All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=613192

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgjOCAsgoM


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

android-galaxy-s5

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


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev    N  Good?
chromium@391406                  17.1775  0.0699179  5  good
chromium@391423                  17.0687  0.0277465  5  good
chromium@391432                  17.1576  0.0679201  5  good
chromium@391434                  17.1742  0.0572899  5  good
chromium@391435                  17.1498  0.080313   5  good
chromium@391435,skia@00d44e014c  17.1048  0.0316999  5  good
chromium@391436                  17.5026  0.118652   5  bad
chromium@391440                  17.5424  0.181775   5  bad

Bisect job ran on: android_s5_perf_bisect
Bug ID: 613192

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: frame_times/http___www.polymer-project.org_components_paper-tabs_demo.html
Relative Change: 2.12%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_s5_perf_bisect/builds/670
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9012219106210659424


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6441739615731712

| 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!

crrev.com/391436  is a skia roll which is suspect for this raster metric regression:
https://chromium.googlesource.com/skia.git/+log/02125d10d5d0..0736f3386820

Assigning to ericrk@ because the culprit is most likely:
"2016-05-03 ericrk Allow stencils to be attached to render targets created via SkSurface::MakeFromBackendTextureAsRenderTarget"

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


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev    N  Good?
chromium@391406                  23.8115  0.0693668  6  good
chromium@391423                  23.7591  0.0265066  4  good
chromium@391432                  23.7255  0.053177   5  good
chromium@391434                  23.7309  0.0393171  6  good
chromium@391435                  23.7589  0.0416861  5  good
chromium@391435,skia@00d44e014c  23.7795  0.0520373  4  good
chromium@391436                  24.1368  0.0353429  5  bad
chromium@391440                  24.2261  0.299501   5  bad

Bisect job ran on: android_s5_perf_bisect
Bug ID: 613192

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: frame_times/frame_times
Relative Change: 1.72%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_s5_perf_bisect/builds/669
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9012219157976867040


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5841855896354816

| 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!

This is as expected - a previous change had broken functionality, causing us to render incorrectly (but more cheaply) - this was commit 389065, and is in the area where the linked graph shows an improvement. My change restores this functionality, bringing us back to the previous performance levels.