Hello,
is it possible with the driver API to find out which type at which position in a CUDA function parameter has?
Something like this:
__device__ float M[1000];
extern "C" __global__ void TestFunc(float A, byte B
{
M[0]=(float)A;
M[1]=(float)B;
}
It would be great to have something like this, in order to create a transparent interface for my C# implementations.
Thank you
Martin