https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=7135
Evan Huus <eapache@xxxxxxxxx> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #8367| |review_for_checkin?
Flags| |
--- Comment #8 from Evan Huus <eapache@xxxxxxxxx> 2012-05-03 20:48:10 PDT ---
Created attachment 8367
--> https://bugs.wireshark.org/bugzilla/attachment.cgi?id=8367
Fix up handling of link layer types in capture_dlg.c
This one gets every case I've found (including the one that spawned this bug)
and throws in a deadlock fix for good measure.
Briefly:
- Make sure that unsupported link-layer types *always* have a dlt of -1 in the
`device.links` list so that we can tell if they're unsupported.
- When building the combo-box of link-layer types, make the unsupported ones
insensitive (disabled), and be sure to default to a supported one if we have
one.
- Compiling a capture filter requires a supported link-layer type, so disable
the 'Compile BPF' button if there are none.
- It is not always a programming error to have no active pointer in the
combo-box, as this can happen when all the available entries are unsupported
(and thus disabled). Don't abort with g_assert_not_reached() when this happens.
- Unlock the `pcap_compile_mtx` mutex in all cases after compiling a packet
filter. This was causing a deadlock (complete program hang) if you tried to
compile two different filters in a row on the same interface.
--
Configure bugmail: https://bugs.wireshark.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.