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

Issue 677423 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner: ----
Closed: Aug 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

1.7%-3.6% regression in smoothness.sync_scroll.key_mobile_sites_smooth at 439532:439603

Project Member Reported by alexclarke@chromium.org, Dec 29 2016

Issue description

See the link to graphs below.
 
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Dec 29 2016

Cc: yangguo@chromium.org
Owner: yangguo@chromium.org

=== PERF REGRESSION ===


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

Hi yangguo@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


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


===== SUSPECTED CL(s) =====
Subject : Fix memory leak in logging-unittest.
Author  : yangguo
Commit description:
  
R=cbruni@chromium.org
BUG= chromium:662388 

Review-Url: https://codereview.chromium.org/2586203002
Cr-Commit-Position: refs/heads/master@{#41799}
Commit  : 2a19ad3603f355c0069f15840ec3d7b2c53fe414
Date    : Mon Dec 19 11:58:28 2016


===== TESTED REVISIONS =====
Revision                       Mean     Std Dev  N     Good?
chromium@439535                17.0598  118.617  1664  good
chromium@439568                17.0934  117.303  1658  good
chromium@439571                17.1737  148.539  1652  good
chromium@439571,v8@81dd9847cf  17.0938  143.862  1659  good
chromium@439571,v8@52702e55aa  17.29    169.073  1646  good
chromium@439571,v8@2a19ad3603  17.339   190.189  1650  bad    <--
chromium@439571,v8@8ac9e55aa6  17.3877  200.542  1650  bad
chromium@439572                17.4415  210.48   1647  bad
chromium@439573                17.2786  201.807  1654  bad
chromium@439577                17.3371  186.052  1649  bad
chromium@439585                17.4328  208.594  1646  bad
chromium@439601                17.4671  224.497  1641  bad

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 677423

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.wowwiki.com.World.of.Warcraft..Mists.of.Pandaria smoothness.sync_scroll.key_mobile_sites_smooth
Test Metric: frame_times/http___www.wowwiki.com_World_of_Warcraft__Mists_of_Pandaria
Relative Change: 2.39%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/4487
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8991948643677161792


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4504658268127232

| 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!
Cc: -yangguo@chromium.org
Owner: alexclarke@chromium.org
Very unlikely. The bisected CL only affects a unit test, and not production code.
I agree the bisect probably found the wrong patch.  The graph is super noisy but there does seem to be a regression, trying again.
Project Member

Comment 9 by 42576172...@developer.gserviceaccount.com, Dec 30 2016


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


=== Bisection aborted ===
The bisect was aborted because Bisect cannot identify a culprit: Bisect failed to reproduce the regression with enough confidence.
Please contact the the team (see below) if you believe this is in error.

===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N     Good?
chromium@439500  17.093   75.4622  1386  good
chromium@439650  17.3777  193.518  1362  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 677423

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Wikipedia..1.tab. smoothness.sync_scroll.key_mobile_sites_smooth
Test Metric: frame_times/Wikipedia (1 tab)
Relative Change: 1.67%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2860
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8991857759964853152


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5274680070504448

| 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, Dec 30 2016

Cc: wenj...@amazon.com
Owner: wenj...@amazon.com

=== PERF REGRESSION ===


=== Auto-CCing suspected CL author wenjinm@amazon.com ===

Hi wenjinm@amazon.com, the bisect results pointed to your CL, please take a look at the
results.


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


===== SUSPECTED CL(s) =====
Subject : Allow deferred intent when the screen is locked
Author  : wenjinm
Commit description:
  
The intent might be ignored if the screen is locked and user just input
pin to unlock it. Instead of simply discard the intent, add a deferred
intent and hold it for a short period of time. Once the USER_PRESENT has
been broadcast, check back the deferred intent and move forward with it.

The other option to handle this scenario would be set the intent aside,
and start an async task to continuously for the screen status, and react
once the screen unlocked. This would be less optimal option since it
requires extra background task while screen is locked.

BUG=455126
TEST=out/Default/bin/run_chrome_public_test_apk -f \*IntentHandlerTest\*
     also verified by file an intent when screen is locked and then
     unlcok screen with pin.

Review-Url: https://codereview.chromium.org/2568843003
Cr-Commit-Position: refs/heads/master@{#439566}
Commit  : 8dec237dce0b0430a9f9c5d7f50bce4943be5554
Date    : Mon Dec 19 21:23:55 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N     Good?
chromium@439500  16.9208  50.6722  1399  good
chromium@439538  16.9633  58.4082  1396  good
chromium@439557  16.9999  64.9276  1396  good
chromium@439562  16.9737  64.9616  1391  good
chromium@439564  16.9721  64.9561  1395  good
chromium@439565  16.9603  63.303   1396  good
chromium@439566  16.942   53.0547  1398  bad    <--
chromium@439575  17.2644  205.578  1370  bad
chromium@439650  17.2072  179.076  1374  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 677423

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Wikipedia..1.tab....delayed.scroll.start smoothness.sync_scroll.key_mobile_sites_smooth
Test Metric: frame_times/Wikipedia (1 tab) - delayed scroll start
Relative Change: 1.69%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2859
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8991857776285011680


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6345316364713984

| 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, Dec 30 2016

Cc: trchen@chromium.org
Owner: trchen@chromium.org

=== PERF REGRESSION ===


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

Hi trchen@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


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


===== SUSPECTED CL(s) =====
Subject : [SPv2] Use PassRefPtr in PropertyTreeState setter and remove meaningless DCHECK
Author  : trchen
Commit description:
  
The common pattern is to accept PassRefPtr in setters because it will allow
zero cost ownership transfer. Also the DCHECK for non-unique ownership is
removed, because it never worked as intended.
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2572423002
Cr-Commit-Position: refs/heads/master@{#439626}
Commit  : 078212bbd956d7d72333d75ca3c2d2da5fff1190
Date    : Tue Dec 20 00:00:53 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N     Good?
chromium@439000  17.044   126.651  1665  good
chromium@439500  17.1738  150.296  1651  good
chromium@439625  17.2179  179.692  1658  good
chromium@439626  17.4423  237.023  1648  bad    <--
chromium@439627  17.3405  177.418  1648  bad
chromium@439629  17.3661  193.872  1651  bad
chromium@439633  17.3668  207.002  1650  bad
chromium@439641  17.4196  201.324  1644  bad
chromium@439657  17.3974  213.56   1642  bad
chromium@439688  17.5038  225.054  1646  bad
chromium@439750  17.5197  234.507  1639  bad
chromium@440000  17.3542  171.801  1646  bad

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 677423

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.wowwiki.com.World.of.Warcraft..Mists.of.Pandaria smoothness.sync_scroll.key_mobile_sites_smooth
Test Metric: frame_times/http___www.wowwiki.com_World_of_Warcraft__Mists_of_Pandaria
Relative Change: 1.82%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/4492
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8991857799303614464


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5851328083918848

| 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!

Comment 12 by wenj...@amazon.com, Dec 30 2016

The CL for BUG=455126 should not impact scroll performance. It only takes effect when an intent was fired when the screen was locked or off.
Status: Assigned (was: Untriaged)
Owner: ----
Status: WontFix (was: Assigned)
Bisect identified 3 different culprits, this regression is too far into the noise to make progress.

Sign in to add a comment