1 /* $NetBSD: raw_ip.c,v 1.60 2001/12/21 02:51:47 itojun 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) 1982, 1986, 1988, 1993 34 * The Regents of the University of California. All rights reserved. 35 * 36 * Redistribution and use in source and binary forms, with or without 37 * modification, are permitted provided that the following conditions 38 * are met: 39 * 1. Redistributions of source code must retain the above copyright 40 * notice, this list of conditions and the following disclaimer. 41 * 2. Redistributions in binary form must reproduce the above copyright 42 * notice, this list of conditions and the following disclaimer in the 43 * documentation and/or other materials provided with the distribution. 44 * 3. All advertising materials mentioning features or use of this software 45 * must display the following acknowledgement: 46 * This product includes software developed by the University of 47 * California, Berkeley and its contributors. 48 * 4. Neither the name of the University nor the names of its contributors 49 * may be used to endorse or promote products derived from this software 50 * without specific prior written permission. 51 * 52 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 53 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 54 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 55 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 56 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 57 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 58 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 59 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 60 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 61 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 62 * SUCH DAMAGE. 63 * 64 * @(#)raw_ip.c 8.7 (Berkeley) 5/15/95 65 */ 66 67 #include <sys/cdefs.h> 68 __KERNEL_RCSID(0, "$NetBSD: raw_ip.c,v 1.60 2001/12/21 02:51:47 itojun Exp $"); 69 70 #include "opt_ipsec.h" 71 #include "opt_mrouting.h" 72 73 #include <sys/param.h> 74 #include <sys/malloc.h> 75 #include <sys/mbuf.h> 76 #include <sys/socket.h> 77 #include <sys/protosw.h> 78 #include <sys/socketvar.h> 79 #include <sys/errno.h> 80 #include <sys/systm.h> 81 #include <sys/proc.h> 82 83 #include <net/if.h> 84 #include <net/route.h> 85 86 #include <netinet/in.h> 87 #include <netinet/in_systm.h> 88 #include <netinet/ip.h> 89 #include <netinet/ip_var.h> 90 #include <netinet/ip_mroute.h> 91 #include <netinet/ip_icmp.h> 92 #include <netinet/in_pcb.h> 93 #include <netinet/in_var.h> 94 95 #include <machine/stdarg.h> 96 97 #ifdef IPSEC 98 #include <netinet6/ipsec.h> 99 #endif /*IPSEC*/ 100 101 struct inpcbtable rawcbtable; 102 103 int rip_pcbnotify __P((struct inpcbtable *, struct in_addr, 104 struct in_addr, int, int, void (*) __P((struct inpcb *, int)))); 105 int rip_bind __P((struct inpcb *, struct mbuf *)); 106 int rip_connect __P((struct inpcb *, struct mbuf *)); 107 void rip_disconnect __P((struct inpcb *)); 108 109 /* 110 * Nominal space allocated to a raw ip socket. 111 */ 112 #define RIPSNDQ 8192 113 #define RIPRCVQ 8192 114 115 /* 116 * Raw interface to IP protocol. 117 */ 118 119 /* 120 * Initialize raw connection block q. 121 */ 122 void 123 rip_init() 124 { 125 126 in_pcbinit(&rawcbtable, 1, 1); 127 } 128 129 static struct sockaddr_in ripsrc = { sizeof(ripsrc), AF_INET }; 130 131 /* 132 * Setup generic address and protocol structures 133 * for raw_input routine, then pass them along with 134 * mbuf chain. 135 */ 136 void 137 #if __STDC__ 138 rip_input(struct mbuf *m, ...) 139 #else 140 rip_input(m, va_alist) 141 struct mbuf *m; 142 va_dcl 143 #endif 144 { 145 int off, proto; 146 struct ip *ip = mtod(m, struct ip *); 147 struct inpcb *inp; 148 struct inpcb *last = 0; 149 struct mbuf *opts = 0; 150 struct sockaddr_in ripsrc; 151 va_list ap; 152 153 va_start(ap, m); 154 off = va_arg(ap, int); 155 proto = va_arg(ap, int); 156 va_end(ap); 157 158 ripsrc.sin_family = AF_INET; 159 ripsrc.sin_len = sizeof(struct sockaddr_in); 160 ripsrc.sin_addr = ip->ip_src; 161 ripsrc.sin_port = 0; 162 bzero((caddr_t)ripsrc.sin_zero, sizeof(ripsrc.sin_zero)); 163 164 /* 165 * XXX Compatibility: programs using raw IP expect ip_len 166 * XXX to have the header length subtracted. 167 */ 168 ip->ip_len -= ip->ip_hl << 2; 169 170 CIRCLEQ_FOREACH(inp, &rawcbtable.inpt_queue, inp_queue) { 171 if (inp->inp_ip.ip_p && inp->inp_ip.ip_p != proto) 172 continue; 173 if (!in_nullhost(inp->inp_laddr) && 174 !in_hosteq(inp->inp_laddr, ip->ip_dst)) 175 continue; 176 if (!in_nullhost(inp->inp_faddr) && 177 !in_hosteq(inp->inp_faddr, ip->ip_src)) 178 continue; 179 if (last) { 180 struct mbuf *n; 181 182 #ifdef IPSEC 183 /* check AH/ESP integrity. */ 184 if (ipsec4_in_reject_so(m, last->inp_socket)) { 185 ipsecstat.in_polvio++; 186 /* do not inject data to pcb */ 187 } else 188 #endif /*IPSEC*/ 189 if ((n = m_copy(m, 0, (int)M_COPYALL)) != NULL) { 190 if (last->inp_flags & INP_CONTROLOPTS || 191 last->inp_socket->so_options & SO_TIMESTAMP) 192 ip_savecontrol(last, &opts, ip, n); 193 if (sbappendaddr(&last->inp_socket->so_rcv, 194 sintosa(&ripsrc), n, opts) == 0) { 195 /* should notify about lost packet */ 196 m_freem(n); 197 if (opts) 198 m_freem(opts); 199 } else 200 sorwakeup(last->inp_socket); 201 opts = NULL; 202 } 203 } 204 last = inp; 205 } 206 #ifdef IPSEC 207 /* check AH/ESP integrity. */ 208 if (last && ipsec4_in_reject_so(m, last->inp_socket)) { 209 m_freem(m); 210 ipsecstat.in_polvio++; 211 ipstat.ips_delivered--; 212 /* do not inject data to pcb */ 213 } else 214 #endif /*IPSEC*/ 215 if (last) { 216 if (last->inp_flags & INP_CONTROLOPTS || 217 last->inp_socket->so_options & SO_TIMESTAMP) 218 ip_savecontrol(last, &opts, ip, m); 219 if (sbappendaddr(&last->inp_socket->so_rcv, 220 sintosa(&ripsrc), m, opts) == 0) { 221 m_freem(m); 222 if (opts) 223 m_freem(opts); 224 } else 225 sorwakeup(last->inp_socket); 226 } else { 227 if (inetsw[ip_protox[ip->ip_p]].pr_input == rip_input) { 228 icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_PROTOCOL, 229 0, 0); 230 ipstat.ips_noproto++; 231 ipstat.ips_delivered--; 232 } else 233 m_freem(m); 234 } 235 return; 236 } 237 238 int 239 rip_pcbnotify(table, faddr, laddr, proto, errno, notify) 240 struct inpcbtable *table; 241 struct in_addr faddr, laddr; 242 int proto; 243 int errno; 244 void (*notify) __P((struct inpcb *, int)); 245 { 246 struct inpcb *inp, *ninp; 247 int nmatch; 248 249 nmatch = 0; 250 for (inp = CIRCLEQ_FIRST(&table->inpt_queue); 251 inp != (struct inpcb *)&table->inpt_queue; 252 inp = ninp) { 253 ninp = inp->inp_queue.cqe_next; 254 if (inp->inp_ip.ip_p && inp->inp_ip.ip_p != proto) 255 continue; 256 if (in_hosteq(inp->inp_faddr, faddr) && 257 in_hosteq(inp->inp_laddr, laddr)) { 258 (*notify)(inp, errno); 259 nmatch++; 260 } 261 } 262 263 return nmatch; 264 } 265 266 void * 267 rip_ctlinput(cmd, sa, v) 268 int cmd; 269 struct sockaddr *sa; 270 void *v; 271 { 272 struct ip *ip = v; 273 void (*notify) __P((struct inpcb *, int)) = in_rtchange; 274 int errno; 275 276 if (sa->sa_family != AF_INET || 277 sa->sa_len != sizeof(struct sockaddr_in)) 278 return NULL; 279 if ((unsigned)cmd >= PRC_NCMDS) 280 return NULL; 281 errno = inetctlerrmap[cmd]; 282 if (PRC_IS_REDIRECT(cmd)) 283 notify = in_rtchange, ip = 0; 284 else if (cmd == PRC_HOSTDEAD) 285 ip = 0; 286 else if (errno == 0) 287 return NULL; 288 if (ip) { 289 rip_pcbnotify(&rawcbtable, satosin(sa)->sin_addr, 290 ip->ip_src, ip->ip_p, errno, notify); 291 292 /* XXX mapped address case */ 293 } else 294 in_pcbnotifyall(&rawcbtable, satosin(sa)->sin_addr, errno, 295 notify); 296 return NULL; 297 } 298 299 /* 300 * Generate IP header and pass packet to ip_output. 301 * Tack on options user may have setup with control call. 302 */ 303 int 304 #if __STDC__ 305 rip_output(struct mbuf *m, ...) 306 #else 307 rip_output(m, va_alist) 308 struct mbuf *m; 309 va_dcl 310 #endif 311 { 312 struct inpcb *inp; 313 struct ip *ip; 314 struct mbuf *opts; 315 int flags; 316 va_list ap; 317 318 va_start(ap, m); 319 inp = va_arg(ap, struct inpcb *); 320 va_end(ap); 321 322 flags = 323 (inp->inp_socket->so_options & SO_DONTROUTE) | IP_ALLOWBROADCAST 324 | IP_RETURNMTU; 325 326 /* 327 * If the user handed us a complete IP packet, use it. 328 * Otherwise, allocate an mbuf for a header and fill it in. 329 */ 330 if ((inp->inp_flags & INP_HDRINCL) == 0) { 331 if ((m->m_pkthdr.len + sizeof(struct ip)) > IP_MAXPACKET) { 332 m_freem(m); 333 return (EMSGSIZE); 334 } 335 M_PREPEND(m, sizeof(struct ip), M_WAIT); 336 ip = mtod(m, struct ip *); 337 ip->ip_tos = 0; 338 ip->ip_off = 0; 339 ip->ip_p = inp->inp_ip.ip_p; 340 ip->ip_len = m->m_pkthdr.len; 341 ip->ip_src = inp->inp_laddr; 342 ip->ip_dst = inp->inp_faddr; 343 ip->ip_ttl = MAXTTL; 344 opts = inp->inp_options; 345 } else { 346 if (m->m_pkthdr.len > IP_MAXPACKET) { 347 m_freem(m); 348 return (EMSGSIZE); 349 } 350 ip = mtod(m, struct ip *); 351 if (m->m_pkthdr.len != ip->ip_len) { 352 m_freem(m); 353 return (EINVAL); 354 } 355 if (ip->ip_id == 0) 356 ip->ip_id = htons(ip_id++); 357 opts = NULL; 358 /* XXX prevent ip_output from overwriting header fields */ 359 flags |= IP_RAWOUTPUT; 360 ipstat.ips_rawout++; 361 } 362 #ifdef IPSEC 363 if (ipsec_setsocket(m, inp->inp_socket) != 0) { 364 m_freem(m); 365 return ENOBUFS; 366 } 367 #endif /*IPSEC*/ 368 return (ip_output(m, opts, &inp->inp_route, flags, inp->inp_moptions, &inp->inp_errormtu)); 369 } 370 371 /* 372 * Raw IP socket option processing. 373 */ 374 int 375 rip_ctloutput(op, so, level, optname, m) 376 int op; 377 struct socket *so; 378 int level, optname; 379 struct mbuf **m; 380 { 381 struct inpcb *inp = sotoinpcb(so); 382 int error = 0; 383 384 if (level != IPPROTO_IP) { 385 error = ENOPROTOOPT; 386 if (op == PRCO_SETOPT && *m != 0) 387 (void) m_free(*m); 388 } else switch (op) { 389 390 case PRCO_SETOPT: 391 switch (optname) { 392 case IP_HDRINCL: 393 if (*m == 0 || (*m)->m_len < sizeof (int)) 394 error = EINVAL; 395 else { 396 if (*mtod(*m, int *)) 397 inp->inp_flags |= INP_HDRINCL; 398 else 399 inp->inp_flags &= ~INP_HDRINCL; 400 } 401 if (*m != 0) 402 (void) m_free(*m); 403 break; 404 405 #ifdef MROUTING 406 case MRT_INIT: 407 case MRT_DONE: 408 case MRT_ADD_VIF: 409 case MRT_DEL_VIF: 410 case MRT_ADD_MFC: 411 case MRT_DEL_MFC: 412 case MRT_ASSERT: 413 error = ip_mrouter_set(so, optname, m); 414 break; 415 #endif 416 417 default: 418 error = ip_ctloutput(op, so, level, optname, m); 419 break; 420 } 421 break; 422 423 case PRCO_GETOPT: 424 switch (optname) { 425 case IP_HDRINCL: 426 *m = m_get(M_WAIT, M_SOOPTS); 427 (*m)->m_len = sizeof (int); 428 *mtod(*m, int *) = inp->inp_flags & INP_HDRINCL ? 1 : 0; 429 break; 430 431 #ifdef MROUTING 432 case MRT_VERSION: 433 case MRT_ASSERT: 434 error = ip_mrouter_get(so, optname, m); 435 break; 436 #endif 437 438 default: 439 error = ip_ctloutput(op, so, level, optname, m); 440 break; 441 } 442 break; 443 } 444 return (error); 445 } 446 447 int 448 rip_bind(inp, nam) 449 struct inpcb *inp; 450 struct mbuf *nam; 451 { 452 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *); 453 454 if (nam->m_len != sizeof(*addr)) 455 return (EINVAL); 456 if (TAILQ_FIRST(&ifnet) == 0) 457 return (EADDRNOTAVAIL); 458 if (addr->sin_family != AF_INET && 459 addr->sin_family != AF_IMPLINK) 460 return (EAFNOSUPPORT); 461 if (!in_nullhost(addr->sin_addr) && 462 ifa_ifwithaddr(sintosa(addr)) == 0) 463 return (EADDRNOTAVAIL); 464 inp->inp_laddr = addr->sin_addr; 465 return (0); 466 } 467 468 int 469 rip_connect(inp, nam) 470 struct inpcb *inp; 471 struct mbuf *nam; 472 { 473 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *); 474 475 if (nam->m_len != sizeof(*addr)) 476 return (EINVAL); 477 if (TAILQ_FIRST(&ifnet) == 0) 478 return (EADDRNOTAVAIL); 479 if (addr->sin_family != AF_INET && 480 addr->sin_family != AF_IMPLINK) 481 return (EAFNOSUPPORT); 482 inp->inp_faddr = addr->sin_addr; 483 return (0); 484 } 485 486 void 487 rip_disconnect(inp) 488 struct inpcb *inp; 489 { 490 491 inp->inp_faddr = zeroin_addr; 492 } 493 494 u_long rip_sendspace = RIPSNDQ; 495 u_long rip_recvspace = RIPRCVQ; 496 497 /*ARGSUSED*/ 498 int 499 rip_usrreq(so, req, m, nam, control, p) 500 struct socket *so; 501 int req; 502 struct mbuf *m, *nam, *control; 503 struct proc *p; 504 { 505 struct inpcb *inp; 506 int s; 507 int error = 0; 508 #ifdef MROUTING 509 extern struct socket *ip_mrouter; 510 #endif 511 512 if (req == PRU_CONTROL) 513 return (in_control(so, (long)m, (caddr_t)nam, 514 (struct ifnet *)control, p)); 515 516 if (req == PRU_PURGEIF) { 517 in_pcbpurgeif0(&rawcbtable, (struct ifnet *)control); 518 in_purgeif((struct ifnet *)control); 519 in_pcbpurgeif(&rawcbtable, (struct ifnet *)control); 520 return (0); 521 } 522 523 s = splsoftnet(); 524 inp = sotoinpcb(so); 525 #ifdef DIAGNOSTIC 526 if (req != PRU_SEND && req != PRU_SENDOOB && control) 527 panic("rip_usrreq: unexpected control mbuf"); 528 #endif 529 if (inp == 0 && req != PRU_ATTACH) { 530 error = EINVAL; 531 goto release; 532 } 533 534 switch (req) { 535 536 case PRU_ATTACH: 537 if (inp != 0) { 538 error = EISCONN; 539 break; 540 } 541 if (p == 0 || (error = suser(p->p_ucred, &p->p_acflag))) { 542 error = EACCES; 543 break; 544 } 545 if (so->so_snd.sb_hiwat == 0 || so->so_rcv.sb_hiwat == 0) { 546 error = soreserve(so, rip_sendspace, rip_recvspace); 547 if (error) 548 break; 549 } 550 error = in_pcballoc(so, &rawcbtable); 551 if (error) 552 break; 553 inp = sotoinpcb(so); 554 inp->inp_ip.ip_p = (long)nam; 555 break; 556 557 case PRU_DETACH: 558 #ifdef MROUTING 559 if (so == ip_mrouter) 560 ip_mrouter_done(); 561 #endif 562 in_pcbdetach(inp); 563 break; 564 565 case PRU_BIND: 566 error = rip_bind(inp, nam); 567 break; 568 569 case PRU_LISTEN: 570 error = EOPNOTSUPP; 571 break; 572 573 case PRU_CONNECT: 574 error = rip_connect(inp, nam); 575 if (error) 576 break; 577 soisconnected(so); 578 break; 579 580 case PRU_CONNECT2: 581 error = EOPNOTSUPP; 582 break; 583 584 case PRU_DISCONNECT: 585 soisdisconnected(so); 586 rip_disconnect(inp); 587 break; 588 589 /* 590 * Mark the connection as being incapable of further input. 591 */ 592 case PRU_SHUTDOWN: 593 socantsendmore(so); 594 break; 595 596 case PRU_RCVD: 597 error = EOPNOTSUPP; 598 break; 599 600 /* 601 * Ship a packet out. The appropriate raw output 602 * routine handles any massaging necessary. 603 */ 604 case PRU_SEND: 605 if (control && control->m_len) { 606 m_freem(control); 607 m_freem(m); 608 error = EINVAL; 609 break; 610 } 611 { 612 if (nam) { 613 if ((so->so_state & SS_ISCONNECTED) != 0) { 614 error = EISCONN; 615 goto die; 616 } 617 error = rip_connect(inp, nam); 618 if (error) { 619 die: 620 m_freem(m); 621 break; 622 } 623 } else { 624 if ((so->so_state & SS_ISCONNECTED) == 0) { 625 error = ENOTCONN; 626 goto die; 627 } 628 } 629 error = rip_output(m, inp); 630 if (nam) 631 rip_disconnect(inp); 632 } 633 break; 634 635 case PRU_SENSE: 636 /* 637 * stat: don't bother with a blocksize. 638 */ 639 splx(s); 640 return (0); 641 642 case PRU_RCVOOB: 643 error = EOPNOTSUPP; 644 break; 645 646 case PRU_SENDOOB: 647 m_freem(control); 648 m_freem(m); 649 error = EOPNOTSUPP; 650 break; 651 652 case PRU_SOCKADDR: 653 in_setsockaddr(inp, nam); 654 break; 655 656 case PRU_PEERADDR: 657 in_setpeeraddr(inp, nam); 658 break; 659 660 default: 661 panic("rip_usrreq"); 662 } 663 664 release: 665 splx(s); 666 return (error); 667 } 668