[bpf] xsk: fix registration of Rx-only sockets
diff mbox series

Message ID 1571645818-16244-1-git-send-email-magnus.karlsson@intel.com
State Accepted
Delegated to: BPF Maintainers
Headers show
Series
  • [bpf] xsk: fix registration of Rx-only sockets
Related show

Commit Message

Magnus Karlsson Oct. 21, 2019, 8:16 a.m. UTC
Having Rx-only AF_XDP sockets can potentially lead to a crash in the
system by a NULL pointer dereference in xsk_umem_consume_tx(). This
function iterates through a list of all sockets tied to a umem and
checks if there are any packets to send on the Tx ring. Rx-only
sockets do not have a Tx ring, so this will cause a NULL pointer
dereference. This will happen if you have registered one or more
Rx-only sockets to a umem and the driver is checking the Tx ring even
on Rx, or if the XDP_SHARED_UMEM mode is used and there is a mix of
Rx-only and other sockets tied to the same umem.

Fixed by only putting sockets with a Tx component on the list that
xsk_umem_consume_tx() iterates over.

Fixes: ac98d8aab61b ("xsk: wire upp Tx zero-copy functions")
Reported-by: Kal Cutter Conley <kal.conley@dectris.com>
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
---
 net/xdp/xdp_umem.c | 6 ++++++
 1 file changed, 6 insertions(+)

Comments

Jonathan Lemon Oct. 23, 2019, 11:03 p.m. UTC | #1
On 21 Oct 2019, at 1:16, Magnus Karlsson wrote:

> Having Rx-only AF_XDP sockets can potentially lead to a crash in the
> system by a NULL pointer dereference in xsk_umem_consume_tx(). This
> function iterates through a list of all sockets tied to a umem and
> checks if there are any packets to send on the Tx ring. Rx-only
> sockets do not have a Tx ring, so this will cause a NULL pointer
> dereference. This will happen if you have registered one or more
> Rx-only sockets to a umem and the driver is checking the Tx ring even
> on Rx, or if the XDP_SHARED_UMEM mode is used and there is a mix of
> Rx-only and other sockets tied to the same umem.
>
> Fixed by only putting sockets with a Tx component on the list that
> xsk_umem_consume_tx() iterates over.

A future improvement might be renaming umem->xsk_list to umem->xsk_tx_list
or similar, in order to make it clear that the list is only used on the
TX path.

>
> Fixes: ac98d8aab61b ("xsk: wire upp Tx zero-copy functions")
> Reported-by: Kal Cutter Conley <kal.conley@dectris.com>
> Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>

Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Alexei Starovoitov Oct. 24, 2019, 3:22 a.m. UTC | #2
On Wed, Oct 23, 2019 at 4:03 PM Jonathan Lemon <jonathan.lemon@gmail.com> wrote:
>
> On 21 Oct 2019, at 1:16, Magnus Karlsson wrote:
>
> > Having Rx-only AF_XDP sockets can potentially lead to a crash in the
> > system by a NULL pointer dereference in xsk_umem_consume_tx(). This
> > function iterates through a list of all sockets tied to a umem and
> > checks if there are any packets to send on the Tx ring. Rx-only
> > sockets do not have a Tx ring, so this will cause a NULL pointer
> > dereference. This will happen if you have registered one or more
> > Rx-only sockets to a umem and the driver is checking the Tx ring even
> > on Rx, or if the XDP_SHARED_UMEM mode is used and there is a mix of
> > Rx-only and other sockets tied to the same umem.
> >
> > Fixed by only putting sockets with a Tx component on the list that
> > xsk_umem_consume_tx() iterates over.
>
> A future improvement might be renaming umem->xsk_list to umem->xsk_tx_list
> or similar, in order to make it clear that the list is only used on the
> TX path.
>
> >
> > Fixes: ac98d8aab61b ("xsk: wire upp Tx zero-copy functions")
> > Reported-by: Kal Cutter Conley <kal.conley@dectris.com>
> > Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
>
> Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>

Applied. Thanks
Magnus Karlsson Oct. 24, 2019, 12:53 p.m. UTC | #3
On Thu, Oct 24, 2019 at 12:47 PM Jonathan Lemon
<jonathan.lemon@gmail.com> wrote:
>
> On 21 Oct 2019, at 1:16, Magnus Karlsson wrote:
>
> > Having Rx-only AF_XDP sockets can potentially lead to a crash in the
> > system by a NULL pointer dereference in xsk_umem_consume_tx(). This
> > function iterates through a list of all sockets tied to a umem and
> > checks if there are any packets to send on the Tx ring. Rx-only
> > sockets do not have a Tx ring, so this will cause a NULL pointer
> > dereference. This will happen if you have registered one or more
> > Rx-only sockets to a umem and the driver is checking the Tx ring even
> > on Rx, or if the XDP_SHARED_UMEM mode is used and there is a mix of
> > Rx-only and other sockets tied to the same umem.
> >
> > Fixed by only putting sockets with a Tx component on the list that
> > xsk_umem_consume_tx() iterates over.
>
> A future improvement might be renaming umem->xsk_list to umem->xsk_tx_list
> or similar, in order to make it clear that the list is only used on the
> TX path.

Agreed. Had that exact name in my first internal version of the patch
:-), but that rename touched a lot of places so it obfuscated the fix
and therefore I removed it to make it clearer. But I can submit a
patch with the rename to bpf-next.

> >
> > Fixes: ac98d8aab61b ("xsk: wire upp Tx zero-copy functions")
> > Reported-by: Kal Cutter Conley <kal.conley@dectris.com>
> > Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
>
> Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>
>
> --
> Jonathan
Alexei Starovoitov Oct. 24, 2019, 3:19 p.m. UTC | #4
On Thu, Oct 24, 2019 at 5:53 AM Magnus Karlsson
<magnus.karlsson@gmail.com> wrote:
>
> On Thu, Oct 24, 2019 at 12:47 PM Jonathan Lemon
> <jonathan.lemon@gmail.com> wrote:
> >
> > On 21 Oct 2019, at 1:16, Magnus Karlsson wrote:
> >
> > > Having Rx-only AF_XDP sockets can potentially lead to a crash in the
> > > system by a NULL pointer dereference in xsk_umem_consume_tx(). This
> > > function iterates through a list of all sockets tied to a umem and
> > > checks if there are any packets to send on the Tx ring. Rx-only
> > > sockets do not have a Tx ring, so this will cause a NULL pointer
> > > dereference. This will happen if you have registered one or more
> > > Rx-only sockets to a umem and the driver is checking the Tx ring even
> > > on Rx, or if the XDP_SHARED_UMEM mode is used and there is a mix of
> > > Rx-only and other sockets tied to the same umem.
> > >
> > > Fixed by only putting sockets with a Tx component on the list that
> > > xsk_umem_consume_tx() iterates over.
> >
> > A future improvement might be renaming umem->xsk_list to umem->xsk_tx_list
> > or similar, in order to make it clear that the list is only used on the
> > TX path.
>
> Agreed. Had that exact name in my first internal version of the patch
> :-), but that rename touched a lot of places so it obfuscated the fix
> and therefore I removed it to make it clearer. But I can submit a
> patch with the rename to bpf-next.

please do so after bpf fixes will travel to net->linus->net-next->bpf-next
Dealing with merge conflicts is not worth it otherwise.

Patch
diff mbox series

diff --git a/net/xdp/xdp_umem.c b/net/xdp/xdp_umem.c
index 16d5f35..3049af2 100644
--- a/net/xdp/xdp_umem.c
+++ b/net/xdp/xdp_umem.c
@@ -27,6 +27,9 @@  void xdp_add_sk_umem(struct xdp_umem *umem, struct xdp_sock *xs)
 {
 	unsigned long flags;
 
+	if (!xs->tx)
+		return;
+
 	spin_lock_irqsave(&umem->xsk_list_lock, flags);
 	list_add_rcu(&xs->list, &umem->xsk_list);
 	spin_unlock_irqrestore(&umem->xsk_list_lock, flags);
@@ -36,6 +39,9 @@  void xdp_del_sk_umem(struct xdp_umem *umem, struct xdp_sock *xs)
 {
 	unsigned long flags;
 
+	if (!xs->tx)
+		return;
+
 	spin_lock_irqsave(&umem->xsk_list_lock, flags);
 	list_del_rcu(&xs->list);
 	spin_unlock_irqrestore(&umem->xsk_list_lock, flags);