floatformat.c
14 KB
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
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
/* IEEE floating point support routines, for GDB, the GNU Debugger.
Copyright (C) 1991, 1994, 1999, 2000, 2003 Free Software Foundation, Inc.
This file is part of GDB.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
#include "ansidecl.h"
#include "floatformat.h"
#include <math.h> /* ldexp */
#ifdef ANSI_PROTOTYPES
#include <stddef.h>
extern void *memcpy (void *s1, const void *s2, size_t n);
extern void *memset (void *s, int c, size_t n);
#else
extern char *memcpy ();
extern char *memset ();
#endif
/* The odds that CHAR_BIT will be anything but 8 are low enough that I'm not
going to bother with trying to muck around with whether it is defined in
a system header, what we do if not, etc. */
#define FLOATFORMAT_CHAR_BIT 8
/* floatformats for IEEE single and double, big and little endian. */
const struct floatformat floatformat_ieee_single_big =
{
floatformat_big, 32, 0, 1, 8, 127, 255, 9, 23,
floatformat_intbit_no,
"floatformat_ieee_single_big"
};
const struct floatformat floatformat_ieee_single_little =
{
floatformat_little, 32, 0, 1, 8, 127, 255, 9, 23,
floatformat_intbit_no,
"floatformat_ieee_single_little"
};
const struct floatformat floatformat_ieee_double_big =
{
floatformat_big, 64, 0, 1, 11, 1023, 2047, 12, 52,
floatformat_intbit_no,
"floatformat_ieee_double_big"
};
const struct floatformat floatformat_ieee_double_little =
{
floatformat_little, 64, 0, 1, 11, 1023, 2047, 12, 52,
floatformat_intbit_no,
"floatformat_ieee_double_little"
};
/* floatformat for IEEE double, little endian byte order, with big endian word
ordering, as on the ARM. */
const struct floatformat floatformat_ieee_double_littlebyte_bigword =
{
floatformat_littlebyte_bigword, 64, 0, 1, 11, 1023, 2047, 12, 52,
floatformat_intbit_no,
"floatformat_ieee_double_littlebyte_bigword"
};
const struct floatformat floatformat_i387_ext =
{
floatformat_little, 80, 0, 1, 15, 0x3fff, 0x7fff, 16, 64,
floatformat_intbit_yes,
"floatformat_i387_ext"
};
const struct floatformat floatformat_m68881_ext =
{
/* Note that the bits from 16 to 31 are unused. */
floatformat_big, 96, 0, 1, 15, 0x3fff, 0x7fff, 32, 64,
floatformat_intbit_yes,
"floatformat_m68881_ext"
};
const struct floatformat floatformat_i960_ext =
{
/* Note that the bits from 0 to 15 are unused. */
floatformat_little, 96, 16, 17, 15, 0x3fff, 0x7fff, 32, 64,
floatformat_intbit_yes,
"floatformat_i960_ext"
};
const struct floatformat floatformat_m88110_ext =
{
floatformat_big, 80, 0, 1, 15, 0x3fff, 0x7fff, 16, 64,
floatformat_intbit_yes,
"floatformat_m88110_ext"
};
const struct floatformat floatformat_m88110_harris_ext =
{
/* Harris uses raw format 128 bytes long, but the number is just an ieee
double, and the last 64 bits are wasted. */
floatformat_big,128, 0, 1, 11, 0x3ff, 0x7ff, 12, 52,
floatformat_intbit_no,
"floatformat_m88110_ext_harris"
};
const struct floatformat floatformat_arm_ext_big =
{
/* Bits 1 to 16 are unused. */
floatformat_big, 96, 0, 17, 15, 0x3fff, 0x7fff, 32, 64,
floatformat_intbit_yes,
"floatformat_arm_ext_big"
};
const struct floatformat floatformat_arm_ext_littlebyte_bigword =
{
/* Bits 1 to 16 are unused. */
floatformat_littlebyte_bigword, 96, 0, 17, 15, 0x3fff, 0x7fff, 32, 64,
floatformat_intbit_yes,
"floatformat_arm_ext_littlebyte_bigword"
};
const struct floatformat floatformat_ia64_spill_big =
{
floatformat_big, 128, 0, 1, 17, 65535, 0x1ffff, 18, 64,
floatformat_intbit_yes,
"floatformat_ia64_spill_big"
};
const struct floatformat floatformat_ia64_spill_little =
{
floatformat_little, 128, 0, 1, 17, 65535, 0x1ffff, 18, 64,
floatformat_intbit_yes,
"floatformat_ia64_spill_little"
};
const struct floatformat floatformat_ia64_quad_big =
{
floatformat_big, 128, 0, 1, 15, 16383, 0x7fff, 16, 112,
floatformat_intbit_no,
"floatformat_ia64_quad_big"
};
const struct floatformat floatformat_ia64_quad_little =
{
floatformat_little, 128, 0, 1, 15, 16383, 0x7fff, 16, 112,
floatformat_intbit_no,
"floatformat_ia64_quad_little"
};
static unsigned long get_field PARAMS ((const unsigned char *,
enum floatformat_byteorders,
unsigned int,
unsigned int,
unsigned int));
/* Extract a field which starts at START and is LEN bits long. DATA and
TOTAL_LEN are the thing we are extracting it from, in byteorder ORDER. */
static unsigned long
get_field (data, order, total_len, start, len)
const unsigned char *data;
enum floatformat_byteorders order;
unsigned int total_len;
unsigned int start;
unsigned int len;
{
unsigned long result;
unsigned int cur_byte;
int cur_bitshift;
/* Start at the least significant part of the field. */
cur_byte = (start + len) / FLOATFORMAT_CHAR_BIT;
if (order == floatformat_little)
cur_byte = (total_len / FLOATFORMAT_CHAR_BIT) - cur_byte - 1;
cur_bitshift =
((start + len) % FLOATFORMAT_CHAR_BIT) - FLOATFORMAT_CHAR_BIT;
result = *(data + cur_byte) >> (-cur_bitshift);
cur_bitshift += FLOATFORMAT_CHAR_BIT;
if (order == floatformat_little)
++cur_byte;
else
--cur_byte;
/* Move towards the most significant part of the field. */
while ((unsigned int) cur_bitshift < len)
{
if (len - cur_bitshift < FLOATFORMAT_CHAR_BIT)
/* This is the last byte; zero out the bits which are not part of
this field. */
result |=
(*(data + cur_byte) & ((1 << (len - cur_bitshift)) - 1))
<< cur_bitshift;
else
result |= *(data + cur_byte) << cur_bitshift;
cur_bitshift += FLOATFORMAT_CHAR_BIT;
if (order == floatformat_little)
++cur_byte;
else
--cur_byte;
}
return result;
}
#ifndef min
#define min(a, b) ((a) < (b) ? (a) : (b))
#endif
/* Convert from FMT to a double.
FROM is the address of the extended float.
Store the double in *TO. */
void
floatformat_to_double (fmt, from, to)
const struct floatformat *fmt;
const char *from;
double *to;
{
const unsigned char *ufrom = (const unsigned char *)from;
double dto;
long exponent;
unsigned long mant;
unsigned int mant_bits, mant_off;
int mant_bits_left;
int special_exponent; /* It's a NaN, denorm or zero */
exponent = get_field (ufrom, fmt->byteorder, fmt->totalsize,
fmt->exp_start, fmt->exp_len);
/* Note that if exponent indicates a NaN, we can't really do anything useful
(not knowing if the host has NaN's, or how to build one). So it will
end up as an infinity or something close; that is OK. */
mant_bits_left = fmt->man_len;
mant_off = fmt->man_start;
dto = 0.0;
special_exponent = exponent == 0 || (unsigned long) exponent == fmt->exp_nan;
/* Don't bias zero's, denorms or NaNs. */
if (!special_exponent)
exponent -= fmt->exp_bias;
/* Build the result algebraically. Might go infinite, underflow, etc;
who cares. */
/* If this format uses a hidden bit, explicitly add it in now. Otherwise,
increment the exponent by one to account for the integer bit. */
if (!special_exponent)
{
if (fmt->intbit == floatformat_intbit_no)
dto = ldexp (1.0, exponent);
else
exponent++;
}
while (mant_bits_left > 0)
{
mant_bits = min (mant_bits_left, 32);
mant = get_field (ufrom, fmt->byteorder, fmt->totalsize,
mant_off, mant_bits);
dto += ldexp ((double)mant, exponent - mant_bits);
exponent -= mant_bits;
mant_off += mant_bits;
mant_bits_left -= mant_bits;
}
/* Negate it if negative. */
if (get_field (ufrom, fmt->byteorder, fmt->totalsize, fmt->sign_start, 1))
dto = -dto;
*to = dto;
}
static void put_field PARAMS ((unsigned char *, enum floatformat_byteorders,
unsigned int,
unsigned int,
unsigned int,
unsigned long));
/* Set a field which starts at START and is LEN bits long. DATA and
TOTAL_LEN are the thing we are extracting it from, in byteorder ORDER. */
static void
put_field (data, order, total_len, start, len, stuff_to_put)
unsigned char *data;
enum floatformat_byteorders order;
unsigned int total_len;
unsigned int start;
unsigned int len;
unsigned long stuff_to_put;
{
unsigned int cur_byte;
int cur_bitshift;
/* Start at the least significant part of the field. */
cur_byte = (start + len) / FLOATFORMAT_CHAR_BIT;
if (order == floatformat_little)
cur_byte = (total_len / FLOATFORMAT_CHAR_BIT) - cur_byte - 1;
cur_bitshift =
((start + len) % FLOATFORMAT_CHAR_BIT) - FLOATFORMAT_CHAR_BIT;
*(data + cur_byte) &=
~(((1 << ((start + len) % FLOATFORMAT_CHAR_BIT)) - 1) << (-cur_bitshift));
*(data + cur_byte) |=
(stuff_to_put & ((1 << FLOATFORMAT_CHAR_BIT) - 1)) << (-cur_bitshift);
cur_bitshift += FLOATFORMAT_CHAR_BIT;
if (order == floatformat_little)
++cur_byte;
else
--cur_byte;
/* Move towards the most significant part of the field. */
while ((unsigned int) cur_bitshift < len)
{
if (len - cur_bitshift < FLOATFORMAT_CHAR_BIT)
{
/* This is the last byte. */
*(data + cur_byte) &=
~((1 << (len - cur_bitshift)) - 1);
*(data + cur_byte) |= (stuff_to_put >> cur_bitshift);
}
else
*(data + cur_byte) = ((stuff_to_put >> cur_bitshift)
& ((1 << FLOATFORMAT_CHAR_BIT) - 1));
cur_bitshift += FLOATFORMAT_CHAR_BIT;
if (order == floatformat_little)
++cur_byte;
else
--cur_byte;
}
}
/* The converse: convert the double *FROM to an extended float
and store where TO points. Neither FROM nor TO have any alignment
restrictions. */
void
floatformat_from_double (fmt, from, to)
const struct floatformat *fmt;
const double *from;
char *to;
{
double dfrom;
int exponent;
double mant;
unsigned int mant_bits, mant_off;
int mant_bits_left;
unsigned char *uto = (unsigned char *)to;
memcpy (&dfrom, from, sizeof (dfrom));
memset (uto, 0, fmt->totalsize / FLOATFORMAT_CHAR_BIT);
if (dfrom == 0)
return; /* Result is zero */
if (dfrom != dfrom)
{
/* From is NaN */
put_field (uto, fmt->byteorder, fmt->totalsize, fmt->exp_start,
fmt->exp_len, fmt->exp_nan);
/* Be sure it's not infinity, but NaN value is irrel */
put_field (uto, fmt->byteorder, fmt->totalsize, fmt->man_start,
32, 1);
return;
}
/* If negative, set the sign bit. */
if (dfrom < 0)
{
put_field (uto, fmt->byteorder, fmt->totalsize, fmt->sign_start, 1, 1);
dfrom = -dfrom;
}
/* How to tell an infinity from an ordinary number? FIXME-someday */
mant = frexp (dfrom, &exponent);
put_field (uto, fmt->byteorder, fmt->totalsize, fmt->exp_start, fmt->exp_len,
exponent + fmt->exp_bias - 1);
mant_bits_left = fmt->man_len;
mant_off = fmt->man_start;
while (mant_bits_left > 0)
{
unsigned long mant_long;
mant_bits = mant_bits_left < 32 ? mant_bits_left : 32;
mant *= 4294967296.0;
mant_long = (unsigned long)mant;
mant -= mant_long;
/* If the integer bit is implicit, then we need to discard it.
If we are discarding a zero, we should be (but are not) creating
a denormalized number which means adjusting the exponent
(I think). */
if ((unsigned int) mant_bits_left == fmt->man_len
&& fmt->intbit == floatformat_intbit_no)
{
mant_long &= 0x7fffffff;
mant_bits -= 1;
}
else if (mant_bits < 32)
{
/* The bits we want are in the most significant MANT_BITS bits of
mant_long. Move them to the least significant. */
mant_long >>= 32 - mant_bits;
}
put_field (uto, fmt->byteorder, fmt->totalsize,
mant_off, mant_bits, mant_long);
mant_off += mant_bits;
mant_bits_left -= mant_bits;
}
}
/* Return non-zero iff the data at FROM is a valid number in format FMT. */
int
floatformat_is_valid (fmt, from)
const struct floatformat *fmt;
const char *from;
{
if (fmt == &floatformat_i387_ext)
{
/* In the i387 double-extended format, if the exponent is all
ones, then the integer bit must be set. If the exponent
is neither 0 nor ~0, the intbit must also be set. Only
if the exponent is zero can it be zero, and then it must
be zero. */
unsigned long exponent, int_bit;
const unsigned char *ufrom = (const unsigned char *) from;
exponent = get_field (ufrom, fmt->byteorder, fmt->totalsize,
fmt->exp_start, fmt->exp_len);
int_bit = get_field (ufrom, fmt->byteorder, fmt->totalsize,
fmt->man_start, 1);
if ((exponent == 0) != (int_bit == 0))
return 0;
else
return 1;
}
/* Other formats with invalid representations should be added
here. */
return 1;
}
#ifdef IEEE_DEBUG
/* This is to be run on a host which uses IEEE floating point. */
void
ieee_test (n)
double n;
{
double result;
char exten[16];
floatformat_to_double (&floatformat_ieee_double_big, &n, &result);
if (n != result)
printf ("Differ(to): %.20g -> %.20g\n", n, result);
floatformat_from_double (&floatformat_ieee_double_big, &n, &result);
if (n != result)
printf ("Differ(from): %.20g -> %.20g\n", n, result);
floatformat_from_double (&floatformat_m68881_ext, &n, exten);
floatformat_to_double (&floatformat_m68881_ext, exten, &result);
if (n != result)
printf ("Differ(to+from): %.20g -> %.20g\n", n, result);
#if IEEE_DEBUG > 1
/* This is to be run on a host which uses 68881 format. */
{
long double ex = *(long double *)exten;
if (ex != n)
printf ("Differ(from vs. extended): %.20g\n", n);
}
#endif
}
int
main ()
{
ieee_test (0.5);
ieee_test (256.0);
ieee_test (0.12345);
ieee_test (234235.78907234);
ieee_test (-512.0);
ieee_test (-0.004321);
return 0;
}
#endif