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
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
|
/*============================================================================
This C header file is part of the SoftFloat IEEE Floating-Point Arithmetic
Package, Release 3e, by John R. Hauser.
Copyright 2011, 2012, 2013, 2014, 2015, 2016, 2017 The Regents of the
University of California. All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
1. Redistributions of source code must retain the above copyright notice,
this list of conditions, and the following disclaimer.
2. 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.
3. Neither the name of the University 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 REGENTS 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 REGENTS 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.
=============================================================================*/
/*============================================================================
| Note: If SoftFloat is made available as a general library for programs to
| use, it is strongly recommended that a platform-specific version of this
| header, "softfloat.h", be created that folds in "softfloat_types.h" and that
| eliminates all dependencies on compile-time macros.
*============================================================================*/
#ifndef softfloat_h
#define softfloat_h 1
#include <stdbool.h>
#include <stdint.h>
#include "softfloat_types.h"
#ifndef THREAD_LOCAL
#define THREAD_LOCAL
#endif
/*----------------------------------------------------------------------------
| Software floating-point underflow tininess-detection mode.
*----------------------------------------------------------------------------*/
extern THREAD_LOCAL uint_fast8_t softfloat_detectTininess;
enum {
softfloat_tininess_beforeRounding = 0,
softfloat_tininess_afterRounding = 1
};
/*----------------------------------------------------------------------------
| Software floating-point rounding mode. (Mode "odd" is supported only if
| SoftFloat is compiled with macro 'SOFTFLOAT_ROUND_ODD' defined.)
*----------------------------------------------------------------------------*/
extern THREAD_LOCAL uint_fast8_t softfloat_roundingMode;
enum {
softfloat_round_near_even = 0,
softfloat_round_minMag = 1,
softfloat_round_min = 2,
softfloat_round_max = 3,
softfloat_round_near_maxMag = 4,
softfloat_round_odd = 6
};
/*----------------------------------------------------------------------------
| Software floating-point exception flags.
*----------------------------------------------------------------------------*/
extern THREAD_LOCAL uint_fast8_t softfloat_exceptionFlags;
typedef enum {
softfloat_flag_inexact = 1,
softfloat_flag_underflow = 2,
softfloat_flag_overflow = 4,
softfloat_flag_infinite = 8,
softfloat_flag_invalid = 16
} exceptionFlag_t;
/*----------------------------------------------------------------------------
| Routine to raise any or all of the software floating-point exception flags.
*----------------------------------------------------------------------------*/
void softfloat_raiseFlags( uint_fast8_t );
/*----------------------------------------------------------------------------
| Integer-to-floating-point conversion routines.
*----------------------------------------------------------------------------*/
float16_t ui32_to_f16( uint32_t );
float32_t ui32_to_f32( uint32_t );
float64_t ui32_to_f64( uint32_t );
#ifdef SOFTFLOAT_FAST_INT64
extFloat80_t ui32_to_extF80( uint32_t );
float128_t ui32_to_f128( uint32_t );
#endif
void ui32_to_extF80M( uint32_t, extFloat80_t * );
void ui32_to_f128M( uint32_t, float128_t * );
float16_t ui64_to_f16( uint64_t );
float32_t ui64_to_f32( uint64_t );
float64_t ui64_to_f64( uint64_t );
#ifdef SOFTFLOAT_FAST_INT64
extFloat80_t ui64_to_extF80( uint64_t );
float128_t ui64_to_f128( uint64_t );
#endif
void ui64_to_extF80M( uint64_t, extFloat80_t * );
void ui64_to_f128M( uint64_t, float128_t * );
float16_t i32_to_f16( int32_t );
float32_t i32_to_f32( int32_t );
float64_t i32_to_f64( int32_t );
#ifdef SOFTFLOAT_FAST_INT64
extFloat80_t i32_to_extF80( int32_t );
float128_t i32_to_f128( int32_t );
#endif
void i32_to_extF80M( int32_t, extFloat80_t * );
void i32_to_f128M( int32_t, float128_t * );
float16_t i64_to_f16( int64_t );
float32_t i64_to_f32( int64_t );
float64_t i64_to_f64( int64_t );
#ifdef SOFTFLOAT_FAST_INT64
extFloat80_t i64_to_extF80( int64_t );
float128_t i64_to_f128( int64_t );
#endif
void i64_to_extF80M( int64_t, extFloat80_t * );
void i64_to_f128M( int64_t, float128_t * );
/*----------------------------------------------------------------------------
| 16-bit (half-precision) floating-point operations.
*----------------------------------------------------------------------------*/
uint_fast32_t f16_to_ui32( float16_t, uint_fast8_t, bool );
uint_fast64_t f16_to_ui64( float16_t, uint_fast8_t, bool );
int_fast32_t f16_to_i32( float16_t, uint_fast8_t, bool );
int_fast64_t f16_to_i64( float16_t, uint_fast8_t, bool );
uint_fast32_t f16_to_ui32_r_minMag( float16_t, bool );
uint_fast64_t f16_to_ui64_r_minMag( float16_t, bool );
int_fast32_t f16_to_i32_r_minMag( float16_t, bool );
int_fast64_t f16_to_i64_r_minMag( float16_t, bool );
float32_t f16_to_f32( float16_t );
float64_t f16_to_f64( float16_t );
#ifdef SOFTFLOAT_FAST_INT64
extFloat80_t f16_to_extF80( float16_t );
float128_t f16_to_f128( float16_t );
#endif
void f16_to_extF80M( float16_t, extFloat80_t * );
void f16_to_f128M( float16_t, float128_t * );
float16_t f16_roundToInt( float16_t, uint_fast8_t, bool );
float16_t f16_add( float16_t, float16_t );
float16_t f16_sub( float16_t, float16_t );
float16_t f16_mul( float16_t, float16_t );
float16_t f16_mulAdd( float16_t, float16_t, float16_t );
float16_t f16_div( float16_t, float16_t );
float16_t f16_rem( float16_t, float16_t );
float16_t f16_sqrt( float16_t );
bool f16_eq( float16_t, float16_t );
bool f16_le( float16_t, float16_t );
bool f16_lt( float16_t, float16_t );
bool f16_eq_signaling( float16_t, float16_t );
bool f16_le_quiet( float16_t, float16_t );
bool f16_lt_quiet( float16_t, float16_t );
bool f16_isSignalingNaN( float16_t );
/*----------------------------------------------------------------------------
| 16-bit (brain float 16) floating-point operations.
*----------------------------------------------------------------------------*/
float32_t bf16_to_f32( bfloat16_t );
bfloat16_t f32_to_bf16( float32_t );
bool bf16_isSignalingNaN( bfloat16_t );
/*----------------------------------------------------------------------------
| 32-bit (single-precision) floating-point operations.
*----------------------------------------------------------------------------*/
uint_fast32_t f32_to_ui32( float32_t, uint_fast8_t, bool );
uint_fast64_t f32_to_ui64( float32_t, uint_fast8_t, bool );
int_fast32_t f32_to_i32( float32_t, uint_fast8_t, bool );
int_fast64_t f32_to_i64( float32_t, uint_fast8_t, bool );
uint_fast32_t f32_to_ui32_r_minMag( float32_t, bool );
uint_fast64_t f32_to_ui64_r_minMag( float32_t, bool );
int_fast32_t f32_to_i32_r_minMag( float32_t, bool );
int_fast64_t f32_to_i64_r_minMag( float32_t, bool );
float16_t f32_to_f16( float32_t );
float64_t f32_to_f64( float32_t );
#ifdef SOFTFLOAT_FAST_INT64
extFloat80_t f32_to_extF80( float32_t );
float128_t f32_to_f128( float32_t );
#endif
void f32_to_extF80M( float32_t, extFloat80_t * );
void f32_to_f128M( float32_t, float128_t * );
float32_t f32_roundToInt( float32_t, uint_fast8_t, bool );
float32_t f32_add( float32_t, float32_t );
float32_t f32_sub( float32_t, float32_t );
float32_t f32_mul( float32_t, float32_t );
float32_t f32_mulAdd( float32_t, float32_t, float32_t );
float32_t f32_div( float32_t, float32_t );
float32_t f32_rem( float32_t, float32_t );
float32_t f32_sqrt( float32_t );
bool f32_eq( float32_t, float32_t );
bool f32_le( float32_t, float32_t );
bool f32_lt( float32_t, float32_t );
bool f32_eq_signaling( float32_t, float32_t );
bool f32_le_quiet( float32_t, float32_t );
bool f32_lt_quiet( float32_t, float32_t );
bool f32_isSignalingNaN( float32_t );
/*----------------------------------------------------------------------------
| 64-bit (double-precision) floating-point operations.
*----------------------------------------------------------------------------*/
uint_fast32_t f64_to_ui32( float64_t, uint_fast8_t, bool );
uint_fast64_t f64_to_ui64( float64_t, uint_fast8_t, bool );
int_fast32_t f64_to_i32( float64_t, uint_fast8_t, bool );
int_fast64_t f64_to_i64( float64_t, uint_fast8_t, bool );
uint_fast32_t f64_to_ui32_r_minMag( float64_t, bool );
uint_fast64_t f64_to_ui64_r_minMag( float64_t, bool );
int_fast32_t f64_to_i32_r_minMag( float64_t, bool );
int_fast64_t f64_to_i64_r_minMag( float64_t, bool );
float16_t f64_to_f16( float64_t );
float32_t f64_to_f32( float64_t );
#ifdef SOFTFLOAT_FAST_INT64
extFloat80_t f64_to_extF80( float64_t );
float128_t f64_to_f128( float64_t );
#endif
void f64_to_extF80M( float64_t, extFloat80_t * );
void f64_to_f128M( float64_t, float128_t * );
float64_t f64_roundToInt( float64_t, uint_fast8_t, bool );
float64_t f64_add( float64_t, float64_t );
float64_t f64_sub( float64_t, float64_t );
float64_t f64_mul( float64_t, float64_t );
float64_t f64_mulAdd( float64_t, float64_t, float64_t );
float64_t f64_div( float64_t, float64_t );
float64_t f64_rem( float64_t, float64_t );
float64_t f64_sqrt( float64_t );
bool f64_eq( float64_t, float64_t );
bool f64_le( float64_t, float64_t );
bool f64_lt( float64_t, float64_t );
bool f64_eq_signaling( float64_t, float64_t );
bool f64_le_quiet( float64_t, float64_t );
bool f64_lt_quiet( float64_t, float64_t );
bool f64_isSignalingNaN( float64_t );
/*----------------------------------------------------------------------------
| Rounding precision for 80-bit extended double-precision floating-point.
| Valid values are 32, 64, and 80.
*----------------------------------------------------------------------------*/
extern THREAD_LOCAL uint_fast8_t extF80_roundingPrecision;
/*----------------------------------------------------------------------------
| 80-bit extended double-precision floating-point operations.
*----------------------------------------------------------------------------*/
#ifdef SOFTFLOAT_FAST_INT64
uint_fast32_t extF80_to_ui32( extFloat80_t, uint_fast8_t, bool );
uint_fast64_t extF80_to_ui64( extFloat80_t, uint_fast8_t, bool );
int_fast32_t extF80_to_i32( extFloat80_t, uint_fast8_t, bool );
int_fast64_t extF80_to_i64( extFloat80_t, uint_fast8_t, bool );
uint_fast32_t extF80_to_ui32_r_minMag( extFloat80_t, bool );
uint_fast64_t extF80_to_ui64_r_minMag( extFloat80_t, bool );
int_fast32_t extF80_to_i32_r_minMag( extFloat80_t, bool );
int_fast64_t extF80_to_i64_r_minMag( extFloat80_t, bool );
float16_t extF80_to_f16( extFloat80_t );
float32_t extF80_to_f32( extFloat80_t );
float64_t extF80_to_f64( extFloat80_t );
float128_t extF80_to_f128( extFloat80_t );
extFloat80_t extF80_roundToInt( extFloat80_t, uint_fast8_t, bool );
extFloat80_t extF80_add( extFloat80_t, extFloat80_t );
extFloat80_t extF80_sub( extFloat80_t, extFloat80_t );
extFloat80_t extF80_mul( extFloat80_t, extFloat80_t );
extFloat80_t extF80_div( extFloat80_t, extFloat80_t );
extFloat80_t extF80_rem( extFloat80_t, extFloat80_t );
extFloat80_t extF80_sqrt( extFloat80_t );
bool extF80_eq( extFloat80_t, extFloat80_t );
bool extF80_le( extFloat80_t, extFloat80_t );
bool extF80_lt( extFloat80_t, extFloat80_t );
bool extF80_eq_signaling( extFloat80_t, extFloat80_t );
bool extF80_le_quiet( extFloat80_t, extFloat80_t );
bool extF80_lt_quiet( extFloat80_t, extFloat80_t );
bool extF80_isSignalingNaN( extFloat80_t );
#endif
uint_fast32_t extF80M_to_ui32( const extFloat80_t *, uint_fast8_t, bool );
uint_fast64_t extF80M_to_ui64( const extFloat80_t *, uint_fast8_t, bool );
int_fast32_t extF80M_to_i32( const extFloat80_t *, uint_fast8_t, bool );
int_fast64_t extF80M_to_i64( const extFloat80_t *, uint_fast8_t, bool );
uint_fast32_t extF80M_to_ui32_r_minMag( const extFloat80_t *, bool );
uint_fast64_t extF80M_to_ui64_r_minMag( const extFloat80_t *, bool );
int_fast32_t extF80M_to_i32_r_minMag( const extFloat80_t *, bool );
int_fast64_t extF80M_to_i64_r_minMag( const extFloat80_t *, bool );
float16_t extF80M_to_f16( const extFloat80_t * );
float32_t extF80M_to_f32( const extFloat80_t * );
float64_t extF80M_to_f64( const extFloat80_t * );
void extF80M_to_f128M( const extFloat80_t *, float128_t * );
void
extF80M_roundToInt(
const extFloat80_t *, uint_fast8_t, bool, extFloat80_t * );
void extF80M_add( const extFloat80_t *, const extFloat80_t *, extFloat80_t * );
void extF80M_sub( const extFloat80_t *, const extFloat80_t *, extFloat80_t * );
void extF80M_mul( const extFloat80_t *, const extFloat80_t *, extFloat80_t * );
void extF80M_div( const extFloat80_t *, const extFloat80_t *, extFloat80_t * );
void extF80M_rem( const extFloat80_t *, const extFloat80_t *, extFloat80_t * );
void extF80M_sqrt( const extFloat80_t *, extFloat80_t * );
bool extF80M_eq( const extFloat80_t *, const extFloat80_t * );
bool extF80M_le( const extFloat80_t *, const extFloat80_t * );
bool extF80M_lt( const extFloat80_t *, const extFloat80_t * );
bool extF80M_eq_signaling( const extFloat80_t *, const extFloat80_t * );
bool extF80M_le_quiet( const extFloat80_t *, const extFloat80_t * );
bool extF80M_lt_quiet( const extFloat80_t *, const extFloat80_t * );
bool extF80M_isSignalingNaN( const extFloat80_t * );
/*----------------------------------------------------------------------------
| 128-bit (quadruple-precision) floating-point operations.
*----------------------------------------------------------------------------*/
#ifdef SOFTFLOAT_FAST_INT64
uint_fast32_t f128_to_ui32( float128_t, uint_fast8_t, bool );
uint_fast64_t f128_to_ui64( float128_t, uint_fast8_t, bool );
int_fast32_t f128_to_i32( float128_t, uint_fast8_t, bool );
int_fast64_t f128_to_i64( float128_t, uint_fast8_t, bool );
uint_fast32_t f128_to_ui32_r_minMag( float128_t, bool );
uint_fast64_t f128_to_ui64_r_minMag( float128_t, bool );
int_fast32_t f128_to_i32_r_minMag( float128_t, bool );
int_fast64_t f128_to_i64_r_minMag( float128_t, bool );
float16_t f128_to_f16( float128_t );
float32_t f128_to_f32( float128_t );
float64_t f128_to_f64( float128_t );
extFloat80_t f128_to_extF80( float128_t );
float128_t f128_roundToInt( float128_t, uint_fast8_t, bool );
float128_t f128_add( float128_t, float128_t );
float128_t f128_sub( float128_t, float128_t );
float128_t f128_mul( float128_t, float128_t );
float128_t f128_mulAdd( float128_t, float128_t, float128_t );
float128_t f128_div( float128_t, float128_t );
float128_t f128_rem( float128_t, float128_t );
float128_t f128_sqrt( float128_t );
bool f128_eq( float128_t, float128_t );
bool f128_le( float128_t, float128_t );
bool f128_lt( float128_t, float128_t );
bool f128_eq_signaling( float128_t, float128_t );
bool f128_le_quiet( float128_t, float128_t );
bool f128_lt_quiet( float128_t, float128_t );
bool f128_isSignalingNaN( float128_t );
#endif
uint_fast32_t f128M_to_ui32( const float128_t *, uint_fast8_t, bool );
uint_fast64_t f128M_to_ui64( const float128_t *, uint_fast8_t, bool );
int_fast32_t f128M_to_i32( const float128_t *, uint_fast8_t, bool );
int_fast64_t f128M_to_i64( const float128_t *, uint_fast8_t, bool );
uint_fast32_t f128M_to_ui32_r_minMag( const float128_t *, bool );
uint_fast64_t f128M_to_ui64_r_minMag( const float128_t *, bool );
int_fast32_t f128M_to_i32_r_minMag( const float128_t *, bool );
int_fast64_t f128M_to_i64_r_minMag( const float128_t *, bool );
float16_t f128M_to_f16( const float128_t * );
float32_t f128M_to_f32( const float128_t * );
float64_t f128M_to_f64( const float128_t * );
void f128M_to_extF80M( const float128_t *, extFloat80_t * );
void f128M_roundToInt( const float128_t *, uint_fast8_t, bool, float128_t * );
void f128M_add( const float128_t *, const float128_t *, float128_t * );
void f128M_sub( const float128_t *, const float128_t *, float128_t * );
void f128M_mul( const float128_t *, const float128_t *, float128_t * );
void
f128M_mulAdd(
const float128_t *, const float128_t *, const float128_t *, float128_t *
);
void f128M_div( const float128_t *, const float128_t *, float128_t * );
void f128M_rem( const float128_t *, const float128_t *, float128_t * );
void f128M_sqrt( const float128_t *, float128_t * );
bool f128M_eq( const float128_t *, const float128_t * );
bool f128M_le( const float128_t *, const float128_t * );
bool f128M_lt( const float128_t *, const float128_t * );
bool f128M_eq_signaling( const float128_t *, const float128_t * );
bool f128M_le_quiet( const float128_t *, const float128_t * );
bool f128M_lt_quiet( const float128_t *, const float128_t * );
bool f128M_isSignalingNaN( const float128_t * );
#endif
|