I have an uml machine in which I do: ip -6 addr add 2001:a:b::1/64 dev eth0
then on the host machine I do: (tap0 is connected to uml eth0 with a of uml_switch-like software) ip -6 addr add 2001:a:b::2/64 dev tap0 then ping6 2001:a:b::2 from within the uml works ping6 2001:a:b::1 from the host works then if you want to go into the uml as a default route from the host: do on the host ip -6 route add default via 2001:a:b::1 This tells the host machine to send a packet to 2001:a:b::1 if no other route has been found, of course, the uml machine is suposed to know then what to do with the packet. You can also put in your uml: ip -6 route add default via 2001:a:b::2 Any packet for which no route has been found will go to the host. A machine has to have a destination address where it can do either an arp for ipv4 or an icmp neigbor solicitation for ipv6, otherwise it cannot send the packet because it has not resolved mac address. The "via" on the default route gives a target address and permits the mac resolution. a route from the host to the uml, Jeff Dike wrote: > On Mon, Feb 11, 2008 at 12:04:27AM -0500, Matt Dunford wrote: > >> I'd like to play with ipv6 on my guest UMLs. I have a native /64 from >> my ISP. Setting up IPv4 is pretty straight forward. On the host >> server: >> >> echo 1 > /proc/sys/net/ipv4/ip_forward >> route add -host 192.168.0.253 dev tap0 >> echo 1 > /proc/sys/net/ipv4/conf/tap0/proxy_arp >> arp -Ds 192.168.0.253 eth0 pub >> >> Then pass the tap device to the guest UML via eth0=tuntap,tap0 >> >> Is there an equivalent for this for IPv6? I'm trying various things, >> but I'm unable to route traffic to the guest UMLs. >> This is how I thought it should work. On the host server: >> >> echo 1 >/proc/sys/net/ipv6/conf/all/forwarding >> ip -6 addr add 2001:YYY:XXX:c::100/64 dev tap0 >> ip -6 route add 2001:YYY:XXX:c::100 dev tap0 >> >> But pings to 2001:YYY:XXX:c::100 go nowhere. Any ideas would be much >> appreciated. >> > > You're missing ARP, I think. I don't understand IPV6 at all, but I've > been told that it uses some basically different mechanism to figure > out its local network. > > Maybe you know this already, but maybe it's helpful if you didn't. > > Jeff > > ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ User-mode-linux-user mailing list User-mode-linux-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user