Issue metadata
Sign in to add a comment
|
2.2% regression in thread_times.key_idle_power_cases at 447312:447361 |
||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Feb 8 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8988183393451960320
,
Feb 9 2017
=== Auto-CCing suspected CL author ericrk@chromium.org === Hi ericrk@chromium.org, the bisect results pointed to your CL, please take a look at the results. === BISECT JOB RESULTS === Perf regression found with culprit Suspected Commit Author : ericrk Commit : 7f6a27f1dbe4d8cafffb919445bc93bb8867a8fe Date : Tue Jan 31 22:34:32 2017 Subject: Delay activation/draw on GPU tile work completion Bisect Details Configuration: android_nexus6_perf_bisect Benchmark : thread_times.key_idle_power_cases Metric : tasks_per_second_total_all/css-animation.html Change : 2.15% | 1973.89336902 -> 2016.24798899 Revision Result N chromium@447311 1973.89 +- 8.35926 6 good chromium@447336 1973.78 +- 10.3792 6 good chromium@447343 1973.22 +- 9.09666 6 good chromium@447345 1972.27 +- 8.20581 6 good chromium@447346 2016.14 +- 4.44534 6 bad <-- chromium@447349 2013.22 +- 6.26504 6 bad chromium@447361 2016.25 +- 6.88581 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 --story-filter=css.animation.html thread_times.key_idle_power_cases Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8988183393451960320 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5819984237821952 | 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!
,
Feb 10 2017
Eric, I'm assuming the increase in tasks run is expected. There was also a small increase in CPU usage. From https://chromeperf.appspot.com/group_report?rev=447346, it doesn't look like there was a direct effect on frame_times or latency. Any room for improvement? Or should we close as WontFix?
,
Feb 13 2017
I think this is expected, I'll do some benchmarking this week to make sure.
,
Mar 11 2017
,
Apr 24 2017
Ping for ericrk c#5. Can we close this?
,
Apr 24 2017
Yeah, let's close this out. |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by briander...@chromium.org
, Feb 8 2017