All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=762225

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=f952fdfb7bdf1f3b6346037b96dcfd6f2e0428ab44e74529bc9024ea9f3b40cf


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

android-webview-nexus5X
chromium-rel-win7-gpu-ati
chromium-rel-win7-gpu-intel

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969255948984934768

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

Hi dtapuska@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 : Dave Tapuska
  Commit : 801cc948b453e8e4b4afc1a08d0a83df0e51aa82
  Date   : Thu Aug 31 14:57:29 2017
  Subject: Enable mojo input messages on all platforms other than android.

Bisect Details
  Configuration: winx64intel_perf_bisect
  Benchmark    : smoothness.top_25_smooth
  Metric       : first_gesture_scroll_update_latency/Facebook
  Change       : 499.17% | 1.66833333333 -> 9.99622222222

Revision             Result                   N
chromium@498827      1.66833 +- 0.19719       9      good
chromium@498860      1.62644 +- 0.174626      9      good
chromium@498863      3.12944 +- 12.7618       9      good
chromium@498864      1.65917 +- 0.133165      6      good
chromium@498865      12.8502 +- 8.93785       6      bad       <--
chromium@498869      10.2777 +- 11.123        6      bad
chromium@498877      9.826 +- 16.9134         6      bad
chromium@498893      9.99622 +- 16.0076       9      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Facebook smoothness.top_25_smooth

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969255948984934768


For feedback, file a bug with component Speed>Bisection

The bisect is likely wrong for the Nexus device because this feature wasn't enabled for Android at all.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969247339323226096

=== Auto-CCing suspected CL author tobiasjs@google.com ===

Hi tobiasjs@google.com, 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 : Tobias Sargeant
  Commit : 416174f07b4d1d930fee687ddd0eae67fbc24ee1
  Date   : Fri Sep 01 01:18:09 2017
  Subject: Disable ScrollTopLeftInterop in WebView.

Bisect Details
  Configuration: android_webview_arm64_aosp_perf_bisect
  Benchmark    : smoothness.top_25_smooth
  Metric       : frame_times/LinkedIn
  Change       : 1.43% | 16.4808565891 -> 16.7170386989

Revision             Result                    N
chromium@498815      16.4809 +- 0.0795109      6       good
chromium@498982      16.4731 +- 0.0473365      6       good
chromium@499066      16.454 +- 0.101025        6       good
chromium@499087      16.486 +- 0.0561013       6       good
chromium@499098      16.4634 +- 0.125342       9       good
chromium@499100      16.4944 +- 0.0903074      14      good
chromium@499101      16.6455 +- 0.553824       14      bad       <--
chromium@499103      16.6384 +- 0.473736       9       bad
chromium@499108      16.7569 +- 0.0268128      6       bad
chromium@499149      16.717 +- 0.211059        6       bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-webview --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=LinkedIn smoothness.top_25_smooth

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969247339323226096


For feedback, file a bug with component Speed>Bisection

Looking at the graphs I think what's happened here is that there was a temporary improvement in the stability of measurements as a result of enabling ScrollTopLeftInterop, which I disabled again, bringing back the old behaviour. There was some secondary regression in the middle of the period between enable and disable.

The CL I landed was probably the cause of this regression, but there's nothing that can be done about it, because we can't enable ScrollTopLeftInterop for WebView.

 Issue 762293  has been merged into this issue.