Bug 871091

Summary: [abrt] wireshark-gnome-1.8.3-2.fc18: __strlen_sse2_pminub: Process /usr/sbin/wireshark was killed by signal 11 (SIGSEGV)
Product: [Fedora] Fedora Reporter: Michele Baldessari <michele>
Component: wiresharkAssignee: Peter Hatina <phatina>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 18CC: c.david86, fweimer, hancockrwd, huzaifas, jlayton, jsafrane, maci, mapcode.mind, michele, ms, ndevos, paulo.fidalgo.pt, phatina, redhat, rvokal, tsmetana, zman0900
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Unspecified   
Whiteboard: abrt_hash:f2dc699efb48f7dec947aecec6a75c75c12f3dd6
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-03-20 07:52:37 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
File: core_backtrace
none
File: environ
none
File: limits
none
File: backtrace
none
File: cgroup
none
File: smolt_data
none
File: executable
none
File: maps
none
File: dso_list
none
File: build_ids
none
File: proc_pid_status
none
File: var_log_messages
none
File: open_fds
none
"bt f" from a crashed wireshark-1.8.4-1.fc18.x86_64 none

Description Michele Baldessari 2012-10-29 15:28:18 UTC
Description of problem:
nfs analysis

Version-Release number of selected component:
wireshark-gnome-1.8.3-2.fc18

Additional info:
libreport version: 2.0.17
abrt_version:   2.0.17
backtrace_rating: 4
cmdline:        wireshark tcpdump_hanging.pcap
crash_function: __strlen_sse2_pminub
kernel:         3.6.3-3.fc18.x86_64

truncated backtrace:
:Thread no. 1 (10 frames)
: #0 __strlen_sse2_pminub at ../sysdeps/x86_64/multiarch/strlen-sse2-pminub.S:38
: #1 construct_match_selected_string at proto.c:6906
: #2 proto_can_match_selected at proto.c:7130
: #3 set_menus_for_selected_tree_row at main_menubar.c:5425
: #4 tree_view_selection_changed_cb at main.c:829
: #5 _g_closure_invoke_va at gclosure.c:840
: #8 _gtk_tree_selection_internal_select_node at gtktreeselection.c:1604
: #9 gtk_tree_view_real_set_cursor at gtktreeview.c:13208
: #10 gtk_tree_view_row_deleted at gtktreeview.c:9120
: #15 gtk_tree_model_row_deleted at gtktreemodel.c:1868

Comment 1 Michele Baldessari 2012-10-29 15:28:22 UTC
Created attachment 635053 [details]
File: core_backtrace

Comment 2 Michele Baldessari 2012-10-29 15:28:25 UTC
Created attachment 635054 [details]
File: environ

Comment 3 Michele Baldessari 2012-10-29 15:28:30 UTC
Created attachment 635055 [details]
File: limits

Comment 4 Michele Baldessari 2012-10-29 15:28:34 UTC
Created attachment 635056 [details]
File: backtrace

Comment 5 Michele Baldessari 2012-10-29 15:28:36 UTC
Created attachment 635057 [details]
File: cgroup

Comment 6 Michele Baldessari 2012-10-29 15:28:40 UTC
Created attachment 635058 [details]
File: smolt_data

Comment 7 Michele Baldessari 2012-10-29 15:28:45 UTC
Created attachment 635059 [details]
File: executable

Comment 8 Michele Baldessari 2012-10-29 15:28:51 UTC
Created attachment 635060 [details]
File: maps

Comment 9 Michele Baldessari 2012-10-29 15:28:54 UTC
Created attachment 635061 [details]
File: dso_list

Comment 10 Michele Baldessari 2012-10-29 15:28:57 UTC
Created attachment 635062 [details]
File: build_ids

Comment 11 Michele Baldessari 2012-10-29 15:28:59 UTC
Created attachment 635063 [details]
File: proc_pid_status

Comment 12 Michele Baldessari 2012-10-29 15:29:03 UTC
Created attachment 635064 [details]
File: var_log_messages

Comment 13 Michele Baldessari 2012-10-29 15:29:06 UTC
Created attachment 635065 [details]
File: open_fds

Comment 14 Niels de Vos 2012-11-14 10:59:09 UTC
Michele, do you have stepsand a capture to reproduce this?

I think the problem is similar to one that a user of my GlusterFS dissector mentioned:
- https://github.com/nixpanic/gluster-wireshark-1.4/issues/21

At the moment, I think the GTK3 implementation is to blame. With a gtk2 version of the package I can not reproduce the issue anymore. Care to test that out?

GTK2 packages are available from here:
- http://repos.fedorapeople.org/repos/devos/wireshark-gluster/fedora-18/x86_64/

Comment 15 Michele Baldessari 2012-11-14 14:17:53 UTC
Hi Niels ;)

nope I don't really have exact steps to reproduce yet. I will spend some more 
time into digging the exact issue. It seems to happen when I work with bigger
dumps but haven't pinpointed it yet.

I also think this is gtk3 related. I will try some patches that I saw flowing
upstream. Let's see if I get lucky

cheers,
Michele

Comment 16 Niels de Vos 2012-11-15 10:50:50 UTC
I am unable to reproduce the crash that was reported to me (see comment #14) with the svn version of wireshark from 20121114.

The RPMs (wireshark-1.8.3-2.fc18.0.svn20121114) are also available from my fedora-18 repository (http://repos.fedorapeople.org/repos/devos/wireshark-gluster/fedora-18/x86_64/).

I think it is likely that this issue has been fixed in upstream wireshark.

Comment 17 Peter Hatina 2012-12-03 09:44:09 UTC
Please, try wireshark-1.8.4-1.fc18, which is available in koji, soon in repositories.

Comment 18 Niels de Vos 2012-12-03 10:05:30 UTC
Created attachment 656518 [details]
"bt f" from a crashed wireshark-1.8.4-1.fc18.x86_64

Sorry, wireshark-1.8.4-1.fc18 from koji does not fix this.

Download the .pcap from https://github.com/nixpanic/gluster-wireshark-1.4/issues/21

1. wireshark /root/glusterfs-dissector-crash.pcap
2. Click on Packet No. 4 (V2 GETSPEC Call) and expand "GlusterFS Handshake" treeview in the payload pane.
3. Click on Packet No. 5 and wireshark will segfault.

(It may be that the GlusterFS protocols are not detected correctly, disabling the incorrectly used dissectors may be required, see https://github.com/nixpanic/gluster-wireshark-1.4/wiki/Disable-non-Gluster-protocols-in-Wireshark)

Comment 19 Jeff Layton 2012-12-20 13:04:10 UTC
Was looking at fields in a SMB capture.

backtrace_rating: 4
Package: wireshark-gnome-1.8.4-1.fc18
OS Release: Fedora release 18 (Spherical Cow)

Comment 20 Dan Ziemba 2013-01-30 07:07:27 UTC
While examining packet capture available at http://media.packetlife.net/media/blog/attachments/424/TCP_example.cap I clicked packet number 37 and wireshare crashed instantly.  

backtrace_rating: 4
Package: wireshark-gnome-1.8.4-1.fc18
OS Release: Fedora release 18 (Spherical Cow)

Comment 21 Paulo Fidalgo 2013-02-18 14:05:26 UTC
I was just capturing over my wireless interface. I had one virtualbox instance working with bride adapter and disable it.

backtrace_rating: 4
Package: wireshark-gnome-1.8.5-2.fc18
OS Release: Fedora release 18 (Spherical Cow)

Comment 22 Robert Hancock 2013-03-06 04:39:05 UTC
Clicked on packet containing a TLS message

backtrace_rating: 4
Package: wireshark-gnome-1.8.5-2.fc18
OS Release: Fedora release 18 (Spherical Cow)

Comment 23 Florian Weimer 2013-03-08 09:52:35 UTC
I was analyzing SSL network traffic.

backtrace_rating: 4
Package: wireshark-gnome-1.8.5-2.fc18
OS Release: Fedora release 18 (Spherical Cow)

Comment 24 Peter Hatina 2013-03-20 07:52:37 UTC
The crash was caused by broken gtk2 vs. gtk3 calls. Wireshark was reverted back to use gtk2 as a gui library. This one can't be reproduced anymore, closing.