I grinded this for a couple of nights, all the tips would be helpful:
I installed linux (Debian) on an old laptop and wrote an OpenGL application with it. The cruise circuit is about 80 frames per second when drawing a texture on a small part of the screen. When I paint a texture on a larger and larger part of the screen, the frame rate drops sharply. Full screen texture reduces fps to 12.
I know that hardware can run at a faster speed. Even if openGL runs in software mode, it still seems rather slow.
Any pointers to diagnosing this will be appreciated, thanks. I am new to OpenGL, so this may be something obvious.
Things I tried:
glxinfo informed me that the laptop is capable of rendering hardware accelerated rendering. Although I'm not sure that it does not work in software, I am sure that the laptop is configured to work in hardware.
I removed the depth buffer and I only request a 24-bit screen buffer in the resolution supported by the laptop. (With backbuffer still).
Moving from GL_TEXTURE_MIN_FILTER, GL_LINEAR to GL_TEXTURE_MIN_FILTER, GL_NEAREST bought me a few frames, but he soars just over 20 frames per second, which is still too slow.
I switched from glShadeModel (GL_SMOOTH) to glShadeModel (GL_FLAT). I am not accelerating.
I tried to display the texture as a vertex pointer and as a call to glBegin (GL_QUADS).
Thanks in advance, ~ W
source
share