Issue metadata
Sign in to add a comment
|
Regression: Black traces of cube is seen in 'Cube 3D' experiment.
Reported by
rk...@etouch.net,
Apr 29 2016
|
||||||||||||||||||||||||
Issue descriptionChrome Version: 52.0.2720.0 Revision 0fcb776029fb1380c8e8c9d4cbdcaf7a217ddb5c-refs/heads/master@{#390547} OS: Windows (7,8,8.1,10),Linux (14.04 LTS),Mac URL: http://paulrhayes.com/experiments/cube-3d/ What steps will reproduce the problem? 1. Launch chrome and go to above URL. 2. Press up/down/left/right key from keyboard and simultaneously click on the cube ,observe. Black traces on cube is observed. No such black traces should seen. This is a regression issue, broken in 'M-51', below is bisect info: Good Build: 51.0.2664.0 Bad Build: 51.0.2665.0 Narrow Bisect: https://chromium.googlesource.com/chromium/src/+log/5d6a2e1871be2f4f8cc4eacb366a5abf8766fbfc..e5db881b268bc26ae3989a9f6f6a3325d7323baf?pretty=fuller&n=100 Suspecting: r378490 Could you please help me to reassign this issue if your change is not cause for it?
,
Apr 29 2016
I will have a look on it, thanks for starting the research.
,
Apr 29 2016
@rkote: the commit you pointed only changed the css/html of a few tests, but not the way borders are rendered. The changes introduced targeted only to remove the deprecation warning (which should only print to the console and not after layout or anything else). The change on border-image came later on: https://codereview.chromium.org/1815043003/ Are you using border-image (or prefix form) in this demo?
,
Apr 29 2016
,
May 3 2016
Just to update : Able to reproduce the issue on win8.1 chrome version 52.0.2723.0
,
May 4 2016
cavalcantii@: Is there any work in-progress for this stable blocker issue?
,
May 4 2016
This repros on 51 in chromeos
,
May 5 2016
Just to double check, I wrote a patch to revert the border-image change and applied it on top of current chromium/blink and the test case wasn't affected as I suspected. That means that probably something else broke rendering for the reported test case. To make things more interesting, it seems that the top of master no longer has the rendering artifacts (at least on mac retina). I tried Canary (Version 52.0.2724.0) and was unable to reproduce. I'm unavailable to dive deeper on this matter with the urgency required, so I'm making this bug available to someone to resume the work.
,
May 5 2016
Unable to repro on Linux with a local check out of e5db881b268bc26ae3989a9f6f6a3325d7323baf. From the video this looks like a paint invalidation issue.
,
May 5 2016
,
May 5 2016
Ali, is this a dup of 607423?
,
May 6 2016
,
May 6 2016
|
|||||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||||
Comment 1 by msrchandra@chromium.org
, Apr 29 2016