[Bug 770953] gstclock: segmentation fault when unschedule

[GStreamer (GNOME Bugzilla)]

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

--- Comment #13 from Sebastian Dröge (slomo) <slomo at coaxion.net> ---
(In reply to GstBlub from comment #11)
> (In reply to Sebastian Dröge (slomo) from comment #10)
> > Review of attachment 335003 [details] [review] [review]:
> > 
> > ::: gst/gstclock.h
> > @@ -347,3 @@
> > - * Get the owner clock of the entry
> > - */
> > -#define GST_CLOCK_ENTRY_CLOCK(entry)    ((entry)->clock)
> > 
> > Don't remove public API
> 
> The problem with keeping it is that it would have to be change to obtain a
> strong reference, which means that the user of this API would have to unref
> it.  I would rather remove this broken API, no one should really be using it
> anyway.

Deprecate this one (and in #ifndef GST_DISABLE_DEPRECATED guards), and add a
function that gives a strong ref :)

> > 
> > @@ +558,3 @@
> >  gint                    gst_clock_id_compare_func       (gconstpointer id1,
> > gconstpointer id2);
> >  
> > +gboolean                gst_clock_id_is_clock           (GstClockID id,
> > GstClock * clock);
> > 
> > Maybe gst_clock_id_belongs_to_clock()? gst_clock_id_get_clock()?
> 
> How about gst_clock_id_uses_clock()?

Ack

> We could add gst_clock_id_get_clock() as a replacement for
> GST_CLOCK_ENTRY_CLOCK(), although the semantics would be slightly different.

Yes, add a new function for that and deprecate the macro

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