All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=666309

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgy7qRqAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgy7jP-AoM


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

chromium-rel-mac-hdd

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8995735994917963488

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

Hi zmo@chromium.org, the bisect results pointed to your CL below as possibly
causing a regression. Please have a look at this info and see whether
your CL be related.


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


===== SUSPECTED CL(s) =====
Subject : Don't forbid GPU process if WebGL/WebGL2 is disabled on GPU but not browser.
Author  : zmo
Commit description:
  
BUG= 662170 
TEST=media.tough_video_cases,content_unittests
R=kbr@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2505473002
Cr-Commit-Position: refs/heads/master@{#432073}
Commit  : 5881e9b3bc7c96dc8a0d5ec823078e3b4d80a85c
Date    : Tue Nov 15 03:42:58 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@431940  41.406   8.54528  8  good
chromium@432035  41.8724  6.49604  8  good
chromium@432059  43.8374  10.8831  8  good
chromium@432071  41.0747  14.4997  8  good
chromium@432072  42.6604  3.24702  8  good
chromium@432073  84.0462  5.73008  5  bad    <--
chromium@432074  87.3848  7.91437  5  bad
chromium@432077  83.64    4.50972  5  bad
chromium@432082  83.893   4.91064  5  bad
chromium@432129  82.192   7.06248  5  bad

Bisect job ran on: mac_hdd_perf_bisect
Bug ID: 666309

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=https...plus.google.com.110031535020051778989.posts smoothness.top_25_smooth
Test Metric: mean_input_event_latency/https___plus.google.com_110031535020051778989_posts
Relative Change: 98.50%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_hdd_perf_bisect/builds/887
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8995735994917963488


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6436170785882112

| 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!

I confirm it is my CL because my CL switched the bot from running on CPU (GPU being blacklisted) to running on GPU.

I would have assumed running on GPU will be a perf gain, but apparently not on that bot.

I don't really know what to do here. It seems we should just set a new base now we run on a different path.

Adding vmiura

WontFix is fine with me if we switched the bot to a different codepath.

I'll leave it to vmiura to decide

Discussing with zmo@, it sounds like the telemetry test was showing only cold start performance, before we cached the GPU device information.  We are now testing closer to actual user experience.

This marks a regression but the performance is still withing the bounds of other Mac devices.

I think WontFix is appropriate.