Announcement

Collapse
No announcement yet.

MLAA For Mesa Is Ready For Testing

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by Drago View Post
    So...this MLAA is entirely in software/shader, and not actual GPU anti-aliasing method?
    Yes, MLAA is typically implemented as a post-processing filter. Ideally I guess you would run it after rendering the game imagery but before rendering text (eg HUD).

    Comment


    • #17
      problem?

      I get the following output, and the sphere is definitely not aa.

      I'm on mesa 7.10.3. Would switching to latest svn help?

      [stephen@goodt60 MLAA-test-app]$ ./mlaatest
      Irrlicht Engine version 1.7.2
      Linux 2.6.39-ARCH #1 SMP PREEMPT Mon Jun 27 22:01:13 CEST 2011 i686
      ATTENTION: default value of option vblank_mode overridden by environment.
      ATTENTION: default value of option vblank_mode overridden by environment.
      Using renderer: OpenGL 2.1
      Gallium 0.4 on ATI RV515: X.Org R300 Project
      OpenGL driver version is 1.2 or better.
      GLSL version: 1.2
      Loaded texture: AreaMap33
      Unsupported texture format
      GLSL shader failed to compile
      0:4(12): warning: extension `GL_ARB_shader_texture_lod' unsupported in fragment shader
      0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
      0:16(86): error: `return' with wrong type , in function `tex2Doffset' returning vec4
      0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
      0:72(48): error: type mismatch
      0:72(48): error: `return' with wrong type , in function `Area' returning vec2
      0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
      0:88(54): error: type mismatch
      0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
      0:89(54): error: type mismatch
      0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
      0:103(54): error: type mismatch
      0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
      0:104(54): error: type mismatch

      Quit message received.

      Ran 28.004s, average fps 198.50, min 1

      Comment


      • #18
        Originally posted by TheCycoONE View Post
        I get the following output, and the sphere is definitely not aa.

        I'm on mesa 7.10.3. Would switching to latest svn help?

        [stephen@goodt60 MLAA-test-app]$ ./mlaatest
        Irrlicht Engine version 1.7.2
        Linux 2.6.39-ARCH #1 SMP PREEMPT Mon Jun 27 22:01:13 CEST 2011 i686
        ATTENTION: default value of option vblank_mode overridden by environment.
        ATTENTION: default value of option vblank_mode overridden by environment.
        Using renderer: OpenGL 2.1
        Gallium 0.4 on ATI RV515: X.Org R300 Project
        OpenGL driver version is 1.2 or better.
        GLSL version: 1.2
        Loaded texture: AreaMap33
        Unsupported texture format
        GLSL shader failed to compile
        0:4(12): warning: extension `GL_ARB_shader_texture_lod' unsupported in fragment shader
        0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
        0:16(86): error: `return' with wrong type , in function `tex2Doffset' returning vec4
        0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
        0:72(48): error: type mismatch
        0:72(48): error: `return' with wrong type , in function `Area' returning vec2
        0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
        0:88(54): error: type mismatch
        0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
        0:89(54): error: type mismatch
        0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
        0:103(54): error: type mismatch
        0:0(0): error: no matching function for call to `texture2DLod(sampler2D, vec2, float)'
        0:104(54): error: type mismatch

        Quit message received.

        Ran 28.004s, average fps 198.50, min 1
        "It requires OpenGL 2.0 and GL_ARB_shader_texture_lod (Mesa 7.11, 7.10 does not have that). [...]"

        Please reread the announcement

        Comment


        • #19
          Code:
          Irrlicht Engine version 1.8.0-alpha
          Linux 3.0.0-drm #1 SMP PREEMPT Tue Jul 5 18:40:40 CEST 2011 x86_64
          Using renderer: OpenGL 2.1
          Gallium 0.4 on AMD JUNIPER: X.Org
          OpenGL driver version is 1.2 or better.
          GLSL version: 1.2
          Could not load sprite bank because the file does not exist: #DefaultFont
          Loaded texture: AreaMap33
          Quit message received.
          
          Ran 122.932s, average fps 274.87, min 1
          
          Average on fps 3033.53, average off fps 235.15
          
          MLAA took on average 0.3ms
          that output is obviously wrong.

          Beside this error: I'm wondering if the FPS output with the sphere is really right, but the sphere shows more FPS if the window is in the lower right corner (~600FPS). In the upper left corner I get ~250FPS. Both without MLAA. With MLAA it's ~77FPS and ~74FPS so it's not that visible but still the case.

          Comment


          • #20
            Again...

            Ok, I've compiled the latest mesa/ati-dri/libgl from git. My display is pretty wonkey so I won't be keeping this setup long (colours are all wrong e.g.)

            However, here is the new output:
            [stephen@goodt60 MLAA-test-app]$ ./mlaatest
            Irrlicht Engine version 1.7.2
            Linux 2.6.39-ARCH #1 SMP PREEMPT Mon Jun 27 22:01:13 CEST 2011 i686
            ATTENTION: default value of option vblank_mode overridden by environment.
            Using renderer: OpenGL 2.1
            Gallium 0.4 on llvmpipe: VMware, Inc.
            OpenGL driver version is 1.2 or better.
            GLSL version: 1.2
            Loaded texture: AreaMap33
            Unsupported texture format
            Quit message received.

            Ran 16.368s, average fps 82.66, min 1

            The title bar said MPAA: Off

            Of course with such wonky display I couldn't really tell how it looked.

            Comment


            • #21
              Originally posted by TheCycoONE View Post
              Using renderer: OpenGL 2.1
              Gallium 0.4 on llvmpipe: VMware, Inc.
              You're running the software renderer there, instead of the r300g driver.

              Comment


              • #22
                Originally posted by droste View Post
                that output is obviously wrong.

                Beside this error: I'm wondering if the FPS output with the sphere is really right, but the sphere shows more FPS if the window is in the lower right corner (~600FPS). In the upper left corner I get ~250FPS. Both without MLAA. With MLAA it's ~77FPS and ~74FPS so it's not that visible but still the case.
                Window position anomaly - maybe caused by your compositor?

                Timing wrong - possibly. Could also be that you use a newer irrlicht, as the title bar fps is taken from irr.

                Comment


                • #23
                  Originally posted by Drago View Post
                  So...this MLAA is entirely in software/shader, and not actual GPU anti-aliasing method?
                  Yeah. But it's entirely on the GPU, no CPU software part.

                  Comment


                  • #24
                    Irrlicht Engine version 1.7.2
                    Linux 2.6.39-ARCH #1 SMP PREEMPT Mon Jun 27 21:26:22 CEST 2011 x86_64
                    Using renderer: OpenGL 4.1.10834
                    ATI Radeon HD 5700 Series: ATI Technologies Inc.
                    OpenGL driver version is 1.2 or better.
                    GLSL version: 4.1
                    Loaded texture: AreaMap33
                    Unsupported texture format
                    Quit message received.

                    Ran 204.882s, average fps 225.12, min 1

                    Average on fps 1055.90, average off fps 181.66

                    MLAA took on average 0.9ms

                    mesa
                    Version : 7.10.99.git20110706-1
                    Radeon 5770

                    Comment


                    • #25
                      I'm also getting some really odd numbers with the latest version of the test, and this is using the binaries you provided.

                      Comment


                      • #26
                        Originally posted by smitty3268 View Post
                        You're running the software renderer there, instead of the r300g driver.
                        Ops, good catch. Hopefully that explains the really broken display too. So next up is diagnosing why it didn't go with r300g. Sorry for all the noise.

                        Comment


                        • #27
                          Okay, trying again. It's funny how llvmpipe gave me correct numbers with the previous commit; let's see if the latest improves things on real hw.

                          Comment


                          • #28
                            Yeah, the new results seems to be okay:

                            Code:
                            Ran 41.635s, average fps 288.01, min 1
                            
                            Average on fps 130.86, average off fps 583.15
                            (The first test I ran was with power_profile set to low, so quite far off the mark, sorry about that!)

                            Comment


                            • #29
                              No, it was interesting to see how much the low profile affected the performance. Almost exactly four times worse with low.

                              Comment


                              • #30
                                Originally posted by curaga View Post
                                Window position anomaly - maybe caused by your compositor?

                                Timing wrong - possibly. Could also be that you use a newer irrlicht, as the title bar fps is taken from irr.
                                You were right, it was KWins fault. The average FPS is now also correct with the new version.

                                Results with r600g and hd5770 (juniper):

                                low: Average on fps 60.43, average off fps 312.62
                                high: Average on fps 60.38, average off fps 526.58

                                with MLAA on, the FPS was somehow capped at 60FPS. That's probably because SwappbuffersWait is turned on, but why isn't the FPS capped without MLAA?

                                Comment

                                Working...
                                X