1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
|
/* -------------------------------------------------------------- */
/* (C)Copyright 2006,2007, */
/* International Business Machines Corporation, */
/* Sony Computer Entertainment, Incorporated, */
/* Toshiba 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. */
/* 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 _EXP2D2_H_
#define _EXP2D2_H_ 1
#include <spu_intrinsics.h>
/*
* FUNCTION
* vector double _exp2d2(vector double x)
*
* DESCRIPTION
* _exp2d2 computes 2 raised to the input x for each
* of the double word elements of x. Computation is
* performed by observing the 2^(a+b) = 2^a * 2^b.
* We decompose x into a and b (above) by letting.
* a = ceil(x), b = x - a;
*
* 2^a is easily computed by placing a into the exponent
* or a floating point number whose mantissa is all zeros.
*
* 2^b is computed using the polynomial approximation.
*
* __13_
* \
* \
* 2^x = / Ci*x^i
* /____
* i=0
*
* for x in the range 0.0 to 1.0.
*
*/
#define EXP_C00 1.0
#define EXP_C01 6.93147180559945286227e-01
#define EXP_C02 2.40226506959100694072e-01
#define EXP_C03 5.55041086648215761801e-02
#define EXP_C04 9.61812910762847687873e-03
#define EXP_C05 1.33335581464284411157e-03
#define EXP_C06 1.54035303933816060656e-04
#define EXP_C07 1.52527338040598376946e-05
#define EXP_C08 1.32154867901443052734e-06
#define EXP_C09 1.01780860092396959520e-07
#define EXP_C10 7.05491162080112087744e-09
#define EXP_C11 4.44553827187081007394e-10
#define EXP_C12 2.56784359934881958182e-11
#define EXP_C13 1.36914888539041240648e-12
static __inline vector double _exp2d2(vector double vx)
{
vec_int4 ix, exp;
vec_uint4 overflow, underflow;
vec_float4 vxf;
vec_double2 p1, p2, x2, x4, x8;
vec_double2 vy, vxw, out_of_range;
/* Compute: vxw = x - ceil(x)
*/
vxw = spu_add(vx, spu_splats(0.5));
vxf = spu_roundtf(vxw);
ix = spu_convts(vxf, 0);
ix = spu_add(ix, (vec_int4)spu_andc(spu_cmpgt(spu_splats(0.0f), vxf), spu_cmpeq(ix, spu_splats((int)0x80000000))));
vxf = spu_convtf(ix, 0);
vxw = spu_sub(vx, spu_extend(vxf));
/* Detect overflow and underflow. If overflow, force the result
* to infinity (at the end).
*/
exp = spu_shuffle(ix, ix, ((vec_uchar16) { 0,1,2,3, 0,1,2,3, 8,9,10,11, 8,9,10,11 }));
overflow = spu_cmpgt(exp, 1023);
underflow = spu_cmpgt(exp, -1023);
out_of_range = (vec_double2)spu_and(overflow, ((vec_uint4) { 0x7FF00000, 0, 0x7FF00000, 0 }));
/* Calculate the result by evaluating the 13th order polynomial.
* For efficiency, the polynomial is broken into two parts and
* evaluate then using nested
*
* result = (((((c13*x + c12)*x + c11)*x + c10)*x + c9)*x + c8)*x^8 +
* ((((((c7*x + c6)*x + c5)*x + c4)*x + c3)*x + c2)*x + c1)*x + c0
*/
p2 = spu_madd(spu_splats(EXP_C07), vxw, spu_splats(EXP_C06));
p1 = spu_madd(spu_splats(EXP_C13), vxw, spu_splats(EXP_C12));
x2 = spu_mul(vxw, vxw);
p2 = spu_madd(vxw, p2, spu_splats(EXP_C05));
p1 = spu_madd(vxw, p1, spu_splats(EXP_C11));
x4 = spu_mul(x2, x2);
p2 = spu_madd(vxw, p2, spu_splats(EXP_C04));
p1 = spu_madd(vxw, p1, spu_splats(EXP_C10));
p2 = spu_madd(vxw, p2, spu_splats(EXP_C03));
p1 = spu_madd(vxw, p1, spu_splats(EXP_C09));
x8 = spu_mul(x4, x4);
p2 = spu_madd(vxw, p2, spu_splats(EXP_C02));
p1 = spu_madd(vxw, p1, spu_splats(EXP_C08));
p2 = spu_madd(vxw, p2, spu_splats(EXP_C01));
p2 = spu_madd(vxw, p2, spu_splats(EXP_C00));
vy = spu_madd(x8, p1, p2);
/* Align the integer integer portion of x with the exponent.
*/
ix = spu_sl(ix, ((vec_uint4) { 20, 32, 20, 32 }));
vy = (vec_double2)spu_add((vec_int4)vy, ix);
/* Select the result if not overflow or underflow. Otherwise select the
* the out of range value.
*/
return (spu_sel(vy, out_of_range, (vec_ullong2)spu_orc(overflow, underflow)));
}
#endif /* _EXP2D2_H_ */
#endif /* __SPU__ */
|
