1
0
mirror of https://sourceware.org/git/glibc.git synced 2025-08-01 10:06:57 +03:00
Files
glibc/sysdeps/ieee754/flt-32/e_acoshf.c
Adhemerval Zanella 994007ff29 math: Consolidate acoshf and asinhf internal tables
The libm size improvement built with "--enable-stack-protector=strong
--enable-bind-now=yes --enable-fortify-source=2":

Before:

   text    data     bss     dec     hex filename
 587304     860      12  588176   8f990 aarch64-linux-gnu-master/math/libm.so
 962855    1068      12  963935   eb55f x86_64-linux-gnu-master/math/libm.so
1191222    5544     368 1197134  12444e powerpc64le-linux-gnu-master/math/libm.so

After:

   text    data     bss     dec     hex filename
 585192     860      12  586064   8f150 aarch64-linux-gnu/math/libm.so
 960775    1068      12  961855   ead3f x86_64-linux-gnu/math/libm.so
1189174    5544     368 1195086  123c4e powerpc64le-linux-gnu/math/libm.so

The are small code changes for x86_64 and powerpc64le, which do not
affect performance; but on aarch64 with gcc-14 I see a slight better
code generation due the usage of ldq for floating point constant loading.
Reviewed-by: Andreas K. Huettel <dilfridge@gentoo.org>
2025-02-12 16:31:57 -03:00

109 lines
3.5 KiB
C

/* Correctly-rounded inverse hyperbolic cosine function for binary32 format.
Copyright (c) 2023-2024 Alexei Sibidanov.
The original version of this file was copied from the CORE-MATH
project (file src/binary32/acosh/acoshf.c, revision d0b9ddd).
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/
#include <math.h>
#include <stdint.h>
#include <libm-alias-finite.h>
#include "math_config.h"
#include "s_asincoshf_data.h"
static __attribute__ ((noinline)) float
as_special (float x)
{
uint32_t t = asuint (x);
if (t == 0x3f800000u)
return 0.0f;
if ((t << 1) > 0xff000000u)
return x + x; /* nan */
if (t == 0x7f800000u)
return x; /* inf */
return __math_invalidf (x);
}
float
__ieee754_acoshf (float x)
{
uint32_t t = asuint (x);
if (__glibc_unlikely (t <= 0x3f800000u))
return as_special (x);
else if (__glibc_unlikely (t <= 0x3f99db23u)) /* x <= 0x1.33b646p+0 */
{
float zf = x - 1.0f;
double z = zf;
double a = sqrt (2.0 * z);
static const double c[] =
{
-0x1.555555555491ep-4, 0x1.333333319c2p-6,
-0x1.6db6da26e5e0ep-8, 0x1.f1c698b7100cep-10,
-0x1.6e7ca9e152dc2p-11, 0x1.1b4cf187158bdp-12,
-0x1.b5137f11a88e5p-14, 0x1.056c6c57152b2p-15
};
double z2 = z * z;
double z4 = z2 * z2;
double f = ((c[0] + z * c[1]) + z2 * (c[2] + z * c[3]))
+ z4 * ((c[4] + z * c[5]) + z2 * (c[6] + z * c[7]));
double r = a + (a * z) * f;
return r;
}
else if (__glibc_likely (t < 0x7f800000u))
{
double xd = x;
double x2 = xd * xd;
uint64_t tp = asuint64 (xd + sqrt (x2 - 1));
uint64_t m = tp & (~UINT64_C(0) >> 12);
int j = (m + (UINT64_C(1) << (52 - 8))) >> (52 - 7);
int e = (tp >> 52) - 0x3ff;
double w = asdouble (m | UINT64_C(0x3ff) << 52);
double z = w * IX[j] - 1.0;
double z2 = z * z;
double r = ((LIX[128] * e + LIX[j]) + z * C[0])
+ z2 * (C[1] + z * C[2]);
if (__glibc_unlikely (((asuint64 (r) + 259000) & INT64_C(0xfffffff))
< 260000)) /* accurate path */
{
z2 = z * z;
double c0 = CP[0] + z * CP[1];
double c2 = CP[2] + z * CP[3];
double c4 = CP[4] + z * CP[5];
c0 += z2 * (c2 + z2 * c4);
const double ln2l = 0x1.7f7d1cf79abcap-20;
const double ln2h = 0x1.62e4p-1;
double Lh = ln2h * e;
double Ll = ln2l * e;
r = fma (z, c0, Ll + LIX[j]) + Lh;
if (__glibc_unlikely ((asuint64 (r) & UINT64_C(0xfffffff)) == 0))
{
double h = fma (z, c0, Ll + LIX[j]) + (Lh - r);
r = r + 64.0 * h;
}
}
return r;
}
else
return as_special (x);
}
libm_alias_finite (__ieee754_acoshf, __acoshf)