New issue
Advanced search Search tips

Issue 648887 link

Starred by 2 users

Issue metadata

Status: WontFix
Owner:
Closed: Oct 2016
Cc:
EstimatedDays: ----
NextAction: ----
OS: Mac
Pri: 2
Type: Bug-Regression



Sign in to add a comment

46.9% regression in smoothness.top_25_smooth at 418772:418827

Project Member Reported by toyoshim@chromium.org, Sep 21 2016

Issue description

See the link to graphs below.
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=648887

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg-bSTpQkM


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

chromium-rel-mac10
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Sep 21 2016


===== 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@418771  11.2391  2.15226  18  good
chromium@418827  11.4902  2.27392  18  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 5.17%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2391
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000922409228839536


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5787072500072448

| 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!
Started with a wider range including newer revisions.
Results look bipolarization. It this another bisect fails, this could be closed as WontFix.
Project Member

Comment 6 by 42576172...@developer.gserviceaccount.com, Sep 22 2016

Cc: hirosh...@chromium.org
Owner: hirosh...@chromium.org

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

Hi hiroshige@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 : Remove EventSender from HTMLLinkElement
Author  : hiroshige
Commit description:
  
This CL replaces EventSender in HTMLLinkElement with postTask():
Instead of calling HTMLLinkElement::dispatchPendingLoadEvents() in
Document::implicitClose() to enforce <link>'s onload events to be executed
before the document's onload event, this CL blocks document's onload until
<link>'s onload event using IncrementLoadEventDelayCount.

This CL also modifies SimTest:
This CL potentially makes onload events of <link> and documents async and
delayed, and thus SimTest can be destructed before document's onload event,
which cause assertion failure.
This CL calls testing::runPendingTasks() in ~SimTest() to flush scheduled
onload events to make Document to be loaded before SimTest's destruction.

BUG= 624697 

Review-Url: https://codereview.chromium.org/2275493002
Cr-Commit-Position: refs/heads/master@{#418835}
Commit  : aaf5ece46ff3b24c2f5ff0054600900f60c70ce4
Date    : Thu Sep 15 11:20:43 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev   N   Good?
chromium@418771  11.5174  2.10551   12  good
chromium@418817  11.6764  2.43881   17  good
chromium@418829  10.2833  0.702408  12  good
chromium@418832  11.8718  2.68444   12  good
chromium@418834  11.8391  2.79939   12  good
chromium@418835  15.8472  0.717976  12  bad    <--
chromium@418840  15.4302  0.570387  18  bad
chromium@418862  15.7811  0.886067  18  bad
chromium@418952  16.5844  0.926594  5   bad
chromium@419133  15.8606  0.612041  5   bad
chromium@419494  15.0692  0.91846   5   bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 34.65%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2394
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000841448402391152


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5828568326602752

| 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!
Project Member

Comment 10 by 42576172...@developer.gserviceaccount.com, Sep 29 2016


===== 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@418771  12.5377  1.95749  18  good
chromium@418827  13.5505  2.47393  18  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 7.63%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2404
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000206162252883248


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5263236746706944

| 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!
Project Member

Comment 11 by 42576172...@developer.gserviceaccount.com, Sep 29 2016


===== 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@418501  11.594   2.49325   18  good
chromium@418541  10.2249  0.456181  18  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 14.98%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2403
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000206170806337424


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6166159802499072

| 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!
Project Member

Comment 12 by 42576172...@developer.gserviceaccount.com, Sep 29 2016


===== 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@418771  24.2082  11.465   18  good
chromium@418827  17.3556  8.5045   18  bad

Bisect job ran on: mac_10_11_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 34.80%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_11_perf_bisect/builds/927
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000206147980004912


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5801042720063488

| 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!
Project Member

Comment 15 by 42576172...@developer.gserviceaccount.com, Sep 30 2016


===== 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@418771  23.2001  13.92    18  good
chromium@419494  27.7266  12.0647  18  bad

Bisect job ran on: mac_10_11_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 21.58%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_11_perf_bisect/builds/931
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000115735658720688


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6683606590685184

| 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!

===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev   N   Good?
chromium@418771  11.6723  2.68605   18  good
chromium@418817  11.0471  2.22592   27  good
chromium@418829  12.9706  2.96007   27  bad
chromium@418840  15.5339  1.20773   32  bad
chromium@418862  15.9007  0.661251  9   bad
chromium@418952  15.1567  1.38587   18  bad
chromium@419133  15.5952  0.948314  18  bad
chromium@419494  15.2757  0.866014  18  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 23.06%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2407
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000115745889389040


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5301604595531776

| 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!
Hmm, this looks like Mac10.10-only. Mac10.11 bots doesn't show clear regressions.
Labels: OS-Mac

===== 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@418829  25.2457  14.1794  18  good
chromium@418840  27.8166  14.0574  18  bad

Bisect job ran on: mac_10_11_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 12.77%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_11_perf_bisect/builds/940
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8999832066753076224


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5770926140948480

| 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!

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


===== SUSPECTED CL(s) =====
Subject : Remove EventSender from HTMLLinkElement
Author  : hiroshige
Commit description:
  
This CL replaces EventSender in HTMLLinkElement with postTask():
Instead of calling HTMLLinkElement::dispatchPendingLoadEvents() in
Document::implicitClose() to enforce <link>'s onload events to be executed
before the document's onload event, this CL blocks document's onload until
<link>'s onload event using IncrementLoadEventDelayCount.

This CL also modifies SimTest:
This CL potentially makes onload events of <link> and documents async and
delayed, and thus SimTest can be destructed before document's onload event,
which cause assertion failure.
This CL calls testing::runPendingTasks() in ~SimTest() to flush scheduled
onload events to make Document to be loaded before SimTest's destruction.

BUG= 624697 

Review-Url: https://codereview.chromium.org/2275493002
Cr-Commit-Position: refs/heads/master@{#418835}
Commit  : aaf5ece46ff3b24c2f5ff0054600900f60c70ce4
Date    : Thu Sep 15 11:20:43 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N   Good?
chromium@418829  11.3456  2.36115  12  good
chromium@418832  11.0443  2.54458  12  good
chromium@418834  11.4926  2.63782  8   good
chromium@418835  15.4324  1.87128  12  bad    <--
chromium@418840  15.5394  1.17594  8   bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 648887

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth
Test Metric: mean_input_event_latency/Blogger
Relative Change: 32.10%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2414
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8999832077780207360


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5902180341514240

| 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!
Status: WontFix (was: Assigned)
This has recovered.
Project Member

Comment 24 by bugdroid1@chromium.org, Jan 20 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/46649a2f0bf126330d95f47b6c8773823451d23f

commit 46649a2f0bf126330d95f47b6c8773823451d23f
Author: hiroshige <hiroshige@chromium.org>
Date: Fri Jan 20 03:18:25 2017

Call checkCompleted() synchronously in HTMLLinkElement::dispatchPendingEvent()

Previously, there was an asynchronous hop from <link>'s load event to
document's load event, because HTMLLinkElement::dispatchPendingEvent()
scheduled checkCompleted() asynchronously via IncrementLoadEventDelayCount.

This CL makes HTMLLinkElement::dispatchPendingEvent() to call checkCompleted()
synchronously, by using
IncrementLoadEventDelayCount::clearAndCheckLoadEvent().

Because checkCompleted() is called at the end of asynchronously called
HTMLLinkElement::dispatchPendingEvent(), this CL doesn't increase synchronous
timing dependencies around Document load events.

This CL fixes the performance regression of smoothness.top_25_smooth on
Mac 10.10 caused by https://codereview.chromium.org/2275493002/.

BUG= 648887 ,  624697 

Review-Url: https://codereview.chromium.org/2391923002
Cr-Commit-Position: refs/heads/master@{#444965}

[modify] https://crrev.com/46649a2f0bf126330d95f47b6c8773823451d23f/third_party/WebKit/Source/core/html/HTMLLinkElement.cpp
[modify] https://crrev.com/46649a2f0bf126330d95f47b6c8773823451d23f/third_party/WebKit/Source/core/html/HTMLLinkElement.h

Project Member

Comment 26 by 42576172...@developer.gserviceaccount.com, Apr 11 2017


=== BISECT JOB RESULTS ===
Perf regression found but unable to narrow commit range

Build failures prevented the bisect from narrowing the range further.


Bisect Details
  Configuration: mac_10_10_perf_bisect
  Benchmark    : smoothness.top_25_smooth
  Metric       : mean_input_event_latency/Blogger

Suspected Commit Range
  1 commits in range
  https://chromium.googlesource.com/chromium/src/+log/b75661d3d5c5924cffc942945d79d0430b812e1c..eddc51a943211fc44d93d05ae1f84d79d12619ed


Revision             Result                   N
chromium@418771      11.6723 +- 2.68605       18      good
chromium@418817      11.0471 +- 2.22592       27      good
chromium@418829      12.9706 +- 2.96007       27      bad
chromium@418840      15.5339 +- 1.20773       32      bad
chromium@418862      15.9007 +- 0.661251      9       bad
chromium@418952      15.1567 +- 1.38587       18      bad
chromium@419133      15.5952 +- 0.948314      18      bad
chromium@419494      15.2757 +- 0.866014      18      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.top_25_smooth

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8982624142538590480

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5301604595531776


| 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!

Sign in to add a comment