On Thu, Aug 13, 2026 at 10:49:36AM -0300, Jason Gunthorpe wrote:
> On Tue, Aug 11, 2026 at 11:50:20PM +0800, Yu Zhang wrote:
>
> > +static void hv_iommu_release_device(struct device *dev)
> > +{
> > + struct hv_iommu_endpoint *vdev = dev_iommu_priv_get(dev);
> > + struct pci_dev *pdev = to_pci_dev(dev);
> > +
> > + if (pdev->ats_enabled)
> > + pci_disable_ats(pdev);
>
> Since attaching blocking does this and you set blocking as your
> release domain, this is redundant.
>
Indeed, it is redundant. Will remove. Thank you, Jason.
B.R.
Yu
> The new flow looks alot better, removing the detach operation entirely
> makes sense if the HV can support attach on attach
>
> Jason
>
