ip_input.c revision 1.69
1/* $NetBSD: ip_input.c,v 1.69 1998/08/09 08:58:19 mrg Exp $ */ 2 3/*- 4 * Copyright (c) 1998 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Public Access Networks Corporation ("Panix"). It was developed under 9 * contract to Panix by Eric Haszlakiewicz and Thor Lancelot Simon. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 3. All advertising materials mentioning features or use of this software 20 * must display the following acknowledgement: 21 * This product includes software developed by the NetBSD 22 * Foundation, Inc. and its contributors. 23 * 4. Neither the name of The NetBSD Foundation nor the names of its 24 * contributors may be used to endorse or promote products derived 25 * from this software without specific prior written permission. 26 * 27 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 28 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 29 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 30 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 31 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 32 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 33 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 34 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 35 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 36 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 37 * POSSIBILITY OF SUCH DAMAGE. 38 */ 39 40/* 41 * Copyright (c) 1982, 1986, 1988, 1993 42 * The Regents of the University of California. All rights reserved. 43 * 44 * Redistribution and use in source and binary forms, with or without 45 * modification, are permitted provided that the following conditions 46 * are met: 47 * 1. Redistributions of source code must retain the above copyright 48 * notice, this list of conditions and the following disclaimer. 49 * 2. Redistributions in binary form must reproduce the above copyright 50 * notice, this list of conditions and the following disclaimer in the 51 * documentation and/or other materials provided with the distribution. 52 * 3. All advertising materials mentioning features or use of this software 53 * must display the following acknowledgement: 54 * This product includes software developed by the University of 55 * California, Berkeley and its contributors. 56 * 4. Neither the name of the University nor the names of its contributors 57 * may be used to endorse or promote products derived from this software 58 * without specific prior written permission. 59 * 60 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 61 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 62 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 63 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 64 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 65 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 66 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 67 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 68 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 69 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 70 * SUCH DAMAGE. 71 * 72 * @(#)ip_input.c 8.2 (Berkeley) 1/4/94 73 */ 74 75#include "opt_gateway.h" 76#include "opt_pfil_hooks.h" 77#include "opt_mrouting.h" 78 79#include <sys/param.h> 80#include <sys/systm.h> 81#include <sys/malloc.h> 82#include <sys/mbuf.h> 83#include <sys/domain.h> 84#include <sys/protosw.h> 85#include <sys/socket.h> 86#include <sys/socketvar.h> 87#include <sys/errno.h> 88#include <sys/time.h> 89#include <sys/kernel.h> 90#include <sys/proc.h> 91 92#include <vm/vm.h> 93#include <sys/sysctl.h> 94 95#include <net/if.h> 96#include <net/if_dl.h> 97#include <net/route.h> 98#include <net/pfil.h> 99 100#include <netinet/in.h> 101#include <netinet/in_systm.h> 102#include <netinet/ip.h> 103#include <netinet/in_pcb.h> 104#include <netinet/in_var.h> 105#include <netinet/ip_var.h> 106#include <netinet/ip_icmp.h> 107 108#ifndef IPFORWARDING 109#ifdef GATEWAY 110#define IPFORWARDING 1 /* forward IP packets not for us */ 111#else /* GATEWAY */ 112#define IPFORWARDING 0 /* don't forward IP packets not for us */ 113#endif /* GATEWAY */ 114#endif /* IPFORWARDING */ 115#ifndef IPSENDREDIRECTS 116#define IPSENDREDIRECTS 1 117#endif 118#ifndef IPFORWSRCRT 119#define IPFORWSRCRT 1 /* forward source-routed packets */ 120#endif 121#ifndef IPALLOWSRCRT 122#define IPALLOWSRCRT 1 /* allow source-routed packets */ 123#endif 124#ifndef IPMTUDISC 125#define IPMTUDISC 0 126#endif 127#ifndef IPMTUDISCTIMEOUT 128#define IPMTUDISCTIMEOUT (10 * 60) /* as per RFC 1191 */ 129#endif 130 131/* 132 * Note: DIRECTED_BROADCAST is handled this way so that previous 133 * configuration using this option will Just Work. 134 */ 135#ifndef IPDIRECTEDBCAST 136#ifdef DIRECTED_BROADCAST 137#define IPDIRECTEDBCAST 1 138#else 139#define IPDIRECTEDBCAST 0 140#endif /* DIRECTED_BROADCAST */ 141#endif /* IPDIRECTEDBCAST */ 142int ipforwarding = IPFORWARDING; 143int ipsendredirects = IPSENDREDIRECTS; 144int ip_defttl = IPDEFTTL; 145int ip_forwsrcrt = IPFORWSRCRT; 146int ip_directedbcast = IPDIRECTEDBCAST; 147int ip_allowsrcrt = IPALLOWSRCRT; 148int ip_mtudisc = IPMTUDISC; 149u_int ip_mtudisc_timeout = IPMTUDISCTIMEOUT; 150#ifdef DIAGNOSTIC 151int ipprintfs = 0; 152#endif 153 154struct rttimer_queue *ip_mtudisc_timeout_q = NULL; 155 156extern struct domain inetdomain; 157extern struct protosw inetsw[]; 158u_char ip_protox[IPPROTO_MAX]; 159int ipqmaxlen = IFQ_MAXLEN; 160struct in_ifaddrhead in_ifaddr; 161struct in_ifaddrhashhead *in_ifaddrhashtbl; 162struct ifqueue ipintrq; 163struct ipstat ipstat; 164u_int16_t ip_id; 165int ip_defttl; 166struct ipqhead ipq; 167 168/* 169 * We need to save the IP options in case a protocol wants to respond 170 * to an incoming packet over the same route if the packet got here 171 * using IP source routing. This allows connection establishment and 172 * maintenance when the remote end is on a network that is not known 173 * to us. 174 */ 175int ip_nhops = 0; 176static struct ip_srcrt { 177 struct in_addr dst; /* final destination */ 178 char nop; /* one NOP to align */ 179 char srcopt[IPOPT_OFFSET + 1]; /* OPTVAL, OLEN and OFFSET */ 180 struct in_addr route[MAX_IPOPTLEN/sizeof(struct in_addr)]; 181} ip_srcrt; 182 183static void save_rte __P((u_char *, struct in_addr)); 184 185/* 186 * IP initialization: fill in IP protocol switch table. 187 * All protocols not implemented in kernel go to raw IP protocol handler. 188 */ 189void 190ip_init() 191{ 192 register struct protosw *pr; 193 register int i; 194 195 pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW); 196 if (pr == 0) 197 panic("ip_init"); 198 for (i = 0; i < IPPROTO_MAX; i++) 199 ip_protox[i] = pr - inetsw; 200 for (pr = inetdomain.dom_protosw; 201 pr < inetdomain.dom_protoswNPROTOSW; pr++) 202 if (pr->pr_domain->dom_family == PF_INET && 203 pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW) 204 ip_protox[pr->pr_protocol] = pr - inetsw; 205 LIST_INIT(&ipq); 206 ip_id = time.tv_sec & 0xffff; 207 ipintrq.ifq_maxlen = ipqmaxlen; 208 TAILQ_INIT(&in_ifaddr); 209 in_ifaddrhashtbl = 210 hashinit(IN_IFADDR_HASH_SIZE, M_IFADDR, M_WAITOK, &in_ifaddrhash); 211 if (ip_mtudisc != 0) 212 ip_mtudisc_timeout_q = 213 rt_timer_queue_create(ip_mtudisc_timeout); 214} 215 216struct sockaddr_in ipaddr = { sizeof(ipaddr), AF_INET }; 217struct route ipforward_rt; 218 219/* 220 * Ip input routine. Checksum and byte swap header. If fragmented 221 * try to reassemble. Process options. Pass to next level. 222 */ 223void 224ipintr() 225{ 226 register struct ip *ip = NULL; 227 register struct mbuf *m; 228 register struct ipq *fp; 229 register struct in_ifaddr *ia; 230 register struct ifaddr *ifa; 231 struct ipqent *ipqe; 232 int hlen = 0, mff, len, s; 233#ifdef PFIL_HOOKS 234 struct packet_filter_hook *pfh; 235 struct mbuf *m0; 236 int rv; 237#endif /* PFIL_HOOKS */ 238 239next: 240 /* 241 * Get next datagram off input queue and get IP header 242 * in first mbuf. 243 */ 244 s = splimp(); 245 IF_DEQUEUE(&ipintrq, m); 246 splx(s); 247 if (m == 0) 248 return; 249#ifdef DIAGNOSTIC 250 if ((m->m_flags & M_PKTHDR) == 0) 251 panic("ipintr no HDR"); 252#endif 253 /* 254 * If no IP addresses have been set yet but the interfaces 255 * are receiving, can't do anything with incoming packets yet. 256 */ 257 if (in_ifaddr.tqh_first == 0) 258 goto bad; 259 ipstat.ips_total++; 260 if (m->m_len < sizeof (struct ip) && 261 (m = m_pullup(m, sizeof (struct ip))) == 0) { 262 ipstat.ips_toosmall++; 263 goto next; 264 } 265 ip = mtod(m, struct ip *); 266 if (ip->ip_v != IPVERSION) { 267 ipstat.ips_badvers++; 268 goto bad; 269 } 270 hlen = ip->ip_hl << 2; 271 if (hlen < sizeof(struct ip)) { /* minimum header length */ 272 ipstat.ips_badhlen++; 273 goto bad; 274 } 275 if (hlen > m->m_len) { 276 if ((m = m_pullup(m, hlen)) == 0) { 277 ipstat.ips_badhlen++; 278 goto next; 279 } 280 ip = mtod(m, struct ip *); 281 } 282 if ((ip->ip_sum = in_cksum(m, hlen)) != 0) { 283 ipstat.ips_badsum++; 284 goto bad; 285 } 286 287 /* 288 * Convert fields to host representation. 289 */ 290 NTOHS(ip->ip_len); 291 NTOHS(ip->ip_id); 292 NTOHS(ip->ip_off); 293 len = ip->ip_len; 294 295 /* 296 * Check that the amount of data in the buffers 297 * is as at least much as the IP header would have us expect. 298 * Trim mbufs if longer than we expect. 299 * Drop packet if shorter than we expect. 300 */ 301 if (m->m_pkthdr.len < len) { 302 ipstat.ips_tooshort++; 303 goto bad; 304 } 305 if (m->m_pkthdr.len > len) { 306 if (m->m_len == m->m_pkthdr.len) { 307 m->m_len = len; 308 m->m_pkthdr.len = len; 309 } else 310 m_adj(m, len - m->m_pkthdr.len); 311 } 312 313 /* 314 * Assume that we can create a fast-forward IP flow entry 315 * based on this packet. 316 */ 317 m->m_flags |= M_CANFASTFWD; 318 319#ifdef PFIL_HOOKS 320 /* 321 * Run through list of hooks for input packets. If there are any 322 * filters which require that additional packets in the flow are 323 * not fast-forwarded, they must clear the M_CANFASTFWD flag. 324 * Note that filters must _never_ set this flag, as another filter 325 * in the list may have previously cleared it. 326 */ 327 m0 = m; 328 for (pfh = pfil_hook_get(PFIL_IN); pfh; pfh = pfh->pfil_link.tqe_next) 329 if (pfh->pfil_func) { 330 rv = pfh->pfil_func(ip, hlen, m->m_pkthdr.rcvif, 0, &m0); 331 if (rv) 332 goto next; 333 m = m0; 334 if (m == NULL) 335 goto next; 336 ip = mtod(m, struct ip *); 337 } 338#endif /* PFIL_HOOKS */ 339 340 /* 341 * Process options and, if not destined for us, 342 * ship it on. ip_dooptions returns 1 when an 343 * error was detected (causing an icmp message 344 * to be sent and the original packet to be freed). 345 */ 346 ip_nhops = 0; /* for source routed packets */ 347 if (hlen > sizeof (struct ip) && ip_dooptions(m)) 348 goto next; 349 350 /* 351 * Check our list of addresses, to see if the packet is for us. 352 */ 353 INADDR_TO_IA(ip->ip_dst, ia); 354 if (ia != NULL) goto ours; 355 if (m->m_pkthdr.rcvif->if_flags & IFF_BROADCAST) { 356 for (ifa = m->m_pkthdr.rcvif->if_addrlist.tqh_first; 357 ifa != NULL; ifa = ifa->ifa_list.tqe_next) { 358 if (ifa->ifa_addr->sa_family != AF_INET) continue; 359 ia = ifatoia(ifa); 360 if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) || 361 in_hosteq(ip->ip_dst, ia->ia_netbroadcast) || 362 /* 363 * Look for all-0's host part (old broadcast addr), 364 * either for subnet or net. 365 */ 366 ip->ip_dst.s_addr == ia->ia_subnet || 367 ip->ip_dst.s_addr == ia->ia_net) 368 goto ours; 369 /* 370 * An interface with IP address zero accepts 371 * all packets that arrive on that interface. 372 */ 373 if (in_nullhost(ia->ia_addr.sin_addr)) 374 goto ours; 375 } 376 } 377 if (IN_MULTICAST(ip->ip_dst.s_addr)) { 378 struct in_multi *inm; 379#ifdef MROUTING 380 extern struct socket *ip_mrouter; 381 382 if (m->m_flags & M_EXT) { 383 if ((m = m_pullup(m, hlen)) == 0) { 384 ipstat.ips_toosmall++; 385 goto next; 386 } 387 ip = mtod(m, struct ip *); 388 } 389 390 if (ip_mrouter) { 391 /* 392 * If we are acting as a multicast router, all 393 * incoming multicast packets are passed to the 394 * kernel-level multicast forwarding function. 395 * The packet is returned (relatively) intact; if 396 * ip_mforward() returns a non-zero value, the packet 397 * must be discarded, else it may be accepted below. 398 * 399 * (The IP ident field is put in the same byte order 400 * as expected when ip_mforward() is called from 401 * ip_output().) 402 */ 403 ip->ip_id = htons(ip->ip_id); 404 if (ip_mforward(m, m->m_pkthdr.rcvif) != 0) { 405 ipstat.ips_cantforward++; 406 m_freem(m); 407 goto next; 408 } 409 ip->ip_id = ntohs(ip->ip_id); 410 411 /* 412 * The process-level routing demon needs to receive 413 * all multicast IGMP packets, whether or not this 414 * host belongs to their destination groups. 415 */ 416 if (ip->ip_p == IPPROTO_IGMP) 417 goto ours; 418 ipstat.ips_forward++; 419 } 420#endif 421 /* 422 * See if we belong to the destination multicast group on the 423 * arrival interface. 424 */ 425 IN_LOOKUP_MULTI(ip->ip_dst, m->m_pkthdr.rcvif, inm); 426 if (inm == NULL) { 427 ipstat.ips_cantforward++; 428 m_freem(m); 429 goto next; 430 } 431 goto ours; 432 } 433 if (ip->ip_dst.s_addr == INADDR_BROADCAST || 434 in_nullhost(ip->ip_dst)) 435 goto ours; 436 437 /* 438 * Not for us; forward if possible and desirable. 439 */ 440 if (ipforwarding == 0) { 441 ipstat.ips_cantforward++; 442 m_freem(m); 443 } else 444 ip_forward(m, 0); 445 goto next; 446 447ours: 448 /* 449 * If offset or IP_MF are set, must reassemble. 450 * Otherwise, nothing need be done. 451 * (We could look in the reassembly queue to see 452 * if the packet was previously fragmented, 453 * but it's not worth the time; just let them time out.) 454 */ 455 if (ip->ip_off & ~(IP_DF|IP_RF)) { 456 /* 457 * Look for queue of fragments 458 * of this datagram. 459 */ 460 for (fp = ipq.lh_first; fp != NULL; fp = fp->ipq_q.le_next) 461 if (ip->ip_id == fp->ipq_id && 462 in_hosteq(ip->ip_src, fp->ipq_src) && 463 in_hosteq(ip->ip_dst, fp->ipq_dst) && 464 ip->ip_p == fp->ipq_p) 465 goto found; 466 fp = 0; 467found: 468 469 /* 470 * Adjust ip_len to not reflect header, 471 * set ipqe_mff if more fragments are expected, 472 * convert offset of this to bytes. 473 */ 474 ip->ip_len -= hlen; 475 mff = (ip->ip_off & IP_MF) != 0; 476 if (mff) { 477 /* 478 * Make sure that fragments have a data length 479 * that's a non-zero multiple of 8 bytes. 480 */ 481 if (ip->ip_len == 0 || (ip->ip_len & 0x7) != 0) { 482 ipstat.ips_badfrags++; 483 goto bad; 484 } 485 } 486 ip->ip_off <<= 3; 487 488 /* 489 * If datagram marked as having more fragments 490 * or if this is not the first fragment, 491 * attempt reassembly; if it succeeds, proceed. 492 */ 493 if (mff || ip->ip_off) { 494 ipstat.ips_fragments++; 495 MALLOC(ipqe, struct ipqent *, sizeof (struct ipqent), 496 M_IPQ, M_NOWAIT); 497 if (ipqe == NULL) { 498 ipstat.ips_rcvmemdrop++; 499 goto bad; 500 } 501 ipqe->ipqe_mff = mff; 502 ipqe->ipqe_m = m; 503 ipqe->ipqe_ip = ip; 504 m = ip_reass(ipqe, fp); 505 if (m == 0) 506 goto next; 507 ipstat.ips_reassembled++; 508 ip = mtod(m, struct ip *); 509 } else 510 if (fp) 511 ip_freef(fp); 512 } else 513 ip->ip_len -= hlen; 514 515 /* 516 * Switch out to protocol's input routine. 517 */ 518 ipstat.ips_delivered++; 519 (*inetsw[ip_protox[ip->ip_p]].pr_input)(m, hlen); 520 goto next; 521bad: 522 m_freem(m); 523 goto next; 524} 525 526/* 527 * Take incoming datagram fragment and try to 528 * reassemble it into whole datagram. If a chain for 529 * reassembly of this datagram already exists, then it 530 * is given as fp; otherwise have to make a chain. 531 */ 532struct mbuf * 533ip_reass(ipqe, fp) 534 register struct ipqent *ipqe; 535 register struct ipq *fp; 536{ 537 register struct mbuf *m = ipqe->ipqe_m; 538 register struct ipqent *nq, *p, *q; 539 struct ip *ip; 540 struct mbuf *t; 541 int hlen = ipqe->ipqe_ip->ip_hl << 2; 542 int i, next; 543 544 /* 545 * Presence of header sizes in mbufs 546 * would confuse code below. 547 */ 548 m->m_data += hlen; 549 m->m_len -= hlen; 550 551 /* 552 * If first fragment to arrive, create a reassembly queue. 553 */ 554 if (fp == 0) { 555 MALLOC(fp, struct ipq *, sizeof (struct ipq), 556 M_FTABLE, M_NOWAIT); 557 if (fp == NULL) 558 goto dropfrag; 559 LIST_INSERT_HEAD(&ipq, fp, ipq_q); 560 fp->ipq_ttl = IPFRAGTTL; 561 fp->ipq_p = ipqe->ipqe_ip->ip_p; 562 fp->ipq_id = ipqe->ipqe_ip->ip_id; 563 LIST_INIT(&fp->ipq_fragq); 564 fp->ipq_src = ipqe->ipqe_ip->ip_src; 565 fp->ipq_dst = ipqe->ipqe_ip->ip_dst; 566 p = NULL; 567 goto insert; 568 } 569 570 /* 571 * Find a segment which begins after this one does. 572 */ 573 for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; 574 p = q, q = q->ipqe_q.le_next) 575 if (q->ipqe_ip->ip_off > ipqe->ipqe_ip->ip_off) 576 break; 577 578 /* 579 * If there is a preceding segment, it may provide some of 580 * our data already. If so, drop the data from the incoming 581 * segment. If it provides all of our data, drop us. 582 */ 583 if (p != NULL) { 584 i = p->ipqe_ip->ip_off + p->ipqe_ip->ip_len - 585 ipqe->ipqe_ip->ip_off; 586 if (i > 0) { 587 if (i >= ipqe->ipqe_ip->ip_len) 588 goto dropfrag; 589 m_adj(ipqe->ipqe_m, i); 590 ipqe->ipqe_ip->ip_off += i; 591 ipqe->ipqe_ip->ip_len -= i; 592 } 593 } 594 595 /* 596 * While we overlap succeeding segments trim them or, 597 * if they are completely covered, dequeue them. 598 */ 599 for (; q != NULL && ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len > 600 q->ipqe_ip->ip_off; q = nq) { 601 i = (ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len) - 602 q->ipqe_ip->ip_off; 603 if (i < q->ipqe_ip->ip_len) { 604 q->ipqe_ip->ip_len -= i; 605 q->ipqe_ip->ip_off += i; 606 m_adj(q->ipqe_m, i); 607 break; 608 } 609 nq = q->ipqe_q.le_next; 610 m_freem(q->ipqe_m); 611 LIST_REMOVE(q, ipqe_q); 612 FREE(q, M_IPQ); 613 } 614 615insert: 616 /* 617 * Stick new segment in its place; 618 * check for complete reassembly. 619 */ 620 if (p == NULL) { 621 LIST_INSERT_HEAD(&fp->ipq_fragq, ipqe, ipqe_q); 622 } else { 623 LIST_INSERT_AFTER(p, ipqe, ipqe_q); 624 } 625 next = 0; 626 for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; 627 p = q, q = q->ipqe_q.le_next) { 628 if (q->ipqe_ip->ip_off != next) 629 return (0); 630 next += q->ipqe_ip->ip_len; 631 } 632 if (p->ipqe_mff) 633 return (0); 634 635 /* 636 * Reassembly is complete. Check for a bogus message size and 637 * concatenate fragments. 638 */ 639 q = fp->ipq_fragq.lh_first; 640 ip = q->ipqe_ip; 641 if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) { 642 ipstat.ips_toolong++; 643 ip_freef(fp); 644 return (0); 645 } 646 m = q->ipqe_m; 647 t = m->m_next; 648 m->m_next = 0; 649 m_cat(m, t); 650 nq = q->ipqe_q.le_next; 651 FREE(q, M_IPQ); 652 for (q = nq; q != NULL; q = nq) { 653 t = q->ipqe_m; 654 nq = q->ipqe_q.le_next; 655 FREE(q, M_IPQ); 656 m_cat(m, t); 657 } 658 659 /* 660 * Create header for new ip packet by 661 * modifying header of first packet; 662 * dequeue and discard fragment reassembly header. 663 * Make header visible. 664 */ 665 ip->ip_len = next; 666 ip->ip_src = fp->ipq_src; 667 ip->ip_dst = fp->ipq_dst; 668 LIST_REMOVE(fp, ipq_q); 669 FREE(fp, M_FTABLE); 670 m->m_len += (ip->ip_hl << 2); 671 m->m_data -= (ip->ip_hl << 2); 672 /* some debugging cruft by sklower, below, will go away soon */ 673 if (m->m_flags & M_PKTHDR) { /* XXX this should be done elsewhere */ 674 register int plen = 0; 675 for (t = m; t; t = t->m_next) 676 plen += t->m_len; 677 m->m_pkthdr.len = plen; 678 } 679 return (m); 680 681dropfrag: 682 ipstat.ips_fragdropped++; 683 m_freem(m); 684 FREE(ipqe, M_IPQ); 685 return (0); 686} 687 688/* 689 * Free a fragment reassembly header and all 690 * associated datagrams. 691 */ 692void 693ip_freef(fp) 694 struct ipq *fp; 695{ 696 register struct ipqent *q, *p; 697 698 for (q = fp->ipq_fragq.lh_first; q != NULL; q = p) { 699 p = q->ipqe_q.le_next; 700 m_freem(q->ipqe_m); 701 LIST_REMOVE(q, ipqe_q); 702 FREE(q, M_IPQ); 703 } 704 LIST_REMOVE(fp, ipq_q); 705 FREE(fp, M_FTABLE); 706} 707 708/* 709 * IP timer processing; 710 * if a timer expires on a reassembly 711 * queue, discard it. 712 */ 713void 714ip_slowtimo() 715{ 716 register struct ipq *fp, *nfp; 717 int s = splsoftnet(); 718 719 for (fp = ipq.lh_first; fp != NULL; fp = nfp) { 720 nfp = fp->ipq_q.le_next; 721 if (--fp->ipq_ttl == 0) { 722 ipstat.ips_fragtimeout++; 723 ip_freef(fp); 724 } 725 } 726#ifdef GATEWAY 727 ipflow_slowtimo(); 728#endif 729 splx(s); 730} 731 732/* 733 * Drain off all datagram fragments. 734 */ 735void 736ip_drain() 737{ 738 739 while (ipq.lh_first != NULL) { 740 ipstat.ips_fragdropped++; 741 ip_freef(ipq.lh_first); 742 } 743} 744 745/* 746 * Do option processing on a datagram, 747 * possibly discarding it if bad options are encountered, 748 * or forwarding it if source-routed. 749 * Returns 1 if packet has been forwarded/freed, 750 * 0 if the packet should be processed further. 751 */ 752int 753ip_dooptions(m) 754 struct mbuf *m; 755{ 756 register struct ip *ip = mtod(m, struct ip *); 757 register u_char *cp; 758 register struct ip_timestamp *ipt; 759 register struct in_ifaddr *ia; 760 int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0; 761 struct in_addr *sin, dst; 762 n_time ntime; 763 764 dst = ip->ip_dst; 765 cp = (u_char *)(ip + 1); 766 cnt = (ip->ip_hl << 2) - sizeof (struct ip); 767 for (; cnt > 0; cnt -= optlen, cp += optlen) { 768 opt = cp[IPOPT_OPTVAL]; 769 if (opt == IPOPT_EOL) 770 break; 771 if (opt == IPOPT_NOP) 772 optlen = 1; 773 else { 774 optlen = cp[IPOPT_OLEN]; 775 if (optlen <= 0 || optlen > cnt) { 776 code = &cp[IPOPT_OLEN] - (u_char *)ip; 777 goto bad; 778 } 779 } 780 switch (opt) { 781 782 default: 783 break; 784 785 /* 786 * Source routing with record. 787 * Find interface with current destination address. 788 * If none on this machine then drop if strictly routed, 789 * or do nothing if loosely routed. 790 * Record interface address and bring up next address 791 * component. If strictly routed make sure next 792 * address is on directly accessible net. 793 */ 794 case IPOPT_LSRR: 795 case IPOPT_SSRR: 796 if (ip_allowsrcrt == 0) { 797 type = ICMP_UNREACH; 798 code = ICMP_UNREACH_NET_PROHIB; 799 goto bad; 800 } 801 if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) { 802 code = &cp[IPOPT_OFFSET] - (u_char *)ip; 803 goto bad; 804 } 805 ipaddr.sin_addr = ip->ip_dst; 806 ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))); 807 if (ia == 0) { 808 if (opt == IPOPT_SSRR) { 809 type = ICMP_UNREACH; 810 code = ICMP_UNREACH_SRCFAIL; 811 goto bad; 812 } 813 /* 814 * Loose routing, and not at next destination 815 * yet; nothing to do except forward. 816 */ 817 break; 818 } 819 off--; /* 0 origin */ 820 if (off > optlen - sizeof(struct in_addr)) { 821 /* 822 * End of source route. Should be for us. 823 */ 824 save_rte(cp, ip->ip_src); 825 break; 826 } 827 /* 828 * locate outgoing interface 829 */ 830 bcopy((caddr_t)(cp + off), (caddr_t)&ipaddr.sin_addr, 831 sizeof(ipaddr.sin_addr)); 832 if (opt == IPOPT_SSRR) { 833#define INA struct in_ifaddr * 834#define SA struct sockaddr * 835 ia = (INA)ifa_ifwithladdr((SA)&ipaddr); 836 } else 837 ia = ip_rtaddr(ipaddr.sin_addr); 838 if (ia == 0) { 839 type = ICMP_UNREACH; 840 code = ICMP_UNREACH_SRCFAIL; 841 goto bad; 842 } 843 ip->ip_dst = ipaddr.sin_addr; 844 bcopy((caddr_t)&ia->ia_addr.sin_addr, 845 (caddr_t)(cp + off), sizeof(struct in_addr)); 846 cp[IPOPT_OFFSET] += sizeof(struct in_addr); 847 /* 848 * Let ip_intr's mcast routing check handle mcast pkts 849 */ 850 forward = !IN_MULTICAST(ip->ip_dst.s_addr); 851 break; 852 853 case IPOPT_RR: 854 if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) { 855 code = &cp[IPOPT_OFFSET] - (u_char *)ip; 856 goto bad; 857 } 858 /* 859 * If no space remains, ignore. 860 */ 861 off--; /* 0 origin */ 862 if (off > optlen - sizeof(struct in_addr)) 863 break; 864 bcopy((caddr_t)(&ip->ip_dst), (caddr_t)&ipaddr.sin_addr, 865 sizeof(ipaddr.sin_addr)); 866 /* 867 * locate outgoing interface; if we're the destination, 868 * use the incoming interface (should be same). 869 */ 870 if ((ia = (INA)ifa_ifwithaddr((SA)&ipaddr)) == 0 && 871 (ia = ip_rtaddr(ipaddr.sin_addr)) == 0) { 872 type = ICMP_UNREACH; 873 code = ICMP_UNREACH_HOST; 874 goto bad; 875 } 876 bcopy((caddr_t)&ia->ia_addr.sin_addr, 877 (caddr_t)(cp + off), sizeof(struct in_addr)); 878 cp[IPOPT_OFFSET] += sizeof(struct in_addr); 879 break; 880 881 case IPOPT_TS: 882 code = cp - (u_char *)ip; 883 ipt = (struct ip_timestamp *)cp; 884 if (ipt->ipt_len < 5) 885 goto bad; 886 if (ipt->ipt_ptr > ipt->ipt_len - sizeof (int32_t)) { 887 if (++ipt->ipt_oflw == 0) 888 goto bad; 889 break; 890 } 891 sin = (struct in_addr *)(cp + ipt->ipt_ptr - 1); 892 switch (ipt->ipt_flg) { 893 894 case IPOPT_TS_TSONLY: 895 break; 896 897 case IPOPT_TS_TSANDADDR: 898 if (ipt->ipt_ptr - 1 + sizeof(n_time) + 899 sizeof(struct in_addr) > ipt->ipt_len) 900 goto bad; 901 ipaddr.sin_addr = dst; 902 ia = (INA)ifaof_ifpforaddr((SA)&ipaddr, 903 m->m_pkthdr.rcvif); 904 if (ia == 0) 905 continue; 906 bcopy((caddr_t)&ia->ia_addr.sin_addr, 907 (caddr_t)sin, sizeof(struct in_addr)); 908 ipt->ipt_ptr += sizeof(struct in_addr); 909 break; 910 911 case IPOPT_TS_PRESPEC: 912 if (ipt->ipt_ptr - 1 + sizeof(n_time) + 913 sizeof(struct in_addr) > ipt->ipt_len) 914 goto bad; 915 bcopy((caddr_t)sin, (caddr_t)&ipaddr.sin_addr, 916 sizeof(struct in_addr)); 917 if (ifa_ifwithaddr((SA)&ipaddr) == 0) 918 continue; 919 ipt->ipt_ptr += sizeof(struct in_addr); 920 break; 921 922 default: 923 goto bad; 924 } 925 ntime = iptime(); 926 bcopy((caddr_t)&ntime, (caddr_t)cp + ipt->ipt_ptr - 1, 927 sizeof(n_time)); 928 ipt->ipt_ptr += sizeof(n_time); 929 } 930 } 931 if (forward) { 932 if (ip_forwsrcrt == 0) { 933 type = ICMP_UNREACH; 934 code = ICMP_UNREACH_SRCFAIL; 935 goto bad; 936 } 937 ip_forward(m, 1); 938 return (1); 939 } 940 return (0); 941bad: 942 ip->ip_len -= ip->ip_hl << 2; /* XXX icmp_error adds in hdr length */ 943 icmp_error(m, type, code, 0, 0); 944 ipstat.ips_badoptions++; 945 return (1); 946} 947 948/* 949 * Given address of next destination (final or next hop), 950 * return internet address info of interface to be used to get there. 951 */ 952struct in_ifaddr * 953ip_rtaddr(dst) 954 struct in_addr dst; 955{ 956 register struct sockaddr_in *sin; 957 958 sin = satosin(&ipforward_rt.ro_dst); 959 960 if (ipforward_rt.ro_rt == 0 || !in_hosteq(dst, sin->sin_addr)) { 961 if (ipforward_rt.ro_rt) { 962 RTFREE(ipforward_rt.ro_rt); 963 ipforward_rt.ro_rt = 0; 964 } 965 sin->sin_family = AF_INET; 966 sin->sin_len = sizeof(*sin); 967 sin->sin_addr = dst; 968 969 rtalloc(&ipforward_rt); 970 } 971 if (ipforward_rt.ro_rt == 0) 972 return ((struct in_ifaddr *)0); 973 return (ifatoia(ipforward_rt.ro_rt->rt_ifa)); 974} 975 976/* 977 * Save incoming source route for use in replies, 978 * to be picked up later by ip_srcroute if the receiver is interested. 979 */ 980void 981save_rte(option, dst) 982 u_char *option; 983 struct in_addr dst; 984{ 985 unsigned olen; 986 987 olen = option[IPOPT_OLEN]; 988#ifdef DIAGNOSTIC 989 if (ipprintfs) 990 printf("save_rte: olen %d\n", olen); 991#endif 992 if (olen > sizeof(ip_srcrt) - (1 + sizeof(dst))) 993 return; 994 bcopy((caddr_t)option, (caddr_t)ip_srcrt.srcopt, olen); 995 ip_nhops = (olen - IPOPT_OFFSET - 1) / sizeof(struct in_addr); 996 ip_srcrt.dst = dst; 997} 998 999/* 1000 * Retrieve incoming source route for use in replies, 1001 * in the same form used by setsockopt. 1002 * The first hop is placed before the options, will be removed later. 1003 */ 1004struct mbuf * 1005ip_srcroute() 1006{ 1007 register struct in_addr *p, *q; 1008 register struct mbuf *m; 1009 1010 if (ip_nhops == 0) 1011 return ((struct mbuf *)0); 1012 m = m_get(M_DONTWAIT, MT_SOOPTS); 1013 if (m == 0) 1014 return ((struct mbuf *)0); 1015 1016#define OPTSIZ (sizeof(ip_srcrt.nop) + sizeof(ip_srcrt.srcopt)) 1017 1018 /* length is (nhops+1)*sizeof(addr) + sizeof(nop + srcrt header) */ 1019 m->m_len = ip_nhops * sizeof(struct in_addr) + sizeof(struct in_addr) + 1020 OPTSIZ; 1021#ifdef DIAGNOSTIC 1022 if (ipprintfs) 1023 printf("ip_srcroute: nhops %d mlen %d", ip_nhops, m->m_len); 1024#endif 1025 1026 /* 1027 * First save first hop for return route 1028 */ 1029 p = &ip_srcrt.route[ip_nhops - 1]; 1030 *(mtod(m, struct in_addr *)) = *p--; 1031#ifdef DIAGNOSTIC 1032 if (ipprintfs) 1033 printf(" hops %x", ntohl(mtod(m, struct in_addr *)->s_addr)); 1034#endif 1035 1036 /* 1037 * Copy option fields and padding (nop) to mbuf. 1038 */ 1039 ip_srcrt.nop = IPOPT_NOP; 1040 ip_srcrt.srcopt[IPOPT_OFFSET] = IPOPT_MINOFF; 1041 bcopy((caddr_t)&ip_srcrt.nop, 1042 mtod(m, caddr_t) + sizeof(struct in_addr), OPTSIZ); 1043 q = (struct in_addr *)(mtod(m, caddr_t) + 1044 sizeof(struct in_addr) + OPTSIZ); 1045#undef OPTSIZ 1046 /* 1047 * Record return path as an IP source route, 1048 * reversing the path (pointers are now aligned). 1049 */ 1050 while (p >= ip_srcrt.route) { 1051#ifdef DIAGNOSTIC 1052 if (ipprintfs) 1053 printf(" %x", ntohl(q->s_addr)); 1054#endif 1055 *q++ = *p--; 1056 } 1057 /* 1058 * Last hop goes to final destination. 1059 */ 1060 *q = ip_srcrt.dst; 1061#ifdef DIAGNOSTIC 1062 if (ipprintfs) 1063 printf(" %x\n", ntohl(q->s_addr)); 1064#endif 1065 return (m); 1066} 1067 1068/* 1069 * Strip out IP options, at higher 1070 * level protocol in the kernel. 1071 * Second argument is buffer to which options 1072 * will be moved, and return value is their length. 1073 * XXX should be deleted; last arg currently ignored. 1074 */ 1075void 1076ip_stripoptions(m, mopt) 1077 register struct mbuf *m; 1078 struct mbuf *mopt; 1079{ 1080 register int i; 1081 struct ip *ip = mtod(m, struct ip *); 1082 register caddr_t opts; 1083 int olen; 1084 1085 olen = (ip->ip_hl<<2) - sizeof (struct ip); 1086 opts = (caddr_t)(ip + 1); 1087 i = m->m_len - (sizeof (struct ip) + olen); 1088 bcopy(opts + olen, opts, (unsigned)i); 1089 m->m_len -= olen; 1090 if (m->m_flags & M_PKTHDR) 1091 m->m_pkthdr.len -= olen; 1092 ip->ip_hl = sizeof(struct ip) >> 2; 1093} 1094 1095int inetctlerrmap[PRC_NCMDS] = { 1096 0, 0, 0, 0, 1097 0, EMSGSIZE, EHOSTDOWN, EHOSTUNREACH, 1098 EHOSTUNREACH, EHOSTUNREACH, ECONNREFUSED, ECONNREFUSED, 1099 EMSGSIZE, EHOSTUNREACH, 0, 0, 1100 0, 0, 0, 0, 1101 ENOPROTOOPT 1102}; 1103 1104/* 1105 * Forward a packet. If some error occurs return the sender 1106 * an icmp packet. Note we can't always generate a meaningful 1107 * icmp message because icmp doesn't have a large enough repertoire 1108 * of codes and types. 1109 * 1110 * If not forwarding, just drop the packet. This could be confusing 1111 * if ipforwarding was zero but some routing protocol was advancing 1112 * us as a gateway to somewhere. However, we must let the routing 1113 * protocol deal with that. 1114 * 1115 * The srcrt parameter indicates whether the packet is being forwarded 1116 * via a source route. 1117 */ 1118void 1119ip_forward(m, srcrt) 1120 struct mbuf *m; 1121 int srcrt; 1122{ 1123 register struct ip *ip = mtod(m, struct ip *); 1124 register struct sockaddr_in *sin; 1125 register struct rtentry *rt; 1126 int error, type = 0, code = 0; 1127 struct mbuf *mcopy; 1128 n_long dest; 1129 struct ifnet *destifp; 1130 1131 dest = 0; 1132#ifdef DIAGNOSTIC 1133 if (ipprintfs) 1134 printf("forward: src %x dst %x ttl %x\n", 1135 ip->ip_src.s_addr, ip->ip_dst.s_addr, ip->ip_ttl); 1136#endif 1137 if (m->m_flags & M_BCAST || in_canforward(ip->ip_dst) == 0) { 1138 ipstat.ips_cantforward++; 1139 m_freem(m); 1140 return; 1141 } 1142 HTONS(ip->ip_id); 1143 if (ip->ip_ttl <= IPTTLDEC) { 1144 icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0); 1145 return; 1146 } 1147 ip->ip_ttl -= IPTTLDEC; 1148 1149 sin = satosin(&ipforward_rt.ro_dst); 1150 if ((rt = ipforward_rt.ro_rt) == 0 || 1151 !in_hosteq(ip->ip_dst, sin->sin_addr)) { 1152 if (ipforward_rt.ro_rt) { 1153 RTFREE(ipforward_rt.ro_rt); 1154 ipforward_rt.ro_rt = 0; 1155 } 1156 sin->sin_family = AF_INET; 1157 sin->sin_len = sizeof(struct sockaddr_in); 1158 sin->sin_addr = ip->ip_dst; 1159 1160 rtalloc(&ipforward_rt); 1161 if (ipforward_rt.ro_rt == 0) { 1162 icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, dest, 0); 1163 return; 1164 } 1165 rt = ipforward_rt.ro_rt; 1166 } 1167 1168 /* 1169 * Save at most 68 bytes of the packet in case 1170 * we need to generate an ICMP message to the src. 1171 */ 1172 mcopy = m_copy(m, 0, imin((int)ip->ip_len, 68)); 1173 1174 /* 1175 * If forwarding packet using same interface that it came in on, 1176 * perhaps should send a redirect to sender to shortcut a hop. 1177 * Only send redirect if source is sending directly to us, 1178 * and if packet was not source routed (or has any options). 1179 * Also, don't send redirect if forwarding using a default route 1180 * or a route modified by a redirect. 1181 */ 1182 if (rt->rt_ifp == m->m_pkthdr.rcvif && 1183 (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 && 1184 !in_nullhost(satosin(rt_key(rt))->sin_addr) && 1185 ipsendredirects && !srcrt) { 1186 if (rt->rt_ifa && 1187 (ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) == 1188 ifatoia(rt->rt_ifa)->ia_subnet) { 1189 if (rt->rt_flags & RTF_GATEWAY) 1190 dest = satosin(rt->rt_gateway)->sin_addr.s_addr; 1191 else 1192 dest = ip->ip_dst.s_addr; 1193 /* Router requirements says to only send host redirects */ 1194 type = ICMP_REDIRECT; 1195 code = ICMP_REDIRECT_HOST; 1196#ifdef DIAGNOSTIC 1197 if (ipprintfs) 1198 printf("redirect (%d) to %x\n", code, (u_int32_t)dest); 1199#endif 1200 } 1201 } 1202 1203 error = ip_output(m, (struct mbuf *)0, &ipforward_rt, 1204 (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)), 0); 1205 if (error) 1206 ipstat.ips_cantforward++; 1207 else { 1208 ipstat.ips_forward++; 1209 if (type) 1210 ipstat.ips_redirectsent++; 1211 else { 1212 if (mcopy) { 1213#ifdef GATEWAY 1214 if (mcopy->m_flags & M_CANFASTFWD) 1215 ipflow_create(&ipforward_rt, mcopy); 1216#endif 1217 m_freem(mcopy); 1218 } 1219 return; 1220 } 1221 } 1222 if (mcopy == NULL) 1223 return; 1224 destifp = NULL; 1225 1226 switch (error) { 1227 1228 case 0: /* forwarded, but need redirect */ 1229 /* type, code set above */ 1230 break; 1231 1232 case ENETUNREACH: /* shouldn't happen, checked above */ 1233 case EHOSTUNREACH: 1234 case ENETDOWN: 1235 case EHOSTDOWN: 1236 default: 1237 type = ICMP_UNREACH; 1238 code = ICMP_UNREACH_HOST; 1239 break; 1240 1241 case EMSGSIZE: 1242 type = ICMP_UNREACH; 1243 code = ICMP_UNREACH_NEEDFRAG; 1244 if (ipforward_rt.ro_rt) 1245 destifp = ipforward_rt.ro_rt->rt_ifp; 1246 ipstat.ips_cantfrag++; 1247 break; 1248 1249 case ENOBUFS: 1250 type = ICMP_SOURCEQUENCH; 1251 code = 0; 1252 break; 1253 } 1254 icmp_error(mcopy, type, code, dest, destifp); 1255} 1256 1257void 1258ip_savecontrol(inp, mp, ip, m) 1259 register struct inpcb *inp; 1260 register struct mbuf **mp; 1261 register struct ip *ip; 1262 register struct mbuf *m; 1263{ 1264 1265 if (inp->inp_socket->so_options & SO_TIMESTAMP) { 1266 struct timeval tv; 1267 1268 microtime(&tv); 1269 *mp = sbcreatecontrol((caddr_t) &tv, sizeof(tv), 1270 SCM_TIMESTAMP, SOL_SOCKET); 1271 if (*mp) 1272 mp = &(*mp)->m_next; 1273 } 1274 if (inp->inp_flags & INP_RECVDSTADDR) { 1275 *mp = sbcreatecontrol((caddr_t) &ip->ip_dst, 1276 sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP); 1277 if (*mp) 1278 mp = &(*mp)->m_next; 1279 } 1280#ifdef notyet 1281 /* 1282 * XXX 1283 * Moving these out of udp_input() made them even more broken 1284 * than they already were. 1285 * - fenner@parc.xerox.com 1286 */ 1287 /* options were tossed already */ 1288 if (inp->inp_flags & INP_RECVOPTS) { 1289 *mp = sbcreatecontrol((caddr_t) opts_deleted_above, 1290 sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP); 1291 if (*mp) 1292 mp = &(*mp)->m_next; 1293 } 1294 /* ip_srcroute doesn't do what we want here, need to fix */ 1295 if (inp->inp_flags & INP_RECVRETOPTS) { 1296 *mp = sbcreatecontrol((caddr_t) ip_srcroute(), 1297 sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP); 1298 if (*mp) 1299 mp = &(*mp)->m_next; 1300 } 1301#endif 1302 if (inp->inp_flags & INP_RECVIF) { 1303 struct sockaddr_dl sdl; 1304 1305 sdl.sdl_len = offsetof(struct sockaddr_dl, sdl_data[0]); 1306 sdl.sdl_family = AF_LINK; 1307 sdl.sdl_index = m->m_pkthdr.rcvif ? 1308 m->m_pkthdr.rcvif->if_index : 0; 1309 sdl.sdl_nlen = sdl.sdl_alen = sdl.sdl_slen = 0; 1310 *mp = sbcreatecontrol((caddr_t) &sdl, sdl.sdl_len, 1311 IP_RECVIF, IPPROTO_IP); 1312 if (*mp) 1313 mp = &(*mp)->m_next; 1314 } 1315} 1316 1317int 1318ip_sysctl(name, namelen, oldp, oldlenp, newp, newlen) 1319 int *name; 1320 u_int namelen; 1321 void *oldp; 1322 size_t *oldlenp; 1323 void *newp; 1324 size_t newlen; 1325{ 1326 extern int subnetsarelocal; 1327 1328 int error, old; 1329 1330 /* All sysctl names at this level are terminal. */ 1331 if (namelen != 1) 1332 return (ENOTDIR); 1333 1334 switch (name[0]) { 1335 case IPCTL_FORWARDING: 1336 return (sysctl_int(oldp, oldlenp, newp, newlen, &ipforwarding)); 1337 case IPCTL_SENDREDIRECTS: 1338 return (sysctl_int(oldp, oldlenp, newp, newlen, 1339 &ipsendredirects)); 1340 case IPCTL_DEFTTL: 1341 return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_defttl)); 1342#ifdef notyet 1343 case IPCTL_DEFMTU: 1344 return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_mtu)); 1345#endif 1346 case IPCTL_FORWSRCRT: 1347 /* Don't allow this to change in a secure environment. */ 1348 if (securelevel > 0) 1349 return (sysctl_rdint(oldp, oldlenp, newp, 1350 ip_forwsrcrt)); 1351 else 1352 return (sysctl_int(oldp, oldlenp, newp, newlen, 1353 &ip_forwsrcrt)); 1354 case IPCTL_DIRECTEDBCAST: 1355 return (sysctl_int(oldp, oldlenp, newp, newlen, 1356 &ip_directedbcast)); 1357 case IPCTL_ALLOWSRCRT: 1358 return (sysctl_int(oldp, oldlenp, newp, newlen, 1359 &ip_allowsrcrt)); 1360 case IPCTL_SUBNETSARELOCAL: 1361 return (sysctl_int(oldp, oldlenp, newp, newlen, 1362 &subnetsarelocal)); 1363 case IPCTL_MTUDISC: 1364 error = sysctl_int(oldp, oldlenp, newp, newlen, 1365 &ip_mtudisc); 1366 if (ip_mtudisc != 0 && ip_mtudisc_timeout_q == NULL) { 1367 ip_mtudisc_timeout_q = 1368 rt_timer_queue_create(ip_mtudisc_timeout); 1369 } else if (ip_mtudisc == 0 && ip_mtudisc_timeout_q != NULL) { 1370 rt_timer_queue_destroy(ip_mtudisc_timeout_q, TRUE); 1371 ip_mtudisc_timeout_q = NULL; 1372 } 1373 return error; 1374 case IPCTL_ANONPORTMIN: 1375 old = anonportmin; 1376 error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmin); 1377 if (anonportmin >= anonportmax || anonportmin > 65535 1378#ifndef IPNOPRIVPORTS 1379 || anonportmin < IPPORT_RESERVED 1380#endif 1381 ) { 1382 anonportmin = old; 1383 return (EINVAL); 1384 } 1385 return (error); 1386 case IPCTL_ANONPORTMAX: 1387 old = anonportmax; 1388 error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmax); 1389 if (anonportmin >= anonportmax || anonportmax > 65535 1390#ifndef IPNOPRIVPORTS 1391 || anonportmax < IPPORT_RESERVED 1392#endif 1393 ) { 1394 anonportmax = old; 1395 return (EINVAL); 1396 } 1397 return (error); 1398 case IPCTL_MTUDISCTIMEOUT: 1399 error = sysctl_int(oldp, oldlenp, newp, newlen, 1400 &ip_mtudisc_timeout); 1401 if (ip_mtudisc_timeout_q != NULL) 1402 rt_timer_queue_change(ip_mtudisc_timeout_q, 1403 ip_mtudisc_timeout); 1404 return (error); 1405#ifdef GATEWAY 1406 case IPCTL_MAXFLOWS: 1407 { 1408 int s; 1409 1410 error = sysctl_int(oldp, oldlenp, newp, newlen, 1411 &ip_maxflows); 1412 s = splsoftnet(); 1413 ipflow_reap(0); 1414 splx(s); 1415 return (error); 1416 } 1417#endif 1418 default: 1419 return (EOPNOTSUPP); 1420 } 1421 /* NOTREACHED */ 1422} 1423