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

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

From: Will Deacon <redacted>

commit b92b1b89a33c172c075edccf6afb0edc41d851fd upstream.

Virtio devices may attempt to add descriptors to a virtqueue from atomic
context using GFP_ATOMIC allocation. This is problematic because such
allocations can fall outside of the lowmem mapping, causing virt_to_phys
to report bogus physical addresses which are subsequently passed to
userspace via the buffers for the virtual device.

This patch masks out __GFP_HIGH and __GFP_HIGHMEM from the requested
flags when allocating descriptors for a virtqueue. If an atomic
allocation is requested and later fails, we will return -ENOSPC which
will be handled by the driver.

Cc: Sasha Levin <redacted>
Signed-off-by: Will Deacon <redacted>
Signed-off-by: Rusty Russell <redacted>
Signed-off-by: Ben Hutchings <redacted>
---
 drivers/virtio/virtio_ring.c |    7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
index e639584..286c30c 100644
--- a/drivers/virtio/virtio_ring.c
+++ b/drivers/virtio/virtio_ring.c

@@ -135,6 +135,13 @@ static int vring_add_indirect(struct vring_virtqueue *vq,
 	unsigned head;
 	int i;
 
+	/*
+	 * We require lowmem mappings for the descriptors because
+	 * otherwise virt_to_phys will give us bogus addresses in the
+	 * virtqueue.
+	 */
+	gfp &= ~(__GFP_HIGHMEM | __GFP_HIGH);
+
 	desc = kmalloc((out + in) * sizeof(struct vring_desc), gfp);
 	if (!desc)
 		return -ENOMEM;