Hi, Alan.
I am able to reproduce it, and the root cause is correct. I will try to
fix it. Thanks for an additional information.
On 29.11.15 22:39, Alan Snyder wrote:
I believe I have encountered this previously reported bug and understand the
cause. It is a timing problem that occurs when native code asks for pixels (
drawInCGLContext) at the same time that LWWindowPeer is trying to replace the
surface data. The problem occurs if the request for pixels happens after a new
CGLSurfaceData has been created but before it is initialized with a background
color. In most cases, this error results in a transient flicker, but if the
window is transparent, then the garbage pixels will be painted over by the
background color and remain visible. Unfortunately, I have code that frequently
runs into this problem.
I have attached a test case that is reliable (for me) on 1.8.0_66.
I hope with this new information the bug will be fixed soon.
Alan
--
Best regards, Sergey.
