Bug 208185

Summary: fc6t3 virt-manager segfaults on launch
Product: [Fedora] Fedora Reporter: das_deniz
Component: virt-managerAssignee: Daniel Berrangé <berrange>
Status: CLOSED RAWHIDE QA Contact: Brian Brock <bbrock>
Severity: medium Docs Contact:
Priority: medium    
Version: 6CC: steved
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2006-10-06 14:53:15 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
requested backtrace attached
none
similar backtrace for newer version. none

Description das_deniz 2006-09-26 20:53:11 UTC
Description of problem:

virtual manager segfaults on launch

Version-Release number of selected component (if applicable):

virt-manager-0.2.2-1.fc6

How reproducible:

always

Steps to Reproduce:
1. install fc6t3 + xen + updates for Tue Sep 26 15:00 EDT
2. Applications -> System Tools -> Virtual Machine Manager
3. or try virt-manager in terminal as root
  
Actual results:

# virt-manager 
dbus_bindings.DBusException Unable to determine the address of the message bus
(try 'man dbus-launch' and 'man dbus-daemon' for help)
Could not connection to session bus, disabling DBus service
/usr/share/virt-manager/virtManager/manager.py:486: Warning: value
"((GValueArray*) 0x94c9e68)" of type `GValueArray' is invalid or out of range
for property `data-array' of type `GValueArray'
  cell.set_property('data_array', data)
/usr/share/virt-manager/virtManager/manager.py:486: Warning: value
"((GValueArray*) 0x94cd0d8)" of type `GValueArray' is invalid or out of range
for property `data-array' of type `GValueArray'
  cell.set_property('data_array', data)
Segmentation fault

Expected results:

python gnome window to create/manage virtual machines... 

Additional info:

host is 2.6.18-1.2693.fc6xen #1 SMP Fri Sep 22 18:38:57 EDT 2006 i686 i686 i386
GNU/Linux

no current vms created - had created one before updates - but removed it...

also did 'mv /var/lib/xen /local'
and 'ln -s /local/xen /var/lib/xen'
to get xen vm filesystem to partition with room.

Comment 1 Daniel Berrangé 2006-09-26 21:01:48 UTC
I can't reproduce this problem here anymore - the two segfaults i knew of are
both solved in latest gnome-vfs & gtk RPMs from about a week ago. Could you try
and capture a stack trace for the crash. You'll need to do something like this:

 # yum --enablerepo=development-debuginfo  install python-debuginfo,
gtk2-debuginfo & virt-manager-debuginfo  
 # ulimit -c unlimited
 # virt-manager
 # gdb /usr/bin/python core.XYZ


Comment 2 das_deniz 2006-09-27 17:43:49 UTC
Created attachment 137232 [details]
requested backtrace attached

fc6t3 + all updates till sep 27th 13:00 EDT

Comment 3 Daniel Berrangé 2006-09-27 20:09:00 UTC
Thanks for the stack trace. I'm investigating the problem & have a couple of
leads which look promsing....

Comment 4 Daniel Berrangé 2006-09-28 21:14:38 UTC
Committed two fixes to upstream repo for this issue

http://hg.et.redhat.com/virt/applications/virt-manager--devel?cs=bf53891faa22
http://hg.et.redhat.com/virt/applications/virt-manager--devel?cs=1cb01dfe5d34

Will add patches these to next RAWHIDE update of virt-manager.

Comment 5 das_deniz 2006-09-29 22:01:15 UTC
i was going to write that not only does it no longer segfault - it now just
exits silently - but that is because i wasn't running xen on my host... (which
would be a nice message after clicking "Connect")

still - i now have virt-manager-0.2.3-1.fc6 - and it still segfaults. will
attach backtrace.

Comment 6 das_deniz 2006-09-29 22:02:18 UTC
Created attachment 137448 [details]
similar backtrace for newer version.

Comment 7 Daniel Berrangé 2006-09-29 22:14:19 UTC
The fix was not in version 0.2.3-1.fc6, but I have just built a version
0.2.3-2.fc6 which will have the neccessary fix. The reason it is a sporadic
failure is because it is based on a subtle timing issue which causes the
calculated CPU utilization to go a little over 100%, which break sthe sparkline
validation, in turn resulting in a null data set which then makes the rendering
code SEGVs. Anyway, check out 0.2.3-2.fc6 once it hits rawhide (hopefully tomorrow).


Comment 8 Steve Dickson 2006-10-02 18:47:37 UTC
*** Bug 208916 has been marked as a duplicate of this bug. ***

Comment 9 Daniel Berrangé 2006-10-06 14:53:15 UTC
Closing on the basis that bug 208916 showing the same problem has reported  
0.2.3-2.fc6  fixes the issue.


Comment 10 das_deniz 2006-10-06 16:23:55 UTC
yep sorry about that - confirmed fix with the release mentioned. thank you.