kernel_optimize_test/arch/x86/lib/memset_64.S
H. Peter Anvin 83a7a2ad2a x86, alternatives: Use 16-bit numbers for cpufeature index
We already have cpufeature indicies above 255, so use a 16-bit number
for the alternatives index.  This consumes a padding field and so
doesn't add any size, but it means that abusing the padding field to
create assembly errors on overflow no longer works.  We can retain the
test simply by redirecting it to the .discard section, however.

[ v3: updated to include open-coded locations ]

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
LKML-Reference: <tip-f88731e3068f9d1392ba71cc9f50f035d26a0d4f@git.kernel.org>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
2010-07-07 10:36:28 -07:00

128 lines
2.2 KiB
ArmAsm

/* Copyright 2002 Andi Kleen, SuSE Labs */
#include <linux/linkage.h>
#include <asm/dwarf2.h>
/*
* ISO C memset - set a memory block to a byte value.
*
* rdi destination
* rsi value (char)
* rdx count (bytes)
*
* rax original destination
*/
.section .altinstr_replacement, "ax", @progbits
.Lmemset_c:
movq %rdi,%r9
movl %edx,%r8d
andl $7,%r8d
movl %edx,%ecx
shrl $3,%ecx
/* expand byte value */
movzbl %sil,%esi
movabs $0x0101010101010101,%rax
mulq %rsi /* with rax, clobbers rdx */
rep stosq
movl %r8d,%ecx
rep stosb
movq %r9,%rax
ret
.Lmemset_e:
.previous
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 */
/* align dst */
movl %edi,%r9d
andl $7,%r9d
jnz .Lbad_alignment
CFI_REMEMBER_STATE
.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
CFI_RESTORE_STATE
.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
.Lfinal:
CFI_ENDPROC
ENDPROC(memset)
ENDPROC(__memset)
/* Some CPUs 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 .Lmemset_c
.word X86_FEATURE_REP_GOOD
.byte .Lfinal - memset
.byte .Lmemset_e - .Lmemset_c
.previous