GitLab

Decoding a bitstream in VA has two steps: one, submitting the parsed
parameters and encoded chunk as VABufferIDs, and two, executing the
decode. This CL transfers managing the lifetime of the VABufferIDs to
VP8VaapiVideoDecoderDelegate, avoiding vaCreateBuffer() calls.

vaapi_utils' FillVP8DataStructures() needs to be changed to operate
on VA-API structs passed as parameter, which in turn forces a bit
of changes in the webp decoder.

This is verified via chrome:tracing and codepen.io/full/qzYXba that
plays 4 1280x572 VP8 videos at the same time. Tracing is captured for
a few seconds, and basically we avoid calling the vaCreateBuffer(),
which takes ~50us every time, reducing the overall decode time from
~0.439ms to ~0.307ms on my CML (kohaku), ~25% reduction. Improvements
are of course extremely small, the advantages of this CL are in
reducing lock/unlock churn and associated contention. This benefit
grows with the amount of decodes (e.g. Meet grid scenarios).

Bug: b/166646505
Change-Id: I6b0e2dc21ad94e2e1c9ab3696fba037d5b88830b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2412132
Commit-Queue: Miguel Casas <mcasas@chromium.org>
Reviewed-by: Andres Calderon Jaramillo <andrescj@chromium.org>
Cr-Commit-Position: refs/heads/master@{#809676}