Bug 669849

Summary: [abrt] python-2.7-8.fc14.1: malloc_consolidate: Process /usr/bin/python2.7 was killed by signal 11 (SIGSEGV)
Product: [Fedora] Fedora Reporter: Robert Park <rbpark>
Component: pythonAssignee: Dave Malcolm <dmalcolm>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: medium Docs Contact:
Priority: low    
Version: 14CC: dmalcolm, ivazqueznet, james.antill, jonathansteffan
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Unspecified   
Whiteboard: abrt_hash:924c7ba946aadbd3d0295c22180fb4b7e941f414
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-01-17 17:41:18 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
File: backtrace none

Description Robert Park 2011-01-15 01:58:31 UTC
abrt version: 1.1.14
architecture: x86_64
Attached file: backtrace
cmdline: /usr/bin/python2.7 ./gottengeography
component: python
crash_function: malloc_consolidate
executable: /usr/bin/python2.7
kernel: 2.6.35.10-74.fc14.x86_64
package: python-2.7-8.fc14.1
rating: 4
reason: Process /usr/bin/python2.7 was killed by signal 11 (SIGSEGV)
release: Fedora release 14 (Laughlin)
How to reproduce: I closed a PyGobject/introspection based python app. 
time: 1295056633
uid: 500

Comment 1 Robert Park 2011-01-15 01:58:33 UTC
Created attachment 473621 [details]
File: backtrace

Comment 2 Robert Park 2011-01-15 03:50:15 UTC
Package: python-2.7-8.fc14.1
Architecture: x86_64
OS Release: Fedora release 14 (Laughlin)


How to reproduce
-----
something something pygobject introspection blah blah

Comment 3 Dave Malcolm 2011-01-17 17:41:18 UTC
Thanks for filing this bug report.

Looking at the backtrace, it looks like something within the process has corrupted the heap.  This leads to a crash, in thread #0, when pulseaudio attempts to allocate memory (I don't think it's puleaudio's fault; I think the heap has been trashed by something else).

Looking at the memory map in attachment 473621 [details], I see numerous libraries below /opt/gnome/ are mapped into memory.

It looks like you're running a non-standard build of the gnome libraries, perhaps you built these locally?

As is, this is likely to be a bug in one of those libraries.

This kind of thing is hard to track down.

I'd recommend using valgrind to figure things out further.  

Given that this problem is most likely in a non-Fedora build of Gnome, I'm going to close this bug out as NOTABUG (on the grounds that the bug is not in Fedora).

Hope that makes sense.

Comment 4 Robert Park 2011-01-17 21:01:57 UTC
Yeah, I am running GNOME from a jhbuild environment, sorry for the noise.

I'm developing a simple python script that should not be doing anything terribly difficult or unusual, but I'm doing it with all the latest git pygobject and gnome-instrospection toys. Do you think this report would be valuable upstream, or do you think this would be a fairly obvious problem for them to fix on their own without my report?

Thanks.

Comment 5 Dave Malcolm 2011-01-17 21:32:45 UTC
This seems like a bug for upstream, but I don't think it's going to be very useful to them without more information.

My recommendation is that you should try to run python under valgrind when replicating this.
  valgrind /usr/bin/python PATH_TO_SCRIPT

Hopefully that should show the error (along with quite a bit of other noise, alas).