Merge branch 'x86-asm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86/asm changes from Ingo Molnar

* 'x86-asm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86: Include probe_roms.h in probe_roms.c
  x86/32: Print control and debug registers for kerenel context
  x86: Tighten dependencies of CPU_SUP_*_32
  x86/numa: Improve internode cache alignment
  x86: Fix the NMI nesting comments
  x86-64: Improve insn scheduling in SAVE_ARGS_IRQ
  x86-64: Fix CFI annotations for NMI nesting code
  bitops: Add missing parentheses to new get_order macro
  bitops: Optimise get_order()
  bitops: Adjust the comment on get_order() to describe the size==0 case
  x86/spinlocks: Eliminate TICKET_MASK
  x86-64: Handle byte-wise tail copying in memcpy() without a loop
  x86-64: Fix memcpy() to support sizes of 4Gb and above
  x86-64: Fix memset() to support sizes of 4Gb and above
  x86-64: Slightly shorten copy_page()

3 files changed, 39 insertions, 50 deletions

diff --git a/arch/x86/lib/copy_page_64.S b/arch/x86/lib/copy_page_64.S
index 01c805b..6b34d04 100644
--- a/arch/x86/lib/copy_page_64.S
+++ b/arch/x86/lib/copy_page_64.S
@@ -20,14 +20,12 @@ ENDPROC(copy_page_c)
 
 ENTRY(copy_page)
 	CFI_STARTPROC
-	subq	$3*8,%rsp
-	CFI_ADJUST_CFA_OFFSET 3*8
+	subq	$2*8,%rsp
+	CFI_ADJUST_CFA_OFFSET 2*8
 	movq	%rbx,(%rsp)
 	CFI_REL_OFFSET rbx, 0
 	movq	%r12,1*8(%rsp)
 	CFI_REL_OFFSET r12, 1*8
-	movq	%r13,2*8(%rsp)
-	CFI_REL_OFFSET r13, 2*8
 
 	movl	$(4096/64)-5,%ecx
 	.p2align 4
@@ -91,10 +89,8 @@ ENTRY(copy_page)
 	CFI_RESTORE rbx
 	movq	1*8(%rsp),%r12
 	CFI_RESTORE r12
-	movq	2*8(%rsp),%r13
-	CFI_RESTORE r13
-	addq	$3*8,%rsp
-	CFI_ADJUST_CFA_OFFSET -3*8
+	addq	$2*8,%rsp
+	CFI_ADJUST_CFA_OFFSET -2*8
 	ret
 .Lcopy_page_end:
 	CFI_ENDPROC
diff --git a/arch/x86/lib/memcpy_64.S b/arch/x86/lib/memcpy_64.S
index efbf2a0..1c273be 100644
--- a/arch/x86/lib/memcpy_64.S
+++ b/arch/x86/lib/memcpy_64.S
@@ -27,9 +27,8 @@
 	.section .altinstr_replacement, "ax", @progbits
 .Lmemcpy_c:
 	movq %rdi, %rax
-
-	movl %edx, %ecx
-	shrl $3, %ecx
+	movq %rdx, %rcx
+	shrq $3, %rcx
 	andl $7, %edx
 	rep movsq
 	movl %edx, %ecx
@@ -48,8 +47,7 @@
 	.section .altinstr_replacement, "ax", @progbits
 .Lmemcpy_c_e:
 	movq %rdi, %rax
-
-	movl %edx, %ecx
+	movq %rdx, %rcx
 	rep movsb
 	ret
 .Lmemcpy_e_e:
@@ -60,10 +58,7 @@ ENTRY(memcpy)
 	CFI_STARTPROC
 	movq %rdi, %rax
 
-	/*
-	 * Use 32bit CMP here to avoid long NOP padding.
-	 */
-	cmp  $0x20, %edx
+	cmpq $0x20, %rdx
 	jb .Lhandle_tail
 
 	/*
@@ -72,7 +67,7 @@ ENTRY(memcpy)
 	 */
 	cmp  %dil, %sil
 	jl .Lcopy_backward
-	subl $0x20, %edx
+	subq $0x20, %rdx
 .Lcopy_forward_loop:
 	subq $0x20,	%rdx
 
@@ -91,7 +86,7 @@ ENTRY(memcpy)
 	movq %r11,	3*8(%rdi)
 	leaq 4*8(%rdi),	%rdi
 	jae  .Lcopy_forward_loop
-	addq $0x20,	%rdx
+	addl $0x20,	%edx
 	jmp  .Lhandle_tail
 
 .Lcopy_backward:
@@ -123,11 +118,11 @@ ENTRY(memcpy)
 	/*
 	 * Calculate copy position to head.
 	 */
-	addq $0x20,	%rdx
+	addl $0x20,	%edx
 	subq %rdx,	%rsi
 	subq %rdx,	%rdi
 .Lhandle_tail:
-	cmpq $16,	%rdx
+	cmpl $16,	%edx
 	jb   .Lless_16bytes
 
 	/*
@@ -144,7 +139,7 @@ ENTRY(memcpy)
 	retq
 	.p2align 4
 .Lless_16bytes:
-	cmpq $8,	%rdx
+	cmpl $8,	%edx
 	jb   .Lless_8bytes
 	/*
 	 * Move data from 8 bytes to 15 bytes.
@@ -156,7 +151,7 @@ ENTRY(memcpy)
 	retq
 	.p2align 4
 .Lless_8bytes:
-	cmpq $4,	%rdx
+	cmpl $4,	%edx
 	jb   .Lless_3bytes
 
 	/*
@@ -169,18 +164,19 @@ ENTRY(memcpy)
 	retq
 	.p2align 4
 .Lless_3bytes:
-	cmpl $0, %edx
-	je .Lend
+	subl $1, %edx
+	jb .Lend
 	/*
 	 * Move data from 1 bytes to 3 bytes.
 	 */
-.Lloop_1:
-	movb (%rsi), %r8b
-	movb %r8b, (%rdi)
-	incq %rdi
-	incq %rsi
-	decl %edx
-	jnz .Lloop_1
+	movzbl (%rsi), %ecx
+	jz .Lstore_1byte
+	movzbq 1(%rsi), %r8
+	movzbq (%rsi, %rdx), %r9
+	movb %r8b, 1(%rdi)
+	movb %r9b, (%rdi, %rdx)
+.Lstore_1byte:
+	movb %cl, (%rdi)
 
 .Lend:
 	retq
diff --git a/arch/x86/lib/memset_64.S b/arch/x86/lib/memset_64.S
index 79bd454..2dcb380 100644
--- a/arch/x86/lib/memset_64.S
+++ b/arch/x86/lib/memset_64.S
@@ -19,16 +19,15 @@
 	.section .altinstr_replacement, "ax", @progbits
 .Lmemset_c:
 	movq %rdi,%r9
-	movl %edx,%r8d
-	andl $7,%r8d
-	movl %edx,%ecx
-	shrl $3,%ecx
+	movq %rdx,%rcx
+	andl $7,%edx
+	shrq $3,%rcx
 	/* expand byte value  */
 	movzbl %sil,%esi
 	movabs $0x0101010101010101,%rax
-	mulq %rsi		/* with rax, clobbers rdx */
+	imulq %rsi,%rax
 	rep stosq
-	movl %r8d,%ecx
+	movl %edx,%ecx
 	rep stosb
 	movq %r9,%rax
 	ret
@@ -50,7 +49,7 @@
 .Lmemset_c_e:
 	movq %rdi,%r9
 	movb %sil,%al
-	movl %edx,%ecx
+	movq %rdx,%rcx
 	rep stosb
 	movq %r9,%rax
 	ret
@@ -61,12 +60,11 @@ ENTRY(memset)
 ENTRY(__memset)
 	CFI_STARTPROC
 	movq %rdi,%r10
-	movq %rdx,%r11
 
 	/* expand byte value  */
 	movzbl %sil,%ecx
 	movabs $0x0101010101010101,%rax
-	mul    %rcx		/* with rax, clobbers rdx */
+	imulq  %rcx,%rax
 
 	/* align dst */
 	movl  %edi,%r9d
@@ -75,13 +73,13 @@ ENTRY(__memset)
 	CFI_REMEMBER_STATE
 .Lafter_bad_alignment:
 
-	movl %r11d,%ecx
-	shrl $6,%ecx
+	movq  %rdx,%rcx
+	shrq  $6,%rcx
 	jz	 .Lhandle_tail
 
 	.p2align 4
 .Lloop_64:
-	decl   %ecx
+	decq  %rcx
 	movq  %rax,(%rdi)
 	movq  %rax,8(%rdi)
 	movq  %rax,16(%rdi)
@@ -97,7 +95,7 @@ ENTRY(__memset)
 	   to predict jump tables. */
 	.p2align 4
 .Lhandle_tail:
-	movl	%r11d,%ecx
+	movl	%edx,%ecx
 	andl    $63&(~7),%ecx
 	jz 		.Lhandle_7
 	shrl	$3,%ecx
@@ -109,12 +107,11 @@ ENTRY(__memset)
 	jnz    .Lloop_8
 
 .Lhandle_7:
-	movl	%r11d,%ecx
-	andl	$7,%ecx
+	andl	$7,%edx
 	jz      .Lende
 	.p2align 4
 .Lloop_1:
-	decl    %ecx
+	decl    %edx
 	movb 	%al,(%rdi)
 	leaq	1(%rdi),%rdi
 	jnz     .Lloop_1
@@ -125,13 +122,13 @@ ENTRY(__memset)
 
 	CFI_RESTORE_STATE
 .Lbad_alignment:
-	cmpq $7,%r11
+	cmpq $7,%rdx
 	jbe	.Lhandle_7
 	movq %rax,(%rdi)	/* unaligned store */
 	movq $8,%r8
 	subq %r9,%r8
 	addq %r8,%rdi
-	subq %r8,%r11
+	subq %r8,%rdx
 	jmp .Lafter_bad_alignment
 .Lfinal:
 	CFI_ENDPROC