GitLab

Right now the slower thread throughput is simply taking the lower
throughput between the main and compositor throughput.

This CL improves that. Specifically, when we are expecting and
producing main frames, then we count the frames produced from the
main thread towards the slower thread. Otherwise, use the numbers
from the compositor thread towards the slower thread.

Bug: 1002994
Change-Id: Ie2de3ff958f920e6afa5234b7cf40aea05930ab9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2022618Reviewed-by: Annie Sullivan <sullivan@chromium.org>
Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Reviewed-by: Robert Flack <flackr@chromium.org>
Commit-Queue: Xida Chen <xidachen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#754547}