Cuda 袖口双到复杂
我想用CuFFT库做一个从double到std::complex的FFT。我的代码看起来像Cuda 袖口双到复杂,cuda,cufft,Cuda,Cufft,我想用CuFFT库做一个从double到std::complex的FFT。我的代码看起来像 #include <complex> #include <iostream> #include <cufft.h> #include <cuda_runtime_api.h> typedef std::complex<double> Complex; using namespace std; int main(){ int n = 100
#include <complex>
#include <iostream>
#include <cufft.h>
#include <cuda_runtime_api.h>
typedef std::complex<double> Complex;
using namespace std;
int main(){
int n = 100;
double* in;
Complex* out;
in = (double*) malloc(sizeof(double) * n);
out = (Complex*) malloc(sizeof(Complex) * n/2+1);
for(int i=0; i<n; i++){
in[i] = 1;
}
cufftHandle plan;
plan = cufftPlan1d(&plan, n, CUFFT_D2Z, 1);
unsigned int mem_size = sizeof(double)*n;
cufftDoubleReal *d_in;
cufftDoubleComplex *d_out;
cudaMalloc((void **)&d_in, mem_size);
cudaMalloc((void **)&d_out, mem_size);
cudaMemcpy(d_in, in, mem_size, cudaMemcpyHostToDevice);
cudaMemcpy(d_out, out, mem_size, cudaMemcpyHostToDevice);
int succes = cufftExecD2Z(plan,(cufftDoubleReal *) d_in,(cufftDoubleComplex *) d_out);
cout << succes << endl;
cudaMemcpy(out, d_out, mem_size, cudaMemcpyDeviceToHost);
for(int i=0; i<n/2; i++){
cout << "out: " << i << " " << out[i].real() << " " << out[i].imag() << endl;
}
return 0;
}
#包括
#包括
#包括
#包括
typedef-std::复杂复合体;
使用名称空间std;
int main(){
int n=100;
双*英寸;
复杂*out;
in=(双*)malloc(sizeof(双)*n);
out=(复合物*)malloc(复合物的尺寸)*n/2+1);
对于(inti=0;i,您的代码有各种错误。您可能应该查看示例代码
您应该对所有API返回值执行适当的cuda错误检查和适当的CUFT错误检查
cufftPlan1d
函数的返回值不进入计划:
plan = cufftPlan1d(&plan, n, CUFFT_D2Z, 1);
函数本身设置计划(这就是为什么您将&plan
传递给函数),然后当您将返回值赋给计划时,它会破坏函数设置的计划
您正确地标识了输出的大小可以是((N/2)+1)
,但是您没有在主机端为其正确分配空间:
out = (Complex*) malloc(sizeof(Complex) * n/2+1);
或在设备端:
unsigned int mem_size = sizeof(double)*n;
...
cudaMalloc((void **)&d_out, mem_size);
下面的代码修复了上面的一些问题,足以得到您想要的结果(100,0,0,…)
#包括
#包括
#包括
#包括
#定义cudaCheckErrors(msg)\
做{\
cudaError\u t\u err=cudaGetLastError()\
如果(_err!=cudaSuccess){\
fprintf(标准,“致命错误:%s(%s位于%s:%d)\n”\
msg,cudaGetErrorString(_err)\
__文件(行)\
fprintf(stderr,“***失败-中止\n”)\
出口(1)\
} \
}而(0)
typedef-std::复杂复合体;
使用名称空间std;
int main(){
int n=100;
双*英寸;
复杂*out;
#ifdef到位
in=(双*)malloc(sizeof(Complex)*(n/2+1));
out=(复数*)in;
#否则
in=(双*)malloc(sizeof(双)*n);
out=(复合物*)malloc(复合物)*(n/2+1)的大小);
#恩迪夫
对于(int i=0;iThanks),这非常有用。你是对的,我在错误检查方面非常懒惰。第2点非常糟糕,因为我今天早些时候犯了同样的错误,知道这是错误的,但没有注意到这个错误。再次感谢你,我以后会做得更好。
#include <complex>
#include <iostream>
#include <cufft.h>
#include <cuda_runtime_api.h>
#define cudaCheckErrors(msg) \
do { \
cudaError_t __err = cudaGetLastError(); \
if (__err != cudaSuccess) { \
fprintf(stderr, "Fatal error: %s (%s at %s:%d)\n", \
msg, cudaGetErrorString(__err), \
__FILE__, __LINE__); \
fprintf(stderr, "*** FAILED - ABORTING\n"); \
exit(1); \
} \
} while (0)
typedef std::complex<double> Complex;
using namespace std;
int main(){
int n = 100;
double* in;
Complex* out;
#ifdef IN_PLACE
in = (double*) malloc(sizeof(Complex) * (n/2+1));
out = (Complex*)in;
#else
in = (double*) malloc(sizeof(double) * n);
out = (Complex*) malloc(sizeof(Complex) * (n/2+1));
#endif
for(int i=0; i<n; i++){
in[i] = 1;
}
cufftHandle plan;
cufftResult res = cufftPlan1d(&plan, n, CUFFT_D2Z, 1);
if (res != CUFFT_SUCCESS) {cout << "cufft plan error: " << res << endl; return 1;}
cufftDoubleReal *d_in;
cufftDoubleComplex *d_out;
unsigned int out_mem_size = (n/2 + 1)*sizeof(cufftDoubleComplex);
#ifdef IN_PLACE
unsigned int in_mem_size = out_mem_size;
cudaMalloc((void **)&d_in, in_mem_size);
d_out = (cufftDoubleComplex *)d_in;
#else
unsigned int in_mem_size = sizeof(cufftDoubleReal)*n;
cudaMalloc((void **)&d_in, in_mem_size);
cudaMalloc((void **)&d_out, out_mem_size);
#endif
cudaCheckErrors("cuda malloc fail");
cudaMemcpy(d_in, in, in_mem_size, cudaMemcpyHostToDevice);
cudaCheckErrors("cuda memcpy H2D fail");
res = cufftExecD2Z(plan,d_in, d_out);
if (res != CUFFT_SUCCESS) {cout << "cufft exec error: " << res << endl; return 1;}
cudaMemcpy(out, d_out, out_mem_size, cudaMemcpyDeviceToHost);
cudaCheckErrors("cuda memcpy D2H fail");
for(int i=0; i<n/2; i++){
cout << "out: " << i << " " << out[i].real() << " " << out[i].imag() << endl;
}
return 0;
}