Usage of global vs. constant memory in CUDA

Question

Hey there, I have the following piece of code:

#if USE_CONST == 1
    __constant__ double PNT[ SIZE ];    
#else
    __device__ double *PNT;
#endif
         


        

Answer 1

The correct usage of cudaMemcpyToSymbol prior to CUDA 4.0 is:

cudaMemcpyToSymbol("PNT", point, sizeof(double)*SIZE)

or alternatively:

double *cpnt;
cudaGetSymbolAddress((void **)&cpnt, "PNT");
cudaMemcpy(cpnt, point, sizeof(double)*SIZE, cudaMemcpyHostToDevice);

which might be a bit faster if you are planning to access the symbol from the host API more than once.

EDIT: misunderstood the question. For the global memory version, do something similar to the second version for constant memory

double *gpnt;
cudaGetSymbolAddress((void **)&gpnt, "PNT");
cudaMemcpy(gpnt, point, sizeof(double)*SIZE.  cudaMemcpyHostToDevice););