1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
|
/* $OpenBSD: rarpd.c,v 1.50 2010/08/26 17:07:46 jsg Exp $ */
/* $NetBSD: rarpd.c,v 1.25 1998/04/23 02:48:33 mrg Exp $ */
/*
* Copyright (c) 1990 The Regents of the University of California.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that: (1) source code distributions
* retain the above copyright notice and this paragraph in its entirety, (2)
* distributions including binary code include the above copyright notice and
* this paragraph in its entirety in the documentation or other materials
* provided with the distribution, and (3) all advertising materials mentioning
* features or use of this software display the following acknowledgement:
* ``This product includes software developed by the University of California,
* Lawrence Berkeley Laboratory and its contributors.'' Neither the name of
* the University nor the names of its contributors may be used to endorse
* or promote products derived from this software without specific prior
* written permission.
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
* WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
*/
/*
* rarpd - Reverse ARP Daemon
*
* usage: rarpd [-adflt] interface
*/
#include <stdio.h>
#include <stdlib.h>
#include <syslog.h>
#include <string.h>
#include <stdarg.h>
#include <sys/param.h>
#include <unistd.h>
#include <sys/time.h>
#include <net/bpf.h>
#include <sys/socket.h>
#include <sys/ioctl.h>
#include <net/if.h>
#include <net/if_dl.h>
#include <net/if_types.h>
#include <netinet/in.h>
#include <netinet/if_ether.h>
#include <sys/file.h>
#include <errno.h>
#include <netdb.h>
#include <arpa/inet.h>
#include <dirent.h>
#include <util.h>
#include <ifaddrs.h>
#include <paths.h>
#define FATAL 1 /* fatal error occurred */
#define NONFATAL 0 /* non fatal error occurred */
/*
* The structures for each interface.
*/
struct if_addr {
in_addr_t ia_ipaddr; /* IP address of this interface */
in_addr_t ia_netmask; /* subnet or net mask */
struct if_addr *ia_next;
};
struct if_info {
int ii_fd; /* BPF file descriptor */
char ii_name[IFNAMSIZ]; /* if name, e.g. "en0" */
u_char ii_eaddr[ETHER_ADDR_LEN]; /* Ethernet address of this iface */
struct if_addr *ii_addrs; /* Networks this interface is on */
struct if_info *ii_next;
};
/*
* The list of all interfaces that are being listened to. rarp_loop()
* "selects" on the descriptors in this list.
*/
struct if_info *iflist;
int rarp_open(char *);
void init_one(char *);
void init_all(void);
void rarp_loop(void);
void lookup_addrs(char *, struct if_info *);
void usage(void);
void rarp_process(struct if_info *, u_char *);
void rarp_reply(struct if_info *, struct if_addr *,
struct ether_header *, u_int32_t, struct hostent *);
void update_arptab(u_char *, u_int32_t);
void error(int, const char *,...);
void debug(const char *,...);
u_int32_t ipaddrtonetmask(u_int32_t);
int rarp_bootable(u_int32_t);
int aflag = 0; /* listen on "all" interfaces */
int dflag = 0; /* print debugging messages */
int fflag = 0; /* don't fork */
int lflag = 0; /* log all replies */
int tflag = 0; /* tftpboot check */
int
main(int argc, char *argv[])
{
extern char *__progname;
extern int optind, opterr;
int op, devnull, f;
char *ifname;
pid_t pid;
/* All error reporting is done through syslogs. */
openlog(__progname, LOG_PID | LOG_CONS, LOG_DAEMON);
opterr = 0;
while ((op = getopt(argc, argv, "adflt")) != -1) {
switch (op) {
case 'a':
++aflag;
break;
case 'd':
++dflag;
break;
case 'f':
++fflag;
break;
case 'l':
++lflag;
break;
case 't':
++tflag;
break;
default:
usage();
/* NOTREACHED */
}
}
argc -= optind;
argv += optind;
if ((aflag && argc > 0) || (!aflag && argc == 0))
usage();
if (aflag)
init_all();
else
while (argc > 0) {
init_one(argv[0]);
argc--;
argv++;
}
if ((!fflag) && (!dflag)) {
pid = fork();
if (pid > 0)
/* Parent exits, leaving child in background. */
exit(0);
else
if (pid == -1) {
error(FATAL, "cannot fork");
/* NOTREACHED */
}
/* write pid file */
pidfile(NULL);
/* Fade into the background */
f = open(_PATH_TTY, O_RDWR);
if (f >= 0) {
if (ioctl(f, TIOCNOTTY, 0) < 0) {
error(FATAL, "TIOCNOTTY: %s", strerror(errno));
/* NOTREACHED */
}
(void) close(f);
}
(void) chdir("/");
(void) setpgrp(0, getpid());
devnull = open(_PATH_DEVNULL, O_RDWR);
if (devnull >= 0) {
(void) dup2(devnull, STDIN_FILENO);
(void) dup2(devnull, STDOUT_FILENO);
(void) dup2(devnull, STDERR_FILENO);
if (devnull > 2)
(void) close(devnull);
}
}
rarp_loop();
exit(0);
}
/*
* Add 'ifname' to the interface list. Lookup its IP address and network
* mask and Ethernet address, and open a BPF file for it.
*/
void
init_one(char *ifname)
{
struct if_info *p;
int fd;
/* first check to see if this "if" was already opened? */
for (p = iflist; p; p = p->ii_next)
if (!strncmp(p->ii_name, ifname, IFNAMSIZ))
return;
fd = rarp_open(ifname);
if (fd < 0)
return;
p = (struct if_info *)malloc(sizeof(*p));
if (p == 0) {
error(FATAL, "malloc: %s", strerror(errno));
/* NOTREACHED */
}
p->ii_next = iflist;
iflist = p;
p->ii_fd = fd;
strncpy(p->ii_name, ifname, IFNAMSIZ);
p->ii_addrs = NULL;
lookup_addrs(ifname, p);
}
/*
* Initialize all "candidate" interfaces that are in the system
* configuration list. A "candidate" is up, not loopback and not
* point to point.
*/
void
init_all(void)
{
struct ifaddrs *ifap, *ifa;
struct sockaddr_dl *sdl;
if (getifaddrs(&ifap) != 0) {
error(FATAL, "getifaddrs: %s", strerror(errno));
/* NOTREACHED */
}
for (ifa = ifap; ifa; ifa = ifa->ifa_next) {
sdl = (struct sockaddr_dl *)ifa->ifa_addr;
if (sdl->sdl_family != AF_LINK || sdl->sdl_type != IFT_ETHER ||
sdl->sdl_alen != 6)
continue;
if ((ifa->ifa_flags &
(IFF_UP | IFF_LOOPBACK | IFF_POINTOPOINT)) != IFF_UP)
continue;
init_one(ifa->ifa_name);
}
freeifaddrs(ifap);
}
void
usage(void)
{
(void) fprintf(stderr, "usage: rarpd [-adflt] if0 [... ifN]\n");
exit(1);
}
static int
bpf_open(void)
{
int fd, n = 0;
char device[sizeof "/dev/bpf0000000000"];
/* Go through all the minors and find one that isn't in use. */
do {
(void) snprintf(device, sizeof device, "/dev/bpf%d", n++);
fd = open(device, O_RDWR);
} while (fd < 0 && errno == EBUSY);
if (fd < 0) {
error(FATAL, "%s: %s", device, strerror(errno));
/* NOTREACHED */
}
return fd;
}
static struct bpf_insn insns[] = {
BPF_STMT(BPF_LD | BPF_H | BPF_ABS, 12),
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, ETHERTYPE_REVARP, 0, 3),
BPF_STMT(BPF_LD | BPF_H | BPF_ABS, 20),
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, ARPOP_REVREQUEST, 0, 1),
BPF_STMT(BPF_RET | BPF_K, sizeof(struct ether_arp) +
sizeof(struct ether_header)),
BPF_STMT(BPF_RET | BPF_K, 0),
};
static struct bpf_program filter = {
sizeof insns / sizeof(insns[0]),
insns
};
/*
* Open a BPF file and attach it to the interface named 'device'.
* Set immediate mode, and set a filter that accepts only RARP requests.
*/
int
rarp_open(char *device)
{
int fd, immediate;
struct ifreq ifr;
u_int dlt;
fd = bpf_open();
/* Set immediate mode so packets are processed as they arrive. */
immediate = 1;
if (ioctl(fd, BIOCIMMEDIATE, &immediate) < 0) {
error(FATAL, "BIOCIMMEDIATE: %s", strerror(errno));
/* NOTREACHED */
}
(void) strncpy(ifr.ifr_name, device, sizeof ifr.ifr_name);
if (ioctl(fd, BIOCSETIF, (caddr_t)&ifr) < 0) {
if (aflag) { /* for -a skip not ethernet interfaces */
close(fd);
return -1;
}
error(FATAL, "BIOCSETIF: %s", strerror(errno));
/* NOTREACHED */
}
/*
* Check that the data link layer is an Ethernet; this code
* won't work with anything else.
*/
if (ioctl(fd, BIOCGDLT, (caddr_t) &dlt) < 0) {
error(FATAL, "BIOCGDLT: %s", strerror(errno));
/* NOTREACHED */
}
if (dlt != DLT_EN10MB) {
if (aflag) { /* for -a skip not ethernet interfaces */
close(fd);
return -1;
}
error(FATAL, "%s is not an ethernet", device);
/* NOTREACHED */
}
/* Set filter program. */
if (ioctl(fd, BIOCSETF, (caddr_t)&filter) < 0) {
error(FATAL, "BIOCSETF: %s", strerror(errno));
/* NOTREACHED */
}
return fd;
}
/*
* Perform various sanity checks on the RARP request packet. Return
* false on failure and log the reason.
*/
static int
rarp_check(u_char *p, int len)
{
struct ether_header *ep = (struct ether_header *) p;
struct ether_arp *ap = (struct ether_arp *) (p + sizeof(*ep));
(void) debug("got a packet");
if (len < sizeof(*ep) + sizeof(*ap)) {
error(NONFATAL, "truncated request");
return 0;
}
/* XXX This test might be better off broken out... */
if (ntohs (ep->ether_type) != ETHERTYPE_REVARP ||
ntohs (ap->arp_hrd) != ARPHRD_ETHER ||
ntohs (ap->arp_op) != ARPOP_REVREQUEST ||
ntohs (ap->arp_pro) != ETHERTYPE_IP ||
ap->arp_hln != 6 || ap->arp_pln != 4) {
error(NONFATAL, "request fails sanity check");
return 0;
}
if (memcmp((char *) &ep->ether_shost, (char *) &ap->arp_sha, 6) != 0) {
error(NONFATAL, "ether/arp sender address mismatch");
return 0;
}
if (memcmp((char *) &ap->arp_sha, (char *) &ap->arp_tha, 6) != 0) {
error(NONFATAL, "ether/arp target address mismatch");
return 0;
}
return 1;
}
/*
* Loop indefinitely listening for RARP requests on the
* interfaces in 'iflist'.
*/
void
rarp_loop(void)
{
int cc, fd, fdsn, bufsize, maxfd = 0;
fd_set *fdsp, *lfdsp;
u_char *buf, *bp, *ep;
struct if_info *ii;
if (iflist == 0) {
error(FATAL, "no interfaces");
/* NOTREACHED */
}
if (ioctl(iflist->ii_fd, BIOCGBLEN, (caddr_t)&bufsize) < 0) {
error(FATAL, "BIOCGBLEN: %s", strerror(errno));
/* NOTREACHED */
}
buf = (u_char *) malloc((unsigned) bufsize);
if (buf == 0) {
error(FATAL, "malloc: %s", strerror(errno));
/* NOTREACHED */
}
/*
* Find the highest numbered file descriptor for select().
* Initialize the set of descriptors to listen to.
*/
for (ii = iflist; ii; ii = ii->ii_next)
if (ii->ii_fd > maxfd)
maxfd = ii->ii_fd;
fdsn = howmany(maxfd+1, NFDBITS) * sizeof(fd_mask);
if ((fdsp = (fd_set *)malloc(fdsn)) == NULL)
error(FATAL, "malloc");
if ((lfdsp = (fd_set *)malloc(fdsn)) == NULL)
error(FATAL, "malloc");
memset(fdsp, 0, fdsn);
for (ii = iflist; ii; ii = ii->ii_next)
FD_SET(ii->ii_fd, fdsp);
while (1) {
memcpy(lfdsp, fdsp, fdsn);
if (select(maxfd + 1, lfdsp, (fd_set *) 0,
(fd_set *) 0, (struct timeval *) 0) < 0) {
error(FATAL, "select: %s", strerror(errno));
/* NOTREACHED */
}
for (ii = iflist; ii; ii = ii->ii_next) {
fd = ii->ii_fd;
if (!FD_ISSET(fd, lfdsp))
continue;
again:
cc = read(fd, (char *) buf, bufsize);
/* Don't choke when we get ptraced */
if (cc < 0 && errno == EINTR)
goto again;
/* Due to a SunOS bug, after 2^31 bytes, the file
* offset overflows and read fails with EINVAL. The
* lseek() to 0 will fix things. */
if (cc < 0) {
if (errno == EINVAL &&
(lseek(fd, (off_t)0, SEEK_CUR) + bufsize) < 0) {
(void) lseek(fd, (off_t)0, SEEK_SET);
goto again;
}
error(FATAL, "read: %s", strerror(errno));
/* NOTREACHED */
}
/* Loop through the packet(s) */
#define bhp ((struct bpf_hdr *)bp)
bp = buf;
ep = bp + cc;
while (bp < ep) {
int caplen, hdrlen;
caplen = bhp->bh_caplen;
hdrlen = bhp->bh_hdrlen;
if (rarp_check(bp + hdrlen, caplen))
rarp_process(ii, bp + hdrlen);
bp += BPF_WORDALIGN(hdrlen + caplen);
}
}
}
}
#ifndef TFTP_DIR
#define TFTP_DIR "/tftpboot"
#endif
/*
* True if this server can boot the host whose IP address is 'addr'.
* This check is made by looking in the tftp directory for the
* configuration file.
*/
int
rarp_bootable(u_int32_t addr)
{
struct dirent *dent;
char ipname[40];
static DIR *dd = 0;
DIR *d;
(void) snprintf(ipname, sizeof ipname, "%08X", addr);
/* If directory is already open, rewind it. Otherwise, open it. */
if ((d = dd))
rewinddir(d);
else {
if (chdir(TFTP_DIR) == -1) {
error(FATAL, "chdir: %s", strerror(errno));
/* NOTREACHED */
}
d = opendir(".");
if (d == 0) {
error(FATAL, "opendir: %s", strerror(errno));
/* NOTREACHED */
}
dd = d;
}
while ((dent = readdir(d)))
if (strncmp(dent->d_name, ipname, 8) == 0)
return 1;
return 0;
}
/*
* Given a list of IP addresses, 'alist', return the first address that
* is on network 'net'; 'netmask' is a mask indicating the network portion
* of the address.
*/
static u_int32_t
choose_ipaddr(u_int32_t **alist, u_int32_t net, u_int32_t netmask)
{
for (; *alist; ++alist) {
if ((**alist & netmask) == net)
return **alist;
}
return 0;
}
/*
* Answer the RARP request in 'pkt', on the interface 'ii'. 'pkt' has
* already been checked for validity. The reply is overlaid on the request.
*/
void
rarp_process(struct if_info *ii, u_char *pkt)
{
char ename[MAXHOSTNAMELEN];
u_int32_t target_ipaddr;
struct ether_header *ep;
struct ether_addr *ea;
struct hostent *hp;
struct in_addr in;
struct if_addr *ia;
ep = (struct ether_header *) pkt;
ea = (struct ether_addr *) &ep->ether_shost;
debug("%s", ether_ntoa(ea));
if (ether_ntohost(ename, ea) != 0) {
debug("ether_ntohost failed");
return;
}
if ((hp = gethostbyname(ename)) == 0) {
debug("gethostbyname (%s) failed", ename);
return;
}
/* Choose correct address from list. */
if (hp->h_addrtype != AF_INET) {
error(FATAL, "cannot handle non IP addresses");
/* NOTREACHED */
}
for (target_ipaddr = 0, ia = ii->ii_addrs; ia; ia = ia->ia_next) {
target_ipaddr = choose_ipaddr((u_int32_t **) hp->h_addr_list,
ia->ia_ipaddr & ia->ia_netmask, ia->ia_netmask);
if (target_ipaddr)
break;
}
if (target_ipaddr == 0) {
for (ia = ii->ii_addrs; ia; ia = ia->ia_next) {
in.s_addr = ia->ia_ipaddr & ia->ia_netmask;
error(NONFATAL, "cannot find %s on net %s",
ename, inet_ntoa(in));
}
return;
}
if (tflag == 0 || rarp_bootable(htonl(target_ipaddr)))
rarp_reply(ii, ia, ep, target_ipaddr, hp);
debug("reply sent");
}
/*
* Lookup the ethernet address of the interface attached to the BPF
* file descriptor 'fd'; return it in 'eaddr'.
*/
void
lookup_addrs(char *ifname, struct if_info *p)
{
struct ifaddrs *ifap, *ifa;
struct sockaddr_dl *sdl;
u_char *eaddr = p->ii_eaddr;
struct if_addr *ia, **iap = &p->ii_addrs;
struct in_addr in;
int found = 0;
if (getifaddrs(&ifap) != 0) {
error(FATAL, "getifaddrs: %s", strerror(errno));
/* NOTREACHED */
}
for (ifa = ifap; ifa; ifa = ifa->ifa_next) {
if (strcmp(ifa->ifa_name, ifname))
continue;
sdl = (struct sockaddr_dl *) ifa->ifa_addr;
if (sdl->sdl_family == AF_LINK &&
sdl->sdl_type == IFT_ETHER && sdl->sdl_alen == 6) {
memcpy((caddr_t)eaddr, (caddr_t)LLADDR(sdl), 6);
if (dflag)
fprintf(stderr, "%s: %x:%x:%x:%x:%x:%x\n",
ifa->ifa_name,
eaddr[0], eaddr[1], eaddr[2],
eaddr[3], eaddr[4], eaddr[5]);
found = 1;
} else if (sdl->sdl_family == AF_INET) {
ia = malloc (sizeof (struct if_addr));
if (ia == NULL)
error(FATAL, "lookup_addrs: malloc: %s",
strerror(errno));
ia->ia_next = NULL;
ia->ia_ipaddr =
((struct sockaddr_in *) ifa->ifa_addr)->
sin_addr.s_addr;
ia->ia_netmask =
((struct sockaddr_in *) ifa->ifa_netmask)->
sin_addr.s_addr;
/* If SIOCGIFNETMASK didn't work,
figure out a mask from the IP
address class. */
if (ia->ia_netmask == 0)
ia->ia_netmask =
ipaddrtonetmask(ia->ia_ipaddr);
if (dflag) {
in.s_addr = ia->ia_ipaddr;
fprintf(stderr, "\t%s\n",
inet_ntoa(in));
}
*iap = ia;
iap = &ia->ia_next;
}
}
freeifaddrs(ifap);
if (!found)
error(FATAL, "lookup_addrs: Never saw interface `%s'!", ifname);
}
int arptab_set(u_char *eaddr, u_int32_t host);
/*
* Poke the kernel arp tables with the ethernet/ip address combinataion
* given. When processing a reply, we must do this so that the booting
* host (i.e. the guy running rarpd), won't try to ARP for the hardware
* address of the guy being booted (he cannot answer the ARP).
*/
void
update_arptab(u_char *ep, u_int32_t ipaddr)
{
#ifdef SIOCSARP
struct sockaddr_in *sin;
struct arpreq request;
u_int32_t host;
u_char *eaddr;
int s;
request.arp_flags = 0;
sin = (struct sockaddr_in *)&request.arp_pa;
sin->sin_family = AF_INET;
sin->sin_addr.s_addr = ipaddr;
request.arp_ha.sa_family = AF_UNSPEC;
/* This is needed #if defined(COMPAT_43) && BYTE_ORDER != BIG_ENDIAN,
because AF_UNSPEC is zero and the kernel assumes that a zero
sa_family means that the real sa_family value is in sa_len. */
request.arp_ha.sa_len = 16; /* XXX */
memcpy((char *) request.arp_ha.sa_data, (char *) ep, 6);
s = socket(AF_INET, SOCK_DGRAM, 0);
if (s < 0) {
error(NONFATAL, "socket: %s", strerror(errno));
} else {
if (ioctl(s, SIOCSARP, (caddr_t)&request) < 0)
error(NONFATAL, "SIOCSARP: %s", strerror(errno));
(void) close(s);
}
#else
if (arptab_set(ep, ipaddr) > 0)
syslog(LOG_ERR, "couldn't update arp table");
#endif
}
/*
* Build a reverse ARP packet and sent it out on the interface.
* 'ep' points to a valid ARPOP_REVREQUEST. The ARPOP_REVREPLY is built
* on top of the request, then written to the network.
*
* RFC 903 defines the ether_arp fields as follows. The following comments
* are taken (more or less) straight from this document.
*
* ARPOP_REVREQUEST
*
* arp_sha is the hardware address of the sender of the packet.
* arp_spa is undefined.
* arp_tha is the 'target' hardware address.
* In the case where the sender wishes to determine his own
* protocol address, this, like arp_sha, will be the hardware
* address of the sender.
* arp_tpa is undefined.
*
* ARPOP_REVREPLY
*
* arp_sha is the hardware address of the responder (the sender of the
* reply packet).
* arp_spa is the protocol address of the responder (see the note below).
* arp_tha is the hardware address of the target, and should be the same as
* that which was given in the request.
* arp_tpa is the protocol address of the target, that is, the desired address.
*
* Note that the requirement that arp_spa be filled in with the responder's
* protocol is purely for convenience. For instance, if a system were to use
* both ARP and RARP, then the inclusion of the valid protocol-hardware
* address pair (arp_spa, arp_sha) may eliminate the need for a subsequent
* ARP request.
*/
void
rarp_reply(struct if_info *ii, struct if_addr *ia, struct ether_header *ep,
u_int32_t ipaddr, struct hostent *hp)
{
struct ether_arp *ap = (struct ether_arp *) (ep + 1);
int len, n;
update_arptab((u_char *)&ap->arp_sha, ipaddr);
/* Build the rarp reply by modifying the rarp request in place. */
ep->ether_type = htons(ETHERTYPE_REVARP);
ap->ea_hdr.ar_hrd = htons(ARPHRD_ETHER);
ap->ea_hdr.ar_pro = htons(ETHERTYPE_IP);
ap->arp_op = htons(ARPOP_REVREPLY);
memcpy((char *) &ep->ether_dhost, (char *) &ap->arp_sha, 6);
memcpy((char *) &ep->ether_shost, (char *) ii->ii_eaddr, 6);
memcpy((char *) &ap->arp_sha, (char *) ii->ii_eaddr, 6);
memcpy((char *) ap->arp_tpa, (char *) &ipaddr, 4);
/* Target hardware is unchanged. */
memcpy((char *) ap->arp_spa, (char *) &ia->ia_ipaddr, 4);
if (lflag) {
struct ether_addr ea;
memcpy(&ea.ether_addr_octet, &ap->arp_sha, 6);
syslog(LOG_INFO, "%s asked; %s replied", hp->h_name,
ether_ntoa(&ea));
}
len = sizeof(*ep) + sizeof(*ap);
n = write(ii->ii_fd, (char *) ep, len);
if (n != len)
error(NONFATAL, "write: only %d of %d bytes written", n, len);
}
/*
* Get the netmask of an IP address. This routine is used if
* SIOCGIFNETMASK doesn't work.
*/
u_int32_t
ipaddrtonetmask(u_int32_t addr)
{
if (IN_CLASSA(addr))
return IN_CLASSA_NET;
if (IN_CLASSB(addr))
return IN_CLASSB_NET;
if (IN_CLASSC(addr))
return IN_CLASSC_NET;
error(FATAL, "unknown IP address class: %08X", addr);
/* NOTREACHED */
}
void
error(int fatal, const char *fmt,...)
{
va_list ap;
if (dflag) {
if (fatal)
(void) fprintf(stderr, "rarpd: error: ");
else
(void) fprintf(stderr, "rarpd: warning: ");
va_start(ap, fmt);
(void) vfprintf(stderr, fmt, ap);
va_end(ap);
(void) fprintf(stderr, "\n");
}
va_start(ap, fmt);
vsyslog(LOG_ERR, fmt, ap);
va_end(ap);
if (fatal)
exit(1);
/* NOTREACHED */
}
void
debug(const char *fmt,...)
{
va_list ap;
if (dflag) {
va_start(ap, fmt);
(void) fprintf(stderr, "rarpd: ");
(void) vfprintf(stderr, fmt, ap);
va_end(ap);
(void) fprintf(stderr, "\n");
}
}
|