From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265777AbUGDVIr (ORCPT ); Sun, 4 Jul 2004 17:08:47 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265781AbUGDVIr (ORCPT ); Sun, 4 Jul 2004 17:08:47 -0400 Received: from agminet02.oracle.com ([141.146.126.229]:59542 "EHLO agminet02.oracle.com") by vger.kernel.org with ESMTP id S265777AbUGDVIa (ORCPT ); Sun, 4 Jul 2004 17:08:30 -0400 Message-ID: <40E87277.4040206@oracle.com> Date: Sun, 04 Jul 2004 23:11:19 +0200 From: Alessandro Suardi User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4.2) Gecko/20040323 X-Accept-Language: en-us, en MIME-Version: 1.0 To: linux-kernel Subject: TCP window scaling still bad in 2.6.7-bk17 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I have a tcpdump and a description of what happens here http://xoomer.virgilio.it/incident/tcpdump.out with www.kernel.org, but I'm mostly shut off the entire 'net. The _only_ site I found I can browse without disabling TCP window scaling is http://www.google.it. My system: RedHat 9 base distro 2.6.7-bk17 plus ACX100 out-of-kernel wireless driver (version 0.2.0-pre8_plus_fixes_13) from acx100.sf.net built with binutils 2.15.91.0.1 and GCC 3.4.0 System behaved properly until 2.6.7-bk1, was bad since 2.6.7-bk7 (haven't tried in-between kernels as I was on holiday). I also tried walking up to my router and connecting via eth0 instead of wlan0 (to rule out the wireless driver) - still the connection to www.kernel.org hangs (always -bk17). I'm available for further digging, please CC: me on replies as I only read lkml from the USSG archives. Thanks, --alessandro "Practice is more important than theory. A _lot_ more important." (Linus Torvalds on lkml, 1 June 2004)