From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: [PATCH v2 2/3] net_sched: Add accessor function for packet length for qdiscs Date: Wed, 30 Jul 2008 13:21:22 +0200 Message-ID: <48904EB2.8060402@trash.net> References: <20080725105748.GB10399@ff.dom.local> <20080725.035712.33516738.davem@davemloft.net> <20080725113757.GC10399@ff.dom.local> <20080725.044928.261427957.davem@davemloft.net> <488B24E6.4020003@trash.net> <20080726141844.GB2873@ami.dom.local> <489046C3.5030208@trash.net> <20080730111915.GA6362@ff.dom.local> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: David Miller , jussi.kivilinna@mbnet.fi, netdev@vger.kernel.org To: Jarek Poplawski Return-path: Received: from stinky.trash.net ([213.144.137.162]:45529 "EHLO stinky.trash.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753927AbYG3LWM (ORCPT ); Wed, 30 Jul 2008 07:22:12 -0400 In-Reply-To: <20080730111915.GA6362@ff.dom.local> Sender: netdev-owner@vger.kernel.org List-ID: Jarek Poplawski wrote: > On Wed, Jul 30, 2008 at 12:47:31PM +0200, Patrick McHardy wrote: > >> Jarek Poplawski wrote: >> >>> On Sat, Jul 26, 2008 at 03:21:42PM +0200, Patrick McHardy wrote: >>> > ... > >>>> The other problem that affects all qdiscs supporting actions is >>>> TC_ACT_QUEUED/TC_ACT_STOLEN getting mapped to NET_XMIT_SUCCESS >>>> even though the packet is not queued, corrupting upper qdiscs' >>>> qlen counters. >>>> >>> Why can't we (temporarily) simply check such cl->un.leaf.q->q.qlen >>> before and after enqueing? >>> >> Thats really ugly, why not simply fix it correctly by >> not lying to upper qdiscs? >> > > I thought it needs to wait for your audit. (Considering current > state of NET_XMIT statuses it's simple and reliable.) Otherwise > I'd prefer nice methods too. > Well, the problems are already clear, someone just needs to fix them :) I won't be able to do this until next week.