Bug 373641 - Twinkle crash
Summary: Twinkle crash
Keywords:
Status: CLOSED INSUFFICIENT_DATA
Alias: None
Product: Fedora
Classification: Fedora
Component: twinkle
Version: rawhide
Hardware: All
OS: Linux
high
medium
Target Milestone: ---
Assignee: Kevin Fenzi
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2007-11-09 20:15 UTC by Marek Mahut
Modified: 2012-08-06 14:54 UTC (History)
3 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2008-02-21 07:04:28 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
backtrace from gdb (3.86 KB, text/plain)
2007-11-09 20:17 UTC, Marek Mahut
no flags Details
backstrace from gdb with ccrtp-debuginfo (4.84 KB, text/plain)
2007-11-28 12:13 UTC, Marek Mahut
no flags Details
twinkle.log (16.16 KB, text/plain)
2007-11-28 12:13 UTC, Marek Mahut
no flags Details

Description Marek Mahut 2007-11-09 20:15:08 UTC
Description of problem:

Twinkle crahes with SIGSEGV, after ending a call.

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

twinkle-1.1-3.fc8


How reproducible:

always

Steps to Reproduce:
1. call someone
2. press "end" button
3. twinkle crashes
  
Actual results:

crash of twinkle

Expected results:

twinkle ends the call without problem

Additional info:

Comment 1 Marek Mahut 2007-11-09 20:17:24 UTC
Created attachment 253341 [details]
backtrace from gdb

Comment 2 Marek Mahut 2007-11-09 20:29:42 UTC
Looks like a null dereference. 

Comment 3 Kevin Fenzi 2007-11-10 20:54:50 UTC
Odd. ok. I will take a look into it... 

i386? I see your gdb is el5... but this is on f8, right? 

Is there anything in your ~/.twinkle/twinkle.log that pertains to this? 

Comment 4 Marek Mahut 2007-11-11 17:37:47 UTC
Yes, it's i386. The gdb output is from el5, but the same behavior applies for f8
too.

No, nothing in twinkle.log.

Comment 5 Kevin Fenzi 2007-11-11 21:48:07 UTC
Looks like this might be something in the ccrtp libs. 

What version of ccrtp do you have installed?

can you install the ccrtp debuginfo and re-run gdb? 

Can you generate a twinkle.log with a failed session and attach it anyhow? 
(Upstream asks for that on all bug reports). 


Comment 6 Kevin Fenzi 2007-11-20 21:39:36 UTC
I take it you are only seeing this in devel, not F8? 

Can you try the twinkle from this scratch build (as soon as it's done)?

http://koji.fedoraproject.org/koji/taskinfo?taskID=251349



Comment 7 Marek Mahut 2007-11-28 12:10:11 UTC
Houps, sorry for late answer. I need to reorganize my bugzilla mails I think.

My ccrtp version is ccrtp-1.5.1-1.

As the scratch build does not preserve the source rpm I can't try it. I'm using
this twinkle package under RHEL, recompiled from Fedora.

Comment 8 Marek Mahut 2007-11-28 12:13:10 UTC
Created attachment 271191 [details]
backstrace from gdb with ccrtp-debuginfo

Comment 9 Marek Mahut 2007-11-28 12:13:47 UTC
Created attachment 271211 [details]
twinkle.log

Comment 10 Andreas Thienemann 2007-12-05 17:11:50 UTC
Comparing straces and an educated guess shows that switching from ALSA to OSS
seems to fix the hang on clicking BYE. This might be heavily related to this
problem.

Comment 11 Marek Mahut 2007-12-05 17:29:23 UTC
True, this solves the problem. I just found another user having same problem.

Comment 12 Andreas Thienemann 2007-12-06 07:40:45 UTC
Marek,

(In reply to comment #11)
> True, this solves the problem. I just found another user having same problem.

I'm not seeing this problem myself. What Headset are you using when calling with
twinkle? Is there any correlation? I'm on a Logitech headset with USB adapter.

regards,
 andreas

Comment 13 Marek Mahut 2007-12-06 07:49:11 UTC
Andreas,

no headset, using my build-in sound card in T60.

Comment 14 Andreas Thienemann 2007-12-06 09:14:45 UTC
Marek,

what about the others? Internal card as well?

Comment 15 Marek Mahut 2007-12-30 20:39:47 UTC
Yes, however after I've reinstalled my computer I don't see this error. Andreas
can you please work with Igor Donev to get this resolved?

Comment 16 Kevin Fenzi 2008-01-02 01:16:24 UTC
You no longer see the issue? Do we have someone who can duplicate it? 
I still don't see it here at all...

Comment 17 Kevin Fenzi 2008-02-21 03:19:05 UTC
Any further news here? Or should we just close this out? 


Comment 18 Marek Mahut 2008-02-21 07:04:28 UTC
(In reply to comment #17)
> Any further news here? Or should we just close this out? 

Kevin, closing. Thank you for your time, this bug report itself has not been
very informative from my side. If I see the issue again I'll make sure to
troubleshoot it and get proper backtrace as well.


Note You need to log in before you can comment on or make changes to this bug.