Commit graph

98 commits

Author SHA1 Message Date
Seungha Yang
10ade466ef nvenc: Use CUDA stream of memory if exists
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:43 +00:00
Seungha Yang
7a8bb85523 cudaupload, cudadownload: Update for shared CUDA stream
Use CUDA stream of memory if exists

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:43 +00:00
Seungha Yang
aabcba16db nvdecoder: Skip synchronization if downstream buffer holds CUDA stream
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:42 +00:00
Seungha Yang
ddc5f1d425 cudaconvertscale: Add support for shared CUDA stream
If CUDA stream is shared by upstream/downstream, don't sync at
convert element

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:42 +00:00
Seungha Yang
1cb47d549b cudaconverter: Don't sync per conversion
Caller should take care of synchronization

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:42 +00:00
Seungha Yang
8d14194fdd nvencoder: Add support for shared CUDA stream
Sets CUDA stream on CUDA buffer pool so that CUDA stream object
can be shared

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:42 +00:00
Seungha Yang
0a81c8deb3 nvcodec: Port to GstCudaStream
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:42 +00:00
Seungha Yang
661b5f60c6 cuda: Provide single header include entry point
Add "gstcuda.h" header file

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3629>
2023-02-03 15:27:42 +00:00
Seungha Yang
bc0708eafb nvencoder: Fix min buffers parameter of allocation query in auto GPU mode
At the time when propose_allocation() get called, encoder session
would not be initialized yet.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3888>
2023-02-03 22:19:22 +09:00
Seungha Yang
29bd8de052 nvvp9dec: Fix return value
It should return GstFlowReturn value, not boolean

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3870>
2023-02-02 05:07:44 +09:00
Nirbheek Chauhan
eff9455ada nvcodec: Log readable errors when initializing CUDA
It is really difficult for people to figure out why nvcodec has
0 features. Even the debug log is cryptic. Also make sure the errors
go to the ERROR log level, which is more likely to be enabled by
default.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3776>
2023-01-25 10:08:10 +00:00
Nirbheek Chauhan
43ae6b421b nvcodec: Fix reporting of CuDeviceGetCount error
cuda_ret is was always going to be CUDA_SUCCESS in the error log.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3776>
2023-01-25 10:08:10 +00:00
Seungha Yang
040473f295 nvcudaenc: Don't use default CUDA stream
Set non-default CUDA stream via NvEncSetIOCudaStreams() if possible,
so that NVENC's internal kernel function can run on the given CUDA
stream instead of default CUDA stream

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3615>
2022-12-22 15:01:52 +00:00
Seungha Yang
15b2cd6565 cudabasetransform: Update CUDA stream on context update
CUDA stream must be associated with updated CUDA context

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3618>
2022-12-21 16:18:36 +00:00
Seungha Yang
14062c06a6 nvdecoder: Use own CUDA stream in GL output path
Use the same CUDA stream passed to CuvidMapVideoFrame()

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3611>
2022-12-20 17:14:10 +00:00
Seungha Yang
9914ff9b4c nvdec: Don't use default CUDA stream
NVDEC launches CUDA kernel function (ConvertNV12BLtoNV12 or so)
when CuvidMapVideoFrame() is called. Which seems to be
NVDEC's internal post-processing kernel function, maybe
to convert tiled YUV to linear YUV format or something similar.

A problem if we don't pass CUDA stream to the CuvidMapVideoFrame()
call is that the NVDEC's internel kernel function will use default CUDA stream.
Then lots of the other CUDA API calls will be blocked/serialized.

To avoid the unnecessary blocking, we should pass our own
CUDA stream object to the CuvidMapVideoFrame() call

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3605>
2022-12-19 20:27:35 +00:00
Seungha Yang
e6585c89ea cuda: Hide memory copy util function
The method was intended to be used by only cudaupload/download elements
and not ready to be a part of public API

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3545>
2022-12-08 19:52:23 +00:00
Seungha Yang
0e4c520cf3 nvdecoder: Handle input caps change
Update output caps if it's notified by baseclass

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3328>
2022-11-16 13:12:38 +00:00
Seungha Yang
b4c19a96b7 nvdecoder: Don't hold decoder output state
It's not referenced by this implementation

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3328>
2022-11-16 13:12:38 +00:00
Seungha Yang
9d30f9dd59 cudaconvertscale, cudascale: Add "add-borders" property and support 8bits RGB planar formats
Adding "add-borders" property which is identical to that of
videoscale and this will be enabled by default.
And adding RGBP/BGRP/GBR/GBRA format support.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3389>
2022-11-15 16:25:44 +00:00
Seungha Yang
c11f8fa930 cuda: Rewrite colorspace/rescale object
Rewriting GstCudaConverter object, since the old implementation was not
well organized and it's hard to add new features.
Moreover, the conversion operations were not very optimized.

Major change of this implementation:
* Remove redundant intermediate conversion operations such as
  any RGB -> ARGB(64) conversion or any YUV -> Y444 (or 16bits Y444).
  That's not required most of cases. The only required case is
  converting 24bits (such as RGB/BGR) packed format to 32bits format
  because CUDA texture object does not support sampling 24bits format
* Use normalized sample fetching (i.e., [0, 1] range float value)
  and also normalized coordinates system for CUDA texture.
  It's consistent with the other graphics APIs such as Direct3D
  and OpenGL, that makes sampling operations much easier.
* Support a kind of viewport and adopt math for colorspace conversion
  from GstD3D11 implementation

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3389>
2022-11-15 16:25:44 +00:00
Seungha Yang
c1efa9ac4b cudaupload,cudadownload: Add support for planar 8bits RGB formats
Defines RGBP, BGRP, and GBR formats, which have the same memory
layout as already supported Y444. And defines GBRA format as well
which needs just one additional alpha plane

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3389>
2022-11-15 16:25:44 +00:00
Seungha Yang
950dad201d cudaconvert, cudascale: Port to GstCudaBaseCovert baseclass
Don't need to hold duplicated code in the source tree

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3389>
2022-11-15 16:25:44 +00:00
Seungha Yang
eb2ef5a4fd cuda: Add convertscale element
GstCudaConverter object can do colorspace conversion and scale at once.
Adding new element "cudaconvertscale" to do that, this can
save unnecessary GPU operation if colorspace conversion and
rescale is required for given input stream format.

Most of codes are taken from d3d11convert element

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3389>
2022-11-15 16:25:44 +00:00
Seungha Yang
722f097b9d nvcodec: Update for documentation
* Use GST_PARAM_DOC_SHOW_DEFAULT flags for GPU ID related
  properties
* Fix some typos
* Add since markers

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3250>
2022-10-24 12:42:51 +00:00
Stéphane Cerveau
fb09c028e3 h265parse: fix typo in member of GstH265SPS
Rename sps_extnsion_params to sps_extension_params

Fix comment about vui_parameters_present_flag

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3072>
2022-10-04 10:01:12 +00:00
Seungha Yang
b6ca76eb68 nvh265sldec: Sync up with d3d11h265dec implementation
Each RefPicSetStCurrBefore/RefPicSetStCurrAfter/RefPicSetLtCurr array
might have empty element (i.e., reference pictures might not be stored
sequentially). Don't error out for the empty element case,
but instead, iterates each array and fill NVDEC's reference list
as much as possible

Fixes: https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/1441
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3039>
2022-09-16 20:25:23 +00:00
Seungha Yang
c91d72e677 nvdec: Fix for HEVC decoding when coded resolution is larger than display resolution
As documented in the SDK header, we should set coded width/height
values to the corresponding decoder configuration option,
instead of display resolution

Fixes: https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/1438
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3022>
2022-09-14 05:59:20 +09:00
Seungha Yang
ca47012a55 cudaupload,cudadownload: Fix document
* Fix typo, NVIDA -> NVIDIA
* Add cudadownload doc to the source file

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3018>
2022-09-13 16:40:51 +00:00
Seungha Yang
7ea3c55549 cudaupload,cudadownload: Use shared GstD3D11Device context if possible
Handle d3d11 device context in set_context() method with
additional device compatibility check so that only NVIDIA GPU
associated d3d11 device can be configured in the element.
And clear old d3d11 device per set_info() for d3d11 device to be
updated as well.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3018>
2022-09-13 16:40:51 +00:00
Thibault Saunier
6a4425e46a meson: Call pkgconfig.generate in the loop where we declare plugins dependencies
Removing some copy pasted code

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2970>
2022-09-01 21:17:35 +00:00
Seungha Yang
75e8f80999 cuda: Hide runtime compiler related header and symbols
That's already abstracted via gst_cuda_nvrtc_compile() method
and therefore, we do not need to expose such symbols yet.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2884>
2022-08-18 05:20:08 +09:00
Seungha Yang
812bb14f28 cudaupload,cudadownload: Don't simplify caps in transform_caps
The simplified caps might not be a subset of filter caps
and basetransform will complain about it.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2883>
2022-08-15 18:44:14 +00:00
Corentin Damman
0d5d94aabd nvcodec: fix caps leaks in nvh264/h265encoder
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2850>
2022-08-10 12:24:37 +00:00
Seungha Yang
6fff608f27 d3d11: Port to GstD3D11Fence
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2790>
2022-07-23 16:53:14 +00:00
Seungha Yang
30c8dbe032 nvencoder: Add support for dynamic GPU device selection
Adding nvautogpu{h264,h265}enc class which will accept upstream logical
GPU device object (GstCudaContext or GstD3D11Device) instead of
using pre-assigned GPU instance.

If upstream logical GPU device object is not NVENC compatible
(e.g., D3D11 device of non-NVIDIA GPU) or it's system memory,
then user specified "cuda-device-id" or "adapter-luid" property
will be used for GPU device selection.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2666>
2022-06-30 12:38:08 +00:00
Seungha Yang
ce82c0570e cuda: Add YUY2 and UYVY formats for GL interop
Those formats are not supported conversion formats by CUDA convert/scale elements
but would be useful for GL <-> CUDA interop use case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2691>
2022-06-30 11:17:15 +00:00
Seungha Yang
d9f08e17f1 cudadownload: Passthrough in case of CUDA -> SYSTEM
GstCudaMemory supports CPU access via CUDA pinned host memory already
and it would show faster memory transfer performance between
GPU and CPU than copying from/to normal system memory.
If downstream supports video meta, we can passthrough CUDA memory.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2690>
2022-06-30 10:45:45 +00:00
Corentin Damman
ae781610c8 cudaformat: add support for planar 4:2:2 YUV formats in CUDA D3D11 interop
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2269>
2022-06-30 07:30:09 +00:00
Corentin Damman
8678fc999a cudamemory: add support for planar 4:2:2 YUV formats
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2269>
2022-06-30 07:30:09 +00:00
Seungha Yang
85fe9fb61d nvcodec: Add cuda-device-id read-only property to stateless decoders
... and remove unnecessary intermediate subclass from class hierarchy

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2665>
2022-06-30 01:54:17 +09:00
Seungha Yang
1e11c05591 nvcodec: Add cuda-device-id read-only property to CUVID decoders
Similar to the other hardware decoder plugins, provides assigned
device id via property

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2665>
2022-06-30 01:45:20 +09:00
Seungha Yang
aca71fd692 d3d11: Update gst_d3d11_allocation_params_new() interface
Define GST_D3D11_ALLOCATION_FLAG_DEFAULT (0) value instead of
casting enum type everywhere. And pass D3D11_RESOURCE_MISC_FLAG value

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2646>
2022-06-27 19:33:57 +00:00
Seungha Yang
16955ba672 cudamemorycopy: Make sure writable caps before removing fields
The caps to be modified may not be writable when D3D11/GL/NVMM
are all disabled.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2664>
2022-06-27 20:28:31 +09:00
Seungha Yang
58c94eed84 nvh264dec,nvh265dec: Don't realloc bitstream buffer per slice
Allocated memory size has not been updated which results in
realloc per slice. Fixing it and also release bitstream buffer
on ::close(), not finalize.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2361>
2022-05-05 12:15:27 +00:00
Seungha Yang
33e77782c6 nvcodec: Add AV1 decoder
Adding GstCodecs based AV1 decoder element

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1453>
2022-05-04 01:23:25 +09:00
Seungha Yang
1444dd91c5 meson: nvcodec: Remove unnecessary override option and fix build with non-MSVC
cpp_std=c++11 was hack for macOS build but we don't build this plugin
for Apple device anymore. And add "-Wno-deprecated-declarations"
compile option for gcc/clang

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1453>
2022-05-04 01:10:07 +09:00
Seungha Yang
fe4d6bb469 nvcodec: Bump SDK header version to 11.1
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1453>
2022-05-03 23:56:34 +09:00
Seungha Yang
29aab904a8 nvdecoder: Handle DPB size margin in a single place
... instead of each subclass

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2343>
2022-05-03 14:17:49 +00:00
Seungha Yang
6a4c42c03f codecs: Signal required DPB size for AV1,MPEG2,VP8, and VP9 via new_sequence()
Make all codecs consistent so that subclass can know additional DPB
size requirement depending on render-delay configuration regardless
of codec. Note that render-delay feature is not implemented for AV1
yet but it's planned.

Also, consider new_sequence() is mandatory requirement, not optional

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2343>
2022-05-03 14:17:49 +00:00