From: Baolu Lu baolu.lu@linux.intel.com Sent: Wednesday, August 2, 2023 4:12 PM
On 2023/8/2 16:09, Baolu Lu wrote:
On 2023/8/2 15:10, Tian, Kevin wrote:
+ pasid_set_slptr(pte, virt_to_phys(pgd)); + pasid_set_fault_enable(pte); + pasid_set_domain_id(pte, did); + pasid_set_address_width(pte, s2_domain->agaw); + pasid_set_page_snoop(pte, !!ecap_smpwc(iommu->ecap)); + pasid_set_translation_type(pte, PASID_ENTRY_PGTT_NESTED); + pasid_set_present(pte); + spin_unlock(&iommu->lock);
this lacks of handling of force_snooping
If stage-2 domain has force_snooping attribute set, then we should set the bit field in the pasid entry, right?
How about below additional change?
yes
diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c index 19ac4084913b..86db81ec91f1 100644 --- a/drivers/iommu/intel/pasid.c +++ b/drivers/iommu/intel/pasid.c @@ -827,6 +827,9 @@ int intel_pasid_setup_nested(struct intel_iommu *iommu, struct device *dev, if (s1_cfg->flags & IOMMU_VTD_S1_EAFE) pasid_set_eafe(pte);
+ if (s2_domain>force_snooping)
if (s2_domain->force_snooping)Sorry for typo.
+ pasid_set_pgsnp(pte);
pasid_set_slptr(pte, virt_to_phys(pgd)); pasid_set_fault_enable(pte); pasid_set_domain_id(pte, did);
Best regards, baolu