1 files changed, 385 insertions, 339 deletions

diff --git a/drivers/iommu/amd/iommu.c b/drivers/iommu/amd/iommu.c
index 95bd7c25ba..fcc987f5d4 100644
--- a/drivers/iommu/amd/iommu.c
+++ b/drivers/iommu/amd/iommu.c
@@ -37,6 +37,7 @@
 #include <asm/iommu.h>
 #include <asm/gart.h>
 #include <asm/dma.h>
+#include <uapi/linux/iommufd.h>
 
 #include "amd_iommu.h"
 #include "../dma-iommu.h"
@@ -44,8 +45,6 @@
 
 #define CMD_SET_TYPE(cmd, t) ((cmd)->data[1] |= ((t) << 28))
 
-#define LOOP_TIMEOUT	100000
-
 /* IO virtual address start page frame number */
 #define IOVA_START_PFN		(1)
 #define IOVA_PFN(addr)		((addr) >> PAGE_SHIFT)
@@ -65,8 +64,8 @@ LIST_HEAD(hpet_map);
 LIST_HEAD(acpihid_map);
 
 const struct iommu_ops amd_iommu_ops;
+const struct iommu_dirty_ops amd_dirty_ops;
 
-static ATOMIC_NOTIFIER_HEAD(ppr_notifier);
 int amd_iommu_max_glx_val = -1;
 
 /*
@@ -79,7 +78,6 @@ struct iommu_cmd {
 struct kmem_cache *amd_iommu_irq_cache;
 
 static void detach_device(struct device *dev);
-static int domain_enable_v2(struct protection_domain *domain, int pasids);
 
 /****************************************************************************
  *
@@ -322,24 +320,141 @@ static struct iommu_group *acpihid_device_group(struct device *dev)
 	return entry->group;
 }
 
-static bool pci_iommuv2_capable(struct pci_dev *pdev)
+static inline bool pdev_pasid_supported(struct iommu_dev_data *dev_data)
 {
-	static const int caps[] = {
-		PCI_EXT_CAP_ID_PRI,
-		PCI_EXT_CAP_ID_PASID,
-	};
-	int i, pos;
+	return (dev_data->flags & AMD_IOMMU_DEVICE_FLAG_PASID_SUP);
+}
 
-	if (!pci_ats_supported(pdev))
-		return false;
+static u32 pdev_get_caps(struct pci_dev *pdev)
+{
+	int features;
+	u32 flags = 0;
+
+	if (pci_ats_supported(pdev))
+		flags |= AMD_IOMMU_DEVICE_FLAG_ATS_SUP;
 
-	for (i = 0; i < 2; ++i) {
-		pos = pci_find_ext_capability(pdev, caps[i]);
-		if (pos == 0)
-			return false;
+	if (pci_pri_supported(pdev))
+		flags |= AMD_IOMMU_DEVICE_FLAG_PRI_SUP;
+
+	features = pci_pasid_features(pdev);
+	if (features >= 0) {
+		flags |= AMD_IOMMU_DEVICE_FLAG_PASID_SUP;
+
+		if (features & PCI_PASID_CAP_EXEC)
+			flags |= AMD_IOMMU_DEVICE_FLAG_EXEC_SUP;
+
+		if (features & PCI_PASID_CAP_PRIV)
+			flags |= AMD_IOMMU_DEVICE_FLAG_PRIV_SUP;
 	}
 
-	return true;
+	return flags;
+}
+
+static inline int pdev_enable_cap_ats(struct pci_dev *pdev)
+{
+	struct iommu_dev_data *dev_data = dev_iommu_priv_get(&pdev->dev);
+	int ret = -EINVAL;
+
+	if (dev_data->ats_enabled)
+		return 0;
+
+	if (amd_iommu_iotlb_sup &&
+	    (dev_data->flags & AMD_IOMMU_DEVICE_FLAG_ATS_SUP)) {
+		ret = pci_enable_ats(pdev, PAGE_SHIFT);
+		if (!ret) {
+			dev_data->ats_enabled = 1;
+			dev_data->ats_qdep    = pci_ats_queue_depth(pdev);
+		}
+	}
+
+	return ret;
+}
+
+static inline void pdev_disable_cap_ats(struct pci_dev *pdev)
+{
+	struct iommu_dev_data *dev_data = dev_iommu_priv_get(&pdev->dev);
+
+	if (dev_data->ats_enabled) {
+		pci_disable_ats(pdev);
+		dev_data->ats_enabled = 0;
+	}
+}
+
+int amd_iommu_pdev_enable_cap_pri(struct pci_dev *pdev)
+{
+	struct iommu_dev_data *dev_data = dev_iommu_priv_get(&pdev->dev);
+	int ret = -EINVAL;
+
+	if (dev_data->pri_enabled)
+		return 0;
+
+	if (dev_data->flags & AMD_IOMMU_DEVICE_FLAG_PRI_SUP) {
+		/*
+		 * First reset the PRI state of the device.
+		 * FIXME: Hardcode number of outstanding requests for now
+		 */
+		if (!pci_reset_pri(pdev) && !pci_enable_pri(pdev, 32)) {
+			dev_data->pri_enabled = 1;
+			dev_data->pri_tlp     = pci_prg_resp_pasid_required(pdev);
+
+			ret = 0;
+		}
+	}
+
+	return ret;
+}
+
+void amd_iommu_pdev_disable_cap_pri(struct pci_dev *pdev)
+{
+	struct iommu_dev_data *dev_data = dev_iommu_priv_get(&pdev->dev);
+
+	if (dev_data->pri_enabled) {
+		pci_disable_pri(pdev);
+		dev_data->pri_enabled = 0;
+	}
+}
+
+static inline int pdev_enable_cap_pasid(struct pci_dev *pdev)
+{
+	struct iommu_dev_data *dev_data = dev_iommu_priv_get(&pdev->dev);
+	int ret = -EINVAL;
+
+	if (dev_data->pasid_enabled)
+		return 0;
+
+	if (dev_data->flags & AMD_IOMMU_DEVICE_FLAG_PASID_SUP) {
+		/* Only allow access to user-accessible pages */
+		ret = pci_enable_pasid(pdev, 0);
+		if (!ret)
+			dev_data->pasid_enabled = 1;
+	}
+
+	return ret;
+}
+
+static inline void pdev_disable_cap_pasid(struct pci_dev *pdev)
+{
+	struct iommu_dev_data *dev_data = dev_iommu_priv_get(&pdev->dev);
+
+	if (dev_data->pasid_enabled) {
+		pci_disable_pasid(pdev);
+		dev_data->pasid_enabled = 0;
+	}
+}
+
+static void pdev_enable_caps(struct pci_dev *pdev)
+{
+	pdev_enable_cap_ats(pdev);
+	pdev_enable_cap_pasid(pdev);
+	amd_iommu_pdev_enable_cap_pri(pdev);
+
+}
+
+static void pdev_disable_caps(struct pci_dev *pdev)
+{
+	pdev_disable_cap_ats(pdev);
+	pdev_disable_cap_pasid(pdev);
+	amd_iommu_pdev_disable_cap_pri(pdev);
 }
 
 /*
@@ -399,8 +514,8 @@ static int iommu_init_device(struct amd_iommu *iommu, struct device *dev)
 	 * it'll be forced to go into translation mode.
 	 */
 	if ((iommu_default_passthrough() || !amd_iommu_force_isolation) &&
-	    dev_is_pci(dev) && pci_iommuv2_capable(to_pci_dev(dev))) {
-		dev_data->iommu_v2 = iommu->is_iommu_v2;
+	    dev_is_pci(dev) && amd_iommu_gt_ppr_supported()) {
+		dev_data->flags = pdev_get_caps(to_pci_dev(dev));
 	}
 
 	dev_iommu_priv_set(dev, dev_data);
@@ -701,24 +816,6 @@ static void iommu_poll_events(struct amd_iommu *iommu)
 	writel(head, iommu->mmio_base + MMIO_EVT_HEAD_OFFSET);
 }
 
-static void iommu_handle_ppr_entry(struct amd_iommu *iommu, u64 *raw)
-{
-	struct amd_iommu_fault fault;
-
-	if (PPR_REQ_TYPE(raw[0]) != PPR_REQ_FAULT) {
-		pr_err_ratelimited("Unknown PPR request received\n");
-		return;
-	}
-
-	fault.address   = raw[1];
-	fault.pasid     = PPR_PASID(raw[0]);
-	fault.sbdf      = PCI_SEG_DEVID_TO_SBDF(iommu->pci_seg->id, PPR_DEVID(raw[0]));
-	fault.tag       = PPR_TAG(raw[0]);
-	fault.flags     = PPR_FLAGS(raw[0]);
-
-	atomic_notifier_call_chain(&ppr_notifier, 0, &fault);
-}
-
 static void iommu_poll_ppr_log(struct amd_iommu *iommu)
 {
 	u32 head, tail;
@@ -764,8 +861,7 @@ static void iommu_poll_ppr_log(struct amd_iommu *iommu)
 		head = (head + PPR_ENTRY_SIZE) % PPR_LOG_SIZE;
 		writel(head, iommu->mmio_base + MMIO_PPR_HEAD_OFFSET);
 
-		/* Handle PPR entry */
-		iommu_handle_ppr_entry(iommu, entry);
+		/* TODO: PPR Handler will be added when we add IOPF support */
 
 		/* Refresh ring-buffer information */
 		head = readl(iommu->mmio_base + MMIO_PPR_HEAD_OFFSET);
@@ -1094,7 +1190,7 @@ static void build_inv_iotlb_pasid(struct iommu_cmd *cmd, u16 devid, u32 pasid,
 }
 
 static void build_complete_ppr(struct iommu_cmd *cmd, u16 devid, u32 pasid,
-			       int status, int tag, bool gn)
+			       int status, int tag, u8 gn)
 {
 	memset(cmd, 0, sizeof(*cmd));
 
@@ -1298,7 +1394,7 @@ static void amd_iommu_flush_irt_all(struct amd_iommu *iommu)
 
 void iommu_flush_all_caches(struct amd_iommu *iommu)
 {
-	if (iommu_feature(iommu, FEATURE_IA)) {
+	if (check_feature(FEATURE_IA)) {
 		amd_iommu_flush_all(iommu);
 	} else {
 		amd_iommu_flush_dte_all(iommu);
@@ -1317,7 +1413,7 @@ static int device_flush_iotlb(struct iommu_dev_data *dev_data,
 	struct iommu_cmd cmd;
 	int qdep;
 
-	qdep     = dev_data->ats.qdep;
+	qdep     = dev_data->ats_qdep;
 	iommu    = rlookup_amd_iommu(dev_data->dev);
 	if (!iommu)
 		return -EINVAL;
@@ -1368,7 +1464,7 @@ static int device_flush_dte(struct iommu_dev_data *dev_data)
 			return ret;
 	}
 
-	if (dev_data->ats.enabled)
+	if (dev_data->ats_enabled)
 		ret = device_flush_iotlb(dev_data, 0, ~0UL);
 
 	return ret;
@@ -1401,7 +1497,7 @@ static void __domain_flush_pages(struct protection_domain *domain,
 
 	list_for_each_entry(dev_data, &domain->dev_list, list) {
 
-		if (!dev_data->ats.enabled)
+		if (!dev_data->ats_enabled)
 			continue;
 
 		ret |= device_flush_iotlb(dev_data, address, size);
@@ -1577,6 +1673,42 @@ static void free_gcr3_table(struct protection_domain *domain)
 	free_page((unsigned long)domain->gcr3_tbl);
 }
 
+/*
+ * Number of GCR3 table levels required. Level must be 4-Kbyte
+ * page and can contain up to 512 entries.
+ */
+static int get_gcr3_levels(int pasids)
+{
+	int levels;
+
+	if (pasids == -1)
+		return amd_iommu_max_glx_val;
+
+	levels = get_count_order(pasids);
+
+	return levels ? (DIV_ROUND_UP(levels, 9) - 1) : levels;
+}
+
+/* Note: This function expects iommu_domain->lock to be held prior calling the function. */
+static int setup_gcr3_table(struct protection_domain *domain, int pasids)
+{
+	int levels = get_gcr3_levels(pasids);
+
+	if (levels > amd_iommu_max_glx_val)
+		return -EINVAL;
+
+	domain->gcr3_tbl = alloc_pgtable_page(domain->nid, GFP_ATOMIC);
+	if (domain->gcr3_tbl == NULL)
+		return -ENOMEM;
+
+	domain->glx      = levels;
+	domain->flags   |= PD_IOMMUV2_MASK;
+
+	amd_iommu_domain_update(domain);
+
+	return 0;
+}
+
 static void set_dte_entry(struct amd_iommu *iommu, u16 devid,
 			  struct protection_domain *domain, bool ats, bool ppr)
 {
@@ -1605,10 +1737,11 @@ static void set_dte_entry(struct amd_iommu *iommu, u16 devid,
 	if (ats)
 		flags |= DTE_FLAG_IOTLB;
 
-	if (ppr) {
-		if (iommu_feature(iommu, FEATURE_EPHSUP))
-			pte_root |= 1ULL << DEV_ENTRY_PPR;
-	}
+	if (ppr)
+		pte_root |= 1ULL << DEV_ENTRY_PPR;
+
+	if (domain->dirty_tracking)
+		pte_root |= DTE_FLAG_HAD;
 
 	if (domain->flags & PD_IOMMUV2_MASK) {
 		u64 gcr3 = iommu_virt_to_phys(domain->gcr3_tbl);
@@ -1685,7 +1818,7 @@ static void do_attach(struct iommu_dev_data *dev_data,
 	iommu = rlookup_amd_iommu(dev_data->dev);
 	if (!iommu)
 		return;
-	ats   = dev_data->ats.enabled;
+	ats   = dev_data->ats_enabled;
 
 	/* Update data structures */
 	dev_data->domain = domain;
@@ -1701,7 +1834,7 @@ static void do_attach(struct iommu_dev_data *dev_data,
 
 	/* Update device table */
 	set_dte_entry(iommu, dev_data->devid, domain,
-		      ats, dev_data->iommu_v2);
+		      ats, dev_data->ppr);
 	clone_aliases(iommu, dev_data->dev);
 
 	device_flush_dte(dev_data);
@@ -1736,48 +1869,6 @@ static void do_detach(struct iommu_dev_data *dev_data)
 	domain->dev_cnt                 -= 1;
 }
 
-static void pdev_iommuv2_disable(struct pci_dev *pdev)
-{
-	pci_disable_ats(pdev);
-	pci_disable_pri(pdev);
-	pci_disable_pasid(pdev);
-}
-
-static int pdev_pri_ats_enable(struct pci_dev *pdev)
-{
-	int ret;
-
-	/* Only allow access to user-accessible pages */
-	ret = pci_enable_pasid(pdev, 0);
-	if (ret)
-		return ret;
-
-	/* First reset the PRI state of the device */
-	ret = pci_reset_pri(pdev);
-	if (ret)
-		goto out_err_pasid;
-
-	/* Enable PRI */
-	/* FIXME: Hardcode number of outstanding requests for now */
-	ret = pci_enable_pri(pdev, 32);
-	if (ret)
-		goto out_err_pasid;
-
-	ret = pci_enable_ats(pdev, PAGE_SHIFT);
-	if (ret)
-		goto out_err_pri;
-
-	return 0;
-
-out_err_pri:
-	pci_disable_pri(pdev);
-
-out_err_pasid:
-	pci_disable_pasid(pdev);
-
-	return ret;
-}
-
 /*
  * If a device is not yet associated with a domain, this function makes the
  * device visible in the domain
@@ -1786,9 +1877,8 @@ static int attach_device(struct device *dev,
 			 struct protection_domain *domain)
 {
 	struct iommu_dev_data *dev_data;
-	struct pci_dev *pdev;
 	unsigned long flags;
-	int ret;
+	int ret = 0;
 
 	spin_lock_irqsave(&domain->lock, flags);
 
@@ -1796,45 +1886,13 @@ static int attach_device(struct device *dev,
 
 	spin_lock(&dev_data->lock);
 
-	ret = -EBUSY;
-	if (dev_data->domain != NULL)
+	if (dev_data->domain != NULL) {
+		ret = -EBUSY;
 		goto out;
-
-	if (!dev_is_pci(dev))
-		goto skip_ats_check;
-
-	pdev = to_pci_dev(dev);
-	if (domain->flags & PD_IOMMUV2_MASK) {
-		struct iommu_domain *def_domain = iommu_get_dma_domain(dev);
-
-		ret = -EINVAL;
-
-		/*
-		 * In case of using AMD_IOMMU_V1 page table mode and the device
-		 * is enabling for PPR/ATS support (using v2 table),
-		 * we need to make sure that the domain type is identity map.
-		 */
-		if ((amd_iommu_pgtable == AMD_IOMMU_V1) &&
-		    def_domain->type != IOMMU_DOMAIN_IDENTITY) {
-			goto out;
-		}
-
-		if (dev_data->iommu_v2) {
-			if (pdev_pri_ats_enable(pdev) != 0)
-				goto out;
-
-			dev_data->ats.enabled = true;
-			dev_data->ats.qdep    = pci_ats_queue_depth(pdev);
-			dev_data->pri_tlp     = pci_prg_resp_pasid_required(pdev);
-		}
-	} else if (amd_iommu_iotlb_sup &&
-		   pci_enable_ats(pdev, PAGE_SHIFT) == 0) {
-		dev_data->ats.enabled = true;
-		dev_data->ats.qdep    = pci_ats_queue_depth(pdev);
 	}
 
-skip_ats_check:
-	ret = 0;
+	if (dev_is_pci(dev))
+		pdev_enable_caps(to_pci_dev(dev));
 
 	do_attach(dev_data, domain);
 
@@ -1882,15 +1940,8 @@ static void detach_device(struct device *dev)
 
 	do_detach(dev_data);
 
-	if (!dev_is_pci(dev))
-		goto out;
-
-	if (domain->flags & PD_IOMMUV2_MASK && dev_data->iommu_v2)
-		pdev_iommuv2_disable(to_pci_dev(dev));
-	else if (dev_data->ats.enabled)
-		pci_disable_ats(to_pci_dev(dev));
-
-	dev_data->ats.enabled = false;
+	if (dev_is_pci(dev))
+		pdev_disable_caps(to_pci_dev(dev));
 
 out:
 	spin_unlock(&dev_data->lock);
@@ -1980,7 +2031,7 @@ static void update_device_table(struct protection_domain *domain)
 		if (!iommu)
 			continue;
 		set_dte_entry(iommu, dev_data->devid, domain,
-			      dev_data->ats.enabled, dev_data->iommu_v2);
+			      dev_data->ats_enabled, dev_data->ppr);
 		clone_aliases(iommu, dev_data->dev);
 	}
 }
@@ -2014,9 +2065,11 @@ void amd_iommu_domain_update(struct protection_domain *domain)
 static void cleanup_domain(struct protection_domain *domain)
 {
 	struct iommu_dev_data *entry;
-	unsigned long flags;
 
-	spin_lock_irqsave(&domain->lock, flags);
+	lockdep_assert_held(&domain->lock);
+
+	if (!domain->dev_cnt)
+		return;
 
 	while (!list_empty(&domain->dev_list)) {
 		entry = list_first_entry(&domain->dev_list,
@@ -2024,8 +2077,7 @@ static void cleanup_domain(struct protection_domain *domain)
 		BUG_ON(!entry->domain);
 		do_detach(entry);
 	}
-
-	spin_unlock_irqrestore(&domain->lock, flags);
+	WARN_ON(domain->dev_cnt != 0);
 }
 
 static void protection_domain_free(struct protection_domain *domain)
@@ -2036,6 +2088,12 @@ static void protection_domain_free(struct protection_domain *domain)
 	if (domain->iop.pgtbl_cfg.tlb)
 		free_io_pgtable_ops(&domain->iop.iop.ops);
 
+	if (domain->flags & PD_IOMMUV2_MASK)
+		free_gcr3_table(domain);
+
+	if (domain->iop.root)
+		free_page((unsigned long)domain->iop.root);
+
 	if (domain->id)
 		domain_id_free(domain->id);
 
@@ -2048,18 +2106,10 @@ static int protection_domain_init_v1(struct protection_domain *domain, int mode)
 
 	BUG_ON(mode < PAGE_MODE_NONE || mode > PAGE_MODE_6_LEVEL);
 
-	spin_lock_init(&domain->lock);
-	domain->id = domain_id_alloc();
-	if (!domain->id)
-		return -ENOMEM;
-	INIT_LIST_HEAD(&domain->dev_list);
-
 	if (mode != PAGE_MODE_NONE) {
 		pt_root = (void *)get_zeroed_page(GFP_KERNEL);
-		if (!pt_root) {
-			domain_id_free(domain->id);
+		if (!pt_root)
 			return -ENOMEM;
-		}
 	}
 
 	amd_iommu_domain_set_pgtable(domain, pt_root, mode);
@@ -2069,20 +2119,12 @@ static int protection_domain_init_v1(struct protection_domain *domain, int mode)
 
 static int protection_domain_init_v2(struct protection_domain *domain)
 {
-	spin_lock_init(&domain->lock);
-	domain->id = domain_id_alloc();
-	if (!domain->id)
-		return -ENOMEM;
-	INIT_LIST_HEAD(&domain->dev_list);
-
 	domain->flags |= PD_GIOV_MASK;
 
 	domain->domain.pgsize_bitmap = AMD_IOMMU_PGSIZES_V2;
 
-	if (domain_enable_v2(domain, 1)) {
-		domain_id_free(domain->id);
+	if (setup_gcr3_table(domain, 1))
 		return -ENOMEM;
-	}
 
 	return 0;
 }
@@ -2092,57 +2134,60 @@ static struct protection_domain *protection_domain_alloc(unsigned int type)
 	struct io_pgtable_ops *pgtbl_ops;
 	struct protection_domain *domain;
 	int pgtable;
-	int mode = DEFAULT_PGTABLE_LEVEL;
 	int ret;
 
+	domain = kzalloc(sizeof(*domain), GFP_KERNEL);
+	if (!domain)
+		return NULL;
+
+	domain->id = domain_id_alloc();
+	if (!domain->id)
+		goto out_err;
+
+	spin_lock_init(&domain->lock);
+	INIT_LIST_HEAD(&domain->dev_list);
+	domain->nid = NUMA_NO_NODE;
+
+	switch (type) {
+	/* No need to allocate io pgtable ops in passthrough mode */
+	case IOMMU_DOMAIN_IDENTITY:
+		return domain;
+	case IOMMU_DOMAIN_DMA:
+		pgtable = amd_iommu_pgtable;
+		break;
 	/*
-	 * Force IOMMU v1 page table when iommu=pt and
-	 * when allocating domain for pass-through devices.
+	 * Force IOMMU v1 page table when allocating
+	 * domain for pass-through devices.
 	 */
-	if (type == IOMMU_DOMAIN_IDENTITY) {
-		pgtable = AMD_IOMMU_V1;
-		mode = PAGE_MODE_NONE;
-	} else if (type == IOMMU_DOMAIN_UNMANAGED) {
+	case IOMMU_DOMAIN_UNMANAGED:
 		pgtable = AMD_IOMMU_V1;
-	} else if (type == IOMMU_DOMAIN_DMA || type == IOMMU_DOMAIN_DMA_FQ) {
-		pgtable = amd_iommu_pgtable;
-	} else {
-		return NULL;
+		break;
+	default:
+		goto out_err;
 	}
 
-	domain = kzalloc(sizeof(*domain), GFP_KERNEL);
-	if (!domain)
-		return NULL;
-
 	switch (pgtable) {
 	case AMD_IOMMU_V1:
-		ret = protection_domain_init_v1(domain, mode);
+		ret = protection_domain_init_v1(domain, DEFAULT_PGTABLE_LEVEL);
 		break;
 	case AMD_IOMMU_V2:
 		ret = protection_domain_init_v2(domain);
 		break;
 	default:
 		ret = -EINVAL;
+		break;
 	}
 
 	if (ret)
 		goto out_err;
 
-	/* No need to allocate io pgtable ops in passthrough mode */
-	if (type == IOMMU_DOMAIN_IDENTITY)
-		return domain;
-
-	domain->nid = NUMA_NO_NODE;
-
 	pgtbl_ops = alloc_io_pgtable_ops(pgtable, &domain->iop.pgtbl_cfg, domain);
-	if (!pgtbl_ops) {
-		domain_id_free(domain->id);
+	if (!pgtbl_ops)
 		goto out_err;
-	}
 
 	return domain;
 out_err:
-	kfree(domain);
+	protection_domain_free(domain);
 	return NULL;
 }
 
@@ -2155,44 +2200,94 @@ static inline u64 dma_max_address(void)
 	return ((1ULL << PM_LEVEL_SHIFT(amd_iommu_gpt_level)) - 1);
 }
 
-static struct iommu_domain *amd_iommu_domain_alloc(unsigned type)
+static bool amd_iommu_hd_support(struct amd_iommu *iommu)
+{
+	return iommu && (iommu->features & FEATURE_HDSUP);
+}
+
+static struct iommu_domain *do_iommu_domain_alloc(unsigned int type,
+						  struct device *dev, u32 flags)
 {
+	bool dirty_tracking = flags & IOMMU_HWPT_ALLOC_DIRTY_TRACKING;
 	struct protection_domain *domain;
+	struct amd_iommu *iommu = NULL;
+
+	if (dev) {
+		iommu = rlookup_amd_iommu(dev);
+		if (!iommu)
+			return ERR_PTR(-ENODEV);
+	}
 
 	/*
 	 * Since DTE[Mode]=0 is prohibited on SNP-enabled system,
 	 * default to use IOMMU_DOMAIN_DMA[_FQ].
 	 */
 	if (amd_iommu_snp_en && (type == IOMMU_DOMAIN_IDENTITY))
-		return NULL;
+		return ERR_PTR(-EINVAL);
+
+	if (dirty_tracking && !amd_iommu_hd_support(iommu))
+		return ERR_PTR(-EOPNOTSUPP);
 
 	domain = protection_domain_alloc(type);
 	if (!domain)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	domain->domain.geometry.aperture_start = 0;
 	domain->domain.geometry.aperture_end   = dma_max_address();
 	domain->domain.geometry.force_aperture = true;
 
+	if (iommu) {
+		domain->domain.type = type;
+		domain->domain.pgsize_bitmap = iommu->iommu.ops->pgsize_bitmap;
+		domain->domain.ops = iommu->iommu.ops->default_domain_ops;
+
+		if (dirty_tracking)
+			domain->domain.dirty_ops = &amd_dirty_ops;
+	}
+
 	return &domain->domain;
 }
 
-static void amd_iommu_domain_free(struct iommu_domain *dom)
+static struct iommu_domain *amd_iommu_domain_alloc(unsigned int type)
 {
-	struct protection_domain *domain;
+	struct iommu_domain *domain;
 
-	domain = to_pdomain(dom);
+	domain = do_iommu_domain_alloc(type, NULL, 0);
+	if (IS_ERR(domain))
+		return NULL;
+
+	return domain;
+}
+
+static struct iommu_domain *
+amd_iommu_domain_alloc_user(struct device *dev, u32 flags,
+			    struct iommu_domain *parent,
+			    const struct iommu_user_data *user_data)
+
+{
+	unsigned int type = IOMMU_DOMAIN_UNMANAGED;
 
-	if (domain->dev_cnt > 0)
-		cleanup_domain(domain);
+	if ((flags & ~IOMMU_HWPT_ALLOC_DIRTY_TRACKING) || parent || user_data)
+		return ERR_PTR(-EOPNOTSUPP);
 
-	BUG_ON(domain->dev_cnt != 0);
+	return do_iommu_domain_alloc(type, dev, flags);
+}
+
+static void amd_iommu_domain_free(struct iommu_domain *dom)
+{
+	struct protection_domain *domain;
+	unsigned long flags;
 
 	if (!dom)
 		return;
 
-	if (domain->flags & PD_IOMMUV2_MASK)
-		free_gcr3_table(domain);
+	domain = to_pdomain(dom);
+
+	spin_lock_irqsave(&domain->lock, flags);
+
+	cleanup_domain(domain);
+
+	spin_unlock_irqrestore(&domain->lock, flags);
 
 	protection_domain_free(domain);
 }
@@ -2214,6 +2309,13 @@ static int amd_iommu_attach_device(struct iommu_domain *dom,
 
 	dev_data->defer_attach = false;
 
+	/*
+	 * Restrict to devices with compatible IOMMU hardware support
+	 * when enforcement of dirty tracking is enabled.
+	 */
+	if (dom->dirty_ops && !amd_iommu_hd_support(iommu))
+		return -EINVAL;
+
 	if (dev_data->domain)
 		detach_device(dev);
 
@@ -2233,14 +2335,15 @@ static int amd_iommu_attach_device(struct iommu_domain *dom,
 	return ret;
 }
 
-static void amd_iommu_iotlb_sync_map(struct iommu_domain *dom,
-				     unsigned long iova, size_t size)
+static int amd_iommu_iotlb_sync_map(struct iommu_domain *dom,
+				    unsigned long iova, size_t size)
 {
 	struct protection_domain *domain = to_pdomain(dom);
 	struct io_pgtable_ops *ops = &domain->iop.iop.ops;
 
 	if (ops->map_pages)
 		domain_flush_np_cache(domain, iova, size);
+	return 0;
 }
 
 static int amd_iommu_map_pages(struct iommu_domain *dom, unsigned long iova,
@@ -2332,6 +2435,11 @@ static bool amd_iommu_capable(struct device *dev, enum iommu_cap cap)
 		return true;
 	case IOMMU_CAP_DEFERRED_FLUSH:
 		return true;
+	case IOMMU_CAP_DIRTY_TRACKING: {
+		struct amd_iommu *iommu = rlookup_amd_iommu(dev);
+
+		return amd_iommu_hd_support(iommu);
+	}
 	default:
 		break;
 	}
@@ -2339,6 +2447,73 @@ static bool amd_iommu_capable(struct device *dev, enum iommu_cap cap)
 	return false;
 }
 
+static int amd_iommu_set_dirty_tracking(struct iommu_domain *domain,
+					bool enable)
+{
+	struct protection_domain *pdomain = to_pdomain(domain);
+	struct dev_table_entry *dev_table;
+	struct iommu_dev_data *dev_data;
+	bool domain_flush = false;
+	struct amd_iommu *iommu;
+	unsigned long flags;
+	u64 pte_root;
+
+	spin_lock_irqsave(&pdomain->lock, flags);
+	if (!(pdomain->dirty_tracking ^ enable)) {
+		spin_unlock_irqrestore(&pdomain->lock, flags);
+		return 0;
+	}
+
+	list_for_each_entry(dev_data, &pdomain->dev_list, list) {
+		iommu = rlookup_amd_iommu(dev_data->dev);
+		if (!iommu)
+			continue;
+
+		dev_table = get_dev_table(iommu);
+		pte_root = dev_table[dev_data->devid].data[0];
+
+		pte_root = (enable ? pte_root | DTE_FLAG_HAD :
+				     pte_root & ~DTE_FLAG_HAD);
+
+		/* Flush device DTE */
+		dev_table[dev_data->devid].data[0] = pte_root;
+		device_flush_dte(dev_data);
+		domain_flush = true;
+	}
+
+	/* Flush IOTLB to mark IOPTE dirty on the next translation(s) */
+	if (domain_flush) {
+		amd_iommu_domain_flush_tlb_pde(pdomain);
+		amd_iommu_domain_flush_complete(pdomain);
+	}
+	pdomain->dirty_tracking = enable;
+	spin_unlock_irqrestore(&pdomain->lock, flags);
+
+	return 0;
+}
+
+static int amd_iommu_read_and_clear_dirty(struct iommu_domain *domain,
+					  unsigned long iova, size_t size,
+					  unsigned long flags,
+					  struct iommu_dirty_bitmap *dirty)
+{
+	struct protection_domain *pdomain = to_pdomain(domain);
+	struct io_pgtable_ops *ops = &pdomain->iop.iop.ops;
+	unsigned long lflags;
+
+	if (!ops || !ops->read_and_clear_dirty)
+		return -EOPNOTSUPP;
+
+	spin_lock_irqsave(&pdomain->lock, lflags);
+	if (!pdomain->dirty_tracking && dirty->bitmap) {
+		spin_unlock_irqrestore(&pdomain->lock, lflags);
+		return -EINVAL;
+	}
+	spin_unlock_irqrestore(&pdomain->lock, lflags);
+
+	return ops->read_and_clear_dirty(ops, iova, size, flags, dirty);
+}
+
 static void amd_iommu_get_resv_regions(struct device *dev,
 				       struct list_head *head)
 {
@@ -2406,7 +2581,6 @@ bool amd_iommu_is_attach_deferred(struct device *dev)
 
 	return dev_data->defer_attach;
 }
-EXPORT_SYMBOL_GPL(amd_iommu_is_attach_deferred);
 
 static void amd_iommu_flush_iotlb_all(struct iommu_domain *domain)
 {
@@ -2446,7 +2620,7 @@ static int amd_iommu_def_domain_type(struct device *dev)
 	 *    and require remapping.
 	 *  - SNP is enabled, because it prohibits DTE[Mode]=0.
 	 */
-	if (dev_data->iommu_v2 &&
+	if (pdev_pasid_supported(dev_data) &&
 	    !cc_platform_has(CC_ATTR_MEM_ENCRYPT) &&
 	    !amd_iommu_snp_en) {
 		return IOMMU_DOMAIN_IDENTITY;
@@ -2461,9 +2635,15 @@ static bool amd_iommu_enforce_cache_coherency(struct iommu_domain *domain)
 	return true;
 }
 
+const struct iommu_dirty_ops amd_dirty_ops = {
+	.set_dirty_tracking = amd_iommu_set_dirty_tracking,
+	.read_and_clear_dirty = amd_iommu_read_and_clear_dirty,
+};
+
 const struct iommu_ops amd_iommu_ops = {
 	.capable = amd_iommu_capable,
 	.domain_alloc = amd_iommu_domain_alloc,
+	.domain_alloc_user = amd_iommu_domain_alloc_user,
 	.probe_device = amd_iommu_probe_device,
 	.release_device = amd_iommu_release_device,
 	.probe_finalize = amd_iommu_probe_finalize,
@@ -2485,93 +2665,6 @@ const struct iommu_ops amd_iommu_ops = {
 	}
 };
 
-/*****************************************************************************
- *
- * The next functions do a basic initialization of IOMMU for pass through
- * mode
- *
- * In passthrough mode the IOMMU is initialized and enabled but not used for
- * DMA-API translation.
- *
- *****************************************************************************/
-
-/* IOMMUv2 specific functions */
-int amd_iommu_register_ppr_notifier(struct notifier_block *nb)
-{
-	return atomic_notifier_chain_register(&ppr_notifier, nb);
-}
-EXPORT_SYMBOL(amd_iommu_register_ppr_notifier);
-
-int amd_iommu_unregister_ppr_notifier(struct notifier_block *nb)
-{
-	return atomic_notifier_chain_unregister(&ppr_notifier, nb);
-}
-EXPORT_SYMBOL(amd_iommu_unregister_ppr_notifier);
-
-void amd_iommu_domain_direct_map(struct iommu_domain *dom)
-{
-	struct protection_domain *domain = to_pdomain(dom);
-	unsigned long flags;
-
-	spin_lock_irqsave(&domain->lock, flags);
-
-	if (domain->iop.pgtbl_cfg.tlb)
-		free_io_pgtable_ops(&domain->iop.iop.ops);
-
-	spin_unlock_irqrestore(&domain->lock, flags);
-}
-EXPORT_SYMBOL(amd_iommu_domain_direct_map);
-
-/* Note: This function expects iommu_domain->lock to be held prior calling the function. */
-static int domain_enable_v2(struct protection_domain *domain, int pasids)
-{
-	int levels;
-
-	/* Number of GCR3 table levels required */
-	for (levels = 0; (pasids - 1) & ~0x1ff; pasids >>= 9)
-		levels += 1;
-
-	if (levels > amd_iommu_max_glx_val)
-		return -EINVAL;
-
-	domain->gcr3_tbl = (void *)get_zeroed_page(GFP_ATOMIC);
-	if (domain->gcr3_tbl == NULL)
-		return -ENOMEM;
-
-	domain->glx      = levels;
-	domain->flags   |= PD_IOMMUV2_MASK;
-
-	amd_iommu_domain_update(domain);
-
-	return 0;
-}
-
-int amd_iommu_domain_enable_v2(struct iommu_domain *dom, int pasids)
-{
-	struct protection_domain *pdom = to_pdomain(dom);
-	unsigned long flags;
-	int ret;
-
-	spin_lock_irqsave(&pdom->lock, flags);
-
-	/*
-	 * Save us all sanity checks whether devices already in the
-	 * domain support IOMMUv2. Just force that the domain has no
-	 * devices attached when it is switched into IOMMUv2 mode.
-	 */
-	ret = -EBUSY;
-	if (pdom->dev_cnt > 0 || pdom->flags & PD_IOMMUV2_MASK)
-		goto out;
-
-	if (!pdom->gcr3_tbl)
-		ret = domain_enable_v2(pdom, pasids);
-
-out:
-	spin_unlock_irqrestore(&pdom->lock, flags);
-	return ret;
-}
-EXPORT_SYMBOL(amd_iommu_domain_enable_v2);
-
 static int __flush_pasid(struct protection_domain *domain, u32 pasid,
 			 u64 address, bool size)
 {
@@ -2609,10 +2702,10 @@ static int __flush_pasid(struct protection_domain *domain, u32 pasid,
 		   There might be non-IOMMUv2 capable devices in an IOMMUv2
 		 * domain.
 		 */
-		if (!dev_data->ats.enabled)
+		if (!dev_data->ats_enabled)
 			continue;
 
-		qdep  = dev_data->ats.qdep;
+		qdep  = dev_data->ats_qdep;
 		iommu = rlookup_amd_iommu(dev_data->dev);
 		if (!iommu)
 			continue;
@@ -2653,7 +2746,6 @@ int amd_iommu_flush_page(struct iommu_domain *dom, u32 pasid,
 
 	return ret;
 }
-EXPORT_SYMBOL(amd_iommu_flush_page);
 
 static int __amd_iommu_flush_tlb(struct protection_domain *domain, u32 pasid)
 {
@@ -2673,7 +2765,6 @@ int amd_iommu_flush_tlb(struct iommu_domain *dom, u32 pasid)
 
 	return ret;
 }
-EXPORT_SYMBOL(amd_iommu_flush_tlb);
 
 static u64 *__get_gcr3_pte(u64 *root, int level, u32 pasid, bool alloc)
 {
@@ -2753,7 +2844,6 @@ int amd_iommu_domain_set_gcr3(struct iommu_domain *dom, u32 pasid,
 
 	return ret;
 }
-EXPORT_SYMBOL(amd_iommu_domain_set_gcr3);
 
 int amd_iommu_domain_clear_gcr3(struct iommu_domain *dom, u32 pasid)
 {
@@ -2767,7 +2857,6 @@ int amd_iommu_domain_clear_gcr3(struct iommu_domain *dom, u32 pasid)
 
 	return ret;
 }
-EXPORT_SYMBOL(amd_iommu_domain_clear_gcr3);
 
 int amd_iommu_complete_ppr(struct pci_dev *pdev, u32 pasid,
 			   int status, int tag)
@@ -2786,49 +2875,6 @@ int amd_iommu_complete_ppr(struct pci_dev *pdev, u32 pasid,
 
 	return iommu_queue_command(iommu, &cmd);
 }
-EXPORT_SYMBOL(amd_iommu_complete_ppr);
-
-int amd_iommu_device_info(struct pci_dev *pdev,
-                          struct amd_iommu_device_info *info)
-{
-	int max_pasids;
-	int pos;
-
-	if (pdev == NULL || info == NULL)
-		return -EINVAL;
-
-	if (!amd_iommu_v2_supported())
-		return -EINVAL;
-
-	memset(info, 0, sizeof(*info));
-
-	if (pci_ats_supported(pdev))
-		info->flags |= AMD_IOMMU_DEVICE_FLAG_ATS_SUP;
-
-	pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_PRI);
-	if (pos)
-		info->flags |= AMD_IOMMU_DEVICE_FLAG_PRI_SUP;
-
-	pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_PASID);
-	if (pos) {
-		int features;
-
-		max_pasids = 1 << (9 * (amd_iommu_max_glx_val + 1));
-		max_pasids = min(max_pasids, (1 << 20));
-
-		info->flags |= AMD_IOMMU_DEVICE_FLAG_PASID_SUP;
-		info->max_pasids = min(pci_max_pasids(pdev), max_pasids);
-
-		features = pci_pasid_features(pdev);
-		if (features & PCI_PASID_CAP_EXEC)
-			info->flags |= AMD_IOMMU_DEVICE_FLAG_EXEC_SUP;
-		if (features & PCI_PASID_CAP_PRIV)
-			info->flags |= AMD_IOMMU_DEVICE_FLAG_PRIV_SUP;
-	}
-
-	return 0;
-}
-EXPORT_SYMBOL(amd_iommu_device_info);
 
 #ifdef CONFIG_IRQ_REMAP