这是 WSL2 的默认 nvidia 驱动程序吗？

noob@LAPTOP-DNCQ5AAC:/mnt/d/gatechFall2022/ihpc/lab1fa22/part2$ nvidia-smi
Thu Apr 20 00:04:03 2023       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.100      Driver Version: 528.76       CUDA Version: 12.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  On   | 00000000:01:00.0 Off |                  N/A |
| N/A   41C    P8     2W /  50W |      0MiB /  4096MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A       227      G   /Xwayland                       N/A      |
+-----------------------------------------------------------------------------+

它说的是，“小心不要安装另一个覆盖默认设置的”在这里https://docs.nvidia.com/cuda/wsl-user-guide/index.html

在我开始尝试在我的 Windows 11 机器上再次安装 cuda 之前，这是我的默认设置吗？有没有人有关于如何在 WSL2 上安装 cuda 的好答案的链接？
我可能不得不搜索我自己的问题，但请记住它从未真正按预期工作，我不得不使用 Windows 系统。

我想用 Cuda 执行的任务如下：

/*
My Chatgpt prompt:
"implement helman-jaja list rank where the function signature is "void cudaListRank (long head, const long* next, long* rank, size_t n)" where the first parameter head is the index into array representation of a linked list and points to its head,  the second parameter next is a "next array" that gives the index of the next node in the array representation of a linked list, the third parameter rank is an array holding the rank of each node in the array representation of a linked list and the fourth parameter n is the length of the array representation of a linked list."
The code written by ChatGPT below:
*/

#include <unistd.h>
#include "listutils.h"                        
#include <math.h>
#include <cuda.h>
#include <curand.h>

void cudaListRank(long head, const long* next, long* rank, size_t n) {
    // Allocate memory for the linked list nodes and next array on the device
    long* devNext;
    long* devRank;
    cudaMalloc((void**)&devNext, n * sizeof(long));
    cudaMalloc((void**)&devRank, n * sizeof(long));

    // Copy the input linked list next array from host to device
    cudaMemcpy(devNext, next, n * sizeof(long), cudaMemcpyHostToDevice);

    // Initialize rank array on the device to all 1's
    cudaMemset(devRank, 1, n * sizeof(long));

    // Launch a CUDA kernel that performs the scan phase of Helman-Jaja list rank
    int block_size = 256;
    int num_blocks = (n + block_size - 1) / block_size;
    helmanJajaScan<<<num_blocks, block_size>>>(head, devNext, devRank);

    // Launch another CUDA kernel that performs the increment phase of Helman-Jaja list rank
    helmanJajaIncrement<<<num_blocks, block_size>>>(devNext, devRank);

    // Copy the updated rank array from device to host
    cudaMemcpy(rank, devRank, n * sizeof(long), cudaMemcpyDeviceToHost);

    // Free the allocated memory on the device
    cudaFree(devNext);
    cudaFree(devRank);
}


//more kernels
/*
Note that in helmanJajaScan, we're traversing the linked list one node at a time and accumulating the ranks. In helmanJajaIncrement, we're traversing the list in reverse order and adding the accumulated ranks to each node's rank.

Also note that cudaMemcpyHostToDevice and cudaMemcpyDeviceToHost are used for transferring data between host and device memory. Finally, cudaMalloc and cudaFree are used to allocate and free device memory respectively.
*/
__global__ void helmanJajaScan(long head, const long* next, long* rank) {
    int tid = threadIdx.x + blockIdx.x * blockDim.x;

    // Traverse the linked list
    while (head != -1) {
        long curr = head;
        for (int i = 0; i < tid && curr != -1; i++) {
            curr = next[curr];
        }

        if (curr == -1 || next[curr] == -1) break;

        rank[next[curr]] += rank[curr];

        __syncthreads();
    }
}

__global__ void helmanJajaIncrement(const long* next, long* rank) {
    int tid = threadIdx.x + blockIdx.x * blockDim.x;

    // Traverse the linked list
    while (tid < gridDim.x * blockDim.x) {
        long curr = tid;
        for (int i = 0; i < blockDim.x && curr != -1; i++) {
            curr = next[curr];
        }

        if (curr == -1) break;

        rank[curr] += rank[next[curr]];

        tid += gridDim.x * blockDim.x;
    }
}

现在，当我尝试使用 Make 文件进行编译时，我得到了

noob@LAPTOP-DNCQ5AAC:/mnt/d/gatechFall2022/ihpc/lab1fa22/part2$ make correctness IMPL=hj
nvcc -Iutils -O0 -g -std=c++11 -o student/cuda_hj.o -c student/cuda_hj.cu
make: nvcc: Not a directory
make: *** [Makefile:8: student/cuda_hj.o] Error 127

此外，我在 VSCode 中看到与 cuda 相关的导入语句的红色波浪下划线，但不确定这是否表明我的 WSL-remote-VSCode 和默认 Cuda 是否运行良好。

我还从 Nvidia 下载并尝试安装两个不同的 cuda 包（WSL 版本）进行安装，但都没有完成工作（我从下载页面复制并粘贴了每个终端指令）：

cuda_12.1.0_530.30.02_linux.run
cuda-repo-wsl-ubuntu-12-1-local_12.1.0-1_amd64.deb