/* Copyright (C) 1991, 1993, 1996-1997, 1999-2000, 2003-2004, 2006, 2008-2010
|
/* Copyright (C) 1991, 1993, 1996-1997, 1999-2000, 2003-2004, 2006, 2008-2010
|
Free Software Foundation, Inc.
|
Free Software Foundation, Inc.
|
|
|
Based on strlen implementation by Torbjorn Granlund (tege@sics.se),
|
Based on strlen implementation by Torbjorn Granlund (tege@sics.se),
|
with help from Dan Sahlin (dan@sics.se) and
|
with help from Dan Sahlin (dan@sics.se) and
|
commentary by Jim Blandy (jimb@ai.mit.edu);
|
commentary by Jim Blandy (jimb@ai.mit.edu);
|
adaptation to memchr suggested by Dick Karpinski (dick@cca.ucsf.edu),
|
adaptation to memchr suggested by Dick Karpinski (dick@cca.ucsf.edu),
|
and implemented by Roland McGrath (roland@ai.mit.edu).
|
and implemented by Roland McGrath (roland@ai.mit.edu).
|
|
|
NOTE: The canonical source of this file is maintained with the GNU C Library.
|
NOTE: The canonical source of this file is maintained with the GNU C Library.
|
Bugs can be reported to bug-glibc@prep.ai.mit.edu.
|
Bugs can be reported to bug-glibc@prep.ai.mit.edu.
|
|
|
This program is free software: you can redistribute it and/or modify it
|
This program is free software: you can redistribute it and/or modify it
|
under the terms of the GNU General Public License as published by the
|
under the terms of the GNU General Public License as published by the
|
Free Software Foundation; either version 3 of the License, or any
|
Free Software Foundation; either version 3 of the License, or any
|
later version.
|
later version.
|
|
|
This program is distributed in the hope that it will be useful,
|
This program is distributed in the hope that it will be useful,
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
GNU General Public License for more details.
|
GNU General Public License for more details.
|
|
|
You should have received a copy of the GNU General Public License
|
You should have received a copy of the GNU General Public License
|
along with this program. If not, see <http://www.gnu.org/licenses/>. */
|
along with this program. If not, see <http://www.gnu.org/licenses/>. */
|
|
|
#ifndef _LIBC
|
#ifndef _LIBC
|
# include <config.h>
|
# include <config.h>
|
#endif
|
#endif
|
|
|
#include <string.h>
|
#include <string.h>
|
|
|
#include <stddef.h>
|
#include <stddef.h>
|
|
|
#if defined _LIBC
|
#if defined _LIBC
|
# include <memcopy.h>
|
# include <memcopy.h>
|
#else
|
#else
|
# define reg_char char
|
# define reg_char char
|
#endif
|
#endif
|
|
|
#include <limits.h>
|
#include <limits.h>
|
|
|
#if HAVE_BP_SYM_H || defined _LIBC
|
#if HAVE_BP_SYM_H || defined _LIBC
|
# include <bp-sym.h>
|
# include <bp-sym.h>
|
#else
|
#else
|
# define BP_SYM(sym) sym
|
# define BP_SYM(sym) sym
|
#endif
|
#endif
|
|
|
#undef __memchr
|
#undef __memchr
|
#ifdef _LIBC
|
#ifdef _LIBC
|
# undef memchr
|
# undef memchr
|
#endif
|
#endif
|
|
|
#ifndef weak_alias
|
#ifndef weak_alias
|
# define __memchr memchr
|
# define __memchr memchr
|
#endif
|
#endif
|
|
|
/* Search no more than N bytes of S for C. */
|
/* Search no more than N bytes of S for C. */
|
void *
|
void *
|
__memchr (void const *s, int c_in, size_t n)
|
__memchr (void const *s, int c_in, size_t n)
|
{
|
{
|
/* On 32-bit hardware, choosing longword to be a 32-bit unsigned
|
/* On 32-bit hardware, choosing longword to be a 32-bit unsigned
|
long instead of a 64-bit uintmax_t tends to give better
|
long instead of a 64-bit uintmax_t tends to give better
|
performance. On 64-bit hardware, unsigned long is generally 64
|
performance. On 64-bit hardware, unsigned long is generally 64
|
bits already. Change this typedef to experiment with
|
bits already. Change this typedef to experiment with
|
performance. */
|
performance. */
|
typedef unsigned long int longword;
|
typedef unsigned long int longword;
|
|
|
const unsigned char *char_ptr;
|
const unsigned char *char_ptr;
|
const longword *longword_ptr;
|
const longword *longword_ptr;
|
longword repeated_one;
|
longword repeated_one;
|
longword repeated_c;
|
longword repeated_c;
|
unsigned reg_char c;
|
unsigned reg_char c;
|
|
|
c = (unsigned char) c_in;
|
c = (unsigned char) c_in;
|
|
|
/* Handle the first few bytes by reading one byte at a time.
|
/* Handle the first few bytes by reading one byte at a time.
|
Do this until CHAR_PTR is aligned on a longword boundary. */
|
Do this until CHAR_PTR is aligned on a longword boundary. */
|
for (char_ptr = (const unsigned char *) s;
|
for (char_ptr = (const unsigned char *) s;
|
n > 0 && (size_t) char_ptr % sizeof (longword) != 0;
|
n > 0 && (size_t) char_ptr % sizeof (longword) != 0;
|
--n, ++char_ptr)
|
--n, ++char_ptr)
|
if (*char_ptr == c)
|
if (*char_ptr == c)
|
return (void *) char_ptr;
|
return (void *) char_ptr;
|
|
|
longword_ptr = (const longword *) char_ptr;
|
longword_ptr = (const longword *) char_ptr;
|
|
|
/* All these elucidatory comments refer to 4-byte longwords,
|
/* All these elucidatory comments refer to 4-byte longwords,
|
but the theory applies equally well to any size longwords. */
|
but the theory applies equally well to any size longwords. */
|
|
|
/* Compute auxiliary longword values:
|
/* Compute auxiliary longword values:
|
repeated_one is a value which has a 1 in every byte.
|
repeated_one is a value which has a 1 in every byte.
|
repeated_c has c in every byte. */
|
repeated_c has c in every byte. */
|
repeated_one = 0x01010101;
|
repeated_one = 0x01010101;
|
repeated_c = c | (c << 8);
|
repeated_c = c | (c << 8);
|
repeated_c |= repeated_c << 16;
|
repeated_c |= repeated_c << 16;
|
if (0xffffffffU < (longword) -1)
|
if (0xffffffffU < (longword) -1)
|
{
|
{
|
repeated_one |= repeated_one << 31 << 1;
|
repeated_one |= repeated_one << 31 << 1;
|
repeated_c |= repeated_c << 31 << 1;
|
repeated_c |= repeated_c << 31 << 1;
|
if (8 < sizeof (longword))
|
if (8 < sizeof (longword))
|
{
|
{
|
size_t i;
|
size_t i;
|
|
|
for (i = 64; i < sizeof (longword) * 8; i *= 2)
|
for (i = 64; i < sizeof (longword) * 8; i *= 2)
|
{
|
{
|
repeated_one |= repeated_one << i;
|
repeated_one |= repeated_one << i;
|
repeated_c |= repeated_c << i;
|
repeated_c |= repeated_c << i;
|
}
|
}
|
}
|
}
|
}
|
}
|
|
|
/* Instead of the traditional loop which tests each byte, we will test a
|
/* Instead of the traditional loop which tests each byte, we will test a
|
longword at a time. The tricky part is testing if *any of the four*
|
longword at a time. The tricky part is testing if *any of the four*
|
bytes in the longword in question are equal to c. We first use an xor
|
bytes in the longword in question are equal to c. We first use an xor
|
with repeated_c. This reduces the task to testing whether *any of the
|
with repeated_c. This reduces the task to testing whether *any of the
|
four* bytes in longword1 is zero.
|
four* bytes in longword1 is zero.
|
|
|
We compute tmp =
|
We compute tmp =
|
((longword1 - repeated_one) & ~longword1) & (repeated_one << 7).
|
((longword1 - repeated_one) & ~longword1) & (repeated_one << 7).
|
That is, we perform the following operations:
|
That is, we perform the following operations:
|
1. Subtract repeated_one.
|
1. Subtract repeated_one.
|
2. & ~longword1.
|
2. & ~longword1.
|
3. & a mask consisting of 0x80 in every byte.
|
3. & a mask consisting of 0x80 in every byte.
|
Consider what happens in each byte:
|
Consider what happens in each byte:
|
- If a byte of longword1 is zero, step 1 and 2 transform it into 0xff,
|
- If a byte of longword1 is zero, step 1 and 2 transform it into 0xff,
|
and step 3 transforms it into 0x80. A carry can also be propagated
|
and step 3 transforms it into 0x80. A carry can also be propagated
|
to more significant bytes.
|
to more significant bytes.
|
- If a byte of longword1 is nonzero, let its lowest 1 bit be at
|
- If a byte of longword1 is nonzero, let its lowest 1 bit be at
|
position k (0 <= k <= 7); so the lowest k bits are 0. After step 1,
|
position k (0 <= k <= 7); so the lowest k bits are 0. After step 1,
|
the byte ends in a single bit of value 0 and k bits of value 1.
|
the byte ends in a single bit of value 0 and k bits of value 1.
|
After step 2, the result is just k bits of value 1: 2^k - 1. After
|
After step 2, the result is just k bits of value 1: 2^k - 1. After
|
step 3, the result is 0. And no carry is produced.
|
step 3, the result is 0. And no carry is produced.
|
So, if longword1 has only non-zero bytes, tmp is zero.
|
So, if longword1 has only non-zero bytes, tmp is zero.
|
Whereas if longword1 has a zero byte, call j the position of the least
|
Whereas if longword1 has a zero byte, call j the position of the least
|
significant zero byte. Then the result has a zero at positions 0, ...,
|
significant zero byte. Then the result has a zero at positions 0, ...,
|
j-1 and a 0x80 at position j. We cannot predict the result at the more
|
j-1 and a 0x80 at position j. We cannot predict the result at the more
|
significant bytes (positions j+1..3), but it does not matter since we
|
significant bytes (positions j+1..3), but it does not matter since we
|
already have a non-zero bit at position 8*j+7.
|
already have a non-zero bit at position 8*j+7.
|
|
|
So, the test whether any byte in longword1 is zero is equivalent to
|
So, the test whether any byte in longword1 is zero is equivalent to
|
testing whether tmp is nonzero. */
|
testing whether tmp is nonzero. */
|
|
|
while (n >= sizeof (longword))
|
while (n >= sizeof (longword))
|
{
|
{
|
longword longword1 = *longword_ptr ^ repeated_c;
|
longword longword1 = *longword_ptr ^ repeated_c;
|
|
|
if ((((longword1 - repeated_one) & ~longword1)
|
if ((((longword1 - repeated_one) & ~longword1)
|
& (repeated_one << 7)) != 0)
|
& (repeated_one << 7)) != 0)
|
break;
|
break;
|
longword_ptr++;
|
longword_ptr++;
|
n -= sizeof (longword);
|
n -= sizeof (longword);
|
}
|
}
|
|
|
char_ptr = (const unsigned char *) longword_ptr;
|
char_ptr = (const unsigned char *) longword_ptr;
|
|
|
/* At this point, we know that either n < sizeof (longword), or one of the
|
/* At this point, we know that either n < sizeof (longword), or one of the
|
sizeof (longword) bytes starting at char_ptr is == c. On little-endian
|
sizeof (longword) bytes starting at char_ptr is == c. On little-endian
|
machines, we could determine the first such byte without any further
|
machines, we could determine the first such byte without any further
|
memory accesses, just by looking at the tmp result from the last loop
|
memory accesses, just by looking at the tmp result from the last loop
|
iteration. But this does not work on big-endian machines. Choose code
|
iteration. But this does not work on big-endian machines. Choose code
|
that works in both cases. */
|
that works in both cases. */
|
|
|
for (; n > 0; --n, ++char_ptr)
|
for (; n > 0; --n, ++char_ptr)
|
{
|
{
|
if (*char_ptr == c)
|
if (*char_ptr == c)
|
return (void *) char_ptr;
|
return (void *) char_ptr;
|
}
|
}
|
|
|
return NULL;
|
return NULL;
|
}
|
}
|
#ifdef weak_alias
|
#ifdef weak_alias
|
weak_alias (__memchr, BP_SYM (memchr))
|
weak_alias (__memchr, BP_SYM (memchr))
|
#endif
|
#endif
|
|
|
