Hm, that CL does look plausibly related. OTOH, the test is already marked as Timeout on Windows, see bug 666993.

Flakiness dashboard: https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=webkit_layout_tests&tests=external/wpt/requestidlecallback/callback-idle-periods.html

Summary: The test has been timing out on Win Dbg for a long time. It recently started timing out on Mac Dbg (reported here), and there are a few recent timeouts on Linux Dbg as well.
So maybe we should just change the timeout expectation from "Win" to "Debug"? Though it does seem like this CL has regressed it.

pdr@, any input? Could your CL have caused the recent timeouts?

Updating the test expectation at https://chromium-review.googlesource.com/c/chromium/src/+/1297146.

Assigning to pdr@ to investigate whether/how the "culprit" CL actually made this worse on Mac.

Looking into this now.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/17293893a9023f50e31d7981237a1dc27e24ee6c

commit 17293893a9023f50e31d7981237a1dc27e24ee6c
Author: Marc Treib <treib@chromium.org>
Date: Wed Oct 24 15:01:56 2018

external/wpt/requestidlecallback/callback-idle-periods.html flakily times out in Debug

The test was previously marked as timing out on Windows. However:
- It does occasionally pass.
- It doesn't actually time out on Windows Release.
- It does time out in Debug builds on other platforms though.

TBR=pdr@chromium.org

Bug: 666993, 898456
Change-Id: I5071af34d4fdb8bed911e864b28ce81eb954fff0
Reviewed-on: https://chromium-review.googlesource.com/c/1297146
Reviewed-by: Marc Treib <treib@chromium.org>
Commit-Queue: Marc Treib <treib@chromium.org>
Cr-Commit-Position: refs/heads/master@{#602339}
[modify] https://crrev.com/17293893a9023f50e31d7981237a1dc27e24ee6c/third_party/WebKit/LayoutTests/TestExpectations

Thanks!
Taking out of the sheriff queue.

I'm unable to reproduce this locally (1000 runs passed) so it's difficult to say what has increased the flakiness. I suspect my patch changed timing so that the existing test flakiness started showing up on other bots. If that's true, I'd like to assign this to the original test author. It's possible that my patch has some pathological performance issues with a no-op raf loop, so I'd like to leave this assigned to me temporarily as I work through the other failures related to my change.