raw_ip.c revision 1.74
1 /* $NetBSD: raw_ip.c,v 1.74 2003/08/22 22:00:37 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 1.74 2003/08/22 22:00:37 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 *inp; 146 struct inpcb *last = 0; 147 struct mbuf *opts = 0; 148 struct sockaddr_in ripsrc; 149 va_list ap; 150 151 va_start(ap, m); 152 (void)va_arg(ap, int); /* ignore value, advance ap */ 153 proto = va_arg(ap, int); 154 va_end(ap); 155 156 ripsrc.sin_family = AF_INET; 157 ripsrc.sin_len = sizeof(struct sockaddr_in); 158 ripsrc.sin_addr = ip->ip_src; 159 ripsrc.sin_port = 0; 160 bzero((caddr_t)ripsrc.sin_zero, sizeof(ripsrc.sin_zero)); 161 162 /* 163 * XXX Compatibility: programs using raw IP expect ip_len 164 * XXX to have the header length subtracted, and in host order. 165 * XXX ip_off is also expected to be host order. 166 */ 167 ip->ip_len = ntohs(ip->ip_len) - (ip->ip_hl << 2); 168 NTOHS(ip->ip_off); 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 #if defined(IPSEC) || defined(FAST_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 #if defined(IPSEC) || defined(FAST_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_DONTWAIT); 336 if (!m) 337 return (ENOBUFS); 338 ip = mtod(m, struct ip *); 339 ip->ip_tos = 0; 340 ip->ip_off = htons(0); 341 ip->ip_p = inp->inp_ip.ip_p; 342 ip->ip_len = htons(m->m_pkthdr.len); 343 ip->ip_src = inp->inp_laddr; 344 ip->ip_dst = inp->inp_faddr; 345 ip->ip_ttl = MAXTTL; 346 opts = inp->inp_options; 347 } else { 348 if (m->m_pkthdr.len > IP_MAXPACKET) { 349 m_freem(m); 350 return (EMSGSIZE); 351 } 352 ip = mtod(m, struct ip *); 353 354 /* 355 * If the mbuf is read-only, we need to allocate 356 * a new mbuf for the header, since we need to 357 * modify the header. 358 */ 359 if (M_READONLY(m)) { 360 int hlen = ip->ip_hl << 2; 361 362 m = m_copyup(m, hlen, (max_linkhdr + 3) & ~3); 363 if (m == NULL) 364 return (ENOMEM); /* XXX */ 365 ip = mtod(m, struct ip *); 366 } 367 368 /* XXX userland passes ip_len and ip_off in host order */ 369 if (m->m_pkthdr.len != ip->ip_len) { 370 m_freem(m); 371 return (EINVAL); 372 } 373 HTONS(ip->ip_len); 374 HTONS(ip->ip_off); 375 if (ip->ip_id == 0) 376 ip->ip_id = htons(ip_id++); 377 opts = NULL; 378 /* XXX prevent ip_output from overwriting header fields */ 379 flags |= IP_RAWOUTPUT; 380 ipstat.ips_rawout++; 381 } 382 return (ip_output(m, opts, &inp->inp_route, flags, inp->inp_moptions, 383 inp->inp_socket, &inp->inp_errormtu)); 384 } 385 386 /* 387 * Raw IP socket option processing. 388 */ 389 int 390 rip_ctloutput(op, so, level, optname, m) 391 int op; 392 struct socket *so; 393 int level, optname; 394 struct mbuf **m; 395 { 396 struct inpcb *inp = sotoinpcb(so); 397 int error = 0; 398 399 if (level != IPPROTO_IP) { 400 error = ENOPROTOOPT; 401 if (op == PRCO_SETOPT && *m != 0) 402 (void) m_free(*m); 403 } else switch (op) { 404 405 case PRCO_SETOPT: 406 switch (optname) { 407 case IP_HDRINCL: 408 if (*m == 0 || (*m)->m_len < sizeof (int)) 409 error = EINVAL; 410 else { 411 if (*mtod(*m, int *)) 412 inp->inp_flags |= INP_HDRINCL; 413 else 414 inp->inp_flags &= ~INP_HDRINCL; 415 } 416 if (*m != 0) 417 (void) m_free(*m); 418 break; 419 420 #ifdef MROUTING 421 case MRT_INIT: 422 case MRT_DONE: 423 case MRT_ADD_VIF: 424 case MRT_DEL_VIF: 425 case MRT_ADD_MFC: 426 case MRT_DEL_MFC: 427 case MRT_ASSERT: 428 error = ip_mrouter_set(so, optname, m); 429 break; 430 #endif 431 432 default: 433 error = ip_ctloutput(op, so, level, optname, m); 434 break; 435 } 436 break; 437 438 case PRCO_GETOPT: 439 switch (optname) { 440 case IP_HDRINCL: 441 *m = m_get(M_WAIT, MT_SOOPTS); 442 MCLAIM((*m), so->so_mowner); 443 (*m)->m_len = sizeof (int); 444 *mtod(*m, int *) = inp->inp_flags & INP_HDRINCL ? 1 : 0; 445 break; 446 447 #ifdef MROUTING 448 case MRT_VERSION: 449 case MRT_ASSERT: 450 error = ip_mrouter_get(so, optname, m); 451 break; 452 #endif 453 454 default: 455 error = ip_ctloutput(op, so, level, optname, m); 456 break; 457 } 458 break; 459 } 460 return (error); 461 } 462 463 int 464 rip_bind(inp, nam) 465 struct inpcb *inp; 466 struct mbuf *nam; 467 { 468 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *); 469 470 if (nam->m_len != sizeof(*addr)) 471 return (EINVAL); 472 if (TAILQ_FIRST(&ifnet) == 0) 473 return (EADDRNOTAVAIL); 474 if (addr->sin_family != AF_INET && 475 addr->sin_family != AF_IMPLINK) 476 return (EAFNOSUPPORT); 477 if (!in_nullhost(addr->sin_addr) && 478 ifa_ifwithaddr(sintosa(addr)) == 0) 479 return (EADDRNOTAVAIL); 480 inp->inp_laddr = addr->sin_addr; 481 return (0); 482 } 483 484 int 485 rip_connect(inp, nam) 486 struct inpcb *inp; 487 struct mbuf *nam; 488 { 489 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *); 490 491 if (nam->m_len != sizeof(*addr)) 492 return (EINVAL); 493 if (TAILQ_FIRST(&ifnet) == 0) 494 return (EADDRNOTAVAIL); 495 if (addr->sin_family != AF_INET && 496 addr->sin_family != AF_IMPLINK) 497 return (EAFNOSUPPORT); 498 inp->inp_faddr = addr->sin_addr; 499 return (0); 500 } 501 502 void 503 rip_disconnect(inp) 504 struct inpcb *inp; 505 { 506 507 inp->inp_faddr = zeroin_addr; 508 } 509 510 u_long rip_sendspace = RIPSNDQ; 511 u_long rip_recvspace = RIPRCVQ; 512 513 /*ARGSUSED*/ 514 int 515 rip_usrreq(so, req, m, nam, control, p) 516 struct socket *so; 517 int req; 518 struct mbuf *m, *nam, *control; 519 struct proc *p; 520 { 521 struct inpcb *inp; 522 int s; 523 int error = 0; 524 #ifdef MROUTING 525 extern struct socket *ip_mrouter; 526 #endif 527 528 if (req == PRU_CONTROL) 529 return (in_control(so, (long)m, (caddr_t)nam, 530 (struct ifnet *)control, p)); 531 532 if (req == PRU_PURGEIF) { 533 in_pcbpurgeif0(&rawcbtable, (struct ifnet *)control); 534 in_purgeif((struct ifnet *)control); 535 in_pcbpurgeif(&rawcbtable, (struct ifnet *)control); 536 return (0); 537 } 538 539 s = splsoftnet(); 540 inp = sotoinpcb(so); 541 #ifdef DIAGNOSTIC 542 if (req != PRU_SEND && req != PRU_SENDOOB && control) 543 panic("rip_usrreq: unexpected control mbuf"); 544 #endif 545 if (inp == 0 && req != PRU_ATTACH) { 546 error = EINVAL; 547 goto release; 548 } 549 550 switch (req) { 551 552 case PRU_ATTACH: 553 if (inp != 0) { 554 error = EISCONN; 555 break; 556 } 557 if (p == 0 || (error = suser(p->p_ucred, &p->p_acflag))) { 558 error = EACCES; 559 break; 560 } 561 if (so->so_snd.sb_hiwat == 0 || so->so_rcv.sb_hiwat == 0) { 562 error = soreserve(so, rip_sendspace, rip_recvspace); 563 if (error) 564 break; 565 } 566 error = in_pcballoc(so, &rawcbtable); 567 if (error) 568 break; 569 inp = sotoinpcb(so); 570 inp->inp_ip.ip_p = (long)nam; 571 break; 572 573 case PRU_DETACH: 574 #ifdef MROUTING 575 if (so == ip_mrouter) 576 ip_mrouter_done(); 577 #endif 578 in_pcbdetach(inp); 579 break; 580 581 case PRU_BIND: 582 error = rip_bind(inp, nam); 583 break; 584 585 case PRU_LISTEN: 586 error = EOPNOTSUPP; 587 break; 588 589 case PRU_CONNECT: 590 error = rip_connect(inp, nam); 591 if (error) 592 break; 593 soisconnected(so); 594 break; 595 596 case PRU_CONNECT2: 597 error = EOPNOTSUPP; 598 break; 599 600 case PRU_DISCONNECT: 601 soisdisconnected(so); 602 rip_disconnect(inp); 603 break; 604 605 /* 606 * Mark the connection as being incapable of further input. 607 */ 608 case PRU_SHUTDOWN: 609 socantsendmore(so); 610 break; 611 612 case PRU_RCVD: 613 error = EOPNOTSUPP; 614 break; 615 616 /* 617 * Ship a packet out. The appropriate raw output 618 * routine handles any massaging necessary. 619 */ 620 case PRU_SEND: 621 if (control && control->m_len) { 622 m_freem(control); 623 m_freem(m); 624 error = EINVAL; 625 break; 626 } 627 { 628 if (nam) { 629 if ((so->so_state & SS_ISCONNECTED) != 0) { 630 error = EISCONN; 631 goto die; 632 } 633 error = rip_connect(inp, nam); 634 if (error) { 635 die: 636 m_freem(m); 637 break; 638 } 639 } else { 640 if ((so->so_state & SS_ISCONNECTED) == 0) { 641 error = ENOTCONN; 642 goto die; 643 } 644 } 645 error = rip_output(m, inp); 646 if (nam) 647 rip_disconnect(inp); 648 } 649 break; 650 651 case PRU_SENSE: 652 /* 653 * stat: don't bother with a blocksize. 654 */ 655 splx(s); 656 return (0); 657 658 case PRU_RCVOOB: 659 error = EOPNOTSUPP; 660 break; 661 662 case PRU_SENDOOB: 663 m_freem(control); 664 m_freem(m); 665 error = EOPNOTSUPP; 666 break; 667 668 case PRU_SOCKADDR: 669 in_setsockaddr(inp, nam); 670 break; 671 672 case PRU_PEERADDR: 673 in_setpeeraddr(inp, nam); 674 break; 675 676 default: 677 panic("rip_usrreq"); 678 } 679 680 release: 681 splx(s); 682 return (error); 683 } 684
Indexes created Mon Sep 29 21:09:56 GMT 2025