4.7-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Sakari Ailus <sakari.ai...@linux.intel.com>

commit 126f40298446a82116e1f92a1aaf72b8c8228fae upstream.

An earlier patch fixing an input validation issue introduced another
issue: vb2_core_dqbuf() is called with pb argument value NULL in some
cases, causing a NULL pointer dereference. Fix this by skipping the
verification as there's nothing to verify.

Fixes: e7e0c3e26587 ("[media] videobuf2-core: Check user space planes array in 
dqbuf")

Signed-off-by: David R <da...@unsolicited.net>
Signed-off-by: Sakari Ailus <sakari.ai...@linux.intel.com>
Reviewed-by: Hans Verkuil <hans.verk...@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mche...@s-opensource.com>
Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org>

---
 drivers/media/v4l2-core/videobuf2-core.c |   10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

--- a/drivers/media/v4l2-core/videobuf2-core.c
+++ b/drivers/media/v4l2-core/videobuf2-core.c
@@ -1648,7 +1648,7 @@ static int __vb2_get_done_vb(struct vb2_
                             void *pb, int nonblocking)
 {
        unsigned long flags;
-       int ret;
+       int ret = 0;
 
        /*
         * Wait for at least one buffer to become available on the done_list.
@@ -1664,10 +1664,12 @@ static int __vb2_get_done_vb(struct vb2_
        spin_lock_irqsave(&q->done_lock, flags);
        *vb = list_first_entry(&q->done_list, struct vb2_buffer, done_entry);
        /*
-        * Only remove the buffer from done_list if v4l2_buffer can handle all
-        * the planes.
+        * Only remove the buffer from done_list if all planes can be
+        * handled. Some cases such as V4L2 file I/O and DVB have pb
+        * == NULL; skip the check then as there's nothing to verify.
         */
-       ret = call_bufop(q, verify_planes_array, *vb, pb);
+       if (pb)
+               ret = call_bufop(q, verify_planes_array, *vb, pb);
        if (!ret)
                list_del(&(*vb)->done_entry);
        spin_unlock_irqrestore(&q->done_lock, flags);


Reply via email to