socket.c 199.8 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) 2001-2003 Intel Corp.
 * Copyright (c) 2001-2002 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 13 14 15 16 17
 *
 * These functions interface with the sockets layer to implement the
 * SCTP Extensions for the Sockets API.
 *
 * Note that the descriptions from the specification are USER level
 * functions--this file is the functions which populate the struct proto
 * for SCTP which is the BOTTOM of the sockets interface.
 *
18
 * This SCTP implementation is free software;
L
Linus Torvalds 已提交
19 20 21 22 23
 * 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.
 *
24
 * This SCTP implementation is distributed in the hope that it
L
Linus Torvalds 已提交
25 26 27 28 29 30
 * 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
31 32
 * along with GNU CC; see the file COPYING.  If not, see
 * <http://www.gnu.org/licenses/>.
L
Linus Torvalds 已提交
33 34 35
 *
 * Please send any bug reports or fixes you make to the
 * email address(es):
36
 *    lksctp developers <linux-sctp@vger.kernel.org>
L
Linus Torvalds 已提交
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
 *
 * Written or modified by:
 *    La Monte H.P. Yarroll <piggy@acm.org>
 *    Narasimha Budihal     <narsi@refcode.org>
 *    Karl Knutson          <karl@athena.chicago.il.us>
 *    Jon Grimm             <jgrimm@us.ibm.com>
 *    Xingang Guo           <xingang.guo@intel.com>
 *    Daisy Chang           <daisyc@us.ibm.com>
 *    Sridhar Samudrala     <samudrala@us.ibm.com>
 *    Inaky Perez-Gonzalez  <inaky.gonzalez@intel.com>
 *    Ardelle Fan	    <ardelle.fan@intel.com>
 *    Ryan Layer	    <rmlayer@us.ibm.com>
 *    Anup Pemmaiah         <pemmaiah@cc.usu.edu>
 *    Kevin Gao             <kevin.gao@intel.com>
 */

53 54
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

L
Linus Torvalds 已提交
55 56 57 58 59
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/wait.h>
#include <linux/time.h>
#include <linux/ip.h>
60
#include <linux/capability.h>
L
Linus Torvalds 已提交
61 62 63 64
#include <linux/fcntl.h>
#include <linux/poll.h>
#include <linux/init.h>
#include <linux/crypto.h>
65
#include <linux/slab.h>
66
#include <linux/file.h>
L
Linus Torvalds 已提交
67 68 69 70 71 72 73 74

#include <net/ip.h>
#include <net/icmp.h>
#include <net/route.h>
#include <net/ipv6.h>
#include <net/inet_common.h>

#include <linux/socket.h> /* for sa_family_t */
75
#include <linux/export.h>
L
Linus Torvalds 已提交
76 77 78 79 80 81 82 83 84 85 86 87 88
#include <net/sock.h>
#include <net/sctp/sctp.h>
#include <net/sctp/sm.h>

/* Forward declarations for internal helper functions. */
static int sctp_writeable(struct sock *sk);
static void sctp_wfree(struct sk_buff *skb);
static int sctp_wait_for_sndbuf(struct sctp_association *, long *timeo_p,
				size_t msg_len);
static int sctp_wait_for_packet(struct sock * sk, int *err, long *timeo_p);
static int sctp_wait_for_connect(struct sctp_association *, long *timeo_p);
static int sctp_wait_for_accept(struct sock *sk, long timeo);
static void sctp_wait_for_close(struct sock *sk, long timeo);
89
static void sctp_destruct_sock(struct sock *sk);
L
Linus Torvalds 已提交
90 91 92 93 94 95 96 97 98 99 100 101 102
static struct sctp_af *sctp_sockaddr_af(struct sctp_sock *opt,
					union sctp_addr *addr, int len);
static int sctp_bindx_add(struct sock *, struct sockaddr *, int);
static int sctp_bindx_rem(struct sock *, struct sockaddr *, int);
static int sctp_send_asconf_add_ip(struct sock *, struct sockaddr *, int);
static int sctp_send_asconf_del_ip(struct sock *, struct sockaddr *, int);
static int sctp_send_asconf(struct sctp_association *asoc,
			    struct sctp_chunk *chunk);
static int sctp_do_bind(struct sock *, union sctp_addr *, int);
static int sctp_autobind(struct sock *sk);
static void sctp_sock_migrate(struct sock *, struct sock *,
			      struct sctp_association *, sctp_socket_type_t);

103
extern struct kmem_cache *sctp_bucket_cachep;
E
Eric Dumazet 已提交
104
extern long sysctl_sctp_mem[3];
105 106 107
extern int sysctl_sctp_rmem[3];
extern int sysctl_sctp_wmem[3];

108
static int sctp_memory_pressure;
E
Eric Dumazet 已提交
109
static atomic_long_t sctp_memory_allocated;
110
struct percpu_counter sctp_sockets_allocated;
111

112
static void sctp_enter_memory_pressure(struct sock *sk)
113 114 115 116 117
{
	sctp_memory_pressure = 1;
}


L
Linus Torvalds 已提交
118 119 120
/* Get the sndbuf space available at the time on the association.  */
static inline int sctp_wspace(struct sctp_association *asoc)
{
121
	int amt;
L
Linus Torvalds 已提交
122

123 124 125
	if (asoc->ep->sndbuf_policy)
		amt = asoc->sndbuf_used;
	else
126
		amt = sk_wmem_alloc_get(asoc->base.sk);
127 128 129 130 131 132 133 134 135

	if (amt >= asoc->base.sk->sk_sndbuf) {
		if (asoc->base.sk->sk_userlocks & SOCK_SNDBUF_LOCK)
			amt = 0;
		else {
			amt = sk_stream_wspace(asoc->base.sk);
			if (amt < 0)
				amt = 0;
		}
136
	} else {
137
		amt = asoc->base.sk->sk_sndbuf - amt;
138
	}
L
Linus Torvalds 已提交
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158
	return amt;
}

/* Increment the used sndbuf space count of the corresponding association by
 * the size of the outgoing data chunk.
 * Also, set the skb destructor for sndbuf accounting later.
 *
 * Since it is always 1-1 between chunk and skb, and also a new skb is always
 * allocated for chunk bundling in sctp_packet_transmit(), we can use the
 * destructor in the data chunk skb for the purpose of the sndbuf space
 * tracking.
 */
static inline void sctp_set_owner_w(struct sctp_chunk *chunk)
{
	struct sctp_association *asoc = chunk->asoc;
	struct sock *sk = asoc->base.sk;

	/* The sndbuf space is tracked per association.  */
	sctp_association_hold(asoc);

159 160
	skb_set_owner_w(chunk->skb, sk);

L
Linus Torvalds 已提交
161 162 163 164
	chunk->skb->destructor = sctp_wfree;
	/* Save the chunk pointer in skb for sctp_wfree to use later.  */
	*((struct sctp_chunk **)(chunk->skb->cb)) = chunk;

165 166 167 168 169
	asoc->sndbuf_used += SCTP_DATA_SNDSIZE(chunk) +
				sizeof(struct sk_buff) +
				sizeof(struct sctp_chunk);

	atomic_add(sizeof(struct sctp_chunk), &sk->sk_wmem_alloc);
170 171
	sk->sk_wmem_queued += chunk->skb->truesize;
	sk_mem_charge(sk, chunk->skb->truesize);
L
Linus Torvalds 已提交
172 173 174 175 176 177 178 179 180 181 182 183 184 185
}

/* Verify that this is a valid address. */
static inline int sctp_verify_addr(struct sock *sk, union sctp_addr *addr,
				   int len)
{
	struct sctp_af *af;

	/* Verify basic sockaddr. */
	af = sctp_sockaddr_af(sctp_sk(sk), addr, len);
	if (!af)
		return -EINVAL;

	/* Is this a valid SCTP address?  */
186
	if (!af->addr_valid(addr, sctp_sk(sk), NULL))
L
Linus Torvalds 已提交
187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244
		return -EINVAL;

	if (!sctp_sk(sk)->pf->send_verify(sctp_sk(sk), (addr)))
		return -EINVAL;

	return 0;
}

/* Look up the association by its id.  If this is not a UDP-style
 * socket, the ID field is always ignored.
 */
struct sctp_association *sctp_id2assoc(struct sock *sk, sctp_assoc_t id)
{
	struct sctp_association *asoc = NULL;

	/* If this is not a UDP-style socket, assoc id should be ignored. */
	if (!sctp_style(sk, UDP)) {
		/* Return NULL if the socket state is not ESTABLISHED. It
		 * could be a TCP-style listening socket or a socket which
		 * hasn't yet called connect() to establish an association.
		 */
		if (!sctp_sstate(sk, ESTABLISHED))
			return NULL;

		/* Get the first and the only association from the list. */
		if (!list_empty(&sctp_sk(sk)->ep->asocs))
			asoc = list_entry(sctp_sk(sk)->ep->asocs.next,
					  struct sctp_association, asocs);
		return asoc;
	}

	/* Otherwise this is a UDP-style socket. */
	if (!id || (id == (sctp_assoc_t)-1))
		return NULL;

	spin_lock_bh(&sctp_assocs_id_lock);
	asoc = (struct sctp_association *)idr_find(&sctp_assocs_id, (int)id);
	spin_unlock_bh(&sctp_assocs_id_lock);

	if (!asoc || (asoc->base.sk != sk) || asoc->base.dead)
		return NULL;

	return asoc;
}

/* Look up the transport from an address and an assoc id. If both address and
 * id are specified, the associations matching the address and the id should be
 * the same.
 */
static struct sctp_transport *sctp_addr_id2transport(struct sock *sk,
					      struct sockaddr_storage *addr,
					      sctp_assoc_t id)
{
	struct sctp_association *addr_asoc = NULL, *id_asoc = NULL;
	struct sctp_transport *transport;
	union sctp_addr *laddr = (union sctp_addr *)addr;

	addr_asoc = sctp_endpoint_lookup_assoc(sctp_sk(sk)->ep,
245
					       laddr,
L
Linus Torvalds 已提交
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
					       &transport);

	if (!addr_asoc)
		return NULL;

	id_asoc = sctp_id2assoc(sk, id);
	if (id_asoc && (id_asoc != addr_asoc))
		return NULL;

	sctp_get_pf_specific(sk->sk_family)->addr_v4map(sctp_sk(sk),
						(union sctp_addr *)addr);

	return transport;
}

/* API 3.1.2 bind() - UDP Style Syntax
 * The syntax of bind() is,
 *
 *   ret = bind(int sd, struct sockaddr *addr, int addrlen);
 *
 *   sd      - the socket descriptor returned by socket().
 *   addr    - the address structure (struct sockaddr_in or struct
 *             sockaddr_in6 [RFC 2553]),
 *   addr_len - the size of the address structure.
 */
271
static int sctp_bind(struct sock *sk, struct sockaddr *addr, int addr_len)
L
Linus Torvalds 已提交
272 273 274 275 276
{
	int retval = 0;

	sctp_lock_sock(sk);

277 278
	pr_debug("%s: sk:%p, addr:%p, addr_len:%d\n", __func__, sk,
		 addr, addr_len);
L
Linus Torvalds 已提交
279 280 281

	/* Disallow binding twice. */
	if (!sctp_sk(sk)->ep->base.bind_addr.port)
F
Frank Filz 已提交
282
		retval = sctp_do_bind(sk, (union sctp_addr *)addr,
L
Linus Torvalds 已提交
283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
				      addr_len);
	else
		retval = -EINVAL;

	sctp_release_sock(sk);

	return retval;
}

static long sctp_get_port_local(struct sock *, union sctp_addr *);

/* Verify this is a valid sockaddr. */
static struct sctp_af *sctp_sockaddr_af(struct sctp_sock *opt,
					union sctp_addr *addr, int len)
{
	struct sctp_af *af;

	/* Check minimum size.  */
	if (len < sizeof (struct sockaddr))
		return NULL;

304 305 306 307 308 309 310 311 312 313
	/* V4 mapped address are really of AF_INET family */
	if (addr->sa.sa_family == AF_INET6 &&
	    ipv6_addr_v4mapped(&addr->v6.sin6_addr)) {
		if (!opt->pf->af_supported(AF_INET, opt))
			return NULL;
	} else {
		/* Does this PF support this AF? */
		if (!opt->pf->af_supported(addr->sa.sa_family, opt))
			return NULL;
	}
L
Linus Torvalds 已提交
314 315 316 317 318 319 320 321 322 323 324

	/* If we get this far, af is valid. */
	af = sctp_get_af_specific(addr->sa.sa_family);

	if (len < af->sockaddr_len)
		return NULL;

	return af;
}

/* Bind a local address either to an endpoint or to an association.  */
325
static int sctp_do_bind(struct sock *sk, union sctp_addr *addr, int len)
L
Linus Torvalds 已提交
326
{
327
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
328 329 330 331 332 333 334 335 336
	struct sctp_sock *sp = sctp_sk(sk);
	struct sctp_endpoint *ep = sp->ep;
	struct sctp_bind_addr *bp = &ep->base.bind_addr;
	struct sctp_af *af;
	unsigned short snum;
	int ret = 0;

	/* Common sockaddr verification. */
	af = sctp_sockaddr_af(sp, addr, len);
F
Frank Filz 已提交
337
	if (!af) {
338 339
		pr_debug("%s: sk:%p, newaddr:%p, len:%d EINVAL\n",
			 __func__, sk, addr, len);
L
Linus Torvalds 已提交
340
		return -EINVAL;
F
Frank Filz 已提交
341 342 343 344
	}

	snum = ntohs(addr->v4.sin_port);

345 346
	pr_debug("%s: sk:%p, new addr:%pISc, port:%d, new port:%d, len:%d\n",
		 __func__, sk, &addr->sa, bp->port, snum, len);
L
Linus Torvalds 已提交
347 348 349 350 351

	/* PF specific bind() address verification. */
	if (!sp->pf->bind_verify(sp, addr))
		return -EADDRNOTAVAIL;

352 353 354 355 356 357 358 359
	/* We must either be unbound, or bind to the same port.
	 * It's OK to allow 0 ports if we are already bound.
	 * We'll just inhert an already bound port in this case
	 */
	if (bp->port) {
		if (!snum)
			snum = bp->port;
		else if (snum != bp->port) {
360 361
			pr_debug("%s: new port %d doesn't match existing port "
				 "%d\n", __func__, snum, bp->port);
362 363
			return -EINVAL;
		}
L
Linus Torvalds 已提交
364 365
	}

366 367
	if (snum && snum < PROT_SOCK &&
	    !ns_capable(net->user_ns, CAP_NET_BIND_SERVICE))
L
Linus Torvalds 已提交
368 369
		return -EACCES;

370 371 372 373 374 375
	/* See if the address matches any of the addresses we may have
	 * already bound before checking against other endpoints.
	 */
	if (sctp_bind_addr_match(bp, addr, sp))
		return -EINVAL;

L
Linus Torvalds 已提交
376 377 378 379
	/* Make sure we are allowed to bind here.
	 * The function sctp_get_port_local() does duplicate address
	 * detection.
	 */
380
	addr->v4.sin_port = htons(snum);
L
Linus Torvalds 已提交
381
	if ((ret = sctp_get_port_local(sk, addr))) {
382
		return -EADDRINUSE;
L
Linus Torvalds 已提交
383 384 385 386
	}

	/* Refresh ephemeral port.  */
	if (!bp->port)
E
Eric Dumazet 已提交
387
		bp->port = inet_sk(sk)->inet_num;
L
Linus Torvalds 已提交
388

389 390 391
	/* Add the address to the bind address list.
	 * Use GFP_ATOMIC since BHs will be disabled.
	 */
392
	ret = sctp_add_bind_addr(bp, addr, SCTP_ADDR_SRC, GFP_ATOMIC);
L
Linus Torvalds 已提交
393 394 395

	/* Copy back into socket for getsockname() use. */
	if (!ret) {
E
Eric Dumazet 已提交
396
		inet_sk(sk)->inet_sport = htons(inet_sk(sk)->inet_num);
L
Linus Torvalds 已提交
397 398 399 400 401 402 403 404
		af->to_sk_saddr(addr, sk);
	}

	return ret;
}

 /* ADDIP Section 4.1.1 Congestion Control of ASCONF Chunks
 *
405
 * R1) One and only one ASCONF Chunk MAY be in transit and unacknowledged
L
Linus Torvalds 已提交
406
 * at any one time.  If a sender, after sending an ASCONF chunk, decides
407
 * it needs to transfer another ASCONF Chunk, it MUST wait until the
L
Linus Torvalds 已提交
408
 * ASCONF-ACK Chunk returns from the previous ASCONF Chunk before sending a
409 410
 * subsequent ASCONF. Note this restriction binds each side, so at any
 * time two ASCONF may be in-transit on any given association (one sent
L
Linus Torvalds 已提交
411 412 413 414 415
 * from each endpoint).
 */
static int sctp_send_asconf(struct sctp_association *asoc,
			    struct sctp_chunk *chunk)
{
416
	struct net 	*net = sock_net(asoc->base.sk);
L
Linus Torvalds 已提交
417 418 419 420
	int		retval = 0;

	/* If there is an outstanding ASCONF chunk, queue it for later
	 * transmission.
421
	 */
L
Linus Torvalds 已提交
422
	if (asoc->addip_last_asconf) {
423
		list_add_tail(&chunk->list, &asoc->addip_chunk_list);
424
		goto out;
L
Linus Torvalds 已提交
425 426 427 428
	}

	/* Hold the chunk until an ASCONF_ACK is received. */
	sctp_chunk_hold(chunk);
429
	retval = sctp_primitive_ASCONF(net, asoc, chunk);
L
Linus Torvalds 已提交
430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450
	if (retval)
		sctp_chunk_free(chunk);
	else
		asoc->addip_last_asconf = chunk;

out:
	return retval;
}

/* Add a list of addresses as bind addresses to local endpoint or
 * association.
 *
 * Basically run through each address specified in the addrs/addrcnt
 * array/length pair, determine if it is IPv6 or IPv4 and call
 * sctp_do_bind() on it.
 *
 * If any of them fails, then the operation will be reversed and the
 * ones that were added will be removed.
 *
 * Only sctp_setsockopt_bindx() is supposed to call this function.
 */
451
static int sctp_bindx_add(struct sock *sk, struct sockaddr *addrs, int addrcnt)
L
Linus Torvalds 已提交
452 453 454 455 456 457 458
{
	int cnt;
	int retval = 0;
	void *addr_buf;
	struct sockaddr *sa_addr;
	struct sctp_af *af;

459 460
	pr_debug("%s: sk:%p, addrs:%p, addrcnt:%d\n", __func__, sk,
		 addrs, addrcnt);
L
Linus Torvalds 已提交
461 462 463 464 465 466

	addr_buf = addrs;
	for (cnt = 0; cnt < addrcnt; cnt++) {
		/* The list may contain either IPv4 or IPv6 address;
		 * determine the address length for walking thru the list.
		 */
J
Joe Perches 已提交
467
		sa_addr = addr_buf;
L
Linus Torvalds 已提交
468 469 470 471 472 473
		af = sctp_get_af_specific(sa_addr->sa_family);
		if (!af) {
			retval = -EINVAL;
			goto err_bindx_add;
		}

474
		retval = sctp_do_bind(sk, (union sctp_addr *)sa_addr,
L
Linus Torvalds 已提交
475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494
				      af->sockaddr_len);

		addr_buf += af->sockaddr_len;

err_bindx_add:
		if (retval < 0) {
			/* Failed. Cleanup the ones that have been added */
			if (cnt > 0)
				sctp_bindx_rem(sk, addrs, cnt);
			return retval;
		}
	}

	return retval;
}

/* Send an ASCONF chunk with Add IP address parameters to all the peers of the
 * associations that are part of the endpoint indicating that a list of local
 * addresses are added to the endpoint.
 *
495
 * If any of the addresses is already in the bind address list of the
L
Linus Torvalds 已提交
496 497 498 499 500
 * association, we do not send the chunk for that association.  But it will not
 * affect other associations.
 *
 * Only sctp_setsockopt_bindx() is supposed to call this function.
 */
501
static int sctp_send_asconf_add_ip(struct sock		*sk,
L
Linus Torvalds 已提交
502 503 504
				   struct sockaddr	*addrs,
				   int 			addrcnt)
{
505
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
506 507 508 509 510 511 512
	struct sctp_sock		*sp;
	struct sctp_endpoint		*ep;
	struct sctp_association		*asoc;
	struct sctp_bind_addr		*bp;
	struct sctp_chunk		*chunk;
	struct sctp_sockaddr_entry	*laddr;
	union sctp_addr			*addr;
513
	union sctp_addr			saveaddr;
L
Linus Torvalds 已提交
514 515 516 517 518 519
	void				*addr_buf;
	struct sctp_af			*af;
	struct list_head		*p;
	int 				i;
	int 				retval = 0;

520
	if (!net->sctp.addip_enable)
L
Linus Torvalds 已提交
521 522 523 524 525
		return retval;

	sp = sctp_sk(sk);
	ep = sp->ep;

526 527
	pr_debug("%s: sk:%p, addrs:%p, addrcnt:%d\n",
		 __func__, sk, addrs, addrcnt);
L
Linus Torvalds 已提交
528

529
	list_for_each_entry(asoc, &ep->asocs, asocs) {
L
Linus Torvalds 已提交
530 531 532 533 534 535 536 537 538 539
		if (!asoc->peer.asconf_capable)
			continue;

		if (asoc->peer.addip_disabled_mask & SCTP_PARAM_ADD_IP)
			continue;

		if (!sctp_state(asoc, ESTABLISHED))
			continue;

		/* Check if any address in the packed array of addresses is
540 541
		 * in the bind address list of the association. If so,
		 * do not send the asconf chunk to its peer, but continue with
L
Linus Torvalds 已提交
542 543 544 545
		 * other associations.
		 */
		addr_buf = addrs;
		for (i = 0; i < addrcnt; i++) {
J
Joe Perches 已提交
546
			addr = addr_buf;
L
Linus Torvalds 已提交
547 548 549 550 551 552 553 554 555 556 557 558 559 560
			af = sctp_get_af_specific(addr->v4.sin_family);
			if (!af) {
				retval = -EINVAL;
				goto out;
			}

			if (sctp_assoc_lookup_laddr(asoc, addr))
				break;

			addr_buf += af->sockaddr_len;
		}
		if (i < addrcnt)
			continue;

561 562
		/* Use the first valid address in bind addr list of
		 * association as Address Parameter of ASCONF CHUNK.
L
Linus Torvalds 已提交
563 564 565 566
		 */
		bp = &asoc->base.bind_addr;
		p = bp->address_list.next;
		laddr = list_entry(p, struct sctp_sockaddr_entry, list);
567
		chunk = sctp_make_asconf_update_ip(asoc, &laddr->a, addrs,
L
Linus Torvalds 已提交
568 569 570 571 572 573
						   addrcnt, SCTP_PARAM_ADD_IP);
		if (!chunk) {
			retval = -ENOMEM;
			goto out;
		}

574 575
		/* Add the new addresses to the bind address list with
		 * use_as_src set to 0.
L
Linus Torvalds 已提交
576
		 */
577 578
		addr_buf = addrs;
		for (i = 0; i < addrcnt; i++) {
J
Joe Perches 已提交
579
			addr = addr_buf;
580 581
			af = sctp_get_af_specific(addr->v4.sin_family);
			memcpy(&saveaddr, addr, af->sockaddr_len);
582 583
			retval = sctp_add_bind_addr(bp, &saveaddr,
						    SCTP_ADDR_NEW, GFP_ATOMIC);
584 585
			addr_buf += af->sockaddr_len;
		}
586 587 588 589 590 591 592 593 594 595 596
		if (asoc->src_out_of_asoc_ok) {
			struct sctp_transport *trans;

			list_for_each_entry(trans,
			    &asoc->peer.transport_addr_list, transports) {
				/* Clear the source and route cache */
				dst_release(trans->dst);
				trans->cwnd = min(4*asoc->pathmtu, max_t(__u32,
				    2*asoc->pathmtu, 4380));
				trans->ssthresh = asoc->peer.i.a_rwnd;
				trans->rto = asoc->rto_initial;
597
				sctp_max_rto(asoc, trans);
598 599 600 601 602 603
				trans->rtt = trans->srtt = trans->rttvar = 0;
				sctp_transport_route(trans, NULL,
				    sctp_sk(asoc->base.sk));
			}
		}
		retval = sctp_send_asconf(asoc, chunk);
L
Linus Torvalds 已提交
604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624
	}

out:
	return retval;
}

/* Remove a list of addresses from bind addresses list.  Do not remove the
 * last address.
 *
 * Basically run through each address specified in the addrs/addrcnt
 * array/length pair, determine if it is IPv6 or IPv4 and call
 * sctp_del_bind() on it.
 *
 * If any of them fails, then the operation will be reversed and the
 * ones that were removed will be added back.
 *
 * At least one address has to be left; if only one address is
 * available, the operation will return -EBUSY.
 *
 * Only sctp_setsockopt_bindx() is supposed to call this function.
 */
625
static int sctp_bindx_rem(struct sock *sk, struct sockaddr *addrs, int addrcnt)
L
Linus Torvalds 已提交
626 627 628 629 630 631 632
{
	struct sctp_sock *sp = sctp_sk(sk);
	struct sctp_endpoint *ep = sp->ep;
	int cnt;
	struct sctp_bind_addr *bp = &ep->base.bind_addr;
	int retval = 0;
	void *addr_buf;
633
	union sctp_addr *sa_addr;
L
Linus Torvalds 已提交
634 635
	struct sctp_af *af;

636 637
	pr_debug("%s: sk:%p, addrs:%p, addrcnt:%d\n",
		 __func__, sk, addrs, addrcnt);
L
Linus Torvalds 已提交
638 639 640 641 642 643 644 645 646 647 648 649 650

	addr_buf = addrs;
	for (cnt = 0; cnt < addrcnt; cnt++) {
		/* If the bind address list is empty or if there is only one
		 * bind address, there is nothing more to be removed (we need
		 * at least one address here).
		 */
		if (list_empty(&bp->address_list) ||
		    (sctp_list_single_entry(&bp->address_list))) {
			retval = -EBUSY;
			goto err_bindx_rem;
		}

J
Joe Perches 已提交
651
		sa_addr = addr_buf;
652
		af = sctp_get_af_specific(sa_addr->sa.sa_family);
L
Linus Torvalds 已提交
653 654 655 656
		if (!af) {
			retval = -EINVAL;
			goto err_bindx_rem;
		}
657 658 659 660 661 662

		if (!af->addr_valid(sa_addr, sp, NULL)) {
			retval = -EADDRNOTAVAIL;
			goto err_bindx_rem;
		}

663 664
		if (sa_addr->v4.sin_port &&
		    sa_addr->v4.sin_port != htons(bp->port)) {
L
Linus Torvalds 已提交
665 666 667 668
			retval = -EINVAL;
			goto err_bindx_rem;
		}

669 670 671
		if (!sa_addr->v4.sin_port)
			sa_addr->v4.sin_port = htons(bp->port);

L
Linus Torvalds 已提交
672 673 674 675 676 677 678
		/* FIXME - There is probably a need to check if sk->sk_saddr and
		 * sk->sk_rcv_addr are currently set to one of the addresses to
		 * be removed. This is something which needs to be looked into
		 * when we are fixing the outstanding issues with multi-homing
		 * socket routing and failover schemes. Refer to comments in
		 * sctp_do_bind(). -daisy
		 */
679
		retval = sctp_del_bind_addr(bp, sa_addr);
L
Linus Torvalds 已提交
680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697

		addr_buf += af->sockaddr_len;
err_bindx_rem:
		if (retval < 0) {
			/* Failed. Add the ones that has been removed back */
			if (cnt > 0)
				sctp_bindx_add(sk, addrs, cnt);
			return retval;
		}
	}

	return retval;
}

/* Send an ASCONF chunk with Delete IP address parameters to all the peers of
 * the associations that are part of the endpoint indicating that a list of
 * local addresses are removed from the endpoint.
 *
698
 * If any of the addresses is already in the bind address list of the
L
Linus Torvalds 已提交
699 700 701 702 703 704 705 706 707
 * association, we do not send the chunk for that association.  But it will not
 * affect other associations.
 *
 * Only sctp_setsockopt_bindx() is supposed to call this function.
 */
static int sctp_send_asconf_del_ip(struct sock		*sk,
				   struct sockaddr	*addrs,
				   int			addrcnt)
{
708
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
709 710 711
	struct sctp_sock	*sp;
	struct sctp_endpoint	*ep;
	struct sctp_association	*asoc;
712
	struct sctp_transport	*transport;
L
Linus Torvalds 已提交
713 714 715 716 717
	struct sctp_bind_addr	*bp;
	struct sctp_chunk	*chunk;
	union sctp_addr		*laddr;
	void			*addr_buf;
	struct sctp_af		*af;
718
	struct sctp_sockaddr_entry *saddr;
L
Linus Torvalds 已提交
719 720
	int 			i;
	int 			retval = 0;
721
	int			stored = 0;
L
Linus Torvalds 已提交
722

723
	chunk = NULL;
724
	if (!net->sctp.addip_enable)
L
Linus Torvalds 已提交
725 726 727 728 729
		return retval;

	sp = sctp_sk(sk);
	ep = sp->ep;

730 731
	pr_debug("%s: sk:%p, addrs:%p, addrcnt:%d\n",
		 __func__, sk, addrs, addrcnt);
L
Linus Torvalds 已提交
732

733
	list_for_each_entry(asoc, &ep->asocs, asocs) {
L
Linus Torvalds 已提交
734 735 736 737 738 739 740 741 742 743 744

		if (!asoc->peer.asconf_capable)
			continue;

		if (asoc->peer.addip_disabled_mask & SCTP_PARAM_DEL_IP)
			continue;

		if (!sctp_state(asoc, ESTABLISHED))
			continue;

		/* Check if any address in the packed array of addresses is
745
		 * not present in the bind address list of the association.
L
Linus Torvalds 已提交
746 747 748 749 750
		 * If so, do not send the asconf chunk to its peer, but
		 * continue with other associations.
		 */
		addr_buf = addrs;
		for (i = 0; i < addrcnt; i++) {
J
Joe Perches 已提交
751
			laddr = addr_buf;
L
Linus Torvalds 已提交
752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773
			af = sctp_get_af_specific(laddr->v4.sin_family);
			if (!af) {
				retval = -EINVAL;
				goto out;
			}

			if (!sctp_assoc_lookup_laddr(asoc, laddr))
				break;

			addr_buf += af->sockaddr_len;
		}
		if (i < addrcnt)
			continue;

		/* Find one address in the association's bind address list
		 * that is not in the packed array of addresses. This is to
		 * make sure that we do not delete all the addresses in the
		 * association.
		 */
		bp = &asoc->base.bind_addr;
		laddr = sctp_find_unmatch_addr(bp, (union sctp_addr *)addrs,
					       addrcnt, sp);
774 775 776 777 778
		if ((laddr == NULL) && (addrcnt == 1)) {
			if (asoc->asconf_addr_del_pending)
				continue;
			asoc->asconf_addr_del_pending =
			    kzalloc(sizeof(union sctp_addr), GFP_ATOMIC);
779 780 781 782
			if (asoc->asconf_addr_del_pending == NULL) {
				retval = -ENOMEM;
				goto out;
			}
783 784 785 786 787 788 789 790 791 792 793 794 795
			asoc->asconf_addr_del_pending->sa.sa_family =
				    addrs->sa_family;
			asoc->asconf_addr_del_pending->v4.sin_port =
				    htons(bp->port);
			if (addrs->sa_family == AF_INET) {
				struct sockaddr_in *sin;

				sin = (struct sockaddr_in *)addrs;
				asoc->asconf_addr_del_pending->v4.sin_addr.s_addr = sin->sin_addr.s_addr;
			} else if (addrs->sa_family == AF_INET6) {
				struct sockaddr_in6 *sin6;

				sin6 = (struct sockaddr_in6 *)addrs;
A
Alexey Dobriyan 已提交
796
				asoc->asconf_addr_del_pending->v6.sin6_addr = sin6->sin6_addr;
797
			}
798 799 800 801 802

			pr_debug("%s: keep the last address asoc:%p %pISc at %p\n",
				 __func__, asoc, &asoc->asconf_addr_del_pending->sa,
				 asoc->asconf_addr_del_pending);

803 804 805 806
			asoc->src_out_of_asoc_ok = 1;
			stored = 1;
			goto skip_mkasconf;
		}
L
Linus Torvalds 已提交
807

808 809 810
		if (laddr == NULL)
			return -EINVAL;

811 812 813 814
		/* We do not need RCU protection throughout this loop
		 * because this is done under a socket lock from the
		 * setsockopt call.
		 */
L
Linus Torvalds 已提交
815 816 817 818 819 820 821
		chunk = sctp_make_asconf_update_ip(asoc, laddr, addrs, addrcnt,
						   SCTP_PARAM_DEL_IP);
		if (!chunk) {
			retval = -ENOMEM;
			goto out;
		}

822
skip_mkasconf:
823 824 825 826 827
		/* Reset use_as_src flag for the addresses in the bind address
		 * list that are to be deleted.
		 */
		addr_buf = addrs;
		for (i = 0; i < addrcnt; i++) {
J
Joe Perches 已提交
828
			laddr = addr_buf;
829
			af = sctp_get_af_specific(laddr->v4.sin_family);
830
			list_for_each_entry(saddr, &bp->address_list, list) {
831
				if (sctp_cmp_addr_exact(&saddr->a, laddr))
832
					saddr->state = SCTP_ADDR_DEL;
833 834 835
			}
			addr_buf += af->sockaddr_len;
		}
L
Linus Torvalds 已提交
836

837 838 839
		/* Update the route and saddr entries for all the transports
		 * as some of the addresses in the bind address list are
		 * about to be deleted and cannot be used as source addresses.
L
Linus Torvalds 已提交
840
		 */
841 842
		list_for_each_entry(transport, &asoc->peer.transport_addr_list,
					transports) {
843 844 845 846 847
			dst_release(transport->dst);
			sctp_transport_route(transport, NULL,
					     sctp_sk(asoc->base.sk));
		}

848 849 850
		if (stored)
			/* We don't need to transmit ASCONF */
			continue;
851
		retval = sctp_send_asconf(asoc, chunk);
L
Linus Torvalds 已提交
852 853 854 855 856
	}
out:
	return retval;
}

857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878
/* set addr events to assocs in the endpoint.  ep and addr_wq must be locked */
int sctp_asconf_mgmt(struct sctp_sock *sp, struct sctp_sockaddr_entry *addrw)
{
	struct sock *sk = sctp_opt2sk(sp);
	union sctp_addr *addr;
	struct sctp_af *af;

	/* It is safe to write port space in caller. */
	addr = &addrw->a;
	addr->v4.sin_port = htons(sp->ep->base.bind_addr.port);
	af = sctp_get_af_specific(addr->sa.sa_family);
	if (!af)
		return -EINVAL;
	if (sctp_verify_addr(sk, addr, af->sockaddr_len))
		return -EINVAL;

	if (addrw->state == SCTP_ADDR_NEW)
		return sctp_send_asconf_add_ip(sk, (struct sockaddr *)addr, 1);
	else
		return sctp_send_asconf_del_ip(sk, (struct sockaddr *)addr, 1);
}

L
Linus Torvalds 已提交
879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894
/* Helper for tunneling sctp_bindx() requests through sctp_setsockopt()
 *
 * API 8.1
 * int sctp_bindx(int sd, struct sockaddr *addrs, int addrcnt,
 *                int flags);
 *
 * If sd is an IPv4 socket, the addresses passed must be IPv4 addresses.
 * If the sd is an IPv6 socket, the addresses passed can either be IPv4
 * or IPv6 addresses.
 *
 * A single address may be specified as INADDR_ANY or IN6ADDR_ANY, see
 * Section 3.1.2 for this usage.
 *
 * addrs is a pointer to an array of one or more socket addresses. Each
 * address is contained in its appropriate structure (i.e. struct
 * sockaddr_in or struct sockaddr_in6) the family of the address type
V
Ville Nuorvala 已提交
895
 * must be used to distinguish the address length (note that this
L
Linus Torvalds 已提交
896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933
 * representation is termed a "packed array" of addresses). The caller
 * specifies the number of addresses in the array with addrcnt.
 *
 * On success, sctp_bindx() returns 0. On failure, sctp_bindx() returns
 * -1, and sets errno to the appropriate error code.
 *
 * For SCTP, the port given in each socket address must be the same, or
 * sctp_bindx() will fail, setting errno to EINVAL.
 *
 * The flags parameter is formed from the bitwise OR of zero or more of
 * the following currently defined flags:
 *
 * SCTP_BINDX_ADD_ADDR
 *
 * SCTP_BINDX_REM_ADDR
 *
 * SCTP_BINDX_ADD_ADDR directs SCTP to add the given addresses to the
 * association, and SCTP_BINDX_REM_ADDR directs SCTP to remove the given
 * addresses from the association. The two flags are mutually exclusive;
 * if both are given, sctp_bindx() will fail with EINVAL. A caller may
 * not remove all addresses from an association; sctp_bindx() will
 * reject such an attempt with EINVAL.
 *
 * An application can use sctp_bindx(SCTP_BINDX_ADD_ADDR) to associate
 * additional addresses with an endpoint after calling bind().  Or use
 * sctp_bindx(SCTP_BINDX_REM_ADDR) to remove some addresses a listening
 * socket is associated with so that no new association accepted will be
 * associated with those addresses. If the endpoint supports dynamic
 * address a SCTP_BINDX_REM_ADDR or SCTP_BINDX_ADD_ADDR may cause a
 * endpoint to send the appropriate message to the peer to change the
 * peers address lists.
 *
 * Adding and removing addresses from a connected association is
 * optional functionality. Implementations that do not support this
 * functionality should return EOPNOTSUPP.
 *
 * Basically do nothing but copying the addresses from user to kernel
 * land and invoking either sctp_bindx_add() or sctp_bindx_rem() on the sk.
F
Frank Filz 已提交
934 935
 * This is used for tunneling the sctp_bindx() request through sctp_setsockopt()
 * from userspace.
L
Linus Torvalds 已提交
936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954
 *
 * We don't use copy_from_user() for optimization: we first do the
 * sanity checks (buffer size -fast- and access check-healthy
 * pointer); if all of those succeed, then we can alloc the memory
 * (expensive operation) needed to copy the data to kernel. Then we do
 * the copying without checking the user space area
 * (__copy_from_user()).
 *
 * On exit there is no need to do sockfd_put(), sys_setsockopt() does
 * it.
 *
 * sk        The sk of the socket
 * addrs     The pointer to the addresses in user land
 * addrssize Size of the addrs buffer
 * op        Operation to perform (add or remove, see the flags of
 *           sctp_bindx)
 *
 * Returns 0 if ok, <0 errno code on error.
 */
955 956 957
static int sctp_setsockopt_bindx(struct sock* sk,
				 struct sockaddr __user *addrs,
				 int addrs_size, int op)
L
Linus Torvalds 已提交
958 959 960 961 962 963 964 965 966
{
	struct sockaddr *kaddrs;
	int err;
	int addrcnt = 0;
	int walk_size = 0;
	struct sockaddr *sa_addr;
	void *addr_buf;
	struct sctp_af *af;

967 968
	pr_debug("%s: sk:%p addrs:%p addrs_size:%d opt:%d\n",
		 __func__, sk, addrs, addrs_size, op);
L
Linus Torvalds 已提交
969 970 971 972 973 974 975 976 977

	if (unlikely(addrs_size <= 0))
		return -EINVAL;

	/* Check the user passed a healthy pointer.  */
	if (unlikely(!access_ok(VERIFY_READ, addrs, addrs_size)))
		return -EFAULT;

	/* Alloc space for the address array in kernel memory.  */
978
	kaddrs = kmalloc(addrs_size, GFP_KERNEL);
L
Linus Torvalds 已提交
979 980 981 982 983 984 985 986
	if (unlikely(!kaddrs))
		return -ENOMEM;

	if (__copy_from_user(kaddrs, addrs, addrs_size)) {
		kfree(kaddrs);
		return -EFAULT;
	}

987
	/* Walk through the addrs buffer and count the number of addresses. */
L
Linus Torvalds 已提交
988 989
	addr_buf = kaddrs;
	while (walk_size < addrs_size) {
990 991 992 993 994
		if (walk_size + sizeof(sa_family_t) > addrs_size) {
			kfree(kaddrs);
			return -EINVAL;
		}

J
Joe Perches 已提交
995
		sa_addr = addr_buf;
L
Linus Torvalds 已提交
996 997 998 999
		af = sctp_get_af_specific(sa_addr->sa_family);

		/* If the address family is not supported or if this address
		 * causes the address buffer to overflow return EINVAL.
1000
		 */
L
Linus Torvalds 已提交
1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
		if (!af || (walk_size + af->sockaddr_len) > addrs_size) {
			kfree(kaddrs);
			return -EINVAL;
		}
		addrcnt++;
		addr_buf += af->sockaddr_len;
		walk_size += af->sockaddr_len;
	}

	/* Do the work. */
	switch (op) {
	case SCTP_BINDX_ADD_ADDR:
		err = sctp_bindx_add(sk, kaddrs, addrcnt);
		if (err)
			goto out;
		err = sctp_send_asconf_add_ip(sk, kaddrs, addrcnt);
		break;

	case SCTP_BINDX_REM_ADDR:
		err = sctp_bindx_rem(sk, kaddrs, addrcnt);
		if (err)
			goto out;
		err = sctp_send_asconf_del_ip(sk, kaddrs, addrcnt);
		break;

	default:
		err = -EINVAL;
		break;
1029
	}
L
Linus Torvalds 已提交
1030 1031 1032 1033 1034 1035 1036

out:
	kfree(kaddrs);

	return err;
}

F
Frank Filz 已提交
1037 1038 1039 1040 1041 1042 1043
/* __sctp_connect(struct sock* sk, struct sockaddr *kaddrs, int addrs_size)
 *
 * Common routine for handling connect() and sctp_connectx().
 * Connect will come in with just a single address.
 */
static int __sctp_connect(struct sock* sk,
			  struct sockaddr *kaddrs,
1044 1045
			  int addrs_size,
			  sctp_assoc_t *assoc_id)
F
Frank Filz 已提交
1046
{
1047
	struct net *net = sock_net(sk);
F
Frank Filz 已提交
1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059
	struct sctp_sock *sp;
	struct sctp_endpoint *ep;
	struct sctp_association *asoc = NULL;
	struct sctp_association *asoc2;
	struct sctp_transport *transport;
	union sctp_addr to;
	struct sctp_af *af;
	sctp_scope_t scope;
	long timeo;
	int err = 0;
	int addrcnt = 0;
	int walk_size = 0;
1060
	union sctp_addr *sa_addr = NULL;
F
Frank Filz 已提交
1061
	void *addr_buf;
1062
	unsigned short port;
1063
	unsigned int f_flags = 0;
F
Frank Filz 已提交
1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081

	sp = sctp_sk(sk);
	ep = sp->ep;

	/* connect() cannot be done on a socket that is already in ESTABLISHED
	 * state - UDP-style peeled off socket or a TCP-style socket that
	 * is already connected.
	 * It cannot be done even on a TCP-style listening socket.
	 */
	if (sctp_sstate(sk, ESTABLISHED) ||
	    (sctp_style(sk, TCP) && sctp_sstate(sk, LISTENING))) {
		err = -EISCONN;
		goto out_free;
	}

	/* Walk through the addrs buffer and count the number of addresses. */
	addr_buf = kaddrs;
	while (walk_size < addrs_size) {
1082 1083 1084 1085 1086
		if (walk_size + sizeof(sa_family_t) > addrs_size) {
			err = -EINVAL;
			goto out_free;
		}

J
Joe Perches 已提交
1087
		sa_addr = addr_buf;
1088
		af = sctp_get_af_specific(sa_addr->sa.sa_family);
F
Frank Filz 已提交
1089 1090 1091 1092 1093 1094 1095 1096 1097

		/* If the address family is not supported or if this address
		 * causes the address buffer to overflow return EINVAL.
		 */
		if (!af || (walk_size + af->sockaddr_len) > addrs_size) {
			err = -EINVAL;
			goto out_free;
		}

1098 1099
		port = ntohs(sa_addr->v4.sin_port);

1100 1101 1102 1103
		/* Save current address so we can work with it */
		memcpy(&to, sa_addr, af->sockaddr_len);

		err = sctp_verify_addr(sk, &to, af->sockaddr_len);
F
Frank Filz 已提交
1104 1105 1106
		if (err)
			goto out_free;

1107 1108 1109
		/* Make sure the destination port is correctly set
		 * in all addresses.
		 */
1110 1111
		if (asoc && asoc->peer.port && asoc->peer.port != port) {
			err = -EINVAL;
1112
			goto out_free;
1113
		}
F
Frank Filz 已提交
1114 1115 1116 1117

		/* Check if there already is a matching association on the
		 * endpoint (other than the one created here).
		 */
1118
		asoc2 = sctp_endpoint_lookup_assoc(ep, &to, &transport);
F
Frank Filz 已提交
1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130
		if (asoc2 && asoc2 != asoc) {
			if (asoc2->state >= SCTP_STATE_ESTABLISHED)
				err = -EISCONN;
			else
				err = -EALREADY;
			goto out_free;
		}

		/* If we could not find a matching association on the endpoint,
		 * make sure that there is no peeled-off association matching
		 * the peer address even on another socket.
		 */
1131
		if (sctp_endpoint_is_peeled_off(ep, &to)) {
F
Frank Filz 已提交
1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146
			err = -EADDRNOTAVAIL;
			goto out_free;
		}

		if (!asoc) {
			/* If a bind() or sctp_bindx() is not called prior to
			 * an sctp_connectx() call, the system picks an
			 * ephemeral port and will choose an address set
			 * equivalent to binding with a wildcard address.
			 */
			if (!ep->base.bind_addr.port) {
				if (sctp_autobind(sk)) {
					err = -EAGAIN;
					goto out_free;
				}
1147 1148
			} else {
				/*
1149 1150 1151 1152
				 * If an unprivileged user inherits a 1-many
				 * style socket with open associations on a
				 * privileged port, it MAY be permitted to
				 * accept new associations, but it SHOULD NOT
1153 1154 1155
				 * be permitted to open new associations.
				 */
				if (ep->base.bind_addr.port < PROT_SOCK &&
1156
				    !ns_capable(net->user_ns, CAP_NET_BIND_SERVICE)) {
1157 1158 1159
					err = -EACCES;
					goto out_free;
				}
F
Frank Filz 已提交
1160 1161
			}

1162
			scope = sctp_scope(&to);
F
Frank Filz 已提交
1163 1164 1165 1166 1167
			asoc = sctp_association_new(ep, sk, scope, GFP_KERNEL);
			if (!asoc) {
				err = -ENOMEM;
				goto out_free;
			}
1168 1169 1170 1171 1172 1173 1174

			err = sctp_assoc_set_bind_addr_from_ep(asoc, scope,
							      GFP_KERNEL);
			if (err < 0) {
				goto out_free;
			}

F
Frank Filz 已提交
1175 1176 1177
		}

		/* Prime the peer's transport structures.  */
1178
		transport = sctp_assoc_add_peer(asoc, &to, GFP_KERNEL,
F
Frank Filz 已提交
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189
						SCTP_UNKNOWN);
		if (!transport) {
			err = -ENOMEM;
			goto out_free;
		}

		addrcnt++;
		addr_buf += af->sockaddr_len;
		walk_size += af->sockaddr_len;
	}

1190 1191 1192 1193 1194 1195 1196 1197 1198
	/* In case the user of sctp_connectx() wants an association
	 * id back, assign one now.
	 */
	if (assoc_id) {
		err = sctp_assoc_set_id(asoc, GFP_KERNEL);
		if (err < 0)
			goto out_free;
	}

1199
	err = sctp_primitive_ASSOCIATE(net, asoc, NULL);
F
Frank Filz 已提交
1200 1201 1202 1203 1204
	if (err < 0) {
		goto out_free;
	}

	/* Initialize sk's dport and daddr for getpeername() */
E
Eric Dumazet 已提交
1205
	inet_sk(sk)->inet_dport = htons(asoc->peer.port);
1206 1207
	af = sctp_get_af_specific(sa_addr->sa.sa_family);
	af->to_sk_daddr(sa_addr, sk);
1208
	sk->sk_err = 0;
F
Frank Filz 已提交
1209

1210 1211 1212 1213 1214 1215 1216 1217
	/* in-kernel sockets don't generally have a file allocated to them
	 * if all they do is call sock_create_kern().
	 */
	if (sk->sk_socket->file)
		f_flags = sk->sk_socket->file->f_flags;

	timeo = sock_sndtimeo(sk, f_flags & O_NONBLOCK);

F
Frank Filz 已提交
1218
	err = sctp_wait_for_connect(asoc, &timeo);
1219
	if ((err == 0 || err == -EINPROGRESS) && assoc_id)
1220
		*assoc_id = asoc->assoc_id;
F
Frank Filz 已提交
1221 1222 1223 1224 1225

	/* Don't free association on exit. */
	asoc = NULL;

out_free:
1226 1227
	pr_debug("%s: took out_free path with asoc:%p kaddrs:%p err:%d\n",
		 __func__, asoc, kaddrs, err);
F
Frank Filz 已提交
1228

1229 1230 1231 1232 1233 1234
	if (asoc) {
		/* sctp_primitive_ASSOCIATE may have added this association
		 * To the hash table, try to unhash it, just in case, its a noop
		 * if it wasn't hashed so we're safe
		 */
		sctp_unhash_established(asoc);
F
Frank Filz 已提交
1235
		sctp_association_free(asoc);
1236
	}
F
Frank Filz 已提交
1237 1238 1239 1240 1241 1242
	return err;
}

/* Helper for tunneling sctp_connectx() requests through sctp_setsockopt()
 *
 * API 8.9
1243 1244
 * int sctp_connectx(int sd, struct sockaddr *addrs, int addrcnt,
 * 			sctp_assoc_t *asoc);
F
Frank Filz 已提交
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259
 *
 * If sd is an IPv4 socket, the addresses passed must be IPv4 addresses.
 * If the sd is an IPv6 socket, the addresses passed can either be IPv4
 * or IPv6 addresses.
 *
 * A single address may be specified as INADDR_ANY or IN6ADDR_ANY, see
 * Section 3.1.2 for this usage.
 *
 * addrs is a pointer to an array of one or more socket addresses. Each
 * address is contained in its appropriate structure (i.e. struct
 * sockaddr_in or struct sockaddr_in6) the family of the address type
 * must be used to distengish the address length (note that this
 * representation is termed a "packed array" of addresses). The caller
 * specifies the number of addresses in the array with addrcnt.
 *
1260 1261 1262 1263
 * On success, sctp_connectx() returns 0. It also sets the assoc_id to
 * the association id of the new association.  On failure, sctp_connectx()
 * returns -1, and sets errno to the appropriate error code.  The assoc_id
 * is not touched by the kernel.
F
Frank Filz 已提交
1264 1265 1266 1267 1268 1269 1270 1271
 *
 * For SCTP, the port given in each socket address must be the same, or
 * sctp_connectx() will fail, setting errno to EINVAL.
 *
 * An application can use sctp_connectx to initiate an association with
 * an endpoint that is multi-homed.  Much like sctp_bindx() this call
 * allows a caller to specify multiple addresses at which a peer can be
 * reached.  The way the SCTP stack uses the list of addresses to set up
L
Lucas De Marchi 已提交
1272
 * the association is implementation dependent.  This function only
F
Frank Filz 已提交
1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299
 * specifies that the stack will try to make use of all the addresses in
 * the list when needed.
 *
 * Note that the list of addresses passed in is only used for setting up
 * the association.  It does not necessarily equal the set of addresses
 * the peer uses for the resulting association.  If the caller wants to
 * find out the set of peer addresses, it must use sctp_getpaddrs() to
 * retrieve them after the association has been set up.
 *
 * Basically do nothing but copying the addresses from user to kernel
 * land and invoking either sctp_connectx(). This is used for tunneling
 * the sctp_connectx() request through sctp_setsockopt() from userspace.
 *
 * We don't use copy_from_user() for optimization: we first do the
 * sanity checks (buffer size -fast- and access check-healthy
 * pointer); if all of those succeed, then we can alloc the memory
 * (expensive operation) needed to copy the data to kernel. Then we do
 * the copying without checking the user space area
 * (__copy_from_user()).
 *
 * On exit there is no need to do sockfd_put(), sys_setsockopt() does
 * it.
 *
 * sk        The sk of the socket
 * addrs     The pointer to the addresses in user land
 * addrssize Size of the addrs buffer
 *
1300
 * Returns >=0 if ok, <0 errno code on error.
F
Frank Filz 已提交
1301
 */
1302
static int __sctp_setsockopt_connectx(struct sock* sk,
F
Frank Filz 已提交
1303
				      struct sockaddr __user *addrs,
1304 1305
				      int addrs_size,
				      sctp_assoc_t *assoc_id)
F
Frank Filz 已提交
1306 1307 1308 1309
{
	int err = 0;
	struct sockaddr *kaddrs;

1310 1311
	pr_debug("%s: sk:%p addrs:%p addrs_size:%d\n",
		 __func__, sk, addrs, addrs_size);
F
Frank Filz 已提交
1312 1313 1314 1315 1316 1317 1318 1319 1320

	if (unlikely(addrs_size <= 0))
		return -EINVAL;

	/* Check the user passed a healthy pointer.  */
	if (unlikely(!access_ok(VERIFY_READ, addrs, addrs_size)))
		return -EFAULT;

	/* Alloc space for the address array in kernel memory.  */
1321
	kaddrs = kmalloc(addrs_size, GFP_KERNEL);
F
Frank Filz 已提交
1322 1323 1324 1325 1326 1327
	if (unlikely(!kaddrs))
		return -ENOMEM;

	if (__copy_from_user(kaddrs, addrs, addrs_size)) {
		err = -EFAULT;
	} else {
1328
		err = __sctp_connect(sk, kaddrs, addrs_size, assoc_id);
F
Frank Filz 已提交
1329 1330 1331
	}

	kfree(kaddrs);
1332

F
Frank Filz 已提交
1333 1334 1335
	return err;
}

1336 1337 1338 1339
/*
 * This is an older interface.  It's kept for backward compatibility
 * to the option that doesn't provide association id.
 */
1340 1341 1342
static int sctp_setsockopt_connectx_old(struct sock* sk,
					struct sockaddr __user *addrs,
					int addrs_size)
1343 1344 1345 1346 1347 1348 1349 1350 1351 1352
{
	return __sctp_setsockopt_connectx(sk, addrs, addrs_size, NULL);
}

/*
 * New interface for the API.  The since the API is done with a socket
 * option, to make it simple we feed back the association id is as a return
 * indication to the call.  Error is always negative and association id is
 * always positive.
 */
1353 1354 1355
static int sctp_setsockopt_connectx(struct sock* sk,
				    struct sockaddr __user *addrs,
				    int addrs_size)
1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
{
	sctp_assoc_t assoc_id = 0;
	int err = 0;

	err = __sctp_setsockopt_connectx(sk, addrs, addrs_size, &assoc_id);

	if (err)
		return err;
	else
		return assoc_id;
}

1368
/*
1369 1370 1371 1372 1373 1374
 * New (hopefully final) interface for the API.
 * We use the sctp_getaddrs_old structure so that use-space library
 * can avoid any unnecessary allocations.   The only defferent part
 * is that we store the actual length of the address buffer into the
 * addrs_num structure member.  That way we can re-use the existing
 * code.
1375
 */
1376 1377 1378
static int sctp_getsockopt_connectx3(struct sock* sk, int len,
				     char __user *optval,
				     int __user *optlen)
1379
{
1380
	struct sctp_getaddrs_old param;
1381 1382 1383
	sctp_assoc_t assoc_id = 0;
	int err = 0;

1384
	if (len < sizeof(param))
1385 1386
		return -EINVAL;

1387 1388 1389
	if (copy_from_user(&param, optval, sizeof(param)))
		return -EFAULT;

1390
	err = __sctp_setsockopt_connectx(sk,
1391 1392
			(struct sockaddr __user *)param.addrs,
			param.addr_num, &assoc_id);
1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403

	if (err == 0 || err == -EINPROGRESS) {
		if (copy_to_user(optval, &assoc_id, sizeof(assoc_id)))
			return -EFAULT;
		if (put_user(sizeof(assoc_id), optlen))
			return -EFAULT;
	}

	return err;
}

L
Linus Torvalds 已提交
1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453
/* API 3.1.4 close() - UDP Style Syntax
 * Applications use close() to perform graceful shutdown (as described in
 * Section 10.1 of [SCTP]) on ALL the associations currently represented
 * by a UDP-style socket.
 *
 * The syntax is
 *
 *   ret = close(int sd);
 *
 *   sd      - the socket descriptor of the associations to be closed.
 *
 * To gracefully shutdown a specific association represented by the
 * UDP-style socket, an application should use the sendmsg() call,
 * passing no user data, but including the appropriate flag in the
 * ancillary data (see Section xxxx).
 *
 * If sd in the close() call is a branched-off socket representing only
 * one association, the shutdown is performed on that association only.
 *
 * 4.1.6 close() - TCP Style Syntax
 *
 * Applications use close() to gracefully close down an association.
 *
 * The syntax is:
 *
 *    int close(int sd);
 *
 *      sd      - the socket descriptor of the association to be closed.
 *
 * After an application calls close() on a socket descriptor, no further
 * socket operations will succeed on that descriptor.
 *
 * API 7.1.4 SO_LINGER
 *
 * An application using the TCP-style socket can use this option to
 * perform the SCTP ABORT primitive.  The linger option structure is:
 *
 *  struct  linger {
 *     int     l_onoff;                // option on/off
 *     int     l_linger;               // linger time
 * };
 *
 * To enable the option, set l_onoff to 1.  If the l_linger value is set
 * to 0, calling close() is the same as the ABORT primitive.  If the
 * value is set to a negative value, the setsockopt() call will return
 * an error.  If the value is set to a positive value linger_time, the
 * close() can be blocked for at most linger_time ms.  If the graceful
 * shutdown phase does not finish during this period, close() will
 * return but the graceful shutdown phase continues in the system.
 */
1454
static void sctp_close(struct sock *sk, long timeout)
L
Linus Torvalds 已提交
1455
{
1456
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
1457 1458 1459
	struct sctp_endpoint *ep;
	struct sctp_association *asoc;
	struct list_head *pos, *temp;
1460
	unsigned int data_was_unread;
L
Linus Torvalds 已提交
1461

1462
	pr_debug("%s: sk:%p, timeout:%ld\n", __func__, sk, timeout);
L
Linus Torvalds 已提交
1463 1464 1465

	sctp_lock_sock(sk);
	sk->sk_shutdown = SHUTDOWN_MASK;
1466
	sk->sk_state = SCTP_SS_CLOSING;
L
Linus Torvalds 已提交
1467 1468 1469

	ep = sctp_sk(sk)->ep;

1470 1471 1472 1473
	/* Clean up any skbs sitting on the receive queue.  */
	data_was_unread = sctp_queue_purge_ulpevents(&sk->sk_receive_queue);
	data_was_unread += sctp_queue_purge_ulpevents(&sctp_sk(sk)->pd_lobby);

1474
	/* Walk all associations on an endpoint.  */
L
Linus Torvalds 已提交
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486
	list_for_each_safe(pos, temp, &ep->asocs) {
		asoc = list_entry(pos, struct sctp_association, asocs);

		if (sctp_style(sk, TCP)) {
			/* A closed association can still be in the list if
			 * it belongs to a TCP-style listening socket that is
			 * not yet accepted. If so, free it. If not, send an
			 * ABORT or SHUTDOWN based on the linger options.
			 */
			if (sctp_state(asoc, CLOSED)) {
				sctp_unhash_established(asoc);
				sctp_association_free(asoc);
1487 1488 1489
				continue;
			}
		}
L
Linus Torvalds 已提交
1490

1491 1492 1493
		if (data_was_unread || !skb_queue_empty(&asoc->ulpq.lobby) ||
		    !skb_queue_empty(&asoc->ulpq.reasm) ||
		    (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime)) {
1494 1495 1496 1497
			struct sctp_chunk *chunk;

			chunk = sctp_make_abort_user(asoc, NULL, 0);
			if (chunk)
1498
				sctp_primitive_ABORT(net, asoc, chunk);
1499
		} else
1500
			sctp_primitive_SHUTDOWN(net, asoc, NULL);
L
Linus Torvalds 已提交
1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562
	}

	/* On a TCP-style socket, block for at most linger_time if set. */
	if (sctp_style(sk, TCP) && timeout)
		sctp_wait_for_close(sk, timeout);

	/* This will run the backlog queue.  */
	sctp_release_sock(sk);

	/* Supposedly, no process has access to the socket, but
	 * the net layers still may.
	 */
	sctp_local_bh_disable();
	sctp_bh_lock_sock(sk);

	/* Hold the sock, since sk_common_release() will put sock_put()
	 * and we have just a little more cleanup.
	 */
	sock_hold(sk);
	sk_common_release(sk);

	sctp_bh_unlock_sock(sk);
	sctp_local_bh_enable();

	sock_put(sk);

	SCTP_DBG_OBJCNT_DEC(sock);
}

/* Handle EPIPE error. */
static int sctp_error(struct sock *sk, int flags, int err)
{
	if (err == -EPIPE)
		err = sock_error(sk) ? : -EPIPE;
	if (err == -EPIPE && !(flags & MSG_NOSIGNAL))
		send_sig(SIGPIPE, current, 0);
	return err;
}

/* API 3.1.3 sendmsg() - UDP Style Syntax
 *
 * An application uses sendmsg() and recvmsg() calls to transmit data to
 * and receive data from its peer.
 *
 *  ssize_t sendmsg(int socket, const struct msghdr *message,
 *                  int flags);
 *
 *  socket  - the socket descriptor of the endpoint.
 *  message - pointer to the msghdr structure which contains a single
 *            user message and possibly some ancillary data.
 *
 *            See Section 5 for complete description of the data
 *            structures.
 *
 *  flags   - flags sent or received with the user message, see Section
 *            5 for complete description of the flags.
 *
 * Note:  This function could use a rewrite especially when explicit
 * connect support comes in.
 */
/* BUG:  We do not implement the equivalent of sk_stream_wait_memory(). */

1563
static int sctp_msghdr_parse(const struct msghdr *, sctp_cmsgs_t *);
L
Linus Torvalds 已提交
1564

1565 1566
static int sctp_sendmsg(struct kiocb *iocb, struct sock *sk,
			struct msghdr *msg, size_t msg_len)
L
Linus Torvalds 已提交
1567
{
1568
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
1569 1570
	struct sctp_sock *sp;
	struct sctp_endpoint *ep;
1571
	struct sctp_association *new_asoc = NULL, *asoc = NULL;
L
Linus Torvalds 已提交
1572 1573
	struct sctp_transport *transport, *chunk_tp;
	struct sctp_chunk *chunk;
1574
	union sctp_addr to;
L
Linus Torvalds 已提交
1575
	struct sockaddr *msg_name = NULL;
1576
	struct sctp_sndrcvinfo default_sinfo;
L
Linus Torvalds 已提交
1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591
	struct sctp_sndrcvinfo *sinfo;
	struct sctp_initmsg *sinit;
	sctp_assoc_t associd = 0;
	sctp_cmsgs_t cmsgs = { NULL };
	int err;
	sctp_scope_t scope;
	long timeo;
	__u16 sinfo_flags = 0;
	struct sctp_datamsg *datamsg;
	int msg_flags = msg->msg_flags;

	err = 0;
	sp = sctp_sk(sk);
	ep = sp->ep;

1592 1593
	pr_debug("%s: sk:%p, msg:%p, msg_len:%zu ep:%p\n", __func__, sk,
		 msg, msg_len, ep);
L
Linus Torvalds 已提交
1594 1595 1596 1597 1598 1599 1600 1601 1602 1603

	/* We cannot send a message over a TCP-style listening socket. */
	if (sctp_style(sk, TCP) && sctp_sstate(sk, LISTENING)) {
		err = -EPIPE;
		goto out_nounlock;
	}

	/* Parse out the SCTP CMSGs.  */
	err = sctp_msghdr_parse(msg, &cmsgs);
	if (err) {
1604
		pr_debug("%s: msghdr parse err:%x\n", __func__, err);
L
Linus Torvalds 已提交
1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635
		goto out_nounlock;
	}

	/* Fetch the destination address for this packet.  This
	 * address only selects the association--it is not necessarily
	 * the address we will send to.
	 * For a peeled-off socket, msg_name is ignored.
	 */
	if (!sctp_style(sk, UDP_HIGH_BANDWIDTH) && msg->msg_name) {
		int msg_namelen = msg->msg_namelen;

		err = sctp_verify_addr(sk, (union sctp_addr *)msg->msg_name,
				       msg_namelen);
		if (err)
			return err;

		if (msg_namelen > sizeof(to))
			msg_namelen = sizeof(to);
		memcpy(&to, msg->msg_name, msg_namelen);
		msg_name = msg->msg_name;
	}

	sinfo = cmsgs.info;
	sinit = cmsgs.init;

	/* Did the user specify SNDRCVINFO?  */
	if (sinfo) {
		sinfo_flags = sinfo->sinfo_flags;
		associd = sinfo->sinfo_assoc_id;
	}

1636 1637
	pr_debug("%s: msg_len:%zu, sinfo_flags:0x%x\n", __func__,
		 msg_len, sinfo_flags);
L
Linus Torvalds 已提交
1638

1639 1640
	/* SCTP_EOF or SCTP_ABORT cannot be set on a TCP-style socket. */
	if (sctp_style(sk, TCP) && (sinfo_flags & (SCTP_EOF | SCTP_ABORT))) {
L
Linus Torvalds 已提交
1641 1642 1643 1644
		err = -EINVAL;
		goto out_nounlock;
	}

1645 1646 1647
	/* If SCTP_EOF is set, no data can be sent. Disallow sending zero
	 * length messages when SCTP_EOF|SCTP_ABORT is not set.
	 * If SCTP_ABORT is set, the message length could be non zero with
L
Linus Torvalds 已提交
1648
	 * the msg_iov set to the user abort reason.
1649
	 */
1650 1651
	if (((sinfo_flags & SCTP_EOF) && (msg_len > 0)) ||
	    (!(sinfo_flags & (SCTP_EOF|SCTP_ABORT)) && (msg_len == 0))) {
L
Linus Torvalds 已提交
1652 1653 1654 1655
		err = -EINVAL;
		goto out_nounlock;
	}

1656
	/* If SCTP_ADDR_OVER is set, there must be an address
L
Linus Torvalds 已提交
1657 1658
	 * specified in msg_name.
	 */
1659
	if ((sinfo_flags & SCTP_ADDR_OVER) && (!msg->msg_name)) {
L
Linus Torvalds 已提交
1660 1661 1662 1663 1664 1665
		err = -EINVAL;
		goto out_nounlock;
	}

	transport = NULL;

1666
	pr_debug("%s: about to look up association\n", __func__);
L
Linus Torvalds 已提交
1667 1668 1669 1670 1671 1672

	sctp_lock_sock(sk);

	/* If a msg_name has been specified, assume this is to be used.  */
	if (msg_name) {
		/* Look for a matching association on the endpoint. */
1673
		asoc = sctp_endpoint_lookup_assoc(ep, &to, &transport);
L
Linus Torvalds 已提交
1674 1675 1676 1677 1678 1679 1680 1681
		if (!asoc) {
			/* If we could not find a matching association on the
			 * endpoint, make sure that it is not a TCP-style
			 * socket that already has an association or there is
			 * no peeled-off association on another socket.
			 */
			if ((sctp_style(sk, TCP) &&
			     sctp_sstate(sk, ESTABLISHED)) ||
1682
			    sctp_endpoint_is_peeled_off(ep, &to)) {
L
Linus Torvalds 已提交
1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695
				err = -EADDRNOTAVAIL;
				goto out_unlock;
			}
		}
	} else {
		asoc = sctp_id2assoc(sk, associd);
		if (!asoc) {
			err = -EPIPE;
			goto out_unlock;
		}
	}

	if (asoc) {
1696
		pr_debug("%s: just looked up association:%p\n", __func__, asoc);
L
Linus Torvalds 已提交
1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707

		/* We cannot send a message on a TCP-style SCTP_SS_ESTABLISHED
		 * socket that has an association in CLOSED state. This can
		 * happen when an accepted socket has an association that is
		 * already CLOSED.
		 */
		if (sctp_state(asoc, CLOSED) && sctp_style(sk, TCP)) {
			err = -EPIPE;
			goto out_unlock;
		}

1708
		if (sinfo_flags & SCTP_EOF) {
1709 1710 1711
			pr_debug("%s: shutting down association:%p\n",
				 __func__, asoc);

1712
			sctp_primitive_SHUTDOWN(net, asoc, NULL);
L
Linus Torvalds 已提交
1713 1714 1715
			err = 0;
			goto out_unlock;
		}
1716
		if (sinfo_flags & SCTP_ABORT) {
1717 1718 1719 1720 1721 1722 1723

			chunk = sctp_make_abort_user(asoc, msg, msg_len);
			if (!chunk) {
				err = -ENOMEM;
				goto out_unlock;
			}

1724 1725 1726
			pr_debug("%s: aborting association:%p\n",
				 __func__, asoc);

1727
			sctp_primitive_ABORT(net, asoc, chunk);
L
Linus Torvalds 已提交
1728 1729 1730 1731 1732 1733 1734
			err = 0;
			goto out_unlock;
		}
	}

	/* Do we need to create the association?  */
	if (!asoc) {
1735
		pr_debug("%s: there is no association yet\n", __func__);
L
Linus Torvalds 已提交
1736

1737
		if (sinfo_flags & (SCTP_EOF | SCTP_ABORT)) {
L
Linus Torvalds 已提交
1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774
			err = -EINVAL;
			goto out_unlock;
		}

		/* Check for invalid stream against the stream counts,
		 * either the default or the user specified stream counts.
		 */
		if (sinfo) {
			if (!sinit || (sinit && !sinit->sinit_num_ostreams)) {
				/* Check against the defaults. */
				if (sinfo->sinfo_stream >=
				    sp->initmsg.sinit_num_ostreams) {
					err = -EINVAL;
					goto out_unlock;
				}
			} else {
				/* Check against the requested.  */
				if (sinfo->sinfo_stream >=
				    sinit->sinit_num_ostreams) {
					err = -EINVAL;
					goto out_unlock;
				}
			}
		}

		/*
		 * API 3.1.2 bind() - UDP Style Syntax
		 * If a bind() or sctp_bindx() is not called prior to a
		 * sendmsg() call that initiates a new association, the
		 * system picks an ephemeral port and will choose an address
		 * set equivalent to binding with a wildcard address.
		 */
		if (!ep->base.bind_addr.port) {
			if (sctp_autobind(sk)) {
				err = -EAGAIN;
				goto out_unlock;
			}
1775 1776 1777 1778 1779 1780 1781 1782 1783
		} else {
			/*
			 * If an unprivileged user inherits a one-to-many
			 * style socket with open associations on a privileged
			 * port, it MAY be permitted to accept new associations,
			 * but it SHOULD NOT be permitted to open new
			 * associations.
			 */
			if (ep->base.bind_addr.port < PROT_SOCK &&
1784
			    !ns_capable(net->user_ns, CAP_NET_BIND_SERVICE)) {
1785 1786 1787
				err = -EACCES;
				goto out_unlock;
			}
L
Linus Torvalds 已提交
1788 1789 1790 1791 1792 1793 1794 1795 1796
		}

		scope = sctp_scope(&to);
		new_asoc = sctp_association_new(ep, sk, scope, GFP_KERNEL);
		if (!new_asoc) {
			err = -ENOMEM;
			goto out_unlock;
		}
		asoc = new_asoc;
1797 1798 1799 1800 1801
		err = sctp_assoc_set_bind_addr_from_ep(asoc, scope, GFP_KERNEL);
		if (err < 0) {
			err = -ENOMEM;
			goto out_free;
		}
L
Linus Torvalds 已提交
1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819

		/* If the SCTP_INIT ancillary data is specified, set all
		 * the association init values accordingly.
		 */
		if (sinit) {
			if (sinit->sinit_num_ostreams) {
				asoc->c.sinit_num_ostreams =
					sinit->sinit_num_ostreams;
			}
			if (sinit->sinit_max_instreams) {
				asoc->c.sinit_max_instreams =
					sinit->sinit_max_instreams;
			}
			if (sinit->sinit_max_attempts) {
				asoc->max_init_attempts
					= sinit->sinit_max_attempts;
			}
			if (sinit->sinit_max_init_timeo) {
1820
				asoc->max_init_timeo =
L
Linus Torvalds 已提交
1821 1822 1823 1824 1825
				 msecs_to_jiffies(sinit->sinit_max_init_timeo);
			}
		}

		/* Prime the peer's transport structures.  */
1826
		transport = sctp_assoc_add_peer(asoc, &to, GFP_KERNEL, SCTP_UNKNOWN);
L
Linus Torvalds 已提交
1827 1828 1829 1830 1831 1832 1833
		if (!transport) {
			err = -ENOMEM;
			goto out_free;
		}
	}

	/* ASSERT: we have a valid association at this point.  */
1834
	pr_debug("%s: we have a valid association\n", __func__);
L
Linus Torvalds 已提交
1835 1836 1837 1838 1839

	if (!sinfo) {
		/* If the user didn't specify SNDRCVINFO, make up one with
		 * some defaults.
		 */
1840
		memset(&default_sinfo, 0, sizeof(default_sinfo));
L
Linus Torvalds 已提交
1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857
		default_sinfo.sinfo_stream = asoc->default_stream;
		default_sinfo.sinfo_flags = asoc->default_flags;
		default_sinfo.sinfo_ppid = asoc->default_ppid;
		default_sinfo.sinfo_context = asoc->default_context;
		default_sinfo.sinfo_timetolive = asoc->default_timetolive;
		default_sinfo.sinfo_assoc_id = sctp_assoc2id(asoc);
		sinfo = &default_sinfo;
	}

	/* API 7.1.7, the sndbuf size per association bounds the
	 * maximum size of data that can be sent in a single send call.
	 */
	if (msg_len > sk->sk_sndbuf) {
		err = -EMSGSIZE;
		goto out_free;
	}

1858
	if (asoc->pmtu_pending)
1859
		sctp_assoc_pending_pmtu(sk, asoc);
1860

L
Linus Torvalds 已提交
1861 1862 1863 1864 1865 1866 1867 1868 1869 1870
	/* If fragmentation is disabled and the message length exceeds the
	 * association fragmentation point, return EMSGSIZE.  The I-D
	 * does not specify what this error is, but this looks like
	 * a great fit.
	 */
	if (sctp_sk(sk)->disable_fragments && (msg_len > asoc->frag_point)) {
		err = -EMSGSIZE;
		goto out_free;
	}

1871 1872 1873 1874
	/* Check for invalid stream. */
	if (sinfo->sinfo_stream >= asoc->c.sinit_num_ostreams) {
		err = -EINVAL;
		goto out_free;
L
Linus Torvalds 已提交
1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885
	}

	timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
	if (!sctp_wspace(asoc)) {
		err = sctp_wait_for_sndbuf(asoc, &timeo, msg_len);
		if (err)
			goto out_free;
	}

	/* If an address is passed with the sendto/sendmsg call, it is used
	 * to override the primary destination address in the TCP model, or
1886
	 * when SCTP_ADDR_OVER flag is set in the UDP model.
L
Linus Torvalds 已提交
1887 1888
	 */
	if ((sctp_style(sk, TCP) && msg_name) ||
1889
	    (sinfo_flags & SCTP_ADDR_OVER)) {
1890
		chunk_tp = sctp_assoc_lookup_paddr(asoc, &to);
L
Linus Torvalds 已提交
1891 1892 1893 1894 1895 1896 1897 1898 1899
		if (!chunk_tp) {
			err = -EINVAL;
			goto out_free;
		}
	} else
		chunk_tp = NULL;

	/* Auto-connect, if we aren't connected already. */
	if (sctp_state(asoc, CLOSED)) {
1900
		err = sctp_primitive_ASSOCIATE(net, asoc, NULL);
L
Linus Torvalds 已提交
1901 1902
		if (err < 0)
			goto out_free;
1903 1904

		pr_debug("%s: we associated primitively\n", __func__);
L
Linus Torvalds 已提交
1905 1906 1907 1908
	}

	/* Break the message into multiple chunks of maximum size. */
	datamsg = sctp_datamsg_from_user(asoc, sinfo, msg, msg_len);
1909 1910
	if (IS_ERR(datamsg)) {
		err = PTR_ERR(datamsg);
L
Linus Torvalds 已提交
1911 1912 1913 1914
		goto out_free;
	}

	/* Now send the (possibly) fragmented message. */
1915
	list_for_each_entry(chunk, &datamsg->chunks, frag_list) {
1916
		sctp_chunk_hold(chunk);
L
Linus Torvalds 已提交
1917 1918 1919 1920 1921 1922 1923

		/* Do accounting for the write space.  */
		sctp_set_owner_w(chunk);

		chunk->transport = chunk_tp;
	}

1924 1925 1926 1927 1928
	/* Send it to the lower layers.  Note:  all chunks
	 * must either fail or succeed.   The lower layer
	 * works that way today.  Keep it that way or this
	 * breaks.
	 */
1929
	err = sctp_primitive_SEND(net, asoc, datamsg);
1930
	/* Did the lower layer accept the chunk? */
1931
	if (err) {
1932
		sctp_datamsg_free(datamsg);
1933 1934
		goto out_free;
	}
1935

1936
	pr_debug("%s: we sent primitively\n", __func__);
1937

1938 1939
	sctp_datamsg_put(datamsg);
	err = msg_len;
L
Linus Torvalds 已提交
1940 1941 1942 1943 1944 1945 1946

	/* If we are already past ASSOCIATE, the lower
	 * layers are responsible for association cleanup.
	 */
	goto out_unlock;

out_free:
1947 1948
	if (new_asoc) {
		sctp_unhash_established(asoc);
L
Linus Torvalds 已提交
1949
		sctp_association_free(asoc);
1950
	}
L
Linus Torvalds 已提交
1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991
out_unlock:
	sctp_release_sock(sk);

out_nounlock:
	return sctp_error(sk, msg_flags, err);

#if 0
do_sock_err:
	if (msg_len)
		err = msg_len;
	else
		err = sock_error(sk);
	goto out;

do_interrupted:
	if (msg_len)
		err = msg_len;
	goto out;
#endif /* 0 */
}

/* This is an extended version of skb_pull() that removes the data from the
 * start of a skb even when data is spread across the list of skb's in the
 * frag_list. len specifies the total amount of data that needs to be removed.
 * when 'len' bytes could be removed from the skb, it returns 0.
 * If 'len' exceeds the total skb length,  it returns the no. of bytes that
 * could not be removed.
 */
static int sctp_skb_pull(struct sk_buff *skb, int len)
{
	struct sk_buff *list;
	int skb_len = skb_headlen(skb);
	int rlen;

	if (len <= skb_len) {
		__skb_pull(skb, len);
		return 0;
	}
	len -= skb_len;
	__skb_pull(skb, skb_len);

1992
	skb_walk_frags(skb, list) {
L
Linus Torvalds 已提交
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
		rlen = sctp_skb_pull(list, len);
		skb->len -= (len-rlen);
		skb->data_len -= (len-rlen);

		if (!rlen)
			return 0;

		len = rlen;
	}

	return len;
}

/* API 3.1.3  recvmsg() - UDP Style Syntax
 *
 *  ssize_t recvmsg(int socket, struct msghdr *message,
 *                    int flags);
 *
 *  socket  - the socket descriptor of the endpoint.
 *  message - pointer to the msghdr structure which contains a single
 *            user message and possibly some ancillary data.
 *
 *            See Section 5 for complete description of the data
 *            structures.
 *
 *  flags   - flags sent or received with the user message, see Section
 *            5 for complete description of the flags.
 */
static struct sk_buff *sctp_skb_recv_datagram(struct sock *, int, int, int *);

2023 2024 2025
static int sctp_recvmsg(struct kiocb *iocb, struct sock *sk,
			struct msghdr *msg, size_t len, int noblock,
			int flags, int *addr_len)
L
Linus Torvalds 已提交
2026 2027 2028 2029 2030 2031 2032 2033
{
	struct sctp_ulpevent *event = NULL;
	struct sctp_sock *sp = sctp_sk(sk);
	struct sk_buff *skb;
	int copied;
	int err = 0;
	int skb_len;

2034 2035 2036
	pr_debug("%s: sk:%p, msghdr:%p, len:%zd, noblock:%d, flags:0x%x, "
		 "addr_len:%p)\n", __func__, sk, msg, len, noblock, flags,
		 addr_len);
L
Linus Torvalds 已提交
2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064

	sctp_lock_sock(sk);

	if (sctp_style(sk, TCP) && !sctp_sstate(sk, ESTABLISHED)) {
		err = -ENOTCONN;
		goto out;
	}

	skb = sctp_skb_recv_datagram(sk, flags, noblock, &err);
	if (!skb)
		goto out;

	/* Get the total length of the skb including any skb's in the
	 * frag_list.
	 */
	skb_len = skb->len;

	copied = skb_len;
	if (copied > len)
		copied = len;

	err = skb_copy_datagram_iovec(skb, 0, msg->msg_iov, copied);

	event = sctp_skb2event(skb);

	if (err)
		goto out_free;

2065
	sock_recv_ts_and_drops(msg, sk, skb);
L
Linus Torvalds 已提交
2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098
	if (sctp_ulpevent_is_notification(event)) {
		msg->msg_flags |= MSG_NOTIFICATION;
		sp->pf->event_msgname(event, msg->msg_name, addr_len);
	} else {
		sp->pf->skb_msgname(skb, msg->msg_name, addr_len);
	}

	/* Check if we allow SCTP_SNDRCVINFO. */
	if (sp->subscribe.sctp_data_io_event)
		sctp_ulpevent_read_sndrcvinfo(event, msg);
#if 0
	/* FIXME: we should be calling IP/IPv6 layers.  */
	if (sk->sk_protinfo.af_inet.cmsg_flags)
		ip_cmsg_recv(msg, skb);
#endif

	err = copied;

	/* If skb's length exceeds the user's buffer, update the skb and
	 * push it back to the receive_queue so that the next call to
	 * recvmsg() will return the remaining data. Don't set MSG_EOR.
	 */
	if (skb_len > copied) {
		msg->msg_flags &= ~MSG_EOR;
		if (flags & MSG_PEEK)
			goto out_free;
		sctp_skb_pull(skb, copied);
		skb_queue_head(&sk->sk_receive_queue, skb);

		/* When only partial message is copied to the user, increase
		 * rwnd by that amount. If all the data in the skb is read,
		 * rwnd is updated when the event is freed.
		 */
2099 2100
		if (!sctp_ulpevent_is_notification(event))
			sctp_assoc_rwnd_increase(event->asoc, copied);
L
Linus Torvalds 已提交
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
		goto out;
	} else if ((event->msg_flags & MSG_NOTIFICATION) ||
		   (event->msg_flags & MSG_EOR))
		msg->msg_flags |= MSG_EOR;
	else
		msg->msg_flags &= ~MSG_EOR;

out_free:
	if (flags & MSG_PEEK) {
		/* Release the skb reference acquired after peeking the skb in
		 * sctp_skb_recv_datagram().
		 */
		kfree_skb(skb);
	} else {
		/* Free the event which includes releasing the reference to
		 * the owner of the skb, freeing the skb and updating the
		 * rwnd.
		 */
		sctp_ulpevent_free(event);
	}
out:
	sctp_release_sock(sk);
	return err;
}

/* 7.1.12 Enable/Disable message fragmentation (SCTP_DISABLE_FRAGMENTS)
 *
 * This option is a on/off flag.  If enabled no SCTP message
 * fragmentation will be performed.  Instead if a message being sent
 * exceeds the current PMTU size, the message will NOT be sent and
 * instead a error will be indicated to the user.
 */
static int sctp_setsockopt_disable_fragments(struct sock *sk,
2134 2135
					     char __user *optval,
					     unsigned int optlen)
L
Linus Torvalds 已提交
2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150
{
	int val;

	if (optlen < sizeof(int))
		return -EINVAL;

	if (get_user(val, (int __user *)optval))
		return -EFAULT;

	sctp_sk(sk)->disable_fragments = (val == 0) ? 0 : 1;

	return 0;
}

static int sctp_setsockopt_events(struct sock *sk, char __user *optval,
2151
				  unsigned int optlen)
L
Linus Torvalds 已提交
2152
{
2153 2154 2155
	struct sctp_association *asoc;
	struct sctp_ulpevent *event;

2156
	if (optlen > sizeof(struct sctp_event_subscribe))
L
Linus Torvalds 已提交
2157 2158 2159
		return -EINVAL;
	if (copy_from_user(&sctp_sk(sk)->subscribe, optval, optlen))
		return -EFAULT;
2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179

	/*
	 * At the time when a user app subscribes to SCTP_SENDER_DRY_EVENT,
	 * if there is no data to be sent or retransmit, the stack will
	 * immediately send up this notification.
	 */
	if (sctp_ulpevent_type_enabled(SCTP_SENDER_DRY_EVENT,
				       &sctp_sk(sk)->subscribe)) {
		asoc = sctp_id2assoc(sk, 0);

		if (asoc && sctp_outq_is_empty(&asoc->outqueue)) {
			event = sctp_ulpevent_make_sender_dry_event(asoc,
					GFP_ATOMIC);
			if (!event)
				return -ENOMEM;

			sctp_ulpq_tail_event(&asoc->ulpq, event);
		}
	}

L
Linus Torvalds 已提交
2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194
	return 0;
}

/* 7.1.8 Automatic Close of associations (SCTP_AUTOCLOSE)
 *
 * This socket option is applicable to the UDP-style socket only.  When
 * set it will cause associations that are idle for more than the
 * specified number of seconds to automatically close.  An association
 * being idle is defined an association that has NOT sent or received
 * user data.  The special value of '0' indicates that no automatic
 * close of any associations should be performed.  The option expects an
 * integer defining the number of seconds of idle time before an
 * association is closed.
 */
static int sctp_setsockopt_autoclose(struct sock *sk, char __user *optval,
2195
				     unsigned int optlen)
L
Linus Torvalds 已提交
2196 2197
{
	struct sctp_sock *sp = sctp_sk(sk);
2198
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
2199 2200 2201 2202 2203 2204 2205 2206 2207

	/* Applicable to UDP-style socket only */
	if (sctp_style(sk, TCP))
		return -EOPNOTSUPP;
	if (optlen != sizeof(int))
		return -EINVAL;
	if (copy_from_user(&sp->autoclose, optval, optlen))
		return -EFAULT;

2208 2209 2210
	if (sp->autoclose > net->sctp.max_autoclose)
		sp->autoclose = net->sctp.max_autoclose;

L
Linus Torvalds 已提交
2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223
	return 0;
}

/* 7.1.13 Peer Address Parameters (SCTP_PEER_ADDR_PARAMS)
 *
 * Applications can enable or disable heartbeats for any peer address of
 * an association, modify an address's heartbeat interval, force a
 * heartbeat to be sent immediately, and adjust the address's maximum
 * number of retransmissions sent before an address is considered
 * unreachable.  The following structure is used to access and modify an
 * address's parameters:
 *
 *  struct sctp_paddrparams {
2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235
 *     sctp_assoc_t            spp_assoc_id;
 *     struct sockaddr_storage spp_address;
 *     uint32_t                spp_hbinterval;
 *     uint16_t                spp_pathmaxrxt;
 *     uint32_t                spp_pathmtu;
 *     uint32_t                spp_sackdelay;
 *     uint32_t                spp_flags;
 * };
 *
 *   spp_assoc_id    - (one-to-many style socket) This is filled in the
 *                     application, and identifies the association for
 *                     this query.
L
Linus Torvalds 已提交
2236 2237
 *   spp_address     - This specifies which address is of interest.
 *   spp_hbinterval  - This contains the value of the heartbeat interval,
2238 2239 2240
 *                     in milliseconds.  If a  value of zero
 *                     is present in this field then no changes are to
 *                     be made to this parameter.
L
Linus Torvalds 已提交
2241 2242
 *   spp_pathmaxrxt  - This contains the maximum number of
 *                     retransmissions before this address shall be
2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280
 *                     considered unreachable. If a  value of zero
 *                     is present in this field then no changes are to
 *                     be made to this parameter.
 *   spp_pathmtu     - When Path MTU discovery is disabled the value
 *                     specified here will be the "fixed" path mtu.
 *                     Note that if the spp_address field is empty
 *                     then all associations on this address will
 *                     have this fixed path mtu set upon them.
 *
 *   spp_sackdelay   - When delayed sack is enabled, this value specifies
 *                     the number of milliseconds that sacks will be delayed
 *                     for. This value will apply to all addresses of an
 *                     association if the spp_address field is empty. Note
 *                     also, that if delayed sack is enabled and this
 *                     value is set to 0, no change is made to the last
 *                     recorded delayed sack timer value.
 *
 *   spp_flags       - These flags are used to control various features
 *                     on an association. The flag field may contain
 *                     zero or more of the following options.
 *
 *                     SPP_HB_ENABLE  - Enable heartbeats on the
 *                     specified address. Note that if the address
 *                     field is empty all addresses for the association
 *                     have heartbeats enabled upon them.
 *
 *                     SPP_HB_DISABLE - Disable heartbeats on the
 *                     speicifed address. Note that if the address
 *                     field is empty all addresses for the association
 *                     will have their heartbeats disabled. Note also
 *                     that SPP_HB_ENABLE and SPP_HB_DISABLE are
 *                     mutually exclusive, only one of these two should
 *                     be specified. Enabling both fields will have
 *                     undetermined results.
 *
 *                     SPP_HB_DEMAND - Request a user initiated heartbeat
 *                     to be made immediately.
 *
2281 2282 2283 2284
 *                     SPP_HB_TIME_IS_ZERO - Specify's that the time for
 *                     heartbeat delayis to be set to the value of 0
 *                     milliseconds.
 *
2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309
 *                     SPP_PMTUD_ENABLE - This field will enable PMTU
 *                     discovery upon the specified address. Note that
 *                     if the address feild is empty then all addresses
 *                     on the association are effected.
 *
 *                     SPP_PMTUD_DISABLE - This field will disable PMTU
 *                     discovery upon the specified address. Note that
 *                     if the address feild is empty then all addresses
 *                     on the association are effected. Not also that
 *                     SPP_PMTUD_ENABLE and SPP_PMTUD_DISABLE are mutually
 *                     exclusive. Enabling both will have undetermined
 *                     results.
 *
 *                     SPP_SACKDELAY_ENABLE - Setting this flag turns
 *                     on delayed sack. The time specified in spp_sackdelay
 *                     is used to specify the sack delay for this address. Note
 *                     that if spp_address is empty then all addresses will
 *                     enable delayed sack and take on the sack delay
 *                     value specified in spp_sackdelay.
 *                     SPP_SACKDELAY_DISABLE - Setting this flag turns
 *                     off delayed sack. If the spp_address field is blank then
 *                     delayed sack is disabled for the entire association. Note
 *                     also that this field is mutually exclusive to
 *                     SPP_SACKDELAY_ENABLE, setting both will have undefined
 *                     results.
L
Linus Torvalds 已提交
2310
 */
A
Adrian Bunk 已提交
2311 2312 2313 2314 2315 2316 2317
static int sctp_apply_peer_addr_params(struct sctp_paddrparams *params,
				       struct sctp_transport   *trans,
				       struct sctp_association *asoc,
				       struct sctp_sock        *sp,
				       int                      hb_change,
				       int                      pmtud_change,
				       int                      sackdelay_change)
2318 2319 2320 2321
{
	int error;

	if (params->spp_flags & SPP_HB_DEMAND && trans) {
2322 2323 2324
		struct net *net = sock_net(trans->asoc->base.sk);

		error = sctp_primitive_REQUESTHEARTBEAT(net, trans->asoc, trans);
2325 2326 2327 2328
		if (error)
			return error;
	}

2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352
	/* Note that unless the spp_flag is set to SPP_HB_ENABLE the value of
	 * this field is ignored.  Note also that a value of zero indicates
	 * the current setting should be left unchanged.
	 */
	if (params->spp_flags & SPP_HB_ENABLE) {

		/* Re-zero the interval if the SPP_HB_TIME_IS_ZERO is
		 * set.  This lets us use 0 value when this flag
		 * is set.
		 */
		if (params->spp_flags & SPP_HB_TIME_IS_ZERO)
			params->spp_hbinterval = 0;

		if (params->spp_hbinterval ||
		    (params->spp_flags & SPP_HB_TIME_IS_ZERO)) {
			if (trans) {
				trans->hbinterval =
				    msecs_to_jiffies(params->spp_hbinterval);
			} else if (asoc) {
				asoc->hbinterval =
				    msecs_to_jiffies(params->spp_hbinterval);
			} else {
				sp->hbinterval = params->spp_hbinterval;
			}
2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368
		}
	}

	if (hb_change) {
		if (trans) {
			trans->param_flags =
				(trans->param_flags & ~SPP_HB) | hb_change;
		} else if (asoc) {
			asoc->param_flags =
				(asoc->param_flags & ~SPP_HB) | hb_change;
		} else {
			sp->param_flags =
				(sp->param_flags & ~SPP_HB) | hb_change;
		}
	}

2369 2370 2371 2372 2373 2374
	/* When Path MTU discovery is disabled the value specified here will
	 * be the "fixed" path mtu (i.e. the value of the spp_flags field must
	 * include the flag SPP_PMTUD_DISABLE for this field to have any
	 * effect).
	 */
	if ((params->spp_flags & SPP_PMTUD_DISABLE) && params->spp_pathmtu) {
2375 2376
		if (trans) {
			trans->pathmtu = params->spp_pathmtu;
2377
			sctp_assoc_sync_pmtu(sctp_opt2sk(sp), asoc);
2378 2379
		} else if (asoc) {
			asoc->pathmtu = params->spp_pathmtu;
2380
			sctp_frag_point(asoc, params->spp_pathmtu);
2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392
		} else {
			sp->pathmtu = params->spp_pathmtu;
		}
	}

	if (pmtud_change) {
		if (trans) {
			int update = (trans->param_flags & SPP_PMTUD_DISABLE) &&
				(params->spp_flags & SPP_PMTUD_ENABLE);
			trans->param_flags =
				(trans->param_flags & ~SPP_PMTUD) | pmtud_change;
			if (update) {
2393
				sctp_transport_pmtu(trans, sctp_opt2sk(sp));
2394
				sctp_assoc_sync_pmtu(sctp_opt2sk(sp), asoc);
2395 2396 2397 2398 2399 2400 2401 2402 2403 2404
			}
		} else if (asoc) {
			asoc->param_flags =
				(asoc->param_flags & ~SPP_PMTUD) | pmtud_change;
		} else {
			sp->param_flags =
				(sp->param_flags & ~SPP_PMTUD) | pmtud_change;
		}
	}

2405 2406 2407 2408 2409
	/* Note that unless the spp_flag is set to SPP_SACKDELAY_ENABLE the
	 * value of this field is ignored.  Note also that a value of zero
	 * indicates the current setting should be left unchanged.
	 */
	if ((params->spp_flags & SPP_SACKDELAY_ENABLE) && params->spp_sackdelay) {
2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436
		if (trans) {
			trans->sackdelay =
				msecs_to_jiffies(params->spp_sackdelay);
		} else if (asoc) {
			asoc->sackdelay =
				msecs_to_jiffies(params->spp_sackdelay);
		} else {
			sp->sackdelay = params->spp_sackdelay;
		}
	}

	if (sackdelay_change) {
		if (trans) {
			trans->param_flags =
				(trans->param_flags & ~SPP_SACKDELAY) |
				sackdelay_change;
		} else if (asoc) {
			asoc->param_flags =
				(asoc->param_flags & ~SPP_SACKDELAY) |
				sackdelay_change;
		} else {
			sp->param_flags =
				(sp->param_flags & ~SPP_SACKDELAY) |
				sackdelay_change;
		}
	}

2437 2438
	/* Note that a value of zero indicates the current setting should be
	   left unchanged.
2439
	 */
2440
	if (params->spp_pathmaxrxt) {
2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452
		if (trans) {
			trans->pathmaxrxt = params->spp_pathmaxrxt;
		} else if (asoc) {
			asoc->pathmaxrxt = params->spp_pathmaxrxt;
		} else {
			sp->pathmaxrxt = params->spp_pathmaxrxt;
		}
	}

	return 0;
}

L
Linus Torvalds 已提交
2453
static int sctp_setsockopt_peer_addr_params(struct sock *sk,
2454 2455
					    char __user *optval,
					    unsigned int optlen)
L
Linus Torvalds 已提交
2456
{
2457 2458 2459 2460
	struct sctp_paddrparams  params;
	struct sctp_transport   *trans = NULL;
	struct sctp_association *asoc = NULL;
	struct sctp_sock        *sp = sctp_sk(sk);
L
Linus Torvalds 已提交
2461
	int error;
2462
	int hb_change, pmtud_change, sackdelay_change;
L
Linus Torvalds 已提交
2463 2464

	if (optlen != sizeof(struct sctp_paddrparams))
2465
		return -EINVAL;
2466

L
Linus Torvalds 已提交
2467 2468 2469
	if (copy_from_user(&params, optval, optlen))
		return -EFAULT;

2470 2471 2472 2473 2474 2475 2476 2477 2478
	/* Validate flags and value parameters. */
	hb_change        = params.spp_flags & SPP_HB;
	pmtud_change     = params.spp_flags & SPP_PMTUD;
	sackdelay_change = params.spp_flags & SPP_SACKDELAY;

	if (hb_change        == SPP_HB ||
	    pmtud_change     == SPP_PMTUD ||
	    sackdelay_change == SPP_SACKDELAY ||
	    params.spp_sackdelay > 500 ||
2479 2480
	    (params.spp_pathmtu &&
	     params.spp_pathmtu < SCTP_DEFAULT_MINSEGMENT))
2481
		return -EINVAL;
L
Linus Torvalds 已提交
2482

2483 2484 2485
	/* If an address other than INADDR_ANY is specified, and
	 * no transport is found, then the request is invalid.
	 */
2486
	if (!sctp_is_any(sk, (union sctp_addr *)&params.spp_address)) {
2487 2488 2489
		trans = sctp_addr_id2transport(sk, &params.spp_address,
					       params.spp_assoc_id);
		if (!trans)
L
Linus Torvalds 已提交
2490 2491 2492
			return -EINVAL;
	}

2493 2494 2495 2496 2497 2498
	/* Get association, if assoc_id != 0 and the socket is a one
	 * to many style socket, and an association was not found, then
	 * the id was invalid.
	 */
	asoc = sctp_id2assoc(sk, params.spp_assoc_id);
	if (!asoc && params.spp_assoc_id && sctp_style(sk, UDP))
L
Linus Torvalds 已提交
2499 2500
		return -EINVAL;

2501 2502
	/* Heartbeat demand can only be sent on a transport or
	 * association, but not a socket.
L
Linus Torvalds 已提交
2503
	 */
2504 2505 2506 2507 2508 2509 2510
	if (params.spp_flags & SPP_HB_DEMAND && !trans && !asoc)
		return -EINVAL;

	/* Process parameters. */
	error = sctp_apply_peer_addr_params(&params, trans, asoc, sp,
					    hb_change, pmtud_change,
					    sackdelay_change);
L
Linus Torvalds 已提交
2511

2512 2513 2514 2515 2516
	if (error)
		return error;

	/* If changes are for association, also apply parameters to each
	 * transport.
L
Linus Torvalds 已提交
2517
	 */
2518
	if (!trans && asoc) {
2519 2520
		list_for_each_entry(trans, &asoc->peer.transport_addr_list,
				transports) {
2521 2522 2523 2524 2525
			sctp_apply_peer_addr_params(&params, trans, asoc, sp,
						    hb_change, pmtud_change,
						    sackdelay_change);
		}
	}
L
Linus Torvalds 已提交
2526 2527 2528 2529

	return 0;
}

2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563
/*
 * 7.1.23.  Get or set delayed ack timer (SCTP_DELAYED_SACK)
 *
 * This option will effect the way delayed acks are performed.  This
 * option allows you to get or set the delayed ack time, in
 * milliseconds.  It also allows changing the delayed ack frequency.
 * Changing the frequency to 1 disables the delayed sack algorithm.  If
 * the assoc_id is 0, then this sets or gets the endpoints default
 * values.  If the assoc_id field is non-zero, then the set or get
 * effects the specified association for the one to many model (the
 * assoc_id field is ignored by the one to one model).  Note that if
 * sack_delay or sack_freq are 0 when setting this option, then the
 * current values will remain unchanged.
 *
 * struct sctp_sack_info {
 *     sctp_assoc_t            sack_assoc_id;
 *     uint32_t                sack_delay;
 *     uint32_t                sack_freq;
 * };
 *
 * sack_assoc_id -  This parameter, indicates which association the user
 *    is performing an action upon.  Note that if this field's value is
 *    zero then the endpoints default value is changed (effecting future
 *    associations only).
 *
 * sack_delay -  This parameter contains the number of milliseconds that
 *    the user is requesting the delayed ACK timer be set to.  Note that
 *    this value is defined in the standard to be between 200 and 500
 *    milliseconds.
 *
 * sack_freq -  This parameter contains the number of packets that must
 *    be received before a sack is sent without waiting for the delay
 *    timer to expire.  The default value for this is 2, setting this
 *    value to 1 will disable the delayed sack algorithm.
2564 2565
 */

2566
static int sctp_setsockopt_delayed_ack(struct sock *sk,
2567
				       char __user *optval, unsigned int optlen)
2568
{
2569
	struct sctp_sack_info    params;
2570 2571 2572 2573
	struct sctp_transport   *trans = NULL;
	struct sctp_association *asoc = NULL;
	struct sctp_sock        *sp = sctp_sk(sk);

2574 2575 2576
	if (optlen == sizeof(struct sctp_sack_info)) {
		if (copy_from_user(&params, optval, optlen))
			return -EFAULT;
2577

2578 2579 2580
		if (params.sack_delay == 0 && params.sack_freq == 0)
			return 0;
	} else if (optlen == sizeof(struct sctp_assoc_value)) {
2581 2582
		pr_warn("Use of struct sctp_assoc_value in delayed_ack socket option deprecated\n");
		pr_warn("Use struct sctp_sack_info instead\n");
2583 2584 2585 2586 2587 2588 2589 2590
		if (copy_from_user(&params, optval, optlen))
			return -EFAULT;

		if (params.sack_delay == 0)
			params.sack_freq = 1;
		else
			params.sack_freq = 0;
	} else
2591
		return -EINVAL;
2592 2593

	/* Validate value parameter. */
2594
	if (params.sack_delay > 500)
2595 2596
		return -EINVAL;

2597
	/* Get association, if sack_assoc_id != 0 and the socket is a one
2598 2599
	 * to many style socket, and an association was not found, then
	 * the id was invalid.
2600
	 */
2601 2602
	asoc = sctp_id2assoc(sk, params.sack_assoc_id);
	if (!asoc && params.sack_assoc_id && sctp_style(sk, UDP))
2603 2604
		return -EINVAL;

2605
	if (params.sack_delay) {
2606 2607
		if (asoc) {
			asoc->sackdelay =
2608
				msecs_to_jiffies(params.sack_delay);
2609
			asoc->param_flags =
2610 2611 2612
				(asoc->param_flags & ~SPP_SACKDELAY) |
				SPP_SACKDELAY_ENABLE;
		} else {
2613
			sp->sackdelay = params.sack_delay;
2614
			sp->param_flags =
2615 2616 2617
				(sp->param_flags & ~SPP_SACKDELAY) |
				SPP_SACKDELAY_ENABLE;
		}
2618 2619 2620
	}

	if (params.sack_freq == 1) {
2621
		if (asoc) {
2622
			asoc->param_flags =
2623 2624 2625
				(asoc->param_flags & ~SPP_SACKDELAY) |
				SPP_SACKDELAY_DISABLE;
		} else {
2626
			sp->param_flags =
2627 2628 2629
				(sp->param_flags & ~SPP_SACKDELAY) |
				SPP_SACKDELAY_DISABLE;
		}
2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641
	} else if (params.sack_freq > 1) {
		if (asoc) {
			asoc->sackfreq = params.sack_freq;
			asoc->param_flags =
				(asoc->param_flags & ~SPP_SACKDELAY) |
				SPP_SACKDELAY_ENABLE;
		} else {
			sp->sackfreq = params.sack_freq;
			sp->param_flags =
				(sp->param_flags & ~SPP_SACKDELAY) |
				SPP_SACKDELAY_ENABLE;
		}
2642 2643 2644 2645
	}

	/* If change is for association, also apply to each transport. */
	if (asoc) {
2646 2647
		list_for_each_entry(trans, &asoc->peer.transport_addr_list,
				transports) {
2648
			if (params.sack_delay) {
2649
				trans->sackdelay =
2650
					msecs_to_jiffies(params.sack_delay);
2651
				trans->param_flags =
2652 2653
					(trans->param_flags & ~SPP_SACKDELAY) |
					SPP_SACKDELAY_ENABLE;
2654
			}
2655
			if (params.sack_freq == 1) {
2656
				trans->param_flags =
2657 2658
					(trans->param_flags & ~SPP_SACKDELAY) |
					SPP_SACKDELAY_DISABLE;
2659 2660 2661 2662 2663
			} else if (params.sack_freq > 1) {
				trans->sackfreq = params.sack_freq;
				trans->param_flags =
					(trans->param_flags & ~SPP_SACKDELAY) |
					SPP_SACKDELAY_ENABLE;
2664 2665 2666
			}
		}
	}
2667

2668 2669 2670
	return 0;
}

L
Linus Torvalds 已提交
2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681
/* 7.1.3 Initialization Parameters (SCTP_INITMSG)
 *
 * Applications can specify protocol parameters for the default association
 * initialization.  The option name argument to setsockopt() and getsockopt()
 * is SCTP_INITMSG.
 *
 * Setting initialization parameters is effective only on an unconnected
 * socket (for UDP-style sockets only future associations are effected
 * by the change).  With TCP-style sockets, this option is inherited by
 * sockets derived from a listener socket.
 */
2682
static int sctp_setsockopt_initmsg(struct sock *sk, char __user *optval, unsigned int optlen)
L
Linus Torvalds 已提交
2683 2684 2685 2686 2687 2688 2689 2690 2691 2692
{
	struct sctp_initmsg sinit;
	struct sctp_sock *sp = sctp_sk(sk);

	if (optlen != sizeof(struct sctp_initmsg))
		return -EINVAL;
	if (copy_from_user(&sinit, optval, optlen))
		return -EFAULT;

	if (sinit.sinit_num_ostreams)
2693
		sp->initmsg.sinit_num_ostreams = sinit.sinit_num_ostreams;
L
Linus Torvalds 已提交
2694
	if (sinit.sinit_max_instreams)
2695
		sp->initmsg.sinit_max_instreams = sinit.sinit_max_instreams;
L
Linus Torvalds 已提交
2696
	if (sinit.sinit_max_attempts)
2697
		sp->initmsg.sinit_max_attempts = sinit.sinit_max_attempts;
L
Linus Torvalds 已提交
2698
	if (sinit.sinit_max_init_timeo)
2699
		sp->initmsg.sinit_max_init_timeo = sinit.sinit_max_init_timeo;
L
Linus Torvalds 已提交
2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718

	return 0;
}

/*
 * 7.1.14 Set default send parameters (SCTP_DEFAULT_SEND_PARAM)
 *
 *   Applications that wish to use the sendto() system call may wish to
 *   specify a default set of parameters that would normally be supplied
 *   through the inclusion of ancillary data.  This socket option allows
 *   such an application to set the default sctp_sndrcvinfo structure.
 *   The application that wishes to use this socket option simply passes
 *   in to this call the sctp_sndrcvinfo structure defined in Section
 *   5.2.2) The input parameters accepted by this call include
 *   sinfo_stream, sinfo_flags, sinfo_ppid, sinfo_context,
 *   sinfo_timetolive.  The user must provide the sinfo_assoc_id field in
 *   to this call if the caller is using the UDP model.
 */
static int sctp_setsockopt_default_send_param(struct sock *sk,
2719 2720
					      char __user *optval,
					      unsigned int optlen)
L
Linus Torvalds 已提交
2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758
{
	struct sctp_sndrcvinfo info;
	struct sctp_association *asoc;
	struct sctp_sock *sp = sctp_sk(sk);

	if (optlen != sizeof(struct sctp_sndrcvinfo))
		return -EINVAL;
	if (copy_from_user(&info, optval, optlen))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, info.sinfo_assoc_id);
	if (!asoc && info.sinfo_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	if (asoc) {
		asoc->default_stream = info.sinfo_stream;
		asoc->default_flags = info.sinfo_flags;
		asoc->default_ppid = info.sinfo_ppid;
		asoc->default_context = info.sinfo_context;
		asoc->default_timetolive = info.sinfo_timetolive;
	} else {
		sp->default_stream = info.sinfo_stream;
		sp->default_flags = info.sinfo_flags;
		sp->default_ppid = info.sinfo_ppid;
		sp->default_context = info.sinfo_context;
		sp->default_timetolive = info.sinfo_timetolive;
	}

	return 0;
}

/* 7.1.10 Set Primary Address (SCTP_PRIMARY_ADDR)
 *
 * Requests that the local SCTP stack use the enclosed peer address as
 * the association primary.  The enclosed address must be one of the
 * association peer's addresses.
 */
static int sctp_setsockopt_primary_addr(struct sock *sk, char __user *optval,
2759
					unsigned int optlen)
L
Linus Torvalds 已提交
2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787
{
	struct sctp_prim prim;
	struct sctp_transport *trans;

	if (optlen != sizeof(struct sctp_prim))
		return -EINVAL;

	if (copy_from_user(&prim, optval, sizeof(struct sctp_prim)))
		return -EFAULT;

	trans = sctp_addr_id2transport(sk, &prim.ssp_addr, prim.ssp_assoc_id);
	if (!trans)
		return -EINVAL;

	sctp_assoc_set_primary(trans->asoc, trans);

	return 0;
}

/*
 * 7.1.5 SCTP_NODELAY
 *
 * Turn on/off any Nagle-like algorithm.  This means that packets are
 * generally sent as soon as possible and no unnecessary delays are
 * introduced, at the cost of more packets in the network.  Expects an
 *  integer boolean flag.
 */
static int sctp_setsockopt_nodelay(struct sock *sk, char __user *optval,
2788
				   unsigned int optlen)
L
Linus Torvalds 已提交
2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812
{
	int val;

	if (optlen < sizeof(int))
		return -EINVAL;
	if (get_user(val, (int __user *)optval))
		return -EFAULT;

	sctp_sk(sk)->nodelay = (val == 0) ? 0 : 1;
	return 0;
}

/*
 *
 * 7.1.1 SCTP_RTOINFO
 *
 * The protocol parameters used to initialize and bound retransmission
 * timeout (RTO) are tunable. sctp_rtoinfo structure is used to access
 * and modify these parameters.
 * All parameters are time values, in milliseconds.  A value of 0, when
 * modifying the parameters, indicates that the current value should not
 * be changed.
 *
 */
2813 2814
static int sctp_setsockopt_rtoinfo(struct sock *sk, char __user *optval, unsigned int optlen)
{
L
Linus Torvalds 已提交
2815 2816
	struct sctp_rtoinfo rtoinfo;
	struct sctp_association *asoc;
2817 2818
	unsigned long rto_min, rto_max;
	struct sctp_sock *sp = sctp_sk(sk);
L
Linus Torvalds 已提交
2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831

	if (optlen != sizeof (struct sctp_rtoinfo))
		return -EINVAL;

	if (copy_from_user(&rtoinfo, optval, optlen))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, rtoinfo.srto_assoc_id);

	/* Set the values to the specific association */
	if (!asoc && rtoinfo.srto_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847
	rto_max = rtoinfo.srto_max;
	rto_min = rtoinfo.srto_min;

	if (rto_max)
		rto_max = asoc ? msecs_to_jiffies(rto_max) : rto_max;
	else
		rto_max = asoc ? asoc->rto_max : sp->rtoinfo.srto_max;

	if (rto_min)
		rto_min = asoc ? msecs_to_jiffies(rto_min) : rto_min;
	else
		rto_min = asoc ? asoc->rto_min : sp->rtoinfo.srto_min;

	if (rto_min > rto_max)
		return -EINVAL;

L
Linus Torvalds 已提交
2848 2849
	if (asoc) {
		if (rtoinfo.srto_initial != 0)
2850
			asoc->rto_initial =
L
Linus Torvalds 已提交
2851
				msecs_to_jiffies(rtoinfo.srto_initial);
2852 2853
		asoc->rto_max = rto_max;
		asoc->rto_min = rto_min;
L
Linus Torvalds 已提交
2854 2855 2856 2857 2858 2859
	} else {
		/* If there is no association or the association-id = 0
		 * set the values to the endpoint.
		 */
		if (rtoinfo.srto_initial != 0)
			sp->rtoinfo.srto_initial = rtoinfo.srto_initial;
2860 2861
		sp->rtoinfo.srto_max = rto_max;
		sp->rtoinfo.srto_min = rto_min;
L
Linus Torvalds 已提交
2862 2863 2864 2865 2866 2867 2868 2869 2870
	}

	return 0;
}

/*
 *
 * 7.1.2 SCTP_ASSOCINFO
 *
2871
 * This option is used to tune the maximum retransmission attempts
L
Linus Torvalds 已提交
2872 2873 2874 2875 2876 2877
 * of the association.
 * Returns an error if the new association retransmission value is
 * greater than the sum of the retransmission value  of the peer.
 * See [SCTP] for more information.
 *
 */
2878
static int sctp_setsockopt_associnfo(struct sock *sk, char __user *optval, unsigned int optlen)
L
Linus Torvalds 已提交
2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895
{

	struct sctp_assocparams assocparams;
	struct sctp_association *asoc;

	if (optlen != sizeof(struct sctp_assocparams))
		return -EINVAL;
	if (copy_from_user(&assocparams, optval, optlen))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, assocparams.sasoc_assoc_id);

	if (!asoc && assocparams.sasoc_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	/* Set the values to the specific association */
	if (asoc) {
2896 2897 2898 2899 2900
		if (assocparams.sasoc_asocmaxrxt != 0) {
			__u32 path_sum = 0;
			int   paths = 0;
			struct sctp_transport *peer_addr;

2901 2902
			list_for_each_entry(peer_addr, &asoc->peer.transport_addr_list,
					transports) {
2903 2904 2905 2906
				path_sum += peer_addr->pathmaxrxt;
				paths++;
			}

2907
			/* Only validate asocmaxrxt if we have more than
2908 2909 2910 2911 2912 2913 2914 2915
			 * one path/transport.  We do this because path
			 * retransmissions are only counted when we have more
			 * then one path.
			 */
			if (paths > 1 &&
			    assocparams.sasoc_asocmaxrxt > path_sum)
				return -EINVAL;

L
Linus Torvalds 已提交
2916
			asoc->max_retrans = assocparams.sasoc_asocmaxrxt;
2917 2918
		}

2919 2920
		if (assocparams.sasoc_cookie_life != 0)
			asoc->cookie_life = ms_to_ktime(assocparams.sasoc_cookie_life);
L
Linus Torvalds 已提交
2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944
	} else {
		/* Set the values to the endpoint */
		struct sctp_sock *sp = sctp_sk(sk);

		if (assocparams.sasoc_asocmaxrxt != 0)
			sp->assocparams.sasoc_asocmaxrxt =
						assocparams.sasoc_asocmaxrxt;
		if (assocparams.sasoc_cookie_life != 0)
			sp->assocparams.sasoc_cookie_life =
						assocparams.sasoc_cookie_life;
	}
	return 0;
}

/*
 * 7.1.16 Set/clear IPv4 mapped addresses (SCTP_I_WANT_MAPPED_V4_ADDR)
 *
 * This socket option is a boolean flag which turns on or off mapped V4
 * addresses.  If this option is turned on and the socket is type
 * PF_INET6, then IPv4 addresses will be mapped to V6 representation.
 * If this option is turned off, then no mapping will be done of V4
 * addresses and a user will receive both PF_INET6 and PF_INET type
 * addresses on the socket.
 */
2945
static int sctp_setsockopt_mappedv4(struct sock *sk, char __user *optval, unsigned int optlen)
L
Linus Torvalds 已提交
2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962
{
	int val;
	struct sctp_sock *sp = sctp_sk(sk);

	if (optlen < sizeof(int))
		return -EINVAL;
	if (get_user(val, (int __user *)optval))
		return -EFAULT;
	if (val)
		sp->v4mapped = 1;
	else
		sp->v4mapped = 0;

	return 0;
}

/*
2963 2964 2965
 * 8.1.16.  Get or Set the Maximum Fragmentation Size (SCTP_MAXSEG)
 * This option will get or set the maximum size to put in any outgoing
 * SCTP DATA chunk.  If a message is larger than this size it will be
L
Linus Torvalds 已提交
2966 2967 2968
 * fragmented by SCTP into the specified size.  Note that the underlying
 * SCTP implementation may fragment into smaller sized chunks when the
 * PMTU of the underlying association is smaller than the value set by
2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987
 * the user.  The default value for this option is '0' which indicates
 * the user is NOT limiting fragmentation and only the PMTU will effect
 * SCTP's choice of DATA chunk size.  Note also that values set larger
 * than the maximum size of an IP datagram will effectively let SCTP
 * control fragmentation (i.e. the same as setting this option to 0).
 *
 * The following structure is used to access and modify this parameter:
 *
 * struct sctp_assoc_value {
 *   sctp_assoc_t assoc_id;
 *   uint32_t assoc_value;
 * };
 *
 * assoc_id:  This parameter is ignored for one-to-one style sockets.
 *    For one-to-many style sockets this parameter indicates which
 *    association the user is performing an action upon.  Note that if
 *    this field's value is zero then the endpoints default value is
 *    changed (effecting future associations only).
 * assoc_value:  This parameter specifies the maximum size in bytes.
L
Linus Torvalds 已提交
2988
 */
2989
static int sctp_setsockopt_maxseg(struct sock *sk, char __user *optval, unsigned int optlen)
L
Linus Torvalds 已提交
2990
{
2991
	struct sctp_assoc_value params;
L
Linus Torvalds 已提交
2992 2993 2994 2995
	struct sctp_association *asoc;
	struct sctp_sock *sp = sctp_sk(sk);
	int val;

2996
	if (optlen == sizeof(int)) {
2997 2998
		pr_warn("Use of int in maxseg socket option deprecated\n");
		pr_warn("Use struct sctp_assoc_value instead\n");
2999 3000 3001 3002 3003 3004 3005 3006
		if (copy_from_user(&val, optval, optlen))
			return -EFAULT;
		params.assoc_id = 0;
	} else if (optlen == sizeof(struct sctp_assoc_value)) {
		if (copy_from_user(&params, optval, optlen))
			return -EFAULT;
		val = params.assoc_value;
	} else
L
Linus Torvalds 已提交
3007
		return -EINVAL;
3008

3009
	if ((val != 0) && ((val < 8) || (val > SCTP_MAX_CHUNK_LEN)))
L
Linus Torvalds 已提交
3010 3011
		return -EINVAL;

3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022
	asoc = sctp_id2assoc(sk, params.assoc_id);
	if (!asoc && params.assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	if (asoc) {
		if (val == 0) {
			val = asoc->pathmtu;
			val -= sp->pf->af->net_header_len;
			val -= sizeof(struct sctphdr) +
					sizeof(struct sctp_data_chunk);
		}
3023 3024
		asoc->user_frag = val;
		asoc->frag_point = sctp_frag_point(asoc, asoc->pathmtu);
3025 3026
	} else {
		sp->user_frag = val;
L
Linus Torvalds 已提交
3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041
	}

	return 0;
}


/*
 *  7.1.9 Set Peer Primary Address (SCTP_SET_PEER_PRIMARY_ADDR)
 *
 *   Requests that the peer mark the enclosed address as the association
 *   primary. The enclosed address must be one of the association's
 *   locally bound addresses. The following structure is used to make a
 *   set primary request:
 */
static int sctp_setsockopt_peer_primary_addr(struct sock *sk, char __user *optval,
3042
					     unsigned int optlen)
L
Linus Torvalds 已提交
3043
{
3044
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
3045 3046 3047 3048
	struct sctp_sock	*sp;
	struct sctp_association	*asoc = NULL;
	struct sctp_setpeerprim	prim;
	struct sctp_chunk	*chunk;
3049
	struct sctp_af		*af;
L
Linus Torvalds 已提交
3050 3051 3052 3053
	int 			err;

	sp = sctp_sk(sk);

3054
	if (!net->sctp.addip_enable)
L
Linus Torvalds 已提交
3055 3056 3057 3058 3059 3060 3061 3062 3063
		return -EPERM;

	if (optlen != sizeof(struct sctp_setpeerprim))
		return -EINVAL;

	if (copy_from_user(&prim, optval, optlen))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, prim.sspp_assoc_id);
3064
	if (!asoc)
L
Linus Torvalds 已提交
3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075
		return -EINVAL;

	if (!asoc->peer.asconf_capable)
		return -EPERM;

	if (asoc->peer.addip_disabled_mask & SCTP_PARAM_SET_PRIMARY)
		return -EPERM;

	if (!sctp_state(asoc, ESTABLISHED))
		return -ENOTCONN;

3076 3077 3078 3079 3080 3081 3082
	af = sctp_get_af_specific(prim.sspp_addr.ss_family);
	if (!af)
		return -EINVAL;

	if (!af->addr_valid((union sctp_addr *)&prim.sspp_addr, sp, NULL))
		return -EADDRNOTAVAIL;

L
Linus Torvalds 已提交
3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093
	if (!sctp_assoc_lookup_laddr(asoc, (union sctp_addr *)&prim.sspp_addr))
		return -EADDRNOTAVAIL;

	/* Create an ASCONF chunk with SET_PRIMARY parameter	*/
	chunk = sctp_make_asconf_set_prim(asoc,
					  (union sctp_addr *)&prim.sspp_addr);
	if (!chunk)
		return -ENOMEM;

	err = sctp_send_asconf(asoc, chunk);

3094
	pr_debug("%s: we set peer primary addr primitively\n", __func__);
L
Linus Torvalds 已提交
3095 3096 3097 3098

	return err;
}

3099
static int sctp_setsockopt_adaptation_layer(struct sock *sk, char __user *optval,
3100
					    unsigned int optlen)
L
Linus Torvalds 已提交
3101
{
3102
	struct sctp_setadaptation adaptation;
L
Linus Torvalds 已提交
3103

3104
	if (optlen != sizeof(struct sctp_setadaptation))
L
Linus Torvalds 已提交
3105
		return -EINVAL;
3106
	if (copy_from_user(&adaptation, optval, optlen))
L
Linus Torvalds 已提交
3107 3108
		return -EFAULT;

3109
	sctp_sk(sk)->adaptation_ind = adaptation.ssb_adaptation_ind;
L
Linus Torvalds 已提交
3110 3111 3112 3113

	return 0;
}

3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128
/*
 * 7.1.29.  Set or Get the default context (SCTP_CONTEXT)
 *
 * The context field in the sctp_sndrcvinfo structure is normally only
 * used when a failed message is retrieved holding the value that was
 * sent down on the actual send call.  This option allows the setting of
 * a default context on an association basis that will be received on
 * reading messages from the peer.  This is especially helpful in the
 * one-2-many model for an application to keep some reference to an
 * internal state machine that is processing messages on the
 * association.  Note that the setting of this value only effects
 * received messages from the peer and does not effect the value that is
 * saved with outbound messages.
 */
static int sctp_setsockopt_context(struct sock *sk, char __user *optval,
3129
				   unsigned int optlen)
3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153
{
	struct sctp_assoc_value params;
	struct sctp_sock *sp;
	struct sctp_association *asoc;

	if (optlen != sizeof(struct sctp_assoc_value))
		return -EINVAL;
	if (copy_from_user(&params, optval, optlen))
		return -EFAULT;

	sp = sctp_sk(sk);

	if (params.assoc_id != 0) {
		asoc = sctp_id2assoc(sk, params.assoc_id);
		if (!asoc)
			return -EINVAL;
		asoc->default_rcv_context = params.assoc_value;
	} else {
		sp->default_rcv_context = params.assoc_value;
	}

	return 0;
}

3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179
/*
 * 7.1.24.  Get or set fragmented interleave (SCTP_FRAGMENT_INTERLEAVE)
 *
 * This options will at a minimum specify if the implementation is doing
 * fragmented interleave.  Fragmented interleave, for a one to many
 * socket, is when subsequent calls to receive a message may return
 * parts of messages from different associations.  Some implementations
 * may allow you to turn this value on or off.  If so, when turned off,
 * no fragment interleave will occur (which will cause a head of line
 * blocking amongst multiple associations sharing the same one to many
 * socket).  When this option is turned on, then each receive call may
 * come from a different association (thus the user must receive data
 * with the extended calls (e.g. sctp_recvmsg) to keep track of which
 * association each receive belongs to.
 *
 * This option takes a boolean value.  A non-zero value indicates that
 * fragmented interleave is on.  A value of zero indicates that
 * fragmented interleave is off.
 *
 * Note that it is important that an implementation that allows this
 * option to be turned on, have it off by default.  Otherwise an unaware
 * application using the one to many model may become confused and act
 * incorrectly.
 */
static int sctp_setsockopt_fragment_interleave(struct sock *sk,
					       char __user *optval,
3180
					       unsigned int optlen)
3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193
{
	int val;

	if (optlen != sizeof(int))
		return -EINVAL;
	if (get_user(val, (int __user *)optval))
		return -EFAULT;

	sctp_sk(sk)->frag_interleave = (val == 0) ? 0 : 1;

	return 0;
}

3194
/*
3195
 * 8.1.21.  Set or Get the SCTP Partial Delivery Point
3196
 *       (SCTP_PARTIAL_DELIVERY_POINT)
3197
 *
3198 3199 3200
 * This option will set or get the SCTP partial delivery point.  This
 * point is the size of a message where the partial delivery API will be
 * invoked to help free up rwnd space for the peer.  Setting this to a
3201
 * lower value will cause partial deliveries to happen more often.  The
3202
 * calls argument is an integer that sets or gets the partial delivery
3203 3204 3205 3206 3207 3208 3209
 * point.  Note also that the call will fail if the user attempts to set
 * this value larger than the socket receive buffer size.
 *
 * Note that any single message having a length smaller than or equal to
 * the SCTP partial delivery point will be delivered in one single read
 * call as long as the user provided buffer is large enough to hold the
 * message.
3210 3211 3212
 */
static int sctp_setsockopt_partial_delivery_point(struct sock *sk,
						  char __user *optval,
3213
						  unsigned int optlen)
3214 3215 3216 3217 3218 3219 3220 3221
{
	u32 val;

	if (optlen != sizeof(u32))
		return -EINVAL;
	if (get_user(val, (int __user *)optval))
		return -EFAULT;

3222 3223 3224 3225 3226 3227
	/* Note: We double the receive buffer from what the user sets
	 * it to be, also initial rwnd is based on rcvbuf/2.
	 */
	if (val > (sk->sk_rcvbuf >> 1))
		return -EINVAL;

3228 3229 3230 3231 3232
	sctp_sk(sk)->pd_point = val;

	return 0; /* is this the right error code? */
}

3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245
/*
 * 7.1.28.  Set or Get the maximum burst (SCTP_MAX_BURST)
 *
 * This option will allow a user to change the maximum burst of packets
 * that can be emitted by this association.  Note that the default value
 * is 4, and some implementations may restrict this setting so that it
 * can only be lowered.
 *
 * NOTE: This text doesn't seem right.  Do this on a socket basis with
 * future associations inheriting the socket value.
 */
static int sctp_setsockopt_maxburst(struct sock *sk,
				    char __user *optval,
3246
				    unsigned int optlen)
3247
{
3248 3249 3250
	struct sctp_assoc_value params;
	struct sctp_sock *sp;
	struct sctp_association *asoc;
3251
	int val;
3252
	int assoc_id = 0;
3253

3254
	if (optlen == sizeof(int)) {
3255 3256
		pr_warn("Use of int in max_burst socket option deprecated\n");
		pr_warn("Use struct sctp_assoc_value instead\n");
3257 3258 3259 3260 3261 3262 3263 3264
		if (copy_from_user(&val, optval, optlen))
			return -EFAULT;
	} else if (optlen == sizeof(struct sctp_assoc_value)) {
		if (copy_from_user(&params, optval, optlen))
			return -EFAULT;
		val = params.assoc_value;
		assoc_id = params.assoc_id;
	} else
3265 3266
		return -EINVAL;

3267 3268 3269 3270 3271 3272 3273 3274 3275
	sp = sctp_sk(sk);

	if (assoc_id != 0) {
		asoc = sctp_id2assoc(sk, assoc_id);
		if (!asoc)
			return -EINVAL;
		asoc->max_burst = val;
	} else
		sp->max_burst = val;
3276 3277 3278 3279

	return 0;
}

3280 3281 3282 3283 3284 3285 3286 3287
/*
 * 7.1.18.  Add a chunk that must be authenticated (SCTP_AUTH_CHUNK)
 *
 * This set option adds a chunk type that the user is requesting to be
 * received only in an authenticated way.  Changes to the list of chunks
 * will only effect future associations on the socket.
 */
static int sctp_setsockopt_auth_chunk(struct sock *sk,
3288 3289
				      char __user *optval,
				      unsigned int optlen)
3290
{
3291
	struct net *net = sock_net(sk);
3292 3293
	struct sctp_authchunk val;

3294
	if (!net->sctp.auth_enable)
3295 3296
		return -EACCES;

3297 3298 3299 3300 3301 3302
	if (optlen != sizeof(struct sctp_authchunk))
		return -EINVAL;
	if (copy_from_user(&val, optval, optlen))
		return -EFAULT;

	switch (val.sauth_chunk) {
J
Joe Perches 已提交
3303 3304 3305 3306 3307
	case SCTP_CID_INIT:
	case SCTP_CID_INIT_ACK:
	case SCTP_CID_SHUTDOWN_COMPLETE:
	case SCTP_CID_AUTH:
		return -EINVAL;
3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320
	}

	/* add this chunk id to the endpoint */
	return sctp_auth_ep_add_chunkid(sctp_sk(sk)->ep, val.sauth_chunk);
}

/*
 * 7.1.19.  Get or set the list of supported HMAC Identifiers (SCTP_HMAC_IDENT)
 *
 * This option gets or sets the list of HMAC algorithms that the local
 * endpoint requires the peer to use.
 */
static int sctp_setsockopt_hmac_ident(struct sock *sk,
3321 3322
				      char __user *optval,
				      unsigned int optlen)
3323
{
3324
	struct net *net = sock_net(sk);
3325
	struct sctp_hmacalgo *hmacs;
3326
	u32 idents;
3327 3328
	int err;

3329
	if (!net->sctp.auth_enable)
3330 3331
		return -EACCES;

3332 3333 3334
	if (optlen < sizeof(struct sctp_hmacalgo))
		return -EINVAL;

3335
	hmacs = memdup_user(optval, optlen);
3336 3337
	if (IS_ERR(hmacs))
		return PTR_ERR(hmacs);
3338

3339 3340 3341
	idents = hmacs->shmac_num_idents;
	if (idents == 0 || idents > SCTP_AUTH_NUM_HMACS ||
	    (idents * sizeof(u16)) > (optlen - sizeof(struct sctp_hmacalgo))) {
3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359
		err = -EINVAL;
		goto out;
	}

	err = sctp_auth_ep_set_hmacs(sctp_sk(sk)->ep, hmacs);
out:
	kfree(hmacs);
	return err;
}

/*
 * 7.1.20.  Set a shared key (SCTP_AUTH_KEY)
 *
 * This option will set a shared secret key which is used to build an
 * association shared key.
 */
static int sctp_setsockopt_auth_key(struct sock *sk,
				    char __user *optval,
3360
				    unsigned int optlen)
3361
{
3362
	struct net *net = sock_net(sk);
3363 3364 3365 3366
	struct sctp_authkey *authkey;
	struct sctp_association *asoc;
	int ret;

3367
	if (!net->sctp.auth_enable)
3368 3369
		return -EACCES;

3370 3371 3372
	if (optlen <= sizeof(struct sctp_authkey))
		return -EINVAL;

3373
	authkey = memdup_user(optval, optlen);
3374 3375
	if (IS_ERR(authkey))
		return PTR_ERR(authkey);
3376

3377
	if (authkey->sca_keylength > optlen - sizeof(struct sctp_authkey)) {
3378 3379 3380 3381
		ret = -EINVAL;
		goto out;
	}

3382 3383 3384 3385 3386 3387 3388 3389
	asoc = sctp_id2assoc(sk, authkey->sca_assoc_id);
	if (!asoc && authkey->sca_assoc_id && sctp_style(sk, UDP)) {
		ret = -EINVAL;
		goto out;
	}

	ret = sctp_auth_set_key(sctp_sk(sk)->ep, asoc, authkey);
out:
3390
	kzfree(authkey);
3391 3392 3393 3394 3395 3396 3397 3398 3399 3400
	return ret;
}

/*
 * 7.1.21.  Get or set the active shared key (SCTP_AUTH_ACTIVE_KEY)
 *
 * This option will get or set the active shared key to be used to build
 * the association shared key.
 */
static int sctp_setsockopt_active_key(struct sock *sk,
3401 3402
				      char __user *optval,
				      unsigned int optlen)
3403
{
3404
	struct net *net = sock_net(sk);
3405 3406 3407
	struct sctp_authkeyid val;
	struct sctp_association *asoc;

3408
	if (!net->sctp.auth_enable)
3409 3410
		return -EACCES;

3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429
	if (optlen != sizeof(struct sctp_authkeyid))
		return -EINVAL;
	if (copy_from_user(&val, optval, optlen))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, val.scact_assoc_id);
	if (!asoc && val.scact_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	return sctp_auth_set_active_key(sctp_sk(sk)->ep, asoc,
					val.scact_keynumber);
}

/*
 * 7.1.22.  Delete a shared key (SCTP_AUTH_DELETE_KEY)
 *
 * This set option will delete a shared secret key from use.
 */
static int sctp_setsockopt_del_key(struct sock *sk,
3430 3431
				   char __user *optval,
				   unsigned int optlen)
3432
{
3433
	struct net *net = sock_net(sk);
3434 3435 3436
	struct sctp_authkeyid val;
	struct sctp_association *asoc;

3437
	if (!net->sctp.auth_enable)
3438 3439
		return -EACCES;

3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453
	if (optlen != sizeof(struct sctp_authkeyid))
		return -EINVAL;
	if (copy_from_user(&val, optval, optlen))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, val.scact_assoc_id);
	if (!asoc && val.scact_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	return sctp_auth_del_key_id(sctp_sk(sk)->ep, asoc,
				    val.scact_keynumber);

}

3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487
/*
 * 8.1.23 SCTP_AUTO_ASCONF
 *
 * This option will enable or disable the use of the automatic generation of
 * ASCONF chunks to add and delete addresses to an existing association.  Note
 * that this option has two caveats namely: a) it only affects sockets that
 * are bound to all addresses available to the SCTP stack, and b) the system
 * administrator may have an overriding control that turns the ASCONF feature
 * off no matter what setting the socket option may have.
 * This option expects an integer boolean flag, where a non-zero value turns on
 * the option, and a zero value turns off the option.
 * Note. In this implementation, socket operation overrides default parameter
 * being set by sysctl as well as FreeBSD implementation
 */
static int sctp_setsockopt_auto_asconf(struct sock *sk, char __user *optval,
					unsigned int optlen)
{
	int val;
	struct sctp_sock *sp = sctp_sk(sk);

	if (optlen < sizeof(int))
		return -EINVAL;
	if (get_user(val, (int __user *)optval))
		return -EFAULT;
	if (!sctp_is_ep_boundall(sk) && val)
		return -EINVAL;
	if ((val && sp->do_auto_asconf) || (!val && !sp->do_auto_asconf))
		return 0;

	if (val == 0 && sp->do_auto_asconf) {
		list_del(&sp->auto_asconf_list);
		sp->do_auto_asconf = 0;
	} else if (val && !sp->do_auto_asconf) {
		list_add_tail(&sp->auto_asconf_list,
3488
		    &sock_net(sk)->sctp.auto_asconf_splist);
3489 3490 3491 3492 3493
		sp->do_auto_asconf = 1;
	}
	return 0;
}

3494

3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544
/*
 * SCTP_PEER_ADDR_THLDS
 *
 * This option allows us to alter the partially failed threshold for one or all
 * transports in an association.  See Section 6.1 of:
 * http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
 */
static int sctp_setsockopt_paddr_thresholds(struct sock *sk,
					    char __user *optval,
					    unsigned int optlen)
{
	struct sctp_paddrthlds val;
	struct sctp_transport *trans;
	struct sctp_association *asoc;

	if (optlen < sizeof(struct sctp_paddrthlds))
		return -EINVAL;
	if (copy_from_user(&val, (struct sctp_paddrthlds __user *)optval,
			   sizeof(struct sctp_paddrthlds)))
		return -EFAULT;


	if (sctp_is_any(sk, (const union sctp_addr *)&val.spt_address)) {
		asoc = sctp_id2assoc(sk, val.spt_assoc_id);
		if (!asoc)
			return -ENOENT;
		list_for_each_entry(trans, &asoc->peer.transport_addr_list,
				    transports) {
			if (val.spt_pathmaxrxt)
				trans->pathmaxrxt = val.spt_pathmaxrxt;
			trans->pf_retrans = val.spt_pathpfthld;
		}

		if (val.spt_pathmaxrxt)
			asoc->pathmaxrxt = val.spt_pathmaxrxt;
		asoc->pf_retrans = val.spt_pathpfthld;
	} else {
		trans = sctp_addr_id2transport(sk, &val.spt_address,
					       val.spt_assoc_id);
		if (!trans)
			return -ENOENT;

		if (val.spt_pathmaxrxt)
			trans->pathmaxrxt = val.spt_pathmaxrxt;
		trans->pf_retrans = val.spt_pathpfthld;
	}

	return 0;
}

L
Linus Torvalds 已提交
3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563
/* API 6.2 setsockopt(), getsockopt()
 *
 * Applications use setsockopt() and getsockopt() to set or retrieve
 * socket options.  Socket options are used to change the default
 * behavior of sockets calls.  They are described in Section 7.
 *
 * The syntax is:
 *
 *   ret = getsockopt(int sd, int level, int optname, void __user *optval,
 *                    int __user *optlen);
 *   ret = setsockopt(int sd, int level, int optname, const void __user *optval,
 *                    int optlen);
 *
 *   sd      - the socket descript.
 *   level   - set to IPPROTO_SCTP for all SCTP options.
 *   optname - the option name.
 *   optval  - the buffer to store the value of the option.
 *   optlen  - the size of the buffer.
 */
3564 3565
static int sctp_setsockopt(struct sock *sk, int level, int optname,
			   char __user *optval, unsigned int optlen)
L
Linus Torvalds 已提交
3566 3567 3568
{
	int retval = 0;

3569
	pr_debug("%s: sk:%p, optname:%d\n", __func__, sk, optname);
L
Linus Torvalds 已提交
3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597

	/* I can hardly begin to describe how wrong this is.  This is
	 * so broken as to be worse than useless.  The API draft
	 * REALLY is NOT helpful here...  I am not convinced that the
	 * semantics of setsockopt() with a level OTHER THAN SOL_SCTP
	 * are at all well-founded.
	 */
	if (level != SOL_SCTP) {
		struct sctp_af *af = sctp_sk(sk)->pf->af;
		retval = af->setsockopt(sk, level, optname, optval, optlen);
		goto out_nounlock;
	}

	sctp_lock_sock(sk);

	switch (optname) {
	case SCTP_SOCKOPT_BINDX_ADD:
		/* 'optlen' is the size of the addresses buffer. */
		retval = sctp_setsockopt_bindx(sk, (struct sockaddr __user *)optval,
					       optlen, SCTP_BINDX_ADD_ADDR);
		break;

	case SCTP_SOCKOPT_BINDX_REM:
		/* 'optlen' is the size of the addresses buffer. */
		retval = sctp_setsockopt_bindx(sk, (struct sockaddr __user *)optval,
					       optlen, SCTP_BINDX_REM_ADDR);
		break;

3598 3599 3600 3601 3602 3603 3604
	case SCTP_SOCKOPT_CONNECTX_OLD:
		/* 'optlen' is the size of the addresses buffer. */
		retval = sctp_setsockopt_connectx_old(sk,
					    (struct sockaddr __user *)optval,
					    optlen);
		break;

F
Frank Filz 已提交
3605 3606
	case SCTP_SOCKOPT_CONNECTX:
		/* 'optlen' is the size of the addresses buffer. */
3607 3608 3609
		retval = sctp_setsockopt_connectx(sk,
					    (struct sockaddr __user *)optval,
					    optlen);
F
Frank Filz 已提交
3610 3611
		break;

L
Linus Torvalds 已提交
3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627
	case SCTP_DISABLE_FRAGMENTS:
		retval = sctp_setsockopt_disable_fragments(sk, optval, optlen);
		break;

	case SCTP_EVENTS:
		retval = sctp_setsockopt_events(sk, optval, optlen);
		break;

	case SCTP_AUTOCLOSE:
		retval = sctp_setsockopt_autoclose(sk, optval, optlen);
		break;

	case SCTP_PEER_ADDR_PARAMS:
		retval = sctp_setsockopt_peer_addr_params(sk, optval, optlen);
		break;

3628
	case SCTP_DELAYED_SACK:
3629
		retval = sctp_setsockopt_delayed_ack(sk, optval, optlen);
3630
		break;
3631 3632 3633
	case SCTP_PARTIAL_DELIVERY_POINT:
		retval = sctp_setsockopt_partial_delivery_point(sk, optval, optlen);
		break;
3634

L
Linus Torvalds 已提交
3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662
	case SCTP_INITMSG:
		retval = sctp_setsockopt_initmsg(sk, optval, optlen);
		break;
	case SCTP_DEFAULT_SEND_PARAM:
		retval = sctp_setsockopt_default_send_param(sk, optval,
							    optlen);
		break;
	case SCTP_PRIMARY_ADDR:
		retval = sctp_setsockopt_primary_addr(sk, optval, optlen);
		break;
	case SCTP_SET_PEER_PRIMARY_ADDR:
		retval = sctp_setsockopt_peer_primary_addr(sk, optval, optlen);
		break;
	case SCTP_NODELAY:
		retval = sctp_setsockopt_nodelay(sk, optval, optlen);
		break;
	case SCTP_RTOINFO:
		retval = sctp_setsockopt_rtoinfo(sk, optval, optlen);
		break;
	case SCTP_ASSOCINFO:
		retval = sctp_setsockopt_associnfo(sk, optval, optlen);
		break;
	case SCTP_I_WANT_MAPPED_V4_ADDR:
		retval = sctp_setsockopt_mappedv4(sk, optval, optlen);
		break;
	case SCTP_MAXSEG:
		retval = sctp_setsockopt_maxseg(sk, optval, optlen);
		break;
3663 3664
	case SCTP_ADAPTATION_LAYER:
		retval = sctp_setsockopt_adaptation_layer(sk, optval, optlen);
L
Linus Torvalds 已提交
3665
		break;
3666 3667 3668
	case SCTP_CONTEXT:
		retval = sctp_setsockopt_context(sk, optval, optlen);
		break;
3669 3670 3671
	case SCTP_FRAGMENT_INTERLEAVE:
		retval = sctp_setsockopt_fragment_interleave(sk, optval, optlen);
		break;
3672 3673 3674
	case SCTP_MAX_BURST:
		retval = sctp_setsockopt_maxburst(sk, optval, optlen);
		break;
3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689
	case SCTP_AUTH_CHUNK:
		retval = sctp_setsockopt_auth_chunk(sk, optval, optlen);
		break;
	case SCTP_HMAC_IDENT:
		retval = sctp_setsockopt_hmac_ident(sk, optval, optlen);
		break;
	case SCTP_AUTH_KEY:
		retval = sctp_setsockopt_auth_key(sk, optval, optlen);
		break;
	case SCTP_AUTH_ACTIVE_KEY:
		retval = sctp_setsockopt_active_key(sk, optval, optlen);
		break;
	case SCTP_AUTH_DELETE_KEY:
		retval = sctp_setsockopt_del_key(sk, optval, optlen);
		break;
3690 3691 3692
	case SCTP_AUTO_ASCONF:
		retval = sctp_setsockopt_auto_asconf(sk, optval, optlen);
		break;
3693 3694 3695
	case SCTP_PEER_ADDR_THLDS:
		retval = sctp_setsockopt_paddr_thresholds(sk, optval, optlen);
		break;
L
Linus Torvalds 已提交
3696 3697 3698
	default:
		retval = -ENOPROTOOPT;
		break;
3699
	}
L
Linus Torvalds 已提交
3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722

	sctp_release_sock(sk);

out_nounlock:
	return retval;
}

/* API 3.1.6 connect() - UDP Style Syntax
 *
 * An application may use the connect() call in the UDP model to initiate an
 * association without sending data.
 *
 * The syntax is:
 *
 * ret = connect(int sd, const struct sockaddr *nam, socklen_t len);
 *
 * sd: the socket descriptor to have a new association added to.
 *
 * nam: the address structure (either struct sockaddr_in or struct
 *    sockaddr_in6 defined in RFC2553 [7]).
 *
 * len: the size of the address.
 */
3723 3724
static int sctp_connect(struct sock *sk, struct sockaddr *addr,
			int addr_len)
L
Linus Torvalds 已提交
3725 3726
{
	int err = 0;
F
Frank Filz 已提交
3727
	struct sctp_af *af;
L
Linus Torvalds 已提交
3728 3729 3730

	sctp_lock_sock(sk);

3731 3732
	pr_debug("%s: sk:%p, sockaddr:%p, addr_len:%d\n", __func__, sk,
		 addr, addr_len);
L
Linus Torvalds 已提交
3733

F
Frank Filz 已提交
3734 3735 3736 3737 3738 3739 3740 3741
	/* Validate addr_len before calling common connect/connectx routine. */
	af = sctp_get_af_specific(addr->sa_family);
	if (!af || addr_len < af->sockaddr_len) {
		err = -EINVAL;
	} else {
		/* Pass correct addr len to common routine (so it knows there
		 * is only one address being passed.
		 */
3742
		err = __sctp_connect(sk, addr, af->sockaddr_len, NULL);
L
Linus Torvalds 已提交
3743 3744 3745 3746 3747 3748 3749
	}

	sctp_release_sock(sk);
	return err;
}

/* FIXME: Write comments. */
3750
static int sctp_disconnect(struct sock *sk, int flags)
L
Linus Torvalds 已提交
3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761
{
	return -EOPNOTSUPP; /* STUB */
}

/* 4.1.4 accept() - TCP Style Syntax
 *
 * Applications use accept() call to remove an established SCTP
 * association from the accept queue of the endpoint.  A new socket
 * descriptor will be returned from accept() to represent the newly
 * formed association.
 */
3762
static struct sock *sctp_accept(struct sock *sk, int flags, int *err)
L
Linus Torvalds 已提交
3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785
{
	struct sctp_sock *sp;
	struct sctp_endpoint *ep;
	struct sock *newsk = NULL;
	struct sctp_association *asoc;
	long timeo;
	int error = 0;

	sctp_lock_sock(sk);

	sp = sctp_sk(sk);
	ep = sp->ep;

	if (!sctp_style(sk, TCP)) {
		error = -EOPNOTSUPP;
		goto out;
	}

	if (!sctp_sstate(sk, LISTENING)) {
		error = -EINVAL;
		goto out;
	}

3786
	timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK);
L
Linus Torvalds 已提交
3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809

	error = sctp_wait_for_accept(sk, timeo);
	if (error)
		goto out;

	/* We treat the list of associations on the endpoint as the accept
	 * queue and pick the first association on the list.
	 */
	asoc = list_entry(ep->asocs.next, struct sctp_association, asocs);

	newsk = sp->pf->create_accept_sk(sk, asoc);
	if (!newsk) {
		error = -ENOMEM;
		goto out;
	}

	/* Populate the fields of the newsk from the oldsk and migrate the
	 * asoc to the newsk.
	 */
	sctp_sock_migrate(sk, newsk, asoc, SCTP_SOCKET_TCP);

out:
	sctp_release_sock(sk);
3810
	*err = error;
L
Linus Torvalds 已提交
3811 3812 3813 3814
	return newsk;
}

/* The SCTP ioctl handler. */
3815
static int sctp_ioctl(struct sock *sk, int cmd, unsigned long arg)
L
Linus Torvalds 已提交
3816
{
3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842
	int rc = -ENOTCONN;

	sctp_lock_sock(sk);

	/*
	 * SEQPACKET-style sockets in LISTENING state are valid, for
	 * SCTP, so only discard TCP-style sockets in LISTENING state.
	 */
	if (sctp_style(sk, TCP) && sctp_sstate(sk, LISTENING))
		goto out;

	switch (cmd) {
	case SIOCINQ: {
		struct sk_buff *skb;
		unsigned int amount = 0;

		skb = skb_peek(&sk->sk_receive_queue);
		if (skb != NULL) {
			/*
			 * We will only return the amount of this packet since
			 * that is all that will be read.
			 */
			amount = skb->len;
		}
		rc = put_user(amount, (int __user *)arg);
		break;
3843
	}
3844 3845 3846 3847 3848 3849 3850
	default:
		rc = -ENOIOCTLCMD;
		break;
	}
out:
	sctp_release_sock(sk);
	return rc;
L
Linus Torvalds 已提交
3851 3852 3853 3854 3855 3856
}

/* This is the function which gets called during socket creation to
 * initialized the SCTP-specific portion of the sock.
 * The sock structure should already be zero-filled memory.
 */
3857
static int sctp_init_sock(struct sock *sk)
L
Linus Torvalds 已提交
3858
{
3859
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
3860 3861
	struct sctp_sock *sp;

3862
	pr_debug("%s: sk:%p\n", __func__, sk);
L
Linus Torvalds 已提交
3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886

	sp = sctp_sk(sk);

	/* Initialize the SCTP per socket area.  */
	switch (sk->sk_type) {
	case SOCK_SEQPACKET:
		sp->type = SCTP_SOCKET_UDP;
		break;
	case SOCK_STREAM:
		sp->type = SCTP_SOCKET_TCP;
		break;
	default:
		return -ESOCKTNOSUPPORT;
	}

	/* Initialize default send parameters. These parameters can be
	 * modified with the SCTP_DEFAULT_SEND_PARAM socket option.
	 */
	sp->default_stream = 0;
	sp->default_ppid = 0;
	sp->default_flags = 0;
	sp->default_context = 0;
	sp->default_timetolive = 0;

3887
	sp->default_rcv_context = 0;
3888
	sp->max_burst = net->sctp.max_burst;
3889

3890 3891
	sp->sctp_hmac_alg = net->sctp.sctp_hmac_alg;

L
Linus Torvalds 已提交
3892 3893 3894 3895 3896 3897
	/* Initialize default setup parameters. These parameters
	 * can be modified with the SCTP_INITMSG socket option or
	 * overridden by the SCTP_INIT CMSG.
	 */
	sp->initmsg.sinit_num_ostreams   = sctp_max_outstreams;
	sp->initmsg.sinit_max_instreams  = sctp_max_instreams;
3898 3899
	sp->initmsg.sinit_max_attempts   = net->sctp.max_retrans_init;
	sp->initmsg.sinit_max_init_timeo = net->sctp.rto_max;
L
Linus Torvalds 已提交
3900 3901 3902 3903

	/* Initialize default RTO related parameters.  These parameters can
	 * be modified for with the SCTP_RTOINFO socket option.
	 */
3904 3905 3906
	sp->rtoinfo.srto_initial = net->sctp.rto_initial;
	sp->rtoinfo.srto_max     = net->sctp.rto_max;
	sp->rtoinfo.srto_min     = net->sctp.rto_min;
L
Linus Torvalds 已提交
3907 3908 3909 3910

	/* Initialize default association related parameters. These parameters
	 * can be modified with the SCTP_ASSOCINFO socket option.
	 */
3911
	sp->assocparams.sasoc_asocmaxrxt = net->sctp.max_retrans_association;
L
Linus Torvalds 已提交
3912 3913 3914
	sp->assocparams.sasoc_number_peer_destinations = 0;
	sp->assocparams.sasoc_peer_rwnd = 0;
	sp->assocparams.sasoc_local_rwnd = 0;
3915
	sp->assocparams.sasoc_cookie_life = net->sctp.valid_cookie_life;
L
Linus Torvalds 已提交
3916 3917

	/* Initialize default event subscriptions. By default, all the
3918
	 * options are off.
L
Linus Torvalds 已提交
3919 3920 3921 3922 3923 3924
	 */
	memset(&sp->subscribe, 0, sizeof(struct sctp_event_subscribe));

	/* Default Peer Address Parameters.  These defaults can
	 * be modified via SCTP_PEER_ADDR_PARAMS
	 */
3925 3926
	sp->hbinterval  = net->sctp.hb_interval;
	sp->pathmaxrxt  = net->sctp.max_retrans_path;
3927
	sp->pathmtu     = 0; // allow default discovery
3928
	sp->sackdelay   = net->sctp.sack_timeout;
3929
	sp->sackfreq	= 2;
3930
	sp->param_flags = SPP_HB_ENABLE |
3931 3932
			  SPP_PMTUD_ENABLE |
			  SPP_SACKDELAY_ENABLE;
L
Linus Torvalds 已提交
3933 3934 3935 3936 3937 3938

	/* If enabled no SCTP message fragmentation will be performed.
	 * Configure through SCTP_DISABLE_FRAGMENTS socket option.
	 */
	sp->disable_fragments = 0;

3939 3940
	/* Enable Nagle algorithm by default.  */
	sp->nodelay           = 0;
L
Linus Torvalds 已提交
3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954

	/* Enable by default. */
	sp->v4mapped          = 1;

	/* Auto-close idle associations after the configured
	 * number of seconds.  A value of 0 disables this
	 * feature.  Configure through the SCTP_AUTOCLOSE socket option,
	 * for UDP-style sockets only.
	 */
	sp->autoclose         = 0;

	/* User specified fragmentation limit. */
	sp->user_frag         = 0;

3955
	sp->adaptation_ind = 0;
L
Linus Torvalds 已提交
3956 3957 3958 3959

	sp->pf = sctp_get_pf_specific(sk->sk_family);

	/* Control variables for partial data delivery. */
3960
	atomic_set(&sp->pd_mode, 0);
L
Linus Torvalds 已提交
3961
	skb_queue_head_init(&sp->pd_lobby);
3962
	sp->frag_interleave = 0;
L
Linus Torvalds 已提交
3963 3964 3965 3966 3967

	/* Create a per socket endpoint structure.  Even if we
	 * change the data structure relationships, this may still
	 * be useful for storing pre-connect address information.
	 */
3968 3969
	sp->ep = sctp_endpoint_new(sk, GFP_KERNEL);
	if (!sp->ep)
L
Linus Torvalds 已提交
3970 3971 3972 3973
		return -ENOMEM;

	sp->hmac = NULL;

3974 3975
	sk->sk_destruct = sctp_destruct_sock;

L
Linus Torvalds 已提交
3976
	SCTP_DBG_OBJCNT_INC(sock);
3977 3978

	local_bh_disable();
3979
	percpu_counter_inc(&sctp_sockets_allocated);
3980 3981
	sock_prot_inuse_add(net, sk->sk_prot, 1);
	if (net->sctp.default_auto_asconf) {
3982
		list_add_tail(&sp->auto_asconf_list,
3983
		    &net->sctp.auto_asconf_splist);
3984 3985 3986
		sp->do_auto_asconf = 1;
	} else
		sp->do_auto_asconf = 0;
3987 3988
	local_bh_enable();

L
Linus Torvalds 已提交
3989 3990 3991 3992
	return 0;
}

/* Cleanup any SCTP per socket resources.  */
3993
static void sctp_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
3994
{
3995
	struct sctp_sock *sp;
L
Linus Torvalds 已提交
3996

3997
	pr_debug("%s: sk:%p\n", __func__, sk);
L
Linus Torvalds 已提交
3998 3999

	/* Release our hold on the endpoint. */
4000
	sp = sctp_sk(sk);
4001 4002 4003 4004 4005 4006
	/* This could happen during socket init, thus we bail out
	 * early, since the rest of the below is not setup either.
	 */
	if (sp->ep == NULL)
		return;

4007 4008 4009 4010 4011
	if (sp->do_auto_asconf) {
		sp->do_auto_asconf = 0;
		list_del(&sp->auto_asconf_list);
	}
	sctp_endpoint_free(sp->ep);
4012
	local_bh_disable();
4013
	percpu_counter_dec(&sctp_sockets_allocated);
4014
	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
4015
	local_bh_enable();
L
Linus Torvalds 已提交
4016 4017
}

4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028
/* Triggered when there are no references on the socket anymore */
static void sctp_destruct_sock(struct sock *sk)
{
	struct sctp_sock *sp = sctp_sk(sk);

	/* Free up the HMAC transform. */
	crypto_free_hash(sp->hmac);

	inet_sock_destruct(sk);
}

L
Linus Torvalds 已提交
4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044
/* API 4.1.7 shutdown() - TCP Style Syntax
 *     int shutdown(int socket, int how);
 *
 *     sd      - the socket descriptor of the association to be closed.
 *     how     - Specifies the type of shutdown.  The  values  are
 *               as follows:
 *               SHUT_RD
 *                     Disables further receive operations. No SCTP
 *                     protocol action is taken.
 *               SHUT_WR
 *                     Disables further send operations, and initiates
 *                     the SCTP shutdown sequence.
 *               SHUT_RDWR
 *                     Disables further send  and  receive  operations
 *                     and initiates the SCTP shutdown sequence.
 */
4045
static void sctp_shutdown(struct sock *sk, int how)
L
Linus Torvalds 已提交
4046
{
4047
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058
	struct sctp_endpoint *ep;
	struct sctp_association *asoc;

	if (!sctp_style(sk, TCP))
		return;

	if (how & SEND_SHUTDOWN) {
		ep = sctp_sk(sk)->ep;
		if (!list_empty(&ep->asocs)) {
			asoc = list_entry(ep->asocs.next,
					  struct sctp_association, asocs);
4059
			sctp_primitive_SHUTDOWN(net, asoc, NULL);
L
Linus Torvalds 已提交
4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080
		}
	}
}

/* 7.2.1 Association Status (SCTP_STATUS)

 * Applications can retrieve current status information about an
 * association, including association state, peer receiver window size,
 * number of unacked data chunks, and number of data chunks pending
 * receipt.  This information is read-only.
 */
static int sctp_getsockopt_sctp_status(struct sock *sk, int len,
				       char __user *optval,
				       int __user *optlen)
{
	struct sctp_status status;
	struct sctp_association *asoc = NULL;
	struct sctp_transport *transport;
	sctp_assoc_t associd;
	int retval = 0;

4081
	if (len < sizeof(status)) {
L
Linus Torvalds 已提交
4082 4083 4084 4085
		retval = -EINVAL;
		goto out;
	}

4086 4087
	len = sizeof(status);
	if (copy_from_user(&status, optval, len)) {
L
Linus Torvalds 已提交
4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110
		retval = -EFAULT;
		goto out;
	}

	associd = status.sstat_assoc_id;
	asoc = sctp_id2assoc(sk, associd);
	if (!asoc) {
		retval = -EINVAL;
		goto out;
	}

	transport = asoc->peer.primary_path;

	status.sstat_assoc_id = sctp_assoc2id(asoc);
	status.sstat_state = asoc->state;
	status.sstat_rwnd =  asoc->peer.rwnd;
	status.sstat_unackdata = asoc->unack_data;

	status.sstat_penddata = sctp_tsnmap_pending(&asoc->peer.tsn_map);
	status.sstat_instrms = asoc->c.sinit_max_instreams;
	status.sstat_outstrms = asoc->c.sinit_num_ostreams;
	status.sstat_fragmentation_point = asoc->frag_point;
	status.sstat_primary.spinfo_assoc_id = sctp_assoc2id(transport->asoc);
4111 4112
	memcpy(&status.sstat_primary.spinfo_address, &transport->ipaddr,
			transport->af_specific->sockaddr_len);
L
Linus Torvalds 已提交
4113 4114 4115
	/* Map ipv4 address into v4-mapped-on-v6 address.  */
	sctp_get_pf_specific(sk->sk_family)->addr_v4map(sctp_sk(sk),
		(union sctp_addr *)&status.sstat_primary.spinfo_address);
F
Frank Filz 已提交
4116
	status.sstat_primary.spinfo_state = transport->state;
L
Linus Torvalds 已提交
4117 4118 4119
	status.sstat_primary.spinfo_cwnd = transport->cwnd;
	status.sstat_primary.spinfo_srtt = transport->srtt;
	status.sstat_primary.spinfo_rto = jiffies_to_msecs(transport->rto);
4120
	status.sstat_primary.spinfo_mtu = transport->pathmtu;
L
Linus Torvalds 已提交
4121

F
Frank Filz 已提交
4122 4123 4124
	if (status.sstat_primary.spinfo_state == SCTP_UNKNOWN)
		status.sstat_primary.spinfo_state = SCTP_ACTIVE;

L
Linus Torvalds 已提交
4125 4126 4127 4128 4129
	if (put_user(len, optlen)) {
		retval = -EFAULT;
		goto out;
	}

4130 4131 4132
	pr_debug("%s: len:%d, state:%d, rwnd:%d, assoc_id:%d\n",
		 __func__, len, status.sstat_state, status.sstat_rwnd,
		 status.sstat_assoc_id);
L
Linus Torvalds 已提交
4133 4134 4135 4136 4137 4138 4139

	if (copy_to_user(optval, &status, len)) {
		retval = -EFAULT;
		goto out;
	}

out:
E
Eric Dumazet 已提交
4140
	return retval;
L
Linus Torvalds 已提交
4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158
}


/* 7.2.2 Peer Address Information (SCTP_GET_PEER_ADDR_INFO)
 *
 * Applications can retrieve information about a specific peer address
 * of an association, including its reachability state, congestion
 * window, and retransmission timer values.  This information is
 * read-only.
 */
static int sctp_getsockopt_peer_addr_info(struct sock *sk, int len,
					  char __user *optval,
					  int __user *optlen)
{
	struct sctp_paddrinfo pinfo;
	struct sctp_transport *transport;
	int retval = 0;

4159
	if (len < sizeof(pinfo)) {
L
Linus Torvalds 已提交
4160 4161 4162 4163
		retval = -EINVAL;
		goto out;
	}

4164 4165
	len = sizeof(pinfo);
	if (copy_from_user(&pinfo, optval, len)) {
L
Linus Torvalds 已提交
4166 4167 4168 4169 4170 4171 4172 4173 4174 4175
		retval = -EFAULT;
		goto out;
	}

	transport = sctp_addr_id2transport(sk, &pinfo.spinfo_address,
					   pinfo.spinfo_assoc_id);
	if (!transport)
		return -EINVAL;

	pinfo.spinfo_assoc_id = sctp_assoc2id(transport->asoc);
F
Frank Filz 已提交
4176
	pinfo.spinfo_state = transport->state;
L
Linus Torvalds 已提交
4177 4178 4179
	pinfo.spinfo_cwnd = transport->cwnd;
	pinfo.spinfo_srtt = transport->srtt;
	pinfo.spinfo_rto = jiffies_to_msecs(transport->rto);
4180
	pinfo.spinfo_mtu = transport->pathmtu;
L
Linus Torvalds 已提交
4181

F
Frank Filz 已提交
4182 4183 4184
	if (pinfo.spinfo_state == SCTP_UNKNOWN)
		pinfo.spinfo_state = SCTP_ACTIVE;

L
Linus Torvalds 已提交
4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195
	if (put_user(len, optlen)) {
		retval = -EFAULT;
		goto out;
	}

	if (copy_to_user(optval, &pinfo, len)) {
		retval = -EFAULT;
		goto out;
	}

out:
E
Eric Dumazet 已提交
4196
	return retval;
L
Linus Torvalds 已提交
4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230
}

/* 7.1.12 Enable/Disable message fragmentation (SCTP_DISABLE_FRAGMENTS)
 *
 * This option is a on/off flag.  If enabled no SCTP message
 * fragmentation will be performed.  Instead if a message being sent
 * exceeds the current PMTU size, the message will NOT be sent and
 * instead a error will be indicated to the user.
 */
static int sctp_getsockopt_disable_fragments(struct sock *sk, int len,
					char __user *optval, int __user *optlen)
{
	int val;

	if (len < sizeof(int))
		return -EINVAL;

	len = sizeof(int);
	val = (sctp_sk(sk)->disable_fragments == 1);
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;
	return 0;
}

/* 7.1.15 Set notification and ancillary events (SCTP_EVENTS)
 *
 * This socket option is used to specify various notifications and
 * ancillary data the user wishes to receive.
 */
static int sctp_getsockopt_events(struct sock *sk, int len, char __user *optval,
				  int __user *optlen)
{
4231
	if (len <= 0)
L
Linus Torvalds 已提交
4232
		return -EINVAL;
4233 4234
	if (len > sizeof(struct sctp_event_subscribe))
		len = sizeof(struct sctp_event_subscribe);
4235 4236
	if (put_user(len, optlen))
		return -EFAULT;
L
Linus Torvalds 已提交
4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257
	if (copy_to_user(optval, &sctp_sk(sk)->subscribe, len))
		return -EFAULT;
	return 0;
}

/* 7.1.8 Automatic Close of associations (SCTP_AUTOCLOSE)
 *
 * This socket option is applicable to the UDP-style socket only.  When
 * set it will cause associations that are idle for more than the
 * specified number of seconds to automatically close.  An association
 * being idle is defined an association that has NOT sent or received
 * user data.  The special value of '0' indicates that no automatic
 * close of any associations should be performed.  The option expects an
 * integer defining the number of seconds of idle time before an
 * association is closed.
 */
static int sctp_getsockopt_autoclose(struct sock *sk, int len, char __user *optval, int __user *optlen)
{
	/* Applicable to UDP-style socket only */
	if (sctp_style(sk, TCP))
		return -EOPNOTSUPP;
4258
	if (len < sizeof(int))
L
Linus Torvalds 已提交
4259
		return -EINVAL;
4260 4261 4262 4263
	len = sizeof(int);
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &sctp_sk(sk)->autoclose, sizeof(int)))
L
Linus Torvalds 已提交
4264 4265 4266 4267 4268
		return -EFAULT;
	return 0;
}

/* Helper routine to branch off an association to a new socket.  */
B
Benjamin Poirier 已提交
4269
int sctp_do_peeloff(struct sock *sk, sctp_assoc_t id, struct socket **sockp)
L
Linus Torvalds 已提交
4270
{
B
Benjamin Poirier 已提交
4271
	struct sctp_association *asoc = sctp_id2assoc(sk, id);
L
Linus Torvalds 已提交
4272
	struct socket *sock;
4273
	struct sctp_af *af;
L
Linus Torvalds 已提交
4274 4275
	int err = 0;

B
Benjamin Poirier 已提交
4276 4277 4278
	if (!asoc)
		return -EINVAL;

L
Linus Torvalds 已提交
4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289
	/* An association cannot be branched off from an already peeled-off
	 * socket, nor is this supported for tcp style sockets.
	 */
	if (!sctp_style(sk, UDP))
		return -EINVAL;

	/* Create a new socket.  */
	err = sock_create(sk->sk_family, SOCK_SEQPACKET, IPPROTO_SCTP, &sock);
	if (err < 0)
		return err;

4290
	sctp_copy_sock(sock->sk, sk, asoc);
4291 4292 4293 4294

	/* Make peeled-off sockets more like 1-1 accepted sockets.
	 * Set the daddr and initialize id to something more random
	 */
4295 4296
	af = sctp_get_af_specific(asoc->peer.primary_addr.sa.sa_family);
	af->to_sk_daddr(&asoc->peer.primary_addr, sk);
4297 4298 4299 4300 4301

	/* Populate the fields of the newsk from the oldsk and migrate the
	 * asoc to the newsk.
	 */
	sctp_sock_migrate(sk, sock->sk, asoc, SCTP_SOCKET_UDP_HIGH_BANDWIDTH);
4302

L
Linus Torvalds 已提交
4303 4304 4305 4306
	*sockp = sock;

	return err;
}
B
Benjamin Poirier 已提交
4307
EXPORT_SYMBOL(sctp_do_peeloff);
L
Linus Torvalds 已提交
4308 4309 4310 4311 4312

static int sctp_getsockopt_peeloff(struct sock *sk, int len, char __user *optval, int __user *optlen)
{
	sctp_peeloff_arg_t peeloff;
	struct socket *newsock;
4313
	struct file *newfile;
L
Linus Torvalds 已提交
4314 4315
	int retval = 0;

4316
	if (len < sizeof(sctp_peeloff_arg_t))
L
Linus Torvalds 已提交
4317
		return -EINVAL;
4318
	len = sizeof(sctp_peeloff_arg_t);
L
Linus Torvalds 已提交
4319 4320 4321
	if (copy_from_user(&peeloff, optval, len))
		return -EFAULT;

B
Benjamin Poirier 已提交
4322
	retval = sctp_do_peeloff(sk, peeloff.associd, &newsock);
L
Linus Torvalds 已提交
4323 4324 4325 4326
	if (retval < 0)
		goto out;

	/* Map the socket to an unused fd that can be returned to the user.  */
4327
	retval = get_unused_fd_flags(0);
L
Linus Torvalds 已提交
4328 4329 4330 4331 4332
	if (retval < 0) {
		sock_release(newsock);
		goto out;
	}

4333
	newfile = sock_alloc_file(newsock, 0, NULL);
4334 4335 4336 4337 4338 4339
	if (unlikely(IS_ERR(newfile))) {
		put_unused_fd(retval);
		sock_release(newsock);
		return PTR_ERR(newfile);
	}

4340 4341
	pr_debug("%s: sk:%p, newsk:%p, sd:%d\n", __func__, sk, newsock->sk,
		 retval);
L
Linus Torvalds 已提交
4342 4343

	/* Return the fd mapped to the new socket.  */
4344 4345 4346 4347 4348
	if (put_user(len, optlen)) {
		fput(newfile);
		put_unused_fd(retval);
		return -EFAULT;
	}
L
Linus Torvalds 已提交
4349
	peeloff.sd = retval;
4350 4351 4352
	if (copy_to_user(optval, &peeloff, len)) {
		fput(newfile);
		put_unused_fd(retval);
4353
		return -EFAULT;
4354 4355
	}
	fd_install(retval, newfile);
L
Linus Torvalds 已提交
4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369
out:
	return retval;
}

/* 7.1.13 Peer Address Parameters (SCTP_PEER_ADDR_PARAMS)
 *
 * Applications can enable or disable heartbeats for any peer address of
 * an association, modify an address's heartbeat interval, force a
 * heartbeat to be sent immediately, and adjust the address's maximum
 * number of retransmissions sent before an address is considered
 * unreachable.  The following structure is used to access and modify an
 * address's parameters:
 *
 *  struct sctp_paddrparams {
4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381
 *     sctp_assoc_t            spp_assoc_id;
 *     struct sockaddr_storage spp_address;
 *     uint32_t                spp_hbinterval;
 *     uint16_t                spp_pathmaxrxt;
 *     uint32_t                spp_pathmtu;
 *     uint32_t                spp_sackdelay;
 *     uint32_t                spp_flags;
 * };
 *
 *   spp_assoc_id    - (one-to-many style socket) This is filled in the
 *                     application, and identifies the association for
 *                     this query.
L
Linus Torvalds 已提交
4382 4383
 *   spp_address     - This specifies which address is of interest.
 *   spp_hbinterval  - This contains the value of the heartbeat interval,
4384 4385 4386
 *                     in milliseconds.  If a  value of zero
 *                     is present in this field then no changes are to
 *                     be made to this parameter.
L
Linus Torvalds 已提交
4387 4388
 *   spp_pathmaxrxt  - This contains the maximum number of
 *                     retransmissions before this address shall be
4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451
 *                     considered unreachable. If a  value of zero
 *                     is present in this field then no changes are to
 *                     be made to this parameter.
 *   spp_pathmtu     - When Path MTU discovery is disabled the value
 *                     specified here will be the "fixed" path mtu.
 *                     Note that if the spp_address field is empty
 *                     then all associations on this address will
 *                     have this fixed path mtu set upon them.
 *
 *   spp_sackdelay   - When delayed sack is enabled, this value specifies
 *                     the number of milliseconds that sacks will be delayed
 *                     for. This value will apply to all addresses of an
 *                     association if the spp_address field is empty. Note
 *                     also, that if delayed sack is enabled and this
 *                     value is set to 0, no change is made to the last
 *                     recorded delayed sack timer value.
 *
 *   spp_flags       - These flags are used to control various features
 *                     on an association. The flag field may contain
 *                     zero or more of the following options.
 *
 *                     SPP_HB_ENABLE  - Enable heartbeats on the
 *                     specified address. Note that if the address
 *                     field is empty all addresses for the association
 *                     have heartbeats enabled upon them.
 *
 *                     SPP_HB_DISABLE - Disable heartbeats on the
 *                     speicifed address. Note that if the address
 *                     field is empty all addresses for the association
 *                     will have their heartbeats disabled. Note also
 *                     that SPP_HB_ENABLE and SPP_HB_DISABLE are
 *                     mutually exclusive, only one of these two should
 *                     be specified. Enabling both fields will have
 *                     undetermined results.
 *
 *                     SPP_HB_DEMAND - Request a user initiated heartbeat
 *                     to be made immediately.
 *
 *                     SPP_PMTUD_ENABLE - This field will enable PMTU
 *                     discovery upon the specified address. Note that
 *                     if the address feild is empty then all addresses
 *                     on the association are effected.
 *
 *                     SPP_PMTUD_DISABLE - This field will disable PMTU
 *                     discovery upon the specified address. Note that
 *                     if the address feild is empty then all addresses
 *                     on the association are effected. Not also that
 *                     SPP_PMTUD_ENABLE and SPP_PMTUD_DISABLE are mutually
 *                     exclusive. Enabling both will have undetermined
 *                     results.
 *
 *                     SPP_SACKDELAY_ENABLE - Setting this flag turns
 *                     on delayed sack. The time specified in spp_sackdelay
 *                     is used to specify the sack delay for this address. Note
 *                     that if spp_address is empty then all addresses will
 *                     enable delayed sack and take on the sack delay
 *                     value specified in spp_sackdelay.
 *                     SPP_SACKDELAY_DISABLE - Setting this flag turns
 *                     off delayed sack. If the spp_address field is blank then
 *                     delayed sack is disabled for the entire association. Note
 *                     also that this field is mutually exclusive to
 *                     SPP_SACKDELAY_ENABLE, setting both will have undefined
 *                     results.
L
Linus Torvalds 已提交
4452 4453
 */
static int sctp_getsockopt_peer_addr_params(struct sock *sk, int len,
4454
					    char __user *optval, int __user *optlen)
L
Linus Torvalds 已提交
4455
{
4456 4457 4458 4459
	struct sctp_paddrparams  params;
	struct sctp_transport   *trans = NULL;
	struct sctp_association *asoc = NULL;
	struct sctp_sock        *sp = sctp_sk(sk);
L
Linus Torvalds 已提交
4460

4461
	if (len < sizeof(struct sctp_paddrparams))
L
Linus Torvalds 已提交
4462
		return -EINVAL;
4463
	len = sizeof(struct sctp_paddrparams);
L
Linus Torvalds 已提交
4464 4465 4466
	if (copy_from_user(&params, optval, len))
		return -EFAULT;

4467 4468
	/* If an address other than INADDR_ANY is specified, and
	 * no transport is found, then the request is invalid.
L
Linus Torvalds 已提交
4469
	 */
4470
	if (!sctp_is_any(sk, (union sctp_addr *)&params.spp_address)) {
4471 4472 4473
		trans = sctp_addr_id2transport(sk, &params.spp_address,
					       params.spp_assoc_id);
		if (!trans) {
4474
			pr_debug("%s: failed no transport\n", __func__);
4475 4476
			return -EINVAL;
		}
L
Linus Torvalds 已提交
4477 4478
	}

4479 4480 4481 4482 4483 4484
	/* Get association, if assoc_id != 0 and the socket is a one
	 * to many style socket, and an association was not found, then
	 * the id was invalid.
	 */
	asoc = sctp_id2assoc(sk, params.spp_assoc_id);
	if (!asoc && params.spp_assoc_id && sctp_style(sk, UDP)) {
4485
		pr_debug("%s: failed no association\n", __func__);
L
Linus Torvalds 已提交
4486
		return -EINVAL;
4487
	}
L
Linus Torvalds 已提交
4488

4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512
	if (trans) {
		/* Fetch transport values. */
		params.spp_hbinterval = jiffies_to_msecs(trans->hbinterval);
		params.spp_pathmtu    = trans->pathmtu;
		params.spp_pathmaxrxt = trans->pathmaxrxt;
		params.spp_sackdelay  = jiffies_to_msecs(trans->sackdelay);

		/*draft-11 doesn't say what to return in spp_flags*/
		params.spp_flags      = trans->param_flags;
	} else if (asoc) {
		/* Fetch association values. */
		params.spp_hbinterval = jiffies_to_msecs(asoc->hbinterval);
		params.spp_pathmtu    = asoc->pathmtu;
		params.spp_pathmaxrxt = asoc->pathmaxrxt;
		params.spp_sackdelay  = jiffies_to_msecs(asoc->sackdelay);

		/*draft-11 doesn't say what to return in spp_flags*/
		params.spp_flags      = asoc->param_flags;
	} else {
		/* Fetch socket values. */
		params.spp_hbinterval = sp->hbinterval;
		params.spp_pathmtu    = sp->pathmtu;
		params.spp_sackdelay  = sp->sackdelay;
		params.spp_pathmaxrxt = sp->pathmaxrxt;
L
Linus Torvalds 已提交
4513

4514 4515 4516
		/*draft-11 doesn't say what to return in spp_flags*/
		params.spp_flags      = sp->param_flags;
	}
L
Linus Torvalds 已提交
4517 4518 4519 4520 4521 4522 4523 4524 4525 4526

	if (copy_to_user(optval, &params, len))
		return -EFAULT;

	if (put_user(len, optlen))
		return -EFAULT;

	return 0;
}

4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545
/*
 * 7.1.23.  Get or set delayed ack timer (SCTP_DELAYED_SACK)
 *
 * This option will effect the way delayed acks are performed.  This
 * option allows you to get or set the delayed ack time, in
 * milliseconds.  It also allows changing the delayed ack frequency.
 * Changing the frequency to 1 disables the delayed sack algorithm.  If
 * the assoc_id is 0, then this sets or gets the endpoints default
 * values.  If the assoc_id field is non-zero, then the set or get
 * effects the specified association for the one to many model (the
 * assoc_id field is ignored by the one to one model).  Note that if
 * sack_delay or sack_freq are 0 when setting this option, then the
 * current values will remain unchanged.
 *
 * struct sctp_sack_info {
 *     sctp_assoc_t            sack_assoc_id;
 *     uint32_t                sack_delay;
 *     uint32_t                sack_freq;
 * };
4546
 *
4547 4548 4549 4550
 * sack_assoc_id -  This parameter, indicates which association the user
 *    is performing an action upon.  Note that if this field's value is
 *    zero then the endpoints default value is changed (effecting future
 *    associations only).
4551
 *
4552 4553 4554 4555
 * sack_delay -  This parameter contains the number of milliseconds that
 *    the user is requesting the delayed ACK timer be set to.  Note that
 *    this value is defined in the standard to be between 200 and 500
 *    milliseconds.
4556
 *
4557 4558 4559 4560
 * sack_freq -  This parameter contains the number of packets that must
 *    be received before a sack is sent without waiting for the delay
 *    timer to expire.  The default value for this is 2, setting this
 *    value to 1 will disable the delayed sack algorithm.
4561
 */
4562
static int sctp_getsockopt_delayed_ack(struct sock *sk, int len,
4563 4564 4565
					    char __user *optval,
					    int __user *optlen)
{
4566
	struct sctp_sack_info    params;
4567 4568 4569
	struct sctp_association *asoc = NULL;
	struct sctp_sock        *sp = sctp_sk(sk);

4570 4571
	if (len >= sizeof(struct sctp_sack_info)) {
		len = sizeof(struct sctp_sack_info);
4572

4573 4574 4575
		if (copy_from_user(&params, optval, len))
			return -EFAULT;
	} else if (len == sizeof(struct sctp_assoc_value)) {
4576 4577
		pr_warn("Use of struct sctp_assoc_value in delayed_ack socket option deprecated\n");
		pr_warn("Use struct sctp_sack_info instead\n");
4578 4579 4580
		if (copy_from_user(&params, optval, len))
			return -EFAULT;
	} else
4581
		return -EINVAL;
4582

4583
	/* Get association, if sack_assoc_id != 0 and the socket is a one
4584 4585
	 * to many style socket, and an association was not found, then
	 * the id was invalid.
4586
	 */
4587 4588
	asoc = sctp_id2assoc(sk, params.sack_assoc_id);
	if (!asoc && params.sack_assoc_id && sctp_style(sk, UDP))
4589 4590 4591 4592
		return -EINVAL;

	if (asoc) {
		/* Fetch association values. */
4593 4594
		if (asoc->param_flags & SPP_SACKDELAY_ENABLE) {
			params.sack_delay = jiffies_to_msecs(
4595
				asoc->sackdelay);
4596 4597 4598 4599 4600 4601
			params.sack_freq = asoc->sackfreq;

		} else {
			params.sack_delay = 0;
			params.sack_freq = 1;
		}
4602 4603
	} else {
		/* Fetch socket values. */
4604 4605 4606 4607 4608 4609 4610
		if (sp->param_flags & SPP_SACKDELAY_ENABLE) {
			params.sack_delay  = sp->sackdelay;
			params.sack_freq = sp->sackfreq;
		} else {
			params.sack_delay  = 0;
			params.sack_freq = 1;
		}
4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621
	}

	if (copy_to_user(optval, &params, len))
		return -EFAULT;

	if (put_user(len, optlen))
		return -EFAULT;

	return 0;
}

L
Linus Torvalds 已提交
4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634
/* 7.1.3 Initialization Parameters (SCTP_INITMSG)
 *
 * Applications can specify protocol parameters for the default association
 * initialization.  The option name argument to setsockopt() and getsockopt()
 * is SCTP_INITMSG.
 *
 * Setting initialization parameters is effective only on an unconnected
 * socket (for UDP-style sockets only future associations are effected
 * by the change).  With TCP-style sockets, this option is inherited by
 * sockets derived from a listener socket.
 */
static int sctp_getsockopt_initmsg(struct sock *sk, int len, char __user *optval, int __user *optlen)
{
4635
	if (len < sizeof(struct sctp_initmsg))
L
Linus Torvalds 已提交
4636
		return -EINVAL;
4637 4638 4639
	len = sizeof(struct sctp_initmsg);
	if (put_user(len, optlen))
		return -EFAULT;
L
Linus Torvalds 已提交
4640 4641 4642 4643 4644
	if (copy_to_user(optval, &sctp_sk(sk)->initmsg, len))
		return -EFAULT;
	return 0;
}

4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670

static int sctp_getsockopt_peer_addrs(struct sock *sk, int len,
				      char __user *optval, int __user *optlen)
{
	struct sctp_association *asoc;
	int cnt = 0;
	struct sctp_getaddrs getaddrs;
	struct sctp_transport *from;
	void __user *to;
	union sctp_addr temp;
	struct sctp_sock *sp = sctp_sk(sk);
	int addrlen;
	size_t space_left;
	int bytes_copied;

	if (len < sizeof(struct sctp_getaddrs))
		return -EINVAL;

	if (copy_from_user(&getaddrs, optval, sizeof(struct sctp_getaddrs)))
		return -EFAULT;

	/* For UDP-style sockets, id specifies the association to query.  */
	asoc = sctp_id2assoc(sk, getaddrs.assoc_id);
	if (!asoc)
		return -EINVAL;

4671 4672
	to = optval + offsetof(struct sctp_getaddrs, addrs);
	space_left = len - offsetof(struct sctp_getaddrs, addrs);
4673

4674 4675
	list_for_each_entry(from, &asoc->peer.transport_addr_list,
				transports) {
4676
		memcpy(&temp, &from->ipaddr, sizeof(temp));
4677
		sctp_get_pf_specific(sk->sk_family)->addr_v4map(sp, &temp);
4678
		addrlen = sctp_get_af_specific(temp.sa.sa_family)->sockaddr_len;
4679
		if (space_left < addrlen)
4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691
			return -ENOMEM;
		if (copy_to_user(to, &temp, addrlen))
			return -EFAULT;
		to += addrlen;
		cnt++;
		space_left -= addrlen;
	}

	if (put_user(cnt, &((struct sctp_getaddrs __user *)optval)->addr_num))
		return -EFAULT;
	bytes_copied = ((char __user *)to) - optval;
	if (put_user(bytes_copied, optlen))
L
Linus Torvalds 已提交
4692 4693 4694 4695 4696
		return -EFAULT;

	return 0;
}

4697 4698
static int sctp_copy_laddrs(struct sock *sk, __u16 port, void *to,
			    size_t space_left, int *bytes_copied)
4699 4700 4701 4702 4703
{
	struct sctp_sockaddr_entry *addr;
	union sctp_addr temp;
	int cnt = 0;
	int addrlen;
4704
	struct net *net = sock_net(sk);
4705

4706
	rcu_read_lock();
4707
	list_for_each_entry_rcu(addr, &net->sctp.local_addr_list, list) {
4708 4709 4710
		if (!addr->valid)
			continue;

4711
		if ((PF_INET == sk->sk_family) &&
4712
		    (AF_INET6 == addr->a.sa.sa_family))
4713
			continue;
4714 4715 4716 4717
		if ((PF_INET6 == sk->sk_family) &&
		    inet_v6_ipv6only(sk) &&
		    (AF_INET == addr->a.sa.sa_family))
			continue;
4718
		memcpy(&temp, &addr->a, sizeof(temp));
4719 4720 4721
		if (!temp.v4.sin_port)
			temp.v4.sin_port = htons(port);

4722 4723 4724
		sctp_get_pf_specific(sk->sk_family)->addr_v4map(sctp_sk(sk),
								&temp);
		addrlen = sctp_get_af_specific(temp.sa.sa_family)->sockaddr_len;
4725 4726 4727 4728
		if (space_left < addrlen) {
			cnt =  -ENOMEM;
			break;
		}
4729
		memcpy(to, &temp, addrlen);
4730

4731
		to += addrlen;
4732
		cnt++;
4733
		space_left -= addrlen;
4734
		*bytes_copied += addrlen;
4735
	}
4736
	rcu_read_unlock();
4737 4738 4739 4740

	return cnt;
}

L
Linus Torvalds 已提交
4741

4742 4743 4744 4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755
static int sctp_getsockopt_local_addrs(struct sock *sk, int len,
				       char __user *optval, int __user *optlen)
{
	struct sctp_bind_addr *bp;
	struct sctp_association *asoc;
	int cnt = 0;
	struct sctp_getaddrs getaddrs;
	struct sctp_sockaddr_entry *addr;
	void __user *to;
	union sctp_addr temp;
	struct sctp_sock *sp = sctp_sk(sk);
	int addrlen;
	int err = 0;
	size_t space_left;
4756 4757
	int bytes_copied = 0;
	void *addrs;
4758
	void *buf;
4759

4760
	if (len < sizeof(struct sctp_getaddrs))
4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774 4775 4776 4777 4778 4779 4780
		return -EINVAL;

	if (copy_from_user(&getaddrs, optval, sizeof(struct sctp_getaddrs)))
		return -EFAULT;

	/*
	 *  For UDP-style sockets, id specifies the association to query.
	 *  If the id field is set to the value '0' then the locally bound
	 *  addresses are returned without regard to any particular
	 *  association.
	 */
	if (0 == getaddrs.assoc_id) {
		bp = &sctp_sk(sk)->ep->base.bind_addr;
	} else {
		asoc = sctp_id2assoc(sk, getaddrs.assoc_id);
		if (!asoc)
			return -EINVAL;
		bp = &asoc->base.bind_addr;
	}

4781 4782
	to = optval + offsetof(struct sctp_getaddrs, addrs);
	space_left = len - offsetof(struct sctp_getaddrs, addrs);
4783

4784 4785 4786
	addrs = kmalloc(space_left, GFP_KERNEL);
	if (!addrs)
		return -ENOMEM;
4787 4788 4789 4790 4791 4792 4793

	/* If the endpoint is bound to 0.0.0.0 or ::0, get the valid
	 * addresses from the global local address list.
	 */
	if (sctp_list_single_entry(&bp->address_list)) {
		addr = list_entry(bp->address_list.next,
				  struct sctp_sockaddr_entry, list);
4794
		if (sctp_is_any(sk, &addr->a)) {
4795 4796
			cnt = sctp_copy_laddrs(sk, bp->port, addrs,
						space_left, &bytes_copied);
4797 4798
			if (cnt < 0) {
				err = cnt;
4799
				goto out;
4800
			}
4801
			goto copy_getaddrs;
4802 4803 4804
		}
	}

4805
	buf = addrs;
4806 4807 4808 4809 4810
	/* Protection on the bound address list is not needed since
	 * in the socket option context we hold a socket lock and
	 * thus the bound address list can't change.
	 */
	list_for_each_entry(addr, &bp->address_list, list) {
4811
		memcpy(&temp, &addr->a, sizeof(temp));
4812 4813
		sctp_get_pf_specific(sk->sk_family)->addr_v4map(sp, &temp);
		addrlen = sctp_get_af_specific(temp.sa.sa_family)->sockaddr_len;
4814 4815
		if (space_left < addrlen) {
			err =  -ENOMEM; /*fixme: right error?*/
4816
			goto out;
4817
		}
4818 4819
		memcpy(buf, &temp, addrlen);
		buf += addrlen;
4820
		bytes_copied += addrlen;
4821
		cnt++;
4822 4823 4824 4825
		space_left -= addrlen;
	}

copy_getaddrs:
4826 4827
	if (copy_to_user(to, addrs, bytes_copied)) {
		err = -EFAULT;
S
Sebastian Siewior 已提交
4828
		goto out;
4829
	}
4830 4831
	if (put_user(cnt, &((struct sctp_getaddrs __user *)optval)->addr_num)) {
		err = -EFAULT;
S
Sebastian Siewior 已提交
4832
		goto out;
4833
	}
4834
	if (put_user(bytes_copied, optlen))
4835
		err = -EFAULT;
S
Sebastian Siewior 已提交
4836
out:
4837
	kfree(addrs);
4838 4839 4840
	return err;
}

L
Linus Torvalds 已提交
4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853
/* 7.1.10 Set Primary Address (SCTP_PRIMARY_ADDR)
 *
 * Requests that the local SCTP stack use the enclosed peer address as
 * the association primary.  The enclosed address must be one of the
 * association peer's addresses.
 */
static int sctp_getsockopt_primary_addr(struct sock *sk, int len,
					char __user *optval, int __user *optlen)
{
	struct sctp_prim prim;
	struct sctp_association *asoc;
	struct sctp_sock *sp = sctp_sk(sk);

4854
	if (len < sizeof(struct sctp_prim))
L
Linus Torvalds 已提交
4855 4856
		return -EINVAL;

4857 4858 4859
	len = sizeof(struct sctp_prim);

	if (copy_from_user(&prim, optval, len))
L
Linus Torvalds 已提交
4860 4861 4862 4863 4864 4865 4866 4867
		return -EFAULT;

	asoc = sctp_id2assoc(sk, prim.ssp_assoc_id);
	if (!asoc)
		return -EINVAL;

	if (!asoc->peer.primary_path)
		return -ENOTCONN;
4868

4869 4870
	memcpy(&prim.ssp_addr, &asoc->peer.primary_path->ipaddr,
		asoc->peer.primary_path->af_specific->sockaddr_len);
L
Linus Torvalds 已提交
4871 4872 4873 4874

	sctp_get_pf_specific(sk->sk_family)->addr_v4map(sp,
			(union sctp_addr *)&prim.ssp_addr);

4875 4876 4877
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &prim, len))
L
Linus Torvalds 已提交
4878 4879 4880 4881 4882 4883
		return -EFAULT;

	return 0;
}

/*
4884
 * 7.1.11  Set Adaptation Layer Indicator (SCTP_ADAPTATION_LAYER)
L
Linus Torvalds 已提交
4885
 *
4886
 * Requests that the local endpoint set the specified Adaptation Layer
L
Linus Torvalds 已提交
4887 4888
 * Indication parameter for all future INIT and INIT-ACK exchanges.
 */
4889
static int sctp_getsockopt_adaptation_layer(struct sock *sk, int len,
L
Linus Torvalds 已提交
4890 4891
				  char __user *optval, int __user *optlen)
{
4892
	struct sctp_setadaptation adaptation;
L
Linus Torvalds 已提交
4893

4894
	if (len < sizeof(struct sctp_setadaptation))
L
Linus Torvalds 已提交
4895 4896
		return -EINVAL;

4897 4898
	len = sizeof(struct sctp_setadaptation);

4899
	adaptation.ssb_adaptation_ind = sctp_sk(sk)->adaptation_ind;
4900 4901 4902

	if (put_user(len, optlen))
		return -EFAULT;
4903
	if (copy_to_user(optval, &adaptation, len))
L
Linus Torvalds 已提交
4904
		return -EFAULT;
4905

L
Linus Torvalds 已提交
4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935
	return 0;
}

/*
 *
 * 7.1.14 Set default send parameters (SCTP_DEFAULT_SEND_PARAM)
 *
 *   Applications that wish to use the sendto() system call may wish to
 *   specify a default set of parameters that would normally be supplied
 *   through the inclusion of ancillary data.  This socket option allows
 *   such an application to set the default sctp_sndrcvinfo structure.


 *   The application that wishes to use this socket option simply passes
 *   in to this call the sctp_sndrcvinfo structure defined in Section
 *   5.2.2) The input parameters accepted by this call include
 *   sinfo_stream, sinfo_flags, sinfo_ppid, sinfo_context,
 *   sinfo_timetolive.  The user must provide the sinfo_assoc_id field in
 *   to this call if the caller is using the UDP model.
 *
 *   For getsockopt, it get the default sctp_sndrcvinfo structure.
 */
static int sctp_getsockopt_default_send_param(struct sock *sk,
					int len, char __user *optval,
					int __user *optlen)
{
	struct sctp_sndrcvinfo info;
	struct sctp_association *asoc;
	struct sctp_sock *sp = sctp_sk(sk);

4936
	if (len < sizeof(struct sctp_sndrcvinfo))
L
Linus Torvalds 已提交
4937
		return -EINVAL;
4938 4939 4940 4941

	len = sizeof(struct sctp_sndrcvinfo);

	if (copy_from_user(&info, optval, len))
L
Linus Torvalds 已提交
4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961
		return -EFAULT;

	asoc = sctp_id2assoc(sk, info.sinfo_assoc_id);
	if (!asoc && info.sinfo_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	if (asoc) {
		info.sinfo_stream = asoc->default_stream;
		info.sinfo_flags = asoc->default_flags;
		info.sinfo_ppid = asoc->default_ppid;
		info.sinfo_context = asoc->default_context;
		info.sinfo_timetolive = asoc->default_timetolive;
	} else {
		info.sinfo_stream = sp->default_stream;
		info.sinfo_flags = sp->default_flags;
		info.sinfo_ppid = sp->default_ppid;
		info.sinfo_context = sp->default_context;
		info.sinfo_timetolive = sp->default_timetolive;
	}

4962 4963 4964
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &info, len))
L
Linus Torvalds 已提交
4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014
		return -EFAULT;

	return 0;
}

/*
 *
 * 7.1.5 SCTP_NODELAY
 *
 * Turn on/off any Nagle-like algorithm.  This means that packets are
 * generally sent as soon as possible and no unnecessary delays are
 * introduced, at the cost of more packets in the network.  Expects an
 * integer boolean flag.
 */

static int sctp_getsockopt_nodelay(struct sock *sk, int len,
				   char __user *optval, int __user *optlen)
{
	int val;

	if (len < sizeof(int))
		return -EINVAL;

	len = sizeof(int);
	val = (sctp_sk(sk)->nodelay == 1);
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;
	return 0;
}

/*
 *
 * 7.1.1 SCTP_RTOINFO
 *
 * The protocol parameters used to initialize and bound retransmission
 * timeout (RTO) are tunable. sctp_rtoinfo structure is used to access
 * and modify these parameters.
 * All parameters are time values, in milliseconds.  A value of 0, when
 * modifying the parameters, indicates that the current value should not
 * be changed.
 *
 */
static int sctp_getsockopt_rtoinfo(struct sock *sk, int len,
				char __user *optval,
				int __user *optlen) {
	struct sctp_rtoinfo rtoinfo;
	struct sctp_association *asoc;

5015
	if (len < sizeof (struct sctp_rtoinfo))
L
Linus Torvalds 已提交
5016 5017
		return -EINVAL;

5018 5019 5020
	len = sizeof(struct sctp_rtoinfo);

	if (copy_from_user(&rtoinfo, optval, len))
L
Linus Torvalds 已提交
5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054
		return -EFAULT;

	asoc = sctp_id2assoc(sk, rtoinfo.srto_assoc_id);

	if (!asoc && rtoinfo.srto_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	/* Values corresponding to the specific association. */
	if (asoc) {
		rtoinfo.srto_initial = jiffies_to_msecs(asoc->rto_initial);
		rtoinfo.srto_max = jiffies_to_msecs(asoc->rto_max);
		rtoinfo.srto_min = jiffies_to_msecs(asoc->rto_min);
	} else {
		/* Values corresponding to the endpoint. */
		struct sctp_sock *sp = sctp_sk(sk);

		rtoinfo.srto_initial = sp->rtoinfo.srto_initial;
		rtoinfo.srto_max = sp->rtoinfo.srto_max;
		rtoinfo.srto_min = sp->rtoinfo.srto_min;
	}

	if (put_user(len, optlen))
		return -EFAULT;

	if (copy_to_user(optval, &rtoinfo, len))
		return -EFAULT;

	return 0;
}

/*
 *
 * 7.1.2 SCTP_ASSOCINFO
 *
5055
 * This option is used to tune the maximum retransmission attempts
L
Linus Torvalds 已提交
5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071
 * of the association.
 * Returns an error if the new association retransmission value is
 * greater than the sum of the retransmission value  of the peer.
 * See [SCTP] for more information.
 *
 */
static int sctp_getsockopt_associnfo(struct sock *sk, int len,
				     char __user *optval,
				     int __user *optlen)
{

	struct sctp_assocparams assocparams;
	struct sctp_association *asoc;
	struct list_head *pos;
	int cnt = 0;

5072
	if (len < sizeof (struct sctp_assocparams))
L
Linus Torvalds 已提交
5073 5074
		return -EINVAL;

5075 5076 5077
	len = sizeof(struct sctp_assocparams);

	if (copy_from_user(&assocparams, optval, len))
L
Linus Torvalds 已提交
5078 5079 5080 5081 5082 5083 5084 5085
		return -EFAULT;

	asoc = sctp_id2assoc(sk, assocparams.sasoc_assoc_id);

	if (!asoc && assocparams.sasoc_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	/* Values correspoinding to the specific association */
5086
	if (asoc) {
L
Linus Torvalds 已提交
5087 5088 5089
		assocparams.sasoc_asocmaxrxt = asoc->max_retrans;
		assocparams.sasoc_peer_rwnd = asoc->peer.rwnd;
		assocparams.sasoc_local_rwnd = asoc->a_rwnd;
5090
		assocparams.sasoc_cookie_life = ktime_to_ms(asoc->cookie_life);
L
Linus Torvalds 已提交
5091 5092

		list_for_each(pos, &asoc->peer.transport_addr_list) {
5093
			cnt++;
L
Linus Torvalds 已提交
5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139 5140 5141 5142 5143 5144 5145 5146 5147 5148
		}

		assocparams.sasoc_number_peer_destinations = cnt;
	} else {
		/* Values corresponding to the endpoint */
		struct sctp_sock *sp = sctp_sk(sk);

		assocparams.sasoc_asocmaxrxt = sp->assocparams.sasoc_asocmaxrxt;
		assocparams.sasoc_peer_rwnd = sp->assocparams.sasoc_peer_rwnd;
		assocparams.sasoc_local_rwnd = sp->assocparams.sasoc_local_rwnd;
		assocparams.sasoc_cookie_life =
					sp->assocparams.sasoc_cookie_life;
		assocparams.sasoc_number_peer_destinations =
					sp->assocparams.
					sasoc_number_peer_destinations;
	}

	if (put_user(len, optlen))
		return -EFAULT;

	if (copy_to_user(optval, &assocparams, len))
		return -EFAULT;

	return 0;
}

/*
 * 7.1.16 Set/clear IPv4 mapped addresses (SCTP_I_WANT_MAPPED_V4_ADDR)
 *
 * This socket option is a boolean flag which turns on or off mapped V4
 * addresses.  If this option is turned on and the socket is type
 * PF_INET6, then IPv4 addresses will be mapped to V6 representation.
 * If this option is turned off, then no mapping will be done of V4
 * addresses and a user will receive both PF_INET6 and PF_INET type
 * addresses on the socket.
 */
static int sctp_getsockopt_mappedv4(struct sock *sk, int len,
				    char __user *optval, int __user *optlen)
{
	int val;
	struct sctp_sock *sp = sctp_sk(sk);

	if (len < sizeof(int))
		return -EINVAL;

	len = sizeof(int);
	val = sp->v4mapped;
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;

	return 0;
}

5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159
/*
 * 7.1.29.  Set or Get the default context (SCTP_CONTEXT)
 * (chapter and verse is quoted at sctp_setsockopt_context())
 */
static int sctp_getsockopt_context(struct sock *sk, int len,
				   char __user *optval, int __user *optlen)
{
	struct sctp_assoc_value params;
	struct sctp_sock *sp;
	struct sctp_association *asoc;

5160
	if (len < sizeof(struct sctp_assoc_value))
5161 5162
		return -EINVAL;

5163 5164
	len = sizeof(struct sctp_assoc_value);

5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186
	if (copy_from_user(&params, optval, len))
		return -EFAULT;

	sp = sctp_sk(sk);

	if (params.assoc_id != 0) {
		asoc = sctp_id2assoc(sk, params.assoc_id);
		if (!asoc)
			return -EINVAL;
		params.assoc_value = asoc->default_rcv_context;
	} else {
		params.assoc_value = sp->default_rcv_context;
	}

	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &params, len))
		return -EFAULT;

	return 0;
}

L
Linus Torvalds 已提交
5187
/*
5188 5189 5190
 * 8.1.16.  Get or Set the Maximum Fragmentation Size (SCTP_MAXSEG)
 * This option will get or set the maximum size to put in any outgoing
 * SCTP DATA chunk.  If a message is larger than this size it will be
L
Linus Torvalds 已提交
5191 5192 5193
 * fragmented by SCTP into the specified size.  Note that the underlying
 * SCTP implementation may fragment into smaller sized chunks when the
 * PMTU of the underlying association is smaller than the value set by
5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212
 * the user.  The default value for this option is '0' which indicates
 * the user is NOT limiting fragmentation and only the PMTU will effect
 * SCTP's choice of DATA chunk size.  Note also that values set larger
 * than the maximum size of an IP datagram will effectively let SCTP
 * control fragmentation (i.e. the same as setting this option to 0).
 *
 * The following structure is used to access and modify this parameter:
 *
 * struct sctp_assoc_value {
 *   sctp_assoc_t assoc_id;
 *   uint32_t assoc_value;
 * };
 *
 * assoc_id:  This parameter is ignored for one-to-one style sockets.
 *    For one-to-many style sockets this parameter indicates which
 *    association the user is performing an action upon.  Note that if
 *    this field's value is zero then the endpoints default value is
 *    changed (effecting future associations only).
 * assoc_value:  This parameter specifies the maximum size in bytes.
L
Linus Torvalds 已提交
5213 5214 5215 5216
 */
static int sctp_getsockopt_maxseg(struct sock *sk, int len,
				  char __user *optval, int __user *optlen)
{
5217 5218
	struct sctp_assoc_value params;
	struct sctp_association *asoc;
L
Linus Torvalds 已提交
5219

5220
	if (len == sizeof(int)) {
5221 5222
		pr_warn("Use of int in maxseg socket option deprecated\n");
		pr_warn("Use struct sctp_assoc_value instead\n");
5223 5224 5225 5226 5227 5228
		params.assoc_id = 0;
	} else if (len >= sizeof(struct sctp_assoc_value)) {
		len = sizeof(struct sctp_assoc_value);
		if (copy_from_user(&params, optval, sizeof(params)))
			return -EFAULT;
	} else
L
Linus Torvalds 已提交
5229 5230
		return -EINVAL;

5231 5232 5233 5234 5235 5236 5237 5238
	asoc = sctp_id2assoc(sk, params.assoc_id);
	if (!asoc && params.assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	if (asoc)
		params.assoc_value = asoc->frag_point;
	else
		params.assoc_value = sctp_sk(sk)->user_frag;
L
Linus Torvalds 已提交
5239 5240 5241

	if (put_user(len, optlen))
		return -EFAULT;
5242 5243 5244 5245 5246 5247 5248
	if (len == sizeof(int)) {
		if (copy_to_user(optval, &params.assoc_value, len))
			return -EFAULT;
	} else {
		if (copy_to_user(optval, &params, len))
			return -EFAULT;
	}
L
Linus Torvalds 已提交
5249 5250 5251 5252

	return 0;
}

5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275
/*
 * 7.1.24.  Get or set fragmented interleave (SCTP_FRAGMENT_INTERLEAVE)
 * (chapter and verse is quoted at sctp_setsockopt_fragment_interleave())
 */
static int sctp_getsockopt_fragment_interleave(struct sock *sk, int len,
					       char __user *optval, int __user *optlen)
{
	int val;

	if (len < sizeof(int))
		return -EINVAL;

	len = sizeof(int);

	val = sctp_sk(sk)->frag_interleave;
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;

	return 0;
}

5276 5277 5278 5279 5280 5281 5282 5283
/*
 * 7.1.25.  Set or Get the sctp partial delivery point
 * (chapter and verse is quoted at sctp_setsockopt_partial_delivery_point())
 */
static int sctp_getsockopt_partial_delivery_point(struct sock *sk, int len,
						  char __user *optval,
						  int __user *optlen)
{
5284
	u32 val;
5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296

	if (len < sizeof(u32))
		return -EINVAL;

	len = sizeof(u32);

	val = sctp_sk(sk)->pd_point;
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;

5297
	return 0;
5298 5299
}

5300 5301 5302 5303 5304 5305 5306 5307
/*
 * 7.1.28.  Set or Get the maximum burst (SCTP_MAX_BURST)
 * (chapter and verse is quoted at sctp_setsockopt_maxburst())
 */
static int sctp_getsockopt_maxburst(struct sock *sk, int len,
				    char __user *optval,
				    int __user *optlen)
{
5308 5309 5310
	struct sctp_assoc_value params;
	struct sctp_sock *sp;
	struct sctp_association *asoc;
5311

5312
	if (len == sizeof(int)) {
5313 5314
		pr_warn("Use of int in max_burst socket option deprecated\n");
		pr_warn("Use struct sctp_assoc_value instead\n");
5315
		params.assoc_id = 0;
5316 5317
	} else if (len >= sizeof(struct sctp_assoc_value)) {
		len = sizeof(struct sctp_assoc_value);
5318 5319 5320 5321
		if (copy_from_user(&params, optval, len))
			return -EFAULT;
	} else
		return -EINVAL;
5322

5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336 5337 5338 5339 5340 5341
	sp = sctp_sk(sk);

	if (params.assoc_id != 0) {
		asoc = sctp_id2assoc(sk, params.assoc_id);
		if (!asoc)
			return -EINVAL;
		params.assoc_value = asoc->max_burst;
	} else
		params.assoc_value = sp->max_burst;

	if (len == sizeof(int)) {
		if (copy_to_user(optval, &params.assoc_value, len))
			return -EFAULT;
	} else {
		if (copy_to_user(optval, &params, len))
			return -EFAULT;
	}

	return 0;
5342 5343 5344

}

5345 5346 5347
static int sctp_getsockopt_hmac_ident(struct sock *sk, int len,
				    char __user *optval, int __user *optlen)
{
5348
	struct net *net = sock_net(sk);
5349
	struct sctp_hmacalgo  __user *p = (void __user *)optval;
5350
	struct sctp_hmac_algo_param *hmacs;
5351 5352 5353
	__u16 data_len = 0;
	u32 num_idents;

5354
	if (!net->sctp.auth_enable)
5355
		return -EACCES;
5356 5357

	hmacs = sctp_sk(sk)->ep->auth_hmacs_list;
5358
	data_len = ntohs(hmacs->param_hdr.length) - sizeof(sctp_paramhdr_t);
5359

5360
	if (len < sizeof(struct sctp_hmacalgo) + data_len)
5361
		return -EINVAL;
5362 5363 5364 5365

	len = sizeof(struct sctp_hmacalgo) + data_len;
	num_idents = data_len / sizeof(u16);

5366 5367
	if (put_user(len, optlen))
		return -EFAULT;
5368 5369 5370
	if (put_user(num_idents, &p->shmac_num_idents))
		return -EFAULT;
	if (copy_to_user(p->shmac_idents, hmacs->hmac_ids, data_len))
5371 5372 5373 5374 5375 5376 5377
		return -EFAULT;
	return 0;
}

static int sctp_getsockopt_active_key(struct sock *sk, int len,
				    char __user *optval, int __user *optlen)
{
5378
	struct net *net = sock_net(sk);
5379 5380 5381
	struct sctp_authkeyid val;
	struct sctp_association *asoc;

5382
	if (!net->sctp.auth_enable)
5383 5384
		return -EACCES;

5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398
	if (len < sizeof(struct sctp_authkeyid))
		return -EINVAL;
	if (copy_from_user(&val, optval, sizeof(struct sctp_authkeyid)))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, val.scact_assoc_id);
	if (!asoc && val.scact_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	if (asoc)
		val.scact_keynumber = asoc->active_key_id;
	else
		val.scact_keynumber = sctp_sk(sk)->ep->active_key_id;

5399 5400 5401 5402 5403 5404
	len = sizeof(struct sctp_authkeyid);
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;

5405 5406 5407 5408 5409 5410
	return 0;
}

static int sctp_getsockopt_peer_auth_chunks(struct sock *sk, int len,
				    char __user *optval, int __user *optlen)
{
5411
	struct net *net = sock_net(sk);
A
Al Viro 已提交
5412
	struct sctp_authchunks __user *p = (void __user *)optval;
5413 5414 5415
	struct sctp_authchunks val;
	struct sctp_association *asoc;
	struct sctp_chunks_param *ch;
5416
	u32    num_chunks = 0;
5417 5418
	char __user *to;

5419
	if (!net->sctp.auth_enable)
5420 5421 5422
		return -EACCES;

	if (len < sizeof(struct sctp_authchunks))
5423 5424
		return -EINVAL;

5425
	if (copy_from_user(&val, optval, sizeof(struct sctp_authchunks)))
5426 5427
		return -EFAULT;

A
Al Viro 已提交
5428
	to = p->gauth_chunks;
5429 5430 5431 5432 5433
	asoc = sctp_id2assoc(sk, val.gauth_assoc_id);
	if (!asoc)
		return -EINVAL;

	ch = asoc->peer.peer_chunks;
5434 5435
	if (!ch)
		goto num;
5436 5437

	/* See if the user provided enough room for all the data */
5438 5439
	num_chunks = ntohs(ch->param_hdr.length) - sizeof(sctp_paramhdr_t);
	if (len < num_chunks)
5440 5441
		return -EINVAL;

5442
	if (copy_to_user(to, ch->chunks, num_chunks))
5443
		return -EFAULT;
5444 5445 5446
num:
	len = sizeof(struct sctp_authchunks) + num_chunks;
	if (put_user(len, optlen)) return -EFAULT;
5447 5448
	if (put_user(num_chunks, &p->gauth_number_of_chunks))
		return -EFAULT;
5449 5450 5451 5452 5453 5454
	return 0;
}

static int sctp_getsockopt_local_auth_chunks(struct sock *sk, int len,
				    char __user *optval, int __user *optlen)
{
5455
	struct net *net = sock_net(sk);
A
Al Viro 已提交
5456
	struct sctp_authchunks __user *p = (void __user *)optval;
5457 5458 5459
	struct sctp_authchunks val;
	struct sctp_association *asoc;
	struct sctp_chunks_param *ch;
5460
	u32    num_chunks = 0;
5461 5462
	char __user *to;

5463
	if (!net->sctp.auth_enable)
5464 5465 5466
		return -EACCES;

	if (len < sizeof(struct sctp_authchunks))
5467 5468
		return -EINVAL;

5469
	if (copy_from_user(&val, optval, sizeof(struct sctp_authchunks)))
5470 5471
		return -EFAULT;

A
Al Viro 已提交
5472
	to = p->gauth_chunks;
5473 5474 5475 5476 5477 5478 5479 5480 5481
	asoc = sctp_id2assoc(sk, val.gauth_assoc_id);
	if (!asoc && val.gauth_assoc_id && sctp_style(sk, UDP))
		return -EINVAL;

	if (asoc)
		ch = (struct sctp_chunks_param*)asoc->c.auth_chunks;
	else
		ch = sctp_sk(sk)->ep->auth_chunk_list;

5482 5483 5484
	if (!ch)
		goto num;

5485
	num_chunks = ntohs(ch->param_hdr.length) - sizeof(sctp_paramhdr_t);
5486
	if (len < sizeof(struct sctp_authchunks) + num_chunks)
5487 5488
		return -EINVAL;

5489 5490 5491 5492
	if (copy_to_user(to, ch->chunks, num_chunks))
		return -EFAULT;
num:
	len = sizeof(struct sctp_authchunks) + num_chunks;
5493 5494
	if (put_user(len, optlen))
		return -EFAULT;
5495 5496
	if (put_user(num_chunks, &p->gauth_number_of_chunks))
		return -EFAULT;
5497 5498 5499 5500

	return 0;
}

5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523 5524 5525 5526 5527 5528 5529 5530 5531 5532
/*
 * 8.2.5.  Get the Current Number of Associations (SCTP_GET_ASSOC_NUMBER)
 * This option gets the current number of associations that are attached
 * to a one-to-many style socket.  The option value is an uint32_t.
 */
static int sctp_getsockopt_assoc_number(struct sock *sk, int len,
				    char __user *optval, int __user *optlen)
{
	struct sctp_sock *sp = sctp_sk(sk);
	struct sctp_association *asoc;
	u32 val = 0;

	if (sctp_style(sk, TCP))
		return -EOPNOTSUPP;

	if (len < sizeof(u32))
		return -EINVAL;

	len = sizeof(u32);

	list_for_each_entry(asoc, &(sp->ep->asocs), asocs) {
		val++;
	}

	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;

	return 0;
}

5533 5534 5535 5536 5537 5538 5539 5540 5541 5542 5543 5544 5545 5546 5547 5548 5549 5550 5551 5552 5553 5554
/*
 * 8.1.23 SCTP_AUTO_ASCONF
 * See the corresponding setsockopt entry as description
 */
static int sctp_getsockopt_auto_asconf(struct sock *sk, int len,
				   char __user *optval, int __user *optlen)
{
	int val = 0;

	if (len < sizeof(int))
		return -EINVAL;

	len = sizeof(int);
	if (sctp_sk(sk)->do_auto_asconf && sctp_is_ep_boundall(sk))
		val = 1;
	if (put_user(len, optlen))
		return -EFAULT;
	if (copy_to_user(optval, &val, len))
		return -EFAULT;
	return 0;
}

5555 5556 5557 5558 5559 5560 5561 5562 5563 5564 5565 5566 5567 5568 5569 5570 5571 5572 5573 5574 5575 5576 5577 5578 5579 5580 5581 5582 5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596 5597 5598 5599 5600 5601 5602 5603
/*
 * 8.2.6. Get the Current Identifiers of Associations
 *        (SCTP_GET_ASSOC_ID_LIST)
 *
 * This option gets the current list of SCTP association identifiers of
 * the SCTP associations handled by a one-to-many style socket.
 */
static int sctp_getsockopt_assoc_ids(struct sock *sk, int len,
				    char __user *optval, int __user *optlen)
{
	struct sctp_sock *sp = sctp_sk(sk);
	struct sctp_association *asoc;
	struct sctp_assoc_ids *ids;
	u32 num = 0;

	if (sctp_style(sk, TCP))
		return -EOPNOTSUPP;

	if (len < sizeof(struct sctp_assoc_ids))
		return -EINVAL;

	list_for_each_entry(asoc, &(sp->ep->asocs), asocs) {
		num++;
	}

	if (len < sizeof(struct sctp_assoc_ids) + sizeof(sctp_assoc_t) * num)
		return -EINVAL;

	len = sizeof(struct sctp_assoc_ids) + sizeof(sctp_assoc_t) * num;

	ids = kmalloc(len, GFP_KERNEL);
	if (unlikely(!ids))
		return -ENOMEM;

	ids->gaids_number_of_ids = num;
	num = 0;
	list_for_each_entry(asoc, &(sp->ep->asocs), asocs) {
		ids->gaids_assoc_id[num++] = asoc->assoc_id;
	}

	if (put_user(len, optlen) || copy_to_user(optval, ids, len)) {
		kfree(ids);
		return -EFAULT;
	}

	kfree(ids);
	return 0;
}

5604 5605 5606 5607 5608 5609 5610 5611 5612 5613 5614 5615 5616 5617 5618 5619 5620 5621 5622 5623 5624 5625 5626 5627 5628 5629 5630 5631 5632 5633 5634 5635 5636 5637 5638 5639 5640 5641 5642 5643 5644 5645 5646 5647 5648
/*
 * SCTP_PEER_ADDR_THLDS
 *
 * This option allows us to fetch the partially failed threshold for one or all
 * transports in an association.  See Section 6.1 of:
 * http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
 */
static int sctp_getsockopt_paddr_thresholds(struct sock *sk,
					    char __user *optval,
					    int len,
					    int __user *optlen)
{
	struct sctp_paddrthlds val;
	struct sctp_transport *trans;
	struct sctp_association *asoc;

	if (len < sizeof(struct sctp_paddrthlds))
		return -EINVAL;
	len = sizeof(struct sctp_paddrthlds);
	if (copy_from_user(&val, (struct sctp_paddrthlds __user *)optval, len))
		return -EFAULT;

	if (sctp_is_any(sk, (const union sctp_addr *)&val.spt_address)) {
		asoc = sctp_id2assoc(sk, val.spt_assoc_id);
		if (!asoc)
			return -ENOENT;

		val.spt_pathpfthld = asoc->pf_retrans;
		val.spt_pathmaxrxt = asoc->pathmaxrxt;
	} else {
		trans = sctp_addr_id2transport(sk, &val.spt_address,
					       val.spt_assoc_id);
		if (!trans)
			return -ENOENT;

		val.spt_pathmaxrxt = trans->pathmaxrxt;
		val.spt_pathpfthld = trans->pf_retrans;
	}

	if (put_user(len, optlen) || copy_to_user(optval, &val, len))
		return -EFAULT;

	return 0;
}

5649 5650 5651 5652 5653 5654 5655 5656 5657 5658 5659 5660 5661 5662 5663 5664 5665
/*
 * SCTP_GET_ASSOC_STATS
 *
 * This option retrieves local per endpoint statistics. It is modeled
 * after OpenSolaris' implementation
 */
static int sctp_getsockopt_assoc_stats(struct sock *sk, int len,
				       char __user *optval,
				       int __user *optlen)
{
	struct sctp_assoc_stats sas;
	struct sctp_association *asoc = NULL;

	/* User must provide at least the assoc id */
	if (len < sizeof(sctp_assoc_t))
		return -EINVAL;

5666 5667 5668
	/* Allow the struct to grow and fill in as much as possible */
	len = min_t(size_t, len, sizeof(sas));

5669 5670 5671 5672 5673 5674 5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696 5697 5698 5699 5700 5701 5702 5703 5704
	if (copy_from_user(&sas, optval, len))
		return -EFAULT;

	asoc = sctp_id2assoc(sk, sas.sas_assoc_id);
	if (!asoc)
		return -EINVAL;

	sas.sas_rtxchunks = asoc->stats.rtxchunks;
	sas.sas_gapcnt = asoc->stats.gapcnt;
	sas.sas_outofseqtsns = asoc->stats.outofseqtsns;
	sas.sas_osacks = asoc->stats.osacks;
	sas.sas_isacks = asoc->stats.isacks;
	sas.sas_octrlchunks = asoc->stats.octrlchunks;
	sas.sas_ictrlchunks = asoc->stats.ictrlchunks;
	sas.sas_oodchunks = asoc->stats.oodchunks;
	sas.sas_iodchunks = asoc->stats.iodchunks;
	sas.sas_ouodchunks = asoc->stats.ouodchunks;
	sas.sas_iuodchunks = asoc->stats.iuodchunks;
	sas.sas_idupchunks = asoc->stats.idupchunks;
	sas.sas_opackets = asoc->stats.opackets;
	sas.sas_ipackets = asoc->stats.ipackets;

	/* New high max rto observed, will return 0 if not a single
	 * RTO update took place. obs_rto_ipaddr will be bogus
	 * in such a case
	 */
	sas.sas_maxrto = asoc->stats.max_obs_rto;
	memcpy(&sas.sas_obs_rto_ipaddr, &asoc->stats.obs_rto_ipaddr,
		sizeof(struct sockaddr_storage));

	/* Mark beginning of a new observation period */
	asoc->stats.max_obs_rto = asoc->rto_min;

	if (put_user(len, optlen))
		return -EFAULT;

5705
	pr_debug("%s: len:%d, assoc_id:%d\n", __func__, len, sas.sas_assoc_id);
5706 5707 5708 5709 5710 5711 5712

	if (copy_to_user(optval, &sas, len))
		return -EFAULT;

	return 0;
}

5713 5714
static int sctp_getsockopt(struct sock *sk, int level, int optname,
			   char __user *optval, int __user *optlen)
L
Linus Torvalds 已提交
5715 5716 5717 5718
{
	int retval = 0;
	int len;

5719
	pr_debug("%s: sk:%p, optname:%d\n", __func__, sk, optname);
L
Linus Torvalds 已提交
5720 5721 5722 5723 5724 5725 5726 5727 5728 5729 5730 5731 5732 5733 5734 5735 5736 5737 5738 5739 5740 5741 5742 5743 5744 5745 5746 5747 5748 5749 5750 5751 5752 5753 5754 5755 5756 5757 5758 5759

	/* I can hardly begin to describe how wrong this is.  This is
	 * so broken as to be worse than useless.  The API draft
	 * REALLY is NOT helpful here...  I am not convinced that the
	 * semantics of getsockopt() with a level OTHER THAN SOL_SCTP
	 * are at all well-founded.
	 */
	if (level != SOL_SCTP) {
		struct sctp_af *af = sctp_sk(sk)->pf->af;

		retval = af->getsockopt(sk, level, optname, optval, optlen);
		return retval;
	}

	if (get_user(len, optlen))
		return -EFAULT;

	sctp_lock_sock(sk);

	switch (optname) {
	case SCTP_STATUS:
		retval = sctp_getsockopt_sctp_status(sk, len, optval, optlen);
		break;
	case SCTP_DISABLE_FRAGMENTS:
		retval = sctp_getsockopt_disable_fragments(sk, len, optval,
							   optlen);
		break;
	case SCTP_EVENTS:
		retval = sctp_getsockopt_events(sk, len, optval, optlen);
		break;
	case SCTP_AUTOCLOSE:
		retval = sctp_getsockopt_autoclose(sk, len, optval, optlen);
		break;
	case SCTP_SOCKOPT_PEELOFF:
		retval = sctp_getsockopt_peeloff(sk, len, optval, optlen);
		break;
	case SCTP_PEER_ADDR_PARAMS:
		retval = sctp_getsockopt_peer_addr_params(sk, len, optval,
							  optlen);
		break;
5760
	case SCTP_DELAYED_SACK:
5761
		retval = sctp_getsockopt_delayed_ack(sk, len, optval,
5762 5763
							  optlen);
		break;
L
Linus Torvalds 已提交
5764 5765 5766 5767 5768 5769 5770 5771 5772 5773 5774
	case SCTP_INITMSG:
		retval = sctp_getsockopt_initmsg(sk, len, optval, optlen);
		break;
	case SCTP_GET_PEER_ADDRS:
		retval = sctp_getsockopt_peer_addrs(sk, len, optval,
						    optlen);
		break;
	case SCTP_GET_LOCAL_ADDRS:
		retval = sctp_getsockopt_local_addrs(sk, len, optval,
						     optlen);
		break;
5775 5776 5777
	case SCTP_SOCKOPT_CONNECTX3:
		retval = sctp_getsockopt_connectx3(sk, len, optval, optlen);
		break;
L
Linus Torvalds 已提交
5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796 5797 5798 5799 5800 5801 5802 5803
	case SCTP_DEFAULT_SEND_PARAM:
		retval = sctp_getsockopt_default_send_param(sk, len,
							    optval, optlen);
		break;
	case SCTP_PRIMARY_ADDR:
		retval = sctp_getsockopt_primary_addr(sk, len, optval, optlen);
		break;
	case SCTP_NODELAY:
		retval = sctp_getsockopt_nodelay(sk, len, optval, optlen);
		break;
	case SCTP_RTOINFO:
		retval = sctp_getsockopt_rtoinfo(sk, len, optval, optlen);
		break;
	case SCTP_ASSOCINFO:
		retval = sctp_getsockopt_associnfo(sk, len, optval, optlen);
		break;
	case SCTP_I_WANT_MAPPED_V4_ADDR:
		retval = sctp_getsockopt_mappedv4(sk, len, optval, optlen);
		break;
	case SCTP_MAXSEG:
		retval = sctp_getsockopt_maxseg(sk, len, optval, optlen);
		break;
	case SCTP_GET_PEER_ADDR_INFO:
		retval = sctp_getsockopt_peer_addr_info(sk, len, optval,
							optlen);
		break;
5804 5805
	case SCTP_ADAPTATION_LAYER:
		retval = sctp_getsockopt_adaptation_layer(sk, len, optval,
L
Linus Torvalds 已提交
5806 5807
							optlen);
		break;
5808 5809 5810
	case SCTP_CONTEXT:
		retval = sctp_getsockopt_context(sk, len, optval, optlen);
		break;
5811 5812 5813 5814
	case SCTP_FRAGMENT_INTERLEAVE:
		retval = sctp_getsockopt_fragment_interleave(sk, len, optval,
							     optlen);
		break;
5815 5816 5817 5818
	case SCTP_PARTIAL_DELIVERY_POINT:
		retval = sctp_getsockopt_partial_delivery_point(sk, len, optval,
								optlen);
		break;
5819 5820 5821
	case SCTP_MAX_BURST:
		retval = sctp_getsockopt_maxburst(sk, len, optval, optlen);
		break;
5822 5823 5824 5825 5826 5827 5828 5829 5830 5831 5832 5833 5834 5835 5836 5837 5838 5839 5840
	case SCTP_AUTH_KEY:
	case SCTP_AUTH_CHUNK:
	case SCTP_AUTH_DELETE_KEY:
		retval = -EOPNOTSUPP;
		break;
	case SCTP_HMAC_IDENT:
		retval = sctp_getsockopt_hmac_ident(sk, len, optval, optlen);
		break;
	case SCTP_AUTH_ACTIVE_KEY:
		retval = sctp_getsockopt_active_key(sk, len, optval, optlen);
		break;
	case SCTP_PEER_AUTH_CHUNKS:
		retval = sctp_getsockopt_peer_auth_chunks(sk, len, optval,
							optlen);
		break;
	case SCTP_LOCAL_AUTH_CHUNKS:
		retval = sctp_getsockopt_local_auth_chunks(sk, len, optval,
							optlen);
		break;
5841 5842 5843
	case SCTP_GET_ASSOC_NUMBER:
		retval = sctp_getsockopt_assoc_number(sk, len, optval, optlen);
		break;
5844 5845 5846
	case SCTP_GET_ASSOC_ID_LIST:
		retval = sctp_getsockopt_assoc_ids(sk, len, optval, optlen);
		break;
5847 5848 5849
	case SCTP_AUTO_ASCONF:
		retval = sctp_getsockopt_auto_asconf(sk, len, optval, optlen);
		break;
5850 5851 5852
	case SCTP_PEER_ADDR_THLDS:
		retval = sctp_getsockopt_paddr_thresholds(sk, optval, len, optlen);
		break;
5853 5854 5855
	case SCTP_GET_ASSOC_STATS:
		retval = sctp_getsockopt_assoc_stats(sk, len, optval, optlen);
		break;
L
Linus Torvalds 已提交
5856 5857 5858
	default:
		retval = -ENOPROTOOPT;
		break;
5859
	}
L
Linus Torvalds 已提交
5860 5861 5862 5863 5864 5865 5866 5867 5868 5869 5870 5871 5872 5873 5874 5875 5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887

	sctp_release_sock(sk);
	return retval;
}

static void sctp_hash(struct sock *sk)
{
	/* STUB */
}

static void sctp_unhash(struct sock *sk)
{
	/* STUB */
}

/* Check if port is acceptable.  Possibly find first available port.
 *
 * The port hash table (contained in the 'global' SCTP protocol storage
 * returned by struct sctp_protocol *sctp_get_protocol()). The hash
 * table is an array of 4096 lists (sctp_bind_hashbucket). Each
 * list (the list number is the port number hashed out, so as you
 * would expect from a hash function, all the ports in a given list have
 * such a number that hashes out to the same list number; you were
 * expecting that, right?); so each list has a set of ports, with a
 * link to the socket (struct sock) that uses it, the port number and
 * a fastreuse flag (FIXME: NPI ipg).
 */
static struct sctp_bind_bucket *sctp_bucket_create(
5888
	struct sctp_bind_hashbucket *head, struct net *, unsigned short snum);
L
Linus Torvalds 已提交
5889 5890 5891 5892

static long sctp_get_port_local(struct sock *sk, union sctp_addr *addr)
{
	struct sctp_bind_hashbucket *head; /* hash list */
5893
	struct sctp_bind_bucket *pp;
L
Linus Torvalds 已提交
5894 5895 5896
	unsigned short snum;
	int ret;

A
Al Viro 已提交
5897
	snum = ntohs(addr->v4.sin_port);
L
Linus Torvalds 已提交
5898

5899 5900
	pr_debug("%s: begins, snum:%d\n", __func__, snum);

L
Linus Torvalds 已提交
5901 5902 5903
	sctp_local_bh_disable();

	if (snum == 0) {
S
Stephen Hemminger 已提交
5904
		/* Search for an available port. */
5905 5906 5907
		int low, high, remaining, index;
		unsigned int rover;

5908
		inet_get_local_port_range(sock_net(sk), &low, &high);
5909 5910
		remaining = (high - low) + 1;
		rover = net_random() % remaining + low;
L
Linus Torvalds 已提交
5911 5912 5913 5914 5915

		do {
			rover++;
			if ((rover < low) || (rover > high))
				rover = low;
5916 5917
			if (inet_is_reserved_local_port(rover))
				continue;
5918
			index = sctp_phashfn(sock_net(sk), rover);
L
Linus Torvalds 已提交
5919 5920
			head = &sctp_port_hashtable[index];
			sctp_spin_lock(&head->lock);
5921
			sctp_for_each_hentry(pp, &head->chain)
5922 5923
				if ((pp->port == rover) &&
				    net_eq(sock_net(sk), pp->net))
L
Linus Torvalds 已提交
5924 5925 5926 5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943 5944 5945 5946
					goto next;
			break;
		next:
			sctp_spin_unlock(&head->lock);
		} while (--remaining > 0);

		/* Exhausted local port range during search? */
		ret = 1;
		if (remaining <= 0)
			goto fail;

		/* OK, here is the one we will use.  HEAD (the port
		 * hash table list entry) is non-NULL and we hold it's
		 * mutex.
		 */
		snum = rover;
	} else {
		/* We are given an specific port number; we verify
		 * that it is not being used. If it is used, we will
		 * exahust the search in the hash list corresponding
		 * to the port number (snum) - we detect that with the
		 * port iterator, pp being NULL.
		 */
5947
		head = &sctp_port_hashtable[sctp_phashfn(sock_net(sk), snum)];
L
Linus Torvalds 已提交
5948
		sctp_spin_lock(&head->lock);
5949
		sctp_for_each_hentry(pp, &head->chain) {
5950
			if ((pp->port == snum) && net_eq(pp->net, sock_net(sk)))
L
Linus Torvalds 已提交
5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965
				goto pp_found;
		}
	}
	pp = NULL;
	goto pp_not_found;
pp_found:
	if (!hlist_empty(&pp->owner)) {
		/* We had a port hash table hit - there is an
		 * available port (pp != NULL) and it is being
		 * used by other socket (pp->owner not empty); that other
		 * socket is going to be sk2.
		 */
		int reuse = sk->sk_reuse;
		struct sock *sk2;

5966 5967
		pr_debug("%s: found a possible match\n", __func__);

5968 5969
		if (pp->fastreuse && sk->sk_reuse &&
			sk->sk_state != SCTP_SS_LISTENING)
L
Linus Torvalds 已提交
5970 5971 5972 5973 5974 5975 5976 5977 5978 5979 5980 5981
			goto success;

		/* Run through the list of sockets bound to the port
		 * (pp->port) [via the pointers bind_next and
		 * bind_pprev in the struct sock *sk2 (pp->sk)]. On each one,
		 * we get the endpoint they describe and run through
		 * the endpoint's list of IP (v4 or v6) addresses,
		 * comparing each of the addresses with the address of
		 * the socket sk. If we find a match, then that means
		 * that this port/socket (sk) combination are already
		 * in an endpoint.
		 */
5982
		sk_for_each_bound(sk2, &pp->owner) {
L
Linus Torvalds 已提交
5983 5984 5985
			struct sctp_endpoint *ep2;
			ep2 = sctp_sk(sk2)->ep;

5986 5987 5988
			if (sk == sk2 ||
			    (reuse && sk2->sk_reuse &&
			     sk2->sk_state != SCTP_SS_LISTENING))
L
Linus Torvalds 已提交
5989 5990
				continue;

5991 5992
			if (sctp_bind_addr_conflict(&ep2->base.bind_addr, addr,
						 sctp_sk(sk2), sctp_sk(sk))) {
L
Linus Torvalds 已提交
5993 5994 5995 5996
				ret = (long)sk2;
				goto fail_unlock;
			}
		}
5997 5998

		pr_debug("%s: found a match\n", __func__);
L
Linus Torvalds 已提交
5999 6000 6001 6002
	}
pp_not_found:
	/* If there was a hash table miss, create a new port.  */
	ret = 1;
6003
	if (!pp && !(pp = sctp_bucket_create(head, sock_net(sk), snum)))
L
Linus Torvalds 已提交
6004 6005 6006 6007 6008 6009
		goto fail_unlock;

	/* In either case (hit or miss), make sure fastreuse is 1 only
	 * if sk->sk_reuse is too (that is, if the caller requested
	 * SO_REUSEADDR on this socket -sk-).
	 */
6010 6011 6012 6013 6014 6015 6016
	if (hlist_empty(&pp->owner)) {
		if (sk->sk_reuse && sk->sk_state != SCTP_SS_LISTENING)
			pp->fastreuse = 1;
		else
			pp->fastreuse = 0;
	} else if (pp->fastreuse &&
		(!sk->sk_reuse || sk->sk_state == SCTP_SS_LISTENING))
L
Linus Torvalds 已提交
6017 6018 6019 6020 6021 6022 6023 6024
		pp->fastreuse = 0;

	/* We are set, so fill up all the data in the hash table
	 * entry, tie the socket list information with the rest of the
	 * sockets FIXME: Blurry, NPI (ipg).
	 */
success:
	if (!sctp_sk(sk)->bind_hash) {
E
Eric Dumazet 已提交
6025
		inet_sk(sk)->inet_num = snum;
L
Linus Torvalds 已提交
6026 6027 6028 6029 6030 6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043 6044 6045 6046 6047 6048 6049 6050 6051
		sk_add_bind_node(sk, &pp->owner);
		sctp_sk(sk)->bind_hash = pp;
	}
	ret = 0;

fail_unlock:
	sctp_spin_unlock(&head->lock);

fail:
	sctp_local_bh_enable();
	return ret;
}

/* Assign a 'snum' port to the socket.  If snum == 0, an ephemeral
 * port is requested.
 */
static int sctp_get_port(struct sock *sk, unsigned short snum)
{
	union sctp_addr addr;
	struct sctp_af *af = sctp_sk(sk)->pf->af;

	/* Set up a dummy address struct from the sk. */
	af->from_sk(&addr, sk);
	addr.v4.sin_port = htons(snum);

	/* Note: sk->sk_num gets filled in if ephemeral port request. */
6052
	return !!sctp_get_port_local(sk, &addr);
L
Linus Torvalds 已提交
6053 6054 6055
}

/*
6056
 *  Move a socket to LISTENING state.
L
Linus Torvalds 已提交
6057
 */
6058
static int sctp_listen_start(struct sock *sk, int backlog)
L
Linus Torvalds 已提交
6059 6060 6061
{
	struct sctp_sock *sp = sctp_sk(sk);
	struct sctp_endpoint *ep = sp->ep;
6062
	struct crypto_hash *tfm = NULL;
6063
	char alg[32];
L
Linus Torvalds 已提交
6064

6065
	/* Allocate HMAC for generating cookie. */
6066 6067 6068
	if (!sp->hmac && sp->sctp_hmac_alg) {
		sprintf(alg, "hmac(%s)", sp->sctp_hmac_alg);
		tfm = crypto_alloc_hash(alg, 0, CRYPTO_ALG_ASYNC);
6069
		if (IS_ERR(tfm)) {
6070
			net_info_ratelimited("failed to load transform for %s: %ld\n",
6071
					     sp->sctp_hmac_alg, PTR_ERR(tfm));
6072 6073 6074
			return -ENOSYS;
		}
		sctp_sk(sk)->hmac = tfm;
L
Linus Torvalds 已提交
6075 6076 6077 6078 6079 6080 6081 6082 6083 6084 6085
	}

	/*
	 * If a bind() or sctp_bindx() is not called prior to a listen()
	 * call that allows new associations to be accepted, the system
	 * picks an ephemeral port and will choose an address set equivalent
	 * to binding with a wildcard address.
	 *
	 * This is not currently spelled out in the SCTP sockets
	 * extensions draft, but follows the practice as seen in TCP
	 * sockets.
6086
	 *
L
Linus Torvalds 已提交
6087
	 */
6088
	sk->sk_state = SCTP_SS_LISTENING;
L
Linus Torvalds 已提交
6089 6090 6091
	if (!ep->base.bind_addr.port) {
		if (sctp_autobind(sk))
			return -EAGAIN;
6092
	} else {
E
Eric Dumazet 已提交
6093
		if (sctp_get_port(sk, inet_sk(sk)->inet_num)) {
6094 6095 6096 6097
			sk->sk_state = SCTP_SS_CLOSED;
			return -EADDRINUSE;
		}
	}
6098

L
Linus Torvalds 已提交
6099 6100 6101 6102 6103 6104
	sk->sk_max_ack_backlog = backlog;
	sctp_hash_endpoint(ep);
	return 0;
}

/*
6105 6106 6107 6108 6109 6110 6111 6112 6113 6114 6115
 * 4.1.3 / 5.1.3 listen()
 *
 *   By default, new associations are not accepted for UDP style sockets.
 *   An application uses listen() to mark a socket as being able to
 *   accept new associations.
 *
 *   On TCP style sockets, applications use listen() to ready the SCTP
 *   endpoint for accepting inbound associations.
 *
 *   On both types of endpoints a backlog of '0' disables listening.
 *
L
Linus Torvalds 已提交
6116 6117 6118 6119 6120
 *  Move a socket to LISTENING state.
 */
int sctp_inet_listen(struct socket *sock, int backlog)
{
	struct sock *sk = sock->sk;
6121
	struct sctp_endpoint *ep = sctp_sk(sk)->ep;
L
Linus Torvalds 已提交
6122 6123 6124
	int err = -EINVAL;

	if (unlikely(backlog < 0))
6125
		return err;
L
Linus Torvalds 已提交
6126 6127 6128

	sctp_lock_sock(sk);

6129 6130 6131 6132
	/* Peeled-off sockets are not allowed to listen().  */
	if (sctp_style(sk, UDP_HIGH_BANDWIDTH))
		goto out;

L
Linus Torvalds 已提交
6133 6134 6135
	if (sock->state != SS_UNCONNECTED)
		goto out;

6136 6137 6138
	/* If backlog is zero, disable listening. */
	if (!backlog) {
		if (sctp_sstate(sk, CLOSED))
L
Linus Torvalds 已提交
6139 6140
			goto out;

6141 6142 6143 6144 6145 6146
		err = 0;
		sctp_unhash_endpoint(ep);
		sk->sk_state = SCTP_SS_CLOSED;
		if (sk->sk_reuse)
			sctp_sk(sk)->bind_hash->fastreuse = 1;
		goto out;
6147 6148
	}

6149 6150 6151 6152 6153 6154 6155 6156
	/* If we are already listening, just update the backlog */
	if (sctp_sstate(sk, LISTENING))
		sk->sk_max_ack_backlog = backlog;
	else {
		err = sctp_listen_start(sk, backlog);
		if (err)
			goto out;
	}
L
Linus Torvalds 已提交
6157

6158
	err = 0;
L
Linus Torvalds 已提交
6159 6160 6161 6162 6163 6164 6165 6166 6167 6168
out:
	sctp_release_sock(sk);
	return err;
}

/*
 * This function is done by modeling the current datagram_poll() and the
 * tcp_poll().  Note that, based on these implementations, we don't
 * lock the socket in this function, even though it seems that,
 * ideally, locking or some other mechanisms can be used to ensure
6169
 * the integrity of the counters (sndbuf and wmem_alloc) used
L
Linus Torvalds 已提交
6170 6171 6172 6173 6174 6175 6176 6177 6178 6179 6180 6181 6182
 * in this place.  We assume that we don't need locks either until proven
 * otherwise.
 *
 * Another thing to note is that we include the Async I/O support
 * here, again, by modeling the current TCP/UDP code.  We don't have
 * a good way to test with it yet.
 */
unsigned int sctp_poll(struct file *file, struct socket *sock, poll_table *wait)
{
	struct sock *sk = sock->sk;
	struct sctp_sock *sp = sctp_sk(sk);
	unsigned int mask;

E
Eric Dumazet 已提交
6183
	poll_wait(file, sk_sleep(sk), wait);
L
Linus Torvalds 已提交
6184 6185 6186 6187 6188 6189

	/* A TCP-style listening socket becomes readable when the accept queue
	 * is not empty.
	 */
	if (sctp_style(sk, TCP) && sctp_sstate(sk, LISTENING))
		return (!list_empty(&sp->ep->asocs)) ?
6190
			(POLLIN | POLLRDNORM) : 0;
L
Linus Torvalds 已提交
6191 6192 6193 6194 6195

	mask = 0;

	/* Is there any exceptional events?  */
	if (sk->sk_err || !skb_queue_empty(&sk->sk_error_queue))
6196
		mask |= POLLERR |
6197
			(sock_flag(sk, SOCK_SELECT_ERR_QUEUE) ? POLLPRI : 0);
6198
	if (sk->sk_shutdown & RCV_SHUTDOWN)
E
Eric Dumazet 已提交
6199
		mask |= POLLRDHUP | POLLIN | POLLRDNORM;
L
Linus Torvalds 已提交
6200 6201 6202 6203
	if (sk->sk_shutdown == SHUTDOWN_MASK)
		mask |= POLLHUP;

	/* Is it readable?  Reconsider this code with TCP-style support.  */
E
Eric Dumazet 已提交
6204
	if (!skb_queue_empty(&sk->sk_receive_queue))
L
Linus Torvalds 已提交
6205 6206 6207 6208 6209 6210 6211 6212 6213 6214 6215 6216 6217 6218 6219 6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231 6232 6233 6234
		mask |= POLLIN | POLLRDNORM;

	/* The association is either gone or not ready.  */
	if (!sctp_style(sk, UDP) && sctp_sstate(sk, CLOSED))
		return mask;

	/* Is it writable?  */
	if (sctp_writeable(sk)) {
		mask |= POLLOUT | POLLWRNORM;
	} else {
		set_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags);
		/*
		 * Since the socket is not locked, the buffer
		 * might be made available after the writeable check and
		 * before the bit is set.  This could cause a lost I/O
		 * signal.  tcp_poll() has a race breaker for this race
		 * condition.  Based on their implementation, we put
		 * in the following code to cover it as well.
		 */
		if (sctp_writeable(sk))
			mask |= POLLOUT | POLLWRNORM;
	}
	return mask;
}

/********************************************************************
 * 2nd Level Abstractions
 ********************************************************************/

static struct sctp_bind_bucket *sctp_bucket_create(
6235
	struct sctp_bind_hashbucket *head, struct net *net, unsigned short snum)
L
Linus Torvalds 已提交
6236 6237 6238
{
	struct sctp_bind_bucket *pp;

6239
	pp = kmem_cache_alloc(sctp_bucket_cachep, GFP_ATOMIC);
L
Linus Torvalds 已提交
6240
	if (pp) {
6241
		SCTP_DBG_OBJCNT_INC(bind_bucket);
L
Linus Torvalds 已提交
6242 6243 6244
		pp->port = snum;
		pp->fastreuse = 0;
		INIT_HLIST_HEAD(&pp->owner);
6245
		pp->net = net;
6246
		hlist_add_head(&pp->node, &head->chain);
L
Linus Torvalds 已提交
6247 6248 6249 6250 6251 6252 6253
	}
	return pp;
}

/* Caller must hold hashbucket lock for this tb with local BH disabled */
static void sctp_bucket_destroy(struct sctp_bind_bucket *pp)
{
6254
	if (pp && hlist_empty(&pp->owner)) {
6255
		__hlist_del(&pp->node);
L
Linus Torvalds 已提交
6256 6257 6258 6259 6260 6261 6262 6263 6264
		kmem_cache_free(sctp_bucket_cachep, pp);
		SCTP_DBG_OBJCNT_DEC(bind_bucket);
	}
}

/* Release this socket's reference to a local port.  */
static inline void __sctp_put_port(struct sock *sk)
{
	struct sctp_bind_hashbucket *head =
6265 6266
		&sctp_port_hashtable[sctp_phashfn(sock_net(sk),
						  inet_sk(sk)->inet_num)];
L
Linus Torvalds 已提交
6267 6268 6269 6270 6271 6272
	struct sctp_bind_bucket *pp;

	sctp_spin_lock(&head->lock);
	pp = sctp_sk(sk)->bind_hash;
	__sk_del_bind_node(sk);
	sctp_sk(sk)->bind_hash = NULL;
E
Eric Dumazet 已提交
6273
	inet_sk(sk)->inet_num = 0;
L
Linus Torvalds 已提交
6274 6275 6276 6277 6278 6279 6280 6281 6282 6283 6284 6285 6286 6287 6288 6289 6290 6291 6292 6293 6294
	sctp_bucket_destroy(pp);
	sctp_spin_unlock(&head->lock);
}

void sctp_put_port(struct sock *sk)
{
	sctp_local_bh_disable();
	__sctp_put_port(sk);
	sctp_local_bh_enable();
}

/*
 * The system picks an ephemeral port and choose an address set equivalent
 * to binding with a wildcard address.
 * One of those addresses will be the primary address for the association.
 * This automatically enables the multihoming capability of SCTP.
 */
static int sctp_autobind(struct sock *sk)
{
	union sctp_addr autoaddr;
	struct sctp_af *af;
A
Al Viro 已提交
6295
	__be16 port;
L
Linus Torvalds 已提交
6296 6297 6298 6299

	/* Initialize a local sockaddr structure to INADDR_ANY. */
	af = sctp_sk(sk)->pf->af;

E
Eric Dumazet 已提交
6300
	port = htons(inet_sk(sk)->inet_num);
L
Linus Torvalds 已提交
6301 6302 6303 6304 6305 6306 6307 6308 6309 6310 6311 6312 6313 6314 6315 6316 6317 6318 6319 6320 6321 6322 6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334 6335 6336 6337 6338 6339 6340 6341 6342 6343 6344
	af->inaddr_any(&autoaddr, port);

	return sctp_do_bind(sk, &autoaddr, af->sockaddr_len);
}

/* Parse out IPPROTO_SCTP CMSG headers.  Perform only minimal validation.
 *
 * From RFC 2292
 * 4.2 The cmsghdr Structure *
 *
 * When ancillary data is sent or received, any number of ancillary data
 * objects can be specified by the msg_control and msg_controllen members of
 * the msghdr structure, because each object is preceded by
 * a cmsghdr structure defining the object's length (the cmsg_len member).
 * Historically Berkeley-derived implementations have passed only one object
 * at a time, but this API allows multiple objects to be
 * passed in a single call to sendmsg() or recvmsg(). The following example
 * shows two ancillary data objects in a control buffer.
 *
 *   |<--------------------------- msg_controllen -------------------------->|
 *   |                                                                       |
 *
 *   |<----- ancillary data object ----->|<----- ancillary data object ----->|
 *
 *   |<---------- CMSG_SPACE() --------->|<---------- CMSG_SPACE() --------->|
 *   |                                   |                                   |
 *
 *   |<---------- cmsg_len ---------->|  |<--------- cmsg_len ----------->|  |
 *
 *   |<--------- CMSG_LEN() --------->|  |<-------- CMSG_LEN() ---------->|  |
 *   |                                |  |                                |  |
 *
 *   +-----+-----+-----+--+-----------+--+-----+-----+-----+--+-----------+--+
 *   |cmsg_|cmsg_|cmsg_|XX|           |XX|cmsg_|cmsg_|cmsg_|XX|           |XX|
 *
 *   |len  |level|type |XX|cmsg_data[]|XX|len  |level|type |XX|cmsg_data[]|XX|
 *
 *   +-----+-----+-----+--+-----------+--+-----+-----+-----+--+-----------+--+
 *    ^
 *    |
 *
 * msg_control
 * points here
 */
6345
static int sctp_msghdr_parse(const struct msghdr *msg, sctp_cmsgs_t *cmsgs)
L
Linus Torvalds 已提交
6346 6347
{
	struct cmsghdr *cmsg;
6348
	struct msghdr *my_msg = (struct msghdr *)msg;
L
Linus Torvalds 已提交
6349 6350 6351

	for (cmsg = CMSG_FIRSTHDR(msg);
	     cmsg != NULL;
6352 6353
	     cmsg = CMSG_NXTHDR(my_msg, cmsg)) {
		if (!CMSG_OK(my_msg, cmsg))
L
Linus Torvalds 已提交
6354 6355 6356 6357 6358 6359 6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373 6374 6375 6376 6377 6378 6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389 6390 6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401 6402
			return -EINVAL;

		/* Should we parse this header or ignore?  */
		if (cmsg->cmsg_level != IPPROTO_SCTP)
			continue;

		/* Strictly check lengths following example in SCM code.  */
		switch (cmsg->cmsg_type) {
		case SCTP_INIT:
			/* SCTP Socket API Extension
			 * 5.2.1 SCTP Initiation Structure (SCTP_INIT)
			 *
			 * This cmsghdr structure provides information for
			 * initializing new SCTP associations with sendmsg().
			 * The SCTP_INITMSG socket option uses this same data
			 * structure.  This structure is not used for
			 * recvmsg().
			 *
			 * cmsg_level    cmsg_type      cmsg_data[]
			 * ------------  ------------   ----------------------
			 * IPPROTO_SCTP  SCTP_INIT      struct sctp_initmsg
			 */
			if (cmsg->cmsg_len !=
			    CMSG_LEN(sizeof(struct sctp_initmsg)))
				return -EINVAL;
			cmsgs->init = (struct sctp_initmsg *)CMSG_DATA(cmsg);
			break;

		case SCTP_SNDRCV:
			/* SCTP Socket API Extension
			 * 5.2.2 SCTP Header Information Structure(SCTP_SNDRCV)
			 *
			 * This cmsghdr structure specifies SCTP options for
			 * sendmsg() and describes SCTP header information
			 * about a received message through recvmsg().
			 *
			 * cmsg_level    cmsg_type      cmsg_data[]
			 * ------------  ------------   ----------------------
			 * IPPROTO_SCTP  SCTP_SNDRCV    struct sctp_sndrcvinfo
			 */
			if (cmsg->cmsg_len !=
			    CMSG_LEN(sizeof(struct sctp_sndrcvinfo)))
				return -EINVAL;

			cmsgs->info =
				(struct sctp_sndrcvinfo *)CMSG_DATA(cmsg);

			/* Minimally, validate the sinfo_flags. */
			if (cmsgs->info->sinfo_flags &
6403 6404
			    ~(SCTP_UNORDERED | SCTP_ADDR_OVER |
			      SCTP_ABORT | SCTP_EOF))
L
Linus Torvalds 已提交
6405 6406 6407 6408 6409
				return -EINVAL;
			break;

		default:
			return -EINVAL;
6410
		}
L
Linus Torvalds 已提交
6411 6412 6413 6414 6415 6416 6417 6418 6419 6420 6421 6422 6423 6424
	}
	return 0;
}

/*
 * Wait for a packet..
 * Note: This function is the same function as in core/datagram.c
 * with a few modifications to make lksctp work.
 */
static int sctp_wait_for_packet(struct sock * sk, int *err, long *timeo_p)
{
	int error;
	DEFINE_WAIT(wait);

E
Eric Dumazet 已提交
6425
	prepare_to_wait_exclusive(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
L
Linus Torvalds 已提交
6426 6427 6428 6429 6430 6431 6432 6433 6434 6435 6436 6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460 6461

	/* Socket errors? */
	error = sock_error(sk);
	if (error)
		goto out;

	if (!skb_queue_empty(&sk->sk_receive_queue))
		goto ready;

	/* Socket shut down?  */
	if (sk->sk_shutdown & RCV_SHUTDOWN)
		goto out;

	/* Sequenced packets can come disconnected.  If so we report the
	 * problem.
	 */
	error = -ENOTCONN;

	/* Is there a good reason to think that we may receive some data?  */
	if (list_empty(&sctp_sk(sk)->ep->asocs) && !sctp_sstate(sk, LISTENING))
		goto out;

	/* Handle signals.  */
	if (signal_pending(current))
		goto interrupted;

	/* Let another process have a go.  Since we are going to sleep
	 * anyway.  Note: This may cause odd behaviors if the message
	 * does not fit in the user's buffer, but this seems to be the
	 * only way to honor MSG_DONTWAIT realistically.
	 */
	sctp_release_sock(sk);
	*timeo_p = schedule_timeout(*timeo_p);
	sctp_lock_sock(sk);

ready:
E
Eric Dumazet 已提交
6462
	finish_wait(sk_sleep(sk), &wait);
L
Linus Torvalds 已提交
6463 6464 6465 6466 6467 6468
	return 0;

interrupted:
	error = sock_intr_errno(*timeo_p);

out:
E
Eric Dumazet 已提交
6469
	finish_wait(sk_sleep(sk), &wait);
L
Linus Torvalds 已提交
6470 6471 6472 6473 6474 6475 6476 6477 6478 6479 6480 6481 6482 6483 6484 6485 6486
	*err = error;
	return error;
}

/* Receive a datagram.
 * Note: This is pretty much the same routine as in core/datagram.c
 * with a few changes to make lksctp work.
 */
static struct sk_buff *sctp_skb_recv_datagram(struct sock *sk, int flags,
					      int noblock, int *err)
{
	int error;
	struct sk_buff *skb;
	long timeo;

	timeo = sock_rcvtimeo(sk, noblock);

6487 6488
	pr_debug("%s: timeo:%ld, max:%ld\n", __func__, timeo,
		 MAX_SCHEDULE_TIMEOUT);
L
Linus Torvalds 已提交
6489 6490 6491 6492 6493 6494 6495

	do {
		/* Again only user level code calls this function,
		 * so nothing interrupt level
		 * will suddenly eat the receive_queue.
		 *
		 *  Look at current nfs client by the way...
6496
		 *  However, this function was correct in any case. 8)
L
Linus Torvalds 已提交
6497 6498
		 */
		if (flags & MSG_PEEK) {
6499
			spin_lock_bh(&sk->sk_receive_queue.lock);
L
Linus Torvalds 已提交
6500 6501 6502
			skb = skb_peek(&sk->sk_receive_queue);
			if (skb)
				atomic_inc(&skb->users);
6503
			spin_unlock_bh(&sk->sk_receive_queue.lock);
L
Linus Torvalds 已提交
6504 6505 6506 6507 6508 6509 6510
		} else {
			skb = skb_dequeue(&sk->sk_receive_queue);
		}

		if (skb)
			return skb;

6511 6512 6513 6514 6515
		/* Caller is allowed not to check sk->sk_err before calling. */
		error = sock_error(sk);
		if (error)
			goto no_packet;

L
Linus Torvalds 已提交
6516 6517 6518 6519 6520 6521 6522 6523 6524 6525 6526 6527 6528 6529 6530 6531 6532 6533 6534 6535 6536 6537 6538 6539 6540 6541 6542
		if (sk->sk_shutdown & RCV_SHUTDOWN)
			break;

		/* User doesn't want to wait.  */
		error = -EAGAIN;
		if (!timeo)
			goto no_packet;
	} while (sctp_wait_for_packet(sk, err, &timeo) == 0);

	return NULL;

no_packet:
	*err = error;
	return NULL;
}

/* If sndbuf has changed, wake up per association sndbuf waiters.  */
static void __sctp_write_space(struct sctp_association *asoc)
{
	struct sock *sk = asoc->base.sk;
	struct socket *sock = sk->sk_socket;

	if ((sctp_wspace(asoc) > 0) && sock) {
		if (waitqueue_active(&asoc->wait))
			wake_up_interruptible(&asoc->wait);

		if (sctp_writeable(sk)) {
6543 6544 6545 6546
			wait_queue_head_t *wq = sk_sleep(sk);

			if (wq && waitqueue_active(wq))
				wake_up_interruptible(wq);
L
Linus Torvalds 已提交
6547 6548 6549 6550 6551

			/* Note that we try to include the Async I/O support
			 * here by modeling from the current TCP/UDP code.
			 * We have not tested with it yet.
			 */
6552
			if (!(sk->sk_shutdown & SEND_SHUTDOWN))
6553 6554
				sock_wake_async(sock,
						SOCK_WAKE_SPACE, POLL_OUT);
L
Linus Torvalds 已提交
6555 6556 6557 6558 6559 6560 6561 6562 6563 6564 6565 6566 6567 6568 6569 6570 6571 6572
		}
	}
}

/* Do accounting for the sndbuf space.
 * Decrement the used sndbuf space of the corresponding association by the
 * data size which was just transmitted(freed).
 */
static void sctp_wfree(struct sk_buff *skb)
{
	struct sctp_association *asoc;
	struct sctp_chunk *chunk;
	struct sock *sk;

	/* Get the saved chunk pointer.  */
	chunk = *((struct sctp_chunk **)(skb->cb));
	asoc = chunk->asoc;
	sk = asoc->base.sk;
6573 6574 6575 6576 6577 6578
	asoc->sndbuf_used -= SCTP_DATA_SNDSIZE(chunk) +
				sizeof(struct sk_buff) +
				sizeof(struct sctp_chunk);

	atomic_sub(sizeof(struct sctp_chunk), &sk->sk_wmem_alloc);

6579
	/*
6580
	 * This undoes what is done via sctp_set_owner_w and sk_mem_charge
6581 6582
	 */
	sk->sk_wmem_queued   -= skb->truesize;
6583
	sk_mem_uncharge(sk, skb->truesize);
6584

6585
	sock_wfree(skb);
L
Linus Torvalds 已提交
6586 6587 6588 6589 6590
	__sctp_write_space(asoc);

	sctp_association_put(asoc);
}

6591 6592 6593 6594 6595 6596 6597 6598 6599 6600 6601
/* Do accounting for the receive space on the socket.
 * Accounting for the association is done in ulpevent.c
 * We set this as a destructor for the cloned data skbs so that
 * accounting is done at the correct time.
 */
void sctp_sock_rfree(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;
	struct sctp_ulpevent *event = sctp_skb2event(skb);

	atomic_sub(event->rmem_len, &sk->sk_rmem_alloc);
6602 6603

	/*
6604
	 * Mimic the behavior of sock_rfree
6605
	 */
6606
	sk_mem_uncharge(sk, event->rmem_len);
6607 6608 6609
}


L
Linus Torvalds 已提交
6610 6611 6612 6613 6614 6615 6616 6617 6618
/* Helper function to wait for space in the sndbuf.  */
static int sctp_wait_for_sndbuf(struct sctp_association *asoc, long *timeo_p,
				size_t msg_len)
{
	struct sock *sk = asoc->base.sk;
	int err = 0;
	long current_timeo = *timeo_p;
	DEFINE_WAIT(wait);

6619 6620
	pr_debug("%s: asoc:%p, timeo:%ld, msg_len:%zu\n", __func__, asoc,
		 *timeo_p, msg_len);
L
Linus Torvalds 已提交
6621 6622 6623 6624 6625 6626 6627 6628 6629 6630 6631 6632 6633 6634 6635 6636 6637 6638 6639 6640 6641 6642 6643

	/* Increment the association's refcnt.  */
	sctp_association_hold(asoc);

	/* Wait on the association specific sndbuf space. */
	for (;;) {
		prepare_to_wait_exclusive(&asoc->wait, &wait,
					  TASK_INTERRUPTIBLE);
		if (!*timeo_p)
			goto do_nonblock;
		if (sk->sk_err || asoc->state >= SCTP_STATE_SHUTDOWN_PENDING ||
		    asoc->base.dead)
			goto do_error;
		if (signal_pending(current))
			goto do_interrupted;
		if (msg_len <= sctp_wspace(asoc))
			break;

		/* Let another process have a go.  Since we are going
		 * to sleep anyway.
		 */
		sctp_release_sock(sk);
		current_timeo = schedule_timeout(current_timeo);
6644
		BUG_ON(sk != asoc->base.sk);
L
Linus Torvalds 已提交
6645 6646 6647 6648 6649 6650 6651 6652 6653 6654 6655 6656 6657 6658 6659 6660 6661 6662 6663 6664 6665 6666 6667 6668 6669 6670
		sctp_lock_sock(sk);

		*timeo_p = current_timeo;
	}

out:
	finish_wait(&asoc->wait, &wait);

	/* Release the association's refcnt.  */
	sctp_association_put(asoc);

	return err;

do_error:
	err = -EPIPE;
	goto out;

do_interrupted:
	err = sock_intr_errno(*timeo_p);
	goto out;

do_nonblock:
	err = -EAGAIN;
	goto out;
}

6671 6672
void sctp_data_ready(struct sock *sk, int len)
{
6673 6674 6675 6676 6677 6678
	struct socket_wq *wq;

	rcu_read_lock();
	wq = rcu_dereference(sk->sk_wq);
	if (wq_has_sleeper(wq))
		wake_up_interruptible_sync_poll(&wq->wait, POLLIN |
6679 6680
						POLLRDNORM | POLLRDBAND);
	sk_wake_async(sk, SOCK_WAKE_WAITD, POLL_IN);
6681
	rcu_read_unlock();
6682 6683
}

L
Linus Torvalds 已提交
6684 6685 6686 6687 6688 6689
/* If socket sndbuf has changed, wake up all per association waiters.  */
void sctp_write_space(struct sock *sk)
{
	struct sctp_association *asoc;

	/* Wake up the tasks in each wait queue.  */
6690
	list_for_each_entry(asoc, &((sctp_sk(sk))->ep->asocs), asocs) {
L
Linus Torvalds 已提交
6691 6692 6693 6694 6695 6696
		__sctp_write_space(asoc);
	}
}

/* Is there any sndbuf space available on the socket?
 *
6697
 * Note that sk_wmem_alloc is the sum of the send buffers on all of the
L
Linus Torvalds 已提交
6698 6699 6700 6701 6702 6703 6704 6705 6706 6707 6708 6709
 * associations on the same socket.  For a UDP-style socket with
 * multiple associations, it is possible for it to be "unwriteable"
 * prematurely.  I assume that this is acceptable because
 * a premature "unwriteable" is better than an accidental "writeable" which
 * would cause an unwanted block under certain circumstances.  For the 1-1
 * UDP-style sockets or TCP-style sockets, this code should work.
 *  - Daisy
 */
static int sctp_writeable(struct sock *sk)
{
	int amt = 0;

6710
	amt = sk->sk_sndbuf - sk_wmem_alloc_get(sk);
L
Linus Torvalds 已提交
6711 6712 6713 6714 6715 6716 6717 6718 6719 6720 6721 6722 6723 6724 6725
	if (amt < 0)
		amt = 0;
	return amt;
}

/* Wait for an association to go into ESTABLISHED state. If timeout is 0,
 * returns immediately with EINPROGRESS.
 */
static int sctp_wait_for_connect(struct sctp_association *asoc, long *timeo_p)
{
	struct sock *sk = asoc->base.sk;
	int err = 0;
	long current_timeo = *timeo_p;
	DEFINE_WAIT(wait);

6726
	pr_debug("%s: asoc:%p, timeo:%ld\n", __func__, asoc, *timeo_p);
L
Linus Torvalds 已提交
6727 6728 6729 6730 6731 6732 6733 6734 6735 6736 6737 6738 6739 6740 6741 6742 6743 6744 6745 6746 6747 6748 6749 6750 6751 6752 6753 6754 6755 6756 6757 6758 6759 6760 6761 6762 6763 6764 6765

	/* Increment the association's refcnt.  */
	sctp_association_hold(asoc);

	for (;;) {
		prepare_to_wait_exclusive(&asoc->wait, &wait,
					  TASK_INTERRUPTIBLE);
		if (!*timeo_p)
			goto do_nonblock;
		if (sk->sk_shutdown & RCV_SHUTDOWN)
			break;
		if (sk->sk_err || asoc->state >= SCTP_STATE_SHUTDOWN_PENDING ||
		    asoc->base.dead)
			goto do_error;
		if (signal_pending(current))
			goto do_interrupted;

		if (sctp_state(asoc, ESTABLISHED))
			break;

		/* Let another process have a go.  Since we are going
		 * to sleep anyway.
		 */
		sctp_release_sock(sk);
		current_timeo = schedule_timeout(current_timeo);
		sctp_lock_sock(sk);

		*timeo_p = current_timeo;
	}

out:
	finish_wait(&asoc->wait, &wait);

	/* Release the association's refcnt.  */
	sctp_association_put(asoc);

	return err;

do_error:
6766
	if (asoc->init_err_counter + 1 > asoc->max_init_attempts)
L
Linus Torvalds 已提交
6767 6768 6769 6770 6771 6772 6773 6774 6775 6776 6777 6778 6779 6780 6781 6782 6783 6784 6785 6786 6787 6788 6789 6790
		err = -ETIMEDOUT;
	else
		err = -ECONNREFUSED;
	goto out;

do_interrupted:
	err = sock_intr_errno(*timeo_p);
	goto out;

do_nonblock:
	err = -EINPROGRESS;
	goto out;
}

static int sctp_wait_for_accept(struct sock *sk, long timeo)
{
	struct sctp_endpoint *ep;
	int err = 0;
	DEFINE_WAIT(wait);

	ep = sctp_sk(sk)->ep;


	for (;;) {
E
Eric Dumazet 已提交
6791
		prepare_to_wait_exclusive(sk_sleep(sk), &wait,
L
Linus Torvalds 已提交
6792 6793 6794 6795 6796 6797 6798 6799 6800 6801 6802 6803 6804 6805 6806 6807 6808 6809 6810 6811 6812 6813 6814 6815 6816
					  TASK_INTERRUPTIBLE);

		if (list_empty(&ep->asocs)) {
			sctp_release_sock(sk);
			timeo = schedule_timeout(timeo);
			sctp_lock_sock(sk);
		}

		err = -EINVAL;
		if (!sctp_sstate(sk, LISTENING))
			break;

		err = 0;
		if (!list_empty(&ep->asocs))
			break;

		err = sock_intr_errno(timeo);
		if (signal_pending(current))
			break;

		err = -EAGAIN;
		if (!timeo)
			break;
	}

E
Eric Dumazet 已提交
6817
	finish_wait(sk_sleep(sk), &wait);
L
Linus Torvalds 已提交
6818 6819 6820 6821

	return err;
}

6822
static void sctp_wait_for_close(struct sock *sk, long timeout)
L
Linus Torvalds 已提交
6823 6824 6825 6826
{
	DEFINE_WAIT(wait);

	do {
E
Eric Dumazet 已提交
6827
		prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
L
Linus Torvalds 已提交
6828 6829 6830 6831 6832 6833 6834
		if (list_empty(&sctp_sk(sk)->ep->asocs))
			break;
		sctp_release_sock(sk);
		timeout = schedule_timeout(timeout);
		sctp_lock_sock(sk);
	} while (!signal_pending(current) && timeout);

E
Eric Dumazet 已提交
6835
	finish_wait(sk_sleep(sk), &wait);
L
Linus Torvalds 已提交
6836 6837
}

6838 6839 6840 6841 6842 6843 6844 6845
static void sctp_skb_set_owner_r_frag(struct sk_buff *skb, struct sock *sk)
{
	struct sk_buff *frag;

	if (!skb->data_len)
		goto done;

	/* Don't forget the fragments. */
6846
	skb_walk_frags(skb, frag)
6847 6848 6849 6850 6851 6852
		sctp_skb_set_owner_r_frag(frag, sk);

done:
	sctp_skb_set_owner_r(skb, sk);
}

6853 6854 6855 6856
void sctp_copy_sock(struct sock *newsk, struct sock *sk,
		    struct sctp_association *asoc)
{
	struct inet_sock *inet = inet_sk(sk);
J
Julia Lawall 已提交
6857
	struct inet_sock *newinet;
6858 6859 6860 6861 6862 6863 6864 6865

	newsk->sk_type = sk->sk_type;
	newsk->sk_bound_dev_if = sk->sk_bound_dev_if;
	newsk->sk_flags = sk->sk_flags;
	newsk->sk_no_check = sk->sk_no_check;
	newsk->sk_reuse = sk->sk_reuse;

	newsk->sk_shutdown = sk->sk_shutdown;
6866
	newsk->sk_destruct = sctp_destruct_sock;
6867 6868 6869 6870 6871 6872 6873 6874 6875 6876 6877 6878 6879 6880
	newsk->sk_family = sk->sk_family;
	newsk->sk_protocol = IPPROTO_SCTP;
	newsk->sk_backlog_rcv = sk->sk_prot->backlog_rcv;
	newsk->sk_sndbuf = sk->sk_sndbuf;
	newsk->sk_rcvbuf = sk->sk_rcvbuf;
	newsk->sk_lingertime = sk->sk_lingertime;
	newsk->sk_rcvtimeo = sk->sk_rcvtimeo;
	newsk->sk_sndtimeo = sk->sk_sndtimeo;

	newinet = inet_sk(newsk);

	/* Initialize sk's sport, dport, rcv_saddr and daddr for
	 * getsockname() and getpeername()
	 */
E
Eric Dumazet 已提交
6881 6882 6883 6884
	newinet->inet_sport = inet->inet_sport;
	newinet->inet_saddr = inet->inet_saddr;
	newinet->inet_rcv_saddr = inet->inet_rcv_saddr;
	newinet->inet_dport = htons(asoc->peer.port);
6885
	newinet->pmtudisc = inet->pmtudisc;
E
Eric Dumazet 已提交
6886
	newinet->inet_id = asoc->next_tsn ^ jiffies;
6887 6888 6889 6890 6891 6892 6893 6894

	newinet->uc_ttl = inet->uc_ttl;
	newinet->mc_loop = 1;
	newinet->mc_ttl = 1;
	newinet->mc_index = 0;
	newinet->mc_list = NULL;
}

L
Linus Torvalds 已提交
6895 6896 6897 6898 6899 6900 6901 6902 6903 6904 6905 6906 6907
/* Populate the fields of the newsk from the oldsk and migrate the assoc
 * and its messages to the newsk.
 */
static void sctp_sock_migrate(struct sock *oldsk, struct sock *newsk,
			      struct sctp_association *assoc,
			      sctp_socket_type_t type)
{
	struct sctp_sock *oldsp = sctp_sk(oldsk);
	struct sctp_sock *newsp = sctp_sk(newsk);
	struct sctp_bind_bucket *pp; /* hash list port iterator */
	struct sctp_endpoint *newep = newsp->ep;
	struct sk_buff *skb, *tmp;
	struct sctp_ulpevent *event;
6908
	struct sctp_bind_hashbucket *head;
6909
	struct list_head tmplist;
L
Linus Torvalds 已提交
6910 6911 6912 6913 6914 6915 6916

	/* Migrate socket buffer sizes and all the socket level options to the
	 * new socket.
	 */
	newsk->sk_sndbuf = oldsk->sk_sndbuf;
	newsk->sk_rcvbuf = oldsk->sk_rcvbuf;
	/* Brute force copy old sctp opt. */
6917 6918 6919 6920 6921 6922
	if (oldsp->do_auto_asconf) {
		memcpy(&tmplist, &newsp->auto_asconf_list, sizeof(tmplist));
		inet_sk_copy_descendant(newsk, oldsk);
		memcpy(&newsp->auto_asconf_list, &tmplist, sizeof(tmplist));
	} else
		inet_sk_copy_descendant(newsk, oldsk);
L
Linus Torvalds 已提交
6923 6924 6925 6926 6927 6928 6929 6930

	/* Restore the ep value that was overwritten with the above structure
	 * copy.
	 */
	newsp->ep = newep;
	newsp->hmac = NULL;

	/* Hook this new socket in to the bind_hash list. */
6931 6932
	head = &sctp_port_hashtable[sctp_phashfn(sock_net(oldsk),
						 inet_sk(oldsk)->inet_num)];
6933 6934
	sctp_local_bh_disable();
	sctp_spin_lock(&head->lock);
L
Linus Torvalds 已提交
6935 6936 6937
	pp = sctp_sk(oldsk)->bind_hash;
	sk_add_bind_node(newsk, &pp->owner);
	sctp_sk(newsk)->bind_hash = pp;
E
Eric Dumazet 已提交
6938
	inet_sk(newsk)->inet_num = inet_sk(oldsk)->inet_num;
6939 6940
	sctp_spin_unlock(&head->lock);
	sctp_local_bh_enable();
L
Linus Torvalds 已提交
6941

6942 6943 6944
	/* Copy the bind_addr list from the original endpoint to the new
	 * endpoint so that we can handle restarts properly
	 */
6945 6946
	sctp_bind_addr_dup(&newsp->ep->base.bind_addr,
				&oldsp->ep->base.bind_addr, GFP_KERNEL);
6947

L
Linus Torvalds 已提交
6948 6949 6950 6951 6952 6953
	/* Move any messages in the old socket's receive queue that are for the
	 * peeled off association to the new socket's receive queue.
	 */
	sctp_skb_for_each(skb, &oldsk->sk_receive_queue, tmp) {
		event = sctp_skb2event(skb);
		if (event->asoc == assoc) {
D
David S. Miller 已提交
6954
			__skb_unlink(skb, &oldsk->sk_receive_queue);
L
Linus Torvalds 已提交
6955
			__skb_queue_tail(&newsk->sk_receive_queue, skb);
6956
			sctp_skb_set_owner_r_frag(skb, newsk);
L
Linus Torvalds 已提交
6957 6958 6959 6960 6961 6962 6963 6964 6965 6966
		}
	}

	/* Clean up any messages pending delivery due to partial
	 * delivery.   Three cases:
	 * 1) No partial deliver;  no work.
	 * 2) Peeling off partial delivery; keep pd_lobby in new pd_lobby.
	 * 3) Peeling off non-partial delivery; move pd_lobby to receive_queue.
	 */
	skb_queue_head_init(&newsp->pd_lobby);
6967
	atomic_set(&sctp_sk(newsk)->pd_mode, assoc->ulpq.pd_mode);
L
Linus Torvalds 已提交
6968

6969
	if (atomic_read(&sctp_sk(oldsk)->pd_mode)) {
L
Linus Torvalds 已提交
6970 6971 6972 6973 6974 6975 6976 6977 6978 6979 6980 6981 6982 6983
		struct sk_buff_head *queue;

		/* Decide which queue to move pd_lobby skbs to. */
		if (assoc->ulpq.pd_mode) {
			queue = &newsp->pd_lobby;
		} else
			queue = &newsk->sk_receive_queue;

		/* Walk through the pd_lobby, looking for skbs that
		 * need moved to the new socket.
		 */
		sctp_skb_for_each(skb, &oldsp->pd_lobby, tmp) {
			event = sctp_skb2event(skb);
			if (event->asoc == assoc) {
D
David S. Miller 已提交
6984
				__skb_unlink(skb, &oldsp->pd_lobby);
L
Linus Torvalds 已提交
6985
				__skb_queue_tail(queue, skb);
6986
				sctp_skb_set_owner_r_frag(skb, newsk);
L
Linus Torvalds 已提交
6987 6988 6989 6990 6991 6992 6993
			}
		}

		/* Clear up any skbs waiting for the partial
		 * delivery to finish.
		 */
		if (assoc->ulpq.pd_mode)
6994
			sctp_clear_pd(oldsk, NULL);
L
Linus Torvalds 已提交
6995 6996 6997

	}

6998
	sctp_skb_for_each(skb, &assoc->ulpq.reasm, tmp)
6999 7000
		sctp_skb_set_owner_r_frag(skb, newsk);

7001
	sctp_skb_for_each(skb, &assoc->ulpq.lobby, tmp)
7002 7003
		sctp_skb_set_owner_r_frag(skb, newsk);

L
Linus Torvalds 已提交
7004 7005 7006 7007 7008 7009
	/* Set the type of socket to indicate that it is peeled off from the
	 * original UDP-style socket or created with the accept() call on a
	 * TCP-style socket..
	 */
	newsp->type = type;

7010 7011 7012 7013 7014
	/* Mark the new socket "in-use" by the user so that any packets
	 * that may arrive on the association after we've moved it are
	 * queued to the backlog.  This prevents a potential race between
	 * backlog processing on the old socket and new-packet processing
	 * on the new socket.
7015 7016 7017
	 *
	 * The caller has just allocated newsk so we can guarantee that other
	 * paths won't try to lock it and then oldsk.
7018
	 */
7019
	lock_sock_nested(newsk, SINGLE_DEPTH_NESTING);
L
Linus Torvalds 已提交
7020 7021 7022 7023 7024 7025 7026 7027 7028
	sctp_assoc_migrate(assoc, newsk);

	/* If the association on the newsk is already closed before accept()
	 * is called, set RCV_SHUTDOWN flag.
	 */
	if (sctp_state(assoc, CLOSED) && sctp_style(newsk, TCP))
		newsk->sk_shutdown |= RCV_SHUTDOWN;

	newsk->sk_state = SCTP_SS_ESTABLISHED;
7029
	sctp_release_sock(newsk);
L
Linus Torvalds 已提交
7030 7031
}

7032

L
Linus Torvalds 已提交
7033 7034 7035 7036 7037 7038 7039 7040 7041 7042 7043 7044 7045 7046 7047 7048 7049 7050 7051 7052 7053 7054
/* This proto struct describes the ULP interface for SCTP.  */
struct proto sctp_prot = {
	.name        =	"SCTP",
	.owner       =	THIS_MODULE,
	.close       =	sctp_close,
	.connect     =	sctp_connect,
	.disconnect  =	sctp_disconnect,
	.accept      =	sctp_accept,
	.ioctl       =	sctp_ioctl,
	.init        =	sctp_init_sock,
	.destroy     =	sctp_destroy_sock,
	.shutdown    =	sctp_shutdown,
	.setsockopt  =	sctp_setsockopt,
	.getsockopt  =	sctp_getsockopt,
	.sendmsg     =	sctp_sendmsg,
	.recvmsg     =	sctp_recvmsg,
	.bind        =	sctp_bind,
	.backlog_rcv =	sctp_backlog_rcv,
	.hash        =	sctp_hash,
	.unhash      =	sctp_unhash,
	.get_port    =	sctp_get_port,
	.obj_size    =  sizeof(struct sctp_sock),
7055 7056 7057 7058 7059 7060
	.sysctl_mem  =  sysctl_sctp_mem,
	.sysctl_rmem =  sysctl_sctp_rmem,
	.sysctl_wmem =  sysctl_sctp_wmem,
	.memory_pressure = &sctp_memory_pressure,
	.enter_memory_pressure = sctp_enter_memory_pressure,
	.memory_allocated = &sctp_memory_allocated,
7061
	.sockets_allocated = &sctp_sockets_allocated,
L
Linus Torvalds 已提交
7062 7063
};

E
Eric Dumazet 已提交
7064
#if IS_ENABLED(CONFIG_IPV6)
7065

L
Linus Torvalds 已提交
7066 7067 7068 7069 7070 7071 7072 7073 7074 7075 7076 7077 7078 7079 7080 7081 7082 7083 7084 7085 7086
struct proto sctpv6_prot = {
	.name		= "SCTPv6",
	.owner		= THIS_MODULE,
	.close		= sctp_close,
	.connect	= sctp_connect,
	.disconnect	= sctp_disconnect,
	.accept		= sctp_accept,
	.ioctl		= sctp_ioctl,
	.init		= sctp_init_sock,
	.destroy	= sctp_destroy_sock,
	.shutdown	= sctp_shutdown,
	.setsockopt	= sctp_setsockopt,
	.getsockopt	= sctp_getsockopt,
	.sendmsg	= sctp_sendmsg,
	.recvmsg	= sctp_recvmsg,
	.bind		= sctp_bind,
	.backlog_rcv	= sctp_backlog_rcv,
	.hash		= sctp_hash,
	.unhash		= sctp_unhash,
	.get_port	= sctp_get_port,
	.obj_size	= sizeof(struct sctp6_sock),
7087 7088 7089 7090 7091 7092
	.sysctl_mem	= sysctl_sctp_mem,
	.sysctl_rmem	= sysctl_sctp_rmem,
	.sysctl_wmem	= sysctl_sctp_wmem,
	.memory_pressure = &sctp_memory_pressure,
	.enter_memory_pressure = sctp_enter_memory_pressure,
	.memory_allocated = &sctp_memory_allocated,
7093
	.sockets_allocated = &sctp_sockets_allocated,
L
Linus Torvalds 已提交
7094
};
E
Eric Dumazet 已提交
7095
#endif /* IS_ENABLED(CONFIG_IPV6) */