output.c 21.6 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 44 45 46 47 48 49 50 51 52
 * 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.
 */

#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>
#include <net/inet_ecn.h>
#include <net/icmp.h>
53
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
54 55 56 57 58 59 60 61 62 63

#ifndef TEST_FRAME
#include <net/tcp.h>
#endif /* TEST_FRAME (not defined) */

#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 67 68 69 70 71 72 73 74 75 76 77

/* Forward declarations for private helpers. */
static sctp_xmit_t sctp_packet_append_data(struct sctp_packet *packet,
					   struct sctp_chunk *chunk);

/* 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;

78
	SCTP_DEBUG_PRINTK("%s: packet:%p vtag:0x%x\n", __func__,
L
Linus Torvalds 已提交
79 80 81 82 83
			  packet, vtag);

	packet->vtag = vtag;
	packet->has_cookie_echo = 0;
	packet->has_sack = 0;
84
	packet->has_auth = 0;
85
	packet->has_data = 0;
L
Linus Torvalds 已提交
86
	packet->ipfragok = 0;
87
	packet->auth = NULL;
L
Linus Torvalds 已提交
88 89 90 91 92

	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
93 94
		 * any other chunks get appended.
		 */
L
Linus Torvalds 已提交
95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
		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;

110
	SCTP_DEBUG_PRINTK("%s: packet:%p transport:%p\n", __func__,
L
Linus Torvalds 已提交
111 112 113 114 115
			  packet, transport);

	packet->transport = transport;
	packet->source_port = sport;
	packet->destination_port = dport;
116
	INIT_LIST_HEAD(&packet->chunk_list);
L
Linus Torvalds 已提交
117
	if (asoc) {
118 119
		struct sctp_sock *sp = sctp_sk(asoc->base.sk);
		overhead = sp->pf->af->net_header_len;
L
Linus Torvalds 已提交
120 121 122 123 124 125 126 127 128
	} else {
		overhead = sizeof(struct ipv6hdr);
	}
	overhead += sizeof(struct sctphdr);
	packet->overhead = overhead;
	packet->size = overhead;
	packet->vtag = 0;
	packet->has_cookie_echo = 0;
	packet->has_sack = 0;
129
	packet->has_auth = 0;
130
	packet->has_data = 0;
L
Linus Torvalds 已提交
131 132
	packet->ipfragok = 0;
	packet->malloced = 0;
133
	packet->auth = NULL;
L
Linus Torvalds 已提交
134 135 136 137 138 139
	return packet;
}

/* Free a packet.  */
void sctp_packet_free(struct sctp_packet *packet)
{
140
	struct sctp_chunk *chunk, *tmp;
L
Linus Torvalds 已提交
141

142
	SCTP_DEBUG_PRINTK("%s: packet:%p\n", __func__, packet);
L
Linus Torvalds 已提交
143

144 145
	list_for_each_entry_safe(chunk, tmp, &packet->chunk_list, list) {
		list_del_init(&chunk->list);
L
Linus Torvalds 已提交
146
		sctp_chunk_free(chunk);
147
	}
L
Linus Torvalds 已提交
148 149 150 151 152 153 154 155 156 157 158 159 160

	if (packet->malloced)
		kfree(packet);
}

/* 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,
161 162
				       struct sctp_chunk *chunk,
				       int one_packet)
L
Linus Torvalds 已提交
163 164 165 166
{
	sctp_xmit_t retval;
	int error = 0;

167
	SCTP_DEBUG_PRINTK("%s: packet:%p chunk:%p\n", __func__,
L
Linus Torvalds 已提交
168 169 170 171 172 173 174 175 176 177 178 179
			  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.
			 */
180 181 182
			if (!one_packet)
				retval = sctp_packet_append_chunk(packet,
								  chunk);
L
Linus Torvalds 已提交
183 184 185 186 187 188 189
		}
		break;

	case SCTP_XMIT_RWND_FULL:
	case SCTP_XMIT_OK:
	case SCTP_XMIT_NAGLE_DELAY:
		break;
190
	}
L
Linus Torvalds 已提交
191 192 193 194

	return retval;
}

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
/* 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.
	 */
	if (chunk->chunk_hdr->type == SCTP_CID_AUTH || pkt->auth)
		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;

	retval = sctp_packet_append_chunk(pkt, auth);

	return retval;
}

L
Linus Torvalds 已提交
228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
/* 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;
		asoc = pkt->transport->asoc;

		if (asoc->a_rwnd > asoc->rwnd) {
			struct sctp_chunk *sack;
			asoc->a_rwnd = asoc->rwnd;
			sack = sctp_make_sack(asoc);
			if (sack) {
				struct timer_list *timer;
				retval = sctp_packet_append_chunk(pkt, sack);
				asoc->peer.sack_needed = 0;
				timer = &asoc->timers[SCTP_EVENT_TIMEOUT_SACK];
				if (timer_pending(timer) && del_timer(timer))
					sctp_association_put(asoc);
			}
		}
	}
	return retval;
}

/* 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;
	__u16 chunk_len = WORD_ROUND(ntohs(chunk->chunk_hdr->length));
	size_t psize;
	size_t pmtu;
	int too_big;

271
	SCTP_DEBUG_PRINTK("%s: packet:%p chunk:%p\n", __func__, packet,
L
Linus Torvalds 已提交
272 273
			  chunk);

274 275 276 277
	/* Try to bundle AUTH chunk */
	retval = sctp_packet_bundle_auth(packet, chunk);
	if (retval != SCTP_XMIT_OK)
		goto finish;
L
Linus Torvalds 已提交
278

279 280
	/* Try to bundle SACK chunk */
	retval = sctp_packet_bundle_sack(packet, chunk);
L
Linus Torvalds 已提交
281 282 283
	if (retval != SCTP_XMIT_OK)
		goto finish;

284
	psize = packet->size;
L
Linus Torvalds 已提交
285
	pmtu  = ((packet->transport->asoc) ?
286 287
		 (packet->transport->asoc->pathmtu) :
		 (packet->transport->pathmtu));
L
Linus Torvalds 已提交
288 289 290 291 292

	too_big = (psize + chunk_len > pmtu);

	/* Decide if we need to fragment or resubmit later. */
	if (too_big) {
293 294 295 296 297 298 299
		/* 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.
L
Linus Torvalds 已提交
300
		 */
301 302
		if (sctp_packet_empty(packet) || !sctp_chunk_is_data(chunk) ||
		    (!packet->has_data && chunk->auth)) {
L
Linus Torvalds 已提交
303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
			/* We no longer do re-fragmentation.
			 * Just fragment at the IP layer, if we
			 * actually hit this condition
			 */
			packet->ipfragok = 1;
			goto append;

		} else {
			retval = SCTP_XMIT_PMTU_FULL;
			goto finish;
		}
	}

append:
	/* We believe that this chunk is OK to add to the packet (as
	 * long as we have the cwnd for it).
	 */

	/* DATA is a special case since we must examine both rwnd and cwnd
	 * before we send DATA.
	 */
324 325
	switch (chunk->chunk_hdr->type) {
	    case SCTP_CID_DATA:
L
Linus Torvalds 已提交
326 327 328
		retval = sctp_packet_append_data(packet, chunk);
		/* Disallow SACK bundling after DATA. */
		packet->has_sack = 1;
329 330 331 332
		/* Disallow AUTH bundling after DATA */
		packet->has_auth = 1;
		/* Let it be knows that packet has DATA in it */
		packet->has_data = 1;
L
Linus Torvalds 已提交
333 334
		if (SCTP_XMIT_OK != retval)
			goto finish;
335 336
		break;
	    case SCTP_CID_COOKIE_ECHO:
L
Linus Torvalds 已提交
337
		packet->has_cookie_echo = 1;
338 339 340
		break;

	    case SCTP_CID_SACK:
L
Linus Torvalds 已提交
341
		packet->has_sack = 1;
342 343 344 345 346 347 348
		break;

	    case SCTP_CID_AUTH:
		packet->has_auth = 1;
		packet->auth = chunk;
		break;
	}
L
Linus Torvalds 已提交
349 350

	/* It is OK to send this chunk.  */
351
	list_add_tail(&chunk->list, &packet->chunk_list);
L
Linus Torvalds 已提交
352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367
	packet->size += chunk_len;
	chunk->transport = packet->transport;
finish:
	return retval;
}

/* 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;
368
	__be32 crc32 = __constant_cpu_to_be32(0);
L
Linus Torvalds 已提交
369
	struct sk_buff *nskb;
370
	struct sctp_chunk *chunk, *tmp;
L
Linus Torvalds 已提交
371 372 373 374
	struct sock *sk;
	int err = 0;
	int padding;		/* How much padding do we need?  */
	__u8 has_data = 0;
375
	struct dst_entry *dst = tp->dst;
376 377
	unsigned char *auth = NULL;	/* pointer to auth in skb data */
	__u32 cksum_buf_len = sizeof(struct sctphdr);
L
Linus Torvalds 已提交
378

379
	SCTP_DEBUG_PRINTK("%s: packet:%p\n", __func__, packet);
L
Linus Torvalds 已提交
380 381

	/* Do NOT generate a chunkless packet. */
382
	if (list_empty(&packet->chunk_list))
L
Linus Torvalds 已提交
383 384 385
		return err;

	/* Set up convenience variables... */
386
	chunk = list_entry(packet->chunk_list.next, struct sctp_chunk, list);
L
Linus Torvalds 已提交
387 388 389
	sk = chunk->skb->sk;

	/* Allocate the new skb.  */
390
	nskb = alloc_skb(packet->size + LL_MAX_HEADER, GFP_ATOMIC);
L
Linus Torvalds 已提交
391 392 393 394
	if (!nskb)
		goto nomem;

	/* Make sure the outbound skb has enough header room reserved. */
395
	skb_reserve(nskb, packet->overhead + LL_MAX_HEADER);
L
Linus Torvalds 已提交
396 397 398 399 400 401

	/* Set the owning socket so that we know where to get the
	 * destination IP address.
	 */
	skb_set_owner_w(nskb, sk);

402 403 404 405 406 407 408 409 410 411 412 413 414
	/* The 'obsolete' field of dst is set to 2 when a dst is freed. */
	if (!dst || (dst->obsolete > 1)) {
		dst_release(dst);
		sctp_transport_route(tp, NULL, sctp_sk(sk));
		if (asoc && (asoc->param_flags & SPP_PMTUD_ENABLE)) {
			sctp_assoc_sync_pmtu(asoc);
		}
	}
	nskb->dst = dst_clone(tp->dst);
	if (!nskb->dst)
		goto no_route;
	dst = nskb->dst;

L
Linus Torvalds 已提交
415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450
	/* Build the SCTP header.  */
	sh = (struct sctphdr *)skb_push(nskb, sizeof(struct sctphdr));
	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");
451 452
	list_for_each_entry_safe(chunk, tmp, &packet->chunk_list, list) {
		list_del_init(&chunk->list);
L
Linus Torvalds 已提交
453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480
		if (sctp_chunk_is_data(chunk)) {

			if (!chunk->has_tsn) {
				sctp_chunk_assign_ssn(chunk);
				sctp_chunk_assign_tsn(chunk);

			/* 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.
			 */

				if (!tp->rto_pending) {
					chunk->rtt_in_progress = 1;
					tp->rto_pending = 1;
				}
			} else
				chunk->resent = 1;

			chunk->sent_at = jiffies;
			has_data = 1;
		}

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

481 482 483 484 485 486 487 488 489
		/* 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),
490
			       chunk->skb->data, chunk->skb->len);
L
Linus Torvalds 已提交
491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508

		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))
509
			sctp_chunk_free(chunk);
L
Linus Torvalds 已提交
510 511
	}

512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534
	/* 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>.
	 */
	if (!(dst->dev->features & NETIF_F_NO_CSUM)) {
		crc32 = sctp_start_cksum((__u8 *)sh, cksum_buf_len);
535
		crc32 = sctp_end_cksum(crc32);
536 537
	} else
		nskb->ip_summed = CHECKSUM_UNNECESSARY;
L
Linus Torvalds 已提交
538 539 540 541

	/* 3) Put the resultant value into the checksum field in the
	 *    common header, and leave the rest of the bits unchanged.
	 */
542
	sh->checksum = crc32;
L
Linus Torvalds 已提交
543 544 545 546 547 548 549 550 551 552 553 554 555

	/* 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
556
	(*tp->af_specific->ecn_capable)(nskb->sk);
L
Linus Torvalds 已提交
557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589

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

	/* Dump that on IP!  */
	if (asoc && 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;
	}

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

		tp->last_time_used = jiffies;

		/* 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);

590 591
	nskb->local_df = packet->ipfragok;
	(*tp->af_specific->sctp_xmit)(nskb, tp);
L
Linus Torvalds 已提交
592 593 594 595 596 597

out:
	packet->size = packet->overhead;
	return err;
no_route:
	kfree_skb(nskb);
598
	IP_INC_STATS_BH(&init_net, IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
599 600 601 602 603 604 605 606 607 608 609 610 611 612

	/* 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.
	 */

613 614
	list_for_each_entry_safe(chunk, tmp, &packet->chunk_list, list) {
		list_del_init(&chunk->list);
L
Linus Torvalds 已提交
615
		if (!sctp_chunk_is_data(chunk))
616
			sctp_chunk_free(chunk);
L
Linus Torvalds 已提交
617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675
	}
	goto out;
nomem:
	err = -ENOMEM;
	goto err;
}

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

/* This private function handles the specifics of appending DATA chunks.  */
static sctp_xmit_t sctp_packet_append_data(struct sctp_packet *packet,
					   struct sctp_chunk *chunk)
{
	sctp_xmit_t retval = SCTP_XMIT_OK;
	size_t datasize, rwnd, inflight;
	struct sctp_transport *transport = packet->transport;
	__u32 max_burst_bytes;
	struct sctp_association *asoc = transport->asoc;
	struct sctp_sock *sp = sctp_sk(asoc->base.sk);
	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;
	inflight = asoc->outqueue.outstanding_bytes;

	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;
		}
	}

	/* sctpimpguide-05 2.14.2
	 * D) When the time comes for the sender to
	 * transmit new DATA chunks, the protocol parameter Max.Burst MUST
	 * first be applied to limit how many new DATA chunks may be sent.
	 * The limit is applied by adjusting cwnd as follows:
	 * 	if ((flightsize + Max.Burst * MTU) < cwnd)
	 *		cwnd = flightsize + Max.Burst * MTU
	 */
676
	max_burst_bytes = asoc->max_burst * asoc->pathmtu;
L
Linus Torvalds 已提交
677 678 679 680 681 682
	if ((transport->flight_size + max_burst_bytes) < transport->cwnd) {
		transport->cwnd = transport->flight_size + max_burst_bytes;
		SCTP_DEBUG_PRINTK("%s: cwnd limited by max_burst: "
				  "transport: %p, cwnd: %d, "
				  "ssthresh: %d, flight_size: %d, "
				  "pba: %d\n",
683
				  __func__, transport,
L
Linus Torvalds 已提交
684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701
				  transport->cwnd,
				  transport->ssthresh,
				  transport->flight_size,
				  transport->partial_bytes_acked);
	}

	/* 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.
	 */
702
	if (chunk->fast_retransmit <= 0)
L
Linus Torvalds 已提交
703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720
		if (transport->flight_size >= transport->cwnd) {
			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.
	 */
	if (!sp->nodelay && sctp_packet_empty(packet) &&
	    q->outstanding_bytes && sctp_state(asoc, ESTABLISHED)) {
		unsigned len = datasize + q->out_qlen;

		/* Check whether this chunk and all the rest of pending
		 * data will fit or delay in hopes of bundling a full
		 * sized packet.
		 */
721
		if (len < asoc->frag_point) {
L
Linus Torvalds 已提交
722 723 724 725 726 727 728 729 730 731 732
			retval = SCTP_XMIT_NAGLE_DELAY;
			goto finish;
		}
	}

	/* 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;

733 734 735 736 737 738 739
	/* Update our view of the receiver's rwnd. Include sk_buff overhead
	 * while updating peer.rwnd so that it reduces the chances of a
	 * receiver running out of receive buffer space even when receive
	 * window is still open. This can happen when a sender is sending
	 * sending small messages.
	 */
	datasize += sizeof(struct sk_buff);
L
Linus Torvalds 已提交
740 741 742 743 744 745 746 747 748 749 750 751 752
	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;

finish:
	return retval;
}