GitLab

The thread cache uses batched allocation from the central
allocator. Rather than using the cache hit rate to assess contention on
the main lock, the number of batch fill requests is the one to track.

This CL records this metric, and reports it in memory dumps and in UMA.

Bug: 998048
Change-Id: Ie6a833c7a0ece66d3c138757c8fb3f49a1690c7f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2593630Reviewed-by: Kentaro Hara <haraken@chromium.org>
Reviewed-by: ssid <ssid@chromium.org>
Reviewed-by: Jesse Doherty <jwd@chromium.org>
Commit-Queue: Benoit L <lizeb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#838015}