a855debfb4
* libquadmath.texi (FLT128_DIG, FLT128_MIN_10_EXP, FLT128_MAX_10_EXP): Document. (strtoflt128): Remove obsolete comment. * configure.ac (HAVE_STRTOULL): New check. * printf/gmp-impl.h (mpn_construct_float128): New prototype, define. * printf/mul_n.c: Include <config.h>. * printf/add_n.c: Likewise. * printf/cmp.c: Likewise. * printf/fpioconst.c: Likewise. * printf/mul_1.c: Likewise. * printf/rshift.c: Likewise. * printf/lshift.c: Likewise. * printf/submul_1.c: Likewise. * printf/sub_n.c: Likewise. * printf/divrem.c: Likewise. * printf/addmul_1.c: Likewise. * printf/mul.c: Likewise. * printf/quadmath-printf.h (isupper, isdigit, tolower): Change to avoid evaluating argument multiple times. (isxdigit): Redefine. * strtod/strtoflt128.c: New file. * strtod/strtod_l.c: New file. * strtod/mpn2flt128.c: New file. * strtod/grouping.h: New file. * strtod/tens_in_limb.c: New file. * gdtoa/arith.h: Removed. * gdtoa/gd_qnan.h: Removed. * gdtoa/gdtoa_fltrnds.h: Removed. * gdtoa/gdtoa.h: Removed. * gdtoa/gdtoaimp.h: Removed. * gdtoa/gethex.c: Removed. * gdtoa/gmisc.c: Removed. * gdtoa/hd_init.c: Removed. * gdtoa/hexnan.c: Removed. * gdtoa/makefile: Removed. * gdtoa/misc.c: Removed. * gdtoa/README.gdtoa: Removed. * gdtoa/smisc.c: Removed. * gdtoa/strtodg.c: Removed. * gdtoa/strtopQ.c: Removed. * gdtoa/sum.c: Removed. * quadmath.h (FLT128_DIG, FLT128_MIN_10_EXP, FLT128_MAX_10_EXP): Define. * Makefile.am (libquadmath_la_SOURCES): Remove gdtoa/*, add strtod/strtoflt128.c, strtod/mpn2flt128.c and strtod/tens_in_limb.c. * config.h.in: Regenerated. * configure: Regenerated. * Makefile.in: Regenerated. From-SVN: r170254
182 lines
5.5 KiB
C
182 lines
5.5 KiB
C
/* Include file for internal GNU MP types and definitions.
|
|
|
|
Copyright (C) 1991, 1993, 1994, 1995, 1996 Free Software Foundation, Inc.
|
|
|
|
This file is part of the GNU MP Library.
|
|
|
|
The GNU MP 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 MP 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 MP Library; see the file COPYING.LIB. If not, write to
|
|
the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
|
|
MA 02111-1307, USA. */
|
|
|
|
#include <stdlib.h>
|
|
#include "quadmath-imp.h"
|
|
|
|
#undef alloca
|
|
#define alloca __builtin_alloca
|
|
|
|
#define ABS(x) (x >= 0 ? x : -x)
|
|
#ifndef MIN
|
|
#define MIN(l,o) ((l) < (o) ? (l) : (o))
|
|
#endif
|
|
#ifndef MAX
|
|
#define MAX(h,i) ((h) > (i) ? (h) : (i))
|
|
#endif
|
|
|
|
#define BITS_PER_MP_LIMB (__SIZEOF_LONG__ * __CHAR_BIT__)
|
|
#define BYTES_PER_MP_LIMB (BITS_PER_MP_LIMB / __CHAR_BIT__)
|
|
typedef unsigned long int mp_limb_t;
|
|
typedef long int mp_limb_signed_t;
|
|
|
|
typedef mp_limb_t * mp_ptr;
|
|
typedef const mp_limb_t * mp_srcptr;
|
|
typedef long int mp_size_t;
|
|
typedef long int mp_exp_t;
|
|
|
|
/* Define stuff for longlong.h. */
|
|
typedef unsigned int UQItype __attribute__ ((mode (QI)));
|
|
typedef int SItype __attribute__ ((mode (SI)));
|
|
typedef unsigned int USItype __attribute__ ((mode (SI)));
|
|
typedef int DItype __attribute__ ((mode (DI)));
|
|
typedef unsigned int UDItype __attribute__ ((mode (DI)));
|
|
|
|
typedef mp_limb_t UWtype;
|
|
typedef unsigned int UHWtype;
|
|
#define W_TYPE_SIZE BITS_PER_MP_LIMB
|
|
|
|
#ifdef HAVE_HIDDEN_VISIBILITY
|
|
#define attribute_hidden __attribute__((__visibility__ ("hidden")))
|
|
#else
|
|
#define attribute_hidden
|
|
#endif
|
|
|
|
#include "../../gcc/longlong.h"
|
|
|
|
/* Copy NLIMBS *limbs* from SRC to DST. */
|
|
#define MPN_COPY_INCR(DST, SRC, NLIMBS) \
|
|
do { \
|
|
mp_size_t __i; \
|
|
for (__i = 0; __i < (NLIMBS); __i++) \
|
|
(DST)[__i] = (SRC)[__i]; \
|
|
} while (0)
|
|
#define MPN_COPY_DECR(DST, SRC, NLIMBS) \
|
|
do { \
|
|
mp_size_t __i; \
|
|
for (__i = (NLIMBS) - 1; __i >= 0; __i--) \
|
|
(DST)[__i] = (SRC)[__i]; \
|
|
} while (0)
|
|
#define MPN_COPY MPN_COPY_INCR
|
|
|
|
/* Zero NLIMBS *limbs* AT DST. */
|
|
#define MPN_ZERO(DST, NLIMBS) \
|
|
do { \
|
|
mp_size_t __i; \
|
|
for (__i = 0; __i < (NLIMBS); __i++) \
|
|
(DST)[__i] = 0; \
|
|
} while (0)
|
|
|
|
#define MPN_MUL_N_RECURSE(prodp, up, vp, size, tspace) \
|
|
do { \
|
|
if ((size) < KARATSUBA_THRESHOLD) \
|
|
impn_mul_n_basecase (prodp, up, vp, size); \
|
|
else \
|
|
impn_mul_n (prodp, up, vp, size, tspace); \
|
|
} while (0);
|
|
|
|
#define __MPN(x) __quadmath_mpn_##x
|
|
|
|
/* Internal mpn calls */
|
|
#define impn_mul_n_basecase __MPN(impn_mul_n_basecase)
|
|
#define impn_mul_n __MPN(impn_mul_n)
|
|
|
|
/* Prototypes for internal mpn calls. */
|
|
void impn_mul_n_basecase (mp_ptr prodp, mp_srcptr up, mp_srcptr vp,
|
|
mp_size_t size) attribute_hidden;
|
|
void impn_mul_n (mp_ptr prodp, mp_srcptr up, mp_srcptr vp, mp_size_t size,
|
|
mp_ptr tspace) attribute_hidden;
|
|
|
|
#define mpn_add_n __MPN(add_n)
|
|
#define mpn_addmul_1 __MPN(addmul_1)
|
|
#define mpn_cmp __MPN(cmp)
|
|
#define mpn_divrem __MPN(divrem)
|
|
#define mpn_lshift __MPN(lshift)
|
|
#define mpn_mul __MPN(mul)
|
|
#define mpn_mul_1 __MPN(mul_1)
|
|
#define mpn_rshift __MPN(rshift)
|
|
#define mpn_sub_n __MPN(sub_n)
|
|
#define mpn_submul_1 __MPN(submul_1)
|
|
|
|
mp_limb_t mpn_add_n (mp_ptr, mp_srcptr, mp_srcptr, mp_size_t)
|
|
attribute_hidden;
|
|
mp_limb_t mpn_addmul_1 (mp_ptr, mp_srcptr, mp_size_t, mp_limb_t)
|
|
attribute_hidden;
|
|
int mpn_cmp (mp_srcptr, mp_srcptr, mp_size_t) attribute_hidden;
|
|
mp_limb_t mpn_divrem (mp_ptr, mp_size_t, mp_ptr, mp_size_t, mp_srcptr,
|
|
mp_size_t) attribute_hidden;
|
|
mp_limb_t mpn_lshift (mp_ptr, mp_srcptr, mp_size_t, unsigned int)
|
|
attribute_hidden;
|
|
mp_limb_t mpn_mul (mp_ptr, mp_srcptr, mp_size_t, mp_srcptr, mp_size_t)
|
|
attribute_hidden;
|
|
mp_limb_t mpn_mul_1 (mp_ptr, mp_srcptr, mp_size_t, mp_limb_t)
|
|
attribute_hidden;
|
|
mp_limb_t mpn_rshift (mp_ptr, mp_srcptr, mp_size_t, unsigned int)
|
|
attribute_hidden;
|
|
mp_limb_t mpn_sub_n (mp_ptr, mp_srcptr, mp_srcptr, mp_size_t)
|
|
attribute_hidden;
|
|
mp_limb_t mpn_submul_1 (mp_ptr, mp_srcptr, mp_size_t, mp_limb_t)
|
|
attribute_hidden;
|
|
|
|
#define mpn_extract_flt128 __MPN(extract_flt128)
|
|
mp_size_t mpn_extract_flt128 (mp_ptr res_ptr, mp_size_t size, int *expt,
|
|
int *is_neg, __float128 value) attribute_hidden;
|
|
|
|
#define mpn_construct_float128 __MPN(construct_float128)
|
|
__float128 mpn_construct_float128 (mp_srcptr frac_ptr, int expt, int sign)
|
|
attribute_hidden;
|
|
|
|
#define mpn_divmod(qp,np,nsize,dp,dsize) mpn_divrem (qp,0,np,nsize,dp,dsize)
|
|
|
|
static inline mp_limb_t
|
|
mpn_add_1 (register mp_ptr res_ptr,
|
|
register mp_srcptr s1_ptr,
|
|
register mp_size_t s1_size,
|
|
register mp_limb_t s2_limb)
|
|
{
|
|
register mp_limb_t x;
|
|
|
|
x = *s1_ptr++;
|
|
s2_limb = x + s2_limb;
|
|
*res_ptr++ = s2_limb;
|
|
if (s2_limb < x)
|
|
{
|
|
while (--s1_size != 0)
|
|
{
|
|
x = *s1_ptr++ + 1;
|
|
*res_ptr++ = x;
|
|
if (x != 0)
|
|
goto fin;
|
|
}
|
|
|
|
return 1;
|
|
}
|
|
|
|
fin:
|
|
if (res_ptr != s1_ptr)
|
|
{
|
|
mp_size_t i;
|
|
for (i = 0; i < s1_size - 1; i++)
|
|
res_ptr[i] = s1_ptr[i];
|
|
}
|
|
return 0;
|
|
}
|