All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=687743

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgyMLKvQsM


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

chromium-rel-mac11

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8988818361108092848

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

Hi mmenke@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 : mmenke
  Commit : 87f5c77aaa3541bee2f8d3b6b64d3879b8e81957
  Date   : Tue Jan 31 16:11:26 2017
  Subject: Refactor ResourceHandler API.

Bisect Details
  Configuration: mac_10_11_perf_bisect
  Benchmark    : service_worker.service_worker_micro_benchmark
  Metric       : concurrent_100_response_99_percentile/concurrent_100_response_99_percentile
  Change       : 7.78% | 56.6907142857 -> 61.1016666667

Revision             Result                  N
chromium@447228      56.6907 +- 10.5718      14      good
chromium@447236      55.1217 +- 3.67539      6       good
chromium@447238      54.4475 +- 1.68442      6       good
chromium@447239      54.7242 +- 2.82428      6       good
chromium@447240      60.5611 +- 7.99838      9       bad       <--
chromium@447244      62.6775 +- 2.42961      6       bad
chromium@447260      61.5154 +- 10.7699      14      bad
chromium@447291      61.1017 +- 7.66575      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 service_worker.service_worker_micro_benchmark

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8988818361108092848

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5228688238444544


| 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!

Going to investigate this and  issue 673625  as the same issue (Same benchmark, related CLs).  Need to figure out if the issue is extra post tasks, or something else is going on.  (Notice that this is 2x the change of the earlier one, so could be 1 extra post task there, 2 here).  One extra is probably expected - there was a case of re-entrancy I addressed in the CL, but not sure why there'd be two.  Or could be something else entirely.