xref: /freebsd/usr.sbin/traceroute6/traceroute6.c (revision 23f282aa31e9b6fceacd449020e936e98d6f2298)
1 /*
2  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
3  * All rights reserved.
4  *
5  * Redistribution and use in source and binary forms, with or without
6  * modification, are permitted provided that the following conditions
7  * are met:
8  * 1. Redistributions of source code must retain the above copyright
9  *    notice, this list of conditions and the following disclaimer.
10  * 2. Redistributions in binary form must reproduce the above copyright
11  *    notice, this list of conditions and the following disclaimer in the
12  *    documentation and/or other materials provided with the distribution.
13  * 3. Neither the name of the project nor the names of its contributors
14  *    may be used to endorse or promote products derived from this software
15  *    without specific prior written permission.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  *
29  * $FreeBSD$
30  */
31 
32 /*-
33  * Copyright (c) 1990, 1993
34  *	The Regents of the University of California.  All rights reserved.
35  *
36  * This code is derived from software contributed to Berkeley by
37  * Van Jacobson.
38  *
39  * Redistribution and use in source and binary forms, with or without
40  * modification, are permitted provided that the following conditions
41  * are met:
42  * 1. Redistributions of source code must retain the above copyright
43  *    notice, this list of conditions and the following disclaimer.
44  * 2. Redistributions in binary form must reproduce the above copyright
45  *    notice, this list of conditions and the following disclaimer in the
46  *    documentation and/or other materials provided with the distribution.
47  * 3. All advertising materials mentioning features or use of this software
48  *    must display the following acknowledgement:
49  *	This product includes software developed by the University of
50  *	California, Berkeley and its contributors.
51  * 4. Neither the name of the University nor the names of its contributors
52  *    may be used to endorse or promote products derived from this software
53  *    without specific prior written permission.
54  *
55  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
56  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
57  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
58  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
59  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
60  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
61  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
62  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
63  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
64  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
65  * SUCH DAMAGE.
66  */
67 
68 #ifndef lint
69 static char copyright[] =
70 "@(#) Copyright (c) 1990, 1993\n\
71 	The Regents of the University of California.  All rights reserved.\n";
72 #endif /* not lint */
73 
74 #ifndef lint
75 static char sccsid[] = "@(#)traceroute.c	8.1 (Berkeley) 6/6/93";
76 #endif /* not lint */
77 
78 /*
79  * traceroute host  - trace the route ip packets follow going to "host".
80  *
81  * Attempt to trace the route an ip packet would follow to some
82  * internet host.  We find out intermediate hops by launching probe
83  * packets with a small ttl (time to live) then listening for an
84  * icmp "time exceeded" reply from a gateway.  We start our probes
85  * with a ttl of one and increase by one until we get an icmp "port
86  * unreachable" (which means we got to "host") or hit a max (which
87  * defaults to 30 hops & can be changed with the -m flag).  Three
88  * probes (change with -q flag) are sent at each ttl setting and a
89  * line is printed showing the ttl, address of the gateway and
90  * round trip time of each probe.  If the probe answers come from
91  * different gateways, the address of each responding system will
92  * be printed.  If there is no response within a 5 sec. timeout
93  * interval (changed with the -w flag), a "*" is printed for that
94  * probe.
95  *
96  * Probe packets are UDP format.  We don't want the destination
97  * host to process them so the destination port is set to an
98  * unlikely value (if some clod on the destination is using that
99  * value, it can be changed with the -p flag).
100  *
101  * A sample use might be:
102  *
103  *     [yak 71]% traceroute nis.nsf.net.
104  *     traceroute to nis.nsf.net (35.1.1.48), 30 hops max, 56 byte packet
105  *      1  helios.ee.lbl.gov (128.3.112.1)  19 ms  19 ms  0 ms
106  *      2  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  39 ms  19 ms
107  *      3  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  39 ms  19 ms
108  *      4  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  39 ms  40 ms  39 ms
109  *      5  ccn-nerif22.Berkeley.EDU (128.32.168.22)  39 ms  39 ms  39 ms
110  *      6  128.32.197.4 (128.32.197.4)  40 ms  59 ms  59 ms
111  *      7  131.119.2.5 (131.119.2.5)  59 ms  59 ms  59 ms
112  *      8  129.140.70.13 (129.140.70.13)  99 ms  99 ms  80 ms
113  *      9  129.140.71.6 (129.140.71.6)  139 ms  239 ms  319 ms
114  *     10  129.140.81.7 (129.140.81.7)  220 ms  199 ms  199 ms
115  *     11  nic.merit.edu (35.1.1.48)  239 ms  239 ms  239 ms
116  *
117  * Note that lines 2 & 3 are the same.  This is due to a buggy
118  * kernel on the 2nd hop system -- lbl-csam.arpa -- that forwards
119  * packets with a zero ttl.
120  *
121  * A more interesting example is:
122  *
123  *     [yak 72]% traceroute allspice.lcs.mit.edu.
124  *     traceroute to allspice.lcs.mit.edu (18.26.0.115), 30 hops max
125  *      1  helios.ee.lbl.gov (128.3.112.1)  0 ms  0 ms  0 ms
126  *      2  lilac-dmc.Berkeley.EDU (128.32.216.1)  19 ms  19 ms  19 ms
127  *      3  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  19 ms  19 ms
128  *      4  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  19 ms  39 ms  39 ms
129  *      5  ccn-nerif22.Berkeley.EDU (128.32.168.22)  20 ms  39 ms  39 ms
130  *      6  128.32.197.4 (128.32.197.4)  59 ms  119 ms  39 ms
131  *      7  131.119.2.5 (131.119.2.5)  59 ms  59 ms  39 ms
132  *      8  129.140.70.13 (129.140.70.13)  80 ms  79 ms  99 ms
133  *      9  129.140.71.6 (129.140.71.6)  139 ms  139 ms  159 ms
134  *     10  129.140.81.7 (129.140.81.7)  199 ms  180 ms  300 ms
135  *     11  129.140.72.17 (129.140.72.17)  300 ms  239 ms  239 ms
136  *     12  * * *
137  *     13  128.121.54.72 (128.121.54.72)  259 ms  499 ms  279 ms
138  *     14  * * *
139  *     15  * * *
140  *     16  * * *
141  *     17  * * *
142  *     18  ALLSPICE.LCS.MIT.EDU (18.26.0.115)  339 ms  279 ms  279 ms
143  *
144  * (I start to see why I'm having so much trouble with mail to
145  * MIT.)  Note that the gateways 12, 14, 15, 16 & 17 hops away
146  * either don't send ICMP "time exceeded" messages or send them
147  * with a ttl too small to reach us.  14 - 17 are running the
148  * MIT C Gateway code that doesn't send "time exceeded"s.  God
149  * only knows what's going on with 12.
150  *
151  * The silent gateway 12 in the above may be the result of a bug in
152  * the 4.[23]BSD network code (and its derivatives):  4.x (x <= 3)
153  * sends an unreachable message using whatever ttl remains in the
154  * original datagram.  Since, for gateways, the remaining ttl is
155  * zero, the icmp "time exceeded" is guaranteed to not make it back
156  * to us.  The behavior of this bug is slightly more interesting
157  * when it appears on the destination system:
158  *
159  *      1  helios.ee.lbl.gov (128.3.112.1)  0 ms  0 ms  0 ms
160  *      2  lilac-dmc.Berkeley.EDU (128.32.216.1)  39 ms  19 ms  39 ms
161  *      3  lilac-dmc.Berkeley.EDU (128.32.216.1)  19 ms  39 ms  19 ms
162  *      4  ccngw-ner-cc.Berkeley.EDU (128.32.136.23)  39 ms  40 ms  19 ms
163  *      5  ccn-nerif35.Berkeley.EDU (128.32.168.35)  39 ms  39 ms  39 ms
164  *      6  csgw.Berkeley.EDU (128.32.133.254)  39 ms  59 ms  39 ms
165  *      7  * * *
166  *      8  * * *
167  *      9  * * *
168  *     10  * * *
169  *     11  * * *
170  *     12  * * *
171  *     13  rip.Berkeley.EDU (128.32.131.22)  59 ms !  39 ms !  39 ms !
172  *
173  * Notice that there are 12 "gateways" (13 is the final
174  * destination) and exactly the last half of them are "missing".
175  * What's really happening is that rip (a Sun-3 running Sun OS3.5)
176  * is using the ttl from our arriving datagram as the ttl in its
177  * icmp reply.  So, the reply will time out on the return path
178  * (with no notice sent to anyone since icmp's aren't sent for
179  * icmp's) until we probe with a ttl that's at least twice the path
180  * length.  I.e., rip is really only 7 hops away.  A reply that
181  * returns with a ttl of 1 is a clue this problem exists.
182  * Traceroute prints a "!" after the time if the ttl is <= 1.
183  * Since vendors ship a lot of obsolete (DEC's Ultrix, Sun 3.x) or
184  * non-standard (HPUX) software, expect to see this problem
185  * frequently and/or take care picking the target host of your
186  * probes.
187  *
188  * Other possible annotations after the time are !H, !N, !P (got a host,
189  * network or protocol unreachable, respectively), !S or !F (source
190  * route failed or fragmentation needed -- neither of these should
191  * ever occur and the associated gateway is busted if you see one).  If
192  * almost all the probes result in some kind of unreachable, traceroute
193  * will give up and exit.
194  *
195  * Notes
196  * -----
197  * This program must be run by root or be setuid.  (I suggest that
198  * you *don't* make it setuid -- casual use could result in a lot
199  * of unnecessary traffic on our poor, congested nets.)
200  *
201  * This program requires a kernel mod that does not appear in any
202  * system available from Berkeley:  A raw ip socket using proto
203  * IPPROTO_RAW must interpret the data sent as an ip datagram (as
204  * opposed to data to be wrapped in a ip datagram).  See the README
205  * file that came with the source to this program for a description
206  * of the mods I made to /sys/netinet/raw_ip.c.  Your mileage may
207  * vary.  But, again, ANY 4.x (x < 4) BSD KERNEL WILL HAVE TO BE
208  * MODIFIED TO RUN THIS PROGRAM.
209  *
210  * The udp port usage may appear bizarre (well, ok, it is bizarre).
211  * The problem is that an icmp message only contains 8 bytes of
212  * data from the original datagram.  8 bytes is the size of a udp
213  * header so, if we want to associate replies with the original
214  * datagram, the necessary information must be encoded into the
215  * udp header (the ip id could be used but there's no way to
216  * interlock with the kernel's assignment of ip id's and, anyway,
217  * it would have taken a lot more kernel hacking to allow this
218  * code to set the ip id).  So, to allow two or more users to
219  * use traceroute simultaneously, we use this task's pid as the
220  * source port (the high bit is set to move the port number out
221  * of the "likely" range).  To keep track of which probe is being
222  * replied to (so times and/or hop counts don't get confused by a
223  * reply that was delayed in transit), we increment the destination
224  * port number before each probe.
225  *
226  * Don't use this as a coding example.  I was trying to find a
227  * routing problem and this code sort-of popped out after 48 hours
228  * without sleep.  I was amazed it ever compiled, much less ran.
229  *
230  * I stole the idea for this program from Steve Deering.  Since
231  * the first release, I've learned that had I attended the right
232  * IETF working group meetings, I also could have stolen it from Guy
233  * Almes or Matt Mathis.  I don't know (or care) who came up with
234  * the idea first.  I envy the originators' perspicacity and I'm
235  * glad they didn't keep the idea a secret.
236  *
237  * Tim Seaver, Ken Adelman and C. Philip Wood provided bug fixes and/or
238  * enhancements to the original distribution.
239  *
240  * I've hacked up a round-trip-route version of this that works by
241  * sending a loose-source-routed udp datagram through the destination
242  * back to yourself.  Unfortunately, SO many gateways botch source
243  * routing, the thing is almost worthless.  Maybe one day...
244  *
245  *  -- Van Jacobson (van@helios.ee.lbl.gov)
246  *     Tue Dec 20 03:50:13 PST 1988
247  */
248 
249 #include <sys/param.h>
250 #include <sys/time.h>
251 #include <sys/socket.h>
252 #include <sys/uio.h>
253 #include <sys/file.h>
254 #include <sys/ioctl.h>
255 
256 #include <netinet/in.h>
257 
258 #include <arpa/inet.h>
259 
260 #include <netdb.h>
261 #include <stdio.h>
262 #include <err.h>
263 #include <errno.h>
264 #include <stdlib.h>
265 #include <string.h>
266 #include <unistd.h>
267 
268 #include <netinet/ip6.h>
269 #include <netinet/icmp6.h>
270 #include <netinet/udp.h>
271 
272 #ifdef IPSEC
273 #include <net/route.h>
274 #include <netinet6/ipsec.h>
275 #endif
276 
277 #define	freehostent(hp)
278 #define	DUMMY_PORT 10010
279 
280 #define	MAXPACKET	65535	/* max ip packet size */
281 #ifndef MAXHOSTNAMELEN
282 #define	MAXHOSTNAMELEN	64
283 #endif
284 
285 #ifndef FD_SET
286 #define	NFDBITS         (8*sizeof(fd_set))
287 #define	FD_SETSIZE      NFDBITS
288 #define	FD_SET(n, p)    ((p)->fds_bits[(n)/NFDBITS] |= (1 << ((n) % NFDBITS)))
289 #define	FD_CLR(n, p)    ((p)->fds_bits[(n)/NFDBITS] &= ~(1 << ((n) % NFDBITS)))
290 #define	FD_ISSET(n, p)  ((p)->fds_bits[(n)/NFDBITS] & (1 << ((n) % NFDBITS)))
291 #define	FD_ZERO(p)      bzero((char *)(p), sizeof(*(p)))
292 #endif
293 
294 #define	Fprintf (void)fprintf
295 #define	Sprintf (void)sprintf
296 #define	Printf (void)printf
297 
298 /*
299  * format of a (udp) probe packet.
300  */
301 struct opacket {
302 	u_char seq;		/* sequence number of this packet */
303 	u_char hops;		/* hop limit of the packet */
304 	struct timeval tv;	/* time packet left */
305 };
306 
307 u_char	packet[512];		/* last inbound (icmp) packet */
308 struct opacket	*outpacket;	/* last output (udp) packet */
309 
310 int	main __P((int, char *[]));
311 int	wait_for_reply __P((int, struct msghdr *));
312 int	setpolicy __P((int so, char *policy));
313 void	send_probe __P((int, int));
314 struct udphdr *get_udphdr __P((struct ip6_hdr *, u_char *));
315 int	get_hoplim __P((struct msghdr *));
316 double	deltaT __P((struct timeval *, struct timeval *));
317 char	*pr_type __P((int));
318 int	packet_ok __P((struct msghdr *, int, int));
319 void	print __P((struct msghdr *, int));
320 void	tvsub __P((struct timeval *, struct timeval *));
321 char	*inetname __P((struct in6_addr *));
322 void	usage __P((void));
323 
324 int rcvsock;			/* receive (icmp) socket file descriptor */
325 int sndsock;			/* send (udp) socket file descriptor */
326 struct timezone tz;		/* leftover */
327 
328 struct msghdr rcvmhdr;
329 struct iovec rcviov[2];
330 int rcvhlim;
331 struct in6_pktinfo *rcvpktinfo;
332 
333 struct sockaddr_in6 Src, Dst, Rcv;
334 struct sockaddr_in6 *src = &Src, *dst = &Dst, *rcv = &Rcv;
335 int datalen;			/* How much data */
336 char rtbuf[1024];	/*XXX*/
337 struct cmsghdr *cmsg;
338 
339 char *source = 0;
340 char *hostname;
341 
342 int nprobes = 3;
343 int max_hops = 30;
344 u_short ident;
345 u_short port = 32768+666;	/* start udp dest port # for probe packets */
346 int options;			/* socket options */
347 int verbose;
348 int waittime = 5;		/* time to wait for response (in seconds) */
349 int nflag;			/* print addresses numerically */
350 int lflag;			/* print both numerical address & hostname */
351 
352 char ntop_buf[INET6_ADDRSTRLEN]; /* for inet_ntop() */
353 
354 int
355 main(argc, argv)
356 	int argc;
357 	char *argv[];
358 {
359 	extern char *optarg;
360 	extern int optind;
361 	struct hostent *hp;
362 	struct addrinfo hints, *res;
363 	int error;
364 	int ch, i, on, probe, seq, hops;
365 	static u_char rcvcmsgbuf[CMSG_SPACE(sizeof(struct in6_pktinfo))
366 				+ CMSG_SPACE(sizeof(int))];
367 
368 	/*
369 	 * Receive ICMP
370 	 */
371 	if ((rcvsock = socket(AF_INET6, SOCK_RAW, IPPROTO_ICMPV6)) < 0) {
372 		perror("traceroute6: icmp socket");
373 		exit(5);
374 	}
375 	setuid(getuid());
376 
377 	on = 1;
378 	seq = 0;
379 
380 	while ((ch = getopt(argc, argv, "dlm:np:q:rs:w:vg:")) != EOF)
381 		switch(ch) {
382 		case 'd':
383 			options |= SO_DEBUG;
384 			break;
385 		case 'l':
386 			lflag++;
387 			break;
388 		case 'g':
389 			hp = getipnodebyname(optarg, AF_INET6, 0, &h_errno);
390 			if (hp == NULL) {
391 				Fprintf(stderr,
392 				    "traceroute6: unknown host %s\n", optarg);
393 				exit(1);
394 			}
395 			if (cmsg == NULL)
396 				cmsg = inet6_rthdr_init(rtbuf, IPV6_RTHDR_TYPE_0);
397 			inet6_rthdr_add(cmsg, (struct in6_addr *)hp->h_addr, IPV6_RTHDR_LOOSE);
398 			break;
399 		case 'm':
400 			max_hops = atoi(optarg);
401 			if (max_hops <= 1) {
402 				Fprintf(stderr,
403 				    "traceroute6: max hoplimit must be >1.\n");
404 				exit(1);
405 			}
406 			break;
407 		case 'n':
408 			nflag++;
409 			break;
410 		case 'p':
411 			port = atoi(optarg);
412 			if (port < 1) {
413 				Fprintf(stderr,
414 				    "traceroute6: port must be >0.\n");
415 				exit(1);
416 			}
417 			break;
418 		case 'q':
419 			nprobes = atoi(optarg);
420 			if (nprobes < 1) {
421 				Fprintf(stderr,
422 				    "traceroute6: nprobes must be >0.\n");
423 				exit(1);
424 			}
425 			break;
426 		case 'r':
427 			options |= SO_DONTROUTE;
428 			break;
429 		case 's':
430 			/*
431 			 * set the ip source address of the outbound
432 			 * probe (e.g., on a multi-homed host).
433 			 */
434 			source = optarg;
435 			break;
436 		case 'v':
437 			verbose++;
438 			break;
439 		case 'w':
440 			waittime = atoi(optarg);
441 			if (waittime <= 1) {
442 				Fprintf(stderr,
443 				    "traceroute6: wait must be >1 sec.\n");
444 				exit(1);
445 			}
446 			break;
447 		default:
448 			usage();
449 		}
450 	argc -= optind;
451 	argv += optind;
452 
453 	if (argc < 1)
454 		usage();
455 
456 	setvbuf(stdout, NULL, _IOLBF, BUFSIZ);
457 
458 	memset(&hints, 0, sizeof(hints));
459 	hints.ai_family = PF_INET6;
460 	hints.ai_socktype = SOCK_RAW;
461 	hints.ai_protocol = IPPROTO_ICMPV6;
462 	hints.ai_flags = AI_CANONNAME;
463 	error = getaddrinfo(*argv, NULL, &hints, &res);
464 	if (error) {
465 		(void)fprintf(stderr,
466 			      "traceroute6: %s\n", gai_strerror(error));
467 		if (error == EAI_SYSTEM)
468 			(void)fprintf(stderr,
469 				      "traceroute6: %s\n", strerror(errno));
470 		exit(1);
471 	}
472 	if (!res->ai_addr)
473 		errx(1, "getaddrinfo failed");
474 	memcpy(&Dst, res->ai_addr, res->ai_addrlen);
475 	hostname = res->ai_canonname ? strdup(res->ai_canonname) : *argv;
476 	freeaddrinfo(res);
477 
478 	if (*++argv)
479 		datalen = atoi(*argv);
480 	if (datalen < 0 || datalen >= MAXPACKET - sizeof(struct opacket)) {
481 		Fprintf(stderr,
482 		    "traceroute6: packet size must be 0 <= s < %ld.\n",
483 		    (long)(MAXPACKET - sizeof(struct opacket)));
484 		exit(1);
485 	}
486 	datalen += sizeof(struct opacket);
487 	outpacket = (struct opacket *)malloc((unsigned)datalen);
488 	if (! outpacket) {
489 		perror("traceroute6: malloc");
490 		exit(1);
491 	}
492 	(void) bzero((char *)outpacket, datalen);
493 
494 	/* initialize msghdr for receiving packets */
495 	rcviov[0].iov_base = (caddr_t)packet;
496 	rcviov[0].iov_len = sizeof(packet);
497 	rcvmhdr.msg_name = (caddr_t)rcv;
498 	rcvmhdr.msg_namelen = sizeof(*rcv);
499 	rcvmhdr.msg_iov = rcviov;
500 	rcvmhdr.msg_iovlen = 1;
501 	rcvmhdr.msg_control = (caddr_t) rcvcmsgbuf;
502 	rcvmhdr.msg_controllen = sizeof(rcvcmsgbuf);
503 
504 	/* specify to tell receiving interface */
505 	if (setsockopt(rcvsock, IPPROTO_IPV6, IPV6_PKTINFO, &on,
506 		       sizeof(on)) < 0)
507 		err(1, "setsockopt(IPV6_PKTINFO)");
508 
509 	/* specify to tell value of hoplimit field of received IP6 hdr */
510 	if (setsockopt(rcvsock, IPPROTO_IPV6, IPV6_HOPLIMIT, &on,
511 		       sizeof(on)) < 0)
512 		err(1, "setsockopt(IPV6_HOPLIMIT)");
513 
514 	if (options & SO_DEBUG)
515 		(void) setsockopt(rcvsock, SOL_SOCKET, SO_DEBUG,
516 				  (char *)&on, sizeof(on));
517 	if (options & SO_DONTROUTE)
518 		(void) setsockopt(rcvsock, SOL_SOCKET, SO_DONTROUTE,
519 				  (char *)&on, sizeof(on));
520 #ifdef IPSEC
521 #ifdef IPSEC_POLICY_IPSEC
522 	/*
523 	 * do not raise error even if setsockopt fails, kernel may have ipsec
524 	 * turned off.
525 	 */
526 	if (setpolicy(rcvsock, "in bypass") < 0)
527 		errx(1, ipsec_strerror());
528 	if (setpolicy(rcvsock, "out bypass") < 0)
529 		errx(1, ipsec_strerror());
530 #endif /*IPSEC_POLICY_IPSEC*/
531 #endif /*IPSEC*/
532 
533 	/*
534 	 * Send UDP
535 	 */
536 	if ((sndsock = socket(AF_INET6, SOCK_DGRAM, 0)) < 0) {
537 		perror("traceroute6: udp socket");
538 		exit(5);
539 	}
540 	if (setsockopt(sndsock, SOL_SOCKET, SO_SNDBUF, (char *)&datalen,
541 		       sizeof(datalen)) < 0) {
542 		perror("traceroute6: SO_SNDBUF");
543 		exit(6);
544 	}
545 	if (options & SO_DEBUG)
546 		(void) setsockopt(sndsock, SOL_SOCKET, SO_DEBUG,
547 				  (char *)&on, sizeof(on));
548 	if (options & SO_DONTROUTE)
549 		(void) setsockopt(sndsock, SOL_SOCKET, SO_DONTROUTE,
550 				  (char *)&on, sizeof(on));
551 	if (cmsg != NULL) {
552 		inet6_rthdr_lasthop(cmsg, IPV6_RTHDR_LOOSE);
553 		(void) setsockopt(sndsock, IPPROTO_IPV6, IPV6_PKTOPTIONS,
554 				  rtbuf, cmsg->cmsg_len);
555 	}
556 #ifdef IPSEC
557 #ifdef IPSEC_POLICY_IPSEC
558 	/*
559 	 * do not raise error even if setsockopt fails, kernel may have ipsec
560 	 * turned off.
561 	 */
562 	if (setpolicy(sndsock, "in bypass") < 0)
563 		errx(1, ipsec_strerror());
564 	if (setpolicy(sndsock, "out bypass") < 0)
565 		errx(1, ipsec_strerror());
566 #endif /*IPSEC_POLICY_IPSEC*/
567 #endif /*IPSEC*/
568 
569 	/*
570 	 * Source selection
571 	 */
572 	bzero((char *)src, sizeof(Src));
573 	if (source) {
574 		if (inet_pton(AF_INET6, source, &Src.sin6_addr) != 1) {
575 			Printf("traceroute6: unknown host %s\n", source);
576 			exit(1);
577 		}
578 	} else {
579 		struct sockaddr_in6 Nxt;
580 		int dummy, len;
581 
582 		len = sizeof(Src);
583 		Nxt = Dst;
584 		Nxt.sin6_port = htons(DUMMY_PORT);
585 		if (cmsg != NULL)
586 			bcopy(inet6_rthdr_getaddr(cmsg, 1), &Nxt.sin6_addr,
587 			      sizeof(Nxt.sin6_addr));
588 		if ((dummy = socket(AF_INET6, SOCK_DGRAM, 0)) < 0) {
589 			perror("socket") ;
590 		}
591 		if(-1 == connect(dummy, (struct sockaddr *)&Nxt, sizeof(Nxt)))
592 			perror("connect");
593 		if(-1 == getsockname(dummy, (struct sockaddr *)src, &len)) {
594 			perror("getsockname");
595 			printf("%d\n", errno);
596 		}
597 		close(dummy) ;
598 	}
599 	ident = (getpid() & 0xffff) | 0x8000;
600 	Src.sin6_family = AF_INET6;
601 	Src.sin6_port = htons(ident);
602 	if (bind(sndsock, (struct sockaddr *)src, sizeof(Src))  < 0){
603 		perror ("traceroute6: bind:");
604 		exit (1);
605 	}
606 
607 	/*
608 	 * Message to users
609 	 */
610 	Fprintf(stderr, "traceroute to %s (%s)", hostname,
611 		inet_ntop(AF_INET6, &Dst.sin6_addr,
612 			  ntop_buf, sizeof(ntop_buf)));
613 	if (source)
614 		Fprintf(stderr, " from %s", source);
615 	Fprintf(stderr, ", %d hops max, %d byte packets\n", max_hops, datalen);
616 	(void) fflush(stderr);
617 
618 	/*
619 	 * Main loop
620 	 */
621 	for (hops = 1; hops <= max_hops; ++hops) {
622 		struct in6_addr lastaddr;
623 		int got_there = 0;
624 		int unreachable = 0;
625 
626 		Printf("%2d ", hops);
627 		bzero(&lastaddr, sizeof(lastaddr));
628 		for (probe = 0; probe < nprobes; ++probe) {
629 			int cc;
630 			struct timeval t1, t2;
631 			struct timezone tz;
632 
633 			(void) gettimeofday(&t1, &tz);
634 			send_probe(++seq, hops);
635 			while ((cc = wait_for_reply(rcvsock, &rcvmhdr))) {
636 				(void) gettimeofday(&t2, &tz);
637 				if ((i = packet_ok(&rcvmhdr, cc, seq))) {
638 					if (! IN6_ARE_ADDR_EQUAL(&Rcv.sin6_addr,
639 							    &lastaddr)) {
640 						print(&rcvmhdr, cc);
641 						lastaddr = Rcv.sin6_addr;
642 					}
643 					Printf("  %g ms", deltaT(&t1, &t2));
644 					switch(i - 1) {
645 					case ICMP6_DST_UNREACH_NOROUTE:
646 						++unreachable;
647 						Printf(" !N");
648 						break;
649 					case ICMP6_DST_UNREACH_ADMIN:
650 						++unreachable;
651 						Printf(" !P");
652 						break;
653 					case ICMP6_DST_UNREACH_NOTNEIGHBOR:
654 						++unreachable;
655 						Printf(" !S");
656 						break;
657 					case ICMP6_DST_UNREACH_ADDR:
658 						++unreachable;
659 						Printf(" !A");
660 						break;
661 					case ICMP6_DST_UNREACH_NOPORT:
662 						if (rcvhlim >= 0 &&
663 						    rcvhlim <= 1)
664 							Printf(" !");
665 						++got_there;
666 						break;
667 					}
668 					break;
669 				}
670 			}
671 			if (cc == 0)
672 				Printf(" *");
673 			(void) fflush(stdout);
674 		}
675 		putchar('\n');
676 		if (got_there ||
677 		    (unreachable > 0 && unreachable >= ((nprobes + 1) / 2))) {
678 			exit(0);
679 		}
680 	}
681 
682 	exit(0);
683 }
684 
685 int
686 wait_for_reply(sock, mhdr)
687 	int sock;
688 	struct msghdr *mhdr;
689 {
690 	fd_set fds;
691 	struct timeval wait;
692 	int cc = 0;
693 
694 	FD_ZERO(&fds);
695 	FD_SET(sock, &fds);
696 	wait.tv_sec = waittime; wait.tv_usec = 0;
697 
698 	if (select(sock+1, &fds, (fd_set *)0, (fd_set *)0, &wait) > 0)
699 		cc = recvmsg(rcvsock, mhdr, 0);
700 
701 	return(cc);
702 }
703 
704 #ifdef IPSEC
705 #ifdef IPSEC_POLICY_IPSEC
706 int
707 setpolicy(so, policy)
708 	int so;
709 	char *policy;
710 {
711 	char *buf;
712 
713 	buf = ipsec_set_policy(policy, strlen(policy));
714 	if (buf == NULL) {
715 		warnx(ipsec_strerror());
716 		return -1;
717 	}
718 	(void)setsockopt(so, IPPROTO_IPV6, IPV6_IPSEC_POLICY,
719 		buf, ipsec_get_policylen(buf));
720 
721 	free(buf);
722 
723 	return 0;
724 }
725 #endif
726 #endif
727 
728 void
729 send_probe(seq, hops)
730 	int seq, hops;
731 {
732 	struct opacket *op = outpacket;
733 	int i;
734 
735 	if(setsockopt(sndsock, IPPROTO_IPV6, IPV6_UNICAST_HOPS,
736 		      (char *)&hops, sizeof(hops)) < 0) {
737 		perror("setsockopt IPV6_UNICAST_HOPS");
738 	}
739 
740 	Dst.sin6_port = htons(port + seq);
741 
742 	op->seq = seq;
743 	op->hops = hops;
744 	(void) gettimeofday(&op->tv, &tz);
745 
746 	i = sendto(sndsock, (char *)outpacket, datalen , 0,
747 		   (struct sockaddr *)dst, sizeof(Dst));
748 	if (i < 0 || i != datalen)  {
749 		if (i<0)
750 			perror("sendto");
751 		Printf("traceroute6: wrote %s %d chars, ret=%d\n", hostname,
752 		       datalen, i);
753 		(void) fflush(stdout);
754 	}
755 }
756 
757 int
758 get_hoplim(mhdr)
759 	struct msghdr *mhdr;
760 {
761 	struct cmsghdr *cm;
762 
763 	for (cm = (struct cmsghdr *)CMSG_FIRSTHDR(mhdr); cm;
764 	     cm = (struct cmsghdr *)CMSG_NXTHDR(mhdr, cm)) {
765 		if (cm->cmsg_level == IPPROTO_IPV6 &&
766 		    cm->cmsg_type == IPV6_HOPLIMIT &&
767 		    cm->cmsg_len == CMSG_LEN(sizeof(int)))
768 			return(*(int *)CMSG_DATA(cm));
769 	}
770 
771 	return(-1);
772 }
773 
774 double
775 deltaT(t1p, t2p)
776 	struct timeval *t1p, *t2p;
777 {
778 	register double dt;
779 
780 	dt = (double)(t2p->tv_sec - t1p->tv_sec) * 1000.0 +
781 	     (double)(t2p->tv_usec - t1p->tv_usec) / 1000.0;
782 	return (dt);
783 }
784 
785 
786 /*
787  * Convert an ICMP "type" field to a printable string.
788  */
789 char *
790 pr_type(t0)
791 	int t0;
792 {
793 	u_char t = t0 & 0xff;
794 	char *cp;
795 
796 	switch (t) {
797 	case ICMP6_DST_UNREACH:
798 		cp = "Destination Unreachable";
799 		break;
800 	case ICMP6_PACKET_TOO_BIG:
801 		cp = "Pakcet Too Big";
802 		break;
803 	case ICMP6_TIME_EXCEEDED:
804 		cp = "Time Exceeded";
805 		break;
806 	case ICMP6_PARAM_PROB:
807 		cp = "Parameter Problem";
808 		break;
809 	case ICMP6_ECHO_REQUEST:
810 		cp = "Echo Request";
811 		break;
812 	case ICMP6_ECHO_REPLY:
813 		cp = "Echo Reply";
814 		break;
815 	case ICMP6_MEMBERSHIP_QUERY:
816 		cp = "Group Membership Query";
817 		break;
818 	case ICMP6_MEMBERSHIP_REPORT:
819 		cp = "Group Membership Report";
820 		break;
821 	case ICMP6_MEMBERSHIP_REDUCTION:
822 		cp = "Group Membership Reduction";
823 		break;
824 	case ND_ROUTER_SOLICIT:
825 		cp = "Router Solicitation";
826 		break;
827 	case ND_ROUTER_ADVERT:
828 		cp = "Router Advertisement";
829 		break;
830 	case ND_NEIGHBOR_SOLICIT:
831 		cp = "Neighbor Solicitation";
832 		break;
833 	case ND_NEIGHBOR_ADVERT:
834 		cp = "Neighbor Advertisement";
835 		break;
836 	case ND_REDIRECT:
837 		cp = "Ridirect";
838 		break;
839 	default:
840 		cp = "Unknown";
841 		break;
842 	}
843 	return cp;
844 }
845 
846 
847 int
848 packet_ok(mhdr, cc, seq)
849 	struct msghdr *mhdr;
850 	int cc;
851 	int seq;
852 {
853 	register struct icmp6_hdr *icp;
854 	struct sockaddr_in6 *from = (struct sockaddr_in6 *)mhdr->msg_name;
855 	u_char type, code;
856 	char *buf = (char *)mhdr->msg_iov[0].iov_base;
857 	struct cmsghdr *cm;
858 	int *hlimp;
859 
860 	if (cc < sizeof(struct icmp6_hdr)) {
861 		if (verbose)
862 			Printf("data too short (%d bytes) from %s\n", cc,
863 			       inet_ntop(AF_INET6, &from->sin6_addr,
864 					 ntop_buf, sizeof(ntop_buf)));
865 		return(0);
866 	}
867 	icp = (struct icmp6_hdr *)buf;
868 	/* get optional information via advanced API */
869 	rcvpktinfo = NULL;
870 	hlimp = NULL;
871 	for (cm = (struct cmsghdr *)CMSG_FIRSTHDR(mhdr); cm;
872 	     cm = (struct cmsghdr *)CMSG_NXTHDR(mhdr, cm)) {
873 		if (cm->cmsg_level == IPPROTO_IPV6 &&
874 		    cm->cmsg_type == IPV6_PKTINFO &&
875 		    cm->cmsg_len ==
876 		    CMSG_LEN(sizeof(struct in6_pktinfo)))
877 			rcvpktinfo = (struct in6_pktinfo *)(CMSG_DATA(cm));
878 
879 		if (cm->cmsg_level == IPPROTO_IPV6 &&
880 		    cm->cmsg_type == IPV6_HOPLIMIT &&
881 		    cm->cmsg_len == CMSG_LEN(sizeof(int)))
882 			hlimp = (int *)CMSG_DATA(cm);
883 	}
884 	if (rcvpktinfo == NULL || hlimp == NULL) {
885 		warnx("failed to get received hop limit or packet info");
886 		return(0);
887 	}
888 	rcvhlim = *hlimp;
889 
890 	type = icp->icmp6_type;
891 	code = icp->icmp6_code;
892 	if ((type == ICMP6_TIME_EXCEEDED && code == ICMP6_TIME_EXCEED_TRANSIT)
893 	 || type == ICMP6_DST_UNREACH) {
894 		struct ip6_hdr *hip;
895 		struct udphdr *up;
896 
897 		hip = (struct ip6_hdr *)(icp + 1);
898 		if ((up = get_udphdr(hip, (u_char *)(buf + cc))) == NULL) {
899 			if (verbose)
900 				warnx("failed to get upper layer header");
901 			return(0);
902 		}
903 		if (up->uh_sport == htons(ident) &&
904 		    up->uh_dport == htons(port+seq))
905 			return (type == ICMP6_TIME_EXCEEDED ? -1 : code + 1);
906 	}
907 	if (verbose) {
908 		int i;
909 		u_long *lp = (u_long *)(icp + 1);
910 		char sbuf[INET6_ADDRSTRLEN+1], dbuf[INET6_ADDRSTRLEN];
911 
912 		Printf("\n%d bytes from %s to %s", cc,
913 			inet_ntop(AF_INET6, &from->sin6_addr,
914 				   sbuf, sizeof(sbuf)),
915 			inet_ntop(AF_INET6, &rcvpktinfo->ipi6_addr,
916 				   dbuf, sizeof(dbuf)));
917 		Printf(": icmp type %d (%s) code %d\n", type, pr_type(type),
918 		       icp->icmp6_code);
919 		for (i = 4; i < cc ; i += sizeof(long))
920 			Printf("%2d: %8.8x\n", i, (u_int32_t)ntohl(*lp++));
921 	}
922 	return(0);
923 }
924 
925 /*
926  * Increment pointer until find the UDP header.
927  */
928 struct udphdr *
929 get_udphdr(ip6, lim)
930 	struct ip6_hdr *ip6;
931 	u_char *lim;
932 {
933 	u_char *cp = (u_char *)ip6, nh;
934 	int hlen;
935 
936 	if (cp + sizeof(*ip6) >= lim)
937 		return(NULL);
938 
939 	nh = ip6->ip6_nxt;
940 	cp += sizeof(struct ip6_hdr);
941 
942 	while(lim - cp >= 8) {
943 		switch(nh) {
944 		 case IPPROTO_ESP:
945 		 case IPPROTO_TCP:
946 		 case IPPROTO_ICMPV6:
947 			 return(NULL);
948 		 case IPPROTO_UDP:
949 			 return((struct udphdr *)cp);
950 		 case IPPROTO_FRAGMENT:
951 			 hlen = sizeof(struct ip6_frag);
952 			 nh = ((struct ip6_frag *)cp)->ip6f_nxt;
953 			 break;
954 		 case IPPROTO_AH:
955 			 hlen = (((struct ip6_ext *)cp)->ip6e_len + 2) << 2;
956 			 nh = ((struct ip6_ext *)cp)->ip6e_nxt;
957 			 break;
958 		  default:
959 			  hlen = (((struct ip6_ext *)cp)->ip6e_len + 1) << 3;
960 			  nh = ((struct ip6_ext *)cp)->ip6e_nxt;
961 			  break;
962 		}
963 
964 		cp += hlen;
965 	}
966 
967 	return(NULL);
968 }
969 
970 void
971 print(mhdr, cc)
972 	struct msghdr *mhdr;
973 	int cc;
974 {
975 	struct sockaddr_in6 *from = (struct sockaddr_in6 *)mhdr->msg_name;
976 
977 	if (nflag) {
978 		Printf(" %s", inet_ntop(AF_INET6, &from->sin6_addr,
979 					 ntop_buf, sizeof(ntop_buf)));
980 	}
981 	else if (lflag) {
982 		Printf(" %s (%s)", inetname(&from->sin6_addr),
983 		       inet_ntop(AF_INET6, &from->sin6_addr,
984 				  ntop_buf, sizeof(ntop_buf)));
985 	}
986 	else {
987 		Printf(" %s", inetname(&from->sin6_addr));
988 	}
989 
990 	if (verbose) {
991 		Printf(" %d bytes of data to %s", cc,
992 		       inet_ntop(AF_INET6, &rcvpktinfo->ipi6_addr,
993 				 ntop_buf, sizeof(ntop_buf)));
994 	}
995 }
996 
997 /*
998  * Subtract 2 timeval structs:  out = out - in.
999  * Out is assumed to be >= in.
1000  */
1001 void
1002 tvsub(out, in)
1003 	register struct timeval *out, *in;
1004 {
1005 	if ((out->tv_usec -= in->tv_usec) < 0)   {
1006 		out->tv_sec--;
1007 		out->tv_usec += 1000000;
1008 	}
1009 	out->tv_sec -= in->tv_sec;
1010 }
1011 
1012 
1013 /*
1014  * Construct an Internet address representation.
1015  * If the nflag has been supplied, give
1016  * numeric value, otherwise try for symbolic name.
1017  */
1018 char *
1019 inetname(in)
1020 	struct in6_addr *in;
1021 {
1022 	register char *cp;
1023 	static char line[50];
1024 	struct hostent *hp;
1025 	static char domain[MAXHOSTNAMELEN + 1];
1026 	static int first = 1;
1027 	int herr;
1028 
1029 	if (first && !nflag) {
1030 		first = 0;
1031 		if (gethostname(domain, MAXHOSTNAMELEN) == 0 &&
1032 		    (cp = index(domain, '.')))
1033 			(void) strcpy(domain, cp + 1);
1034 		else
1035 			domain[0] = 0;
1036 	}
1037 	cp = 0;
1038 	if (!nflag) {
1039 		/* hp = (struct hostent *)addr2hostname(in, sizeof(*in), AF_INET6, &herr); */
1040 	  hp = (struct hostent *)getipnodebyaddr((const char *)in, sizeof(*in),
1041 						 AF_INET6, &herr);
1042 		if (hp) {
1043 			if ((cp = index(hp->h_name, '.')) &&
1044 			    !strcmp(cp + 1, domain))
1045 				*cp = 0;
1046 			cp = hp->h_name;
1047 #undef freehostent(hp)
1048 			freehostent(hp);
1049 #define	freehostent(hp)
1050 		}
1051 	}
1052 	if (cp)
1053 		(void) strcpy(line, cp);
1054 	else {
1055 		(void)inet_ntop(AF_INET6, in, line, sizeof(line));
1056 	}
1057 	return (line);
1058 }
1059 
1060 void
1061 usage()
1062 {
1063 	(void)fprintf(stderr,
1064 "usage: traceroute6 [-dlnrv] [-m max_hops] [-p port#] [-q nqueries]\n\t\
1065 [-s src_addr] [-g gateway] [-w wait] host [data size]\n");
1066 	exit(1);
1067 }
1068