From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Hutchings Subject: Re: [PATCH] don't allow CAP_NET_ADMIN to load non-netdev kernel modules Date: Fri, 25 Feb 2011 19:07:59 +0000 Message-ID: <1298660879.2554.23.camel@bwh-desktop> References: <20110225151414.GA5211@albatros> <20110225.104720.71110261.davem@davemloft.net> <20110225190205.GA4541@albatros> <20110225.110529.39178636.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: segoon@openwall.com, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, kuznet@ms2.inr.ac.ru, pekkas@netcore.fi, jmorris@namei.org, yoshfuji@linux-ipv6.org, kaber@trash.net, eric.dumazet@gmail.com, therbert@google.com, xiaosuo@gmail.com, jesse@nicira.com, kees.cook@canonical.com, eugene@redhat.com, dan.j.rosenberg@gmail.com, akpm@linux-foundation.org To: David Miller Return-path: In-Reply-To: <20110225.110529.39178636.davem@davemloft.net> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Fri, 2011-02-25 at 11:05 -0800, David Miller wrote: > From: Vasiliy Kulikov > Date: Fri, 25 Feb 2011 22:02:05 +0300 > > > On Fri, Feb 25, 2011 at 10:47 -0800, David Miller wrote: > >> From: Vasiliy Kulikov > >> Date: Fri, 25 Feb 2011 18:14:14 +0300 > >> > >> > Since a8f80e8ff94ecba629542d9b4b5f5a8ee3eb565c any process with > >> > CAP_NET_ADMIN may load any module from /lib/modules/. This doesn't mean > >> > that CAP_NET_ADMIN is a superset of CAP_SYS_MODULE as modules are limited > >> > to /lib/modules/**. However, CAP_NET_ADMIN capability shouldn't allow > >> > anybody load any module not related to networking. > >> > >> Why go through this naming change, which does break things, instead of > >> simply adding a capability mask tag or similar to modules somehow. You > >> could stick it into a special elf section or similar. > >> > >> Doesn't that make tons more sense than this? > > > > This is not "simply", adding special section for a single workaround > > seems like an overkill for me - this touches the core (modules' > > internals), which is not related to the initial CAP_* problem at all. > > > > I'd be happy with not breaking anything, but I don't see any acceptable > > solution. > > I think it's warranted given that it allows us to avoid breaking things. > > I don't understand there is resistence in response to the first idea > I've seen proprosed that actually allows to fix the problem and not > break anything at the same time. > > That seems silly. You realise that module loading doesn't actually run in the context of request_module(), right? Ben. -- Ben Hutchings, Senior Software Engineer, Solarflare Communications Not speaking for my employer; that's the marketing department's job. They asked us to note that Solarflare product names are trademarked.