[Top] [All Lists]

Re: [PATCH 4/7] netpoll: fix ->poll() locking

To: "David S. Miller" <davem@xxxxxxxxxxxxx>
Subject: Re: [PATCH 4/7] netpoll: fix ->poll() locking
From: Matt Mackall <mpm@xxxxxxxxxxx>
Date: Fri, 22 Apr 2005 19:14:40 -0700
Cc: jmoyer@xxxxxxxxxx, jgarzik@xxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to: <20050422155958.474a150c.davem@xxxxxxxxxxxxx>
References: <4.454130102@xxxxxxxxxxx> <5.454130102@xxxxxxxxxxx> <17001.31150.194263.732284@xxxxxxxxxxxxxxxxxxxxxxxxxx> <20050422155218.223fa84d.davem@xxxxxxxxxxxxx> <17001.33427.462497.856777@xxxxxxxxxxxxxxxxxxxxxxxxxx> <20050422155958.474a150c.davem@xxxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mutt/1.5.6+20040907i
On Fri, Apr 22, 2005 at 03:59:58PM -0700, David S. Miller wrote:
> On Fri, 22 Apr 2005 19:02:43 -0400
> Jeff Moyer <jmoyer@xxxxxxxxxx> wrote:
> > Oh yes, of course.  Somehow I managed to forget that we now squirrel away
> > the struct netpoll pointer in the net_device.  Previously you could
> > register multiple netpoll clients to one device, and this was useful for
> > say doing netconsole and netdump over the same interface.  If we've removed
> > this ability, this is a bad thing.  Oh dear...

Hrm, it appears I forgot about that when I did the recent rework.

> In such a configuration, how did the netpoll code "tell" who the
> receive packets were for or did it send them to all registered
> netpoll clients for that device?

It walked a global list of clients.

Mathematics is the supreme nostalgia of our time.

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