After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 166073 - [threads] wrong error handling
[threads] wrong error handling
Status: RESOLVED FIXED
Product: GStreamer
Classification: Platform
Component: gstreamer (core)
git master
Other Linux
: Normal critical
: 0.8.10
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2005-02-02 19:10 UTC by Ronald Bultje
Modified: 2005-02-08 14:04 UTC
See Also:
GNOME target: ---
GNOME version: ---

Attachments
crappy fix (2.79 KB, patch)
2005-02-02 19:11 UTC, Ronald Bultje 		none Details | Review

Description Ronald Bultje 2005-02-02 19:10:40 UTC 
GstThread will not revert its state on a failed parent state change return. This
means that we lose thread references, which leads again to 100% CPU use. This
easily reproduced by { fakesrc ! fakesink }, where fakesrc emits an error on
NULL->READY, I can create a testcase if needed. Attached patch fixes this.
Comment 1 Ronald Bultje 2005-02-02 19:11:20 UTC 
Created attachment 36878 [details] [review]
crappy fix

This fix was coded in less than a minute and shows how to solve it. Someone
with more motivation may want to rewrite it to look better.
Comment 2 Ronald Bultje 2005-02-04 09:03:37 UTC 
Thomas, I want to nominate this for a showstopped because the patch is very
simple and straightforward, has little to no chance of breaking anything and
fixes a very clear and simple issue (that is a daily practice in cupid). I'd
appreciate this patch in 0.8.9.
Comment 3 Thomas Vander Stichele 2005-02-07 14:43:03 UTC 
ronald,

any patch where
- locks are involved
- gotos are used
- more than ten lines are touched
are neither simple nor straightforward.

I really don't know what to do with this.  If this is considered a real blocker,
we just have to start all over again, and we're screwing ourselves over again. 
and again.
Comment 4 Ronald Bultje 2005-02-07 14:46:36 UTC 
K, skipping to 0.8.10 then.
Comment 5 Ronald Bultje 2005-02-08 14:04:44 UTC 
Done.