All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=595906

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgoMngtwoM


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

chromium-rel-mac-hdd

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

Hi erikchen@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 : mac: Use IOSurfaces in Canvas2DLayerBridge.
Author  : erikchen
Commit description:
  
This CL adds the method prepareIOSurfaceMailboxFromImage(). By default, Canvas2D
now emits mailboxes that contain textures backed by IOSurfaces. This CL adds the
switch kDisable2dCanvasImageChromium, which disables this new behavior.

This CL allows webpages that have canvas 2D elements to enter the new
CoreAnimation compositing mode on OS X.

Overview of new behavior in prepareMailbox():
1. On prepareMailbox(), grab an SkImage snapshot of the underlying SKCanvas.
2. Copy the snapshot into a new IOSurface backed GL texture, and pass that to
prepareMailbox().
3. Discard the SkImage. Skia does not perform any internal copies.

BUG= 579664 
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Review URL: https://codereview.chromium.org/1752083003

Cr-Commit-Position: refs/heads/master@{#380670}
Commit  : f3d656148aa51e025094c93290d1fbf889064ca8
Date    : Fri Mar 11 18:18:20 2016


===== TESTED REVISIONS =====
Revision                Mean Value  Std. Dev.   Num Values  Good?
chromium@380439         272.0       5.070926    8           good
chromium@380589         279.125     8.724964    8           good
chromium@380664         272.875     7.120142    8           good
chromium@380669         272.375     10.676242   8           good
chromium@380670         309.2       9.364828    5           bad
chromium@380671         305.5       13.897585   8           bad
chromium@380672         314.2       4.086563    5           bad
chromium@380674         305.4       13.011533   5           bad
chromium@380683         303.8       16.709279   5           bad
chromium@380701         303.2       13.971399   5           bad
chromium@380738         303.8       9.833616    5           bad
chromium@381037         306.428571  7.412987    7           bad

Bisect job ran on: mac_hdd_perf_bisect
Bug ID: 595906

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --also-run-disabled-tests media.tough_video_cases
Test Metric: idle_wakeups_browser/idle_wakeups_browser
Relative Change: 11.93%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_hdd_perf_bisect/builds/453
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9017902394534148400


| 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 label Cr-Tests-AutoBisect.  Thank you!

We expect IOSurface Canvas2D to be a slight performance improvement in the lab, and a significant battery improvement in the wild.

I ran a bunch of tests on the try-bots, and determined that these metrics are really noisy:
https://codereview.chromium.org/1752083003/#msg19

For another example, notice that turning off this feature also causes an even larger regression in idle_wakeups_gpu/idle_wakeups_gpu:
https://bugs.chromium.org/p/chromium/issues/detail?id=595515

This is a WontFix.

 Issue 595515  has been merged into this issue.