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 134583 - Race on gst_pad_link_free
Race on gst_pad_link_free
Status: RESOLVED FIXED
Product: GStreamer
Classification: Platform
Component: gstreamer (core)
git master
Other Linux
: Normal normal
: 0.7.5
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2004-02-16 23:59 UTC by Julien MOUTTE
Modified: 2004-12-22 21:47 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Julien MOUTTE 2004-02-16 23:59:23 UTC
When calling gst_pad_try_set_caps multilple times (user dragging the corner
of the player window) on a slow machine (my ppc laptop) i guess that a
second call to gst_pad_try_set_caps overlap a current pad unlinking and
somehow fucks up the links...

I have the following backtrace :

Program received signal SIGSEGV, Segmentation fault.

Thread 32771 (LWP 19635)

  • #0 g_type_init
    from /usr/lib/libgobject-2.0.so.0
  • #1 g_type_check_value
    from /usr/lib/libgobject-2.0.so.0
  • #2 gst_structure_free
    at gststructure.c line 245
  • #3 gst_caps_free
    at gstcaps.c line 268
  • #4 gst_pad_link_free
    at gstpad.c line 1088
  • #5 gst_pad_link_try
    at gstpad.c line 1323
  • #6 gst_ximagesink_renegotiate_size
    at ximagesink.c line 419
  • #7 gst_ximagesink_handle_xevents
    at ximagesink.c line 492
  • #8 gst_pad_push
    at gstpad.c line 3035
  • #9 gst_videoscale_chain
    at gstvideoscale.c line 387
  • #10 gst_opt_scheduler_chain_wrapper

You can clearly see here that the argument passed to gst_pad_link_free is
invalid. On a very fast machine i never manage to trigger that bug.
Comment 1 Benjamin Otte (Company) 2004-02-18 17:26:07 UTC
Are you sure this isn't/wasn't caused by some overlapping 
try_set_caps calls?
Comment 2 Julien MOUTTE 2004-02-18 18:07:24 UTC
Yeah i found the race already...

Closing the bug.