CUDA二维矩阵加法

#include<cuda_runtime_api.h>
#include<device_launch_parameters.h>
#include<stdio.h>
__global__ void add(int a[][2], int b[][2], int c[][2])
{
	int i = threadIdx.x + blockIdx.x * blockDim.x;
	int j = threadIdx.y + blockIdx.y * blockDim.y;
	if (i < 2 && j < 2)
	{
		c[i][j] = a[i][j] + b[i][j];
	}
}
int main()
{
	int (*a)[2] = new int[2][2];
	int (*b)[2] = new int[2][2];
	int (*c)[2] = new int[2][2];

	a[0][0] = 1; a[0][1] = 2; a[1][0] = 3; a[1][1] = 4;
	b[0][0] = 1; b[0][1] = 2; b[1][0] = 3; b[1][1] =4;

	cudaError_t error = cudaSuccess;

	int (*device_a)[2],(*device_b)[2],(*device_c)[2];

	error = cudaMalloc((void **)&device_a, sizeof(int)* 4);
	error = cudaMalloc((void **)&device_b, sizeof(int)* 4);
	error = cudaMalloc((void **)&device_c, sizeof(int)* 4);

	cudaMemcpy(device_a, a, sizeof(int)* 4, cudaMemcpyHostToDevice);
	cudaMemcpy(device_b, b, sizeof(int)* 4, cudaMemcpyHostToDevice);

	//    dim3 threadsPerBlock(1, 1);
	//    dim3 numBlocks(2 / threadsPerBlock.x, 2 / threadsPerBlock.y);
	dim3 blocks(1, 1);
	dim3 threads(2, 2);
	add << <blocks, threads>> >(device_a, device_b, device_c);

	cudaMemcpy(c, device_c, sizeof(int)* 4, cudaMemcpyDeviceToHost);

	for (int i = 0; i < 2; i++)
	{
	for (int j = 0; j < 2; j++)
		printf("%10d", c[i][j]);
	printf("\n");
	}

	return 0;
}
<img src="https://img-blog.csdn.net/20150914122105544?watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQv/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/Center" height="355" width="555" alt="" />