[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20140522.121215.16105419184813584.davem@davemloft.net>
Date: Thu, 22 May 2014 12:12:15 -0400 (EDT)
From: David Miller <davem@...emloft.net>
To: huizhang@...vell.com
Cc: netdev@...r.kernel.org, alan@...rguk.ukuu.org.uk,
nickcave.zhang@...il.com
Subject: Re: [PATCH] net: ipv6: Fixed up ipsec packet be re-routing issue
From: huizhang <huizhang@...vell.com>
Date: Thu, 22 May 2014 15:05:46 +0800
> diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
> index 5f8e128..869b68b 100644
> --- a/net/ipv6/xfrm6_policy.c
> +++ b/net/ipv6/xfrm6_policy.c
> @@ -134,6 +134,8 @@ _decode_session6(struct sk_buff *skb, struct flowi *fl, int reverse)
> const struct ipv6hdr *hdr = ipv6_hdr(skb);
> struct ipv6_opt_hdr *exthdr;
> const unsigned char *nh = skb_network_header(skb);
> + if(IP6CB(skb)->nhoff==0)
> + IP6CB(skb)->nhoff = offsetof(struct ipv6hdr, nexthdr);
> u8 nexthdr = nh[IP6CB(skb)->nhoff];
> int oif = 0;
Never put actual statements in the middle of a series of variable
declarations.
Also, it would probably be better to do this assignment in
__ip_local_out().
That's the bug, we only set nhoff in the input paths, we need
to set it in the output paths too if reaching _decode_session6
is possible for output packets.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists