protocol.c 41.4 KB
Newer Older
1
/* SCTP kernel implementation
L
Linus Torvalds 已提交
2 3 4 5 6 7 8
 * (C) Copyright IBM Corp. 2001, 2004
 * Copyright (c) 1999-2000 Cisco, Inc.
 * Copyright (c) 1999-2001 Motorola, Inc.
 * Copyright (c) 2001 Intel Corp.
 * Copyright (c) 2001 Nokia, Inc.
 * Copyright (c) 2001 La Monte H.P. Yarroll
 *
9
 * This file is part of the SCTP kernel implementation
L
Linus Torvalds 已提交
10 11 12
 *
 * Initialization/cleanup for SCTP protocol support.
 *
13
 * This SCTP implementation is free software;
L
Linus Torvalds 已提交
14 15 16 17 18
 * 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, or (at your option)
 * any later version.
 *
19
 * This SCTP implementation is distributed in the hope that it
L
Linus Torvalds 已提交
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
 * will be useful, but WITHOUT ANY WARRANTY; without even the implied
 *                 ************************
 * warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with GNU CC; see the file COPYING.  If not, write to
 * the Free Software Foundation, 59 Temple Place - Suite 330,
 * Boston, MA 02111-1307, USA.
 *
 * Please send any bug reports or fixes you make to the
 * email address(es):
 *    lksctp developers <lksctp-developers@lists.sourceforge.net>
 *
 * Or submit a bug report through the following website:
 *    http://www.sf.net/projects/lksctp
 *
 * Written or modified by:
 *    La Monte H.P. Yarroll <piggy@acm.org>
 *    Karl Knutson <karl@athena.chicago.il.us>
 *    Jon Grimm <jgrimm@us.ibm.com>
 *    Sridhar Samudrala <sri@us.ibm.com>
 *    Daisy Chang <daisyc@us.ibm.com>
 *    Ardelle Fan <ardelle.fan@intel.com>
 *
 * Any bugs reported given to us we will try to fix... any fixes shared will
 * be incorporated into the next SCTP release.
 */

49 50
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

L
Linus Torvalds 已提交
51 52 53 54 55
#include <linux/module.h>
#include <linux/init.h>
#include <linux/netdevice.h>
#include <linux/inetdevice.h>
#include <linux/seq_file.h>
56
#include <linux/bootmem.h>
57 58
#include <linux/highmem.h>
#include <linux/swap.h>
59
#include <linux/slab.h>
60
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
61 62 63
#include <net/protocol.h>
#include <net/ip.h>
#include <net/ipv6.h>
64
#include <net/route.h>
L
Linus Torvalds 已提交
65 66 67 68 69 70
#include <net/sctp/sctp.h>
#include <net/addrconf.h>
#include <net/inet_common.h>
#include <net/inet_ecn.h>

/* Global data structures. */
71
struct sctp_globals sctp_globals __read_mostly;
L
Linus Torvalds 已提交
72 73 74 75 76 77 78 79 80

struct idr sctp_assocs_id;
DEFINE_SPINLOCK(sctp_assocs_id_lock);

static struct sctp_pf *sctp_pf_inet6_specific;
static struct sctp_pf *sctp_pf_inet_specific;
static struct sctp_af *sctp_af_v4_specific;
static struct sctp_af *sctp_af_v6_specific;

81 82
struct kmem_cache *sctp_chunk_cachep __read_mostly;
struct kmem_cache *sctp_bucket_cachep __read_mostly;
L
Linus Torvalds 已提交
83

E
Eric Dumazet 已提交
84
long sysctl_sctp_mem[3];
85 86
int sysctl_sctp_rmem[3];
int sysctl_sctp_wmem[3];
87

L
Linus Torvalds 已提交
88
/* Set up the proc fs entry for the SCTP protocol. */
89
static __net_init int sctp_proc_init(struct net *net)
L
Linus Torvalds 已提交
90
{
91
#ifdef CONFIG_PROC_FS
92 93 94 95
	net->sctp.proc_net_sctp = proc_net_mkdir(net, "sctp", net->proc_net);
	if (!net->sctp.proc_net_sctp)
		goto out_proc_net_sctp;
	if (sctp_snmp_proc_init(net))
96
		goto out_snmp_proc_init;
97
	if (sctp_eps_proc_init(net))
98
		goto out_eps_proc_init;
99
	if (sctp_assocs_proc_init(net))
100
		goto out_assocs_proc_init;
101
	if (sctp_remaddr_proc_init(net))
102
		goto out_remaddr_proc_init;
L
Linus Torvalds 已提交
103 104 105

	return 0;

106
out_remaddr_proc_init:
107
	sctp_assocs_proc_exit(net);
108
out_assocs_proc_init:
109
	sctp_eps_proc_exit(net);
110
out_eps_proc_init:
111
	sctp_snmp_proc_exit(net);
112
out_snmp_proc_init:
113 114 115
	remove_proc_entry("sctp", net->proc_net);
	net->sctp.proc_net_sctp = NULL;
out_proc_net_sctp:
116
	return -ENOMEM;
117 118
#endif /* CONFIG_PROC_FS */
	return 0;
L
Linus Torvalds 已提交
119 120
}

121
/* Clean up the proc fs entry for the SCTP protocol.
L
Linus Torvalds 已提交
122 123 124
 * Note: Do not make this __exit as it is used in the init error
 * path.
 */
125
static void sctp_proc_exit(struct net *net)
L
Linus Torvalds 已提交
126
{
127
#ifdef CONFIG_PROC_FS
128 129 130 131 132 133 134
	sctp_snmp_proc_exit(net);
	sctp_eps_proc_exit(net);
	sctp_assocs_proc_exit(net);
	sctp_remaddr_proc_exit(net);

	remove_proc_entry("sctp", net->proc_net);
	net->sctp.proc_net_sctp = NULL;
135
#endif
L
Linus Torvalds 已提交
136 137 138 139 140 141 142 143 144 145 146 147 148
}

/* Private helper to extract ipv4 address and stash them in
 * the protocol structure.
 */
static void sctp_v4_copy_addrlist(struct list_head *addrlist,
				  struct net_device *dev)
{
	struct in_device *in_dev;
	struct in_ifaddr *ifa;
	struct sctp_sockaddr_entry *addr;

	rcu_read_lock();
149
	if ((in_dev = __in_dev_get_rcu(dev)) == NULL) {
L
Linus Torvalds 已提交
150 151 152 153 154 155 156 157
		rcu_read_unlock();
		return;
	}

	for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) {
		/* Add the address to the local list.  */
		addr = t_new(struct sctp_sockaddr_entry, GFP_ATOMIC);
		if (addr) {
158 159 160
			addr->a.v4.sin_family = AF_INET;
			addr->a.v4.sin_port = 0;
			addr->a.v4.sin_addr.s_addr = ifa->ifa_local;
161 162
			addr->valid = 1;
			INIT_LIST_HEAD(&addr->list);
L
Linus Torvalds 已提交
163 164 165 166 167 168 169 170 171 172
			list_add_tail(&addr->list, addrlist);
		}
	}

	rcu_read_unlock();
}

/* Extract our IP addresses from the system and stash them in the
 * protocol structure.
 */
173
static void sctp_get_local_addr_list(struct net *net)
L
Linus Torvalds 已提交
174 175 176 177 178
{
	struct net_device *dev;
	struct list_head *pos;
	struct sctp_af *af;

179
	rcu_read_lock();
180
	for_each_netdev_rcu(net, dev) {
L
Linus Torvalds 已提交
181 182
		__list_for_each(pos, &sctp_address_families) {
			af = list_entry(pos, struct sctp_af, list);
183
			af->copy_addrlist(&net->sctp.local_addr_list, dev);
L
Linus Torvalds 已提交
184 185
		}
	}
186
	rcu_read_unlock();
L
Linus Torvalds 已提交
187 188 189
}

/* Free the existing local addresses.  */
190
static void sctp_free_local_addr_list(struct net *net)
L
Linus Torvalds 已提交
191 192 193 194
{
	struct sctp_sockaddr_entry *addr;
	struct list_head *pos, *temp;

195
	list_for_each_safe(pos, temp, &net->sctp.local_addr_list) {
L
Linus Torvalds 已提交
196 197 198 199 200 201 202
		addr = list_entry(pos, struct sctp_sockaddr_entry, list);
		list_del(pos);
		kfree(addr);
	}
}

/* Copy the local addresses which are valid for 'scope' into 'bp'.  */
203 204
int sctp_copy_local_addr_list(struct net *net, struct sctp_bind_addr *bp,
			      sctp_scope_t scope, gfp_t gfp, int copy_flags)
L
Linus Torvalds 已提交
205 206 207 208
{
	struct sctp_sockaddr_entry *addr;
	int error = 0;

209
	rcu_read_lock();
210
	list_for_each_entry_rcu(addr, &net->sctp.local_addr_list, list) {
211 212
		if (!addr->valid)
			continue;
213
		if (sctp_in_scope(net, &addr->a, scope)) {
L
Linus Torvalds 已提交
214 215 216 217
			/* Now that the address is in scope, check to see if
			 * the address type is really supported by the local
			 * sock as well as the remote peer.
			 */
218
			if ((((AF_INET == addr->a.sa.sa_family) &&
L
Linus Torvalds 已提交
219
			      (copy_flags & SCTP_ADDR4_PEERSUPP))) ||
220
			    (((AF_INET6 == addr->a.sa.sa_family) &&
L
Linus Torvalds 已提交
221 222
			      (copy_flags & SCTP_ADDR6_ALLOWED) &&
			      (copy_flags & SCTP_ADDR6_PEERSUPP)))) {
223 224
				error = sctp_add_bind_addr(bp, &addr->a,
						    SCTP_ADDR_SRC, GFP_ATOMIC);
L
Linus Torvalds 已提交
225 226 227 228 229 230 231
				if (error)
					goto end_copy;
			}
		}
	}

end_copy:
232
	rcu_read_unlock();
L
Linus Torvalds 已提交
233 234 235 236 237 238 239 240
	return error;
}

/* Initialize a sctp_addr from in incoming skb.  */
static void sctp_v4_from_skb(union sctp_addr *addr, struct sk_buff *skb,
			     int is_saddr)
{
	void *from;
241
	__be16 *port;
L
Linus Torvalds 已提交
242 243 244 245 246
	struct sctphdr *sh;

	port = &addr->v4.sin_port;
	addr->v4.sin_family = AF_INET;

247
	sh = sctp_hdr(skb);
L
Linus Torvalds 已提交
248
	if (is_saddr) {
249
		*port  = sh->source;
250
		from = &ip_hdr(skb)->saddr;
L
Linus Torvalds 已提交
251
	} else {
252
		*port = sh->dest;
253
		from = &ip_hdr(skb)->daddr;
L
Linus Torvalds 已提交
254 255 256 257 258 259 260 261
	}
	memcpy(&addr->v4.sin_addr.s_addr, from, sizeof(struct in_addr));
}

/* Initialize an sctp_addr from a socket. */
static void sctp_v4_from_sk(union sctp_addr *addr, struct sock *sk)
{
	addr->v4.sin_family = AF_INET;
262
	addr->v4.sin_port = 0;
E
Eric Dumazet 已提交
263
	addr->v4.sin_addr.s_addr = inet_sk(sk)->inet_rcv_saddr;
L
Linus Torvalds 已提交
264 265 266 267 268
}

/* Initialize sk->sk_rcv_saddr from sctp_addr. */
static void sctp_v4_to_sk_saddr(union sctp_addr *addr, struct sock *sk)
{
E
Eric Dumazet 已提交
269
	inet_sk(sk)->inet_rcv_saddr = addr->v4.sin_addr.s_addr;
L
Linus Torvalds 已提交
270 271 272 273 274
}

/* Initialize sk->sk_daddr from sctp_addr. */
static void sctp_v4_to_sk_daddr(union sctp_addr *addr, struct sock *sk)
{
E
Eric Dumazet 已提交
275
	inet_sk(sk)->inet_daddr = addr->v4.sin_addr.s_addr;
L
Linus Torvalds 已提交
276 277 278 279 280
}

/* Initialize a sctp_addr from an address parameter. */
static void sctp_v4_from_addr_param(union sctp_addr *addr,
				    union sctp_addr_param *param,
281
				    __be16 port, int iif)
L
Linus Torvalds 已提交
282 283 284 285 286 287 288 289 290 291 292 293 294 295 296
{
	addr->v4.sin_family = AF_INET;
	addr->v4.sin_port = port;
	addr->v4.sin_addr.s_addr = param->v4.addr.s_addr;
}

/* Initialize an address parameter from a sctp_addr and return the length
 * of the address parameter.
 */
static int sctp_v4_to_addr_param(const union sctp_addr *addr,
				 union sctp_addr_param *param)
{
	int length = sizeof(sctp_ipv4addr_param_t);

	param->v4.param_hdr.type = SCTP_PARAM_IPV4_ADDRESS;
297
	param->v4.param_hdr.length = htons(length);
298
	param->v4.addr.s_addr = addr->v4.sin_addr.s_addr;
L
Linus Torvalds 已提交
299 300 301 302 303

	return length;
}

/* Initialize a sctp_addr from a dst_entry. */
304
static void sctp_v4_dst_saddr(union sctp_addr *saddr, struct flowi4 *fl4,
A
Al Viro 已提交
305
			      __be16 port)
L
Linus Torvalds 已提交
306 307 308
{
	saddr->v4.sin_family = AF_INET;
	saddr->v4.sin_port = port;
309
	saddr->v4.sin_addr.s_addr = fl4->saddr;
L
Linus Torvalds 已提交
310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326
}

/* Compare two addresses exactly. */
static int sctp_v4_cmp_addr(const union sctp_addr *addr1,
			    const union sctp_addr *addr2)
{
	if (addr1->sa.sa_family != addr2->sa.sa_family)
		return 0;
	if (addr1->v4.sin_port != addr2->v4.sin_port)
		return 0;
	if (addr1->v4.sin_addr.s_addr != addr2->v4.sin_addr.s_addr)
		return 0;

	return 1;
}

/* Initialize addr struct to INADDR_ANY. */
A
Al Viro 已提交
327
static void sctp_v4_inaddr_any(union sctp_addr *addr, __be16 port)
L
Linus Torvalds 已提交
328 329
{
	addr->v4.sin_family = AF_INET;
A
Al Viro 已提交
330
	addr->v4.sin_addr.s_addr = htonl(INADDR_ANY);
L
Linus Torvalds 已提交
331 332 333 334 335 336
	addr->v4.sin_port = port;
}

/* Is this a wildcard address? */
static int sctp_v4_is_any(const union sctp_addr *addr)
{
A
Al Viro 已提交
337
	return htonl(INADDR_ANY) == addr->v4.sin_addr.s_addr;
L
Linus Torvalds 已提交
338 339 340 341 342 343 344 345 346
}

/* This function checks if the address is a valid address to be used for
 * SCTP binding.
 *
 * Output:
 * Return 0 - If the address is a non-unicast or an illegal address.
 * Return 1 - If the address is a unicast.
 */
347 348 349
static int sctp_v4_addr_valid(union sctp_addr *addr,
			      struct sctp_sock *sp,
			      const struct sk_buff *skb)
L
Linus Torvalds 已提交
350
{
351 352 353 354
	/* IPv4 addresses not allowed */
	if (sp && ipv6_only_sock(sctp_opt2sk(sp)))
		return 0;

L
Linus Torvalds 已提交
355
	/* Is this a non-unicast address or a unusable SCTP address? */
J
Joe Perches 已提交
356
	if (IS_IPV4_UNUSABLE_ADDRESS(addr->v4.sin_addr.s_addr))
L
Linus Torvalds 已提交
357 358
		return 0;

359
	/* Is this a broadcast address? */
E
Eric Dumazet 已提交
360
	if (skb && skb_rtable(skb)->rt_flags & RTCF_BROADCAST)
361
		return 0;
362

L
Linus Torvalds 已提交
363 364 365 366 367 368
	return 1;
}

/* Should this be available for binding?   */
static int sctp_v4_available(union sctp_addr *addr, struct sctp_sock *sp)
{
369 370
	struct net *net = sock_net(&sp->inet.sk);
	int ret = inet_addr_type(net, addr->v4.sin_addr.s_addr);
L
Linus Torvalds 已提交
371 372


A
Al Viro 已提交
373
	if (addr->v4.sin_addr.s_addr != htonl(INADDR_ANY) &&
374 375 376
	   ret != RTN_LOCAL &&
	   !sp->inet.freebind &&
	   !sysctl_ip_nonlocal_bind)
L
Linus Torvalds 已提交
377
		return 0;
378

379 380 381
	if (ipv6_only_sock(sctp_opt2sk(sp)))
		return 0;

L
Linus Torvalds 已提交
382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397
	return 1;
}

/* Checking the loopback, private and other address scopes as defined in
 * RFC 1918.   The IPv4 scoping is based on the draft for SCTP IPv4
 * scoping <draft-stewart-tsvwg-sctp-ipv4-00.txt>.
 *
 * Level 0 - unusable SCTP addresses
 * Level 1 - loopback address
 * Level 2 - link-local addresses
 * Level 3 - private addresses.
 * Level 4 - global addresses
 * For INIT and INIT-ACK address list, let L be the level of
 * of requested destination address, sender and receiver
 * SHOULD include all of its addresses with level greater
 * than or equal to L.
398 399 400
 *
 * IPv4 scoping can be controlled through sysctl option
 * net.sctp.addr_scope_policy
L
Linus Torvalds 已提交
401 402 403 404 405 406
 */
static sctp_scope_t sctp_v4_scope(union sctp_addr *addr)
{
	sctp_scope_t retval;

	/* Check for unusable SCTP addresses. */
J
Joe Perches 已提交
407
	if (IS_IPV4_UNUSABLE_ADDRESS(addr->v4.sin_addr.s_addr)) {
L
Linus Torvalds 已提交
408
		retval =  SCTP_SCOPE_UNUSABLE;
J
Joe Perches 已提交
409
	} else if (ipv4_is_loopback(addr->v4.sin_addr.s_addr)) {
L
Linus Torvalds 已提交
410
		retval = SCTP_SCOPE_LOOPBACK;
J
Joe Perches 已提交
411
	} else if (ipv4_is_linklocal_169(addr->v4.sin_addr.s_addr)) {
L
Linus Torvalds 已提交
412
		retval = SCTP_SCOPE_LINK;
J
Joe Perches 已提交
413 414 415
	} else if (ipv4_is_private_10(addr->v4.sin_addr.s_addr) ||
		   ipv4_is_private_172(addr->v4.sin_addr.s_addr) ||
		   ipv4_is_private_192(addr->v4.sin_addr.s_addr)) {
L
Linus Torvalds 已提交
416 417 418 419 420 421 422 423 424 425 426 427
		retval = SCTP_SCOPE_PRIVATE;
	} else {
		retval = SCTP_SCOPE_GLOBAL;
	}

	return retval;
}

/* Returns a valid dst cache entry for the given source and destination ip
 * addresses. If an association is passed, trys to get a dst entry with a
 * source address that matches an address in the bind address list.
 */
428 429
static void sctp_v4_get_dst(struct sctp_transport *t, union sctp_addr *saddr,
				struct flowi *fl, struct sock *sk)
L
Linus Torvalds 已提交
430
{
431
	struct sctp_association *asoc = t->asoc;
L
Linus Torvalds 已提交
432
	struct rtable *rt;
433
	struct flowi4 *fl4 = &fl->u.ip4;
L
Linus Torvalds 已提交
434 435 436
	struct sctp_bind_addr *bp;
	struct sctp_sockaddr_entry *laddr;
	struct dst_entry *dst = NULL;
437
	union sctp_addr *daddr = &t->ipaddr;
L
Linus Torvalds 已提交
438 439
	union sctp_addr dst_saddr;

440 441 442 443
	memset(fl4, 0x0, sizeof(struct flowi4));
	fl4->daddr  = daddr->v4.sin_addr.s_addr;
	fl4->fl4_dport = daddr->v4.sin_port;
	fl4->flowi4_proto = IPPROTO_SCTP;
L
Linus Torvalds 已提交
444
	if (asoc) {
445 446 447
		fl4->flowi4_tos = RT_CONN_FLAGS(asoc->base.sk);
		fl4->flowi4_oif = asoc->base.sk->sk_bound_dev_if;
		fl4->fl4_sport = htons(asoc->base.bind_addr.port);
L
Linus Torvalds 已提交
448
	}
449
	if (saddr) {
450 451
		fl4->saddr = saddr->v4.sin_addr.s_addr;
		fl4->fl4_sport = saddr->v4.sin_port;
452
	}
L
Linus Torvalds 已提交
453

H
Harvey Harrison 已提交
454
	SCTP_DEBUG_PRINTK("%s: DST:%pI4, SRC:%pI4 - ",
455
			  __func__, &fl4->daddr, &fl4->saddr);
L
Linus Torvalds 已提交
456

457
	rt = ip_route_output_key(sock_net(sk), fl4);
458
	if (!IS_ERR(rt))
459
		dst = &rt->dst;
L
Linus Torvalds 已提交
460 461 462 463 464 465 466 467 468 469 470 471 472

	/* If there is no association or if a source address is passed, no
	 * more validation is required.
	 */
	if (!asoc || saddr)
		goto out;

	bp = &asoc->base.bind_addr;

	if (dst) {
		/* Walk through the bind address list and look for a bind
		 * address that matches the source address of the returned dst.
		 */
473
		sctp_v4_dst_saddr(&dst_saddr, fl4, htons(bp->port));
474 475
		rcu_read_lock();
		list_for_each_entry_rcu(laddr, &bp->address_list, list) {
476 477 478
			if (!laddr->valid || (laddr->state == SCTP_ADDR_DEL) ||
			    (laddr->state != SCTP_ADDR_SRC &&
			    !asoc->src_out_of_asoc_ok))
479
				continue;
A
Al Viro 已提交
480
			if (sctp_v4_cmp_addr(&dst_saddr, &laddr->a))
L
Linus Torvalds 已提交
481 482
				goto out_unlock;
		}
483
		rcu_read_unlock();
L
Linus Torvalds 已提交
484 485 486 487 488 489 490 491 492 493 494

		/* None of the bound addresses match the source address of the
		 * dst. So release it.
		 */
		dst_release(dst);
		dst = NULL;
	}

	/* Walk through the bind address list and try to get a dst that
	 * matches a bind address as the source address.
	 */
495 496 497 498
	rcu_read_lock();
	list_for_each_entry_rcu(laddr, &bp->address_list, list) {
		if (!laddr->valid)
			continue;
499
		if ((laddr->state == SCTP_ADDR_SRC) &&
500
		    (AF_INET == laddr->a.sa.sa_family)) {
501 502
			fl4->saddr = laddr->a.v4.sin_addr.s_addr;
			fl4->fl4_sport = laddr->a.v4.sin_port;
503
			rt = ip_route_output_key(sock_net(sk), fl4);
504
			if (!IS_ERR(rt)) {
505
				dst = &rt->dst;
L
Linus Torvalds 已提交
506 507 508 509 510 511
				goto out_unlock;
			}
		}
	}

out_unlock:
512
	rcu_read_unlock();
L
Linus Torvalds 已提交
513
out:
514
	t->dst = dst;
L
Linus Torvalds 已提交
515
	if (dst)
H
Harvey Harrison 已提交
516
		SCTP_DEBUG_PRINTK("rt_dst:%pI4, rt_src:%pI4\n",
517
				  &fl4->daddr, &fl4->saddr);
L
Linus Torvalds 已提交
518 519 520 521 522 523 524
	else
		SCTP_DEBUG_PRINTK("NO ROUTE\n");
}

/* For v4, the source address is cached in the route entry(dst). So no need
 * to cache it separately and hence this is an empty routine.
 */
525
static void sctp_v4_get_saddr(struct sctp_sock *sk,
526 527
			      struct sctp_transport *t,
			      struct flowi *fl)
L
Linus Torvalds 已提交
528
{
529 530
	union sctp_addr *saddr = &t->saddr;
	struct rtable *rt = (struct rtable *)t->dst;
531

L
Linus Torvalds 已提交
532 533
	if (rt) {
		saddr->v4.sin_family = AF_INET;
534
		saddr->v4.sin_addr.s_addr = fl->u.ip4.saddr;
L
Linus Torvalds 已提交
535 536 537 538 539 540
	}
}

/* What interface did this skb arrive on? */
static int sctp_v4_skb_iif(const struct sk_buff *skb)
{
541
	return inet_iif(skb);
L
Linus Torvalds 已提交
542 543 544 545 546
}

/* Was this packet marked by Explicit Congestion Notification? */
static int sctp_v4_is_ce(const struct sk_buff *skb)
{
547
	return INET_ECN_is_ce(ip_hdr(skb)->tos);
L
Linus Torvalds 已提交
548 549 550 551 552 553
}

/* Create and initialize a new sk for the socket returned by accept(). */
static struct sock *sctp_v4_create_accept_sk(struct sock *sk,
					     struct sctp_association *asoc)
{
554
	struct sock *newsk = sk_alloc(sock_net(sk), PF_INET, GFP_KERNEL,
555
			sk->sk_prot);
556
	struct inet_sock *newinet;
L
Linus Torvalds 已提交
557 558 559 560 561 562

	if (!newsk)
		goto out;

	sock_init_data(NULL, newsk);

563
	sctp_copy_sock(newsk, sk, asoc);
L
Linus Torvalds 已提交
564 565 566 567
	sock_reset_flag(newsk, SOCK_ZAPPED);

	newinet = inet_sk(newsk);

E
Eric Dumazet 已提交
568
	newinet->inet_daddr = asoc->peer.primary_addr.v4.sin_addr.s_addr;
L
Linus Torvalds 已提交
569

570
	sk_refcnt_debug_inc(newsk);
L
Linus Torvalds 已提交
571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589

	if (newsk->sk_prot->init(newsk)) {
		sk_common_release(newsk);
		newsk = NULL;
	}

out:
	return newsk;
}

/* Map address, empty for v4 family */
static void sctp_v4_addr_v4map(struct sctp_sock *sp, union sctp_addr *addr)
{
	/* Empty */
}

/* Dump the v4 addr to the seq file. */
static void sctp_v4_seq_dump_addr(struct seq_file *seq, union sctp_addr *addr)
{
H
Harvey Harrison 已提交
590
	seq_printf(seq, "%pI4 ", &addr->v4.sin_addr);
L
Linus Torvalds 已提交
591 592
}

V
Vlad Yasevich 已提交
593 594 595 596 597
static void sctp_v4_ecn_capable(struct sock *sk)
{
	INET_ECN_xmit(sk);
}

598 599
void sctp_addr_wq_timeout_handler(unsigned long arg)
{
600
	struct net *net = (struct net *)arg;
601 602 603
	struct sctp_sockaddr_entry *addrw, *temp;
	struct sctp_sock *sp;

604
	spin_lock_bh(&net->sctp.addr_wq_lock);
605

606
	list_for_each_entry_safe(addrw, temp, &net->sctp.addr_waitq, list) {
607
		SCTP_DEBUG_PRINTK_IPADDR("sctp_addrwq_timo_handler: the first ent in wq %p is ",
608
		    " for cmd %d at entry %p\n", &net->sctp.addr_waitq, &addrw->a, addrw->state,
609 610
		    addrw);

E
Eric Dumazet 已提交
611
#if IS_ENABLED(CONFIG_IPV6)
612 613 614 615 616 617 618 619 620 621
		/* Now we send an ASCONF for each association */
		/* Note. we currently don't handle link local IPv6 addressees */
		if (addrw->a.sa.sa_family == AF_INET6) {
			struct in6_addr *in6;

			if (ipv6_addr_type(&addrw->a.v6.sin6_addr) &
			    IPV6_ADDR_LINKLOCAL)
				goto free_next;

			in6 = (struct in6_addr *)&addrw->a.v6.sin6_addr;
622
			if (ipv6_chk_addr(net, in6, NULL, 0) == 0 &&
623 624 625 626 627 628 629
			    addrw->state == SCTP_ADDR_NEW) {
				unsigned long timeo_val;

				SCTP_DEBUG_PRINTK("sctp_timo_handler: this is on DAD, trying %d sec later\n",
				    SCTP_ADDRESS_TICK_DELAY);
				timeo_val = jiffies;
				timeo_val += msecs_to_jiffies(SCTP_ADDRESS_TICK_DELAY);
630
				mod_timer(&net->sctp.addr_wq_timer, timeo_val);
631 632 633
				break;
			}
		}
634
#endif
635
		list_for_each_entry(sp, &net->sctp.auto_asconf_splist, auto_asconf_list) {
636 637 638 639 640 641 642 643 644 645 646
			struct sock *sk;

			sk = sctp_opt2sk(sp);
			/* ignore bound-specific endpoints */
			if (!sctp_is_ep_boundall(sk))
				continue;
			sctp_bh_lock_sock(sk);
			if (sctp_asconf_mgmt(sp, addrw) < 0)
				SCTP_DEBUG_PRINTK("sctp_addrwq_timo_handler: sctp_asconf_mgmt failed\n");
			sctp_bh_unlock_sock(sk);
		}
647
#if IS_ENABLED(CONFIG_IPV6)
648
free_next:
649
#endif
650 651 652
		list_del(&addrw->list);
		kfree(addrw);
	}
653
	spin_unlock_bh(&net->sctp.addr_wq_lock);
654 655
}

656
static void sctp_free_addr_wq(struct net *net)
657 658 659 660
{
	struct sctp_sockaddr_entry *addrw;
	struct sctp_sockaddr_entry *temp;

661 662 663
	spin_lock_bh(&net->sctp.addr_wq_lock);
	del_timer(&net->sctp.addr_wq_timer);
	list_for_each_entry_safe(addrw, temp, &net->sctp.addr_waitq, list) {
664 665 666
		list_del(&addrw->list);
		kfree(addrw);
	}
667
	spin_unlock_bh(&net->sctp.addr_wq_lock);
668 669 670 671 672
}

/* lookup the entry for the same address in the addr_waitq
 * sctp_addr_wq MUST be locked
 */
673 674
static struct sctp_sockaddr_entry *sctp_addr_wq_lookup(struct net *net,
					struct sctp_sockaddr_entry *addr)
675 676 677
{
	struct sctp_sockaddr_entry *addrw;

678
	list_for_each_entry(addrw, &net->sctp.addr_waitq, list) {
679 680 681 682 683 684 685 686 687 688 689 690 691 692 693
		if (addrw->a.sa.sa_family != addr->a.sa.sa_family)
			continue;
		if (addrw->a.sa.sa_family == AF_INET) {
			if (addrw->a.v4.sin_addr.s_addr ==
			    addr->a.v4.sin_addr.s_addr)
				return addrw;
		} else if (addrw->a.sa.sa_family == AF_INET6) {
			if (ipv6_addr_equal(&addrw->a.v6.sin6_addr,
			    &addr->a.v6.sin6_addr))
				return addrw;
		}
	}
	return NULL;
}

694
void sctp_addr_wq_mgmt(struct net *net, struct sctp_sockaddr_entry *addr, int cmd)
695 696 697 698 699 700 701 702 703 704
{
	struct sctp_sockaddr_entry *addrw;
	unsigned long timeo_val;

	/* first, we check if an opposite message already exist in the queue.
	 * If we found such message, it is removed.
	 * This operation is a bit stupid, but the DHCP client attaches the
	 * new address after a couple of addition and deletion of that address
	 */

705
	spin_lock_bh(&net->sctp.addr_wq_lock);
706
	/* Offsets existing events in addr_wq */
707
	addrw = sctp_addr_wq_lookup(net, addr);
708 709 710 711
	if (addrw) {
		if (addrw->state != cmd) {
			SCTP_DEBUG_PRINTK_IPADDR("sctp_addr_wq_mgmt offsets existing entry for %d ",
			    " in wq %p\n", addrw->state, &addrw->a,
712
			    &net->sctp.addr_waitq);
713 714 715
			list_del(&addrw->list);
			kfree(addrw);
		}
716
		spin_unlock_bh(&net->sctp.addr_wq_lock);
717 718 719 720 721 722
		return;
	}

	/* OK, we have to add the new address to the wait queue */
	addrw = kmemdup(addr, sizeof(struct sctp_sockaddr_entry), GFP_ATOMIC);
	if (addrw == NULL) {
723
		spin_unlock_bh(&net->sctp.addr_wq_lock);
724 725 726
		return;
	}
	addrw->state = cmd;
727
	list_add_tail(&addrw->list, &net->sctp.addr_waitq);
728
	SCTP_DEBUG_PRINTK_IPADDR("sctp_addr_wq_mgmt add new entry for cmd:%d ",
729
	    " in wq %p\n", addrw->state, &addrw->a, &net->sctp.addr_waitq);
730

731
	if (!timer_pending(&net->sctp.addr_wq_timer)) {
732 733
		timeo_val = jiffies;
		timeo_val += msecs_to_jiffies(SCTP_ADDRESS_TICK_DELAY);
734
		mod_timer(&net->sctp.addr_wq_timer, timeo_val);
735
	}
736
	spin_unlock_bh(&net->sctp.addr_wq_lock);
737 738
}

739 740 741 742 743 744
/* Event handler for inet address addition/deletion events.
 * The sctp_local_addr_list needs to be protocted by a spin lock since
 * multiple notifiers (say IPv4 and IPv6) may be running at the same
 * time and thus corrupt the list.
 * The reader side is protected with RCU.
 */
A
Adrian Bunk 已提交
745 746
static int sctp_inetaddr_event(struct notifier_block *this, unsigned long ev,
			       void *ptr)
L
Linus Torvalds 已提交
747
{
748
	struct in_ifaddr *ifa = (struct in_ifaddr *)ptr;
749 750
	struct sctp_sockaddr_entry *addr = NULL;
	struct sctp_sockaddr_entry *temp;
751
	struct net *net = dev_net(ifa->ifa_dev->dev);
752
	int found = 0;
L
Linus Torvalds 已提交
753

754 755 756 757 758 759 760
	switch (ev) {
	case NETDEV_UP:
		addr = kmalloc(sizeof(struct sctp_sockaddr_entry), GFP_ATOMIC);
		if (addr) {
			addr->a.v4.sin_family = AF_INET;
			addr->a.v4.sin_port = 0;
			addr->a.v4.sin_addr.s_addr = ifa->ifa_local;
761
			addr->valid = 1;
762 763 764 765
			spin_lock_bh(&net->sctp.local_addr_lock);
			list_add_tail_rcu(&addr->list, &net->sctp.local_addr_list);
			sctp_addr_wq_mgmt(net, addr, SCTP_ADDR_NEW);
			spin_unlock_bh(&net->sctp.local_addr_lock);
766 767 768
		}
		break;
	case NETDEV_DOWN:
769
		spin_lock_bh(&net->sctp.local_addr_lock);
770
		list_for_each_entry_safe(addr, temp,
771
					&net->sctp.local_addr_list, list) {
772 773 774
			if (addr->a.sa.sa_family == AF_INET &&
					addr->a.v4.sin_addr.s_addr ==
					ifa->ifa_local) {
775
				sctp_addr_wq_mgmt(net, addr, SCTP_ADDR_DEL);
776
				found = 1;
777 778
				addr->valid = 0;
				list_del_rcu(&addr->list);
779 780 781
				break;
			}
		}
782
		spin_unlock_bh(&net->sctp.local_addr_lock);
783
		if (found)
784
			kfree_rcu(addr, rcu);
785 786
		break;
	}
L
Linus Torvalds 已提交
787 788 789 790 791 792 793 794

	return NOTIFY_DONE;
}

/*
 * Initialize the control inode/socket with a control endpoint data
 * structure.  This endpoint is reserved exclusively for the OOTB processing.
 */
795
static int sctp_ctl_sock_init(struct net *net)
L
Linus Torvalds 已提交
796 797
{
	int err;
798
	sa_family_t family = PF_INET;
L
Linus Torvalds 已提交
799 800 801 802

	if (sctp_get_pf_specific(PF_INET6))
		family = PF_INET6;

803 804
	err = inet_ctl_sock_create(&net->sctp.ctl_sock, family,
				   SOCK_SEQPACKET, IPPROTO_SCTP, net);
805 806 807

	/* If IPv6 socket could not be created, try the IPv4 socket */
	if (err < 0 && family == PF_INET6)
808
		err = inet_ctl_sock_create(&net->sctp.ctl_sock, AF_INET,
809
					   SOCK_SEQPACKET, IPPROTO_SCTP,
810
					   net);
811

L
Linus Torvalds 已提交
812
	if (err < 0) {
813
		pr_err("Failed to create the SCTP control socket\n");
L
Linus Torvalds 已提交
814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889
		return err;
	}
	return 0;
}

/* Register address family specific functions. */
int sctp_register_af(struct sctp_af *af)
{
	switch (af->sa_family) {
	case AF_INET:
		if (sctp_af_v4_specific)
			return 0;
		sctp_af_v4_specific = af;
		break;
	case AF_INET6:
		if (sctp_af_v6_specific)
			return 0;
		sctp_af_v6_specific = af;
		break;
	default:
		return 0;
	}

	INIT_LIST_HEAD(&af->list);
	list_add_tail(&af->list, &sctp_address_families);
	return 1;
}

/* Get the table of functions for manipulating a particular address
 * family.
 */
struct sctp_af *sctp_get_af_specific(sa_family_t family)
{
	switch (family) {
	case AF_INET:
		return sctp_af_v4_specific;
	case AF_INET6:
		return sctp_af_v6_specific;
	default:
		return NULL;
	}
}

/* Common code to initialize a AF_INET msg_name. */
static void sctp_inet_msgname(char *msgname, int *addr_len)
{
	struct sockaddr_in *sin;

	sin = (struct sockaddr_in *)msgname;
	*addr_len = sizeof(struct sockaddr_in);
	sin->sin_family = AF_INET;
	memset(sin->sin_zero, 0, sizeof(sin->sin_zero));
}

/* Copy the primary address of the peer primary address as the msg_name. */
static void sctp_inet_event_msgname(struct sctp_ulpevent *event, char *msgname,
				    int *addr_len)
{
	struct sockaddr_in *sin, *sinfrom;

	if (msgname) {
		struct sctp_association *asoc;

		asoc = event->asoc;
		sctp_inet_msgname(msgname, addr_len);
		sin = (struct sockaddr_in *)msgname;
		sinfrom = &asoc->peer.primary_addr.v4;
		sin->sin_port = htons(asoc->peer.port);
		sin->sin_addr.s_addr = sinfrom->sin_addr.s_addr;
	}
}

/* Initialize and copy out a msgname from an inbound skb. */
static void sctp_inet_skb_msgname(struct sk_buff *skb, char *msgname, int *len)
{
	if (msgname) {
890 891 892
		struct sctphdr *sh = sctp_hdr(skb);
		struct sockaddr_in *sin = (struct sockaddr_in *)msgname;

L
Linus Torvalds 已提交
893 894
		sctp_inet_msgname(msgname, len);
		sin->sin_port = sh->source;
895
		sin->sin_addr.s_addr = ip_hdr(skb)->saddr;
L
Linus Torvalds 已提交
896 897 898 899 900 901 902
	}
}

/* Do we support this AF? */
static int sctp_inet_af_supported(sa_family_t family, struct sctp_sock *sp)
{
	/* PF_INET only supports AF_INET addresses. */
E
Eric Dumazet 已提交
903
	return AF_INET == family;
L
Linus Torvalds 已提交
904 905 906 907 908 909 910 911 912 913
}

/* Address matching with wildcards allowed. */
static int sctp_inet_cmp_addr(const union sctp_addr *addr1,
			      const union sctp_addr *addr2,
			      struct sctp_sock *opt)
{
	/* PF_INET only supports AF_INET addresses. */
	if (addr1->sa.sa_family != addr2->sa.sa_family)
		return 0;
A
Al Viro 已提交
914 915
	if (htonl(INADDR_ANY) == addr1->v4.sin_addr.s_addr ||
	    htonl(INADDR_ANY) == addr2->v4.sin_addr.s_addr)
L
Linus Torvalds 已提交
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
		return 1;
	if (addr1->v4.sin_addr.s_addr == addr2->v4.sin_addr.s_addr)
		return 1;

	return 0;
}

/* Verify that provided sockaddr looks bindable.  Common verification has
 * already been taken care of.
 */
static int sctp_inet_bind_verify(struct sctp_sock *opt, union sctp_addr *addr)
{
	return sctp_v4_available(addr, opt);
}

/* Verify that sockaddr looks sendable.  Common verification has already
 * been taken care of.
 */
static int sctp_inet_send_verify(struct sctp_sock *opt, union sctp_addr *addr)
{
	return 1;
}

/* Fill in Supported Address Type information for INIT and INIT-ACK
 * chunks.  Returns number of addresses supported.
 */
static int sctp_inet_supported_addrs(const struct sctp_sock *opt,
A
Al Viro 已提交
943
				     __be16 *types)
L
Linus Torvalds 已提交
944 945 946 947 948 949 950
{
	types[0] = SCTP_PARAM_IPV4_ADDRESS;
	return 1;
}

/* Wrapper routine that calls the ip transmit routine. */
static inline int sctp_v4_xmit(struct sk_buff *skb,
951
			       struct sctp_transport *transport)
L
Linus Torvalds 已提交
952
{
953 954
	struct inet_sock *inet = inet_sk(skb->sk);

H
Harvey Harrison 已提交
955
	SCTP_DEBUG_PRINTK("%s: skb:%p, len:%d, src:%pI4, dst:%pI4\n",
956
			  __func__, skb, skb->len,
D
David S. Miller 已提交
957 958
			  &transport->fl.u.ip4.saddr,
			  &transport->fl.u.ip4.daddr);
L
Linus Torvalds 已提交
959

960 961 962
	inet->pmtudisc = transport->param_flags & SPP_PMTUD_ENABLE ?
			 IP_PMTUDISC_DO : IP_PMTUDISC_DONT;

963
	SCTP_INC_STATS(sock_net(&inet->sk), SCTP_MIB_OUTSCTPPACKS);
964
	return ip_queue_xmit(skb, &transport->fl);
L
Linus Torvalds 已提交
965 966
}

967
static struct sctp_af sctp_af_inet;
L
Linus Torvalds 已提交
968 969 970 971 972 973 974 975 976 977 978

static struct sctp_pf sctp_pf_inet = {
	.event_msgname = sctp_inet_event_msgname,
	.skb_msgname   = sctp_inet_skb_msgname,
	.af_supported  = sctp_inet_af_supported,
	.cmp_addr      = sctp_inet_cmp_addr,
	.bind_verify   = sctp_inet_bind_verify,
	.send_verify   = sctp_inet_send_verify,
	.supported_addrs = sctp_inet_supported_addrs,
	.create_accept_sk = sctp_v4_create_accept_sk,
	.addr_v4map	= sctp_v4_addr_v4map,
979
	.af            = &sctp_af_inet
L
Linus Torvalds 已提交
980 981 982 983 984 985 986 987
};

/* Notifier for inetaddr addition/deletion events.  */
static struct notifier_block sctp_inetaddr_notifier = {
	.notifier_call = sctp_inetaddr_event,
};

/* Socket operations.  */
988
static const struct proto_ops inet_seqpacket_ops = {
989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006
	.family		   = PF_INET,
	.owner		   = THIS_MODULE,
	.release	   = inet_release,	/* Needs to be wrapped... */
	.bind		   = inet_bind,
	.connect	   = inet_dgram_connect,
	.socketpair	   = sock_no_socketpair,
	.accept		   = inet_accept,
	.getname	   = inet_getname,	/* Semantics are different.  */
	.poll		   = sctp_poll,
	.ioctl		   = inet_ioctl,
	.listen		   = sctp_inet_listen,
	.shutdown	   = inet_shutdown,	/* Looks harmless.  */
	.setsockopt	   = sock_common_setsockopt, /* IP_SOL IP_OPTION is a problem */
	.getsockopt	   = sock_common_getsockopt,
	.sendmsg	   = inet_sendmsg,
	.recvmsg	   = sock_common_recvmsg,
	.mmap		   = sock_no_mmap,
	.sendpage	   = sock_no_sendpage,
1007
#ifdef CONFIG_COMPAT
1008 1009
	.compat_setsockopt = compat_sock_common_setsockopt,
	.compat_getsockopt = compat_sock_common_getsockopt,
1010
#endif
L
Linus Torvalds 已提交
1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
};

/* Registration with AF_INET family.  */
static struct inet_protosw sctp_seqpacket_protosw = {
	.type       = SOCK_SEQPACKET,
	.protocol   = IPPROTO_SCTP,
	.prot       = &sctp_prot,
	.ops        = &inet_seqpacket_ops,
	.no_check   = 0,
	.flags      = SCTP_PROTOSW_FLAG
};
static struct inet_protosw sctp_stream_protosw = {
	.type       = SOCK_STREAM,
	.protocol   = IPPROTO_SCTP,
	.prot       = &sctp_prot,
	.ops        = &inet_seqpacket_ops,
	.no_check   = 0,
	.flags      = SCTP_PROTOSW_FLAG
};

/* Register with IP layer.  */
1032
static const struct net_protocol sctp_protocol = {
L
Linus Torvalds 已提交
1033 1034 1035
	.handler     = sctp_rcv,
	.err_handler = sctp_v4_err,
	.no_policy   = 1,
1036
	.netns_ok    = 1,
L
Linus Torvalds 已提交
1037 1038 1039
};

/* IPv4 address related functions.  */
1040
static struct sctp_af sctp_af_inet = {
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
	.sa_family	   = AF_INET,
	.sctp_xmit	   = sctp_v4_xmit,
	.setsockopt	   = ip_setsockopt,
	.getsockopt	   = ip_getsockopt,
	.get_dst	   = sctp_v4_get_dst,
	.get_saddr	   = sctp_v4_get_saddr,
	.copy_addrlist	   = sctp_v4_copy_addrlist,
	.from_skb	   = sctp_v4_from_skb,
	.from_sk	   = sctp_v4_from_sk,
	.to_sk_saddr	   = sctp_v4_to_sk_saddr,
	.to_sk_daddr	   = sctp_v4_to_sk_daddr,
	.from_addr_param   = sctp_v4_from_addr_param,
	.to_addr_param	   = sctp_v4_to_addr_param,
	.cmp_addr	   = sctp_v4_cmp_addr,
	.addr_valid	   = sctp_v4_addr_valid,
	.inaddr_any	   = sctp_v4_inaddr_any,
	.is_any		   = sctp_v4_is_any,
	.available	   = sctp_v4_available,
	.scope		   = sctp_v4_scope,
	.skb_iif	   = sctp_v4_skb_iif,
	.is_ce		   = sctp_v4_is_ce,
	.seq_dump_addr	   = sctp_v4_seq_dump_addr,
V
Vlad Yasevich 已提交
1063
	.ecn_capable	   = sctp_v4_ecn_capable,
1064 1065
	.net_header_len	   = sizeof(struct iphdr),
	.sockaddr_len	   = sizeof(struct sockaddr_in),
1066
#ifdef CONFIG_COMPAT
1067 1068
	.compat_setsockopt = compat_ip_setsockopt,
	.compat_getsockopt = compat_ip_getsockopt,
1069
#endif
L
Linus Torvalds 已提交
1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103
};

struct sctp_pf *sctp_get_pf_specific(sa_family_t family) {

	switch (family) {
	case PF_INET:
		return sctp_pf_inet_specific;
	case PF_INET6:
		return sctp_pf_inet6_specific;
	default:
		return NULL;
	}
}

/* Register the PF specific function table.  */
int sctp_register_pf(struct sctp_pf *pf, sa_family_t family)
{
	switch (family) {
	case PF_INET:
		if (sctp_pf_inet_specific)
			return 0;
		sctp_pf_inet_specific = pf;
		break;
	case PF_INET6:
		if (sctp_pf_inet6_specific)
			return 0;
		sctp_pf_inet6_specific = pf;
		break;
	default:
		return 0;
	}
	return 1;
}

1104
static inline int init_sctp_mibs(struct net *net)
1105
{
1106
	return snmp_mib_init((void __percpu **)net->sctp.sctp_statistics,
1107 1108
			     sizeof(struct sctp_mib),
			     __alignof__(struct sctp_mib));
L
Linus Torvalds 已提交
1109 1110
}

1111
static inline void cleanup_sctp_mibs(struct net *net)
L
Linus Torvalds 已提交
1112
{
1113
	snmp_mib_free((void __percpu **)net->sctp.sctp_statistics);
L
Linus Torvalds 已提交
1114 1115
}

1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167
static void sctp_v4_pf_init(void)
{
	/* Initialize the SCTP specific PF functions. */
	sctp_register_pf(&sctp_pf_inet, PF_INET);
	sctp_register_af(&sctp_af_inet);
}

static void sctp_v4_pf_exit(void)
{
	list_del(&sctp_af_inet.list);
}

static int sctp_v4_protosw_init(void)
{
	int rc;

	rc = proto_register(&sctp_prot, 1);
	if (rc)
		return rc;

	/* Register SCTP(UDP and TCP style) with socket layer.  */
	inet_register_protosw(&sctp_seqpacket_protosw);
	inet_register_protosw(&sctp_stream_protosw);

	return 0;
}

static void sctp_v4_protosw_exit(void)
{
	inet_unregister_protosw(&sctp_stream_protosw);
	inet_unregister_protosw(&sctp_seqpacket_protosw);
	proto_unregister(&sctp_prot);
}

static int sctp_v4_add_protocol(void)
{
	/* Register notifier for inet address additions/deletions. */
	register_inetaddr_notifier(&sctp_inetaddr_notifier);

	/* Register SCTP with inet layer.  */
	if (inet_add_protocol(&sctp_protocol, IPPROTO_SCTP) < 0)
		return -EAGAIN;

	return 0;
}

static void sctp_v4_del_protocol(void)
{
	inet_del_protocol(&sctp_protocol, IPPROTO_SCTP);
	unregister_inetaddr_notifier(&sctp_inetaddr_notifier);
}

1168 1169
static int sctp_net_init(struct net *net)
{
1170 1171
	int status;

1172 1173 1174 1175
	status = sctp_sysctl_net_register(net);
	if (status)
		goto err_sysctl_register;

1176 1177 1178 1179 1180
	/* Allocate and initialise sctp mibs.  */
	status = init_sctp_mibs(net);
	if (status)
		goto err_init_mibs;

1181 1182 1183 1184 1185 1186 1187
	/* Initialize proc fs directory.  */
	status = sctp_proc_init(net);
	if (status)
		goto err_init_proc;

	sctp_dbg_objcnt_init(net);

1188 1189 1190 1191 1192 1193
	/* Initialize the control inode/socket for handling OOTB packets.  */
	if ((status = sctp_ctl_sock_init(net))) {
		pr_err("Failed to initialize the SCTP control sock\n");
		goto err_ctl_sock_init;
	}

1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207
	/* Initialize the local address list. */
	INIT_LIST_HEAD(&net->sctp.local_addr_list);
	spin_lock_init(&net->sctp.local_addr_lock);
	sctp_get_local_addr_list(net);

	/* Initialize the address event list */
	INIT_LIST_HEAD(&net->sctp.addr_waitq);
	INIT_LIST_HEAD(&net->sctp.auto_asconf_splist);
	spin_lock_init(&net->sctp.addr_wq_lock);
	net->sctp.addr_wq_timer.expires = 0;
	setup_timer(&net->sctp.addr_wq_timer, sctp_addr_wq_timeout_handler,
		    (unsigned long)net);

	return 0;
1208 1209

err_ctl_sock_init:
1210 1211 1212
	sctp_dbg_objcnt_exit(net);
	sctp_proc_exit(net);
err_init_proc:
1213 1214
	cleanup_sctp_mibs(net);
err_init_mibs:
1215 1216
	sctp_sysctl_net_unregister(net);
err_sysctl_register:
1217
	return status;
1218 1219 1220 1221 1222 1223 1224
}

static void sctp_net_exit(struct net *net)
{
	/* Free the local address list */
	sctp_free_addr_wq(net);
	sctp_free_local_addr_list(net);
1225 1226 1227

	/* Free the control endpoint.  */
	inet_ctl_sock_destroy(net->sctp.ctl_sock);
1228 1229 1230 1231

	sctp_dbg_objcnt_exit(net);

	sctp_proc_exit(net);
1232
	cleanup_sctp_mibs(net);
1233
	sctp_sysctl_net_unregister(net);
1234 1235 1236 1237 1238 1239 1240
}

static struct pernet_operations sctp_net_ops = {
	.init = sctp_net_init,
	.exit = sctp_net_exit,
};

L
Linus Torvalds 已提交
1241 1242 1243 1244 1245 1246
/* Initialize the universe into something sensible.  */
SCTP_STATIC __init int sctp_init(void)
{
	int i;
	int status = -EINVAL;
	unsigned long goal;
1247 1248
	unsigned long limit;
	int max_share;
L
Linus Torvalds 已提交
1249 1250 1251 1252 1253 1254
	int order;

	/* SCTP_DEBUG sanity check. */
	if (!sctp_sanity_check())
		goto out;

1255
	/* Allocate bind_bucket and chunk caches. */
L
Linus Torvalds 已提交
1256 1257 1258 1259
	status = -ENOBUFS;
	sctp_bucket_cachep = kmem_cache_create("sctp_bind_bucket",
					       sizeof(struct sctp_bind_bucket),
					       0, SLAB_HWCACHE_ALIGN,
1260
					       NULL);
L
Linus Torvalds 已提交
1261
	if (!sctp_bucket_cachep)
1262
		goto out;
L
Linus Torvalds 已提交
1263 1264 1265 1266

	sctp_chunk_cachep = kmem_cache_create("sctp_chunk",
					       sizeof(struct sctp_chunk),
					       0, SLAB_HWCACHE_ALIGN,
1267
					       NULL);
L
Linus Torvalds 已提交
1268 1269 1270
	if (!sctp_chunk_cachep)
		goto err_chunk_cachep;

1271 1272 1273 1274
	status = percpu_counter_init(&sctp_sockets_allocated, 0);
	if (status)
		goto err_percpu_counter_init;

L
Linus Torvalds 已提交
1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290
	/*
	 * 14. Suggested SCTP Protocol Parameter Values
	 */
	/* The following protocol parameters are RECOMMENDED:  */
	/* RTO.Initial              - 3  seconds */
	sctp_rto_initial		= SCTP_RTO_INITIAL;
	/* RTO.Min                  - 1  second */
	sctp_rto_min	 		= SCTP_RTO_MIN;
	/* RTO.Max                 -  60 seconds */
	sctp_rto_max 			= SCTP_RTO_MAX;
	/* RTO.Alpha                - 1/8 */
	sctp_rto_alpha	        	= SCTP_RTO_ALPHA;
	/* RTO.Beta                 - 1/4 */
	sctp_rto_beta			= SCTP_RTO_BETA;

	/* Valid.Cookie.Life        - 60  seconds */
1291
	sctp_valid_cookie_life		= SCTP_DEFAULT_COOKIE_LIFE;
L
Linus Torvalds 已提交
1292 1293 1294 1295 1296

	/* Whether Cookie Preservative is enabled(1) or not(0) */
	sctp_cookie_preserve_enable 	= 1;

	/* Max.Burst		    - 4 */
1297
	sctp_max_burst 			= SCTP_DEFAULT_MAX_BURST;
L
Linus Torvalds 已提交
1298 1299 1300 1301 1302 1303 1304 1305 1306

	/* Association.Max.Retrans  - 10 attempts
	 * Path.Max.Retrans         - 5  attempts (per destination address)
	 * Max.Init.Retransmits     - 8  attempts
	 */
	sctp_max_retrans_association 	= 10;
	sctp_max_retrans_path		= 5;
	sctp_max_retrans_init		= 8;

1307 1308 1309
	/* Sendbuffer growth	    - do per-socket accounting */
	sctp_sndbuf_policy		= 0;

1310 1311 1312
	/* Rcvbuffer growth	    - do per-socket accounting */
	sctp_rcvbuf_policy		= 0;

L
Linus Torvalds 已提交
1313
	/* HB.interval              - 30 seconds */
1314 1315 1316 1317
	sctp_hb_interval		= SCTP_DEFAULT_TIMEOUT_HEARTBEAT;

	/* delayed SACK timeout */
	sctp_sack_timeout		= SCTP_DEFAULT_TIMEOUT_SACK;
L
Linus Torvalds 已提交
1318 1319 1320 1321 1322 1323 1324

	/* Implementation specific variables. */

	/* Initialize default stream count setup information. */
	sctp_max_instreams    		= SCTP_DEFAULT_INSTREAMS;
	sctp_max_outstreams   		= SCTP_DEFAULT_OUTSTREAMS;

1325 1326 1327
	/* Initialize maximum autoclose timeout. */
	sctp_max_autoclose		= INT_MAX / HZ;

L
Linus Torvalds 已提交
1328 1329 1330
	/* Initialize handle used for association ids. */
	idr_init(&sctp_assocs_id);

1331
	limit = nr_free_buffer_pages() / 8;
1332 1333 1334 1335 1336 1337 1338 1339 1340
	limit = max(limit, 128UL);
	sysctl_sctp_mem[0] = limit / 4 * 3;
	sysctl_sctp_mem[1] = limit;
	sysctl_sctp_mem[2] = sysctl_sctp_mem[0] * 2;

	/* Set per-socket limits to no more than 1/128 the pressure threshold*/
	limit = (sysctl_sctp_mem[1]) << (PAGE_SHIFT - 7);
	max_share = min(4UL*1024*1024, limit);

1341
	sysctl_sctp_rmem[0] = SK_MEM_QUANTUM; /* give each asoc 1 page min */
E
Eric Dumazet 已提交
1342
	sysctl_sctp_rmem[1] = 1500 * SKB_TRUESIZE(1);
1343 1344
	sysctl_sctp_rmem[2] = max(sysctl_sctp_rmem[1], max_share);

1345
	sysctl_sctp_wmem[0] = SK_MEM_QUANTUM;
1346 1347 1348
	sysctl_sctp_wmem[1] = 16*1024;
	sysctl_sctp_wmem[2] = max(64*1024, max_share);

L
Linus Torvalds 已提交
1349 1350 1351
	/* Size and allocate the association hash table.
	 * The methodology is similar to that of the tcp hash tables.
	 */
1352 1353
	if (totalram_pages >= (128 * 1024))
		goal = totalram_pages >> (22 - PAGE_SHIFT);
L
Linus Torvalds 已提交
1354
	else
1355
		goal = totalram_pages >> (24 - PAGE_SHIFT);
L
Linus Torvalds 已提交
1356 1357 1358 1359 1360 1361 1362 1363 1364 1365

	for (order = 0; (1UL << order) < goal; order++)
		;

	do {
		sctp_assoc_hashsize = (1UL << order) * PAGE_SIZE /
					sizeof(struct sctp_hashbucket);
		if ((sctp_assoc_hashsize > (64 * 1024)) && order > 0)
			continue;
		sctp_assoc_hashtable = (struct sctp_hashbucket *)
1366
			__get_free_pages(GFP_ATOMIC|__GFP_NOWARN, order);
L
Linus Torvalds 已提交
1367 1368
	} while (!sctp_assoc_hashtable && --order > 0);
	if (!sctp_assoc_hashtable) {
1369
		pr_err("Failed association hash alloc\n");
L
Linus Torvalds 已提交
1370 1371 1372 1373 1374
		status = -ENOMEM;
		goto err_ahash_alloc;
	}
	for (i = 0; i < sctp_assoc_hashsize; i++) {
		rwlock_init(&sctp_assoc_hashtable[i].lock);
1375
		INIT_HLIST_HEAD(&sctp_assoc_hashtable[i].chain);
L
Linus Torvalds 已提交
1376 1377 1378 1379 1380 1381 1382
	}

	/* Allocate and initialize the endpoint hash table.  */
	sctp_ep_hashsize = 64;
	sctp_ep_hashtable = (struct sctp_hashbucket *)
		kmalloc(64 * sizeof(struct sctp_hashbucket), GFP_KERNEL);
	if (!sctp_ep_hashtable) {
1383
		pr_err("Failed endpoint_hash alloc\n");
L
Linus Torvalds 已提交
1384 1385 1386 1387 1388
		status = -ENOMEM;
		goto err_ehash_alloc;
	}
	for (i = 0; i < sctp_ep_hashsize; i++) {
		rwlock_init(&sctp_ep_hashtable[i].lock);
1389
		INIT_HLIST_HEAD(&sctp_ep_hashtable[i].chain);
L
Linus Torvalds 已提交
1390 1391 1392 1393 1394 1395 1396 1397 1398
	}

	/* Allocate and initialize the SCTP port hash table.  */
	do {
		sctp_port_hashsize = (1UL << order) * PAGE_SIZE /
					sizeof(struct sctp_bind_hashbucket);
		if ((sctp_port_hashsize > (64 * 1024)) && order > 0)
			continue;
		sctp_port_hashtable = (struct sctp_bind_hashbucket *)
1399
			__get_free_pages(GFP_ATOMIC|__GFP_NOWARN, order);
L
Linus Torvalds 已提交
1400 1401
	} while (!sctp_port_hashtable && --order > 0);
	if (!sctp_port_hashtable) {
1402
		pr_err("Failed bind hash alloc\n");
L
Linus Torvalds 已提交
1403 1404 1405 1406 1407
		status = -ENOMEM;
		goto err_bhash_alloc;
	}
	for (i = 0; i < sctp_port_hashsize; i++) {
		spin_lock_init(&sctp_port_hashtable[i].lock);
1408
		INIT_HLIST_HEAD(&sctp_port_hashtable[i].chain);
L
Linus Torvalds 已提交
1409 1410
	}

1411
	pr_info("Hash tables configured (established %d bind %d)\n",
L
Linus Torvalds 已提交
1412 1413 1414 1415
		sctp_assoc_hashsize, sctp_port_hashsize);

	/* Disable ADDIP by default. */
	sctp_addip_enable = 0;
1416
	sctp_addip_noauth = 0;
1417
	sctp_default_auto_asconf = 0;
L
Linus Torvalds 已提交
1418 1419 1420 1421

	/* Enable PR-SCTP by default. */
	sctp_prsctp_enable = 1;

1422 1423 1424
	/* Disable AUTH by default. */
	sctp_auth_enable = 0;

1425 1426 1427
	/* Set SCOPE policy to enabled */
	sctp_scope_policy = SCTP_SCOPE_POLICY_ENABLE;

1428 1429 1430
	/* Set the default rwnd update threshold */
	sctp_rwnd_upd_shift		= SCTP_DEFAULT_RWND_SHIFT;

L
Linus Torvalds 已提交
1431 1432 1433
	sctp_sysctl_register();

	INIT_LIST_HEAD(&sctp_address_families);
1434 1435
	sctp_v4_pf_init();
	sctp_v6_pf_init();
L
Linus Torvalds 已提交
1436

1437
	status = sctp_v4_protosw_init();
1438

L
Linus Torvalds 已提交
1439
	if (status)
1440 1441 1442 1443 1444
		goto err_protosw_init;

	status = sctp_v6_protosw_init();
	if (status)
		goto err_v6_protosw_init;
L
Linus Torvalds 已提交
1445

1446 1447 1448 1449
	status = register_pernet_subsys(&sctp_net_ops);
	if (status)
		goto err_register_pernet_subsys;

1450 1451
	status = sctp_v4_add_protocol();
	if (status)
1452 1453 1454 1455 1456 1457 1458
		goto err_add_protocol;

	/* Register SCTP with inet6 layer.  */
	status = sctp_v6_add_protocol();
	if (status)
		goto err_v6_add_protocol;

L
Linus Torvalds 已提交
1459 1460 1461
	status = 0;
out:
	return status;
1462
err_v6_add_protocol:
1463
	sctp_v4_del_protocol();
1464
err_add_protocol:
1465 1466
	unregister_pernet_subsys(&sctp_net_ops);
err_register_pernet_subsys:
1467 1468 1469 1470 1471 1472
	sctp_v6_protosw_exit();
err_v6_protosw_init:
	sctp_v4_protosw_exit();
err_protosw_init:
	sctp_v4_pf_exit();
	sctp_v6_pf_exit();
L
Linus Torvalds 已提交
1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483
	sctp_sysctl_unregister();
	free_pages((unsigned long)sctp_port_hashtable,
		   get_order(sctp_port_hashsize *
			     sizeof(struct sctp_bind_hashbucket)));
err_bhash_alloc:
	kfree(sctp_ep_hashtable);
err_ehash_alloc:
	free_pages((unsigned long)sctp_assoc_hashtable,
		   get_order(sctp_assoc_hashsize *
			     sizeof(struct sctp_hashbucket)));
err_ahash_alloc:
1484 1485
	percpu_counter_destroy(&sctp_sockets_allocated);
err_percpu_counter_init:
L
Linus Torvalds 已提交
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
	kmem_cache_destroy(sctp_chunk_cachep);
err_chunk_cachep:
	kmem_cache_destroy(sctp_bucket_cachep);
	goto out;
}

/* Exit handler for the SCTP protocol.  */
SCTP_STATIC __exit void sctp_exit(void)
{
	/* BUG.  This should probably do something useful like clean
	 * up all the remaining associations and all that memory.
	 */

1499 1500
	/* Unregister with inet6/inet layers. */
	sctp_v6_del_protocol();
1501
	sctp_v4_del_protocol();
L
Linus Torvalds 已提交
1502

1503 1504
	unregister_pernet_subsys(&sctp_net_ops);

1505 1506 1507 1508
	/* Free protosw registrations */
	sctp_v6_protosw_exit();
	sctp_v4_protosw_exit();

1509
	/* Unregister with socket layer. */
1510 1511
	sctp_v6_pf_exit();
	sctp_v4_pf_exit();
1512

L
Linus Torvalds 已提交
1513 1514 1515 1516 1517 1518 1519 1520 1521 1522
	sctp_sysctl_unregister();

	free_pages((unsigned long)sctp_assoc_hashtable,
		   get_order(sctp_assoc_hashsize *
			     sizeof(struct sctp_hashbucket)));
	kfree(sctp_ep_hashtable);
	free_pages((unsigned long)sctp_port_hashtable,
		   get_order(sctp_port_hashsize *
			     sizeof(struct sctp_bind_hashbucket)));

1523
	percpu_counter_destroy(&sctp_sockets_allocated);
L
Linus Torvalds 已提交
1524

1525 1526
	rcu_barrier(); /* Wait for completion of call_rcu()'s */

1527 1528
	kmem_cache_destroy(sctp_chunk_cachep);
	kmem_cache_destroy(sctp_bucket_cachep);
L
Linus Torvalds 已提交
1529 1530 1531 1532 1533
}

module_init(sctp_init);
module_exit(sctp_exit);

1534 1535 1536 1537
/*
 * __stringify doesn't likes enums, so use IPPROTO_SCTP value (132) directly.
 */
MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-proto-132");
1538
MODULE_ALIAS("net-pf-" __stringify(PF_INET6) "-proto-132");
L
Linus Torvalds 已提交
1539 1540
MODULE_AUTHOR("Linux Kernel SCTP developers <lksctp-developers@lists.sourceforge.net>");
MODULE_DESCRIPTION("Support for the SCTP protocol (RFC2960)");
1541 1542
module_param_named(no_checksums, sctp_checksum_disable, bool, 0644);
MODULE_PARM_DESC(no_checksums, "Disable checksums computing and verification");
L
Linus Torvalds 已提交
1543
MODULE_LICENSE("GPL");