Pull vtd-iommu into release branch

Conflicts:
	arch/ia64/kernel/acpi.c

arch/ia64/Kconfig

@@ -145,13 +145,15 @@ config IA64_GENERIC
 	select NUMA
 	select ACPI_NUMA
 	select SWIOTLB
+	select PCI_MSI
 	help
 	  This selects the system type of your hardware.  A "generic" kernel
 	  will run on any supported IA-64 system.  However, if you configure
 	  a kernel for your specific system, it will be faster and smaller.
 
 	  generic		For any supported IA-64 system
 	  DIG-compliant		For DIG ("Developer's Interface Guide") compliant systems
+	  DIG+Intel+IOMMU	For DIG systems with Intel IOMMU
 	  HP-zx1/sx1000		For HP systems
 	  HP-zx1/sx1000+swiotlb	For HP systems with (broken) DMA-constrained devices.
 	  SGI-SN2		For SGI Altix systems
@@ -165,6 +167,11 @@ config IA64_DIG
 	bool "DIG-compliant"
 	select SWIOTLB
 
+config IA64_DIG_VTD
+	bool "DIG+Intel+IOMMU"
+	select DMAR
+	select PCI_MSI
+
 config IA64_HP_ZX1
 	bool "HP-zx1/sx1000"
 	help
@@ -614,6 +621,16 @@ source "drivers/pci/hotplug/Kconfig"
 
 source "drivers/pcmcia/Kconfig"
 
+config DMAR
+        bool "Support for DMA Remapping Devices (EXPERIMENTAL)"
+        depends on IA64_GENERIC && ACPI && EXPERIMENTAL
+	help
+	  DMA remapping (DMAR) devices support enables independent address
+	  translations for Direct Memory Access (DMA) from devices.
+	  These DMA remapping devices are reported via ACPI tables
+	  and include PCI device scope covered by these DMA
+	  remapping devices.
+
 endmenu
 
 endif

arch/ia64/Makefile

@@ -53,6 +53,7 @@ libs-y				+= arch/ia64/lib/
 core-y				+= arch/ia64/kernel/ arch/ia64/mm/
 core-$(CONFIG_IA32_SUPPORT)	+= arch/ia64/ia32/
 core-$(CONFIG_IA64_DIG) 	+= arch/ia64/dig/
+core-$(CONFIG_IA64_DIG_VTD) 	+= arch/ia64/dig/
 core-$(CONFIG_IA64_GENERIC) 	+= arch/ia64/dig/
 core-$(CONFIG_IA64_HP_ZX1)	+= arch/ia64/dig/
 core-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += arch/ia64/dig/

arch/ia64/configs/generic_defconfig

@@ -233,6 +233,8 @@ CONFIG_DMIID=y
 CONFIG_BINFMT_ELF=y
 CONFIG_BINFMT_MISC=m
 
+# CONFIG_DMAR is not set
+
 #
 # Power management and ACPI
 #

arch/ia64/configs/tiger_defconfig

@@ -172,6 +172,8 @@ CONFIG_DMIID=y
 CONFIG_BINFMT_ELF=y
 CONFIG_BINFMT_MISC=m
 
+# CONFIG_DMAR is not set
+
 #
 # Power management and ACPI
 #

arch/ia64/dig/Makefile

@@ -6,4 +6,9 @@
 #
 
 obj-y := setup.o
+ifeq ($(CONFIG_DMAR), y)
+obj-$(CONFIG_IA64_GENERIC) += machvec.o machvec_vtd.o dig_vtd_iommu.o
+else
 obj-$(CONFIG_IA64_GENERIC) += machvec.o
+endif
+obj-$(CONFIG_IA64_DIG_VTD) += dig_vtd_iommu.o

arch/ia64/dig/dig_vtd_iommu.c

@@ -0,0 +1,59 @@
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/intel-iommu.h>
+
+void *
+vtd_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle,
+		 gfp_t flags)
+{
+	return intel_alloc_coherent(dev, size, dma_handle, flags);
+}
+EXPORT_SYMBOL_GPL(vtd_alloc_coherent);
+
+void
+vtd_free_coherent(struct device *dev, size_t size, void *vaddr,
+		 dma_addr_t dma_handle)
+{
+	intel_free_coherent(dev, size, vaddr, dma_handle);
+}
+EXPORT_SYMBOL_GPL(vtd_free_coherent);
+
+dma_addr_t
+vtd_map_single_attrs(struct device *dev, void *addr, size_t size,
+		     int dir, struct dma_attrs *attrs)
+{
+	return intel_map_single(dev, (phys_addr_t)addr, size, dir);
+}
+EXPORT_SYMBOL_GPL(vtd_map_single_attrs);
+
+void
+vtd_unmap_single_attrs(struct device *dev, dma_addr_t iova, size_t size,
+		       int dir, struct dma_attrs *attrs)
+{
+	intel_unmap_single(dev, iova, size, dir);
+}
+EXPORT_SYMBOL_GPL(vtd_unmap_single_attrs);
+
+int
+vtd_map_sg_attrs(struct device *dev, struct scatterlist *sglist, int nents,
+		 int dir, struct dma_attrs *attrs)
+{
+	return intel_map_sg(dev, sglist, nents, dir);
+}
+EXPORT_SYMBOL_GPL(vtd_map_sg_attrs);
+
+void
+vtd_unmap_sg_attrs(struct device *dev, struct scatterlist *sglist,
+		   int nents, int dir, struct dma_attrs *attrs)
+{
+	intel_unmap_sg(dev, sglist, nents, dir);
+}
+EXPORT_SYMBOL_GPL(vtd_unmap_sg_attrs);
+
+int
+vtd_dma_mapping_error(struct device *dev, dma_addr_t dma_addr)
+{
+	return 0;
+}
+EXPORT_SYMBOL_GPL(vtd_dma_mapping_error);

arch/ia64/dig/machvec_vtd.c

@@ -0,0 +1,3 @@
+#define MACHVEC_PLATFORM_NAME		dig_vtd
+#define MACHVEC_PLATFORM_HEADER		<asm/machvec_dig_vtd.h>
+#include <asm/machvec_init.h>

arch/ia64/include/asm/cacheflush.h

@@ -34,6 +34,8 @@ do {						\
 #define flush_dcache_mmap_unlock(mapping)	do { } while (0)
 
 extern void flush_icache_range (unsigned long start, unsigned long end);
+extern void clflush_cache_range(void *addr, int size);
+
 
 #define flush_icache_user_range(vma, page, user_addr, len)					\
 do {												\

arch/ia64/include/asm/device.h

@@ -10,6 +10,9 @@ struct dev_archdata {
 #ifdef CONFIG_ACPI
 	void	*acpi_handle;
 #endif
+#ifdef CONFIG_DMAR
+	void *iommu; /* hook for IOMMU specific extension */
+#endif
 };
 
 #endif /* _ASM_IA64_DEVICE_H */

arch/ia64/include/asm/dma-mapping.h

@@ -7,6 +7,49 @@
  */
 #include <asm/machvec.h>
 #include <linux/scatterlist.h>
+#include <asm/swiotlb.h>
+
+struct dma_mapping_ops {
+	int             (*mapping_error)(struct device *dev,
+					 dma_addr_t dma_addr);
+	void*           (*alloc_coherent)(struct device *dev, size_t size,
+				dma_addr_t *dma_handle, gfp_t gfp);
+	void            (*free_coherent)(struct device *dev, size_t size,
+				void *vaddr, dma_addr_t dma_handle);
+	dma_addr_t      (*map_single)(struct device *hwdev, unsigned long ptr,
+				size_t size, int direction);
+	void            (*unmap_single)(struct device *dev, dma_addr_t addr,
+				size_t size, int direction);
+	void            (*sync_single_for_cpu)(struct device *hwdev,
+				dma_addr_t dma_handle, size_t size,
+				int direction);
+	void            (*sync_single_for_device)(struct device *hwdev,
+				dma_addr_t dma_handle, size_t size,
+				int direction);
+	void            (*sync_single_range_for_cpu)(struct device *hwdev,
+				dma_addr_t dma_handle, unsigned long offset,
+				size_t size, int direction);
+	void            (*sync_single_range_for_device)(struct device *hwdev,
+				dma_addr_t dma_handle, unsigned long offset,
+				size_t size, int direction);
+	void            (*sync_sg_for_cpu)(struct device *hwdev,
+				struct scatterlist *sg, int nelems,
+				int direction);
+	void            (*sync_sg_for_device)(struct device *hwdev,
+				struct scatterlist *sg, int nelems,
+				int direction);
+	int             (*map_sg)(struct device *hwdev, struct scatterlist *sg,
+				int nents, int direction);
+	void            (*unmap_sg)(struct device *hwdev,
+				struct scatterlist *sg, int nents,
+				int direction);
+	int             (*dma_supported_op)(struct device *hwdev, u64 mask);
+	int		is_phys;
+};
+
+extern struct dma_mapping_ops *dma_ops;
+extern struct ia64_machine_vector ia64_mv;
+extern void set_iommu_machvec(void);
 
 #define dma_alloc_coherent(dev, size, handle, gfp)	\
 	platform_dma_alloc_coherent(dev, size, handle, (gfp) | GFP_DMA)
@@ -96,4 +139,11 @@ dma_cache_sync (struct device *dev, void *vaddr, size_t size,
 
 #define dma_is_consistent(d, h)	(1)	/* all we do is coherent memory... */
 
+static inline struct dma_mapping_ops *get_dma_ops(struct device *dev)
+{
+	return dma_ops;
+}
+
+
+
 #endif /* _ASM_IA64_DMA_MAPPING_H */

arch/ia64/include/asm/iommu.h

@@ -0,0 +1,16 @@
+#ifndef _ASM_IA64_IOMMU_H
+#define _ASM_IA64_IOMMU_H 1
+
+#define cpu_has_x2apic 0
+/* 10 seconds */
+#define DMAR_OPERATION_TIMEOUT (((cycles_t) local_cpu_data->itc_freq)*10)
+
+extern void pci_iommu_shutdown(void);
+extern void no_iommu_init(void);
+extern int force_iommu, no_iommu;
+extern int iommu_detected;
+extern void iommu_dma_init(void);
+extern void machvec_init(const char *name);
+extern int forbid_dac;
+
+#endif

arch/ia64/include/asm/machvec.h

@@ -120,6 +120,8 @@ extern void machvec_tlb_migrate_finish (struct mm_struct *);
 #  include <asm/machvec_hpsim.h>
 # elif defined (CONFIG_IA64_DIG)
 #  include <asm/machvec_dig.h>
+# elif defined(CONFIG_IA64_DIG_VTD)
+#  include <asm/machvec_dig_vtd.h>
 # elif defined (CONFIG_IA64_HP_ZX1)
 #  include <asm/machvec_hpzx1.h>
 # elif defined (CONFIG_IA64_HP_ZX1_SWIOTLB)

arch/ia64/include/asm/machvec_dig_vtd.h

@@ -0,0 +1,38 @@
+#ifndef _ASM_IA64_MACHVEC_DIG_VTD_h
+#define _ASM_IA64_MACHVEC_DIG_VTD_h
+
+extern ia64_mv_setup_t			dig_setup;
+extern ia64_mv_dma_alloc_coherent	vtd_alloc_coherent;
+extern ia64_mv_dma_free_coherent	vtd_free_coherent;
+extern ia64_mv_dma_map_single_attrs	vtd_map_single_attrs;
+extern ia64_mv_dma_unmap_single_attrs	vtd_unmap_single_attrs;
+extern ia64_mv_dma_map_sg_attrs		vtd_map_sg_attrs;
+extern ia64_mv_dma_unmap_sg_attrs	vtd_unmap_sg_attrs;
+extern ia64_mv_dma_supported		iommu_dma_supported;
+extern ia64_mv_dma_mapping_error	vtd_dma_mapping_error;
+extern ia64_mv_dma_init			pci_iommu_alloc;
+
+/*
+ * This stuff has dual use!
+ *
+ * For a generic kernel, the macros are used to initialize the
+ * platform's machvec structure.  When compiling a non-generic kernel,
+ * the macros are used directly.
+ */
+#define platform_name				"dig_vtd"
+#define platform_setup				dig_setup
+#define platform_dma_init			pci_iommu_alloc
+#define platform_dma_alloc_coherent		vtd_alloc_coherent
+#define platform_dma_free_coherent		vtd_free_coherent
+#define platform_dma_map_single_attrs		vtd_map_single_attrs
+#define platform_dma_unmap_single_attrs		vtd_unmap_single_attrs
+#define platform_dma_map_sg_attrs		vtd_map_sg_attrs
+#define platform_dma_unmap_sg_attrs		vtd_unmap_sg_attrs
+#define platform_dma_sync_single_for_cpu	machvec_dma_sync_single
+#define platform_dma_sync_sg_for_cpu		machvec_dma_sync_sg
+#define platform_dma_sync_single_for_device	machvec_dma_sync_single
+#define platform_dma_sync_sg_for_device		machvec_dma_sync_sg
+#define platform_dma_supported			iommu_dma_supported
+#define platform_dma_mapping_error		vtd_dma_mapping_error
+
+#endif /* _ASM_IA64_MACHVEC_DIG_VTD_h */

arch/ia64/include/asm/machvec_init.h

@@ -1,3 +1,4 @@
+#include <asm/iommu.h>
 #include <asm/machvec.h>
 
 extern ia64_mv_send_ipi_t ia64_send_ipi;

arch/ia64/include/asm/pci.h

@@ -164,4 +164,7 @@ static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel)
 	return channel ? isa_irq_to_vector(15) : isa_irq_to_vector(14);
 }
 
+#ifdef CONFIG_DMAR
+extern void pci_iommu_alloc(void);
+#endif
 #endif /* _ASM_IA64_PCI_H */

arch/ia64/include/asm/swiotlb.h

@@ -0,0 +1,56 @@
+#ifndef ASM_IA64__SWIOTLB_H
+#define ASM_IA64__SWIOTLB_H
+
+#include <linux/dma-mapping.h>
+
+/* SWIOTLB interface */
+
+extern dma_addr_t swiotlb_map_single(struct device *hwdev, void *ptr,
+				     size_t size, int dir);
+extern void *swiotlb_alloc_coherent(struct device *hwdev, size_t size,
+				    dma_addr_t *dma_handle, gfp_t flags);
+extern void swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr,
+				 size_t size, int dir);
+extern void swiotlb_sync_single_for_cpu(struct device *hwdev,
+					dma_addr_t dev_addr,
+					size_t size, int dir);
+extern void swiotlb_sync_single_for_device(struct device *hwdev,
+					   dma_addr_t dev_addr,
+					   size_t size, int dir);
+extern void swiotlb_sync_single_range_for_cpu(struct device *hwdev,
+					      dma_addr_t dev_addr,
+					      unsigned long offset,
+					      size_t size, int dir);
+extern void swiotlb_sync_single_range_for_device(struct device *hwdev,
+						 dma_addr_t dev_addr,
+						 unsigned long offset,
+						 size_t size, int dir);
+extern void swiotlb_sync_sg_for_cpu(struct device *hwdev,
+				    struct scatterlist *sg, int nelems,
+				    int dir);
+extern void swiotlb_sync_sg_for_device(struct device *hwdev,
+				       struct scatterlist *sg, int nelems,
+				       int dir);
+extern int swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg,
+			  int nents, int direction);
+extern void swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg,
+			     int nents, int direction);
+extern int swiotlb_dma_mapping_error(struct device *hwdev, dma_addr_t dma_addr);
+extern void swiotlb_free_coherent(struct device *hwdev, size_t size,
+				  void *vaddr, dma_addr_t dma_handle);
+extern int swiotlb_dma_supported(struct device *hwdev, u64 mask);
+extern void swiotlb_init(void);
+
+extern int swiotlb_force;
+
+#ifdef CONFIG_SWIOTLB
+extern int swiotlb;
+extern void pci_swiotlb_init(void);
+#else
+#define swiotlb 0
+static inline void pci_swiotlb_init(void)
+{
+}
+#endif
+
+#endif /* ASM_IA64__SWIOTLB_H */

arch/ia64/kernel/Makefile

@@ -42,6 +42,10 @@ obj-$(CONFIG_IA64_ESI)		+= esi.o
 ifneq ($(CONFIG_IA64_ESI),)
 obj-y				+= esi_stub.o	# must be in kernel proper
 endif
+obj-$(CONFIG_DMAR)		+= pci-dma.o
+ifeq ($(CONFIG_DMAR), y)
+obj-$(CONFIG_SWIOTLB)		+= pci-swiotlb.o
+endif
 
 # The gate DSO image is built using a special linker script.
 targets += gate.so gate-syms.o