All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=760567

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=f8f21f09886e88b3e82dafe91438676702b989215f71c50642bc418db6d7986d


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

android-nexus7v2

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969831196831481056

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

Hi dtapuska@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 : Dave Tapuska
  Commit : f238a1a8f2f73d226bdb38394a6e085a59cf0e41
  Date   : Mon Aug 28 21:24:39 2017
  Subject: Move Input Injection from Chrome IPC to mojo.

Bisect Details
  Configuration: android_nexus7_perf_bisect
  Benchmark    : thread_times.key_idle_power_cases
  Metric       : tasks_per_second_total_all/request-animation-frame.html
  Change       : 11.66% | 4.50278481754 -> 5.02772868391

Revision             Result                    N
chromium@497870      4.50278 +- 0.204323       6      good
chromium@497877      4.58584 +- 0.349479       6      good
chromium@497879      4.46135 +- 0.194108       6      good
chromium@497880      5.06844 +- 0.220474       6      bad       <--
chromium@497881      4.99401 +- 0.118604       6      bad
chromium@497884      5.06853 +- 0.314075       6      bad
chromium@497897      4.94332 +- 0.0998198      6      bad
chromium@497924      5.12641 +- 0.371471       6      bad
chromium@497977      5.02773 +- 0.238134       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 --story-filter=request.animation.frame.html thread_times.key_idle_power_cases

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969831196831481056


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969713678488785472

=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Dave Tapuska
  Commit : f238a1a8f2f73d226bdb38394a6e085a59cf0e41
  Date   : Mon Aug 28 21:24:39 2017
  Subject: Move Input Injection from Chrome IPC to mojo.

Bisect Details
  Configuration: android_nexus7_perf_bisect
  Benchmark    : thread_times.key_idle_power_cases
  Metric       : tasks_per_second_total_all/request-animation-frame.html
  Change       : 8.58% | 4.66112130709 -> 5.06098038013

Revision             Result                   N
chromium@497800      4.66112 +- 0.338607      6      good
chromium@497850      4.528 +- 0.115104        6      good
chromium@497875      4.51112 +- 0.335788      6      good
chromium@497878      4.62783 +- 0.152506      6      good
chromium@497879      4.54439 +- 0.222489      6      good
chromium@497880      5.08515 +- 0.147242      6      bad       <--
chromium@497881      5.03535 +- 0.12997       6      bad
chromium@497887      5.02697 +- 0.312807      6      bad
chromium@497899      5.00294 +- 0.278003      6      bad
chromium@497997      5.06098 +- 0.330944      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 --story-filter=request.animation.frame.html thread_times.key_idle_power_cases

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969713678488785472


For feedback, file a bug with component Speed>Bisection

 Issue 761727  has been merged into this issue.

 Issue 761732  has been merged into this issue.

 Issue 761728  has been merged into this issue.

 Issue 761940  has been merged into this issue.

 Issue 761729  has been merged into this issue.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b75607c70c4a4ad0e34d7d2ff1a2ea7aa0a50662

commit b75607c70c4a4ad0e34d7d2ff1a2ea7aa0a50662
Author: Dave Tapuska <dtapuska@chromium.org>
Date: Tue Sep 05 22:11:11 2017

Ensure that we only connect the remote gpu mojo interface when necessary.

There was some slight regressions in some the v8 benchmarks and they
were bisected to
https://crrev.com/f238a1a8f2f73d226bdb38394a6e085a59cf0e41.

This change tries to eliminate the regressions by connecting the mojo
interface when only necessary. I believe it caused the regressions in
telemetry because of the number of nested iframes that don't end up
using the interface in the end.

BUG= 760567 

Change-Id: I9311498186ceed32281454a589c66380707843e3
Reviewed-on: https://chromium-review.googlesource.com/651212
Reviewed-by: Antoine Labour <piman@chromium.org>
Commit-Queue: Dave Tapuska <dtapuska@chromium.org>
Cr-Commit-Position: refs/heads/master@{#499766}
[modify] https://crrev.com/b75607c70c4a4ad0e34d7d2ff1a2ea7aa0a50662/content/renderer/gpu/gpu_benchmarking_extension.cc
[modify] https://crrev.com/b75607c70c4a4ad0e34d7d2ff1a2ea7aa0a50662/content/renderer/gpu/gpu_benchmarking_extension.h

Graphs show that connecting the mojo interface only when necessary has removed the regressions.

 Issue 762304  has been merged into this issue.