Issue 774169  has been merged into this issue.

I think the bug is actually in the test. It's what tdresser@ guessed in  crbug.com/774169 . Mojofication surfaced the bug. Fix is in https://chromium-review.googlesource.com/c/chromium/src/+/721762

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/980e4eb5a6b3846c331bf82cb00674ec4dce300b

commit 980e4eb5a6b3846c331bf82cb00674ec4dce300b
Author: Ehsan Chiniforooshan <chiniforooshan@chromium.org>
Date: Mon Oct 16 19:11:14 2017

Fix a mouse latency browsertest bug

The quit closure of a runloop is passed to the StartTracing call but
the runloop is never actually run. So, the test is not properly
waiting for the completion of StartTracing. This was not a problem
before 507716 since the tracing controller would immidiately start
tracing in the browser process. However the controller starts tracing
in the browser by a mojo call now which will create a PostTask.

BUG= 775040 

Change-Id: Ib74267c66a45eeba5065170a6727dfaa7e81846f
Reviewed-on: https://chromium-review.googlesource.com/721762
Commit-Queue: Ehsan Chiniforooshan <chiniforooshan@chromium.org>
Reviewed-by: Timothy Dresser <tdresser@chromium.org>
Cr-Commit-Position: refs/heads/master@{#509127}
[modify] https://crrev.com/980e4eb5a6b3846c331bf82cb00674ec4dce300b/content/browser/renderer_host/input/mouse_latency_browsertest.cc