Merge branch 'master' into for-upstream

Conflicts:

	Documentation/ABI/testing/sysfs-bus-usb
	drivers/Makefile

23 files changed, 321 insertions, 292 deletions

diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
index 48e496f..3b7aa38 100644
--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -60,14 +60,6 @@ config RWSEM_XCHGADD_ALGORITHM
 	bool
 	default y
 
-config ARCH_HAS_ILOG2_U32
-	bool
-	default n
-
-config ARCH_HAS_ILOG2_U64
-	bool
-	default n
-
 config HUGETLB_PAGE_SIZE_VARIABLE
 	bool
 	depends on HUGETLB_PAGE
diff --git a/arch/ia64/ia32/binfmt_elf32.c b/arch/ia64/ia32/binfmt_elf32.c
index 4f0c30c..f92bdaa 100644
--- a/arch/ia64/ia32/binfmt_elf32.c
+++ b/arch/ia64/ia32/binfmt_elf32.c
@@ -41,7 +41,7 @@ randomize_stack_top(unsigned long stack_top);
 #define elf_map				elf32_map
 
 #undef SET_PERSONALITY
-#define SET_PERSONALITY(ex, ibcs2)	elf32_set_personality()
+#define SET_PERSONALITY(ex)	elf32_set_personality()
 
 #define elf_read_implies_exec(ex, have_pt_gnu_stack)	(!(have_pt_gnu_stack))
 
diff --git a/arch/ia64/ia32/ia32_entry.S b/arch/ia64/ia32/ia32_entry.S
index ff88c48..53505bb 100644
--- a/arch/ia64/ia32/ia32_entry.S
+++ b/arch/ia64/ia32/ia32_entry.S
@@ -251,8 +251,8 @@ ia32_syscall_table:
 	data8 compat_sys_setrlimit	  /* 75 */
 	data8 compat_sys_old_getrlimit
 	data8 compat_sys_getrusage
-	data8 sys32_gettimeofday
-	data8 sys32_settimeofday
+	data8 compat_sys_gettimeofday
+	data8 compat_sys_settimeofday
 	data8 sys32_getgroups16	  /* 80 */
 	data8 sys32_setgroups16
 	data8 sys32_old_select
diff --git a/arch/ia64/ia32/ia32priv.h b/arch/ia64/ia32/ia32priv.h
index dd0c536..0f15349 100644
--- a/arch/ia64/ia32/ia32priv.h
+++ b/arch/ia64/ia32/ia32priv.h
@@ -332,8 +332,8 @@ void ia64_elf32_init(struct pt_regs *regs);
 #define ELF_PLATFORM	NULL
 
 #ifdef __KERNEL__
-# define SET_PERSONALITY(EX,IBCS2)				\
-	(current->personality = (IBCS2) ? PER_SVR4 : PER_LINUX)
+# define SET_PERSONALITY(EX)				\
+	(current->personality = PER_LINUX)
 #endif
 
 #define IA32_EFLAG	0x200
diff --git a/arch/ia64/ia32/sys_ia32.c b/arch/ia64/ia32/sys_ia32.c
index bf196cb..f4430bb 100644
--- a/arch/ia64/ia32/sys_ia32.c
+++ b/arch/ia64/ia32/sys_ia32.c
@@ -118,41 +118,6 @@ sys32_execve (char __user *name, compat_uptr_t __user *argv, compat_uptr_t __use
 	return error;
 }
 
-int cp_compat_stat(struct kstat *stat, struct compat_stat __user *ubuf)
-{
-	compat_ino_t ino;
-	int err;
-
-	if ((u64) stat->size > MAX_NON_LFS ||
-	    !old_valid_dev(stat->dev) ||
-	    !old_valid_dev(stat->rdev))
-		return -EOVERFLOW;
-
-	ino = stat->ino;
-	if (sizeof(ino) < sizeof(stat->ino) && ino != stat->ino)
-		return -EOVERFLOW;
-
-	if (clear_user(ubuf, sizeof(*ubuf)))
-		return -EFAULT;
-
-	err  = __put_user(old_encode_dev(stat->dev), &ubuf->st_dev);
-	err |= __put_user(ino, &ubuf->st_ino);
-	err |= __put_user(stat->mode, &ubuf->st_mode);
-	err |= __put_user(stat->nlink, &ubuf->st_nlink);
-	err |= __put_user(high2lowuid(stat->uid), &ubuf->st_uid);
-	err |= __put_user(high2lowgid(stat->gid), &ubuf->st_gid);
-	err |= __put_user(old_encode_dev(stat->rdev), &ubuf->st_rdev);
-	err |= __put_user(stat->size, &ubuf->st_size);
-	err |= __put_user(stat->atime.tv_sec, &ubuf->st_atime);
-	err |= __put_user(stat->atime.tv_nsec, &ubuf->st_atime_nsec);
-	err |= __put_user(stat->mtime.tv_sec, &ubuf->st_mtime);
-	err |= __put_user(stat->mtime.tv_nsec, &ubuf->st_mtime_nsec);
-	err |= __put_user(stat->ctime.tv_sec, &ubuf->st_ctime);
-	err |= __put_user(stat->ctime.tv_nsec, &ubuf->st_ctime_nsec);
-	err |= __put_user(stat->blksize, &ubuf->st_blksize);
-	err |= __put_user(stat->blocks, &ubuf->st_blocks);
-	return err;
-}
 
 #if PAGE_SHIFT > IA32_PAGE_SHIFT
 
@@ -1148,68 +1113,12 @@ sys32_pipe (int __user *fd)
 	return retval;
 }
 
-static inline long
-get_tv32 (struct timeval *o, struct compat_timeval __user *i)
-{
-	return (!access_ok(VERIFY_READ, i, sizeof(*i)) ||
-		(__get_user(o->tv_sec, &i->tv_sec) | __get_user(o->tv_usec, &i->tv_usec)));
-}
-
-static inline long
-put_tv32 (struct compat_timeval __user *o, struct timeval *i)
-{
-	return (!access_ok(VERIFY_WRITE, o, sizeof(*o)) ||
-		(__put_user(i->tv_sec, &o->tv_sec) | __put_user(i->tv_usec, &o->tv_usec)));
-}
-
 asmlinkage unsigned long
 sys32_alarm (unsigned int seconds)
 {
 	return alarm_setitimer(seconds);
 }
 
-/* Translations due to time_t size differences.  Which affects all
-   sorts of things, like timeval and itimerval.  */
-
-extern struct timezone sys_tz;
-
-asmlinkage long
-sys32_gettimeofday (struct compat_timeval __user *tv, struct timezone __user *tz)
-{
-	if (tv) {
-		struct timeval ktv;
-		do_gettimeofday(&ktv);
-		if (put_tv32(tv, &ktv))
-			return -EFAULT;
-	}
-	if (tz) {
-		if (copy_to_user(tz, &sys_tz, sizeof(sys_tz)))
-			return -EFAULT;
-	}
-	return 0;
-}
-
-asmlinkage long
-sys32_settimeofday (struct compat_timeval __user *tv, struct timezone __user *tz)
-{
-	struct timeval ktv;
-	struct timespec kts;
-	struct timezone ktz;
-
-	if (tv) {
-		if (get_tv32(&ktv, tv))
-			return -EFAULT;
-		kts.tv_sec = ktv.tv_sec;
-		kts.tv_nsec = ktv.tv_usec * 1000;
-	}
-	if (tz) {
-		if (copy_from_user(&ktz, tz, sizeof(ktz)))
-			return -EFAULT;
-	}
-
-	return do_sys_settimeofday(tv ? &kts : NULL, tz ? &ktz : NULL);
-}
-
 struct sel_arg_struct {
 	unsigned int n;
 	unsigned int inp;
diff --git a/arch/ia64/include/asm/a.out.h b/arch/ia64/include/asm/a.out.h
deleted file mode 100644
index 193dcfb..0000000
--- a/arch/ia64/include/asm/a.out.h
+++ /dev/null
@@ -1,32 +0,0 @@
-#ifndef _ASM_IA64_A_OUT_H
-#define _ASM_IA64_A_OUT_H
-
-/*
- * No a.out format has been (or should be) defined so this file is
- * just a dummy that allows us to get binfmt_elf compiled.  It
- * probably would be better to clean up binfmt_elf.c so it does not
- * necessarily depend on there being a.out support.
- *
- * Modified 1998-2002
- *	David Mosberger-Tang <davidm@hpl.hp.com>, Hewlett-Packard Co.
- */
-
-#include <linux/types.h>
-
-struct exec {
-	unsigned long a_info;
-	unsigned long a_text;
-	unsigned long a_data;
-	unsigned long a_bss;
-	unsigned long a_entry;
-};
-
-#define N_TXTADDR(x)	0
-#define N_DATADDR(x)	0
-#define N_BSSADDR(x)	0
-#define N_DRSIZE(x)	0
-#define N_TRSIZE(x)	0
-#define N_SYMSIZE(x)	0
-#define N_TXTOFF(x)	0
-
-#endif /* _ASM_IA64_A_OUT_H */
diff --git a/arch/ia64/include/asm/dma-mapping.h b/arch/ia64/include/asm/dma-mapping.h
index 9f0df9b..06ff1ba 100644
--- a/arch/ia64/include/asm/dma-mapping.h
+++ b/arch/ia64/include/asm/dma-mapping.h
@@ -8,7 +8,9 @@
 #include <asm/machvec.h>
 #include <linux/scatterlist.h>
 
-#define dma_alloc_coherent	platform_dma_alloc_coherent
+#define dma_alloc_coherent(dev, size, handle, gfp)	\
+	platform_dma_alloc_coherent(dev, size, handle, (gfp) | GFP_DMA)
+
 /* coherent mem. is cheap */
 static inline void *
 dma_alloc_noncoherent(struct device *dev, size_t size, dma_addr_t *dma_handle,
diff --git a/arch/ia64/include/asm/elf.h b/arch/ia64/include/asm/elf.h
index 2acb6b6..86eddee 100644
--- a/arch/ia64/include/asm/elf.h
+++ b/arch/ia64/include/asm/elf.h
@@ -202,7 +202,7 @@ extern void ia64_elf_core_copy_regs (struct pt_regs *src, elf_gregset_t dst);
    relevant until we have real hardware to play with... */
 #define ELF_PLATFORM	NULL
 
-#define SET_PERSONALITY(ex, ibcs2)	set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex)	set_personality(PER_LINUX)
 #define elf_read_implies_exec(ex, executable_stack)					\
 	((executable_stack!=EXSTACK_DISABLE_X) && ((ex).e_flags & EF_IA_64_LINUX_EXECUTABLE_STACK) != 0)
 
diff --git a/arch/ia64/include/asm/kvm_host.h b/arch/ia64/include/asm/kvm_host.h
index 1efe513..85db124 100644
--- a/arch/ia64/include/asm/kvm_host.h
+++ b/arch/ia64/include/asm/kvm_host.h
@@ -132,7 +132,7 @@
 #define GPFN_IOSAPIC        (4UL << 60) /* IOSAPIC base */
 #define GPFN_LEGACY_IO      (5UL << 60) /* Legacy I/O base */
 #define GPFN_GFW        (6UL << 60) /* Guest Firmware */
-#define GPFN_HIGH_MMIO      (7UL << 60) /* High MMIO range */
+#define GPFN_PHYS_MMIO      (7UL << 60) /* Directed MMIO Range */
 
 #define GPFN_IO_MASK        (7UL << 60) /* Guest pfn is I/O type */
 #define GPFN_INV_MASK       (1UL << 63) /* Guest pfn is invalid */
@@ -413,6 +413,10 @@ struct kvm_arch {
 	struct kvm_ioapic *vioapic;
 	struct kvm_vm_stat stat;
 	struct kvm_sal_data rdv_sal_data;
+
+	struct list_head assigned_dev_head;
+	struct dmar_domain *intel_iommu_domain;
+	struct hlist_head irq_ack_notifier_list;
 };
 
 union cpuid3_t {
diff --git a/arch/ia64/include/asm/siginfo.h b/arch/ia64/include/asm/siginfo.h
index 9294e4b..118d429 100644
--- a/arch/ia64/include/asm/siginfo.h
+++ b/arch/ia64/include/asm/siginfo.h
@@ -113,11 +113,6 @@ typedef struct siginfo {
 #undef NSIGSEGV
 #define NSIGSEGV	3
 
-/*
- * SIGTRAP si_codes
- */
-#define TRAP_BRANCH	(__SI_FAULT|3)	/* process taken branch trap */
-#define TRAP_HWBKPT	(__SI_FAULT|4)	/* hardware breakpoint or watchpoint */
 #undef NSIGTRAP
 #define NSIGTRAP	4
 
diff --git a/arch/ia64/include/asm/statfs.h b/arch/ia64/include/asm/statfs.h
index 8110979..1e58966 100644
--- a/arch/ia64/include/asm/statfs.h
+++ b/arch/ia64/include/asm/statfs.h
@@ -8,55 +8,13 @@
  *	David Mosberger-Tang <davidm@hpl.hp.com>, Hewlett-Packard Co
  */
 
-#ifndef __KERNEL_STRICT_NAMES
-# include <linux/types.h>
-typedef __kernel_fsid_t	fsid_t;
-#endif
-
 /*
- * This is ugly --- we're already 64-bit, so just duplicate the definitions
+ * We need compat_statfs64 to be packed, because the i386 ABI won't
+ * add padding at the end to bring it to a multiple of 8 bytes, but
+ * the IA64 ABI will.
  */
-struct statfs {
-	long f_type;
-	long f_bsize;
-	long f_blocks;
-	long f_bfree;
-	long f_bavail;
-	long f_files;
-	long f_ffree;
-	__kernel_fsid_t f_fsid;
-	long f_namelen;
-	long f_frsize;
-	long f_spare[5];
-};
-
-
-struct statfs64 {
-	long f_type;
-	long f_bsize;
-	long f_blocks;
-	long f_bfree;
-	long f_bavail;
-	long f_files;
-	long f_ffree;
-	__kernel_fsid_t f_fsid;
-	long f_namelen;
-	long f_frsize;
-	long f_spare[5];
-};
+#define ARCH_PACK_COMPAT_STATFS64 __attribute__((packed,aligned(4)))
 
-struct compat_statfs64 {
-	__u32 f_type;
-	__u32 f_bsize;
-	__u64 f_blocks;
-	__u64 f_bfree;
-	__u64 f_bavail;
-	__u64 f_files;
-	__u64 f_ffree;
-	__kernel_fsid_t f_fsid;
-	__u32 f_namelen;
-	__u32 f_frsize;
-	__u32 f_spare[5];
-} __attribute__((packed));
+#include <asm-generic/statfs.h>
 
 #endif /* _ASM_IA64_STATFS_H */
diff --git a/arch/ia64/kernel/smpboot.c b/arch/ia64/kernel/smpboot.c
index d8f05e5..1dcbb85 100644
--- a/arch/ia64/kernel/smpboot.c
+++ b/arch/ia64/kernel/smpboot.c
@@ -401,6 +401,7 @@ smp_callin (void)
 	spin_lock(&vector_lock);
 	/* Setup the per cpu irq handling data structures */
 	__setup_vector_irq(cpuid);
+	notify_cpu_starting(cpuid);
 	cpu_set(cpuid, cpu_online_map);
 	per_cpu(cpu_state, cpuid) = CPU_ONLINE;
 	spin_unlock(&vector_lock);
diff --git a/arch/ia64/kvm/Kconfig b/arch/ia64/kvm/Kconfig
index 7914e48..8e99fed 100644
--- a/arch/ia64/kvm/Kconfig
+++ b/arch/ia64/kvm/Kconfig
@@ -46,4 +46,6 @@ config KVM_INTEL
 config KVM_TRACE
        bool
 
+source drivers/virtio/Kconfig
+
 endif # VIRTUALIZATION
diff --git a/arch/ia64/kvm/Makefile b/arch/ia64/kvm/Makefile
index bf22fb9..cf37f8f 100644
--- a/arch/ia64/kvm/Makefile
+++ b/arch/ia64/kvm/Makefile
@@ -44,7 +44,11 @@ EXTRA_CFLAGS += -Ivirt/kvm -Iarch/ia64/kvm/
 EXTRA_AFLAGS += -Ivirt/kvm -Iarch/ia64/kvm/
 
 common-objs = $(addprefix ../../../virt/kvm/, kvm_main.o ioapic.o \
-		coalesced_mmio.o)
+		coalesced_mmio.o irq_comm.o)
+
+ifeq ($(CONFIG_DMAR),y)
+common-objs += $(addprefix ../../../virt/kvm/, vtd.o)
+endif
 
 kvm-objs := $(common-objs) kvm-ia64.o kvm_fw.o
 obj-$(CONFIG_KVM) += kvm.o
diff --git a/arch/ia64/kvm/irq.h b/arch/ia64/kvm/irq.h
new file mode 100644
index 0000000..c6786e8
--- /dev/null
+++ b/arch/ia64/kvm/irq.h
@@ -0,0 +1,31 @@
+/*
+ * irq.h: In-kernel interrupt controller related definitions
+ * Copyright (c) 2008, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc., 59 Temple
+ * Place - Suite 330, Boston, MA 02111-1307 USA.
+ *
+ * Authors:
+ *   Xiantao Zhang <xiantao.zhang@intel.com>
+ *
+ */
+
+#ifndef __IRQ_H
+#define __IRQ_H
+
+static inline int irqchip_in_kernel(struct kvm *kvm)
+{
+	return 1;
+}
+
+#endif
diff --git a/arch/ia64/kvm/kvm-ia64.c b/arch/ia64/kvm/kvm-ia64.c
index cd0d1a7..c0699f0 100644
--- a/arch/ia64/kvm/kvm-ia64.c
+++ b/arch/ia64/kvm/kvm-ia64.c
@@ -31,6 +31,7 @@
 #include <linux/bitops.h>
 #include <linux/hrtimer.h>
 #include <linux/uaccess.h>
+#include <linux/intel-iommu.h>
 
 #include <asm/pgtable.h>
 #include <asm/gcc_intrin.h>
@@ -45,6 +46,7 @@
 #include "iodev.h"
 #include "ioapic.h"
 #include "lapic.h"
+#include "irq.h"
 
 static unsigned long kvm_vmm_base;
 static unsigned long kvm_vsa_base;
@@ -179,12 +181,16 @@ int kvm_dev_ioctl_check_extension(long ext)
 	switch (ext) {
 	case KVM_CAP_IRQCHIP:
 	case KVM_CAP_USER_MEMORY:
+	case KVM_CAP_MP_STATE:
 
 		r = 1;
 		break;
 	case KVM_CAP_COALESCED_MMIO:
 		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
 		break;
+	case KVM_CAP_IOMMU:
+		r = intel_iommu_found();
+		break;
 	default:
 		r = 0;
 	}
@@ -771,6 +777,7 @@ static void kvm_init_vm(struct kvm *kvm)
 	 */
 	kvm_build_io_pmt(kvm);
 
+	INIT_LIST_HEAD(&kvm->arch.assigned_dev_head);
 }
 
 struct  kvm *kvm_arch_create_vm(void)
@@ -1334,6 +1341,10 @@ static void kvm_release_vm_pages(struct kvm *kvm)
 
 void kvm_arch_destroy_vm(struct kvm *kvm)
 {
+	kvm_iommu_unmap_guest(kvm);
+#ifdef  KVM_CAP_DEVICE_ASSIGNMENT
+	kvm_free_all_assigned_devices(kvm);
+#endif
 	kfree(kvm->arch.vioapic);
 	kvm_release_vm_pages(kvm);
 	kvm_free_physmem(kvm);
@@ -1435,17 +1446,24 @@ int kvm_arch_set_memory_region(struct kvm *kvm,
 		int user_alloc)
 {
 	unsigned long i;
-	struct page *page;
+	unsigned long pfn;
 	int npages = mem->memory_size >> PAGE_SHIFT;
 	struct kvm_memory_slot *memslot = &kvm->memslots[mem->slot];
 	unsigned long base_gfn = memslot->base_gfn;
 
 	for (i = 0; i < npages; i++) {
-		page = gfn_to_page(kvm, base_gfn + i);
-		kvm_set_pmt_entry(kvm, base_gfn + i,
-				page_to_pfn(page) << PAGE_SHIFT,
-				_PAGE_AR_RWX|_PAGE_MA_WB);
-		memslot->rmap[i] = (unsigned long)page;
+		pfn = gfn_to_pfn(kvm, base_gfn + i);
+		if (!kvm_is_mmio_pfn(pfn)) {
+			kvm_set_pmt_entry(kvm, base_gfn + i,
+					pfn << PAGE_SHIFT,
+				_PAGE_AR_RWX | _PAGE_MA_WB);
+			memslot->rmap[i] = (unsigned long)pfn_to_page(pfn);
+		} else {
+			kvm_set_pmt_entry(kvm, base_gfn + i,
+					GPFN_PHYS_MMIO | (pfn << PAGE_SHIFT),
+					_PAGE_MA_UC);
+			memslot->rmap[i] = 0;
+			}
 	}
 
 	return 0;
@@ -1789,11 +1807,43 @@ int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
-	return -EINVAL;
+	vcpu_load(vcpu);
+	mp_state->mp_state = vcpu->arch.mp_state;
+	vcpu_put(vcpu);
+	return 0;
+}
+
+static int vcpu_reset(struct kvm_vcpu *vcpu)
+{
+	int r;
+	long psr;
+	local_irq_save(psr);
+	r = kvm_insert_vmm_mapping(vcpu);
+	if (r)
+		goto fail;
+
+	vcpu->arch.launched = 0;
+	kvm_arch_vcpu_uninit(vcpu);
+	r = kvm_arch_vcpu_init(vcpu);
+	if (r)
+		goto fail;
+
+	kvm_purge_vmm_mapping(vcpu);
+	r = 0;
+fail:
+	local_irq_restore(psr);
+	return r;
 }
 
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
-	return -EINVAL;
+	int r = 0;
+
+	vcpu_load(vcpu);
+	vcpu->arch.mp_state = mp_state->mp_state;
+	if (vcpu->arch.mp_state == KVM_MP_STATE_UNINITIALIZED)
+		r = vcpu_reset(vcpu);
+	vcpu_put(vcpu);
+	return r;
 }
diff --git a/arch/ia64/kvm/kvm_minstate.h b/arch/ia64/kvm/kvm_minstate.h
index 13980d9..2cc41d1 100644
--- a/arch/ia64/kvm/kvm_minstate.h
+++ b/arch/ia64/kvm/kvm_minstate.h
@@ -50,27 +50,18 @@
 
 #define PAL_VSA_SYNC_READ						\
 	/* begin to call pal vps sync_read */				\
+{.mii;									\
 	add r25 = VMM_VPD_BASE_OFFSET, r21;				\
-	adds r20 = VMM_VCPU_VSA_BASE_OFFSET, r21;  /* entry point */	\
+	nop 0x0;							\
+	mov r24=ip;							\
 	;;								\
+}									\
+{.mmb									\
+	add r24=0x20, r24;						\
 	ld8 r25 = [r25];      /* read vpd base */			\
-	ld8 r20 = [r20];						\
-	;;								\
-	add r20 = PAL_VPS_SYNC_READ,r20;				\
-	;;								\
-{ .mii;									\
-	nop 0x0;							\
-	mov r24 = ip;							\
-	mov b0 = r20;							\
+	br.cond.sptk kvm_vps_sync_read;		/*call the service*/	\
 	;;								\
 };									\
-{ .mmb;									\
-	add r24 = 0x20, r24;						\
-	nop 0x0;							\
-	br.cond.sptk b0;        /*  call the service */			\
-	;;								\
-};
-
 
 
 #define KVM_MINSTATE_GET_CURRENT(reg)   mov reg=r21
diff --git a/arch/ia64/kvm/optvfault.S b/arch/ia64/kvm/optvfault.S
index e4f15d6..634abad 100644
--- a/arch/ia64/kvm/optvfault.S
+++ b/arch/ia64/kvm/optvfault.S
@@ -1,9 +1,12 @@
 /*
- * arch/ia64/vmx/optvfault.S
+ * arch/ia64/kvm/optvfault.S
  * optimize virtualization fault handler
  *
  * Copyright (C) 2006 Intel Co
  *	Xuefei Xu (Anthony Xu) <anthony.xu@intel.com>
+ * Copyright (C) 2008 Intel Co
+ *      Add the support for Tukwila processors.
+ *	Xiantao Zhang <xiantao.zhang@intel.com>
  */
 
 #include <asm/asmmacro.h>
@@ -20,6 +23,98 @@
 #define ACCE_MOV_TO_PSR
 #define ACCE_THASH
 
+#define VMX_VPS_SYNC_READ			\
+	add r16=VMM_VPD_BASE_OFFSET,r21;	\
+	mov r17 = b0;				\
+	mov r18 = r24;				\
+	mov r19 = r25;				\
+	mov r20 = r31;				\
+	;;					\
+{.mii;						\
+	ld8 r16 = [r16];			\
+	nop 0x0;				\
+	mov r24 = ip;				\
+	;;					\
+};						\
+{.mmb;						\
+	add r24=0x20, r24;			\
+	mov r25 =r16;				\
+	br.sptk.many kvm_vps_sync_read;		\
+};						\
+	mov b0 = r17;				\
+	mov r24 = r18;				\
+	mov r25 = r19;				\
+	mov r31 = r20
+
+ENTRY(kvm_vps_entry)
+	adds r29 = VMM_VCPU_VSA_BASE_OFFSET,r21
+	;;
+	ld8 r29 = [r29]
+	;;
+	add r29 = r29, r30
+	;;
+	mov b0 = r29
+	br.sptk.many b0
+END(kvm_vps_entry)
+
+/*
+ *	Inputs:
+ *	r24 : return address
+ *  	r25 : vpd
+ *	r29 : scratch
+ *
+ */
+GLOBAL_ENTRY(kvm_vps_sync_read)
+	movl r30 = PAL_VPS_SYNC_READ
+	;;
+	br.sptk.many kvm_vps_entry
+END(kvm_vps_sync_read)
+
+/*
+ *	Inputs:
+ *	r24 : return address
+ *  	r25 : vpd
+ *	r29 : scratch
+ *
+ */
+GLOBAL_ENTRY(kvm_vps_sync_write)
+	movl r30 = PAL_VPS_SYNC_WRITE
+	;;
+	br.sptk.many kvm_vps_entry
+END(kvm_vps_sync_write)
+
+/*
+ *	Inputs:
+ *	r23 : pr
+ *	r24 : guest b0
+ *  	r25 : vpd
+ *
+ */
+GLOBAL_ENTRY(kvm_vps_resume_normal)
+	movl r30 = PAL_VPS_RESUME_NORMAL
+	;;
+	mov pr=r23,-2
+	br.sptk.many kvm_vps_entry
+END(kvm_vps_resume_normal)
+
+/*
+ *	Inputs:
+ *	r23 : pr
+ *	r24 : guest b0
+ *  	r25 : vpd
+ *	r17 : isr
+ */
+GLOBAL_ENTRY(kvm_vps_resume_handler)
+	movl r30 = PAL_VPS_RESUME_HANDLER
+	;;
+	ld8 r27=[r25]
+	shr r17=r17,IA64_ISR_IR_BIT
+	;;
+	dep r27=r17,r27,63,1   // bit 63 of r27 indicate whether enable CFLE
+	mov pr=r23,-2
+	br.sptk.many kvm_vps_entry
+END(kvm_vps_resume_handler)
+
 //mov r1=ar3
 GLOBAL_ENTRY(kvm_asm_mov_from_ar)
 #ifndef ACCE_MOV_FROM_AR
@@ -157,11 +252,11 @@ GLOBAL_ENTRY(kvm_asm_rsm)
 #ifndef ACCE_RSM
 	br.many kvm_virtualization_fault_back
 #endif
-	add r16=VMM_VPD_BASE_OFFSET,r21
+	VMX_VPS_SYNC_READ
+	;;
 	extr.u r26=r25,6,21
 	extr.u r27=r25,31,2
 	;;
-	ld8 r16=[r16]
 	extr.u r28=r25,36,1
 	dep r26=r27,r26,21,2
 	;;
@@ -196,7 +291,7 @@ GLOBAL_ENTRY(kvm_asm_rsm)
 	tbit.nz p6,p0=r23,0
 	;;
 	tbit.z.or p6,p0=r26,IA64_PSR_DT_BIT
-	(p6) br.dptk kvm_resume_to_guest
+	(p6) br.dptk kvm_resume_to_guest_with_sync
 	;;
 	add r26=VMM_VCPU_META_RR0_OFFSET,r21
 	add r27=VMM_VCPU_META_RR0_OFFSET+8,r21
@@ -212,7 +307,7 @@ GLOBAL_ENTRY(kvm_asm_rsm)
 	mov rr[r28]=r27
 	;;
 	srlz.d
-	br.many kvm_resume_to_guest
+	br.many kvm_resume_to_guest_with_sync
 END(kvm_asm_rsm)
 
 
@@ -221,11 +316,11 @@ GLOBAL_ENTRY(kvm_asm_ssm)
 #ifndef ACCE_SSM
 	br.many kvm_virtualization_fault_back
 #endif
-	add r16=VMM_VPD_BASE_OFFSET,r21
+	VMX_VPS_SYNC_READ
+	;;
 	extr.u r26=r25,6,21
 	extr.u r27=r25,31,2
 	;;
-	ld8 r16=[r16]
 	extr.u r28=r25,36,1
 	dep r26=r27,r26,21,2
 	;;  //r26 is imm24
@@ -271,7 +366,7 @@ kvm_asm_ssm_1:
 	tbit.nz p6,p0=r29,IA64_PSR_I_BIT
 	;;
 	tbit.z.or p6,p0=r19,IA64_PSR_I_BIT
-	(p6) br.dptk kvm_resume_to_guest
+	(p6) br.dptk kvm_resume_to_guest_with_sync
 	;;
 	add r29=VPD_VTPR_START_OFFSET,r16
 	add r30=VPD_VHPI_START_OFFSET,r16
@@ -286,7 +381,7 @@ kvm_asm_ssm_1:
 	;;
 	cmp.gt p6,p0=r30,r17
 	(p6) br.dpnt.few kvm_asm_dispatch_vexirq
-	br.many kvm_resume_to_guest
+	br.many kvm_resume_to_guest_with_sync
 END(kvm_asm_ssm)
 
 
@@ -295,10 +390,9 @@ GLOBAL_ENTRY(kvm_asm_mov_to_psr)
 #ifndef ACCE_MOV_TO_PSR
 	br.many kvm_virtualization_fault_back
 #endif
-	add r16=VMM_VPD_BASE_OFFSET,r21
-	extr.u r26=r25,13,7 //r2
+	VMX_VPS_SYNC_READ
 	;;
-	ld8 r16=[r16]
+	extr.u r26=r25,13,7 //r2
 	addl r20=@gprel(asm_mov_from_reg),gp
 	;;
 	adds r30=kvm_asm_mov_to_psr_back-asm_mov_from_reg,r20
@@ -374,7 +468,7 @@ kvm_asm_mov_to_psr_1:
 	;;
 	tbit.nz.or p6,p0=r17,IA64_PSR_I_BIT
 	tbit.z.or p6,p0=r30,IA64_PSR_I_BIT
-	(p6) br.dpnt.few kvm_resume_to_guest
+	(p6) br.dpnt.few kvm_resume_to_guest_with_sync
 	;;
 	add r29=VPD_VTPR_START_OFFSET,r16
 	add r30=VPD_VHPI_START_OFFSET,r16
@@ -389,13 +483,29 @@ kvm_asm_mov_to_psr_1:
 	;;
 	cmp.gt p6,p0=r30,r17
 	(p6) br.dpnt.few kvm_asm_dispatch_vexirq
-	br.many kvm_resume_to_guest
+	br.many kvm_resume_to_guest_with_sync
 END(kvm_asm_mov_to_psr)
 
 
 ENTRY(kvm_asm_dispatch_vexirq)
 //increment iip
+	mov r17 = b0
+	mov r18 = r31
+{.mii
+	add r25=VMM_VPD_BASE_OFFSET,r21
+	nop 0x0
+	mov r24 = ip
+	;;
+}
+{.mmb
+	add r24 = 0x20, r24
+	ld8 r25 = [r25]
+	br.sptk.many kvm_vps_sync_write
+}
+	mov b0 =r17
 	mov r16=cr.ipsr
+	mov r31 = r18
+	mov r19 = 37
 	;;
 	extr.u r17=r16,IA64_PSR_RI_BIT,2
 	tbit.nz p6,p7=r16,IA64_PSR_RI_BIT+1
@@ -435,25 +545,31 @@ GLOBAL_ENTRY(kvm_asm_thash)
 	;;
 kvm_asm_thash_back1:
 	shr.u r23=r19,61		// get RR number
-	adds r25=VMM_VCPU_VRR0_OFFSET,r21	// get vcpu->arch.vrr[0]'s addr
+	adds r28=VMM_VCPU_VRR0_OFFSET,r21	// get vcpu->arch.vrr[0]'s addr
 	adds r16=VMM_VPD_VPTA_OFFSET,r16	// get vpta
 	;;
-	shladd r27=r23,3,r25	// get vcpu->arch.vrr[r23]'s addr
+	shladd r27=r23,3,r28	// get vcpu->arch.vrr[r23]'s addr
 	ld8 r17=[r16]		// get PTA
 	mov r26=1
 	;;
-	extr.u r29=r17,2,6		// get pta.size
-	ld8 r25=[r27]		// get vcpu->arch.vrr[r23]'s value
+	extr.u r29=r17,2,6	// get pta.size
+	ld8 r28=[r27]		// get vcpu->arch.vrr[r23]'s value
 	;;
-	extr.u r25=r25,2,6		// get rr.ps
+	mov b0=r24
+	//Fallback to C if pta.vf is set
+	tbit.nz p6,p0=r17, 8
+	;;
+	(p6) mov r24=EVENT_THASH
+	(p6) br.cond.dpnt.many kvm_virtualization_fault_back
+	extr.u r28=r28,2,6	// get rr.ps
 	shl r22=r26,r29		// 1UL << pta.size
 	;;
-	shr.u r23=r19,r25		// vaddr >> rr.ps
+	shr.u r23=r19,r28	// vaddr >> rr.ps
 	adds r26=3,r29		// pta.size + 3
 	shl r27=r17,3		// pta << 3
 	;;
 	shl r23=r23,3		// (vaddr >> rr.ps) << 3
-	shr.u r27=r27,r26		// (pta << 3) >> (pta.size+3)
+	shr.u r27=r27,r26	// (pta << 3) >> (pta.size+3)
 	movl r16=7<<61
 	;;
 	adds r22=-1,r22		// (1UL << pta.size) - 1
@@ -724,6 +840,29 @@ END(asm_mov_from_reg)
  * r31: pr
  * r24: b0
  */
+ENTRY(kvm_resume_to_guest_with_sync)
+	adds r19=VMM_VPD_BASE_OFFSET,r21
+	mov r16 = r31
+	mov r17 = r24
+	;;
+{.mii
+	ld8 r25 =[r19]
+	nop 0x0
+	mov r24 = ip
+	;;
+}
+{.mmb
+	add r24 =0x20, r24
+	nop 0x0
+	br.sptk.many kvm_vps_sync_write
+}
+
+	mov r31 = r16
+	mov r24 =r17
+	;;
+	br.sptk.many kvm_resume_to_guest
+END(kvm_resume_to_guest_with_sync)
+
 ENTRY(kvm_resume_to_guest)
 	adds r16 = VMM_VCPU_SAVED_GP_OFFSET,r21
 	;;
diff --git a/arch/ia64/kvm/process.c b/arch/ia64/kvm/process.c
index 5a33f7e..3417783 100644
--- a/arch/ia64/kvm/process.c
+++ b/arch/ia64/kvm/process.c
@@ -962,9 +962,9 @@ static void kvm_do_resume_op(struct kvm_vcpu *vcpu)
 void vmm_transition(struct kvm_vcpu *vcpu)
 {
 	ia64_call_vsa(PAL_VPS_SAVE, (unsigned long)vcpu->arch.vpd,
-			0, 0, 0, 0, 0, 0);
+			1, 0, 0, 0, 0, 0);
 	vmm_trampoline(&vcpu->arch.guest, &vcpu->arch.host);
 	ia64_call_vsa(PAL_VPS_RESTORE, (unsigned long)vcpu->arch.vpd,
-						0, 0, 0, 0, 0, 0);
+						1, 0, 0, 0, 0, 0);
 	kvm_do_resume_op(vcpu);
 }
diff --git a/arch/ia64/kvm/vcpu.h b/arch/ia64/kvm/vcpu.h
index b0fcfb6..341e3fe 100644
--- a/arch/ia64/kvm/vcpu.h
+++ b/arch/ia64/kvm/vcpu.h
@@ -313,21 +313,21 @@ static inline void vcpu_set_tr(struct thash_data *trp, u64 pte, u64 itir,
 	trp->rid = rid;
 }
 
-extern u64 kvm_lookup_mpa(u64 gpfn);
-extern u64 kvm_gpa_to_mpa(u64 gpa);
-
-/* Return I/O type if trye */
-#define __gpfn_is_io(gpfn)			\
-	({						\
-	 u64 pte, ret = 0;			\
-	 pte = kvm_lookup_mpa(gpfn);		\
-	 if (!(pte & GPFN_INV_MASK))		\
-	 ret = pte & GPFN_IO_MASK;	\
-	 ret;					\
-	 })
+extern u64 kvm_get_mpt_entry(u64 gpfn);
 
+/* Return I/ */
+static inline u64 __gpfn_is_io(u64 gpfn)
+{
+	u64  pte;
+	pte = kvm_get_mpt_entry(gpfn);
+	if (!(pte & GPFN_INV_MASK)) {
+		pte = pte & GPFN_IO_MASK;
+		if (pte != GPFN_PHYS_MMIO)
+			return pte;
+	}
+	return 0;
+}
 #endif
-
 #define IA64_NO_FAULT	0
 #define IA64_FAULT	1
 
diff --git a/arch/ia64/kvm/vmm_ivt.S b/arch/ia64/kvm/vmm_ivt.S
index 3ee5f48..c1d7251 100644
--- a/arch/ia64/kvm/vmm_ivt.S
+++ b/arch/ia64/kvm/vmm_ivt.S
@@ -1261,11 +1261,6 @@ kvm_rse_clear_invalid:
     adds r19=VMM_VPD_VPSR_OFFSET,r18
     ;;
     ld8 r19=[r19]        //vpsr
-    adds r20=VMM_VCPU_VSA_BASE_OFFSET,r21
-    ;;
-    ld8 r20=[r20]
-    ;;
-//vsa_sync_write_start
     mov r25=r18
     adds r16= VMM_VCPU_GP_OFFSET,r21
     ;;
@@ -1274,10 +1269,7 @@ kvm_rse_clear_invalid:
     ;;
     add  r24=r24,r16
     ;;
-    add r16=PAL_VPS_SYNC_WRITE,r20
-    ;;
-    mov b0=r16
-    br.cond.sptk b0         // call the service
+    br.sptk.many  kvm_vps_sync_write       // call the service
     ;;
 END(ia64_leave_hypervisor)
 // fall through
@@ -1288,28 +1280,15 @@ GLOBAL_ENTRY(ia64_vmm_entry)
  *  r17:cr.isr
  *  r18:vpd
  *  r19:vpsr
- *  r20:__vsa_base
  *  r22:b0
  *  r23:predicate
  */
     mov r24=r22
     mov r25=r18
     tbit.nz p1,p2 = r19,IA64_PSR_IC_BIT        // p1=vpsr.ic
+    (p1) br.cond.sptk.few kvm_vps_resume_normal
+    (p2) br.cond.sptk.many kvm_vps_resume_handler
     ;;
-    (p1) add r29=PAL_VPS_RESUME_NORMAL,r20
-    (p1) br.sptk.many ia64_vmm_entry_out
-    ;;
-    tbit.nz p1,p2 = r17,IA64_ISR_IR_BIT		//p1=cr.isr.ir
-    ;;
-    (p1) add r29=PAL_VPS_RESUME_NORMAL,r20
-    (p2) add r29=PAL_VPS_RESUME_HANDLER,r20
-    (p2) ld8 r26=[r25]
-    ;;
-ia64_vmm_entry_out:
-    mov pr=r23,-2
-    mov b0=r29
-    ;;
-    br.cond.sptk b0             // call pal service
 END(ia64_vmm_entry)
 
 
@@ -1376,6 +1355,9 @@ GLOBAL_ENTRY(vmm_reset_entry)
     //set up ipsr, iip, vpd.vpsr, dcr
     // For IPSR: it/dt/rt=1, i/ic=1, si=1, vm/bn=1
     // For DCR: all bits 0
+    bsw.0
+    ;;
+    mov r21 =r13
     adds r14=-VMM_PT_REGS_SIZE, r12
     ;;
     movl r6=0x501008826000      // IPSR dt/rt/it:1;i/ic:1, si:1, vm/bn:1
@@ -1387,12 +1369,6 @@ GLOBAL_ENTRY(vmm_reset_entry)
     ;;
     srlz.i
     ;;
-    bsw.0
-    ;;
-    mov r21 =r13
-    ;;
-    bsw.1
-    ;;
     mov ar.rsc = 0
     ;;
     flushrs
@@ -1406,12 +1382,9 @@ GLOBAL_ENTRY(vmm_reset_entry)
     ld8 r1 = [r20]
     ;;
     mov cr.iip=r4
-    ;;
     adds r16=VMM_VPD_BASE_OFFSET,r13
-    adds r20=VMM_VCPU_VSA_BASE_OFFSET,r13
     ;;
     ld8 r18=[r16]
-    ld8 r20=[r20]
     ;;
     adds r19=VMM_VPD_VPSR_OFFSET,r18
     ;;
diff --git a/arch/ia64/kvm/vtlb.c b/arch/ia64/kvm/vtlb.c
index def4576..e22b933 100644
--- a/arch/ia64/kvm/vtlb.c
+++ b/arch/ia64/kvm/vtlb.c
@@ -390,7 +390,7 @@ void thash_purge_entries_remote(struct kvm_vcpu *v, u64 va, u64 ps)
 
 u64 translate_phy_pte(u64 *pte, u64 itir, u64 va)
 {
-	u64 ps, ps_mask, paddr, maddr;
+	u64 ps, ps_mask, paddr, maddr, io_mask;
 	union pte_flags phy_pte;
 
 	ps = itir_ps(itir);
@@ -398,8 +398,9 @@ u64 translate_phy_pte(u64 *pte, u64 itir, u64 va)
 	phy_pte.val = *pte;
 	paddr = *pte;
 	paddr = ((paddr & _PAGE_PPN_MASK) & ps_mask) | (va & ~ps_mask);
-	maddr = kvm_lookup_mpa(paddr >> PAGE_SHIFT);
-	if (maddr & GPFN_IO_MASK) {
+	maddr = kvm_get_mpt_entry(paddr >> PAGE_SHIFT);
+	io_mask = maddr & GPFN_IO_MASK;
+	if (io_mask && (io_mask != GPFN_PHYS_MMIO)) {
 		*pte |= VTLB_PTE_IO;
 		return -1;
 	}
@@ -418,7 +419,7 @@ int thash_purge_and_insert(struct kvm_vcpu *v, u64 pte, u64 itir,
 						u64 ifa, int type)
 {
 	u64 ps;
-	u64 phy_pte;
+	u64 phy_pte, io_mask, index;
 	union ia64_rr vrr, mrr;
 	int ret = 0;
 
@@ -426,13 +427,16 @@ int thash_purge_and_insert(struct kvm_vcpu *v, u64 pte, u64 itir,
 	vrr.val = vcpu_get_rr(v, ifa);
 	mrr.val = ia64_get_rr(ifa);
 
+	index = (pte & _PAGE_PPN_MASK) >> PAGE_SHIFT;
+	io_mask = kvm_get_mpt_entry(index) & GPFN_IO_MASK;
 	phy_pte = translate_phy_pte(&pte, itir, ifa);
 
 	/* Ensure WB attribute if pte is related to a normal mem page,
 	 * which is required by vga acceleration since qemu maps shared
 	 * vram buffer with WB.
 	 */
-	if (!(pte & VTLB_PTE_IO) && ((pte & _PAGE_MA_MASK) != _PAGE_MA_NAT)) {
+	if (!(pte & VTLB_PTE_IO) && ((pte & _PAGE_MA_MASK) != _PAGE_MA_NAT) &&
+			io_mask != GPFN_PHYS_MMIO) {
 		pte &= ~_PAGE_MA_MASK;
 		phy_pte &= ~_PAGE_MA_MASK;
 	}
@@ -566,12 +570,19 @@ void thash_init(struct thash_cb *hcb, u64 sz)
 	}
 }
 
-u64 kvm_lookup_mpa(u64 gpfn)
+u64 kvm_get_mpt_entry(u64 gpfn)
 {
 	u64 *base = (u64 *) KVM_P2M_BASE;
 	return *(base + gpfn);
 }
 
+u64 kvm_lookup_mpa(u64 gpfn)
+{
+	u64 maddr;
+	maddr = kvm_get_mpt_entry(gpfn);
+	return maddr&_PAGE_PPN_MASK;
+}
+
 u64 kvm_gpa_to_mpa(u64 gpa)
 {
 	u64 pte = kvm_lookup_mpa(gpa >> PAGE_SHIFT);
diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c
index 200100e..f482a90 100644
--- a/arch/ia64/mm/init.c
+++ b/arch/ia64/mm/init.c
@@ -21,7 +21,6 @@
 #include <linux/bitops.h>
 #include <linux/kexec.h>
 
-#include <asm/a.out.h>
 #include <asm/dma.h>
 #include <asm/ia32.h>
 #include <asm/io.h>