[PATCH net-next] IP_GRE: Fix IP-Identification.

[PATCH net-next] IP_GRE: Fix IP-Identification.

[Pravin B Shelar]

Use ip_select_ident_more() to generate and account ids for all
segments that can be used for segments of a gso packet.

Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
---
 net/ipv4/af_inet.c |    6 ++++--
 net/ipv4/ip_gre.c  |    4 +++-
 2 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index e225a4e..2992f37 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1327,8 +1327,10 @@ static struct sk_buff *inet_gso_segment(struct sk_buff *skb,
 			if (skb->next != NULL)
 				iph->frag_off |= htons(IP_MF);
 			offset += (skb->len - skb->mac_len - iph->ihl * 4);
-		} else
-			iph->id = htons(id++);
+		} else  {
+			if (!(iph->frag_off & htons(IP_DF)))
+				iph->id = htons(id++);
+		}
 		iph->tot_len = htons(skb->len - skb->mac_len);
 		iph->check = 0;
 		iph->check = ip_fast_csum(skb_network_header(skb), iph->ihl);
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index 5ef4da7..3c6f2e2 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -970,7 +970,9 @@ static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev
 	iph->daddr		=	fl4.daddr;
 	iph->saddr		=	fl4.saddr;
 	iph->ttl		=	ttl;
-	iph->id			=	0;
+
+	ip_select_ident_more(iph, &rt->dst, NULL,
+			     (skb_shinfo(skb)->gso_segs ?: 1) - 1);
 
 	if (ttl == 0) {
 		if (skb->protocol == htons(ETH_P_IP))
-- 
1.7.1

Re: [PATCH net-next] IP_GRE: Fix IP-Identification.

[Eric Dumazet]

On Thu, 2013-02-21 at 18:18 -0800, Pravin B Shelar wrote:
> Use ip_select_ident_more() to generate and account ids for all
> segments that can be used for segments of a gso packet.
> 
> Signed-off-by: Pravin B Shelar <pshelar@nicira.com>

I am not sure it makes any sense to do that.

Why is it needed ?

Re: [PATCH net-next] IP_GRE: Fix IP-Identification.

[Pravin Shelar]

On Thu, Feb 21, 2013 at 6:35 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Thu, 2013-02-21 at 18:18 -0800, Pravin B Shelar wrote:
>> Use ip_select_ident_more() to generate and account ids for all
>> segments that can be used for segments of a gso packet.
>>
>> Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
>
> I am not sure it makes any sense to do that.
>
> Why is it needed ?
>

Without this patch GRE-GSO generates ip fragments with id 0,2,3,4...
for every GSO packet, which is not correct. This patch fixes it.

Re: [PATCH net-next] IP_GRE: Fix IP-Identification.

[Eric Dumazet]

On Thu, 2013-02-21 at 19:42 -0800, Pravin Shelar wrote:

> Without this patch GRE-GSO generates ip fragments with id 0,2,3,4...
> for every GSO packet, which is not correct. This patch fixes it.
> --

I guess my question was more : Why dont we inherit ID from the inner
packet ?

ip_select_ident_more() is a performance killer, as all cpus are going to
access a shared counter in the dst.

Re: [PATCH net-next] IP_GRE: Fix IP-Identification.

[pravin]

On Thu, Feb 21, 2013 at 10:06 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Thu, 2013-02-21 at 19:42 -0800, Pravin Shelar wrote:
>
>> Without this patch GRE-GSO generates ip fragments with id 0,2,3,4...
>> for every GSO packet, which is not correct. This patch fixes it.
>> --
>
> I guess my question was more : Why dont we inherit ID from the inner
> packet ?
>
Its good idea. But we still need to get counter from dst if inner
packet is not ipv4.


> ip_select_ident_more() is a performance killer, as all cpus are going to
> access a shared counter in the dst.
>
>
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html