[src/trunk]: src/sys/netinet Apply Revision 220794 from FreeBSD to avoid dup ...

[christos <christos%NetBSD.org@localhost>]

details:   https://anonhg.NetBSD.org/src/rev/ebc90aee12b4
branches:  trunk
changeset: 807930:ebc90aee12b4
user:      christos <christos%NetBSD.org@localhost>
date:      Mon Apr 27 16:50:17 2015 +0000

description:
Apply Revision 220794 from FreeBSD to avoid dup ACKs:

When checking to see if a window update should be sent to the remote peer,
don't force a window update if the window would not actually grow due to
window scaling.  Specifically, if the window scaling factor is larger than
2 * MSS, then after the local reader has drained 2 * MSS bytes from the
socket, a window update can end up advertising the same window.  If this
happens, the supposed window update actually ends up being a duplicate ACK.
This can result in an excessive number of duplicate ACKs when using a
higher maximum socket buffer size.

Pointed out by Ricky Charlet, in tech-net.

diffstat:

 sys/netinet/tcp_output.c |  12 ++++++++++--
 1 files changed, 10 insertions(+), 2 deletions(-)

diffs (38 lines):

diff -r 24d2ad3f01d2 -r ebc90aee12b4 sys/netinet/tcp_output.c
--- a/sys/netinet/tcp_output.c  Mon Apr 27 12:26:35 2015 +0000
+++ b/sys/netinet/tcp_output.c  Mon Apr 27 16:50:17 2015 +0000
@@ -1,4 +1,4 @@
-/*     $NetBSD: tcp_output.c,v 1.181 2015/04/27 02:59:44 ozaki-r Exp $ */
+/*     $NetBSD: tcp_output.c,v 1.182 2015/04/27 16:50:17 christos Exp $        */
 
 /*
  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
@@ -135,7 +135,7 @@
  */
 
 #include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: tcp_output.c,v 1.181 2015/04/27 02:59:44 ozaki-r Exp $");
+__KERNEL_RCSID(0, "$NetBSD: tcp_output.c,v 1.182 2015/04/27 16:50:17 christos Exp $");
 
 #include "opt_inet.h"
 #include "opt_ipsec.h"
@@ -1020,11 +1020,19 @@
                long adv = min(win, (long)TCP_MAXWIN << tp->rcv_scale) -
                        (tp->rcv_adv - tp->rcv_nxt);
 
+               /*
+                * If the new window size ends up being the same as the old
+                * size when it is scaled, then don't force a window update.
+                */
+               if ((tp->rcv_adv - tp->rcv_nxt) >> tp->rcv_scale ==
+                   (adv + tp->rcv_adv - tp->rcv_nxt) >> tp->rcv_scale)
+                       goto dontupdate;
                if (adv >= (long) (2 * rxsegsize))
                        goto send;
                if (2 * adv >= (long) so->so_rcv.sb_hiwat)
                        goto send;
        }
+dontupdate:
 
        /*
         * Send if we owe peer an ACK.