1 /*- 2 * SPDX-License-Identifier: BSD-3-Clause 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) 1982, 1986, 1988, 1993 34 * The Regents of the University of California. 35 * All rights reserved. 36 * 37 * Redistribution and use in source and binary forms, with or without 38 * modification, are permitted provided that the following conditions 39 * are met: 40 * 1. Redistributions of source code must retain the above copyright 41 * notice, this list of conditions and the following disclaimer. 42 * 2. Redistributions in binary form must reproduce the above copyright 43 * notice, this list of conditions and the following disclaimer in the 44 * documentation and/or other materials provided with the distribution. 45 * 3. Neither the name of the University nor the names of its contributors 46 * may be used to endorse or promote products derived from this software 47 * without specific prior written permission. 48 * 49 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 50 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 51 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 52 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 53 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 54 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 55 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 56 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 57 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 58 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 59 * SUCH DAMAGE. 60 * 61 * @(#)raw_ip.c 8.2 (Berkeley) 1/4/94 62 */ 63 64 #include <sys/cdefs.h> 65 __FBSDID("$FreeBSD$"); 66 67 #include "opt_ipsec.h" 68 #include "opt_inet6.h" 69 70 #include <sys/param.h> 71 #include <sys/errno.h> 72 #include <sys/jail.h> 73 #include <sys/kernel.h> 74 #include <sys/lock.h> 75 #include <sys/malloc.h> 76 #include <sys/mbuf.h> 77 #include <sys/priv.h> 78 #include <sys/proc.h> 79 #include <sys/protosw.h> 80 #include <sys/signalvar.h> 81 #include <sys/socket.h> 82 #include <sys/socketvar.h> 83 #include <sys/sx.h> 84 #include <sys/syslog.h> 85 86 #include <net/if.h> 87 #include <net/if_var.h> 88 #include <net/if_types.h> 89 #include <net/route.h> 90 #include <net/vnet.h> 91 92 #include <netinet/in.h> 93 #include <netinet/in_var.h> 94 #include <netinet/in_systm.h> 95 #include <netinet/in_pcb.h> 96 97 #include <netinet/icmp6.h> 98 #include <netinet/ip6.h> 99 #include <netinet/ip_var.h> 100 #include <netinet6/ip6protosw.h> 101 #include <netinet6/ip6_mroute.h> 102 #include <netinet6/in6_pcb.h> 103 #include <netinet6/ip6_var.h> 104 #include <netinet6/nd6.h> 105 #include <netinet6/raw_ip6.h> 106 #include <netinet6/scope6_var.h> 107 #include <netinet6/send.h> 108 109 #include <netipsec/ipsec_support.h> 110 111 #include <machine/stdarg.h> 112 113 #define satosin6(sa) ((struct sockaddr_in6 *)(sa)) 114 #define ifatoia6(ifa) ((struct in6_ifaddr *)(ifa)) 115 116 /* 117 * Raw interface to IP6 protocol. 118 */ 119 120 VNET_DECLARE(struct inpcbhead, ripcb); 121 VNET_DECLARE(struct inpcbinfo, ripcbinfo); 122 #define V_ripcb VNET(ripcb) 123 #define V_ripcbinfo VNET(ripcbinfo) 124 125 extern u_long rip_sendspace; 126 extern u_long rip_recvspace; 127 128 VNET_PCPUSTAT_DEFINE(struct rip6stat, rip6stat); 129 VNET_PCPUSTAT_SYSINIT(rip6stat); 130 131 #ifdef VIMAGE 132 VNET_PCPUSTAT_SYSUNINIT(rip6stat); 133 #endif /* VIMAGE */ 134 135 /* 136 * Hooks for multicast routing. They all default to NULL, so leave them not 137 * initialized and rely on BSS being set to 0. 138 */ 139 140 /* 141 * The socket used to communicate with the multicast routing daemon. 142 */ 143 VNET_DEFINE(struct socket *, ip6_mrouter); 144 145 /* 146 * The various mrouter functions. 147 */ 148 int (*ip6_mrouter_set)(struct socket *, struct sockopt *); 149 int (*ip6_mrouter_get)(struct socket *, struct sockopt *); 150 int (*ip6_mrouter_done)(void); 151 int (*ip6_mforward)(struct ip6_hdr *, struct ifnet *, struct mbuf *); 152 int (*mrt6_ioctl)(u_long, caddr_t); 153 154 /* 155 * Setup generic address and protocol structures for raw_input routine, then 156 * pass them along with mbuf chain. 157 */ 158 int 159 rip6_input(struct mbuf **mp, int *offp, int proto) 160 { 161 struct ifnet *ifp; 162 struct mbuf *m = *mp; 163 struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *); 164 struct inpcb *in6p; 165 struct inpcb *last = NULL; 166 struct mbuf *opts = NULL; 167 struct sockaddr_in6 fromsa; 168 struct epoch_tracker et; 169 170 RIP6STAT_INC(rip6s_ipackets); 171 172 init_sin6(&fromsa, m, 0); /* general init */ 173 174 ifp = m->m_pkthdr.rcvif; 175 176 INP_INFO_RLOCK_ET(&V_ripcbinfo, et); 177 CK_LIST_FOREACH(in6p, &V_ripcb, inp_list) { 178 /* XXX inp locking */ 179 if ((in6p->inp_vflag & INP_IPV6) == 0) 180 continue; 181 if (in6p->inp_ip_p && 182 in6p->inp_ip_p != proto) 183 continue; 184 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_laddr) && 185 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_laddr, &ip6->ip6_dst)) 186 continue; 187 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_faddr) && 188 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_faddr, &ip6->ip6_src)) 189 continue; 190 if (last != NULL) { 191 struct mbuf *n = m_copym(m, 0, M_COPYALL, M_NOWAIT); 192 193 #if defined(IPSEC) || defined(IPSEC_SUPPORT) 194 /* 195 * Check AH/ESP integrity. 196 */ 197 if (IPSEC_ENABLED(ipv6)) { 198 if (n != NULL && 199 IPSEC_CHECK_POLICY(ipv6, n, last) != 0) { 200 m_freem(n); 201 /* Do not inject data into pcb. */ 202 n = NULL; 203 } 204 } 205 #endif /* IPSEC */ 206 if (n) { 207 if (last->inp_flags & INP_CONTROLOPTS || 208 last->inp_socket->so_options & SO_TIMESTAMP) 209 ip6_savecontrol(last, n, &opts); 210 /* strip intermediate headers */ 211 m_adj(n, *offp); 212 if (sbappendaddr(&last->inp_socket->so_rcv, 213 (struct sockaddr *)&fromsa, 214 n, opts) == 0) { 215 m_freem(n); 216 if (opts) 217 m_freem(opts); 218 RIP6STAT_INC(rip6s_fullsock); 219 } else 220 sorwakeup(last->inp_socket); 221 opts = NULL; 222 } 223 INP_RUNLOCK(last); 224 last = NULL; 225 } 226 INP_RLOCK(in6p); 227 if (__predict_false(in6p->inp_flags2 & INP_FREED)) 228 goto skip_2; 229 if (jailed_without_vnet(in6p->inp_cred)) { 230 /* 231 * Allow raw socket in jail to receive multicast; 232 * assume process had PRIV_NETINET_RAW at attach, 233 * and fall through into normal filter path if so. 234 */ 235 if (!IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst) && 236 prison_check_ip6(in6p->inp_cred, 237 &ip6->ip6_dst) != 0) 238 goto skip_2; 239 } 240 if (in6p->in6p_cksum != -1) { 241 RIP6STAT_INC(rip6s_isum); 242 if (m->m_pkthdr.len - (*offp + in6p->in6p_cksum) < 2 || 243 in6_cksum(m, proto, *offp, 244 m->m_pkthdr.len - *offp)) { 245 RIP6STAT_INC(rip6s_badsum); 246 /* 247 * Drop the received message, don't send an 248 * ICMP6 message. Set proto to IPPROTO_NONE 249 * to achieve that. 250 */ 251 proto = IPPROTO_NONE; 252 goto skip_2; 253 } 254 } 255 /* 256 * If this raw socket has multicast state, and we 257 * have received a multicast, check if this socket 258 * should receive it, as multicast filtering is now 259 * the responsibility of the transport layer. 260 */ 261 if (in6p->in6p_moptions && 262 IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) { 263 /* 264 * If the incoming datagram is for MLD, allow it 265 * through unconditionally to the raw socket. 266 * 267 * Use the M_RTALERT_MLD flag to check for MLD 268 * traffic without having to inspect the mbuf chain 269 * more deeply, as all MLDv1/v2 host messages MUST 270 * contain the Router Alert option. 271 * 272 * In the case of MLDv1, we may not have explicitly 273 * joined the group, and may have set IFF_ALLMULTI 274 * on the interface. im6o_mc_filter() may discard 275 * control traffic we actually need to see. 276 * 277 * Userland multicast routing daemons should continue 278 * filter the control traffic appropriately. 279 */ 280 int blocked; 281 282 blocked = MCAST_PASS; 283 if ((m->m_flags & M_RTALERT_MLD) == 0) { 284 struct sockaddr_in6 mcaddr; 285 286 bzero(&mcaddr, sizeof(struct sockaddr_in6)); 287 mcaddr.sin6_len = sizeof(struct sockaddr_in6); 288 mcaddr.sin6_family = AF_INET6; 289 mcaddr.sin6_addr = ip6->ip6_dst; 290 291 blocked = im6o_mc_filter(in6p->in6p_moptions, 292 ifp, 293 (struct sockaddr *)&mcaddr, 294 (struct sockaddr *)&fromsa); 295 } 296 if (blocked != MCAST_PASS) { 297 IP6STAT_INC(ip6s_notmember); 298 goto skip_2; 299 } 300 } 301 last = in6p; 302 continue; 303 skip_2: 304 INP_RUNLOCK(in6p); 305 } 306 INP_INFO_RUNLOCK_ET(&V_ripcbinfo, et); 307 #if defined(IPSEC) || defined(IPSEC_SUPPORT) 308 /* 309 * Check AH/ESP integrity. 310 */ 311 if (IPSEC_ENABLED(ipv6) && last != NULL && 312 IPSEC_CHECK_POLICY(ipv6, m, last) != 0) { 313 m_freem(m); 314 IP6STAT_DEC(ip6s_delivered); 315 /* Do not inject data into pcb. */ 316 INP_RUNLOCK(last); 317 } else 318 #endif /* IPSEC */ 319 if (last != NULL) { 320 if (last->inp_flags & INP_CONTROLOPTS || 321 last->inp_socket->so_options & SO_TIMESTAMP) 322 ip6_savecontrol(last, m, &opts); 323 /* Strip intermediate headers. */ 324 m_adj(m, *offp); 325 if (sbappendaddr(&last->inp_socket->so_rcv, 326 (struct sockaddr *)&fromsa, m, opts) == 0) { 327 m_freem(m); 328 if (opts) 329 m_freem(opts); 330 RIP6STAT_INC(rip6s_fullsock); 331 } else 332 sorwakeup(last->inp_socket); 333 INP_RUNLOCK(last); 334 } else { 335 RIP6STAT_INC(rip6s_nosock); 336 if (m->m_flags & M_MCAST) 337 RIP6STAT_INC(rip6s_nosockmcast); 338 if (proto == IPPROTO_NONE) 339 m_freem(m); 340 else 341 icmp6_error(m, ICMP6_PARAM_PROB, 342 ICMP6_PARAMPROB_NEXTHEADER, 343 ip6_get_prevhdr(m, *offp)); 344 IP6STAT_DEC(ip6s_delivered); 345 } 346 return (IPPROTO_DONE); 347 } 348 349 void 350 rip6_ctlinput(int cmd, struct sockaddr *sa, void *d) 351 { 352 struct ip6ctlparam *ip6cp = NULL; 353 const struct sockaddr_in6 *sa6_src = NULL; 354 void *cmdarg; 355 struct inpcb *(*notify)(struct inpcb *, int) = in6_rtchange; 356 357 if (sa->sa_family != AF_INET6 || 358 sa->sa_len != sizeof(struct sockaddr_in6)) 359 return; 360 361 if ((unsigned)cmd >= PRC_NCMDS) 362 return; 363 if (PRC_IS_REDIRECT(cmd)) 364 notify = in6_rtchange, d = NULL; 365 else if (cmd == PRC_HOSTDEAD) 366 d = NULL; 367 else if (inet6ctlerrmap[cmd] == 0) 368 return; 369 370 /* 371 * If the parameter is from icmp6, decode it. 372 */ 373 if (d != NULL) { 374 ip6cp = (struct ip6ctlparam *)d; 375 cmdarg = ip6cp->ip6c_cmdarg; 376 sa6_src = ip6cp->ip6c_src; 377 } else { 378 cmdarg = NULL; 379 sa6_src = &sa6_any; 380 } 381 382 (void) in6_pcbnotify(&V_ripcbinfo, sa, 0, 383 (const struct sockaddr *)sa6_src, 0, cmd, cmdarg, notify); 384 } 385 386 /* 387 * Generate IPv6 header and pass packet to ip6_output. Tack on options user 388 * may have setup with control call. 389 */ 390 int 391 rip6_output(struct mbuf *m, struct socket *so, ...) 392 { 393 struct mbuf *control; 394 struct m_tag *mtag; 395 struct sockaddr_in6 *dstsock; 396 struct ip6_hdr *ip6; 397 struct inpcb *in6p; 398 u_int plen = m->m_pkthdr.len; 399 int error = 0; 400 struct ip6_pktopts opt, *optp; 401 struct ifnet *oifp = NULL; 402 int type = 0, code = 0; /* for ICMPv6 output statistics only */ 403 int scope_ambiguous = 0; 404 int use_defzone = 0; 405 int hlim = 0; 406 struct in6_addr in6a; 407 va_list ap; 408 409 va_start(ap, so); 410 dstsock = va_arg(ap, struct sockaddr_in6 *); 411 control = va_arg(ap, struct mbuf *); 412 va_end(ap); 413 414 in6p = sotoinpcb(so); 415 INP_WLOCK(in6p); 416 417 if (control != NULL) { 418 if ((error = ip6_setpktopts(control, &opt, 419 in6p->in6p_outputopts, so->so_cred, 420 so->so_proto->pr_protocol)) != 0) { 421 goto bad; 422 } 423 optp = &opt; 424 } else 425 optp = in6p->in6p_outputopts; 426 427 /* 428 * Check and convert scope zone ID into internal form. 429 * 430 * XXX: we may still need to determine the zone later. 431 */ 432 if (!(so->so_state & SS_ISCONNECTED)) { 433 if (!optp || !optp->ip6po_pktinfo || 434 !optp->ip6po_pktinfo->ipi6_ifindex) 435 use_defzone = V_ip6_use_defzone; 436 if (dstsock->sin6_scope_id == 0 && !use_defzone) 437 scope_ambiguous = 1; 438 if ((error = sa6_embedscope(dstsock, use_defzone)) != 0) 439 goto bad; 440 } 441 442 /* 443 * For an ICMPv6 packet, we should know its type and code to update 444 * statistics. 445 */ 446 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) { 447 struct icmp6_hdr *icmp6; 448 if (m->m_len < sizeof(struct icmp6_hdr) && 449 (m = m_pullup(m, sizeof(struct icmp6_hdr))) == NULL) { 450 error = ENOBUFS; 451 goto bad; 452 } 453 icmp6 = mtod(m, struct icmp6_hdr *); 454 type = icmp6->icmp6_type; 455 code = icmp6->icmp6_code; 456 } 457 458 M_PREPEND(m, sizeof(*ip6), M_NOWAIT); 459 if (m == NULL) { 460 error = ENOBUFS; 461 goto bad; 462 } 463 ip6 = mtod(m, struct ip6_hdr *); 464 465 /* 466 * Source address selection. 467 */ 468 error = in6_selectsrc_socket(dstsock, optp, in6p, so->so_cred, 469 scope_ambiguous, &in6a, &hlim); 470 471 if (error) 472 goto bad; 473 error = prison_check_ip6(in6p->inp_cred, &in6a); 474 if (error != 0) 475 goto bad; 476 ip6->ip6_src = in6a; 477 478 ip6->ip6_dst = dstsock->sin6_addr; 479 480 /* 481 * Fill in the rest of the IPv6 header fields. 482 */ 483 ip6->ip6_flow = (ip6->ip6_flow & ~IPV6_FLOWINFO_MASK) | 484 (in6p->inp_flow & IPV6_FLOWINFO_MASK); 485 ip6->ip6_vfc = (ip6->ip6_vfc & ~IPV6_VERSION_MASK) | 486 (IPV6_VERSION & IPV6_VERSION_MASK); 487 488 /* 489 * ip6_plen will be filled in ip6_output, so not fill it here. 490 */ 491 ip6->ip6_nxt = in6p->inp_ip_p; 492 ip6->ip6_hlim = hlim; 493 494 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6 || 495 in6p->in6p_cksum != -1) { 496 struct mbuf *n; 497 int off; 498 u_int16_t *p; 499 500 /* Compute checksum. */ 501 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) 502 off = offsetof(struct icmp6_hdr, icmp6_cksum); 503 else 504 off = in6p->in6p_cksum; 505 if (plen < off + 2) { 506 error = EINVAL; 507 goto bad; 508 } 509 off += sizeof(struct ip6_hdr); 510 511 n = m; 512 while (n && n->m_len <= off) { 513 off -= n->m_len; 514 n = n->m_next; 515 } 516 if (!n) 517 goto bad; 518 p = (u_int16_t *)(mtod(n, caddr_t) + off); 519 *p = 0; 520 *p = in6_cksum(m, ip6->ip6_nxt, sizeof(*ip6), plen); 521 } 522 523 /* 524 * Send RA/RS messages to user land for protection, before sending 525 * them to rtadvd/rtsol. 526 */ 527 if ((send_sendso_input_hook != NULL) && 528 so->so_proto->pr_protocol == IPPROTO_ICMPV6) { 529 switch (type) { 530 case ND_ROUTER_ADVERT: 531 case ND_ROUTER_SOLICIT: 532 mtag = m_tag_get(PACKET_TAG_ND_OUTGOING, 533 sizeof(unsigned short), M_NOWAIT); 534 if (mtag == NULL) 535 goto bad; 536 m_tag_prepend(m, mtag); 537 } 538 } 539 540 error = ip6_output(m, optp, NULL, 0, in6p->in6p_moptions, &oifp, in6p); 541 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) { 542 if (oifp) 543 icmp6_ifoutstat_inc(oifp, type, code); 544 ICMP6STAT_INC(icp6s_outhist[type]); 545 } else 546 RIP6STAT_INC(rip6s_opackets); 547 548 goto freectl; 549 550 bad: 551 if (m) 552 m_freem(m); 553 554 freectl: 555 if (control != NULL) { 556 ip6_clearpktopts(&opt, -1); 557 m_freem(control); 558 } 559 INP_WUNLOCK(in6p); 560 return (error); 561 } 562 563 /* 564 * Raw IPv6 socket option processing. 565 */ 566 int 567 rip6_ctloutput(struct socket *so, struct sockopt *sopt) 568 { 569 struct inpcb *inp; 570 int error; 571 572 if (sopt->sopt_level == IPPROTO_ICMPV6) 573 /* 574 * XXX: is it better to call icmp6_ctloutput() directly 575 * from protosw? 576 */ 577 return (icmp6_ctloutput(so, sopt)); 578 else if (sopt->sopt_level != IPPROTO_IPV6) { 579 if (sopt->sopt_level == SOL_SOCKET && 580 sopt->sopt_name == SO_SETFIB) { 581 inp = sotoinpcb(so); 582 INP_WLOCK(inp); 583 inp->inp_inc.inc_fibnum = so->so_fibnum; 584 INP_WUNLOCK(inp); 585 return (0); 586 } 587 return (EINVAL); 588 } 589 590 error = 0; 591 592 switch (sopt->sopt_dir) { 593 case SOPT_GET: 594 switch (sopt->sopt_name) { 595 case MRT6_INIT: 596 case MRT6_DONE: 597 case MRT6_ADD_MIF: 598 case MRT6_DEL_MIF: 599 case MRT6_ADD_MFC: 600 case MRT6_DEL_MFC: 601 case MRT6_PIM: 602 error = ip6_mrouter_get ? ip6_mrouter_get(so, sopt) : 603 EOPNOTSUPP; 604 break; 605 case IPV6_CHECKSUM: 606 error = ip6_raw_ctloutput(so, sopt); 607 break; 608 default: 609 error = ip6_ctloutput(so, sopt); 610 break; 611 } 612 break; 613 614 case SOPT_SET: 615 switch (sopt->sopt_name) { 616 case MRT6_INIT: 617 case MRT6_DONE: 618 case MRT6_ADD_MIF: 619 case MRT6_DEL_MIF: 620 case MRT6_ADD_MFC: 621 case MRT6_DEL_MFC: 622 case MRT6_PIM: 623 error = ip6_mrouter_set ? ip6_mrouter_set(so, sopt) : 624 EOPNOTSUPP; 625 break; 626 case IPV6_CHECKSUM: 627 error = ip6_raw_ctloutput(so, sopt); 628 break; 629 default: 630 error = ip6_ctloutput(so, sopt); 631 break; 632 } 633 break; 634 } 635 636 return (error); 637 } 638 639 static int 640 rip6_attach(struct socket *so, int proto, struct thread *td) 641 { 642 struct inpcb *inp; 643 struct icmp6_filter *filter; 644 int error; 645 646 inp = sotoinpcb(so); 647 KASSERT(inp == NULL, ("rip6_attach: inp != NULL")); 648 649 error = priv_check(td, PRIV_NETINET_RAW); 650 if (error) 651 return (error); 652 error = soreserve(so, rip_sendspace, rip_recvspace); 653 if (error) 654 return (error); 655 filter = malloc(sizeof(struct icmp6_filter), M_PCB, M_NOWAIT); 656 if (filter == NULL) 657 return (ENOMEM); 658 INP_INFO_WLOCK(&V_ripcbinfo); 659 error = in_pcballoc(so, &V_ripcbinfo); 660 if (error) { 661 INP_INFO_WUNLOCK(&V_ripcbinfo); 662 free(filter, M_PCB); 663 return (error); 664 } 665 inp = (struct inpcb *)so->so_pcb; 666 INP_INFO_WUNLOCK(&V_ripcbinfo); 667 inp->inp_vflag |= INP_IPV6; 668 inp->inp_ip_p = (long)proto; 669 inp->in6p_hops = -1; /* use kernel default */ 670 inp->in6p_cksum = -1; 671 inp->in6p_icmp6filt = filter; 672 ICMP6_FILTER_SETPASSALL(inp->in6p_icmp6filt); 673 INP_WUNLOCK(inp); 674 return (0); 675 } 676 677 static void 678 rip6_detach(struct socket *so) 679 { 680 struct inpcb *inp; 681 682 inp = sotoinpcb(so); 683 KASSERT(inp != NULL, ("rip6_detach: inp == NULL")); 684 685 if (so == V_ip6_mrouter && ip6_mrouter_done) 686 ip6_mrouter_done(); 687 /* xxx: RSVP */ 688 INP_INFO_WLOCK(&V_ripcbinfo); 689 INP_WLOCK(inp); 690 free(inp->in6p_icmp6filt, M_PCB); 691 in_pcbdetach(inp); 692 in_pcbfree(inp); 693 INP_INFO_WUNLOCK(&V_ripcbinfo); 694 } 695 696 /* XXXRW: This can't ever be called. */ 697 static void 698 rip6_abort(struct socket *so) 699 { 700 struct inpcb *inp; 701 702 inp = sotoinpcb(so); 703 KASSERT(inp != NULL, ("rip6_abort: inp == NULL")); 704 705 soisdisconnected(so); 706 } 707 708 static void 709 rip6_close(struct socket *so) 710 { 711 struct inpcb *inp; 712 713 inp = sotoinpcb(so); 714 KASSERT(inp != NULL, ("rip6_close: inp == NULL")); 715 716 soisdisconnected(so); 717 } 718 719 static int 720 rip6_disconnect(struct socket *so) 721 { 722 struct inpcb *inp; 723 724 inp = sotoinpcb(so); 725 KASSERT(inp != NULL, ("rip6_disconnect: inp == NULL")); 726 727 if ((so->so_state & SS_ISCONNECTED) == 0) 728 return (ENOTCONN); 729 inp->in6p_faddr = in6addr_any; 730 rip6_abort(so); 731 return (0); 732 } 733 734 static int 735 rip6_bind(struct socket *so, struct sockaddr *nam, struct thread *td) 736 { 737 struct epoch_tracker et; 738 struct inpcb *inp; 739 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam; 740 struct ifaddr *ifa = NULL; 741 int error = 0; 742 743 inp = sotoinpcb(so); 744 KASSERT(inp != NULL, ("rip6_bind: inp == NULL")); 745 746 if (nam->sa_len != sizeof(*addr)) 747 return (EINVAL); 748 if ((error = prison_check_ip6(td->td_ucred, &addr->sin6_addr)) != 0) 749 return (error); 750 if (CK_STAILQ_EMPTY(&V_ifnet) || addr->sin6_family != AF_INET6) 751 return (EADDRNOTAVAIL); 752 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0) 753 return (error); 754 755 NET_EPOCH_ENTER(et); 756 if (!IN6_IS_ADDR_UNSPECIFIED(&addr->sin6_addr) && 757 (ifa = ifa_ifwithaddr((struct sockaddr *)addr)) == NULL) { 758 NET_EPOCH_EXIT(et); 759 return (EADDRNOTAVAIL); 760 } 761 if (ifa != NULL && 762 ((struct in6_ifaddr *)ifa)->ia6_flags & 763 (IN6_IFF_ANYCAST|IN6_IFF_NOTREADY| 764 IN6_IFF_DETACHED|IN6_IFF_DEPRECATED)) { 765 NET_EPOCH_EXIT(et); 766 return (EADDRNOTAVAIL); 767 } 768 NET_EPOCH_EXIT(et); 769 INP_INFO_WLOCK(&V_ripcbinfo); 770 INP_WLOCK(inp); 771 inp->in6p_laddr = addr->sin6_addr; 772 INP_WUNLOCK(inp); 773 INP_INFO_WUNLOCK(&V_ripcbinfo); 774 return (0); 775 } 776 777 static int 778 rip6_connect(struct socket *so, struct sockaddr *nam, struct thread *td) 779 { 780 struct inpcb *inp; 781 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam; 782 struct in6_addr in6a; 783 int error = 0, scope_ambiguous = 0; 784 785 inp = sotoinpcb(so); 786 KASSERT(inp != NULL, ("rip6_connect: inp == NULL")); 787 788 if (nam->sa_len != sizeof(*addr)) 789 return (EINVAL); 790 if (CK_STAILQ_EMPTY(&V_ifnet)) 791 return (EADDRNOTAVAIL); 792 if (addr->sin6_family != AF_INET6) 793 return (EAFNOSUPPORT); 794 795 /* 796 * Application should provide a proper zone ID or the use of default 797 * zone IDs should be enabled. Unfortunately, some applications do 798 * not behave as it should, so we need a workaround. Even if an 799 * appropriate ID is not determined, we'll see if we can determine 800 * the outgoing interface. If we can, determine the zone ID based on 801 * the interface below. 802 */ 803 if (addr->sin6_scope_id == 0 && !V_ip6_use_defzone) 804 scope_ambiguous = 1; 805 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0) 806 return (error); 807 808 INP_INFO_WLOCK(&V_ripcbinfo); 809 INP_WLOCK(inp); 810 /* Source address selection. XXX: need pcblookup? */ 811 error = in6_selectsrc_socket(addr, inp->in6p_outputopts, 812 inp, so->so_cred, scope_ambiguous, &in6a, NULL); 813 if (error) { 814 INP_WUNLOCK(inp); 815 INP_INFO_WUNLOCK(&V_ripcbinfo); 816 return (error); 817 } 818 819 inp->in6p_faddr = addr->sin6_addr; 820 inp->in6p_laddr = in6a; 821 soisconnected(so); 822 INP_WUNLOCK(inp); 823 INP_INFO_WUNLOCK(&V_ripcbinfo); 824 return (0); 825 } 826 827 static int 828 rip6_shutdown(struct socket *so) 829 { 830 struct inpcb *inp; 831 832 inp = sotoinpcb(so); 833 KASSERT(inp != NULL, ("rip6_shutdown: inp == NULL")); 834 835 INP_WLOCK(inp); 836 socantsendmore(so); 837 INP_WUNLOCK(inp); 838 return (0); 839 } 840 841 static int 842 rip6_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam, 843 struct mbuf *control, struct thread *td) 844 { 845 struct inpcb *inp; 846 struct sockaddr_in6 tmp; 847 struct sockaddr_in6 *dst; 848 int ret; 849 850 inp = sotoinpcb(so); 851 KASSERT(inp != NULL, ("rip6_send: inp == NULL")); 852 853 /* Always copy sockaddr to avoid overwrites. */ 854 /* Unlocked read. */ 855 if (so->so_state & SS_ISCONNECTED) { 856 if (nam) { 857 m_freem(m); 858 return (EISCONN); 859 } 860 /* XXX */ 861 bzero(&tmp, sizeof(tmp)); 862 tmp.sin6_family = AF_INET6; 863 tmp.sin6_len = sizeof(struct sockaddr_in6); 864 INP_RLOCK(inp); 865 bcopy(&inp->in6p_faddr, &tmp.sin6_addr, 866 sizeof(struct in6_addr)); 867 INP_RUNLOCK(inp); 868 dst = &tmp; 869 } else { 870 if (nam == NULL) { 871 m_freem(m); 872 return (ENOTCONN); 873 } 874 if (nam->sa_len != sizeof(struct sockaddr_in6)) { 875 m_freem(m); 876 return (EINVAL); 877 } 878 tmp = *(struct sockaddr_in6 *)nam; 879 dst = &tmp; 880 881 if (dst->sin6_family == AF_UNSPEC) { 882 /* 883 * XXX: we allow this case for backward 884 * compatibility to buggy applications that 885 * rely on old (and wrong) kernel behavior. 886 */ 887 log(LOG_INFO, "rip6 SEND: address family is " 888 "unspec. Assume AF_INET6\n"); 889 dst->sin6_family = AF_INET6; 890 } else if (dst->sin6_family != AF_INET6) { 891 m_freem(m); 892 return(EAFNOSUPPORT); 893 } 894 } 895 ret = rip6_output(m, so, dst, control); 896 return (ret); 897 } 898 899 struct pr_usrreqs rip6_usrreqs = { 900 .pru_abort = rip6_abort, 901 .pru_attach = rip6_attach, 902 .pru_bind = rip6_bind, 903 .pru_connect = rip6_connect, 904 .pru_control = in6_control, 905 .pru_detach = rip6_detach, 906 .pru_disconnect = rip6_disconnect, 907 .pru_peeraddr = in6_getpeeraddr, 908 .pru_send = rip6_send, 909 .pru_shutdown = rip6_shutdown, 910 .pru_sockaddr = in6_getsockaddr, 911 .pru_close = rip6_close, 912 }; 913