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 (jailed_without_vnet(in6p->inp_cred)) { 191 /* 192 * Allow raw socket in jail to receive multicast; 193 * assume process had PRIV_NETINET_RAW at attach, 194 * and fall through into normal filter path if so. 195 */ 196 if (!IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst) && 197 prison_check_ip6(in6p->inp_cred, 198 &ip6->ip6_dst) != 0) 199 continue; 200 } 201 INP_RLOCK(in6p); 202 if (in6p->in6p_cksum != -1) { 203 RIP6STAT_INC(rip6s_isum); 204 if (in6_cksum(m, proto, *offp, 205 m->m_pkthdr.len - *offp)) { 206 INP_RUNLOCK(in6p); 207 RIP6STAT_INC(rip6s_badsum); 208 continue; 209 } 210 } 211 /* 212 * If this raw socket has multicast state, and we 213 * have received a multicast, check if this socket 214 * should receive it, as multicast filtering is now 215 * the responsibility of the transport layer. 216 */ 217 if (in6p->in6p_moptions && 218 IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) { 219 /* 220 * If the incoming datagram is for MLD, allow it 221 * through unconditionally to the raw socket. 222 * 223 * Use the M_RTALERT_MLD flag to check for MLD 224 * traffic without having to inspect the mbuf chain 225 * more deeply, as all MLDv1/v2 host messages MUST 226 * contain the Router Alert option. 227 * 228 * In the case of MLDv1, we may not have explicitly 229 * joined the group, and may have set IFF_ALLMULTI 230 * on the interface. im6o_mc_filter() may discard 231 * control traffic we actually need to see. 232 * 233 * Userland multicast routing daemons should continue 234 * filter the control traffic appropriately. 235 */ 236 int blocked; 237 238 blocked = MCAST_PASS; 239 if ((m->m_flags & M_RTALERT_MLD) == 0) { 240 struct sockaddr_in6 mcaddr; 241 242 bzero(&mcaddr, sizeof(struct sockaddr_in6)); 243 mcaddr.sin6_len = sizeof(struct sockaddr_in6); 244 mcaddr.sin6_family = AF_INET6; 245 mcaddr.sin6_addr = ip6->ip6_dst; 246 247 blocked = im6o_mc_filter(in6p->in6p_moptions, 248 ifp, 249 (struct sockaddr *)&mcaddr, 250 (struct sockaddr *)&fromsa); 251 } 252 if (blocked != MCAST_PASS) { 253 IP6STAT_INC(ip6s_notmember); 254 INP_RUNLOCK(in6p); 255 continue; 256 } 257 } 258 if (last != NULL) { 259 struct mbuf *n = m_copym(m, 0, M_COPYALL, M_NOWAIT); 260 261 #if defined(IPSEC) || defined(IPSEC_SUPPORT) 262 /* 263 * Check AH/ESP integrity. 264 */ 265 if (IPSEC_ENABLED(ipv6)) { 266 if (n != NULL && 267 IPSEC_CHECK_POLICY(ipv6, n, last) != 0) { 268 m_freem(n); 269 /* Do not inject data into pcb. */ 270 n = NULL; 271 } 272 } 273 #endif /* IPSEC */ 274 if (n) { 275 if (last->inp_flags & INP_CONTROLOPTS || 276 last->inp_socket->so_options & SO_TIMESTAMP) 277 ip6_savecontrol(last, n, &opts); 278 /* strip intermediate headers */ 279 m_adj(n, *offp); 280 if (sbappendaddr(&last->inp_socket->so_rcv, 281 (struct sockaddr *)&fromsa, 282 n, opts) == 0) { 283 m_freem(n); 284 if (opts) 285 m_freem(opts); 286 RIP6STAT_INC(rip6s_fullsock); 287 } else 288 sorwakeup(last->inp_socket); 289 opts = NULL; 290 } 291 INP_RUNLOCK(last); 292 } 293 last = in6p; 294 } 295 INP_INFO_RUNLOCK_ET(&V_ripcbinfo, et); 296 #if defined(IPSEC) || defined(IPSEC_SUPPORT) 297 /* 298 * Check AH/ESP integrity. 299 */ 300 if (IPSEC_ENABLED(ipv6) && last != NULL && 301 IPSEC_CHECK_POLICY(ipv6, m, last) != 0) { 302 m_freem(m); 303 IP6STAT_DEC(ip6s_delivered); 304 /* Do not inject data into pcb. */ 305 INP_RUNLOCK(last); 306 } else 307 #endif /* IPSEC */ 308 if (last != NULL) { 309 if (last->inp_flags & INP_CONTROLOPTS || 310 last->inp_socket->so_options & SO_TIMESTAMP) 311 ip6_savecontrol(last, m, &opts); 312 /* Strip intermediate headers. */ 313 m_adj(m, *offp); 314 if (sbappendaddr(&last->inp_socket->so_rcv, 315 (struct sockaddr *)&fromsa, m, opts) == 0) { 316 m_freem(m); 317 if (opts) 318 m_freem(opts); 319 RIP6STAT_INC(rip6s_fullsock); 320 } else 321 sorwakeup(last->inp_socket); 322 INP_RUNLOCK(last); 323 } else { 324 RIP6STAT_INC(rip6s_nosock); 325 if (m->m_flags & M_MCAST) 326 RIP6STAT_INC(rip6s_nosockmcast); 327 if (proto == IPPROTO_NONE) 328 m_freem(m); 329 else 330 icmp6_error(m, ICMP6_PARAM_PROB, 331 ICMP6_PARAMPROB_NEXTHEADER, 332 ip6_get_prevhdr(m, *offp)); 333 IP6STAT_DEC(ip6s_delivered); 334 } 335 return (IPPROTO_DONE); 336 } 337 338 void 339 rip6_ctlinput(int cmd, struct sockaddr *sa, void *d) 340 { 341 struct ip6ctlparam *ip6cp = NULL; 342 const struct sockaddr_in6 *sa6_src = NULL; 343 void *cmdarg; 344 struct inpcb *(*notify)(struct inpcb *, int) = in6_rtchange; 345 346 if (sa->sa_family != AF_INET6 || 347 sa->sa_len != sizeof(struct sockaddr_in6)) 348 return; 349 350 if ((unsigned)cmd >= PRC_NCMDS) 351 return; 352 if (PRC_IS_REDIRECT(cmd)) 353 notify = in6_rtchange, d = NULL; 354 else if (cmd == PRC_HOSTDEAD) 355 d = NULL; 356 else if (inet6ctlerrmap[cmd] == 0) 357 return; 358 359 /* 360 * If the parameter is from icmp6, decode it. 361 */ 362 if (d != NULL) { 363 ip6cp = (struct ip6ctlparam *)d; 364 cmdarg = ip6cp->ip6c_cmdarg; 365 sa6_src = ip6cp->ip6c_src; 366 } else { 367 cmdarg = NULL; 368 sa6_src = &sa6_any; 369 } 370 371 (void) in6_pcbnotify(&V_ripcbinfo, sa, 0, 372 (const struct sockaddr *)sa6_src, 0, cmd, cmdarg, notify); 373 } 374 375 /* 376 * Generate IPv6 header and pass packet to ip6_output. Tack on options user 377 * may have setup with control call. 378 */ 379 int 380 rip6_output(struct mbuf *m, struct socket *so, ...) 381 { 382 struct mbuf *control; 383 struct m_tag *mtag; 384 struct sockaddr_in6 *dstsock; 385 struct ip6_hdr *ip6; 386 struct inpcb *in6p; 387 u_int plen = m->m_pkthdr.len; 388 int error = 0; 389 struct ip6_pktopts opt, *optp; 390 struct ifnet *oifp = NULL; 391 int type = 0, code = 0; /* for ICMPv6 output statistics only */ 392 int scope_ambiguous = 0; 393 int use_defzone = 0; 394 int hlim = 0; 395 struct in6_addr in6a; 396 va_list ap; 397 398 va_start(ap, so); 399 dstsock = va_arg(ap, struct sockaddr_in6 *); 400 control = va_arg(ap, struct mbuf *); 401 va_end(ap); 402 403 in6p = sotoinpcb(so); 404 INP_WLOCK(in6p); 405 406 if (control != NULL) { 407 if ((error = ip6_setpktopts(control, &opt, 408 in6p->in6p_outputopts, so->so_cred, 409 so->so_proto->pr_protocol)) != 0) { 410 goto bad; 411 } 412 optp = &opt; 413 } else 414 optp = in6p->in6p_outputopts; 415 416 /* 417 * Check and convert scope zone ID into internal form. 418 * 419 * XXX: we may still need to determine the zone later. 420 */ 421 if (!(so->so_state & SS_ISCONNECTED)) { 422 if (!optp || !optp->ip6po_pktinfo || 423 !optp->ip6po_pktinfo->ipi6_ifindex) 424 use_defzone = V_ip6_use_defzone; 425 if (dstsock->sin6_scope_id == 0 && !use_defzone) 426 scope_ambiguous = 1; 427 if ((error = sa6_embedscope(dstsock, use_defzone)) != 0) 428 goto bad; 429 } 430 431 /* 432 * For an ICMPv6 packet, we should know its type and code to update 433 * statistics. 434 */ 435 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) { 436 struct icmp6_hdr *icmp6; 437 if (m->m_len < sizeof(struct icmp6_hdr) && 438 (m = m_pullup(m, sizeof(struct icmp6_hdr))) == NULL) { 439 error = ENOBUFS; 440 goto bad; 441 } 442 icmp6 = mtod(m, struct icmp6_hdr *); 443 type = icmp6->icmp6_type; 444 code = icmp6->icmp6_code; 445 } 446 447 M_PREPEND(m, sizeof(*ip6), M_NOWAIT); 448 if (m == NULL) { 449 error = ENOBUFS; 450 goto bad; 451 } 452 ip6 = mtod(m, struct ip6_hdr *); 453 454 /* 455 * Source address selection. 456 */ 457 error = in6_selectsrc_socket(dstsock, optp, in6p, so->so_cred, 458 scope_ambiguous, &in6a, &hlim); 459 460 if (error) 461 goto bad; 462 error = prison_check_ip6(in6p->inp_cred, &in6a); 463 if (error != 0) 464 goto bad; 465 ip6->ip6_src = in6a; 466 467 ip6->ip6_dst = dstsock->sin6_addr; 468 469 /* 470 * Fill in the rest of the IPv6 header fields. 471 */ 472 ip6->ip6_flow = (ip6->ip6_flow & ~IPV6_FLOWINFO_MASK) | 473 (in6p->inp_flow & IPV6_FLOWINFO_MASK); 474 ip6->ip6_vfc = (ip6->ip6_vfc & ~IPV6_VERSION_MASK) | 475 (IPV6_VERSION & IPV6_VERSION_MASK); 476 477 /* 478 * ip6_plen will be filled in ip6_output, so not fill it here. 479 */ 480 ip6->ip6_nxt = in6p->inp_ip_p; 481 ip6->ip6_hlim = hlim; 482 483 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6 || 484 in6p->in6p_cksum != -1) { 485 struct mbuf *n; 486 int off; 487 u_int16_t *p; 488 489 /* Compute checksum. */ 490 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) 491 off = offsetof(struct icmp6_hdr, icmp6_cksum); 492 else 493 off = in6p->in6p_cksum; 494 if (plen < off + 1) { 495 error = EINVAL; 496 goto bad; 497 } 498 off += sizeof(struct ip6_hdr); 499 500 n = m; 501 while (n && n->m_len <= off) { 502 off -= n->m_len; 503 n = n->m_next; 504 } 505 if (!n) 506 goto bad; 507 p = (u_int16_t *)(mtod(n, caddr_t) + off); 508 *p = 0; 509 *p = in6_cksum(m, ip6->ip6_nxt, sizeof(*ip6), plen); 510 } 511 512 /* 513 * Send RA/RS messages to user land for protection, before sending 514 * them to rtadvd/rtsol. 515 */ 516 if ((send_sendso_input_hook != NULL) && 517 so->so_proto->pr_protocol == IPPROTO_ICMPV6) { 518 switch (type) { 519 case ND_ROUTER_ADVERT: 520 case ND_ROUTER_SOLICIT: 521 mtag = m_tag_get(PACKET_TAG_ND_OUTGOING, 522 sizeof(unsigned short), M_NOWAIT); 523 if (mtag == NULL) 524 goto bad; 525 m_tag_prepend(m, mtag); 526 } 527 } 528 529 error = ip6_output(m, optp, NULL, 0, in6p->in6p_moptions, &oifp, in6p); 530 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) { 531 if (oifp) 532 icmp6_ifoutstat_inc(oifp, type, code); 533 ICMP6STAT_INC(icp6s_outhist[type]); 534 } else 535 RIP6STAT_INC(rip6s_opackets); 536 537 goto freectl; 538 539 bad: 540 if (m) 541 m_freem(m); 542 543 freectl: 544 if (control != NULL) { 545 ip6_clearpktopts(&opt, -1); 546 m_freem(control); 547 } 548 INP_WUNLOCK(in6p); 549 return (error); 550 } 551 552 /* 553 * Raw IPv6 socket option processing. 554 */ 555 int 556 rip6_ctloutput(struct socket *so, struct sockopt *sopt) 557 { 558 struct inpcb *inp; 559 int error; 560 561 if (sopt->sopt_level == IPPROTO_ICMPV6) 562 /* 563 * XXX: is it better to call icmp6_ctloutput() directly 564 * from protosw? 565 */ 566 return (icmp6_ctloutput(so, sopt)); 567 else if (sopt->sopt_level != IPPROTO_IPV6) { 568 if (sopt->sopt_level == SOL_SOCKET && 569 sopt->sopt_name == SO_SETFIB) { 570 inp = sotoinpcb(so); 571 INP_WLOCK(inp); 572 inp->inp_inc.inc_fibnum = so->so_fibnum; 573 INP_WUNLOCK(inp); 574 return (0); 575 } 576 return (EINVAL); 577 } 578 579 error = 0; 580 581 switch (sopt->sopt_dir) { 582 case SOPT_GET: 583 switch (sopt->sopt_name) { 584 case MRT6_INIT: 585 case MRT6_DONE: 586 case MRT6_ADD_MIF: 587 case MRT6_DEL_MIF: 588 case MRT6_ADD_MFC: 589 case MRT6_DEL_MFC: 590 case MRT6_PIM: 591 error = ip6_mrouter_get ? ip6_mrouter_get(so, sopt) : 592 EOPNOTSUPP; 593 break; 594 case IPV6_CHECKSUM: 595 error = ip6_raw_ctloutput(so, sopt); 596 break; 597 default: 598 error = ip6_ctloutput(so, sopt); 599 break; 600 } 601 break; 602 603 case SOPT_SET: 604 switch (sopt->sopt_name) { 605 case MRT6_INIT: 606 case MRT6_DONE: 607 case MRT6_ADD_MIF: 608 case MRT6_DEL_MIF: 609 case MRT6_ADD_MFC: 610 case MRT6_DEL_MFC: 611 case MRT6_PIM: 612 error = ip6_mrouter_set ? ip6_mrouter_set(so, sopt) : 613 EOPNOTSUPP; 614 break; 615 case IPV6_CHECKSUM: 616 error = ip6_raw_ctloutput(so, sopt); 617 break; 618 default: 619 error = ip6_ctloutput(so, sopt); 620 break; 621 } 622 break; 623 } 624 625 return (error); 626 } 627 628 static int 629 rip6_attach(struct socket *so, int proto, struct thread *td) 630 { 631 struct inpcb *inp; 632 struct icmp6_filter *filter; 633 int error; 634 635 inp = sotoinpcb(so); 636 KASSERT(inp == NULL, ("rip6_attach: inp != NULL")); 637 638 error = priv_check(td, PRIV_NETINET_RAW); 639 if (error) 640 return (error); 641 error = soreserve(so, rip_sendspace, rip_recvspace); 642 if (error) 643 return (error); 644 filter = malloc(sizeof(struct icmp6_filter), M_PCB, M_NOWAIT); 645 if (filter == NULL) 646 return (ENOMEM); 647 INP_INFO_WLOCK(&V_ripcbinfo); 648 error = in_pcballoc(so, &V_ripcbinfo); 649 if (error) { 650 INP_INFO_WUNLOCK(&V_ripcbinfo); 651 free(filter, M_PCB); 652 return (error); 653 } 654 inp = (struct inpcb *)so->so_pcb; 655 INP_INFO_WUNLOCK(&V_ripcbinfo); 656 inp->inp_vflag |= INP_IPV6; 657 inp->inp_ip_p = (long)proto; 658 inp->in6p_hops = -1; /* use kernel default */ 659 inp->in6p_cksum = -1; 660 inp->in6p_icmp6filt = filter; 661 ICMP6_FILTER_SETPASSALL(inp->in6p_icmp6filt); 662 INP_WUNLOCK(inp); 663 return (0); 664 } 665 666 static void 667 rip6_detach(struct socket *so) 668 { 669 struct inpcb *inp; 670 671 inp = sotoinpcb(so); 672 KASSERT(inp != NULL, ("rip6_detach: inp == NULL")); 673 674 if (so == V_ip6_mrouter && ip6_mrouter_done) 675 ip6_mrouter_done(); 676 /* xxx: RSVP */ 677 INP_INFO_WLOCK(&V_ripcbinfo); 678 INP_WLOCK(inp); 679 free(inp->in6p_icmp6filt, M_PCB); 680 in_pcbdetach(inp); 681 in_pcbfree(inp); 682 INP_INFO_WUNLOCK(&V_ripcbinfo); 683 } 684 685 /* XXXRW: This can't ever be called. */ 686 static void 687 rip6_abort(struct socket *so) 688 { 689 struct inpcb *inp; 690 691 inp = sotoinpcb(so); 692 KASSERT(inp != NULL, ("rip6_abort: inp == NULL")); 693 694 soisdisconnected(so); 695 } 696 697 static void 698 rip6_close(struct socket *so) 699 { 700 struct inpcb *inp; 701 702 inp = sotoinpcb(so); 703 KASSERT(inp != NULL, ("rip6_close: inp == NULL")); 704 705 soisdisconnected(so); 706 } 707 708 static int 709 rip6_disconnect(struct socket *so) 710 { 711 struct inpcb *inp; 712 713 inp = sotoinpcb(so); 714 KASSERT(inp != NULL, ("rip6_disconnect: inp == NULL")); 715 716 if ((so->so_state & SS_ISCONNECTED) == 0) 717 return (ENOTCONN); 718 inp->in6p_faddr = in6addr_any; 719 rip6_abort(so); 720 return (0); 721 } 722 723 static int 724 rip6_bind(struct socket *so, struct sockaddr *nam, struct thread *td) 725 { 726 struct inpcb *inp; 727 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam; 728 struct ifaddr *ifa = NULL; 729 int error = 0; 730 731 inp = sotoinpcb(so); 732 KASSERT(inp != NULL, ("rip6_bind: inp == NULL")); 733 734 if (nam->sa_len != sizeof(*addr)) 735 return (EINVAL); 736 if ((error = prison_check_ip6(td->td_ucred, &addr->sin6_addr)) != 0) 737 return (error); 738 if (CK_STAILQ_EMPTY(&V_ifnet) || addr->sin6_family != AF_INET6) 739 return (EADDRNOTAVAIL); 740 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0) 741 return (error); 742 743 NET_EPOCH_ENTER(); 744 if (!IN6_IS_ADDR_UNSPECIFIED(&addr->sin6_addr) && 745 (ifa = ifa_ifwithaddr((struct sockaddr *)addr)) == NULL) { 746 NET_EPOCH_EXIT(); 747 return (EADDRNOTAVAIL); 748 } 749 if (ifa != NULL && 750 ((struct in6_ifaddr *)ifa)->ia6_flags & 751 (IN6_IFF_ANYCAST|IN6_IFF_NOTREADY| 752 IN6_IFF_DETACHED|IN6_IFF_DEPRECATED)) { 753 NET_EPOCH_EXIT(); 754 return (EADDRNOTAVAIL); 755 } 756 NET_EPOCH_EXIT(); 757 INP_INFO_WLOCK(&V_ripcbinfo); 758 INP_WLOCK(inp); 759 inp->in6p_laddr = addr->sin6_addr; 760 INP_WUNLOCK(inp); 761 INP_INFO_WUNLOCK(&V_ripcbinfo); 762 return (0); 763 } 764 765 static int 766 rip6_connect(struct socket *so, struct sockaddr *nam, struct thread *td) 767 { 768 struct inpcb *inp; 769 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam; 770 struct in6_addr in6a; 771 int error = 0, scope_ambiguous = 0; 772 773 inp = sotoinpcb(so); 774 KASSERT(inp != NULL, ("rip6_connect: inp == NULL")); 775 776 if (nam->sa_len != sizeof(*addr)) 777 return (EINVAL); 778 if (CK_STAILQ_EMPTY(&V_ifnet)) 779 return (EADDRNOTAVAIL); 780 if (addr->sin6_family != AF_INET6) 781 return (EAFNOSUPPORT); 782 783 /* 784 * Application should provide a proper zone ID or the use of default 785 * zone IDs should be enabled. Unfortunately, some applications do 786 * not behave as it should, so we need a workaround. Even if an 787 * appropriate ID is not determined, we'll see if we can determine 788 * the outgoing interface. If we can, determine the zone ID based on 789 * the interface below. 790 */ 791 if (addr->sin6_scope_id == 0 && !V_ip6_use_defzone) 792 scope_ambiguous = 1; 793 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0) 794 return (error); 795 796 INP_INFO_WLOCK(&V_ripcbinfo); 797 INP_WLOCK(inp); 798 /* Source address selection. XXX: need pcblookup? */ 799 error = in6_selectsrc_socket(addr, inp->in6p_outputopts, 800 inp, so->so_cred, scope_ambiguous, &in6a, NULL); 801 if (error) { 802 INP_WUNLOCK(inp); 803 INP_INFO_WUNLOCK(&V_ripcbinfo); 804 return (error); 805 } 806 807 inp->in6p_faddr = addr->sin6_addr; 808 inp->in6p_laddr = in6a; 809 soisconnected(so); 810 INP_WUNLOCK(inp); 811 INP_INFO_WUNLOCK(&V_ripcbinfo); 812 return (0); 813 } 814 815 static int 816 rip6_shutdown(struct socket *so) 817 { 818 struct inpcb *inp; 819 820 inp = sotoinpcb(so); 821 KASSERT(inp != NULL, ("rip6_shutdown: inp == NULL")); 822 823 INP_WLOCK(inp); 824 socantsendmore(so); 825 INP_WUNLOCK(inp); 826 return (0); 827 } 828 829 static int 830 rip6_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam, 831 struct mbuf *control, struct thread *td) 832 { 833 struct inpcb *inp; 834 struct sockaddr_in6 tmp; 835 struct sockaddr_in6 *dst; 836 int ret; 837 838 inp = sotoinpcb(so); 839 KASSERT(inp != NULL, ("rip6_send: inp == NULL")); 840 841 /* Always copy sockaddr to avoid overwrites. */ 842 /* Unlocked read. */ 843 if (so->so_state & SS_ISCONNECTED) { 844 if (nam) { 845 m_freem(m); 846 return (EISCONN); 847 } 848 /* XXX */ 849 bzero(&tmp, sizeof(tmp)); 850 tmp.sin6_family = AF_INET6; 851 tmp.sin6_len = sizeof(struct sockaddr_in6); 852 INP_RLOCK(inp); 853 bcopy(&inp->in6p_faddr, &tmp.sin6_addr, 854 sizeof(struct in6_addr)); 855 INP_RUNLOCK(inp); 856 dst = &tmp; 857 } else { 858 if (nam == NULL) { 859 m_freem(m); 860 return (ENOTCONN); 861 } 862 if (nam->sa_len != sizeof(struct sockaddr_in6)) { 863 m_freem(m); 864 return (EINVAL); 865 } 866 tmp = *(struct sockaddr_in6 *)nam; 867 dst = &tmp; 868 869 if (dst->sin6_family == AF_UNSPEC) { 870 /* 871 * XXX: we allow this case for backward 872 * compatibility to buggy applications that 873 * rely on old (and wrong) kernel behavior. 874 */ 875 log(LOG_INFO, "rip6 SEND: address family is " 876 "unspec. Assume AF_INET6\n"); 877 dst->sin6_family = AF_INET6; 878 } else if (dst->sin6_family != AF_INET6) { 879 m_freem(m); 880 return(EAFNOSUPPORT); 881 } 882 } 883 ret = rip6_output(m, so, dst, control); 884 return (ret); 885 } 886 887 struct pr_usrreqs rip6_usrreqs = { 888 .pru_abort = rip6_abort, 889 .pru_attach = rip6_attach, 890 .pru_bind = rip6_bind, 891 .pru_connect = rip6_connect, 892 .pru_control = in6_control, 893 .pru_detach = rip6_detach, 894 .pru_disconnect = rip6_disconnect, 895 .pru_peeraddr = in6_getpeeraddr, 896 .pru_send = rip6_send, 897 .pru_shutdown = rip6_shutdown, 898 .pru_sockaddr = in6_getsockaddr, 899 .pru_close = rip6_close, 900 }; 901