ipv6: Fix default route failover when CONFIG_IPV6_ROUTER_PREF=n
authorPaul Marks <pmarks@google.com>
Mon, 3 Dec 2012 10:26:54 +0000 (10:26 +0000)
committerDavid S. Miller <davem@davemloft.net>
Mon, 3 Dec 2012 20:34:47 +0000 (15:34 -0500)
commita5a81f0b9025867efb999d14a8dfc1907c5a4c3b
tree4b2de223b5e08e0dcd4604d3619c738ff45eb2d3
parent5d097109257c03a71845729f8db6b5770c4bbedc
ipv6: Fix default route failover when CONFIG_IPV6_ROUTER_PREF=n

I believe this commit from 2008 was incorrect:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commitdiff;h=398bcbebb6f721ac308df1e3d658c0029bb74503

When CONFIG_IPV6_ROUTER_PREF is disabled, the kernel should follow
RFC4861 section 6.3.6: if no route is NUD_VALID, then traffic should be
sprayed across all routers (indirectly triggering NUD) until one of them
becomes NUD_VALID.

However, the following experiment demonstrates that this does not work:

1) Connect to an IPv6 network.
2) Change the router's MAC (and link-local) address.

The kernel will lock onto the first router and never try the new one, even
if the first becomes unreachable.  This patch fixes the problem by
allowing rt6_check_neigh() to return 0; if all routers return 0, then
rt6_select() will fall back to round-robin behavior.

This patch should have no effect when CONFIG_IPV6_ROUTER_PREF=y.

Note that rt6_check_neigh() is only used in a boolean context, so I've
changed its return type accordingly.

Signed-off-by: Paul Marks <pmarks@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv6/route.c