Allow SkQP to be shardable
Per b/134522828 CtsSkQPTestCases used to not be properly shardable, but
that was before it was reworked. It seems something may have changed, as
upon investigation the tests seem to be properly distributed between
shards, without overlap (spot checked).
This should allow us more headroom for the 15m timeouts we're
encountering in TEST_MAPPING runs.
Test: compare Perfetto traces of different shards in TEST_MAPPING run
Reviewed-by: Leon Scroggins <email@example.com>
Auto-Submit: Nolan Scobie <firstname.lastname@example.org>
Commit-Queue: Nolan Scobie <email@example.com>
1 file changed