Hi All,
Please find below a brief summary of the progress Collabora have made last week on Weston/Wayland.
Weston / Wayland
Marius:
- Spent most of my time debugging and investigating the
cluster-receiver application
(SPEC-3382); the receiver application is now able to use
waylandsink. Both Qt and
Gstreamer make use of independent threads to receive and handle
events. While I thought
initially this was a problem related to the way the parent surface
(the one created by QtWayland)
is not updated/committed in sync with the one created by
Gstreamer, it seems that there is a slight
race when QtWayland actually creates and attaches the buffer to
the parent surface. Gstreamer
is using that surface to create its own sub-surface to render the
incoming stream. Temporary solution
for the moment is to install a timer such that Qt is able to do
all its things underneath and
then start the gstreamer pipeline. In this way when gstramer
creates it's sub-subsurface
it does a desync commit on the parent surface.
- On the same item, the receiver requires some definitions found
in waylandsink. Used to be exported
with autotools, but now Gstreamer switch to meson, adapted and
push a MR for that.
- Found, fixed and pushed a corner case issue for the compisitor
related to pending surfaces (SPEC-3448)
-
Things currently looking into:
- The cluster-receiver changes are targeted at waylandsink but on
other systems (x86_64) the vaapi
sink is in use. Need to determine how my changes work out with
that and if that is a real use case.
Kind regards,
Nick
Links:
You receive all messages sent to this group.
View/Reply Online (#8472) |
Reply To Group
| Reply To Sender
|
Mute This Topic
| New Topic
Your Subscription |
Contact Group Owner |
Unsubscribe
[list-automotive-discussions82@xxxxxxxxxxx]
