(usagi-users 03606) [PATCH] [NET]: NETFILTER: remove duplicated operation and fix order in skb_clone().

(usagi-users 03606) [PATCH] [NET]: NETFILTER: remove duplicated operation and fix order in skb_clone().

[YOSHIFUJI Hideaki / 吉藤英明]

Hello.

In article <43F5A7B2.90803@6wind.com> (at Fri, 17 Feb 2006 11:38:42 +0100), Olivier MATZ <olivier.matz@6wind.com> says:

> in skb_clone(), the following code is done twice :
> 
> #if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
> 	C(nfct_reasm);
> 	nf_conntrack_get_reasm(skb->nfct_reasm);
> #endif

David, please apply the fillowing patch.

Thank you.

---
[NET]: NETFILTER: remove duplicated lines and fix order in skb_clone().

Some of netfilter-related members are initalized / copied twice in
skb_clone(). Remove one. Pointed out by Olivier MATZ <olivier.matz@6wind.com>.

And this patch also fixes order of copying / clearing members.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 6766f11..2144952 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -411,6 +411,9 @@ struct sk_buff *skb_clone(struct sk_buff
 	C(pkt_type);
 	C(ip_summed);
 	C(priority);
+#if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE)
+	C(ipvs_property);
+#endif
 	C(protocol);
 	n->destructor = NULL;
 #ifdef CONFIG_NETFILTER
@@ -422,13 +425,6 @@ struct sk_buff *skb_clone(struct sk_buff
 	C(nfct_reasm);
 	nf_conntrack_get_reasm(skb->nfct_reasm);
 #endif
-#if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE)
-	C(ipvs_property);
-#endif
-#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
-	C(nfct_reasm);
-	nf_conntrack_get_reasm(skb->nfct_reasm);
-#endif
 #ifdef CONFIG_BRIDGE_NETFILTER
 	C(nf_bridge);
 	nf_bridge_get(skb->nf_bridge);

-- 
YOSHIFUJI Hideaki @ USAGI Project  <yoshfuji@linux-ipv6.org>
GPG-FP  : 9022 65EB 1ECF 3AD1 0BDF  80D8 4807 F894 E062 0EEA

(usagi-users 03609) IPv6 setsockopt software MTU patch

[hoerdt]

Hi,

In my precedent mails, i pointed out that IPV6_MTU setsockopt
was ineffective because the field frag_size was not used further
in the ipv6 fragment code of the kernel. You can check out
that this is the case by this simple proof of concept code :

http://clarinet.u-strasbg.fr/~hoerdt/mtu.c

The fix is very simple, it only grab the value of frag_size and
use it in ip6_fragment file, could you please review the patch which
fix that problem at :

http://clarinet.u-strasbg.fr/~hoerdt/ip6_output.c.softfrag.patch

Thanks for the attention,

Hoerdt Mickaël

(usagi-users 03610) Re: IPv6 setsockopt software MTU patch

[YOSHIFUJI Hideaki / 吉藤英明]

In article <20060219134709.GC16304@clarinet.u-strasbg.fr> (at Sun, 19 Feb 2006 14:47:09 +0100), hoerdt@clarinet.u-strasbg.fr says:

> The fix is very simple, it only grab the value of frag_size and
> use it in ip6_fragment file, could you please review the patch which
> fix that problem at :
> 
> http://clarinet.u-strasbg.fr/~hoerdt/ip6_output.c.softfrag.patch

Thanks for the patch.
I think even we have np->frag_size, we should consider dst_mtu(&rt->u.dst)
as well. No?

--yoshfuji

(usagi-users 03611) Re: IPv6 setsockopt software MTU patch

[hoerdt]

Yes, Hugo Santos suggested me to use the minimum of both, the new
patch is available there :

http://clarinet.u-strasbg.fr/~hoerdt/ip6_output.c.softfrag.patch2

Thanks,

Hoerdt Mickaël


On Sun, Feb 19, 2006 at 11:13:44PM +0900, YOSHIFUJI Hideaki / ?$B5HF#1QL@ wrote:
> In article <20060219134709.GC16304@clarinet.u-strasbg.fr> (at Sun, 19 Feb 2006 14:47:09 +0100), hoerdt@clarinet.u-strasbg.fr says:
> 
> > The fix is very simple, it only grab the value of frag_size and
> > use it in ip6_fragment file, could you please review the patch which
> > fix that problem at :
> > 
> > http://clarinet.u-strasbg.fr/~hoerdt/ip6_output.c.softfrag.patch
> 
> Thanks for the patch.
> I think even we have np->frag_size, we should consider dst_mtu(&rt->u.dst)
> as well. No?
> 
> --yoshfuji
> 
> 

(usagi-users 03612) Re: IPv6 setsockopt software MTU patch

[YOSHIFUJI Hideaki / 吉藤英明]

In article <20060219145413.GD16304@clarinet.u-strasbg.fr> (at Sun, 19 Feb 2006 15:54:13 +0100), hoerdt@clarinet.u-strasbg.fr says:

> Yes, Hugo Santos suggested me to use the minimum of both, the new
> patch is available there :
> 
> http://clarinet.u-strasbg.fr/~hoerdt/ip6_output.c.softfrag.patch2

Do you mean like this?

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index efa3e72..3264740 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -494,6 +494,7 @@ static int ip6_fragment(struct sk_buff *
 	struct net_device *dev;
 	struct sk_buff *frag;
 	struct rt6_info *rt = (struct rt6_info*)skb->dst;
+	struct ipv6_pinfo *np = skb->sk ? inet6_sk(skb->sk) : NULL;
 	struct ipv6hdr *tmp_hdr;
 	struct frag_hdr *fh;
 	unsigned int mtu, hlen, left, len;
@@ -506,6 +507,8 @@ static int ip6_fragment(struct sk_buff *
 	nexthdr = *prevhdr;
 
 	mtu = dst_mtu(&rt->u.dst) - hlen - sizeof(struct frag_hdr);
+	if (np && np->frag_size && np->frag_size < dst_mtu(&rt->u.dst))
+		mtu = np->frag_size - hlen - sizeof(struct frag_hdr);
 
 	if (skb_shinfo(skb)->frag_list) {
 		int first_len = skb_pagelen(skb);


-- 
YOSHIFUJI Hideaki @ USAGI Project  <yoshfuji@linux-ipv6.org>
GPG-FP  : 9022 65EB 1ECF 3AD1 0BDF  80D8 4807 F894 E062 0EEA

(usagi-users 03613) Re: [PATCH] [NET]: NETFILTER: remove duplicated operation and fix order in skb_clone().

[David S. Miller]

From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Fri, 17 Feb 2006 21:51:01 +0900 (JST)

> [NET]: NETFILTER: remove duplicated lines and fix order in skb_clone().
> 
> Some of netfilter-related members are initalized / copied twice in
> skb_clone(). Remove one. Pointed out by Olivier MATZ <olivier.matz@6wind.com>.
> 
> And this patch also fixes order of copying / clearing members.
> 
> Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

Patch applied to net-2.6, thank you very much.