Problem with glNewObjectBufferATI. . .

[Ostsol]

Ok, I've look all over and found out how to use extensions. However, I still have a problem getting one of the functions to work: glNewObjectBufferATI. Using the debugger I've found that wglGetProcAddress can't seem to find the address of the function when I'm initializing it. The result always comes out as zero -- which I know can't be right.

// in a header:
typedef GLuint (APIENTRY * PFNGLNEWOBJECTBUFFERATIPROC) (GLsizei size, const GLvoid *pointer, GLenum usage);

PFNGLNEWOBJECTBUFFERATIPROC glNewObjectBufferATI = NULL;

// in the .cpp:
glNewObjectBufferATI = (PFNGLNEWOBJECTBUFFERATIPROC) wglGetProcAddress ("glNewObjectBufferATI");

if (glNewObjectBufferATI)
gluiVertexObject = glNewObjectBufferATI (objectsize, polygonlist.vertlist, GL_STATIC_ATI);

Using the same method, I can get glArrayObjectATI to work. Is there something I'm missing?

 

[NitroGL]

Re: Problem with glNewObjectBufferATI. . .

Ok, I've look all over and found out how to use extensions. However, I still have a problem getting one of the functions to work: glNewObjectBufferATI. Using the debugger I've found that wglGetProcAddress can't seem to find the address of the function when I'm initializing it. The result always comes out as zero -- which I know can't be right.
Using the same method, I can get glArrayObjectATI to work. Is there something I'm missing?

Stupid question: Do you have a proper OpenGL rendering context?

 

[Ostsol]

Re: Re: Problem with glNewObjectBufferATI. . .

Re: Re: Problem with glNewObjectBufferATI. . .

Stupid question: Do you have a proper OpenGL rendering context?

Hmm, okay. . . I had some stuff in the wrong order (created the rendering context -after- trying to initialize the glNewObjectArrayATI) and fixing that, I can get the address of glNewObjectArrayATI.

Now, however, I have the bug I mentioned in the Radeon 8500 General Discussion forum: glNewObjectArrayATI always returns zero! The result of that is an Access Violation error when I get to glDrawElements (...). It even happens in the ATI demos that use this extension. I think it might be a driver bug (which is why on the other forum I asked for help in confirming it). If it is a driver bug, I guess it's back to 9013's for me. . .

 

[ehart]

Have you double checked to make sure that the extension string is exported? (This extension wasn't available for a while on Win98)

If this checks out, you should contact developer relations to report the issue. ([email protected])

-Evan

 

[Ostsol]

Have you double checked to make sure that the extension string is exported? (This extension wasn't available for a while on Win98)

If this checks out, you should contact developer relations to report the issue. ([email protected])

-Evan

Yep, everything was coded correctly, as far as I can tell.

I've just switched back to the recently released, WHQL certified 9016 drivers and it now works. I receive back the correct address for glNewObjectBufferATI and the result it gives back to me is no longer 0. The ATi demos that wouldn't work under the 9021s work under the 9016s, too.