test_macros.c

#define CUDA_1D_KERNEL_LOOP(i, n)                            \
  for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; \
       i += blockDim.x * gridDim.x)

CUDA_1D_KERNEL_LOOP(index, nthreads){

}