All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=690215

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6KGF_wgM


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

chromium-rel-win8-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8988182867837764176

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

Hi nhiroki@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 : nhiroki
  Commit : 04055dd2786e27cde8d210b68e7a726fab6d0c42
  Date   : Sun Feb 05 04:05:05 2017
  Subject: Scheduler: Enqueue non-JS-timer tasks into the unthrottled task runner

Bisect Details
  Configuration: win_8_perf_bisect
  Benchmark    : smoothness.key_desktop_move_cases
  Metric       : mean_frame_time/mean_frame_time
  Change       : 1.67% | 17.6409444444 -> 17.9350555556

Revision             Result                   N
chromium@448158      17.6409 +- 0.784472      9      good
chromium@448168      17.4197 +- 0.736637      9      good
chromium@448171      17.4901 +- 0.505611      6      good
chromium@448172      17.3118 +- 0.347605      6      good
chromium@448173      18.0719 +- 0.415997      6      bad       <--
chromium@448178      17.9351 +- 0.340402      9      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 smoothness.key_desktop_move_cases

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8988182867837764176

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6685640533475328


| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Tests>AutoBisect.  Thank you!

nhiroki: Was this regression expected?

Yes, this was expected. Some tasks in a renderer process were accidentally posted to a task runner that can be throttled. To mitigate this, I moved them to the unthrottled task runner. Please let me know if you think this regression is not acceptable.