From: Andy Lutomirski <luto@amacapital.net>
To: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Linux Containers <containers@lists.linux-foundation.org>,
Josh Triplett <josh@joshtriplett.org>,
Andrew Morton <akpm@linux-foundation.org>,
Kees Cook <keescook@chromium.org>,
Michael Kerrisk-manpages <mtk.manpages@gmail.com>,
Linux API <linux-api@vger.kernel.org>,
linux-man <linux-man@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
LSM <linux-security-module@vger.kernel.org>,
Casey Schaufler <casey@schaufler-ca.com>,
"Serge E. Hallyn" <serge@hallyn.com>,
Richard Weinberger <richard@nod.at>,
Kenton Varda <kenton@sandstorm.io>,
stable <stable@vger.kernel.org>
Subject: Re: [CFT][PATCH 2/7] userns: Don't allow setgroups until a gid mapping has been setablished
Date: Mon, 8 Dec 2014 14:11:27 -0800 [thread overview]
Message-ID: <CALCETrVczpN8k_yenjGnocv4jYyNy+EXEW7h5NOhqy2XMAB=_Q@mail.gmail.com> (raw)
In-Reply-To: <87bnndre2h.fsf_-_@x220.int.ebiederm.org>
On Mon, Dec 8, 2014 at 2:07 PM, Eric W. Biederman <ebiederm@xmission.com> wrote:
>
> setgroups is unique in not needing a valid mapping before it can be called,
> in the case of setgroups(0, NULL) which drops all supplemental groups.
>
> The design of the user namespace assumes that CAP_SETGID can not actually
> be used until a gid mapping is established. Therefore add a helper function
> to see if the user namespace gid mapping has been established and call
> that function in the setgroups permission check.
>
> This is part of the fix for CVE-2014-8989, being able to drop groups
> without privilege using user namespaces.
>
> Cc: stable@vger.kernel.org
> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
> ---
> include/linux/user_namespace.h | 9 +++++++++
> kernel/groups.c | 7 ++++++-
> 2 files changed, 15 insertions(+), 1 deletion(-)
>
> diff --git a/include/linux/user_namespace.h b/include/linux/user_namespace.h
> index e95372654f09..41cc26e5a350 100644
> --- a/include/linux/user_namespace.h
> +++ b/include/linux/user_namespace.h
> @@ -37,6 +37,15 @@ struct user_namespace {
>
> extern struct user_namespace init_user_ns;
>
> +static inline bool userns_gid_mappings_established(const struct user_namespace *ns)
> +{
> + bool established;
> + smp_mb__before_atomic();
> + established = ACCESS_ONCE(ns->gid_map.nr_extents) != 0;
> + smp_mb__after_atomic();
> + return established;
> +}
I don't think this works on all platforms. ACCESS_ONCE is not atomic
in the smp_mb__before_atomic sense.
> +
> #ifdef CONFIG_USER_NS
>
> static inline struct user_namespace *get_user_ns(struct user_namespace *ns)
> diff --git a/kernel/groups.c b/kernel/groups.c
> index 02d8a251c476..e0335e44f76a 100644
> --- a/kernel/groups.c
> +++ b/kernel/groups.c
> @@ -6,6 +6,7 @@
> #include <linux/slab.h>
> #include <linux/security.h>
> #include <linux/syscalls.h>
> +#include <linux/user_namespace.h>
> #include <asm/uaccess.h>
>
> /* init to 2 - one for init_task, one to ensure it is never freed */
> @@ -217,7 +218,11 @@ bool may_setgroups(void)
> {
> struct user_namespace *user_ns = current_user_ns();
>
> - return ns_capable(user_ns, CAP_SETGID);
> + /* It is not safe to use setgroups until a gid mapping in
> + * the user namespace has been established.
> + */
> + return userns_gid_mappings_established(user_ns) &&
> + ns_capable(user_ns, CAP_SETGID);
> }
>
> /*
> --
> 1.9.1
>
--Andy
--
Andy Lutomirski
AMA Capital Management, LLC
next prev parent reply other threads:[~2014-12-08 22:11 UTC|newest]
Thread overview: 79+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-11-29 17:26 [PATCH v2] userns: Disallow setgroups unless the gid_map writer is privileged Andy Lutomirski
2014-12-02 12:09 ` Eric W. Biederman
2014-12-02 18:53 ` Andy Lutomirski
2014-12-02 19:45 ` Eric W. Biederman
2014-12-02 20:13 ` Andy Lutomirski
2014-12-02 20:25 ` [CFT][PATCH 1/3] userns: Avoid problems with negative groups Eric W. Biederman
2014-12-02 20:28 ` [CFT][PATCH 2/3] userns: Add a knob to disable setgroups on a per user namespace basis Eric W. Biederman
2014-12-02 20:30 ` [CFT][PATCH 3/3] userns: Unbreak the unprivileged remount tests Eric W. Biederman
2014-12-02 21:05 ` [CFT][PATCH 2/3] userns: Add a knob to disable setgroups on a per user namespace basis Andy Lutomirski
2014-12-02 21:45 ` Eric W. Biederman
2014-12-02 22:17 ` Andy Lutomirski
2014-12-02 23:07 ` Eric W. Biederman
2014-12-02 23:17 ` Andy Lutomirski
2014-12-08 22:06 ` [CFT][PATCH 1/7] userns: Document what the invariant required for safe unprivileged mappings Eric W. Biederman
2014-12-08 22:07 ` [CFT][PATCH 2/7] userns: Don't allow setgroups until a gid mapping has been setablished Eric W. Biederman
2014-12-08 22:11 ` Andy Lutomirski [this message]
[not found] ` <87h9x5ok0h.fsf@x220.int.ebiederm.org>
2014-12-08 22:33 ` Andy Lutomirski
2014-12-08 22:17 ` Richard Weinberger
2014-12-08 22:25 ` Andy Lutomirski
2014-12-08 22:27 ` Richard Weinberger
[not found] ` <874mt5ojfh.fsf@x220.int.ebiederm.org>
2014-12-08 22:47 ` Andy Lutomirski
2014-12-08 22:07 ` [CFT][PATCH 3/7] userns: Don't allow unprivileged creation of gid mappings Eric W. Biederman
2014-12-08 22:08 ` [CFT][PATCH 4/7] userns: Check euid no fsuid when establishing an unprivileged uid mapping Eric W. Biederman
2014-12-08 22:12 ` Andy Lutomirski
2014-12-08 22:10 ` [CFT][PATCH 5/7] userns: Only allow the creator of the userns unprivileged mappings Eric W. Biederman
2014-12-08 22:15 ` Andy Lutomirski
2014-12-08 22:11 ` [CFT][PATCH 6/7] userns: Add a knob to disable setgroups on a per user namespace basis Eric W. Biederman
2014-12-08 22:21 ` Andy Lutomirski
2014-12-08 22:44 ` Eric W. Biederman
2014-12-08 22:48 ` Andy Lutomirski
2014-12-08 23:30 ` Eric W. Biederman
2014-12-09 19:31 ` Eric W. Biederman
2014-12-09 20:36 ` [CFT][PATCH 1/8] userns: Document what the invariant required for safe unprivileged mappings Eric W. Biederman
2014-12-09 20:38 ` [CFT][PATCH 2/8] userns: Don't allow setgroups until a gid mapping has been setablished Eric W. Biederman
2014-12-09 22:49 ` Andy Lutomirski
2014-12-09 20:39 ` [CFT][PATCH 3/8] userns: Don't allow unprivileged creation of gid mappings Eric W. Biederman
2014-12-09 23:00 ` Andy Lutomirski
2014-12-09 20:39 ` [CFT][PATCH 4/8] userns: Check euid no fsuid when establishing an unprivileged uid mapping Eric W. Biederman
2014-12-09 20:41 ` [CFT][PATCH 5/8] userns: Only allow the creator of the userns unprivileged mappings Eric W. Biederman
2014-12-09 20:41 ` [CFT][PATCH 6/8] userns: Rename id_map_mutex to userns_state_mutex Eric W. Biederman
2014-12-09 22:49 ` Andy Lutomirski
2014-12-09 20:42 ` [CFT][PATCH 7/8] userns: Add a knob to disable setgroups on a per user namespace basis Eric W. Biederman
2014-12-09 22:28 ` Andy Lutomirski
[not found] ` <971ad3f6-90fd-4e3f-916c-8988af3c826d@email.android.com>
2014-12-10 0:21 ` Andy Lutomirski
[not found] ` <87wq5zf83t.fsf@x220.int.ebiederm.org>
[not found] ` <87iohh3c9c.fsf@x220.int.ebiederm.org>
2014-12-12 1:30 ` Andy Lutomirski
[not found] ` <8761dh3b7k.fsf_-_@x220.int.ebiederm.org>
[not found] ` <878uicy1r9.fsf_-_@x220.int.ebiederm.org>
2014-12-12 21:54 ` [PATCH 1/2] proc.5: Document /proc/[pid]/setgroups Eric W. Biederman
2015-02-02 15:36 ` Michael Kerrisk (man-pages)
2015-02-11 8:01 ` Michael Kerrisk (man-pages)
2015-02-11 13:51 ` Eric W. Biederman
2015-02-12 13:53 ` Michael Kerrisk (man-pages)
2015-02-21 7:57 ` Michael Kerrisk (man-pages)
2015-03-03 11:39 ` Michael Kerrisk (man-pages)
2014-12-12 21:54 ` [PATCH 2/2] user_namespaces.7: Update the documention to reflect the fixes for negative groups Eric W. Biederman
2015-02-02 15:37 ` Michael Kerrisk (man-pages)
2015-02-11 8:02 ` Michael Kerrisk (man-pages)
2015-02-11 14:01 ` Eric W. Biederman
2015-02-12 10:11 ` Michael Kerrisk (man-pages)
2015-02-02 21:31 ` Alban Crequy
2015-03-04 14:00 ` Michael Kerrisk (man-pages)
2014-12-09 20:43 ` [CFT][PATCH 8/8] userns: Allow setting gid_maps without privilege when setgroups is disabled Eric W. Biederman
2014-12-10 16:39 ` [CFT] Can I get some Tested-By's on this series? Eric W. Biederman
2014-12-10 22:48 ` Serge Hallyn
2014-12-10 22:50 ` Richard Weinberger
2014-12-10 23:19 ` Eric W. Biederman
2014-12-11 19:27 ` Richard Weinberger
2014-12-12 6:56 ` Chen, Hanxiao
2014-12-13 22:31 ` serge
[not found] ` <87lhmcy2et.fsf@x220.int.ebiederm.org>
[not found] ` <20141212220840.GF22091@castiana.ipv6.teksavvy.com>
[not found] ` <8761dgze56.fsf@x220.int.ebiederm.org>
2014-12-15 19:38 ` Serge Hallyn
2014-12-15 20:11 ` Eric W. Biederman
2014-12-15 20:49 ` Serge Hallyn
2014-12-16 2:05 ` Andy Lutomirski
2014-12-16 9:23 ` Richard Weinberger
2014-12-08 22:14 ` [CFT][PATCH 7/7] userns: Allow setting gid_maps without privilege when setgroups is disabled Eric W. Biederman
2014-12-08 22:26 ` Andy Lutomirski
2014-12-02 20:58 ` [CFT][PATCH 1/3] userns: Avoid problems with negative groups Andy Lutomirski
2014-12-02 21:26 ` Eric W. Biederman
2014-12-02 22:09 ` Andy Lutomirski
2014-12-02 22:48 ` Eric W. Biederman
2014-12-02 22:56 ` Andy Lutomirski
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='CALCETrVczpN8k_yenjGnocv4jYyNy+EXEW7h5NOhqy2XMAB=_Q@mail.gmail.com' \
--to=luto@amacapital.net \
--cc=akpm@linux-foundation.org \
--cc=casey@schaufler-ca.com \
--cc=containers@lists.linux-foundation.org \
--cc=ebiederm@xmission.com \
--cc=josh@joshtriplett.org \
--cc=keescook@chromium.org \
--cc=kenton@sandstorm.io \
--cc=linux-api@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-man@vger.kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=mtk.manpages@gmail.com \
--cc=richard@nod.at \
--cc=serge@hallyn.com \
--cc=stable@vger.kernel.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;
as well as URLs for NNTP newsgroup(s).