Commit graph

25 commits

Author SHA1 Message Date
Seungha Yang
afb62e98c7 cuda: Enable x86 NVMM support again
It was broken since memory copy helper function was moved to gst-libs.
Also, adding "cuda-nvmm" and "cuda-nvmm-include-path" build options
to en/disable NVMM support in gstcuda library

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6978>
2024-06-06 12:16:50 +00:00
Seungha Yang
15c24abf5d nvcodec: Add AV1 encoder
Adding CUDA mode "nvav1enc", D3D11 mode "nvd3d11av1enc" and auto GPU
mode "nvautogpuav1enc" elements

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6754>
2024-05-15 14:35:52 +00:00
Seungha Yang
b74422dcbc nvcodec: Remove old nvenc implementation
Stop shipping deprecated implementation

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6754>
2024-05-15 14:35:52 +00:00
Seungha Yang
6dc902a962 nvcodec: Add JPEG encoder
Adding nvJPEG library based JPEG encoder element

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6021>
2024-01-31 12:51:28 +00:00
Seungha Yang
fc4b136ac8 nvcodec: Move runtime compiler check to plugin init function
Required information for other elements

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6021>
2024-01-31 12:51:28 +00:00
Seungha Yang
7b6023d9cf nvcodec: Add support for CUDA IPC
Adding cudaipc{src,sink} element for CUDA IPC support.

Implementation note:
* For the communication between end points, Win32 named-pipe
and unix domain socket will be used on Windows and Linux respectively.

* cudaipcsink behaves as a server, and all GPU resources will be owned by
the server process and exported for other processes, then cudaipcsrc
(client) will import each exported handle.

* User can select IPC mode via "ipc-mode" property of cudaipcsink.
There are two IPC mode, one is "legacy" which uses legacy CUDA IPC
method and the other is "mmap" which uses CUDA virtual memory API
with OS's resource handle sharing method such as DuplicateHandle()
on Windows. The "mmap" mode might be better than "legacy" in terms
of stability since it relies on OS's resource management but
it would consume more GPU memory than "legacy" mode.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4510>
2023-08-14 13:41:01 +00:00
Seungha Yang
59f359eb99 cuda: Rename macro HAVE_NVCODEC_GST_GL -> HAVE_CUDA_GST_GL
... and always use #ifdef instead of #if

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3992>
2023-02-20 01:49:31 +09:00
Seungha Yang
eb0fca4180 nvencoder: Reuse input resource
Call input resource map functions (i.e., nvEncRegisterResource,
nvEncUnregisterResource, nvEncMapInputResource, and
nvEncUnmapInputResource) only once and reuse the mapped resources,
instead of per input frame map/unmap

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3884>
2023-02-16 17:49:54 +00:00
Seungha Yang
03425bc702 nvdecoder: Add support for CUDA zero-copy in stateless decoder
Wrap mapped decoder output surface using GstCudaMemory and
output without any copy operation. Also, for application to be able to
control the number of zero-copyable output surfaces,
"num-output-surfaces" property is added.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3884>
2023-02-16 17:49:54 +00:00
Seungha Yang
a94af552f5 nvdecoder: Port to C++
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3884>
2023-02-16 17:49:54 +00:00
Seungha Yang
992406cf4f cuda, nvcodec: Make GstD3D11 dependency mandatory
GstD3D11 build-time dependencies should be always available on Windows already
and runtime dependencies as well, since required external
(non-GStreamer) depends are all system DLLs

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3884>
2023-02-16 17:49:54 +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
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
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
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
205a23f456 nvcodec: Stop building for non-{Windows,Linux} target
The other platforms, specifically macOS have not been supported already
because this plugin loads so or dll. Moreover, NVIDIA dropped
support for macOS as of CUDA 11.0. See also
https://developer.nvidia.com/nvidia-cuda-toolkit-11_0_0-developer-tools-mac-hosts

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2283>
2022-04-25 15:50:52 +00:00
Thibault Saunier
c033f8fad2 cuda: Factor out a public GstCUDA library
So applications and elements implemented outside GStreamer can reuse
our infrastructure

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1472>
2022-04-19 16:39:43 +00:00
Seungha Yang
60735deded nvcodec: Add new Direct3D11/CUDA mode encoder implementation
Adding new encoder elements nvd3d11{h264,h265}enc for Direct3D11
input support and re-written nvcuda{h264,h265}enc elements.
Newly writeen elements have some differences compared with old
nv{h264,h265}enc including non-backward compatible changes.

* RGBA is not a supported input format any more:
  New elements will support only YUV formats to avoid implicit conversion
  done by hardware. Ideally it should be done by upstream element
  in order to have more control on it. Moreover, RGBA support can cause
  redundant RGBA -> YUV conversion if multiple encoders are
  used for the same RGBA input
* Subsampled planar format support is dropped:
  I420 and YV12 format are not supported formats for Direct3D11.
  Although it's supported in CUDA mode, it's not a hardware friendly
  memory layout and it will waste GPU memory since UV planes
  will have large padding due to the memory layout requirement of NVENC.
* GL support is dropped: Similar to the RGBA case,
  GL support in encoder would be suboptimal if GL input is
  used by multiple encoders, because each encoder will copy GL memory
  into CUDA memory.
  Upstream cudaupload element can be used for GL <-> CUDA
  interop instead.
* No more pre-allocation of encoder input surfaces. New implementation
  will use input CUDA memory without copy (zero-copy) or
  will copy into a NVENC's input buffer struct in case of
  system memory input.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1997>
2022-03-23 13:48:53 +00:00
Corentin Damman
1568db2c3e cudacontext: find associated DXGI Adapter LUID
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1807>
2022-03-10 18:08:10 +00:00
Corentin Damman
895f11401d cudautils: add support of D3D11 resource as Cuda graphics resource type
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1807>
2022-03-10 18:08:10 +00:00
Seungha Yang
e5132a8508 cudaupload,cudadownload: Add support for dGPU NVMM
Implement NVMM <-> CUDA, GL, SYSTEM memory conversion. Jetson is
not supported yet. Note that NVMM <-> GL interop on Jetson platform
is supported by GstGL

Some example pipelines are:
- Convert NVMM to GstGL memory
  nvv4l2decoder ! "video/x-raw(memory:NVMM)" ! cudadownload ! "video/x-raw(memory:GLMemory)" ! glimagesink

- Upload system memory to NVMM and encode
  video/x-raw,format=NV12 ! cudaupload ! "video/x-raw(memory:NVMM)" ! nvv4l2h264enc

- Convert NVMM to GstCUDA memory and encode
  nvvideoconvert ! "video/x-raw(memory:NVMM)" ! cudaupload ! "video/x-raw(memory:CUDAMemory)" ! nvh264enc

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1865>
2022-03-07 19:05:29 +00:00
Seungha Yang
c31bf2db4d nvcodec: Refactor cudaupload/download and add support for GL memory
* Implement new baseclass GstCudaMemoryCopy to remove duplicated
  cudaupload/download code
* Add support for CUDA <-> GL memory conversion via cudaupload/download

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1834>
2022-03-06 15:39:04 +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
Renamed from sys/nvcodec/meson.build (Browse further)