[Bug 770953] gstclock: segmentation fault when unschedule

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=770953

--- Comment #7 from GstBlub <gstblub at gmail.com> ---
(In reply to Sebastian Dröge (slomo) from comment #3)
> Comment on attachment 334928 [details] [review]
> gstclock: Keep weak reference to underlying clock
> 
> While this patch makes sense (except for the change in common/, please omit
> that :) ), it shouldn't be necessary according to the comments in the clock
> code.

Oops, that change was not intentional (switching branches too often I guess)

> Question is what and why the clock is gone before the clock ids are gone.
> This sounds like an element getting rid of the clock but not its clock ids.

Yeah, sounds like an element is doing something wrong, BUT at the same time the
Clock IDs should either ref the clock so it never goes away until all IDs are
gone, or they should at least hold a weak reference.  I chose the latter as
holding a strong reference would change behaviour.  Keeping a week reference
handles the incorrect use more gracefully, though at the expense of some
overhead.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.