All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=714907

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg4srM9wgM


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

chromium-rel-win7-gpu-intel

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8981385122063000528

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: winx64intel_perf_bisect
  Benchmark    : blink_perf.events
  Metric       : hit-test-lots-of-layers/hit-test-lots-of-layers

Revision             Result                     N
chromium@466304      0.925627 +- 0.0258104      21      good
chromium@466349      0.927899 +- 0.0425087      21      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 blink_perf.events

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981385122063000528

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5825132649512960


| 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 Speed>Bisection.  Thank you!

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8981376087144356384

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

Hi thakis@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 : thakis
  Commit : 9cb3b1d088bdd435de0292a58966c4c514acd0a1
  Date   : Fri Apr 21 14:48:55 2017
  Subject: Address some TODO(GYP_GONE)s

Bisect Details
  Configuration: winx64intel_perf_bisect
  Benchmark    : blink_perf.events
  Metric       : hit-test-lots-of-layers/hit-test-lots-of-layers
  Change       : 0.44% | 0.838510926346 -> 0.842201023843

Revision             Result                      N
chromium@466304      0.838511 +- 0.0312513       21      good
chromium@466327      0.839038 +- 0.00564689      14      good
chromium@466330      0.835311 +- 0.0102265       9       good
chromium@466331      0.840306 +- 0.0116192       9       bad       <--
chromium@466332      0.840935 +- 0.0088877       9       bad
chromium@466333      0.840147 +- 0.0383748       21      bad
chromium@466338      0.84178 +- 0.0134727        21      bad
chromium@466349      0.842201 +- 0.0140772       21      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 blink_perf.events

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981376087144356384

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5825132649512960


| 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 Speed>Bisection.  Thank you!

That's a behavior-preserving CL and can't cause regressions. I've reported the bisect as incorrec.t

From the regression range (https://chromium.googlesource.com/chromium/src/+log/27d529dd566202cbb4d347d5ccf5d5e0aa198a63%5E..6e1f3c35d3d0d34ea488a59b3990c5770a701e8e?pretty=fuller) here are a few CLs that look vaguely related.

https://codereview.chromium.org/2813683002
https://codereview.chromium.org/2834553002
https://codereview.chromium.org/2831943002
https://codereview.chromium.org/2773893003

cc'd folks: Please check if your CL could plausibly have caused this regression.

https://codereview.chromium.org/2834553002 doesn't change behaviors. It is 
unlikely to be the culprit.

I think https://codereview.chromium.org/2831943002 is OK too: hit-test-lots-of-layers.html doesn't seem to exercise the JS-sequence conversion code path.

https://codereview.chromium.org/2813683002 is unlikely because hit-test-lots-of-layers.html doesn't process any input. It just executes elementFromPoint over and over.

https://codereview.chromium.org/2773893003 is unlikely because hi-test-lots-of-layers.html doesn't do any scrolls and the patch added logic only in "scroll begin".

This alert was found before M-60 branched. Closing as WontFix as this is believed to either be invalid or non-reproducible.