Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970718561904532432

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

Hi michaelpg@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 : Michael Giuffrida
  Commit : 2e59602847f8073dc1f2a58d15697c470a710327
  Date   : Sat Jul 22 03:16:57 2017
  Subject: Split MessageService::MessagePort into own file

Bisect Details
  Configuration: android_nexus5_perf_bisect
  Benchmark    : thread_times.tough_compositor_cases
  Metric       : thread_raster_cpu_time_per_frame/http___jsbin.com_beqojupo_1_quiet?JS_FULL_SCREEN_INVALIDATION
  Change       : 2.73% | 9.67372264836 -> 9.93751621544

Revision             Result                    N
chromium@488097      9.67372 +- 0.0643162      6       good
chromium@488478      9.66828 +- 0.145333       6       good
chromium@488669      9.77749 +- 0.0686703      6       good
chromium@488764      9.81012 +- 0.0446377      6       good
chromium@488812      9.82326 +- 0.0977624      6       good
chromium@488836      9.86633 +- 0.146076       14      good
chromium@488843      9.83598 +- 0.12035        9       good
chromium@488846      9.83543 +- 0.151924       14      good
chromium@488847      9.84285 +- 0.0862117      9       good
chromium@488848      9.89855 +- 0.173883       14      bad       <--
chromium@488859      9.93752 +- 0.0922531      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=http...jsbin.com.beqojupo.1.quiet.JS.FULL.SCREEN.INVALIDATION thread_times.tough_compositor_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970718561904532432


For feedback, file a bug with component Speed>Bisection

This seems to be too noisy to bisect.  I have no confidence in what the bisect at comment #3 points to.  Re-assigning back to myself.

agreed, a perf regression from that CL is unlikely to begin, with but there's no way it could've affected android.

Too small to bisect, closing.