Issue metadata
Sign in to add a comment
|
5.4% regression in rasterize_and_record_micro.top_25 at 544614:544647 |
||||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Mar 23 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/12f2452d440000
,
Mar 23 2018
📍 Found a significant difference after 1 commit. https://pinpoint-dot-chromeperf.appspot.com/job/12f2452d440000 cc: Support animated images in PaintFilters. by khushalsagar@chromium.org https://chromium.googlesource.com/chromium/src/+/0c568dc759c39d7598cfd811f7f2096daebc26c0 Understanding performance regressions: http://g.co/ChromePerformanceRegressions
,
Mar 24 2018
,
Mar 29 2018
There are multiple changes to discardable image analysis, which is included in record_time, that went into that change. We include non-draw ops during the analysis, and also track HasDiscardableImages state for them correctly. Excluding both these changes locally and testing on Android Nexus5, I was able to recover the regression. But these things are necessary to support the use-case addressed in that patch. Also the regression is minor (~0.05ms in my local test). |
|||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||
Comment 1 by 42576172...@developer.gserviceaccount.com
, Mar 23 2018