Linux Virtualization - Sep 2020 - [PATCH] vhost: Don't call vq_access_ok() when using IOTLB

On Mon, Sep 28, 2020 at 02:35:04PM +0200, Greg Kurz
wrote:
> When the IOTLB device is enabled, the vring addresses we get from
> userspace are GIOVAs. It is thus wrong to pass them to vq_access_ok()
> which only takes HVAs. The IOTLB map is likely empty at this stage,
> so there isn't much that can be done with these GIOVAs. Access
validation
> will be performed at IOTLB prefetch time anyway.
> 
> BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1883084
> Fixes: 6b1e6cc7855b ("vhost: new device IOTLB API")
> Cc: jasowang at redhat.com
> CC: stable at vger.kernel.org # 4.14+
> Signed-off-by: Greg Kurz <groug at kaod.org>
> ---
>  drivers/vhost/vhost.c |    5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
> index b45519ca66a7..6296e33df31d 100644
> --- a/drivers/vhost/vhost.c
> +++ b/drivers/vhost/vhost.c
> @@ -1509,7 +1509,10 @@ static long vhost_vring_set_addr(struct vhost_dev
*d,
>  	 * If it is not, we don't as size might not have been setup.
>  	 * We will verify when backend is configured. */
>  	if (vq->private_data) {
> -		if (!vq_access_ok(vq, vq->num,
> +		/* If an IOTLB device is present, the vring addresses are
> +		 * GIOVAs. Access will be validated during IOTLB prefetch. */
> +		if (!vq->iotlb &&
> +		    !vq_access_ok(vq, vq->num,
>  			(void __user *)(unsigned long)a.desc_user_addr,
>  			(void __user *)(unsigned long)a.avail_user_addr,
>  			(void __user *)(unsigned long)a.used_user_addr))
OK I think you are right here.

Jason, can you ack pls?

However, I think a cleaner way to check this is by moving
the following check from vhost_vq_access_ok to vq_access_ok:

        /* Access validation occurs at prefetch time with IOTLB */
        if (vq->iotlb)
                return true;

>