2011-01-12 18:59:09 +00:00
|
|
|
/* Optimized memcpy implementation for PowerPC476.
|
2024-01-01 18:12:26 +00:00
|
|
|
Copyright (C) 2010-2024 Free Software Foundation, Inc.
|
2011-01-12 18:59:09 +00:00
|
|
|
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
|
2012-03-09 23:56:38 +00:00
|
|
|
License along with the GNU C Library. If not, see
|
Prefer https to http for gnu.org and fsf.org URLs
Also, change sources.redhat.com to sourceware.org.
This patch was automatically generated by running the following shell
script, which uses GNU sed, and which avoids modifying files imported
from upstream:
sed -ri '
s,(http|ftp)(://(.*\.)?(gnu|fsf|sourceware)\.org($|[^.]|\.[^a-z])),https\2,g
s,(http|ftp)(://(.*\.)?)sources\.redhat\.com($|[^.]|\.[^a-z]),https\2sourceware.org\4,g
' \
$(find $(git ls-files) -prune -type f \
! -name '*.po' \
! -name 'ChangeLog*' \
! -path COPYING ! -path COPYING.LIB \
! -path manual/fdl-1.3.texi ! -path manual/lgpl-2.1.texi \
! -path manual/texinfo.tex ! -path scripts/config.guess \
! -path scripts/config.sub ! -path scripts/install-sh \
! -path scripts/mkinstalldirs ! -path scripts/move-if-change \
! -path INSTALL ! -path locale/programs/charmap-kw.h \
! -path po/libc.pot ! -path sysdeps/gnu/errlist.c \
! '(' -name configure \
-execdir test -f configure.ac -o -f configure.in ';' ')' \
! '(' -name preconfigure \
-execdir test -f preconfigure.ac ';' ')' \
-print)
and then by running 'make dist-prepare' to regenerate files built
from the altered files, and then executing the following to cleanup:
chmod a+x sysdeps/unix/sysv/linux/riscv/configure
# Omit irrelevant whitespace and comment-only changes,
# perhaps from a slightly-different Autoconf version.
git checkout -f \
sysdeps/csky/configure \
sysdeps/hppa/configure \
sysdeps/riscv/configure \
sysdeps/unix/sysv/linux/csky/configure
# Omit changes that caused a pre-commit check to fail like this:
# remote: *** error: sysdeps/powerpc/powerpc64/ppc-mcount.S: trailing lines
git checkout -f \
sysdeps/powerpc/powerpc64/ppc-mcount.S \
sysdeps/unix/sysv/linux/s390/s390-64/syscall.S
# Omit change that caused a pre-commit check to fail like this:
# remote: *** error: sysdeps/sparc/sparc64/multiarch/memcpy-ultra3.S: last line does not end in newline
git checkout -f sysdeps/sparc/sparc64/multiarch/memcpy-ultra3.S
2019-09-07 05:40:42 +00:00
|
|
|
<https://www.gnu.org/licenses/>. */
|
2011-01-12 18:59:09 +00:00
|
|
|
|
|
|
|
|
#include <sysdep.h>
|
|
|
|
|
|
|
|
|
|
/* memcpy
|
|
|
|
|
|
|
|
|
|
r0:return address
|
|
|
|
|
r3:destination address
|
|
|
|
|
r4:source address
|
|
|
|
|
r5:byte count
|
|
|
|
|
|
|
|
|
|
Save return address in r0.
|
2023-05-23 12:11:30 +00:00
|
|
|
If destination and source are unaligned and copy count is greater than 256
|
2011-01-12 18:59:09 +00:00
|
|
|
then copy 0-3 bytes to make destination aligned.
|
|
|
|
|
If 32 or more bytes to copy we use 32 byte copy loop.
|
2023-05-23 12:11:30 +00:00
|
|
|
Finally we copy 0-31 extra bytes. */
|
2011-01-12 18:59:09 +00:00
|
|
|
|
2013-02-28 21:23:47 +00:00
|
|
|
EALIGN (memcpy, 5, 0)
|
2011-01-12 18:59:09 +00:00
|
|
|
/* Check if bytes to copy are greater than 256 and if
|
|
|
|
|
source and destination are unaligned */
|
|
|
|
|
cmpwi r5,0x0100
|
|
|
|
|
addi r0,r3,0
|
|
|
|
|
ble L(string_count_loop)
|
|
|
|
|
neg r6,r3
|
|
|
|
|
clrlwi. r6,r6,30
|
|
|
|
|
beq L(string_count_loop)
|
|
|
|
|
neg r6,r4
|
|
|
|
|
clrlwi. r6,r6,30
|
|
|
|
|
beq L(string_count_loop)
|
|
|
|
|
mtctr r6
|
|
|
|
|
subf r5,r6,r5
|
|
|
|
|
|
|
|
|
|
L(unaligned_bytecopy_loop): /* Align destination by coping 0-3 bytes */
|
|
|
|
|
lbz r8,0x0(r4)
|
|
|
|
|
addi r4,r4,1
|
|
|
|
|
stb r8,0x0(r3)
|
|
|
|
|
addi r3,r3,1
|
|
|
|
|
bdnz L(unaligned_bytecopy_loop)
|
|
|
|
|
srwi. r7,r5,5
|
|
|
|
|
beq L(preword2_count_loop)
|
|
|
|
|
mtctr r7
|
|
|
|
|
|
|
|
|
|
L(word8_count_loop_no_dcbt): /* Copy 32 bytes at a time */
|
|
|
|
|
lwz r6,0(r4)
|
|
|
|
|
lwz r7,4(r4)
|
|
|
|
|
lwz r8,8(r4)
|
|
|
|
|
lwz r9,12(r4)
|
|
|
|
|
subi r5,r5,0x20
|
|
|
|
|
stw r6,0(r3)
|
|
|
|
|
stw r7,4(r3)
|
|
|
|
|
stw r8,8(r3)
|
|
|
|
|
stw r9,12(r3)
|
|
|
|
|
lwz r6,16(r4)
|
|
|
|
|
lwz r7,20(r4)
|
|
|
|
|
lwz r8,24(r4)
|
|
|
|
|
lwz r9,28(r4)
|
|
|
|
|
addi r4,r4,0x20
|
|
|
|
|
stw r6,16(r3)
|
|
|
|
|
stw r7,20(r3)
|
|
|
|
|
stw r8,24(r3)
|
|
|
|
|
stw r9,28(r3)
|
|
|
|
|
addi r3,r3,0x20
|
|
|
|
|
bdnz L(word8_count_loop_no_dcbt)
|
|
|
|
|
|
|
|
|
|
L(preword2_count_loop): /* Copy remaining 0-31 bytes */
|
|
|
|
|
clrlwi. r12,r5,27
|
|
|
|
|
beq L(end_memcpy)
|
|
|
|
|
mtxer r12
|
|
|
|
|
lswx r5,0,r4
|
|
|
|
|
stswx r5,0,r3
|
|
|
|
|
mr r3,r0
|
|
|
|
|
blr
|
|
|
|
|
|
|
|
|
|
L(string_count_loop): /* Copy odd 0-31 bytes */
|
|
|
|
|
clrlwi. r12,r5,28
|
|
|
|
|
add r3,r3,r5
|
|
|
|
|
add r4,r4,r5
|
|
|
|
|
beq L(pre_string_copy)
|
|
|
|
|
mtxer r12
|
|
|
|
|
subf r4,r12,r4
|
|
|
|
|
subf r3,r12,r3
|
|
|
|
|
lswx r6,0,r4
|
|
|
|
|
stswx r6,0,r3
|
|
|
|
|
|
2013-08-30 12:32:07 +00:00
|
|
|
L(pre_string_copy): /* Check how many 32 byte chunks to copy */
|
2011-01-12 18:59:09 +00:00
|
|
|
srwi. r7,r5,4
|
|
|
|
|
beq L(end_memcpy)
|
|
|
|
|
mtctr r7
|
|
|
|
|
|
|
|
|
|
L(word4_count_loop_no_dcbt): /* Copy 32 bytes at a time */
|
|
|
|
|
lwz r6,-4(r4)
|
|
|
|
|
lwz r7,-8(r4)
|
|
|
|
|
lwz r8,-12(r4)
|
|
|
|
|
lwzu r9,-16(r4)
|
|
|
|
|
stw r6,-4(r3)
|
|
|
|
|
stw r7,-8(r3)
|
|
|
|
|
stw r8,-12(r3)
|
|
|
|
|
stwu r9,-16(r3)
|
|
|
|
|
bdz L(end_memcpy)
|
|
|
|
|
lwz r6,-4(r4)
|
|
|
|
|
lwz r7,-8(r4)
|
|
|
|
|
lwz r8,-12(r4)
|
|
|
|
|
lwzu r9,-16(r4)
|
|
|
|
|
stw r6,-4(r3)
|
|
|
|
|
stw r7,-8(r3)
|
|
|
|
|
stw r8,-12(r3)
|
|
|
|
|
stwu r9,-16(r3)
|
|
|
|
|
bdnz L(word4_count_loop_no_dcbt)
|
|
|
|
|
|
|
|
|
|
L(end_memcpy):
|
|
|
|
|
mr r3,r0
|
|
|
|
|
blr
|
2013-02-28 21:23:47 +00:00
|
|
|
END (memcpy)
|
2011-01-12 18:59:09 +00:00
|
|
|
libc_hidden_builtin_def (memcpy)
|
