Skip to content

Running Linux on QEMU

Jump to bottom
Yuriy Kolerov edited this page Aug 4, 2023 · 2 revisions

Preface

It's possible to run on QEMU the same Linux images as for nSIM/HAPS. A process of building of Linux images is completely same.

Running a Kernel for ARC HS 3x/4x

$ qemu-system-arc -M virt -cpu archs -display none -nographic -monitor none -m 2G -kernel vmlinux

Running a Kernel for ARC HS 5x

ℹ️ Linux kernels for ARCv3 are loaded differently in comparison to kernels for ARCv1 and ARCv2. Kernels for ARCv3 are always loaded from a physical address 0x0 instead of 0x80000000. A custom loader image is generated to distinguish it from a default vmlinux (vmlinux uses 0x80000000 as a starting address in physical and virtual spaces). Thus, it's necessary to run QEMU with -M virt,ram_start=0 parameter.

$ qemu-system-arc -M virt,ram_start=0 -cpu hs5x -m 2G -display none -nographic -monitor none -kernel loader

Running a Kernel for ARC HS 6x

$ qemu-system-arc64 -M virt,ram_start=0 -cpu hs6x -m 2G -display none -nographic -monitor none -kernel loader

Networking in QEMU Using a User Level Network Interface

Configuring Network

Firstly, it's necessary to enable BR2_SYSTEM_DHCP option in Buildroot configuration. Just use BR2_SYSTEM_DHCP=eth0 in your configuration file or use this option in a configuration menu:

System configuration -> Network interface to configure through DHCP -> eth0

You can connect a QEMU machine to a network using a user level network interface by passing these additional options:

-netdev user,id=net0 -device virtio-net-device,netdev=net0 -device virtio-rng-pci

You can enable port forwarding for a set of ports. Here is an example for forwarding ports for FTP and SSH:

-netdev user,id=net0,hostfwd=tcp::2021-:21,hostfwd=tcp::2022-:22 -device virtio-net-device,netdev=net0 -device virtio-rng-pci

Here is a full command line for running QEMU with support of networking and forwarding of ports for FTP and SSH:

qemu-system-arc -M virt -cpu archs -m 2G -display none -nographic -monitor none -kernel vmlinux -device virtio-rng-pci \
                -netdev user,id=net0,hostfwd=tcp::2021-:21,hostfwd=tcp::2022-:22 -device virtio-net-device,netdev=net0

In boot log output you will see these lines:

Starting network: udhcpc: started, v1.35.0
udhcpc: broadcasting discover
udhcpc: broadcasting select for 10.0.2.15, server 10.0.2.2
udhcpc: lease of 10.0.2.15 obtained from 10.0.2.2, lease time 86400
deleting routers
adding dns 10.0.2.3

Use ifconfig or ping utility to ensure that an Ethernet interface is initialized successfully:

# ifconfig
eth0      Link encap:Ethernet  HWaddr 52:54:00:12:34:56
          inet addr:10.0.2.15  Bcast:10.0.2.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:2 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:1180 (1.1 KiB)  TX bytes:689 (689.0 B)

lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

# ping 8.8.8.8
PING 8.8.8.8 (8.8.8.8): 56 data bytes
64 bytes from 8.8.8.8: seq=0 ttl=255 time=15.418 ms
64 bytes from 8.8.8.8: seq=1 ttl=255 time=0.701 ms
64 bytes from 8.8.8.8: seq=2 ttl=255 time=0.558 ms
64 bytes from 8.8.8.8: seq=3 ttl=255 time=1.122 ms

If you don't configure Buildroot with BR2_SYSTEM_DHCP=eth0 option, then you will to initialize an network device manually after Linux is booted:

# ifconfig eth0 up
# udhcpc
udhcpc: started, v1.35.0
udhcpc: broadcasting discover
udhcpc: broadcasting select for 10.0.2.15, server 10.0.2.2
udhcpc: lease of 10.0.2.15 obtained from 10.0.2.2, lease time 86400
deleting routers
adding dns 10.0.2.3

Using SSH

You can connect to the Linux kernel through SSH using a custom port 2022 (it is forwarded to 22 port inside of QEMU machine). To add SSH server to the Linux image use BR2_PACKAGE_OPENSSH=y and BR2_PACKAGE_HAVEGED=y options for Buildroot:

Target packages -> Networking applications -> [*] openssh
                -> Miscellaneous -> [*] haveged

Run Linux and create a user user and crate a home directory for the user:

# adduser user
# mkdir -p /home/user

Here is an example for connecting to the Linux using SSH:

$ ssh -p 2022 user@127.0.0.1
user@127.0.0.1's password:
$ pwd
/home/user

Networking in QEMU Using TUN/TAP Network Interface

TUN/TAP network interface allows interacting of the target with your host in both directions. For example, you can mount host’s NFS directories inside of the target.

Configure TUN/TAP interface on host’s side (choose your own IP address instead of 10.42.0.1):

$ sudo ip tuntap add tap1 mode tap
$ sudo ip addr add 10.42.0.1/24 dev tap1
$ sudo ip link set tap1 up

Run QEMU

$ qemu-system-arc -M virt -cpu archs -m 2G -display none -nographic -monitor none -kernel vmlinux \
                  -netdev tap,id=net0,ifname=tap1,script=no,downscript=no -device virtio-net-device,netdev=net0

Configure a network interface on target’s side:

# ifconfig eth0 10.42.0.100

After that you can connect to any Linux service without port forwarding. Here is an example of connecting to the Linux using SSH with TUN/TAP interface configured:

$ ssh user@10.42.0.100

Mounting an External Filesystem Image

It's possible to run Linux with an external filesystem image. Firstly, turn off linking an initial RAM filesystem into Linux kernel (BR2_TARGET_ROOTFS_INITRAMFS=n) and add an option for building Ext2 filesystem image (BR2_TARGET_ROOTFS_EXT2=y) in Buildroot:

Filesystem images -> [*] ext2/3/4 root filesystem
                  -> [ ] initial RAM filesystem linked into linux kernel

Run the Linux kernel and initialize and external storage:

$ qemu-system-arc -M virt -cpu archs -m 2G -display none -nographic -monitor none -kernel vmlinux \
                  -append "root=/dev/vda ro" -drive file=images/rootfs.ext2,format=raw,id=hd0 \
                  -device virtio-blk-device,drive=hd0

Mounting a Shared Folder

If you need shared folder access, add the following arguments to QEMU's command line:

-drive format=vvfat,id=hd0,file=fat:rw:/path/to/directory -device virtio-blk-device,drive=hd0

And the on guest's side:

# mount /dev/vda1
# mount /dev/vda1 /mnt
# touch /mnt/test.txt

Target will see contents on your directory as if it's a FAT partition on some drive without symbolic links, special attributes, etc. Target may do changes to the contents of this folder (with read/write permissions above). Host must not do any changes to the folder while target is using this shared folder, otherwise things will go seriously wrong (QEMU creates FAT table on start and then it's target which manipulates it but not host).

Clone this wiki locally