On 02/28/2013 05:33 AM, Thomas Graf wrote:

On 02/28/13 at 02:01pm, Cong Wang wrote:

quoted

WARNING: This is just an RFC patch, it only compiles!!

This patch adds IPv6 support to vxlan device, as the new version
RFC already mentioned it:

	http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03

I am not sure if the following trick will work:

	+       if (nla_len(nla) == sizeof(__be32)) {
	+               ip->ip4 = nla_get_be32(nla);
	+               ip->proto = htons(ETH_P_IP);
	+       }
	+#if IS_ENABLED(CONFIG_IPV6)
	+       else if (nla_len(nla) == sizeof(struct in6_addr)) {
	+               nla_memcpy(&ip->ip6, nla, sizeof(struct in6_addr));
	+               ip->proto = htons(ETH_P_IPV6);
	+       }
	+#endif

otherwise we have to introduce some new netlink attributes for
IPv6 addresses? Any comments?

How about setting ndm_family to AF_INET6 and parse NDA_DST based on
that?

Not sure if you can do that without registering your own fdb netlink 
handler.  rtnl_fdb_add is registered for PF_BRIDGE family.

You could do that, or use NDA_DST similar to how neighbor table uses it.
There the data inside of NDA_DST holds an address family and other data.

-vlad

Assuming that nla_len(NDA_DST) > 4 == ip6 is a bit weak as like with
any Netlink attribute we only ever enforce a minimum length and never
a maximum or exact length.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html