Re: [PATCH RFC] ipv6: fix route selection if kernel is not compiled with CONFIG_IPV6_ROUTER_PREF

[Nicolas Dichtel <nicolas.dichtel@6wind.com>]

Le 10/07/2013 18:35, Hannes Frederic Sowa a écrit :
> On Wed, Jul 10, 2013 at 05:59:04PM +0200, Hannes Frederic Sowa wrote:
>> On Wed, Jul 10, 2013 at 05:20:01PM +0200, Hannes Frederic Sowa wrote:
>>> On Wed, Jul 10, 2013 at 04:10:58PM +0200, Nicolas Dichtel wrote:
>>>> Le 10/07/2013 15:21, Hannes Frederic Sowa a écrit :
>>>>> On Wed, Jul 10, 2013 at 02:22:55PM +0200, Nicolas Dichtel wrote:
>>>>>> Le 10/07/2013 12:53, Hannes Frederic Sowa a écrit :
>>>>> The route with rt6i_gateway does actually expire because I got it from
>>>>> autoconf and ip -6 r l confirms this, too. It seems this is only the cached
>>>>> route (I will confirm shortly). Is this still ok?
>>>> I wonder why expires is 0. Even if this route is cached, the flag
>>>> RTF_EXPIRES should be set. Am I wrong?
>>>
>>> It seems it is possible cached route gets its expiration updated. As
>>> such it is not counted in the iteration but it is found as a sibling and its
>>> nsiblings count is updated again.
>>
>> ip6_link_failure is the problem. We need to remove the route directly instead
>> of calling rt6_update_expires:
>>
>> static void ip6_link_failure(struct sk_buff *skb)
>> {
>>          struct rt6_info *rt;
>>
>>          icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_ADDR_UNREACH, 0);
>>
>>          rt = (struct rt6_info *) skb_dst(skb);
>>          if (rt) {
>>                  if (rt->rt6i_flags & RTF_CACHE)
>>                          rt6_update_expires(rt, 0);
>>                  else if (rt->rt6i_node && (rt->rt6i_flags & RTF_DEFAULT))
>>                          rt->rt6i_node->fn_sernum = -1;
>>          }
>> }
>
> With this patch I do not get the panic any more and the routing table
> stabilizes and discarded all NUD nonavailable routes. I'll need to review
> it myself, first. Thanks for your help!

Good catch!
Thank you.