Patchwork [Lucid,CVE-2013-4483] ipc/msg: fix race around refcount

login
register
mail settings
Submitter Luis Henriques
Date May 9, 2014, 2:35 p.m.
Message ID <1399646109-16362-2-git-send-email-luis.henriques@canonical.com>
Download mbox | patch
Permalink /patch/347431/
State New
Headers show

Comments

Luis Henriques - May 9, 2014, 2:35 p.m.
From: Konstantin Khlebnikov <k.khlebnikov@samsung.com>

In older kernels (before v3.10) ipc_rcu_hdr->refcount was non-atomic int.
There was possuble double-free bug: do_msgsnd() calls ipc_rcu_putref() under
msq->q_perm->lock and RCU, while freequeue() calls it while it holds only
'rw_mutex', so there is no sinchronization between them. Two function
decrements '2' non-atomically, they both can get '0' as result.

do_msgsnd()					freequeue()

msq = msg_lock_check(ns, msqid);
...
ipc_rcu_getref(msq);
msg_unlock(msq);
schedule();
						(caller locks spinlock)
						expunge_all(msq, -EIDRM);
						ss_wakeup(&msq->q_senders, 1);
						msg_rmid(ns, msq);
						msg_unlock(msq);
ipc_lock_by_ptr(&msq->q_perm);
ipc_rcu_putref(msq);				ipc_rcu_putref(msq);
< both may get get --(...)->refcount == 0 >

This patch locks ipc_lock and RCU around ipc_rcu_putref in freequeue.
( RCU protects memory for spin_unlock() )

Similar bugs might be in other users of ipc_rcu_putref().

In the mainline this has been fixed in v3.10 indirectly in commmit
6062a8dc0517bce23e3c2f7d2fea5e22411269a3
("ipc,sem: fine grained locking for semtimedop") by Rik van Riel.
That commit optimized locking and converted refcount into atomic.

I'm not sure that anybody should care about this bug: it's very-very unlikely
and no longer exists in actual mainline. I've found this just by looking into
the code, probably this never happens in real life.

Signed-off-by: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
(cherry picked from 91182754daa6ca26dd2e97ee0b0f6e9e37d33324 3.2.y)
CVE-2013-4483
BugLink: http://bugs.launchpad.net/bugs/1248713
Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
---
 ipc/msg.c | 2 ++
 1 file changed, 2 insertions(+)
Andy Whitcroft - May 9, 2014, 3:08 p.m.
On Fri, May 09, 2014 at 03:35:09PM +0100, Luis Henriques wrote:
> From: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
> 
> In older kernels (before v3.10) ipc_rcu_hdr->refcount was non-atomic int.
> There was possuble double-free bug: do_msgsnd() calls ipc_rcu_putref() under
> msq->q_perm->lock and RCU, while freequeue() calls it while it holds only
> 'rw_mutex', so there is no sinchronization between them. Two function
> decrements '2' non-atomically, they both can get '0' as result.
> 
> do_msgsnd()					freequeue()
> 
> msq = msg_lock_check(ns, msqid);
> ...
> ipc_rcu_getref(msq);
> msg_unlock(msq);
> schedule();
> 						(caller locks spinlock)
> 						expunge_all(msq, -EIDRM);
> 						ss_wakeup(&msq->q_senders, 1);
> 						msg_rmid(ns, msq);
> 						msg_unlock(msq);
> ipc_lock_by_ptr(&msq->q_perm);
> ipc_rcu_putref(msq);				ipc_rcu_putref(msq);
> < both may get get --(...)->refcount == 0 >
> 
> This patch locks ipc_lock and RCU around ipc_rcu_putref in freequeue.
> ( RCU protects memory for spin_unlock() )
> 
> Similar bugs might be in other users of ipc_rcu_putref().
> 
> In the mainline this has been fixed in v3.10 indirectly in commmit
> 6062a8dc0517bce23e3c2f7d2fea5e22411269a3
> ("ipc,sem: fine grained locking for semtimedop") by Rik van Riel.
> That commit optimized locking and converted refcount into atomic.
> 
> I'm not sure that anybody should care about this bug: it's very-very unlikely
> and no longer exists in actual mainline. I've found this just by looking into
> the code, probably this never happens in real life.
> 
> Signed-off-by: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
> Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
> (cherry picked from 91182754daa6ca26dd2e97ee0b0f6e9e37d33324 3.2.y)
> CVE-2013-4483
> BugLink: http://bugs.launchpad.net/bugs/1248713
> Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
> ---
>  ipc/msg.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/ipc/msg.c b/ipc/msg.c
> index 779f762..75075fc 100644
> --- a/ipc/msg.c
> +++ b/ipc/msg.c
> @@ -297,7 +297,9 @@ static void freeque(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
>  	}
>  	atomic_sub(msq->q_cbytes, &ns->msg_bytes);
>  	security_msg_queue_free(msq);
> +	ipc_lock_by_ptr(&msq->q_perm);
>  	ipc_rcu_putref(msq);
> +	ipc_unlock(&msq->q_perm);
>  }
>  
>  /*

Matches the fix as applied to P & Q.  

Acked-by: Andy Whitcroft <apw@canonical.com>

-apw
Brad Figg - May 9, 2014, 3:13 p.m.
On 05/09/2014 09:35 AM, Luis Henriques wrote:
> From: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
> 
> In older kernels (before v3.10) ipc_rcu_hdr->refcount was non-atomic int.
> There was possuble double-free bug: do_msgsnd() calls ipc_rcu_putref() under
> msq->q_perm->lock and RCU, while freequeue() calls it while it holds only
> 'rw_mutex', so there is no sinchronization between them. Two function
> decrements '2' non-atomically, they both can get '0' as result.
> 
> do_msgsnd()					freequeue()
> 
> msq = msg_lock_check(ns, msqid);
> ...
> ipc_rcu_getref(msq);
> msg_unlock(msq);
> schedule();
> 						(caller locks spinlock)
> 						expunge_all(msq, -EIDRM);
> 						ss_wakeup(&msq->q_senders, 1);
> 						msg_rmid(ns, msq);
> 						msg_unlock(msq);
> ipc_lock_by_ptr(&msq->q_perm);
> ipc_rcu_putref(msq);				ipc_rcu_putref(msq);
> < both may get get --(...)->refcount == 0 >
> 
> This patch locks ipc_lock and RCU around ipc_rcu_putref in freequeue.
> ( RCU protects memory for spin_unlock() )
> 
> Similar bugs might be in other users of ipc_rcu_putref().
> 
> In the mainline this has been fixed in v3.10 indirectly in commmit
> 6062a8dc0517bce23e3c2f7d2fea5e22411269a3
> ("ipc,sem: fine grained locking for semtimedop") by Rik van Riel.
> That commit optimized locking and converted refcount into atomic.
> 
> I'm not sure that anybody should care about this bug: it's very-very unlikely
> and no longer exists in actual mainline. I've found this just by looking into
> the code, probably this never happens in real life.
> 
> Signed-off-by: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
> Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
> (cherry picked from 91182754daa6ca26dd2e97ee0b0f6e9e37d33324 3.2.y)
> CVE-2013-4483
> BugLink: http://bugs.launchpad.net/bugs/1248713
> Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
> ---
>  ipc/msg.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/ipc/msg.c b/ipc/msg.c
> index 779f762..75075fc 100644
> --- a/ipc/msg.c
> +++ b/ipc/msg.c
> @@ -297,7 +297,9 @@ static void freeque(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
>  	}
>  	atomic_sub(msq->q_cbytes, &ns->msg_bytes);
>  	security_msg_queue_free(msq);
> +	ipc_lock_by_ptr(&msq->q_perm);
>  	ipc_rcu_putref(msq);
> +	ipc_unlock(&msq->q_perm);
>  }
>  
>  /*
>
Luis Henriques - May 9, 2014, 3:17 p.m.
On Fri, May 09, 2014 at 03:35:09PM +0100, Luis Henriques wrote:
> From: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
> 
> In older kernels (before v3.10) ipc_rcu_hdr->refcount was non-atomic int.
> There was possuble double-free bug: do_msgsnd() calls ipc_rcu_putref() under
> msq->q_perm->lock and RCU, while freequeue() calls it while it holds only
> 'rw_mutex', so there is no sinchronization between them. Two function
> decrements '2' non-atomically, they both can get '0' as result.
> 
> do_msgsnd()					freequeue()
> 
> msq = msg_lock_check(ns, msqid);
> ...
> ipc_rcu_getref(msq);
> msg_unlock(msq);
> schedule();
> 						(caller locks spinlock)
> 						expunge_all(msq, -EIDRM);
> 						ss_wakeup(&msq->q_senders, 1);
> 						msg_rmid(ns, msq);
> 						msg_unlock(msq);
> ipc_lock_by_ptr(&msq->q_perm);
> ipc_rcu_putref(msq);				ipc_rcu_putref(msq);
> < both may get get --(...)->refcount == 0 >
> 
> This patch locks ipc_lock and RCU around ipc_rcu_putref in freequeue.
> ( RCU protects memory for spin_unlock() )
> 
> Similar bugs might be in other users of ipc_rcu_putref().
> 
> In the mainline this has been fixed in v3.10 indirectly in commmit
> 6062a8dc0517bce23e3c2f7d2fea5e22411269a3
> ("ipc,sem: fine grained locking for semtimedop") by Rik van Riel.
> That commit optimized locking and converted refcount into atomic.
> 
> I'm not sure that anybody should care about this bug: it's very-very unlikely
> and no longer exists in actual mainline. I've found this just by looking into
> the code, probably this never happens in real life.
> 
> Signed-off-by: Konstantin Khlebnikov <k.khlebnikov@samsung.com>
> Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
> (cherry picked from 91182754daa6ca26dd2e97ee0b0f6e9e37d33324 3.2.y)
> CVE-2013-4483
> BugLink: http://bugs.launchpad.net/bugs/1248713
> Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
> ---
>  ipc/msg.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/ipc/msg.c b/ipc/msg.c
> index 779f762..75075fc 100644
> --- a/ipc/msg.c
> +++ b/ipc/msg.c
> @@ -297,7 +297,9 @@ static void freeque(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
>  	}
>  	atomic_sub(msq->q_cbytes, &ns->msg_bytes);
>  	security_msg_queue_free(msq);
> +	ipc_lock_by_ptr(&msq->q_perm);
>  	ipc_rcu_putref(msq);
> +	ipc_unlock(&msq->q_perm);
>  }
>  
>  /*
> -- 
> 1.9.1
> 
> -- 
> kernel-team mailing list
> kernel-team@lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/kernel-team

This fix should be also applied to Precise/lts-backport-raring (thanks
for noticing Andy!).  As the author of the patch stated, it applies to
kernels <= 3.10, which includes Raring 3.8 :-)

Cheers,
--
Luís

Patch

diff --git a/ipc/msg.c b/ipc/msg.c
index 779f762..75075fc 100644
--- a/ipc/msg.c
+++ b/ipc/msg.c
@@ -297,7 +297,9 @@  static void freeque(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
 	}
 	atomic_sub(msq->q_cbytes, &ns->msg_bytes);
 	security_msg_queue_free(msq);
+	ipc_lock_by_ptr(&msq->q_perm);
 	ipc_rcu_putref(msq);
+	ipc_unlock(&msq->q_perm);
 }
 
 /*