All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=663155

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgo_Hx7QsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg47ultgoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg4_elowkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg44fUvQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgk8aFpQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgo9O39goM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgk6H7tQkM


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

chromium-rel-win10
chromium-rel-win7-dual
chromium-rel-win7-x64-dual
chromium-rel-win8-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8996606633091606448

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

Hi flackr@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 : Enable compositing opaque fixed position elements which will maintain LCD text.
Author  : flackr
Commit description:
  
BUG= 661754 

Review-Url: https://codereview.chromium.org/2468323002
Cr-Commit-Position: refs/heads/master@{#429804}
Commit  : 5516863851ccc8158564461bb1c15af7f16436ac
Date    : Fri Nov 04 06:02:02 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev   N  Good?
chromium@429803  17.0128  0.570266  6  good
chromium@429804  23.7509  0.209034  4  bad    <--
chromium@429805  23.6253  0.157591  5  bad
chromium@429806  23.6462  0.163778  5  bad
chromium@429809  23.9635  0.263379  5  bad

Bisect job ran on: win_8_perf_bisect
Bug ID: 663155

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests smoothness.key_desktop_move_cases
Test Metric: mean_input_event_latency/mean_input_event_latency
Relative Change: 43.00%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/win_8_perf_bisect/builds/2274
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8996606633091606448


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5320913824776192

| 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!

+ssid, test owner.

@flacker: Was the regression in frame_times and latency expected for key_desktop_move cases?

The original CL was reverted and the graph returned to normal. I'll mark this as blocking re-enabling.

Changing type as this is not a current bug / regression, but needs to be tracked when relanding 5516863851ccc8158564461bb1c15af7f16436ac

It looks like gmail has changed a lot since this recording. In the recorded version they have several full-screen opaque fixed position elements. On low dpi these didn't used to be promoted but when promoted they cause promotion of several of the e-mail elements during the scroll due to overlap.

I think we should update the recording so that the test is representative of the current real world experience.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/8d097cf30a2dee37de4586de97f06269ed03da09

commit 8d097cf30a2dee37de4586de97f06269ed03da09
Author: Robert Flack <flackr@chromium.org>
Date: Fri Jun 30 16:42:01 2017

Update google_login helper and gmail recordings.

Support old and new google login fields. Update the gmail recordings used for
smoothness metrics. Removes the old custom login behavior in GmailSmoothPage.

Bug:  590766 , 663155 
Change-Id: I7c6d22b48e7cb16501dc393564d7e4df097490db
Reviewed-on: https://chromium-review.googlesource.com/546615
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Commit-Queue: Robert Flack <flackr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#483732}
[modify] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/data/key_desktop_move_cases.json
[add] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/data/key_desktop_move_cases_002.wpr.sha1
[add] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/data/top_25_006.wpr.sha1
[modify] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/data/top_25_smooth.json
[modify] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/key_desktop_move_cases.py
[modify] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/login_helpers/google_login.py
[modify] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/top_25_smooth.py
[modify] https://crrev.com/8d097cf30a2dee37de4586de97f06269ed03da09/tools/perf/page_sets/top_pages.py

With the up to date recording, we should no longer see the spike when enabling compositing opaque fixed position. While the old site regressing may be concerning, it's worth noting that fixed position promotion is default on high DPI devices, so they would have always had the slower performance.