New issue
Advanced search Search tips

Issue 617469 link

Starred by 0 users

Issue metadata

Status: Duplicate
Merged: issue 617427
Owner:
Closed: Jun 2016
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

29.7% regression in tab_switching.top_10 at 397797:397815

Project Member Reported by rmcilroy@chromium.org, Jun 5 2016

Issue description

See the link to graphs below.
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=617469

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg3PzTowkM


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

chromium-rel-mac10

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


=== Bisection aborted ===
The bisect was aborted because The metric values for the initial "good" and "bad" revisions do not represent a clear regression.
Please contact the the team (see below) if you believe this is in error.

=== Warnings ===
The following warnings were raised by the bisect job:

 * Bisect failed to reproduce the regression with enough confidence.

===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N   Good?
chromium@397796  1333.33  106.914  18  good
chromium@397815  1345.33  84.0308  18  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 617469

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests tab_switching.top_10
Test Metric: idle_wakeups_total/idle_wakeups_total
Relative Change: 9.49%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2132
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9010662184513112080


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5269365243510784

| 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!
Mergedinto: 617427
Status: Duplicate (was: Assigned)

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


===== SUSPECTED CL(s) =====
Subject : mac: ensure ui::Compositor exists for visible RWHVs
Author  : enne
Commit description:
  
RenderWidgetHostViewMac is careful about creating ui::Compositor and so
currently only does this in Show and SwapCompositorFrame.  However, if
it the RenderWidgetHostView is created while the RenderWidgetHost is
visible, then Show will early out and never create a ui::Compositor.

This causes problems with begin frame scheduling, which wants to have
the Display (which is created as part of creating the ui::Compositor's
output surface currently) own the real begin frame source that drives
everything.  In that world, if no ui::Compositor exists, no frames ever
are sent or swapped.  So, the ui::Compositor needs to be created first.

Also, if the ui::Compositor is correctly created during Show (or the
constructor), then there's no need to ensure that it exists during
swap compositor frame, and so that can be safely removed.

R=asvitkine@chromium.org,ccameron@chromium.org

Review-Url: https://codereview.chromium.org/2005013002
Cr-Commit-Position: refs/heads/master@{#397755}
Commit  : 9af6c23da7070f9423d5cf25856ed27f7d5d8e5e
Date    : Fri Jun 03 18:13:13 2016


===== TESTED REVISIONS =====
Revision         Mean    Std Dev  N  Good?
chromium@397752  1037.8  27.6803  5  good
chromium@397754  1033.8  30.244   5  good
chromium@397755  1377.2  29.9449  5  bad    <--
chromium@397756  1291.0  109.474  5  bad
chromium@397760  1263.4  115.556  5  bad
chromium@397768  1302.4  123.269  5  bad
chromium@397784  1390.0  31.3209  5  bad
chromium@397815  1306.6  109.96   5  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 617469

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests tab_switching.top_10
Test Metric: idle_wakeups_total/idle_wakeups_total
Relative Change: 25.90%
Score: 99.0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2133
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9010662132027855904


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5839684958158848

| 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!

Sign in to add a comment