Merge tag 'iommu-updates-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu

Pull iommu updates from Joerg Roedel:

 - ARM SMMU and Mediatek updates from Will Deacon:
     - Support for MT8192 IOMMU from Mediatek
     - Arm v7s io-pgtable extensions for MT8192
     - Removal of TLBI_ON_MAP quirk
     - New Qualcomm compatible strings
     - Allow SVA without hardware broadcast TLB maintenance on SMMUv3
     - Virtualization Host Extension support for SMMUv3 (SVA)
     - Allow SMMUv3 PMU perf driver to be built independently from IOMMU

 - Some tidy-up in IOVA and core code

 - Conversion of the AMD IOMMU code to use the generic IO-page-table
   framework

 - Intel VT-d updates from Lu Baolu:
     - Audit capability consistency among different IOMMUs
     - Add SATC reporting structure support
     - Add iotlb_sync_map callback support

 - SDHI support for Renesas IOMMU driver

 - Misc cleanups and other small improvments

* tag 'iommu-updates-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu: (94 commits)
  iommu/amd: Fix performance counter initialization
  MAINTAINERS: repair file pattern in MEDIATEK IOMMU DRIVER
  iommu/mediatek: Fix error code in probe()
  iommu/mediatek: Fix unsigned domid comparison with less than zero
  iommu/vt-d: Parse SATC reporting structure
  iommu/vt-d: Add new enum value and structure for SATC
  iommu/vt-d: Add iotlb_sync_map callback
  iommu/vt-d: Move capability check code to cap_audit files
  iommu/vt-d: Audit IOMMU Capabilities and add helper functions
  iommu/vt-d: Fix 'physical' typos
  iommu: Properly pass gfp_t in _iommu_map() to avoid atomic sleeping
  iommu/vt-d: Fix compile error [-Werror=implicit-function-declaration]
  driver/perf: Remove ARM_SMMU_V3_PMU dependency on ARM_SMMU_V3
  MAINTAINERS: Add entry for MediaTek IOMMU
  iommu/mediatek: Add mt8192 support
  iommu/mediatek: Remove unnecessary check in attach_device
  iommu/mediatek: Support master use iova over 32bit
  iommu/mediatek: Add iova reserved function
  iommu/mediatek: Support for multi domains
  iommu/mediatek: Add get_domain_id from dev->dma_range_map
  ...

1 files changed, 21 insertions, 52 deletions

diff --git a/drivers/iommu/intel/svm.c b/drivers/iommu/intel/svm.c
index 18a9f05df407..574a7e657a9a 100644
--- a/drivers/iommu/intel/svm.c
+++ b/drivers/iommu/intel/svm.c
@@ -123,53 +123,16 @@ static void __flush_svm_range_dev(struct intel_svm *svm,
 				  unsigned long address,
 				  unsigned long pages, int ih)
 {
-	struct qi_desc desc;
+	struct device_domain_info *info = get_domain_info(sdev->dev);
 
-	if (pages == -1) {
-		desc.qw0 = QI_EIOTLB_PASID(svm->pasid) |
-			QI_EIOTLB_DID(sdev->did) |
-			QI_EIOTLB_GRAN(QI_GRAN_NONG_PASID) |
-			QI_EIOTLB_TYPE;
-		desc.qw1 = 0;
-	} else {
-		int mask = ilog2(__roundup_pow_of_two(pages));
-
-		desc.qw0 = QI_EIOTLB_PASID(svm->pasid) |
-				QI_EIOTLB_DID(sdev->did) |
-				QI_EIOTLB_GRAN(QI_GRAN_PSI_PASID) |
-				QI_EIOTLB_TYPE;
-		desc.qw1 = QI_EIOTLB_ADDR(address) |
-				QI_EIOTLB_IH(ih) |
-				QI_EIOTLB_AM(mask);
-	}
-	desc.qw2 = 0;
-	desc.qw3 = 0;
-	qi_submit_sync(sdev->iommu, &desc, 1, 0);
-
-	if (sdev->dev_iotlb) {
-		desc.qw0 = QI_DEV_EIOTLB_PASID(svm->pasid) |
-				QI_DEV_EIOTLB_SID(sdev->sid) |
-				QI_DEV_EIOTLB_QDEP(sdev->qdep) |
-				QI_DEIOTLB_TYPE;
-		if (pages == -1) {
-			desc.qw1 = QI_DEV_EIOTLB_ADDR(-1ULL >> 1) |
-					QI_DEV_EIOTLB_SIZE;
-		} else if (pages > 1) {
-			/* The least significant zero bit indicates the size. So,
-			 * for example, an "address" value of 0x12345f000 will
-			 * flush from 0x123440000 to 0x12347ffff (256KiB). */
-			unsigned long last = address + ((unsigned long)(pages - 1) << VTD_PAGE_SHIFT);
-			unsigned long mask = __rounddown_pow_of_two(address ^ last);
-
-			desc.qw1 = QI_DEV_EIOTLB_ADDR((address & ~mask) |
-					(mask - 1)) | QI_DEV_EIOTLB_SIZE;
-		} else {
-			desc.qw1 = QI_DEV_EIOTLB_ADDR(address);
-		}
-		desc.qw2 = 0;
-		desc.qw3 = 0;
-		qi_submit_sync(sdev->iommu, &desc, 1, 0);
-	}
+	if (WARN_ON(!pages))
+		return;
+
+	qi_flush_piotlb(sdev->iommu, sdev->did, svm->pasid, address, pages, ih);
+	if (info->ats_enabled)
+		qi_flush_dev_iotlb_pasid(sdev->iommu, sdev->sid, info->pfsid,
+					 svm->pasid, sdev->qdep, address,
+					 order_base_2(pages));
 }
 
 static void intel_flush_svm_range_dev(struct intel_svm *svm,
@@ -948,10 +911,8 @@ static irqreturn_t prq_event_thread(int irq, void *d)
 		u64 address;
 
 		handled = 1;
-
 		req = &iommu->prq[head / sizeof(*req)];
-
-		result = QI_RESP_FAILURE;
+		result = QI_RESP_INVALID;
 		address = (u64)req->addr << VTD_PAGE_SHIFT;
 		if (!req->pasid_present) {
 			pr_err("%s: Page request without PASID: %08llx %08llx\n",
@@ -989,7 +950,6 @@ static irqreturn_t prq_event_thread(int irq, void *d)
 			rcu_read_unlock();
 		}
 
-		result = QI_RESP_INVALID;
 		/* Since we're using init_mm.pgd directly, we should never take
 		 * any faults on kernel addresses. */
 		if (!svm->mm)
@@ -1079,8 +1039,17 @@ prq_advance:
 	 * Clear the page request overflow bit and wake up all threads that
 	 * are waiting for the completion of this handling.
 	 */
-	if (readl(iommu->reg + DMAR_PRS_REG) & DMA_PRS_PRO)
-		writel(DMA_PRS_PRO, iommu->reg + DMAR_PRS_REG);
+	if (readl(iommu->reg + DMAR_PRS_REG) & DMA_PRS_PRO) {
+		pr_info_ratelimited("IOMMU: %s: PRQ overflow detected\n",
+				    iommu->name);
+		head = dmar_readq(iommu->reg + DMAR_PQH_REG) & PRQ_RING_MASK;
+		tail = dmar_readq(iommu->reg + DMAR_PQT_REG) & PRQ_RING_MASK;
+		if (head == tail) {
+			writel(DMA_PRS_PRO, iommu->reg + DMAR_PRS_REG);
+			pr_info_ratelimited("IOMMU: %s: PRQ overflow cleared",
+					    iommu->name);
+		}
+	}
 
 	if (!completion_done(&iommu->prq_complete))
 		complete(&iommu->prq_complete);