All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=766572

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=bf58386093b7dfddd1fd197a2a67cb22e086c9c066569cba129ddc3ba481dce7


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

linux-release

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8968030474617579600

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

Hi bpastene@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 : Benjamin Pastene
  Commit : ef0f6ece2dcbcfdbfe15c8dbb6178fa324826b0a
  Date   : Fri Sep 15 23:31:28 2017
  Subject: Remove android_devices dimension from perf android swarming tests.

Bisect Details
  Configuration: linux_perf_bisect
  Benchmark    : thread_times.tough_scrolling_cases
  Metric       : thread_raster_cpu_time_per_frame/text_constant_full_page_raster_90000_pixels_per_second
  Change       : 33.45% | 4.79374828518 -> 6.39708888889

Revision             Result                   N
chromium@502250      4.79375 +- 0.390147      6      good
chromium@502345      5.47473 +- 1.50722       6      good
chromium@502392      5.0532 +- 1.17612        6      good
chromium@502416      5.46583 +- 0.636434      6      good
chromium@502428      5.43934 +- 0.780359      6      good
chromium@502434      5.27602 +- 0.890702      6      good
chromium@502437      5.31176 +- 0.578643      6      good
chromium@502438      5.48943 +- 0.583349      6      good
chromium@502439      6.39709 +- 0.512074      6      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 --story-filter=text.constant.full.page.raster.90000.pixels.per.second thread_times.tough_scrolling_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8968030474617579600


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8968019144464722032

+martiniss, eyaich: I started another job to double-check this. Could this CL affect performance?

=== BISECT JOB RESULTS ===
Perf regression found but unable to narrow commit range

Build failures prevented the bisect from narrowing the range further.


Bisect Details
  Configuration: linux_perf_bisect
  Benchmark    : thread_times.tough_scrolling_cases
  Metric       : thread_raster_cpu_time_per_frame/text_constant_full_page_raster_90000_pixels_per_second
  Change       : 34.00% | 4.89817536998 -> 6.56347504894

Suspected Commit Range
  3 commits in range
  https://chromium.googlesource.com/chromium/src/+log/387cce38aa459ed49bc4e4a0bd7ec390d77563a0..b56be486bfbc02d9421435603e77231ae7cb924b


Revision             Result                   N
chromium@502250      4.89818 +- 0.451236      6        good
chromium@502253      5.4308 +- 1.24607        14       good
chromium@502254      ---                      ---      build failure
chromium@502255      ---                      ---      build failure
chromium@502256      5.52346 +- 1.43902       14       bad
chromium@502262      5.84723 +- 0.514245      6        bad
chromium@502274      5.95784 +- 0.783245      6        bad
chromium@502298      5.88293 +- 0.808613      6        bad
chromium@502345      5.81012 +- 0.726763      6        bad
chromium@502439      6.56348 +- 1.11365       6        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 --story-filter=text.constant.full.page.raster.90000.pixels.per.second thread_times.tough_scrolling_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8968019144464722032


For feedback, file a bug with component Speed>Bisection

+Ned: the other bisect didn't complete, but in both we see a jump at r502439, "Remove android_devices dimension from perf android swarming tests". Any idea what could be going on here?

That change definitely should have been an effective no-op (the same tests are still running on the same bots) But it's safe to revert if we want to try that.

Ben is right that the change should be no-op. I think it's just due to chance that we have a jump  at r502439. Is it possible to rerun bisect on that revision to confirm?