On Thu, Oct 31, 2024 at 10:14:59AM -0300, Jason Gunthorpe wrote: > On Wed, Oct 30, 2024 at 02:34:33PM -0700, Nicolin Chen wrote: > > +static struct iommufd_hwpt_nested * > > +iommufd_viommu_alloc_hwpt_nested(struct iommufd_viommu *viommu, u32 flags, > > + const struct iommu_user_data *user_data) > > +{ > > + struct iommufd_hwpt_nested *hwpt_nested; > > + struct iommufd_hw_pagetable *hwpt; > > + int rc; > > + > > + if ((flags & ~IOMMU_HWPT_FAULT_ID_VALID) || !user_data->len) > > + return ERR_PTR(-EOPNOTSUPP); > > Now that flags exists we are expecting the driver to validate it, so > this check on flags should be dropped? > > Or we go back to hiding FAULT_ID_VALID from the driver (and keep the flags)? If we plan to fix the HWPT-based side, I think we should just drop this check on flags here. Thanks Nicolin