1 files changed, 118 insertions, 20 deletions

diff --git a/drivers/iommu/tegra-smmu.c b/drivers/iommu/tegra-smmu.c
index 124c8848ab7e..0becdbfea306 100644
--- a/drivers/iommu/tegra-smmu.c
+++ b/drivers/iommu/tegra-smmu.c
@@ -12,6 +12,7 @@
 #include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
+#include <linux/spinlock.h>
 #include <linux/dma-mapping.h>
 
 #include <soc/tegra/ahb.h>
@@ -19,8 +20,10 @@
 
 struct tegra_smmu_group {
 	struct list_head list;
+	struct tegra_smmu *smmu;
 	const struct tegra_smmu_group_soc *soc;
 	struct iommu_group *group;
+	unsigned int swgroup;
 };
 
 struct tegra_smmu {
@@ -49,6 +52,7 @@ struct tegra_smmu_as {
 	struct iommu_domain domain;
 	struct tegra_smmu *smmu;
 	unsigned int use_count;
+	spinlock_t lock;
 	u32 *count;
 	struct page **pts;
 	struct page *pd;
@@ -127,6 +131,11 @@ static inline u32 smmu_readl(struct tegra_smmu *smmu, unsigned long offset)
 #define SMMU_PDE_SHIFT 22
 #define SMMU_PTE_SHIFT 12
 
+#define SMMU_PAGE_MASK		(~(SMMU_SIZE_PT-1))
+#define SMMU_OFFSET_IN_PAGE(x)	((unsigned long)(x) & ~SMMU_PAGE_MASK)
+#define SMMU_PFN_PHYS(x)	((phys_addr_t)(x) << SMMU_PTE_SHIFT)
+#define SMMU_PHYS_PFN(x)	((unsigned long)((x) >> SMMU_PTE_SHIFT))
+
 #define SMMU_PD_READABLE	(1 << 31)
 #define SMMU_PD_WRITABLE	(1 << 30)
 #define SMMU_PD_NONSECURE	(1 << 29)
@@ -308,6 +317,8 @@ static struct iommu_domain *tegra_smmu_domain_alloc(unsigned type)
 		return NULL;
 	}
 
+	spin_lock_init(&as->lock);
+
 	/* setup aperture */
 	as->domain.geometry.aperture_start = 0;
 	as->domain.geometry.aperture_end = 0xffffffff;
@@ -569,19 +580,14 @@ static u32 *tegra_smmu_pte_lookup(struct tegra_smmu_as *as, unsigned long iova,
 }
 
 static u32 *as_get_pte(struct tegra_smmu_as *as, dma_addr_t iova,
-		       dma_addr_t *dmap)
+		       dma_addr_t *dmap, struct page *page)
 {
 	unsigned int pde = iova_pd_index(iova);
 	struct tegra_smmu *smmu = as->smmu;
 
 	if (!as->pts[pde]) {
-		struct page *page;
 		dma_addr_t dma;
 
-		page = alloc_page(GFP_KERNEL | __GFP_DMA | __GFP_ZERO);
-		if (!page)
-			return NULL;
-
 		dma = dma_map_page(smmu->dev, page, 0, SMMU_SIZE_PT,
 				   DMA_TO_DEVICE);
 		if (dma_mapping_error(smmu->dev, dma)) {
@@ -644,7 +650,7 @@ static void tegra_smmu_set_pte(struct tegra_smmu_as *as, unsigned long iova,
 			       u32 *pte, dma_addr_t pte_dma, u32 val)
 {
 	struct tegra_smmu *smmu = as->smmu;
-	unsigned long offset = offset_in_page(pte);
+	unsigned long offset = SMMU_OFFSET_IN_PAGE(pte);
 
 	*pte = val;
 
@@ -655,15 +661,61 @@ static void tegra_smmu_set_pte(struct tegra_smmu_as *as, unsigned long iova,
 	smmu_flush(smmu);
 }
 
-static int tegra_smmu_map(struct iommu_domain *domain, unsigned long iova,
-			  phys_addr_t paddr, size_t size, int prot, gfp_t gfp)
+static struct page *as_get_pde_page(struct tegra_smmu_as *as,
+				    unsigned long iova, gfp_t gfp,
+				    unsigned long *flags)
+{
+	unsigned int pde = iova_pd_index(iova);
+	struct page *page = as->pts[pde];
+
+	/* at first check whether allocation needs to be done at all */
+	if (page)
+		return page;
+
+	/*
+	 * In order to prevent exhaustion of the atomic memory pool, we
+	 * allocate page in a sleeping context if GFP flags permit. Hence
+	 * spinlock needs to be unlocked and re-locked after allocation.
+	 */
+	if (!(gfp & __GFP_ATOMIC))
+		spin_unlock_irqrestore(&as->lock, *flags);
+
+	page = alloc_page(gfp | __GFP_DMA | __GFP_ZERO);
+
+	if (!(gfp & __GFP_ATOMIC))
+		spin_lock_irqsave(&as->lock, *flags);
+
+	/*
+	 * In a case of blocking allocation, a concurrent mapping may win
+	 * the PDE allocation. In this case the allocated page isn't needed
+	 * if allocation succeeded and the allocation failure isn't fatal.
+	 */
+	if (as->pts[pde]) {
+		if (page)
+			__free_page(page);
+
+		page = as->pts[pde];
+	}
+
+	return page;
+}
+
+static int
+__tegra_smmu_map(struct iommu_domain *domain, unsigned long iova,
+		 phys_addr_t paddr, size_t size, int prot, gfp_t gfp,
+		 unsigned long *flags)
 {
 	struct tegra_smmu_as *as = to_smmu_as(domain);
 	dma_addr_t pte_dma;
+	struct page *page;
 	u32 pte_attrs;
 	u32 *pte;
 
-	pte = as_get_pte(as, iova, &pte_dma);
+	page = as_get_pde_page(as, iova, gfp, flags);
+	if (!page)
+		return -ENOMEM;
+
+	pte = as_get_pte(as, iova, &pte_dma, page);
 	if (!pte)
 		return -ENOMEM;
 
@@ -680,13 +732,14 @@ static int tegra_smmu_map(struct iommu_domain *domain, unsigned long iova,
 		pte_attrs |= SMMU_PTE_WRITABLE;
 
 	tegra_smmu_set_pte(as, iova, pte, pte_dma,
-			   __phys_to_pfn(paddr) | pte_attrs);
+			   SMMU_PHYS_PFN(paddr) | pte_attrs);
 
 	return 0;
 }
 
-static size_t tegra_smmu_unmap(struct iommu_domain *domain, unsigned long iova,
-			       size_t size, struct iommu_iotlb_gather *gather)
+static size_t
+__tegra_smmu_unmap(struct iommu_domain *domain, unsigned long iova,
+		   size_t size, struct iommu_iotlb_gather *gather)
 {
 	struct tegra_smmu_as *as = to_smmu_as(domain);
 	dma_addr_t pte_dma;
@@ -702,6 +755,33 @@ static size_t tegra_smmu_unmap(struct iommu_domain *domain, unsigned long iova,
 	return size;
 }
 
+static int tegra_smmu_map(struct iommu_domain *domain, unsigned long iova,
+			  phys_addr_t paddr, size_t size, int prot, gfp_t gfp)
+{
+	struct tegra_smmu_as *as = to_smmu_as(domain);
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&as->lock, flags);
+	ret = __tegra_smmu_map(domain, iova, paddr, size, prot, gfp, &flags);
+	spin_unlock_irqrestore(&as->lock, flags);
+
+	return ret;
+}
+
+static size_t tegra_smmu_unmap(struct iommu_domain *domain, unsigned long iova,
+			       size_t size, struct iommu_iotlb_gather *gather)
+{
+	struct tegra_smmu_as *as = to_smmu_as(domain);
+	unsigned long flags;
+
+	spin_lock_irqsave(&as->lock, flags);
+	size = __tegra_smmu_unmap(domain, iova, size, gather);
+	spin_unlock_irqrestore(&as->lock, flags);
+
+	return size;
+}
+
 static phys_addr_t tegra_smmu_iova_to_phys(struct iommu_domain *domain,
 					   dma_addr_t iova)
 {
@@ -716,7 +796,7 @@ static phys_addr_t tegra_smmu_iova_to_phys(struct iommu_domain *domain,
 
 	pfn = *pte & as->smmu->pfn_mask;
 
-	return PFN_PHYS(pfn);
+	return SMMU_PFN_PHYS(pfn) + SMMU_OFFSET_IN_PAGE(iova);
 }
 
 static struct tegra_smmu *tegra_smmu_find(struct device_node *np)
@@ -813,22 +893,34 @@ tegra_smmu_find_group(struct tegra_smmu *smmu, unsigned int swgroup)
 	return NULL;
 }
 
+static void tegra_smmu_group_release(void *iommu_data)
+{
+	struct tegra_smmu_group *group = iommu_data;
+	struct tegra_smmu *smmu = group->smmu;
+
+	mutex_lock(&smmu->lock);
+	list_del(&group->list);
+	mutex_unlock(&smmu->lock);
+}
+
 static struct iommu_group *tegra_smmu_group_get(struct tegra_smmu *smmu,
 						unsigned int swgroup)
 {
 	const struct tegra_smmu_group_soc *soc;
 	struct tegra_smmu_group *group;
+	struct iommu_group *grp;
 
+	/* Find group_soc associating with swgroup */
 	soc = tegra_smmu_find_group(smmu, swgroup);
-	if (!soc)
-		return NULL;
 
 	mutex_lock(&smmu->lock);
 
+	/* Find existing iommu_group associating with swgroup or group_soc */
 	list_for_each_entry(group, &smmu->groups, list)
-		if (group->soc == soc) {
+		if ((group->swgroup == swgroup) || (soc && group->soc == soc)) {
+			grp = iommu_group_ref_get(group->group);
 			mutex_unlock(&smmu->lock);
-			return group->group;
+			return grp;
 		}
 
 	group = devm_kzalloc(smmu->dev, sizeof(*group), GFP_KERNEL);
@@ -838,6 +930,8 @@ static struct iommu_group *tegra_smmu_group_get(struct tegra_smmu *smmu,
 	}
 
 	INIT_LIST_HEAD(&group->list);
+	group->swgroup = swgroup;
+	group->smmu = smmu;
 	group->soc = soc;
 
 	group->group = iommu_group_alloc();
@@ -847,6 +941,9 @@ static struct iommu_group *tegra_smmu_group_get(struct tegra_smmu *smmu,
 		return NULL;
 	}
 
+	iommu_group_set_iommudata(group->group, group, tegra_smmu_group_release);
+	if (soc)
+		iommu_group_set_name(group->group, soc->name);
 	list_add_tail(&group->list, &smmu->groups);
 	mutex_unlock(&smmu->lock);
 
@@ -1019,10 +1116,11 @@ struct tegra_smmu *tegra_smmu_probe(struct device *dev,
 	smmu->dev = dev;
 	smmu->mc = mc;
 
-	smmu->pfn_mask = BIT_MASK(mc->soc->num_address_bits - PAGE_SHIFT) - 1;
+	smmu->pfn_mask =
+		BIT_MASK(mc->soc->num_address_bits - SMMU_PTE_SHIFT) - 1;
 	dev_dbg(dev, "address bits: %u, PFN mask: %#lx\n",
 		mc->soc->num_address_bits, smmu->pfn_mask);
-	smmu->tlb_mask = (smmu->soc->num_tlb_lines << 1) - 1;
+	smmu->tlb_mask = (1 << fls(smmu->soc->num_tlb_lines)) - 1;
 	dev_dbg(dev, "TLB lines: %u, mask: %#lx\n", smmu->soc->num_tlb_lines,
 		smmu->tlb_mask);