[PATCH] x86_64: Remove optimization for B stepping AMD K8

B stepping were the first shipping Opterons. memcpy/memset/copy_page/
clear_page had special optimized version for them. These are really
old and in the minority now and the difference to the generic versions
(using rep microcode) is not that big anyways. So just remove them.

TODO: figure out optimized versions for Intel Netburst based EM64T

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

4 files changed, 2 insertions, 310 deletions

diff --git a/arch/x86_64/lib/clear_page.S b/arch/x86_64/lib/clear_page.S
index 30a9da4..43d9fa1 100644
--- a/arch/x86_64/lib/clear_page.S
+++ b/arch/x86_64/lib/clear_page.S
@@ -5,46 +5,8 @@
 	.globl clear_page
 	.p2align 4
 clear_page:
-	xorl   %eax,%eax
-	movl   $4096/64,%ecx
-	.p2align 4
-.Lloop:
-	decl	%ecx
-#define PUT(x) movq %rax,x*8(%rdi) 
-	movq %rax,(%rdi)
-	PUT(1)
-	PUT(2)
-	PUT(3)
-	PUT(4)
-	PUT(5)
-	PUT(6)
-	PUT(7)
-	leaq	64(%rdi),%rdi
-	jnz	.Lloop
-	nop
-	ret
-clear_page_end:	
-	
-	/* C stepping K8 run faster using the string instructions.
-	   It is also a lot simpler. Use this when possible */
-	
-#include <asm/cpufeature.h>
-	    	
-	.section .altinstructions,"a"
-	.align 8
-	.quad  clear_page
-	.quad  clear_page_c
-	.byte  X86_FEATURE_K8_C
-	.byte  clear_page_end-clear_page	
-	.byte  clear_page_c_end-clear_page_c
-	.previous
-
-	.section .altinstr_replacement,"ax"
-clear_page_c:
 	movl $4096/8,%ecx
 	xorl %eax,%eax
 	rep 
 	stosq
 	ret
-clear_page_c_end:
-	.previous
diff --git a/arch/x86_64/lib/copy_page.S b/arch/x86_64/lib/copy_page.S
index dd3aa47..621a197 100644
--- a/arch/x86_64/lib/copy_page.S
+++ b/arch/x86_64/lib/copy_page.S
@@ -8,94 +8,7 @@
 	.globl copy_page
 	.p2align 4
 copy_page:
-	subq	$3*8,%rsp
-	movq	%rbx,(%rsp)
-	movq	%r12,1*8(%rsp)
-	movq	%r13,2*8(%rsp)
-			
-	movl	$(4096/64)-5,%ecx
-	.p2align 4
-.Loop64:	
-  	dec     %rcx
-
-	movq        (%rsi), %rax
-	movq      8 (%rsi), %rbx
-	movq     16 (%rsi), %rdx
-	movq     24 (%rsi), %r8
-	movq     32 (%rsi), %r9
-	movq     40 (%rsi), %r10
-	movq     48 (%rsi), %r11
-	movq     56 (%rsi), %r12
-
-	prefetcht0 5*64(%rsi)
-
-	movq     %rax,    (%rdi)
-	movq     %rbx,  8 (%rdi)
-	movq     %rdx, 16 (%rdi)
-	movq     %r8,  24 (%rdi)
-	movq     %r9,  32 (%rdi)
-	movq     %r10, 40 (%rdi)
-	movq     %r11, 48 (%rdi)
-	movq     %r12, 56 (%rdi)
-
-	leaq    64 (%rsi), %rsi
-	leaq    64 (%rdi), %rdi
-
-	jnz     .Loop64
-
-	movl	$5,%ecx
-	.p2align 4
-.Loop2:	
-	decl   %ecx
-
-	movq        (%rsi), %rax
-	movq      8 (%rsi), %rbx
-	movq     16 (%rsi), %rdx
-	movq     24 (%rsi), %r8
-	movq     32 (%rsi), %r9
-	movq     40 (%rsi), %r10
-	movq     48 (%rsi), %r11
-	movq     56 (%rsi), %r12
-
-	movq     %rax,    (%rdi)
-	movq     %rbx,  8 (%rdi)
-	movq     %rdx, 16 (%rdi)
-	movq     %r8,  24 (%rdi)
-	movq     %r9,  32 (%rdi)
-	movq     %r10, 40 (%rdi)
-	movq     %r11, 48 (%rdi)
-	movq     %r12, 56 (%rdi)
-	
-	leaq	64(%rdi),%rdi			
-	leaq	64(%rsi),%rsi			
-	
-	jnz	.Loop2		
-	
-	movq	(%rsp),%rbx
-	movq	1*8(%rsp),%r12
-	movq	2*8(%rsp),%r13
-	addq	$3*8,%rsp
-	ret
-	
-	/* C stepping K8 run faster using the string copy instructions.
-	   It is also a lot simpler. Use this when possible */
-
-#include <asm/cpufeature.h>		
-		
-	.section .altinstructions,"a"
-	.align 8
-	.quad  copy_page
-	.quad  copy_page_c
-	.byte  X86_FEATURE_K8_C
-	.byte  copy_page_c_end-copy_page_c
-	.byte  copy_page_c_end-copy_page_c
-	.previous
-
-	.section .altinstr_replacement,"ax"
-copy_page_c:
 	movl $4096/8,%ecx
 	rep 
 	movsq 
 	ret
-copy_page_c_end:
-	.previous
diff --git a/arch/x86_64/lib/memcpy.S b/arch/x86_64/lib/memcpy.S
index c6c4649..92dd805 100644
--- a/arch/x86_64/lib/memcpy.S
+++ b/arch/x86_64/lib/memcpy.S
@@ -11,6 +11,8 @@
  * 
  * Output:
  * rax original destination
+ * 
+ * TODO: check best memcpy for PSC
  */	
 
  	.globl __memcpy
@@ -18,95 +20,6 @@
 	.p2align 4
 __memcpy:
 memcpy:		
-	pushq %rbx
-	movq %rdi,%rax
-
-	movl %edx,%ecx
-	shrl $6,%ecx
-	jz .Lhandle_tail
-	
-	.p2align 4
-.Lloop_64:
-	decl %ecx
-	
-	movq (%rsi),%r11
-	movq 8(%rsi),%r8
-
-	movq %r11,(%rdi)
-	movq %r8,1*8(%rdi)
-
-	movq 2*8(%rsi),%r9
-	movq 3*8(%rsi),%r10
-
-	movq %r9,2*8(%rdi)
-	movq %r10,3*8(%rdi)
-		
-	movq 4*8(%rsi),%r11
-	movq 5*8(%rsi),%r8
-
-	movq %r11,4*8(%rdi)
-	movq %r8,5*8(%rdi)
-
-	movq 6*8(%rsi),%r9
-	movq 7*8(%rsi),%r10
-
-	movq %r9,6*8(%rdi)
-	movq %r10,7*8(%rdi)
-
-	leaq 64(%rsi),%rsi
-	leaq 64(%rdi),%rdi
-	jnz  .Lloop_64
-
-.Lhandle_tail:
-	movl %edx,%ecx
-	andl $63,%ecx
-	shrl $3,%ecx
-	jz   .Lhandle_7
-	.p2align 4
-.Lloop_8: 
-	decl %ecx
-	movq (%rsi),%r8
-	movq %r8,(%rdi) 
-	leaq 8(%rdi),%rdi
-	leaq 8(%rsi),%rsi
-	jnz  .Lloop_8
-
-.Lhandle_7:
-	movl %edx,%ecx
-	andl $7,%ecx
-	jz .Lende
-	.p2align 4
-.Lloop_1:
-	movb (%rsi),%r8b
-	movb %r8b,(%rdi) 
-	incq %rdi
-	incq %rsi
-	decl %ecx
-	jnz .Lloop_1
-	
-.Lende: 	
-	popq %rbx
-	ret
-.Lfinal:
-	
-	/* C stepping K8 run faster using the string copy instructions.
-	   It is also a lot simpler. Use this when possible */
-	
-	.section .altinstructions,"a"
-	.align 8
-	.quad  memcpy
-	.quad  memcpy_c
-	.byte  X86_FEATURE_K8_C
-	.byte  .Lfinal-memcpy
-	.byte  memcpy_c_end-memcpy_c	
-	.previous
-
-	.section .altinstr_replacement,"ax"
- /* rdi	destination
-  * rsi source
-  * rdx count
-  */			
-memcpy_c:
 	movq %rdi,%rax
 	movl %edx,%ecx
 	shrl $3,%ecx
@@ -117,5 +30,3 @@ memcpy_c:
 	rep
 	movsb
 	ret
-memcpy_c_end:
-	.previous
diff --git a/arch/x86_64/lib/memset.S b/arch/x86_64/lib/memset.S
index 4b4c406..2aa48f2 100644
--- a/arch/x86_64/lib/memset.S
+++ b/arch/x86_64/lib/memset.S
@@ -13,98 +13,6 @@
 	.p2align 4
 memset:	
 __memset:
-	movq %rdi,%r10
-	movq %rdx,%r11
-
-	/* expand byte value  */
-	movzbl %sil,%ecx
-	movabs $0x0101010101010101,%rax
-	mul    %rcx		/* with rax, clobbers rdx */
-
-	/* align dst */
-	movl  %edi,%r9d		
-	andl  $7,%r9d	
-	jnz  .Lbad_alignment
-.Lafter_bad_alignment:
-	
-	movl %r11d,%ecx
-	shrl $6,%ecx
-	jz	 .Lhandle_tail
-
-	.p2align 4
-.Lloop_64:	
-	decl   %ecx
-	movq  %rax,(%rdi) 
-	movq  %rax,8(%rdi) 
-	movq  %rax,16(%rdi) 
-	movq  %rax,24(%rdi) 
-	movq  %rax,32(%rdi) 
-	movq  %rax,40(%rdi) 
-	movq  %rax,48(%rdi) 
-	movq  %rax,56(%rdi) 
-	leaq  64(%rdi),%rdi
-	jnz    .Lloop_64
-
-	/* Handle tail in loops. The loops should be faster than hard
-	   to predict jump tables. */ 
-	.p2align 4	   
-.Lhandle_tail:
-	movl	%r11d,%ecx
-	andl    $63&(~7),%ecx
-	jz 		.Lhandle_7
-	shrl	$3,%ecx
-	.p2align 4
-.Lloop_8:
-	decl   %ecx
-	movq  %rax,(%rdi)
-	leaq  8(%rdi),%rdi
-	jnz    .Lloop_8
-
-.Lhandle_7:
-	movl	%r11d,%ecx
-	andl	$7,%ecx
-	jz      .Lende
-	.p2align 4
-.Lloop_1:
-	decl    %ecx
-	movb 	%al,(%rdi)
-	leaq	1(%rdi),%rdi
-	jnz     .Lloop_1
-	
-.Lende:	
-	movq	%r10,%rax
-	ret
-
-.Lbad_alignment:
-	cmpq $7,%r11
-	jbe	.Lhandle_7
-	movq %rax,(%rdi)	/* unaligned store */
-	movq $8,%r8			
-	subq %r9,%r8 
-	addq %r8,%rdi
-	subq %r8,%r11
-	jmp .Lafter_bad_alignment
-
-	/* C stepping K8 run faster using the string instructions.
-	   It is also a lot simpler. Use this when possible */
-
-#include <asm/cpufeature.h>	
-		
-	.section .altinstructions,"a"
-	.align 8
-	.quad  memset
-	.quad  memset_c
-	.byte  X86_FEATURE_K8_C
-	.byte  memset_c_end-memset_c
-	.byte  memset_c_end-memset_c
-	.previous
-
-	.section .altinstr_replacement,"ax"
- /* rdi	destination
-  * rsi value
-  * rdx count
-  */			
-memset_c:	
 	movq %rdi,%r9
 	movl %edx,%r8d
 	andl $7,%r8d		
@@ -121,5 +29,3 @@ memset_c:
 	stosb
 	movq %r9,%rax
 	ret
-memset_c_end:
-	.previous