Version: ToT
OS: All (probably)
What steps will reproduce the problem?
(1) Look at the test result for virtual/gpu-rasterization/fast/images/pixel-crack-image-background-webkit-transform-scale.html with something that will let you see that color FFFEFF is not the same as FFFFFF
What is the expected output? A completely FFFFFF image, or at least something that is the same across the entire tiled output.
What do you see instead? Tile artifacts causing colors to differ by 1 in one or more channels.
It's hard to say when this regressed, or whether it has always been this way, because the test has been marked as Timeout and apparently has not had pixel expectations updated, or otherwise not been marked as failing.
Comment 1 by bugdroid1@chromium.org
, Dec 22 2016