[Bug 763491] Typefind behavior has changed on have-type signal
GStreamer (GNOME Bugzilla)
bugzilla at gnome.org
Sat Mar 12 15:25:58 UTC 2016
https://bugzilla.gnome.org/show_bug.cgi?id=763491
--- Comment #13 from Sebastian Dröge (slomo) <slomo at coaxion.net> ---
For reference, IRC discussion / monologue:
<thaytan> slomo_, I don't understand what's happened there in decodebin
<thaytan> I thought the bug report was that typefind used to expose caps
before-hand, and this was just restoring that behaviour
<slomo_> thaytan: and then v added a change to decodebin that expected typefind
to give the caps later
<slomo_> thaytan: specifically his change expects that no element linked after
the typefind in decodebin will directly add sometimes pads when started inside
decodebin
<slomo_> thaytan: the call stack is type_found() taking the expose lock ->
analyze_new_pad() -> connect_pad() -> gst_pad_sticky_events_foreach() ->
pad_added_cb() taking the expose lock
<slomo_> thaytan: with typefind sending caps before type_found() (because then
the caps are part of the sticky events, otherwise not)
<slomo_> thaytan: imho the decodebin code is wrong. especially because the
locking is inconsistent and the above can also happen in other cases in theory
--
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.
More information about the gstreamer-bugs
mailing list