tcp_sack.c revision 1.20
1/* $NetBSD: tcp_sack.c,v 1.20 2006/10/20 13:11:09 reinoud Exp $ */ 2 3/* 4 * Copyright (c) 2005 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Kentaro A. Kurahone. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. All advertising materials mentioning features or use of this software 19 * must display the following acknowledgement: 20 * This product includes software developed by the NetBSD 21 * Foundation, Inc. and its contributors. 22 * 4. Neither the name of The NetBSD Foundation nor the names of its 23 * contributors may be used to endorse or promote products derived 24 * from this software without specific prior written permission. 25 * 26 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 27 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 28 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 29 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 30 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 31 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 32 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 33 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 34 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 36 * POSSIBILITY OF SUCH DAMAGE. 37 */ 38 39/* 40 * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1994, 1995 41 * The Regents of the University of California. All rights reserved. 42 * 43 * Redistribution and use in source and binary forms, with or without 44 * modification, are permitted provided that the following conditions 45 * are met: 46 * 1. Redistributions of source code must retain the above copyright 47 * notice, this list of conditions and the following disclaimer. 48 * 2. Redistributions in binary form must reproduce the above copyright 49 * notice, this list of conditions and the following disclaimer in the 50 * documentation and/or other materials provided with the distribution. 51 * 4. Neither the name of the University nor the names of its contributors 52 * may be used to endorse or promote products derived from this software 53 * without specific prior written permission. 54 * 55 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 56 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 57 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 58 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 59 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 60 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 61 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 62 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 63 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 64 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 65 * SUCH DAMAGE. 66 * 67 * @(#)tcp_sack.c 8.12 (Berkeley) 5/24/95 68 * $FreeBSD: src/sys/netinet/tcp_sack.c,v 1.3.2.2 2004/12/25 23:02:57 rwatson Exp $ 69 */ 70 71/* 72 * @@(#)COPYRIGHT 1.1 (NRL) 17 January 1995 73 * 74 * NRL grants permission for redistribution and use in source and binary 75 * forms, with or without modification, of the software and documentation 76 * created at NRL provided that the following conditions are met: 77 * 78 * 1. Redistributions of source code must retain the above copyright 79 * notice, this list of conditions and the following disclaimer. 80 * 2. Redistributions in binary form must reproduce the above copyright 81 * notice, this list of conditions and the following disclaimer in the 82 * documentation and/or other materials provided with the distribution. 83 * 3. All advertising materials mentioning features or use of this software 84 * must display the following acknowledgements: 85 * This product includes software developed by the University of 86 * California, Berkeley and its contributors. 87 * This product includes software developed at the Information 88 * Technology Division, US Naval Research Laboratory. 89 * 4. Neither the name of the NRL nor the names of its contributors 90 * may be used to endorse or promote products derived from this software 91 * without specific prior written permission. 92 * 93 * THE SOFTWARE PROVIDED BY NRL IS PROVIDED BY NRL AND CONTRIBUTORS ``AS 94 * IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 95 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A 96 * PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL NRL OR 97 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 98 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 99 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR 100 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF 101 * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING 102 * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 103 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 104 * 105 * The views and conclusions contained in the software and documentation 106 * are those of the authors and should not be interpreted as representing 107 * official policies, either expressed or implied, of the US Naval 108 * Research Laboratory (NRL). 109 */ 110 111#include <sys/cdefs.h> 112__KERNEL_RCSID(0, "$NetBSD: tcp_sack.c,v 1.20 2006/10/20 13:11:09 reinoud Exp $"); 113 114#include "opt_inet.h" 115#include "opt_ipsec.h" 116#include "opt_inet_csum.h" 117#include "opt_tcp_debug.h" 118 119#include <sys/param.h> 120#include <sys/systm.h> 121#include <sys/malloc.h> 122#include <sys/mbuf.h> 123#include <sys/protosw.h> 124#include <sys/socket.h> 125#include <sys/socketvar.h> 126#include <sys/errno.h> 127#include <sys/syslog.h> 128#include <sys/pool.h> 129#include <sys/domain.h> 130#include <sys/kernel.h> 131 132#include <net/if.h> 133#include <net/route.h> 134#include <net/if_types.h> 135 136#include <netinet/in.h> 137#include <netinet/in_systm.h> 138#include <netinet/ip.h> 139#include <netinet/in_pcb.h> 140#include <netinet/in_var.h> 141#include <netinet/ip_var.h> 142 143#ifdef INET6 144#ifndef INET 145#include <netinet/in.h> 146#endif 147#include <netinet/ip6.h> 148#include <netinet6/ip6_var.h> 149#include <netinet6/in6_pcb.h> 150#include <netinet6/ip6_var.h> 151#include <netinet6/in6_var.h> 152#include <netinet/icmp6.h> 153#include <netinet6/nd6.h> 154#endif 155 156#ifndef INET6 157/* always need ip6.h for IP6_EXTHDR_GET */ 158#include <netinet/ip6.h> 159#endif 160 161#include <netinet/tcp.h> 162#include <netinet/tcp_fsm.h> 163#include <netinet/tcp_seq.h> 164#include <netinet/tcp_timer.h> 165#include <netinet/tcp_var.h> 166#include <netinet/tcpip.h> 167#include <netinet/tcp_debug.h> 168 169#include <machine/stdarg.h> 170 171/* SACK block pool. */ 172static POOL_INIT(sackhole_pool, sizeof(struct sackhole), 0, 0, 0, "sackholepl", 173 NULL); 174 175static struct sackhole * 176sack_allochole(struct tcpcb *tp) 177{ 178 struct sackhole *hole; 179 180 if (tp->snd_numholes >= tcp_sack_tp_maxholes || 181 tcp_sack_globalholes >= tcp_sack_globalmaxholes) { 182 return NULL; 183 } 184 hole = pool_get(&sackhole_pool, PR_NOWAIT); 185 if (hole == NULL) { 186 return NULL; 187 } 188 tp->snd_numholes++; 189 tcp_sack_globalholes++; 190 191 return hole; 192} 193 194static struct sackhole * 195sack_inserthole(struct tcpcb *tp, tcp_seq start, tcp_seq end, 196 struct sackhole *prev) 197{ 198 struct sackhole *hole; 199 200 hole = sack_allochole(tp); 201 if (hole == NULL) { 202 return NULL; 203 } 204 hole->start = hole->rxmit = start; 205 hole->end = end; 206 if (prev != NULL) { 207 TAILQ_INSERT_AFTER(&tp->snd_holes, prev, hole, sackhole_q); 208 } else { 209 TAILQ_INSERT_TAIL(&tp->snd_holes, hole, sackhole_q); 210 } 211 return hole; 212} 213 214static struct sackhole * 215sack_removehole(struct tcpcb *tp, struct sackhole *hole) 216{ 217 struct sackhole *next; 218 219 next = TAILQ_NEXT(hole, sackhole_q); 220 tp->snd_numholes--; 221 tcp_sack_globalholes--; 222 TAILQ_REMOVE(&tp->snd_holes, hole, sackhole_q); 223 pool_put(&sackhole_pool, hole); 224 225 return next; 226} 227 228void 229tcp_new_dsack(struct tcpcb *tp, tcp_seq seq, u_int32_t len) 230{ 231 if (TCP_SACK_ENABLED(tp)) { 232 tp->rcv_dsack_block.left = seq; 233 tp->rcv_dsack_block.right = seq + len; 234 tp->rcv_sack_flags |= TCPSACK_HAVED; 235 } 236} 237 238void 239tcp_sack_option(struct tcpcb *tp, struct tcphdr *th, u_char *cp, int optlen) 240{ 241 struct sackblk 242 t_sack_block[(MAX_TCPOPTLEN - 2) / (sizeof(u_int32_t) * 2)]; 243 struct sackblk *sack = NULL; 244 struct sackhole *cur = NULL; 245 struct sackhole *tmp = NULL; 246 char *lp = cp + 2; 247 int i, j, num_sack_blks; 248 tcp_seq left, right, acked; 249 250 /* 251 * If we aren't processing SACK responses, this is not an ACK 252 * or the peer sends us a sack option with invalid length, don't 253 * update the scoreboard. 254 */ 255 if (!TCP_SACK_ENABLED(tp) || ((th->th_flags & TH_ACK) == 0) || 256 (optlen % 8 != 2 || optlen < 10)) { 257 return; 258 } 259 260 /* 261 * If we don't want any SACK holes to be allocated, just return. 262 */ 263 if (tcp_sack_globalmaxholes == 0 || tcp_sack_tp_maxholes == 0) { 264 return; 265 } 266 267 /* If the ACK is outside [snd_una, snd_max], ignore the SACK options. */ 268 if (SEQ_LT(th->th_ack, tp->snd_una) || SEQ_GT(th->th_ack, tp->snd_max)) 269 return; 270 271 /* 272 * Extract SACK blocks. 273 * 274 * Note that t_sack_block is sorted so that we only need to do 275 * one pass over the sequence number space. (SACK "fast-path") 276 */ 277 num_sack_blks = optlen / 8; 278 acked = (SEQ_GT(th->th_ack, tp->snd_una)) ? th->th_ack : tp->snd_una; 279 for (i = 0; i < num_sack_blks; i++, lp += sizeof(uint32_t) * 2) { 280 memcpy(&left, lp, sizeof(uint32_t)); 281 memcpy(&right, lp + sizeof(uint32_t), sizeof(uint32_t)); 282 left = ntohl(left); 283 right = ntohl(right); 284 285 if (SEQ_LEQ(right, acked) || SEQ_GT(right, tp->snd_max) || 286 SEQ_GEQ(left, right)) { 287 /* SACK entry that's old, or invalid. */ 288 i--; 289 num_sack_blks--; 290 continue; 291 } 292 293 /* Insertion sort. */ 294 for (j = i; (j > 0) && SEQ_LT(left, t_sack_block[j - 1].left); 295 j--) { 296 t_sack_block[j].left = t_sack_block[j - 1].left; 297 t_sack_block[j].right = t_sack_block[j - 1].right; 298 } 299 t_sack_block[j].left = left; 300 t_sack_block[j].right = right; 301 } 302 303 /* Update the scoreboard. */ 304 cur = TAILQ_FIRST(&tp->snd_holes); 305 for (i = 0; i < num_sack_blks; i++) { 306 sack = &t_sack_block[i]; 307 /* 308 * FACK TCP. Update snd_fack so we can enter Fast 309 * Recovery early. 310 */ 311 if (SEQ_GEQ(sack->right, tp->snd_fack)) 312 tp->snd_fack = sack->right; 313 314 if (TAILQ_EMPTY(&tp->snd_holes)) { 315 /* First hole. */ 316 cur = sack_inserthole(tp, th->th_ack, sack->left, NULL); 317 if (cur == NULL) { 318 /* ENOBUFS, bail out*/ 319 return; 320 } 321 tp->rcv_lastsack = sack->right; 322 continue; /* With next sack block */ 323 } 324 325 /* Go through the list of holes. */ 326 while (cur) { 327 if (SEQ_LEQ(sack->right, cur->start)) 328 /* SACKs data before the current hole */ 329 break; /* No use going through more holes */ 330 331 if (SEQ_GEQ(sack->left, cur->end)) { 332 /* SACKs data beyond the current hole */ 333 cur = TAILQ_NEXT(cur, sackhole_q); 334 continue; 335 } 336 337 if (SEQ_LEQ(sack->left, cur->start)) { 338 /* Data acks at least the beginning of hole */ 339 if (SEQ_GEQ(sack->right, cur->end)) { 340 /* Acks entire hole, so delete hole */ 341 cur = sack_removehole(tp, cur); 342 break; 343 } 344 345 /* Otherwise, move start of hole forward */ 346 cur->start = sack->right; 347 cur->rxmit = SEQ_MAX(cur->rxmit, cur->start); 348 break; 349 } 350 351 if (SEQ_GEQ(sack->right, cur->end)) { 352 /* Move end of hole backward. */ 353 cur->end = sack->left; 354 cur->rxmit = SEQ_MIN(cur->rxmit, cur->end); 355 cur = TAILQ_NEXT(cur, sackhole_q); 356 break; 357 } 358 359 if (SEQ_LT(cur->start, sack->left) && 360 SEQ_GT(cur->end, sack->right)) { 361 /* 362 * ACKs some data in middle of a hole; need to 363 * split current hole 364 */ 365 tmp = sack_inserthole(tp, sack->right, cur->end, 366 cur); 367 if (tmp == NULL) { 368 return; 369 } 370 tmp->rxmit = SEQ_MAX(cur->rxmit, tmp->start); 371 cur->end = sack->left; 372 cur->rxmit = SEQ_MIN(cur->rxmit, cur->end); 373 cur = tmp; 374 break; 375 } 376 } 377 378 /* At this point, we have reached the tail of the list. */ 379 if (SEQ_LT(tp->rcv_lastsack, sack->left)) { 380 /* 381 * Need to append new hole at end. 382 */ 383 cur = sack_inserthole(tp, tp->rcv_lastsack, sack->left, 384 NULL); 385 if (cur == NULL) { 386 return; 387 } 388 } 389 if (SEQ_LT(tp->rcv_lastsack, sack->right)) { 390 tp->rcv_lastsack = sack->right; 391 } 392 } 393} 394 395void 396tcp_del_sackholes(struct tcpcb *tp, struct tcphdr *th) 397{ 398 /* Max because this could be an older ack that just arrived. */ 399 tcp_seq lastack = SEQ_GT(th->th_ack, tp->snd_una) ? 400 th->th_ack : tp->snd_una; 401 struct sackhole *cur = TAILQ_FIRST(&tp->snd_holes); 402 403 while (cur) { 404 if (SEQ_LEQ(cur->end, lastack)) { 405 cur = sack_removehole(tp, cur); 406 } else if (SEQ_LT(cur->start, lastack)) { 407 cur->start = lastack; 408 if (SEQ_LT(cur->rxmit, cur->start)) 409 cur->rxmit = cur->start; 410 break; 411 } else 412 break; 413 } 414} 415 416void 417tcp_free_sackholes(struct tcpcb *tp) 418{ 419 struct sackhole *sack; 420 421 /* Free up the SACK hole list. */ 422 while ((sack = TAILQ_FIRST(&tp->snd_holes)) != NULL) { 423 sack_removehole(tp, sack); 424 } 425 KASSERT(tp->snd_numholes == 0); 426} 427 428/* 429 * Implements the SACK response to a new ack, checking for partial acks 430 * in fast recovery. 431 */ 432void 433tcp_sack_newack(struct tcpcb *tp, struct tcphdr *th) 434{ 435 if (tp->t_partialacks < 0) { 436 /* 437 * Not in fast recovery. Reset the duplicate ack 438 * counter. 439 */ 440 tp->t_dupacks = 0; 441 } else if (SEQ_LT(th->th_ack, tp->snd_recover)) { 442 /* 443 * Partial ack handling within a sack recovery episode. 444 * Keeping this very simple for now. When a partial ack 445 * is received, force snd_cwnd to a value that will allow 446 * the sender to transmit no more than 2 segments. 447 * If necessary, a fancier scheme can be adopted at a 448 * later point, but for now, the goal is to prevent the 449 * sender from bursting a large amount of data in the midst 450 * of sack recovery. 451 */ 452 int num_segs = 1; 453 int sack_bytes_rxmt = 0; 454 455 tp->t_partialacks++; 456 TCP_TIMER_DISARM(tp, TCPT_REXMT); 457 tp->t_rtttime = 0; 458 459 /* 460 * send one or 2 segments based on how much new data was acked 461 */ 462 if (((th->th_ack - tp->snd_una) / tp->t_segsz) > 2) 463 num_segs = 2; 464 (void)tcp_sack_output(tp, &sack_bytes_rxmt); 465 tp->snd_cwnd = sack_bytes_rxmt + 466 (tp->snd_nxt - tp->sack_newdata) + num_segs * tp->t_segsz; 467 tp->t_flags |= TF_ACKNOW; 468 (void) tcp_output(tp); 469 } else { 470 /* 471 * Complete ack, inflate the congestion window to 472 * ssthresh and exit fast recovery. 473 * 474 * Window inflation should have left us with approx. 475 * snd_ssthresh outstanding data. But in case we 476 * would be inclined to send a burst, better to do 477 * it via the slow start mechanism. 478 */ 479 if (SEQ_SUB(tp->snd_max, th->th_ack) < tp->snd_ssthresh) 480 tp->snd_cwnd = SEQ_SUB(tp->snd_max, th->th_ack) 481 + tp->t_segsz; 482 else 483 tp->snd_cwnd = tp->snd_ssthresh; 484 tp->t_partialacks = -1; 485 tp->t_dupacks = 0; 486 if (SEQ_GT(th->th_ack, tp->snd_fack)) 487 tp->snd_fack = th->th_ack; 488 } 489} 490 491/* 492 * Returns pointer to a sackhole if there are any pending retransmissions; 493 * NULL otherwise. 494 */ 495struct sackhole * 496tcp_sack_output(struct tcpcb *tp, int *sack_bytes_rexmt) 497{ 498 struct sackhole *cur = NULL; 499 500 if (!TCP_SACK_ENABLED(tp)) 501 return (NULL); 502 503 *sack_bytes_rexmt = 0; 504 TAILQ_FOREACH(cur, &tp->snd_holes, sackhole_q) { 505 if (SEQ_LT(cur->rxmit, cur->end)) { 506 if (SEQ_LT(cur->rxmit, tp->snd_una)) { 507 /* old SACK hole */ 508 continue; 509 } 510 *sack_bytes_rexmt += (cur->rxmit - cur->start); 511 break; 512 } 513 *sack_bytes_rexmt += (cur->rxmit - cur->start); 514 } 515 516 return (cur); 517} 518 519/* 520 * After a timeout, the SACK list may be rebuilt. This SACK information 521 * should be used to avoid retransmitting SACKed data. This function 522 * traverses the SACK list to see if snd_nxt should be moved forward. 523 */ 524void 525tcp_sack_adjust(struct tcpcb *tp) 526{ 527 struct sackhole *cur = TAILQ_FIRST(&tp->snd_holes); 528 struct sackhole *n = NULL; 529 530 if (TAILQ_EMPTY(&tp->snd_holes)) 531 return; /* No holes */ 532 if (SEQ_GEQ(tp->snd_nxt, tp->rcv_lastsack)) 533 return; /* We're already beyond any SACKed blocks */ 534 535 /* 536 * Two cases for which we want to advance snd_nxt: 537 * i) snd_nxt lies between end of one hole and beginning of another 538 * ii) snd_nxt lies between end of last hole and rcv_lastsack 539 */ 540 while ((n = TAILQ_NEXT(cur, sackhole_q)) != NULL) { 541 if (SEQ_LT(tp->snd_nxt, cur->end)) 542 return; 543 if (SEQ_GEQ(tp->snd_nxt, n->start)) 544 cur = n; 545 else { 546 tp->snd_nxt = n->start; 547 return; 548 } 549 } 550 if (SEQ_LT(tp->snd_nxt, cur->end)) 551 return; 552 tp->snd_nxt = tp->rcv_lastsack; 553 554 return; 555} 556 557int 558tcp_sack_numblks(const struct tcpcb *tp) 559{ 560 int numblks; 561 562 if (!TCP_SACK_ENABLED(tp)) { 563 return 0; 564 } 565 566 numblks = (((tp->rcv_sack_flags & TCPSACK_HAVED) != 0) ? 1 : 0) + 567 tp->t_segqlen; 568 569 if (numblks == 0) { 570 return 0; 571 } 572 573 if (numblks > TCP_SACK_MAX) { 574 numblks = TCP_SACK_MAX; 575 } 576 577 return numblks; 578} 579