From mboxrd@z Thu Jan 1 00:00:00 1970 From: "J. K. Cliburn" Subject: Re: sendfile() broken with 2.6.26 + Apache 2 ? Date: Wed, 16 Jul 2008 08:38:52 -0500 Message-ID: <487DF9EC.9060202@bellsouth.net> References: <487CD7A7.2080800@jeffray.co.uk> <487D893A.5080207@cosmosbay.com> <487DA388.3030104@jeffray.co.uk> <487DBA74.3090800@cosmosbay.com> Reply-To: jacliburn@bellsouth.net Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Ian Jeffray , linux-kernel@vger.kernel.org, Linux Netdev List , csnook@redhat.com, jeff@garzik.org To: Eric Dumazet Return-path: Received: from py-out-1112.google.com ([64.233.166.182]:55619 "EHLO py-out-1112.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756339AbYGPNi7 (ORCPT ); Wed, 16 Jul 2008 09:38:59 -0400 Received: by py-out-1112.google.com with SMTP id p76so3351349pyb.10 for ; Wed, 16 Jul 2008 06:38:58 -0700 (PDT) In-Reply-To: <487DBA74.3090800@cosmosbay.com> Sender: netdev-owner@vger.kernel.org List-ID: Eric Dumazet wrote: > Ian Jeffray a =E9crit : >> Hi Eric, >> >> Thanks for directing me to a better list. >> >> Further responses below: >> >> Eric Dumazet wrote: >>> CC to netdev where this report might find better answers >>> >>> Ian Jeffray a =E9crit : >>>> All, >>>> >>>> I moved from kernel 2.6.25.4 to 2.6.26 yesterday and observed that >>>> large files sent via Apache2 are partially corrupt. >>>> >>>> This appears to be linked to sendfile() -- disabling the use of >>>> sendfile in the apache config (EnableSendfile Off) allows it to >>>> function as normal. >>> >>> >>> What kind of network adapter are you using ? (lspci | grep -i ether= ) >> >> 02:00.0 Ethernet controller: Attansic Technology Corp. L1 Gigabit=20 >> Ethernet Adapter (rev b0) >> >> >>> If you disable TCP segmentation offload on this NIC (ethtool -K eth= 0=20 >>> tso off) , is this problem still present ? >> >> Wow. That 'solves' the problem! Great. >> >> Does this therefore point to an attansic driver issue? >> >=20 >=20 > Yes, maybe related to commit 9d90fb1ac9d97da86e24d9ea947bf2a2f333829a > In this patch, Jay Cliburn enabled TSO by default for atl1 driver. >=20 > This might be a driver problem, or a generic sendfile() problem, I do= nt=20 > know... I'm currently traveling and unable to delve into this issue and its=20 relation to the atl1 driver. I should be able to look at it this=20 weekend when I get back home. Jay