[uClibc-cvs] uClibc/libc/string/arm Makefile, NONE, 1.1 _memcpy.S, NONE, 1.1 bcopy.S, NONE, 1.1 bzero.S, NONE, 1.1 memcmp.S, NONE, 1.1 memcpy.S, NONE, 1.1 memmove.S, NONE, 1.1 memset.S, NONE, 1.1 strcmp.S, NONE, 1.1 strlen.S, NONE, 1.1 strncmp.S, NONE, 1.1
Erik Andersen
andersen at uclibc.org
Thu Nov 20 15:23:08 UTC 2003
Update of /var/cvs/uClibc/libc/string/arm
In directory winder:/home/andersen/CVS/uClibc/libc/string/arm
Added Files:
Makefile _memcpy.S bcopy.S bzero.S memcmp.S memcpy.S memmove.S
memset.S strcmp.S strlen.S strncmp.S
Log Message:
Add an initial set of arm optimized string functions. These
do seem to make noticable speed improvement...
-Erik
--- NEW FILE: memset.S ---
/* Copyright (C) 1998 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Contributed by Philip Blundell <philb at gnu.org>
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, write to the Free
Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
02111-1307 USA. */
#include <sys/syscall.h>
.text
.global memset;
.type memset,%function
.align 4; \
memset:
mov a4, a1
cmp a3, $8 @ at least 8 bytes to do?
blt 2f
orr a2, a2, a2, lsl $8
orr a2, a2, a2, lsl $16
1:
tst a4, $3 @ aligned yet?
strneb a2, [a4], $1
subne a3, a3, $1
bne 1b
mov ip, a2
1:
cmp a3, $8 @ 8 bytes still to do?
blt 2f
stmia a4!, {a2, ip}
sub a3, a3, $8
cmp a3, $8 @ 8 bytes still to do?
blt 2f
stmia a4!, {a2, ip}
sub a3, a3, $8
cmp a3, $8 @ 8 bytes still to do?
blt 2f
stmia a4!, {a2, ip}
sub a3, a3, $8
cmp a3, $8 @ 8 bytes still to do?
stmgeia a4!, {a2, ip}
subge a3, a3, $8
bge 1b
2:
movs a3, a3 @ anything left?
moveq pc, lr @ nope
rsb a3, a3, $7
add pc, pc, a3, lsl $2
mov r0, r0
strb a2, [a4], $1
strb a2, [a4], $1
strb a2, [a4], $1
strb a2, [a4], $1
strb a2, [a4], $1
strb a2, [a4], $1
strb a2, [a4], $1
mov pc, lr
.size memset,.-memset;
--- NEW FILE: strncmp.S ---
/*
* Copyright (c) 2002 ARM Ltd
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. The name of the company may not be used to endorse or promote
* products derived from this software without specific prior written
* permission.
*
* THIS SOFTWARE IS PROVIDED BY ARM LTD ``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 ARM LTD 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.
*
* Adapted for uClibc from NetBSD strncmp.S, version 1.2 2003/04/05
* by Erik Andersen <andersen at codepoet.org>
*/
.text
.global strncmp;
.type strncmp,%function
.align 4; \
strncmp:
/* if ((len - 1) < 0) return 0 */
subs r2, r2, #1
movmi r0, #0
movmi pc, lr
/* ip == last src address to compare */
add ip, r0, r2
1:
ldrb r2, [r0], #1
ldrb r3, [r1], #1
cmp ip, r0
cmpcs r2, #1
cmpcs r2, r3
beq 1b
sub r0, r2, r3
mov pc, lr
--- NEW FILE: strcmp.S ---
/*
* Copyright (c) 2002 ARM Ltd
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. The name of the company may not be used to endorse or promote
* products derived from this software without specific prior written
* permission.
*
* THIS SOFTWARE IS PROVIDED BY ARM LTD ``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 ARM LTD 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.
*
* Adapted for uClibc from NetBSD strcmp.S, version 1.3 2003/04/05
* by Erik Andersen <andersen at codepoet.org>
*/
.text
.global strcmp;
.type strcmp,%function
.align 4; \
strcmp:
1:
ldrb r2, [r0], #1
ldrb r3, [r1], #1
cmp r2, #1
cmpcs r2, r3
beq 1b
sub r0, r2, r3
mov pc, lr
.weak strcoll;
strcoll = strcmp
--- NEW FILE: memcpy.S ---
/*-
* Copyright (c) 1997 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Neil A. Carson and Mark Brinicombe
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the NetBSD
* Foundation, Inc. and its contributors.
* 4. Neither the name of The NetBSD Foundation 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 NETBSD FOUNDATION, INC. 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 FOUNDATION 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.
*
* Adapted for uClibc from NetBSD memcpy.S,v 1.3 2003/04/05
* by Erik Andersen <andersen at codepoet.org>
*/
.text
.global memcpy;
.type memcpy,%function
.align 4; \
memcpy:
stmfd sp!, {r0, lr}
bl _memcpy (PLT)
ldmfd sp!, {r0, pc}
--- NEW FILE: Makefile ---
# Makefile for uClibc
#
# Copyright (C) 2000-2003 Erik Andersen <andersen at uclibc.org>
#
# This program is free software; you can redistribute it and/or modify it under
# the terms of the GNU Library General Public License as published by the Free
# Software Foundation; either version 2 of the License, or (at your option) any
# later version.
#
# This program 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 Library General Public License for more
# details.
#
# You should have received a copy of the GNU Library General Public License
# along with this program; if not, write to the Free Software Foundation, Inc.,
# 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
TOPDIR=../../../
include $(TOPDIR)Rules.mak
SSRC= _memcpy.S bcopy.S bzero.S memcmp.S memcpy.S memmove.S memset.S \
strcmp.S strlen.S strncmp.S
SOBJS=$(patsubst %.S,%.o, $(SSRC))
all: $(SOBJS) $(LIBC)
$(LIBC): ar-target
ar-target: $(SOBJS)
$(AR) $(ARFLAGS) $(LIBC) $(SOBJS)
$(SOBJS): %.o : %.S
$(CC) $(CFLAGS) -c $< -o $@
$(STRIPTOOL) -x -R .note -R .comment $*.o
clean:
$(RM) *.[oa] *~ core
--- NEW FILE: bzero.S ---
/*-
* Copyright (c) 1997 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Neil A. Carson and Mark Brinicombe
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the NetBSD
* Foundation, Inc. and its contributors.
* 4. Neither the name of The NetBSD Foundation 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 NETBSD FOUNDATION, INC. 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 FOUNDATION 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.
*
* Adapted for uClibc from NetBSD bzero.S,v 1.2 1997/10/09
* by Erik Andersen <andersen at codepoet.org>
*/
.text
.global bzero;
.type bzero,%function
.align 4; \
bzero:
mov r2, r1
mov r1, #0
b memset (PLT)
--- NEW FILE: bcopy.S ---
/*-
* Copyright (c) 1997 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Neil A. Carson and Mark Brinicombe
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the NetBSD
* Foundation, Inc. and its contributors.
* 4. Neither the name of The NetBSD Foundation 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 NETBSD FOUNDATION, INC. 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 FOUNDATION 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.
*
* Adapted for uClibc from NetBSD bcopy.S,v 1.2 1997/10/09
* by Erik Andersen <andersen at codepoet.org>
*/
/* bcopy = memcpy/memmove with arguments reversed. */
.text
.global bcopy;
.type bcopy,%function
.align 4; \
bcopy:
/* switch the source and destination registers */
eor r0, r1, r0
eor r1, r0, r1
eor r0, r1, r0
b _memcpy (PLT)
--- NEW FILE: _memcpy.S ---
/*-
* Copyright (c) 1997 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Neil A. Carson and Mark Brinicombe
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the NetBSD
* Foundation, Inc. and its contributors.
* 4. Neither the name of The NetBSD Foundation 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 NETBSD FOUNDATION, INC. 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 FOUNDATION 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.
*
* Adapted for uClibc from NetBSD _memcpy.S,v 1.6 2003/10/09
* by Erik Andersen <andersen at codepoet.org>
*/
#include <endian.h>
/*
* This is one fun bit of code ...
* Some easy listening music is suggested while trying to understand this
* code e.g. Iron Maiden
*
* For anyone attempting to understand it :
*
* The core code is implemented here with simple stubs for memcpy()
* memmove() and bcopy().
*
* All local labels are prefixed with Lmemcpy_
* Following the prefix a label starting f is used in the forward copy code
* while a label using b is used in the backwards copy code
* The source and destination addresses determine whether a forward or
* backward copy is performed.
* Separate bits of code are used to deal with the following situations
* for both the forward and backwards copy.
* unaligned source address
* unaligned destination address
* Separate copy routines are used to produce an optimised result for each
* of these cases.
* The copy code will use LDM/STM instructions to copy up to 32 bytes at
* a time where possible.
*
* Note: r12 (aka ip) can be trashed during the function along with
* r0-r3 although r0-r2 have defined uses i.e. src, dest, len through out.
* Additional registers are preserved prior to use i.e. r4, r5 & lr
*
* Apologies for the state of the comments ;-)
*/
.text
.global _memcpy;
.type _memcpy,%function
.align 4; \
_memcpy:
/* Determine copy direction */
cmp r1, r0
bcc .Lmemcpy_backwards
moveq r0, #0 /* Quick abort for len=0 */
moveq pc, lr
stmdb sp!, {r0, lr} /* memcpy() returns dest addr */
subs r2, r2, #4
blt .Lmemcpy_fl4 /* less than 4 bytes */
ands r12, r0, #3
bne .Lmemcpy_fdestul /* oh unaligned destination addr */
ands r12, r1, #3
bne .Lmemcpy_fsrcul /* oh unaligned source addr */
.Lmemcpy_ft8:
/* We have aligned source and destination */
subs r2, r2, #8
blt .Lmemcpy_fl12 /* less than 12 bytes (4 from above) */
subs r2, r2, #0x14
blt .Lmemcpy_fl32 /* less than 32 bytes (12 from above) */
stmdb sp!, {r4} /* borrow r4 */
/* blat 32 bytes at a time */
/* XXX for really big copies perhaps we should use more registers */
.Lmemcpy_floop32:
ldmia r1!, {r3, r4, r12, lr}
stmia r0!, {r3, r4, r12, lr}
ldmia r1!, {r3, r4, r12, lr}
stmia r0!, {r3, r4, r12, lr}
subs r2, r2, #0x20
bge .Lmemcpy_floop32
cmn r2, #0x10
ldmgeia r1!, {r3, r4, r12, lr} /* blat a remaining 16 bytes */
stmgeia r0!, {r3, r4, r12, lr}
subge r2, r2, #0x10
ldmia sp!, {r4} /* return r4 */
.Lmemcpy_fl32:
adds r2, r2, #0x14
/* blat 12 bytes at a time */
.Lmemcpy_floop12:
ldmgeia r1!, {r3, r12, lr}
stmgeia r0!, {r3, r12, lr}
subges r2, r2, #0x0c
bge .Lmemcpy_floop12
.Lmemcpy_fl12:
adds r2, r2, #8
blt .Lmemcpy_fl4
subs r2, r2, #4
ldrlt r3, [r1], #4
strlt r3, [r0], #4
ldmgeia r1!, {r3, r12}
stmgeia r0!, {r3, r12}
subge r2, r2, #4
.Lmemcpy_fl4:
/* less than 4 bytes to go */
adds r2, r2, #4
ldmeqia sp!, {r0, pc} /* done */
/* copy the crud byte at a time */
cmp r2, #2
ldrb r3, [r1], #1
strb r3, [r0], #1
ldrgeb r3, [r1], #1
strgeb r3, [r0], #1
ldrgtb r3, [r1], #1
strgtb r3, [r0], #1
ldmia sp!, {r0, pc}
/* erg - unaligned destination */
.Lmemcpy_fdestul:
rsb r12, r12, #4
cmp r12, #2
/* align destination with byte copies */
ldrb r3, [r1], #1
strb r3, [r0], #1
ldrgeb r3, [r1], #1
strgeb r3, [r0], #1
ldrgtb r3, [r1], #1
strgtb r3, [r0], #1
subs r2, r2, r12
blt .Lmemcpy_fl4 /* less the 4 bytes */
ands r12, r1, #3
beq .Lmemcpy_ft8 /* we have an aligned source */
/* erg - unaligned source */
/* This is where it gets nasty ... */
.Lmemcpy_fsrcul:
bic r1, r1, #3
ldr lr, [r1], #4
cmp r12, #2
bgt .Lmemcpy_fsrcul3
beq .Lmemcpy_fsrcul2
cmp r2, #0x0c
blt .Lmemcpy_fsrcul1loop4
sub r2, r2, #0x0c
stmdb sp!, {r4, r5}
.Lmemcpy_fsrcul1loop16:
#if BYTE_ORDER == BIG_ENDIAN
mov r3, lr, lsl #8
ldmia r1!, {r4, r5, r12, lr}
orr r3, r3, r4, lsr #24
mov r4, r4, lsl #8
orr r4, r4, r5, lsr #24
mov r5, r5, lsl #8
orr r5, r5, r12, lsr #24
mov r12, r12, lsl #8
orr r12, r12, lr, lsr #24
#else
mov r3, lr, lsr #8
ldmia r1!, {r4, r5, r12, lr}
orr r3, r3, r4, lsl #24
mov r4, r4, lsr #8
orr r4, r4, r5, lsl #24
mov r5, r5, lsr #8
orr r5, r5, r12, lsl #24
mov r12, r12, lsr #8
orr r12, r12, lr, lsl #24
#endif
stmia r0!, {r3-r5, r12}
subs r2, r2, #0x10
bge .Lmemcpy_fsrcul1loop16
ldmia sp!, {r4, r5}
adds r2, r2, #0x0c
blt .Lmemcpy_fsrcul1l4
.Lmemcpy_fsrcul1loop4:
#if BYTE_ORDER == BIG_ENDIAN
mov r12, lr, lsl #8
ldr lr, [r1], #4
orr r12, r12, lr, lsr #24
#else
mov r12, lr, lsr #8
ldr lr, [r1], #4
orr r12, r12, lr, lsl #24
#endif
str r12, [r0], #4
subs r2, r2, #4
bge .Lmemcpy_fsrcul1loop4
.Lmemcpy_fsrcul1l4:
sub r1, r1, #3
b .Lmemcpy_fl4
.Lmemcpy_fsrcul2:
cmp r2, #0x0c
blt .Lmemcpy_fsrcul2loop4
sub r2, r2, #0x0c
stmdb sp!, {r4, r5}
.Lmemcpy_fsrcul2loop16:
#if BYTE_ORDER == BIG_ENDIAN
mov r3, lr, lsl #16
ldmia r1!, {r4, r5, r12, lr}
orr r3, r3, r4, lsr #16
mov r4, r4, lsl #16
orr r4, r4, r5, lsr #16
mov r5, r5, lsl #16
orr r5, r5, r12, lsr #16
mov r12, r12, lsl #16
orr r12, r12, lr, lsr #16
#else
mov r3, lr, lsr #16
ldmia r1!, {r4, r5, r12, lr}
orr r3, r3, r4, lsl #16
mov r4, r4, lsr #16
orr r4, r4, r5, lsl #16
mov r5, r5, lsr #16
orr r5, r5, r12, lsl #16
mov r12, r12, lsr #16
orr r12, r12, lr, lsl #16
#endif
stmia r0!, {r3-r5, r12}
subs r2, r2, #0x10
bge .Lmemcpy_fsrcul2loop16
ldmia sp!, {r4, r5}
adds r2, r2, #0x0c
blt .Lmemcpy_fsrcul2l4
.Lmemcpy_fsrcul2loop4:
#if BYTE_ORDER == BIG_ENDIAN
mov r12, lr, lsl #16
ldr lr, [r1], #4
orr r12, r12, lr, lsr #16
#else
mov r12, lr, lsr #16
ldr lr, [r1], #4
orr r12, r12, lr, lsl #16
#endif
str r12, [r0], #4
subs r2, r2, #4
bge .Lmemcpy_fsrcul2loop4
.Lmemcpy_fsrcul2l4:
sub r1, r1, #2
b .Lmemcpy_fl4
.Lmemcpy_fsrcul3:
cmp r2, #0x0c
blt .Lmemcpy_fsrcul3loop4
sub r2, r2, #0x0c
stmdb sp!, {r4, r5}
.Lmemcpy_fsrcul3loop16:
#if BYTE_ORDER == BIG_ENDIAN
mov r3, lr, lsl #24
ldmia r1!, {r4, r5, r12, lr}
orr r3, r3, r4, lsr #8
mov r4, r4, lsl #24
orr r4, r4, r5, lsr #8
mov r5, r5, lsl #24
orr r5, r5, r12, lsr #8
mov r12, r12, lsl #24
orr r12, r12, lr, lsr #8
#else
mov r3, lr, lsr #24
ldmia r1!, {r4, r5, r12, lr}
orr r3, r3, r4, lsl #8
mov r4, r4, lsr #24
orr r4, r4, r5, lsl #8
mov r5, r5, lsr #24
orr r5, r5, r12, lsl #8
mov r12, r12, lsr #24
orr r12, r12, lr, lsl #8
#endif
stmia r0!, {r3-r5, r12}
subs r2, r2, #0x10
bge .Lmemcpy_fsrcul3loop16
ldmia sp!, {r4, r5}
adds r2, r2, #0x0c
blt .Lmemcpy_fsrcul3l4
.Lmemcpy_fsrcul3loop4:
#if BYTE_ORDER == BIG_ENDIAN
mov r12, lr, lsl #24
ldr lr, [r1], #4
orr r12, r12, lr, lsr #8
#else
mov r12, lr, lsr #24
ldr lr, [r1], #4
orr r12, r12, lr, lsl #8
#endif
str r12, [r0], #4
subs r2, r2, #4
bge .Lmemcpy_fsrcul3loop4
.Lmemcpy_fsrcul3l4:
sub r1, r1, #1
b .Lmemcpy_fl4
.Lmemcpy_backwards:
add r1, r1, r2
add r0, r0, r2
subs r2, r2, #4
blt .Lmemcpy_bl4 /* less than 4 bytes */
ands r12, r0, #3
bne .Lmemcpy_bdestul /* oh unaligned destination addr */
ands r12, r1, #3
bne .Lmemcpy_bsrcul /* oh unaligned source addr */
.Lmemcpy_bt8:
/* We have aligned source and destination */
subs r2, r2, #8
blt .Lmemcpy_bl12 /* less than 12 bytes (4 from above) */
stmdb sp!, {r4, lr}
subs r2, r2, #0x14 /* less than 32 bytes (12 from above) */
blt .Lmemcpy_bl32
/* blat 32 bytes at a time */
/* XXX for really big copies perhaps we should use more registers */
.Lmemcpy_bloop32:
ldmdb r1!, {r3, r4, r12, lr}
stmdb r0!, {r3, r4, r12, lr}
ldmdb r1!, {r3, r4, r12, lr}
stmdb r0!, {r3, r4, r12, lr}
subs r2, r2, #0x20
bge .Lmemcpy_bloop32
.Lmemcpy_bl32:
cmn r2, #0x10
ldmgedb r1!, {r3, r4, r12, lr} /* blat a remaining 16 bytes */
stmgedb r0!, {r3, r4, r12, lr}
subge r2, r2, #0x10
adds r2, r2, #0x14
ldmgedb r1!, {r3, r12, lr} /* blat a remaining 12 bytes */
stmgedb r0!, {r3, r12, lr}
subge r2, r2, #0x0c
ldmia sp!, {r4, lr}
.Lmemcpy_bl12:
adds r2, r2, #8
blt .Lmemcpy_bl4
subs r2, r2, #4
ldrlt r3, [r1, #-4]!
strlt r3, [r0, #-4]!
ldmgedb r1!, {r3, r12}
stmgedb r0!, {r3, r12}
subge r2, r2, #4
.Lmemcpy_bl4:
/* less than 4 bytes to go */
adds r2, r2, #4
moveq pc, lr /* done */
/* copy the crud byte at a time */
cmp r2, #2
ldrb r3, [r1, #-1]!
strb r3, [r0, #-1]!
ldrgeb r3, [r1, #-1]!
strgeb r3, [r0, #-1]!
ldrgtb r3, [r1, #-1]!
strgtb r3, [r0, #-1]!
mov pc, lr
/* erg - unaligned destination */
.Lmemcpy_bdestul:
cmp r12, #2
/* align destination with byte copies */
ldrb r3, [r1, #-1]!
strb r3, [r0, #-1]!
ldrgeb r3, [r1, #-1]!
strgeb r3, [r0, #-1]!
ldrgtb r3, [r1, #-1]!
strgtb r3, [r0, #-1]!
subs r2, r2, r12
blt .Lmemcpy_bl4 /* less than 4 bytes to go */
ands r12, r1, #3
beq .Lmemcpy_bt8 /* we have an aligned source */
/* erg - unaligned source */
/* This is where it gets nasty ... */
.Lmemcpy_bsrcul:
bic r1, r1, #3
ldr r3, [r1, #0]
cmp r12, #2
blt .Lmemcpy_bsrcul1
beq .Lmemcpy_bsrcul2
cmp r2, #0x0c
blt .Lmemcpy_bsrcul3loop4
sub r2, r2, #0x0c
stmdb sp!, {r4, r5, lr}
.Lmemcpy_bsrcul3loop16:
#if BYTE_ORDER == BIG_ENDIAN
mov lr, r3, lsr #8
ldmdb r1!, {r3-r5, r12}
orr lr, lr, r12, lsl #24
mov r12, r12, lsr #8
orr r12, r12, r5, lsl #24
mov r5, r5, lsr #8
orr r5, r5, r4, lsl #24
mov r4, r4, lsr #8
orr r4, r4, r3, lsl #24
#else
mov lr, r3, lsl #8
ldmdb r1!, {r3-r5, r12}
orr lr, lr, r12, lsr #24
mov r12, r12, lsl #8
orr r12, r12, r5, lsr #24
mov r5, r5, lsl #8
orr r5, r5, r4, lsr #24
mov r4, r4, lsl #8
orr r4, r4, r3, lsr #24
#endif
stmdb r0!, {r4, r5, r12, lr}
subs r2, r2, #0x10
bge .Lmemcpy_bsrcul3loop16
ldmia sp!, {r4, r5, lr}
adds r2, r2, #0x0c
blt .Lmemcpy_bsrcul3l4
.Lmemcpy_bsrcul3loop4:
#if BYTE_ORDER == BIG_ENDIAN
mov r12, r3, lsr #8
ldr r3, [r1, #-4]!
orr r12, r12, r3, lsl #24
#else
mov r12, r3, lsl #8
ldr r3, [r1, #-4]!
orr r12, r12, r3, lsr #24
#endif
str r12, [r0, #-4]!
subs r2, r2, #4
bge .Lmemcpy_bsrcul3loop4
.Lmemcpy_bsrcul3l4:
add r1, r1, #3
b .Lmemcpy_bl4
.Lmemcpy_bsrcul2:
cmp r2, #0x0c
blt .Lmemcpy_bsrcul2loop4
sub r2, r2, #0x0c
stmdb sp!, {r4, r5, lr}
.Lmemcpy_bsrcul2loop16:
#if BYTE_ORDER == BIG_ENDIAN
mov lr, r3, lsr #16
ldmdb r1!, {r3-r5, r12}
orr lr, lr, r12, lsl #16
mov r12, r12, lsr #16
orr r12, r12, r5, lsl #16
mov r5, r5, lsr #16
orr r5, r5, r4, lsl #16
mov r4, r4, lsr #16
orr r4, r4, r3, lsl #16
#else
mov lr, r3, lsl #16
ldmdb r1!, {r3-r5, r12}
orr lr, lr, r12, lsr #16
mov r12, r12, lsl #16
orr r12, r12, r5, lsr #16
mov r5, r5, lsl #16
orr r5, r5, r4, lsr #16
mov r4, r4, lsl #16
orr r4, r4, r3, lsr #16
#endif
stmdb r0!, {r4, r5, r12, lr}
subs r2, r2, #0x10
bge .Lmemcpy_bsrcul2loop16
ldmia sp!, {r4, r5, lr}
adds r2, r2, #0x0c
blt .Lmemcpy_bsrcul2l4
.Lmemcpy_bsrcul2loop4:
#if BYTE_ORDER == BIG_ENDIAN
mov r12, r3, lsr #16
ldr r3, [r1, #-4]!
orr r12, r12, r3, lsl #16
#else
mov r12, r3, lsl #16
ldr r3, [r1, #-4]!
orr r12, r12, r3, lsr #16
#endif
str r12, [r0, #-4]!
subs r2, r2, #4
bge .Lmemcpy_bsrcul2loop4
.Lmemcpy_bsrcul2l4:
add r1, r1, #2
b .Lmemcpy_bl4
.Lmemcpy_bsrcul1:
cmp r2, #0x0c
blt .Lmemcpy_bsrcul1loop4
sub r2, r2, #0x0c
stmdb sp!, {r4, r5, lr}
.Lmemcpy_bsrcul1loop32:
#if BYTE_ORDER == BIG_ENDIAN
mov lr, r3, lsr #24
ldmdb r1!, {r3-r5, r12}
orr lr, lr, r12, lsl #8
mov r12, r12, lsr #24
orr r12, r12, r5, lsl #8
mov r5, r5, lsr #24
orr r5, r5, r4, lsl #8
mov r4, r4, lsr #24
orr r4, r4, r3, lsl #8
#else
mov lr, r3, lsl #24
ldmdb r1!, {r3-r5, r12}
orr lr, lr, r12, lsr #8
mov r12, r12, lsl #24
orr r12, r12, r5, lsr #8
mov r5, r5, lsl #24
orr r5, r5, r4, lsr #8
mov r4, r4, lsl #24
orr r4, r4, r3, lsr #8
#endif
stmdb r0!, {r4, r5, r12, lr}
subs r2, r2, #0x10
bge .Lmemcpy_bsrcul1loop32
ldmia sp!, {r4, r5, lr}
adds r2, r2, #0x0c
blt .Lmemcpy_bsrcul1l4
.Lmemcpy_bsrcul1loop4:
#if BYTE_ORDER == BIG_ENDIAN
mov r12, r3, lsr #24
ldr r3, [r1, #-4]!
orr r12, r12, r3, lsl #8
#else
mov r12, r3, lsl #24
ldr r3, [r1, #-4]!
orr r12, r12, r3, lsr #8
#endif
str r12, [r0, #-4]!
subs r2, r2, #4
bge .Lmemcpy_bsrcul1loop4
.Lmemcpy_bsrcul1l4:
add r1, r1, #1
b .Lmemcpy_bl4
--- NEW FILE: memmove.S ---
/*-
* Copyright (c) 1997 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Neil A. Carson and Mark Brinicombe
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the NetBSD
* Foundation, Inc. and its contributors.
* 4. Neither the name of The NetBSD Foundation 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 NETBSD FOUNDATION, INC. 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 FOUNDATION 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.
*
* Adapted for uClibc from NetBSD memmove.S,v 1.3 2003/04/05
* by Erik Andersen <andersen at codepoet.org>
*/
.text
.global memmove;
.type memmove,%function
.align 4; \
memmove:
stmfd sp!, {r0, lr}
bl _memcpy (PLT)
ldmfd sp!, {r0, pc}
--- NEW FILE: memcmp.S ---
/*
* Copyright (c) 2002 ARM Ltd
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. The name of the company may not be used to endorse or promote
* products derived from this software without specific prior written
* permission.
*
* THIS SOFTWARE IS PROVIDED BY ARM LTD ``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 ARM LTD 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.
*
* Adapted for uClibc from NetBSD memcmp.S, version 1.2 2003/04/05
* by Erik Andersen <andersen at codepoet.org>
*/
.text
.global memcmp;
.type memcmp,%function
.align 4; \
memcmp:
/* if ((len - 1) < 0) return 0 */
subs r2, r2, #1
movmi r0, #0
movmi pc, lr
/* ip == last src address to compare */
add ip, r0, r2
1:
ldrb r2, [r0], #1
ldrb r3, [r1], #1
cmp ip, r0
cmpcs r2, r3
beq 1b
sub r0, r2, r3
mov pc, lr
--- NEW FILE: strlen.S ---
/* Copyright (C) 1998 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Code contributed by Matthew Wilcox <willy at odie.barnet.ac.uk>
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, write to the Free
Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
02111-1307 USA. */
#include <endian.h>
#include <sys/syscall.h>
/* size_t strlen(const char *S)
* entry: r0 -> string
* exit: r0 = len
*/
.text
.global strlen;
.type strlen,%function
.align 4; \
strlen:
bic r1, r0, $3 @ addr of word containing first byte
ldr r2, [r1], $4 @ get the first word
ands r3, r0, $3 @ how many bytes are duff?
rsb r0, r3, $0 @ get - that number into counter.
beq Laligned @ skip into main check routine if no
@ more
#if __BYTE_ORDER == __BIG_ENDIAN
orr r2, r2, $0xff000000 @ set this byte to non-zero
subs r3, r3, $1 @ any more to do?
orrgt r2, r2, $0x00ff0000 @ if so, set this byte
subs r3, r3, $1 @ more?
orrgt r2, r2, $0x0000ff00 @ then set.
#else
orr r2, r2, $0x000000ff @ set this byte to non-zero
subs r3, r3, $1 @ any more to do?
orrgt r2, r2, $0x0000ff00 @ if so, set this byte
subs r3, r3, $1 @ more?
orrgt r2, r2, $0x00ff0000 @ then set.
#endif
Laligned: @ here, we have a word in r2. Does it
tst r2, $0x000000ff @ contain any zeroes?
tstne r2, $0x0000ff00 @
tstne r2, $0x00ff0000 @
tstne r2, $0xff000000 @
addne r0, r0, $4 @ if not, the string is 4 bytes longer
ldrne r2, [r1], $4 @ and we continue to the next word
bne Laligned @
Llastword: @ drop through to here once we find a
tst r2, $0x000000ff @ word that has a zero byte in it
addne r0, r0, $1 @
tstne r2, $0x0000ff00 @ and add up to 3 bytes on to it
addne r0, r0, $1 @
tstne r2, $0x00ff0000 @ (if first three all non-zero, 4th
addne r0, r0, $1 @ must be zero)
mov pc,lr
.size strlen,.-strlen;
More information about the uClibc-cvs
mailing list