Issue metadata
Sign in to add a comment
|
1.3%-875% regression in media.android.tough_video_cases at 455708:455885 |
||||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Mar 10 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8985467808649859376
,
Mar 11 2017
=== BISECT JOB RESULTS === NO Perf regression found Bisect Details Configuration: android_nexus5X_perf_bisect Benchmark : media.android.tough_video_cases Metric : vm_shared_dirty_delta/crowd1080_vp9.webm_browser Revision Result N chromium@455776 -96.8571 +- 776.802 21 good chromium@455844 -142.476 +- 1342.1 21 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 media.android.tough_video_cases Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8985467808649859376 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=6695854460633088 | 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!
,
Mar 13 2017
Frank, I notice these graphs follow a pattern of restoring the metric to a previous high value after all temporarily dropping down for a few hundred revisions. I'll bisect the initial dip just for curiosity, but I suspect its either a transient issue (improvement?) or a commit that landed/reverted.
,
Mar 13 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8985214568514745744
,
Mar 14 2017
=== BISECT JOB RESULTS === Perf regression found with culprit Suspected Commit Author : pdr Commit : a20ad732333dc9db2b545886acd16a11abdc5842 Date : Tue Mar 07 17:11:15 2017 Subject: Remove indirection: setup scrollbar scroll layers in the scrollbar constructor Bisect Details Configuration: android_nexus5X_perf_bisect Benchmark : media.android.tough_video_cases Metric : vm_shared_dirty_delta/tulip2.ogv_browser Change : 127.27% | -22.0 -> -50.0 Revision Result N chromium@455068 -22.0 +- 20.199 6 good chromium@455091 -3.33333 +- 30.6812 6 good chromium@455102 8.0 +- 13.8564 6 good chromium@455104 8.0 +- 13.8564 6 good chromium@455105 -34.6667 +- 14.6059 6 bad <-- chromium@455108 -26.0 +- 14.6969 6 bad chromium@455113 -50.0 +- 14.6969 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 media.android.tough_video_cases Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8985214568514745744 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5341580405243904 | 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 |
|||||||||||||||||||||||
Comment 1 by liberato@google.com
, Mar 10 2017