On Tue, 15 Sep 2026 20:15:28 GMT, Phil Race <[email protected]> wrote:
> The idea of this test is to verify that hardware pipelines match the software > pipelines for some specific rendering primitives. > > Getting all these primitives to paint identically across different pipelines > is not easy, and in particular D3D and GDI are problems, but the test made it > even harder for itself in ways that muddied the picture. > > This test would fail even where it should pass, because it wasn't always > getting the window contents it should have done. > Rounded borders and ncomplete paints were two specific problems. > And although it was intended to test OpenGL it didn't actually ask for OpenGL. > > The test updates here can't fix the actual pipeline rendering, but give the > test a chance to pass where it can. > And the problem list now more narrowly identifies where it fails. > > --------- > - [x] I confirm that I make this contribution in accordance with the [OpenJDK > Interim AI Policy](https://openjdk.org/legal/ai). Good idea to use insets and uiScale 1 in these scenarios to make sure Java2D is doing things correctly and avoid sampling and rounded corner issues on specific platforms. ------------- Marked as reviewed by jdv (Reviewer). PR Review: https://git.openjdk.org/jdk/pull/32891#pullrequestreview-5219251990
