Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6

230 files changed, 4611 insertions, 1779 deletions

diff --git a/arch/alpha/include/asm/machvec.h b/arch/alpha/include/asm/machvec.h
index fea4ea7..13cd427 100644
--- a/arch/alpha/include/asm/machvec.h
+++ b/arch/alpha/include/asm/machvec.h
@@ -80,7 +80,7 @@ struct alpha_machine_vector
 	void (*update_irq_hw)(unsigned long, unsigned long, int);
 	void (*ack_irq)(unsigned long);
 	void (*device_interrupt)(unsigned long vector);
-	void (*machine_check)(u64 vector, u64 la);
+	void (*machine_check)(unsigned long vector, unsigned long la);
 
 	void (*smp_callin)(void);
 	void (*init_arch)(void);
diff --git a/arch/alpha/include/asm/pci.h b/arch/alpha/include/asm/pci.h
index 2a14302..cb04eaa 100644
--- a/arch/alpha/include/asm/pci.h
+++ b/arch/alpha/include/asm/pci.h
@@ -273,4 +273,18 @@ struct pci_dev *alpha_gendev_to_pci(struct device *dev);
 
 extern struct pci_dev *isa_bridge;
 
+extern int pci_legacy_read(struct pci_bus *bus, loff_t port, u32 *val,
+			   size_t count);
+extern int pci_legacy_write(struct pci_bus *bus, loff_t port, u32 val,
+			    size_t count);
+extern int pci_mmap_legacy_page_range(struct pci_bus *bus,
+				      struct vm_area_struct *vma,
+				      enum pci_mmap_state mmap_state);
+extern void pci_adjust_legacy_attr(struct pci_bus *bus,
+				   enum pci_mmap_state mmap_type);
+#define HAVE_PCI_LEGACY	1
+
+extern int pci_create_resource_files(struct pci_dev *dev);
+extern void pci_remove_resource_files(struct pci_dev *dev);
+
 #endif /* __ALPHA_PCI_H */
diff --git a/arch/alpha/include/asm/system.h b/arch/alpha/include/asm/system.h
index afe20fa..5aa40cc 100644
--- a/arch/alpha/include/asm/system.h
+++ b/arch/alpha/include/asm/system.h
@@ -309,518 +309,71 @@ extern int __min_ipl;
 #define tbia()		__tbi(-2, /* no second argument */)
 
 /*
- * Atomic exchange.
- * Since it can be used to implement critical sections
- * it must clobber "memory" (also for interrupts in UP).
+ * Atomic exchange routines.
  */
 
-static inline unsigned long
-__xchg_u8(volatile char *m, unsigned long val)
-{
-	unsigned long ret, tmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%4,7,%3\n"
-	"	insbl	%1,%4,%1\n"
-	"1:	ldq_l	%2,0(%3)\n"
-	"	extbl	%2,%4,%0\n"
-	"	mskbl	%2,%4,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%3)\n"
-	"	beq	%2,2f\n"
-#ifdef CONFIG_SMP
-	"	mb\n"
-#endif
-	".subsection 2\n"
-	"2:	br	1b\n"
-	".previous"
-	: "=&r" (ret), "=&r" (val), "=&r" (tmp), "=&r" (addr64)
-	: "r" ((long)m), "1" (val) : "memory");
-
-	return ret;
-}
-
-static inline unsigned long
-__xchg_u16(volatile short *m, unsigned long val)
-{
-	unsigned long ret, tmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%4,7,%3\n"
-	"	inswl	%1,%4,%1\n"
-	"1:	ldq_l	%2,0(%3)\n"
-	"	extwl	%2,%4,%0\n"
-	"	mskwl	%2,%4,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%3)\n"
-	"	beq	%2,2f\n"
-#ifdef CONFIG_SMP
-	"	mb\n"
-#endif
-	".subsection 2\n"
-	"2:	br	1b\n"
-	".previous"
-	: "=&r" (ret), "=&r" (val), "=&r" (tmp), "=&r" (addr64)
-	: "r" ((long)m), "1" (val) : "memory");
-
-	return ret;
-}
-
-static inline unsigned long
-__xchg_u32(volatile int *m, unsigned long val)
-{
-	unsigned long dummy;
-
-	__asm__ __volatile__(
-	"1:	ldl_l %0,%4\n"
-	"	bis $31,%3,%1\n"
-	"	stl_c %1,%2\n"
-	"	beq %1,2f\n"
-#ifdef CONFIG_SMP
-	"	mb\n"
-#endif
-	".subsection 2\n"
-	"2:	br 1b\n"
-	".previous"
-	: "=&r" (val), "=&r" (dummy), "=m" (*m)
-	: "rI" (val), "m" (*m) : "memory");
-
-	return val;
-}
-
-static inline unsigned long
-__xchg_u64(volatile long *m, unsigned long val)
-{
-	unsigned long dummy;
-
-	__asm__ __volatile__(
-	"1:	ldq_l %0,%4\n"
-	"	bis $31,%3,%1\n"
-	"	stq_c %1,%2\n"
-	"	beq %1,2f\n"
-#ifdef CONFIG_SMP
-	"	mb\n"
-#endif
-	".subsection 2\n"
-	"2:	br 1b\n"
-	".previous"
-	: "=&r" (val), "=&r" (dummy), "=m" (*m)
-	: "rI" (val), "m" (*m) : "memory");
+#define __ASM__MB
+#define ____xchg(type, args...)		__xchg ## type ## _local(args)
+#define ____cmpxchg(type, args...)	__cmpxchg ## type ## _local(args)
+#include <asm/xchg.h>
 
-	return val;
-}
-
-/* This function doesn't exist, so you'll get a linker error
-   if something tries to do an invalid xchg().  */
-extern void __xchg_called_with_bad_pointer(void);
-
-#define __xchg(ptr, x, size) \
-({ \
-	unsigned long __xchg__res; \
-	volatile void *__xchg__ptr = (ptr); \
-	switch (size) { \
-		case 1: __xchg__res = __xchg_u8(__xchg__ptr, x); break; \
-		case 2: __xchg__res = __xchg_u16(__xchg__ptr, x); break; \
-		case 4: __xchg__res = __xchg_u32(__xchg__ptr, x); break; \
-		case 8: __xchg__res = __xchg_u64(__xchg__ptr, x); break; \
-		default: __xchg_called_with_bad_pointer(); __xchg__res = x; \
-	} \
-	__xchg__res; \
-})
-
-#define xchg(ptr,x)							     \
-  ({									     \
-     __typeof__(*(ptr)) _x_ = (x);					     \
-     (__typeof__(*(ptr))) __xchg((ptr), (unsigned long)_x_, sizeof(*(ptr))); \
+#define xchg_local(ptr,x)						\
+  ({									\
+     __typeof__(*(ptr)) _x_ = (x);					\
+     (__typeof__(*(ptr))) __xchg_local((ptr), (unsigned long)_x_,	\
+				       sizeof(*(ptr)));			\
   })
 
-static inline unsigned long
-__xchg_u8_local(volatile char *m, unsigned long val)
-{
-	unsigned long ret, tmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%4,7,%3\n"
-	"	insbl	%1,%4,%1\n"
-	"1:	ldq_l	%2,0(%3)\n"
-	"	extbl	%2,%4,%0\n"
-	"	mskbl	%2,%4,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%3)\n"
-	"	beq	%2,2f\n"
-	".subsection 2\n"
-	"2:	br	1b\n"
-	".previous"
-	: "=&r" (ret), "=&r" (val), "=&r" (tmp), "=&r" (addr64)
-	: "r" ((long)m), "1" (val) : "memory");
-
-	return ret;
-}
-
-static inline unsigned long
-__xchg_u16_local(volatile short *m, unsigned long val)
-{
-	unsigned long ret, tmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%4,7,%3\n"
-	"	inswl	%1,%4,%1\n"
-	"1:	ldq_l	%2,0(%3)\n"
-	"	extwl	%2,%4,%0\n"
-	"	mskwl	%2,%4,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%3)\n"
-	"	beq	%2,2f\n"
-	".subsection 2\n"
-	"2:	br	1b\n"
-	".previous"
-	: "=&r" (ret), "=&r" (val), "=&r" (tmp), "=&r" (addr64)
-	: "r" ((long)m), "1" (val) : "memory");
-
-	return ret;
-}
-
-static inline unsigned long
-__xchg_u32_local(volatile int *m, unsigned long val)
-{
-	unsigned long dummy;
-
-	__asm__ __volatile__(
-	"1:	ldl_l %0,%4\n"
-	"	bis $31,%3,%1\n"
-	"	stl_c %1,%2\n"
-	"	beq %1,2f\n"
-	".subsection 2\n"
-	"2:	br 1b\n"
-	".previous"
-	: "=&r" (val), "=&r" (dummy), "=m" (*m)
-	: "rI" (val), "m" (*m) : "memory");
-
-	return val;
-}
-
-static inline unsigned long
-__xchg_u64_local(volatile long *m, unsigned long val)
-{
-	unsigned long dummy;
-
-	__asm__ __volatile__(
-	"1:	ldq_l %0,%4\n"
-	"	bis $31,%3,%1\n"
-	"	stq_c %1,%2\n"
-	"	beq %1,2f\n"
-	".subsection 2\n"
-	"2:	br 1b\n"
-	".previous"
-	: "=&r" (val), "=&r" (dummy), "=m" (*m)
-	: "rI" (val), "m" (*m) : "memory");
-
-	return val;
-}
-
-#define __xchg_local(ptr, x, size) \
-({ \
-	unsigned long __xchg__res; \
-	volatile void *__xchg__ptr = (ptr); \
-	switch (size) { \
-		case 1: __xchg__res = __xchg_u8_local(__xchg__ptr, x); break; \
-		case 2: __xchg__res = __xchg_u16_local(__xchg__ptr, x); break; \
-		case 4: __xchg__res = __xchg_u32_local(__xchg__ptr, x); break; \
-		case 8: __xchg__res = __xchg_u64_local(__xchg__ptr, x); break; \
-		default: __xchg_called_with_bad_pointer(); __xchg__res = x; \
-	} \
-	__xchg__res; \
-})
-
-#define xchg_local(ptr,x)						     \
-  ({									     \
-     __typeof__(*(ptr)) _x_ = (x);					     \
-     (__typeof__(*(ptr))) __xchg_local((ptr), (unsigned long)_x_,	     \
-     		sizeof(*(ptr))); \
+#define cmpxchg_local(ptr, o, n)					\
+  ({									\
+     __typeof__(*(ptr)) _o_ = (o);					\
+     __typeof__(*(ptr)) _n_ = (n);					\
+     (__typeof__(*(ptr))) __cmpxchg_local((ptr), (unsigned long)_o_,	\
+					  (unsigned long)_n_,		\
+					  sizeof(*(ptr)));		\
   })
 
-/* 
- * Atomic compare and exchange.  Compare OLD with MEM, if identical,
- * store NEW in MEM.  Return the initial value in MEM.  Success is
- * indicated by comparing RETURN with OLD.
- *
- * The memory barrier should be placed in SMP only when we actually
- * make the change. If we don't change anything (so if the returned
- * prev is equal to old) then we aren't acquiring anything new and
- * we don't need any memory barrier as far I can tell.
- */
-
-#define __HAVE_ARCH_CMPXCHG 1
-
-static inline unsigned long
-__cmpxchg_u8(volatile char *m, long old, long new)
-{
-	unsigned long prev, tmp, cmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%5,7,%4\n"
-	"	insbl	%1,%5,%1\n"
-	"1:	ldq_l	%2,0(%4)\n"
-	"	extbl	%2,%5,%0\n"
-	"	cmpeq	%0,%6,%3\n"
-	"	beq	%3,2f\n"
-	"	mskbl	%2,%5,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%4)\n"
-	"	beq	%2,3f\n"
-#ifdef CONFIG_SMP
-	"	mb\n"
-#endif
-	"2:\n"
-	".subsection 2\n"
-	"3:	br	1b\n"
-	".previous"
-	: "=&r" (prev), "=&r" (new), "=&r" (tmp), "=&r" (cmp), "=&r" (addr64)
-	: "r" ((long)m), "Ir" (old), "1" (new) : "memory");
-
-	return prev;
-}
-
-static inline unsigned long
-__cmpxchg_u16(volatile short *m, long old, long new)
-{
-	unsigned long prev, tmp, cmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%5,7,%4\n"
-	"	inswl	%1,%5,%1\n"
-	"1:	ldq_l	%2,0(%4)\n"
-	"	extwl	%2,%5,%0\n"
-	"	cmpeq	%0,%6,%3\n"
-	"	beq	%3,2f\n"
-	"	mskwl	%2,%5,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%4)\n"
-	"	beq	%2,3f\n"
-#ifdef CONFIG_SMP
-	"	mb\n"
-#endif
-	"2:\n"
-	".subsection 2\n"
-	"3:	br	1b\n"
-	".previous"
-	: "=&r" (prev), "=&r" (new), "=&r" (tmp), "=&r" (cmp), "=&r" (addr64)
-	: "r" ((long)m), "Ir" (old), "1" (new) : "memory");
-
-	return prev;
-}
-
-static inline unsigned long
-__cmpxchg_u32(volatile int *m, int old, int new)
-{
-	unsigned long prev, cmp;
-
-	__asm__ __volatile__(
-	"1:	ldl_l %0,%5\n"
-	"	cmpeq %0,%3,%1\n"
-	"	beq %1,2f\n"
-	"	mov %4,%1\n"
-	"	stl_c %1,%2\n"
-	"	beq %1,3f\n"
-#ifdef CONFIG_SMP
-	"	mb\n"
-#endif
-	"2:\n"
-	".subsection 2\n"
-	"3:	br 1b\n"
-	".previous"
-	: "=&r"(prev), "=&r"(cmp), "=m"(*m)
-	: "r"((long) old), "r"(new), "m"(*m) : "memory");
-
-	return prev;
-}
+#define cmpxchg64_local(ptr, o, n)					\
+  ({									\
+	BUILD_BUG_ON(sizeof(*(ptr)) != 8);				\
+	cmpxchg_local((ptr), (o), (n));					\
+  })
 
-static inline unsigned long
-__cmpxchg_u64(volatile long *m, unsigned long old, unsigned long new)
-{
-	unsigned long prev, cmp;
-
-	__asm__ __volatile__(
-	"1:	ldq_l %0,%5\n"
-	"	cmpeq %0,%3,%1\n"
-	"	beq %1,2f\n"
-	"	mov %4,%1\n"
-	"	stq_c %1,%2\n"
-	"	beq %1,3f\n"
 #ifdef CONFIG_SMP
-	"	mb\n"
+#undef __ASM__MB
+#define __ASM__MB	"\tmb\n"
 #endif
-	"2:\n"
-	".subsection 2\n"
-	"3:	br 1b\n"
-	".previous"
-	: "=&r"(prev), "=&r"(cmp), "=m"(*m)
-	: "r"((long) old), "r"(new), "m"(*m) : "memory");
-
-	return prev;
-}
-
-/* This function doesn't exist, so you'll get a linker error
-   if something tries to do an invalid cmpxchg().  */
-extern void __cmpxchg_called_with_bad_pointer(void);
-
-static __always_inline unsigned long
-__cmpxchg(volatile void *ptr, unsigned long old, unsigned long new, int size)
-{
-	switch (size) {
-		case 1:
-			return __cmpxchg_u8(ptr, old, new);
-		case 2:
-			return __cmpxchg_u16(ptr, old, new);
-		case 4:
-			return __cmpxchg_u32(ptr, old, new);
-		case 8:
-			return __cmpxchg_u64(ptr, old, new);
-	}
-	__cmpxchg_called_with_bad_pointer();
-	return old;
-}
-
-#define cmpxchg(ptr, o, n)						 \
-  ({									 \
-     __typeof__(*(ptr)) _o_ = (o);					 \
-     __typeof__(*(ptr)) _n_ = (n);					 \
-     (__typeof__(*(ptr))) __cmpxchg((ptr), (unsigned long)_o_,		 \
-				    (unsigned long)_n_, sizeof(*(ptr))); \
+#undef ____xchg
+#undef ____cmpxchg
+#define ____xchg(type, args...)		__xchg ##type(args)
+#define ____cmpxchg(type, args...)	__cmpxchg ##type(args)
+#include <asm/xchg.h>
+
+#define xchg(ptr,x)							\
+  ({									\
+     __typeof__(*(ptr)) _x_ = (x);					\
+     (__typeof__(*(ptr))) __xchg((ptr), (unsigned long)_x_,		\
+				 sizeof(*(ptr)));			\
   })
-#define cmpxchg64(ptr, o, n)						 \
-  ({									 \
-	BUILD_BUG_ON(sizeof(*(ptr)) != 8);				 \
-	cmpxchg((ptr), (o), (n));					 \
-  })
-
-static inline unsigned long
-__cmpxchg_u8_local(volatile char *m, long old, long new)
-{
-	unsigned long prev, tmp, cmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%5,7,%4\n"
-	"	insbl	%1,%5,%1\n"
-	"1:	ldq_l	%2,0(%4)\n"
-	"	extbl	%2,%5,%0\n"
-	"	cmpeq	%0,%6,%3\n"
-	"	beq	%3,2f\n"
-	"	mskbl	%2,%5,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%4)\n"
-	"	beq	%2,3f\n"
-	"2:\n"
-	".subsection 2\n"
-	"3:	br	1b\n"
-	".previous"
-	: "=&r" (prev), "=&r" (new), "=&r" (tmp), "=&r" (cmp), "=&r" (addr64)
-	: "r" ((long)m), "Ir" (old), "1" (new) : "memory");
-
-	return prev;
-}
-
-static inline unsigned long
-__cmpxchg_u16_local(volatile short *m, long old, long new)
-{
-	unsigned long prev, tmp, cmp, addr64;
-
-	__asm__ __volatile__(
-	"	andnot	%5,7,%4\n"
-	"	inswl	%1,%5,%1\n"
-	"1:	ldq_l	%2,0(%4)\n"
-	"	extwl	%2,%5,%0\n"
-	"	cmpeq	%0,%6,%3\n"
-	"	beq	%3,2f\n"
-	"	mskwl	%2,%5,%2\n"
-	"	or	%1,%2,%2\n"
-	"	stq_c	%2,0(%4)\n"
-	"	beq	%2,3f\n"
-	"2:\n"
-	".subsection 2\n"
-	"3:	br	1b\n"
-	".previous"
-	: "=&r" (prev), "=&r" (new), "=&r" (tmp), "=&r" (cmp), "=&r" (addr64)
-	: "r" ((long)m), "Ir" (old), "1" (new) : "memory");
-
-	return prev;
-}
-
-static inline unsigned long
-__cmpxchg_u32_local(volatile int *m, int old, int new)
-{
-	unsigned long prev, cmp;
-
-	__asm__ __volatile__(
-	"1:	ldl_l %0,%5\n"
-	"	cmpeq %0,%3,%1\n"
-	"	beq %1,2f\n"
-	"	mov %4,%1\n"
-	"	stl_c %1,%2\n"
-	"	beq %1,3f\n"
-	"2:\n"
-	".subsection 2\n"
-	"3:	br 1b\n"
-	".previous"
-	: "=&r"(prev), "=&r"(cmp), "=m"(*m)
-	: "r"((long) old), "r"(new), "m"(*m) : "memory");
-
-	return prev;
-}
-
-static inline unsigned long
-__cmpxchg_u64_local(volatile long *m, unsigned long old, unsigned long new)
-{
-	unsigned long prev, cmp;
-
-	__asm__ __volatile__(
-	"1:	ldq_l %0,%5\n"
-	"	cmpeq %0,%3,%1\n"
-	"	beq %1,2f\n"
-	"	mov %4,%1\n"
-	"	stq_c %1,%2\n"
-	"	beq %1,3f\n"
-	"2:\n"
-	".subsection 2\n"
-	"3:	br 1b\n"
-	".previous"
-	: "=&r"(prev), "=&r"(cmp), "=m"(*m)
-	: "r"((long) old), "r"(new), "m"(*m) : "memory");
-
-	return prev;
-}
-
-static __always_inline unsigned long
-__cmpxchg_local(volatile void *ptr, unsigned long old, unsigned long new,
-		int size)
-{
-	switch (size) {
-		case 1:
-			return __cmpxchg_u8_local(ptr, old, new);
-		case 2:
-			return __cmpxchg_u16_local(ptr, old, new);
-		case 4:
-			return __cmpxchg_u32_local(ptr, old, new);
-		case 8:
-			return __cmpxchg_u64_local(ptr, old, new);
-	}
-	__cmpxchg_called_with_bad_pointer();
-	return old;
-}
 
-#define cmpxchg_local(ptr, o, n)					 \
-  ({									 \
-     __typeof__(*(ptr)) _o_ = (o);					 \
-     __typeof__(*(ptr)) _n_ = (n);					 \
-     (__typeof__(*(ptr))) __cmpxchg_local((ptr), (unsigned long)_o_,	 \
-				    (unsigned long)_n_, sizeof(*(ptr))); \
+#define cmpxchg(ptr, o, n)						\
+  ({									\
+     __typeof__(*(ptr)) _o_ = (o);					\
+     __typeof__(*(ptr)) _n_ = (n);					\
+     (__typeof__(*(ptr))) __cmpxchg((ptr), (unsigned long)_o_,		\
+				    (unsigned long)_n_,	sizeof(*(ptr)));\
   })
-#define cmpxchg64_local(ptr, o, n)					 \
-  ({									 \
-	BUILD_BUG_ON(sizeof(*(ptr)) != 8);				 \
-	cmpxchg_local((ptr), (o), (n));					 \
+
+#define cmpxchg64(ptr, o, n)						\
+  ({									\
+	BUILD_BUG_ON(sizeof(*(ptr)) != 8);				\
+	cmpxchg((ptr), (o), (n));					\
   })
 
+#undef __ASM__MB
+#undef ____cmpxchg
+
+#define __HAVE_ARCH_CMPXCHG 1
 
 #endif /* __ASSEMBLY__ */
 
diff --git a/arch/alpha/include/asm/types.h b/arch/alpha/include/asm/types.h
index c154135..f072f34 100644
--- a/arch/alpha/include/asm/types.h
+++ b/arch/alpha/include/asm/types.h
@@ -8,7 +8,12 @@
  * not a major issue.  However, for interoperability, libraries still
  * need to be careful to avoid a name clashes.
  */
+
+#ifdef __KERNEL__
+#include <asm-generic/int-ll64.h>
+#else
 #include <asm-generic/int-l64.h>
+#endif
 
 #ifndef __ASSEMBLY__
 
diff --git a/arch/alpha/include/asm/uaccess.h b/arch/alpha/include/asm/uaccess.h
index 22de3b4..163f305 100644
--- a/arch/alpha/include/asm/uaccess.h
+++ b/arch/alpha/include/asm/uaccess.h
@@ -498,13 +498,13 @@ struct exception_table_entry
 };
 
 /* Returns the new pc */
-#define fixup_exception(map_reg, fixup, pc)			\
+#define fixup_exception(map_reg, _fixup, pc)			\
 ({								\
-	if ((fixup)->fixup.bits.valreg != 31)			\
-		map_reg((fixup)->fixup.bits.valreg) = 0;	\
-	if ((fixup)->fixup.bits.errreg != 31)			\
-		map_reg((fixup)->fixup.bits.errreg) = -EFAULT;	\
-	(pc) + (fixup)->fixup.bits.nextinsn;			\
+	if ((_fixup)->fixup.bits.valreg != 31)			\
+		map_reg((_fixup)->fixup.bits.valreg) = 0;	\
+	if ((_fixup)->fixup.bits.errreg != 31)			\
+		map_reg((_fixup)->fixup.bits.errreg) = -EFAULT;	\
+	(pc) + (_fixup)->fixup.bits.nextinsn;			\
 })
 
 
diff --git a/arch/alpha/include/asm/xchg.h b/arch/alpha/include/asm/xchg.h
new file mode 100644
index 0000000..beba1b8
--- /dev/null
+++ b/arch/alpha/include/asm/xchg.h
@@ -0,0 +1,258 @@
+#ifndef __ALPHA_SYSTEM_H
+#error Do not include xchg.h directly!
+#else
+/*
+ * xchg/xchg_local and cmpxchg/cmpxchg_local share the same code
+ * except that local version do not have the expensive memory barrier.
+ * So this file is included twice from asm/system.h.
+ */
+
+/*
+ * Atomic exchange.
+ * Since it can be used to implement critical sections
+ * it must clobber "memory" (also for interrupts in UP).
+ */
+
+static inline unsigned long
+____xchg(_u8, volatile char *m, unsigned long val)
+{
+	unsigned long ret, tmp, addr64;
+
+	__asm__ __volatile__(
+	"	andnot	%4,7,%3\n"
+	"	insbl	%1,%4,%1\n"
+	"1:	ldq_l	%2,0(%3)\n"
+	"	extbl	%2,%4,%0\n"
+	"	mskbl	%2,%4,%2\n"
+	"	or	%1,%2,%2\n"
+	"	stq_c	%2,0(%3)\n"
+	"	beq	%2,2f\n"
+		__ASM__MB
+	".subsection 2\n"
+	"2:	br	1b\n"
+	".previous"
+	: "=&r" (ret), "=&r" (val), "=&r" (tmp), "=&r" (addr64)
+	: "r" ((long)m), "1" (val) : "memory");
+
+	return ret;
+}
+
+static inline unsigned long
+____xchg(_u16, volatile short *m, unsigned long val)
+{
+	unsigned long ret, tmp, addr64;
+
+	__asm__ __volatile__(
+	"	andnot	%4,7,%3\n"
+	"	inswl	%1,%4,%1\n"
+	"1:	ldq_l	%2,0(%3)\n"
+	"	extwl	%2,%4,%0\n"
+	"	mskwl	%2,%4,%2\n"
+	"	or	%1,%2,%2\n"
+	"	stq_c	%2,0(%3)\n"
+	"	beq	%2,2f\n"
+		__ASM__MB
+	".subsection 2\n"
+	"2:	br	1b\n"
+	".previous"
+	: "=&r" (ret), "=&r" (val), "=&r" (tmp), "=&r" (addr64)
+	: "r" ((long)m), "1" (val) : "memory");
+
+	return ret;
+}
+
+static inline unsigned long
+____xchg(_u32, volatile int *m, unsigned long val)
+{
+	unsigned long dummy;
+
+	__asm__ __volatile__(
+	"1:	ldl_l %0,%4\n"
+	"	bis $31,%3,%1\n"
+	"	stl_c %1,%2\n"
+	"	beq %1,2f\n"
+		__ASM__MB
+	".subsection 2\n"
+	"2:	br 1b\n"
+	".previous"
+	: "=&r" (val), "=&r" (dummy), "=m" (*m)
+	: "rI" (val), "m" (*m) : "memory");
+
+	return val;
+}
+
+static inline unsigned long
+____xchg(_u64, volatile long *m, unsigned long val)
+{
+	unsigned long dummy;
+
+	__asm__ __volatile__(
+	"1:	ldq_l %0,%4\n"
+	"	bis $31,%3,%1\n"
+	"	stq_c %1,%2\n"
+	"	beq %1,2f\n"
+		__ASM__MB
+	".subsection 2\n"
+	"2:	br 1b\n"
+	".previous"
+	: "=&r" (val), "=&r" (dummy), "=m" (*m)
+	: "rI" (val), "m" (*m) : "memory");
+
+	return val;
+}
+
+/* This function doesn't exist, so you'll get a linker error
+   if something tries to do an invalid xchg().  */
+extern void __xchg_called_with_bad_pointer(void);
+
+static __always_inline unsigned long
+____xchg(, volatile void *ptr, unsigned long x, int size)
+{
+	switch (size) {
+		case 1:
+			return ____xchg(_u8, ptr, x);
+		case 2:
+			return ____xchg(_u16, ptr, x);
+		case 4:
+			return ____xchg(_u32, ptr, x);
+		case 8:
+			return ____xchg(_u64, ptr, x);
+	}
+	__xchg_called_with_bad_pointer();
+	return x;
+}
+
+/*
+ * Atomic compare and exchange.  Compare OLD with MEM, if identical,
+ * store NEW in MEM.  Return the initial value in MEM.  Success is
+ * indicated by comparing RETURN with OLD.
+ *
+ * The memory barrier should be placed in SMP only when we actually
+ * make the change. If we don't change anything (so if the returned
+ * prev is equal to old) then we aren't acquiring anything new and
+ * we don't need any memory barrier as far I can tell.
+ */
+
+static inline unsigned long
+____cmpxchg(_u8, volatile char *m, unsigned char old, unsigned char new)
+{
+	unsigned long prev, tmp, cmp, addr64;
+
+	__asm__ __volatile__(
+	"	andnot	%5,7,%4\n"
+	"	insbl	%1,%5,%1\n"
+	"1:	ldq_l	%2,0(%4)\n"
+	"	extbl	%2,%5,%0\n"
+	"	cmpeq	%0,%6,%3\n"
+	"	beq	%3,2f\n"
+	"	mskbl	%2,%5,%2\n"
+	"	or	%1,%2,%2\n"
+	"	stq_c	%2,0(%4)\n"
+	"	beq	%2,3f\n"
+		__ASM__MB
+	"2:\n"
+	".subsection 2\n"
+	"3:	br	1b\n"
+	".previous"
+	: "=&r" (prev), "=&r" (new), "=&r" (tmp), "=&r" (cmp), "=&r" (addr64)
+	: "r" ((long)m), "Ir" (old), "1" (new) : "memory");
+
+	return prev;
+}
+
+static inline unsigned long
+____cmpxchg(_u16, volatile short *m, unsigned short old, unsigned short new)
+{
+	unsigned long prev, tmp, cmp, addr64;
+
+	__asm__ __volatile__(
+	"	andnot	%5,7,%4\n"
+	"	inswl	%1,%5,%1\n"
+	"1:	ldq_l	%2,0(%4)\n"
+	"	extwl	%2,%5,%0\n"
+	"	cmpeq	%0,%6,%3\n"
+	"	beq	%3,2f\n"
+	"	mskwl	%2,%5,%2\n"
+	"	or	%1,%2,%2\n"
+	"	stq_c	%2,0(%4)\n"
+	"	beq	%2,3f\n"
+		__ASM__MB
+	"2:\n"
+	".subsection 2\n"
+	"3:	br	1b\n"
+	".previous"
+	: "=&r" (prev), "=&r" (new), "=&r" (tmp), "=&r" (cmp), "=&r" (addr64)
+	: "r" ((long)m), "Ir" (old), "1" (new) : "memory");
+
+	return prev;
+}
+
+static inline unsigned long
+____cmpxchg(_u32, volatile int *m, int old, int new)
+{
+	unsigned long prev, cmp;
+
+	__asm__ __volatile__(
+	"1:	ldl_l %0,%5\n"
+	"	cmpeq %0,%3,%1\n"
+	"	beq %1,2f\n"
+	"	mov %4,%1\n"
+	"	stl_c %1,%2\n"
+	"	beq %1,3f\n"
+		__ASM__MB
+	"2:\n"
+	".subsection 2\n"
+	"3:	br 1b\n"
+	".previous"
+	: "=&r"(prev), "=&r"(cmp), "=m"(*m)
+	: "r"((long) old), "r"(new), "m"(*m) : "memory");
+
+	return prev;
+}
+
+static inline unsigned long
+____cmpxchg(_u64, volatile long *m, unsigned long old, unsigned long new)
+{
+	unsigned long prev, cmp;
+
+	__asm__ __volatile__(
+	"1:	ldq_l %0,%5\n"
+	"	cmpeq %0,%3,%1\n"
+	"	beq %1,2f\n"
+	"	mov %4,%1\n"
+	"	stq_c %1,%2\n"
+	"	beq %1,3f\n"
+		__ASM__MB
+	"2:\n"
+	".subsection 2\n"
+	"3:	br 1b\n"
+	".previous"
+	: "=&r"(prev), "=&r"(cmp), "=m"(*m)
+	: "r"((long) old), "r"(new), "m"(*m) : "memory");
+
+	return prev;
+}
+
+/* This function doesn't exist, so you'll get a linker error
+   if something tries to do an invalid cmpxchg().  */
+extern void __cmpxchg_called_with_bad_pointer(void);
+
+static __always_inline unsigned long
+____cmpxchg(, volatile void *ptr, unsigned long old, unsigned long new,
+	      int size)
+{
+	switch (size) {
+		case 1:
+			return ____cmpxchg(_u8, ptr, old, new);
+		case 2:
+			return ____cmpxchg(_u16, ptr, old, new);
+		case 4:
+			return ____cmpxchg(_u32, ptr, old, new);
+		case 8:
+			return ____cmpxchg(_u64, ptr, old, new);
+	}
+	__cmpxchg_called_with_bad_pointer();
+	return old;
+}
+
+#endif
diff --git a/arch/alpha/kernel/Makefile b/arch/alpha/kernel/Makefile
index b469775..a427538 100644
--- a/arch/alpha/kernel/Makefile
+++ b/arch/alpha/kernel/Makefile
@@ -12,7 +12,7 @@ obj-y    := entry.o traps.o process.o init_task.o osf_sys.o irq.o \
 
 obj-$(CONFIG_VGA_HOSE)	+= console.o
 obj-$(CONFIG_SMP)	+= smp.o
-obj-$(CONFIG_PCI)	+= pci.o pci_iommu.o
+obj-$(CONFIG_PCI)	+= pci.o pci_iommu.o pci-sysfs.o
 obj-$(CONFIG_SRM_ENV)	+= srm_env.o
 obj-$(CONFIG_MODULES)	+= module.o
 
diff --git a/arch/alpha/kernel/err_ev6.c b/arch/alpha/kernel/err_ev6.c
index 11aee01..985e5c1 100644
--- a/arch/alpha/kernel/err_ev6.c
+++ b/arch/alpha/kernel/err_ev6.c
@@ -157,8 +157,8 @@ ev6_parse_cbox(u64 c_addr, u64 c1_syn, u64 c2_syn,
 		       err_print_prefix,
 		       streamname[stream], bitsname[bits], sourcename[source]);
 
-	printk("%s    Address: 0x%016lx\n"
-	         "    Syndrome[upper.lower]: %02lx.%02lx\n", 
+	printk("%s    Address: 0x%016llx\n"
+	         "    Syndrome[upper.lower]: %02llx.%02llx\n",
 	       err_print_prefix,
 	       c_addr,
 	       c2_syn, c1_syn);
diff --git a/arch/alpha/kernel/err_ev7.c b/arch/alpha/kernel/err_ev7.c
index 68cd493..73770c6 100644
--- a/arch/alpha/kernel/err_ev7.c
+++ b/arch/alpha/kernel/err_ev7.c
@@ -246,13 +246,13 @@ ev7_process_pal_subpacket(struct el_subpacket *header)
 
 	switch(header->type) {
 	case EL_TYPE__PAL__LOGOUT_FRAME:
-		printk("%s*** MCHK occurred on LPID %ld (RBOX %lx)\n",
+		printk("%s*** MCHK occurred on LPID %ld (RBOX %llx)\n",
 		       err_print_prefix,
 		       packet->by_type.logout.whami, 
 		       packet->by_type.logout.rbox_whami);
 		el_print_timestamp(&packet->by_type.logout.timestamp);
-		printk("%s  EXC_ADDR: %016lx\n"
-		         "  HALT_CODE: %lx\n", 
+		printk("%s  EXC_ADDR: %016llx\n"
+		         "  HALT_CODE: %llx\n",
 		       err_print_prefix,
 		       packet->by_type.logout.exc_addr,
 		       packet->by_type.logout.halt_code);
diff --git a/arch/alpha/kernel/err_marvel.c b/arch/alpha/kernel/err_marvel.c
index 413bf37..6bfd243 100644
--- a/arch/alpha/kernel/err_marvel.c
+++ b/arch/alpha/kernel/err_marvel.c
@@ -129,7 +129,7 @@ marvel_print_po7_crrct_sym(u64 crrct_sym)
 
 
 	printk("%s      Correctable Error Symptoms:\n"
-	       "%s        Syndrome: 0x%lx\n",
+	       "%s        Syndrome: 0x%llx\n",
 	       err_print_prefix,
 	       err_print_prefix, EXTRACT(crrct_sym, IO7__PO7_CRRCT_SYM__SYN));
 	marvel_print_err_cyc(EXTRACT(crrct_sym, IO7__PO7_CRRCT_SYM__ERR_CYC));
@@ -186,7 +186,7 @@ marvel_print_po7_uncrr_sym(u64 uncrr_sym, u64 valid_mask)
 	uncrr_sym &= valid_mask;
 
 	if (EXTRACT(valid_mask, IO7__PO7_UNCRR_SYM__SYN))
-		printk("%s        Syndrome: 0x%lx\n",
+		printk("%s        Syndrome: 0x%llx\n",
 		       err_print_prefix, 
 		       EXTRACT(uncrr_sym, IO7__PO7_UNCRR_SYM__SYN));
 
@@ -307,7 +307,7 @@ marvel_print_po7_ugbge_sym(u64 ugbge_sym)
 		sprintf(opcode_str, "BlkIO");
 		break;
 	default:
-		sprintf(opcode_str, "0x%lx\n", 
+		sprintf(opcode_str, "0x%llx\n",
 			EXTRACT(ugbge_sym, IO7__PO7_UGBGE_SYM__UPH_OPCODE));
 		break;
 	}
@@ -321,7 +321,7 @@ marvel_print_po7_ugbge_sym(u64 ugbge_sym)
 	       opcode_str);
 
 	if (0xC5 != EXTRACT(ugbge_sym, IO7__PO7_UGBGE_SYM__UPH_OPCODE))
-		printk("%s        Packet Offset 0x%08lx\n",
+		printk("%s        Packet Offset 0x%08llx\n",
 		       err_print_prefix,
 		       EXTRACT(ugbge_sym, IO7__PO7_UGBGE_SYM__UPH_PKT_OFF));
 }
@@ -480,8 +480,8 @@ marvel_print_po7_err_sum(struct ev7_pal_io_subpacket *io)
 		printk("%s    Lost Error\n", err_print_prefix);
 
 	printk("%s    Failing Packet:\n"
-	       "%s      Cycle 1: %016lx\n"
-	       "%s      Cycle 2: %016lx\n",
+	       "%s      Cycle 1: %016llx\n"
+	       "%s      Cycle 2: %016llx\n",
 	       err_print_prefix,
 	       err_print_prefix, io->po7_err_pkt0,
 	       err_print_prefix, io->po7_err_pkt1);
@@ -515,9 +515,9 @@ marvel_print_pox_tlb_err(u64 tlb_err)
 	if (!(tlb_err & IO7__POX_TLBERR__ERR_VALID))
 		return;
 
-	printk("%s      TLB Error on index 0x%lx:\n"
+	printk("%s      TLB Error on index 0x%llx:\n"
 	       "%s        - %s\n"
-	       "%s        - Addr: 0x%016lx\n",
+	       "%s        - Addr: 0x%016llx\n",
 	       err_print_prefix,
 	       EXTRACT(tlb_err, IO7__POX_TLBERR__ERR_TLB_PTR),
 	       err_print_prefix,
@@ -579,7 +579,7 @@ marvel_print_pox_spl_cmplt(u64 spl_cmplt)
 		sprintf(message, "Uncorrectable Split Write Data Error");
 		break;
 	default:
-		sprintf(message, "%08lx\n", 
+		sprintf(message, "%08llx\n",
 			EXTRACT(spl_cmplt, IO7__POX_SPLCMPLT__MESSAGE));
 		break;
 	}
@@ -620,9 +620,9 @@ marvel_print_pox_trans_sum(u64 trans_sum)
 		return;
 
 	printk("%s      Transaction Summary:\n"
-	       "%s        Command: 0x%lx - %s\n"
-	       "%s        Address: 0x%016lx%s\n"
-	       "%s        PCI-X Master Slot: 0x%lx\n",
+	       "%s        Command: 0x%llx - %s\n"
+	       "%s        Address: 0x%016llx%s\n"
+	       "%s        PCI-X Master Slot: 0x%llx\n",
 	       err_print_prefix, 
 	       err_print_prefix, 
 	       EXTRACT(trans_sum, IO7__POX_TRANSUM__PCIX_CMD),
@@ -964,12 +964,12 @@ marvel_process_io_error(struct ev7_lf_subpackets *lf_subpackets, int print)
 
 #if 0
 		printk("%s  PORT 7 ERROR:\n"
-		       "%s    PO7_ERROR_SUM: %016lx\n"
-		       "%s    PO7_UNCRR_SYM: %016lx\n"
-		       "%s    PO7_CRRCT_SYM: %016lx\n"
-		       "%s    PO7_UGBGE_SYM: %016lx\n"
-		       "%s    PO7_ERR_PKT0:  %016lx\n"
-		       "%s    PO7_ERR_PKT1:  %016lx\n",
+		       "%s    PO7_ERROR_SUM: %016llx\n"
+		       "%s    PO7_UNCRR_SYM: %016llx\n"
+		       "%s    PO7_CRRCT_SYM: %016llx\n"
+		       "%s    PO7_UGBGE_SYM: %016llx\n"
+		       "%s    PO7_ERR_PKT0:  %016llx\n"
+		       "%s    PO7_ERR_PKT1:  %016llx\n",
 		       err_print_prefix,
 		       err_print_prefix, io->po7_error_sum,
 		       err_print_prefix, io->po7_uncrr_sym,
@@ -987,12 +987,12 @@ marvel_process_io_error(struct ev7_lf_subpackets *lf_subpackets, int print)
 		if (!MARVEL_IO_ERR_VALID(io->ports[i].pox_err_sum))
 			continue;
 
-		printk("%s  PID %u PORT %d POx_ERR_SUM: %016lx\n", 
+		printk("%s  PID %u PORT %d POx_ERR_SUM: %016llx\n",
 		       err_print_prefix, 
 		       lf_subpackets->io_pid, i, io->ports[i].pox_err_sum);
 		marvel_print_pox_err(io->ports[i].pox_err_sum, &io->ports[i]);
 
-		printk("%s  [ POx_FIRST_ERR: %016lx ]\n", 
+		printk("%s  [ POx_FIRST_ERR: %016llx ]\n",
 		       err_print_prefix, io->ports[i].pox_first_err);
 		marvel_print_pox_err(io->ports[i].pox_first_err, 
 				     &io->ports[i]);
diff --git a/arch/alpha/kernel/err_titan.c b/arch/alpha/kernel/err_titan.c
index 257449e..c7e28a8 100644
--- a/arch/alpha/kernel/err_titan.c
+++ b/arch/alpha/kernel/err_titan.c
@@ -107,12 +107,12 @@ titan_parse_p_serror(int which, u64 serror, int print)
 	if (!print)
 		return status;
 
-	printk("%s  PChip %d SERROR: %016lx\n", 
+	printk("%s  PChip %d SERROR: %016llx\n",
 	       err_print_prefix, which, serror);
 	if (serror & TITAN__PCHIP_SERROR__ECCMASK) {
 		printk("%s    %sorrectable ECC Error:\n"
 		       "      Source: %-6s  Command: %-8s  Syndrome: 0x%08x\n"
-		       "      Address: 0x%lx\n", 
+		       "      Address: 0x%llx\n",
 		       err_print_prefix,
 		       (serror & TITAN__PCHIP_SERROR__UECC) ? "Unc" : "C",
 		       serror_src[EXTRACT(serror, TITAN__PCHIP_SERROR__SRC)],
@@ -223,7 +223,7 @@ titan_parse_p_perror(int which, int port, u64 perror, int print)
 	if (!print) 
 		return status;
 
-	printk("%s  PChip %d %cPERROR: %016lx\n", 
+	printk("%s  PChip %d %cPERROR: %016llx\n",
 	       err_print_prefix, which, 
 	       port ? 'A' : 'G', perror);
 	if (perror & TITAN__PCHIP_PERROR__IPTPW)
@@ -316,7 +316,7 @@ titan_parse_p_agperror(int which, u64 agperror, int print)
 	addr = EXTRACT(agperror, TITAN__PCHIP_AGPERROR__ADDR) << 3;
 	len = EXTRACT(agperror, TITAN__PCHIP_AGPERROR__LEN);
 
-	printk("%s  PChip %d AGPERROR: %016lx\n", err_print_prefix,
+	printk("%s  PChip %d AGPERROR: %016llx\n", err_print_prefix,
 	       which, agperror);
 	if (agperror & TITAN__PCHIP_AGPERROR__NOWINDOW)
 		printk("%s    No Window\n", err_print_prefix);
@@ -597,16 +597,16 @@ privateer_process_680_frame(struct el_common *mchk_header, int print)
 		return status;
 
 	/* TODO - decode instead of just dumping... */
-	printk("%s  Summary Flags:         %016lx\n"
- 	         "  CChip DIRx:            %016lx\n"
-		 "  System Management IR:  %016lx\n"
-		 "  CPU IR:                %016lx\n"
-		 "  Power Supply IR:       %016lx\n"
-		 "  LM78 Fault Status:     %016lx\n"
-		 "  System Doors:          %016lx\n"
-		 "  Temperature Warning:   %016lx\n"
-		 "  Fan Control:           %016lx\n"
-		 "  Fatal Power Down Code: %016lx\n",
+	printk("%s  Summary Flags:         %016llx\n"
+ 	         "  CChip DIRx:            %016llx\n"
+		 "  System Management IR:  %016llx\n"
+		 "  CPU IR:                %016llx\n"
+		 "  Power Supply IR:       %016llx\n"
+		 "  LM78 Fault Status:     %016llx\n"
+		 "  System Doors:          %016llx\n"
+		 "  Temperature Warning:   %016llx\n"
+		 "  Fan Control:           %016llx\n"
+		 "  Fatal Power Down Code: %016llx\n",
 	       err_print_prefix,
 	       emchk->summary,
 	       emchk->c_dirx,
diff --git a/arch/alpha/kernel/pci-sysfs.c b/arch/alpha/kernel/pci-sysfs.c
new file mode 100644
index 0000000..6ea822e
--- /dev/null
+++ b/arch/alpha/kernel/pci-sysfs.c
@@ -0,0 +1,366 @@
+/*
+ * arch/alpha/kernel/pci-sysfs.c
+ *
+ * Copyright (C) 2009 Ivan Kokshaysky
+ *
+ * Alpha PCI resource files.
+ *
+ * Loosely based on generic HAVE_PCI_MMAP implementation in
+ * drivers/pci/pci-sysfs.c
+ */
+
+#include <linux/sched.h>
+#include <linux/pci.h>
+
+static int hose_mmap_page_range(struct pci_controller *hose,
+				struct vm_area_struct *vma,
+				enum pci_mmap_state mmap_type, int sparse)
+{
+	unsigned long base;
+
+	if (mmap_type == pci_mmap_mem)
+		base = sparse ? hose->sparse_mem_base : hose->dense_mem_base;
+	else
+		base = sparse ? hose->sparse_io_base : hose->dense_io_base;
+
+	vma->vm_pgoff += base >> PAGE_SHIFT;
+	vma->vm_flags |= (VM_IO | VM_RESERVED);
+
+	return io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff,
+				  vma->vm_end - vma->vm_start,
+				  vma->vm_page_prot);
+}
+
+static int __pci_mmap_fits(struct pci_dev *pdev, int num,
+			   struct vm_area_struct *vma, int sparse)
+{
+	unsigned long nr, start, size;
+	int shift = sparse ? 5 : 0;
+
+	nr = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
+	start = vma->vm_pgoff;
+	size = ((pci_resource_len(pdev, num) - 1) >> (PAGE_SHIFT - shift)) + 1;
+
+	if (start < size && size - start >= nr)
+		return 1;
+	WARN(1, "process \"%s\" tried to map%s 0x%08lx-0x%08lx on %s BAR %d "
+		"(size 0x%08lx)\n",
+		current->comm, sparse ? " sparse" : "", start, start + nr,
+		pci_name(pdev), num, size);
+	return 0;
+}
+
+/**
+ * pci_mmap_resource - map a PCI resource into user memory space
+ * @kobj: kobject for mapping
+ * @attr: struct bin_attribute for the file being mapped
+ * @vma: struct vm_area_struct passed into the mmap
+ * @sparse: address space type
+ *
+ * Use the bus mapping routines to map a PCI resource into userspace.
+ */
+static int pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr,
+			     struct vm_area_struct *vma, int sparse)
+{
+	struct pci_dev *pdev = to_pci_dev(container_of(kobj,
+						       struct device, kobj));
+	struct resource *res = (struct resource *)attr->private;
+	enum pci_mmap_state mmap_type;
+	struct pci_bus_region bar;
+	int i;
+
+	for (i = 0; i < PCI_ROM_RESOURCE; i++)
+		if (res == &pdev->resource[i])
+			break;
+	if (i >= PCI_ROM_RESOURCE)
+		return -ENODEV;
+
+	if (!__pci_mmap_fits(pdev, i, vma, sparse))
+		return -EINVAL;
+
+	if (iomem_is_exclusive(res->start))
+		return -EINVAL;
+
+	pcibios_resource_to_bus(pdev, &bar, res);
+	vma->vm_pgoff += bar.start >> (PAGE_SHIFT - (sparse ? 5 : 0));
+	mmap_type = res->flags & IORESOURCE_MEM ? pci_mmap_mem : pci_mmap_io;
+
+	return hose_mmap_page_range(pdev->sysdata, vma, mmap_type, sparse);
+}
+
+static int pci_mmap_resource_sparse(struct kobject *kobj,
+				    struct bin_attribute *attr,
+				    struct vm_area_struct *vma)
+{
+	return pci_mmap_resource(kobj, attr, vma, 1);
+}
+
+static int pci_mmap_resource_dense(struct kobject *kobj,
+				   struct bin_attribute *attr,
+				   struct vm_area_struct *vma)
+{
+	return pci_mmap_resource(kobj, attr, vma, 0);
+}
+
+/**
+ * pci_remove_resource_files - cleanup resource files
+ * @dev: dev to cleanup
+ *
+ * If we created resource files for @dev, remove them from sysfs and
+ * free their resources.
+ */
+void pci_remove_resource_files(struct pci_dev *pdev)
+{
+	int i;
+
+	for (i = 0; i < PCI_ROM_RESOURCE; i++) {
+		struct bin_attribute *res_attr;
+
+		res_attr = pdev->res_attr[i];
+		if (res_attr) {
+			sysfs_remove_bin_file(&pdev->dev.kobj, res_attr);
+			kfree(res_attr);
+		}
+
+		res_attr = pdev->res_attr_wc[i];
+		if (res_attr) {
+			sysfs_remove_bin_file(&pdev->dev.kobj, res_attr);
+			kfree(res_attr);
+		}
+	}
+}
+
+static int sparse_mem_mmap_fits(struct pci_dev *pdev, int num)
+{
+	struct pci_bus_region bar;
+	struct pci_controller *hose = pdev->sysdata;
+	long dense_offset;
+	unsigned long sparse_size;
+
+	pcibios_resource_to_bus(pdev, &bar, &pdev->resource[num]);
+
+	/* All core logic chips have 4G sparse address space, except
+	   CIA which has 16G (see xxx_SPARSE_MEM and xxx_DENSE_MEM
+	   definitions in asm/core_xxx.h files). This corresponds
+	   to 128M or 512M of the bus space. */
+	dense_offset = (long)(hose->dense_mem_base - hose->sparse_mem_base);
+	sparse_size = dense_offset >= 0x400000000UL ? 0x20000000 : 0x8000000;
+
+	return bar.end < sparse_size;
+}
+
+static int pci_create_one_attr(struct pci_dev *pdev, int num, char *name,
+			       char *suffix, struct bin_attribute *res_attr,
+			       unsigned long sparse)
+{
+	size_t size = pci_resource_len(pdev, num);
+
+	sprintf(name, "resource%d%s", num, suffix);
+	res_attr->mmap = sparse ? pci_mmap_resource_sparse :
+				  pci_mmap_resource_dense;
+	res_attr->attr.name = name;
+	res_attr->attr.mode = S_IRUSR | S_IWUSR;
+	res_attr->size = sparse ? size << 5 : size;
+	res_attr->private = &pdev->resource[num];
+	return sysfs_create_bin_file(&pdev->dev.kobj, res_attr);
+}
+
+static int pci_create_attr(struct pci_dev *pdev, int num)
+{
+	/* allocate attribute structure, piggyback attribute name */
+	int retval, nlen1, nlen2 = 0, res_count = 1;
+	unsigned long sparse_base, dense_base;
+	struct bin_attribute *attr;
+	struct pci_controller *hose = pdev->sysdata;
+	char *suffix, *attr_name;
+
+	suffix = "";	/* Assume bwx machine, normal resourceN files. */
+	nlen1 = 10;
+
+	if (pdev->resource[num].flags & IORESOURCE_MEM) {
+		sparse_base = hose->sparse_mem_base;
+		dense_base = hose->dense_mem_base;
+		if (sparse_base && !sparse_mem_mmap_fits(pdev, num)) {
+			sparse_base = 0;
+			suffix = "_dense";
+			nlen1 = 16;	/* resourceN_dense */
+		}
+	} else {
+		sparse_base = hose->sparse_io_base;
+		dense_base = hose->dense_io_base;
+	}
+
+	if (sparse_base) {
+		suffix = "_sparse";
+		nlen1 = 17;
+		if (dense_base) {
+			nlen2 = 16;	/* resourceN_dense */
+			res_count = 2;
+		}
+	}
+
+	attr = kzalloc(sizeof(*attr) * res_count + nlen1 + nlen2, GFP_ATOMIC);
+	if (!attr)
+		return -ENOMEM;
+
+	/* Create bwx, sparse or single dense file */
+	attr_name = (char *)(attr + res_count);
+	pdev->res_attr[num] = attr;
+	retval = pci_create_one_attr(pdev, num, attr_name, suffix, attr,
+				     sparse_base);
+	if (retval || res_count == 1)
+		return retval;
+
+	/* Create dense file */
+	attr_name += nlen1;
+	attr++;
+	pdev->res_attr_wc[num] = attr;
+	return pci_create_one_attr(pdev, num, attr_name, "_dense", attr, 0);
+}
+
+/**
+ * pci_create_resource_files - create resource files in sysfs for @dev
+ * @dev: dev in question
+ *
+ * Walk the resources in @dev creating files for each resource available.
+ */
+int pci_create_resource_files(struct pci_dev *pdev)
+{
+	int i;
+	int retval;
+
+	/* Expose the PCI resources from this device as files */
+	for (i = 0; i < PCI_ROM_RESOURCE; i++) {
+
+		/* skip empty resources */
+		if (!pci_resource_len(pdev, i))
+			continue;
+
+		retval = pci_create_attr(pdev, i);
+		if (retval) {
+			pci_remove_resource_files(pdev);
+			return retval;
+		}
+	}
+	return 0;
+}
+
+/* Legacy I/O bus mapping stuff. */
+
+static int __legacy_mmap_fits(struct pci_controller *hose,
+			      struct vm_area_struct *vma,
+			      unsigned long res_size, int sparse)
+{
+	unsigned long nr, start, size;
+
+	nr = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
+	start = vma->vm_pgoff;
+	size = ((res_size - 1) >> PAGE_SHIFT) + 1;
+
+	if (start < size && size - start >= nr)
+		return 1;
+	WARN(1, "process \"%s\" tried to map%s 0x%08lx-0x%08lx on hose %d "
+		"(size 0x%08lx)\n",
+		current->comm, sparse ? " sparse" : "", start, start + nr,
+		hose->index, size);
+	return 0;
+}
+
+static inline int has_sparse(struct pci_controller *hose,
+			     enum pci_mmap_state mmap_type)
+{
+	unsigned long base;
+
+	base = (mmap_type == pci_mmap_mem) ? hose->sparse_mem_base :
+					     hose->sparse_io_base;
+
+	return base != 0;
+}
+
+int pci_mmap_legacy_page_range(struct pci_bus *bus, struct vm_area_struct *vma,
+			       enum pci_mmap_state mmap_type)
+{
+	struct pci_controller *hose = bus->sysdata;
+	int sparse = has_sparse(hose, mmap_type);
+	unsigned long res_size;
+
+	res_size = (mmap_type == pci_mmap_mem) ? bus->legacy_mem->size :
+						 bus->legacy_io->size;
+	if (!__legacy_mmap_fits(hose, vma, res_size, sparse))
+		return -EINVAL;
+
+	return hose_mmap_page_range(hose, vma, mmap_type, sparse);
+}
+
+/**
+ * pci_adjust_legacy_attr - adjustment of legacy file attributes
+ * @b: bus to create files under
+ * @mmap_type: I/O port or memory
+ *
+ * Adjust file name and size for sparse mappings.
+ */
+void pci_adjust_legacy_attr(struct pci_bus *bus, enum pci_mmap_state mmap_type)
+{
+	struct pci_controller *hose = bus->sysdata;
+
+	if (!has_sparse(hose, mmap_type))
+		return;
+
+	if (mmap_type == pci_mmap_mem) {
+		bus->legacy_mem->attr.name = "legacy_mem_sparse";
+		bus->legacy_mem->size <<= 5;
+	} else {
+		bus->legacy_io->attr.name = "legacy_io_sparse";
+		bus->legacy_io->size <<= 5;
+	}
+	return;
+}
+
+/* Legacy I/O bus read/write functions */
+int pci_legacy_read(struct pci_bus *bus, loff_t port, u32 *val, size_t size)
+{
+	struct pci_controller *hose = bus->sysdata;
+
+	port += hose->io_space->start;
+
+	switch(size) {
+	case 1:
+		*((u8 *)val) = inb(port);
+		return 1;
+	case 2:
+		if (port & 1)
+			return -EINVAL;
+		*((u16 *)val) = inw(port);
+		return 2;
+	case 4:
+		if (port & 3)
+			return -EINVAL;
+		*((u32 *)val) = inl(port);
+		return 4;
+	}
+	return -EINVAL;
+}
+
+int pci_legacy_write(struct pci_bus *bus, loff_t port, u32 val, size_t size)
+{
+	struct pci_controller *hose = bus->sysdata;
+
+	port += hose->io_space->start;
+
+	switch(size) {
+	case 1:
+		outb(port, val);
+		return 1;
+	case 2:
+		if (port & 1)
+			return -EINVAL;
+		outw(port, val);
+		return 2;
+	case 4:
+		if (port & 3)
+			return -EINVAL;
+		outl(port, val);
+		return 4;
+	}
+	return -EINVAL;
+}
diff --git a/arch/alpha/kernel/pci.c b/arch/alpha/kernel/pci.c
index a3b9388..a91ba28 100644
--- a/arch/alpha/kernel/pci.c
+++ b/arch/alpha/kernel/pci.c
@@ -168,7 +168,7 @@ pcibios_align_resource(void *data, struct resource *res,
 		 */
 
 		/* Align to multiple of size of minimum base.  */
-		alignto = max(0x1000UL, align);
+		alignto = max_t(resource_size_t, 0x1000, align);
 		start = ALIGN(start, alignto);
 		if (hose->sparse_mem_base && size <= 7 * 16*MB) {
 			if (((start / (16*MB)) & 0x7) == 0) {
diff --git a/arch/alpha/kernel/pci_iommu.c b/arch/alpha/kernel/pci_iommu.c
index b9094da..bfb880a 100644
--- a/arch/alpha/kernel/pci_iommu.c
+++ b/arch/alpha/kernel/pci_iommu.c
@@ -247,7 +247,7 @@ pci_map_single_1(struct pci_dev *pdev, void *cpu_addr, size_t size,
 	    && paddr + size <= __direct_map_size) {
 		ret = paddr + __direct_map_base;
 
-		DBGA2("pci_map_single: [%p,%lx] -> direct %lx from %p\n",
+		DBGA2("pci_map_single: [%p,%zx] -> direct %llx from %p\n",
 		      cpu_addr, size, ret, __builtin_return_address(0));
 
 		return ret;
@@ -258,7 +258,7 @@ pci_map_single_1(struct pci_dev *pdev, void *cpu_addr, size_t size,
 	if (dac_allowed) {
 		ret = paddr + alpha_mv.pci_dac_offset;
 
-		DBGA2("pci_map_single: [%p,%lx] -> DAC %lx from %p\n",
+		DBGA2("pci_map_single: [%p,%zx] -> DAC %llx from %p\n",
 		      cpu_addr, size, ret, __builtin_return_address(0));
 
 		return ret;
@@ -299,7 +299,7 @@ pci_map_single_1(struct pci_dev *pdev, void *cpu_addr, size_t size,
 	ret = arena->dma_base + dma_ofs * PAGE_SIZE;
 	ret += (unsigned long)cpu_addr & ~PAGE_MASK;
 
-	DBGA2("pci_map_single: [%p,%lx] np %ld -> sg %lx from %p\n",
+	DBGA2("pci_map_single: [%p,%zx] np %ld -> sg %llx from %p\n",
 	      cpu_addr, size, npages, ret, __builtin_return_address(0));
 
 	return ret;
@@ -355,14 +355,14 @@ pci_unmap_single(struct pci_dev *pdev, dma_addr_t dma_addr, size_t size,
 	    && dma_addr < __direct_map_base + __direct_map_size) {
 		/* Nothing to do.  */
 
-		DBGA2("pci_unmap_single: direct [%lx,%lx] from %p\n",
+		DBGA2("pci_unmap_single: direct [%llx,%zx] from %p\n",
 		      dma_addr, size, __builtin_return_address(0));
 
 		return;
 	}
 
 	if (dma_addr > 0xffffffff) {
-		DBGA2("pci64_unmap_single: DAC [%lx,%lx] from %p\n",
+		DBGA2("pci64_unmap_single: DAC [%llx,%zx] from %p\n",
 		      dma_addr, size, __builtin_return_address(0));
 		return;
 	}
@@ -373,9 +373,9 @@ pci_unmap_single(struct pci_dev *pdev, dma_addr_t dma_addr, size_t size,
 
 	dma_ofs = (dma_addr - arena->dma_base) >> PAGE_SHIFT;
 	if (dma_ofs * PAGE_SIZE >= arena->size) {
-		printk(KERN_ERR "Bogus pci_unmap_single: dma_addr %lx "
-		       " base %lx size %x\n", dma_addr, arena->dma_base,
-		       arena->size);
+		printk(KERN_ERR "Bogus pci_unmap_single: dma_addr %llx "
+		       " base %llx size %x\n",
+		       dma_addr, arena->dma_base, arena->size);
 		return;
 		BUG();
 	}
@@ -394,7 +394,7 @@ pci_unmap_single(struct pci_dev *pdev, dma_addr_t dma_addr, size_t size,
 
 	spin_unlock_irqrestore(&arena->lock, flags);
 
-	DBGA2("pci_unmap_single: sg [%lx,%lx] np %ld from %p\n",
+	DBGA2("pci_unmap_single: sg [%llx,%zx] np %ld from %p\n",
 	      dma_addr, size, npages, __builtin_return_address(0));
 }
 EXPORT_SYMBOL(pci_unmap_single);
@@ -444,7 +444,7 @@ try_again:
 		goto try_again;
 	}
 		
-	DBGA2("pci_alloc_consistent: %lx -> [%p,%x] from %p\n",
+	DBGA2("pci_alloc_consistent: %zx -> [%p,%llx] from %p\n",
 	      size, cpu_addr, *dma_addrp, __builtin_return_address(0));
 
 	return cpu_addr;
@@ -464,7 +464,7 @@ pci_free_consistent(struct pci_dev *pdev, size_t size, void *cpu_addr,
 	pci_unmap_single(pdev, dma_addr, size, PCI_DMA_BIDIRECTIONAL);
 	free_pages((unsigned long)cpu_addr, get_order(size));
 
-	DBGA2("pci_free_consistent: [%x,%lx] from %p\n",
+	DBGA2("pci_free_consistent: [%llx,%zx] from %p\n",
 	      dma_addr, size, __builtin_return_address(0));
 }
 EXPORT_SYMBOL(pci_free_consistent);
@@ -551,7 +551,7 @@ sg_fill(struct device *dev, struct scatterlist *leader, struct scatterlist *end,
 		out->dma_address = paddr + __direct_map_base;
 		out->dma_length = size;
 
-		DBGA("    sg_fill: [%p,%lx] -> direct %lx\n",
+		DBGA("    sg_fill: [%p,%lx] -> direct %llx\n",
 		     __va(paddr), size, out->dma_address);
 
 		return 0;
@@ -563,7 +563,7 @@ sg_fill(struct device *dev, struct scatterlist *leader, struct scatterlist *end,
 		out->dma_address = paddr + alpha_mv.pci_dac_offset;
 		out->dma_length = size;
 
-		DBGA("    sg_fill: [%p,%lx] -> DAC %lx\n",
+		DBGA("    sg_fill: [%p,%lx] -> DAC %llx\n",
 		     __va(paddr), size, out->dma_address);
 
 		return 0;
@@ -589,7 +589,7 @@ sg_fill(struct device *dev, struct scatterlist *leader, struct scatterlist *end,
 	out->dma_address = arena->dma_base + dma_ofs*PAGE_SIZE + paddr;
 	out->dma_length = size;
 
-	DBGA("    sg_fill: [%p,%lx] -> sg %lx np %ld\n",
+	DBGA("    sg_fill: [%p,%lx] -> sg %llx np %ld\n",
 	     __va(paddr), size, out->dma_address, npages);
 
 	/* All virtually contiguous.  We need to find the length of each
@@ -752,7 +752,7 @@ pci_unmap_sg(struct pci_dev *pdev, struct scatterlist *sg, int nents,
 
 		if (addr > 0xffffffff) {
 			/* It's a DAC address -- nothing to do.  */
-			DBGA("    (%ld) DAC [%lx,%lx]\n",
+			DBGA("    (%ld) DAC [%llx,%zx]\n",
 			      sg - end + nents, addr, size);
 			continue;
 		}
@@ -760,12 +760,12 @@ pci_unmap_sg(struct pci_dev *pdev, struct scatterlist *sg, int nents,
 		if (addr >= __direct_map_base
 		    && addr < __direct_map_base + __direct_map_size) {
 			/* Nothing to do.  */
-			DBGA("    (%ld) direct [%lx,%lx]\n",
+			DBGA("    (%ld) direct [%llx,%zx]\n",
 			      sg - end + nents, addr, size);
 			continue;
 		}
 
-		DBGA("    (%ld) sg [%lx,%lx]\n",
+		DBGA("    (%ld) sg [%llx,%zx]\n",
 		     sg - end + nents, addr, size);
 
 		npages = iommu_num_pages(addr, size, PAGE_SIZE);
diff --git a/arch/alpha/kernel/proto.h b/arch/alpha/kernel/proto.h
index fe14c67..567f259 100644
--- a/arch/alpha/kernel/proto.h
+++ b/arch/alpha/kernel/proto.h
@@ -20,7 +20,7 @@ struct pci_controller;
 extern struct pci_ops apecs_pci_ops;
 extern void apecs_init_arch(void);
 extern void apecs_pci_clr_err(void);
-extern void apecs_machine_check(u64, u64);
+extern void apecs_machine_check(unsigned long vector, unsigned long la_ptr);
 extern void apecs_pci_tbi(struct pci_controller *, dma_addr_t, dma_addr_t);
 
 /* core_cia.c */
@@ -29,7 +29,7 @@ extern void cia_init_pci(void);
 extern void cia_init_arch(void);
 extern void pyxis_init_arch(void);
 extern void cia_kill_arch(int);
-extern void cia_machine_check(u64, u64);
+extern void cia_machine_check(unsigned long vector, unsigned long la_ptr);
 extern void cia_pci_tbi(struct pci_controller *, dma_addr_t, dma_addr_t);
 
 /* core_irongate.c */
@@ -42,7 +42,7 @@ extern void irongate_machine_check(u64, u64);
 /* core_lca.c */
 extern struct pci_ops lca_pci_ops;
 extern void lca_init_arch(void);
-extern void lca_machine_check(u64, u64);
+extern void lca_machine_check(unsigned long vector, unsigned long la_ptr);
 extern void lca_pci_tbi(struct pci_controller *, dma_addr_t, dma_addr_t);
 
 /* core_marvel.c */
@@ -64,7 +64,7 @@ void io7_clear_errors(struct io7 *io7);
 extern struct pci_ops mcpcia_pci_ops;
 extern void mcpcia_init_arch(void);
 extern void mcpcia_init_hoses(void);
-extern void mcpcia_machine_check(u64, u64);
+extern void mcpcia_machine_check(unsigned long vector, unsigned long la_ptr);
 extern void mcpcia_pci_tbi(struct pci_controller *, dma_addr_t, dma_addr_t);
 
 /* core_polaris.c */
@@ -72,14 +72,14 @@ extern struct pci_ops polaris_pci_ops;
 extern int polaris_read_config_dword(struct pci_dev *, int, u32 *);
 extern int polaris_write_config_dword(struct pci_dev *, int, u32);
 extern void polaris_init_arch(void);
-extern void polaris_machine_check(u64, u64);
+extern void polaris_machine_check(unsigned long vector, unsigned long la_ptr);
 #define polaris_pci_tbi ((void *)0)
 
 /* core_t2.c */
 extern struct pci_ops t2_pci_ops;
 extern void t2_init_arch(void);
 extern void t2_kill_arch(int);
-extern void t2_machine_check(u64, u64);
+extern void t2_machine_check(unsigned long vector, unsigned long la_ptr);
 extern void t2_pci_tbi(struct pci_controller *, dma_addr_t, dma_addr_t);
 
 /* core_titan.c */
@@ -94,14 +94,14 @@ extern struct _alpha_agp_info *titan_agp_info(void);
 extern struct pci_ops tsunami_pci_ops;
 extern void tsunami_init_arch(void);
 extern void tsunami_kill_arch(int);
-extern void tsunami_machine_check(u64, u64);
+extern void tsunami_machine_check(unsigned long vector, unsigned long la_ptr);
 extern void tsunami_pci_tbi(struct pci_controller *, dma_addr_t, dma_addr_t);
 
 /* core_wildfire.c */
 extern struct pci_ops wildfire_pci_ops;
 extern void wildfire_init_arch(void);
 extern void wildfire_kill_arch(int);
-extern void wildfire_machine_check(u64, u64);
+extern void wildfire_machine_check(unsigned long vector, unsigned long la_ptr);
 extern void wildfire_pci_tbi(struct pci_controller *, dma_addr_t, dma_addr_t);
 extern int wildfire_pa_to_nid(unsigned long);
 extern int wildfire_cpuid_to_nid(int);
diff --git a/arch/alpha/kernel/setup.c b/arch/alpha/kernel/setup.c
index 02bee69..80df86c 100644
--- a/arch/alpha/kernel/setup.c
+++ b/arch/alpha/kernel/setup.c
@@ -1255,7 +1255,7 @@ show_cpuinfo(struct seq_file *f, void *slot)
 		       platform_string(), nr_processors);
 
 #ifdef CONFIG_SMP
-	seq_printf(f, "cpus active\t\t: %d\n"
+	seq_printf(f, "cpus active\t\t: %u\n"
 		      "cpu active mask\t\t: %016lx\n",
 		       num_online_cpus(), cpus_addr(cpu_possible_map)[0]);
 #endif
diff --git a/arch/alpha/kernel/smc37c669.c b/arch/alpha/kernel/smc37c669.c
index fd467b2..bca5bda 100644
--- a/arch/alpha/kernel/smc37c669.c
+++ b/arch/alpha/kernel/smc37c669.c
@@ -2542,8 +2542,8 @@ void __init SMC669_Init ( int index )
         SMC37c669_display_device_info( );
 #endif
 	local_irq_restore(flags);
-        printk( "SMC37c669 Super I/O Controller found @ 0x%lx\n",
-		(unsigned long) SMC_base );
+        printk( "SMC37c669 Super I/O Controller found @ 0x%p\n",
+		SMC_base );
     }
     else {
 	local_irq_restore(flags);
diff --git a/arch/alpha/kernel/srm_env.c b/arch/alpha/kernel/srm_env.c
index 78ad7cd..d12af47 100644
--- a/arch/alpha/kernel/srm_env.c
+++ b/arch/alpha/kernel/srm_env.c
@@ -218,7 +218,6 @@ srm_env_init(void)
 				BASE_DIR);
 		goto cleanup;
 	}
-	base_dir->owner = THIS_MODULE;
 
 	/*
 	 * Create per-name subdirectory
@@ -229,7 +228,6 @@ srm_env_init(void)
 				BASE_DIR, NAMED_DIR);
 		goto cleanup;
 	}
-	named_dir->owner = THIS_MODULE;
 
 	/*
 	 * Create per-number subdirectory
@@ -241,7 +239,6 @@ srm_env_init(void)
 		goto cleanup;
 
 	}
-	numbered_dir->owner = THIS_MODULE;
 
 	/*
 	 * Create all named nodes
@@ -254,7 +251,6 @@ srm_env_init(void)
 			goto cleanup;
 
 		entry->proc_entry->data		= (void *) entry;
-		entry->proc_entry->owner	= THIS_MODULE;
 		entry->proc_entry->read_proc	= srm_env_read;
 		entry->proc_entry->write_proc	= srm_env_write;
 
@@ -275,7 +271,6 @@ srm_env_init(void)
 
 		entry->id			= var_num;
 		entry->proc_entry->data		= (void *) entry;
-		entry->proc_entry->owner	= THIS_MODULE;
 		entry->proc_entry->read_proc	= srm_env_read;
 		entry->proc_entry->write_proc	= srm_env_write;
 	}
diff --git a/arch/alpha/kernel/sys_jensen.c b/arch/alpha/kernel/sys_jensen.c
index e2516f9..2b5caf3 100644
--- a/arch/alpha/kernel/sys_jensen.c
+++ b/arch/alpha/kernel/sys_jensen.c
@@ -244,12 +244,11 @@ jensen_init_arch(void)
 }
 
 static void
-jensen_machine_check (u64 vector, u64 la)
+jensen_machine_check(unsigned long vector, unsigned long la)
 {
 	printk(KERN_CRIT "Machine check\n");
 }
 
-
 /*
  * The System Vector
  */
diff --git a/arch/alpha/kernel/sys_sable.c b/arch/alpha/kernel/sys_sable.c
index d232e42..9e26325 100644
--- a/arch/alpha/kernel/sys_sable.c
+++ b/arch/alpha/kernel/sys_sable.c
@@ -453,7 +453,7 @@ sable_lynx_enable_irq(unsigned int irq)
 	sable_lynx_irq_swizzle->update_irq_hw(bit, mask);
 	spin_unlock(&sable_lynx_irq_lock);
 #if 0
-	printk("%s: mask 0x%lx bit 0x%x irq 0x%x\n",
+	printk("%s: mask 0x%lx bit 0x%lx irq 0x%x\n",
 	       __func__, mask, bit, irq);
 #endif
 }
@@ -469,7 +469,7 @@ sable_lynx_disable_irq(unsigned int irq)
 	sable_lynx_irq_swizzle->update_irq_hw(bit, mask);
 	spin_unlock(&sable_lynx_irq_lock);
 #if 0
-	printk("%s: mask 0x%lx bit 0x%x irq 0x%x\n",
+	printk("%s: mask 0x%lx bit 0x%lx irq 0x%x\n",
 	       __func__, mask, bit, irq);
 #endif
 }
diff --git a/arch/alpha/kernel/traps.c b/arch/alpha/kernel/traps.c
index cefc5a3..6ee7655 100644
--- a/arch/alpha/kernel/traps.c
+++ b/arch/alpha/kernel/traps.c
@@ -623,7 +623,7 @@ do_entUna(void * va, unsigned long opcode, unsigned long reg,
 	}
 
 	lock_kernel();
-	printk("Bad unaligned kernel access at %016lx: %p %lx %ld\n",
+	printk("Bad unaligned kernel access at %016lx: %p %lx %lu\n",
 		pc, va, opcode, reg);
 	do_exit(SIGSEGV);
 
diff --git a/arch/avr32/mm/fault.c b/arch/avr32/mm/fault.c
index ce4e429..62d4abb 100644
--- a/arch/avr32/mm/fault.c
+++ b/arch/avr32/mm/fault.c
@@ -250,21 +250,3 @@ asmlinkage void do_bus_error(unsigned long addr, int write_access,
 	dump_dtlb();
 	die("Bus Error", regs, SIGKILL);
 }
-
-/*
- * This functionality is currently not possible to implement because
- * we're using segmentation to ensure a fixed mapping of the kernel
- * virtual address space.
- *
- * It would be possible to implement this, but it would require us to
- * disable segmentation at startup and load the kernel mappings into
- * the TLB like any other pages. There will be lots of trickery to
- * avoid recursive invocation of the TLB miss handler, though...
- */
-#ifdef CONFIG_DEBUG_PAGEALLOC
-void kernel_map_pages(struct page *page, int numpages, int enable)
-{
-
-}
-EXPORT_SYMBOL(kernel_map_pages);
-#endif
diff --git a/arch/blackfin/mm/sram-alloc.c b/arch/blackfin/mm/sram-alloc.c
index 834cab7..530d139 100644
--- a/arch/blackfin/mm/sram-alloc.c
+++ b/arch/blackfin/mm/sram-alloc.c
@@ -854,7 +854,6 @@ static int __init sram_proc_init(void)
 		printk(KERN_WARNING "unable to create /proc/sram\n");
 		return -1;
 	}
-	ptr->owner = THIS_MODULE;
 	ptr->read_proc = sram_proc_read;
 	return 0;
 }
diff --git a/arch/cris/arch-v10/kernel/time.c b/arch/cris/arch-v10/kernel/time.c
index c685ba4..2b73c7a 100644
--- a/arch/cris/arch-v10/kernel/time.c
+++ b/arch/cris/arch-v10/kernel/time.c
@@ -261,7 +261,6 @@ timer_interrupt(int irq, void *dev_id)
 static struct irqaction irq2  = {
 	.handler = timer_interrupt,
 	.flags = IRQF_SHARED | IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "timer",
 };
 
diff --git a/arch/cris/arch-v32/kernel/smp.c b/arch/cris/arch-v32/kernel/smp.c
index 9dac173..f59a973 100644
--- a/arch/cris/arch-v32/kernel/smp.c
+++ b/arch/cris/arch-v32/kernel/smp.c
@@ -65,7 +65,6 @@ static int send_ipi(int vector, int wait, cpumask_t cpu_mask);
 static struct irqaction irq_ipi  = {
 	.handler = crisv32_ipi_interrupt,
 	.flags = IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "ipi",
 };
 
diff --git a/arch/cris/arch-v32/kernel/time.c b/arch/cris/arch-v32/kernel/time.c
index 3a13dd6..65633d0 100644
--- a/arch/cris/arch-v32/kernel/time.c
+++ b/arch/cris/arch-v32/kernel/time.c
@@ -267,7 +267,6 @@ timer_interrupt(int irq, void *dev_id)
 static struct irqaction irq_timer = {
 	.handler = timer_interrupt,
 	.flags = IRQF_SHARED | IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "timer"
 };
 
diff --git a/arch/frv/kernel/irq-mb93091.c b/arch/frv/kernel/irq-mb93091.c
index 9e38f99..4dd9ada 100644
--- a/arch/frv/kernel/irq-mb93091.c
+++ b/arch/frv/kernel/irq-mb93091.c
@@ -109,28 +109,24 @@ static struct irqaction fpga_irq[4]  = {
 	[0] = {
 		.handler	= fpga_interrupt,
 		.flags		= IRQF_DISABLED | IRQF_SHARED,
-		.mask		= CPU_MASK_NONE,
 		.name		= "fpga.0",
 		.dev_id		= (void *) 0x0028UL,
 	},
 	[1] = {
 		.handler	= fpga_interrupt,
 		.flags		= IRQF_DISABLED | IRQF_SHARED,
-		.mask		= CPU_MASK_NONE,
 		.name		= "fpga.1",
 		.dev_id		= (void *) 0x0050UL,
 	},
 	[2] = {
 		.handler	= fpga_interrupt,
 		.flags		= IRQF_DISABLED | IRQF_SHARED,
-		.mask		= CPU_MASK_NONE,
 		.name		= "fpga.2",
 		.dev_id		= (void *) 0x1c00UL,
 	},
 	[3] = {
 		.handler	= fpga_interrupt,
 		.flags		= IRQF_DISABLED | IRQF_SHARED,
-		.mask		= CPU_MASK_NONE,
 		.name		= "fpga.3",
 		.dev_id		= (void *) 0x6386UL,
 	}
diff --git a/arch/frv/kernel/irq-mb93093.c b/arch/frv/kernel/irq-mb93093.c
index 3c2752c..e452090 100644
--- a/arch/frv/kernel/irq-mb93093.c
+++ b/arch/frv/kernel/irq-mb93093.c
@@ -108,7 +108,6 @@ static struct irqaction fpga_irq[1]  = {
 	[0] = {
 		.handler	= fpga_interrupt,
 		.flags		= IRQF_DISABLED,
-		.mask		= CPU_MASK_NONE,
 		.name		= "fpga.0",
 		.dev_id		= (void *) 0x0700UL,
 	}
diff --git a/arch/frv/kernel/irq-mb93493.c b/arch/frv/kernel/irq-mb93493.c
index 7754c73..ba55ecd 100644
--- a/arch/frv/kernel/irq-mb93493.c
+++ b/arch/frv/kernel/irq-mb93493.c
@@ -120,14 +120,12 @@ static struct irqaction mb93493_irq[2]  = {
 	[0] = {
 		.handler	= mb93493_interrupt,
 		.flags		= IRQF_DISABLED | IRQF_SHARED,
-		.mask		= CPU_MASK_NONE,
 		.name		= "mb93493.0",
 		.dev_id		= (void *) __addr_MB93493_IQSR(0),
 	},
 	[1] = {
 		.handler	= mb93493_interrupt,
 		.flags		= IRQF_DISABLED | IRQF_SHARED,
-		.mask		= CPU_MASK_NONE,
 		.name		= "mb93493.1",
 		.dev_id		= (void *) __addr_MB93493_IQSR(1),
 	}
diff --git a/arch/frv/kernel/time.c b/arch/frv/kernel/time.c
index 69f6a4e..fb0ce75 100644
--- a/arch/frv/kernel/time.c
+++ b/arch/frv/kernel/time.c
@@ -45,7 +45,6 @@ static irqreturn_t timer_interrupt(int irq, void *dummy);
 static struct irqaction timer_irq  = {
 	.handler = timer_interrupt,
 	.flags = IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "timer",
 };
 
diff --git a/arch/h8300/kernel/timer/itu.c b/arch/h8300/kernel/timer/itu.c
index d1c9265..4883ba7 100644
--- a/arch/h8300/kernel/timer/itu.c
+++ b/arch/h8300/kernel/timer/itu.c
@@ -60,7 +60,6 @@ static struct irqaction itu_irq = {
 	.name		= "itu",
 	.handler	= timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
-	.mask		= CPU_MASK_NONE,
 };
 
 static const int __initdata divide_rate[] = {1, 2, 4, 8};
diff --git a/arch/h8300/kernel/timer/timer16.c b/arch/h8300/kernel/timer/timer16.c
index e14271b..042dbb5 100644
--- a/arch/h8300/kernel/timer/timer16.c
+++ b/arch/h8300/kernel/timer/timer16.c
@@ -55,7 +55,6 @@ static struct irqaction timer16_irq = {
 	.name		= "timer-16",
 	.handler	= timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
-	.mask		= CPU_MASK_NONE,
 };
 
 static const int __initdata divide_rate[] = {1, 2, 4, 8};
diff --git a/arch/h8300/kernel/timer/timer8.c b/arch/h8300/kernel/timer/timer8.c
index 0556d7c..38be0ca 100644
--- a/arch/h8300/kernel/timer/timer8.c
+++ b/arch/h8300/kernel/timer/timer8.c
@@ -75,7 +75,6 @@ static struct irqaction timer8_irq = {
 	.name		= "timer-8",
 	.handler	= timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
-	.mask		= CPU_MASK_NONE,
 };
 
 static const int __initdata divide_rate[] = {8, 64, 8192};
diff --git a/arch/h8300/kernel/timer/tpu.c b/arch/h8300/kernel/timer/tpu.c
index df7f453..ad383ca 100644
--- a/arch/h8300/kernel/timer/tpu.c
+++ b/arch/h8300/kernel/timer/tpu.c
@@ -65,7 +65,6 @@ static struct irqaction tpu_irq = {
 	.name		= "tpu",
 	.handler	= timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
-	.mask		= CPU_MASK_NONE,
 };
 
 const static int __initdata divide_rate[] = {
diff --git a/arch/ia64/hp/sim/simserial.c b/arch/ia64/hp/sim/simserial.c
index 24b1ad5..2bef526 100644
--- a/arch/ia64/hp/sim/simserial.c
+++ b/arch/ia64/hp/sim/simserial.c
@@ -24,6 +24,7 @@
 #include <linux/major.h>
 #include <linux/fcntl.h>
 #include <linux/mm.h>
+#include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/capability.h>
 #include <linux/console.h>
@@ -848,38 +849,36 @@ static int rs_open(struct tty_struct *tty, struct file * filp)
  * /proc fs routines....
  */
 
-static inline int line_info(char *buf, struct serial_state *state)
+static inline void line_info(struct seq_file *m, struct serial_state *state)
 {
-	return sprintf(buf, "%d: uart:%s port:%lX irq:%d\n",
+	seq_printf(m, "%d: uart:%s port:%lX irq:%d\n",
 		       state->line, uart_config[state->type].name,
 		       state->port, state->irq);
 }
 
-static int rs_read_proc(char *page, char **start, off_t off, int count,
-		 int *eof, void *data)
+static int rs_proc_show(struct seq_file *m, void *v)
 {
-	int i, len = 0, l;
-	off_t	begin = 0;
-
-	len += sprintf(page, "simserinfo:1.0 driver:%s\n", serial_version);
-	for (i = 0; i < NR_PORTS && len < 4000; i++) {
-		l = line_info(page + len, &rs_table[i]);
-		len += l;
-		if (len+begin > off+count)
-			goto done;
-		if (len+begin < off) {
-			begin += len;
-			len = 0;
-		}
-	}
-	*eof = 1;
-done:
-	if (off >= len+begin)
-		return 0;
-	*start = page + (begin-off);
-	return ((count < begin+len-off) ? count : begin+len-off);
+	int i;
+
+	seq_printf(m, "simserinfo:1.0 driver:%s\n", serial_version);
+	for (i = 0; i < NR_PORTS; i++)
+		line_info(m, &rs_table[i]);
+	return 0;
 }
 
+static int rs_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, rs_proc_show, NULL);
+}
+
+static const struct file_operations rs_proc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= rs_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
 /*
  * ---------------------------------------------------------------------
  * rs_init() and friends
@@ -917,7 +916,7 @@ static const struct tty_operations hp_ops = {
 	.start = rs_start,
 	.hangup = rs_hangup,
 	.wait_until_sent = rs_wait_until_sent,
-	.read_proc = rs_read_proc,
+	.proc_fops = &rs_proc_fops,
 };
 
 /*
diff --git a/arch/ia64/include/asm/intrinsics.h b/arch/ia64/include/asm/intrinsics.h
index c47830e..111ed52 100644
--- a/arch/ia64/include/asm/intrinsics.h
+++ b/arch/ia64/include/asm/intrinsics.h
@@ -202,7 +202,11 @@ extern long ia64_cmpxchg_called_with_bad_pointer (void);
 
 #ifndef __ASSEMBLY__
 #if defined(CONFIG_PARAVIRT) && defined(__KERNEL__)
-#define IA64_INTRINSIC_API(name)	pv_cpu_ops.name
+#ifdef ASM_SUPPORTED
+# define IA64_INTRINSIC_API(name)	paravirt_ ## name
+#else
+# define IA64_INTRINSIC_API(name)	pv_cpu_ops.name
+#endif
 #define IA64_INTRINSIC_MACRO(name)	paravirt_ ## name
 #else
 #define IA64_INTRINSIC_API(name)	ia64_native_ ## name
diff --git a/arch/ia64/include/asm/mmu_context.h b/arch/ia64/include/asm/mmu_context.h
index 040bc87..7f2a456 100644
--- a/arch/ia64/include/asm/mmu_context.h
+++ b/arch/ia64/include/asm/mmu_context.h
@@ -87,7 +87,7 @@ get_mmu_context (struct mm_struct *mm)
 	/* re-check, now that we've got the lock: */
 	context = mm->context;
 	if (context == 0) {
-		cpus_clear(mm->cpu_vm_mask);
+		cpumask_clear(mm_cpumask(mm));
 		if (ia64_ctx.next >= ia64_ctx.limit) {
 			ia64_ctx.next = find_next_zero_bit(ia64_ctx.bitmap,
 					ia64_ctx.max_ctx, ia64_ctx.next);
@@ -166,8 +166,8 @@ activate_context (struct mm_struct *mm)
 
 	do {
 		context = get_mmu_context(mm);
-		if (!cpu_isset(smp_processor_id(), mm->cpu_vm_mask))
-			cpu_set(smp_processor_id(), mm->cpu_vm_mask);
+		if (!cpumask_test_cpu(smp_processor_id(), mm_cpumask(mm)))
+			cpumask_set_cpu(smp_processor_id(), mm_cpumask(mm));
 		reload_context(context);
 		/*
 		 * in the unlikely event of a TLB-flush by another thread,
diff --git a/arch/ia64/include/asm/module.h b/arch/ia64/include/asm/module.h
index d2da61e..908eaef 100644
--- a/arch/ia64/include/asm/module.h
+++ b/arch/ia64/include/asm/module.h
@@ -16,6 +16,12 @@ struct mod_arch_specific {
 	struct elf64_shdr *got;		/* global offset table */
 	struct elf64_shdr *opd;		/* official procedure descriptors */
 	struct elf64_shdr *unwind;	/* unwind-table section */
+#ifdef CONFIG_PARAVIRT
+	struct elf64_shdr *paravirt_bundles;
+					/* paravirt_alt_bundle_patch table */
+	struct elf64_shdr *paravirt_insts;
+					/* paravirt_alt_inst_patch table */
+#endif
 	unsigned long gp;		/* global-pointer for module */
 
 	void *core_unw_table;		/* core unwind-table cookie returned by unwinder */
diff --git a/arch/ia64/include/asm/native/inst.h b/arch/ia64/include/asm/native/inst.h
index 0a1026c..d2d46ef 100644
--- a/arch/ia64/include/asm/native/inst.h
+++ b/arch/ia64/include/asm/native/inst.h
@@ -30,6 +30,9 @@
 #define __paravirt_work_processed_syscall_target \
 						ia64_work_processed_syscall
 
+#define paravirt_fsyscall_table			ia64_native_fsyscall_table
+#define paravirt_fsys_bubble_down		ia64_native_fsys_bubble_down
+
 #ifdef CONFIG_PARAVIRT_GUEST_ASM_CLOBBER_CHECK
 # define PARAVIRT_POISON	0xdeadbeefbaadf00d
 # define CLOBBER(clob)				\
@@ -74,6 +77,11 @@
 (pred)	mov reg = psr			\
 	CLOBBER(clob)
 
+#define MOV_FROM_ITC(pred, pred_clob, reg, clob)	\
+(pred)	mov reg = ar.itc				\
+	CLOBBER(clob)					\
+	CLOBBER_PRED(pred_clob)
+
 #define MOV_TO_IFA(reg, clob)	\
 	mov cr.ifa = reg	\
 	CLOBBER(clob)
@@ -158,6 +166,11 @@
 #define RSM_PSR_DT		\
 	rsm psr.dt
 
+#define RSM_PSR_BE_I(clob0, clob1)	\
+	rsm psr.be | psr.i		\
+	CLOBBER(clob0)			\
+	CLOBBER(clob1)
+
 #define SSM_PSR_DT_AND_SRLZ_I	\
 	ssm psr.dt		\
 	;;			\
diff --git a/arch/ia64/include/asm/native/patchlist.h b/arch/ia64/include/asm/native/patchlist.h
new file mode 100644
index 0000000..be16ca9
--- /dev/null
+++ b/arch/ia64/include/asm/native/patchlist.h
@@ -0,0 +1,38 @@
+/******************************************************************************
+ * arch/ia64/include/asm/native/inst.h
+ *
+ * Copyright (c) 2008 Isaku Yamahata <yamahata at valinux co jp>
+ *                    VA Linux Systems Japan K.K.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that 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
+ *
+ */
+
+#define __paravirt_start_gate_fsyscall_patchlist		\
+	__ia64_native_start_gate_fsyscall_patchlist
+#define __paravirt_end_gate_fsyscall_patchlist			\
+	__ia64_native_end_gate_fsyscall_patchlist
+#define __paravirt_start_gate_brl_fsys_bubble_down_patchlist	\
+	__ia64_native_start_gate_brl_fsys_bubble_down_patchlist
+#define __paravirt_end_gate_brl_fsys_bubble_down_patchlist	\
+	__ia64_native_end_gate_brl_fsys_bubble_down_patchlist
+#define __paravirt_start_gate_vtop_patchlist			\
+	__ia64_native_start_gate_vtop_patchlist
+#define __paravirt_end_gate_vtop_patchlist			\
+	__ia64_native_end_gate_vtop_patchlist
+#define __paravirt_start_gate_mckinley_e9_patchlist		\
+	__ia64_native_start_gate_mckinley_e9_patchlist
+#define __paravirt_end_gate_mckinley_e9_patchlist		\
+	__ia64_native_end_gate_mckinley_e9_patchlist
diff --git a/arch/ia64/include/asm/native/pvchk_inst.h b/arch/ia64/include/asm/native/pvchk_inst.h
index b8e6eb1..8d72962 100644
--- a/arch/ia64/include/asm/native/pvchk_inst.h
+++ b/arch/ia64/include/asm/native/pvchk_inst.h
@@ -180,6 +180,11 @@
 	IS_PRED_IN(pred)			\
 	IS_RREG_OUT(reg)			\
 	IS_RREG_CLOB(clob)
+#define MOV_FROM_ITC(pred, pred_clob, reg, clob)	\
+	IS_PRED_IN(pred)				\
+	IS_PRED_CLOB(pred_clob)				\
+	IS_RREG_OUT(reg)				\
+	IS_RREG_CLOB(clob)
 #define MOV_TO_IFA(reg, clob)			\
 	IS_RREG_IN(reg)				\
 	IS_RREG_CLOB(clob)
@@ -246,6 +251,9 @@
 	IS_RREG_CLOB(clob2)
 #define RSM_PSR_DT				\
 	nop 0
+#define RSM_PSR_BE_I(clob0, clob1)		\
+	IS_RREG_CLOB(clob0)			\
+	IS_RREG_CLOB(clob1)
 #define SSM_PSR_DT_AND_SRLZ_I			\
 	nop 0
 #define BSW_0(clob0, clob1, clob2)		\
diff --git a/arch/ia64/include/asm/paravirt.h b/arch/ia64/include/asm/paravirt.h
index 2bf3636..2eb0a98 100644
--- a/arch/ia64/include/asm/paravirt.h
+++ b/arch/ia64/include/asm/paravirt.h
@@ -22,6 +22,56 @@
 #ifndef __ASM_PARAVIRT_H
 #define __ASM_PARAVIRT_H
 
+#ifndef __ASSEMBLY__
+/******************************************************************************
+ * fsys related addresses
+ */
+struct pv_fsys_data {
+	unsigned long *fsyscall_table;
+	void *fsys_bubble_down;
+};
+
+extern struct pv_fsys_data pv_fsys_data;
+
+unsigned long *paravirt_get_fsyscall_table(void);
+char *paravirt_get_fsys_bubble_down(void);
+
+/******************************************************************************
+ * patchlist addresses for gate page
+ */
+enum pv_gate_patchlist {
+	PV_GATE_START_FSYSCALL,
+	PV_GATE_END_FSYSCALL,
+
+	PV_GATE_START_BRL_FSYS_BUBBLE_DOWN,
+	PV_GATE_END_BRL_FSYS_BUBBLE_DOWN,
+
+	PV_GATE_START_VTOP,
+	PV_GATE_END_VTOP,
+
+	PV_GATE_START_MCKINLEY_E9,
+	PV_GATE_END_MCKINLEY_E9,
+};
+
+struct pv_patchdata {
+	unsigned long start_fsyscall_patchlist;
+	unsigned long end_fsyscall_patchlist;
+	unsigned long start_brl_fsys_bubble_down_patchlist;
+	unsigned long end_brl_fsys_bubble_down_patchlist;
+	unsigned long start_vtop_patchlist;
+	unsigned long end_vtop_patchlist;
+	unsigned long start_mckinley_e9_patchlist;
+	unsigned long end_mckinley_e9_patchlist;
+
+	void *gate_section;
+};
+
+extern struct pv_patchdata pv_patchdata;
+
+unsigned long paravirt_get_gate_patchlist(enum pv_gate_patchlist type);
+void *paravirt_get_gate_section(void);
+#endif
+
 #ifdef CONFIG_PARAVIRT_GUEST
 
 #define PARAVIRT_HYPERVISOR_TYPE_DEFAULT	0
@@ -68,6 +118,14 @@ struct pv_init_ops {
 	int (*arch_setup_nomca)(void);
 
 	void (*post_smp_prepare_boot_cpu)(void);
+
+#ifdef ASM_SUPPORTED
+	unsigned long (*patch_bundle)(void *sbundle, void *ebundle,
+				      unsigned long type);
+	unsigned long (*patch_inst)(unsigned long stag, unsigned long etag,
+				    unsigned long type);
+#endif
+	void (*patch_branch)(unsigned long tag, unsigned long type);
 };
 
 extern struct pv_init_ops pv_init_ops;
@@ -210,6 +268,8 @@ struct pv_time_ops {
 	int (*do_steal_accounting)(unsigned long *new_itm);
 
 	void (*clocksource_resume)(void);
+
+	unsigned long long (*sched_clock)(void);
 };
 
 extern struct pv_time_ops pv_time_ops;
@@ -227,6 +287,11 @@ paravirt_do_steal_accounting(unsigned long *new_itm)
 	return pv_time_ops.do_steal_accounting(new_itm);
 }
 
+static inline unsigned long long paravirt_sched_clock(void)
+{
+	return pv_time_ops.sched_clock();
+}
+
 #endif /* !__ASSEMBLY__ */
 
 #else
diff --git a/arch/ia64/include/asm/paravirt_patch.h b/arch/ia64/include/asm/paravirt_patch.h
new file mode 100644
index 0000000..128ff5d
--- /dev/null
+++ b/arch/ia64/include/asm/paravirt_patch.h
@@ -0,0 +1,143 @@
+/******************************************************************************
+ * Copyright (c) 2008 Isaku Yamahata <yamahata at valinux co jp>
+ *                    VA Linux Systems Japan K.K.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that 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
+ *
+ */
+
+#ifndef __ASM_PARAVIRT_PATCH_H
+#define __ASM_PARAVIRT_PATCH_H
+
+#ifdef __ASSEMBLY__
+
+	.section .paravirt_branches, "a"
+	.previous
+#define PARAVIRT_PATCH_SITE_BR(type)		\
+	{					\
+	[1:] ;					\
+	br.cond.sptk.many 2f ;			\
+	nop.b 0 ;				\
+	nop.b 0;; ;				\
+	} ;					\
+	2:					\
+	.xdata8 ".paravirt_branches", 1b, type
+
+#else
+
+#include <linux/stringify.h>
+#include <asm/intrinsics.h>
+
+/* for binary patch */
+struct paravirt_patch_site_bundle {
+	void		*sbundle;
+	void		*ebundle;
+	unsigned long	type;
+};
+
+/* label means the beginning of new bundle */
+#define paravirt_alt_bundle(instr, privop)				\
+	"\t998:\n"							\
+	"\t" instr "\n"							\
+	"\t999:\n"							\
+	"\t.pushsection .paravirt_bundles, \"a\"\n"			\
+	"\t.popsection\n"						\
+	"\t.xdata8 \".paravirt_bundles\", 998b, 999b, "			\
+	__stringify(privop) "\n"
+
+
+struct paravirt_patch_bundle_elem {
+	const void	*sbundle;
+	const void	*ebundle;
+	unsigned long	type;
+};
+
+
+struct paravirt_patch_site_inst {
+	unsigned long	stag;
+	unsigned long	etag;
+	unsigned long	type;
+};
+
+#define paravirt_alt_inst(instr, privop)				\
+	"\t[998:]\n"							\
+	"\t" instr "\n"							\
+	"\t[999:]\n"							\
+	"\t.pushsection .paravirt_insts, \"a\"\n"			\
+	"\t.popsection\n"						\
+	"\t.xdata8 \".paravirt_insts\", 998b, 999b, "			\
+	__stringify(privop) "\n"
+
+struct paravirt_patch_site_branch {
+	unsigned long	tag;
+	unsigned long	type;
+};
+
+struct paravirt_patch_branch_target {
+	const void	*entry;
+	unsigned long	type;
+};
+
+void
+__paravirt_patch_apply_branch(
+	unsigned long tag, unsigned long type,
+	const struct paravirt_patch_branch_target *entries,
+	unsigned int nr_entries);
+
+void
+paravirt_patch_reloc_br(unsigned long tag, const void *target);
+
+void
+paravirt_patch_reloc_brl(unsigned long tag, const void *target);
+
+
+#if defined(ASM_SUPPORTED) && defined(CONFIG_PARAVIRT)
+unsigned long
+ia64_native_patch_bundle(void *sbundle, void *ebundle, unsigned long type);
+
+unsigned long
+__paravirt_patch_apply_bundle(void *sbundle, void *ebundle, unsigned long type,
+			      const struct paravirt_patch_bundle_elem *elems,
+			      unsigned long nelems,
+			      const struct paravirt_patch_bundle_elem **found);
+
+void
+paravirt_patch_apply_bundle(const struct paravirt_patch_site_bundle *start,
+			    const struct paravirt_patch_site_bundle *end);
+
+void
+paravirt_patch_apply_inst(const struct paravirt_patch_site_inst *start,
+			  const struct paravirt_patch_site_inst *end);
+
+void paravirt_patch_apply(void);
+#else
+#define paravirt_patch_apply_bundle(start, end)	do { } while (0)
+#define paravirt_patch_apply_inst(start, end)	do { } while (0)
+#define paravirt_patch_apply()			do { } while (0)
+#endif
+
+#endif /* !__ASSEMBLEY__ */
+
+#endif /* __ASM_PARAVIRT_PATCH_H */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-set-style: "linux"
+ * c-basic-offset: 8
+ * tab-width: 8
+ * indent-tabs-mode: t
+ * End:
+ */
diff --git a/arch/ia64/include/asm/paravirt_privop.h b/arch/ia64/include/asm/paravirt_privop.h
index 33c8e55..3d29511 100644
--- a/arch/ia64/include/asm/paravirt_privop.h
+++ b/arch/ia64/include/asm/paravirt_privop.h
@@ -33,7 +33,7 @@
  */
 
 struct pv_cpu_ops {
-	void (*fc)(unsigned long addr);
+	void (*fc)(void *addr);
 	unsigned long (*thash)(unsigned long addr);
 	unsigned long (*get_cpuid)(int index);
 	unsigned long (*get_pmd)(int index);
@@ -60,12 +60,18 @@ extern unsigned long ia64_native_getreg_func(int regnum);
 /* Instructions paravirtualized for performance */
 /************************************************/
 
+#ifndef ASM_SUPPORTED
+#define paravirt_ssm_i()	pv_cpu_ops.ssm_i()
+#define paravirt_rsm_i()	pv_cpu_ops.rsm_i()
+#define __paravirt_getreg()	pv_cpu_ops.getreg()
+#endif
+
 /* mask for ia64_native_ssm/rsm() must be constant.("i" constraing).
  * static inline function doesn't satisfy it. */
 #define paravirt_ssm(mask)			\
 	do {					\
 		if ((mask) == IA64_PSR_I)	\
-			pv_cpu_ops.ssm_i();	\
+			paravirt_ssm_i();	\
 		else				\
 			ia64_native_ssm(mask);	\
 	} while (0)
@@ -73,7 +79,7 @@ extern unsigned long ia64_native_getreg_func(int regnum);
 #define paravirt_rsm(mask)			\
 	do {					\
 		if ((mask) == IA64_PSR_I)	\
-			pv_cpu_ops.rsm_i();	\
+			paravirt_rsm_i();	\
 		else				\
 			ia64_native_rsm(mask);	\
 	} while (0)
@@ -86,7 +92,7 @@ extern unsigned long ia64_native_getreg_func(int regnum);
 		if ((reg) == _IA64_REG_IP)			\
 			res = ia64_native_getreg(_IA64_REG_IP); \
 		else						\
-			res = pv_cpu_ops.getreg(reg);		\
+			res = __paravirt_getreg(reg);		\
 		res;						\
 	})
 
@@ -112,6 +118,12 @@ void paravirt_cpu_asm_init(const struct pv_cpu_asm_switch *cpu_asm_switch);
 
 #endif /* CONFIG_PARAVIRT */
 
+#if defined(CONFIG_PARAVIRT) && defined(ASM_SUPPORTED)
+#define paravirt_dv_serialize_data()	ia64_dv_serialize_data()
+#else
+#define paravirt_dv_serialize_data()	/* nothing */
+#endif
+
 /* these routines utilize privilege-sensitive or performance-sensitive
  * privileged instructions so the code must be replaced with
  * paravirtualized versions */
@@ -121,4 +133,349 @@ void paravirt_cpu_asm_init(const struct pv_cpu_asm_switch *cpu_asm_switch);
 	IA64_PARAVIRT_ASM_FUNC(work_processed_syscall)
 #define ia64_leave_kernel		IA64_PARAVIRT_ASM_FUNC(leave_kernel)
 
+
+#if defined(CONFIG_PARAVIRT)
+/******************************************************************************
+ * binary patching infrastructure
+ */
+#define PARAVIRT_PATCH_TYPE_FC				1
+#define PARAVIRT_PATCH_TYPE_THASH			2
+#define PARAVIRT_PATCH_TYPE_GET_CPUID			3
+#define PARAVIRT_PATCH_TYPE_GET_PMD			4
+#define PARAVIRT_PATCH_TYPE_PTCGA			5
+#define PARAVIRT_PATCH_TYPE_GET_RR			6
+#define PARAVIRT_PATCH_TYPE_SET_RR			7
+#define PARAVIRT_PATCH_TYPE_SET_RR0_TO_RR4		8
+#define PARAVIRT_PATCH_TYPE_SSM_I			9
+#define PARAVIRT_PATCH_TYPE_RSM_I			10
+#define PARAVIRT_PATCH_TYPE_GET_PSR_I			11
+#define PARAVIRT_PATCH_TYPE_INTRIN_LOCAL_IRQ_RESTORE	12
+
+/* PARAVIRT_PATY_TYPE_[GS]ETREG + _IA64_REG_xxx */
+#define PARAVIRT_PATCH_TYPE_GETREG			0x10000000
+#define PARAVIRT_PATCH_TYPE_SETREG			0x20000000
+
+/*
+ * struct task_struct* (*ia64_switch_to)(void* next_task);
+ * void *ia64_leave_syscall;
+ * void *ia64_work_processed_syscall
+ * void *ia64_leave_kernel;
+ */
+
+#define PARAVIRT_PATCH_TYPE_BR_START			0x30000000
+#define PARAVIRT_PATCH_TYPE_BR_SWITCH_TO		\
+	(PARAVIRT_PATCH_TYPE_BR_START + 0)
+#define PARAVIRT_PATCH_TYPE_BR_LEAVE_SYSCALL		\
+	(PARAVIRT_PATCH_TYPE_BR_START + 1)
+#define PARAVIRT_PATCH_TYPE_BR_WORK_PROCESSED_SYSCALL	\
+	(PARAVIRT_PATCH_TYPE_BR_START + 2)
+#define PARAVIRT_PATCH_TYPE_BR_LEAVE_KERNEL		\
+	(PARAVIRT_PATCH_TYPE_BR_START + 3)
+
+#ifdef ASM_SUPPORTED
+#include <asm/paravirt_patch.h>
+
+/*
+ * pv_cpu_ops calling stub.
+ * normal function call convension can't be written by gcc
+ * inline assembly.
+ *
+ * from the caller's point of view,
+ * the following registers will be clobbered.
+ * r2, r3
+ * r8-r15
+ * r16, r17
+ * b6, b7
+ * p6-p15
+ * ar.ccv
+ *
+ * from the callee's point of view ,
+ * the following registers can be used.
+ * r2, r3: scratch
+ * r8: scratch, input argument0 and return value
+ * r0-r15: scratch, input argument1-5
+ * b6: return pointer
+ * b7: scratch
+ * p6-p15: scratch
+ * ar.ccv: scratch
+ *
+ * other registers must not be changed. especially
+ * b0: rp: preserved. gcc ignores b0 in clobbered register.
+ * r16: saved gp
+ */
+/* 5 bundles */
+#define __PARAVIRT_BR							\
+	";;\n"								\
+	"{ .mlx\n"							\
+	"nop 0\n"							\
+	"movl r2 = %[op_addr]\n"/* get function pointer address */	\
+	";;\n"								\
+	"}\n"								\
+	"1:\n"								\
+	"{ .mii\n"							\
+	"ld8 r2 = [r2]\n"	/* load function descriptor address */	\
+	"mov r17 = ip\n"	/* get ip to calc return address */	\
+	"mov r16 = gp\n"	/* save gp */				\
+	";;\n"								\
+	"}\n"								\
+	"{ .mii\n"							\
+	"ld8 r3 = [r2], 8\n"	/* load entry address */		\
+	"adds r17 =  1f - 1b, r17\n"	/* calculate return address */	\
+	";;\n"								\
+	"mov b7 = r3\n"		/* set entry address */			\
+	"}\n"								\
+	"{ .mib\n"							\
+	"ld8 gp = [r2]\n"	/* load gp value */			\
+	"mov b6 = r17\n"	/* set return address */		\
+	"br.cond.sptk.few b7\n"	/* intrinsics are very short isns */	\
+	"}\n"								\
+	"1:\n"								\
+	"{ .mii\n"							\
+	"mov gp = r16\n"	/* restore gp value */			\
+	"nop 0\n"							\
+	"nop 0\n"							\
+	";;\n"								\
+	"}\n"
+
+#define PARAVIRT_OP(op)				\
+	[op_addr] "i"(&pv_cpu_ops.op)
+
+#define PARAVIRT_TYPE(type)			\
+	PARAVIRT_PATCH_TYPE_ ## type
+
+#define PARAVIRT_REG_CLOBBERS0					\
+	"r2", "r3", /*"r8",*/ "r9", "r10", "r11", "r14",	\
+		"r15", "r16", "r17"
+
+#define PARAVIRT_REG_CLOBBERS1					\
+	"r2","r3", /*"r8",*/ "r9", "r10", "r11", "r14",		\
+		"r15", "r16", "r17"
+
+#define PARAVIRT_REG_CLOBBERS2					\
+	"r2", "r3", /*"r8", "r9",*/ "r10", "r11", "r14",	\
+		"r15", "r16", "r17"
+
+#define PARAVIRT_REG_CLOBBERS5					\
+	"r2", "r3", /*"r8", "r9", "r10", "r11", "r14",*/	\
+		"r15", "r16", "r17"
+
+#define PARAVIRT_BR_CLOBBERS			\
+	"b6", "b7"
+
+#define PARAVIRT_PR_CLOBBERS						\
+	"p6", "p7", "p8", "p9", "p10", "p11", "p12", "p13", "p14", "p15"
+
+#define PARAVIRT_AR_CLOBBERS			\
+	"ar.ccv"
+
+#define PARAVIRT_CLOBBERS0			\
+		PARAVIRT_REG_CLOBBERS0,		\
+		PARAVIRT_BR_CLOBBERS,		\
+		PARAVIRT_PR_CLOBBERS,		\
+		PARAVIRT_AR_CLOBBERS,		\
+		"memory"
+
+#define PARAVIRT_CLOBBERS1			\
+		PARAVIRT_REG_CLOBBERS1,		\
+		PARAVIRT_BR_CLOBBERS,		\
+		PARAVIRT_PR_CLOBBERS,		\
+		PARAVIRT_AR_CLOBBERS,		\
+		"memory"
+
+#define PARAVIRT_CLOBBERS2			\
+		PARAVIRT_REG_CLOBBERS2,		\
+		PARAVIRT_BR_CLOBBERS,		\
+		PARAVIRT_PR_CLOBBERS,		\
+		PARAVIRT_AR_CLOBBERS,		\
+		"memory"
+
+#define PARAVIRT_CLOBBERS5			\
+		PARAVIRT_REG_CLOBBERS5,		\
+		PARAVIRT_BR_CLOBBERS,		\
+		PARAVIRT_PR_CLOBBERS,		\
+		PARAVIRT_AR_CLOBBERS,		\
+		"memory"
+
+#define PARAVIRT_BR0(op, type)					\
+	register unsigned long ia64_clobber asm ("r8");		\
+	asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,	\
+					  PARAVIRT_TYPE(type))	\
+		      :	"=r"(ia64_clobber)			\
+		      : PARAVIRT_OP(op)				\
+		      : PARAVIRT_CLOBBERS0)
+
+#define PARAVIRT_BR0_RET(op, type)				\
+	register unsigned long ia64_intri_res asm ("r8");	\
+	asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,	\
+					  PARAVIRT_TYPE(type))	\
+		      : "=r"(ia64_intri_res)			\
+		      : PARAVIRT_OP(op)				\
+		      : PARAVIRT_CLOBBERS0)
+
+#define PARAVIRT_BR1(op, type, arg1)				\
+	register unsigned long __##arg1 asm ("r8") = arg1;	\
+	register unsigned long ia64_clobber asm ("r8");		\
+	asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,	\
+					  PARAVIRT_TYPE(type))	\
+		      :	"=r"(ia64_clobber)			\
+		      : PARAVIRT_OP(op), "0"(__##arg1)		\
+		      : PARAVIRT_CLOBBERS1)
+
+#define PARAVIRT_BR1_RET(op, type, arg1)			\
+	register unsigned long ia64_intri_res asm ("r8");	\
+	register unsigned long __##arg1 asm ("r8") = arg1;	\
+	asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,	\
+					  PARAVIRT_TYPE(type))	\
+		      : "=r"(ia64_intri_res)			\
+		      : PARAVIRT_OP(op), "0"(__##arg1)		\
+		      : PARAVIRT_CLOBBERS1)
+
+#define PARAVIRT_BR1_VOID(op, type, arg1)			\
+	register void *__##arg1 asm ("r8") = arg1;		\
+	register unsigned long ia64_clobber asm ("r8");		\
+	asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,	\
+					  PARAVIRT_TYPE(type))	\
+		      :	"=r"(ia64_clobber)			\
+		      : PARAVIRT_OP(op), "0"(__##arg1)		\
+		      : PARAVIRT_CLOBBERS1)
+
+#define PARAVIRT_BR2(op, type, arg1, arg2)				\
+	register unsigned long __##arg1 asm ("r8") = arg1;		\
+	register unsigned long __##arg2 asm ("r9") = arg2;		\
+	register unsigned long ia64_clobber1 asm ("r8");		\
+	register unsigned long ia64_clobber2 asm ("r9");		\
+	asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,		\
+					  PARAVIRT_TYPE(type))		\
+		      : "=r"(ia64_clobber1), "=r"(ia64_clobber2)	\
+		      : PARAVIRT_OP(op), "0"(__##arg1), "1"(__##arg2)	\
+		      : PARAVIRT_CLOBBERS2)
+
+
+#define PARAVIRT_DEFINE_CPU_OP0(op, type)		\
+	static inline void				\
+	paravirt_ ## op (void)				\
+	{						\
+		PARAVIRT_BR0(op, type);			\
+	}
+
+#define PARAVIRT_DEFINE_CPU_OP0_RET(op, type)		\
+	static inline unsigned long			\
+	paravirt_ ## op (void)				\
+	{						\
+		PARAVIRT_BR0_RET(op, type);		\
+		return ia64_intri_res;			\
+	}
+
+#define PARAVIRT_DEFINE_CPU_OP1_VOID(op, type)		\
+	static inline void				\
+	paravirt_ ## op (void *arg1)			\
+	{						\
+		PARAVIRT_BR1_VOID(op, type, arg1);	\
+	}
+
+#define PARAVIRT_DEFINE_CPU_OP1(op, type)		\
+	static inline void				\
+	paravirt_ ## op (unsigned long arg1)		\
+	{						\
+		PARAVIRT_BR1(op, type, arg1);		\
+	}
+
+#define PARAVIRT_DEFINE_CPU_OP1_RET(op, type)		\
+	static inline unsigned long			\
+	paravirt_ ## op (unsigned long arg1)		\
+	{						\
+		PARAVIRT_BR1_RET(op, type, arg1);	\
+		return ia64_intri_res;			\
+	}
+
+#define PARAVIRT_DEFINE_CPU_OP2(op, type)		\
+	static inline void				\
+	paravirt_ ## op (unsigned long arg1,		\
+			 unsigned long arg2)		\
+	{						\
+		PARAVIRT_BR2(op, type, arg1, arg2);	\
+	}
+
+
+PARAVIRT_DEFINE_CPU_OP1_VOID(fc, FC);
+PARAVIRT_DEFINE_CPU_OP1_RET(thash, THASH)
+PARAVIRT_DEFINE_CPU_OP1_RET(get_cpuid, GET_CPUID)
+PARAVIRT_DEFINE_CPU_OP1_RET(get_pmd, GET_PMD)
+PARAVIRT_DEFINE_CPU_OP2(ptcga, PTCGA)
+PARAVIRT_DEFINE_CPU_OP1_RET(get_rr, GET_RR)
+PARAVIRT_DEFINE_CPU_OP2(set_rr, SET_RR)
+PARAVIRT_DEFINE_CPU_OP0(ssm_i, SSM_I)
+PARAVIRT_DEFINE_CPU_OP0(rsm_i, RSM_I)
+PARAVIRT_DEFINE_CPU_OP0_RET(get_psr_i, GET_PSR_I)
+PARAVIRT_DEFINE_CPU_OP1(intrin_local_irq_restore, INTRIN_LOCAL_IRQ_RESTORE)
+
+static inline void
+paravirt_set_rr0_to_rr4(unsigned long val0, unsigned long val1,
+			unsigned long val2, unsigned long val3,
+			unsigned long val4)
+{
+	register unsigned long __val0 asm ("r8") = val0;
+	register unsigned long __val1 asm ("r9") = val1;
+	register unsigned long __val2 asm ("r10") = val2;
+	register unsigned long __val3 asm ("r11") = val3;
+	register unsigned long __val4 asm ("r14") = val4;
+
+	register unsigned long ia64_clobber0 asm ("r8");
+	register unsigned long ia64_clobber1 asm ("r9");
+	register unsigned long ia64_clobber2 asm ("r10");
+	register unsigned long ia64_clobber3 asm ("r11");
+	register unsigned long ia64_clobber4 asm ("r14");
+
+	asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,
+					  PARAVIRT_TYPE(SET_RR0_TO_RR4))
+		      : "=r"(ia64_clobber0),
+			"=r"(ia64_clobber1),
+			"=r"(ia64_clobber2),
+			"=r"(ia64_clobber3),
+			"=r"(ia64_clobber4)
+		      : PARAVIRT_OP(set_rr0_to_rr4),
+			"0"(__val0), "1"(__val1), "2"(__val2),
+			"3"(__val3), "4"(__val4)
+		      : PARAVIRT_CLOBBERS5);
+}
+
+/* unsigned long paravirt_getreg(int reg) */
+#define __paravirt_getreg(reg)						\
+	({								\
+		register unsigned long ia64_intri_res asm ("r8");	\
+		register unsigned long __reg asm ("r8") = (reg);	\
+									\
+		BUILD_BUG_ON(!__builtin_constant_p(reg));		\
+		asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,	\
+						  PARAVIRT_TYPE(GETREG) \
+						  + (reg))		\
+			      : "=r"(ia64_intri_res)			\
+			      : PARAVIRT_OP(getreg), "0"(__reg)		\
+			      : PARAVIRT_CLOBBERS1);			\
+									\
+		ia64_intri_res;						\
+	})
+
+/* void paravirt_setreg(int reg, unsigned long val) */
+#define paravirt_setreg(reg, val)					\
+	do {								\
+		register unsigned long __val asm ("r8") = val;		\
+		register unsigned long __reg asm ("r9") = reg;		\
+		register unsigned long ia64_clobber1 asm ("r8");	\
+		register unsigned long ia64_clobber2 asm ("r9");	\
+									\
+		BUILD_BUG_ON(!__builtin_constant_p(reg));		\
+		asm volatile (paravirt_alt_bundle(__PARAVIRT_BR,	\
+						  PARAVIRT_TYPE(SETREG) \
+						  + (reg))		\
+			      : "=r"(ia64_clobber1),			\
+				"=r"(ia64_clobber2)			\
+			      : PARAVIRT_OP(setreg),			\
+				"1"(__reg), "0"(__val)			\
+			      : PARAVIRT_CLOBBERS2);			\
+	} while (0)
+
+#endif /* ASM_SUPPORTED */
+#endif /* CONFIG_PARAVIRT && ASM_SUPPOTED */
+
 #endif /* _ASM_IA64_PARAVIRT_PRIVOP_H */
diff --git a/arch/ia64/include/asm/smp.h b/arch/ia64/include/asm/smp.h
index 21c4023..5984083 100644
--- a/arch/ia64/include/asm/smp.h
+++ b/arch/ia64/include/asm/smp.h
@@ -126,7 +126,8 @@ extern void identify_siblings (struct cpuinfo_ia64 *);
 extern int is_multithreading_enabled(void);
 
 extern void arch_send_call_function_single_ipi(int cpu);
-extern void arch_send_call_function_ipi(cpumask_t mask);
+extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
+#define arch_send_call_function_ipi_mask arch_send_call_function_ipi_mask
 
 #else /* CONFIG_SMP */
 
diff --git a/arch/ia64/include/asm/timex.h b/arch/ia64/include/asm/timex.h
index 4e03cfe..86c7db8 100644
--- a/arch/ia64/include/asm/timex.h
+++ b/arch/ia64/include/asm/timex.h
@@ -40,5 +40,6 @@ get_cycles (void)
 }
 
 extern void ia64_cpu_local_tick (void);
+extern unsigned long long ia64_native_sched_clock (void);
 
 #endif /* _ASM_IA64_TIMEX_H */
diff --git a/arch/ia64/include/asm/topology.h b/arch/ia64/include/asm/topology.h
index 3193f44..7b4c8c7 100644
--- a/arch/ia64/include/asm/topology.h
+++ b/arch/ia64/include/asm/topology.h
@@ -44,11 +44,6 @@
 #define parent_node(nid) (nid)
 
 /*
- * Returns the number of the first CPU on Node 'node'.
- */
-#define node_to_first_cpu(node) (cpumask_first(cpumask_of_node(node)))
-
-/*
  * Determines the node for a given pci bus
  */
 #define pcibus_to_node(bus) PCI_CONTROLLER(bus)->node
@@ -117,11 +112,6 @@ void build_cpu_to_node_map(void);
 
 extern void arch_fix_phys_package_id(int num, u32 slot);
 
-#define pcibus_to_cpumask(bus)	(pcibus_to_node(bus) == -1 ? \
-					CPU_MASK_ALL : \
-					node_to_cpumask(pcibus_to_node(bus)) \
-				)
-
 #define cpumask_of_pcibus(bus)	(pcibus_to_node(bus) == -1 ?		\
 				 cpu_all_mask :				\
 				 cpumask_of_node(pcibus_to_node(bus)))
diff --git a/arch/ia64/include/asm/xen/hypervisor.h b/arch/ia64/include/asm/xen/hypervisor.h
index 7a804e8..e425227 100644
--- a/arch/ia64/include/asm/xen/hypervisor.h
+++ b/arch/ia64/include/asm/xen/hypervisor.h
@@ -33,9 +33,6 @@
 #ifndef _ASM_IA64_XEN_HYPERVISOR_H
 #define _ASM_IA64_XEN_HYPERVISOR_H
 
-#ifdef CONFIG_XEN
-
-#include <linux/init.h>
 #include <xen/interface/xen.h>
 #include <xen/interface/version.h>	/* to compile feature.c */
 #include <xen/features.h>		/* to comiple xen-netfront.c */
@@ -43,22 +40,32 @@
 
 /* xen_domain_type is set before executing any C code by early_xen_setup */
 enum xen_domain_type {
-	XEN_NATIVE,
-	XEN_PV_DOMAIN,
-	XEN_HVM_DOMAIN,
+	XEN_NATIVE,	/* running on bare hardware */
+	XEN_PV_DOMAIN,	/* running in a PV domain */
+	XEN_HVM_DOMAIN,	/* running in a Xen hvm domain*/
 };
 
+#ifdef CONFIG_XEN
 extern enum xen_domain_type xen_domain_type;
+#else
+#define xen_domain_type		XEN_NATIVE
+#endif
 
 #define xen_domain()		(xen_domain_type != XEN_NATIVE)
-#define xen_pv_domain()		(xen_domain_type == XEN_PV_DOMAIN)
-#define xen_initial_domain()	(xen_pv_domain() && \
+#define xen_pv_domain()		(xen_domain() &&			\
+				 xen_domain_type == XEN_PV_DOMAIN)
+#define xen_hvm_domain()	(xen_domain() &&			\
+				 xen_domain_type == XEN_HVM_DOMAIN)
+
+#ifdef CONFIG_XEN_DOM0
+#define xen_initial_domain()	(xen_pv_domain() &&			\
 				 (xen_start_info->flags & SIF_INITDOMAIN))
-#define xen_hvm_domain()	(xen_domain_type == XEN_HVM_DOMAIN)
+#else
+#define xen_initial_domain()	(0)
+#endif
 
-/* deprecated. remove this */
-#define is_running_on_xen()	(xen_domain_type == XEN_PV_DOMAIN)
 
+#ifdef CONFIG_XEN
 extern struct shared_info *HYPERVISOR_shared_info;
 extern struct start_info *xen_start_info;
 
@@ -74,16 +81,6 @@ void force_evtchn_callback(void);
 
 /* For setup_arch() in arch/ia64/kernel/setup.c */
 void xen_ia64_enable_opt_feature(void);
-
-#else /* CONFIG_XEN */
-
-#define xen_domain()		(0)
-#define xen_pv_domain()		(0)
-#define xen_initial_domain()	(0)
-#define xen_hvm_domain()	(0)
-#define is_running_on_xen()	(0)	/* deprecated. remove this */
 #endif
 
-#define is_initial_xendomain()	(0)	/* deprecated. remove this */
-
 #endif /* _ASM_IA64_XEN_HYPERVISOR_H */
diff --git a/arch/ia64/include/asm/xen/inst.h b/arch/ia64/include/asm/xen/inst.h
index 19c2ae1..c53a476 100644
--- a/arch/ia64/include/asm/xen/inst.h
+++ b/arch/ia64/include/asm/xen/inst.h
@@ -33,6 +33,9 @@
 #define __paravirt_work_processed_syscall_target \
 						xen_work_processed_syscall
 
+#define paravirt_fsyscall_table			xen_fsyscall_table
+#define paravirt_fsys_bubble_down		xen_fsys_bubble_down
+
 #define MOV_FROM_IFA(reg)	\
 	movl reg = XSI_IFA;	\
 	;;			\
@@ -110,6 +113,27 @@
 .endm
 #define MOV_FROM_PSR(pred, reg, clob)	__MOV_FROM_PSR pred, reg, clob
 
+/* assuming ar.itc is read with interrupt disabled. */
+#define MOV_FROM_ITC(pred, pred_clob, reg, clob)		\
+(pred)	movl clob = XSI_ITC_OFFSET;				\
+	;;							\
+(pred)	ld8 clob = [clob];					\
+(pred)	mov reg = ar.itc;					\
+	;;							\
+(pred)	add reg = reg, clob;					\
+	;;							\
+(pred)	movl clob = XSI_ITC_LAST;				\
+	;;							\
+(pred)	ld8 clob = [clob];					\
+	;;							\
+(pred)	cmp.geu.unc pred_clob, p0 = clob, reg;			\
+	;;							\
+(pred_clob)	add reg = 1, clob;				\
+	;;							\
+(pred)	movl clob = XSI_ITC_LAST;				\
+	;;							\
+(pred)	st8 [clob] = reg
+
 
 #define MOV_TO_IFA(reg, clob)	\
 	movl clob = XSI_IFA;	\
@@ -362,6 +386,10 @@
 #define RSM_PSR_DT		\
 	XEN_HYPER_RSM_PSR_DT
 
+#define RSM_PSR_BE_I(clob0, clob1)	\
+	RSM_PSR_I(p0, clob0, clob1);	\
+	rum psr.be
+
 #define SSM_PSR_DT_AND_SRLZ_I	\
 	XEN_HYPER_SSM_PSR_DT
 
diff --git a/arch/ia64/include/asm/xen/interface.h b/arch/ia64/include/asm/xen/interface.h
index f00fab40..e951e74 100644
--- a/arch/ia64/include/asm/xen/interface.h
+++ b/arch/ia64/include/asm/xen/interface.h
@@ -209,6 +209,15 @@ struct mapped_regs {
 			unsigned long krs[8];	/* kernel registers */
 			unsigned long tmp[16];	/* temp registers
 						   (e.g. for hyperprivops) */
+
+			/* itc paravirtualization
+			 * vAR.ITC = mAR.ITC + itc_offset
+			 * itc_last is one which was lastly passed to
+			 * the guest OS in order to prevent it from
+			 * going backwords.
+			 */
+			unsigned long itc_offset;
+			unsigned long itc_last;
 		};
 	};
 };
diff --git a/arch/ia64/include/asm/xen/minstate.h b/arch/ia64/include/asm/xen/minstate.h
index 4d92d9b..c57fa91 100644
--- a/arch/ia64/include/asm/xen/minstate.h
+++ b/arch/ia64/include/asm/xen/minstate.h
@@ -1,3 +1,12 @@
+
+#ifdef CONFIG_VIRT_CPU_ACCOUNTING
+/* read ar.itc in advance, and use it before leaving bank 0 */
+#define XEN_ACCOUNT_GET_STAMP		\
+	MOV_FROM_ITC(pUStk, p6, r20, r2);
+#else
+#define XEN_ACCOUNT_GET_STAMP
+#endif
+
 /*
  * DO_SAVE_MIN switches to the kernel stacks (if necessary) and saves
  * the minimum state necessary that allows us to turn psr.ic back
@@ -123,7 +132,7 @@
 	;;											\
 .mem.offset 0,0; st8.spill [r16]=r2,16;								\
 .mem.offset 8,0; st8.spill [r17]=r3,16;								\
-	ACCOUNT_GET_STAMP									\
+	XEN_ACCOUNT_GET_STAMP									\
 	adds r2=IA64_PT_REGS_R16_OFFSET,r1;							\
 	;;											\
 	EXTRA;											\
diff --git a/arch/ia64/include/asm/xen/patchlist.h b/arch/ia64/include/asm/xen/patchlist.h
new file mode 100644
index 0000000..eae944e
--- /dev/null
+++ b/arch/ia64/include/asm/xen/patchlist.h
@@ -0,0 +1,38 @@
+/******************************************************************************
+ * arch/ia64/include/asm/xen/patchlist.h
+ *
+ * Copyright (c) 2008 Isaku Yamahata <yamahata at valinux co jp>
+ *                    VA Linux Systems Japan K.K.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that 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
+ *
+ */
+
+#define __paravirt_start_gate_fsyscall_patchlist		\
+	__xen_start_gate_fsyscall_patchlist
+#define __paravirt_end_gate_fsyscall_patchlist			\
+	__xen_end_gate_fsyscall_patchlist
+#define __paravirt_start_gate_brl_fsys_bubble_down_patchlist	\
+	__xen_start_gate_brl_fsys_bubble_down_patchlist
+#define __paravirt_end_gate_brl_fsys_bubble_down_patchlist	\
+	__xen_end_gate_brl_fsys_bubble_down_patchlist
+#define __paravirt_start_gate_vtop_patchlist			\
+	__xen_start_gate_vtop_patchlist
+#define __paravirt_end_gate_vtop_patchlist			\
+	__xen_end_gate_vtop_patchlist
+#define __paravirt_start_gate_mckinley_e9_patchlist		\
+	__xen_start_gate_mckinley_e9_patchlist
+#define __paravirt_end_gate_mckinley_e9_patchlist		\
+	__xen_end_gate_mckinley_e9_patchlist
diff --git a/arch/ia64/include/asm/xen/privop.h b/arch/ia64/include/asm/xen/privop.h
index 71ec754..fb4ec5e 100644
--- a/arch/ia64/include/asm/xen/privop.h
+++ b/arch/ia64/include/asm/xen/privop.h
@@ -55,6 +55,8 @@
 #define XSI_BANK1_R16			(XSI_BASE + XSI_BANK1_R16_OFS)
 #define XSI_BANKNUM			(XSI_BASE + XSI_BANKNUM_OFS)
 #define XSI_IHA				(XSI_BASE + XSI_IHA_OFS)
+#define XSI_ITC_OFFSET			(XSI_BASE + XSI_ITC_OFFSET_OFS)
+#define XSI_ITC_LAST			(XSI_BASE + XSI_ITC_LAST_OFS)
 #endif
 
 #ifndef __ASSEMBLY__
@@ -67,7 +69,7 @@
  *  may have different semantics depending on whether they are executed
  *  at PL0 vs PL!=0.  When paravirtualized, these instructions mustn't
  *  be allowed to execute directly, lest incorrect semantics result. */
-extern void xen_fc(unsigned long addr);
+extern void xen_fc(void *addr);
 extern unsigned long xen_thash(unsigned long addr);
 
 /* Note that "ttag" and "cover" are also privilege-sensitive; "ttag"
@@ -80,8 +82,10 @@ extern unsigned long xen_thash(unsigned long addr);
 extern unsigned long xen_get_cpuid(int index);
 extern unsigned long xen_get_pmd(int index);
 
+#ifndef ASM_SUPPORTED
 extern unsigned long xen_get_eflag(void);	/* see xen_ia64_getreg */
 extern void xen_set_eflag(unsigned long);	/* see xen_ia64_setreg */
+#endif
 
 /************************************************/
 /* Instructions paravirtualized for performance */
@@ -106,6 +110,7 @@ extern void xen_set_eflag(unsigned long);	/* see xen_ia64_setreg */
 #define xen_get_virtual_pend()		\
 	(*(((uint8_t *)XEN_MAPPEDREGS->interrupt_mask_addr) - 1))
 
+#ifndef ASM_SUPPORTED
 /* Although all privileged operations can be left to trap and will
  * be properly handled by Xen, some are frequent enough that we use
  * hyperprivops for performance. */
@@ -123,6 +128,7 @@ extern void xen_set_rr0_to_rr4(unsigned long val0, unsigned long val1,
 			       unsigned long val4);
 extern void xen_set_kr(unsigned long index, unsigned long val);
 extern void xen_ptcga(unsigned long addr, unsigned long size);
+#endif /* !ASM_SUPPORTED */
 
 #endif /* !__ASSEMBLY__ */
 
diff --git a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile
index f2778f2..5628e9a 100644
--- a/arch/ia64/kernel/Makefile
+++ b/arch/ia64/kernel/Makefile
@@ -5,7 +5,7 @@
 extra-y	:= head.o init_task.o vmlinux.lds
 
 obj-y := acpi.o entry.o efi.o efi_stub.o gate-data.o fsys.o ia64_ksyms.o irq.o irq_ia64.o	\
-	 irq_lsapic.o ivt.o machvec.o pal.o patch.o process.o perfmon.o ptrace.o sal.o		\
+	 irq_lsapic.o ivt.o machvec.o pal.o paravirt_patchlist.o patch.o process.o perfmon.o ptrace.o sal.o		\
 	 salinfo.o setup.o signal.o sys_ia64.o time.o traps.o unaligned.o \
 	 unwind.o mca.o mca_asm.o topology.o dma-mapping.o
 
@@ -36,7 +36,8 @@ obj-$(CONFIG_PCI_MSI)		+= msi_ia64.o
 mca_recovery-y			+= mca_drv.o mca_drv_asm.o
 obj-$(CONFIG_IA64_MC_ERR_INJECT)+= err_inject.o
 
-obj-$(CONFIG_PARAVIRT)		+= paravirt.o paravirtentry.o
+obj-$(CONFIG_PARAVIRT)		+= paravirt.o paravirtentry.o \
+				   paravirt_patch.o
 
 obj-$(CONFIG_IA64_ESI)		+= esi.o
 ifneq ($(CONFIG_IA64_ESI),)
@@ -45,35 +46,13 @@ endif
 obj-$(CONFIG_DMAR)		+= pci-dma.o
 obj-$(CONFIG_SWIOTLB)		+= pci-swiotlb.o
 
-# The gate DSO image is built using a special linker script.
-targets += gate.so gate-syms.o
-
-extra-y += gate.so gate-syms.o gate.lds gate.o
-
 # fp_emulate() expects f2-f5,f16-f31 to contain the user-level state.
 CFLAGS_traps.o  += -mfixed-range=f2-f5,f16-f31
 
-CPPFLAGS_gate.lds := -P -C -U$(ARCH)
-
-quiet_cmd_gate = GATE $@
-      cmd_gate = $(CC) -nostdlib $(GATECFLAGS_$(@F)) -Wl,-T,$(filter-out FORCE,$^) -o $@
-
-GATECFLAGS_gate.so = -shared -s -Wl,-soname=linux-gate.so.1 \
-		     $(call ld-option, -Wl$(comma)--hash-style=sysv)
-$(obj)/gate.so: $(obj)/gate.lds $(obj)/gate.o FORCE
-	$(call if_changed,gate)
-
-$(obj)/built-in.o: $(obj)/gate-syms.o
-$(obj)/built-in.o: ld_flags += -R $(obj)/gate-syms.o
-
-GATECFLAGS_gate-syms.o = -r
-$(obj)/gate-syms.o: $(obj)/gate.lds $(obj)/gate.o FORCE
-	$(call if_changed,gate)
-
-# gate-data.o contains the gate DSO image as data in section .data.gate.
-# We must build gate.so before we can assemble it.
-# Note: kbuild does not track this dependency due to usage of .incbin
-$(obj)/gate-data.o: $(obj)/gate.so
+# The gate DSO image is built using a special linker script.
+include $(srctree)/arch/ia64/kernel/Makefile.gate
+# tell compiled for native
+CPPFLAGS_gate.lds += -D__IA64_GATE_PARAVIRTUALIZED_NATIVE
 
 # Calculate NR_IRQ = max(IA64_NATIVE_NR_IRQS, XEN_NR_IRQS, ...) based on config
 define sed-y
@@ -109,9 +88,9 @@ include/asm-ia64/nr-irqs.h: arch/$(SRCARCH)/kernel/nr-irqs.s
 clean-files += $(objtree)/include/asm-ia64/nr-irqs.h
 
 #
-# native ivt.S and entry.S
+# native ivt.S, entry.S and fsys.S
 #
-ASM_PARAVIRT_OBJS = ivt.o entry.o
+ASM_PARAVIRT_OBJS = ivt.o entry.o fsys.o
 define paravirtualized_native
 AFLAGS_$(1) += -D__IA64_ASM_PARAVIRTUALIZED_NATIVE
 AFLAGS_pvchk-sed-$(1) += -D__IA64_ASM_PARAVIRTUALIZED_PVCHECK
diff --git a/arch/ia64/kernel/Makefile.gate b/arch/ia64/kernel/Makefile.gate
new file mode 100644
index 0000000..1d87f84
--- /dev/null
+++ b/arch/ia64/kernel/Makefile.gate
@@ -0,0 +1,27 @@
+# The gate DSO image is built using a special linker script.
+
+targets += gate.so gate-syms.o
+
+extra-y += gate.so gate-syms.o gate.lds gate.o
+
+CPPFLAGS_gate.lds := -P -C -U$(ARCH)
+
+quiet_cmd_gate = GATE $@
+      cmd_gate = $(CC) -nostdlib $(GATECFLAGS_$(@F)) -Wl,-T,$(filter-out FORCE,$^) -o $@
+
+GATECFLAGS_gate.so = -shared -s -Wl,-soname=linux-gate.so.1 \
+		     $(call ld-option, -Wl$(comma)--hash-style=sysv)
+$(obj)/gate.so: $(obj)/gate.lds $(obj)/gate.o FORCE
+	$(call if_changed,gate)
+
+$(obj)/built-in.o: $(obj)/gate-syms.o
+$(obj)/built-in.o: ld_flags += -R $(obj)/gate-syms.o
+
+GATECFLAGS_gate-syms.o = -r
+$(obj)/gate-syms.o: $(obj)/gate.lds $(obj)/gate.o FORCE
+	$(call if_changed,gate)
+
+# gate-data.o contains the gate DSO image as data in section .data.gate.
+# We must build gate.so before we can assemble it.
+# Note: kbuild does not track this dependency due to usage of .incbin
+$(obj)/gate-data.o: $(obj)/gate.so
diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c
index bdef2ce..5510317 100644
--- a/arch/ia64/kernel/acpi.c
+++ b/arch/ia64/kernel/acpi.c
@@ -890,7 +890,7 @@ __init void prefill_possible_map(void)
 		possible, max((possible - available_cpus), 0));
 
 	for (i = 0; i < possible; i++)
-		cpu_set(i, cpu_possible_map);
+		set_cpu_possible(i, true);
 }
 
 int acpi_map_lsapic(acpi_handle handle, int *pcpu)
@@ -928,9 +928,9 @@ int acpi_map_lsapic(acpi_handle handle, int *pcpu)
 	buffer.length = ACPI_ALLOCATE_BUFFER;
 	buffer.pointer = NULL;
 
-	cpus_complement(tmp_map, cpu_present_map);
-	cpu = first_cpu(tmp_map);
-	if (cpu >= NR_CPUS)
+	cpumask_complement(&tmp_map, cpu_present_mask);
+	cpu = cpumask_first(&tmp_map);
+	if (cpu >= nr_cpu_ids)
 		return -EINVAL;
 
 	acpi_map_cpu2node(handle, cpu, physid);
diff --git a/arch/ia64/kernel/asm-offsets.c b/arch/ia64/kernel/asm-offsets.c
index 742dbb1..af56501 100644
--- a/arch/ia64/kernel/asm-offsets.c
+++ b/arch/ia64/kernel/asm-offsets.c
@@ -316,5 +316,7 @@ void foo(void)
 	DEFINE_MAPPED_REG_OFS(XSI_BANK1_R16_OFS, bank1_regs[0]);
 	DEFINE_MAPPED_REG_OFS(XSI_B0NATS_OFS, vbnat);
 	DEFINE_MAPPED_REG_OFS(XSI_B1NATS_OFS, vnat);
+	DEFINE_MAPPED_REG_OFS(XSI_ITC_OFFSET_OFS, itc_offset);
+	DEFINE_MAPPED_REG_OFS(XSI_ITC_LAST_OFS, itc_last);
 #endif /* CONFIG_XEN */
 }
diff --git a/arch/ia64/kernel/efi.c b/arch/ia64/kernel/efi.c
index efaff15..7ef80e8 100644
--- a/arch/ia64/kernel/efi.c
+++ b/arch/ia64/kernel/efi.c
@@ -456,6 +456,7 @@ efi_map_pal_code (void)
 		 GRANULEROUNDDOWN((unsigned long) pal_vaddr),
 		 pte_val(pfn_pte(__pa(pal_vaddr) >> PAGE_SHIFT, PAGE_KERNEL)),
 		 IA64_GRANULE_SHIFT);
+	paravirt_dv_serialize_data();
 	ia64_set_psr(psr);		/* restore psr */
 }
 
diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S
index e5341e2..ccfdeee 100644
--- a/arch/ia64/kernel/entry.S
+++ b/arch/ia64/kernel/entry.S
@@ -735,7 +735,7 @@ GLOBAL_ENTRY(__paravirt_leave_syscall)
 __paravirt_work_processed_syscall:
 #ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	adds r2=PT(LOADRS)+16,r12
-(pUStk)	mov.m r22=ar.itc			// fetch time at leave
+	MOV_FROM_ITC(pUStk, p9, r22, r19)	// fetch time at leave
 	adds r18=TI_FLAGS+IA64_TASK_SIZE,r13
 	;;
 (p6)	ld4 r31=[r18]				// load current_thread_info()->flags
@@ -984,7 +984,7 @@ GLOBAL_ENTRY(__paravirt_leave_kernel)
 #ifdef CONFIG_VIRT_CPU_ACCOUNTING
 	.pred.rel.mutex pUStk,pKStk
 	MOV_FROM_PSR(pKStk, r22, r29)	// M2 read PSR now that interrupts are disabled
-(pUStk)	mov.m r22=ar.itc	// M  fetch time at leave
+	MOV_FROM_ITC(pUStk, p9, r22, r29)	// M  fetch time at leave
 	nop.i 0
 	;;
 #else
diff --git a/arch/ia64/kernel/fsys.S b/arch/ia64/kernel/fsys.S
index c1625c7..3567d54 100644
--- a/arch/ia64/kernel/fsys.S
+++ b/arch/ia64/kernel/fsys.S
@@ -25,6 +25,7 @@
 #include <asm/unistd.h>
 
 #include "entry.h"
+#include "paravirt_inst.h"
 
 /*
  * See Documentation/ia64/fsys.txt for details on fsyscalls.
@@ -279,7 +280,7 @@ ENTRY(fsys_gettimeofday)
 (p9)	cmp.eq p13,p0 = 0,r30	// if mmio_ptr, clear p13 jitter control
 	;;
 	.pred.rel.mutex p8,p9
-(p8)	mov r2 = ar.itc		// CPU_TIMER. 36 clocks latency!!!
+	MOV_FROM_ITC(p8, p6, r2, r10)	// CPU_TIMER. 36 clocks latency!!!
 (p9)	ld8 r2 = [r30]		// MMIO_TIMER. Could also have latency issues..
 (p13)	ld8 r25 = [r19]		// get itc_lastcycle value
 	ld8 r9 = [r22],IA64_TIMESPEC_TV_NSEC_OFFSET	// tv_sec
@@ -418,7 +419,7 @@ EX(.fail_efault, ld8 r14=[r33])			// r14 <- *set
 	mov r17=(1 << (SIGKILL - 1)) | (1 << (SIGSTOP - 1))
 	;;
 
-	rsm psr.i				// mask interrupt delivery
+	RSM_PSR_I(p0, r18, r19)			// mask interrupt delivery
 	mov ar.ccv=0
 	andcm r14=r14,r17			// filter out SIGKILL & SIGSTOP
 
@@ -491,7 +492,7 @@ EX(.fail_efault, ld8 r14=[r33])			// r14 <- *set
 #ifdef CONFIG_SMP
 	st4.rel [r31]=r0			// release the lock
 #endif
-	ssm psr.i
+	SSM_PSR_I(p0, p9, r31)
 	;;
 
 	srlz.d					// ensure psr.i is set again
@@ -513,7 +514,7 @@ EX(.fail_efault, (p15) st8 [r34]=r3)
 #ifdef CONFIG_SMP
 	st4.rel [r31]=r0			// release the lock
 #endif
-	ssm psr.i
+	SSM_PSR_I(p0, p9, r17)
 	;;
 	srlz.d
 	br.sptk.many fsys_fallback_syscall	// with signal pending, do the heavy-weight syscall
@@ -521,7 +522,7 @@ EX(.fail_efault, (p15) st8 [r34]=r3)
 #ifdef CONFIG_SMP
 .lock_contention:
 	/* Rather than spinning here, fall back on doing a heavy-weight syscall.  */
-	ssm psr.i
+	SSM_PSR_I(p0, p9, r17)
 	;;
 	srlz.d
 	br.sptk.many fsys_fallback_syscall
@@ -592,17 +593,17 @@ ENTRY(fsys_fallback_syscall)
 	adds r17=-1024,r15
 	movl r14=sys_call_table
 	;;
-	rsm psr.i
+	RSM_PSR_I(p0, r26, r27)
 	shladd r18=r17,3,r14
 	;;
 	ld8 r18=[r18]				// load normal (heavy-weight) syscall entry-point
-	mov r29=psr				// read psr (12 cyc load latency)
+	MOV_FROM_PSR(p0, r29, r26)		// read psr (12 cyc load latency)
 	mov r27=ar.rsc
 	mov r21=ar.fpsr
 	mov r26=ar.pfs
 END(fsys_fallback_syscall)
 	/* FALL THROUGH */
-GLOBAL_ENTRY(fsys_bubble_down)
+GLOBAL_ENTRY(paravirt_fsys_bubble_down)
 	.prologue
 	.altrp b6
 	.body
@@ -640,7 +641,7 @@ GLOBAL_ENTRY(fsys_bubble_down)
 	 *
 	 * PSR.BE : already is turned off in __kernel_syscall_via_epc()
 	 * PSR.AC : don't care (kernel normally turns PSR.AC on)
-	 * PSR.I  : already turned off by the time fsys_bubble_down gets
+	 * PSR.I  : already turned off by the time paravirt_fsys_bubble_down gets
 	 *	    invoked
 	 * PSR.DFL: always 0 (kernel never turns it on)
 	 * PSR.DFH: don't care --- kernel never touches f32-f127 on its own
@@ -650,7 +651,7 @@ GLOBAL_ENTRY(fsys_bubble_down)
 	 * PSR.DB : don't care --- kernel never enables kernel-level
 	 *	    breakpoints
 	 * PSR.TB : must be 0 already; if it wasn't zero on entry to
-	 *          __kernel_syscall_via_epc, the branch to fsys_bubble_down
+	 *          __kernel_syscall_via_epc, the branch to paravirt_fsys_bubble_down
 	 *          will trigger a taken branch; the taken-trap-handler then
 	 *          converts the syscall into a break-based system-call.
 	 */
@@ -683,7 +684,7 @@ GLOBAL_ENTRY(fsys_bubble_down)
 	;;
 	mov ar.rsc=0				// M2   set enforced lazy mode, pl 0, LE, loadrs=0
 #ifdef CONFIG_VIRT_CPU_ACCOUNTING
-	mov.m r30=ar.itc			// M    get cycle for accounting
+	MOV_FROM_ITC(p0, p6, r30, r23)		// M    get cycle for accounting
 #else
 	nop.m 0
 #endif
@@ -734,21 +735,21 @@ GLOBAL_ENTRY(fsys_bubble_down)
 	mov rp=r14				// I0   set the real return addr
 	and r3=_TIF_SYSCALL_TRACEAUDIT,r3	// A
 	;;
-	ssm psr.i				// M2   we're on kernel stacks now, reenable irqs
+	SSM_PSR_I(p0, p6, r22)			// M2   we're on kernel stacks now, reenable irqs
 	cmp.eq p8,p0=r3,r0			// A
 (p10)	br.cond.spnt.many ia64_ret_from_syscall	// B    return if bad call-frame or r15 is a NaT
 
 	nop.m 0
 (p8)	br.call.sptk.many b6=b6			// B    (ignore return address)
 	br.cond.spnt ia64_trace_syscall		// B
-END(fsys_bubble_down)
+END(paravirt_fsys_bubble_down)
 
 	.rodata
 	.align 8
-	.globl fsyscall_table
+	.globl paravirt_fsyscall_table
 
-	data8 fsys_bubble_down
-fsyscall_table:
+	data8 paravirt_fsys_bubble_down
+paravirt_fsyscall_table:
 	data8 fsys_ni_syscall
 	data8 0				// exit			// 1025
 	data8 0				// read
@@ -1033,4 +1034,4 @@ fsyscall_table:
 
 	// fill in zeros for the remaining entries
 	.zero:
-	.space fsyscall_table + 8*NR_syscalls - .zero, 0
+	.space paravirt_fsyscall_table + 8*NR_syscalls - .zero, 0
diff --git a/arch/ia64/kernel/gate.S b/arch/ia64/kernel/gate.S
index 74b1ccc..cf5e0a1 100644
--- a/arch/ia64/kernel/gate.S
+++ b/arch/ia64/kernel/gate.S
@@ -13,6 +13,7 @@
 #include <asm/sigcontext.h>
 #include <asm/system.h>
 #include <asm/unistd.h>
+#include "paravirt_inst.h"
 
 /*
  * We can't easily refer to symbols inside the kernel.  To avoid full runtime relocation,
@@ -48,87 +49,6 @@ GLOBAL_ENTRY(__kernel_syscall_via_break)
 }
 END(__kernel_syscall_via_break)
 
-/*
- * On entry:
- *	r11 = saved ar.pfs
- *	r15 = system call #
- *	b0  = saved return address
- *	b6  = return address
- * On exit:
- *	r11 = saved ar.pfs
- *	r15 = system call #
- *	b0  = saved return address
- *	all other "scratch" registers:	undefined
- *	all "preserved" registers:	same as on entry
- */
-
-GLOBAL_ENTRY(__kernel_syscall_via_epc)
-	.prologue
-	.altrp b6
-	.body
-{
-	/*
-	 * Note: the kernel cannot assume that the first two instructions in this
-	 * bundle get executed.  The remaining code must be safe even if
-	 * they do not get executed.
-	 */
-	adds r17=-1024,r15			// A
-	mov r10=0				// A    default to successful syscall execution
-	epc					// B	causes split-issue
-}
-	;;
-	rsm psr.be | psr.i			// M2 (5 cyc to srlz.d)
-	LOAD_FSYSCALL_TABLE(r14)		// X
-	;;
-	mov r16=IA64_KR(CURRENT)		// M2 (12 cyc)
-	shladd r18=r17,3,r14			// A
-	mov r19=NR_syscalls-1			// A
-	;;
-	lfetch [r18]				// M0|1
-	mov r29=psr				// M2 (12 cyc)
-	// If r17 is a NaT, p6 will be zero
-	cmp.geu p6,p7=r19,r17			// A    (sysnr > 0 && sysnr < 1024+NR_syscalls)?
-	;;
-	mov r21=ar.fpsr				// M2 (12 cyc)
-	tnat.nz p10,p9=r15			// I0
-	mov.i r26=ar.pfs			// I0 (would stall anyhow due to srlz.d...)
-	;;
-	srlz.d					// M0 (forces split-issue) ensure PSR.BE==0
-(p6)	ld8 r18=[r18]				// M0|1
-	nop.i 0
-	;;
-	nop.m 0
-(p6)	tbit.z.unc p8,p0=r18,0			// I0 (dual-issues with "mov b7=r18"!)
-	nop.i 0
-	;;
-(p8)	ssm psr.i
-(p6)	mov b7=r18				// I0
-(p8)	br.dptk.many b7				// B
-
-	mov r27=ar.rsc				// M2 (12 cyc)
-/*
- * brl.cond doesn't work as intended because the linker would convert this branch
- * into a branch to a PLT.  Perhaps there will be a way to avoid this with some
- * future version of the linker.  In the meantime, we just use an indirect branch
- * instead.
- */
-#ifdef CONFIG_ITANIUM
-(p6)	add r14=-8,r14				// r14 <- addr of fsys_bubble_down entry
-	;;
-(p6)	ld8 r14=[r14]				// r14 <- fsys_bubble_down
-	;;
-(p6)	mov b7=r14
-(p6)	br.sptk.many b7
-#else
-	BRL_COND_FSYS_BUBBLE_DOWN(p6)
-#endif
-	ssm psr.i
-	mov r10=-1
-(p10)	mov r8=EINVAL
-(p9)	mov r8=ENOSYS
-	FSYS_RETURN
-END(__kernel_syscall_via_epc)
-
 #	define ARG0_OFF		(16 + IA64_SIGFRAME_ARG0_OFFSET)
 #	define ARG1_OFF		(16 + IA64_SIGFRAME_ARG1_OFFSET)
 #	define ARG2_OFF		(16 + IA64_SIGFRAME_ARG2_OFFSET)
@@ -374,3 +294,92 @@ restore_rbs:
 	// invala not necessary as that will happen when returning to user-mode
 	br.cond.sptk back_from_restore_rbs
 END(__kernel_sigtramp)
+
+/*
+ * On entry:
+ *	r11 = saved ar.pfs
+ *	r15 = system call #
+ *	b0  = saved return address
+ *	b6  = return address
+ * On exit:
+ *	r11 = saved ar.pfs
+ *	r15 = system call #
+ *	b0  = saved return address
+ *	all other "scratch" registers:	undefined
+ *	all "preserved" registers:	same as on entry
+ */
+
+GLOBAL_ENTRY(__kernel_syscall_via_epc)
+	.prologue
+	.altrp b6
+	.body
+{
+	/*
+	 * Note: the kernel cannot assume that the first two instructions in this
+	 * bundle get executed.  The remaining code must be safe even if
+	 * they do not get executed.
+	 */
+	adds r17=-1024,r15			// A
+	mov r10=0				// A    default to successful syscall execution
+	epc					// B	causes split-issue
+}
+	;;
+	RSM_PSR_BE_I(r20, r22)			// M2 (5 cyc to srlz.d)
+	LOAD_FSYSCALL_TABLE(r14)		// X
+	;;
+	mov r16=IA64_KR(CURRENT)		// M2 (12 cyc)
+	shladd r18=r17,3,r14			// A
+	mov r19=NR_syscalls-1			// A
+	;;
+	lfetch [r18]				// M0|1
+	MOV_FROM_PSR(p0, r29, r8)		// M2 (12 cyc)
+	// If r17 is a NaT, p6 will be zero
+	cmp.geu p6,p7=r19,r17			// A    (sysnr > 0 && sysnr < 1024+NR_syscalls)?
+	;;
+	mov r21=ar.fpsr				// M2 (12 cyc)
+	tnat.nz p10,p9=r15			// I0
+	mov.i r26=ar.pfs			// I0 (would stall anyhow due to srlz.d...)
+	;;
+	srlz.d					// M0 (forces split-issue) ensure PSR.BE==0
+(p6)	ld8 r18=[r18]				// M0|1
+	nop.i 0
+	;;
+	nop.m 0
+(p6)	tbit.z.unc p8,p0=r18,0			// I0 (dual-issues with "mov b7=r18"!)
+	nop.i 0
+	;;
+	SSM_PSR_I(p8, p14, r25)
+(p6)	mov b7=r18				// I0
+(p8)	br.dptk.many b7				// B
+
+	mov r27=ar.rsc				// M2 (12 cyc)
+/*
+ * brl.cond doesn't work as intended because the linker would convert this branch
+ * into a branch to a PLT.  Perhaps there will be a way to avoid this with some
+ * future version of the linker.  In the meantime, we just use an indirect branch
+ * instead.
+ */
+#ifdef CONFIG_ITANIUM
+(p6)	add r14=-8,r14				// r14 <- addr of fsys_bubble_down entry
+	;;
+(p6)	ld8 r14=[r14]				// r14 <- fsys_bubble_down
+	;;
+(p6)	mov b7=r14
+(p6)	br.sptk.many b7
+#else
+	BRL_COND_FSYS_BUBBLE_DOWN(p6)
+#endif
+	SSM_PSR_I(p0, p14, r10)
+	mov r10=-1
+(p10)	mov r8=EINVAL
+(p9)	mov r8=ENOSYS
+	FSYS_RETURN
+
+#ifdef CONFIG_PARAVIRT
+	/*
+	 * padd to make the size of this symbol constant
+	 * independent of paravirtualization.
+	 */
+	.align PAGE_SIZE / 8
+#endif
+END(__kernel_syscall_via_epc)
diff --git a/arch/ia64/kernel/gate.lds.S b/arch/ia64/kernel/gate.lds.S
index 3cb1abc..88c64ed 100644
--- a/arch/ia64/kernel/gate.lds.S
+++ b/arch/ia64/kernel/gate.lds.S
@@ -7,6 +7,7 @@
 
 
 #include <asm/system.h>
+#include "paravirt_patchlist.h"
 
 SECTIONS
 {
@@ -33,21 +34,21 @@ SECTIONS
 	. = GATE_ADDR + 0x600;
 
 	.data.patch		: {
-		__start_gate_mckinley_e9_patchlist = .;
+		__paravirt_start_gate_mckinley_e9_patchlist = .;
 		*(.data.patch.mckinley_e9)
-		__end_gate_mckinley_e9_patchlist = .;
+		__paravirt_end_gate_mckinley_e9_patchlist = .;
 
-		__start_gate_vtop_patchlist = .;
+		__paravirt_start_gate_vtop_patchlist = .;
 		*(.data.patch.vtop)
-		__end_gate_vtop_patchlist = .;
+		__paravirt_end_gate_vtop_patchlist = .;
 
-		__start_gate_fsyscall_patchlist = .;
+		__paravirt_start_gate_fsyscall_patchlist = .;
 		*(.data.patch.fsyscall_table)
-		__end_gate_fsyscall_patchlist = .;
+		__paravirt_end_gate_fsyscall_patchlist = .;
 
-		__start_gate_brl_fsys_bubble_down_patchlist = .;
+		__paravirt_start_gate_brl_fsys_bubble_down_patchlist = .;
 		*(.data.patch.brl_fsys_bubble_down)
-		__end_gate_brl_fsys_bubble_down_patchlist = .;
+		__paravirt_end_gate_brl_fsys_bubble_down_patchlist = .;
 	}						:readable
 
 	.IA_64.unwind_info	: { *(.IA_64.unwind_info*) }
diff --git a/arch/ia64/kernel/head.S b/arch/ia64/kernel/head.S
index 59301c4..23f846d 100644
--- a/arch/ia64/kernel/head.S
+++ b/arch/ia64/kernel/head.S
@@ -1050,7 +1050,7 @@ END(ia64_delay_loop)
  * except that the multiplication and the shift are done with 128-bit
  * intermediate precision so that we can produce a full 64-bit result.
  */
-GLOBAL_ENTRY(sched_clock)
+GLOBAL_ENTRY(ia64_native_sched_clock)
 	addl r8=THIS_CPU(cpu_info) + IA64_CPUINFO_NSEC_PER_CYC_OFFSET,r0
 	mov.m r9=ar.itc		// fetch cycle-counter				(35 cyc)
 	;;
@@ -1066,7 +1066,13 @@ GLOBAL_ENTRY(sched_clock)
 	;;
 	shrp r8=r9,r8,IA64_NSEC_PER_CYC_SHIFT
 	br.ret.sptk.many rp
-END(sched_clock)
+END(ia64_native_sched_clock)
+#ifndef CONFIG_PARAVIRT
+	//unsigned long long
+	//sched_clock(void) __attribute__((alias("ia64_native_sched_clock")));
+	.global sched_clock
+sched_clock = ia64_native_sched_clock
+#endif
 
 #ifdef CONFIG_VIRT_CPU_ACCOUNTING
 GLOBAL_ENTRY(cycle_to_cputime)
diff --git a/arch/ia64/kernel/ivt.S b/arch/ia64/kernel/ivt.S
index f675d8e3..ec9a5fd 100644
--- a/arch/ia64/kernel/ivt.S
+++ b/arch/ia64/kernel/ivt.S
@@ -804,7 +804,7 @@ ENTRY(break_fault)
 ///////////////////////////////////////////////////////////////////////
 	st1 [r16]=r0				// M2|3 clear current->thread.on_ustack flag
 #ifdef CONFIG_VIRT_CPU_ACCOUNTING
-	mov.m r30=ar.itc			// M    get cycle for accounting
+	MOV_FROM_ITC(p0, p14, r30, r18)		// M    get cycle for accounting
 #else
 	mov b6=r30				// I0   setup syscall handler branch reg early
 #endif
diff --git a/arch/ia64/kernel/mca.c b/arch/ia64/kernel/mca.c
index bab1de2..8f33a88 100644
--- a/arch/ia64/kernel/mca.c
+++ b/arch/ia64/kernel/mca.c
@@ -1456,9 +1456,9 @@ ia64_mca_cmc_int_caller(int cmc_irq, void *arg)
 
 	ia64_mca_cmc_int_handler(cmc_irq, arg);
 
-	for (++cpuid ; cpuid < NR_CPUS && !cpu_online(cpuid) ; cpuid++);
+	cpuid = cpumask_next(cpuid+1, cpu_online_mask);
 
-	if (cpuid < NR_CPUS) {
+	if (cpuid < nr_cpu_ids) {
 		platform_send_ipi(cpuid, IA64_CMCP_VECTOR, IA64_IPI_DM_INT, 0);
 	} else {
 		/* If no log record, switch out of polling mode */
@@ -1525,7 +1525,7 @@ ia64_mca_cpe_int_caller(int cpe_irq, void *arg)
 
 	ia64_mca_cpe_int_handler(cpe_irq, arg);
 
-	for (++cpuid ; cpuid < NR_CPUS && !cpu_online(cpuid) ; cpuid++);
+	cpuid = cpumask_next(cpuid+1, cpu_online_mask);
 
 	if (cpuid < NR_CPUS) {
 		platform_send_ipi(cpuid, IA64_CPEP_VECTOR, IA64_IPI_DM_INT, 0);
diff --git a/arch/ia64/kernel/module.c b/arch/ia64/kernel/module.c
index aaa7d9015..da3b0cf 100644
--- a/arch/ia64/kernel/module.c
+++ b/arch/ia64/kernel/module.c
@@ -446,6 +446,14 @@ module_frob_arch_sections (Elf_Ehdr *ehdr, Elf_Shdr *sechdrs, char *secstrings,
 			mod->arch.opd = s;
 		else if (strcmp(".IA_64.unwind", secstrings + s->sh_name) == 0)
 			mod->arch.unwind = s;
+#ifdef CONFIG_PARAVIRT
+		else if (strcmp(".paravirt_bundles",
+				secstrings + s->sh_name) == 0)
+			mod->arch.paravirt_bundles = s;
+		else if (strcmp(".paravirt_insts",
+				secstrings + s->sh_name) == 0)
+			mod->arch.paravirt_insts = s;
+#endif
 
 	if (!mod->arch.core_plt || !mod->arch.init_plt || !mod->arch.got || !mod->arch.opd) {
 		printk(KERN_ERR "%s: sections missing\n", mod->name);
@@ -525,8 +533,7 @@ get_ltoff (struct module *mod, uint64_t value, int *okp)
 			goto found;
 
 	/* Not enough GOT entries? */
-	if (e >= (struct got_entry *) (mod->arch.got->sh_addr + mod->arch.got->sh_size))
-		BUG();
+	BUG_ON(e >= (struct got_entry *) (mod->arch.got->sh_addr + mod->arch.got->sh_size));
 
 	e->val = value;
 	++mod->arch.next_got_entry;
@@ -921,6 +928,30 @@ module_finalize (const Elf_Ehdr *hdr, const Elf_Shdr *sechdrs, struct module *mo
 	DEBUGP("%s: init: entry=%p\n", __func__, mod->init);
 	if (mod->arch.unwind)
 		register_unwind_table(mod);
+#ifdef CONFIG_PARAVIRT
+        if (mod->arch.paravirt_bundles) {
+                struct paravirt_patch_site_bundle *start =
+                        (struct paravirt_patch_site_bundle *)
+                        mod->arch.paravirt_bundles->sh_addr;
+                struct paravirt_patch_site_bundle *end =
+                        (struct paravirt_patch_site_bundle *)
+                        (mod->arch.paravirt_bundles->sh_addr +
+                         mod->arch.paravirt_bundles->sh_size);
+
+                paravirt_patch_apply_bundle(start, end);
+        }
+        if (mod->arch.paravirt_insts) {
+                struct paravirt_patch_site_inst *start =
+                        (struct paravirt_patch_site_inst *)
+                        mod->arch.paravirt_insts->sh_addr;
+                struct paravirt_patch_site_inst *end =
+                        (struct paravirt_patch_site_inst *)
+                        (mod->arch.paravirt_insts->sh_addr +
+                         mod->arch.paravirt_insts->sh_size);
+
+                paravirt_patch_apply_inst(start, end);
+        }
+#endif
 	return 0;
 }
 
diff --git a/arch/ia64/kernel/palinfo.c b/arch/ia64/kernel/palinfo.c
index e5c57f4..a4f19c7 100644
--- a/arch/ia64/kernel/palinfo.c
+++ b/arch/ia64/kernel/palinfo.c
@@ -1002,8 +1002,6 @@ create_palinfo_proc_entries(unsigned int cpu)
 		*pdir = create_proc_read_entry(
 				palinfo_entries[j].name, 0, cpu_dir,
 				palinfo_read_entry, (void *)f.value);
-		if (*pdir)
-			(*pdir)->owner = THIS_MODULE;
 		pdir++;
 	}
 }
diff --git a/arch/ia64/kernel/paravirt.c b/arch/ia64/kernel/paravirt.c
index 9f14c16..a21d7bb 100644
--- a/arch/ia64/kernel/paravirt.c
+++ b/arch/ia64/kernel/paravirt.c
@@ -46,13 +46,23 @@ struct pv_info pv_info = {
  * initialization hooks.
  */
 
-struct pv_init_ops pv_init_ops;
+static void __init
+ia64_native_patch_branch(unsigned long tag, unsigned long type);
+
+struct pv_init_ops pv_init_ops =
+{
+#ifdef ASM_SUPPORTED
+	.patch_bundle = ia64_native_patch_bundle,
+#endif
+	.patch_branch = ia64_native_patch_branch,
+};
 
 /***************************************************************************
  * pv_cpu_ops
  * intrinsics hooks.
  */
 
+#ifndef ASM_SUPPORTED
 /* ia64_native_xxx are macros so that we have to make them real functions */
 
 #define DEFINE_VOID_FUNC1(name)					\
@@ -60,7 +70,14 @@ struct pv_init_ops pv_init_ops;
 	ia64_native_ ## name ## _func(unsigned long arg)	\
 	{							\
 		ia64_native_ ## name(arg);			\
-	}							\
+	}
+
+#define DEFINE_VOID_FUNC1_VOID(name)				\
+	static void						\
+	ia64_native_ ## name ## _func(void *arg)		\
+	{							\
+		ia64_native_ ## name(arg);			\
+	}
 
 #define DEFINE_VOID_FUNC2(name)					\
 	static void						\
@@ -68,7 +85,7 @@ struct pv_init_ops pv_init_ops;
 				      unsigned long arg1)	\
 	{							\
 		ia64_native_ ## name(arg0, arg1);		\
-	}							\
+	}
 
 #define DEFINE_FUNC0(name)			\
 	static unsigned long			\
@@ -84,7 +101,7 @@ struct pv_init_ops pv_init_ops;
 		return ia64_native_ ## name(arg);	\
 	}						\
 
-DEFINE_VOID_FUNC1(fc);
+DEFINE_VOID_FUNC1_VOID(fc);
 DEFINE_VOID_FUNC1(intrin_local_irq_restore);
 
 DEFINE_VOID_FUNC2(ptcga);
@@ -274,6 +291,266 @@ ia64_native_setreg_func(int regnum, unsigned long val)
 		break;
 	}
 }
+#else
+
+#define __DEFINE_FUNC(name, code)					\
+	extern const char ia64_native_ ## name ## _direct_start[];	\
+	extern const char ia64_native_ ## name ## _direct_end[];	\
+	asm (".align 32\n"						\
+	     ".proc ia64_native_" #name "_func\n"			\
+	     "ia64_native_" #name "_func:\n"				\
+	     "ia64_native_" #name "_direct_start:\n"			\
+	     code							\
+	     "ia64_native_" #name "_direct_end:\n"			\
+	     "br.cond.sptk.many b6\n"					\
+	     ".endp ia64_native_" #name "_func\n")
+
+#define DEFINE_VOID_FUNC0(name, code)				\
+	extern void						\
+	ia64_native_ ## name ## _func(void);			\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_VOID_FUNC1(name, code)				\
+	extern void						\
+	ia64_native_ ## name ## _func(unsigned long arg);	\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_VOID_FUNC1_VOID(name, code)			\
+	extern void						\
+	ia64_native_ ## name ## _func(void *arg);		\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_VOID_FUNC2(name, code)				\
+	extern void						\
+	ia64_native_ ## name ## _func(unsigned long arg0,	\
+				      unsigned long arg1);	\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_FUNC0(name, code)		\
+	extern unsigned long			\
+	ia64_native_ ## name ## _func(void);	\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_FUNC1(name, type, code)			\
+	extern unsigned long				\
+	ia64_native_ ## name ## _func(type arg);	\
+	__DEFINE_FUNC(name, code)
+
+DEFINE_VOID_FUNC1_VOID(fc,
+		       "fc r8\n");
+DEFINE_VOID_FUNC1(intrin_local_irq_restore,
+		  ";;\n"
+		  "     cmp.ne p6, p7 = r8, r0\n"
+		  ";;\n"
+		  "(p6) ssm psr.i\n"
+		  "(p7) rsm psr.i\n"
+		  ";;\n"
+		  "(p6) srlz.d\n");
+
+DEFINE_VOID_FUNC2(ptcga,
+		  "ptc.ga r8, r9\n");
+DEFINE_VOID_FUNC2(set_rr,
+		  "mov rr[r8] = r9\n");
+
+/* ia64_native_getreg(_IA64_REG_PSR) & IA64_PSR_I */
+DEFINE_FUNC0(get_psr_i,
+	     "mov r2 = " __stringify(1 << IA64_PSR_I_BIT) "\n"
+	     "mov r8 = psr\n"
+	     ";;\n"
+	     "and r8 = r2, r8\n");
+
+DEFINE_FUNC1(thash, unsigned long,
+	     "thash r8 = r8\n");
+DEFINE_FUNC1(get_cpuid, int,
+	     "mov r8 = cpuid[r8]\n");
+DEFINE_FUNC1(get_pmd, int,
+	     "mov r8 = pmd[r8]\n");
+DEFINE_FUNC1(get_rr, unsigned long,
+	     "mov r8 = rr[r8]\n");
+
+DEFINE_VOID_FUNC0(ssm_i,
+		  "ssm psr.i\n");
+DEFINE_VOID_FUNC0(rsm_i,
+		  "rsm psr.i\n");
+
+extern void
+ia64_native_set_rr0_to_rr4_func(unsigned long val0, unsigned long val1,
+				unsigned long val2, unsigned long val3,
+				unsigned long val4);
+__DEFINE_FUNC(set_rr0_to_rr4,
+	      "mov rr[r0] = r8\n"
+	      "movl r2 = 0x2000000000000000\n"
+	      ";;\n"
+	      "mov rr[r2] = r9\n"
+	      "shl r3 = r2, 1\n"	/* movl r3 = 0x4000000000000000 */
+	      ";;\n"
+	      "add r2 = r2, r3\n"	/* movl r2 = 0x6000000000000000 */
+	      "mov rr[r3] = r10\n"
+	      ";;\n"
+	      "mov rr[r2] = r11\n"
+	      "shl r3 = r3, 1\n"	/* movl r3 = 0x8000000000000000 */
+	      ";;\n"
+	      "mov rr[r3] = r14\n");
+
+extern unsigned long ia64_native_getreg_func(int regnum);
+asm(".global ia64_native_getreg_func\n");
+#define __DEFINE_GET_REG(id, reg)			\
+	"mov r2 = " __stringify(_IA64_REG_ ## id) "\n"	\
+	";;\n"						\
+	"cmp.eq p6, p0 = r2, r8\n"			\
+	";;\n"						\
+	"(p6) mov r8 = " #reg "\n"			\
+	"(p6) br.cond.sptk.many b6\n"			\
+	";;\n"
+#define __DEFINE_GET_AR(id, reg)	__DEFINE_GET_REG(AR_ ## id, ar.reg)
+#define __DEFINE_GET_CR(id, reg)	__DEFINE_GET_REG(CR_ ## id, cr.reg)
+
+__DEFINE_FUNC(getreg,
+	      __DEFINE_GET_REG(GP, gp)
+	      /*__DEFINE_GET_REG(IP, ip)*/ /* returned ip value shouldn't be constant */
+	      __DEFINE_GET_REG(PSR, psr)
+	      __DEFINE_GET_REG(TP, tp)
+	      __DEFINE_GET_REG(SP, sp)
+
+	      __DEFINE_GET_REG(AR_KR0, ar0)
+	      __DEFINE_GET_REG(AR_KR1, ar1)
+	      __DEFINE_GET_REG(AR_KR2, ar2)
+	      __DEFINE_GET_REG(AR_KR3, ar3)
+	      __DEFINE_GET_REG(AR_KR4, ar4)
+	      __DEFINE_GET_REG(AR_KR5, ar5)
+	      __DEFINE_GET_REG(AR_KR6, ar6)
+	      __DEFINE_GET_REG(AR_KR7, ar7)
+	      __DEFINE_GET_AR(RSC, rsc)
+	      __DEFINE_GET_AR(BSP, bsp)
+	      __DEFINE_GET_AR(BSPSTORE, bspstore)
+	      __DEFINE_GET_AR(RNAT, rnat)
+	      __DEFINE_GET_AR(FCR, fcr)
+	      __DEFINE_GET_AR(EFLAG, eflag)
+	      __DEFINE_GET_AR(CSD, csd)
+	      __DEFINE_GET_AR(SSD, ssd)
+	      __DEFINE_GET_REG(AR_CFLAG, ar27)
+	      __DEFINE_GET_AR(FSR, fsr)
+	      __DEFINE_GET_AR(FIR, fir)
+	      __DEFINE_GET_AR(FDR, fdr)
+	      __DEFINE_GET_AR(CCV, ccv)
+	      __DEFINE_GET_AR(UNAT, unat)
+	      __DEFINE_GET_AR(FPSR, fpsr)
+	      __DEFINE_GET_AR(ITC, itc)
+	      __DEFINE_GET_AR(PFS, pfs)
+	      __DEFINE_GET_AR(LC, lc)
+	      __DEFINE_GET_AR(EC, ec)
+
+	      __DEFINE_GET_CR(DCR, dcr)
+	      __DEFINE_GET_CR(ITM, itm)
+	      __DEFINE_GET_CR(IVA, iva)
+	      __DEFINE_GET_CR(PTA, pta)
+	      __DEFINE_GET_CR(IPSR, ipsr)
+	      __DEFINE_GET_CR(ISR, isr)
+	      __DEFINE_GET_CR(IIP, iip)
+	      __DEFINE_GET_CR(IFA, ifa)
+	      __DEFINE_GET_CR(ITIR, itir)
+	      __DEFINE_GET_CR(IIPA, iipa)
+	      __DEFINE_GET_CR(IFS, ifs)
+	      __DEFINE_GET_CR(IIM, iim)
+	      __DEFINE_GET_CR(IHA, iha)
+	      __DEFINE_GET_CR(LID, lid)
+	      __DEFINE_GET_CR(IVR, ivr)
+	      __DEFINE_GET_CR(TPR, tpr)
+	      __DEFINE_GET_CR(EOI, eoi)
+	      __DEFINE_GET_CR(IRR0, irr0)
+	      __DEFINE_GET_CR(IRR1, irr1)
+	      __DEFINE_GET_CR(IRR2, irr2)
+	      __DEFINE_GET_CR(IRR3, irr3)
+	      __DEFINE_GET_CR(ITV, itv)
+	      __DEFINE_GET_CR(PMV, pmv)
+	      __DEFINE_GET_CR(CMCV, cmcv)
+	      __DEFINE_GET_CR(LRR0, lrr0)
+	      __DEFINE_GET_CR(LRR1, lrr1)
+
+	      "mov r8 = -1\n"	/* unsupported case */
+	);
+
+extern void ia64_native_setreg_func(int regnum, unsigned long val);
+asm(".global ia64_native_setreg_func\n");
+#define __DEFINE_SET_REG(id, reg)			\
+	"mov r2 = " __stringify(_IA64_REG_ ## id) "\n"	\
+	";;\n"						\
+	"cmp.eq p6, p0 = r2, r9\n"			\
+	";;\n"						\
+	"(p6) mov " #reg " = r8\n"			\
+	"(p6) br.cond.sptk.many b6\n"			\
+	";;\n"
+#define __DEFINE_SET_AR(id, reg)	__DEFINE_SET_REG(AR_ ## id, ar.reg)
+#define __DEFINE_SET_CR(id, reg)	__DEFINE_SET_REG(CR_ ## id, cr.reg)
+__DEFINE_FUNC(setreg,
+	      "mov r2 = " __stringify(_IA64_REG_PSR_L) "\n"
+	      ";;\n"
+	      "cmp.eq p6, p0 = r2, r9\n"
+	      ";;\n"
+	      "(p6) mov psr.l = r8\n"
+#ifdef HAVE_SERIALIZE_DIRECTIVE
+	      ".serialize.data\n"
+#endif
+	      "(p6) br.cond.sptk.many b6\n"
+	      __DEFINE_SET_REG(GP, gp)
+	      __DEFINE_SET_REG(SP, sp)
+
+	      __DEFINE_SET_REG(AR_KR0, ar0)
+	      __DEFINE_SET_REG(AR_KR1, ar1)
+	      __DEFINE_SET_REG(AR_KR2, ar2)
+	      __DEFINE_SET_REG(AR_KR3, ar3)
+	      __DEFINE_SET_REG(AR_KR4, ar4)
+	      __DEFINE_SET_REG(AR_KR5, ar5)
+	      __DEFINE_SET_REG(AR_KR6, ar6)
+	      __DEFINE_SET_REG(AR_KR7, ar7)
+	      __DEFINE_SET_AR(RSC, rsc)
+	      __DEFINE_SET_AR(BSP, bsp)
+	      __DEFINE_SET_AR(BSPSTORE, bspstore)
+	      __DEFINE_SET_AR(RNAT, rnat)
+	      __DEFINE_SET_AR(FCR, fcr)
+	      __DEFINE_SET_AR(EFLAG, eflag)
+	      __DEFINE_SET_AR(CSD, csd)
+	      __DEFINE_SET_AR(SSD, ssd)
+	      __DEFINE_SET_REG(AR_CFLAG, ar27)
+	      __DEFINE_SET_AR(FSR, fsr)
+	      __DEFINE_SET_AR(FIR, fir)
+	      __DEFINE_SET_AR(FDR, fdr)
+	      __DEFINE_SET_AR(CCV, ccv)
+	      __DEFINE_SET_AR(UNAT, unat)
+	      __DEFINE_SET_AR(FPSR, fpsr)
+	      __DEFINE_SET_AR(ITC, itc)
+	      __DEFINE_SET_AR(PFS, pfs)
+	      __DEFINE_SET_AR(LC, lc)
+	      __DEFINE_SET_AR(EC, ec)
+
+	      __DEFINE_SET_CR(DCR, dcr)
+	      __DEFINE_SET_CR(ITM, itm)
+	      __DEFINE_SET_CR(IVA, iva)
+	      __DEFINE_SET_CR(PTA, pta)
+	      __DEFINE_SET_CR(IPSR, ipsr)
+	      __DEFINE_SET_CR(ISR, isr)
+	      __DEFINE_SET_CR(IIP, iip)
+	      __DEFINE_SET_CR(IFA, ifa)
+	      __DEFINE_SET_CR(ITIR, itir)
+	      __DEFINE_SET_CR(IIPA, iipa)
+	      __DEFINE_SET_CR(IFS, ifs)
+	      __DEFINE_SET_CR(IIM, iim)
+	      __DEFINE_SET_CR(IHA, iha)
+	      __DEFINE_SET_CR(LID, lid)
+	      __DEFINE_SET_CR(IVR, ivr)
+	      __DEFINE_SET_CR(TPR, tpr)
+	      __DEFINE_SET_CR(EOI, eoi)
+	      __DEFINE_SET_CR(IRR0, irr0)
+	      __DEFINE_SET_CR(IRR1, irr1)
+	      __DEFINE_SET_CR(IRR2, irr2)
+	      __DEFINE_SET_CR(IRR3, irr3)
+	      __DEFINE_SET_CR(ITV, itv)
+	      __DEFINE_SET_CR(PMV, pmv)
+	      __DEFINE_SET_CR(CMCV, cmcv)
+	      __DEFINE_SET_CR(LRR0, lrr0)
+	      __DEFINE_SET_CR(LRR1, lrr1)
+	);
+#endif
 
 struct pv_cpu_ops pv_cpu_ops = {
 	.fc		= ia64_native_fc_func,
@@ -366,4 +643,258 @@ ia64_native_do_steal_accounting(unsigned long *new_itm)
 
 struct pv_time_ops pv_time_ops = {
 	.do_steal_accounting = ia64_native_do_steal_accounting,
+	.sched_clock = ia64_native_sched_clock,
+};
+
+/***************************************************************************
+ * binary pacthing
+ * pv_init_ops.patch_bundle
+ */
+
+#ifdef ASM_SUPPORTED
+#define IA64_NATIVE_PATCH_DEFINE_GET_REG(name, reg)	\
+	__DEFINE_FUNC(get_ ## name,			\
+		      ";;\n"				\
+		      "mov r8 = " #reg "\n"		\
+		      ";;\n")
+
+#define IA64_NATIVE_PATCH_DEFINE_SET_REG(name, reg)	\
+	__DEFINE_FUNC(set_ ## name,			\
+		      ";;\n"				\
+		      "mov " #reg " = r8\n"		\
+		      ";;\n")
+
+#define IA64_NATIVE_PATCH_DEFINE_REG(name, reg)		\
+	IA64_NATIVE_PATCH_DEFINE_GET_REG(name, reg);	\
+	IA64_NATIVE_PATCH_DEFINE_SET_REG(name, reg)	\
+
+#define IA64_NATIVE_PATCH_DEFINE_AR(name, reg)			\
+	IA64_NATIVE_PATCH_DEFINE_REG(ar_ ## name, ar.reg)
+
+#define IA64_NATIVE_PATCH_DEFINE_CR(name, reg)			\
+	IA64_NATIVE_PATCH_DEFINE_REG(cr_ ## name, cr.reg)
+
+
+IA64_NATIVE_PATCH_DEFINE_GET_REG(psr, psr);
+IA64_NATIVE_PATCH_DEFINE_GET_REG(tp, tp);
+
+/* IA64_NATIVE_PATCH_DEFINE_SET_REG(psr_l, psr.l); */
+__DEFINE_FUNC(set_psr_l,
+	      ";;\n"
+	      "mov psr.l = r8\n"
+#ifdef HAVE_SERIALIZE_DIRECTIVE
+	      ".serialize.data\n"
+#endif
+	      ";;\n");
+
+IA64_NATIVE_PATCH_DEFINE_REG(gp, gp);
+IA64_NATIVE_PATCH_DEFINE_REG(sp, sp);
+
+IA64_NATIVE_PATCH_DEFINE_REG(kr0, ar0);
+IA64_NATIVE_PATCH_DEFINE_REG(kr1, ar1);
+IA64_NATIVE_PATCH_DEFINE_REG(kr2, ar2);
+IA64_NATIVE_PATCH_DEFINE_REG(kr3, ar3);
+IA64_NATIVE_PATCH_DEFINE_REG(kr4, ar4);
+IA64_NATIVE_PATCH_DEFINE_REG(kr5, ar5);
+IA64_NATIVE_PATCH_DEFINE_REG(kr6, ar6);
+IA64_NATIVE_PATCH_DEFINE_REG(kr7, ar7);
+
+IA64_NATIVE_PATCH_DEFINE_AR(rsc, rsc);
+IA64_NATIVE_PATCH_DEFINE_AR(bsp, bsp);
+IA64_NATIVE_PATCH_DEFINE_AR(bspstore, bspstore);
+IA64_NATIVE_PATCH_DEFINE_AR(rnat, rnat);
+IA64_NATIVE_PATCH_DEFINE_AR(fcr, fcr);
+IA64_NATIVE_PATCH_DEFINE_AR(eflag, eflag);
+IA64_NATIVE_PATCH_DEFINE_AR(csd, csd);
+IA64_NATIVE_PATCH_DEFINE_AR(ssd, ssd);
+IA64_NATIVE_PATCH_DEFINE_REG(ar27, ar27);
+IA64_NATIVE_PATCH_DEFINE_AR(fsr, fsr);
+IA64_NATIVE_PATCH_DEFINE_AR(fir, fir);
+IA64_NATIVE_PATCH_DEFINE_AR(fdr, fdr);
+IA64_NATIVE_PATCH_DEFINE_AR(ccv, ccv);
+IA64_NATIVE_PATCH_DEFINE_AR(unat, unat);
+IA64_NATIVE_PATCH_DEFINE_AR(fpsr, fpsr);
+IA64_NATIVE_PATCH_DEFINE_AR(itc, itc);
+IA64_NATIVE_PATCH_DEFINE_AR(pfs, pfs);
+IA64_NATIVE_PATCH_DEFINE_AR(lc, lc);
+IA64_NATIVE_PATCH_DEFINE_AR(ec, ec);
+
+IA64_NATIVE_PATCH_DEFINE_CR(dcr, dcr);
+IA64_NATIVE_PATCH_DEFINE_CR(itm, itm);
+IA64_NATIVE_PATCH_DEFINE_CR(iva, iva);
+IA64_NATIVE_PATCH_DEFINE_CR(pta, pta);
+IA64_NATIVE_PATCH_DEFINE_CR(ipsr, ipsr);
+IA64_NATIVE_PATCH_DEFINE_CR(isr, isr);
+IA64_NATIVE_PATCH_DEFINE_CR(iip, iip);
+IA64_NATIVE_PATCH_DEFINE_CR(ifa, ifa);
+IA64_NATIVE_PATCH_DEFINE_CR(itir, itir);
+IA64_NATIVE_PATCH_DEFINE_CR(iipa, iipa);
+IA64_NATIVE_PATCH_DEFINE_CR(ifs, ifs);
+IA64_NATIVE_PATCH_DEFINE_CR(iim, iim);
+IA64_NATIVE_PATCH_DEFINE_CR(iha, iha);
+IA64_NATIVE_PATCH_DEFINE_CR(lid, lid);
+IA64_NATIVE_PATCH_DEFINE_CR(ivr, ivr);
+IA64_NATIVE_PATCH_DEFINE_CR(tpr, tpr);
+IA64_NATIVE_PATCH_DEFINE_CR(eoi, eoi);
+IA64_NATIVE_PATCH_DEFINE_CR(irr0, irr0);
+IA64_NATIVE_PATCH_DEFINE_CR(irr1, irr1);
+IA64_NATIVE_PATCH_DEFINE_CR(irr2, irr2);
+IA64_NATIVE_PATCH_DEFINE_CR(irr3, irr3);
+IA64_NATIVE_PATCH_DEFINE_CR(itv, itv);
+IA64_NATIVE_PATCH_DEFINE_CR(pmv, pmv);
+IA64_NATIVE_PATCH_DEFINE_CR(cmcv, cmcv);
+IA64_NATIVE_PATCH_DEFINE_CR(lrr0, lrr0);
+IA64_NATIVE_PATCH_DEFINE_CR(lrr1, lrr1);
+
+static const struct paravirt_patch_bundle_elem ia64_native_patch_bundle_elems[]
+__initdata_or_module =
+{
+#define IA64_NATIVE_PATCH_BUNDLE_ELEM(name, type)		\
+	{							\
+		(void*)ia64_native_ ## name ## _direct_start,	\
+		(void*)ia64_native_ ## name ## _direct_end,	\
+		PARAVIRT_PATCH_TYPE_ ## type,			\
+	}
+
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(fc, FC),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(thash, THASH),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(get_cpuid, GET_CPUID),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(get_pmd, GET_PMD),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(ptcga, PTCGA),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(get_rr, GET_RR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(set_rr, SET_RR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(set_rr0_to_rr4, SET_RR0_TO_RR4),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(ssm_i, SSM_I),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(rsm_i, RSM_I),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(get_psr_i, GET_PSR_I),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM(intrin_local_irq_restore,
+				      INTRIN_LOCAL_IRQ_RESTORE),
+
+#define IA64_NATIVE_PATCH_BUNDLE_ELEM_GETREG(name, reg)			\
+	{								\
+		(void*)ia64_native_get_ ## name ## _direct_start,	\
+		(void*)ia64_native_get_ ## name ## _direct_end,		\
+		PARAVIRT_PATCH_TYPE_GETREG + _IA64_REG_ ## reg,		\
+	}
+
+#define IA64_NATIVE_PATCH_BUNDLE_ELEM_SETREG(name, reg)			\
+	{								\
+		(void*)ia64_native_set_ ## name ## _direct_start,	\
+		(void*)ia64_native_set_ ## name ## _direct_end,		\
+		PARAVIRT_PATCH_TYPE_SETREG + _IA64_REG_ ## reg,		\
+	}
+
+#define IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(name, reg)		\
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_GETREG(name, reg),	\
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_SETREG(name, reg)		\
+
+#define IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(name, reg)		\
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(ar_ ## name, AR_ ## reg)
+
+#define IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(name, reg)		\
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(cr_ ## name, CR_ ## reg)
+
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_GETREG(psr, PSR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_GETREG(tp, TP),
+
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_SETREG(psr_l, PSR_L),
+
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(gp, GP),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(sp, SP),
+
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr0, AR_KR0),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr1, AR_KR1),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr2, AR_KR2),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr3, AR_KR3),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr4, AR_KR4),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr5, AR_KR5),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr6, AR_KR6),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(kr7, AR_KR7),
+
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(rsc, RSC),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(bsp, BSP),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(bspstore, BSPSTORE),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(rnat, RNAT),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(fcr, FCR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(eflag, EFLAG),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(csd, CSD),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(ssd, SSD),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_REG(ar27, AR_CFLAG),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(fsr, FSR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(fir, FIR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(fdr, FDR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(ccv, CCV),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(unat, UNAT),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(fpsr, FPSR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(itc, ITC),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(pfs, PFS),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(lc, LC),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_AR(ec, EC),
+
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(dcr, DCR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(itm, ITM),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(iva, IVA),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(pta, PTA),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(ipsr, IPSR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(isr, ISR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(iip, IIP),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(ifa, IFA),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(itir, ITIR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(iipa, IIPA),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(ifs, IFS),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(iim, IIM),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(iha, IHA),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(lid, LID),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(ivr, IVR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(tpr, TPR),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(eoi, EOI),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(irr0, IRR0),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(irr1, IRR1),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(irr2, IRR2),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(irr3, IRR3),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(itv, ITV),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(pmv, PMV),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(cmcv, CMCV),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(lrr0, LRR0),
+	IA64_NATIVE_PATCH_BUNDLE_ELEM_CR(lrr1, LRR1),
 };
+
+unsigned long __init_or_module
+ia64_native_patch_bundle(void *sbundle, void *ebundle, unsigned long type)
+{
+	const unsigned long nelems = sizeof(ia64_native_patch_bundle_elems) /
+		sizeof(ia64_native_patch_bundle_elems[0]);
+
+	return __paravirt_patch_apply_bundle(sbundle, ebundle, type,
+					      ia64_native_patch_bundle_elems,
+					      nelems, NULL);
+}
+#endif /* ASM_SUPPOTED */
+
+extern const char ia64_native_switch_to[];
+extern const char ia64_native_leave_syscall[];
+extern const char ia64_native_work_processed_syscall[];
+extern const char ia64_native_leave_kernel[];
+
+const struct paravirt_patch_branch_target ia64_native_branch_target[]
+__initconst = {
+#define PARAVIRT_BR_TARGET(name, type)			\
+	{						\
+		ia64_native_ ## name,			\
+		PARAVIRT_PATCH_TYPE_BR_ ## type,	\
+	}
+	PARAVIRT_BR_TARGET(switch_to, SWITCH_TO),
+	PARAVIRT_BR_TARGET(leave_syscall, LEAVE_SYSCALL),
+	PARAVIRT_BR_TARGET(work_processed_syscall, WORK_PROCESSED_SYSCALL),
+	PARAVIRT_BR_TARGET(leave_kernel, LEAVE_KERNEL),
+};
+
+static void __init
+ia64_native_patch_branch(unsigned long tag, unsigned long type)
+{
+	const unsigned long nelem =
+		sizeof(ia64_native_branch_target) /
+		sizeof(ia64_native_branch_target[0]);
+	__paravirt_patch_apply_branch(tag, type,
+				      ia64_native_branch_target, nelem);
+}
diff --git a/arch/ia64/kernel/paravirt_patch.c b/arch/ia64/kernel/paravirt_patch.c
new file mode 100644
index 0000000..bfdfef1
--- /dev/null
+++ b/arch/ia64/kernel/paravirt_patch.c
@@ -0,0 +1,514 @@
+/******************************************************************************
+ * linux/arch/ia64/xen/paravirt_patch.c
+ *
+ * Copyright (c) 2008 Isaku Yamahata <yamahata at valinux co jp>
+ *                    VA Linux Systems Japan K.K.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that 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
+ *
+ */
+
+#include <linux/init.h>
+#include <asm/intrinsics.h>
+#include <asm/kprobes.h>
+#include <asm/paravirt.h>
+#include <asm/paravirt_patch.h>
+
+typedef union ia64_inst {
+        struct {
+		unsigned long long qp : 6;
+		unsigned long long : 31;
+		unsigned long long opcode : 4;
+		unsigned long long reserved : 23;
+        } generic;
+        unsigned long long l;
+} ia64_inst_t;
+
+/*
+ * flush_icache_range() can't be used here.
+ * we are here before cpu_init() which initializes
+ * ia64_i_cache_stride_shift. flush_icache_range() uses it.
+ */
+void __init_or_module
+paravirt_flush_i_cache_range(const void *instr, unsigned long size)
+{
+	extern void paravirt_fc_i(const void *addr);
+	unsigned long i;
+
+	for (i = 0; i < size; i += sizeof(bundle_t))
+		paravirt_fc_i(instr + i);
+}
+
+bundle_t* __init_or_module
+paravirt_get_bundle(unsigned long tag)
+{
+	return (bundle_t *)(tag & ~3UL);
+}
+
+unsigned long __init_or_module
+paravirt_get_slot(unsigned long tag)
+{
+	return tag & 3UL;
+}
+
+unsigned long __init_or_module
+paravirt_get_num_inst(unsigned long stag, unsigned long etag)
+{
+	bundle_t *sbundle = paravirt_get_bundle(stag);
+	unsigned long sslot = paravirt_get_slot(stag);
+	bundle_t *ebundle = paravirt_get_bundle(etag);
+	unsigned long eslot = paravirt_get_slot(etag);
+
+	return (ebundle - sbundle) * 3 + eslot - sslot + 1;
+}
+
+unsigned long __init_or_module
+paravirt_get_next_tag(unsigned long tag)
+{
+	unsigned long slot = paravirt_get_slot(tag);
+
+	switch (slot) {
+	case 0:
+	case 1:
+		return tag + 1;
+	case 2: {
+		bundle_t *bundle = paravirt_get_bundle(tag);
+		return (unsigned long)(bundle + 1);
+	}
+	default:
+		BUG();
+	}
+	/* NOTREACHED */
+}
+
+ia64_inst_t __init_or_module
+paravirt_read_slot0(const bundle_t *bundle)
+{
+	ia64_inst_t inst;
+	inst.l = bundle->quad0.slot0;
+	return inst;
+}
+
+ia64_inst_t __init_or_module
+paravirt_read_slot1(const bundle_t *bundle)
+{
+	ia64_inst_t inst;
+	inst.l = bundle->quad0.slot1_p0 |
+		((unsigned long long)bundle->quad1.slot1_p1 << 18UL);
+	return inst;
+}
+
+ia64_inst_t __init_or_module
+paravirt_read_slot2(const bundle_t *bundle)
+{
+	ia64_inst_t inst;
+	inst.l = bundle->quad1.slot2;
+	return inst;
+}
+
+ia64_inst_t __init_or_module
+paravirt_read_inst(unsigned long tag)
+{
+	bundle_t *bundle = paravirt_get_bundle(tag);
+	unsigned long slot = paravirt_get_slot(tag);
+
+	switch (slot) {
+	case 0:
+		return paravirt_read_slot0(bundle);
+	case 1:
+		return paravirt_read_slot1(bundle);
+	case 2:
+		return paravirt_read_slot2(bundle);
+	default:
+		BUG();
+	}
+	/* NOTREACHED */
+}
+
+void __init_or_module
+paravirt_write_slot0(bundle_t *bundle, ia64_inst_t inst)
+{
+	bundle->quad0.slot0 = inst.l;
+}
+
+void __init_or_module
+paravirt_write_slot1(bundle_t *bundle, ia64_inst_t inst)
+{
+	bundle->quad0.slot1_p0 = inst.l;
+	bundle->quad1.slot1_p1 = inst.l >> 18UL;
+}
+
+void __init_or_module
+paravirt_write_slot2(bundle_t *bundle, ia64_inst_t inst)
+{
+	bundle->quad1.slot2 = inst.l;
+}
+
+void __init_or_module
+paravirt_write_inst(unsigned long tag, ia64_inst_t inst)
+{
+	bundle_t *bundle = paravirt_get_bundle(tag);
+	unsigned long slot = paravirt_get_slot(tag);
+
+	switch (slot) {
+	case 0:
+		paravirt_write_slot0(bundle, inst);
+		break;
+	case 1:
+		paravirt_write_slot1(bundle, inst);
+		break;
+	case 2:
+		paravirt_write_slot2(bundle, inst);
+		break;
+	default:
+		BUG();
+		break;
+	}
+	paravirt_flush_i_cache_range(bundle, sizeof(*bundle));
+}
+
+/* for debug */
+void
+paravirt_print_bundle(const bundle_t *bundle)
+{
+	const unsigned long *quad = (const unsigned long *)bundle;
+	ia64_inst_t slot0 = paravirt_read_slot0(bundle);
+	ia64_inst_t slot1 = paravirt_read_slot1(bundle);
+	ia64_inst_t slot2 = paravirt_read_slot2(bundle);
+
+	printk(KERN_DEBUG
+	       "bundle 0x%p 0x%016lx 0x%016lx\n", bundle, quad[0], quad[1]);
+	printk(KERN_DEBUG
+	       "bundle template 0x%x\n",
+	       bundle->quad0.template);
+	printk(KERN_DEBUG
+	       "slot0 0x%lx slot1_p0 0x%lx slot1_p1 0x%lx slot2 0x%lx\n",
+	       (unsigned long)bundle->quad0.slot0,
+	       (unsigned long)bundle->quad0.slot1_p0,
+	       (unsigned long)bundle->quad1.slot1_p1,
+	       (unsigned long)bundle->quad1.slot2);
+	printk(KERN_DEBUG
+	       "slot0 0x%016llx slot1 0x%016llx slot2 0x%016llx\n",
+	       slot0.l, slot1.l, slot2.l);
+}
+
+static int noreplace_paravirt __init_or_module = 0;
+
+static int __init setup_noreplace_paravirt(char *str)
+{
+	noreplace_paravirt = 1;
+	return 1;
+}
+__setup("noreplace-paravirt", setup_noreplace_paravirt);
+
+#ifdef ASM_SUPPORTED
+static void __init_or_module
+fill_nop_bundle(void *sbundle, void *ebundle)
+{
+	extern const char paravirt_nop_bundle[];
+	extern const unsigned long paravirt_nop_bundle_size;
+
+	void *bundle = sbundle;
+
+	BUG_ON((((unsigned long)sbundle) % sizeof(bundle_t)) != 0);
+	BUG_ON((((unsigned long)ebundle) % sizeof(bundle_t)) != 0);
+
+	while (bundle < ebundle) {
+		memcpy(bundle, paravirt_nop_bundle, paravirt_nop_bundle_size);
+
+		bundle += paravirt_nop_bundle_size;
+	}
+}
+
+/* helper function */
+unsigned long __init_or_module
+__paravirt_patch_apply_bundle(void *sbundle, void *ebundle, unsigned long type,
+			      const struct paravirt_patch_bundle_elem *elems,
+			      unsigned long nelems,
+			      const struct paravirt_patch_bundle_elem **found)
+{
+	unsigned long used = 0;
+	unsigned long i;
+
+	BUG_ON((((unsigned long)sbundle) % sizeof(bundle_t)) != 0);
+	BUG_ON((((unsigned long)ebundle) % sizeof(bundle_t)) != 0);
+
+	found = NULL;
+	for (i = 0; i < nelems; i++) {
+		const struct paravirt_patch_bundle_elem *p = &elems[i];
+		if (p->type == type) {
+			unsigned long need = p->ebundle - p->sbundle;
+			unsigned long room = ebundle - sbundle;
+
+			if (found != NULL)
+				*found = p;
+
+			if (room < need) {
+				/* no room to replace. skip it */
+				printk(KERN_DEBUG
+				       "the space is too small to put "
+				       "bundles. type %ld need %ld room %ld\n",
+				       type, need, room);
+				break;
+			}
+
+			used = need;
+			memcpy(sbundle, p->sbundle, used);
+			break;
+		}
+	}
+
+	return used;
+}
+
+void __init_or_module
+paravirt_patch_apply_bundle(const struct paravirt_patch_site_bundle *start,
+			    const struct paravirt_patch_site_bundle *end)
+{
+	const struct paravirt_patch_site_bundle *p;
+
+	if (noreplace_paravirt)
+		return;
+	if (pv_init_ops.patch_bundle == NULL)
+		return;
+
+	for (p = start; p < end; p++) {
+		unsigned long used;
+
+		used = (*pv_init_ops.patch_bundle)(p->sbundle, p->ebundle,
+						   p->type);
+		if (used == 0)
+			continue;
+
+		fill_nop_bundle(p->sbundle + used, p->ebundle);
+		paravirt_flush_i_cache_range(p->sbundle,
+					     p->ebundle - p->sbundle);
+	}
+	ia64_sync_i();
+	ia64_srlz_i();
+}
+
+/*
+ * nop.i, nop.m, nop.f instruction are same format.
+ * but nop.b has differennt format.
+ * This doesn't support nop.b for now.
+ */
+static void __init_or_module
+fill_nop_inst(unsigned long stag, unsigned long etag)
+{
+	extern const bundle_t paravirt_nop_mfi_inst_bundle[];
+	unsigned long tag;
+	const ia64_inst_t nop_inst =
+		paravirt_read_slot0(paravirt_nop_mfi_inst_bundle);
+
+	for (tag = stag; tag < etag; tag = paravirt_get_next_tag(tag))
+		paravirt_write_inst(tag, nop_inst);
+}
+
+void __init_or_module
+paravirt_patch_apply_inst(const struct paravirt_patch_site_inst *start,
+			  const struct paravirt_patch_site_inst *end)
+{
+	const struct paravirt_patch_site_inst *p;
+
+	if (noreplace_paravirt)
+		return;
+	if (pv_init_ops.patch_inst == NULL)
+		return;
+
+	for (p = start; p < end; p++) {
+		unsigned long tag;
+		bundle_t *sbundle;
+		bundle_t *ebundle;
+
+		tag = (*pv_init_ops.patch_inst)(p->stag, p->etag, p->type);
+		if (tag == p->stag)
+			continue;
+
+		fill_nop_inst(tag, p->etag);
+		sbundle = paravirt_get_bundle(p->stag);
+		ebundle = paravirt_get_bundle(p->etag) + 1;
+		paravirt_flush_i_cache_range(sbundle, (ebundle - sbundle) *
+					     sizeof(bundle_t));
+	}
+	ia64_sync_i();
+	ia64_srlz_i();
+}
+#endif /* ASM_SUPPOTED */
+
+/* brl.cond.sptk.many <target64> X3 */
+typedef union inst_x3_op {
+	ia64_inst_t inst;
+	struct {
+		unsigned long qp: 6;
+		unsigned long btyp: 3;
+		unsigned long unused: 3;
+		unsigned long p: 1;
+		unsigned long imm20b: 20;
+		unsigned long wh: 2;
+		unsigned long d: 1;
+		unsigned long i: 1;
+		unsigned long opcode: 4;
+	};
+	unsigned long l;
+} inst_x3_op_t;
+
+typedef union inst_x3_imm {
+	ia64_inst_t inst;
+	struct {
+		unsigned long unused: 2;
+		unsigned long imm39: 39;
+	};
+	unsigned long l;
+} inst_x3_imm_t;
+
+void __init_or_module
+paravirt_patch_reloc_brl(unsigned long tag, const void *target)
+{
+	unsigned long tag_op = paravirt_get_next_tag(tag);
+	unsigned long tag_imm = tag;
+	bundle_t *bundle = paravirt_get_bundle(tag);
+
+	ia64_inst_t inst_op = paravirt_read_inst(tag_op);
+	ia64_inst_t inst_imm = paravirt_read_inst(tag_imm);
+
+	inst_x3_op_t inst_x3_op = { .l = inst_op.l };
+	inst_x3_imm_t inst_x3_imm = { .l = inst_imm.l };
+
+	unsigned long imm60 =
+		((unsigned long)target - (unsigned long)bundle) >> 4;
+
+	BUG_ON(paravirt_get_slot(tag) != 1); /* MLX */
+	BUG_ON(((unsigned long)target & (sizeof(bundle_t) - 1)) != 0);
+
+	/* imm60[59] 1bit */
+	inst_x3_op.i = (imm60 >> 59) & 1;
+	/* imm60[19:0] 20bit */
+	inst_x3_op.imm20b = imm60 & ((1UL << 20) - 1);
+	/* imm60[58:20] 39bit */
+	inst_x3_imm.imm39 = (imm60 >> 20) & ((1UL << 39) - 1);
+
+	inst_op.l = inst_x3_op.l;
+	inst_imm.l = inst_x3_imm.l;
+
+	paravirt_write_inst(tag_op, inst_op);
+	paravirt_write_inst(tag_imm, inst_imm);
+}
+
+/* br.cond.sptk.many <target25>	B1 */
+typedef union inst_b1 {
+	ia64_inst_t inst;
+	struct {
+		unsigned long qp: 6;
+		unsigned long btype: 3;
+		unsigned long unused: 3;
+		unsigned long p: 1;
+		unsigned long imm20b: 20;
+		unsigned long wh: 2;
+		unsigned long d: 1;
+		unsigned long s: 1;
+		unsigned long opcode: 4;
+	};
+	unsigned long l;
+} inst_b1_t;
+
+void __init
+paravirt_patch_reloc_br(unsigned long tag, const void *target)
+{
+	bundle_t *bundle = paravirt_get_bundle(tag);
+	ia64_inst_t inst = paravirt_read_inst(tag);
+	unsigned long target25 = (unsigned long)target - (unsigned long)bundle;
+	inst_b1_t inst_b1;
+
+	BUG_ON(((unsigned long)target & (sizeof(bundle_t) - 1)) != 0);
+
+	inst_b1.l = inst.l;
+	if (target25 & (1UL << 63))
+		inst_b1.s = 1;
+	else
+		inst_b1.s = 0;
+
+	inst_b1.imm20b = target25 >> 4;
+	inst.l = inst_b1.l;
+
+	paravirt_write_inst(tag, inst);
+}
+
+void __init
+__paravirt_patch_apply_branch(
+	unsigned long tag, unsigned long type,
+	const struct paravirt_patch_branch_target *entries,
+	unsigned int nr_entries)
+{
+	unsigned int i;
+	for (i = 0; i < nr_entries; i++) {
+		if (entries[i].type == type) {
+			paravirt_patch_reloc_br(tag, entries[i].entry);
+			break;
+		}
+	}
+}
+
+static void __init
+paravirt_patch_apply_branch(const struct paravirt_patch_site_branch *start,
+			    const struct paravirt_patch_site_branch *end)
+{
+	const struct paravirt_patch_site_branch *p;
+
+	if (noreplace_paravirt)
+		return;
+	if (pv_init_ops.patch_branch == NULL)
+		return;
+
+	for (p = start; p < end; p++)
+		(*pv_init_ops.patch_branch)(p->tag, p->type);
+
+	ia64_sync_i();
+	ia64_srlz_i();
+}
+
+void __init
+paravirt_patch_apply(void)
+{
+	extern const char __start_paravirt_bundles[];
+	extern const char __stop_paravirt_bundles[];
+	extern const char __start_paravirt_insts[];
+	extern const char __stop_paravirt_insts[];
+	extern const char __start_paravirt_branches[];
+	extern const char __stop_paravirt_branches[];
+
+	paravirt_patch_apply_bundle((const struct paravirt_patch_site_bundle *)
+				    __start_paravirt_bundles,
+				    (const struct paravirt_patch_site_bundle *)
+				    __stop_paravirt_bundles);
+	paravirt_patch_apply_inst((const struct paravirt_patch_site_inst *)
+				  __start_paravirt_insts,
+				  (const struct paravirt_patch_site_inst *)
+				  __stop_paravirt_insts);
+	paravirt_patch_apply_branch((const struct paravirt_patch_site_branch *)
+				    __start_paravirt_branches,
+				    (const struct paravirt_patch_site_branch *)
+				    __stop_paravirt_branches);
+}
+
+/*
+ * Local variables:
+ * mode: C
+ * c-set-style: "linux"
+ * c-basic-offset: 8
+ * tab-width: 8
+ * indent-tabs-mode: t
+ * End:
+ */
diff --git a/arch/ia64/kernel/paravirt_patchlist.c b/arch/ia64/kernel/paravirt_patchlist.c
new file mode 100644
index 0000000..b28082a
--- /dev/null
+++ b/arch/ia64/kernel/paravirt_patchlist.c
@@ -0,0 +1,79 @@
+/******************************************************************************
+ * Copyright (c) 2008 Isaku Yamahata <yamahata at valinux co jp>
+ *                    VA Linux Systems Japan K.K.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that 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
+ *
+ */
+
+#include <linux/bug.h>
+#include <asm/paravirt.h>
+
+#define DECLARE(name)						\
+	extern unsigned long					\
+		__ia64_native_start_gate_##name##_patchlist[];	\
+	extern unsigned long					\
+		__ia64_native_end_gate_##name##_patchlist[]
+
+DECLARE(fsyscall);
+DECLARE(brl_fsys_bubble_down);
+DECLARE(vtop);
+DECLARE(mckinley_e9);
+
+extern unsigned long __start_gate_section[];
+
+#define ASSIGN(name)							    \
+	.start_##name##_patchlist =					    \
+		(unsigned long)__ia64_native_start_gate_##name##_patchlist, \
+	.end_##name##_patchlist =					    \
+		(unsigned long)__ia64_native_end_gate_##name##_patchlist
+
+struct pv_patchdata pv_patchdata __initdata = {
+	ASSIGN(fsyscall),
+	ASSIGN(brl_fsys_bubble_down),
+	ASSIGN(vtop),
+	ASSIGN(mckinley_e9),
+
+	.gate_section = (void*)__start_gate_section,
+};
+
+
+unsigned long __init
+paravirt_get_gate_patchlist(enum pv_gate_patchlist type)
+{
+
+#define CASE(NAME, name)					\
+	case PV_GATE_START_##NAME:				\
+		return pv_patchdata.start_##name##_patchlist;	\
+	case PV_GATE_END_##NAME:				\
+		return pv_patchdata.end_##name##_patchlist;	\
+
+	switch (type) {
+		CASE(FSYSCALL, fsyscall);
+		CASE(BRL_FSYS_BUBBLE_DOWN, brl_fsys_bubble_down);
+		CASE(VTOP, vtop);
+		CASE(MCKINLEY_E9, mckinley_e9);
+	default:
+		BUG();
+		break;
+	}
+	return 0;
+}
+
+void * __init
+paravirt_get_gate_section(void)
+{
+	return pv_patchdata.gate_section;
+}
diff --git a/arch/ia64/kernel/paravirt_patchlist.h b/arch/ia64/kernel/paravirt_patchlist.h
new file mode 100644
index 0000000..0684aa6
--- /dev/null
+++ b/arch/ia64/kernel/paravirt_patchlist.h
@@ -0,0 +1,28 @@
+/******************************************************************************
+ * linux/arch/ia64/xen/paravirt_patchlist.h
+ *
+ * Copyright (c) 2008 Isaku Yamahata <yamahata at valinux co jp>
+ *                    VA Linux Systems Japan K.K.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that 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
+ *
+ */
+
+#if defined(__IA64_GATE_PARAVIRTUALIZED_XEN)
+#include <asm/xen/patchlist.h>
+#else
+#include <asm/native/patchlist.h>
+#endif
+
diff --git a/arch/ia64/kernel/paravirtentry.S b/arch/ia64/kernel/paravirtentry.S
index 2f42fcb..6158560 100644
--- a/arch/ia64/kernel/paravirtentry.S
+++ b/arch/ia64/kernel/paravirtentry.S
@@ -20,8 +20,11 @@
  *
  */
 
+#include <linux/init.h>
 #include <asm/asmmacro.h>
 #include <asm/asm-offsets.h>
+#include <asm/paravirt_privop.h>
+#include <asm/paravirt_patch.h>
 #include "entry.h"
 
 #define DATA8(sym, init_value)			\
@@ -32,29 +35,87 @@
 	data8 init_value ;			\
 	.popsection
 
-#define BRANCH(targ, reg, breg)		\
-	movl reg=targ ;			\
-	;;				\
-	ld8 reg=[reg] ;			\
-	;;				\
-	mov breg=reg ;			\
+#define BRANCH(targ, reg, breg, type)					\
+	PARAVIRT_PATCH_SITE_BR(PARAVIRT_PATCH_TYPE_BR_ ## type) ;	\
+	;;								\
+	movl reg=targ ;							\
+	;;								\
+	ld8 reg=[reg] ;							\
+	;;								\
+	mov breg=reg ;							\
 	br.cond.sptk.many breg
 
-#define BRANCH_PROC(sym, reg, breg)				\
-	DATA8(paravirt_ ## sym ## _targ, ia64_native_ ## sym) ; \
-	GLOBAL_ENTRY(paravirt_ ## sym) ;			\
-		BRANCH(paravirt_ ## sym ## _targ, reg, breg) ;	\
+#define BRANCH_PROC(sym, reg, breg, type)				\
+	DATA8(paravirt_ ## sym ## _targ, ia64_native_ ## sym) ;		\
+	GLOBAL_ENTRY(paravirt_ ## sym) ;				\
+		BRANCH(paravirt_ ## sym ## _targ, reg, breg, type) ;	\
 	END(paravirt_ ## sym)
 
-#define BRANCH_PROC_UNWINFO(sym, reg, breg)			\
-	DATA8(paravirt_ ## sym ## _targ, ia64_native_ ## sym) ; \
-	GLOBAL_ENTRY(paravirt_ ## sym) ;			\
-		PT_REGS_UNWIND_INFO(0) ;			\
-		BRANCH(paravirt_ ## sym ## _targ, reg, breg) ;	\
+#define BRANCH_PROC_UNWINFO(sym, reg, breg, type)			\
+	DATA8(paravirt_ ## sym ## _targ, ia64_native_ ## sym) ;		\
+	GLOBAL_ENTRY(paravirt_ ## sym) ;				\
+		PT_REGS_UNWIND_INFO(0) ;				\
+		BRANCH(paravirt_ ## sym ## _targ, reg, breg, type) ;	\
 	END(paravirt_ ## sym)
 
 
-BRANCH_PROC(switch_to, r22, b7)
-BRANCH_PROC_UNWINFO(leave_syscall, r22, b7)
-BRANCH_PROC(work_processed_syscall, r2, b7)
-BRANCH_PROC_UNWINFO(leave_kernel, r22, b7)
+BRANCH_PROC(switch_to, r22, b7, SWITCH_TO)
+BRANCH_PROC_UNWINFO(leave_syscall, r22, b7, LEAVE_SYSCALL)
+BRANCH_PROC(work_processed_syscall, r2, b7, WORK_PROCESSED_SYSCALL)
+BRANCH_PROC_UNWINFO(leave_kernel, r22, b7, LEAVE_KERNEL)
+
+
+#ifdef CONFIG_MODULES
+#define __INIT_OR_MODULE	.text
+#define __INITDATA_OR_MODULE	.data
+#else
+#define __INIT_OR_MODULE	__INIT
+#define __INITDATA_OR_MODULE	__INITDATA
+#endif /* CONFIG_MODULES */
+
+	__INIT_OR_MODULE
+	GLOBAL_ENTRY(paravirt_fc_i)
+	fc.i r32
+	br.ret.sptk.many rp
+	END(paravirt_fc_i)
+	__FINIT
+
+	__INIT_OR_MODULE
+	.align 32
+	GLOBAL_ENTRY(paravirt_nop_b_inst_bundle)
+	{
+		nop.b 0
+		nop.b 0
+		nop.b 0
+	}
+	END(paravirt_nop_b_inst_bundle)
+	__FINIT
+
+	/* NOTE: nop.[mfi] has same format */
+	__INIT_OR_MODULE
+	GLOBAL_ENTRY(paravirt_nop_mfi_inst_bundle)
+	{
+		nop.m 0
+		nop.f 0
+		nop.i 0
+	}
+	END(paravirt_nop_mfi_inst_bundle)
+	__FINIT
+
+	__INIT_OR_MODULE
+	GLOBAL_ENTRY(paravirt_nop_bundle)
+paravirt_nop_bundle_start:
+	{
+		nop 0
+		nop 0
+		nop 0
+	}
+paravirt_nop_bundle_end:
+	END(paravirt_nop_bundle)
+	__FINIT
+
+	__INITDATA_OR_MODULE
+	.align 8
+	.global paravirt_nop_bundle_size
+paravirt_nop_bundle_size:
+	data8	paravirt_nop_bundle_end - paravirt_nop_bundle_start
diff --git a/arch/ia64/kernel/patch.c b/arch/ia64/kernel/patch.c
index b83b2c5..68a1311 100644
--- a/arch/ia64/kernel/patch.c
+++ b/arch/ia64/kernel/patch.c
@@ -7,6 +7,7 @@
 #include <linux/init.h>
 #include <linux/string.h>
 
+#include <asm/paravirt.h>
 #include <asm/patch.h>
 #include <asm/processor.h>
 #include <asm/sections.h>
@@ -169,16 +170,35 @@ ia64_patch_mckinley_e9 (unsigned long start, unsigned long end)
 	ia64_srlz_i();
 }
 
+extern unsigned long ia64_native_fsyscall_table[NR_syscalls];
+extern char ia64_native_fsys_bubble_down[];
+struct pv_fsys_data pv_fsys_data __initdata = {
+	.fsyscall_table = (unsigned long *)ia64_native_fsyscall_table,
+	.fsys_bubble_down = (void *)ia64_native_fsys_bubble_down,
+};
+
+unsigned long * __init
+paravirt_get_fsyscall_table(void)
+{
+	return pv_fsys_data.fsyscall_table;
+}
+
+char * __init
+paravirt_get_fsys_bubble_down(void)
+{
+	return pv_fsys_data.fsys_bubble_down;
+}
+
 static void __init
 patch_fsyscall_table (unsigned long start, unsigned long end)
 {
-	extern unsigned long fsyscall_table[NR_syscalls];
+	u64 fsyscall_table = (u64)paravirt_get_fsyscall_table();
 	s32 *offp = (s32 *) start;
 	u64 ip;
 
 	while (offp < (s32 *) end) {
 		ip = (u64) ia64_imva((char *) offp + *offp);
-		ia64_patch_imm64(ip, (u64) fsyscall_table);
+		ia64_patch_imm64(ip, fsyscall_table);
 		ia64_fc((void *) ip);
 		++offp;
 	}
@@ -189,7 +209,7 @@ patch_fsyscall_table (unsigned long start, unsigned long end)
 static void __init
 patch_brl_fsys_bubble_down (unsigned long start, unsigned long end)
 {
-	extern char fsys_bubble_down[];
+	u64 fsys_bubble_down = (u64)paravirt_get_fsys_bubble_down();
 	s32 *offp = (s32 *) start;
 	u64 ip;
 
@@ -207,13 +227,13 @@ patch_brl_fsys_bubble_down (unsigned long start, unsigned long end)
 void __init
 ia64_patch_gate (void)
 {
-#	define START(name)	((unsigned long) __start_gate_##name##_patchlist)
-#	define END(name)	((unsigned long)__end_gate_##name##_patchlist)
+#	define START(name)	paravirt_get_gate_patchlist(PV_GATE_START_##name)
+#	define END(name)	paravirt_get_gate_patchlist(PV_GATE_END_##name)
 
-	patch_fsyscall_table(START(fsyscall), END(fsyscall));
-	patch_brl_fsys_bubble_down(START(brl_fsys_bubble_down), END(brl_fsys_bubble_down));
-	ia64_patch_vtop(START(vtop), END(vtop));
-	ia64_patch_mckinley_e9(START(mckinley_e9), END(mckinley_e9));
+	patch_fsyscall_table(START(FSYSCALL), END(FSYSCALL));
+	patch_brl_fsys_bubble_down(START(BRL_FSYS_BUBBLE_DOWN), END(BRL_FSYS_BUBBLE_DOWN));
+	ia64_patch_vtop(START(VTOP), END(VTOP));
+	ia64_patch_mckinley_e9(START(MCKINLEY_E9), END(MCKINLEY_E9));
 }
 
 void ia64_patch_phys_stack_reg(unsigned long val)
@@ -229,7 +249,7 @@ void ia64_patch_phys_stack_reg(unsigned long val)
 	while (offp < end) {
 		ip = (u64) offp + *offp;
 		ia64_patch(ip, mask, imm);
-		ia64_fc(ip);
+		ia64_fc((void *)ip);
 		++offp;
 	}
 	ia64_sync_i();
diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c
index 5c0f408..8a06dc48 100644
--- a/arch/ia64/kernel/perfmon.c
+++ b/arch/ia64/kernel/perfmon.c
@@ -5603,7 +5603,7 @@ pfm_interrupt_handler(int irq, void *arg)
  * /proc/perfmon interface, for debug only
  */
 
-#define PFM_PROC_SHOW_HEADER	((void *)NR_CPUS+1)
+#define PFM_PROC_SHOW_HEADER	((void *)nr_cpu_ids+1)
 
 static void *
 pfm_proc_start(struct seq_file *m, loff_t *pos)
@@ -5612,7 +5612,7 @@ pfm_proc_start(struct seq_file *m, loff_t *pos)
 		return PFM_PROC_SHOW_HEADER;
 	}
 
-	while (*pos <= NR_CPUS) {
+	while (*pos <= nr_cpu_ids) {
 		if (cpu_online(*pos - 1)) {
 			return (void *)*pos;
 		}
diff --git a/arch/ia64/kernel/salinfo.c b/arch/ia64/kernel/salinfo.c
index ecb9eb7..7053c55 100644
--- a/arch/ia64/kernel/salinfo.c
+++ b/arch/ia64/kernel/salinfo.c
@@ -317,7 +317,7 @@ retry:
 	}
 
 	n = data->cpu_check;
-	for (i = 0; i < NR_CPUS; i++) {
+	for (i = 0; i < nr_cpu_ids; i++) {
 		if (cpu_isset(n, data->cpu_event)) {
 			if (!cpu_online(n)) {
 				cpu_clear(n, data->cpu_event);
@@ -326,7 +326,7 @@ retry:
 			cpu = n;
 			break;
 		}
-		if (++n == NR_CPUS)
+		if (++n == nr_cpu_ids)
 			n = 0;
 	}
 
@@ -337,7 +337,7 @@ retry:
 
 	/* for next read, start checking at next CPU */
 	data->cpu_check = cpu;
-	if (++data->cpu_check == NR_CPUS)
+	if (++data->cpu_check == nr_cpu_ids)
 		data->cpu_check = 0;
 
 	snprintf(cmd, sizeof(cmd), "read %d\n", cpu);
diff --git a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c
index 865af27..714066a 100644
--- a/arch/ia64/kernel/setup.c
+++ b/arch/ia64/kernel/setup.c
@@ -52,6 +52,7 @@
 #include <asm/meminit.h>
 #include <asm/page.h>
 #include <asm/paravirt.h>
+#include <asm/paravirt_patch.h>
 #include <asm/patch.h>
 #include <asm/pgtable.h>
 #include <asm/processor.h>
@@ -537,6 +538,7 @@ setup_arch (char **cmdline_p)
 	paravirt_arch_setup_early();
 
 	ia64_patch_vtop((u64) __start___vtop_patchlist, (u64) __end___vtop_patchlist);
+	paravirt_patch_apply();
 
 	*cmdline_p = __va(ia64_boot_param->command_line);
 	strlcpy(boot_command_line, *cmdline_p, COMMAND_LINE_SIZE);
@@ -730,10 +732,10 @@ static void *
 c_start (struct seq_file *m, loff_t *pos)
 {
 #ifdef CONFIG_SMP
-	while (*pos < NR_CPUS && !cpu_isset(*pos, cpu_online_map))
+	while (*pos < nr_cpu_ids && !cpu_online(*pos))
 		++*pos;
 #endif
-	return *pos < NR_CPUS ? cpu_data(*pos) : NULL;
+	return *pos < nr_cpu_ids ? cpu_data(*pos) : NULL;
 }
 
 static void *
@@ -1016,8 +1018,7 @@ cpu_init (void)
 					| IA64_DCR_DA | IA64_DCR_DD | IA64_DCR_LC));
 	atomic_inc(&init_mm.mm_count);
 	current->active_mm = &init_mm;
-	if (current->mm)
-		BUG();
+	BUG_ON(current->mm);
 
 	ia64_mmu_init(ia64_imva(cpu_data));
 	ia64_mca_cpu_init(ia64_imva(cpu_data));
diff --git a/arch/ia64/kernel/smp.c b/arch/ia64/kernel/smp.c
index da8f020..2ea4199 100644
--- a/arch/ia64/kernel/smp.c
+++ b/arch/ia64/kernel/smp.c
@@ -166,11 +166,11 @@ send_IPI_allbutself (int op)
  * Called with preemption disabled.
  */
 static inline void
-send_IPI_mask(cpumask_t mask, int op)
+send_IPI_mask(const struct cpumask *mask, int op)
 {
 	unsigned int cpu;
 
-	for_each_cpu_mask(cpu, mask) {
+	for_each_cpu(cpu, mask) {
 			send_IPI_single(cpu, op);
 	}
 }
@@ -316,7 +316,7 @@ void arch_send_call_function_single_ipi(int cpu)
 	send_IPI_single(cpu, IPI_CALL_FUNC_SINGLE);
 }
 
-void arch_send_call_function_ipi(cpumask_t mask)
+void arch_send_call_function_ipi_mask(const struct cpumask *mask)
 {
 	send_IPI_mask(mask, IPI_CALL_FUNC);
 }
diff --git a/arch/ia64/kernel/smpboot.c b/arch/ia64/kernel/smpboot.c
index 5229054..7700e23 100644
--- a/arch/ia64/kernel/smpboot.c
+++ b/arch/ia64/kernel/smpboot.c
@@ -581,14 +581,14 @@ smp_build_cpu_map (void)
 
 	ia64_cpu_to_sapicid[0] = boot_cpu_id;
 	cpus_clear(cpu_present_map);
-	cpu_set(0, cpu_present_map);
-	cpu_set(0, cpu_possible_map);
+	set_cpu_present(0, true);
+	set_cpu_possible(0, true);
 	for (cpu = 1, i = 0; i < smp_boot_data.cpu_count; i++) {
 		sapicid = smp_boot_data.cpu_phys_id[i];
 		if (sapicid == boot_cpu_id)
 			continue;
-		cpu_set(cpu, cpu_present_map);
-		cpu_set(cpu, cpu_possible_map);
+		set_cpu_present(cpu, true);
+		set_cpu_possible(cpu, true);
 		ia64_cpu_to_sapicid[cpu] = sapicid;
 		cpu++;
 	}
@@ -626,12 +626,9 @@ smp_prepare_cpus (unsigned int max_cpus)
 	 */
 	if (!max_cpus) {
 		printk(KERN_INFO "SMP mode deactivated.\n");
-		cpus_clear(cpu_online_map);
-		cpus_clear(cpu_present_map);
-		cpus_clear(cpu_possible_map);
-		cpu_set(0, cpu_online_map);
-		cpu_set(0, cpu_present_map);
-		cpu_set(0, cpu_possible_map);
+		init_cpu_online(cpumask_of(0));
+		init_cpu_present(cpumask_of(0));
+		init_cpu_possible(cpumask_of(0));
 		return;
 	}
 }
diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c
index f0ebb34..641c8b6 100644
--- a/arch/ia64/kernel/time.c
+++ b/arch/ia64/kernel/time.c
@@ -20,6 +20,7 @@
 #include <linux/efi.h>
 #include <linux/timex.h>
 #include <linux/clocksource.h>
+#include <linux/platform_device.h>
 
 #include <asm/machvec.h>
 #include <asm/delay.h>
@@ -50,6 +51,15 @@ EXPORT_SYMBOL(last_cli_ip);
 #endif
 
 #ifdef CONFIG_PARAVIRT
+/* We need to define a real function for sched_clock, to override the
+   weak default version */
+unsigned long long sched_clock(void)
+{
+        return paravirt_sched_clock();
+}
+#endif
+
+#ifdef CONFIG_PARAVIRT
 static void
 paravirt_clocksource_resume(void)
 {
@@ -405,6 +415,21 @@ static struct irqaction timer_irqaction = {
 	.name =		"timer"
 };
 
+static struct platform_device rtc_efi_dev = {
+	.name = "rtc-efi",
+	.id = -1,
+};
+
+static int __init rtc_init(void)
+{
+	if (platform_device_register(&rtc_efi_dev) < 0)
+		printk(KERN_ERR "unable to register rtc device...\n");
+
+	/* not necessarily an error */
+	return 0;
+}
+module_init(rtc_init);
+
 void __init
 time_init (void)
 {
diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S
index 3765efc..4a95e86 100644
--- a/arch/ia64/kernel/vmlinux.lds.S
+++ b/arch/ia64/kernel/vmlinux.lds.S
@@ -169,6 +169,30 @@ SECTIONS
 	  __end___mckinley_e9_bundles = .;
 	}
 
+#if defined(CONFIG_PARAVIRT)
+  . = ALIGN(16);
+  .paravirt_bundles : AT(ADDR(.paravirt_bundles) - LOAD_OFFSET)
+	{
+	  __start_paravirt_bundles = .;
+          *(.paravirt_bundles)
+	  __stop_paravirt_bundles = .;
+	}
+  . = ALIGN(16);
+  .paravirt_insts : AT(ADDR(.paravirt_insts) - LOAD_OFFSET)
+	{
+	  __start_paravirt_insts = .;
+          *(.paravirt_insts)
+	  __stop_paravirt_insts = .;
+	}
+  . = ALIGN(16);
+  .paravirt_branches : AT(ADDR(.paravirt_branches) - LOAD_OFFSET)
+	{
+	  __start_paravirt_branches = .;
+	  *(.paravirt_branches)
+	  __stop_paravirt_branches = .;
+	}
+#endif
+
 #if defined(CONFIG_IA64_GENERIC)
   /* Machine Vector */
   . = ALIGN(16);
@@ -201,6 +225,12 @@ SECTIONS
 	  __start_gate_section = .;
 	  *(.data.gate)
 	  __stop_gate_section = .;
+#ifdef CONFIG_XEN
+	  . = ALIGN(PAGE_SIZE);
+	  __xen_start_gate_section = .;
+	  *(.data.gate.xen)
+	  __xen_stop_gate_section = .;
+#endif
 	}
   . = ALIGN(PAGE_SIZE);		/* make sure the gate page doesn't expose
   				 * kernel data
diff --git a/arch/ia64/kvm/kvm-ia64.c b/arch/ia64/kvm/kvm-ia64.c
index 076b00d..28af6a7 100644
--- a/arch/ia64/kvm/kvm-ia64.c
+++ b/arch/ia64/kvm/kvm-ia64.c
@@ -70,7 +70,7 @@ static void kvm_flush_icache(unsigned long start, unsigned long len)
 	int l;
 
 	for (l = 0; l < (len + 32); l += 32)
-		ia64_fc(start + l);
+		ia64_fc((void *)(start + l));
 
 	ia64_sync_i();
 	ia64_srlz_i();
diff --git a/arch/ia64/kvm/vcpu.c b/arch/ia64/kvm/vcpu.c
index d4d2805..a18ee17 100644
--- a/arch/ia64/kvm/vcpu.c
+++ b/arch/ia64/kvm/vcpu.c
@@ -386,7 +386,7 @@ void set_rse_reg(struct kvm_pt_regs *regs, unsigned long r1,
 		else
 			*rnat_addr = (*rnat_addr) & (~nat_mask);
 
-		ia64_setreg(_IA64_REG_AR_BSPSTORE, bspstore);
+		ia64_setreg(_IA64_REG_AR_BSPSTORE, (unsigned long)bspstore);
 		ia64_setreg(_IA64_REG_AR_RNAT, rnat);
 	}
 	local_irq_restore(psr);
diff --git a/arch/ia64/kvm/vtlb.c b/arch/ia64/kvm/vtlb.c
index 38232b3..2c2501f 100644
--- a/arch/ia64/kvm/vtlb.c
+++ b/arch/ia64/kvm/vtlb.c
@@ -210,6 +210,7 @@ void thash_vhpt_insert(struct kvm_vcpu *v, u64 pte, u64 itir, u64 va, int type)
 		phy_pte  &= ~PAGE_FLAGS_RV_MASK;
 		psr = ia64_clear_ic();
 		ia64_itc(type, va, phy_pte, itir_ps(itir));
+		paravirt_dv_serialize_data();
 		ia64_set_psr(psr);
 	}
 
@@ -456,6 +457,7 @@ void  thash_purge_and_insert(struct kvm_vcpu *v, u64 pte, u64 itir,
 		phy_pte  &= ~PAGE_FLAGS_RV_MASK;
 		psr = ia64_clear_ic();
 		ia64_itc(type, ifa, phy_pte, ps);
+		paravirt_dv_serialize_data();
 		ia64_set_psr(psr);
 	}
 	if (!(pte&VTLB_PTE_IO))
diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c
index 56e1290..c0f3bee 100644
--- a/arch/ia64/mm/init.c
+++ b/arch/ia64/mm/init.c
@@ -35,6 +35,7 @@
 #include <asm/uaccess.h>
 #include <asm/unistd.h>
 #include <asm/mca.h>
+#include <asm/paravirt.h>
 
 DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
 
@@ -259,6 +260,7 @@ put_kernel_page (struct page *page, unsigned long address, pgprot_t pgprot)
 static void __init
 setup_gate (void)
 {
+	void *gate_section;
 	struct page *page;
 
 	/*
@@ -266,10 +268,11 @@ setup_gate (void)
 	 * headers etc. and once execute-only page to enable
 	 * privilege-promotion via "epc":
 	 */
-	page = virt_to_page(ia64_imva(__start_gate_section));
+	gate_section = paravirt_get_gate_section();
+	page = virt_to_page(ia64_imva(gate_section));
 	put_kernel_page(page, GATE_ADDR, PAGE_READONLY);
 #ifdef HAVE_BUGGY_SEGREL
-	page = virt_to_page(ia64_imva(__start_gate_section + PAGE_SIZE));
+	page = virt_to_page(ia64_imva(gate_section + PAGE_SIZE));
 	put_kernel_page(page, GATE_ADDR + PAGE_SIZE, PAGE_GATE);
 #else
 	put_kernel_page(page, GATE_ADDR + PERCPU_PAGE_SIZE, PAGE_GATE);
@@ -633,8 +636,7 @@ mem_init (void)
 #endif
 
 #ifdef CONFIG_FLATMEM
-	if (!mem_map)
-		BUG();
+	BUG_ON(!mem_map);
 	max_mapnr = max_low_pfn;
 #endif
 
@@ -667,8 +669,8 @@ mem_init (void)
 	 * code can tell them apart.
 	 */
 	for (i = 0; i < NR_syscalls; ++i) {
-		extern unsigned long fsyscall_table[NR_syscalls];
 		extern unsigned long sys_call_table[NR_syscalls];
+		unsigned long *fsyscall_table = paravirt_get_fsyscall_table();
 
 		if (!fsyscall_table[i] || nolwsys)
 			fsyscall_table[i] = sys_call_table[i] | 1;
diff --git a/arch/ia64/mm/tlb.c b/arch/ia64/mm/tlb.c
index bd9818a..b9f3d7b 100644
--- a/arch/ia64/mm/tlb.c
+++ b/arch/ia64/mm/tlb.c
@@ -309,7 +309,7 @@ flush_tlb_range (struct vm_area_struct *vma, unsigned long start,
 
 	preempt_disable();
 #ifdef CONFIG_SMP
-	if (mm != current->active_mm || cpus_weight(mm->cpu_vm_mask) != 1) {
+	if (mm != current->active_mm || cpumask_weight(mm_cpumask(mm)) != 1) {
 		platform_global_tlb_purge(mm, start, end, nbits);
 		preempt_enable();
 		return;
diff --git a/arch/ia64/scripts/pvcheck.sed b/arch/ia64/scripts/pvcheck.sed
index ba66ac2..e59809a 100644
--- a/arch/ia64/scripts/pvcheck.sed
+++ b/arch/ia64/scripts/pvcheck.sed
@@ -17,6 +17,7 @@ s/mov.*=.*cr\.iip/.warning \"cr.iip should not used directly\"/g
 s/mov.*=.*cr\.ivr/.warning \"cr.ivr should not used directly\"/g
 s/mov.*=[^\.]*psr/.warning \"psr should not used directly\"/g	# avoid ar.fpsr
 s/mov.*=.*ar\.eflags/.warning \"ar.eflags should not used directly\"/g
+s/mov.*=.*ar\.itc.*/.warning \"ar.itc should not used directly\"/g
 s/mov.*cr\.ifa.*=.*/.warning \"cr.ifa should not used directly\"/g
 s/mov.*cr\.itir.*=.*/.warning \"cr.itir should not used directly\"/g
 s/mov.*cr\.iha.*=.*/.warning \"cr.iha should not used directly\"/g
diff --git a/arch/ia64/sn/kernel/io_common.c b/arch/ia64/sn/kernel/io_common.c
index 0d4ffa4..57f280d 100644
--- a/arch/ia64/sn/kernel/io_common.c
+++ b/arch/ia64/sn/kernel/io_common.c
@@ -135,8 +135,7 @@ static s64 sn_device_fixup_war(u64 nasid, u64 widget, int device,
 	}
 
 	war_list = kzalloc(DEV_PER_WIDGET * sizeof(*war_list), GFP_KERNEL);
-	if (!war_list)
-		BUG();
+	BUG_ON(!war_list);
 
 	SAL_CALL_NOLOCK(isrv, SN_SAL_IOIF_GET_WIDGET_DMAFLUSH_LIST,
 			nasid, widget, __pa(war_list), 0, 0, 0 ,0);
@@ -180,23 +179,20 @@ sn_common_hubdev_init(struct hubdev_info *hubdev)
 		sizeof(struct sn_flush_device_kernel *);
 	hubdev->hdi_flush_nasid_list.widget_p =
 		kzalloc(size, GFP_KERNEL);
-	if (!hubdev->hdi_flush_nasid_list.widget_p)
-		BUG();
+	BUG_ON(!hubdev->hdi_flush_nasid_list.widget_p);
 
 	for (widget = 0; widget <= HUB_WIDGET_ID_MAX; widget++) {
 		size = DEV_PER_WIDGET *
 			sizeof(struct sn_flush_device_kernel);
 		sn_flush_device_kernel = kzalloc(size, GFP_KERNEL);
-		if (!sn_flush_device_kernel)
-			BUG();
+		BUG_ON(!sn_flush_device_kernel);
 
 		dev_entry = sn_flush_device_kernel;
 		for (device = 0; device < DEV_PER_WIDGET;
 		     device++, dev_entry++) {
 			size = sizeof(struct sn_flush_device_common);
 			dev_entry->common = kzalloc(size, GFP_KERNEL);
-			if (!dev_entry->common)
-				BUG();
+			BUG_ON(!dev_entry->common);
 			if (sn_prom_feature_available(PRF_DEVICE_FLUSH_LIST))
 				status = sal_get_device_dmaflush_list(
 					     hubdev->hdi_nasid, widget, device,
@@ -326,8 +322,7 @@ sn_common_bus_fixup(struct pci_bus *bus,
 	 */
 	controller->platform_data = kzalloc(sizeof(struct sn_platform_data),
 					    GFP_KERNEL);
-	if (controller->platform_data == NULL)
-		BUG();
+	BUG_ON(controller->platform_data == NULL);
 	sn_platform_data =
 			(struct sn_platform_data *) controller->platform_data;
 	sn_platform_data->provider_soft = provider_soft;
diff --git a/arch/ia64/sn/kernel/io_init.c b/arch/ia64/sn/kernel/io_init.c
index e2eb2da..ee774c3 100644
--- a/arch/ia64/sn/kernel/io_init.c
+++ b/arch/ia64/sn/kernel/io_init.c
@@ -128,8 +128,7 @@ sn_legacy_pci_window_fixup(struct pci_controller *controller,
 {
 		controller->window = kcalloc(2, sizeof(struct pci_window),
 					     GFP_KERNEL);
-		if (controller->window == NULL)
-			BUG();
+		BUG_ON(controller->window == NULL);
 		controller->window[0].offset = legacy_io;
 		controller->window[0].resource.name = "legacy_io";
 		controller->window[0].resource.flags = IORESOURCE_IO;
@@ -168,8 +167,7 @@ sn_pci_window_fixup(struct pci_dev *dev, unsigned int count,
 	idx = controller->windows;
 	new_count = controller->windows + count;
 	new_window = kcalloc(new_count, sizeof(struct pci_window), GFP_KERNEL);
-	if (new_window == NULL)
-		BUG();
+	BUG_ON(new_window == NULL);
 	if (controller->window) {
 		memcpy(new_window, controller->window,
 		       sizeof(struct pci_window) * controller->windows);
@@ -222,8 +220,7 @@ sn_io_slot_fixup(struct pci_dev *dev)
 		(u64) __pa(pcidev_info),
 		(u64) __pa(sn_irq_info));
 
-	if (status)
-		BUG(); /* Cannot get platform pci device information */
+	BUG_ON(status); /* Cannot get platform pci device information */
 
 
 	/* Copy over PIO Mapped Addresses */
@@ -307,8 +304,7 @@ sn_pci_controller_fixup(int segment, int busnum, struct pci_bus *bus)
 	prom_bussoft_ptr = __va(prom_bussoft_ptr);
 
 	controller = kzalloc(sizeof(*controller), GFP_KERNEL);
-	if (!controller)
-		BUG();
+	BUG_ON(!controller);
 	controller->segment = segment;
 
 	/*
diff --git a/arch/ia64/sn/kernel/setup.c b/arch/ia64/sn/kernel/setup.c
index 02c5b8a..e456f06 100644
--- a/arch/ia64/sn/kernel/setup.c
+++ b/arch/ia64/sn/kernel/setup.c
@@ -732,8 +732,7 @@ void __init build_cnode_tables(void)
 		kl_config_hdr_t *klgraph_header;
 		nasid = cnodeid_to_nasid(node);
 		klgraph_header = ia64_sn_get_klconfig_addr(nasid);
-		if (klgraph_header == NULL)
-			BUG();
+		BUG_ON(klgraph_header == NULL);
 		brd = NODE_OFFSET_TO_LBOARD(nasid, klgraph_header->ch_board_info);
 		while (brd) {
 			if (board_needs_cnode(brd->brd_type) && physical_node_map[brd->brd_nasid] < 0) {
@@ -750,7 +749,7 @@ nasid_slice_to_cpuid(int nasid, int slice)
 {
 	long cpu;
 
-	for (cpu = 0; cpu < NR_CPUS; cpu++)
+	for (cpu = 0; cpu < nr_cpu_ids; cpu++)
 		if (cpuid_to_nasid(cpu) == nasid &&
 					cpuid_to_slice(cpu) == slice)
 			return cpu;
diff --git a/arch/ia64/sn/kernel/sn2/prominfo_proc.c b/arch/ia64/sn/kernel/sn2/prominfo_proc.c
index 4dcce3d..e633288 100644
--- a/arch/ia64/sn/kernel/sn2/prominfo_proc.c
+++ b/arch/ia64/sn/kernel/sn2/prominfo_proc.c
@@ -225,7 +225,6 @@ static struct proc_dir_entry *sgi_prominfo_entry;
 int __init prominfo_init(void)
 {
 	struct proc_dir_entry **entp;
-	struct proc_dir_entry *p;
 	cnodeid_t cnodeid;
 	unsigned long nasid;
 	int size;
@@ -246,14 +245,10 @@ int __init prominfo_init(void)
 		sprintf(name, "node%d", cnodeid);
 		*entp = proc_mkdir(name, sgi_prominfo_entry);
 		nasid = cnodeid_to_nasid(cnodeid);
-		p = create_proc_read_entry("fit", 0, *entp, read_fit_entry,
+		create_proc_read_entry("fit", 0, *entp, read_fit_entry,
 					   (void *)nasid);
-		if (p)
-			p->owner = THIS_MODULE;
-		p = create_proc_read_entry("version", 0, *entp,
+		create_proc_read_entry("version", 0, *entp,
 					   read_version_entry, (void *)nasid);
-		if (p)
-			p->owner = THIS_MODULE;
 		entp++;
 	}
 
diff --git a/arch/ia64/sn/kernel/sn2/sn2_smp.c b/arch/ia64/sn/kernel/sn2/sn2_smp.c
index e585f9a..1176506 100644
--- a/arch/ia64/sn/kernel/sn2/sn2_smp.c
+++ b/arch/ia64/sn/kernel/sn2/sn2_smp.c
@@ -133,7 +133,7 @@ sn2_ipi_flush_all_tlb(struct mm_struct *mm)
 	unsigned long itc;
 
 	itc = ia64_get_itc();
-	smp_flush_tlb_cpumask(mm->cpu_vm_mask);
+	smp_flush_tlb_cpumask(*mm_cpumask(mm));
 	itc = ia64_get_itc() - itc;
 	__get_cpu_var(ptcstats).shub_ipi_flushes_itc_clocks += itc;
 	__get_cpu_var(ptcstats).shub_ipi_flushes++;
@@ -182,7 +182,7 @@ sn2_global_tlb_purge(struct mm_struct *mm, unsigned long start,
 	nodes_clear(nodes_flushed);
 	i = 0;
 
-	for_each_cpu_mask(cpu, mm->cpu_vm_mask) {
+	for_each_cpu(cpu, mm_cpumask(mm)) {
 		cnode = cpu_to_node(cpu);
 		node_set(cnode, nodes_flushed);
 		lcpu = cpu;
@@ -461,7 +461,7 @@ bool sn_cpu_disable_allowed(int cpu)
 
 static void *sn2_ptc_seq_start(struct seq_file *file, loff_t * offset)
 {
-	if (*offset < NR_CPUS)
+	if (*offset < nr_cpu_ids)
 		return offset;
 	return NULL;
 }
@@ -469,7 +469,7 @@ static void *sn2_ptc_seq_start(struct seq_file *file, loff_t * offset)
 static void *sn2_ptc_seq_next(struct seq_file *file, void *data, loff_t * offset)
 {
 	(*offset)++;
-	if (*offset < NR_CPUS)
+	if (*offset < nr_cpu_ids)
 		return offset;
 	return NULL;
 }
@@ -491,7 +491,7 @@ static int sn2_ptc_seq_show(struct seq_file *file, void *data)
 		seq_printf(file, "# ptctest %d, flushopt %d\n", sn2_ptctest, sn2_flush_opt);
 	}
 
-	if (cpu < NR_CPUS && cpu_online(cpu)) {
+	if (cpu < nr_cpu_ids && cpu_online(cpu)) {
 		stat = &per_cpu(ptcstats, cpu);
 		seq_printf(file, "cpu %d %ld %ld %ld %ld %ld %ld %ld %ld %ld %ld %ld %ld\n", cpu, stat->ptc_l,
 				stat->change_rid, stat->shub_ptc_flushes, stat->nodes_flushed,
@@ -554,7 +554,7 @@ static int __init sn2_ptc_init(void)
 
 	proc_sn2_ptc = proc_create(PTC_BASENAME, 0444,
 				   NULL, &proc_sn2_ptc_operations);
-	if (!&proc_sn2_ptc_operations) {
+	if (!proc_sn2_ptc) {
 		printk(KERN_ERR "unable to create %s proc entry", PTC_BASENAME);
 		return -EINVAL;
 	}
diff --git a/arch/ia64/sn/kernel/sn2/sn_hwperf.c b/arch/ia64/sn/kernel/sn2/sn_hwperf.c
index be33947..9e6491c 100644
--- a/arch/ia64/sn/kernel/sn2/sn_hwperf.c
+++ b/arch/ia64/sn/kernel/sn2/sn_hwperf.c
@@ -275,8 +275,7 @@ static int sn_hwperf_get_nearest_node_objdata(struct sn_hwperf_object_info *objb
 
 	/* get it's interconnect topology */
 	sz = op->ports * sizeof(struct sn_hwperf_port_info);
-	if (sz > sizeof(ptdata))
-		BUG();
+	BUG_ON(sz > sizeof(ptdata));
 	e = ia64_sn_hwperf_op(sn_hwperf_master_nasid,
 			      SN_HWPERF_ENUM_PORTS, nodeobj->id, sz,
 			      (u64)&ptdata, 0, 0, NULL);
@@ -310,8 +309,7 @@ static int sn_hwperf_get_nearest_node_objdata(struct sn_hwperf_object_info *objb
 	if (router && (!found_cpu || !found_mem)) {
 		/* search for a node connected to the same router */
 		sz = router->ports * sizeof(struct sn_hwperf_port_info);
-		if (sz > sizeof(ptdata))
-			BUG();
+		BUG_ON(sz > sizeof(ptdata));
 		e = ia64_sn_hwperf_op(sn_hwperf_master_nasid,
 				      SN_HWPERF_ENUM_PORTS, router->id, sz,
 				      (u64)&ptdata, 0, 0, NULL);
@@ -612,7 +610,7 @@ static int sn_hwperf_op_cpu(struct sn_hwperf_op_info *op_info)
 	op_info->a->arg &= SN_HWPERF_ARG_OBJID_MASK;
 
 	if (cpu != SN_HWPERF_ARG_ANY_CPU) {
-		if (cpu >= NR_CPUS || !cpu_online(cpu)) {
+		if (cpu >= nr_cpu_ids || !cpu_online(cpu)) {
 			r = -EINVAL;
 			goto out;
 		}
diff --git a/arch/ia64/sn/pci/pcibr/pcibr_dma.c b/arch/ia64/sn/pci/pcibr/pcibr_dma.c
index 060df4a..c659ad5 100644
--- a/arch/ia64/sn/pci/pcibr/pcibr_dma.c
+++ b/arch/ia64/sn/pci/pcibr/pcibr_dma.c
@@ -256,9 +256,7 @@ void sn_dma_flush(u64 addr)
 
 	hubinfo = (NODEPDA(nasid_to_cnodeid(nasid)))->pdinfo;
 
-	if (!hubinfo) {
-		BUG();
-	}
+	BUG_ON(!hubinfo);
 
 	flush_nasid_list = &hubinfo->hdi_flush_nasid_list;
 	if (flush_nasid_list->widget_p == NULL)
diff --git a/arch/ia64/xen/Makefile b/arch/ia64/xen/Makefile
index 0ad0224..e6f4a0a 100644
--- a/arch/ia64/xen/Makefile
+++ b/arch/ia64/xen/Makefile
@@ -3,14 +3,29 @@
 #
 
 obj-y := hypercall.o xenivt.o xensetup.o xen_pv_ops.o irq_xen.o \
-	 hypervisor.o xencomm.o xcom_hcall.o grant-table.o time.o suspend.o
+	 hypervisor.o xencomm.o xcom_hcall.o grant-table.o time.o suspend.o \
+	 gate-data.o
 
 obj-$(CONFIG_IA64_GENERIC) += machvec.o
 
+# The gate DSO image is built using a special linker script.
+include $(srctree)/arch/ia64/kernel/Makefile.gate
+
+# tell compiled for xen
+CPPFLAGS_gate.lds += -D__IA64_GATE_PARAVIRTUALIZED_XEN
+AFLAGS_gate.o += -D__IA64_ASM_PARAVIRTUALIZED_XEN -D__IA64_GATE_PARAVIRTUALIZED_XEN
+
+# use same file of native.
+$(obj)/gate.o: $(src)/../kernel/gate.S FORCE
+	$(call if_changed_dep,as_o_S)
+$(obj)/gate.lds: $(src)/../kernel/gate.lds.S FORCE
+	$(call if_changed_dep,cpp_lds_S)
+
+
 AFLAGS_xenivt.o += -D__IA64_ASM_PARAVIRTUALIZED_XEN
 
 # xen multi compile
-ASM_PARAVIRT_MULTI_COMPILE_SRCS = ivt.S entry.S
+ASM_PARAVIRT_MULTI_COMPILE_SRCS = ivt.S entry.S fsys.S
 ASM_PARAVIRT_OBJS = $(addprefix xen-,$(ASM_PARAVIRT_MULTI_COMPILE_SRCS:.S=.o))
 obj-y += $(ASM_PARAVIRT_OBJS)
 define paravirtualized_xen
diff --git a/arch/ia64/xen/gate-data.S b/arch/ia64/xen/gate-data.S
new file mode 100644
index 0000000..7d4830a
--- /dev/null
+++ b/arch/ia64/xen/gate-data.S
@@ -0,0 +1,3 @@
+	.section .data.gate.xen, "aw"
+
+	.incbin "arch/ia64/xen/gate.so"
diff --git a/arch/ia64/xen/hypercall.S b/arch/ia64/xen/hypercall.S
index 45e02bb..e32dae4 100644
--- a/arch/ia64/xen/hypercall.S
+++ b/arch/ia64/xen/hypercall.S
@@ -9,6 +9,7 @@
 #include <asm/intrinsics.h>
 #include <asm/xen/privop.h>
 
+#ifdef __INTEL_COMPILER
 /*
  * Hypercalls without parameter.
  */
@@ -72,6 +73,7 @@ GLOBAL_ENTRY(xen_set_rr0_to_rr4)
 	br.ret.sptk.many rp
 	;;
 END(xen_set_rr0_to_rr4)
+#endif
 
 GLOBAL_ENTRY(xen_send_ipi)
 	mov r14=r32
diff --git a/arch/ia64/xen/time.c b/arch/ia64/xen/time.c
index 68d6204..fb83326 100644
--- a/arch/ia64/xen/time.c
+++ b/arch/ia64/xen/time.c
@@ -175,10 +175,58 @@ static void xen_itc_jitter_data_reset(void)
 	} while (unlikely(ret != lcycle));
 }
 
+/* based on xen_sched_clock() in arch/x86/xen/time.c. */
+/*
+ * This relies on HAVE_UNSTABLE_SCHED_CLOCK. If it can't be defined,
+ * something similar logic should be implemented here.
+ */
+/*
+ * Xen sched_clock implementation.  Returns the number of unstolen
+ * nanoseconds, which is nanoseconds the VCPU spent in RUNNING+BLOCKED
+ * states.
+ */
+static unsigned long long xen_sched_clock(void)
+{
+	struct vcpu_runstate_info runstate;
+
+	unsigned long long now;
+	unsigned long long offset;
+	unsigned long long ret;
+
+	/*
+	 * Ideally sched_clock should be called on a per-cpu basis
+	 * anyway, so preempt should already be disabled, but that's
+	 * not current practice at the moment.
+	 */
+	preempt_disable();
+
+	/*
+	 * both ia64_native_sched_clock() and xen's runstate are
+	 * based on mAR.ITC. So difference of them makes sense.
+	 */
+	now = ia64_native_sched_clock();
+
+	get_runstate_snapshot(&runstate);
+
+	WARN_ON(runstate.state != RUNSTATE_running);
+
+	offset = 0;
+	if (now > runstate.state_entry_time)
+		offset = now - runstate.state_entry_time;
+	ret = runstate.time[RUNSTATE_blocked] +
+		runstate.time[RUNSTATE_running] +
+		offset;
+
+	preempt_enable();
+
+	return ret;
+}
+
 struct pv_time_ops xen_time_ops __initdata = {
 	.init_missing_ticks_accounting	= xen_init_missing_ticks_accounting,
 	.do_steal_accounting		= xen_do_steal_accounting,
 	.clocksource_resume		= xen_itc_jitter_data_reset,
+	.sched_clock			= xen_sched_clock,
 };
 
 /* Called after suspend, to resume time.  */
diff --git a/arch/ia64/xen/xen_pv_ops.c b/arch/ia64/xen/xen_pv_ops.c
index 936cff3..5e2270a 100644
--- a/arch/ia64/xen/xen_pv_ops.c
+++ b/arch/ia64/xen/xen_pv_ops.c
@@ -24,6 +24,7 @@
 #include <linux/irq.h>
 #include <linux/kernel.h>
 #include <linux/pm.h>
+#include <linux/unistd.h>
 
 #include <asm/xen/hypervisor.h>
 #include <asm/xen/xencomm.h>
@@ -153,6 +154,13 @@ xen_post_smp_prepare_boot_cpu(void)
 	xen_setup_vcpu_info_placement();
 }
 
+#ifdef ASM_SUPPORTED
+static unsigned long __init_or_module
+xen_patch_bundle(void *sbundle, void *ebundle, unsigned long type);
+#endif
+static void __init
+xen_patch_branch(unsigned long tag, unsigned long type);
+
 static const struct pv_init_ops xen_init_ops __initconst = {
 	.banner = xen_banner,
 
@@ -163,6 +171,53 @@ static const struct pv_init_ops xen_init_ops __initconst = {
 	.arch_setup_nomca = xen_arch_setup_nomca,
 
 	.post_smp_prepare_boot_cpu = xen_post_smp_prepare_boot_cpu,
+#ifdef ASM_SUPPORTED
+	.patch_bundle = xen_patch_bundle,
+#endif
+	.patch_branch = xen_patch_branch,
+};
+
+/***************************************************************************
+ * pv_fsys_data
+ * addresses for fsys
+ */
+
+extern unsigned long xen_fsyscall_table[NR_syscalls];
+extern char xen_fsys_bubble_down[];
+struct pv_fsys_data xen_fsys_data __initdata = {
+	.fsyscall_table = (unsigned long *)xen_fsyscall_table,
+	.fsys_bubble_down = (void *)xen_fsys_bubble_down,
+};
+
+/***************************************************************************
+ * pv_patchdata
+ * patchdata addresses
+ */
+
+#define DECLARE(name)							\
+	extern unsigned long __xen_start_gate_##name##_patchlist[];	\
+	extern unsigned long __xen_end_gate_##name##_patchlist[]
+
+DECLARE(fsyscall);
+DECLARE(brl_fsys_bubble_down);
+DECLARE(vtop);
+DECLARE(mckinley_e9);
+
+extern unsigned long __xen_start_gate_section[];
+
+#define ASSIGN(name)							\
+	.start_##name##_patchlist =					\
+		(unsigned long)__xen_start_gate_##name##_patchlist,	\
+	.end_##name##_patchlist =					\
+		(unsigned long)__xen_end_gate_##name##_patchlist
+
+static struct pv_patchdata xen_patchdata __initdata = {
+	ASSIGN(fsyscall),
+	ASSIGN(brl_fsys_bubble_down),
+	ASSIGN(vtop),
+	ASSIGN(mckinley_e9),
+
+	.gate_section = (void*)__xen_start_gate_section,
 };
 
 /***************************************************************************
@@ -170,6 +225,76 @@ static const struct pv_init_ops xen_init_ops __initconst = {
  * intrinsics hooks.
  */
 
+#ifndef ASM_SUPPORTED
+static void
+xen_set_itm_with_offset(unsigned long val)
+{
+	/* ia64_cpu_local_tick() calls this with interrupt enabled. */
+	/* WARN_ON(!irqs_disabled()); */
+	xen_set_itm(val - XEN_MAPPEDREGS->itc_offset);
+}
+
+static unsigned long
+xen_get_itm_with_offset(void)
+{
+	/* unused at this moment */
+	printk(KERN_DEBUG "%s is called.\n", __func__);
+
+	WARN_ON(!irqs_disabled());
+	return ia64_native_getreg(_IA64_REG_CR_ITM) +
+		XEN_MAPPEDREGS->itc_offset;
+}
+
+/* ia64_set_itc() is only called by
+ * cpu_init() with ia64_set_itc(0) and ia64_sync_itc().
+ * So XEN_MAPPEDRESG->itc_offset cal be considered as almost constant.
+ */
+static void
+xen_set_itc(unsigned long val)
+{
+	unsigned long mitc;
+
+	WARN_ON(!irqs_disabled());
+	mitc = ia64_native_getreg(_IA64_REG_AR_ITC);
+	XEN_MAPPEDREGS->itc_offset = val - mitc;
+	XEN_MAPPEDREGS->itc_last = val;
+}
+
+static unsigned long
+xen_get_itc(void)
+{
+	unsigned long res;
+	unsigned long itc_offset;
+	unsigned long itc_last;
+	unsigned long ret_itc_last;
+
+	itc_offset = XEN_MAPPEDREGS->itc_offset;
+	do {
+		itc_last = XEN_MAPPEDREGS->itc_last;
+		res = ia64_native_getreg(_IA64_REG_AR_ITC);
+		res += itc_offset;
+		if (itc_last >= res)
+			res = itc_last + 1;
+		ret_itc_last = cmpxchg(&XEN_MAPPEDREGS->itc_last,
+				       itc_last, res);
+	} while (unlikely(ret_itc_last != itc_last));
+	return res;
+
+#if 0
+	/* ia64_itc_udelay() calls ia64_get_itc() with interrupt enabled.
+	   Should it be paravirtualized instead? */
+	WARN_ON(!irqs_disabled());
+	itc_offset = XEN_MAPPEDREGS->itc_offset;
+	itc_last = XEN_MAPPEDREGS->itc_last;
+	res = ia64_native_getreg(_IA64_REG_AR_ITC);
+	res += itc_offset;
+	if (itc_last >= res)
+		res = itc_last + 1;
+	XEN_MAPPEDREGS->itc_last = res;
+	return res;
+#endif
+}
+
 static void xen_setreg(int regnum, unsigned long val)
 {
 	switch (regnum) {
@@ -181,11 +306,14 @@ static void xen_setreg(int regnum, unsigned long val)
 		xen_set_eflag(val);
 		break;
 #endif
+	case _IA64_REG_AR_ITC:
+		xen_set_itc(val);
+		break;
 	case _IA64_REG_CR_TPR:
 		xen_set_tpr(val);
 		break;
 	case _IA64_REG_CR_ITM:
-		xen_set_itm(val);
+		xen_set_itm_with_offset(val);
 		break;
 	case _IA64_REG_CR_EOI:
 		xen_eoi(val);
@@ -209,6 +337,12 @@ static unsigned long xen_getreg(int regnum)
 		res = xen_get_eflag();
 		break;
 #endif
+	case _IA64_REG_AR_ITC:
+		res = xen_get_itc();
+		break;
+	case _IA64_REG_CR_ITM:
+		res = xen_get_itm_with_offset();
+		break;
 	case _IA64_REG_CR_IVR:
 		res = xen_get_ivr();
 		break;
@@ -259,8 +393,417 @@ xen_intrin_local_irq_restore(unsigned long mask)
 	else
 		xen_rsm_i();
 }
+#else
+#define __DEFINE_FUNC(name, code)					\
+	extern const char xen_ ## name ## _direct_start[];		\
+	extern const char xen_ ## name ## _direct_end[];		\
+	asm (".align 32\n"						\
+	     ".proc xen_" #name "\n"					\
+	     "xen_" #name ":\n"						\
+	     "xen_" #name "_direct_start:\n"				\
+	     code							\
+	     "xen_" #name "_direct_end:\n"				\
+	     "br.cond.sptk.many b6\n"					\
+	     ".endp xen_" #name "\n")
+
+#define DEFINE_VOID_FUNC0(name, code)		\
+	extern void				\
+	xen_ ## name (void);			\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_VOID_FUNC1(name, code)		\
+	extern void				\
+	xen_ ## name (unsigned long arg);	\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_VOID_FUNC1_VOID(name, code)	\
+	extern void				\
+	xen_ ## name (void *arg);		\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_VOID_FUNC2(name, code)		\
+	extern void				\
+	xen_ ## name (unsigned long arg0,	\
+		      unsigned long arg1);	\
+	__DEFINE_FUNC(name, code)
 
-static const struct pv_cpu_ops xen_cpu_ops __initdata = {
+#define DEFINE_FUNC0(name, code)		\
+	extern unsigned long			\
+	xen_ ## name (void);			\
+	__DEFINE_FUNC(name, code)
+
+#define DEFINE_FUNC1(name, type, code)		\
+	extern unsigned long			\
+	xen_ ## name (type arg);		\
+	__DEFINE_FUNC(name, code)
+
+#define XEN_PSR_I_ADDR_ADDR     (XSI_BASE + XSI_PSR_I_ADDR_OFS)
+
+/*
+ * static void xen_set_itm_with_offset(unsigned long val)
+ *        xen_set_itm(val - XEN_MAPPEDREGS->itc_offset);
+ */
+/* 2 bundles */
+DEFINE_VOID_FUNC1(set_itm_with_offset,
+		  "mov r2 = " __stringify(XSI_BASE) " + "
+		  __stringify(XSI_ITC_OFFSET_OFS) "\n"
+		  ";;\n"
+		  "ld8 r3 = [r2]\n"
+		  ";;\n"
+		  "sub r8 = r8, r3\n"
+		  "break " __stringify(HYPERPRIVOP_SET_ITM) "\n");
+
+/*
+ * static unsigned long xen_get_itm_with_offset(void)
+ *    return ia64_native_getreg(_IA64_REG_CR_ITM) + XEN_MAPPEDREGS->itc_offset;
+ */
+/* 2 bundles */
+DEFINE_FUNC0(get_itm_with_offset,
+	     "mov r2 = " __stringify(XSI_BASE) " + "
+	     __stringify(XSI_ITC_OFFSET_OFS) "\n"
+	     ";;\n"
+	     "ld8 r3 = [r2]\n"
+	     "mov r8 = cr.itm\n"
+	     ";;\n"
+	     "add r8 = r8, r2\n");
+
+/*
+ * static void xen_set_itc(unsigned long val)
+ *	unsigned long mitc;
+ *
+ *	WARN_ON(!irqs_disabled());
+ *	mitc = ia64_native_getreg(_IA64_REG_AR_ITC);
+ *	XEN_MAPPEDREGS->itc_offset = val - mitc;
+ *	XEN_MAPPEDREGS->itc_last = val;
+ */
+/* 2 bundles */
+DEFINE_VOID_FUNC1(set_itc,
+		  "mov r2 = " __stringify(XSI_BASE) " + "
+		  __stringify(XSI_ITC_LAST_OFS) "\n"
+		  "mov r3 = ar.itc\n"
+		  ";;\n"
+		  "sub r3 = r8, r3\n"
+		  "st8 [r2] = r8, "
+		  __stringify(XSI_ITC_LAST_OFS) " - "
+		  __stringify(XSI_ITC_OFFSET_OFS) "\n"
+		  ";;\n"
+		  "st8 [r2] = r3\n");
+
+/*
+ * static unsigned long xen_get_itc(void)
+ *	unsigned long res;
+ *	unsigned long itc_offset;
+ *	unsigned long itc_last;
+ *	unsigned long ret_itc_last;
+ *
+ *	itc_offset = XEN_MAPPEDREGS->itc_offset;
+ *	do {
+ *		itc_last = XEN_MAPPEDREGS->itc_last;
+ *		res = ia64_native_getreg(_IA64_REG_AR_ITC);
+ *		res += itc_offset;
+ *		if (itc_last >= res)
+ *			res = itc_last + 1;
+ *		ret_itc_last = cmpxchg(&XEN_MAPPEDREGS->itc_last,
+ *				       itc_last, res);
+ *	} while (unlikely(ret_itc_last != itc_last));
+ *	return res;
+ */
+/* 5 bundles */
+DEFINE_FUNC0(get_itc,
+	     "mov r2 = " __stringify(XSI_BASE) " + "
+	     __stringify(XSI_ITC_OFFSET_OFS) "\n"
+	     ";;\n"
+	     "ld8 r9 = [r2], " __stringify(XSI_ITC_LAST_OFS) " - "
+	     __stringify(XSI_ITC_OFFSET_OFS) "\n"
+					/* r9 = itc_offset */
+					/* r2 = XSI_ITC_OFFSET */
+	     "888:\n"
+	     "mov r8 = ar.itc\n"	/* res = ar.itc */
+	     ";;\n"
+	     "ld8 r3 = [r2]\n"		/* r3 = itc_last */
+	     "add r8 = r8, r9\n"	/* res = ar.itc + itc_offset */
+	     ";;\n"
+	     "cmp.gtu p6, p0 = r3, r8\n"
+	     ";;\n"
+	     "(p6) add r8 = 1, r3\n"	/* if (itc_last > res) itc_last + 1 */
+	     ";;\n"
+	     "mov ar.ccv = r8\n"
+	     ";;\n"
+	     "cmpxchg8.acq r10 = [r2], r8, ar.ccv\n"
+	     ";;\n"
+	     "cmp.ne p6, p0 = r10, r3\n"
+	     "(p6) hint @pause\n"
+	     "(p6) br.cond.spnt 888b\n");
+
+DEFINE_VOID_FUNC1_VOID(fc,
+		       "break " __stringify(HYPERPRIVOP_FC) "\n");
+
+/*
+ * psr_i_addr_addr = XEN_PSR_I_ADDR_ADDR
+ * masked_addr = *psr_i_addr_addr
+ * pending_intr_addr = masked_addr - 1
+ * if (val & IA64_PSR_I) {
+ *   masked = *masked_addr
+ *   *masked_addr = 0:xen_set_virtual_psr_i(1)
+ *   compiler barrier
+ *   if (masked) {
+ *      uint8_t pending = *pending_intr_addr;
+ *      if (pending)
+ *              XEN_HYPER_SSM_I
+ *   }
+ * } else {
+ *   *masked_addr = 1:xen_set_virtual_psr_i(0)
+ * }
+ */
+/* 6 bundles */
+DEFINE_VOID_FUNC1(intrin_local_irq_restore,
+		  /* r8 = input value: 0 or IA64_PSR_I
+		   * p6 =  (flags & IA64_PSR_I)
+		   *    = if clause
+		   * p7 = !(flags & IA64_PSR_I)
+		   *    = else clause
+		   */
+		  "cmp.ne p6, p7 = r8, r0\n"
+		  "mov r9 = " __stringify(XEN_PSR_I_ADDR_ADDR) "\n"
+		  ";;\n"
+		  /* r9 = XEN_PSR_I_ADDR */
+		  "ld8 r9 = [r9]\n"
+		  ";;\n"
+
+		  /* r10 = masked previous value */
+		  "(p6)	ld1.acq r10 = [r9]\n"
+		  ";;\n"
+
+		  /* p8 = !masked interrupt masked previously? */
+		  "(p6)	cmp.ne.unc p8, p0 = r10, r0\n"
+
+		  /* p7 = else clause */
+		  "(p7)	mov r11 = 1\n"
+		  ";;\n"
+		  /* masked = 1 */
+		  "(p7)	st1.rel [r9] = r11\n"
+
+		  /* p6 = if clause */
+		  /* masked = 0
+		   * r9 = masked_addr - 1
+		   *    = pending_intr_addr
+		   */
+		  "(p8)	st1.rel [r9] = r0, -1\n"
+		  ";;\n"
+		  /* r8 = pending_intr */
+		  "(p8)	ld1.acq r11 = [r9]\n"
+		  ";;\n"
+		  /* p9 = interrupt pending? */
+		  "(p8)	cmp.ne.unc p9, p10 = r11, r0\n"
+		  ";;\n"
+		  "(p10) mf\n"
+		  /* issue hypercall to trigger interrupt */
+		  "(p9)	break " __stringify(HYPERPRIVOP_SSM_I) "\n");
+
+DEFINE_VOID_FUNC2(ptcga,
+		  "break " __stringify(HYPERPRIVOP_PTC_GA) "\n");
+DEFINE_VOID_FUNC2(set_rr,
+		  "break " __stringify(HYPERPRIVOP_SET_RR) "\n");
+
+/*
+ * tmp = XEN_MAPPEDREGS->interrupt_mask_addr = XEN_PSR_I_ADDR_ADDR;
+ * tmp = *tmp
+ * tmp = *tmp;
+ * psr_i = tmp? 0: IA64_PSR_I;
+ */
+/* 4 bundles */
+DEFINE_FUNC0(get_psr_i,
+	     "mov r9 = " __stringify(XEN_PSR_I_ADDR_ADDR) "\n"
+	     ";;\n"
+	     "ld8 r9 = [r9]\n"			/* r9 = XEN_PSR_I_ADDR */
+	     "mov r8 = 0\n"			/* psr_i = 0 */
+	     ";;\n"
+	     "ld1.acq r9 = [r9]\n"		/* r9 = XEN_PSR_I */
+	     ";;\n"
+	     "cmp.eq.unc p6, p0 = r9, r0\n"	/* p6 = (XEN_PSR_I != 0) */
+	     ";;\n"
+	     "(p6) mov r8 = " __stringify(1 << IA64_PSR_I_BIT) "\n");
+
+DEFINE_FUNC1(thash, unsigned long,
+	     "break " __stringify(HYPERPRIVOP_THASH) "\n");
+DEFINE_FUNC1(get_cpuid, int,
+	     "break " __stringify(HYPERPRIVOP_GET_CPUID) "\n");
+DEFINE_FUNC1(get_pmd, int,
+	     "break " __stringify(HYPERPRIVOP_GET_PMD) "\n");
+DEFINE_FUNC1(get_rr, unsigned long,
+	     "break " __stringify(HYPERPRIVOP_GET_RR) "\n");
+
+/*
+ * void xen_privop_ssm_i(void)
+ *
+ * int masked = !xen_get_virtual_psr_i();
+ *	// masked = *(*XEN_MAPPEDREGS->interrupt_mask_addr)
+ * xen_set_virtual_psr_i(1)
+ *	// *(*XEN_MAPPEDREGS->interrupt_mask_addr) = 0
+ * // compiler barrier
+ * if (masked) {
+ *	uint8_t* pend_int_addr =
+ *		(uint8_t*)(*XEN_MAPPEDREGS->interrupt_mask_addr) - 1;
+ *	uint8_t pending = *pend_int_addr;
+ *	if (pending)
+ *		XEN_HYPER_SSM_I
+ * }
+ */
+/* 4 bundles */
+DEFINE_VOID_FUNC0(ssm_i,
+		  "mov r8 = " __stringify(XEN_PSR_I_ADDR_ADDR) "\n"
+		  ";;\n"
+		  "ld8 r8 = [r8]\n"		/* r8 = XEN_PSR_I_ADDR */
+		  ";;\n"
+		  "ld1.acq r9 = [r8]\n"		/* r9 = XEN_PSR_I */
+		  ";;\n"
+		  "st1.rel [r8] = r0, -1\n"	/* psr_i = 0. enable interrupt
+						 * r8 = XEN_PSR_I_ADDR - 1
+						 *    = pend_int_addr
+						 */
+		  "cmp.eq.unc p0, p6 = r9, r0\n"/* p6 = !XEN_PSR_I
+						 * previously interrupt
+						 * masked?
+						 */
+		  ";;\n"
+		  "(p6) ld1.acq r8 = [r8]\n"	/* r8 = xen_pend_int */
+		  ";;\n"
+		  "(p6) cmp.eq.unc p6, p7 = r8, r0\n"	/*interrupt pending?*/
+		  ";;\n"
+		  /* issue hypercall to get interrupt */
+		  "(p7) break " __stringify(HYPERPRIVOP_SSM_I) "\n"
+		  ";;\n");
+
+/*
+ * psr_i_addr_addr = XEN_MAPPEDREGS->interrupt_mask_addr
+ *		   = XEN_PSR_I_ADDR_ADDR;
+ * psr_i_addr = *psr_i_addr_addr;
+ * *psr_i_addr = 1;
+ */
+/* 2 bundles */
+DEFINE_VOID_FUNC0(rsm_i,
+		  "mov r8 = " __stringify(XEN_PSR_I_ADDR_ADDR) "\n"
+						/* r8 = XEN_PSR_I_ADDR */
+		  "mov r9 = 1\n"
+		  ";;\n"
+		  "ld8 r8 = [r8]\n"		/* r8 = XEN_PSR_I */
+		  ";;\n"
+		  "st1.rel [r8] = r9\n");	/* XEN_PSR_I = 1 */
+
+extern void
+xen_set_rr0_to_rr4(unsigned long val0, unsigned long val1,
+		   unsigned long val2, unsigned long val3,
+		   unsigned long val4);
+__DEFINE_FUNC(set_rr0_to_rr4,
+	      "break " __stringify(HYPERPRIVOP_SET_RR0_TO_RR4) "\n");
+
+
+extern unsigned long xen_getreg(int regnum);
+#define __DEFINE_GET_REG(id, privop)					\
+	"mov r2 = " __stringify(_IA64_REG_ ## id) "\n"			\
+	";;\n"								\
+	"cmp.eq p6, p0 = r2, r8\n"					\
+	";;\n"								\
+	"(p6) break " __stringify(HYPERPRIVOP_GET_ ## privop) "\n"	\
+	"(p6) br.cond.sptk.many b6\n"					\
+	";;\n"
+
+__DEFINE_FUNC(getreg,
+	      __DEFINE_GET_REG(PSR, PSR)
+#ifdef CONFIG_IA32_SUPPORT
+	      __DEFINE_GET_REG(AR_EFLAG, EFLAG)
+#endif
+
+	      /* get_itc */
+	      "mov r2 = " __stringify(_IA64_REG_AR_ITC) "\n"
+	      ";;\n"
+	      "cmp.eq p6, p0 = r2, r8\n"
+	      ";;\n"
+	      "(p6) br.cond.spnt xen_get_itc\n"
+	      ";;\n"
+
+	      /* get itm */
+	      "mov r2 = " __stringify(_IA64_REG_CR_ITM) "\n"
+	      ";;\n"
+	      "cmp.eq p6, p0 = r2, r8\n"
+	      ";;\n"
+	      "(p6) br.cond.spnt xen_get_itm_with_offset\n"
+	      ";;\n"
+
+	      __DEFINE_GET_REG(CR_IVR, IVR)
+	      __DEFINE_GET_REG(CR_TPR, TPR)
+
+	      /* fall back */
+	      "movl r2 = ia64_native_getreg_func\n"
+	      ";;\n"
+	      "mov b7 = r2\n"
+	      ";;\n"
+	      "br.cond.sptk.many b7\n");
+
+extern void xen_setreg(int regnum, unsigned long val);
+#define __DEFINE_SET_REG(id, privop)					\
+	"mov r2 = " __stringify(_IA64_REG_ ## id) "\n"			\
+	";;\n"								\
+	"cmp.eq p6, p0 = r2, r9\n"					\
+	";;\n"								\
+	"(p6) break " __stringify(HYPERPRIVOP_ ## privop) "\n"		\
+	"(p6) br.cond.sptk.many b6\n"					\
+	";;\n"
+
+__DEFINE_FUNC(setreg,
+	      /* kr0 .. kr 7*/
+	      /*
+	       * if (_IA64_REG_AR_KR0 <= regnum &&
+	       *     regnum <= _IA64_REG_AR_KR7) {
+	       *     register __index asm ("r8") = regnum - _IA64_REG_AR_KR0
+	       *     register __val asm ("r9") = val
+	       *    "break HYPERPRIVOP_SET_KR"
+	       * }
+	       */
+	      "mov r17 = r9\n"
+	      "mov r2 = " __stringify(_IA64_REG_AR_KR0) "\n"
+	      ";;\n"
+	      "cmp.ge p6, p0 = r9, r2\n"
+	      "sub r17 = r17, r2\n"
+	      ";;\n"
+	      "(p6) cmp.ge.unc p7, p0 = "
+	      __stringify(_IA64_REG_AR_KR7) " - " __stringify(_IA64_REG_AR_KR0)
+	      ", r17\n"
+	      ";;\n"
+	      "(p7) mov r9 = r8\n"
+	      ";;\n"
+	      "(p7) mov r8 = r17\n"
+	      "(p7) break " __stringify(HYPERPRIVOP_SET_KR) "\n"
+
+	      /* set itm */
+	      "mov r2 = " __stringify(_IA64_REG_CR_ITM) "\n"
+	      ";;\n"
+	      "cmp.eq p6, p0 = r2, r8\n"
+	      ";;\n"
+	      "(p6) br.cond.spnt xen_set_itm_with_offset\n"
+
+	      /* set itc */
+	      "mov r2 = " __stringify(_IA64_REG_AR_ITC) "\n"
+	      ";;\n"
+	      "cmp.eq p6, p0 = r2, r8\n"
+	      ";;\n"
+	      "(p6) br.cond.spnt xen_set_itc\n"
+
+#ifdef CONFIG_IA32_SUPPORT
+	      __DEFINE_SET_REG(AR_EFLAG, SET_EFLAG)
+#endif
+	      __DEFINE_SET_REG(CR_TPR, SET_TPR)
+	      __DEFINE_SET_REG(CR_EOI, EOI)
+
+	      /* fall back */
+	      "movl r2 = ia64_native_setreg_func\n"
+	      ";;\n"
+	      "mov b7 = r2\n"
+	      ";;\n"
+	      "br.cond.sptk.many b7\n");
+#endif
+
+static const struct pv_cpu_ops xen_cpu_ops __initconst = {
 	.fc		= xen_fc,
 	.thash		= xen_thash,
 	.get_cpuid	= xen_get_cpuid,
@@ -337,7 +880,7 @@ xen_iosapic_write(char __iomem *iosapic, unsigned int reg, u32 val)
 	HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op);
 }
 
-static const struct pv_iosapic_ops xen_iosapic_ops __initconst = {
+static struct pv_iosapic_ops xen_iosapic_ops __initdata = {
 	.pcat_compat_init = xen_pcat_compat_init,
 	.__get_irq_chip = xen_iosapic_get_irq_chip,
 
@@ -355,6 +898,8 @@ xen_setup_pv_ops(void)
 	xen_info_init();
 	pv_info = xen_info;
 	pv_init_ops = xen_init_ops;
+	pv_fsys_data = xen_fsys_data;
+	pv_patchdata = xen_patchdata;
 	pv_cpu_ops = xen_cpu_ops;
 	pv_iosapic_ops = xen_iosapic_ops;
 	pv_irq_ops = xen_irq_ops;
@@ -362,3 +907,252 @@ xen_setup_pv_ops(void)
 
 	paravirt_cpu_asm_init(&xen_cpu_asm_switch);
 }
+
+#ifdef ASM_SUPPORTED
+/***************************************************************************
+ * binary pacthing
+ * pv_init_ops.patch_bundle
+ */
+
+#define DEFINE_FUNC_GETREG(name, privop)				\
+	DEFINE_FUNC0(get_ ## name,					\
+		     "break "__stringify(HYPERPRIVOP_GET_ ## privop) "\n")
+
+DEFINE_FUNC_GETREG(psr, PSR);
+DEFINE_FUNC_GETREG(eflag, EFLAG);
+DEFINE_FUNC_GETREG(ivr, IVR);
+DEFINE_FUNC_GETREG(tpr, TPR);
+
+#define DEFINE_FUNC_SET_KR(n)						\
+	DEFINE_VOID_FUNC0(set_kr ## n,					\
+			  ";;\n"					\
+			  "mov r9 = r8\n"				\
+			  "mov r8 = " #n "\n"				\
+			  "break " __stringify(HYPERPRIVOP_SET_KR) "\n")
+
+DEFINE_FUNC_SET_KR(0);
+DEFINE_FUNC_SET_KR(1);
+DEFINE_FUNC_SET_KR(2);
+DEFINE_FUNC_SET_KR(3);
+DEFINE_FUNC_SET_KR(4);
+DEFINE_FUNC_SET_KR(5);
+DEFINE_FUNC_SET_KR(6);
+DEFINE_FUNC_SET_KR(7);
+
+#define __DEFINE_FUNC_SETREG(name, privop)				\
+	DEFINE_VOID_FUNC0(name,						\
+			  "break "__stringify(HYPERPRIVOP_ ## privop) "\n")
+
+#define DEFINE_FUNC_SETREG(name, privop)			\
+	__DEFINE_FUNC_SETREG(set_ ## name, SET_ ## privop)
+
+DEFINE_FUNC_SETREG(eflag, EFLAG);
+DEFINE_FUNC_SETREG(tpr, TPR);
+__DEFINE_FUNC_SETREG(eoi, EOI);
+
+extern const char xen_check_events[];
+extern const char __xen_intrin_local_irq_restore_direct_start[];
+extern const char __xen_intrin_local_irq_restore_direct_end[];
+extern const unsigned long __xen_intrin_local_irq_restore_direct_reloc;
+
+asm (
+	".align 32\n"
+	".proc xen_check_events\n"
+	"xen_check_events:\n"
+	/* masked = 0
+	 * r9 = masked_addr - 1
+	 *    = pending_intr_addr
+	 */
+	"st1.rel [r9] = r0, -1\n"
+	";;\n"
+	/* r8 = pending_intr */
+	"ld1.acq r11 = [r9]\n"
+	";;\n"
+	/* p9 = interrupt pending? */
+	"cmp.ne p9, p10 = r11, r0\n"
+	";;\n"
+	"(p10) mf\n"
+	/* issue hypercall to trigger interrupt */
+	"(p9) break " __stringify(HYPERPRIVOP_SSM_I) "\n"
+	"br.cond.sptk.many b6\n"
+	".endp xen_check_events\n"
+	"\n"
+	".align 32\n"
+	".proc __xen_intrin_local_irq_restore_direct\n"
+	"__xen_intrin_local_irq_restore_direct:\n"
+	"__xen_intrin_local_irq_restore_direct_start:\n"
+	"1:\n"
+	"{\n"
+	"cmp.ne p6, p7 = r8, r0\n"
+	"mov r17 = ip\n" /* get ip to calc return address */
+	"mov r9 = "__stringify(XEN_PSR_I_ADDR_ADDR) "\n"
+	";;\n"
+	"}\n"
+	"{\n"
+	/* r9 = XEN_PSR_I_ADDR */
+	"ld8 r9 = [r9]\n"
+	";;\n"
+	/* r10 = masked previous value */
+	"(p6) ld1.acq r10 = [r9]\n"
+	"adds r17 =  1f - 1b, r17\n" /* calculate return address */
+	";;\n"
+	"}\n"
+	"{\n"
+	/* p8 = !masked interrupt masked previously? */
+	"(p6) cmp.ne.unc p8, p0 = r10, r0\n"
+	"\n"
+	/* p7 = else clause */
+	"(p7) mov r11 = 1\n"
+	";;\n"
+	"(p8) mov b6 = r17\n" /* set return address */
+	"}\n"
+	"{\n"
+	/* masked = 1 */
+	"(p7) st1.rel [r9] = r11\n"
+	"\n"
+	"[99:]\n"
+	"(p8) brl.cond.dptk.few xen_check_events\n"
+	"}\n"
+	/* pv calling stub is 5 bundles. fill nop to adjust return address */
+	"{\n"
+	"nop 0\n"
+	"nop 0\n"
+	"nop 0\n"
+	"}\n"
+	"1:\n"
+	"__xen_intrin_local_irq_restore_direct_end:\n"
+	".endp __xen_intrin_local_irq_restore_direct\n"
+	"\n"
+	".align 8\n"
+	"__xen_intrin_local_irq_restore_direct_reloc:\n"
+	"data8 99b\n"
+);
+
+static struct paravirt_patch_bundle_elem xen_patch_bundle_elems[]
+__initdata_or_module =
+{
+#define XEN_PATCH_BUNDLE_ELEM(name, type)		\
+	{						\
+		(void*)xen_ ## name ## _direct_start,	\
+		(void*)xen_ ## name ## _direct_end,	\
+		PARAVIRT_PATCH_TYPE_ ## type,		\
+	}
+
+	XEN_PATCH_BUNDLE_ELEM(fc, FC),
+	XEN_PATCH_BUNDLE_ELEM(thash, THASH),
+	XEN_PATCH_BUNDLE_ELEM(get_cpuid, GET_CPUID),
+	XEN_PATCH_BUNDLE_ELEM(get_pmd, GET_PMD),
+	XEN_PATCH_BUNDLE_ELEM(ptcga, PTCGA),
+	XEN_PATCH_BUNDLE_ELEM(get_rr, GET_RR),
+	XEN_PATCH_BUNDLE_ELEM(set_rr, SET_RR),
+	XEN_PATCH_BUNDLE_ELEM(set_rr0_to_rr4, SET_RR0_TO_RR4),
+	XEN_PATCH_BUNDLE_ELEM(ssm_i, SSM_I),
+	XEN_PATCH_BUNDLE_ELEM(rsm_i, RSM_I),
+	XEN_PATCH_BUNDLE_ELEM(get_psr_i, GET_PSR_I),
+	{
+		(void*)__xen_intrin_local_irq_restore_direct_start,
+		(void*)__xen_intrin_local_irq_restore_direct_end,
+		PARAVIRT_PATCH_TYPE_INTRIN_LOCAL_IRQ_RESTORE,
+	},
+
+#define XEN_PATCH_BUNDLE_ELEM_GETREG(name, reg)			\
+	{							\
+		xen_get_ ## name ## _direct_start,		\
+		xen_get_ ## name ## _direct_end,		\
+		PARAVIRT_PATCH_TYPE_GETREG + _IA64_REG_ ## reg, \
+	}
+
+	XEN_PATCH_BUNDLE_ELEM_GETREG(psr, PSR),
+	XEN_PATCH_BUNDLE_ELEM_GETREG(eflag, AR_EFLAG),
+
+	XEN_PATCH_BUNDLE_ELEM_GETREG(ivr, CR_IVR),
+	XEN_PATCH_BUNDLE_ELEM_GETREG(tpr, CR_TPR),
+
+	XEN_PATCH_BUNDLE_ELEM_GETREG(itc, AR_ITC),
+	XEN_PATCH_BUNDLE_ELEM_GETREG(itm_with_offset, CR_ITM),
+
+
+#define __XEN_PATCH_BUNDLE_ELEM_SETREG(name, reg)		\
+	{							\
+		xen_ ## name ## _direct_start,			\
+		xen_ ## name ## _direct_end,			\
+		PARAVIRT_PATCH_TYPE_SETREG + _IA64_REG_ ## reg, \
+	}
+
+#define XEN_PATCH_BUNDLE_ELEM_SETREG(name, reg)			\
+	__XEN_PATCH_BUNDLE_ELEM_SETREG(set_ ## name, reg)
+
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr0, AR_KR0),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr1, AR_KR1),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr2, AR_KR2),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr3, AR_KR3),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr4, AR_KR4),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr5, AR_KR5),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr6, AR_KR6),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(kr7, AR_KR7),
+
+	XEN_PATCH_BUNDLE_ELEM_SETREG(eflag, AR_EFLAG),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(tpr, CR_TPR),
+	__XEN_PATCH_BUNDLE_ELEM_SETREG(eoi, CR_EOI),
+
+	XEN_PATCH_BUNDLE_ELEM_SETREG(itc, AR_ITC),
+	XEN_PATCH_BUNDLE_ELEM_SETREG(itm_with_offset, CR_ITM),
+};
+
+static unsigned long __init_or_module
+xen_patch_bundle(void *sbundle, void *ebundle, unsigned long type)
+{
+	const unsigned long nelems = sizeof(xen_patch_bundle_elems) /
+		sizeof(xen_patch_bundle_elems[0]);
+	unsigned long used;
+	const struct paravirt_patch_bundle_elem *found;
+
+	used = __paravirt_patch_apply_bundle(sbundle, ebundle, type,
+					     xen_patch_bundle_elems, nelems,
+					     &found);
+
+	if (found == NULL)
+		/* fallback */
+		return ia64_native_patch_bundle(sbundle, ebundle, type);
+	if (used == 0)
+		return used;
+
+	/* relocation */
+	switch (type) {
+	case PARAVIRT_PATCH_TYPE_INTRIN_LOCAL_IRQ_RESTORE: {
+		unsigned long reloc =
+			__xen_intrin_local_irq_restore_direct_reloc;
+		unsigned long reloc_offset = reloc - (unsigned long)
+			__xen_intrin_local_irq_restore_direct_start;
+		unsigned long tag = (unsigned long)sbundle + reloc_offset;
+		paravirt_patch_reloc_brl(tag, xen_check_events);
+		break;
+	}
+	default:
+		/* nothing */
+		break;
+	}
+	return used;
+}
+#endif /* ASM_SUPPOTED */
+
+const struct paravirt_patch_branch_target xen_branch_target[]
+__initconst = {
+#define PARAVIRT_BR_TARGET(name, type)			\
+	{						\
+		&xen_ ## name,				\
+		PARAVIRT_PATCH_TYPE_BR_ ## type,	\
+	}
+	PARAVIRT_BR_TARGET(switch_to, SWITCH_TO),
+	PARAVIRT_BR_TARGET(leave_syscall, LEAVE_SYSCALL),
+	PARAVIRT_BR_TARGET(work_processed_syscall, WORK_PROCESSED_SYSCALL),
+	PARAVIRT_BR_TARGET(leave_kernel, LEAVE_KERNEL),
+};
+
+static void __init
+xen_patch_branch(unsigned long tag, unsigned long type)
+{
+	const unsigned long nelem =
+		sizeof(xen_branch_target) / sizeof(xen_branch_target[0]);
+	__paravirt_patch_apply_branch(tag, type, xen_branch_target, nelem);
+}
diff --git a/arch/m32r/kernel/time.c b/arch/m32r/kernel/time.c
index 6ea0177..cada3ba 100644
--- a/arch/m32r/kernel/time.c
+++ b/arch/m32r/kernel/time.c
@@ -230,7 +230,6 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id)
 static struct irqaction irq0 = {
 	.handler = timer_interrupt,
 	.flags = IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "MFT2",
 };
 
diff --git a/arch/mips/cobalt/irq.c b/arch/mips/cobalt/irq.c
index ac4fb91..cb9bf82 100644
--- a/arch/mips/cobalt/irq.c
+++ b/arch/mips/cobalt/irq.c
@@ -47,7 +47,6 @@ asmlinkage void plat_irq_dispatch(void)
 
 static struct irqaction cascade = {
 	.handler	= no_action,
-	.mask		= CPU_MASK_NONE,
 	.name		= "cascade",
 };
 
diff --git a/arch/mips/emma/markeins/irq.c b/arch/mips/emma/markeins/irq.c
index 2bbc41a..43828ae 100644
--- a/arch/mips/emma/markeins/irq.c
+++ b/arch/mips/emma/markeins/irq.c
@@ -186,7 +186,6 @@ void emma2rh_gpio_irq_init(void)
 static struct irqaction irq_cascade = {
 	   .handler = no_action,
 	   .flags = 0,
-	   .mask = CPU_MASK_NONE,
 	   .name = "cascade",
 	   .dev_id = NULL,
 	   .next = NULL,
diff --git a/arch/mips/include/asm/mach-bcm47xx/gpio.h b/arch/mips/include/asm/mach-bcm47xx/gpio.h
index d8ff4cd..1784fde 100644
--- a/arch/mips/include/asm/mach-bcm47xx/gpio.h
+++ b/arch/mips/include/asm/mach-bcm47xx/gpio.h
@@ -31,24 +31,28 @@ static inline void gpio_set_value(unsigned gpio, int value)
 
 static inline int gpio_direction_input(unsigned gpio)
 {
-	return ssb_gpio_outen(&ssb_bcm47xx, 1 << gpio, 0);
+	ssb_gpio_outen(&ssb_bcm47xx, 1 << gpio, 0);
+	return 0;
 }
 
 static inline int gpio_direction_output(unsigned gpio, int value)
 {
-	return ssb_gpio_outen(&ssb_bcm47xx, 1 << gpio, 1 << gpio);
+	ssb_gpio_outen(&ssb_bcm47xx, 1 << gpio, 1 << gpio);
+	return 0;
 }
 
-static int gpio_intmask(unsigned gpio, int value)
+static inline int gpio_intmask(unsigned gpio, int value)
 {
-	return ssb_gpio_intmask(&ssb_bcm47xx, 1 << gpio,
-				value ? 1 << gpio : 0);
+	ssb_gpio_intmask(&ssb_bcm47xx, 1 << gpio,
+			 value ? 1 << gpio : 0);
+	return 0;
 }
 
-static int gpio_polarity(unsigned gpio, int value)
+static inline int gpio_polarity(unsigned gpio, int value)
 {
-	return ssb_gpio_polarity(&ssb_bcm47xx, 1 << gpio,
-				 value ? 1 << gpio : 0);
+	ssb_gpio_polarity(&ssb_bcm47xx, 1 << gpio,
+			  value ? 1 << gpio : 0);
+	return 0;
 }
 
 
diff --git a/arch/mips/include/asm/mach-ip27/topology.h b/arch/mips/include/asm/mach-ip27/topology.h
index 55d4815..0754723 100644
--- a/arch/mips/include/asm/mach-ip27/topology.h
+++ b/arch/mips/include/asm/mach-ip27/topology.h
@@ -26,7 +26,6 @@ extern struct cpuinfo_ip27 sn_cpu_info[NR_CPUS];
 #define parent_node(node)	(node)
 #define node_to_cpumask(node)	(hub_data(node)->h_cpus)
 #define cpumask_of_node(node)	(&hub_data(node)->h_cpus)
-#define node_to_first_cpu(node)	(cpumask_first(cpumask_of_node(node)))
 struct pci_bus;
 extern int pcibus_to_node(struct pci_bus *);
 
diff --git a/arch/mips/jazz/irq.c b/arch/mips/jazz/irq.c
index 03965cb..d9b6a5b 100644
--- a/arch/mips/jazz/irq.c
+++ b/arch/mips/jazz/irq.c
@@ -134,7 +134,6 @@ static irqreturn_t r4030_timer_interrupt(int irq, void *dev_id)
 static struct irqaction r4030_timer_irqaction = {
 	.handler	= r4030_timer_interrupt,
 	.flags		= IRQF_DISABLED,
-	.mask		= CPU_MASK_CPU0,
 	.name		= "R4030 timer",
 };
 
diff --git a/arch/mips/kernel/cevt-bcm1480.c b/arch/mips/kernel/cevt-bcm1480.c
index b820661..a5182a2 100644
--- a/arch/mips/kernel/cevt-bcm1480.c
+++ b/arch/mips/kernel/cevt-bcm1480.c
@@ -144,7 +144,6 @@ void __cpuinit sb1480_clockevent_init(void)
 
 	action->handler	= sibyte_counter_handler;
 	action->flags	= IRQF_DISABLED | IRQF_PERCPU;
-	action->mask	= cpumask_of_cpu(cpu);
 	action->name	= name;
 	action->dev_id	= cd;
 
diff --git a/arch/mips/kernel/cevt-sb1250.c b/arch/mips/kernel/cevt-sb1250.c
index a2eebaa..340f53e 100644
--- a/arch/mips/kernel/cevt-sb1250.c
+++ b/arch/mips/kernel/cevt-sb1250.c
@@ -143,7 +143,6 @@ void __cpuinit sb1250_clockevent_init(void)
 
 	action->handler	= sibyte_counter_handler;
 	action->flags	= IRQF_DISABLED | IRQF_PERCPU;
-	action->mask	= cpumask_of_cpu(cpu);
 	action->name	= name;
 	action->dev_id	= cd;
 
diff --git a/arch/mips/kernel/i8253.c b/arch/mips/kernel/i8253.c
index f4d1878..689719e 100644
--- a/arch/mips/kernel/i8253.c
+++ b/arch/mips/kernel/i8253.c
@@ -98,7 +98,6 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id)
 static struct irqaction irq0  = {
 	.handler = timer_interrupt,
 	.flags = IRQF_DISABLED | IRQF_NOBALANCING,
-	.mask = CPU_MASK_NONE,
 	.name = "timer"
 };
 
@@ -121,7 +120,6 @@ void __init setup_pit_timer(void)
 	cd->min_delta_ns = clockevent_delta2ns(0xF, cd);
 	clockevents_register_device(cd);
 
-	irq0.mask = cpumask_of_cpu(cpu);
 	setup_irq(0, &irq0);
 }
 
diff --git a/arch/mips/kernel/i8259.c b/arch/mips/kernel/i8259.c
index 413bd1d..01c0885 100644
--- a/arch/mips/kernel/i8259.c
+++ b/arch/mips/kernel/i8259.c
@@ -306,7 +306,6 @@ static void init_8259A(int auto_eoi)
  */
 static struct irqaction irq2 = {
 	.handler = no_action,
-	.mask = CPU_MASK_NONE,
 	.name = "cascade",
 };
 
diff --git a/arch/mips/lasat/interrupt.c b/arch/mips/lasat/interrupt.c
index d1ac7a2..1353fb1 100644
--- a/arch/mips/lasat/interrupt.c
+++ b/arch/mips/lasat/interrupt.c
@@ -104,7 +104,6 @@ asmlinkage void plat_irq_dispatch(void)
 
 static struct irqaction cascade = {
 	.handler	= no_action,
-	.mask		= CPU_MASK_NONE,
 	.name		= "cascade",
 };
 
diff --git a/arch/mips/lemote/lm2e/irq.c b/arch/mips/lemote/lm2e/irq.c
index 3e0b7be..1d0a09f 100644
--- a/arch/mips/lemote/lm2e/irq.c
+++ b/arch/mips/lemote/lm2e/irq.c
@@ -92,7 +92,6 @@ asmlinkage void plat_irq_dispatch(void)
 
 static struct irqaction cascade_irqaction = {
 	.handler = no_action,
-	.mask = CPU_MASK_NONE,
 	.name = "cascade",
 };
 
diff --git a/arch/mips/mm/highmem.c b/arch/mips/mm/highmem.c
index 060d28d..4481656 100644
--- a/arch/mips/mm/highmem.c
+++ b/arch/mips/mm/highmem.c
@@ -42,6 +42,7 @@ void *__kmap_atomic(struct page *page, enum km_type type)
 	if (!PageHighMem(page))
 		return page_address(page);
 
+	debug_kmap_atomic(type);
 	idx = type + KM_TYPE_NR*smp_processor_id();
 	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
 #ifdef CONFIG_DEBUG_HIGHMEM
@@ -88,6 +89,7 @@ void *kmap_atomic_pfn(unsigned long pfn, enum km_type type)
 
 	pagefault_disable();
 
+	debug_kmap_atomic(type);
 	idx = type + KM_TYPE_NR*smp_processor_id();
 	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
 	set_pte(kmap_pte-idx, pfn_pte(pfn, kmap_prot));
diff --git a/arch/mips/sgi-ip27/ip27-nmi.c b/arch/mips/sgi-ip27/ip27-nmi.c
index a1f21d9..6c5a630 100644
--- a/arch/mips/sgi-ip27/ip27-nmi.c
+++ b/arch/mips/sgi-ip27/ip27-nmi.c
@@ -219,7 +219,7 @@ cont_nmi_dump(void)
 		if (i == 1000) {
 			for_each_online_node(node)
 				if (NODEPDA(node)->dump_count == 0) {
-					cpu = node_to_first_cpu(node);
+					cpu = cpumask_first(cpumask_of_node(node));
 					for (n=0; n < CNODE_NUM_CPUS(node); cpu++, n++) {
 						CPUMASK_SETB(nmied_cpus, cpu);
 						/*
diff --git a/arch/mips/sgi-ip32/ip32-irq.c b/arch/mips/sgi-ip32/ip32-irq.c
index 0aefc53..83a0b3c 100644
--- a/arch/mips/sgi-ip32/ip32-irq.c
+++ b/arch/mips/sgi-ip32/ip32-irq.c
@@ -115,14 +115,12 @@ extern irqreturn_t crime_cpuerr_intr(int irq, void *dev_id);
 struct irqaction memerr_irq = {
 	.handler = crime_memerr_intr,
 	.flags = IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "CRIME memory error",
 };
 
 struct irqaction cpuerr_irq = {
 	.handler = crime_cpuerr_intr,
 	.flags = IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "CRIME CPU error",
 };
 
diff --git a/arch/mips/sni/rm200.c b/arch/mips/sni/rm200.c
index b4352a0..5e68781 100644
--- a/arch/mips/sni/rm200.c
+++ b/arch/mips/sni/rm200.c
@@ -359,7 +359,8 @@ void sni_rm200_init_8259A(void)
  * IRQ2 is cascade interrupt to second interrupt controller
  */
 static struct irqaction sni_rm200_irq2 = {
-	no_action, 0, CPU_MASK_NONE, "cascade", NULL, NULL
+	.handler = no_action,
+	.name = "cascade",
 };
 
 static struct resource sni_rm200_pic1_resource = {
diff --git a/arch/mips/vr41xx/common/irq.c b/arch/mips/vr41xx/common/irq.c
index 92dd1a0..9cc3891 100644
--- a/arch/mips/vr41xx/common/irq.c
+++ b/arch/mips/vr41xx/common/irq.c
@@ -32,7 +32,6 @@ static irq_cascade_t irq_cascade[NR_IRQS] __cacheline_aligned;
 
 static struct irqaction cascade_irqaction = {
 	.handler	= no_action,
-	.mask		= CPU_MASK_NONE,
 	.name		= "cascade",
 };
 
diff --git a/arch/mn10300/kernel/time.c b/arch/mn10300/kernel/time.c
index e460658..395caf0 100644
--- a/arch/mn10300/kernel/time.c
+++ b/arch/mn10300/kernel/time.c
@@ -37,7 +37,6 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id);
 static struct irqaction timer_irq = {
 	.handler	= timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_SHARED | IRQF_TIMER,
-	.mask		= CPU_MASK_NONE,
 	.name		= "timer",
 };
 
diff --git a/arch/parisc/kernel/time.c b/arch/parisc/kernel/time.c
index 9d46c43..e75cae6 100644
--- a/arch/parisc/kernel/time.c
+++ b/arch/parisc/kernel/time.c
@@ -216,17 +216,14 @@ void __init start_cpu_itimer(void)
 	per_cpu(cpu_data, cpu).it_value = next_tick;
 }
 
-struct platform_device rtc_parisc_dev = {
+static struct platform_device rtc_parisc_dev = {
 	.name = "rtc-parisc",
 	.id = -1,
 };
 
 static int __init rtc_init(void)
 {
-	int ret;
-
-	ret = platform_device_register(&rtc_parisc_dev);
-	if (ret < 0)
+	if (platform_device_register(&rtc_parisc_dev) < 0)
 		printk(KERN_ERR "unable to register rtc device...\n");
 
 	/* not necessarily an error */
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index ad6b1c0..45192dc 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -228,6 +228,9 @@ config PPC_OF_PLATFORM_PCI
 	depends on PPC64 # not supported on 32 bits yet
 	default n
 
+config ARCH_SUPPORTS_DEBUG_PAGEALLOC
+	def_bool y
+
 source "init/Kconfig"
 
 source "kernel/Kconfig.freezer"
diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug
index 22091bb..6aa0b5e 100644
--- a/arch/powerpc/Kconfig.debug
+++ b/arch/powerpc/Kconfig.debug
@@ -30,6 +30,7 @@ config DEBUG_STACK_USAGE
 config DEBUG_PAGEALLOC
         bool "Debug page memory allocations"
         depends on DEBUG_KERNEL && !HIBERNATION
+	depends on ARCH_SUPPORTS_DEBUG_PAGEALLOC
         help
           Unmap pages from the kernel linear mapping after free_pages().
           This results in a large slowdown, but helps to find certain types
diff --git a/arch/powerpc/boot/dts/mpc832x_rdb.dts b/arch/powerpc/boot/dts/mpc832x_rdb.dts
index dea3091..4319bd7 100644
--- a/arch/powerpc/boot/dts/mpc832x_rdb.dts
+++ b/arch/powerpc/boot/dts/mpc832x_rdb.dts
@@ -152,10 +152,21 @@
 		};
 
 		par_io@1400 {
+			#address-cells = <1>;
+			#size-cells = <1>;
 			reg = <0x1400 0x100>;
+			ranges = <3 0x1448 0x18>;
+			compatible = "fsl,mpc8323-qe-pario";
 			device_type = "par_io";
 			num-ports = <7>;
 
+			qe_pio_d: gpio-controller@1448 {
+				#gpio-cells = <2>;
+				compatible = "fsl,mpc8323-qe-pario-bank";
+				reg = <3 0x18>;
+				gpio-controller;
+			};
+
 			ucc2pio:ucc_pin@02 {
 				pio-map = <
 			/* port  pin  dir  open_drain  assignment  has_irq */
@@ -225,12 +236,25 @@
 		};
 
 		spi@4c0 {
+			#address-cells = <1>;
+			#size-cells = <0>;
 			cell-index = <0>;
 			compatible = "fsl,spi";
 			reg = <0x4c0 0x40>;
 			interrupts = <2>;
 			interrupt-parent = <&qeic>;
+			gpios = <&qe_pio_d 13 0>;
 			mode = "cpu-qe";
+
+			mmc-slot@0 {
+				compatible = "fsl,mpc8323rdb-mmc-slot",
+					     "mmc-spi-slot";
+				reg = <0>;
+				gpios = <&qe_pio_d 14 1
+					 &qe_pio_d 15 0>;
+				voltage-ranges = <3300 3300>;
+				spi-max-frequency = <50000000>;
+			};
 		};
 
 		spi@500 {
diff --git a/arch/powerpc/include/asm/highmem.h b/arch/powerpc/include/asm/highmem.h
index 545028f..684a73f 100644
--- a/arch/powerpc/include/asm/highmem.h
+++ b/arch/powerpc/include/asm/highmem.h
@@ -24,6 +24,7 @@
 
 #include <linux/init.h>
 #include <linux/interrupt.h>
+#include <linux/highmem.h>
 #include <asm/kmap_types.h>
 #include <asm/tlbflush.h>
 #include <asm/page.h>
@@ -94,6 +95,7 @@ static inline void *kmap_atomic_prot(struct page *page, enum km_type type, pgpro
 	if (!PageHighMem(page))
 		return page_address(page);
 
+	debug_kmap_atomic(type);
 	idx = type + KM_TYPE_NR*smp_processor_id();
 	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
 #ifdef CONFIG_DEBUG_HIGHMEM
diff --git a/arch/powerpc/include/asm/pci.h b/arch/powerpc/include/asm/pci.h
index 3548159..ba17d5d 100644
--- a/arch/powerpc/include/asm/pci.h
+++ b/arch/powerpc/include/asm/pci.h
@@ -114,6 +114,10 @@ extern int pci_domain_nr(struct pci_bus *bus);
 /* Decide whether to display the domain number in /proc */
 extern int pci_proc_domain(struct pci_bus *bus);
 
+/* MSI arch hooks */
+#define arch_setup_msi_irqs arch_setup_msi_irqs
+#define arch_teardown_msi_irqs arch_teardown_msi_irqs
+#define arch_msi_check_device arch_msi_check_device
 
 struct vm_area_struct;
 /* Map a range of PCI memory or I/O space for a device into user space */
diff --git a/arch/powerpc/include/asm/suspend.h b/arch/powerpc/include/asm/suspend.h
index cbf2c94..c6efc34 100644
--- a/arch/powerpc/include/asm/suspend.h
+++ b/arch/powerpc/include/asm/suspend.h
@@ -3,7 +3,4 @@
 
 static inline int arch_prepare_suspend(void) { return 0; }
 
-void save_processor_state(void);
-void restore_processor_state(void);
-
 #endif /* __ASM_POWERPC_SUSPEND_H */
diff --git a/arch/powerpc/include/asm/topology.h b/arch/powerpc/include/asm/topology.h
index 3752585..054a16d 100644
--- a/arch/powerpc/include/asm/topology.h
+++ b/arch/powerpc/include/asm/topology.h
@@ -24,11 +24,6 @@ static inline cpumask_t node_to_cpumask(int node)
 
 #define cpumask_of_node(node) (&numa_cpumask_lookup_table[node])
 
-static inline int node_to_first_cpu(int node)
-{
-	return cpumask_first(cpumask_of_node(node));
-}
-
 int of_node_to_nid(struct device_node *device);
 
 struct pci_bus;
diff --git a/arch/powerpc/kernel/msi.c b/arch/powerpc/kernel/msi.c
index 3bb7d3d..8bbc12d 100644
--- a/arch/powerpc/kernel/msi.c
+++ b/arch/powerpc/kernel/msi.c
@@ -9,6 +9,7 @@
 
 #include <linux/kernel.h>
 #include <linux/msi.h>
+#include <linux/pci.h>
 
 #include <asm/machdep.h>
 
@@ -19,6 +20,10 @@ int arch_msi_check_device(struct pci_dev* dev, int nvec, int type)
 		return -ENOSYS;
 	}
 
+	/* PowerPC doesn't support multiple MSI yet */
+	if (type == PCI_CAP_ID_MSI && nvec > 1)
+		return 1;
+
 	if (ppc_md.msi_check_device) {
 		pr_debug("msi: Using platform check routine.\n");
 		return ppc_md.msi_check_device(dev, nvec, type);
diff --git a/arch/powerpc/kernel/rtas_flash.c b/arch/powerpc/kernel/rtas_flash.c
index 149cb11..13011a9 100644
--- a/arch/powerpc/kernel/rtas_flash.c
+++ b/arch/powerpc/kernel/rtas_flash.c
@@ -669,7 +669,6 @@ static void remove_flash_pde(struct proc_dir_entry *dp)
 {
 	if (dp) {
 		kfree(dp->data);
-		dp->owner = NULL;
 		remove_proc_entry(dp->name, dp->parent);
 	}
 }
diff --git a/arch/powerpc/platforms/83xx/mpc832x_rdb.c b/arch/powerpc/platforms/83xx/mpc832x_rdb.c
index 2a1295f..567ded7 100644
--- a/arch/powerpc/platforms/83xx/mpc832x_rdb.c
+++ b/arch/powerpc/platforms/83xx/mpc832x_rdb.c
@@ -20,6 +20,7 @@
 #include <linux/spi/mmc_spi.h>
 #include <linux/mmc/host.h>
 #include <linux/of_platform.h>
+#include <linux/fsl_devices.h>
 
 #include <asm/time.h>
 #include <asm/ipic.h>
@@ -39,16 +40,116 @@
 #endif
 
 #ifdef CONFIG_QUICC_ENGINE
-static void mpc83xx_spi_activate_cs(u8 cs, u8 polarity)
+static int __init of_fsl_spi_probe(char *type, char *compatible, u32 sysclk,
+				   struct spi_board_info *board_infos,
+				   unsigned int num_board_infos,
+				   void (*cs_control)(struct spi_device *dev,
+						      bool on))
 {
-	pr_debug("%s %d %d\n", __func__, cs, polarity);
-	par_io_data_set(3, 13, polarity);
+	struct device_node *np;
+	unsigned int i = 0;
+
+	for_each_compatible_node(np, type, compatible) {
+		int ret;
+		unsigned int j;
+		const void *prop;
+		struct resource res[2];
+		struct platform_device *pdev;
+		struct fsl_spi_platform_data pdata = {
+			.cs_control = cs_control,
+		};
+
+		memset(res, 0, sizeof(res));
+
+		pdata.sysclk = sysclk;
+
+		prop = of_get_property(np, "reg", NULL);
+		if (!prop)
+			goto err;
+		pdata.bus_num = *(u32 *)prop;
+
+		prop = of_get_property(np, "cell-index", NULL);
+		if (prop)
+			i = *(u32 *)prop;
+
+		prop = of_get_property(np, "mode", NULL);
+		if (prop && !strcmp(prop, "cpu-qe"))
+			pdata.qe_mode = 1;
+
+		for (j = 0; j < num_board_infos; j++) {
+			if (board_infos[j].bus_num == pdata.bus_num)
+				pdata.max_chipselect++;
+		}
+
+		if (!pdata.max_chipselect)
+			continue;
+
+		ret = of_address_to_resource(np, 0, &res[0]);
+		if (ret)
+			goto err;
+
+		ret = of_irq_to_resource(np, 0, &res[1]);
+		if (ret == NO_IRQ)
+			goto err;
+
+		pdev = platform_device_alloc("mpc83xx_spi", i);
+		if (!pdev)
+			goto err;
+
+		ret = platform_device_add_data(pdev, &pdata, sizeof(pdata));
+		if (ret)
+			goto unreg;
+
+		ret = platform_device_add_resources(pdev, res,
+						    ARRAY_SIZE(res));
+		if (ret)
+			goto unreg;
+
+		ret = platform_device_add(pdev);
+		if (ret)
+			goto unreg;
+
+		goto next;
+unreg:
+		platform_device_del(pdev);
+err:
+		pr_err("%s: registration failed\n", np->full_name);
+next:
+		i++;
+	}
+
+	return i;
 }
 
-static void mpc83xx_spi_deactivate_cs(u8 cs, u8 polarity)
+static int __init fsl_spi_init(struct spi_board_info *board_infos,
+			       unsigned int num_board_infos,
+			       void (*cs_control)(struct spi_device *spi,
+						  bool on))
 {
-	pr_debug("%s %d %d\n", __func__, cs, polarity);
-	par_io_data_set(3, 13, !polarity);
+	u32 sysclk = -1;
+	int ret;
+
+	/* SPI controller is either clocked from QE or SoC clock */
+	sysclk = get_brgfreq();
+	if (sysclk == -1) {
+		sysclk = fsl_get_sys_freq();
+		if (sysclk == -1)
+			return -ENODEV;
+	}
+
+	ret = of_fsl_spi_probe(NULL, "fsl,spi", sysclk, board_infos,
+			       num_board_infos, cs_control);
+	if (!ret)
+		of_fsl_spi_probe("spi", "fsl_spi", sysclk, board_infos,
+				 num_board_infos, cs_control);
+
+	return spi_register_board_info(board_infos, num_board_infos);
+}
+
+static void mpc83xx_spi_cs_control(struct spi_device *spi, bool on)
+{
+	pr_debug("%s %d %d\n", __func__, spi->chip_select, on);
+	par_io_data_set(3, 13, on);
 }
 
 static struct mmc_spi_platform_data mpc832x_mmc_pdata = {
@@ -74,9 +175,13 @@ static int __init mpc832x_spi_init(void)
 	par_io_config_pin(3, 14, 2, 0, 0, 0); /* SD_INSERT, I */
 	par_io_config_pin(3, 15, 2, 0, 0, 0); /* SD_PROTECT,I */
 
-	return fsl_spi_init(&mpc832x_spi_boardinfo, 1,
-			    mpc83xx_spi_activate_cs,
-			    mpc83xx_spi_deactivate_cs);
+	/*
+	 * Don't bother with legacy stuff when device tree contains
+	 * mmc-spi-slot node.
+	 */
+	if (of_find_compatible_node(NULL, NULL, "mmc-spi-slot"))
+		return 0;
+	return fsl_spi_init(&mpc832x_spi_boardinfo, 1, mpc83xx_spi_cs_control);
 }
 machine_device_initcall(mpc832x_rdb, mpc832x_spi_init);
 #endif /* CONFIG_QUICC_ENGINE */
diff --git a/arch/powerpc/platforms/85xx/mpc85xx_cds.c b/arch/powerpc/platforms/85xx/mpc85xx_cds.c
index 0a9e491..458d91f 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_cds.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_cds.c
@@ -179,7 +179,6 @@ static irqreturn_t mpc85xx_8259_cascade_action(int irq, void *dev_id)
 static struct irqaction mpc85xxcds_8259_irqaction = {
 	.handler = mpc85xx_8259_cascade_action,
 	.flags = IRQF_SHARED,
-	.mask = CPU_MASK_NONE,
 	.name = "8259 cascade",
 };
 #endif /* PPC_I8259 */
diff --git a/arch/powerpc/platforms/8xx/m8xx_setup.c b/arch/powerpc/platforms/8xx/m8xx_setup.c
index 0d9f75c..385acfc 100644
--- a/arch/powerpc/platforms/8xx/m8xx_setup.c
+++ b/arch/powerpc/platforms/8xx/m8xx_setup.c
@@ -44,7 +44,6 @@ static irqreturn_t timebase_interrupt(int irq, void *dev)
 
 static struct irqaction tbint_irqaction = {
 	.handler = timebase_interrupt,
-	.mask = CPU_MASK_NONE,
 	.name = "tbint",
 };
 
diff --git a/arch/powerpc/platforms/chrp/setup.c b/arch/powerpc/platforms/chrp/setup.c
index 272d79a..cd4ad9a 100644
--- a/arch/powerpc/platforms/chrp/setup.c
+++ b/arch/powerpc/platforms/chrp/setup.c
@@ -472,7 +472,6 @@ static void __init chrp_find_openpic(void)
 #if defined(CONFIG_VT) && defined(CONFIG_INPUT_ADBHID) && defined(CONFIG_XMON)
 static struct irqaction xmon_irqaction = {
 	.handler = xmon_irq,
-	.mask = CPU_MASK_NONE,
 	.name = "XMON break",
 };
 #endif
diff --git a/arch/powerpc/platforms/powermac/pic.c b/arch/powerpc/platforms/powermac/pic.c
index 6d149ae..7039d8f 100644
--- a/arch/powerpc/platforms/powermac/pic.c
+++ b/arch/powerpc/platforms/powermac/pic.c
@@ -266,7 +266,6 @@ static unsigned int pmac_pic_get_irq(void)
 static struct irqaction xmon_action = {
 	.handler	= xmon_irq,
 	.flags		= 0,
-	.mask		= CPU_MASK_NONE,
 	.name		= "NMI - XMON"
 };
 #endif
@@ -274,7 +273,6 @@ static struct irqaction xmon_action = {
 static struct irqaction gatwick_cascade_action = {
 	.handler	= gatwick_action,
 	.flags		= IRQF_DISABLED,
-	.mask		= CPU_MASK_NONE,
 	.name		= "cascade",
 };
 
diff --git a/arch/powerpc/platforms/powermac/smp.c b/arch/powerpc/platforms/powermac/smp.c
index bd8817b..cf1dbe7 100644
--- a/arch/powerpc/platforms/powermac/smp.c
+++ b/arch/powerpc/platforms/powermac/smp.c
@@ -385,7 +385,6 @@ static void __init psurge_dual_sync_tb(int cpu_nr)
 static struct irqaction psurge_irqaction = {
 	.handler = psurge_primary_intr,
 	.flags = IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "primary IPI",
 };
 
diff --git a/arch/powerpc/sysdev/cpm1.c b/arch/powerpc/sysdev/cpm1.c
index 490473c..82424cd 100644
--- a/arch/powerpc/sysdev/cpm1.c
+++ b/arch/powerpc/sysdev/cpm1.c
@@ -119,7 +119,6 @@ static irqreturn_t cpm_error_interrupt(int irq, void *dev)
 
 static struct irqaction cpm_error_irqaction = {
 	.handler = cpm_error_interrupt,
-	.mask = CPU_MASK_NONE,
 	.name = "error",
 };
 
diff --git a/arch/powerpc/sysdev/fsl_soc.c b/arch/powerpc/sysdev/fsl_soc.c
index a01c89d..afe8dbc 100644
--- a/arch/powerpc/sysdev/fsl_soc.c
+++ b/arch/powerpc/sysdev/fsl_soc.c
@@ -417,115 +417,6 @@ err:
 
 arch_initcall(fsl_usb_of_init);
 
-static int __init of_fsl_spi_probe(char *type, char *compatible, u32 sysclk,
-				   struct spi_board_info *board_infos,
-				   unsigned int num_board_infos,
-				   void (*activate_cs)(u8 cs, u8 polarity),
-				   void (*deactivate_cs)(u8 cs, u8 polarity))
-{
-	struct device_node *np;
-	unsigned int i = 0;
-
-	for_each_compatible_node(np, type, compatible) {
-		int ret;
-		unsigned int j;
-		const void *prop;
-		struct resource res[2];
-		struct platform_device *pdev;
-		struct fsl_spi_platform_data pdata = {
-			.activate_cs = activate_cs,
-			.deactivate_cs = deactivate_cs,
-		};
-
-		memset(res, 0, sizeof(res));
-
-		pdata.sysclk = sysclk;
-
-		prop = of_get_property(np, "reg", NULL);
-		if (!prop)
-			goto err;
-		pdata.bus_num = *(u32 *)prop;
-
-		prop = of_get_property(np, "cell-index", NULL);
-		if (prop)
-			i = *(u32 *)prop;
-
-		prop = of_get_property(np, "mode", NULL);
-		if (prop && !strcmp(prop, "cpu-qe"))
-			pdata.qe_mode = 1;
-
-		for (j = 0; j < num_board_infos; j++) {
-			if (board_infos[j].bus_num == pdata.bus_num)
-				pdata.max_chipselect++;
-		}
-
-		if (!pdata.max_chipselect)
-			continue;
-
-		ret = of_address_to_resource(np, 0, &res[0]);
-		if (ret)
-			goto err;
-
-		ret = of_irq_to_resource(np, 0, &res[1]);
-		if (ret == NO_IRQ)
-			goto err;
-
-		pdev = platform_device_alloc("mpc83xx_spi", i);
-		if (!pdev)
-			goto err;
-
-		ret = platform_device_add_data(pdev, &pdata, sizeof(pdata));
-		if (ret)
-			goto unreg;
-
-		ret = platform_device_add_resources(pdev, res,
-						    ARRAY_SIZE(res));
-		if (ret)
-			goto unreg;
-
-		ret = platform_device_add(pdev);
-		if (ret)
-			goto unreg;
-
-		goto next;
-unreg:
-		platform_device_del(pdev);
-err:
-		pr_err("%s: registration failed\n", np->full_name);
-next:
-		i++;
-	}
-
-	return i;
-}
-
-int __init fsl_spi_init(struct spi_board_info *board_infos,
-			unsigned int num_board_infos,
-			void (*activate_cs)(u8 cs, u8 polarity),
-			void (*deactivate_cs)(u8 cs, u8 polarity))
-{
-	u32 sysclk = -1;
-	int ret;
-
-#ifdef CONFIG_QUICC_ENGINE
-	/* SPI controller is either clocked from QE or SoC clock */
-	sysclk = get_brgfreq();
-#endif
-	if (sysclk == -1) {
-		sysclk = fsl_get_sys_freq();
-		if (sysclk == -1)
-			return -ENODEV;
-	}
-
-	ret = of_fsl_spi_probe(NULL, "fsl,spi", sysclk, board_infos,
-			       num_board_infos, activate_cs, deactivate_cs);
-	if (!ret)
-		of_fsl_spi_probe("spi", "fsl_spi", sysclk, board_infos,
-				 num_board_infos, activate_cs, deactivate_cs);
-
-	return spi_register_board_info(board_infos, num_board_infos);
-}
-
 #if defined(CONFIG_PPC_85xx) || defined(CONFIG_PPC_86xx)
 static __be32 __iomem *rstcr;
 
diff --git a/arch/powerpc/sysdev/fsl_soc.h b/arch/powerpc/sysdev/fsl_soc.h
index 9c744e4..42381bb 100644
--- a/arch/powerpc/sysdev/fsl_soc.h
+++ b/arch/powerpc/sysdev/fsl_soc.h
@@ -4,6 +4,8 @@
 
 #include <asm/mmu.h>
 
+struct spi_device;
+
 extern phys_addr_t get_immrbase(void);
 #if defined(CONFIG_CPM2) || defined(CONFIG_QUICC_ENGINE) || defined(CONFIG_8xx)
 extern u32 get_brgfreq(void);
@@ -17,11 +19,6 @@ extern u32 fsl_get_sys_freq(void);
 struct spi_board_info;
 struct device_node;
 
-extern int fsl_spi_init(struct spi_board_info *board_infos,
-			unsigned int num_board_infos,
-			void (*activate_cs)(u8 cs, u8 polarity),
-			void (*deactivate_cs)(u8 cs, u8 polarity));
-
 extern void fsl_rstcr_restart(char *cmd);
 
 #if defined(CONFIG_FB_FSL_DIU) || defined(CONFIG_FB_FSL_DIU_MODULE)
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index 2a8af5e..dcb667c 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -72,6 +72,9 @@ config PGSTE
 config VIRT_CPU_ACCOUNTING
 	def_bool y
 
+config ARCH_SUPPORTS_DEBUG_PAGEALLOC
+	def_bool y
+
 mainmenu "Linux Kernel Configuration"
 
 config S390
diff --git a/arch/s390/Kconfig.debug b/arch/s390/Kconfig.debug
index 4599fa0..7e297a3 100644
--- a/arch/s390/Kconfig.debug
+++ b/arch/s390/Kconfig.debug
@@ -9,6 +9,7 @@ source "lib/Kconfig.debug"
 config DEBUG_PAGEALLOC
 	bool "Debug page memory allocations"
 	depends on DEBUG_KERNEL
+	depends on ARCH_SUPPORTS_DEBUG_PAGEALLOC
 	help
 	  Unmap pages from the kernel linear mapping after free_pages().
 	  This results in a slowdown, but helps to find certain types of
diff --git a/arch/s390/hypfs/hypfs_diag.c b/arch/s390/hypfs/hypfs_diag.c
index b1e892a..704dd39 100644
--- a/arch/s390/hypfs/hypfs_diag.c
+++ b/arch/s390/hypfs/hypfs_diag.c
@@ -12,6 +12,8 @@
 
 #include <linux/types.h>
 #include <linux/errno.h>
+#include <linux/gfp.h>
+#include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/vmalloc.h>
 #include <asm/ebcdic.h>
diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h
index 6dccb07..619bf94 100644
--- a/arch/s390/include/asm/cio.h
+++ b/arch/s390/include/asm/cio.h
@@ -456,6 +456,8 @@ struct ciw {
 #define CIO_OPER       0x0004
 /* Sick revalidation of device. */
 #define CIO_REVALIDATE 0x0008
+/* Device did not respond in time. */
+#define CIO_BOXED      0x0010
 
 /**
  * struct ccw_dev_id - unique identifier for ccw devices
diff --git a/arch/sh/include/asm/topology.h b/arch/sh/include/asm/topology.h
index 066f0fb..a3f2395 100644
--- a/arch/sh/include/asm/topology.h
+++ b/arch/sh/include/asm/topology.h
@@ -33,7 +33,6 @@
 
 #define node_to_cpumask(node)	((void)node, cpu_online_map)
 #define cpumask_of_node(node)	((void)node, cpu_online_mask)
-#define node_to_first_cpu(node)	((void)(node),0)
 
 #define pcibus_to_node(bus)	((void)(bus), -1)
 #define pcibus_to_cpumask(bus)	(pcibus_to_node(bus) == -1 ? \
diff --git a/arch/sh/kernel/time_64.c b/arch/sh/kernel/time_64.c
index 59d2a03..988c77c 100644
--- a/arch/sh/kernel/time_64.c
+++ b/arch/sh/kernel/time_64.c
@@ -284,7 +284,6 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id)
 static struct irqaction irq0  = {
 	.handler = timer_interrupt,
 	.flags = IRQF_DISABLED,
-	.mask = CPU_MASK_NONE,
 	.name = "timer",
 };
 
diff --git a/arch/sh/kernel/timers/timer-cmt.c b/arch/sh/kernel/timers/timer-cmt.c
index c127293..9aa3486 100644
--- a/arch/sh/kernel/timers/timer-cmt.c
+++ b/arch/sh/kernel/timers/timer-cmt.c
@@ -109,7 +109,6 @@ static struct irqaction cmt_irq = {
 	.name		= "timer",
 	.handler	= cmt_timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL,
-	.mask		= CPU_MASK_NONE,
 };
 
 static void cmt_clk_init(struct clk *clk)
diff --git a/arch/sh/kernel/timers/timer-mtu2.c b/arch/sh/kernel/timers/timer-mtu2.c
index 9a77ae8..9b0ef01 100644
--- a/arch/sh/kernel/timers/timer-mtu2.c
+++ b/arch/sh/kernel/timers/timer-mtu2.c
@@ -115,7 +115,6 @@ static struct irqaction mtu2_irq = {
 	.name		= "timer",
 	.handler	= mtu2_timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL,
-	.mask		= CPU_MASK_NONE,
 };
 
 static unsigned int divisors[] = { 1, 4, 16, 64, 1, 1, 256 };
diff --git a/arch/sh/kernel/timers/timer-tmu.c b/arch/sh/kernel/timers/timer-tmu.c
index 10b5a6f..c5d3396 100644
--- a/arch/sh/kernel/timers/timer-tmu.c
+++ b/arch/sh/kernel/timers/timer-tmu.c
@@ -162,7 +162,6 @@ static struct irqaction tmu0_irq = {
 	.name		= "periodic/oneshot timer",
 	.handler	= tmu_timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL,
-	.mask		= CPU_MASK_NONE,
 };
 
 static void __init tmu_clk_init(struct clk *clk)
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index c3ea215..cc12cd4 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -124,6 +124,9 @@ config ARCH_NO_VIRT_TO_BUS
 config OF
 	def_bool y
 
+config ARCH_SUPPORTS_DEBUG_PAGEALLOC
+	def_bool y if SPARC64
+
 source "init/Kconfig"
 
 source "kernel/Kconfig.freezer"
diff --git a/arch/sparc/Kconfig.debug b/arch/sparc/Kconfig.debug
index b8a15e2..d001b42 100644
--- a/arch/sparc/Kconfig.debug
+++ b/arch/sparc/Kconfig.debug
@@ -24,7 +24,8 @@ config STACK_DEBUG
 
 config DEBUG_PAGEALLOC
 	bool "Debug page memory allocations"
-	depends on SPARC64 && DEBUG_KERNEL && !HIBERNATION
+	depends on DEBUG_KERNEL && !HIBERNATION
+	depends on ARCH_SUPPORTS_DEBUG_PAGEALLOC
 	help
 	  Unmap pages from the kernel linear mapping after free_pages().
 	  This results in a large slowdown, but helps to find certain types
diff --git a/arch/sparc/include/asm/mmu_context_64.h b/arch/sparc/include/asm/mmu_context_64.h
index 5693ab4..666a73fe 100644
--- a/arch/sparc/include/asm/mmu_context_64.h
+++ b/arch/sparc/include/asm/mmu_context_64.h
@@ -121,8 +121,8 @@ static inline void switch_mm(struct mm_struct *old_mm, struct mm_struct *mm, str
 	 * local TLB.
 	 */
 	cpu = smp_processor_id();
-	if (!ctx_valid || !cpu_isset(cpu, mm->cpu_vm_mask)) {
-		cpu_set(cpu, mm->cpu_vm_mask);
+	if (!ctx_valid || !cpumask_test_cpu(cpu, mm_cpumask(mm))) {
+		cpumask_set_cpu(cpu, mm_cpumask(mm));
 		__flush_tlb_mm(CTX_HWBITS(mm->context),
 			       SECONDARY_CONTEXT);
 	}
@@ -141,8 +141,8 @@ static inline void activate_mm(struct mm_struct *active_mm, struct mm_struct *mm
 	if (!CTX_VALID(mm->context))
 		get_new_mmu_context(mm);
 	cpu = smp_processor_id();
-	if (!cpu_isset(cpu, mm->cpu_vm_mask))
-		cpu_set(cpu, mm->cpu_vm_mask);
+	if (!cpumask_test_cpu(cpu, mm_cpumask(mm)))
+		cpumask_set_cpu(cpu, mm_cpumask(mm));
 
 	load_secondary_context(mm);
 	__flush_tlb_mm(CTX_HWBITS(mm->context), SECONDARY_CONTEXT);
diff --git a/arch/sparc/include/asm/smp_64.h b/arch/sparc/include/asm/smp_64.h
index 57224dd..becb6bf 100644
--- a/arch/sparc/include/asm/smp_64.h
+++ b/arch/sparc/include/asm/smp_64.h
@@ -35,7 +35,8 @@ extern cpumask_t cpu_core_map[NR_CPUS];
 extern int sparc64_multi_core;
 
 extern void arch_send_call_function_single_ipi(int cpu);
-extern void arch_send_call_function_ipi(cpumask_t mask);
+extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
+#define arch_send_call_function_ipi_mask arch_send_call_function_ipi_mask
 
 /*
  *	General functions that each host system must provide.
diff --git a/arch/sparc/include/asm/system_32.h b/arch/sparc/include/asm/system_32.h
index 79c1ae2..751c8c1 100644
--- a/arch/sparc/include/asm/system_32.h
+++ b/arch/sparc/include/asm/system_32.h
@@ -126,7 +126,7 @@ extern void flushw_all(void);
 #define switch_to(prev, next, last) do {						\
 	SWITCH_ENTER(prev);								\
 	SWITCH_DO_LAZY_FPU(next);							\
-	cpu_set(smp_processor_id(), next->active_mm->cpu_vm_mask);			\
+	cpumask_set_cpu(smp_processor_id(), mm_cpumask(next->active_mm));		\
 	__asm__ __volatile__(								\
 	"sethi	%%hi(here - 0x8), %%o7\n\t"						\
 	"mov	%%g6, %%g3\n\t"								\
diff --git a/arch/sparc/include/asm/topology_64.h b/arch/sparc/include/asm/topology_64.h
index 5bc0b8f..e5ea8d3 100644
--- a/arch/sparc/include/asm/topology_64.h
+++ b/arch/sparc/include/asm/topology_64.h
@@ -28,11 +28,6 @@ static inline cpumask_t node_to_cpumask(int node)
 #define node_to_cpumask_ptr_next(v, node)	\
 			   v = &(numa_cpumask_lookup_table[node])
 
-static inline int node_to_first_cpu(int node)
-{
-	return cpumask_first(cpumask_of_node(node));
-}
-
 struct pci_bus;
 #ifdef CONFIG_PCI
 extern int pcibus_to_node(struct pci_bus *pbus);
@@ -43,13 +38,9 @@ static inline int pcibus_to_node(struct pci_bus *pbus)
 }
 #endif
 
-#define pcibus_to_cpumask(bus)	\
-	(pcibus_to_node(bus) == -1 ? \
-	 CPU_MASK_ALL : \
-	 node_to_cpumask(pcibus_to_node(bus)))
 #define cpumask_of_pcibus(bus)	\
 	(pcibus_to_node(bus) == -1 ? \
-	 CPU_MASK_ALL_PTR : \
+	 cpu_all_mask : \
 	 cpumask_of_node(pcibus_to_node(bus)))
 
 #define SD_NODE_INIT (struct sched_domain) {		\
@@ -89,7 +80,6 @@ static inline int pcibus_to_node(struct pci_bus *pbus)
 #define smt_capable()				(sparc64_multi_core)
 #endif /* CONFIG_SMP */
 
-#define cpu_coregroup_map(cpu)			(cpu_core_map[cpu])
 #define cpu_coregroup_mask(cpu)			(&cpu_core_map[cpu])
 
 #endif /* _ASM_SPARC64_TOPOLOGY_H */
diff --git a/arch/sparc/kernel/ds.c b/arch/sparc/kernel/ds.c
index 57c3984..90350f83 100644
--- a/arch/sparc/kernel/ds.c
+++ b/arch/sparc/kernel/ds.c
@@ -653,7 +653,7 @@ static void __cpuinit dr_cpu_data(struct ds_info *dp,
 		if (cpu_list[i] == CPU_SENTINEL)
 			continue;
 
-		if (cpu_list[i] < NR_CPUS)
+		if (cpu_list[i] < nr_cpu_ids)
 			cpu_set(cpu_list[i], mask);
 	}
 
diff --git a/arch/sparc/kernel/irq_32.c b/arch/sparc/kernel/irq_32.c
index 44dd5ee..ad800b8 100644
--- a/arch/sparc/kernel/irq_32.c
+++ b/arch/sparc/kernel/irq_32.c
@@ -439,7 +439,6 @@ static int request_fast_irq(unsigned int irq,
 	flush_cache_all();
 
 	action->flags = irqflags;
-	cpus_clear(action->mask);
 	action->name = devname;
 	action->dev_id = NULL;
 	action->next = NULL;
@@ -574,7 +573,6 @@ int request_irq(unsigned int irq,
 
 	action->handler = handler;
 	action->flags = irqflags;
-	cpus_clear(action->mask);
 	action->name = devname;
 	action->next = NULL;
 	action->dev_id = dev_id;
diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c
index d0d6a51..5deabe9 100644
--- a/arch/sparc/kernel/irq_64.c
+++ b/arch/sparc/kernel/irq_64.c
@@ -266,12 +266,12 @@ static int irq_choose_cpu(unsigned int virt_irq)
 		spin_lock_irqsave(&irq_rover_lock, flags);
 
 		while (!cpu_online(irq_rover)) {
-			if (++irq_rover >= NR_CPUS)
+			if (++irq_rover >= nr_cpu_ids)
 				irq_rover = 0;
 		}
 		cpuid = irq_rover;
 		do {
-			if (++irq_rover >= NR_CPUS)
+			if (++irq_rover >= nr_cpu_ids)
 				irq_rover = 0;
 		} while (!cpu_online(irq_rover));
 
diff --git a/arch/sparc/kernel/led.c b/arch/sparc/kernel/led.c
index adaaed4..00d034e 100644
--- a/arch/sparc/kernel/led.c
+++ b/arch/sparc/kernel/led.c
@@ -126,7 +126,6 @@ static int __init led_init(void)
 	led = proc_create("led", 0, NULL, &led_proc_fops);
 	if (!led)
 		return -ENOMEM;
-	led->owner = THIS_MODULE;
 
 	printk(KERN_INFO
 	       "led: version %s, Lars Kotthoff <metalhead@metalhead.ws>\n",
diff --git a/arch/sparc/kernel/mdesc.c b/arch/sparc/kernel/mdesc.c
index 3f79f0c..f0e6ed2 100644
--- a/arch/sparc/kernel/mdesc.c
+++ b/arch/sparc/kernel/mdesc.c
@@ -567,7 +567,7 @@ static void __init report_platform_properties(void)
 			max_cpu = NR_CPUS;
 		}
 		for (i = 0; i < max_cpu; i++)
-			cpu_set(i, cpu_possible_map);
+			set_cpu_possible(i, true);
 	}
 #endif
 
diff --git a/arch/sparc/kernel/nmi.c b/arch/sparc/kernel/nmi.c
index f357722..2c0cc72 100644
--- a/arch/sparc/kernel/nmi.c
+++ b/arch/sparc/kernel/nmi.c
@@ -13,6 +13,7 @@
 #include <linux/module.h>
 #include <linux/kprobes.h>
 #include <linux/kernel_stat.h>
+#include <linux/reboot.h>
 #include <linux/slab.h>
 #include <linux/kdebug.h>
 #include <linux/delay.h>
@@ -206,13 +207,33 @@ void nmi_adjust_hz(unsigned int new_hz)
 }
 EXPORT_SYMBOL_GPL(nmi_adjust_hz);
 
+static int nmi_shutdown(struct notifier_block *nb, unsigned long cmd, void *p)
+{
+	on_each_cpu(stop_watchdog, NULL, 1);
+	return 0;
+}
+
+static struct notifier_block nmi_reboot_notifier = {
+	.notifier_call = nmi_shutdown,
+};
+
 int __init nmi_init(void)
 {
+	int err;
+
 	nmi_usable = 1;
 
 	on_each_cpu(start_watchdog, NULL, 1);
 
-	return check_nmi_watchdog();
+	err = check_nmi_watchdog();
+	if (!err) {
+		err = register_reboot_notifier(&nmi_reboot_notifier);
+		if (err) {
+			nmi_usable = 0;
+			on_each_cpu(stop_watchdog, NULL, 1);
+		}
+	}
+	return err;
 }
 
 static int __init setup_nmi_watchdog(char *str)
diff --git a/arch/sparc/kernel/prom_64.c b/arch/sparc/kernel/prom_64.c
index edecca7..ca55c70 100644
--- a/arch/sparc/kernel/prom_64.c
+++ b/arch/sparc/kernel/prom_64.c
@@ -518,8 +518,8 @@ void __init of_fill_in_cpu_data(void)
 		}
 
 #ifdef CONFIG_SMP
-		cpu_set(cpuid, cpu_present_map);
-		cpu_set(cpuid, cpu_possible_map);
+		set_cpu_present(cpuid, true);
+		set_cpu_possible(cpuid, true);
 #endif
 	}
 
diff --git a/arch/sparc/kernel/smp_32.c b/arch/sparc/kernel/smp_32.c
index 1e5ac4e2..132d81f 100644
--- a/arch/sparc/kernel/smp_32.c
+++ b/arch/sparc/kernel/smp_32.c
@@ -70,13 +70,12 @@ void __init smp_cpus_done(unsigned int max_cpus)
 	extern void smp4m_smp_done(void);
 	extern void smp4d_smp_done(void);
 	unsigned long bogosum = 0;
-	int cpu, num;
+	int cpu, num = 0;
 
-	for (cpu = 0, num = 0; cpu < NR_CPUS; cpu++)
-		if (cpu_online(cpu)) {
-			num++;
-			bogosum += cpu_data(cpu).udelay_val;
-		}
+	for_each_online_cpu(cpu) {
+		num++;
+		bogosum += cpu_data(cpu).udelay_val;
+	}
 
 	printk("Total of %d processors activated (%lu.%02lu BogoMIPS).\n",
 		num, bogosum/(500000/HZ),
@@ -144,7 +143,7 @@ void smp_flush_tlb_all(void)
 void smp_flush_cache_mm(struct mm_struct *mm)
 {
 	if(mm->context != NO_CONTEXT) {
-		cpumask_t cpu_mask = mm->cpu_vm_mask;
+		cpumask_t cpu_mask = *mm_cpumask(mm);
 		cpu_clear(smp_processor_id(), cpu_mask);
 		if (!cpus_empty(cpu_mask))
 			xc1((smpfunc_t) BTFIXUP_CALL(local_flush_cache_mm), (unsigned long) mm);
@@ -155,12 +154,13 @@ void smp_flush_cache_mm(struct mm_struct *mm)
 void smp_flush_tlb_mm(struct mm_struct *mm)
 {
 	if(mm->context != NO_CONTEXT) {
-		cpumask_t cpu_mask = mm->cpu_vm_mask;
+		cpumask_t cpu_mask = *mm_cpumask(mm);
 		cpu_clear(smp_processor_id(), cpu_mask);
 		if (!cpus_empty(cpu_mask)) {
 			xc1((smpfunc_t) BTFIXUP_CALL(local_flush_tlb_mm), (unsigned long) mm);
 			if(atomic_read(&mm->mm_users) == 1 && current->active_mm == mm)
-				mm->cpu_vm_mask = cpumask_of_cpu(smp_processor_id());
+				cpumask_copy(mm_cpumask(mm),
+					     cpumask_of(smp_processor_id()));
 		}
 		local_flush_tlb_mm(mm);
 	}
@@ -172,7 +172,7 @@ void smp_flush_cache_range(struct vm_area_struct *vma, unsigned long start,
 	struct mm_struct *mm = vma->vm_mm;
 
 	if (mm->context != NO_CONTEXT) {
-		cpumask_t cpu_mask = mm->cpu_vm_mask;
+		cpumask_t cpu_mask = *mm_cpumask(mm);
 		cpu_clear(smp_processor_id(), cpu_mask);
 		if (!cpus_empty(cpu_mask))
 			xc3((smpfunc_t) BTFIXUP_CALL(local_flush_cache_range), (unsigned long) vma, start, end);
@@ -186,7 +186,7 @@ void smp_flush_tlb_range(struct vm_area_struct *vma, unsigned long start,
 	struct mm_struct *mm = vma->vm_mm;
 
 	if (mm->context != NO_CONTEXT) {
-		cpumask_t cpu_mask = mm->cpu_vm_mask;
+		cpumask_t cpu_mask = *mm_cpumask(mm);
 		cpu_clear(smp_processor_id(), cpu_mask);
 		if (!cpus_empty(cpu_mask))
 			xc3((smpfunc_t) BTFIXUP_CALL(local_flush_tlb_range), (unsigned long) vma, start, end);
@@ -199,7 +199,7 @@ void smp_flush_cache_page(struct vm_area_struct *vma, unsigned long page)
 	struct mm_struct *mm = vma->vm_mm;
 
 	if(mm->context != NO_CONTEXT) {
-		cpumask_t cpu_mask = mm->cpu_vm_mask;
+		cpumask_t cpu_mask = *mm_cpumask(mm);
 		cpu_clear(smp_processor_id(), cpu_mask);
 		if (!cpus_empty(cpu_mask))
 			xc2((smpfunc_t) BTFIXUP_CALL(local_flush_cache_page), (unsigned long) vma, page);
@@ -212,7 +212,7 @@ void smp_flush_tlb_page(struct vm_area_struct *vma, unsigned long page)
 	struct mm_struct *mm = vma->vm_mm;
 
 	if(mm->context != NO_CONTEXT) {
-		cpumask_t cpu_mask = mm->cpu_vm_mask;
+		cpumask_t cpu_mask = *mm_cpumask(mm);
 		cpu_clear(smp_processor_id(), cpu_mask);
 		if (!cpus_empty(cpu_mask))
 			xc2((smpfunc_t) BTFIXUP_CALL(local_flush_tlb_page), (unsigned long) vma, page);
@@ -241,7 +241,7 @@ void smp_flush_page_to_ram(unsigned long page)
 
 void smp_flush_sig_insns(struct mm_struct *mm, unsigned long insn_addr)
 {
-	cpumask_t cpu_mask = mm->cpu_vm_mask;
+	cpumask_t cpu_mask = *mm_cpumask(mm);
 	cpu_clear(smp_processor_id(), cpu_mask);
 	if (!cpus_empty(cpu_mask))
 		xc2((smpfunc_t) BTFIXUP_CALL(local_flush_sig_insns), (unsigned long) mm, insn_addr);
@@ -332,8 +332,8 @@ void __init smp_setup_cpu_possible_map(void)
 	instance = 0;
 	while (!cpu_find_by_instance(instance, NULL, &mid)) {
 		if (mid < NR_CPUS) {
-			cpu_set(mid, cpu_possible_map);
-			cpu_set(mid, cpu_present_map);
+			set_cpu_possible(mid, true);
+			set_cpu_present(mid, true);
 		}
 		instance++;
 	}
@@ -351,8 +351,8 @@ void __init smp_prepare_boot_cpu(void)
 		printk("boot cpu id != 0, this could work but is untested\n");
 
 	current_thread_info()->cpu = cpuid;
-	cpu_set(cpuid, cpu_online_map);
-	cpu_set(cpuid, cpu_possible_map);
+	set_cpu_online(cpuid, true);
+	set_cpu_possible(cpuid, true);
 }
 
 int __cpuinit __cpu_up(unsigned int cpu)
diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c
index 79457f6..708e12a 100644
--- a/arch/sparc/kernel/smp_64.c
+++ b/arch/sparc/kernel/smp_64.c
@@ -808,9 +808,9 @@ static void smp_start_sync_tick_client(int cpu)
 
 extern unsigned long xcall_call_function;
 
-void arch_send_call_function_ipi(cpumask_t mask)
+void arch_send_call_function_ipi_mask(const struct cpumask *mask)
 {
-	xcall_deliver((u64) &xcall_call_function, 0, 0, &mask);
+	xcall_deliver((u64) &xcall_call_function, 0, 0, mask);
 }
 
 extern unsigned long xcall_call_function_single;
@@ -850,7 +850,7 @@ static void tsb_sync(void *info)
 
 void smp_tsb_sync(struct mm_struct *mm)
 {
-	smp_call_function_mask(mm->cpu_vm_mask, tsb_sync, mm, 1);
+	smp_call_function_many(mm_cpumask(mm), tsb_sync, mm, 1);
 }
 
 extern unsigned long xcall_flush_tlb_mm;
@@ -1055,13 +1055,13 @@ void smp_flush_tlb_mm(struct mm_struct *mm)
 	int cpu = get_cpu();
 
 	if (atomic_read(&mm->mm_users) == 1) {
-		mm->cpu_vm_mask = cpumask_of_cpu(cpu);
+		cpumask_copy(mm_cpumask(mm), cpumask_of(cpu));
 		goto local_flush_and_out;
 	}
 
 	smp_cross_call_masked(&xcall_flush_tlb_mm,
 			      ctx, 0, 0,
-			      &mm->cpu_vm_mask);
+			      mm_cpumask(mm));
 
 local_flush_and_out:
 	__flush_tlb_mm(ctx, SECONDARY_CONTEXT);
@@ -1075,11 +1075,11 @@ void smp_flush_tlb_pending(struct mm_struct *mm, unsigned long nr, unsigned long
 	int cpu = get_cpu();
 
 	if (mm == current->mm && atomic_read(&mm->mm_users) == 1)
-		mm->cpu_vm_mask = cpumask_of_cpu(cpu);
+		cpumask_copy(mm_cpumask(mm), cpumask_of(cpu));
 	else
 		smp_cross_call_masked(&xcall_flush_tlb_pending,
 				      ctx, nr, (unsigned long) vaddrs,
-				      &mm->cpu_vm_mask);
+				      mm_cpumask(mm));
 
 	__flush_tlb_pending(ctx, nr, vaddrs);
 
diff --git a/arch/sparc/kernel/sun4d_irq.c b/arch/sparc/kernel/sun4d_irq.c
index 3369fef..ab036a7 100644
--- a/arch/sparc/kernel/sun4d_irq.c
+++ b/arch/sparc/kernel/sun4d_irq.c
@@ -326,7 +326,6 @@ int sun4d_request_irq(unsigned int irq,
 
 	action->handler = handler;
 	action->flags = irqflags;
-	cpus_clear(action->mask);
 	action->name = devname;
 	action->next = NULL;
 	action->dev_id = dev_id;
diff --git a/arch/sparc/kernel/sun4d_smp.c b/arch/sparc/kernel/sun4d_smp.c
index 50afaed..54fb024 100644
--- a/arch/sparc/kernel/sun4d_smp.c
+++ b/arch/sparc/kernel/sun4d_smp.c
@@ -150,7 +150,7 @@ void __cpuinit smp4d_callin(void)
 	spin_lock_irqsave(&sun4d_imsk_lock, flags);
 	cc_set_imsk(cc_get_imsk() & ~0x4000); /* Allow PIL 14 as well */
 	spin_unlock_irqrestore(&sun4d_imsk_lock, flags);
-	cpu_set(cpuid, cpu_online_map);
+	set_cpu_online(cpuid, true);
 
 }
 
@@ -228,11 +228,10 @@ void __init smp4d_smp_done(void)
 	/* setup cpu list for irq rotation */
 	first = 0;
 	prev = &first;
-	for (i = 0; i < NR_CPUS; i++)
-		if (cpu_online(i)) {
-			*prev = i;
-			prev = &cpu_data(i).next;
-		}
+	for_each_online_cpu(i) {
+		*prev = i;
+		prev = &cpu_data(i).next;
+	}
 	*prev = first;
 	local_flush_cache_all();
 
diff --git a/arch/sparc/kernel/sun4m_smp.c b/arch/sparc/kernel/sun4m_smp.c
index 8040376..960b113 100644
--- a/arch/sparc/kernel/sun4m_smp.c
+++ b/arch/sparc/kernel/sun4m_smp.c
@@ -113,7 +113,7 @@ void __cpuinit smp4m_callin(void)
 
 	local_irq_enable();
 
-	cpu_set(cpuid, cpu_online_map);
+	set_cpu_online(cpuid, true);
 }
 
 /*
@@ -186,11 +186,9 @@ void __init smp4m_smp_done(void)
 	/* setup cpu list for irq rotation */
 	first = 0;
 	prev = &first;
-	for (i = 0; i < NR_CPUS; i++) {
-		if (cpu_online(i)) {
-			*prev = i;
-			prev = &cpu_data(i).next;
-		}
+	for_each_online_cpu(i) {
+		*prev = i;
+		prev = &cpu_data(i).next;
 	}
 	*prev = first;
 	local_flush_cache_all();
diff --git a/arch/sparc/mm/highmem.c b/arch/sparc/mm/highmem.c
index 752d0c9..7916feb 100644
--- a/arch/sparc/mm/highmem.c
+++ b/arch/sparc/mm/highmem.c
@@ -39,6 +39,7 @@ void *kmap_atomic(struct page *page, enum km_type type)
 	if (!PageHighMem(page))
 		return page_address(page);
 
+	debug_kmap_atomic(type);
 	idx = type + KM_TYPE_NR*smp_processor_id();
 	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
 
diff --git a/arch/sparc/mm/init_64.c b/arch/sparc/mm/init_64.c
index 00373ce..2c8dfeb 100644
--- a/arch/sparc/mm/init_64.c
+++ b/arch/sparc/mm/init_64.c
@@ -1092,7 +1092,7 @@ static void __init numa_parse_mdesc_group_cpus(struct mdesc_handle *md,
 		if (strcmp(name, "cpu"))
 			continue;
 		id = mdesc_get_property(md, target, "id", NULL);
-		if (*id < NR_CPUS)
+		if (*id < nr_cpu_ids)
 			cpu_set(*id, *mask);
 	}
 }
diff --git a/arch/sparc/mm/srmmu.c b/arch/sparc/mm/srmmu.c
index fe7ed08..06c9a7d 100644
--- a/arch/sparc/mm/srmmu.c
+++ b/arch/sparc/mm/srmmu.c
@@ -1425,7 +1425,7 @@ static void __init init_vac_layout(void)
 				min_line_size = vac_line_size;
 			//FIXME: cpus not contiguous!!
 			cpu++;
-			if (cpu >= NR_CPUS || !cpu_online(cpu))
+			if (cpu >= nr_cpu_ids || !cpu_online(cpu))
 				break;
 #else
 			break;
diff --git a/arch/um/drivers/pcap_user.h b/arch/um/drivers/pcap_user.h
index 96b80b5..d8ba615 100644
--- a/arch/um/drivers/pcap_user.h
+++ b/arch/um/drivers/pcap_user.h
@@ -19,13 +19,3 @@ extern const struct net_user_info pcap_user_info;
 
 extern int pcap_user_read(int fd, void *buf, int len, struct pcap_data *pri);
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/drivers/port.h b/arch/um/drivers/port.h
index 9117609..372a80c 100644
--- a/arch/um/drivers/port.h
+++ b/arch/um/drivers/port.h
@@ -18,13 +18,3 @@ extern void port_remove_dev(void *d);
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/drivers/ssl.h b/arch/um/drivers/ssl.h
index 98412aa..314d177 100644
--- a/arch/um/drivers/ssl.h
+++ b/arch/um/drivers/ssl.h
@@ -11,13 +11,3 @@ extern void ssl_receive_char(int line, char ch);
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/drivers/stdio_console.h b/arch/um/drivers/stdio_console.h
index 505a3d5..6d8275f 100644
--- a/arch/um/drivers/stdio_console.h
+++ b/arch/um/drivers/stdio_console.h
@@ -9,13 +9,3 @@
 extern void save_console_flags(void);
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/drivers/ubd_kern.c b/arch/um/drivers/ubd_kern.c
index 0a86811..d42f826 100644
--- a/arch/um/drivers/ubd_kern.c
+++ b/arch/um/drivers/ubd_kern.c
@@ -17,7 +17,6 @@
  * James McMechan
  */
 
-#define MAJOR_NR UBD_MAJOR
 #define UBD_SHIFT 4
 
 #include "linux/kernel.h"
@@ -115,7 +114,7 @@ static struct block_device_operations ubd_blops = {
 };
 
 /* Protected by ubd_lock */
-static int fake_major = MAJOR_NR;
+static int fake_major = UBD_MAJOR;
 static struct gendisk *ubd_gendisk[MAX_DEV];
 static struct gendisk *fake_gendisk[MAX_DEV];
 
@@ -299,7 +298,7 @@ static int ubd_setup_common(char *str, int *index_out, char **error_out)
 		}
 
 		mutex_lock(&ubd_lock);
-		if(fake_major != MAJOR_NR){
+		if (fake_major != UBD_MAJOR) {
 			*error_out = "Can't assign a fake major twice";
 			goto out1;
 		}
@@ -818,13 +817,13 @@ static int ubd_disk_register(int major, u64 size, int unit,
 	disk->first_minor = unit << UBD_SHIFT;
 	disk->fops = &ubd_blops;
 	set_capacity(disk, size / 512);
-	if(major == MAJOR_NR)
+	if (major == UBD_MAJOR)
 		sprintf(disk->disk_name, "ubd%c", 'a' + unit);
 	else
 		sprintf(disk->disk_name, "ubd_fake%d", unit);
 
 	/* sysfs register (not for ide fake devices) */
-	if (major == MAJOR_NR) {
+	if (major == UBD_MAJOR) {
 		ubd_devs[unit].pdev.id   = unit;
 		ubd_devs[unit].pdev.name = DRIVER_NAME;
 		ubd_devs[unit].pdev.dev.release = ubd_device_release;
@@ -871,13 +870,13 @@ static int ubd_add(int n, char **error_out)
 	ubd_dev->queue->queuedata = ubd_dev;
 
 	blk_queue_max_hw_segments(ubd_dev->queue, MAX_SG);
-	err = ubd_disk_register(MAJOR_NR, ubd_dev->size, n, &ubd_gendisk[n]);
+	err = ubd_disk_register(UBD_MAJOR, ubd_dev->size, n, &ubd_gendisk[n]);
 	if(err){
 		*error_out = "Failed to register device";
 		goto out_cleanup;
 	}
 
-	if(fake_major != MAJOR_NR)
+	if (fake_major != UBD_MAJOR)
 		ubd_disk_register(fake_major, ubd_dev->size, n,
 				  &fake_gendisk[n]);
 
@@ -1059,10 +1058,10 @@ static int __init ubd_init(void)
 	char *error;
 	int i, err;
 
-	if (register_blkdev(MAJOR_NR, "ubd"))
+	if (register_blkdev(UBD_MAJOR, "ubd"))
 		return -1;
 
-	if (fake_major != MAJOR_NR) {
+	if (fake_major != UBD_MAJOR) {
 		char name[sizeof("ubd_nnn\0")];
 
 		snprintf(name, sizeof(name), "ubd_%d", fake_major);
diff --git a/arch/um/drivers/xterm.h b/arch/um/drivers/xterm.h
index f33a6e7..56b9c4a 100644
--- a/arch/um/drivers/xterm.h
+++ b/arch/um/drivers/xterm.h
@@ -10,13 +10,3 @@ extern int xterm_fd(int socket, int *pid_out);
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/asm/irq_vectors.h b/arch/um/include/asm/irq_vectors.h
index 62ddba6..272a81e 100644
--- a/arch/um/include/asm/irq_vectors.h
+++ b/arch/um/include/asm/irq_vectors.h
@@ -8,13 +8,3 @@
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/asm/mmu.h b/arch/um/include/asm/mmu.h
index 2cf35c2..cf259de 100644
--- a/arch/um/include/asm/mmu.h
+++ b/arch/um/include/asm/mmu.h
@@ -10,13 +10,3 @@
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/asm/pda.h b/arch/um/include/asm/pda.h
index 0d8bf33..ddcd774 100644
--- a/arch/um/include/asm/pda.h
+++ b/arch/um/include/asm/pda.h
@@ -19,13 +19,3 @@ extern struct foo me;
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/asm/pgalloc.h b/arch/um/include/asm/pgalloc.h
index 9062a6e..7189843 100644
--- a/arch/um/include/asm/pgalloc.h
+++ b/arch/um/include/asm/pgalloc.h
@@ -60,13 +60,3 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd)
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/asm/pgtable-3level.h b/arch/um/include/asm/pgtable-3level.h
index 0446f45..084de4a 100644
--- a/arch/um/include/asm/pgtable-3level.h
+++ b/arch/um/include/asm/pgtable-3level.h
@@ -134,13 +134,3 @@ static inline pmd_t pfn_pmd(pfn_t page_nr, pgprot_t pgprot)
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/shared/frame_kern.h b/arch/um/include/shared/frame_kern.h
index ce9514f..7607849 100644
--- a/arch/um/include/shared/frame_kern.h
+++ b/arch/um/include/shared/frame_kern.h
@@ -20,13 +20,3 @@ extern int setup_signal_stack_si(unsigned long stack_top, int sig,
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/shared/initrd.h b/arch/um/include/shared/initrd.h
index 439b9a8..22673bc 100644
--- a/arch/um/include/shared/initrd.h
+++ b/arch/um/include/shared/initrd.h
@@ -10,13 +10,3 @@ extern int load_initrd(char *filename, void *buf, int size);
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/shared/irq_kern.h b/arch/um/include/shared/irq_kern.h
index fba3895..b05d22f 100644
--- a/arch/um/include/shared/irq_kern.h
+++ b/arch/um/include/shared/irq_kern.h
@@ -16,13 +16,3 @@ extern int um_request_irq(unsigned int irq, int fd, int type,
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/shared/mem_kern.h b/arch/um/include/shared/mem_kern.h
index cb7e196..69be0fd 100644
--- a/arch/um/include/shared/mem_kern.h
+++ b/arch/um/include/shared/mem_kern.h
@@ -18,13 +18,3 @@ extern void register_remapper(struct remapper *info);
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/include/shared/ubd_user.h b/arch/um/include/shared/ubd_user.h
index bb66517..3845051 100644
--- a/arch/um/include/shared/ubd_user.h
+++ b/arch/um/include/shared/ubd_user.h
@@ -14,13 +14,3 @@ extern int kernel_fd;
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/kernel/Makefile b/arch/um/kernel/Makefile
index 499e5e9..388ec0a 100644
--- a/arch/um/kernel/Makefile
+++ b/arch/um/kernel/Makefile
@@ -28,7 +28,7 @@ $(obj)/config.tmp: $(objtree)/.config FORCE
 	$(call if_changed,quote1)
 
 quiet_cmd_quote1 = QUOTE   $@
-      cmd_quote1 = sed -e 's/"/\\"/g' -e 's/^/"/' -e 's/$$/\\n"/' \
+      cmd_quote1 = sed -e 's/"/\\"/g' -e 's/^/"/' -e 's/$$/\\n",/' \
 		   $< > $@
 
 $(obj)/config.c: $(src)/config.c.in $(obj)/config.tmp FORCE
@@ -36,9 +36,9 @@ $(obj)/config.c: $(src)/config.c.in $(obj)/config.tmp FORCE
 
 quiet_cmd_quote2 = QUOTE   $@
       cmd_quote2 = sed -e '/CONFIG/{'          \
-		  -e 's/"CONFIG"\;/""/'        \
+		  -e 's/"CONFIG"//'            \
 		  -e 'r $(obj)/config.tmp'     \
 		  -e 'a \'                     \
-		  -e '""\;'                    \
+		  -e '""'                      \
 		  -e '}'                       \
 		  $< > $@
diff --git a/arch/um/kernel/config.c.in b/arch/um/kernel/config.c.in
index c062cbf..b7a43fe 100644
--- a/arch/um/kernel/config.c.in
+++ b/arch/um/kernel/config.c.in
@@ -7,11 +7,15 @@
 #include <stdlib.h>
 #include "init.h"
 
-static __initdata char *config = "CONFIG";
+static __initdata const char *config[] = {
+"CONFIG"
+};
 
 static int __init print_config(char *line, int *add)
 {
-	printf("%s", config);
+	int i;
+	for (i = 0; i < sizeof(config)/sizeof(config[0]); i++)
+		printf("%s", config[i]);
 	exit(0);
 }
 
@@ -20,13 +24,3 @@ __uml_setup("--showconfig", print_config,
 "    Prints the config file that this UML binary was generated from.\n\n"
 );
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/os-Linux/start_up.c b/arch/um/os-Linux/start_up.c
index 183db26..02ee9ad 100644
--- a/arch/um/os-Linux/start_up.c
+++ b/arch/um/os-Linux/start_up.c
@@ -244,7 +244,7 @@ static void __init check_sysemu(void)
 
 	if ((ptrace(PTRACE_OLDSETOPTIONS, pid, 0,
 		   (void *) PTRACE_O_TRACESYSGOOD) < 0))
-		fatal_perror("check_ptrace: PTRACE_OLDSETOPTIONS failed");
+		fatal_perror("check_sysemu: PTRACE_OLDSETOPTIONS failed");
 
 	while (1) {
 		count++;
@@ -252,12 +252,12 @@ static void __init check_sysemu(void)
 			goto fail;
 		CATCH_EINTR(n = waitpid(pid, &status, WUNTRACED));
 		if (n < 0)
-			fatal_perror("check_ptrace : wait failed");
+			fatal_perror("check_sysemu: wait failed");
 
 		if (WIFSTOPPED(status) &&
 		    (WSTOPSIG(status) == (SIGTRAP|0x80))) {
 			if (!count) {
-				non_fatal("check_ptrace : SYSEMU_SINGLESTEP "
+				non_fatal("check_sysemu: SYSEMU_SINGLESTEP "
 					  "doesn't singlestep");
 				goto fail;
 			}
@@ -271,7 +271,7 @@ static void __init check_sysemu(void)
 		else if (WIFSTOPPED(status) && (WSTOPSIG(status) == SIGTRAP))
 			count++;
 		else {
-			non_fatal("check_ptrace : expected SIGTRAP or "
+			non_fatal("check_sysemu: expected SIGTRAP or "
 				  "(SIGTRAP | 0x80), got status = %d\n",
 				  status);
 			goto fail;
diff --git a/arch/um/sys-i386/asm/archparam.h b/arch/um/sys-i386/asm/archparam.h
index 93fd723..2a18a88 100644
--- a/arch/um/sys-i386/asm/archparam.h
+++ b/arch/um/sys-i386/asm/archparam.h
@@ -14,13 +14,3 @@
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-i386/shared/sysdep/checksum.h b/arch/um/sys-i386/shared/sysdep/checksum.h
index 0cb4645..ed47445 100644
--- a/arch/um/sys-i386/shared/sysdep/checksum.h
+++ b/arch/um/sys-i386/shared/sysdep/checksum.h
@@ -199,13 +199,3 @@ static __inline__ __wsum csum_and_copy_to_user(const void *src,
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ia64/sysdep/ptrace.h b/arch/um/sys-ia64/sysdep/ptrace.h
index 42dd8fb..0f0f4e6 100644
--- a/arch/um/sys-ia64/sysdep/ptrace.h
+++ b/arch/um/sys-ia64/sysdep/ptrace.h
@@ -14,13 +14,3 @@ struct sys_pt_regs {
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ia64/sysdep/sigcontext.h b/arch/um/sys-ia64/sysdep/sigcontext.h
index f15fb25..76b4316 100644
--- a/arch/um/sys-ia64/sysdep/sigcontext.h
+++ b/arch/um/sys-ia64/sysdep/sigcontext.h
@@ -8,13 +8,3 @@
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ia64/sysdep/syscalls.h b/arch/um/sys-ia64/sysdep/syscalls.h
index 4a1f46e..5f6700c 100644
--- a/arch/um/sys-ia64/sysdep/syscalls.h
+++ b/arch/um/sys-ia64/sysdep/syscalls.h
@@ -8,13 +8,3 @@
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ppc/miscthings.c b/arch/um/sys-ppc/miscthings.c
index 373061c..1c11aed 100644
--- a/arch/um/sys-ppc/miscthings.c
+++ b/arch/um/sys-ppc/miscthings.c
@@ -40,14 +40,3 @@ void shove_aux_table(unsigned long sp)
 }
 /* END stuff taken from arch/ppc/kernel/process.c */
 
-
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ppc/ptrace.c b/arch/um/sys-ppc/ptrace.c
index 8e71b47..66ef155 100644
--- a/arch/um/sys-ppc/ptrace.c
+++ b/arch/um/sys-ppc/ptrace.c
@@ -56,13 +56,3 @@ int peek_user(struct task_struct *child, long addr, long data)
 	return put_user(tmp, (unsigned long *) data);
 }
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ppc/ptrace_user.c b/arch/um/sys-ppc/ptrace_user.c
index ff0b9c0..224d240 100644
--- a/arch/um/sys-ppc/ptrace_user.c
+++ b/arch/um/sys-ppc/ptrace_user.c
@@ -27,13 +27,3 @@ int ptrace_setregs(long pid, unsigned long *regs_in)
     }
     return 0;
 }
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ppc/shared/sysdep/ptrace.h b/arch/um/sys-ppc/shared/sysdep/ptrace.h
index df2397d..0e3230e 100644
--- a/arch/um/sys-ppc/shared/sysdep/ptrace.h
+++ b/arch/um/sys-ppc/shared/sysdep/ptrace.h
@@ -91,13 +91,3 @@ extern void shove_aux_table(unsigned long sp);
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ppc/shared/sysdep/sigcontext.h b/arch/um/sys-ppc/shared/sysdep/sigcontext.h
index f20d965..b7286f0 100644
--- a/arch/um/sys-ppc/shared/sysdep/sigcontext.h
+++ b/arch/um/sys-ppc/shared/sysdep/sigcontext.h
@@ -50,13 +50,3 @@
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ppc/shared/sysdep/syscalls.h b/arch/um/sys-ppc/shared/sysdep/syscalls.h
index 679df35..1ff8155 100644
--- a/arch/um/sys-ppc/shared/sysdep/syscalls.h
+++ b/arch/um/sys-ppc/shared/sysdep/syscalls.h
@@ -41,13 +41,3 @@ int old_mmap(unsigned long addr, unsigned long len,
 
 #define LAST_ARCH_SYSCALL __NR_fadvise64
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-ppc/sigcontext.c b/arch/um/sys-ppc/sigcontext.c
index 4bdc15c..40694d0 100644
--- a/arch/um/sys-ppc/sigcontext.c
+++ b/arch/um/sys-ppc/sigcontext.c
@@ -2,13 +2,3 @@
 #include "asm/sigcontext.h"
 #include "sysdep/ptrace.h"
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-x86_64/asm/archparam.h b/arch/um/sys-x86_64/asm/archparam.h
index 270ed95..6c08366 100644
--- a/arch/um/sys-x86_64/asm/archparam.h
+++ b/arch/um/sys-x86_64/asm/archparam.h
@@ -14,13 +14,3 @@
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-x86_64/asm/module.h b/arch/um/sys-x86_64/asm/module.h
index 35b5491..8eb79c2 100644
--- a/arch/um/sys-x86_64/asm/module.h
+++ b/arch/um/sys-x86_64/asm/module.h
@@ -18,13 +18,3 @@ struct mod_arch_specific
 
 #endif
 
-/*
- * Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/um/sys-x86_64/mem.c b/arch/um/sys-x86_64/mem.c
index 3f59a0a..3f8df8a 100644
--- a/arch/um/sys-x86_64/mem.c
+++ b/arch/um/sys-x86_64/mem.c
@@ -14,12 +14,3 @@ unsigned long vm_data_default_flags = __VM_DATA_DEFAULT_FLAGS;
 unsigned long vm_data_default_flags32 = __VM_DATA_DEFAULT_FLAGS;
 unsigned long vm_force_exec32 = PROT_EXEC;
 
-/* Overrides for Emacs so that we follow Linus's tabbing style.
- * Emacs will notice this stuff at the end of the file and automatically
- * adjust the settings for this buffer only.  This must remain at the end
- * of the file.
- * ---------------------------------------------------------------------------
- * Local variables:
- * c-file-style: "linux"
- * End:
- */
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 45161b8..748e50a 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -165,6 +165,9 @@ config AUDIT_ARCH
 config ARCH_SUPPORTS_OPTIMIZED_INLINING
 	def_bool y
 
+config ARCH_SUPPORTS_DEBUG_PAGEALLOC
+	def_bool y
+
 # Use the generic interrupt handling code in kernel/irq/:
 config GENERIC_HARDIRQS
 	bool
diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
index fdb45df..a345cb5 100644
--- a/arch/x86/Kconfig.debug
+++ b/arch/x86/Kconfig.debug
@@ -75,6 +75,7 @@ config DEBUG_STACK_USAGE
 config DEBUG_PAGEALLOC
 	bool "Debug page memory allocations"
 	depends on DEBUG_KERNEL
+	depends on ARCH_SUPPORTS_DEBUG_PAGEALLOC
 	---help---
 	  Unmap pages from the kernel linear mapping after free_pages().
 	  This results in a large slowdown, but helps to find certain types
diff --git a/arch/x86/boot/memory.c b/arch/x86/boot/memory.c
index 8c3c25f..5054c2d 100644
--- a/arch/x86/boot/memory.c
+++ b/arch/x86/boot/memory.c
@@ -2,6 +2,7 @@
  *
  *   Copyright (C) 1991, 1992 Linus Torvalds
  *   Copyright 2007 rPath, Inc. - All Rights Reserved
+ *   Copyright 2009 Intel Corporation; author H. Peter Anvin
  *
  *   This file is part of the Linux kernel, and is made available under
  *   the terms of the GNU General Public License version 2.
@@ -16,24 +17,38 @@
 
 #define SMAP	0x534d4150	/* ASCII "SMAP" */
 
+struct e820_ext_entry {
+	struct e820entry std;
+	u32 ext_flags;
+} __attribute__((packed));
+
 static int detect_memory_e820(void)
 {
 	int count = 0;
 	u32 next = 0;
-	u32 size, id;
+	u32 size, id, edi;
 	u8 err;
 	struct e820entry *desc = boot_params.e820_map;
+	static struct e820_ext_entry buf; /* static so it is zeroed */
+
+	/*
+	 * Set this here so that if the BIOS doesn't change this field
+	 * but still doesn't change %ecx, we're still okay...
+	 */
+	buf.ext_flags = 1;
 
 	do {
-		size = sizeof(struct e820entry);
+		size = sizeof buf;
 
-		/* Important: %edx is clobbered by some BIOSes,
-		   so it must be either used for the error output
-		   or explicitly marked clobbered. */
-		asm("int $0x15; setc %0"
+		/* Important: %edx and %esi are clobbered by some BIOSes,
+		   so they must be either used for the error output
+		   or explicitly marked clobbered.  Given that, assume there
+		   is something out there clobbering %ebp and %edi, too. */
+		asm("pushl %%ebp; int $0x15; popl %%ebp; setc %0"
 		    : "=d" (err), "+b" (next), "=a" (id), "+c" (size),
-		      "=m" (*desc)
-		    : "D" (desc), "d" (SMAP), "a" (0xe820));
+		      "=D" (edi), "+m" (buf)
+		    : "D" (&buf), "d" (SMAP), "a" (0xe820)
+		    : "esi");
 
 		/* BIOSes which terminate the chain with CF = 1 as opposed
 		   to %ebx = 0 don't always report the SMAP signature on
@@ -51,8 +66,14 @@ static int detect_memory_e820(void)
 			break;
 		}
 
+		/* ACPI 3.0 added the extended flags support.  If bit 0
+		   in the extended flags is zero, we're supposed to simply
+		   ignore the entry -- a backwards incompatible change! */
+		if (size > 20 && !(buf.ext_flags & 1))
+			continue;
+
+		*desc++ = buf.std;
 		count++;
-		desc++;
 	} while (next && count < ARRAY_SIZE(boot_params.e820_map));
 
 	return boot_params.e820_entries = count;
diff --git a/arch/x86/include/asm/lguest_hcall.h b/arch/x86/include/asm/lguest_hcall.h
index 4389442..0f4ee71 100644
--- a/arch/x86/include/asm/lguest_hcall.h
+++ b/arch/x86/include/asm/lguest_hcall.h
@@ -26,36 +26,20 @@
 
 #ifndef __ASSEMBLY__
 #include <asm/hw_irq.h>
+#include <asm/kvm_para.h>
 
 /*G:031 But first, how does our Guest contact the Host to ask for privileged
  * operations?  There are two ways: the direct way is to make a "hypercall",
  * to make requests of the Host Itself.
  *
- * Our hypercall mechanism uses the highest unused trap code (traps 32 and
- * above are used by real hardware interrupts).  Fifteen hypercalls are
+ * We use the KVM hypercall mechanism. Eighteen hypercalls are
  * available: the hypercall number is put in the %eax register, and the
- * arguments (when required) are placed in %edx, %ebx and %ecx.  If a return
+ * arguments (when required) are placed in %ebx, %ecx and %edx.  If a return
  * value makes sense, it's returned in %eax.
  *
  * Grossly invalid calls result in Sudden Death at the hands of the vengeful
  * Host, rather than returning failure.  This reflects Winston Churchill's
  * definition of a gentleman: "someone who is only rude intentionally". */
-static inline unsigned long
-hcall(unsigned long call,
-      unsigned long arg1, unsigned long arg2, unsigned long arg3)
-{
-	/* "int" is the Intel instruction to trigger a trap. */
-	asm volatile("int $" __stringify(LGUEST_TRAP_ENTRY)
-		     /* The call in %eax (aka "a") might be overwritten */
-		     : "=a"(call)
-		       /* The arguments are in %eax, %edx, %ebx & %ecx */
-		     : "a"(call), "d"(arg1), "b"(arg2), "c"(arg3)
-		       /* "memory" means this might write somewhere in memory.
-			* This isn't true for all calls, but it's safe to tell
-			* gcc that it might happen so it doesn't get clever. */
-		     : "memory");
-	return call;
-}
 /*:*/
 
 /* Can't use our min() macro here: needs to be a constant */
@@ -64,7 +48,7 @@ hcall(unsigned long call,
 #define LHCALL_RING_SIZE 64
 struct hcall_args {
 	/* These map directly onto eax, ebx, ecx, edx in struct lguest_regs */
-	unsigned long arg0, arg2, arg3, arg1;
+	unsigned long arg0, arg1, arg2, arg3;
 };
 
 #endif /* !__ASSEMBLY__ */
diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h
index a977de2..a0301bf 100644
--- a/arch/x86/include/asm/pci.h
+++ b/arch/x86/include/asm/pci.h
@@ -86,6 +86,9 @@ static inline void early_quirks(void) { }
 
 extern void pci_iommu_alloc(void);
 
+/* MSI arch hook */
+#define arch_setup_msi_irqs arch_setup_msi_irqs
+
 #endif  /* __KERNEL__ */
 
 #ifdef CONFIG_X86_32
diff --git a/arch/x86/include/asm/suspend_32.h b/arch/x86/include/asm/suspend_32.h
index a5074bd..48dcfa6 100644
--- a/arch/x86/include/asm/suspend_32.h
+++ b/arch/x86/include/asm/suspend_32.h
@@ -24,28 +24,4 @@ struct saved_context {
 	unsigned long return_address;
 } __attribute__((packed));
 
-#ifdef CONFIG_ACPI
-extern unsigned long saved_eip;
-extern unsigned long saved_esp;
-extern unsigned long saved_ebp;
-extern unsigned long saved_ebx;
-extern unsigned long saved_esi;
-extern unsigned long saved_edi;
-
-static inline void acpi_save_register_state(unsigned long return_point)
-{
-	saved_eip = return_point;
-	asm volatile("movl %%esp,%0" : "=m" (saved_esp));
-	asm volatile("movl %%ebp,%0" : "=m" (saved_ebp));
-	asm volatile("movl %%ebx,%0" : "=m" (saved_ebx));
-	asm volatile("movl %%edi,%0" : "=m" (saved_edi));
-	asm volatile("movl %%esi,%0" : "=m" (saved_esi));
-}
-
-#define acpi_restore_register_state()  do {} while (0)
-
-/* routines for saving/restoring kernel state */
-extern int acpi_save_state_mem(void);
-#endif
-
 #endif /* _ASM_X86_SUSPEND_32_H */
diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h
index 77cfb2c..744299c 100644
--- a/arch/x86/include/asm/topology.h
+++ b/arch/x86/include/asm/topology.h
@@ -217,10 +217,6 @@ static inline cpumask_t node_to_cpumask(int node)
 {
 	return cpu_online_map;
 }
-static inline int node_to_first_cpu(int node)
-{
-	return first_cpu(cpu_online_map);
-}
 
 static inline void setup_node_to_cpumask_map(void) { }
 
@@ -237,14 +233,6 @@ static inline void setup_node_to_cpumask_map(void) { }
 
 #include <asm-generic/topology.h>
 
-#ifdef CONFIG_NUMA
-/* Returns the number of the first CPU on Node 'node'. */
-static inline int node_to_first_cpu(int node)
-{
-	return cpumask_first(cpumask_of_node(node));
-}
-#endif
-
 extern cpumask_t cpu_coregroup_map(int cpu);
 extern const struct cpumask *cpu_coregroup_mask(int cpu);
 
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index da99ffc..1bb5c6c 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -3468,6 +3468,10 @@ int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
 	struct intel_iommu *iommu = NULL;
 	int index = 0;
 
+	/* x86 doesn't support multiple MSI yet */
+	if (type == PCI_CAP_ID_MSI && nvec > 1)
+		return 1;
+
 	irq_want = nr_irqs_gsi;
 	sub_handle = 0;
 	list_for_each_entry(msidesc, &dev->msi_list, list) {
diff --git a/arch/x86/kernel/asm-offsets_32.c b/arch/x86/kernel/asm-offsets_32.c
index fbf2f33..5a6aa1c 100644
--- a/arch/x86/kernel/asm-offsets_32.c
+++ b/arch/x86/kernel/asm-offsets_32.c
@@ -18,6 +18,7 @@
 #include <asm/thread_info.h>
 #include <asm/bootparam.h>
 #include <asm/elf.h>
+#include <asm/suspend.h>
 
 #include <xen/interface/xen.h>
 
diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c
index 8793ab3..e72f062 100644
--- a/arch/x86/kernel/asm-offsets_64.c
+++ b/arch/x86/kernel/asm-offsets_64.c
@@ -16,6 +16,7 @@
 #include <asm/thread_info.h>
 #include <asm/ia32.h>
 #include <asm/bootparam.h>
+#include <asm/suspend.h>
 
 #include <xen/interface/xen.h>
 
diff --git a/arch/x86/kernel/cpu/mtrr/if.c b/arch/x86/kernel/cpu/mtrr/if.c
index 4c42146..fb73a52 100644
--- a/arch/x86/kernel/cpu/mtrr/if.c
+++ b/arch/x86/kernel/cpu/mtrr/if.c
@@ -377,10 +377,6 @@ static const struct file_operations mtrr_fops = {
 	.release = mtrr_close,
 };
 
-
-static struct proc_dir_entry *proc_root_mtrr;
-
-
 static int mtrr_seq_show(struct seq_file *seq, void *offset)
 {
 	char factor;
@@ -423,11 +419,7 @@ static int __init mtrr_if_init(void)
 	    (!cpu_has(c, X86_FEATURE_CENTAUR_MCR)))
 		return -ENODEV;
 
-	proc_root_mtrr =
-		proc_create("mtrr", S_IWUSR | S_IRUGO, NULL, &mtrr_fops);
-
-	if (proc_root_mtrr)
-		proc_root_mtrr->owner = THIS_MODULE;
+	proc_create("mtrr", S_IWUSR | S_IRUGO, NULL, &mtrr_fops);
 	return 0;
 }
 
diff --git a/arch/x86/kernel/irqinit_32.c b/arch/x86/kernel/irqinit_32.c
index bc13261..368b0a8 100644
--- a/arch/x86/kernel/irqinit_32.c
+++ b/arch/x86/kernel/irqinit_32.c
@@ -50,7 +50,6 @@ static irqreturn_t math_error_irq(int cpl, void *dev_id)
  */
 static struct irqaction fpu_irq = {
 	.handler = math_error_irq,
-	.mask = CPU_MASK_NONE,
 	.name = "fpu",
 };
 
@@ -83,7 +82,6 @@ void __init init_ISA_irqs(void)
  */
 static struct irqaction irq2 = {
 	.handler = no_action,
-	.mask = CPU_MASK_NONE,
 	.name = "cascade",
 };
 
diff --git a/arch/x86/kernel/irqinit_64.c b/arch/x86/kernel/irqinit_64.c
index c7a49e0..8cd1053 100644
--- a/arch/x86/kernel/irqinit_64.c
+++ b/arch/x86/kernel/irqinit_64.c
@@ -45,7 +45,6 @@
 
 static struct irqaction irq2 = {
 	.handler = no_action,
-	.mask = CPU_MASK_NONE,
 	.name = "cascade",
 };
 DEFINE_PER_CPU(vector_irq_t, vector_irq) = {
diff --git a/arch/x86/kernel/mfgpt_32.c b/arch/x86/kernel/mfgpt_32.c
index 8815f3c..846510b 100644
--- a/arch/x86/kernel/mfgpt_32.c
+++ b/arch/x86/kernel/mfgpt_32.c
@@ -348,7 +348,6 @@ static irqreturn_t mfgpt_tick(int irq, void *dev_id)
 static struct irqaction mfgptirq  = {
 	.handler = mfgpt_tick,
 	.flags = IRQF_DISABLED | IRQF_NOBALANCING,
-	.mask = CPU_MASK_NONE,
 	.name = "mfgpt-timer"
 };
 
diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
index c7c4776..90f5b9e 100644
--- a/arch/x86/kernel/pci-dma.c
+++ b/arch/x86/kernel/pci-dma.c
@@ -300,8 +300,7 @@ fs_initcall(pci_iommu_init);
 static __devinit void via_no_dac(struct pci_dev *dev)
 {
 	if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI && forbid_dac == 0) {
-		printk(KERN_INFO
-			"PCI: VIA PCI bridge detected. Disabling DAC.\n");
+		dev_info(&dev->dev, "disabling DAC on VIA PCI bridge\n");
 		forbid_dac = 1;
 	}
 }
diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
index a0d2623..b415843 100644
--- a/arch/x86/kernel/setup.c
+++ b/arch/x86/kernel/setup.c
@@ -1049,7 +1049,6 @@ void __init x86_quirk_trap_init(void)
 static struct irqaction irq0  = {
 	.handler = timer_interrupt,
 	.flags = IRQF_DISABLED | IRQF_NOBALANCING | IRQF_IRQPOLL | IRQF_TIMER,
-	.mask = CPU_MASK_NONE,
 	.name = "timer"
 };
 
diff --git a/arch/x86/kernel/time_64.c b/arch/x86/kernel/time_64.c
index 241ec39..5ba343e 100644
--- a/arch/x86/kernel/time_64.c
+++ b/arch/x86/kernel/time_64.c
@@ -116,7 +116,6 @@ unsigned long __init calibrate_cpu(void)
 static struct irqaction irq0 = {
 	.handler	= timer_interrupt,
 	.flags		= IRQF_DISABLED | IRQF_IRQPOLL | IRQF_NOBALANCING | IRQF_TIMER,
-	.mask		= CPU_MASK_NONE,
 	.name		= "timer"
 };
 
@@ -125,7 +124,6 @@ void __init hpet_time_init(void)
 	if (!hpet_enable())
 		setup_pit_timer();
 
-	irq0.mask = cpumask_of_cpu(0);
 	setup_irq(0, &irq0);
 }
 
diff --git a/arch/x86/kernel/vmiclock_32.c b/arch/x86/kernel/vmiclock_32.c
index 33a788d..d303369 100644
--- a/arch/x86/kernel/vmiclock_32.c
+++ b/arch/x86/kernel/vmiclock_32.c
@@ -202,7 +202,6 @@ static struct irqaction vmi_clock_action  = {
 	.name 		= "vmi-timer",
 	.handler 	= vmi_timer_interrupt,
 	.flags 		= IRQF_DISABLED | IRQF_NOBALANCING | IRQF_TIMER,
-	.mask 		= CPU_MASK_ALL,
 };
 
 static void __devinit vmi_time_init_clockevent(void)
diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c
index 90e44a1..e94a11e 100644
--- a/arch/x86/lguest/boot.c
+++ b/arch/x86/lguest/boot.c
@@ -107,7 +107,7 @@ static void async_hcall(unsigned long call, unsigned long arg1,
 	local_irq_save(flags);
 	if (lguest_data.hcall_status[next_call] != 0xFF) {
 		/* Table full, so do normal hcall which will flush table. */
-		hcall(call, arg1, arg2, arg3);
+		kvm_hypercall3(call, arg1, arg2, arg3);
 	} else {
 		lguest_data.hcalls[next_call].arg0 = call;
 		lguest_data.hcalls[next_call].arg1 = arg1;
@@ -134,13 +134,32 @@ static void async_hcall(unsigned long call, unsigned long arg1,
  *
  * So, when we're in lazy mode, we call async_hcall() to store the call for
  * future processing: */
-static void lazy_hcall(unsigned long call,
+static void lazy_hcall1(unsigned long call,
+		       unsigned long arg1)
+{
+	if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_NONE)
+		kvm_hypercall1(call, arg1);
+	else
+		async_hcall(call, arg1, 0, 0);
+}
+
+static void lazy_hcall2(unsigned long call,
+		       unsigned long arg1,
+		       unsigned long arg2)
+{
+	if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_NONE)
+		kvm_hypercall2(call, arg1, arg2);
+	else
+		async_hcall(call, arg1, arg2, 0);
+}
+
+static void lazy_hcall3(unsigned long call,
 		       unsigned long arg1,
 		       unsigned long arg2,
 		       unsigned long arg3)
 {
 	if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_NONE)
-		hcall(call, arg1, arg2, arg3);
+		kvm_hypercall3(call, arg1, arg2, arg3);
 	else
 		async_hcall(call, arg1, arg2, arg3);
 }
@@ -150,7 +169,7 @@ static void lazy_hcall(unsigned long call,
 static void lguest_leave_lazy_mode(void)
 {
 	paravirt_leave_lazy(paravirt_get_lazy_mode());
-	hcall(LHCALL_FLUSH_ASYNC, 0, 0, 0);
+	kvm_hypercall0(LHCALL_FLUSH_ASYNC);
 }
 
 /*G:033
@@ -229,7 +248,7 @@ static void lguest_write_idt_entry(gate_desc *dt,
 	/* Keep the local copy up to date. */
 	native_write_idt_entry(dt, entrynum, g);
 	/* Tell Host about this new entry. */
-	hcall(LHCALL_LOAD_IDT_ENTRY, entrynum, desc[0], desc[1]);
+	kvm_hypercall3(LHCALL_LOAD_IDT_ENTRY, entrynum, desc[0], desc[1]);
 }
 
 /* Changing to a different IDT is very rare: we keep the IDT up-to-date every
@@ -241,7 +260,7 @@ static void lguest_load_idt(const struct desc_ptr *desc)
 	struct desc_struct *idt = (void *)desc->address;
 
 	for (i = 0; i < (desc->size+1)/8; i++)
-		hcall(LHCALL_LOAD_IDT_ENTRY, i, idt[i].a, idt[i].b);
+		kvm_hypercall3(LHCALL_LOAD_IDT_ENTRY, i, idt[i].a, idt[i].b);
 }
 
 /*
@@ -261,8 +280,8 @@ static void lguest_load_idt(const struct desc_ptr *desc)
  */
 static void lguest_load_gdt(const struct desc_ptr *desc)
 {
-	BUG_ON((desc->size+1)/8 != GDT_ENTRIES);
-	hcall(LHCALL_LOAD_GDT, __pa(desc->address), GDT_ENTRIES, 0);
+	BUG_ON((desc->size + 1) / 8 != GDT_ENTRIES);
+	kvm_hypercall2(LHCALL_LOAD_GDT, __pa(desc->address), GDT_ENTRIES);
 }
 
 /* For a single GDT entry which changes, we do the lazy thing: alter our GDT,
@@ -272,7 +291,7 @@ static void lguest_write_gdt_entry(struct desc_struct *dt, int entrynum,
 				   const void *desc, int type)
 {
 	native_write_gdt_entry(dt, entrynum, desc, type);
-	hcall(LHCALL_LOAD_GDT, __pa(dt), GDT_ENTRIES, 0);
+	kvm_hypercall2(LHCALL_LOAD_GDT, __pa(dt), GDT_ENTRIES);
 }
 
 /* OK, I lied.  There are three "thread local storage" GDT entries which change
@@ -284,7 +303,7 @@ static void lguest_load_tls(struct thread_struct *t, unsigned int cpu)
 	 * can't handle us removing entries we're currently using.  So we clear
 	 * the GS register here: if it's needed it'll be reloaded anyway. */
 	lazy_load_gs(0);
-	lazy_hcall(LHCALL_LOAD_TLS, __pa(&t->tls_array), cpu, 0);
+	lazy_hcall2(LHCALL_LOAD_TLS, __pa(&t->tls_array), cpu);
 }
 
 /*G:038 That's enough excitement for now, back to ploughing through each of
@@ -382,7 +401,7 @@ static void lguest_cpuid(unsigned int *ax, unsigned int *bx,
 static unsigned long current_cr0;
 static void lguest_write_cr0(unsigned long val)
 {
-	lazy_hcall(LHCALL_TS, val & X86_CR0_TS, 0, 0);
+	lazy_hcall1(LHCALL_TS, val & X86_CR0_TS);
 	current_cr0 = val;
 }
 
@@ -396,7 +415,7 @@ static unsigned long lguest_read_cr0(void)
  * the vowels have been optimized out. */
 static void lguest_clts(void)
 {
-	lazy_hcall(LHCALL_TS, 0, 0, 0);
+	lazy_hcall1(LHCALL_TS, 0);
 	current_cr0 &= ~X86_CR0_TS;
 }
 
@@ -418,7 +437,7 @@ static bool cr3_changed = false;
 static void lguest_write_cr3(unsigned long cr3)
 {
 	lguest_data.pgdir = cr3;
-	lazy_hcall(LHCALL_NEW_PGTABLE, cr3, 0, 0);
+	lazy_hcall1(LHCALL_NEW_PGTABLE, cr3);
 	cr3_changed = true;
 }
 
@@ -490,11 +509,17 @@ static void lguest_write_cr4(unsigned long val)
  * into a process' address space.  We set the entry then tell the Host the
  * toplevel and address this corresponds to.  The Guest uses one pagetable per
  * process, so we need to tell the Host which one we're changing (mm->pgd). */
+static void lguest_pte_update(struct mm_struct *mm, unsigned long addr,
+			       pte_t *ptep)
+{
+	lazy_hcall3(LHCALL_SET_PTE, __pa(mm->pgd), addr, ptep->pte_low);
+}
+
 static void lguest_set_pte_at(struct mm_struct *mm, unsigned long addr,
 			      pte_t *ptep, pte_t pteval)
 {
 	*ptep = pteval;
-	lazy_hcall(LHCALL_SET_PTE, __pa(mm->pgd), addr, pteval.pte_low);
+	lguest_pte_update(mm, addr, ptep);
 }
 
 /* The Guest calls this to set a top-level entry.  Again, we set the entry then
@@ -503,8 +528,8 @@ static void lguest_set_pte_at(struct mm_struct *mm, unsigned long addr,
 static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval)
 {
 	*pmdp = pmdval;
-	lazy_hcall(LHCALL_SET_PMD, __pa(pmdp)&PAGE_MASK,
-		   (__pa(pmdp)&(PAGE_SIZE-1))/4, 0);
+	lazy_hcall2(LHCALL_SET_PMD, __pa(pmdp) & PAGE_MASK,
+		   (__pa(pmdp) & (PAGE_SIZE - 1)) / 4);
 }
 
 /* There are a couple of legacy places where the kernel sets a PTE, but we
@@ -520,7 +545,7 @@ static void lguest_set_pte(pte_t *ptep, pte_t pteval)
 {
 	*ptep = pteval;
 	if (cr3_changed)
-		lazy_hcall(LHCALL_FLUSH_TLB, 1, 0, 0);
+		lazy_hcall1(LHCALL_FLUSH_TLB, 1);
 }
 
 /* Unfortunately for Lguest, the pv_mmu_ops for page tables were based on
@@ -536,7 +561,7 @@ static void lguest_set_pte(pte_t *ptep, pte_t pteval)
 static void lguest_flush_tlb_single(unsigned long addr)
 {
 	/* Simply set it to zero: if it was not, it will fault back in. */
-	lazy_hcall(LHCALL_SET_PTE, lguest_data.pgdir, addr, 0);
+	lazy_hcall3(LHCALL_SET_PTE, lguest_data.pgdir, addr, 0);
 }
 
 /* This is what happens after the Guest has removed a large number of entries.
@@ -544,7 +569,7 @@ static void lguest_flush_tlb_single(unsigned long addr)
  * have changed, ie. virtual addresses below PAGE_OFFSET. */
 static void lguest_flush_tlb_user(void)
 {
-	lazy_hcall(LHCALL_FLUSH_TLB, 0, 0, 0);
+	lazy_hcall1(LHCALL_FLUSH_TLB, 0);
 }
 
 /* This is called when the kernel page tables have changed.  That's not very
@@ -552,7 +577,7 @@ static void lguest_flush_tlb_user(void)
  * slow), so it's worth separating this from the user flushing above. */
 static void lguest_flush_tlb_kernel(void)
 {
-	lazy_hcall(LHCALL_FLUSH_TLB, 1, 0, 0);
+	lazy_hcall1(LHCALL_FLUSH_TLB, 1);
 }
 
 /*
@@ -689,7 +714,7 @@ static int lguest_clockevent_set_next_event(unsigned long delta,
 	}
 
 	/* Please wake us this far in the future. */
-	hcall(LHCALL_SET_CLOCKEVENT, delta, 0, 0);
+	kvm_hypercall1(LHCALL_SET_CLOCKEVENT, delta);
 	return 0;
 }
 
@@ -700,7 +725,7 @@ static void lguest_clockevent_set_mode(enum clock_event_mode mode,
 	case CLOCK_EVT_MODE_UNUSED:
 	case CLOCK_EVT_MODE_SHUTDOWN:
 		/* A 0 argument shuts the clock down. */
-		hcall(LHCALL_SET_CLOCKEVENT, 0, 0, 0);
+		kvm_hypercall0(LHCALL_SET_CLOCKEVENT);
 		break;
 	case CLOCK_EVT_MODE_ONESHOT:
 		/* This is what we expect. */
@@ -775,8 +800,8 @@ static void lguest_time_init(void)
 static void lguest_load_sp0(struct tss_struct *tss,
 			    struct thread_struct *thread)
 {
-	lazy_hcall(LHCALL_SET_STACK, __KERNEL_DS|0x1, thread->sp0,
-		   THREAD_SIZE/PAGE_SIZE);
+	lazy_hcall3(LHCALL_SET_STACK, __KERNEL_DS | 0x1, thread->sp0,
+		   THREAD_SIZE / PAGE_SIZE);
 }
 
 /* Let's just say, I wouldn't do debugging under a Guest. */
@@ -849,7 +874,7 @@ static void set_lguest_basic_apic_ops(void)
 /* STOP!  Until an interrupt comes in. */
 static void lguest_safe_halt(void)
 {
-	hcall(LHCALL_HALT, 0, 0, 0);
+	kvm_hypercall0(LHCALL_HALT);
 }
 
 /* The SHUTDOWN hypercall takes a string to describe what's happening, and
@@ -859,7 +884,8 @@ static void lguest_safe_halt(void)
  * rather than virtual addresses, so we use __pa() here. */
 static void lguest_power_off(void)
 {
-	hcall(LHCALL_SHUTDOWN, __pa("Power down"), LGUEST_SHUTDOWN_POWEROFF, 0);
+	kvm_hypercall2(LHCALL_SHUTDOWN, __pa("Power down"),
+					LGUEST_SHUTDOWN_POWEROFF);
 }
 
 /*
@@ -869,7 +895,7 @@ static void lguest_power_off(void)
  */
 static int lguest_panic(struct notifier_block *nb, unsigned long l, void *p)
 {
-	hcall(LHCALL_SHUTDOWN, __pa(p), LGUEST_SHUTDOWN_POWEROFF, 0);
+	kvm_hypercall2(LHCALL_SHUTDOWN, __pa(p), LGUEST_SHUTDOWN_POWEROFF);
 	/* The hcall won't return, but to keep gcc happy, we're "done". */
 	return NOTIFY_DONE;
 }
@@ -910,7 +936,7 @@ static __init int early_put_chars(u32 vtermno, const char *buf, int count)
 		len = sizeof(scratch) - 1;
 	scratch[len] = '\0';
 	memcpy(scratch, buf, len);
-	hcall(LHCALL_NOTIFY, __pa(scratch), 0, 0);
+	kvm_hypercall1(LHCALL_NOTIFY, __pa(scratch));
 
 	/* This routine returns the number of bytes actually written. */
 	return len;
@@ -920,7 +946,7 @@ static __init int early_put_chars(u32 vtermno, const char *buf, int count)
  * Launcher to reboot us. */
 static void lguest_restart(char *reason)
 {
-	hcall(LHCALL_SHUTDOWN, __pa(reason), LGUEST_SHUTDOWN_RESTART, 0);
+	kvm_hypercall2(LHCALL_SHUTDOWN, __pa(reason), LGUEST_SHUTDOWN_RESTART);
 }
 
 /*G:050
@@ -1040,6 +1066,8 @@ __init void lguest_init(void)
 	pv_mmu_ops.read_cr3 = lguest_read_cr3;
 	pv_mmu_ops.lazy_mode.enter = paravirt_enter_lazy_mmu;
 	pv_mmu_ops.lazy_mode.leave = lguest_leave_lazy_mode;
+	pv_mmu_ops.pte_update = lguest_pte_update;
+	pv_mmu_ops.pte_update_defer = lguest_pte_update;
 
 #ifdef CONFIG_X86_LOCAL_APIC
 	/* apic read/write intercepts */
diff --git a/arch/x86/lguest/i386_head.S b/arch/x86/lguest/i386_head.S
index 10b9bd3..f795419 100644
--- a/arch/x86/lguest/i386_head.S
+++ b/arch/x86/lguest/i386_head.S
@@ -27,8 +27,8 @@ ENTRY(lguest_entry)
 	/* We make the "initialization" hypercall now to tell the Host about
 	 * us, and also find out where it put our page tables. */
 	movl $LHCALL_LGUEST_INIT, %eax
-	movl $lguest_data - __PAGE_OFFSET, %edx
-	int $LGUEST_TRAP_ENTRY
+	movl $lguest_data - __PAGE_OFFSET, %ebx
+	.byte 0x0f,0x01,0xc1 /* KVM_HYPERCALL */
 
 	/* Set up the initial stack so we can run C code. */
 	movl $(init_thread_union+THREAD_SIZE),%esp
diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
index 522db5e..5bc5d16 100644
--- a/arch/x86/mm/highmem_32.c
+++ b/arch/x86/mm/highmem_32.c
@@ -19,49 +19,6 @@ void kunmap(struct page *page)
 	kunmap_high(page);
 }
 
-static void debug_kmap_atomic_prot(enum km_type type)
-{
-#ifdef CONFIG_DEBUG_HIGHMEM
-	static unsigned warn_count = 10;
-
-	if (unlikely(warn_count == 0))
-		return;
-
-	if (unlikely(in_interrupt())) {
-		if (in_irq()) {
-			if (type != KM_IRQ0 && type != KM_IRQ1 &&
-			    type != KM_BIO_SRC_IRQ && type != KM_BIO_DST_IRQ &&
-			    type != KM_BOUNCE_READ) {
-				WARN_ON(1);
-				warn_count--;
-			}
-		} else if (!irqs_disabled()) {	/* softirq */
-			if (type != KM_IRQ0 && type != KM_IRQ1 &&
-			    type != KM_SOFTIRQ0 && type != KM_SOFTIRQ1 &&
-			    type != KM_SKB_SUNRPC_DATA &&
-			    type != KM_SKB_DATA_SOFTIRQ &&
-			    type != KM_BOUNCE_READ) {
-				WARN_ON(1);
-				warn_count--;
-			}
-		}
-	}
-
-	if (type == KM_IRQ0 || type == KM_IRQ1 || type == KM_BOUNCE_READ ||
-			type == KM_BIO_SRC_IRQ || type == KM_BIO_DST_IRQ) {
-		if (!irqs_disabled()) {
-			WARN_ON(1);
-			warn_count--;
-		}
-	} else if (type == KM_SOFTIRQ0 || type == KM_SOFTIRQ1) {
-		if (irq_count() == 0 && !irqs_disabled()) {
-			WARN_ON(1);
-			warn_count--;
-		}
-	}
-#endif
-}
-
 /*
  * kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
  * no global lock is needed and because the kmap code must perform a global TLB
@@ -81,8 +38,9 @@ void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot)
 	if (!PageHighMem(page))
 		return page_address(page);
 
-	debug_kmap_atomic_prot(type);
+	debug_kmap_atomic(type);
 
+	debug_kmap_atomic(type);
 	idx = type + KM_TYPE_NR*smp_processor_id();
 	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
 	BUG_ON(!pte_none(*(kmap_pte-idx)));
diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c
index 699c9b2..bff0c90 100644
--- a/arch/x86/mm/iomap_32.c
+++ b/arch/x86/mm/iomap_32.c
@@ -19,6 +19,7 @@
 #include <asm/iomap.h>
 #include <asm/pat.h>
 #include <linux/module.h>
+#include <linux/highmem.h>
 
 int is_io_mapping_possible(resource_size_t base, unsigned long size)
 {
@@ -71,6 +72,7 @@ iounmap_atomic(void *kvaddr, enum km_type type)
 	unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
 	enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id();
 
+	debug_kmap_atomic(type);
 	/*
 	 * Force other mappings to Oops if they'll try to access this pte
 	 * without first remap it.  Keeping stale mappings around is a bad idea
diff --git a/arch/x86/pci/early.c b/arch/x86/pci/early.c
index f6adf2c..aaf26ae 100644
--- a/arch/x86/pci/early.c
+++ b/arch/x86/pci/early.c
@@ -69,11 +69,12 @@ void early_dump_pci_device(u8 bus, u8 slot, u8 func)
 	int j;
 	u32 val;
 
-	printk(KERN_INFO "PCI: %02x:%02x:%02x", bus, slot, func);
+	printk(KERN_INFO "pci 0000:%02x:%02x.%d config space:",
+	       bus, slot, func);
 
 	for (i = 0; i < 256; i += 4) {
 		if (!(i & 0x0f))
-			printk("\n%04x:",i);
+			printk("\n  %02x:",i);
 
 		val = read_pci_config(bus, slot, func, i);
 		for (j = 0; j < 4; j++) {
@@ -96,20 +97,22 @@ void early_dump_pci_devices(void)
 			for (func = 0; func < 8; func++) {
 				u32 class;
 				u8 type;
+
 				class = read_pci_config(bus, slot, func,
 							PCI_CLASS_REVISION);
 				if (class == 0xffffffff)
-					break;
+					continue;
 
 				early_dump_pci_device(bus, slot, func);
 
-				/* No multi-function device? */
-				type = read_pci_config_byte(bus, slot, func,
+				if (func == 0) {
+					type = read_pci_config_byte(bus, slot,
+								    func,
 							       PCI_HEADER_TYPE);
-				if (!(type & 0x80))
-					break;
+					if (!(type & 0x80))
+						break;
+				}
 			}
 		}
 	}
 }
-
diff --git a/arch/x86/pci/fixup.c b/arch/x86/pci/fixup.c
index 9c49919..6dd8955 100644
--- a/arch/x86/pci/fixup.c
+++ b/arch/x86/pci/fixup.c
@@ -495,26 +495,6 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SIEMENS, 0x0015,
 			  pci_siemens_interrupt_controller);
 
 /*
- * Regular PCI devices have 256 bytes, but AMD Family 10h/11h CPUs have
- * 4096 bytes configuration space for each function of their processor
- * configuration space.
- */
-static void amd_cpu_pci_cfg_space_size(struct pci_dev *dev)
-{
-	dev->cfg_size = pci_cfg_space_size_ext(dev);
-}
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1200, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1201, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1202, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1203, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1204, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1300, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1301, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1302, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1303, amd_cpu_pci_cfg_space_size);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, 0x1304, amd_cpu_pci_cfg_space_size);
-
-/*
  * SB600: Disable BAR1 on device 14.0 to avoid HPET resources from
  * confusing the PCI engine:
  */
diff --git a/arch/x86/pci/legacy.c b/arch/x86/pci/legacy.c
index f1065b1..4061bb0 100644
--- a/arch/x86/pci/legacy.c
+++ b/arch/x86/pci/legacy.c
@@ -50,8 +50,6 @@ static int __init pci_legacy_init(void)
 	if (pci_root_bus)
 		pci_bus_add_devices(pci_root_bus);
 
-	pcibios_fixup_peer_bridges();
-
 	return 0;
 }
 
@@ -67,6 +65,7 @@ int __init pci_subsys_init(void)
 	pci_visws_init();
 #endif
 	pci_legacy_init();
+	pcibios_fixup_peer_bridges();
 	pcibios_irq_init();
 	pcibios_init();
 
diff --git a/arch/x86/pci/mmconfig-shared.c b/arch/x86/pci/mmconfig-shared.c
index 89bf924..905bb52 100644
--- a/arch/x86/pci/mmconfig-shared.c
+++ b/arch/x86/pci/mmconfig-shared.c
@@ -14,6 +14,7 @@
 #include <linux/init.h>
 #include <linux/acpi.h>
 #include <linux/bitmap.h>
+#include <linux/sort.h>
 #include <asm/e820.h>
 #include <asm/pci_x86.h>
 
@@ -24,24 +25,49 @@
 /* Indicate if the mmcfg resources have been placed into the resource table. */
 static int __initdata pci_mmcfg_resources_inserted;
 
+static __init int extend_mmcfg(int num)
+{
+	struct acpi_mcfg_allocation *new;
+	int new_num = pci_mmcfg_config_num + num;
+
+	new = kzalloc(sizeof(pci_mmcfg_config[0]) * new_num, GFP_KERNEL);
+	if (!new)
+		return -1;
+
+	if (pci_mmcfg_config) {
+		memcpy(new, pci_mmcfg_config,
+			 sizeof(pci_mmcfg_config[0]) * new_num);
+		kfree(pci_mmcfg_config);
+	}
+	pci_mmcfg_config = new;
+
+	return 0;
+}
+
+static __init void fill_one_mmcfg(u64 addr, int segment, int start, int end)
+{
+	int i = pci_mmcfg_config_num;
+
+	pci_mmcfg_config_num++;
+	pci_mmcfg_config[i].address = addr;
+	pci_mmcfg_config[i].pci_segment = segment;
+	pci_mmcfg_config[i].start_bus_number = start;
+	pci_mmcfg_config[i].end_bus_number = end;
+}
+
 static const char __init *pci_mmcfg_e7520(void)
 {
 	u32 win;
 	raw_pci_ops->read(0, 0, PCI_DEVFN(0, 0), 0xce, 2, &win);
 
 	win = win & 0xf000;
-	if(win == 0x0000 || win == 0xf000)
-		pci_mmcfg_config_num = 0;
-	else {
-		pci_mmcfg_config_num = 1;
-		pci_mmcfg_config = kzalloc(sizeof(pci_mmcfg_config[0]), GFP_KERNEL);
-		if (!pci_mmcfg_config)
-			return NULL;
-		pci_mmcfg_config[0].address = win << 16;
-		pci_mmcfg_config[0].pci_segment = 0;
-		pci_mmcfg_config[0].start_bus_number = 0;
-		pci_mmcfg_config[0].end_bus_number = 255;
-	}
+	if (win == 0x0000 || win == 0xf000)
+		return NULL;
+
+	if (extend_mmcfg(1) == -1)
+		return NULL;
+
+	fill_one_mmcfg(win << 16, 0, 0, 255);
 
 	return "Intel Corporation E7520 Memory Controller Hub";
 }
@@ -50,13 +76,11 @@ static const char __init *pci_mmcfg_intel_945(void)
 {
 	u32 pciexbar, mask = 0, len = 0;
 
-	pci_mmcfg_config_num = 1;
-
 	raw_pci_ops->read(0, 0, PCI_DEVFN(0, 0), 0x48, 4, &pciexbar);
 
 	/* Enable bit */
 	if (!(pciexbar & 1))
-		pci_mmcfg_config_num = 0;
+		return NULL;
 
 	/* Size bits */
 	switch ((pciexbar >> 1) & 3) {
@@ -73,28 +97,23 @@ static const char __init *pci_mmcfg_intel_945(void)
 		len  = 0x04000000U;
 		break;
 	default:
-		pci_mmcfg_config_num = 0;
+		return NULL;
 	}
 
 	/* Errata #2, things break when not aligned on a 256Mb boundary */
 	/* Can only happen in 64M/128M mode */
 
 	if ((pciexbar & mask) & 0x0fffffffU)
-		pci_mmcfg_config_num = 0;
+		return NULL;
 
 	/* Don't hit the APIC registers and their friends */
 	if ((pciexbar & mask) >= 0xf0000000U)
-		pci_mmcfg_config_num = 0;
-
-	if (pci_mmcfg_config_num) {
-		pci_mmcfg_config = kzalloc(sizeof(pci_mmcfg_config[0]), GFP_KERNEL);
-		if (!pci_mmcfg_config)
-			return NULL;
-		pci_mmcfg_config[0].address = pciexbar & mask;
-		pci_mmcfg_config[0].pci_segment = 0;
-		pci_mmcfg_config[0].start_bus_number = 0;
-		pci_mmcfg_config[0].end_bus_number = (len >> 20) - 1;
-	}
+		return NULL;
+
+	if (extend_mmcfg(1) == -1)
+		return NULL;
+
+	fill_one_mmcfg(pciexbar & mask, 0, 0, (len >> 20) - 1);
 
 	return "Intel Corporation 945G/GZ/P/PL Express Memory Controller Hub";
 }
@@ -138,22 +157,77 @@ static const char __init *pci_mmcfg_amd_fam10h(void)
 		busnbits = 8;
 	}
 
-	pci_mmcfg_config_num = (1 << segnbits);
-	pci_mmcfg_config = kzalloc(sizeof(pci_mmcfg_config[0]) *
-				   pci_mmcfg_config_num, GFP_KERNEL);
-	if (!pci_mmcfg_config)
+	if (extend_mmcfg(1 << segnbits) == -1)
 		return NULL;
 
-	for (i = 0; i < (1 << segnbits); i++) {
-		pci_mmcfg_config[i].address = base + (1<<28) * i;
-		pci_mmcfg_config[i].pci_segment = i;
-		pci_mmcfg_config[i].start_bus_number = 0;
-		pci_mmcfg_config[i].end_bus_number = (1 << busnbits) - 1;
-	}
+	for (i = 0; i < (1 << segnbits); i++)
+		fill_one_mmcfg(base + (1<<28) * i, i, 0, (1 << busnbits) - 1);
 
 	return "AMD Family 10h NB";
 }
 
+static bool __initdata mcp55_checked;
+static const char __init *pci_mmcfg_nvidia_mcp55(void)
+{
+	int bus;
+	int mcp55_mmconf_found = 0;
+
+	static const u32 extcfg_regnum		= 0x90;
+	static const u32 extcfg_regsize		= 4;
+	static const u32 extcfg_enable_mask	= 1<<31;
+	static const u32 extcfg_start_mask	= 0xff<<16;
+	static const int extcfg_start_shift	= 16;
+	static const u32 extcfg_size_mask	= 0x3<<28;
+	static const int extcfg_size_shift	= 28;
+	static const int extcfg_sizebus[]	= {0x100, 0x80, 0x40, 0x20};
+	static const u32 extcfg_base_mask[]	= {0x7ff8, 0x7ffc, 0x7ffe, 0x7fff};
+	static const int extcfg_base_lshift	= 25;
+
+	/*
+	 * do check if amd fam10h already took over
+	 */
+	if (!acpi_disabled || pci_mmcfg_config_num || mcp55_checked)
+		return NULL;
+
+	mcp55_checked = true;
+	for (bus = 0; bus < 256; bus++) {
+		u64 base;
+		u32 l, extcfg;
+		u16 vendor, device;
+		int start, size_index, end;
+
+		raw_pci_ops->read(0, bus, PCI_DEVFN(0, 0), 0, 4, &l);
+		vendor = l & 0xffff;
+		device = (l >> 16) & 0xffff;
+
+		if (PCI_VENDOR_ID_NVIDIA != vendor || 0x0369 != device)
+			continue;
+
+		raw_pci_ops->read(0, bus, PCI_DEVFN(0, 0), extcfg_regnum,
+				  extcfg_regsize, &extcfg);
+
+		if (!(extcfg & extcfg_enable_mask))
+			continue;
+
+		if (extend_mmcfg(1) == -1)
+			continue;
+
+		size_index = (extcfg & extcfg_size_mask) >> extcfg_size_shift;
+		base = extcfg & extcfg_base_mask[size_index];
+		/* base could > 4G */
+		base <<= extcfg_base_lshift;
+		start = (extcfg & extcfg_start_mask) >> extcfg_start_shift;
+		end = start + extcfg_sizebus[size_index] - 1;
+		fill_one_mmcfg(base, 0, start, end);
+		mcp55_mmconf_found++;
+	}
+
+	if (!mcp55_mmconf_found)
+		return NULL;
+
+	return "nVidia MCP55";
+}
+
 struct pci_mmcfg_hostbridge_probe {
 	u32 bus;
 	u32 devfn;
@@ -171,8 +245,52 @@ static struct pci_mmcfg_hostbridge_probe pci_mmcfg_probes[] __initdata = {
 	  0x1200, pci_mmcfg_amd_fam10h },
 	{ 0xff, PCI_DEVFN(0, 0), PCI_VENDOR_ID_AMD,
 	  0x1200, pci_mmcfg_amd_fam10h },
+	{ 0, PCI_DEVFN(0, 0), PCI_VENDOR_ID_NVIDIA,
+	  0x0369, pci_mmcfg_nvidia_mcp55 },
 };
 
+static int __init cmp_mmcfg(const void *x1, const void *x2)
+{
+	const typeof(pci_mmcfg_config[0]) *m1 = x1;
+	const typeof(pci_mmcfg_config[0]) *m2 = x2;
+	int start1, start2;
+
+	start1 = m1->start_bus_number;
+	start2 = m2->start_bus_number;
+
+	return start1 - start2;
+}
+
+static void __init pci_mmcfg_check_end_bus_number(void)
+{
+	int i;
+	typeof(pci_mmcfg_config[0]) *cfg, *cfgx;
+
+	/* sort them at first */
+	sort(pci_mmcfg_config, pci_mmcfg_config_num,
+		 sizeof(pci_mmcfg_config[0]), cmp_mmcfg, NULL);
+
+	/* last one*/
+	if (pci_mmcfg_config_num > 0) {
+		i = pci_mmcfg_config_num - 1;
+		cfg = &pci_mmcfg_config[i];
+		if (cfg->end_bus_number < cfg->start_bus_number)
+			cfg->end_bus_number = 255;
+	}
+
+	/* don't overlap please */
+	for (i = 0; i < pci_mmcfg_config_num - 1; i++) {
+		cfg = &pci_mmcfg_config[i];
+		cfgx = &pci_mmcfg_config[i+1];
+
+		if (cfg->end_bus_number < cfg->start_bus_number)
+			cfg->end_bus_number = 255;
+
+		if (cfg->end_bus_number >= cfgx->start_bus_number)
+			cfg->end_bus_number = cfgx->start_bus_number - 1;
+	}
+}
+
 static int __init pci_mmcfg_check_hostbridge(void)
 {
 	u32 l;
@@ -186,31 +304,33 @@ static int __init pci_mmcfg_check_hostbridge(void)
 
 	pci_mmcfg_config_num = 0;
 	pci_mmcfg_config = NULL;
-	name = NULL;
 
-	for (i = 0; !name && i < ARRAY_SIZE(pci_mmcfg_probes); i++) {
+	for (i = 0; i < ARRAY_SIZE(pci_mmcfg_probes); i++) {
 		bus =  pci_mmcfg_probes[i].bus;
 		devfn = pci_mmcfg_probes[i].devfn;
 		raw_pci_ops->read(0, bus, devfn, 0, 4, &l);
 		vendor = l & 0xffff;
 		device = (l >> 16) & 0xffff;
 
+		name = NULL;
 		if (pci_mmcfg_probes[i].vendor == vendor &&
 		    pci_mmcfg_probes[i].device == device)
 			name = pci_mmcfg_probes[i].probe();
-	}
 
-	if (name) {
-		printk(KERN_INFO "PCI: Found %s %s MMCONFIG support.\n",
-		       name, pci_mmcfg_config_num ? "with" : "without");
+		if (name)
+			printk(KERN_INFO "PCI: Found %s with MMCONFIG support.\n",
+			       name);
 	}
 
-	return name != NULL;
+	/* some end_bus_number is crazy, fix it */
+	pci_mmcfg_check_end_bus_number();
+
+	return pci_mmcfg_config_num != 0;
 }
 
 static void __init pci_mmcfg_insert_resources(void)
 {
-#define PCI_MMCFG_RESOURCE_NAME_LEN 19
+#define PCI_MMCFG_RESOURCE_NAME_LEN 24
 	int i;
 	struct resource *res;
 	char *names;
@@ -228,9 +348,10 @@ static void __init pci_mmcfg_insert_resources(void)
 		struct acpi_mcfg_allocation *cfg = &pci_mmcfg_config[i];
 		num_buses = cfg->end_bus_number - cfg->start_bus_number + 1;
 		res->name = names;
-		snprintf(names, PCI_MMCFG_RESOURCE_NAME_LEN, "PCI MMCONFIG %u",
-			 cfg->pci_segment);
-		res->start = cfg->address;
+		snprintf(names, PCI_MMCFG_RESOURCE_NAME_LEN,
+			 "PCI MMCONFIG %u [%02x-%02x]", cfg->pci_segment,
+			 cfg->start_bus_number, cfg->end_bus_number);
+		res->start = cfg->address + (cfg->start_bus_number << 20);
 		res->end = res->start + (num_buses << 20) - 1;
 		res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
 		insert_resource(&iomem_resource, res);
@@ -354,8 +475,6 @@ static void __init pci_mmcfg_reject_broken(int early)
 	    (pci_mmcfg_config[0].address == 0))
 		return;
 
-	cfg = &pci_mmcfg_config[0];
-
 	for (i = 0; i < pci_mmcfg_config_num; i++) {
 		int valid = 0;
 		u64 addr, size;
@@ -423,10 +542,10 @@ static void __init __pci_mmcfg_init(int early)
 			known_bridge = 1;
 	}
 
-	if (!known_bridge) {
+	if (!known_bridge)
 		acpi_table_parse(ACPI_SIG_MCFG, acpi_parse_mcfg);
-		pci_mmcfg_reject_broken(early);
-	}
+
+	pci_mmcfg_reject_broken(early);
 
 	if ((pci_mmcfg_config_num == 0) ||
 	    (pci_mmcfg_config == NULL) ||
diff --git a/arch/x86/pci/mmconfig_64.c b/arch/x86/pci/mmconfig_64.c
index 30007ff..94349f8 100644
--- a/arch/x86/pci/mmconfig_64.c
+++ b/arch/x86/pci/mmconfig_64.c
@@ -112,13 +112,18 @@ static struct pci_raw_ops pci_mmcfg = {
 static void __iomem * __init mcfg_ioremap(struct acpi_mcfg_allocation *cfg)
 {
 	void __iomem *addr;
-	u32 size;
-
-	size = (cfg->end_bus_number + 1) << 20;
-	addr = ioremap_nocache(cfg->address, size);
+	u64 start, size;
+
+	start = cfg->start_bus_number;
+	start <<= 20;
+	start += cfg->address;
+	size = cfg->end_bus_number + 1 - cfg->start_bus_number;
+	size <<= 20;
+	addr = ioremap_nocache(start, size);
 	if (addr) {
 		printk(KERN_INFO "PCI: Using MMCONFIG at %Lx - %Lx\n",
-		       cfg->address, cfg->address + size - 1);
+		       start, start + size - 1);
+		addr -= cfg->start_bus_number << 20;
 	}
 	return addr;
 }
@@ -157,7 +162,7 @@ void __init pci_mmcfg_arch_free(void)
 
 	for (i = 0; i < pci_mmcfg_config_num; ++i) {
 		if (pci_mmcfg_virt[i].virt) {
-			iounmap(pci_mmcfg_virt[i].virt);
+			iounmap(pci_mmcfg_virt[i].virt + (pci_mmcfg_virt[i].cfg->start_bus_number << 20));
 			pci_mmcfg_virt[i].virt = NULL;
 			pci_mmcfg_virt[i].cfg = NULL;
 		}
diff --git a/arch/x86/power/cpu_32.c b/arch/x86/power/cpu_32.c
index 274d060..ce702c5 100644
--- a/arch/x86/power/cpu_32.c
+++ b/arch/x86/power/cpu_32.c
@@ -12,6 +12,7 @@
 #include <asm/mtrr.h>
 #include <asm/mce.h>
 #include <asm/xcr.h>
+#include <asm/suspend.h>
 
 static struct saved_context saved_context;
 
diff --git a/arch/x86/power/cpu_64.c b/arch/x86/power/cpu_64.c
index e3b6cf7..5343540 100644
--- a/arch/x86/power/cpu_64.c
+++ b/arch/x86/power/cpu_64.c
@@ -15,6 +15,7 @@
 #include <asm/pgtable.h>
 #include <asm/mtrr.h>
 #include <asm/xcr.h>
+#include <asm/suspend.h>
 
 static void fix_processor_context(void);
 
diff --git a/arch/x86/power/hibernate_64.c b/arch/x86/power/hibernate_64.c
index 6dd000d..65fdc86 100644
--- a/arch/x86/power/hibernate_64.c
+++ b/arch/x86/power/hibernate_64.c
@@ -14,6 +14,7 @@
 #include <asm/page.h>
 #include <asm/pgtable.h>
 #include <asm/mtrr.h>
+#include <asm/suspend.h>
 
 /* References to section boundaries */
 extern const void __nosave_begin, __nosave_end;
diff --git a/arch/xtensa/platforms/iss/console.c b/arch/xtensa/platforms/iss/console.c
index 25d46c8..4c559cf 100644
--- a/arch/xtensa/platforms/iss/console.c
+++ b/arch/xtensa/platforms/iss/console.c
@@ -18,6 +18,7 @@
 #include <linux/mm.h>
 #include <linux/major.h>
 #include <linux/param.h>
+#include <linux/seq_file.h>
 #include <linux/serial.h>
 #include <linux/serialP.h>
 
@@ -176,22 +177,24 @@ static void rs_wait_until_sent(struct tty_struct *tty, int timeout)
 	/* Stub, once again.. */
 }
 
-static int rs_read_proc(char *page, char **start, off_t off, int count,
-			int *eof, void *data)
+static int rs_proc_show(struct seq_file *m, void *v)
 {
-	int len = 0;
-	off_t begin = 0;
-
-	len += sprintf(page, "serinfo:1.0 driver:%s\n", serial_version);
-	*eof = 1;
-
-	if (off >= len + begin)
-		return 0;
+	seq_printf(m, "serinfo:1.0 driver:%s\n", serial_version);
+	return 0;
+}
 
-	*start = page + (off - begin);
-	return ((count < begin + len - off) ? count : begin + len - off);
+static int rs_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, rs_proc_show, NULL);
 }
 
+static const struct file_operations rs_proc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= rs_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
 
 static struct tty_operations serial_ops = {
 	.open = rs_open,
@@ -203,7 +206,7 @@ static struct tty_operations serial_ops = {
 	.chars_in_buffer = rs_chars_in_buffer,
 	.hangup = rs_hangup,
 	.wait_until_sent = rs_wait_until_sent,
-	.read_proc = rs_read_proc
+	.proc_fops = &rs_proc_fops,
 };
 
 int __init rs_init(void)