Skip to content

Latest commit

 

History

History
103 lines (87 loc) · 5.44 KB

video_decoder_test_usage.md

File metadata and controls

103 lines (87 loc) · 5.44 KB

Video Decoder tests

The video decoder tests are a set of tests used to validate various video decoder implementations. Multiple scenarios are tested, and the resulting decoded frames are validated against known checksums. These tests run directly on top of the video decoder implementation, and don't require the full Chrome browser stack. They can be very useful when adding support for a new codec or platform, or to make sure code changes don't break existing functionality. They are build on top of the GoogleTest framework.

[TOC]

Running from Tast

The Tast framework provides an easy way to run the video decoder tests from a ChromeOS chroot. Test data is automatically deployed to the device being tested. To run all video decoder tests use:

tast run $HOST video.DecodeAccelH264* video.DecodeAccelVP*

Wildcards can be used to run specific sets of tests:

  • Run all VP8 tests: tast run $HOST video.DecodeAccelVP8*
  • Run all VP9 profile 2 tests: tast run $HOST video.DecodeAccelVP92*

Check the tast video folder for a list of all available tests. See the Tast quickstart guide for more information about the Tast framework.

Running manually

To run the video decoder tests manually the video_decode_accelerator_tests target needs to be built and deployed to the device being tested. Running the video decoder tests can be done by executing:

./video_decode_accelerator_tests [<video path>] [<video metadata path>]

e.g.: ./video_decode_accelerator_tests test-25fps.h264

Test videos: Test videos are present for multiple codecs in the media/test/data folder in Chromium's source tree (e.g. test-25fps.vp8). If no video is specified test-25fps.h264 will be used.

Video Metadata: These videos also have an accompanying metadata .json file that needs to be deployed alongside the test video. They can also be found in the media/test/data folder (e.g. test-25fps.h264.json). If no metadata file is specified <video path>.json will be used. The video metadata file contains info about the video such as its codec profile, dimensions, number of frames and a list of md5 frame checksums to validate decoded frames. These frame checksums can be generated using ffmpeg, e.g.: ffmpeg -i test-25fps.h264 -f framemd5 test-25fps.h264.frames.md5.

Command line options

Multiple command line arguments can be given to the command:

-v                    enable verbose mode, e.g. -v=2.
--vmodule             enable verbose mode for the specified module,
                      e.g. --vmodule=*media/gpu*=2.

--validator_type      validate decoded frames, possible values are
                      md5 (default, compare against md5hash of expected
                      frames), ssim (compute SSIM against expected
                      frames, currently allowed for AV1 streams only)
                      and none (disable frame validation).
--use-legacy          use the legacy VDA-based video decoders.
--use_vd_vda          use the new VD-based video decoders with a
                      wrapper that translates to the VDA interface,
                      used to test interaction with older components
--linear_output       use linear buffers as the final output of the
                      decoder which may require the use of an image
                      processor internally. This flag only works in
                      conjunction with --use_vd_vda.
                      Disabled by default.
--output_frames       write the selected video frames to disk, possible
                      values are "all|corrupt".
--output_format       set the format of frames saved to disk, supported
                      formats are "png" (default) and "yuv".
--output_limit        limit the number of frames saved to disk.
--output_folder       set the folder used to store frames, defaults to
                      "<testname>".
--use-gl              specify which GPU backend to use, possible values
                      include desktop (GLX), egl (GLES w/ ANGLE), and
                      swiftshader (software rendering)
--ozone-platform      specify which Ozone platform to use, possible values
                      depend on build configuration but normally include
                      x11, drm, wayland, and headless
--disable_vaapi_lock  disable the global VA-API lock if applicable,
                      i.e., only on devices that use the VA-API with a libva
                      backend that's known to be thread-safe and only in
                      portions of the Chrome stack that should be able to
                      deal with the absence of the lock
                      (not the VaapiVideoDecodeAccelerator).

--gtest_help          display the gtest help and exit.
--help                display this help and exit.

Source code

See the video decoder tests source code.