Why is this cuda program printing 0? - cuda

Question

This code is printing "Result is 0", it seems not to be copying the results from the device. Why is this ?
__global__ void add_arrays(int *a, int *b, int *c, int size){
int index = threadIdx.x + blockIdx.x * blockDim.x;
//avoid accessing beyond the end of the array
if (index < size)
{ c[index] = a[index] + b[index]; }
}
#define N (2048 * 2048)
#define THREADS_PER_BLOCK 512
int main()
{
int *a, *b, *c; // local (host) copy
int *d_a, *d_b, *d_c; // device copy
int size = N * sizeof(int);
//allocating space for device copies ON THE DEVICE
CUDA_CHECK_RETURN(cudaMalloc((void **)&d_a, size));
CUDA_CHECK_RETURN(cudaMalloc((void **)&d_b, size));
CUDA_CHECK_RETURN(cudaMalloc((void **)&d_c, size));
//allocating space for local copies ON THE HOST and initialize: a, b
a = (int *)malloc(size); random_ints(a, N);
b = (int *)malloc(size); random_ints(b, N);
c = (int *)malloc(size); //random_ints(c, N);
//copy inputs to device
CUDA_CHECK_RETURN(cudaMemcpy(d_a, &a, size, cudaMemcpyHostToDevice));
CUDA_CHECK_RETURN(cudaMemcpy(d_b, &b, size, cudaMemcpyHostToDevice));
int num_block = (N + THREADS_PER_BLOCK - 1) / THREADS_PER_BLOCK;
add_arrays<<<num_block, THREADS_PER_BLOCK>>>(d_a, d_b, d_c, N);
// Copy result back from Device to Host
CUDA_CHECK_RETURN(cudaMemcpy(c, d_c, size, cudaMemcpyDeviceToHost));
for (int i =0; i<N; ++i) {
printf("Result is %d \n", *(c+i));
}

Answer 1

You may want to consider posting a complete reproducible example with your question so that it contains your definition of CUDA_CHECK_RETURN, random_ints, included headers, etc. As mentioned in the comments, the obvious problem in your code though is that
cudaMemcpy(d_a, &a, size, cudaMemcpyHostToDevice)
should be
cudaMemcpy(d_a, a, size, cudaMemcpyHostToDevice)
and similarly for b as a and b are pointers.