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
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
|
/*
* Copyright 2017 Advanced Micro Devices, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
* Authors: AMD
*
*/
#include "dm_services.h"
#include "custom_float.h"
static bool build_custom_float(
struct fixed31_32 value,
const struct custom_float_format *format,
bool *negative,
uint32_t *mantissa,
uint32_t *exponenta)
{
uint32_t exp_offset = (1 << (format->exponenta_bits - 1)) - 1;
const struct fixed31_32 mantissa_constant_plus_max_fraction =
dc_fixpt_from_fraction(
(1LL << (format->mantissa_bits + 1)) - 1,
1LL << format->mantissa_bits);
struct fixed31_32 mantiss;
if (dc_fixpt_eq(
value,
dc_fixpt_zero)) {
*negative = false;
*mantissa = 0;
*exponenta = 0;
return true;
}
if (dc_fixpt_lt(
value,
dc_fixpt_zero)) {
*negative = format->sign;
value = dc_fixpt_neg(value);
} else {
*negative = false;
}
if (dc_fixpt_lt(
value,
dc_fixpt_one)) {
uint32_t i = 1;
do {
value = dc_fixpt_shl(value, 1);
++i;
} while (dc_fixpt_lt(
value,
dc_fixpt_one));
--i;
if (exp_offset <= i) {
*mantissa = 0;
*exponenta = 0;
return true;
}
*exponenta = exp_offset - i;
} else if (dc_fixpt_le(
mantissa_constant_plus_max_fraction,
value)) {
uint32_t i = 1;
do {
value = dc_fixpt_shr(value, 1);
++i;
} while (dc_fixpt_lt(
mantissa_constant_plus_max_fraction,
value));
*exponenta = exp_offset + i - 1;
} else {
*exponenta = exp_offset;
}
mantiss = dc_fixpt_sub(
value,
dc_fixpt_one);
if (dc_fixpt_lt(
mantiss,
dc_fixpt_zero) ||
dc_fixpt_lt(
dc_fixpt_one,
mantiss))
mantiss = dc_fixpt_zero;
else
mantiss = dc_fixpt_shl(
mantiss,
format->mantissa_bits);
*mantissa = dc_fixpt_floor(mantiss);
return true;
}
static bool setup_custom_float(
const struct custom_float_format *format,
bool negative,
uint32_t mantissa,
uint32_t exponenta,
uint32_t *result)
{
uint32_t i = 0;
uint32_t j = 0;
uint32_t value = 0;
/* verification code:
* once calculation is ok we can remove it
*/
const uint32_t mantissa_mask =
(1 << (format->mantissa_bits + 1)) - 1;
const uint32_t exponenta_mask =
(1 << (format->exponenta_bits + 1)) - 1;
if (mantissa & ~mantissa_mask) {
BREAK_TO_DEBUGGER();
mantissa = mantissa_mask;
}
if (exponenta & ~exponenta_mask) {
BREAK_TO_DEBUGGER();
exponenta = exponenta_mask;
}
/* end of verification code */
while (i < format->mantissa_bits) {
uint32_t mask = 1 << i;
if (mantissa & mask)
value |= mask;
++i;
}
while (j < format->exponenta_bits) {
uint32_t mask = 1 << j;
if (exponenta & mask)
value |= mask << i;
++j;
}
if (negative && format->sign)
value |= 1 << (i + j);
*result = value;
return true;
}
bool convert_to_custom_float_format(
struct fixed31_32 value,
const struct custom_float_format *format,
uint32_t *result)
{
uint32_t mantissa;
uint32_t exponenta;
bool negative;
return build_custom_float(
value, format, &negative, &mantissa, &exponenta) &&
setup_custom_float(
format, negative, mantissa, exponenta, result);
}
|
