mirror of
https://sourceware.org/git/glibc.git
synced 2025-01-08 10:20:15 +00:00
e74d959862
Based on the glibc microbenchmark, strnlen-aligned implementation could reduce the runtime more than 10%, strnlen-lsx implementation could reduce the runtime about 50%-78%, strnlen-lasx implementation could reduce the runtime about 50%-88%.
103 lines
2.4 KiB
ArmAsm
103 lines
2.4 KiB
ArmAsm
/* Optimized strnlen implementation using basic Loongarch instructions.
|
|
Copyright (C) 2023 Free Software Foundation, Inc.
|
|
|
|
This file is part of the GNU C Library.
|
|
|
|
The GNU C Library is free software; you can redistribute it and/or
|
|
modify it under the terms of the GNU Lesser General Public
|
|
License as published by the Free Software Foundation; either
|
|
version 2.1 of the License, or (at your option) any later version.
|
|
|
|
The GNU C Library is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
Lesser General Public License for more details.
|
|
|
|
You should have received a copy of the GNU Lesser General Public
|
|
License along with the GNU C Library. If not, see
|
|
<https://www.gnu.org/licenses/>. */
|
|
|
|
#include <sysdep.h>
|
|
#include <sys/regdef.h>
|
|
#include <sys/asm.h>
|
|
|
|
#if IS_IN (libc)
|
|
# define STRNLEN __strnlen_aligned
|
|
#else
|
|
# define STRNLEN __strnlen
|
|
#endif
|
|
|
|
LEAF(STRNLEN, 6)
|
|
beqz a1, L(out)
|
|
lu12i.w a2, 0x01010
|
|
andi t1, a0, 0x7
|
|
move t4, a0
|
|
|
|
bstrins.d a0, zero, 2, 0
|
|
ori a2, a2, 0x101
|
|
li.w t0, -1
|
|
ld.d t2, a0, 0
|
|
|
|
slli.d t3, t1, 3
|
|
bstrins.d a2, a2, 63, 32
|
|
li.w t5, 8
|
|
slli.d a3, a2, 7
|
|
|
|
sub.w t1, t5, t1
|
|
sll.d t0, t0, t3
|
|
orn t2, t2, t0
|
|
sub.d t0, t2, a2
|
|
|
|
|
|
andn t3, a3, t2
|
|
and t0, t0, t3
|
|
bnez t0, L(count_pos)
|
|
sub.d t5, a1, t1
|
|
|
|
bgeu t1, a1, L(out)
|
|
addi.d a0, a0, 8
|
|
L(loop):
|
|
ld.d t2, a0, 0
|
|
sub.d t0, t2, a2
|
|
|
|
andn t1, a3, t2
|
|
sltui t6, t5, 9
|
|
and t0, t0, t1
|
|
or t7, t0, t6
|
|
|
|
bnez t7, L(count_pos)
|
|
ld.d t2, a0, 8
|
|
addi.d a0, a0, 16
|
|
sub.d t0, t2, a2
|
|
|
|
|
|
andn t1, a3, t2
|
|
sltui t6, t5, 17
|
|
and t0, t0, t1
|
|
addi.d t5, t5, -16
|
|
|
|
or t7, t0, t6
|
|
beqz t7, L(loop)
|
|
addi.d a0, a0, -8
|
|
L(count_pos):
|
|
ctz.d t1, t0
|
|
|
|
sub.d a0, a0, t4
|
|
srli.d t1, t1, 3
|
|
add.d a0, t1, a0
|
|
sltu t0, a0, a1
|
|
|
|
masknez t1, a1, t0
|
|
maskeqz a0, a0, t0
|
|
or a0, a0, t1
|
|
jr ra
|
|
|
|
|
|
L(out):
|
|
move a0, a1
|
|
jr ra
|
|
END(STRNLEN)
|
|
|
|
weak_alias (STRNLEN, strnlen)
|
|
libc_hidden_builtin_def (STRNLEN)
|