addrconf.c 132.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5
/*
 *	IPv6 Address [auto]configuration
 *	Linux INET6 implementation
 *
 *	Authors:
6
 *	Pedro Roque		<roque@di.fc.ul.pt>
L
Linus Torvalds 已提交
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
 *	Alexey Kuznetsov	<kuznet@ms2.inr.ac.ru>
 *
 *	This program is free software; you can redistribute it and/or
 *      modify it under the terms of the GNU General Public License
 *      as published by the Free Software Foundation; either version
 *      2 of the License, or (at your option) any later version.
 */

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

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

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

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

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

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

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

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

#define	INFINITY_LIFE_TIME	0xFFFFFFFF
104 105 106 107 108

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

#ifdef CONFIG_SYSCTL
111
static int addrconf_sysctl_register(struct inet6_dev *idev);
112 113
static void addrconf_sysctl_unregister(struct inet6_dev *idev);
#else
114
static inline int addrconf_sysctl_register(struct inet6_dev *idev)
115
{
116
	return 0;
117 118 119 120 121
}

static inline void addrconf_sysctl_unregister(struct inet6_dev *idev)
{
}
L
Linus Torvalds 已提交
122 123
#endif

S
Sorin Dumitru 已提交
124 125
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
126 127
static void ipv6_regen_rndid(unsigned long data);

128
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev);
L
Linus Torvalds 已提交
129 130 131 132 133
static int ipv6_count_addresses(struct inet6_dev *idev);

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

137 138 139
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
140

141 142
static struct workqueue_struct *addrconf_wq;
static DECLARE_DELAYED_WORK(addr_chk_work, addrconf_verify_work);
L
Linus Torvalds 已提交
143 144 145 146

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

147 148
static void addrconf_type_change(struct net_device *dev,
				 unsigned long event);
L
Linus Torvalds 已提交
149 150
static int addrconf_ifdown(struct net_device *dev, int how);

151 152 153 154 155
static struct rt6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
						  int plen,
						  const struct net_device *dev,
						  u32 flags, u32 noflags);

156
static void addrconf_dad_start(struct inet6_ifaddr *ifp);
157
static void addrconf_dad_work(struct work_struct *w);
L
Linus Torvalds 已提交
158
static void addrconf_dad_completed(struct inet6_ifaddr *ifp);
159
static void addrconf_dad_run(struct inet6_dev *idev);
L
Linus Torvalds 已提交
160 161 162 163
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);

164
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
165
				struct prefix_info *pinfo);
166 167
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev);
L
Linus Torvalds 已提交
168

169
static struct ipv6_devconf ipv6_devconf __read_mostly = {
L
Linus Torvalds 已提交
170 171 172 173 174 175 176
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
	.force_mld_version	= 0,
177 178
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
179 180 181 182
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
183
	.use_tempaddr		= 0,
L
Linus Torvalds 已提交
184 185 186 187 188
	.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,
189
	.accept_ra_defrtr	= 1,
190
	.accept_ra_from_local	= 0,
191
	.accept_ra_pinfo	= 1,
192 193
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
194
	.rtr_probe_interval	= 60 * HZ,
195 196 197
#ifdef CONFIG_IPV6_ROUTE_INFO
	.accept_ra_rt_info_max_plen = 0,
#endif
198
#endif
199
	.proxy_ndp		= 0,
200
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
201
	.disable_ipv6		= 0,
202
	.accept_dad		= 1,
203
	.suppress_frag_ndisc	= 1,
L
Linus Torvalds 已提交
204 205
};

206
static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
L
Linus Torvalds 已提交
207 208 209 210 211 212
	.forwarding		= 0,
	.hop_limit		= IPV6_DEFAULT_HOPLIMIT,
	.mtu6			= IPV6_MIN_MTU,
	.accept_ra		= 1,
	.accept_redirects	= 1,
	.autoconf		= 1,
213 214 215
	.force_mld_version	= 0,
	.mldv1_unsolicited_report_interval = 10 * HZ,
	.mldv2_unsolicited_report_interval = HZ,
L
Linus Torvalds 已提交
216 217 218 219 220 221 222 223 224 225
	.dad_transmits		= 1,
	.rtr_solicits		= MAX_RTR_SOLICITATIONS,
	.rtr_solicit_interval	= RTR_SOLICITATION_INTERVAL,
	.rtr_solicit_delay	= MAX_RTR_SOLICITATION_DELAY,
	.use_tempaddr		= 0,
	.temp_valid_lft		= TEMP_VALID_LIFETIME,
	.temp_prefered_lft	= TEMP_PREFERRED_LIFETIME,
	.regen_max_retry	= REGEN_MAX_RETRY,
	.max_desync_factor	= MAX_DESYNC_FACTOR,
	.max_addresses		= IPV6_MAX_ADDRESSES,
226
	.accept_ra_defrtr	= 1,
227
	.accept_ra_from_local	= 0,
228
	.accept_ra_pinfo	= 1,
229 230
#ifdef CONFIG_IPV6_ROUTER_PREF
	.accept_ra_rtr_pref	= 1,
231
	.rtr_probe_interval	= 60 * HZ,
232 233 234
#ifdef CONFIG_IPV6_ROUTE_INFO
	.accept_ra_rt_info_max_plen = 0,
#endif
235
#endif
236
	.proxy_ndp		= 0,
237
	.accept_source_route	= 0,	/* we do not accept RH0 by default. */
238
	.disable_ipv6		= 0,
239
	.accept_dad		= 1,
240
	.suppress_frag_ndisc	= 1,
L
Linus Torvalds 已提交
241 242
};

243
/* Check if a valid qdisc is available */
244
static inline bool addrconf_qdisc_ok(const struct net_device *dev)
245
{
246
	return !qdisc_tx_is_noop(dev);
247 248
}

249
static void addrconf_del_rs_timer(struct inet6_dev *idev)
L
Linus Torvalds 已提交
250
{
251 252 253 254
	if (del_timer(&idev->rs_timer))
		__in6_dev_put(idev);
}

255
static void addrconf_del_dad_work(struct inet6_ifaddr *ifp)
256
{
257
	if (cancel_delayed_work(&ifp->dad_work))
L
Linus Torvalds 已提交
258 259 260
		__in6_ifa_put(ifp);
}

261 262 263 264 265 266 267
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
268

269 270
static void addrconf_mod_dad_work(struct inet6_ifaddr *ifp,
				   unsigned long delay)
L
Linus Torvalds 已提交
271
{
272
	if (!delayed_work_pending(&ifp->dad_work))
L
Linus Torvalds 已提交
273
		in6_ifa_hold(ifp);
274
	mod_delayed_work(addrconf_wq, &ifp->dad_work, delay);
L
Linus Torvalds 已提交
275 276
}

277 278
static int snmp6_alloc_dev(struct inet6_dev *idev)
{
279 280
	int i;

W
WANG Cong 已提交
281 282
	idev->stats.ipv6 = alloc_percpu(struct ipstats_mib);
	if (!idev->stats.ipv6)
283
		goto err_ip;
284 285 286

	for_each_possible_cpu(i) {
		struct ipstats_mib *addrconf_stats;
W
WANG Cong 已提交
287
		addrconf_stats = per_cpu_ptr(idev->stats.ipv6, i);
288 289 290 291
		u64_stats_init(&addrconf_stats->syncp);
	}


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

	return 0;

303
err_icmpmsg:
304
	kfree(idev->stats.icmpv6dev);
305
err_icmp:
W
WANG Cong 已提交
306
	free_percpu(idev->stats.ipv6);
307
err_ip:
308
	return -ENOMEM;
309 310
}

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

	ASSERT_RTNL();

	if (dev->mtu < IPV6_MIN_MTU)
319
		return ERR_PTR(-EINVAL);
L
Linus Torvalds 已提交
320

321 322
	ndev = kzalloc(sizeof(struct inet6_dev), GFP_KERNEL);
	if (ndev == NULL)
323
		return ERR_PTR(err);
324 325 326

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

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

353
	if (snmp6_register_dev(ndev) < 0) {
354
		ADBG(KERN_WARNING
355
			"%s: cannot create /proc/net/dev_snmp6/%s\n",
356
			__func__, dev->name);
357
		goto err_release;
358 359 360 361 362 363
	}

	/* One reference from device.  We must do this before
	 * we invoke __ipv6_regen_rndid().
	 */
	in6_dev_hold(ndev);
L
Linus Torvalds 已提交
364

365 366 367
	if (dev->flags & (IFF_NOARP | IFF_LOOPBACK))
		ndev->cnf.accept_dad = -1;

A
Amerigo Wang 已提交
368
#if IS_ENABLED(CONFIG_IPV6_SIT)
369
	if (dev->type == ARPHRD_SIT && (dev->priv_flags & IFF_ISATAP)) {
370
		pr_info("%s: Disabled Multicast RS\n", dev->name);
371 372 373 374
		ndev->cnf.rtr_solicits = 0;
	}
#endif

375
	INIT_LIST_HEAD(&ndev->tempaddr_list);
376
	setup_timer(&ndev->regen_timer, ipv6_regen_rndid, (unsigned long)ndev);
377 378
	if ((dev->flags&IFF_LOOPBACK) ||
	    dev->type == ARPHRD_TUNNEL ||
379
	    dev->type == ARPHRD_TUNNEL6 ||
380 381
	    dev->type == ARPHRD_SIT ||
	    dev->type == ARPHRD_NONE) {
382 383 384 385 386
		ndev->cnf.use_tempaddr = -1;
	} else {
		in6_dev_hold(ndev);
		ipv6_regen_rndid((unsigned long) ndev);
	}
387

388
	ndev->token = in6addr_any;
L
Linus Torvalds 已提交
389

390
	if (netif_running(dev) && addrconf_qdisc_ok(dev))
391 392
		ndev->if_flags |= IF_READY;

393 394
	ipv6_mc_init_dev(ndev);
	ndev->tstamp = jiffies;
395 396 397 398 399 400
	err = addrconf_sysctl_register(ndev);
	if (err) {
		ipv6_mc_destroy_dev(ndev);
		del_timer(&ndev->regen_timer);
		goto err_release;
	}
401
	/* protected by rtnl_lock */
402
	rcu_assign_pointer(dev->ip6_ptr, ndev);
403

404 405 406
	/* Join interface-local all-node multicast group */
	ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allnodes);

407
	/* Join all-node multicast group */
408
	ipv6_dev_mc_inc(dev, &in6addr_linklocal_allnodes);
409

410
	/* Join all-router multicast group if forwarding is set */
L
Li Wei 已提交
411
	if (ndev->cnf.forwarding && (dev->flags & IFF_MULTICAST))
412 413
		ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);

L
Linus Torvalds 已提交
414
	return ndev;
415 416 417 418 419 420

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
421 422
}

423
static struct inet6_dev *ipv6_find_idev(struct net_device *dev)
L
Linus Torvalds 已提交
424 425 426 427 428
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
429 430 431
	idev = __in6_dev_get(dev);
	if (!idev) {
		idev = ipv6_add_dev(dev);
432
		if (IS_ERR(idev))
L
Linus Torvalds 已提交
433 434
			return NULL;
	}
435

L
Linus Torvalds 已提交
436 437 438 439 440
	if (dev->flags&IFF_UP)
		ipv6_mc_up(idev);
	return idev;
}

441 442 443 444 445
static int inet6_netconf_msgsize_devconf(int type)
{
	int size =  NLMSG_ALIGN(sizeof(struct netconfmsg))
		    + nla_total_size(4);	/* NETCONFA_IFINDEX */

446 447
	/* type -1 is used for ALL */
	if (type == -1 || type == NETCONFA_FORWARDING)
448
		size += nla_total_size(4);
449
#ifdef CONFIG_IPV6_MROUTE
450 451
	if (type == -1 || type == NETCONFA_MC_FORWARDING)
		size += nla_total_size(4);
452
#endif
453 454
	if (type == -1 || type == NETCONFA_PROXY_NEIGH)
		size += nla_total_size(4);
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477

	return size;
}

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

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

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

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

478 479
	/* type -1 is used for ALL */
	if ((type == -1 || type == NETCONFA_FORWARDING) &&
480 481
	    nla_put_s32(skb, NETCONFA_FORWARDING, devconf->forwarding) < 0)
		goto nla_put_failure;
482
#ifdef CONFIG_IPV6_MROUTE
483 484 485 486
	if ((type == -1 || type == NETCONFA_MC_FORWARDING) &&
	    nla_put_s32(skb, NETCONFA_MC_FORWARDING,
			devconf->mc_forwarding) < 0)
		goto nla_put_failure;
487
#endif
488 489 490 491
	if ((type == -1 || type == NETCONFA_PROXY_NEIGH) &&
	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
		goto nla_put_failure;

492 493 494 495 496 497 498
	return nlmsg_end(skb, nlh);

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

499 500
void inet6_netconf_notify_devconf(struct net *net, int type, int ifindex,
				  struct ipv6_devconf *devconf)
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519
{
	struct sk_buff *skb;
	int err = -ENOBUFS;

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

	err = inet6_netconf_fill_devconf(skb, ifindex, devconf, 0, 0,
					 RTM_NEWNETCONF, 0, type);
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_NETCONF, NULL, GFP_ATOMIC);
	return;
errout:
520
	rtnl_set_sk_err(net, RTNLGRP_IPV6_NETCONF, err);
521 522
}

523 524 525
static const struct nla_policy devconf_ipv6_policy[NETCONFA_MAX+1] = {
	[NETCONFA_IFINDEX]	= { .len = sizeof(int) },
	[NETCONFA_FORWARDING]	= { .len = sizeof(int) },
526
	[NETCONFA_PROXY_NEIGH]	= { .len = sizeof(int) },
527 528 529
};

static int inet6_netconf_get_devconf(struct sk_buff *in_skb,
530
				     struct nlmsghdr *nlh)
531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589
{
	struct net *net = sock_net(in_skb->sk);
	struct nlattr *tb[NETCONFA_MAX+1];
	struct netconfmsg *ncm;
	struct sk_buff *skb;
	struct ipv6_devconf *devconf;
	struct inet6_dev *in6_dev;
	struct net_device *dev;
	int ifindex;
	int err;

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

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

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

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

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

590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606
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();
607 608
		cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^
			  net->dev_base_seq;
609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;

			if (inet6_netconf_fill_devconf(skb, dev->ifindex,
						       &idev->cnf,
						       NETLINK_CB(cb->skb).portid,
						       cb->nlh->nlmsg_seq,
						       RTM_NEWNETCONF,
						       NLM_F_MULTI,
						       -1) <= 0) {
				rcu_read_unlock();
				goto done;
			}
626
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
cont:
			idx++;
		}
		rcu_read_unlock();
	}
	if (h == NETDEV_HASHENTRIES) {
		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_ALL,
					       net->ipv6.devconf_all,
					       NETLINK_CB(cb->skb).portid,
					       cb->nlh->nlmsg_seq,
					       RTM_NEWNETCONF, NLM_F_MULTI,
					       -1) <= 0)
			goto done;
		else
			h++;
	}
	if (h == NETDEV_HASHENTRIES + 1) {
		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_DEFAULT,
					       net->ipv6.devconf_dflt,
					       NETLINK_CB(cb->skb).portid,
					       cb->nlh->nlmsg_seq,
					       RTM_NEWNETCONF, NLM_F_MULTI,
					       -1) <= 0)
			goto done;
		else
			h++;
	}
done:
	cb->args[0] = h;
	cb->args[1] = idx;

	return skb->len;
}

L
Linus Torvalds 已提交
661 662 663 664 665 666 667 668 669
#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;
670 671
	if (idev->cnf.forwarding)
		dev_disable_lro(dev);
A
Amerigo Wang 已提交
672
	if (dev->flags & IFF_MULTICAST) {
673
		if (idev->cnf.forwarding) {
674
			ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);
675 676 677
			ipv6_dev_mc_inc(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_inc(dev, &in6addr_sitelocal_allrouters);
		} else {
678
			ipv6_dev_mc_dec(dev, &in6addr_linklocal_allrouters);
679 680 681
			ipv6_dev_mc_dec(dev, &in6addr_interfacelocal_allrouters);
			ipv6_dev_mc_dec(dev, &in6addr_sitelocal_allrouters);
		}
L
Linus Torvalds 已提交
682
	}
683 684

	list_for_each_entry(ifa, &idev->addr_list, if_list) {
M
Michal Wrobel 已提交
685 686
		if (ifa->flags&IFA_F_TENTATIVE)
			continue;
L
Linus Torvalds 已提交
687 688 689 690 691
		if (idev->cnf.forwarding)
			addrconf_join_anycast(ifa);
		else
			addrconf_leave_anycast(ifa);
	}
692 693
	inet6_netconf_notify_devconf(dev_net(dev), NETCONFA_FORWARDING,
				     dev->ifindex, &idev->cnf);
L
Linus Torvalds 已提交
694 695 696
}


697
static void addrconf_forward_change(struct net *net, __s32 newf)
L
Linus Torvalds 已提交
698 699 700 701
{
	struct net_device *dev;
	struct inet6_dev *idev;

702
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
703 704
		idev = __in6_dev_get(dev);
		if (idev) {
705 706
			int changed = (!idev->cnf.forwarding) ^ (!newf);
			idev->cnf.forwarding = newf;
L
Linus Torvalds 已提交
707 708 709 710 711
			if (changed)
				dev_forward_change(idev);
		}
	}
}
712

713
static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
714
{
715
	struct net *net;
716 717 718 719
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
720 721

	net = (struct net *)table->extra2;
722 723
	old = *p;
	*p = newf;
724

725
	if (p == &net->ipv6.devconf_dflt->forwarding) {
726 727 728 729
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net, NETCONFA_FORWARDING,
						     NETCONFA_IFINDEX_DEFAULT,
						     net->ipv6.devconf_dflt);
730 731
		rtnl_unlock();
		return 0;
E
Eric W. Biederman 已提交
732
	}
733

734 735 736
	if (p == &net->ipv6.devconf_all->forwarding) {
		net->ipv6.devconf_dflt->forwarding = newf;
		addrconf_forward_change(net, newf);
737 738 739 740
		if ((!newf) ^ (!old))
			inet6_netconf_notify_devconf(net, NETCONFA_FORWARDING,
						     NETCONFA_IFINDEX_ALL,
						     net->ipv6.devconf_all);
741
	} else if ((!newf) ^ (!old))
742
		dev_forward_change((struct inet6_dev *)table->extra1);
743
	rtnl_unlock();
744

745
	if (newf)
746
		rt6_purge_dflt_routers(net);
747
	return 1;
748
}
L
Linus Torvalds 已提交
749 750
#endif

751
/* Nobody refers to this ifaddr, destroy it */
L
Linus Torvalds 已提交
752 753
void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp)
{
754
	WARN_ON(!hlist_unhashed(&ifp->addr_lst));
755

L
Linus Torvalds 已提交
756
#ifdef NET_REFCNT_DEBUG
757
	pr_debug("%s\n", __func__);
L
Linus Torvalds 已提交
758 759 760 761
#endif

	in6_dev_put(ifp->idev);

762 763 764
	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
765

766
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
767
		pr_warn("Freeing alive inet6 address %p\n", ifp);
L
Linus Torvalds 已提交
768 769
		return;
	}
A
Amerigo Wang 已提交
770
	ip6_rt_put(ifp->rt);
L
Linus Torvalds 已提交
771

772
	kfree_rcu(ifp, rcu);
L
Linus Torvalds 已提交
773 774
}

B
Brian Haley 已提交
775 776 777
static void
ipv6_link_dev_addr(struct inet6_dev *idev, struct inet6_ifaddr *ifp)
{
778
	struct list_head *p;
779
	int ifp_scope = ipv6_addr_src_scope(&ifp->addr);
B
Brian Haley 已提交
780 781 782 783 784

	/*
	 * Each device address list is sorted in order of scope -
	 * global before linklocal.
	 */
785 786 787
	list_for_each(p, &idev->addr_list) {
		struct inet6_ifaddr *ifa
			= list_entry(p, struct inet6_ifaddr, if_list);
788
		if (ifp_scope >= ipv6_addr_src_scope(&ifa->addr))
B
Brian Haley 已提交
789 790 791
			break;
	}

792
	list_add_tail(&ifp->if_list, p);
B
Brian Haley 已提交
793 794
}

E
Eric Dumazet 已提交
795
static u32 inet6_addr_hash(const struct in6_addr *addr)
796
{
E
Eric Dumazet 已提交
797
	return hash_32(ipv6_addr_hash(addr), IN6_ADDR_HSIZE_SHIFT);
798 799
}

L
Linus Torvalds 已提交
800 801 802
/* On success it returns ifp with increased reference count */

static struct inet6_ifaddr *
803 804
ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
	      const struct in6_addr *peer_addr, int pfxlen,
805
	      int scope, u32 flags, u32 valid_lft, u32 prefered_lft)
L
Linus Torvalds 已提交
806 807 808
{
	struct inet6_ifaddr *ifa = NULL;
	struct rt6_info *rt;
809
	unsigned int hash;
L
Linus Torvalds 已提交
810
	int err = 0;
811 812 813 814 815 816 817
	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
818

819
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
820 821 822 823 824
	if (idev->dead) {
		err = -ENODEV;			/*XXX*/
		goto out2;
	}

825
	if (idev->cnf.disable_ipv6) {
826 827 828 829
		err = -EACCES;
		goto out2;
	}

830
	spin_lock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
831 832

	/* Ignore adding duplicate addresses on an interface */
833
	if (ipv6_chk_same_addr(dev_net(idev->dev), addr, idev->dev)) {
834
		ADBG("ipv6_add_addr: already assigned\n");
L
Linus Torvalds 已提交
835 836 837 838
		err = -EEXIST;
		goto out;
	}

839
	ifa = kzalloc(sizeof(struct inet6_ifaddr), GFP_ATOMIC);
L
Linus Torvalds 已提交
840 841

	if (ifa == NULL) {
842
		ADBG("ipv6_add_addr: malloc failed\n");
L
Linus Torvalds 已提交
843 844 845 846
		err = -ENOBUFS;
		goto out;
	}

847
	rt = addrconf_dst_alloc(idev, addr, false);
L
Linus Torvalds 已提交
848 849 850 851 852
	if (IS_ERR(rt)) {
		err = PTR_ERR(rt);
		goto out;
	}

853 854
	neigh_parms_data_state_setall(idev->nd_parms);

A
Alexey Dobriyan 已提交
855
	ifa->addr = *addr;
856 857
	if (peer_addr)
		ifa->peer_addr = *peer_addr;
L
Linus Torvalds 已提交
858 859

	spin_lock_init(&ifa->lock);
860
	spin_lock_init(&ifa->state_lock);
861
	INIT_DELAYED_WORK(&ifa->dad_work, addrconf_dad_work);
862
	INIT_HLIST_NODE(&ifa->addr_lst);
L
Linus Torvalds 已提交
863 864 865
	ifa->scope = scope;
	ifa->prefix_len = pfxlen;
	ifa->flags = flags | IFA_F_TENTATIVE;
866 867
	ifa->valid_lft = valid_lft;
	ifa->prefered_lft = prefered_lft;
L
Linus Torvalds 已提交
868
	ifa->cstamp = ifa->tstamp = jiffies;
869
	ifa->tokenized = false;
L
Linus Torvalds 已提交
870

871 872
	ifa->rt = rt;

L
Linus Torvalds 已提交
873 874 875 876 877 878
	ifa->idev = idev;
	in6_dev_hold(idev);
	/* For caller */
	in6_ifa_hold(ifa);

	/* Add to big hash table */
E
Eric Dumazet 已提交
879
	hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
880

881 882
	hlist_add_head_rcu(&ifa->addr_lst, &inet6_addr_lst[hash]);
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
883 884 885

	write_lock(&idev->lock);
	/* Add to inet6_dev unicast addr list. */
B
Brian Haley 已提交
886
	ipv6_link_dev_addr(idev, ifa);
L
Linus Torvalds 已提交
887 888

	if (ifa->flags&IFA_F_TEMPORARY) {
889
		list_add(&ifa->tmp_list, &idev->tempaddr_list);
L
Linus Torvalds 已提交
890 891 892 893 894 895
		in6_ifa_hold(ifa);
	}

	in6_ifa_hold(ifa);
	write_unlock(&idev->lock);
out2:
896
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
897

898
	if (likely(err == 0))
899
		inet6addr_notifier_call_chain(NETDEV_UP, ifa);
L
Linus Torvalds 已提交
900 901 902 903 904 905 906
	else {
		kfree(ifa);
		ifa = ERR_PTR(err);
	}

	return ifa;
out:
907
	spin_unlock(&addrconf_hash_lock);
L
Linus Torvalds 已提交
908 909 910
	goto out2;
}

911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992
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
993 994 995 996
/* This function wants to get referenced ifp and releases it before return */

static void ipv6_del_addr(struct inet6_ifaddr *ifp)
{
997
	int state;
998 999
	enum cleanup_prefix_rt_t action = CLEANUP_PREFIX_RT_NOP;
	unsigned long expires;
L
Linus Torvalds 已提交
1000

1001 1002
	ASSERT_RTNL();

1003 1004
	spin_lock_bh(&ifp->state_lock);
	state = ifp->state;
1005
	ifp->state = INET6_IFADDR_STATE_DEAD;
1006 1007 1008 1009
	spin_unlock_bh(&ifp->state_lock);

	if (state == INET6_IFADDR_STATE_DEAD)
		goto out;
L
Linus Torvalds 已提交
1010

1011 1012 1013
	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
1014

1015
	write_lock_bh(&ifp->idev->lock);
1016

L
Linus Torvalds 已提交
1017
	if (ifp->flags&IFA_F_TEMPORARY) {
1018 1019 1020 1021
		list_del(&ifp->tmp_list);
		if (ifp->ifpub) {
			in6_ifa_put(ifp->ifpub);
			ifp->ifpub = NULL;
L
Linus Torvalds 已提交
1022
		}
1023
		__in6_ifa_put(ifp);
L
Linus Torvalds 已提交
1024 1025
	}

1026 1027
	if (ifp->flags & IFA_F_PERMANENT && !(ifp->flags & IFA_F_NOPREFIXROUTE))
		action = check_cleanup_prefix_route(ifp, &expires);
1028

1029 1030 1031 1032
	list_del_init(&ifp->if_list);
	__in6_ifa_put(ifp);

	write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
1033

1034
	addrconf_del_dad_work(ifp);
1035

L
Linus Torvalds 已提交
1036 1037
	ipv6_ifa_notify(RTM_DELADDR, ifp);

1038
	inet6addr_notifier_call_chain(NETDEV_DOWN, ifp);
L
Linus Torvalds 已提交
1039

1040 1041 1042
	if (action != CLEANUP_PREFIX_RT_NOP) {
		cleanup_prefix_route(ifp, expires,
			action == CLEANUP_PREFIX_RT_DEL);
L
Linus Torvalds 已提交
1043 1044
	}

1045 1046
	/* clean up prefsrc entries */
	rt6_remove_prefsrc(ifp);
1047
out:
L
Linus Torvalds 已提交
1048 1049 1050 1051 1052 1053 1054
	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;
1055
	unsigned long tmp_prefered_lft, tmp_valid_lft, tmp_tstamp, age;
1056
	unsigned long regen_advance;
L
Linus Torvalds 已提交
1057 1058
	int tmp_plen;
	int ret = 0;
1059
	u32 addr_flags;
1060
	unsigned long now = jiffies;
L
Linus Torvalds 已提交
1061

1062
	write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073
	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) {
1074
		write_unlock_bh(&idev->lock);
1075
		pr_info("%s: use_tempaddr is disabled\n", __func__);
L
Linus Torvalds 已提交
1076 1077 1078 1079 1080 1081 1082 1083
		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);
1084
		write_unlock_bh(&idev->lock);
1085 1086
		pr_warn("%s: regeneration time exceeded - disabled temporary address support\n",
			__func__);
L
Linus Torvalds 已提交
1087 1088 1089 1090 1091 1092
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}
	in6_ifa_hold(ifp);
	memcpy(addr.s6_addr, ifp->addr.s6_addr, 8);
S
Sorin Dumitru 已提交
1093
	__ipv6_try_regen_rndid(idev, tmpaddr);
L
Linus Torvalds 已提交
1094
	memcpy(&addr.s6_addr[8], idev->rndid, 8);
1095
	age = (now - ifp->tstamp) / HZ;
L
Linus Torvalds 已提交
1096 1097
	tmp_valid_lft = min_t(__u32,
			      ifp->valid_lft,
1098
			      idev->cnf.temp_valid_lft + age);
1099 1100
	tmp_prefered_lft = min_t(__u32,
				 ifp->prefered_lft,
1101
				 idev->cnf.temp_prefered_lft + age -
1102
				 idev->cnf.max_desync_factor);
L
Linus Torvalds 已提交
1103 1104 1105 1106
	tmp_plen = ifp->prefix_len;
	tmp_tstamp = ifp->tstamp;
	spin_unlock_bh(&ifp->lock);

1107
	regen_advance = idev->cnf.regen_max_retry *
1108 1109
			idev->cnf.dad_transmits *
			NEIGH_VAR(idev->nd_parms, RETRANS_TIME) / HZ;
1110
	write_unlock_bh(&idev->lock);
1111

1112 1113 1114 1115
	/* 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.
1116 1117
	 * Use age calculation as in addrconf_verify to avoid unnecessary
	 * temporary addresses being generated.
1118
	 */
1119 1120
	age = (now - tmp_tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
	if (tmp_prefered_lft <= regen_advance + age) {
1121 1122 1123 1124 1125 1126
		in6_ifa_put(ifp);
		in6_dev_put(idev);
		ret = -1;
		goto out;
	}

1127 1128 1129 1130 1131
	addr_flags = IFA_F_TEMPORARY;
	/* set in addrconf_prefix_rcv() */
	if (ifp->flags & IFA_F_OPTIMISTIC)
		addr_flags |= IFA_F_OPTIMISTIC;

1132 1133 1134 1135
	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
1136 1137
		in6_ifa_put(ifp);
		in6_dev_put(idev);
1138
		pr_info("%s: retry temporary address regeneration\n", __func__);
L
Linus Torvalds 已提交
1139
		tmpaddr = &addr;
1140
		write_lock_bh(&idev->lock);
L
Linus Torvalds 已提交
1141 1142 1143 1144 1145
		goto retry;
	}

	spin_lock_bh(&ift->lock);
	ift->ifpub = ifp;
1146
	ift->cstamp = now;
L
Linus Torvalds 已提交
1147 1148 1149
	ift->tstamp = tmp_tstamp;
	spin_unlock_bh(&ift->lock);

1150
	addrconf_dad_start(ift);
L
Linus Torvalds 已提交
1151 1152 1153 1154 1155 1156 1157
	in6_ifa_put(ift);
	in6_dev_put(idev);
out:
	return ret;
}

/*
1158
 *	Choose an appropriate source address (RFC3484)
L
Linus Torvalds 已提交
1159
 */
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172
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,
1173 1174 1175
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	IPV6_SADDR_RULE_NOT_OPTIMISTIC,
#endif
1176 1177 1178
	IPV6_SADDR_RULE_MAX
};

1179
struct ipv6_saddr_score {
1180 1181 1182 1183 1184 1185
	int			rule;
	int			addr_type;
	struct inet6_ifaddr	*ifa;
	DECLARE_BITMAP(scorebits, IPV6_SADDR_RULE_MAX);
	int			scopedist;
	int			matchlen;
1186 1187
};

1188
struct ipv6_saddr_dst {
1189
	const struct in6_addr *addr;
1190 1191 1192
	int ifindex;
	int scope;
	int label;
1193
	unsigned int prefs;
1194
};
1195

D
Dave Jones 已提交
1196
static inline int ipv6_saddr_preferred(int type)
L
Linus Torvalds 已提交
1197
{
U
Ulrich Weber 已提交
1198
	if (type & (IPV6_ADDR_MAPPED|IPV6_ADDR_COMPATv4|IPV6_ADDR_LOOPBACK))
1199 1200
		return 1;
	return 0;
L
Linus Torvalds 已提交
1201 1202
}

1203 1204 1205 1206 1207 1208 1209 1210 1211
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
}

1212 1213
static int ipv6_get_saddr_eval(struct net *net,
			       struct ipv6_saddr_score *score,
1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252
			       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
1253
		 *  B-15 |    \        if scope is enough for destination.
1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271
		 *       |             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:
1272
	    {
1273
		/* Rule 3: Avoid deprecated and optimistic addresses */
1274 1275 1276 1277
		u8 avoid = IFA_F_DEPRECATED;

		if (!ipv6_use_optimistic_addr(score->ifa->idev))
			avoid |= IFA_F_OPTIMISTIC;
1278
		ret = ipv6_saddr_preferred(score->addr_type) ||
1279
		      !(score->ifa->flags & avoid);
1280
		break;
1281
	    }
1282 1283
#ifdef CONFIG_IPV6_MIP6
	case IPV6_SADDR_RULE_HOA:
1284
	    {
1285
		/* Rule 4: Prefer home address */
1286 1287
		int prefhome = !(dst->prefs & IPV6_PREFER_SRC_COA);
		ret = !(score->ifa->flags & IFA_F_HOMEADDRESS) ^ prefhome;
1288
		break;
1289
	    }
1290 1291 1292 1293 1294 1295 1296 1297
#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 */
1298 1299
		ret = ipv6_addr_label(net,
				      &score->ifa->addr, score->addr_type,
1300 1301 1302
				      score->ifa->idev->dev->ifindex) == dst->label;
		break;
	case IPV6_SADDR_RULE_PRIVACY:
1303
	    {
1304
		/* Rule 7: Prefer public address
L
Lucas De Marchi 已提交
1305
		 * Note: prefer temporary address if use_tempaddr >= 2
1306
		 */
1307 1308 1309 1310
		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;
1311
		break;
1312
	    }
1313 1314 1315 1316 1317 1318 1319 1320 1321
	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 */
1322 1323 1324 1325
		ret = ipv6_addr_diff(&score->ifa->addr, dst->addr);
		if (ret > score->ifa->prefix_len)
			ret = score->ifa->prefix_len;
		score->matchlen = ret;
1326
		break;
1327 1328 1329 1330 1331 1332 1333 1334
#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
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345
	default:
		ret = 0;
	}

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

1346
int ipv6_dev_get_saddr(struct net *net, const struct net_device *dst_dev,
1347
		       const struct in6_addr *daddr, unsigned int prefs,
1348
		       struct in6_addr *saddr)
L
Linus Torvalds 已提交
1349
{
1350 1351 1352
	struct ipv6_saddr_score scores[2],
				*score = &scores[0], *hiscore = &scores[1];
	struct ipv6_saddr_dst dst;
1353
	struct net_device *dev;
1354
	int dst_type;
L
Linus Torvalds 已提交
1355

1356 1357 1358 1359
	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);
1360
	dst.label = ipv6_addr_label(net, daddr, dst_type, dst.ifindex);
1361
	dst.prefs = prefs;
1362 1363 1364

	hiscore->rule = -1;
	hiscore->ifa = NULL;
L
Linus Torvalds 已提交
1365

1366
	rcu_read_lock();
L
Linus Torvalds 已提交
1367

1368
	for_each_netdev_rcu(net, dev) {
1369 1370
		struct inet6_dev *idev;

1371
		/* Candidate Source Address (section 4)
1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382
		 *  - 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.)
		 */
1383 1384 1385
		if (((dst_type & IPV6_ADDR_MULTICAST) ||
		     dst.scope <= IPV6_ADDR_SCOPE_LINKLOCAL) &&
		    dst.ifindex && dev->ifindex != dst.ifindex)
1386
			continue;
L
Linus Torvalds 已提交
1387 1388

		idev = __in6_dev_get(dev);
1389 1390 1391 1392
		if (!idev)
			continue;

		read_lock_bh(&idev->lock);
1393
		list_for_each_entry(score->ifa, &idev->addr_list, if_list) {
1394
			int i;
1395

1396
			/*
1397 1398 1399
			 * - Tentative Address (RFC2462 section 5.4)
			 *  - A tentative address is not considered
			 *    "assigned to an interface" in the traditional
1400
			 *    sense, unless it is also flagged as optimistic.
1401
			 * - Candidate Source Address (section 4)
1402 1403 1404 1405
			 *  - In any case, anycast addresses, multicast
			 *    addresses, and the unspecified address MUST
			 *    NOT be included in a candidate set.
			 */
1406 1407
			if ((score->ifa->flags & IFA_F_TENTATIVE) &&
			    (!(score->ifa->flags & IFA_F_OPTIMISTIC)))
1408
				continue;
1409 1410 1411 1412 1413

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

			if (unlikely(score->addr_type == IPV6_ADDR_ANY ||
				     score->addr_type & IPV6_ADDR_MULTICAST)) {
1414
				LIMIT_NETDEBUG(KERN_DEBUG
J
Joe Perches 已提交
1415
					       "ADDRCONF: unspecified / multicast address "
1416 1417 1418 1419 1420
					       "assigned as unicast address on %s",
					       dev->name);
				continue;
			}

1421 1422 1423 1424 1425 1426
			score->rule = -1;
			bitmap_zero(score->scorebits, IPV6_SADDR_RULE_MAX);

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

1427 1428
				minihiscore = ipv6_get_saddr_eval(net, hiscore, &dst, i);
				miniscore = ipv6_get_saddr_eval(net, score, &dst, i);
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449

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

					in6_ifa_hold(score->ifa);

1450
					swap(hiscore, score);
1451 1452 1453 1454 1455

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

					break;
L
Linus Torvalds 已提交
1456
				}
1457
			}
L
Linus Torvalds 已提交
1458
		}
1459
try_nextdev:
1460
		read_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
1461
	}
1462
	rcu_read_unlock();
L
Linus Torvalds 已提交
1463

1464
	if (!hiscore->ifa)
1465
		return -EADDRNOTAVAIL;
1466

A
Alexey Dobriyan 已提交
1467
	*saddr = hiscore->ifa->addr;
1468
	in6_ifa_put(hiscore->ifa);
1469
	return 0;
L
Linus Torvalds 已提交
1470
}
1471
EXPORT_SYMBOL(ipv6_dev_get_saddr);
L
Linus Torvalds 已提交
1472

1473
int __ipv6_get_lladdr(struct inet6_dev *idev, struct in6_addr *addr,
1474
		      u32 banned_flags)
1475 1476 1477 1478
{
	struct inet6_ifaddr *ifp;
	int err = -EADDRNOTAVAIL;

1479 1480 1481
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
1482 1483 1484 1485 1486 1487 1488 1489 1490 1491
		if (ifp->scope == IFA_LINK &&
		    !(ifp->flags & banned_flags)) {
			*addr = ifp->addr;
			err = 0;
			break;
		}
	}
	return err;
}

1492
int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr,
1493
		    u32 banned_flags)
L
Linus Torvalds 已提交
1494 1495 1496 1497
{
	struct inet6_dev *idev;
	int err = -EADDRNOTAVAIL;

1498
	rcu_read_lock();
S
Stephen Hemminger 已提交
1499 1500
	idev = __in6_dev_get(dev);
	if (idev) {
L
Linus Torvalds 已提交
1501
		read_lock_bh(&idev->lock);
1502
		err = __ipv6_get_lladdr(idev, addr, banned_flags);
L
Linus Torvalds 已提交
1503 1504
		read_unlock_bh(&idev->lock);
	}
1505
	rcu_read_unlock();
L
Linus Torvalds 已提交
1506 1507 1508 1509 1510 1511 1512 1513 1514
	return err;
}

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

	read_lock_bh(&idev->lock);
1515
	list_for_each_entry(ifp, &idev->addr_list, if_list)
L
Linus Torvalds 已提交
1516 1517 1518 1519 1520
		cnt++;
	read_unlock_bh(&idev->lock);
	return cnt;
}

1521
int ipv6_chk_addr(struct net *net, const struct in6_addr *addr,
1522
		  const struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1523
{
1524
	struct inet6_ifaddr *ifp;
E
Eric Dumazet 已提交
1525
	unsigned int hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
1526

1527
	rcu_read_lock_bh();
1528
	hlist_for_each_entry_rcu(ifp, &inet6_addr_lst[hash], addr_lst) {
1529
		if (!net_eq(dev_net(ifp->idev->dev), net))
1530
			continue;
L
Linus Torvalds 已提交
1531
		if (ipv6_addr_equal(&ifp->addr, addr) &&
1532 1533 1534 1535 1536
		    !(ifp->flags&IFA_F_TENTATIVE) &&
		    (dev == NULL || ifp->idev->dev == dev ||
		     !(ifp->scope&(IFA_LINK|IFA_HOST) || strict))) {
			rcu_read_unlock_bh();
			return 1;
L
Linus Torvalds 已提交
1537 1538
		}
	}
1539

1540 1541
	rcu_read_unlock_bh();
	return 0;
L
Linus Torvalds 已提交
1542
}
1543 1544
EXPORT_SYMBOL(ipv6_chk_addr);

1545 1546
static bool ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
			       struct net_device *dev)
L
Linus Torvalds 已提交
1547
{
E
Eric Dumazet 已提交
1548
	unsigned int hash = inet6_addr_hash(addr);
1549
	struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
1550

1551
	hlist_for_each_entry(ifp, &inet6_addr_lst[hash], addr_lst) {
1552
		if (!net_eq(dev_net(ifp->idev->dev), net))
1553
			continue;
L
Linus Torvalds 已提交
1554 1555
		if (ipv6_addr_equal(&ifp->addr, addr)) {
			if (dev == NULL || ifp->idev->dev == dev)
1556
				return true;
L
Linus Torvalds 已提交
1557 1558
		}
	}
1559
	return false;
L
Linus Torvalds 已提交
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
/* 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);

1589
int ipv6_chk_prefix(const struct in6_addr *addr, struct net_device *dev)
1590 1591 1592 1593 1594 1595 1596 1597 1598 1599
{
	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);
1600
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612
			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);

1613
struct inet6_ifaddr *ipv6_get_ifaddr(struct net *net, const struct in6_addr *addr,
1614
				     struct net_device *dev, int strict)
L
Linus Torvalds 已提交
1615
{
1616
	struct inet6_ifaddr *ifp, *result = NULL;
E
Eric Dumazet 已提交
1617
	unsigned int hash = inet6_addr_hash(addr);
L
Linus Torvalds 已提交
1618

1619
	rcu_read_lock_bh();
1620
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
1621
		if (!net_eq(dev_net(ifp->idev->dev), net))
1622
			continue;
L
Linus Torvalds 已提交
1623 1624 1625
		if (ipv6_addr_equal(&ifp->addr, addr)) {
			if (dev == NULL || ifp->idev->dev == dev ||
			    !(ifp->scope&(IFA_LINK|IFA_HOST) || strict)) {
1626
				result = ifp;
L
Linus Torvalds 已提交
1627 1628 1629 1630 1631
				in6_ifa_hold(ifp);
				break;
			}
		}
	}
1632
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1633

1634
	return result;
L
Linus Torvalds 已提交
1635 1636 1637 1638
}

/* Gets referenced address, destroys ifaddr */

B
Brian Haley 已提交
1639
static void addrconf_dad_stop(struct inet6_ifaddr *ifp, int dad_failed)
L
Linus Torvalds 已提交
1640 1641 1642
{
	if (ifp->flags&IFA_F_PERMANENT) {
		spin_lock_bh(&ifp->lock);
1643
		addrconf_del_dad_work(ifp);
L
Linus Torvalds 已提交
1644
		ifp->flags |= IFA_F_TENTATIVE;
B
Brian Haley 已提交
1645 1646
		if (dad_failed)
			ifp->flags |= IFA_F_DADFAILED;
L
Linus Torvalds 已提交
1647
		spin_unlock_bh(&ifp->lock);
1648 1649
		if (dad_failed)
			ipv6_ifa_notify(0, ifp);
L
Linus Torvalds 已提交
1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663
		in6_ifa_put(ifp);
	} else if (ifp->flags&IFA_F_TEMPORARY) {
		struct inet6_ifaddr *ifpub;
		spin_lock_bh(&ifp->lock);
		ifpub = ifp->ifpub;
		if (ifpub) {
			in6_ifa_hold(ifpub);
			spin_unlock_bh(&ifp->lock);
			ipv6_create_tempaddr(ifpub, ifp);
			in6_ifa_put(ifpub);
		} else {
			spin_unlock_bh(&ifp->lock);
		}
		ipv6_del_addr(ifp);
1664
	} else {
L
Linus Torvalds 已提交
1665
		ipv6_del_addr(ifp);
1666
	}
L
Linus Torvalds 已提交
1667 1668
}

H
Herbert Xu 已提交
1669 1670 1671 1672
static int addrconf_dad_end(struct inet6_ifaddr *ifp)
{
	int err = -ENOENT;

1673
	spin_lock_bh(&ifp->state_lock);
H
Herbert Xu 已提交
1674 1675 1676 1677
	if (ifp->state == INET6_IFADDR_STATE_DAD) {
		ifp->state = INET6_IFADDR_STATE_POSTDAD;
		err = 0;
	}
1678
	spin_unlock_bh(&ifp->state_lock);
H
Herbert Xu 已提交
1679 1680 1681 1682

	return err;
}

1683 1684
void addrconf_dad_failure(struct inet6_ifaddr *ifp)
{
1685
	struct inet6_dev *idev = ifp->idev;
1686

1687 1688
	if (addrconf_dad_end(ifp)) {
		in6_ifa_put(ifp);
H
Herbert Xu 已提交
1689
		return;
1690
	}
H
Herbert Xu 已提交
1691

1692 1693
	net_info_ratelimited("%s: IPv6 duplicate address %pI6c detected!\n",
			     ifp->idev->dev->name, &ifp->addr);
1694

1695 1696 1697 1698 1699 1700 1701 1702 1703 1704
	if (idev->cnf.accept_dad > 1 && !idev->cnf.disable_ipv6) {
		struct in6_addr addr;

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

		if (!ipv6_generate_eui64(addr.s6_addr + 8, idev->dev) &&
		    ipv6_addr_equal(&ifp->addr, &addr)) {
			/* DAD failed for link-local based on MAC address */
			idev->cnf.disable_ipv6 = 1;
1705

1706
			pr_info("%s: IPv6 being disabled!\n",
1707
				ifp->idev->dev->name);
1708 1709 1710
		}
	}

1711 1712 1713 1714 1715 1716
	spin_lock_bh(&ifp->state_lock);
	/* transition from _POSTDAD to _ERRDAD */
	ifp->state = INET6_IFADDR_STATE_ERRDAD;
	spin_unlock_bh(&ifp->state_lock);

	addrconf_mod_dad_work(ifp, 0);
1717
}
L
Linus Torvalds 已提交
1718

1719 1720
/* Join to solicited addr multicast group.
 * caller must hold RTNL */
1721
void addrconf_join_solict(struct net_device *dev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
1722 1723 1724 1725 1726 1727 1728 1729 1730 1731
{
	struct in6_addr maddr;

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

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

1732
/* caller must hold RTNL */
1733
void addrconf_leave_solict(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
1734 1735 1736 1737 1738 1739 1740 1741 1742 1743
{
	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);
}

1744
/* caller must hold RTNL */
1745
static void addrconf_join_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
1746 1747
{
	struct in6_addr addr;
1748

1749
	if (ifp->prefix_len >= 127) /* RFC 6164 */
1750
		return;
L
Linus Torvalds 已提交
1751 1752 1753
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
W
WANG Cong 已提交
1754
	__ipv6_dev_ac_inc(ifp->idev, &addr);
L
Linus Torvalds 已提交
1755 1756
}

1757
/* caller must hold RTNL */
1758
static void addrconf_leave_anycast(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
1759 1760
{
	struct in6_addr addr;
1761

1762
	if (ifp->prefix_len >= 127) /* RFC 6164 */
1763
		return;
L
Linus Torvalds 已提交
1764 1765 1766 1767 1768 1769
	ipv6_addr_prefix(&addr, &ifp->addr, ifp->prefix_len);
	if (ipv6_addr_any(&addr))
		return;
	__ipv6_dev_ac_dec(ifp->idev, &addr);
}

1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800
static int addrconf_ifid_eui48(u8 *eui, struct net_device *dev)
{
	if (dev->addr_len != ETH_ALEN)
		return -1;
	memcpy(eui, dev->dev_addr, 3);
	memcpy(eui + 5, dev->dev_addr + 3, 3);

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

1801 1802 1803 1804 1805
static int addrconf_ifid_eui64(u8 *eui, struct net_device *dev)
{
	if (dev->addr_len != IEEE802154_ADDR_LEN)
		return -1;
	memcpy(eui, dev->dev_addr, 8);
1806
	eui[0] ^= 2;
1807 1808 1809
	return 0;
}

1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823
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;
}

1824 1825 1826 1827 1828 1829
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);
1830
	eui[7] = *(u8 *)dev->dev_addr;
1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842
	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;
}

1843
static int __ipv6_isatap_ifid(u8 *eui, __be32 addr)
1844
{
1845 1846
	if (addr == 0)
		return -1;
1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866
	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;
}

1867 1868 1869 1870 1871
static int addrconf_ifid_gre(u8 *eui, struct net_device *dev)
{
	return __ipv6_isatap_ifid(eui, *(__be32 *)dev->dev_addr);
}

1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
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
1882 1883 1884 1885 1886
static int ipv6_generate_eui64(u8 *eui, struct net_device *dev)
{
	switch (dev->type) {
	case ARPHRD_ETHER:
	case ARPHRD_FDDI:
1887
		return addrconf_ifid_eui48(eui, dev);
L
Linus Torvalds 已提交
1888
	case ARPHRD_ARCNET:
1889
		return addrconf_ifid_arcnet(eui, dev);
L
Linus Torvalds 已提交
1890
	case ARPHRD_INFINIBAND:
1891
		return addrconf_ifid_infiniband(eui, dev);
F
Fred L. Templin 已提交
1892
	case ARPHRD_SIT:
1893
		return addrconf_ifid_sit(eui, dev);
1894 1895
	case ARPHRD_IPGRE:
		return addrconf_ifid_gre(eui, dev);
1896
	case ARPHRD_6LOWPAN:
1897 1898
	case ARPHRD_IEEE802154:
		return addrconf_ifid_eui64(eui, dev);
1899 1900
	case ARPHRD_IEEE1394:
		return addrconf_ifid_ieee1394(eui, dev);
1901 1902
	case ARPHRD_TUNNEL6:
		return addrconf_ifid_ip6tnl(eui, dev);
L
Linus Torvalds 已提交
1903 1904 1905 1906 1907 1908 1909 1910 1911 1912
	}
	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);
1913 1914 1915
	list_for_each_entry_reverse(ifp, &idev->addr_list, if_list) {
		if (ifp->scope > IFA_LINK)
			break;
L
Linus Torvalds 已提交
1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926
		if (ifp->scope == IFA_LINK && !(ifp->flags&IFA_F_TENTATIVE)) {
			memcpy(eui, ifp->addr.s6_addr+8, 8);
			err = 0;
			break;
		}
	}
	read_unlock_bh(&idev->lock);
	return err;
}

/* (re)generation of randomized interface identifier (RFC 3041 3.2, 3.5) */
S
Sorin Dumitru 已提交
1927
static void __ipv6_regen_rndid(struct inet6_dev *idev)
L
Linus Torvalds 已提交
1928 1929
{
regen:
1930
	get_random_bytes(idev->rndid, sizeof(idev->rndid));
L
Linus Torvalds 已提交
1931 1932 1933 1934 1935 1936 1937 1938
	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
1939
	 *  - ISATAP (RFC4214) 6.1
L
Linus Torvalds 已提交
1940 1941 1942 1943
	 *	00-00-5E-FE-xx-xx-xx-xx
	 *  - value 0
	 *  - XXX: already assigned to an address on the device
	 */
1944
	if (idev->rndid[0] == 0xfd &&
L
Linus Torvalds 已提交
1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960
	    (idev->rndid[1]&idev->rndid[2]&idev->rndid[3]&idev->rndid[4]&idev->rndid[5]&idev->rndid[6]) == 0xff &&
	    (idev->rndid[7]&0x80))
		goto regen;
	if ((idev->rndid[0]|idev->rndid[1]) == 0) {
		if (idev->rndid[2] == 0x5e && idev->rndid[3] == 0xfe)
			goto regen;
		if ((idev->rndid[2]|idev->rndid[3]|idev->rndid[4]|idev->rndid[5]|idev->rndid[6]|idev->rndid[7]) == 0x00)
			goto regen;
	}
}

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

1961
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
1962 1963 1964 1965 1966
	write_lock_bh(&idev->lock);

	if (idev->dead)
		goto out;

S
Sorin Dumitru 已提交
1967
	__ipv6_regen_rndid(idev);
1968

L
Linus Torvalds 已提交
1969
	expires = jiffies +
1970
		idev->cnf.temp_prefered_lft * HZ -
J
Jiri Pirko 已提交
1971 1972
		idev->cnf.regen_max_retry * idev->cnf.dad_transmits *
		NEIGH_VAR(idev->nd_parms, RETRANS_TIME) -
1973
		idev->cnf.max_desync_factor * HZ;
L
Linus Torvalds 已提交
1974
	if (time_before(expires, jiffies)) {
1975 1976
		pr_warn("%s: too short regeneration interval; timer disabled for %s\n",
			__func__, idev->dev->name);
L
Linus Torvalds 已提交
1977 1978 1979 1980 1981 1982 1983 1984
		goto out;
	}

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

out:
	write_unlock_bh(&idev->lock);
1985
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1986 1987 1988
	in6_dev_put(idev);
}

S
Sorin Dumitru 已提交
1989
static void  __ipv6_try_regen_rndid(struct inet6_dev *idev, struct in6_addr *tmpaddr)
1990
{
L
Linus Torvalds 已提交
1991
	if (tmpaddr && memcmp(idev->rndid, &tmpaddr->s6_addr[8], 8) == 0)
S
Sorin Dumitru 已提交
1992
		__ipv6_regen_rndid(idev);
L
Linus Torvalds 已提交
1993 1994 1995 1996 1997 1998 1999 2000
}

/*
 *	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
2001
		      unsigned long expires, u32 flags)
L
Linus Torvalds 已提交
2002
{
2003 2004 2005 2006 2007 2008 2009
	struct fib6_config cfg = {
		.fc_table = RT6_TABLE_PREFIX,
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_expires = expires,
		.fc_dst_len = plen,
		.fc_flags = RTF_UP | flags,
2010
		.fc_nlinfo.nl_net = dev_net(dev),
2011
		.fc_protocol = RTPROT_KERNEL,
2012
	};
L
Linus Torvalds 已提交
2013

A
Alexey Dobriyan 已提交
2014
	cfg.fc_dst = *pfx;
L
Linus Torvalds 已提交
2015 2016 2017 2018 2019

	/* 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
2020
#if IS_ENABLED(CONFIG_IPV6_SIT)
2021 2022
	if (dev->type == ARPHRD_SIT && (dev->flags & IFF_POINTOPOINT))
		cfg.fc_flags |= RTF_NONEXTHOP;
2023
#endif
L
Linus Torvalds 已提交
2024

2025
	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2026 2027
}

2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041

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

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

2042
	read_lock_bh(&table->tb6_lock);
2043 2044 2045 2046
	fn = fib6_locate(&table->tb6_root, pfx, plen, NULL, 0);
	if (!fn)
		goto out;
	for (rt = fn->leaf; rt; rt = rt->dst.rt6_next) {
2047
		if (rt->dst.dev->ifindex != dev->ifindex)
2048 2049 2050
			continue;
		if ((rt->rt6i_flags & flags) != flags)
			continue;
2051
		if ((rt->rt6i_flags & noflags) != 0)
2052 2053 2054 2055 2056
			continue;
		dst_hold(&rt->dst);
		break;
	}
out:
2057
	read_unlock_bh(&table->tb6_lock);
2058 2059 2060 2061
	return rt;
}


L
Linus Torvalds 已提交
2062 2063 2064 2065
/* Create "default" multicast route to the interface */

static void addrconf_add_mroute(struct net_device *dev)
{
2066 2067 2068 2069 2070 2071
	struct fib6_config cfg = {
		.fc_table = RT6_TABLE_LOCAL,
		.fc_metric = IP6_RT_PRIO_ADDRCONF,
		.fc_ifindex = dev->ifindex,
		.fc_dst_len = 8,
		.fc_flags = RTF_UP,
2072
		.fc_nlinfo.nl_net = dev_net(dev),
2073 2074 2075 2076 2077
	};

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

	ip6_route_add(&cfg);
L
Linus Torvalds 已提交
2078 2079 2080 2081 2082 2083 2084 2085
}

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

	ASSERT_RTNL();

S
Stephen Hemminger 已提交
2086 2087
	idev = ipv6_find_idev(dev);
	if (!idev)
2088 2089 2090 2091
		return ERR_PTR(-ENOBUFS);

	if (idev->cnf.disable_ipv6)
		return ERR_PTR(-EACCES);
L
Linus Torvalds 已提交
2092 2093

	/* Add default multicast route */
2094 2095
	if (!(dev->flags & IFF_LOOPBACK))
		addrconf_add_mroute(dev);
L
Linus Torvalds 已提交
2096 2097 2098 2099

	return idev;
}

2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166
static void manage_tempaddrs(struct inet6_dev *idev,
			     struct inet6_ifaddr *ifp,
			     __u32 valid_lft, __u32 prefered_lft,
			     bool create, unsigned long now)
{
	u32 flags;
	struct inet6_ifaddr *ift;

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

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

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

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

		if (valid_lft > max_valid)
			valid_lft = max_valid;

		if (prefered_lft > max_prefered)
			prefered_lft = max_prefered;

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

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

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

2167
void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
L
Linus Torvalds 已提交
2168 2169 2170 2171 2172 2173
{
	struct prefix_info *pinfo;
	__u32 valid_lft;
	__u32 prefered_lft;
	int addr_type;
	struct inet6_dev *in6_dev;
2174
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
2175 2176

	pinfo = (struct prefix_info *) opt;
2177

L
Linus Torvalds 已提交
2178
	if (len < sizeof(struct prefix_info)) {
2179
		ADBG("addrconf: prefix option too short\n");
L
Linus Torvalds 已提交
2180 2181
		return;
	}
2182

L
Linus Torvalds 已提交
2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195
	/*
	 *	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) {
2196
		net_warn_ratelimited("addrconf: prefix option has invalid lifetime\n");
L
Linus Torvalds 已提交
2197 2198 2199 2200 2201 2202
		return;
	}

	in6_dev = in6_dev_get(dev);

	if (in6_dev == NULL) {
2203 2204
		net_dbg_ratelimited("addrconf: device %s not configured\n",
				    dev->name);
L
Linus Torvalds 已提交
2205 2206 2207 2208 2209 2210 2211 2212 2213
		return;
	}

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

2214 2215 2216 2217
	if (pinfo->onlink) {
		struct rt6_info *rt;
		unsigned long rt_expires;

2218 2219 2220 2221 2222
		/* 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.
		 */
2223 2224 2225 2226
		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
2227

2228 2229
		if (addrconf_finite_timeout(rt_expires))
			rt_expires *= HZ;
L
Linus Torvalds 已提交
2230

2231 2232 2233 2234 2235
		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
2236

2237
		if (rt) {
2238 2239 2240 2241
			/* Autoconf prefix route */
			if (valid_lft == 0) {
				ip6_del_rt(rt);
				rt = NULL;
2242
			} else if (addrconf_finite_timeout(rt_expires)) {
2243
				/* not infinity */
2244
				rt6_set_expires(rt, jiffies + rt_expires);
2245
			} else {
2246
				rt6_clean_expires(rt);
L
Linus Torvalds 已提交
2247 2248
			}
		} else if (valid_lft) {
2249
			clock_t expires = 0;
2250 2251
			int flags = RTF_ADDRCONF | RTF_PREFIX_RT;
			if (addrconf_finite_timeout(rt_expires)) {
2252 2253 2254 2255
				/* not infinity */
				flags |= RTF_EXPIRES;
				expires = jiffies_to_clock_t(rt_expires);
			}
L
Linus Torvalds 已提交
2256
			addrconf_prefix_route(&pinfo->prefix, pinfo->prefix_len,
2257
					      dev, expires, flags);
L
Linus Torvalds 已提交
2258
		}
A
Amerigo Wang 已提交
2259
		ip6_rt_put(rt);
L
Linus Torvalds 已提交
2260 2261 2262 2263 2264
	}

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

	if (pinfo->autoconf && in6_dev->cnf.autoconf) {
2265
		struct inet6_ifaddr *ifp;
L
Linus Torvalds 已提交
2266 2267
		struct in6_addr addr;
		int create = 0, update_lft = 0;
2268
		bool tokenized = false;
L
Linus Torvalds 已提交
2269 2270 2271

		if (pinfo->prefix_len == 64) {
			memcpy(&addr, &pinfo->prefix, 8);
2272 2273 2274 2275 2276 2277

			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);
2278
				tokenized = true;
2279 2280
			} else if (ipv6_generate_eui64(addr.s6_addr + 8, dev) &&
				   ipv6_inherit_eui64(addr.s6_addr + 8, in6_dev)) {
L
Linus Torvalds 已提交
2281 2282 2283 2284 2285
				in6_dev_put(in6_dev);
				return;
			}
			goto ok;
		}
2286 2287
		net_dbg_ratelimited("IPv6 addrconf: prefix with wrong length %d\n",
				    pinfo->prefix_len);
L
Linus Torvalds 已提交
2288 2289 2290 2291 2292
		in6_dev_put(in6_dev);
		return;

ok:

2293
		ifp = ipv6_get_ifaddr(net, &addr, dev, 1);
L
Linus Torvalds 已提交
2294 2295 2296

		if (ifp == NULL && valid_lft) {
			int max_addresses = in6_dev->cnf.max_addresses;
2297 2298 2299 2300
			u32 addr_flags = 0;

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
			if (in6_dev->cnf.optimistic_dad &&
2301
			    !net->ipv6.devconf_all->forwarding && sllao)
2302 2303
				addr_flags = IFA_F_OPTIMISTIC;
#endif
L
Linus Torvalds 已提交
2304 2305 2306 2307 2308 2309

			/* 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)
2310 2311
				ifp = ipv6_add_addr(in6_dev, &addr, NULL,
						    pinfo->prefix_len,
2312
						    addr_type&IPV6_ADDR_SCOPE_MASK,
2313 2314
						    addr_flags, valid_lft,
						    prefered_lft);
L
Linus Torvalds 已提交
2315

2316
			if (IS_ERR_OR_NULL(ifp)) {
L
Linus Torvalds 已提交
2317 2318 2319 2320
				in6_dev_put(in6_dev);
				return;
			}

2321 2322
			update_lft = 0;
			create = 1;
2323 2324
			spin_lock_bh(&ifp->lock);
			ifp->flags |= IFA_F_MANAGETEMPADDR;
L
Linus Torvalds 已提交
2325
			ifp->cstamp = jiffies;
2326
			ifp->tokenized = tokenized;
2327
			spin_unlock_bh(&ifp->lock);
2328
			addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
2329 2330 2331
		}

		if (ifp) {
2332
			u32 flags;
L
Linus Torvalds 已提交
2333 2334 2335 2336 2337 2338 2339 2340 2341 2342
			unsigned long now;
			u32 stored_lft;

			/* update lifetime (RFC2462 5.5.3 e) */
			spin_lock(&ifp->lock);
			now = jiffies;
			if (ifp->valid_lft > (now - ifp->tstamp) / HZ)
				stored_lft = ifp->valid_lft - (now - ifp->tstamp) / HZ;
			else
				stored_lft = 0;
2343
			if (!update_lft && !create && stored_lft) {
2344 2345
				const u32 minimum_lft = min_t(u32,
					stored_lft, MIN_VALID_LIFETIME);
2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358
				valid_lft = max(valid_lft, minimum_lft);

				/* RFC4862 Section 5.5.3e:
				 * "Note that the preferred lifetime of the
				 *  corresponding address is always reset to
				 *  the Preferred Lifetime in the received
				 *  Prefix Information option, regardless of
				 *  whether the valid lifetime is also reset or
				 *  ignored."
				 *
				 * So we should always update prefered_lft here.
				 */
				update_lft = 1;
L
Linus Torvalds 已提交
2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373
			}

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

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

2374 2375
			manage_tempaddrs(in6_dev, ifp, valid_lft, prefered_lft,
					 create, now);
2376

L
Linus Torvalds 已提交
2377
			in6_ifa_put(ifp);
2378
			addrconf_verify();
L
Linus Torvalds 已提交
2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389
		}
	}
	inet6_prefix_notify(RTM_NEWPREFIX, in6_dev, pinfo);
	in6_dev_put(in6_dev);
}

/*
 *	Set destination address.
 *	Special case for SIT interfaces where we create a new "virtual"
 *	device.
 */
2390
int addrconf_set_dstaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401
{
	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;

2402
	dev = __dev_get_by_index(net, ireq.ifr6_ifindex);
L
Linus Torvalds 已提交
2403 2404 2405 2406 2407

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

A
Amerigo Wang 已提交
2408
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2409
	if (dev->type == ARPHRD_SIT) {
2410
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424
		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;
2425
		ifr.ifr_ifru.ifru_data = (__force void __user *)&p;
L
Linus Torvalds 已提交
2426

2427 2428 2429 2430 2431 2432 2433 2434
		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
2435 2436 2437

		if (err == 0) {
			err = -ENOBUFS;
2438 2439
			dev = __dev_get_by_name(net, p.name);
			if (!dev)
L
Linus Torvalds 已提交
2440 2441 2442 2443
				goto err_exit;
			err = dev_open(dev);
		}
	}
2444
#endif
L
Linus Torvalds 已提交
2445 2446 2447 2448 2449 2450 2451 2452 2453

err_exit:
	rtnl_unlock();
	return err;
}

/*
 *	Manual configuration of address on an interface
 */
2454 2455
static int inet6_addr_add(struct net *net, int ifindex,
			  const struct in6_addr *pfx,
2456
			  const struct in6_addr *peer_pfx,
2457 2458
			  unsigned int plen, __u32 ifa_flags,
			  __u32 prefered_lft, __u32 valid_lft)
L
Linus Torvalds 已提交
2459 2460 2461 2462 2463
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
	int scope;
2464 2465
	u32 flags;
	clock_t expires;
2466
	unsigned long timeout;
L
Linus Torvalds 已提交
2467 2468

	ASSERT_RTNL();
2469

2470 2471 2472
	if (plen > 128)
		return -EINVAL;

2473 2474 2475 2476
	/* check the lifetime */
	if (!valid_lft || prefered_lft > valid_lft)
		return -EINVAL;

2477 2478 2479
	if (ifa_flags & IFA_F_MANAGETEMPADDR && plen != 64)
		return -EINVAL;

2480 2481
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2482
		return -ENODEV;
2483

2484 2485 2486
	idev = addrconf_add_dev(dev);
	if (IS_ERR(idev))
		return PTR_ERR(idev);
L
Linus Torvalds 已提交
2487 2488 2489

	scope = ipv6_addr_scope(pfx);

2490 2491 2492 2493
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
2494
		flags = RTF_EXPIRES;
2495 2496 2497 2498
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
2499
	}
2500

2501 2502 2503 2504 2505 2506
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
2507

2508 2509
	ifp = ipv6_add_addr(idev, pfx, peer_pfx, plen, scope, ifa_flags,
			    valid_lft, prefered_lft);
2510

L
Linus Torvalds 已提交
2511
	if (!IS_ERR(ifp)) {
2512 2513 2514 2515 2516
		if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
			addrconf_prefix_route(&ifp->addr, ifp->prefix_len, dev,
					      expires, flags);
		}

2517 2518 2519 2520 2521
		/*
		 * Note that section 3.1 of RFC 4429 indicates
		 * that the Optimistic flag should not be set for
		 * manually configured addresses
		 */
2522
		addrconf_dad_start(ifp);
2523 2524 2525
		if (ifa_flags & IFA_F_MANAGETEMPADDR)
			manage_tempaddrs(idev, ifp, valid_lft, prefered_lft,
					 true, jiffies);
L
Linus Torvalds 已提交
2526
		in6_ifa_put(ifp);
2527
		addrconf_verify_rtnl();
L
Linus Torvalds 已提交
2528 2529 2530 2531 2532 2533
		return 0;
	}

	return PTR_ERR(ifp);
}

2534 2535
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
2536 2537 2538 2539
{
	struct inet6_ifaddr *ifp;
	struct inet6_dev *idev;
	struct net_device *dev;
2540

2541 2542 2543
	if (plen > 128)
		return -EINVAL;

2544 2545
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
L
Linus Torvalds 已提交
2546 2547
		return -ENODEV;

2548
	if ((idev = __in6_dev_get(dev)) == NULL)
L
Linus Torvalds 已提交
2549 2550 2551
		return -ENXIO;

	read_lock_bh(&idev->lock);
2552
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
L
Linus Torvalds 已提交
2553 2554 2555 2556
		if (ifp->prefix_len == plen &&
		    ipv6_addr_equal(pfx, &ifp->addr)) {
			in6_ifa_hold(ifp);
			read_unlock_bh(&idev->lock);
2557

2558 2559 2560 2561
			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
2562
			ipv6_del_addr(ifp);
2563
			addrconf_verify_rtnl();
L
Linus Torvalds 已提交
2564 2565 2566 2567 2568 2569 2570 2571
			return 0;
		}
	}
	read_unlock_bh(&idev->lock);
	return -EADDRNOTAVAIL;
}


2572
int addrconf_add_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2573 2574 2575
{
	struct in6_ifreq ireq;
	int err;
2576

2577
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2578
		return -EPERM;
2579

L
Linus Torvalds 已提交
2580 2581 2582 2583
	if (copy_from_user(&ireq, arg, sizeof(struct in6_ifreq)))
		return -EFAULT;

	rtnl_lock();
2584
	err = inet6_addr_add(net, ireq.ifr6_ifindex, &ireq.ifr6_addr, NULL,
2585 2586
			     ireq.ifr6_prefixlen, IFA_F_PERMANENT,
			     INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
2587 2588 2589 2590
	rtnl_unlock();
	return err;
}

2591
int addrconf_del_ifaddr(struct net *net, void __user *arg)
L
Linus Torvalds 已提交
2592 2593 2594
{
	struct in6_ifreq ireq;
	int err;
2595

2596
	if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
L
Linus Torvalds 已提交
2597 2598 2599 2600 2601 2602
		return -EPERM;

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

	rtnl_lock();
2603
	err = inet6_addr_del(net, ireq.ifr6_ifindex, 0, &ireq.ifr6_addr,
2604
			     ireq.ifr6_prefixlen);
L
Linus Torvalds 已提交
2605 2606 2607 2608
	rtnl_unlock();
	return err;
}

2609 2610 2611 2612 2613
static void add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
		     int plen, int scope)
{
	struct inet6_ifaddr *ifp;

2614
	ifp = ipv6_add_addr(idev, addr, NULL, plen,
2615 2616
			    scope, IFA_F_PERMANENT,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
2617 2618 2619 2620 2621 2622 2623 2624 2625
	if (!IS_ERR(ifp)) {
		spin_lock_bh(&ifp->lock);
		ifp->flags &= ~IFA_F_TENTATIVE;
		spin_unlock_bh(&ifp->lock);
		ipv6_ifa_notify(RTM_NEWADDR, ifp);
		in6_ifa_put(ifp);
	}
}

A
Amerigo Wang 已提交
2626
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2627 2628 2629 2630
static void sit_add_v4_addrs(struct inet6_dev *idev)
{
	struct in6_addr addr;
	struct net_device *dev;
2631
	struct net *net = dev_net(idev->dev);
2632
	int scope, plen;
2633
	u32 pflags = 0;
L
Linus Torvalds 已提交
2634 2635 2636 2637 2638 2639 2640 2641 2642

	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;
2643
		plen = 64;
L
Linus Torvalds 已提交
2644 2645
	} else {
		scope = IPV6_ADDR_COMPATv4;
2646
		plen = 96;
2647
		pflags |= RTF_NONEXTHOP;
L
Linus Torvalds 已提交
2648 2649 2650
	}

	if (addr.s6_addr32[3]) {
2651
		add_addr(idev, &addr, plen, scope);
2652
		addrconf_prefix_route(&addr, plen, idev->dev, 0, pflags);
L
Linus Torvalds 已提交
2653 2654 2655
		return;
	}

2656
	for_each_netdev(net, dev) {
2657
		struct in_device *in_dev = __in_dev_get_rtnl(dev);
L
Linus Torvalds 已提交
2658
		if (in_dev && (dev->flags & IFF_UP)) {
2659
			struct in_ifaddr *ifa;
L
Linus Torvalds 已提交
2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674

			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;
				}

2675
				add_addr(idev, &addr, plen, flag);
2676 2677
				addrconf_prefix_route(&addr, plen, idev->dev, 0,
						      pflags);
L
Linus Torvalds 已提交
2678 2679
			}
		}
2680
	}
L
Linus Torvalds 已提交
2681
}
2682
#endif
L
Linus Torvalds 已提交
2683 2684 2685 2686

static void init_loopback(struct net_device *dev)
{
	struct inet6_dev  *idev;
2687 2688 2689
	struct net_device *sp_dev;
	struct inet6_ifaddr *sp_ifa;
	struct rt6_info *sp_rt;
L
Linus Torvalds 已提交
2690 2691 2692 2693 2694

	/* ::1 */

	ASSERT_RTNL();

2695
	if ((idev = ipv6_find_idev(dev)) == NULL) {
2696
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
2697 2698 2699
		return;
	}

2700
	add_addr(idev, &in6addr_loopback, 128, IFA_HOST);
2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716

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

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

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

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

2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728
			if (sp_ifa->rt) {
				/* This dst has been added to garbage list when
				 * lo device down, release this obsolete dst and
				 * reallocate a new router for ifa.
				 */
				if (sp_ifa->rt->dst.obsolete > 0) {
					ip6_rt_put(sp_ifa->rt);
					sp_ifa->rt = NULL;
				} else {
					continue;
				}
			}
2729

2730
			sp_rt = addrconf_dst_alloc(idev, &sp_ifa->addr, false);
2731 2732

			/* Failure cases are ignored */
2733 2734
			if (!IS_ERR(sp_rt)) {
				sp_ifa->rt = sp_rt;
2735
				ip6_ins_rt(sp_rt);
2736
			}
2737 2738 2739
		}
		read_unlock_bh(&idev->lock);
	}
L
Linus Torvalds 已提交
2740 2741
}

2742
static void addrconf_add_linklocal(struct inet6_dev *idev, const struct in6_addr *addr)
L
Linus Torvalds 已提交
2743
{
2744
	struct inet6_ifaddr *ifp;
2745 2746 2747 2748
	u32 addr_flags = IFA_F_PERMANENT;

#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	if (idev->cnf.optimistic_dad &&
2749
	    !dev_net(idev->dev)->ipv6.devconf_all->forwarding)
2750 2751
		addr_flags |= IFA_F_OPTIMISTIC;
#endif
L
Linus Torvalds 已提交
2752

2753

2754 2755
	ifp = ipv6_add_addr(idev, addr, NULL, 64, IFA_LINK, addr_flags,
			    INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
L
Linus Torvalds 已提交
2756
	if (!IS_ERR(ifp)) {
2757
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, idev->dev, 0, 0);
2758
		addrconf_dad_start(ifp);
L
Linus Torvalds 已提交
2759 2760 2761 2762
		in6_ifa_put(ifp);
	}
}

2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779
static void addrconf_addr_gen(struct inet6_dev *idev, bool prefix_route)
{
	if (idev->addr_gen_mode == IN6_ADDR_GEN_MODE_EUI64) {
		struct in6_addr addr;

		ipv6_addr_set(&addr,  htonl(0xFE800000), 0, 0, 0);
		/* 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)
			addrconf_add_linklocal(idev, &addr);
		else if (prefix_route)
			addrconf_prefix_route(&addr, 64, idev->dev, 0, 0);
	}
}

L
Linus Torvalds 已提交
2780 2781
static void addrconf_dev_config(struct net_device *dev)
{
2782
	struct inet6_dev *idev;
L
Linus Torvalds 已提交
2783 2784 2785

	ASSERT_RTNL();

2786 2787 2788
	if ((dev->type != ARPHRD_ETHER) &&
	    (dev->type != ARPHRD_FDDI) &&
	    (dev->type != ARPHRD_ARCNET) &&
2789
	    (dev->type != ARPHRD_INFINIBAND) &&
2790
	    (dev->type != ARPHRD_IEEE802154) &&
2791
	    (dev->type != ARPHRD_IEEE1394) &&
2792 2793
	    (dev->type != ARPHRD_TUNNEL6) &&
	    (dev->type != ARPHRD_6LOWPAN)) {
2794 2795 2796 2797
		/* Alas, we support only Ethernet autoconfiguration. */
		return;
	}

L
Linus Torvalds 已提交
2798
	idev = addrconf_add_dev(dev);
2799
	if (IS_ERR(idev))
L
Linus Torvalds 已提交
2800 2801
		return;

2802
	addrconf_addr_gen(idev, false);
L
Linus Torvalds 已提交
2803 2804
}

A
Amerigo Wang 已提交
2805
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2806 2807 2808 2809 2810 2811
static void addrconf_sit_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

2812 2813 2814
	/*
	 * Configure the tunnel with one of our IPv4
	 * addresses... we should configure all of
L
Linus Torvalds 已提交
2815 2816 2817
	 * our v4 addrs in the tunnel
	 */

2818
	if ((idev = ipv6_find_idev(dev)) == NULL) {
2819
		pr_debug("%s: add_dev failed\n", __func__);
L
Linus Torvalds 已提交
2820 2821 2822
		return;
	}

F
Fred L. Templin 已提交
2823
	if (dev->priv_flags & IFF_ISATAP) {
2824
		addrconf_addr_gen(idev, false);
F
Fred L. Templin 已提交
2825 2826 2827
		return;
	}

L
Linus Torvalds 已提交
2828 2829
	sit_add_v4_addrs(idev);

2830
	if (dev->flags&IFF_POINTOPOINT)
L
Linus Torvalds 已提交
2831 2832
		addrconf_add_mroute(dev);
}
2833
#endif
L
Linus Torvalds 已提交
2834

A
Amerigo Wang 已提交
2835
#if IS_ENABLED(CONFIG_NET_IPGRE)
2836 2837 2838 2839 2840 2841
static void addrconf_gre_config(struct net_device *dev)
{
	struct inet6_dev *idev;

	ASSERT_RTNL();

2842
	if ((idev = ipv6_find_idev(dev)) == NULL) {
2843
		pr_debug("%s: add_dev failed\n", __func__);
2844 2845 2846
		return;
	}

2847
	addrconf_addr_gen(idev, true);
2848 2849 2850
}
#endif

2851
static int addrconf_notify(struct notifier_block *this, unsigned long event,
2852
			   void *ptr)
L
Linus Torvalds 已提交
2853
{
2854
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
2855
	struct inet6_dev *idev = __in6_dev_get(dev);
2856
	int run_pending = 0;
2857
	int err;
L
Linus Torvalds 已提交
2858

S
Stephen Hemminger 已提交
2859
	switch (event) {
2860
	case NETDEV_REGISTER:
2861
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
2862
			idev = ipv6_add_dev(dev);
2863 2864
			if (IS_ERR(idev))
				return notifier_from_errno(PTR_ERR(idev));
2865 2866
		}
		break;
S
Stephen Hemminger 已提交
2867

L
Linus Torvalds 已提交
2868
	case NETDEV_UP:
2869
	case NETDEV_CHANGE:
2870 2871 2872
		if (dev->flags & IFF_SLAVE)
			break;

2873 2874 2875
		if (idev && idev->cnf.disable_ipv6)
			break;

2876
		if (event == NETDEV_UP) {
2877
			if (!addrconf_qdisc_ok(dev)) {
2878
				/* device is not ready yet. */
2879
				pr_info("ADDRCONF(NETDEV_UP): %s: link is not ready\n",
2880 2881 2882
					dev->name);
				break;
			}
2883

2884 2885 2886
			if (!idev && dev->mtu >= IPV6_MIN_MTU)
				idev = ipv6_add_dev(dev);

2887
			if (!IS_ERR_OR_NULL(idev)) {
2888
				idev->if_flags |= IF_READY;
2889 2890
				run_pending = 1;
			}
2891
		} else {
2892
			if (!addrconf_qdisc_ok(dev)) {
2893 2894 2895 2896 2897
				/* device is still not ready. */
				break;
			}

			if (idev) {
S
Stephen Hemminger 已提交
2898
				if (idev->if_flags & IF_READY)
2899 2900 2901 2902 2903
					/* device is already configured. */
					break;
				idev->if_flags |= IF_READY;
			}

2904 2905
			pr_info("ADDRCONF(NETDEV_CHANGE): %s: link becomes ready\n",
				dev->name);
2906

2907
			run_pending = 1;
2908 2909
		}

S
Stephen Hemminger 已提交
2910
		switch (dev->type) {
A
Amerigo Wang 已提交
2911
#if IS_ENABLED(CONFIG_IPV6_SIT)
L
Linus Torvalds 已提交
2912 2913 2914
		case ARPHRD_SIT:
			addrconf_sit_config(dev);
			break;
2915
#endif
A
Amerigo Wang 已提交
2916
#if IS_ENABLED(CONFIG_NET_IPGRE)
2917 2918 2919
		case ARPHRD_IPGRE:
			addrconf_gre_config(dev);
			break;
2920
#endif
L
Linus Torvalds 已提交
2921 2922 2923 2924 2925 2926 2927
		case ARPHRD_LOOPBACK:
			init_loopback(dev);
			break;

		default:
			addrconf_dev_config(dev);
			break;
2928
		}
S
Stephen Hemminger 已提交
2929

2930
		if (!IS_ERR_OR_NULL(idev)) {
2931 2932 2933
			if (run_pending)
				addrconf_dad_run(idev);

S
Stephen Hemminger 已提交
2934 2935 2936 2937
			/*
			 * 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
2938
			 */
S
Stephen Hemminger 已提交
2939 2940
			if (idev->cnf.mtu6 != dev->mtu &&
			    dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
2941 2942 2943 2944 2945
				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
2946 2947 2948 2949

			/*
			 * If the changed mtu during down is lower than
			 * IPV6_MIN_MTU stop IPv6 on this interface.
L
Linus Torvalds 已提交
2950 2951
			 */
			if (dev->mtu < IPV6_MIN_MTU)
S
Stephen Hemminger 已提交
2952
				addrconf_ifdown(dev, 1);
L
Linus Torvalds 已提交
2953 2954 2955 2956
		}
		break;

	case NETDEV_CHANGEMTU:
2957
		if (idev && dev->mtu >= IPV6_MIN_MTU) {
L
Linus Torvalds 已提交
2958 2959 2960 2961 2962
			rt6_mtu_change(dev, dev->mtu);
			idev->cnf.mtu6 = dev->mtu;
			break;
		}

2963 2964
		if (!idev && dev->mtu >= IPV6_MIN_MTU) {
			idev = ipv6_add_dev(dev);
2965
			if (!IS_ERR(idev))
2966 2967 2968
				break;
		}

S
Stephen Hemminger 已提交
2969
		/*
2970
		 * if MTU under IPV6_MIN_MTU.
S
Stephen Hemminger 已提交
2971 2972
		 * Stop IPv6 on this interface.
		 */
L
Linus Torvalds 已提交
2973 2974 2975 2976 2977 2978 2979 2980

	case NETDEV_DOWN:
	case NETDEV_UNREGISTER:
		/*
		 *	Remove all addresses from this interface.
		 */
		addrconf_ifdown(dev, event != NETDEV_DOWN);
		break;
2981

L
Linus Torvalds 已提交
2982 2983
	case NETDEV_CHANGENAME:
		if (idev) {
2984
			snmp6_unregister_dev(idev);
2985
			addrconf_sysctl_unregister(idev);
2986
			err = addrconf_sysctl_register(idev);
2987 2988
			if (err)
				return notifier_from_errno(err);
2989 2990 2991 2992 2993
			err = snmp6_register_dev(idev);
			if (err) {
				addrconf_sysctl_unregister(idev);
				return notifier_from_errno(err);
			}
2994
		}
L
Linus Torvalds 已提交
2995
		break;
S
Stephen Hemminger 已提交
2996

2997 2998 2999
	case NETDEV_PRE_TYPE_CHANGE:
	case NETDEV_POST_TYPE_CHANGE:
		addrconf_type_change(dev, event);
3000
		break;
3001
	}
L
Linus Torvalds 已提交
3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012

	return NOTIFY_OK;
}

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

3013
static void addrconf_type_change(struct net_device *dev, unsigned long event)
3014 3015 3016 3017 3018 3019
{
	struct inet6_dev *idev;
	ASSERT_RTNL();

	idev = __in6_dev_get(dev);

3020
	if (event == NETDEV_POST_TYPE_CHANGE)
3021
		ipv6_mc_remap(idev);
3022
	else if (event == NETDEV_PRE_TYPE_CHANGE)
3023 3024 3025
		ipv6_mc_unmap(idev);
}

L
Linus Torvalds 已提交
3026 3027
static int addrconf_ifdown(struct net_device *dev, int how)
{
3028
	struct net *net = dev_net(dev);
3029 3030
	struct inet6_dev *idev;
	struct inet6_ifaddr *ifa;
3031
	int state, i;
L
Linus Torvalds 已提交
3032 3033 3034

	ASSERT_RTNL();

3035 3036
	rt6_ifdown(net, dev);
	neigh_ifdown(&nd_tbl, dev);
L
Linus Torvalds 已提交
3037 3038 3039 3040 3041

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

S
Stephen Hemminger 已提交
3042 3043 3044
	/*
	 * Step 1: remove reference to ipv6 device from parent device.
	 *	   Do not dev_put!
L
Linus Torvalds 已提交
3045
	 */
3046
	if (how) {
L
Linus Torvalds 已提交
3047
		idev->dead = 1;
3048 3049

		/* protected by rtnl_lock */
3050
		RCU_INIT_POINTER(dev->ip6_ptr, NULL);
L
Linus Torvalds 已提交
3051 3052 3053 3054 3055 3056

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

	}

3057 3058 3059 3060 3061
	/* 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);
3062
restart:
3063
		hlist_for_each_entry_rcu(ifa, h, addr_lst) {
3064 3065
			if (ifa->idev == idev) {
				hlist_del_init_rcu(&ifa->addr_lst);
3066
				addrconf_del_dad_work(ifa);
3067 3068 3069 3070 3071 3072
				goto restart;
			}
		}
		spin_unlock_bh(&addrconf_hash_lock);
	}

L
Linus Torvalds 已提交
3073 3074
	write_lock_bh(&idev->lock);

3075 3076
	addrconf_del_rs_timer(idev);

S
Stephen Hemminger 已提交
3077
	/* Step 2: clear flags for stateless addrconf */
3078
	if (!how)
3079
		idev->if_flags &= ~(IF_RS_SENT|IF_RA_RCVD|IF_READY);
L
Linus Torvalds 已提交
3080

3081
	if (how && del_timer(&idev->regen_timer))
L
Linus Torvalds 已提交
3082 3083
		in6_dev_put(idev);

S
Stephen Hemminger 已提交
3084
	/* Step 3: clear tempaddr list */
3085 3086 3087 3088
	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
3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099
		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);
	}
3100

3101 3102 3103
	while (!list_empty(&idev->addr_list)) {
		ifa = list_first_entry(&idev->addr_list,
				       struct inet6_ifaddr, if_list);
3104
		addrconf_del_dad_work(ifa);
3105

3106
		list_del(&ifa->if_list);
3107

3108
		write_unlock_bh(&idev->lock);
3109

3110 3111 3112 3113
		spin_lock_bh(&ifa->state_lock);
		state = ifa->state;
		ifa->state = INET6_IFADDR_STATE_DEAD;
		spin_unlock_bh(&ifa->state_lock);
3114

3115 3116
		if (state != INET6_IFADDR_STATE_DEAD) {
			__ipv6_ifa_notify(RTM_DELADDR, ifa);
3117
			inet6addr_notifier_call_chain(NETDEV_DOWN, ifa);
3118
		}
3119
		in6_ifa_put(ifa);
3120

3121 3122
		write_lock_bh(&idev->lock);
	}
3123

L
Linus Torvalds 已提交
3124 3125
	write_unlock_bh(&idev->lock);

3126 3127 3128
	/* Step 5: Discard anycast and multicast list */
	if (how) {
		ipv6_ac_destroy_dev(idev);
L
Linus Torvalds 已提交
3129
		ipv6_mc_destroy_dev(idev);
3130
	} else {
L
Linus Torvalds 已提交
3131
		ipv6_mc_down(idev);
3132
	}
L
Linus Torvalds 已提交
3133 3134

	idev->tstamp = jiffies;
3135

S
Stephen Hemminger 已提交
3136
	/* Last: Shot the device (if unregistered) */
3137
	if (how) {
3138
		addrconf_sysctl_unregister(idev);
L
Linus Torvalds 已提交
3139 3140 3141 3142 3143 3144 3145 3146 3147
		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)
{
3148
	struct inet6_dev *idev = (struct inet6_dev *)data;
3149
	struct net_device *dev = idev->dev;
3150
	struct in6_addr lladdr;
L
Linus Torvalds 已提交
3151

3152
	write_lock(&idev->lock);
S
stephen hemminger 已提交
3153
	if (idev->dead || !(idev->if_flags & IF_READY))
L
Linus Torvalds 已提交
3154 3155
		goto out;

3156
	if (!ipv6_accept_ra(idev))
S
stephen hemminger 已提交
3157 3158 3159 3160
		goto out;

	/* Announcement received after solicitation was sent */
	if (idev->if_flags & IF_RA_RCVD)
L
Linus Torvalds 已提交
3161 3162
		goto out;

3163
	if (idev->rs_probes++ < idev->cnf.rtr_solicits) {
3164 3165 3166
		write_unlock(&idev->lock);
		if (!ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
			ndisc_send_rs(dev, &lladdr,
3167 3168
				      &in6addr_linklocal_allrouters);
		else
3169
			goto put;
L
Linus Torvalds 已提交
3170

3171
		write_lock(&idev->lock);
3172 3173 3174 3175 3176
		/* The wait after the last probe can be shorter */
		addrconf_mod_rs_timer(idev, (idev->rs_probes ==
					     idev->cnf.rtr_solicits) ?
				      idev->cnf.rtr_solicit_delay :
				      idev->cnf.rtr_solicit_interval);
L
Linus Torvalds 已提交
3177 3178 3179 3180 3181
	} else {
		/*
		 * Note: we do not support deprecated "all on-link"
		 * assumption any longer.
		 */
3182
		pr_debug("%s: no IPv6 routers present\n", idev->dev->name);
L
Linus Torvalds 已提交
3183 3184 3185
	}

out:
3186
	write_unlock(&idev->lock);
3187
put:
3188
	in6_dev_put(idev);
L
Linus Torvalds 已提交
3189 3190 3191 3192 3193
}

/*
 *	Duplicate Address Detection
 */
3194 3195 3196 3197 3198
static void addrconf_dad_kick(struct inet6_ifaddr *ifp)
{
	unsigned long rand_num;
	struct inet6_dev *idev = ifp->idev;

3199 3200 3201
	if (ifp->flags & IFA_F_OPTIMISTIC)
		rand_num = 0;
	else
3202
		rand_num = prandom_u32() % (idev->cnf.rtr_solicit_delay ? : 1);
3203

3204
	ifp->dad_probes = idev->cnf.dad_transmits;
3205
	addrconf_mod_dad_work(ifp, rand_num);
3206 3207
}

3208
static void addrconf_dad_begin(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3209 3210 3211 3212 3213 3214
{
	struct inet6_dev *idev = ifp->idev;
	struct net_device *dev = idev->dev;

	addrconf_join_solict(dev, &ifp->addr);

3215
	prandom_seed((__force u32) ifp->addr.s6_addr32[3]);
L
Linus Torvalds 已提交
3216 3217

	read_lock_bh(&idev->lock);
3218
	spin_lock(&ifp->lock);
3219
	if (ifp->state == INET6_IFADDR_STATE_DEAD)
L
Linus Torvalds 已提交
3220 3221 3222
		goto out;

	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
3223
	    idev->cnf.accept_dad < 1 ||
3224 3225
	    !(ifp->flags&IFA_F_TENTATIVE) ||
	    ifp->flags & IFA_F_NODAD) {
B
Brian Haley 已提交
3226
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3227
		spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3228 3229 3230 3231 3232 3233
		read_unlock_bh(&idev->lock);

		addrconf_dad_completed(ifp);
		return;
	}

3234
	if (!(idev->if_flags & IF_READY)) {
3235
		spin_unlock(&ifp->lock);
3236
		read_unlock_bh(&idev->lock);
3237
		/*
3238
		 * If the device is not ready:
3239 3240 3241 3242
		 * - keep it tentative if it is a permanent address.
		 * - otherwise, kill it.
		 */
		in6_ifa_hold(ifp);
B
Brian Haley 已提交
3243
		addrconf_dad_stop(ifp, 0);
3244
		return;
3245
	}
3246 3247 3248 3249 3250

	/*
	 * Optimistic nodes can start receiving
	 * Frames right away
	 */
3251
	if (ifp->flags & IFA_F_OPTIMISTIC) {
3252
		ip6_ins_rt(ifp->rt);
3253 3254 3255 3256 3257 3258 3259
		if (ipv6_use_optimistic_addr(idev)) {
			/* Because optimistic nodes can use this address,
			 * notify listeners. If DAD fails, RTM_DELADDR is sent.
			 */
			ipv6_ifa_notify(RTM_NEWADDR, ifp);
		}
	}
3260

3261
	addrconf_dad_kick(ifp);
L
Linus Torvalds 已提交
3262
out:
3263
	spin_unlock(&ifp->lock);
L
Linus Torvalds 已提交
3264 3265 3266
	read_unlock_bh(&idev->lock);
}

3267
static void addrconf_dad_start(struct inet6_ifaddr *ifp)
L
Linus Torvalds 已提交
3268
{
3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286
	bool begin_dad = false;

	spin_lock_bh(&ifp->state_lock);
	if (ifp->state != INET6_IFADDR_STATE_DEAD) {
		ifp->state = INET6_IFADDR_STATE_PREDAD;
		begin_dad = true;
	}
	spin_unlock_bh(&ifp->state_lock);

	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
3287 3288 3289
	struct inet6_dev *idev = ifp->idev;
	struct in6_addr mcaddr;

3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315
	enum {
		DAD_PROCESS,
		DAD_BEGIN,
		DAD_ABORT,
	} action = DAD_PROCESS;

	rtnl_lock();

	spin_lock_bh(&ifp->state_lock);
	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;
	}
	spin_unlock_bh(&ifp->state_lock);

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

3316
	if (!ifp->dad_probes && addrconf_dad_end(ifp))
H
Herbert Xu 已提交
3317 3318
		goto out;

3319
	write_lock_bh(&idev->lock);
3320
	if (idev->dead || !(idev->if_flags & IF_READY)) {
3321
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3322 3323
		goto out;
	}
3324 3325

	spin_lock(&ifp->lock);
3326 3327
	if (ifp->state == INET6_IFADDR_STATE_DEAD) {
		spin_unlock(&ifp->lock);
3328
		write_unlock_bh(&idev->lock);
3329 3330 3331
		goto out;
	}

3332
	if (ifp->dad_probes == 0) {
L
Linus Torvalds 已提交
3333 3334 3335 3336
		/*
		 * DAD was successful
		 */

B
Brian Haley 已提交
3337
		ifp->flags &= ~(IFA_F_TENTATIVE|IFA_F_OPTIMISTIC|IFA_F_DADFAILED);
3338
		spin_unlock(&ifp->lock);
3339
		write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3340 3341 3342 3343 3344 3345

		addrconf_dad_completed(ifp);

		goto out;
	}

3346
	ifp->dad_probes--;
3347 3348
	addrconf_mod_dad_work(ifp,
			      NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME));
3349
	spin_unlock(&ifp->lock);
3350
	write_unlock_bh(&idev->lock);
L
Linus Torvalds 已提交
3351 3352 3353

	/* send a neighbour solicitation for our addr */
	addrconf_addr_solict_mult(&ifp->addr, &mcaddr);
3354
	ndisc_send_ns(ifp->idev->dev, NULL, &ifp->addr, &mcaddr, &in6addr_any);
L
Linus Torvalds 已提交
3355 3356
out:
	in6_ifa_put(ifp);
3357
	rtnl_unlock();
L
Linus Torvalds 已提交
3358 3359
}

3360 3361 3362 3363 3364 3365
/* 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;

3366 3367 3368
	list_for_each_entry_reverse(ifpiter, &idev->addr_list, if_list) {
		if (ifpiter->scope > IFA_LINK)
			break;
3369 3370 3371 3372 3373 3374 3375 3376 3377
		if (ifp != ifpiter && ifpiter->scope == IFA_LINK &&
		    (ifpiter->flags & (IFA_F_PERMANENT|IFA_F_TENTATIVE|
				       IFA_F_OPTIMISTIC|IFA_F_DADFAILED)) ==
		    IFA_F_PERMANENT)
			return false;
	}
	return true;
}

L
Linus Torvalds 已提交
3378 3379
static void addrconf_dad_completed(struct inet6_ifaddr *ifp)
{
S
Stephen Hemminger 已提交
3380
	struct net_device *dev = ifp->idev->dev;
3381
	struct in6_addr lladdr;
3382
	bool send_rs, send_mld;
3383

3384
	addrconf_del_dad_work(ifp);
L
Linus Torvalds 已提交
3385 3386 3387 3388 3389 3390 3391

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

	ipv6_ifa_notify(RTM_NEWADDR, ifp);

3392 3393
	/* If added prefix is link local and we are prepared to process
	   router advertisements, start sending router solicitations.
L
Linus Torvalds 已提交
3394 3395
	 */

3396
	read_lock_bh(&ifp->idev->lock);
3397
	send_mld = ifp->scope == IFA_LINK && ipv6_lonely_lladdr(ifp);
3398 3399
	send_rs = send_mld &&
		  ipv6_accept_ra(ifp->idev) &&
3400
		  ifp->idev->cnf.rtr_solicits > 0 &&
3401
		  (dev->flags&IFF_LOOPBACK) == 0;
3402 3403
	read_unlock_bh(&ifp->idev->lock);

3404 3405 3406 3407 3408 3409
	/* 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);

3410
	if (send_rs) {
L
Linus Torvalds 已提交
3411 3412 3413 3414 3415
		/*
		 *	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
		 */
3416
		if (ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
3417
			return;
3418
		ndisc_send_rs(dev, &lladdr, &in6addr_linklocal_allrouters);
L
Linus Torvalds 已提交
3419

3420 3421 3422
		write_lock_bh(&ifp->idev->lock);
		spin_lock(&ifp->lock);
		ifp->idev->rs_probes = 1;
L
Linus Torvalds 已提交
3423
		ifp->idev->if_flags |= IF_RS_SENT;
3424 3425 3426 3427
		addrconf_mod_rs_timer(ifp->idev,
				      ifp->idev->cnf.rtr_solicit_interval);
		spin_unlock(&ifp->lock);
		write_unlock_bh(&ifp->idev->lock);
L
Linus Torvalds 已提交
3428 3429 3430
	}
}

S
Stephen Hemminger 已提交
3431 3432
static void addrconf_dad_run(struct inet6_dev *idev)
{
3433 3434 3435
	struct inet6_ifaddr *ifp;

	read_lock_bh(&idev->lock);
3436
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
3437
		spin_lock(&ifp->lock);
H
Herbert Xu 已提交
3438 3439 3440
		if (ifp->flags & IFA_F_TENTATIVE &&
		    ifp->state == INET6_IFADDR_STATE_DAD)
			addrconf_dad_kick(ifp);
3441
		spin_unlock(&ifp->lock);
3442 3443 3444 3445
	}
	read_unlock_bh(&idev->lock);
}

L
Linus Torvalds 已提交
3446 3447
#ifdef CONFIG_PROC_FS
struct if6_iter_state {
3448
	struct seq_net_private p;
L
Linus Torvalds 已提交
3449
	int bucket;
3450
	int offset;
L
Linus Torvalds 已提交
3451 3452
};

3453
static struct inet6_ifaddr *if6_get_first(struct seq_file *seq, loff_t pos)
L
Linus Torvalds 已提交
3454 3455 3456
{
	struct inet6_ifaddr *ifa = NULL;
	struct if6_iter_state *state = seq->private;
3457
	struct net *net = seq_file_net(seq);
3458
	int p = 0;
L
Linus Torvalds 已提交
3459

3460 3461 3462 3463 3464 3465 3466
	/* initial bucket if pos is 0 */
	if (pos == 0) {
		state->bucket = 0;
		state->offset = 0;
	}

	for (; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
3467
		hlist_for_each_entry_rcu_bh(ifa, &inet6_addr_lst[state->bucket],
3468
					 addr_lst) {
3469 3470
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
3471 3472 3473 3474 3475 3476
			/* sync with offset */
			if (p < state->offset) {
				p++;
				continue;
			}
			state->offset++;
3477
			return ifa;
3478 3479 3480 3481 3482
		}

		/* prepare for next bucket */
		state->offset = 0;
		p = 0;
L
Linus Torvalds 已提交
3483
	}
3484
	return NULL;
L
Linus Torvalds 已提交
3485 3486
}

3487 3488
static struct inet6_ifaddr *if6_get_next(struct seq_file *seq,
					 struct inet6_ifaddr *ifa)
L
Linus Torvalds 已提交
3489 3490
{
	struct if6_iter_state *state = seq->private;
3491
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
3492

3493
	hlist_for_each_entry_continue_rcu_bh(ifa, addr_lst) {
3494 3495
		if (!net_eq(dev_net(ifa->idev->dev), net))
			continue;
3496
		state->offset++;
3497
		return ifa;
3498
	}
3499

3500
	while (++state->bucket < IN6_ADDR_HSIZE) {
3501
		state->offset = 0;
3502
		hlist_for_each_entry_rcu_bh(ifa,
3503
				     &inet6_addr_lst[state->bucket], addr_lst) {
3504 3505
			if (!net_eq(dev_net(ifa->idev->dev), net))
				continue;
3506
			state->offset++;
3507
			return ifa;
3508
		}
L
Linus Torvalds 已提交
3509
	}
3510

3511
	return NULL;
L
Linus Torvalds 已提交
3512 3513 3514
}

static void *if6_seq_start(struct seq_file *seq, loff_t *pos)
E
Eric Dumazet 已提交
3515
	__acquires(rcu_bh)
L
Linus Torvalds 已提交
3516
{
3517
	rcu_read_lock_bh();
3518
	return if6_get_first(seq, *pos);
L
Linus Torvalds 已提交
3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530
}

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
3531
	__releases(rcu_bh)
L
Linus Torvalds 已提交
3532
{
3533
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3534 3535 3536 3537 3538
}

static int if6_seq_show(struct seq_file *seq, void *v)
{
	struct inet6_ifaddr *ifp = (struct inet6_ifaddr *)v;
3539
	seq_printf(seq, "%pi6 %02x %02x %02x %02x %8s\n",
3540
		   &ifp->addr,
L
Linus Torvalds 已提交
3541 3542 3543
		   ifp->idev->dev->ifindex,
		   ifp->prefix_len,
		   ifp->scope,
3544
		   (u8) ifp->flags,
L
Linus Torvalds 已提交
3545 3546 3547 3548
		   ifp->idev->dev->name);
	return 0;
}

3549
static const struct seq_operations if6_seq_ops = {
L
Linus Torvalds 已提交
3550 3551 3552 3553 3554 3555 3556 3557
	.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)
{
3558 3559
	return seq_open_net(inode, file, &if6_seq_ops,
			    sizeof(struct if6_iter_state));
L
Linus Torvalds 已提交
3560 3561
}

3562
static const struct file_operations if6_fops = {
L
Linus Torvalds 已提交
3563 3564 3565 3566
	.owner		= THIS_MODULE,
	.open		= if6_seq_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
3567
	.release	= seq_release_net,
L
Linus Torvalds 已提交
3568 3569
};

3570
static int __net_init if6_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
3571
{
3572
	if (!proc_create("if_inet6", S_IRUGO, net->proc_net, &if6_fops))
L
Linus Torvalds 已提交
3573 3574 3575 3576
		return -ENOMEM;
	return 0;
}

3577
static void __net_exit if6_proc_net_exit(struct net *net)
3578
{
3579
	remove_proc_entry("if_inet6", net->proc_net);
3580 3581 3582
}

static struct pernet_operations if6_proc_net_ops = {
3583 3584
	.init = if6_proc_net_init,
	.exit = if6_proc_net_exit,
3585 3586 3587 3588 3589 3590 3591
};

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

L
Linus Torvalds 已提交
3592 3593
void if6_proc_exit(void)
{
3594
	unregister_pernet_subsys(&if6_proc_net_ops);
L
Linus Torvalds 已提交
3595 3596 3597
}
#endif	/* CONFIG_PROC_FS */

A
Amerigo Wang 已提交
3598
#if IS_ENABLED(CONFIG_IPV6_MIP6)
3599
/* Check if address is a home address configured on any interface. */
3600
int ipv6_chk_home_addr(struct net *net, const struct in6_addr *addr)
3601 3602
{
	int ret = 0;
3603
	struct inet6_ifaddr *ifp = NULL;
E
Eric Dumazet 已提交
3604
	unsigned int hash = inet6_addr_hash(addr);
3605

3606
	rcu_read_lock_bh();
3607
	hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[hash], addr_lst) {
3608
		if (!net_eq(dev_net(ifp->idev->dev), net))
3609
			continue;
3610
		if (ipv6_addr_equal(&ifp->addr, addr) &&
3611 3612 3613 3614 3615
		    (ifp->flags & IFA_F_HOMEADDRESS)) {
			ret = 1;
			break;
		}
	}
3616
	rcu_read_unlock_bh();
3617 3618 3619 3620
	return ret;
}
#endif

L
Linus Torvalds 已提交
3621 3622 3623 3624
/*
 *	Periodic address status verification
 */

3625
static void addrconf_verify_rtnl(void)
L
Linus Torvalds 已提交
3626
{
3627
	unsigned long now, next, next_sec, next_sched;
L
Linus Torvalds 已提交
3628 3629 3630
	struct inet6_ifaddr *ifp;
	int i;

3631 3632
	ASSERT_RTNL();

3633
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
3634
	now = jiffies;
3635
	next = round_jiffies_up(now + ADDR_CHECK_FREQUENCY);
L
Linus Torvalds 已提交
3636

3637
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
3638

S
Stephen Hemminger 已提交
3639
	for (i = 0; i < IN6_ADDR_HSIZE; i++) {
L
Linus Torvalds 已提交
3640
restart:
3641
		hlist_for_each_entry_rcu_bh(ifp, &inet6_addr_lst[i], addr_lst) {
L
Linus Torvalds 已提交
3642 3643
			unsigned long age;

3644 3645 3646 3647 3648 3649
			/* 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
3650 3651 3652
				continue;

			spin_lock(&ifp->lock);
3653 3654
			/* We try to batch several events at once. */
			age = (now - ifp->tstamp + ADDRCONF_TIMER_FUZZ_MINUS) / HZ;
L
Linus Torvalds 已提交
3655

3656 3657
			if (ifp->valid_lft != INFINITY_LIFE_TIME &&
			    age >= ifp->valid_lft) {
L
Linus Torvalds 已提交
3658 3659 3660 3661
				spin_unlock(&ifp->lock);
				in6_ifa_hold(ifp);
				ipv6_del_addr(ifp);
				goto restart;
3662 3663 3664
			} else if (ifp->prefered_lft == INFINITY_LIFE_TIME) {
				spin_unlock(&ifp->lock);
				continue;
L
Linus Torvalds 已提交
3665
			} else if (age >= ifp->prefered_lft) {
3666
				/* jiffies - ifp->tstamp > age >= ifp->prefered_lft */
L
Linus Torvalds 已提交
3667 3668 3669 3670 3671 3672 3673
				int deprecate = 0;

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

3674 3675
				if ((ifp->valid_lft != INFINITY_LIFE_TIME) &&
				    (time_before(ifp->tstamp + ifp->valid_lft * HZ, next)))
L
Linus Torvalds 已提交
3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688
					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)) {
3689 3690
				unsigned long regen_advance = ifp->idev->cnf.regen_max_retry *
					ifp->idev->cnf.dad_transmits *
J
Jiri Pirko 已提交
3691
					NEIGH_VAR(ifp->idev->nd_parms, RETRANS_TIME) / HZ;
3692

L
Linus Torvalds 已提交
3693 3694 3695 3696 3697 3698 3699 3700 3701
				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);
3702

3703 3704 3705
						spin_lock(&ifpub->lock);
						ifpub->regen_count = 0;
						spin_unlock(&ifpub->lock);
L
Linus Torvalds 已提交
3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722
						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);
			}
		}
	}

3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733
	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;

3734 3735
	ADBG(KERN_DEBUG "now = %lu, schedule = %lu, rounded schedule = %lu => %lu\n",
	      now, next, next_sec, next_sched);
3736
	mod_delayed_work(addrconf_wq, &addr_chk_work, next_sched - now);
3737
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3738 3739
}

3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751
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);
}

3752 3753
static struct in6_addr *extract_addr(struct nlattr *addr, struct nlattr *local,
				     struct in6_addr **peer_pfx)
3754 3755 3756
{
	struct in6_addr *pfx = NULL;

3757 3758
	*peer_pfx = NULL;

3759 3760 3761 3762 3763
	if (addr)
		pfx = nla_data(addr);

	if (local) {
		if (pfx && nla_memcmp(local, pfx, sizeof(*pfx)))
3764 3765
			*peer_pfx = pfx;
		pfx = nla_data(local);
3766 3767 3768 3769 3770
	}

	return pfx;
}

3771
static const struct nla_policy ifa_ipv6_policy[IFA_MAX+1] = {
3772 3773 3774
	[IFA_ADDRESS]		= { .len = sizeof(struct in6_addr) },
	[IFA_LOCAL]		= { .len = sizeof(struct in6_addr) },
	[IFA_CACHEINFO]		= { .len = sizeof(struct ifa_cacheinfo) },
3775
	[IFA_FLAGS]		= { .len = sizeof(u32) },
3776 3777
};

L
Linus Torvalds 已提交
3778
static int
3779
inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
3780
{
3781
	struct net *net = sock_net(skb->sk);
3782 3783
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
3784
	struct in6_addr *pfx, *peer_pfx;
3785
	u32 ifa_flags;
3786
	int err;
L
Linus Torvalds 已提交
3787

3788 3789 3790 3791 3792
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
3793
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
L
Linus Torvalds 已提交
3794 3795 3796
	if (pfx == NULL)
		return -EINVAL;

3797 3798 3799 3800 3801 3802 3803
	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
3804 3805
}

3806
static int inet6_addr_modify(struct inet6_ifaddr *ifp, u32 ifa_flags,
3807
			     u32 prefered_lft, u32 valid_lft)
3808
{
3809 3810
	u32 flags;
	clock_t expires;
3811
	unsigned long timeout;
3812
	bool was_managetempaddr;
3813
	bool had_prefixroute;
3814

3815 3816
	ASSERT_RTNL();

3817 3818 3819
	if (!valid_lft || (prefered_lft > valid_lft))
		return -EINVAL;

3820 3821 3822 3823
	if (ifa_flags & IFA_F_MANAGETEMPADDR &&
	    (ifp->flags & IFA_F_TEMPORARY || ifp->prefix_len != 64))
		return -EINVAL;

3824 3825 3826 3827
	timeout = addrconf_timeout_fixup(valid_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		expires = jiffies_to_clock_t(timeout * HZ);
		valid_lft = timeout;
3828
		flags = RTF_EXPIRES;
3829 3830 3831 3832
	} else {
		expires = 0;
		flags = 0;
		ifa_flags |= IFA_F_PERMANENT;
3833
	}
3834

3835 3836 3837 3838 3839 3840
	timeout = addrconf_timeout_fixup(prefered_lft, HZ);
	if (addrconf_finite_timeout(timeout)) {
		if (timeout == 0)
			ifa_flags |= IFA_F_DEPRECATED;
		prefered_lft = timeout;
	}
3841 3842

	spin_lock_bh(&ifp->lock);
3843
	was_managetempaddr = ifp->flags & IFA_F_MANAGETEMPADDR;
3844 3845
	had_prefixroute = ifp->flags & IFA_F_PERMANENT &&
			  !(ifp->flags & IFA_F_NOPREFIXROUTE);
3846
	ifp->flags &= ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD |
3847 3848
			IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
			IFA_F_NOPREFIXROUTE);
3849
	ifp->flags |= ifa_flags;
3850 3851 3852 3853 3854 3855 3856 3857
	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);

3858 3859 3860
	if (!(ifa_flags & IFA_F_NOPREFIXROUTE)) {
		addrconf_prefix_route(&ifp->addr, ifp->prefix_len, ifp->idev->dev,
				      expires, flags);
3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872
	} 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);
		}
3873
	}
3874 3875 3876 3877 3878 3879 3880 3881

	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);
	}

3882
	addrconf_verify_rtnl();
3883 3884 3885 3886

	return 0;
}

L
Linus Torvalds 已提交
3887
static int
3888
inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh)
L
Linus Torvalds 已提交
3889
{
3890
	struct net *net = sock_net(skb->sk);
3891 3892
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
3893
	struct in6_addr *pfx, *peer_pfx;
3894 3895
	struct inet6_ifaddr *ifa;
	struct net_device *dev;
3896
	u32 valid_lft = INFINITY_LIFE_TIME, preferred_lft = INFINITY_LIFE_TIME;
3897
	u32 ifa_flags;
3898
	int err;
L
Linus Torvalds 已提交
3899

3900 3901 3902 3903 3904
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		return err;

	ifm = nlmsg_data(nlh);
3905
	pfx = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer_pfx);
L
Linus Torvalds 已提交
3906 3907 3908
	if (pfx == NULL)
		return -EINVAL;

3909
	if (tb[IFA_CACHEINFO]) {
3910
		struct ifa_cacheinfo *ci;
3911 3912

		ci = nla_data(tb[IFA_CACHEINFO]);
3913
		valid_lft = ci->ifa_valid;
3914 3915 3916 3917
		preferred_lft = ci->ifa_prefered;
	} else {
		preferred_lft = INFINITY_LIFE_TIME;
		valid_lft = INFINITY_LIFE_TIME;
3918 3919
	}

3920
	dev =  __dev_get_by_index(net, ifm->ifa_index);
3921 3922 3923
	if (dev == NULL)
		return -ENODEV;

3924 3925
	ifa_flags = tb[IFA_FLAGS] ? nla_get_u32(tb[IFA_FLAGS]) : ifm->ifa_flags;

3926
	/* We ignore other flags so far. */
3927 3928
	ifa_flags &= IFA_F_NODAD | IFA_F_HOMEADDRESS | IFA_F_MANAGETEMPADDR |
		     IFA_F_NOPREFIXROUTE;
3929

3930
	ifa = ipv6_get_ifaddr(net, pfx, dev, 1);
3931 3932 3933
	if (ifa == NULL) {
		/*
		 * It would be best to check for !NLM_F_CREATE here but
3934
		 * userspace already relies on not having to provide this.
3935
		 */
3936
		return inet6_addr_add(net, ifm->ifa_index, pfx, peer_pfx,
3937 3938
				      ifm->ifa_prefixlen, ifa_flags,
				      preferred_lft, valid_lft);
3939 3940
	}

3941 3942 3943 3944
	if (nlh->nlmsg_flags & NLM_F_EXCL ||
	    !(nlh->nlmsg_flags & NLM_F_REPLACE))
		err = -EEXIST;
	else
3945
		err = inet6_addr_modify(ifa, ifa_flags, preferred_lft, valid_lft);
3946 3947 3948 3949

	in6_ifa_put(ifa);

	return err;
L
Linus Torvalds 已提交
3950 3951
}

3952
static void put_ifaddrmsg(struct nlmsghdr *nlh, u8 prefixlen, u32 flags,
3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964
			  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;
}

3965 3966 3967 3968 3969
static int put_cacheinfo(struct sk_buff *skb, unsigned long cstamp,
			 unsigned long tstamp, u32 preferred, u32 valid)
{
	struct ifa_cacheinfo ci;

3970 3971
	ci.cstamp = cstamp_delta(cstamp);
	ci.tstamp = cstamp_delta(tstamp);
3972 3973 3974 3975 3976 3977
	ci.ifa_prefered = preferred;
	ci.ifa_valid = valid;

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

3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989
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;
}

3990 3991
static inline int inet6_ifaddr_msgsize(void)
{
3992
	return NLMSG_ALIGN(sizeof(struct ifaddrmsg))
3993
	       + nla_total_size(16) /* IFA_LOCAL */
3994
	       + nla_total_size(16) /* IFA_ADDRESS */
3995 3996
	       + nla_total_size(sizeof(struct ifa_cacheinfo))
	       + nla_total_size(4)  /* IFA_FLAGS */;
3997
}
3998

L
Linus Torvalds 已提交
3999
static int inet6_fill_ifaddr(struct sk_buff *skb, struct inet6_ifaddr *ifa,
4000
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4001 4002
{
	struct nlmsghdr  *nlh;
4003
	u32 preferred, valid;
L
Linus Torvalds 已提交
4004

4005
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4006
	if (nlh == NULL)
4007
		return -EMSGSIZE;
4008

4009 4010 4011
	put_ifaddrmsg(nlh, ifa->prefix_len, ifa->flags, rt_scope(ifa->scope),
		      ifa->idev->dev->ifindex);

4012 4013
	if (!((ifa->flags&IFA_F_PERMANENT) &&
	      (ifa->prefered_lft == INFINITY_LIFE_TIME))) {
4014 4015 4016
		preferred = ifa->prefered_lft;
		valid = ifa->valid_lft;
		if (preferred != INFINITY_LIFE_TIME) {
L
Linus Torvalds 已提交
4017
			long tval = (jiffies - ifa->tstamp)/HZ;
4018 4019 4020 4021
			if (preferred > tval)
				preferred -= tval;
			else
				preferred = 0;
4022 4023 4024 4025 4026 4027
			if (valid != INFINITY_LIFE_TIME) {
				if (valid > tval)
					valid -= tval;
				else
					valid = 0;
			}
L
Linus Torvalds 已提交
4028 4029
		}
	} else {
4030 4031 4032 4033
		preferred = INFINITY_LIFE_TIME;
		valid = INFINITY_LIFE_TIME;
	}

C
Cong Wang 已提交
4034
	if (!ipv6_addr_any(&ifa->peer_addr)) {
4035 4036 4037 4038 4039 4040 4041 4042 4043
		if (nla_put(skb, IFA_LOCAL, 16, &ifa->addr) < 0 ||
		    nla_put(skb, IFA_ADDRESS, 16, &ifa->peer_addr) < 0)
			goto error;
	} else
		if (nla_put(skb, IFA_ADDRESS, 16, &ifa->addr) < 0)
			goto error;

	if (put_cacheinfo(skb, ifa->cstamp, ifa->tstamp, preferred, valid) < 0)
		goto error;
L
Linus Torvalds 已提交
4044

4045 4046 4047
	if (nla_put_u32(skb, IFA_FLAGS, ifa->flags) < 0)
		goto error;

4048
	return nlmsg_end(skb, nlh);
4049 4050 4051 4052

error:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4053 4054 4055
}

static int inet6_fill_ifmcaddr(struct sk_buff *skb, struct ifmcaddr6 *ifmca,
4056
				u32 portid, u32 seq, int event, u16 flags)
L
Linus Torvalds 已提交
4057 4058
{
	struct nlmsghdr  *nlh;
4059 4060 4061 4062 4063
	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
4064

4065
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4066
	if (nlh == NULL)
4067
		return -EMSGSIZE;
4068

4069 4070 4071
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
	if (nla_put(skb, IFA_MULTICAST, 16, &ifmca->mca_addr) < 0 ||
	    put_cacheinfo(skb, ifmca->mca_cstamp, ifmca->mca_tstamp,
4072 4073 4074 4075
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
4076

4077
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
4078 4079 4080
}

static int inet6_fill_ifacaddr(struct sk_buff *skb, struct ifacaddr6 *ifaca,
4081
				u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4082 4083
{
	struct nlmsghdr  *nlh;
4084 4085 4086 4087 4088
	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
4089

4090
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct ifaddrmsg), flags);
4091
	if (nlh == NULL)
4092
		return -EMSGSIZE;
4093

4094 4095 4096
	put_ifaddrmsg(nlh, 128, IFA_F_PERMANENT, scope, ifindex);
	if (nla_put(skb, IFA_ANYCAST, 16, &ifaca->aca_addr) < 0 ||
	    put_cacheinfo(skb, ifaca->aca_cstamp, ifaca->aca_tstamp,
4097 4098 4099 4100
			  INFINITY_LIFE_TIME, INFINITY_LIFE_TIME) < 0) {
		nlmsg_cancel(skb, nlh);
		return -EMSGSIZE;
	}
L
Linus Torvalds 已提交
4101

4102
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
4103 4104
}

S
Stephen Hemminger 已提交
4105
enum addr_type_t {
L
Linus Torvalds 已提交
4106 4107 4108 4109 4110
	UNICAST_ADDR,
	MULTICAST_ADDR,
	ANYCAST_ADDR,
};

E
Eric Dumazet 已提交
4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122
/* 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) {
4123 4124 4125
	case UNICAST_ADDR: {
		struct inet6_ifaddr *ifa;

E
Eric Dumazet 已提交
4126
		/* unicast address incl. temp addr */
4127 4128
		list_for_each_entry(ifa, &idev->addr_list, if_list) {
			if (++ip_idx < s_ip_idx)
E
Eric Dumazet 已提交
4129 4130
				continue;
			err = inet6_fill_ifaddr(skb, ifa,
4131
						NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4132 4133 4134 4135 4136
						cb->nlh->nlmsg_seq,
						RTM_NEWADDR,
						NLM_F_MULTI);
			if (err <= 0)
				break;
4137
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
E
Eric Dumazet 已提交
4138 4139
		}
		break;
4140
	}
E
Eric Dumazet 已提交
4141 4142 4143 4144 4145 4146 4147
	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,
4148
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162
						  cb->nlh->nlmsg_seq,
						  RTM_GETMULTICAST,
						  NLM_F_MULTI);
			if (err <= 0)
				break;
		}
		break;
	case ANYCAST_ADDR:
		/* anycast address */
		for (ifaca = idev->ac_list; ifaca;
		     ifaca = ifaca->aca_next, ip_idx++) {
			if (ip_idx < s_ip_idx)
				continue;
			err = inet6_fill_ifacaddr(skb, ifaca,
4163
						  NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178
						  cb->nlh->nlmsg_seq,
						  RTM_GETANYCAST,
						  NLM_F_MULTI);
			if (err <= 0)
				break;
		}
		break;
	default:
		break;
	}
	read_unlock_bh(&idev->lock);
	*p_ip_idx = ip_idx;
	return err;
}

L
Linus Torvalds 已提交
4179 4180 4181
static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
			   enum addr_type_t type)
{
E
Eric Dumazet 已提交
4182 4183
	struct net *net = sock_net(skb->sk);
	int h, s_h;
L
Linus Torvalds 已提交
4184 4185 4186
	int idx, ip_idx;
	int s_idx, s_ip_idx;
	struct net_device *dev;
E
Eric Dumazet 已提交
4187 4188
	struct inet6_dev *idev;
	struct hlist_head *head;
4189

E
Eric Dumazet 已提交
4190 4191 4192
	s_h = cb->args[0];
	s_idx = idx = cb->args[1];
	s_ip_idx = ip_idx = cb->args[2];
4193

E
Eric Dumazet 已提交
4194
	rcu_read_lock();
4195
	cb->seq = atomic_read(&net->ipv6.dev_addr_genid) ^ net->dev_base_seq;
E
Eric Dumazet 已提交
4196 4197 4198
	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
4199
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
4200 4201
			if (idx < s_idx)
				goto cont;
4202
			if (h > s_h || idx > s_idx)
E
Eric Dumazet 已提交
4203 4204
				s_ip_idx = 0;
			ip_idx = 0;
S
Stephen Hemminger 已提交
4205 4206
			idev = __in6_dev_get(dev);
			if (!idev)
E
Eric Dumazet 已提交
4207 4208 4209 4210 4211
				goto cont;

			if (in6_dump_addrs(idev, skb, cb, type,
					   s_ip_idx, &ip_idx) <= 0)
				goto done;
4212
cont:
E
Eric Dumazet 已提交
4213 4214
			idx++;
		}
L
Linus Torvalds 已提交
4215
	}
E
Eric Dumazet 已提交
4216 4217 4218 4219 4220 4221
done:
	rcu_read_unlock();
	cb->args[0] = h;
	cb->args[1] = idx;
	cb->args[2] = ip_idx;

L
Linus Torvalds 已提交
4222 4223 4224 4225 4226 4227
	return skb->len;
}

static int inet6_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
	enum addr_type_t type = UNICAST_ADDR;
4228

L
Linus Torvalds 已提交
4229 4230 4231 4232 4233 4234
	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;
4235

L
Linus Torvalds 已提交
4236 4237 4238 4239 4240 4241 4242
	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;
4243

L
Linus Torvalds 已提交
4244 4245 4246
	return inet6_dump_addr(skb, cb, type);
}

4247
static int inet6_rtm_getaddr(struct sk_buff *in_skb, struct nlmsghdr *nlh)
4248
{
4249
	struct net *net = sock_net(in_skb->sk);
4250 4251
	struct ifaddrmsg *ifm;
	struct nlattr *tb[IFA_MAX+1];
4252
	struct in6_addr *addr = NULL, *peer;
4253 4254 4255 4256 4257
	struct net_device *dev = NULL;
	struct inet6_ifaddr *ifa;
	struct sk_buff *skb;
	int err;

4258 4259 4260 4261
	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
	if (err < 0)
		goto errout;

4262
	addr = extract_addr(tb[IFA_ADDRESS], tb[IFA_LOCAL], &peer);
4263 4264 4265
	if (addr == NULL) {
		err = -EINVAL;
		goto errout;
4266 4267
	}

4268
	ifm = nlmsg_data(nlh);
4269
	if (ifm->ifa_index)
4270
		dev = __dev_get_by_index(net, ifm->ifa_index);
4271

S
Stephen Hemminger 已提交
4272 4273
	ifa = ipv6_get_ifaddr(net, addr, dev, 1);
	if (!ifa) {
4274 4275 4276
		err = -EADDRNOTAVAIL;
		goto errout;
	}
4277

S
Stephen Hemminger 已提交
4278 4279
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_KERNEL);
	if (!skb) {
4280
		err = -ENOBUFS;
4281
		goto errout_ifa;
4282 4283
	}

4284
	err = inet6_fill_ifaddr(skb, ifa, NETLINK_CB(in_skb).portid,
4285
				nlh->nlmsg_seq, RTM_NEWADDR, 0);
4286 4287 4288 4289 4290 4291
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout_ifa;
	}
4292
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
4293
errout_ifa:
4294
	in6_ifa_put(ifa);
4295
errout:
4296 4297 4298
	return err;
}

L
Linus Torvalds 已提交
4299 4300 4301
static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa)
{
	struct sk_buff *skb;
4302
	struct net *net = dev_net(ifa->idev->dev);
4303
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
4304

4305
	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_ATOMIC);
4306 4307 4308 4309
	if (skb == NULL)
		goto errout;

	err = inet6_fill_ifaddr(skb, ifa, 0, 0, event, 0);
4310 4311 4312 4313 4314 4315
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_ifaddr_msgsize() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4316 4317
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFADDR, NULL, GFP_ATOMIC);
	return;
4318 4319
errout:
	if (err < 0)
4320
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFADDR, err);
L
Linus Torvalds 已提交
4321 4322
}

D
Dave Jones 已提交
4323
static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
L
Linus Torvalds 已提交
4324 4325
				__s32 *array, int bytes)
{
4326 4327
	BUG_ON(bytes < (DEVCONF_MAX * 4));

L
Linus Torvalds 已提交
4328 4329 4330 4331 4332 4333 4334 4335 4336
	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;
4337 4338 4339 4340
	array[DEVCONF_RTR_SOLICIT_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_solicit_interval);
	array[DEVCONF_RTR_SOLICIT_DELAY] =
		jiffies_to_msecs(cnf->rtr_solicit_delay);
L
Linus Torvalds 已提交
4341
	array[DEVCONF_FORCE_MLD_VERSION] = cnf->force_mld_version;
4342 4343 4344 4345
	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
4346 4347 4348 4349 4350 4351
	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;
4352
	array[DEVCONF_ACCEPT_RA_DEFRTR] = cnf->accept_ra_defrtr;
4353
	array[DEVCONF_ACCEPT_RA_PINFO] = cnf->accept_ra_pinfo;
4354 4355
#ifdef CONFIG_IPV6_ROUTER_PREF
	array[DEVCONF_ACCEPT_RA_RTR_PREF] = cnf->accept_ra_rtr_pref;
4356 4357
	array[DEVCONF_RTR_PROBE_INTERVAL] =
		jiffies_to_msecs(cnf->rtr_probe_interval);
N
Neil Horman 已提交
4358
#ifdef CONFIG_IPV6_ROUTE_INFO
4359 4360
	array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
#endif
4361
#endif
4362
	array[DEVCONF_PROXY_NDP] = cnf->proxy_ndp;
4363
	array[DEVCONF_ACCEPT_SOURCE_ROUTE] = cnf->accept_source_route;
4364 4365
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
4366
	array[DEVCONF_USE_OPTIMISTIC] = cnf->use_optimistic;
4367
#endif
4368 4369 4370
#ifdef CONFIG_IPV6_MROUTE
	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
#endif
4371
	array[DEVCONF_DISABLE_IPV6] = cnf->disable_ipv6;
4372
	array[DEVCONF_ACCEPT_DAD] = cnf->accept_dad;
4373
	array[DEVCONF_FORCE_TLLAO] = cnf->force_tllao;
4374
	array[DEVCONF_NDISC_NOTIFY] = cnf->ndisc_notify;
4375
	array[DEVCONF_SUPPRESS_FRAG_NDISC] = cnf->suppress_frag_ndisc;
4376
	array[DEVCONF_ACCEPT_RA_FROM_LOCAL] = cnf->accept_ra_from_local;
L
Linus Torvalds 已提交
4377 4378
}

T
Thomas Graf 已提交
4379 4380 4381 4382 4383 4384
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 */
4385 4386
	     + 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
4387 4388
}

4389 4390 4391 4392 4393 4394 4395
static inline size_t inet6_if_nlmsg_size(void)
{
	return NLMSG_ALIGN(sizeof(struct ifinfomsg))
	       + nla_total_size(IFNAMSIZ) /* IFLA_IFNAME */
	       + nla_total_size(MAX_ADDR_LEN) /* IFLA_ADDRESS */
	       + nla_total_size(4) /* IFLA_MTU */
	       + nla_total_size(4) /* IFLA_LINK */
T
Thomas Graf 已提交
4396
	       + nla_total_size(inet6_ifla6_size()); /* IFLA_PROTINFO */
4397
}
4398

4399
static inline void __snmp6_fill_statsdev(u64 *stats, atomic_long_t *mib,
4400
				      int items, int bytes)
4401 4402 4403 4404 4405 4406 4407 4408
{
	int i;
	int pad = bytes - sizeof(u64) * items;
	BUG_ON(pad < 0);

	/* Use put_unaligned() because stats may not be aligned for u64. */
	put_unaligned(items, &stats[0]);
	for (i = 1; i < items; i++)
4409
		put_unaligned(atomic_long_read(&mib[i]), &stats[i]);
4410 4411 4412 4413

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

W
WANG Cong 已提交
4414
static inline void __snmp6_fill_stats64(u64 *stats, void __percpu *mib,
4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428
				      int items, int bytes, size_t syncpoff)
{
	int i;
	int pad = bytes - sizeof(u64) * items;
	BUG_ON(pad < 0);

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

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

4429 4430 4431
static void snmp6_fill_stats(u64 *stats, struct inet6_dev *idev, int attrtype,
			     int bytes)
{
S
Stephen Hemminger 已提交
4432
	switch (attrtype) {
4433
	case IFLA_INET6_STATS:
W
WANG Cong 已提交
4434
		__snmp6_fill_stats64(stats, idev->stats.ipv6,
4435
				     IPSTATS_MIB_MAX, bytes, offsetof(struct ipstats_mib, syncp));
4436 4437
		break;
	case IFLA_INET6_ICMP6STATS:
4438
		__snmp6_fill_statsdev(stats, idev->stats.icmpv6dev->mibs, ICMP6_MIB_MAX, bytes);
4439 4440 4441 4442
		break;
	}
}

T
Thomas Graf 已提交
4443 4444 4445 4446 4447
static int inet6_fill_ifla6_attrs(struct sk_buff *skb, struct inet6_dev *idev)
{
	struct nlattr *nla;
	struct ifla_cacheinfo ci;

D
David S. Miller 已提交
4448 4449
	if (nla_put_u32(skb, IFLA_INET6_FLAGS, idev->if_flags))
		goto nla_put_failure;
T
Thomas Graf 已提交
4450
	ci.max_reasm_len = IPV6_MAXPLEN;
4451 4452
	ci.tstamp = cstamp_delta(idev->tstamp);
	ci.reachable_time = jiffies_to_msecs(idev->nd_parms->reachable_time);
J
Jiri Pirko 已提交
4453
	ci.retrans_time = jiffies_to_msecs(NEIGH_VAR(idev->nd_parms, RETRANS_TIME));
D
David S. Miller 已提交
4454 4455
	if (nla_put(skb, IFLA_INET6_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
T
Thomas Graf 已提交
4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472
	nla = nla_reserve(skb, IFLA_INET6_CONF, DEVCONF_MAX * sizeof(s32));
	if (nla == NULL)
		goto nla_put_failure;
	ipv6_store_devconf(&idev->cnf, nla_data(nla), nla_len(nla));

	/* XXX - MC not implemented */

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

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

4473 4474 4475
	nla = nla_reserve(skb, IFLA_INET6_TOKEN, sizeof(struct in6_addr));
	if (nla == NULL)
		goto nla_put_failure;
4476 4477 4478 4479

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

4480 4481 4482 4483
	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
4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510
	return 0;

nla_put_failure:
	return -EMSGSIZE;
}

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

	return inet6_ifla6_size();
}

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

	if (!idev)
		return -ENODATA;

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

	return 0;
}

4511 4512 4513 4514
static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token)
{
	struct inet6_ifaddr *ifp;
	struct net_device *dev = idev->dev;
4515
	bool update_rs = false;
4516
	struct in6_addr ll_addr;
4517

4518 4519
	ASSERT_RTNL();

4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537
	if (token == NULL)
		return -EINVAL;
	if (ipv6_addr_any(token))
		return -EINVAL;
	if (dev->flags & (IFF_LOOPBACK | IFF_NOARP))
		return -EINVAL;
	if (!ipv6_accept_ra(idev))
		return -EINVAL;
	if (idev->cnf.rtr_solicits <= 0)
		return -EINVAL;

	write_lock_bh(&idev->lock);

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

	write_unlock_bh(&idev->lock);

4538 4539 4540
	if (!idev->dead && (idev->if_flags & IF_READY) &&
	    !ipv6_get_lladdr(dev, &ll_addr, IFA_F_TENTATIVE |
			     IFA_F_OPTIMISTIC)) {
4541 4542 4543 4544 4545 4546 4547

		/* 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;
	}
4548 4549

	write_lock_bh(&idev->lock);
4550

4551
	if (update_rs) {
4552
		idev->if_flags |= IF_RS_SENT;
4553 4554 4555
		idev->rs_probes = 1;
		addrconf_mod_rs_timer(idev, idev->cnf.rtr_solicit_interval);
	}
4556 4557 4558 4559

	/* Well, that's kinda nasty ... */
	list_for_each_entry(ifp, &idev->addr_list, if_list) {
		spin_lock(&ifp->lock);
4560
		if (ifp->tokenized) {
4561 4562 4563 4564 4565 4566 4567
			ifp->valid_lft = 0;
			ifp->prefered_lft = 0;
		}
		spin_unlock(&ifp->lock);
	}

	write_unlock_bh(&idev->lock);
4568
	inet6_ifinfo_notify(RTM_NEWLINK, idev);
4569
	addrconf_verify_rtnl();
4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584
	return 0;
}

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

	if (!idev)
		return -EAFNOSUPPORT;

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

4585
	if (tb[IFLA_INET6_TOKEN]) {
4586
		err = inet6_set_iftoken(idev, nla_data(tb[IFLA_INET6_TOKEN]));
4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599
		if (err)
			return err;
	}

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

		if (mode != IN6_ADDR_GEN_MODE_EUI64 &&
		    mode != IN6_ADDR_GEN_MODE_NONE)
			return -EINVAL;
		idev->addr_gen_mode = mode;
		err = 0;
	}
4600 4601 4602 4603

	return err;
}

4604
static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
4605
			     u32 portid, u32 seq, int event, unsigned int flags)
L
Linus Torvalds 已提交
4606
{
4607 4608 4609 4610 4611
	struct net_device *dev = idev->dev;
	struct ifinfomsg *hdr;
	struct nlmsghdr *nlh;
	void *protoinfo;

4612
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*hdr), flags);
4613
	if (nlh == NULL)
4614
		return -EMSGSIZE;
4615 4616 4617 4618 4619 4620 4621 4622 4623

	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
4624 4625 4626 4627 4628 4629 4630
	if (nla_put_string(skb, IFLA_IFNAME, dev->name) ||
	    (dev->addr_len &&
	     nla_put(skb, IFLA_ADDRESS, dev->addr_len, dev->dev_addr)) ||
	    nla_put_u32(skb, IFLA_MTU, dev->mtu) ||
	    (dev->ifindex != dev->iflink &&
	     nla_put_u32(skb, IFLA_LINK, dev->iflink)))
		goto nla_put_failure;
4631 4632 4633
	protoinfo = nla_nest_start(skb, IFLA_PROTINFO);
	if (protoinfo == NULL)
		goto nla_put_failure;
L
Linus Torvalds 已提交
4634

T
Thomas Graf 已提交
4635
	if (inet6_fill_ifla6_attrs(skb, idev) < 0)
4636
		goto nla_put_failure;
L
Linus Torvalds 已提交
4637

4638 4639
	nla_nest_end(skb, protoinfo);
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
4640

4641
nla_put_failure:
4642 4643
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4644 4645 4646 4647
}

static int inet6_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
{
4648
	struct net *net = sock_net(skb->sk);
E
Eric Dumazet 已提交
4649
	int h, s_h;
4650
	int idx = 0, s_idx;
L
Linus Torvalds 已提交
4651 4652
	struct net_device *dev;
	struct inet6_dev *idev;
E
Eric Dumazet 已提交
4653
	struct hlist_head *head;
L
Linus Torvalds 已提交
4654

E
Eric Dumazet 已提交
4655 4656 4657 4658 4659 4660 4661
	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];
4662
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
E
Eric Dumazet 已提交
4663 4664 4665 4666 4667 4668
			if (idx < s_idx)
				goto cont;
			idev = __in6_dev_get(dev);
			if (!idev)
				goto cont;
			if (inet6_fill_ifinfo(skb, idev,
4669
					      NETLINK_CB(cb->skb).portid,
E
Eric Dumazet 已提交
4670 4671 4672
					      cb->nlh->nlmsg_seq,
					      RTM_NEWLINK, NLM_F_MULTI) <= 0)
				goto out;
4673
cont:
E
Eric Dumazet 已提交
4674 4675
			idx++;
		}
L
Linus Torvalds 已提交
4676
	}
E
Eric Dumazet 已提交
4677 4678 4679 4680
out:
	rcu_read_unlock();
	cb->args[1] = idx;
	cb->args[0] = h;
L
Linus Torvalds 已提交
4681 4682 4683 4684 4685 4686 4687

	return skb->len;
}

void inet6_ifinfo_notify(int event, struct inet6_dev *idev)
{
	struct sk_buff *skb;
4688
	struct net *net = dev_net(idev->dev);
4689
	int err = -ENOBUFS;
4690

4691
	skb = nlmsg_new(inet6_if_nlmsg_size(), GFP_ATOMIC);
4692 4693 4694 4695
	if (skb == NULL)
		goto errout;

	err = inet6_fill_ifinfo(skb, idev, 0, 0, event, 0);
4696 4697 4698 4699 4700 4701
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_if_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4702
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_IFINFO, NULL, GFP_ATOMIC);
4703
	return;
4704 4705
errout:
	if (err < 0)
4706
		rtnl_set_sk_err(net, RTNLGRP_IPV6_IFINFO, err);
L
Linus Torvalds 已提交
4707 4708
}

4709 4710 4711 4712 4713 4714
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));
}
4715

L
Linus Torvalds 已提交
4716
static int inet6_fill_prefix(struct sk_buff *skb, struct inet6_dev *idev,
4717
			     struct prefix_info *pinfo, u32 portid, u32 seq,
4718
			     int event, unsigned int flags)
L
Linus Torvalds 已提交
4719
{
4720 4721
	struct prefixmsg *pmsg;
	struct nlmsghdr *nlh;
L
Linus Torvalds 已提交
4722 4723
	struct prefix_cacheinfo	ci;

4724
	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*pmsg), flags);
4725
	if (nlh == NULL)
4726
		return -EMSGSIZE;
4727 4728

	pmsg = nlmsg_data(nlh);
L
Linus Torvalds 已提交
4729
	pmsg->prefix_family = AF_INET6;
4730 4731
	pmsg->prefix_pad1 = 0;
	pmsg->prefix_pad2 = 0;
L
Linus Torvalds 已提交
4732 4733 4734
	pmsg->prefix_ifindex = idev->dev->ifindex;
	pmsg->prefix_len = pinfo->prefix_len;
	pmsg->prefix_type = pinfo->type;
4735
	pmsg->prefix_pad3 = 0;
L
Linus Torvalds 已提交
4736 4737 4738 4739 4740 4741
	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
4742 4743
	if (nla_put(skb, PREFIX_ADDRESS, sizeof(pinfo->prefix), &pinfo->prefix))
		goto nla_put_failure;
L
Linus Torvalds 已提交
4744 4745
	ci.preferred_time = ntohl(pinfo->prefered);
	ci.valid_time = ntohl(pinfo->valid);
D
David S. Miller 已提交
4746 4747
	if (nla_put(skb, PREFIX_CACHEINFO, sizeof(ci), &ci))
		goto nla_put_failure;
4748
	return nlmsg_end(skb, nlh);
L
Linus Torvalds 已提交
4749

4750
nla_put_failure:
4751 4752
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
L
Linus Torvalds 已提交
4753 4754
}

4755
static void inet6_prefix_notify(int event, struct inet6_dev *idev,
L
Linus Torvalds 已提交
4756 4757 4758
			 struct prefix_info *pinfo)
{
	struct sk_buff *skb;
4759
	struct net *net = dev_net(idev->dev);
4760
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
4761

4762
	skb = nlmsg_new(inet6_prefix_nlmsg_size(), GFP_ATOMIC);
4763 4764 4765 4766
	if (skb == NULL)
		goto errout;

	err = inet6_fill_prefix(skb, idev, pinfo, 0, 0, event, 0);
4767 4768 4769 4770 4771 4772
	if (err < 0) {
		/* -EMSGSIZE implies BUG in inet6_prefix_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
4773 4774
	rtnl_notify(skb, net, 0, RTNLGRP_IPV6_PREFIX, NULL, GFP_ATOMIC);
	return;
4775 4776
errout:
	if (err < 0)
4777
		rtnl_set_sk_err(net, RTNLGRP_IPV6_PREFIX, err);
L
Linus Torvalds 已提交
4778 4779 4780 4781
}

static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
4782 4783
	struct net *net = dev_net(ifp->idev->dev);

4784 4785 4786
	if (event)
		ASSERT_RTNL();

L
Linus Torvalds 已提交
4787 4788 4789 4790
	inet6_ifa_notify(event ? : RTM_NEWADDR, ifp);

	switch (event) {
	case RTM_NEWADDR:
4791 4792 4793 4794 4795 4796 4797 4798
		/*
		 * 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
4799 4800
		if (ifp->idev->cnf.forwarding)
			addrconf_join_anycast(ifp);
C
Cong Wang 已提交
4801
		if (!ipv6_addr_any(&ifp->peer_addr))
4802 4803
			addrconf_prefix_route(&ifp->peer_addr, 128,
					      ifp->idev->dev, 0, 0);
L
Linus Torvalds 已提交
4804 4805 4806 4807 4808
		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
4809
		if (!ipv6_addr_any(&ifp->peer_addr)) {
4810 4811
			struct rt6_info *rt;

4812 4813
			rt = addrconf_get_prefix_route(&ifp->peer_addr, 128,
						       ifp->idev->dev, 0, 0);
4814 4815
			if (rt && ip6_del_rt(rt))
				dst_free(&rt->dst);
4816
		}
4817
		dst_hold(&ifp->rt->dst);
4818

4819
		if (ip6_del_rt(ifp->rt))
4820
			dst_free(&ifp->rt->dst);
H
Hannes Frederic Sowa 已提交
4821 4822

		rt_genid_bump_ipv6(net);
L
Linus Torvalds 已提交
4823 4824
		break;
	}
4825
	atomic_inc(&net->ipv6.dev_addr_genid);
L
Linus Torvalds 已提交
4826 4827 4828 4829
}

static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
{
4830
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
4831 4832
	if (likely(ifp->idev->dead == 0))
		__ipv6_ifa_notify(event, ifp);
4833
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
4834 4835 4836 4837 4838
}

#ifdef CONFIG_SYSCTL

static
4839
int addrconf_sysctl_forward(struct ctl_table *ctl, int write,
L
Linus Torvalds 已提交
4840 4841 4842 4843
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
4844
	loff_t pos = *ppos;
4845
	struct ctl_table lctl;
L
Linus Torvalds 已提交
4846 4847
	int ret;

4848 4849 4850 4851 4852 4853 4854 4855
	/*
	 * 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
4856

4857
	if (write)
4858
		ret = addrconf_fixup_forwarding(ctl, valp, val);
E
Eric W. Biederman 已提交
4859 4860
	if (ret)
		*ppos = pos;
4861
	return ret;
L
Linus Torvalds 已提交
4862 4863
}

4864 4865
static void dev_disable_change(struct inet6_dev *idev)
{
4866 4867
	struct netdev_notifier_info info;

4868 4869 4870
	if (!idev || !idev->dev)
		return;

4871
	netdev_notifier_info_init(&info, idev->dev);
4872
	if (idev->cnf.disable_ipv6)
4873
		addrconf_notify(NULL, NETDEV_DOWN, &info);
4874
	else
4875
		addrconf_notify(NULL, NETDEV_UP, &info);
4876 4877 4878 4879 4880 4881 4882
}

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

4883 4884
	rcu_read_lock();
	for_each_netdev_rcu(net, dev) {
4885 4886 4887 4888 4889 4890 4891 4892
		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);
		}
	}
4893
	rcu_read_unlock();
4894 4895
}

4896
static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
4897 4898
{
	struct net *net;
4899 4900 4901 4902
	int old;

	if (!rtnl_trylock())
		return restart_syscall();
4903 4904

	net = (struct net *)table->extra2;
4905 4906
	old = *p;
	*p = newf;
4907

4908 4909
	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
		rtnl_unlock();
4910
		return 0;
E
Eric W. Biederman 已提交
4911
	}
4912 4913 4914 4915

	if (p == &net->ipv6.devconf_all->disable_ipv6) {
		net->ipv6.devconf_dflt->disable_ipv6 = newf;
		addrconf_disable_change(net, newf);
4916
	} else if ((!newf) ^ (!old))
4917 4918 4919 4920 4921 4922 4923
		dev_disable_change((struct inet6_dev *)table->extra1);

	rtnl_unlock();
	return 0;
}

static
4924
int addrconf_sysctl_disable(struct ctl_table *ctl, int write,
4925 4926 4927 4928
			    void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int val = *valp;
E
Eric W. Biederman 已提交
4929
	loff_t pos = *ppos;
4930
	struct ctl_table lctl;
4931 4932
	int ret;

4933 4934 4935 4936 4937 4938 4939 4940
	/*
	 * 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);
4941 4942 4943

	if (write)
		ret = addrconf_disable_ipv6(ctl, valp, val);
E
Eric W. Biederman 已提交
4944 4945
	if (ret)
		*ppos = pos;
4946 4947 4948
	return ret;
}

4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988
static
int addrconf_sysctl_proxy_ndp(struct ctl_table *ctl, int write,
			      void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = ctl->data;
	int ret;
	int old, new;

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

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

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

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

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

	return ret;
}


L
Linus Torvalds 已提交
4989 4990 4991
static struct addrconf_sysctl_table
{
	struct ctl_table_header *sysctl_header;
4992
	struct ctl_table addrconf_vars[DEVCONF_MAX+1];
4993
} addrconf_sysctl __read_mostly = {
L
Linus Torvalds 已提交
4994 4995
	.sysctl_header = NULL,
	.addrconf_vars = {
4996
		{
S
Stephen Hemminger 已提交
4997 4998 4999 5000 5001
			.procname	= "forwarding",
			.data		= &ipv6_devconf.forwarding,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= addrconf_sysctl_forward,
L
Linus Torvalds 已提交
5002 5003
		},
		{
S
Stephen Hemminger 已提交
5004 5005 5006 5007 5008
			.procname	= "hop_limit",
			.data		= &ipv6_devconf.hop_limit,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5009 5010
		},
		{
S
Stephen Hemminger 已提交
5011 5012 5013 5014 5015
			.procname	= "mtu",
			.data		= &ipv6_devconf.mtu6,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5016 5017
		},
		{
S
Stephen Hemminger 已提交
5018 5019 5020 5021 5022
			.procname	= "accept_ra",
			.data		= &ipv6_devconf.accept_ra,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5023 5024
		},
		{
S
Stephen Hemminger 已提交
5025 5026 5027 5028 5029
			.procname	= "accept_redirects",
			.data		= &ipv6_devconf.accept_redirects,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5030 5031
		},
		{
S
Stephen Hemminger 已提交
5032 5033 5034 5035 5036
			.procname	= "autoconf",
			.data		= &ipv6_devconf.autoconf,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5037 5038
		},
		{
S
Stephen Hemminger 已提交
5039 5040 5041 5042 5043
			.procname	= "dad_transmits",
			.data		= &ipv6_devconf.dad_transmits,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5044 5045
		},
		{
S
Stephen Hemminger 已提交
5046 5047 5048 5049 5050
			.procname	= "router_solicitations",
			.data		= &ipv6_devconf.rtr_solicits,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5051 5052
		},
		{
S
Stephen Hemminger 已提交
5053 5054 5055 5056 5057
			.procname	= "router_solicitation_interval",
			.data		= &ipv6_devconf.rtr_solicit_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
5058 5059
		},
		{
S
Stephen Hemminger 已提交
5060 5061 5062 5063 5064
			.procname	= "router_solicitation_delay",
			.data		= &ipv6_devconf.rtr_solicit_delay,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
5065 5066
		},
		{
S
Stephen Hemminger 已提交
5067 5068 5069 5070 5071
			.procname	= "force_mld_version",
			.data		= &ipv6_devconf.force_mld_version,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5072
		},
5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088
		{
			.procname	= "mldv1_unsolicited_report_interval",
			.data		=
				&ipv6_devconf.mldv1_unsolicited_report_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_ms_jiffies,
		},
		{
			.procname	= "mldv2_unsolicited_report_interval",
			.data		=
				&ipv6_devconf.mldv2_unsolicited_report_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_ms_jiffies,
		},
L
Linus Torvalds 已提交
5089
		{
S
Stephen Hemminger 已提交
5090 5091 5092 5093 5094
			.procname	= "use_tempaddr",
			.data		= &ipv6_devconf.use_tempaddr,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5095 5096
		},
		{
S
Stephen Hemminger 已提交
5097 5098 5099 5100 5101
			.procname	= "temp_valid_lft",
			.data		= &ipv6_devconf.temp_valid_lft,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5102 5103
		},
		{
S
Stephen Hemminger 已提交
5104 5105 5106 5107 5108
			.procname	= "temp_prefered_lft",
			.data		= &ipv6_devconf.temp_prefered_lft,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5109 5110
		},
		{
S
Stephen Hemminger 已提交
5111 5112 5113 5114 5115
			.procname	= "regen_max_retry",
			.data		= &ipv6_devconf.regen_max_retry,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5116 5117
		},
		{
S
Stephen Hemminger 已提交
5118 5119 5120 5121 5122
			.procname	= "max_desync_factor",
			.data		= &ipv6_devconf.max_desync_factor,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5123 5124
		},
		{
S
Stephen Hemminger 已提交
5125 5126 5127 5128 5129
			.procname	= "max_addresses",
			.data		= &ipv6_devconf.max_addresses,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
5130
		},
5131
		{
S
Stephen Hemminger 已提交
5132 5133 5134 5135 5136
			.procname	= "accept_ra_defrtr",
			.data		= &ipv6_devconf.accept_ra_defrtr,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5137
		},
5138
		{
S
Stephen Hemminger 已提交
5139 5140 5141 5142 5143
			.procname	= "accept_ra_pinfo",
			.data		= &ipv6_devconf.accept_ra_pinfo,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5144
		},
5145 5146
#ifdef CONFIG_IPV6_ROUTER_PREF
		{
S
Stephen Hemminger 已提交
5147 5148 5149 5150 5151
			.procname	= "accept_ra_rtr_pref",
			.data		= &ipv6_devconf.accept_ra_rtr_pref,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5152
		},
5153
		{
S
Stephen Hemminger 已提交
5154 5155 5156 5157 5158
			.procname	= "router_probe_interval",
			.data		= &ipv6_devconf.rtr_probe_interval,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec_jiffies,
5159
		},
N
Neil Horman 已提交
5160
#ifdef CONFIG_IPV6_ROUTE_INFO
5161
		{
S
Stephen Hemminger 已提交
5162 5163 5164 5165 5166
			.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,
5167 5168
		},
#endif
5169
#endif
5170
		{
S
Stephen Hemminger 已提交
5171 5172 5173 5174
			.procname	= "proxy_ndp",
			.data		= &ipv6_devconf.proxy_ndp,
			.maxlen		= sizeof(int),
			.mode		= 0644,
5175
			.proc_handler	= addrconf_sysctl_proxy_ndp,
5176
		},
5177
		{
S
Stephen Hemminger 已提交
5178 5179 5180 5181 5182
			.procname	= "accept_source_route",
			.data		= &ipv6_devconf.accept_source_route,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5183
		},
5184 5185
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
		{
S
Stephen Hemminger 已提交
5186 5187 5188 5189 5190
			.procname       = "optimistic_dad",
			.data           = &ipv6_devconf.optimistic_dad,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec,
5191 5192

		},
5193 5194 5195 5196 5197 5198 5199 5200
		{
			.procname       = "use_optimistic",
			.data           = &ipv6_devconf.use_optimistic,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec,

		},
5201 5202 5203
#endif
#ifdef CONFIG_IPV6_MROUTE
		{
S
Stephen Hemminger 已提交
5204 5205 5206 5207 5208
			.procname	= "mc_forwarding",
			.data		= &ipv6_devconf.mc_forwarding,
			.maxlen		= sizeof(int),
			.mode		= 0444,
			.proc_handler	= proc_dointvec,
5209
		},
5210
#endif
5211
		{
S
Stephen Hemminger 已提交
5212 5213 5214 5215 5216
			.procname	= "disable_ipv6",
			.data		= &ipv6_devconf.disable_ipv6,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= addrconf_sysctl_disable,
5217
		},
5218
		{
S
Stephen Hemminger 已提交
5219 5220 5221 5222 5223
			.procname	= "accept_dad",
			.data		= &ipv6_devconf.accept_dad,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
5224
		},
5225 5226 5227 5228 5229 5230 5231
		{
			.procname       = "force_tllao",
			.data           = &ipv6_devconf.force_tllao,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec
		},
5232 5233 5234 5235 5236 5237 5238
		{
			.procname       = "ndisc_notify",
			.data           = &ipv6_devconf.ndisc_notify,
			.maxlen         = sizeof(int),
			.mode           = 0644,
			.proc_handler   = proc_dointvec
		},
5239 5240 5241 5242 5243 5244 5245
		{
			.procname	= "suppress_frag_ndisc",
			.data		= &ipv6_devconf.suppress_frag_ndisc,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec
		},
5246 5247 5248 5249 5250 5251 5252
		{
			.procname	= "accept_ra_from_local",
			.data		= &ipv6_devconf.accept_ra_from_local,
			.maxlen		= sizeof(int),
			.mode		= 0644,
			.proc_handler	= proc_dointvec,
		},
L
Linus Torvalds 已提交
5253
		{
5254
			/* sentinel */
L
Linus Torvalds 已提交
5255 5256 5257 5258
		}
	},
};

5259
static int __addrconf_sysctl_register(struct net *net, char *dev_name,
5260
		struct inet6_dev *idev, struct ipv6_devconf *p)
L
Linus Torvalds 已提交
5261 5262 5263
{
	int i;
	struct addrconf_sysctl_table *t;
5264
	char path[sizeof("net/ipv6/conf/") + IFNAMSIZ];
5265

A
Arnaldo Carvalho de Melo 已提交
5266
	t = kmemdup(&addrconf_sysctl, sizeof(*t), GFP_KERNEL);
L
Linus Torvalds 已提交
5267
	if (t == NULL)
5268 5269
		goto out;

S
Stephen Hemminger 已提交
5270
	for (i = 0; t->addrconf_vars[i].data; i++) {
S
Stephen Hemminger 已提交
5271
		t->addrconf_vars[i].data += (char *)p - (char *)&ipv6_devconf;
L
Linus Torvalds 已提交
5272
		t->addrconf_vars[i].extra1 = idev; /* embedded; no ref */
5273
		t->addrconf_vars[i].extra2 = net;
L
Linus Torvalds 已提交
5274 5275
	}

5276
	snprintf(path, sizeof(path), "net/ipv6/conf/%s", dev_name);
L
Linus Torvalds 已提交
5277

5278
	t->sysctl_header = register_net_sysctl(net, path, t->addrconf_vars);
L
Linus Torvalds 已提交
5279
	if (t->sysctl_header == NULL)
5280
		goto free;
5281 5282

	p->sysctl = t;
5283
	return 0;
L
Linus Torvalds 已提交
5284

5285
free:
L
Linus Torvalds 已提交
5286
	kfree(t);
5287
out:
5288
	return -ENOBUFS;
L
Linus Torvalds 已提交
5289 5290
}

5291 5292 5293 5294 5295 5296 5297 5298 5299
static void __addrconf_sysctl_unregister(struct ipv6_devconf *p)
{
	struct addrconf_sysctl_table *t;

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

	t = p->sysctl;
	p->sysctl = NULL;
5300
	unregister_net_sysctl_table(t->sysctl_header);
5301 5302 5303
	kfree(t);
}

5304
static int addrconf_sysctl_register(struct inet6_dev *idev)
5305
{
5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320
	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;
5321 5322
}

5323
static void addrconf_sysctl_unregister(struct inet6_dev *idev)
L
Linus Torvalds 已提交
5324
{
5325 5326
	__addrconf_sysctl_unregister(&idev->cnf);
	neigh_sysctl_unregister(idev->nd_parms);
L
Linus Torvalds 已提交
5327 5328 5329 5330 5331
}


#endif

5332
static int __net_init addrconf_init_net(struct net *net)
5333
{
5334
	int err = -ENOMEM;
5335 5336
	struct ipv6_devconf *all, *dflt;

5337 5338 5339
	all = kmemdup(&ipv6_devconf, sizeof(ipv6_devconf), GFP_KERNEL);
	if (all == NULL)
		goto err_alloc_all;
5340

5341 5342 5343
	dflt = kmemdup(&ipv6_devconf_dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
	if (dflt == NULL)
		goto err_alloc_dflt;
5344

5345 5346 5347
	/* these will be inherited by all namespaces */
	dflt->autoconf = ipv6_defaults.autoconf;
	dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
5348 5349 5350 5351 5352

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

#ifdef CONFIG_SYSCTL
5353
	err = __addrconf_sysctl_register(net, "all", NULL, all);
5354 5355 5356
	if (err < 0)
		goto err_reg_all;

5357
	err = __addrconf_sysctl_register(net, "default", NULL, dflt);
5358 5359 5360 5361 5362 5363 5364 5365 5366 5367 5368 5369 5370 5371 5372 5373 5374
	if (err < 0)
		goto err_reg_dflt;
#endif
	return 0;

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

5375
static void __net_exit addrconf_exit_net(struct net *net)
5376 5377 5378 5379 5380
{
#ifdef CONFIG_SYSCTL
	__addrconf_sysctl_unregister(net->ipv6.devconf_dflt);
	__addrconf_sysctl_unregister(net->ipv6.devconf_all);
#endif
O
Octavian Purdila 已提交
5381
	if (!net_eq(net, &init_net)) {
5382 5383 5384 5385 5386 5387 5388 5389 5390 5391
		kfree(net->ipv6.devconf_dflt);
		kfree(net->ipv6.devconf_all);
	}
}

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

T
Thomas Graf 已提交
5392 5393 5394 5395
static struct rtnl_af_ops inet6_ops = {
	.family		  = AF_INET6,
	.fill_link_af	  = inet6_fill_link_af,
	.get_link_af_size = inet6_get_link_af_size,
5396
	.set_link_af	  = inet6_set_link_af,
T
Thomas Graf 已提交
5397 5398
};

L
Linus Torvalds 已提交
5399 5400 5401 5402 5403 5404
/*
 *	Init / cleanup code
 */

int __init addrconf_init(void)
{
5405
	struct inet6_dev *idev;
5406
	int i, err;
5407

S
Stephen Hemminger 已提交
5408 5409
	err = ipv6_addr_label_init();
	if (err < 0) {
5410 5411
		pr_crit("%s: cannot initialize default policy table: %d\n",
			__func__, err);
5412
		goto out;
5413
	}
L
Linus Torvalds 已提交
5414

5415 5416 5417
	err = register_pernet_subsys(&addrconf_ops);
	if (err < 0)
		goto out_addrlabel;
5418

5419 5420 5421 5422 5423 5424
	addrconf_wq = create_workqueue("ipv6_addrconf");
	if (!addrconf_wq) {
		err = -ENOMEM;
		goto out_nowq;
	}

L
Linus Torvalds 已提交
5425 5426 5427 5428 5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443
	/* 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();
5444
	idev = ipv6_add_dev(init_net.loopback_dev);
L
Linus Torvalds 已提交
5445
	rtnl_unlock();
5446 5447
	if (IS_ERR(idev)) {
		err = PTR_ERR(idev);
5448
		goto errlo;
5449
	}
L
Linus Torvalds 已提交
5450

5451 5452 5453
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		INIT_HLIST_HEAD(&inet6_addr_lst[i]);

L
Linus Torvalds 已提交
5454 5455
	register_netdevice_notifier(&ipv6_dev_notf);

5456
	addrconf_verify();
5457

5458
	rtnl_af_register(&inet6_ops);
T
Thomas Graf 已提交
5459

5460 5461
	err = __rtnl_register(PF_INET6, RTM_GETLINK, NULL, inet6_dump_ifinfo,
			      NULL);
5462 5463 5464 5465
	if (err < 0)
		goto errout;

	/* Only the first call to __rtnl_register can fail */
5466 5467 5468 5469 5470 5471 5472 5473
	__rtnl_register(PF_INET6, RTM_NEWADDR, inet6_rtm_newaddr, NULL, NULL);
	__rtnl_register(PF_INET6, RTM_DELADDR, inet6_rtm_deladdr, NULL, NULL);
	__rtnl_register(PF_INET6, RTM_GETADDR, inet6_rtm_getaddr,
			inet6_dump_ifaddr, NULL);
	__rtnl_register(PF_INET6, RTM_GETMULTICAST, NULL,
			inet6_dump_ifmcaddr, NULL);
	__rtnl_register(PF_INET6, RTM_GETANYCAST, NULL,
			inet6_dump_ifacaddr, NULL);
5474
	__rtnl_register(PF_INET6, RTM_GETNETCONF, inet6_netconf_get_devconf,
5475
			inet6_netconf_dump_devconf, NULL);
5476

5477 5478
	ipv6_addr_label_rtnl_register();

L
Linus Torvalds 已提交
5479
	return 0;
5480
errout:
T
Thomas Graf 已提交
5481
	rtnl_af_unregister(&inet6_ops);
5482
	unregister_netdevice_notifier(&ipv6_dev_notf);
5483
errlo:
5484 5485
	destroy_workqueue(addrconf_wq);
out_nowq:
5486
	unregister_pernet_subsys(&addrconf_ops);
5487 5488 5489
out_addrlabel:
	ipv6_addr_label_cleanup();
out:
5490
	return err;
L
Linus Torvalds 已提交
5491 5492
}

5493
void addrconf_cleanup(void)
L
Linus Torvalds 已提交
5494
{
5495
	struct net_device *dev;
L
Linus Torvalds 已提交
5496 5497 5498
	int i;

	unregister_netdevice_notifier(&ipv6_dev_notf);
5499
	unregister_pernet_subsys(&addrconf_ops);
5500
	ipv6_addr_label_cleanup();
L
Linus Torvalds 已提交
5501 5502 5503

	rtnl_lock();

T
Thomas Graf 已提交
5504 5505
	__rtnl_af_unregister(&inet6_ops);

5506 5507 5508 5509 5510 5511 5512 5513
	/* 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
5514 5515 5516
	/*
	 *	Check hash table.
	 */
5517
	spin_lock_bh(&addrconf_hash_lock);
5518 5519
	for (i = 0; i < IN6_ADDR_HSIZE; i++)
		WARN_ON(!hlist_empty(&inet6_addr_lst[i]));
5520
	spin_unlock_bh(&addrconf_hash_lock);
5521
	cancel_delayed_work(&addr_chk_work);
L
Linus Torvalds 已提交
5522
	rtnl_unlock();
5523 5524

	destroy_workqueue(addrconf_wq);
L
Linus Torvalds 已提交
5525
}