Issue metadata
Sign in to add a comment
|
1.2%-92.5% regression in media.tough_video_cases_tbmv2 at 476975:477745 |
||||||||||||||||||||||
Issue descriptionLikely a dupe of Issue 731841 , though the range is a little tighter here.
,
Jun 9 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8977235375420086352
,
Jun 10 2017
=== Auto-CCing suspected CL author ulan@chromium.org === Hi ulan@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 : Ulan Degenbaev Commit : c8e6cdfdcee0319fec69926e074d047f668c6f8b Date : Tue Jun 06 14:16:55 2017 Subject: Revert "[heap] Use partial free when shrinking instead of uncommitting" Bisect Details Configuration: winx64_high_dpi_perf_bisect Benchmark : media.tough_video_cases_tbmv2 Metric : memory:chrome:all_processes:reported_by_chrome:v8:heap:map_space:effective_size_avg/video.html?src_crowd.ogg_type_audio Change : 92.48% | 544768.0 -> 1048576.0 Revision Result N chromium@477369 544768 +- 0.0 6 good chromium@477411 544768 +- 0.0 6 good chromium@477432 544768 +- 0.0 6 good chromium@477443 544768 +- 0.0 6 good chromium@477448 544768 +- 0.0 6 good chromium@477448,v8@8a15026270 544768 +- 0.0 6 good chromium@477448,v8@c8e6cdfdce 1048576 +- 0.0 6 bad <-- chromium@477448,v8@5005faed5c 1048576 +- 0.0 6 bad chromium@477449 1048576 +- 0.0 6 bad chromium@477450 1048576 +- 0.0 6 bad chromium@477451 1048576 +- 0.0 6 bad chromium@477453 1048576 +- 0.0 6 bad chromium@477536 1048576 +- 0.0 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=video.html.src.crowd.ogg.type.audio media.tough_video_cases_tbmv2 Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8977235375420086352 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5261201304453120 | 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 10 2017
Issue 731872 has been merged into this issue.
,
Jun 10 2017
Issue 731630 has been merged into this issue.
,
Jun 10 2017
,
Jun 12 2017
Issue 732478 has been merged into this issue.
,
Jun 13 2017
Michael is working on relanding the patch that improved memory usage.
,
Jun 13 2017
,
Jun 14 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976799152342715024
,
Jun 14 2017
The patch relanded but didn't recover most of the graphs. By now I don't believe that the majority of regressions was actually caused by reverting my patch. In the range I also see 24fca6f Enable parallel decode requests for VpxVideoDecoder when offloading. by Dale Curtis ยท 7 days ago Assigning back that the other regressions can be sorted out.
,
Jun 14 2017
=== BISECT JOB RESULTS === Perf regression found with culprit Suspected Commit Author : Dale Curtis Commit : 24fca6fb8a70b95cae28843751485529e02378c0 Date : Wed Jun 07 03:50:37 2017 Subject: Enable parallel decode requests for VpxVideoDecoder when offloading. Bisect Details Configuration: mac_air_perf_bisect Benchmark : media.tough_video_cases_tbmv2 Metric : memory:chrome:renderer_processes:reported_by_chrome:malloc:allocated_objects_size_avg/video.html?src_smpte_3840x2160_60fps_vp9.webm_seek Change : 22.61% | 172339013.333 -> 211309277.333 Revision Result N chromium@477513 172339013 +- 385745 6 good chromium@477531 172303333 +- 362911 6 good chromium@477536 172221861 +- 732980 6 good chromium@477537 211227440 +- 1069970 6 bad <-- chromium@477538 211462904 +- 698553 6 bad chromium@477540 211372621 +- 830931 6 bad chromium@477548 211342920 +- 760542 6 bad chromium@477583 211309277 +- 429439 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=video.html.src.smpte.3840x2160.60fps.vp9.webm.seek media.tough_video_cases_tbmv2 Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8976799152342715024 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5041915910684672 | 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 chcunningham@chromium.org
, Jun 9 2017