Issue metadata
Sign in to add a comment
|
8.4% regression in thread_times.key_idle_power_cases at 471252:471296 |
||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
May 16 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8979407825353529408
,
May 16 2017
=== Auto-CCing suspected CL author mlippautz@chromium.org === Hi mlippautz@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 : mlippautz Commit : f86985ad116d75255ddf78b3a4df2590543fd8cd Date : Wed May 10 15:30:36 2017 Subject: [heap] Bound number of tasks by embedder limit instead of artificially capping Bisect Details Configuration: android_nexus5X_perf_bisect Benchmark : thread_times.key_idle_power_cases Metric : tasks_per_second_total_all/request-animation-frame.html Change : 6.48% | 5.39306760635 -> 5.74242853018 Revision Result N chromium@471251 5.39307 +- 0.367304 6 good chromium@471274 5.28444 +- 0.269597 6 good chromium@471280 5.39827 +- 0.498082 9 good chromium@471280,v8@5c66d6fcd4 5.4175 +- 0.320858 6 good chromium@471280,v8@cf37556f0e 5.37691 +- 0.209341 6 good chromium@471280,v8@69c6970fae 5.39875 +- 0.473581 9 good chromium@471280,v8@2026d5cb79 5.35047 +- 0.216444 6 good chromium@471280,v8@11a211ff1b 5.36704 +- 0.208936 6 good chromium@471280,v8@f86985ad11 5.77483 +- 0.32858 6 bad <-- chromium@471281 5.77552 +- 0.304364 6 bad chromium@471282 5.95888 +- 0.424537 6 bad chromium@471283 5.78536 +- 0.249713 6 bad chromium@471285 5.82727 +- 0.392804 6 bad chromium@471296 5.74243 +- 0.291751 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=request.animation.frame.html thread_times.key_idle_power_cases Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8979407825353529408 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5696384210042880 | 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!
,
May 19 2017
The following revision refers to this bug: https://chromium.googlesource.com/v8/v8.git/+/34c1c76d8ab1c09799d6f5c9089ee1a7b6487943 commit 34c1c76d8ab1c09799d6f5c9089ee1a7b6487943 Author: Michael Lippautz <mlippautz@chromium.org> Date: Fri May 19 13:02:24 2017 [heap] Limit the number of pointer updating tasks (again) Task creation often dominates the actual work that is being done. Bug: chromium:722989 Change-Id: Ibdd6ffa6f3154f17dc6ccbd30475710b97e802e7 Reviewed-on: https://chromium-review.googlesource.com/508783 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/master@{#45425} [modify] https://crrev.com/34c1c76d8ab1c09799d6f5c9089ee1a7b6487943/src/heap/mark-compact.cc
,
May 21 2017
,
May 29 2017
The CL in #4 put the limits back in place. |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by lanwei@chromium.org
, May 16 2017