On Thu, Dec 11, 2025 at 03:05:49AM -0600, Aaron Lo wrote:
> The VirtIO specification (section 5.5.2) states that the stats queue
> is only present if the VIRTIO_BALLOON_F_STATS_VQ feature is
> negotiated. QEMU currently creates the statsq unconditionally.
> 
> This patch guards statsq creation so it occurs only when the
> feature bit is enabled.
> 
> Resolves: https://gitlab.com/qemu-project/qemu/-/issues/3188
> 
> Signed-off-by: Aaron Lo <[email protected]>


did you test this with linux guests and with
free page hinting?



The issue is that sadly everyone is out of spec here.


We will have to fix the spec.


> ---
>  hw/virtio/virtio-balloon.c | 9 +++++++--
>  1 file changed, 7 insertions(+), 2 deletions(-)
> 
> diff --git a/hw/virtio/virtio-balloon.c b/hw/virtio/virtio-balloon.c
> index 02cdd807d7..f5d4d5f60c 100644
> --- a/hw/virtio/virtio-balloon.c
> +++ b/hw/virtio/virtio-balloon.c
> @@ -892,7 +892,10 @@ static void virtio_balloon_device_realize(DeviceState 
> *dev, Error **errp)
>  
>      s->ivq = virtio_add_queue(vdev, 128, virtio_balloon_handle_output);
>      s->dvq = virtio_add_queue(vdev, 128, virtio_balloon_handle_output);
> -    s->svq = virtio_add_queue(vdev, 128, virtio_balloon_receive_stats);
> +
> +    if (virtio_has_feature(s->host_features, VIRTIO_BALLOON_F_STATS_VQ)) {
> +        s->svq = virtio_add_queue(vdev, 128, virtio_balloon_receive_stats);
> +    }
>  
>      if (virtio_has_feature(s->host_features, 
> VIRTIO_BALLOON_F_FREE_PAGE_HINT)) {
>          s->free_page_vq = virtio_add_queue(vdev, VIRTQUEUE_MAX_SIZE,
> @@ -932,7 +935,9 @@ static void virtio_balloon_device_unrealize(DeviceState 
> *dev)
>  
>      virtio_delete_queue(s->ivq);
>      virtio_delete_queue(s->dvq);
> -    virtio_delete_queue(s->svq);
> +    if (s->svq) {
> +        virtio_delete_queue(s->svq);
> +    }
>      if (s->free_page_vq) {
>          virtio_delete_queue(s->free_page_vq);
>      }
> 
> ---
> base-commit: 9c23f2a7b0b45277693a14074b1aaa827eecdb92
> change-id: 20251211-balloon-check-stats-feature-7ea658e038ce
> 
> Best regards,
> -- 
> Aaron Lo <[email protected]>


Reply via email to