[PATCH net] xen-netback: avoid race that can lead to NULL pointer dereference

[PATCH net] xen-netback: avoid race that can lead to NULL pointer dereference

[Paul Durrant]

Commit 2ac061ce97f4 ("xen/netback: cleanup init and deinit code")
introduced a problem. In function xenvif_disconnect_queue(), the value of
queue->rx_irq is zeroed *before* queue->task is stopped. Unfortunately that
task may call notify_remote_via_irq(queue->rx_irq) and calling that
function with a zero value results in a NULL pointer dereference in
evtchn_from_irq().

This patch simply re-orders things, stopping all tasks before zero-ing the
irq values, thereby avoiding the possibility of the race.

Signed-off-by: Paul Durrant <pdurrant@amazon.com>
---
Cc: Juergen Gross <jgross@suse.com>
Cc: Jakub Kicinski <jakub.kicinski@netronome.com>
Cc: Wei Liu <wei.liu@kernel.org>
Cc: "David S. Miller" <davem@davemloft.net>
---
 drivers/net/xen-netback/interface.c | 24 ++++++++++++------------
 1 file changed, 12 insertions(+), 12 deletions(-)

diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index 68dd7bb07ca6..f15ba3de6195 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -628,18 +628,6 @@ int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref,
 
 static void xenvif_disconnect_queue(struct xenvif_queue *queue)
 {
-	if (queue->tx_irq) {
-		unbind_from_irqhandler(queue->tx_irq, queue);
-		if (queue->tx_irq == queue->rx_irq)
-			queue->rx_irq = 0;
-		queue->tx_irq = 0;
-	}
-
-	if (queue->rx_irq) {
-		unbind_from_irqhandler(queue->rx_irq, queue);
-		queue->rx_irq = 0;
-	}
-
 	if (queue->task) {
 		kthread_stop(queue->task);
 		queue->task = NULL;
@@ -655,6 +643,18 @@ static void xenvif_disconnect_queue(struct xenvif_queue *queue)
 		queue->napi.poll = NULL;
 	}
 
+	if (queue->tx_irq) {
+		unbind_from_irqhandler(queue->tx_irq, queue);
+		if (queue->tx_irq == queue->rx_irq)
+			queue->rx_irq = 0;
+		queue->tx_irq = 0;
+	}
+
+	if (queue->rx_irq) {
+		unbind_from_irqhandler(queue->rx_irq, queue);
+		queue->rx_irq = 0;
+	}
+
 	xenvif_unmap_frontend_data_rings(queue);
 }
 
-- 
2.20.1

Re: [PATCH net] xen-netback: avoid race that can lead to NULL pointer dereference

[Wei Liu]

On Thu, Dec 12, 2019 at 12:37:23PM +0000, Paul Durrant wrote:
> Commit 2ac061ce97f4 ("xen/netback: cleanup init and deinit code")
> introduced a problem. In function xenvif_disconnect_queue(), the value of
> queue->rx_irq is zeroed *before* queue->task is stopped. Unfortunately that
> task may call notify_remote_via_irq(queue->rx_irq) and calling that
> function with a zero value results in a NULL pointer dereference in
> evtchn_from_irq().
> 
> This patch simply re-orders things, stopping all tasks before zero-ing the
> irq values, thereby avoiding the possibility of the race.
> 
> Signed-off-by: Paul Durrant <pdurrant@amazon.com>

Acked-by: Wei Liu <wei.liu@kernel.org>

Re: [PATCH net] xen-netback: avoid race that can lead to NULL pointer dereference

[David Miller]

From: Paul Durrant <pdurrant@amazon.com>
Date: Thu, 12 Dec 2019 12:37:23 +0000

> Commit 2ac061ce97f4 ("xen/netback: cleanup init and deinit code")
> introduced a problem. In function xenvif_disconnect_queue(), the value of
> queue->rx_irq is zeroed *before* queue->task is stopped. Unfortunately that
> task may call notify_remote_via_irq(queue->rx_irq) and calling that
> function with a zero value results in a NULL pointer dereference in
> evtchn_from_irq().
> 
> This patch simply re-orders things, stopping all tasks before zero-ing the
> irq values, thereby avoiding the possibility of the race.
> 
> Signed-off-by: Paul Durrant <pdurrant@amazon.com>

Please repost this with an appropriate Fixes: tag.

And then you can removed the explicit commit reference from the log message
and simply say "The commit mentioned in the Fixes tag introduced a problen ..."