Merge remote-tracking branch 'origin' into spi/next

Pull in Linus' tree to pick up changes required for the langwell gpio fixes

11 files changed, 329 insertions, 334 deletions

diff --git a/arch/x86/lib/Makefile b/arch/x86/lib/Makefile
index e10cf070ede..f2479f19ddd 100644
--- a/arch/x86/lib/Makefile
+++ b/arch/x86/lib/Makefile
@@ -42,4 +42,5 @@ else
         lib-y += memmove_64.o memset_64.o
         lib-y += copy_user_64.o rwlock_64.o copy_user_nocache_64.o
 	lib-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem_64.o
+	lib-y += cmpxchg16b_emu.o
 endif
diff --git a/arch/x86/lib/atomic64_386_32.S b/arch/x86/lib/atomic64_386_32.S
index 2cda60a06e6..e8e7e0d06f4 100644
--- a/arch/x86/lib/atomic64_386_32.S
+++ b/arch/x86/lib/atomic64_386_32.S
@@ -15,14 +15,12 @@
 
 /* if you want SMP support, implement these with real spinlocks */
 .macro LOCK reg
-	pushfl
-	CFI_ADJUST_CFA_OFFSET 4
+	pushfl_cfi
 	cli
 .endm
 
 .macro UNLOCK reg
-	popfl
-	CFI_ADJUST_CFA_OFFSET -4
+	popfl_cfi
 .endm
 
 #define BEGIN(op) \
diff --git a/arch/x86/lib/atomic64_cx8_32.S b/arch/x86/lib/atomic64_cx8_32.S
index 71e080de335..391a083674b 100644
--- a/arch/x86/lib/atomic64_cx8_32.S
+++ b/arch/x86/lib/atomic64_cx8_32.S
@@ -14,14 +14,12 @@
 #include <asm/dwarf2.h>
 
 .macro SAVE reg
-	pushl %\reg
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %\reg
 	CFI_REL_OFFSET \reg, 0
 .endm
 
 .macro RESTORE reg
-	popl %\reg
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %\reg
 	CFI_RESTORE \reg
 .endm
 
diff --git a/arch/x86/lib/checksum_32.S b/arch/x86/lib/checksum_32.S
index adbccd0bbb7..78d16a554db 100644
--- a/arch/x86/lib/checksum_32.S
+++ b/arch/x86/lib/checksum_32.S
@@ -50,11 +50,9 @@ unsigned int csum_partial(const unsigned char * buff, int len, unsigned int sum)
 	   */		
 ENTRY(csum_partial)
 	CFI_STARTPROC
-	pushl %esi
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %esi
 	CFI_REL_OFFSET esi, 0
-	pushl %ebx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ebx
 	CFI_REL_OFFSET ebx, 0
 	movl 20(%esp),%eax	# Function arg: unsigned int sum
 	movl 16(%esp),%ecx	# Function arg: int len
@@ -132,11 +130,9 @@ ENTRY(csum_partial)
 	jz 8f
 	roll $8, %eax
 8:
-	popl %ebx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %ebx
 	CFI_RESTORE ebx
-	popl %esi
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %esi
 	CFI_RESTORE esi
 	ret
 	CFI_ENDPROC
@@ -148,11 +144,9 @@ ENDPROC(csum_partial)
 
 ENTRY(csum_partial)
 	CFI_STARTPROC
-	pushl %esi
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %esi
 	CFI_REL_OFFSET esi, 0
-	pushl %ebx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ebx
 	CFI_REL_OFFSET ebx, 0
 	movl 20(%esp),%eax	# Function arg: unsigned int sum
 	movl 16(%esp),%ecx	# Function arg: int len
@@ -260,11 +254,9 @@ ENTRY(csum_partial)
 	jz 90f
 	roll $8, %eax
 90: 
-	popl %ebx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %ebx
 	CFI_RESTORE ebx
-	popl %esi
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %esi
 	CFI_RESTORE esi
 	ret
 	CFI_ENDPROC
@@ -309,14 +301,11 @@ ENTRY(csum_partial_copy_generic)
 	CFI_STARTPROC
 	subl  $4,%esp	
 	CFI_ADJUST_CFA_OFFSET 4
-	pushl %edi
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %edi
 	CFI_REL_OFFSET edi, 0
-	pushl %esi
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %esi
 	CFI_REL_OFFSET esi, 0
-	pushl %ebx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ebx
 	CFI_REL_OFFSET ebx, 0
 	movl ARGBASE+16(%esp),%eax	# sum
 	movl ARGBASE+12(%esp),%ecx	# len
@@ -426,17 +415,13 @@ DST(	movb %cl, (%edi)	)
 
 .previous
 
-	popl %ebx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %ebx
 	CFI_RESTORE ebx
-	popl %esi
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %esi
 	CFI_RESTORE esi
-	popl %edi
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %edi
 	CFI_RESTORE edi
-	popl %ecx			# equivalent to addl $4,%esp
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %ecx			# equivalent to addl $4,%esp
 	ret	
 	CFI_ENDPROC
 ENDPROC(csum_partial_copy_generic)
@@ -459,14 +444,11 @@ ENDPROC(csum_partial_copy_generic)
 		
 ENTRY(csum_partial_copy_generic)
 	CFI_STARTPROC
-	pushl %ebx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ebx
 	CFI_REL_OFFSET ebx, 0
-	pushl %edi
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %edi
 	CFI_REL_OFFSET edi, 0
-	pushl %esi
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %esi
 	CFI_REL_OFFSET esi, 0
 	movl ARGBASE+4(%esp),%esi	#src
 	movl ARGBASE+8(%esp),%edi	#dst	
@@ -527,14 +509,11 @@ DST(	movb %dl, (%edi)         )
 	jmp  7b			
 .previous				
 
-	popl %esi
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %esi
 	CFI_RESTORE esi
-	popl %edi
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %edi
 	CFI_RESTORE edi
-	popl %ebx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %ebx
 	CFI_RESTORE ebx
 	ret
 	CFI_ENDPROC
diff --git a/arch/x86/lib/cmpxchg16b_emu.S b/arch/x86/lib/cmpxchg16b_emu.S
new file mode 100644
index 00000000000..3e8b08a6de2
--- /dev/null
+++ b/arch/x86/lib/cmpxchg16b_emu.S
@@ -0,0 +1,59 @@
+/*
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; version 2
+ *	of the License.
+ *
+ */
+#include <linux/linkage.h>
+#include <asm/alternative-asm.h>
+#include <asm/frame.h>
+#include <asm/dwarf2.h>
+
+.text
+
+/*
+ * Inputs:
+ * %rsi : memory location to compare
+ * %rax : low 64 bits of old value
+ * %rdx : high 64 bits of old value
+ * %rbx : low 64 bits of new value
+ * %rcx : high 64 bits of new value
+ * %al  : Operation successful
+ */
+ENTRY(this_cpu_cmpxchg16b_emu)
+CFI_STARTPROC
+
+#
+# Emulate 'cmpxchg16b %gs:(%rsi)' except we return the result in %al not
+# via the ZF.  Caller will access %al to get result.
+#
+# Note that this is only useful for a cpuops operation.  Meaning that we
+# do *not* have a fully atomic operation but just an operation that is
+# *atomic* on a single cpu (as provided by the this_cpu_xx class of
+# macros).
+#
+this_cpu_cmpxchg16b_emu:
+	pushf
+	cli
+
+	cmpq %gs:(%rsi), %rax
+	jne not_same
+	cmpq %gs:8(%rsi), %rdx
+	jne not_same
+
+	movq %rbx, %gs:(%rsi)
+	movq %rcx, %gs:8(%rsi)
+
+	popf
+	mov $1, %al
+	ret
+
+ not_same:
+	popf
+	xor %al,%al
+	ret
+
+CFI_ENDPROC
+
+ENDPROC(this_cpu_cmpxchg16b_emu)
diff --git a/arch/x86/lib/memmove_64.S b/arch/x86/lib/memmove_64.S
new file mode 100644
index 00000000000..0ecb8433e5a
--- /dev/null
+++ b/arch/x86/lib/memmove_64.S
@@ -0,0 +1,197 @@
+/*
+ * Normally compiler builtins are used, but sometimes the compiler calls out
+ * of line code. Based on asm-i386/string.h.
+ *
+ * This assembly file is re-written from memmove_64.c file.
+ *	- Copyright 2011 Fenghua Yu <fenghua.yu@intel.com>
+ */
+#define _STRING_C
+#include <linux/linkage.h>
+#include <asm/dwarf2.h>
+
+#undef memmove
+
+/*
+ * Implement memmove(). This can handle overlap between src and dst.
+ *
+ * Input:
+ * rdi: dest
+ * rsi: src
+ * rdx: count
+ *
+ * Output:
+ * rax: dest
+ */
+ENTRY(memmove)
+	CFI_STARTPROC
+	/* Handle more 32bytes in loop */
+	mov %rdi, %rax
+	cmp $0x20, %rdx
+	jb	1f
+
+	/* Decide forward/backward copy mode */
+	cmp %rdi, %rsi
+	jb	2f
+
+	/*
+	 * movsq instruction have many startup latency
+	 * so we handle small size by general register.
+	 */
+	cmp  $680, %rdx
+	jb	3f
+	/*
+	 * movsq instruction is only good for aligned case.
+	 */
+
+	cmpb %dil, %sil
+	je 4f
+3:
+	sub $0x20, %rdx
+	/*
+	 * We gobble 32byts forward in each loop.
+	 */
+5:
+	sub $0x20, %rdx
+	movq 0*8(%rsi), %r11
+	movq 1*8(%rsi), %r10
+	movq 2*8(%rsi), %r9
+	movq 3*8(%rsi), %r8
+	leaq 4*8(%rsi), %rsi
+
+	movq %r11, 0*8(%rdi)
+	movq %r10, 1*8(%rdi)
+	movq %r9, 2*8(%rdi)
+	movq %r8, 3*8(%rdi)
+	leaq 4*8(%rdi), %rdi
+	jae 5b
+	addq $0x20, %rdx
+	jmp 1f
+	/*
+	 * Handle data forward by movsq.
+	 */
+	.p2align 4
+4:
+	movq %rdx, %rcx
+	movq -8(%rsi, %rdx), %r11
+	lea -8(%rdi, %rdx), %r10
+	shrq $3, %rcx
+	rep movsq
+	movq %r11, (%r10)
+	jmp 13f
+	/*
+	 * Handle data backward by movsq.
+	 */
+	.p2align 4
+7:
+	movq %rdx, %rcx
+	movq (%rsi), %r11
+	movq %rdi, %r10
+	leaq -8(%rsi, %rdx), %rsi
+	leaq -8(%rdi, %rdx), %rdi
+	shrq $3, %rcx
+	std
+	rep movsq
+	cld
+	movq %r11, (%r10)
+	jmp 13f
+
+	/*
+	 * Start to prepare for backward copy.
+	 */
+	.p2align 4
+2:
+	cmp $680, %rdx
+	jb 6f
+	cmp %dil, %sil
+	je 7b
+6:
+	/*
+	 * Calculate copy position to tail.
+	 */
+	addq %rdx, %rsi
+	addq %rdx, %rdi
+	subq $0x20, %rdx
+	/*
+	 * We gobble 32byts backward in each loop.
+	 */
+8:
+	subq $0x20, %rdx
+	movq -1*8(%rsi), %r11
+	movq -2*8(%rsi), %r10
+	movq -3*8(%rsi), %r9
+	movq -4*8(%rsi), %r8
+	leaq -4*8(%rsi), %rsi
+
+	movq %r11, -1*8(%rdi)
+	movq %r10, -2*8(%rdi)
+	movq %r9, -3*8(%rdi)
+	movq %r8, -4*8(%rdi)
+	leaq -4*8(%rdi), %rdi
+	jae 8b
+	/*
+	 * Calculate copy position to head.
+	 */
+	addq $0x20, %rdx
+	subq %rdx, %rsi
+	subq %rdx, %rdi
+1:
+	cmpq $16, %rdx
+	jb 9f
+	/*
+	 * Move data from 16 bytes to 31 bytes.
+	 */
+	movq 0*8(%rsi), %r11
+	movq 1*8(%rsi), %r10
+	movq -2*8(%rsi, %rdx), %r9
+	movq -1*8(%rsi, %rdx), %r8
+	movq %r11, 0*8(%rdi)
+	movq %r10, 1*8(%rdi)
+	movq %r9, -2*8(%rdi, %rdx)
+	movq %r8, -1*8(%rdi, %rdx)
+	jmp 13f
+	.p2align 4
+9:
+	cmpq $8, %rdx
+	jb 10f
+	/*
+	 * Move data from 8 bytes to 15 bytes.
+	 */
+	movq 0*8(%rsi), %r11
+	movq -1*8(%rsi, %rdx), %r10
+	movq %r11, 0*8(%rdi)
+	movq %r10, -1*8(%rdi, %rdx)
+	jmp 13f
+10:
+	cmpq $4, %rdx
+	jb 11f
+	/*
+	 * Move data from 4 bytes to 7 bytes.
+	 */
+	movl (%rsi), %r11d
+	movl -4(%rsi, %rdx), %r10d
+	movl %r11d, (%rdi)
+	movl %r10d, -4(%rdi, %rdx)
+	jmp 13f
+11:
+	cmp $2, %rdx
+	jb 12f
+	/*
+	 * Move data from 2 bytes to 3 bytes.
+	 */
+	movw (%rsi), %r11w
+	movw -2(%rsi, %rdx), %r10w
+	movw %r11w, (%rdi)
+	movw %r10w, -2(%rdi, %rdx)
+	jmp 13f
+12:
+	cmp $1, %rdx
+	jb 13f
+	/*
+	 * Move data for 1 byte.
+	 */
+	movb (%rsi), %r11b
+	movb %r11b, (%rdi)
+13:
+	retq
+	CFI_ENDPROC
+ENDPROC(memmove)
diff --git a/arch/x86/lib/memmove_64.c b/arch/x86/lib/memmove_64.c
deleted file mode 100644
index 6d0f0ec41b3..00000000000
--- a/arch/x86/lib/memmove_64.c
+++ /dev/null
@@ -1,192 +0,0 @@
-/* Normally compiler builtins are used, but sometimes the compiler calls out
-   of line code. Based on asm-i386/string.h.
- */
-#define _STRING_C
-#include <linux/string.h>
-#include <linux/module.h>
-
-#undef memmove
-void *memmove(void *dest, const void *src, size_t count)
-{
-	unsigned long d0,d1,d2,d3,d4,d5,d6,d7;
-	char *ret;
-
-	__asm__ __volatile__(
-		/* Handle more 32bytes in loop */
-		"mov %2, %3\n\t"
-		"cmp $0x20, %0\n\t"
-		"jb	1f\n\t"
-
-		/* Decide forward/backward copy mode */
-		"cmp %2, %1\n\t"
-		"jb	2f\n\t"
-
-		/*
-		 * movsq instruction have many startup latency
-		 * so we handle small size by general register.
-		 */
-		"cmp  $680, %0\n\t"
-		"jb 3f\n\t"
-		/*
-		 * movsq instruction is only good for aligned case.
-		 */
-		"cmpb %%dil, %%sil\n\t"
-		"je 4f\n\t"
-		"3:\n\t"
-		"sub $0x20, %0\n\t"
-		/*
-		 * We gobble 32byts forward in each loop.
-		 */
-		"5:\n\t"
-		"sub $0x20, %0\n\t"
-		"movq 0*8(%1), %4\n\t"
-		"movq 1*8(%1), %5\n\t"
-		"movq 2*8(%1), %6\n\t"
-		"movq 3*8(%1), %7\n\t"
-		"leaq 4*8(%1), %1\n\t"
-
-		"movq %4, 0*8(%2)\n\t"
-		"movq %5, 1*8(%2)\n\t"
-		"movq %6, 2*8(%2)\n\t"
-		"movq %7, 3*8(%2)\n\t"
-		"leaq 4*8(%2), %2\n\t"
-		"jae 5b\n\t"
-		"addq $0x20, %0\n\t"
-		"jmp 1f\n\t"
-		/*
-		 * Handle data forward by movsq.
-		 */
-		".p2align 4\n\t"
-		"4:\n\t"
-		"movq %0, %8\n\t"
-		"movq -8(%1, %0), %4\n\t"
-		"lea -8(%2, %0), %5\n\t"
-		"shrq $3, %8\n\t"
-		"rep movsq\n\t"
-		"movq %4, (%5)\n\t"
-		"jmp 13f\n\t"
-		/*
-		 * Handle data backward by movsq.
-		 */
-		".p2align 4\n\t"
-		"7:\n\t"
-		"movq %0, %8\n\t"
-		"movq (%1), %4\n\t"
-		"movq %2, %5\n\t"
-		"leaq -8(%1, %0), %1\n\t"
-		"leaq -8(%2, %0), %2\n\t"
-		"shrq $3, %8\n\t"
-		"std\n\t"
-		"rep movsq\n\t"
-		"cld\n\t"
-		"movq %4, (%5)\n\t"
-		"jmp 13f\n\t"
-
-		/*
-		 * Start to prepare for backward copy.
-		 */
-		".p2align 4\n\t"
-		"2:\n\t"
-		"cmp $680, %0\n\t"
-		"jb 6f \n\t"
-		"cmp %%dil, %%sil\n\t"
-		"je 7b \n\t"
-		"6:\n\t"
-		/*
-		 * Calculate copy position to tail.
-		 */
-		"addq %0, %1\n\t"
-		"addq %0, %2\n\t"
-		"subq $0x20, %0\n\t"
-		/*
-		 * We gobble 32byts backward in each loop.
-		 */
-		"8:\n\t"
-		"subq $0x20, %0\n\t"
-		"movq -1*8(%1), %4\n\t"
-		"movq -2*8(%1), %5\n\t"
-		"movq -3*8(%1), %6\n\t"
-		"movq -4*8(%1), %7\n\t"
-		"leaq -4*8(%1), %1\n\t"
-
-		"movq %4, -1*8(%2)\n\t"
-		"movq %5, -2*8(%2)\n\t"
-		"movq %6, -3*8(%2)\n\t"
-		"movq %7, -4*8(%2)\n\t"
-		"leaq -4*8(%2), %2\n\t"
-		"jae 8b\n\t"
-		/*
-		 * Calculate copy position to head.
-		 */
-		"addq $0x20, %0\n\t"
-		"subq %0, %1\n\t"
-		"subq %0, %2\n\t"
-		"1:\n\t"
-		"cmpq $16, %0\n\t"
-		"jb 9f\n\t"
-		/*
-		 * Move data from 16 bytes to 31 bytes.
-		 */
-		"movq 0*8(%1), %4\n\t"
-		"movq 1*8(%1), %5\n\t"
-		"movq -2*8(%1, %0), %6\n\t"
-		"movq -1*8(%1, %0), %7\n\t"
-		"movq %4, 0*8(%2)\n\t"
-		"movq %5, 1*8(%2)\n\t"
-		"movq %6, -2*8(%2, %0)\n\t"
-		"movq %7, -1*8(%2, %0)\n\t"
-		"jmp 13f\n\t"
-		".p2align 4\n\t"
-		"9:\n\t"
-		"cmpq $8, %0\n\t"
-		"jb 10f\n\t"
-		/*
-		 * Move data from 8 bytes to 15 bytes.
-		 */
-		"movq 0*8(%1), %4\n\t"
-		"movq -1*8(%1, %0), %5\n\t"
-		"movq %4, 0*8(%2)\n\t"
-		"movq %5, -1*8(%2, %0)\n\t"
-		"jmp 13f\n\t"
-		"10:\n\t"
-		"cmpq $4, %0\n\t"
-		"jb 11f\n\t"
-		/*
-		 * Move data from 4 bytes to 7 bytes.
-		 */
-		"movl (%1), %4d\n\t"
-		"movl -4(%1, %0), %5d\n\t"
-		"movl %4d, (%2)\n\t"
-		"movl %5d, -4(%2, %0)\n\t"
-		"jmp 13f\n\t"
-		"11:\n\t"
-		"cmp $2, %0\n\t"
-		"jb 12f\n\t"
-		/*
-		 * Move data from 2 bytes to 3 bytes.
-		 */
-		"movw (%1), %4w\n\t"
-		"movw -2(%1, %0), %5w\n\t"
-		"movw %4w, (%2)\n\t"
-		"movw %5w, -2(%2, %0)\n\t"
-		"jmp 13f\n\t"
-		"12:\n\t"
-		"cmp $1, %0\n\t"
-		"jb 13f\n\t"
-		/*
-		 * Move data for 1 byte.
-		 */
-		"movb (%1), %4b\n\t"
-		"movb %4b, (%2)\n\t"
-		"13:\n\t"
-		: "=&d" (d0), "=&S" (d1), "=&D" (d2), "=&a" (ret) ,
-		  "=r"(d3), "=r"(d4), "=r"(d5), "=r"(d6), "=&c" (d7)
-		:"0" (count),
-		 "1" (src),
-		 "2" (dest)
-		:"memory");
-
-		return ret;
-
-}
-EXPORT_SYMBOL(memmove);
diff --git a/arch/x86/lib/rwsem_64.S b/arch/x86/lib/rwsem_64.S
index 41fcf00e49d..67743977398 100644
--- a/arch/x86/lib/rwsem_64.S
+++ b/arch/x86/lib/rwsem_64.S
@@ -23,43 +23,50 @@
 #include <asm/dwarf2.h>
 
 #define save_common_regs \
-	pushq %rdi; \
-	pushq %rsi; \
-	pushq %rcx; \
-	pushq %r8; \
-	pushq %r9; \
-	pushq %r10; \
-	pushq %r11
+	pushq_cfi %rdi; CFI_REL_OFFSET rdi, 0; \
+	pushq_cfi %rsi; CFI_REL_OFFSET rsi, 0; \
+	pushq_cfi %rcx; CFI_REL_OFFSET rcx, 0; \
+	pushq_cfi %r8;  CFI_REL_OFFSET r8,  0; \
+	pushq_cfi %r9;  CFI_REL_OFFSET r9,  0; \
+	pushq_cfi %r10; CFI_REL_OFFSET r10, 0; \
+	pushq_cfi %r11; CFI_REL_OFFSET r11, 0
 
 #define restore_common_regs \
-	popq %r11; \
-	popq %r10; \
-	popq %r9; \
-	popq %r8; \
-	popq %rcx; \
-	popq %rsi; \
-	popq %rdi
+	popq_cfi %r11; CFI_RESTORE r11; \
+	popq_cfi %r10; CFI_RESTORE r10; \
+	popq_cfi %r9;  CFI_RESTORE r9; \
+	popq_cfi %r8;  CFI_RESTORE r8; \
+	popq_cfi %rcx; CFI_RESTORE rcx; \
+	popq_cfi %rsi; CFI_RESTORE rsi; \
+	popq_cfi %rdi; CFI_RESTORE rdi
 
 /* Fix up special calling conventions */
 ENTRY(call_rwsem_down_read_failed)
+	CFI_STARTPROC
 	save_common_regs
-	pushq %rdx
+	pushq_cfi %rdx
+	CFI_REL_OFFSET rdx, 0
 	movq %rax,%rdi
 	call rwsem_down_read_failed
-	popq %rdx
+	popq_cfi %rdx
+	CFI_RESTORE rdx
 	restore_common_regs
 	ret
-	ENDPROC(call_rwsem_down_read_failed)
+	CFI_ENDPROC
+ENDPROC(call_rwsem_down_read_failed)
 
 ENTRY(call_rwsem_down_write_failed)
+	CFI_STARTPROC
 	save_common_regs
 	movq %rax,%rdi
 	call rwsem_down_write_failed
 	restore_common_regs
 	ret
-	ENDPROC(call_rwsem_down_write_failed)
+	CFI_ENDPROC
+ENDPROC(call_rwsem_down_write_failed)
 
 ENTRY(call_rwsem_wake)
+	CFI_STARTPROC
 	decl %edx	/* do nothing if still outstanding active readers */
 	jnz 1f
 	save_common_regs
@@ -67,15 +74,20 @@ ENTRY(call_rwsem_wake)
 	call rwsem_wake
 	restore_common_regs
 1:	ret
-	ENDPROC(call_rwsem_wake)
+	CFI_ENDPROC
+ENDPROC(call_rwsem_wake)
 
 /* Fix up special calling conventions */
 ENTRY(call_rwsem_downgrade_wake)
+	CFI_STARTPROC
 	save_common_regs
-	pushq %rdx
+	pushq_cfi %rdx
+	CFI_REL_OFFSET rdx, 0
 	movq %rax,%rdi
 	call rwsem_downgrade_wake
-	popq %rdx
+	popq_cfi %rdx
+	CFI_RESTORE rdx
 	restore_common_regs
 	ret
-	ENDPROC(call_rwsem_downgrade_wake)
+	CFI_ENDPROC
+ENDPROC(call_rwsem_downgrade_wake)
diff --git a/arch/x86/lib/semaphore_32.S b/arch/x86/lib/semaphore_32.S
index 648fe474178..06691daa410 100644
--- a/arch/x86/lib/semaphore_32.S
+++ b/arch/x86/lib/semaphore_32.S
@@ -36,7 +36,7 @@
  */
 #ifdef CONFIG_SMP
 ENTRY(__write_lock_failed)
-	CFI_STARTPROC simple
+	CFI_STARTPROC
 	FRAME
 2: 	LOCK_PREFIX
 	addl	$ RW_LOCK_BIAS,(%eax)
@@ -74,29 +74,23 @@ ENTRY(__read_lock_failed)
 /* Fix up special calling conventions */
 ENTRY(call_rwsem_down_read_failed)
 	CFI_STARTPROC
-	push %ecx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ecx
 	CFI_REL_OFFSET ecx,0
-	push %edx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %edx
 	CFI_REL_OFFSET edx,0
 	call rwsem_down_read_failed
-	pop %edx
-	CFI_ADJUST_CFA_OFFSET -4
-	pop %ecx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %edx
+	popl_cfi %ecx
 	ret
 	CFI_ENDPROC
 	ENDPROC(call_rwsem_down_read_failed)
 
 ENTRY(call_rwsem_down_write_failed)
 	CFI_STARTPROC
-	push %ecx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ecx
 	CFI_REL_OFFSET ecx,0
 	calll rwsem_down_write_failed
-	pop %ecx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %ecx
 	ret
 	CFI_ENDPROC
 	ENDPROC(call_rwsem_down_write_failed)
@@ -105,12 +99,10 @@ ENTRY(call_rwsem_wake)
 	CFI_STARTPROC
 	decw %dx    /* do nothing if still outstanding active readers */
 	jnz 1f
-	push %ecx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ecx
 	CFI_REL_OFFSET ecx,0
 	call rwsem_wake
-	pop %ecx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %ecx
 1:	ret
 	CFI_ENDPROC
 	ENDPROC(call_rwsem_wake)
@@ -118,17 +110,13 @@ ENTRY(call_rwsem_wake)
 /* Fix up special calling conventions */
 ENTRY(call_rwsem_downgrade_wake)
 	CFI_STARTPROC
-	push %ecx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %ecx
 	CFI_REL_OFFSET ecx,0
-	push %edx
-	CFI_ADJUST_CFA_OFFSET 4
+	pushl_cfi %edx
 	CFI_REL_OFFSET edx,0
 	call rwsem_downgrade_wake
-	pop %edx
-	CFI_ADJUST_CFA_OFFSET -4
-	pop %ecx
-	CFI_ADJUST_CFA_OFFSET -4
+	popl_cfi %edx
+	popl_cfi %ecx
 	ret
 	CFI_ENDPROC
 	ENDPROC(call_rwsem_downgrade_wake)
diff --git a/arch/x86/lib/thunk_32.S b/arch/x86/lib/thunk_32.S
index 650b11e00ec..2930ae05d77 100644
--- a/arch/x86/lib/thunk_32.S
+++ b/arch/x86/lib/thunk_32.S
@@ -7,24 +7,6 @@
 
 	#include <linux/linkage.h>
 
-#define ARCH_TRACE_IRQS_ON			\
-	pushl %eax;				\
-	pushl %ecx;				\
-	pushl %edx;				\
-	call trace_hardirqs_on;			\
-	popl %edx;				\
-	popl %ecx;				\
-	popl %eax;
-
-#define ARCH_TRACE_IRQS_OFF			\
-	pushl %eax;				\
-	pushl %ecx;				\
-	pushl %edx;				\
-	call trace_hardirqs_off;		\
-	popl %edx;				\
-	popl %ecx;				\
-	popl %eax;
-
 #ifdef CONFIG_TRACE_IRQFLAGS
 	/* put return address in eax (arg1) */
 	.macro thunk_ra name,func
diff --git a/arch/x86/lib/thunk_64.S b/arch/x86/lib/thunk_64.S
index bf9a7d5a542..782b082c9ff 100644
--- a/arch/x86/lib/thunk_64.S
+++ b/arch/x86/lib/thunk_64.S
@@ -22,26 +22,6 @@
 	CFI_ENDPROC
 	.endm
 
-	/* rdi:	arg1 ... normal C conventions. rax is passed from C. */ 	
-	.macro thunk_retrax name,func
-	.globl \name
-\name:	
-	CFI_STARTPROC
-	SAVE_ARGS
-	call \func
-	jmp  restore_norax
-	CFI_ENDPROC
-	.endm
-	
-
-	.section .sched.text, "ax"
-#ifdef CONFIG_RWSEM_XCHGADD_ALGORITHM
-	thunk rwsem_down_read_failed_thunk,rwsem_down_read_failed
-	thunk rwsem_down_write_failed_thunk,rwsem_down_write_failed
-	thunk rwsem_wake_thunk,rwsem_wake
-	thunk rwsem_downgrade_thunk,rwsem_downgrade_wake
-#endif	
-	
 #ifdef CONFIG_TRACE_IRQFLAGS
 	/* put return address in rdi (arg1) */
 	.macro thunk_ra name,func
@@ -72,10 +52,3 @@ restore:
 	RESTORE_ARGS
 	ret	
 	CFI_ENDPROC
-	
-	CFI_STARTPROC
-	SAVE_ARGS
-restore_norax:	
-	RESTORE_ARGS 1
-	ret
-	CFI_ENDPROC