If the vulkan plugin was compiled against a newer version than the supported
vulkan runtime instance or device, then it was possible for format retrieval to
fail. Failure was due to unconditionally using newer extensions and features
without runtime checking them.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8554>
The API version exposed by a particular device can be completely different from
what is exported by the parent instance. Since Vulkan 1.1 it is also possible
to use newer device API than supported by the instance API version (with the
appropriate version checks).
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8554>
Using #if instead of #ifdef was causing some issues when cross-compiling, like:
../ext/curl/gstcurlsmtpsink.c:54:5: error: "HAVE_SYS_SOCKET_H" is not
defined, evaluates to 0 [-Werror=undef]
54 | #if HAVE_SYS_SOCKET_H
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8589>
If a plugin gets disabled due to a `disabler()` dependency, the plugin
docs build itself will get disabled because `all_plugins_paths` will
become a disabler.
This was actually happening with opencv on systems that don't have
opencv available, and could happen with libsoup too if the build files
change in the future.
Let's avoid wasting hours of debugging for people. A not-found
dependency has the same effect.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8582>
The stream was keeping a reference to the sink, preventing
it to be removed properly by the pipleline bin.
Clean up and simplify the code to get the stream from the pad.
Add more mutex protection against add/remove requested pad.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/7916>
This change ensures that the `GST_EVENT_EOS`,
`GST_EVENT_FLUSH_START` and `GST_EVENT_FLUSH_STOP`
events are forwarded to the sink downstream.
The logging message for `GST_EVENT_FLUSH_START`
has also been fixed.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8576>
This patch refactors gst_wl_display_callback_destroy() to use the
recently introduced gst_wl_display_object_destroy() helper. Previously,
the function manually handled wl_callback destruction with explicit
lock/unlock calls and direct invocation of wl_callback_destroy().
Switching to gst_wl_display_object_destroy() unifies the destruction
process across similar objects, reducing code duplication and potential
errors.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8242>
when received xdg config event fron wayland server,
gst_wl_display_thread_run will call handle_xdg_surface_configure
which protected by priv->sync_mutex.
and in handle_xdg_surface_configure, configure_mutex also is locked
but if waylandsink set state from paused to ready, that will dispose
wlwindow, which will try to clear configure_mutex, and try to destroy
xdg_surface,
that do not proteced by anything.
so, problem is:
1) if clear configure_mutex(with locked state), clear lock will abort
2) after xdg_surface destroy, handle_xdg_surface_config may still call
ack_configure, that will lead wayland server go wrong
so, this patch updates gst_wl_window_finalize to use the new
destruction function for xdg_toplevel and xdg_surface, ensuring all
destruction operations are properly synchronized.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8242>
When the driver prefered format is not picked by downstream, the
decoders needs to select another format from the list. The selection
was currently unsorted, resulting in 10bit data often being stripped
to 8bit.
To solve this, reorder the formats in an HW preference order. This order
deviates slightly from the preferred order in libgstvideo. This is to
prefer bandwidth saving over better CPU alignment. As an example NV15 is
prefered over P010. We also prefer tiled over linear.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8522>
Add new property to support application allocated GstCudaMemory.
CUDA memory alloc/free is a global device synchronization point
as if launching CUDA kernel on default CUDA stream. To avoid the global
synchronization, we added stream-ordered allocation support
which allocates CUDA memory asynchronously.
However, NVENC does not allow registering the stream-ordered
allocated memory. Thus encoder was allocating normal CUDA
memory in case that input CUDA memory is stream-ordered type.
In this commit, newly introduced property will allow application
to provide encoder with GstCudaBufferPool. Application can
preallocate sufficient amount of CUDA memory in advance
to avoid global device synchronization while pipeline is running.
For now, this pool is used only if input CUDA memory is allocated
via stream-ordered-allocation
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8516>
The parent class will allow the handle to be reused at the end of the function.
If we are still modifying the released fence, then another thread can acquire
the fence while we are still clearing some of its data and produce a data race
or a leaked fence depending on which thread wins.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8491>
