All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=794993

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=604bf9c4c80e0614496a72abb0dc419c9dd7762bf6329be50f5c695b0c8475f9


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

android-nexus5

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12c438b8840000

Not sure why bisect didn't kick off originally; kicked off another bisect.

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/12c438b8840000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/16c1b46a840000

+dtu, simonhatch: more bisects not producing values.

I *think* this was the gitiles issue, based on the date and that the new jobs should get further.

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12dfe9dc440000

I re-started the bisect. simonbhatch, dtu: note it's still in "running" state.

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/12dfe9dc440000

Enable composited border radius scrolling by sunxd@chromium.org
https://chromium.googlesource.com/chromium/src/+/269a3a39c49562eac9e10b8fd3623fc93d3a874a

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

Thanks, yeah we're trying to tackle the cases where the job gets left in the running state forever, https://github.com/catapult-project/catapult/issues/4248 as an example. Believe Dave and I had talked about detecting these jobs frozen in the running state and changing them to errors as well.