couldn't find Aa suitable framebuffer configuration with ATI 9800 pro

OpenGL Development JOGL Development
#1

I have a jogl app that uses GLJPanel and works fine on my Inspiron 5150 laptop with an NVIDIA card. I recently started working on it on another machine with a ATI 9800 PRO with fedora and the fglrx-8.23.7-4 drivers. When I start up the app now it givces me the following message:

exception in QueueFlusher:
javax.media.opengl.GLException: context creation error: couldn’t find a suitable frame buffer configuration
at com.sun.opengl.impl.x11.X11ExternalGLDrawable$Context.create(X11ExternalGLDrawable.java:180)
at com.sun.opengl.impl.x11.X11ExternalGLDrawable$Context.makeCurrentImpl(X11ExternalGLDrawable.java:123)
at com.sun.opengl.impl.GLContextImpl.makeCurrent(GLContextImpl.java:118)
at com.sun.opengl.impl.GLDrawableHelper.invokeGL(GLDrawableHelper.java:182)
at javax.media.opengl.GLJPanel$1.run(GLJPanel.java:585)
at sun.java2d.opengl.OGLRenderQueue$QueueFlusher.run(OGLRenderQueue.java:203)

Any idea what I can try. I get the following GLCapabilities:
JOGL GLCapabilities [DoubleBuffered: true, Stereo: false, HardwareAccelerated: true, DepthBits: 24, StencilBits: 0, Red: 8, Green: 8, Blue: 8, Alpha: 8, Red Accum: 0, Green Accum: 0, Blue Accum: 0, Alpha Accum: 0, Multisample: false

#2

Is it possible for you to check out and build the jogl sources?

Something is going wrong in some low-level X11/GLX calls which may ultimately be driver issues.

I note that you’re using Mustang with the Java2D/OpenGL bridge enabled. Have you tried not using the Java2D/OpenGL pipeline (will impact performance, but may improve the stability)?

#3

I turned off the Java2D/OpenGL pipeline and it dosen’t give me that error anymore, but I get one related to the vertex buffer objects that I use (NullPointerException …). I haven’t checked out what that issue is, but it seems to find the correct framebuffer configuration alright. I’m using the beta 4 jogl, I haven’t had the time to checkout the jogl sources yet. Is there anything different since beta 4. I also updated to the kmod-fglrx-8.24.8-1.2.6.16_1.2096_FC5 from the kmod-fglrx-8.23.8-1.2.6.16_1.2096_FC5 driver.

#4

No, there’s nothing different from beta 4, but we don’t have enough tracing information in the source base at this point to figure out what is going on. I could add this and you could test with a subsequent nightly build if that would be easier.

Does it look like the NullPointerException is coming from within JOGL? Could you please post the stack trace?

#5

The tracing nightly build would be nice. Let me know when its available.

No the NullPointerException is to do with the ByteBuffer that I get mapping to the vertex buffer object.

// map buffer into client memory, update buffer, and unmap buffer again
ByteBuffer mappedBuffer = gl.glMapBuffer(GL.GL_PIXEL_UNPACK_BUFFER_ARB, GL.GL_WRITE_ONLY);

#6

Actually now when I look at it, it is the PixelBufferObjects byte byffer that returrns null. The machine with the ATI card is at home so I can’t verify this until later. I’ll post some more details then.

#7

Opps :-[
My code incorrectly tested the available extensions and so it was trying to use pixel buffer objects when the driver did not support it. Both
GL_ARB_pixel_buffer_object and GL_EXT_pixel_buffer_object returned false.

I don’t have code to deal with the alternate case yet so I can’t check that everything else is ok, but I guess it should be. I would have thought that a ATI 9800 Pro with the latest ATI driver would have supported pixel buffer objects, but I guess not.