Merge branch 'core' into arm/tegra

14 files changed, 429 insertions, 499 deletions

diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index 354cbd6392cd..5a61cf5a4443 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -137,20 +137,7 @@ struct kmem_cache *amd_iommu_irq_cache;
 static void update_domain(struct protection_domain *domain);
 static int protection_domain_init(struct protection_domain *domain);
 static void detach_device(struct device *dev);
-
-#define FLUSH_QUEUE_SIZE 256
-
-struct flush_queue_entry {
-	unsigned long iova_pfn;
-	unsigned long pages;
-	u64 counter; /* Flush counter when this entry was added to the queue */
-};
-
-struct flush_queue {
-	struct flush_queue_entry *entries;
-	unsigned head, tail;
-	spinlock_t lock;
-};
+static void iova_domain_flush_tlb(struct iova_domain *iovad);
 
 /*
  * Data container for a dma_ops specific protection domain
@@ -161,36 +148,6 @@ struct dma_ops_domain {
 
 	/* IOVA RB-Tree */
 	struct iova_domain iovad;
-
-	struct flush_queue __percpu *flush_queue;
-
-	/*
-	 * We need two counter here to be race-free wrt. IOTLB flushing and
-	 * adding entries to the flush queue.
-	 *
-	 * The flush_start_cnt is incremented _before_ the IOTLB flush starts.
-	 * New entries added to the flush ring-buffer get their 'counter' value
-	 * from here. This way we can make sure that entries added to the queue
-	 * (or other per-cpu queues of the same domain) while the TLB is about
-	 * to be flushed are not considered to be flushed already.
-	 */
-	atomic64_t flush_start_cnt;
-
-	/*
-	 * The flush_finish_cnt is incremented when an IOTLB flush is complete.
-	 * This value is always smaller than flush_start_cnt. The queue_add
-	 * function frees all IOVAs that have a counter value smaller than
-	 * flush_finish_cnt. This makes sure that we only free IOVAs that are
-	 * flushed out of the IOTLB of the domain.
-	 */
-	atomic64_t flush_finish_cnt;
-
-	/*
-	 * Timer to make sure we don't keep IOVAs around unflushed
-	 * for too long
-	 */
-	struct timer_list flush_timer;
-	atomic_t flush_timer_on;
 };
 
 static struct iova_domain reserved_iova_ranges;
@@ -1788,178 +1745,19 @@ static void free_gcr3_table(struct protection_domain *domain)
 	free_page((unsigned long)domain->gcr3_tbl);
 }
 
-static void dma_ops_domain_free_flush_queue(struct dma_ops_domain *dom)
-{
-	int cpu;
-
-	for_each_possible_cpu(cpu) {
-		struct flush_queue *queue;
-
-		queue = per_cpu_ptr(dom->flush_queue, cpu);
-		kfree(queue->entries);
-	}
-
-	free_percpu(dom->flush_queue);
-
-	dom->flush_queue = NULL;
-}
-
-static int dma_ops_domain_alloc_flush_queue(struct dma_ops_domain *dom)
-{
-	int cpu;
-
-	atomic64_set(&dom->flush_start_cnt,  0);
-	atomic64_set(&dom->flush_finish_cnt, 0);
-
-	dom->flush_queue = alloc_percpu(struct flush_queue);
-	if (!dom->flush_queue)
-		return -ENOMEM;
-
-	/* First make sure everything is cleared */
-	for_each_possible_cpu(cpu) {
-		struct flush_queue *queue;
-
-		queue = per_cpu_ptr(dom->flush_queue, cpu);
-		queue->head    = 0;
-		queue->tail    = 0;
-		queue->entries = NULL;
-	}
-
-	/* Now start doing the allocation */
-	for_each_possible_cpu(cpu) {
-		struct flush_queue *queue;
-
-		queue = per_cpu_ptr(dom->flush_queue, cpu);
-		queue->entries = kzalloc(FLUSH_QUEUE_SIZE * sizeof(*queue->entries),
-					 GFP_KERNEL);
-		if (!queue->entries) {
-			dma_ops_domain_free_flush_queue(dom);
-			return -ENOMEM;
-		}
-
-		spin_lock_init(&queue->lock);
-	}
-
-	return 0;
-}
-
 static void dma_ops_domain_flush_tlb(struct dma_ops_domain *dom)
 {
-	atomic64_inc(&dom->flush_start_cnt);
 	domain_flush_tlb(&dom->domain);
 	domain_flush_complete(&dom->domain);
-	atomic64_inc(&dom->flush_finish_cnt);
 }
 
-static inline bool queue_ring_full(struct flush_queue *queue)
+static void iova_domain_flush_tlb(struct iova_domain *iovad)
 {
-	assert_spin_locked(&queue->lock);
-
-	return (((queue->tail + 1) % FLUSH_QUEUE_SIZE) == queue->head);
-}
+	struct dma_ops_domain *dom;
 
-#define queue_ring_for_each(i, q) \
-	for (i = (q)->head; i != (q)->tail; i = (i + 1) % FLUSH_QUEUE_SIZE)
-
-static inline unsigned queue_ring_add(struct flush_queue *queue)
-{
-	unsigned idx = queue->tail;
-
-	assert_spin_locked(&queue->lock);
-	queue->tail = (idx + 1) % FLUSH_QUEUE_SIZE;
-
-	return idx;
-}
-
-static inline void queue_ring_remove_head(struct flush_queue *queue)
-{
-	assert_spin_locked(&queue->lock);
-	queue->head = (queue->head + 1) % FLUSH_QUEUE_SIZE;
-}
-
-static void queue_ring_free_flushed(struct dma_ops_domain *dom,
-				    struct flush_queue *queue)
-{
-	u64 counter = atomic64_read(&dom->flush_finish_cnt);
-	int idx;
-
-	queue_ring_for_each(idx, queue) {
-		/*
-		 * This assumes that counter values in the ring-buffer are
-		 * monotonously rising.
-		 */
-		if (queue->entries[idx].counter >= counter)
-			break;
-
-		free_iova_fast(&dom->iovad,
-			       queue->entries[idx].iova_pfn,
-			       queue->entries[idx].pages);
-
-		queue_ring_remove_head(queue);
-	}
-}
-
-static void queue_add(struct dma_ops_domain *dom,
-		      unsigned long address, unsigned long pages)
-{
-	struct flush_queue *queue;
-	unsigned long flags;
-	int idx;
-
-	pages     = __roundup_pow_of_two(pages);
-	address >>= PAGE_SHIFT;
-
-	queue = get_cpu_ptr(dom->flush_queue);
-	spin_lock_irqsave(&queue->lock, flags);
-
-	/*
-	 * First remove the enries from the ring-buffer that are already
-	 * flushed to make the below queue_ring_full() check less likely
-	 */
-	queue_ring_free_flushed(dom, queue);
-
-	/*
-	 * When ring-queue is full, flush the entries from the IOTLB so
-	 * that we can free all entries with queue_ring_free_flushed()
-	 * below.
-	 */
-	if (queue_ring_full(queue)) {
-		dma_ops_domain_flush_tlb(dom);
-		queue_ring_free_flushed(dom, queue);
-	}
-
-	idx = queue_ring_add(queue);
-
-	queue->entries[idx].iova_pfn = address;
-	queue->entries[idx].pages    = pages;
-	queue->entries[idx].counter  = atomic64_read(&dom->flush_start_cnt);
-
-	spin_unlock_irqrestore(&queue->lock, flags);
-
-	if (atomic_cmpxchg(&dom->flush_timer_on, 0, 1) == 0)
-		mod_timer(&dom->flush_timer, jiffies + msecs_to_jiffies(10));
-
-	put_cpu_ptr(dom->flush_queue);
-}
-
-static void queue_flush_timeout(unsigned long data)
-{
-	struct dma_ops_domain *dom = (struct dma_ops_domain *)data;
-	int cpu;
-
-	atomic_set(&dom->flush_timer_on, 0);
+	dom = container_of(iovad, struct dma_ops_domain, iovad);
 
 	dma_ops_domain_flush_tlb(dom);
-
-	for_each_possible_cpu(cpu) {
-		struct flush_queue *queue;
-		unsigned long flags;
-
-		queue = per_cpu_ptr(dom->flush_queue, cpu);
-		spin_lock_irqsave(&queue->lock, flags);
-		queue_ring_free_flushed(dom, queue);
-		spin_unlock_irqrestore(&queue->lock, flags);
-	}
 }
 
 /*
@@ -1973,11 +1771,6 @@ static void dma_ops_domain_free(struct dma_ops_domain *dom)
 
 	del_domain_from_list(&dom->domain);
 
-	if (timer_pending(&dom->flush_timer))
-		del_timer(&dom->flush_timer);
-
-	dma_ops_domain_free_flush_queue(dom);
-
 	put_iova_domain(&dom->iovad);
 
 	free_pagetable(&dom->domain);
@@ -2013,16 +1806,11 @@ static struct dma_ops_domain *dma_ops_domain_alloc(void)
 	init_iova_domain(&dma_dom->iovad, PAGE_SIZE,
 			 IOVA_START_PFN, DMA_32BIT_PFN);
 
-	/* Initialize reserved ranges */
-	copy_reserved_iova(&reserved_iova_ranges, &dma_dom->iovad);
-
-	if (dma_ops_domain_alloc_flush_queue(dma_dom))
+	if (init_iova_flush_queue(&dma_dom->iovad, iova_domain_flush_tlb, NULL))
 		goto free_dma_dom;
 
-	setup_timer(&dma_dom->flush_timer, queue_flush_timeout,
-		    (unsigned long)dma_dom);
-
-	atomic_set(&dma_dom->flush_timer_on, 0);
+	/* Initialize reserved ranges */
+	copy_reserved_iova(&reserved_iova_ranges, &dma_dom->iovad);
 
 	add_domain_to_list(&dma_dom->domain);
 
@@ -2619,7 +2407,8 @@ static void __unmap_single(struct dma_ops_domain *dma_dom,
 		domain_flush_tlb(&dma_dom->domain);
 		domain_flush_complete(&dma_dom->domain);
 	} else {
-		queue_add(dma_dom, dma_addr, pages);
+		pages = __roundup_pow_of_two(pages);
+		queue_iova(&dma_dom->iovad, dma_addr >> PAGE_SHIFT, pages, 0);
 	}
 }
 
diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c
index c8b0329c85d2..ca5ebaeafd6a 100644
--- a/drivers/iommu/dmar.c
+++ b/drivers/iommu/dmar.c
@@ -1343,7 +1343,7 @@ void qi_flush_dev_iotlb(struct intel_iommu *iommu, u16 sid, u16 qdep,
 
 	if (mask) {
 		BUG_ON(addr & ((1 << (VTD_PAGE_SHIFT + mask)) - 1));
-		addr |= (1 << (VTD_PAGE_SHIFT + mask - 1)) - 1;
+		addr |= (1ULL << (VTD_PAGE_SHIFT + mask - 1)) - 1;
 		desc.high = QI_DEV_IOTLB_ADDR(addr) | QI_DEV_IOTLB_SIZE;
 	} else
 		desc.high = QI_DEV_IOTLB_ADDR(addr);
diff --git a/drivers/iommu/fsl_pamu.c b/drivers/iommu/fsl_pamu.c
index a34355fca37a..919ad9045ac4 100644
--- a/drivers/iommu/fsl_pamu.c
+++ b/drivers/iommu/fsl_pamu.c
@@ -530,8 +530,8 @@ u32 get_stash_id(u32 stash_dest_hint, u32 vcpu)
 		if (node) {
 			prop = of_get_property(node, "cache-stash-id", NULL);
 			if (!prop) {
-				pr_debug("missing cache-stash-id at %s\n",
-					 node->full_name);
+				pr_debug("missing cache-stash-id at %pOF\n",
+					 node);
 				of_node_put(node);
 				return ~(u32)0;
 			}
@@ -557,8 +557,8 @@ found_cpu_node:
 		if (stash_dest_hint == cache_level) {
 			prop = of_get_property(node, "cache-stash-id", NULL);
 			if (!prop) {
-				pr_debug("missing cache-stash-id at %s\n",
-					 node->full_name);
+				pr_debug("missing cache-stash-id at %pOF\n",
+					 node);
 				of_node_put(node);
 				return ~(u32)0;
 			}
@@ -568,8 +568,7 @@ found_cpu_node:
 
 		prop = of_get_property(node, "next-level-cache", NULL);
 		if (!prop) {
-			pr_debug("can't find next-level-cache at %s\n",
-				 node->full_name);
+			pr_debug("can't find next-level-cache at %pOF\n", node);
 			of_node_put(node);
 			return ~(u32)0;  /* can't traverse any further */
 		}
@@ -1063,8 +1062,7 @@ static int fsl_pamu_probe(struct platform_device *pdev)
 
 	guts_node = of_find_matching_node(NULL, guts_device_ids);
 	if (!guts_node) {
-		dev_err(dev, "could not find GUTS node %s\n",
-			dev->of_node->full_name);
+		dev_err(dev, "could not find GUTS node %pOF\n", dev->of_node);
 		ret = -ENODEV;
 		goto error;
 	}
@@ -1246,8 +1244,7 @@ static __init int fsl_pamu_init(void)
 
 	pdev = platform_device_alloc("fsl-of-pamu", 0);
 	if (!pdev) {
-		pr_err("could not allocate device %s\n",
-		       np->full_name);
+		pr_err("could not allocate device %pOF\n", np);
 		ret = -ENOMEM;
 		goto error_device_alloc;
 	}
@@ -1259,8 +1256,7 @@ static __init int fsl_pamu_init(void)
 
 	ret = platform_device_add(pdev);
 	if (ret) {
-		pr_err("could not add device %s (err=%i)\n",
-		       np->full_name, ret);
+		pr_err("could not add device %pOF (err=%i)\n", np, ret);
 		goto error_device_add;
 	}
 
diff --git a/drivers/iommu/fsl_pamu_domain.c b/drivers/iommu/fsl_pamu_domain.c
index da0e1e30ef37..01c73479345d 100644
--- a/drivers/iommu/fsl_pamu_domain.c
+++ b/drivers/iommu/fsl_pamu_domain.c
@@ -619,8 +619,8 @@ static int handle_attach_device(struct fsl_dma_domain *dma_domain,
 	for (i = 0; i < num; i++) {
 		/* Ensure that LIODN value is valid */
 		if (liodn[i] >= PAACE_NUMBER_ENTRIES) {
-			pr_debug("Invalid liodn %d, attach device failed for %s\n",
-				 liodn[i], dev->of_node->full_name);
+			pr_debug("Invalid liodn %d, attach device failed for %pOF\n",
+				 liodn[i], dev->of_node);
 			ret = -EINVAL;
 			break;
 		}
@@ -684,8 +684,7 @@ static int fsl_pamu_attach_device(struct iommu_domain *domain,
 		liodn_cnt = len / sizeof(u32);
 		ret = handle_attach_device(dma_domain, dev, liodn, liodn_cnt);
 	} else {
-		pr_debug("missing fsl,liodn property at %s\n",
-			 dev->of_node->full_name);
+		pr_debug("missing fsl,liodn property at %pOF\n", dev->of_node);
 		ret = -EINVAL;
 	}
 
@@ -720,8 +719,7 @@ static void fsl_pamu_detach_device(struct iommu_domain *domain,
 	if (prop)
 		detach_device(dev, dma_domain);
 	else
-		pr_debug("missing fsl,liodn property at %s\n",
-			 dev->of_node->full_name);
+		pr_debug("missing fsl,liodn property at %pOF\n", dev->of_node);
 }
 
 static  int configure_domain_geometry(struct iommu_domain *domain, void *data)
diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
index 687f18f65cea..d5e8b8628a1a 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;
 
@@ -1309,6 +1284,13 @@ 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 +1604,25 @@ 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 +1933,16 @@ 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 +1994,6 @@ 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);
@@ -3206,7 +3206,7 @@ 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 +3239,6 @@ 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 +3421,9 @@ 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 +3628,6 @@ 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 +3663,8 @@ 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...
@@ -4721,7 +4601,6 @@ 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;
 }
 
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 3f6ea160afed..af69bf7e035a 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -1005,11 +1005,10 @@ struct iommu_group *iommu_group_get_for_dev(struct device *dev)
 	if (group)
 		return group;
 
-	group = ERR_PTR(-EINVAL);
-
-	if (ops && ops->device_group)
-		group = ops->device_group(dev);
+	if (!ops)
+		return ERR_PTR(-EINVAL);
 
+	group = ops->device_group(dev);
 	if (WARN_ON_ONCE(group == NULL))
 		return ERR_PTR(-EINVAL);
 
@@ -1298,12 +1297,8 @@ int iommu_attach_device(struct iommu_domain *domain, struct device *dev)
 	int ret;
 
 	group = iommu_group_get(dev);
-	/* FIXME: Remove this when groups a mandatory for iommu drivers */
-	if (group == NULL)
-		return __iommu_attach_device(domain, dev);
-
 	/*
-	 * We have a group - lock it to make sure the device-count doesn't
+	 * Lock the group to make sure the device-count doesn't
 	 * change while we are attaching
 	 */
 	mutex_lock(&group->mutex);
@@ -1336,9 +1331,6 @@ void iommu_detach_device(struct iommu_domain *domain, struct device *dev)
 	struct iommu_group *group;
 
 	group = iommu_group_get(dev);
-	/* FIXME: Remove this when groups a mandatory for iommu drivers */
-	if (group == NULL)
-		return __iommu_detach_device(domain, dev);
 
 	mutex_lock(&group->mutex);
 	if (iommu_group_device_count(group) != 1) {
@@ -1360,9 +1352,6 @@ struct iommu_domain *iommu_get_domain_for_dev(struct device *dev)
 	struct iommu_group *group;
 
 	group = iommu_group_get(dev);
-	/* FIXME: Remove this when groups a mandatory for iommu drivers */
-	if (group == NULL)
-		return NULL;
 
 	domain = group->domain;
 
diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c
index 246f14c83944..33edfa794ae9 100644
--- a/drivers/iommu/iova.c
+++ b/drivers/iommu/iova.c
@@ -32,6 +32,8 @@ static unsigned long iova_rcache_get(struct iova_domain *iovad,
 				     unsigned long limit_pfn);
 static void init_iova_rcaches(struct iova_domain *iovad);
 static void free_iova_rcaches(struct iova_domain *iovad);
+static void fq_destroy_all_entries(struct iova_domain *iovad);
+static void fq_flush_timeout(unsigned long data);
 
 void
 init_iova_domain(struct iova_domain *iovad, unsigned long granule,
@@ -50,10 +52,61 @@ init_iova_domain(struct iova_domain *iovad, unsigned long granule,
 	iovad->granule = granule;
 	iovad->start_pfn = start_pfn;
 	iovad->dma_32bit_pfn = pfn_32bit + 1;
+	iovad->flush_cb = NULL;
+	iovad->fq = NULL;
 	init_iova_rcaches(iovad);
 }
 EXPORT_SYMBOL_GPL(init_iova_domain);
 
+static void free_iova_flush_queue(struct iova_domain *iovad)
+{
+	if (!iovad->fq)
+		return;
+
+	if (timer_pending(&iovad->fq_timer))
+		del_timer(&iovad->fq_timer);
+
+	fq_destroy_all_entries(iovad);
+
+	free_percpu(iovad->fq);
+
+	iovad->fq         = NULL;
+	iovad->flush_cb   = NULL;
+	iovad->entry_dtor = NULL;
+}
+
+int init_iova_flush_queue(struct iova_domain *iovad,
+			  iova_flush_cb flush_cb, iova_entry_dtor entry_dtor)
+{
+	int cpu;
+
+	atomic64_set(&iovad->fq_flush_start_cnt,  0);
+	atomic64_set(&iovad->fq_flush_finish_cnt, 0);
+
+	iovad->fq = alloc_percpu(struct iova_fq);
+	if (!iovad->fq)
+		return -ENOMEM;
+
+	iovad->flush_cb   = flush_cb;
+	iovad->entry_dtor = entry_dtor;
+
+	for_each_possible_cpu(cpu) {
+		struct iova_fq *fq;
+
+		fq = per_cpu_ptr(iovad->fq, cpu);
+		fq->head = 0;
+		fq->tail = 0;
+
+		spin_lock_init(&fq->lock);
+	}
+
+	setup_timer(&iovad->fq_timer, fq_flush_timeout, (unsigned long)iovad);
+	atomic_set(&iovad->fq_timer_on, 0);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(init_iova_flush_queue);
+
 static struct rb_node *
 __get_cached_rbnode(struct iova_domain *iovad, unsigned long *limit_pfn)
 {
@@ -423,6 +476,135 @@ free_iova_fast(struct iova_domain *iovad, unsigned long pfn, unsigned long size)
 }
 EXPORT_SYMBOL_GPL(free_iova_fast);
 
+#define fq_ring_for_each(i, fq) \
+	for ((i) = (fq)->head; (i) != (fq)->tail; (i) = ((i) + 1) % IOVA_FQ_SIZE)
+
+static inline bool fq_full(struct iova_fq *fq)
+{
+	assert_spin_locked(&fq->lock);
+	return (((fq->tail + 1) % IOVA_FQ_SIZE) == fq->head);
+}
+
+static inline unsigned fq_ring_add(struct iova_fq *fq)
+{
+	unsigned idx = fq->tail;
+
+	assert_spin_locked(&fq->lock);
+
+	fq->tail = (idx + 1) % IOVA_FQ_SIZE;
+
+	return idx;
+}
+
+static void fq_ring_free(struct iova_domain *iovad, struct iova_fq *fq)
+{
+	u64 counter = atomic64_read(&iovad->fq_flush_finish_cnt);
+	unsigned idx;
+
+	assert_spin_locked(&fq->lock);
+
+	fq_ring_for_each(idx, fq) {
+
+		if (fq->entries[idx].counter >= counter)
+			break;
+
+		if (iovad->entry_dtor)
+			iovad->entry_dtor(fq->entries[idx].data);
+
+		free_iova_fast(iovad,
+			       fq->entries[idx].iova_pfn,
+			       fq->entries[idx].pages);
+
+		fq->head = (fq->head + 1) % IOVA_FQ_SIZE;
+	}
+}
+
+static void iova_domain_flush(struct iova_domain *iovad)
+{
+	atomic64_inc(&iovad->fq_flush_start_cnt);
+	iovad->flush_cb(iovad);
+	atomic64_inc(&iovad->fq_flush_finish_cnt);
+}
+
+static void fq_destroy_all_entries(struct iova_domain *iovad)
+{
+	int cpu;
+
+	/*
+	 * This code runs when the iova_domain is being detroyed, so don't
+	 * bother to free iovas, just call the entry_dtor on all remaining
+	 * entries.
+	 */
+	if (!iovad->entry_dtor)
+		return;
+
+	for_each_possible_cpu(cpu) {
+		struct iova_fq *fq = per_cpu_ptr(iovad->fq, cpu);
+		int idx;
+
+		fq_ring_for_each(idx, fq)
+			iovad->entry_dtor(fq->entries[idx].data);
+	}
+}
+
+static void fq_flush_timeout(unsigned long data)
+{
+	struct iova_domain *iovad = (struct iova_domain *)data;
+	int cpu;
+
+	atomic_set(&iovad->fq_timer_on, 0);
+	iova_domain_flush(iovad);
+
+	for_each_possible_cpu(cpu) {
+		unsigned long flags;
+		struct iova_fq *fq;
+
+		fq = per_cpu_ptr(iovad->fq, cpu);
+		spin_lock_irqsave(&fq->lock, flags);
+		fq_ring_free(iovad, fq);
+		spin_unlock_irqrestore(&fq->lock, flags);
+	}
+}
+
+void queue_iova(struct iova_domain *iovad,
+		unsigned long pfn, unsigned long pages,
+		unsigned long data)
+{
+	struct iova_fq *fq = get_cpu_ptr(iovad->fq);
+	unsigned long flags;
+	unsigned idx;
+
+	spin_lock_irqsave(&fq->lock, flags);
+
+	/*
+	 * First remove all entries from the flush queue that have already been
+	 * flushed out on another CPU. This makes the fq_full() check below less
+	 * likely to be true.
+	 */
+	fq_ring_free(iovad, fq);
+
+	if (fq_full(fq)) {
+		iova_domain_flush(iovad);
+		fq_ring_free(iovad, fq);
+	}
+
+	idx = fq_ring_add(fq);
+
+	fq->entries[idx].iova_pfn = pfn;
+	fq->entries[idx].pages    = pages;
+	fq->entries[idx].data     = data;
+	fq->entries[idx].counter  = atomic64_read(&iovad->fq_flush_start_cnt);
+
+	spin_unlock_irqrestore(&fq->lock, flags);
+
+	if (atomic_cmpxchg(&iovad->fq_timer_on, 0, 1) == 0)
+		mod_timer(&iovad->fq_timer,
+			  jiffies + msecs_to_jiffies(IOVA_FQ_TIMEOUT));
+
+	put_cpu_ptr(iovad->fq);
+}
+EXPORT_SYMBOL_GPL(queue_iova);
+
 /**
  * put_iova_domain - destroys the iova doamin
  * @iovad: - iova domain in question.
@@ -433,6 +615,7 @@ void put_iova_domain(struct iova_domain *iovad)
 	struct rb_node *node;
 	unsigned long flags;
 
+	free_iova_flush_queue(iovad);
 	free_iova_rcaches(iovad);
 	spin_lock_irqsave(&iovad->iova_rbtree_lock, flags);
 	node = rb_first(&iovad->rbroot);
diff --git a/drivers/iommu/msm_iommu.c b/drivers/iommu/msm_iommu.c
index d0448353d501..04f4d51ffacb 100644
--- a/drivers/iommu/msm_iommu.c
+++ b/drivers/iommu/msm_iommu.c
@@ -393,6 +393,7 @@ static struct msm_iommu_dev *find_iommu_for_dev(struct device *dev)
 static int msm_iommu_add_device(struct device *dev)
 {
 	struct msm_iommu_dev *iommu;
+	struct iommu_group *group;
 	unsigned long flags;
 	int ret = 0;
 
@@ -406,7 +407,16 @@ static int msm_iommu_add_device(struct device *dev)
 
 	spin_unlock_irqrestore(&msm_iommu_lock, flags);
 
-	return ret;
+	if (ret)
+		return ret;
+
+	group = iommu_group_get_for_dev(dev);
+	if (IS_ERR(group))
+		return PTR_ERR(group);
+
+	iommu_group_put(group);
+
+	return 0;
 }
 
 static void msm_iommu_remove_device(struct device *dev)
@@ -421,6 +431,8 @@ static void msm_iommu_remove_device(struct device *dev)
 		iommu_device_unlink(&iommu->iommu, dev);
 
 	spin_unlock_irqrestore(&msm_iommu_lock, flags);
+
+	iommu_group_remove_device(dev);
 }
 
 static int msm_iommu_attach_dev(struct iommu_domain *domain, struct device *dev)
@@ -700,6 +712,7 @@ static struct iommu_ops msm_iommu_ops = {
 	.iova_to_phys = msm_iommu_iova_to_phys,
 	.add_device = msm_iommu_add_device,
 	.remove_device = msm_iommu_remove_device,
+	.device_group = generic_device_group,
 	.pgsize_bitmap = MSM_IOMMU_PGSIZES,
 	.of_xlate = qcom_iommu_of_xlate,
 };
diff --git a/drivers/iommu/of_iommu.c b/drivers/iommu/of_iommu.c
index 8cb60829a7a1..e60e3dba85a0 100644
--- a/drivers/iommu/of_iommu.c
+++ b/drivers/iommu/of_iommu.c
@@ -25,6 +25,8 @@
 #include <linux/of_pci.h>
 #include <linux/slab.h>
 
+#define NO_IOMMU	1
+
 static const struct of_device_id __iommu_of_table_sentinel
 	__used __section(__iommu_of_table_end);
 
@@ -109,8 +111,8 @@ static bool of_iommu_driver_present(struct device_node *np)
 	return of_match_node(&__iommu_of_table, np);
 }
 
-static const struct iommu_ops
-*of_iommu_xlate(struct device *dev, struct of_phandle_args *iommu_spec)
+static int of_iommu_xlate(struct device *dev,
+			  struct of_phandle_args *iommu_spec)
 {
 	const struct iommu_ops *ops;
 	struct fwnode_handle *fwnode = &iommu_spec->np->fwnode;
@@ -120,95 +122,53 @@ static const struct iommu_ops
 	if ((ops && !ops->of_xlate) ||
 	    !of_device_is_available(iommu_spec->np) ||
 	    (!ops && !of_iommu_driver_present(iommu_spec->np)))
-		return NULL;
+		return NO_IOMMU;
 
 	err = iommu_fwspec_init(dev, &iommu_spec->np->fwnode, ops);
 	if (err)
-		return ERR_PTR(err);
+		return err;
 	/*
 	 * The otherwise-empty fwspec handily serves to indicate the specific
 	 * IOMMU device we're waiting for, which will be useful if we ever get
 	 * a proper probe-ordering dependency mechanism in future.
 	 */
 	if (!ops)
-		return ERR_PTR(-EPROBE_DEFER);
-
-	err = ops->of_xlate(dev, iommu_spec);
-	if (err)
-		return ERR_PTR(err);
+		return -EPROBE_DEFER;
 
-	return ops;
+	return ops->of_xlate(dev, iommu_spec);
 }
 
-static int __get_pci_rid(struct pci_dev *pdev, u16 alias, void *data)
-{
-	struct of_phandle_args *iommu_spec = data;
-
-	iommu_spec->args[0] = alias;
-	return iommu_spec->np == pdev->bus->dev.of_node;
-}
+struct of_pci_iommu_alias_info {
+	struct device *dev;
+	struct device_node *np;
+};
 
-static const struct iommu_ops
-*of_pci_iommu_init(struct pci_dev *pdev, struct device_node *bridge_np)
+static int of_pci_iommu_init(struct pci_dev *pdev, u16 alias, void *data)
 {
-	const struct iommu_ops *ops;
-	struct of_phandle_args iommu_spec;
+	struct of_pci_iommu_alias_info *info = data;
+	struct of_phandle_args iommu_spec = { .args_count = 1 };
 	int err;
 
-	/*
-	 * Start by tracing the RID alias down the PCI topology as
-	 * far as the host bridge whose OF node we have...
-	 * (we're not even attempting to handle multi-alias devices yet)
-	 */
-	iommu_spec.args_count = 1;
-	iommu_spec.np = bridge_np;
-	pci_for_each_dma_alias(pdev, __get_pci_rid, &iommu_spec);
-	/*
-	 * ...then find out what that becomes once it escapes the PCI
-	 * bus into the system beyond, and which IOMMU it ends up at.
-	 */
-	iommu_spec.np = NULL;
-	err = of_pci_map_rid(bridge_np, iommu_spec.args[0], "iommu-map",
+	err = of_pci_map_rid(info->np, alias, "iommu-map",
 			     "iommu-map-mask", &iommu_spec.np,
 			     iommu_spec.args);
 	if (err)
-		return err == -ENODEV ? NULL : ERR_PTR(err);
-
-	ops = of_iommu_xlate(&pdev->dev, &iommu_spec);
+		return err == -ENODEV ? NO_IOMMU : err;
 
+	err = of_iommu_xlate(info->dev, &iommu_spec);
 	of_node_put(iommu_spec.np);
-	return ops;
-}
-
-static const struct iommu_ops
-*of_platform_iommu_init(struct device *dev, struct device_node *np)
-{
-	struct of_phandle_args iommu_spec;
-	const struct iommu_ops *ops = NULL;
-	int idx = 0;
-
-	/*
-	 * We don't currently walk up the tree looking for a parent IOMMU.
-	 * See the `Notes:' section of
-	 * Documentation/devicetree/bindings/iommu/iommu.txt
-	 */
-	while (!of_parse_phandle_with_args(np, "iommus", "#iommu-cells",
-					   idx, &iommu_spec)) {
-		ops = of_iommu_xlate(dev, &iommu_spec);
-		of_node_put(iommu_spec.np);
-		idx++;
-		if (IS_ERR_OR_NULL(ops))
-			break;
-	}
+	if (err)
+		return err;
 
-	return ops;
+	return info->np == pdev->bus->dev.of_node;
 }
 
 const struct iommu_ops *of_iommu_configure(struct device *dev,
 					   struct device_node *master_np)
 {
-	const struct iommu_ops *ops;
+	const struct iommu_ops *ops = NULL;
 	struct iommu_fwspec *fwspec = dev->iommu_fwspec;
+	int err = NO_IOMMU;
 
 	if (!master_np)
 		return NULL;
@@ -221,25 +181,54 @@ const struct iommu_ops *of_iommu_configure(struct device *dev,
 		iommu_fwspec_free(dev);
 	}
 
-	if (dev_is_pci(dev))
-		ops = of_pci_iommu_init(to_pci_dev(dev), master_np);
-	else
-		ops = of_platform_iommu_init(dev, master_np);
+	/*
+	 * We don't currently walk up the tree looking for a parent IOMMU.
+	 * See the `Notes:' section of
+	 * Documentation/devicetree/bindings/iommu/iommu.txt
+	 */
+	if (dev_is_pci(dev)) {
+		struct of_pci_iommu_alias_info info = {
+			.dev = dev,
+			.np = master_np,
+		};
+
+		err = pci_for_each_dma_alias(to_pci_dev(dev),
+					     of_pci_iommu_init, &info);
+	} else {
+		struct of_phandle_args iommu_spec;
+		int idx = 0;
+
+		while (!of_parse_phandle_with_args(master_np, "iommus",
+						   "#iommu-cells",
+						   idx, &iommu_spec)) {
+			err = of_iommu_xlate(dev, &iommu_spec);
+			of_node_put(iommu_spec.np);
+			idx++;
+			if (err)
+				break;
+		}
+	}
+
+	/*
+	 * Two success conditions can be represented by non-negative err here:
+	 * >0 : there is no IOMMU, or one was unavailable for non-fatal reasons
+	 *  0 : we found an IOMMU, and dev->fwspec is initialised appropriately
+	 * <0 : any actual error
+	 */
+	if (!err)
+		ops = dev->iommu_fwspec->ops;
 	/*
 	 * If we have reason to believe the IOMMU driver missed the initial
 	 * add_device callback for dev, replay it to get things in order.
 	 */
-	if (!IS_ERR_OR_NULL(ops) && ops->add_device &&
-	    dev->bus && !dev->iommu_group) {
-		int err = ops->add_device(dev);
-
-		if (err)
-			ops = ERR_PTR(err);
-	}
+	if (ops && ops->add_device && dev->bus && !dev->iommu_group)
+		err = ops->add_device(dev);
 
 	/* Ignore all other errors apart from EPROBE_DEFER */
-	if (IS_ERR(ops) && (PTR_ERR(ops) != -EPROBE_DEFER)) {
-		dev_dbg(dev, "Adding to IOMMU failed: %ld\n", PTR_ERR(ops));
+	if (err == -EPROBE_DEFER) {
+		ops = ERR_PTR(err);
+	} else if (err < 0) {
+		dev_dbg(dev, "Adding to IOMMU failed: %d\n", err);
 		ops = NULL;
 	}
 
@@ -255,8 +244,7 @@ static int __init of_iommu_init(void)
 		const of_iommu_init_fn init_fn = match->data;
 
 		if (init_fn && init_fn(np))
-			pr_err("Failed to initialise IOMMU %s\n",
-				of_node_full_name(np));
+			pr_err("Failed to initialise IOMMU %pOF\n", np);
 	}
 
 	return 0;
diff --git a/drivers/iommu/rockchip-iommu.c b/drivers/iommu/rockchip-iommu.c
index 4ba48a26b389..1b8155dada26 100644
--- a/drivers/iommu/rockchip-iommu.c
+++ b/drivers/iommu/rockchip-iommu.c
@@ -1008,20 +1008,20 @@ static int rk_iommu_group_set_iommudata(struct iommu_group *group,
 	ret = of_parse_phandle_with_args(np, "iommus", "#iommu-cells", 0,
 					 &args);
 	if (ret) {
-		dev_err(dev, "of_parse_phandle_with_args(%s) => %d\n",
-			np->full_name, ret);
+		dev_err(dev, "of_parse_phandle_with_args(%pOF) => %d\n",
+			np, ret);
 		return ret;
 	}
 	if (args.args_count != 0) {
-		dev_err(dev, "incorrect number of iommu params found for %s (found %d, expected 0)\n",
-			args.np->full_name, args.args_count);
+		dev_err(dev, "incorrect number of iommu params found for %pOF (found %d, expected 0)\n",
+			args.np, args.args_count);
 		return -EINVAL;
 	}
 
 	pd = of_find_device_by_node(args.np);
 	of_node_put(args.np);
 	if (!pd) {
-		dev_err(dev, "iommu %s not found\n", args.np->full_name);
+		dev_err(dev, "iommu %pOF not found\n", args.np);
 		return -EPROBE_DEFER;
 	}
 
diff --git a/drivers/iommu/tegra-gart.c b/drivers/iommu/tegra-gart.c
index 37e708fdbb5a..29bafc6e82ae 100644
--- a/drivers/iommu/tegra-gart.c
+++ b/drivers/iommu/tegra-gart.c
@@ -334,12 +334,31 @@ static bool gart_iommu_capable(enum iommu_cap cap)
 	return false;
 }
 
+static int gart_iommu_add_device(struct device *dev)
+{
+	struct iommu_group *group = iommu_group_get_for_dev(dev);
+
+	if (IS_ERR(group))
+		return PTR_ERR(group);
+
+	iommu_group_put(group);
+	return 0;
+}
+
+static void gart_iommu_remove_device(struct device *dev)
+{
+	iommu_group_remove_device(dev);
+}
+
 static const struct iommu_ops gart_iommu_ops = {
 	.capable	= gart_iommu_capable,
 	.domain_alloc	= gart_iommu_domain_alloc,
 	.domain_free	= gart_iommu_domain_free,
 	.attach_dev	= gart_iommu_attach_dev,
 	.detach_dev	= gart_iommu_detach_dev,
+	.add_device	= gart_iommu_add_device,
+	.remove_device	= gart_iommu_remove_device,
+	.device_group	= generic_device_group,
 	.map		= gart_iommu_map,
 	.map_sg		= default_iommu_map_sg,
 	.unmap		= gart_iommu_unmap,
diff --git a/drivers/iommu/tegra-smmu.c b/drivers/iommu/tegra-smmu.c
index eeb19f560a05..faa9c1e70482 100644
--- a/drivers/iommu/tegra-smmu.c
+++ b/drivers/iommu/tegra-smmu.c
@@ -704,6 +704,7 @@ static struct tegra_smmu *tegra_smmu_find(struct device_node *np)
 static int tegra_smmu_add_device(struct device *dev)
 {
 	struct device_node *np = dev->of_node;
+	struct iommu_group *group;
 	struct of_phandle_args args;
 	unsigned int index = 0;
 
@@ -725,12 +726,19 @@ static int tegra_smmu_add_device(struct device *dev)
 		index++;
 	}
 
+	group = iommu_group_get_for_dev(dev);
+	if (IS_ERR(group))
+		return PTR_ERR(group);
+
+	iommu_group_put(group);
+
 	return 0;
 }
 
 static void tegra_smmu_remove_device(struct device *dev)
 {
 	dev->archdata.iommu = NULL;
+	iommu_group_remove_device(dev);
 }
 
 static const struct iommu_ops tegra_smmu_ops = {
@@ -741,6 +749,7 @@ static const struct iommu_ops tegra_smmu_ops = {
 	.detach_dev = tegra_smmu_detach_dev,
 	.add_device = tegra_smmu_add_device,
 	.remove_device = tegra_smmu_remove_device,
+	.device_group = generic_device_group,
 	.map = tegra_smmu_map,
 	.unmap = tegra_smmu_unmap,
 	.map_sg = default_iommu_map_sg,
diff --git a/include/linux/iommu.h b/include/linux/iommu.h
index 2cb54adc4a33..f1ce8e517d8d 100644
--- a/include/linux/iommu.h
+++ b/include/linux/iommu.h
@@ -425,13 +425,13 @@ static inline struct iommu_domain *iommu_get_domain_for_dev(struct device *dev)
 }
 
 static inline int iommu_map(struct iommu_domain *domain, unsigned long iova,
-			    phys_addr_t paddr, int gfp_order, int prot)
+			    phys_addr_t paddr, size_t size, int prot)
 {
 	return -ENODEV;
 }
 
 static inline int iommu_unmap(struct iommu_domain *domain, unsigned long iova,
-			      int gfp_order)
+			      size_t size)
 {
 	return -ENODEV;
 }
diff --git a/include/linux/iova.h b/include/linux/iova.h
index e0a892ae45c0..d179b9bf7814 100644
--- a/include/linux/iova.h
+++ b/include/linux/iova.h
@@ -14,6 +14,7 @@
 #include <linux/types.h>
 #include <linux/kernel.h>
 #include <linux/rbtree.h>
+#include <linux/atomic.h>
 #include <linux/dma-mapping.h>
 
 /* iova structure */
@@ -36,6 +37,35 @@ struct iova_rcache {
 	struct iova_cpu_rcache __percpu *cpu_rcaches;
 };
 
+struct iova_domain;
+
+/* Call-Back from IOVA code into IOMMU drivers */
+typedef void (* iova_flush_cb)(struct iova_domain *domain);
+
+/* Destructor for per-entry data */
+typedef void (* iova_entry_dtor)(unsigned long data);
+
+/* Number of entries per Flush Queue */
+#define IOVA_FQ_SIZE	256
+
+/* Timeout (in ms) after which entries are flushed from the Flush-Queue */
+#define IOVA_FQ_TIMEOUT	10
+
+/* Flush Queue entry for defered flushing */
+struct iova_fq_entry {
+	unsigned long iova_pfn;
+	unsigned long pages;
+	unsigned long data;
+	u64 counter; /* Flush counter when this entrie was added */
+};
+
+/* Per-CPU Flush Queue structure */
+struct iova_fq {
+	struct iova_fq_entry entries[IOVA_FQ_SIZE];
+	unsigned head, tail;
+	spinlock_t lock;
+};
+
 /* holds all the iova translations for a domain */
 struct iova_domain {
 	spinlock_t	iova_rbtree_lock; /* Lock to protect update of rbtree */
@@ -45,6 +75,25 @@ struct iova_domain {
 	unsigned long	start_pfn;	/* Lower limit for this domain */
 	unsigned long	dma_32bit_pfn;
 	struct iova_rcache rcaches[IOVA_RANGE_CACHE_MAX_SIZE];	/* IOVA range caches */
+
+	iova_flush_cb	flush_cb;	/* Call-Back function to flush IOMMU
+					   TLBs */
+
+	iova_entry_dtor entry_dtor;	/* IOMMU driver specific destructor for
+					   iova entry */
+
+	struct iova_fq __percpu *fq;	/* Flush Queue */
+
+	atomic64_t	fq_flush_start_cnt;	/* Number of TLB flushes that
+						   have been started */
+
+	atomic64_t	fq_flush_finish_cnt;	/* Number of TLB flushes that
+						   have been finished */
+
+	struct timer_list fq_timer;		/* Timer to regularily empty the
+						   flush-queues */
+	atomic_t fq_timer_on;			/* 1 when timer is active, 0
+						   when not */
 };
 
 static inline unsigned long iova_size(struct iova *iova)
@@ -95,6 +144,9 @@ struct iova *alloc_iova(struct iova_domain *iovad, unsigned long size,
 	bool size_aligned);
 void free_iova_fast(struct iova_domain *iovad, unsigned long pfn,
 		    unsigned long size);
+void queue_iova(struct iova_domain *iovad,
+		unsigned long pfn, unsigned long pages,
+		unsigned long data);
 unsigned long alloc_iova_fast(struct iova_domain *iovad, unsigned long size,
 			      unsigned long limit_pfn);
 struct iova *reserve_iova(struct iova_domain *iovad, unsigned long pfn_lo,
@@ -102,6 +154,8 @@ struct iova *reserve_iova(struct iova_domain *iovad, unsigned long pfn_lo,
 void copy_reserved_iova(struct iova_domain *from, struct iova_domain *to);
 void init_iova_domain(struct iova_domain *iovad, unsigned long granule,
 	unsigned long start_pfn, unsigned long pfn_32bit);
+int init_iova_flush_queue(struct iova_domain *iovad,
+			  iova_flush_cb flush_cb, iova_entry_dtor entry_dtor);
 struct iova *find_iova(struct iova_domain *iovad, unsigned long pfn);
 void put_iova_domain(struct iova_domain *iovad);
 struct iova *split_and_remove_iova(struct iova_domain *iovad,
@@ -148,6 +202,12 @@ static inline void free_iova_fast(struct iova_domain *iovad,
 {
 }
 
+static inline void queue_iova(struct iova_domain *iovad,
+			      unsigned long pfn, unsigned long pages,
+			      unsigned long data)
+{
+}
+
 static inline unsigned long alloc_iova_fast(struct iova_domain *iovad,
 					    unsigned long size,
 					    unsigned long limit_pfn)
@@ -174,6 +234,13 @@ static inline void init_iova_domain(struct iova_domain *iovad,
 {
 }
 
+static inline int init_iova_flush_queue(struct iova_domain *iovad,
+					iova_flush_cb flush_cb,
+					iova_entry_dtor entry_dtor)
+{
+	return -ENODEV;
+}
+
 static inline struct iova *find_iova(struct iova_domain *iovad,
 				     unsigned long pfn)
 {