metze/wireshark/wip.git
10 years agoRemove a no-longer-used variable.
Jeff Morriss [Thu, 27 Jun 2013 18:44:11 +0000 (18:44 -0000)]
Remove a no-longer-used variable.

svn path=/trunk/; revision=50194

10 years agocleanup int/uint fill proto functions.
Jakub Zawadzki [Thu, 27 Jun 2013 18:40:08 +0000 (18:40 -0000)]
cleanup int/uint fill proto functions.

- hfinfo_uint_value_format(), hfinfo_int_value_format() formats value instead of returning proper format
- remove hfinfo_int_format(), hfinfo_uint_format

svn path=/trunk/; revision=50193

10 years agoWindows changes for main_window_update ....
Luis Ontanon [Thu, 27 Jun 2013 17:45:07 +0000 (17:45 -0000)]
Windows changes for main_window_update ....

svn path=/trunk/; revision=50192

10 years agoforgot from the las commit...
Luis Ontanon [Thu, 27 Jun 2013 17:27:41 +0000 (17:27 -0000)]
forgot from the las commit...

svn path=/trunk/; revision=50191

10 years agoAdd functions hf_try_val_to_str[_const]
Jakub Zawadzki [Thu, 27 Jun 2013 17:26:23 +0000 (17:26 -0000)]
Add functions hf_try_val_to_str[_const]

svn path=/trunk/; revision=50190

10 years agocleanup, use packet_list_iter_nth_child() in another place.
Jakub Zawadzki [Thu, 27 Jun 2013 17:12:23 +0000 (17:12 -0000)]
cleanup, use packet_list_iter_nth_child() in another place.

svn path=/trunk/; revision=50189

10 years agoget main_window_update out of the way... pass an update_cb to the capture_sync stuff
Luis Ontanon [Thu, 27 Jun 2013 17:10:50 +0000 (17:10 -0000)]
get main_window_update out of the way... pass an update_cb to the capture_sync stuff
... as per the XXX comment removed from tshark.c this was a mess to keep the linker
happy... I couldn't!

I did this without even understanding whether calling main_window_update was realy
necessary in most cases. I guess nothing or more specific update cbs would be best.

svn path=/trunk/; revision=50188

10 years agoFor now, don't turn on -Wcast-align by default. There are at least some
Guy Harris [Thu, 27 Jun 2013 10:35:06 +0000 (10:35 -0000)]
For now, don't turn on -Wcast-align by default.  There are at least some
bugs it points out that probably mean the code won't work on machines
that require alignment (e.g., SPARC machines), but we'll turn it on once
we fix them.  (clang is fussier than GCC about this.)

svn path=/trunk/; revision=50187

10 years agoDon't assume the system include files are under /usr/include; trick the
Guy Harris [Thu, 27 Jun 2013 10:30:33 +0000 (10:30 -0000)]
Don't assume the system include files are under /usr/include; trick the
compiler into saying where they are, and use that.

Patch a bug in PortAudio pa_stable_v19_20111121 that clang found.

svn path=/trunk/; revision=50186

10 years agoFix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8857 :
Pascal Quantin [Thu, 27 Jun 2013 10:17:39 +0000 (10:17 -0000)]
Fix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8857 :
Do not assume that we must have at least 1 character present before \r\n

svn path=/trunk/; revision=50185

10 years ago*sigh*
Evan Huus [Thu, 27 Jun 2013 03:44:48 +0000 (03:44 -0000)]
*sigh*

From the comment above wmem_tree_insert32_array():
 * If you use ...32_array() calls you MUST make sure that every single node
 * you add to a specific tree always has a key of exactly the same number of
 * keylen words or things will most likely crash. Or at least that every single
 * item that sits behind the same top level node always have exactly the same
 * number of words.

So clearly generating thousands of keys with random lengths while testing is
going to cause problems. Generate a set of random lengths, then use those
lengths consistently (but still generating random keys of those lengths).

Should hopefully fix the intermittent build-bot failures.

(unfortunately this does not manifest nicely, and I cannot see an easy way to
assert it so that we catch other people trying to use different-length key
subtrees)

svn path=/trunk/; revision=50184

10 years agoMS: Pong from the dispatcher!
Luis Ontanon [Thu, 27 Jun 2013 03:41:48 +0000 (03:41 -0000)]
MS: Pong from the dispatcher!

svn path=/trunk/; revision=50183

10 years agoMisc little cleanups to wmem tree (comments, whitespace, assertions)
Evan Huus [Thu, 27 Jun 2013 03:01:53 +0000 (03:01 -0000)]
Misc little cleanups to wmem tree (comments, whitespace, assertions)

svn path=/trunk/; revision=50182

10 years agoRemove debugging lines from wmem_test, I think I've figured out the problem.
Evan Huus [Thu, 27 Jun 2013 03:00:33 +0000 (03:00 -0000)]
Remove debugging lines from wmem_test, I think I've figured out the problem.

svn path=/trunk/; revision=50181

10 years agoFrom Audric Schiltknecht, bug 8521 (https://bugs.wireshark.org/bugzilla/show_bug...
Michael Mann [Thu, 27 Jun 2013 02:34:57 +0000 (02:34 -0000)]
From Audric Schiltknecht, bug 8521 (https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8521)

According to ETSI TS 102 771 (GSE implementation guidelines), "mandatory extension headers" - when GSE's protocol type field is (strictly) less than 256 (0x100) - are of 'pre-defined length (and format) that must be known by all GSE receivers'.

svn path=/trunk/; revision=50180

10 years agoBugfix implementation of RFC 4379. Reported through bug 8835 (https://bugs.wireshark...
Michael Mann [Thu, 27 Jun 2013 01:15:24 +0000 (01:15 -0000)]
Bugfix implementation of RFC 4379.  Reported through bug 8835 (https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8835)

svn path=/trunk/; revision=50179

10 years agoDon't assume that compilers not named "clang" will, by default, report
Guy Harris [Thu, 27 Jun 2013 00:45:39 +0000 (00:45 -0000)]
Don't assume that compilers not named "clang" will, by default, report
an error, or not issue warnings, by default if you give them an unknown
-f flag.  Instead, test that flag with all compilers, and use -Werror to
force it to error out.

As with C/C++ flags, so with C++-only flags.

svn path=/trunk/; revision=50178

10 years agoDon't assume that compilers that need an extra option to report errors,
Guy Harris [Wed, 26 Jun 2013 22:43:51 +0000 (22:43 -0000)]
Don't assume that compilers that need an extra option to report errors,
rather than just warnings, for unknown warning options are named
"clang"; they might not be.

svn path=/trunk/; revision=50177

10 years ago#gtk #packet-list simplify, use gtk_tree_view_column_add_attribute() for text attribute.
Jakub Zawadzki [Wed, 26 Jun 2013 22:03:57 +0000 (22:03 -0000)]
#gtk #packet-list simplify, use gtk_tree_view_column_add_attribute() for text attribute.

svn path=/trunk/; revision=50176

10 years agoAnother clang issue to work around:
Guy Harris [Wed, 26 Jun 2013 21:29:38 +0000 (21:29 -0000)]
Another clang issue to work around:

libgcrypt expects gnu89-style semantics for "inline"; GCC defaults to
gnu89, but clang defaults to gnu99.

svn path=/trunk/; revision=50175

10 years agoSuppress -Wformat-nonliteral when building GLib, to avoid a clang bug
Guy Harris [Wed, 26 Jun 2013 19:38:47 +0000 (19:38 -0000)]
Suppress -Wformat-nonliteral when building GLib, to avoid a clang bug
where it issues bogus warnings.

svn path=/trunk/; revision=50174

10 years ago(Attempt to) fix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8813 :
Jeff Morriss [Wed, 26 Jun 2013 18:19:43 +0000 (18:19 -0000)]
(Attempt to) fix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8813 :

Enable INET6 checking on Solaris versions other than 5.8.

(Yes, theoretically this should be checking for Solaris 5.7 or later.  I'm
going to go out on a limb and assume that no one is actually still using 5.7
or, if they are, they aren't trying to compile a modern Wireshark on it...)

svn path=/trunk/; revision=50173

10 years agopacket-editor: Allow focus for byteview, it should fix problems with field hexeditor...
Jakub Zawadzki [Wed, 26 Jun 2013 17:15:25 +0000 (17:15 -0000)]
packet-editor: Allow focus for byteview, it should fix problems with field hexeditor not getting key-press-event.

Generally it should be also safe for builds without packet-editor, just in case ifdef it.

svn path=/trunk/; revision=50172

10 years agoFix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8853 :
Pascal Quantin [Wed, 26 Jun 2013 16:23:54 +0000 (16:23 -0000)]
Fix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8853 :
Add a missing subtree initialization

svn path=/trunk/; revision=50171

10 years agoDon't link wireshark against the Qt libraries and don't like wireshark-qt
Jeff Morriss [Wed, 26 Jun 2013 15:12:50 +0000 (15:12 -0000)]
Don't link wireshark against the Qt libraries and don't like wireshark-qt
against the Gtk+ libraries.

(I didn't notice this before because the resulting programs aren't, on my
system, actually linked against the libraries they don't use--presumably
because of the "-Wl,--as-needed" flag.)

svn path=/trunk/; revision=50170

10 years agoAdjust when name_len offset is used. Part of bug 8776 (https://bugs.wireshark.org...
Michael Mann [Wed, 26 Jun 2013 14:09:10 +0000 (14:09 -0000)]
Adjust when name_len offset is used.  Part of bug 8776 (https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8776)

From Dario Lombardo

svn path=/trunk/; revision=50169

10 years agoechld is not an optional package, it's an optional Wireshark feature: use
Jeff Morriss [Wed, 26 Jun 2013 13:52:52 +0000 (13:52 -0000)]
echld is not an optional package, it's an optional Wireshark feature: use
--enable-echld rather than --with-echld to control whether it is built.

svn path=/trunk/; revision=50168

10 years agoIf the user enables gtk3, (silently) disable gtk2 rather than forcing the user
Jeff Morriss [Wed, 26 Jun 2013 13:26:25 +0000 (13:26 -0000)]
If the user enables gtk3, (silently) disable gtk2 rather than forcing the user
to also specify --without-gtk2 or complaining that they have specified both.

svn path=/trunk/; revision=50167

10 years agoopenSAFETY: CRC Maintenance, new CRC, new expert interface implemented. Bug 8847...
Michael Mann [Wed, 26 Jun 2013 12:11:08 +0000 (12:11 -0000)]
openSAFETY: CRC Maintenance, new CRC, new expert interface implemented.  Bug 8847 (https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8847)

From Roland Knall

svn path=/trunk/; revision=50166

10 years agoMove tap-*-common.{h,c} into ui/ since that's the home for common UI functions.
Jeff Morriss [Wed, 26 Jun 2013 02:12:12 +0000 (02:12 -0000)]
Move tap-*-common.{h,c} into ui/ since that's the home for common UI functions.

svn path=/trunk/; revision=50165

10 years agoMove tempfile.{h,c} into wsutil.
Jeff Morriss [Wed, 26 Jun 2013 01:14:35 +0000 (01:14 -0000)]
Move tempfile.{h,c} into wsutil.

svn path=/trunk/; revision=50163

10 years agoNothing uses the HAVE_GTK #define, so get rid of it. (Keep the
Guy Harris [Wed, 26 Jun 2013 00:48:58 +0000 (00:48 -0000)]
Nothing uses the HAVE_GTK #define, so get rid of it.  (Keep the
Makefile.am conditional, however.)

svn path=/trunk/; revision=50162

10 years agoeditcap no longer has to link against nstime.o, it's in libwsutil now.
Jeff Morriss [Wed, 26 Jun 2013 00:45:38 +0000 (00:45 -0000)]
editcap no longer has to link against nstime.o, it's in libwsutil now.

svn path=/trunk/; revision=50161

10 years agoThis file is in the top-level directory. Don't try to include things from ../
Jeff Morriss [Wed, 26 Jun 2013 00:20:56 +0000 (00:20 -0000)]
This file is in the top-level directory. Don't try to include things from ../

svn path=/trunk/; revision=50160

10 years agoAs pointed out by Guy: timestats uses packet_info so it belongs in epan
Jeff Morriss [Wed, 26 Jun 2013 00:18:44 +0000 (00:18 -0000)]
As pointed out by Guy: timestats uses packet_info so it belongs in epan
not wsutil.

svn path=/trunk/; revision=50159

10 years agoPut wsutil/*.h into the wireshark-dev package (in a blind attempt to fix the debian...
Jeff Morriss [Tue, 25 Jun 2013 23:41:06 +0000 (23:41 -0000)]
Put wsutil/*.h into the wireshark-dev package (in a blind attempt to fix the debian packaging step).  Why do we allow/encourage a wirshark-dev package?  IMO Wireshark (C language) development should only be done from the source (a tarball or better yet SVN).

svn path=/trunk/; revision=50158

10 years agoThere's no need to compile timestats.c any more--it's now in libwsutil.
Jeff Morriss [Tue, 25 Jun 2013 23:34:07 +0000 (23:34 -0000)]
There's no need to compile timestats.c any more--it's now in libwsutil.

svn path=/trunk/; revision=50157

10 years agoArrays seems to be where the failure is, so remove excess verbiage around other
Evan Huus [Tue, 25 Jun 2013 22:32:53 +0000 (22:32 -0000)]
Arrays seems to be where the failure is, so remove excess verbiage around other
tests, and add more details to that section.

Now we wait for the buildbots to fail again...

svn path=/trunk/; revision=50156

10 years agoLink this plugin with libwsutil too (in order to pick up time symbols).
Jeff Morriss [Tue, 25 Jun 2013 22:15:11 +0000 (22:15 -0000)]
Link this plugin with libwsutil too (in order to pick up time symbols).

svn path=/trunk/; revision=50155

10 years agoMove a couple of time-related modules into wsutil.
Jeff Morriss [Tue, 25 Jun 2013 22:02:20 +0000 (22:02 -0000)]
Move a couple of time-related modules into wsutil.

A bunch of files didn't really need to include these header files so remove
the include line rather than changing it.

svn path=/trunk/; revision=50154

10 years agoFix build with packet editor enabled. (Fix a typo while we're at it.)
Guy Harris [Tue, 25 Jun 2013 21:45:56 +0000 (21:45 -0000)]
Fix build with packet editor enabled.  (Fix a typo while we're at it.)

svn path=/trunk/; revision=50153

10 years agoLast changes to autofoo, rename the files as they live now in their own dir...
Luis Ontanon [Tue, 25 Jun 2013 21:09:29 +0000 (21:09 -0000)]
Last changes to autofoo, rename the files as they live now in their own dir...

svn path=/trunk/; revision=50152

10 years agoFrom Reinhard Speyerer via https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8851:
Pascal Quantin [Tue, 25 Jun 2013 21:05:28 +0000 (21:05 -0000)]
From Reinhard Speyerer via https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8851:
ansi_637_tele dissector displays MSB as MBS for Call-Back Number

svn path=/trunk/; revision=50151

10 years agoLook for moc-qt4 before moc and uic-qt4 before uic, because (at least on my
Evan Huus [Tue, 25 Jun 2013 20:49:16 +0000 (20:49 -0000)]
Look for moc-qt4 before moc and uic-qt4 before uic, because (at least on my
machine, with both qt4 and qt5 development tools installed), moc and uic are the
qt5 versions, but the rest of the configuration process grabs qt4 headers and
link paths, leading to a lot of weird errors.

svn path=/trunk/; revision=50150

10 years agoOf course I forgot to test the normal/default case (where no --with arguments
Jeff Morriss [Tue, 25 Jun 2013 20:06:33 +0000 (20:06 -0000)]
Of course I forgot to test the normal/default case (where no --with arguments
are given)...

Actually default --with-gtk2 to "yes".

svn path=/trunk/; revision=50149

10 years agoAllow the Gtk+ and Qt versions to be built simultaneously (with autotools).
Jeff Morriss [Tue, 25 Jun 2013 19:51:53 +0000 (19:51 -0000)]
Allow the Gtk+ and Qt versions to be built simultaneously (with autotools).

./configure now accepts:
--with-gtk2 :  enabled by default; exclusive of --with-gtk3
--with-gtk3 : disabled by default; exclusive of --with-gtk2
--with-qt   : disabled by default, can be specified with gtk
--enable-wireshark : controls whether *any* GUI is built

If Qt is enabled then a new program is created: "wireshark-qt".  "wireshark"
remains the Gtk+ version.

svn path=/trunk/; revision=50147

10 years agoIndentation/Whitespace cleanup;
Bill Meier [Tue, 25 Jun 2013 19:38:15 +0000 (19:38 -0000)]
Indentation/Whitespace cleanup;
Add editor modelines.

svn path=/trunk/; revision=50146

10 years agoBugfix iec104 dissector, bug 8849 (https://bugs.wireshark.org/bugzilla/show_bug.cgi...
Michael Mann [Tue, 25 Jun 2013 19:32:39 +0000 (19:32 -0000)]
Bugfix iec104 dissector, bug 8849 (https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8849)

From Jiří Engelthaler

1) Wrong bits definitions for SIQ.BL, SIQ.SB, SIQ.NT, SIQ.IV, QDS.BL, QDS.SB, QDS.NT, QDS.IV
2) Invalid field abbrev for VTI Transient
3) Wrong bit size for SCO.QU, DCO.QU, RCO.QU
4) Changes from BASE_DEC to BASE_HEX
5) Several code style changes

svn path=/trunk/; revision=50145

10 years agoAllocate memory for col_text[_len] when really needed
Jakub Zawadzki [Tue, 25 Jun 2013 18:27:31 +0000 (18:27 -0000)]
Allocate memory for col_text[_len] when really needed

remove columnized bit, now the same can be done with col_text.

svn path=/trunk/; revision=50144

10 years agopacket_list_iter_children() simplify, just call packet_list_iter_nth_child(..., 0)
Jakub Zawadzki [Tue, 25 Jun 2013 18:07:33 +0000 (18:07 -0000)]
packet_list_iter_children() simplify, just call packet_list_iter_nth_child(..., 0)

svn path=/trunk/; revision=50143

10 years agoremove the Makefile.nmake target to avoid errors in dist...
Luis Ontanon [Tue, 25 Jun 2013 17:55:47 +0000 (17:55 -0000)]
remove the Makefile.nmake target to avoid errors in dist...

svn path=/trunk/; revision=50142

10 years agoThis should get cmake to ignore echld...
Luis Ontanon [Tue, 25 Jun 2013 17:43:52 +0000 (17:43 -0000)]
This should get cmake to ignore echld...

svn path=/trunk/; revision=50141

10 years agomakefile.nmake -> Makefile.nmake and minor whitespace (spaces -> tabs) changes.
Chris Maynard [Tue, 25 Jun 2013 15:32:49 +0000 (15:32 -0000)]
makefile.nmake -> Makefile.nmake and minor whitespace (spaces -> tabs) changes.

svn path=/trunk/; revision=50140

10 years agoFrom Dirk Jagdmann: Use a guint32 for IPv4 addresses. Fixes compilation
Gerald Combs [Tue, 25 Jun 2013 13:31:19 +0000 (13:31 -0000)]
From Dirk Jagdmann: Use a guint32 for IPv4 addresses. Fixes compilation
on OS X + Cmake.

svn path=/trunk/; revision=50139

10 years agoRevert --with-echld until I find out how to get it to work with Cmake...
Luis Ontanon [Tue, 25 Jun 2013 04:50:36 +0000 (04:50 -0000)]
Revert --with-echld until I find out how to get it to work with Cmake...
:

svn path=/trunk/; revision=50138

10 years ago--with-echld
Luis Ontanon [Tue, 25 Jun 2013 04:20:08 +0000 (04:20 -0000)]
--with-echld

svn path=/trunk/; revision=50137

10 years agoanother iteration, prepared for automake.
Luis Ontanon [Tue, 25 Jun 2013 03:33:02 +0000 (03:33 -0000)]
another iteration, prepared for automake.

svn path=/trunk/; revision=50136

10 years agomore stuff...
Luis Ontanon [Tue, 25 Jun 2013 00:31:01 +0000 (00:31 -0000)]
more stuff...

svn path=/trunk/; revision=50135

10 years agoBased on initial patch by Michal Labedzki
Evan Huus [Mon, 24 Jun 2013 23:26:56 +0000 (23:26 -0000)]
Based on initial patch by Michal Labedzki

Don't return a negative value from get_type_length, it can get us into bad
loops. Fixes https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8831

svn path=/trunk/; revision=50134

10 years agoFrom Fabio Tarabelloni via
Evan Huus [Mon, 24 Jun 2013 22:59:02 +0000 (22:59 -0000)]
From Fabio Tarabelloni via
https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8718

More zigbee dissection:
- illuminance measurement cluster
- temperature measurement cluster
- pressure measurement cluster
- relative humidity measurement cluster

svn path=/trunk/; revision=50133

10 years agoFix wmem_tree.h ifdef guards, hopefully fixing the check-abi buildbot.
Evan Huus [Mon, 24 Jun 2013 22:37:04 +0000 (22:37 -0000)]
Fix wmem_tree.h ifdef guards, hopefully fixing the check-abi buildbot.

svn path=/trunk/; revision=50132

10 years agoSo it comes to this...
Evan Huus [Mon, 24 Jun 2013 21:59:02 +0000 (21:59 -0000)]
So it comes to this...

Put in a whole bunch of stderr output in the wmem tree tests in the hopes that
the next time one of the buildbots randomly (and irreproducibly) fails on this
step we'll have at least a bit of a hint as to where it happened.

svn path=/trunk/; revision=50131

10 years agoAdd an option to dissect RRC-Container IE
Pascal Quantin [Mon, 24 Jun 2013 21:31:25 +0000 (21:31 -0000)]
Add an option to dissect RRC-Container IE

svn path=/trunk/; revision=50130

10 years agoA follow-up to r50127 to highlight the correct 4 bytes of the Ethernet FCS for the...
Chris Maynard [Mon, 24 Jun 2013 19:27:47 +0000 (19:27 -0000)]
A follow-up to r50127 to highlight the correct 4 bytes of the Ethernet FCS for the hf_eth_fcs_good and hf_eth_fcs_bad fields.

svn path=/trunk/; revision=50128

10 years agoThe Ethernet FCS is 32 bits, not 16 bits.
Guy Harris [Mon, 24 Jun 2013 19:00:55 +0000 (19:00 -0000)]
The Ethernet FCS is 32 bits, not 16 bits.

svn path=/trunk/; revision=50127

10 years agoDumpcap can write raw packets to stdout
Balint Reczey [Mon, 24 Jun 2013 14:49:19 +0000 (14:49 -0000)]
Dumpcap can write raw packets to stdout

svn path=/trunk/; revision=50126

10 years agoDisplay RSSI in dB instead of dBm. Fixes the last remaining issue reported in https...
Chris Maynard [Mon, 24 Jun 2013 14:30:00 +0000 (14:30 -0000)]
Display RSSI in dB instead of dBm.  Fixes the last remaining issue reported in https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=4507
#BACKPORT(1.10,1.8)

svn path=/trunk/; revision=50125

10 years agoMake Authentification String filterable
Alexis La Goutte [Sun, 23 Jun 2013 20:41:32 +0000 (20:41 -0000)]
Make Authentification String filterable

svn path=/trunk/; revision=50124

10 years agoFrom chaitanya via https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8837 Add suppo...
Alexis La Goutte [Sun, 23 Jun 2013 20:41:18 +0000 (20:41 -0000)]
From chaitanya via https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8837 Add support for VHT Transmit Power Envelope IE

Added support for VHT TPE IE and also tested it

Also fixed a small typo from "IEEE Stc" to "IEEE Std" for all 802.11ac references.

From me:
* Remove some trailing whitespace
* Fix bitmask PWR Info Unit
* Modify expert_info to display error in PWR Info Count
* Fix (possibility) loop
* Fix value_string (not need to have Reserved...)

svn path=/trunk/; revision=50123

10 years agoEnhance VRRP Dissector * Make checksum field filterable * Add also a checksum_bad...
Alexis La Goutte [Sun, 23 Jun 2013 20:41:09 +0000 (20:41 -0000)]
Enhance VRRP Dissector * Make checksum field filterable * Add also a checksum_bad filter and display a expert info when checksum is bad

svn path=/trunk/; revision=50122

10 years agoAlphabetically sort in file-mp4.c
Jörg Mayer [Sun, 23 Jun 2013 19:13:25 +0000 (19:13 -0000)]
Alphabetically sort in file-mp4.c

svn path=/trunk/; revision=50121

10 years agoSmall whitespace fixes
Jörg Mayer [Sun, 23 Jun 2013 19:12:50 +0000 (19:12 -0000)]
Small whitespace fixes

svn path=/trunk/; revision=50120

10 years ago[Automatic manuf, services and enterprise-numbers update for 2013-06-23]
Gerald Combs [Sun, 23 Jun 2013 14:03:31 +0000 (14:03 -0000)]
[Automatic manuf, services and enterprise-numbers update for 2013-06-23]

svn path=/trunk/; revision=50117

10 years agoOnly do a tree lookup if we have a tree to look up in.
Evan Huus [Sat, 22 Jun 2013 21:01:03 +0000 (21:01 -0000)]
Only do a tree lookup if we have a tree to look up in.
Fixes https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8833

svn path=/trunk/; revision=50116

10 years agoMake assertions in tree lookup more useful and fix formatting.
Evan Huus [Sat, 22 Jun 2013 21:00:36 +0000 (21:00 -0000)]
Make assertions in tree lookup more useful and fix formatting.

Should help debugging https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8833

svn path=/trunk/; revision=50115

10 years agoNot yet running but almost there with the dummy
Luis Ontanon [Sat, 22 Jun 2013 20:26:13 +0000 (20:26 -0000)]
Not yet running  but almost there with the dummy

svn path=/trunk/; revision=50114

10 years agoMore proper fix (per advice of bug 8515 (https://bugs.wireshark.org/bugzilla/show_bug...
Michael Mann [Sat, 22 Jun 2013 15:08:18 +0000 (15:08 -0000)]
More proper fix (per advice of bug 8515 (https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8515)) of global variables complained about in bug 7060 (https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=7060).  Not sure if algorithm_id of x509af dissector could also be a actx->external.direct_reference, so left it alone.

svn path=/trunk/; revision=50113

10 years agoMake TCP segment length a generated field instead of a hidden field.
Ronnie Sahlberg [Sat, 22 Jun 2013 02:00:03 +0000 (02:00 -0000)]
Make TCP segment length a generated field instead of a hidden field.

We should deprecate the use of hidden fields, at least for fields that arei
useful in filters. To make it easier for users to discover and use the fields.

Change the highly useful field for TCP segment payload length from
being a hidden field to be a generated field instead.

svn path=/trunk/; revision=50112

10 years agoAdd Uli Heilmeier
Michael Mann [Fri, 21 Jun 2013 23:23:08 +0000 (23:23 -0000)]
Add Uli Heilmeier

svn path=/trunk/; revision=50111

10 years agoAF_INET and AF_INET6 undefined in textpcap.c when compiling on FreeBSD
Stephen Fisher [Fri, 21 Jun 2013 22:43:24 +0000 (22:43 -0000)]
AF_INET and AF_INET6 undefined in textpcap.c when compiling on FreeBSD
9.1-RELEASE with non-standard gcc 4.6, so pull in sys/socket.h when available.

svn path=/trunk/; revision=50110

10 years agoCast away another implicit conversion not allowed in C++ warning
Stephen Fisher [Fri, 21 Jun 2013 22:37:40 +0000 (22:37 -0000)]
Cast away another implicit conversion not allowed in C++ warning

svn path=/trunk/; revision=50109

10 years agoCast away some errors of these types:
Stephen Fisher [Fri, 21 Jun 2013 22:01:37 +0000 (22:01 -0000)]
Cast away some errors of these types:

    request for implicit conversion from 'gpointer' to 'char *' not permitted
    in C++ [-Werror=c++-compat]

and

    enum conversion when passing argument 3 of 'krb5_crypto_init' is invalid
    in C++ [-Werror=c++-compat]

svn path=/trunk/; revision=50108

10 years agoSort the man pages. Add capinfos and reordercap. Should we also any others (i.e...
Chris Maynard [Fri, 21 Jun 2013 15:30:08 +0000 (15:30 -0000)]
Sort the man pages.  Add capinfos and reordercap.  Should we also any others (i.e., asn2deb, dftest, idl2deb, idl2wrs, randpkt)?

svn path=/trunk/; revision=50107

10 years agoget_ax25_name() is *NOT* the right routine to use to translate a
Guy Harris [Fri, 21 Jun 2013 05:24:59 +0000 (05:24 -0000)]
get_ax25_name() is *NOT* the right routine to use to translate a
human-readable representing an AX.25 subaddress (e.g. "KA9Q-01") into
the binary form of an AX.25 address, because what it does is translate
the binary form of an AX.25 address to the human-readable form!

We currently have no routine that does the right thing and, even if we
did, given that some bits in the AX.25 subaddress format are used for
purposes other than representing the call sign and substation ID, so the
matching routines for AX.25 addresses need to ignore certain bits.

For now, we just remove the call to get_ax25_name() (which squelches the
pointer-signedness warning that made me look at this code, and find the
problem, in the first place) and replace it with a comment discussing
the problem and a failure.

The other pointer-signedness warning brought up a question of what to do
with G_REGEX_RAW in the g_regex_match_full() call; it didn't bring up an
immediately obvious *answer*, so we throw a cast at the warning and add
another comment.  (We fix up alignment while we're at it.)

svn path=/trunk/; revision=50106

10 years agorename packet-mp4.c to file-mp4.c
Martin Kaiser [Fri, 21 Jun 2013 05:18:45 +0000 (05:18 -0000)]
rename packet-mp4.c to file-mp4.c

svn path=/trunk/; revision=50105

10 years agoFix some pointer-signedness (well, signedness of what the pointer points
Guy Harris [Fri, 21 Jun 2013 02:59:54 +0000 (02:59 -0000)]
Fix some pointer-signedness (well, signedness of what the pointer points
to) warnings.

svn path=/trunk/; revision=50104

10 years agoDelete the moved files
Luis Ontanon [Fri, 21 Jun 2013 01:30:39 +0000 (01:30 -0000)]
Delete the moved files

svn path=/trunk/; revision=50103

10 years agomove echld to final dest...
Luis Ontanon [Fri, 21 Jun 2013 01:27:28 +0000 (01:27 -0000)]
move echld to final dest...

svn path=/trunk/; revision=50102

10 years agoCompiles, still incomplete...
Luis Ontanon [Fri, 21 Jun 2013 01:25:22 +0000 (01:25 -0000)]
Compiles, still incomplete...

svn path=/trunk/; revision=50101

10 years agoOne more fix for the protocol grey highlighting.
Evan Huus [Fri, 21 Jun 2013 00:49:35 +0000 (00:49 -0000)]
One more fix for the protocol grey highlighting.

svn path=/trunk/; revision=50100

10 years agoYet another fix, put spaces in insensitive or normal state.
Jakub Zawadzki [Fri, 21 Jun 2013 00:42:19 +0000 (00:42 -0000)]
Yet another fix, put spaces in insensitive or normal state.

svn path=/trunk/; revision=50099

10 years agofix for r50095 don't fetch parent hfinfo, traverse tree to get top one.
Jakub Zawadzki [Fri, 21 Jun 2013 00:23:55 +0000 (00:23 -0000)]
fix for r50095 don't fetch parent hfinfo, traverse tree to get top one.

svn path=/trunk/; revision=50098

10 years agocreate libfiledissectors.a during the autotools build
Martin Kaiser [Fri, 21 Jun 2013 00:08:20 +0000 (00:08 -0000)]
create libfiledissectors.a during the autotools build

svn path=/trunk/; revision=50097

10 years agoFix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8799 :
Pascal Quantin [Thu, 20 Jun 2013 23:37:16 +0000 (23:37 -0000)]
Fix https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8799 :
Display certificate as UINT64 and add support for a 64bits unsigned value in BER dissector

svn path=/trunk/; revision=50096

10 years agoHighlight protocol bytes in grey, idea stolen from qtshark
Jakub Zawadzki [Thu, 20 Jun 2013 23:28:35 +0000 (23:28 -0000)]
Highlight protocol bytes in grey, idea stolen from qtshark

svn path=/trunk/; revision=50095

10 years agoSave and restore private_data pointer instead of freeing the old one, since this
Evan Huus [Thu, 20 Jun 2013 23:27:16 +0000 (23:27 -0000)]
Save and restore private_data pointer instead of freeing the old one, since this
function can be called nestedly. Also use wmem memory so it doesn't leak.

Part of https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8828

svn path=/trunk/; revision=50094

10 years agoRegenerate DAP dissector that was changed in r50073 for some reason
Pascal Quantin [Thu, 20 Jun 2013 23:14:38 +0000 (23:14 -0000)]
Regenerate DAP dissector that was changed in r50073 for some reason

svn path=/trunk/; revision=50093

10 years agoSplit README.developer into two: README.developer and README.dissector. All the
Evan Huus [Thu, 20 Jun 2013 22:46:08 +0000 (22:46 -0000)]
Split README.developer into two: README.developer and README.dissector. All the
documentation specific to dissectors should now live in README.dissector - what
remains in README.developer should be useful to anybody coding on Wireshark
regardless of if you're working on the GUI, wiretap, dissectors, etc.

This first pass I did a fairly dumb split of copy-pasting relevant chunks from
one file to the other. There are probably fragments that aren't in the right
file anymore, so cleanup welcome.

svn path=/trunk/; revision=50092

10 years agoinform expert info layer about the current number of packet comments
Martin Kaiser [Thu, 20 Jun 2013 22:25:33 +0000 (22:25 -0000)]
inform expert info layer about the current number of packet comments
when the last comment is removed and we have no other expert info,
the maximum severity is changed from comment to none

svn path=/trunk/; revision=50091

10 years agoFix bug https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8829 :
Pascal Quantin [Thu, 20 Jun 2013 22:21:24 +0000 (22:21 -0000)]
Fix bug https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=8829 :
Prevent an out of bounds access in 802.11 Radiotap dissector

svn path=/trunk/; revision=50090