From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nishit Shah Subject: TSO issue - "DNAT/mark based routing/MASQUEARDE on locally genrated traffic" Date: Wed, 3 Jun 2009 14:25:49 +0530 Message-ID: Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: QUOTED-PRINTABLE To: netfilter-devel@vger.kernel.org Return-path: Received: from mail-qy0-f104.google.com ([209.85.221.104]:39335 "EHLO mail-qy0-f104.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753082AbZFCIzs convert rfc822-to-8bit (ORCPT ); Wed, 3 Jun 2009 04:55:48 -0400 Received: by qyk2 with SMTP id 2so1675674qyk.33 for ; Wed, 03 Jun 2009 01:55:50 -0700 (PDT) Sender: netfilter-devel-owner@vger.kernel.org List-ID: Hi, Whenever there is a output device change for locally generated traffic due to, DNAT/mark based routing in OUTPUT nat/mangle MASQURADE in POSTROUTING nat NETIF_F_TSO flag is not unset from sk->sk_route_caps. So, there are chances that, output device returned at TCP/IP layer routing supports NETIF_F_TSO and changed output device due to DNAT/mark based routing/MASQUERADE doesn=92t support NETIF_F_TSO. In some of my experiments for some interfaces, I got =93NETDEV_WATCHDOG= : eth1 transmit timeout=94 messages and for some others packets doesn=92t= go out of the box. Currently I have disabled TSO on all interfaces through ethtool. Is there any other way to solve this issue (something like resetting the flag in route_me_harder() if device doesn=92t support TSO ) ? Rgds, Nishit Shah. -- To unsubscribe from this list: send the line "unsubscribe netfilter-dev= el" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html