iommu/vt-d: Add missing device iotlb flush for parent domain

ATS-capable devices cache the result of nested translation. This result
relies on the mappings in s2 domain (a.k.a. parent). When there are
modifications in the s2 domain, the related nested translation caches on
the device should be flushed. This includes the devices that are attached
to the s1 domain. However, the existing code ignores this fact to only
loops its own devices.

As there is no easy way to identify the exact set of nested translations
affected by the change of s2 domain. So, this just flushes the entire
device iotlb on the device.

As above, driver loops the s2 domain's s1_domains list and loops the
devices list of each s1_domain to flush the entire device iotlb on the
devices.

Fixes: b41e38e ("iommu/vt-d: Add nested domain allocation")
Signed-off-by: Yi Liu <[email protected]>
Reviewed-by: Kevin Tian <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Lu Baolu <[email protected]>
Signed-off-by: Joerg Roedel <[email protected]>

drivers/iommu/intel/iommu.c

@@ -1461,12 +1461,30 @@ static void parent_domain_flush(struct dmar_domain *domain,
 
 	spin_lock(&domain->s1_lock);
 	list_for_each_entry(s1_domain, &domain->s1_domains, s2_link) {
+		struct device_domain_info *device_info;
 		struct iommu_domain_info *info;
+		unsigned long flags;
 		unsigned long i;
 
 		xa_for_each(&s1_domain->iommu_array, i, info)
 			__iommu_flush_iotlb_psi(info->iommu, info->did,
 						pfn, pages, ih);
+
+		if (!s1_domain->has_iotlb_device)
+			continue;
+
+		spin_lock_irqsave(&s1_domain->lock, flags);
+		list_for_each_entry(device_info, &s1_domain->devices, link)
+			/*
+			 * Address translation cache in device side caches the
+			 * result of nested translation. There is no easy way
+			 * to identify the exact set of nested translations
+			 * affected by a change in S2. So just flush the entire
+			 * device cache.
+			 */
+			__iommu_flush_dev_iotlb(device_info, 0,
+						MAX_AGAW_PFN_WIDTH);
+		spin_unlock_irqrestore(&s1_domain->lock, flags);
 	}
 	spin_unlock(&domain->s1_lock);
 }