Commit graph

129 commits

Author SHA1 Message Date
Seungha Yang
9518f15309 d3d11upload,d3d11download: Use staging buffer pool
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1643>
2022-02-06 21:58:48 +09:00
Seungha Yang
b22aa14e43 d3d11decoder: Use staging buffer pool for non-d3d11 downstream
In specific condition, this can show about 10% ~ 30%
performance gain with non-d3d11 downstream element.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1643>
2022-02-06 21:57:17 +09:00
Seungha Yang
ea06439c39 d3d11decoder: Fix typo in doc
s/elemenet/element/g

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1598>
2022-01-29 04:47:16 +09:00
Seungha Yang
7b1683f6b2 d3d11av1dec: Fix typo in debug message
Fixing copy and paste mistake, It's AV1 decoder not VP8

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1576>
2022-01-27 03:11:33 +09:00
Seungha Yang
f97ab9edf7 d3d11deinterlace: Do not restrict minimum resolution to 64x64
The value 64 was completely arbitrary one, and this element
will be able to support smaller resolutions

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1528>
2022-01-16 02:23:55 +09:00
Seungha Yang
e63a3a7f31 d3d11compositor: Don't try to read empty buffer
The queued buffer may not be readable buffer in case that
upstream sends GAP event or so.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1497>
2022-01-13 20:27:50 +00:00
Seungha Yang
5fefc689a0 d3d11decoder: Negotiate again on the first output buffer
... unconditionally. There may be updated field in sinkpad caps
after the new_sequence() call (HDR related ones for example),
then we should signal the information to downstream.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1474>
2021-12-30 11:09:45 +00:00
Seungha Yang
91484ce2ac d3d11av1dec: Update sinkpad template for profile
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1456>
2021-12-21 22:20:34 +09:00
Seungha Yang
1564567c3e d3d11av1dec: Fix for Cdef param
av1parser will increase the sec_strength values by 1 if parsed
values were equal to 3 as defined in spec. But DXVA wants unmodified
ones.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1455>
2021-12-17 23:28:29 +09:00
Seungha Yang
146815bbb6 d3d11av1dec: Sync DXVA AV1 data structure with released header
Update AV1 data structure based on Windows 11 SDK header

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1455>
2021-12-17 23:27:24 +09:00
Seungha Yang
f10867dfb5 d3d11videosink: Use only tested color space for swapchain
We are querying supported swapchain colorspace via
CheckColorSpaceSupport() but it doesn't seem to be reliable.
Use only tested full-range RGB formats which are:
- sRGB
- BT709 primaries with linear RGB
- BT2020 primaries with PQ gamma

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1433>
2021-12-12 11:00:24 +00:00
Seungha Yang
e8c923ebde d3d11window: Remove hack related to color space selection
Use input GstVideoColorPrimaries without any special case handling
otherwise rendered image color would be very wrong.

The hack was added to work around an issue that some Intel driver
couldn't handle wide color gamut image without HDR10 metadata, specifically PQ image.
But device capability can be checked via a method added in
https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/merge_requests/1723
so there's no issue now.

Fixes: https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/1175
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1430>
2021-12-09 20:12:12 +09:00
Seungha Yang
1a12883993 d3d11window: Fix typo in debug message
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1430>
2021-12-09 19:51:20 +09:00
Seungha Yang
e87791592e d3d11vp9dec: Drop frames on non-keyframe format change
... in case of NVIDIA GPU

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1407>
2021-12-02 19:19:59 +00:00
Seungha Yang
7fe08d8bed d3d11: Stop doc for non-default decoder and deinterlacer elements
Just skip doc for non-default decoder/deinterlacer elements
since there are multiple element in case that system has
multiple GPUs.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1360>
2021-11-19 13:38:19 +00:00
Wonchul Lee
ae7a365274 d3d11: Fix deadlock while doing unprepare
ShowWindow() could be blocked while doing gst_d3d11_window_win32_unprepare
when external window handle provided to d3d11videosink in multi-threaded
environment.
The condition that issue happened is, UI thread is waiting for a
background thread that changes d3d11videosink state to NULL, and the
background thread would try to send a window message to the queue.
The queue is already occupied by the UI thread, so the background
thread will be blocked.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1366>
2021-11-19 13:42:53 +09:00
Seungha Yang
be9cf6c9cd d3d11: d3d11{convert,scale}: Add add-borders property
Functionally identical to that of videoscale element.

When disabled, d3d11convert or d3d11scale element will scale
image without adding borders, meaning that display aspect ratio
will not be preserved.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1170>
2021-10-17 12:36:28 +00:00
Seungha Yang
b3710aa68d d3d11: Rename screen capture element
Old name "desktopdup" may confuse users. Now it's renamed to
"screencapture"

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1136>
2021-10-14 15:15:56 +00:00
Seungha Yang
4b85704b88 d3d11: Add device provider for d3d11desktopdupsrc
... and add support for multi-GPU/multi-monitor

By using newly added "monitor-handle" property, user can specify a
monitor to be captured via HMONITOR handle.

Fixes: https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/1673
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1103>
2021-10-11 15:21:26 +00:00
Seungha Yang
5707e487cf d3d11: d3d11decoder: Use DXGI adapter LUID
... instead of index of DXGI adapter.

The order of IDXGIAdapter1 enumerated via IDXGIFactory1::EnumAdapters1
can be varying even there's no rebooting in case that GPU preference order
is updated by user (for example, it can be done by using NVIDIA Control Panel
in case of multi-GPU laptop system) and eGPU is another possible case.

So, for an element which requires fixed target GPU requirement,
index based device enumeration is unreliable.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1098>
2021-10-10 03:23:58 +09:00
Seungha Yang
d8eff2623d d3d11: d3d11device: Remove "allow-tearing" property
Plugin can query DXGI_FEATURE_PRESENT_ALLOW_TEARING without d3d11device
help

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1098>
2021-10-10 01:01:44 +09:00
Seungha Yang
7286c8e302 d3d11: d3d11desktopdupsrc: Add support for non-D3D11 downstream element
By this commit, application doesn't need to configure d3d11download
element for software pipeline which will make things simpler

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1082>
2021-10-08 16:13:46 +09:00
Seungha Yang
7c557c2d65 codecs: mpeg2decoder: Use GstFlowReturn everywhere
boolean return value is not sufficient for representing the reason
of error in most cases. For instance, any errors around new_sequence()
would mean negotiation error, not just *ERROR*.
And some subclasses will allocate buffer/memory/surface on new_picture()
but it could be failed because of expected error, likely flushing

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1019>
2021-10-04 20:56:46 +00:00
Seungha Yang
e322745763 codecs: h264decoder: Use GstFlowReturn everywhere
boolean return value is not sufficient for representing the reason
of error in most cases. For instance, any errors around new_sequence()
would mean negotiation error, not just *ERROR*.
And some subclasses will allocate buffer/memory/surface on new_picture()
but it could be failed because of expected error, likely flushing

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1019>
2021-10-04 20:56:46 +00:00
Seungha Yang
5b405d1585 codecs: h265decoder: Use GstFlowReturn everywhere
boolean return value is not sufficient for representing the reason
of error in most cases. For instance, any errors around new_sequence()
would mean negotiation error, not just *ERROR*.
And some subclasses will allocate buffer/memory/surface on new_picture()
but it could be failed because of expected error, likely flushing

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1019>
2021-10-04 20:56:46 +00:00
Seungha Yang
6a756b9f46 d3d11: d3d11vp9dec: Fix use_prev_in_find_mv_refs value setting
"last_show_frame" should be updated based on
GstVp9FrameHeader::show_frame, not show_existing_frame

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1021>
2021-10-03 02:25:50 +09:00
Seungha Yang
052983014e d3d11videosink: Add support for crop meta
... when upstream element is d3d11.

Note that, if upstream element is not d3d11, crop meta is almost
pointless since d3d11videosink will upload video frame to GPU memory
in any case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/933>
2021-09-27 16:29:09 +00:00
Seungha Yang
0fa5a06ed9 d3d11videosink: Perform propose_allocation() even when we have no configured window
In order to support d3d11 device update, d3d11videosink will configure
window on the first buffer. So, there might not be configured
window when propose_allocation() is required.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/933>
2021-09-27 16:29:09 +00:00
Thibault Saunier
019971a3c7 Move files from gst-plugins-bad into the "subprojects/gst-plugins-bad/" subdir 2021-09-24 16:14:36 -03:00