On Thu, Aug 05, 2021 at 11:33:11AM -0600, Alex Williamson wrote:

quoted

+static int vfio_pci_is_device_in_set(struct pci_dev *pdev, void *data)
+{
+	struct vfio_device_set *dev_set = data;
+	struct vfio_device *cur;
+
+	lockdep_assert_held(&dev_set->lock);
+
+	list_for_each_entry(cur, &dev_set->device_list, dev_set_list)
+		if (cur->dev == &pdev->dev)
+			return 0;
+	return -EBUSY;
+}
+
+static bool vfio_pci_dev_set_needs_reset(struct vfio_device_set *dev_set)

Slight nit on the name here since we're essentially combining
needs_reset along with the notion of the device being unused.  I'm not
sure, maybe "should_reset"?  Otherwise it looks ok.  Thanks,

What I did is add a new function vfio_pci_dev_set_resettable() which
pulls in three parts of logic that can be be shared with the
VFIO_DEVICE_PCI_HOT_RESET change in the next patch. That leaves this
function as purely needs_reset.

In turn the VFIO_DEVICE_PCI_HOT_RESET patch gets the same treatment
where it becomes a dev_set centric API just like this.

I'll send it as a v4.

Thanks,
Jason