1 /* $NetBSD: tcp_subr.c,v 1.112 2001/06/12 15:17:28 wiz 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 (c) 1997, 1998, 2000, 2001 The NetBSD Foundation, Inc. 34 * All rights reserved. 35 * 36 * This code is derived from software contributed to The NetBSD Foundation 37 * by Jason R. Thorpe and Kevin M. Lahey of the Numerical Aerospace Simulation 38 * Facility, NASA Ames Research Center. 39 * 40 * Redistribution and use in source and binary forms, with or without 41 * modification, are permitted provided that the following conditions 42 * are met: 43 * 1. Redistributions of source code must retain the above copyright 44 * notice, this list of conditions and the following disclaimer. 45 * 2. Redistributions in binary form must reproduce the above copyright 46 * notice, this list of conditions and the following disclaimer in the 47 * documentation and/or other materials provided with the distribution. 48 * 3. All advertising materials mentioning features or use of this software 49 * must display the following acknowledgement: 50 * This product includes software developed by the NetBSD 51 * Foundation, Inc. and its contributors. 52 * 4. Neither the name of The NetBSD Foundation nor the names of its 53 * contributors may be used to endorse or promote products derived 54 * from this software without specific prior written permission. 55 * 56 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 57 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 58 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 59 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 60 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 61 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 62 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 63 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 64 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 65 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 66 * POSSIBILITY OF SUCH DAMAGE. 67 */ 68 69 /* 70 * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1995 71 * The Regents of the University of California. All rights reserved. 72 * 73 * Redistribution and use in source and binary forms, with or without 74 * modification, are permitted provided that the following conditions 75 * are met: 76 * 1. Redistributions of source code must retain the above copyright 77 * notice, this list of conditions and the following disclaimer. 78 * 2. Redistributions in binary form must reproduce the above copyright 79 * notice, this list of conditions and the following disclaimer in the 80 * documentation and/or other materials provided with the distribution. 81 * 3. All advertising materials mentioning features or use of this software 82 * must display the following acknowledgement: 83 * This product includes software developed by the University of 84 * California, Berkeley and its contributors. 85 * 4. Neither the name of the University nor the names of its contributors 86 * may be used to endorse or promote products derived from this software 87 * without specific prior written permission. 88 * 89 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 90 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 91 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 92 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 93 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 94 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 95 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 96 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 97 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 98 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 99 * SUCH DAMAGE. 100 * 101 * @(#)tcp_subr.c 8.2 (Berkeley) 5/24/95 102 */ 103 104 #include "opt_inet.h" 105 #include "opt_ipsec.h" 106 #include "opt_tcp_compat_42.h" 107 #include "opt_inet_csum.h" 108 #include "rnd.h" 109 110 #include <sys/param.h> 111 #include <sys/proc.h> 112 #include <sys/systm.h> 113 #include <sys/malloc.h> 114 #include <sys/mbuf.h> 115 #include <sys/socket.h> 116 #include <sys/socketvar.h> 117 #include <sys/protosw.h> 118 #include <sys/errno.h> 119 #include <sys/kernel.h> 120 #include <sys/pool.h> 121 #if NRND > 0 122 #include <sys/md5.h> 123 #include <sys/rnd.h> 124 #endif 125 126 #include <net/route.h> 127 #include <net/if.h> 128 129 #include <netinet/in.h> 130 #include <netinet/in_systm.h> 131 #include <netinet/ip.h> 132 #include <netinet/in_pcb.h> 133 #include <netinet/ip_var.h> 134 #include <netinet/ip_icmp.h> 135 136 #ifdef INET6 137 #ifndef INET 138 #include <netinet/in.h> 139 #endif 140 #include <netinet/ip6.h> 141 #include <netinet6/in6_pcb.h> 142 #include <netinet6/ip6_var.h> 143 #include <netinet6/in6_var.h> 144 #include <netinet6/ip6protosw.h> 145 #include <netinet/icmp6.h> 146 #endif 147 148 #include <netinet/tcp.h> 149 #include <netinet/tcp_fsm.h> 150 #include <netinet/tcp_seq.h> 151 #include <netinet/tcp_timer.h> 152 #include <netinet/tcp_var.h> 153 #include <netinet/tcpip.h> 154 155 #ifdef IPSEC 156 #include <netinet6/ipsec.h> 157 #endif /*IPSEC*/ 158 159 #ifdef INET6 160 struct in6pcb tcb6; 161 #endif 162 163 /* patchable/settable parameters for tcp */ 164 int tcp_mssdflt = TCP_MSS; 165 int tcp_rttdflt = TCPTV_SRTTDFLT / PR_SLOWHZ; 166 int tcp_do_rfc1323 = 1; /* window scaling / timestamps (obsolete) */ 167 #if NRND > 0 168 int tcp_do_rfc1948 = 0; /* ISS by cryptographic hash */ 169 #endif 170 int tcp_do_sack = 1; /* selective acknowledgement */ 171 int tcp_do_win_scale = 1; /* RFC1323 window scaling */ 172 int tcp_do_timestamps = 1; /* RFC1323 timestamps */ 173 int tcp_do_newreno = 0; /* Use the New Reno algorithms */ 174 int tcp_ack_on_push = 0; /* set to enable immediate ACK-on-PUSH */ 175 int tcp_init_win = 1; 176 int tcp_mss_ifmtu = 0; 177 #ifdef TCP_COMPAT_42 178 int tcp_compat_42 = 1; 179 #else 180 int tcp_compat_42 = 0; 181 #endif 182 int tcp_rst_ppslim = 100; /* 100pps */ 183 184 /* tcb hash */ 185 #ifndef TCBHASHSIZE 186 #define TCBHASHSIZE 128 187 #endif 188 int tcbhashsize = TCBHASHSIZE; 189 190 /* syn hash parameters */ 191 #define TCP_SYN_HASH_SIZE 293 192 #define TCP_SYN_BUCKET_SIZE 35 193 int tcp_syn_cache_size = TCP_SYN_HASH_SIZE; 194 int tcp_syn_cache_limit = TCP_SYN_HASH_SIZE*TCP_SYN_BUCKET_SIZE; 195 int tcp_syn_bucket_limit = 3*TCP_SYN_BUCKET_SIZE; 196 struct syn_cache_head tcp_syn_cache[TCP_SYN_HASH_SIZE]; 197 int tcp_syn_cache_interval = 1; /* runs timer twice a second */ 198 199 int tcp_freeq __P((struct tcpcb *)); 200 201 #ifdef INET 202 void tcp_mtudisc_callback __P((struct in_addr)); 203 #endif 204 #ifdef INET6 205 void tcp6_mtudisc_callback __P((struct in6_addr *)); 206 #endif 207 208 void tcp_mtudisc __P((struct inpcb *, int)); 209 #ifdef INET6 210 void tcp6_mtudisc __P((struct in6pcb *, int)); 211 #endif 212 213 struct pool tcpcb_pool; 214 215 #ifdef TCP_CSUM_COUNTERS 216 #include <sys/device.h> 217 218 struct evcnt tcp_hwcsum_bad = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 219 NULL, "tcp", "hwcsum bad"); 220 struct evcnt tcp_hwcsum_ok = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 221 NULL, "tcp", "hwcsum ok"); 222 struct evcnt tcp_hwcsum_data = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 223 NULL, "tcp", "hwcsum data"); 224 struct evcnt tcp_swcsum = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 225 NULL, "tcp", "swcsum"); 226 #endif /* TCP_CSUM_COUNTERS */ 227 228 /* 229 * Tcp initialization 230 */ 231 void 232 tcp_init() 233 { 234 int hlen; 235 236 pool_init(&tcpcb_pool, sizeof(struct tcpcb), 0, 0, 0, "tcpcbpl", 237 0, NULL, NULL, M_PCB); 238 in_pcbinit(&tcbtable, tcbhashsize, tcbhashsize); 239 #ifdef INET6 240 tcb6.in6p_next = tcb6.in6p_prev = &tcb6; 241 #endif 242 LIST_INIT(&tcp_delacks); 243 244 hlen = sizeof(struct ip) + sizeof(struct tcphdr); 245 #ifdef INET6 246 if (sizeof(struct ip) < sizeof(struct ip6_hdr)) 247 hlen = sizeof(struct ip6_hdr) + sizeof(struct tcphdr); 248 #endif 249 if (max_protohdr < hlen) 250 max_protohdr = hlen; 251 if (max_linkhdr + hlen > MHLEN) 252 panic("tcp_init"); 253 254 #ifdef INET 255 icmp_mtudisc_callback_register(tcp_mtudisc_callback); 256 #endif 257 #ifdef INET6 258 icmp6_mtudisc_callback_register(tcp6_mtudisc_callback); 259 #endif 260 261 /* Initialize the compressed state engine. */ 262 syn_cache_init(); 263 264 #ifdef TCP_CSUM_COUNTERS 265 evcnt_attach_static(&tcp_hwcsum_bad); 266 evcnt_attach_static(&tcp_hwcsum_ok); 267 evcnt_attach_static(&tcp_hwcsum_data); 268 evcnt_attach_static(&tcp_swcsum); 269 #endif /* TCP_CSUM_COUNTERS */ 270 } 271 272 /* 273 * Create template to be used to send tcp packets on a connection. 274 * Call after host entry created, allocates an mbuf and fills 275 * in a skeletal tcp/ip header, minimizing the amount of work 276 * necessary when the connection is used. 277 */ 278 struct mbuf * 279 tcp_template(tp) 280 struct tcpcb *tp; 281 { 282 struct inpcb *inp = tp->t_inpcb; 283 #ifdef INET6 284 struct in6pcb *in6p = tp->t_in6pcb; 285 #endif 286 struct tcphdr *n; 287 struct mbuf *m; 288 int hlen; 289 290 switch (tp->t_family) { 291 case AF_INET: 292 hlen = sizeof(struct ip); 293 if (inp) 294 break; 295 #ifdef INET6 296 if (in6p) { 297 /* mapped addr case */ 298 if (IN6_IS_ADDR_V4MAPPED(&in6p->in6p_laddr) 299 && IN6_IS_ADDR_V4MAPPED(&in6p->in6p_faddr)) 300 break; 301 } 302 #endif 303 return NULL; /*EINVAL*/ 304 #ifdef INET6 305 case AF_INET6: 306 hlen = sizeof(struct ip6_hdr); 307 if (in6p) { 308 /* more sainty check? */ 309 break; 310 } 311 return NULL; /*EINVAL*/ 312 #endif 313 default: 314 hlen = 0; /*pacify gcc*/ 315 return NULL; /*EAFNOSUPPORT*/ 316 } 317 #ifdef DIAGNOSTIC 318 if (hlen + sizeof(struct tcphdr) > MCLBYTES) 319 panic("mclbytes too small for t_template"); 320 #endif 321 m = tp->t_template; 322 if (m && m->m_len == hlen + sizeof(struct tcphdr)) 323 ; 324 else { 325 if (m) 326 m_freem(m); 327 m = tp->t_template = NULL; 328 MGETHDR(m, M_DONTWAIT, MT_HEADER); 329 if (m && hlen + sizeof(struct tcphdr) > MHLEN) { 330 MCLGET(m, M_DONTWAIT); 331 if ((m->m_flags & M_EXT) == 0) { 332 m_free(m); 333 m = NULL; 334 } 335 } 336 if (m == NULL) 337 return NULL; 338 m->m_pkthdr.len = m->m_len = hlen + sizeof(struct tcphdr); 339 } 340 341 bzero(mtod(m, caddr_t), m->m_len); 342 343 n = (struct tcphdr *)(mtod(m, caddr_t) + hlen); 344 345 switch (tp->t_family) { 346 case AF_INET: 347 { 348 struct ipovly *ipov; 349 mtod(m, struct ip *)->ip_v = 4; 350 ipov = mtod(m, struct ipovly *); 351 ipov->ih_pr = IPPROTO_TCP; 352 ipov->ih_len = htons(sizeof(struct tcphdr)); 353 if (inp) { 354 ipov->ih_src = inp->inp_laddr; 355 ipov->ih_dst = inp->inp_faddr; 356 } 357 #ifdef INET6 358 else if (in6p) { 359 /* mapped addr case */ 360 bcopy(&in6p->in6p_laddr.s6_addr32[3], &ipov->ih_src, 361 sizeof(ipov->ih_src)); 362 bcopy(&in6p->in6p_faddr.s6_addr32[3], &ipov->ih_dst, 363 sizeof(ipov->ih_dst)); 364 } 365 #endif 366 /* 367 * Compute the pseudo-header portion of the checksum 368 * now. We incrementally add in the TCP option and 369 * payload lengths later, and then compute the TCP 370 * checksum right before the packet is sent off onto 371 * the wire. 372 */ 373 n->th_sum = in_cksum_phdr(ipov->ih_src.s_addr, 374 ipov->ih_dst.s_addr, 375 htons(sizeof(struct tcphdr) + IPPROTO_TCP)); 376 break; 377 } 378 #ifdef INET6 379 case AF_INET6: 380 { 381 struct ip6_hdr *ip6; 382 mtod(m, struct ip *)->ip_v = 6; 383 ip6 = mtod(m, struct ip6_hdr *); 384 ip6->ip6_nxt = IPPROTO_TCP; 385 ip6->ip6_plen = htons(sizeof(struct tcphdr)); 386 ip6->ip6_src = in6p->in6p_laddr; 387 ip6->ip6_dst = in6p->in6p_faddr; 388 ip6->ip6_flow = in6p->in6p_flowinfo & IPV6_FLOWINFO_MASK; 389 if (ip6_auto_flowlabel) { 390 ip6->ip6_flow &= ~IPV6_FLOWLABEL_MASK; 391 ip6->ip6_flow |= 392 (htonl(ip6_flow_seq++) & IPV6_FLOWLABEL_MASK); 393 } 394 ip6->ip6_vfc &= ~IPV6_VERSION_MASK; 395 ip6->ip6_vfc |= IPV6_VERSION; 396 397 /* 398 * Compute the pseudo-header portion of the checksum 399 * now. We incrementally add in the TCP option and 400 * payload lengths later, and then compute the TCP 401 * checksum right before the packet is sent off onto 402 * the wire. 403 */ 404 n->th_sum = in6_cksum_phdr(&in6p->in6p_laddr, 405 &in6p->in6p_faddr, htonl(sizeof(struct tcphdr)), 406 htonl(IPPROTO_TCP)); 407 break; 408 } 409 #endif 410 } 411 if (inp) { 412 n->th_sport = inp->inp_lport; 413 n->th_dport = inp->inp_fport; 414 } 415 #ifdef INET6 416 else if (in6p) { 417 n->th_sport = in6p->in6p_lport; 418 n->th_dport = in6p->in6p_fport; 419 } 420 #endif 421 n->th_seq = 0; 422 n->th_ack = 0; 423 n->th_x2 = 0; 424 n->th_off = 5; 425 n->th_flags = 0; 426 n->th_win = 0; 427 n->th_urp = 0; 428 return (m); 429 } 430 431 /* 432 * Send a single message to the TCP at address specified by 433 * the given TCP/IP header. If m == 0, then we make a copy 434 * of the tcpiphdr at ti and send directly to the addressed host. 435 * This is used to force keep alive messages out using the TCP 436 * template for a connection tp->t_template. If flags are given 437 * then we send a message back to the TCP which originated the 438 * segment ti, and discard the mbuf containing it and any other 439 * attached mbufs. 440 * 441 * In any case the ack and sequence number of the transmitted 442 * segment are as specified by the parameters. 443 */ 444 int 445 tcp_respond(tp, template, m, th0, ack, seq, flags) 446 struct tcpcb *tp; 447 struct mbuf *template; 448 struct mbuf *m; 449 struct tcphdr *th0; 450 tcp_seq ack, seq; 451 int flags; 452 { 453 struct route *ro; 454 int error, tlen, win = 0; 455 int hlen; 456 struct ip *ip; 457 #ifdef INET6 458 struct ip6_hdr *ip6; 459 #endif 460 int family; /* family on packet, not inpcb/in6pcb! */ 461 struct tcphdr *th; 462 463 if (tp != NULL && (flags & TH_RST) == 0) { 464 #ifdef DIAGNOSTIC 465 if (tp->t_inpcb && tp->t_in6pcb) 466 panic("tcp_respond: both t_inpcb and t_in6pcb are set"); 467 #endif 468 #ifdef INET 469 if (tp->t_inpcb) 470 win = sbspace(&tp->t_inpcb->inp_socket->so_rcv); 471 #endif 472 #ifdef INET6 473 if (tp->t_in6pcb) 474 win = sbspace(&tp->t_in6pcb->in6p_socket->so_rcv); 475 #endif 476 } 477 478 ip = NULL; 479 #ifdef INET6 480 ip6 = NULL; 481 #endif 482 if (m == 0) { 483 if (!template) 484 return EINVAL; 485 486 /* get family information from template */ 487 switch (mtod(template, struct ip *)->ip_v) { 488 case 4: 489 family = AF_INET; 490 hlen = sizeof(struct ip); 491 break; 492 #ifdef INET6 493 case 6: 494 family = AF_INET6; 495 hlen = sizeof(struct ip6_hdr); 496 break; 497 #endif 498 default: 499 return EAFNOSUPPORT; 500 } 501 502 MGETHDR(m, M_DONTWAIT, MT_HEADER); 503 if (m) { 504 MCLGET(m, M_DONTWAIT); 505 if ((m->m_flags & M_EXT) == 0) { 506 m_free(m); 507 m = NULL; 508 } 509 } 510 if (m == NULL) 511 return (ENOBUFS); 512 513 if (tcp_compat_42) 514 tlen = 1; 515 else 516 tlen = 0; 517 518 m->m_data += max_linkhdr; 519 bcopy(mtod(template, caddr_t), mtod(m, caddr_t), 520 template->m_len); 521 switch (family) { 522 case AF_INET: 523 ip = mtod(m, struct ip *); 524 th = (struct tcphdr *)(ip + 1); 525 break; 526 #ifdef INET6 527 case AF_INET6: 528 ip6 = mtod(m, struct ip6_hdr *); 529 th = (struct tcphdr *)(ip6 + 1); 530 break; 531 #endif 532 #if 0 533 default: 534 /* noone will visit here */ 535 m_freem(m); 536 return EAFNOSUPPORT; 537 #endif 538 } 539 flags = TH_ACK; 540 } else { 541 542 if ((m->m_flags & M_PKTHDR) == 0) { 543 #if 0 544 printf("non PKTHDR to tcp_respond\n"); 545 #endif 546 m_freem(m); 547 return EINVAL; 548 } 549 #ifdef DIAGNOSTIC 550 if (!th0) 551 panic("th0 == NULL in tcp_respond"); 552 #endif 553 554 /* get family information from m */ 555 switch (mtod(m, struct ip *)->ip_v) { 556 case 4: 557 family = AF_INET; 558 hlen = sizeof(struct ip); 559 ip = mtod(m, struct ip *); 560 break; 561 #ifdef INET6 562 case 6: 563 family = AF_INET6; 564 hlen = sizeof(struct ip6_hdr); 565 ip6 = mtod(m, struct ip6_hdr *); 566 break; 567 #endif 568 default: 569 m_freem(m); 570 return EAFNOSUPPORT; 571 } 572 if ((flags & TH_SYN) == 0 || sizeof(*th0) > (th0->th_off << 2)) 573 tlen = sizeof(*th0); 574 else 575 tlen = th0->th_off << 2; 576 577 if (m->m_len > hlen + tlen && (m->m_flags & M_EXT) == 0 && 578 mtod(m, caddr_t) + hlen == (caddr_t)th0) { 579 m->m_len = hlen + tlen; 580 m_freem(m->m_next); 581 m->m_next = NULL; 582 } else { 583 struct mbuf *n; 584 585 #ifdef DIAGNOSTIC 586 if (max_linkhdr + hlen + tlen > MCLBYTES) { 587 m_freem(m); 588 return EMSGSIZE; 589 } 590 #endif 591 MGETHDR(n, M_DONTWAIT, MT_HEADER); 592 if (n && max_linkhdr + hlen + tlen > MHLEN) { 593 MCLGET(n, M_DONTWAIT); 594 if ((n->m_flags & M_EXT) == 0) { 595 m_freem(n); 596 n = NULL; 597 } 598 } 599 if (!n) { 600 m_freem(m); 601 return ENOBUFS; 602 } 603 604 n->m_data += max_linkhdr; 605 n->m_len = hlen + tlen; 606 m_copyback(n, 0, hlen, mtod(m, caddr_t)); 607 m_copyback(n, hlen, tlen, (caddr_t)th0); 608 609 m_freem(m); 610 m = n; 611 n = NULL; 612 } 613 614 #define xchg(a,b,type) { type t; t=a; a=b; b=t; } 615 switch (family) { 616 case AF_INET: 617 ip = mtod(m, struct ip *); 618 th = (struct tcphdr *)(ip + 1); 619 ip->ip_p = IPPROTO_TCP; 620 xchg(ip->ip_dst, ip->ip_src, struct in_addr); 621 ip->ip_p = IPPROTO_TCP; 622 break; 623 #ifdef INET6 624 case AF_INET6: 625 ip6 = mtod(m, struct ip6_hdr *); 626 th = (struct tcphdr *)(ip6 + 1); 627 ip6->ip6_nxt = IPPROTO_TCP; 628 xchg(ip6->ip6_dst, ip6->ip6_src, struct in6_addr); 629 ip6->ip6_nxt = IPPROTO_TCP; 630 break; 631 #endif 632 #if 0 633 default: 634 /* noone will visit here */ 635 m_freem(m); 636 return EAFNOSUPPORT; 637 #endif 638 } 639 xchg(th->th_dport, th->th_sport, u_int16_t); 640 #undef xchg 641 tlen = 0; /*be friendly with the following code*/ 642 } 643 th->th_seq = htonl(seq); 644 th->th_ack = htonl(ack); 645 th->th_x2 = 0; 646 if ((flags & TH_SYN) == 0) { 647 if (tp) 648 win >>= tp->rcv_scale; 649 if (win > TCP_MAXWIN) 650 win = TCP_MAXWIN; 651 th->th_win = htons((u_int16_t)win); 652 th->th_off = sizeof (struct tcphdr) >> 2; 653 tlen += sizeof(*th); 654 } else 655 tlen += th->th_off << 2; 656 m->m_len = hlen + tlen; 657 m->m_pkthdr.len = hlen + tlen; 658 m->m_pkthdr.rcvif = (struct ifnet *) 0; 659 th->th_flags = flags; 660 th->th_urp = 0; 661 662 switch (family) { 663 #ifdef INET 664 case AF_INET: 665 { 666 struct ipovly *ipov = (struct ipovly *)ip; 667 bzero(ipov->ih_x1, sizeof ipov->ih_x1); 668 ipov->ih_len = htons((u_int16_t)tlen); 669 670 th->th_sum = 0; 671 th->th_sum = in_cksum(m, hlen + tlen); 672 ip->ip_len = hlen + tlen; /*will be flipped on output*/ 673 ip->ip_ttl = ip_defttl; 674 break; 675 } 676 #endif 677 #ifdef INET6 678 case AF_INET6: 679 { 680 th->th_sum = 0; 681 th->th_sum = in6_cksum(m, IPPROTO_TCP, sizeof(struct ip6_hdr), 682 tlen); 683 ip6->ip6_plen = ntohs(tlen); 684 if (tp && tp->t_in6pcb) { 685 struct ifnet *oifp; 686 ro = (struct route *)&tp->t_in6pcb->in6p_route; 687 oifp = ro->ro_rt ? ro->ro_rt->rt_ifp : NULL; 688 ip6->ip6_hlim = in6_selecthlim(tp->t_in6pcb, oifp); 689 } else 690 ip6->ip6_hlim = ip6_defhlim; 691 ip6->ip6_flow &= ~IPV6_FLOWINFO_MASK; 692 if (ip6_auto_flowlabel) { 693 ip6->ip6_flow |= 694 (htonl(ip6_flow_seq++) & IPV6_FLOWLABEL_MASK); 695 } 696 break; 697 } 698 #endif 699 } 700 701 #ifdef IPSEC 702 (void)ipsec_setsocket(m, NULL); 703 #endif /*IPSEC*/ 704 705 if (tp != NULL && tp->t_inpcb != NULL) { 706 ro = &tp->t_inpcb->inp_route; 707 #ifdef IPSEC 708 if (ipsec_setsocket(m, tp->t_inpcb->inp_socket) != 0) { 709 m_freem(m); 710 return ENOBUFS; 711 } 712 #endif 713 #ifdef DIAGNOSTIC 714 if (family != AF_INET) 715 panic("tcp_respond: address family mismatch"); 716 if (!in_hosteq(ip->ip_dst, tp->t_inpcb->inp_faddr)) { 717 panic("tcp_respond: ip_dst %x != inp_faddr %x", 718 ntohl(ip->ip_dst.s_addr), 719 ntohl(tp->t_inpcb->inp_faddr.s_addr)); 720 } 721 #endif 722 } 723 #ifdef INET6 724 else if (tp != NULL && tp->t_in6pcb != NULL) { 725 ro = (struct route *)&tp->t_in6pcb->in6p_route; 726 #ifdef IPSEC 727 if (ipsec_setsocket(m, tp->t_in6pcb->in6p_socket) != 0) { 728 m_freem(m); 729 return ENOBUFS; 730 } 731 #endif 732 #ifdef DIAGNOSTIC 733 if (family == AF_INET) { 734 if (!IN6_IS_ADDR_V4MAPPED(&tp->t_in6pcb->in6p_faddr)) 735 panic("tcp_respond: not mapped addr"); 736 if (bcmp(&ip->ip_dst, 737 &tp->t_in6pcb->in6p_faddr.s6_addr32[3], 738 sizeof(ip->ip_dst)) != 0) { 739 panic("tcp_respond: ip_dst != in6p_faddr"); 740 } 741 } else if (family == AF_INET6) { 742 if (!IN6_ARE_ADDR_EQUAL(&ip6->ip6_dst, &tp->t_in6pcb->in6p_faddr)) 743 panic("tcp_respond: ip6_dst != in6p_faddr"); 744 } else 745 panic("tcp_respond: address family mismatch"); 746 #endif 747 } 748 #endif 749 else 750 ro = NULL; 751 752 switch (family) { 753 #ifdef INET 754 case AF_INET: 755 error = ip_output(m, NULL, ro, 756 (ip_mtudisc ? IP_MTUDISC : 0), 757 NULL); 758 break; 759 #endif 760 #ifdef INET6 761 case AF_INET6: 762 error = ip6_output(m, NULL, (struct route_in6 *)ro, 0, NULL, 763 NULL); 764 break; 765 #endif 766 default: 767 error = EAFNOSUPPORT; 768 break; 769 } 770 771 return (error); 772 } 773 774 /* 775 * Create a new TCP control block, making an 776 * empty reassembly queue and hooking it to the argument 777 * protocol control block. 778 */ 779 struct tcpcb * 780 tcp_newtcpcb(family, aux) 781 int family; /* selects inpcb, or in6pcb */ 782 void *aux; 783 { 784 struct tcpcb *tp; 785 786 switch (family) { 787 case PF_INET: 788 break; 789 #ifdef INET6 790 case PF_INET6: 791 break; 792 #endif 793 default: 794 return NULL; 795 } 796 797 tp = pool_get(&tcpcb_pool, PR_NOWAIT); 798 if (tp == NULL) 799 return (NULL); 800 bzero((caddr_t)tp, sizeof(struct tcpcb)); 801 LIST_INIT(&tp->segq); 802 LIST_INIT(&tp->timeq); 803 tp->t_family = family; /* may be overridden later on */ 804 tp->t_peermss = tcp_mssdflt; 805 tp->t_ourmss = tcp_mssdflt; 806 tp->t_segsz = tcp_mssdflt; 807 LIST_INIT(&tp->t_sc); 808 809 tp->t_flags = 0; 810 if (tcp_do_rfc1323 && tcp_do_win_scale) 811 tp->t_flags |= TF_REQ_SCALE; 812 if (tcp_do_rfc1323 && tcp_do_timestamps) 813 tp->t_flags |= TF_REQ_TSTMP; 814 if (tcp_do_sack == 2) 815 tp->t_flags |= TF_WILL_SACK; 816 else if (tcp_do_sack == 1) 817 tp->t_flags |= TF_WILL_SACK|TF_IGNR_RXSACK; 818 tp->t_flags |= TF_CANT_TXSACK; 819 switch (family) { 820 case PF_INET: 821 tp->t_inpcb = (struct inpcb *)aux; 822 break; 823 #ifdef INET6 824 case PF_INET6: 825 tp->t_in6pcb = (struct in6pcb *)aux; 826 break; 827 #endif 828 } 829 /* 830 * Init srtt to TCPTV_SRTTBASE (0), so we can tell that we have no 831 * rtt estimate. Set rttvar so that srtt + 2 * rttvar gives 832 * reasonable initial retransmit time. 833 */ 834 tp->t_srtt = TCPTV_SRTTBASE; 835 tp->t_rttvar = tcp_rttdflt * PR_SLOWHZ << (TCP_RTTVAR_SHIFT + 2 - 1); 836 tp->t_rttmin = TCPTV_MIN; 837 TCPT_RANGESET(tp->t_rxtcur, TCP_REXMTVAL(tp), 838 TCPTV_MIN, TCPTV_REXMTMAX); 839 tp->snd_cwnd = TCP_MAXWIN << TCP_MAX_WINSHIFT; 840 tp->snd_ssthresh = TCP_MAXWIN << TCP_MAX_WINSHIFT; 841 if (family == AF_INET) { 842 struct inpcb *inp = (struct inpcb *)aux; 843 inp->inp_ip.ip_ttl = ip_defttl; 844 inp->inp_ppcb = (caddr_t)tp; 845 } 846 #ifdef INET6 847 else if (family == AF_INET6) { 848 struct in6pcb *in6p = (struct in6pcb *)aux; 849 in6p->in6p_ip6.ip6_hlim = in6_selecthlim(in6p, 850 in6p->in6p_route.ro_rt ? in6p->in6p_route.ro_rt->rt_ifp 851 : NULL); 852 in6p->in6p_ppcb = (caddr_t)tp; 853 } 854 #endif 855 856 /* 857 * Initialize our timebase. When we send timestamps, we take 858 * the delta from tcp_now -- this means each connection always 859 * gets a timebase of 0, which makes it, among other things, 860 * more difficult to determine how long a system has been up, 861 * and thus how many TCP sequence increments have occurred. 862 */ 863 tp->ts_timebase = tcp_now; 864 865 return (tp); 866 } 867 868 /* 869 * Drop a TCP connection, reporting 870 * the specified error. If connection is synchronized, 871 * then send a RST to peer. 872 */ 873 struct tcpcb * 874 tcp_drop(tp, errno) 875 struct tcpcb *tp; 876 int errno; 877 { 878 struct socket *so = NULL; 879 880 #ifdef DIAGNOSTIC 881 if (tp->t_inpcb && tp->t_in6pcb) 882 panic("tcp_drop: both t_inpcb and t_in6pcb are set"); 883 #endif 884 #ifdef INET 885 if (tp->t_inpcb) 886 so = tp->t_inpcb->inp_socket; 887 #endif 888 #ifdef INET6 889 if (tp->t_in6pcb) 890 so = tp->t_in6pcb->in6p_socket; 891 #endif 892 if (!so) 893 return NULL; 894 895 if (TCPS_HAVERCVDSYN(tp->t_state)) { 896 tp->t_state = TCPS_CLOSED; 897 (void) tcp_output(tp); 898 tcpstat.tcps_drops++; 899 } else 900 tcpstat.tcps_conndrops++; 901 if (errno == ETIMEDOUT && tp->t_softerror) 902 errno = tp->t_softerror; 903 so->so_error = errno; 904 return (tcp_close(tp)); 905 } 906 907 /* 908 * Close a TCP control block: 909 * discard all space held by the tcp 910 * discard internet protocol block 911 * wake up any sleepers 912 */ 913 struct tcpcb * 914 tcp_close(tp) 915 struct tcpcb *tp; 916 { 917 struct inpcb *inp; 918 #ifdef INET6 919 struct in6pcb *in6p; 920 #endif 921 struct socket *so; 922 #ifdef RTV_RTT 923 struct rtentry *rt; 924 #endif 925 struct route *ro; 926 927 inp = tp->t_inpcb; 928 #ifdef INET6 929 in6p = tp->t_in6pcb; 930 #endif 931 so = NULL; 932 ro = NULL; 933 if (inp) { 934 so = inp->inp_socket; 935 ro = &inp->inp_route; 936 } 937 #ifdef INET6 938 else if (in6p) { 939 so = in6p->in6p_socket; 940 ro = (struct route *)&in6p->in6p_route; 941 } 942 #endif 943 944 #ifdef RTV_RTT 945 /* 946 * If we sent enough data to get some meaningful characteristics, 947 * save them in the routing entry. 'Enough' is arbitrarily 948 * defined as the sendpipesize (default 4K) * 16. This would 949 * give us 16 rtt samples assuming we only get one sample per 950 * window (the usual case on a long haul net). 16 samples is 951 * enough for the srtt filter to converge to within 5% of the correct 952 * value; fewer samples and we could save a very bogus rtt. 953 * 954 * Don't update the default route's characteristics and don't 955 * update anything that the user "locked". 956 */ 957 if (SEQ_LT(tp->iss + so->so_snd.sb_hiwat * 16, tp->snd_max) && 958 ro && (rt = ro->ro_rt) && 959 !in_nullhost(satosin(rt_key(rt))->sin_addr)) { 960 u_long i = 0; 961 962 if ((rt->rt_rmx.rmx_locks & RTV_RTT) == 0) { 963 i = tp->t_srtt * 964 ((RTM_RTTUNIT / PR_SLOWHZ) >> (TCP_RTT_SHIFT + 2)); 965 if (rt->rt_rmx.rmx_rtt && i) 966 /* 967 * filter this update to half the old & half 968 * the new values, converting scale. 969 * See route.h and tcp_var.h for a 970 * description of the scaling constants. 971 */ 972 rt->rt_rmx.rmx_rtt = 973 (rt->rt_rmx.rmx_rtt + i) / 2; 974 else 975 rt->rt_rmx.rmx_rtt = i; 976 } 977 if ((rt->rt_rmx.rmx_locks & RTV_RTTVAR) == 0) { 978 i = tp->t_rttvar * 979 ((RTM_RTTUNIT / PR_SLOWHZ) >> (TCP_RTTVAR_SHIFT + 2)); 980 if (rt->rt_rmx.rmx_rttvar && i) 981 rt->rt_rmx.rmx_rttvar = 982 (rt->rt_rmx.rmx_rttvar + i) / 2; 983 else 984 rt->rt_rmx.rmx_rttvar = i; 985 } 986 /* 987 * update the pipelimit (ssthresh) if it has been updated 988 * already or if a pipesize was specified & the threshhold 989 * got below half the pipesize. I.e., wait for bad news 990 * before we start updating, then update on both good 991 * and bad news. 992 */ 993 if (((rt->rt_rmx.rmx_locks & RTV_SSTHRESH) == 0 && 994 (i = tp->snd_ssthresh) && rt->rt_rmx.rmx_ssthresh) || 995 i < (rt->rt_rmx.rmx_sendpipe / 2)) { 996 /* 997 * convert the limit from user data bytes to 998 * packets then to packet data bytes. 999 */ 1000 i = (i + tp->t_segsz / 2) / tp->t_segsz; 1001 if (i < 2) 1002 i = 2; 1003 i *= (u_long)(tp->t_segsz + sizeof (struct tcpiphdr)); 1004 if (rt->rt_rmx.rmx_ssthresh) 1005 rt->rt_rmx.rmx_ssthresh = 1006 (rt->rt_rmx.rmx_ssthresh + i) / 2; 1007 else 1008 rt->rt_rmx.rmx_ssthresh = i; 1009 } 1010 } 1011 #endif /* RTV_RTT */ 1012 /* free the reassembly queue, if any */ 1013 TCP_REASS_LOCK(tp); 1014 (void) tcp_freeq(tp); 1015 TCP_REASS_UNLOCK(tp); 1016 1017 TCP_CLEAR_DELACK(tp); 1018 syn_cache_cleanup(tp); 1019 1020 if (tp->t_template) { 1021 m_free(tp->t_template); 1022 tp->t_template = NULL; 1023 } 1024 pool_put(&tcpcb_pool, tp); 1025 if (inp) { 1026 inp->inp_ppcb = 0; 1027 soisdisconnected(so); 1028 in_pcbdetach(inp); 1029 } 1030 #ifdef INET6 1031 else if (in6p) { 1032 in6p->in6p_ppcb = 0; 1033 soisdisconnected(so); 1034 in6_pcbdetach(in6p); 1035 } 1036 #endif 1037 tcpstat.tcps_closed++; 1038 return ((struct tcpcb *)0); 1039 } 1040 1041 int 1042 tcp_freeq(tp) 1043 struct tcpcb *tp; 1044 { 1045 struct ipqent *qe; 1046 int rv = 0; 1047 #ifdef TCPREASS_DEBUG 1048 int i = 0; 1049 #endif 1050 1051 TCP_REASS_LOCK_CHECK(tp); 1052 1053 while ((qe = tp->segq.lh_first) != NULL) { 1054 #ifdef TCPREASS_DEBUG 1055 printf("tcp_freeq[%p,%d]: %u:%u(%u) 0x%02x\n", 1056 tp, i++, qe->ipqe_seq, qe->ipqe_seq + qe->ipqe_len, 1057 qe->ipqe_len, qe->ipqe_flags & (TH_SYN|TH_FIN|TH_RST)); 1058 #endif 1059 LIST_REMOVE(qe, ipqe_q); 1060 LIST_REMOVE(qe, ipqe_timeq); 1061 m_freem(qe->ipqe_m); 1062 pool_put(&ipqent_pool, qe); 1063 rv = 1; 1064 } 1065 return (rv); 1066 } 1067 1068 /* 1069 * Protocol drain routine. Called when memory is in short supply. 1070 */ 1071 void 1072 tcp_drain() 1073 { 1074 struct inpcb *inp; 1075 struct tcpcb *tp; 1076 1077 /* 1078 * Free the sequence queue of all TCP connections. 1079 */ 1080 inp = tcbtable.inpt_queue.cqh_first; 1081 if (inp) /* XXX */ 1082 for (; inp != (struct inpcb *)&tcbtable.inpt_queue; 1083 inp = inp->inp_queue.cqe_next) { 1084 if ((tp = intotcpcb(inp)) != NULL) { 1085 /* 1086 * We may be called from a device's interrupt 1087 * context. If the tcpcb is already busy, 1088 * just bail out now. 1089 */ 1090 if (tcp_reass_lock_try(tp) == 0) 1091 continue; 1092 if (tcp_freeq(tp)) 1093 tcpstat.tcps_connsdrained++; 1094 TCP_REASS_UNLOCK(tp); 1095 } 1096 } 1097 } 1098 1099 /* 1100 * Notify a tcp user of an asynchronous error; 1101 * store error as soft error, but wake up user 1102 * (for now, won't do anything until can select for soft error). 1103 */ 1104 void 1105 tcp_notify(inp, error) 1106 struct inpcb *inp; 1107 int error; 1108 { 1109 struct tcpcb *tp = (struct tcpcb *)inp->inp_ppcb; 1110 struct socket *so = inp->inp_socket; 1111 1112 /* 1113 * Ignore some errors if we are hooked up. 1114 * If connection hasn't completed, has retransmitted several times, 1115 * and receives a second error, give up now. This is better 1116 * than waiting a long time to establish a connection that 1117 * can never complete. 1118 */ 1119 if (tp->t_state == TCPS_ESTABLISHED && 1120 (error == EHOSTUNREACH || error == ENETUNREACH || 1121 error == EHOSTDOWN)) { 1122 return; 1123 } else if (TCPS_HAVEESTABLISHED(tp->t_state) == 0 && 1124 tp->t_rxtshift > 3 && tp->t_softerror) 1125 so->so_error = error; 1126 else 1127 tp->t_softerror = error; 1128 wakeup((caddr_t) &so->so_timeo); 1129 sorwakeup(so); 1130 sowwakeup(so); 1131 } 1132 1133 #ifdef INET6 1134 void 1135 tcp6_notify(in6p, error) 1136 struct in6pcb *in6p; 1137 int error; 1138 { 1139 struct tcpcb *tp = (struct tcpcb *)in6p->in6p_ppcb; 1140 struct socket *so = in6p->in6p_socket; 1141 1142 /* 1143 * Ignore some errors if we are hooked up. 1144 * If connection hasn't completed, has retransmitted several times, 1145 * and receives a second error, give up now. This is better 1146 * than waiting a long time to establish a connection that 1147 * can never complete. 1148 */ 1149 if (tp->t_state == TCPS_ESTABLISHED && 1150 (error == EHOSTUNREACH || error == ENETUNREACH || 1151 error == EHOSTDOWN)) { 1152 return; 1153 } else if (TCPS_HAVEESTABLISHED(tp->t_state) == 0 && 1154 tp->t_rxtshift > 3 && tp->t_softerror) 1155 so->so_error = error; 1156 else 1157 tp->t_softerror = error; 1158 wakeup((caddr_t) &so->so_timeo); 1159 sorwakeup(so); 1160 sowwakeup(so); 1161 } 1162 #endif 1163 1164 #ifdef INET6 1165 void 1166 tcp6_ctlinput(cmd, sa, d) 1167 int cmd; 1168 struct sockaddr *sa; 1169 void *d; 1170 { 1171 struct tcphdr th; 1172 void (*notify) __P((struct in6pcb *, int)) = tcp6_notify; 1173 int nmatch; 1174 struct ip6_hdr *ip6; 1175 const struct sockaddr_in6 *sa6_src = NULL; 1176 struct sockaddr_in6 *sa6 = (struct sockaddr_in6 *)sa; 1177 struct mbuf *m; 1178 int off; 1179 1180 if (sa->sa_family != AF_INET6 || 1181 sa->sa_len != sizeof(struct sockaddr_in6)) 1182 return; 1183 if ((unsigned)cmd >= PRC_NCMDS) 1184 return; 1185 else if (cmd == PRC_QUENCH) { 1186 /* XXX there's no PRC_QUENCH in IPv6 */ 1187 notify = tcp6_quench; 1188 } else if (PRC_IS_REDIRECT(cmd)) 1189 notify = in6_rtchange, d = NULL; 1190 else if (cmd == PRC_MSGSIZE) 1191 ; /* special code is present, see below */ 1192 else if (cmd == PRC_HOSTDEAD) 1193 d = NULL; 1194 else if (inet6ctlerrmap[cmd] == 0) 1195 return; 1196 1197 /* if the parameter is from icmp6, decode it. */ 1198 if (d != NULL) { 1199 struct ip6ctlparam *ip6cp = (struct ip6ctlparam *)d; 1200 m = ip6cp->ip6c_m; 1201 ip6 = ip6cp->ip6c_ip6; 1202 off = ip6cp->ip6c_off; 1203 sa6_src = ip6cp->ip6c_src; 1204 } else { 1205 m = NULL; 1206 ip6 = NULL; 1207 sa6_src = &sa6_any; 1208 } 1209 1210 if (ip6) { 1211 /* 1212 * XXX: We assume that when ip6 is non NULL, 1213 * M and OFF are valid. 1214 */ 1215 1216 /* check if we can safely examine src and dst ports */ 1217 if (m->m_pkthdr.len < off + sizeof(th)) { 1218 if (cmd == PRC_MSGSIZE) 1219 icmp6_mtudisc_update((struct ip6ctlparam *)d, 0); 1220 return; 1221 } 1222 1223 bzero(&th, sizeof(th)); 1224 m_copydata(m, off, sizeof(th), (caddr_t)&th); 1225 1226 if (cmd == PRC_MSGSIZE) { 1227 int valid = 0; 1228 1229 /* 1230 * Check to see if we have a valid TCP connection 1231 * corresponding to the address in the ICMPv6 message 1232 * payload. 1233 */ 1234 if (in6_pcblookup_connect(&tcb6, &sa6->sin6_addr, 1235 th.th_dport, (struct in6_addr *)&sa6_src->sin6_addr, 1236 th.th_sport, 0)) 1237 valid++; 1238 1239 /* 1240 * Depending on the value of "valid" and routing table 1241 * size (mtudisc_{hi,lo}wat), we will: 1242 * - recalcurate the new MTU and create the 1243 * corresponding routing entry, or 1244 * - ignore the MTU change notification. 1245 */ 1246 icmp6_mtudisc_update((struct ip6ctlparam *)d, valid); 1247 1248 /* 1249 * no need to call in6_pcbnotify, it should have been 1250 * called via callback if necessary 1251 */ 1252 return; 1253 } 1254 1255 nmatch = in6_pcbnotify(&tcb6, sa, th.th_dport, 1256 (struct sockaddr *)sa6_src, th.th_sport, cmd, NULL, notify); 1257 if (nmatch == 0 && syn_cache_count && 1258 (inet6ctlerrmap[cmd] == EHOSTUNREACH || 1259 inet6ctlerrmap[cmd] == ENETUNREACH || 1260 inet6ctlerrmap[cmd] == EHOSTDOWN)) 1261 syn_cache_unreach((struct sockaddr *)sa6_src, 1262 sa, &th); 1263 } else { 1264 (void) in6_pcbnotify(&tcb6, sa, 0, (struct sockaddr *)sa6_src, 1265 0, cmd, NULL, notify); 1266 } 1267 } 1268 #endif 1269 1270 #ifdef INET 1271 /* assumes that ip header and tcp header are contiguous on mbuf */ 1272 void * 1273 tcp_ctlinput(cmd, sa, v) 1274 int cmd; 1275 struct sockaddr *sa; 1276 void *v; 1277 { 1278 struct ip *ip = v; 1279 struct tcphdr *th; 1280 struct icmp *icp; 1281 extern int inetctlerrmap[]; 1282 void (*notify) __P((struct inpcb *, int)) = tcp_notify; 1283 int errno; 1284 int nmatch; 1285 1286 if (sa->sa_family != AF_INET || 1287 sa->sa_len != sizeof(struct sockaddr_in)) 1288 return NULL; 1289 if ((unsigned)cmd >= PRC_NCMDS) 1290 return NULL; 1291 errno = inetctlerrmap[cmd]; 1292 if (cmd == PRC_QUENCH) 1293 notify = tcp_quench; 1294 else if (PRC_IS_REDIRECT(cmd)) 1295 notify = in_rtchange, ip = 0; 1296 else if (cmd == PRC_MSGSIZE && ip_mtudisc && ip && ip->ip_v == 4) { 1297 /* 1298 * Check to see if we have a valid TCP connection 1299 * corresponding to the address in the ICMP message 1300 * payload. 1301 * 1302 * Boundary check is made in icmp_input(), with ICMP_ADVLENMIN. 1303 */ 1304 th = (struct tcphdr *)((caddr_t)ip + (ip->ip_hl << 2)); 1305 if (in_pcblookup_connect(&tcbtable, 1306 ip->ip_dst, th->th_dport, 1307 ip->ip_src, th->th_sport) == NULL) 1308 return NULL; 1309 1310 /* 1311 * Now that we've validated that we are actually communicating 1312 * with the host indicated in the ICMP message, locate the 1313 * ICMP header, recalculate the new MTU, and create the 1314 * corresponding routing entry. 1315 */ 1316 icp = (struct icmp *)((caddr_t)ip - 1317 offsetof(struct icmp, icmp_ip)); 1318 icmp_mtudisc(icp, ip->ip_dst); 1319 1320 return NULL; 1321 } else if (cmd == PRC_HOSTDEAD) 1322 ip = 0; 1323 else if (errno == 0) 1324 return NULL; 1325 if (ip && ip->ip_v == 4 && sa->sa_family == AF_INET) { 1326 th = (struct tcphdr *)((caddr_t)ip + (ip->ip_hl << 2)); 1327 nmatch = in_pcbnotify(&tcbtable, satosin(sa)->sin_addr, 1328 th->th_dport, ip->ip_src, th->th_sport, errno, notify); 1329 if (nmatch == 0 && syn_cache_count && 1330 (inetctlerrmap[cmd] == EHOSTUNREACH || 1331 inetctlerrmap[cmd] == ENETUNREACH || 1332 inetctlerrmap[cmd] == EHOSTDOWN)) { 1333 struct sockaddr_in sin; 1334 bzero(&sin, sizeof(sin)); 1335 sin.sin_len = sizeof(sin); 1336 sin.sin_family = AF_INET; 1337 sin.sin_port = th->th_sport; 1338 sin.sin_addr = ip->ip_src; 1339 syn_cache_unreach((struct sockaddr *)&sin, sa, th); 1340 } 1341 1342 /* XXX mapped address case */ 1343 } else 1344 in_pcbnotifyall(&tcbtable, satosin(sa)->sin_addr, errno, 1345 notify); 1346 return NULL; 1347 } 1348 1349 /* 1350 * When a source quence is received, we are being notifed of congestion. 1351 * Close the congestion window down to the Loss Window (one segment). 1352 * We will gradually open it again as we proceed. 1353 */ 1354 void 1355 tcp_quench(inp, errno) 1356 struct inpcb *inp; 1357 int errno; 1358 { 1359 struct tcpcb *tp = intotcpcb(inp); 1360 1361 if (tp) 1362 tp->snd_cwnd = tp->t_segsz; 1363 } 1364 #endif 1365 1366 #ifdef INET6 1367 void 1368 tcp6_quench(in6p, errno) 1369 struct in6pcb *in6p; 1370 int errno; 1371 { 1372 struct tcpcb *tp = in6totcpcb(in6p); 1373 1374 if (tp) 1375 tp->snd_cwnd = tp->t_segsz; 1376 } 1377 #endif 1378 1379 #ifdef INET 1380 /* 1381 * Path MTU Discovery handlers. 1382 */ 1383 void 1384 tcp_mtudisc_callback(faddr) 1385 struct in_addr faddr; 1386 { 1387 1388 in_pcbnotifyall(&tcbtable, faddr, EMSGSIZE, tcp_mtudisc); 1389 } 1390 1391 /* 1392 * On receipt of path MTU corrections, flush old route and replace it 1393 * with the new one. Retransmit all unacknowledged packets, to ensure 1394 * that all packets will be received. 1395 */ 1396 void 1397 tcp_mtudisc(inp, errno) 1398 struct inpcb *inp; 1399 int errno; 1400 { 1401 struct tcpcb *tp = intotcpcb(inp); 1402 struct rtentry *rt = in_pcbrtentry(inp); 1403 1404 if (tp != 0) { 1405 if (rt != 0) { 1406 /* 1407 * If this was not a host route, remove and realloc. 1408 */ 1409 if ((rt->rt_flags & RTF_HOST) == 0) { 1410 in_rtchange(inp, errno); 1411 if ((rt = in_pcbrtentry(inp)) == 0) 1412 return; 1413 } 1414 1415 /* 1416 * Slow start out of the error condition. We 1417 * use the MTU because we know it's smaller 1418 * than the previously transmitted segment. 1419 * 1420 * Note: This is more conservative than the 1421 * suggestion in draft-floyd-incr-init-win-03. 1422 */ 1423 if (rt->rt_rmx.rmx_mtu != 0) 1424 tp->snd_cwnd = 1425 TCP_INITIAL_WINDOW(tcp_init_win, 1426 rt->rt_rmx.rmx_mtu); 1427 } 1428 1429 /* 1430 * Resend unacknowledged packets. 1431 */ 1432 tp->snd_nxt = tp->snd_una; 1433 tcp_output(tp); 1434 } 1435 } 1436 #endif 1437 1438 #ifdef INET6 1439 /* 1440 * Path MTU Discovery handlers. 1441 */ 1442 void 1443 tcp6_mtudisc_callback(faddr) 1444 struct in6_addr *faddr; 1445 { 1446 struct sockaddr_in6 sin6; 1447 1448 bzero(&sin6, sizeof(sin6)); 1449 sin6.sin6_family = AF_INET6; 1450 sin6.sin6_len = sizeof(struct sockaddr_in6); 1451 sin6.sin6_addr = *faddr; 1452 (void) in6_pcbnotify(&tcb6, (struct sockaddr *)&sin6, 0, 1453 (struct sockaddr *)&sa6_any, 0, PRC_MSGSIZE, NULL, tcp6_mtudisc); 1454 } 1455 1456 void 1457 tcp6_mtudisc(in6p, errno) 1458 struct in6pcb *in6p; 1459 int errno; 1460 { 1461 struct tcpcb *tp = in6totcpcb(in6p); 1462 struct rtentry *rt = in6_pcbrtentry(in6p); 1463 1464 if (tp != 0) { 1465 if (rt != 0) { 1466 /* 1467 * If this was not a host route, remove and realloc. 1468 */ 1469 if ((rt->rt_flags & RTF_HOST) == 0) { 1470 in6_rtchange(in6p, errno); 1471 if ((rt = in6_pcbrtentry(in6p)) == 0) 1472 return; 1473 } 1474 1475 /* 1476 * Slow start out of the error condition. We 1477 * use the MTU because we know it's smaller 1478 * than the previously transmitted segment. 1479 * 1480 * Note: This is more conservative than the 1481 * suggestion in draft-floyd-incr-init-win-03. 1482 */ 1483 if (rt->rt_rmx.rmx_mtu != 0) 1484 tp->snd_cwnd = 1485 TCP_INITIAL_WINDOW(tcp_init_win, 1486 rt->rt_rmx.rmx_mtu); 1487 } 1488 1489 /* 1490 * Resend unacknowledged packets. 1491 */ 1492 tp->snd_nxt = tp->snd_una; 1493 tcp_output(tp); 1494 } 1495 } 1496 #endif /* INET6 */ 1497 1498 /* 1499 * Compute the MSS to advertise to the peer. Called only during 1500 * the 3-way handshake. If we are the server (peer initiated 1501 * connection), we are called with a pointer to the interface 1502 * on which the SYN packet arrived. If we are the client (we 1503 * initiated connection), we are called with a pointer to the 1504 * interface out which this connection should go. 1505 * 1506 * NOTE: Do not subtract IP option/extension header size nor IPsec 1507 * header size from MSS advertisement. MSS option must hold the maximum 1508 * segment size we can accept, so it must always be: 1509 * max(if mtu) - ip header - tcp header 1510 */ 1511 u_long 1512 tcp_mss_to_advertise(ifp, af) 1513 const struct ifnet *ifp; 1514 int af; 1515 { 1516 extern u_long in_maxmtu; 1517 u_long mss = 0; 1518 u_long hdrsiz; 1519 1520 /* 1521 * In order to avoid defeating path MTU discovery on the peer, 1522 * we advertise the max MTU of all attached networks as our MSS, 1523 * per RFC 1191, section 3.1. 1524 * 1525 * We provide the option to advertise just the MTU of 1526 * the interface on which we hope this connection will 1527 * be receiving. If we are responding to a SYN, we 1528 * will have a pretty good idea about this, but when 1529 * initiating a connection there is a bit more doubt. 1530 * 1531 * We also need to ensure that loopback has a large enough 1532 * MSS, as the loopback MTU is never included in in_maxmtu. 1533 */ 1534 1535 if (ifp != NULL) 1536 mss = ifp->if_mtu; 1537 1538 if (tcp_mss_ifmtu == 0) 1539 mss = max(in_maxmtu, mss); 1540 1541 switch (af) { 1542 case AF_INET: 1543 hdrsiz = sizeof(struct ip); 1544 break; 1545 #ifdef INET6 1546 case AF_INET6: 1547 hdrsiz = sizeof(struct ip6_hdr); 1548 break; 1549 #endif 1550 default: 1551 hdrsiz = 0; 1552 break; 1553 } 1554 hdrsiz += sizeof(struct tcphdr); 1555 if (mss > hdrsiz) 1556 mss -= hdrsiz; 1557 1558 mss = max(tcp_mssdflt, mss); 1559 return (mss); 1560 } 1561 1562 /* 1563 * Set connection variables based on the peer's advertised MSS. 1564 * We are passed the TCPCB for the actual connection. If we 1565 * are the server, we are called by the compressed state engine 1566 * when the 3-way handshake is complete. If we are the client, 1567 * we are called when we receive the SYN,ACK from the server. 1568 * 1569 * NOTE: Our advertised MSS value must be initialized in the TCPCB 1570 * before this routine is called! 1571 */ 1572 void 1573 tcp_mss_from_peer(tp, offer) 1574 struct tcpcb *tp; 1575 int offer; 1576 { 1577 struct socket *so; 1578 #if defined(RTV_SPIPE) || defined(RTV_SSTHRESH) 1579 struct rtentry *rt; 1580 #endif 1581 u_long bufsize; 1582 int mss; 1583 1584 #ifdef DIAGNOSTIC 1585 if (tp->t_inpcb && tp->t_in6pcb) 1586 panic("tcp_mss_from_peer: both t_inpcb and t_in6pcb are set"); 1587 #endif 1588 so = NULL; 1589 rt = NULL; 1590 #ifdef INET 1591 if (tp->t_inpcb) { 1592 so = tp->t_inpcb->inp_socket; 1593 #if defined(RTV_SPIPE) || defined(RTV_SSTHRESH) 1594 rt = in_pcbrtentry(tp->t_inpcb); 1595 #endif 1596 } 1597 #endif 1598 #ifdef INET6 1599 if (tp->t_in6pcb) { 1600 so = tp->t_in6pcb->in6p_socket; 1601 #if defined(RTV_SPIPE) || defined(RTV_SSTHRESH) 1602 rt = in6_pcbrtentry(tp->t_in6pcb); 1603 #endif 1604 } 1605 #endif 1606 1607 /* 1608 * As per RFC1122, use the default MSS value, unless they 1609 * sent us an offer. Do not accept offers less than 32 bytes. 1610 */ 1611 mss = tcp_mssdflt; 1612 if (offer) 1613 mss = offer; 1614 mss = max(mss, 32); /* sanity */ 1615 tp->t_peermss = mss; 1616 mss -= tcp_optlen(tp); 1617 #ifdef INET 1618 if (tp->t_inpcb) 1619 mss -= ip_optlen(tp->t_inpcb); 1620 #endif 1621 #ifdef INET6 1622 if (tp->t_in6pcb) 1623 mss -= ip6_optlen(tp->t_in6pcb); 1624 #endif 1625 1626 /* 1627 * If there's a pipesize, change the socket buffer to that size. 1628 * Make the socket buffer an integral number of MSS units. If 1629 * the MSS is larger than the socket buffer, artificially decrease 1630 * the MSS. 1631 */ 1632 #ifdef RTV_SPIPE 1633 if (rt != NULL && rt->rt_rmx.rmx_sendpipe != 0) 1634 bufsize = rt->rt_rmx.rmx_sendpipe; 1635 else 1636 #endif 1637 bufsize = so->so_snd.sb_hiwat; 1638 if (bufsize < mss) 1639 mss = bufsize; 1640 else { 1641 bufsize = roundup(bufsize, mss); 1642 if (bufsize > sb_max) 1643 bufsize = sb_max; 1644 (void) sbreserve(&so->so_snd, bufsize); 1645 } 1646 tp->t_segsz = mss; 1647 1648 #ifdef RTV_SSTHRESH 1649 if (rt != NULL && rt->rt_rmx.rmx_ssthresh) { 1650 /* 1651 * There's some sort of gateway or interface buffer 1652 * limit on the path. Use this to set the slow 1653 * start threshold, but set the threshold to no less 1654 * than 2 * MSS. 1655 */ 1656 tp->snd_ssthresh = max(2 * mss, rt->rt_rmx.rmx_ssthresh); 1657 } 1658 #endif 1659 } 1660 1661 /* 1662 * Processing necessary when a TCP connection is established. 1663 */ 1664 void 1665 tcp_established(tp) 1666 struct tcpcb *tp; 1667 { 1668 struct socket *so; 1669 #ifdef RTV_RPIPE 1670 struct rtentry *rt; 1671 #endif 1672 u_long bufsize; 1673 1674 #ifdef DIAGNOSTIC 1675 if (tp->t_inpcb && tp->t_in6pcb) 1676 panic("tcp_established: both t_inpcb and t_in6pcb are set"); 1677 #endif 1678 so = NULL; 1679 rt = NULL; 1680 #ifdef INET 1681 if (tp->t_inpcb) { 1682 so = tp->t_inpcb->inp_socket; 1683 #if defined(RTV_RPIPE) 1684 rt = in_pcbrtentry(tp->t_inpcb); 1685 #endif 1686 } 1687 #endif 1688 #ifdef INET6 1689 if (tp->t_in6pcb) { 1690 so = tp->t_in6pcb->in6p_socket; 1691 #if defined(RTV_RPIPE) 1692 rt = in6_pcbrtentry(tp->t_in6pcb); 1693 #endif 1694 } 1695 #endif 1696 1697 tp->t_state = TCPS_ESTABLISHED; 1698 TCP_TIMER_ARM(tp, TCPT_KEEP, tcp_keepidle); 1699 1700 #ifdef RTV_RPIPE 1701 if (rt != NULL && rt->rt_rmx.rmx_recvpipe != 0) 1702 bufsize = rt->rt_rmx.rmx_recvpipe; 1703 else 1704 #endif 1705 bufsize = so->so_rcv.sb_hiwat; 1706 if (bufsize > tp->t_ourmss) { 1707 bufsize = roundup(bufsize, tp->t_ourmss); 1708 if (bufsize > sb_max) 1709 bufsize = sb_max; 1710 (void) sbreserve(&so->so_rcv, bufsize); 1711 } 1712 } 1713 1714 /* 1715 * Check if there's an initial rtt or rttvar. Convert from the 1716 * route-table units to scaled multiples of the slow timeout timer. 1717 * Called only during the 3-way handshake. 1718 */ 1719 void 1720 tcp_rmx_rtt(tp) 1721 struct tcpcb *tp; 1722 { 1723 #ifdef RTV_RTT 1724 struct rtentry *rt = NULL; 1725 int rtt; 1726 1727 #ifdef DIAGNOSTIC 1728 if (tp->t_inpcb && tp->t_in6pcb) 1729 panic("tcp_rmx_rtt: both t_inpcb and t_in6pcb are set"); 1730 #endif 1731 #ifdef INET 1732 if (tp->t_inpcb) 1733 rt = in_pcbrtentry(tp->t_inpcb); 1734 #endif 1735 #ifdef INET6 1736 if (tp->t_in6pcb) 1737 rt = in6_pcbrtentry(tp->t_in6pcb); 1738 #endif 1739 if (rt == NULL) 1740 return; 1741 1742 if (tp->t_srtt == 0 && (rtt = rt->rt_rmx.rmx_rtt)) { 1743 /* 1744 * XXX The lock bit for MTU indicates that the value 1745 * is also a minimum value; this is subject to time. 1746 */ 1747 if (rt->rt_rmx.rmx_locks & RTV_RTT) 1748 TCPT_RANGESET(tp->t_rttmin, 1749 rtt / (RTM_RTTUNIT / PR_SLOWHZ), 1750 TCPTV_MIN, TCPTV_REXMTMAX); 1751 tp->t_srtt = rtt / 1752 ((RTM_RTTUNIT / PR_SLOWHZ) >> (TCP_RTT_SHIFT + 2)); 1753 if (rt->rt_rmx.rmx_rttvar) { 1754 tp->t_rttvar = rt->rt_rmx.rmx_rttvar / 1755 ((RTM_RTTUNIT / PR_SLOWHZ) >> 1756 (TCP_RTTVAR_SHIFT + 2)); 1757 } else { 1758 /* Default variation is +- 1 rtt */ 1759 tp->t_rttvar = 1760 tp->t_srtt >> (TCP_RTT_SHIFT - TCP_RTTVAR_SHIFT); 1761 } 1762 TCPT_RANGESET(tp->t_rxtcur, 1763 ((tp->t_srtt >> 2) + tp->t_rttvar) >> (1 + 2), 1764 tp->t_rttmin, TCPTV_REXMTMAX); 1765 } 1766 #endif 1767 } 1768 1769 tcp_seq tcp_iss_seq = 0; /* tcp initial seq # */ 1770 #if NRND > 0 1771 u_int8_t tcp_iss_secret[16]; /* 128 bits; should be plenty */ 1772 #endif 1773 1774 /* 1775 * Get a new sequence value given a tcp control block 1776 */ 1777 tcp_seq 1778 tcp_new_iss(struct tcpcb *tp, tcp_seq addin) 1779 { 1780 1781 #ifdef INET 1782 if (tp->t_inpcb != NULL) { 1783 return (tcp_new_iss1(&tp->t_inpcb->inp_laddr, 1784 &tp->t_inpcb->inp_faddr, tp->t_inpcb->inp_lport, 1785 tp->t_inpcb->inp_fport, sizeof(tp->t_inpcb->inp_laddr), 1786 addin)); 1787 } 1788 #endif 1789 #ifdef INET6 1790 if (tp->t_in6pcb != NULL) { 1791 return (tcp_new_iss1(&tp->t_in6pcb->in6p_laddr, 1792 &tp->t_in6pcb->in6p_faddr, tp->t_in6pcb->in6p_lport, 1793 tp->t_in6pcb->in6p_fport, sizeof(tp->t_in6pcb->in6p_laddr), 1794 addin)); 1795 } 1796 #endif 1797 /* Not possible. */ 1798 panic("tcp_new_iss"); 1799 } 1800 1801 /* 1802 * This routine actually generates a new TCP initial sequence number. 1803 */ 1804 tcp_seq 1805 tcp_new_iss1(void *laddr, void *faddr, u_int16_t lport, u_int16_t fport, 1806 size_t addrsz, tcp_seq addin) 1807 { 1808 tcp_seq tcp_iss; 1809 1810 #if NRND > 0 1811 static int beenhere; 1812 1813 /* 1814 * If we haven't been here before, initialize our cryptographic 1815 * hash secret. 1816 */ 1817 if (beenhere == 0) { 1818 rnd_extract_data(tcp_iss_secret, sizeof(tcp_iss_secret), 1819 RND_EXTRACT_ANY); 1820 beenhere = 1; 1821 } 1822 1823 if (tcp_do_rfc1948) { 1824 MD5_CTX ctx; 1825 u_int8_t hash[16]; /* XXX MD5 knowledge */ 1826 1827 /* 1828 * Compute the base value of the ISS. It is a hash 1829 * of (saddr, sport, daddr, dport, secret). 1830 */ 1831 MD5Init(&ctx); 1832 1833 MD5Update(&ctx, (u_char *) laddr, addrsz); 1834 MD5Update(&ctx, (u_char *) &lport, sizeof(lport)); 1835 1836 MD5Update(&ctx, (u_char *) faddr, addrsz); 1837 MD5Update(&ctx, (u_char *) &fport, sizeof(fport)); 1838 1839 MD5Update(&ctx, tcp_iss_secret, sizeof(tcp_iss_secret)); 1840 1841 MD5Final(hash, &ctx); 1842 1843 memcpy(&tcp_iss, hash, sizeof(tcp_iss)); 1844 1845 /* 1846 * Now increment our "timer", and add it in to 1847 * the computed value. 1848 * 1849 * XXX Use `addin'? 1850 * XXX TCP_ISSINCR too large to use? 1851 */ 1852 tcp_iss_seq += TCP_ISSINCR; 1853 #ifdef TCPISS_DEBUG 1854 printf("ISS hash 0x%08x, ", tcp_iss); 1855 #endif 1856 tcp_iss += tcp_iss_seq + addin; 1857 #ifdef TCPISS_DEBUG 1858 printf("new ISS 0x%08x\n", tcp_iss); 1859 #endif 1860 } else 1861 #endif /* NRND > 0 */ 1862 { 1863 /* 1864 * Randomize. 1865 */ 1866 #if NRND > 0 1867 rnd_extract_data(&tcp_iss, sizeof(tcp_iss), RND_EXTRACT_ANY); 1868 #else 1869 tcp_iss = random(); 1870 #endif 1871 1872 /* 1873 * If we were asked to add some amount to a known value, 1874 * we will take a random value obtained above, mask off 1875 * the upper bits, and add in the known value. We also 1876 * add in a constant to ensure that we are at least a 1877 * certain distance from the original value. 1878 * 1879 * This is used when an old connection is in timed wait 1880 * and we have a new one coming in, for instance. 1881 */ 1882 if (addin != 0) { 1883 #ifdef TCPISS_DEBUG 1884 printf("Random %08x, ", tcp_iss); 1885 #endif 1886 tcp_iss &= TCP_ISS_RANDOM_MASK; 1887 tcp_iss += addin + TCP_ISSINCR; 1888 #ifdef TCPISS_DEBUG 1889 printf("Old ISS %08x, ISS %08x\n", addin, tcp_iss); 1890 #endif 1891 } else { 1892 tcp_iss &= TCP_ISS_RANDOM_MASK; 1893 tcp_iss += tcp_iss_seq; 1894 tcp_iss_seq += TCP_ISSINCR; 1895 #ifdef TCPISS_DEBUG 1896 printf("ISS %08x\n", tcp_iss); 1897 #endif 1898 } 1899 } 1900 1901 if (tcp_compat_42) { 1902 /* 1903 * Limit it to the positive range for really old TCP 1904 * implementations. 1905 */ 1906 if (tcp_iss >= 0x80000000) 1907 tcp_iss &= 0x7fffffff; /* XXX */ 1908 } 1909 1910 return (tcp_iss); 1911 } 1912 1913 #ifdef IPSEC 1914 /* compute ESP/AH header size for TCP, including outer IP header. */ 1915 size_t 1916 ipsec4_hdrsiz_tcp(tp) 1917 struct tcpcb *tp; 1918 { 1919 struct inpcb *inp; 1920 size_t hdrsiz; 1921 1922 /* XXX mapped addr case (tp->t_in6pcb) */ 1923 if (!tp || !tp->t_template || !(inp = tp->t_inpcb)) 1924 return 0; 1925 switch (tp->t_family) { 1926 case AF_INET: 1927 /* XXX: should use currect direction. */ 1928 hdrsiz = ipsec4_hdrsiz(tp->t_template, IPSEC_DIR_OUTBOUND, inp); 1929 break; 1930 default: 1931 hdrsiz = 0; 1932 break; 1933 } 1934 1935 return hdrsiz; 1936 } 1937 1938 #ifdef INET6 1939 size_t 1940 ipsec6_hdrsiz_tcp(tp) 1941 struct tcpcb *tp; 1942 { 1943 struct in6pcb *in6p; 1944 size_t hdrsiz; 1945 1946 if (!tp || !tp->t_template || !(in6p = tp->t_in6pcb)) 1947 return 0; 1948 switch (tp->t_family) { 1949 case AF_INET6: 1950 /* XXX: should use currect direction. */ 1951 hdrsiz = ipsec6_hdrsiz(tp->t_template, IPSEC_DIR_OUTBOUND, in6p); 1952 break; 1953 case AF_INET: 1954 /* mapped address case - tricky */ 1955 default: 1956 hdrsiz = 0; 1957 break; 1958 } 1959 1960 return hdrsiz; 1961 } 1962 #endif 1963 #endif /*IPSEC*/ 1964 1965 /* 1966 * Determine the length of the TCP options for this connection. 1967 * 1968 * XXX: What do we do for SACK, when we add that? Just reserve 1969 * all of the space? Otherwise we can't exactly be incrementing 1970 * cwnd by an amount that varies depending on the amount we last 1971 * had to SACK! 1972 */ 1973 1974 u_int 1975 tcp_optlen(tp) 1976 struct tcpcb *tp; 1977 { 1978 if ((tp->t_flags & (TF_REQ_TSTMP|TF_RCVD_TSTMP|TF_NOOPT)) == 1979 (TF_REQ_TSTMP | TF_RCVD_TSTMP)) 1980 return TCPOLEN_TSTAMP_APPA; 1981 else 1982 return 0; 1983 } 1984