All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=818671

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=07aef59bdb219baa891d308a011a2619913456dc2121334603792e19cf68d150


Bot(s) for this bug's original alert(s):

android-one
chromium-rel-win8-dual

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14c2f0dc440000

📍 Found significant differences after each of 3 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/14c2f0dc440000

predictors: Remove ResourcePrefetchPredictor by alexilin@chromium.org
https://chromium.googlesource.com/chromium/src/+/b5ea1fe654f7e0e71704dc278ccd0b5c5cab7a8a

Oilpan: Emit write barrier on collection backing assignment by keishi@chromium.org
https://chromium.googlesource.com/chromium/src/+/d0e0bb6ae37606fdb8b19bab4714706af7580088

Revert "Revert "webauthn: support the appid extension."" by agl@chromium.org
https://chromium.googlesource.com/chromium/src/+/dc639a910dd43a1c1ff88715f092669e5f57878a

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

From the pinpoint page, only "Oilpan: Emit write barrier on collection backing assignment" appears to have a visually-significant impact. But the impact is downwards so I'm not sure whether it's good or bad.

The other two nominated changes (which includes mine) appear to be in the noise from what I can tell. "predictors: Remove ResourcePrefetchPredictor" is accompanied by an abnormally wide error bar and mine ("Revert "Revert "webauthn: support the appid extension."") drifts back up to the previous level over the next few revisions.

Thus I'm reassigning to keishi because their change is the dominant one.