Issue metadata
Sign in to add a comment
|
31.1%-58.9% regression in loading.desktop at 483666:483828 |
||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Jul 4 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8974997067689496576
,
Jul 5 2017
=== Auto-CCing suspected CL author yhirano@chromium.org === Hi yhirano@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 : Yutaka Hirano Commit : 3a0905c1392858d49ee03a2081500351d4568320 Date : Fri Jun 30 16:09:00 2017 Subject: Remove PingLoaderImpl Bisect Details Configuration: win_perf_bisect Benchmark : loading.desktop Metric : timeToFirstContentfulPaint_avg/pcv1-warm/money.cnn Change : 48.68% | 259.2105 -> 385.3905 Revision Result N chromium@483665 259.211 +- 20.8702 6 good chromium@483693 263.66 +- 4.37838 6 good chromium@483707 261.869 +- 8.85965 6 good chromium@483714 281.824 +- 134.263 9 good chromium@483718 262.607 +- 5.88797 6 good chromium@483719 265.557 +- 10.627 6 good chromium@483720 389.009 +- 78.6868 6 bad <-- chromium@483721 385.39 +- 79.6965 6 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=money.cnn loading.desktop More information on addressing performance regressions: http://g.co/ChromePerformanceRegressions Debug information about this bisect: https://chromeperf.appspot.com/buildbucket_job_status/8974997067689496576 For feedback, file a bug with component Speed>Bisection
,
Jul 5 2017
All reports are for money.cnn. I confirmed that PingLoader::LoadImage is used by the story.
,
Jul 5 2017
,
Jul 13 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/7c45454f3655609e883505dc3269b2a6140302a6 commit 7c45454f3655609e883505dc3269b2a6140302a6 Author: yhirano <yhirano@chromium.org> Date: Thu Jul 13 04:19:56 2017 Set "ping" requests priority as lowest Requests from PingLoader had the lowest priority, but I changed it unintentionally. This CL restores the original behavior. BUG= 739161 Review-Url: https://codereview.chromium.org/2982523002 Cr-Commit-Position: refs/heads/master@{#486267} [modify] https://crrev.com/7c45454f3655609e883505dc3269b2a6140302a6/third_party/WebKit/Source/core/loader/PingLoaderTest.cpp [modify] https://crrev.com/7c45454f3655609e883505dc3269b2a6140302a6/third_party/WebKit/Source/platform/loader/fetch/ResourceFetcher.cpp
,
Jul 13 2017
,
Jul 14 2017
Fixed by #6 |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by rmcilroy@chromium.org
, Jul 4 2017