obnox/wireshark/wip.git
24 years agoMove calls to "dissector_add()" out of the register routines for TCP and
guy [Sat, 8 Apr 2000 07:07:42 +0000 (07:07 +0000)]
Move calls to "dissector_add()" out of the register routines for TCP and
UDP and into the handoff registration routines for the protocols in
question.

Make the dissectors for those protocols static if they're not called
outside the dissector's source file.

Get rid of header files if all they did was declare dissectors that are
now static; remove declarations of now-static dissectors from header
files that do more than just declare the dissector.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1823 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoJeff Foster's rlogin dissector, and changes to the TCP dissector to
guy [Sat, 8 Apr 2000 03:32:10 +0000 (03:32 +0000)]
Jeff Foster's rlogin dissector, and changes to the TCP dissector to
export the urgent pointer (as the rlogin dissector needs it).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1822 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMinor update to SMB browse display code ...
sharpe [Sat, 8 Apr 2000 01:28:13 +0000 (01:28 +0000)]
Minor update to SMB browse display code ...

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1821 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoDeclare, in "wtap.h", Richard's new "wtap_read()" function.
guy [Sat, 8 Apr 2000 01:16:52 +0000 (01:16 +0000)]
Declare, in "wtap.h", Richard's new "wtap_read()" function.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1820 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoSome stuff, like libdencode, needs a wiretap routine that allows
sharpe [Sat, 8 Apr 2000 00:33:04 +0000 (00:33 +0000)]
Some stuff, like libdencode, needs a wiretap routine that allows
them to simply read through the capture file ...

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1819 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAvoid using zero-length arrays - they're an extension that only some
guy [Fri, 7 Apr 2000 19:10:51 +0000 (19:10 +0000)]
Avoid using zero-length arrays - they're an extension that only some
compilers support, and at least some versions of HP's PA-RISC C
compiler, for example (which people do use to compile Ethereal), don't
support them.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1818 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoDisable "Update list of packets in real tiem" and
gram [Fri, 7 Apr 2000 15:37:59 +0000 (15:37 +0000)]
Disable "Update list of packets in real tiem" and
"Automatic scrolling in live capture" for Win32.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1817 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPatch for multiple %doc lines in RPM spec file.
gram [Fri, 7 Apr 2000 10:22:20 +0000 (10:22 +0000)]
Patch for multiple %doc lines in RPM spec file.
From Patrick Higgins <phiggins@transzap.com>

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1816 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoWhen searching for a frame that matches a filter expression, we
guy [Fri, 7 Apr 2000 08:00:34 +0000 (08:00 +0000)]
When searching for a frame that matches a filter expression, we
shouldn't stop the search as soon as we get back to the starting frame,
we should stop the search if the filter expression doesn't match and the
frame we tried it on was the starting frame - it's OK if we find the
starting frame, it's just not OK to continue if we don't find that
frame.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1815 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPlug a memory leak in the "Find Frame" code - we weren't freeing the
guy [Fri, 7 Apr 2000 07:48:15 +0000 (07:48 +0000)]
Plug a memory leak in the "Find Frame" code - we weren't freeing the
protocol tree after checking whether it matched the filter we supplied.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1814 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMove to version 0.8.6.
gram [Fri, 7 Apr 2000 03:18:08 +0000 (03:18 +0000)]
Move to version 0.8.6.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1812 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoDefine "popup_menu_object" only in "gtk/menu.c", and just declare it as
guy [Thu, 6 Apr 2000 06:52:10 +0000 (06:52 +0000)]
Define "popup_menu_object" only in "gtk/menu.c", and just declare it as
an "extern" in "gtk/menu.h", to avoid complaints from the linker on some
platforms.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1810 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoUpdate the document to reflect Gilbert's changes to split
guy [Thu, 6 Apr 2000 06:38:24 +0000 (06:38 +0000)]
Update the document to reflect Gilbert's changes to split
"proto_tree_add_item_format()" into multiple routines for different item
types, and to note that a subtree can be added under any item.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1809 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoUse the "-o" flag, rather than using the "-t" flag and redirecting the
guy [Thu, 6 Apr 2000 06:19:42 +0000 (06:19 +0000)]
Use the "-o" flag, rather than using the "-t" flag and redirecting the
standard output, in the rules to get Flex to produce scanner code; that
way, if Flex fails to run for some reason, we don't leave around a
zero-length or otherwise incorrect "XXX-scanner.c" file that might
keep a subsequent make from thinking it has to generate that file.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1808 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoFix a typo in the rules to make "dfilter-grammar.c" and
guy [Thu, 6 Apr 2000 06:18:50 +0000 (06:18 +0000)]
Fix a typo in the rules to make "dfilter-grammar.c" and
"dfilter-grammar.h".

Use the "-o" flag, rather than using the "-t" flag and redirecting the
standard output, in the rules to get Flex to produce scanner code; that
way, if Flex fails to run for some reason, we don't leave around a
zero-length or otherwise incorrect "XXX-scanner.c" file that might
keep a subsequent make from thinking it has to generate that file.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1807 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAfter destroying the mem chunk for the packet list, null out the pointer
guy [Thu, 6 Apr 2000 06:04:24 +0000 (06:04 +0000)]
After destroying the mem chunk for the packet list, null out the pointer
pointing to it, so that if we call "close_cap_file()" again, we don't
try to destroy an already-destroyed mem chunk.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1806 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd in an include of "snprintf.h" on platforms where it's necessary.
guy [Thu, 6 Apr 2000 03:59:28 +0000 (03:59 +0000)]
Add in an include of "snprintf.h" on platforms where it's necessary.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1805 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd Attribute Tunnel-Password, from Florian Lohoff.
gram [Wed, 5 Apr 2000 16:01:39 +0000 (16:01 +0000)]
Add Attribute Tunnel-Password, from Florian Lohoff.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1804 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoSpelling fix from Florian Lohoff
gram [Wed, 5 Apr 2000 15:57:04 +0000 (15:57 +0000)]
Spelling fix from Florian Lohoff

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1803 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoUpdate version and plugin directory.
gram [Wed, 5 Apr 2000 11:53:30 +0000 (11:53 +0000)]
Update version and plugin directory.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1801 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPlugin API is at version 0.8.5 (reflected in directory name for plugins)
gram [Wed, 5 Apr 2000 11:45:50 +0000 (11:45 +0000)]
Plugin API is at version 0.8.5 (reflected in directory name for plugins)

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1800 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPackaging fixes.
gram [Wed, 5 Apr 2000 11:04:46 +0000 (11:04 +0000)]
Packaging fixes.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1799 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd a test to check if there is at least one enabled plugin before searching
oabad [Tue, 4 Apr 2000 22:26:36 +0000 (22:26 +0000)]
Add a test to check if there is at least one enabled plugin before searching
the plugin list.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1798 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd "dissector_add()" to the list of routines exported to plugins via
guy [Tue, 4 Apr 2000 21:46:29 +0000 (21:46 +0000)]
Add "dissector_add()" to the list of routines exported to plugins via
the Big Function Pointer Table.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1797 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd some additional notes about
guy [Tue, 4 Apr 2000 21:07:33 +0000 (21:07 +0000)]
Add some additional notes about

building GTK+/GLib with 64-bit integer support when using HP's C
compiler;

libpcap;

from Jost Martin.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1796 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMove to version 0.8.5
gram [Tue, 4 Apr 2000 19:47:20 +0000 (19:47 +0000)]
Move to version 0.8.5

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1795 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd assertions from Ben Fowler <wapdev@leedsnet.com>.
gram [Tue, 4 Apr 2000 17:07:07 +0000 (17:07 +0000)]
Add assertions from Ben Fowler <wapdev@leedsnet.com>.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1794 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd #include "plugins.h" to get definition of HAVE_PLUGINS
gram [Tue, 4 Apr 2000 16:33:57 +0000 (16:33 +0000)]
Add #include "plugins.h" to get definition of HAVE_PLUGINS

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1793 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd "clean" rules in subdirectories, and run subdirectory "nmake -f
guy [Tue, 4 Apr 2000 08:25:35 +0000 (08:25 +0000)]
Add "clean" rules in subdirectories, and run subdirectory "nmake -f
Makefile.nmake clean" when "nmake -f Makefile.nmake clean" is done in
the top-level directory, so that "nmake -f Makefile.nmake clean" cleans
everything up.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1792 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoThe Win32 port of libpcap requires that programs with which it's built
guy [Tue, 4 Apr 2000 08:22:13 +0000 (08:22 +0000)]
The Win32 port of libpcap requires that programs with which it's built
be built as multi-threaded programs; add "/MT" to the list of compiler
flags.

Add "clean" rules in subdirectories, and run subdirectory "nmake -f
Makefile.nmake clean" when "nmake -f Makefile.nmake clean" is done in
the top-level directory, so that "nmake -f Makefile.nmake clean" cleans
everything up.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1791 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoTurn "ethereal_proto_init()" and "ethereal_proto_cleanup()" into
guy [Tue, 4 Apr 2000 07:03:07 +0000 (07:03 +0000)]
Turn "ethereal_proto_init()" and "ethereal_proto_cleanup()" into
"dissect_init()" and "dissect_cleanup()", in "packet.c", so that we
don't duplicate those routines in Ethereal and Tethereal (and so on),
and don't have to remember to update N different versions of them if we
have to change the way we do one-time initialization and cleanup.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1790 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoUse the new split between protocol registration and protocol handoff
guy [Tue, 4 Apr 2000 06:46:41 +0000 (06:46 +0000)]
Use the new split between protocol registration and protocol handoff
registration routines to get rid of the special handling of ONC RPC
protocols - dissectors for ONC RPC-based protocols should register their
protocol, fields, and ETT values in a protocol registration routine, and
register themselves with the ONC RPC dissector in their protocol handoff
routine, so that the latter is done after the ONC RPC dissector's
protocol registration routine is called, so that the data structures
needed when dissectors for ONC RPC protocols register themselves with
the ONC RPC dissector have been initialized.

Get rid of "init_dissect_rpc()", which initializes said data structures;
do that in "proto_register_rpc()" instead.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1789 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMake "make-reg-dotc" generate a "register_all_protocol_handoffs()"
guy [Tue, 4 Apr 2000 06:17:30 +0000 (06:17 +0000)]
Make "make-reg-dotc" generate a "register_all_protocol_handoffs()"
routine, which calls all routines found in the dissector source files
with names that match " proto_reg_handoff_[a-z_0-9A-Z]*".

Call "register_all_protocol_handoffs()" after calling
"register_all_protocols()" - "register_all_protocols()" needs to be
called first, so that all protocols can register their fields, because
registering a dissector as being called if field "proto.port" is equal
to N requires that "proto.port" be a registered field.

Give DNS a handoff registration routine, and register its dissector to
be called if "udp.port" is UDP_PORT_DNS; remove the registration of DNS
from "packet-udp.c", and make "dissect_dns()" static (as nobody else
need know that it exists).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1788 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMake "register.c" depend on "$(srcdir)/make-reg-dotc", so that if the
guy [Tue, 4 Apr 2000 06:10:44 +0000 (06:10 +0000)]
Make "register.c" depend on "$(srcdir)/make-reg-dotc", so that if the
"make-reg-dotc" script is changed, "register.c" is remade (as the change
may change what it generates).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1787 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoDo all the UDP port numbers that we can, and that don't require special
guy [Tue, 4 Apr 2000 05:54:59 +0000 (05:54 +0000)]
Do all the UDP port numbers that we can, and that don't require special
processing (as TFTP does), and don't have comments suggesting that extra
checks are needed, with the port table.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1786 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMake a routine that takes a dissector table, a port number, and
guy [Tue, 4 Apr 2000 05:37:36 +0000 (05:37 +0000)]
Make a routine that takes a dissector table, a port number, and
pd/offset/fd/tree arguments, looks up the port number in the dissector
table, and:

if it finds it, call the corresponding dissector routine with
the pd/offset/fd/tree arguments, and return TRUE;

if it doesn't find it, return FALSE.

Use that in the TCP and UDP dissectors.

Don't add arbitrary UDP ports for which a dissector is found in the
table as ports that should be dissected as TFTP; this should only be
done if we find a packet going from port XXX to the official TFTP port.

Don't register TFTP in UDP's dissector table, as it has to be handled
specially (i.e., we have to add the source port as a TFTP port, although
we really should register the source port *and* IP address); eventually,
we should move that registration to the TFTP dissector itself, at which
point we can register TFTP normally.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1785 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAllocate the new mem chunk for the capture in "open_cap_file()", rather
guy [Tue, 4 Apr 2000 05:14:48 +0000 (05:14 +0000)]
Allocate the new mem chunk for the capture in "open_cap_file()", rather
than duplicating the allocation code in "read_cap_file()" and
"start_tail_cap_file()".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1784 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoRename find_hfinfo_record() to proto_registrar_get_nth() since
gram [Tue, 4 Apr 2000 02:34:40 +0000 (02:34 +0000)]
Rename find_hfinfo_record() to proto_registrar_get_nth() since
all the other public functions in proto.c start with "proto_".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1783 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoSet cf->plist_chunk in the other place where needed. Guy's experiencing
gram [Mon, 3 Apr 2000 22:28:51 +0000 (22:28 +0000)]
Set cf->plist_chunk in the other place where needed. Guy's experiencing
some network problems today, so I checked this in for him.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1782 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMove the creation of, and registration of protocols known to UDP in, the
guy [Mon, 3 Apr 2000 09:41:31 +0000 (09:41 +0000)]
Move the creation of, and registration of protocols known to UDP in, the
hash table attached to "udp.port" out of "init_dissect_udp()" into
"proto_register_udp()", so that it's done the way TCP does it, and then
get rid of "init_dissect_udp()".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1781 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoFix a compile error introduced by the previous checkin, which also moved
guy [Mon, 3 Apr 2000 09:37:39 +0000 (09:37 +0000)]
Fix a compile error introduced by the previous checkin, which also moved
the check for plugins after the check for ONC RPC protocols, so that we
do the checks in the same order for TCP and UDP (ONC RPC first, as we
expect the RPC heuristics not to get false hits, and ONC RPC protocols
could well use ports that are nominally assigned to other protocols).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1780 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoJeff Foster's patch to support attaching a hash table to a protocol
guy [Mon, 3 Apr 2000 09:24:12 +0000 (09:24 +0000)]
Jeff Foster's patch to support attaching a hash table to a protocol
field, to allow dissectors to register their dissection routine in a
particular field's hash table with a particular "port" value, and to
make the TCP and UDP dissectors support that for their "port" field and
to look up ports in that hash table.

This replaces the hash table that the UDP dissector was using.

There's still more work needed to make this useful - right now, the hash
tables are attached to the protocol field in the register routines for
the TCP and UDP protocols, which means that the register routines for
protocols that run atop TCP and UDP can't use this unless their register
routines happen to be called after those for TCP and/or UDP, and several
other protocols need to attach hash tables to fields, and there's no
single global field for Ethernet types so we can't even attach a hash
table to such a field to allow protocols to register themselves with a
particular Ethertype - but it's a start.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1779 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAnother update from Doug Nazar.
guy [Mon, 3 Apr 2000 09:00:31 +0000 (09:00 +0000)]
Another update from Doug Nazar.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1778 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoNote that it didn't seem to save a significant amount of time or space
guy [Mon, 3 Apr 2000 08:57:17 +0000 (08:57 +0000)]
Note that it didn't seem to save a significant amount of time or space
to use a G_ALLOC_ONLY memory chunk rather than a G_ALLOC_AND_FREE memory
chunk.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1777 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoUse a GMemChunk for allocating and freeing "frame_data" structures; it
guy [Mon, 3 Apr 2000 08:42:45 +0000 (08:42 +0000)]
Use a GMemChunk for allocating and freeing "frame_data" structures; it
reduces the memory required for one big capture file I have by about 10
megabytes, and speeds up the reading in of that file, presumably because
it amortizes the memory-allocator space and CPU time overhead over
a set of several "frame_data" structures.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1776 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoImprovements from Doug Nazar.
guy [Mon, 3 Apr 2000 07:48:55 +0000 (07:48 +0000)]
Improvements from Doug Nazar.

One of those improvements adds an "exit now" after we've filled in the
protocol and info columns if we're not building a protocol tree,
obviating the need for the "parse only the first request in the sequence
if we're not building a protocol tree" code I stuck in, so remove the
latter code.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1775 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoSplit "filter_dialog_cb()" into "filter_dialog_cb()", which pops up a
guy [Sat, 1 Apr 2000 12:03:42 +0000 (12:03 +0000)]
Split "filter_dialog_cb()" into "filter_dialog_cb()", which pops up a
"global" dialog box when "Edit:Filters" is selected, so that the list of
filters can be edited, and "filter_browse_cb()", which pops up a dialog
box associated with a "Filter:" button and a text entry widget attached
to that button, so that a filter can be selected or saved (although it
also supports the same editing that the "global" dialog box does).

Have "filter_dialog_cb()" connect the window in which the "Filter:"
button lives and the filter dialog box, so that:

if the window in which the "Filter:" button lives goes away, so
does the filter dialog box (as it no longer has a text widget
into which it can stuff the selected filter);

if the "Filter:" button is clicked when there's already a filter
dialog box open, we just reactivate that existing dialog box
rather than popping up a new one.

Also keep a pointer to the "global" filter dialog box, so that we also
arrange that there's only one of them (by reactivating the existing on
if "Edit:Filters" is selected when there's already a "global" filter
dialog box open).

Keep around pointers to the dialog boxes that contain the "Filter:"
buttons, so that we can arrange that there be only one of them (that was
a side-effect of an earlier attempt at fixing the problems described
above, but it's still useful for keeping multiple competing dialog boxes
from being open - there's more of that to be done).

Make the pointer to the "Open Capture File" dialog box widget static to
"file_dlg.c" - nobody outside of "file_dlg.c cares about it.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1774 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoThere's no need to catch the "delete_event" signal on "Follow TCP
guy [Sat, 1 Apr 2000 11:30:53 +0000 (11:30 +0000)]
There's no need to catch the "delete_event" signal on "Follow TCP
Stream" windows - the window should always be deleted in that situation,
so there's no need for a signal handler that might return TRUE (meaning
"don't delete the window"), and the "destroy" handler gets called when
the window actually gets destroyed, so there's no need to do any cleanup
in the "delete_event" handler.

Catch the "delete_event" signal on the main window in a routine with the
right signature, and that returns FALSE so that the window actually gets
deleted.

Call "close_cap_file()" in the callback for the "File:Quit" menu item
(which is also called by the "delete_event" handler for the main
window), rather than calling it after "gtk_main()" returns -
"close_cap_file()" manipulates stuff in the main window, and if we do so
after "gtk_main()" returns, it appears that the main window may have
disappeared (if we are exiting because the user deleted the main
window), in which case we can get crashes or other errors when
"close_cap_file()" tries to manipulate stuff in the main window.

There's no need to catch the "destroy" signal on the main window - we do
some of the cleanup in the handler for "delete_event" (we have to, for
reasons described above), and we do the rest of it after the main
routine returns.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1773 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoExit from Ethereal by making the main loop exit, so that any quit
guy [Sat, 1 Apr 2000 10:23:01 +0000 (10:23 +0000)]
Exit from Ethereal by making the main loop exit, so that any quit
routines we register get called.

Register the "follow TCP stream" windows with "gtk_quit_add_destroy()",
so that, when Ethereal exits, they all get destroyed; this means that
their destroy callbacks get called, which means that they get to delete
their temporary files, so that, if you exit Ethereal without manually
closing the "follow TCP stream" windows, the temporary files don't get
left around.

Exit from Ethereal's "main()" function by calling "gtk_exit()", rather
than "exit()", so that we do whatever cleanup GTK+ requires.  (We used
to call "gtk_exit()" in the callback for the "File:Quit" menu item and
the "delete" callback for the main window, but we now just call
"gtk_main_quit()" so that we exit from the main loop.)

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1772 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoNo work need be done when a summary dialog box is destroyed, or when the
guy [Sat, 1 Apr 2000 09:16:49 +0000 (09:16 +0000)]
No work need be done when a summary dialog box is destroyed, or when the
window manager tries to delete it, so no callback is needed for the
"destroy" or "delete_event" signals (grabs are, at least in GTK+ 1.2.7,
removed when a widget is destroyed, and there's no need for the destroy
callback to destroy the widget itself; the delete event handler *could*,
for example, pop up a dialog box saying "Do you really want to close
this?", and allow the user to back out of the operation, but there's
no unsaved work that would be lost by closing the window, so there's no
point in having a delete event handler that does that).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1771 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd a counter : "enabled_plugins_number", to record how many plugins are
oabad [Fri, 31 Mar 2000 21:42:24 +0000 (21:42 +0000)]
Add a counter : "enabled_plugins_number", to record how many plugins are
enabled. The counter is incremented in enable_plugin() and decremented
in disable_plugin().
In add_packet_to_packet_list(), we check this counter (instead of
plugin_list) to see if there is at least one enabled plugin. If this is
the case, we must build the protocol tree.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1770 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoRick Jones of HP says that the "-Ae" flag to the HP ANSI C compiler
guy [Fri, 31 Mar 2000 20:45:29 +0000 (20:45 +0000)]
Rick Jones of HP says that the "-Ae" flag to the HP ANSI C compiler
gives you everything that "-Aa" does, plus the extensions that you get
only with "-Ae", and that there's no need to specify "-Aa" if you've
specified "-Ae".  Therefore, we get rid of "-Aa".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1769 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPut in the beginnings of checks for ASN.1 dissection errors.
guy [Fri, 31 Mar 2000 10:22:24 +0000 (10:22 +0000)]
Put in the beginnings of checks for ASN.1 dissection errors.

The "present" choice in the type Filter is, in LDAP V2,
AttributeType, and, in LDAP V3, it's AttributeDescription.  Both of
those are just LDAPString, which is, in turn, OCTET STRING, so it should
be required to have the primitive representation (unless and until we
add support for the constructed representation of octet strings - but
RFC 1777, the LDAP V2 spec, says

   (2)  Bitstrings and octet strings and all character string types
        will be encoded in the primitive form only.

and RFC 2251, the LDAP V3 spec, says

   (2) OCTET STRING values will be encoded in the primitive form only.

so we shouldn't ever see it with the constructed representation), and be
parsed with "asn1_octet_string_value_decode()", as, by that point, we've
already dissected the ASN.1 id and length.

Put the bind authorization type into the protocol tree before switching
on the type, so that it goes in even if it's not something we yet
dissect, and actually pass it as an argument to "proto_tree_add_item()"
(alas, "proto_tree_add_item()" is a varargs function, so this error
couldn't have been detected by the compiler).

When not constructing a protocol tree, quit "dissect_ldap()" after
dissecting the first operation - we don't need to dissect the others.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1768 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoJochen Friedrich <jochen@nwe.de>
gram [Thu, 30 Mar 2000 21:41:11 +0000 (21:41 +0000)]
Jochen Friedrich <jochen@nwe.de>
protocol type 0x0c in AIX iptrace is used for the IBM SP switch
devices. Encoding is RAW IP...

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1767 f5534014-38df-0310-8fa8-9805f1628bb7

24 years ago"add_rr_to_tree()" is exported from "packet-dns.c" for use by the NBNS
guy [Thu, 30 Mar 2000 01:52:40 +0000 (01:52 +0000)]
"add_rr_to_tree()" is exported from "packet-dns.c" for use by the NBNS
protocol, which is DNS-derived; hopefully, Microsoft won't shovel any
more stuff into NBNS (I suspect that they ultimately want to make DNS
replace it completely), so it won't pick up stuff such as OPT RRs.

As such, we don't need to export "add_opt_rr_to_tree()", so make it
static to "packet-dns.c".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1766 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoadd dissector for OPT pseudo-RR, in RFC2671 section 4.
itojun [Thu, 30 Mar 2000 01:33:10 +0000 (01:33 +0000)]
add dissector for OPT pseudo-RR, in RFC2671 section 4.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1765 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoGet rid of the "Find Frame" and "Go To Frame" items under "Display"; we
guy [Wed, 29 Mar 2000 22:39:19 +0000 (22:39 +0000)]
Get rid of the "Find Frame" and "Go To Frame" items under "Display"; we
put them under "Edit" a while ago.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1764 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoUse the "-o" flag, rather than using the "-t" flag and redirecting the
guy [Wed, 29 Mar 2000 19:55:51 +0000 (19:55 +0000)]
Use the "-o" flag, rather than using the "-t" flag and redirecting the
standard output, in the rules to get Flex to produce scanner code; that
way, if Flex fails to run for some reason, we don't leave around a
zero-length or otherwise incorrect "XXX-scanner.c" file that might
keep a subsequent make from thinking it has to generate that file.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1763 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd ID for Lucent, and sort lines.
gram [Wed, 29 Mar 2000 16:55:08 +0000 (16:55 +0000)]
Add ID for Lucent, and sort lines.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1762 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoFix from Michael Johnston to check the correct bit when testing the
guy [Wed, 29 Mar 2000 09:29:16 +0000 (09:29 +0000)]
Fix from Michael Johnston to check the correct bit when testing the
broadcast flag in DHCP packets.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1761 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoUpdates from Doug Nazar.
guy [Wed, 29 Mar 2000 09:25:21 +0000 (09:25 +0000)]
Updates from Doug Nazar.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1760 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoupdate RFC1966 (BGP route reflection) support.
itojun [Wed, 29 Mar 2000 06:27:54 +0000 (06:27 +0000)]
update RFC1966 (BGP route reflection) support.
From: Greg Hankins <gregh@twoguys.org>

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1759 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd getopt.c (from GNU libc) for use on Win32.
gram [Tue, 28 Mar 2000 20:20:11 +0000 (20:20 +0000)]
Add getopt.c (from GNU libc) for use on Win32.
Ethereal/win32 now supports command-line options.
Tethereal now compiles on win32, except for the fact that I haven't
put the Makefile.nmake changes in for that yet.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1758 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPatches from Andreas Sikkema:
guy [Tue, 28 Mar 2000 08:11:52 +0000 (08:11 +0000)]
Patches from Andreas Sikkema:

On Win32, always save a temporary capture file by copying -
Win32 systems don't allow you to rename a file that is open, and
we have the temporary file open.

When saving by copying the raw bytes of a capture file, create
the target file with "open()", using the O_BINARY flag, rather
than with "creat()"; on Win32 systems, "creat()" apparently
opens the file as a text file rather than a binary file.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1757 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoDoug Nazar's LDAP dissector.
guy [Tue, 28 Mar 2000 07:12:36 +0000 (07:12 +0000)]
Doug Nazar's LDAP dissector.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1756 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoRalf Schneider's change to boost the maximum size of a line in the
guy [Tue, 28 Mar 2000 06:20:08 +0000 (06:20 +0000)]
Ralf Schneider's change to boost the maximum size of a line in the
"filters" file (we also make it a #define so you only have to change it
in one place).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1755 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoFix from Ralf Schneider to the IS Neighbor dissection.
guy [Tue, 28 Mar 2000 06:12:15 +0000 (06:12 +0000)]
Fix from Ralf Schneider to the IS Neighbor dissection.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1754 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoChange dissect_ppp() to accept offset.
gram [Mon, 27 Mar 2000 17:53:20 +0000 (17:53 +0000)]
Change dissect_ppp() to accept offset.
Change GRE dissector to call dissect_ppp() instead of dissect_payload_ppp().

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1753 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd type tag values for the REAL and ENUMERATED types.
guy [Mon, 27 Mar 2000 08:51:05 +0000 (08:51 +0000)]
Add type tag values for the REAL and ENUMERATED types.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1752 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoNote that we may want to add support for the constructed encoding of
guy [Mon, 27 Mar 2000 08:26:23 +0000 (08:26 +0000)]
Note that we may want to add support for the constructed encoding of
octet strings, as well as the primitive encoding (if we ever run into a
protocol where the constructed encoding is allowed).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1751 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdding definitions of routines ... Will want to get rid of malloc and use
sharpe [Sun, 26 Mar 2000 07:59:47 +0000 (07:59 +0000)]
Adding definitions of routines ... Will want to get rid of malloc and use
the glib equivalents ...

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1750 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoSet the per frame data pointer to NULL when a new frame is read in.
sharpe [Sun, 26 Mar 2000 07:03:52 +0000 (07:03 +0000)]
Set the per frame data pointer to NULL when a new frame is read in.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1749 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoFirst pass at per frame proto data. Keep each proto block as a GSList list.
sharpe [Sun, 26 Mar 2000 06:57:41 +0000 (06:57 +0000)]
First pass at per frame proto data. Keep each proto block as a GSList list.

Use glib as far as possible. Currently have data structures and routines
defined ... Next will write the routines ...

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1748 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoGet rid of some unnecessary variables.
guy [Thu, 23 Mar 2000 10:49:33 +0000 (10:49 +0000)]
Get rid of some unnecessary variables.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1747 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoDon't use temporary buffers for strings; use the data in the packet
guy [Thu, 23 Mar 2000 10:25:38 +0000 (10:25 +0000)]
Don't use temporary buffers for strings; use the data in the packet
buffer directly.

Don't use "strlen()" to determine how much Telnet data there is; we have
the number of bytes of Telnet data in a variable, and using "strlen()"
makes it not correctly handle Telnet data with '\0' in it.

Correctly handle Telnet options we don't know about.

General cleanup.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1746 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoRemove small memory leak when converting strings that represent
gram [Thu, 23 Mar 2000 05:43:57 +0000 (05:43 +0000)]
Remove small memory leak when converting strings that represent
byte sequences in display filters to byte arrays. This was caused
by a duplicate g_strdup() in my code.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1745 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoWhen dissecting an NFS file handle, show the raw bytes of the file
guy [Thu, 23 Mar 2000 00:38:11 +0000 (00:38 +0000)]
When dissecting an NFS file handle, show the raw bytes of the file
handle data (snoop does this, and it's somewhat convenient if you're
trying to see whether, say, an ACCESS call is acting on the file whose
file handle you got back from a previous MOUNT request).

Boost the maximum number of bytes worth of hex data shown by
"bytes_to_str()" from 10 to 16, so that we can show the file handle 16
bytes per line (as snoop does).

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1744 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoFlorian Lohoff's fixes to L2TP result and error code dissection.
guy [Wed, 22 Mar 2000 23:49:20 +0000 (23:49 +0000)]
Florian Lohoff's fixes to L2TP result and error code dissection.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1743 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoMake the previous checkin actually compile on big-endian machines.
guy [Wed, 22 Mar 2000 23:47:28 +0000 (23:47 +0000)]
Make the previous checkin actually compile on big-endian machines.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1742 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoByte-swap the frame table on big-endian machines.
guy [Wed, 22 Mar 2000 09:52:21 +0000 (09:52 +0000)]
Byte-swap the frame table on big-endian machines.

Free it as soon as we're at the end of the sequential pass through the
file; that way, if we keep the capture file open with Wiretap even after
that's done (as I may do as part of some stuff I'm working on), we
at least aren't hanging on to the frame table memory after that point.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1741 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoIn a Network Monitor capture file, get the starting offsets of frames
guy [Wed, 22 Mar 2000 07:06:59 +0000 (07:06 +0000)]
In a Network Monitor capture file, get the starting offsets of frames
from the frame table - Network Monitor 2.x, at least, doesn't always
write frame N+1 right after frame N.

To do that, we need to mallocate a big array to hold the frame table,
and free it when we close the capture file; this requires that we have
capture-file-type-specific close routines as well as
capture-file-type-specific read routines - we let it the pointer to that
routine be null if it's not needed.  Given that, we might as well get
rid of the switch statement in "wtap_close()", in favor of using
capture-file-type-specific close routines, as per the comment before
that switch statement.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1740 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoSome cleanup of the RADIUS AVP output (Tunnel Tagging Tag), and a bit of
guy [Tue, 21 Mar 2000 19:20:41 +0000 (19:20 +0000)]
Some cleanup of the RADIUS AVP output (Tunnel Tagging Tag), and a bit of
Vendor-specific decoding (shows at least the Vendor), from Florian
Lohoff.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1739 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAndreas Sikkema's patch to make the SMB dissector not blithely pass a
guy [Tue, 21 Mar 2000 10:15:50 +0000 (10:15 +0000)]
Andreas Sikkema's patch to make the SMB dissector not blithely pass a
null "request_val->last_transact_command" to "col_add_fstr()".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1738 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPaul Welchinski's changes to, on Win32 systems:
guy [Tue, 21 Mar 2000 06:52:13 +0000 (06:52 +0000)]
Paul Welchinski's changes to, on Win32 systems:

properly handle ASCII vs. Unicode in the list of interfaces;

initialize Winsock before starting a capture, so that the code
in the Win32 libpcap to get the IP address and netmask by
translating the host name to an IP address works.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1737 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoCreate the tree entry for resource records in DNS replies, and add the
guy [Tue, 21 Mar 2000 06:21:37 +0000 (06:21 +0000)]
Create the tree entry for resource records in DNS replies, and add the
record type to the info column, before parsing the reply; add the text
for the entry, and any additional information for the info column, in
the code that handles the reply.

Don't use stuff from the resource record if you ran past the end of the
packet.

Fix some bit-scanning code.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1736 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd support for TXT and WKS records to the DNS dissector.
guy [Tue, 21 Mar 2000 05:15:12 +0000 (05:15 +0000)]
Add support for TXT and WKS records to the DNS dissector.

Add EIGRP and VINES to the list of protocols "ipprotostr()" knows about.

Get rid of the "proto_vals" table in "packet-ip.c" - it's not used, and
the two entries it had that weren't in the table in "ipproto.c" have
been moved there.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1735 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoJochen Friedrich's fix to IPv6 fragment handling.
guy [Tue, 21 Mar 2000 04:15:14 +0000 (04:15 +0000)]
Jochen Friedrich's fix to IPv6 fragment handling.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1734 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoChange dfilter_init() to check for empty-string abbreviations and for
gram [Mon, 20 Mar 2000 22:52:48 +0000 (22:52 +0000)]
Change dfilter_init() to check for empty-string abbreviations and for
duplicate abbreviations. All mods to packet-*.c files are fixes to remove
those cases.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1733 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdd fields tr.addr, fddi.addr, and isl.addr that act like eth.addr, matching
gram [Mon, 20 Mar 2000 22:22:45 +0000 (22:22 +0000)]
Add fields tr.addr, fddi.addr, and isl.addr that act like eth.addr, matching
either *.src or *.dst

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1732 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdded Novell option decoding patch from Fredrik Sorensson <fsn@volvo.se>.
gerald [Mon, 20 Mar 2000 21:39:00 +0000 (21:39 +0000)]
Added Novell option decoding patch from Fredrik Sorensson <fsn@volvo.se>.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1731 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoAdded Jeff Foster's patch to add an eth.addr field.
gram [Mon, 20 Mar 2000 21:21:33 +0000 (21:21 +0000)]
Added Jeff Foster's patch to add an eth.addr field.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1730 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoFlorian Lohoff's changes for RADIUS tunnel attributes and for the
guy [Mon, 20 Mar 2000 18:30:59 +0000 (18:30 +0000)]
Florian Lohoff's changes for RADIUS tunnel attributes and for the
Internet Draft for RADIUS tunnel attribute tagging.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1729 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoIn "add_packet_to_packet_list()", always use the "buf" argument to refer
guy [Mon, 20 Mar 2000 04:55:10 +0000 (04:55 +0000)]
In "add_packet_to_packet_list()", always use the "buf" argument to refer
to the raw packet data, don't use "cf->pd" - "cf->pd" doesn't contain
the packet data when the file is being read in, it only contains it when
we're re-reading packets.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1728 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoChange the type of the "v120.control" protocol field from FT_STRING to
guy [Fri, 17 Mar 2000 05:21:48 +0000 (05:21 +0000)]
Change the type of the "v120.control" protocol field from FT_STRING to
FT_UINT16; the xDLC code requires control fields to be FT_UINTn, as it
uses "proto_tree_add_uint_format()" to add that field to the protocol
tree.

Always pass TRUE as the "is_extended" argument to
"dissect_xdlc_control()" - the V.120 spec says "The data link control
sublayer provides the procedures and formats of fields for data link
layer peer-to-peer communication.  ...  For formats and the elements of
procedures, see Recommendation Q.922", and Q.922 seems to at least allow
for modulo 128 operation; it also speaks of SABME, as does V.120, and I
*think* the "extended" in "Set asynchronous balanced mode extended"
means "modulo 128".

Dissect the terminal adaptation sublayer header field after dissecting
the control field, so that the V.120 tree shows the data in sequential
order.

Calculates the length of the V.120 header based on the control field
type (using the "XDLC_CONTROL_LEN()" macro) and the length of the
terminal adaptation sublayer header ("dissect_v120_header()" now returns
the length of what it dissects).

Call "dissect_data()" on the V.120 payload, so you at least know how
much data there is, and can click on an item to see it highlighted.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1727 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoPut the EIGRP packet type into the INFO column in the packet list entry
guy [Thu, 16 Mar 2000 08:23:21 +0000 (08:23 +0000)]
Put the EIGRP packet type into the INFO column in the packet list entry
for an EIGRP packet.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1726 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoChanged the type of the lapb.control protocol field from FT_STRING to
oabad [Wed, 15 Mar 2000 19:17:08 +0000 (19:17 +0000)]
Changed the type of the lapb.control protocol field from FT_STRING to
FT_UINT8.
The lapb dissector was broken by the changes to `proto_tree_add_uint_format'
in xdlc.c because of the type of this field.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1725 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoGive all the Ethereal APIs available to plugins their own function
guy [Wed, 15 Mar 2000 19:09:52 +0000 (19:09 +0000)]
Give all the Ethereal APIs available to plugins their own function
pointers, and call the APIs by calling through the function pointers -
the old technique of using wrappers didn't work, as the underlying
function they all called no longer exists.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1724 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoRename the local variable "pi" to "ti" in several places - "pi" is also
guy [Wed, 15 Mar 2000 18:38:58 +0000 (18:38 +0000)]
Rename the local variable "pi" to "ti" in several places - "pi" is also
the name of the global variable containing packet information, and, on
platforms where dynamically-loaded code can't access global variables
and functions in the main program, we #define "pi" in plugins to be
"(*p_pi)", which causes problems even if a plugin with a local variable
named "pi" *isn't* accessing the global "pi".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1723 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoGive the entry box in the "Find Frame" and "Go To Frame" dialog boxes
guy [Wed, 15 Mar 2000 08:54:24 +0000 (08:54 +0000)]
Give the entry box in the "Find Frame" and "Go To Frame" dialog boxes
the initial focus, so you can start typing immediately.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1722 f5534014-38df-0310-8fa8-9805f1628bb7

24 years agoTreat the two integers after the request ID in a BulkPDU as
guy [Wed, 15 Mar 2000 07:12:55 +0000 (07:12 +0000)]
Treat the two integers after the request ID in a BulkPDU as
"non-repeaters" and "max-repetitions" rather than as "error status" and
"error index".

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@1721 f5534014-38df-0310-8fa8-9805f1628bb7