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

Issue 641347 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Nov 2016
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

76.3%-131.9% regression in smoothness.scrolling_tough_ad_cases at 414469:414544

Project Member Reported by mustaq@chromium.org, Aug 26 2016

Issue description

See the link to graphs below.
 
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Aug 26 2016

Cc: sahel@chromium.org
Owner: sahel@chromium.org

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

Hi sahel@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 : Touchpad scroll latching enabled for Mac behind flag.
Author  : sahel
Commit description:
  
To test it, use the --enable-features=TouchpadScrollLatching runtime flag.

BUG=526463
TEST=MouseWheelEventQueueTests/MouseWheelEventQueueTest.GestureSendingWithPhaseInformation/[0/1]
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Review-Url: https://codereview.chromium.org/2256733003
Cr-Commit-Position: refs/heads/master@{#414515}
Commit  : 7adc4a771383af81e2411e755bdce689c803b939
Date    : Thu Aug 25 19:56:17 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@414503  21.2829  8.96943  8  good
chromium@414509  14.205   2.07178  5  good
chromium@414512  15.5725  2.03145  5  good
chromium@414514  16.3376  2.6924   5  good
chromium@414515  34.3123  9.65123  8  bad    <--

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 641347

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.scrolling_tough_ad_cases
Test Metric: first_gesture_scroll_update_latency/http___www.theguardian.com_uk
Relative Change: 71.74%
Score: 99.8

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/4058
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9003254832617643056


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5901846359572480

| 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 4 by mustaq@chromium.org, Aug 26 2016

Cc: tdres...@chromium.org
Tim, please check if the regression is expected and/or transient.
This was not expected. The behavior is currently behind a flag, so that patch shouldn't have any performance impact.

Did we accidentally start sending more events than we should have?

Sahel, can you take a look at this? You'll probably want to come up with a local reproduction of the regression.
Labels: BadBisect
Owner: robert...@chromium.org
Status: Started (was: Assigned)
The signal is quite noisy, noice how the std dev is pretty large for the reference range.

Revision       Values         
chromium@414503[11.703800000000001, 39.7279, 15.169900000000002, 25.673400000000004, 16.723, 17.057, 17.948600000000003, 26.2592]
chromium@414504[]             
chromium@414505[]             
chromium@414506[]             
chromium@414507[]             
chromium@414508[]             
chromium@414509[15.838399999999998, 16.3587, 11.714800000000002, 12.369000000000003, 14.744200000000001]
chromium@414510[]             
chromium@414511[]             
chromium@414512[15.814599999999999, 15.774599999999998, 13.9058, 18.721799999999998, 13.645900000000001]
chromium@414513[]             
chromium@414514[18.333000000000002, 16.808299999999996, 18.7275, 15.8188, 12.000300000000001]
chromium@414515[27.043100000000003, 43.299299999999995, 45.724, 27.381099999999996, 43.7417, 39.738800000000005, 24.0506, 23.5195]
LKGR values: [15.814599999999999, 15.774599999999998, 13.9058, 18.721799999999998, 13.645900000000001]
FKBR values: [27.043100000000003, 43.299299999999995, 45.724, 27.381099999999996, 43.7417, 39.738800000000005, 24.0506, 23.5195]

I'll trigger a bisect on a different metric

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


===== SUSPECTED CL(s) =====
Subject : Touchpad scroll latching enabled for Mac behind flag.
Author  : sahel
Commit description:
  
To test it, use the --enable-features=TouchpadScrollLatching runtime flag.

BUG=526463
TEST=MouseWheelEventQueueTests/MouseWheelEventQueueTest.GestureSendingWithPhaseInformation/[0/1]
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Review-Url: https://codereview.chromium.org/2256733003
Cr-Commit-Position: refs/heads/master@{#414515}
Commit  : 7adc4a771383af81e2411e755bdce689c803b939
Date    : Thu Aug 25 19:56:17 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@414503  15.1579  2.31253  5  good
chromium@414514  12.3009  2.03629  5  good
chromium@414515  28.2984  4.73385  5  bad    <--
chromium@414516  28.1944  4.12822  5  bad
chromium@414517  28.3643  3.97521  5  bad
chromium@414519  29.0004  2.89559  5  bad
chromium@414524  29.3513  4.704    5  bad
chromium@414544  27.4086  6.12422  5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 641347

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.scrolling_tough_ad_cases
Test Metric: first_gesture_scroll_update_latency/http___www.theguardian.com_uk
Relative Change: 80.82%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/2100
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9002039796981008816


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5320716772179968

| 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: robert...@chromium.org
Labels: -BadBisect
Owner: sahel@chromium.org
Status: Assigned (was: Started)
I guess bisect seems to be consistent about finding sahel's patch as culprit.

sahel@ do you need any help reproducing this locally?

Comment 10 by sahel@chromium.org, Sep 12 2016

I thought it will be solved after submitting this patch which is related to always scrolling on main thread. I'll check to see if it still exists or not, and work on it if the regression still exists.

https://codereview.chromium.org/2290233004
Project Member

Comment 15 by 42576172...@developer.gserviceaccount.com, Sep 20 2016


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


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@414514  17.9044  2.7716   5  bad

Bisect job ran on: android_nexus5X_perf_bisect
Bug ID: 641347

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.scrolling_tough_ad_cases
Test Metric: first_gesture_scroll_update_latency/http___www.theguardian.com_uk
Relative Change: None
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5X_perf_bisect/builds/685
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000970211488258656


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6374407373586432

| 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 16 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@414503  16.3078  2.48955  18  good
chromium@414514  15.7908  1.84028  18  bad

Bisect job ran on: android_nexus5X_perf_bisect
Bug ID: 641347

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.scrolling_tough_ad_cases
Test Metric: first_gesture_scroll_update_latency/http___www.theguardian.com_uk
Relative Change: 6.40%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5X_perf_bisect/builds/686
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000969301563387328


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5878970011090944

| 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 17 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@414470  16.0053  2.27685  18  good
chromium@414514  15.7024  1.71211  18  bad

Bisect job ran on: android_nexus5X_perf_bisect
Bug ID: 641347

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.scrolling_tough_ad_cases
Test Metric: first_gesture_scroll_update_latency/http___www.theguardian.com_uk
Relative Change: 9.86%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5X_perf_bisect/builds/684
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000970647620283968


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5327590968000512

| 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 18 by 42576172...@developer.gserviceaccount.com, Sep 21 2016


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


===== SUSPECTED CL(s) =====
Subject : Touchpad scroll latching enabled for Mac behind flag.
Author  : sahel
Commit description:
  
To test it, use the --enable-features=TouchpadScrollLatching runtime flag.

BUG=526463
TEST=MouseWheelEventQueueTests/MouseWheelEventQueueTest.GestureSendingWithPhaseInformation/[0/1]
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Review-Url: https://codereview.chromium.org/2256733003
Cr-Commit-Position: refs/heads/master@{#414515}
Commit  : 7adc4a771383af81e2411e755bdce689c803b939
Date    : Thu Aug 25 19:56:17 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@414375  16.0187  2.96182  5  good
chromium@414469  15.9559  2.55503  4  good
chromium@414503  14.7365  1.98161  5  good
chromium@414511  16.9836  1.8101   4  good
chromium@414514  15.8368  2.89201  5  good
chromium@414515  27.9323  2.10426  5  bad    <--
chromium@414516  27.0231  2.1825   5  bad
chromium@414517  27.6523  2.7458   5  bad
chromium@414530  26.4919  1.62663  5  bad

Bisect job ran on: android_nexus5X_perf_bisect
Bug ID: 641347

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.scrolling_tough_ad_cases
Test Metric: first_gesture_scroll_update_latency/http___www.theguardian.com_uk
Relative Change: 65.38%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5X_perf_bisect/builds/683
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000971718474928688


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5895675311554560

| 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: Started (was: Assigned)
sahel says
I have a fix, I am testing to see if the perf regression goes away or not.
The blamed cl is reverted in m54 though
Status: Fixed (was: Started)
The blamed patch is reverted.

https://codereview.chromium.org/2467023003

Sign in to add a comment