netbsd10 + qemu in bridge mode

push

Hi,

I would like to run qemu on netbsd10 in bridge mode.
According to manual tap interface is replaced with vether on netbsd 10.
I created interface vether0, added to bridge, but I don't how to call
qemu to use vether0 interface. BTW, no device vether0 created automatically.
Tried to create something manually with mknod, no luck.
qemu otherwise works.

push

My ifconfig.vether0

create
inet 192.168.0.35 netmask 255.255.255.0
! ifconfig bridge0 create
! ifconfig bridge0 up
! brconfig bridge0 add vether0 add iwm0

and script to start qemu

qemu-system-x86_64 -accel nvmm \
-cpu max -smp cpus=2 -m 1G \
-drive file=netbsd.qcow2,if=none,id=hd0 \
-device virtio-blk-pci,drive=hd0 \
-audiodev oss,id=oss,out.dev=/dev/audio,in.dev=/dev/audio \
-netdev tap,id=vether0,ifname=vether0,script=no -device virtio-net-pci,netdev=vether0 \
-device ac97,audiodev=oss \
-display sdl,gl=on \

// -netdev tap,id=vether0,ifname=vether0,script=no -device virtio-net-pci,netdev=vether0 \
// -netdev user,id=vioif0 -device virtio-net-pci,netdev=vioif0 \
// -boot once=d -cdrom iso/NetBSD-10.0_RC3-amd64.iso \

ip forwarding is enabled
nvmm loaded

and when the script is run:

qemu-system-x86_64: -netdev tap,id=vether0,ifname=vether0,script=no: could not open /dev/vether0: No such file or directory

push

Are you able to run qemu in bridge mode on NetBSD 10 RC3 amd64?

Jay

Check if this helps : https://www.netbsd.org/docs/guide/en/chap-virt.html#chap-virt-net-bridge

push

Well, there is link to chapter 24 on this page,
where is noted:

Note

In NetBSD 10.0 and later, it will become necessary to use vether instead of tap as a bridge endpoint. vether is unavailable in previous releases. It looks like it means
real 10 release, not candidate.

Anyway, back to tap0. vioif0 interface was created on the guest,
It is possible to ssh from guest to host and back.
Suppose for more routes should be created, on host.
I don't need this now.

Thank you

Vitalii

push

Hello, I don't think vether will work even after 10.0 is released because qemu needs a physical network device in /dev for bridging
For tap0 it works because /dev/tap* exists