Bug 488769 (Noran)

Summary:	Emesene Trunk Crashes because of python's lock in threads
Product:	[Fedora] Fedora	Reporter:	Franck Waechter <xeoslaenor>
Component:	python	Assignee:	James Antill <james.antill>
Status:	CLOSED NOTABUG	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	high	Docs Contact:
Priority:	low
Version:	10	CC:	ivazqueznet, james.antill, jonathansteffan, katzj, kylev
Target Milestone:	---
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2009-03-05 18:01:08 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:

Description Franck Waechter 2009-03-05 17:23:53 UTC

Description of problem:

Emesene Trunk Crashes because of python's lock in threads with message :

python: Python/pystate.c:612: PyGILState_Release: Assertion `oldstate == PyGILState_UNLOCKED' failed.
Aborted

it's working on others distributions


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

python-2.5.2-1.fc10.x86_64 and emesene trunk.
Note : same crash in rawhide fedora 11


How reproducible:

Go to www.emesene.org and get the last svn trunk. Try to login, it crashes !


Steps to Reproduce:
1. Go to www.emesene.org
2. Get the last revision of trunk version (see wiki page)
3. Try to login, it crashes !
  
Actual results:

emesene crashes !

Expected results:

emesene must login without python's errors

Additional info:

See http://forum.emesene.org/index.php/topic,1853.0.html

Comment 1 James Antill 2009-03-05 18:01:08 UTC

Which others?
How are the working vs. non-working python's compiled?
Are there any differences in the HW on the working vs. non-working machines?
What makes you suspect that it's a bug in our version of python?
Have you tried upstream python?

Realistically the Fedora python maintainers don't have the available time to debug other people's python code, sorry. So unless you have significant proof that it's our python at fault, then I'm going to assume otherwise.

Comment 2 Kyle VanderBeek 2009-05-28 22:41:41 UTC

This appears to be a emesene problem, not a Fedora one.  I poked around in the SVN trunk code, and PyGTK threads are initialized kind of late (by the Msnp object created on Controller.login().

The FAQ entry on PyGTK seems to indicate that you should do this before gtk.main():

http://faq.pygtk.org/index.py?req=show&file=faq20.001.htp