All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=759697

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=dbc75b50aeb79c90016ce84401e4fdb7c0e0a90303e16b5bb40a53973a146d8e


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

chromium-rel-mac11-air

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969999935559130624

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: mac_air_perf_bisect
  Benchmark    : loading.desktop
  Metric       : timeToFirstContentfulPaint_avg/pcv1-cold/Haraj

Revision             Result                  N
chromium@495286      652.881 +- 628.957      21      good
chromium@495420      676.903 +- 563.376      21      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Haraj loading.desktop

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969999935559130624


For feedback, file a bug with component Speed>Bisection

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14a95b79780000

Trying a pinpoint bisect.

📍 Found significant differences after each of 2 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/14a95b79780000

Revert "Ensure overlapping compositor locks can time out"
By ccameron@chromium.org · Fri Aug 11 04:59:46 2017
chromium @ 28e0a3c183499e50b1b89231d4ed40bafb8883fe

Convert loading.desktop page set to WprGo
By ksakamoto@chromium.org · Tue Aug 15 04:08:53 2017
chromium @ ca2e6be7bac44462efdd864ef346311b9a6033de

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

I think there's something going on on the right side of that Pinpoint job, too. A repeat_count of 20 would have detected something.

The big causes of the regression are known (a revert, and a test case re-record) and not things we're going to act on, so WontFix-ing this one.