linux-kernel - Re: [PATCH 1/3] ifb: make device count build-time configurable

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <1454684532.2779.5.camel@v3.sk>
Date:	Fri, 05 Feb 2016 16:02:12 +0100
From:	Lubomir Rintel <lkundrak@...sk>
To:	David Miller <davem@...emloft.net>, stephen@...workplumber.org
Cc:	jhs@...atatu.com, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/3] ifb: make device count build-time configurable

On Tue, 2016-01-12 at 15:54 -0500, David Miller wrote:
> From: Stephen Hemminger <stephen@...workplumber.org>
> Date: Tue, 12 Jan 2016 10:44:37 -0800
> 
> > On Tue, 12 Jan 2016 07:55:22 -0500
> > Jamal Hadi Salim <jhs@...atatu.com> wrote:
> > 
> >> On 16-01-12 06:56 AM, Lubomir Rintel wrote:
> >> > The devices can be created at run-time for quite some time
> already and the
> >> > load-time device creation collides with attempts to create the
> device of
> >> > the same name:
> >> >
> >> >    # rmmod ifb
> >> >    # ip link add ifb0 type ifb
> >> >    RTNETLINK answers: File exists
> >> >
> >> > This is pretty much the same situation as was with the block
> loop devices
> >> > which was solved by adding a build-time configuration that the
> >> > distributions could use as they deem fit while keeping the
> default for
> >> > compatibility.
> >> >
> >> > Let's do that here as well.
> >> >
> >> > Signed-off-by: Lubomir Rintel <lkundrak@...sk>
> >> 
> >> I guess module options are frowned upon. so:
> > 
> > I would prefer that this were done with a module parameter, the
> same as dummy.
> > Only developers build their own configured kernels. Having the
> value set later
> > at module load time is preferable.
> 
> I like this even less, it means tools behave significantly
> differently
> based upon what module options were passed to the kernel.
> 
> Module options really should not change kernel behavior like this..

The module option is already there. It's defaults (creating the devices
noone asked for and that potentially collide with what the user tried
to create) are what we find bothersome.

Lubo