Re: [Resend PATCH] RDS: fix race condition when sending a message on unbound socket

[santosh shilimkar]

On 11/25/2015 4:52 AM, Quentin Casasnovas wrote:
> On Wed, Nov 25, 2015 at 12:21:45PM +0000, David Laight wrote:
> > From: Santosh Shilimkar
> > > Sent: 24 November 2015 22:13
> > ...
> > > Sasha's found a NULL pointer dereference in the RDS connection code when
> > > sending a message to an apparently unbound socket.  The problem is caused
> > > by the code checking if the socket is bound in rds_sendmsg(), which checks
> > > the rs_bound_addr field without taking a lock on the socket.  This opens a
> > > race where rs_bound_addr is temporarily set but where the transport is not
> > > in rds_bind(), leading to a NULL pointer dereference when trying to
> > > dereference 'trans' in __rds_conn_create().
> > >
> > > Vegard wrote a reproducer for this issue, so kindly ask him to share if
> > > you're interested.
> > ...
> > > diff --git a/net/rds/send.c b/net/rds/send.c
> > > index 827155c..c9cdb35 100644
> > > --- a/net/rds/send.c
> > > +++ b/net/rds/send.c
> > > @@ -1013,11 +1013,13 @@ int rds_sendmsg(struct socket *sock, struct msghdr *msg, size_t payload_len)
> > >   		release_sock(sk);
> >
> > This is falling though into an unconditional lock_sock().
> > No need to unlock and relock immediately.
> >
> > >   	}
> > >
> > > -	/* racing with another thread binding seems ok here */
> > > +	lock_sock(sk);
> > >   	if (daddr == 0 || rs->rs_bound_addr == 0) {
> > > +		release_sock(sk);
> > >   		ret = -ENOTCONN; /* XXX not a great errno */
> > >   		goto out;
> > >   	}
> > > +	release_sock(sk);
> >
> > On the face of it the above looks somewhat dubious.
> > Locks usually tie together two action (eg a test and use of a value),
> > In this case you only have a test inside the lock.
> > That either means that the state can change after you release the lock
> > (ie rs->rs_bound_addr = 0 is executed somewhere), or you don't
> > really need the lock.
If you see this patch in isolation, what you said looks very obvious
but as indicated by Quentin below, the update is protected by
lock and check wasn't which lead to the issue on bind() failures.

> If you look at rds_bind(), you'll see that it does something like the
> following:
>
>      lock_sock(sk);
>      ...
> 1:  rds_add_bound();  # This sets rs->rs_bound_addr
>      ...
>      if (!trans) {
>              ...
> 2:          rds_remove_bound(rs);  # This unsets rs->rs_bound_addr
>    ...
>    release_sock(sk);
>
> So any code checking rs_bound_addr without taking that lock could
> potentially think the socket is bound, when in fact rds_bind() has failed.
> This can happen if checking rs_bound_addr happens exactly between [1] and
> [2] above.  So the usage of the lock in this particular case is to get a
> consistent view of the sk.
>
> The only other case where rs_bound_addr is cleared is on socket release, so
> I didn't _think_ there was a problem here but maybe you can see another
> race?
I will be curious as well to know.

Regards,
Santosh
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/