mirror of
https://github.com/fluencelabs/musl
synced 2025-05-19 02:31:29 +00:00
e216951f50
When FLT_EVAL_METHOD!=0 (only i386 with x87 fp) the excess
precision of an expression must be removed in an assignment.
(gcc needs -fexcess-precision=standard or -std=c99 for this)
This is done by extra load/store instructions which adds code
bloat when lot of temporaries are used and it makes the result
less precise in many cases.
Using double_t and float_t avoids these issues on i386 and
it makes no difference on other archs.
For now only a few functions are modified where the excess
precision is clearly beneficial (mostly polynomial evaluations
with temporaries).
object size differences on i386, gcc-4.8:
old new
__cosdf.o 123 95
__cos.o 199 169
__sindf.o 131 95
__sin.o 225 203
__tandf.o 207 151
__tan.o 605 499
erff.o 1470 1416
erf.o 1703 1649
j0f.o 1779 1745
j0.o 2308 2274
j1f.o 1602 1568
j1.o 2286 2252
tgamma.o 1431 1424
math/*.o 64164 63635
72 lines
1.5 KiB
C
72 lines
1.5 KiB
C
/* origin: FreeBSD /usr/src/lib/msun/src/e_acosf.c */
|
|
/*
|
|
* Conversion to float by Ian Lance Taylor, Cygnus Support, ian@cygnus.com.
|
|
*/
|
|
/*
|
|
* ====================================================
|
|
* Copyright (C) 1993 by Sun Microsystems, Inc. All rights reserved.
|
|
*
|
|
* Developed at SunPro, a Sun Microsystems, Inc. business.
|
|
* Permission to use, copy, modify, and distribute this
|
|
* software is freely granted, provided that this notice
|
|
* is preserved.
|
|
* ====================================================
|
|
*/
|
|
|
|
#include "libm.h"
|
|
|
|
static const float
|
|
pio2_hi = 1.5707962513e+00, /* 0x3fc90fda */
|
|
pio2_lo = 7.5497894159e-08, /* 0x33a22168 */
|
|
pS0 = 1.6666586697e-01,
|
|
pS1 = -4.2743422091e-02,
|
|
pS2 = -8.6563630030e-03,
|
|
qS1 = -7.0662963390e-01;
|
|
|
|
static float R(float z)
|
|
{
|
|
float_t p, q;
|
|
p = z*(pS0+z*(pS1+z*pS2));
|
|
q = 1.0f+z*qS1;
|
|
return p/q;
|
|
}
|
|
|
|
float acosf(float x)
|
|
{
|
|
float z,w,s,c,df;
|
|
uint32_t hx,ix;
|
|
|
|
GET_FLOAT_WORD(hx, x);
|
|
ix = hx & 0x7fffffff;
|
|
/* |x| >= 1 or nan */
|
|
if (ix >= 0x3f800000) {
|
|
if (ix == 0x3f800000) {
|
|
if (hx >> 31)
|
|
return 2*pio2_hi + 0x1p-120f;
|
|
return 0;
|
|
}
|
|
return 0/(x-x);
|
|
}
|
|
/* |x| < 0.5 */
|
|
if (ix < 0x3f000000) {
|
|
if (ix <= 0x32800000) /* |x| < 2**-26 */
|
|
return pio2_hi + 0x1p-120f;
|
|
return pio2_hi - (x - (pio2_lo-x*R(x*x)));
|
|
}
|
|
/* x < -0.5 */
|
|
if (hx >> 31) {
|
|
z = (1+x)*0.5f;
|
|
s = sqrtf(z);
|
|
w = R(z)*s-pio2_lo;
|
|
return 2*(pio2_hi - (s+w));
|
|
}
|
|
/* x > 0.5 */
|
|
z = (1-x)*0.5f;
|
|
s = sqrtf(z);
|
|
GET_FLOAT_WORD(hx,s);
|
|
SET_FLOAT_WORD(df,hx&0xfffff000);
|
|
c = (z-df*df)/(s+df);
|
|
w = R(z)*s+c;
|
|
return 2*(df+w);
|
|
}
|