New issue
Advanced search Search tips

Issue 611237 link

Starred by 0 users

Issue metadata

Status: Duplicate
Owner:
Closed: May 2016
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

28.1%-265.5% regression in media.tough_video_cases at 381693:392695

Project Member Reported by sande...@chromium.org, May 11 2016

Issue description

Started after enabling this bot again ( issue 595665 ). It's just the one metric alerting though, there may not be much to find.
 
Project Member

Comment 2 by 42576172...@developer.gserviceaccount.com, May 12 2016

Mergedinto: 603244
Status: Duplicate (was: Assigned)

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


===== SUSPECTED CL(s) =====
Subject : Blacklist CALayer overlays on 10.10
Author  : ccameron
Commit description:
  
This is causing hangs on 10.10. Forcing fallback to the GL renderer
makes the issue go away according to people who have seen the
issue.

So far all reports of this issue are on the following dual-GPU
combination:
  iGPU: 0x8086, 0x0166
  dGPU: 0x10de, 0x0fd5
If this improves histograms, we will try to limit the scope of
the blacklist to just this combination.

BUG= 580616 

Review URL: https://codereview.chromium.org/1873283002

Cr-Commit-Position: refs/heads/master@{#386783}
Commit  : d5c58b5bfba2721be8681f67e784279e2030c423
Date    : Tue Apr 12 19:55:31 2016


===== TESTED REVISIONS =====
Revision         Mean   Std Dev  N  Good?
chromium@381693  182.0  1.58114  5  good
chromium@384444  357.2  8.75785  5  good
chromium@385823  202.6  1.34164  5  good
chromium@386507  176.0  18.1108  5  good
chromium@386680  175.8  13.3679  5  good
chromium@386775  128.6  3.91152  5  good
chromium@386779  126.4  7.89303  5  good
chromium@386781  124.4  10.6442  5  good
chromium@386782  130.0  3.937    5  good
chromium@386783  689.8  31.8857  5  bad    <--
chromium@386790  689.8  27.49    5  bad
chromium@386811  698.2  10.849   5  bad
chromium@386859  691.6  25.5597  5  bad
chromium@387193  690.4  11.5888  5  bad
chromium@392685  661.2  19.5243  5  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 611237

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests media.tough_video_cases
Test Metric: idle_wakeups_gpu/idle_wakeups_gpu
Relative Change: 263.30%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2115
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9012918375096719264


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5909741562232832

| 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 3 by 42576172...@developer.gserviceaccount.com, May 12 2016


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


===== SUSPECTED CL(s) =====
Subject : Blacklist CALayer overlays on 10.10
Author  : ccameron
Commit description:
  
This is causing hangs on 10.10. Forcing fallback to the GL renderer
makes the issue go away according to people who have seen the
issue.

So far all reports of this issue are on the following dual-GPU
combination:
  iGPU: 0x8086, 0x0166
  dGPU: 0x10de, 0x0fd5
If this improves histograms, we will try to limit the scope of
the blacklist to just this combination.

BUG= 580616 

Review URL: https://codereview.chromium.org/1873283002

Cr-Commit-Position: refs/heads/master@{#386783}
Commit  : d5c58b5bfba2721be8681f67e784279e2030c423
Date    : Tue Apr 12 19:55:31 2016


===== TESTED REVISIONS =====
Revision         Mean   Std Dev  N  Good?
chromium@381693  182.8  1.09545  5  good
chromium@384444  360.6  8.70632  5  good
chromium@385823  196.4  14.8088  5  good
chromium@386507  185.0  3.16228  5  good
chromium@386680  180.4  5.85662  5  good
chromium@386775  127.0  3.16228  5  good
chromium@386779  128.8  6.45755  5  good
chromium@386781  131.4  3.84708  5  good
chromium@386782  130.4  3.36155  5  good
chromium@386783  707.0  22.7596  5  bad    <--
chromium@386790  712.4  19.9324  5  bad
chromium@386811  711.0  17.9025  5  bad
chromium@386859  716.2  13.7368  5  bad
chromium@387193  688.0  29.6816  5  bad
chromium@392685  699.4  8.87694  5  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 611237

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests media.tough_video_cases
Test Metric: idle_wakeups_gpu/idle_wakeups_gpu
Relative Change: 282.60%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2116
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9012918296245087584


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5878412997033984

| 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