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 752754 - validate: valgrind errors
validate: valgrind errors
Status: RESOLVED FIXED
Product: GStreamer
Classification: Platform
Component: gst-devtools
git master
Other Linux
: Normal normal
: 1.7.1
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2015-07-23 04:36 UTC by Vineeth
Modified: 2015-10-02 15:48 UTC
See Also:
GNOME target: ---
GNOME version: ---

Attachments
set errors-for-leak-kinds as definite (1.09 KB, patch)
2015-07-23 04:42 UTC, Vineeth 		none Details | Review
fix memory leak in override registry (1001 bytes, patch)
2015-07-23 04:42 UTC, Vineeth 		committed Details | Review
set errors-for-leak-kinds as definite (1.26 KB, patch)
2015-07-26 23:48 UTC, Vineeth 		committed Details | Review

Description Vineeth 2015-07-23 04:36:47 UTC 
1) errors-for-leak-kinds should be set to definite, because almost every test case, will have possibly lost memory, which may or may not be a leak. And throwing error for all these cases doesn't seem to be correct.

2) memory leak in override-registry.c
Comment 1 Vineeth 2015-07-23 04:42:15 UTC 
Created attachment 307958 [details] [review]
set errors-for-leak-kinds as definite
Comment 2 Vineeth 2015-07-23 04:42:48 UTC 
Created attachment 307959 [details] [review]
fix memory leak in override registry
Comment 3 Thibault Saunier 2015-07-23 08:41:14 UTC 
Review of attachment 307959 [details] [review]:

OK
Comment 4 Thibault Saunier 2015-07-23 08:46:18 UTC 
Review of attachment 307958 [details] [review]:

The idea until now is that they should be fixed or added to the suppression file if they were false positive. I guess it is OK to just ignore them for now though (no strong opinion about that).
Comment 5 Guillaume Desmottes 2015-07-24 13:30:43 UTC 
Review of attachment 307958 [details] [review]:

I'm fine adding this for now but please add a TODO saying it would be good to get rid of it using suppressions at some point.
Comment 6 Thibault Saunier 2015-07-24 14:05:39 UTC 
Review of attachment 307958 [details] [review]:

Or maybe we should make that list configurable by the testsuite?
Comment 7 Thibault Saunier 2015-07-24 14:06:27 UTC 
Review of attachment 307958 [details] [review]:

Marking as needs-work as per Guillaume's comment
Comment 8 Vineeth 2015-07-26 23:48:10 UTC 
Created attachment 308191 [details] [review]
set errors-for-leak-kinds as definite

Adding a TODO to make sure the false positives are added to suppression files
Comment 9 Nicolas Dufresne (ndufresne) 2015-08-05 21:23:03 UTC 
Comment on attachment 307959 [details] [review]
fix memory leak in override registry

commit 185b3b2d7e01d3cb0189fedb0fe7b54f764fb558
Author: Vineeth TM <vineeth.tm@samsung.com>
Date:   Thu Jul 23 13:35:04 2015 +0900

    validate:override-registry: fix memory leak
    
    mutex is being initialized but not cleared.
    
    https://bugzilla.gnome.org/show_bug.cgi?id=752754
Comment 10 Thibault Saunier 2015-10-02 15:48:02 UTC 
commit aa2c93c3d4f1663a0efe8c421fc95c0134092943
Author: Vineeth TM <vineeth.tm@samsung.com>
Date:   Mon Jul 27 08:46:01 2015 +0900

    validate:launcher: throw valgrind error only for definite loss
    
    errors-for-leak-kinds should be set to definite, because almost every test case
    , will have possibly lost memory, which may or may not be a leak.
    And throwing error for all these cases doesn't seem to be correct.
    
    https://bugzilla.gnome.org/show_bug.cgi?id=752754