来自CUDA Thrust的过程inclusive_scan不适用于浮点数的设备向量

Question

我正在尝试从Thrust库计算device_vector上的就地前缀和。以下代码段不起作用：

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

#include <iostream>
#include <stdio.h>
#include <thrust/scan.h>
#include <thrust/device_vector.h>

int main()
{
    thrust::device_vector<float> vector(4);
    vector[0] = 1.0;
    vector[1] = -3.0;
    vector[2] = 4.0;
    vector[3] = 5.0;

    thrust::inclusive_scan(vector.begin(), vector.end(), vector.begin());
    thrust::host_vector<float> h_vector(vector.begin(), vector.end());

    for (size_t i = 0; i < 4; ++i) {
        std::cout << h_vector[i] << std::endl;
    }
    return 0;
}

它在带有inclusive_scan调用的行上抛出一些异常。我做错了什么？