Merge branch 'linus' into locking/core, to pick up fixes

Signed-off-by: Ingo Molnar <mingo@kernel.org>

32 files changed, 1400 insertions, 723 deletions

diff --git a/arch/sparc/include/asm/cpudata_64.h b/arch/sparc/include/asm/cpudata_64.h
index a6cfdabb6054..5b0ed48e5b0c 100644
--- a/arch/sparc/include/asm/cpudata_64.h
+++ b/arch/sparc/include/asm/cpudata_64.h
@@ -24,9 +24,10 @@ typedef struct {
 	unsigned int	icache_line_size;
 	unsigned int	ecache_size;
 	unsigned int	ecache_line_size;
-	unsigned short	sock_id;
+	unsigned short	sock_id;	/* physical package */
 	unsigned short	core_id;
-	int		proc_id;
+	unsigned short  max_cache_id;	/* groupings of highest shared cache */
+	unsigned short	proc_id;	/* strand (aka HW thread) id */
 } cpuinfo_sparc;
 
 DECLARE_PER_CPU(cpuinfo_sparc, __cpu_data);
diff --git a/arch/sparc/include/asm/spinlock_32.h b/arch/sparc/include/asm/spinlock_32.h
index d9c5876c6121..8011e79f59c9 100644
--- a/arch/sparc/include/asm/spinlock_32.h
+++ b/arch/sparc/include/asm/spinlock_32.h
@@ -134,7 +134,7 @@ static inline void arch_write_lock(arch_rwlock_t *rw)
 	*(volatile __u32 *)&lp->lock = ~0U;
 }
 
-static void inline arch_write_unlock(arch_rwlock_t *lock)
+static inline void arch_write_unlock(arch_rwlock_t *lock)
 {
 	__asm__ __volatile__(
 "	st		%%g0, [%0]"
diff --git a/arch/sparc/include/asm/spinlock_64.h b/arch/sparc/include/asm/spinlock_64.h
index 87990b7c6b0d..07c9f2e9bf57 100644
--- a/arch/sparc/include/asm/spinlock_64.h
+++ b/arch/sparc/include/asm/spinlock_64.h
@@ -96,7 +96,7 @@ static inline void arch_spin_lock_flags(arch_spinlock_t *lock, unsigned long fla
 
 /* Multi-reader locks, these are much saner than the 32-bit Sparc ones... */
 
-static void inline arch_read_lock(arch_rwlock_t *lock)
+static inline void arch_read_lock(arch_rwlock_t *lock)
 {
 	unsigned long tmp1, tmp2;
 
@@ -119,7 +119,7 @@ static void inline arch_read_lock(arch_rwlock_t *lock)
 	: "memory");
 }
 
-static int inline arch_read_trylock(arch_rwlock_t *lock)
+static inline int arch_read_trylock(arch_rwlock_t *lock)
 {
 	int tmp1, tmp2;
 
@@ -140,7 +140,7 @@ static int inline arch_read_trylock(arch_rwlock_t *lock)
 	return tmp1;
 }
 
-static void inline arch_read_unlock(arch_rwlock_t *lock)
+static inline void arch_read_unlock(arch_rwlock_t *lock)
 {
 	unsigned long tmp1, tmp2;
 
@@ -156,7 +156,7 @@ static void inline arch_read_unlock(arch_rwlock_t *lock)
 	: "memory");
 }
 
-static void inline arch_write_lock(arch_rwlock_t *lock)
+static inline void arch_write_lock(arch_rwlock_t *lock)
 {
 	unsigned long mask, tmp1, tmp2;
 
@@ -181,7 +181,7 @@ static void inline arch_write_lock(arch_rwlock_t *lock)
 	: "memory");
 }
 
-static void inline arch_write_unlock(arch_rwlock_t *lock)
+static inline void arch_write_unlock(arch_rwlock_t *lock)
 {
 	__asm__ __volatile__(
 "	stw		%%g0, [%0]"
@@ -190,7 +190,7 @@ static void inline arch_write_unlock(arch_rwlock_t *lock)
 	: "memory");
 }
 
-static int inline arch_write_trylock(arch_rwlock_t *lock)
+static inline int arch_write_trylock(arch_rwlock_t *lock)
 {
 	unsigned long mask, tmp1, tmp2, result;
 
diff --git a/arch/sparc/include/asm/topology_64.h b/arch/sparc/include/asm/topology_64.h
index bec481aaca16..7b4898a36eee 100644
--- a/arch/sparc/include/asm/topology_64.h
+++ b/arch/sparc/include/asm/topology_64.h
@@ -44,14 +44,20 @@ int __node_distance(int, int);
 #define topology_physical_package_id(cpu)	(cpu_data(cpu).proc_id)
 #define topology_core_id(cpu)			(cpu_data(cpu).core_id)
 #define topology_core_cpumask(cpu)		(&cpu_core_sib_map[cpu])
+#define topology_core_cache_cpumask(cpu)	(&cpu_core_sib_cache_map[cpu])
 #define topology_sibling_cpumask(cpu)		(&per_cpu(cpu_sibling_map, cpu))
 #endif /* CONFIG_SMP */
 
 extern cpumask_t cpu_core_map[NR_CPUS];
 extern cpumask_t cpu_core_sib_map[NR_CPUS];
+extern cpumask_t cpu_core_sib_cache_map[NR_CPUS];
+
+/**
+ * Return cores that shares the last level cache.
+ */
 static inline const struct cpumask *cpu_coregroup_mask(int cpu)
 {
-        return &cpu_core_map[cpu];
+	return &cpu_core_sib_cache_map[cpu];
 }
 
 #endif /* _ASM_SPARC64_TOPOLOGY_H */
diff --git a/arch/sparc/include/asm/uaccess_64.h b/arch/sparc/include/asm/uaccess_64.h
index b68acc563235..5373136c412b 100644
--- a/arch/sparc/include/asm/uaccess_64.h
+++ b/arch/sparc/include/asm/uaccess_64.h
@@ -82,7 +82,6 @@ static inline int access_ok(int type, const void __user * addr, unsigned long si
 	return 1;
 }
 
-void __ret_efault(void);
 void __retl_efault(void);
 
 /* Uh, these should become the main single-value transfer routines..
@@ -189,55 +188,34 @@ int __get_user_bad(void);
 unsigned long __must_check ___copy_from_user(void *to,
 					     const void __user *from,
 					     unsigned long size);
-unsigned long copy_from_user_fixup(void *to, const void __user *from,
-				   unsigned long size);
 static inline unsigned long __must_check
 copy_from_user(void *to, const void __user *from, unsigned long size)
 {
-	unsigned long ret;
-
 	check_object_size(to, size, false);
 
-	ret = ___copy_from_user(to, from, size);
-	if (unlikely(ret))
-		ret = copy_from_user_fixup(to, from, size);
-
-	return ret;
+	return ___copy_from_user(to, from, size);
 }
 #define __copy_from_user copy_from_user
 
 unsigned long __must_check ___copy_to_user(void __user *to,
 					   const void *from,
 					   unsigned long size);
-unsigned long copy_to_user_fixup(void __user *to, const void *from,
-				 unsigned long size);
 static inline unsigned long __must_check
 copy_to_user(void __user *to, const void *from, unsigned long size)
 {
-	unsigned long ret;
-
 	check_object_size(from, size, true);
 
-	ret = ___copy_to_user(to, from, size);
-	if (unlikely(ret))
-		ret = copy_to_user_fixup(to, from, size);
-	return ret;
+	return ___copy_to_user(to, from, size);
 }
 #define __copy_to_user copy_to_user
 
 unsigned long __must_check ___copy_in_user(void __user *to,
 					   const void __user *from,
 					   unsigned long size);
-unsigned long copy_in_user_fixup(void __user *to, void __user *from,
-				 unsigned long size);
 static inline unsigned long __must_check
 copy_in_user(void __user *to, void __user *from, unsigned long size)
 {
-	unsigned long ret = ___copy_in_user(to, from, size);
-
-	if (unlikely(ret))
-		ret = copy_in_user_fixup(to, from, size);
-	return ret;
+	return ___copy_in_user(to, from, size);
 }
 #define __copy_in_user copy_in_user
 
diff --git a/arch/sparc/kernel/head_64.S b/arch/sparc/kernel/head_64.S
index beba6c11554c..6aa3da152c20 100644
--- a/arch/sparc/kernel/head_64.S
+++ b/arch/sparc/kernel/head_64.S
@@ -926,48 +926,11 @@ tlb_type:	.word	0	/* Must NOT end up in BSS */
 EXPORT_SYMBOL(tlb_type)
 	.section	".fixup",#alloc,#execinstr
 
-	.globl	__ret_efault, __retl_efault, __ret_one, __retl_one
-ENTRY(__ret_efault)
-	ret
-	 restore %g0, -EFAULT, %o0
-ENDPROC(__ret_efault)
-EXPORT_SYMBOL(__ret_efault)
-
 ENTRY(__retl_efault)
 	retl
 	 mov	-EFAULT, %o0
 ENDPROC(__retl_efault)
 
-ENTRY(__retl_one)
-	retl
-	 mov	1, %o0
-ENDPROC(__retl_one)
-
-ENTRY(__retl_one_fp)
-	VISExitHalf
-	retl
-	 mov	1, %o0
-ENDPROC(__retl_one_fp)
-
-ENTRY(__ret_one_asi)
-	wr	%g0, ASI_AIUS, %asi
-	ret
-	 restore %g0, 1, %o0
-ENDPROC(__ret_one_asi)
-
-ENTRY(__retl_one_asi)
-	wr	%g0, ASI_AIUS, %asi
-	retl
-	 mov	1, %o0
-ENDPROC(__retl_one_asi)
-
-ENTRY(__retl_one_asi_fp)
-	wr	%g0, ASI_AIUS, %asi
-	VISExitHalf
-	retl
-	 mov	1, %o0
-ENDPROC(__retl_one_asi_fp)
-
 ENTRY(__retl_o1)
 	retl
 	 mov	%o1, %o0
diff --git a/arch/sparc/kernel/jump_label.c b/arch/sparc/kernel/jump_label.c
index 59bbeff55024..07933b9e9ce0 100644
--- a/arch/sparc/kernel/jump_label.c
+++ b/arch/sparc/kernel/jump_label.c
@@ -13,19 +13,30 @@
 void arch_jump_label_transform(struct jump_entry *entry,
 			       enum jump_label_type type)
 {
-	u32 val;
 	u32 *insn = (u32 *) (unsigned long) entry->code;
+	u32 val;
 
 	if (type == JUMP_LABEL_JMP) {
 		s32 off = (s32)entry->target - (s32)entry->code;
+		bool use_v9_branch = false;
+
+		BUG_ON(off & 3);
 
 #ifdef CONFIG_SPARC64
-		/* ba,pt %xcc, . + (off << 2) */
-		val = 0x10680000 | ((u32) off >> 2);
-#else
-		/* ba . + (off << 2) */
-		val = 0x10800000 | ((u32) off >> 2);
+		if (off <= 0xfffff && off >= -0x100000)
+			use_v9_branch = true;
 #endif
+		if (use_v9_branch) {
+			/* WDISP19 - target is . + immed << 2 */
+			/* ba,pt %xcc, . + off */
+			val = 0x10680000 | (((u32) off >> 2) & 0x7ffff);
+		} else {
+			/* WDISP22 - target is . + immed << 2 */
+			BUG_ON(off > 0x7fffff);
+			BUG_ON(off < -0x800000);
+			/* ba . + off */
+			val = 0x10800000 | (((u32) off >> 2) & 0x3fffff);
+		}
 	} else {
 		val = 0x01000000;
 	}
diff --git a/arch/sparc/kernel/mdesc.c b/arch/sparc/kernel/mdesc.c
index 11228861d9b4..8a6982dfd733 100644
--- a/arch/sparc/kernel/mdesc.c
+++ b/arch/sparc/kernel/mdesc.c
@@ -645,13 +645,20 @@ static void __mark_core_id(struct mdesc_handle *hp, u64 node,
 		cpu_data(*id).core_id = core_id;
 }
 
-static void __mark_sock_id(struct mdesc_handle *hp, u64 node,
-			   int sock_id)
+static void __mark_max_cache_id(struct mdesc_handle *hp, u64 node,
+				int max_cache_id)
 {
 	const u64 *id = mdesc_get_property(hp, node, "id", NULL);
 
-	if (*id < num_possible_cpus())
-		cpu_data(*id).sock_id = sock_id;
+	if (*id < num_possible_cpus()) {
+		cpu_data(*id).max_cache_id = max_cache_id;
+
+		/**
+		 * On systems without explicit socket descriptions socket
+		 * is max_cache_id
+		 */
+		cpu_data(*id).sock_id = max_cache_id;
+	}
 }
 
 static void mark_core_ids(struct mdesc_handle *hp, u64 mp,
@@ -660,10 +667,11 @@ static void mark_core_ids(struct mdesc_handle *hp, u64 mp,
 	find_back_node_value(hp, mp, "cpu", __mark_core_id, core_id, 10);
 }
 
-static void mark_sock_ids(struct mdesc_handle *hp, u64 mp,
-			  int sock_id)
+static void mark_max_cache_ids(struct mdesc_handle *hp, u64 mp,
+			       int max_cache_id)
 {
-	find_back_node_value(hp, mp, "cpu", __mark_sock_id, sock_id, 10);
+	find_back_node_value(hp, mp, "cpu", __mark_max_cache_id,
+			     max_cache_id, 10);
 }
 
 static void set_core_ids(struct mdesc_handle *hp)
@@ -694,14 +702,15 @@ static void set_core_ids(struct mdesc_handle *hp)
 	}
 }
 
-static int set_sock_ids_by_cache(struct mdesc_handle *hp, int level)
+static int set_max_cache_ids_by_cache(struct mdesc_handle *hp, int level)
 {
 	u64 mp;
 	int idx = 1;
 	int fnd = 0;
 
-	/* Identify unique sockets by looking for cpus backpointed to by
-	 * shared level n caches.
+	/**
+	 * Identify unique highest level of shared cache by looking for cpus
+	 * backpointed to by shared level N caches.
 	 */
 	mdesc_for_each_node_by_name(hp, mp, "cache") {
 		const u64 *cur_lvl;
@@ -709,8 +718,7 @@ static int set_sock_ids_by_cache(struct mdesc_handle *hp, int level)
 		cur_lvl = mdesc_get_property(hp, mp, "level", NULL);
 		if (*cur_lvl != level)
 			continue;
-
-		mark_sock_ids(hp, mp, idx);
+		mark_max_cache_ids(hp, mp, idx);
 		idx++;
 		fnd = 1;
 	}
@@ -745,15 +753,17 @@ static void set_sock_ids(struct mdesc_handle *hp)
 {
 	u64 mp;
 
-	/* If machine description exposes sockets data use it.
-	 * Otherwise fallback to use shared L3 or L2 caches.
+	/**
+	 * Find the highest level of shared cache which pre-T7 is also
+	 * the socket.
 	 */
+	if (!set_max_cache_ids_by_cache(hp, 3))
+		set_max_cache_ids_by_cache(hp, 2);
+
+	/* If machine description exposes sockets data use it.*/
 	mp = mdesc_node_by_name(hp, MDESC_NODE_NULL, "sockets");
 	if (mp != MDESC_NODE_NULL)
-		return set_sock_ids_by_socket(hp, mp);
-
-	if (!set_sock_ids_by_cache(hp, 3))
-		set_sock_ids_by_cache(hp, 2);
+		set_sock_ids_by_socket(hp, mp);
 }
 
 static void mark_proc_ids(struct mdesc_handle *hp, u64 mp, int proc_id)
diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c
index d3035ba6cd31..8182f7caf5b1 100644
--- a/arch/sparc/kernel/smp_64.c
+++ b/arch/sparc/kernel/smp_64.c
@@ -63,9 +63,13 @@ cpumask_t cpu_core_map[NR_CPUS] __read_mostly =
 cpumask_t cpu_core_sib_map[NR_CPUS] __read_mostly = {
 	[0 ... NR_CPUS-1] = CPU_MASK_NONE };
 
+cpumask_t cpu_core_sib_cache_map[NR_CPUS] __read_mostly = {
+	[0 ... NR_CPUS - 1] = CPU_MASK_NONE };
+
 EXPORT_PER_CPU_SYMBOL(cpu_sibling_map);
 EXPORT_SYMBOL(cpu_core_map);
 EXPORT_SYMBOL(cpu_core_sib_map);
+EXPORT_SYMBOL(cpu_core_sib_cache_map);
 
 static cpumask_t smp_commenced_mask;
 
@@ -1265,6 +1269,10 @@ void smp_fill_in_sib_core_maps(void)
 		unsigned int j;
 
 		for_each_present_cpu(j)  {
+			if (cpu_data(i).max_cache_id ==
+			    cpu_data(j).max_cache_id)
+				cpumask_set_cpu(j, &cpu_core_sib_cache_map[i]);
+
 			if (cpu_data(i).sock_id == cpu_data(j).sock_id)
 				cpumask_set_cpu(j, &cpu_core_sib_map[i]);
 		}
diff --git a/arch/sparc/lib/GENcopy_from_user.S b/arch/sparc/lib/GENcopy_from_user.S
index b7d0bd6b1406..69a439fa2fc1 100644
--- a/arch/sparc/lib/GENcopy_from_user.S
+++ b/arch/sparc/lib/GENcopy_from_user.S
@@ -3,11 +3,11 @@
  * Copyright (C) 2007 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_LD(x)		\
+#define EX_LD(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one;	\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/GENcopy_to_user.S b/arch/sparc/lib/GENcopy_to_user.S
index 780550e1afc7..9947427ce354 100644
--- a/arch/sparc/lib/GENcopy_to_user.S
+++ b/arch/sparc/lib/GENcopy_to_user.S
@@ -3,11 +3,11 @@
  * Copyright (C) 2007 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_ST(x)		\
+#define EX_ST(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one;	\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/GENmemcpy.S b/arch/sparc/lib/GENmemcpy.S
index 89358ee94851..059ea24ad73d 100644
--- a/arch/sparc/lib/GENmemcpy.S
+++ b/arch/sparc/lib/GENmemcpy.S
@@ -4,21 +4,18 @@
  */
 
 #ifdef __KERNEL__
+#include <linux/linkage.h>
 #define GLOBAL_SPARE	%g7
 #else
 #define GLOBAL_SPARE	%g5
 #endif
 
 #ifndef EX_LD
-#define EX_LD(x)	x
+#define EX_LD(x,y)	x
 #endif
 
 #ifndef EX_ST
-#define EX_ST(x)	x
-#endif
-
-#ifndef EX_RETVAL
-#define EX_RETVAL(x)	x
+#define EX_ST(x,y)	x
 #endif
 
 #ifndef LOAD
@@ -45,6 +42,29 @@
 	.register	%g3,#scratch
 
 	.text
+
+#ifndef EX_RETVAL
+#define EX_RETVAL(x)	x
+ENTRY(GEN_retl_o4_1)
+	add	%o4, %o2, %o4
+	retl
+	 add	%o4, 1, %o0
+ENDPROC(GEN_retl_o4_1)
+ENTRY(GEN_retl_g1_8)
+	add	%g1, %o2, %g1
+	retl
+	 add	%g1, 8, %o0
+ENDPROC(GEN_retl_g1_8)
+ENTRY(GEN_retl_o2_4)
+	retl
+	 add	%o2, 4, %o0
+ENDPROC(GEN_retl_o2_4)
+ENTRY(GEN_retl_o2_1)
+	retl
+	 add	%o2, 1, %o0
+ENDPROC(GEN_retl_o2_1)
+#endif
+
 	.align		64
 
 	.globl	FUNC_NAME
@@ -73,8 +93,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	sub		%g0, %o4, %o4
 	sub		%o2, %o4, %o2
 1:	subcc		%o4, 1, %o4
-	EX_LD(LOAD(ldub, %o1, %g1))
-	EX_ST(STORE(stb, %g1, %o0))
+	EX_LD(LOAD(ldub, %o1, %g1),GEN_retl_o4_1)
+	EX_ST(STORE(stb, %g1, %o0),GEN_retl_o4_1)
 	add		%o1, 1, %o1
 	bne,pt		%XCC, 1b
 	add		%o0, 1, %o0
@@ -82,8 +102,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	andn		%o2, 0x7, %g1
 	sub		%o2, %g1, %o2
 1:	subcc		%g1, 0x8, %g1
-	EX_LD(LOAD(ldx, %o1, %g2))
-	EX_ST(STORE(stx, %g2, %o0))
+	EX_LD(LOAD(ldx, %o1, %g2),GEN_retl_g1_8)
+	EX_ST(STORE(stx, %g2, %o0),GEN_retl_g1_8)
 	add		%o1, 0x8, %o1
 	bne,pt		%XCC, 1b
 	 add		%o0, 0x8, %o0
@@ -100,8 +120,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 1:
 	subcc		%o2, 4, %o2
-	EX_LD(LOAD(lduw, %o1, %g1))
-	EX_ST(STORE(stw, %g1, %o1 + %o3))
+	EX_LD(LOAD(lduw, %o1, %g1),GEN_retl_o2_4)
+	EX_ST(STORE(stw, %g1, %o1 + %o3),GEN_retl_o2_4)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 4, %o1
 
@@ -111,8 +131,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	.align		32
 90:
 	subcc		%o2, 1, %o2
-	EX_LD(LOAD(ldub, %o1, %g1))
-	EX_ST(STORE(stb, %g1, %o1 + %o3))
+	EX_LD(LOAD(ldub, %o1, %g1),GEN_retl_o2_1)
+	EX_ST(STORE(stb, %g1, %o1 + %o3),GEN_retl_o2_1)
 	bgu,pt		%XCC, 90b
 	 add		%o1, 1, %o1
 	retl
diff --git a/arch/sparc/lib/Makefile b/arch/sparc/lib/Makefile
index 885f00e81d1a..69912d2f8b54 100644
--- a/arch/sparc/lib/Makefile
+++ b/arch/sparc/lib/Makefile
@@ -38,7 +38,7 @@ lib-$(CONFIG_SPARC64) +=  NG4patch.o NG4copy_page.o NG4clear_page.o NG4memset.o
 lib-$(CONFIG_SPARC64) += GENmemcpy.o GENcopy_from_user.o GENcopy_to_user.o
 lib-$(CONFIG_SPARC64) += GENpatch.o GENpage.o GENbzero.o
 
-lib-$(CONFIG_SPARC64) += copy_in_user.o user_fixup.o memmove.o
+lib-$(CONFIG_SPARC64) += copy_in_user.o memmove.o
 lib-$(CONFIG_SPARC64) += mcount.o ipcsum.o xor.o hweight.o ffs.o
 
 obj-$(CONFIG_SPARC64) += iomap.o
diff --git a/arch/sparc/lib/NG2copy_from_user.S b/arch/sparc/lib/NG2copy_from_user.S
index d5242b8c4f94..b79a6998d87c 100644
--- a/arch/sparc/lib/NG2copy_from_user.S
+++ b/arch/sparc/lib/NG2copy_from_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 2007 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_LD(x)		\
+#define EX_LD(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_LD_FP(x)		\
+#define EX_LD_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi_fp;\
+	.word 98b, y##_fp;	\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/NG2copy_to_user.S b/arch/sparc/lib/NG2copy_to_user.S
index 4e962d993b10..dcec55f254ab 100644
--- a/arch/sparc/lib/NG2copy_to_user.S
+++ b/arch/sparc/lib/NG2copy_to_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 2007 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_ST(x)		\
+#define EX_ST(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_ST_FP(x)		\
+#define EX_ST_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi_fp;\
+	.word 98b, y##_fp;	\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/NG2memcpy.S b/arch/sparc/lib/NG2memcpy.S
index d5f585df2f3f..c629dbd121b6 100644
--- a/arch/sparc/lib/NG2memcpy.S
+++ b/arch/sparc/lib/NG2memcpy.S
@@ -4,6 +4,7 @@
  */
 
 #ifdef __KERNEL__
+#include <linux/linkage.h>
 #include <asm/visasm.h>
 #include <asm/asi.h>
 #define GLOBAL_SPARE	%g7
@@ -32,21 +33,17 @@
 #endif
 
 #ifndef EX_LD
-#define EX_LD(x)	x
+#define EX_LD(x,y)	x
 #endif
 #ifndef EX_LD_FP
-#define EX_LD_FP(x)	x
+#define EX_LD_FP(x,y)	x
 #endif
 
 #ifndef EX_ST
-#define EX_ST(x)	x
+#define EX_ST(x,y)	x
 #endif
 #ifndef EX_ST_FP
-#define EX_ST_FP(x)	x
-#endif
-
-#ifndef EX_RETVAL
-#define EX_RETVAL(x)	x
+#define EX_ST_FP(x,y)	x
 #endif
 
 #ifndef LOAD
@@ -140,45 +137,110 @@
 	fsrc2		%x6, %f12; \
 	fsrc2		%x7, %f14;
 #define FREG_LOAD_1(base, x0) \
-	EX_LD_FP(LOAD(ldd, base + 0x00, %x0))
+	EX_LD_FP(LOAD(ldd, base + 0x00, %x0), NG2_retl_o2_plus_g1)
 #define FREG_LOAD_2(base, x0, x1) \
-	EX_LD_FP(LOAD(ldd, base + 0x00, %x0)); \
-	EX_LD_FP(LOAD(ldd, base + 0x08, %x1));
+	EX_LD_FP(LOAD(ldd, base + 0x00, %x0), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x08, %x1), NG2_retl_o2_plus_g1);
 #define FREG_LOAD_3(base, x0, x1, x2) \
-	EX_LD_FP(LOAD(ldd, base + 0x00, %x0)); \
-	EX_LD_FP(LOAD(ldd, base + 0x08, %x1)); \
-	EX_LD_FP(LOAD(ldd, base + 0x10, %x2));
+	EX_LD_FP(LOAD(ldd, base + 0x00, %x0), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x08, %x1), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x10, %x2), NG2_retl_o2_plus_g1);
 #define FREG_LOAD_4(base, x0, x1, x2, x3) \
-	EX_LD_FP(LOAD(ldd, base + 0x00, %x0)); \
-	EX_LD_FP(LOAD(ldd, base + 0x08, %x1)); \
-	EX_LD_FP(LOAD(ldd, base + 0x10, %x2)); \
-	EX_LD_FP(LOAD(ldd, base + 0x18, %x3));
+	EX_LD_FP(LOAD(ldd, base + 0x00, %x0), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x08, %x1), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x10, %x2), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x18, %x3), NG2_retl_o2_plus_g1);
 #define FREG_LOAD_5(base, x0, x1, x2, x3, x4) \
-	EX_LD_FP(LOAD(ldd, base + 0x00, %x0)); \
-	EX_LD_FP(LOAD(ldd, base + 0x08, %x1)); \
-	EX_LD_FP(LOAD(ldd, base + 0x10, %x2)); \
-	EX_LD_FP(LOAD(ldd, base + 0x18, %x3)); \
-	EX_LD_FP(LOAD(ldd, base + 0x20, %x4));
+	EX_LD_FP(LOAD(ldd, base + 0x00, %x0), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x08, %x1), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x10, %x2), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x18, %x3), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x20, %x4), NG2_retl_o2_plus_g1);
 #define FREG_LOAD_6(base, x0, x1, x2, x3, x4, x5) \
-	EX_LD_FP(LOAD(ldd, base + 0x00, %x0)); \
-	EX_LD_FP(LOAD(ldd, base + 0x08, %x1)); \
-	EX_LD_FP(LOAD(ldd, base + 0x10, %x2)); \
-	EX_LD_FP(LOAD(ldd, base + 0x18, %x3)); \
-	EX_LD_FP(LOAD(ldd, base + 0x20, %x4)); \
-	EX_LD_FP(LOAD(ldd, base + 0x28, %x5));
+	EX_LD_FP(LOAD(ldd, base + 0x00, %x0), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x08, %x1), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x10, %x2), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x18, %x3), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x20, %x4), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x28, %x5), NG2_retl_o2_plus_g1);
 #define FREG_LOAD_7(base, x0, x1, x2, x3, x4, x5, x6) \
-	EX_LD_FP(LOAD(ldd, base + 0x00, %x0)); \
-	EX_LD_FP(LOAD(ldd, base + 0x08, %x1)); \
-	EX_LD_FP(LOAD(ldd, base + 0x10, %x2)); \
-	EX_LD_FP(LOAD(ldd, base + 0x18, %x3)); \
-	EX_LD_FP(LOAD(ldd, base + 0x20, %x4)); \
-	EX_LD_FP(LOAD(ldd, base + 0x28, %x5)); \
-	EX_LD_FP(LOAD(ldd, base + 0x30, %x6));
+	EX_LD_FP(LOAD(ldd, base + 0x00, %x0), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x08, %x1), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x10, %x2), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x18, %x3), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x20, %x4), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x28, %x5), NG2_retl_o2_plus_g1); \
+	EX_LD_FP(LOAD(ldd, base + 0x30, %x6), NG2_retl_o2_plus_g1);
 
 	.register	%g2,#scratch
 	.register	%g3,#scratch
 
 	.text
+#ifndef EX_RETVAL
+#define EX_RETVAL(x)	x
+__restore_fp:
+	VISExitHalf
+__restore_asi:
+	retl
+	 wr	%g0, ASI_AIUS, %asi
+ENTRY(NG2_retl_o2)
+	ba,pt	%xcc, __restore_asi
+	 mov	%o2, %o0
+ENDPROC(NG2_retl_o2)
+ENTRY(NG2_retl_o2_plus_1)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, 1, %o0
+ENDPROC(NG2_retl_o2_plus_1)
+ENTRY(NG2_retl_o2_plus_4)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, 4, %o0
+ENDPROC(NG2_retl_o2_plus_4)
+ENTRY(NG2_retl_o2_plus_8)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, 8, %o0
+ENDPROC(NG2_retl_o2_plus_8)
+ENTRY(NG2_retl_o2_plus_o4_plus_1)
+	add	%o4, 1, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG2_retl_o2_plus_o4_plus_1)
+ENTRY(NG2_retl_o2_plus_o4_plus_8)
+	add	%o4, 8, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG2_retl_o2_plus_o4_plus_8)
+ENTRY(NG2_retl_o2_plus_o4_plus_16)
+	add	%o4, 16, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG2_retl_o2_plus_o4_plus_16)
+ENTRY(NG2_retl_o2_plus_g1_fp)
+	ba,pt	%xcc, __restore_fp
+	 add	%o2, %g1, %o0
+ENDPROC(NG2_retl_o2_plus_g1_fp)
+ENTRY(NG2_retl_o2_plus_g1_plus_64_fp)
+	add	%g1, 64, %g1
+	ba,pt	%xcc, __restore_fp
+	 add	%o2, %g1, %o0
+ENDPROC(NG2_retl_o2_plus_g1_plus_64_fp)
+ENTRY(NG2_retl_o2_plus_g1_plus_1)
+	add	%g1, 1, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %g1, %o0
+ENDPROC(NG2_retl_o2_plus_g1_plus_1)
+ENTRY(NG2_retl_o2_and_7_plus_o4)
+	and	%o2, 7, %o2
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG2_retl_o2_and_7_plus_o4)
+ENTRY(NG2_retl_o2_and_7_plus_o4_plus_8)
+	and	%o2, 7, %o2
+	add	%o4, 8, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG2_retl_o2_and_7_plus_o4_plus_8)
+#endif
+
 	.align		64
 
 	.globl	FUNC_NAME
@@ -230,8 +292,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	sub		%g0, %o4, %o4	! bytes to align dst
 	sub		%o2, %o4, %o2
 1:	subcc		%o4, 1, %o4
-	EX_LD(LOAD(ldub, %o1, %g1))
-	EX_ST(STORE(stb, %g1, %o0))
+	EX_LD(LOAD(ldub, %o1, %g1), NG2_retl_o2_plus_o4_plus_1)
+	EX_ST(STORE(stb, %g1, %o0), NG2_retl_o2_plus_o4_plus_1)
 	add		%o1, 1, %o1
 	bne,pt		%XCC, 1b
 	add		%o0, 1, %o0
@@ -281,11 +343,11 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	 nop
 	/* fall through for 0 < low bits < 8 */
 110:	sub		%o4, 64, %g2
-	EX_LD_FP(LOAD_BLK(%g2, %f0))
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+	EX_LD_FP(LOAD_BLK(%g2, %f0), NG2_retl_o2_plus_g1)
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f2, f4, f6, f8, f10, f12, f14, f16)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_8(f16, f18, f20, f22, f24, f26, f28, f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -296,10 +358,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 120:	sub		%o4, 56, %g2
 	FREG_LOAD_7(%g2, f0, f2, f4, f6, f8, f10, f12)
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f2, f4, f6, f8, f10, f12, f16, f18)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_7(f18, f20, f22, f24, f26, f28, f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -310,10 +372,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 130:	sub		%o4, 48, %g2
 	FREG_LOAD_6(%g2, f0, f2, f4, f6, f8, f10)
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f2, f4, f6, f8, f10, f16, f18, f20)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_6(f20, f22, f24, f26, f28, f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -324,10 +386,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 140:	sub		%o4, 40, %g2
 	FREG_LOAD_5(%g2, f0, f2, f4, f6, f8)
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f2, f4, f6, f8, f16, f18, f20, f22)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_5(f22, f24, f26, f28, f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -338,10 +400,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 150:	sub		%o4, 32, %g2
 	FREG_LOAD_4(%g2, f0, f2, f4, f6)
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f2, f4, f6, f16, f18, f20, f22, f24)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_4(f24, f26, f28, f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -352,10 +414,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 160:	sub		%o4, 24, %g2
 	FREG_LOAD_3(%g2, f0, f2, f4)
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f2, f4, f16, f18, f20, f22, f24, f26)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_3(f26, f28, f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -366,10 +428,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 170:	sub		%o4, 16, %g2
 	FREG_LOAD_2(%g2, f0, f2)
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f2, f16, f18, f20, f22, f24, f26, f28)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_2(f28, f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -380,10 +442,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 180:	sub		%o4, 8, %g2
 	FREG_LOAD_1(%g2, f0)
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
-	EX_LD_FP(LOAD_BLK(%o4, %f16))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
+	EX_LD_FP(LOAD_BLK(%o4, %f16), NG2_retl_o2_plus_g1)
 	FREG_FROB(f0, f16, f18, f20, f22, f24, f26, f28, f30)
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	FREG_MOVE_1(f30)
 	subcc		%g1, 64, %g1
 	add		%o4, 64, %o4
@@ -393,10 +455,10 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	 nop
 
 190:
-1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3))
+1:	EX_ST_FP(STORE_INIT(%g0, %o4 + %g3), NG2_retl_o2_plus_g1)
 	subcc		%g1, 64, %g1
-	EX_LD_FP(LOAD_BLK(%o4, %f0))
-	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3))
+	EX_LD_FP(LOAD_BLK(%o4, %f0), NG2_retl_o2_plus_g1_plus_64)
+	EX_ST_FP(STORE_BLK(%f0, %o4 + %g3), NG2_retl_o2_plus_g1_plus_64)
 	add		%o4, 64, %o4
 	bne,pt		%xcc, 1b
 	 LOAD(prefetch, %o4 + 64, #one_read)
@@ -423,28 +485,28 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	andn		%o2, 0xf, %o4
 	and		%o2, 0xf, %o2
 1:	subcc		%o4, 0x10, %o4
-	EX_LD(LOAD(ldx, %o1, %o5))
+	EX_LD(LOAD(ldx, %o1, %o5), NG2_retl_o2_plus_o4_plus_16)
 	add		%o1, 0x08, %o1
-	EX_LD(LOAD(ldx, %o1, %g1))
+	EX_LD(LOAD(ldx, %o1, %g1), NG2_retl_o2_plus_o4_plus_16)
 	sub		%o1, 0x08, %o1
-	EX_ST(STORE(stx, %o5, %o1 + GLOBAL_SPARE))
+	EX_ST(STORE(stx, %o5, %o1 + GLOBAL_SPARE), NG2_retl_o2_plus_o4_plus_16)
 	add		%o1, 0x8, %o1
-	EX_ST(STORE(stx, %g1, %o1 + GLOBAL_SPARE))
+	EX_ST(STORE(stx, %g1, %o1 + GLOBAL_SPARE), NG2_retl_o2_plus_o4_plus_8)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 0x8, %o1
 73:	andcc		%o2, 0x8, %g0
 	be,pt		%XCC, 1f
 	 nop
 	sub		%o2, 0x8, %o2
-	EX_LD(LOAD(ldx, %o1, %o5))
-	EX_ST(STORE(stx, %o5, %o1 + GLOBAL_SPARE))
+	EX_LD(LOAD(ldx, %o1, %o5), NG2_retl_o2_plus_8)
+	EX_ST(STORE(stx, %o5, %o1 + GLOBAL_SPARE), NG2_retl_o2_plus_8)
 	add		%o1, 0x8, %o1
 1:	andcc		%o2, 0x4, %g0
 	be,pt		%XCC, 1f
 	 nop
 	sub		%o2, 0x4, %o2
-	EX_LD(LOAD(lduw, %o1, %o5))
-	EX_ST(STORE(stw, %o5, %o1 + GLOBAL_SPARE))
+	EX_LD(LOAD(lduw, %o1, %o5), NG2_retl_o2_plus_4)
+	EX_ST(STORE(stw, %o5, %o1 + GLOBAL_SPARE), NG2_retl_o2_plus_4)
 	add		%o1, 0x4, %o1
 1:	cmp		%o2, 0
 	be,pt		%XCC, 85f
@@ -460,8 +522,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	sub		%o2, %g1, %o2
 
 1:	subcc		%g1, 1, %g1
-	EX_LD(LOAD(ldub, %o1, %o5))
-	EX_ST(STORE(stb, %o5, %o1 + GLOBAL_SPARE))
+	EX_LD(LOAD(ldub, %o1, %o5), NG2_retl_o2_plus_g1_plus_1)
+	EX_ST(STORE(stb, %o5, %o1 + GLOBAL_SPARE), NG2_retl_o2_plus_g1_plus_1)
 	bgu,pt		%icc, 1b
 	 add		%o1, 1, %o1
 
@@ -477,16 +539,16 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 8:	mov		64, GLOBAL_SPARE
 	andn		%o1, 0x7, %o1
-	EX_LD(LOAD(ldx, %o1, %g2))
+	EX_LD(LOAD(ldx, %o1, %g2), NG2_retl_o2)
 	sub		GLOBAL_SPARE, %g1, GLOBAL_SPARE
 	andn		%o2, 0x7, %o4
 	sllx		%g2, %g1, %g2
 1:	add		%o1, 0x8, %o1
-	EX_LD(LOAD(ldx, %o1, %g3))
+	EX_LD(LOAD(ldx, %o1, %g3), NG2_retl_o2_and_7_plus_o4)
 	subcc		%o4, 0x8, %o4
 	srlx		%g3, GLOBAL_SPARE, %o5
 	or		%o5, %g2, %o5
-	EX_ST(STORE(stx, %o5, %o0))
+	EX_ST(STORE(stx, %o5, %o0), NG2_retl_o2_and_7_plus_o4_plus_8)
 	add		%o0, 0x8, %o0
 	bgu,pt		%icc, 1b
 	 sllx		%g3, %g1, %g2
@@ -506,8 +568,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 1:
 	subcc		%o2, 4, %o2
-	EX_LD(LOAD(lduw, %o1, %g1))
-	EX_ST(STORE(stw, %g1, %o1 + GLOBAL_SPARE))
+	EX_LD(LOAD(lduw, %o1, %g1), NG2_retl_o2_plus_4)
+	EX_ST(STORE(stw, %g1, %o1 + GLOBAL_SPARE), NG2_retl_o2_plus_4)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 4, %o1
 
@@ -517,8 +579,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	.align		32
 90:
 	subcc		%o2, 1, %o2
-	EX_LD(LOAD(ldub, %o1, %g1))
-	EX_ST(STORE(stb, %g1, %o1 + GLOBAL_SPARE))
+	EX_LD(LOAD(ldub, %o1, %g1), NG2_retl_o2_plus_1)
+	EX_ST(STORE(stb, %g1, %o1 + GLOBAL_SPARE), NG2_retl_o2_plus_1)
 	bgu,pt		%XCC, 90b
 	 add		%o1, 1, %o1
 	retl
diff --git a/arch/sparc/lib/NG4copy_from_user.S b/arch/sparc/lib/NG4copy_from_user.S
index 2e8ee7ad07a9..16a286c1a528 100644
--- a/arch/sparc/lib/NG4copy_from_user.S
+++ b/arch/sparc/lib/NG4copy_from_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 2012 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_LD(x)		\
+#define EX_LD(x, y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_LD_FP(x)		\
+#define EX_LD_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi_fp;\
+	.word 98b, y##_fp;	\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/NG4copy_to_user.S b/arch/sparc/lib/NG4copy_to_user.S
index be0bf4590df8..6b0276ffc858 100644
--- a/arch/sparc/lib/NG4copy_to_user.S
+++ b/arch/sparc/lib/NG4copy_to_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 2012 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_ST(x)		\
+#define EX_ST(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_ST_FP(x)		\
+#define EX_ST_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_asi_fp;\
+	.word 98b, y##_fp;	\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/NG4memcpy.S b/arch/sparc/lib/NG4memcpy.S
index 8e13ee1f4454..75bb93b1437f 100644
--- a/arch/sparc/lib/NG4memcpy.S
+++ b/arch/sparc/lib/NG4memcpy.S
@@ -4,6 +4,7 @@
  */
 
 #ifdef __KERNEL__
+#include <linux/linkage.h>
 #include <asm/visasm.h>
 #include <asm/asi.h>
 #define GLOBAL_SPARE	%g7
@@ -46,22 +47,19 @@
 #endif
 
 #ifndef EX_LD
-#define EX_LD(x)	x
+#define EX_LD(x,y)	x
 #endif
 #ifndef EX_LD_FP
-#define EX_LD_FP(x)	x
+#define EX_LD_FP(x,y)	x
 #endif
 
 #ifndef EX_ST
-#define EX_ST(x)	x
+#define EX_ST(x,y)	x
 #endif
 #ifndef EX_ST_FP
-#define EX_ST_FP(x)	x
+#define EX_ST_FP(x,y)	x
 #endif
 
-#ifndef EX_RETVAL
-#define EX_RETVAL(x)	x
-#endif
 
 #ifndef LOAD
 #define LOAD(type,addr,dest)	type [addr], dest
@@ -94,6 +92,158 @@
 	.register	%g3,#scratch
 
 	.text
+#ifndef EX_RETVAL
+#define EX_RETVAL(x)	x
+__restore_asi_fp:
+	VISExitHalf
+__restore_asi:
+	retl
+	 wr	%g0, ASI_AIUS, %asi
+
+ENTRY(NG4_retl_o2)
+	ba,pt	%xcc, __restore_asi
+	 mov	%o2, %o0
+ENDPROC(NG4_retl_o2)
+ENTRY(NG4_retl_o2_plus_1)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, 1, %o0
+ENDPROC(NG4_retl_o2_plus_1)
+ENTRY(NG4_retl_o2_plus_4)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, 4, %o0
+ENDPROC(NG4_retl_o2_plus_4)
+ENTRY(NG4_retl_o2_plus_o5)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o5, %o0
+ENDPROC(NG4_retl_o2_plus_o5)
+ENTRY(NG4_retl_o2_plus_o5_plus_4)
+	add	%o5, 4, %o5
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o5, %o0
+ENDPROC(NG4_retl_o2_plus_o5_plus_4)
+ENTRY(NG4_retl_o2_plus_o5_plus_8)
+	add	%o5, 8, %o5
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o5, %o0
+ENDPROC(NG4_retl_o2_plus_o5_plus_8)
+ENTRY(NG4_retl_o2_plus_o5_plus_16)
+	add	%o5, 16, %o5
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o5, %o0
+ENDPROC(NG4_retl_o2_plus_o5_plus_16)
+ENTRY(NG4_retl_o2_plus_o5_plus_24)
+	add	%o5, 24, %o5
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o5, %o0
+ENDPROC(NG4_retl_o2_plus_o5_plus_24)
+ENTRY(NG4_retl_o2_plus_o5_plus_32)
+	add	%o5, 32, %o5
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o5, %o0
+ENDPROC(NG4_retl_o2_plus_o5_plus_32)
+ENTRY(NG4_retl_o2_plus_g1)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %g1, %o0
+ENDPROC(NG4_retl_o2_plus_g1)
+ENTRY(NG4_retl_o2_plus_g1_plus_1)
+	add	%g1, 1, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %g1, %o0
+ENDPROC(NG4_retl_o2_plus_g1_plus_1)
+ENTRY(NG4_retl_o2_plus_g1_plus_8)
+	add	%g1, 8, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %g1, %o0
+ENDPROC(NG4_retl_o2_plus_g1_plus_8)
+ENTRY(NG4_retl_o2_plus_o4)
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4)
+ENTRY(NG4_retl_o2_plus_o4_plus_8)
+	add	%o4, 8, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_8)
+ENTRY(NG4_retl_o2_plus_o4_plus_16)
+	add	%o4, 16, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_16)
+ENTRY(NG4_retl_o2_plus_o4_plus_24)
+	add	%o4, 24, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_24)
+ENTRY(NG4_retl_o2_plus_o4_plus_32)
+	add	%o4, 32, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_32)
+ENTRY(NG4_retl_o2_plus_o4_plus_40)
+	add	%o4, 40, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_40)
+ENTRY(NG4_retl_o2_plus_o4_plus_48)
+	add	%o4, 48, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_48)
+ENTRY(NG4_retl_o2_plus_o4_plus_56)
+	add	%o4, 56, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_56)
+ENTRY(NG4_retl_o2_plus_o4_plus_64)
+	add	%o4, 64, %o4
+	ba,pt	%xcc, __restore_asi
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_64)
+ENTRY(NG4_retl_o2_plus_o4_fp)
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_8_fp)
+	add	%o4, 8, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_8_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_16_fp)
+	add	%o4, 16, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_16_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_24_fp)
+	add	%o4, 24, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_24_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_32_fp)
+	add	%o4, 32, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_32_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_40_fp)
+	add	%o4, 40, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_40_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_48_fp)
+	add	%o4, 48, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_48_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_56_fp)
+	add	%o4, 56, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_56_fp)
+ENTRY(NG4_retl_o2_plus_o4_plus_64_fp)
+	add	%o4, 64, %o4
+	ba,pt	%xcc, __restore_asi_fp
+	 add	%o2, %o4, %o0
+ENDPROC(NG4_retl_o2_plus_o4_plus_64_fp)
+#endif
 	.align		64
 
 	.globl	FUNC_NAME
@@ -124,12 +274,13 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	brz,pt		%g1, 51f
 	 sub		%o2, %g1, %o2
 
-1:	EX_LD(LOAD(ldub, %o1 + 0x00, %g2))
+
+1:	EX_LD(LOAD(ldub, %o1 + 0x00, %g2), NG4_retl_o2_plus_g1)
 	add		%o1, 1, %o1
 	subcc		%g1, 1, %g1
 	add		%o0, 1, %o0
 	bne,pt		%icc, 1b
-	 EX_ST(STORE(stb, %g2, %o0 - 0x01))
+	 EX_ST(STORE(stb, %g2, %o0 - 0x01), NG4_retl_o2_plus_g1_plus_1)
 
 51:	LOAD(prefetch, %o1 + 0x040, #n_reads_strong)
 	LOAD(prefetch, %o1 + 0x080, #n_reads_strong)
@@ -154,43 +305,43 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	brz,pt		%g1, .Llarge_aligned
 	 sub		%o2, %g1, %o2
 
-1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g2))
+1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g2), NG4_retl_o2_plus_g1)
 	add		%o1, 8, %o1
 	subcc		%g1, 8, %g1
 	add		%o0, 8, %o0
 	bne,pt		%icc, 1b
-	 EX_ST(STORE(stx, %g2, %o0 - 0x08))
+	 EX_ST(STORE(stx, %g2, %o0 - 0x08), NG4_retl_o2_plus_g1_plus_8)
 
 .Llarge_aligned:
 	/* len >= 0x80 && src 8-byte aligned && dest 8-byte aligned */
 	andn		%o2, 0x3f, %o4
 	sub		%o2, %o4, %o2
 
-1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g1))
+1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g1), NG4_retl_o2_plus_o4)
 	add		%o1, 0x40, %o1
-	EX_LD(LOAD(ldx, %o1 - 0x38, %g2))
+	EX_LD(LOAD(ldx, %o1 - 0x38, %g2), NG4_retl_o2_plus_o4)
 	subcc		%o4, 0x40, %o4
-	EX_LD(LOAD(ldx, %o1 - 0x30, %g3))
-	EX_LD(LOAD(ldx, %o1 - 0x28, GLOBAL_SPARE))
-	EX_LD(LOAD(ldx, %o1 - 0x20, %o5))
-	EX_ST(STORE_INIT(%g1, %o0))
+	EX_LD(LOAD(ldx, %o1 - 0x30, %g3), NG4_retl_o2_plus_o4_plus_64)
+	EX_LD(LOAD(ldx, %o1 - 0x28, GLOBAL_SPARE), NG4_retl_o2_plus_o4_plus_64)
+	EX_LD(LOAD(ldx, %o1 - 0x20, %o5), NG4_retl_o2_plus_o4_plus_64)
+	EX_ST(STORE_INIT(%g1, %o0), NG4_retl_o2_plus_o4_plus_64)
 	add		%o0, 0x08, %o0
-	EX_ST(STORE_INIT(%g2, %o0))
+	EX_ST(STORE_INIT(%g2, %o0), NG4_retl_o2_plus_o4_plus_56)
 	add		%o0, 0x08, %o0
-	EX_LD(LOAD(ldx, %o1 - 0x18, %g2))
-	EX_ST(STORE_INIT(%g3, %o0))
+	EX_LD(LOAD(ldx, %o1 - 0x18, %g2), NG4_retl_o2_plus_o4_plus_48)
+	EX_ST(STORE_INIT(%g3, %o0), NG4_retl_o2_plus_o4_plus_48)
 	add		%o0, 0x08, %o0
-	EX_LD(LOAD(ldx, %o1 - 0x10, %g3))
-	EX_ST(STORE_INIT(GLOBAL_SPARE, %o0))
+	EX_LD(LOAD(ldx, %o1 - 0x10, %g3), NG4_retl_o2_plus_o4_plus_40)
+	EX_ST(STORE_INIT(GLOBAL_SPARE, %o0), NG4_retl_o2_plus_o4_plus_40)
 	add		%o0, 0x08, %o0
-	EX_LD(LOAD(ldx, %o1 - 0x08, GLOBAL_SPARE))
-	EX_ST(STORE_INIT(%o5, %o0))
+	EX_LD(LOAD(ldx, %o1 - 0x08, GLOBAL_SPARE), NG4_retl_o2_plus_o4_plus_32)
+	EX_ST(STORE_INIT(%o5, %o0), NG4_retl_o2_plus_o4_plus_32)
 	add		%o0, 0x08, %o0
-	EX_ST(STORE_INIT(%g2, %o0))
+	EX_ST(STORE_INIT(%g2, %o0), NG4_retl_o2_plus_o4_plus_24)
 	add		%o0, 0x08, %o0
-	EX_ST(STORE_INIT(%g3, %o0))
+	EX_ST(STORE_INIT(%g3, %o0), NG4_retl_o2_plus_o4_plus_16)
 	add		%o0, 0x08, %o0
-	EX_ST(STORE_INIT(GLOBAL_SPARE, %o0))
+	EX_ST(STORE_INIT(GLOBAL_SPARE, %o0), NG4_retl_o2_plus_o4_plus_8)
 	add		%o0, 0x08, %o0
 	bne,pt		%icc, 1b
 	 LOAD(prefetch, %o1 + 0x200, #n_reads_strong)
@@ -216,17 +367,17 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	sub		%o2, %o4, %o2
 	alignaddr	%o1, %g0, %g1
 	add		%o1, %o4, %o1
-	EX_LD_FP(LOAD(ldd, %g1 + 0x00, %f0))
-1:	EX_LD_FP(LOAD(ldd, %g1 + 0x08, %f2))
+	EX_LD_FP(LOAD(ldd, %g1 + 0x00, %f0), NG4_retl_o2_plus_o4)
+1:	EX_LD_FP(LOAD(ldd, %g1 + 0x08, %f2), NG4_retl_o2_plus_o4)
 	subcc		%o4, 0x40, %o4
-	EX_LD_FP(LOAD(ldd, %g1 + 0x10, %f4))
-	EX_LD_FP(LOAD(ldd, %g1 + 0x18, %f6))
-	EX_LD_FP(LOAD(ldd, %g1 + 0x20, %f8))
-	EX_LD_FP(LOAD(ldd, %g1 + 0x28, %f10))
-	EX_LD_FP(LOAD(ldd, %g1 + 0x30, %f12))
-	EX_LD_FP(LOAD(ldd, %g1 + 0x38, %f14))
+	EX_LD_FP(LOAD(ldd, %g1 + 0x10, %f4), NG4_retl_o2_plus_o4_plus_64)
+	EX_LD_FP(LOAD(ldd, %g1 + 0x18, %f6), NG4_retl_o2_plus_o4_plus_64)
+	EX_LD_FP(LOAD(ldd, %g1 + 0x20, %f8), NG4_retl_o2_plus_o4_plus_64)
+	EX_LD_FP(LOAD(ldd, %g1 + 0x28, %f10), NG4_retl_o2_plus_o4_plus_64)
+	EX_LD_FP(LOAD(ldd, %g1 + 0x30, %f12), NG4_retl_o2_plus_o4_plus_64)
+	EX_LD_FP(LOAD(ldd, %g1 + 0x38, %f14), NG4_retl_o2_plus_o4_plus_64)
 	faligndata	%f0, %f2, %f16
-	EX_LD_FP(LOAD(ldd, %g1 + 0x40, %f0))
+	EX_LD_FP(LOAD(ldd, %g1 + 0x40, %f0), NG4_retl_o2_plus_o4_plus_64)
 	faligndata	%f2, %f4, %f18
 	add		%g1, 0x40, %g1
 	faligndata	%f4, %f6, %f20
@@ -235,14 +386,14 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	faligndata	%f10, %f12, %f26
 	faligndata	%f12, %f14, %f28
 	faligndata	%f14, %f0, %f30
-	EX_ST_FP(STORE(std, %f16, %o0 + 0x00))
-	EX_ST_FP(STORE(std, %f18, %o0 + 0x08))
-	EX_ST_FP(STORE(std, %f20, %o0 + 0x10))
-	EX_ST_FP(STORE(std, %f22, %o0 + 0x18))
-	EX_ST_FP(STORE(std, %f24, %o0 + 0x20))
-	EX_ST_FP(STORE(std, %f26, %o0 + 0x28))
-	EX_ST_FP(STORE(std, %f28, %o0 + 0x30))
-	EX_ST_FP(STORE(std, %f30, %o0 + 0x38))
+	EX_ST_FP(STORE(std, %f16, %o0 + 0x00), NG4_retl_o2_plus_o4_plus_64)
+	EX_ST_FP(STORE(std, %f18, %o0 + 0x08), NG4_retl_o2_plus_o4_plus_56)
+	EX_ST_FP(STORE(std, %f20, %o0 + 0x10), NG4_retl_o2_plus_o4_plus_48)
+	EX_ST_FP(STORE(std, %f22, %o0 + 0x18), NG4_retl_o2_plus_o4_plus_40)
+	EX_ST_FP(STORE(std, %f24, %o0 + 0x20), NG4_retl_o2_plus_o4_plus_32)
+	EX_ST_FP(STORE(std, %f26, %o0 + 0x28), NG4_retl_o2_plus_o4_plus_24)
+	EX_ST_FP(STORE(std, %f28, %o0 + 0x30), NG4_retl_o2_plus_o4_plus_16)
+	EX_ST_FP(STORE(std, %f30, %o0 + 0x38), NG4_retl_o2_plus_o4_plus_8)
 	add		%o0, 0x40, %o0
 	bne,pt		%icc, 1b
 	 LOAD(prefetch, %g1 + 0x200, #n_reads_strong)
@@ -270,37 +421,38 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	andncc		%o2, 0x20 - 1, %o5
 	be,pn		%icc, 2f
 	 sub		%o2, %o5, %o2
-1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g1))
-	EX_LD(LOAD(ldx, %o1 + 0x08, %g2))
-	EX_LD(LOAD(ldx, %o1 + 0x10, GLOBAL_SPARE))
-	EX_LD(LOAD(ldx, %o1 + 0x18, %o4))
+1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g1), NG4_retl_o2_plus_o5)
+	EX_LD(LOAD(ldx, %o1 + 0x08, %g2), NG4_retl_o2_plus_o5)
+	EX_LD(LOAD(ldx, %o1 + 0x10, GLOBAL_SPARE), NG4_retl_o2_plus_o5)
+	EX_LD(LOAD(ldx, %o1 + 0x18, %o4), NG4_retl_o2_plus_o5)
 	add		%o1, 0x20, %o1
 	subcc		%o5, 0x20, %o5
-	EX_ST(STORE(stx, %g1, %o0 + 0x00))
-	EX_ST(STORE(stx, %g2, %o0 + 0x08))
-	EX_ST(STORE(stx, GLOBAL_SPARE, %o0 + 0x10))
-	EX_ST(STORE(stx, %o4, %o0 + 0x18))
+	EX_ST(STORE(stx, %g1, %o0 + 0x00), NG4_retl_o2_plus_o5_plus_32)
+	EX_ST(STORE(stx, %g2, %o0 + 0x08), NG4_retl_o2_plus_o5_plus_24)
+	EX_ST(STORE(stx, GLOBAL_SPARE, %o0 + 0x10), NG4_retl_o2_plus_o5_plus_24)
+	EX_ST(STORE(stx, %o4, %o0 + 0x18), NG4_retl_o2_plus_o5_plus_8)
 	bne,pt		%icc, 1b
 	 add		%o0, 0x20, %o0
 2:	andcc		%o2, 0x18, %o5
 	be,pt		%icc, 3f
 	 sub		%o2, %o5, %o2
-1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g1))
+
+1:	EX_LD(LOAD(ldx, %o1 + 0x00, %g1), NG4_retl_o2_plus_o5)
 	add		%o1, 0x08, %o1
 	add		%o0, 0x08, %o0
 	subcc		%o5, 0x08, %o5
 	bne,pt		%icc, 1b
-	 EX_ST(STORE(stx, %g1, %o0 - 0x08))
+	 EX_ST(STORE(stx, %g1, %o0 - 0x08), NG4_retl_o2_plus_o5_plus_8)
 3:	brz,pt		%o2, .Lexit
 	 cmp		%o2, 0x04
 	bl,pn		%icc, .Ltiny
 	 nop
-	EX_LD(LOAD(lduw, %o1 + 0x00, %g1))
+	EX_LD(LOAD(lduw, %o1 + 0x00, %g1), NG4_retl_o2)
 	add		%o1, 0x04, %o1
 	add		%o0, 0x04, %o0
 	subcc		%o2, 0x04, %o2
 	bne,pn		%icc, .Ltiny
-	 EX_ST(STORE(stw, %g1, %o0 - 0x04))
+	 EX_ST(STORE(stw, %g1, %o0 - 0x04), NG4_retl_o2_plus_4)
 	ba,a,pt		%icc, .Lexit
 .Lmedium_unaligned:
 	/* First get dest 8 byte aligned.  */
@@ -309,12 +461,12 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	brz,pt		%g1, 2f
 	 sub		%o2, %g1, %o2
 
-1:	EX_LD(LOAD(ldub, %o1 + 0x00, %g2))
+1:	EX_LD(LOAD(ldub, %o1 + 0x00, %g2), NG4_retl_o2_plus_g1)
 	add		%o1, 1, %o1
 	subcc		%g1, 1, %g1
 	add		%o0, 1, %o0
 	bne,pt		%icc, 1b
-	 EX_ST(STORE(stb, %g2, %o0 - 0x01))
+	 EX_ST(STORE(stb, %g2, %o0 - 0x01), NG4_retl_o2_plus_g1_plus_1)
 2:
 	and		%o1, 0x7, %g1
 	brz,pn		%g1, .Lmedium_noprefetch
@@ -322,16 +474,16 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	mov		64, %g2
 	sub		%g2, %g1, %g2
 	andn		%o1, 0x7, %o1
-	EX_LD(LOAD(ldx, %o1 + 0x00, %o4))
+	EX_LD(LOAD(ldx, %o1 + 0x00, %o4), NG4_retl_o2)
 	sllx		%o4, %g1, %o4
 	andn		%o2, 0x08 - 1, %o5
 	sub		%o2, %o5, %o2
-1:	EX_LD(LOAD(ldx, %o1 + 0x08, %g3))
+1:	EX_LD(LOAD(ldx, %o1 + 0x08, %g3), NG4_retl_o2_plus_o5)
 	add		%o1, 0x08, %o1
 	subcc		%o5, 0x08, %o5
 	srlx		%g3, %g2, GLOBAL_SPARE
 	or		GLOBAL_SPARE, %o4, GLOBAL_SPARE
-	EX_ST(STORE(stx, GLOBAL_SPARE, %o0 + 0x00))
+	EX_ST(STORE(stx, GLOBAL_SPARE, %o0 + 0x00), NG4_retl_o2_plus_o5_plus_8)
 	add		%o0, 0x08, %o0
 	bne,pt		%icc, 1b
 	 sllx		%g3, %g1, %o4
@@ -342,17 +494,17 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	ba,pt		%icc, .Lsmall_unaligned
 
 .Ltiny:
-	EX_LD(LOAD(ldub, %o1 + 0x00, %g1))
+	EX_LD(LOAD(ldub, %o1 + 0x00, %g1), NG4_retl_o2)
 	subcc		%o2, 1, %o2
 	be,pn		%icc, .Lexit
-	 EX_ST(STORE(stb, %g1, %o0 + 0x00))
-	EX_LD(LOAD(ldub, %o1 + 0x01, %g1))
+	 EX_ST(STORE(stb, %g1, %o0 + 0x00), NG4_retl_o2_plus_1)
+	EX_LD(LOAD(ldub, %o1 + 0x01, %g1), NG4_retl_o2)
 	subcc		%o2, 1, %o2
 	be,pn		%icc, .Lexit
-	 EX_ST(STORE(stb, %g1, %o0 + 0x01))
-	EX_LD(LOAD(ldub, %o1 + 0x02, %g1))
+	 EX_ST(STORE(stb, %g1, %o0 + 0x01), NG4_retl_o2_plus_1)
+	EX_LD(LOAD(ldub, %o1 + 0x02, %g1), NG4_retl_o2)
 	ba,pt		%icc, .Lexit
-	 EX_ST(STORE(stb, %g1, %o0 + 0x02))
+	 EX_ST(STORE(stb, %g1, %o0 + 0x02), NG4_retl_o2)
 
 .Lsmall:
 	andcc		%g2, 0x3, %g0
@@ -360,22 +512,22 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	 andn		%o2, 0x4 - 1, %o5
 	sub		%o2, %o5, %o2
 1:
-	EX_LD(LOAD(lduw, %o1 + 0x00, %g1))
+	EX_LD(LOAD(lduw, %o1 + 0x00, %g1), NG4_retl_o2_plus_o5)
 	add		%o1, 0x04, %o1
 	subcc		%o5, 0x04, %o5
 	add		%o0, 0x04, %o0
 	bne,pt		%icc, 1b
-	 EX_ST(STORE(stw, %g1, %o0 - 0x04))
+	 EX_ST(STORE(stw, %g1, %o0 - 0x04), NG4_retl_o2_plus_o5_plus_4)
 	brz,pt		%o2, .Lexit
 	 nop
 	ba,a,pt		%icc, .Ltiny
 
 .Lsmall_unaligned:
-1:	EX_LD(LOAD(ldub, %o1 + 0x00, %g1))
+1:	EX_LD(LOAD(ldub, %o1 + 0x00, %g1), NG4_retl_o2)
 	add		%o1, 1, %o1
 	add		%o0, 1, %o0
 	subcc		%o2, 1, %o2
 	bne,pt		%icc, 1b
-	 EX_ST(STORE(stb, %g1, %o0 - 0x01))
+	 EX_ST(STORE(stb, %g1, %o0 - 0x01), NG4_retl_o2_plus_1)
 	ba,a,pt		%icc, .Lexit
 	.size		FUNC_NAME, .-FUNC_NAME
diff --git a/arch/sparc/lib/NGcopy_from_user.S b/arch/sparc/lib/NGcopy_from_user.S
index 5d1e4d1ac21e..9cd42fcbc781 100644
--- a/arch/sparc/lib/NGcopy_from_user.S
+++ b/arch/sparc/lib/NGcopy_from_user.S
@@ -3,11 +3,11 @@
  * Copyright (C) 2006, 2007 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_LD(x)		\
+#define EX_LD(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __ret_one_asi;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/NGcopy_to_user.S b/arch/sparc/lib/NGcopy_to_user.S
index ff630dcb273c..5c358afd464e 100644
--- a/arch/sparc/lib/NGcopy_to_user.S
+++ b/arch/sparc/lib/NGcopy_to_user.S
@@ -3,11 +3,11 @@
  * Copyright (C) 2006, 2007 David S. Miller (davem@davemloft.net)
  */
 
-#define EX_ST(x)		\
+#define EX_ST(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __ret_one_asi;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/NGmemcpy.S b/arch/sparc/lib/NGmemcpy.S
index 96a14caf6966..d88c4ed50a00 100644
--- a/arch/sparc/lib/NGmemcpy.S
+++ b/arch/sparc/lib/NGmemcpy.S
@@ -4,6 +4,7 @@
  */
 
 #ifdef __KERNEL__
+#include <linux/linkage.h>
 #include <asm/asi.h>
 #include <asm/thread_info.h>
 #define GLOBAL_SPARE	%g7
@@ -27,15 +28,11 @@
 #endif
 
 #ifndef EX_LD
-#define EX_LD(x)	x
+#define EX_LD(x,y)	x
 #endif
 
 #ifndef EX_ST
-#define EX_ST(x)	x
-#endif
-
-#ifndef EX_RETVAL
-#define EX_RETVAL(x)	x
+#define EX_ST(x,y)	x
 #endif
 
 #ifndef LOAD
@@ -79,6 +76,92 @@
 	.register	%g3,#scratch
 
 	.text
+#ifndef EX_RETVAL
+#define EX_RETVAL(x)	x
+__restore_asi:
+	ret
+	wr	%g0, ASI_AIUS, %asi
+	 restore
+ENTRY(NG_ret_i2_plus_i4_plus_1)
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %i5, %i0
+ENDPROC(NG_ret_i2_plus_i4_plus_1)
+ENTRY(NG_ret_i2_plus_g1)
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1)
+ENTRY(NG_ret_i2_plus_g1_minus_8)
+	sub	%g1, 8, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_minus_8)
+ENTRY(NG_ret_i2_plus_g1_minus_16)
+	sub	%g1, 16, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_minus_16)
+ENTRY(NG_ret_i2_plus_g1_minus_24)
+	sub	%g1, 24, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_minus_24)
+ENTRY(NG_ret_i2_plus_g1_minus_32)
+	sub	%g1, 32, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_minus_32)
+ENTRY(NG_ret_i2_plus_g1_minus_40)
+	sub	%g1, 40, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_minus_40)
+ENTRY(NG_ret_i2_plus_g1_minus_48)
+	sub	%g1, 48, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_minus_48)
+ENTRY(NG_ret_i2_plus_g1_minus_56)
+	sub	%g1, 56, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_minus_56)
+ENTRY(NG_ret_i2_plus_i4)
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %i4, %i0
+ENDPROC(NG_ret_i2_plus_i4)
+ENTRY(NG_ret_i2_plus_i4_minus_8)
+	sub	%i4, 8, %i4
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %i4, %i0
+ENDPROC(NG_ret_i2_plus_i4_minus_8)
+ENTRY(NG_ret_i2_plus_8)
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, 8, %i0
+ENDPROC(NG_ret_i2_plus_8)
+ENTRY(NG_ret_i2_plus_4)
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, 4, %i0
+ENDPROC(NG_ret_i2_plus_4)
+ENTRY(NG_ret_i2_plus_1)
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, 1, %i0
+ENDPROC(NG_ret_i2_plus_1)
+ENTRY(NG_ret_i2_plus_g1_plus_1)
+	add	%g1, 1, %g1
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %g1, %i0
+ENDPROC(NG_ret_i2_plus_g1_plus_1)
+ENTRY(NG_ret_i2)
+	ba,pt	%xcc, __restore_asi
+	 mov	%i2, %i0
+ENDPROC(NG_ret_i2)
+ENTRY(NG_ret_i2_and_7_plus_i4)
+	and	%i2, 7, %i2
+	ba,pt	%xcc, __restore_asi
+	 add	%i2, %i4, %i0
+ENDPROC(NG_ret_i2_and_7_plus_i4)
+#endif
+
 	.align		64
 
 	.globl	FUNC_NAME
@@ -126,8 +209,8 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	sub		%g0, %i4, %i4	! bytes to align dst
 	sub		%i2, %i4, %i2
 1:	subcc		%i4, 1, %i4
-	EX_LD(LOAD(ldub, %i1, %g1))
-	EX_ST(STORE(stb, %g1, %o0))
+	EX_LD(LOAD(ldub, %i1, %g1), NG_ret_i2_plus_i4_plus_1)
+	EX_ST(STORE(stb, %g1, %o0), NG_ret_i2_plus_i4_plus_1)
 	add		%i1, 1, %i1
 	bne,pt		%XCC, 1b
 	add		%o0, 1, %o0
@@ -160,7 +243,7 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	and		%i4, 0x7, GLOBAL_SPARE
 	sll		GLOBAL_SPARE, 3, GLOBAL_SPARE
 	mov		64, %i5
-	EX_LD(LOAD_TWIN(%i1, %g2, %g3))
+	EX_LD(LOAD_TWIN(%i1, %g2, %g3), NG_ret_i2_plus_g1)
 	sub		%i5, GLOBAL_SPARE, %i5
 	mov		16, %o4
 	mov		32, %o5
@@ -178,31 +261,31 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	srlx		WORD3, PRE_SHIFT, TMP; \
 	or		WORD2, TMP, WORD2;
 
-8:	EX_LD(LOAD_TWIN(%i1 + %o4, %o2, %o3))
+8:	EX_LD(LOAD_TWIN(%i1 + %o4, %o2, %o3), NG_ret_i2_plus_g1)
 	MIX_THREE_WORDS(%g2, %g3, %o2, %i5, GLOBAL_SPARE, %o1)
 	LOAD(prefetch, %i1 + %i3, #one_read)
 
-	EX_ST(STORE_INIT(%g2, %o0 + 0x00))
-	EX_ST(STORE_INIT(%g3, %o0 + 0x08))
+	EX_ST(STORE_INIT(%g2, %o0 + 0x00), NG_ret_i2_plus_g1)
+	EX_ST(STORE_INIT(%g3, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8)
 
-	EX_LD(LOAD_TWIN(%i1 + %o5, %g2, %g3))
+	EX_LD(LOAD_TWIN(%i1 + %o5, %g2, %g3), NG_ret_i2_plus_g1_minus_16)
 	MIX_THREE_WORDS(%o2, %o3, %g2, %i5, GLOBAL_SPARE, %o1)
 
-	EX_ST(STORE_INIT(%o2, %o0 + 0x10))
-	EX_ST(STORE_INIT(%o3, %o0 + 0x18))
+	EX_ST(STORE_INIT(%o2, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16)
+	EX_ST(STORE_INIT(%o3, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24)
 
-	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3))
+	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1_minus_32)
 	MIX_THREE_WORDS(%g2, %g3, %o2, %i5, GLOBAL_SPARE, %o1)
 
-	EX_ST(STORE_INIT(%g2, %o0 + 0x20))
-	EX_ST(STORE_INIT(%g3, %o0 + 0x28))
+	EX_ST(STORE_INIT(%g2, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32)
+	EX_ST(STORE_INIT(%g3, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40)
 
-	EX_LD(LOAD_TWIN(%i1 + %i3, %g2, %g3))
+	EX_LD(LOAD_TWIN(%i1 + %i3, %g2, %g3), NG_ret_i2_plus_g1_minus_48)
 	add		%i1, 64, %i1
 	MIX_THREE_WORDS(%o2, %o3, %g2, %i5, GLOBAL_SPARE, %o1)
 
-	EX_ST(STORE_INIT(%o2, %o0 + 0x30))
-	EX_ST(STORE_INIT(%o3, %o0 + 0x38))
+	EX_ST(STORE_INIT(%o2, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48)
+	EX_ST(STORE_INIT(%o3, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56)
 
 	subcc		%g1, 64, %g1
 	bne,pt		%XCC, 8b
@@ -211,31 +294,31 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	ba,pt		%XCC, 60f
 	 add		%i1, %i4, %i1
 
-9:	EX_LD(LOAD_TWIN(%i1 + %o4, %o2, %o3))
+9:	EX_LD(LOAD_TWIN(%i1 + %o4, %o2, %o3), NG_ret_i2_plus_g1)
 	MIX_THREE_WORDS(%g3, %o2, %o3, %i5, GLOBAL_SPARE, %o1)
 	LOAD(prefetch, %i1 + %i3, #one_read)
 
-	EX_ST(STORE_INIT(%g3, %o0 + 0x00))
-	EX_ST(STORE_INIT(%o2, %o0 + 0x08))
+	EX_ST(STORE_INIT(%g3, %o0 + 0x00), NG_ret_i2_plus_g1)
+	EX_ST(STORE_INIT(%o2, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8)
 
-	EX_LD(LOAD_TWIN(%i1 + %o5, %g2, %g3))
+	EX_LD(LOAD_TWIN(%i1 + %o5, %g2, %g3), NG_ret_i2_plus_g1_minus_16)
 	MIX_THREE_WORDS(%o3, %g2, %g3, %i5, GLOBAL_SPARE, %o1)
 
-	EX_ST(STORE_INIT(%o3, %o0 + 0x10))
-	EX_ST(STORE_INIT(%g2, %o0 + 0x18))
+	EX_ST(STORE_INIT(%o3, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16)
+	EX_ST(STORE_INIT(%g2, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24)
 
-	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3))
+	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1_minus_32)
 	MIX_THREE_WORDS(%g3, %o2, %o3, %i5, GLOBAL_SPARE, %o1)
 
-	EX_ST(STORE_INIT(%g3, %o0 + 0x20))
-	EX_ST(STORE_INIT(%o2, %o0 + 0x28))
+	EX_ST(STORE_INIT(%g3, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32)
+	EX_ST(STORE_INIT(%o2, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40)
 
-	EX_LD(LOAD_TWIN(%i1 + %i3, %g2, %g3))
+	EX_LD(LOAD_TWIN(%i1 + %i3, %g2, %g3), NG_ret_i2_plus_g1_minus_48)
 	add		%i1, 64, %i1
 	MIX_THREE_WORDS(%o3, %g2, %g3, %i5, GLOBAL_SPARE, %o1)
 
-	EX_ST(STORE_INIT(%o3, %o0 + 0x30))
-	EX_ST(STORE_INIT(%g2, %o0 + 0x38))
+	EX_ST(STORE_INIT(%o3, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48)
+	EX_ST(STORE_INIT(%g2, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56)
 
 	subcc		%g1, 64, %g1
 	bne,pt		%XCC, 9b
@@ -249,25 +332,25 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	 * one twin load ahead, then add 8 back into source when
 	 * we finish the loop.
 	 */
-	EX_LD(LOAD_TWIN(%i1, %o4, %o5))
+	EX_LD(LOAD_TWIN(%i1, %o4, %o5), NG_ret_i2_plus_g1)
 	mov	16, %o7
 	mov	32, %g2
 	mov	48, %g3
 	mov	64, %o1
-1:	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3))
+1:	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1)
 	LOAD(prefetch, %i1 + %o1, #one_read)
-	EX_ST(STORE_INIT(%o5, %o0 + 0x00))	! initializes cache line
-	EX_ST(STORE_INIT(%o2, %o0 + 0x08))
-	EX_LD(LOAD_TWIN(%i1 + %g2, %o4, %o5))
-	EX_ST(STORE_INIT(%o3, %o0 + 0x10))
-	EX_ST(STORE_INIT(%o4, %o0 + 0x18))
-	EX_LD(LOAD_TWIN(%i1 + %g3, %o2, %o3))
-	EX_ST(STORE_INIT(%o5, %o0 + 0x20))
-	EX_ST(STORE_INIT(%o2, %o0 + 0x28))
-	EX_LD(LOAD_TWIN(%i1 + %o1, %o4, %o5))
+	EX_ST(STORE_INIT(%o5, %o0 + 0x00), NG_ret_i2_plus_g1)	! initializes cache line
+	EX_ST(STORE_INIT(%o2, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8)
+	EX_LD(LOAD_TWIN(%i1 + %g2, %o4, %o5), NG_ret_i2_plus_g1_minus_16)
+	EX_ST(STORE_INIT(%o3, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16)
+	EX_ST(STORE_INIT(%o4, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24)
+	EX_LD(LOAD_TWIN(%i1 + %g3, %o2, %o3), NG_ret_i2_plus_g1_minus_32)
+	EX_ST(STORE_INIT(%o5, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32)
+	EX_ST(STORE_INIT(%o2, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40)
+	EX_LD(LOAD_TWIN(%i1 + %o1, %o4, %o5), NG_ret_i2_plus_g1_minus_48)
 	add		%i1, 64, %i1
-	EX_ST(STORE_INIT(%o3, %o0 + 0x30))
-	EX_ST(STORE_INIT(%o4, %o0 + 0x38))
+	EX_ST(STORE_INIT(%o3, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48)
+	EX_ST(STORE_INIT(%o4, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56)
 	subcc		%g1, 64, %g1
 	bne,pt		%XCC, 1b
 	 add		%o0, 64, %o0
@@ -282,20 +365,20 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	mov	32, %g2
 	mov	48, %g3
 	mov	64, %o1
-1:	EX_LD(LOAD_TWIN(%i1 + %g0, %o4, %o5))
-	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3))
+1:	EX_LD(LOAD_TWIN(%i1 + %g0, %o4, %o5), NG_ret_i2_plus_g1)
+	EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1)
 	LOAD(prefetch, %i1 + %o1, #one_read)
-	EX_ST(STORE_INIT(%o4, %o0 + 0x00))	! initializes cache line
-	EX_ST(STORE_INIT(%o5, %o0 + 0x08))
-	EX_LD(LOAD_TWIN(%i1 + %g2, %o4, %o5))
-	EX_ST(STORE_INIT(%o2, %o0 + 0x10))
-	EX_ST(STORE_INIT(%o3, %o0 + 0x18))
-	EX_LD(LOAD_TWIN(%i1 + %g3, %o2, %o3))
+	EX_ST(STORE_INIT(%o4, %o0 + 0x00), NG_ret_i2_plus_g1)	! initializes cache line
+	EX_ST(STORE_INIT(%o5, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8)
+	EX_LD(LOAD_TWIN(%i1 + %g2, %o4, %o5), NG_ret_i2_plus_g1_minus_16)
+	EX_ST(STORE_INIT(%o2, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16)
+	EX_ST(STORE_INIT(%o3, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24)
+	EX_LD(LOAD_TWIN(%i1 + %g3, %o2, %o3), NG_ret_i2_plus_g1_minus_32)
 	add	%i1, 64, %i1
-	EX_ST(STORE_INIT(%o4, %o0 + 0x20))
-	EX_ST(STORE_INIT(%o5, %o0 + 0x28))
-	EX_ST(STORE_INIT(%o2, %o0 + 0x30))
-	EX_ST(STORE_INIT(%o3, %o0 + 0x38))
+	EX_ST(STORE_INIT(%o4, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32)
+	EX_ST(STORE_INIT(%o5, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40)
+	EX_ST(STORE_INIT(%o2, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48)
+	EX_ST(STORE_INIT(%o3, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56)
 	subcc	%g1, 64, %g1
 	bne,pt	%XCC, 1b
 	 add	%o0, 64, %o0
@@ -321,28 +404,28 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	andn		%i2, 0xf, %i4
 	and		%i2, 0xf, %i2
 1:	subcc		%i4, 0x10, %i4
-	EX_LD(LOAD(ldx, %i1, %o4))
+	EX_LD(LOAD(ldx, %i1, %o4), NG_ret_i2_plus_i4)
 	add		%i1, 0x08, %i1
-	EX_LD(LOAD(ldx, %i1, %g1))
+	EX_LD(LOAD(ldx, %i1, %g1), NG_ret_i2_plus_i4)
 	sub		%i1, 0x08, %i1
-	EX_ST(STORE(stx, %o4, %i1 + %i3))
+	EX_ST(STORE(stx, %o4, %i1 + %i3), NG_ret_i2_plus_i4)
 	add		%i1, 0x8, %i1
-	EX_ST(STORE(stx, %g1, %i1 + %i3))
+	EX_ST(STORE(stx, %g1, %i1 + %i3), NG_ret_i2_plus_i4_minus_8)
 	bgu,pt		%XCC, 1b
 	 add		%i1, 0x8, %i1
 73:	andcc		%i2, 0x8, %g0
 	be,pt		%XCC, 1f
 	 nop
 	sub		%i2, 0x8, %i2
-	EX_LD(LOAD(ldx, %i1, %o4))
-	EX_ST(STORE(stx, %o4, %i1 + %i3))
+	EX_LD(LOAD(ldx, %i1, %o4), NG_ret_i2_plus_8)
+	EX_ST(STORE(stx, %o4, %i1 + %i3), NG_ret_i2_plus_8)
 	add		%i1, 0x8, %i1
 1:	andcc		%i2, 0x4, %g0
 	be,pt		%XCC, 1f
 	 nop
 	sub		%i2, 0x4, %i2
-	EX_LD(LOAD(lduw, %i1, %i5))
-	EX_ST(STORE(stw, %i5, %i1 + %i3))
+	EX_LD(LOAD(lduw, %i1, %i5), NG_ret_i2_plus_4)
+	EX_ST(STORE(stw, %i5, %i1 + %i3), NG_ret_i2_plus_4)
 	add		%i1, 0x4, %i1
 1:	cmp		%i2, 0
 	be,pt		%XCC, 85f
@@ -358,8 +441,8 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	sub		%i2, %g1, %i2
 
 1:	subcc		%g1, 1, %g1
-	EX_LD(LOAD(ldub, %i1, %i5))
-	EX_ST(STORE(stb, %i5, %i1 + %i3))
+	EX_LD(LOAD(ldub, %i1, %i5), NG_ret_i2_plus_g1_plus_1)
+	EX_ST(STORE(stb, %i5, %i1 + %i3), NG_ret_i2_plus_g1_plus_1)
 	bgu,pt		%icc, 1b
 	 add		%i1, 1, %i1
 
@@ -375,16 +458,16 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 
 8:	mov		64, %i3
 	andn		%i1, 0x7, %i1
-	EX_LD(LOAD(ldx, %i1, %g2))
+	EX_LD(LOAD(ldx, %i1, %g2), NG_ret_i2)
 	sub		%i3, %g1, %i3
 	andn		%i2, 0x7, %i4
 	sllx		%g2, %g1, %g2
 1:	add		%i1, 0x8, %i1
-	EX_LD(LOAD(ldx, %i1, %g3))
+	EX_LD(LOAD(ldx, %i1, %g3), NG_ret_i2_and_7_plus_i4)
 	subcc		%i4, 0x8, %i4
 	srlx		%g3, %i3, %i5
 	or		%i5, %g2, %i5
-	EX_ST(STORE(stx, %i5, %o0))
+	EX_ST(STORE(stx, %i5, %o0), NG_ret_i2_and_7_plus_i4)
 	add		%o0, 0x8, %o0
 	bgu,pt		%icc, 1b
 	 sllx		%g3, %g1, %g2
@@ -404,8 +487,8 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 
 1:
 	subcc		%i2, 4, %i2
-	EX_LD(LOAD(lduw, %i1, %g1))
-	EX_ST(STORE(stw, %g1, %i1 + %i3))
+	EX_LD(LOAD(lduw, %i1, %g1), NG_ret_i2_plus_4)
+	EX_ST(STORE(stw, %g1, %i1 + %i3), NG_ret_i2_plus_4)
 	bgu,pt		%XCC, 1b
 	 add		%i1, 4, %i1
 
@@ -415,8 +498,8 @@ FUNC_NAME:	/* %i0=dst, %i1=src, %i2=len */
 	.align		32
 90:
 	subcc		%i2, 1, %i2
-	EX_LD(LOAD(ldub, %i1, %g1))
-	EX_ST(STORE(stb, %g1, %i1 + %i3))
+	EX_LD(LOAD(ldub, %i1, %g1), NG_ret_i2_plus_1)
+	EX_ST(STORE(stb, %g1, %i1 + %i3), NG_ret_i2_plus_1)
 	bgu,pt		%XCC, 90b
 	 add		%i1, 1, %i1
 	ret
diff --git a/arch/sparc/lib/U1copy_from_user.S b/arch/sparc/lib/U1copy_from_user.S
index ecc5692fa2b4..bb6ff73229e3 100644
--- a/arch/sparc/lib/U1copy_from_user.S
+++ b/arch/sparc/lib/U1copy_from_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 1999, 2000, 2004 David S. Miller (davem@redhat.com)
  */
 
-#define EX_LD(x)		\
+#define EX_LD(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one;	\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_LD_FP(x)		\
+#define EX_LD_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_fp;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/U1copy_to_user.S b/arch/sparc/lib/U1copy_to_user.S
index 9eea392e44d4..ed92ce739558 100644
--- a/arch/sparc/lib/U1copy_to_user.S
+++ b/arch/sparc/lib/U1copy_to_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 1999, 2000, 2004 David S. Miller (davem@redhat.com)
  */
 
-#define EX_ST(x)		\
+#define EX_ST(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one;	\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_ST_FP(x)		\
+#define EX_ST_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_fp;\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/U1memcpy.S b/arch/sparc/lib/U1memcpy.S
index 97e1b211090c..4f0d50b33a72 100644
--- a/arch/sparc/lib/U1memcpy.S
+++ b/arch/sparc/lib/U1memcpy.S
@@ -5,6 +5,7 @@
  */
 
 #ifdef __KERNEL__
+#include <linux/linkage.h>
 #include <asm/visasm.h>
 #include <asm/asi.h>
 #include <asm/export.h>
@@ -24,21 +25,17 @@
 #endif
 
 #ifndef EX_LD
-#define EX_LD(x)	x
+#define EX_LD(x,y)	x
 #endif
 #ifndef EX_LD_FP
-#define EX_LD_FP(x)	x
+#define EX_LD_FP(x,y)	x
 #endif
 
 #ifndef EX_ST
-#define EX_ST(x)	x
+#define EX_ST(x,y)	x
 #endif
 #ifndef EX_ST_FP
-#define EX_ST_FP(x)	x
-#endif
-
-#ifndef EX_RETVAL
-#define EX_RETVAL(x)	x
+#define EX_ST_FP(x,y)	x
 #endif
 
 #ifndef LOAD
@@ -79,53 +76,169 @@
 	faligndata		%f7, %f8, %f60;			\
 	faligndata		%f8, %f9, %f62;
 
-#define MAIN_LOOP_CHUNK(src, dest, fdest, fsrc, len, jmptgt)	\
-	EX_LD_FP(LOAD_BLK(%src, %fdest));				\
-	EX_ST_FP(STORE_BLK(%fsrc, %dest));				\
-	add			%src, 0x40, %src;		\
-	subcc			%len, 0x40, %len;		\
-	be,pn			%xcc, jmptgt;			\
-	 add			%dest, 0x40, %dest;		\
-
-#define LOOP_CHUNK1(src, dest, len, branch_dest)		\
-	MAIN_LOOP_CHUNK(src, dest, f0,  f48, len, branch_dest)
-#define LOOP_CHUNK2(src, dest, len, branch_dest)		\
-	MAIN_LOOP_CHUNK(src, dest, f16, f48, len, branch_dest)
-#define LOOP_CHUNK3(src, dest, len, branch_dest)		\
-	MAIN_LOOP_CHUNK(src, dest, f32, f48, len, branch_dest)
+#define MAIN_LOOP_CHUNK(src, dest, fdest, fsrc, jmptgt)			\
+	EX_LD_FP(LOAD_BLK(%src, %fdest), U1_gs_80_fp);			\
+	EX_ST_FP(STORE_BLK(%fsrc, %dest), U1_gs_80_fp);			\
+	add			%src, 0x40, %src;			\
+	subcc			%GLOBAL_SPARE, 0x40, %GLOBAL_SPARE;	\
+	be,pn			%xcc, jmptgt;				\
+	 add			%dest, 0x40, %dest;			\
+
+#define LOOP_CHUNK1(src, dest, branch_dest)		\
+	MAIN_LOOP_CHUNK(src, dest, f0,  f48, branch_dest)
+#define LOOP_CHUNK2(src, dest, branch_dest)		\
+	MAIN_LOOP_CHUNK(src, dest, f16, f48, branch_dest)
+#define LOOP_CHUNK3(src, dest, branch_dest)		\
+	MAIN_LOOP_CHUNK(src, dest, f32, f48, branch_dest)
 
 #define DO_SYNC			membar	#Sync;
 #define STORE_SYNC(dest, fsrc)				\
-	EX_ST_FP(STORE_BLK(%fsrc, %dest));			\
+	EX_ST_FP(STORE_BLK(%fsrc, %dest), U1_gs_80_fp);	\
 	add			%dest, 0x40, %dest;	\
 	DO_SYNC
 
 #define STORE_JUMP(dest, fsrc, target)			\
-	EX_ST_FP(STORE_BLK(%fsrc, %dest));			\
+	EX_ST_FP(STORE_BLK(%fsrc, %dest), U1_gs_40_fp);	\
 	add			%dest, 0x40, %dest;	\
 	ba,pt			%xcc, target;		\
 	 nop;
 
-#define FINISH_VISCHUNK(dest, f0, f1, left)	\
-	subcc			%left, 8, %left;\
-	bl,pn			%xcc, 95f;	\
-	 faligndata		%f0, %f1, %f48;	\
-	EX_ST_FP(STORE(std, %f48, %dest));		\
+#define FINISH_VISCHUNK(dest, f0, f1)			\
+	subcc			%g3, 8, %g3;		\
+	bl,pn			%xcc, 95f;		\
+	 faligndata		%f0, %f1, %f48;		\
+	EX_ST_FP(STORE(std, %f48, %dest), U1_g3_8_fp);	\
 	add			%dest, 8, %dest;
 
-#define UNEVEN_VISCHUNK_LAST(dest, f0, f1, left)	\
-	subcc			%left, 8, %left;	\
-	bl,pn			%xcc, 95f;		\
+#define UNEVEN_VISCHUNK_LAST(dest, f0, f1)	\
+	subcc			%g3, 8, %g3;	\
+	bl,pn			%xcc, 95f;	\
 	 fsrc2			%f0, %f1;
 
-#define UNEVEN_VISCHUNK(dest, f0, f1, left)		\
-	UNEVEN_VISCHUNK_LAST(dest, f0, f1, left)	\
+#define UNEVEN_VISCHUNK(dest, f0, f1)		\
+	UNEVEN_VISCHUNK_LAST(dest, f0, f1)	\
 	ba,a,pt			%xcc, 93f;
 
 	.register	%g2,#scratch
 	.register	%g3,#scratch
 
 	.text
+#ifndef EX_RETVAL
+#define EX_RETVAL(x)	x
+ENTRY(U1_g1_1_fp)
+	VISExitHalf
+	add		%g1, 1, %g1
+	add		%g1, %g2, %g1
+	retl
+	 add		%g1, %o2, %o0
+ENDPROC(U1_g1_1_fp)
+ENTRY(U1_g2_0_fp)
+	VISExitHalf
+	retl
+	 add		%g2, %o2, %o0
+ENDPROC(U1_g2_0_fp)
+ENTRY(U1_g2_8_fp)
+	VISExitHalf
+	add		%g2, 8, %g2
+	retl
+	 add		%g2, %o2, %o0
+ENDPROC(U1_g2_8_fp)
+ENTRY(U1_gs_0_fp)
+	VISExitHalf
+	add		%GLOBAL_SPARE, %g3, %o0
+	retl
+	 add		%o0, %o2, %o0
+ENDPROC(U1_gs_0_fp)
+ENTRY(U1_gs_80_fp)
+	VISExitHalf
+	add		%GLOBAL_SPARE, 0x80, %GLOBAL_SPARE
+	add		%GLOBAL_SPARE, %g3, %o0
+	retl
+	 add		%o0, %o2, %o0
+ENDPROC(U1_gs_80_fp)
+ENTRY(U1_gs_40_fp)
+	VISExitHalf
+	add		%GLOBAL_SPARE, 0x40, %GLOBAL_SPARE
+	add		%GLOBAL_SPARE, %g3, %o0
+	retl
+	 add		%o0, %o2, %o0
+ENDPROC(U1_gs_40_fp)
+ENTRY(U1_g3_0_fp)
+	VISExitHalf
+	retl
+	 add		%g3, %o2, %o0
+ENDPROC(U1_g3_0_fp)
+ENTRY(U1_g3_8_fp)
+	VISExitHalf
+	add		%g3, 8, %g3
+	retl
+	 add		%g3, %o2, %o0
+ENDPROC(U1_g3_8_fp)
+ENTRY(U1_o2_0_fp)
+	VISExitHalf
+	retl
+	 mov		%o2, %o0
+ENDPROC(U1_o2_0_fp)
+ENTRY(U1_o2_1_fp)
+	VISExitHalf
+	retl
+	 add		%o2, 1, %o0
+ENDPROC(U1_o2_1_fp)
+ENTRY(U1_gs_0)
+	VISExitHalf
+	retl
+	 add		%GLOBAL_SPARE, %o2, %o0
+ENDPROC(U1_gs_0)
+ENTRY(U1_gs_8)
+	VISExitHalf
+	add		%GLOBAL_SPARE, %o2, %GLOBAL_SPARE
+	retl
+	 add		%GLOBAL_SPARE, 0x8, %o0
+ENDPROC(U1_gs_8)
+ENTRY(U1_gs_10)
+	VISExitHalf
+	add		%GLOBAL_SPARE, %o2, %GLOBAL_SPARE
+	retl
+	 add		%GLOBAL_SPARE, 0x10, %o0
+ENDPROC(U1_gs_10)
+ENTRY(U1_o2_0)
+	retl
+	 mov		%o2, %o0
+ENDPROC(U1_o2_0)
+ENTRY(U1_o2_8)
+	retl
+	 add		%o2, 8, %o0
+ENDPROC(U1_o2_8)
+ENTRY(U1_o2_4)
+	retl
+	 add		%o2, 4, %o0
+ENDPROC(U1_o2_4)
+ENTRY(U1_o2_1)
+	retl
+	 add		%o2, 1, %o0
+ENDPROC(U1_o2_1)
+ENTRY(U1_g1_0)
+	retl
+	 add		%g1, %o2, %o0
+ENDPROC(U1_g1_0)
+ENTRY(U1_g1_1)
+	add		%g1, 1, %g1
+	retl
+	 add		%g1, %o2, %o0
+ENDPROC(U1_g1_1)
+ENTRY(U1_gs_0_o2_adj)
+	and		%o2, 7, %o2
+	retl
+	 add		%GLOBAL_SPARE, %o2, %o0
+ENDPROC(U1_gs_0_o2_adj)
+ENTRY(U1_gs_8_o2_adj)
+	and		%o2, 7, %o2
+	add		%GLOBAL_SPARE, 8, %GLOBAL_SPARE
+	retl
+	 add		%GLOBAL_SPARE, %o2, %o0
+ENDPROC(U1_gs_8_o2_adj)
+#endif
+
 	.align		64
 
 	.globl		FUNC_NAME
@@ -167,8 +280,8 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	 and		%g2, 0x38, %g2
 
 1:	subcc		%g1, 0x1, %g1
-	EX_LD_FP(LOAD(ldub, %o1 + 0x00, %o3))
-	EX_ST_FP(STORE(stb, %o3, %o1 + %GLOBAL_SPARE))
+	EX_LD_FP(LOAD(ldub, %o1 + 0x00, %o3), U1_g1_1_fp)
+	EX_ST_FP(STORE(stb, %o3, %o1 + %GLOBAL_SPARE), U1_g1_1_fp)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 0x1, %o1
 
@@ -179,20 +292,20 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	be,pt		%icc, 3f
 	 alignaddr	%o1, %g0, %o1
 
-	EX_LD_FP(LOAD(ldd, %o1, %f4))
-1:	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f6))
+	EX_LD_FP(LOAD(ldd, %o1, %f4), U1_g2_0_fp)
+1:	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f6), U1_g2_0_fp)
 	add		%o1, 0x8, %o1
 	subcc		%g2, 0x8, %g2
 	faligndata	%f4, %f6, %f0
-	EX_ST_FP(STORE(std, %f0, %o0))
+	EX_ST_FP(STORE(std, %f0, %o0), U1_g2_8_fp)
 	be,pn		%icc, 3f
 	 add		%o0, 0x8, %o0
 
-	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f4))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f4), U1_g2_0_fp)
 	add		%o1, 0x8, %o1
 	subcc		%g2, 0x8, %g2
 	faligndata	%f6, %f4, %f0
-	EX_ST_FP(STORE(std, %f0, %o0))
+	EX_ST_FP(STORE(std, %f0, %o0), U1_g2_8_fp)
 	bne,pt		%icc, 1b
 	 add		%o0, 0x8, %o0
 
@@ -215,13 +328,13 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	add		%g1, %GLOBAL_SPARE, %g1
 	subcc		%o2, %g3, %o2
 
-	EX_LD_FP(LOAD_BLK(%o1, %f0))
+	EX_LD_FP(LOAD_BLK(%o1, %f0), U1_gs_0_fp)
 	add		%o1, 0x40, %o1
 	add		%g1, %g3, %g1
-	EX_LD_FP(LOAD_BLK(%o1, %f16))
+	EX_LD_FP(LOAD_BLK(%o1, %f16), U1_gs_0_fp)
 	add		%o1, 0x40, %o1
 	sub		%GLOBAL_SPARE, 0x80, %GLOBAL_SPARE
-	EX_LD_FP(LOAD_BLK(%o1, %f32))
+	EX_LD_FP(LOAD_BLK(%o1, %f32), U1_gs_80_fp)
 	add		%o1, 0x40, %o1
 
 	/* There are 8 instances of the unrolled loop,
@@ -241,11 +354,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 
 	.align		64
 1:	FREG_FROB(f0, f2, f4, f6, f8, f10,f12,f14,f16)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f16,f18,f20,f22,f24,f26,f28,f30,f32)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f32,f34,f36,f38,f40,f42,f44,f46,f0)
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f0, %f2, %f48
 1:	FREG_FROB(f16,f18,f20,f22,f24,f26,f28,f30,f32)
@@ -262,11 +375,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	STORE_JUMP(o0, f48, 56f)
 
 1:	FREG_FROB(f2, f4, f6, f8, f10,f12,f14,f16,f18)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f18,f20,f22,f24,f26,f28,f30,f32,f34)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f34,f36,f38,f40,f42,f44,f46,f0, f2)
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f2, %f4, %f48
 1:	FREG_FROB(f18,f20,f22,f24,f26,f28,f30,f32,f34)
@@ -283,11 +396,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	STORE_JUMP(o0, f48, 57f)
 
 1:	FREG_FROB(f4, f6, f8, f10,f12,f14,f16,f18,f20)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f20,f22,f24,f26,f28,f30,f32,f34,f36)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f36,f38,f40,f42,f44,f46,f0, f2, f4)
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f4, %f6, %f48
 1:	FREG_FROB(f20,f22,f24,f26,f28,f30,f32,f34,f36)
@@ -304,11 +417,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	STORE_JUMP(o0, f48, 58f)
 
 1:	FREG_FROB(f6, f8, f10,f12,f14,f16,f18,f20,f22)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f22,f24,f26,f28,f30,f32,f34,f36,f38)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f38,f40,f42,f44,f46,f0, f2, f4, f6) 
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f6, %f8, %f48
 1:	FREG_FROB(f22,f24,f26,f28,f30,f32,f34,f36,f38)
@@ -325,11 +438,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	STORE_JUMP(o0, f48, 59f)
 
 1:	FREG_FROB(f8, f10,f12,f14,f16,f18,f20,f22,f24)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f24,f26,f28,f30,f32,f34,f36,f38,f40)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f40,f42,f44,f46,f0, f2, f4, f6, f8)
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f8, %f10, %f48
 1:	FREG_FROB(f24,f26,f28,f30,f32,f34,f36,f38,f40)
@@ -346,11 +459,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	STORE_JUMP(o0, f48, 60f)
 
 1:	FREG_FROB(f10,f12,f14,f16,f18,f20,f22,f24,f26)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f26,f28,f30,f32,f34,f36,f38,f40,f42)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f42,f44,f46,f0, f2, f4, f6, f8, f10)
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f10, %f12, %f48
 1:	FREG_FROB(f26,f28,f30,f32,f34,f36,f38,f40,f42)
@@ -367,11 +480,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	STORE_JUMP(o0, f48, 61f)
 
 1:	FREG_FROB(f12,f14,f16,f18,f20,f22,f24,f26,f28)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f28,f30,f32,f34,f36,f38,f40,f42,f44)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f44,f46,f0, f2, f4, f6, f8, f10,f12)
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f12, %f14, %f48
 1:	FREG_FROB(f28,f30,f32,f34,f36,f38,f40,f42,f44)
@@ -388,11 +501,11 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	STORE_JUMP(o0, f48, 62f)
 
 1:	FREG_FROB(f14,f16,f18,f20,f22,f24,f26,f28,f30)
-	LOOP_CHUNK1(o1, o0, GLOBAL_SPARE, 1f)
+	LOOP_CHUNK1(o1, o0, 1f)
 	FREG_FROB(f30,f32,f34,f36,f38,f40,f42,f44,f46)
-	LOOP_CHUNK2(o1, o0, GLOBAL_SPARE, 2f)
+	LOOP_CHUNK2(o1, o0, 2f)
 	FREG_FROB(f46,f0, f2, f4, f6, f8, f10,f12,f14)
-	LOOP_CHUNK3(o1, o0, GLOBAL_SPARE, 3f)
+	LOOP_CHUNK3(o1, o0, 3f)
 	ba,pt		%xcc, 1b+4
 	 faligndata	%f14, %f16, %f48
 1:	FREG_FROB(f30,f32,f34,f36,f38,f40,f42,f44,f46)
@@ -408,53 +521,53 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	FREG_FROB(f30,f32,f34,f36,f38,f40,f42,f44,f46)
 	STORE_JUMP(o0, f48, 63f)
 
-40:	FINISH_VISCHUNK(o0, f0,  f2,  g3)
-41:	FINISH_VISCHUNK(o0, f2,  f4,  g3)
-42:	FINISH_VISCHUNK(o0, f4,  f6,  g3)
-43:	FINISH_VISCHUNK(o0, f6,  f8,  g3)
-44:	FINISH_VISCHUNK(o0, f8,  f10, g3)
-45:	FINISH_VISCHUNK(o0, f10, f12, g3)
-46:	FINISH_VISCHUNK(o0, f12, f14, g3)
-47:	UNEVEN_VISCHUNK(o0, f14, f0,  g3)
-48:	FINISH_VISCHUNK(o0, f16, f18, g3)
-49:	FINISH_VISCHUNK(o0, f18, f20, g3)
-50:	FINISH_VISCHUNK(o0, f20, f22, g3)
-51:	FINISH_VISCHUNK(o0, f22, f24, g3)
-52:	FINISH_VISCHUNK(o0, f24, f26, g3)
-53:	FINISH_VISCHUNK(o0, f26, f28, g3)
-54:	FINISH_VISCHUNK(o0, f28, f30, g3)
-55:	UNEVEN_VISCHUNK(o0, f30, f0,  g3)
-56:	FINISH_VISCHUNK(o0, f32, f34, g3)
-57:	FINISH_VISCHUNK(o0, f34, f36, g3)
-58:	FINISH_VISCHUNK(o0, f36, f38, g3)
-59:	FINISH_VISCHUNK(o0, f38, f40, g3)
-60:	FINISH_VISCHUNK(o0, f40, f42, g3)
-61:	FINISH_VISCHUNK(o0, f42, f44, g3)
-62:	FINISH_VISCHUNK(o0, f44, f46, g3)
-63:	UNEVEN_VISCHUNK_LAST(o0, f46, f0,  g3)
-
-93:	EX_LD_FP(LOAD(ldd, %o1, %f2))
+40:	FINISH_VISCHUNK(o0, f0,  f2)
+41:	FINISH_VISCHUNK(o0, f2,  f4)
+42:	FINISH_VISCHUNK(o0, f4,  f6)
+43:	FINISH_VISCHUNK(o0, f6,  f8)
+44:	FINISH_VISCHUNK(o0, f8,  f10)
+45:	FINISH_VISCHUNK(o0, f10, f12)
+46:	FINISH_VISCHUNK(o0, f12, f14)
+47:	UNEVEN_VISCHUNK(o0, f14, f0)
+48:	FINISH_VISCHUNK(o0, f16, f18)
+49:	FINISH_VISCHUNK(o0, f18, f20)
+50:	FINISH_VISCHUNK(o0, f20, f22)
+51:	FINISH_VISCHUNK(o0, f22, f24)
+52:	FINISH_VISCHUNK(o0, f24, f26)
+53:	FINISH_VISCHUNK(o0, f26, f28)
+54:	FINISH_VISCHUNK(o0, f28, f30)
+55:	UNEVEN_VISCHUNK(o0, f30, f0)
+56:	FINISH_VISCHUNK(o0, f32, f34)
+57:	FINISH_VISCHUNK(o0, f34, f36)
+58:	FINISH_VISCHUNK(o0, f36, f38)
+59:	FINISH_VISCHUNK(o0, f38, f40)
+60:	FINISH_VISCHUNK(o0, f40, f42)
+61:	FINISH_VISCHUNK(o0, f42, f44)
+62:	FINISH_VISCHUNK(o0, f44, f46)
+63:	UNEVEN_VISCHUNK_LAST(o0, f46, f0)
+
+93:	EX_LD_FP(LOAD(ldd, %o1, %f2), U1_g3_0_fp)
 	add		%o1, 8, %o1
 	subcc		%g3, 8, %g3
 	faligndata	%f0, %f2, %f8
-	EX_ST_FP(STORE(std, %f8, %o0))
+	EX_ST_FP(STORE(std, %f8, %o0), U1_g3_8_fp)
 	bl,pn		%xcc, 95f
 	 add		%o0, 8, %o0
-	EX_LD_FP(LOAD(ldd, %o1, %f0))
+	EX_LD_FP(LOAD(ldd, %o1, %f0), U1_g3_0_fp)
 	add		%o1, 8, %o1
 	subcc		%g3, 8, %g3
 	faligndata	%f2, %f0, %f8
-	EX_ST_FP(STORE(std, %f8, %o0))
+	EX_ST_FP(STORE(std, %f8, %o0), U1_g3_8_fp)
 	bge,pt		%xcc, 93b
 	 add		%o0, 8, %o0
 
 95:	brz,pt		%o2, 2f
 	 mov		%g1, %o1
 
-1:	EX_LD_FP(LOAD(ldub, %o1, %o3))
+1:	EX_LD_FP(LOAD(ldub, %o1, %o3), U1_o2_0_fp)
 	add		%o1, 1, %o1
 	subcc		%o2, 1, %o2
-	EX_ST_FP(STORE(stb, %o3, %o0))
+	EX_ST_FP(STORE(stb, %o3, %o0), U1_o2_1_fp)
 	bne,pt		%xcc, 1b
 	 add		%o0, 1, %o0
 
@@ -470,27 +583,27 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 
 72:	andn		%o2, 0xf, %GLOBAL_SPARE
 	and		%o2, 0xf, %o2
-1:	EX_LD(LOAD(ldx, %o1 + 0x00, %o5))
-	EX_LD(LOAD(ldx, %o1 + 0x08, %g1))
+1:	EX_LD(LOAD(ldx, %o1 + 0x00, %o5), U1_gs_0)
+	EX_LD(LOAD(ldx, %o1 + 0x08, %g1), U1_gs_0)
 	subcc		%GLOBAL_SPARE, 0x10, %GLOBAL_SPARE
-	EX_ST(STORE(stx, %o5, %o1 + %o3))
+	EX_ST(STORE(stx, %o5, %o1 + %o3), U1_gs_10)
 	add		%o1, 0x8, %o1
-	EX_ST(STORE(stx, %g1, %o1 + %o3))
+	EX_ST(STORE(stx, %g1, %o1 + %o3), U1_gs_8)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 0x8, %o1
 73:	andcc		%o2, 0x8, %g0
 	be,pt		%XCC, 1f
 	 nop
-	EX_LD(LOAD(ldx, %o1, %o5))
+	EX_LD(LOAD(ldx, %o1, %o5), U1_o2_0)
 	sub		%o2, 0x8, %o2
-	EX_ST(STORE(stx, %o5, %o1 + %o3))
+	EX_ST(STORE(stx, %o5, %o1 + %o3), U1_o2_8)
 	add		%o1, 0x8, %o1
 1:	andcc		%o2, 0x4, %g0
 	be,pt		%XCC, 1f
 	 nop
-	EX_LD(LOAD(lduw, %o1, %o5))
+	EX_LD(LOAD(lduw, %o1, %o5), U1_o2_0)
 	sub		%o2, 0x4, %o2
-	EX_ST(STORE(stw, %o5, %o1 + %o3))
+	EX_ST(STORE(stw, %o5, %o1 + %o3), U1_o2_4)
 	add		%o1, 0x4, %o1
 1:	cmp		%o2, 0
 	be,pt		%XCC, 85f
@@ -504,9 +617,9 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	 sub		%g0, %g1, %g1
 	sub		%o2, %g1, %o2
 
-1:	EX_LD(LOAD(ldub, %o1, %o5))
+1:	EX_LD(LOAD(ldub, %o1, %o5), U1_g1_0)
 	subcc		%g1, 1, %g1
-	EX_ST(STORE(stb, %o5, %o1 + %o3))
+	EX_ST(STORE(stb, %o5, %o1 + %o3), U1_g1_1)
 	bgu,pt		%icc, 1b
 	 add		%o1, 1, %o1
 
@@ -522,16 +635,16 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 
 8:	mov		64, %o3
 	andn		%o1, 0x7, %o1
-	EX_LD(LOAD(ldx, %o1, %g2))
+	EX_LD(LOAD(ldx, %o1, %g2), U1_o2_0)
 	sub		%o3, %g1, %o3
 	andn		%o2, 0x7, %GLOBAL_SPARE
 	sllx		%g2, %g1, %g2
-1:	EX_LD(LOAD(ldx, %o1 + 0x8, %g3))
+1:	EX_LD(LOAD(ldx, %o1 + 0x8, %g3), U1_gs_0_o2_adj)
 	subcc		%GLOBAL_SPARE, 0x8, %GLOBAL_SPARE
 	add		%o1, 0x8, %o1
 	srlx		%g3, %o3, %o5
 	or		%o5, %g2, %o5
-	EX_ST(STORE(stx, %o5, %o0))
+	EX_ST(STORE(stx, %o5, %o0), U1_gs_8_o2_adj)
 	add		%o0, 0x8, %o0
 	bgu,pt		%icc, 1b
 	 sllx		%g3, %g1, %g2
@@ -549,9 +662,9 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	bne,pn		%XCC, 90f
 	 sub		%o0, %o1, %o3
 
-1:	EX_LD(LOAD(lduw, %o1, %g1))
+1:	EX_LD(LOAD(lduw, %o1, %g1), U1_o2_0)
 	subcc		%o2, 4, %o2
-	EX_ST(STORE(stw, %g1, %o1 + %o3))
+	EX_ST(STORE(stw, %g1, %o1 + %o3), U1_o2_4)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 4, %o1
 
@@ -559,9 +672,9 @@ FUNC_NAME:		/* %o0=dst, %o1=src, %o2=len */
 	 mov		EX_RETVAL(%o4), %o0
 
 	.align		32
-90:	EX_LD(LOAD(ldub, %o1, %g1))
+90:	EX_LD(LOAD(ldub, %o1, %g1), U1_o2_0)
 	subcc		%o2, 1, %o2
-	EX_ST(STORE(stb, %g1, %o1 + %o3))
+	EX_ST(STORE(stb, %g1, %o1 + %o3), U1_o2_1)
 	bgu,pt		%XCC, 90b
 	 add		%o1, 1, %o1
 	retl
diff --git a/arch/sparc/lib/U3copy_from_user.S b/arch/sparc/lib/U3copy_from_user.S
index 88ad73d86fe4..db73010a1af8 100644
--- a/arch/sparc/lib/U3copy_from_user.S
+++ b/arch/sparc/lib/U3copy_from_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 1999, 2000, 2004 David S. Miller (davem@redhat.com)
  */
 
-#define EX_LD(x)		\
+#define EX_LD(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one;	\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_LD_FP(x)		\
+#define EX_LD_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_fp;\
+	.word 98b, y##_fp;	\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/U3copy_to_user.S b/arch/sparc/lib/U3copy_to_user.S
index 845139d75537..c4ee858e352a 100644
--- a/arch/sparc/lib/U3copy_to_user.S
+++ b/arch/sparc/lib/U3copy_to_user.S
@@ -3,19 +3,19 @@
  * Copyright (C) 1999, 2000, 2004 David S. Miller (davem@redhat.com)
  */
 
-#define EX_ST(x)		\
+#define EX_ST(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one;	\
+	.word 98b, y;		\
 	.text;			\
 	.align 4;
 
-#define EX_ST_FP(x)		\
+#define EX_ST_FP(x,y)		\
 98:	x;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one_fp;\
+	.word 98b, y##_fp;	\
 	.text;			\
 	.align 4;
 
diff --git a/arch/sparc/lib/U3memcpy.S b/arch/sparc/lib/U3memcpy.S
index 491ee69e4995..54f98706b03b 100644
--- a/arch/sparc/lib/U3memcpy.S
+++ b/arch/sparc/lib/U3memcpy.S
@@ -4,6 +4,7 @@
  */
 
 #ifdef __KERNEL__
+#include <linux/linkage.h>
 #include <asm/visasm.h>
 #include <asm/asi.h>
 #define GLOBAL_SPARE	%g7
@@ -22,21 +23,17 @@
 #endif
 
 #ifndef EX_LD
-#define EX_LD(x)	x
+#define EX_LD(x,y)	x
 #endif
 #ifndef EX_LD_FP
-#define EX_LD_FP(x)	x
+#define EX_LD_FP(x,y)	x
 #endif
 
 #ifndef EX_ST
-#define EX_ST(x)	x
+#define EX_ST(x,y)	x
 #endif
 #ifndef EX_ST_FP
-#define EX_ST_FP(x)	x
-#endif
-
-#ifndef EX_RETVAL
-#define EX_RETVAL(x)	x
+#define EX_ST_FP(x,y)	x
 #endif
 
 #ifndef LOAD
@@ -77,6 +74,87 @@
 	 */
 
 	.text
+#ifndef EX_RETVAL
+#define EX_RETVAL(x)	x
+__restore_fp:
+	VISExitHalf
+	retl
+	 nop
+ENTRY(U3_retl_o2_plus_g2_plus_g1_plus_1_fp)
+	add	%g1, 1, %g1
+	add	%g2, %g1, %g2
+	ba,pt	%xcc, __restore_fp
+	 add	%o2, %g2, %o0
+ENDPROC(U3_retl_o2_plus_g2_plus_g1_plus_1_fp)
+ENTRY(U3_retl_o2_plus_g2_fp)
+	ba,pt	%xcc, __restore_fp
+	 add	%o2, %g2, %o0
+ENDPROC(U3_retl_o2_plus_g2_fp)
+ENTRY(U3_retl_o2_plus_g2_plus_8_fp)
+	add	%g2, 8, %g2
+	ba,pt	%xcc, __restore_fp
+	 add	%o2, %g2, %o0
+ENDPROC(U3_retl_o2_plus_g2_plus_8_fp)
+ENTRY(U3_retl_o2)
+	retl
+	 mov	%o2, %o0
+ENDPROC(U3_retl_o2)
+ENTRY(U3_retl_o2_plus_1)
+	retl
+	 add	%o2, 1, %o0
+ENDPROC(U3_retl_o2_plus_1)
+ENTRY(U3_retl_o2_plus_4)
+	retl
+	 add	%o2, 4, %o0
+ENDPROC(U3_retl_o2_plus_4)
+ENTRY(U3_retl_o2_plus_8)
+	retl
+	 add	%o2, 8, %o0
+ENDPROC(U3_retl_o2_plus_8)
+ENTRY(U3_retl_o2_plus_g1_plus_1)
+	add	%g1, 1, %g1
+	retl
+	 add	%o2, %g1, %o0
+ENDPROC(U3_retl_o2_plus_g1_plus_1)
+ENTRY(U3_retl_o2_fp)
+	ba,pt	%xcc, __restore_fp
+	 mov	%o2, %o0
+ENDPROC(U3_retl_o2_fp)
+ENTRY(U3_retl_o2_plus_o3_sll_6_plus_0x80_fp)
+	sll	%o3, 6, %o3
+	add	%o3, 0x80, %o3
+	ba,pt	%xcc, __restore_fp
+	 add	%o2, %o3, %o0
+ENDPROC(U3_retl_o2_plus_o3_sll_6_plus_0x80_fp)
+ENTRY(U3_retl_o2_plus_o3_sll_6_plus_0x40_fp)
+	sll	%o3, 6, %o3
+	add	%o3, 0x40, %o3
+	ba,pt	%xcc, __restore_fp
+	 add	%o2, %o3, %o0
+ENDPROC(U3_retl_o2_plus_o3_sll_6_plus_0x40_fp)
+ENTRY(U3_retl_o2_plus_GS_plus_0x10)
+	add	GLOBAL_SPARE, 0x10, GLOBAL_SPARE
+	retl
+	 add	%o2, GLOBAL_SPARE, %o0
+ENDPROC(U3_retl_o2_plus_GS_plus_0x10)
+ENTRY(U3_retl_o2_plus_GS_plus_0x08)
+	add	GLOBAL_SPARE, 0x08, GLOBAL_SPARE
+	retl
+	 add	%o2, GLOBAL_SPARE, %o0
+ENDPROC(U3_retl_o2_plus_GS_plus_0x08)
+ENTRY(U3_retl_o2_and_7_plus_GS)
+	and	%o2, 7, %o2
+	retl
+	 add	%o2, GLOBAL_SPARE, %o2
+ENDPROC(U3_retl_o2_and_7_plus_GS)
+ENTRY(U3_retl_o2_and_7_plus_GS_plus_8)
+	add	GLOBAL_SPARE, 8, GLOBAL_SPARE
+	and	%o2, 7, %o2
+	retl
+	 add	%o2, GLOBAL_SPARE, %o2
+ENDPROC(U3_retl_o2_and_7_plus_GS_plus_8)
+#endif
+
 	.align		64
 
 	/* The cheetah's flexible spine, oversized liver, enlarged heart,
@@ -126,8 +204,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	 and		%g2, 0x38, %g2
 
 1:	subcc		%g1, 0x1, %g1
-	EX_LD_FP(LOAD(ldub, %o1 + 0x00, %o3))
-	EX_ST_FP(STORE(stb, %o3, %o1 + GLOBAL_SPARE))
+	EX_LD_FP(LOAD(ldub, %o1 + 0x00, %o3), U3_retl_o2_plus_g2_plus_g1_plus_1)
+	EX_ST_FP(STORE(stb, %o3, %o1 + GLOBAL_SPARE), U3_retl_o2_plus_g2_plus_g1_plus_1)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 0x1, %o1
 
@@ -138,20 +216,20 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	be,pt		%icc, 3f
 	 alignaddr	%o1, %g0, %o1
 
-	EX_LD_FP(LOAD(ldd, %o1, %f4))
-1:	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f6))
+	EX_LD_FP(LOAD(ldd, %o1, %f4), U3_retl_o2_plus_g2)
+1:	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f6), U3_retl_o2_plus_g2)
 	add		%o1, 0x8, %o1
 	subcc		%g2, 0x8, %g2
 	faligndata	%f4, %f6, %f0
-	EX_ST_FP(STORE(std, %f0, %o0))
+	EX_ST_FP(STORE(std, %f0, %o0), U3_retl_o2_plus_g2_plus_8)
 	be,pn		%icc, 3f
 	 add		%o0, 0x8, %o0
 
-	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f4))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x8, %f4), U3_retl_o2_plus_g2)
 	add		%o1, 0x8, %o1
 	subcc		%g2, 0x8, %g2
 	faligndata	%f6, %f4, %f2
-	EX_ST_FP(STORE(std, %f2, %o0))
+	EX_ST_FP(STORE(std, %f2, %o0), U3_retl_o2_plus_g2_plus_8)
 	bne,pt		%icc, 1b
 	 add		%o0, 0x8, %o0
 
@@ -161,25 +239,25 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	LOAD(prefetch, %o1 + 0x080, #one_read)
 	LOAD(prefetch, %o1 + 0x0c0, #one_read)
 	LOAD(prefetch, %o1 + 0x100, #one_read)
-	EX_LD_FP(LOAD(ldd, %o1 + 0x000, %f0))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x000, %f0), U3_retl_o2)
 	LOAD(prefetch, %o1 + 0x140, #one_read)
-	EX_LD_FP(LOAD(ldd, %o1 + 0x008, %f2))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x008, %f2), U3_retl_o2)
 	LOAD(prefetch, %o1 + 0x180, #one_read)
-	EX_LD_FP(LOAD(ldd, %o1 + 0x010, %f4))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x010, %f4), U3_retl_o2)
 	LOAD(prefetch, %o1 + 0x1c0, #one_read)
 	faligndata	%f0, %f2, %f16
-	EX_LD_FP(LOAD(ldd, %o1 + 0x018, %f6))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x018, %f6), U3_retl_o2)
 	faligndata	%f2, %f4, %f18
-	EX_LD_FP(LOAD(ldd, %o1 + 0x020, %f8))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x020, %f8), U3_retl_o2)
 	faligndata	%f4, %f6, %f20
-	EX_LD_FP(LOAD(ldd, %o1 + 0x028, %f10))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x028, %f10), U3_retl_o2)
 	faligndata	%f6, %f8, %f22
 
-	EX_LD_FP(LOAD(ldd, %o1 + 0x030, %f12))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x030, %f12), U3_retl_o2)
 	faligndata	%f8, %f10, %f24
-	EX_LD_FP(LOAD(ldd, %o1 + 0x038, %f14))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x038, %f14), U3_retl_o2)
 	faligndata	%f10, %f12, %f26
-	EX_LD_FP(LOAD(ldd, %o1 + 0x040, %f0))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x040, %f0), U3_retl_o2)
 
 	subcc		GLOBAL_SPARE, 0x80, GLOBAL_SPARE
 	add		%o1, 0x40, %o1
@@ -190,26 +268,26 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 	.align		64
 1:
-	EX_LD_FP(LOAD(ldd, %o1 + 0x008, %f2))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x008, %f2), U3_retl_o2_plus_o3_sll_6_plus_0x80)
 	faligndata	%f12, %f14, %f28
-	EX_LD_FP(LOAD(ldd, %o1 + 0x010, %f4))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x010, %f4), U3_retl_o2_plus_o3_sll_6_plus_0x80)
 	faligndata	%f14, %f0, %f30
-	EX_ST_FP(STORE_BLK(%f16, %o0))
-	EX_LD_FP(LOAD(ldd, %o1 + 0x018, %f6))
+	EX_ST_FP(STORE_BLK(%f16, %o0), U3_retl_o2_plus_o3_sll_6_plus_0x80)
+	EX_LD_FP(LOAD(ldd, %o1 + 0x018, %f6), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f0, %f2, %f16
 	add		%o0, 0x40, %o0
 
-	EX_LD_FP(LOAD(ldd, %o1 + 0x020, %f8))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x020, %f8), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f2, %f4, %f18
-	EX_LD_FP(LOAD(ldd, %o1 + 0x028, %f10))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x028, %f10), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f4, %f6, %f20
-	EX_LD_FP(LOAD(ldd, %o1 + 0x030, %f12))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x030, %f12), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	subcc		%o3, 0x01, %o3
 	faligndata	%f6, %f8, %f22
-	EX_LD_FP(LOAD(ldd, %o1 + 0x038, %f14))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x038, %f14), U3_retl_o2_plus_o3_sll_6_plus_0x80)
 
 	faligndata	%f8, %f10, %f24
-	EX_LD_FP(LOAD(ldd, %o1 + 0x040, %f0))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x040, %f0), U3_retl_o2_plus_o3_sll_6_plus_0x80)
 	LOAD(prefetch, %o1 + 0x1c0, #one_read)
 	faligndata	%f10, %f12, %f26
 	bg,pt		%XCC, 1b
@@ -217,29 +295,29 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 	/* Finally we copy the last full 64-byte block. */
 2:
-	EX_LD_FP(LOAD(ldd, %o1 + 0x008, %f2))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x008, %f2), U3_retl_o2_plus_o3_sll_6_plus_0x80)
 	faligndata	%f12, %f14, %f28
-	EX_LD_FP(LOAD(ldd, %o1 + 0x010, %f4))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x010, %f4), U3_retl_o2_plus_o3_sll_6_plus_0x80)
 	faligndata	%f14, %f0, %f30
-	EX_ST_FP(STORE_BLK(%f16, %o0))
-	EX_LD_FP(LOAD(ldd, %o1 + 0x018, %f6))
+	EX_ST_FP(STORE_BLK(%f16, %o0), U3_retl_o2_plus_o3_sll_6_plus_0x80)
+	EX_LD_FP(LOAD(ldd, %o1 + 0x018, %f6), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f0, %f2, %f16
-	EX_LD_FP(LOAD(ldd, %o1 + 0x020, %f8))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x020, %f8), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f2, %f4, %f18
-	EX_LD_FP(LOAD(ldd, %o1 + 0x028, %f10))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x028, %f10), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f4, %f6, %f20
-	EX_LD_FP(LOAD(ldd, %o1 + 0x030, %f12))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x030, %f12), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f6, %f8, %f22
-	EX_LD_FP(LOAD(ldd, %o1 + 0x038, %f14))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x038, %f14), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	faligndata	%f8, %f10, %f24
 	cmp		%g1, 0
 	be,pt		%XCC, 1f
 	 add		%o0, 0x40, %o0
-	EX_LD_FP(LOAD(ldd, %o1 + 0x040, %f0))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x040, %f0), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 1:	faligndata	%f10, %f12, %f26
 	faligndata	%f12, %f14, %f28
 	faligndata	%f14, %f0, %f30
-	EX_ST_FP(STORE_BLK(%f16, %o0))
+	EX_ST_FP(STORE_BLK(%f16, %o0), U3_retl_o2_plus_o3_sll_6_plus_0x40)
 	add		%o0, 0x40, %o0
 	add		%o1, 0x40, %o1
 	membar		#Sync
@@ -259,20 +337,20 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 	sub		%o2, %g2, %o2
 	be,a,pt		%XCC, 1f
-	 EX_LD_FP(LOAD(ldd, %o1 + 0x00, %f0))
+	 EX_LD_FP(LOAD(ldd, %o1 + 0x00, %f0), U3_retl_o2_plus_g2)
 
-1:	EX_LD_FP(LOAD(ldd, %o1 + 0x08, %f2))
+1:	EX_LD_FP(LOAD(ldd, %o1 + 0x08, %f2), U3_retl_o2_plus_g2)
 	add		%o1, 0x8, %o1
 	subcc		%g2, 0x8, %g2
 	faligndata	%f0, %f2, %f8
-	EX_ST_FP(STORE(std, %f8, %o0))
+	EX_ST_FP(STORE(std, %f8, %o0), U3_retl_o2_plus_g2_plus_8)
 	be,pn		%XCC, 2f
 	 add		%o0, 0x8, %o0
-	EX_LD_FP(LOAD(ldd, %o1 + 0x08, %f0))
+	EX_LD_FP(LOAD(ldd, %o1 + 0x08, %f0), U3_retl_o2_plus_g2)
 	add		%o1, 0x8, %o1
 	subcc		%g2, 0x8, %g2
 	faligndata	%f2, %f0, %f8
-	EX_ST_FP(STORE(std, %f8, %o0))
+	EX_ST_FP(STORE(std, %f8, %o0), U3_retl_o2_plus_g2_plus_8)
 	bne,pn		%XCC, 1b
 	 add		%o0, 0x8, %o0
 
@@ -292,30 +370,33 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	 andcc		%o2, 0x8, %g0
 	be,pt		%icc, 1f
 	 nop
-	EX_LD(LOAD(ldx, %o1, %o5))
-	EX_ST(STORE(stx, %o5, %o1 + %o3))
+	EX_LD(LOAD(ldx, %o1, %o5), U3_retl_o2)
+	EX_ST(STORE(stx, %o5, %o1 + %o3), U3_retl_o2)
 	add		%o1, 0x8, %o1
+	sub		%o2, 8, %o2
 
 1:	andcc		%o2, 0x4, %g0
 	be,pt		%icc, 1f
 	 nop
-	EX_LD(LOAD(lduw, %o1, %o5))
-	EX_ST(STORE(stw, %o5, %o1 + %o3))
+	EX_LD(LOAD(lduw, %o1, %o5), U3_retl_o2)
+	EX_ST(STORE(stw, %o5, %o1 + %o3), U3_retl_o2)
 	add		%o1, 0x4, %o1
+	sub		%o2, 4, %o2
 
 1:	andcc		%o2, 0x2, %g0
 	be,pt		%icc, 1f
 	 nop
-	EX_LD(LOAD(lduh, %o1, %o5))
-	EX_ST(STORE(sth, %o5, %o1 + %o3))
+	EX_LD(LOAD(lduh, %o1, %o5), U3_retl_o2)
+	EX_ST(STORE(sth, %o5, %o1 + %o3), U3_retl_o2)
 	add		%o1, 0x2, %o1
+	sub		%o2, 2, %o2
 
 1:	andcc		%o2, 0x1, %g0
 	be,pt		%icc, 85f
 	 nop
-	EX_LD(LOAD(ldub, %o1, %o5))
+	EX_LD(LOAD(ldub, %o1, %o5), U3_retl_o2)
 	ba,pt		%xcc, 85f
-	 EX_ST(STORE(stb, %o5, %o1 + %o3))
+	 EX_ST(STORE(stb, %o5, %o1 + %o3), U3_retl_o2)
 
 	.align		64
 70: /* 16 < len <= 64 */
@@ -326,26 +407,26 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	andn		%o2, 0xf, GLOBAL_SPARE
 	and		%o2, 0xf, %o2
 1:	subcc		GLOBAL_SPARE, 0x10, GLOBAL_SPARE
-	EX_LD(LOAD(ldx, %o1 + 0x00, %o5))
-	EX_LD(LOAD(ldx, %o1 + 0x08, %g1))
-	EX_ST(STORE(stx, %o5, %o1 + %o3))
+	EX_LD(LOAD(ldx, %o1 + 0x00, %o5), U3_retl_o2_plus_GS_plus_0x10)
+	EX_LD(LOAD(ldx, %o1 + 0x08, %g1), U3_retl_o2_plus_GS_plus_0x10)
+	EX_ST(STORE(stx, %o5, %o1 + %o3), U3_retl_o2_plus_GS_plus_0x10)
 	add		%o1, 0x8, %o1
-	EX_ST(STORE(stx, %g1, %o1 + %o3))
+	EX_ST(STORE(stx, %g1, %o1 + %o3), U3_retl_o2_plus_GS_plus_0x08)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 0x8, %o1
 73:	andcc		%o2, 0x8, %g0
 	be,pt		%XCC, 1f
 	 nop
 	sub		%o2, 0x8, %o2
-	EX_LD(LOAD(ldx, %o1, %o5))
-	EX_ST(STORE(stx, %o5, %o1 + %o3))
+	EX_LD(LOAD(ldx, %o1, %o5), U3_retl_o2_plus_8)
+	EX_ST(STORE(stx, %o5, %o1 + %o3), U3_retl_o2_plus_8)
 	add		%o1, 0x8, %o1
 1:	andcc		%o2, 0x4, %g0
 	be,pt		%XCC, 1f
 	 nop
 	sub		%o2, 0x4, %o2
-	EX_LD(LOAD(lduw, %o1, %o5))
-	EX_ST(STORE(stw, %o5, %o1 + %o3))
+	EX_LD(LOAD(lduw, %o1, %o5), U3_retl_o2_plus_4)
+	EX_ST(STORE(stw, %o5, %o1 + %o3), U3_retl_o2_plus_4)
 	add		%o1, 0x4, %o1
 1:	cmp		%o2, 0
 	be,pt		%XCC, 85f
@@ -361,8 +442,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	sub		%o2, %g1, %o2
 
 1:	subcc		%g1, 1, %g1
-	EX_LD(LOAD(ldub, %o1, %o5))
-	EX_ST(STORE(stb, %o5, %o1 + %o3))
+	EX_LD(LOAD(ldub, %o1, %o5), U3_retl_o2_plus_g1_plus_1)
+	EX_ST(STORE(stb, %o5, %o1 + %o3), U3_retl_o2_plus_g1_plus_1)
 	bgu,pt		%icc, 1b
 	 add		%o1, 1, %o1
 
@@ -378,16 +459,16 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 8:	mov		64, %o3
 	andn		%o1, 0x7, %o1
-	EX_LD(LOAD(ldx, %o1, %g2))
+	EX_LD(LOAD(ldx, %o1, %g2), U3_retl_o2)
 	sub		%o3, %g1, %o3
 	andn		%o2, 0x7, GLOBAL_SPARE
 	sllx		%g2, %g1, %g2
-1:	EX_LD(LOAD(ldx, %o1 + 0x8, %g3))
+1:	EX_LD(LOAD(ldx, %o1 + 0x8, %g3), U3_retl_o2_and_7_plus_GS)
 	subcc		GLOBAL_SPARE, 0x8, GLOBAL_SPARE
 	add		%o1, 0x8, %o1
 	srlx		%g3, %o3, %o5
 	or		%o5, %g2, %o5
-	EX_ST(STORE(stx, %o5, %o0))
+	EX_ST(STORE(stx, %o5, %o0), U3_retl_o2_and_7_plus_GS_plus_8)
 	add		%o0, 0x8, %o0
 	bgu,pt		%icc, 1b
 	 sllx		%g3, %g1, %g2
@@ -407,8 +488,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 
 1:
 	subcc		%o2, 4, %o2
-	EX_LD(LOAD(lduw, %o1, %g1))
-	EX_ST(STORE(stw, %g1, %o1 + %o3))
+	EX_LD(LOAD(lduw, %o1, %g1), U3_retl_o2_plus_4)
+	EX_ST(STORE(stw, %g1, %o1 + %o3), U3_retl_o2_plus_4)
 	bgu,pt		%XCC, 1b
 	 add		%o1, 4, %o1
 
@@ -418,8 +499,8 @@ FUNC_NAME:	/* %o0=dst, %o1=src, %o2=len */
 	.align		32
 90:
 	subcc		%o2, 1, %o2
-	EX_LD(LOAD(ldub, %o1, %g1))
-	EX_ST(STORE(stb, %g1, %o1 + %o3))
+	EX_LD(LOAD(ldub, %o1, %g1), U3_retl_o2_plus_1)
+	EX_ST(STORE(stb, %g1, %o1 + %o3), U3_retl_o2_plus_1)
 	bgu,pt		%XCC, 90b
 	 add		%o1, 1, %o1
 	retl
diff --git a/arch/sparc/lib/copy_in_user.S b/arch/sparc/lib/copy_in_user.S
index 482de093bdae..0252b218de45 100644
--- a/arch/sparc/lib/copy_in_user.S
+++ b/arch/sparc/lib/copy_in_user.S
@@ -9,18 +9,33 @@
 
 #define XCC xcc
 
-#define EX(x,y)			\
+#define EX(x,y,z)		\
 98:	x,y;			\
 	.section __ex_table,"a";\
 	.align 4;		\
-	.word 98b, __retl_one;	\
+	.word 98b, z;		\
 	.text;			\
 	.align 4;
 
+#define EX_O4(x,y) EX(x,y,__retl_o4_plus_8)
+#define EX_O2_4(x,y) EX(x,y,__retl_o2_plus_4)
+#define EX_O2_1(x,y) EX(x,y,__retl_o2_plus_1)
+
 	.register	%g2,#scratch
 	.register	%g3,#scratch
 
 	.text
+__retl_o4_plus_8:
+	add	%o4, %o2, %o4
+	retl
+	 add	%o4, 8, %o0
+__retl_o2_plus_4:
+	retl
+	 add	%o2, 4, %o0
+__retl_o2_plus_1:
+	retl
+	 add	%o2, 1, %o0
+
 	.align	32
 
 	/* Don't try to get too fancy here, just nice and
@@ -45,8 +60,8 @@ ENTRY(___copy_in_user)	/* %o0=dst, %o1=src, %o2=len */
 	andn		%o2, 0x7, %o4
 	and		%o2, 0x7, %o2
 1:	subcc		%o4, 0x8, %o4
-	EX(ldxa [%o1] %asi, %o5)
-	EX(stxa %o5, [%o0] %asi)
+	EX_O4(ldxa [%o1] %asi, %o5)
+	EX_O4(stxa %o5, [%o0] %asi)
 	add		%o1, 0x8, %o1
 	bgu,pt		%XCC, 1b
 	 add		%o0, 0x8, %o0
@@ -54,8 +69,8 @@ ENTRY(___copy_in_user)	/* %o0=dst, %o1=src, %o2=len */
 	be,pt		%XCC, 1f
 	 nop
 	sub		%o2, 0x4, %o2
-	EX(lduwa [%o1] %asi, %o5)
-	EX(stwa %o5, [%o0] %asi)
+	EX_O2_4(lduwa [%o1] %asi, %o5)
+	EX_O2_4(stwa %o5, [%o0] %asi)
 	add		%o1, 0x4, %o1
 	add		%o0, 0x4, %o0
 1:	cmp		%o2, 0
@@ -71,8 +86,8 @@ ENTRY(___copy_in_user)	/* %o0=dst, %o1=src, %o2=len */
 
 82:
 	subcc		%o2, 4, %o2
-	EX(lduwa [%o1] %asi, %g1)
-	EX(stwa %g1, [%o0] %asi)
+	EX_O2_4(lduwa [%o1] %asi, %g1)
+	EX_O2_4(stwa %g1, [%o0] %asi)
 	add		%o1, 4, %o1
 	bgu,pt		%XCC, 82b
 	 add		%o0, 4, %o0
@@ -83,8 +98,8 @@ ENTRY(___copy_in_user)	/* %o0=dst, %o1=src, %o2=len */
 	.align	32
 90:
 	subcc		%o2, 1, %o2
-	EX(lduba [%o1] %asi, %g1)
-	EX(stba %g1, [%o0] %asi)
+	EX_O2_1(lduba [%o1] %asi, %g1)
+	EX_O2_1(stba %g1, [%o0] %asi)
 	add		%o1, 1, %o1
 	bgu,pt		%XCC, 90b
 	 add		%o0, 1, %o0
diff --git a/arch/sparc/lib/user_fixup.c b/arch/sparc/lib/user_fixup.c
deleted file mode 100644
index ac96ae236709..000000000000
--- a/arch/sparc/lib/user_fixup.c
+++ /dev/null
@@ -1,71 +0,0 @@
-/* user_fixup.c: Fix up user copy faults.
- *
- * Copyright (C) 2004 David S. Miller <davem@redhat.com>
- */
-
-#include <linux/compiler.h>
-#include <linux/kernel.h>
-#include <linux/string.h>
-#include <linux/errno.h>
-#include <linux/module.h>
-
-#include <asm/uaccess.h>
-
-/* Calculating the exact fault address when using
- * block loads and stores can be very complicated.
- *
- * Instead of trying to be clever and handling all
- * of the cases, just fix things up simply here.
- */
-
-static unsigned long compute_size(unsigned long start, unsigned long size, unsigned long *offset)
-{
-	unsigned long fault_addr = current_thread_info()->fault_address;
-	unsigned long end = start + size;
-
-	if (fault_addr < start || fault_addr >= end) {
-		*offset = 0;
-	} else {
-		*offset = fault_addr - start;
-		size = end - fault_addr;
-	}
-	return size;
-}
-
-unsigned long copy_from_user_fixup(void *to, const void __user *from, unsigned long size)
-{
-	unsigned long offset;
-
-	size = compute_size((unsigned long) from, size, &offset);
-	if (likely(size))
-		memset(to + offset, 0, size);
-
-	return size;
-}
-EXPORT_SYMBOL(copy_from_user_fixup);
-
-unsigned long copy_to_user_fixup(void __user *to, const void *from, unsigned long size)
-{
-	unsigned long offset;
-
-	return compute_size((unsigned long) to, size, &offset);
-}
-EXPORT_SYMBOL(copy_to_user_fixup);
-
-unsigned long copy_in_user_fixup(void __user *to, void __user *from, unsigned long size)
-{
-	unsigned long fault_addr = current_thread_info()->fault_address;
-	unsigned long start = (unsigned long) to;
-	unsigned long end = start + size;
-
-	if (fault_addr >= start && fault_addr < end)
-		return end - fault_addr;
-
-	start = (unsigned long) from;
-	end = start + size;
-	if (fault_addr >= start && fault_addr < end)
-		return end - fault_addr;
-
-	return size;
-}
-EXPORT_SYMBOL(copy_in_user_fixup);
diff --git a/arch/sparc/mm/tsb.c b/arch/sparc/mm/tsb.c
index f2b77112e9d8..e20fbbafb0b0 100644
--- a/arch/sparc/mm/tsb.c
+++ b/arch/sparc/mm/tsb.c
@@ -27,6 +27,20 @@ static inline int tag_compare(unsigned long tag, unsigned long vaddr)
 	return (tag == (vaddr >> 22));
 }
 
+static void flush_tsb_kernel_range_scan(unsigned long start, unsigned long end)
+{
+	unsigned long idx;
+
+	for (idx = 0; idx < KERNEL_TSB_NENTRIES; idx++) {
+		struct tsb *ent = &swapper_tsb[idx];
+		unsigned long match = idx << 13;
+
+		match |= (ent->tag << 22);
+		if (match >= start && match < end)
+			ent->tag = (1UL << TSB_TAG_INVALID_BIT);
+	}
+}
+
 /* TSB flushes need only occur on the processor initiating the address
  * space modification, not on each cpu the address space has run on.
  * Only the TLB flush needs that treatment.
@@ -36,6 +50,9 @@ void flush_tsb_kernel_range(unsigned long start, unsigned long end)
 {
 	unsigned long v;
 
+	if ((end - start) >> PAGE_SHIFT >= 2 * KERNEL_TSB_NENTRIES)
+		return flush_tsb_kernel_range_scan(start, end);
+
 	for (v = start; v < end; v += PAGE_SIZE) {
 		unsigned long hash = tsb_hash(v, PAGE_SHIFT,
 					      KERNEL_TSB_NENTRIES);
diff --git a/arch/sparc/mm/ultra.S b/arch/sparc/mm/ultra.S
index b4f4733abc6e..5d2fd6cd3189 100644
--- a/arch/sparc/mm/ultra.S
+++ b/arch/sparc/mm/ultra.S
@@ -30,7 +30,7 @@
 	.text
 	.align		32
 	.globl		__flush_tlb_mm
-__flush_tlb_mm:		/* 18 insns */
+__flush_tlb_mm:		/* 19 insns */
 	/* %o0=(ctx & TAG_CONTEXT_BITS), %o1=SECONDARY_CONTEXT */
 	ldxa		[%o1] ASI_DMMU, %g2
 	cmp		%g2, %o0
@@ -81,7 +81,7 @@ __flush_tlb_page:	/* 22 insns */
 
 	.align		32
 	.globl		__flush_tlb_pending
-__flush_tlb_pending:	/* 26 insns */
+__flush_tlb_pending:	/* 27 insns */
 	/* %o0 = context, %o1 = nr, %o2 = vaddrs[] */
 	rdpr		%pstate, %g7
 	sllx		%o1, 3, %o1
@@ -113,12 +113,14 @@ __flush_tlb_pending:	/* 26 insns */
 
 	.align		32
 	.globl		__flush_tlb_kernel_range
-__flush_tlb_kernel_range:	/* 16 insns */
+__flush_tlb_kernel_range:	/* 31 insns */
 	/* %o0=start, %o1=end */
 	cmp		%o0, %o1
 	be,pn		%xcc, 2f
+	 sub		%o1, %o0, %o3
+	srlx		%o3, 18, %o4
+	brnz,pn		%o4, __spitfire_flush_tlb_kernel_range_slow
 	 sethi		%hi(PAGE_SIZE), %o4
-	sub		%o1, %o0, %o3
 	sub		%o3, %o4, %o3
 	or		%o0, 0x20, %o0		! Nucleus
 1:	stxa		%g0, [%o0 + %o3] ASI_DMMU_DEMAP
@@ -131,6 +133,41 @@ __flush_tlb_kernel_range:	/* 16 insns */
 	retl
 	 nop
 	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+
+__spitfire_flush_tlb_kernel_range_slow:
+	mov		63 * 8, %o4
+1:	ldxa		[%o4] ASI_ITLB_DATA_ACCESS, %o3
+	andcc		%o3, 0x40, %g0			/* _PAGE_L_4U */
+	bne,pn		%xcc, 2f
+	 mov		TLB_TAG_ACCESS, %o3
+	stxa		%g0, [%o3] ASI_IMMU
+	stxa		%g0, [%o4] ASI_ITLB_DATA_ACCESS
+	membar		#Sync
+2:	ldxa		[%o4] ASI_DTLB_DATA_ACCESS, %o3
+	andcc		%o3, 0x40, %g0
+	bne,pn		%xcc, 2f
+	 mov		TLB_TAG_ACCESS, %o3
+	stxa		%g0, [%o3] ASI_DMMU
+	stxa		%g0, [%o4] ASI_DTLB_DATA_ACCESS
+	membar		#Sync
+2:	sub		%o4, 8, %o4
+	brgez,pt	%o4, 1b
+	 nop
+	retl
+	 nop
 
 __spitfire_flush_tlb_mm_slow:
 	rdpr		%pstate, %g1
@@ -285,6 +322,40 @@ __cheetah_flush_tlb_pending:	/* 27 insns */
 	retl
 	 wrpr		%g7, 0x0, %pstate
 
+__cheetah_flush_tlb_kernel_range:	/* 31 insns */
+	/* %o0=start, %o1=end */
+	cmp		%o0, %o1
+	be,pn		%xcc, 2f
+	 sub		%o1, %o0, %o3
+	srlx		%o3, 18, %o4
+	brnz,pn		%o4, 3f
+	 sethi		%hi(PAGE_SIZE), %o4
+	sub		%o3, %o4, %o3
+	or		%o0, 0x20, %o0		! Nucleus
+1:	stxa		%g0, [%o0 + %o3] ASI_DMMU_DEMAP
+	stxa		%g0, [%o0 + %o3] ASI_IMMU_DEMAP
+	membar		#Sync
+	brnz,pt		%o3, 1b
+	 sub		%o3, %o4, %o3
+2:	sethi		%hi(KERNBASE), %o3
+	flush		%o3
+	retl
+	 nop
+3:	mov		0x80, %o4
+	stxa		%g0, [%o4] ASI_DMMU_DEMAP
+	membar		#Sync
+	stxa		%g0, [%o4] ASI_IMMU_DEMAP
+	membar		#Sync
+	retl
+	 nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+
 #ifdef DCACHE_ALIASING_POSSIBLE
 __cheetah_flush_dcache_page: /* 11 insns */
 	sethi		%hi(PAGE_OFFSET), %g1
@@ -309,19 +380,28 @@ __hypervisor_tlb_tl0_error:
 	ret
 	 restore
 
-__hypervisor_flush_tlb_mm: /* 10 insns */
+__hypervisor_flush_tlb_mm: /* 19 insns */
 	mov		%o0, %o2	/* ARG2: mmu context */
 	mov		0, %o0		/* ARG0: CPU lists unimplemented */
 	mov		0, %o1		/* ARG1: CPU lists unimplemented */
 	mov		HV_MMU_ALL, %o3	/* ARG3: flags */
 	mov		HV_FAST_MMU_DEMAP_CTX, %o5
 	ta		HV_FAST_TRAP
-	brnz,pn		%o0, __hypervisor_tlb_tl0_error
+	brnz,pn		%o0, 1f
 	 mov		HV_FAST_MMU_DEMAP_CTX, %o1
 	retl
 	 nop
+1:	sethi		%hi(__hypervisor_tlb_tl0_error), %o5
+	jmpl		%o5 + %lo(__hypervisor_tlb_tl0_error), %g0
+	 nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
 
-__hypervisor_flush_tlb_page: /* 11 insns */
+__hypervisor_flush_tlb_page: /* 22 insns */
 	/* %o0 = context, %o1 = vaddr */
 	mov		%o0, %g2
 	mov		%o1, %o0              /* ARG0: vaddr + IMMU-bit */
@@ -330,12 +410,23 @@ __hypervisor_flush_tlb_page: /* 11 insns */
 	srlx		%o0, PAGE_SHIFT, %o0
 	sllx		%o0, PAGE_SHIFT, %o0
 	ta		HV_MMU_UNMAP_ADDR_TRAP
-	brnz,pn		%o0, __hypervisor_tlb_tl0_error
+	brnz,pn		%o0, 1f
 	 mov		HV_MMU_UNMAP_ADDR_TRAP, %o1
 	retl
 	 nop
+1:	sethi		%hi(__hypervisor_tlb_tl0_error), %o2
+	jmpl		%o2 + %lo(__hypervisor_tlb_tl0_error), %g0
+	 nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
 
-__hypervisor_flush_tlb_pending: /* 16 insns */
+__hypervisor_flush_tlb_pending: /* 27 insns */
 	/* %o0 = context, %o1 = nr, %o2 = vaddrs[] */
 	sllx		%o1, 3, %g1
 	mov		%o2, %g2
@@ -347,31 +438,57 @@ __hypervisor_flush_tlb_pending: /* 16 insns */
 	srlx		%o0, PAGE_SHIFT, %o0
 	sllx		%o0, PAGE_SHIFT, %o0
 	ta		HV_MMU_UNMAP_ADDR_TRAP
-	brnz,pn		%o0, __hypervisor_tlb_tl0_error
+	brnz,pn		%o0, 1f
 	 mov		HV_MMU_UNMAP_ADDR_TRAP, %o1
 	brnz,pt		%g1, 1b
 	 nop
 	retl
 	 nop
+1:	sethi		%hi(__hypervisor_tlb_tl0_error), %o2
+	jmpl		%o2 + %lo(__hypervisor_tlb_tl0_error), %g0
+	 nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
 
-__hypervisor_flush_tlb_kernel_range: /* 16 insns */
+__hypervisor_flush_tlb_kernel_range: /* 31 insns */
 	/* %o0=start, %o1=end */
 	cmp		%o0, %o1
 	be,pn		%xcc, 2f
-	 sethi		%hi(PAGE_SIZE), %g3
-	mov		%o0, %g1
-	sub		%o1, %g1, %g2
+	 sub		%o1, %o0, %g2
+	srlx		%g2, 18, %g3
+	brnz,pn		%g3, 4f
+	 mov		%o0, %g1
+	sethi		%hi(PAGE_SIZE), %g3
 	sub		%g2, %g3, %g2
 1:	add		%g1, %g2, %o0	/* ARG0: virtual address */
 	mov		0, %o1		/* ARG1: mmu context */
 	mov		HV_MMU_ALL, %o2	/* ARG2: flags */
 	ta		HV_MMU_UNMAP_ADDR_TRAP
-	brnz,pn		%o0, __hypervisor_tlb_tl0_error
+	brnz,pn		%o0, 3f
 	 mov		HV_MMU_UNMAP_ADDR_TRAP, %o1
 	brnz,pt		%g2, 1b
 	 sub		%g2, %g3, %g2
 2:	retl
 	 nop
+3:	sethi		%hi(__hypervisor_tlb_tl0_error), %o2
+	jmpl		%o2 + %lo(__hypervisor_tlb_tl0_error), %g0
+	 nop
+4:	mov		0, %o0		/* ARG0: CPU lists unimplemented */
+	mov		0, %o1		/* ARG1: CPU lists unimplemented */
+	mov		0, %o2		/* ARG2: mmu context == nucleus */
+	mov		HV_MMU_ALL, %o3	/* ARG3: flags */
+	mov		HV_FAST_MMU_DEMAP_CTX, %o5
+	ta		HV_FAST_TRAP
+	brnz,pn		%o0, 3b
+	 mov		HV_FAST_MMU_DEMAP_CTX, %o1
+	retl
+	 nop
 
 #ifdef DCACHE_ALIASING_POSSIBLE
 	/* XXX Niagara and friends have an 8K cache, so no aliasing is
@@ -394,43 +511,6 @@ tlb_patch_one:
 	retl
 	 nop
 
-	.globl		cheetah_patch_cachetlbops
-cheetah_patch_cachetlbops:
-	save		%sp, -128, %sp
-
-	sethi		%hi(__flush_tlb_mm), %o0
-	or		%o0, %lo(__flush_tlb_mm), %o0
-	sethi		%hi(__cheetah_flush_tlb_mm), %o1
-	or		%o1, %lo(__cheetah_flush_tlb_mm), %o1
-	call		tlb_patch_one
-	 mov		19, %o2
-
-	sethi		%hi(__flush_tlb_page), %o0
-	or		%o0, %lo(__flush_tlb_page), %o0
-	sethi		%hi(__cheetah_flush_tlb_page), %o1
-	or		%o1, %lo(__cheetah_flush_tlb_page), %o1
-	call		tlb_patch_one
-	 mov		22, %o2
-
-	sethi		%hi(__flush_tlb_pending), %o0
-	or		%o0, %lo(__flush_tlb_pending), %o0
-	sethi		%hi(__cheetah_flush_tlb_pending), %o1
-	or		%o1, %lo(__cheetah_flush_tlb_pending), %o1
-	call		tlb_patch_one
-	 mov		27, %o2
-
-#ifdef DCACHE_ALIASING_POSSIBLE
-	sethi		%hi(__flush_dcache_page), %o0
-	or		%o0, %lo(__flush_dcache_page), %o0
-	sethi		%hi(__cheetah_flush_dcache_page), %o1
-	or		%o1, %lo(__cheetah_flush_dcache_page), %o1
-	call		tlb_patch_one
-	 mov		11, %o2
-#endif /* DCACHE_ALIASING_POSSIBLE */
-
-	ret
-	 restore
-
 #ifdef CONFIG_SMP
 	/* These are all called by the slaves of a cross call, at
 	 * trap level 1, with interrupts fully disabled.
@@ -447,7 +527,7 @@ cheetah_patch_cachetlbops:
 	 */
 	.align		32
 	.globl		xcall_flush_tlb_mm
-xcall_flush_tlb_mm:	/* 21 insns */
+xcall_flush_tlb_mm:	/* 24 insns */
 	mov		PRIMARY_CONTEXT, %g2
 	ldxa		[%g2] ASI_DMMU, %g3
 	srlx		%g3, CTX_PGSZ1_NUC_SHIFT, %g4
@@ -469,9 +549,12 @@ xcall_flush_tlb_mm:	/* 21 insns */
 	nop
 	nop
 	nop
+	nop
+	nop
+	nop
 
 	.globl		xcall_flush_tlb_page
-xcall_flush_tlb_page:	/* 17 insns */
+xcall_flush_tlb_page:	/* 20 insns */
 	/* %g5=context, %g1=vaddr */
 	mov		PRIMARY_CONTEXT, %g4
 	ldxa		[%g4] ASI_DMMU, %g2
@@ -490,15 +573,20 @@ xcall_flush_tlb_page:	/* 17 insns */
 	retry
 	nop
 	nop
+	nop
+	nop
+	nop
 
 	.globl		xcall_flush_tlb_kernel_range
-xcall_flush_tlb_kernel_range:	/* 25 insns */
+xcall_flush_tlb_kernel_range:	/* 44 insns */
 	sethi		%hi(PAGE_SIZE - 1), %g2
 	or		%g2, %lo(PAGE_SIZE - 1), %g2
 	andn		%g1, %g2, %g1
 	andn		%g7, %g2, %g7
 	sub		%g7, %g1, %g3
-	add		%g2, 1, %g2
+	srlx		%g3, 18, %g2
+	brnz,pn		%g2, 2f
+	 add		%g2, 1, %g2
 	sub		%g3, %g2, %g3
 	or		%g1, 0x20, %g1		! Nucleus
 1:	stxa		%g0, [%g1 + %g3] ASI_DMMU_DEMAP
@@ -507,8 +595,25 @@ xcall_flush_tlb_kernel_range:	/* 25 insns */
 	brnz,pt		%g3, 1b
 	 sub		%g3, %g2, %g3
 	retry
-	nop
-	nop
+2:	mov		63 * 8, %g1
+1:	ldxa		[%g1] ASI_ITLB_DATA_ACCESS, %g2
+	andcc		%g2, 0x40, %g0			/* _PAGE_L_4U */
+	bne,pn		%xcc, 2f
+	 mov		TLB_TAG_ACCESS, %g2
+	stxa		%g0, [%g2] ASI_IMMU
+	stxa		%g0, [%g1] ASI_ITLB_DATA_ACCESS
+	membar		#Sync
+2:	ldxa		[%g1] ASI_DTLB_DATA_ACCESS, %g2
+	andcc		%g2, 0x40, %g0
+	bne,pn		%xcc, 2f
+	 mov		TLB_TAG_ACCESS, %g2
+	stxa		%g0, [%g2] ASI_DMMU
+	stxa		%g0, [%g1] ASI_DTLB_DATA_ACCESS
+	membar		#Sync
+2:	sub		%g1, 8, %g1
+	brgez,pt	%g1, 1b
+	 nop
+	retry
 	nop
 	nop
 	nop
@@ -637,6 +742,52 @@ xcall_fetch_glob_pmu_n4:
 
 	retry
 
+__cheetah_xcall_flush_tlb_kernel_range:	/* 44 insns */
+	sethi		%hi(PAGE_SIZE - 1), %g2
+	or		%g2, %lo(PAGE_SIZE - 1), %g2
+	andn		%g1, %g2, %g1
+	andn		%g7, %g2, %g7
+	sub		%g7, %g1, %g3
+	srlx		%g3, 18, %g2
+	brnz,pn		%g2, 2f
+	 add		%g2, 1, %g2
+	sub		%g3, %g2, %g3
+	or		%g1, 0x20, %g1		! Nucleus
+1:	stxa		%g0, [%g1 + %g3] ASI_DMMU_DEMAP
+	stxa		%g0, [%g1 + %g3] ASI_IMMU_DEMAP
+	membar		#Sync
+	brnz,pt		%g3, 1b
+	 sub		%g3, %g2, %g3
+	retry
+2:	mov		0x80, %g2
+	stxa		%g0, [%g2] ASI_DMMU_DEMAP
+	membar		#Sync
+	stxa		%g0, [%g2] ASI_IMMU_DEMAP
+	membar		#Sync
+	retry
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+
 #ifdef DCACHE_ALIASING_POSSIBLE
 	.align		32
 	.globl		xcall_flush_dcache_page_cheetah
@@ -700,7 +851,7 @@ __hypervisor_tlb_xcall_error:
 	ba,a,pt	%xcc, rtrap
 
 	.globl		__hypervisor_xcall_flush_tlb_mm
-__hypervisor_xcall_flush_tlb_mm: /* 21 insns */
+__hypervisor_xcall_flush_tlb_mm: /* 24 insns */
 	/* %g5=ctx, g1,g2,g3,g4,g7=scratch, %g6=unusable */
 	mov		%o0, %g2
 	mov		%o1, %g3
@@ -714,7 +865,7 @@ __hypervisor_xcall_flush_tlb_mm: /* 21 insns */
 	mov		HV_FAST_MMU_DEMAP_CTX, %o5
 	ta		HV_FAST_TRAP
 	mov		HV_FAST_MMU_DEMAP_CTX, %g6
-	brnz,pn		%o0, __hypervisor_tlb_xcall_error
+	brnz,pn		%o0, 1f
 	 mov		%o0, %g5
 	mov		%g2, %o0
 	mov		%g3, %o1
@@ -723,9 +874,12 @@ __hypervisor_xcall_flush_tlb_mm: /* 21 insns */
 	mov		%g7, %o5
 	membar		#Sync
 	retry
+1:	sethi		%hi(__hypervisor_tlb_xcall_error), %g4
+	jmpl		%g4 + %lo(__hypervisor_tlb_xcall_error), %g0
+	 nop
 
 	.globl		__hypervisor_xcall_flush_tlb_page
-__hypervisor_xcall_flush_tlb_page: /* 17 insns */
+__hypervisor_xcall_flush_tlb_page: /* 20 insns */
 	/* %g5=ctx, %g1=vaddr */
 	mov		%o0, %g2
 	mov		%o1, %g3
@@ -737,42 +891,64 @@ __hypervisor_xcall_flush_tlb_page: /* 17 insns */
 	sllx		%o0, PAGE_SHIFT, %o0
 	ta		HV_MMU_UNMAP_ADDR_TRAP
 	mov		HV_MMU_UNMAP_ADDR_TRAP, %g6
-	brnz,a,pn	%o0, __hypervisor_tlb_xcall_error
+	brnz,a,pn	%o0, 1f
 	 mov		%o0, %g5
 	mov		%g2, %o0
 	mov		%g3, %o1
 	mov		%g4, %o2
 	membar		#Sync
 	retry
+1:	sethi		%hi(__hypervisor_tlb_xcall_error), %g4
+	jmpl		%g4 + %lo(__hypervisor_tlb_xcall_error), %g0
+	 nop
 
 	.globl		__hypervisor_xcall_flush_tlb_kernel_range
-__hypervisor_xcall_flush_tlb_kernel_range: /* 25 insns */
+__hypervisor_xcall_flush_tlb_kernel_range: /* 44 insns */
 	/* %g1=start, %g7=end, g2,g3,g4,g5,g6=scratch */
 	sethi		%hi(PAGE_SIZE - 1), %g2
 	or		%g2, %lo(PAGE_SIZE - 1), %g2
 	andn		%g1, %g2, %g1
 	andn		%g7, %g2, %g7
 	sub		%g7, %g1, %g3
+	srlx		%g3, 18, %g7
 	add		%g2, 1, %g2
 	sub		%g3, %g2, %g3
 	mov		%o0, %g2
 	mov		%o1, %g4
-	mov		%o2, %g7
+	brnz,pn		%g7, 2f
+	 mov		%o2, %g7
 1:	add		%g1, %g3, %o0	/* ARG0: virtual address */
 	mov		0, %o1		/* ARG1: mmu context */
 	mov		HV_MMU_ALL, %o2	/* ARG2: flags */
 	ta		HV_MMU_UNMAP_ADDR_TRAP
 	mov		HV_MMU_UNMAP_ADDR_TRAP, %g6
-	brnz,pn		%o0, __hypervisor_tlb_xcall_error
+	brnz,pn		%o0, 1f
 	 mov		%o0, %g5
 	sethi		%hi(PAGE_SIZE), %o2
 	brnz,pt		%g3, 1b
 	 sub		%g3, %o2, %g3
-	mov		%g2, %o0
+5:	mov		%g2, %o0
 	mov		%g4, %o1
 	mov		%g7, %o2
 	membar		#Sync
 	retry
+1:	sethi		%hi(__hypervisor_tlb_xcall_error), %g4
+	jmpl		%g4 + %lo(__hypervisor_tlb_xcall_error), %g0
+	 nop
+2:	mov		%o3, %g1
+	mov		%o5, %g3
+	mov		0, %o0		/* ARG0: CPU lists unimplemented */
+	mov		0, %o1		/* ARG1: CPU lists unimplemented */
+	mov		0, %o2		/* ARG2: mmu context == nucleus */
+	mov		HV_MMU_ALL, %o3	/* ARG3: flags */
+	mov		HV_FAST_MMU_DEMAP_CTX, %o5
+	ta		HV_FAST_TRAP
+	mov		%g1, %o3
+	brz,pt		%o0, 5b
+	 mov		%g3, %o5
+	mov		HV_FAST_MMU_DEMAP_CTX, %g6
+	ba,pt		%xcc, 1b
+	 clr		%g5
 
 	/* These just get rescheduled to PIL vectors. */
 	.globl		xcall_call_function
@@ -809,6 +985,58 @@ xcall_kgdb_capture:
 
 #endif /* CONFIG_SMP */
 
+	.globl		cheetah_patch_cachetlbops
+cheetah_patch_cachetlbops:
+	save		%sp, -128, %sp
+
+	sethi		%hi(__flush_tlb_mm), %o0
+	or		%o0, %lo(__flush_tlb_mm), %o0
+	sethi		%hi(__cheetah_flush_tlb_mm), %o1
+	or		%o1, %lo(__cheetah_flush_tlb_mm), %o1
+	call		tlb_patch_one
+	 mov		19, %o2
+
+	sethi		%hi(__flush_tlb_page), %o0
+	or		%o0, %lo(__flush_tlb_page), %o0
+	sethi		%hi(__cheetah_flush_tlb_page), %o1
+	or		%o1, %lo(__cheetah_flush_tlb_page), %o1
+	call		tlb_patch_one
+	 mov		22, %o2
+
+	sethi		%hi(__flush_tlb_pending), %o0
+	or		%o0, %lo(__flush_tlb_pending), %o0
+	sethi		%hi(__cheetah_flush_tlb_pending), %o1
+	or		%o1, %lo(__cheetah_flush_tlb_pending), %o1
+	call		tlb_patch_one
+	 mov		27, %o2
+
+	sethi		%hi(__flush_tlb_kernel_range), %o0
+	or		%o0, %lo(__flush_tlb_kernel_range), %o0
+	sethi		%hi(__cheetah_flush_tlb_kernel_range), %o1
+	or		%o1, %lo(__cheetah_flush_tlb_kernel_range), %o1
+	call		tlb_patch_one
+	 mov		31, %o2
+
+#ifdef DCACHE_ALIASING_POSSIBLE
+	sethi		%hi(__flush_dcache_page), %o0
+	or		%o0, %lo(__flush_dcache_page), %o0
+	sethi		%hi(__cheetah_flush_dcache_page), %o1
+	or		%o1, %lo(__cheetah_flush_dcache_page), %o1
+	call		tlb_patch_one
+	 mov		11, %o2
+#endif /* DCACHE_ALIASING_POSSIBLE */
+
+#ifdef CONFIG_SMP
+	sethi		%hi(xcall_flush_tlb_kernel_range), %o0
+	or		%o0, %lo(xcall_flush_tlb_kernel_range), %o0
+	sethi		%hi(__cheetah_xcall_flush_tlb_kernel_range), %o1
+	or		%o1, %lo(__cheetah_xcall_flush_tlb_kernel_range), %o1
+	call		tlb_patch_one
+	 mov		44, %o2
+#endif /* CONFIG_SMP */
+
+	ret
+	 restore
 
 	.globl		hypervisor_patch_cachetlbops
 hypervisor_patch_cachetlbops:
@@ -819,28 +1047,28 @@ hypervisor_patch_cachetlbops:
 	sethi		%hi(__hypervisor_flush_tlb_mm), %o1
 	or		%o1, %lo(__hypervisor_flush_tlb_mm), %o1
 	call		tlb_patch_one
-	 mov		10, %o2
+	 mov		19, %o2
 
 	sethi		%hi(__flush_tlb_page), %o0
 	or		%o0, %lo(__flush_tlb_page), %o0
 	sethi		%hi(__hypervisor_flush_tlb_page), %o1
 	or		%o1, %lo(__hypervisor_flush_tlb_page), %o1
 	call		tlb_patch_one
-	 mov		11, %o2
+	 mov		22, %o2
 
 	sethi		%hi(__flush_tlb_pending), %o0
 	or		%o0, %lo(__flush_tlb_pending), %o0
 	sethi		%hi(__hypervisor_flush_tlb_pending), %o1
 	or		%o1, %lo(__hypervisor_flush_tlb_pending), %o1
 	call		tlb_patch_one
-	 mov		16, %o2
+	 mov		27, %o2
 
 	sethi		%hi(__flush_tlb_kernel_range), %o0
 	or		%o0, %lo(__flush_tlb_kernel_range), %o0
 	sethi		%hi(__hypervisor_flush_tlb_kernel_range), %o1
 	or		%o1, %lo(__hypervisor_flush_tlb_kernel_range), %o1
 	call		tlb_patch_one
-	 mov		16, %o2
+	 mov		31, %o2
 
 #ifdef DCACHE_ALIASING_POSSIBLE
 	sethi		%hi(__flush_dcache_page), %o0
@@ -857,21 +1085,21 @@ hypervisor_patch_cachetlbops:
 	sethi		%hi(__hypervisor_xcall_flush_tlb_mm), %o1
 	or		%o1, %lo(__hypervisor_xcall_flush_tlb_mm), %o1
 	call		tlb_patch_one
-	 mov		21, %o2
+	 mov		24, %o2
 
 	sethi		%hi(xcall_flush_tlb_page), %o0
 	or		%o0, %lo(xcall_flush_tlb_page), %o0
 	sethi		%hi(__hypervisor_xcall_flush_tlb_page), %o1
 	or		%o1, %lo(__hypervisor_xcall_flush_tlb_page), %o1
 	call		tlb_patch_one
-	 mov		17, %o2
+	 mov		20, %o2
 
 	sethi		%hi(xcall_flush_tlb_kernel_range), %o0
 	or		%o0, %lo(xcall_flush_tlb_kernel_range), %o0
 	sethi		%hi(__hypervisor_xcall_flush_tlb_kernel_range), %o1
 	or		%o1, %lo(__hypervisor_xcall_flush_tlb_kernel_range), %o1
 	call		tlb_patch_one
-	 mov		25, %o2
+	 mov		44, %o2
 #endif /* CONFIG_SMP */
 
 	ret