This project provides a simple, virtual network driver for the Linux kernel. It is primarily intended for educational purposes and demonstrates various concepts such as kernel module programming, network device operations, and ethtool operations.
The virtual network driver can be loaded and unloaded using standard Linux kernel module utilities, such as insmod and rmmod. It also provides basic networking functionality, allowing users to interact with the virtual network device using standard network commands like ifconfig and iwconfig.
To compile a Loadable Kernel Module, you need the kernel headers matching the kernel version you are working with. To resolve this issue, install the kernel headers and development tools for your Linux distribution. For Fedora, you can use the following command:
sudo dnf install kernel-devel kernel-headers gcc make iw hostapd
To build the virtual network driver module, update the Makefile to use the correct path for the kernel build directory. The build directory is typically found under /lib/modules/$(uname -r)/build, where $(uname -r) returns the currently running kernel version. Finally, run the make command from the project directory.
make
This will generate a kernel module file called virt_net_driver.ko.
Load the cfg80211 kernel module:
sudo modprobe cfg80211
Load the virtual network driver module using insmod:
sudo insmod virt_net_driver.ko
To check if the module has been loaded successfully, run:
lsmod | grep virt_net_driver
Check if the virtual interfaces were created:
ifconfig
You should see three new interfaces like the example below:
vif0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 00:76:69:66:30:00 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
vif1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 00:76:69:66:31:00 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
vif2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 00:76:69:66:32:00 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
Check if the wifi devices have been properly configured. You should see the same interface names as the ifonfig output.
sudo iw dev
sudo iw list
Check if the get_station function works:
sudo iw dev vif0 station get 00:76:69:66:30:00
The output should look something like:
Station 00:76:69:66:30:00 (on vif0)
rx bytes: 1
rx packets: 1
tx bytes: 1
tx packets: 1
tx failed: 1
current time: 1682978276569 ms
To test if the scan and connect functions work properly, we must put each interface in their own network namespaces.
Add network namespaces:
sudo ip netns add v0
sudo ip netns add v1
sudo ip netns add v2
run sudo iw dev again and look and note the physical device numbers (phy#{NUMBER}). There should be three consecutive physical numbers (e.g. phy#0 phy#1 phy#2). Use these physical device numbers to add each device to its respective namespace:
sudo iw phy phy0 set netns name v0
sudo iw phy phy1 set netns name v1
sudo iw phy phy2 set netns name v2
Ensure all the interfaces are UP:
sudo ip netns exec v0 ip link set vif0 up
sudo ip netns exec v1 ip link set vif1 up
sudo ip netns exec v2 ip link set vif2 up
Configure vif0 to be an access point:
sudo ip netns exec v0 hostapd -B hostapd.conf
Check if vif0 has been successfully configured as an AP:
sudo ip netns exec v0 iw dev
The output should say ssid VAP and type AP:
phy#0
Interface vif0
ifindex 3
wdev 0x1
addr 00:76:69:66:30:00
ssid VAP
type AP
Give each interface an IP:
sudo ip netns exec v0 ip addr add 10.10.10.1/24 dev vif0
sudo ip netns exec v1 ip addr add 10.10.10.2/24 dev vif1
sudo ip netns exec v2 ip addr add 10.10.10.3/24 dev vif2
Perform a scan from vif1:
sudo ip netns exec v1 iw dev vif1 scan
Connect to vif0 from vif1:
sudo ip netns exec v1 iw dev vif1 connect VAP
Check vif1’s link status:
sudo ip netns exec v1 iw dev vif1 link
Remove the virtual network driver module:
sudo rmmod virt_net_driver
Test packet transfer functionality. We did not implement packet transfers between APs and STAs, but packet transfer between two STAs on the same namespace do work.
Re-insert the module:
sudo insmod virt_net_driver.ko
Give each interface each IP
sudo ip addr add 10.10.10.1/24 dev vif0
sudo ip addr add 10.10.10.2/24 dev vif1
sudo ip addr add 10.10.10.3/24 dev vif2
Ping v2 from v1:
ping -I 10.10.10.2 10.10.10.1
This project was developed and tested using Visual Studio Code on Fedora Core 38 with Linux Kernel 6.2.11. To set up your Visual Studio Code environment for C/C++ development, debugging, and linting, follow these steps:
C/C++ (by Microsoft)C/C++ Advanced Lint (by JBenden)Native Debug (by WebFreak)c_cpp_properties.json file in the .vscode directory of your workspace. Add the following content to the file, adjusting the compilerPath and includePath as needed:{
"configurations": [
{
"name": "Linux",
"includePath": [
"${workspaceFolder}/**",
"/usr/lib/gcc/x86_64-linux-gnu/9/include"
],
"defines": [],
"compilerPath": "/usr/bin/gcc",
"cStandard": "c11",
"cppStandard": "c++17",
"intelliSenseMode": "gcc-x64"
}
],
"version": 4
}
launch.json file in the .vscode directory. Add the following content to the file, adjusting the program and miDebuggerPath as needed:{
"version": "0.2.0",
"configurations": [
{
"name": "Debug Kernel Module",
"type": "cppdbg",
"request": "launch",
"program": "/usr/src/linux-headers-$(uname -r)/scripts/gdb/vmlinux-gdb.py",
"args": [],
"stopAtEntry": false,
"cwd": "${workspaceFolder}",
"environment": [],
"externalConsole": false,
"MIMode": "gdb",
"miDebuggerPath": "/usr/bin/gdb",
"setupCommands": [
{
"description": "Enable pretty-printing for gdb",
"text": "-enable-pretty-printing",
"ignoreFailures": true
}
],
"preLaunchTask": "loadModule"
}
]
}
tasks.json file in the .vscode directory. Add the following content to the file:{
"version": "2.0.0",
"tasks": [
{
"label": "build",
"type": "shell",
"command": "make",
"group": {
"kind": "build",
"isDefault": true
},
"problemMatcher": []
},
{
"label": "loadModule",
"type": "shell",
"command": "sudo insmod virt_net_driver.ko",
"dependsOn": "build",
"problemMatcher": []
}
]
}
F5 in Visual Studio Code.If you encounter issues with the virtual network driver, follow these basic debugging steps:
dmesg command for any error messages or warnings related to the virtual network driver.printk function in the virtual network driver source code to output debug information to the kernel log. Make sure to recompile and reload the module after adding or modifying any printk statements.modinfo command to inspect the virtual network driver module and ensure that it was built with the correct kernel version:modinfo virt_net_driver.ko