All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=706391

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg7My36QsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg7OWkswsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg7MzApgkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg7LnW6goM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgnIy4twkM


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

win-zenbook

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8983782315116467312

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

Hi pkotwicz@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 : pkotwicz
  Commit : 0e44ebc7dcf7c343d45694b97b2edc455f1a6256
  Date   : Mon Mar 27 16:25:49 2017
  Subject: [Android WebAPKs] Don't navigate WebAPK when launching it from launcher

Bisect Details
  Configuration: winx64_zen_perf_bisect
  Benchmark    : speedometer
  Metric       : EmberJS-TodoMVC/EmberJS-TodoMVC
  Change       : 5.62% | 1230.61942857 -> 1263.19841667

Revision             Result                  N
chromium@459798      1230.62 +- 221.23       14      good
chromium@459799      1311.96 +- 269.026      9       bad       <--
chromium@459800      1287.01 +- 111.15       9       bad
chromium@459801      1317.68 +- 182.315      9       bad
chromium@459804      1289.44 +- 107.646      9       bad
chromium@459809      1274.31 +- 218.387      21      bad
chromium@459822      1274.48 +- 158.081      14      bad
chromium@459844      1232.08 +- 111.767      14      bad
chromium@459890      1250.27 +- 215.756      9       bad
chromium@459982      1263.2 +- 101.416       6       bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests speedometer

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8983782315116467312

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6453990407012352


| 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!

It is impossible for my CL to be the cause. My CL modifies Android only code (which is run only in a super weird configuration of Chrome for Android)

Reassigning to  tebbi@ to do bisect. The bisect results look super duper suspicious. For instance revisions 459798 and 459844 have a very similar result (which landed before and after my CL respectively)

Ref build is also having the same spike => bot problem.