[Bug 733661] glimagesink navigation interface causes hangs with X11/GMainLoop event thread

[GStreamer (bugzilla.gnome.org)]

https://bugzilla.gnome.org/show_bug.cgi?id=733661
  GStreamer | gst-plugins-bad | git

Matthew Waters <ystreet00> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #284623|none                        |reviewed
             status|                            |

--- Comment #2 from Matthew Waters <ystreet00 at gmail.com> 2014-09-04 07:36:17 UTC ---
Review of attachment 284623:
 --> (https://bugzilla.gnome.org/review?bug=733661&attachment=284623)

Looks good, some comments below.

(In reply to comment #1)
> The navigation thread is currently part of GstGLWindow and not implemented in
> separate subclasses / backends yet. Perhaps this is needed?

It will be needed at some point :)

> Also commit 4dacc4ba introduced get_surface_dimensions method. The problem for
> the X11 backend, is that now this function is called from the navigation
> thread, resulting in xlib aborting due to multithreaded access:

For x11 it would make sense to store the window dimensions in the subclass and
return the cached values in get_surface_dimensions ().  The other backends will
most likely be fine with all of this.

If you have the GST_GL_XINITHTHREADS environment variable set, then
XInitThreads() will be called on plugin initialization.  However it would be
nice to avoid requiring XInitThreads in the basic gst-launch case.

-- 
Configure bugmail: https://bugzilla.gnome.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
You are the assignee for the bug.