All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=627083

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgst69qwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg8sD0pQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgstXouAoM


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

chromium-rel-win7-dual
chromium-rel-win7-single
linux-release

Kicked off another bisect.

I've put CL https://codereview.chromium.org/2142963002/ up for review to hopefully address this. ( A perf tryjob is also running here: https://codereview.chromium.org/2141063002/)

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

Hi henrika@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 : Fixes crash in AudioManagerMac at shutdown.
Author  : henrika
Commit description:
  
Takes a copy of the list of input streams before deleting the stream.
The reason is that when a stream is deleted it also removes itself from the list and we are iterating over that same list. Should be resolved by taking a copy of the list first.

BUG=626512

Review-Url: https://codereview.chromium.org/2133703002
Cr-Commit-Position: refs/heads/master@{#404349}
Commit  : fe1a694e498267bfa2597f5db820148f4eb38bd6
Date    : Fri Jul 08 14:06:17 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev    N  Good?
chromium@404341  22.0615  0.165524   5  good
chromium@404348  21.9437  0.185952   5  good
chromium@404349  26.509   0.227235   5  bad    <--
chromium@404350  23.4184  0.163524   5  bad
chromium@404352  24.3282  0.213712   5  bad
chromium@404355  24.204   0.11273    5  bad
chromium@404368  22.9715  0.122328   5  bad
chromium@404387  23.0599  0.0983284  5  bad

Bisect job ran on: win_perf_bisect
Bug ID: 627083

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests scheduler.tough_scheduling_cases
Test Metric: mean_input_event_latency/mean_input_event_latency
Relative Change: 4.53%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/win_perf_bisect/builds/6686
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9007336522640487936


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5789877744959488

| 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!

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/799f1ac42b56c604bf24c6e603bccebae3026d61

commit 799f1ac42b56c604bf24c6e603bccebae3026d61
Author: dtapuska <dtapuska@chromium.org>
Date: Wed Jul 13 01:38:45 2016

Add passive: false to touch_scheduling_cases to avoid interventions.

The page sets are specifically tracking prevent defaulting events
and causing scrolling on the main thread. We should set them as
passive: false to avoid the root scroller intervention.

BUG= 627083 

Review-Url: https://codereview.chromium.org/2142963002
Cr-Commit-Position: refs/heads/master@{#404880}

[modify] https://crrev.com/799f1ac42b56c604bf24c6e603bccebae3026d61/tools/perf/page_sets/tough_scheduling_cases/raf_touch_animation.html
[modify] https://crrev.com/799f1ac42b56c604bf24c6e603bccebae3026d61/tools/perf/page_sets/tough_scheduling_cases/touch_handler_scrolling.html

My change mentioned in #7 can not be causing this since it only touches Mac and Linux. Removing myself as owner. Seems to be be picked up by dtapuska@.