All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=751828

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=32733c825b8ea2c38f09123254f81155fd393a6b6ce07fe3ceddd6b9ed39df0b


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

chromium-rel-mac11

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8972338403528493904

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

Hi fdoray@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 : Francois Doray
  Commit : 26486308a3f4f2443a15f4c5813c0cab0b53c8d8
  Date   : Mon Jul 31 23:15:23 2017
  Subject: Remove usage of ScopedTaskScheduler from PipelineIntegrationTestBase.

Bisect Details
  Configuration: mac_10_11_perf_bisect
  Benchmark    : media_perftests
  Metric       : clockless_video_playback_theora/bear_silent.ogv
  Change       : 1.22% | 258.965707274 -> 255.795473905

Revision             Result                  N
chromium@490808      258.966 +- 7.33773      9      good
chromium@490813      260.329 +- 4.20918      6      good
chromium@490816      260.257 +- 2.59785      6      good
chromium@490817      256.982 +- 2.06941      6      bad       <--
chromium@490818      256.28 +- 3.54182       9      bad
chromium@490828      255.814 +- 3.88192      9      bad
chromium@490849      255.795 +- 2.86374      9      bad

To Run This Test
  ./src/out/Release/media_perftests --single-process-tests

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8972338403528493904


For feedback, file a bug with component Speed>Bisection

Previously, this perf test used ScopedTaskScheduler which runs all tasks on the main thread. Now, it uses ScopedTaskEnvironment which runs tasks in a real TaskScheduler.

This isn't a perf regression for production code. It's just a change that makes the test run in an environment which is closer from the production environment.