Using RKNN-ToolKit2 in WSL
1. Installing WSL on Windows Host
- You can refer to the following link for installing WSL: https://learn.microsoft.com/en-us/windows/wsl/install
2. Using RKNN-Toolkit2 in WSL
- Refer to the document《Rockchip_RKNPU_Quick_Start》 to install the RKNN-ToolKit2 environment in WSL
- Refer to the document《Rockchip_RKNPU_User_Guide》 for model conversion, quantization, and other operations in WSL.
3. Using RKNN-Toolkit2 for Board Debugging in WSL
3.1. Installing adb in WSL Terminal
sudo apt update
sudo apt install adb
3.2 Connecting the Device in WSL
You can choose to connect the device through Ethernet or USB.
3.2.1 Ethernet Connection
# 1. Connect the device using Ethernet cable.
# 2. In WSL, use adb connect to connect to the device.
adb connect <IP address:port>
IP address
is the IP address of the board.
3.2.2 USB Connection
# 1. Connect the device to Windows via USB.
# 2. Use the adbkit tool in Windows to convert USB to TCP.
npm install --save adbkit
adbkit usb-device-to-tcp <device_id> -p <port>
# Note: Configure WSL to access Windows network.
# 3. In WSL, use adb connect to connect to the device.
adb connect <IP address:port>
device_id
can be obtained by running theadb devices
command in Windows (adb tool must be installed in Windows).IP address
is the IP address of the Windows host.
3.3 Using RKNN-Toolkit2 for Board Debugging
Refer to the document 《Rockchip_RKNPU_User_Guide》 for board inference, accuracy analysis, and other operations in WSL.
NOte:
- It is recommended to install WSL2 with Ubuntu version 22.04, which has been verified to work (other versions are theoretically feasible but not tested).
- If you encounter the "ImportError: libGL.so.1: cannot open shared object file: No such file or directory" error while using RKNN-ToolKit2 in WSL, execute the following code to resolve it:
1. Install the required library:
sudo apt update
sudo apt install libgl1-mesa-glx
2. Set the environment variable:
echo 'export LD_LIBRARY_PATH=/usr/lib/x86_64-linux-gnu/mesa' >> ~/.bashrc
source ~/.bashrc