[PATCH v4 09/12] iommu/arm-smmu-v3: Support IOMMU_DOMAIN_NESTED
Jason Gunthorpe
jgg at nvidia.com
Mon Nov 4 09:19:31 PST 2024
On Thu, Oct 31, 2024 at 02:21:11PM +0800, Zhangfei Gao wrote:
> > +static struct iommu_domain *
> > +arm_vsmmu_alloc_domain_nested(struct iommufd_viommu *viommu, u32 flags,
> > + const struct iommu_user_data *user_data)
> > +{
> > + struct arm_vsmmu *vsmmu = container_of(viommu, struct arm_vsmmu, core);
> > + struct arm_smmu_nested_domain *nested_domain;
> > + struct iommu_hwpt_arm_smmuv3 arg;
> > + int ret;
> > +
> > + if (flags)
> > + return ERR_PTR(-EOPNOTSUPP);
>
> This check fails when using user page fault, with flags =
> IOMMU_HWPT_FAULT_ID_VALID (4)
> Strange, the check is not exist in last version?
>
> iommufd_viommu_alloc_hwpt_nested ->
> viommu->ops->alloc_domain_nested(viommu, flags, user_data) ->
> arm_vsmmu_alloc_domain_nested
It should permit IOMMU_HWPT_FAULT_ID_VALID, I'll add this hunk:
--- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3-iommufd.c
+++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3-iommufd.c
@@ -178,12 +178,18 @@ arm_vsmmu_alloc_domain_nested(struct iommufd_viommu *viommu, u32 flags,
const struct iommu_user_data *user_data)
{
struct arm_vsmmu *vsmmu = container_of(viommu, struct arm_vsmmu, core);
+ const u32 SUPPORTED_FLAGS = IOMMU_HWPT_FAULT_ID_VALID;
struct arm_smmu_nested_domain *nested_domain;
struct iommu_hwpt_arm_smmuv3 arg;
bool enable_ats = false;
int ret;
- if (flags)
+ /*
+ * Faults delivered to the nested domain are faults that originated by
+ * the S1 in the domain. The core code will match all PASIDs when
+ * delivering the fault due to user_pasid_table
+ */
+ if (flags & ~SUPPORTED_FLAGS)
return ERR_PTR(-EOPNOTSUPP);
ret = iommu_copy_struct_from_user(&arg, user_data,
Thanks,
Jason
More information about the linux-arm-kernel
mailing list