1 /* $OpenBSD: traceroute.c,v 1.163 2020/02/11 18:41:39 deraadt Exp $ */ 2 /* $NetBSD: traceroute.c,v 1.10 1995/05/21 15:50:45 mycroft Exp $ */ 3 4 /* 5 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project. 6 * All rights reserved. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 3. Neither the name of the project nor the names of its contributors 17 * may be used to endorse or promote products derived from this software 18 * without specific prior written permission. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND 21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 23 * ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE 24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 30 * SUCH DAMAGE. 31 */ 32 33 /*- 34 * Copyright (c) 1990, 1993 35 * The Regents of the University of California. All rights reserved. 36 * 37 * This code is derived from software contributed to Berkeley by 38 * Van Jacobson. 39 * 40 * Redistribution and use in source and binary forms, with or without 41 * modification, are permitted provided that the following conditions 42 * are met: 43 * 1. Redistributions of source code must retain the above copyright 44 * notice, this list of conditions and the following disclaimer. 45 * 2. Redistributions in binary form must reproduce the above copyright 46 * notice, this list of conditions and the following disclaimer in the 47 * documentation and/or other materials provided with the distribution. 48 * 3. 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 65 /* 66 * traceroute host - trace the route ip packets follow going to "host". 67 * 68 * Attempt to trace the route an ip packet would follow to some 69 * internet host. We find out intermediate hops by launching probe 70 * packets with a small ttl (time to live) then listening for an 71 * icmp "time exceeded" reply from a gateway. We start our probes 72 * with a ttl of one and increase by one until we get an icmp "port 73 * unreachable" (which means we got to "host") or hit a max (which 74 * defaults to 64 hops & can be changed with the -m flag). Three 75 * probes (change with -q flag) are sent at each ttl setting and a 76 * line is printed showing the ttl, address of the gateway and 77 * round trip time of each probe. If the probe answers come from 78 * different gateways, the address of each responding system will 79 * be printed. If there is no response within a 5 sec. timeout 80 * interval (changed with the -w flag), a "*" is printed for that 81 * probe. 82 * 83 * Probe packets are UDP format. We don't want the destination 84 * host to process them so the destination port is set to an 85 * unlikely value (if some clod on the destination is using that 86 * value, it can be changed with the -p flag). 87 * 88 * A sample use might be: 89 * 90 * [yak 71]% traceroute nis.nsf.net. 91 * traceroute to nis.nsf.net (35.1.1.48), 64 hops max, 56 byte packet 92 * 1 helios.ee.lbl.gov (128.3.112.1) 19 ms 19 ms 0 ms 93 * 2 lilac-dmc.Berkeley.EDU (128.32.216.1) 39 ms 39 ms 19 ms 94 * 3 lilac-dmc.Berkeley.EDU (128.32.216.1) 39 ms 39 ms 19 ms 95 * 4 ccngw-ner-cc.Berkeley.EDU (128.32.136.23) 39 ms 40 ms 39 ms 96 * 5 ccn-nerif22.Berkeley.EDU (128.32.168.22) 39 ms 39 ms 39 ms 97 * 6 128.32.197.4 (128.32.197.4) 40 ms 59 ms 59 ms 98 * 7 131.119.2.5 (131.119.2.5) 59 ms 59 ms 59 ms 99 * 8 129.140.70.13 (129.140.70.13) 99 ms 99 ms 80 ms 100 * 9 129.140.71.6 (129.140.71.6) 139 ms 239 ms 319 ms 101 * 10 129.140.81.7 (129.140.81.7) 220 ms 199 ms 199 ms 102 * 11 nic.merit.edu (35.1.1.48) 239 ms 239 ms 239 ms 103 * 104 * Note that lines 2 & 3 are the same. This is due to a buggy 105 * kernel on the 2nd hop system -- lbl-csam.arpa -- that forwards 106 * packets with a zero ttl. 107 * 108 * A more interesting example is: 109 * 110 * [yak 72]% traceroute allspice.lcs.mit.edu. 111 * traceroute to allspice.lcs.mit.edu (18.26.0.115), 64 hops max 112 * 1 helios.ee.lbl.gov (128.3.112.1) 0 ms 0 ms 0 ms 113 * 2 lilac-dmc.Berkeley.EDU (128.32.216.1) 19 ms 19 ms 19 ms 114 * 3 lilac-dmc.Berkeley.EDU (128.32.216.1) 39 ms 19 ms 19 ms 115 * 4 ccngw-ner-cc.Berkeley.EDU (128.32.136.23) 19 ms 39 ms 39 ms 116 * 5 ccn-nerif22.Berkeley.EDU (128.32.168.22) 20 ms 39 ms 39 ms 117 * 6 128.32.197.4 (128.32.197.4) 59 ms 119 ms 39 ms 118 * 7 131.119.2.5 (131.119.2.5) 59 ms 59 ms 39 ms 119 * 8 129.140.70.13 (129.140.70.13) 80 ms 79 ms 99 ms 120 * 9 129.140.71.6 (129.140.71.6) 139 ms 139 ms 159 ms 121 * 10 129.140.81.7 (129.140.81.7) 199 ms 180 ms 300 ms 122 * 11 129.140.72.17 (129.140.72.17) 300 ms 239 ms 239 ms 123 * 12 * * * 124 * 13 128.121.54.72 (128.121.54.72) 259 ms 499 ms 279 ms 125 * 14 * * * 126 * 15 * * * 127 * 16 * * * 128 * 17 * * * 129 * 18 ALLSPICE.LCS.MIT.EDU (18.26.0.115) 339 ms 279 ms 279 ms 130 * 131 * (I start to see why I'm having so much trouble with mail to 132 * MIT.) Note that the gateways 12, 14, 15, 16 & 17 hops away 133 * either don't send ICMP "time exceeded" messages or send them 134 * with a ttl too small to reach us. 14 - 17 are running the 135 * MIT C Gateway code that doesn't send "time exceeded"s. God 136 * only knows what's going on with 12. 137 * 138 * The silent gateway 12 in the above may be the result of a bug in 139 * the 4.[23]BSD network code (and its derivatives): 4.x (x <= 3) 140 * sends an unreachable message using whatever ttl remains in the 141 * original datagram. Since, for gateways, the remaining ttl is 142 * zero, the icmp "time exceeded" is guaranteed to not make it back 143 * to us. The behavior of this bug is slightly more interesting 144 * when it appears on the destination system: 145 * 146 * 1 helios.ee.lbl.gov (128.3.112.1) 0 ms 0 ms 0 ms 147 * 2 lilac-dmc.Berkeley.EDU (128.32.216.1) 39 ms 19 ms 39 ms 148 * 3 lilac-dmc.Berkeley.EDU (128.32.216.1) 19 ms 39 ms 19 ms 149 * 4 ccngw-ner-cc.Berkeley.EDU (128.32.136.23) 39 ms 40 ms 19 ms 150 * 5 ccn-nerif35.Berkeley.EDU (128.32.168.35) 39 ms 39 ms 39 ms 151 * 6 csgw.Berkeley.EDU (128.32.133.254) 39 ms 59 ms 39 ms 152 * 7 * * * 153 * 8 * * * 154 * 9 * * * 155 * 10 * * * 156 * 11 * * * 157 * 12 * * * 158 * 13 rip.Berkeley.EDU (128.32.131.22) 59 ms ! 39 ms ! 39 ms ! 159 * 160 * Notice that there are 12 "gateways" (13 is the final 161 * destination) and exactly the last half of them are "missing". 162 * What's really happening is that rip (a Sun-3 running Sun OS3.5) 163 * is using the ttl from our arriving datagram as the ttl in its 164 * icmp reply. So, the reply will time out on the return path 165 * (with no notice sent to anyone since icmp's aren't sent for 166 * icmp's) until we probe with a ttl that's at least twice the path 167 * length. I.e., rip is really only 7 hops away. A reply that 168 * returns with a ttl of 1 is a clue this problem exists. 169 * Traceroute prints a "!" after the time if the ttl is <= 1. 170 * Since vendors ship a lot of obsolete (DEC's Ultrix, Sun 3.x) or 171 * non-standard (HPUX) software, expect to see this problem 172 * frequently and/or take care picking the target host of your 173 * probes. 174 * 175 * Other possible annotations after the time are !H, !N, !P (got a host, 176 * network or protocol unreachable, respectively), !S or !F (source 177 * route failed or fragmentation needed -- neither of these should 178 * ever occur and the associated gateway is busted if you see one). If 179 * almost all the probes result in some kind of unreachable, traceroute 180 * will give up and exit. 181 * 182 * Notes 183 * ----- 184 * This program must be run by root or be setuid. (I suggest that 185 * you *don't* make it setuid -- casual use could result in a lot 186 * of unnecessary traffic on our poor, congested nets.) 187 * 188 * This program requires a kernel mod that does not appear in any 189 * system available from Berkeley: A raw ip socket using proto 190 * IPPROTO_RAW must interpret the data sent as an ip datagram (as 191 * opposed to data to be wrapped in a ip datagram). See the README 192 * file that came with the source to this program for a description 193 * of the mods I made to /sys/netinet/raw_ip.c. Your mileage may 194 * vary. But, again, ANY 4.x (x < 4) BSD KERNEL WILL HAVE TO BE 195 * MODIFIED TO RUN THIS PROGRAM. 196 * 197 * The udp port usage may appear bizarre (well, ok, it is bizarre). 198 * The problem is that an icmp message only contains 8 bytes of 199 * data from the original datagram. 8 bytes is the size of a udp 200 * header so, if we want to associate replies with the original 201 * datagram, the necessary information must be encoded into the 202 * udp header (the ip id could be used but there's no way to 203 * interlock with the kernel's assignment of ip id's and, anyway, 204 * it would have taken a lot more kernel hacking to allow this 205 * code to set the ip id). So, to allow two or more users to 206 * use traceroute simultaneously, we use this task's pid as the 207 * source port (the high bit is set to move the port number out 208 * of the "likely" range). To keep track of which probe is being 209 * replied to (so times and/or hop counts don't get confused by a 210 * reply that was delayed in transit), we increment the destination 211 * port number before each probe. 212 * 213 * Don't use this as a coding example. I was trying to find a 214 * routing problem and this code sort-of popped out after 48 hours 215 * without sleep. I was amazed it ever compiled, much less ran. 216 * 217 * I stole the idea for this program from Steve Deering. Since 218 * the first release, I've learned that had I attended the right 219 * IETF working group meetings, I also could have stolen it from Guy 220 * Almes or Matt Mathis. I don't know (or care) who came up with 221 * the idea first. I envy the originators' perspicacity and I'm 222 * glad they didn't keep the idea a secret. 223 * 224 * Tim Seaver, Ken Adelman and C. Philip Wood provided bug fixes and/or 225 * enhancements to the original distribution. 226 * 227 * I've hacked up a round-trip-route version of this that works by 228 * sending a loose-source-routed udp datagram through the destination 229 * back to yourself. Unfortunately, SO many gateways botch source 230 * routing, the thing is almost worthless. Maybe one day... 231 * 232 * -- Van Jacobson (van@helios.ee.lbl.gov) 233 * Tue Dec 20 03:50:13 PST 1988 234 */ 235 236 237 #include <arpa/inet.h> 238 #include <endian.h> 239 #include <err.h> 240 #include <errno.h> 241 #include <limits.h> 242 #include <netdb.h> 243 #include <netinet/icmp6.h> 244 #include <netinet/in.h> 245 #include <netinet/ip.h> 246 #include <netinet/ip6.h> 247 #include <netinet/ip_icmp.h> 248 #include <netinet/udp.h> 249 #include <pwd.h> 250 #include <stdio.h> 251 #include <stdlib.h> 252 #include <string.h> 253 #include <sys/socket.h> 254 #include <sys/sysctl.h> 255 #include <sys/time.h> 256 #include <sys/uio.h> 257 #include <unistd.h> 258 259 #include "traceroute.h" 260 261 int32_t sec_perturb; 262 int32_t usec_perturb; 263 264 u_char packet[512]; 265 u_char *outpacket; /* last inbound (icmp) packet */ 266 267 int rcvsock; /* receive (icmp) socket file descriptor */ 268 int sndsock; /* send (udp) socket file descriptor */ 269 270 int rcvhlim; 271 struct in6_pktinfo *rcvpktinfo; 272 273 int datalen; /* How much data */ 274 275 char *hostname; 276 277 u_int16_t srcport; 278 279 void usage(int); 280 281 #define TRACEROUTE_USER "_traceroute" 282 283 int 284 main(int argc, char *argv[]) 285 { 286 int mib[4] = { CTL_NET, PF_INET, IPPROTO_IP, IPCTL_DEFTTL }; 287 char hbuf[NI_MAXHOST]; 288 289 struct tr_conf *conf; /* configuration defaults */ 290 struct sockaddr_in from4, to4; 291 struct sockaddr_in6 from6, to6; 292 struct sockaddr *from, *to; 293 struct addrinfo hints, *res; 294 struct hostent *hp; 295 struct ip *ip = NULL; 296 struct iovec rcviov[2]; 297 struct msghdr rcvmhdr; 298 static u_char *rcvcmsgbuf; 299 struct passwd *pw; 300 301 long l; 302 socklen_t len; 303 size_t size; 304 305 int ch; 306 int on = 1; 307 int seq = 0; 308 int error; 309 int curwaittime; 310 int headerlen; /* How long packet's header is */ 311 int i; 312 int last_tos = 0; 313 int packetlen; 314 int probe; 315 int rcvcmsglen; 316 int rcvsock4, rcvsock6; 317 int sndsock4, sndsock6; 318 u_int32_t tmprnd; 319 int v4sock_errno, v6sock_errno; 320 int v6flag = 0; 321 int xflag = 0; /* show ICMP extension header */ 322 323 char *dest; 324 const char *errstr; 325 u_int8_t ttl; 326 327 uid_t ouid, uid; 328 gid_t gid; 329 330 /* Cannot pledge due to special setsockopt()s below */ 331 if (unveil("/", "r") == -1) 332 err(1, "unveil"); 333 if (unveil(NULL, NULL) == -1) 334 err(1, "unveil"); 335 336 if ((conf = calloc(1, sizeof(*conf))) == NULL) 337 err(1,NULL); 338 339 conf->incflag = 1; 340 conf->first_ttl = 1; 341 conf->proto = IPPROTO_UDP; 342 conf->max_ttl = IPDEFTTL; 343 conf->nprobes = 3; 344 345 /* start udp dest port # for probe packets */ 346 conf->port = 32768+666; 347 348 memset(&rcvmhdr, 0, sizeof(rcvmhdr)); 349 memset(&rcviov, 0, sizeof(rcviov)); 350 351 rcvsock4 = rcvsock6 = sndsock4 = sndsock6 = -1; 352 v4sock_errno = v6sock_errno = 0; 353 354 conf->waittime = 5 * 1000; 355 356 if ((rcvsock6 = socket(AF_INET6, SOCK_RAW, IPPROTO_ICMPV6)) == -1) 357 v6sock_errno = errno; 358 else if ((sndsock6 = socket(AF_INET6, SOCK_DGRAM, 0)) == -1) 359 v6sock_errno = errno; 360 361 if ((rcvsock4 = socket(AF_INET, SOCK_RAW, IPPROTO_ICMP)) == -1) 362 v4sock_errno = errno; 363 else if ((sndsock4 = socket(AF_INET, SOCK_RAW, IPPROTO_RAW)) == -1) 364 v4sock_errno = errno; 365 366 /* revoke privs */ 367 ouid = getuid(); 368 if (ouid == 0 && (pw = getpwnam(TRACEROUTE_USER)) != NULL) { 369 uid = pw->pw_uid; 370 gid = pw->pw_gid; 371 } else { 372 uid = getuid(); 373 gid = getgid(); 374 } 375 if (ouid && (setgroups(1, &gid) || 376 setresgid(gid, gid, gid) || 377 setresuid(uid, uid, uid))) 378 err(1, "unable to revoke privs"); 379 380 if (strcmp("traceroute6", __progname) == 0) { 381 v6flag = 1; 382 if (v6sock_errno != 0) 383 errc(5, v6sock_errno, rcvsock6 < 0 ? "socket(ICMPv6)" : 384 "socket(SOCK_DGRAM)"); 385 rcvsock = rcvsock6; 386 sndsock = sndsock6; 387 if (rcvsock4 >= 0) 388 close(rcvsock4); 389 if (sndsock4 >= 0) 390 close(sndsock4); 391 } else { 392 if (v4sock_errno != 0) 393 errc(5, v4sock_errno, rcvsock4 < 0 ? "icmp socket" : 394 "raw socket"); 395 rcvsock = rcvsock4; 396 sndsock = sndsock4; 397 if (rcvsock6 >= 0) 398 close(rcvsock6); 399 if (sndsock6 >= 0) 400 close(sndsock6); 401 } 402 403 if (v6flag) { 404 mib[1] = PF_INET6; 405 mib[2] = IPPROTO_IPV6; 406 mib[3] = IPV6CTL_DEFHLIM; 407 /* specify to tell receiving interface */ 408 if (setsockopt(rcvsock, IPPROTO_IPV6, IPV6_RECVPKTINFO, &on, 409 sizeof(on)) == -1) 410 err(1, "setsockopt(IPV6_RECVPKTINFO)"); 411 412 /* specify to tell hoplimit field of received IP6 hdr */ 413 if (setsockopt(rcvsock, IPPROTO_IPV6, IPV6_RECVHOPLIMIT, &on, 414 sizeof(on)) == -1) 415 err(1, "setsockopt(IPV6_RECVHOPLIMIT)"); 416 } 417 418 size = sizeof(i); 419 if (sysctl(mib, sizeof(mib)/sizeof(mib[0]), &i, &size, NULL, 0) == -1) 420 err(1, "sysctl"); 421 conf->max_ttl = i; 422 423 while ((ch = getopt(argc, argv, v6flag ? "AcDdf:Ilm:np:q:Ss:t:w:vV:" : 424 "AcDdf:g:Ilm:nP:p:q:Ss:t:V:vw:x")) != -1) 425 switch (ch) { 426 case 'A': 427 conf->Aflag = 1; 428 break; 429 case 'c': 430 conf->incflag = 0; 431 break; 432 case 'd': 433 conf->dflag = 1; 434 break; 435 case 'D': 436 conf->dump = 1; 437 break; 438 case 'f': 439 conf->first_ttl = strtonum(optarg, 1, conf->max_ttl, 440 &errstr); 441 if (errstr) 442 errx(1, "min ttl must be 1 to %u.", 443 conf->max_ttl); 444 break; 445 case 'g': 446 if (conf->lsrr >= MAX_LSRR) 447 errx(1, "too many gateways; max %d", MAX_LSRR); 448 if (inet_aton(optarg, &conf->gateway[conf->lsrr]) == 449 0) { 450 hp = gethostbyname(optarg); 451 if (hp == 0) 452 errx(1, "unknown host %s", optarg); 453 memcpy(&conf->gateway[conf->lsrr], hp->h_addr, 454 hp->h_length); 455 } 456 if (++conf->lsrr == 1) 457 conf->lsrrlen = 4; 458 conf->lsrrlen += 4; 459 break; 460 case 'I': 461 if (conf->protoset) 462 errx(1, "protocol already set with -P"); 463 conf->protoset = 1; 464 conf->proto = IPPROTO_ICMP; 465 break; 466 case 'l': 467 conf->ttl_flag = 1; 468 break; 469 case 'm': 470 conf->max_ttl = strtonum(optarg, conf->first_ttl, 471 MAXTTL, &errstr); 472 if (errstr) 473 errx(1, "max ttl must be %u to %u.", 474 conf->first_ttl, MAXTTL); 475 break; 476 case 'n': 477 conf->nflag = 1; 478 break; 479 case 'p': 480 conf->port = strtonum(optarg, 1, 65535, &errstr); 481 if (errstr) 482 errx(1, "port must be >0, <65536."); 483 break; 484 case 'P': 485 if (conf->protoset) 486 errx(1, "protocol already set with -I"); 487 conf->protoset = 1; 488 conf->proto = strtonum(optarg, 1, IPPROTO_MAX - 1, 489 &errstr); 490 if (errstr) { 491 struct protoent *pent; 492 493 pent = getprotobyname(optarg); 494 if (pent) 495 conf->proto = pent->p_proto; 496 else 497 errx(1, "proto must be >=1, or a " 498 "name."); 499 } 500 break; 501 case 'q': 502 conf->nprobes = strtonum(optarg, 1, INT_MAX, &errstr); 503 if (errstr) 504 errx(1, "nprobes must be >0."); 505 break; 506 case 's': 507 /* 508 * set the ip source address of the outbound 509 * probe (e.g., on a multi-homed host). 510 */ 511 conf->source = optarg; 512 break; 513 case 'S': 514 conf->sump = 1; 515 break; 516 case 't': 517 if (!map_tos(optarg, &conf->tos)) { 518 if (strlen(optarg) > 1 && optarg[0] == '0' && 519 optarg[1] == 'x') { 520 char *ep; 521 errno = 0; 522 ep = NULL; 523 l = strtol(optarg, &ep, 16); 524 if (errno || !*optarg || *ep || 525 l < 0 || l > 255) 526 errx(1, "illegal tos value %s", 527 optarg); 528 conf->tos = (int)l; 529 } else { 530 conf->tos = strtonum(optarg, 0, 255, 531 &errstr); 532 if (errstr) 533 errx(1, "illegal tos value %s", 534 optarg); 535 } 536 } 537 conf->tflag = 1; 538 last_tos = conf->tos; 539 break; 540 case 'v': 541 conf->verbose = 1; 542 break; 543 case 'V': 544 conf->rtableid = (unsigned int)strtonum(optarg, 0, 545 RT_TABLEID_MAX, &errstr); 546 if (errstr) 547 errx(1, "rtable value is %s: %s", 548 errstr, optarg); 549 if (setsockopt(sndsock, SOL_SOCKET, SO_RTABLE, 550 &conf->rtableid, sizeof(conf->rtableid)) == -1) 551 err(1, "setsockopt SO_RTABLE"); 552 if (setsockopt(rcvsock, SOL_SOCKET, SO_RTABLE, 553 &conf->rtableid, sizeof(conf->rtableid)) == -1) 554 err(1, "setsockopt SO_RTABLE"); 555 break; 556 case 'w': 557 conf->waittime = strtonum(optarg, 2, INT_MAX, &errstr); 558 if (errstr) 559 errx(1, "wait must be >1 sec."); 560 conf->waittime *= 1000; 561 break; 562 case 'x': 563 xflag = 1; 564 break; 565 default: 566 usage(v6flag); 567 } 568 569 if (ouid == 0 && (setgroups(1, &gid) || 570 setresgid(gid, gid, gid) || 571 setresuid(uid, uid, uid))) 572 err(1, "unable to revoke privs"); 573 574 argc -= optind; 575 argv += optind; 576 577 if (argc < 1 || argc > 2) 578 usage(v6flag); 579 580 setvbuf(stdout, NULL, _IOLBF, 0); 581 582 conf->ident = (getpid() & 0xffff) | 0x8000; 583 tmprnd = arc4random(); 584 sec_perturb = (tmprnd & 0x80000000) ? -(tmprnd & 0x7ff) : 585 (tmprnd & 0x7ff); 586 usec_perturb = arc4random(); 587 588 memset(&to4, 0, sizeof(to4)); 589 memset(&to6, 0, sizeof(to6)); 590 591 dest = *argv; 592 593 memset(&hints, 0, sizeof(hints)); 594 hints.ai_family = v6flag ? PF_INET6 : PF_INET; 595 hints.ai_socktype = SOCK_RAW; 596 hints.ai_protocol = 0; 597 hints.ai_flags = AI_CANONNAME; 598 if ((error = getaddrinfo(dest, NULL, &hints, &res))) 599 errx(1, "%s", gai_strerror(error)); 600 601 switch (res->ai_family) { 602 case AF_INET: 603 to = (struct sockaddr *)&to4; 604 from = (struct sockaddr *)&from4; 605 break; 606 case AF_INET6: 607 to = (struct sockaddr *)&to6; 608 from = (struct sockaddr *)&from6; 609 break; 610 default: 611 errx(1, "unsupported AF: %d", res->ai_family); 612 break; 613 } 614 615 memcpy(to, res->ai_addr, res->ai_addrlen); 616 617 if (!hostname) { 618 hostname = res->ai_canonname ? strdup(res->ai_canonname) : dest; 619 if (!hostname) 620 errx(1, "malloc"); 621 } 622 623 if (res->ai_next) { 624 if (getnameinfo(res->ai_addr, res->ai_addrlen, hbuf, 625 sizeof(hbuf), NULL, 0, NI_NUMERICHOST) != 0) 626 strlcpy(hbuf, "?", sizeof(hbuf)); 627 warnx("Warning: %s has multiple " 628 "addresses; using %s", hostname, hbuf); 629 } 630 freeaddrinfo(res); 631 632 if (*++argv) { 633 datalen = strtonum(*argv, 0, INT_MAX, &errstr); 634 if (errstr) 635 errx(1, "datalen out of range"); 636 } 637 638 switch (to->sa_family) { 639 case AF_INET: 640 switch (conf->proto) { 641 case IPPROTO_UDP: 642 headerlen = (sizeof(struct ip) + conf->lsrrlen + 643 sizeof(struct udphdr) + sizeof(struct packetdata)); 644 break; 645 case IPPROTO_ICMP: 646 headerlen = (sizeof(struct ip) + conf->lsrrlen + 647 sizeof(struct icmp) + sizeof(struct packetdata)); 648 break; 649 default: 650 headerlen = (sizeof(struct ip) + conf->lsrrlen + 651 sizeof(struct packetdata)); 652 } 653 654 if (datalen < 0 || datalen > IP_MAXPACKET - headerlen) 655 errx(1, "packet size must be 0 to %d.", 656 IP_MAXPACKET - headerlen); 657 658 datalen += headerlen; 659 660 if ((outpacket = calloc(1, datalen)) == NULL) 661 err(1, "calloc"); 662 663 rcviov[0].iov_base = (caddr_t)packet; 664 rcviov[0].iov_len = sizeof(packet); 665 rcvmhdr.msg_name = (caddr_t)&from4; 666 rcvmhdr.msg_namelen = sizeof(from4); 667 rcvmhdr.msg_iov = rcviov; 668 rcvmhdr.msg_iovlen = 1; 669 rcvmhdr.msg_control = NULL; 670 rcvmhdr.msg_controllen = 0; 671 672 ip = (struct ip *)outpacket; 673 if (conf->lsrr != 0) { 674 u_char *p = (u_char *)(ip + 1); 675 676 *p++ = IPOPT_NOP; 677 *p++ = IPOPT_LSRR; 678 *p++ = conf->lsrrlen - 1; 679 *p++ = IPOPT_MINOFF; 680 conf->gateway[conf->lsrr] = to4.sin_addr; 681 for (i = 1; i <= conf->lsrr; i++) { 682 memcpy(p, &conf->gateway[i], 683 sizeof(struct in_addr)); 684 p += sizeof(struct in_addr); 685 } 686 ip->ip_dst = conf->gateway[0]; 687 } else 688 ip->ip_dst = to4.sin_addr; 689 ip->ip_off = htons(0); 690 ip->ip_hl = (sizeof(struct ip) + conf->lsrrlen) >> 2; 691 ip->ip_p = conf->proto; 692 ip->ip_v = IPVERSION; 693 ip->ip_tos = conf->tos; 694 695 if (setsockopt(sndsock, IPPROTO_IP, IP_HDRINCL, 696 &on, sizeof(on)) == -1) 697 err(6, "IP_HDRINCL"); 698 699 if (conf->source) { 700 memset(&from4, 0, sizeof(from4)); 701 from4.sin_family = AF_INET; 702 if (inet_aton(conf->source, &from4.sin_addr) == 0) 703 errx(1, "unknown host %s", conf->source); 704 ip->ip_src = from4.sin_addr; 705 if (ouid != 0 && 706 (ntohl(from4.sin_addr.s_addr) & 0xff000000U) == 707 0x7f000000U && (ntohl(to4.sin_addr.s_addr) & 708 0xff000000U) != 0x7f000000U) 709 errx(1, "source is on 127/8, destination is" 710 " not"); 711 if (ouid && bind(sndsock, (struct sockaddr *)&from4, 712 sizeof(from4)) == -1) 713 err(1, "bind"); 714 } 715 packetlen = datalen; 716 break; 717 case AF_INET6: 718 /* 719 * packetlen is the size of the complete IP packet sent and 720 * reported in the first line of output. 721 * For IPv4 this is equal to datalen since we are constructing 722 * a raw packet. 723 * For IPv6 we need to always add the size of the IP6 header 724 * and for UDP packets the size of the UDP header since they 725 * are prepended to the packet by the kernel 726 */ 727 packetlen = sizeof(struct ip6_hdr); 728 switch (conf->proto) { 729 case IPPROTO_UDP: 730 headerlen = sizeof(struct packetdata); 731 packetlen += sizeof(struct udphdr); 732 break; 733 case IPPROTO_ICMP: 734 headerlen = sizeof(struct icmp6_hdr) + 735 sizeof(struct packetdata); 736 break; 737 default: 738 errx(1, "Unsupported proto: %hhu", conf->proto); 739 break; 740 } 741 742 if (datalen < 0 || datalen > IP_MAXPACKET - headerlen) 743 errx(1, "packet size must be 0 to %d.", 744 IP_MAXPACKET - headerlen); 745 746 datalen += headerlen; 747 packetlen += datalen; 748 749 if ((outpacket = calloc(1, datalen)) == NULL) 750 err(1, "calloc"); 751 752 /* initialize msghdr for receiving packets */ 753 rcviov[0].iov_base = (caddr_t)packet; 754 rcviov[0].iov_len = sizeof(packet); 755 rcvmhdr.msg_name = (caddr_t)&from6; 756 rcvmhdr.msg_namelen = sizeof(from6); 757 rcvmhdr.msg_iov = rcviov; 758 rcvmhdr.msg_iovlen = 1; 759 rcvcmsglen = CMSG_SPACE(sizeof(struct in6_pktinfo)) + 760 CMSG_SPACE(sizeof(int)); 761 762 if ((rcvcmsgbuf = malloc(rcvcmsglen)) == NULL) 763 errx(1, "malloc"); 764 rcvmhdr.msg_control = (caddr_t) rcvcmsgbuf; 765 rcvmhdr.msg_controllen = rcvcmsglen; 766 767 /* 768 * Send UDP or ICMP 769 */ 770 if (conf->proto == IPPROTO_ICMP) { 771 close(sndsock); 772 sndsock = rcvsock; 773 } 774 775 /* 776 * Source selection 777 */ 778 memset(&from6, 0, sizeof(from6)); 779 if (conf->source) { 780 memset(&hints, 0, sizeof(hints)); 781 hints.ai_family = AF_INET6; 782 hints.ai_socktype = SOCK_DGRAM; /*dummy*/ 783 hints.ai_flags = AI_NUMERICHOST; 784 if ((error = getaddrinfo(conf->source, "0", &hints, 785 &res))) 786 errx(1, "%s: %s", conf->source, 787 gai_strerror(error)); 788 memcpy(&from6, res->ai_addr, res->ai_addrlen); 789 freeaddrinfo(res); 790 } else { 791 struct sockaddr_in6 nxt; 792 int dummy; 793 794 nxt = to6; 795 nxt.sin6_port = htons(DUMMY_PORT); 796 if ((dummy = socket(AF_INET6, SOCK_DGRAM, 0)) == -1) 797 err(1, "socket"); 798 if (conf->rtableid > 0 && 799 setsockopt(dummy, SOL_SOCKET, SO_RTABLE, 800 &conf->rtableid, sizeof(conf->rtableid)) == -1) 801 err(1, "setsockopt(SO_RTABLE)"); 802 if (connect(dummy, (struct sockaddr *)&nxt, 803 nxt.sin6_len) == -1) 804 err(1, "connect"); 805 len = sizeof(from6); 806 if (getsockname(dummy, (struct sockaddr *)&from6, 807 &len) == -1) 808 err(1, "getsockname"); 809 close(dummy); 810 } 811 812 from6.sin6_port = htons(0); 813 if (bind(sndsock, (struct sockaddr *)&from6, from6.sin6_len) == -1) 814 err(1, "bind sndsock"); 815 816 if (conf->tflag) { 817 if (setsockopt(sndsock, IPPROTO_IPV6, IPV6_TCLASS, 818 &conf->tos, sizeof(conf->tos)) == -1) 819 err(6, "IPV6_TCLASS"); 820 } 821 822 len = sizeof(from6); 823 if (getsockname(sndsock, (struct sockaddr *)&from6, &len) == -1) 824 err(1, "getsockname"); 825 srcport = ntohs(from6.sin6_port); 826 break; 827 default: 828 errx(1, "unsupported AF: %d", to->sa_family); 829 break; 830 } 831 832 if (conf->dflag) { 833 (void) setsockopt(rcvsock, SOL_SOCKET, SO_DEBUG, 834 &on, sizeof(on)); 835 (void) setsockopt(sndsock, SOL_SOCKET, SO_DEBUG, 836 &on, sizeof(on)); 837 } 838 839 if (setsockopt(sndsock, SOL_SOCKET, SO_SNDBUF, 840 &datalen, sizeof(datalen)) == -1) 841 err(6, "SO_SNDBUF"); 842 843 if (conf->nflag && !conf->Aflag) { 844 if (pledge("stdio inet", NULL) == -1) 845 err(1, "pledge"); 846 } else { 847 if (pledge("stdio inet dns", NULL) == -1) 848 err(1, "pledge"); 849 } 850 851 if (getnameinfo(to, to->sa_len, hbuf, 852 sizeof(hbuf), NULL, 0, NI_NUMERICHOST)) 853 strlcpy(hbuf, "(invalid)", sizeof(hbuf)); 854 fprintf(stderr, "%s to %s (%s)", __progname, hostname, hbuf); 855 if (conf->source) 856 fprintf(stderr, " from %s", conf->source); 857 fprintf(stderr, ", %u hops max, %d byte packets\n", conf->max_ttl, 858 packetlen); 859 (void) fflush(stderr); 860 861 if (conf->first_ttl > 1) 862 printf("Skipping %u intermediate hops\n", conf->first_ttl - 1); 863 864 for (ttl = conf->first_ttl; ttl && ttl <= conf->max_ttl; ++ttl) { 865 int got_there = 0, unreachable = 0, timeout = 0, loss; 866 in_addr_t lastaddr = 0; 867 struct in6_addr lastaddr6; 868 869 printf("%2u ", ttl); 870 memset(&lastaddr6, 0, sizeof(lastaddr6)); 871 for (probe = 0, loss = 0; probe < conf->nprobes; ++probe) { 872 int cc; 873 struct timeval t1, t2; 874 875 gettime(&t1); 876 send_probe(conf, ++seq, ttl, conf->incflag, to); 877 curwaittime = conf->waittime; 878 while ((cc = wait_for_reply(rcvsock, &rcvmhdr, 879 curwaittime))) { 880 gettime(&t2); 881 i = packet_ok(conf, to->sa_family, &rcvmhdr, 882 cc, seq, conf->incflag); 883 /* Skip wrong packet */ 884 if (i == 0) { 885 curwaittime = conf->waittime - 886 ((t2.tv_sec - t1.tv_sec) * 1000 + 887 (t2.tv_usec - t1.tv_usec) / 1000); 888 if (curwaittime < 0) 889 curwaittime = 0; 890 continue; 891 } 892 if (to->sa_family == AF_INET) { 893 ip = (struct ip *)packet; 894 if (from4.sin_addr.s_addr != lastaddr) { 895 print(conf, from, 896 cc - (ip->ip_hl << 2), 897 inet_ntop(AF_INET, 898 &ip->ip_dst, hbuf, 899 sizeof(hbuf))); 900 lastaddr = 901 from4.sin_addr.s_addr; 902 } 903 } else if (to->sa_family == AF_INET6) { 904 if (!IN6_ARE_ADDR_EQUAL( 905 &from6.sin6_addr, &lastaddr6)) { 906 print(conf, from, cc, 907 rcvpktinfo ? 908 inet_ntop( AF_INET6, 909 &rcvpktinfo->ipi6_addr, 910 hbuf, sizeof(hbuf)) : "?"); 911 lastaddr6 = from6.sin6_addr; 912 } 913 } else 914 errx(1, "unsupported AF: %d", 915 to->sa_family); 916 917 printf(" %g ms", deltaT(&t1, &t2)); 918 if (conf->ttl_flag) 919 printf(" (%u)", v6flag ? rcvhlim : 920 ip->ip_ttl); 921 if (to->sa_family == AF_INET) { 922 if (i == -2) { 923 if (ip->ip_ttl <= 1) 924 printf(" !"); 925 ++got_there; 926 break; 927 } 928 929 if (conf->tflag) 930 check_tos(ip, &last_tos); 931 } 932 933 /* time exceeded in transit */ 934 if (i == -1) 935 break; 936 icmp_code(to->sa_family, i - 1, &got_there, 937 &unreachable); 938 break; 939 } 940 if (cc == 0) { 941 printf(" *"); 942 timeout++; 943 loss++; 944 } else if (cc && probe == conf->nprobes - 1 && 945 (xflag || conf->verbose)) 946 print_exthdr(packet, cc); 947 (void) fflush(stdout); 948 } 949 if (conf->sump) 950 printf(" (%d%% loss)", (loss * 100) / conf->nprobes); 951 putchar('\n'); 952 if (got_there || 953 (unreachable && (unreachable + timeout) >= conf->nprobes)) 954 break; 955 } 956 exit(0); 957 } 958 959 void 960 usage(int v6flag) 961 { 962 if (v6flag) { 963 fprintf(stderr, "usage: %s " 964 "[-AcDdIlnSv] [-f first_hop] [-m max_hop] [-p port]\n" 965 "\t[-q nqueries] [-s sourceaddr] [-t toskeyword] [-V rtable] " 966 "[-w waittime]\n\thost [datalen]\n", __progname); 967 } else { 968 fprintf(stderr, 969 "usage: %s [-AcDdIlnSvx] [-f first_ttl] [-g gateway_addr] " 970 "[-m max_ttl]\n" 971 "\t[-P proto] [-p port] [-q nqueries] [-s sourceaddr]\n" 972 "\t[-t toskeyword] " 973 "[-V rtable] [-w waittime] host [datalen]\n", 974 __progname); 975 } 976 exit(1); 977 } 978