From mboxrd@z Thu Jan  1 00:00:00 1970
From: Greg KH <greg@kroah.com>
Subject: Re: [PATCH] sysfs: Don't allow the creation of symlinks we can't
 remove
Date: Thu, 8 Jul 2010 16:06:01 -0700
Message-ID: <20100708230601.GB17002@kroah.com>
References: <20100621150826.762ac9f2.akpm@linux-foundation.org>
 <m1mxuodo0n.fsf@fess.ebiederm.org>
 <20100621222938.GA20583@suse.de>
 <m1eig0dmhk.fsf@fess.ebiederm.org>
 <20100621231058.GA1066@suse.de>
 <m14ogwc4ou.fsf@fess.ebiederm.org>
 <20100622035631.GA3755@suse.de>
 <m1630q7x5v.fsf_-_@fess.ebiederm.org>
 <20100708211930.GA15385@kroah.com>
 <m1r5jdbobe.fsf@fess.ebiederm.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Greg KH <gregkh@suse.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	"Rafael J. Wysocki" <rjw@sisk.pl>,
	"Maciej W. Rozycki" <macro@linux-mips.org>,
	Kay Sievers <kay.sievers@vrfy.org>,
	Johannes Berg <johannes@sipsolutions.net>,
	netdev <netdev@vger.kernel.org>
To: "Eric W. Biederman" <ebiederm@xmission.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from kroah.org ([198.145.64.141]:46398 "EHLO coco.kroah.org"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1757178Ab0GHXGT (ORCPT <rfc822;netdev@vger.kernel.org>);
	Thu, 8 Jul 2010 19:06:19 -0400
Content-Disposition: inline
In-Reply-To: <m1r5jdbobe.fsf@fess.ebiederm.org>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Thu, Jul 08, 2010 at 03:28:53PM -0700, Eric W. Biederman wrote:
> Greg KH <greg@kroah.com> writes:
> 
> > With this patch, how does the existing code fail as the drivers aren't
> > fixed up?
> >
> > I like this change, just worried it will cause problems if it gets into
> > .35, without your RFC patch.  Will it?
> 
> It don't expect this patch to be worse than where we are current at.
> Network devices are renamed, and come and go enough that both the
> mac80211_hwsim and the bnep driver are currently unusable with a
> failure only the rename and remove.
> 
> This patch simply moves the failure into creation where we are a little more
> prepared to deal with problems, and this patch is limited to mac80211_hwsim,
> bnep, and any hypothetically undiscovered other network devices that
> have the same problem.
> 
> mac80211_hwsim with just this patch becomes somewhat usable as it's primary
> network device gets registered and the module can be loaded and unloaded.  It
> just doesn't create the wlan0 and wlan1 interfaces for the wifi interfaces.

Fair enough, I've commited it now, let's see what happens :)

> On a slightly unrelated note, what got us on trying to convert
> mac80211_hwsim was that everything happens in a single module which
> makes avoiding cleanup races hard.  I am hoping bnep is structured
> enough differently so we can convert it to a bus device using the
> existing infrastructure.  I haven't figured out bnep yet though ;(

Yeah, good luck with that...

greg k-h