Deflake the SubprocessMetricsProviderBrowserTest
There are two issues:
It seemed that triggering render process crash isn't reliable,
RenderProcessExit sometime wasn't sent on navigation return, as
the render process crash isn't critical case we want to test,
this patch uses the simulation to avoid the flaky.
The specific histogram the test checked wasn't always written after
the navigation, this patch picked a random histogram from render
process allocator and verifies if it was copied to the global
allocator.
Bug: 1102807,1102852
Change-Id: If1456b3fc87b0255c549e12350074e318737dcf8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2285319
Commit-Queue: Tao Bai <michaelbai@chromium.org>
Reviewed-by: Brian White <bcwhite@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#786877}
Cr-Mirrored-From: https://chromium.googlesource.com/chromium/src
Cr-Mirrored-Commit: d1006d308caa55dba3dbe99d86f5f84494540f25
2 files changed