Issue metadata
Sign in to add a comment
|
4.5%-21.6% regression in thread_times.key_silk_cases at 531700:531788 |
||||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Jan 26 2018
๐ Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/107dd992840000
,
Jan 27 2018
๐ Found a significant difference after 1 commit. https://pinpoint-dot-chromeperf.appspot.com/job/107dd992840000 Revert "Use new Skia resource purging API when we go idle" By torne@chromium.org ยท Wed Jan 24 22:03:51 2018 chromium @ 1fd722308b8c3d4533b19ad2df5d40752e945e01 Understanding performance regressions: http://g.co/ChromePerformanceRegressions
,
Feb 5 2018
,
Mar 19 2018
ericrk@, this bisect pointed out my sheriff revert of your change - handing this to you to take a look, since I have no idea (and didn't notice this for some time, whoops)
,
Apr 18 2018
My change had a bug which caused us to preserve caches longer than intended. It makes sense that reverting it (and throwing out caches) would result in more work in certain cases. I'm not too concerned as all the "regressions were just bringing us back to the previous levels before my buggy CL landed. |
|||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||
Comment 1 by 42576172...@developer.gserviceaccount.com
, Jan 26 2018