tcp_input.c revision 1.346
1 /* $NetBSD: tcp_input.c,v 1.346 2016/06/10 13:27:16 ozaki-r Exp $ */ 2 3 /* 4 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project. 5 * All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 3. Neither the name of the project nor the names of its contributors 16 * may be used to endorse or promote products derived from this software 17 * without specific prior written permission. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND 20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 22 * ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE 23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 29 * SUCH DAMAGE. 30 */ 31 32 /* 33 * @(#)COPYRIGHT 1.1 (NRL) 17 January 1995 34 * 35 * NRL grants permission for redistribution and use in source and binary 36 * forms, with or without modification, of the software and documentation 37 * created at NRL provided that the following conditions are met: 38 * 39 * 1. Redistributions of source code must retain the above copyright 40 * notice, this list of conditions and the following disclaimer. 41 * 2. Redistributions in binary form must reproduce the above copyright 42 * notice, this list of conditions and the following disclaimer in the 43 * documentation and/or other materials provided with the distribution. 44 * 3. All advertising materials mentioning features or use of this software 45 * must display the following acknowledgements: 46 * This product includes software developed by the University of 47 * California, Berkeley and its contributors. 48 * This product includes software developed at the Information 49 * Technology Division, US Naval Research Laboratory. 50 * 4. Neither the name of the NRL nor the names of its contributors 51 * may be used to endorse or promote products derived from this software 52 * without specific prior written permission. 53 * 54 * THE SOFTWARE PROVIDED BY NRL IS PROVIDED BY NRL AND CONTRIBUTORS ``AS 55 * IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 56 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A 57 * PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL NRL OR 58 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 59 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 60 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR 61 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF 62 * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING 63 * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 64 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 65 * 66 * The views and conclusions contained in the software and documentation 67 * are those of the authors and should not be interpreted as representing 68 * official policies, either expressed or implied, of the US Naval 69 * Research Laboratory (NRL). 70 */ 71 72 /*- 73 * Copyright (c) 1997, 1998, 1999, 2001, 2005, 2006, 74 * 2011 The NetBSD Foundation, Inc. 75 * All rights reserved. 76 * 77 * This code is derived from software contributed to The NetBSD Foundation 78 * by Coyote Point Systems, Inc. 79 * This code is derived from software contributed to The NetBSD Foundation 80 * by Jason R. Thorpe and Kevin M. Lahey of the Numerical Aerospace Simulation 81 * Facility, NASA Ames Research Center. 82 * This code is derived from software contributed to The NetBSD Foundation 83 * by Charles M. Hannum. 84 * This code is derived from software contributed to The NetBSD Foundation 85 * by Rui Paulo. 86 * 87 * Redistribution and use in source and binary forms, with or without 88 * modification, are permitted provided that the following conditions 89 * are met: 90 * 1. Redistributions of source code must retain the above copyright 91 * notice, this list of conditions and the following disclaimer. 92 * 2. Redistributions in binary form must reproduce the above copyright 93 * notice, this list of conditions and the following disclaimer in the 94 * documentation and/or other materials provided with the distribution. 95 * 96 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 97 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 98 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 99 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 100 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 101 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 102 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 103 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 104 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 105 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 106 * POSSIBILITY OF SUCH DAMAGE. 107 */ 108 109 /* 110 * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1994, 1995 111 * The Regents of the University of California. All rights reserved. 112 * 113 * Redistribution and use in source and binary forms, with or without 114 * modification, are permitted provided that the following conditions 115 * are met: 116 * 1. Redistributions of source code must retain the above copyright 117 * notice, this list of conditions and the following disclaimer. 118 * 2. Redistributions in binary form must reproduce the above copyright 119 * notice, this list of conditions and the following disclaimer in the 120 * documentation and/or other materials provided with the distribution. 121 * 3. Neither the name of the University nor the names of its contributors 122 * may be used to endorse or promote products derived from this software 123 * without specific prior written permission. 124 * 125 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 126 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 127 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 128 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 129 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 130 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 131 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 132 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 133 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 134 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 135 * SUCH DAMAGE. 136 * 137 * @(#)tcp_input.c 8.12 (Berkeley) 5/24/95 138 */ 139 140 /* 141 * TODO list for SYN cache stuff: 142 * 143 * Find room for a "state" field, which is needed to keep a 144 * compressed state for TIME_WAIT TCBs. It's been noted already 145 * that this is fairly important for very high-volume web and 146 * mail servers, which use a large number of short-lived 147 * connections. 148 */ 149 150 #include <sys/cdefs.h> 151 __KERNEL_RCSID(0, "$NetBSD: tcp_input.c,v 1.346 2016/06/10 13:27:16 ozaki-r Exp $"); 152 153 #ifdef _KERNEL_OPT 154 #include "opt_inet.h" 155 #include "opt_ipsec.h" 156 #include "opt_inet_csum.h" 157 #include "opt_tcp_debug.h" 158 #endif 159 160 #include <sys/param.h> 161 #include <sys/systm.h> 162 #include <sys/malloc.h> 163 #include <sys/mbuf.h> 164 #include <sys/protosw.h> 165 #include <sys/socket.h> 166 #include <sys/socketvar.h> 167 #include <sys/errno.h> 168 #include <sys/syslog.h> 169 #include <sys/pool.h> 170 #include <sys/domain.h> 171 #include <sys/kernel.h> 172 #ifdef TCP_SIGNATURE 173 #include <sys/md5.h> 174 #endif 175 #include <sys/lwp.h> /* for lwp0 */ 176 #include <sys/cprng.h> 177 178 #include <net/if.h> 179 #include <net/if_types.h> 180 181 #include <netinet/in.h> 182 #include <netinet/in_systm.h> 183 #include <netinet/ip.h> 184 #include <netinet/in_pcb.h> 185 #include <netinet/in_var.h> 186 #include <netinet/ip_var.h> 187 #include <netinet/in_offload.h> 188 189 #ifdef INET6 190 #ifndef INET 191 #include <netinet/in.h> 192 #endif 193 #include <netinet/ip6.h> 194 #include <netinet6/ip6_var.h> 195 #include <netinet6/in6_pcb.h> 196 #include <netinet6/ip6_var.h> 197 #include <netinet6/in6_var.h> 198 #include <netinet/icmp6.h> 199 #include <netinet6/nd6.h> 200 #ifdef TCP_SIGNATURE 201 #include <netinet6/scope6_var.h> 202 #endif 203 #endif 204 205 #ifndef INET6 206 /* always need ip6.h for IP6_EXTHDR_GET */ 207 #include <netinet/ip6.h> 208 #endif 209 210 #include <netinet/tcp.h> 211 #include <netinet/tcp_fsm.h> 212 #include <netinet/tcp_seq.h> 213 #include <netinet/tcp_timer.h> 214 #include <netinet/tcp_var.h> 215 #include <netinet/tcp_private.h> 216 #include <netinet/tcpip.h> 217 #include <netinet/tcp_congctl.h> 218 #include <netinet/tcp_debug.h> 219 220 #ifdef INET6 221 #include "faith.h" 222 #if defined(NFAITH) && NFAITH > 0 223 #include <net/if_faith.h> 224 #endif 225 #endif /* INET6 */ 226 227 #ifdef IPSEC 228 #include <netipsec/ipsec.h> 229 #include <netipsec/ipsec_var.h> 230 #include <netipsec/ipsec_private.h> 231 #include <netipsec/key.h> 232 #ifdef INET6 233 #include <netipsec/ipsec6.h> 234 #endif 235 #endif /* IPSEC*/ 236 237 #include <netinet/tcp_vtw.h> 238 239 int tcprexmtthresh = 3; 240 int tcp_log_refused; 241 242 int tcp_do_autorcvbuf = 1; 243 int tcp_autorcvbuf_inc = 16 * 1024; 244 int tcp_autorcvbuf_max = 256 * 1024; 245 int tcp_msl = (TCPTV_MSL / PR_SLOWHZ); 246 247 static int tcp_rst_ppslim_count = 0; 248 static struct timeval tcp_rst_ppslim_last; 249 static int tcp_ackdrop_ppslim_count = 0; 250 static struct timeval tcp_ackdrop_ppslim_last; 251 252 #define TCP_PAWS_IDLE (24U * 24 * 60 * 60 * PR_SLOWHZ) 253 254 /* for modulo comparisons of timestamps */ 255 #define TSTMP_LT(a,b) ((int)((a)-(b)) < 0) 256 #define TSTMP_GEQ(a,b) ((int)((a)-(b)) >= 0) 257 258 /* 259 * Neighbor Discovery, Neighbor Unreachability Detection Upper layer hint. 260 */ 261 #ifdef INET6 262 static inline void 263 nd6_hint(struct tcpcb *tp) 264 { 265 struct rtentry *rt; 266 267 if (tp != NULL && tp->t_in6pcb != NULL && tp->t_family == AF_INET6 && 268 (rt = rtcache_validate(&tp->t_in6pcb->in6p_route)) != NULL) 269 nd6_nud_hint(rt); 270 } 271 #else 272 static inline void 273 nd6_hint(struct tcpcb *tp) 274 { 275 } 276 #endif 277 278 /* 279 * Compute ACK transmission behavior. Delay the ACK unless 280 * we have already delayed an ACK (must send an ACK every two segments). 281 * We also ACK immediately if we received a PUSH and the ACK-on-PUSH 282 * option is enabled. 283 */ 284 static void 285 tcp_setup_ack(struct tcpcb *tp, const struct tcphdr *th) 286 { 287 288 if (tp->t_flags & TF_DELACK || 289 (tcp_ack_on_push && th->th_flags & TH_PUSH)) 290 tp->t_flags |= TF_ACKNOW; 291 else 292 TCP_SET_DELACK(tp); 293 } 294 295 static void 296 icmp_check(struct tcpcb *tp, const struct tcphdr *th, int acked) 297 { 298 299 /* 300 * If we had a pending ICMP message that refers to data that have 301 * just been acknowledged, disregard the recorded ICMP message. 302 */ 303 if ((tp->t_flags & TF_PMTUD_PEND) && 304 SEQ_GT(th->th_ack, tp->t_pmtud_th_seq)) 305 tp->t_flags &= ~TF_PMTUD_PEND; 306 307 /* 308 * Keep track of the largest chunk of data 309 * acknowledged since last PMTU update 310 */ 311 if (tp->t_pmtud_mss_acked < acked) 312 tp->t_pmtud_mss_acked = acked; 313 } 314 315 /* 316 * Convert TCP protocol fields to host order for easier processing. 317 */ 318 static void 319 tcp_fields_to_host(struct tcphdr *th) 320 { 321 322 NTOHL(th->th_seq); 323 NTOHL(th->th_ack); 324 NTOHS(th->th_win); 325 NTOHS(th->th_urp); 326 } 327 328 /* 329 * ... and reverse the above. 330 */ 331 static void 332 tcp_fields_to_net(struct tcphdr *th) 333 { 334 335 HTONL(th->th_seq); 336 HTONL(th->th_ack); 337 HTONS(th->th_win); 338 HTONS(th->th_urp); 339 } 340 341 #ifdef TCP_CSUM_COUNTERS 342 #include <sys/device.h> 343 344 #if defined(INET) 345 extern struct evcnt tcp_hwcsum_ok; 346 extern struct evcnt tcp_hwcsum_bad; 347 extern struct evcnt tcp_hwcsum_data; 348 extern struct evcnt tcp_swcsum; 349 #endif /* defined(INET) */ 350 #if defined(INET6) 351 extern struct evcnt tcp6_hwcsum_ok; 352 extern struct evcnt tcp6_hwcsum_bad; 353 extern struct evcnt tcp6_hwcsum_data; 354 extern struct evcnt tcp6_swcsum; 355 #endif /* defined(INET6) */ 356 357 #define TCP_CSUM_COUNTER_INCR(ev) (ev)->ev_count++ 358 359 #else 360 361 #define TCP_CSUM_COUNTER_INCR(ev) /* nothing */ 362 363 #endif /* TCP_CSUM_COUNTERS */ 364 365 #ifdef TCP_REASS_COUNTERS 366 #include <sys/device.h> 367 368 extern struct evcnt tcp_reass_; 369 extern struct evcnt tcp_reass_empty; 370 extern struct evcnt tcp_reass_iteration[8]; 371 extern struct evcnt tcp_reass_prependfirst; 372 extern struct evcnt tcp_reass_prepend; 373 extern struct evcnt tcp_reass_insert; 374 extern struct evcnt tcp_reass_inserttail; 375 extern struct evcnt tcp_reass_append; 376 extern struct evcnt tcp_reass_appendtail; 377 extern struct evcnt tcp_reass_overlaptail; 378 extern struct evcnt tcp_reass_overlapfront; 379 extern struct evcnt tcp_reass_segdup; 380 extern struct evcnt tcp_reass_fragdup; 381 382 #define TCP_REASS_COUNTER_INCR(ev) (ev)->ev_count++ 383 384 #else 385 386 #define TCP_REASS_COUNTER_INCR(ev) /* nothing */ 387 388 #endif /* TCP_REASS_COUNTERS */ 389 390 static int tcp_reass(struct tcpcb *, const struct tcphdr *, struct mbuf *, 391 int *); 392 static int tcp_dooptions(struct tcpcb *, const u_char *, int, 393 struct tcphdr *, struct mbuf *, int, struct tcp_opt_info *); 394 395 #ifdef INET 396 static void tcp4_log_refused(const struct ip *, const struct tcphdr *); 397 #endif 398 #ifdef INET6 399 static void tcp6_log_refused(const struct ip6_hdr *, const struct tcphdr *); 400 #endif 401 402 #define TRAVERSE(x) while ((x)->m_next) (x) = (x)->m_next 403 404 #if defined(MBUFTRACE) 405 struct mowner tcp_reass_mowner = MOWNER_INIT("tcp", "reass"); 406 #endif /* defined(MBUFTRACE) */ 407 408 static struct pool tcpipqent_pool; 409 410 void 411 tcpipqent_init(void) 412 { 413 414 pool_init(&tcpipqent_pool, sizeof(struct ipqent), 0, 0, 0, "tcpipqepl", 415 NULL, IPL_VM); 416 } 417 418 struct ipqent * 419 tcpipqent_alloc(void) 420 { 421 struct ipqent *ipqe; 422 int s; 423 424 s = splvm(); 425 ipqe = pool_get(&tcpipqent_pool, PR_NOWAIT); 426 splx(s); 427 428 return ipqe; 429 } 430 431 void 432 tcpipqent_free(struct ipqent *ipqe) 433 { 434 int s; 435 436 s = splvm(); 437 pool_put(&tcpipqent_pool, ipqe); 438 splx(s); 439 } 440 441 static int 442 tcp_reass(struct tcpcb *tp, const struct tcphdr *th, struct mbuf *m, int *tlen) 443 { 444 struct ipqent *p, *q, *nq, *tiqe = NULL; 445 struct socket *so = NULL; 446 int pkt_flags; 447 tcp_seq pkt_seq; 448 unsigned pkt_len; 449 u_long rcvpartdupbyte = 0; 450 u_long rcvoobyte; 451 #ifdef TCP_REASS_COUNTERS 452 u_int count = 0; 453 #endif 454 uint64_t *tcps; 455 456 if (tp->t_inpcb) 457 so = tp->t_inpcb->inp_socket; 458 #ifdef INET6 459 else if (tp->t_in6pcb) 460 so = tp->t_in6pcb->in6p_socket; 461 #endif 462 463 TCP_REASS_LOCK_CHECK(tp); 464 465 /* 466 * Call with th==0 after become established to 467 * force pre-ESTABLISHED data up to user socket. 468 */ 469 if (th == 0) 470 goto present; 471 472 m_claimm(m, &tcp_reass_mowner); 473 474 rcvoobyte = *tlen; 475 /* 476 * Copy these to local variables because the tcpiphdr 477 * gets munged while we are collapsing mbufs. 478 */ 479 pkt_seq = th->th_seq; 480 pkt_len = *tlen; 481 pkt_flags = th->th_flags; 482 483 TCP_REASS_COUNTER_INCR(&tcp_reass_); 484 485 if ((p = TAILQ_LAST(&tp->segq, ipqehead)) != NULL) { 486 /* 487 * When we miss a packet, the vast majority of time we get 488 * packets that follow it in order. So optimize for that. 489 */ 490 if (pkt_seq == p->ipqe_seq + p->ipqe_len) { 491 p->ipqe_len += pkt_len; 492 p->ipqe_flags |= pkt_flags; 493 m_cat(p->ipre_mlast, m); 494 TRAVERSE(p->ipre_mlast); 495 m = NULL; 496 tiqe = p; 497 TAILQ_REMOVE(&tp->timeq, p, ipqe_timeq); 498 TCP_REASS_COUNTER_INCR(&tcp_reass_appendtail); 499 goto skip_replacement; 500 } 501 /* 502 * While we're here, if the pkt is completely beyond 503 * anything we have, just insert it at the tail. 504 */ 505 if (SEQ_GT(pkt_seq, p->ipqe_seq + p->ipqe_len)) { 506 TCP_REASS_COUNTER_INCR(&tcp_reass_inserttail); 507 goto insert_it; 508 } 509 } 510 511 q = TAILQ_FIRST(&tp->segq); 512 513 if (q != NULL) { 514 /* 515 * If this segment immediately precedes the first out-of-order 516 * block, simply slap the segment in front of it and (mostly) 517 * skip the complicated logic. 518 */ 519 if (pkt_seq + pkt_len == q->ipqe_seq) { 520 q->ipqe_seq = pkt_seq; 521 q->ipqe_len += pkt_len; 522 q->ipqe_flags |= pkt_flags; 523 m_cat(m, q->ipqe_m); 524 q->ipqe_m = m; 525 q->ipre_mlast = m; /* last mbuf may have changed */ 526 TRAVERSE(q->ipre_mlast); 527 tiqe = q; 528 TAILQ_REMOVE(&tp->timeq, q, ipqe_timeq); 529 TCP_REASS_COUNTER_INCR(&tcp_reass_prependfirst); 530 goto skip_replacement; 531 } 532 } else { 533 TCP_REASS_COUNTER_INCR(&tcp_reass_empty); 534 } 535 536 /* 537 * Find a segment which begins after this one does. 538 */ 539 for (p = NULL; q != NULL; q = nq) { 540 nq = TAILQ_NEXT(q, ipqe_q); 541 #ifdef TCP_REASS_COUNTERS 542 count++; 543 #endif 544 /* 545 * If the received segment is just right after this 546 * fragment, merge the two together and then check 547 * for further overlaps. 548 */ 549 if (q->ipqe_seq + q->ipqe_len == pkt_seq) { 550 #ifdef TCPREASS_DEBUG 551 printf("tcp_reass[%p]: concat %u:%u(%u) to %u:%u(%u)\n", 552 tp, pkt_seq, pkt_seq + pkt_len, pkt_len, 553 q->ipqe_seq, q->ipqe_seq + q->ipqe_len, q->ipqe_len); 554 #endif 555 pkt_len += q->ipqe_len; 556 pkt_flags |= q->ipqe_flags; 557 pkt_seq = q->ipqe_seq; 558 m_cat(q->ipre_mlast, m); 559 TRAVERSE(q->ipre_mlast); 560 m = q->ipqe_m; 561 TCP_REASS_COUNTER_INCR(&tcp_reass_append); 562 goto free_ipqe; 563 } 564 /* 565 * If the received segment is completely past this 566 * fragment, we need to go the next fragment. 567 */ 568 if (SEQ_LT(q->ipqe_seq + q->ipqe_len, pkt_seq)) { 569 p = q; 570 continue; 571 } 572 /* 573 * If the fragment is past the received segment, 574 * it (or any following) can't be concatenated. 575 */ 576 if (SEQ_GT(q->ipqe_seq, pkt_seq + pkt_len)) { 577 TCP_REASS_COUNTER_INCR(&tcp_reass_insert); 578 break; 579 } 580 581 /* 582 * We've received all the data in this segment before. 583 * mark it as a duplicate and return. 584 */ 585 if (SEQ_LEQ(q->ipqe_seq, pkt_seq) && 586 SEQ_GEQ(q->ipqe_seq + q->ipqe_len, pkt_seq + pkt_len)) { 587 tcps = TCP_STAT_GETREF(); 588 tcps[TCP_STAT_RCVDUPPACK]++; 589 tcps[TCP_STAT_RCVDUPBYTE] += pkt_len; 590 TCP_STAT_PUTREF(); 591 tcp_new_dsack(tp, pkt_seq, pkt_len); 592 m_freem(m); 593 if (tiqe != NULL) { 594 tcpipqent_free(tiqe); 595 } 596 TCP_REASS_COUNTER_INCR(&tcp_reass_segdup); 597 goto out; 598 } 599 /* 600 * Received segment completely overlaps this fragment 601 * so we drop the fragment (this keeps the temporal 602 * ordering of segments correct). 603 */ 604 if (SEQ_GEQ(q->ipqe_seq, pkt_seq) && 605 SEQ_LEQ(q->ipqe_seq + q->ipqe_len, pkt_seq + pkt_len)) { 606 rcvpartdupbyte += q->ipqe_len; 607 m_freem(q->ipqe_m); 608 TCP_REASS_COUNTER_INCR(&tcp_reass_fragdup); 609 goto free_ipqe; 610 } 611 /* 612 * RX'ed segment extends past the end of the 613 * fragment. Drop the overlapping bytes. Then 614 * merge the fragment and segment then treat as 615 * a longer received packet. 616 */ 617 if (SEQ_LT(q->ipqe_seq, pkt_seq) && 618 SEQ_GT(q->ipqe_seq + q->ipqe_len, pkt_seq)) { 619 int overlap = q->ipqe_seq + q->ipqe_len - pkt_seq; 620 #ifdef TCPREASS_DEBUG 621 printf("tcp_reass[%p]: trim starting %d bytes of %u:%u(%u)\n", 622 tp, overlap, 623 pkt_seq, pkt_seq + pkt_len, pkt_len); 624 #endif 625 m_adj(m, overlap); 626 rcvpartdupbyte += overlap; 627 m_cat(q->ipre_mlast, m); 628 TRAVERSE(q->ipre_mlast); 629 m = q->ipqe_m; 630 pkt_seq = q->ipqe_seq; 631 pkt_len += q->ipqe_len - overlap; 632 rcvoobyte -= overlap; 633 TCP_REASS_COUNTER_INCR(&tcp_reass_overlaptail); 634 goto free_ipqe; 635 } 636 /* 637 * RX'ed segment extends past the front of the 638 * fragment. Drop the overlapping bytes on the 639 * received packet. The packet will then be 640 * contatentated with this fragment a bit later. 641 */ 642 if (SEQ_GT(q->ipqe_seq, pkt_seq) && 643 SEQ_LT(q->ipqe_seq, pkt_seq + pkt_len)) { 644 int overlap = pkt_seq + pkt_len - q->ipqe_seq; 645 #ifdef TCPREASS_DEBUG 646 printf("tcp_reass[%p]: trim trailing %d bytes of %u:%u(%u)\n", 647 tp, overlap, 648 pkt_seq, pkt_seq + pkt_len, pkt_len); 649 #endif 650 m_adj(m, -overlap); 651 pkt_len -= overlap; 652 rcvpartdupbyte += overlap; 653 TCP_REASS_COUNTER_INCR(&tcp_reass_overlapfront); 654 rcvoobyte -= overlap; 655 } 656 /* 657 * If the received segment immediates precedes this 658 * fragment then tack the fragment onto this segment 659 * and reinsert the data. 660 */ 661 if (q->ipqe_seq == pkt_seq + pkt_len) { 662 #ifdef TCPREASS_DEBUG 663 printf("tcp_reass[%p]: append %u:%u(%u) to %u:%u(%u)\n", 664 tp, q->ipqe_seq, q->ipqe_seq + q->ipqe_len, q->ipqe_len, 665 pkt_seq, pkt_seq + pkt_len, pkt_len); 666 #endif 667 pkt_len += q->ipqe_len; 668 pkt_flags |= q->ipqe_flags; 669 m_cat(m, q->ipqe_m); 670 TAILQ_REMOVE(&tp->segq, q, ipqe_q); 671 TAILQ_REMOVE(&tp->timeq, q, ipqe_timeq); 672 tp->t_segqlen--; 673 KASSERT(tp->t_segqlen >= 0); 674 KASSERT(tp->t_segqlen != 0 || 675 (TAILQ_EMPTY(&tp->segq) && 676 TAILQ_EMPTY(&tp->timeq))); 677 if (tiqe == NULL) { 678 tiqe = q; 679 } else { 680 tcpipqent_free(q); 681 } 682 TCP_REASS_COUNTER_INCR(&tcp_reass_prepend); 683 break; 684 } 685 /* 686 * If the fragment is before the segment, remember it. 687 * When this loop is terminated, p will contain the 688 * pointer to fragment that is right before the received 689 * segment. 690 */ 691 if (SEQ_LEQ(q->ipqe_seq, pkt_seq)) 692 p = q; 693 694 continue; 695 696 /* 697 * This is a common operation. It also will allow 698 * to save doing a malloc/free in most instances. 699 */ 700 free_ipqe: 701 TAILQ_REMOVE(&tp->segq, q, ipqe_q); 702 TAILQ_REMOVE(&tp->timeq, q, ipqe_timeq); 703 tp->t_segqlen--; 704 KASSERT(tp->t_segqlen >= 0); 705 KASSERT(tp->t_segqlen != 0 || 706 (TAILQ_EMPTY(&tp->segq) && TAILQ_EMPTY(&tp->timeq))); 707 if (tiqe == NULL) { 708 tiqe = q; 709 } else { 710 tcpipqent_free(q); 711 } 712 } 713 714 #ifdef TCP_REASS_COUNTERS 715 if (count > 7) 716 TCP_REASS_COUNTER_INCR(&tcp_reass_iteration[0]); 717 else if (count > 0) 718 TCP_REASS_COUNTER_INCR(&tcp_reass_iteration[count]); 719 #endif 720 721 insert_it: 722 723 /* 724 * Allocate a new queue entry since the received segment did not 725 * collapse onto any other out-of-order block; thus we are allocating 726 * a new block. If it had collapsed, tiqe would not be NULL and 727 * we would be reusing it. 728 * XXX If we can't, just drop the packet. XXX 729 */ 730 if (tiqe == NULL) { 731 tiqe = tcpipqent_alloc(); 732 if (tiqe == NULL) { 733 TCP_STATINC(TCP_STAT_RCVMEMDROP); 734 m_freem(m); 735 goto out; 736 } 737 } 738 739 /* 740 * Update the counters. 741 */ 742 tp->t_rcvoopack++; 743 tcps = TCP_STAT_GETREF(); 744 tcps[TCP_STAT_RCVOOPACK]++; 745 tcps[TCP_STAT_RCVOOBYTE] += rcvoobyte; 746 if (rcvpartdupbyte) { 747 tcps[TCP_STAT_RCVPARTDUPPACK]++; 748 tcps[TCP_STAT_RCVPARTDUPBYTE] += rcvpartdupbyte; 749 } 750 TCP_STAT_PUTREF(); 751 752 /* 753 * Insert the new fragment queue entry into both queues. 754 */ 755 tiqe->ipqe_m = m; 756 tiqe->ipre_mlast = m; 757 tiqe->ipqe_seq = pkt_seq; 758 tiqe->ipqe_len = pkt_len; 759 tiqe->ipqe_flags = pkt_flags; 760 if (p == NULL) { 761 TAILQ_INSERT_HEAD(&tp->segq, tiqe, ipqe_q); 762 #ifdef TCPREASS_DEBUG 763 if (tiqe->ipqe_seq != tp->rcv_nxt) 764 printf("tcp_reass[%p]: insert %u:%u(%u) at front\n", 765 tp, pkt_seq, pkt_seq + pkt_len, pkt_len); 766 #endif 767 } else { 768 TAILQ_INSERT_AFTER(&tp->segq, p, tiqe, ipqe_q); 769 #ifdef TCPREASS_DEBUG 770 printf("tcp_reass[%p]: insert %u:%u(%u) after %u:%u(%u)\n", 771 tp, pkt_seq, pkt_seq + pkt_len, pkt_len, 772 p->ipqe_seq, p->ipqe_seq + p->ipqe_len, p->ipqe_len); 773 #endif 774 } 775 tp->t_segqlen++; 776 777 skip_replacement: 778 779 TAILQ_INSERT_HEAD(&tp->timeq, tiqe, ipqe_timeq); 780 781 present: 782 /* 783 * Present data to user, advancing rcv_nxt through 784 * completed sequence space. 785 */ 786 if (TCPS_HAVEESTABLISHED(tp->t_state) == 0) 787 goto out; 788 q = TAILQ_FIRST(&tp->segq); 789 if (q == NULL || q->ipqe_seq != tp->rcv_nxt) 790 goto out; 791 if (tp->t_state == TCPS_SYN_RECEIVED && q->ipqe_len) 792 goto out; 793 794 tp->rcv_nxt += q->ipqe_len; 795 pkt_flags = q->ipqe_flags & TH_FIN; 796 nd6_hint(tp); 797 798 TAILQ_REMOVE(&tp->segq, q, ipqe_q); 799 TAILQ_REMOVE(&tp->timeq, q, ipqe_timeq); 800 tp->t_segqlen--; 801 KASSERT(tp->t_segqlen >= 0); 802 KASSERT(tp->t_segqlen != 0 || 803 (TAILQ_EMPTY(&tp->segq) && TAILQ_EMPTY(&tp->timeq))); 804 if (so->so_state & SS_CANTRCVMORE) 805 m_freem(q->ipqe_m); 806 else 807 sbappendstream(&so->so_rcv, q->ipqe_m); 808 tcpipqent_free(q); 809 TCP_REASS_UNLOCK(tp); 810 sorwakeup(so); 811 return (pkt_flags); 812 out: 813 TCP_REASS_UNLOCK(tp); 814 return (0); 815 } 816 817 #ifdef INET6 818 int 819 tcp6_input(struct mbuf **mp, int *offp, int proto) 820 { 821 struct mbuf *m = *mp; 822 823 /* 824 * draft-itojun-ipv6-tcp-to-anycast 825 * better place to put this in? 826 */ 827 if (m->m_flags & M_ANYCAST6) { 828 struct ip6_hdr *ip6; 829 if (m->m_len < sizeof(struct ip6_hdr)) { 830 if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) { 831 TCP_STATINC(TCP_STAT_RCVSHORT); 832 return IPPROTO_DONE; 833 } 834 } 835 ip6 = mtod(m, struct ip6_hdr *); 836 icmp6_error(m, ICMP6_DST_UNREACH, ICMP6_DST_UNREACH_ADDR, 837 (char *)&ip6->ip6_dst - (char *)ip6); 838 return IPPROTO_DONE; 839 } 840 841 tcp_input(m, *offp, proto); 842 return IPPROTO_DONE; 843 } 844 #endif 845 846 #ifdef INET 847 static void 848 tcp4_log_refused(const struct ip *ip, const struct tcphdr *th) 849 { 850 char src[INET_ADDRSTRLEN]; 851 char dst[INET_ADDRSTRLEN]; 852 853 if (ip) { 854 in_print(src, sizeof(src), &ip->ip_src); 855 in_print(dst, sizeof(dst), &ip->ip_dst); 856 } 857 else { 858 strlcpy(src, "(unknown)", sizeof(src)); 859 strlcpy(dst, "(unknown)", sizeof(dst)); 860 } 861 log(LOG_INFO, 862 "Connection attempt to TCP %s:%d from %s:%d\n", 863 dst, ntohs(th->th_dport), 864 src, ntohs(th->th_sport)); 865 } 866 #endif 867 868 #ifdef INET6 869 static void 870 tcp6_log_refused(const struct ip6_hdr *ip6, const struct tcphdr *th) 871 { 872 char src[INET6_ADDRSTRLEN]; 873 char dst[INET6_ADDRSTRLEN]; 874 875 if (ip6) { 876 in6_print(src, sizeof(src), &ip6->ip6_src); 877 in6_print(dst, sizeof(dst), &ip6->ip6_dst); 878 } 879 else { 880 strlcpy(src, "(unknown v6)", sizeof(src)); 881 strlcpy(dst, "(unknown v6)", sizeof(dst)); 882 } 883 log(LOG_INFO, 884 "Connection attempt to TCP [%s]:%d from [%s]:%d\n", 885 dst, ntohs(th->th_dport), 886 src, ntohs(th->th_sport)); 887 } 888 #endif 889 890 /* 891 * Checksum extended TCP header and data. 892 */ 893 int 894 tcp_input_checksum(int af, struct mbuf *m, const struct tcphdr *th, 895 int toff, int off, int tlen) 896 { 897 898 /* 899 * XXX it's better to record and check if this mbuf is 900 * already checked. 901 */ 902 903 switch (af) { 904 #ifdef INET 905 case AF_INET: 906 switch (m->m_pkthdr.csum_flags & 907 ((m->m_pkthdr.rcvif->if_csum_flags_rx & M_CSUM_TCPv4) | 908 M_CSUM_TCP_UDP_BAD | M_CSUM_DATA)) { 909 case M_CSUM_TCPv4|M_CSUM_TCP_UDP_BAD: 910 TCP_CSUM_COUNTER_INCR(&tcp_hwcsum_bad); 911 goto badcsum; 912 913 case M_CSUM_TCPv4|M_CSUM_DATA: { 914 u_int32_t hw_csum = m->m_pkthdr.csum_data; 915 916 TCP_CSUM_COUNTER_INCR(&tcp_hwcsum_data); 917 if (m->m_pkthdr.csum_flags & M_CSUM_NO_PSEUDOHDR) { 918 const struct ip *ip = 919 mtod(m, const struct ip *); 920 921 hw_csum = in_cksum_phdr(ip->ip_src.s_addr, 922 ip->ip_dst.s_addr, 923 htons(hw_csum + tlen + off + IPPROTO_TCP)); 924 } 925 if ((hw_csum ^ 0xffff) != 0) 926 goto badcsum; 927 break; 928 } 929 930 case M_CSUM_TCPv4: 931 /* Checksum was okay. */ 932 TCP_CSUM_COUNTER_INCR(&tcp_hwcsum_ok); 933 break; 934 935 default: 936 /* 937 * Must compute it ourselves. Maybe skip checksum 938 * on loopback interfaces. 939 */ 940 if (__predict_true(!(m->m_pkthdr.rcvif->if_flags & 941 IFF_LOOPBACK) || 942 tcp_do_loopback_cksum)) { 943 TCP_CSUM_COUNTER_INCR(&tcp_swcsum); 944 if (in4_cksum(m, IPPROTO_TCP, toff, 945 tlen + off) != 0) 946 goto badcsum; 947 } 948 break; 949 } 950 break; 951 #endif /* INET4 */ 952 953 #ifdef INET6 954 case AF_INET6: 955 switch (m->m_pkthdr.csum_flags & 956 ((m->m_pkthdr.rcvif->if_csum_flags_rx & M_CSUM_TCPv6) | 957 M_CSUM_TCP_UDP_BAD | M_CSUM_DATA)) { 958 case M_CSUM_TCPv6|M_CSUM_TCP_UDP_BAD: 959 TCP_CSUM_COUNTER_INCR(&tcp6_hwcsum_bad); 960 goto badcsum; 961 962 #if 0 /* notyet */ 963 case M_CSUM_TCPv6|M_CSUM_DATA: 964 #endif 965 966 case M_CSUM_TCPv6: 967 /* Checksum was okay. */ 968 TCP_CSUM_COUNTER_INCR(&tcp6_hwcsum_ok); 969 break; 970 971 default: 972 /* 973 * Must compute it ourselves. Maybe skip checksum 974 * on loopback interfaces. 975 */ 976 if (__predict_true((m->m_flags & M_LOOP) == 0 || 977 tcp_do_loopback_cksum)) { 978 TCP_CSUM_COUNTER_INCR(&tcp6_swcsum); 979 if (in6_cksum(m, IPPROTO_TCP, toff, 980 tlen + off) != 0) 981 goto badcsum; 982 } 983 } 984 break; 985 #endif /* INET6 */ 986 } 987 988 return 0; 989 990 badcsum: 991 TCP_STATINC(TCP_STAT_RCVBADSUM); 992 return -1; 993 } 994 995 /* When a packet arrives addressed to a vestigial tcpbp, we 996 * nevertheless have to respond to it per the spec. 997 */ 998 static void tcp_vtw_input(struct tcphdr *th, vestigial_inpcb_t *vp, 999 struct mbuf *m, int tlen, int multicast) 1000 { 1001 int tiflags; 1002 int todrop; 1003 uint32_t t_flags = 0; 1004 uint64_t *tcps; 1005 1006 tiflags = th->th_flags; 1007 todrop = vp->rcv_nxt - th->th_seq; 1008 1009 if (todrop > 0) { 1010 if (tiflags & TH_SYN) { 1011 tiflags &= ~TH_SYN; 1012 ++th->th_seq; 1013 if (th->th_urp > 1) 1014 --th->th_urp; 1015 else { 1016 tiflags &= ~TH_URG; 1017 th->th_urp = 0; 1018 } 1019 --todrop; 1020 } 1021 if (todrop > tlen || 1022 (todrop == tlen && (tiflags & TH_FIN) == 0)) { 1023 /* 1024 * Any valid FIN or RST must be to the left of the 1025 * window. At this point the FIN or RST must be a 1026 * duplicate or out of sequence; drop it. 1027 */ 1028 if (tiflags & TH_RST) 1029 goto drop; 1030 tiflags &= ~(TH_FIN|TH_RST); 1031 /* 1032 * Send an ACK to resynchronize and drop any data. 1033 * But keep on processing for RST or ACK. 1034 */ 1035 t_flags |= TF_ACKNOW; 1036 todrop = tlen; 1037 tcps = TCP_STAT_GETREF(); 1038 tcps[TCP_STAT_RCVDUPPACK] += 1; 1039 tcps[TCP_STAT_RCVDUPBYTE] += todrop; 1040 TCP_STAT_PUTREF(); 1041 } else if ((tiflags & TH_RST) 1042 && th->th_seq != vp->rcv_nxt) { 1043 /* 1044 * Test for reset before adjusting the sequence 1045 * number for overlapping data. 1046 */ 1047 goto dropafterack_ratelim; 1048 } else { 1049 tcps = TCP_STAT_GETREF(); 1050 tcps[TCP_STAT_RCVPARTDUPPACK] += 1; 1051 tcps[TCP_STAT_RCVPARTDUPBYTE] += todrop; 1052 TCP_STAT_PUTREF(); 1053 } 1054 1055 // tcp_new_dsack(tp, th->th_seq, todrop); 1056 // hdroptlen += todrop; /*drop from head afterwards*/ 1057 1058 th->th_seq += todrop; 1059 tlen -= todrop; 1060 1061 if (th->th_urp > todrop) 1062 th->th_urp -= todrop; 1063 else { 1064 tiflags &= ~TH_URG; 1065 th->th_urp = 0; 1066 } 1067 } 1068 1069 /* 1070 * If new data are received on a connection after the 1071 * user processes are gone, then RST the other end. 1072 */ 1073 if (tlen) { 1074 TCP_STATINC(TCP_STAT_RCVAFTERCLOSE); 1075 goto dropwithreset; 1076 } 1077 1078 /* 1079 * If segment ends after window, drop trailing data 1080 * (and PUSH and FIN); if nothing left, just ACK. 1081 */ 1082 todrop = (th->th_seq + tlen) - (vp->rcv_nxt+vp->rcv_wnd); 1083 1084 if (todrop > 0) { 1085 TCP_STATINC(TCP_STAT_RCVPACKAFTERWIN); 1086 if (todrop >= tlen) { 1087 /* 1088 * The segment actually starts after the window. 1089 * th->th_seq + tlen - vp->rcv_nxt - vp->rcv_wnd >= tlen 1090 * th->th_seq - vp->rcv_nxt - vp->rcv_wnd >= 0 1091 * th->th_seq >= vp->rcv_nxt + vp->rcv_wnd 1092 */ 1093 TCP_STATADD(TCP_STAT_RCVBYTEAFTERWIN, tlen); 1094 /* 1095 * If a new connection request is received 1096 * while in TIME_WAIT, drop the old connection 1097 * and start over if the sequence numbers 1098 * are above the previous ones. 1099 */ 1100 if ((tiflags & TH_SYN) 1101 && SEQ_GT(th->th_seq, vp->rcv_nxt)) { 1102 /* We only support this in the !NOFDREF case, which 1103 * is to say: not here. 1104 */ 1105 goto dropwithreset; 1106 } 1107 /* 1108 * If window is closed can only take segments at 1109 * window edge, and have to drop data and PUSH from 1110 * incoming segments. Continue processing, but 1111 * remember to ack. Otherwise, drop segment 1112 * and (if not RST) ack. 1113 */ 1114 if (vp->rcv_wnd == 0 && th->th_seq == vp->rcv_nxt) { 1115 t_flags |= TF_ACKNOW; 1116 TCP_STATINC(TCP_STAT_RCVWINPROBE); 1117 } else 1118 goto dropafterack; 1119 } else 1120 TCP_STATADD(TCP_STAT_RCVBYTEAFTERWIN, todrop); 1121 m_adj(m, -todrop); 1122 tlen -= todrop; 1123 tiflags &= ~(TH_PUSH|TH_FIN); 1124 } 1125 1126 if (tiflags & TH_RST) { 1127 if (th->th_seq != vp->rcv_nxt) 1128 goto dropafterack_ratelim; 1129 1130 vtw_del(vp->ctl, vp->vtw); 1131 goto drop; 1132 } 1133 1134 /* 1135 * If the ACK bit is off we drop the segment and return. 1136 */ 1137 if ((tiflags & TH_ACK) == 0) { 1138 if (t_flags & TF_ACKNOW) 1139 goto dropafterack; 1140 else 1141 goto drop; 1142 } 1143 1144 /* 1145 * In TIME_WAIT state the only thing that should arrive 1146 * is a retransmission of the remote FIN. Acknowledge 1147 * it and restart the finack timer. 1148 */ 1149 vtw_restart(vp); 1150 goto dropafterack; 1151 1152 dropafterack: 1153 /* 1154 * Generate an ACK dropping incoming segment if it occupies 1155 * sequence space, where the ACK reflects our state. 1156 */ 1157 if (tiflags & TH_RST) 1158 goto drop; 1159 goto dropafterack2; 1160 1161 dropafterack_ratelim: 1162 /* 1163 * We may want to rate-limit ACKs against SYN/RST attack. 1164 */ 1165 if (ppsratecheck(&tcp_ackdrop_ppslim_last, &tcp_ackdrop_ppslim_count, 1166 tcp_ackdrop_ppslim) == 0) { 1167 /* XXX stat */ 1168 goto drop; 1169 } 1170 /* ...fall into dropafterack2... */ 1171 1172 dropafterack2: 1173 (void)tcp_respond(0, m, m, th, th->th_seq + tlen, th->th_ack, 1174 TH_ACK); 1175 return; 1176 1177 dropwithreset: 1178 /* 1179 * Generate a RST, dropping incoming segment. 1180 * Make ACK acceptable to originator of segment. 1181 */ 1182 if (tiflags & TH_RST) 1183 goto drop; 1184 1185 if (tiflags & TH_ACK) 1186 tcp_respond(0, m, m, th, (tcp_seq)0, th->th_ack, TH_RST); 1187 else { 1188 if (tiflags & TH_SYN) 1189 ++tlen; 1190 (void)tcp_respond(0, m, m, th, th->th_seq + tlen, (tcp_seq)0, 1191 TH_RST|TH_ACK); 1192 } 1193 return; 1194 drop: 1195 m_freem(m); 1196 } 1197 1198 /* 1199 * TCP input routine, follows pages 65-76 of RFC 793 very closely. 1200 */ 1201 void 1202 tcp_input(struct mbuf *m, ...) 1203 { 1204 struct tcphdr *th; 1205 struct ip *ip; 1206 struct inpcb *inp; 1207 #ifdef INET6 1208 struct ip6_hdr *ip6; 1209 struct in6pcb *in6p; 1210 #endif 1211 u_int8_t *optp = NULL; 1212 int optlen = 0; 1213 int len, tlen, toff, hdroptlen = 0; 1214 struct tcpcb *tp = 0; 1215 int tiflags; 1216 struct socket *so = NULL; 1217 int todrop, acked, ourfinisacked, needoutput = 0; 1218 bool dupseg; 1219 #ifdef TCP_DEBUG 1220 short ostate = 0; 1221 #endif 1222 u_long tiwin; 1223 struct tcp_opt_info opti; 1224 int off, iphlen; 1225 va_list ap; 1226 int af; /* af on the wire */ 1227 struct mbuf *tcp_saveti = NULL; 1228 uint32_t ts_rtt; 1229 uint8_t iptos; 1230 uint64_t *tcps; 1231 vestigial_inpcb_t vestige; 1232 1233 vestige.valid = 0; 1234 1235 MCLAIM(m, &tcp_rx_mowner); 1236 va_start(ap, m); 1237 toff = va_arg(ap, int); 1238 (void)va_arg(ap, int); /* ignore value, advance ap */ 1239 va_end(ap); 1240 1241 TCP_STATINC(TCP_STAT_RCVTOTAL); 1242 1243 memset(&opti, 0, sizeof(opti)); 1244 opti.ts_present = 0; 1245 opti.maxseg = 0; 1246 1247 /* 1248 * RFC1122 4.2.3.10, p. 104: discard bcast/mcast SYN. 1249 * 1250 * TCP is, by definition, unicast, so we reject all 1251 * multicast outright. 1252 * 1253 * Note, there are additional src/dst address checks in 1254 * the AF-specific code below. 1255 */ 1256 if (m->m_flags & (M_BCAST|M_MCAST)) { 1257 /* XXX stat */ 1258 goto drop; 1259 } 1260 #ifdef INET6 1261 if (m->m_flags & M_ANYCAST6) { 1262 /* XXX stat */ 1263 goto drop; 1264 } 1265 #endif 1266 1267 /* 1268 * Get IP and TCP header. 1269 * Note: IP leaves IP header in first mbuf. 1270 */ 1271 ip = mtod(m, struct ip *); 1272 switch (ip->ip_v) { 1273 #ifdef INET 1274 case 4: 1275 #ifdef INET6 1276 ip6 = NULL; 1277 #endif 1278 af = AF_INET; 1279 iphlen = sizeof(struct ip); 1280 IP6_EXTHDR_GET(th, struct tcphdr *, m, toff, 1281 sizeof(struct tcphdr)); 1282 if (th == NULL) { 1283 TCP_STATINC(TCP_STAT_RCVSHORT); 1284 return; 1285 } 1286 /* We do the checksum after PCB lookup... */ 1287 len = ntohs(ip->ip_len); 1288 tlen = len - toff; 1289 iptos = ip->ip_tos; 1290 break; 1291 #endif 1292 #ifdef INET6 1293 case 6: 1294 ip = NULL; 1295 iphlen = sizeof(struct ip6_hdr); 1296 af = AF_INET6; 1297 ip6 = mtod(m, struct ip6_hdr *); 1298 IP6_EXTHDR_GET(th, struct tcphdr *, m, toff, 1299 sizeof(struct tcphdr)); 1300 if (th == NULL) { 1301 TCP_STATINC(TCP_STAT_RCVSHORT); 1302 return; 1303 } 1304 1305 /* Be proactive about malicious use of IPv4 mapped address */ 1306 if (IN6_IS_ADDR_V4MAPPED(&ip6->ip6_src) || 1307 IN6_IS_ADDR_V4MAPPED(&ip6->ip6_dst)) { 1308 /* XXX stat */ 1309 goto drop; 1310 } 1311 1312 /* 1313 * Be proactive about unspecified IPv6 address in source. 1314 * As we use all-zero to indicate unbounded/unconnected pcb, 1315 * unspecified IPv6 address can be used to confuse us. 1316 * 1317 * Note that packets with unspecified IPv6 destination is 1318 * already dropped in ip6_input. 1319 */ 1320 if (IN6_IS_ADDR_UNSPECIFIED(&ip6->ip6_src)) { 1321 /* XXX stat */ 1322 goto drop; 1323 } 1324 1325 /* 1326 * Make sure destination address is not multicast. 1327 * Source address checked in ip6_input(). 1328 */ 1329 if (IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) { 1330 /* XXX stat */ 1331 goto drop; 1332 } 1333 1334 /* We do the checksum after PCB lookup... */ 1335 len = m->m_pkthdr.len; 1336 tlen = len - toff; 1337 iptos = (ntohl(ip6->ip6_flow) >> 20) & 0xff; 1338 break; 1339 #endif 1340 default: 1341 m_freem(m); 1342 return; 1343 } 1344 1345 KASSERT(TCP_HDR_ALIGNED_P(th)); 1346 1347 /* 1348 * Check that TCP offset makes sense, 1349 * pull out TCP options and adjust length. XXX 1350 */ 1351 off = th->th_off << 2; 1352 if (off < sizeof (struct tcphdr) || off > tlen) { 1353 TCP_STATINC(TCP_STAT_RCVBADOFF); 1354 goto drop; 1355 } 1356 tlen -= off; 1357 1358 /* 1359 * tcp_input() has been modified to use tlen to mean the TCP data 1360 * length throughout the function. Other functions can use 1361 * m->m_pkthdr.len as the basis for calculating the TCP data length. 1362 * rja 1363 */ 1364 1365 if (off > sizeof (struct tcphdr)) { 1366 IP6_EXTHDR_GET(th, struct tcphdr *, m, toff, off); 1367 if (th == NULL) { 1368 TCP_STATINC(TCP_STAT_RCVSHORT); 1369 return; 1370 } 1371 /* 1372 * NOTE: ip/ip6 will not be affected by m_pulldown() 1373 * (as they're before toff) and we don't need to update those. 1374 */ 1375 KASSERT(TCP_HDR_ALIGNED_P(th)); 1376 optlen = off - sizeof (struct tcphdr); 1377 optp = ((u_int8_t *)th) + sizeof(struct tcphdr); 1378 /* 1379 * Do quick retrieval of timestamp options ("options 1380 * prediction?"). If timestamp is the only option and it's 1381 * formatted as recommended in RFC 1323 appendix A, we 1382 * quickly get the values now and not bother calling 1383 * tcp_dooptions(), etc. 1384 */ 1385 if ((optlen == TCPOLEN_TSTAMP_APPA || 1386 (optlen > TCPOLEN_TSTAMP_APPA && 1387 optp[TCPOLEN_TSTAMP_APPA] == TCPOPT_EOL)) && 1388 *(u_int32_t *)optp == htonl(TCPOPT_TSTAMP_HDR) && 1389 (th->th_flags & TH_SYN) == 0) { 1390 opti.ts_present = 1; 1391 opti.ts_val = ntohl(*(u_int32_t *)(optp + 4)); 1392 opti.ts_ecr = ntohl(*(u_int32_t *)(optp + 8)); 1393 optp = NULL; /* we've parsed the options */ 1394 } 1395 } 1396 tiflags = th->th_flags; 1397 1398 /* 1399 * Checksum extended TCP header and data 1400 */ 1401 if (tcp_input_checksum(af, m, th, toff, off, tlen)) 1402 goto badcsum; 1403 1404 /* 1405 * Locate pcb for segment. 1406 */ 1407 findpcb: 1408 inp = NULL; 1409 #ifdef INET6 1410 in6p = NULL; 1411 #endif 1412 switch (af) { 1413 #ifdef INET 1414 case AF_INET: 1415 inp = in_pcblookup_connect(&tcbtable, ip->ip_src, th->th_sport, 1416 ip->ip_dst, th->th_dport, 1417 &vestige); 1418 if (inp == 0 && !vestige.valid) { 1419 TCP_STATINC(TCP_STAT_PCBHASHMISS); 1420 inp = in_pcblookup_bind(&tcbtable, ip->ip_dst, th->th_dport); 1421 } 1422 #ifdef INET6 1423 if (inp == 0 && !vestige.valid) { 1424 struct in6_addr s, d; 1425 1426 /* mapped addr case */ 1427 in6_in_2_v4mapin6(&ip->ip_src, &s); 1428 in6_in_2_v4mapin6(&ip->ip_dst, &d); 1429 in6p = in6_pcblookup_connect(&tcbtable, &s, 1430 th->th_sport, &d, th->th_dport, 1431 0, &vestige); 1432 if (in6p == 0 && !vestige.valid) { 1433 TCP_STATINC(TCP_STAT_PCBHASHMISS); 1434 in6p = in6_pcblookup_bind(&tcbtable, &d, 1435 th->th_dport, 0); 1436 } 1437 } 1438 #endif 1439 #ifndef INET6 1440 if (inp == 0 && !vestige.valid) 1441 #else 1442 if (inp == 0 && in6p == 0 && !vestige.valid) 1443 #endif 1444 { 1445 TCP_STATINC(TCP_STAT_NOPORT); 1446 if (tcp_log_refused && 1447 (tiflags & (TH_RST|TH_ACK|TH_SYN)) == TH_SYN) { 1448 tcp4_log_refused(ip, th); 1449 } 1450 tcp_fields_to_host(th); 1451 goto dropwithreset_ratelim; 1452 } 1453 #if defined(IPSEC) 1454 if (ipsec_used) { 1455 if (inp && 1456 (inp->inp_socket->so_options & SO_ACCEPTCONN) == 0 1457 && ipsec4_in_reject(m, inp)) { 1458 IPSEC_STATINC(IPSEC_STAT_IN_POLVIO); 1459 goto drop; 1460 } 1461 #ifdef INET6 1462 else if (in6p && 1463 (in6p->in6p_socket->so_options & SO_ACCEPTCONN) == 0 1464 && ipsec6_in_reject_so(m, in6p->in6p_socket)) { 1465 IPSEC_STATINC(IPSEC_STAT_IN_POLVIO); 1466 goto drop; 1467 } 1468 #endif 1469 } 1470 #endif /*IPSEC*/ 1471 break; 1472 #endif /*INET*/ 1473 #ifdef INET6 1474 case AF_INET6: 1475 { 1476 int faith; 1477 1478 #if defined(NFAITH) && NFAITH > 0 1479 faith = faithprefix(&ip6->ip6_dst); 1480 #else 1481 faith = 0; 1482 #endif 1483 in6p = in6_pcblookup_connect(&tcbtable, &ip6->ip6_src, 1484 th->th_sport, &ip6->ip6_dst, th->th_dport, faith, &vestige); 1485 if (!in6p && !vestige.valid) { 1486 TCP_STATINC(TCP_STAT_PCBHASHMISS); 1487 in6p = in6_pcblookup_bind(&tcbtable, &ip6->ip6_dst, 1488 th->th_dport, faith); 1489 } 1490 if (!in6p && !vestige.valid) { 1491 TCP_STATINC(TCP_STAT_NOPORT); 1492 if (tcp_log_refused && 1493 (tiflags & (TH_RST|TH_ACK|TH_SYN)) == TH_SYN) { 1494 tcp6_log_refused(ip6, th); 1495 } 1496 tcp_fields_to_host(th); 1497 goto dropwithreset_ratelim; 1498 } 1499 #if defined(IPSEC) 1500 if (ipsec_used && in6p 1501 && (in6p->in6p_socket->so_options & SO_ACCEPTCONN) == 0 1502 && ipsec6_in_reject(m, in6p)) { 1503 IPSEC6_STATINC(IPSEC_STAT_IN_POLVIO); 1504 goto drop; 1505 } 1506 #endif /*IPSEC*/ 1507 break; 1508 } 1509 #endif 1510 } 1511 1512 /* 1513 * If the state is CLOSED (i.e., TCB does not exist) then 1514 * all data in the incoming segment is discarded. 1515 * If the TCB exists but is in CLOSED state, it is embryonic, 1516 * but should either do a listen or a connect soon. 1517 */ 1518 tp = NULL; 1519 so = NULL; 1520 if (inp) { 1521 /* Check the minimum TTL for socket. */ 1522 if (ip->ip_ttl < inp->inp_ip_minttl) 1523 goto drop; 1524 1525 tp = intotcpcb(inp); 1526 so = inp->inp_socket; 1527 } 1528 #ifdef INET6 1529 else if (in6p) { 1530 tp = in6totcpcb(in6p); 1531 so = in6p->in6p_socket; 1532 } 1533 #endif 1534 else if (vestige.valid) { 1535 int mc = 0; 1536 1537 /* We do not support the resurrection of vtw tcpcps. 1538 */ 1539 if (tcp_input_checksum(af, m, th, toff, off, tlen)) 1540 goto badcsum; 1541 1542 switch (af) { 1543 #ifdef INET6 1544 case AF_INET6: 1545 mc = IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst); 1546 break; 1547 #endif 1548 1549 case AF_INET: 1550 mc = (IN_MULTICAST(ip->ip_dst.s_addr) 1551 || in_broadcast(ip->ip_dst, m->m_pkthdr.rcvif)); 1552 break; 1553 } 1554 1555 tcp_fields_to_host(th); 1556 tcp_vtw_input(th, &vestige, m, tlen, mc); 1557 m = 0; 1558 goto drop; 1559 } 1560 1561 if (tp == 0) { 1562 tcp_fields_to_host(th); 1563 goto dropwithreset_ratelim; 1564 } 1565 if (tp->t_state == TCPS_CLOSED) 1566 goto drop; 1567 1568 KASSERT(so->so_lock == softnet_lock); 1569 KASSERT(solocked(so)); 1570 1571 tcp_fields_to_host(th); 1572 1573 /* Unscale the window into a 32-bit value. */ 1574 if ((tiflags & TH_SYN) == 0) 1575 tiwin = th->th_win << tp->snd_scale; 1576 else 1577 tiwin = th->th_win; 1578 1579 #ifdef INET6 1580 /* save packet options if user wanted */ 1581 if (in6p && (in6p->in6p_flags & IN6P_CONTROLOPTS)) { 1582 if (in6p->in6p_options) { 1583 m_freem(in6p->in6p_options); 1584 in6p->in6p_options = 0; 1585 } 1586 KASSERT(ip6 != NULL); 1587 ip6_savecontrol(in6p, &in6p->in6p_options, ip6, m); 1588 } 1589 #endif 1590 1591 if (so->so_options & (SO_DEBUG|SO_ACCEPTCONN)) { 1592 union syn_cache_sa src; 1593 union syn_cache_sa dst; 1594 1595 memset(&src, 0, sizeof(src)); 1596 memset(&dst, 0, sizeof(dst)); 1597 switch (af) { 1598 #ifdef INET 1599 case AF_INET: 1600 src.sin.sin_len = sizeof(struct sockaddr_in); 1601 src.sin.sin_family = AF_INET; 1602 src.sin.sin_addr = ip->ip_src; 1603 src.sin.sin_port = th->th_sport; 1604 1605 dst.sin.sin_len = sizeof(struct sockaddr_in); 1606 dst.sin.sin_family = AF_INET; 1607 dst.sin.sin_addr = ip->ip_dst; 1608 dst.sin.sin_port = th->th_dport; 1609 break; 1610 #endif 1611 #ifdef INET6 1612 case AF_INET6: 1613 src.sin6.sin6_len = sizeof(struct sockaddr_in6); 1614 src.sin6.sin6_family = AF_INET6; 1615 src.sin6.sin6_addr = ip6->ip6_src; 1616 src.sin6.sin6_port = th->th_sport; 1617 1618 dst.sin6.sin6_len = sizeof(struct sockaddr_in6); 1619 dst.sin6.sin6_family = AF_INET6; 1620 dst.sin6.sin6_addr = ip6->ip6_dst; 1621 dst.sin6.sin6_port = th->th_dport; 1622 break; 1623 #endif /* INET6 */ 1624 default: 1625 goto badsyn; /*sanity*/ 1626 } 1627 1628 if (so->so_options & SO_DEBUG) { 1629 #ifdef TCP_DEBUG 1630 ostate = tp->t_state; 1631 #endif 1632 1633 tcp_saveti = NULL; 1634 if (iphlen + sizeof(struct tcphdr) > MHLEN) 1635 goto nosave; 1636 1637 if (m->m_len > iphlen && (m->m_flags & M_EXT) == 0) { 1638 tcp_saveti = m_copym(m, 0, iphlen, M_DONTWAIT); 1639 if (!tcp_saveti) 1640 goto nosave; 1641 } else { 1642 MGETHDR(tcp_saveti, M_DONTWAIT, MT_HEADER); 1643 if (!tcp_saveti) 1644 goto nosave; 1645 MCLAIM(m, &tcp_mowner); 1646 tcp_saveti->m_len = iphlen; 1647 m_copydata(m, 0, iphlen, 1648 mtod(tcp_saveti, void *)); 1649 } 1650 1651 if (M_TRAILINGSPACE(tcp_saveti) < sizeof(struct tcphdr)) { 1652 m_freem(tcp_saveti); 1653 tcp_saveti = NULL; 1654 } else { 1655 tcp_saveti->m_len += sizeof(struct tcphdr); 1656 memcpy(mtod(tcp_saveti, char *) + iphlen, th, 1657 sizeof(struct tcphdr)); 1658 } 1659 nosave:; 1660 } 1661 if (so->so_options & SO_ACCEPTCONN) { 1662 if ((tiflags & (TH_RST|TH_ACK|TH_SYN)) != TH_SYN) { 1663 if (tiflags & TH_RST) { 1664 syn_cache_reset(&src.sa, &dst.sa, th); 1665 } else if ((tiflags & (TH_ACK|TH_SYN)) == 1666 (TH_ACK|TH_SYN)) { 1667 /* 1668 * Received a SYN,ACK. This should 1669 * never happen while we are in 1670 * LISTEN. Send an RST. 1671 */ 1672 goto badsyn; 1673 } else if (tiflags & TH_ACK) { 1674 so = syn_cache_get(&src.sa, &dst.sa, 1675 th, toff, tlen, so, m); 1676 if (so == NULL) { 1677 /* 1678 * We don't have a SYN for 1679 * this ACK; send an RST. 1680 */ 1681 goto badsyn; 1682 } else if (so == 1683 (struct socket *)(-1)) { 1684 /* 1685 * We were unable to create 1686 * the connection. If the 1687 * 3-way handshake was 1688 * completed, and RST has 1689 * been sent to the peer. 1690 * Since the mbuf might be 1691 * in use for the reply, 1692 * do not free it. 1693 */ 1694 m = NULL; 1695 } else { 1696 /* 1697 * We have created a 1698 * full-blown connection. 1699 */ 1700 tp = NULL; 1701 inp = NULL; 1702 #ifdef INET6 1703 in6p = NULL; 1704 #endif 1705 switch (so->so_proto->pr_domain->dom_family) { 1706 #ifdef INET 1707 case AF_INET: 1708 inp = sotoinpcb(so); 1709 tp = intotcpcb(inp); 1710 break; 1711 #endif 1712 #ifdef INET6 1713 case AF_INET6: 1714 in6p = sotoin6pcb(so); 1715 tp = in6totcpcb(in6p); 1716 break; 1717 #endif 1718 } 1719 if (tp == NULL) 1720 goto badsyn; /*XXX*/ 1721 tiwin <<= tp->snd_scale; 1722 goto after_listen; 1723 } 1724 } else { 1725 /* 1726 * None of RST, SYN or ACK was set. 1727 * This is an invalid packet for a 1728 * TCB in LISTEN state. Send a RST. 1729 */ 1730 goto badsyn; 1731 } 1732 } else { 1733 /* 1734 * Received a SYN. 1735 * 1736 * RFC1122 4.2.3.10, p. 104: discard bcast/mcast SYN 1737 */ 1738 if (m->m_flags & (M_BCAST|M_MCAST)) 1739 goto drop; 1740 1741 switch (af) { 1742 #ifdef INET6 1743 case AF_INET6: 1744 if (IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) 1745 goto drop; 1746 break; 1747 #endif /* INET6 */ 1748 case AF_INET: 1749 if (IN_MULTICAST(ip->ip_dst.s_addr) || 1750 in_broadcast(ip->ip_dst, m->m_pkthdr.rcvif)) 1751 goto drop; 1752 break; 1753 } 1754 1755 #ifdef INET6 1756 /* 1757 * If deprecated address is forbidden, we do 1758 * not accept SYN to deprecated interface 1759 * address to prevent any new inbound 1760 * connection from getting established. 1761 * When we do not accept SYN, we send a TCP 1762 * RST, with deprecated source address (instead 1763 * of dropping it). We compromise it as it is 1764 * much better for peer to send a RST, and 1765 * RST will be the final packet for the 1766 * exchange. 1767 * 1768 * If we do not forbid deprecated addresses, we 1769 * accept the SYN packet. RFC2462 does not 1770 * suggest dropping SYN in this case. 1771 * If we decipher RFC2462 5.5.4, it says like 1772 * this: 1773 * 1. use of deprecated addr with existing 1774 * communication is okay - "SHOULD continue 1775 * to be used" 1776 * 2. use of it with new communication: 1777 * (2a) "SHOULD NOT be used if alternate 1778 * address with sufficient scope is 1779 * available" 1780 * (2b) nothing mentioned otherwise. 1781 * Here we fall into (2b) case as we have no 1782 * choice in our source address selection - we 1783 * must obey the peer. 1784 * 1785 * The wording in RFC2462 is confusing, and 1786 * there are multiple description text for 1787 * deprecated address handling - worse, they 1788 * are not exactly the same. I believe 5.5.4 1789 * is the best one, so we follow 5.5.4. 1790 */ 1791 if (af == AF_INET6 && !ip6_use_deprecated) { 1792 struct in6_ifaddr *ia6; 1793 if ((ia6 = in6ifa_ifpwithaddr(m->m_pkthdr.rcvif, 1794 &ip6->ip6_dst)) && 1795 (ia6->ia6_flags & IN6_IFF_DEPRECATED)) { 1796 tp = NULL; 1797 goto dropwithreset; 1798 } 1799 } 1800 #endif 1801 1802 #if defined(IPSEC) 1803 if (ipsec_used) { 1804 switch (af) { 1805 #ifdef INET 1806 case AF_INET: 1807 if (!ipsec4_in_reject_so(m, so)) 1808 break; 1809 IPSEC_STATINC( 1810 IPSEC_STAT_IN_POLVIO); 1811 tp = NULL; 1812 goto dropwithreset; 1813 #endif 1814 #ifdef INET6 1815 case AF_INET6: 1816 if (!ipsec6_in_reject_so(m, so)) 1817 break; 1818 IPSEC6_STATINC( 1819 IPSEC_STAT_IN_POLVIO); 1820 tp = NULL; 1821 goto dropwithreset; 1822 #endif /*INET6*/ 1823 } 1824 } 1825 #endif /*IPSEC*/ 1826 1827 /* 1828 * LISTEN socket received a SYN 1829 * from itself? This can't possibly 1830 * be valid; drop the packet. 1831 */ 1832 if (th->th_sport == th->th_dport) { 1833 int i; 1834 1835 switch (af) { 1836 #ifdef INET 1837 case AF_INET: 1838 i = in_hosteq(ip->ip_src, ip->ip_dst); 1839 break; 1840 #endif 1841 #ifdef INET6 1842 case AF_INET6: 1843 i = IN6_ARE_ADDR_EQUAL(&ip6->ip6_src, &ip6->ip6_dst); 1844 break; 1845 #endif 1846 default: 1847 i = 1; 1848 } 1849 if (i) { 1850 TCP_STATINC(TCP_STAT_BADSYN); 1851 goto drop; 1852 } 1853 } 1854 1855 /* 1856 * SYN looks ok; create compressed TCP 1857 * state for it. 1858 */ 1859 if (so->so_qlen <= so->so_qlimit && 1860 syn_cache_add(&src.sa, &dst.sa, th, tlen, 1861 so, m, optp, optlen, &opti)) 1862 m = NULL; 1863 } 1864 goto drop; 1865 } 1866 } 1867 1868 after_listen: 1869 #ifdef DIAGNOSTIC 1870 /* 1871 * Should not happen now that all embryonic connections 1872 * are handled with compressed state. 1873 */ 1874 if (tp->t_state == TCPS_LISTEN) 1875 panic("tcp_input: TCPS_LISTEN"); 1876 #endif 1877 1878 /* 1879 * Segment received on connection. 1880 * Reset idle time and keep-alive timer. 1881 */ 1882 tp->t_rcvtime = tcp_now; 1883 if (TCPS_HAVEESTABLISHED(tp->t_state)) 1884 TCP_TIMER_ARM(tp, TCPT_KEEP, tp->t_keepidle); 1885 1886 /* 1887 * Process options. 1888 */ 1889 #ifdef TCP_SIGNATURE 1890 if (optp || (tp->t_flags & TF_SIGNATURE)) 1891 #else 1892 if (optp) 1893 #endif 1894 if (tcp_dooptions(tp, optp, optlen, th, m, toff, &opti) < 0) 1895 goto drop; 1896 1897 if (TCP_SACK_ENABLED(tp)) { 1898 tcp_del_sackholes(tp, th); 1899 } 1900 1901 if (TCP_ECN_ALLOWED(tp)) { 1902 if (tiflags & TH_CWR) { 1903 tp->t_flags &= ~TF_ECN_SND_ECE; 1904 } 1905 switch (iptos & IPTOS_ECN_MASK) { 1906 case IPTOS_ECN_CE: 1907 tp->t_flags |= TF_ECN_SND_ECE; 1908 TCP_STATINC(TCP_STAT_ECN_CE); 1909 break; 1910 case IPTOS_ECN_ECT0: 1911 TCP_STATINC(TCP_STAT_ECN_ECT); 1912 break; 1913 case IPTOS_ECN_ECT1: 1914 /* XXX: ignore for now -- rpaulo */ 1915 break; 1916 } 1917 /* 1918 * Congestion experienced. 1919 * Ignore if we are already trying to recover. 1920 */ 1921 if ((tiflags & TH_ECE) && SEQ_GEQ(tp->snd_una, tp->snd_recover)) 1922 tp->t_congctl->cong_exp(tp); 1923 } 1924 1925 if (opti.ts_present && opti.ts_ecr) { 1926 /* 1927 * Calculate the RTT from the returned time stamp and the 1928 * connection's time base. If the time stamp is later than 1929 * the current time, or is extremely old, fall back to non-1323 1930 * RTT calculation. Since ts_rtt is unsigned, we can test both 1931 * at the same time. 1932 * 1933 * Note that ts_rtt is in units of slow ticks (500 1934 * ms). Since most earthbound RTTs are < 500 ms, 1935 * observed values will have large quantization noise. 1936 * Our smoothed RTT is then the fraction of observed 1937 * samples that are 1 tick instead of 0 (times 500 1938 * ms). 1939 * 1940 * ts_rtt is increased by 1 to denote a valid sample, 1941 * with 0 indicating an invalid measurement. This 1942 * extra 1 must be removed when ts_rtt is used, or 1943 * else an an erroneous extra 500 ms will result. 1944 */ 1945 ts_rtt = TCP_TIMESTAMP(tp) - opti.ts_ecr + 1; 1946 if (ts_rtt > TCP_PAWS_IDLE) 1947 ts_rtt = 0; 1948 } else { 1949 ts_rtt = 0; 1950 } 1951 1952 /* 1953 * Header prediction: check for the two common cases 1954 * of a uni-directional data xfer. If the packet has 1955 * no control flags, is in-sequence, the window didn't 1956 * change and we're not retransmitting, it's a 1957 * candidate. If the length is zero and the ack moved 1958 * forward, we're the sender side of the xfer. Just 1959 * free the data acked & wake any higher level process 1960 * that was blocked waiting for space. If the length 1961 * is non-zero and the ack didn't move, we're the 1962 * receiver side. If we're getting packets in-order 1963 * (the reassembly queue is empty), add the data to 1964 * the socket buffer and note that we need a delayed ack. 1965 */ 1966 if (tp->t_state == TCPS_ESTABLISHED && 1967 (tiflags & (TH_SYN|TH_FIN|TH_RST|TH_URG|TH_ECE|TH_CWR|TH_ACK)) 1968 == TH_ACK && 1969 (!opti.ts_present || TSTMP_GEQ(opti.ts_val, tp->ts_recent)) && 1970 th->th_seq == tp->rcv_nxt && 1971 tiwin && tiwin == tp->snd_wnd && 1972 tp->snd_nxt == tp->snd_max) { 1973 1974 /* 1975 * If last ACK falls within this segment's sequence numbers, 1976 * record the timestamp. 1977 * NOTE that the test is modified according to the latest 1978 * proposal of the tcplw (at) cray.com list (Braden 1993/04/26). 1979 * 1980 * note that we already know 1981 * TSTMP_GEQ(opti.ts_val, tp->ts_recent) 1982 */ 1983 if (opti.ts_present && 1984 SEQ_LEQ(th->th_seq, tp->last_ack_sent)) { 1985 tp->ts_recent_age = tcp_now; 1986 tp->ts_recent = opti.ts_val; 1987 } 1988 1989 if (tlen == 0) { 1990 /* Ack prediction. */ 1991 if (SEQ_GT(th->th_ack, tp->snd_una) && 1992 SEQ_LEQ(th->th_ack, tp->snd_max) && 1993 tp->snd_cwnd >= tp->snd_wnd && 1994 tp->t_partialacks < 0) { 1995 /* 1996 * this is a pure ack for outstanding data. 1997 */ 1998 if (ts_rtt) 1999 tcp_xmit_timer(tp, ts_rtt - 1); 2000 else if (tp->t_rtttime && 2001 SEQ_GT(th->th_ack, tp->t_rtseq)) 2002 tcp_xmit_timer(tp, 2003 tcp_now - tp->t_rtttime); 2004 acked = th->th_ack - tp->snd_una; 2005 tcps = TCP_STAT_GETREF(); 2006 tcps[TCP_STAT_PREDACK]++; 2007 tcps[TCP_STAT_RCVACKPACK]++; 2008 tcps[TCP_STAT_RCVACKBYTE] += acked; 2009 TCP_STAT_PUTREF(); 2010 nd6_hint(tp); 2011 2012 if (acked > (tp->t_lastoff - tp->t_inoff)) 2013 tp->t_lastm = NULL; 2014 sbdrop(&so->so_snd, acked); 2015 tp->t_lastoff -= acked; 2016 2017 icmp_check(tp, th, acked); 2018 2019 tp->snd_una = th->th_ack; 2020 tp->snd_fack = tp->snd_una; 2021 if (SEQ_LT(tp->snd_high, tp->snd_una)) 2022 tp->snd_high = tp->snd_una; 2023 m_freem(m); 2024 2025 /* 2026 * If all outstanding data are acked, stop 2027 * retransmit timer, otherwise restart timer 2028 * using current (possibly backed-off) value. 2029 * If process is waiting for space, 2030 * wakeup/selnotify/signal. If data 2031 * are ready to send, let tcp_output 2032 * decide between more output or persist. 2033 */ 2034 if (tp->snd_una == tp->snd_max) 2035 TCP_TIMER_DISARM(tp, TCPT_REXMT); 2036 else if (TCP_TIMER_ISARMED(tp, 2037 TCPT_PERSIST) == 0) 2038 TCP_TIMER_ARM(tp, TCPT_REXMT, 2039 tp->t_rxtcur); 2040 2041 sowwakeup(so); 2042 if (so->so_snd.sb_cc) { 2043 KERNEL_LOCK(1, NULL); 2044 (void) tcp_output(tp); 2045 KERNEL_UNLOCK_ONE(NULL); 2046 } 2047 if (tcp_saveti) 2048 m_freem(tcp_saveti); 2049 return; 2050 } 2051 } else if (th->th_ack == tp->snd_una && 2052 TAILQ_FIRST(&tp->segq) == NULL && 2053 tlen <= sbspace(&so->so_rcv)) { 2054 int newsize = 0; /* automatic sockbuf scaling */ 2055 2056 /* 2057 * this is a pure, in-sequence data packet 2058 * with nothing on the reassembly queue and 2059 * we have enough buffer space to take it. 2060 */ 2061 tp->rcv_nxt += tlen; 2062 tcps = TCP_STAT_GETREF(); 2063 tcps[TCP_STAT_PREDDAT]++; 2064 tcps[TCP_STAT_RCVPACK]++; 2065 tcps[TCP_STAT_RCVBYTE] += tlen; 2066 TCP_STAT_PUTREF(); 2067 nd6_hint(tp); 2068 2069 /* 2070 * Automatic sizing enables the performance of large buffers 2071 * and most of the efficiency of small ones by only allocating 2072 * space when it is needed. 2073 * 2074 * On the receive side the socket buffer memory is only rarely 2075 * used to any significant extent. This allows us to be much 2076 * more aggressive in scaling the receive socket buffer. For 2077 * the case that the buffer space is actually used to a large 2078 * extent and we run out of kernel memory we can simply drop 2079 * the new segments; TCP on the sender will just retransmit it 2080 * later. Setting the buffer size too big may only consume too 2081 * much kernel memory if the application doesn't read() from 2082 * the socket or packet loss or reordering makes use of the 2083 * reassembly queue. 2084 * 2085 * The criteria to step up the receive buffer one notch are: 2086 * 1. the number of bytes received during the time it takes 2087 * one timestamp to be reflected back to us (the RTT); 2088 * 2. received bytes per RTT is within seven eighth of the 2089 * current socket buffer size; 2090 * 3. receive buffer size has not hit maximal automatic size; 2091 * 2092 * This algorithm does one step per RTT at most and only if 2093 * we receive a bulk stream w/o packet losses or reorderings. 2094 * Shrinking the buffer during idle times is not necessary as 2095 * it doesn't consume any memory when idle. 2096 * 2097 * TODO: Only step up if the application is actually serving 2098 * the buffer to better manage the socket buffer resources. 2099 */ 2100 if (tcp_do_autorcvbuf && 2101 opti.ts_ecr && 2102 (so->so_rcv.sb_flags & SB_AUTOSIZE)) { 2103 if (opti.ts_ecr > tp->rfbuf_ts && 2104 opti.ts_ecr - tp->rfbuf_ts < PR_SLOWHZ) { 2105 if (tp->rfbuf_cnt > 2106 (so->so_rcv.sb_hiwat / 8 * 7) && 2107 so->so_rcv.sb_hiwat < 2108 tcp_autorcvbuf_max) { 2109 newsize = 2110 min(so->so_rcv.sb_hiwat + 2111 tcp_autorcvbuf_inc, 2112 tcp_autorcvbuf_max); 2113 } 2114 /* Start over with next RTT. */ 2115 tp->rfbuf_ts = 0; 2116 tp->rfbuf_cnt = 0; 2117 } else 2118 tp->rfbuf_cnt += tlen; /* add up */ 2119 } 2120 2121 /* 2122 * Drop TCP, IP headers and TCP options then add data 2123 * to socket buffer. 2124 */ 2125 if (so->so_state & SS_CANTRCVMORE) 2126 m_freem(m); 2127 else { 2128 /* 2129 * Set new socket buffer size. 2130 * Give up when limit is reached. 2131 */ 2132 if (newsize) 2133 if (!sbreserve(&so->so_rcv, 2134 newsize, so)) 2135 so->so_rcv.sb_flags &= ~SB_AUTOSIZE; 2136 m_adj(m, toff + off); 2137 sbappendstream(&so->so_rcv, m); 2138 } 2139 sorwakeup(so); 2140 tcp_setup_ack(tp, th); 2141 if (tp->t_flags & TF_ACKNOW) { 2142 KERNEL_LOCK(1, NULL); 2143 (void) tcp_output(tp); 2144 KERNEL_UNLOCK_ONE(NULL); 2145 } 2146 if (tcp_saveti) 2147 m_freem(tcp_saveti); 2148 return; 2149 } 2150 } 2151 2152 /* 2153 * Compute mbuf offset to TCP data segment. 2154 */ 2155 hdroptlen = toff + off; 2156 2157 /* 2158 * Calculate amount of space in receive window, 2159 * and then do TCP input processing. 2160 * Receive window is amount of space in rcv queue, 2161 * but not less than advertised window. 2162 */ 2163 { int win; 2164 2165 win = sbspace(&so->so_rcv); 2166 if (win < 0) 2167 win = 0; 2168 tp->rcv_wnd = imax(win, (int)(tp->rcv_adv - tp->rcv_nxt)); 2169 } 2170 2171 /* Reset receive buffer auto scaling when not in bulk receive mode. */ 2172 tp->rfbuf_ts = 0; 2173 tp->rfbuf_cnt = 0; 2174 2175 switch (tp->t_state) { 2176 /* 2177 * If the state is SYN_SENT: 2178 * if seg contains an ACK, but not for our SYN, drop the input. 2179 * if seg contains a RST, then drop the connection. 2180 * if seg does not contain SYN, then drop it. 2181 * Otherwise this is an acceptable SYN segment 2182 * initialize tp->rcv_nxt and tp->irs 2183 * if seg contains ack then advance tp->snd_una 2184 * if seg contains a ECE and ECN support is enabled, the stream 2185 * is ECN capable. 2186 * if SYN has been acked change to ESTABLISHED else SYN_RCVD state 2187 * arrange for segment to be acked (eventually) 2188 * continue processing rest of data/controls, beginning with URG 2189 */ 2190 case TCPS_SYN_SENT: 2191 if ((tiflags & TH_ACK) && 2192 (SEQ_LEQ(th->th_ack, tp->iss) || 2193 SEQ_GT(th->th_ack, tp->snd_max))) 2194 goto dropwithreset; 2195 if (tiflags & TH_RST) { 2196 if (tiflags & TH_ACK) 2197 tp = tcp_drop(tp, ECONNREFUSED); 2198 goto drop; 2199 } 2200 if ((tiflags & TH_SYN) == 0) 2201 goto drop; 2202 if (tiflags & TH_ACK) { 2203 tp->snd_una = th->th_ack; 2204 if (SEQ_LT(tp->snd_nxt, tp->snd_una)) 2205 tp->snd_nxt = tp->snd_una; 2206 if (SEQ_LT(tp->snd_high, tp->snd_una)) 2207 tp->snd_high = tp->snd_una; 2208 TCP_TIMER_DISARM(tp, TCPT_REXMT); 2209 2210 if ((tiflags & TH_ECE) && tcp_do_ecn) { 2211 tp->t_flags |= TF_ECN_PERMIT; 2212 TCP_STATINC(TCP_STAT_ECN_SHS); 2213 } 2214 2215 } 2216 tp->irs = th->th_seq; 2217 tcp_rcvseqinit(tp); 2218 tp->t_flags |= TF_ACKNOW; 2219 tcp_mss_from_peer(tp, opti.maxseg); 2220 2221 /* 2222 * Initialize the initial congestion window. If we 2223 * had to retransmit the SYN, we must initialize cwnd 2224 * to 1 segment (i.e. the Loss Window). 2225 */ 2226 if (tp->t_flags & TF_SYN_REXMT) 2227 tp->snd_cwnd = tp->t_peermss; 2228 else { 2229 int ss = tcp_init_win; 2230 #ifdef INET 2231 if (inp != NULL && in_localaddr(inp->inp_faddr)) 2232 ss = tcp_init_win_local; 2233 #endif 2234 #ifdef INET6 2235 if (in6p != NULL && in6_localaddr(&in6p->in6p_faddr)) 2236 ss = tcp_init_win_local; 2237 #endif 2238 tp->snd_cwnd = TCP_INITIAL_WINDOW(ss, tp->t_peermss); 2239 } 2240 2241 tcp_rmx_rtt(tp); 2242 if (tiflags & TH_ACK) { 2243 TCP_STATINC(TCP_STAT_CONNECTS); 2244 /* 2245 * move tcp_established before soisconnected 2246 * because upcall handler can drive tcp_output 2247 * functionality. 2248 * XXX we might call soisconnected at the end of 2249 * all processing 2250 */ 2251 tcp_established(tp); 2252 soisconnected(so); 2253 /* Do window scaling on this connection? */ 2254 if ((tp->t_flags & (TF_RCVD_SCALE|TF_REQ_SCALE)) == 2255 (TF_RCVD_SCALE|TF_REQ_SCALE)) { 2256 tp->snd_scale = tp->requested_s_scale; 2257 tp->rcv_scale = tp->request_r_scale; 2258 } 2259 TCP_REASS_LOCK(tp); 2260 (void) tcp_reass(tp, NULL, NULL, &tlen); 2261 /* 2262 * if we didn't have to retransmit the SYN, 2263 * use its rtt as our initial srtt & rtt var. 2264 */ 2265 if (tp->t_rtttime) 2266 tcp_xmit_timer(tp, tcp_now - tp->t_rtttime); 2267 } else 2268 tp->t_state = TCPS_SYN_RECEIVED; 2269 2270 /* 2271 * Advance th->th_seq to correspond to first data byte. 2272 * If data, trim to stay within window, 2273 * dropping FIN if necessary. 2274 */ 2275 th->th_seq++; 2276 if (tlen > tp->rcv_wnd) { 2277 todrop = tlen - tp->rcv_wnd; 2278 m_adj(m, -todrop); 2279 tlen = tp->rcv_wnd; 2280 tiflags &= ~TH_FIN; 2281 tcps = TCP_STAT_GETREF(); 2282 tcps[TCP_STAT_RCVPACKAFTERWIN]++; 2283 tcps[TCP_STAT_RCVBYTEAFTERWIN] += todrop; 2284 TCP_STAT_PUTREF(); 2285 } 2286 tp->snd_wl1 = th->th_seq - 1; 2287 tp->rcv_up = th->th_seq; 2288 goto step6; 2289 2290 /* 2291 * If the state is SYN_RECEIVED: 2292 * If seg contains an ACK, but not for our SYN, drop the input 2293 * and generate an RST. See page 36, rfc793 2294 */ 2295 case TCPS_SYN_RECEIVED: 2296 if ((tiflags & TH_ACK) && 2297 (SEQ_LEQ(th->th_ack, tp->iss) || 2298 SEQ_GT(th->th_ack, tp->snd_max))) 2299 goto dropwithreset; 2300 break; 2301 } 2302 2303 /* 2304 * States other than LISTEN or SYN_SENT. 2305 * First check timestamp, if present. 2306 * Then check that at least some bytes of segment are within 2307 * receive window. If segment begins before rcv_nxt, 2308 * drop leading data (and SYN); if nothing left, just ack. 2309 * 2310 * RFC 1323 PAWS: If we have a timestamp reply on this segment 2311 * and it's less than ts_recent, drop it. 2312 */ 2313 if (opti.ts_present && (tiflags & TH_RST) == 0 && tp->ts_recent && 2314 TSTMP_LT(opti.ts_val, tp->ts_recent)) { 2315 2316 /* Check to see if ts_recent is over 24 days old. */ 2317 if (tcp_now - tp->ts_recent_age > TCP_PAWS_IDLE) { 2318 /* 2319 * Invalidate ts_recent. If this segment updates 2320 * ts_recent, the age will be reset later and ts_recent 2321 * will get a valid value. If it does not, setting 2322 * ts_recent to zero will at least satisfy the 2323 * requirement that zero be placed in the timestamp 2324 * echo reply when ts_recent isn't valid. The 2325 * age isn't reset until we get a valid ts_recent 2326 * because we don't want out-of-order segments to be 2327 * dropped when ts_recent is old. 2328 */ 2329 tp->ts_recent = 0; 2330 } else { 2331 tcps = TCP_STAT_GETREF(); 2332 tcps[TCP_STAT_RCVDUPPACK]++; 2333 tcps[TCP_STAT_RCVDUPBYTE] += tlen; 2334 tcps[TCP_STAT_PAWSDROP]++; 2335 TCP_STAT_PUTREF(); 2336 tcp_new_dsack(tp, th->th_seq, tlen); 2337 goto dropafterack; 2338 } 2339 } 2340 2341 todrop = tp->rcv_nxt - th->th_seq; 2342 dupseg = false; 2343 if (todrop > 0) { 2344 if (tiflags & TH_SYN) { 2345 tiflags &= ~TH_SYN; 2346 th->th_seq++; 2347 if (th->th_urp > 1) 2348 th->th_urp--; 2349 else { 2350 tiflags &= ~TH_URG; 2351 th->th_urp = 0; 2352 } 2353 todrop--; 2354 } 2355 if (todrop > tlen || 2356 (todrop == tlen && (tiflags & TH_FIN) == 0)) { 2357 /* 2358 * Any valid FIN or RST must be to the left of the 2359 * window. At this point the FIN or RST must be a 2360 * duplicate or out of sequence; drop it. 2361 */ 2362 if (tiflags & TH_RST) 2363 goto drop; 2364 tiflags &= ~(TH_FIN|TH_RST); 2365 /* 2366 * Send an ACK to resynchronize and drop any data. 2367 * But keep on processing for RST or ACK. 2368 */ 2369 tp->t_flags |= TF_ACKNOW; 2370 todrop = tlen; 2371 dupseg = true; 2372 tcps = TCP_STAT_GETREF(); 2373 tcps[TCP_STAT_RCVDUPPACK]++; 2374 tcps[TCP_STAT_RCVDUPBYTE] += todrop; 2375 TCP_STAT_PUTREF(); 2376 } else if ((tiflags & TH_RST) && 2377 th->th_seq != tp->rcv_nxt) { 2378 /* 2379 * Test for reset before adjusting the sequence 2380 * number for overlapping data. 2381 */ 2382 goto dropafterack_ratelim; 2383 } else { 2384 tcps = TCP_STAT_GETREF(); 2385 tcps[TCP_STAT_RCVPARTDUPPACK]++; 2386 tcps[TCP_STAT_RCVPARTDUPBYTE] += todrop; 2387 TCP_STAT_PUTREF(); 2388 } 2389 tcp_new_dsack(tp, th->th_seq, todrop); 2390 hdroptlen += todrop; /*drop from head afterwards*/ 2391 th->th_seq += todrop; 2392 tlen -= todrop; 2393 if (th->th_urp > todrop) 2394 th->th_urp -= todrop; 2395 else { 2396 tiflags &= ~TH_URG; 2397 th->th_urp = 0; 2398 } 2399 } 2400 2401 /* 2402 * If new data are received on a connection after the 2403 * user processes are gone, then RST the other end. 2404 */ 2405 if ((so->so_state & SS_NOFDREF) && 2406 tp->t_state > TCPS_CLOSE_WAIT && tlen) { 2407 tp = tcp_close(tp); 2408 TCP_STATINC(TCP_STAT_RCVAFTERCLOSE); 2409 goto dropwithreset; 2410 } 2411 2412 /* 2413 * If segment ends after window, drop trailing data 2414 * (and PUSH and FIN); if nothing left, just ACK. 2415 */ 2416 todrop = (th->th_seq + tlen) - (tp->rcv_nxt+tp->rcv_wnd); 2417 if (todrop > 0) { 2418 TCP_STATINC(TCP_STAT_RCVPACKAFTERWIN); 2419 if (todrop >= tlen) { 2420 /* 2421 * The segment actually starts after the window. 2422 * th->th_seq + tlen - tp->rcv_nxt - tp->rcv_wnd >= tlen 2423 * th->th_seq - tp->rcv_nxt - tp->rcv_wnd >= 0 2424 * th->th_seq >= tp->rcv_nxt + tp->rcv_wnd 2425 */ 2426 TCP_STATADD(TCP_STAT_RCVBYTEAFTERWIN, tlen); 2427 /* 2428 * If a new connection request is received 2429 * while in TIME_WAIT, drop the old connection 2430 * and start over if the sequence numbers 2431 * are above the previous ones. 2432 * 2433 * NOTE: We will checksum the packet again, and 2434 * so we need to put the header fields back into 2435 * network order! 2436 * XXX This kind of sucks, but we don't expect 2437 * XXX this to happen very often, so maybe it 2438 * XXX doesn't matter so much. 2439 */ 2440 if (tiflags & TH_SYN && 2441 tp->t_state == TCPS_TIME_WAIT && 2442 SEQ_GT(th->th_seq, tp->rcv_nxt)) { 2443 tp = tcp_close(tp); 2444 tcp_fields_to_net(th); 2445 goto findpcb; 2446 } 2447 /* 2448 * If window is closed can only take segments at 2449 * window edge, and have to drop data and PUSH from 2450 * incoming segments. Continue processing, but 2451 * remember to ack. Otherwise, drop segment 2452 * and (if not RST) ack. 2453 */ 2454 if (tp->rcv_wnd == 0 && th->th_seq == tp->rcv_nxt) { 2455 tp->t_flags |= TF_ACKNOW; 2456 TCP_STATINC(TCP_STAT_RCVWINPROBE); 2457 } else 2458 goto dropafterack; 2459 } else 2460 TCP_STATADD(TCP_STAT_RCVBYTEAFTERWIN, todrop); 2461 m_adj(m, -todrop); 2462 tlen -= todrop; 2463 tiflags &= ~(TH_PUSH|TH_FIN); 2464 } 2465 2466 /* 2467 * If last ACK falls within this segment's sequence numbers, 2468 * record the timestamp. 2469 * NOTE: 2470 * 1) That the test incorporates suggestions from the latest 2471 * proposal of the tcplw (at) cray.com list (Braden 1993/04/26). 2472 * 2) That updating only on newer timestamps interferes with 2473 * our earlier PAWS tests, so this check should be solely 2474 * predicated on the sequence space of this segment. 2475 * 3) That we modify the segment boundary check to be 2476 * Last.ACK.Sent <= SEG.SEQ + SEG.Len 2477 * instead of RFC1323's 2478 * Last.ACK.Sent < SEG.SEQ + SEG.Len, 2479 * This modified check allows us to overcome RFC1323's 2480 * limitations as described in Stevens TCP/IP Illustrated 2481 * Vol. 2 p.869. In such cases, we can still calculate the 2482 * RTT correctly when RCV.NXT == Last.ACK.Sent. 2483 */ 2484 if (opti.ts_present && 2485 SEQ_LEQ(th->th_seq, tp->last_ack_sent) && 2486 SEQ_LEQ(tp->last_ack_sent, th->th_seq + tlen + 2487 ((tiflags & (TH_SYN|TH_FIN)) != 0))) { 2488 tp->ts_recent_age = tcp_now; 2489 tp->ts_recent = opti.ts_val; 2490 } 2491 2492 /* 2493 * If the RST bit is set examine the state: 2494 * SYN_RECEIVED STATE: 2495 * If passive open, return to LISTEN state. 2496 * If active open, inform user that connection was refused. 2497 * ESTABLISHED, FIN_WAIT_1, FIN_WAIT2, CLOSE_WAIT STATES: 2498 * Inform user that connection was reset, and close tcb. 2499 * CLOSING, LAST_ACK, TIME_WAIT STATES 2500 * Close the tcb. 2501 */ 2502 if (tiflags & TH_RST) { 2503 if (th->th_seq != tp->rcv_nxt) 2504 goto dropafterack_ratelim; 2505 2506 switch (tp->t_state) { 2507 case TCPS_SYN_RECEIVED: 2508 so->so_error = ECONNREFUSED; 2509 goto close; 2510 2511 case TCPS_ESTABLISHED: 2512 case TCPS_FIN_WAIT_1: 2513 case TCPS_FIN_WAIT_2: 2514 case TCPS_CLOSE_WAIT: 2515 so->so_error = ECONNRESET; 2516 close: 2517 tp->t_state = TCPS_CLOSED; 2518 TCP_STATINC(TCP_STAT_DROPS); 2519 tp = tcp_close(tp); 2520 goto drop; 2521 2522 case TCPS_CLOSING: 2523 case TCPS_LAST_ACK: 2524 case TCPS_TIME_WAIT: 2525 tp = tcp_close(tp); 2526 goto drop; 2527 } 2528 } 2529 2530 /* 2531 * Since we've covered the SYN-SENT and SYN-RECEIVED states above 2532 * we must be in a synchronized state. RFC791 states (under RST 2533 * generation) that any unacceptable segment (an out-of-order SYN 2534 * qualifies) received in a synchronized state must elicit only an 2535 * empty acknowledgment segment ... and the connection remains in 2536 * the same state. 2537 */ 2538 if (tiflags & TH_SYN) { 2539 if (tp->rcv_nxt == th->th_seq) { 2540 tcp_respond(tp, m, m, th, (tcp_seq)0, th->th_ack - 1, 2541 TH_ACK); 2542 if (tcp_saveti) 2543 m_freem(tcp_saveti); 2544 return; 2545 } 2546 2547 goto dropafterack_ratelim; 2548 } 2549 2550 /* 2551 * If the ACK bit is off we drop the segment and return. 2552 */ 2553 if ((tiflags & TH_ACK) == 0) { 2554 if (tp->t_flags & TF_ACKNOW) 2555 goto dropafterack; 2556 else 2557 goto drop; 2558 } 2559 2560 /* 2561 * Ack processing. 2562 */ 2563 switch (tp->t_state) { 2564 2565 /* 2566 * In SYN_RECEIVED state if the ack ACKs our SYN then enter 2567 * ESTABLISHED state and continue processing, otherwise 2568 * send an RST. 2569 */ 2570 case TCPS_SYN_RECEIVED: 2571 if (SEQ_GT(tp->snd_una, th->th_ack) || 2572 SEQ_GT(th->th_ack, tp->snd_max)) 2573 goto dropwithreset; 2574 TCP_STATINC(TCP_STAT_CONNECTS); 2575 soisconnected(so); 2576 tcp_established(tp); 2577 /* Do window scaling? */ 2578 if ((tp->t_flags & (TF_RCVD_SCALE|TF_REQ_SCALE)) == 2579 (TF_RCVD_SCALE|TF_REQ_SCALE)) { 2580 tp->snd_scale = tp->requested_s_scale; 2581 tp->rcv_scale = tp->request_r_scale; 2582 } 2583 TCP_REASS_LOCK(tp); 2584 (void) tcp_reass(tp, NULL, NULL, &tlen); 2585 tp->snd_wl1 = th->th_seq - 1; 2586 /* fall into ... */ 2587 2588 /* 2589 * In ESTABLISHED state: drop duplicate ACKs; ACK out of range 2590 * ACKs. If the ack is in the range 2591 * tp->snd_una < th->th_ack <= tp->snd_max 2592 * then advance tp->snd_una to th->th_ack and drop 2593 * data from the retransmission queue. If this ACK reflects 2594 * more up to date window information we update our window information. 2595 */ 2596 case TCPS_ESTABLISHED: 2597 case TCPS_FIN_WAIT_1: 2598 case TCPS_FIN_WAIT_2: 2599 case TCPS_CLOSE_WAIT: 2600 case TCPS_CLOSING: 2601 case TCPS_LAST_ACK: 2602 case TCPS_TIME_WAIT: 2603 2604 if (SEQ_LEQ(th->th_ack, tp->snd_una)) { 2605 if (tlen == 0 && !dupseg && tiwin == tp->snd_wnd) { 2606 TCP_STATINC(TCP_STAT_RCVDUPACK); 2607 /* 2608 * If we have outstanding data (other than 2609 * a window probe), this is a completely 2610 * duplicate ack (ie, window info didn't 2611 * change), the ack is the biggest we've 2612 * seen and we've seen exactly our rexmt 2613 * threshhold of them, assume a packet 2614 * has been dropped and retransmit it. 2615 * Kludge snd_nxt & the congestion 2616 * window so we send only this one 2617 * packet. 2618 */ 2619 if (TCP_TIMER_ISARMED(tp, TCPT_REXMT) == 0 || 2620 th->th_ack != tp->snd_una) 2621 tp->t_dupacks = 0; 2622 else if (tp->t_partialacks < 0 && 2623 (++tp->t_dupacks == tcprexmtthresh || 2624 TCP_FACK_FASTRECOV(tp))) { 2625 /* 2626 * Do the fast retransmit, and adjust 2627 * congestion control paramenters. 2628 */ 2629 if (tp->t_congctl->fast_retransmit(tp, th)) { 2630 /* False fast retransmit */ 2631 break; 2632 } else 2633 goto drop; 2634 } else if (tp->t_dupacks > tcprexmtthresh) { 2635 tp->snd_cwnd += tp->t_segsz; 2636 KERNEL_LOCK(1, NULL); 2637 (void) tcp_output(tp); 2638 KERNEL_UNLOCK_ONE(NULL); 2639 goto drop; 2640 } 2641 } else { 2642 /* 2643 * If the ack appears to be very old, only 2644 * allow data that is in-sequence. This 2645 * makes it somewhat more difficult to insert 2646 * forged data by guessing sequence numbers. 2647 * Sent an ack to try to update the send 2648 * sequence number on the other side. 2649 */ 2650 if (tlen && th->th_seq != tp->rcv_nxt && 2651 SEQ_LT(th->th_ack, 2652 tp->snd_una - tp->max_sndwnd)) 2653 goto dropafterack; 2654 } 2655 break; 2656 } 2657 /* 2658 * If the congestion window was inflated to account 2659 * for the other side's cached packets, retract it. 2660 */ 2661 tp->t_congctl->fast_retransmit_newack(tp, th); 2662 2663 if (SEQ_GT(th->th_ack, tp->snd_max)) { 2664 TCP_STATINC(TCP_STAT_RCVACKTOOMUCH); 2665 goto dropafterack; 2666 } 2667 acked = th->th_ack - tp->snd_una; 2668 tcps = TCP_STAT_GETREF(); 2669 tcps[TCP_STAT_RCVACKPACK]++; 2670 tcps[TCP_STAT_RCVACKBYTE] += acked; 2671 TCP_STAT_PUTREF(); 2672 2673 /* 2674 * If we have a timestamp reply, update smoothed 2675 * round trip time. If no timestamp is present but 2676 * transmit timer is running and timed sequence 2677 * number was acked, update smoothed round trip time. 2678 * Since we now have an rtt measurement, cancel the 2679 * timer backoff (cf., Phil Karn's retransmit alg.). 2680 * Recompute the initial retransmit timer. 2681 */ 2682 if (ts_rtt) 2683 tcp_xmit_timer(tp, ts_rtt - 1); 2684 else if (tp->t_rtttime && SEQ_GT(th->th_ack, tp->t_rtseq)) 2685 tcp_xmit_timer(tp, tcp_now - tp->t_rtttime); 2686 2687 /* 2688 * If all outstanding data is acked, stop retransmit 2689 * timer and remember to restart (more output or persist). 2690 * If there is more data to be acked, restart retransmit 2691 * timer, using current (possibly backed-off) value. 2692 */ 2693 if (th->th_ack == tp->snd_max) { 2694 TCP_TIMER_DISARM(tp, TCPT_REXMT); 2695 needoutput = 1; 2696 } else if (TCP_TIMER_ISARMED(tp, TCPT_PERSIST) == 0) 2697 TCP_TIMER_ARM(tp, TCPT_REXMT, tp->t_rxtcur); 2698 2699 /* 2700 * New data has been acked, adjust the congestion window. 2701 */ 2702 tp->t_congctl->newack(tp, th); 2703 2704 nd6_hint(tp); 2705 if (acked > so->so_snd.sb_cc) { 2706 tp->snd_wnd -= so->so_snd.sb_cc; 2707 sbdrop(&so->so_snd, (int)so->so_snd.sb_cc); 2708 ourfinisacked = 1; 2709 } else { 2710 if (acked > (tp->t_lastoff - tp->t_inoff)) 2711 tp->t_lastm = NULL; 2712 sbdrop(&so->so_snd, acked); 2713 tp->t_lastoff -= acked; 2714 if (tp->snd_wnd > acked) 2715 tp->snd_wnd -= acked; 2716 else 2717 tp->snd_wnd = 0; 2718 ourfinisacked = 0; 2719 } 2720 sowwakeup(so); 2721 2722 icmp_check(tp, th, acked); 2723 2724 tp->snd_una = th->th_ack; 2725 if (SEQ_GT(tp->snd_una, tp->snd_fack)) 2726 tp->snd_fack = tp->snd_una; 2727 if (SEQ_LT(tp->snd_nxt, tp->snd_una)) 2728 tp->snd_nxt = tp->snd_una; 2729 if (SEQ_LT(tp->snd_high, tp->snd_una)) 2730 tp->snd_high = tp->snd_una; 2731 2732 switch (tp->t_state) { 2733 2734 /* 2735 * In FIN_WAIT_1 STATE in addition to the processing 2736 * for the ESTABLISHED state if our FIN is now acknowledged 2737 * then enter FIN_WAIT_2. 2738 */ 2739 case TCPS_FIN_WAIT_1: 2740 if (ourfinisacked) { 2741 /* 2742 * If we can't receive any more 2743 * data, then closing user can proceed. 2744 * Starting the timer is contrary to the 2745 * specification, but if we don't get a FIN 2746 * we'll hang forever. 2747 */ 2748 if (so->so_state & SS_CANTRCVMORE) { 2749 soisdisconnected(so); 2750 if (tp->t_maxidle > 0) 2751 TCP_TIMER_ARM(tp, TCPT_2MSL, 2752 tp->t_maxidle); 2753 } 2754 tp->t_state = TCPS_FIN_WAIT_2; 2755 } 2756 break; 2757 2758 /* 2759 * In CLOSING STATE in addition to the processing for 2760 * the ESTABLISHED state if the ACK acknowledges our FIN 2761 * then enter the TIME-WAIT state, otherwise ignore 2762 * the segment. 2763 */ 2764 case TCPS_CLOSING: 2765 if (ourfinisacked) { 2766 tp->t_state = TCPS_TIME_WAIT; 2767 tcp_canceltimers(tp); 2768 TCP_TIMER_ARM(tp, TCPT_2MSL, 2 * tp->t_msl); 2769 soisdisconnected(so); 2770 } 2771 break; 2772 2773 /* 2774 * In LAST_ACK, we may still be waiting for data to drain 2775 * and/or to be acked, as well as for the ack of our FIN. 2776 * If our FIN is now acknowledged, delete the TCB, 2777 * enter the closed state and return. 2778 */ 2779 case TCPS_LAST_ACK: 2780 if (ourfinisacked) { 2781 tp = tcp_close(tp); 2782 goto drop; 2783 } 2784 break; 2785 2786 /* 2787 * In TIME_WAIT state the only thing that should arrive 2788 * is a retransmission of the remote FIN. Acknowledge 2789 * it and restart the finack timer. 2790 */ 2791 case TCPS_TIME_WAIT: 2792 TCP_TIMER_ARM(tp, TCPT_2MSL, 2 * tp->t_msl); 2793 goto dropafterack; 2794 } 2795 } 2796 2797 step6: 2798 /* 2799 * Update window information. 2800 * Don't look at window if no ACK: TAC's send garbage on first SYN. 2801 */ 2802 if ((tiflags & TH_ACK) && (SEQ_LT(tp->snd_wl1, th->th_seq) || 2803 (tp->snd_wl1 == th->th_seq && (SEQ_LT(tp->snd_wl2, th->th_ack) || 2804 (tp->snd_wl2 == th->th_ack && tiwin > tp->snd_wnd))))) { 2805 /* keep track of pure window updates */ 2806 if (tlen == 0 && 2807 tp->snd_wl2 == th->th_ack && tiwin > tp->snd_wnd) 2808 TCP_STATINC(TCP_STAT_RCVWINUPD); 2809 tp->snd_wnd = tiwin; 2810 tp->snd_wl1 = th->th_seq; 2811 tp->snd_wl2 = th->th_ack; 2812 if (tp->snd_wnd > tp->max_sndwnd) 2813 tp->max_sndwnd = tp->snd_wnd; 2814 needoutput = 1; 2815 } 2816 2817 /* 2818 * Process segments with URG. 2819 */ 2820 if ((tiflags & TH_URG) && th->th_urp && 2821 TCPS_HAVERCVDFIN(tp->t_state) == 0) { 2822 /* 2823 * This is a kludge, but if we receive and accept 2824 * random urgent pointers, we'll crash in 2825 * soreceive. It's hard to imagine someone 2826 * actually wanting to send this much urgent data. 2827 */ 2828 if (th->th_urp + so->so_rcv.sb_cc > sb_max) { 2829 th->th_urp = 0; /* XXX */ 2830 tiflags &= ~TH_URG; /* XXX */ 2831 goto dodata; /* XXX */ 2832 } 2833 /* 2834 * If this segment advances the known urgent pointer, 2835 * then mark the data stream. This should not happen 2836 * in CLOSE_WAIT, CLOSING, LAST_ACK or TIME_WAIT STATES since 2837 * a FIN has been received from the remote side. 2838 * In these states we ignore the URG. 2839 * 2840 * According to RFC961 (Assigned Protocols), 2841 * the urgent pointer points to the last octet 2842 * of urgent data. We continue, however, 2843 * to consider it to indicate the first octet 2844 * of data past the urgent section as the original 2845 * spec states (in one of two places). 2846 */ 2847 if (SEQ_GT(th->th_seq+th->th_urp, tp->rcv_up)) { 2848 tp->rcv_up = th->th_seq + th->th_urp; 2849 so->so_oobmark = so->so_rcv.sb_cc + 2850 (tp->rcv_up - tp->rcv_nxt) - 1; 2851 if (so->so_oobmark == 0) 2852 so->so_state |= SS_RCVATMARK; 2853 sohasoutofband(so); 2854 tp->t_oobflags &= ~(TCPOOB_HAVEDATA | TCPOOB_HADDATA); 2855 } 2856 /* 2857 * Remove out of band data so doesn't get presented to user. 2858 * This can happen independent of advancing the URG pointer, 2859 * but if two URG's are pending at once, some out-of-band 2860 * data may creep in... ick. 2861 */ 2862 if (th->th_urp <= (u_int16_t) tlen 2863 #ifdef SO_OOBINLINE 2864 && (so->so_options & SO_OOBINLINE) == 0 2865 #endif 2866 ) 2867 tcp_pulloutofband(so, th, m, hdroptlen); 2868 } else 2869 /* 2870 * If no out of band data is expected, 2871 * pull receive urgent pointer along 2872 * with the receive window. 2873 */ 2874 if (SEQ_GT(tp->rcv_nxt, tp->rcv_up)) 2875 tp->rcv_up = tp->rcv_nxt; 2876 dodata: /* XXX */ 2877 2878 /* 2879 * Process the segment text, merging it into the TCP sequencing queue, 2880 * and arranging for acknowledgement of receipt if necessary. 2881 * This process logically involves adjusting tp->rcv_wnd as data 2882 * is presented to the user (this happens in tcp_usrreq.c, 2883 * tcp_rcvd()). If a FIN has already been received on this 2884 * connection then we just ignore the text. 2885 */ 2886 if ((tlen || (tiflags & TH_FIN)) && 2887 TCPS_HAVERCVDFIN(tp->t_state) == 0) { 2888 /* 2889 * Insert segment ti into reassembly queue of tcp with 2890 * control block tp. Return TH_FIN if reassembly now includes 2891 * a segment with FIN. The macro form does the common case 2892 * inline (segment is the next to be received on an 2893 * established connection, and the queue is empty), 2894 * avoiding linkage into and removal from the queue and 2895 * repetition of various conversions. 2896 * Set DELACK for segments received in order, but ack 2897 * immediately when segments are out of order 2898 * (so fast retransmit can work). 2899 */ 2900 /* NOTE: this was TCP_REASS() macro, but used only once */ 2901 TCP_REASS_LOCK(tp); 2902 if (th->th_seq == tp->rcv_nxt && 2903 TAILQ_FIRST(&tp->segq) == NULL && 2904 tp->t_state == TCPS_ESTABLISHED) { 2905 tcp_setup_ack(tp, th); 2906 tp->rcv_nxt += tlen; 2907 tiflags = th->th_flags & TH_FIN; 2908 tcps = TCP_STAT_GETREF(); 2909 tcps[TCP_STAT_RCVPACK]++; 2910 tcps[TCP_STAT_RCVBYTE] += tlen; 2911 TCP_STAT_PUTREF(); 2912 nd6_hint(tp); 2913 if (so->so_state & SS_CANTRCVMORE) 2914 m_freem(m); 2915 else { 2916 m_adj(m, hdroptlen); 2917 sbappendstream(&(so)->so_rcv, m); 2918 } 2919 TCP_REASS_UNLOCK(tp); 2920 sorwakeup(so); 2921 } else { 2922 m_adj(m, hdroptlen); 2923 tiflags = tcp_reass(tp, th, m, &tlen); 2924 tp->t_flags |= TF_ACKNOW; 2925 } 2926 2927 /* 2928 * Note the amount of data that peer has sent into 2929 * our window, in order to estimate the sender's 2930 * buffer size. 2931 */ 2932 len = so->so_rcv.sb_hiwat - (tp->rcv_adv - tp->rcv_nxt); 2933 } else { 2934 m_freem(m); 2935 m = NULL; 2936 tiflags &= ~TH_FIN; 2937 } 2938 2939 /* 2940 * If FIN is received ACK the FIN and let the user know 2941 * that the connection is closing. Ignore a FIN received before 2942 * the connection is fully established. 2943 */ 2944 if ((tiflags & TH_FIN) && TCPS_HAVEESTABLISHED(tp->t_state)) { 2945 if (TCPS_HAVERCVDFIN(tp->t_state) == 0) { 2946 socantrcvmore(so); 2947 tp->t_flags |= TF_ACKNOW; 2948 tp->rcv_nxt++; 2949 } 2950 switch (tp->t_state) { 2951 2952 /* 2953 * In ESTABLISHED STATE enter the CLOSE_WAIT state. 2954 */ 2955 case TCPS_ESTABLISHED: 2956 tp->t_state = TCPS_CLOSE_WAIT; 2957 break; 2958 2959 /* 2960 * If still in FIN_WAIT_1 STATE FIN has not been acked so 2961 * enter the CLOSING state. 2962 */ 2963 case TCPS_FIN_WAIT_1: 2964 tp->t_state = TCPS_CLOSING; 2965 break; 2966 2967 /* 2968 * In FIN_WAIT_2 state enter the TIME_WAIT state, 2969 * starting the time-wait timer, turning off the other 2970 * standard timers. 2971 */ 2972 case TCPS_FIN_WAIT_2: 2973 tp->t_state = TCPS_TIME_WAIT; 2974 tcp_canceltimers(tp); 2975 TCP_TIMER_ARM(tp, TCPT_2MSL, 2 * tp->t_msl); 2976 soisdisconnected(so); 2977 break; 2978 2979 /* 2980 * In TIME_WAIT state restart the 2 MSL time_wait timer. 2981 */ 2982 case TCPS_TIME_WAIT: 2983 TCP_TIMER_ARM(tp, TCPT_2MSL, 2 * tp->t_msl); 2984 break; 2985 } 2986 } 2987 #ifdef TCP_DEBUG 2988 if (so->so_options & SO_DEBUG) 2989 tcp_trace(TA_INPUT, ostate, tp, tcp_saveti, 0); 2990 #endif 2991 2992 /* 2993 * Return any desired output. 2994 */ 2995 if (needoutput || (tp->t_flags & TF_ACKNOW)) { 2996 KERNEL_LOCK(1, NULL); 2997 (void) tcp_output(tp); 2998 KERNEL_UNLOCK_ONE(NULL); 2999 } 3000 if (tcp_saveti) 3001 m_freem(tcp_saveti); 3002 3003 if (tp->t_state == TCPS_TIME_WAIT 3004 && (so->so_state & SS_NOFDREF) 3005 && (tp->t_inpcb || af != AF_INET) 3006 && (tp->t_in6pcb || af != AF_INET6) 3007 && ((af == AF_INET ? tcp4_vtw_enable : tcp6_vtw_enable) & 1) != 0 3008 && TAILQ_EMPTY(&tp->segq) 3009 && vtw_add(af, tp)) { 3010 ; 3011 } 3012 return; 3013 3014 badsyn: 3015 /* 3016 * Received a bad SYN. Increment counters and dropwithreset. 3017 */ 3018 TCP_STATINC(TCP_STAT_BADSYN); 3019 tp = NULL; 3020 goto dropwithreset; 3021 3022 dropafterack: 3023 /* 3024 * Generate an ACK dropping incoming segment if it occupies 3025 * sequence space, where the ACK reflects our state. 3026 */ 3027 if (tiflags & TH_RST) 3028 goto drop; 3029 goto dropafterack2; 3030 3031 dropafterack_ratelim: 3032 /* 3033 * We may want to rate-limit ACKs against SYN/RST attack. 3034 */ 3035 if (ppsratecheck(&tcp_ackdrop_ppslim_last, &tcp_ackdrop_ppslim_count, 3036 tcp_ackdrop_ppslim) == 0) { 3037 /* XXX stat */ 3038 goto drop; 3039 } 3040 /* ...fall into dropafterack2... */ 3041 3042 dropafterack2: 3043 m_freem(m); 3044 tp->t_flags |= TF_ACKNOW; 3045 KERNEL_LOCK(1, NULL); 3046 (void) tcp_output(tp); 3047 KERNEL_UNLOCK_ONE(NULL); 3048 if (tcp_saveti) 3049 m_freem(tcp_saveti); 3050 return; 3051 3052 dropwithreset_ratelim: 3053 /* 3054 * We may want to rate-limit RSTs in certain situations, 3055 * particularly if we are sending an RST in response to 3056 * an attempt to connect to or otherwise communicate with 3057 * a port for which we have no socket. 3058 */ 3059 if (ppsratecheck(&tcp_rst_ppslim_last, &tcp_rst_ppslim_count, 3060 tcp_rst_ppslim) == 0) { 3061 /* XXX stat */ 3062 goto drop; 3063 } 3064 /* ...fall into dropwithreset... */ 3065 3066 dropwithreset: 3067 /* 3068 * Generate a RST, dropping incoming segment. 3069 * Make ACK acceptable to originator of segment. 3070 */ 3071 if (tiflags & TH_RST) 3072 goto drop; 3073 3074 switch (af) { 3075 #ifdef INET6 3076 case AF_INET6: 3077 /* For following calls to tcp_respond */ 3078 if (IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) 3079 goto drop; 3080 break; 3081 #endif /* INET6 */ 3082 case AF_INET: 3083 if (IN_MULTICAST(ip->ip_dst.s_addr) || 3084 in_broadcast(ip->ip_dst, m->m_pkthdr.rcvif)) 3085 goto drop; 3086 } 3087 3088 if (tiflags & TH_ACK) 3089 (void)tcp_respond(tp, m, m, th, (tcp_seq)0, th->th_ack, TH_RST); 3090 else { 3091 if (tiflags & TH_SYN) 3092 tlen++; 3093 (void)tcp_respond(tp, m, m, th, th->th_seq + tlen, (tcp_seq)0, 3094 TH_RST|TH_ACK); 3095 } 3096 if (tcp_saveti) 3097 m_freem(tcp_saveti); 3098 return; 3099 3100 badcsum: 3101 drop: 3102 /* 3103 * Drop space held by incoming segment and return. 3104 */ 3105 if (tp) { 3106 if (tp->t_inpcb) 3107 so = tp->t_inpcb->inp_socket; 3108 #ifdef INET6 3109 else if (tp->t_in6pcb) 3110 so = tp->t_in6pcb->in6p_socket; 3111 #endif 3112 else 3113 so = NULL; 3114 #ifdef TCP_DEBUG 3115 if (so && (so->so_options & SO_DEBUG) != 0) 3116 tcp_trace(TA_DROP, ostate, tp, tcp_saveti, 0); 3117 #endif 3118 } 3119 if (tcp_saveti) 3120 m_freem(tcp_saveti); 3121 m_freem(m); 3122 return; 3123 } 3124 3125 #ifdef TCP_SIGNATURE 3126 int 3127 tcp_signature_apply(void *fstate, void *data, u_int len) 3128 { 3129 3130 MD5Update(fstate, (u_char *)data, len); 3131 return (0); 3132 } 3133 3134 struct secasvar * 3135 tcp_signature_getsav(struct mbuf *m, struct tcphdr *th) 3136 { 3137 struct ip *ip; 3138 struct ip6_hdr *ip6; 3139 3140 ip = mtod(m, struct ip *); 3141 switch (ip->ip_v) { 3142 case 4: 3143 ip = mtod(m, struct ip *); 3144 ip6 = NULL; 3145 break; 3146 case 6: 3147 ip = NULL; 3148 ip6 = mtod(m, struct ip6_hdr *); 3149 break; 3150 default: 3151 return (NULL); 3152 } 3153 3154 #ifdef IPSEC 3155 if (ipsec_used) { 3156 union sockaddr_union dst; 3157 /* Extract the destination from the IP header in the mbuf. */ 3158 memset(&dst, 0, sizeof(union sockaddr_union)); 3159 if (ip != NULL) { 3160 dst.sa.sa_len = sizeof(struct sockaddr_in); 3161 dst.sa.sa_family = AF_INET; 3162 dst.sin.sin_addr = ip->ip_dst; 3163 } else { 3164 dst.sa.sa_len = sizeof(struct sockaddr_in6); 3165 dst.sa.sa_family = AF_INET6; 3166 dst.sin6.sin6_addr = ip6->ip6_dst; 3167 } 3168 3169 /* 3170 * Look up an SADB entry which matches the address of the peer. 3171 */ 3172 return KEY_ALLOCSA(&dst, IPPROTO_TCP, htonl(TCP_SIG_SPI), 0, 0); 3173 } 3174 return NULL; 3175 #else 3176 if (ip) 3177 return key_allocsa(AF_INET, (void *)&ip->ip_src, 3178 (void *)&ip->ip_dst, IPPROTO_TCP, 3179 htonl(TCP_SIG_SPI), 0, 0); 3180 else 3181 return key_allocsa(AF_INET6, (void *)&ip6->ip6_src, 3182 (void *)&ip6->ip6_dst, IPPROTO_TCP, 3183 htonl(TCP_SIG_SPI), 0, 0); 3184 #endif 3185 } 3186 3187 int 3188 tcp_signature(struct mbuf *m, struct tcphdr *th, int thoff, 3189 struct secasvar *sav, char *sig) 3190 { 3191 MD5_CTX ctx; 3192 struct ip *ip; 3193 struct ipovly *ipovly; 3194 #ifdef INET6 3195 struct ip6_hdr *ip6; 3196 struct ip6_hdr_pseudo ip6pseudo; 3197 #endif /* INET6 */ 3198 struct ippseudo ippseudo; 3199 struct tcphdr th0; 3200 int l, tcphdrlen; 3201 3202 if (sav == NULL) 3203 return (-1); 3204 3205 tcphdrlen = th->th_off * 4; 3206 3207 switch (mtod(m, struct ip *)->ip_v) { 3208 case 4: 3209 MD5Init(&ctx); 3210 ip = mtod(m, struct ip *); 3211 memset(&ippseudo, 0, sizeof(ippseudo)); 3212 ipovly = (struct ipovly *)ip; 3213 ippseudo.ippseudo_src = ipovly->ih_src; 3214 ippseudo.ippseudo_dst = ipovly->ih_dst; 3215 ippseudo.ippseudo_pad = 0; 3216 ippseudo.ippseudo_p = IPPROTO_TCP; 3217 ippseudo.ippseudo_len = htons(m->m_pkthdr.len - thoff); 3218 MD5Update(&ctx, (char *)&ippseudo, sizeof(ippseudo)); 3219 break; 3220 #if INET6 3221 case 6: 3222 MD5Init(&ctx); 3223 ip6 = mtod(m, struct ip6_hdr *); 3224 memset(&ip6pseudo, 0, sizeof(ip6pseudo)); 3225 ip6pseudo.ip6ph_src = ip6->ip6_src; 3226 in6_clearscope(&ip6pseudo.ip6ph_src); 3227 ip6pseudo.ip6ph_dst = ip6->ip6_dst; 3228 in6_clearscope(&ip6pseudo.ip6ph_dst); 3229 ip6pseudo.ip6ph_len = htons(m->m_pkthdr.len - thoff); 3230 ip6pseudo.ip6ph_nxt = IPPROTO_TCP; 3231 MD5Update(&ctx, (char *)&ip6pseudo, sizeof(ip6pseudo)); 3232 break; 3233 #endif /* INET6 */ 3234 default: 3235 return (-1); 3236 } 3237 3238 th0 = *th; 3239 th0.th_sum = 0; 3240 MD5Update(&ctx, (char *)&th0, sizeof(th0)); 3241 3242 l = m->m_pkthdr.len - thoff - tcphdrlen; 3243 if (l > 0) 3244 m_apply(m, thoff + tcphdrlen, 3245 m->m_pkthdr.len - thoff - tcphdrlen, 3246 tcp_signature_apply, &ctx); 3247 3248 MD5Update(&ctx, _KEYBUF(sav->key_auth), _KEYLEN(sav->key_auth)); 3249 MD5Final(sig, &ctx); 3250 3251 return (0); 3252 } 3253 #endif 3254 3255 /* 3256 * tcp_dooptions: parse and process tcp options. 3257 * 3258 * returns -1 if this segment should be dropped. (eg. wrong signature) 3259 * otherwise returns 0. 3260 */ 3261 3262 static int 3263 tcp_dooptions(struct tcpcb *tp, const u_char *cp, int cnt, 3264 struct tcphdr *th, 3265 struct mbuf *m, int toff, struct tcp_opt_info *oi) 3266 { 3267 u_int16_t mss; 3268 int opt, optlen = 0; 3269 #ifdef TCP_SIGNATURE 3270 void *sigp = NULL; 3271 char sigbuf[TCP_SIGLEN]; 3272 struct secasvar *sav = NULL; 3273 #endif 3274 3275 for (; cp && cnt > 0; cnt -= optlen, cp += optlen) { 3276 opt = cp[0]; 3277 if (opt == TCPOPT_EOL) 3278 break; 3279 if (opt == TCPOPT_NOP) 3280 optlen = 1; 3281 else { 3282 if (cnt < 2) 3283 break; 3284 optlen = cp[1]; 3285 if (optlen < 2 || optlen > cnt) 3286 break; 3287 } 3288 switch (opt) { 3289 3290 default: 3291 continue; 3292 3293 case TCPOPT_MAXSEG: 3294 if (optlen != TCPOLEN_MAXSEG) 3295 continue; 3296 if (!(th->th_flags & TH_SYN)) 3297 continue; 3298 if (TCPS_HAVERCVDSYN(tp->t_state)) 3299 continue; 3300 bcopy(cp + 2, &mss, sizeof(mss)); 3301 oi->maxseg = ntohs(mss); 3302 break; 3303 3304 case TCPOPT_WINDOW: 3305 if (optlen != TCPOLEN_WINDOW) 3306 continue; 3307 if (!(th->th_flags & TH_SYN)) 3308 continue; 3309 if (TCPS_HAVERCVDSYN(tp->t_state)) 3310 continue; 3311 tp->t_flags |= TF_RCVD_SCALE; 3312 tp->requested_s_scale = cp[2]; 3313 if (tp->requested_s_scale > TCP_MAX_WINSHIFT) { 3314 char buf[INET6_ADDRSTRLEN]; 3315 struct ip *ip = mtod(m, struct ip *); 3316 #ifdef INET6 3317 struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *); 3318 #endif 3319 if (ip) 3320 in_print(buf, sizeof(buf), 3321 &ip->ip_src); 3322 #ifdef INET6 3323 else if (ip6) 3324 in6_print(buf, sizeof(buf), 3325 &ip6->ip6_src); 3326 #endif 3327 else 3328 strlcpy(buf, "(unknown)", sizeof(buf)); 3329 log(LOG_ERR, "TCP: invalid wscale %d from %s, " 3330 "assuming %d\n", 3331 tp->requested_s_scale, buf, 3332 TCP_MAX_WINSHIFT); 3333 tp->requested_s_scale = TCP_MAX_WINSHIFT; 3334 } 3335 break; 3336 3337 case TCPOPT_TIMESTAMP: 3338 if (optlen != TCPOLEN_TIMESTAMP) 3339 continue; 3340 oi->ts_present = 1; 3341 bcopy(cp + 2, &oi->ts_val, sizeof(oi->ts_val)); 3342 NTOHL(oi->ts_val); 3343 bcopy(cp + 6, &oi->ts_ecr, sizeof(oi->ts_ecr)); 3344 NTOHL(oi->ts_ecr); 3345 3346 if (!(th->th_flags & TH_SYN)) 3347 continue; 3348 if (TCPS_HAVERCVDSYN(tp->t_state)) 3349 continue; 3350 /* 3351 * A timestamp received in a SYN makes 3352 * it ok to send timestamp requests and replies. 3353 */ 3354 tp->t_flags |= TF_RCVD_TSTMP; 3355 tp->ts_recent = oi->ts_val; 3356 tp->ts_recent_age = tcp_now; 3357 break; 3358 3359 case TCPOPT_SACK_PERMITTED: 3360 if (optlen != TCPOLEN_SACK_PERMITTED) 3361 continue; 3362 if (!(th->th_flags & TH_SYN)) 3363 continue; 3364 if (TCPS_HAVERCVDSYN(tp->t_state)) 3365 continue; 3366 if (tcp_do_sack) { 3367 tp->t_flags |= TF_SACK_PERMIT; 3368 tp->t_flags |= TF_WILL_SACK; 3369 } 3370 break; 3371 3372 case TCPOPT_SACK: 3373 tcp_sack_option(tp, th, cp, optlen); 3374 break; 3375 #ifdef TCP_SIGNATURE 3376 case TCPOPT_SIGNATURE: 3377 if (optlen != TCPOLEN_SIGNATURE) 3378 continue; 3379 if (sigp && memcmp(sigp, cp + 2, TCP_SIGLEN)) 3380 return (-1); 3381 3382 sigp = sigbuf; 3383 memcpy(sigbuf, cp + 2, TCP_SIGLEN); 3384 tp->t_flags |= TF_SIGNATURE; 3385 break; 3386 #endif 3387 } 3388 } 3389 3390 #ifndef TCP_SIGNATURE 3391 return 0; 3392 #else 3393 if (tp->t_flags & TF_SIGNATURE) { 3394 3395 sav = tcp_signature_getsav(m, th); 3396 3397 if (sav == NULL && tp->t_state == TCPS_LISTEN) 3398 return (-1); 3399 } 3400 3401 if ((sigp ? TF_SIGNATURE : 0) ^ (tp->t_flags & TF_SIGNATURE)) 3402 goto out; 3403 3404 if (sigp) { 3405 char sig[TCP_SIGLEN]; 3406 3407 tcp_fields_to_net(th); 3408 if (tcp_signature(m, th, toff, sav, sig) < 0) { 3409 tcp_fields_to_host(th); 3410 goto out; 3411 } 3412 tcp_fields_to_host(th); 3413 3414 if (memcmp(sig, sigp, TCP_SIGLEN)) { 3415 TCP_STATINC(TCP_STAT_BADSIG); 3416 goto out; 3417 } else 3418 TCP_STATINC(TCP_STAT_GOODSIG); 3419 3420 key_sa_recordxfer(sav, m); 3421 KEY_FREESAV(&sav); 3422 } 3423 return 0; 3424 out: 3425 if (sav != NULL) 3426 KEY_FREESAV(&sav); 3427 return -1; 3428 #endif 3429 } 3430 3431 /* 3432 * Pull out of band byte out of a segment so 3433 * it doesn't appear in the user's data queue. 3434 * It is still reflected in the segment length for 3435 * sequencing purposes. 3436 */ 3437 void 3438 tcp_pulloutofband(struct socket *so, struct tcphdr *th, 3439 struct mbuf *m, int off) 3440 { 3441 int cnt = off + th->th_urp - 1; 3442 3443 while (cnt >= 0) { 3444 if (m->m_len > cnt) { 3445 char *cp = mtod(m, char *) + cnt; 3446 struct tcpcb *tp = sototcpcb(so); 3447 3448 tp->t_iobc = *cp; 3449 tp->t_oobflags |= TCPOOB_HAVEDATA; 3450 bcopy(cp+1, cp, (unsigned)(m->m_len - cnt - 1)); 3451 m->m_len--; 3452 return; 3453 } 3454 cnt -= m->m_len; 3455 m = m->m_next; 3456 if (m == 0) 3457 break; 3458 } 3459 panic("tcp_pulloutofband"); 3460 } 3461 3462 /* 3463 * Collect new round-trip time estimate 3464 * and update averages and current timeout. 3465 * 3466 * rtt is in units of slow ticks (typically 500 ms) -- essentially the 3467 * difference of two timestamps. 3468 */ 3469 void 3470 tcp_xmit_timer(struct tcpcb *tp, uint32_t rtt) 3471 { 3472 int32_t delta; 3473 3474 TCP_STATINC(TCP_STAT_RTTUPDATED); 3475 if (tp->t_srtt != 0) { 3476 /* 3477 * Compute the amount to add to srtt for smoothing, 3478 * *alpha, or 2^(-TCP_RTT_SHIFT). Because 3479 * srtt is stored in 1/32 slow ticks, we conceptually 3480 * shift left 5 bits, subtract srtt to get the 3481 * diference, and then shift right by TCP_RTT_SHIFT 3482 * (3) to obtain 1/8 of the difference. 3483 */ 3484 delta = (rtt << 2) - (tp->t_srtt >> TCP_RTT_SHIFT); 3485 /* 3486 * This can never happen, because delta's lowest 3487 * possible value is 1/8 of t_srtt. But if it does, 3488 * set srtt to some reasonable value, here chosen 3489 * as 1/8 tick. 3490 */ 3491 if ((tp->t_srtt += delta) <= 0) 3492 tp->t_srtt = 1 << 2; 3493 /* 3494 * RFC2988 requires that rttvar be updated first. 3495 * This code is compliant because "delta" is the old 3496 * srtt minus the new observation (scaled). 3497 * 3498 * RFC2988 says: 3499 * rttvar = (1-beta) * rttvar + beta * |srtt-observed| 3500 * 3501 * delta is in units of 1/32 ticks, and has then been 3502 * divided by 8. This is equivalent to being in 1/16s 3503 * units and divided by 4. Subtract from it 1/4 of 3504 * the existing rttvar to form the (signed) amount to 3505 * adjust. 3506 */ 3507 if (delta < 0) 3508 delta = -delta; 3509 delta -= (tp->t_rttvar >> TCP_RTTVAR_SHIFT); 3510 /* 3511 * As with srtt, this should never happen. There is 3512 * no support in RFC2988 for this operation. But 1/4s 3513 * as rttvar when faced with something arguably wrong 3514 * is ok. 3515 */ 3516 if ((tp->t_rttvar += delta) <= 0) 3517 tp->t_rttvar = 1 << 2; 3518 3519 /* 3520 * If srtt exceeds .01 second, ensure we use the 'remote' MSL 3521 * Problem is: it doesn't work. Disabled by defaulting 3522 * tcp_rttlocal to 0; see corresponding code in 3523 * tcp_subr that selects local vs remote in a different way. 3524 * 3525 * The static branch prediction hint here should be removed 3526 * when the rtt estimator is fixed and the rtt_enable code 3527 * is turned back on. 3528 */ 3529 if (__predict_false(tcp_rttlocal) && tcp_msl_enable 3530 && tp->t_srtt > tcp_msl_remote_threshold 3531 && tp->t_msl < tcp_msl_remote) { 3532 tp->t_msl = tcp_msl_remote; 3533 } 3534 } else { 3535 /* 3536 * This is the first measurement. Per RFC2988, 2.2, 3537 * set rtt=R and srtt=R/2. 3538 * For srtt, storage representation is 1/32 ticks, 3539 * so shift left by 5. 3540 * For rttvar, storage representation is 1/16 ticks, 3541 * So shift left by 4, but then right by 1 to halve. 3542 */ 3543 tp->t_srtt = rtt << (TCP_RTT_SHIFT + 2); 3544 tp->t_rttvar = rtt << (TCP_RTTVAR_SHIFT + 2 - 1); 3545 } 3546 tp->t_rtttime = 0; 3547 tp->t_rxtshift = 0; 3548 3549 /* 3550 * the retransmit should happen at rtt + 4 * rttvar. 3551 * Because of the way we do the smoothing, srtt and rttvar 3552 * will each average +1/2 tick of bias. When we compute 3553 * the retransmit timer, we want 1/2 tick of rounding and 3554 * 1 extra tick because of +-1/2 tick uncertainty in the 3555 * firing of the timer. The bias will give us exactly the 3556 * 1.5 tick we need. But, because the bias is 3557 * statistical, we have to test that we don't drop below 3558 * the minimum feasible timer (which is 2 ticks). 3559 */ 3560 TCPT_RANGESET(tp->t_rxtcur, TCP_REXMTVAL(tp), 3561 max(tp->t_rttmin, rtt + 2), TCPTV_REXMTMAX); 3562 3563 /* 3564 * We received an ack for a packet that wasn't retransmitted; 3565 * it is probably safe to discard any error indications we've 3566 * received recently. This isn't quite right, but close enough 3567 * for now (a route might have failed after we sent a segment, 3568 * and the return path might not be symmetrical). 3569 */ 3570 tp->t_softerror = 0; 3571 } 3572 3573 3574 /* 3575 * TCP compressed state engine. Currently used to hold compressed 3576 * state for SYN_RECEIVED. 3577 */ 3578 3579 u_long syn_cache_count; 3580 u_int32_t syn_hash1, syn_hash2; 3581 3582 #define SYN_HASH(sa, sp, dp) \ 3583 ((((sa)->s_addr^syn_hash1)*(((((u_int32_t)(dp))<<16) + \ 3584 ((u_int32_t)(sp)))^syn_hash2))) 3585 #ifndef INET6 3586 #define SYN_HASHALL(hash, src, dst) \ 3587 do { \ 3588 hash = SYN_HASH(&((const struct sockaddr_in *)(src))->sin_addr, \ 3589 ((const struct sockaddr_in *)(src))->sin_port, \ 3590 ((const struct sockaddr_in *)(dst))->sin_port); \ 3591 } while (/*CONSTCOND*/ 0) 3592 #else 3593 #define SYN_HASH6(sa, sp, dp) \ 3594 ((((sa)->s6_addr32[0] ^ (sa)->s6_addr32[3] ^ syn_hash1) * \ 3595 (((((u_int32_t)(dp))<<16) + ((u_int32_t)(sp)))^syn_hash2)) \ 3596 & 0x7fffffff) 3597 3598 #define SYN_HASHALL(hash, src, dst) \ 3599 do { \ 3600 switch ((src)->sa_family) { \ 3601 case AF_INET: \ 3602 hash = SYN_HASH(&((const struct sockaddr_in *)(src))->sin_addr, \ 3603 ((const struct sockaddr_in *)(src))->sin_port, \ 3604 ((const struct sockaddr_in *)(dst))->sin_port); \ 3605 break; \ 3606 case AF_INET6: \ 3607 hash = SYN_HASH6(&((const struct sockaddr_in6 *)(src))->sin6_addr, \ 3608 ((const struct sockaddr_in6 *)(src))->sin6_port, \ 3609 ((const struct sockaddr_in6 *)(dst))->sin6_port); \ 3610 break; \ 3611 default: \ 3612 hash = 0; \ 3613 } \ 3614 } while (/*CONSTCOND*/0) 3615 #endif /* INET6 */ 3616 3617 static struct pool syn_cache_pool; 3618 3619 /* 3620 * We don't estimate RTT with SYNs, so each packet starts with the default 3621 * RTT and each timer step has a fixed timeout value. 3622 */ 3623 #define SYN_CACHE_TIMER_ARM(sc) \ 3624 do { \ 3625 TCPT_RANGESET((sc)->sc_rxtcur, \ 3626 TCPTV_SRTTDFLT * tcp_backoff[(sc)->sc_rxtshift], TCPTV_MIN, \ 3627 TCPTV_REXMTMAX); \ 3628 callout_reset(&(sc)->sc_timer, \ 3629 (sc)->sc_rxtcur * (hz / PR_SLOWHZ), syn_cache_timer, (sc)); \ 3630 } while (/*CONSTCOND*/0) 3631 3632 #define SYN_CACHE_TIMESTAMP(sc) (tcp_now - (sc)->sc_timebase) 3633 3634 static inline void 3635 syn_cache_rm(struct syn_cache *sc) 3636 { 3637 TAILQ_REMOVE(&tcp_syn_cache[sc->sc_bucketidx].sch_bucket, 3638 sc, sc_bucketq); 3639 sc->sc_tp = NULL; 3640 LIST_REMOVE(sc, sc_tpq); 3641 tcp_syn_cache[sc->sc_bucketidx].sch_length--; 3642 callout_stop(&sc->sc_timer); 3643 syn_cache_count--; 3644 } 3645 3646 static inline void 3647 syn_cache_put(struct syn_cache *sc) 3648 { 3649 if (sc->sc_ipopts) 3650 (void) m_free(sc->sc_ipopts); 3651 rtcache_free(&sc->sc_route); 3652 sc->sc_flags |= SCF_DEAD; 3653 if (!callout_invoking(&sc->sc_timer)) 3654 callout_schedule(&(sc)->sc_timer, 1); 3655 } 3656 3657 void 3658 syn_cache_init(void) 3659 { 3660 int i; 3661 3662 pool_init(&syn_cache_pool, sizeof(struct syn_cache), 0, 0, 0, 3663 "synpl", NULL, IPL_SOFTNET); 3664 3665 /* Initialize the hash buckets. */ 3666 for (i = 0; i < tcp_syn_cache_size; i++) 3667 TAILQ_INIT(&tcp_syn_cache[i].sch_bucket); 3668 } 3669 3670 void 3671 syn_cache_insert(struct syn_cache *sc, struct tcpcb *tp) 3672 { 3673 struct syn_cache_head *scp; 3674 struct syn_cache *sc2; 3675 int s; 3676 3677 /* 3678 * If there are no entries in the hash table, reinitialize 3679 * the hash secrets. 3680 */ 3681 if (syn_cache_count == 0) { 3682 syn_hash1 = cprng_fast32(); 3683 syn_hash2 = cprng_fast32(); 3684 } 3685 3686 SYN_HASHALL(sc->sc_hash, &sc->sc_src.sa, &sc->sc_dst.sa); 3687 sc->sc_bucketidx = sc->sc_hash % tcp_syn_cache_size; 3688 scp = &tcp_syn_cache[sc->sc_bucketidx]; 3689 3690 /* 3691 * Make sure that we don't overflow the per-bucket 3692 * limit or the total cache size limit. 3693 */ 3694 s = splsoftnet(); 3695 if (scp->sch_length >= tcp_syn_bucket_limit) { 3696 TCP_STATINC(TCP_STAT_SC_BUCKETOVERFLOW); 3697 /* 3698 * The bucket is full. Toss the oldest element in the 3699 * bucket. This will be the first entry in the bucket. 3700 */ 3701 sc2 = TAILQ_FIRST(&scp->sch_bucket); 3702 #ifdef DIAGNOSTIC 3703 /* 3704 * This should never happen; we should always find an 3705 * entry in our bucket. 3706 */ 3707 if (sc2 == NULL) 3708 panic("syn_cache_insert: bucketoverflow: impossible"); 3709 #endif 3710 syn_cache_rm(sc2); 3711 syn_cache_put(sc2); /* calls pool_put but see spl above */ 3712 } else if (syn_cache_count >= tcp_syn_cache_limit) { 3713 struct syn_cache_head *scp2, *sce; 3714 3715 TCP_STATINC(TCP_STAT_SC_OVERFLOWED); 3716 /* 3717 * The cache is full. Toss the oldest entry in the 3718 * first non-empty bucket we can find. 3719 * 3720 * XXX We would really like to toss the oldest 3721 * entry in the cache, but we hope that this 3722 * condition doesn't happen very often. 3723 */ 3724 scp2 = scp; 3725 if (TAILQ_EMPTY(&scp2->sch_bucket)) { 3726 sce = &tcp_syn_cache[tcp_syn_cache_size]; 3727 for (++scp2; scp2 != scp; scp2++) { 3728 if (scp2 >= sce) 3729 scp2 = &tcp_syn_cache[0]; 3730 if (! TAILQ_EMPTY(&scp2->sch_bucket)) 3731 break; 3732 } 3733 #ifdef DIAGNOSTIC 3734 /* 3735 * This should never happen; we should always find a 3736 * non-empty bucket. 3737 */ 3738 if (scp2 == scp) 3739 panic("syn_cache_insert: cacheoverflow: " 3740 "impossible"); 3741 #endif 3742 } 3743 sc2 = TAILQ_FIRST(&scp2->sch_bucket); 3744 syn_cache_rm(sc2); 3745 syn_cache_put(sc2); /* calls pool_put but see spl above */ 3746 } 3747 3748 /* 3749 * Initialize the entry's timer. 3750 */ 3751 sc->sc_rxttot = 0; 3752 sc->sc_rxtshift = 0; 3753 SYN_CACHE_TIMER_ARM(sc); 3754 3755 /* Link it from tcpcb entry */ 3756 LIST_INSERT_HEAD(&tp->t_sc, sc, sc_tpq); 3757 3758 /* Put it into the bucket. */ 3759 TAILQ_INSERT_TAIL(&scp->sch_bucket, sc, sc_bucketq); 3760 scp->sch_length++; 3761 syn_cache_count++; 3762 3763 TCP_STATINC(TCP_STAT_SC_ADDED); 3764 splx(s); 3765 } 3766 3767 /* 3768 * Walk the timer queues, looking for SYN,ACKs that need to be retransmitted. 3769 * If we have retransmitted an entry the maximum number of times, expire 3770 * that entry. 3771 */ 3772 void 3773 syn_cache_timer(void *arg) 3774 { 3775 struct syn_cache *sc = arg; 3776 3777 mutex_enter(softnet_lock); 3778 KERNEL_LOCK(1, NULL); 3779 callout_ack(&sc->sc_timer); 3780 3781 if (__predict_false(sc->sc_flags & SCF_DEAD)) { 3782 TCP_STATINC(TCP_STAT_SC_DELAYED_FREE); 3783 callout_destroy(&sc->sc_timer); 3784 pool_put(&syn_cache_pool, sc); 3785 KERNEL_UNLOCK_ONE(NULL); 3786 mutex_exit(softnet_lock); 3787 return; 3788 } 3789 3790 if (__predict_false(sc->sc_rxtshift == TCP_MAXRXTSHIFT)) { 3791 /* Drop it -- too many retransmissions. */ 3792 goto dropit; 3793 } 3794 3795 /* 3796 * Compute the total amount of time this entry has 3797 * been on a queue. If this entry has been on longer 3798 * than the keep alive timer would allow, expire it. 3799 */ 3800 sc->sc_rxttot += sc->sc_rxtcur; 3801 if (sc->sc_rxttot >= tcp_keepinit) 3802 goto dropit; 3803 3804 TCP_STATINC(TCP_STAT_SC_RETRANSMITTED); 3805 (void) syn_cache_respond(sc, NULL); 3806 3807 /* Advance the timer back-off. */ 3808 sc->sc_rxtshift++; 3809 SYN_CACHE_TIMER_ARM(sc); 3810 3811 KERNEL_UNLOCK_ONE(NULL); 3812 mutex_exit(softnet_lock); 3813 return; 3814 3815 dropit: 3816 TCP_STATINC(TCP_STAT_SC_TIMED_OUT); 3817 syn_cache_rm(sc); 3818 if (sc->sc_ipopts) 3819 (void) m_free(sc->sc_ipopts); 3820 rtcache_free(&sc->sc_route); 3821 callout_destroy(&sc->sc_timer); 3822 pool_put(&syn_cache_pool, sc); 3823 KERNEL_UNLOCK_ONE(NULL); 3824 mutex_exit(softnet_lock); 3825 } 3826 3827 /* 3828 * Remove syn cache created by the specified tcb entry, 3829 * because this does not make sense to keep them 3830 * (if there's no tcb entry, syn cache entry will never be used) 3831 */ 3832 void 3833 syn_cache_cleanup(struct tcpcb *tp) 3834 { 3835 struct syn_cache *sc, *nsc; 3836 int s; 3837 3838 s = splsoftnet(); 3839 3840 for (sc = LIST_FIRST(&tp->t_sc); sc != NULL; sc = nsc) { 3841 nsc = LIST_NEXT(sc, sc_tpq); 3842 3843 #ifdef DIAGNOSTIC 3844 if (sc->sc_tp != tp) 3845 panic("invalid sc_tp in syn_cache_cleanup"); 3846 #endif 3847 syn_cache_rm(sc); 3848 syn_cache_put(sc); /* calls pool_put but see spl above */ 3849 } 3850 /* just for safety */ 3851 LIST_INIT(&tp->t_sc); 3852 3853 splx(s); 3854 } 3855 3856 /* 3857 * Find an entry in the syn cache. 3858 */ 3859 struct syn_cache * 3860 syn_cache_lookup(const struct sockaddr *src, const struct sockaddr *dst, 3861 struct syn_cache_head **headp) 3862 { 3863 struct syn_cache *sc; 3864 struct syn_cache_head *scp; 3865 u_int32_t hash; 3866 int s; 3867 3868 SYN_HASHALL(hash, src, dst); 3869 3870 scp = &tcp_syn_cache[hash % tcp_syn_cache_size]; 3871 *headp = scp; 3872 s = splsoftnet(); 3873 for (sc = TAILQ_FIRST(&scp->sch_bucket); sc != NULL; 3874 sc = TAILQ_NEXT(sc, sc_bucketq)) { 3875 if (sc->sc_hash != hash) 3876 continue; 3877 if (!memcmp(&sc->sc_src, src, src->sa_len) && 3878 !memcmp(&sc->sc_dst, dst, dst->sa_len)) { 3879 splx(s); 3880 return (sc); 3881 } 3882 } 3883 splx(s); 3884 return (NULL); 3885 } 3886 3887 /* 3888 * This function gets called when we receive an ACK for a 3889 * socket in the LISTEN state. We look up the connection 3890 * in the syn cache, and if its there, we pull it out of 3891 * the cache and turn it into a full-blown connection in 3892 * the SYN-RECEIVED state. 3893 * 3894 * The return values may not be immediately obvious, and their effects 3895 * can be subtle, so here they are: 3896 * 3897 * NULL SYN was not found in cache; caller should drop the 3898 * packet and send an RST. 3899 * 3900 * -1 We were unable to create the new connection, and are 3901 * aborting it. An ACK,RST is being sent to the peer 3902 * (unless we got screwey sequence numbners; see below), 3903 * because the 3-way handshake has been completed. Caller 3904 * should not free the mbuf, since we may be using it. If 3905 * we are not, we will free it. 3906 * 3907 * Otherwise, the return value is a pointer to the new socket 3908 * associated with the connection. 3909 */ 3910 struct socket * 3911 syn_cache_get(struct sockaddr *src, struct sockaddr *dst, 3912 struct tcphdr *th, unsigned int hlen, unsigned int tlen, 3913 struct socket *so, struct mbuf *m) 3914 { 3915 struct syn_cache *sc; 3916 struct syn_cache_head *scp; 3917 struct inpcb *inp = NULL; 3918 #ifdef INET6 3919 struct in6pcb *in6p = NULL; 3920 #endif 3921 struct tcpcb *tp = 0; 3922 int s; 3923 struct socket *oso; 3924 3925 s = splsoftnet(); 3926 if ((sc = syn_cache_lookup(src, dst, &scp)) == NULL) { 3927 splx(s); 3928 return (NULL); 3929 } 3930 3931 /* 3932 * Verify the sequence and ack numbers. Try getting the correct 3933 * response again. 3934 */ 3935 if ((th->th_ack != sc->sc_iss + 1) || 3936 SEQ_LEQ(th->th_seq, sc->sc_irs) || 3937 SEQ_GT(th->th_seq, sc->sc_irs + 1 + sc->sc_win)) { 3938 (void) syn_cache_respond(sc, m); 3939 splx(s); 3940 return ((struct socket *)(-1)); 3941 } 3942 3943 /* Remove this cache entry */ 3944 syn_cache_rm(sc); 3945 splx(s); 3946 3947 /* 3948 * Ok, create the full blown connection, and set things up 3949 * as they would have been set up if we had created the 3950 * connection when the SYN arrived. If we can't create 3951 * the connection, abort it. 3952 */ 3953 /* 3954 * inp still has the OLD in_pcb stuff, set the 3955 * v6-related flags on the new guy, too. This is 3956 * done particularly for the case where an AF_INET6 3957 * socket is bound only to a port, and a v4 connection 3958 * comes in on that port. 3959 * we also copy the flowinfo from the original pcb 3960 * to the new one. 3961 */ 3962 oso = so; 3963 so = sonewconn(so, true); 3964 if (so == NULL) 3965 goto resetandabort; 3966 3967 switch (so->so_proto->pr_domain->dom_family) { 3968 #ifdef INET 3969 case AF_INET: 3970 inp = sotoinpcb(so); 3971 break; 3972 #endif 3973 #ifdef INET6 3974 case AF_INET6: 3975 in6p = sotoin6pcb(so); 3976 break; 3977 #endif 3978 } 3979 switch (src->sa_family) { 3980 #ifdef INET 3981 case AF_INET: 3982 if (inp) { 3983 inp->inp_laddr = ((struct sockaddr_in *)dst)->sin_addr; 3984 inp->inp_lport = ((struct sockaddr_in *)dst)->sin_port; 3985 inp->inp_options = ip_srcroute(); 3986 in_pcbstate(inp, INP_BOUND); 3987 if (inp->inp_options == NULL) { 3988 inp->inp_options = sc->sc_ipopts; 3989 sc->sc_ipopts = NULL; 3990 } 3991 } 3992 #ifdef INET6 3993 else if (in6p) { 3994 /* IPv4 packet to AF_INET6 socket */ 3995 memset(&in6p->in6p_laddr, 0, sizeof(in6p->in6p_laddr)); 3996 in6p->in6p_laddr.s6_addr16[5] = htons(0xffff); 3997 bcopy(&((struct sockaddr_in *)dst)->sin_addr, 3998 &in6p->in6p_laddr.s6_addr32[3], 3999 sizeof(((struct sockaddr_in *)dst)->sin_addr)); 4000 in6p->in6p_lport = ((struct sockaddr_in *)dst)->sin_port; 4001 in6totcpcb(in6p)->t_family = AF_INET; 4002 if (sotoin6pcb(oso)->in6p_flags & IN6P_IPV6_V6ONLY) 4003 in6p->in6p_flags |= IN6P_IPV6_V6ONLY; 4004 else 4005 in6p->in6p_flags &= ~IN6P_IPV6_V6ONLY; 4006 in6_pcbstate(in6p, IN6P_BOUND); 4007 } 4008 #endif 4009 break; 4010 #endif 4011 #ifdef INET6 4012 case AF_INET6: 4013 if (in6p) { 4014 in6p->in6p_laddr = ((struct sockaddr_in6 *)dst)->sin6_addr; 4015 in6p->in6p_lport = ((struct sockaddr_in6 *)dst)->sin6_port; 4016 in6_pcbstate(in6p, IN6P_BOUND); 4017 } 4018 break; 4019 #endif 4020 } 4021 #ifdef INET6 4022 if (in6p && in6totcpcb(in6p)->t_family == AF_INET6 && sotoinpcb(oso)) { 4023 struct in6pcb *oin6p = sotoin6pcb(oso); 4024 /* inherit socket options from the listening socket */ 4025 in6p->in6p_flags |= (oin6p->in6p_flags & IN6P_CONTROLOPTS); 4026 if (in6p->in6p_flags & IN6P_CONTROLOPTS) { 4027 m_freem(in6p->in6p_options); 4028 in6p->in6p_options = 0; 4029 } 4030 ip6_savecontrol(in6p, &in6p->in6p_options, 4031 mtod(m, struct ip6_hdr *), m); 4032 } 4033 #endif 4034 4035 #if defined(IPSEC) 4036 if (ipsec_used) { 4037 /* 4038 * we make a copy of policy, instead of sharing the policy, for 4039 * better behavior in terms of SA lookup and dead SA removal. 4040 */ 4041 if (inp) { 4042 /* copy old policy into new socket's */ 4043 if (ipsec_copy_pcbpolicy(sotoinpcb(oso)->inp_sp, 4044 inp->inp_sp)) 4045 printf("tcp_input: could not copy policy\n"); 4046 } 4047 #ifdef INET6 4048 else if (in6p) { 4049 /* copy old policy into new socket's */ 4050 if (ipsec_copy_pcbpolicy(sotoin6pcb(oso)->in6p_sp, 4051 in6p->in6p_sp)) 4052 printf("tcp_input: could not copy policy\n"); 4053 } 4054 #endif 4055 } 4056 #endif 4057 4058 /* 4059 * Give the new socket our cached route reference. 4060 */ 4061 if (inp) { 4062 rtcache_copy(&inp->inp_route, &sc->sc_route); 4063 rtcache_free(&sc->sc_route); 4064 } 4065 #ifdef INET6 4066 else { 4067 rtcache_copy(&in6p->in6p_route, &sc->sc_route); 4068 rtcache_free(&sc->sc_route); 4069 } 4070 #endif 4071 4072 if (inp) { 4073 struct sockaddr_in sin; 4074 memcpy(&sin, src, src->sa_len); 4075 if (in_pcbconnect(inp, &sin, &lwp0)) { 4076 goto resetandabort; 4077 } 4078 } 4079 #ifdef INET6 4080 else if (in6p) { 4081 struct sockaddr_in6 sin6; 4082 memcpy(&sin6, src, src->sa_len); 4083 if (src->sa_family == AF_INET) { 4084 /* IPv4 packet to AF_INET6 socket */ 4085 in6_sin_2_v4mapsin6((struct sockaddr_in *)src, &sin6); 4086 } 4087 if (in6_pcbconnect(in6p, &sin6, NULL)) { 4088 goto resetandabort; 4089 } 4090 } 4091 #endif 4092 else { 4093 goto resetandabort; 4094 } 4095 4096 if (inp) 4097 tp = intotcpcb(inp); 4098 #ifdef INET6 4099 else if (in6p) 4100 tp = in6totcpcb(in6p); 4101 #endif 4102 else 4103 tp = NULL; 4104 tp->t_flags = sototcpcb(oso)->t_flags & TF_NODELAY; 4105 if (sc->sc_request_r_scale != 15) { 4106 tp->requested_s_scale = sc->sc_requested_s_scale; 4107 tp->request_r_scale = sc->sc_request_r_scale; 4108 tp->snd_scale = sc->sc_requested_s_scale; 4109 tp->rcv_scale = sc->sc_request_r_scale; 4110 tp->t_flags |= TF_REQ_SCALE|TF_RCVD_SCALE; 4111 } 4112 if (sc->sc_flags & SCF_TIMESTAMP) 4113 tp->t_flags |= TF_REQ_TSTMP|TF_RCVD_TSTMP; 4114 tp->ts_timebase = sc->sc_timebase; 4115 4116 tp->t_template = tcp_template(tp); 4117 if (tp->t_template == 0) { 4118 tp = tcp_drop(tp, ENOBUFS); /* destroys socket */ 4119 so = NULL; 4120 m_freem(m); 4121 goto abort; 4122 } 4123 4124 tp->iss = sc->sc_iss; 4125 tp->irs = sc->sc_irs; 4126 tcp_sendseqinit(tp); 4127 tcp_rcvseqinit(tp); 4128 tp->t_state = TCPS_SYN_RECEIVED; 4129 TCP_TIMER_ARM(tp, TCPT_KEEP, tp->t_keepinit); 4130 TCP_STATINC(TCP_STAT_ACCEPTS); 4131 4132 if ((sc->sc_flags & SCF_SACK_PERMIT) && tcp_do_sack) 4133 tp->t_flags |= TF_WILL_SACK; 4134 4135 if ((sc->sc_flags & SCF_ECN_PERMIT) && tcp_do_ecn) 4136 tp->t_flags |= TF_ECN_PERMIT; 4137 4138 #ifdef TCP_SIGNATURE 4139 if (sc->sc_flags & SCF_SIGNATURE) 4140 tp->t_flags |= TF_SIGNATURE; 4141 #endif 4142 4143 /* Initialize tp->t_ourmss before we deal with the peer's! */ 4144 tp->t_ourmss = sc->sc_ourmaxseg; 4145 tcp_mss_from_peer(tp, sc->sc_peermaxseg); 4146 4147 /* 4148 * Initialize the initial congestion window. If we 4149 * had to retransmit the SYN,ACK, we must initialize cwnd 4150 * to 1 segment (i.e. the Loss Window). 4151 */ 4152 if (sc->sc_rxtshift) 4153 tp->snd_cwnd = tp->t_peermss; 4154 else { 4155 int ss = tcp_init_win; 4156 #ifdef INET 4157 if (inp != NULL && in_localaddr(inp->inp_faddr)) 4158 ss = tcp_init_win_local; 4159 #endif 4160 #ifdef INET6 4161 if (in6p != NULL && in6_localaddr(&in6p->in6p_faddr)) 4162 ss = tcp_init_win_local; 4163 #endif 4164 tp->snd_cwnd = TCP_INITIAL_WINDOW(ss, tp->t_peermss); 4165 } 4166 4167 tcp_rmx_rtt(tp); 4168 tp->snd_wl1 = sc->sc_irs; 4169 tp->rcv_up = sc->sc_irs + 1; 4170 4171 /* 4172 * This is what whould have happened in tcp_output() when 4173 * the SYN,ACK was sent. 4174 */ 4175 tp->snd_up = tp->snd_una; 4176 tp->snd_max = tp->snd_nxt = tp->iss+1; 4177 TCP_TIMER_ARM(tp, TCPT_REXMT, tp->t_rxtcur); 4178 if (sc->sc_win > 0 && SEQ_GT(tp->rcv_nxt + sc->sc_win, tp->rcv_adv)) 4179 tp->rcv_adv = tp->rcv_nxt + sc->sc_win; 4180 tp->last_ack_sent = tp->rcv_nxt; 4181 tp->t_partialacks = -1; 4182 tp->t_dupacks = 0; 4183 4184 TCP_STATINC(TCP_STAT_SC_COMPLETED); 4185 s = splsoftnet(); 4186 syn_cache_put(sc); 4187 splx(s); 4188 return (so); 4189 4190 resetandabort: 4191 (void)tcp_respond(NULL, m, m, th, (tcp_seq)0, th->th_ack, TH_RST); 4192 abort: 4193 if (so != NULL) { 4194 (void) soqremque(so, 1); 4195 (void) soabort(so); 4196 mutex_enter(softnet_lock); 4197 } 4198 s = splsoftnet(); 4199 syn_cache_put(sc); 4200 splx(s); 4201 TCP_STATINC(TCP_STAT_SC_ABORTED); 4202 return ((struct socket *)(-1)); 4203 } 4204 4205 /* 4206 * This function is called when we get a RST for a 4207 * non-existent connection, so that we can see if the 4208 * connection is in the syn cache. If it is, zap it. 4209 */ 4210 4211 void 4212 syn_cache_reset(struct sockaddr *src, struct sockaddr *dst, struct tcphdr *th) 4213 { 4214 struct syn_cache *sc; 4215 struct syn_cache_head *scp; 4216 int s = splsoftnet(); 4217 4218 if ((sc = syn_cache_lookup(src, dst, &scp)) == NULL) { 4219 splx(s); 4220 return; 4221 } 4222 if (SEQ_LT(th->th_seq, sc->sc_irs) || 4223 SEQ_GT(th->th_seq, sc->sc_irs+1)) { 4224 splx(s); 4225 return; 4226 } 4227 syn_cache_rm(sc); 4228 TCP_STATINC(TCP_STAT_SC_RESET); 4229 syn_cache_put(sc); /* calls pool_put but see spl above */ 4230 splx(s); 4231 } 4232 4233 void 4234 syn_cache_unreach(const struct sockaddr *src, const struct sockaddr *dst, 4235 struct tcphdr *th) 4236 { 4237 struct syn_cache *sc; 4238 struct syn_cache_head *scp; 4239 int s; 4240 4241 s = splsoftnet(); 4242 if ((sc = syn_cache_lookup(src, dst, &scp)) == NULL) { 4243 splx(s); 4244 return; 4245 } 4246 /* If the sequence number != sc_iss, then it's a bogus ICMP msg */ 4247 if (ntohl (th->th_seq) != sc->sc_iss) { 4248 splx(s); 4249 return; 4250 } 4251 4252 /* 4253 * If we've retransmitted 3 times and this is our second error, 4254 * we remove the entry. Otherwise, we allow it to continue on. 4255 * This prevents us from incorrectly nuking an entry during a 4256 * spurious network outage. 4257 * 4258 * See tcp_notify(). 4259 */ 4260 if ((sc->sc_flags & SCF_UNREACH) == 0 || sc->sc_rxtshift < 3) { 4261 sc->sc_flags |= SCF_UNREACH; 4262 splx(s); 4263 return; 4264 } 4265 4266 syn_cache_rm(sc); 4267 TCP_STATINC(TCP_STAT_SC_UNREACH); 4268 syn_cache_put(sc); /* calls pool_put but see spl above */ 4269 splx(s); 4270 } 4271 4272 /* 4273 * Given a LISTEN socket and an inbound SYN request, add 4274 * this to the syn cache, and send back a segment: 4275 * <SEQ=ISS><ACK=RCV_NXT><CTL=SYN,ACK> 4276 * to the source. 4277 * 4278 * IMPORTANT NOTE: We do _NOT_ ACK data that might accompany the SYN. 4279 * Doing so would require that we hold onto the data and deliver it 4280 * to the application. However, if we are the target of a SYN-flood 4281 * DoS attack, an attacker could send data which would eventually 4282 * consume all available buffer space if it were ACKed. By not ACKing 4283 * the data, we avoid this DoS scenario. 4284 */ 4285 4286 int 4287 syn_cache_add(struct sockaddr *src, struct sockaddr *dst, struct tcphdr *th, 4288 unsigned int hlen, struct socket *so, struct mbuf *m, u_char *optp, 4289 int optlen, struct tcp_opt_info *oi) 4290 { 4291 struct tcpcb tb, *tp; 4292 long win; 4293 struct syn_cache *sc; 4294 struct syn_cache_head *scp; 4295 struct mbuf *ipopts; 4296 struct tcp_opt_info opti; 4297 int s; 4298 4299 tp = sototcpcb(so); 4300 4301 memset(&opti, 0, sizeof(opti)); 4302 4303 /* 4304 * RFC1122 4.2.3.10, p. 104: discard bcast/mcast SYN 4305 * 4306 * Note this check is performed in tcp_input() very early on. 4307 */ 4308 4309 /* 4310 * Initialize some local state. 4311 */ 4312 win = sbspace(&so->so_rcv); 4313 if (win > TCP_MAXWIN) 4314 win = TCP_MAXWIN; 4315 4316 switch (src->sa_family) { 4317 #ifdef INET 4318 case AF_INET: 4319 /* 4320 * Remember the IP options, if any. 4321 */ 4322 ipopts = ip_srcroute(); 4323 break; 4324 #endif 4325 default: 4326 ipopts = NULL; 4327 } 4328 4329 #ifdef TCP_SIGNATURE 4330 if (optp || (tp->t_flags & TF_SIGNATURE)) 4331 #else 4332 if (optp) 4333 #endif 4334 { 4335 tb.t_flags = tcp_do_rfc1323 ? (TF_REQ_SCALE|TF_REQ_TSTMP) : 0; 4336 #ifdef TCP_SIGNATURE 4337 tb.t_flags |= (tp->t_flags & TF_SIGNATURE); 4338 #endif 4339 tb.t_state = TCPS_LISTEN; 4340 if (tcp_dooptions(&tb, optp, optlen, th, m, m->m_pkthdr.len - 4341 sizeof(struct tcphdr) - optlen - hlen, oi) < 0) 4342 return (0); 4343 } else 4344 tb.t_flags = 0; 4345 4346 /* 4347 * See if we already have an entry for this connection. 4348 * If we do, resend the SYN,ACK. We do not count this 4349 * as a retransmission (XXX though maybe we should). 4350 */ 4351 if ((sc = syn_cache_lookup(src, dst, &scp)) != NULL) { 4352 TCP_STATINC(TCP_STAT_SC_DUPESYN); 4353 if (ipopts) { 4354 /* 4355 * If we were remembering a previous source route, 4356 * forget it and use the new one we've been given. 4357 */ 4358 if (sc->sc_ipopts) 4359 (void) m_free(sc->sc_ipopts); 4360 sc->sc_ipopts = ipopts; 4361 } 4362 sc->sc_timestamp = tb.ts_recent; 4363 if (syn_cache_respond(sc, m) == 0) { 4364 uint64_t *tcps = TCP_STAT_GETREF(); 4365 tcps[TCP_STAT_SNDACKS]++; 4366 tcps[TCP_STAT_SNDTOTAL]++; 4367 TCP_STAT_PUTREF(); 4368 } 4369 return (1); 4370 } 4371 4372 s = splsoftnet(); 4373 sc = pool_get(&syn_cache_pool, PR_NOWAIT); 4374 splx(s); 4375 if (sc == NULL) { 4376 if (ipopts) 4377 (void) m_free(ipopts); 4378 return (0); 4379 } 4380 4381 /* 4382 * Fill in the cache, and put the necessary IP and TCP 4383 * options into the reply. 4384 */ 4385 memset(sc, 0, sizeof(struct syn_cache)); 4386 callout_init(&sc->sc_timer, CALLOUT_MPSAFE); 4387 bcopy(src, &sc->sc_src, src->sa_len); 4388 bcopy(dst, &sc->sc_dst, dst->sa_len); 4389 sc->sc_flags = 0; 4390 sc->sc_ipopts = ipopts; 4391 sc->sc_irs = th->th_seq; 4392 switch (src->sa_family) { 4393 #ifdef INET 4394 case AF_INET: 4395 { 4396 struct sockaddr_in *srcin = (void *) src; 4397 struct sockaddr_in *dstin = (void *) dst; 4398 4399 sc->sc_iss = tcp_new_iss1(&dstin->sin_addr, 4400 &srcin->sin_addr, dstin->sin_port, 4401 srcin->sin_port, sizeof(dstin->sin_addr), 0); 4402 break; 4403 } 4404 #endif /* INET */ 4405 #ifdef INET6 4406 case AF_INET6: 4407 { 4408 struct sockaddr_in6 *srcin6 = (void *) src; 4409 struct sockaddr_in6 *dstin6 = (void *) dst; 4410 4411 sc->sc_iss = tcp_new_iss1(&dstin6->sin6_addr, 4412 &srcin6->sin6_addr, dstin6->sin6_port, 4413 srcin6->sin6_port, sizeof(dstin6->sin6_addr), 0); 4414 break; 4415 } 4416 #endif /* INET6 */ 4417 } 4418 sc->sc_peermaxseg = oi->maxseg; 4419 sc->sc_ourmaxseg = tcp_mss_to_advertise(m->m_flags & M_PKTHDR ? 4420 m->m_pkthdr.rcvif : NULL, 4421 sc->sc_src.sa.sa_family); 4422 sc->sc_win = win; 4423 sc->sc_timebase = tcp_now - 1; /* see tcp_newtcpcb() */ 4424 sc->sc_timestamp = tb.ts_recent; 4425 if ((tb.t_flags & (TF_REQ_TSTMP|TF_RCVD_TSTMP)) == 4426 (TF_REQ_TSTMP|TF_RCVD_TSTMP)) 4427 sc->sc_flags |= SCF_TIMESTAMP; 4428 if ((tb.t_flags & (TF_RCVD_SCALE|TF_REQ_SCALE)) == 4429 (TF_RCVD_SCALE|TF_REQ_SCALE)) { 4430 sc->sc_requested_s_scale = tb.requested_s_scale; 4431 sc->sc_request_r_scale = 0; 4432 /* 4433 * Pick the smallest possible scaling factor that 4434 * will still allow us to scale up to sb_max. 4435 * 4436 * We do this because there are broken firewalls that 4437 * will corrupt the window scale option, leading to 4438 * the other endpoint believing that our advertised 4439 * window is unscaled. At scale factors larger than 4440 * 5 the unscaled window will drop below 1500 bytes, 4441 * leading to serious problems when traversing these 4442 * broken firewalls. 4443 * 4444 * With the default sbmax of 256K, a scale factor 4445 * of 3 will be chosen by this algorithm. Those who 4446 * choose a larger sbmax should watch out 4447 * for the compatiblity problems mentioned above. 4448 * 4449 * RFC1323: The Window field in a SYN (i.e., a <SYN> 4450 * or <SYN,ACK>) segment itself is never scaled. 4451 */ 4452 while (sc->sc_request_r_scale < TCP_MAX_WINSHIFT && 4453 (TCP_MAXWIN << sc->sc_request_r_scale) < sb_max) 4454 sc->sc_request_r_scale++; 4455 } else { 4456 sc->sc_requested_s_scale = 15; 4457 sc->sc_request_r_scale = 15; 4458 } 4459 if ((tb.t_flags & TF_SACK_PERMIT) && tcp_do_sack) 4460 sc->sc_flags |= SCF_SACK_PERMIT; 4461 4462 /* 4463 * ECN setup packet recieved. 4464 */ 4465 if ((th->th_flags & (TH_ECE|TH_CWR)) && tcp_do_ecn) 4466 sc->sc_flags |= SCF_ECN_PERMIT; 4467 4468 #ifdef TCP_SIGNATURE 4469 if (tb.t_flags & TF_SIGNATURE) 4470 sc->sc_flags |= SCF_SIGNATURE; 4471 #endif 4472 sc->sc_tp = tp; 4473 if (syn_cache_respond(sc, m) == 0) { 4474 uint64_t *tcps = TCP_STAT_GETREF(); 4475 tcps[TCP_STAT_SNDACKS]++; 4476 tcps[TCP_STAT_SNDTOTAL]++; 4477 TCP_STAT_PUTREF(); 4478 syn_cache_insert(sc, tp); 4479 } else { 4480 s = splsoftnet(); 4481 /* 4482 * syn_cache_put() will try to schedule the timer, so 4483 * we need to initialize it 4484 */ 4485 SYN_CACHE_TIMER_ARM(sc); 4486 syn_cache_put(sc); 4487 splx(s); 4488 TCP_STATINC(TCP_STAT_SC_DROPPED); 4489 } 4490 return (1); 4491 } 4492 4493 /* 4494 * syn_cache_respond: (re)send SYN+ACK. 4495 * 4496 * returns 0 on success. otherwise returns an errno, typically ENOBUFS. 4497 */ 4498 4499 int 4500 syn_cache_respond(struct syn_cache *sc, struct mbuf *m) 4501 { 4502 #ifdef INET6 4503 struct rtentry *rt; 4504 #endif 4505 struct route *ro; 4506 u_int8_t *optp; 4507 int optlen, error; 4508 u_int16_t tlen; 4509 struct ip *ip = NULL; 4510 #ifdef INET6 4511 struct ip6_hdr *ip6 = NULL; 4512 #endif 4513 struct tcpcb *tp = NULL; 4514 struct tcphdr *th; 4515 u_int hlen; 4516 struct socket *so; 4517 4518 ro = &sc->sc_route; 4519 switch (sc->sc_src.sa.sa_family) { 4520 case AF_INET: 4521 hlen = sizeof(struct ip); 4522 break; 4523 #ifdef INET6 4524 case AF_INET6: 4525 hlen = sizeof(struct ip6_hdr); 4526 break; 4527 #endif 4528 default: 4529 if (m) 4530 m_freem(m); 4531 return (EAFNOSUPPORT); 4532 } 4533 4534 /* Compute the size of the TCP options. */ 4535 optlen = 4 + (sc->sc_request_r_scale != 15 ? 4 : 0) + 4536 ((sc->sc_flags & SCF_SACK_PERMIT) ? (TCPOLEN_SACK_PERMITTED + 2) : 0) + 4537 #ifdef TCP_SIGNATURE 4538 ((sc->sc_flags & SCF_SIGNATURE) ? (TCPOLEN_SIGNATURE + 2) : 0) + 4539 #endif 4540 ((sc->sc_flags & SCF_TIMESTAMP) ? TCPOLEN_TSTAMP_APPA : 0); 4541 4542 tlen = hlen + sizeof(struct tcphdr) + optlen; 4543 4544 /* 4545 * Create the IP+TCP header from scratch. 4546 */ 4547 if (m) 4548 m_freem(m); 4549 #ifdef DIAGNOSTIC 4550 if (max_linkhdr + tlen > MCLBYTES) 4551 return (ENOBUFS); 4552 #endif 4553 MGETHDR(m, M_DONTWAIT, MT_DATA); 4554 if (m && (max_linkhdr + tlen) > MHLEN) { 4555 MCLGET(m, M_DONTWAIT); 4556 if ((m->m_flags & M_EXT) == 0) { 4557 m_freem(m); 4558 m = NULL; 4559 } 4560 } 4561 if (m == NULL) 4562 return (ENOBUFS); 4563 MCLAIM(m, &tcp_tx_mowner); 4564 4565 /* Fixup the mbuf. */ 4566 m->m_data += max_linkhdr; 4567 m->m_len = m->m_pkthdr.len = tlen; 4568 if (sc->sc_tp) { 4569 tp = sc->sc_tp; 4570 if (tp->t_inpcb) 4571 so = tp->t_inpcb->inp_socket; 4572 #ifdef INET6 4573 else if (tp->t_in6pcb) 4574 so = tp->t_in6pcb->in6p_socket; 4575 #endif 4576 else 4577 so = NULL; 4578 } else 4579 so = NULL; 4580 m_reset_rcvif(m); 4581 memset(mtod(m, u_char *), 0, tlen); 4582 4583 switch (sc->sc_src.sa.sa_family) { 4584 case AF_INET: 4585 ip = mtod(m, struct ip *); 4586 ip->ip_v = 4; 4587 ip->ip_dst = sc->sc_src.sin.sin_addr; 4588 ip->ip_src = sc->sc_dst.sin.sin_addr; 4589 ip->ip_p = IPPROTO_TCP; 4590 th = (struct tcphdr *)(ip + 1); 4591 th->th_dport = sc->sc_src.sin.sin_port; 4592 th->th_sport = sc->sc_dst.sin.sin_port; 4593 break; 4594 #ifdef INET6 4595 case AF_INET6: 4596 ip6 = mtod(m, struct ip6_hdr *); 4597 ip6->ip6_vfc = IPV6_VERSION; 4598 ip6->ip6_dst = sc->sc_src.sin6.sin6_addr; 4599 ip6->ip6_src = sc->sc_dst.sin6.sin6_addr; 4600 ip6->ip6_nxt = IPPROTO_TCP; 4601 /* ip6_plen will be updated in ip6_output() */ 4602 th = (struct tcphdr *)(ip6 + 1); 4603 th->th_dport = sc->sc_src.sin6.sin6_port; 4604 th->th_sport = sc->sc_dst.sin6.sin6_port; 4605 break; 4606 #endif 4607 default: 4608 th = NULL; 4609 } 4610 4611 th->th_seq = htonl(sc->sc_iss); 4612 th->th_ack = htonl(sc->sc_irs + 1); 4613 th->th_off = (sizeof(struct tcphdr) + optlen) >> 2; 4614 th->th_flags = TH_SYN|TH_ACK; 4615 th->th_win = htons(sc->sc_win); 4616 /* th_sum already 0 */ 4617 /* th_urp already 0 */ 4618 4619 /* Tack on the TCP options. */ 4620 optp = (u_int8_t *)(th + 1); 4621 *optp++ = TCPOPT_MAXSEG; 4622 *optp++ = 4; 4623 *optp++ = (sc->sc_ourmaxseg >> 8) & 0xff; 4624 *optp++ = sc->sc_ourmaxseg & 0xff; 4625 4626 if (sc->sc_request_r_scale != 15) { 4627 *((u_int32_t *)optp) = htonl(TCPOPT_NOP << 24 | 4628 TCPOPT_WINDOW << 16 | TCPOLEN_WINDOW << 8 | 4629 sc->sc_request_r_scale); 4630 optp += 4; 4631 } 4632 4633 if (sc->sc_flags & SCF_TIMESTAMP) { 4634 u_int32_t *lp = (u_int32_t *)(optp); 4635 /* Form timestamp option as shown in appendix A of RFC 1323. */ 4636 *lp++ = htonl(TCPOPT_TSTAMP_HDR); 4637 *lp++ = htonl(SYN_CACHE_TIMESTAMP(sc)); 4638 *lp = htonl(sc->sc_timestamp); 4639 optp += TCPOLEN_TSTAMP_APPA; 4640 } 4641 4642 if (sc->sc_flags & SCF_SACK_PERMIT) { 4643 u_int8_t *p = optp; 4644 4645 /* Let the peer know that we will SACK. */ 4646 p[0] = TCPOPT_SACK_PERMITTED; 4647 p[1] = 2; 4648 p[2] = TCPOPT_NOP; 4649 p[3] = TCPOPT_NOP; 4650 optp += 4; 4651 } 4652 4653 /* 4654 * Send ECN SYN-ACK setup packet. 4655 * Routes can be asymetric, so, even if we receive a packet 4656 * with ECE and CWR set, we must not assume no one will block 4657 * the ECE packet we are about to send. 4658 */ 4659 if ((sc->sc_flags & SCF_ECN_PERMIT) && tp && 4660 SEQ_GEQ(tp->snd_nxt, tp->snd_max)) { 4661 th->th_flags |= TH_ECE; 4662 TCP_STATINC(TCP_STAT_ECN_SHS); 4663 4664 /* 4665 * draft-ietf-tcpm-ecnsyn-00.txt 4666 * 4667 * "[...] a TCP node MAY respond to an ECN-setup 4668 * SYN packet by setting ECT in the responding 4669 * ECN-setup SYN/ACK packet, indicating to routers 4670 * that the SYN/ACK packet is ECN-Capable. 4671 * This allows a congested router along the path 4672 * to mark the packet instead of dropping the 4673 * packet as an indication of congestion." 4674 * 4675 * "[...] There can be a great benefit in setting 4676 * an ECN-capable codepoint in SYN/ACK packets [...] 4677 * Congestion is most likely to occur in 4678 * the server-to-client direction. As a result, 4679 * setting an ECN-capable codepoint in SYN/ACK 4680 * packets can reduce the occurence of three-second 4681 * retransmit timeouts resulting from the drop 4682 * of SYN/ACK packets." 4683 * 4684 * Page 4 and 6, January 2006. 4685 */ 4686 4687 switch (sc->sc_src.sa.sa_family) { 4688 #ifdef INET 4689 case AF_INET: 4690 ip->ip_tos |= IPTOS_ECN_ECT0; 4691 break; 4692 #endif 4693 #ifdef INET6 4694 case AF_INET6: 4695 ip6->ip6_flow |= htonl(IPTOS_ECN_ECT0 << 20); 4696 break; 4697 #endif 4698 } 4699 TCP_STATINC(TCP_STAT_ECN_ECT); 4700 } 4701 4702 #ifdef TCP_SIGNATURE 4703 if (sc->sc_flags & SCF_SIGNATURE) { 4704 struct secasvar *sav; 4705 u_int8_t *sigp; 4706 4707 sav = tcp_signature_getsav(m, th); 4708 4709 if (sav == NULL) { 4710 if (m) 4711 m_freem(m); 4712 return (EPERM); 4713 } 4714 4715 *optp++ = TCPOPT_SIGNATURE; 4716 *optp++ = TCPOLEN_SIGNATURE; 4717 sigp = optp; 4718 memset(optp, 0, TCP_SIGLEN); 4719 optp += TCP_SIGLEN; 4720 *optp++ = TCPOPT_NOP; 4721 *optp++ = TCPOPT_EOL; 4722 4723 (void)tcp_signature(m, th, hlen, sav, sigp); 4724 4725 key_sa_recordxfer(sav, m); 4726 KEY_FREESAV(&sav); 4727 } 4728 #endif 4729 4730 /* Compute the packet's checksum. */ 4731 switch (sc->sc_src.sa.sa_family) { 4732 case AF_INET: 4733 ip->ip_len = htons(tlen - hlen); 4734 th->th_sum = 0; 4735 th->th_sum = in4_cksum(m, IPPROTO_TCP, hlen, tlen - hlen); 4736 break; 4737 #ifdef INET6 4738 case AF_INET6: 4739 ip6->ip6_plen = htons(tlen - hlen); 4740 th->th_sum = 0; 4741 th->th_sum = in6_cksum(m, IPPROTO_TCP, hlen, tlen - hlen); 4742 break; 4743 #endif 4744 } 4745 4746 /* 4747 * Fill in some straggling IP bits. Note the stack expects 4748 * ip_len to be in host order, for convenience. 4749 */ 4750 switch (sc->sc_src.sa.sa_family) { 4751 #ifdef INET 4752 case AF_INET: 4753 ip->ip_len = htons(tlen); 4754 ip->ip_ttl = ip_defttl; 4755 /* XXX tos? */ 4756 break; 4757 #endif 4758 #ifdef INET6 4759 case AF_INET6: 4760 ip6->ip6_vfc &= ~IPV6_VERSION_MASK; 4761 ip6->ip6_vfc |= IPV6_VERSION; 4762 ip6->ip6_plen = htons(tlen - hlen); 4763 /* ip6_hlim will be initialized afterwards */ 4764 /* XXX flowlabel? */ 4765 break; 4766 #endif 4767 } 4768 4769 /* XXX use IPsec policy on listening socket, on SYN ACK */ 4770 tp = sc->sc_tp; 4771 4772 switch (sc->sc_src.sa.sa_family) { 4773 #ifdef INET 4774 case AF_INET: 4775 error = ip_output(m, sc->sc_ipopts, ro, 4776 (ip_mtudisc ? IP_MTUDISC : 0), 4777 NULL, so); 4778 break; 4779 #endif 4780 #ifdef INET6 4781 case AF_INET6: 4782 ip6->ip6_hlim = in6_selecthlim(NULL, 4783 (rt = rtcache_validate(ro)) != NULL ? rt->rt_ifp : NULL); 4784 4785 error = ip6_output(m, NULL /*XXX*/, ro, 0, NULL, so, NULL); 4786 break; 4787 #endif 4788 default: 4789 error = EAFNOSUPPORT; 4790 break; 4791 } 4792 return (error); 4793 } 4794
Indexes created Wed Sep 24 05:09:52 GMT 2025