mirror of
https://sourceware.org/git/glibc.git
synced 2024-12-03 10:21:05 +00:00
8f5b00d375
This patch continues cleaning up math_private.h by moving the math_check_force_underflow set of macros to a separate header math-underflow.h. This header is included by the files that need it rather than from math_private.h. Moving these macros to a separate file removes the math_private.h uses of macros from float.h, so the inclusion of float.h in math_private.h is also removed; files that were depending on that inclusion are fixed to include float.h directly. The inclusion of math-barriers.h from math_private.h will be removed in a separate patch. Tested for x86_64 and x86. Also tested with build-many-glibcs.py that installed stripped shared libraries are unchanged by this patch. * math/math-underflow.h: New file. * sysdeps/generic/math_private.h: Do not include <float.h>. (fabs_tg): Remove macro. Moved to math-underflow.h. (min_of_type_f): Likewise. (min_of_type_): Likewise. (min_of_type_l): Likewise. (min_of_type_f128): Likewise. (min_of_type): Likewise. (math_check_force_underflow): Likewise. (math_check_force_underflow_nonneg): Likewise. (math_check_force_underflow_complex): Likewise. * math/e_exp2_template.c: Include <math-underflow.h>. * math/k_casinh_template.c: Likewise. * math/s_catan_template.c: Likewise. * math/s_catanh_template.c: Likewise. * math/s_ccosh_template.c: Likewise. * math/s_cexp_template.c: Likewise. * math/s_clog10_template.c: Likewise. * math/s_clog_template.c: Likewise. * math/s_csin_template.c: Likewise. * math/s_csinh_template.c: Likewise. * math/s_csqrt_template.c: Likewise. * math/s_ctan_template.c: Likewise. * math/s_ctanh_template.c: Likewise. * sysdeps/ieee754/dbl-64/e_asin.c: Likewise. * sysdeps/ieee754/dbl-64/e_atanh.c: Likewise. * sysdeps/ieee754/dbl-64/e_exp2.c: Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c: Likewise. * sysdeps/ieee754/dbl-64/e_hypot.c: Likewise. * sysdeps/ieee754/dbl-64/e_j1.c: Likewise. * sysdeps/ieee754/dbl-64/e_jn.c: Likewise. * sysdeps/ieee754/dbl-64/e_pow.c: Likewise. * sysdeps/ieee754/dbl-64/e_sinh.c: Likewise. * sysdeps/ieee754/dbl-64/s_asinh.c: Likewise. * sysdeps/ieee754/dbl-64/s_atan.c: Likewise. * sysdeps/ieee754/dbl-64/s_erf.c: Likewise. * sysdeps/ieee754/dbl-64/s_expm1.c: Likewise. * sysdeps/ieee754/dbl-64/s_log1p.c: Likewise. * sysdeps/ieee754/dbl-64/s_sin.c: Likewise. * sysdeps/ieee754/dbl-64/s_sincos.c: Likewise. * sysdeps/ieee754/dbl-64/s_tan.c: Likewise. * sysdeps/ieee754/dbl-64/s_tanh.c: Likewise. * sysdeps/ieee754/flt-32/e_asinf.c: Likewise. * sysdeps/ieee754/flt-32/e_atanhf.c: Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c: Likewise. * sysdeps/ieee754/flt-32/e_j1f.c: Likewise. * sysdeps/ieee754/flt-32/e_jnf.c: Likewise. * sysdeps/ieee754/flt-32/e_sinhf.c: Likewise. * sysdeps/ieee754/flt-32/k_sinf.c: Likewise. * sysdeps/ieee754/flt-32/k_tanf.c: Likewise. * sysdeps/ieee754/flt-32/s_asinhf.c: Likewise. * sysdeps/ieee754/flt-32/s_atanf.c: Likewise. * sysdeps/ieee754/flt-32/s_erff.c: Likewise. * sysdeps/ieee754/flt-32/s_expm1f.c: Likewise. * sysdeps/ieee754/flt-32/s_log1pf.c: Likewise. * sysdeps/ieee754/flt-32/s_tanhf.c: Likewise. * sysdeps/ieee754/ldbl-128/e_asinl.c: Likewise. * sysdeps/ieee754/ldbl-128/e_atanhl.c: Likewise. * sysdeps/ieee754/ldbl-128/e_expl.c: Likewise. * sysdeps/ieee754/ldbl-128/e_gammal_r.c: Likewise. * sysdeps/ieee754/ldbl-128/e_hypotl.c: Likewise. * sysdeps/ieee754/ldbl-128/e_j1l.c: Likewise. * sysdeps/ieee754/ldbl-128/e_jnl.c: Likewise. * sysdeps/ieee754/ldbl-128/e_sinhl.c: Likewise. * sysdeps/ieee754/ldbl-128/k_sincosl.c: Likewise. * sysdeps/ieee754/ldbl-128/k_sinl.c: Likewise. * sysdeps/ieee754/ldbl-128/k_tanl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_asinhl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_atanl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_erfl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_expm1l.c: Likewise. * sysdeps/ieee754/ldbl-128/s_log1pl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_tanhl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_asinl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_atanhl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_hypotl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_j1l.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_jnl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_powl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/e_sinhl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/k_sincosl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/k_sinl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/k_tanl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_asinhl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_atanl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_erfl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fmal.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_tanhl.c: Likewise. * sysdeps/ieee754/ldbl-96/e_asinl.c: Likewise. * sysdeps/ieee754/ldbl-96/e_atanhl.c: Likewise. * sysdeps/ieee754/ldbl-96/e_gammal_r.c: Likewise. * sysdeps/ieee754/ldbl-96/e_hypotl.c: Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c: Likewise. * sysdeps/ieee754/ldbl-96/e_jnl.c: Likewise. * sysdeps/ieee754/ldbl-96/e_sinhl.c: Likewise. * sysdeps/ieee754/ldbl-96/k_sinl.c: Likewise. * sysdeps/ieee754/ldbl-96/k_tanl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_asinhl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_erfl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_tanhl.c: Likewise. * sysdeps/powerpc/fpu/e_hypot.c: Likewise. * sysdeps/x86/fpu/powl_helper.c: Likewise. * sysdeps/ieee754/dbl-64/s_nextup.c: Include <float.h>. * sysdeps/ieee754/flt-32/s_nextupf.c: Likewise. * sysdeps/ieee754/ldbl-128/s_nextupl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nextupl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_nextupl.c: Likewise.
212 lines
6.0 KiB
C
212 lines
6.0 KiB
C
/* Implementation of gamma function according to ISO C.
|
|
Copyright (C) 1997-2018 Free Software Foundation, Inc.
|
|
This file is part of the GNU C Library.
|
|
Contributed by Ulrich Drepper <drepper@cygnus.com>, 1997.
|
|
|
|
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
|
|
<http://www.gnu.org/licenses/>. */
|
|
|
|
#include <math.h>
|
|
#include <math_private.h>
|
|
#include <math-underflow.h>
|
|
#include <float.h>
|
|
|
|
/* Coefficients B_2k / 2k(2k-1) of x^-(2k-1) inside exp in Stirling's
|
|
approximation to gamma function. */
|
|
|
|
static const long double gamma_coeff[] =
|
|
{
|
|
0x1.5555555555555556p-4L,
|
|
-0xb.60b60b60b60b60bp-12L,
|
|
0x3.4034034034034034p-12L,
|
|
-0x2.7027027027027028p-12L,
|
|
0x3.72a3c5631fe46aep-12L,
|
|
-0x7.daac36664f1f208p-12L,
|
|
0x1.a41a41a41a41a41ap-8L,
|
|
-0x7.90a1b2c3d4e5f708p-8L,
|
|
};
|
|
|
|
#define NCOEFF (sizeof (gamma_coeff) / sizeof (gamma_coeff[0]))
|
|
|
|
/* Return gamma (X), for positive X less than 1766, in the form R *
|
|
2^(*EXP2_ADJ), where R is the return value and *EXP2_ADJ is set to
|
|
avoid overflow or underflow in intermediate calculations. */
|
|
|
|
static long double
|
|
gammal_positive (long double x, int *exp2_adj)
|
|
{
|
|
int local_signgam;
|
|
if (x < 0.5L)
|
|
{
|
|
*exp2_adj = 0;
|
|
return __ieee754_expl (__ieee754_lgammal_r (x + 1, &local_signgam)) / x;
|
|
}
|
|
else if (x <= 1.5L)
|
|
{
|
|
*exp2_adj = 0;
|
|
return __ieee754_expl (__ieee754_lgammal_r (x, &local_signgam));
|
|
}
|
|
else if (x < 7.5L)
|
|
{
|
|
/* Adjust into the range for using exp (lgamma). */
|
|
*exp2_adj = 0;
|
|
long double n = __ceill (x - 1.5L);
|
|
long double x_adj = x - n;
|
|
long double eps;
|
|
long double prod = __gamma_productl (x_adj, 0, n, &eps);
|
|
return (__ieee754_expl (__ieee754_lgammal_r (x_adj, &local_signgam))
|
|
* prod * (1.0L + eps));
|
|
}
|
|
else
|
|
{
|
|
long double eps = 0;
|
|
long double x_eps = 0;
|
|
long double x_adj = x;
|
|
long double prod = 1;
|
|
if (x < 13.0L)
|
|
{
|
|
/* Adjust into the range for applying Stirling's
|
|
approximation. */
|
|
long double n = __ceill (13.0L - x);
|
|
x_adj = x + n;
|
|
x_eps = (x - (x_adj - n));
|
|
prod = __gamma_productl (x_adj - n, x_eps, n, &eps);
|
|
}
|
|
/* The result is now gamma (X_ADJ + X_EPS) / (PROD * (1 + EPS)).
|
|
Compute gamma (X_ADJ + X_EPS) using Stirling's approximation,
|
|
starting by computing pow (X_ADJ, X_ADJ) with a power of 2
|
|
factored out. */
|
|
long double exp_adj = -eps;
|
|
long double x_adj_int = __roundl (x_adj);
|
|
long double x_adj_frac = x_adj - x_adj_int;
|
|
int x_adj_log2;
|
|
long double x_adj_mant = __frexpl (x_adj, &x_adj_log2);
|
|
if (x_adj_mant < M_SQRT1_2l)
|
|
{
|
|
x_adj_log2--;
|
|
x_adj_mant *= 2.0L;
|
|
}
|
|
*exp2_adj = x_adj_log2 * (int) x_adj_int;
|
|
long double ret = (__ieee754_powl (x_adj_mant, x_adj)
|
|
* __ieee754_exp2l (x_adj_log2 * x_adj_frac)
|
|
* __ieee754_expl (-x_adj)
|
|
* sqrtl (2 * M_PIl / x_adj)
|
|
/ prod);
|
|
exp_adj += x_eps * __ieee754_logl (x_adj);
|
|
long double bsum = gamma_coeff[NCOEFF - 1];
|
|
long double x_adj2 = x_adj * x_adj;
|
|
for (size_t i = 1; i <= NCOEFF - 1; i++)
|
|
bsum = bsum / x_adj2 + gamma_coeff[NCOEFF - 1 - i];
|
|
exp_adj += bsum / x_adj;
|
|
return ret + ret * __expm1l (exp_adj);
|
|
}
|
|
}
|
|
|
|
long double
|
|
__ieee754_gammal_r (long double x, int *signgamp)
|
|
{
|
|
uint32_t es, hx, lx;
|
|
long double ret;
|
|
|
|
GET_LDOUBLE_WORDS (es, hx, lx, x);
|
|
|
|
if (__glibc_unlikely (((es & 0x7fff) | hx | lx) == 0))
|
|
{
|
|
/* Return value for x == 0 is Inf with divide by zero exception. */
|
|
*signgamp = 0;
|
|
return 1.0 / x;
|
|
}
|
|
if (__glibc_unlikely (es == 0xffffffff && ((hx & 0x7fffffff) | lx) == 0))
|
|
{
|
|
/* x == -Inf. According to ISO this is NaN. */
|
|
*signgamp = 0;
|
|
return x - x;
|
|
}
|
|
if (__glibc_unlikely ((es & 0x7fff) == 0x7fff))
|
|
{
|
|
/* Positive infinity (return positive infinity) or NaN (return
|
|
NaN). */
|
|
*signgamp = 0;
|
|
return x + x;
|
|
}
|
|
if (__builtin_expect ((es & 0x8000) != 0, 0) && __rintl (x) == x)
|
|
{
|
|
/* Return value for integer x < 0 is NaN with invalid exception. */
|
|
*signgamp = 0;
|
|
return (x - x) / (x - x);
|
|
}
|
|
|
|
if (x >= 1756.0L)
|
|
{
|
|
/* Overflow. */
|
|
*signgamp = 0;
|
|
return LDBL_MAX * LDBL_MAX;
|
|
}
|
|
else
|
|
{
|
|
SET_RESTORE_ROUNDL (FE_TONEAREST);
|
|
if (x > 0.0L)
|
|
{
|
|
*signgamp = 0;
|
|
int exp2_adj;
|
|
ret = gammal_positive (x, &exp2_adj);
|
|
ret = __scalbnl (ret, exp2_adj);
|
|
}
|
|
else if (x >= -LDBL_EPSILON / 4.0L)
|
|
{
|
|
*signgamp = 0;
|
|
ret = 1.0L / x;
|
|
}
|
|
else
|
|
{
|
|
long double tx = __truncl (x);
|
|
*signgamp = (tx == 2.0L * __truncl (tx / 2.0L)) ? -1 : 1;
|
|
if (x <= -1766.0L)
|
|
/* Underflow. */
|
|
ret = LDBL_MIN * LDBL_MIN;
|
|
else
|
|
{
|
|
long double frac = tx - x;
|
|
if (frac > 0.5L)
|
|
frac = 1.0L - frac;
|
|
long double sinpix = (frac <= 0.25L
|
|
? __sinl (M_PIl * frac)
|
|
: __cosl (M_PIl * (0.5L - frac)));
|
|
int exp2_adj;
|
|
ret = M_PIl / (-x * sinpix
|
|
* gammal_positive (-x, &exp2_adj));
|
|
ret = __scalbnl (ret, -exp2_adj);
|
|
math_check_force_underflow_nonneg (ret);
|
|
}
|
|
}
|
|
}
|
|
if (isinf (ret) && x != 0)
|
|
{
|
|
if (*signgamp < 0)
|
|
return -(-__copysignl (LDBL_MAX, ret) * LDBL_MAX);
|
|
else
|
|
return __copysignl (LDBL_MAX, ret) * LDBL_MAX;
|
|
}
|
|
else if (ret == 0)
|
|
{
|
|
if (*signgamp < 0)
|
|
return -(-__copysignl (LDBL_MIN, ret) * LDBL_MIN);
|
|
else
|
|
return __copysignl (LDBL_MIN, ret) * LDBL_MIN;
|
|
}
|
|
else
|
|
return ret;
|
|
}
|
|
strong_alias (__ieee754_gammal_r, __gammal_r_finite)
|