All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=610244

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg1LWYtQoM


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

android-nexus6

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


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev   N  Good?
chromium@391400                  36.3539  0.205874  5  good
chromium@391425                  36.2341  0.197188  5  good
chromium@391432                  36.1813  0.146644  5  good
chromium@391435                  36.1761  0.231352  5  good
chromium@391435,skia@00d44e014c  36.246   0.204572  5  good
chromium@391436                  37.5526  0.143853  5  bad
chromium@391437                  37.6111  0.204757  5  bad
chromium@391438                  37.3345  0.436021  5  bad
chromium@391450                  37.5286  0.24169   5  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 610244

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.tough_path_rendering_cases
Test Metric: frame_times/frame_times
Relative Change: 3.23%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2168
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9013151421139797104


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5813092450041856

| 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!

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


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev    N  Good?
chromium@391431                  36.384   0.112489   5  good
chromium@391434                  36.5549  0.100737   5  good
chromium@391435                  36.4876  0.106518   5  good
chromium@391435,skia@00d44e014c  36.3507  0.078427   5  good
chromium@391436                  37.4983  0.116687   5  bad
chromium@391437                  37.6853  0.251357   5  bad
chromium@391443                  37.6237  0.0867798  5  bad
chromium@391454                  37.7696  0.169827   5  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 610244

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.tough_path_rendering_cases
Test Metric: frame_times/frame_times
Relative Change: 3.81%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2167
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9013151437770379776


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5839356040839168

| 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!

Trying with even more iterations.  It might not be possible to track this down with the bots :/

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


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev   N  Good?
chromium@391400                  36.1812  0.299677  5  good
chromium@391425                  36.0762  0.172041  5  good
chromium@391432                  36.2603  0.151404  5  good
chromium@391435                  35.873   0.346125  5  good
chromium@391435,skia@00d44e014c  36.1685  0.218495  5  good
chromium@391436                  37.4698  0.19767   5  bad
chromium@391437                  37.4582  0.108245  5  bad
chromium@391438                  37.3495  0.148281  5  bad
chromium@391450                  37.5536  0.10129   5  bad
chromium@391500                  37.536   0.140444  5  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 610244

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.tough_path_rendering_cases
Test Metric: frame_times/frame_times
Relative Change: 3.74%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2177
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9013146177951375728


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5907560926806016

| 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!

Hi ericrk@ & halcanary@

For some reason the bisects are not able to track down this regression, but it looks like it might be due to a skia roll.  Could you please check if either of your patches might have contributed?  Thanks!

https://chromium.googlesource.com/skia.git/+log/02125d10d5d0..0736f3386820

Kicked off another bisect.  Feedback from ricrk@ & halcanary@ would be appreciated.

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


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev    N  Good?
chromium@391431                  36.1404  0.104242   5  good
chromium@391434                  36.188   0.203387   5  good
chromium@391435                  36.361   0.0802587  5  good
chromium@391435,skia@00d44e014c  36.2768  0.166011   5  good
chromium@391436                  37.4674  0.0757033  5  bad
chromium@391437                  37.5302  0.0741749  5  bad
chromium@391443                  37.5204  0.066907   5  bad
chromium@391454                  37.5524  0.221387   5  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 610244

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.tough_path_rendering_cases
Test Metric: frame_times/frame_times
Relative Change: 3.91%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2200
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9011590086432787904


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5827877992398848

| 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 regression we're seeing is due to my change, 0736f3386820f19c0fe90b5dda2094e253780071. However, my change is just fixing a correctness regression that was landed here: https://chromium.googlesource.com/chromium/src/+/516778f8da98f40c94bfa34702fc167542bcd522

If you look at the graph over a longer period, you'll see that when the original correctness issue landed, it improved performance (as we were rendering less content, incorrectly). My change corrects this correctness issue and restores us to our previous (worse) performance.