CUDA 实现qr分解，并得到Q矩阵和R矩阵，任意尺寸均可,并且实现了Q矩阵和R矩阵格式转换

kernal.cu

#include <stdio.h>
#include <stdlib.h>
#include "cuda_runtime.h"
#include "device_launch_parameters.h"
#include <cusolverDn.h>

extern "C" int GET_QR(int ncpt)
{
	cusolverDnHandle_t cusolverH = NULL;

	const int m = 3;
	const int lda = m;
	float *A = (float*)malloc(lda*m * sizeof(float));

	//for (size_t i = 0; i < m; i++)
	//{
	//	for (size_t j = 0; j < m; j++)
	//	{
	//		A[i*m+j]=j*2;
	//	}
	//}
	A[0] = 12;
	A[1] = 6;
	A[2] = -4;
	A[3] = -51;
	A[4] = 167;
	A[5] = 24;
	A[6] = 4;
	A[7] = -68;
	A[8] = -41;
	float *W;
	W = (float*)malloc(lda*m * sizeof(float)); // eigenvalues最终保存结果
	int info_gpu = 0;//计算状态保存
	// 步骤1：声明句柄
	cusolverDnCreate(&cusolverH);

	// 步骤2：分配显存空间
	float *d_A = NULL; cudaMalloc((void**)&d_A, sizeof(float) * lda * m);//声明Hermite矩阵（与计算后的特征向量为同一空间）
	float *d_W =