From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: [PATCH 3/4] sfc: resolve tx multiqueue bug Date: Tue, 22 Jul 2008 17:29:59 -0400 Message-ID: <48865157.20709@pobox.com> References: <20080718180156.GH10471@solarflare.com> <20080718184925.GJ10471@solarflare.com> <20080722184118.GR10471@solarflare.com> <20080722.140023.127726788.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: bhutchings@solarflare.com, netdev@vger.kernel.org, linux-net-drivers@solarflare.com To: David Miller Return-path: Received: from srv5.dvmed.net ([207.36.208.214]:34082 "EHLO mail.dvmed.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752182AbYGVVam (ORCPT ); Tue, 22 Jul 2008 17:30:42 -0400 In-Reply-To: <20080722.140023.127726788.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: David Miller wrote: > From: Ben Hutchings > Date: Tue, 22 Jul 2008 19:41:19 +0100 > >> So long as a queue is only woken in response to TX completions, the current >> arrangement is fine. However, sfc needs to tear down and restart hardware TX >> queues as part of some reconfiguration, self-test and recovery code, and I >> doubt it's the only such driver. > > Mark the carrier off and packets will stop flowing to the driver. Yep. Though based on looking at a lot of existing driver code, I think this is sometimes unclear to driver writers -- when to manage carrier (netif_carrier_on/off) and when to manage queue flow (netif_queue_start/stop/wake). Jeff