From: Phil Oester <kernel@linuxace.com>
To: Patrick McHardy <kaber@trash.net>
Cc: Harald Welte <laforge@netfilter.org>,
Rusty Russell <rusty@rustcorp.com.au>,
netfilter-devel@lists.netfilter.org
Subject: Re: [PATCH] Fix NAT TCP sequence adjustment
Date: Tue, 5 Apr 2005 21:48:06 -0700 [thread overview]
Message-ID: <20050406044806.GA9711@linuxace.com> (raw)
In-Reply-To: <20050404204716.GA4067@linuxace.com>
[-- Attachment #1: Type: text/plain, Size: 527 bytes --]
On Mon, Apr 04, 2005 at 01:47:16PM -0700, Phil Oester wrote:
> So as it stands, only retransmits on positive sizediff are handled
> properly. But my original check to compensate for this was flawed,
> since seq != correction_pos when entering adjust_tcp_sequence - but
> it does seem clear we cannot blindly -= offset_after.
I've become more convinced that this is the correct approach. Anticipating
that you will too, below is an updated patch with comments added.
Phil
Signed-off-by: Phil Oester <kernel@linuxace.com>
[-- Attachment #2: patch-ftpstable --]
[-- Type: text/plain, Size: 1599 bytes --]
diff -ru linux-orig/net/ipv4/netfilter/ip_nat_helper.c linux-stable/net/ipv4/netfilter/ip_nat_helper.c
--- linux-orig/net/ipv4/netfilter/ip_nat_helper.c 2005-03-02 02:37:49.000000000 -0500
+++ linux-stable/net/ipv4/netfilter/ip_nat_helper.c 2005-04-06 00:42:48.736295496 -0400
@@ -72,15 +72,23 @@
LOCK_BH(&ip_nat_seqofs_lock);
- /* SYN adjust. If it's uninitialized, or this is after last
- * correction, record it: we don't handle more than one
- * adjustment in the window, but do deal with common case of a
- * retransmit */
- if (this_way->offset_before == this_way->offset_after
- || before(this_way->correction_pos, seq)) {
- this_way->correction_pos = seq;
- this_way->offset_before = this_way->offset_after;
- this_way->offset_after += sizediff;
+ /* SYN adjust. In the case of a retransmit, ip_nat_seq_adjust has
+ * added offset_before to client's seq, so subtract it to test
+ * for retransmit here */
+ if (seq - this_way->offset_before != this_way->correction_pos) {
+ /* If it's uninitialized, or this is after last correction,
+ * record it. We don't handle more than one adjustment in
+ * the window */
+ seq -= this_way->offset_after;
+ if (this_way->offset_before == this_way->offset_after
+ || before(this_way->correction_pos, seq)) {
+ /* Since ip_nat_seq_adjust has added
+ * offset_after, subtract it to store
+ * the client's seq in correction_pos */
+ this_way->correction_pos = seq;
+ this_way->offset_before = this_way->offset_after;
+ this_way->offset_after += sizediff;
+ }
}
UNLOCK_BH(&ip_nat_seqofs_lock);
next prev parent reply other threads:[~2005-04-06 4:48 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-04-02 20:24 [PATCH] Fix NAT TCP sequence adjustment Phil Oester
2005-04-03 12:26 ` Milos Wimmer
2005-04-03 19:26 ` Patrick McHardy
2005-04-03 23:53 ` Phil Oester
2005-04-04 4:40 ` Phil Oester
2005-04-04 8:27 ` Patrick McHardy
2005-04-04 20:47 ` Phil Oester
2005-04-05 7:32 ` Patrick McHardy
2005-04-05 13:33 ` Patch lifetime " Amin Azez
2005-04-10 20:49 ` Harald Welte
2005-04-06 4:48 ` Phil Oester [this message]
2005-04-18 1:42 ` Patrick McHardy
2005-04-19 0:58 ` Phil Oester
2005-04-20 15:03 ` Patrick McHardy
2005-04-20 15:53 ` Phil Oester
2005-04-20 16:07 ` Patrick McHardy
2005-04-20 17:24 ` Phil Oester
2005-04-20 17:50 ` Patrick McHardy
2005-04-20 18:25 ` Phil Oester
2005-04-20 21:39 ` Martijn Lievaart
2005-04-21 1:41 ` Patrick McHardy
2005-04-21 1:38 ` Patrick McHardy
2005-04-21 12:31 ` Milos Wimmer
2005-04-21 12:32 ` Patrick McHardy
2005-04-21 13:31 ` Jonas Berlin
2005-04-21 23:01 ` Patrick McHardy
2005-04-27 0:44 ` Rusty Russell
2005-04-27 10:27 ` Patrick McHardy
2005-05-31 9:17 ` Rusty Russell
2005-05-31 13:02 ` Patrick McHardy
2005-05-31 13:48 ` Rusty Russell
2005-05-31 14:35 ` 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=20050406044806.GA9711@linuxace.com \
--to=kernel@linuxace.com \
--cc=kaber@trash.net \
--cc=laforge@netfilter.org \
--cc=netfilter-devel@lists.netfilter.org \
--cc=rusty@rustcorp.com.au \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.