New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 715279 link

Starred by 2 users

Issue metadata

Status: Fixed
Owner:
Closed: Apr 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

2.8% regression in smoothness.sync_scroll.key_mobile_sites_smooth at 466669:466722

Project Member Reported by benhenry@chromium.org, Apr 25 2017

Issue description

See the link to graphs below.
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=715279

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgkonMvgoM


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

android-nexus6
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Apr 25 2017

Cc: skobes@chromium.org
Owner: skobes@chromium.org

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

Hi skobes@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 : skobes
  Commit : 8a4d4531daca4b9a1f6dc1224bcb7134dc053ff4
  Date   : Mon Apr 24 19:19:10 2017
  Subject: Temporarily ignore ScrollbarAnimationController::DidRequestShowFromMainThread.

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : smoothness.sync_scroll.key_mobile_sites_smooth
  Metric       : frame_times/http___mobile-news.sandbox.google.com_news_pt0
  Change       : 2.52% | 16.885299073 -> 17.3103218039

Revision             Result                    N
chromium@466668      16.8853 +- 0.0157128      6       good
chromium@466695      16.9276 +- 0.148251       6       good
chromium@466709      16.9958 +- 0.323017       9       good
chromium@466711      16.8967 +- 0.129284       6       good
chromium@466712      17.2389 +- 0.325893       6       bad       <--
chromium@466713      17.3366 +- 0.758714       14      bad
chromium@466716      17.2376 +- 0.537934       9       bad
chromium@466722      17.3103 +- 0.17814        6       bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...mobile.news.sandbox.google.com.news.pt0 smoothness.sync_scroll.key_mobile_sites_smooth

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981306063119974512

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5770832720166912


| 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 Speed>Bisection.  Thank you!
Labels: Performance-Responsiveness
Status: Assigned (was: Untriaged)

Comment 5 by skobes@chromium.org, Apr 26 2017

Cc: aelias@chromium.org
@aelias, do you think this is just noise?  r466712 should have restored the behavior we had before r460178 to address a different metric in key_mobile_sites_smooth ( issue 712453 ).

If these tests are that unpredictable I wonder if we should not even bother with the viewport size workarounds in http://crrev.com/2834393003.

Comment 6 by aelias@chromium.org, Apr 26 2017

Status: WontFix (was: Assigned)
I don't think it's just noise, the bots only succeed in exact bisect when there's a signal.  I think your change made one one metric better and another worse.  The graph history doesn't bear out your theory that this restores behavior as of before r460178.  However this particular regression is mild enough we can probably overlook it.
Project Member

Comment 7 by 42576172...@developer.gserviceaccount.com, Apr 26 2017

Cc: jgruber@chromium.org
 Issue 715103  has been merged into this issue.
Project Member

Comment 8 by 42576172...@developer.gserviceaccount.com, Apr 26 2017

 Issue 715103  has been merged into this issue.
Which metric has improved? Looks like 500us regression, so I'd like to understand the tradeoff you made when determining to wontfix this.
Project Member

Comment 10 by 42576172...@developer.gserviceaccount.com, Apr 26 2017

 Issue 715103  has been merged into this issue.
Status: Assigned (was: WontFix)
I'm reverting r466712 in http://crrev.com/2834393003.
Project Member

Comment 12 by bugdroid1@chromium.org, Apr 26 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/66aacfa671360982f234c0fbd3b186efe21ee763

commit 66aacfa671360982f234c0fbd3b186efe21ee763
Author: skobes <skobes@chromium.org>
Date: Wed Apr 26 23:40:39 2017

Re-enable ScrollbarAnimationController::DidRequestShowFromMainThread.

Disabling it didn't fix  http://crbug.com/706927 , but caused a new regression in
 http://crbug.com/715279 , while  http://crbug.com/712453  recovered for some other
reason.

BUG= 606395 ,  706927 ,  712453 ,  715279 
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2834393003
Cr-Commit-Position: refs/heads/master@{#467513}

[modify] https://crrev.com/66aacfa671360982f234c0fbd3b186efe21ee763/cc/input/scrollbar_animation_controller.cc

Project Member

Comment 13 by 42576172...@developer.gserviceaccount.com, Apr 27 2017

Issue 715269 has been merged into this issue.
Project Member

Comment 14 by 42576172...@developer.gserviceaccount.com, Apr 27 2017

 Issue 715273  has been merged into this issue.
Status: Fixed (was: Assigned)

Sign in to add a comment