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]
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.
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.
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.
See the video decoder tests source code.