Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:57077 "EHLO sunset.davemloft.net" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1750803AbYGUP01 (ORCPT ); Mon, 21 Jul 2008 11:26:27 -0400 Date: Mon, 21 Jul 2008 08:26:27 -0700 (PDT) Message-Id: <20080721.082627.128593651.davem@davemloft.net> (sfid-20080721_172633_385319_63215AD3) To: herbert@gondor.apana.org.au Cc: hadi@cyberus.ca, kaber@trash.net, netdev@vger.kernel.org, johannes@sipsolutions.net, linux-wireless@vger.kernel.org Subject: Re: [PATCH 20/31]: pkt_sched: Perform bulk of qdisc destruction in RCU. From: David Miller In-Reply-To: <20080721152220.GA12341@gondor.apana.org.au> References: <20080721135846.GA11543@gondor.apana.org.au> <20080721.080901.214889979.davem@davemloft.net> <20080721152220.GA12341@gondor.apana.org.au> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Sender: linux-wireless-owner@vger.kernel.org List-ID: From: Herbert Xu Date: Mon, 21 Jul 2008 23:22:20 +0800 > On Mon, Jul 21, 2008 at 08:09:01AM -0700, David Miller wrote: > > > > > Actually you've hit it on the head, as an alternative to TX hashing > > > on the packet content, we need to allow TX queue selection based on > > > the current CPU ID. > > > > This we should avoid, it would allow reordering within a flow. > > Not if the RX hashing is flow-based... Can you control the process scheduler and where it decides to run the processing running sendmsg() too? That's the problem. It is totally unwise to do CPU based TX hashing.