Re: [OpenJDK 2D-Dev] RFR: 8250658: Performance of ClipFlatOval Renderperf test is very low

2021-05-12 Thread Ajit Ghaisas
On Tue, 11 May 2021 16:06:09 GMT, Alexey Ushakov wrote: >> src/java.desktop/macosx/native/libawt_lwawt/java2d/metal/MTLClip.h line 52: >> >>> 50: @property (readonly) id stencilTextureRef; >>> 51: @property (readonly) BOOL stencilMaskGenerationInProgress; >>> 52: @property (readwrite ) BOOL

Re: [OpenJDK 2D-Dev] RFR: 8250658: Performance of ClipFlatOval Renderperf test is very low

2021-05-12 Thread Ajit Ghaisas
On Fri, 7 May 2021 22:29:53 GMT, Alexey Ushakov wrote: > Implemented indirect rendering (via stencil texture attachment) to stencil > texture I tested on macBook pro 16" with discrete GPU. No regressions were observed in J2DDemo, SwingSet2 and Netbeans IDE with this fix. I also tested

Re: [OpenJDK 2D-Dev] RFR: 8250658: Performance of ClipFlatOval Renderperf test is very low

2021-05-11 Thread Ajit Ghaisas
On Fri, 7 May 2021 22:29:53 GMT, Alexey Ushakov wrote: > Implemented indirect rendering (via stencil texture attachment) to stencil > texture src/java.desktop/macosx/native/libawt_lwawt/java2d/metal/MTLClip.h line 52: > 50: @property (readonly) id stencilTextureRef; > 51: @property (readonly)

Re: [OpenJDK 2D-Dev] RFR: 8250658: Performance of ClipFlatOval Renderperf test is very low

2021-05-11 Thread Jayathirth D V
On Fri, 7 May 2021 22:29:53 GMT, Alexey Ushakov wrote: > Implemented indirect rendering (via stencil texture attachment) to stencil > texture Its good that we have removed intermediate stencil blit operations using this approach. LGTM. jtreg/JCK all test run is fine. J2DDemo, SwingSet2,

[OpenJDK 2D-Dev] RFR: 8250658: Performance of ClipFlatOval Renderperf test is very low

2021-05-07 Thread Alexey Ushakov
Implemented indirect rendering (via stencil texture attachment) to stencil texture - Commit messages: - 8250658: Performance of ClipFlatOval Renderperf test is very low Changes: https://git.openjdk.java.net/jdk/pull/3929/files Webrev: