Issue metadata
Sign in to add a comment
|
74.9%-93.1% regression in blink_perf.canvas at 537243:537521 |
||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Feb 21 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/12d77617840000
,
Mar 5 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/11e2229c440000
,
Mar 6 2018
📍 Found a significant difference after 1 commit. https://pinpoint-dot-chromeperf.appspot.com/job/11e2229c440000 canvas: Avoid unnecessary book-keeping of images in CanvasImageProvider. by khushalsagar@chromium.org https://chromium.googlesource.com/chromium/src/+/cde9d1d1e464957716e4e15f44810fe28753e4df Understanding performance regressions: http://g.co/ChromePerformanceRegressions
,
Mar 6 2018
Looks like things improved when we were keeping these images ref-ed throughout the frame, even though they didn't have anything locked inside cc. So nothing regressed really, things came back to where they were before https://chromium.googlesource.com/chromium/src/+/aa36bffbca291c237487d2374d416a9c59f4de2e. I'll keep this open to get around to investigate before closing this.
,
Jul 25
|
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by 42576172...@developer.gserviceaccount.com
, Feb 21 2018