封面图加载中

WSL2使用指北

在 Win11 22H2 中重装并配置 WSL Ubuntu-24.04：卸载/安装分发，设置网络与代理，开启 SSH 并安装 CUDA 13.1。
  • 本文作者
  • 文章发布日期 2025-12-18 14:08
  • 热度 23
  • 天气 霾 2°
  • 作者心情 有点愉快
  • 地点 奥林匹克森林公园
  • 本文共计
  • 预计阅读

本人使用的是Win11 22H2版本，目前最新的Win11 25H2版本无法安装虚拟化平台，怎么尝试都无法解决，故降级为22H2。

PS：22H2版本的Win11没有严格的文件权限，且可以轻松关闭内存隔离和基于虚拟化的安全性。使用VMware虚拟机时必须要关闭这两项，而25H2很难通过图形化界面手动关闭。

本文的教程从重装开始，若为第一次安装可跳过开头。

卸载乌班图24.04

分步

查看已安装的 WSL 发行版

# 在 PowerShell 或 Windows 终端中运行
wsl --list --verbose
# 或简写
wsl -l -v

停止 Ubuntu 24.04

wsl --terminate Ubuntu-24.04
# 或使用分发名称
wsl -t Ubuntu-24.04

注销（删除）Ubuntu 24.04

wsl --unregister Ubuntu-24.04

注意：​ 这个命令会永久删除Ubuntu及其所有数据，不可恢复！

完整流程

w安装乌班图24.04

重新安装

分步

以管理员身份打开PowerShell

# 直接安装最新版本
wsl --install -d Ubuntu-24.04

完整流程

PS C:\Users\Skyler> wsl --install -d Ubuntu-24.04
正在下载: Ubuntu 24.04 LTS
正在安装: Ubuntu 24.04 LTS
已成功安装分发。可以通过 “wsl.exe -d Ubuntu-24.04” 启动它
正在启动 Ubuntu-24.04...
Provisioning the new WSL instance Ubuntu-24.04
This might take a while...
Create a default Unix user account: skyler
New password:
Retype new password:
passwd: password updated successfully
To run a command as administrator (user "root"), use "sudo <command>".
See "man sudo_root" for details.

skyler@DESKTOP-L2L5NTC:/mnt/c/Users/Skyler$

可选

设置Ubuntu-24.04为默认发行版

# 使用分发名称
wsl --setdefault Ubuntu-24.04
# 或简写
wsl -s Ubuntu-24.04

网络设置

打开WSL Setting

NAT模式下的IP每次启动可能会改变，故选用Mirrored模式固定IP

代理设置

NAT模式下需要将代理软件的允许局域网连接打开，Mirrored则可忽略这一步

编辑~/.bashrc

nano ~/.bashrc

在最后一行添加

# 设置代理环境变量（端口按实际修改）
export http_proxy=http://localhost:63107
export https_proxy=http://localhost:63107
export all_proxy=socks5://localhost:63107

如果是NAT模式则添加

# 获取 Windows 主机 IP
host_ip=$(ip route | awk '/default/ {print $3}')
# 设置戴笠环境变量（端口按实际修改）
export http_proxy=http://$host_ip:63107
export https_proxy=http://$host_ip:63107
export all_proxy=socks5://$host_ip:63107

生效配置

source ~/.bashrc

验证配置

curl -I www.google.com
 
skyler@DESKTOP-L2L5NTC:/mnt/c/Users/Skyler$ curl -I www.google.com
HTTP/1.1 200 OK
Transfer-Encoding: chunked
Cache-Control: private
Connection: keep-alive
Content-Security-Policy-Report-Only: object-src 'none';base-uri 'self';script-src 'nonce-C9-WCfdHe3GHPoGj3dvQqQ' 'strict-dynamic' 'report-sample' 'unsafe-eval' 'unsafe-inline' https: http:;report-uri https://csp.withgoogle.com/csp/gws/other-hp
Content-Type: text/html; charset=ISO-8859-1
Date: Thu, 18 Dec 2025 05:21:14 GMT
Expires: Thu, 18 Dec 2025 05:21:14 GMT
Keep-Alive: timeout=4
P3p: CP="This is not a P3P policy! See g.co/p3phelp for more info."
Proxy-Connection: keep-alive
Server: gws
Set-Cookie: AEC=AaJma5tMq-RxNzziWYeGHtL7uA6sjTI4o0TDrzCpRm-NOJhupDaWPUjS6A; expires=Tue, 16-Jun-2026 05:21:14 GMT; path=/; domain=.google.com; Secure; HttpOnly; SameSite=lax
Set-Cookie: NID=527=Sp-T9PiHMQ5OyeIVr_NY2QkIaJo0awPOAMHbRJOdUj8Qge_zGi0k6uNXJtZiaDVfBzF9xbTscIo3fNBkr-qWt9Y2qd-X25AwyjCMTPnxQaREp4zf3hgjmZDFuyL_JH2u5Sk7rGuVV5wIK28TZOuhdB5xUyZ-2BoL56ea_sory5KQ-Mgfa97Y-nVJZaDbMcIYDgUDcfGHuZmxGIY8RiV_ok8-dGS289RA5Kw; expires=Fri, 19-Jun-2026 05:21:14 GMT; path=/; domain=.google.com; HttpOnly
X-Frame-Options: SAMEORIGIN
X-Xss-Protection: 0

开启SSH

更新软件包并安装 OpenSSH 服务器

sudo apt update
sudo apt install openssh-server -y

启动 SSH 服务并设置开机自启

sudo service ssh start
sudo systemctl enable ssh

修改端口

sudo nano /etc/ssh/sshd_config

找到并修改以下行

# 修改端口
Port 5643

# 确保允许密码认证（如果不需要可禁用）
PasswordAuthentication yes

# 允许 root 登录（根据需要设置）
PermitRootLogin yes

# 允许公钥认证
PubkeyAuthentication yes

# 允许空密码（通常设为 no）
PermitEmptyPasswords no

重启 SSH 服务

sudo systemctl restart ssh

检查 SSH 是否监听新端口

sudo ss -tlnp | grep ssh
 
root@DESKTOP-L2L5NTC:~# sudo ss -tlnp | grep ssh
LISTEN 0      128           0.0.0.0:5643      0.0.0.0:*    users:(("sshd",pid=2701,fd=3))
LISTEN 0      128              [::]:5643         [::]:*    users:(("sshd",pid=2701,fd=4))

安装Cuda

检查英伟达驱动

nvidia-smi
 
skyler@DESKTOP-L2L5NTC:/mnt/c/Users/Skyler$ nvidia-smi
Thu Dec 18 13:33:12 2025
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 590.44.01              Driver Version: 591.44         CUDA Version: 13.1     |
+-----------------------------------------+------------------------+----------------------+
| 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 RTX 5090 ...    On  |   00000000:02:00.0 Off |                  N/A |
| N/A   39C    P4             19W /   40W |       0MiB /  24463MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+
skyler@DESKTOP-L2L5NTC:/mnt/c/Users/Skyler$

安装对应13.1版本的Cuda

https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&Distribution=WSL-Ubuntu&target_version=2.0&target_type=runfile_local

选择如图所示

接着在下方能看见安装命令

复制即可

wget https://developer.download.nvidia.com/compute/cuda/13.1.0/local_installers/cuda_13.1.0_590.44.01_linux.run
sudo sh cuda_13.1.0_590.44.01_linux.run

逐条复制

skyler@DESKTOP-L2L5NTC:~$ wget https://developer.download.nvidia.com/compute/cuda/13.1.0/local_installers/cuda_13.1.0_590.44.01_linux.run
--2025-12-18 13:37:57--  https://developer.download.nvidia.com/compute/cuda/13.1.0/local_installers/cuda_13.1.0_590.44.01_linux.run
Resolving localhost (localhost)... 127.0.0.1
Connecting to localhost (localhost)|127.0.0.1|:63107... connected.
Proxy request sent, awaiting response... 200 OK
Length: 4345889047 (4.0G) [application/octet-stream]
Saving to: ‘cuda_13.1.0_590.44.01_linux.run’

cuda_13.1.0_590.44.01_linux.r 100%[=================================================>]   4.05G  67.2MB/s    in 67s

2025-12-18 13:39:10 (61.9 MB/s) - ‘cuda_13.1.0_590.44.01_linux.run’ saved [4345889047/4345889047]

在运行sh cuda_13.1.0_590.44.01_linux.run 之前要安装编译使用的gcc

sudo apt-get install build-essential -y

运行安装脚本

sudo apt-get install build-essential -y

输入accept并回车

┌──────────────────────────────────────────────────────────────────────────────┐
│  End User License Agreement                                                  │
│  --------------------------                                                  │
│                                                                              │
│  NVIDIA Software License Agreement and CUDA Supplement to                    │
│  Software License Agreement.                                                 │
│                                                                              │
│  The CUDA Toolkit End User License Agreement applies to the                  │
│  NVIDIA CUDA Toolkit, the NVIDIA CUDA Samples, the NVIDIA                    │
│  Display Driver, NVIDIA Nsight tools (Visual Studio Edition),                │
│  and the associated documentation on CUDA APIs, programming                  │
│  model and development tools. If you do not agree with the                   │
│  terms and conditions of the license agreement, then do not                  │
│  download or use the software.                                               │
│                                                                              │
│  Last updated: January 12, 2025                                              │
│                                                                              │
│                                                                              │
│  Preface                                                                     │
│  -------                                                                     │
│                                                                              │
│──────────────────────────────────────────────────────────────────────────────│
│ Do you accept the above EULA? (accept/decline/quit):                         │
│ accept                                                                       │
└──────────────────────────────────────────────────────────────────────────────┘

默认勾选即可，选择Install

──────────────────────────────────────────────────────────────────────────────┐
│ CUDA Installer                                                               │
│ + [X] CUDA Toolkit 13.1                                                      │
│   [ ] CUDA Documentation 13.1                                                │
│ - [ ] Kernel Objects                                                         │
│      [ ] nvidia-fs                                                           │
│   Options                                                                    │
│   Install                                                                    │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│                                                                              │
│ Up/Down: Move | Left/Right: Expand | 'Enter': Select | 'A': Advanced options │
└──────────────────────────────────────────────────────────────────────────────┘

安装后提示

skyler@DESKTOP-L2L5NTC:~$ sudo sh cuda_13.1.0_590.44.01_linux.run
===========
= Summary =
===========

Driver:   Not Selected
Toolkit:  Installed in /usr/local/cuda-13.1/

Please make sure that
 -   PATH includes /usr/local/cuda-13.1/bin
 -   LD_LIBRARY_PATH includes /usr/local/cuda-13.1/lib64, or, add /usr/local/cuda-13.1/lib64 to /etc/ld.so.conf and run ldconfig as root

To uninstall the CUDA Toolkit, run cuda-uninstaller in /usr/local/cuda-13.1/bin
***WARNING: Incomplete installation! This installation did not install the CUDA Driver. A driver of version at least 590.00 is required for CUDA 13.1 functionality to work.
To install the driver using this installer, run the following command, replacing <CudaInstaller> with the name of this run file:
    sudo <CudaInstaller>.run --silent --driver

Logfile is /var/log/cuda-installer.log

使用脚本配置环境

sudo tee /etc/profile.d/cuda.sh << 'EOF'
#!/bin/bash

# CUDA 环境变量配置
# 自动检测并设置 CUDA 路径

# 检查 CUDA 13.1 是否存在
if [ -d "/usr/local/cuda-13.1" ]; then
    export CUDA_HOME=/usr/local/cuda-13.1
    export PATH=$PATH:$CUDA_HOME/bin
    export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_HOME/lib64
    export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_HOME/extras/CUPTI/lib64
    echo "已设置 CUDA 13.1 环境变量"
    
# 检查 CUDA 11.4 是否存在
elif [ -d "/usr/local/cuda-11.4" ]; then
    export CUDA_HOME=/usr/local/cuda-11.4
    export PATH=$PATH:$CUDA_HOME/bin
    export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_HOME/lib64
    export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_HOME/extras/CUPTI/lib64
    echo "已设置 CUDA 11.4 环境变量"
    
# 检查默认 CUDA 链接是否存在
elif [ -L "/usr/local/cuda" ]; then
    export CUDA_HOME=/usr/local/cuda
    export PATH=$PATH:$CUDA_HOME/bin
    export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_HOME/lib64
    export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_HOME/extras/CUPTI/lib64
    CUDA_VERSION=$(basename $(readlink -f /usr/local/cuda))
    echo "已设置 $CUDA_VERSION 环境变量"
fi
EOF

设置权限

sudo chmod +x /etc/profile.d/cuda.sh

重新加载配置

source /etc/profile

验证

nvcc --version
 
skyler@DESKTOP-L2L5NTC:~$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2025 NVIDIA Corporation
Built on Fri_Nov__7_07:23:37_PM_PST_2025
Cuda compilation tools, release 13.1, V13.1.80
Build cuda_13.1.r13.1/compiler.36836380_0

安装cuDNN

打开官网链接

cuDNN 9.17.0 Downloads |

https://developer.nvidia.com/cudnn-downloads?target_os=Linux&target_arch=x86_64&Distribution=Ubuntu&target_version=24.04&target_type=deb_local&Configuration=Full

选择如图

逐行输入命令即可

wget https://developer.download.nvidia.com/compute/cudnn/9.17.0/local_installers/cudnn-local-repo-ubuntu2404-9.17.0_1.0-1_amd64.deb
sudo dpkg -i cudnn-local-repo-ubuntu2404-9.17.0_1.0-1_amd64.deb
sudo cp /var/cudnn-local-repo-ubuntu2404-9.17.0/cudnn-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get -y install cudnn

安装Miniconda

打开官网

https://www.anaconda.com/download/success

下载如图

下载完后复制到WSL，后运行脚本（xxx为文件全名）

bash xxx.sh