net: bridge: fix sign bug

ipv6_skip_exthdr() can return error code that is below zero.
'offset' is unsigned, so it makes no sense.
ipv6_skip_exthdr() returns 'int' so we can painlessly change type of
offset to int.

Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index 27ae946363f16c6b709b5f4267e7ab5a574d5314..85afcdab4921c1b220da6391a581e7bf2199b5ca 100644 (file)
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -1435,7 +1435,7 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
        struct icmp6hdr *icmp6h;
        u8 nexthdr;
        unsigned len;
-       unsigned offset;
+       int offset;
        int err;
 
        if (!pskb_may_pull(skb, sizeof(*ip6h)))