All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=689190

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqIHUowoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqPORqwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqJnhswoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqK37vQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqP7YsAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6ITHswoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6IS9ogoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqJjSlAgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqOSjtAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqNGV5wgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgyLHxhggM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgyOOm4wsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqLDPsQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgyInGgwsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqPbU9goM


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

chromium-rel-mac-retina
chromium-rel-mac11
chromium-rel-mac11-air
chromium-rel-mac11-pro
chromium-rel-mac12
chromium-rel-mac12-mini-8gb
chromium-rel-win10
chromium-rel-win7-dual
chromium-rel-win7-gpu-intel
chromium-rel-win7-gpu-nvidia
chromium-rel-win7-x64-dual
chromium-rel-win8-dual
linux-release
win-high-dpi
win-zenbook

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8988372716673015424

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

Hi csharrison@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 : csharrison
  Commit : d86c35bcf47e34f09f295127796ea246746b5ab1
  Date   : Thu Feb 02 17:41:26 2017
  Subject: Make ResourceScheduler work in OOPIF

Bisect Details
  Configuration: win_8_perf_bisect
  Benchmark    : blink_perf.css
  Metric       : StyleSheetInsert-bootstrap/StyleSheetInsert-bootstrap
  Change       : 20.80% | 50915.8538173 -> 40326.7231837

Revision             Result                  N
chromium@447774      50915.9 +- 681.153      6      good
chromium@447790      51293.5 +- 527.273      6      good
chromium@447791      51268.3 +- 1654.5       6      good
chromium@447792      39972.4 +- 810.941      6      bad       <--
chromium@447794      39916.1 +- 847.529      6      bad
chromium@447798      39441.4 +- 1264.21      6      bad
chromium@447805      40386.1 +- 1052.54      6      bad
chromium@447836      40326.7 +- 607.638      6      bad

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

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8988372716673015424

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4974805675147264


| 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!

This is a real regression. My CL adds an IPC when a subframe inserts the body tag, and this benchmark has that in its critical path.

However, I don't think it is a big deal, as inserting the body tag in a tight loop is not very representative of any workload (as far as I know).

I have a CL in flight [1] to modify this benchmark accordingly.

https://codereview.chromium.org/2684483002/

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d68b4a106ecc41b8b16079a30ac422ad571554ed

commit d68b4a106ecc41b8b16079a30ac422ad571554ed
Author: csharrison <csharrison@chromium.org>
Date: Tue Feb 07 14:30:52 2017

Ensure StyleSheetInsert-bootstrap.html doesn't spam willInsertBody IPCs

A recent change to support ResourceScheduler in OOPIF mode [1] caused
a regression on this benchmark. The reason is that this benchmark
Essentially creates an empty iframe, populates it, and empties it again
in a tight loop.

This triggers us to send a WillInsertBody IPC for every loop iteration,
adding a lot of overhead to the benchmark.

As this idiom is likely not representative, this patch modifies the
benchmark to instead just insert and remove the style node in a tight
loop.

[1] https://codereview.chromium.org/2655393004

BUG= 689190 

Review-Url: https://codereview.chromium.org/2684483002
Cr-Commit-Position: refs/heads/master@{#448621}

[modify] https://crrev.com/d68b4a106ecc41b8b16079a30ac422ad571554ed/third_party/WebKit/PerformanceTests/CSS/StyleSheetInsert-bootstrap.html

Perf test now runs faster than it used to, and it more accurately describes css perf.