platform_bionic/libc/arch-x86/string/sse2-wcslen-atom.S
Liubov Dmitrieva 0a490665a3 bionic/x86: Optimization for string routines
Optimized strcpy, strcat,
strncpy, strncat, strlcpy, strlcat,
memchr, memrchr, strchr, strrchr, index,
strnlen, strlen, wcslen, wmemcmp, wcscmp,
wcschr, wcsrchr, wcscpy, wcscat

Change-Id: I82b29132edf9a2e144e0bb3ee4ff5217df8d2a6d
Signed-off-by: Liubov Dmitrieva <liubov.dmitrieva@intel.com>
2013-05-31 13:37:03 +04:00

306 lines
5.4 KiB
ArmAsm

/*
Copyright (c) 2011 Intel Corporation
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
* Neither the name of Intel Corporation nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#ifndef USE_AS_WCSCAT
# ifndef L
# define L(label) .L##label
# endif
# ifndef cfi_startproc
# define cfi_startproc .cfi_startproc
# endif
# ifndef cfi_endproc
# define cfi_endproc .cfi_endproc
# endif
# ifndef ENTRY
# define ENTRY(name) \
.type name, @function; \
.globl name; \
.p2align 4; \
name: \
cfi_startproc
# endif
# ifndef END
# define END(name) \
cfi_endproc; \
.size name, .-name
# endif
# define PARMS 4
# define STR PARMS
# define RETURN ret
.text
ENTRY (wcslen)
mov STR(%esp), %edx
#endif
cmp $0, (%edx)
jz L(exit_tail0)
cmp $0, 4(%edx)
jz L(exit_tail1)
cmp $0, 8(%edx)
jz L(exit_tail2)
cmp $0, 12(%edx)
jz L(exit_tail3)
cmp $0, 16(%edx)
jz L(exit_tail4)
cmp $0, 20(%edx)
jz L(exit_tail5)
cmp $0, 24(%edx)
jz L(exit_tail6)
cmp $0, 28(%edx)
jz L(exit_tail7)
pxor %xmm0, %xmm0
lea 32(%edx), %eax
lea -16(%eax), %ecx
and $-16, %eax
pcmpeqd (%eax), %xmm0
pmovmskb %xmm0, %edx
pxor %xmm1, %xmm1
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm1
pmovmskb %xmm1, %edx
pxor %xmm2, %xmm2
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm2
pmovmskb %xmm2, %edx
pxor %xmm3, %xmm3
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm3
pmovmskb %xmm3, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm0
pmovmskb %xmm0, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm1
pmovmskb %xmm1, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm2
pmovmskb %xmm2, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm3
pmovmskb %xmm3, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm0
pmovmskb %xmm0, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm1
pmovmskb %xmm1, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm2
pmovmskb %xmm2, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm3
pmovmskb %xmm3, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm0
pmovmskb %xmm0, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm1
pmovmskb %xmm1, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm2
pmovmskb %xmm2, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
pcmpeqd (%eax), %xmm3
pmovmskb %xmm3, %edx
lea 16(%eax), %eax
test %edx, %edx
jnz L(exit)
and $-0x40, %eax
.p2align 4
L(aligned_64_loop):
movaps (%eax), %xmm0
movaps 16(%eax), %xmm1
movaps 32(%eax), %xmm2
movaps 48(%eax), %xmm6
pminub %xmm1, %xmm0
pminub %xmm6, %xmm2
pminub %xmm0, %xmm2
pcmpeqd %xmm3, %xmm2
pmovmskb %xmm2, %edx
lea 64(%eax), %eax
test %edx, %edx
jz L(aligned_64_loop)
pcmpeqd -64(%eax), %xmm3
pmovmskb %xmm3, %edx
lea 48(%ecx), %ecx
test %edx, %edx
jnz L(exit)
pcmpeqd %xmm1, %xmm3
pmovmskb %xmm3, %edx
lea -16(%ecx), %ecx
test %edx, %edx
jnz L(exit)
pcmpeqd -32(%eax), %xmm3
pmovmskb %xmm3, %edx
lea -16(%ecx), %ecx
test %edx, %edx
jnz L(exit)
pcmpeqd %xmm6, %xmm3
pmovmskb %xmm3, %edx
lea -16(%ecx), %ecx
test %edx, %edx
jnz L(exit)
jmp L(aligned_64_loop)
.p2align 4
L(exit):
sub %ecx, %eax
shr $2, %eax
test %dl, %dl
jz L(exit_high)
mov %dl, %cl
and $15, %cl
jz L(exit_1)
RETURN
.p2align 4
L(exit_high):
mov %dh, %ch
and $15, %ch
jz L(exit_3)
add $2, %eax
RETURN
.p2align 4
L(exit_1):
add $1, %eax
RETURN
.p2align 4
L(exit_3):
add $3, %eax
RETURN
.p2align 4
L(exit_tail0):
xor %eax, %eax
RETURN
.p2align 4
L(exit_tail1):
mov $1, %eax
RETURN
.p2align 4
L(exit_tail2):
mov $2, %eax
RETURN
.p2align 4
L(exit_tail3):
mov $3, %eax
RETURN
.p2align 4
L(exit_tail4):
mov $4, %eax
RETURN
.p2align 4
L(exit_tail5):
mov $5, %eax
RETURN
.p2align 4
L(exit_tail6):
mov $6, %eax
RETURN
.p2align 4
L(exit_tail7):
mov $7, %eax
#ifndef USE_AS_WCSCAT
RETURN
END (wcslen)
#endif