048 6049 6050 6051 6052 6053 6054 6055 6056 6057 6058 6059 6060 6061 6062 6063 6064 6065 6066 6067 6068 6069 6070 6071 6072 6073 6074 6075 6076 6077 6078 6079 6080 6081 6082 6083 6084 6085 6086 6087 6088 6089 6090 6091 6092 6093 6094 6095 6096 6097 6098 6099 6100 6101 6102 6103 6104 6105 6106 6107 6108 6109 6110 6111 6112 6113 6114 6115 6116 6117 6118 6119 6120 6121 6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135
	{
		.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,
	},
6136
#ifdef CONFIG_IPV6_ROUTER_PREF
6137 6138 6139 6140 6141 6142 6143 6144 6145 6146 6147 6148 6149 6150
	{
		.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
6151
#ifdef CONFIG_IPV6_ROUTE_INFO
6152 6153 6154 6155 6156 6157 6158
	{
		.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,
	},
6159 6160 6161 6162 6163 6164 6165
	{
		.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,
	},
6166
#endif
6167
#endif
6168 6169 6170 6171 6172 6173 6174 6175 6176 6177 6178 6179 6180 6181
	{
		.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,
	},
6182
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
6183 6184 6185 6186 6187 6188 6189 6190 6191 6192 6193 6194 6195 6196
	{
		.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,
	},
6197 6198
#endif
#ifdef CONFIG_IPV6_MROUTE
6199 6200 6201 6202 6203 6204 6205
	{
		.procname	= "mc_forwarding",
		.data		= &ipv6_devconf.mc_forwarding,
		.maxlen		= sizeof(int),
		.mode		= 0444,
		.proc_handler	= proc_dointvec,
	},
6206
#endif
6207 6208 6209 6210 6211 6212 6213 6214 6215 6216 6217 6218 6219 6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231 6232 6233 6234 6235 6236 6237 6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248 6249 6250 6251 6252 6253 6254 6255 6256 6257 6258 6259 6260 6261 6262 6263 6264 6265 6266 6267 6268 6269 6270 6271 6272 6273 6274 6275 6276 6277 6278 6279 6280 6281 6282 6283 6284 6285 6286 6287 6288 6289 6290 6291 6292 6293 6294 6295 6296 6297 6298
	{
		.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,

	},
6299 6300 6301 6302 6303 6304 6305
	{
		.procname	= "seg6_enabled",
		.data		= &ipv6_devconf.seg6_enabled,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
6306 6307 6308 6309 6310 6311 6312 6313 6314
#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
6315 6316 6317 6318 6319 6320 6321
	{
		.procname       = "enhanced_dad",
		.data           = &ipv6_devconf.enhanced_dad,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = proc_dointvec,
	},
6322 6323 6324 6325 6326 6327 6328
	{
		.procname		= "addr_gen_mode",
		.data			= &ipv6_devconf.addr_gen_mode,
		.maxlen			= sizeof(int),
		.mode			= 0644,
		.proc_handler	= addrconf_sysctl_addr_gen_mode,
	},
6329 6330 6331 6332 6333 6334 6335
	{
		.procname       = "disable_policy",
		.data           = &ipv6_devconf.disable_policy,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = addrconf_sysctl_disable_policy,
	},
6336 6337 6338
	{
		/* sentinel */
	}
L
Linus Torvalds 已提交
6339 6340
};

6341
static int __addrconf_sysctl_register(struct net *net, char *dev_name,
6342
		struct inet6_dev *idev, struct ipv6_devconf *p)
L
Linus Torvalds 已提交
6343
{
6344
	int i, ifindex;
6345
	struct ctl_table *table;
6346
	char path[sizeof("net/ipv6/conf/") + IFNAMSIZ];
6347

6348 6349
	table = kmemdup(addrconf_sysctl, sizeof(addrconf_sysctl), GFP_KERNEL);
	if (!table)
6350 6351
		goto out;

6352 6353
	for (i = 0; table[i].data; i++) {
		table[i].data += (char *)p - (char *)&ipv6_devconf;
6354 6355 6356 6357 6358 6359 6360 6361
		/* 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
6362 6363
	}

6364
	snprintf(path, sizeof(path), "net/ipv6/conf/%s", dev_name);
L
Linus Torvalds 已提交
6365

6366 6367
	p->sysctl_header = register_net_sysctl(net, path, table);
	if (!p->sysctl_header)
6368
		goto free;
6369

6370 6371 6372 6373 6374 6375
	if (!strcmp(dev_name, "all"))
		ifindex = NETCONFA_IFINDEX_ALL;
	else if (!strcmp(dev_name, "default"))
		ifindex = NETCONFA_IFINDEX_DEFAULT;
	else
		ifindex = idev->dev->ifindex;
6376 6377
	inet6_netconf_notify_devconf(net, RTM_NEWNETCONF, NETCONFA_ALL,
				     ifindex, p);
6378
	return 0;
L
Linus Torvalds 已提交
6379

6380
free:
6381
	kfree(table);
6382
out:
6383
	return -ENOBUFS;
L
Linus Torvalds 已提交
6384 6385
}

6386 6387
static void __addrconf_sysctl_unregister(struct net *net,
					 struct ipv6_devconf *p, int ifindex)
6388
{
6389
	struct ctl_table *table;
6390

6391
	if (!p->sysctl_header)
6392 6393
		return;

6394 6395 6396 6397
	table = p->sysctl_header->ctl_table_arg;
	unregister_net_sysctl_table(p->sysctl_header);
	p->sysctl_header = NULL;
	kfree(table);
6398 6399

	inet6_netconf_notify_devconf(net, RTM_DELNETCONF, 0, ifindex, NULL);
6400 6401
}

6402
static int addrconf_sysctl_register(struct inet6_dev *idev)
6403
{
6404 6405 6406 6407 6408 6409 6410 6411 6412 6413 6414 6415 6416 6417 6418
	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;
6419 6420
}

6421
static void addrconf_sysctl_unregister(struct inet6_dev *idev)
L
Linus Torvalds 已提交
6422
{
6423 6424
	__addrconf_sysctl_unregister(dev_net(idev->dev), &idev->cnf,
				     idev->dev->ifindex);
6425
	neigh_sysctl_unregister(idev->nd_parms);
L
Linus Torvalds 已提交
6426 6427 6428 6429 6430
}


#endif

6431
static int __net_init addrconf_init_net(struct net *net)
6432
{
6433
	int err = -ENOMEM;
6434 6435
	struct ipv6_devconf *all, *dflt;

6436
	all = kmemdup(&ipv6_devconf, sizeof(ipv6_devconf), GFP_KERNEL);
6437
	if (!all)
6438
		goto err_alloc_all;
6439

6440
	dflt = kmemdup(&ipv6_devconf_dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
6441
	if (!dflt)
6442
		goto err_alloc_dflt;
6443

6444 6445 6446
	/* these will be inherited by all namespaces */
	dflt->autoconf = ipv6_defaults.autoconf;
	dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
6447

6448 6449 6450
	dflt->stable_secret.initialized = false;
	all->stable_secret.initialized = false;

6451 6452 6453 6454
	net->ipv6.devconf_all = all;
	net->ipv6.devconf_dflt = dflt;

#ifdef CONFIG_SYSCTL
6455
	err = __addrconf_sysctl_register(net, "all", NULL, all);
6456 6457 6458
	if (err < 0)
		goto err_reg_all;

6459
	err = __addrconf_sysctl_register(net, "default", NULL, dflt);
6460 6461 6462 6463 6464 6465 6466
	if (err < 0)
		goto err_reg_dflt;
#endif
	return 0;

#ifdef CONFIG_SYSCTL
err_reg_dflt:
6467
	__addrconf_sysctl_unregister(net, all, NETCONFA_IFINDEX_ALL);
6468 6469 6470 6471 6472 6473 6474 6475 6476
err_reg_all:
	kfree(dflt);
#endif
err_alloc_dflt:
	kfree(all);
err_alloc_all:
	return err;
}

6477
static void __net_exit addrconf_exit_net(struct net *net)
6478 6479
{
#ifdef CONFIG_SYSCTL
6480 6481 6482 6483
	__addrconf_sysctl_unregister(net, net->ipv6.devconf_dflt,
				     NETCONFA_IFINDEX_DEFAULT);
	__addrconf_sysctl_unregister(net, net->ipv6.devconf_all,
				     NETCONFA_IFINDEX_ALL);
6484
#endif
Z
zhuyj 已提交
6485 6486
	kfree(net->ipv6.devconf_dflt);
	kfree(net->ipv6.devconf_all);
6487 6488 6489 6490 6491 6492 6493
}

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

6494
static struct rtnl_af_ops inet6_ops __read_mostly = {
T
Thomas Graf 已提交
6495 6496 6497
	.family		  = AF_INET6,
	.fill_link_af	  = inet6_fill_link_af,
	.get_link_af_size = inet6_get_link_af_size,
6498
	.validate_link_af = inet6_validate_link_af,
6499
	.set_link_af	  = inet6_set_link_af,
T
Thomas Graf 已提交
6500 6501
};

L
Linus Torvalds 已提交
6502 6503 6504 6505 6506 6507
/*
 *	Init / cleanup code
 */

int __init addrconf_init(void)
{
6508
	struct inet6_dev *idev;
6509
	int i, err;
6510

S
Stephen Hemminger 已提交
6511 6512
	err = ipv6_addr_label_init();
	if (err < 0) {
6513 6514
		pr_crit("%s: cannot initialize default policy table: %d\n",
			__func__, err);
6515
		goto out;
6516
	}
L
Linus Torvalds 已提交
6517

6518 6519 6520
	err = register_pernet_subsys(&addrconf_ops);
	if (err < 0)
		goto out_addrlabel;
6521

6522 6523 6524 6525 6526 6527
	addrconf_wq = create_workqueue("ipv6_addrconf");
	if (!addrconf_wq) {
		err = -ENOMEM;
		goto out_nowq;
	}

L
Linus Torvalds 已提交
6528 6529 6530 6531 6532 6533 6534 6535 6536 6537 6538 6539 6540 6541 6542 6543 6544 6545 6546
	/* 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();
6547
	idev = ipv6_add_dev(init_net.loopback_dev);
L
Linus Torvalds 已提交
6548
	rtnl_unlock();
6549 6550
	if (IS_ERR(idev)) {
		err = PTR_ERR(idev);
6551
		goto errlo;
6552
	}
L
Linus Torvalds 已提交
6553

6554 6555
	ip6_route_init_special_entries();

6556 6557 6558
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		INIT_HLIST_HEAD(&inet6_addr_lst[i]);

L
Linus Torvalds 已提交
6559 6560
	register_netdevice_notifier(&ipv6_dev_notf);

6561
	addrconf_verify();
6562

6563
	rtnl_af_register(&inet6_ops);
T
Thomas Graf 已提交
6564

6565
	err = __rtnl_register(PF_INET6, RTM_GETLINK, NULL, inet6_dump_ifinfo,
6566
			      0);
6567 6568 6569 6570
	if (err < 0)
		goto errout;

	/* Only the first call to __rtnl_register can fail */
6571 6572
	__rtnl_register(PF_INET6, RTM_NEWADDR, inet6_rtm_newaddr, NULL, 0);
	__rtnl_register(PF_INET6, RTM_DELADDR, inet6_rtm_deladdr, NULL, 0);
6573
	__rtnl_register(PF_INET6, RTM_GETADDR, inet6_rtm_getaddr,
6574
			inet6_dump_ifaddr, 0);
6575
	__rtnl_register(PF_INET6, RTM_GETMULTICAST, NULL,
6576
			inet6_dump_ifmcaddr, 0);
6577
	__rtnl_register(PF_INET6, RTM_GETANYCAST, NULL,
6578
			inet6_dump_ifacaddr, 0);
6579
	__rtnl_register(PF_INET6, RTM_GETNETCONF, inet6_netconf_get_devconf,
6580
			inet6_netconf_dump_devconf, 0);
6581

6582 6583
	ipv6_addr_label_rtnl_register();

L
Linus Torvalds 已提交
6584
	return 0;
6585
errout:
T
Thomas Graf 已提交
6586
	rtnl_af_unregister(&inet6_ops);
6587
	unregister_netdevice_notifier(&ipv6_dev_notf);
6588
errlo:
6589 6590
	destroy_workqueue(addrconf_wq);
out_nowq:
6591
	unregister_pernet_subsys(&addrconf_ops);
6592 6593 6594
out_addrlabel:
	ipv6_addr_label_cleanup();
out:
6595
	return err;
L
Linus Torvalds 已提交
6596 6597
}

6598
void addrconf_cleanup(void)
L
Linus Torvalds 已提交
6599
{
6600
	struct net_device *dev;
L
Linus Torvalds 已提交
6601 6602 6603
	int i;

	unregister_netdevice_notifier(&ipv6_dev_notf);
6604
	unregister_pernet_subsys(&addrconf_ops);
6605
	ipv6_addr_label_cleanup();
L
Linus Torvalds 已提交
6606 6607 6608

	rtnl_lock();

T
Thomas Graf 已提交
6609 6610
	__rtnl_af_unregister(&inet6_ops);

6611 6612 6613 6614 6615 6616 6617 6618
	/* 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
6619 6620 6621
	/*
	 *	Check hash table.
	 */
6622
	spin_lock_bh(&addrconf_hash_lock);
6623 6624
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		WARN_ON(!hlist_empty(&inet6_addr_lst[i]));
6625
	spin_unlock_bh(&addrconf_hash_lock);
6626
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
6627
	rtnl_unlock();
6628 6629

	destroy_workqueue(addrconf_wq);
L
Linus Torvalds 已提交
6630
}