Fix vhost-user buffer over-read on ram hot-unplug
authorRaphael Norwitz <raphael.norwitz@nutanix.com>
Fri, 17 Jul 2020 04:21:30 +0000 (04:21 +0000)
committerMichael S. Tsirkin <mst@redhat.com>
Mon, 27 Jul 2020 14:28:28 +0000 (10:28 -0400)
The VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS vhost-user protocol
feature introduced a shadow-table, used by the backend to dynamically
determine how a vdev's memory regions have changed since the last
vhost_user_set_mem_table() call. On hot-remove, a memmove() operation
is used to overwrite the removed shadow region descriptor(s). The size
parameter of this memmove was off by 1 such that if a VM with a backend
supporting the VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS filled it's
shadow-table (by performing the maximum number of supported hot-add
operatons) and attempted to remove the last region, Qemu would read an
out of bounds value and potentially crash.

This change fixes the memmove() bounds such that this erroneous read can
never happen.

Signed-off-by: Peter Turschmid <peter.turschm@nutanix.com>
Signed-off-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <1594799958-31356-1-git-send-email-raphael.norwitz@nutanix.com>
Fixes: f1aeb14b0809 ("Transmit vhost-user memory regions individually")
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: qemu-stable@nongnu.org
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
hw/virtio/vhost-user.c

index 31231218dc6a8d9e92a48b6704a4c407875861c3..d7e24237625f791774437add0e710a2b119e7982 100644 (file)
@@ -672,7 +672,7 @@ static int send_remove_regions(struct vhost_dev *dev,
         memmove(&u->shadow_regions[shadow_reg_idx],
                 &u->shadow_regions[shadow_reg_idx + 1],
                 sizeof(struct vhost_memory_region) *
-                (u->num_shadow_regions - shadow_reg_idx));
+                (u->num_shadow_regions - shadow_reg_idx - 1));
         u->num_shadow_regions--;
     }