From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: [PATCH 4/6] pkt_sched: Add qdisc->ops->peek() implementation. Date: Thu, 16 Oct 2008 15:00:29 +0200 Message-ID: <48F73AED.7080109@trash.net> References: <20081016094804.GE19019@ff.dom.local> <48F732F9.8010806@trash.net> <20081016125413.GB20302@ff.dom.local> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit Cc: David Miller , netdev@vger.kernel.org, Herbert Xu To: Jarek Poplawski Return-path: Received: from stinky.trash.net ([213.144.137.162]:33380 "EHLO stinky.trash.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751534AbYJPNAg (ORCPT ); Thu, 16 Oct 2008 09:00:36 -0400 In-Reply-To: <20081016125413.GB20302@ff.dom.local> Sender: netdev-owner@vger.kernel.org List-ID: Jarek Poplawski wrote: > On Thu, Oct 16, 2008 at 02:26:33PM +0200, Patrick McHardy wrote: >> Jarek Poplawski wrote: >>> Add qdisc->ops->peek() implementation for work-conserving qdiscs. >>> >>> diff --git a/net/sched/sch_multiq.c b/net/sched/sch_multiq.c >>> index 915f314..155648d 100644 >>> --- a/net/sched/sch_multiq.c >>> +++ b/net/sched/sch_multiq.c >> I haven't looked at sch_multiq in detail yet, but isn't it supposed to >> be used as root qdisc? In that case we wouldn't need a ->peek operation. >> > > It's recommended, but IMHO, could be useful as a child too, especially > for testing. I've thought if it shouldn't be rather treated as non-work- > conserving, but I don't think there is a reason for preventing such test > or other non-optimal usage. OK, if it works thats fine of course. >>> diff --git a/net/sched/sch_teql.c b/net/sched/sch_teql.c >>> index d35ef05..8d7acd8 100644 >>> --- a/net/sched/sch_teql.c >>> +++ b/net/sched/sch_teql.c >> teql is also meant to be used as root qdisc. It doesn't seem to enfore >> it though. > > Probably like above: there is not much of this code, but it could be > cut, no problem. I'm mainly wondering whether it works at all when not used as a root qdisc. I'm unable to answer this by a quick look at the code, so for now I'd just keep that part too.