Issue metadata
Sign in to add a comment
|
2.7%-10.2% regression in rendering.mobile/avg_surface_fps at 591987:592071 |
||||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Sep 25
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/16657f7b640000
,
Sep 26
📍 Found significant differences after each of 2 commits. https://pinpoint-dot-chromeperf.appspot.com/job/16657f7b640000 implement SkPixmap::erase() with draws by mtklein@google.com https://skia.googlesource.com/skia/+/42f29f11817f9829b92dbf52d77f1447fbfc5ed0 1133 → 1148 (+14.76) Make atomic lists list for bitmaps and paths by herb@google.com https://skia.googlesource.com/skia/+/6c8d242b14355bf66c9137e9e4d6c7861d22168f 1152 → 1250 (+97.67) Understanding performance regressions: http://g.co/ChromePerformanceRegressions Benchmark documentation link: https://bit.ly/blink-perf-benchmarks
,
Sep 26
"implement SkPixmap::erase() with draws" has been in and out but is out now. "Make atomic lists list for bitmaps and paths" is a bug fix. I think this is WAI. |
|||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||
Comment 1 by 42576172...@developer.gserviceaccount.com
, Sep 25