| 1 |
207 |
jeremybenn |
/* -------------------------------------------------------------- */
|
| 2 |
|
|
/* (C)Copyright 2001,2008, */
|
| 3 |
|
|
/* International Business Machines Corporation, */
|
| 4 |
|
|
/* Sony Computer Entertainment, Incorporated, */
|
| 5 |
|
|
/* Toshiba Corporation, */
|
| 6 |
|
|
/* */
|
| 7 |
|
|
/* All Rights Reserved. */
|
| 8 |
|
|
/* */
|
| 9 |
|
|
/* Redistribution and use in source and binary forms, with or */
|
| 10 |
|
|
/* without modification, are permitted provided that the */
|
| 11 |
|
|
/* following conditions are met: */
|
| 12 |
|
|
/* */
|
| 13 |
|
|
/* - Redistributions of source code must retain the above copyright*/
|
| 14 |
|
|
/* notice, this list of conditions and the following disclaimer. */
|
| 15 |
|
|
/* */
|
| 16 |
|
|
/* - Redistributions in binary form must reproduce the above */
|
| 17 |
|
|
/* copyright notice, this list of conditions and the following */
|
| 18 |
|
|
/* disclaimer in the documentation and/or other materials */
|
| 19 |
|
|
/* provided with the distribution. */
|
| 20 |
|
|
/* */
|
| 21 |
|
|
/* - Neither the name of IBM Corporation nor the names of its */
|
| 22 |
|
|
/* contributors may be used to endorse or promote products */
|
| 23 |
|
|
/* derived from this software without specific prior written */
|
| 24 |
|
|
/* permission. */
|
| 25 |
|
|
/* */
|
| 26 |
|
|
/* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND */
|
| 27 |
|
|
/* CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, */
|
| 28 |
|
|
/* INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF */
|
| 29 |
|
|
/* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE */
|
| 30 |
|
|
/* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR */
|
| 31 |
|
|
/* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, */
|
| 32 |
|
|
/* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT */
|
| 33 |
|
|
/* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; */
|
| 34 |
|
|
/* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) */
|
| 35 |
|
|
/* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN */
|
| 36 |
|
|
/* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR */
|
| 37 |
|
|
/* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, */
|
| 38 |
|
|
/* EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */
|
| 39 |
|
|
/* -------------------------------------------------------------- */
|
| 40 |
|
|
/* PROLOG END TAG zYx */
|
| 41 |
|
|
#ifdef __SPU__
|
| 42 |
|
|
#ifndef _EXPF4_H_
|
| 43 |
|
|
#define _EXPF4_H_ 1
|
| 44 |
|
|
|
| 45 |
|
|
|
| 46 |
|
|
#include "floorf4.h"
|
| 47 |
|
|
#include "ldexpf4.h"
|
| 48 |
|
|
|
| 49 |
|
|
/*
|
| 50 |
|
|
* FUNCTION
|
| 51 |
|
|
* vector float _expf4(vector float x)
|
| 52 |
|
|
*
|
| 53 |
|
|
* DESCRIPTION
|
| 54 |
|
|
* The _expf4 function computes e raised to the input x for
|
| 55 |
|
|
* each of the element of the float vector.
|
| 56 |
|
|
*
|
| 57 |
|
|
*/
|
| 58 |
|
|
static __inline vector float _expf4(vector float x)
|
| 59 |
|
|
{
|
| 60 |
|
|
|
| 61 |
|
|
// log2(e)
|
| 62 |
|
|
vec_float4 log2e = spu_splats(1.4426950408889634074f);
|
| 63 |
|
|
|
| 64 |
|
|
// Extra precision for the ln2 multiply
|
| 65 |
|
|
vec_float4 ln2_hi = spu_splats(0.693359375f);
|
| 66 |
|
|
vec_float4 ln2_lo = spu_splats(-2.12194440E-4f);
|
| 67 |
|
|
|
| 68 |
|
|
// Coefficents for the Taylor series
|
| 69 |
|
|
vec_float4 f02 = spu_splats(5.0000000000000000E-1f); // 1/2!
|
| 70 |
|
|
vec_float4 f03 = spu_splats(1.6666666666666667E-1f); // 1/3!
|
| 71 |
|
|
vec_float4 f04 = spu_splats(4.1666666666666667E-2f); // 1/4!
|
| 72 |
|
|
vec_float4 f05 = spu_splats(8.3333333333333333E-3f); // 1/5!
|
| 73 |
|
|
vec_float4 f06 = spu_splats(1.3888888888888889E-3f); // 1/6!
|
| 74 |
|
|
vec_float4 f07 = spu_splats(1.9841269841269841E-4f); // 1/7!
|
| 75 |
|
|
|
| 76 |
|
|
// Range reduce input, so that:
|
| 77 |
|
|
// e^x = e^z * 2^n
|
| 78 |
|
|
// e^x = e^z * e^(n * ln(2))
|
| 79 |
|
|
// e^x = e^(z + (n * ln(2)))
|
| 80 |
|
|
|
| 81 |
|
|
vec_int4 n; // exponent of reduction
|
| 82 |
|
|
vec_float4 q; // range reduced result
|
| 83 |
|
|
|
| 84 |
|
|
vec_float4 z;
|
| 85 |
|
|
vec_float4 r;
|
| 86 |
|
|
|
| 87 |
|
|
z = spu_madd(x,log2e,spu_splats(0.5f));
|
| 88 |
|
|
z = _floorf4(z);
|
| 89 |
|
|
r = spu_nmsub(z,ln2_hi,x);
|
| 90 |
|
|
r = spu_nmsub(z,ln2_lo,r);
|
| 91 |
|
|
n = spu_convts(z,0);
|
| 92 |
|
|
z = spu_mul(r,r);
|
| 93 |
|
|
|
| 94 |
|
|
// Use Horner's method on the Taylor series
|
| 95 |
|
|
q = spu_madd(r,f07,f06);
|
| 96 |
|
|
q = spu_madd(q,r,f05);
|
| 97 |
|
|
q = spu_madd(q,r,f04);
|
| 98 |
|
|
q = spu_madd(q,r,f03);
|
| 99 |
|
|
q = spu_madd(q,r,f02);
|
| 100 |
|
|
q = spu_madd(q,z,r);
|
| 101 |
|
|
q = spu_add(q,spu_splats(1.0f));
|
| 102 |
|
|
|
| 103 |
|
|
// Adjust the result by the range reduction
|
| 104 |
|
|
r = _ldexpf4( q, n );
|
| 105 |
|
|
|
| 106 |
|
|
return(r);
|
| 107 |
|
|
|
| 108 |
|
|
}
|
| 109 |
|
|
|
| 110 |
|
|
#endif /* _EXPF4_H_ */
|
| 111 |
|
|
#endif /* __SPU__ */
|
| 112 |
|
|
|
