netdev
[Top] [All Lists]

Re: [PATCH]: was Re: LLTX and netif_stop_queue

To: "David S. Miller" <davem@xxxxxxxxxxxxx>
Subject: Re: [PATCH]: was Re: LLTX and netif_stop_queue
From: Andi Kleen <ak@xxxxxxx>
Date: Thu, 27 Jan 2005 08:16:45 +0100
Cc: Lennert Buytenhek <buytenh@xxxxxxxxxxxxxx>, shemminger@xxxxxxxx, roland@xxxxxxxxxxx, hadi@xxxxxxxxxx, iod00d@xxxxxx, eric.lemoine@xxxxxxxxx, netdev@xxxxxxxxxxx, ak@xxxxxxx, openib-general@xxxxxxxxxx, kaber@xxxxxxxxx
In-reply-to: <20050126223247.3e4643cc.davem@xxxxxxxxxxxxx>
References: <20050119144711.3fdd3d93.davem@xxxxxxxxxxxxx> <20050119151853.259de49a@xxxxxxxxxxxxxxxxx> <20050119164640.6c67bdfa.davem@xxxxxxxxxxxxx> <52r7kgu5n5.fsf@xxxxxxxxxxx> <20050119230526.393a5184.davem@xxxxxxxxxxxxx> <20050120085611.33f9485e@xxxxxxxxxxxxxxxxx> <20050121105452.GA12988@xxxxxxxxxxxxxxxxx> <20050125222705.1ee878fd.davem@xxxxxxxxxxxxx> <20050126132512.GA18220@xxxxxxxxxxxxxxxxx> <20050126223247.3e4643cc.davem@xxxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
On Wed, Jan 26, 2005 at 10:32:47PM -0800, David S. Miller wrote:
> On Wed, 26 Jan 2005 14:25:12 +0100
> Lennert Buytenhek <buytenh@xxxxxxxxxxxxxx> wrote:
> 
> > I've thought about this a bit, and the only sane way of doing recursion
> > detection that doesn't involve 'struct net_device' would be to keep track
> > of the recursion depth (perhaps per-CPU as you suggest) and tossing the
> > packet when it exceeds some random value, right?
> 
> Yes, that's the idea.

per CPU only works in preemptive kernel if you have preemption
disabled all the time. Do you? 

Seems not likely to me.

-Andi

<Prev in Thread] Current Thread [Next in Thread>