Issue metadata
Sign in to add a comment
|
19.6% regression in blink_perf.css at 455489:455614 |
||||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Mar 13 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8985239145328746160
,
Mar 13 2017
=== Auto-CCing suspected CL author pfeldman@chromium.org === Hi pfeldman@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 : pfeldman Commit : 87cb93767256f83a39625de66e528ef2fb1296d5 Date : Thu Mar 09 00:35:22 2017 Subject: DevTools: move counter-related devtools.timeline trace events into probe:: probes. Bisect Details Configuration: winx64_zen_perf_bisect Benchmark : blink_perf.css Metric : SelectorCountScaling/SelectorCountScaling Change : 2.74% | 102.382364346 -> 99.5754339946 Revision Result N chromium@455488 102.382 +- 12.5764 21 good chromium@455551 102.237 +- 12.1271 14 good chromium@455583 100.965 +- 10.7264 9 good chromium@455599 101.595 +- 10.9219 14 good chromium@455608 100.757 +- 21.3365 14 good chromium@455612 100.734 +- 29.9227 14 good chromium@455613 99.4037 +- 8.81002 14 bad <-- chromium@455614 99.5754 +- 13.8509 14 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 blink_perf.css Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8985239145328746160 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=6095270057607168 | 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
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8985204835218000768
,
Mar 13 2017
The numbers from that bisect look really weird. Kicked off another one with a broader range since the graph went up and down.
,
Mar 14 2017
=== BISECT JOB RESULTS === Perf regression found with culprit Suspected Commit Author : Michael Hablich Commit : 7c3936c67a6361371d9a00d83e2c03ecc603ce76 Date : Thu Mar 09 08:05:42 2017 Subject: Version 5.9.34.1 (Turn on I+TF) Bisect Details Configuration: winx64_zen_perf_bisect Benchmark : blink_perf.css Metric : SelectorCountScaling/SelectorCountScaling Change : 22.08% | 104.511222104 -> 81.4303949068 Revision Result N chromium@455426 104.511 +- 1.38869 6 good chromium@455606 104.868 +- 1.71846 6 good chromium@455696 104.206 +- 2.77189 6 good chromium@455719 104.559 +- 2.20517 6 good chromium@455719,v8@7c3936c67a 80.0514 +- 10.0027 6 bad <-- chromium@455719,v8@fbffc377e3 81.6983 +- 1.42487 6 bad chromium@455720 81.892 +- 2.10971 6 bad chromium@455721 79.2535 +- 11.8999 6 bad chromium@455722 81.6882 +- 1.52704 6 bad chromium@455725 80.7295 +- 6.01941 6 bad chromium@455730 81.2636 +- 2.70197 6 bad chromium@455741 80.3947 +- 3.45085 6 bad chromium@455785 81.4304 +- 6.56767 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 blink_perf.css Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8985204835218000768 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5882572539691008 | 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 hjd@chromium.org
, Mar 13 2017