> From: Jason Gunthorpe <jgg@xxxxxxxxxx>
> Sent: Tuesday, April 23, 2024 8:44 PM
>
> On Fri, Apr 12, 2024 at 01:21:19AM -0700, Yi Liu wrote:
> > +int vfio_iommufd_physical_pasid_attach_ioas(struct vfio_device *vdev,
> > + u32 pasid, u32 *pt_id)
> > +{
> > + int rc;
> > +
> > + lockdep_assert_held(&vdev->dev_set->lock);
> > +
> > + if (WARN_ON(!vdev->iommufd_device))
> > + return -EINVAL;
> > +
> > + rc = ida_get_lowest(&vdev->pasids, pasid, pasid);
>
> A helper inline
>
> bool ida_is_allocate(&ida, id)
>
> Would be nicer for that
>
> > diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
> > index cb5b7f865d58..e0198851ffd2 100644
> > --- a/drivers/vfio/pci/vfio_pci.c
> > +++ b/drivers/vfio/pci/vfio_pci.c
> > @@ -142,6 +142,8 @@ static const struct vfio_device_ops vfio_pci_ops = {
> > .unbind_iommufd = vfio_iommufd_physical_unbind,
> > .attach_ioas = vfio_iommufd_physical_attach_ioas,
> > .detach_ioas = vfio_iommufd_physical_detach_ioas,
> > + .pasid_attach_ioas = vfio_iommufd_physical_pasid_attach_ioas,
> > + .pasid_detach_ioas = vfio_iommufd_physical_pasid_detach_ioas,
> > };
>
> This should be copied into mlx5 and nvgrace-gpu at least as well
>
I'd prefer to the driver owners to add them separately. They know
their hardware and can do proper test.
