All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=718877

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg8sr9lgkM


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

win-high-dpi

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8980426293362501184

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8980426287397830272

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

Bisect Details
  Configuration: winx64_high_dpi_perf_bisect
  Benchmark    : smoothness.gpu_rasterization.tough_path_rendering_cases
  Metric       : frame_times/MotionMark Canvas Stroke Shapes

Revision             Result                  N
chromium@468544      59.5418 +- 13.9833      21      good
chromium@468596      58.4462 +- 4.52363      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 --story-filter=MotionMark.Canvas.Stroke.Shapes smoothness.gpu_rasterization.tough_path_rendering_cases

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8980426293362501184

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5217621319876608


| 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!

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

Bisect Details
  Configuration: winx64_high_dpi_perf_bisect
  Benchmark    : smoothness.gpu_rasterization.tough_path_rendering_cases
  Metric       : frame_times/MotionMark Canvas Stroke Shapes

Revision             Result                  N
chromium@468544      59.2168 +- 10.4144      21      good
chromium@468596      58.6807 +- 4.38937      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 --story-filter=MotionMark.Canvas.Stroke.Shapes smoothness.gpu_rasterization.tough_path_rendering_cases

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8980426287397830272

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5855095884873728


| 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/8980420603523774800

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8980420600554535888

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

Bisect Details
  Configuration: winx64_high_dpi_perf_bisect
  Benchmark    : smoothness.gpu_rasterization.tough_path_rendering_cases
  Metric       : frame_times/MotionMark Canvas Stroke Shapes

Revision             Result                  N
chromium@468544      57.6346 +- 5.45         21      good
chromium@468596      58.4943 +- 10.2444      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 --story-filter=MotionMark.Canvas.Stroke.Shapes smoothness.gpu_rasterization.tough_path_rendering_cases

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8980420603523774800

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5217621319876608


| 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!

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

Hi kouhei@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 : kouhei
  Commit : d087434ab42e298587480260ee4cafbfd002e614
  Date   : Tue May 02 04:15:45 2017
  Subject: [ES6 modules] ModuleTreeLinker::FetchDescendants should not assume urls isn't empty.

Bisect Details
  Configuration: winx64_high_dpi_perf_bisect
  Benchmark    : smoothness.gpu_rasterization.tough_path_rendering_cases
  Metric       : frame_times/MotionMark Canvas Stroke Shapes
  Change       : 1.97% | 58.1627699627 -> 57.01483452

Revision             Result                  N
chromium@468544      58.1628 +- 3.26124      14      good
chromium@468553      58.5915 +- 2.81146      14      good
chromium@468554      57.1139 +- 1.14609      6       bad       <--
chromium@468556      57.327 +- 3.61914       14      bad
chromium@468570      57.622 +- 4.13571       14      bad
chromium@468596      57.0148 +- 2.31059      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=MotionMark.Canvas.Stroke.Shapes smoothness.gpu_rasterization.tough_path_rendering_cases

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8980420600554535888

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5855095884873728


| 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!

I highly doubt this, as the touched codepath is only run under ES6 modules load.
Would you retry bisect?

False alert since ref has gone up as well