From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH] bonding: reset queue mapping prior to transmission to physical device Date: Thu, 02 Jun 2011 13:46:37 -0700 (PDT) Message-ID: <20110602.134637.1101638217263464114.davem@davemloft.net> References: <20110602194621.GB2749@hmsreliant.think-freely.org> <20110602.130448.917724318321380583.davem@davemloft.net> <4DE7EF05.8030505@gmail.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: nhorman@tuxdriver.com, bhutchings@solarflare.com, netdev@vger.kernel.org, fubar@us.ibm.com, andy@greyhouse.net To: nicolas.2p.debian@gmail.com Return-path: Received: from shards.monkeyblade.net ([198.137.202.13]:45525 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753068Ab1FBUr6 convert rfc822-to-8bit (ORCPT ); Thu, 2 Jun 2011 16:47:58 -0400 In-Reply-To: <4DE7EF05.8030505@gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: =46rom: Nicolas de Peslo=FCan Date: Thu, 02 Jun 2011 22:13:57 +0200 > To be more precise, due to the way bonding use queue mapping for slav= e > selection, it is desirable to clear the mapping before sending to the > slave, because the meaning of the mapping for the slave interface > might be really different from the meaning for the bonding > interface. Arguably, this is the mapping usage in bonding which is > "different" from other usages, but... This just confirms my reasoning behind why I wanted to discourage drivers from providing explicit ->ndo_select_queue() methods unless absolutely necessary. Information now gets lost in cases like this bonding issue. Bonding should definitely, as I suggested, remember the original rxhash value and restore it when sending to the slave.