All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=684565

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgoPGT-AgM


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

android-nexus5X

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8989574113900088752

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8989574103851099344

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

Hi mtklein@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 : Mike Klein
  Commit : e71b167dbd4d8da76495ca85db83d1a3b49aaabd
  Date   : Fri Jan 13 13:25:28 2017
  Subject: Attempt 3: SkRasterPipelineBlitter: support A8

Bisect Details
  Configuration: android_nexus5X_perf_bisect
  Benchmark    : webrtc.peerconnection
  Metric       : cpu_utilization/cpu_utilization_browser
  Change       : 21.37% | 15.8078682151 -> 19.1861713851

Revision                             Result                   N
chromium@443552                      15.8079 +- 1.89059       6      good
chromium@443563                      16.396 +- 0.625887       6      good
chromium@443564                      16.3192 +- 1.1245        6      good
chromium@443564,skia@e71b167dbd      18.7519 +- 1.00314       6      bad       <--
chromium@443564,skia@7adde145d3      19.2802 +- 1.20617       6      bad
chromium@443565                      19.2486 +- 1.31168       6      bad
chromium@443566                      19.3283 +- 0.613272      6      bad
chromium@443569                      19.2896 +- 0.833095      6      bad
chromium@443574                      19.2303 +- 1.26169       6      bad
chromium@443595                      19.1862 +- 1.305         6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests webrtc.peerconnection

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8989574113900088752

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5230101567897600


| 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 ===
Perf regression found with culprit

Suspected Commit
  Author : Mike Klein
  Commit : e71b167dbd4d8da76495ca85db83d1a3b49aaabd
  Date   : Fri Jan 13 13:25:28 2017
  Subject: Attempt 3: SkRasterPipelineBlitter: support A8

Bisect Details
  Configuration: android_nexus5X_perf_bisect
  Benchmark    : webrtc.peerconnection
  Metric       : cpu_utilization/cpu_utilization_browser
  Change       : 22.61% | 15.5784932646 -> 19.1003744788

Revision                             Result                   N
chromium@443552                      15.5785 +- 0.903313      6      good
chromium@443563                      16.4325 +- 1.85688       6      good
chromium@443564                      16.1009 +- 1.30445       6      good
chromium@443564,skia@e71b167dbd      19.2138 +- 0.403078      6      bad       <--
chromium@443564,skia@7adde145d3      19.2033 +- 0.895027      6      bad
chromium@443565                      19.1912 +- 0.864368      6      bad
chromium@443566                      19.0976 +- 1.43029       6      bad
chromium@443569                      19.0803 +- 1.5202        6      bad
chromium@443574                      18.9004 +- 1.00885       6      bad
chromium@443595                      19.1004 +- 0.73831       6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests webrtc.peerconnection

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8989574103851099344

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5331088865492992


| 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!

Any performance regression from that CL is intentional in exchange for quality.

Does this benchmark really measure graphics performance?
  Benchmark    : webrtc.peerconnection
  Metric       : cpu_utilization/cpu_utilization_browser

I'd say so. It launches a loopback call in 720p, and displays both the video from the camera, and the video received from the peerconnection.

The page loaded is this one:
https://webrtc.github.io/samples/src/content/peerconnection/constraints/