1 /* $NetBSD: tcp_subr.c,v 1.248 2012/09/08 02:58:13 msaitoh 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, 2008 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 * 49 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 50 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 51 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 52 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 53 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 54 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 55 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 56 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 57 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 58 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 59 * POSSIBILITY OF SUCH DAMAGE. 60 */ 61 62 /* 63 * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1995 64 * The Regents of the University of California. All rights reserved. 65 * 66 * Redistribution and use in source and binary forms, with or without 67 * modification, are permitted provided that the following conditions 68 * are met: 69 * 1. Redistributions of source code must retain the above copyright 70 * notice, this list of conditions and the following disclaimer. 71 * 2. Redistributions in binary form must reproduce the above copyright 72 * notice, this list of conditions and the following disclaimer in the 73 * documentation and/or other materials provided with the distribution. 74 * 3. Neither the name of the University nor the names of its contributors 75 * may be used to endorse or promote products derived from this software 76 * without specific prior written permission. 77 * 78 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 79 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 80 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 81 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 82 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 83 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 84 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 85 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 86 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 87 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 88 * SUCH DAMAGE. 89 * 90 * @(#)tcp_subr.c 8.2 (Berkeley) 5/24/95 91 */ 92 93 #include <sys/cdefs.h> 94 __KERNEL_RCSID(0, "$NetBSD: tcp_subr.c,v 1.248 2012/09/08 02:58:13 msaitoh Exp $"); 95 96 #include "opt_inet.h" 97 #include "opt_ipsec.h" 98 #include "opt_tcp_compat_42.h" 99 #include "opt_inet_csum.h" 100 #include "opt_mbuftrace.h" 101 102 #include <sys/param.h> 103 #include <sys/proc.h> 104 #include <sys/systm.h> 105 #include <sys/malloc.h> 106 #include <sys/mbuf.h> 107 #include <sys/socket.h> 108 #include <sys/socketvar.h> 109 #include <sys/protosw.h> 110 #include <sys/errno.h> 111 #include <sys/kernel.h> 112 #include <sys/pool.h> 113 #include <sys/md5.h> 114 #include <sys/cprng.h> 115 116 #include <net/route.h> 117 #include <net/if.h> 118 119 #include <netinet/in.h> 120 #include <netinet/in_systm.h> 121 #include <netinet/ip.h> 122 #include <netinet/in_pcb.h> 123 #include <netinet/ip_var.h> 124 #include <netinet/ip_icmp.h> 125 126 #ifdef INET6 127 #ifndef INET 128 #include <netinet/in.h> 129 #endif 130 #include <netinet/ip6.h> 131 #include <netinet6/in6_pcb.h> 132 #include <netinet6/ip6_var.h> 133 #include <netinet6/in6_var.h> 134 #include <netinet6/ip6protosw.h> 135 #include <netinet/icmp6.h> 136 #include <netinet6/nd6.h> 137 #endif 138 139 #include <netinet/tcp.h> 140 #include <netinet/tcp_fsm.h> 141 #include <netinet/tcp_seq.h> 142 #include <netinet/tcp_timer.h> 143 #include <netinet/tcp_var.h> 144 #include <netinet/tcp_vtw.h> 145 #include <netinet/tcp_private.h> 146 #include <netinet/tcp_congctl.h> 147 #include <netinet/tcpip.h> 148 149 #ifdef FAST_IPSEC 150 #include <netipsec/ipsec.h> 151 #include <netipsec/xform.h> 152 #ifdef INET6 153 #include <netipsec/ipsec6.h> 154 #endif 155 #include <netipsec/key.h> 156 #endif /* FAST_IPSEC*/ 157 158 159 struct inpcbtable tcbtable; /* head of queue of active tcpcb's */ 160 u_int32_t tcp_now; /* slow ticks, for RFC 1323 timestamps */ 161 162 percpu_t *tcpstat_percpu; 163 164 /* patchable/settable parameters for tcp */ 165 int tcp_mssdflt = TCP_MSS; 166 int tcp_minmss = TCP_MINMSS; 167 int tcp_rttdflt = TCPTV_SRTTDFLT / PR_SLOWHZ; 168 int tcp_do_rfc1323 = 1; /* window scaling / timestamps (obsolete) */ 169 int tcp_do_rfc1948 = 0; /* ISS by cryptographic hash */ 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_ack_on_push = 0; /* set to enable immediate ACK-on-PUSH */ 174 int tcp_do_ecn = 0; /* Explicit Congestion Notification */ 175 #ifndef TCP_INIT_WIN 176 #define TCP_INIT_WIN 0 /* initial slow start window */ 177 #endif 178 #ifndef TCP_INIT_WIN_LOCAL 179 #define TCP_INIT_WIN_LOCAL 4 /* initial slow start window for local nets */ 180 #endif 181 int tcp_init_win = TCP_INIT_WIN; 182 int tcp_init_win_local = TCP_INIT_WIN_LOCAL; 183 int tcp_mss_ifmtu = 0; 184 #ifdef TCP_COMPAT_42 185 int tcp_compat_42 = 1; 186 #else 187 int tcp_compat_42 = 0; 188 #endif 189 int tcp_rst_ppslim = 100; /* 100pps */ 190 int tcp_ackdrop_ppslim = 100; /* 100pps */ 191 int tcp_do_loopback_cksum = 0; 192 int tcp_do_abc = 1; /* RFC3465 Appropriate byte counting. */ 193 int tcp_abc_aggressive = 1; /* 1: L=2*SMSS 0: L=1*SMSS */ 194 int tcp_sack_tp_maxholes = 32; 195 int tcp_sack_globalmaxholes = 1024; 196 int tcp_sack_globalholes = 0; 197 int tcp_ecn_maxretries = 1; 198 int tcp_msl_enable = 1; /* enable TIME_WAIT truncation */ 199 int tcp_msl_loop = PR_SLOWHZ; /* MSL for loopback */ 200 int tcp_msl_local = 5 * PR_SLOWHZ; /* MSL for 'local' */ 201 int tcp_msl_remote = TCPTV_MSL; /* MSL otherwise */ 202 int tcp_msl_remote_threshold = TCPTV_SRTTDFLT; /* RTT threshold */ 203 int tcp_rttlocal = 0; /* Use RTT to decide who's 'local' */ 204 205 int tcp4_vtw_enable = 0; /* 1 to enable */ 206 int tcp6_vtw_enable = 0; /* 1 to enable */ 207 int tcp_vtw_was_enabled = 0; 208 int tcp_vtw_entries = 1 << 16; /* 64K vestigial TIME_WAIT entries */ 209 210 /* tcb hash */ 211 #ifndef TCBHASHSIZE 212 #define TCBHASHSIZE 128 213 #endif 214 int tcbhashsize = TCBHASHSIZE; 215 216 /* syn hash parameters */ 217 #define TCP_SYN_HASH_SIZE 293 218 #define TCP_SYN_BUCKET_SIZE 35 219 int tcp_syn_cache_size = TCP_SYN_HASH_SIZE; 220 int tcp_syn_cache_limit = TCP_SYN_HASH_SIZE*TCP_SYN_BUCKET_SIZE; 221 int tcp_syn_bucket_limit = 3*TCP_SYN_BUCKET_SIZE; 222 struct syn_cache_head tcp_syn_cache[TCP_SYN_HASH_SIZE]; 223 224 int tcp_freeq(struct tcpcb *); 225 226 #ifdef INET 227 void tcp_mtudisc_callback(struct in_addr); 228 #endif 229 #ifdef INET6 230 void tcp6_mtudisc_callback(struct in6_addr *); 231 #endif 232 233 #ifdef INET6 234 void tcp6_mtudisc(struct in6pcb *, int); 235 #endif 236 237 static struct pool tcpcb_pool; 238 239 static int tcp_drainwanted; 240 241 #ifdef TCP_CSUM_COUNTERS 242 #include <sys/device.h> 243 244 #if defined(INET) 245 struct evcnt tcp_hwcsum_bad = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 246 NULL, "tcp", "hwcsum bad"); 247 struct evcnt tcp_hwcsum_ok = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 248 NULL, "tcp", "hwcsum ok"); 249 struct evcnt tcp_hwcsum_data = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 250 NULL, "tcp", "hwcsum data"); 251 struct evcnt tcp_swcsum = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 252 NULL, "tcp", "swcsum"); 253 254 EVCNT_ATTACH_STATIC(tcp_hwcsum_bad); 255 EVCNT_ATTACH_STATIC(tcp_hwcsum_ok); 256 EVCNT_ATTACH_STATIC(tcp_hwcsum_data); 257 EVCNT_ATTACH_STATIC(tcp_swcsum); 258 #endif /* defined(INET) */ 259 260 #if defined(INET6) 261 struct evcnt tcp6_hwcsum_bad = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 262 NULL, "tcp6", "hwcsum bad"); 263 struct evcnt tcp6_hwcsum_ok = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 264 NULL, "tcp6", "hwcsum ok"); 265 struct evcnt tcp6_hwcsum_data = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 266 NULL, "tcp6", "hwcsum data"); 267 struct evcnt tcp6_swcsum = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 268 NULL, "tcp6", "swcsum"); 269 270 EVCNT_ATTACH_STATIC(tcp6_hwcsum_bad); 271 EVCNT_ATTACH_STATIC(tcp6_hwcsum_ok); 272 EVCNT_ATTACH_STATIC(tcp6_hwcsum_data); 273 EVCNT_ATTACH_STATIC(tcp6_swcsum); 274 #endif /* defined(INET6) */ 275 #endif /* TCP_CSUM_COUNTERS */ 276 277 278 #ifdef TCP_OUTPUT_COUNTERS 279 #include <sys/device.h> 280 281 struct evcnt tcp_output_bigheader = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 282 NULL, "tcp", "output big header"); 283 struct evcnt tcp_output_predict_hit = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 284 NULL, "tcp", "output predict hit"); 285 struct evcnt tcp_output_predict_miss = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 286 NULL, "tcp", "output predict miss"); 287 struct evcnt tcp_output_copysmall = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 288 NULL, "tcp", "output copy small"); 289 struct evcnt tcp_output_copybig = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 290 NULL, "tcp", "output copy big"); 291 struct evcnt tcp_output_refbig = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 292 NULL, "tcp", "output reference big"); 293 294 EVCNT_ATTACH_STATIC(tcp_output_bigheader); 295 EVCNT_ATTACH_STATIC(tcp_output_predict_hit); 296 EVCNT_ATTACH_STATIC(tcp_output_predict_miss); 297 EVCNT_ATTACH_STATIC(tcp_output_copysmall); 298 EVCNT_ATTACH_STATIC(tcp_output_copybig); 299 EVCNT_ATTACH_STATIC(tcp_output_refbig); 300 301 #endif /* TCP_OUTPUT_COUNTERS */ 302 303 #ifdef TCP_REASS_COUNTERS 304 #include <sys/device.h> 305 306 struct evcnt tcp_reass_ = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 307 NULL, "tcp_reass", "calls"); 308 struct evcnt tcp_reass_empty = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 309 &tcp_reass_, "tcp_reass", "insert into empty queue"); 310 struct evcnt tcp_reass_iteration[8] = { 311 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", ">7 iterations"), 312 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", "1 iteration"), 313 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", "2 iterations"), 314 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", "3 iterations"), 315 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", "4 iterations"), 316 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", "5 iterations"), 317 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", "6 iterations"), 318 EVCNT_INITIALIZER(EVCNT_TYPE_MISC, &tcp_reass_, "tcp_reass", "7 iterations"), 319 }; 320 struct evcnt tcp_reass_prependfirst = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 321 &tcp_reass_, "tcp_reass", "prepend to first"); 322 struct evcnt tcp_reass_prepend = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 323 &tcp_reass_, "tcp_reass", "prepend"); 324 struct evcnt tcp_reass_insert = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 325 &tcp_reass_, "tcp_reass", "insert"); 326 struct evcnt tcp_reass_inserttail = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 327 &tcp_reass_, "tcp_reass", "insert at tail"); 328 struct evcnt tcp_reass_append = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 329 &tcp_reass_, "tcp_reass", "append"); 330 struct evcnt tcp_reass_appendtail = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 331 &tcp_reass_, "tcp_reass", "append to tail fragment"); 332 struct evcnt tcp_reass_overlaptail = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 333 &tcp_reass_, "tcp_reass", "overlap at end"); 334 struct evcnt tcp_reass_overlapfront = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 335 &tcp_reass_, "tcp_reass", "overlap at start"); 336 struct evcnt tcp_reass_segdup = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 337 &tcp_reass_, "tcp_reass", "duplicate segment"); 338 struct evcnt tcp_reass_fragdup = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 339 &tcp_reass_, "tcp_reass", "duplicate fragment"); 340 341 EVCNT_ATTACH_STATIC(tcp_reass_); 342 EVCNT_ATTACH_STATIC(tcp_reass_empty); 343 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 0); 344 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 1); 345 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 2); 346 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 3); 347 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 4); 348 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 5); 349 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 6); 350 EVCNT_ATTACH_STATIC2(tcp_reass_iteration, 7); 351 EVCNT_ATTACH_STATIC(tcp_reass_prependfirst); 352 EVCNT_ATTACH_STATIC(tcp_reass_prepend); 353 EVCNT_ATTACH_STATIC(tcp_reass_insert); 354 EVCNT_ATTACH_STATIC(tcp_reass_inserttail); 355 EVCNT_ATTACH_STATIC(tcp_reass_append); 356 EVCNT_ATTACH_STATIC(tcp_reass_appendtail); 357 EVCNT_ATTACH_STATIC(tcp_reass_overlaptail); 358 EVCNT_ATTACH_STATIC(tcp_reass_overlapfront); 359 EVCNT_ATTACH_STATIC(tcp_reass_segdup); 360 EVCNT_ATTACH_STATIC(tcp_reass_fragdup); 361 362 #endif /* TCP_REASS_COUNTERS */ 363 364 #ifdef MBUFTRACE 365 struct mowner tcp_mowner = MOWNER_INIT("tcp", ""); 366 struct mowner tcp_rx_mowner = MOWNER_INIT("tcp", "rx"); 367 struct mowner tcp_tx_mowner = MOWNER_INIT("tcp", "tx"); 368 struct mowner tcp_sock_mowner = MOWNER_INIT("tcp", "sock"); 369 struct mowner tcp_sock_rx_mowner = MOWNER_INIT("tcp", "sock rx"); 370 struct mowner tcp_sock_tx_mowner = MOWNER_INIT("tcp", "sock tx"); 371 #endif 372 373 /* 374 * Tcp initialization 375 */ 376 void 377 tcp_init(void) 378 { 379 int hlen; 380 381 in_pcbinit(&tcbtable, tcbhashsize, tcbhashsize); 382 pool_init(&tcpcb_pool, sizeof(struct tcpcb), 0, 0, 0, "tcpcbpl", 383 NULL, IPL_SOFTNET); 384 385 hlen = sizeof(struct ip) + sizeof(struct tcphdr); 386 #ifdef INET6 387 if (sizeof(struct ip) < sizeof(struct ip6_hdr)) 388 hlen = sizeof(struct ip6_hdr) + sizeof(struct tcphdr); 389 #endif 390 if (max_protohdr < hlen) 391 max_protohdr = hlen; 392 if (max_linkhdr + hlen > MHLEN) 393 panic("tcp_init"); 394 395 #ifdef INET 396 icmp_mtudisc_callback_register(tcp_mtudisc_callback); 397 #endif 398 #ifdef INET6 399 icmp6_mtudisc_callback_register(tcp6_mtudisc_callback); 400 #endif 401 402 tcp_usrreq_init(); 403 404 /* Initialize timer state. */ 405 tcp_timer_init(); 406 407 /* Initialize the compressed state engine. */ 408 syn_cache_init(); 409 410 /* Initialize the congestion control algorithms. */ 411 tcp_congctl_init(); 412 413 /* Initialize the TCPCB template. */ 414 tcp_tcpcb_template(); 415 416 /* Initialize reassembly queue */ 417 tcpipqent_init(); 418 419 /* SACK */ 420 tcp_sack_init(); 421 422 MOWNER_ATTACH(&tcp_tx_mowner); 423 MOWNER_ATTACH(&tcp_rx_mowner); 424 MOWNER_ATTACH(&tcp_reass_mowner); 425 MOWNER_ATTACH(&tcp_sock_mowner); 426 MOWNER_ATTACH(&tcp_sock_tx_mowner); 427 MOWNER_ATTACH(&tcp_sock_rx_mowner); 428 MOWNER_ATTACH(&tcp_mowner); 429 430 tcpstat_percpu = percpu_alloc(sizeof(uint64_t) * TCP_NSTATS); 431 432 vtw_earlyinit(); 433 } 434 435 /* 436 * Create template to be used to send tcp packets on a connection. 437 * Call after host entry created, allocates an mbuf and fills 438 * in a skeletal tcp/ip header, minimizing the amount of work 439 * necessary when the connection is used. 440 */ 441 struct mbuf * 442 tcp_template(struct tcpcb *tp) 443 { 444 struct inpcb *inp = tp->t_inpcb; 445 #ifdef INET6 446 struct in6pcb *in6p = tp->t_in6pcb; 447 #endif 448 struct tcphdr *n; 449 struct mbuf *m; 450 int hlen; 451 452 switch (tp->t_family) { 453 case AF_INET: 454 hlen = sizeof(struct ip); 455 if (inp) 456 break; 457 #ifdef INET6 458 if (in6p) { 459 /* mapped addr case */ 460 if (IN6_IS_ADDR_V4MAPPED(&in6p->in6p_laddr) 461 && IN6_IS_ADDR_V4MAPPED(&in6p->in6p_faddr)) 462 break; 463 } 464 #endif 465 return NULL; /*EINVAL*/ 466 #ifdef INET6 467 case AF_INET6: 468 hlen = sizeof(struct ip6_hdr); 469 if (in6p) { 470 /* more sainty check? */ 471 break; 472 } 473 return NULL; /*EINVAL*/ 474 #endif 475 default: 476 hlen = 0; /*pacify gcc*/ 477 return NULL; /*EAFNOSUPPORT*/ 478 } 479 #ifdef DIAGNOSTIC 480 if (hlen + sizeof(struct tcphdr) > MCLBYTES) 481 panic("mclbytes too small for t_template"); 482 #endif 483 m = tp->t_template; 484 if (m && m->m_len == hlen + sizeof(struct tcphdr)) 485 ; 486 else { 487 if (m) 488 m_freem(m); 489 m = tp->t_template = NULL; 490 MGETHDR(m, M_DONTWAIT, MT_HEADER); 491 if (m && hlen + sizeof(struct tcphdr) > MHLEN) { 492 MCLGET(m, M_DONTWAIT); 493 if ((m->m_flags & M_EXT) == 0) { 494 m_free(m); 495 m = NULL; 496 } 497 } 498 if (m == NULL) 499 return NULL; 500 MCLAIM(m, &tcp_mowner); 501 m->m_pkthdr.len = m->m_len = hlen + sizeof(struct tcphdr); 502 } 503 504 memset(mtod(m, void *), 0, m->m_len); 505 506 n = (struct tcphdr *)(mtod(m, char *) + hlen); 507 508 switch (tp->t_family) { 509 case AF_INET: 510 { 511 struct ipovly *ipov; 512 mtod(m, struct ip *)->ip_v = 4; 513 mtod(m, struct ip *)->ip_hl = hlen >> 2; 514 ipov = mtod(m, struct ipovly *); 515 ipov->ih_pr = IPPROTO_TCP; 516 ipov->ih_len = htons(sizeof(struct tcphdr)); 517 if (inp) { 518 ipov->ih_src = inp->inp_laddr; 519 ipov->ih_dst = inp->inp_faddr; 520 } 521 #ifdef INET6 522 else if (in6p) { 523 /* mapped addr case */ 524 bcopy(&in6p->in6p_laddr.s6_addr32[3], &ipov->ih_src, 525 sizeof(ipov->ih_src)); 526 bcopy(&in6p->in6p_faddr.s6_addr32[3], &ipov->ih_dst, 527 sizeof(ipov->ih_dst)); 528 } 529 #endif 530 /* 531 * Compute the pseudo-header portion of the checksum 532 * now. We incrementally add in the TCP option and 533 * payload lengths later, and then compute the TCP 534 * checksum right before the packet is sent off onto 535 * the wire. 536 */ 537 n->th_sum = in_cksum_phdr(ipov->ih_src.s_addr, 538 ipov->ih_dst.s_addr, 539 htons(sizeof(struct tcphdr) + IPPROTO_TCP)); 540 break; 541 } 542 #ifdef INET6 543 case AF_INET6: 544 { 545 struct ip6_hdr *ip6; 546 mtod(m, struct ip *)->ip_v = 6; 547 ip6 = mtod(m, struct ip6_hdr *); 548 ip6->ip6_nxt = IPPROTO_TCP; 549 ip6->ip6_plen = htons(sizeof(struct tcphdr)); 550 ip6->ip6_src = in6p->in6p_laddr; 551 ip6->ip6_dst = in6p->in6p_faddr; 552 ip6->ip6_flow = in6p->in6p_flowinfo & IPV6_FLOWINFO_MASK; 553 if (ip6_auto_flowlabel) { 554 ip6->ip6_flow &= ~IPV6_FLOWLABEL_MASK; 555 ip6->ip6_flow |= 556 (htonl(ip6_randomflowlabel()) & IPV6_FLOWLABEL_MASK); 557 } 558 ip6->ip6_vfc &= ~IPV6_VERSION_MASK; 559 ip6->ip6_vfc |= IPV6_VERSION; 560 561 /* 562 * Compute the pseudo-header portion of the checksum 563 * now. We incrementally add in the TCP option and 564 * payload lengths later, and then compute the TCP 565 * checksum right before the packet is sent off onto 566 * the wire. 567 */ 568 n->th_sum = in6_cksum_phdr(&in6p->in6p_laddr, 569 &in6p->in6p_faddr, htonl(sizeof(struct tcphdr)), 570 htonl(IPPROTO_TCP)); 571 break; 572 } 573 #endif 574 } 575 if (inp) { 576 n->th_sport = inp->inp_lport; 577 n->th_dport = inp->inp_fport; 578 } 579 #ifdef INET6 580 else if (in6p) { 581 n->th_sport = in6p->in6p_lport; 582 n->th_dport = in6p->in6p_fport; 583 } 584 #endif 585 n->th_seq = 0; 586 n->th_ack = 0; 587 n->th_x2 = 0; 588 n->th_off = 5; 589 n->th_flags = 0; 590 n->th_win = 0; 591 n->th_urp = 0; 592 return (m); 593 } 594 595 /* 596 * Send a single message to the TCP at address specified by 597 * the given TCP/IP header. If m == 0, then we make a copy 598 * of the tcpiphdr at ti and send directly to the addressed host. 599 * This is used to force keep alive messages out using the TCP 600 * template for a connection tp->t_template. If flags are given 601 * then we send a message back to the TCP which originated the 602 * segment ti, and discard the mbuf containing it and any other 603 * attached mbufs. 604 * 605 * In any case the ack and sequence number of the transmitted 606 * segment are as specified by the parameters. 607 */ 608 int 609 tcp_respond(struct tcpcb *tp, struct mbuf *template, struct mbuf *m, 610 struct tcphdr *th0, tcp_seq ack, tcp_seq seq, int flags) 611 { 612 #ifdef INET6 613 struct rtentry *rt; 614 #endif 615 struct route *ro; 616 int error, tlen, win = 0; 617 int hlen; 618 struct ip *ip; 619 #ifdef INET6 620 struct ip6_hdr *ip6; 621 #endif 622 int family; /* family on packet, not inpcb/in6pcb! */ 623 struct tcphdr *th; 624 struct socket *so; 625 626 if (tp != NULL && (flags & TH_RST) == 0) { 627 #ifdef DIAGNOSTIC 628 if (tp->t_inpcb && tp->t_in6pcb) 629 panic("tcp_respond: both t_inpcb and t_in6pcb are set"); 630 #endif 631 #ifdef INET 632 if (tp->t_inpcb) 633 win = sbspace(&tp->t_inpcb->inp_socket->so_rcv); 634 #endif 635 #ifdef INET6 636 if (tp->t_in6pcb) 637 win = sbspace(&tp->t_in6pcb->in6p_socket->so_rcv); 638 #endif 639 } 640 641 th = NULL; /* Quell uninitialized warning */ 642 ip = NULL; 643 #ifdef INET6 644 ip6 = NULL; 645 #endif 646 if (m == 0) { 647 if (!template) 648 return EINVAL; 649 650 /* get family information from template */ 651 switch (mtod(template, struct ip *)->ip_v) { 652 case 4: 653 family = AF_INET; 654 hlen = sizeof(struct ip); 655 break; 656 #ifdef INET6 657 case 6: 658 family = AF_INET6; 659 hlen = sizeof(struct ip6_hdr); 660 break; 661 #endif 662 default: 663 return EAFNOSUPPORT; 664 } 665 666 MGETHDR(m, M_DONTWAIT, MT_HEADER); 667 if (m) { 668 MCLAIM(m, &tcp_tx_mowner); 669 MCLGET(m, M_DONTWAIT); 670 if ((m->m_flags & M_EXT) == 0) { 671 m_free(m); 672 m = NULL; 673 } 674 } 675 if (m == NULL) 676 return (ENOBUFS); 677 678 if (tcp_compat_42) 679 tlen = 1; 680 else 681 tlen = 0; 682 683 m->m_data += max_linkhdr; 684 bcopy(mtod(template, void *), mtod(m, void *), 685 template->m_len); 686 switch (family) { 687 case AF_INET: 688 ip = mtod(m, struct ip *); 689 th = (struct tcphdr *)(ip + 1); 690 break; 691 #ifdef INET6 692 case AF_INET6: 693 ip6 = mtod(m, struct ip6_hdr *); 694 th = (struct tcphdr *)(ip6 + 1); 695 break; 696 #endif 697 #if 0 698 default: 699 /* noone will visit here */ 700 m_freem(m); 701 return EAFNOSUPPORT; 702 #endif 703 } 704 flags = TH_ACK; 705 } else { 706 707 if ((m->m_flags & M_PKTHDR) == 0) { 708 #if 0 709 printf("non PKTHDR to tcp_respond\n"); 710 #endif 711 m_freem(m); 712 return EINVAL; 713 } 714 #ifdef DIAGNOSTIC 715 if (!th0) 716 panic("th0 == NULL in tcp_respond"); 717 #endif 718 719 /* get family information from m */ 720 switch (mtod(m, struct ip *)->ip_v) { 721 case 4: 722 family = AF_INET; 723 hlen = sizeof(struct ip); 724 ip = mtod(m, struct ip *); 725 break; 726 #ifdef INET6 727 case 6: 728 family = AF_INET6; 729 hlen = sizeof(struct ip6_hdr); 730 ip6 = mtod(m, struct ip6_hdr *); 731 break; 732 #endif 733 default: 734 m_freem(m); 735 return EAFNOSUPPORT; 736 } 737 /* clear h/w csum flags inherited from rx packet */ 738 m->m_pkthdr.csum_flags = 0; 739 740 if ((flags & TH_SYN) == 0 || sizeof(*th0) > (th0->th_off << 2)) 741 tlen = sizeof(*th0); 742 else 743 tlen = th0->th_off << 2; 744 745 if (m->m_len > hlen + tlen && (m->m_flags & M_EXT) == 0 && 746 mtod(m, char *) + hlen == (char *)th0) { 747 m->m_len = hlen + tlen; 748 m_freem(m->m_next); 749 m->m_next = NULL; 750 } else { 751 struct mbuf *n; 752 753 #ifdef DIAGNOSTIC 754 if (max_linkhdr + hlen + tlen > MCLBYTES) { 755 m_freem(m); 756 return EMSGSIZE; 757 } 758 #endif 759 MGETHDR(n, M_DONTWAIT, MT_HEADER); 760 if (n && max_linkhdr + hlen + tlen > MHLEN) { 761 MCLGET(n, M_DONTWAIT); 762 if ((n->m_flags & M_EXT) == 0) { 763 m_freem(n); 764 n = NULL; 765 } 766 } 767 if (!n) { 768 m_freem(m); 769 return ENOBUFS; 770 } 771 772 MCLAIM(n, &tcp_tx_mowner); 773 n->m_data += max_linkhdr; 774 n->m_len = hlen + tlen; 775 m_copyback(n, 0, hlen, mtod(m, void *)); 776 m_copyback(n, hlen, tlen, (void *)th0); 777 778 m_freem(m); 779 m = n; 780 n = NULL; 781 } 782 783 #define xchg(a,b,type) { type t; t=a; a=b; b=t; } 784 switch (family) { 785 case AF_INET: 786 ip = mtod(m, struct ip *); 787 th = (struct tcphdr *)(ip + 1); 788 ip->ip_p = IPPROTO_TCP; 789 xchg(ip->ip_dst, ip->ip_src, struct in_addr); 790 ip->ip_p = IPPROTO_TCP; 791 break; 792 #ifdef INET6 793 case AF_INET6: 794 ip6 = mtod(m, struct ip6_hdr *); 795 th = (struct tcphdr *)(ip6 + 1); 796 ip6->ip6_nxt = IPPROTO_TCP; 797 xchg(ip6->ip6_dst, ip6->ip6_src, struct in6_addr); 798 ip6->ip6_nxt = IPPROTO_TCP; 799 break; 800 #endif 801 #if 0 802 default: 803 /* noone will visit here */ 804 m_freem(m); 805 return EAFNOSUPPORT; 806 #endif 807 } 808 xchg(th->th_dport, th->th_sport, u_int16_t); 809 #undef xchg 810 tlen = 0; /*be friendly with the following code*/ 811 } 812 th->th_seq = htonl(seq); 813 th->th_ack = htonl(ack); 814 th->th_x2 = 0; 815 if ((flags & TH_SYN) == 0) { 816 if (tp) 817 win >>= tp->rcv_scale; 818 if (win > TCP_MAXWIN) 819 win = TCP_MAXWIN; 820 th->th_win = htons((u_int16_t)win); 821 th->th_off = sizeof (struct tcphdr) >> 2; 822 tlen += sizeof(*th); 823 } else 824 tlen += th->th_off << 2; 825 m->m_len = hlen + tlen; 826 m->m_pkthdr.len = hlen + tlen; 827 m->m_pkthdr.rcvif = NULL; 828 th->th_flags = flags; 829 th->th_urp = 0; 830 831 switch (family) { 832 #ifdef INET 833 case AF_INET: 834 { 835 struct ipovly *ipov = (struct ipovly *)ip; 836 memset(ipov->ih_x1, 0, sizeof ipov->ih_x1); 837 ipov->ih_len = htons((u_int16_t)tlen); 838 839 th->th_sum = 0; 840 th->th_sum = in_cksum(m, hlen + tlen); 841 ip->ip_len = htons(hlen + tlen); 842 ip->ip_ttl = ip_defttl; 843 break; 844 } 845 #endif 846 #ifdef INET6 847 case AF_INET6: 848 { 849 th->th_sum = 0; 850 th->th_sum = in6_cksum(m, IPPROTO_TCP, sizeof(struct ip6_hdr), 851 tlen); 852 ip6->ip6_plen = htons(tlen); 853 if (tp && tp->t_in6pcb) { 854 struct ifnet *oifp; 855 ro = &tp->t_in6pcb->in6p_route; 856 oifp = (rt = rtcache_validate(ro)) != NULL ? rt->rt_ifp 857 : NULL; 858 ip6->ip6_hlim = in6_selecthlim(tp->t_in6pcb, oifp); 859 } else 860 ip6->ip6_hlim = ip6_defhlim; 861 ip6->ip6_flow &= ~IPV6_FLOWINFO_MASK; 862 if (ip6_auto_flowlabel) { 863 ip6->ip6_flow |= 864 (htonl(ip6_randomflowlabel()) & IPV6_FLOWLABEL_MASK); 865 } 866 break; 867 } 868 #endif 869 } 870 871 if (tp && tp->t_inpcb) 872 so = tp->t_inpcb->inp_socket; 873 #ifdef INET6 874 else if (tp && tp->t_in6pcb) 875 so = tp->t_in6pcb->in6p_socket; 876 #endif 877 else 878 so = NULL; 879 880 if (tp != NULL && tp->t_inpcb != NULL) { 881 ro = &tp->t_inpcb->inp_route; 882 #ifdef DIAGNOSTIC 883 if (family != AF_INET) 884 panic("tcp_respond: address family mismatch"); 885 if (!in_hosteq(ip->ip_dst, tp->t_inpcb->inp_faddr)) { 886 panic("tcp_respond: ip_dst %x != inp_faddr %x", 887 ntohl(ip->ip_dst.s_addr), 888 ntohl(tp->t_inpcb->inp_faddr.s_addr)); 889 } 890 #endif 891 } 892 #ifdef INET6 893 else if (tp != NULL && tp->t_in6pcb != NULL) { 894 ro = (struct route *)&tp->t_in6pcb->in6p_route; 895 #ifdef DIAGNOSTIC 896 if (family == AF_INET) { 897 if (!IN6_IS_ADDR_V4MAPPED(&tp->t_in6pcb->in6p_faddr)) 898 panic("tcp_respond: not mapped addr"); 899 if (memcmp(&ip->ip_dst, 900 &tp->t_in6pcb->in6p_faddr.s6_addr32[3], 901 sizeof(ip->ip_dst)) != 0) { 902 panic("tcp_respond: ip_dst != in6p_faddr"); 903 } 904 } else if (family == AF_INET6) { 905 if (!IN6_ARE_ADDR_EQUAL(&ip6->ip6_dst, 906 &tp->t_in6pcb->in6p_faddr)) 907 panic("tcp_respond: ip6_dst != in6p_faddr"); 908 } else 909 panic("tcp_respond: address family mismatch"); 910 #endif 911 } 912 #endif 913 else 914 ro = NULL; 915 916 switch (family) { 917 #ifdef INET 918 case AF_INET: 919 error = ip_output(m, NULL, ro, 920 (tp && tp->t_mtudisc ? IP_MTUDISC : 0), NULL, so); 921 break; 922 #endif 923 #ifdef INET6 924 case AF_INET6: 925 error = ip6_output(m, NULL, ro, 0, NULL, so, NULL); 926 break; 927 #endif 928 default: 929 error = EAFNOSUPPORT; 930 break; 931 } 932 933 return (error); 934 } 935 936 /* 937 * Template TCPCB. Rather than zeroing a new TCPCB and initializing 938 * a bunch of members individually, we maintain this template for the 939 * static and mostly-static components of the TCPCB, and copy it into 940 * the new TCPCB instead. 941 */ 942 static struct tcpcb tcpcb_template = { 943 .t_srtt = TCPTV_SRTTBASE, 944 .t_rttmin = TCPTV_MIN, 945 946 .snd_cwnd = TCP_MAXWIN << TCP_MAX_WINSHIFT, 947 .snd_ssthresh = TCP_MAXWIN << TCP_MAX_WINSHIFT, 948 .snd_numholes = 0, 949 950 .t_partialacks = -1, 951 .t_bytes_acked = 0, 952 }; 953 954 /* 955 * Updates the TCPCB template whenever a parameter that would affect 956 * the template is changed. 957 */ 958 void 959 tcp_tcpcb_template(void) 960 { 961 struct tcpcb *tp = &tcpcb_template; 962 int flags; 963 964 tp->t_peermss = tcp_mssdflt; 965 tp->t_ourmss = tcp_mssdflt; 966 tp->t_segsz = tcp_mssdflt; 967 968 flags = 0; 969 if (tcp_do_rfc1323 && tcp_do_win_scale) 970 flags |= TF_REQ_SCALE; 971 if (tcp_do_rfc1323 && tcp_do_timestamps) 972 flags |= TF_REQ_TSTMP; 973 tp->t_flags = flags; 974 975 /* 976 * Init srtt to TCPTV_SRTTBASE (0), so we can tell that we have no 977 * rtt estimate. Set rttvar so that srtt + 2 * rttvar gives 978 * reasonable initial retransmit time. 979 */ 980 tp->t_rttvar = tcp_rttdflt * PR_SLOWHZ << (TCP_RTTVAR_SHIFT + 2 - 1); 981 TCPT_RANGESET(tp->t_rxtcur, TCP_REXMTVAL(tp), 982 TCPTV_MIN, TCPTV_REXMTMAX); 983 984 /* Keep Alive */ 985 tp->t_keepinit = tcp_keepinit; 986 tp->t_keepidle = tcp_keepidle; 987 tp->t_keepintvl = tcp_keepintvl; 988 tp->t_keepcnt = tcp_keepcnt; 989 tp->t_maxidle = tp->t_keepcnt * tp->t_keepintvl; 990 991 /* MSL */ 992 tp->t_msl = TCPTV_MSL; 993 } 994 995 /* 996 * Create a new TCP control block, making an 997 * empty reassembly queue and hooking it to the argument 998 * protocol control block. 999 */ 1000 /* family selects inpcb, or in6pcb */ 1001 struct tcpcb * 1002 tcp_newtcpcb(int family, void *aux) 1003 { 1004 #ifdef INET6 1005 struct rtentry *rt; 1006 #endif 1007 struct tcpcb *tp; 1008 int i; 1009 1010 /* XXX Consider using a pool_cache for speed. */ 1011 tp = pool_get(&tcpcb_pool, PR_NOWAIT); /* splsoftnet via tcp_usrreq */ 1012 if (tp == NULL) 1013 return (NULL); 1014 memcpy(tp, &tcpcb_template, sizeof(*tp)); 1015 TAILQ_INIT(&tp->segq); 1016 TAILQ_INIT(&tp->timeq); 1017 tp->t_family = family; /* may be overridden later on */ 1018 TAILQ_INIT(&tp->snd_holes); 1019 LIST_INIT(&tp->t_sc); /* XXX can template this */ 1020 1021 /* Don't sweat this loop; hopefully the compiler will unroll it. */ 1022 for (i = 0; i < TCPT_NTIMERS; i++) { 1023 callout_init(&tp->t_timer[i], CALLOUT_MPSAFE); 1024 TCP_TIMER_INIT(tp, i); 1025 } 1026 callout_init(&tp->t_delack_ch, CALLOUT_MPSAFE); 1027 1028 switch (family) { 1029 case AF_INET: 1030 { 1031 struct inpcb *inp = (struct inpcb *)aux; 1032 1033 inp->inp_ip.ip_ttl = ip_defttl; 1034 inp->inp_ppcb = (void *)tp; 1035 1036 tp->t_inpcb = inp; 1037 tp->t_mtudisc = ip_mtudisc; 1038 break; 1039 } 1040 #ifdef INET6 1041 case AF_INET6: 1042 { 1043 struct in6pcb *in6p = (struct in6pcb *)aux; 1044 1045 in6p->in6p_ip6.ip6_hlim = in6_selecthlim(in6p, 1046 (rt = rtcache_validate(&in6p->in6p_route)) != NULL 1047 ? rt->rt_ifp 1048 : NULL); 1049 in6p->in6p_ppcb = (void *)tp; 1050 1051 tp->t_in6pcb = in6p; 1052 /* for IPv6, always try to run path MTU discovery */ 1053 tp->t_mtudisc = 1; 1054 break; 1055 } 1056 #endif /* INET6 */ 1057 default: 1058 for (i = 0; i < TCPT_NTIMERS; i++) 1059 callout_destroy(&tp->t_timer[i]); 1060 callout_destroy(&tp->t_delack_ch); 1061 pool_put(&tcpcb_pool, tp); /* splsoftnet via tcp_usrreq */ 1062 return (NULL); 1063 } 1064 1065 /* 1066 * Initialize our timebase. When we send timestamps, we take 1067 * the delta from tcp_now -- this means each connection always 1068 * gets a timebase of 1, which makes it, among other things, 1069 * more difficult to determine how long a system has been up, 1070 * and thus how many TCP sequence increments have occurred. 1071 * 1072 * We start with 1, because 0 doesn't work with linux, which 1073 * considers timestamp 0 in a SYN packet as a bug and disables 1074 * timestamps. 1075 */ 1076 tp->ts_timebase = tcp_now - 1; 1077 1078 tcp_congctl_select(tp, tcp_congctl_global_name); 1079 1080 return (tp); 1081 } 1082 1083 /* 1084 * Drop a TCP connection, reporting 1085 * the specified error. If connection is synchronized, 1086 * then send a RST to peer. 1087 */ 1088 struct tcpcb * 1089 tcp_drop(struct tcpcb *tp, int errno) 1090 { 1091 struct socket *so = NULL; 1092 1093 #ifdef DIAGNOSTIC 1094 if (tp->t_inpcb && tp->t_in6pcb) 1095 panic("tcp_drop: both t_inpcb and t_in6pcb are set"); 1096 #endif 1097 #ifdef INET 1098 if (tp->t_inpcb) 1099 so = tp->t_inpcb->inp_socket; 1100 #endif 1101 #ifdef INET6 1102 if (tp->t_in6pcb) 1103 so = tp->t_in6pcb->in6p_socket; 1104 #endif 1105 if (!so) 1106 return NULL; 1107 1108 if (TCPS_HAVERCVDSYN(tp->t_state)) { 1109 tp->t_state = TCPS_CLOSED; 1110 (void) tcp_output(tp); 1111 TCP_STATINC(TCP_STAT_DROPS); 1112 } else 1113 TCP_STATINC(TCP_STAT_CONNDROPS); 1114 if (errno == ETIMEDOUT && tp->t_softerror) 1115 errno = tp->t_softerror; 1116 so->so_error = errno; 1117 return (tcp_close(tp)); 1118 } 1119 1120 /* 1121 * Close a TCP control block: 1122 * discard all space held by the tcp 1123 * discard internet protocol block 1124 * wake up any sleepers 1125 */ 1126 struct tcpcb * 1127 tcp_close(struct tcpcb *tp) 1128 { 1129 struct inpcb *inp; 1130 #ifdef INET6 1131 struct in6pcb *in6p; 1132 #endif 1133 struct socket *so; 1134 #ifdef RTV_RTT 1135 struct rtentry *rt; 1136 #endif 1137 struct route *ro; 1138 int j; 1139 1140 inp = tp->t_inpcb; 1141 #ifdef INET6 1142 in6p = tp->t_in6pcb; 1143 #endif 1144 so = NULL; 1145 ro = NULL; 1146 if (inp) { 1147 so = inp->inp_socket; 1148 ro = &inp->inp_route; 1149 } 1150 #ifdef INET6 1151 else if (in6p) { 1152 so = in6p->in6p_socket; 1153 ro = (struct route *)&in6p->in6p_route; 1154 } 1155 #endif 1156 1157 #ifdef RTV_RTT 1158 /* 1159 * If we sent enough data to get some meaningful characteristics, 1160 * save them in the routing entry. 'Enough' is arbitrarily 1161 * defined as the sendpipesize (default 4K) * 16. This would 1162 * give us 16 rtt samples assuming we only get one sample per 1163 * window (the usual case on a long haul net). 16 samples is 1164 * enough for the srtt filter to converge to within 5% of the correct 1165 * value; fewer samples and we could save a very bogus rtt. 1166 * 1167 * Don't update the default route's characteristics and don't 1168 * update anything that the user "locked". 1169 */ 1170 if (SEQ_LT(tp->iss + so->so_snd.sb_hiwat * 16, tp->snd_max) && 1171 ro && (rt = rtcache_validate(ro)) != NULL && 1172 !in_nullhost(satocsin(rt_getkey(rt))->sin_addr)) { 1173 u_long i = 0; 1174 1175 if ((rt->rt_rmx.rmx_locks & RTV_RTT) == 0) { 1176 i = tp->t_srtt * 1177 ((RTM_RTTUNIT / PR_SLOWHZ) >> (TCP_RTT_SHIFT + 2)); 1178 if (rt->rt_rmx.rmx_rtt && i) 1179 /* 1180 * filter this update to half the old & half 1181 * the new values, converting scale. 1182 * See route.h and tcp_var.h for a 1183 * description of the scaling constants. 1184 */ 1185 rt->rt_rmx.rmx_rtt = 1186 (rt->rt_rmx.rmx_rtt + i) / 2; 1187 else 1188 rt->rt_rmx.rmx_rtt = i; 1189 } 1190 if ((rt->rt_rmx.rmx_locks & RTV_RTTVAR) == 0) { 1191 i = tp->t_rttvar * 1192 ((RTM_RTTUNIT / PR_SLOWHZ) >> (TCP_RTTVAR_SHIFT + 2)); 1193 if (rt->rt_rmx.rmx_rttvar && i) 1194 rt->rt_rmx.rmx_rttvar = 1195 (rt->rt_rmx.rmx_rttvar + i) / 2; 1196 else 1197 rt->rt_rmx.rmx_rttvar = i; 1198 } 1199 /* 1200 * update the pipelimit (ssthresh) if it has been updated 1201 * already or if a pipesize was specified & the threshhold 1202 * got below half the pipesize. I.e., wait for bad news 1203 * before we start updating, then update on both good 1204 * and bad news. 1205 */ 1206 if (((rt->rt_rmx.rmx_locks & RTV_SSTHRESH) == 0 && 1207 (i = tp->snd_ssthresh) && rt->rt_rmx.rmx_ssthresh) || 1208 i < (rt->rt_rmx.rmx_sendpipe / 2)) { 1209 /* 1210 * convert the limit from user data bytes to 1211 * packets then to packet data bytes. 1212 */ 1213 i = (i + tp->t_segsz / 2) / tp->t_segsz; 1214 if (i < 2) 1215 i = 2; 1216 i *= (u_long)(tp->t_segsz + sizeof (struct tcpiphdr)); 1217 if (rt->rt_rmx.rmx_ssthresh) 1218 rt->rt_rmx.rmx_ssthresh = 1219 (rt->rt_rmx.rmx_ssthresh + i) / 2; 1220 else 1221 rt->rt_rmx.rmx_ssthresh = i; 1222 } 1223 } 1224 #endif /* RTV_RTT */ 1225 /* free the reassembly queue, if any */ 1226 TCP_REASS_LOCK(tp); 1227 (void) tcp_freeq(tp); 1228 TCP_REASS_UNLOCK(tp); 1229 1230 /* free the SACK holes list. */ 1231 tcp_free_sackholes(tp); 1232 tcp_congctl_release(tp); 1233 syn_cache_cleanup(tp); 1234 1235 if (tp->t_template) { 1236 m_free(tp->t_template); 1237 tp->t_template = NULL; 1238 } 1239 1240 /* 1241 * Detaching the pcb will unlock the socket/tcpcb, and stopping 1242 * the timers can also drop the lock. We need to prevent access 1243 * to the tcpcb as it's half torn down. Flag the pcb as dead 1244 * (prevents access by timers) and only then detach it. 1245 */ 1246 tp->t_flags |= TF_DEAD; 1247 if (inp) { 1248 inp->inp_ppcb = 0; 1249 soisdisconnected(so); 1250 in_pcbdetach(inp); 1251 } 1252 #ifdef INET6 1253 else if (in6p) { 1254 in6p->in6p_ppcb = 0; 1255 soisdisconnected(so); 1256 in6_pcbdetach(in6p); 1257 } 1258 #endif 1259 /* 1260 * pcb is no longer visble elsewhere, so we can safely release 1261 * the lock in callout_halt() if needed. 1262 */ 1263 TCP_STATINC(TCP_STAT_CLOSED); 1264 for (j = 0; j < TCPT_NTIMERS; j++) { 1265 callout_halt(&tp->t_timer[j], softnet_lock); 1266 callout_destroy(&tp->t_timer[j]); 1267 } 1268 callout_halt(&tp->t_delack_ch, softnet_lock); 1269 callout_destroy(&tp->t_delack_ch); 1270 pool_put(&tcpcb_pool, tp); 1271 1272 return NULL; 1273 } 1274 1275 int 1276 tcp_freeq(struct tcpcb *tp) 1277 { 1278 struct ipqent *qe; 1279 int rv = 0; 1280 #ifdef TCPREASS_DEBUG 1281 int i = 0; 1282 #endif 1283 1284 TCP_REASS_LOCK_CHECK(tp); 1285 1286 while ((qe = TAILQ_FIRST(&tp->segq)) != NULL) { 1287 #ifdef TCPREASS_DEBUG 1288 printf("tcp_freeq[%p,%d]: %u:%u(%u) 0x%02x\n", 1289 tp, i++, qe->ipqe_seq, qe->ipqe_seq + qe->ipqe_len, 1290 qe->ipqe_len, qe->ipqe_flags & (TH_SYN|TH_FIN|TH_RST)); 1291 #endif 1292 TAILQ_REMOVE(&tp->segq, qe, ipqe_q); 1293 TAILQ_REMOVE(&tp->timeq, qe, ipqe_timeq); 1294 m_freem(qe->ipqe_m); 1295 tcpipqent_free(qe); 1296 rv = 1; 1297 } 1298 tp->t_segqlen = 0; 1299 KASSERT(TAILQ_EMPTY(&tp->timeq)); 1300 return (rv); 1301 } 1302 1303 void 1304 tcp_fasttimo(void) 1305 { 1306 if (tcp_drainwanted) { 1307 tcp_drain(); 1308 tcp_drainwanted = 0; 1309 } 1310 } 1311 1312 void 1313 tcp_drainstub(void) 1314 { 1315 tcp_drainwanted = 1; 1316 } 1317 1318 /* 1319 * Protocol drain routine. Called when memory is in short supply. 1320 * Called from pr_fasttimo thus a callout context. 1321 */ 1322 void 1323 tcp_drain(void) 1324 { 1325 struct inpcb_hdr *inph; 1326 struct tcpcb *tp; 1327 1328 mutex_enter(softnet_lock); 1329 KERNEL_LOCK(1, NULL); 1330 1331 /* 1332 * Free the sequence queue of all TCP connections. 1333 */ 1334 CIRCLEQ_FOREACH(inph, &tcbtable.inpt_queue, inph_queue) { 1335 switch (inph->inph_af) { 1336 case AF_INET: 1337 tp = intotcpcb((struct inpcb *)inph); 1338 break; 1339 #ifdef INET6 1340 case AF_INET6: 1341 tp = in6totcpcb((struct in6pcb *)inph); 1342 break; 1343 #endif 1344 default: 1345 tp = NULL; 1346 break; 1347 } 1348 if (tp != NULL) { 1349 /* 1350 * We may be called from a device's interrupt 1351 * context. If the tcpcb is already busy, 1352 * just bail out now. 1353 */ 1354 if (tcp_reass_lock_try(tp) == 0) 1355 continue; 1356 if (tcp_freeq(tp)) 1357 TCP_STATINC(TCP_STAT_CONNSDRAINED); 1358 TCP_REASS_UNLOCK(tp); 1359 } 1360 } 1361 1362 KERNEL_UNLOCK_ONE(NULL); 1363 mutex_exit(softnet_lock); 1364 } 1365 1366 /* 1367 * Notify a tcp user of an asynchronous error; 1368 * store error as soft error, but wake up user 1369 * (for now, won't do anything until can select for soft error). 1370 */ 1371 void 1372 tcp_notify(struct inpcb *inp, int error) 1373 { 1374 struct tcpcb *tp = (struct tcpcb *)inp->inp_ppcb; 1375 struct socket *so = inp->inp_socket; 1376 1377 /* 1378 * Ignore some errors if we are hooked up. 1379 * If connection hasn't completed, has retransmitted several times, 1380 * and receives a second error, give up now. This is better 1381 * than waiting a long time to establish a connection that 1382 * can never complete. 1383 */ 1384 if (tp->t_state == TCPS_ESTABLISHED && 1385 (error == EHOSTUNREACH || error == ENETUNREACH || 1386 error == EHOSTDOWN)) { 1387 return; 1388 } else if (TCPS_HAVEESTABLISHED(tp->t_state) == 0 && 1389 tp->t_rxtshift > 3 && tp->t_softerror) 1390 so->so_error = error; 1391 else 1392 tp->t_softerror = error; 1393 cv_broadcast(&so->so_cv); 1394 sorwakeup(so); 1395 sowwakeup(so); 1396 } 1397 1398 #ifdef INET6 1399 void 1400 tcp6_notify(struct in6pcb *in6p, int error) 1401 { 1402 struct tcpcb *tp = (struct tcpcb *)in6p->in6p_ppcb; 1403 struct socket *so = in6p->in6p_socket; 1404 1405 /* 1406 * Ignore some errors if we are hooked up. 1407 * If connection hasn't completed, has retransmitted several times, 1408 * and receives a second error, give up now. This is better 1409 * than waiting a long time to establish a connection that 1410 * can never complete. 1411 */ 1412 if (tp->t_state == TCPS_ESTABLISHED && 1413 (error == EHOSTUNREACH || error == ENETUNREACH || 1414 error == EHOSTDOWN)) { 1415 return; 1416 } else if (TCPS_HAVEESTABLISHED(tp->t_state) == 0 && 1417 tp->t_rxtshift > 3 && tp->t_softerror) 1418 so->so_error = error; 1419 else 1420 tp->t_softerror = error; 1421 cv_broadcast(&so->so_cv); 1422 sorwakeup(so); 1423 sowwakeup(so); 1424 } 1425 #endif 1426 1427 #ifdef INET6 1428 void * 1429 tcp6_ctlinput(int cmd, const struct sockaddr *sa, void *d) 1430 { 1431 struct tcphdr th; 1432 void (*notify)(struct in6pcb *, int) = tcp6_notify; 1433 int nmatch; 1434 struct ip6_hdr *ip6; 1435 const struct sockaddr_in6 *sa6_src = NULL; 1436 const struct sockaddr_in6 *sa6 = (const struct sockaddr_in6 *)sa; 1437 struct mbuf *m; 1438 int off; 1439 1440 if (sa->sa_family != AF_INET6 || 1441 sa->sa_len != sizeof(struct sockaddr_in6)) 1442 return NULL; 1443 if ((unsigned)cmd >= PRC_NCMDS) 1444 return NULL; 1445 else if (cmd == PRC_QUENCH) { 1446 /* 1447 * Don't honor ICMP Source Quench messages meant for 1448 * TCP connections. 1449 */ 1450 return NULL; 1451 } else if (PRC_IS_REDIRECT(cmd)) 1452 notify = in6_rtchange, d = NULL; 1453 else if (cmd == PRC_MSGSIZE) 1454 ; /* special code is present, see below */ 1455 else if (cmd == PRC_HOSTDEAD) 1456 d = NULL; 1457 else if (inet6ctlerrmap[cmd] == 0) 1458 return NULL; 1459 1460 /* if the parameter is from icmp6, decode it. */ 1461 if (d != NULL) { 1462 struct ip6ctlparam *ip6cp = (struct ip6ctlparam *)d; 1463 m = ip6cp->ip6c_m; 1464 ip6 = ip6cp->ip6c_ip6; 1465 off = ip6cp->ip6c_off; 1466 sa6_src = ip6cp->ip6c_src; 1467 } else { 1468 m = NULL; 1469 ip6 = NULL; 1470 sa6_src = &sa6_any; 1471 off = 0; 1472 } 1473 1474 if (ip6) { 1475 /* 1476 * XXX: We assume that when ip6 is non NULL, 1477 * M and OFF are valid. 1478 */ 1479 1480 /* check if we can safely examine src and dst ports */ 1481 if (m->m_pkthdr.len < off + sizeof(th)) { 1482 if (cmd == PRC_MSGSIZE) 1483 icmp6_mtudisc_update((struct ip6ctlparam *)d, 0); 1484 return NULL; 1485 } 1486 1487 memset(&th, 0, sizeof(th)); 1488 m_copydata(m, off, sizeof(th), (void *)&th); 1489 1490 if (cmd == PRC_MSGSIZE) { 1491 int valid = 0; 1492 1493 /* 1494 * Check to see if we have a valid TCP connection 1495 * corresponding to the address in the ICMPv6 message 1496 * payload. 1497 */ 1498 if (in6_pcblookup_connect(&tcbtable, &sa6->sin6_addr, 1499 th.th_dport, 1500 (const struct in6_addr *)&sa6_src->sin6_addr, 1501 th.th_sport, 0, 0)) 1502 valid++; 1503 1504 /* 1505 * Depending on the value of "valid" and routing table 1506 * size (mtudisc_{hi,lo}wat), we will: 1507 * - recalcurate the new MTU and create the 1508 * corresponding routing entry, or 1509 * - ignore the MTU change notification. 1510 */ 1511 icmp6_mtudisc_update((struct ip6ctlparam *)d, valid); 1512 1513 /* 1514 * no need to call in6_pcbnotify, it should have been 1515 * called via callback if necessary 1516 */ 1517 return NULL; 1518 } 1519 1520 nmatch = in6_pcbnotify(&tcbtable, sa, th.th_dport, 1521 (const struct sockaddr *)sa6_src, th.th_sport, cmd, NULL, notify); 1522 if (nmatch == 0 && syn_cache_count && 1523 (inet6ctlerrmap[cmd] == EHOSTUNREACH || 1524 inet6ctlerrmap[cmd] == ENETUNREACH || 1525 inet6ctlerrmap[cmd] == EHOSTDOWN)) 1526 syn_cache_unreach((const struct sockaddr *)sa6_src, 1527 sa, &th); 1528 } else { 1529 (void) in6_pcbnotify(&tcbtable, sa, 0, 1530 (const struct sockaddr *)sa6_src, 0, cmd, NULL, notify); 1531 } 1532 1533 return NULL; 1534 } 1535 #endif 1536 1537 #ifdef INET 1538 /* assumes that ip header and tcp header are contiguous on mbuf */ 1539 void * 1540 tcp_ctlinput(int cmd, const struct sockaddr *sa, void *v) 1541 { 1542 struct ip *ip = v; 1543 struct tcphdr *th; 1544 struct icmp *icp; 1545 extern const int inetctlerrmap[]; 1546 void (*notify)(struct inpcb *, int) = tcp_notify; 1547 int errno; 1548 int nmatch; 1549 struct tcpcb *tp; 1550 u_int mtu; 1551 tcp_seq seq; 1552 struct inpcb *inp; 1553 #ifdef INET6 1554 struct in6pcb *in6p; 1555 struct in6_addr src6, dst6; 1556 #endif 1557 1558 if (sa->sa_family != AF_INET || 1559 sa->sa_len != sizeof(struct sockaddr_in)) 1560 return NULL; 1561 if ((unsigned)cmd >= PRC_NCMDS) 1562 return NULL; 1563 errno = inetctlerrmap[cmd]; 1564 if (cmd == PRC_QUENCH) 1565 /* 1566 * Don't honor ICMP Source Quench messages meant for 1567 * TCP connections. 1568 */ 1569 return NULL; 1570 else if (PRC_IS_REDIRECT(cmd)) 1571 notify = in_rtchange, ip = 0; 1572 else if (cmd == PRC_MSGSIZE && ip && ip->ip_v == 4) { 1573 /* 1574 * Check to see if we have a valid TCP connection 1575 * corresponding to the address in the ICMP message 1576 * payload. 1577 * 1578 * Boundary check is made in icmp_input(), with ICMP_ADVLENMIN. 1579 */ 1580 th = (struct tcphdr *)((char *)ip + (ip->ip_hl << 2)); 1581 #ifdef INET6 1582 memset(&src6, 0, sizeof(src6)); 1583 memset(&dst6, 0, sizeof(dst6)); 1584 src6.s6_addr16[5] = dst6.s6_addr16[5] = 0xffff; 1585 memcpy(&src6.s6_addr32[3], &ip->ip_src, sizeof(struct in_addr)); 1586 memcpy(&dst6.s6_addr32[3], &ip->ip_dst, sizeof(struct in_addr)); 1587 #endif 1588 if ((inp = in_pcblookup_connect(&tcbtable, ip->ip_dst, 1589 th->th_dport, ip->ip_src, th->th_sport, 0)) != NULL) 1590 #ifdef INET6 1591 in6p = NULL; 1592 #else 1593 ; 1594 #endif 1595 #ifdef INET6 1596 else if ((in6p = in6_pcblookup_connect(&tcbtable, &dst6, 1597 th->th_dport, &src6, th->th_sport, 0, 0)) != NULL) 1598 ; 1599 #endif 1600 else 1601 return NULL; 1602 1603 /* 1604 * Now that we've validated that we are actually communicating 1605 * with the host indicated in the ICMP message, locate the 1606 * ICMP header, recalculate the new MTU, and create the 1607 * corresponding routing entry. 1608 */ 1609 icp = (struct icmp *)((char *)ip - 1610 offsetof(struct icmp, icmp_ip)); 1611 if (inp) { 1612 if ((tp = intotcpcb(inp)) == NULL) 1613 return NULL; 1614 } 1615 #ifdef INET6 1616 else if (in6p) { 1617 if ((tp = in6totcpcb(in6p)) == NULL) 1618 return NULL; 1619 } 1620 #endif 1621 else 1622 return NULL; 1623 seq = ntohl(th->th_seq); 1624 if (SEQ_LT(seq, tp->snd_una) || SEQ_GT(seq, tp->snd_max)) 1625 return NULL; 1626 /* 1627 * If the ICMP message advertises a Next-Hop MTU 1628 * equal or larger than the maximum packet size we have 1629 * ever sent, drop the message. 1630 */ 1631 mtu = (u_int)ntohs(icp->icmp_nextmtu); 1632 if (mtu >= tp->t_pmtud_mtu_sent) 1633 return NULL; 1634 if (mtu >= tcp_hdrsz(tp) + tp->t_pmtud_mss_acked) { 1635 /* 1636 * Calculate new MTU, and create corresponding 1637 * route (traditional PMTUD). 1638 */ 1639 tp->t_flags &= ~TF_PMTUD_PEND; 1640 icmp_mtudisc(icp, ip->ip_dst); 1641 } else { 1642 /* 1643 * Record the information got in the ICMP 1644 * message; act on it later. 1645 * If we had already recorded an ICMP message, 1646 * replace the old one only if the new message 1647 * refers to an older TCP segment 1648 */ 1649 if (tp->t_flags & TF_PMTUD_PEND) { 1650 if (SEQ_LT(tp->t_pmtud_th_seq, seq)) 1651 return NULL; 1652 } else 1653 tp->t_flags |= TF_PMTUD_PEND; 1654 tp->t_pmtud_th_seq = seq; 1655 tp->t_pmtud_nextmtu = icp->icmp_nextmtu; 1656 tp->t_pmtud_ip_len = icp->icmp_ip.ip_len; 1657 tp->t_pmtud_ip_hl = icp->icmp_ip.ip_hl; 1658 } 1659 return NULL; 1660 } else if (cmd == PRC_HOSTDEAD) 1661 ip = 0; 1662 else if (errno == 0) 1663 return NULL; 1664 if (ip && ip->ip_v == 4 && sa->sa_family == AF_INET) { 1665 th = (struct tcphdr *)((char *)ip + (ip->ip_hl << 2)); 1666 nmatch = in_pcbnotify(&tcbtable, satocsin(sa)->sin_addr, 1667 th->th_dport, ip->ip_src, th->th_sport, errno, notify); 1668 if (nmatch == 0 && syn_cache_count && 1669 (inetctlerrmap[cmd] == EHOSTUNREACH || 1670 inetctlerrmap[cmd] == ENETUNREACH || 1671 inetctlerrmap[cmd] == EHOSTDOWN)) { 1672 struct sockaddr_in sin; 1673 memset(&sin, 0, sizeof(sin)); 1674 sin.sin_len = sizeof(sin); 1675 sin.sin_family = AF_INET; 1676 sin.sin_port = th->th_sport; 1677 sin.sin_addr = ip->ip_src; 1678 syn_cache_unreach((struct sockaddr *)&sin, sa, th); 1679 } 1680 1681 /* XXX mapped address case */ 1682 } else 1683 in_pcbnotifyall(&tcbtable, satocsin(sa)->sin_addr, errno, 1684 notify); 1685 return NULL; 1686 } 1687 1688 /* 1689 * When a source quench is received, we are being notified of congestion. 1690 * Close the congestion window down to the Loss Window (one segment). 1691 * We will gradually open it again as we proceed. 1692 */ 1693 void 1694 tcp_quench(struct inpcb *inp, int errno) 1695 { 1696 struct tcpcb *tp = intotcpcb(inp); 1697 1698 if (tp) { 1699 tp->snd_cwnd = tp->t_segsz; 1700 tp->t_bytes_acked = 0; 1701 } 1702 } 1703 #endif 1704 1705 #ifdef INET6 1706 void 1707 tcp6_quench(struct in6pcb *in6p, int errno) 1708 { 1709 struct tcpcb *tp = in6totcpcb(in6p); 1710 1711 if (tp) { 1712 tp->snd_cwnd = tp->t_segsz; 1713 tp->t_bytes_acked = 0; 1714 } 1715 } 1716 #endif 1717 1718 #ifdef INET 1719 /* 1720 * Path MTU Discovery handlers. 1721 */ 1722 void 1723 tcp_mtudisc_callback(struct in_addr faddr) 1724 { 1725 #ifdef INET6 1726 struct in6_addr in6; 1727 #endif 1728 1729 in_pcbnotifyall(&tcbtable, faddr, EMSGSIZE, tcp_mtudisc); 1730 #ifdef INET6 1731 memset(&in6, 0, sizeof(in6)); 1732 in6.s6_addr16[5] = 0xffff; 1733 memcpy(&in6.s6_addr32[3], &faddr, sizeof(struct in_addr)); 1734 tcp6_mtudisc_callback(&in6); 1735 #endif 1736 } 1737 1738 /* 1739 * On receipt of path MTU corrections, flush old route and replace it 1740 * with the new one. Retransmit all unacknowledged packets, to ensure 1741 * that all packets will be received. 1742 */ 1743 void 1744 tcp_mtudisc(struct inpcb *inp, int errno) 1745 { 1746 struct tcpcb *tp = intotcpcb(inp); 1747 struct rtentry *rt = in_pcbrtentry(inp); 1748 1749 if (tp != 0) { 1750 if (rt != 0) { 1751 /* 1752 * If this was not a host route, remove and realloc. 1753 */ 1754 if ((rt->rt_flags & RTF_HOST) == 0) { 1755 in_rtchange(inp, errno); 1756 if ((rt = in_pcbrtentry(inp)) == 0) 1757 return; 1758 } 1759 1760 /* 1761 * Slow start out of the error condition. We 1762 * use the MTU because we know it's smaller 1763 * than the previously transmitted segment. 1764 * 1765 * Note: This is more conservative than the 1766 * suggestion in draft-floyd-incr-init-win-03. 1767 */ 1768 if (rt->rt_rmx.rmx_mtu != 0) 1769 tp->snd_cwnd = 1770 TCP_INITIAL_WINDOW(tcp_init_win, 1771 rt->rt_rmx.rmx_mtu); 1772 } 1773 1774 /* 1775 * Resend unacknowledged packets. 1776 */ 1777 tp->snd_nxt = tp->sack_newdata = tp->snd_una; 1778 tcp_output(tp); 1779 } 1780 } 1781 #endif 1782 1783 #ifdef INET6 1784 /* 1785 * Path MTU Discovery handlers. 1786 */ 1787 void 1788 tcp6_mtudisc_callback(struct in6_addr *faddr) 1789 { 1790 struct sockaddr_in6 sin6; 1791 1792 memset(&sin6, 0, sizeof(sin6)); 1793 sin6.sin6_family = AF_INET6; 1794 sin6.sin6_len = sizeof(struct sockaddr_in6); 1795 sin6.sin6_addr = *faddr; 1796 (void) in6_pcbnotify(&tcbtable, (struct sockaddr *)&sin6, 0, 1797 (const struct sockaddr *)&sa6_any, 0, PRC_MSGSIZE, NULL, tcp6_mtudisc); 1798 } 1799 1800 void 1801 tcp6_mtudisc(struct in6pcb *in6p, int errno) 1802 { 1803 struct tcpcb *tp = in6totcpcb(in6p); 1804 struct rtentry *rt = in6_pcbrtentry(in6p); 1805 1806 if (tp != 0) { 1807 if (rt != 0) { 1808 /* 1809 * If this was not a host route, remove and realloc. 1810 */ 1811 if ((rt->rt_flags & RTF_HOST) == 0) { 1812 in6_rtchange(in6p, errno); 1813 if ((rt = in6_pcbrtentry(in6p)) == 0) 1814 return; 1815 } 1816 1817 /* 1818 * Slow start out of the error condition. We 1819 * use the MTU because we know it's smaller 1820 * than the previously transmitted segment. 1821 * 1822 * Note: This is more conservative than the 1823 * suggestion in draft-floyd-incr-init-win-03. 1824 */ 1825 if (rt->rt_rmx.rmx_mtu != 0) 1826 tp->snd_cwnd = 1827 TCP_INITIAL_WINDOW(tcp_init_win, 1828 rt->rt_rmx.rmx_mtu); 1829 } 1830 1831 /* 1832 * Resend unacknowledged packets. 1833 */ 1834 tp->snd_nxt = tp->sack_newdata = tp->snd_una; 1835 tcp_output(tp); 1836 } 1837 } 1838 #endif /* INET6 */ 1839 1840 /* 1841 * Compute the MSS to advertise to the peer. Called only during 1842 * the 3-way handshake. If we are the server (peer initiated 1843 * connection), we are called with a pointer to the interface 1844 * on which the SYN packet arrived. If we are the client (we 1845 * initiated connection), we are called with a pointer to the 1846 * interface out which this connection should go. 1847 * 1848 * NOTE: Do not subtract IP option/extension header size nor IPsec 1849 * header size from MSS advertisement. MSS option must hold the maximum 1850 * segment size we can accept, so it must always be: 1851 * max(if mtu) - ip header - tcp header 1852 */ 1853 u_long 1854 tcp_mss_to_advertise(const struct ifnet *ifp, int af) 1855 { 1856 extern u_long in_maxmtu; 1857 u_long mss = 0; 1858 u_long hdrsiz; 1859 1860 /* 1861 * In order to avoid defeating path MTU discovery on the peer, 1862 * we advertise the max MTU of all attached networks as our MSS, 1863 * per RFC 1191, section 3.1. 1864 * 1865 * We provide the option to advertise just the MTU of 1866 * the interface on which we hope this connection will 1867 * be receiving. If we are responding to a SYN, we 1868 * will have a pretty good idea about this, but when 1869 * initiating a connection there is a bit more doubt. 1870 * 1871 * We also need to ensure that loopback has a large enough 1872 * MSS, as the loopback MTU is never included in in_maxmtu. 1873 */ 1874 1875 if (ifp != NULL) 1876 switch (af) { 1877 case AF_INET: 1878 mss = ifp->if_mtu; 1879 break; 1880 #ifdef INET6 1881 case AF_INET6: 1882 mss = IN6_LINKMTU(ifp); 1883 break; 1884 #endif 1885 } 1886 1887 if (tcp_mss_ifmtu == 0) 1888 switch (af) { 1889 case AF_INET: 1890 mss = max(in_maxmtu, mss); 1891 break; 1892 #ifdef INET6 1893 case AF_INET6: 1894 mss = max(in6_maxmtu, mss); 1895 break; 1896 #endif 1897 } 1898 1899 switch (af) { 1900 case AF_INET: 1901 hdrsiz = sizeof(struct ip); 1902 break; 1903 #ifdef INET6 1904 case AF_INET6: 1905 hdrsiz = sizeof(struct ip6_hdr); 1906 break; 1907 #endif 1908 default: 1909 hdrsiz = 0; 1910 break; 1911 } 1912 hdrsiz += sizeof(struct tcphdr); 1913 if (mss > hdrsiz) 1914 mss -= hdrsiz; 1915 1916 mss = max(tcp_mssdflt, mss); 1917 return (mss); 1918 } 1919 1920 /* 1921 * Set connection variables based on the peer's advertised MSS. 1922 * We are passed the TCPCB for the actual connection. If we 1923 * are the server, we are called by the compressed state engine 1924 * when the 3-way handshake is complete. If we are the client, 1925 * we are called when we receive the SYN,ACK from the server. 1926 * 1927 * NOTE: Our advertised MSS value must be initialized in the TCPCB 1928 * before this routine is called! 1929 */ 1930 void 1931 tcp_mss_from_peer(struct tcpcb *tp, int offer) 1932 { 1933 struct socket *so; 1934 #if defined(RTV_SPIPE) || defined(RTV_SSTHRESH) 1935 struct rtentry *rt; 1936 #endif 1937 u_long bufsize; 1938 int mss; 1939 1940 #ifdef DIAGNOSTIC 1941 if (tp->t_inpcb && tp->t_in6pcb) 1942 panic("tcp_mss_from_peer: both t_inpcb and t_in6pcb are set"); 1943 #endif 1944 so = NULL; 1945 rt = NULL; 1946 #ifdef INET 1947 if (tp->t_inpcb) { 1948 so = tp->t_inpcb->inp_socket; 1949 #if defined(RTV_SPIPE) || defined(RTV_SSTHRESH) 1950 rt = in_pcbrtentry(tp->t_inpcb); 1951 #endif 1952 } 1953 #endif 1954 #ifdef INET6 1955 if (tp->t_in6pcb) { 1956 so = tp->t_in6pcb->in6p_socket; 1957 #if defined(RTV_SPIPE) || defined(RTV_SSTHRESH) 1958 rt = in6_pcbrtentry(tp->t_in6pcb); 1959 #endif 1960 } 1961 #endif 1962 1963 /* 1964 * As per RFC1122, use the default MSS value, unless they 1965 * sent us an offer. Do not accept offers less than 256 bytes. 1966 */ 1967 mss = tcp_mssdflt; 1968 if (offer) 1969 mss = offer; 1970 mss = max(mss, 256); /* sanity */ 1971 tp->t_peermss = mss; 1972 mss -= tcp_optlen(tp); 1973 #ifdef INET 1974 if (tp->t_inpcb) 1975 mss -= ip_optlen(tp->t_inpcb); 1976 #endif 1977 #ifdef INET6 1978 if (tp->t_in6pcb) 1979 mss -= ip6_optlen(tp->t_in6pcb); 1980 #endif 1981 1982 /* 1983 * If there's a pipesize, change the socket buffer to that size. 1984 * Make the socket buffer an integral number of MSS units. If 1985 * the MSS is larger than the socket buffer, artificially decrease 1986 * the MSS. 1987 */ 1988 #ifdef RTV_SPIPE 1989 if (rt != NULL && rt->rt_rmx.rmx_sendpipe != 0) 1990 bufsize = rt->rt_rmx.rmx_sendpipe; 1991 else 1992 #endif 1993 { 1994 KASSERT(so != NULL); 1995 bufsize = so->so_snd.sb_hiwat; 1996 } 1997 if (bufsize < mss) 1998 mss = bufsize; 1999 else { 2000 bufsize = roundup(bufsize, mss); 2001 if (bufsize > sb_max) 2002 bufsize = sb_max; 2003 (void) sbreserve(&so->so_snd, bufsize, so); 2004 } 2005 tp->t_segsz = mss; 2006 2007 #ifdef RTV_SSTHRESH 2008 if (rt != NULL && rt->rt_rmx.rmx_ssthresh) { 2009 /* 2010 * There's some sort of gateway or interface buffer 2011 * limit on the path. Use this to set the slow 2012 * start threshold, but set the threshold to no less 2013 * than 2 * MSS. 2014 */ 2015 tp->snd_ssthresh = max(2 * mss, rt->rt_rmx.rmx_ssthresh); 2016 } 2017 #endif 2018 } 2019 2020 /* 2021 * Processing necessary when a TCP connection is established. 2022 */ 2023 void 2024 tcp_established(struct tcpcb *tp) 2025 { 2026 struct socket *so; 2027 #ifdef RTV_RPIPE 2028 struct rtentry *rt; 2029 #endif 2030 u_long bufsize; 2031 2032 #ifdef DIAGNOSTIC 2033 if (tp->t_inpcb && tp->t_in6pcb) 2034 panic("tcp_established: both t_inpcb and t_in6pcb are set"); 2035 #endif 2036 so = NULL; 2037 rt = NULL; 2038 #ifdef INET 2039 /* This is a while() to reduce the dreadful stairstepping below */ 2040 while (tp->t_inpcb) { 2041 so = tp->t_inpcb->inp_socket; 2042 #if defined(RTV_RPIPE) 2043 rt = in_pcbrtentry(tp->t_inpcb); 2044 #endif 2045 if (__predict_true(tcp_msl_enable)) { 2046 if (tp->t_inpcb->inp_laddr.s_addr == INADDR_LOOPBACK) { 2047 tp->t_msl = tcp_msl_loop ? tcp_msl_loop : (TCPTV_MSL >> 2); 2048 break; 2049 } 2050 2051 if (__predict_false(tcp_rttlocal)) { 2052 /* This may be adjusted by tcp_input */ 2053 tp->t_msl = tcp_msl_local ? tcp_msl_local : (TCPTV_MSL >> 1); 2054 break; 2055 } 2056 if (in_localaddr(tp->t_inpcb->inp_faddr)) { 2057 tp->t_msl = tcp_msl_local ? tcp_msl_local : (TCPTV_MSL >> 1); 2058 break; 2059 } 2060 } 2061 tp->t_msl = tcp_msl_remote ? tcp_msl_remote : TCPTV_MSL; 2062 break; 2063 } 2064 #endif 2065 #ifdef INET6 2066 /* The !tp->t_inpcb lets the compiler know it can't be v4 *and* v6 */ 2067 while (!tp->t_inpcb && tp->t_in6pcb) { 2068 so = tp->t_in6pcb->in6p_socket; 2069 #if defined(RTV_RPIPE) 2070 rt = in6_pcbrtentry(tp->t_in6pcb); 2071 #endif 2072 if (__predict_true(tcp_msl_enable)) { 2073 extern const struct in6_addr in6addr_loopback; 2074 2075 if (IN6_ARE_ADDR_EQUAL(&tp->t_in6pcb->in6p_laddr, 2076 &in6addr_loopback)) { 2077 tp->t_msl = tcp_msl_loop ? tcp_msl_loop : (TCPTV_MSL >> 2); 2078 break; 2079 } 2080 2081 if (__predict_false(tcp_rttlocal)) { 2082 /* This may be adjusted by tcp_input */ 2083 tp->t_msl = tcp_msl_local ? tcp_msl_local : (TCPTV_MSL >> 1); 2084 break; 2085 } 2086 if (in6_localaddr(&tp->t_in6pcb->in6p_faddr)) { 2087 tp->t_msl = tcp_msl_local ? tcp_msl_local : (TCPTV_MSL >> 1); 2088 break; 2089 } 2090 } 2091 tp->t_msl = tcp_msl_remote ? tcp_msl_remote : TCPTV_MSL; 2092 break; 2093 } 2094 #endif 2095 2096 tp->t_state = TCPS_ESTABLISHED; 2097 TCP_TIMER_ARM(tp, TCPT_KEEP, tp->t_keepidle); 2098 2099 #ifdef RTV_RPIPE 2100 if (rt != NULL && rt->rt_rmx.rmx_recvpipe != 0) 2101 bufsize = rt->rt_rmx.rmx_recvpipe; 2102 else 2103 #endif 2104 { 2105 KASSERT(so != NULL); 2106 bufsize = so->so_rcv.sb_hiwat; 2107 } 2108 if (bufsize > tp->t_ourmss) { 2109 bufsize = roundup(bufsize, tp->t_ourmss); 2110 if (bufsize > sb_max) 2111 bufsize = sb_max; 2112 (void) sbreserve(&so->so_rcv, bufsize, so); 2113 } 2114 } 2115 2116 /* 2117 * Check if there's an initial rtt or rttvar. Convert from the 2118 * route-table units to scaled multiples of the slow timeout timer. 2119 * Called only during the 3-way handshake. 2120 */ 2121 void 2122 tcp_rmx_rtt(struct tcpcb *tp) 2123 { 2124 #ifdef RTV_RTT 2125 struct rtentry *rt = NULL; 2126 int rtt; 2127 2128 #ifdef DIAGNOSTIC 2129 if (tp->t_inpcb && tp->t_in6pcb) 2130 panic("tcp_rmx_rtt: both t_inpcb and t_in6pcb are set"); 2131 #endif 2132 #ifdef INET 2133 if (tp->t_inpcb) 2134 rt = in_pcbrtentry(tp->t_inpcb); 2135 #endif 2136 #ifdef INET6 2137 if (tp->t_in6pcb) 2138 rt = in6_pcbrtentry(tp->t_in6pcb); 2139 #endif 2140 if (rt == NULL) 2141 return; 2142 2143 if (tp->t_srtt == 0 && (rtt = rt->rt_rmx.rmx_rtt)) { 2144 /* 2145 * XXX The lock bit for MTU indicates that the value 2146 * is also a minimum value; this is subject to time. 2147 */ 2148 if (rt->rt_rmx.rmx_locks & RTV_RTT) 2149 TCPT_RANGESET(tp->t_rttmin, 2150 rtt / (RTM_RTTUNIT / PR_SLOWHZ), 2151 TCPTV_MIN, TCPTV_REXMTMAX); 2152 tp->t_srtt = rtt / 2153 ((RTM_RTTUNIT / PR_SLOWHZ) >> (TCP_RTT_SHIFT + 2)); 2154 if (rt->rt_rmx.rmx_rttvar) { 2155 tp->t_rttvar = rt->rt_rmx.rmx_rttvar / 2156 ((RTM_RTTUNIT / PR_SLOWHZ) >> 2157 (TCP_RTTVAR_SHIFT + 2)); 2158 } else { 2159 /* Default variation is +- 1 rtt */ 2160 tp->t_rttvar = 2161 tp->t_srtt >> (TCP_RTT_SHIFT - TCP_RTTVAR_SHIFT); 2162 } 2163 TCPT_RANGESET(tp->t_rxtcur, 2164 ((tp->t_srtt >> 2) + tp->t_rttvar) >> (1 + 2), 2165 tp->t_rttmin, TCPTV_REXMTMAX); 2166 } 2167 #endif 2168 } 2169 2170 tcp_seq tcp_iss_seq = 0; /* tcp initial seq # */ 2171 u_int8_t tcp_iss_secret[16]; /* 128 bits; should be plenty */ 2172 2173 /* 2174 * Get a new sequence value given a tcp control block 2175 */ 2176 tcp_seq 2177 tcp_new_iss(struct tcpcb *tp, tcp_seq addin) 2178 { 2179 2180 #ifdef INET 2181 if (tp->t_inpcb != NULL) { 2182 return (tcp_new_iss1(&tp->t_inpcb->inp_laddr, 2183 &tp->t_inpcb->inp_faddr, tp->t_inpcb->inp_lport, 2184 tp->t_inpcb->inp_fport, sizeof(tp->t_inpcb->inp_laddr), 2185 addin)); 2186 } 2187 #endif 2188 #ifdef INET6 2189 if (tp->t_in6pcb != NULL) { 2190 return (tcp_new_iss1(&tp->t_in6pcb->in6p_laddr, 2191 &tp->t_in6pcb->in6p_faddr, tp->t_in6pcb->in6p_lport, 2192 tp->t_in6pcb->in6p_fport, sizeof(tp->t_in6pcb->in6p_laddr), 2193 addin)); 2194 } 2195 #endif 2196 /* Not possible. */ 2197 panic("tcp_new_iss"); 2198 } 2199 2200 /* 2201 * This routine actually generates a new TCP initial sequence number. 2202 */ 2203 tcp_seq 2204 tcp_new_iss1(void *laddr, void *faddr, u_int16_t lport, u_int16_t fport, 2205 size_t addrsz, tcp_seq addin) 2206 { 2207 tcp_seq tcp_iss; 2208 2209 static bool tcp_iss_gotten_secret; 2210 2211 /* 2212 * If we haven't been here before, initialize our cryptographic 2213 * hash secret. 2214 */ 2215 if (tcp_iss_gotten_secret == false) { 2216 cprng_strong(kern_cprng, 2217 tcp_iss_secret, sizeof(tcp_iss_secret), FASYNC); 2218 tcp_iss_gotten_secret = true; 2219 } 2220 2221 if (tcp_do_rfc1948) { 2222 MD5_CTX ctx; 2223 u_int8_t hash[16]; /* XXX MD5 knowledge */ 2224 2225 /* 2226 * Compute the base value of the ISS. It is a hash 2227 * of (saddr, sport, daddr, dport, secret). 2228 */ 2229 MD5Init(&ctx); 2230 2231 MD5Update(&ctx, (u_char *) laddr, addrsz); 2232 MD5Update(&ctx, (u_char *) &lport, sizeof(lport)); 2233 2234 MD5Update(&ctx, (u_char *) faddr, addrsz); 2235 MD5Update(&ctx, (u_char *) &fport, sizeof(fport)); 2236 2237 MD5Update(&ctx, tcp_iss_secret, sizeof(tcp_iss_secret)); 2238 2239 MD5Final(hash, &ctx); 2240 2241 memcpy(&tcp_iss, hash, sizeof(tcp_iss)); 2242 2243 /* 2244 * Now increment our "timer", and add it in to 2245 * the computed value. 2246 * 2247 * XXX Use `addin'? 2248 * XXX TCP_ISSINCR too large to use? 2249 */ 2250 tcp_iss_seq += TCP_ISSINCR; 2251 #ifdef TCPISS_DEBUG 2252 printf("ISS hash 0x%08x, ", tcp_iss); 2253 #endif 2254 tcp_iss += tcp_iss_seq + addin; 2255 #ifdef TCPISS_DEBUG 2256 printf("new ISS 0x%08x\n", tcp_iss); 2257 #endif 2258 } else { 2259 /* 2260 * Randomize. 2261 */ 2262 tcp_iss = cprng_fast32(); 2263 2264 /* 2265 * If we were asked to add some amount to a known value, 2266 * we will take a random value obtained above, mask off 2267 * the upper bits, and add in the known value. We also 2268 * add in a constant to ensure that we are at least a 2269 * certain distance from the original value. 2270 * 2271 * This is used when an old connection is in timed wait 2272 * and we have a new one coming in, for instance. 2273 */ 2274 if (addin != 0) { 2275 #ifdef TCPISS_DEBUG 2276 printf("Random %08x, ", tcp_iss); 2277 #endif 2278 tcp_iss &= TCP_ISS_RANDOM_MASK; 2279 tcp_iss += addin + TCP_ISSINCR; 2280 #ifdef TCPISS_DEBUG 2281 printf("Old ISS %08x, ISS %08x\n", addin, tcp_iss); 2282 #endif 2283 } else { 2284 tcp_iss &= TCP_ISS_RANDOM_MASK; 2285 tcp_iss += tcp_iss_seq; 2286 tcp_iss_seq += TCP_ISSINCR; 2287 #ifdef TCPISS_DEBUG 2288 printf("ISS %08x\n", tcp_iss); 2289 #endif 2290 } 2291 } 2292 2293 if (tcp_compat_42) { 2294 /* 2295 * Limit it to the positive range for really old TCP 2296 * implementations. 2297 * Just AND off the top bit instead of checking if 2298 * is set first - saves a branch 50% of the time. 2299 */ 2300 tcp_iss &= 0x7fffffff; /* XXX */ 2301 } 2302 2303 return (tcp_iss); 2304 } 2305 2306 #if defined(FAST_IPSEC) 2307 /* compute ESP/AH header size for TCP, including outer IP header. */ 2308 size_t 2309 ipsec4_hdrsiz_tcp(struct tcpcb *tp) 2310 { 2311 struct inpcb *inp; 2312 size_t hdrsiz; 2313 2314 /* XXX mapped addr case (tp->t_in6pcb) */ 2315 if (!tp || !tp->t_template || !(inp = tp->t_inpcb)) 2316 return 0; 2317 switch (tp->t_family) { 2318 case AF_INET: 2319 /* XXX: should use currect direction. */ 2320 hdrsiz = ipsec4_hdrsiz(tp->t_template, IPSEC_DIR_OUTBOUND, inp); 2321 break; 2322 default: 2323 hdrsiz = 0; 2324 break; 2325 } 2326 2327 return hdrsiz; 2328 } 2329 2330 #ifdef INET6 2331 size_t 2332 ipsec6_hdrsiz_tcp(struct tcpcb *tp) 2333 { 2334 struct in6pcb *in6p; 2335 size_t hdrsiz; 2336 2337 if (!tp || !tp->t_template || !(in6p = tp->t_in6pcb)) 2338 return 0; 2339 switch (tp->t_family) { 2340 case AF_INET6: 2341 /* XXX: should use currect direction. */ 2342 hdrsiz = ipsec6_hdrsiz(tp->t_template, IPSEC_DIR_OUTBOUND, in6p); 2343 break; 2344 case AF_INET: 2345 /* mapped address case - tricky */ 2346 default: 2347 hdrsiz = 0; 2348 break; 2349 } 2350 2351 return hdrsiz; 2352 } 2353 #endif 2354 #endif /*IPSEC*/ 2355 2356 /* 2357 * Determine the length of the TCP options for this connection. 2358 * 2359 * XXX: What do we do for SACK, when we add that? Just reserve 2360 * all of the space? Otherwise we can't exactly be incrementing 2361 * cwnd by an amount that varies depending on the amount we last 2362 * had to SACK! 2363 */ 2364 2365 u_int 2366 tcp_optlen(struct tcpcb *tp) 2367 { 2368 u_int optlen; 2369 2370 optlen = 0; 2371 if ((tp->t_flags & (TF_REQ_TSTMP|TF_RCVD_TSTMP|TF_NOOPT)) == 2372 (TF_REQ_TSTMP | TF_RCVD_TSTMP)) 2373 optlen += TCPOLEN_TSTAMP_APPA; 2374 2375 #ifdef TCP_SIGNATURE 2376 if (tp->t_flags & TF_SIGNATURE) 2377 optlen += TCPOLEN_SIGNATURE + 2; 2378 #endif /* TCP_SIGNATURE */ 2379 2380 return optlen; 2381 } 2382 2383 u_int 2384 tcp_hdrsz(struct tcpcb *tp) 2385 { 2386 u_int hlen; 2387 2388 switch (tp->t_family) { 2389 #ifdef INET6 2390 case AF_INET6: 2391 hlen = sizeof(struct ip6_hdr); 2392 break; 2393 #endif 2394 case AF_INET: 2395 hlen = sizeof(struct ip); 2396 break; 2397 default: 2398 hlen = 0; 2399 break; 2400 } 2401 hlen += sizeof(struct tcphdr); 2402 2403 if ((tp->t_flags & (TF_REQ_TSTMP|TF_NOOPT)) == TF_REQ_TSTMP && 2404 (tp->t_flags & TF_RCVD_TSTMP) == TF_RCVD_TSTMP) 2405 hlen += TCPOLEN_TSTAMP_APPA; 2406 #ifdef TCP_SIGNATURE 2407 if (tp->t_flags & TF_SIGNATURE) 2408 hlen += TCPOLEN_SIGLEN; 2409 #endif 2410 return hlen; 2411 } 2412 2413 void 2414 tcp_statinc(u_int stat) 2415 { 2416 2417 KASSERT(stat < TCP_NSTATS); 2418 TCP_STATINC(stat); 2419 } 2420 2421 void 2422 tcp_statadd(u_int stat, uint64_t val) 2423 { 2424 2425 KASSERT(stat < TCP_NSTATS); 2426 TCP_STATADD(stat, val); 2427 } 2428