I want to pass an integer as an attribute into a shader, which is then used to index a uniform. However, it seems to fail on my nVidia card and garbage is returned.
However, once i switch to a float-attribute and cast it to an int in the shader, everything works fine.
Doing the same thing in D3D11 works both with integers and floats.
Did anyone encounter similar problems? Are integer attributes problematic in any way? Or could I have forgotten to enable some state, that is not necessary when using floats?