All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=811322

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=ca48cea126d210a8463f68b52933685902b3092a26486d15dd9feb22af8292ef


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

android-one

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12dde03d840000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/16d10d6d840000

Kicked a second bisect with a wider range because the graph goes up over a few revisions.

😿 Pinpoint job stopped with an error.
https://pinpoint-dot-chromeperf.appspot.com/job/12dde03d840000

😿 Pinpoint job stopped with an error.
https://pinpoint-dot-chromeperf.appspot.com/job/16d10d6d840000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14ef767d840000

😿 Pinpoint job stopped with an error.
https://pinpoint-dot-chromeperf.appspot.com/job/14ef767d840000

kouhei: bisect is unfortunately broken over the revision range. As the owner of loading.mobile, do you want to dig into this regression? Revision range is: https://chromium.googlesource.com/chromium/src/+log/3bcb977059e69c264ec1fc799517df422a445d2a%5E..636bbfa7b51f325fda9d105d8a2e2c6b5ae765aa?pretty=fuller&n=1000

I looked into the revision range but can't find obvious culprit. Closing this since I can't find anything actionable.