ipv6: Avoid redoing fib6_lookup() for RTF_CACHE hit case
authorMartin KaFai Lau <kafai@fb.com>
Mon, 20 Oct 2014 20:42:44 +0000 (13:42 -0700)
committerDavid S. Miller <davem@davemloft.net>
Fri, 24 Oct 2014 04:14:39 +0000 (00:14 -0400)
When there is a RTF_CACHE hit, no need to redo fib6_lookup()
with reachable=0.

Cc: David Miller <davem@davemloft.net>
Cc: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv6/route.c

index f1ab2f4..98c523f 100644 (file)
@@ -944,12 +944,12 @@ redo_rt6_select:
                        goto out;
        }
 
-       if (rt->rt6i_flags & RTF_CACHE)
-               goto out;
-
        dst_hold(&rt->dst);
        read_unlock_bh(&table->tb6_lock);
 
+       if (rt->rt6i_flags & RTF_CACHE)
+               goto out2;
+
        if (!(rt->rt6i_flags & (RTF_NONEXTHOP | RTF_GATEWAY)))
                nrt = rt6_alloc_cow(rt, &fl6->daddr, &fl6->saddr);
        else if (!(rt->dst.flags & DST_HOST))