[Qemu-devel] [Bug 533610] Re: Routing net multicast socket over loopback make Ethernet broadcast bounce back to VM.

[Anthony Liguori]

The way you're setting up your route is going to result in an infinite
loop.

I'd suggest you use tap networking to do this.

** Changed in: qemu
       Status: New => Invalid

-- 
Routing net multicast socket over loopback make Ethernet broadcast bounce back 
to VM.
https://bugs.launchpad.net/bugs/533610
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.

Status in QEMU: Invalid

Bug description:
How to reproduce:

Host commands:
ip link set dev lo up
ip addr add 127.0.0.1/8 dev lo
ip route add 230.0.0.0/24 dev lo
qemu-system-x86_64 --kernel bzImage --initrd initrd.gz -net 
nic,model=e1000,vlan=0,macaddr=52:54:00:00:00:01 -net 
socket,vlan=0,mcast=230.0.0.1:1234 -net 
nic,model=e1000,vlan=1,macaddr=52:54:00:00:00:02 -net 
socket,vlan=1,mcast=230.0.0.2:1234

In any of the VM:
ip link set dev eth0 up
ip link set dev eth2 up
brctl addbr br0
brctl addif br0 eth0
brctl addif br0 eth2
brctl stp br0 on
ip link set dev br0 up

As soon as the bridge is up, you will get messages about eth0 and eth2 
receiving their own packets from the bridge, then, be prepared for a 
packet/message storm as it grow infinitly until you end Qemu or probably have 
the guest's kernel hang.