1 files changed, 100 insertions, 191 deletions

diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
index 687f18f65cea..6784a05dd6b2 100644
--- a/drivers/iommu/intel-iommu.c
+++ b/drivers/iommu/intel-iommu.c
@@ -458,31 +458,6 @@ static LIST_HEAD(dmar_rmrr_units);
 #define for_each_rmrr_units(rmrr) \
 	list_for_each_entry(rmrr, &dmar_rmrr_units, list)
 
-static void flush_unmaps_timeout(unsigned long data);
-
-struct deferred_flush_entry {
-	unsigned long iova_pfn;
-	unsigned long nrpages;
-	struct dmar_domain *domain;
-	struct page *freelist;
-};
-
-#define HIGH_WATER_MARK 250
-struct deferred_flush_table {
-	int next;
-	struct deferred_flush_entry entries[HIGH_WATER_MARK];
-};
-
-struct deferred_flush_data {
-	spinlock_t lock;
-	int timer_on;
-	struct timer_list timer;
-	long size;
-	struct deferred_flush_table *tables;
-};
-
-static DEFINE_PER_CPU(struct deferred_flush_data, deferred_flush);
-
 /* bitmap for indexing intel_iommus */
 static int g_num_of_iommus;
 
@@ -901,6 +876,13 @@ static struct intel_iommu *device_to_iommu(struct device *dev, u8 *bus, u8 *devf
 		struct pci_dev *pf_pdev;
 
 		pdev = to_pci_dev(dev);
+
+#ifdef CONFIG_X86
+		/* VMD child devices currently cannot be handled individually */
+		if (is_vmd(pdev->bus))
+			return NULL;
+#endif
+
 		/* VFs aren't listed in scope tables; we need to look up
 		 * the PF instead to find the IOMMU. */
 		pf_pdev = pci_physfn(pdev);
@@ -974,20 +956,6 @@ static int device_context_mapped(struct intel_iommu *iommu, u8 bus, u8 devfn)
 	return ret;
 }
 
-static void clear_context_table(struct intel_iommu *iommu, u8 bus, u8 devfn)
-{
-	struct context_entry *context;
-	unsigned long flags;
-
-	spin_lock_irqsave(&iommu->lock, flags);
-	context = iommu_context_addr(iommu, bus, devfn, 0);
-	if (context) {
-		context_clear_entry(context);
-		__iommu_flush_cache(iommu, context, sizeof(*context));
-	}
-	spin_unlock_irqrestore(&iommu->lock, flags);
-}
-
 static void free_context_table(struct intel_iommu *iommu)
 {
 	int i;
@@ -1137,8 +1105,9 @@ static void dma_pte_clear_range(struct dmar_domain *domain,
 }
 
 static void dma_pte_free_level(struct dmar_domain *domain, int level,
-			       struct dma_pte *pte, unsigned long pfn,
-			       unsigned long start_pfn, unsigned long last_pfn)
+			       int retain_level, struct dma_pte *pte,
+			       unsigned long pfn, unsigned long start_pfn,
+			       unsigned long last_pfn)
 {
 	pfn = max(start_pfn, pfn);
 	pte = &pte[pfn_level_offset(pfn, level)];
@@ -1153,12 +1122,17 @@ static void dma_pte_free_level(struct dmar_domain *domain, int level,
 		level_pfn = pfn & level_mask(level);
 		level_pte = phys_to_virt(dma_pte_addr(pte));
 
-		if (level > 2)
-			dma_pte_free_level(domain, level - 1, level_pte,
-					   level_pfn, start_pfn, last_pfn);
+		if (level > 2) {
+			dma_pte_free_level(domain, level - 1, retain_level,
+					   level_pte, level_pfn, start_pfn,
+					   last_pfn);
+		}
 
-		/* If range covers entire pagetable, free it */
-		if (!(start_pfn > level_pfn ||
+		/*
+		 * Free the page table if we're below the level we want to
+		 * retain and the range covers the entire table.
+		 */
+		if (level < retain_level && !(start_pfn > level_pfn ||
 		      last_pfn < level_pfn + level_size(level) - 1)) {
 			dma_clear_pte(pte);
 			domain_flush_cache(domain, pte, sizeof(*pte));
@@ -1169,10 +1143,14 @@ next:
 	} while (!first_pte_in_page(++pte) && pfn <= last_pfn);
 }
 
-/* clear last level (leaf) ptes and free page table pages. */
+/*
+ * clear last level (leaf) ptes and free page table pages below the
+ * level we wish to keep intact.
+ */
 static void dma_pte_free_pagetable(struct dmar_domain *domain,
 				   unsigned long start_pfn,
-				   unsigned long last_pfn)
+				   unsigned long last_pfn,
+				   int retain_level)
 {
 	BUG_ON(!domain_pfn_supported(domain, start_pfn));
 	BUG_ON(!domain_pfn_supported(domain, last_pfn));
@@ -1181,7 +1159,7 @@ static void dma_pte_free_pagetable(struct dmar_domain *domain,
 	dma_pte_clear_range(domain, start_pfn, last_pfn);
 
 	/* We don't need lock here; nobody else touches the iova range */
-	dma_pte_free_level(domain, agaw_to_level(domain->agaw),
+	dma_pte_free_level(domain, agaw_to_level(domain->agaw), retain_level,
 			   domain->pgd, 0, start_pfn, last_pfn);
 
 	/* free pgd */
@@ -1309,6 +1287,13 @@ static void dma_free_pagelist(struct page *freelist)
 	}
 }
 
+static void iova_entry_free(unsigned long data)
+{
+	struct page *freelist = (struct page *)data;
+
+	dma_free_pagelist(freelist);
+}
+
 /* iommu handling */
 static int iommu_alloc_root_entry(struct intel_iommu *iommu)
 {
@@ -1622,6 +1607,25 @@ static void iommu_flush_iotlb_psi(struct intel_iommu *iommu,
 				      addr, mask);
 }
 
+static void iommu_flush_iova(struct iova_domain *iovad)
+{
+	struct dmar_domain *domain;
+	int idx;
+
+	domain = container_of(iovad, struct dmar_domain, iovad);
+
+	for_each_domain_iommu(idx, domain) {
+		struct intel_iommu *iommu = g_iommus[idx];
+		u16 did = domain->iommu_did[iommu->seq_id];
+
+		iommu->flush.flush_iotlb(iommu, did, 0, 0, DMA_TLB_DSI_FLUSH);
+
+		if (!cap_caching_mode(iommu->cap))
+			iommu_flush_dev_iotlb(get_iommu_domain(iommu, did),
+					      0, MAX_AGAW_PFN_WIDTH);
+	}
+}
+
 static void iommu_disable_protect_mem_regions(struct intel_iommu *iommu)
 {
 	u32 pmen;
@@ -1932,9 +1936,16 @@ static int domain_init(struct dmar_domain *domain, struct intel_iommu *iommu,
 {
 	int adjust_width, agaw;
 	unsigned long sagaw;
+	int err;
 
 	init_iova_domain(&domain->iovad, VTD_PAGE_SIZE, IOVA_START_PFN,
 			DMA_32BIT_PFN);
+
+	err = init_iova_flush_queue(&domain->iovad,
+				    iommu_flush_iova, iova_entry_free);
+	if (err)
+		return err;
+
 	domain_reserve_special_ranges(domain);
 
 	/* calculate AGAW */
@@ -1986,14 +1997,6 @@ static void domain_exit(struct dmar_domain *domain)
 	if (!domain)
 		return;
 
-	/* Flush any lazy unmaps that may reference this domain */
-	if (!intel_iommu_strict) {
-		int cpu;
-
-		for_each_possible_cpu(cpu)
-			flush_unmaps_timeout(cpu);
-	}
-
 	/* Remove associated devices and clear attached or cached domains */
 	rcu_read_lock();
 	domain_remove_dev_info(domain);
@@ -2277,8 +2280,11 @@ static int __domain_mapping(struct dmar_domain *domain, unsigned long iov_pfn,
 				/*
 				 * Ensure that old small page tables are
 				 * removed to make room for superpage(s).
+				 * We're adding new large pages, so make sure
+				 * we don't remove their parent tables.
 				 */
-				dma_pte_free_pagetable(domain, iov_pfn, end_pfn);
+				dma_pte_free_pagetable(domain, iov_pfn, end_pfn,
+						       largepage_lvl + 1);
 			} else {
 				pteval &= ~(uint64_t)DMA_PTE_LARGE_PAGE;
 			}
@@ -2351,13 +2357,33 @@ static inline int domain_pfn_mapping(struct dmar_domain *domain, unsigned long i
 
 static void domain_context_clear_one(struct intel_iommu *iommu, u8 bus, u8 devfn)
 {
+	unsigned long flags;
+	struct context_entry *context;
+	u16 did_old;
+
 	if (!iommu)
 		return;
 
-	clear_context_table(iommu, bus, devfn);
-	iommu->flush.flush_context(iommu, 0, 0, 0,
-					   DMA_CCMD_GLOBAL_INVL);
-	iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH);
+	spin_lock_irqsave(&iommu->lock, flags);
+	context = iommu_context_addr(iommu, bus, devfn, 0);
+	if (!context) {
+		spin_unlock_irqrestore(&iommu->lock, flags);
+		return;
+	}
+	did_old = context_domain_id(context);
+	context_clear_entry(context);
+	__iommu_flush_cache(iommu, context, sizeof(*context));
+	spin_unlock_irqrestore(&iommu->lock, flags);
+	iommu->flush.flush_context(iommu,
+				   did_old,
+				   (((u16)bus) << 8) | devfn,
+				   DMA_CCMD_MASK_NOBIT,
+				   DMA_CCMD_DEVICE_INVL);
+	iommu->flush.flush_iotlb(iommu,
+				 did_old,
+				 0,
+				 0,
+				 DMA_TLB_DSI_FLUSH);
 }
 
 static inline void unlink_domain_info(struct device_domain_info *info)
@@ -3206,7 +3232,7 @@ static int __init init_dmars(void)
 	bool copied_tables = false;
 	struct device *dev;
 	struct intel_iommu *iommu;
-	int i, ret, cpu;
+	int i, ret;
 
 	/*
 	 * for each drhd
@@ -3239,22 +3265,6 @@ static int __init init_dmars(void)
 		goto error;
 	}
 
-	for_each_possible_cpu(cpu) {
-		struct deferred_flush_data *dfd = per_cpu_ptr(&deferred_flush,
-							      cpu);
-
-		dfd->tables = kzalloc(g_num_of_iommus *
-				      sizeof(struct deferred_flush_table),
-				      GFP_KERNEL);
-		if (!dfd->tables) {
-			ret = -ENOMEM;
-			goto free_g_iommus;
-		}
-
-		spin_lock_init(&dfd->lock);
-		setup_timer(&dfd->timer, flush_unmaps_timeout, cpu);
-	}
-
 	for_each_active_iommu(iommu, drhd) {
 		g_iommus[iommu->seq_id] = iommu;
 
@@ -3437,10 +3447,9 @@ free_iommu:
 		disable_dmar_iommu(iommu);
 		free_dmar_iommu(iommu);
 	}
-free_g_iommus:
-	for_each_possible_cpu(cpu)
-		kfree(per_cpu_ptr(&deferred_flush, cpu)->tables);
+
 	kfree(g_iommus);
+
 error:
 	return ret;
 }
@@ -3645,110 +3654,6 @@ static dma_addr_t intel_map_page(struct device *dev, struct page *page,
 				  dir, *dev->dma_mask);
 }
 
-static void flush_unmaps(struct deferred_flush_data *flush_data)
-{
-	int i, j;
-
-	flush_data->timer_on = 0;
-
-	/* just flush them all */
-	for (i = 0; i < g_num_of_iommus; i++) {
-		struct intel_iommu *iommu = g_iommus[i];
-		struct deferred_flush_table *flush_table =
-				&flush_data->tables[i];
-		if (!iommu)
-			continue;
-
-		if (!flush_table->next)
-			continue;
-
-		/* In caching mode, global flushes turn emulation expensive */
-		if (!cap_caching_mode(iommu->cap))
-			iommu->flush.flush_iotlb(iommu, 0, 0, 0,
-					 DMA_TLB_GLOBAL_FLUSH);
-		for (j = 0; j < flush_table->next; j++) {
-			unsigned long mask;
-			struct deferred_flush_entry *entry =
-						&flush_table->entries[j];
-			unsigned long iova_pfn = entry->iova_pfn;
-			unsigned long nrpages = entry->nrpages;
-			struct dmar_domain *domain = entry->domain;
-			struct page *freelist = entry->freelist;
-
-			/* On real hardware multiple invalidations are expensive */
-			if (cap_caching_mode(iommu->cap))
-				iommu_flush_iotlb_psi(iommu, domain,
-					mm_to_dma_pfn(iova_pfn),
-					nrpages, !freelist, 0);
-			else {
-				mask = ilog2(nrpages);
-				iommu_flush_dev_iotlb(domain,
-						(uint64_t)iova_pfn << PAGE_SHIFT, mask);
-			}
-			free_iova_fast(&domain->iovad, iova_pfn, nrpages);
-			if (freelist)
-				dma_free_pagelist(freelist);
-		}
-		flush_table->next = 0;
-	}
-
-	flush_data->size = 0;
-}
-
-static void flush_unmaps_timeout(unsigned long cpuid)
-{
-	struct deferred_flush_data *flush_data = per_cpu_ptr(&deferred_flush, cpuid);
-	unsigned long flags;
-
-	spin_lock_irqsave(&flush_data->lock, flags);
-	flush_unmaps(flush_data);
-	spin_unlock_irqrestore(&flush_data->lock, flags);
-}
-
-static void add_unmap(struct dmar_domain *dom, unsigned long iova_pfn,
-		      unsigned long nrpages, struct page *freelist)
-{
-	unsigned long flags;
-	int entry_id, iommu_id;
-	struct intel_iommu *iommu;
-	struct deferred_flush_entry *entry;
-	struct deferred_flush_data *flush_data;
-
-	flush_data = raw_cpu_ptr(&deferred_flush);
-
-	/* Flush all CPUs' entries to avoid deferring too much.  If
-	 * this becomes a bottleneck, can just flush us, and rely on
-	 * flush timer for the rest.
-	 */
-	if (flush_data->size == HIGH_WATER_MARK) {
-		int cpu;
-
-		for_each_online_cpu(cpu)
-			flush_unmaps_timeout(cpu);
-	}
-
-	spin_lock_irqsave(&flush_data->lock, flags);
-
-	iommu = domain_get_iommu(dom);
-	iommu_id = iommu->seq_id;
-
-	entry_id = flush_data->tables[iommu_id].next;
-	++(flush_data->tables[iommu_id].next);
-
-	entry = &flush_data->tables[iommu_id].entries[entry_id];
-	entry->domain = dom;
-	entry->iova_pfn = iova_pfn;
-	entry->nrpages = nrpages;
-	entry->freelist = freelist;
-
-	if (!flush_data->timer_on) {
-		mod_timer(&flush_data->timer, jiffies + msecs_to_jiffies(10));
-		flush_data->timer_on = 1;
-	}
-	flush_data->size++;
-	spin_unlock_irqrestore(&flush_data->lock, flags);
-}
-
 static void intel_unmap(struct device *dev, dma_addr_t dev_addr, size_t size)
 {
 	struct dmar_domain *domain;
@@ -3784,7 +3689,8 @@ static void intel_unmap(struct device *dev, dma_addr_t dev_addr, size_t size)
 		free_iova_fast(&domain->iovad, iova_pfn, dma_to_mm_pfn(nrpages));
 		dma_free_pagelist(freelist);
 	} else {
-		add_unmap(domain, iova_pfn, nrpages, freelist);
+		queue_iova(&domain->iovad, iova_pfn, nrpages,
+			   (unsigned long)freelist);
 		/*
 		 * queue up the release of the unmap to save the 1/6th of the
 		 * cpu used up by the iotlb flush operation...
@@ -3938,7 +3844,8 @@ static int intel_map_sg(struct device *dev, struct scatterlist *sglist, int nele
 	ret = domain_sg_mapping(domain, start_vpfn, sglist, size, prot);
 	if (unlikely(ret)) {
 		dma_pte_free_pagetable(domain, start_vpfn,
-				       start_vpfn + size - 1);
+				       start_vpfn + size - 1,
+				       agaw_to_level(domain->agaw) + 1);
 		free_iova_fast(&domain->iovad, iova_pfn, dma_to_mm_pfn(size));
 		return 0;
 	}
@@ -4721,7 +4628,6 @@ static void free_all_cpu_cached_iovas(unsigned int cpu)
 static int intel_iommu_cpu_dead(unsigned int cpu)
 {
 	free_all_cpu_cached_iovas(cpu);
-	flush_unmaps_timeout(cpu);
 	return 0;
 }
 
@@ -4736,7 +4642,9 @@ static void intel_disable_iommus(void)
 
 static inline struct intel_iommu *dev_to_intel_iommu(struct device *dev)
 {
-	return container_of(dev, struct intel_iommu, iommu.dev);
+	struct iommu_device *iommu_dev = dev_to_iommu_device(dev);
+
+	return container_of(iommu_dev, struct intel_iommu, iommu);
 }
 
 static ssize_t intel_iommu_show_version(struct device *dev,
@@ -5341,7 +5249,8 @@ int intel_iommu_enable_pasid(struct intel_iommu *iommu, struct intel_svm_dev *sd
 	sdev->sid = PCI_DEVID(info->bus, info->devfn);
 
 	if (!(ctx_lo & CONTEXT_PASIDE)) {
-		context[1].hi = (u64)virt_to_phys(iommu->pasid_state_table);
+		if (iommu->pasid_state_table)
+			context[1].hi = (u64)virt_to_phys(iommu->pasid_state_table);
 		context[1].lo = (u64)virt_to_phys(iommu->pasid_table) |
 			intel_iommu_get_pts(iommu);