newlib-cygwin/newlib/libm/machine/spu/headers/coshd2.h

116 lines
5.1 KiB
C

/* -------------------------------------------------------------- */
/* (C)Copyright 2006,2008, */
/* International Business Machines Corporation */
/* All Rights Reserved. */
/* */
/* Redistribution and use in source and binary forms, with or */
/* without modification, are permitted provided that the */
/* following conditions are met: */
/* */
/* - Redistributions of source code must retain the above copyright*/
/* notice, this list of conditions and the following disclaimer. */
/* */
/* - Redistributions in binary form must reproduce the above */
/* copyright notice, this list of conditions and the following */
/* disclaimer in the documentation and/or other materials */
/* provided with the distribution. */
/* */
/* - Neither the name of IBM Corporation nor the names of its */
/* contributors may be used to endorse or promote products */
/* derived from this software without specific prior written */
/* permission. */
/* */
/* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND */
/* CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, */
/* INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF */
/* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE */
/* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR */
/* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, */
/* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT */
/* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; */
/* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) */
/* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN */
/* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR */
/* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, */
/* EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */
/* -------------------------------------------------------------- */
/* PROLOG END TAG zYx */
#ifdef __SPU__
#ifndef _COSHD2_H_
#define _COSHD2_H_ 1
#include <spu_intrinsics.h>
#include "expd2.h"
#include "recipd2.h"
/*
* FUNCTION
* vector float _coshd2(vector double angle)
*
* DESCRIPTION
* _coshd2 computes the hyperbolic cosines of a vector of angles
* (expressed in radians) to an accuracy of a double precision
* floating point.
*/
static __inline vector double _coshd2(vector double x)
{
// Coefficents for the power series
vec_double2 f02 = spu_splats(5.00000000000000000000E-1); // 1/(2!)
vec_double2 f04 = spu_splats(4.16666666666666666667E-2); // 1/(4!)
vec_double2 f06 = spu_splats(1.38888888888888888889E-3); // 1/(6!)
vec_double2 f08 = spu_splats(2.48015873015873015873E-5); // 1/(8!)
vec_double2 f10 = spu_splats(2.75573192239858906526E-7); // 1/(10!)
vec_double2 f12 = spu_splats(2.08767569878680989792E-9); // 1/(12!)
vec_double2 f14 = spu_splats(1.14707455977297247139E-11); // 1/(14!)
vec_double2 f16 = spu_splats(4.77947733238738529744E-14); // 1/(16!)
vec_double2 f18 = spu_splats(1.56192069685862264622E-16); // 1/(18!)
vec_double2 f20 = spu_splats(4.11031762331216485848E-19); // 1/(20!)
vec_double2 f22 = spu_splats(8.89679139245057328675E-22); // 1/(22!)
// Check if the input is within the range [ -1.0 ... 1.0 ]
// If it is, we want to use the power series, otherwise
// we want to use the 0.5 * (e^x + e^-x)
// round to float, check if within range. Results will be in
// slots 0 and 2, so we rotate right 4 bytes, and "or" with ourself
// to produce 64 bits of all 1's or 0's.
vec_uint4 use_exp = spu_cmpabsgt(spu_roundtf(x),spu_splats(1.0f));
use_exp = spu_or(use_exp,spu_rlmaskqwbyte(use_exp,-4));
// Perform the calculation of the power series using Horner's method
vec_double2 result;
vec_double2 x2 = spu_mul(x,x);
result = spu_madd(x2,f22,f20);
result = spu_madd(x2,result,f18);
result = spu_madd(x2,result,f16);
result = spu_madd(x2,result,f14);
result = spu_madd(x2,result,f12);
result = spu_madd(x2,result,f10);
result = spu_madd(x2,result,f08);
result = spu_madd(x2,result,f06);
result = spu_madd(x2,result,f04);
result = spu_madd(x2,result,f02);
result = spu_madd(x2,result,spu_splats(1.0));
// Perform calculation as a function of 0.5 * (e^x + e^-x)
vec_double2 ex = _expd2(x);
vec_double2 ex_inv = _recipd2(ex);
vec_double2 r2= spu_add(ex,ex_inv);
r2 = spu_mul(r2,f02); // we can reuse f02 here
// Select either the power series or exp version
result = spu_sel(result,r2,(vec_ullong2)use_exp);
return result;
}
#endif /* _COSHD2_H_ */
#endif /* __SPU__ */