On Thu, 17 Sep 2026 07:10:55 GMT, Prasanta Sadhukhan <[email protected]> wrote:
> It fails for me locally in windows11 > > ----------System.out:(11/752)----------^M Testing source: > BufferedImage@3559ef8e: type = 1 DirectColorModel: rmask=ff0000 gmask=ff00 > bmask=ff amask=0 IntegerInterleavedRaster: width = 50 height = 50 #Bands = 3 > xOff = 0 yOff = 0 dataOffset[0] 0^M testing blit rect src: > java.awt.Rectangle[x=0,y=0,width=30,height=30]^M dst: > java.awt.Rectangle[x=-10,y=-10,width=30,height=30]^M calculated dest > rect:java.awt.Rectangle[x=0,y=0,width=20,height=20]^M subtest passed^M > testing blit rect src: java.awt.Rectangle[x=-10,y=-10,width=50,height=50]^M > dst: java.awt.Rectangle[x=0,y=0,width=50,height=50]^M calculated dest > rect:java.awt.Rectangle[x=10,y=10,width=40,height=40]^M Dumped res to: > SourceClippingBlitTest.png^M Test failed at x=9 y=9^M expected: ff00ff00 > got:ff24db00^M ----------System.err:(39/3186)----------^M > java.lang.RuntimeException: Test failed at x=9 y=9^M at > SourceClippingBlitTest.test(SourceClippingBlitTest.java:253) Looks like on your local system you have specific display scaling which is causing blending of pixels at intersection. Also even if we take HiDPI capture, native image can also do this blending. To not hit these failure in future, we should not verifiy edge as well as intersection pixels. ------------- PR Comment: https://git.openjdk.org/jdk/pull/32904#issuecomment-5711312320
