RFR: 8022403: sun/java2d/DirectX/OnScreenRenderingResizeTest/OnScreenRenderingResizeTest.java fails

Phil Race prr at openjdk.org
Tue Nov 15 18:03:02 UTC 2022


On Tue, 15 Nov 2022 08:59:23 GMT, Jayathirth D V <jdv at openjdk.org> wrote:

>> This test was failing because of color difference on boundaries. Updated test case to not check edges and added color tolerance.
>> 
>> Also updated test to:
>> 1) Use standard GraphicsConfiguration bounds for image instead of constant numbers.
>> 2) Use waitForIdle before screen capture and added auto delay
>> 3) Move mouse pointer away from rendering content (In my local Mac it failed every time when mouse pointer was present over test frame)
>> 4) Remove usage of insets as it is undecorated and remove windows closing listener
>> 5) Dispose frame when test fails
>> 
>> With these changes test is not failing in CI with multiple runs on all platforms.
>
> test/jdk/sun/java2d/DirectX/OnScreenRenderingResizeTest/OnScreenRenderingResizeTest.java line 31:
> 
>> 29:  * @author Dmitri.Trembovetski at sun.com: area=Graphics
>> 30:  * @run main/othervm OnScreenRenderingResizeTest
>> 31:  * @run main/othervm -Dsun.java2d.d3d=false OnScreenRenderingResizeTest
> 
> DirectDraw/GDI is default pipeline in Windows(https://docs.oracle.com/javase/10/troubleshoot/java-2d-pipeline-rendering-and-properties.htm#JSTGD428 ) and explicitly mentioning d3d=false should not make any difference that's why i have removed it.
> 
> But i have verified updated test even with d3d=false and there are no issues. If needed i can update the test to run with d3d=false also.

oh boy that doc is SO out of date.
D3D has been the default pipeline since shortly after that doc was written !
It was written for 6 GA and 6u10 made D3D the default.
So if you want to be sure to disable it you DO need d3d=false ..

-------------

PR: https://git.openjdk.org/jdk/pull/11158



More information about the client-libs-dev mailing list