Received: with ECARTIS (v1.0.0; list netdev); Tue, 20 May 2003 01:58:44 -0700 (PDT)
Received: from brmea-mail-4.sun.com (brmea-mail-4.Sun.COM [192.18.98.36])
	by oss.sgi.com (8.12.9/8.12.9) with SMTP id h4K8wb2x026000
	for <netdev@oss.sgi.com>; Tue, 20 May 2003 01:58:37 -0700
Received: from esunmail ([129.147.58.120])
	by brmea-mail-4.sun.com (8.12.9/8.12.9) with ESMTP id h4K8wapd011836
	for <netdev@oss.sgi.com>; Tue, 20 May 2003 02:58:36 -0600 (MDT)
Received: from xpa-fe2 (esunmail [129.147.58.120]) by edgemail1.Central.Sun.COM
 (iPlanet Messaging Server 5.2 HotFix 1.12 (built Feb 13 2003))
 with ESMTP id <0HF600E49GVSZG@edgemail1.Central.Sun.COM> for
 netdev@oss.sgi.com; Tue, 20 May 2003 02:57:28 -0600 (MDT)
Received: from udine ([140.77.13.119])
 by mail.sun.net (iPlanet Messaging Server 5.2 HotFix 1.12 (built Feb 13 2003))
 with ESMTPSA id <0HF600CWOGVPEF@mail.sun.net> for netdev@oss.sgi.com; Tue,
 20 May 2003 02:57:28 -0600 (MDT)
Received: by udine (sSMTP sendmail emulation); Tue, 20 May 2003 10:57:25 +0200
Date: Tue, 20 May 2003 10:57:25 +0200
From: Eric Lemoine <Eric.Lemoine@Sun.COM>
Subject: Re: simple change to qdisc_restart()
In-reply-to: <20030520.012824.85398613.davem@redhat.com>
To: "David S. Miller" <davem@redhat.com>
Cc: Eric.Lemoine@Sun.COM, netdev@oss.sgi.com
Message-id: <20030520085724.GD978@udine>
MIME-version: 1.0
Content-type: text/plain; charset=us-ascii
Content-transfer-encoding: 7BIT
Content-disposition: inline
User-Agent: Mutt/1.3.28i
References: <20030520082217.GC978@udine>
 <20030520.012824.85398613.davem@redhat.com>
X-archive-position: 2592
X-ecartis-version: Ecartis v1.0.0
Sender: netdev-bounce@oss.sgi.com
Errors-to: netdev-bounce@oss.sgi.com
X-original-sender: Eric.Lemoine@Sun.COM
Precedence: bulk
X-list: netdev

>    From: Eric Lemoine <Eric.Lemoine@Sun.COM>
>    Date: Tue, 20 May 2003 10:22:17 +0200
>    
>    Any comments regarding the following patch?
> 
> I understand why it is valid, etc., but why do we even want to do
> this?  It is not like this dead-loop detection stuff is a hot-path or
> anything like that.

I've implemented a prototype that uses per-CPU kernel threads for
processing packets coming in from a single interface. The idea is to
apply multiple CPUs to a single network interface to be able to have
multiple CPUs simultaneously pumping data into the network. So in my
case, I have lots of cpu_collisions and running the tx softirq to do
nothing may lower the performances. Anyway, even though my patch may
help me, it may indeed be irrelevant to the stock kernel.

Thx.

-- 
Eric