All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=709674

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgvPyqqgsM


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

android-webview-nexus6

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8982930034149622640

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

Hi alancutter@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 : alancutter
  Commit : 317a8e0e26459f401312043fa99ed8e17b1b24a4
  Date   : Wed Apr 05 10:09:29 2017
  Subject: Notify Blink to suppress frame requests during BeginMainFrame

Bisect Details
  Configuration: android_webview_nexus6_aosp_perf_bisect
  Benchmark    : page_cycler_v2.intl_hi_ru
  Metric       : timeToFirstContentfulPaint_avg/pcv1-warm/http___www.indiatimes.com_
  Change       : 29.48% | 429.4395 -> 556.031166667

Revision             Result                  N
chromium@462009      429.44 +- 44.9948       6      good
chromium@462016      437.583 +- 26.3951      6      good
chromium@462019      435.823 +- 32.2039      6      good
chromium@462021      426.613 +- 24.4167      6      good
chromium@462022      565.619 +- 29.9839      6      bad       <--
chromium@462035      556.031 +- 34.8556      6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-webview --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.indiatimes.com. page_cycler_v2.intl_hi_ru

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8982930034149622640

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6334608469131264


| 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/8982739093395610800

317a8e0e26459f401312043fa99ed8e17b1b24a4 is intended as a revert of the main thread frame scheduling effects of 157413286770a7ac5a24c446a30c08f749738276.
It looks like the first time this metric drops to the ~470ms range includes 157413286770a7ac5a24c446a30c08f749738276 so it makes sense that 317a8e0e26459f401312043fa99ed8e17b1b24a4 would undo some of the effects it had.

First drop into ~470ms revision range:
http://test-results.appspot.com/revision_range?start=449920&end=449932
Started bisect for that range:
https://chromeperf.appspot.com/buildbucket_job_status/8982739093395610800

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

Hi yosin@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 : yosin
  Commit : 157413286770a7ac5a24c446a30c08f749738276
  Date   : Mon Feb 13 10:55:13 2017
  Subject: Make FrameSelection to hold non-canonicalized DOM positions

Bisect Details
  Configuration: android_webview_nexus6_aosp_perf_bisect
  Benchmark    : page_cycler_v2.intl_hi_ru
  Metric       : timeToFirstContentfulPaint_avg/pcv1-warm/http___www.indiatimes.com_
  Change       : 21.06% | 597.646333333 -> 471.8105

Revision             Result                  N
chromium@449919      597.646 +- 30.7516      6      good
chromium@449926      602.969 +- 51.366       9      good
chromium@449927      604.646 +- 46.1531      6      good
chromium@449928      504.579 +- 133.897      6      bad       <--
chromium@449929      493.813 +- 139.279      9      bad
chromium@449932      471.811 +- 41.9761      6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-webview --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.indiatimes.com. page_cycler_v2.intl_hi_ru

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8982739093395610800

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5283145311584256


| 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!

#7 confirms my suspicions in #6. This is WAI; my change was intended to revert frame related effects of 157413286770a7ac5a24c446a30c08f749738276.

Regarding #c7, http://crrev.com/449928 was reverted then reland in  http://crrev.com/450370

 Issue 710210  has been merged into this issue.