Merge branches 'amd/fixes', 'debug/dma-api', 'arm/omap', 'arm/msm', 'core', 'iommu/fault-reporting' and 'api/iommu-ops-per-bus' into next

Conflicts:
	drivers/iommu/amd_iommu.c
	drivers/iommu/iommu.c
diff --git a/arch/arm/mach-omap2/iommu2.c b/arch/arm/mach-omap2/iommu2.c
index f286012..eefc379 100644
--- a/arch/arm/mach-omap2/iommu2.c
+++ b/arch/arm/mach-omap2/iommu2.c
@@ -66,7 +66,7 @@
 	 ((pgsz) == MMU_CAM_PGSZ_4K)  ? 0xfffff000 : 0)
 
 
-static void __iommu_set_twl(struct iommu *obj, bool on)
+static void __iommu_set_twl(struct omap_iommu *obj, bool on)
 {
 	u32 l = iommu_read_reg(obj, MMU_CNTL);
 
@@ -85,7 +85,7 @@
 }
 
 
-static int omap2_iommu_enable(struct iommu *obj)
+static int omap2_iommu_enable(struct omap_iommu *obj)
 {
 	u32 l, pa;
 	unsigned long timeout;
@@ -127,7 +127,7 @@
 	return 0;
 }
 
-static void omap2_iommu_disable(struct iommu *obj)
+static void omap2_iommu_disable(struct omap_iommu *obj)
 {
 	u32 l = iommu_read_reg(obj, MMU_CNTL);
 
@@ -138,12 +138,12 @@
 	dev_dbg(obj->dev, "%s is shutting down\n", obj->name);
 }
 
-static void omap2_iommu_set_twl(struct iommu *obj, bool on)
+static void omap2_iommu_set_twl(struct omap_iommu *obj, bool on)
 {
 	__iommu_set_twl(obj, false);
 }
 
-static u32 omap2_iommu_fault_isr(struct iommu *obj, u32 *ra)
+static u32 omap2_iommu_fault_isr(struct omap_iommu *obj, u32 *ra)
 {
 	u32 stat, da;
 	u32 errs = 0;
@@ -173,13 +173,13 @@
 	return errs;
 }
 
-static void omap2_tlb_read_cr(struct iommu *obj, struct cr_regs *cr)
+static void omap2_tlb_read_cr(struct omap_iommu *obj, struct cr_regs *cr)
 {
 	cr->cam = iommu_read_reg(obj, MMU_READ_CAM);
 	cr->ram = iommu_read_reg(obj, MMU_READ_RAM);
 }
 
-static void omap2_tlb_load_cr(struct iommu *obj, struct cr_regs *cr)
+static void omap2_tlb_load_cr(struct omap_iommu *obj, struct cr_regs *cr)
 {
 	iommu_write_reg(obj, cr->cam | MMU_CAM_V, MMU_CAM);
 	iommu_write_reg(obj, cr->ram, MMU_RAM);
@@ -193,7 +193,8 @@
 	return cr->cam & mask;
 }
 
-static struct cr_regs *omap2_alloc_cr(struct iommu *obj, struct iotlb_entry *e)
+static struct cr_regs *omap2_alloc_cr(struct omap_iommu *obj,
+						struct iotlb_entry *e)
 {
 	struct cr_regs *cr;
 
@@ -230,7 +231,8 @@
 	return attr;
 }
 
-static ssize_t omap2_dump_cr(struct iommu *obj, struct cr_regs *cr, char *buf)
+static ssize_t
+omap2_dump_cr(struct omap_iommu *obj, struct cr_regs *cr, char *buf)
 {
 	char *p = buf;
 
@@ -254,7 +256,8 @@
 			goto out;					\
 	} while (0)
 
-static ssize_t omap2_iommu_dump_ctx(struct iommu *obj, char *buf, ssize_t len)
+static ssize_t
+omap2_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t len)
 {
 	char *p = buf;
 
@@ -280,7 +283,7 @@
 	return p - buf;
 }
 
-static void omap2_iommu_save_ctx(struct iommu *obj)
+static void omap2_iommu_save_ctx(struct omap_iommu *obj)
 {
 	int i;
 	u32 *p = obj->ctx;
@@ -293,7 +296,7 @@
 	BUG_ON(p[0] != IOMMU_ARCH_VERSION);
 }
 
-static void omap2_iommu_restore_ctx(struct iommu *obj)
+static void omap2_iommu_restore_ctx(struct omap_iommu *obj)
 {
 	int i;
 	u32 *p = obj->ctx;
@@ -343,13 +346,13 @@
 
 static int __init omap2_iommu_init(void)
 {
-	return install_iommu_arch(&omap2_iommu_ops);
+	return omap_install_iommu_arch(&omap2_iommu_ops);
 }
 module_init(omap2_iommu_init);
 
 static void __exit omap2_iommu_exit(void)
 {
-	uninstall_iommu_arch(&omap2_iommu_ops);
+	omap_uninstall_iommu_arch(&omap2_iommu_ops);
 }
 module_exit(omap2_iommu_exit);
 
diff --git a/arch/arm/plat-omap/Kconfig b/arch/arm/plat-omap/Kconfig
index bb8f4a6..fa62037 100644
--- a/arch/arm/plat-omap/Kconfig
+++ b/arch/arm/plat-omap/Kconfig
@@ -132,18 +132,6 @@
 	  This can also be changed at runtime (via the mbox_kfifo_size
 	  module parameter).
 
-config OMAP_IOMMU
-	tristate
-
-config OMAP_IOMMU_DEBUG
-       tristate "Export OMAP IOMMU internals in DebugFS"
-       depends on OMAP_IOMMU && DEBUG_FS
-       help
-         Select this to see extensive information about
-         the internal state of OMAP IOMMU in debugfs.
-
-         Say N unless you know you need this.
-
 config OMAP_IOMMU_IVA2
 	bool
 
diff --git a/arch/arm/plat-omap/Makefile b/arch/arm/plat-omap/Makefile
index f0233e6..9852622 100644
--- a/arch/arm/plat-omap/Makefile
+++ b/arch/arm/plat-omap/Makefile
@@ -18,8 +18,6 @@
 obj-$(CONFIG_ARCH_OMAP4) += omap_device.o
 
 obj-$(CONFIG_OMAP_MCBSP) += mcbsp.o
-obj-$(CONFIG_OMAP_IOMMU) += iommu.o iovmm.o
-obj-$(CONFIG_OMAP_IOMMU_DEBUG) += iommu-debug.o
 
 obj-$(CONFIG_CPU_FREQ) += cpu-omap.o
 obj-$(CONFIG_OMAP_DM_TIMER) += dmtimer.o
diff --git a/arch/arm/plat-omap/include/plat/iommu.h b/arch/arm/plat-omap/include/plat/iommu.h
index 174f1b9..a1d79ee 100644
--- a/arch/arm/plat-omap/include/plat/iommu.h
+++ b/arch/arm/plat-omap/include/plat/iommu.h
@@ -25,16 +25,17 @@
 	};
 };
 
-struct iommu {
+struct omap_iommu {
 	const char	*name;
 	struct module	*owner;
 	struct clk	*clk;
 	void __iomem	*regbase;
 	struct device	*dev;
 	void		*isr_priv;
+	struct iommu_domain *domain;
 
 	unsigned int	refcount;
-	struct mutex	iommu_lock;	/* global for this whole object */
+	spinlock_t	iommu_lock;	/* global for this whole object */
 
 	/*
 	 * We don't change iopgd for a situation like pgd for a task,
@@ -48,8 +49,6 @@
 	struct list_head	mmap;
 	struct mutex		mmap_lock; /* protect mmap */
 
-	int (*isr)(struct iommu *obj, u32 da, u32 iommu_errs, void *priv);
-
 	void *ctx; /* iommu context: registres saved area */
 	u32 da_start;
 	u32 da_end;
@@ -81,25 +80,27 @@
 struct iommu_functions {
 	unsigned long	version;
 
-	int (*enable)(struct iommu *obj);
-	void (*disable)(struct iommu *obj);
-	void (*set_twl)(struct iommu *obj, bool on);
-	u32 (*fault_isr)(struct iommu *obj, u32 *ra);
+	int (*enable)(struct omap_iommu *obj);
+	void (*disable)(struct omap_iommu *obj);
+	void (*set_twl)(struct omap_iommu *obj, bool on);
+	u32 (*fault_isr)(struct omap_iommu *obj, u32 *ra);
 
-	void (*tlb_read_cr)(struct iommu *obj, struct cr_regs *cr);
-	void (*tlb_load_cr)(struct iommu *obj, struct cr_regs *cr);
+	void (*tlb_read_cr)(struct omap_iommu *obj, struct cr_regs *cr);
+	void (*tlb_load_cr)(struct omap_iommu *obj, struct cr_regs *cr);
 
-	struct cr_regs *(*alloc_cr)(struct iommu *obj, struct iotlb_entry *e);
+	struct cr_regs *(*alloc_cr)(struct omap_iommu *obj,
+							struct iotlb_entry *e);
 	int (*cr_valid)(struct cr_regs *cr);
 	u32 (*cr_to_virt)(struct cr_regs *cr);
 	void (*cr_to_e)(struct cr_regs *cr, struct iotlb_entry *e);
-	ssize_t (*dump_cr)(struct iommu *obj, struct cr_regs *cr, char *buf);
+	ssize_t (*dump_cr)(struct omap_iommu *obj, struct cr_regs *cr,
+							char *buf);
 
 	u32 (*get_pte_attr)(struct iotlb_entry *e);
 
-	void (*save_ctx)(struct iommu *obj);
-	void (*restore_ctx)(struct iommu *obj);
-	ssize_t (*dump_ctx)(struct iommu *obj, char *buf, ssize_t len);
+	void (*save_ctx)(struct omap_iommu *obj);
+	void (*restore_ctx)(struct omap_iommu *obj);
+	ssize_t (*dump_ctx)(struct omap_iommu *obj, char *buf, ssize_t len);
 };
 
 struct iommu_platform_data {
@@ -150,40 +151,31 @@
 /*
  * global functions
  */
-extern u32 iommu_arch_version(void);
+extern u32 omap_iommu_arch_version(void);
 
-extern void iotlb_cr_to_e(struct cr_regs *cr, struct iotlb_entry *e);
-extern u32 iotlb_cr_to_virt(struct cr_regs *cr);
+extern void omap_iotlb_cr_to_e(struct cr_regs *cr, struct iotlb_entry *e);
 
-extern int load_iotlb_entry(struct iommu *obj, struct iotlb_entry *e);
-extern void iommu_set_twl(struct iommu *obj, bool on);
-extern void flush_iotlb_page(struct iommu *obj, u32 da);
-extern void flush_iotlb_range(struct iommu *obj, u32 start, u32 end);
-extern void flush_iotlb_all(struct iommu *obj);
+extern int
+omap_iopgtable_store_entry(struct omap_iommu *obj, struct iotlb_entry *e);
 
-extern int iopgtable_store_entry(struct iommu *obj, struct iotlb_entry *e);
-extern void iopgtable_lookup_entry(struct iommu *obj, u32 da, u32 **ppgd,
-				   u32 **ppte);
-extern size_t iopgtable_clear_entry(struct iommu *obj, u32 iova);
-
-extern int iommu_set_da_range(struct iommu *obj, u32 start, u32 end);
-extern struct iommu *iommu_get(const char *name);
-extern void iommu_put(struct iommu *obj);
-extern int iommu_set_isr(const char *name,
-			 int (*isr)(struct iommu *obj, u32 da, u32 iommu_errs,
+extern int omap_iommu_set_isr(const char *name,
+		 int (*isr)(struct omap_iommu *obj, u32 da, u32 iommu_errs,
 				    void *priv),
 			 void *isr_priv);
 
-extern void iommu_save_ctx(struct iommu *obj);
-extern void iommu_restore_ctx(struct iommu *obj);
+extern void omap_iommu_save_ctx(struct omap_iommu *obj);
+extern void omap_iommu_restore_ctx(struct omap_iommu *obj);
 
-extern int install_iommu_arch(const struct iommu_functions *ops);
-extern void uninstall_iommu_arch(const struct iommu_functions *ops);
+extern int omap_install_iommu_arch(const struct iommu_functions *ops);
+extern void omap_uninstall_iommu_arch(const struct iommu_functions *ops);
 
-extern int foreach_iommu_device(void *data,
+extern int omap_foreach_iommu_device(void *data,
 				int (*fn)(struct device *, void *));
 
-extern ssize_t iommu_dump_ctx(struct iommu *obj, char *buf, ssize_t len);
-extern size_t dump_tlb_entries(struct iommu *obj, char *buf, ssize_t len);
+extern ssize_t
+omap_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t len);
+extern size_t
+omap_dump_tlb_entries(struct omap_iommu *obj, char *buf, ssize_t len);
+struct device *omap_find_iommu_device(const char *name);
 
 #endif /* __MACH_IOMMU_H */
diff --git a/arch/arm/plat-omap/include/plat/iommu2.h b/arch/arm/plat-omap/include/plat/iommu2.h
index 10ad05f..d4116b5 100644
--- a/arch/arm/plat-omap/include/plat/iommu2.h
+++ b/arch/arm/plat-omap/include/plat/iommu2.h
@@ -83,12 +83,12 @@
 /*
  * register accessors
  */
-static inline u32 iommu_read_reg(struct iommu *obj, size_t offs)
+static inline u32 iommu_read_reg(struct omap_iommu *obj, size_t offs)
 {
 	return __raw_readl(obj->regbase + offs);
 }
 
-static inline void iommu_write_reg(struct iommu *obj, u32 val, size_t offs)
+static inline void iommu_write_reg(struct omap_iommu *obj, u32 val, size_t offs)
 {
 	__raw_writel(val, obj->regbase + offs);
 }
diff --git a/arch/arm/plat-omap/iopgtable.h b/arch/arm/plat-omap/include/plat/iopgtable.h
similarity index 81%
rename from arch/arm/plat-omap/iopgtable.h
rename to arch/arm/plat-omap/include/plat/iopgtable.h
index c3e93bb..66a8139 100644
--- a/arch/arm/plat-omap/iopgtable.h
+++ b/arch/arm/plat-omap/include/plat/iopgtable.h
@@ -56,6 +56,19 @@
 
 #define IOPAGE_MASK		IOPTE_MASK
 
+/**
+ * omap_iommu_translate() - va to pa translation
+ * @d:		omap iommu descriptor
+ * @va:		virtual address
+ * @mask:	omap iommu descriptor mask
+ *
+ * va to pa translation
+ */
+static inline phys_addr_t omap_iommu_translate(u32 d, u32 va, u32 mask)
+{
+	return (d & mask) | (va & (~mask));
+}
+
 /*
  * some descriptor attributes.
  */
@@ -64,10 +77,15 @@
 #define IOPGD_SUPER		(1 << 18 | 2 << 0)
 
 #define iopgd_is_table(x)	(((x) & 3) == IOPGD_TABLE)
+#define iopgd_is_section(x)	(((x) & (1 << 18 | 3)) == IOPGD_SECTION)
+#define iopgd_is_super(x)	(((x) & (1 << 18 | 3)) == IOPGD_SUPER)
 
 #define IOPTE_SMALL		(2 << 0)
 #define IOPTE_LARGE		(1 << 0)
 
+#define iopte_is_small(x)	(((x) & 2) == IOPTE_SMALL)
+#define iopte_is_large(x)	(((x) & 3) == IOPTE_LARGE)
+
 /* to find an entry in a page-table-directory */
 #define iopgd_index(da)		(((da) >> IOPGD_SHIFT) & (PTRS_PER_IOPGD - 1))
 #define iopgd_offset(obj, da)	((obj)->iopgd + iopgd_index(da))
@@ -97,6 +115,6 @@
 }
 
 #define to_iommu(dev)							\
-	(struct iommu *)platform_get_drvdata(to_platform_device(dev))
+	(struct omap_iommu *)platform_get_drvdata(to_platform_device(dev))
 
 #endif /* __PLAT_OMAP_IOMMU_H */
diff --git a/arch/arm/plat-omap/include/plat/iovmm.h b/arch/arm/plat-omap/include/plat/iovmm.h
index e992b96..6af1a91 100644
--- a/arch/arm/plat-omap/include/plat/iovmm.h
+++ b/arch/arm/plat-omap/include/plat/iovmm.h
@@ -13,8 +13,10 @@
 #ifndef __IOMMU_MMAP_H
 #define __IOMMU_MMAP_H
 
+#include <linux/iommu.h>
+
 struct iovm_struct {
-	struct iommu		*iommu;	/* iommu object which this belongs to */
+	struct omap_iommu	*iommu;	/* iommu object which this belongs to */
 	u32			da_start; /* area definition */
 	u32			da_end;
 	u32			flags; /* IOVMF_: see below */
@@ -70,20 +72,18 @@
 #define IOVMF_DA_FIXED		(1 << (4 + IOVMF_SW_SHIFT))
 
 
-extern struct iovm_struct *find_iovm_area(struct iommu *obj, u32 da);
-extern u32 iommu_vmap(struct iommu *obj, u32 da,
+extern struct iovm_struct *omap_find_iovm_area(struct omap_iommu *obj, u32 da);
+extern u32
+omap_iommu_vmap(struct iommu_domain *domain, struct omap_iommu *obj, u32 da,
 			const struct sg_table *sgt, u32 flags);
-extern struct sg_table *iommu_vunmap(struct iommu *obj, u32 da);
-extern u32 iommu_vmalloc(struct iommu *obj, u32 da, size_t bytes,
-			   u32 flags);
-extern void iommu_vfree(struct iommu *obj, const u32 da);
-extern u32 iommu_kmap(struct iommu *obj, u32 da, u32 pa, size_t bytes,
-			u32 flags);
-extern void iommu_kunmap(struct iommu *obj, u32 da);
-extern u32 iommu_kmalloc(struct iommu *obj, u32 da, size_t bytes,
-			   u32 flags);
-extern void iommu_kfree(struct iommu *obj, u32 da);
-
-extern void *da_to_va(struct iommu *obj, u32 da);
+extern struct sg_table *omap_iommu_vunmap(struct iommu_domain *domain,
+				struct omap_iommu *obj, u32 da);
+extern u32
+omap_iommu_vmalloc(struct iommu_domain *domain, struct omap_iommu *obj,
+				u32 da, size_t bytes, u32 flags);
+extern void
+omap_iommu_vfree(struct iommu_domain *domain, struct omap_iommu *obj,
+				const u32 da);
+extern void *omap_da_to_va(struct omap_iommu *obj, u32 da);
 
 #endif /* __IOMMU_MMAP_H */
diff --git a/arch/ia64/kvm/kvm-ia64.c b/arch/ia64/kvm/kvm-ia64.c
index 8213efe..43f4c92 100644
--- a/arch/ia64/kvm/kvm-ia64.c
+++ b/arch/ia64/kvm/kvm-ia64.c
@@ -33,6 +33,7 @@
 #include <linux/uaccess.h>
 #include <linux/iommu.h>
 #include <linux/intel-iommu.h>
+#include <linux/pci.h>
 
 #include <asm/pgtable.h>
 #include <asm/gcc_intrin.h>
@@ -204,7 +205,7 @@
 		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
 		break;
 	case KVM_CAP_IOMMU:
-		r = iommu_found();
+		r = iommu_present(&pci_bus_type);
 		break;
 	default:
 		r = 0;
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 84a28ea..73c6a42 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -44,6 +44,7 @@
 #include <linux/perf_event.h>
 #include <linux/uaccess.h>
 #include <linux/hash.h>
+#include <linux/pci.h>
 #include <trace/events/kvm.h>
 
 #define CREATE_TRACE_POINTS
@@ -2095,7 +2096,7 @@
 		r = 0;
 		break;
 	case KVM_CAP_IOMMU:
-		r = iommu_found();
+		r = iommu_present(&pci_bus_type);
 		break;
 	case KVM_CAP_MCE:
 		r = KVM_MAX_MCE_BANKS;
diff --git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig
index b57b3fa..ae46776 100644
--- a/drivers/iommu/Kconfig
+++ b/drivers/iommu/Kconfig
@@ -107,4 +107,23 @@
 	  To use x2apic mode in the CPU's which support x2APIC enhancements or
 	  to support platforms with CPU's having > 8 bit APIC ID, say Y.
 
+# OMAP IOMMU support
+config OMAP_IOMMU
+	bool "OMAP IOMMU Support"
+	depends on ARCH_OMAP
+	select IOMMU_API
+
+config OMAP_IOVMM
+	tristate "OMAP IO Virtual Memory Manager Support"
+	depends on OMAP_IOMMU
+
+config OMAP_IOMMU_DEBUG
+       tristate "Export OMAP IOMMU/IOVMM internals in DebugFS"
+       depends on OMAP_IOVMM && DEBUG_FS
+       help
+         Select this to see extensive information about
+         the internal state of OMAP IOMMU/IOVMM in debugfs.
+
+         Say N unless you know you need this.
+
 endif # IOMMU_SUPPORT
diff --git a/drivers/iommu/Makefile b/drivers/iommu/Makefile
index 4d4d77d..f798cdd 100644
--- a/drivers/iommu/Makefile
+++ b/drivers/iommu/Makefile
@@ -3,3 +3,6 @@
 obj-$(CONFIG_AMD_IOMMU) += amd_iommu.o amd_iommu_init.o
 obj-$(CONFIG_DMAR) += dmar.o iova.o intel-iommu.o
 obj-$(CONFIG_INTR_REMAP) += dmar.o intr_remapping.o
+obj-$(CONFIG_OMAP_IOMMU) += omap-iommu.o
+obj-$(CONFIG_OMAP_IOVMM) += omap-iovmm.o
+obj-$(CONFIG_OMAP_IOMMU_DEBUG) += omap-iommu-debug.o
diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index 0e4227f..4ee277a 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -1283,7 +1283,7 @@
 		if (!pte || !IOMMU_PTE_PRESENT(*pte))
 			continue;
 
-		dma_ops_reserve_addresses(dma_dom, i << PAGE_SHIFT, 1);
+		dma_ops_reserve_addresses(dma_dom, i >> PAGE_SHIFT, 1);
 	}
 
 	update_domain(&dma_dom->domain);
@@ -2495,7 +2495,7 @@
 
 void __init amd_iommu_init_api(void)
 {
-	register_iommu(&amd_iommu_ops);
+	bus_set_iommu(&pci_bus_type, &amd_iommu_ops);
 }
 
 int __init amd_iommu_init_dma_ops(void)
diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
index c621c98..2d53c3d 100644
--- a/drivers/iommu/intel-iommu.c
+++ b/drivers/iommu/intel-iommu.c
@@ -3486,7 +3486,7 @@
 
 	init_iommu_pm_ops();
 
-	register_iommu(&intel_iommu_ops);
+	bus_set_iommu(&pci_bus_type, &intel_iommu_ops);
 
 	bus_register_notifier(&pci_bus_type, &device_nb);
 
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 6e6b6a1..2fb2963 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -16,6 +16,8 @@
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
  */
 
+#include <linux/device.h>
+#include <linux/kernel.h>
 #include <linux/bug.h>
 #include <linux/types.h>
 #include <linux/module.h>
@@ -23,32 +25,78 @@
 #include <linux/errno.h>
 #include <linux/iommu.h>
 
-static struct iommu_ops *iommu_ops;
-
-void register_iommu(struct iommu_ops *ops)
+static void iommu_bus_init(struct bus_type *bus, struct iommu_ops *ops)
 {
-	if (iommu_ops)
-		BUG();
-
-	iommu_ops = ops;
 }
 
-bool iommu_found(void)
+/**
+ * bus_set_iommu - set iommu-callbacks for the bus
+ * @bus: bus.
+ * @ops: the callbacks provided by the iommu-driver
+ *
+ * This function is called by an iommu driver to set the iommu methods
+ * used for a particular bus. Drivers for devices on that bus can use
+ * the iommu-api after these ops are registered.
+ * This special function is needed because IOMMUs are usually devices on
+ * the bus itself, so the iommu drivers are not initialized when the bus
+ * is set up. With this function the iommu-driver can set the iommu-ops
+ * afterwards.
+ */
+int bus_set_iommu(struct bus_type *bus, struct iommu_ops *ops)
 {
-	return iommu_ops != NULL;
-}
-EXPORT_SYMBOL_GPL(iommu_found);
+	if (bus->iommu_ops != NULL)
+		return -EBUSY;
 
-struct iommu_domain *iommu_domain_alloc(void)
+	bus->iommu_ops = ops;
+
+	/* Do IOMMU specific setup for this bus-type */
+	iommu_bus_init(bus, ops);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(bus_set_iommu);
+
+bool iommu_present(struct bus_type *bus)
+{
+	return bus->iommu_ops != NULL;
+}
+EXPORT_SYMBOL_GPL(iommu_present);
+
+/**
+ * iommu_set_fault_handler() - set a fault handler for an iommu domain
+ * @domain: iommu domain
+ * @handler: fault handler
+ *
+ * This function should be used by IOMMU users which want to be notified
+ * whenever an IOMMU fault happens.
+ *
+ * The fault handler itself should return 0 on success, and an appropriate
+ * error code otherwise.
+ */
+void iommu_set_fault_handler(struct iommu_domain *domain,
+					iommu_fault_handler_t handler)
+{
+	BUG_ON(!domain);
+
+	domain->handler = handler;
+}
+EXPORT_SYMBOL_GPL(iommu_set_fault_handler);
+
+struct iommu_domain *iommu_domain_alloc(struct bus_type *bus)
 {
 	struct iommu_domain *domain;
 	int ret;
 
+	if (bus == NULL || bus->iommu_ops == NULL)
+		return NULL;
+
 	domain = kmalloc(sizeof(*domain), GFP_KERNEL);
 	if (!domain)
 		return NULL;
 
-	ret = iommu_ops->domain_init(domain);
+	domain->ops = bus->iommu_ops;
+
+	ret = domain->ops->domain_init(domain);
 	if (ret)
 		goto out_free;
 
@@ -63,62 +111,78 @@
 
 void iommu_domain_free(struct iommu_domain *domain)
 {
-	iommu_ops->domain_destroy(domain);
+	if (likely(domain->ops->domain_destroy != NULL))
+		domain->ops->domain_destroy(domain);
+
 	kfree(domain);
 }
 EXPORT_SYMBOL_GPL(iommu_domain_free);
 
 int iommu_attach_device(struct iommu_domain *domain, struct device *dev)
 {
-	return iommu_ops->attach_dev(domain, dev);
+	if (unlikely(domain->ops->attach_dev == NULL))
+		return -ENODEV;
+
+	return domain->ops->attach_dev(domain, dev);
 }
 EXPORT_SYMBOL_GPL(iommu_attach_device);
 
 void iommu_detach_device(struct iommu_domain *domain, struct device *dev)
 {
-	iommu_ops->detach_dev(domain, dev);
+	if (unlikely(domain->ops->detach_dev == NULL))
+		return;
+
+	domain->ops->detach_dev(domain, dev);
 }
 EXPORT_SYMBOL_GPL(iommu_detach_device);
 
 phys_addr_t iommu_iova_to_phys(struct iommu_domain *domain,
 			       unsigned long iova)
 {
-	return iommu_ops->iova_to_phys(domain, iova);
+	if (unlikely(domain->ops->iova_to_phys == NULL))
+		return 0;
+
+	return domain->ops->iova_to_phys(domain, iova);
 }
 EXPORT_SYMBOL_GPL(iommu_iova_to_phys);
 
 int iommu_domain_has_cap(struct iommu_domain *domain,
 			 unsigned long cap)
 {
-	return iommu_ops->domain_has_cap(domain, cap);
+	if (unlikely(domain->ops->domain_has_cap == NULL))
+		return 0;
+
+	return domain->ops->domain_has_cap(domain, cap);
 }
 EXPORT_SYMBOL_GPL(iommu_domain_has_cap);
 
 int iommu_map(struct iommu_domain *domain, unsigned long iova,
 	      phys_addr_t paddr, int gfp_order, int prot)
 {
-	unsigned long invalid_mask;
 	size_t size;
 
-	size         = 0x1000UL << gfp_order;
-	invalid_mask = size - 1;
+	if (unlikely(domain->ops->map == NULL))
+		return -ENODEV;
 
-	BUG_ON((iova | paddr) & invalid_mask);
+	size         = PAGE_SIZE << gfp_order;
 
-	return iommu_ops->map(domain, iova, paddr, gfp_order, prot);
+	BUG_ON(!IS_ALIGNED(iova | paddr, size));
+
+	return domain->ops->map(domain, iova, paddr, gfp_order, prot);
 }
 EXPORT_SYMBOL_GPL(iommu_map);
 
 int iommu_unmap(struct iommu_domain *domain, unsigned long iova, int gfp_order)
 {
-	unsigned long invalid_mask;
 	size_t size;
 
-	size         = 0x1000UL << gfp_order;
-	invalid_mask = size - 1;
+	if (unlikely(domain->ops->unmap == NULL))
+		return -ENODEV;
 
-	BUG_ON(iova & invalid_mask);
+	size         = PAGE_SIZE << gfp_order;
 
-	return iommu_ops->unmap(domain, iova, gfp_order);
+	BUG_ON(!IS_ALIGNED(iova, size));
+
+	return domain->ops->unmap(domain, iova, gfp_order);
 }
 EXPORT_SYMBOL_GPL(iommu_unmap);
diff --git a/drivers/iommu/msm_iommu.c b/drivers/iommu/msm_iommu.c
index 1a584e0..5865dd2 100644
--- a/drivers/iommu/msm_iommu.c
+++ b/drivers/iommu/msm_iommu.c
@@ -543,6 +543,13 @@
 	}
 
 	ret = __flush_iotlb(domain);
+
+	/*
+	 * the IOMMU API requires us to return the order of the unmapped
+	 * page (on success).
+	 */
+	if (!ret)
+		ret = order;
 fail:
 	spin_unlock_irqrestore(&msm_iommu_lock, flags);
 	return ret;
@@ -721,7 +728,7 @@
 static int __init msm_iommu_init(void)
 {
 	setup_iommu_tex_classes();
-	register_iommu(&msm_iommu_ops);
+	bus_set_iommu(&platform_bus_type, &msm_iommu_ops);
 	return 0;
 }
 
diff --git a/arch/arm/plat-omap/iommu-debug.c b/drivers/iommu/omap-iommu-debug.c
similarity index 90%
rename from arch/arm/plat-omap/iommu-debug.c
rename to drivers/iommu/omap-iommu-debug.c
index f07cf2f..9c192e7 100644
--- a/arch/arm/plat-omap/iommu-debug.c
+++ b/drivers/iommu/omap-iommu-debug.c
@@ -21,7 +21,7 @@
 #include <plat/iommu.h>
 #include <plat/iovmm.h>
 
-#include "iopgtable.h"
+#include <plat/iopgtable.h>
 
 #define MAXCOLUMN 100 /* for short messages */
 
@@ -32,7 +32,7 @@
 static ssize_t debug_read_ver(struct file *file, char __user *userbuf,
 			      size_t count, loff_t *ppos)
 {
-	u32 ver = iommu_arch_version();
+	u32 ver = omap_iommu_arch_version();
 	char buf[MAXCOLUMN], *p = buf;
 
 	p += sprintf(p, "H/W version: %d.%d\n", (ver >> 4) & 0xf , ver & 0xf);
@@ -43,7 +43,7 @@
 static ssize_t debug_read_regs(struct file *file, char __user *userbuf,
 			       size_t count, loff_t *ppos)
 {
-	struct iommu *obj = file->private_data;
+	struct omap_iommu *obj = file->private_data;
 	char *p, *buf;
 	ssize_t bytes;
 
@@ -54,7 +54,7 @@
 
 	mutex_lock(&iommu_debug_lock);
 
-	bytes = iommu_dump_ctx(obj, p, count);
+	bytes = omap_iommu_dump_ctx(obj, p, count);
 	bytes = simple_read_from_buffer(userbuf, count, ppos, buf, bytes);
 
 	mutex_unlock(&iommu_debug_lock);
@@ -66,7 +66,7 @@
 static ssize_t debug_read_tlb(struct file *file, char __user *userbuf,
 			      size_t count, loff_t *ppos)
 {
-	struct iommu *obj = file->private_data;
+	struct omap_iommu *obj = file->private_data;
 	char *p, *buf;
 	ssize_t bytes, rest;
 
@@ -80,7 +80,7 @@
 	p += sprintf(p, "%8s %8s\n", "cam:", "ram:");
 	p += sprintf(p, "-----------------------------------------\n");
 	rest = count - (p - buf);
-	p += dump_tlb_entries(obj, p, rest);
+	p += omap_dump_tlb_entries(obj, p, rest);
 
 	bytes = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
 
@@ -96,7 +96,7 @@
 	struct iotlb_entry e;
 	struct cr_regs cr;
 	int err;
-	struct iommu *obj = file->private_data;
+	struct omap_iommu *obj = file->private_data;
 	char buf[MAXCOLUMN], *p = buf;
 
 	count = min(count, sizeof(buf));
@@ -113,8 +113,8 @@
 		return -EINVAL;
 	}
 
-	iotlb_cr_to_e(&cr, &e);
-	err = iopgtable_store_entry(obj, &e);
+	omap_iotlb_cr_to_e(&cr, &e);
+	err = omap_iopgtable_store_entry(obj, &e);
 	if (err)
 		dev_err(obj->dev, "%s: fail to store cr\n", __func__);
 
@@ -136,7 +136,7 @@
 		__err;						\
 	})
 
-static ssize_t dump_ioptable(struct iommu *obj, char *buf, ssize_t len)
+static ssize_t dump_ioptable(struct omap_iommu *obj, char *buf, ssize_t len)
 {
 	int i;
 	u32 *iopgd;
@@ -183,7 +183,7 @@
 static ssize_t debug_read_pagetable(struct file *file, char __user *userbuf,
 				    size_t count, loff_t *ppos)
 {
-	struct iommu *obj = file->private_data;
+	struct omap_iommu *obj = file->private_data;
 	char *p, *buf;
 	size_t bytes;
 
@@ -211,7 +211,7 @@
 static ssize_t debug_read_mmap(struct file *file, char __user *userbuf,
 			       size_t count, loff_t *ppos)
 {
-	struct iommu *obj = file->private_data;
+	struct omap_iommu *obj = file->private_data;
 	char *p, *buf;
 	struct iovm_struct *tmp;
 	int uninitialized_var(i);
@@ -253,7 +253,7 @@
 static ssize_t debug_read_mem(struct file *file, char __user *userbuf,
 			      size_t count, loff_t *ppos)
 {
-	struct iommu *obj = file->private_data;
+	struct omap_iommu *obj = file->private_data;
 	char *p, *buf;
 	struct iovm_struct *area;
 	ssize_t bytes;
@@ -267,7 +267,7 @@
 
 	mutex_lock(&iommu_debug_lock);
 
-	area = find_iovm_area(obj, (u32)ppos);
+	area = omap_find_iovm_area(obj, (u32)ppos);
 	if (IS_ERR(area)) {
 		bytes = -EINVAL;
 		goto err_out;
@@ -286,7 +286,7 @@
 static ssize_t debug_write_mem(struct file *file, const char __user *userbuf,
 			       size_t count, loff_t *ppos)
 {
-	struct iommu *obj = file->private_data;
+	struct omap_iommu *obj = file->private_data;
 	struct iovm_struct *area;
 	char *p, *buf;
 
@@ -304,7 +304,7 @@
 		goto err_out;
 	}
 
-	area = find_iovm_area(obj, (u32)ppos);
+	area = omap_find_iovm_area(obj, (u32)ppos);
 	if (IS_ERR(area)) {
 		count = -EINVAL;
 		goto err_out;
@@ -360,7 +360,7 @@
 static int iommu_debug_register(struct device *dev, void *data)
 {
 	struct platform_device *pdev = to_platform_device(dev);
-	struct iommu *obj = platform_get_drvdata(pdev);
+	struct omap_iommu *obj = platform_get_drvdata(pdev);
 	struct dentry *d, *parent;
 
 	if (!obj || !obj->dev)
@@ -396,7 +396,7 @@
 		return -ENOMEM;
 	iommu_debug_root = d;
 
-	err = foreach_iommu_device(d, iommu_debug_register);
+	err = omap_foreach_iommu_device(d, iommu_debug_register);
 	if (err)
 		goto err_out;
 	return 0;
diff --git a/arch/arm/plat-omap/iommu.c b/drivers/iommu/omap-iommu.c
similarity index 60%
rename from arch/arm/plat-omap/iommu.c
rename to drivers/iommu/omap-iommu.c
index 34fc31e..8f32b2b 100644
--- a/arch/arm/plat-omap/iommu.c
+++ b/drivers/iommu/omap-iommu.c
@@ -18,18 +18,34 @@
 #include <linux/ioport.h>
 #include <linux/clk.h>
 #include <linux/platform_device.h>
+#include <linux/iommu.h>
+#include <linux/mutex.h>
+#include <linux/spinlock.h>
 
 #include <asm/cacheflush.h>
 
 #include <plat/iommu.h>
 
-#include "iopgtable.h"
+#include <plat/iopgtable.h>
 
 #define for_each_iotlb_cr(obj, n, __i, cr)				\
 	for (__i = 0;							\
 	     (__i < (n)) && (cr = __iotlb_read_cr((obj), __i), true);	\
 	     __i++)
 
+/**
+ * struct omap_iommu_domain - omap iommu domain
+ * @pgtable:	the page table
+ * @iommu_dev:	an omap iommu device attached to this domain. only a single
+ *		iommu device can be attached for now.
+ * @lock:	domain lock, should be taken when attaching/detaching
+ */
+struct omap_iommu_domain {
+	u32 *pgtable;
+	struct omap_iommu *iommu_dev;
+	spinlock_t lock;
+};
+
 /* accommodate the difference between omap1 and omap2/3 */
 static const struct iommu_functions *arch_iommu;
 
@@ -37,13 +53,13 @@
 static struct kmem_cache *iopte_cachep;
 
 /**
- * install_iommu_arch - Install archtecure specific iommu functions
+ * omap_install_iommu_arch - Install archtecure specific iommu functions
  * @ops:	a pointer to architecture specific iommu functions
  *
  * There are several kind of iommu algorithm(tlb, pagetable) among
  * omap series. This interface installs such an iommu algorighm.
  **/
-int install_iommu_arch(const struct iommu_functions *ops)
+int omap_install_iommu_arch(const struct iommu_functions *ops)
 {
 	if (arch_iommu)
 		return -EBUSY;
@@ -51,53 +67,53 @@
 	arch_iommu = ops;
 	return 0;
 }
-EXPORT_SYMBOL_GPL(install_iommu_arch);
+EXPORT_SYMBOL_GPL(omap_install_iommu_arch);
 
 /**
- * uninstall_iommu_arch - Uninstall archtecure specific iommu functions
+ * omap_uninstall_iommu_arch - Uninstall archtecure specific iommu functions
  * @ops:	a pointer to architecture specific iommu functions
  *
  * This interface uninstalls the iommu algorighm installed previously.
  **/
-void uninstall_iommu_arch(const struct iommu_functions *ops)
+void omap_uninstall_iommu_arch(const struct iommu_functions *ops)
 {
 	if (arch_iommu != ops)
 		pr_err("%s: not your arch\n", __func__);
 
 	arch_iommu = NULL;
 }
-EXPORT_SYMBOL_GPL(uninstall_iommu_arch);
+EXPORT_SYMBOL_GPL(omap_uninstall_iommu_arch);
 
 /**
- * iommu_save_ctx - Save registers for pm off-mode support
+ * omap_iommu_save_ctx - Save registers for pm off-mode support
  * @obj:	target iommu
  **/
-void iommu_save_ctx(struct iommu *obj)
+void omap_iommu_save_ctx(struct omap_iommu *obj)
 {
 	arch_iommu->save_ctx(obj);
 }
-EXPORT_SYMBOL_GPL(iommu_save_ctx);
+EXPORT_SYMBOL_GPL(omap_iommu_save_ctx);
 
 /**
- * iommu_restore_ctx - Restore registers for pm off-mode support
+ * omap_iommu_restore_ctx - Restore registers for pm off-mode support
  * @obj:	target iommu
  **/
-void iommu_restore_ctx(struct iommu *obj)
+void omap_iommu_restore_ctx(struct omap_iommu *obj)
 {
 	arch_iommu->restore_ctx(obj);
 }
-EXPORT_SYMBOL_GPL(iommu_restore_ctx);
+EXPORT_SYMBOL_GPL(omap_iommu_restore_ctx);
 
 /**
- * iommu_arch_version - Return running iommu arch version
+ * omap_iommu_arch_version - Return running iommu arch version
  **/
-u32 iommu_arch_version(void)
+u32 omap_iommu_arch_version(void)
 {
 	return arch_iommu->version;
 }
-EXPORT_SYMBOL_GPL(iommu_arch_version);
+EXPORT_SYMBOL_GPL(omap_iommu_arch_version);
 
-static int iommu_enable(struct iommu *obj)
+static int iommu_enable(struct omap_iommu *obj)
 {
 	int err;
 
@@ -115,7 +131,7 @@
 	return err;
 }
 
-static void iommu_disable(struct iommu *obj)
+static void iommu_disable(struct omap_iommu *obj)
 {
 	if (!obj)
 		return;
@@ -130,13 +146,13 @@
 /*
  *	TLB operations
  */
-void iotlb_cr_to_e(struct cr_regs *cr, struct iotlb_entry *e)
+void omap_iotlb_cr_to_e(struct cr_regs *cr, struct iotlb_entry *e)
 {
 	BUG_ON(!cr || !e);
 
 	arch_iommu->cr_to_e(cr, e);
 }
-EXPORT_SYMBOL_GPL(iotlb_cr_to_e);
+EXPORT_SYMBOL_GPL(omap_iotlb_cr_to_e);
 
 static inline int iotlb_cr_valid(struct cr_regs *cr)
 {
@@ -146,7 +162,7 @@
 	return arch_iommu->cr_valid(cr);
 }
 
-static inline struct cr_regs *iotlb_alloc_cr(struct iommu *obj,
+static inline struct cr_regs *iotlb_alloc_cr(struct omap_iommu *obj,
 					     struct iotlb_entry *e)
 {
 	if (!e)
@@ -155,23 +171,22 @@
 	return arch_iommu->alloc_cr(obj, e);
 }
 
-u32 iotlb_cr_to_virt(struct cr_regs *cr)
+static u32 iotlb_cr_to_virt(struct cr_regs *cr)
 {
 	return arch_iommu->cr_to_virt(cr);
 }
-EXPORT_SYMBOL_GPL(iotlb_cr_to_virt);
 
 static u32 get_iopte_attr(struct iotlb_entry *e)
 {
 	return arch_iommu->get_pte_attr(e);
 }
 
-static u32 iommu_report_fault(struct iommu *obj, u32 *da)
+static u32 iommu_report_fault(struct omap_iommu *obj, u32 *da)
 {
 	return arch_iommu->fault_isr(obj, da);
 }
 
-static void iotlb_lock_get(struct iommu *obj, struct iotlb_lock *l)
+static void iotlb_lock_get(struct omap_iommu *obj, struct iotlb_lock *l)
 {
 	u32 val;
 
@@ -182,7 +197,7 @@
 
 }
 
-static void iotlb_lock_set(struct iommu *obj, struct iotlb_lock *l)
+static void iotlb_lock_set(struct omap_iommu *obj, struct iotlb_lock *l)
 {
 	u32 val;
 
@@ -192,12 +207,12 @@
 	iommu_write_reg(obj, val, MMU_LOCK);
 }
 
-static void iotlb_read_cr(struct iommu *obj, struct cr_regs *cr)
+static void iotlb_read_cr(struct omap_iommu *obj, struct cr_regs *cr)
 {
 	arch_iommu->tlb_read_cr(obj, cr);
 }
 
-static void iotlb_load_cr(struct iommu *obj, struct cr_regs *cr)
+static void iotlb_load_cr(struct omap_iommu *obj, struct cr_regs *cr)
 {
 	arch_iommu->tlb_load_cr(obj, cr);
 
@@ -211,7 +226,7 @@
  * @cr:		contents of cam and ram register
  * @buf:	output buffer
  **/
-static inline ssize_t iotlb_dump_cr(struct iommu *obj, struct cr_regs *cr,
+static inline ssize_t iotlb_dump_cr(struct omap_iommu *obj, struct cr_regs *cr,
 				    char *buf)
 {
 	BUG_ON(!cr || !buf);
@@ -220,7 +235,7 @@
 }
 
 /* only used in iotlb iteration for-loop */
-static struct cr_regs __iotlb_read_cr(struct iommu *obj, int n)
+static struct cr_regs __iotlb_read_cr(struct omap_iommu *obj, int n)
 {
 	struct cr_regs cr;
 	struct iotlb_lock l;
@@ -238,7 +253,8 @@
  * @obj:	target iommu
  * @e:		an iommu tlb entry info
  **/
-int load_iotlb_entry(struct iommu *obj, struct iotlb_entry *e)
+#ifdef PREFETCH_IOTLB
+static int load_iotlb_entry(struct omap_iommu *obj, struct iotlb_entry *e)
 {
 	int err = 0;
 	struct iotlb_lock l;
@@ -294,7 +310,20 @@
 	clk_disable(obj->clk);
 	return err;
 }
-EXPORT_SYMBOL_GPL(load_iotlb_entry);
+
+#else /* !PREFETCH_IOTLB */
+
+static int load_iotlb_entry(struct omap_iommu *obj, struct iotlb_entry *e)
+{
+	return 0;
+}
+
+#endif /* !PREFETCH_IOTLB */
+
+static int prefetch_iotlb_entry(struct omap_iommu *obj, struct iotlb_entry *e)
+{
+	return load_iotlb_entry(obj, e);
+}
 
 /**
  * flush_iotlb_page - Clear an iommu tlb entry
@@ -303,7 +332,7 @@
  *
  * Clear an iommu tlb entry which includes 'da' address.
  **/
-void flush_iotlb_page(struct iommu *obj, u32 da)
+static void flush_iotlb_page(struct omap_iommu *obj, u32 da)
 {
 	int i;
 	struct cr_regs cr;
@@ -332,33 +361,12 @@
 	if (i == obj->nr_tlb_entries)
 		dev_dbg(obj->dev, "%s: no page for %08x\n", __func__, da);
 }
-EXPORT_SYMBOL_GPL(flush_iotlb_page);
-
-/**
- * flush_iotlb_range - Clear an iommu tlb entries
- * @obj:	target iommu
- * @start:	iommu device virtual address(start)
- * @end:	iommu device virtual address(end)
- *
- * Clear an iommu tlb entry which includes 'da' address.
- **/
-void flush_iotlb_range(struct iommu *obj, u32 start, u32 end)
-{
-	u32 da = start;
-
-	while (da < end) {
-		flush_iotlb_page(obj, da);
-		/* FIXME: Optimize for multiple page size */
-		da += IOPTE_SIZE;
-	}
-}
-EXPORT_SYMBOL_GPL(flush_iotlb_range);
 
 /**
  * flush_iotlb_all - Clear all iommu tlb entries
  * @obj:	target iommu
  **/
-void flush_iotlb_all(struct iommu *obj)
+static void flush_iotlb_all(struct omap_iommu *obj)
 {
 	struct iotlb_lock l;
 
@@ -372,28 +380,10 @@
 
 	clk_disable(obj->clk);
 }
-EXPORT_SYMBOL_GPL(flush_iotlb_all);
 
-/**
- * iommu_set_twl - enable/disable table walking logic
- * @obj:	target iommu
- * @on:		enable/disable
- *
- * Function used to enable/disable TWL. If one wants to work
- * exclusively with locked TLB entries and receive notifications
- * for TLB miss then call this function to disable TWL.
- */
-void iommu_set_twl(struct iommu *obj, bool on)
-{
-	clk_enable(obj->clk);
-	arch_iommu->set_twl(obj, on);
-	clk_disable(obj->clk);
-}
-EXPORT_SYMBOL_GPL(iommu_set_twl);
+#if defined(CONFIG_OMAP_IOMMU_DEBUG) || defined(CONFIG_OMAP_IOMMU_DEBUG_MODULE)
 
-#if defined(CONFIG_OMAP_IOMMU_DEBUG_MODULE)
-
-ssize_t iommu_dump_ctx(struct iommu *obj, char *buf, ssize_t bytes)
+ssize_t omap_iommu_dump_ctx(struct omap_iommu *obj, char *buf, ssize_t bytes)
 {
 	if (!obj || !buf)
 		return -EINVAL;
@@ -406,9 +396,10 @@
 
 	return bytes;
 }
-EXPORT_SYMBOL_GPL(iommu_dump_ctx);
+EXPORT_SYMBOL_GPL(omap_iommu_dump_ctx);
 
-static int __dump_tlb_entries(struct iommu *obj, struct cr_regs *crs, int num)
+static int
+__dump_tlb_entries(struct omap_iommu *obj, struct cr_regs *crs, int num)
 {
 	int i;
 	struct iotlb_lock saved;
@@ -431,11 +422,11 @@
 }
 
 /**
- * dump_tlb_entries - dump cr arrays to given buffer
+ * omap_dump_tlb_entries - dump cr arrays to given buffer
  * @obj:	target iommu
  * @buf:	output buffer
  **/
-size_t dump_tlb_entries(struct iommu *obj, char *buf, ssize_t bytes)
+size_t omap_dump_tlb_entries(struct omap_iommu *obj, char *buf, ssize_t bytes)
 {
 	int i, num;
 	struct cr_regs *cr;
@@ -455,14 +446,14 @@
 
 	return p - buf;
 }
-EXPORT_SYMBOL_GPL(dump_tlb_entries);
+EXPORT_SYMBOL_GPL(omap_dump_tlb_entries);
 
-int foreach_iommu_device(void *data, int (*fn)(struct device *, void *))
+int omap_foreach_iommu_device(void *data, int (*fn)(struct device *, void *))
 {
 	return driver_for_each_device(&omap_iommu_driver.driver,
 				      NULL, data, fn);
 }
-EXPORT_SYMBOL_GPL(foreach_iommu_device);
+EXPORT_SYMBOL_GPL(omap_foreach_iommu_device);
 
 #endif /* CONFIG_OMAP_IOMMU_DEBUG_MODULE */
 
@@ -495,7 +486,7 @@
 	kmem_cache_free(iopte_cachep, iopte);
 }
 
-static u32 *iopte_alloc(struct iommu *obj, u32 *iopgd, u32 da)
+static u32 *iopte_alloc(struct omap_iommu *obj, u32 *iopgd, u32 da)
 {
 	u32 *iopte;
 
@@ -533,7 +524,7 @@
 	return iopte;
 }
 
-static int iopgd_alloc_section(struct iommu *obj, u32 da, u32 pa, u32 prot)
+static int iopgd_alloc_section(struct omap_iommu *obj, u32 da, u32 pa, u32 prot)
 {
 	u32 *iopgd = iopgd_offset(obj, da);
 
@@ -548,7 +539,7 @@
 	return 0;
 }
 
-static int iopgd_alloc_super(struct iommu *obj, u32 da, u32 pa, u32 prot)
+static int iopgd_alloc_super(struct omap_iommu *obj, u32 da, u32 pa, u32 prot)
 {
 	u32 *iopgd = iopgd_offset(obj, da);
 	int i;
@@ -565,7 +556,7 @@
 	return 0;
 }
 
-static int iopte_alloc_page(struct iommu *obj, u32 da, u32 pa, u32 prot)
+static int iopte_alloc_page(struct omap_iommu *obj, u32 da, u32 pa, u32 prot)
 {
 	u32 *iopgd = iopgd_offset(obj, da);
 	u32 *iopte = iopte_alloc(obj, iopgd, da);
@@ -582,7 +573,7 @@
 	return 0;
 }
 
-static int iopte_alloc_large(struct iommu *obj, u32 da, u32 pa, u32 prot)
+static int iopte_alloc_large(struct omap_iommu *obj, u32 da, u32 pa, u32 prot)
 {
 	u32 *iopgd = iopgd_offset(obj, da);
 	u32 *iopte = iopte_alloc(obj, iopgd, da);
@@ -603,9 +594,10 @@
 	return 0;
 }
 
-static int iopgtable_store_entry_core(struct iommu *obj, struct iotlb_entry *e)
+static int
+iopgtable_store_entry_core(struct omap_iommu *obj, struct iotlb_entry *e)
 {
-	int (*fn)(struct iommu *, u32, u32, u32);
+	int (*fn)(struct omap_iommu *, u32, u32, u32);
 	u32 prot;
 	int err;
 
@@ -641,23 +633,21 @@
 }
 
 /**
- * iopgtable_store_entry - Make an iommu pte entry
+ * omap_iopgtable_store_entry - Make an iommu pte entry
  * @obj:	target iommu
  * @e:		an iommu tlb entry info
  **/
-int iopgtable_store_entry(struct iommu *obj, struct iotlb_entry *e)
+int omap_iopgtable_store_entry(struct omap_iommu *obj, struct iotlb_entry *e)
 {
 	int err;
 
 	flush_iotlb_page(obj, e->da);
 	err = iopgtable_store_entry_core(obj, e);
-#ifdef PREFETCH_IOTLB
 	if (!err)
-		load_iotlb_entry(obj, e);
-#endif
+		prefetch_iotlb_entry(obj, e);
 	return err;
 }
-EXPORT_SYMBOL_GPL(iopgtable_store_entry);
+EXPORT_SYMBOL_GPL(omap_iopgtable_store_entry);
 
 /**
  * iopgtable_lookup_entry - Lookup an iommu pte entry
@@ -666,7 +656,8 @@
  * @ppgd:	iommu pgd entry pointer to be returned
  * @ppte:	iommu pte entry pointer to be returned
  **/
-void iopgtable_lookup_entry(struct iommu *obj, u32 da, u32 **ppgd, u32 **ppte)
+static void
+iopgtable_lookup_entry(struct omap_iommu *obj, u32 da, u32 **ppgd, u32 **ppte)
 {
 	u32 *iopgd, *iopte = NULL;
 
@@ -680,9 +671,8 @@
 	*ppgd = iopgd;
 	*ppte = iopte;
 }
-EXPORT_SYMBOL_GPL(iopgtable_lookup_entry);
 
-static size_t iopgtable_clear_entry_core(struct iommu *obj, u32 da)
+static size_t iopgtable_clear_entry_core(struct omap_iommu *obj, u32 da)
 {
 	size_t bytes;
 	u32 *iopgd = iopgd_offset(obj, da);
@@ -735,7 +725,7 @@
  * @obj:	target iommu
  * @da:		iommu device virtual address
  **/
-size_t iopgtable_clear_entry(struct iommu *obj, u32 da)
+static size_t iopgtable_clear_entry(struct omap_iommu *obj, u32 da)
 {
 	size_t bytes;
 
@@ -748,9 +738,8 @@
 
 	return bytes;
 }
-EXPORT_SYMBOL_GPL(iopgtable_clear_entry);
 
-static void iopgtable_clear_entry_all(struct iommu *obj)
+static void iopgtable_clear_entry_all(struct omap_iommu *obj)
 {
 	int i;
 
@@ -785,7 +774,8 @@
 {
 	u32 da, errs;
 	u32 *iopgd, *iopte;
-	struct iommu *obj = data;
+	struct omap_iommu *obj = data;
+	struct iommu_domain *domain = obj->domain;
 
 	if (!obj->refcount)
 		return IRQ_NONE;
@@ -797,7 +787,7 @@
 		return IRQ_HANDLED;
 
 	/* Fault callback or TLB/PTE Dynamic loading */
-	if (obj->isr && !obj->isr(obj, da, errs, obj->isr_priv))
+	if (!report_iommu_fault(domain, obj->dev, da, 0))
 		return IRQ_HANDLED;
 
 	iommu_disable(obj);
@@ -821,7 +811,7 @@
 
 static int device_match_by_alias(struct device *dev, void *data)
 {
-	struct iommu *obj = to_iommu(dev);
+	struct omap_iommu *obj = to_iommu(dev);
 	const char *name = data;
 
 	pr_debug("%s: %s %s\n", __func__, obj->name, name);
@@ -830,57 +820,55 @@
 }
 
 /**
- * iommu_set_da_range - Set a valid device address range
- * @obj:		target iommu
- * @start		Start of valid range
- * @end			End of valid range
- **/
-int iommu_set_da_range(struct iommu *obj, u32 start, u32 end)
+ * omap_find_iommu_device() - find an omap iommu device by name
+ * @name:	name of the iommu device
+ *
+ * The generic iommu API requires the caller to provide the device
+ * he wishes to attach to a certain iommu domain.
+ *
+ * Drivers generally should not bother with this as it should just
+ * be taken care of by the DMA-API using dev_archdata.
+ *
+ * This function is provided as an interim solution until the latter
+ * materializes, and omap3isp is fully migrated to the DMA-API.
+ */
+struct device *omap_find_iommu_device(const char *name)
 {
-
-	if (!obj)
-		return -EFAULT;
-
-	if (end < start || !PAGE_ALIGN(start | end))
-		return -EINVAL;
-
-	obj->da_start = start;
-	obj->da_end = end;
-
-	return 0;
+	return driver_find_device(&omap_iommu_driver.driver, NULL,
+				(void *)name,
+				device_match_by_alias);
 }
-EXPORT_SYMBOL_GPL(iommu_set_da_range);
+EXPORT_SYMBOL_GPL(omap_find_iommu_device);
 
 /**
- * iommu_get - Get iommu handler
- * @name:	target iommu name
+ * omap_iommu_attach() - attach iommu device to an iommu domain
+ * @dev:	target omap iommu device
+ * @iopgd:	page table
  **/
-struct iommu *iommu_get(const char *name)
+static struct omap_iommu *omap_iommu_attach(struct device *dev, u32 *iopgd)
 {
 	int err = -ENOMEM;
-	struct device *dev;
-	struct iommu *obj;
+	struct omap_iommu *obj = to_iommu(dev);
 
-	dev = driver_find_device(&omap_iommu_driver.driver, NULL, (void *)name,
-				 device_match_by_alias);
-	if (!dev)
-		return ERR_PTR(-ENODEV);
+	spin_lock(&obj->iommu_lock);
 
-	obj = to_iommu(dev);
-
-	mutex_lock(&obj->iommu_lock);
-
-	if (obj->refcount++ == 0) {
-		err = iommu_enable(obj);
-		if (err)
-			goto err_enable;
-		flush_iotlb_all(obj);
+	/* an iommu device can only be attached once */
+	if (++obj->refcount > 1) {
+		dev_err(dev, "%s: already attached!\n", obj->name);
+		err = -EBUSY;
+		goto err_enable;
 	}
 
+	obj->iopgd = iopgd;
+	err = iommu_enable(obj);
+	if (err)
+		goto err_enable;
+	flush_iotlb_all(obj);
+
 	if (!try_module_get(obj->owner))
 		goto err_module;
 
-	mutex_unlock(&obj->iommu_lock);
+	spin_unlock(&obj->iommu_lock);
 
 	dev_dbg(obj->dev, "%s: %s\n", __func__, obj->name);
 	return obj;
@@ -890,59 +878,32 @@
 		iommu_disable(obj);
 err_enable:
 	obj->refcount--;
-	mutex_unlock(&obj->iommu_lock);
+	spin_unlock(&obj->iommu_lock);
 	return ERR_PTR(err);
 }
-EXPORT_SYMBOL_GPL(iommu_get);
 
 /**
- * iommu_put - Put back iommu handler
+ * omap_iommu_detach - release iommu device
  * @obj:	target iommu
  **/
-void iommu_put(struct iommu *obj)
+static void omap_iommu_detach(struct omap_iommu *obj)
 {
 	if (!obj || IS_ERR(obj))
 		return;
 
-	mutex_lock(&obj->iommu_lock);
+	spin_lock(&obj->iommu_lock);
 
 	if (--obj->refcount == 0)
 		iommu_disable(obj);
 
 	module_put(obj->owner);
 
-	mutex_unlock(&obj->iommu_lock);
+	obj->iopgd = NULL;
+
+	spin_unlock(&obj->iommu_lock);
 
 	dev_dbg(obj->dev, "%s: %s\n", __func__, obj->name);
 }
-EXPORT_SYMBOL_GPL(iommu_put);
-
-int iommu_set_isr(const char *name,
-		  int (*isr)(struct iommu *obj, u32 da, u32 iommu_errs,
-			     void *priv),
-		  void *isr_priv)
-{
-	struct device *dev;
-	struct iommu *obj;
-
-	dev = driver_find_device(&omap_iommu_driver.driver, NULL, (void *)name,
-				 device_match_by_alias);
-	if (!dev)
-		return -ENODEV;
-
-	obj = to_iommu(dev);
-	mutex_lock(&obj->iommu_lock);
-	if (obj->refcount != 0) {
-		mutex_unlock(&obj->iommu_lock);
-		return -EBUSY;
-	}
-	obj->isr = isr;
-	obj->isr_priv = isr_priv;
-	mutex_unlock(&obj->iommu_lock);
-
-	return 0;
-}
-EXPORT_SYMBOL_GPL(iommu_set_isr);
 
 /*
  *	OMAP Device MMU(IOMMU) detection
@@ -950,9 +911,8 @@
 static int __devinit omap_iommu_probe(struct platform_device *pdev)
 {
 	int err = -ENODEV;
-	void *p;
 	int irq;
-	struct iommu *obj;
+	struct omap_iommu *obj;
 	struct resource *res;
 	struct iommu_platform_data *pdata = pdev->dev.platform_data;
 
@@ -974,7 +934,7 @@
 	obj->da_start = pdata->da_start;
 	obj->da_end = pdata->da_end;
 
-	mutex_init(&obj->iommu_lock);
+	spin_lock_init(&obj->iommu_lock);
 	mutex_init(&obj->mmap_lock);
 	spin_lock_init(&obj->page_table_lock);
 	INIT_LIST_HEAD(&obj->mmap);
@@ -1009,22 +969,9 @@
 		goto err_irq;
 	platform_set_drvdata(pdev, obj);
 
-	p = (void *)__get_free_pages(GFP_KERNEL, get_order(IOPGD_TABLE_SIZE));
-	if (!p) {
-		err = -ENOMEM;
-		goto err_pgd;
-	}
-	memset(p, 0, IOPGD_TABLE_SIZE);
-	clean_dcache_area(p, IOPGD_TABLE_SIZE);
-	obj->iopgd = p;
-
-	BUG_ON(!IS_ALIGNED((unsigned long)obj->iopgd, IOPGD_TABLE_SIZE));
-
 	dev_info(&pdev->dev, "%s registered\n", obj->name);
 	return 0;
 
-err_pgd:
-	free_irq(irq, obj);
 err_irq:
 	iounmap(obj->regbase);
 err_ioremap:
@@ -1040,12 +987,11 @@
 {
 	int irq;
 	struct resource *res;
-	struct iommu *obj = platform_get_drvdata(pdev);
+	struct omap_iommu *obj = platform_get_drvdata(pdev);
 
 	platform_set_drvdata(pdev, NULL);
 
 	iopgtable_clear_entry_all(obj);
-	free_pages((unsigned long)obj->iopgd, get_order(IOPGD_TABLE_SIZE));
 
 	irq = platform_get_irq(pdev, 0);
 	free_irq(irq, obj);
@@ -1072,6 +1018,201 @@
 	clean_dcache_area(iopte, IOPTE_TABLE_SIZE);
 }
 
+static int omap_iommu_map(struct iommu_domain *domain, unsigned long da,
+			 phys_addr_t pa, int order, int prot)
+{
+	struct omap_iommu_domain *omap_domain = domain->priv;
+	struct omap_iommu *oiommu = omap_domain->iommu_dev;
+	struct device *dev = oiommu->dev;
+	size_t bytes = PAGE_SIZE << order;
+	struct iotlb_entry e;
+	int omap_pgsz;
+	u32 ret, flags;
+
+	/* we only support mapping a single iommu page for now */
+	omap_pgsz = bytes_to_iopgsz(bytes);
+	if (omap_pgsz < 0) {
+		dev_err(dev, "invalid size to map: %d\n", bytes);
+		return -EINVAL;
+	}
+
+	dev_dbg(dev, "mapping da 0x%lx to pa 0x%x size 0x%x\n", da, pa, bytes);
+
+	flags = omap_pgsz | prot;
+
+	iotlb_init_entry(&e, da, pa, flags);
+
+	ret = omap_iopgtable_store_entry(oiommu, &e);
+	if (ret)
+		dev_err(dev, "omap_iopgtable_store_entry failed: %d\n", ret);
+
+	return ret;
+}
+
+static int omap_iommu_unmap(struct iommu_domain *domain, unsigned long da,
+			    int order)
+{
+	struct omap_iommu_domain *omap_domain = domain->priv;
+	struct omap_iommu *oiommu = omap_domain->iommu_dev;
+	struct device *dev = oiommu->dev;
+	size_t unmap_size;
+
+	dev_dbg(dev, "unmapping da 0x%lx order %d\n", da, order);
+
+	unmap_size = iopgtable_clear_entry(oiommu, da);
+
+	return unmap_size ? get_order(unmap_size) : -EINVAL;
+}
+
+static int
+omap_iommu_attach_dev(struct iommu_domain *domain, struct device *dev)
+{
+	struct omap_iommu_domain *omap_domain = domain->priv;
+	struct omap_iommu *oiommu;
+	int ret = 0;
+
+	spin_lock(&omap_domain->lock);
+
+	/* only a single device is supported per domain for now */
+	if (omap_domain->iommu_dev) {
+		dev_err(dev, "iommu domain is already attached\n");
+		ret = -EBUSY;
+		goto out;
+	}
+
+	/* get a handle to and enable the omap iommu */
+	oiommu = omap_iommu_attach(dev, omap_domain->pgtable);
+	if (IS_ERR(oiommu)) {
+		ret = PTR_ERR(oiommu);
+		dev_err(dev, "can't get omap iommu: %d\n", ret);
+		goto out;
+	}
+
+	omap_domain->iommu_dev = oiommu;
+	oiommu->domain = domain;
+
+out:
+	spin_unlock(&omap_domain->lock);
+	return ret;
+}
+
+static void omap_iommu_detach_dev(struct iommu_domain *domain,
+				 struct device *dev)
+{
+	struct omap_iommu_domain *omap_domain = domain->priv;
+	struct omap_iommu *oiommu = to_iommu(dev);
+
+	spin_lock(&omap_domain->lock);
+
+	/* only a single device is supported per domain for now */
+	if (omap_domain->iommu_dev != oiommu) {
+		dev_err(dev, "invalid iommu device\n");
+		goto out;
+	}
+
+	iopgtable_clear_entry_all(oiommu);
+
+	omap_iommu_detach(oiommu);
+
+	omap_domain->iommu_dev = NULL;
+
+out:
+	spin_unlock(&omap_domain->lock);
+}
+
+static int omap_iommu_domain_init(struct iommu_domain *domain)
+{
+	struct omap_iommu_domain *omap_domain;
+
+	omap_domain = kzalloc(sizeof(*omap_domain), GFP_KERNEL);
+	if (!omap_domain) {
+		pr_err("kzalloc failed\n");
+		goto out;
+	}
+
+	omap_domain->pgtable = kzalloc(IOPGD_TABLE_SIZE, GFP_KERNEL);
+	if (!omap_domain->pgtable) {
+		pr_err("kzalloc failed\n");
+		goto fail_nomem;
+	}
+
+	/*
+	 * should never fail, but please keep this around to ensure
+	 * we keep the hardware happy
+	 */
+	BUG_ON(!IS_ALIGNED((long)omap_domain->pgtable, IOPGD_TABLE_SIZE));
+
+	clean_dcache_area(omap_domain->pgtable, IOPGD_TABLE_SIZE);
+	spin_lock_init(&omap_domain->lock);
+
+	domain->priv = omap_domain;
+
+	return 0;
+
+fail_nomem:
+	kfree(omap_domain);
+out:
+	return -ENOMEM;
+}
+
+/* assume device was already detached */
+static void omap_iommu_domain_destroy(struct iommu_domain *domain)
+{
+	struct omap_iommu_domain *omap_domain = domain->priv;
+
+	domain->priv = NULL;
+
+	kfree(omap_domain->pgtable);
+	kfree(omap_domain);
+}
+
+static phys_addr_t omap_iommu_iova_to_phys(struct iommu_domain *domain,
+					  unsigned long da)
+{
+	struct omap_iommu_domain *omap_domain = domain->priv;
+	struct omap_iommu *oiommu = omap_domain->iommu_dev;
+	struct device *dev = oiommu->dev;
+	u32 *pgd, *pte;
+	phys_addr_t ret = 0;
+
+	iopgtable_lookup_entry(oiommu, da, &pgd, &pte);
+
+	if (pte) {
+		if (iopte_is_small(*pte))
+			ret = omap_iommu_translate(*pte, da, IOPTE_MASK);
+		else if (iopte_is_large(*pte))
+			ret = omap_iommu_translate(*pte, da, IOLARGE_MASK);
+		else
+			dev_err(dev, "bogus pte 0x%x", *pte);
+	} else {
+		if (iopgd_is_section(*pgd))
+			ret = omap_iommu_translate(*pgd, da, IOSECTION_MASK);
+		else if (iopgd_is_super(*pgd))
+			ret = omap_iommu_translate(*pgd, da, IOSUPER_MASK);
+		else
+			dev_err(dev, "bogus pgd 0x%x", *pgd);
+	}
+
+	return ret;
+}
+
+static int omap_iommu_domain_has_cap(struct iommu_domain *domain,
+				    unsigned long cap)
+{
+	return 0;
+}
+
+static struct iommu_ops omap_iommu_ops = {
+	.domain_init	= omap_iommu_domain_init,
+	.domain_destroy	= omap_iommu_domain_destroy,
+	.attach_dev	= omap_iommu_attach_dev,
+	.detach_dev	= omap_iommu_detach_dev,
+	.map		= omap_iommu_map,
+	.unmap		= omap_iommu_unmap,
+	.iova_to_phys	= omap_iommu_iova_to_phys,
+	.domain_has_cap	= omap_iommu_domain_has_cap,
+};
+
 static int __init omap_iommu_init(void)
 {
 	struct kmem_cache *p;
@@ -1084,6 +1225,8 @@
 		return -ENOMEM;
 	iopte_cachep = p;
 
+	bus_set_iommu(&platform_bus_type, &omap_iommu_ops);
+
 	return platform_driver_register(&omap_iommu_driver);
 }
 module_init(omap_iommu_init);
diff --git a/arch/arm/plat-omap/iovmm.c b/drivers/iommu/omap-iovmm.c
similarity index 61%
rename from arch/arm/plat-omap/iovmm.c
rename to drivers/iommu/omap-iovmm.c
index 79e7fed..e8fdb88 100644
--- a/arch/arm/plat-omap/iovmm.c
+++ b/drivers/iommu/omap-iovmm.c
@@ -15,6 +15,7 @@
 #include <linux/vmalloc.h>
 #include <linux/device.h>
 #include <linux/scatterlist.h>
+#include <linux/iommu.h>
 
 #include <asm/cacheflush.h>
 #include <asm/mach/map.h>
@@ -22,44 +23,19 @@
 #include <plat/iommu.h>
 #include <plat/iovmm.h>
 
-#include "iopgtable.h"
-
-/*
- * A device driver needs to create address mappings between:
- *
- * - iommu/device address
- * - physical address
- * - mpu virtual address
- *
- * There are 4 possible patterns for them:
- *
- *    |iova/			  mapping		iommu_		page
- *    | da	pa	va	(d)-(p)-(v)		function	type
- *  ---------------------------------------------------------------------------
- *  1 | c	c	c	 1 - 1 - 1	  _kmap() / _kunmap()	s
- *  2 | c	c,a	c	 1 - 1 - 1	_kmalloc()/ _kfree()	s
- *  3 | c	d	c	 1 - n - 1	  _vmap() / _vunmap()	s
- *  4 | c	d,a	c	 1 - n - 1	_vmalloc()/ _vfree()	n*
- *
- *
- *	'iova':	device iommu virtual address
- *	'da':	alias of 'iova'
- *	'pa':	physical address
- *	'va':	mpu virtual address
- *
- *	'c':	contiguous memory area
- *	'd':	discontiguous memory area
- *	'a':	anonymous memory allocation
- *	'()':	optional feature
- *
- *	'n':	a normal page(4KB) size is used.
- *	's':	multiple iommu superpage(16MB, 1MB, 64KB, 4KB) size is used.
- *
- *	'*':	not yet, but feasible.
- */
+#include <plat/iopgtable.h>
 
 static struct kmem_cache *iovm_area_cachep;
 
+/* return the offset of the first scatterlist entry in a sg table */
+static unsigned int sgtable_offset(const struct sg_table *sgt)
+{
+	if (!sgt || !sgt->nents)
+		return 0;
+
+	return sgt->sgl->offset;
+}
+
 /* return total bytes of sg buffers */
 static size_t sgtable_len(const struct sg_table *sgt)
 {
@@ -72,11 +48,17 @@
 	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
 		size_t bytes;
 
-		bytes = sg->length;
+		bytes = sg->length + sg->offset;
 
 		if (!iopgsz_ok(bytes)) {
-			pr_err("%s: sg[%d] not iommu pagesize(%x)\n",
-			       __func__, i, bytes);
+			pr_err("%s: sg[%d] not iommu pagesize(%u %u)\n",
+			       __func__, i, bytes, sg->offset);
+			return 0;
+		}
+
+		if (i && sg->offset) {
+			pr_err("%s: sg[%d] offset not allowed in internal "
+					"entries\n", __func__, i);
 			return 0;
 		}
 
@@ -197,8 +179,8 @@
 		u32 pa;
 		int err;
 
-		pa = sg_phys(sg);
-		bytes = sg->length;
+		pa = sg_phys(sg) - sg->offset;
+		bytes = sg->length + sg->offset;
 
 		BUG_ON(bytes != PAGE_SIZE);
 
@@ -224,7 +206,8 @@
 	vunmap(va);
 }
 
-static struct iovm_struct *__find_iovm_area(struct iommu *obj, const u32 da)
+static struct iovm_struct *__find_iovm_area(struct omap_iommu *obj,
+							const u32 da)
 {
 	struct iovm_struct *tmp;
 
@@ -246,12 +229,12 @@
 }
 
 /**
- * find_iovm_area  -  find iovma which includes @da
+ * omap_find_iovm_area  -  find iovma which includes @da
  * @da:		iommu device virtual address
  *
  * Find the existing iovma starting at @da
  */
-struct iovm_struct *find_iovm_area(struct iommu *obj, u32 da)
+struct iovm_struct *omap_find_iovm_area(struct omap_iommu *obj, u32 da)
 {
 	struct iovm_struct *area;
 
@@ -261,13 +244,13 @@
 
 	return area;
 }
-EXPORT_SYMBOL_GPL(find_iovm_area);
+EXPORT_SYMBOL_GPL(omap_find_iovm_area);
 
 /*
  * This finds the hole(area) which fits the requested address and len
  * in iovmas mmap, and returns the new allocated iovma.
  */
-static struct iovm_struct *alloc_iovm_area(struct iommu *obj, u32 da,
+static struct iovm_struct *alloc_iovm_area(struct omap_iommu *obj, u32 da,
 					   size_t bytes, u32 flags)
 {
 	struct iovm_struct *new, *tmp;
@@ -342,7 +325,7 @@
 	return new;
 }
 
-static void free_iovm_area(struct iommu *obj, struct iovm_struct *area)
+static void free_iovm_area(struct omap_iommu *obj, struct iovm_struct *area)
 {
 	size_t bytes;
 
@@ -358,14 +341,14 @@
 }
 
 /**
- * da_to_va - convert (d) to (v)
+ * omap_da_to_va - convert (d) to (v)
  * @obj:	objective iommu
  * @da:		iommu device virtual address
  * @va:		mpu virtual address
  *
  * Returns mpu virtual addr which corresponds to a given device virtual addr
  */
-void *da_to_va(struct iommu *obj, u32 da)
+void *omap_da_to_va(struct omap_iommu *obj, u32 da)
 {
 	void *va = NULL;
 	struct iovm_struct *area;
@@ -383,7 +366,7 @@
 
 	return va;
 }
-EXPORT_SYMBOL_GPL(da_to_va);
+EXPORT_SYMBOL_GPL(omap_da_to_va);
 
 static void sgtable_fill_vmalloc(struct sg_table *sgt, void *_va)
 {
@@ -397,7 +380,7 @@
 		const size_t bytes = PAGE_SIZE;
 
 		/*
-		 * iommu 'superpage' isn't supported with 'iommu_vmalloc()'
+		 * iommu 'superpage' isn't supported with 'omap_iommu_vmalloc()'
 		 */
 		pg = vmalloc_to_page(va);
 		BUG_ON(!pg);
@@ -418,74 +401,39 @@
 	BUG_ON(!sgt);
 }
 
-static void sgtable_fill_kmalloc(struct sg_table *sgt, u32 pa, u32 da,
-								size_t len)
-{
-	unsigned int i;
-	struct scatterlist *sg;
-
-	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
-		unsigned bytes;
-
-		bytes = max_alignment(da | pa);
-		bytes = min_t(unsigned, bytes, iopgsz_max(len));
-
-		BUG_ON(!iopgsz_ok(bytes));
-
-		sg_set_buf(sg, phys_to_virt(pa), bytes);
-		/*
-		 * 'pa' is cotinuous(linear).
-		 */
-		pa += bytes;
-		da += bytes;
-		len -= bytes;
-	}
-	BUG_ON(len);
-}
-
-static inline void sgtable_drain_kmalloc(struct sg_table *sgt)
-{
-	/*
-	 * Actually this is not necessary at all, just exists for
-	 * consistency of the code readability
-	 */
-	BUG_ON(!sgt);
-}
-
 /* create 'da' <-> 'pa' mapping from 'sgt' */
-static int map_iovm_area(struct iommu *obj, struct iovm_struct *new,
-			 const struct sg_table *sgt, u32 flags)
+static int map_iovm_area(struct iommu_domain *domain, struct iovm_struct *new,
+			const struct sg_table *sgt, u32 flags)
 {
 	int err;
 	unsigned int i, j;
 	struct scatterlist *sg;
 	u32 da = new->da_start;
+	int order;
 
-	if (!obj || !sgt)
+	if (!domain || !sgt)
 		return -EINVAL;
 
 	BUG_ON(!sgtable_ok(sgt));
 
 	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
 		u32 pa;
-		int pgsz;
 		size_t bytes;
-		struct iotlb_entry e;
 
-		pa = sg_phys(sg);
-		bytes = sg->length;
+		pa = sg_phys(sg) - sg->offset;
+		bytes = sg->length + sg->offset;
 
 		flags &= ~IOVMF_PGSZ_MASK;
-		pgsz = bytes_to_iopgsz(bytes);
-		if (pgsz < 0)
+
+		if (bytes_to_iopgsz(bytes) < 0)
 			goto err_out;
-		flags |= pgsz;
+
+		order = get_order(bytes);
 
 		pr_debug("%s: [%d] %08x %08x(%x)\n", __func__,
 			 i, da, pa, bytes);
 
-		iotlb_init_entry(&e, da, pa, flags);
-		err = iopgtable_store_entry(obj, &e);
+		err = iommu_map(domain, da, pa, order, flags);
 		if (err)
 			goto err_out;
 
@@ -499,9 +447,11 @@
 	for_each_sg(sgt->sgl, sg, i, j) {
 		size_t bytes;
 
-		bytes = iopgtable_clear_entry(obj, da);
+		bytes = sg->length + sg->offset;
+		order = get_order(bytes);
 
-		BUG_ON(!iopgsz_ok(bytes));
+		/* ignore failures.. we're already handling one */
+		iommu_unmap(domain, da, order);
 
 		da += bytes;
 	}
@@ -509,22 +459,31 @@
 }
 
 /* release 'da' <-> 'pa' mapping */
-static void unmap_iovm_area(struct iommu *obj, struct iovm_struct *area)
+static void unmap_iovm_area(struct iommu_domain *domain, struct omap_iommu *obj,
+						struct iovm_struct *area)
 {
 	u32 start;
 	size_t total = area->da_end - area->da_start;
+	const struct sg_table *sgt = area->sgt;
+	struct scatterlist *sg;
+	int i, err;
 
+	BUG_ON(!sgtable_ok(sgt));
 	BUG_ON((!total) || !IS_ALIGNED(total, PAGE_SIZE));
 
 	start = area->da_start;
-	while (total > 0) {
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
 		size_t bytes;
+		int order;
 
-		bytes = iopgtable_clear_entry(obj, start);
-		if (bytes == 0)
-			bytes = PAGE_SIZE;
-		else
-			dev_dbg(obj->dev, "%s: unmap %08x(%x) %08x\n",
+		bytes = sg->length + sg->offset;
+		order = get_order(bytes);
+
+		err = iommu_unmap(domain, start, order);
+		if (err < 0)
+			break;
+
+		dev_dbg(obj->dev, "%s: unmap %08x(%x) %08x\n",
 				__func__, start, bytes, area->flags);
 
 		BUG_ON(!IS_ALIGNED(bytes, PAGE_SIZE));
@@ -536,7 +495,8 @@
 }
 
 /* template function for all unmapping */
-static struct sg_table *unmap_vm_area(struct iommu *obj, const u32 da,
+static struct sg_table *unmap_vm_area(struct iommu_domain *domain,
+				      struct omap_iommu *obj, const u32 da,
 				      void (*fn)(const void *), u32 flags)
 {
 	struct sg_table *sgt = NULL;
@@ -562,7 +522,7 @@
 	}
 	sgt = (struct sg_table *)area->sgt;
 
-	unmap_iovm_area(obj, area);
+	unmap_iovm_area(domain, obj, area);
 
 	fn(area->va);
 
@@ -577,8 +537,9 @@
 	return sgt;
 }
 
-static u32 map_iommu_region(struct iommu *obj, u32 da,
-	      const struct sg_table *sgt, void *va, size_t bytes, u32 flags)
+static u32 map_iommu_region(struct iommu_domain *domain, struct omap_iommu *obj,
+				u32 da, const struct sg_table *sgt, void *va,
+				size_t bytes, u32 flags)
 {
 	int err = -ENOMEM;
 	struct iovm_struct *new;
@@ -593,7 +554,7 @@
 	new->va = va;
 	new->sgt = sgt;
 
-	if (map_iovm_area(obj, new, sgt, new->flags))
+	if (map_iovm_area(domain, new, sgt, new->flags))
 		goto err_map;
 
 	mutex_unlock(&obj->mmap_lock);
@@ -610,14 +571,16 @@
 	return err;
 }
 
-static inline u32 __iommu_vmap(struct iommu *obj, u32 da,
-		 const struct sg_table *sgt, void *va, size_t bytes, u32 flags)
+static inline u32
+__iommu_vmap(struct iommu_domain *domain, struct omap_iommu *obj,
+				u32 da, const struct sg_table *sgt,
+				void *va, size_t bytes, u32 flags)
 {
-	return map_iommu_region(obj, da, sgt, va, bytes, flags);
+	return map_iommu_region(domain, obj, da, sgt, va, bytes, flags);
 }
 
 /**
- * iommu_vmap  -  (d)-(p)-(v) address mapper
+ * omap_iommu_vmap  -  (d)-(p)-(v) address mapper
  * @obj:	objective iommu
  * @sgt:	address of scatter gather table
  * @flags:	iovma and page property
@@ -625,8 +588,8 @@
  * Creates 1-n-1 mapping with given @sgt and returns @da.
  * All @sgt element must be io page size aligned.
  */
-u32 iommu_vmap(struct iommu *obj, u32 da, const struct sg_table *sgt,
-		 u32 flags)
+u32 omap_iommu_vmap(struct iommu_domain *domain, struct omap_iommu *obj, u32 da,
+		const struct sg_table *sgt, u32 flags)
 {
 	size_t bytes;
 	void *va = NULL;
@@ -648,38 +611,41 @@
 	flags |= IOVMF_DISCONT;
 	flags |= IOVMF_MMIO;
 
-	da = __iommu_vmap(obj, da, sgt, va, bytes, flags);
+	da = __iommu_vmap(domain, obj, da, sgt, va, bytes, flags);
 	if (IS_ERR_VALUE(da))
 		vunmap_sg(va);
 
-	return da;
+	return da + sgtable_offset(sgt);
 }
-EXPORT_SYMBOL_GPL(iommu_vmap);
+EXPORT_SYMBOL_GPL(omap_iommu_vmap);
 
 /**
- * iommu_vunmap  -  release virtual mapping obtained by 'iommu_vmap()'
+ * omap_iommu_vunmap  -  release virtual mapping obtained by 'omap_iommu_vmap()'
  * @obj:	objective iommu
  * @da:		iommu device virtual address
  *
  * Free the iommu virtually contiguous memory area starting at
- * @da, which was returned by 'iommu_vmap()'.
+ * @da, which was returned by 'omap_iommu_vmap()'.
  */
-struct sg_table *iommu_vunmap(struct iommu *obj, u32 da)
+struct sg_table *
+omap_iommu_vunmap(struct iommu_domain *domain, struct omap_iommu *obj, u32 da)
 {
 	struct sg_table *sgt;
 	/*
-	 * 'sgt' is allocated before 'iommu_vmalloc()' is called.
+	 * 'sgt' is allocated before 'omap_iommu_vmalloc()' is called.
 	 * Just returns 'sgt' to the caller to free
 	 */
-	sgt = unmap_vm_area(obj, da, vunmap_sg, IOVMF_DISCONT | IOVMF_MMIO);
+	da &= PAGE_MASK;
+	sgt = unmap_vm_area(domain, obj, da, vunmap_sg,
+					IOVMF_DISCONT | IOVMF_MMIO);
 	if (!sgt)
 		dev_dbg(obj->dev, "%s: No sgt\n", __func__);
 	return sgt;
 }
-EXPORT_SYMBOL_GPL(iommu_vunmap);
+EXPORT_SYMBOL_GPL(omap_iommu_vunmap);
 
 /**
- * iommu_vmalloc  -  (d)-(p)-(v) address allocator and mapper
+ * omap_iommu_vmalloc  -  (d)-(p)-(v) address allocator and mapper
  * @obj:	objective iommu
  * @da:		contiguous iommu virtual memory
  * @bytes:	allocation size
@@ -688,7 +654,9 @@
  * Allocate @bytes linearly and creates 1-n-1 mapping and returns
  * @da again, which might be adjusted if 'IOVMF_DA_FIXED' is not set.
  */
-u32 iommu_vmalloc(struct iommu *obj, u32 da, size_t bytes, u32 flags)
+u32
+omap_iommu_vmalloc(struct iommu_domain *domain, struct omap_iommu *obj, u32 da,
+						size_t bytes, u32 flags)
 {
 	void *va;
 	struct sg_table *sgt;
@@ -712,7 +680,7 @@
 	}
 	sgtable_fill_vmalloc(sgt, va);
 
-	da = __iommu_vmap(obj, da, sgt, va, bytes, flags);
+	da = __iommu_vmap(domain, obj, da, sgt, va, bytes, flags);
 	if (IS_ERR_VALUE(da))
 		goto err_iommu_vmap;
 
@@ -725,158 +693,28 @@
 	vfree(va);
 	return da;
 }
-EXPORT_SYMBOL_GPL(iommu_vmalloc);
+EXPORT_SYMBOL_GPL(omap_iommu_vmalloc);
 
 /**
- * iommu_vfree  -  release memory allocated by 'iommu_vmalloc()'
+ * omap_iommu_vfree  -  release memory allocated by 'omap_iommu_vmalloc()'
  * @obj:	objective iommu
  * @da:		iommu device virtual address
  *
  * Frees the iommu virtually continuous memory area starting at
- * @da, as obtained from 'iommu_vmalloc()'.
+ * @da, as obtained from 'omap_iommu_vmalloc()'.
  */
-void iommu_vfree(struct iommu *obj, const u32 da)
+void omap_iommu_vfree(struct iommu_domain *domain, struct omap_iommu *obj,
+								const u32 da)
 {
 	struct sg_table *sgt;
 
-	sgt = unmap_vm_area(obj, da, vfree, IOVMF_DISCONT | IOVMF_ALLOC);
+	sgt = unmap_vm_area(domain, obj, da, vfree,
+						IOVMF_DISCONT | IOVMF_ALLOC);
 	if (!sgt)
 		dev_dbg(obj->dev, "%s: No sgt\n", __func__);
 	sgtable_free(sgt);
 }
-EXPORT_SYMBOL_GPL(iommu_vfree);
-
-static u32 __iommu_kmap(struct iommu *obj, u32 da, u32 pa, void *va,
-			  size_t bytes, u32 flags)
-{
-	struct sg_table *sgt;
-
-	sgt = sgtable_alloc(bytes, flags, da, pa);
-	if (IS_ERR(sgt))
-		return PTR_ERR(sgt);
-
-	sgtable_fill_kmalloc(sgt, pa, da, bytes);
-
-	da = map_iommu_region(obj, da, sgt, va, bytes, flags);
-	if (IS_ERR_VALUE(da)) {
-		sgtable_drain_kmalloc(sgt);
-		sgtable_free(sgt);
-	}
-
-	return da;
-}
-
-/**
- * iommu_kmap  -  (d)-(p)-(v) address mapper
- * @obj:	objective iommu
- * @da:		contiguous iommu virtual memory
- * @pa:		contiguous physical memory
- * @flags:	iovma and page property
- *
- * Creates 1-1-1 mapping and returns @da again, which can be
- * adjusted if 'IOVMF_DA_FIXED' is not set.
- */
-u32 iommu_kmap(struct iommu *obj, u32 da, u32 pa, size_t bytes,
-		 u32 flags)
-{
-	void *va;
-
-	if (!obj || !obj->dev || !bytes)
-		return -EINVAL;
-
-	bytes = PAGE_ALIGN(bytes);
-
-	va = ioremap(pa, bytes);
-	if (!va)
-		return -ENOMEM;
-
-	flags |= IOVMF_LINEAR;
-	flags |= IOVMF_MMIO;
-
-	da = __iommu_kmap(obj, da, pa, va, bytes, flags);
-	if (IS_ERR_VALUE(da))
-		iounmap(va);
-
-	return da;
-}
-EXPORT_SYMBOL_GPL(iommu_kmap);
-
-/**
- * iommu_kunmap  -  release virtual mapping obtained by 'iommu_kmap()'
- * @obj:	objective iommu
- * @da:		iommu device virtual address
- *
- * Frees the iommu virtually contiguous memory area starting at
- * @da, which was passed to and was returned by'iommu_kmap()'.
- */
-void iommu_kunmap(struct iommu *obj, u32 da)
-{
-	struct sg_table *sgt;
-	typedef void (*func_t)(const void *);
-
-	sgt = unmap_vm_area(obj, da, (func_t)iounmap,
-			    IOVMF_LINEAR | IOVMF_MMIO);
-	if (!sgt)
-		dev_dbg(obj->dev, "%s: No sgt\n", __func__);
-	sgtable_free(sgt);
-}
-EXPORT_SYMBOL_GPL(iommu_kunmap);
-
-/**
- * iommu_kmalloc  -  (d)-(p)-(v) address allocator and mapper
- * @obj:	objective iommu
- * @da:		contiguous iommu virtual memory
- * @bytes:	bytes for allocation
- * @flags:	iovma and page property
- *
- * Allocate @bytes linearly and creates 1-1-1 mapping and returns
- * @da again, which might be adjusted if 'IOVMF_DA_FIXED' is not set.
- */
-u32 iommu_kmalloc(struct iommu *obj, u32 da, size_t bytes, u32 flags)
-{
-	void *va;
-	u32 pa;
-
-	if (!obj || !obj->dev || !bytes)
-		return -EINVAL;
-
-	bytes = PAGE_ALIGN(bytes);
-
-	va = kmalloc(bytes, GFP_KERNEL | GFP_DMA);
-	if (!va)
-		return -ENOMEM;
-	pa = virt_to_phys(va);
-
-	flags |= IOVMF_LINEAR;
-	flags |= IOVMF_ALLOC;
-
-	da = __iommu_kmap(obj, da, pa, va, bytes, flags);
-	if (IS_ERR_VALUE(da))
-		kfree(va);
-
-	return da;
-}
-EXPORT_SYMBOL_GPL(iommu_kmalloc);
-
-/**
- * iommu_kfree  -  release virtual mapping obtained by 'iommu_kmalloc()'
- * @obj:	objective iommu
- * @da:		iommu device virtual address
- *
- * Frees the iommu virtually contiguous memory area starting at
- * @da, which was passed to and was returned by'iommu_kmalloc()'.
- */
-void iommu_kfree(struct iommu *obj, u32 da)
-{
-	struct sg_table *sgt;
-
-	sgt = unmap_vm_area(obj, da, kfree, IOVMF_LINEAR | IOVMF_ALLOC);
-	if (!sgt)
-		dev_dbg(obj->dev, "%s: No sgt\n", __func__);
-	sgtable_free(sgt);
-}
-EXPORT_SYMBOL_GPL(iommu_kfree);
-
+EXPORT_SYMBOL_GPL(omap_iommu_vfree);
 
 static int __init iovmm_init(void)
 {
diff --git a/drivers/media/video/Kconfig b/drivers/media/video/Kconfig
index f574dc0..6201069 100644
--- a/drivers/media/video/Kconfig
+++ b/drivers/media/video/Kconfig
@@ -763,8 +763,7 @@
 
 config VIDEO_OMAP3
 	tristate "OMAP 3 Camera support (EXPERIMENTAL)"
-	select OMAP_IOMMU
-	depends on VIDEO_V4L2 && I2C && VIDEO_V4L2_SUBDEV_API && ARCH_OMAP3 && EXPERIMENTAL
+	depends on OMAP_IOVMM && VIDEO_V4L2 && I2C && VIDEO_V4L2_SUBDEV_API && ARCH_OMAP3 && EXPERIMENTAL
 	---help---
 	  Driver for an OMAP 3 camera controller.
 
diff --git a/drivers/media/video/omap3isp/isp.c b/drivers/media/video/omap3isp/isp.c
index 5cea2bb..a7ed985 100644
--- a/drivers/media/video/omap3isp/isp.c
+++ b/drivers/media/video/omap3isp/isp.c
@@ -80,6 +80,13 @@
 #include "isph3a.h"
 #include "isphist.h"
 
+/*
+ * this is provided as an interim solution until omap3isp doesn't need
+ * any omap-specific iommu API
+ */
+#define to_iommu(dev)							\
+	(struct omap_iommu *)platform_get_drvdata(to_platform_device(dev))
+
 static unsigned int autoidle;
 module_param(autoidle, int, 0444);
 MODULE_PARM_DESC(autoidle, "Enable OMAP3ISP AUTOIDLE support");
@@ -1108,7 +1115,7 @@
 {
 	isp_save_context(isp, isp_reg_list);
 	if (isp->iommu)
-		iommu_save_ctx(isp->iommu);
+		omap_iommu_save_ctx(isp->iommu);
 }
 
 /*
@@ -1122,7 +1129,7 @@
 {
 	isp_restore_context(isp, isp_reg_list);
 	if (isp->iommu)
-		iommu_restore_ctx(isp->iommu);
+		omap_iommu_restore_ctx(isp->iommu);
 	omap3isp_ccdc_restore_context(isp);
 	omap3isp_preview_restore_context(isp);
 }
@@ -1975,7 +1982,8 @@
 	isp_cleanup_modules(isp);
 
 	omap3isp_get(isp);
-	iommu_put(isp->iommu);
+	iommu_detach_device(isp->domain, isp->iommu_dev);
+	iommu_domain_free(isp->domain);
 	omap3isp_put(isp);
 
 	free_irq(isp->irq_num, isp);
@@ -2123,25 +2131,41 @@
 	}
 
 	/* IOMMU */
-	isp->iommu = iommu_get("isp");
-	if (IS_ERR_OR_NULL(isp->iommu)) {
-		isp->iommu = NULL;
+	isp->iommu_dev = omap_find_iommu_device("isp");
+	if (!isp->iommu_dev) {
+		dev_err(isp->dev, "omap_find_iommu_device failed\n");
 		ret = -ENODEV;
 		goto error_isp;
 	}
 
+	/* to be removed once iommu migration is complete */
+	isp->iommu = to_iommu(isp->iommu_dev);
+
+	isp->domain = iommu_domain_alloc(pdev->dev.bus);
+	if (!isp->domain) {
+		dev_err(isp->dev, "can't alloc iommu domain\n");
+		ret = -ENOMEM;
+		goto error_isp;
+	}
+
+	ret = iommu_attach_device(isp->domain, isp->iommu_dev);
+	if (ret) {
+		dev_err(&pdev->dev, "can't attach iommu device: %d\n", ret);
+		goto free_domain;
+	}
+
 	/* Interrupt */
 	isp->irq_num = platform_get_irq(pdev, 0);
 	if (isp->irq_num <= 0) {
 		dev_err(isp->dev, "No IRQ resource\n");
 		ret = -ENODEV;
-		goto error_isp;
+		goto detach_dev;
 	}
 
 	if (request_irq(isp->irq_num, isp_isr, IRQF_SHARED, "OMAP3 ISP", isp)) {
 		dev_err(isp->dev, "Unable to request IRQ\n");
 		ret = -EINVAL;
-		goto error_isp;
+		goto detach_dev;
 	}
 
 	/* Entities */
@@ -2162,8 +2186,11 @@
 	isp_cleanup_modules(isp);
 error_irq:
 	free_irq(isp->irq_num, isp);
+detach_dev:
+	iommu_detach_device(isp->domain, isp->iommu_dev);
+free_domain:
+	iommu_domain_free(isp->domain);
 error_isp:
-	iommu_put(isp->iommu);
 	omap3isp_put(isp);
 error:
 	isp_put_clocks(isp);
diff --git a/drivers/media/video/omap3isp/isp.h b/drivers/media/video/omap3isp/isp.h
index 529e582..81fdd85 100644
--- a/drivers/media/video/omap3isp/isp.h
+++ b/drivers/media/video/omap3isp/isp.h
@@ -32,6 +32,7 @@
 #include <linux/io.h>
 #include <linux/platform_device.h>
 #include <linux/wait.h>
+#include <linux/iommu.h>
 #include <plat/iommu.h>
 #include <plat/iovmm.h>
 
@@ -294,7 +295,9 @@
 	unsigned int sbl_resources;
 	unsigned int subclk_resources;
 
-	struct iommu *iommu;
+	struct omap_iommu *iommu;
+	struct iommu_domain *domain;
+	struct device *iommu_dev;
 
 	struct isp_platform_callback platform_cb;
 };
diff --git a/drivers/media/video/omap3isp/ispccdc.c b/drivers/media/video/omap3isp/ispccdc.c
index 80796eb..8926719 100644
--- a/drivers/media/video/omap3isp/ispccdc.c
+++ b/drivers/media/video/omap3isp/ispccdc.c
@@ -366,7 +366,7 @@
 		dma_unmap_sg(isp->dev, req->iovm->sgt->sgl,
 			     req->iovm->sgt->nents, DMA_TO_DEVICE);
 	if (req->table)
-		iommu_vfree(isp->iommu, req->table);
+		omap_iommu_vfree(isp->domain, isp->iommu, req->table);
 	kfree(req);
 }
 
@@ -438,15 +438,15 @@
 
 		req->enable = 1;
 
-		req->table = iommu_vmalloc(isp->iommu, 0, req->config.size,
-					   IOMMU_FLAG);
+		req->table = omap_iommu_vmalloc(isp->domain, isp->iommu, 0,
+					req->config.size, IOMMU_FLAG);
 		if (IS_ERR_VALUE(req->table)) {
 			req->table = 0;
 			ret = -ENOMEM;
 			goto done;
 		}
 
-		req->iovm = find_iovm_area(isp->iommu, req->table);
+		req->iovm = omap_find_iovm_area(isp->iommu, req->table);
 		if (req->iovm == NULL) {
 			ret = -ENOMEM;
 			goto done;
@@ -462,7 +462,7 @@
 		dma_sync_sg_for_cpu(isp->dev, req->iovm->sgt->sgl,
 				    req->iovm->sgt->nents, DMA_TO_DEVICE);
 
-		table = da_to_va(isp->iommu, req->table);
+		table = omap_da_to_va(isp->iommu, req->table);
 		if (copy_from_user(table, config->lsc, req->config.size)) {
 			ret = -EFAULT;
 			goto done;
@@ -731,18 +731,19 @@
 
 			/*
 			 * table_new must be 64-bytes aligned, but it's
-			 * already done by iommu_vmalloc().
+			 * already done by omap_iommu_vmalloc().
 			 */
 			size = ccdc->fpc.fpnum * 4;
-			table_new = iommu_vmalloc(isp->iommu, 0, size,
-						  IOMMU_FLAG);
+			table_new = omap_iommu_vmalloc(isp->domain, isp->iommu,
+							0, size, IOMMU_FLAG);
 			if (IS_ERR_VALUE(table_new))
 				return -ENOMEM;
 
-			if (copy_from_user(da_to_va(isp->iommu, table_new),
+			if (copy_from_user(omap_da_to_va(isp->iommu, table_new),
 					   (__force void __user *)
 					   ccdc->fpc.fpcaddr, size)) {
-				iommu_vfree(isp->iommu, table_new);
+				omap_iommu_vfree(isp->domain, isp->iommu,
+								table_new);
 				return -EFAULT;
 			}
 
@@ -752,7 +753,7 @@
 
 		ccdc_configure_fpc(ccdc);
 		if (table_old != 0)
-			iommu_vfree(isp->iommu, table_old);
+			omap_iommu_vfree(isp->domain, isp->iommu, table_old);
 	}
 
 	return ccdc_lsc_config(ccdc, ccdc_struct);
@@ -2287,5 +2288,5 @@
 	ccdc_lsc_free_queue(ccdc, &ccdc->lsc.free_queue);
 
 	if (ccdc->fpc.fpcaddr != 0)
-		iommu_vfree(isp->iommu, ccdc->fpc.fpcaddr);
+		omap_iommu_vfree(isp->domain, isp->iommu, ccdc->fpc.fpcaddr);
 }
diff --git a/drivers/media/video/omap3isp/ispstat.c b/drivers/media/video/omap3isp/ispstat.c
index 8080659..7329055 100644
--- a/drivers/media/video/omap3isp/ispstat.c
+++ b/drivers/media/video/omap3isp/ispstat.c
@@ -366,7 +366,8 @@
 				dma_unmap_sg(isp->dev, buf->iovm->sgt->sgl,
 					     buf->iovm->sgt->nents,
 					     DMA_FROM_DEVICE);
-			iommu_vfree(isp->iommu, buf->iommu_addr);
+			omap_iommu_vfree(isp->domain, isp->iommu,
+							buf->iommu_addr);
 		} else {
 			if (!buf->virt_addr)
 				continue;
@@ -399,8 +400,8 @@
 		struct iovm_struct *iovm;
 
 		WARN_ON(buf->dma_addr);
-		buf->iommu_addr = iommu_vmalloc(isp->iommu, 0, size,
-						IOMMU_FLAG);
+		buf->iommu_addr = omap_iommu_vmalloc(isp->domain, isp->iommu, 0,
+							size, IOMMU_FLAG);
 		if (IS_ERR((void *)buf->iommu_addr)) {
 			dev_err(stat->isp->dev,
 				 "%s: Can't acquire memory for "
@@ -409,7 +410,7 @@
 			return -ENOMEM;
 		}
 
-		iovm = find_iovm_area(isp->iommu, buf->iommu_addr);
+		iovm = omap_find_iovm_area(isp->iommu, buf->iommu_addr);
 		if (!iovm ||
 		    !dma_map_sg(isp->dev, iovm->sgt->sgl, iovm->sgt->nents,
 				DMA_FROM_DEVICE)) {
@@ -418,7 +419,7 @@
 		}
 		buf->iovm = iovm;
 
-		buf->virt_addr = da_to_va(stat->isp->iommu,
+		buf->virt_addr = omap_da_to_va(stat->isp->iommu,
 					  (u32)buf->iommu_addr);
 		buf->empty = 1;
 		dev_dbg(stat->isp->dev, "%s: buffer[%d] allocated."
diff --git a/drivers/media/video/omap3isp/ispvideo.c b/drivers/media/video/omap3isp/ispvideo.c
index fd965ad..912ac07 100644
--- a/drivers/media/video/omap3isp/ispvideo.c
+++ b/drivers/media/video/omap3isp/ispvideo.c
@@ -446,7 +446,7 @@
 	sgt->nents = sglen;
 	sgt->orig_nents = sglen;
 
-	da = iommu_vmap(isp->iommu, 0, sgt, IOMMU_FLAG);
+	da = omap_iommu_vmap(isp->domain, isp->iommu, 0, sgt, IOMMU_FLAG);
 	if (IS_ERR_VALUE(da))
 		kfree(sgt);
 
@@ -462,7 +462,7 @@
 {
 	struct sg_table *sgt;
 
-	sgt = iommu_vunmap(isp->iommu, (u32)da);
+	sgt = omap_iommu_vunmap(isp->domain, isp->iommu, (u32)da);
 	kfree(sgt);
 }
 
diff --git a/include/linux/device.h b/include/linux/device.h
index c20dfbf..e838e14 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -33,6 +33,7 @@
 struct subsys_private;
 struct bus_type;
 struct device_node;
+struct iommu_ops;
 
 struct bus_attribute {
 	struct attribute	attr;
@@ -67,6 +68,9 @@
  * @resume:	Called to bring a device on this bus out of sleep mode.
  * @pm:		Power management operations of this bus, callback the specific
  *		device driver's pm-ops.
+ * @iommu_ops   IOMMU specific operations for this bus, used to attach IOMMU
+ *              driver implementations to a bus and allow the driver to do
+ *              bus-specific setup
  * @p:		The private data of the driver core, only the driver core can
  *		touch this.
  *
@@ -96,6 +100,8 @@
 
 	const struct dev_pm_ops *pm;
 
+	struct iommu_ops *iommu_ops;
+
 	struct subsys_private *p;
 };
 
diff --git a/include/linux/iommu.h b/include/linux/iommu.h
index 9940319..432acc4 100644
--- a/include/linux/iommu.h
+++ b/include/linux/iommu.h
@@ -25,15 +25,29 @@
 #define IOMMU_WRITE	(2)
 #define IOMMU_CACHE	(4) /* DMA cache coherency */
 
+struct iommu_ops;
+struct bus_type;
 struct device;
+struct iommu_domain;
+
+/* iommu fault flags */
+#define IOMMU_FAULT_READ	0x0
+#define IOMMU_FAULT_WRITE	0x1
+
+typedef int (*iommu_fault_handler_t)(struct iommu_domain *,
+				struct device *, unsigned long, int);
 
 struct iommu_domain {
+	struct iommu_ops *ops;
 	void *priv;
+	iommu_fault_handler_t handler;
 };
 
 #define IOMMU_CAP_CACHE_COHERENCY	0x1
 #define IOMMU_CAP_INTR_REMAP		0x2	/* isolates device intrs */
 
+#ifdef CONFIG_IOMMU_API
+
 struct iommu_ops {
 	int (*domain_init)(struct iommu_domain *domain);
 	void (*domain_destroy)(struct iommu_domain *domain);
@@ -49,11 +63,9 @@
 			      unsigned long cap);
 };
 
-#ifdef CONFIG_IOMMU_API
-
-extern void register_iommu(struct iommu_ops *ops);
-extern bool iommu_found(void);
-extern struct iommu_domain *iommu_domain_alloc(void);
+extern int bus_set_iommu(struct bus_type *bus, struct iommu_ops *ops);
+extern bool iommu_present(struct bus_type *bus);
+extern struct iommu_domain *iommu_domain_alloc(struct bus_type *bus);
 extern void iommu_domain_free(struct iommu_domain *domain);
 extern int iommu_attach_device(struct iommu_domain *domain,
 			       struct device *dev);
@@ -67,19 +79,58 @@
 				      unsigned long iova);
 extern int iommu_domain_has_cap(struct iommu_domain *domain,
 				unsigned long cap);
+extern void iommu_set_fault_handler(struct iommu_domain *domain,
+					iommu_fault_handler_t handler);
+
+/**
+ * report_iommu_fault() - report about an IOMMU fault to the IOMMU framework
+ * @domain: the iommu domain where the fault has happened
+ * @dev: the device where the fault has happened
+ * @iova: the faulting address
+ * @flags: mmu fault flags (e.g. IOMMU_FAULT_READ/IOMMU_FAULT_WRITE/...)
+ *
+ * This function should be called by the low-level IOMMU implementations
+ * whenever IOMMU faults happen, to allow high-level users, that are
+ * interested in such events, to know about them.
+ *
+ * This event may be useful for several possible use cases:
+ * - mere logging of the event
+ * - dynamic TLB/PTE loading
+ * - if restarting of the faulting device is required
+ *
+ * Returns 0 on success and an appropriate error code otherwise (if dynamic
+ * PTE/TLB loading will one day be supported, implementations will be able
+ * to tell whether it succeeded or not according to this return value).
+ *
+ * Specifically, -ENOSYS is returned if a fault handler isn't installed
+ * (though fault handlers can also return -ENOSYS, in case they want to
+ * elicit the default behavior of the IOMMU drivers).
+ */
+static inline int report_iommu_fault(struct iommu_domain *domain,
+		struct device *dev, unsigned long iova, int flags)
+{
+	int ret = -ENOSYS;
+
+	/*
+	 * if upper layers showed interest and installed a fault handler,
+	 * invoke it.
+	 */
+	if (domain->handler)
+		ret = domain->handler(domain, dev, iova, flags);
+
+	return ret;
+}
 
 #else /* CONFIG_IOMMU_API */
 
-static inline void register_iommu(struct iommu_ops *ops)
-{
-}
+struct iommu_ops {};
 
-static inline bool iommu_found(void)
+static inline bool iommu_present(struct bus_type *bus)
 {
 	return false;
 }
 
-static inline struct iommu_domain *iommu_domain_alloc(void)
+static inline struct iommu_domain *iommu_domain_alloc(struct bus_type *bus)
 {
 	return NULL;
 }
@@ -123,6 +174,11 @@
 	return 0;
 }
 
+static inline void iommu_set_fault_handler(struct iommu_domain *domain,
+					iommu_fault_handler_t handler)
+{
+}
+
 #endif /* CONFIG_IOMMU_API */
 
 #endif /* __LINUX_IOMMU_H */
diff --git a/lib/dma-debug.c b/lib/dma-debug.c
index db07bfd..79700fa 100644
--- a/lib/dma-debug.c
+++ b/lib/dma-debug.c
@@ -62,6 +62,8 @@
 #endif
 };
 
+typedef bool (*match_fn)(struct dma_debug_entry *, struct dma_debug_entry *);
+
 struct hash_bucket {
 	struct list_head list;
 	spinlock_t lock;
@@ -240,18 +242,37 @@
 	spin_unlock_irqrestore(&bucket->lock, __flags);
 }
 
+static bool exact_match(struct dma_debug_entry *a, struct dma_debug_entry *b)
+{
+	return ((a->dev_addr == a->dev_addr) &&
+		(a->dev == b->dev)) ? true : false;
+}
+
+static bool containing_match(struct dma_debug_entry *a,
+			     struct dma_debug_entry *b)
+{
+	if (a->dev != b->dev)
+		return false;
+
+	if ((b->dev_addr <= a->dev_addr) &&
+	    ((b->dev_addr + b->size) >= (a->dev_addr + a->size)))
+		return true;
+
+	return false;
+}
+
 /*
  * Search a given entry in the hash bucket list
  */
-static struct dma_debug_entry *hash_bucket_find(struct hash_bucket *bucket,
-						struct dma_debug_entry *ref)
+static struct dma_debug_entry *__hash_bucket_find(struct hash_bucket *bucket,
+						  struct dma_debug_entry *ref,
+						  match_fn match)
 {
 	struct dma_debug_entry *entry, *ret = NULL;
 	int matches = 0, match_lvl, last_lvl = 0;
 
 	list_for_each_entry(entry, &bucket->list, list) {
-		if ((entry->dev_addr != ref->dev_addr) ||
-		    (entry->dev != ref->dev))
+		if (!match(ref, entry))
 			continue;
 
 		/*
@@ -293,6 +314,39 @@
 	return ret;
 }
 
+static struct dma_debug_entry *bucket_find_exact(struct hash_bucket *bucket,
+						 struct dma_debug_entry *ref)
+{
+	return __hash_bucket_find(bucket, ref, exact_match);
+}
+
+static struct dma_debug_entry *bucket_find_contain(struct hash_bucket **bucket,
+						   struct dma_debug_entry *ref,
+						   unsigned long *flags)
+{
+
+	unsigned int max_range = dma_get_max_seg_size(ref->dev);
+	struct dma_debug_entry *entry, index = *ref;
+	unsigned int range = 0;
+
+	while (range <= max_range) {
+		entry = __hash_bucket_find(*bucket, &index, containing_match);
+
+		if (entry)
+			return entry;
+
+		/*
+		 * Nothing found, go back a hash bucket
+		 */
+		put_hash_bucket(*bucket, flags);
+		range          += (1 << HASH_FN_SHIFT);
+		index.dev_addr -= (1 << HASH_FN_SHIFT);
+		*bucket = get_hash_bucket(&index, flags);
+	}
+
+	return NULL;
+}
+
 /*
  * Add an entry to a hash bucket
  */
@@ -802,7 +856,7 @@
 	}
 
 	bucket = get_hash_bucket(ref, &flags);
-	entry = hash_bucket_find(bucket, ref);
+	entry = bucket_find_exact(bucket, ref);
 
 	if (!entry) {
 		err_printk(ref->dev, NULL, "DMA-API: device driver tries "
@@ -902,7 +956,7 @@
 
 	bucket = get_hash_bucket(ref, &flags);
 
-	entry = hash_bucket_find(bucket, ref);
+	entry = bucket_find_contain(&bucket, ref, &flags);
 
 	if (!entry) {
 		err_printk(dev, NULL, "DMA-API: device driver tries "
@@ -1060,7 +1114,7 @@
 	int mapped_ents;
 
 	bucket       = get_hash_bucket(ref, &flags);
-	entry        = hash_bucket_find(bucket, ref);
+	entry        = bucket_find_exact(bucket, ref);
 	mapped_ents  = 0;
 
 	if (entry)
diff --git a/virt/kvm/iommu.c b/virt/kvm/iommu.c
index 78c80f6..d149940 100644
--- a/virt/kvm/iommu.c
+++ b/virt/kvm/iommu.c
@@ -228,12 +228,12 @@
 {
 	int r;
 
-	if (!iommu_found()) {
+	if (!iommu_present(&pci_bus_type)) {
 		printk(KERN_ERR "%s: iommu not found\n", __func__);
 		return -ENODEV;
 	}
 
-	kvm->arch.iommu_domain = iommu_domain_alloc();
+	kvm->arch.iommu_domain = iommu_domain_alloc(&pci_bus_type);
 	if (!kvm->arch.iommu_domain)
 		return -ENOMEM;