raw_ip.c revision 1.76
1 /* $NetBSD: raw_ip.c,v 1.76 2003/09/06 03:36:31 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. Neither the name of the University nor the names of its contributors 45 * may be used to endorse or promote products derived from this software 46 * without specific prior written permission. 47 * 48 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 49 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 50 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 51 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 52 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 53 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 54 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 55 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 56 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 57 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 58 * SUCH DAMAGE. 59 * 60 * @(#)raw_ip.c 8.7 (Berkeley) 5/15/95 61 */ 62 63 #include <sys/cdefs.h> 64 __KERNEL_RCSID(0, "$NetBSD: raw_ip.c,v 1.76 2003/09/06 03:36:31 itojun Exp $"); 65 66 #include "opt_ipsec.h" 67 #include "opt_mrouting.h" 68 69 #include <sys/param.h> 70 #include <sys/malloc.h> 71 #include <sys/mbuf.h> 72 #include <sys/socket.h> 73 #include <sys/protosw.h> 74 #include <sys/socketvar.h> 75 #include <sys/errno.h> 76 #include <sys/systm.h> 77 #include <sys/proc.h> 78 79 #include <net/if.h> 80 #include <net/route.h> 81 82 #include <netinet/in.h> 83 #include <netinet/in_systm.h> 84 #include <netinet/ip.h> 85 #include <netinet/ip_var.h> 86 #include <netinet/ip_mroute.h> 87 #include <netinet/ip_icmp.h> 88 #include <netinet/in_pcb.h> 89 #include <netinet/in_var.h> 90 91 #include <machine/stdarg.h> 92 93 #ifdef IPSEC 94 #include <netinet6/ipsec.h> 95 #endif /*IPSEC*/ 96 97 #ifdef FAST_IPSEC 98 #include <netipsec/ipsec.h> 99 #endif /* FAST_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 /* 130 * Setup generic address and protocol structures 131 * for raw_input routine, then pass them along with 132 * mbuf chain. 133 */ 134 void 135 #if __STDC__ 136 rip_input(struct mbuf *m, ...) 137 #else 138 rip_input(m, va_alist) 139 struct mbuf *m; 140 va_dcl 141 #endif 142 { 143 int proto; 144 struct ip *ip = mtod(m, struct ip *); 145 struct inpcb_hdr *inph; 146 struct inpcb *inp; 147 struct inpcb *last = 0; 148 struct mbuf *opts = 0; 149 struct sockaddr_in ripsrc; 150 va_list ap; 151 152 va_start(ap, m); 153 (void)va_arg(ap, int); /* ignore value, advance ap */ 154 proto = va_arg(ap, int); 155 va_end(ap); 156 157 ripsrc.sin_family = AF_INET; 158 ripsrc.sin_len = sizeof(struct sockaddr_in); 159 ripsrc.sin_addr = ip->ip_src; 160 ripsrc.sin_port = 0; 161 bzero((caddr_t)ripsrc.sin_zero, sizeof(ripsrc.sin_zero)); 162 163 /* 164 * XXX Compatibility: programs using raw IP expect ip_len 165 * XXX to have the header length subtracted, and in host order. 166 * XXX ip_off is also expected to be host order. 167 */ 168 ip->ip_len = ntohs(ip->ip_len) - (ip->ip_hl << 2); 169 NTOHS(ip->ip_off); 170 171 CIRCLEQ_FOREACH(inph, &rawcbtable.inpt_queue, inph_queue) { 172 inp = (struct inpcb *)inph; 173 if (inp->inp_af != AF_INET) 174 continue; 175 if (inp->inp_ip.ip_p && inp->inp_ip.ip_p != proto) 176 continue; 177 if (!in_nullhost(inp->inp_laddr) && 178 !in_hosteq(inp->inp_laddr, ip->ip_dst)) 179 continue; 180 if (!in_nullhost(inp->inp_faddr) && 181 !in_hosteq(inp->inp_faddr, ip->ip_src)) 182 continue; 183 if (last) { 184 struct mbuf *n; 185 186 #if defined(IPSEC) || defined(FAST_IPSEC) 187 /* check AH/ESP integrity. */ 188 if (ipsec4_in_reject_so(m, last->inp_socket)) { 189 ipsecstat.in_polvio++; 190 /* do not inject data to pcb */ 191 } else 192 #endif /*IPSEC*/ 193 if ((n = m_copy(m, 0, (int)M_COPYALL)) != NULL) { 194 if (last->inp_flags & INP_CONTROLOPTS || 195 last->inp_socket->so_options & SO_TIMESTAMP) 196 ip_savecontrol(last, &opts, ip, n); 197 if (sbappendaddr(&last->inp_socket->so_rcv, 198 sintosa(&ripsrc), n, opts) == 0) { 199 /* should notify about lost packet */ 200 m_freem(n); 201 if (opts) 202 m_freem(opts); 203 } else 204 sorwakeup(last->inp_socket); 205 opts = NULL; 206 } 207 } 208 last = inp; 209 } 210 #if defined(IPSEC) || defined(FAST_IPSEC) 211 /* check AH/ESP integrity. */ 212 if (last && ipsec4_in_reject_so(m, last->inp_socket)) { 213 m_freem(m); 214 ipsecstat.in_polvio++; 215 ipstat.ips_delivered--; 216 /* do not inject data to pcb */ 217 } else 218 #endif /*IPSEC*/ 219 if (last) { 220 if (last->inp_flags & INP_CONTROLOPTS || 221 last->inp_socket->so_options & SO_TIMESTAMP) 222 ip_savecontrol(last, &opts, ip, m); 223 if (sbappendaddr(&last->inp_socket->so_rcv, 224 sintosa(&ripsrc), m, opts) == 0) { 225 m_freem(m); 226 if (opts) 227 m_freem(opts); 228 } else 229 sorwakeup(last->inp_socket); 230 } else { 231 if (inetsw[ip_protox[ip->ip_p]].pr_input == rip_input) { 232 icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_PROTOCOL, 233 0, 0); 234 ipstat.ips_noproto++; 235 ipstat.ips_delivered--; 236 } else 237 m_freem(m); 238 } 239 return; 240 } 241 242 int 243 rip_pcbnotify(table, faddr, laddr, proto, errno, notify) 244 struct inpcbtable *table; 245 struct in_addr faddr, laddr; 246 int proto; 247 int errno; 248 void (*notify) __P((struct inpcb *, int)); 249 { 250 struct inpcb *inp, *ninp; 251 int nmatch; 252 253 nmatch = 0; 254 for (inp = (struct inpcb *)CIRCLEQ_FIRST(&table->inpt_queue); 255 inp != (struct inpcb *)&table->inpt_queue; 256 inp = ninp) { 257 ninp = (struct inpcb *)inp->inp_queue.cqe_next; 258 if (inp->inp_af != AF_INET) 259 continue; 260 if (inp->inp_ip.ip_p && inp->inp_ip.ip_p != proto) 261 continue; 262 if (in_hosteq(inp->inp_faddr, faddr) && 263 in_hosteq(inp->inp_laddr, laddr)) { 264 (*notify)(inp, errno); 265 nmatch++; 266 } 267 } 268 269 return nmatch; 270 } 271 272 void * 273 rip_ctlinput(cmd, sa, v) 274 int cmd; 275 struct sockaddr *sa; 276 void *v; 277 { 278 struct ip *ip = v; 279 void (*notify) __P((struct inpcb *, int)) = in_rtchange; 280 int errno; 281 282 if (sa->sa_family != AF_INET || 283 sa->sa_len != sizeof(struct sockaddr_in)) 284 return NULL; 285 if ((unsigned)cmd >= PRC_NCMDS) 286 return NULL; 287 errno = inetctlerrmap[cmd]; 288 if (PRC_IS_REDIRECT(cmd)) 289 notify = in_rtchange, ip = 0; 290 else if (cmd == PRC_HOSTDEAD) 291 ip = 0; 292 else if (errno == 0) 293 return NULL; 294 if (ip) { 295 rip_pcbnotify(&rawcbtable, satosin(sa)->sin_addr, 296 ip->ip_src, ip->ip_p, errno, notify); 297 298 /* XXX mapped address case */ 299 } else 300 in_pcbnotifyall(&rawcbtable, satosin(sa)->sin_addr, errno, 301 notify); 302 return NULL; 303 } 304 305 /* 306 * Generate IP header and pass packet to ip_output. 307 * Tack on options user may have setup with control call. 308 */ 309 int 310 #if __STDC__ 311 rip_output(struct mbuf *m, ...) 312 #else 313 rip_output(m, va_alist) 314 struct mbuf *m; 315 va_dcl 316 #endif 317 { 318 struct inpcb *inp; 319 struct ip *ip; 320 struct mbuf *opts; 321 int flags; 322 va_list ap; 323 324 va_start(ap, m); 325 inp = va_arg(ap, struct inpcb *); 326 va_end(ap); 327 328 flags = 329 (inp->inp_socket->so_options & SO_DONTROUTE) | IP_ALLOWBROADCAST 330 | IP_RETURNMTU; 331 332 /* 333 * If the user handed us a complete IP packet, use it. 334 * Otherwise, allocate an mbuf for a header and fill it in. 335 */ 336 if ((inp->inp_flags & INP_HDRINCL) == 0) { 337 if ((m->m_pkthdr.len + sizeof(struct ip)) > IP_MAXPACKET) { 338 m_freem(m); 339 return (EMSGSIZE); 340 } 341 M_PREPEND(m, sizeof(struct ip), M_DONTWAIT); 342 if (!m) 343 return (ENOBUFS); 344 ip = mtod(m, struct ip *); 345 ip->ip_tos = 0; 346 ip->ip_off = htons(0); 347 ip->ip_p = inp->inp_ip.ip_p; 348 ip->ip_len = htons(m->m_pkthdr.len); 349 ip->ip_src = inp->inp_laddr; 350 ip->ip_dst = inp->inp_faddr; 351 ip->ip_ttl = MAXTTL; 352 opts = inp->inp_options; 353 } else { 354 if (m->m_pkthdr.len > IP_MAXPACKET) { 355 m_freem(m); 356 return (EMSGSIZE); 357 } 358 ip = mtod(m, struct ip *); 359 360 /* 361 * If the mbuf is read-only, we need to allocate 362 * a new mbuf for the header, since we need to 363 * modify the header. 364 */ 365 if (M_READONLY(m)) { 366 int hlen = ip->ip_hl << 2; 367 368 m = m_copyup(m, hlen, (max_linkhdr + 3) & ~3); 369 if (m == NULL) 370 return (ENOMEM); /* XXX */ 371 ip = mtod(m, struct ip *); 372 } 373 374 /* XXX userland passes ip_len and ip_off in host order */ 375 if (m->m_pkthdr.len != ip->ip_len) { 376 m_freem(m); 377 return (EINVAL); 378 } 379 HTONS(ip->ip_len); 380 HTONS(ip->ip_off); 381 if (ip->ip_id == 0) 382 ip->ip_id = htons(ip_randomid()); 383 opts = NULL; 384 /* XXX prevent ip_output from overwriting header fields */ 385 flags |= IP_RAWOUTPUT; 386 ipstat.ips_rawout++; 387 } 388 return (ip_output(m, opts, &inp->inp_route, flags, inp->inp_moptions, 389 inp->inp_socket, &inp->inp_errormtu)); 390 } 391 392 /* 393 * Raw IP socket option processing. 394 */ 395 int 396 rip_ctloutput(op, so, level, optname, m) 397 int op; 398 struct socket *so; 399 int level, optname; 400 struct mbuf **m; 401 { 402 struct inpcb *inp = sotoinpcb(so); 403 int error = 0; 404 405 if (level != IPPROTO_IP) { 406 error = ENOPROTOOPT; 407 if (op == PRCO_SETOPT && *m != 0) 408 (void) m_free(*m); 409 } else switch (op) { 410 411 case PRCO_SETOPT: 412 switch (optname) { 413 case IP_HDRINCL: 414 if (*m == 0 || (*m)->m_len < sizeof (int)) 415 error = EINVAL; 416 else { 417 if (*mtod(*m, int *)) 418 inp->inp_flags |= INP_HDRINCL; 419 else 420 inp->inp_flags &= ~INP_HDRINCL; 421 } 422 if (*m != 0) 423 (void) m_free(*m); 424 break; 425 426 #ifdef MROUTING 427 case MRT_INIT: 428 case MRT_DONE: 429 case MRT_ADD_VIF: 430 case MRT_DEL_VIF: 431 case MRT_ADD_MFC: 432 case MRT_DEL_MFC: 433 case MRT_ASSERT: 434 error = ip_mrouter_set(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 case PRCO_GETOPT: 445 switch (optname) { 446 case IP_HDRINCL: 447 *m = m_get(M_WAIT, MT_SOOPTS); 448 MCLAIM((*m), so->so_mowner); 449 (*m)->m_len = sizeof (int); 450 *mtod(*m, int *) = inp->inp_flags & INP_HDRINCL ? 1 : 0; 451 break; 452 453 #ifdef MROUTING 454 case MRT_VERSION: 455 case MRT_ASSERT: 456 error = ip_mrouter_get(so, optname, m); 457 break; 458 #endif 459 460 default: 461 error = ip_ctloutput(op, so, level, optname, m); 462 break; 463 } 464 break; 465 } 466 return (error); 467 } 468 469 int 470 rip_bind(inp, nam) 471 struct inpcb *inp; 472 struct mbuf *nam; 473 { 474 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *); 475 476 if (nam->m_len != sizeof(*addr)) 477 return (EINVAL); 478 if (TAILQ_FIRST(&ifnet) == 0) 479 return (EADDRNOTAVAIL); 480 if (addr->sin_family != AF_INET && 481 addr->sin_family != AF_IMPLINK) 482 return (EAFNOSUPPORT); 483 if (!in_nullhost(addr->sin_addr) && 484 ifa_ifwithaddr(sintosa(addr)) == 0) 485 return (EADDRNOTAVAIL); 486 inp->inp_laddr = addr->sin_addr; 487 return (0); 488 } 489 490 int 491 rip_connect(inp, nam) 492 struct inpcb *inp; 493 struct mbuf *nam; 494 { 495 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *); 496 497 if (nam->m_len != sizeof(*addr)) 498 return (EINVAL); 499 if (TAILQ_FIRST(&ifnet) == 0) 500 return (EADDRNOTAVAIL); 501 if (addr->sin_family != AF_INET && 502 addr->sin_family != AF_IMPLINK) 503 return (EAFNOSUPPORT); 504 inp->inp_faddr = addr->sin_addr; 505 return (0); 506 } 507 508 void 509 rip_disconnect(inp) 510 struct inpcb *inp; 511 { 512 513 inp->inp_faddr = zeroin_addr; 514 } 515 516 u_long rip_sendspace = RIPSNDQ; 517 u_long rip_recvspace = RIPRCVQ; 518 519 /*ARGSUSED*/ 520 int 521 rip_usrreq(so, req, m, nam, control, p) 522 struct socket *so; 523 int req; 524 struct mbuf *m, *nam, *control; 525 struct proc *p; 526 { 527 struct inpcb *inp; 528 int s; 529 int error = 0; 530 #ifdef MROUTING 531 extern struct socket *ip_mrouter; 532 #endif 533 534 if (req == PRU_CONTROL) 535 return (in_control(so, (long)m, (caddr_t)nam, 536 (struct ifnet *)control, p)); 537 538 if (req == PRU_PURGEIF) { 539 in_pcbpurgeif0(&rawcbtable, (struct ifnet *)control); 540 in_purgeif((struct ifnet *)control); 541 in_pcbpurgeif(&rawcbtable, (struct ifnet *)control); 542 return (0); 543 } 544 545 s = splsoftnet(); 546 inp = sotoinpcb(so); 547 #ifdef DIAGNOSTIC 548 if (req != PRU_SEND && req != PRU_SENDOOB && control) 549 panic("rip_usrreq: unexpected control mbuf"); 550 #endif 551 if (inp == 0 && req != PRU_ATTACH) { 552 error = EINVAL; 553 goto release; 554 } 555 556 switch (req) { 557 558 case PRU_ATTACH: 559 if (inp != 0) { 560 error = EISCONN; 561 break; 562 } 563 if (p == 0 || (error = suser(p->p_ucred, &p->p_acflag))) { 564 error = EACCES; 565 break; 566 } 567 if (so->so_snd.sb_hiwat == 0 || so->so_rcv.sb_hiwat == 0) { 568 error = soreserve(so, rip_sendspace, rip_recvspace); 569 if (error) 570 break; 571 } 572 error = in_pcballoc(so, &rawcbtable); 573 if (error) 574 break; 575 inp = sotoinpcb(so); 576 inp->inp_ip.ip_p = (long)nam; 577 break; 578 579 case PRU_DETACH: 580 #ifdef MROUTING 581 if (so == ip_mrouter) 582 ip_mrouter_done(); 583 #endif 584 in_pcbdetach(inp); 585 break; 586 587 case PRU_BIND: 588 error = rip_bind(inp, nam); 589 break; 590 591 case PRU_LISTEN: 592 error = EOPNOTSUPP; 593 break; 594 595 case PRU_CONNECT: 596 error = rip_connect(inp, nam); 597 if (error) 598 break; 599 soisconnected(so); 600 break; 601 602 case PRU_CONNECT2: 603 error = EOPNOTSUPP; 604 break; 605 606 case PRU_DISCONNECT: 607 soisdisconnected(so); 608 rip_disconnect(inp); 609 break; 610 611 /* 612 * Mark the connection as being incapable of further input. 613 */ 614 case PRU_SHUTDOWN: 615 socantsendmore(so); 616 break; 617 618 case PRU_RCVD: 619 error = EOPNOTSUPP; 620 break; 621 622 /* 623 * Ship a packet out. The appropriate raw output 624 * routine handles any massaging necessary. 625 */ 626 case PRU_SEND: 627 if (control && control->m_len) { 628 m_freem(control); 629 m_freem(m); 630 error = EINVAL; 631 break; 632 } 633 { 634 if (nam) { 635 if ((so->so_state & SS_ISCONNECTED) != 0) { 636 error = EISCONN; 637 goto die; 638 } 639 error = rip_connect(inp, nam); 640 if (error) { 641 die: 642 m_freem(m); 643 break; 644 } 645 } else { 646 if ((so->so_state & SS_ISCONNECTED) == 0) { 647 error = ENOTCONN; 648 goto die; 649 } 650 } 651 error = rip_output(m, inp); 652 if (nam) 653 rip_disconnect(inp); 654 } 655 break; 656 657 case PRU_SENSE: 658 /* 659 * stat: don't bother with a blocksize. 660 */ 661 splx(s); 662 return (0); 663 664 case PRU_RCVOOB: 665 error = EOPNOTSUPP; 666 break; 667 668 case PRU_SENDOOB: 669 m_freem(control); 670 m_freem(m); 671 error = EOPNOTSUPP; 672 break; 673 674 case PRU_SOCKADDR: 675 in_setsockaddr(inp, nam); 676 break; 677 678 case PRU_PEERADDR: 679 in_setpeeraddr(inp, nam); 680 break; 681 682 default: 683 panic("rip_usrreq"); 684 } 685 686 release: 687 splx(s); 688 return (error); 689 } 690
Indexes created Mon Sep 29 21:09:56 GMT 2025