ip_input.c revision 1.106
1/* $NetBSD: ip_input.c,v 1.106 2000/03/07 05:39:57 thorpej 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) 1998 The NetBSD Foundation, Inc. 34 * All rights reserved. 35 * 36 * This code is derived from software contributed to The NetBSD Foundation 37 * by Public Access Networks Corporation ("Panix"). It was developed under 38 * contract to Panix by Eric Haszlakiewicz and Thor Lancelot Simon. 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, 1993 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 * @(#)ip_input.c 8.2 (Berkeley) 1/4/94 102 */ 103 104#include "opt_gateway.h" 105#include "opt_pfil_hooks.h" 106#include "opt_ipsec.h" 107#include "opt_mrouting.h" 108 109#include <sys/param.h> 110#include <sys/systm.h> 111#include <sys/malloc.h> 112#include <sys/mbuf.h> 113#include <sys/domain.h> 114#include <sys/protosw.h> 115#include <sys/socket.h> 116#include <sys/socketvar.h> 117#include <sys/errno.h> 118#include <sys/time.h> 119#include <sys/kernel.h> 120#include <sys/proc.h> 121#include <sys/pool.h> 122 123#include <vm/vm.h> 124#include <sys/sysctl.h> 125 126#include <net/if.h> 127#include <net/if_dl.h> 128#include <net/route.h> 129#include <net/pfil.h> 130 131#include <netinet/in.h> 132#include <netinet/in_systm.h> 133#include <netinet/ip.h> 134#include <netinet/in_pcb.h> 135#include <netinet/in_var.h> 136#include <netinet/ip_var.h> 137#include <netinet/ip_icmp.h> 138/* just for gif_ttl */ 139#include <netinet/in_gif.h> 140#include "gif.h" 141 142#ifdef IPSEC 143#include <netinet6/ipsec.h> 144#include <netkey/key.h> 145#include <netkey/key_debug.h> 146#endif 147 148#ifndef IPFORWARDING 149#ifdef GATEWAY 150#define IPFORWARDING 1 /* forward IP packets not for us */ 151#else /* GATEWAY */ 152#define IPFORWARDING 0 /* don't forward IP packets not for us */ 153#endif /* GATEWAY */ 154#endif /* IPFORWARDING */ 155#ifndef IPSENDREDIRECTS 156#define IPSENDREDIRECTS 1 157#endif 158#ifndef IPFORWSRCRT 159#define IPFORWSRCRT 1 /* forward source-routed packets */ 160#endif 161#ifndef IPALLOWSRCRT 162#define IPALLOWSRCRT 1 /* allow source-routed packets */ 163#endif 164#ifndef IPMTUDISC 165#define IPMTUDISC 0 166#endif 167#ifndef IPMTUDISCTIMEOUT 168#define IPMTUDISCTIMEOUT (10 * 60) /* as per RFC 1191 */ 169#endif 170 171/* 172 * Note: DIRECTED_BROADCAST is handled this way so that previous 173 * configuration using this option will Just Work. 174 */ 175#ifndef IPDIRECTEDBCAST 176#ifdef DIRECTED_BROADCAST 177#define IPDIRECTEDBCAST 1 178#else 179#define IPDIRECTEDBCAST 0 180#endif /* DIRECTED_BROADCAST */ 181#endif /* IPDIRECTEDBCAST */ 182int ipforwarding = IPFORWARDING; 183int ipsendredirects = IPSENDREDIRECTS; 184int ip_defttl = IPDEFTTL; 185int ip_forwsrcrt = IPFORWSRCRT; 186int ip_directedbcast = IPDIRECTEDBCAST; 187int ip_allowsrcrt = IPALLOWSRCRT; 188int ip_mtudisc = IPMTUDISC; 189u_int ip_mtudisc_timeout = IPMTUDISCTIMEOUT; 190#ifdef DIAGNOSTIC 191int ipprintfs = 0; 192#endif 193 194struct rttimer_queue *ip_mtudisc_timeout_q = NULL; 195 196extern struct domain inetdomain; 197int ipqmaxlen = IFQ_MAXLEN; 198struct in_ifaddrhead in_ifaddr; 199struct in_ifaddrhashhead *in_ifaddrhashtbl; 200struct ifqueue ipintrq; 201struct ipstat ipstat; 202u_int16_t ip_id; 203int ip_defttl; 204 205struct ipqhead ipq; 206int ipq_locked; 207 208static __inline int ipq_lock_try __P((void)); 209static __inline void ipq_unlock __P((void)); 210 211static __inline int 212ipq_lock_try() 213{ 214 int s; 215 216 s = splimp(); 217 if (ipq_locked) { 218 splx(s); 219 return (0); 220 } 221 ipq_locked = 1; 222 splx(s); 223 return (1); 224} 225 226static __inline void 227ipq_unlock() 228{ 229 int s; 230 231 s = splimp(); 232 ipq_locked = 0; 233 splx(s); 234} 235 236#ifdef DIAGNOSTIC 237#define IPQ_LOCK() \ 238do { \ 239 if (ipq_lock_try() == 0) { \ 240 printf("%s:%d: ipq already locked\n", __FILE__, __LINE__); \ 241 panic("ipq_lock"); \ 242 } \ 243} while (0) 244#define IPQ_LOCK_CHECK() \ 245do { \ 246 if (ipq_locked == 0) { \ 247 printf("%s:%d: ipq lock not held\n", __FILE__, __LINE__); \ 248 panic("ipq lock check"); \ 249 } \ 250} while (0) 251#else 252#define IPQ_LOCK() (void) ipq_lock_try() 253#define IPQ_LOCK_CHECK() /* nothing */ 254#endif 255 256#define IPQ_UNLOCK() ipq_unlock() 257 258struct pool ipqent_pool; 259 260/* 261 * We need to save the IP options in case a protocol wants to respond 262 * to an incoming packet over the same route if the packet got here 263 * using IP source routing. This allows connection establishment and 264 * maintenance when the remote end is on a network that is not known 265 * to us. 266 */ 267int ip_nhops = 0; 268static struct ip_srcrt { 269 struct in_addr dst; /* final destination */ 270 char nop; /* one NOP to align */ 271 char srcopt[IPOPT_OFFSET + 1]; /* OPTVAL, OLEN and OFFSET */ 272 struct in_addr route[MAX_IPOPTLEN/sizeof(struct in_addr)]; 273} ip_srcrt; 274 275static void save_rte __P((u_char *, struct in_addr)); 276 277/* 278 * IP initialization: fill in IP protocol switch table. 279 * All protocols not implemented in kernel go to raw IP protocol handler. 280 */ 281void 282ip_init() 283{ 284 register struct protosw *pr; 285 register int i; 286 287 pool_init(&ipqent_pool, sizeof(struct ipqent), 0, 0, 0, "ipqepl", 288 0, NULL, NULL, M_IPQ); 289 290 pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW); 291 if (pr == 0) 292 panic("ip_init"); 293 for (i = 0; i < IPPROTO_MAX; i++) 294 ip_protox[i] = pr - inetsw; 295 for (pr = inetdomain.dom_protosw; 296 pr < inetdomain.dom_protoswNPROTOSW; pr++) 297 if (pr->pr_domain->dom_family == PF_INET && 298 pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW) 299 ip_protox[pr->pr_protocol] = pr - inetsw; 300 LIST_INIT(&ipq); 301 ip_id = time.tv_sec & 0xffff; 302 ipintrq.ifq_maxlen = ipqmaxlen; 303 TAILQ_INIT(&in_ifaddr); 304 in_ifaddrhashtbl = 305 hashinit(IN_IFADDR_HASH_SIZE, M_IFADDR, M_WAITOK, &in_ifaddrhash); 306 if (ip_mtudisc != 0) 307 ip_mtudisc_timeout_q = 308 rt_timer_queue_create(ip_mtudisc_timeout); 309#ifdef GATEWAY 310 ipflow_init(); 311#endif 312} 313 314struct sockaddr_in ipaddr = { sizeof(ipaddr), AF_INET }; 315struct route ipforward_rt; 316 317/* 318 * IP software interrupt routine 319 */ 320void 321ipintr() 322{ 323 int s; 324 struct mbuf *m; 325 326 while (1) { 327 s = splimp(); 328 IF_DEQUEUE(&ipintrq, m); 329 splx(s); 330 if (m == 0) 331 return; 332 ip_input(m); 333 } 334} 335 336/* 337 * Ip input routine. Checksum and byte swap header. If fragmented 338 * try to reassemble. Process options. Pass to next level. 339 */ 340void 341ip_input(struct mbuf *m) 342{ 343 register struct ip *ip = NULL; 344 register struct ipq *fp; 345 register struct in_ifaddr *ia; 346 register struct ifaddr *ifa; 347 struct ipqent *ipqe; 348 int hlen = 0, mff, len; 349 int downmatch; 350#ifdef PFIL_HOOKS 351 struct packet_filter_hook *pfh; 352 struct mbuf *m0; 353 int rv; 354#endif /* PFIL_HOOKS */ 355 356#ifdef DIAGNOSTIC 357 if ((m->m_flags & M_PKTHDR) == 0) 358 panic("ipintr no HDR"); 359#endif 360#ifdef IPSEC 361 /* 362 * should the inner packet be considered authentic? 363 * see comment in ah4_input(). 364 */ 365 if (m) { 366 m->m_flags &= ~M_AUTHIPHDR; 367 m->m_flags &= ~M_AUTHIPDGM; 368 } 369#endif 370 /* 371 * If no IP addresses have been set yet but the interfaces 372 * are receiving, can't do anything with incoming packets yet. 373 */ 374 if (in_ifaddr.tqh_first == 0) 375 goto bad; 376 ipstat.ips_total++; 377 if (m->m_len < sizeof (struct ip) && 378 (m = m_pullup(m, sizeof (struct ip))) == 0) { 379 ipstat.ips_toosmall++; 380 return; 381 } 382 ip = mtod(m, struct ip *); 383 if (ip->ip_v != IPVERSION) { 384 ipstat.ips_badvers++; 385 goto bad; 386 } 387 hlen = ip->ip_hl << 2; 388 if (hlen < sizeof(struct ip)) { /* minimum header length */ 389 ipstat.ips_badhlen++; 390 goto bad; 391 } 392 if (hlen > m->m_len) { 393 if ((m = m_pullup(m, hlen)) == 0) { 394 ipstat.ips_badhlen++; 395 return; 396 } 397 ip = mtod(m, struct ip *); 398 } 399 400 /* 401 * RFC1122: packets with a multicast source address are 402 * not allowed. 403 */ 404 if (IN_MULTICAST(ip->ip_src.s_addr)) { 405 /* XXX stat */ 406 goto bad; 407 } 408 409 if (in_cksum(m, hlen) != 0) { 410 ipstat.ips_badsum++; 411 goto bad; 412 } 413 414 /* 415 * Convert fields to host representation. 416 */ 417 NTOHS(ip->ip_len); 418 NTOHS(ip->ip_off); 419 len = ip->ip_len; 420 421 /* 422 * Check for additional length bogosity 423 */ 424 if (len < hlen) { 425 ipstat.ips_badlen++; 426 goto bad; 427 } 428 429 /* 430 * Check that the amount of data in the buffers 431 * is as at least much as the IP header would have us expect. 432 * Trim mbufs if longer than we expect. 433 * Drop packet if shorter than we expect. 434 */ 435 if (m->m_pkthdr.len < len) { 436 ipstat.ips_tooshort++; 437 goto bad; 438 } 439 if (m->m_pkthdr.len > len) { 440 if (m->m_len == m->m_pkthdr.len) { 441 m->m_len = len; 442 m->m_pkthdr.len = len; 443 } else 444 m_adj(m, len - m->m_pkthdr.len); 445 } 446 447#ifdef IPSEC 448 /* ipflow (IP fast fowarding) is not compatible with IPsec. */ 449 m->m_flags &= ~M_CANFASTFWD; 450#else 451 /* 452 * Assume that we can create a fast-forward IP flow entry 453 * based on this packet. 454 */ 455 m->m_flags |= M_CANFASTFWD; 456#endif 457 458#ifdef PFIL_HOOKS 459 /* 460 * Run through list of hooks for input packets. If there are any 461 * filters which require that additional packets in the flow are 462 * not fast-forwarded, they must clear the M_CANFASTFWD flag. 463 * Note that filters must _never_ set this flag, as another filter 464 * in the list may have previously cleared it. 465 */ 466 m0 = m; 467 pfh = pfil_hook_get(PFIL_IN, &inetsw[ip_protox[IPPROTO_IP]].pr_pfh); 468 for (; pfh; pfh = pfh->pfil_link.tqe_next) 469 if (pfh->pfil_func) { 470 rv = pfh->pfil_func(ip, hlen, 471 m->m_pkthdr.rcvif, 0, &m0); 472 if (rv) 473 return; 474 m = m0; 475 if (m == NULL) 476 return; 477 ip = mtod(m, struct ip *); 478 } 479#endif /* PFIL_HOOKS */ 480 481 /* 482 * Process options and, if not destined for us, 483 * ship it on. ip_dooptions returns 1 when an 484 * error was detected (causing an icmp message 485 * to be sent and the original packet to be freed). 486 */ 487 ip_nhops = 0; /* for source routed packets */ 488 if (hlen > sizeof (struct ip) && ip_dooptions(m)) 489 return; 490 491 /* 492 * Check our list of addresses, to see if the packet is for us. 493 * 494 * Traditional 4.4BSD did not consult IFF_UP at all. 495 * The behavior here is to treat addresses on !IFF_UP interface 496 * as not mine. 497 */ 498 downmatch = 0; 499 for (ia = IN_IFADDR_HASH(ip->ip_dst.s_addr).lh_first; 500 ia != NULL; 501 ia = ia->ia_hash.le_next) { 502 if (in_hosteq(ia->ia_addr.sin_addr, ip->ip_dst)) { 503 if ((ia->ia_ifp->if_flags & IFF_UP) != 0) 504 break; 505 else 506 downmatch++; 507 } 508 } 509 if (ia != NULL) 510 goto ours; 511 if (m->m_pkthdr.rcvif->if_flags & IFF_BROADCAST) { 512 for (ifa = m->m_pkthdr.rcvif->if_addrlist.tqh_first; 513 ifa != NULL; ifa = ifa->ifa_list.tqe_next) { 514 if (ifa->ifa_addr->sa_family != AF_INET) continue; 515 ia = ifatoia(ifa); 516 if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) || 517 in_hosteq(ip->ip_dst, ia->ia_netbroadcast) || 518 /* 519 * Look for all-0's host part (old broadcast addr), 520 * either for subnet or net. 521 */ 522 ip->ip_dst.s_addr == ia->ia_subnet || 523 ip->ip_dst.s_addr == ia->ia_net) 524 goto ours; 525 /* 526 * An interface with IP address zero accepts 527 * all packets that arrive on that interface. 528 */ 529 if (in_nullhost(ia->ia_addr.sin_addr)) 530 goto ours; 531 } 532 } 533 if (IN_MULTICAST(ip->ip_dst.s_addr)) { 534 struct in_multi *inm; 535#ifdef MROUTING 536 extern struct socket *ip_mrouter; 537 538 if (m->m_flags & M_EXT) { 539 if ((m = m_pullup(m, hlen)) == 0) { 540 ipstat.ips_toosmall++; 541 return; 542 } 543 ip = mtod(m, struct ip *); 544 } 545 546 if (ip_mrouter) { 547 /* 548 * If we are acting as a multicast router, all 549 * incoming multicast packets are passed to the 550 * kernel-level multicast forwarding function. 551 * The packet is returned (relatively) intact; if 552 * ip_mforward() returns a non-zero value, the packet 553 * must be discarded, else it may be accepted below. 554 * 555 * (The IP ident field is put in the same byte order 556 * as expected when ip_mforward() is called from 557 * ip_output().) 558 */ 559 if (ip_mforward(m, m->m_pkthdr.rcvif) != 0) { 560 ipstat.ips_cantforward++; 561 m_freem(m); 562 return; 563 } 564 565 /* 566 * The process-level routing demon needs to receive 567 * all multicast IGMP packets, whether or not this 568 * host belongs to their destination groups. 569 */ 570 if (ip->ip_p == IPPROTO_IGMP) 571 goto ours; 572 ipstat.ips_forward++; 573 } 574#endif 575 /* 576 * See if we belong to the destination multicast group on the 577 * arrival interface. 578 */ 579 IN_LOOKUP_MULTI(ip->ip_dst, m->m_pkthdr.rcvif, inm); 580 if (inm == NULL) { 581 ipstat.ips_cantforward++; 582 m_freem(m); 583 return; 584 } 585 goto ours; 586 } 587 if (ip->ip_dst.s_addr == INADDR_BROADCAST || 588 in_nullhost(ip->ip_dst)) 589 goto ours; 590 591 /* 592 * Not for us; forward if possible and desirable. 593 */ 594 if (ipforwarding == 0) { 595 ipstat.ips_cantforward++; 596 m_freem(m); 597 } else { 598 /* 599 * If ip_dst matched any of my address on !IFF_UP interface, 600 * and there's no IFF_UP interface that matches ip_dst, 601 * send icmp unreach. Forwarding it will result in in-kernel 602 * forwarding loop till TTL goes to 0. 603 */ 604 if (downmatch) { 605 icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, 0, 0); 606 ipstat.ips_cantforward++; 607 return; 608 } 609 ip_forward(m, 0); 610 } 611 return; 612 613ours: 614 /* 615 * If offset or IP_MF are set, must reassemble. 616 * Otherwise, nothing need be done. 617 * (We could look in the reassembly queue to see 618 * if the packet was previously fragmented, 619 * but it's not worth the time; just let them time out.) 620 */ 621 if (ip->ip_off & ~(IP_DF|IP_RF)) { 622 /* 623 * Look for queue of fragments 624 * of this datagram. 625 */ 626 IPQ_LOCK(); 627 for (fp = ipq.lh_first; fp != NULL; fp = fp->ipq_q.le_next) 628 if (ip->ip_id == fp->ipq_id && 629 in_hosteq(ip->ip_src, fp->ipq_src) && 630 in_hosteq(ip->ip_dst, fp->ipq_dst) && 631 ip->ip_p == fp->ipq_p) 632 goto found; 633 fp = 0; 634found: 635 636 /* 637 * Adjust ip_len to not reflect header, 638 * set ipqe_mff if more fragments are expected, 639 * convert offset of this to bytes. 640 */ 641 ip->ip_len -= hlen; 642 mff = (ip->ip_off & IP_MF) != 0; 643 if (mff) { 644 /* 645 * Make sure that fragments have a data length 646 * that's a non-zero multiple of 8 bytes. 647 */ 648 if (ip->ip_len == 0 || (ip->ip_len & 0x7) != 0) { 649 ipstat.ips_badfrags++; 650 IPQ_UNLOCK(); 651 goto bad; 652 } 653 } 654 ip->ip_off <<= 3; 655 656 /* 657 * If datagram marked as having more fragments 658 * or if this is not the first fragment, 659 * attempt reassembly; if it succeeds, proceed. 660 */ 661 if (mff || ip->ip_off) { 662 ipstat.ips_fragments++; 663 ipqe = pool_get(&ipqent_pool, PR_NOWAIT); 664 if (ipqe == NULL) { 665 ipstat.ips_rcvmemdrop++; 666 IPQ_UNLOCK(); 667 goto bad; 668 } 669 ipqe->ipqe_mff = mff; 670 ipqe->ipqe_m = m; 671 ipqe->ipqe_ip = ip; 672 m = ip_reass(ipqe, fp); 673 if (m == 0) { 674 IPQ_UNLOCK(); 675 return; 676 } 677 ipstat.ips_reassembled++; 678 ip = mtod(m, struct ip *); 679 hlen = ip->ip_hl << 2; 680 ip->ip_len += hlen; 681 } else 682 if (fp) 683 ip_freef(fp); 684 IPQ_UNLOCK(); 685 } 686 687 /* 688 * Switch out to protocol's input routine. 689 */ 690#if IFA_STATS 691 ia->ia_ifa.ifa_data.ifad_inbytes += ip->ip_len; 692#endif 693 ipstat.ips_delivered++; 694 { 695 int off = hlen, nh = ip->ip_p; 696 697 (*inetsw[ip_protox[nh]].pr_input)(m, off, nh); 698 return; 699 } 700bad: 701 m_freem(m); 702} 703 704/* 705 * Take incoming datagram fragment and try to 706 * reassemble it into whole datagram. If a chain for 707 * reassembly of this datagram already exists, then it 708 * is given as fp; otherwise have to make a chain. 709 */ 710struct mbuf * 711ip_reass(ipqe, fp) 712 register struct ipqent *ipqe; 713 register struct ipq *fp; 714{ 715 register struct mbuf *m = ipqe->ipqe_m; 716 register struct ipqent *nq, *p, *q; 717 struct ip *ip; 718 struct mbuf *t; 719 int hlen = ipqe->ipqe_ip->ip_hl << 2; 720 int i, next; 721 722 IPQ_LOCK_CHECK(); 723 724 /* 725 * Presence of header sizes in mbufs 726 * would confuse code below. 727 */ 728 m->m_data += hlen; 729 m->m_len -= hlen; 730 731 /* 732 * If first fragment to arrive, create a reassembly queue. 733 */ 734 if (fp == 0) { 735 MALLOC(fp, struct ipq *, sizeof (struct ipq), 736 M_FTABLE, M_NOWAIT); 737 if (fp == NULL) 738 goto dropfrag; 739 LIST_INSERT_HEAD(&ipq, fp, ipq_q); 740 fp->ipq_ttl = IPFRAGTTL; 741 fp->ipq_p = ipqe->ipqe_ip->ip_p; 742 fp->ipq_id = ipqe->ipqe_ip->ip_id; 743 LIST_INIT(&fp->ipq_fragq); 744 fp->ipq_src = ipqe->ipqe_ip->ip_src; 745 fp->ipq_dst = ipqe->ipqe_ip->ip_dst; 746 p = NULL; 747 goto insert; 748 } 749 750 /* 751 * Find a segment which begins after this one does. 752 */ 753 for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; 754 p = q, q = q->ipqe_q.le_next) 755 if (q->ipqe_ip->ip_off > ipqe->ipqe_ip->ip_off) 756 break; 757 758 /* 759 * If there is a preceding segment, it may provide some of 760 * our data already. If so, drop the data from the incoming 761 * segment. If it provides all of our data, drop us. 762 */ 763 if (p != NULL) { 764 i = p->ipqe_ip->ip_off + p->ipqe_ip->ip_len - 765 ipqe->ipqe_ip->ip_off; 766 if (i > 0) { 767 if (i >= ipqe->ipqe_ip->ip_len) 768 goto dropfrag; 769 m_adj(ipqe->ipqe_m, i); 770 ipqe->ipqe_ip->ip_off += i; 771 ipqe->ipqe_ip->ip_len -= i; 772 } 773 } 774 775 /* 776 * While we overlap succeeding segments trim them or, 777 * if they are completely covered, dequeue them. 778 */ 779 for (; q != NULL && ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len > 780 q->ipqe_ip->ip_off; q = nq) { 781 i = (ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len) - 782 q->ipqe_ip->ip_off; 783 if (i < q->ipqe_ip->ip_len) { 784 q->ipqe_ip->ip_len -= i; 785 q->ipqe_ip->ip_off += i; 786 m_adj(q->ipqe_m, i); 787 break; 788 } 789 nq = q->ipqe_q.le_next; 790 m_freem(q->ipqe_m); 791 LIST_REMOVE(q, ipqe_q); 792 pool_put(&ipqent_pool, q); 793 } 794 795insert: 796 /* 797 * Stick new segment in its place; 798 * check for complete reassembly. 799 */ 800 if (p == NULL) { 801 LIST_INSERT_HEAD(&fp->ipq_fragq, ipqe, ipqe_q); 802 } else { 803 LIST_INSERT_AFTER(p, ipqe, ipqe_q); 804 } 805 next = 0; 806 for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; 807 p = q, q = q->ipqe_q.le_next) { 808 if (q->ipqe_ip->ip_off != next) 809 return (0); 810 next += q->ipqe_ip->ip_len; 811 } 812 if (p->ipqe_mff) 813 return (0); 814 815 /* 816 * Reassembly is complete. Check for a bogus message size and 817 * concatenate fragments. 818 */ 819 q = fp->ipq_fragq.lh_first; 820 ip = q->ipqe_ip; 821 if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) { 822 ipstat.ips_toolong++; 823 ip_freef(fp); 824 return (0); 825 } 826 m = q->ipqe_m; 827 t = m->m_next; 828 m->m_next = 0; 829 m_cat(m, t); 830 nq = q->ipqe_q.le_next; 831 pool_put(&ipqent_pool, q); 832 for (q = nq; q != NULL; q = nq) { 833 t = q->ipqe_m; 834 nq = q->ipqe_q.le_next; 835 pool_put(&ipqent_pool, q); 836 m_cat(m, t); 837 } 838 839 /* 840 * Create header for new ip packet by 841 * modifying header of first packet; 842 * dequeue and discard fragment reassembly header. 843 * Make header visible. 844 */ 845 ip->ip_len = next; 846 ip->ip_src = fp->ipq_src; 847 ip->ip_dst = fp->ipq_dst; 848 LIST_REMOVE(fp, ipq_q); 849 FREE(fp, M_FTABLE); 850 m->m_len += (ip->ip_hl << 2); 851 m->m_data -= (ip->ip_hl << 2); 852 /* some debugging cruft by sklower, below, will go away soon */ 853 if (m->m_flags & M_PKTHDR) { /* XXX this should be done elsewhere */ 854 register int plen = 0; 855 for (t = m; t; t = t->m_next) 856 plen += t->m_len; 857 m->m_pkthdr.len = plen; 858 } 859 return (m); 860 861dropfrag: 862 ipstat.ips_fragdropped++; 863 m_freem(m); 864 pool_put(&ipqent_pool, ipqe); 865 return (0); 866} 867 868/* 869 * Free a fragment reassembly header and all 870 * associated datagrams. 871 */ 872void 873ip_freef(fp) 874 struct ipq *fp; 875{ 876 register struct ipqent *q, *p; 877 878 IPQ_LOCK_CHECK(); 879 880 for (q = fp->ipq_fragq.lh_first; q != NULL; q = p) { 881 p = q->ipqe_q.le_next; 882 m_freem(q->ipqe_m); 883 LIST_REMOVE(q, ipqe_q); 884 pool_put(&ipqent_pool, q); 885 } 886 LIST_REMOVE(fp, ipq_q); 887 FREE(fp, M_FTABLE); 888} 889 890/* 891 * IP timer processing; 892 * if a timer expires on a reassembly 893 * queue, discard it. 894 */ 895void 896ip_slowtimo() 897{ 898 register struct ipq *fp, *nfp; 899 int s = splsoftnet(); 900 901 IPQ_LOCK(); 902 for (fp = ipq.lh_first; fp != NULL; fp = nfp) { 903 nfp = fp->ipq_q.le_next; 904 if (--fp->ipq_ttl == 0) { 905 ipstat.ips_fragtimeout++; 906 ip_freef(fp); 907 } 908 } 909 IPQ_UNLOCK(); 910#ifdef GATEWAY 911 ipflow_slowtimo(); 912#endif 913 splx(s); 914} 915 916/* 917 * Drain off all datagram fragments. 918 */ 919void 920ip_drain() 921{ 922 923 /* 924 * We may be called from a device's interrupt context. If 925 * the ipq is already busy, just bail out now. 926 */ 927 if (ipq_lock_try() == 0) 928 return; 929 930 while (ipq.lh_first != NULL) { 931 ipstat.ips_fragdropped++; 932 ip_freef(ipq.lh_first); 933 } 934 935 IPQ_UNLOCK(); 936} 937 938/* 939 * Do option processing on a datagram, 940 * possibly discarding it if bad options are encountered, 941 * or forwarding it if source-routed. 942 * Returns 1 if packet has been forwarded/freed, 943 * 0 if the packet should be processed further. 944 */ 945int 946ip_dooptions(m) 947 struct mbuf *m; 948{ 949 register struct ip *ip = mtod(m, struct ip *); 950 register u_char *cp, *cp0; 951 register struct ip_timestamp *ipt; 952 register struct in_ifaddr *ia; 953 int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0; 954 struct in_addr dst; 955 n_time ntime; 956 957 dst = ip->ip_dst; 958 cp = (u_char *)(ip + 1); 959 cnt = (ip->ip_hl << 2) - sizeof (struct ip); 960 for (; cnt > 0; cnt -= optlen, cp += optlen) { 961 opt = cp[IPOPT_OPTVAL]; 962 if (opt == IPOPT_EOL) 963 break; 964 if (opt == IPOPT_NOP) 965 optlen = 1; 966 else { 967 optlen = cp[IPOPT_OLEN]; 968 if (optlen <= 0 || optlen > cnt) { 969 code = &cp[IPOPT_OLEN] - (u_char *)ip; 970 goto bad; 971 } 972 } 973 switch (opt) { 974 975 default: 976 break; 977 978 /* 979 * Source routing with record. 980 * Find interface with current destination address. 981 * If none on this machine then drop if strictly routed, 982 * or do nothing if loosely routed. 983 * Record interface address and bring up next address 984 * component. If strictly routed make sure next 985 * address is on directly accessible net. 986 */ 987 case IPOPT_LSRR: 988 case IPOPT_SSRR: 989 if (ip_allowsrcrt == 0) { 990 type = ICMP_UNREACH; 991 code = ICMP_UNREACH_NET_PROHIB; 992 goto bad; 993 } 994 if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) { 995 code = &cp[IPOPT_OFFSET] - (u_char *)ip; 996 goto bad; 997 } 998 ipaddr.sin_addr = ip->ip_dst; 999 ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))); 1000 if (ia == 0) { 1001 if (opt == IPOPT_SSRR) { 1002 type = ICMP_UNREACH; 1003 code = ICMP_UNREACH_SRCFAIL; 1004 goto bad; 1005 } 1006 /* 1007 * Loose routing, and not at next destination 1008 * yet; nothing to do except forward. 1009 */ 1010 break; 1011 } 1012 off--; /* 0 origin */ 1013 if (off > optlen - sizeof(struct in_addr)) { 1014 /* 1015 * End of source route. Should be for us. 1016 */ 1017 save_rte(cp, ip->ip_src); 1018 break; 1019 } 1020 /* 1021 * locate outgoing interface 1022 */ 1023 bcopy((caddr_t)(cp + off), (caddr_t)&ipaddr.sin_addr, 1024 sizeof(ipaddr.sin_addr)); 1025 if (opt == IPOPT_SSRR) 1026 ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))); 1027 else 1028 ia = ip_rtaddr(ipaddr.sin_addr); 1029 if (ia == 0) { 1030 type = ICMP_UNREACH; 1031 code = ICMP_UNREACH_SRCFAIL; 1032 goto bad; 1033 } 1034 ip->ip_dst = ipaddr.sin_addr; 1035 bcopy((caddr_t)&ia->ia_addr.sin_addr, 1036 (caddr_t)(cp + off), sizeof(struct in_addr)); 1037 cp[IPOPT_OFFSET] += sizeof(struct in_addr); 1038 /* 1039 * Let ip_intr's mcast routing check handle mcast pkts 1040 */ 1041 forward = !IN_MULTICAST(ip->ip_dst.s_addr); 1042 break; 1043 1044 case IPOPT_RR: 1045 if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) { 1046 code = &cp[IPOPT_OFFSET] - (u_char *)ip; 1047 goto bad; 1048 } 1049 /* 1050 * If no space remains, ignore. 1051 */ 1052 off--; /* 0 origin */ 1053 if (off > optlen - sizeof(struct in_addr)) 1054 break; 1055 bcopy((caddr_t)(&ip->ip_dst), (caddr_t)&ipaddr.sin_addr, 1056 sizeof(ipaddr.sin_addr)); 1057 /* 1058 * locate outgoing interface; if we're the destination, 1059 * use the incoming interface (should be same). 1060 */ 1061 if ((ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)))) 1062 == NULL && 1063 (ia = ip_rtaddr(ipaddr.sin_addr)) == NULL) { 1064 type = ICMP_UNREACH; 1065 code = ICMP_UNREACH_HOST; 1066 goto bad; 1067 } 1068 bcopy((caddr_t)&ia->ia_addr.sin_addr, 1069 (caddr_t)(cp + off), sizeof(struct in_addr)); 1070 cp[IPOPT_OFFSET] += sizeof(struct in_addr); 1071 break; 1072 1073 case IPOPT_TS: 1074 code = cp - (u_char *)ip; 1075 ipt = (struct ip_timestamp *)cp; 1076 if (ipt->ipt_len < 5) 1077 goto bad; 1078 if (ipt->ipt_ptr > ipt->ipt_len - sizeof (int32_t)) { 1079 if (++ipt->ipt_oflw == 0) 1080 goto bad; 1081 break; 1082 } 1083 cp0 = (cp + ipt->ipt_ptr - 1); 1084 switch (ipt->ipt_flg) { 1085 1086 case IPOPT_TS_TSONLY: 1087 break; 1088 1089 case IPOPT_TS_TSANDADDR: 1090 if (ipt->ipt_ptr - 1 + sizeof(n_time) + 1091 sizeof(struct in_addr) > ipt->ipt_len) 1092 goto bad; 1093 ipaddr.sin_addr = dst; 1094 ia = ifatoia(ifaof_ifpforaddr(sintosa(&ipaddr), 1095 m->m_pkthdr.rcvif)); 1096 if (ia == 0) 1097 continue; 1098 bcopy(&ia->ia_addr.sin_addr, 1099 cp0, sizeof(struct in_addr)); 1100 ipt->ipt_ptr += sizeof(struct in_addr); 1101 break; 1102 1103 case IPOPT_TS_PRESPEC: 1104 if (ipt->ipt_ptr - 1 + sizeof(n_time) + 1105 sizeof(struct in_addr) > ipt->ipt_len) 1106 goto bad; 1107 bcopy(cp0, &ipaddr.sin_addr, 1108 sizeof(struct in_addr)); 1109 if (ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))) 1110 == NULL) 1111 continue; 1112 ipt->ipt_ptr += sizeof(struct in_addr); 1113 break; 1114 1115 default: 1116 goto bad; 1117 } 1118 ntime = iptime(); 1119 bcopy(&ntime, (caddr_t)cp + ipt->ipt_ptr - 1, 1120 sizeof(n_time)); 1121 ipt->ipt_ptr += sizeof(n_time); 1122 } 1123 } 1124 if (forward) { 1125 if (ip_forwsrcrt == 0) { 1126 type = ICMP_UNREACH; 1127 code = ICMP_UNREACH_SRCFAIL; 1128 goto bad; 1129 } 1130 ip_forward(m, 1); 1131 return (1); 1132 } 1133 return (0); 1134bad: 1135 icmp_error(m, type, code, 0, 0); 1136 ipstat.ips_badoptions++; 1137 return (1); 1138} 1139 1140/* 1141 * Given address of next destination (final or next hop), 1142 * return internet address info of interface to be used to get there. 1143 */ 1144struct in_ifaddr * 1145ip_rtaddr(dst) 1146 struct in_addr dst; 1147{ 1148 register struct sockaddr_in *sin; 1149 1150 sin = satosin(&ipforward_rt.ro_dst); 1151 1152 if (ipforward_rt.ro_rt == 0 || !in_hosteq(dst, sin->sin_addr)) { 1153 if (ipforward_rt.ro_rt) { 1154 RTFREE(ipforward_rt.ro_rt); 1155 ipforward_rt.ro_rt = 0; 1156 } 1157 sin->sin_family = AF_INET; 1158 sin->sin_len = sizeof(*sin); 1159 sin->sin_addr = dst; 1160 1161 rtalloc(&ipforward_rt); 1162 } 1163 if (ipforward_rt.ro_rt == 0) 1164 return ((struct in_ifaddr *)0); 1165 return (ifatoia(ipforward_rt.ro_rt->rt_ifa)); 1166} 1167 1168/* 1169 * Save incoming source route for use in replies, 1170 * to be picked up later by ip_srcroute if the receiver is interested. 1171 */ 1172void 1173save_rte(option, dst) 1174 u_char *option; 1175 struct in_addr dst; 1176{ 1177 unsigned olen; 1178 1179 olen = option[IPOPT_OLEN]; 1180#ifdef DIAGNOSTIC 1181 if (ipprintfs) 1182 printf("save_rte: olen %d\n", olen); 1183#endif /* 0 */ 1184 if (olen > sizeof(ip_srcrt) - (1 + sizeof(dst))) 1185 return; 1186 bcopy((caddr_t)option, (caddr_t)ip_srcrt.srcopt, olen); 1187 ip_nhops = (olen - IPOPT_OFFSET - 1) / sizeof(struct in_addr); 1188 ip_srcrt.dst = dst; 1189} 1190 1191/* 1192 * Retrieve incoming source route for use in replies, 1193 * in the same form used by setsockopt. 1194 * The first hop is placed before the options, will be removed later. 1195 */ 1196struct mbuf * 1197ip_srcroute() 1198{ 1199 register struct in_addr *p, *q; 1200 register struct mbuf *m; 1201 1202 if (ip_nhops == 0) 1203 return ((struct mbuf *)0); 1204 m = m_get(M_DONTWAIT, MT_SOOPTS); 1205 if (m == 0) 1206 return ((struct mbuf *)0); 1207 1208#define OPTSIZ (sizeof(ip_srcrt.nop) + sizeof(ip_srcrt.srcopt)) 1209 1210 /* length is (nhops+1)*sizeof(addr) + sizeof(nop + srcrt header) */ 1211 m->m_len = ip_nhops * sizeof(struct in_addr) + sizeof(struct in_addr) + 1212 OPTSIZ; 1213#ifdef DIAGNOSTIC 1214 if (ipprintfs) 1215 printf("ip_srcroute: nhops %d mlen %d", ip_nhops, m->m_len); 1216#endif 1217 1218 /* 1219 * First save first hop for return route 1220 */ 1221 p = &ip_srcrt.route[ip_nhops - 1]; 1222 *(mtod(m, struct in_addr *)) = *p--; 1223#ifdef DIAGNOSTIC 1224 if (ipprintfs) 1225 printf(" hops %x", ntohl(mtod(m, struct in_addr *)->s_addr)); 1226#endif 1227 1228 /* 1229 * Copy option fields and padding (nop) to mbuf. 1230 */ 1231 ip_srcrt.nop = IPOPT_NOP; 1232 ip_srcrt.srcopt[IPOPT_OFFSET] = IPOPT_MINOFF; 1233 bcopy((caddr_t)&ip_srcrt.nop, 1234 mtod(m, caddr_t) + sizeof(struct in_addr), OPTSIZ); 1235 q = (struct in_addr *)(mtod(m, caddr_t) + 1236 sizeof(struct in_addr) + OPTSIZ); 1237#undef OPTSIZ 1238 /* 1239 * Record return path as an IP source route, 1240 * reversing the path (pointers are now aligned). 1241 */ 1242 while (p >= ip_srcrt.route) { 1243#ifdef DIAGNOSTIC 1244 if (ipprintfs) 1245 printf(" %x", ntohl(q->s_addr)); 1246#endif 1247 *q++ = *p--; 1248 } 1249 /* 1250 * Last hop goes to final destination. 1251 */ 1252 *q = ip_srcrt.dst; 1253#ifdef DIAGNOSTIC 1254 if (ipprintfs) 1255 printf(" %x\n", ntohl(q->s_addr)); 1256#endif 1257 return (m); 1258} 1259 1260/* 1261 * Strip out IP options, at higher 1262 * level protocol in the kernel. 1263 * Second argument is buffer to which options 1264 * will be moved, and return value is their length. 1265 * XXX should be deleted; last arg currently ignored. 1266 */ 1267void 1268ip_stripoptions(m, mopt) 1269 register struct mbuf *m; 1270 struct mbuf *mopt; 1271{ 1272 register int i; 1273 struct ip *ip = mtod(m, struct ip *); 1274 register caddr_t opts; 1275 int olen; 1276 1277 olen = (ip->ip_hl << 2) - sizeof (struct ip); 1278 opts = (caddr_t)(ip + 1); 1279 i = m->m_len - (sizeof (struct ip) + olen); 1280 bcopy(opts + olen, opts, (unsigned)i); 1281 m->m_len -= olen; 1282 if (m->m_flags & M_PKTHDR) 1283 m->m_pkthdr.len -= olen; 1284 ip->ip_len -= olen; 1285 ip->ip_hl = sizeof (struct ip) >> 2; 1286} 1287 1288int inetctlerrmap[PRC_NCMDS] = { 1289 0, 0, 0, 0, 1290 0, EMSGSIZE, EHOSTDOWN, EHOSTUNREACH, 1291 EHOSTUNREACH, EHOSTUNREACH, ECONNREFUSED, ECONNREFUSED, 1292 EMSGSIZE, EHOSTUNREACH, 0, 0, 1293 0, 0, 0, 0, 1294 ENOPROTOOPT 1295}; 1296 1297/* 1298 * Forward a packet. If some error occurs return the sender 1299 * an icmp packet. Note we can't always generate a meaningful 1300 * icmp message because icmp doesn't have a large enough repertoire 1301 * of codes and types. 1302 * 1303 * If not forwarding, just drop the packet. This could be confusing 1304 * if ipforwarding was zero but some routing protocol was advancing 1305 * us as a gateway to somewhere. However, we must let the routing 1306 * protocol deal with that. 1307 * 1308 * The srcrt parameter indicates whether the packet is being forwarded 1309 * via a source route. 1310 */ 1311void 1312ip_forward(m, srcrt) 1313 struct mbuf *m; 1314 int srcrt; 1315{ 1316 register struct ip *ip = mtod(m, struct ip *); 1317 register struct sockaddr_in *sin; 1318 register struct rtentry *rt; 1319 int error, type = 0, code = 0; 1320 struct mbuf *mcopy; 1321 n_long dest; 1322 struct ifnet *destifp; 1323#ifdef IPSEC 1324 struct ifnet dummyifp; 1325#endif 1326 1327 dest = 0; 1328#ifdef DIAGNOSTIC 1329 if (ipprintfs) 1330 printf("forward: src %2.2x dst %2.2x ttl %x\n", 1331 ntohl(ip->ip_src.s_addr), 1332 ntohl(ip->ip_dst.s_addr), ip->ip_ttl); 1333#endif 1334 if (m->m_flags & (M_BCAST|M_MCAST) || in_canforward(ip->ip_dst) == 0) { 1335 ipstat.ips_cantforward++; 1336 m_freem(m); 1337 return; 1338 } 1339 if (ip->ip_ttl <= IPTTLDEC) { 1340 icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0); 1341 return; 1342 } 1343 ip->ip_ttl -= IPTTLDEC; 1344 1345 sin = satosin(&ipforward_rt.ro_dst); 1346 if ((rt = ipforward_rt.ro_rt) == 0 || 1347 !in_hosteq(ip->ip_dst, sin->sin_addr)) { 1348 if (ipforward_rt.ro_rt) { 1349 RTFREE(ipforward_rt.ro_rt); 1350 ipforward_rt.ro_rt = 0; 1351 } 1352 sin->sin_family = AF_INET; 1353 sin->sin_len = sizeof(struct sockaddr_in); 1354 sin->sin_addr = ip->ip_dst; 1355 1356 rtalloc(&ipforward_rt); 1357 if (ipforward_rt.ro_rt == 0) { 1358 icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, dest, 0); 1359 return; 1360 } 1361 rt = ipforward_rt.ro_rt; 1362 } 1363 1364 /* 1365 * Save at most 68 bytes of the packet in case 1366 * we need to generate an ICMP message to the src. 1367 */ 1368 mcopy = m_copy(m, 0, imin((int)ip->ip_len, 68)); 1369 1370 /* 1371 * If forwarding packet using same interface that it came in on, 1372 * perhaps should send a redirect to sender to shortcut a hop. 1373 * Only send redirect if source is sending directly to us, 1374 * and if packet was not source routed (or has any options). 1375 * Also, don't send redirect if forwarding using a default route 1376 * or a route modified by a redirect. 1377 */ 1378 if (rt->rt_ifp == m->m_pkthdr.rcvif && 1379 (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 && 1380 !in_nullhost(satosin(rt_key(rt))->sin_addr) && 1381 ipsendredirects && !srcrt) { 1382 if (rt->rt_ifa && 1383 (ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) == 1384 ifatoia(rt->rt_ifa)->ia_subnet) { 1385 if (rt->rt_flags & RTF_GATEWAY) 1386 dest = satosin(rt->rt_gateway)->sin_addr.s_addr; 1387 else 1388 dest = ip->ip_dst.s_addr; 1389 /* 1390 * Router requirements says to only send host 1391 * redirects. 1392 */ 1393 type = ICMP_REDIRECT; 1394 code = ICMP_REDIRECT_HOST; 1395#ifdef DIAGNOSTIC 1396 if (ipprintfs) 1397 printf("redirect (%d) to %x\n", code, 1398 (u_int32_t)dest); 1399#endif 1400 } 1401 } 1402 1403#ifdef IPSEC 1404 /* Don't lookup socket in forwading case */ 1405 ipsec_setsocket(m, NULL); 1406#endif 1407 error = ip_output(m, (struct mbuf *)0, &ipforward_rt, 1408 (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)), 0); 1409 if (error) 1410 ipstat.ips_cantforward++; 1411 else { 1412 ipstat.ips_forward++; 1413 if (type) 1414 ipstat.ips_redirectsent++; 1415 else { 1416 if (mcopy) { 1417#ifdef GATEWAY 1418 if (mcopy->m_flags & M_CANFASTFWD) 1419 ipflow_create(&ipforward_rt, mcopy); 1420#endif 1421 m_freem(mcopy); 1422 } 1423 return; 1424 } 1425 } 1426 if (mcopy == NULL) 1427 return; 1428 destifp = NULL; 1429 1430 switch (error) { 1431 1432 case 0: /* forwarded, but need redirect */ 1433 /* type, code set above */ 1434 break; 1435 1436 case ENETUNREACH: /* shouldn't happen, checked above */ 1437 case EHOSTUNREACH: 1438 case ENETDOWN: 1439 case EHOSTDOWN: 1440 default: 1441 type = ICMP_UNREACH; 1442 code = ICMP_UNREACH_HOST; 1443 break; 1444 1445 case EMSGSIZE: 1446 type = ICMP_UNREACH; 1447 code = ICMP_UNREACH_NEEDFRAG; 1448#ifndef IPSEC 1449 if (ipforward_rt.ro_rt) 1450 destifp = ipforward_rt.ro_rt->rt_ifp; 1451#else 1452 /* 1453 * If the packet is routed over IPsec tunnel, tell the 1454 * originator the tunnel MTU. 1455 * tunnel MTU = if MTU - sizeof(IP) - ESP/AH hdrsiz 1456 * XXX quickhack!!! 1457 */ 1458 if (ipforward_rt.ro_rt) { 1459 struct secpolicy *sp; 1460 int ipsecerror; 1461 size_t ipsechdr; 1462 struct route *ro; 1463 1464 sp = ipsec4_getpolicybyaddr(mcopy, 1465 IPSEC_DIR_OUTBOUND, 1466 IP_FORWARDING, 1467 &ipsecerror); 1468 1469 if (sp == NULL) 1470 destifp = ipforward_rt.ro_rt->rt_ifp; 1471 else { 1472 /* count IPsec header size */ 1473 ipsechdr = ipsec4_hdrsiz(mcopy, 1474 IPSEC_DIR_OUTBOUND, 1475 NULL); 1476 1477 /* 1478 * find the correct route for outer IPv4 1479 * header, compute tunnel MTU. 1480 * 1481 * XXX BUG ALERT 1482 * The "dummyifp" code relies upon the fact 1483 * that icmp_error() touches only ifp->if_mtu. 1484 */ 1485 /*XXX*/ 1486 destifp = NULL; 1487 if (sp->req != NULL 1488 && sp->req->sav != NULL 1489 && sp->req->sav->sah != NULL) { 1490 ro = &sp->req->sav->sah->sa_route; 1491 if (ro->ro_rt && ro->ro_rt->rt_ifp) { 1492 dummyifp.if_mtu = 1493 ro->ro_rt->rt_ifp->if_mtu; 1494 dummyifp.if_mtu -= ipsechdr; 1495 destifp = &dummyifp; 1496 } 1497 } 1498 1499 key_freesp(sp); 1500 } 1501 } 1502#endif /*IPSEC*/ 1503 ipstat.ips_cantfrag++; 1504 break; 1505 1506 case ENOBUFS: 1507 type = ICMP_SOURCEQUENCH; 1508 code = 0; 1509 break; 1510 } 1511 icmp_error(mcopy, type, code, dest, destifp); 1512} 1513 1514void 1515ip_savecontrol(inp, mp, ip, m) 1516 register struct inpcb *inp; 1517 register struct mbuf **mp; 1518 register struct ip *ip; 1519 register struct mbuf *m; 1520{ 1521 1522 if (inp->inp_socket->so_options & SO_TIMESTAMP) { 1523 struct timeval tv; 1524 1525 microtime(&tv); 1526 *mp = sbcreatecontrol((caddr_t) &tv, sizeof(tv), 1527 SCM_TIMESTAMP, SOL_SOCKET); 1528 if (*mp) 1529 mp = &(*mp)->m_next; 1530 } 1531 if (inp->inp_flags & INP_RECVDSTADDR) { 1532 *mp = sbcreatecontrol((caddr_t) &ip->ip_dst, 1533 sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP); 1534 if (*mp) 1535 mp = &(*mp)->m_next; 1536 } 1537#ifdef notyet 1538 /* 1539 * XXX 1540 * Moving these out of udp_input() made them even more broken 1541 * than they already were. 1542 * - fenner@parc.xerox.com 1543 */ 1544 /* options were tossed already */ 1545 if (inp->inp_flags & INP_RECVOPTS) { 1546 *mp = sbcreatecontrol((caddr_t) opts_deleted_above, 1547 sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP); 1548 if (*mp) 1549 mp = &(*mp)->m_next; 1550 } 1551 /* ip_srcroute doesn't do what we want here, need to fix */ 1552 if (inp->inp_flags & INP_RECVRETOPTS) { 1553 *mp = sbcreatecontrol((caddr_t) ip_srcroute(), 1554 sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP); 1555 if (*mp) 1556 mp = &(*mp)->m_next; 1557 } 1558#endif 1559 if (inp->inp_flags & INP_RECVIF) { 1560 struct sockaddr_dl sdl; 1561 1562 sdl.sdl_len = offsetof(struct sockaddr_dl, sdl_data[0]); 1563 sdl.sdl_family = AF_LINK; 1564 sdl.sdl_index = m->m_pkthdr.rcvif ? 1565 m->m_pkthdr.rcvif->if_index : 0; 1566 sdl.sdl_nlen = sdl.sdl_alen = sdl.sdl_slen = 0; 1567 *mp = sbcreatecontrol((caddr_t) &sdl, sdl.sdl_len, 1568 IP_RECVIF, IPPROTO_IP); 1569 if (*mp) 1570 mp = &(*mp)->m_next; 1571 } 1572} 1573 1574int 1575ip_sysctl(name, namelen, oldp, oldlenp, newp, newlen) 1576 int *name; 1577 u_int namelen; 1578 void *oldp; 1579 size_t *oldlenp; 1580 void *newp; 1581 size_t newlen; 1582{ 1583 extern int subnetsarelocal, hostzeroisbroadcast; 1584 1585 int error, old; 1586 1587 /* All sysctl names at this level are terminal. */ 1588 if (namelen != 1) 1589 return (ENOTDIR); 1590 1591 switch (name[0]) { 1592 case IPCTL_FORWARDING: 1593 return (sysctl_int(oldp, oldlenp, newp, newlen, &ipforwarding)); 1594 case IPCTL_SENDREDIRECTS: 1595 return (sysctl_int(oldp, oldlenp, newp, newlen, 1596 &ipsendredirects)); 1597 case IPCTL_DEFTTL: 1598 return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_defttl)); 1599#ifdef notyet 1600 case IPCTL_DEFMTU: 1601 return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_mtu)); 1602#endif 1603 case IPCTL_FORWSRCRT: 1604 /* Don't allow this to change in a secure environment. */ 1605 if (securelevel > 0) 1606 return (sysctl_rdint(oldp, oldlenp, newp, 1607 ip_forwsrcrt)); 1608 else 1609 return (sysctl_int(oldp, oldlenp, newp, newlen, 1610 &ip_forwsrcrt)); 1611 case IPCTL_DIRECTEDBCAST: 1612 return (sysctl_int(oldp, oldlenp, newp, newlen, 1613 &ip_directedbcast)); 1614 case IPCTL_ALLOWSRCRT: 1615 return (sysctl_int(oldp, oldlenp, newp, newlen, 1616 &ip_allowsrcrt)); 1617 case IPCTL_SUBNETSARELOCAL: 1618 return (sysctl_int(oldp, oldlenp, newp, newlen, 1619 &subnetsarelocal)); 1620 case IPCTL_MTUDISC: 1621 error = sysctl_int(oldp, oldlenp, newp, newlen, 1622 &ip_mtudisc); 1623 if (ip_mtudisc != 0 && ip_mtudisc_timeout_q == NULL) { 1624 ip_mtudisc_timeout_q = 1625 rt_timer_queue_create(ip_mtudisc_timeout); 1626 } else if (ip_mtudisc == 0 && ip_mtudisc_timeout_q != NULL) { 1627 rt_timer_queue_destroy(ip_mtudisc_timeout_q, TRUE); 1628 ip_mtudisc_timeout_q = NULL; 1629 } 1630 return error; 1631 case IPCTL_ANONPORTMIN: 1632 old = anonportmin; 1633 error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmin); 1634 if (anonportmin >= anonportmax || anonportmin > 65535 1635#ifndef IPNOPRIVPORTS 1636 || anonportmin < IPPORT_RESERVED 1637#endif 1638 ) { 1639 anonportmin = old; 1640 return (EINVAL); 1641 } 1642 return (error); 1643 case IPCTL_ANONPORTMAX: 1644 old = anonportmax; 1645 error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmax); 1646 if (anonportmin >= anonportmax || anonportmax > 65535 1647#ifndef IPNOPRIVPORTS 1648 || anonportmax < IPPORT_RESERVED 1649#endif 1650 ) { 1651 anonportmax = old; 1652 return (EINVAL); 1653 } 1654 return (error); 1655 case IPCTL_MTUDISCTIMEOUT: 1656 error = sysctl_int(oldp, oldlenp, newp, newlen, 1657 &ip_mtudisc_timeout); 1658 if (ip_mtudisc_timeout_q != NULL) 1659 rt_timer_queue_change(ip_mtudisc_timeout_q, 1660 ip_mtudisc_timeout); 1661 return (error); 1662#ifdef GATEWAY 1663 case IPCTL_MAXFLOWS: 1664 { 1665 int s; 1666 1667 error = sysctl_int(oldp, oldlenp, newp, newlen, 1668 &ip_maxflows); 1669 s = splsoftnet(); 1670 ipflow_reap(0); 1671 splx(s); 1672 return (error); 1673 } 1674#endif 1675 case IPCTL_HOSTZEROBROADCAST: 1676 return (sysctl_int(oldp, oldlenp, newp, newlen, 1677 &hostzeroisbroadcast)); 1678#if NGIF > 0 1679 case IPCTL_GIF_TTL: 1680 return(sysctl_int(oldp, oldlenp, newp, newlen, 1681 &ip_gif_ttl)); 1682#endif 1683 1684 default: 1685 return (EOPNOTSUPP); 1686 } 1687 /* NOTREACHED */ 1688} 1689