Age | Commit message (Collapse) | Author |
|
|
|
Git will de-duplicate the contents for us anyway, and
symlinks can cause problems with some versions of git
and also on Windows.
https://github.com/mesonbuild/meson/issues/11646
https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4326
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1161>
|
|
|
|
|
|
The first time might return `EAGAIN` if there are pending frames but
there is no decoded frame available yet. The second time it will
actually wait for frames to become available and only start returning
`EAGAIN` again once no more frames are left.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1086>
|
|
|
|
Instead use the videodecoder error handling to allow up to max-errors
consecutive decoding errors, i.e. infinite by default in 1.22 and newer.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1058>
|
|
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1056>
|
|
After flushing there are no frames left anymore that could be dropped.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1056>
|
|
This directly discards all frames and it won't be possible to output
them anymore.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1056>
|
|
decoder requires more before accepting new data
This works around a race condition in dav1d where the decoder deadlocks
if multiple threads are used, and also is generally beneficial as it
allows for proper frame threading.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1056>
|
|
|
|
|
|
|
|
|
|
|
|
The rank of AOM av1dec was demoted as secondary, and thus
primary rank is sufficient.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/1008>
|
|
|
|
|
|
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/977>
|
|
|
|
|
|
It's doing the same thing and is shorter.
|
|
Fixes https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/issues/198
|
|
Fixes https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/issues/238
|
|
See https://gitlab.freedesktop.org/gstreamer/gstreamer-rs/-/merge_requests/1128
|
|
|
|
0.9.0 is the next release, so we can't name things that already.
Also the version in meson.build was 0.13.0, which is completely wrong.
|
|
|
|
|
|
|
|
Which will automatically be integrated in gstreamer documentation
|
|
|
|
Simplify caps creation codes
|
|
|
|
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
This is needed because we may now access state concurrently with and
outside of the streaming thread during a LATENCY query.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
We need a successful LATENCY query to be able to correctly initialise
max frame delay, so let's defer decoder init until we know upstream is
readier for it.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
This is not strictly correct as we might not use the same mechanism that
dav1d is using to detect the number of CPUs. We can drop the num_cpus
based approach once
https://code.videolan.org/videolan/dav1d/-/merge_requests/1407 is merged
upstream.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs/-/merge_requests/698>
|
|
|
|
|
|
This is now automatically implemented.
|
|
|
|
See the details:
https://gitlab.freedesktop.org/gstreamer/gstreamer-rs/-/merge_requests/980
|
|
It's only accessed from the streaming thread.
|
|
|