From: Ben Hutchings <bhutchings@solarflare.com>
To: Vasiliy Kulikov <segoon@openwall.com>
Cc: linux-kernel@vger.kernel.org, mjt@tls.msk.ru, arnd@arndb.de,
mirqus@gmail.com, netdev@vger.kernel.org,
David Miller <davem@davemloft.net>,
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
Subject: Re: [PATCH] net: don't allow CAP_NET_ADMIN to load non-netdev kernel modules
Date: Tue, 01 Mar 2011 20:13:10 +0000 [thread overview]
Message-ID: <1299010390.2529.30.camel@bwh-desktop> (raw)
In-Reply-To: <20110301194845.GA3533@albatros>
On Tue, 2011-03-01 at 22:48 +0300, Vasiliy Kulikov wrote:
> 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.
>
> This patch restricts an ability of autoloading modules to netdev modules
> with explicit aliases. This fixes CVE-2011-1019.
>
> Arnd Bergmann suggested to leave untouched the old pre-v2.6.32 behavior
> of loading netdev modules by name (without any prefix) for processes
> with CAP_SYS_MODULE to maintain the compatibility with network scripts
> that use autoloading netdev modules by aliases like "eth0", "wlan0".
[...]
> diff --git a/net/core/dev.c b/net/core/dev.c
> index 8ae6631..fc6f037 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -1119,8 +1119,16 @@ void dev_load(struct net *net, const char *name)
> dev = dev_get_by_name_rcu(net, name);
> rcu_read_unlock();
>
> - if (!dev && capable(CAP_NET_ADMIN))
> - request_module("%s", name);
> + if (!dev) {
> + if (capable(CAP_NET_ADMIN))
> + request_module("netdev-%s", name);
If this succeeds then the second request_module() should be skipped.
> + if (capable(CAP_SYS_MODULE)) {
> + if (!request_module("%s", name))
> + WARN_ONCE(1, "Loading kernel module for a "
> +"network device with CAP_SYS_MODULE (deprecated). Use CAP_NET_ADMIN and alias "
> +"netdev-%s instead\n", name);
[...]
If this feature is to be deprecated, there should be an error message
for each interface that depends on it. However, use of the feature is
not a bug so WARN is not appropriate. I think pr_err() would be fine.
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.
next prev parent reply other threads:[~2011-03-01 20:13 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-02-24 15:12 module loading with CAP_NET_ADMIN Vasiliy Kulikov
2011-02-24 16:34 ` Ben Hutchings
2011-02-25 12:30 ` Vasiliy Kulikov
2011-02-25 15:14 ` [PATCH] don't allow CAP_NET_ADMIN to load non-netdev kernel modules Vasiliy Kulikov
2011-02-25 17:25 ` Valdis.Kletnieks
2011-02-25 17:47 ` Vasiliy Kulikov
2011-02-25 17:48 ` Ben Hutchings
2011-02-25 18:47 ` David Miller
2011-02-25 19:02 ` Vasiliy Kulikov
2011-02-25 19:05 ` David Miller
2011-02-25 19:07 ` Ben Hutchings
2011-02-25 19:16 ` David Miller
2011-02-25 19:30 ` Ben Hutchings
2011-02-25 19:43 ` David Miller
2011-02-25 19:53 ` Ben Hutchings
2011-02-25 20:37 ` David Miller
2011-02-25 20:38 ` Ben Hutchings
2011-02-25 20:59 ` Michał Mirosław
2011-02-27 20:22 ` Arnd Bergmann
2011-02-28 9:29 ` Michael Tokarev
2011-02-28 9:51 ` Vasiliy Kulikov
2011-02-28 19:23 ` David Miller
2011-03-01 19:48 ` [PATCH] net: " Vasiliy Kulikov
2011-03-01 20:13 ` Ben Hutchings [this message]
2011-03-01 21:33 ` [PATCH v2] " Vasiliy Kulikov
2011-03-02 7:15 ` Michael Tokarev
2011-03-09 22:06 ` Vasiliy Kulikov
2011-03-09 22:09 ` David Miller
2011-03-09 22:53 ` James Morris
2011-03-10 9:49 ` Vasiliy Kulikov
2011-03-02 16:01 ` Kees Cook
2011-03-02 19:39 ` Jake Edge
2011-03-02 19:43 ` Vasiliy Kulikov
2011-03-02 19:49 ` Jake Edge
2011-03-02 20:18 ` Vasiliy Kulikov
2011-03-02 20:38 ` Jake Edge
2011-03-02 20:40 ` Jake Edge
2011-03-22 20:47 ` Eric Paris
2011-03-24 15:37 ` Serge E. Hallyn
2011-03-24 18:03 ` Eric Paris
2011-03-24 18:33 ` Ben Hutchings
2011-03-24 20:26 ` Serge E. Hallyn
2011-03-24 21:39 ` Stephen Hemminger
2011-03-24 21:46 ` David Miller
2011-03-24 21:57 ` Serge E. Hallyn
2011-03-24 22:15 ` Eric Paris
2011-03-24 21:57 ` Greg KH
2011-03-26 10:35 ` Vasiliy Kulikov
2011-02-27 11:44 ` [PATCH] " Vasiliy Kulikov
2011-02-27 23:18 ` David Miller
2011-02-27 23:19 ` David Miller
2011-02-25 15:29 ` module loading with CAP_NET_ADMIN Michael Tokarev
2011-02-25 15:57 ` Vasiliy Kulikov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1299010390.2529.30.camel@bwh-desktop \
--to=bhutchings@solarflare.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=dan.j.rosenberg@gmail.com \
--cc=davem@davemloft.net \
--cc=eric.dumazet@gmail.com \
--cc=eugene@redhat.com \
--cc=jesse@nicira.com \
--cc=jmorris@namei.org \
--cc=kaber@trash.net \
--cc=kees.cook@canonical.com \
--cc=kuznet@ms2.inr.ac.ru \
--cc=linux-kernel@vger.kernel.org \
--cc=mirqus@gmail.com \
--cc=mjt@tls.msk.ru \
--cc=netdev@vger.kernel.org \
--cc=pekkas@netcore.fi \
--cc=segoon@openwall.com \
--cc=therbert@google.com \
--cc=xiaosuo@gmail.com \
--cc=yoshfuji@linux-ipv6.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox