All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=759686

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=943400642310239601180b1ecd2b7a587c583fdfc619244fe8ce4d06a75e08ac


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

android-one

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970000064101797376

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

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : thread_times.key_idle_power_cases
  Metric       : tasks_per_second_total_all/set-timeout.html (Long Idle)

Revision             Result                   N
chromium@494099      4.28543 +- 0.823312      17      good
chromium@494324      4.46147 +- 0.618257      16      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=set.timeout.html..Long.Idle. thread_times.key_idle_power_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970000064101797376


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8966190536321198960

Bisecting on a wider range.

=== Auto-CCing suspected CL author kylechar@chromium.org ===

Hi kylechar@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : kylechar
  Commit : 68156e68656b00d9c963c5b30463a27b9627d14d
  Date   : Tue Aug 15 13:29:52 2017
  Subject: Enable surface references by default.

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : thread_times.key_idle_power_cases
  Metric       : tasks_per_second_total_all/set-timeout.html (Long Idle)
  Change       : 9.95% | 4.48285833422 -> 4.92872726474

Revision             Result                    N
chromium@494015      4.48286 +- 0.166251       6      good
chromium@494210      4.70187 +- 0.194872       6      good
chromium@494307      4.70518 +- 0.0711119      6      good
chromium@494356      4.7539 +- 0.130023        6      good
chromium@494362      4.71315 +- 0.12447        6      good
chromium@494365      4.75035 +- 0.183829       9      good
chromium@494366      4.70505 +- 0.151506       6      good
chromium@494367      4.90783 +- 0.0820104      6      bad       <--
chromium@494368      4.92888 +- 0.184139       6      bad
chromium@494380      4.92341 +- 0.133373       6      bad
chromium@494404      4.92873 +- 0.17518        6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=set.timeout.html..Long.Idle. thread_times.key_idle_power_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8966190536321198960


For feedback, file a bug with component Speed>Bisection

The suspect CL turned on surface references with Android. Surface references were turned off on Android again in https://crrev.com/c/665697 a few weeks later. I'm not sure why the metric didn't go back down at that point, but maybe the suspect CL wasn't the actual cause of the regression? Either way, that original change was reverted for Android.