That sounds fine to me too ..
-phil.
On 10/1/2013 6:57 AM, Andrew Brygin wrote:
Hi Vadim,
the fix looks fine to me.
I am not sure about the test, which will fail on windows.
Probably it would be fine too, if you will submit a new
bug against D3D immediately after the push.
Thanks,
Andrew
On 10/1/2013 2:06 PM, Vadim Pakhnushev wrote:
Hi,
Please review a fix for this bug:
https://bugs.openjdk.java.net/browse/JDK-8024343
It's confidential, so here is a simple explanation:
While rendering with XOR mode set, XRSurfaceData validation method
doesn't detect subsequent changes in foreground (or xor) color if the
Composite itself is not changed.
Proposed fix is to validate pixel value independently.
http://cr.openjdk.java.net/~vadim/8024343/webrev.00/
While creating a regression test, I've found that D3D pipeline
doesn't pass this test, that is drawing in XOR mode into
VolatileImage, while passing the original test which is drawing to
the screen.
So I'm planning to file a new bug for that.
Thanks,
Vadim