All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=609482

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg1MH-rwkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg1L3qugoM


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

android-nexus5
android-nexus5X

===== BISECT JOB RESULTS =====
Status: completed


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev   N  Good?
chromium@391417                  71.7694  0.504506  5  good
chromium@391430                  71.0314  1.44216   5  good
chromium@391433                  71.4782  0.204211  5  good
chromium@391435                  71.6101  0.372646  5  good
chromium@391435,skia@00d44e014c  71.4898  0.297053  5  good
chromium@391436                  82.5099  0.181143  5  bad
chromium@391442                  82.2948  0.247518  5  bad

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 609482

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.tough_filters_cases
Test Metric: frame_times/http___letmespellitoutforyou.com_samples_svg_filter_terrain.svg
Relative Change: 14.67%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3666
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9013495768662781952


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5231335774355456

| 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 Tests>AutoBisect.  Thank you!

The suspect CL is a skia roll with two commits. Cc-ing the authors as I have know clue which commit caused this regression.

ericrk@, halcanary@: Any clue?

can't be me.

I think I know what caused this:

A few weeks ago, kkinnunen@ landed a change (crrev.com/1862043002) which broke our ability to GPU Raster with stencils. This was rolled (crrev.com/1912063002), and caused a perf improvement of 8.6% (Point ID: 389070 Time added: 2016-04-22T14:29:02.000Z).

While this was an improvement, the improvement was due to us rasterizing incorrectly.

I corrected this rasterization bug in the Skia roll mentioned above. This fixed the visual output, but moved our performance back to the previous levels.

We should probably just close this as working as intended.

WontFix per comment 5

 Issue 609483  has been merged into this issue.

 Issue 609788  has been merged into this issue.