1 The Software Porting And Archive Centre for HP-UX, at
3 http://hpux.csc.liv.ac.uk/
5 (and with mirrors in various countries, listed on the Centre's home
6 page) has ported versions, in both source and binary form, for Ethereal,
7 as well as for the "libpcap", GLib, GTK+, "zlib", and CMU SNMP libraries
10 The changes they've made appear largely to be compile option changes; if
11 you've downloaded the source to the latest version of Ethereal (the
12 version on the Centre's site may not necessarily be the latest version),
13 it should be able to compile, perhaps with those changes.
15 They appear to have used HP-UX's "cc" compiler, with the options "-Ae
16 -O"; there's a comment "Add -Dhpux_9 if building under 9.X". It may
19 If you want to use Ethereal to capture packets, you will have to install
20 "libpcap"; the INSTALL file for "libpcap" has several comments about
21 HP-UX, which you should read if you're going to install and use
24 Another note, from a mail message to the "ethereal-users" list:
26 Date: Wed, 22 Dec 1999 09:05:47 -0600 (EST)
27 From: Gerald Combs <gerald@zing.org>
28 To: Lothar Seitter <lothar.seitter@arcormail.de>
29 cc: ethereal-users@zing.org
30 Subject: Re: [ethereal-users] permission problem with capturing
32 On Wed, 22 Dec 1999, Lothar Seitter wrote:
34 > running 'ethereal' under HP-UX 11 with root permission and
35 > /dev/lan0 set to 777, I always get the message:
36 > "There are no network interfaces that can be opened.
37 > Please to make sure you have sufficient permission to
40 > I start ethereal with 'etheral -i lan0' and lan0 is definitely
43 > What am I missing???
45 You may need to reference the card's DLPI device directly. We were having
46 trouble getting Ethereal to capture on an HP-UX 10.20 machine here. I
47 found an article on Deja News that says:
49 "To access a particular interface, you would say "tcpdump -i /dev/dlpiN"
50 where N is the PPA of the interface you wish to use. You get the PPA by
51 looking at the output of lanscan. On 10.20, it is the same value as the
52 NMID. On 11.X, it is the Card Instance number."
54 This didn't help in our case, but it might in yours. The full article is
55 at http://x34.deja.com/[ST_rn=ps]/getdoc.xp?AN=549366486 .
57 Another article by the same author mentions that experimental versions of
58 libpcap and tcpdump are available at
59 ftp://ftp.cup.hp.com/dist/networking/tools/ . The article itself is at
60 http://x34.deja.com/[ST_rn=ps]/getdoc.xp?AN=558665378 .
62 The first of those articles also says:
64 BTW, before you have to make a follow-up post, you will find that
65 unless you have the latest lan common/DLPI/driver patches installed,
66 you will _not_ see the system's own outbound traffic.
68 It appears that a consequence of the fact that HP-UX's DLPI doesn't work
69 like Solaris's, in that, on Solaris, to get at the device "hme0", say,
70 "libpcap" has to open "/dev/hme" and then tell it to use the 0th
71 interface, whilst on HP-UX you have to go through "/dev/dlpi", you won't
72 get a list of interfaces in the dialog box for "Capture:Start" - you'll
73 have to do through the aforementioned song and dance to find the PPA of
74 the interface you want to use, and supply the "dlpiN" name by hand (I
75 think you can omit the "/dev/" in both tcpdump and Ethereal).
77 Here is a patch to "pcap-dlpi.c" in libpcap that, at least on HP-UX
78 11.X, allows the name of the network interface, rather than the "dlpiN"
79 name, to be specified to tcpdump and Ethereal. It has not been tried on
80 HP-UX 10.20; it fixes one bug that could have caused the code in vanilla
81 libpcap not to correctly find the PPA for an interface on HP-UX 10.20,
82 but HP-UX 10.20's DLPI doesn't supply, in the data returned by a
83 DL_HP_PPA_REQ request, network interface names, so, on systems without
84 the network interface names in that data, the code continues to check
85 the major device number.
87 On HP-UX 11.00, the patch allows a network interface to be specified by
88 name as an argument to tcpdump, rather than requiring that you specify a
89 "dlpiN" name (and it should work equally well with Ethereal).
91 If you try this code on HP-UX 10.20, and it doesn't let you specify the
92 interface by name, please send mail to ethereal-dev@zing.org, so that we
93 know that it didn't work - we'll probably send you debugging patches in
94 the hopes of being able to make it work on 10.20 as well.
96 Here's the patch (to vanilla libpcap 0.4; it patches "configure",
97 "aclocal.m4", and "configure.in", to make the configure script check
98 whether your version of HP-UX supplies the interface names in the reply
99 to a DL_HP_PPA_REQ request, and "pcap-dlpi.c"):
101 *** ../libpcap-0.4/configure Sat Jul 25 12:41:51 1998
102 --- configure Fri Jan 14 00:40:04 2000
108 echo $ac_n "checking if unaligned accesses fail""... $ac_c" 1>&6
109 ! echo "configure:1707: checking if unaligned accesses fail" >&5
110 if eval "test \"`echo '$''{'ac_cv_lbl_unaligned_fail'+set}'`\" = set"; then
111 echo $ac_n "(cached) $ac_c" 1>&6
117 + echo $ac_n "checking if dl_hp_ppa_info_t struct has dl_module_id_1 member""... $ac_c" 1>&6
118 + echo "configure:1707: checking if dl_hp_ppa_info_t struct has dl_module_id_1 member" >&5
119 + if eval "test \"`echo '$''{'ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1'+set}'`\" = set"; then
120 + echo $ac_n "(cached) $ac_c" 1>&6
122 + cat > conftest.$ac_ext <<EOF
123 + #line 1712 "configure"
124 + #include "confdefs.h"
126 + # include <sys/types.h>
127 + # include <sys/dlpi.h>
128 + # include <sys/dlpi_ext.h>
130 + u_int i = sizeof(((dl_hp_ppa_info_t *)0)->dl_module_id_1)
133 + if { (eval echo configure:1722: \"$ac_compile\") 1>&5; (eval $ac_compile) 2>&5; }; then
135 + ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1=yes
137 + echo "configure: failed program was:" >&5
138 + cat conftest.$ac_ext >&5
140 + ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1=no
145 + echo "$ac_t""$ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1" 1>&6
146 + if test $ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1 = yes ; then
147 + cat >> confdefs.h <<\EOF
148 + #define HAVE_HP_PPA_INFO_T_DL_MODULE_ID_1 1
153 echo $ac_n "checking if unaligned accesses fail""... $ac_c" 1>&6
154 ! echo "configure:1743: checking if unaligned accesses fail" >&5
155 if eval "test \"`echo '$''{'ac_cv_lbl_unaligned_fail'+set}'`\" = set"; then
156 echo $ac_n "(cached) $ac_c" 1>&6
160 # SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff"
161 # ./install, which can be erroneously created by make from ./install.sh.
162 echo $ac_n "checking for a BSD compatible install""... $ac_c" 1>&6
163 ! echo "configure:1803: checking for a BSD compatible install" >&5
164 if test -z "$INSTALL"; then
165 if eval "test \"`echo '$''{'ac_cv_path_install'+set}'`\" = set"; then
166 echo $ac_n "(cached) $ac_c" 1>&6
168 # SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff"
169 # ./install, which can be erroneously created by make from ./install.sh.
170 echo $ac_n "checking for a BSD compatible install""... $ac_c" 1>&6
171 ! echo "configure:1839: checking for a BSD compatible install" >&5
172 if test -z "$INSTALL"; then
173 if eval "test \"`echo '$''{'ac_cv_path_install'+set}'`\" = set"; then
174 echo $ac_n "(cached) $ac_c" 1>&6
177 *** ../libpcap-0.4/aclocal.m4 Fri Jun 12 03:45:15 1998
178 --- aclocal.m4 Tue Jan 4 21:02:13 2000
185 + dnl Checks to see if the dl_hp_ppa_info_t struct has the HP-UX 11.00
186 + dnl dl_module_id_1 member
190 + dnl AC_LBL_HP_PPA_INFO_T_DL_MODULE_ID_1
194 + dnl HAVE_HP_PPA_INFO_T_DL_MODULE_ID_1 (defined)
196 + dnl NOTE: any compile failure means we conclude that it doesn't have
197 + dnl that member, so if we don't have DLPI, don't have a <sys/dlpi_ext.h>
198 + dnl header, or have one that doesn't declare a dl_hp_ppa_info_t type,
199 + dnl we conclude it doesn't have that member (which is OK, as either we
200 + dnl won't be using code that would use that member, or we wouldn't
201 + dnl compile in any case).
203 + AC_DEFUN(AC_LBL_HP_PPA_INFO_T_DL_MODULE_ID_1,
204 + [AC_MSG_CHECKING(if dl_hp_ppa_info_t struct has dl_module_id_1 member)
205 + AC_CACHE_VAL(ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1,
207 + # include <sys/types.h>
208 + # include <sys/dlpi.h>
209 + # include <sys/dlpi_ext.h>],
210 + [u_int i = sizeof(((dl_hp_ppa_info_t *)0)->dl_module_id_1)],
211 + ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1=yes,
212 + ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1=no))
213 + AC_MSG_RESULT($ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1)
214 + if test $ac_cv_lbl_dl_hp_ppa_info_t_has_dl_module_id_1 = yes ; then
215 + AC_DEFINE(HAVE_HP_PPA_INFO_T_DL_MODULE_ID_1)
219 dnl Checks to see if -R is used
224 *** ../libpcap-0.4/configure.in Sun Jul 27 22:16:22 1997
225 --- configure.in Tue Jan 4 21:02:13 2000
230 AC_LBL_SOCKADDR_SA_LEN
232 + AC_LBL_HP_PPA_INFO_T_DL_MODULE_ID_1
234 AC_LBL_UNALIGNED_ACCESS
236 if test "${srcdir}" = "." ; then
238 *** ../libpcap-0.4/pcap-dlpi.c Wed Oct 15 21:59:34 1997
239 --- pcap-dlpi.c Tue Jan 4 21:02:13 2000
243 memset(p, 0, sizeof(*p));
246 ! ** Determine device and ppa
248 ! cp = strpbrk(device, "0123456789");
250 sprintf(ebuf, "%s missing unit number", device);
254 memset(p, 0, sizeof(*p));
256 + #ifdef HAVE_DEV_DLPI
258 ! ** Remove any "/dev/" on the front of the device.
260 ! cp = strrchr(device, '/');
268 ! * Split the name into a device type and a unit number.
270 ! cp = strpbrk(dname, "0123456789");
272 sprintf(ebuf, "%s missing unit number", device);
276 sprintf(ebuf, "%s bad unit number", device);
280 ! if (*device == '/')
281 ! strcpy(dname, device);
283 ! sprintf(dname, "%s/%s", PCAP_DEV_PREFIX, device);
284 ! #ifdef HAVE_DEV_DLPI
285 ! /* Map network device to /dev/dlpi unit */
287 if ((p->fd = open(cp, O_RDWR)) < 0) {
288 sprintf(ebuf, "%s: %s", cp, pcap_strerror(errno));
291 ! /* Map network interface to /dev/dlpi unit */
292 ppa = get_dlpi_ppa(p->fd, dname, ppa, ebuf);
296 /* Try device without unit number */
297 strcpy(dname2, dname);
298 cp = strchr(dname, *cp);
300 sprintf(ebuf, "%s bad unit number", device);
306 ! * Use "/dev/dlpi" as the device.
308 ! * XXX - HP's DLPI Programmer's Guide for HP-UX 11.00 says that
309 ! * the "dl_mjr_num" field is for the "major number of interface
310 ! * driver"; that's the major of "/dev/dlpi" on the system on
311 ! * which I tried this, but there may be DLPI devices that
312 ! * use a different driver, in which case we may need to
313 ! * search "/dev" for the appropriate device with that major
314 ! * device number, rather than hardwiring "/dev/dlpi".
316 ! * I'm assuming that the code that was used for HP-UX 10.x
317 ! * is valid, and therefore that, in 10.x, DLPIable devices have
318 ! * "/dev" entries corresponding to them and that their
319 ! * major device number is what appears in the "dl_hp_ppa_info_t"
320 ! * structure for that device. We can search by name in 11.x
321 ! * (and we may have to, as it's not clear that DLPIable devices
322 ! * will have those sorts of "/dev" entries), but we can't in
323 ! * 10.x, as 10.x doesn't have the "dl_module_id_1" and
324 ! * "dl_module_id_2" members in the "dl_hp_ppa_info_t" structure.
327 if ((p->fd = open(cp, O_RDWR)) < 0) {
328 sprintf(ebuf, "%s: %s", cp, pcap_strerror(errno));
333 ! * Get a table of all PPAs for that device, and search that
334 ! * table for the specified device type name and unit number.
336 ppa = get_dlpi_ppa(p->fd, dname, ppa, ebuf);
341 + ** Determine device and ppa
343 + cp = strpbrk(device, "0123456789");
345 + sprintf(ebuf, "%s missing unit number", device);
348 + ppa = strtol(cp, &eos, 10);
349 + if (*eos != '\0') {
350 + sprintf(ebuf, "%s bad unit number", device);
354 + if (*device == '/')
355 + strcpy(dname, device);
357 + sprintf(dname, "%s/%s", PCAP_DEV_PREFIX, device);
359 /* Try device without unit number */
360 strcpy(dname2, dname);
361 cp = strchr(dname, *cp);
367 ! sprintf(ebuf, "unknown mac type 0x%lu", infop->dl_mac_type);
375 ! sprintf(ebuf, "unknown mac type 0x%lu",
376 ! (unsigned long)infop->dl_mac_type);
383 #ifdef DL_HP_PPA_ACK_OBS
385 ! * Under HP-UX 10, we can ask for the ppa
391 #ifdef DL_HP_PPA_ACK_OBS
393 ! * Under HP-UX 10 and HP-UX 11, we can ask for the ppa
401 register dl_hp_ppa_ack_t *ap;
402 ! register dl_hp_ppa_info_t *ip;
404 register u_long majdev;
405 - dl_hp_ppa_req_t req;
407 bpf_u_int32 buf[MAXDLBUF];
409 ! if (stat(device, &statbuf) < 0) {
410 ! sprintf(ebuf, "stat: %s: %s", device, pcap_strerror(errno));
413 majdev = major(statbuf.st_rdev);
415 memset((char *)&req, 0, sizeof(req));
416 req.dl_primitive = DL_HP_PPA_REQ;
420 register dl_hp_ppa_ack_t *ap;
421 ! register dl_hp_ppa_info_t *ipstart, *ip;
423 + #ifndef HAVE_HP_PPA_INFO_T_DL_MODULE_ID_1
425 register u_long majdev;
428 + dl_hp_ppa_req_t req;
429 bpf_u_int32 buf[MAXDLBUF];
431 ! #ifndef HAVE_HP_PPA_INFO_T_DL_MODULE_ID_1
433 ! * HP-UX 10 doesn't put the interface name in the
434 ! * "dl_hp_ppa_info_t" structure, so we have to
435 ! * try to get the major device number for the device
436 ! * corresponding to the device and unit name provided to
437 ! * us, and search for the entry with that major device number.
439 ! sprintf(dname, "/dev/%s%d", device, unit);
440 ! if (stat(dname, &statbuf) < 0) {
441 ! sprintf(ebuf, "stat: %s: %s", dname, pcap_strerror(errno));
444 majdev = major(statbuf.st_rdev);
447 memset((char *)&req, 0, sizeof(req));
448 req.dl_primitive = DL_HP_PPA_REQ;
453 ap = (dl_hp_ppa_ack_t *)buf;
454 ! ip = (dl_hp_ppa_info_t *)((u_char *)ap + ap->dl_offset);
456 for(i = 0; i < ap->dl_count; i++) {
457 ! if (ip->dl_mjr_num == majdev && ip->dl_instance_num == unit)
460 ! ip = (dl_hp_ppa_info_t *)((u_char *)ip + ip->dl_next_offset);
462 if (i == ap->dl_count) {
463 ! sprintf(ebuf, "can't find PPA for %s", device);
466 if (ip->dl_hdw_state == HDW_DEAD) {
467 ! sprintf(ebuf, "%s: hardware state: DOWN\n", device);
470 return ((int)ip->dl_ppa);
474 ap = (dl_hp_ppa_ack_t *)buf;
475 ! ipstart = (dl_hp_ppa_info_t *)((u_char *)ap + ap->dl_offset);
478 for(i = 0; i < ap->dl_count; i++) {
479 ! #ifdef HAVE_HP_PPA_INFO_T_DL_MODULE_ID_1
480 ! if ((strcmp(ip->dl_module_id_1, device) == 0 ||
481 ! strcmp(ip->dl_module_id_2, device) == 0) &&
483 ! if (ip->dl_mjr_num == majdev &&
485 ! ip->dl_instance_num == unit)
488 ! ip = (dl_hp_ppa_info_t *)((u_char *)ipstart + ip->dl_next_offset);
490 if (i == ap->dl_count) {
491 ! sprintf(ebuf, "can't find PPA for %s%d", device, unit);
494 if (ip->dl_hdw_state == HDW_DEAD) {
495 ! sprintf(ebuf, "%s%d: hardware state: DOWN\n", device, unit);
498 return ((int)ip->dl_ppa);
504 ! char if_name[sizeof(ifnet.if_name)], tifname[32];
506 cp = strrchr(ifname, '/');
512 ! char if_name[sizeof(ifnet.if_name) + 1];
514 cp = strrchr(ifname, '/');
518 if (dlpi_kread(kd, (off_t)addr,
519 &ifnet, sizeof(ifnet), ebuf) < 0 ||
520 dlpi_kread(kd, (off_t)ifnet.if_name,
521 ! if_name, sizeof(if_name), ebuf) < 0) {
525 ! sprintf(tifname, "%.*s%d",
526 ! (int)sizeof(if_name), if_name, ifnet.if_unit);
527 ! if (strcmp(tifname, ifname) == 0)
528 return (ifnet.if_index);
532 if (dlpi_kread(kd, (off_t)addr,
533 &ifnet, sizeof(ifnet), ebuf) < 0 ||
534 dlpi_kread(kd, (off_t)ifnet.if_name,
535 ! if_name, sizeof(ifnet.if_name), ebuf) < 0) {
539 ! if_name[sizeof(ifnet.if_name)] = '\0';
540 ! if (strcmp(if_name, ifname) == 0 && ifnet.if_unit == unit)
541 return (ifnet.if_index);