All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=808640

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=9b4d17b85c0220f4c968c3542f907751c337812241fa1042d68ceb7c97ac27cc


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

android-nexus6

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14526bce840000

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/14526bce840000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14d530d1840000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/14d530d1840000

Revert "Add UKM metrics for rendering timing measurements" by cbiesinger@chromium.org
chromium @ 772b2bcdbaa8ea5c5b1f47255c6c1562016fb680

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

the nexus5 graph shows this just got things back to the previous state

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/11b347e9840000

Most of the graphs didn't recover so I don't think we should close this yet. Hopefully this new bisect points to a different culprit.

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/11b347e9840000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12c1410d840000

Triggered a bisect with a wider range.

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/12c1410d840000

It looks like both ref and tot had huge changes at the same time. Could this be an infra related issue?

sullivan@ any advice/guidance on how to proceed farther for this kind of regressions?

I would WontFix this because the results all appear to come from the same Nexus6 device, with the ref build affected as well. Seems like it may be a device swap?

Thank you!