RE: [PATCH 1/2] vfio/xe: Reorganize the init to decouple migration from reset

[Tian, Kevin]

> From: Winiarski, Michal <michal.winiarski@xxxxxxxxx>
> Sent: Thursday, April 9, 2026 12:04 AM
> 
> xe_vfio_pci_migration_ops = {
>  static void xe_vfio_pci_migration_init(struct xe_vfio_pci_core_device
> *xe_vdev)
>  {
>  	struct vfio_device *core_vdev = &xe_vdev->core_device.vdev;
> -	struct pci_dev *pdev = to_pci_dev(core_vdev->dev);
> -	struct xe_device *xe = xe_sriov_vfio_get_pf(pdev);
> 
> -	if (!xe)
> -		return;
> -	if (!xe_sriov_vfio_migration_supported(xe))
> +	if (!xe_sriov_vfio_migration_supported(xe_vdev->xe))
>  		return;
> 
>  	mutex_init(&xe_vdev->state_mutex);
>  	spin_lock_init(&xe_vdev->reset_lock);
> 

those two belongs to xe_vfio_pci_vf_init() as a non-migratable
vf still requires them, e.g. in xe_vfio_pci_reset_done()?