ip_output.c revision 1.258
1 /* $NetBSD: ip_output.c,v 1.258 2016/06/21 03:28:27 ozaki-r 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 * 49 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 50 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 51 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 52 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 53 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 54 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 55 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 56 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 57 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 58 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 59 * POSSIBILITY OF SUCH DAMAGE. 60 */ 61 62 /* 63 * Copyright (c) 1982, 1986, 1988, 1990, 1993 64 * The Regents of the University of California. All rights reserved. 65 * 66 * Redistribution and use in source and binary forms, with or without 67 * modification, are permitted provided that the following conditions 68 * are met: 69 * 1. Redistributions of source code must retain the above copyright 70 * notice, this list of conditions and the following disclaimer. 71 * 2. Redistributions in binary form must reproduce the above copyright 72 * notice, this list of conditions and the following disclaimer in the 73 * documentation and/or other materials provided with the distribution. 74 * 3. Neither the name of the University nor the names of its contributors 75 * may be used to endorse or promote products derived from this software 76 * without specific prior written permission. 77 * 78 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 79 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 80 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 81 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 82 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 83 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 84 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 85 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 86 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 87 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 88 * SUCH DAMAGE. 89 * 90 * @(#)ip_output.c 8.3 (Berkeley) 1/21/94 91 */ 92 93 #include <sys/cdefs.h> 94 __KERNEL_RCSID(0, "$NetBSD: ip_output.c,v 1.258 2016/06/21 03:28:27 ozaki-r Exp $"); 95 96 #ifdef _KERNEL_OPT 97 #include "opt_inet.h" 98 #include "opt_ipsec.h" 99 #include "opt_mrouting.h" 100 #include "opt_net_mpsafe.h" 101 #include "opt_mpls.h" 102 #endif 103 104 #include <sys/param.h> 105 #include <sys/kmem.h> 106 #include <sys/mbuf.h> 107 #include <sys/protosw.h> 108 #include <sys/socket.h> 109 #include <sys/socketvar.h> 110 #include <sys/kauth.h> 111 #ifdef IPSEC 112 #include <sys/domain.h> 113 #endif 114 #include <sys/systm.h> 115 116 #include <net/if.h> 117 #include <net/if_types.h> 118 #include <net/route.h> 119 #include <net/pfil.h> 120 121 #include <netinet/in.h> 122 #include <netinet/in_systm.h> 123 #include <netinet/ip.h> 124 #include <netinet/in_pcb.h> 125 #include <netinet/in_var.h> 126 #include <netinet/ip_var.h> 127 #include <netinet/ip_private.h> 128 #include <netinet/in_offload.h> 129 #include <netinet/portalgo.h> 130 #include <netinet/udp.h> 131 132 #ifdef INET6 133 #include <netinet6/ip6_var.h> 134 #endif 135 136 #ifdef MROUTING 137 #include <netinet/ip_mroute.h> 138 #endif 139 140 #ifdef IPSEC 141 #include <netipsec/ipsec.h> 142 #include <netipsec/key.h> 143 #endif 144 145 #ifdef MPLS 146 #include <netmpls/mpls.h> 147 #include <netmpls/mpls_var.h> 148 #endif 149 150 static int ip_pcbopts(struct inpcb *, const struct sockopt *); 151 static struct mbuf *ip_insertoptions(struct mbuf *, struct mbuf *, int *); 152 static struct ifnet *ip_multicast_if(struct in_addr *, int *); 153 static void ip_mloopback(struct ifnet *, struct mbuf *, 154 const struct sockaddr_in *); 155 156 extern pfil_head_t *inet_pfil_hook; /* XXX */ 157 158 int ip_do_loopback_cksum = 0; 159 160 static int 161 ip_mark_mpls(struct ifnet * const ifp, struct mbuf * const m, 162 const struct rtentry *rt) 163 { 164 int error = 0; 165 #ifdef MPLS 166 union mpls_shim msh; 167 168 if (rt == NULL || rt_gettag(rt) == NULL || 169 rt_gettag(rt)->sa_family != AF_MPLS || 170 (m->m_flags & (M_MCAST | M_BCAST)) != 0 || 171 ifp->if_type != IFT_ETHER) 172 return 0; 173 174 msh.s_addr = MPLS_GETSADDR(rt); 175 if (msh.shim.label != MPLS_LABEL_IMPLNULL) { 176 struct m_tag *mtag; 177 /* 178 * XXX tentative solution to tell ether_output 179 * it's MPLS. Need some more efficient solution. 180 */ 181 mtag = m_tag_get(PACKET_TAG_MPLS, 182 sizeof(int) /* dummy */, 183 M_NOWAIT); 184 if (mtag == NULL) 185 return ENOMEM; 186 m_tag_prepend(m, mtag); 187 } 188 #endif 189 return error; 190 } 191 192 /* 193 * Send an IP packet to a host. 194 */ 195 int 196 ip_if_output(struct ifnet * const ifp, struct mbuf * const m, 197 const struct sockaddr * const dst, const struct rtentry *rt) 198 { 199 int error = 0; 200 201 if (rt != NULL) { 202 error = rt_check_reject_route(rt, ifp); 203 if (error != 0) { 204 m_freem(m); 205 return error; 206 } 207 } 208 209 error = ip_mark_mpls(ifp, m, rt); 210 if (error != 0) { 211 m_freem(m); 212 return error; 213 } 214 215 error = if_output_lock(ifp, ifp, m, dst, rt); 216 217 return error; 218 } 219 220 /* 221 * IP output. The packet in mbuf chain m contains a skeletal IP 222 * header (with len, off, ttl, proto, tos, src, dst). 223 * The mbuf chain containing the packet will be freed. 224 * The mbuf opt, if present, will not be freed. 225 */ 226 int 227 ip_output(struct mbuf *m0, struct mbuf *opt, struct route *ro, int flags, 228 struct ip_moptions *imo, struct socket *so) 229 { 230 struct rtentry *rt; 231 struct ip *ip; 232 struct ifnet *ifp, *mifp = NULL; 233 struct mbuf *m = m0; 234 int hlen = sizeof (struct ip); 235 int len, error = 0; 236 struct route iproute; 237 const struct sockaddr_in *dst; 238 struct in_ifaddr *ia; 239 int isbroadcast; 240 int sw_csum; 241 u_long mtu; 242 #ifdef IPSEC 243 struct secpolicy *sp = NULL; 244 #endif 245 bool natt_frag = false; 246 bool rtmtu_nolock; 247 union { 248 struct sockaddr dst; 249 struct sockaddr_in dst4; 250 } u; 251 struct sockaddr *rdst = &u.dst; /* real IP destination, as opposed 252 * to the nexthop 253 */ 254 struct psref psref; 255 int bound; 256 257 len = 0; 258 259 MCLAIM(m, &ip_tx_mowner); 260 261 KASSERT((m->m_flags & M_PKTHDR) != 0); 262 KASSERT((m->m_pkthdr.csum_flags & (M_CSUM_TCPv6|M_CSUM_UDPv6)) == 0); 263 KASSERT((m->m_pkthdr.csum_flags & (M_CSUM_TCPv4|M_CSUM_UDPv4)) != 264 (M_CSUM_TCPv4|M_CSUM_UDPv4)); 265 266 if (opt) { 267 m = ip_insertoptions(m, opt, &len); 268 if (len >= sizeof(struct ip)) 269 hlen = len; 270 } 271 ip = mtod(m, struct ip *); 272 273 /* 274 * Fill in IP header. 275 */ 276 if ((flags & (IP_FORWARDING|IP_RAWOUTPUT)) == 0) { 277 ip->ip_v = IPVERSION; 278 ip->ip_off = htons(0); 279 /* ip->ip_id filled in after we find out source ia */ 280 ip->ip_hl = hlen >> 2; 281 IP_STATINC(IP_STAT_LOCALOUT); 282 } else { 283 hlen = ip->ip_hl << 2; 284 } 285 286 /* 287 * Route packet. 288 */ 289 if (ro == NULL) { 290 memset(&iproute, 0, sizeof(iproute)); 291 ro = &iproute; 292 } 293 sockaddr_in_init(&u.dst4, &ip->ip_dst, 0); 294 dst = satocsin(rtcache_getdst(ro)); 295 296 /* 297 * If there is a cached route, check that it is to the same 298 * destination and is still up. If not, free it and try again. 299 * The address family should also be checked in case of sharing 300 * the cache with IPv6. 301 */ 302 if (dst && (dst->sin_family != AF_INET || 303 !in_hosteq(dst->sin_addr, ip->ip_dst))) 304 rtcache_free(ro); 305 306 if ((rt = rtcache_validate(ro)) == NULL && 307 (rt = rtcache_update(ro, 1)) == NULL) { 308 dst = &u.dst4; 309 error = rtcache_setdst(ro, &u.dst); 310 if (error != 0) 311 goto bad; 312 } 313 314 /* 315 * If routing to interface only, short circuit routing lookup. 316 */ 317 if (flags & IP_ROUTETOIF) { 318 if ((ia = ifatoia(ifa_ifwithladdr(sintocsa(dst)))) == NULL) { 319 IP_STATINC(IP_STAT_NOROUTE); 320 error = ENETUNREACH; 321 goto bad; 322 } 323 ifp = ia->ia_ifp; 324 mtu = ifp->if_mtu; 325 ip->ip_ttl = 1; 326 isbroadcast = in_broadcast(dst->sin_addr, ifp); 327 } else if ((IN_MULTICAST(ip->ip_dst.s_addr) || 328 ip->ip_dst.s_addr == INADDR_BROADCAST) && 329 imo != NULL && imo->imo_multicast_if_index != 0) { 330 bound = curlwp_bind(); 331 ifp = mifp = if_get_byindex(imo->imo_multicast_if_index, &psref); 332 if (ifp == NULL) { 333 curlwp_bindx(bound); 334 IP_STATINC(IP_STAT_NOROUTE); 335 error = ENETUNREACH; 336 goto bad; 337 } 338 mtu = ifp->if_mtu; 339 IFP_TO_IA(ifp, ia); 340 isbroadcast = 0; 341 } else { 342 if (rt == NULL) 343 rt = rtcache_init(ro); 344 if (rt == NULL) { 345 IP_STATINC(IP_STAT_NOROUTE); 346 error = EHOSTUNREACH; 347 goto bad; 348 } 349 ia = ifatoia(rt->rt_ifa); 350 ifp = rt->rt_ifp; 351 if ((mtu = rt->rt_rmx.rmx_mtu) == 0) 352 mtu = ifp->if_mtu; 353 rt->rt_use++; 354 if (rt->rt_flags & RTF_GATEWAY) 355 dst = satosin(rt->rt_gateway); 356 if (rt->rt_flags & RTF_HOST) 357 isbroadcast = rt->rt_flags & RTF_BROADCAST; 358 else 359 isbroadcast = in_broadcast(dst->sin_addr, ifp); 360 } 361 rtmtu_nolock = rt && (rt->rt_rmx.rmx_locks & RTV_MTU) == 0; 362 363 if (IN_MULTICAST(ip->ip_dst.s_addr) || 364 (ip->ip_dst.s_addr == INADDR_BROADCAST)) { 365 bool inmgroup; 366 367 m->m_flags |= (ip->ip_dst.s_addr == INADDR_BROADCAST) ? 368 M_BCAST : M_MCAST; 369 /* 370 * See if the caller provided any multicast options 371 */ 372 if (imo != NULL) 373 ip->ip_ttl = imo->imo_multicast_ttl; 374 else 375 ip->ip_ttl = IP_DEFAULT_MULTICAST_TTL; 376 377 /* 378 * if we don't know the outgoing ifp yet, we can't generate 379 * output 380 */ 381 if (!ifp) { 382 IP_STATINC(IP_STAT_NOROUTE); 383 error = ENETUNREACH; 384 goto bad; 385 } 386 387 /* 388 * If the packet is multicast or broadcast, confirm that 389 * the outgoing interface can transmit it. 390 */ 391 if (((m->m_flags & M_MCAST) && 392 (ifp->if_flags & IFF_MULTICAST) == 0) || 393 ((m->m_flags & M_BCAST) && 394 (ifp->if_flags & (IFF_BROADCAST|IFF_POINTOPOINT)) == 0)) { 395 IP_STATINC(IP_STAT_NOROUTE); 396 error = ENETUNREACH; 397 goto bad; 398 } 399 /* 400 * If source address not specified yet, use an address 401 * of outgoing interface. 402 */ 403 if (in_nullhost(ip->ip_src)) { 404 struct in_ifaddr *xia; 405 struct ifaddr *xifa; 406 407 IFP_TO_IA(ifp, xia); 408 if (!xia) { 409 error = EADDRNOTAVAIL; 410 goto bad; 411 } 412 xifa = &xia->ia_ifa; 413 if (xifa->ifa_getifa != NULL) { 414 xia = ifatoia((*xifa->ifa_getifa)(xifa, rdst)); 415 if (xia == NULL) { 416 error = EADDRNOTAVAIL; 417 goto bad; 418 } 419 } 420 ip->ip_src = xia->ia_addr.sin_addr; 421 } 422 423 inmgroup = in_multi_group(ip->ip_dst, ifp, flags); 424 if (inmgroup && (imo == NULL || imo->imo_multicast_loop)) { 425 /* 426 * If we belong to the destination multicast group 427 * on the outgoing interface, and the caller did not 428 * forbid loopback, loop back a copy. 429 */ 430 ip_mloopback(ifp, m, &u.dst4); 431 } 432 #ifdef MROUTING 433 else { 434 /* 435 * If we are acting as a multicast router, perform 436 * multicast forwarding as if the packet had just 437 * arrived on the interface to which we are about 438 * to send. The multicast forwarding function 439 * recursively calls this function, using the 440 * IP_FORWARDING flag to prevent infinite recursion. 441 * 442 * Multicasts that are looped back by ip_mloopback(), 443 * above, will be forwarded by the ip_input() routine, 444 * if necessary. 445 */ 446 extern struct socket *ip_mrouter; 447 448 if (ip_mrouter && (flags & IP_FORWARDING) == 0) { 449 if (ip_mforward(m, ifp) != 0) { 450 m_freem(m); 451 goto done; 452 } 453 } 454 } 455 #endif 456 /* 457 * Multicasts with a time-to-live of zero may be looped- 458 * back, above, but must not be transmitted on a network. 459 * Also, multicasts addressed to the loopback interface 460 * are not sent -- the above call to ip_mloopback() will 461 * loop back a copy if this host actually belongs to the 462 * destination group on the loopback interface. 463 */ 464 if (ip->ip_ttl == 0 || (ifp->if_flags & IFF_LOOPBACK) != 0) { 465 m_freem(m); 466 goto done; 467 } 468 goto sendit; 469 } 470 471 /* 472 * If source address not specified yet, use address 473 * of outgoing interface. 474 */ 475 if (in_nullhost(ip->ip_src)) { 476 struct ifaddr *xifa; 477 478 xifa = &ia->ia_ifa; 479 if (xifa->ifa_getifa != NULL) { 480 ia = ifatoia((*xifa->ifa_getifa)(xifa, rdst)); 481 if (ia == NULL) { 482 error = EADDRNOTAVAIL; 483 goto bad; 484 } 485 } 486 ip->ip_src = ia->ia_addr.sin_addr; 487 } 488 489 /* 490 * packets with Class-D address as source are not valid per 491 * RFC 1112 492 */ 493 if (IN_MULTICAST(ip->ip_src.s_addr)) { 494 IP_STATINC(IP_STAT_ODROPPED); 495 error = EADDRNOTAVAIL; 496 goto bad; 497 } 498 499 /* 500 * Look for broadcast address and and verify user is allowed to 501 * send such a packet. 502 */ 503 if (isbroadcast) { 504 if ((ifp->if_flags & IFF_BROADCAST) == 0) { 505 error = EADDRNOTAVAIL; 506 goto bad; 507 } 508 if ((flags & IP_ALLOWBROADCAST) == 0) { 509 error = EACCES; 510 goto bad; 511 } 512 /* don't allow broadcast messages to be fragmented */ 513 if (ntohs(ip->ip_len) > ifp->if_mtu) { 514 error = EMSGSIZE; 515 goto bad; 516 } 517 m->m_flags |= M_BCAST; 518 } else 519 m->m_flags &= ~M_BCAST; 520 521 sendit: 522 if ((flags & (IP_FORWARDING|IP_NOIPNEWID)) == 0) { 523 if (m->m_pkthdr.len < IP_MINFRAGSIZE) { 524 ip->ip_id = 0; 525 } else if ((m->m_pkthdr.csum_flags & M_CSUM_TSOv4) == 0) { 526 ip->ip_id = ip_newid(ia); 527 } else { 528 529 /* 530 * TSO capable interfaces (typically?) increment 531 * ip_id for each segment. 532 * "allocate" enough ids here to increase the chance 533 * for them to be unique. 534 * 535 * note that the following calculation is not 536 * needed to be precise. wasting some ip_id is fine. 537 */ 538 539 unsigned int segsz = m->m_pkthdr.segsz; 540 unsigned int datasz = ntohs(ip->ip_len) - hlen; 541 unsigned int num = howmany(datasz, segsz); 542 543 ip->ip_id = ip_newid_range(ia, num); 544 } 545 } 546 547 /* 548 * If we're doing Path MTU Discovery, we need to set DF unless 549 * the route's MTU is locked. 550 */ 551 if ((flags & IP_MTUDISC) != 0 && rtmtu_nolock) { 552 ip->ip_off |= htons(IP_DF); 553 } 554 555 #ifdef IPSEC 556 if (ipsec_used) { 557 bool ipsec_done = false; 558 559 /* Perform IPsec processing, if any. */ 560 error = ipsec4_output(m, so, flags, &sp, &mtu, &natt_frag, 561 &ipsec_done); 562 if (error || ipsec_done) 563 goto done; 564 } 565 #endif 566 567 /* 568 * Run through list of hooks for output packets. 569 */ 570 error = pfil_run_hooks(inet_pfil_hook, &m, ifp, PFIL_OUT); 571 if (error) 572 goto done; 573 if (m == NULL) 574 goto done; 575 576 ip = mtod(m, struct ip *); 577 hlen = ip->ip_hl << 2; 578 579 m->m_pkthdr.csum_data |= hlen << 16; 580 581 #if IFA_STATS 582 /* 583 * search for the source address structure to 584 * maintain output statistics. 585 */ 586 INADDR_TO_IA(ip->ip_src, ia); 587 #endif 588 589 /* Maybe skip checksums on loopback interfaces. */ 590 if (IN_NEED_CHECKSUM(ifp, M_CSUM_IPv4)) { 591 m->m_pkthdr.csum_flags |= M_CSUM_IPv4; 592 } 593 sw_csum = m->m_pkthdr.csum_flags & ~ifp->if_csum_flags_tx; 594 /* 595 * If small enough for mtu of path, or if using TCP segmentation 596 * offload, can just send directly. 597 */ 598 if (ntohs(ip->ip_len) <= mtu || 599 (m->m_pkthdr.csum_flags & M_CSUM_TSOv4) != 0) { 600 const struct sockaddr *sa; 601 602 #if IFA_STATS 603 if (ia) 604 ia->ia_ifa.ifa_data.ifad_outbytes += ntohs(ip->ip_len); 605 #endif 606 /* 607 * Always initialize the sum to 0! Some HW assisted 608 * checksumming requires this. 609 */ 610 ip->ip_sum = 0; 611 612 if ((m->m_pkthdr.csum_flags & M_CSUM_TSOv4) == 0) { 613 /* 614 * Perform any checksums that the hardware can't do 615 * for us. 616 * 617 * XXX Does any hardware require the {th,uh}_sum 618 * XXX fields to be 0? 619 */ 620 if (sw_csum & M_CSUM_IPv4) { 621 KASSERT(IN_NEED_CHECKSUM(ifp, M_CSUM_IPv4)); 622 ip->ip_sum = in_cksum(m, hlen); 623 m->m_pkthdr.csum_flags &= ~M_CSUM_IPv4; 624 } 625 if (sw_csum & (M_CSUM_TCPv4|M_CSUM_UDPv4)) { 626 if (IN_NEED_CHECKSUM(ifp, 627 sw_csum & (M_CSUM_TCPv4|M_CSUM_UDPv4))) { 628 in_delayed_cksum(m); 629 } 630 m->m_pkthdr.csum_flags &= 631 ~(M_CSUM_TCPv4|M_CSUM_UDPv4); 632 } 633 } 634 635 sa = (m->m_flags & M_MCAST) ? sintocsa(rdst) : sintocsa(dst); 636 if (__predict_true( 637 (m->m_pkthdr.csum_flags & M_CSUM_TSOv4) == 0 || 638 (ifp->if_capenable & IFCAP_TSOv4) != 0)) { 639 error = ip_if_output(ifp, m, sa, rt); 640 } else { 641 error = ip_tso_output(ifp, m, sa, rt); 642 } 643 goto done; 644 } 645 646 /* 647 * We can't use HW checksumming if we're about to 648 * to fragment the packet. 649 * 650 * XXX Some hardware can do this. 651 */ 652 if (m->m_pkthdr.csum_flags & (M_CSUM_TCPv4|M_CSUM_UDPv4)) { 653 if (IN_NEED_CHECKSUM(ifp, 654 m->m_pkthdr.csum_flags & (M_CSUM_TCPv4|M_CSUM_UDPv4))) { 655 in_delayed_cksum(m); 656 } 657 m->m_pkthdr.csum_flags &= ~(M_CSUM_TCPv4|M_CSUM_UDPv4); 658 } 659 660 /* 661 * Too large for interface; fragment if possible. 662 * Must be able to put at least 8 bytes per fragment. 663 */ 664 if (ntohs(ip->ip_off) & IP_DF) { 665 if (flags & IP_RETURNMTU) { 666 struct inpcb *inp; 667 668 KASSERT(so && solocked(so)); 669 inp = sotoinpcb(so); 670 inp->inp_errormtu = mtu; 671 } 672 error = EMSGSIZE; 673 IP_STATINC(IP_STAT_CANTFRAG); 674 goto bad; 675 } 676 677 error = ip_fragment(m, ifp, mtu); 678 if (error) { 679 m = NULL; 680 goto bad; 681 } 682 683 for (; m; m = m0) { 684 m0 = m->m_nextpkt; 685 m->m_nextpkt = 0; 686 if (error) { 687 m_freem(m); 688 continue; 689 } 690 #if IFA_STATS 691 if (ia) 692 ia->ia_ifa.ifa_data.ifad_outbytes += ntohs(ip->ip_len); 693 #endif 694 /* 695 * If we get there, the packet has not been handled by 696 * IPsec whereas it should have. Now that it has been 697 * fragmented, re-inject it in ip_output so that IPsec 698 * processing can occur. 699 */ 700 if (natt_frag) { 701 error = ip_output(m, opt, ro, 702 flags | IP_RAWOUTPUT | IP_NOIPNEWID, 703 imo, so); 704 } else { 705 KASSERT((m->m_pkthdr.csum_flags & 706 (M_CSUM_UDPv4 | M_CSUM_TCPv4)) == 0); 707 error = ip_if_output(ifp, m, 708 (m->m_flags & M_MCAST) ? 709 sintocsa(rdst) : sintocsa(dst), rt); 710 } 711 } 712 if (error == 0) { 713 IP_STATINC(IP_STAT_FRAGMENTED); 714 } 715 done: 716 if (ro == &iproute) { 717 rtcache_free(&iproute); 718 } 719 #ifdef IPSEC 720 if (sp) { 721 KEY_FREESP(&sp); 722 } 723 #endif 724 if (mifp != NULL) { 725 if_put(mifp, &psref); 726 curlwp_bindx(bound); 727 } 728 return error; 729 bad: 730 m_freem(m); 731 goto done; 732 } 733 734 int 735 ip_fragment(struct mbuf *m, struct ifnet *ifp, u_long mtu) 736 { 737 struct ip *ip, *mhip; 738 struct mbuf *m0; 739 int len, hlen, off; 740 int mhlen, firstlen; 741 struct mbuf **mnext; 742 int sw_csum = m->m_pkthdr.csum_flags; 743 int fragments = 0; 744 int s; 745 int error = 0; 746 747 ip = mtod(m, struct ip *); 748 hlen = ip->ip_hl << 2; 749 if (ifp != NULL) 750 sw_csum &= ~ifp->if_csum_flags_tx; 751 752 len = (mtu - hlen) &~ 7; 753 if (len < 8) { 754 m_freem(m); 755 return (EMSGSIZE); 756 } 757 758 firstlen = len; 759 mnext = &m->m_nextpkt; 760 761 /* 762 * Loop through length of segment after first fragment, 763 * make new header and copy data of each part and link onto chain. 764 */ 765 m0 = m; 766 mhlen = sizeof (struct ip); 767 for (off = hlen + len; off < ntohs(ip->ip_len); off += len) { 768 MGETHDR(m, M_DONTWAIT, MT_HEADER); 769 if (m == 0) { 770 error = ENOBUFS; 771 IP_STATINC(IP_STAT_ODROPPED); 772 goto sendorfree; 773 } 774 MCLAIM(m, m0->m_owner); 775 *mnext = m; 776 mnext = &m->m_nextpkt; 777 m->m_data += max_linkhdr; 778 mhip = mtod(m, struct ip *); 779 *mhip = *ip; 780 /* we must inherit MCAST and BCAST flags */ 781 m->m_flags |= m0->m_flags & (M_MCAST|M_BCAST); 782 if (hlen > sizeof (struct ip)) { 783 mhlen = ip_optcopy(ip, mhip) + sizeof (struct ip); 784 mhip->ip_hl = mhlen >> 2; 785 } 786 m->m_len = mhlen; 787 mhip->ip_off = ((off - hlen) >> 3) + 788 (ntohs(ip->ip_off) & ~IP_MF); 789 if (ip->ip_off & htons(IP_MF)) 790 mhip->ip_off |= IP_MF; 791 if (off + len >= ntohs(ip->ip_len)) 792 len = ntohs(ip->ip_len) - off; 793 else 794 mhip->ip_off |= IP_MF; 795 HTONS(mhip->ip_off); 796 mhip->ip_len = htons((u_int16_t)(len + mhlen)); 797 m->m_next = m_copym(m0, off, len, M_DONTWAIT); 798 if (m->m_next == 0) { 799 error = ENOBUFS; /* ??? */ 800 IP_STATINC(IP_STAT_ODROPPED); 801 goto sendorfree; 802 } 803 m->m_pkthdr.len = mhlen + len; 804 m_reset_rcvif(m); 805 mhip->ip_sum = 0; 806 KASSERT((m->m_pkthdr.csum_flags & M_CSUM_IPv4) == 0); 807 if (sw_csum & M_CSUM_IPv4) { 808 mhip->ip_sum = in_cksum(m, mhlen); 809 } else { 810 /* 811 * checksum is hw-offloaded or not necessary. 812 */ 813 m->m_pkthdr.csum_flags |= 814 m0->m_pkthdr.csum_flags & M_CSUM_IPv4; 815 m->m_pkthdr.csum_data |= mhlen << 16; 816 KASSERT(!(ifp != NULL && 817 IN_NEED_CHECKSUM(ifp, M_CSUM_IPv4)) || 818 (m->m_pkthdr.csum_flags & M_CSUM_IPv4) != 0); 819 } 820 IP_STATINC(IP_STAT_OFRAGMENTS); 821 fragments++; 822 } 823 /* 824 * Update first fragment by trimming what's been copied out 825 * and updating header, then send each fragment (in order). 826 */ 827 m = m0; 828 m_adj(m, hlen + firstlen - ntohs(ip->ip_len)); 829 m->m_pkthdr.len = hlen + firstlen; 830 ip->ip_len = htons((u_int16_t)m->m_pkthdr.len); 831 ip->ip_off |= htons(IP_MF); 832 ip->ip_sum = 0; 833 if (sw_csum & M_CSUM_IPv4) { 834 ip->ip_sum = in_cksum(m, hlen); 835 m->m_pkthdr.csum_flags &= ~M_CSUM_IPv4; 836 } else { 837 /* 838 * checksum is hw-offloaded or not necessary. 839 */ 840 KASSERT(!(ifp != NULL && IN_NEED_CHECKSUM(ifp, M_CSUM_IPv4)) || 841 (m->m_pkthdr.csum_flags & M_CSUM_IPv4) != 0); 842 KASSERT(M_CSUM_DATA_IPv4_IPHL(m->m_pkthdr.csum_data) >= 843 sizeof(struct ip)); 844 } 845 sendorfree: 846 /* 847 * If there is no room for all the fragments, don't queue 848 * any of them. 849 */ 850 if (ifp != NULL) { 851 s = splnet(); 852 if (ifp->if_snd.ifq_maxlen - ifp->if_snd.ifq_len < fragments && 853 error == 0) { 854 error = ENOBUFS; 855 IP_STATINC(IP_STAT_ODROPPED); 856 IFQ_INC_DROPS(&ifp->if_snd); 857 } 858 splx(s); 859 } 860 if (error) { 861 for (m = m0; m; m = m0) { 862 m0 = m->m_nextpkt; 863 m->m_nextpkt = NULL; 864 m_freem(m); 865 } 866 } 867 return (error); 868 } 869 870 /* 871 * Process a delayed payload checksum calculation. 872 */ 873 void 874 in_delayed_cksum(struct mbuf *m) 875 { 876 struct ip *ip; 877 u_int16_t csum, offset; 878 879 ip = mtod(m, struct ip *); 880 offset = ip->ip_hl << 2; 881 csum = in4_cksum(m, 0, offset, ntohs(ip->ip_len) - offset); 882 if (csum == 0 && (m->m_pkthdr.csum_flags & M_CSUM_UDPv4) != 0) 883 csum = 0xffff; 884 885 offset += M_CSUM_DATA_IPv4_OFFSET(m->m_pkthdr.csum_data); 886 887 if ((offset + sizeof(u_int16_t)) > m->m_len) { 888 /* This happen when ip options were inserted 889 printf("in_delayed_cksum: pullup len %d off %d proto %d\n", 890 m->m_len, offset, ip->ip_p); 891 */ 892 m_copyback(m, offset, sizeof(csum), (void *) &csum); 893 } else 894 *(u_int16_t *)(mtod(m, char *) + offset) = csum; 895 } 896 897 /* 898 * Determine the maximum length of the options to be inserted; 899 * we would far rather allocate too much space rather than too little. 900 */ 901 902 u_int 903 ip_optlen(struct inpcb *inp) 904 { 905 struct mbuf *m = inp->inp_options; 906 907 if (m && m->m_len > offsetof(struct ipoption, ipopt_dst)) { 908 return (m->m_len - offsetof(struct ipoption, ipopt_dst)); 909 } 910 return 0; 911 } 912 913 /* 914 * Insert IP options into preformed packet. 915 * Adjust IP destination as required for IP source routing, 916 * as indicated by a non-zero in_addr at the start of the options. 917 */ 918 static struct mbuf * 919 ip_insertoptions(struct mbuf *m, struct mbuf *opt, int *phlen) 920 { 921 struct ipoption *p = mtod(opt, struct ipoption *); 922 struct mbuf *n; 923 struct ip *ip = mtod(m, struct ip *); 924 unsigned optlen; 925 926 optlen = opt->m_len - sizeof(p->ipopt_dst); 927 if (optlen + ntohs(ip->ip_len) > IP_MAXPACKET) 928 return (m); /* XXX should fail */ 929 if (!in_nullhost(p->ipopt_dst)) 930 ip->ip_dst = p->ipopt_dst; 931 if (M_READONLY(m) || M_LEADINGSPACE(m) < optlen) { 932 MGETHDR(n, M_DONTWAIT, MT_HEADER); 933 if (n == 0) 934 return (m); 935 MCLAIM(n, m->m_owner); 936 M_MOVE_PKTHDR(n, m); 937 m->m_len -= sizeof(struct ip); 938 m->m_data += sizeof(struct ip); 939 n->m_next = m; 940 m = n; 941 m->m_len = optlen + sizeof(struct ip); 942 m->m_data += max_linkhdr; 943 bcopy((void *)ip, mtod(m, void *), sizeof(struct ip)); 944 } else { 945 m->m_data -= optlen; 946 m->m_len += optlen; 947 memmove(mtod(m, void *), ip, sizeof(struct ip)); 948 } 949 m->m_pkthdr.len += optlen; 950 ip = mtod(m, struct ip *); 951 bcopy((void *)p->ipopt_list, (void *)(ip + 1), (unsigned)optlen); 952 *phlen = sizeof(struct ip) + optlen; 953 ip->ip_len = htons(ntohs(ip->ip_len) + optlen); 954 return (m); 955 } 956 957 /* 958 * Copy options from ip to jp, 959 * omitting those not copied during fragmentation. 960 */ 961 int 962 ip_optcopy(struct ip *ip, struct ip *jp) 963 { 964 u_char *cp, *dp; 965 int opt, optlen, cnt; 966 967 cp = (u_char *)(ip + 1); 968 dp = (u_char *)(jp + 1); 969 cnt = (ip->ip_hl << 2) - sizeof (struct ip); 970 for (; cnt > 0; cnt -= optlen, cp += optlen) { 971 opt = cp[0]; 972 if (opt == IPOPT_EOL) 973 break; 974 if (opt == IPOPT_NOP) { 975 /* Preserve for IP mcast tunnel's LSRR alignment. */ 976 *dp++ = IPOPT_NOP; 977 optlen = 1; 978 continue; 979 } 980 981 KASSERT(cnt >= IPOPT_OLEN + sizeof(*cp)); 982 optlen = cp[IPOPT_OLEN]; 983 KASSERT(optlen >= IPOPT_OLEN + sizeof(*cp) && optlen < cnt); 984 985 /* Invalid lengths should have been caught by ip_dooptions. */ 986 if (optlen > cnt) 987 optlen = cnt; 988 if (IPOPT_COPIED(opt)) { 989 bcopy((void *)cp, (void *)dp, (unsigned)optlen); 990 dp += optlen; 991 } 992 } 993 for (optlen = dp - (u_char *)(jp+1); optlen & 0x3; optlen++) 994 *dp++ = IPOPT_EOL; 995 return (optlen); 996 } 997 998 /* 999 * IP socket option processing. 1000 */ 1001 int 1002 ip_ctloutput(int op, struct socket *so, struct sockopt *sopt) 1003 { 1004 struct inpcb *inp = sotoinpcb(so); 1005 struct ip *ip = &inp->inp_ip; 1006 int inpflags = inp->inp_flags; 1007 int optval = 0, error = 0; 1008 1009 if (sopt->sopt_level != IPPROTO_IP) { 1010 if (sopt->sopt_level == SOL_SOCKET && sopt->sopt_name == SO_NOHEADER) 1011 return 0; 1012 return ENOPROTOOPT; 1013 } 1014 1015 switch (op) { 1016 case PRCO_SETOPT: 1017 switch (sopt->sopt_name) { 1018 case IP_OPTIONS: 1019 #ifdef notyet 1020 case IP_RETOPTS: 1021 #endif 1022 error = ip_pcbopts(inp, sopt); 1023 break; 1024 1025 case IP_TOS: 1026 case IP_TTL: 1027 case IP_MINTTL: 1028 case IP_PKTINFO: 1029 case IP_RECVOPTS: 1030 case IP_RECVRETOPTS: 1031 case IP_RECVDSTADDR: 1032 case IP_RECVIF: 1033 case IP_RECVPKTINFO: 1034 case IP_RECVTTL: 1035 error = sockopt_getint(sopt, &optval); 1036 if (error) 1037 break; 1038 1039 switch (sopt->sopt_name) { 1040 case IP_TOS: 1041 ip->ip_tos = optval; 1042 break; 1043 1044 case IP_TTL: 1045 ip->ip_ttl = optval; 1046 break; 1047 1048 case IP_MINTTL: 1049 if (optval > 0 && optval <= MAXTTL) 1050 inp->inp_ip_minttl = optval; 1051 else 1052 error = EINVAL; 1053 break; 1054 #define OPTSET(bit) \ 1055 if (optval) \ 1056 inpflags |= bit; \ 1057 else \ 1058 inpflags &= ~bit; 1059 1060 case IP_PKTINFO: 1061 OPTSET(INP_PKTINFO); 1062 break; 1063 1064 case IP_RECVOPTS: 1065 OPTSET(INP_RECVOPTS); 1066 break; 1067 1068 case IP_RECVPKTINFO: 1069 OPTSET(INP_RECVPKTINFO); 1070 break; 1071 1072 case IP_RECVRETOPTS: 1073 OPTSET(INP_RECVRETOPTS); 1074 break; 1075 1076 case IP_RECVDSTADDR: 1077 OPTSET(INP_RECVDSTADDR); 1078 break; 1079 1080 case IP_RECVIF: 1081 OPTSET(INP_RECVIF); 1082 break; 1083 1084 case IP_RECVTTL: 1085 OPTSET(INP_RECVTTL); 1086 break; 1087 } 1088 break; 1089 #undef OPTSET 1090 1091 case IP_MULTICAST_IF: 1092 case IP_MULTICAST_TTL: 1093 case IP_MULTICAST_LOOP: 1094 case IP_ADD_MEMBERSHIP: 1095 case IP_DROP_MEMBERSHIP: 1096 error = ip_setmoptions(&inp->inp_moptions, sopt); 1097 break; 1098 1099 case IP_PORTRANGE: 1100 error = sockopt_getint(sopt, &optval); 1101 if (error) 1102 break; 1103 1104 switch (optval) { 1105 case IP_PORTRANGE_DEFAULT: 1106 case IP_PORTRANGE_HIGH: 1107 inpflags &= ~(INP_LOWPORT); 1108 break; 1109 1110 case IP_PORTRANGE_LOW: 1111 inpflags |= INP_LOWPORT; 1112 break; 1113 1114 default: 1115 error = EINVAL; 1116 break; 1117 } 1118 break; 1119 1120 case IP_PORTALGO: 1121 error = sockopt_getint(sopt, &optval); 1122 if (error) 1123 break; 1124 1125 error = portalgo_algo_index_select( 1126 (struct inpcb_hdr *)inp, optval); 1127 break; 1128 1129 #if defined(IPSEC) 1130 case IP_IPSEC_POLICY: 1131 if (ipsec_enabled) { 1132 error = ipsec4_set_policy(inp, sopt->sopt_name, 1133 sopt->sopt_data, sopt->sopt_size, 1134 curlwp->l_cred); 1135 break; 1136 } 1137 /*FALLTHROUGH*/ 1138 #endif /* IPSEC */ 1139 1140 default: 1141 error = ENOPROTOOPT; 1142 break; 1143 } 1144 break; 1145 1146 case PRCO_GETOPT: 1147 switch (sopt->sopt_name) { 1148 case IP_OPTIONS: 1149 case IP_RETOPTS: { 1150 struct mbuf *mopts = inp->inp_options; 1151 1152 if (mopts) { 1153 struct mbuf *m; 1154 1155 m = m_copym(mopts, 0, M_COPYALL, M_DONTWAIT); 1156 if (m == NULL) { 1157 error = ENOBUFS; 1158 break; 1159 } 1160 error = sockopt_setmbuf(sopt, m); 1161 } 1162 break; 1163 } 1164 case IP_PKTINFO: 1165 case IP_TOS: 1166 case IP_TTL: 1167 case IP_MINTTL: 1168 case IP_RECVOPTS: 1169 case IP_RECVRETOPTS: 1170 case IP_RECVDSTADDR: 1171 case IP_RECVIF: 1172 case IP_RECVPKTINFO: 1173 case IP_RECVTTL: 1174 case IP_ERRORMTU: 1175 switch (sopt->sopt_name) { 1176 case IP_TOS: 1177 optval = ip->ip_tos; 1178 break; 1179 1180 case IP_TTL: 1181 optval = ip->ip_ttl; 1182 break; 1183 1184 case IP_MINTTL: 1185 optval = inp->inp_ip_minttl; 1186 break; 1187 1188 case IP_ERRORMTU: 1189 optval = inp->inp_errormtu; 1190 break; 1191 1192 #define OPTBIT(bit) (inpflags & bit ? 1 : 0) 1193 1194 case IP_PKTINFO: 1195 optval = OPTBIT(INP_PKTINFO); 1196 break; 1197 1198 case IP_RECVOPTS: 1199 optval = OPTBIT(INP_RECVOPTS); 1200 break; 1201 1202 case IP_RECVPKTINFO: 1203 optval = OPTBIT(INP_RECVPKTINFO); 1204 break; 1205 1206 case IP_RECVRETOPTS: 1207 optval = OPTBIT(INP_RECVRETOPTS); 1208 break; 1209 1210 case IP_RECVDSTADDR: 1211 optval = OPTBIT(INP_RECVDSTADDR); 1212 break; 1213 1214 case IP_RECVIF: 1215 optval = OPTBIT(INP_RECVIF); 1216 break; 1217 1218 case IP_RECVTTL: 1219 optval = OPTBIT(INP_RECVTTL); 1220 break; 1221 } 1222 error = sockopt_setint(sopt, optval); 1223 break; 1224 1225 #if 0 /* defined(IPSEC) */ 1226 case IP_IPSEC_POLICY: 1227 { 1228 struct mbuf *m = NULL; 1229 1230 /* XXX this will return EINVAL as sopt is empty */ 1231 error = ipsec4_get_policy(inp, sopt->sopt_data, 1232 sopt->sopt_size, &m); 1233 if (error == 0) 1234 error = sockopt_setmbuf(sopt, m); 1235 break; 1236 } 1237 #endif /*IPSEC*/ 1238 1239 case IP_MULTICAST_IF: 1240 case IP_MULTICAST_TTL: 1241 case IP_MULTICAST_LOOP: 1242 case IP_ADD_MEMBERSHIP: 1243 case IP_DROP_MEMBERSHIP: 1244 error = ip_getmoptions(inp->inp_moptions, sopt); 1245 break; 1246 1247 case IP_PORTRANGE: 1248 if (inpflags & INP_LOWPORT) 1249 optval = IP_PORTRANGE_LOW; 1250 else 1251 optval = IP_PORTRANGE_DEFAULT; 1252 error = sockopt_setint(sopt, optval); 1253 break; 1254 1255 case IP_PORTALGO: 1256 optval = inp->inp_portalgo; 1257 error = sockopt_setint(sopt, optval); 1258 break; 1259 1260 default: 1261 error = ENOPROTOOPT; 1262 break; 1263 } 1264 break; 1265 } 1266 1267 if (!error) { 1268 inp->inp_flags = inpflags; 1269 } 1270 return error; 1271 } 1272 1273 /* 1274 * Set up IP options in pcb for insertion in output packets. 1275 * Store in mbuf with pointer in pcbopt, adding pseudo-option 1276 * with destination address if source routed. 1277 */ 1278 static int 1279 ip_pcbopts(struct inpcb *inp, const struct sockopt *sopt) 1280 { 1281 struct mbuf *m; 1282 const u_char *cp; 1283 u_char *dp; 1284 int cnt; 1285 1286 /* Turn off any old options. */ 1287 if (inp->inp_options) { 1288 m_free(inp->inp_options); 1289 } 1290 inp->inp_options = NULL; 1291 if ((cnt = sopt->sopt_size) == 0) { 1292 /* Only turning off any previous options. */ 1293 return 0; 1294 } 1295 cp = sopt->sopt_data; 1296 1297 #ifndef __vax__ 1298 if (cnt % sizeof(int32_t)) 1299 return (EINVAL); 1300 #endif 1301 1302 m = m_get(M_DONTWAIT, MT_SOOPTS); 1303 if (m == NULL) 1304 return (ENOBUFS); 1305 1306 dp = mtod(m, u_char *); 1307 memset(dp, 0, sizeof(struct in_addr)); 1308 dp += sizeof(struct in_addr); 1309 m->m_len = sizeof(struct in_addr); 1310 1311 /* 1312 * IP option list according to RFC791. Each option is of the form 1313 * 1314 * [optval] [olen] [(olen - 2) data bytes] 1315 * 1316 * We validate the list and copy options to an mbuf for prepending 1317 * to data packets. The IP first-hop destination address will be 1318 * stored before actual options and is zero if unset. 1319 */ 1320 while (cnt > 0) { 1321 uint8_t optval, olen, offset; 1322 1323 optval = cp[IPOPT_OPTVAL]; 1324 1325 if (optval == IPOPT_EOL || optval == IPOPT_NOP) { 1326 olen = 1; 1327 } else { 1328 if (cnt < IPOPT_OLEN + 1) 1329 goto bad; 1330 1331 olen = cp[IPOPT_OLEN]; 1332 if (olen < IPOPT_OLEN + 1 || olen > cnt) 1333 goto bad; 1334 } 1335 1336 if (optval == IPOPT_LSRR || optval == IPOPT_SSRR) { 1337 /* 1338 * user process specifies route as: 1339 * ->A->B->C->D 1340 * D must be our final destination (but we can't 1341 * check that since we may not have connected yet). 1342 * A is first hop destination, which doesn't appear in 1343 * actual IP option, but is stored before the options. 1344 */ 1345 if (olen < IPOPT_OFFSET + 1 + sizeof(struct in_addr)) 1346 goto bad; 1347 1348 offset = cp[IPOPT_OFFSET]; 1349 memcpy(mtod(m, u_char *), cp + IPOPT_OFFSET + 1, 1350 sizeof(struct in_addr)); 1351 1352 cp += sizeof(struct in_addr); 1353 cnt -= sizeof(struct in_addr); 1354 olen -= sizeof(struct in_addr); 1355 1356 if (m->m_len + olen > MAX_IPOPTLEN + sizeof(struct in_addr)) 1357 goto bad; 1358 1359 memcpy(dp, cp, olen); 1360 dp[IPOPT_OPTVAL] = optval; 1361 dp[IPOPT_OLEN] = olen; 1362 dp[IPOPT_OFFSET] = offset; 1363 break; 1364 } else { 1365 if (m->m_len + olen > MAX_IPOPTLEN + sizeof(struct in_addr)) 1366 goto bad; 1367 1368 memcpy(dp, cp, olen); 1369 break; 1370 } 1371 1372 dp += olen; 1373 m->m_len += olen; 1374 1375 if (optval == IPOPT_EOL) 1376 break; 1377 1378 cp += olen; 1379 cnt -= olen; 1380 } 1381 1382 inp->inp_options = m; 1383 return 0; 1384 bad: 1385 (void)m_free(m); 1386 return EINVAL; 1387 } 1388 1389 /* 1390 * following RFC1724 section 3.3, 0.0.0.0/8 is interpreted as interface index. 1391 */ 1392 static struct ifnet * 1393 ip_multicast_if(struct in_addr *a, int *ifindexp) 1394 { 1395 int ifindex; 1396 struct ifnet *ifp = NULL; 1397 struct in_ifaddr *ia; 1398 1399 if (ifindexp) 1400 *ifindexp = 0; 1401 if (ntohl(a->s_addr) >> 24 == 0) { 1402 ifindex = ntohl(a->s_addr) & 0xffffff; 1403 ifp = if_byindex(ifindex); 1404 if (!ifp) 1405 return NULL; 1406 if (ifindexp) 1407 *ifindexp = ifindex; 1408 } else { 1409 LIST_FOREACH(ia, &IN_IFADDR_HASH(a->s_addr), ia_hash) { 1410 if (in_hosteq(ia->ia_addr.sin_addr, *a) && 1411 (ia->ia_ifp->if_flags & IFF_MULTICAST) != 0) { 1412 ifp = ia->ia_ifp; 1413 break; 1414 } 1415 } 1416 } 1417 return ifp; 1418 } 1419 1420 static int 1421 ip_getoptval(const struct sockopt *sopt, u_int8_t *val, u_int maxval) 1422 { 1423 u_int tval; 1424 u_char cval; 1425 int error; 1426 1427 if (sopt == NULL) 1428 return EINVAL; 1429 1430 switch (sopt->sopt_size) { 1431 case sizeof(u_char): 1432 error = sockopt_get(sopt, &cval, sizeof(u_char)); 1433 tval = cval; 1434 break; 1435 1436 case sizeof(u_int): 1437 error = sockopt_get(sopt, &tval, sizeof(u_int)); 1438 break; 1439 1440 default: 1441 error = EINVAL; 1442 } 1443 1444 if (error) 1445 return error; 1446 1447 if (tval > maxval) 1448 return EINVAL; 1449 1450 *val = tval; 1451 return 0; 1452 } 1453 1454 static int 1455 ip_get_membership(const struct sockopt *sopt, struct ifnet **ifp, 1456 struct in_addr *ia, bool add) 1457 { 1458 int error; 1459 struct ip_mreq mreq; 1460 1461 error = sockopt_get(sopt, &mreq, sizeof(mreq)); 1462 if (error) 1463 return error; 1464 1465 if (!IN_MULTICAST(mreq.imr_multiaddr.s_addr)) 1466 return EINVAL; 1467 1468 memcpy(ia, &mreq.imr_multiaddr, sizeof(*ia)); 1469 1470 if (in_nullhost(mreq.imr_interface)) { 1471 union { 1472 struct sockaddr dst; 1473 struct sockaddr_in dst4; 1474 } u; 1475 struct route ro; 1476 1477 if (!add) { 1478 *ifp = NULL; 1479 return 0; 1480 } 1481 /* 1482 * If no interface address was provided, use the interface of 1483 * the route to the given multicast address. 1484 */ 1485 struct rtentry *rt; 1486 memset(&ro, 0, sizeof(ro)); 1487 1488 sockaddr_in_init(&u.dst4, ia, 0); 1489 error = rtcache_setdst(&ro, &u.dst); 1490 if (error != 0) 1491 return error; 1492 *ifp = (rt = rtcache_init(&ro)) != NULL ? rt->rt_ifp : NULL; 1493 rtcache_free(&ro); 1494 } else { 1495 *ifp = ip_multicast_if(&mreq.imr_interface, NULL); 1496 if (!add && *ifp == NULL) 1497 return EADDRNOTAVAIL; 1498 } 1499 return 0; 1500 } 1501 1502 /* 1503 * Add a multicast group membership. 1504 * Group must be a valid IP multicast address. 1505 */ 1506 static int 1507 ip_add_membership(struct ip_moptions *imo, const struct sockopt *sopt) 1508 { 1509 struct ifnet *ifp = NULL; // XXX: gcc [ppc] 1510 struct in_addr ia; 1511 int i, error; 1512 1513 if (sopt->sopt_size == sizeof(struct ip_mreq)) 1514 error = ip_get_membership(sopt, &ifp, &ia, true); 1515 else 1516 #ifdef INET6 1517 error = ip6_get_membership(sopt, &ifp, &ia, sizeof(ia)); 1518 #else 1519 return EINVAL; 1520 #endif 1521 1522 if (error) 1523 return error; 1524 1525 /* 1526 * See if we found an interface, and confirm that it 1527 * supports multicast. 1528 */ 1529 if (ifp == NULL || (ifp->if_flags & IFF_MULTICAST) == 0) 1530 return EADDRNOTAVAIL; 1531 1532 /* 1533 * See if the membership already exists or if all the 1534 * membership slots are full. 1535 */ 1536 for (i = 0; i < imo->imo_num_memberships; ++i) { 1537 if (imo->imo_membership[i]->inm_ifp == ifp && 1538 in_hosteq(imo->imo_membership[i]->inm_addr, ia)) 1539 break; 1540 } 1541 if (i < imo->imo_num_memberships) 1542 return EADDRINUSE; 1543 1544 if (i == IP_MAX_MEMBERSHIPS) 1545 return ETOOMANYREFS; 1546 1547 /* 1548 * Everything looks good; add a new record to the multicast 1549 * address list for the given interface. 1550 */ 1551 if ((imo->imo_membership[i] = in_addmulti(&ia, ifp)) == NULL) 1552 return ENOBUFS; 1553 1554 ++imo->imo_num_memberships; 1555 return 0; 1556 } 1557 1558 /* 1559 * Drop a multicast group membership. 1560 * Group must be a valid IP multicast address. 1561 */ 1562 static int 1563 ip_drop_membership(struct ip_moptions *imo, const struct sockopt *sopt) 1564 { 1565 struct in_addr ia = { .s_addr = 0 }; // XXX: gcc [ppc] 1566 struct ifnet *ifp = NULL; // XXX: gcc [ppc] 1567 int i, error; 1568 1569 if (sopt->sopt_size == sizeof(struct ip_mreq)) 1570 error = ip_get_membership(sopt, &ifp, &ia, false); 1571 else 1572 #ifdef INET6 1573 error = ip6_get_membership(sopt, &ifp, &ia, sizeof(ia)); 1574 #else 1575 return EINVAL; 1576 #endif 1577 1578 if (error) 1579 return error; 1580 1581 /* 1582 * Find the membership in the membership array. 1583 */ 1584 for (i = 0; i < imo->imo_num_memberships; ++i) { 1585 if ((ifp == NULL || 1586 imo->imo_membership[i]->inm_ifp == ifp) && 1587 in_hosteq(imo->imo_membership[i]->inm_addr, ia)) 1588 break; 1589 } 1590 if (i == imo->imo_num_memberships) 1591 return EADDRNOTAVAIL; 1592 1593 /* 1594 * Give up the multicast address record to which the 1595 * membership points. 1596 */ 1597 in_delmulti(imo->imo_membership[i]); 1598 1599 /* 1600 * Remove the gap in the membership array. 1601 */ 1602 for (++i; i < imo->imo_num_memberships; ++i) 1603 imo->imo_membership[i-1] = imo->imo_membership[i]; 1604 --imo->imo_num_memberships; 1605 return 0; 1606 } 1607 1608 /* 1609 * Set the IP multicast options in response to user setsockopt(). 1610 */ 1611 int 1612 ip_setmoptions(struct ip_moptions **pimo, const struct sockopt *sopt) 1613 { 1614 struct ip_moptions *imo = *pimo; 1615 struct in_addr addr; 1616 struct ifnet *ifp; 1617 int ifindex, error = 0; 1618 1619 if (!imo) { 1620 /* 1621 * No multicast option buffer attached to the pcb; 1622 * allocate one and initialize to default values. 1623 */ 1624 imo = kmem_intr_alloc(sizeof(*imo), KM_NOSLEEP); 1625 if (imo == NULL) 1626 return ENOBUFS; 1627 1628 imo->imo_multicast_if_index = 0; 1629 imo->imo_multicast_addr.s_addr = INADDR_ANY; 1630 imo->imo_multicast_ttl = IP_DEFAULT_MULTICAST_TTL; 1631 imo->imo_multicast_loop = IP_DEFAULT_MULTICAST_LOOP; 1632 imo->imo_num_memberships = 0; 1633 *pimo = imo; 1634 } 1635 1636 switch (sopt->sopt_name) { 1637 case IP_MULTICAST_IF: 1638 /* 1639 * Select the interface for outgoing multicast packets. 1640 */ 1641 error = sockopt_get(sopt, &addr, sizeof(addr)); 1642 if (error) 1643 break; 1644 1645 /* 1646 * INADDR_ANY is used to remove a previous selection. 1647 * When no interface is selected, a default one is 1648 * chosen every time a multicast packet is sent. 1649 */ 1650 if (in_nullhost(addr)) { 1651 imo->imo_multicast_if_index = 0; 1652 break; 1653 } 1654 /* 1655 * The selected interface is identified by its local 1656 * IP address. Find the interface and confirm that 1657 * it supports multicasting. 1658 */ 1659 ifp = ip_multicast_if(&addr, &ifindex); 1660 if (ifp == NULL || (ifp->if_flags & IFF_MULTICAST) == 0) { 1661 error = EADDRNOTAVAIL; 1662 break; 1663 } 1664 imo->imo_multicast_if_index = ifp->if_index; 1665 if (ifindex) 1666 imo->imo_multicast_addr = addr; 1667 else 1668 imo->imo_multicast_addr.s_addr = INADDR_ANY; 1669 break; 1670 1671 case IP_MULTICAST_TTL: 1672 /* 1673 * Set the IP time-to-live for outgoing multicast packets. 1674 */ 1675 error = ip_getoptval(sopt, &imo->imo_multicast_ttl, MAXTTL); 1676 break; 1677 1678 case IP_MULTICAST_LOOP: 1679 /* 1680 * Set the loopback flag for outgoing multicast packets. 1681 * Must be zero or one. 1682 */ 1683 error = ip_getoptval(sopt, &imo->imo_multicast_loop, 1); 1684 break; 1685 1686 case IP_ADD_MEMBERSHIP: /* IPV6_JOIN_GROUP */ 1687 error = ip_add_membership(imo, sopt); 1688 break; 1689 1690 case IP_DROP_MEMBERSHIP: /* IPV6_LEAVE_GROUP */ 1691 error = ip_drop_membership(imo, sopt); 1692 break; 1693 1694 default: 1695 error = EOPNOTSUPP; 1696 break; 1697 } 1698 1699 /* 1700 * If all options have default values, no need to keep the mbuf. 1701 */ 1702 if (imo->imo_multicast_if_index == 0 && 1703 imo->imo_multicast_ttl == IP_DEFAULT_MULTICAST_TTL && 1704 imo->imo_multicast_loop == IP_DEFAULT_MULTICAST_LOOP && 1705 imo->imo_num_memberships == 0) { 1706 kmem_free(imo, sizeof(*imo)); 1707 *pimo = NULL; 1708 } 1709 1710 return error; 1711 } 1712 1713 /* 1714 * Return the IP multicast options in response to user getsockopt(). 1715 */ 1716 int 1717 ip_getmoptions(struct ip_moptions *imo, struct sockopt *sopt) 1718 { 1719 struct in_addr addr; 1720 uint8_t optval; 1721 int error = 0; 1722 1723 switch (sopt->sopt_name) { 1724 case IP_MULTICAST_IF: 1725 if (imo == NULL || imo->imo_multicast_if_index == 0) 1726 addr = zeroin_addr; 1727 else if (imo->imo_multicast_addr.s_addr) { 1728 /* return the value user has set */ 1729 addr = imo->imo_multicast_addr; 1730 } else { 1731 struct ifnet *ifp; 1732 struct in_ifaddr *ia = NULL; 1733 int s = pserialize_read_enter(); 1734 1735 ifp = if_byindex(imo->imo_multicast_if_index); 1736 if (ifp != NULL) { 1737 IFP_TO_IA(ifp, ia); 1738 } 1739 addr = ia ? ia->ia_addr.sin_addr : zeroin_addr; 1740 pserialize_read_exit(s); 1741 } 1742 error = sockopt_set(sopt, &addr, sizeof(addr)); 1743 break; 1744 1745 case IP_MULTICAST_TTL: 1746 optval = imo ? imo->imo_multicast_ttl 1747 : IP_DEFAULT_MULTICAST_TTL; 1748 1749 error = sockopt_set(sopt, &optval, sizeof(optval)); 1750 break; 1751 1752 case IP_MULTICAST_LOOP: 1753 optval = imo ? imo->imo_multicast_loop 1754 : IP_DEFAULT_MULTICAST_LOOP; 1755 1756 error = sockopt_set(sopt, &optval, sizeof(optval)); 1757 break; 1758 1759 default: 1760 error = EOPNOTSUPP; 1761 } 1762 1763 return error; 1764 } 1765 1766 /* 1767 * Discard the IP multicast options. 1768 */ 1769 void 1770 ip_freemoptions(struct ip_moptions *imo) 1771 { 1772 int i; 1773 1774 if (imo != NULL) { 1775 for (i = 0; i < imo->imo_num_memberships; ++i) 1776 in_delmulti(imo->imo_membership[i]); 1777 kmem_free(imo, sizeof(*imo)); 1778 } 1779 } 1780 1781 /* 1782 * Routine called from ip_output() to loop back a copy of an IP multicast 1783 * packet to the input queue of a specified interface. Note that this 1784 * calls the output routine of the loopback "driver", but with an interface 1785 * pointer that might NOT be lo0ifp -- easier than replicating that code here. 1786 */ 1787 static void 1788 ip_mloopback(struct ifnet *ifp, struct mbuf *m, const struct sockaddr_in *dst) 1789 { 1790 struct ip *ip; 1791 struct mbuf *copym; 1792 1793 copym = m_copypacket(m, M_DONTWAIT); 1794 if (copym != NULL && 1795 (copym->m_flags & M_EXT || copym->m_len < sizeof(struct ip))) 1796 copym = m_pullup(copym, sizeof(struct ip)); 1797 if (copym == NULL) 1798 return; 1799 /* 1800 * We don't bother to fragment if the IP length is greater 1801 * than the interface's MTU. Can this possibly matter? 1802 */ 1803 ip = mtod(copym, struct ip *); 1804 1805 if (copym->m_pkthdr.csum_flags & (M_CSUM_TCPv4|M_CSUM_UDPv4)) { 1806 in_delayed_cksum(copym); 1807 copym->m_pkthdr.csum_flags &= 1808 ~(M_CSUM_TCPv4|M_CSUM_UDPv4); 1809 } 1810 1811 ip->ip_sum = 0; 1812 ip->ip_sum = in_cksum(copym, ip->ip_hl << 2); 1813 #ifndef NET_MPSAFE 1814 KERNEL_LOCK(1, NULL); 1815 #endif 1816 (void)looutput(ifp, copym, sintocsa(dst), NULL); 1817 #ifndef NET_MPSAFE 1818 KERNEL_UNLOCK_ONE(NULL); 1819 #endif 1820 } 1821
Indexes created Tue Sep 23 07:09:52 GMT 2025