#include "cuda_runtime.h"
#include "device_launch_parameters.h"

#include <stdio.h>

#define BLOCK_SIZE 2
#define GRID_SIZE  2

__global__ void addKernel(int *x)
{
    x[0]++;
    return;
}

int main(int argc, char** argv)
{
	int i;

	int* x = new int[1];
	int* dev_x;

	int res_cpu;
	int* res_gpu = new int[1];

	cudaMalloc((void**)&dev_x, sizeof(int));

	*x = 4;

	//	Számítás a GPU-n
	cudaMemcpy(dev_x, x, sizeof(int), cudaMemcpyHostToDevice);

	addKernel<<<2, 2>>>(dev_x);

	cudaMemcpy(res_gpu, dev_x, sizeof(int), cudaMemcpyDeviceToHost);

	//	 Számítás a CPU-n
	res_cpu = x[0];
	for(i=0; i<BLOCK_SIZE*GRID_SIZE; i++) {
		res_cpu++;
	}

	printf("Result on GPU: %d\n", *res_gpu);
	printf("Result on CPU: %d\n", res_cpu);
	
	return 0;
}