output.c 22.9 KB
Newer Older
1
/* SCTP kernel implementation
L
Linus Torvalds 已提交
2 3 4 5
 * (C) Copyright IBM Corp. 2001, 2004
 * Copyright (c) 1999-2000 Cisco, Inc.
 * Copyright (c) 1999-2001 Motorola, Inc.
 *
6
 * This file is part of the SCTP kernel implementation
L
Linus Torvalds 已提交
7 8 9
 *
 * These functions handle output processing.
 *
10
 * This SCTP implementation is free software;
L
Linus Torvalds 已提交
11 12 13 14 15
 * 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.
 *
16
 * This SCTP implementation is distributed in the hope that it
L
Linus Torvalds 已提交
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
 * will be useful, but WITHOUT ANY WARRANTY; without even the implied
 *                 ************************
 * warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with GNU CC; see the file COPYING.  If not, write to
 * the Free Software Foundation, 59 Temple Place - Suite 330,
 * Boston, MA 02111-1307, USA.
 *
 * Please send any bug reports or fixes you make to the
 * email address(es):
 *    lksctp developers <lksctp-developers@lists.sourceforge.net>
 *
 * Or submit a bug report through the following website:
 *    http://www.sf.net/projects/lksctp
 *
 * Written or modified by:
 *    La Monte H.P. Yarroll <piggy@acm.org>
 *    Karl Knutson          <karl@athena.chicago.il.us>
 *    Jon Grimm             <jgrimm@austin.ibm.com>
 *    Sridhar Samudrala     <sri@us.ibm.com>
 *
 * Any bugs reported given to us we will try to fix... any fixes shared will
 * be incorporated into the next SCTP release.
 */

44 45
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

L
Linus Torvalds 已提交
46 47 48 49 50 51 52
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/wait.h>
#include <linux/time.h>
#include <linux/ip.h>
#include <linux/ipv6.h>
#include <linux/init.h>
53
#include <linux/slab.h>
L
Linus Torvalds 已提交
54
#include <net/inet_ecn.h>
V
Vlad Yasevich 已提交
55
#include <net/ip.h>
L
Linus Torvalds 已提交
56
#include <net/icmp.h>
57
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
58 59 60 61 62 63

#include <linux/socket.h> /* for sa_family_t */
#include <net/sock.h>

#include <net/sctp/sctp.h>
#include <net/sctp/sm.h>
64
#include <net/sctp/checksum.h>
L
Linus Torvalds 已提交
65 66

/* Forward declarations for private helpers. */
67 68
static sctp_xmit_t __sctp_packet_append_chunk(struct sctp_packet *packet,
					      struct sctp_chunk *chunk);
69
static sctp_xmit_t sctp_packet_can_append_data(struct sctp_packet *packet,
L
Linus Torvalds 已提交
70
					   struct sctp_chunk *chunk);
71 72 73 74 75
static void sctp_packet_append_data(struct sctp_packet *packet,
					   struct sctp_chunk *chunk);
static sctp_xmit_t sctp_packet_will_fit(struct sctp_packet *packet,
					struct sctp_chunk *chunk,
					u16 chunk_len);
L
Linus Torvalds 已提交
76

77 78 79 80 81 82 83 84 85 86 87
static void sctp_packet_reset(struct sctp_packet *packet)
{
	packet->size = packet->overhead;
	packet->has_cookie_echo = 0;
	packet->has_sack = 0;
	packet->has_data = 0;
	packet->has_auth = 0;
	packet->ipfragok = 0;
	packet->auth = NULL;
}

L
Linus Torvalds 已提交
88 89 90 91 92 93 94 95
/* Config a packet.
 * This appears to be a followup set of initializations.
 */
struct sctp_packet *sctp_packet_config(struct sctp_packet *packet,
				       __u32 vtag, int ecn_capable)
{
	struct sctp_chunk *chunk = NULL;

96
	SCTP_DEBUG_PRINTK("%s: packet:%p vtag:0x%x\n", __func__,
L
Linus Torvalds 已提交
97 98 99 100 101 102 103 104
			  packet, vtag);

	packet->vtag = vtag;

	if (ecn_capable && sctp_packet_empty(packet)) {
		chunk = sctp_get_ecne_prepend(packet->transport->asoc);

		/* If there a is a prepend chunk stick it on the list before
105 106
		 * any other chunks get appended.
		 */
L
Linus Torvalds 已提交
107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
		if (chunk)
			sctp_packet_append_chunk(packet, chunk);
	}

	return packet;
}

/* Initialize the packet structure. */
struct sctp_packet *sctp_packet_init(struct sctp_packet *packet,
				     struct sctp_transport *transport,
				     __u16 sport, __u16 dport)
{
	struct sctp_association *asoc = transport->asoc;
	size_t overhead;

122
	SCTP_DEBUG_PRINTK("%s: packet:%p transport:%p\n", __func__,
L
Linus Torvalds 已提交
123 124 125 126 127
			  packet, transport);

	packet->transport = transport;
	packet->source_port = sport;
	packet->destination_port = dport;
128
	INIT_LIST_HEAD(&packet->chunk_list);
L
Linus Torvalds 已提交
129
	if (asoc) {
130 131
		struct sctp_sock *sp = sctp_sk(asoc->base.sk);
		overhead = sp->pf->af->net_header_len;
L
Linus Torvalds 已提交
132 133 134 135 136
	} else {
		overhead = sizeof(struct ipv6hdr);
	}
	overhead += sizeof(struct sctphdr);
	packet->overhead = overhead;
137
	sctp_packet_reset(packet);
L
Linus Torvalds 已提交
138
	packet->vtag = 0;
139

L
Linus Torvalds 已提交
140 141 142 143 144 145
	return packet;
}

/* Free a packet.  */
void sctp_packet_free(struct sctp_packet *packet)
{
146
	struct sctp_chunk *chunk, *tmp;
L
Linus Torvalds 已提交
147

148
	SCTP_DEBUG_PRINTK("%s: packet:%p\n", __func__, packet);
L
Linus Torvalds 已提交
149

150 151
	list_for_each_entry_safe(chunk, tmp, &packet->chunk_list, list) {
		list_del_init(&chunk->list);
L
Linus Torvalds 已提交
152
		sctp_chunk_free(chunk);
153
	}
L
Linus Torvalds 已提交
154 155 156 157 158 159 160 161 162 163
}

/* This routine tries to append the chunk to the offered packet. If adding
 * the chunk causes the packet to exceed the path MTU and COOKIE_ECHO chunk
 * is not present in the packet, it transmits the input packet.
 * Data can be bundled with a packet containing a COOKIE_ECHO chunk as long
 * as it can fit in the packet, but any more data that does not fit in this
 * packet can be sent only after receiving the COOKIE_ACK.
 */
sctp_xmit_t sctp_packet_transmit_chunk(struct sctp_packet *packet,
164 165
				       struct sctp_chunk *chunk,
				       int one_packet)
L
Linus Torvalds 已提交
166 167 168 169
{
	sctp_xmit_t retval;
	int error = 0;

170
	SCTP_DEBUG_PRINTK("%s: packet:%p chunk:%p\n", __func__,
L
Linus Torvalds 已提交
171 172 173 174 175 176 177 178 179 180 181 182
			  packet, chunk);

	switch ((retval = (sctp_packet_append_chunk(packet, chunk)))) {
	case SCTP_XMIT_PMTU_FULL:
		if (!packet->has_cookie_echo) {
			error = sctp_packet_transmit(packet);
			if (error < 0)
				chunk->skb->sk->sk_err = -error;

			/* If we have an empty packet, then we can NOT ever
			 * return PMTU_FULL.
			 */
183 184 185
			if (!one_packet)
				retval = sctp_packet_append_chunk(packet,
								  chunk);
L
Linus Torvalds 已提交
186 187 188 189 190 191 192
		}
		break;

	case SCTP_XMIT_RWND_FULL:
	case SCTP_XMIT_OK:
	case SCTP_XMIT_NAGLE_DELAY:
		break;
193
	}
L
Linus Torvalds 已提交
194 195 196 197

	return retval;
}

198 199 200 201 202 203 204 205 206 207 208 209 210 211 212
/* Try to bundle an auth chunk into the packet. */
static sctp_xmit_t sctp_packet_bundle_auth(struct sctp_packet *pkt,
					   struct sctp_chunk *chunk)
{
	struct sctp_association *asoc = pkt->transport->asoc;
	struct sctp_chunk *auth;
	sctp_xmit_t retval = SCTP_XMIT_OK;

	/* if we don't have an association, we can't do authentication */
	if (!asoc)
		return retval;

	/* See if this is an auth chunk we are bundling or if
	 * auth is already bundled.
	 */
213
	if (chunk->chunk_hdr->type == SCTP_CID_AUTH || pkt->has_auth)
214 215 216 217 218 219 220 221 222 223 224 225
		return retval;

	/* if the peer did not request this chunk to be authenticated,
	 * don't do it
	 */
	if (!chunk->auth)
		return retval;

	auth = sctp_make_auth(asoc);
	if (!auth)
		return retval;

226 227 228 229
	retval = __sctp_packet_append_chunk(pkt, auth);

	if (retval != SCTP_XMIT_OK)
		sctp_chunk_free(auth);
230 231 232 233

	return retval;
}

L
Linus Torvalds 已提交
234 235 236 237 238 239 240 241 242 243 244 245
/* Try to bundle a SACK with the packet. */
static sctp_xmit_t sctp_packet_bundle_sack(struct sctp_packet *pkt,
					   struct sctp_chunk *chunk)
{
	sctp_xmit_t retval = SCTP_XMIT_OK;

	/* If sending DATA and haven't aleady bundled a SACK, try to
	 * bundle one in to the packet.
	 */
	if (sctp_chunk_is_data(chunk) && !pkt->has_sack &&
	    !pkt->has_cookie_echo) {
		struct sctp_association *asoc;
D
Doug Graham 已提交
246
		struct timer_list *timer;
L
Linus Torvalds 已提交
247
		asoc = pkt->transport->asoc;
D
Doug Graham 已提交
248
		timer = &asoc->timers[SCTP_EVENT_TIMEOUT_SACK];
L
Linus Torvalds 已提交
249

D
Doug Graham 已提交
250 251
		/* If the SACK timer is running, we have a pending SACK */
		if (timer_pending(timer)) {
L
Linus Torvalds 已提交
252
			struct sctp_chunk *sack;
253 254 255 256 257

			if (pkt->transport->sack_generation !=
			    pkt->transport->asoc->peer.sack_generation)
				return retval;

L
Linus Torvalds 已提交
258 259 260
			asoc->a_rwnd = asoc->rwnd;
			sack = sctp_make_sack(asoc);
			if (sack) {
261 262 263 264 265
				retval = __sctp_packet_append_chunk(pkt, sack);
				if (retval != SCTP_XMIT_OK) {
					sctp_chunk_free(sack);
					goto out;
				}
L
Linus Torvalds 已提交
266
				asoc->peer.sack_needed = 0;
D
Doug Graham 已提交
267
				if (del_timer(timer))
L
Linus Torvalds 已提交
268 269 270 271
					sctp_association_put(asoc);
			}
		}
	}
272
out:
L
Linus Torvalds 已提交
273 274 275
	return retval;
}

276

L
Linus Torvalds 已提交
277 278 279
/* Append a chunk to the offered packet reporting back any inability to do
 * so.
 */
280 281
static sctp_xmit_t __sctp_packet_append_chunk(struct sctp_packet *packet,
					      struct sctp_chunk *chunk)
L
Linus Torvalds 已提交
282 283 284 285
{
	sctp_xmit_t retval = SCTP_XMIT_OK;
	__u16 chunk_len = WORD_ROUND(ntohs(chunk->chunk_hdr->length));

286 287 288 289
	/* Check to see if this chunk will fit into the packet */
	retval = sctp_packet_will_fit(packet, chunk, chunk_len);
	if (retval != SCTP_XMIT_OK)
		goto finish;
L
Linus Torvalds 已提交
290

291
	/* We believe that this chunk is OK to add to the packet */
292 293
	switch (chunk->chunk_hdr->type) {
	    case SCTP_CID_DATA:
294 295
		/* Account for the data being in the packet */
		sctp_packet_append_data(packet, chunk);
L
Linus Torvalds 已提交
296 297
		/* Disallow SACK bundling after DATA. */
		packet->has_sack = 1;
298 299 300 301
		/* Disallow AUTH bundling after DATA */
		packet->has_auth = 1;
		/* Let it be knows that packet has DATA in it */
		packet->has_data = 1;
302 303
		/* timestamp the chunk for rtx purposes */
		chunk->sent_at = jiffies;
304 305
		break;
	    case SCTP_CID_COOKIE_ECHO:
L
Linus Torvalds 已提交
306
		packet->has_cookie_echo = 1;
307 308 309
		break;

	    case SCTP_CID_SACK:
L
Linus Torvalds 已提交
310
		packet->has_sack = 1;
311 312
		if (chunk->asoc)
			chunk->asoc->stats.osacks++;
313 314 315 316 317 318 319
		break;

	    case SCTP_CID_AUTH:
		packet->has_auth = 1;
		packet->auth = chunk;
		break;
	}
L
Linus Torvalds 已提交
320 321

	/* It is OK to send this chunk.  */
322
	list_add_tail(&chunk->list, &packet->chunk_list);
L
Linus Torvalds 已提交
323 324 325 326 327 328
	packet->size += chunk_len;
	chunk->transport = packet->transport;
finish:
	return retval;
}

329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365
/* Append a chunk to the offered packet reporting back any inability to do
 * so.
 */
sctp_xmit_t sctp_packet_append_chunk(struct sctp_packet *packet,
				     struct sctp_chunk *chunk)
{
	sctp_xmit_t retval = SCTP_XMIT_OK;

	SCTP_DEBUG_PRINTK("%s: packet:%p chunk:%p\n", __func__, packet,
			  chunk);

	/* Data chunks are special.  Before seeing what else we can
	 * bundle into this packet, check to see if we are allowed to
	 * send this DATA.
	 */
	if (sctp_chunk_is_data(chunk)) {
		retval = sctp_packet_can_append_data(packet, chunk);
		if (retval != SCTP_XMIT_OK)
			goto finish;
	}

	/* Try to bundle AUTH chunk */
	retval = sctp_packet_bundle_auth(packet, chunk);
	if (retval != SCTP_XMIT_OK)
		goto finish;

	/* Try to bundle SACK chunk */
	retval = sctp_packet_bundle_sack(packet, chunk);
	if (retval != SCTP_XMIT_OK)
		goto finish;

	retval = __sctp_packet_append_chunk(packet, chunk);

finish:
	return retval;
}

366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384
static void sctp_packet_release_owner(struct sk_buff *skb)
{
	sk_free(skb->sk);
}

static void sctp_packet_set_owner_w(struct sk_buff *skb, struct sock *sk)
{
	skb_orphan(skb);
	skb->sk = sk;
	skb->destructor = sctp_packet_release_owner;

	/*
	 * The data chunks have already been accounted for in sctp_sendmsg(),
	 * therefore only reserve a single byte to keep socket around until
	 * the packet has been transmitted.
	 */
	atomic_inc(&sk->sk_wmem_alloc);
}

L
Linus Torvalds 已提交
385 386 387 388 389 390 391 392 393 394 395
/* All packets are sent to the network through this function from
 * sctp_outq_tail().
 *
 * The return value is a normal kernel error return value.
 */
int sctp_packet_transmit(struct sctp_packet *packet)
{
	struct sctp_transport *tp = packet->transport;
	struct sctp_association *asoc = tp->asoc;
	struct sctphdr *sh;
	struct sk_buff *nskb;
396
	struct sctp_chunk *chunk, *tmp;
L
Linus Torvalds 已提交
397 398 399 400
	struct sock *sk;
	int err = 0;
	int padding;		/* How much padding do we need?  */
	__u8 has_data = 0;
401
	struct dst_entry *dst = tp->dst;
402 403
	unsigned char *auth = NULL;	/* pointer to auth in skb data */
	__u32 cksum_buf_len = sizeof(struct sctphdr);
L
Linus Torvalds 已提交
404

405
	SCTP_DEBUG_PRINTK("%s: packet:%p\n", __func__, packet);
L
Linus Torvalds 已提交
406 407

	/* Do NOT generate a chunkless packet. */
408
	if (list_empty(&packet->chunk_list))
L
Linus Torvalds 已提交
409 410 411
		return err;

	/* Set up convenience variables... */
412
	chunk = list_entry(packet->chunk_list.next, struct sctp_chunk, list);
L
Linus Torvalds 已提交
413 414 415
	sk = chunk->skb->sk;

	/* Allocate the new skb.  */
416
	nskb = alloc_skb(packet->size + LL_MAX_HEADER, GFP_ATOMIC);
L
Linus Torvalds 已提交
417 418 419 420
	if (!nskb)
		goto nomem;

	/* Make sure the outbound skb has enough header room reserved. */
421
	skb_reserve(nskb, packet->overhead + LL_MAX_HEADER);
L
Linus Torvalds 已提交
422 423 424 425

	/* Set the owning socket so that we know where to get the
	 * destination IP address.
	 */
426
	sctp_packet_set_owner_w(nskb, sk);
L
Linus Torvalds 已提交
427

428
	if (!sctp_transport_dst_check(tp)) {
429 430
		sctp_transport_route(tp, NULL, sctp_sk(sk));
		if (asoc && (asoc->param_flags & SPP_PMTUD_ENABLE)) {
431
			sctp_assoc_sync_pmtu(sk, asoc);
432 433
		}
	}
E
Eric Dumazet 已提交
434 435
	dst = dst_clone(tp->dst);
	skb_dst_set(nskb, dst);
436
	if (!dst)
437 438
		goto no_route;

L
Linus Torvalds 已提交
439 440
	/* Build the SCTP header.  */
	sh = (struct sctphdr *)skb_push(nskb, sizeof(struct sctphdr));
441
	skb_reset_transport_header(nskb);
L
Linus Torvalds 已提交
442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475
	sh->source = htons(packet->source_port);
	sh->dest   = htons(packet->destination_port);

	/* From 6.8 Adler-32 Checksum Calculation:
	 * After the packet is constructed (containing the SCTP common
	 * header and one or more control or DATA chunks), the
	 * transmitter shall:
	 *
	 * 1) Fill in the proper Verification Tag in the SCTP common
	 *    header and initialize the checksum field to 0's.
	 */
	sh->vtag     = htonl(packet->vtag);
	sh->checksum = 0;

	/**
	 * 6.10 Bundling
	 *
	 *    An endpoint bundles chunks by simply including multiple
	 *    chunks in one outbound SCTP packet.  ...
	 */

	/**
	 * 3.2  Chunk Field Descriptions
	 *
	 * The total length of a chunk (including Type, Length and
	 * Value fields) MUST be a multiple of 4 bytes.  If the length
	 * of the chunk is not a multiple of 4 bytes, the sender MUST
	 * pad the chunk with all zero bytes and this padding is not
	 * included in the chunk length field.  The sender should
	 * never pad with more than 3 bytes.
	 *
	 * [This whole comment explains WORD_ROUND() below.]
	 */
	SCTP_DEBUG_PRINTK("***sctp_transmit_packet***\n");
476 477
	list_for_each_entry_safe(chunk, tmp, &packet->chunk_list, list) {
		list_del_init(&chunk->list);
L
Linus Torvalds 已提交
478
		if (sctp_chunk_is_data(chunk)) {
479 480 481 482 483 484
			/* 6.3.1 C4) When data is in flight and when allowed
			 * by rule C5, a new RTT measurement MUST be made each
			 * round trip.  Furthermore, new RTT measurements
			 * SHOULD be made no more than once per round-trip
			 * for a given destination transport address.
			 */
L
Linus Torvalds 已提交
485

486 487 488
			if (!tp->rto_pending) {
				chunk->rtt_in_progress = 1;
				tp->rto_pending = 1;
489
			}
L
Linus Torvalds 已提交
490 491 492 493 494 495 496
			has_data = 1;
		}

		padding = WORD_ROUND(chunk->skb->len) - chunk->skb->len;
		if (padding)
			memset(skb_put(chunk->skb, padding), 0, padding);

497 498 499 500 501 502 503 504 505
		/* if this is the auth chunk that we are adding,
		 * store pointer where it will be added and put
		 * the auth into the packet.
		 */
		if (chunk == packet->auth)
			auth = skb_tail_pointer(nskb);

		cksum_buf_len += chunk->skb->len;
		memcpy(skb_put(nskb, chunk->skb->len),
506
			       chunk->skb->data, chunk->skb->len);
L
Linus Torvalds 已提交
507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524

		SCTP_DEBUG_PRINTK("%s %p[%s] %s 0x%x, %s %d, %s %d, %s %d\n",
				  "*** Chunk", chunk,
				  sctp_cname(SCTP_ST_CHUNK(
					  chunk->chunk_hdr->type)),
				  chunk->has_tsn ? "TSN" : "No TSN",
				  chunk->has_tsn ?
				  ntohl(chunk->subh.data_hdr->tsn) : 0,
				  "length", ntohs(chunk->chunk_hdr->length),
				  "chunk->skb->len", chunk->skb->len,
				  "rtt_in_progress", chunk->rtt_in_progress);

		/*
		 * If this is a control chunk, this is our last
		 * reference. Free data chunks after they've been
		 * acknowledged or have failed.
		 */
		if (!sctp_chunk_is_data(chunk))
525
			sctp_chunk_free(chunk);
L
Linus Torvalds 已提交
526 527
	}

528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548
	/* SCTP-AUTH, Section 6.2
	 *    The sender MUST calculate the MAC as described in RFC2104 [2]
	 *    using the hash function H as described by the MAC Identifier and
	 *    the shared association key K based on the endpoint pair shared key
	 *    described by the shared key identifier.  The 'data' used for the
	 *    computation of the AUTH-chunk is given by the AUTH chunk with its
	 *    HMAC field set to zero (as shown in Figure 6) followed by all
	 *    chunks that are placed after the AUTH chunk in the SCTP packet.
	 */
	if (auth)
		sctp_auth_calculate_hmac(asoc, nskb,
					(struct sctp_auth_chunk *)auth,
					GFP_ATOMIC);

	/* 2) Calculate the Adler-32 checksum of the whole packet,
	 *    including the SCTP common header and all the
	 *    chunks.
	 *
	 * Note: Adler-32 is no longer applicable, as has been replaced
	 * by CRC32-C as described in <draft-ietf-tsvwg-sctpcsum-02.txt>.
	 */
549 550 551
	if (!sctp_checksum_disable) {
		if (!(dst->dev->features & NETIF_F_SCTP_CSUM)) {
			__u32 crc32 = sctp_start_cksum((__u8 *)sh, cksum_buf_len);
552

553 554 555 556 557
			/* 3) Put the resultant value into the checksum field in the
			 *    common header, and leave the rest of the bits unchanged.
			 */
			sh->checksum = sctp_end_cksum(crc32);
		} else {
L
Lucas De Marchi 已提交
558
			/* no need to seed pseudo checksum for SCTP */
559 560 561 562 563 564
			nskb->ip_summed = CHECKSUM_PARTIAL;
			nskb->csum_start = (skb_transport_header(nskb) -
			                    nskb->head);
			nskb->csum_offset = offsetof(struct sctphdr, checksum);
		}
	}
L
Linus Torvalds 已提交
565 566 567 568 569 570 571 572 573 574 575 576 577

	/* IP layer ECN support
	 * From RFC 2481
	 *  "The ECN-Capable Transport (ECT) bit would be set by the
	 *   data sender to indicate that the end-points of the
	 *   transport protocol are ECN-capable."
	 *
	 * Now setting the ECT bit all the time, as it should not cause
	 * any problems protocol-wise even if our peer ignores it.
	 *
	 * Note: The works for IPv6 layer checks this bit too later
	 * in transmission.  See IP6_ECN_flow_xmit().
	 */
V
Vlad Yasevich 已提交
578
	(*tp->af_specific->ecn_capable)(nskb->sk);
L
Linus Torvalds 已提交
579 580 581 582 583 584 585

	/* Set up the IP options.  */
	/* BUG: not implemented
	 * For v4 this all lives somewhere in sk->sk_opt...
	 */

	/* Dump that on IP!  */
586 587 588 589 590 591 592
	if (asoc) {
		asoc->stats.opackets++;
		if (asoc->peer.last_sent_to != tp)
			/* Considering the multiple CPU scenario, this is a
			 * "correcter" place for last_sent_to.  --xguo
			 */
			asoc->peer.last_sent_to = tp;
L
Linus Torvalds 已提交
593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611
	}

	if (has_data) {
		struct timer_list *timer;
		unsigned long timeout;

		/* Restart the AUTOCLOSE timer when sending data. */
		if (sctp_state(asoc, ESTABLISHED) && asoc->autoclose) {
			timer = &asoc->timers[SCTP_EVENT_TIMEOUT_AUTOCLOSE];
			timeout = asoc->timeouts[SCTP_EVENT_TIMEOUT_AUTOCLOSE];

			if (!mod_timer(timer, jiffies + timeout))
				sctp_association_hold(asoc);
		}
	}

	SCTP_DEBUG_PRINTK("***sctp_transmit_packet*** skb len %d\n",
			  nskb->len);

612 613
	nskb->local_df = packet->ipfragok;
	(*tp->af_specific->sctp_xmit)(nskb, tp);
L
Linus Torvalds 已提交
614 615

out:
616
	sctp_packet_reset(packet);
L
Linus Torvalds 已提交
617 618 619
	return err;
no_route:
	kfree_skb(nskb);
620
	IP_INC_STATS_BH(sock_net(asoc->base.sk), IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
621 622 623 624 625 626 627 628 629 630 631 632 633 634

	/* FIXME: Returning the 'err' will effect all the associations
	 * associated with a socket, although only one of the paths of the
	 * association is unreachable.
	 * The real failure of a transport or association can be passed on
	 * to the user via notifications. So setting this error may not be
	 * required.
	 */
	 /* err = -EHOSTUNREACH; */
err:
	/* Control chunks are unreliable so just drop them.  DATA chunks
	 * will get resent or dropped later.
	 */

635 636
	list_for_each_entry_safe(chunk, tmp, &packet->chunk_list, list) {
		list_del_init(&chunk->list);
L
Linus Torvalds 已提交
637
		if (!sctp_chunk_is_data(chunk))
638
			sctp_chunk_free(chunk);
L
Linus Torvalds 已提交
639 640 641 642 643 644 645 646 647 648 649
	}
	goto out;
nomem:
	err = -ENOMEM;
	goto err;
}

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

650 651
/* This private function check to see if a chunk can be added */
static sctp_xmit_t sctp_packet_can_append_data(struct sctp_packet *packet,
L
Linus Torvalds 已提交
652 653 654
					   struct sctp_chunk *chunk)
{
	sctp_xmit_t retval = SCTP_XMIT_OK;
655
	size_t datasize, rwnd, inflight, flight_size;
L
Linus Torvalds 已提交
656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673
	struct sctp_transport *transport = packet->transport;
	struct sctp_association *asoc = transport->asoc;
	struct sctp_outq *q = &asoc->outqueue;

	/* RFC 2960 6.1  Transmission of DATA Chunks
	 *
	 * A) At any given time, the data sender MUST NOT transmit new data to
	 * any destination transport address if its peer's rwnd indicates
	 * that the peer has no buffer space (i.e. rwnd is 0, see Section
	 * 6.2.1).  However, regardless of the value of rwnd (including if it
	 * is 0), the data sender can always have one DATA chunk in flight to
	 * the receiver if allowed by cwnd (see rule B below).  This rule
	 * allows the sender to probe for a change in rwnd that the sender
	 * missed due to the SACK having been lost in transit from the data
	 * receiver to the data sender.
	 */

	rwnd = asoc->peer.rwnd;
674 675
	inflight = q->outstanding_bytes;
	flight_size = transport->flight_size;
L
Linus Torvalds 已提交
676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700

	datasize = sctp_data_size(chunk);

	if (datasize > rwnd) {
		if (inflight > 0) {
			/* We have (at least) one data chunk in flight,
			 * so we can't fall back to rule 6.1 B).
			 */
			retval = SCTP_XMIT_RWND_FULL;
			goto finish;
		}
	}

	/* RFC 2960 6.1  Transmission of DATA Chunks
	 *
	 * B) At any given time, the sender MUST NOT transmit new data
	 * to a given transport address if it has cwnd or more bytes
	 * of data outstanding to that transport address.
	 */
	/* RFC 7.2.4 & the Implementers Guide 2.8.
	 *
	 * 3) ...
	 *    When a Fast Retransmit is being performed the sender SHOULD
	 *    ignore the value of cwnd and SHOULD NOT delay retransmission.
	 */
701
	if (chunk->fast_retransmit != SCTP_NEED_FRTX)
702
		if (flight_size >= transport->cwnd) {
L
Linus Torvalds 已提交
703 704 705 706 707 708 709 710 711
			retval = SCTP_XMIT_RWND_FULL;
			goto finish;
		}

	/* Nagle's algorithm to solve small-packet problem:
	 * Inhibit the sending of new chunks when new outgoing data arrives
	 * if any previously transmitted data on the connection remains
	 * unacknowledged.
	 */
712 713
	if (!sctp_sk(asoc->base.sk)->nodelay && sctp_packet_empty(packet) &&
	    inflight && sctp_state(asoc, ESTABLISHED)) {
714 715
		unsigned int max = transport->pathmtu - packet->overhead;
		unsigned int len = chunk->skb->len + q->out_qlen;
L
Linus Torvalds 已提交
716 717 718 719

		/* Check whether this chunk and all the rest of pending
		 * data will fit or delay in hopes of bundling a full
		 * sized packet.
720 721
		 * Don't delay large message writes that may have been
		 * fragmeneted into small peices.
L
Linus Torvalds 已提交
722
		 */
723
		if ((len < max) && chunk->msg->can_delay) {
L
Linus Torvalds 已提交
724 725 726 727 728
			retval = SCTP_XMIT_NAGLE_DELAY;
			goto finish;
		}
	}

729 730 731 732 733 734 735 736 737 738 739 740 741
finish:
	return retval;
}

/* This private function does management things when adding DATA chunk */
static void sctp_packet_append_data(struct sctp_packet *packet,
				struct sctp_chunk *chunk)
{
	struct sctp_transport *transport = packet->transport;
	size_t datasize = sctp_data_size(chunk);
	struct sctp_association *asoc = transport->asoc;
	u32 rwnd = asoc->peer.rwnd;

L
Linus Torvalds 已提交
742 743 744 745 746 747
	/* Keep track of how many bytes are in flight over this transport. */
	transport->flight_size += datasize;

	/* Keep track of how many bytes are in flight to the receiver. */
	asoc->outqueue.outstanding_bytes += datasize;

748
	/* Update our view of the receiver's rwnd. */
L
Linus Torvalds 已提交
749 750 751 752 753 754 755 756 757
	if (datasize < rwnd)
		rwnd -= datasize;
	else
		rwnd = 0;

	asoc->peer.rwnd = rwnd;
	/* Has been accepted for transmission. */
	if (!asoc->peer.prsctp_capable)
		chunk->msg->can_abandon = 0;
758 759
	sctp_chunk_assign_tsn(chunk);
	sctp_chunk_assign_ssn(chunk);
760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798
}

static sctp_xmit_t sctp_packet_will_fit(struct sctp_packet *packet,
					struct sctp_chunk *chunk,
					u16 chunk_len)
{
	size_t psize;
	size_t pmtu;
	int too_big;
	sctp_xmit_t retval = SCTP_XMIT_OK;

	psize = packet->size;
	pmtu  = ((packet->transport->asoc) ?
		(packet->transport->asoc->pathmtu) :
		(packet->transport->pathmtu));

	too_big = (psize + chunk_len > pmtu);

	/* Decide if we need to fragment or resubmit later. */
	if (too_big) {
		/* It's OK to fragmet at IP level if any one of the following
		 * is true:
		 * 	1. The packet is empty (meaning this chunk is greater
		 * 	   the MTU)
		 * 	2. The chunk we are adding is a control chunk
		 * 	3. The packet doesn't have any data in it yet and data
		 * 	requires authentication.
		 */
		if (sctp_packet_empty(packet) || !sctp_chunk_is_data(chunk) ||
		    (!packet->has_data && chunk->auth)) {
			/* We no longer do re-fragmentation.
			 * Just fragment at the IP layer, if we
			 * actually hit this condition
			 */
			packet->ipfragok = 1;
		} else {
			retval = SCTP_XMIT_PMTU_FULL;
		}
	}
L
Linus Torvalds 已提交
799 800 801

	return retval;
}