mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Keir Fraser <Keir.Fraser@cl.cam.ac.uk>
To: linux-kernel@vger.kernel.org
Cc: xen-devel@lists.xensource.com, Keir.Fraser@cl.cam.ac.uk,
	cedric@schieli.dyndns.org, coreteam@netfilter.org
Subject: [PATCH] Avoid unncessary checksum validation in TCP/UDP netfilter
Date: Fri, 27 May 2005 12:03:08 +0100	[thread overview]
Message-ID: <E1DbccR-00063Q-00@mta1.cl.cam.ac.uk> (raw)

[-- Attachment #1: Type: text/plain, Size: 750 bytes --]

The TCP/UDP connection-tracking code in netfilter validates the
checksum of incoming packets, to prevent nastier errors further down
the road. This check is unnecessary if the skb is marked as
CHECKSUM_UNNECESSARY. 

This patch will improve performance for networkinterfaces that perform
CHECKSUM_UNNECESSARY-style checksum offload (that's most modern
ones). More importantly (for me :-)) is that this makes netfilter play
nicely with checksum avoidance in the Xen virtual machine
monitor. Inter-domain traffic has no checksum calculated, so
validation checks will fail even though the packet data is good.
Of course, this avoidance relies on the network stack respecting
CHECKSUM_UNNECESSARY.

Signed-off-by: Keir Fraser <Keir.Fraser@xl.cam.ac.uk>


[-- Attachment #2: Checksum patch --]
[-- Type: text/plain, Size: 1351 bytes --]

diff -ur linux-2.6.11/net/ipv4/netfilter/ip_conntrack_proto_tcp.c linux-2.6.11-csum/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
--- linux-2.6.11/net/ipv4/netfilter/ip_conntrack_proto_tcp.c	2005-05-27 11:47:48 +01:00
+++ linux-2.6.11-csum/net/ipv4/netfilter/ip_conntrack_proto_tcp.c	2005-05-27 11:48:07 +01:00
@@ -803,6 +803,7 @@
 	 */
 	/* FIXME: Source route IP option packets --RR */
 	if (hooknum == NF_IP_PRE_ROUTING
+	    && skb->ip_summed != CHECKSUM_UNNECESSARY
 	    && csum_tcpudp_magic(iph->saddr, iph->daddr, tcplen, IPPROTO_TCP,
 			         skb->ip_summed == CHECKSUM_HW ? skb->csum
 			      	 : skb_checksum(skb, iph->ihl*4, tcplen, 0))) {
diff -ur linux-2.6.11/net/ipv4/netfilter/ip_conntrack_proto_udp.c linux-2.6.11-csum/net/ipv4/netfilter/ip_conntrack_proto_udp.c
--- linux-2.6.11/net/ipv4/netfilter/ip_conntrack_proto_udp.c	2005-05-27 11:47:48 +01:00
+++ linux-2.6.11-csum/net/ipv4/netfilter/ip_conntrack_proto_udp.c	2005-05-27 11:48:07 +01:00
@@ -120,6 +120,7 @@
 	 * and moreover root might send raw packets.
 	 * FIXME: Source route IP option packets --RR */
 	if (hooknum == NF_IP_PRE_ROUTING
+	    && skb->ip_summed != CHECKSUM_UNNECESSARY
 	    && csum_tcpudp_magic(iph->saddr, iph->daddr, udplen, IPPROTO_UDP,
 			         skb->ip_summed == CHECKSUM_HW ? skb->csum
 			      	 : skb_checksum(skb, iph->ihl*4, udplen, 0))) {

             reply	other threads:[~2005-05-27 11:03 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-05-27 11:03 Keir Fraser [this message]
2005-05-27 11:20 ` Phil Oester
2005-05-28  8:48   ` Keir Fraser
2005-06-12 15:06     ` [netfilter-core] " Patrick McHardy

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=E1DbccR-00063Q-00@mta1.cl.cam.ac.uk \
    --to=keir.fraser@cl.cam.ac.uk \
    --cc=cedric@schieli.dyndns.org \
    --cc=coreteam@netfilter.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=xen-devel@lists.xensource.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®