bsd/netinet6/ip6_forward.c

   1 /*
   2  * Copyright (c) 2009-2018 Apple Inc. All rights reserved.
   3  *
   4  * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
   5  *
   6  * This file contains Original Code and/or Modifications of Original Code
   7  * as defined in and that are subject to the Apple Public Source License
   8  * Version 2.0 (the 'License'). You may not use this file except in
   9  * compliance with the License. The rights granted to you under the License
  10  * may not be used to create, or enable the creation or redistribution of,
  11  * unlawful or unlicensed copies of an Apple operating system, or to
  12  * circumvent, violate, or enable the circumvention or violation of, any
  13  * terms of an Apple operating system software license agreement.
  14  *
  15  * Please obtain a copy of the License at
  16  * http://www.opensource.apple.com/apsl/ and read it before using this file.
  17  *
  18  * The Original Code and all software distributed under the License are
  19  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  20  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  21  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  22  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  23  * Please see the License for the specific language governing rights and
  24  * limitations under the License.
  25  *
  26  * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  27  */
  28
  29 /*      $FreeBSD: src/sys/netinet6/ip6_forward.c,v 1.16 2002/10/16 02:25:05 sam Exp $   */
  30 /*      $KAME: ip6_forward.c,v 1.69 2001/05/17 03:48:30 itojun Exp $    */
  31
  32 /*
  33  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
  34  * All rights reserved.
  35  *
  36  * Redistribution and use in source and binary forms, with or without
  37  * modification, are permitted provided that the following conditions
  38  * are met:
  39  * 1. Redistributions of source code must retain the above copyright
  40  *    notice, this list of conditions and the following disclaimer.
  41  * 2. Redistributions in binary form must reproduce the above copyright
  42  *    notice, this list of conditions and the following disclaimer in the
  43  *    documentation and/or other materials provided with the distribution.
  44  * 3. Neither the name of the project nor the names of its contributors
  45  *    may be used to endorse or promote products derived from this software
  46  *    without specific prior written permission.
  47  *
  48  * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
  49  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  50  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  51  * ARE DISCLAIMED.  IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
  52  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  53  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  54  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  55  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  56  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  57  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  58  * SUCH DAMAGE.
  59  */
  60
  61
  62 #include <sys/param.h>
  63 #include <sys/systm.h>
  64 #include <sys/malloc.h>
  65 #include <sys/mbuf.h>
  66 #include <sys/domain.h>
  67 #include <sys/protosw.h>
  68 #include <sys/socket.h>
  69 #include <sys/errno.h>
  70 #include <sys/time.h>
  71 #include <sys/kernel.h>
  72 #include <sys/syslog.h>
  73
  74 #include <net/if.h>
  75 #include <net/route.h>
  76
  77 #include <netinet/in.h>
  78 #include <netinet/in_var.h>
  79 #include <netinet/in_systm.h>
  80 #include <netinet/ip.h>
  81 #include <netinet/ip_var.h>
  82 #include <netinet6/in6_var.h>
  83 #include <netinet/ip6.h>
  84 #include <netinet6/ip6_var.h>
  85 #include <netinet/icmp6.h>
  86 #include <netinet6/nd6.h>
  87 #include <netinet6/scope6_var.h>
  88
  89 #include <netinet/in_pcb.h>
  90
  91 #if IPSEC
  92 #include <netinet6/ipsec.h>
  93 #if INET6
  94 #include <netinet6/ipsec6.h>
  95 #endif
  96 #include <netkey/key.h>
  97 extern int ipsec_bypass;
  98 #endif /* IPSEC */
  99
 100 #include <net/net_osdep.h>
 101
 102 #if DUMMYNET
 103 #include <netinet/ip_fw.h>
 104 #include <netinet/ip_dummynet.h>
 105 #endif /* DUMMYNET */
 106
 107 #if PF
 108 #include <net/pfvar.h>
 109 #endif /* PF */
 110
 111 /*
 112  * Forward a packet.  If some error occurs return the sender
 113  * an icmp packet.  Note we can't always generate a meaningful
 114  * icmp message because icmp doesn't have a large enough repertoire
 115  * of codes and types.
 116  *
 117  * If not forwarding, just drop the packet.  This could be confusing
 118  * if ipforwarding was zero but some routing protocol was advancing
 119  * us as a gateway to somewhere.  However, we must let the routing
 120  * protocol deal with that.
 121  *
 122  */
 123
 124 struct mbuf *
 125 ip6_forward(struct mbuf *m, struct route_in6 *ip6forward_rt,
 126     int srcrt)
 127 {
 128         struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
 129         struct sockaddr_in6 *dst;
 130         struct rtentry *rt;
 131         int error, type = 0, code = 0;
 132         boolean_t proxy = FALSE;
 133         struct mbuf *mcopy = NULL;
 134         struct ifnet *ifp, *rcvifp, *origifp;   /* maybe unnecessary */
 135         u_int32_t inzone, outzone, len = 0, pktcnt = 0;
 136         struct in6_addr src_in6, dst_in6;
 137         uint64_t curtime = net_uptime();
 138 #if IPSEC
 139         struct secpolicy *sp = NULL;
 140 #endif
 141         unsigned int ifscope = IFSCOPE_NONE;
 142 #if PF
 143         struct pf_mtag *pf_mtag;
 144         struct pf_fragment_tag *pf_ftagp, pf_ftag;
 145         boolean_t pf_ftag_valid = FALSE;
 146 #endif /* PF */
 147         uint32_t mpktlen = 0;
 148
 149         /*
 150          * In the prefix proxying case, the route to the proxied node normally
 151          * gets created by nd6_prproxy_ns_output(), as part of forwarding a
 152          * NS (NUD/AR) packet to the proxied node.  In the event that such
 153          * packet did not arrive in time before the correct route gets created,
 154          * ip6_input() would have performed a rtalloc() which most likely will
 155          * create the wrong cloned route; this route points back to the same
 156          * interface as the inbound interface, since the parent non-scoped
 157          * prefix route points there.  Therefore we check if that is the case
 158          * and perform the necessary fixup to get the correct route installed.
 159          */
 160         if (!srcrt && nd6_prproxy &&
 161             (rt = ip6forward_rt->ro_rt) != NULL && (rt->rt_flags & RTF_PROXY)) {
 162                 nd6_proxy_find_fwdroute(m->m_pkthdr.rcvif, ip6forward_rt);
 163                 if ((rt = ip6forward_rt->ro_rt) != NULL) {
 164                         ifscope = rt->rt_ifp->if_index;
 165                 }
 166         }
 167
 168 #if PF
 169         pf_mtag = pf_find_mtag(m);
 170         /*
 171          * save the PF fragmentation metadata as m_copy() removes the
 172          * mbufs tags from the original mbuf.
 173          */
 174         pf_ftagp = pf_find_fragment_tag(m);
 175         if (pf_ftagp != NULL) {
 176                 ASSERT(pf_mtag->pftag_flags & PF_TAG_REASSEMBLED);
 177                 pf_ftag = *pf_ftagp;
 178                 pf_ftag_valid = TRUE;
 179                 mpktlen = pf_ftag.ft_maxlen;
 180                 ASSERT(mpktlen);
 181         }
 182         if (pf_mtag != NULL && pf_mtag->pftag_rtableid != IFSCOPE_NONE) {
 183                 ifscope = pf_mtag->pftag_rtableid;
 184         }
 185         pf_mtag = NULL;
 186         pf_ftagp = NULL;
 187         /*
 188          * If the caller provides a route which is on a different interface
 189          * than the one specified for scoped forwarding, discard the route
 190          * and do a lookup below.
 191          */
 192         if (ifscope != IFSCOPE_NONE && (rt = ip6forward_rt->ro_rt) != NULL) {
 193                 RT_LOCK(rt);
 194                 if (rt->rt_ifp->if_index != ifscope) {
 195                         RT_UNLOCK(rt);
 196                         ROUTE_RELEASE(ip6forward_rt);
 197                         rt = NULL;
 198                 } else {
 199                         RT_UNLOCK(rt);
 200                 }
 201         }
 202 #endif /* PF */
 203
 204 #if IPSEC
 205         /*
 206          * Check AH/ESP integrity.
 207          */
 208         /*
 209          * Don't increment ip6s_cantforward because this is the check
 210          * before forwarding packet actually.
 211          */
 212         if (ipsec_bypass == 0) {
 213                 if (ipsec6_in_reject(m, NULL)) {
 214                         IPSEC_STAT_INCREMENT(ipsec6stat.in_polvio);
 215                         m_freem(m);
 216                         return NULL;
 217                 }
 218         }
 219 #endif /*IPSEC*/
 220
 221         /*
 222          * Do not forward packets to multicast destination.
 223          * Do not forward packets with unspecified source.  It was discussed
 224          * in July 2000, on ipngwg mailing list.
 225          */
 226         if ((m->m_flags & (M_BCAST | M_MCAST)) != 0 ||
 227             IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst) ||
 228             IN6_IS_ADDR_UNSPECIFIED(&ip6->ip6_src)) {
 229                 ip6stat.ip6s_cantforward++;
 230                 /* XXX in6_ifstat_inc(rt->rt_ifp, ifs6_in_discard) */
 231                 if (ip6_log_time + ip6_log_interval < curtime) {
 232                         ip6_log_time = curtime;
 233                         log(LOG_DEBUG,
 234                             "cannot forward "
 235                             "from %s to %s nxt %d received on %s\n",
 236                             ip6_sprintf(&ip6->ip6_src),
 237                             ip6_sprintf(&ip6->ip6_dst),
 238                             ip6->ip6_nxt,
 239                             if_name(m->m_pkthdr.rcvif));
 240                 }
 241                 m_freem(m);
 242                 return NULL;
 243         }
 244
 245         if (ip6->ip6_hlim <= IPV6_HLIMDEC) {
 246                 /* XXX in6_ifstat_inc(rt->rt_ifp, ifs6_in_discard) */
 247                 icmp6_error_flag(m, ICMP6_TIME_EXCEEDED,
 248                     ICMP6_TIME_EXCEED_TRANSIT, 0, 0);
 249                 return NULL;
 250         }
 251
 252         /*
 253          * See if the destination is a proxied address, and if so pretend
 254          * that it's for us.  This is mostly to handle NUD probes against
 255          * the proxied addresses.  We filter for ICMPv6 here and will let
 256          * icmp6_input handle the rest.
 257          */
 258         if (!srcrt && nd6_prproxy) {
 259                 VERIFY(!IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst));
 260                 proxy = nd6_prproxy_isours(m, ip6, ip6forward_rt, ifscope);
 261                 /*
 262                  * Don't update hop limit while proxying; RFC 4389 4.1.
 263                  * Also skip IPsec forwarding path processing as this
 264                  * packet is not to be forwarded.
 265                  */
 266                 if (proxy) {
 267                         goto skip_ipsec;
 268                 }
 269         }
 270
 271         ip6->ip6_hlim -= IPV6_HLIMDEC;
 272
 273         /*
 274          * Save at most ICMPV6_PLD_MAXLEN (= the min IPv6 MTU -
 275          * size of IPv6 + ICMPv6 headers) bytes of the packet in case
 276          * we need to generate an ICMP6 message to the src.
 277          * Thanks to M_EXT, in most cases copy will not occur.
 278          *
 279          * It is important to save it before IPsec processing as IPsec
 280          * processing may modify the mbuf.
 281          */
 282         mcopy = m_copy(m, 0, imin(m->m_pkthdr.len, ICMPV6_PLD_MAXLEN));
 283
 284 #if IPSEC
 285         if (ipsec_bypass != 0) {
 286                 goto skip_ipsec;
 287         }
 288         /* get a security policy for this packet */
 289         sp = ipsec6_getpolicybyaddr(m, IPSEC_DIR_OUTBOUND, IP_FORWARDING,
 290             &error);
 291         if (sp == NULL) {
 292                 IPSEC_STAT_INCREMENT(ipsec6stat.out_inval);
 293                 ip6stat.ip6s_cantforward++;
 294                 if (mcopy) {
 295 #if 0
 296                         /* XXX: what icmp ? */
 297 #else
 298                         m_freem(mcopy);
 299 #endif
 300                 }
 301                 m_freem(m);
 302                 return NULL;
 303         }
 304
 305         error = 0;
 306
 307         /* check policy */
 308         switch (sp->policy) {
 309         case IPSEC_POLICY_DISCARD:
 310         case IPSEC_POLICY_GENERATE:
 311                 /*
 312                  * This packet is just discarded.
 313                  */
 314                 IPSEC_STAT_INCREMENT(ipsec6stat.out_polvio);
 315                 ip6stat.ip6s_cantforward++;
 316                 key_freesp(sp, KEY_SADB_UNLOCKED);
 317                 if (mcopy) {
 318 #if 0
 319                         /* XXX: what icmp ? */
 320 #else
 321                         m_freem(mcopy);
 322 #endif
 323                 }
 324                 m_freem(m);
 325                 return NULL;
 326
 327         case IPSEC_POLICY_BYPASS:
 328         case IPSEC_POLICY_NONE:
 329                 /* no need to do IPsec. */
 330                 key_freesp(sp, KEY_SADB_UNLOCKED);
 331                 goto skip_ipsec;
 332
 333         case IPSEC_POLICY_IPSEC:
 334                 if (sp->req == NULL) {
 335                         /* XXX should be panic ? */
 336                         printf("ip6_forward: No IPsec request specified.\n");
 337                         ip6stat.ip6s_cantforward++;
 338                         key_freesp(sp, KEY_SADB_UNLOCKED);
 339                         if (mcopy) {
 340 #if 0
 341                                 /* XXX: what icmp ? */
 342 #else
 343                                 m_freem(mcopy);
 344 #endif
 345                         }
 346                         m_freem(m);
 347                         return NULL;
 348                 }
 349                 /* do IPsec */
 350                 break;
 351
 352         case IPSEC_POLICY_ENTRUST:
 353         default:
 354                 /* should be panic ?? */
 355                 printf("ip6_forward: Invalid policy found. %d\n", sp->policy);
 356                 key_freesp(sp, KEY_SADB_UNLOCKED);
 357                 goto skip_ipsec;
 358         }
 359
 360         {
 361                 struct ipsec_output_state state;
 362
 363                 /*
 364                  * All the extension headers will become inaccessible
 365                  * (since they can be encrypted).
 366                  * Don't panic, we need no more updates to extension headers
 367                  * on inner IPv6 packet (since they are now encapsulated).
 368                  *
 369                  * IPv6 [ESP|AH] IPv6 [extension headers] payload
 370                  */
 371                 bzero(&state, sizeof(state));
 372                 state.m = m;
 373                 state.dst = NULL; /* update at ipsec6_output_tunnel() */
 374
 375                 error = ipsec6_output_tunnel(&state, sp, 0);
 376                 key_freesp(sp, KEY_SADB_UNLOCKED);
 377                 if (state.tunneled == 4) {
 378                         ROUTE_RELEASE(&state.ro);
 379                         return NULL; /* packet is gone - sent over IPv4 */
 380                 }
 381
 382                 m = state.m;
 383                 ROUTE_RELEASE(&state.ro);
 384
 385                 if (error) {
 386                         /* mbuf is already reclaimed in ipsec6_output_tunnel. */
 387                         switch (error) {
 388                         case EHOSTUNREACH:
 389                         case ENETUNREACH:
 390                         case EMSGSIZE:
 391                         case ENOBUFS:
 392                         case ENOMEM:
 393                                 break;
 394                         default:
 395                                 printf("ip6_output (ipsec): error code %d\n", error);
 396                         /* fall through */
 397                         case ENOENT:
 398                                 /* don't show these error codes to the user */
 399                                 break;
 400                         }
 401                         ip6stat.ip6s_cantforward++;
 402                         if (mcopy) {
 403 #if 0
 404                                 /* XXX: what icmp ? */
 405 #else
 406                                 m_freem(mcopy);
 407 #endif
 408                         }
 409                         m_freem(m);
 410                         return NULL;
 411                 }
 412         }
 413 #endif /* IPSEC */
 414 skip_ipsec:
 415
 416         dst = (struct sockaddr_in6 *)&ip6forward_rt->ro_dst;
 417         if ((rt = ip6forward_rt->ro_rt) != NULL) {
 418                 RT_LOCK(rt);
 419                 /* Take an extra ref for ourselves */
 420                 RT_ADDREF_LOCKED(rt);
 421         }
 422
 423         VERIFY(rt == NULL || rt == ip6forward_rt->ro_rt);
 424         if (!srcrt) {
 425                 /*
 426                  * ip6forward_rt->ro_dst.sin6_addr is equal to ip6->ip6_dst
 427                  */
 428                 if (ROUTE_UNUSABLE(ip6forward_rt)) {
 429                         if (rt != NULL) {
 430                                 /* Release extra ref */
 431                                 RT_REMREF_LOCKED(rt);
 432                                 RT_UNLOCK(rt);
 433                         }
 434                         ROUTE_RELEASE(ip6forward_rt);
 435
 436                         /* this probably fails but give it a try again */
 437                         rtalloc_scoped_ign((struct route *)ip6forward_rt,
 438                             RTF_PRCLONING, ifscope);
 439                         if ((rt = ip6forward_rt->ro_rt) != NULL) {
 440                                 RT_LOCK(rt);
 441                                 /* Take an extra ref for ourselves */
 442                                 RT_ADDREF_LOCKED(rt);
 443                         }
 444                 }
 445
 446                 if (rt == NULL) {
 447                         ip6stat.ip6s_noroute++;
 448                         in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_noroute);
 449                         if (mcopy) {
 450                                 icmp6_error(mcopy, ICMP6_DST_UNREACH,
 451                                     ICMP6_DST_UNREACH_NOROUTE, 0);
 452                         }
 453                         m_freem(m);
 454                         return NULL;
 455                 }
 456                 RT_LOCK_ASSERT_HELD(rt);
 457         } else if (ROUTE_UNUSABLE(ip6forward_rt) ||
 458             !IN6_ARE_ADDR_EQUAL(&ip6->ip6_dst, &dst->sin6_addr)) {
 459                 if (rt != NULL) {
 460                         /* Release extra ref */
 461                         RT_REMREF_LOCKED(rt);
 462                         RT_UNLOCK(rt);
 463                 }
 464                 ROUTE_RELEASE(ip6forward_rt);
 465
 466                 bzero(dst, sizeof(*dst));
 467                 dst->sin6_len = sizeof(struct sockaddr_in6);
 468                 dst->sin6_family = AF_INET6;
 469                 dst->sin6_addr = ip6->ip6_dst;
 470
 471                 rtalloc_scoped_ign((struct route *)ip6forward_rt,
 472                     RTF_PRCLONING, ifscope);
 473                 if ((rt = ip6forward_rt->ro_rt) == NULL) {
 474                         ip6stat.ip6s_noroute++;
 475                         in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_noroute);
 476                         if (mcopy) {
 477                                 icmp6_error(mcopy, ICMP6_DST_UNREACH,
 478                                     ICMP6_DST_UNREACH_NOROUTE, 0);
 479                         }
 480                         m_freem(m);
 481                         return NULL;
 482                 }
 483                 RT_LOCK(rt);
 484                 /* Take an extra ref for ourselves */
 485                 RT_ADDREF_LOCKED(rt);
 486         }
 487
 488         /*
 489          * Source scope check: if a packet can't be delivered to its
 490          * destination for the reason that the destination is beyond the scope
 491          * of the source address, discard the packet and return an icmp6
 492          * destination unreachable error with Code 2 (beyond scope of source
 493          * address) unless we are proxying (source address is link local
 494          * for NUDs.)  We use a local copy of ip6_src, since in6_setscope()
 495          * will possibly modify its first argument.
 496          * [draft-ietf-ipngwg-icmp-v3-04.txt, Section 3.1]
 497          */
 498         src_in6 = ip6->ip6_src;
 499         if (in6_setscope(&src_in6, rt->rt_ifp, &outzone)) {
 500                 RT_REMREF_LOCKED(rt);
 501                 RT_UNLOCK(rt);
 502                 /* XXX: this should not happen */
 503                 ip6stat.ip6s_cantforward++;
 504                 ip6stat.ip6s_badscope++;
 505                 m_freem(m);
 506                 return NULL;
 507         }
 508         if (in6_setscope(&src_in6, m->m_pkthdr.rcvif, &inzone)) {
 509                 RT_REMREF_LOCKED(rt);
 510                 RT_UNLOCK(rt);
 511                 ip6stat.ip6s_cantforward++;
 512                 ip6stat.ip6s_badscope++;
 513                 m_freem(m);
 514                 return NULL;
 515         }
 516
 517         if (inzone != outzone && !proxy) {
 518                 ip6stat.ip6s_cantforward++;
 519                 ip6stat.ip6s_badscope++;
 520                 in6_ifstat_inc(rt->rt_ifp, ifs6_in_discard);
 521
 522                 if (ip6_log_time + ip6_log_interval < curtime) {
 523                         ip6_log_time = curtime;
 524                         log(LOG_DEBUG,
 525                             "cannot forward "
 526                             "src %s, dst %s, nxt %d, rcvif %s, outif %s\n",
 527                             ip6_sprintf(&ip6->ip6_src),
 528                             ip6_sprintf(&ip6->ip6_dst),
 529                             ip6->ip6_nxt,
 530                             if_name(m->m_pkthdr.rcvif), if_name(rt->rt_ifp));
 531                 }
 532                 /* Release extra ref */
 533                 RT_REMREF_LOCKED(rt);
 534                 RT_UNLOCK(rt);
 535                 if (mcopy) {
 536                         icmp6_error(mcopy, ICMP6_DST_UNREACH,
 537                             ICMP6_DST_UNREACH_BEYONDSCOPE, 0);
 538                 }
 539                 m_freem(m);
 540                 return NULL;
 541         }
 542
 543         /*
 544          * Destination scope check: if a packet is going to break the scope
 545          * zone of packet's destination address, discard it.  This case should
 546          * usually be prevented by appropriately-configured routing table, but
 547          * we need an explicit check because we may mistakenly forward the
 548          * packet to a different zone by (e.g.) a default route.
 549          */
 550         dst_in6 = ip6->ip6_dst;
 551         if (in6_setscope(&dst_in6, m->m_pkthdr.rcvif, &inzone) != 0 ||
 552             in6_setscope(&dst_in6, rt->rt_ifp, &outzone) != 0 ||
 553             inzone != outzone) {
 554                 RT_REMREF_LOCKED(rt);
 555                 RT_UNLOCK(rt);
 556                 ip6stat.ip6s_cantforward++;
 557                 ip6stat.ip6s_badscope++;
 558                 m_freem(m);
 559                 return NULL;
 560         }
 561
 562         if (mpktlen == 0) {
 563                 mpktlen = m->m_pkthdr.len;
 564         }
 565
 566         if (mpktlen > rt->rt_ifp->if_mtu) {
 567                 in6_ifstat_inc(rt->rt_ifp, ifs6_in_toobig);
 568                 if (mcopy) {
 569                         uint32_t mtu;
 570 #if IPSEC
 571                         struct secpolicy *sp2;
 572                         int ipsecerror;
 573                         size_t ipsechdrsiz;
 574 #endif
 575
 576                         mtu = rt->rt_ifp->if_mtu;
 577 #if IPSEC
 578                         /*
 579                          * When we do IPsec tunnel ingress, we need to play
 580                          * with the link value (decrement IPsec header size
 581                          * from mtu value).  The code is much simpler than v4
 582                          * case, as we have the outgoing interface for
 583                          * encapsulated packet as "rt->rt_ifp".
 584                          */
 585                         sp2 = ipsec6_getpolicybyaddr(mcopy, IPSEC_DIR_OUTBOUND,
 586                             IP_FORWARDING, &ipsecerror);
 587                         if (sp2) {
 588                                 ipsechdrsiz = ipsec6_hdrsiz(mcopy,
 589                                     IPSEC_DIR_OUTBOUND, NULL);
 590                                 if (ipsechdrsiz < mtu) {
 591                                         mtu -= ipsechdrsiz;
 592                                 }
 593                                 key_freesp(sp2, KEY_SADB_UNLOCKED);
 594                         }
 595                         /*
 596                          * if mtu becomes less than minimum MTU,
 597                          * tell minimum MTU (and I'll need to fragment it).
 598                          */
 599                         if (mtu < IPV6_MMTU) {
 600                                 mtu = IPV6_MMTU;
 601                         }
 602 #endif
 603                         /* Release extra ref */
 604                         RT_REMREF_LOCKED(rt);
 605                         RT_UNLOCK(rt);
 606                         icmp6_error(mcopy, ICMP6_PACKET_TOO_BIG, 0, mtu);
 607                 } else {
 608                         /* Release extra ref */
 609                         RT_REMREF_LOCKED(rt);
 610                         RT_UNLOCK(rt);
 611                 }
 612                 m_freem(m);
 613                 return NULL;
 614         }
 615
 616         if (rt->rt_flags & RTF_GATEWAY) {
 617                 dst = (struct sockaddr_in6 *)(void *)rt->rt_gateway;
 618         }
 619
 620         /*
 621          * If we are to forward the packet using the same interface
 622          * as one we got the packet from, perhaps we should send a redirect
 623          * to sender to shortcut a hop.
 624          * Only send redirect if source is sending directly to us,
 625          * and if packet was not source routed (or has any options).
 626          * Also, don't send redirect if forwarding using a route
 627          * modified by a redirect.
 628          */
 629         if (!proxy &&
 630             ip6_sendredirects && rt->rt_ifp == m->m_pkthdr.rcvif && !srcrt &&
 631             (rt->rt_flags & (RTF_DYNAMIC | RTF_MODIFIED)) == 0) {
 632                 if ((rt->rt_ifp->if_flags & IFF_POINTOPOINT) != 0) {
 633                         /*
 634                          * If the incoming interface is equal to the outgoing
 635                          * one, and the link attached to the interface is
 636                          * point-to-point, then it will be highly probable
 637                          * that a routing loop occurs. Thus, we immediately
 638                          * drop the packet and send an ICMPv6 error message.
 639                          *
 640                          * type/code is based on suggestion by Rich Draves.
 641                          * not sure if it is the best pick.
 642                          */
 643                         RT_REMREF_LOCKED(rt);   /* Release extra ref */
 644                         RT_UNLOCK(rt);
 645                         icmp6_error(mcopy, ICMP6_DST_UNREACH,
 646                             ICMP6_DST_UNREACH_ADDR, 0);
 647                         m_freem(m);
 648                         return NULL;
 649                 }
 650                 type = ND_REDIRECT;
 651         }
 652         /*
 653          * Fake scoped addresses. Note that even link-local source or
 654          * destinaion can appear, if the originating node just sends the
 655          * packet to us (without address resolution for the destination).
 656          * Since both icmp6_error and icmp6_redirect_output fill the embedded
 657          * link identifiers, we can do this stuff after making a copy for
 658          * returning an error.
 659          */
 660         if ((rt->rt_ifp->if_flags & IFF_LOOPBACK) != 0) {
 661                 /*
 662                  * See corresponding comments in ip6_output.
 663                  * XXX: but is it possible that ip6_forward() sends a packet
 664                  *      to a loopback interface? I don't think so, and thus
 665                  *      I bark here. (jinmei@kame.net)
 666                  * XXX: it is common to route invalid packets to loopback.
 667                  *      also, the codepath will be visited on use of ::1 in
 668                  *      rthdr. (itojun)
 669                  */
 670 #if 1
 671                 if ((0))
 672 #else
 673                 if ((rt->rt_flags & (RTF_BLACKHOLE | RTF_REJECT)) == 0)
 674 #endif
 675                 {
 676                         printf("ip6_forward: outgoing interface is loopback. "
 677                             "src %s, dst %s, nxt %d, rcvif %s, outif %s\n",
 678                             ip6_sprintf(&ip6->ip6_src),
 679                             ip6_sprintf(&ip6->ip6_dst),
 680                             ip6->ip6_nxt, if_name(m->m_pkthdr.rcvif),
 681                             if_name(rt->rt_ifp));
 682                 }
 683
 684                 /* we can just use rcvif in forwarding. */
 685                 origifp = rcvifp = m->m_pkthdr.rcvif;
 686         } else if (nd6_prproxy) {
 687                 /*
 688                  * In the prefix proxying case, we need to inform nd6_output()
 689                  * about the inbound interface, so that any subsequent NS
 690                  * packets generated by nd6_prproxy_ns_output() will not be
 691                  * sent back to that same interface.
 692                  */
 693                 origifp = rcvifp = m->m_pkthdr.rcvif;
 694         } else {
 695                 rcvifp = m->m_pkthdr.rcvif;
 696                 origifp = rt->rt_ifp;
 697         }
 698         /*
 699          * clear embedded scope identifiers if necessary.
 700          * in6_clearscope will touch the addresses only when necessary.
 701          */
 702         in6_clearscope(&ip6->ip6_src);
 703         in6_clearscope(&ip6->ip6_dst);
 704
 705         ifp = rt->rt_ifp;
 706         /* Drop the lock but retain the extra ref */
 707         RT_UNLOCK(rt);
 708
 709         /*
 710          * If this is to be processed locally, let ip6_input have it.
 711          */
 712         if (proxy) {
 713                 VERIFY(m->m_pkthdr.pkt_flags & PKTF_PROXY_DST);
 714                 /* Release extra ref */
 715                 RT_REMREF(rt);
 716                 if (mcopy != NULL) {
 717                         m_freem(mcopy);
 718                 }
 719                 return m;
 720         }
 721
 722         /* Mark this packet as being forwarded from another interface */
 723         m->m_pkthdr.pkt_flags |= PKTF_FORWARDED;
 724
 725 #if PF
 726         if (PF_IS_ENABLED) {
 727                 /*
 728                  * PF refragments any packet which it reassembled due to scrub
 729                  * rules, in which case it will set the PF_TAG_REFRAGMENTED
 730                  * flag in PF mbuf tag.
 731                  */
 732                 if (pf_ftag_valid) {
 733                         pf_copy_fragment_tag(m, &pf_ftag, M_DONTWAIT);
 734                 }
 735 #if DUMMYNET
 736                 struct ip_fw_args args;
 737                 bzero(&args, sizeof(args));
 738
 739                 args.fwa_m = m;
 740                 args.fwa_oif = ifp;
 741                 args.fwa_oflags = 0;
 742                 args.fwa_ro6 = ip6forward_rt;
 743                 args.fwa_ro6_pmtu = ip6forward_rt;
 744                 args.fwa_mtu = rt->rt_ifp->if_mtu;
 745                 args.fwa_dst6 = dst;
 746                 args.fwa_origifp = origifp;
 747                 /* Invoke outbound packet filter */
 748                 error = pf_af_hook(ifp, NULL, &m, AF_INET6, FALSE, &args);
 749 #else /* !DUMMYNET */
 750                 error = pf_af_hook(ifp, NULL, &m, AF_INET6, FALSE, NULL);
 751 #endif /* !DUMMYNET */
 752                 if (error != 0 || m == NULL) {
 753                         if (m != NULL) {
 754                                 panic("%s: unexpected packet %p\n", __func__, m);
 755                                 /* NOTREACHED */
 756                         }
 757                         /* Already freed by callee */
 758                         goto senderr;
 759                 }
 760
 761                 pf_mtag = pf_find_mtag(m);
 762                 /*
 763                  * refragmented packets from PF.
 764                  */
 765                 if ((pf_mtag->pftag_flags & PF_TAG_REFRAGMENTED) != 0) {
 766                         struct mbuf *t;
 767
 768                         pf_mtag->pftag_flags &= ~PF_TAG_REFRAGMENTED;
 769                         /* for statistics */
 770                         t = m;
 771                         while (t != NULL) {
 772                                 pktcnt++;
 773                                 len += m_pktlen(t);
 774                                 t = t->m_nextpkt;
 775                         }
 776
 777                         /*
 778                          * nd6_output() frees packetchain in both success and
 779                          * failure cases.
 780                          */
 781                         error = nd6_output(ifp, origifp, m, dst, rt, NULL);
 782                         m = NULL;
 783                         goto sent;
 784                 }
 785                 /*
 786                  * We do not use ip6 header again in the code below,
 787                  * however still adding the bit here so that any new
 788                  * code in future doesn't end up working with the
 789                  * wrong pointer
 790                  */
 791                 ip6 = mtod(m, struct ip6_hdr *);
 792         }
 793 #endif /* PF */
 794
 795         len = m_pktlen(m);
 796         pktcnt = 1;
 797         error = nd6_output(ifp, origifp, m, dst, rt, NULL);
 798 sent:
 799         if (error) {
 800                 in6_ifstat_add(ifp, ifs6_out_discard, pktcnt);
 801                 ip6stat.ip6s_cantforward += pktcnt;
 802         } else {
 803                 /*
 804                  * Increment stats on the source interface; the ones
 805                  * for destination interface has been taken care of
 806                  * during output above by virtue of PKTF_FORWARDED.
 807                  */
 808                 rcvifp->if_fpackets += pktcnt;
 809                 rcvifp->if_fbytes += len;
 810
 811                 ip6stat.ip6s_forward += pktcnt;
 812                 in6_ifstat_add(ifp, ifs6_out_forward, pktcnt);
 813                 if (type) {
 814                         ip6stat.ip6s_redirectsent++;
 815                 } else {
 816                         if (mcopy) {
 817                                 goto freecopy;
 818                         }
 819                 }
 820         }
 821 #if PF
 822 senderr:
 823 #endif /* PF */
 824         if (mcopy == NULL) {
 825                 /* Release extra ref */
 826                 RT_REMREF(rt);
 827                 return NULL;
 828         }
 829         switch (error) {
 830         case 0:
 831 #if 1
 832                 if (type == ND_REDIRECT) {
 833                         icmp6_redirect_output(mcopy, rt);
 834                         /* Release extra ref */
 835                         RT_REMREF(rt);
 836                         return NULL;
 837                 }
 838 #endif
 839                 goto freecopy;
 840
 841         case EMSGSIZE:
 842                 /* xxx MTU is constant in PPP? */
 843                 goto freecopy;
 844
 845         case ENOBUFS:
 846                 /* Tell source to slow down like source quench in IP? */
 847                 goto freecopy;
 848
 849         case ENETUNREACH:       /* shouldn't happen, checked above */
 850         case EHOSTUNREACH:
 851         case ENETDOWN:
 852         case EHOSTDOWN:
 853         default:
 854                 type = ICMP6_DST_UNREACH;
 855                 code = ICMP6_DST_UNREACH_ADDR;
 856                 break;
 857         }
 858         icmp6_error(mcopy, type, code, 0);
 859         /* Release extra ref */
 860         RT_REMREF(rt);
 861         return NULL;
 862
 863 freecopy:
 864         m_freem(mcopy);
 865         /* Release extra ref */
 866         RT_REMREF(rt);
 867         return NULL;
 868 }