Issue metadata
Sign in to add a comment
|
4%-16.7% regression in system_health.common_desktop at 475913:476747 |
||||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Jun 5 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8977591344005996624
,
Jun 6 2017
=== Auto-CCing suspected CL author hansberry@chromium.org === Hi hansberry@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 : hansberry Commit : bdcf8951a275f7f572ea15bdfe921ed222598159 Date : Thu Jun 01 21:23:50 2017 Subject: ChromeCryptAuthService: only perform device sync once enrollment is complete. Bisect Details Configuration: winx64_high_dpi_perf_bisect Benchmark : system_health.common_desktop Metric : load:energy_sum/load_games/load_games_lazors Change : 17.15% | 7.1827023928 -> 8.26956719138 Revision Result N chromium@475912 7.1827 +- 0.781675 6 good chromium@476264 7.17791 +- 1.4755 9 good chromium@476352 7.46984 +- 1.65722 14 good chromium@476396 7.43078 +- 2.19212 21 good chromium@476418 7.50668 +- 1.2047 14 good chromium@476429 7.45743 +- 0.942752 9 good chromium@476435 7.26916 +- 1.17225 9 good chromium@476438 6.97207 +- 0.51739 6 good chromium@476439 7.05618 +- 0.508903 6 good chromium@476440 7.81542 +- 1.37907 14 bad <-- chromium@476616 8.26957 +- 0.913949 9 bad To Run This Test src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=load.games.lazors system_health.common_desktop Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8977591344005996624 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5870613173895168 | 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!
,
Jun 6 2017
Hi miu@, could you help me understand how my CL might have caused this issue? I'm seeing a lot of references to "games" which shouldn't have been affected by my change.
,
Jun 6 2017
It's a little hard to get context from looking at your CL. What platforms does it do the device enrollment on? How did it change the timing of device enrollment exactly? A note about the benchmarks: we restart the browser for every page. So adding delayed work after startup can cause the benchmarks to regress on various pages, which may not be directly related to your code. I'm going to kick off a few more bisects to see if they all point to your CL as a culprit.
,
Jun 6 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8977531093549817856
,
Jun 6 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8977531065961541664
,
Jun 6 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8977531049223022576
,
Jun 6 2017
=== BISECT JOB RESULTS === Perf regression found with culprit Suspected Commit Author : Philippe Hamel Commit : 1bc3d8c1ce10761406e57b07f7a2dc1ad301221e Date : Thu Jun 01 16:57:23 2017 Subject: Add query only fieldtrial for TranslateRankerModel. Bisect Details Configuration: winx64nvidia_perf_bisect Benchmark : smoothness.top_25_smooth Metric : frame_times/LinkedIn Change : 2.94% | 16.2369278695 -> 16.7142612903 Revision Result N chromium@476266 16.2369 +- 0.00593161 6 good chromium@476302 16.2107 +- 0.140907 6 good chromium@476320 16.25 +- 0.0130524 6 good chromium@476325 16.2099 +- 0.00631621 6 good chromium@476327 16.2204 +- 0.190303 6 good chromium@476328 16.7239 +- 0.0698167 6 bad <-- chromium@476329 16.7288 +- 0.0966242 6 bad chromium@476338 16.7265 +- 0.113701 6 bad chromium@476409 16.7143 +- 0.0903569 6 bad To Run This Test src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=LinkedIn smoothness.top_25_smooth Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8977531049223022576 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5210110822973440 | 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!
,
Jun 6 2017
=== BISECT JOB RESULTS === Bisect failed for unknown reasons Please contact the team (see below) and report the error. Bisect Details Configuration: winx64_high_dpi_perf_bisect Benchmark : system_health.common_desktop Metric : timeToFirstContentfulPaint_avg/load_news/load_news_qq To Run This Test src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=load.news.qq system_health.common_desktop Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8977531093549817856 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5771789868728320 | 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!
,
Jun 6 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8977521500961854992
,
Jun 6 2017
=== BISECT JOB RESULTS === Perf regression found with culprit Suspected Commit Author : yoav Commit : 1f18863cc9b2a5fe463671f3eb60de2f95bbf283 Date : Sat Jun 03 00:20:44 2017 Subject: [preload] Mandatory `as` value and related spec alignments Bisect Details Configuration: linux_perf_bisect Benchmark : loading.desktop Metric : timeToFirstContentfulPaint_avg/pcv1-warm/TheOnion Change : 192.08% | 114.14311111 -> 263.637111104 Revision Result N chromium@476650 114.143 +- 162.311 9 good chromium@476761 129.804 +- 171.524 9 good chromium@476816 97.2217 +- 25.1438 6 good chromium@476830 92.598 +- 61.2162 9 good chromium@476837 112.31 +- 171.358 14 good chromium@476839 129.071 +- 282.269 14 good chromium@476840 250.442 +- 193.989 14 bad <-- chromium@476841 252.948 +- 189.829 14 bad chromium@476844 262.894 +- 11.4577 9 bad chromium@476871 263.637 +- 19.2167 9 bad To Run This Test src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=TheOnion loading.desktop Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8977531065961541664 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5791374516944896 | 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!
,
Jun 6 2017
=== BISECT JOB RESULTS === Bisect failed for unknown reasons Please contact the team (see below) and report the error. Bisect Details Configuration: winx64_high_dpi_perf_bisect Benchmark : system_health.common_desktop Metric : timeToFirstContentfulPaint_avg/load_news/load_news_qq To Run This Test src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=load.news.qq system_health.common_desktop Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8977521500961854992 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5771789868728320 | 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 m...@chromium.org
, Jun 5 2017