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
|
/*
* Copyright (c) 2017-2018, NVIDIA CORPORATION. All rights reserved.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
*/
#include <math.h>
#include "fslog_defs.h"
#define FMAF __builtin_fmaf
extern "C" float __fss_log_fma3(float);
//inline float itf(int a)
//{
// return *reinterpret_cast<float*>(&a);
//}
//inline int fti(float a)
//{
// return *reinterpret_cast<int*>(&a);
//}
float __fss_log_fma3(float a_input)
{
unsigned const canonical_nan = CANONICAL_NAN;
unsigned const minus_inf = MINUS_INF;
unsigned exp_offset = EXP_OFFSET;
unsigned a_input_as_uint = *(unsigned*)&a_input;
if (a_input < 0.0f) {
return *(float*)&canonical_nan;;
}
if ( (a_input_as_uint & NAN_INF_MASK) == NAN_INF_MASK ) {
return a_input + a_input;
}
if (a_input == 0.0f) {
return *(float*)&minus_inf;
}
if (a_input < TWO_TO_M126_F) {
a_input *= TWO_TO_24_F;
exp_offset += U24;
}
unsigned a_as_uint = *(unsigned*)&a_input;
int e_int = (a_as_uint>>23) - exp_offset;
float e = (float)e_int;
unsigned m_as_uint = (a_as_uint & BIT_MASK2) + OFFSET;
float m = *(float*)&m_as_uint;
if (m < PARTITION_CONST) {
m = m + m;
e = e - 1.0f;
}
m = m - 1.0f;
float LN2_0 = 0x1.62E43p-01;
float exp = e * LN2_0;
float t = LOG_CA;
t = FMAF(t, m, LOG_C9);
t = FMAF(t, m, LOG_C8);
t = FMAF(t, m, LOG_C7);
t = FMAF(t, m, LOG_C6);
t = FMAF(t, m, LOG_C5);
t = FMAF(t, m, LOG_C4);
t = FMAF(t, m, LOG_C3);
t = FMAF(t, m, LOG_C2);
t = FMAF(t, m, LOG_C1);
float m2 = m * m;
t = FMAF(t, m2, m);
t = t + exp;
return t;
}
|
