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
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
|
// license:BSD-3-Clause
// copyright-holders:Aaron Giles
/***************************************************************************
eminline.h
Definitions for inline functions that can be overridden by OSD-
specific code.
***************************************************************************/
#ifndef MAME_OSD_EMINLINE_H
#define MAME_OSD_EMINLINE_H
#pragma once
#include "osdcomm.h"
#include "osdcore.h"
#if !defined(MAME_NOASM)
#if defined(__GNUC__)
#if defined(__i386__) || defined(__x86_64__)
#include "eigccx86.h"
#elif defined(__ppc__) || defined (__PPC__) || defined(__ppc64__) || defined(__PPC64__)
#include "eigccppc.h"
#elif defined(__arm__) || defined(__aarch64__)
#include "eigccarm.h"
#endif
#include "eigcc.h"
#elif defined(_MSC_VER)
#if defined(_M_IX86) || defined(_M_X64)
#include "eivcx86.h"
#elif defined(_M_ARM) || defined(_M_ARM64)
#include "eivcarm.h"
#endif
#include "eivc.h"
#endif
#endif // !defined(MAME_NOASM)
/***************************************************************************
INLINE MATH FUNCTIONS
***************************************************************************/
/*-------------------------------------------------
mul_32x32 - perform a signed 32 bit x 32 bit
multiply and return the full 64 bit result
-------------------------------------------------*/
#ifndef mul_32x32
constexpr int64_t mul_32x32(int32_t a, int32_t b)
{
return int64_t(a) * int64_t(b);
}
#endif
/*-------------------------------------------------
mulu_32x32 - perform an unsigned 32 bit x
32 bit multiply and return the full 64 bit
result
-------------------------------------------------*/
#ifndef mulu_32x32
constexpr uint64_t mulu_32x32(uint32_t a, uint32_t b)
{
return uint64_t(a) * uint64_t(b);
}
#endif
/*-------------------------------------------------
mul_32x32_hi - perform a signed 32 bit x 32 bit
multiply and return the upper 32 bits of the
result
-------------------------------------------------*/
#ifndef mul_32x32_hi
constexpr int32_t mul_32x32_hi(int32_t a, int32_t b)
{
return uint32_t((int64_t(a) * int64_t(b)) >> 32);
}
#endif
/*-------------------------------------------------
mulu_32x32_hi - perform an unsigned 32 bit x
32 bit multiply and return the upper 32 bits
of the result
-------------------------------------------------*/
#ifndef mulu_32x32_hi
constexpr uint32_t mulu_32x32_hi(uint32_t a, uint32_t b)
{
return uint32_t((uint64_t(a) * uint64_t(b)) >> 32);
}
#endif
/*-------------------------------------------------
mul_32x32_shift - perform a signed 32 bit x
32 bit multiply and shift the result by the
given number of bits before truncating the
result to 32 bits
-------------------------------------------------*/
#ifndef mul_32x32_shift
constexpr int32_t mul_32x32_shift(int32_t a, int32_t b, uint8_t shift)
{
return int32_t((int64_t(a) * int64_t(b)) >> shift);
}
#endif
/*-------------------------------------------------
mulu_32x32_shift - perform an unsigned 32 bit x
32 bit multiply and shift the result by the
given number of bits before truncating the
result to 32 bits
-------------------------------------------------*/
#ifndef mulu_32x32_shift
constexpr uint32_t mulu_32x32_shift(uint32_t a, uint32_t b, uint8_t shift)
{
return uint32_t((uint64_t(a) * uint64_t(b)) >> shift);
}
#endif
/*-------------------------------------------------
div_64x32 - perform a signed 64 bit x 32 bit
divide and return the 32 bit quotient
-------------------------------------------------*/
#ifndef div_64x32
constexpr int32_t div_64x32(int64_t a, int32_t b)
{
return a / int64_t(b);
}
#endif
/*-------------------------------------------------
divu_64x32 - perform an unsigned 64 bit x 32 bit
divide and return the 32 bit quotient
-------------------------------------------------*/
#ifndef divu_64x32
constexpr uint32_t divu_64x32(uint64_t a, uint32_t b)
{
return a / uint64_t(b);
}
#endif
/*-------------------------------------------------
div_64x32_rem - perform a signed 64 bit x 32
bit divide and return the 32 bit quotient and
32 bit remainder
-------------------------------------------------*/
#ifndef div_64x32_rem
inline int32_t div_64x32_rem(int64_t a, int32_t b, int32_t &remainder)
{
int32_t const res(div_64x32(a, b));
remainder = a - (int64_t(b) * res);
return res;
}
#endif
/*-------------------------------------------------
divu_64x32_rem - perform an unsigned 64 bit x
32 bit divide and return the 32 bit quotient
and 32 bit remainder
-------------------------------------------------*/
#ifndef divu_64x32_rem
inline uint32_t divu_64x32_rem(uint64_t a, uint32_t b, uint32_t &remainder)
{
uint32_t const res(divu_64x32(a, b));
remainder = a - (uint64_t(b) * res);
return res;
}
#endif
/*-------------------------------------------------
div_32x32_shift - perform a signed divide of
two 32 bit values, shifting the first before
division, and returning the 32 bit quotient
-------------------------------------------------*/
#ifndef div_32x32_shift
constexpr int32_t div_32x32_shift(int32_t a, int32_t b, uint8_t shift)
{
return (int64_t(a) << shift) / int64_t(b);
}
#endif
/*-------------------------------------------------
divu_32x32_shift - perform an unsigned divide of
two 32 bit values, shifting the first before
division, and returning the 32 bit quotient
-------------------------------------------------*/
#ifndef divu_32x32_shift
constexpr uint32_t divu_32x32_shift(uint32_t a, uint32_t b, uint8_t shift)
{
return (uint64_t(a) << shift) / uint64_t(b);
}
#endif
/*-------------------------------------------------
mod_64x32 - perform a signed 64 bit x 32 bit
divide and return the 32 bit remainder
-------------------------------------------------*/
#ifndef mod_64x32
constexpr int32_t mod_64x32(int64_t a, int32_t b)
{
return a - (b * div_64x32(a, b));
}
#endif
/*-------------------------------------------------
modu_64x32 - perform an unsigned 64 bit x 32 bit
divide and return the 32 bit remainder
-------------------------------------------------*/
#ifndef modu_64x32
constexpr uint32_t modu_64x32(uint64_t a, uint32_t b)
{
return a - (b * divu_64x32(a, b));
}
#endif
/*-------------------------------------------------
recip_approx - compute an approximate floating
point reciprocal
-------------------------------------------------*/
#ifndef recip_approx
constexpr float recip_approx(float value)
{
return 1.0f / value;
}
#endif
/*-------------------------------------------------
mul_64x64 - perform a signed 64 bit x 64 bit
multiply and return the full 128 bit result
-------------------------------------------------*/
#ifndef mul_64x64
inline int64_t mul_64x64(int64_t a, int64_t b, int64_t &hi)
{
uint64_t const a_hi = uint64_t(a) >> 32;
uint64_t const b_hi = uint64_t(b) >> 32;
uint64_t const a_lo = uint32_t(uint64_t(a));
uint64_t const b_lo = uint32_t(uint64_t(b));
uint64_t const ab_lo = a_lo * b_lo;
uint64_t const ab_m1 = a_hi * b_lo;
uint64_t const ab_m2 = a_lo * b_hi;
uint64_t const ab_hi = a_hi * b_hi;
uint64_t const carry = ((ab_lo >> 32) + uint32_t(ab_m1) + uint32_t(ab_m2)) >> 32;
hi = ab_hi + (ab_m1 >> 32) + (ab_m2 >> 32) + carry;
// adjust for sign
if (a < 0)
hi -= b;
if (b < 0)
hi -= a;
return ab_lo + (ab_m1 << 32) + (ab_m2 << 32);
}
#endif
/*-------------------------------------------------
mulu_64x64 - perform an unsigned 64 bit x 64
bit multiply and return the full 128 bit result
-------------------------------------------------*/
#ifndef mulu_64x64
inline uint64_t mulu_64x64(uint64_t a, uint64_t b, uint64_t &hi)
{
uint64_t const a_hi = uint32_t(a >> 32);
uint64_t const b_hi = uint32_t(b >> 32);
uint64_t const a_lo = uint32_t(a);
uint64_t const b_lo = uint32_t(b);
uint64_t const ab_lo = a_lo * b_lo;
uint64_t const ab_m1 = a_hi * b_lo;
uint64_t const ab_m2 = a_lo * b_hi;
uint64_t const ab_hi = a_hi * b_hi;
uint64_t const carry = ((ab_lo >> 32) + uint32_t(ab_m1) + uint32_t(ab_m2)) >> 32;
hi = ab_hi + (ab_m1 >> 32) + (ab_m2 >> 32) + carry;
return ab_lo + (ab_m1 << 32) + (ab_m2 << 32);
}
#endif
/*-------------------------------------------------
addu_32x32_co - perform an unsigned 32 bit + 32
bit addition and return the result with carry
out
-------------------------------------------------*/
#ifndef addu_32x32_co
inline bool addu_32x32_co(uint32_t a, uint32_t b, uint32_t &sum)
{
sum = a + b;
return (a > sum) || (b > sum);
}
#endif
/*-------------------------------------------------
addu_64x64_co - perform an unsigned 64 bit + 64
bit addition and return the result with carry
out
-------------------------------------------------*/
#ifndef addu_64x64_co
inline bool addu_64x64_co(uint64_t a, uint64_t b, uint64_t &sum)
{
sum = a + b;
return (a > sum) || (b > sum);
}
#endif
/***************************************************************************
INLINE BIT MANIPULATION FUNCTIONS
***************************************************************************/
/*-------------------------------------------------
count_leading_zeros_32 - return the number of
leading zero bits in a 32-bit value
-------------------------------------------------*/
#ifndef count_leading_zeros_32
inline uint8_t count_leading_zeros_32(uint32_t val)
{
if (!val) return 32U;
uint8_t count;
for (count = 0; int32_t(val) >= 0; count++) val <<= 1;
return count;
}
#endif
/*-------------------------------------------------
count_leading_ones_32 - return the number of
leading one bits in a 32-bit value
-------------------------------------------------*/
#ifndef count_leading_ones_32
inline uint8_t count_leading_ones_32(uint32_t val)
{
uint8_t count;
for (count = 0; int32_t(val) < 0; count++) val <<= 1;
return count;
}
#endif
/*-------------------------------------------------
count_leading_zeros_64 - return the number of
leading zero bits in a 64-bit value
-------------------------------------------------*/
#ifndef count_leading_zeros_64
inline uint8_t count_leading_zeros_64(uint64_t val)
{
if (!val) return 64U;
uint8_t count;
for (count = 0; int64_t(val) >= 0; count++) val <<= 1;
return count;
}
#endif
/*-------------------------------------------------
count_leading_ones_64 - return the number of
leading one bits in a 64-bit value
-------------------------------------------------*/
#ifndef count_leading_ones_64
inline uint8_t count_leading_ones_64(uint64_t val)
{
uint8_t count;
for (count = 0; int64_t(val) < 0; count++) val <<= 1;
return count;
}
#endif
/*-------------------------------------------------
population_count_32 - return the number of
one bits in a 32-bit value
-------------------------------------------------*/
#ifndef population_count_32
inline unsigned population_count_32(uint32_t val)
{
#if defined(__NetBSD__)
return popcount32(val);
#else
// optimal Hamming weight assuming fast 32*32->32
constexpr uint32_t m1(0x55555555);
constexpr uint32_t m2(0x33333333);
constexpr uint32_t m4(0x0f0f0f0f);
constexpr uint32_t h01(0x01010101);
val -= (val >> 1) & m1;
val = (val & m2) + ((val >> 2) & m2);
val = (val + (val >> 4)) & m4;
return unsigned((val * h01) >> 24);
#endif
}
#endif
/*-------------------------------------------------
population_count_64 - return the number of
one bits in a 64-bit value
-------------------------------------------------*/
#ifndef population_count_64
inline unsigned population_count_64(uint64_t val)
{
#if defined(__NetBSD__)
return popcount64(val);
#else
// guess that architectures with 64-bit pointers have 64-bit multiplier
if (sizeof(void *) >= sizeof(uint64_t))
{
// optimal Hamming weight assuming fast 64*64->64
constexpr uint64_t m1(0x5555555555555555);
constexpr uint64_t m2(0x3333333333333333);
constexpr uint64_t m4(0x0f0f0f0f0f0f0f0f);
constexpr uint64_t h01(0x0101010101010101);
val -= (val >> 1) & m1;
val = (val & m2) + ((val >> 2) & m2);
val = (val + (val >> 4)) & m4;
return unsigned((val * h01) >> 56);
}
else
{
// fall back to two 32-bit operations to avoid slow multiply
return population_count_32(uint32_t(val)) + population_count_32(uint32_t(val >> 32));
}
#endif
}
#endif
/*-------------------------------------------------
rotl_32 - circularly shift a 32-bit value left
by the specified number of bits (modulo 32)
-------------------------------------------------*/
#ifndef rotl_32
constexpr uint32_t rotl_32(uint32_t val, int shift)
{
shift &= 31;
if (shift)
return val << shift | val >> (32 - shift);
else
return val;
}
#endif
/*-------------------------------------------------
rotr_32 - circularly shift a 32-bit value right
by the specified number of bits (modulo 32)
-------------------------------------------------*/
#ifndef rotr_32
constexpr uint32_t rotr_32(uint32_t val, int shift)
{
shift &= 31;
if (shift)
return val >> shift | val << (32 - shift);
else
return val;
}
#endif
/*-------------------------------------------------
rotl_64 - circularly shift a 64-bit value left
by the specified number of bits (modulo 64)
-------------------------------------------------*/
#ifndef rotl_64
constexpr uint64_t rotl_64(uint64_t val, int shift)
{
shift &= 63;
if (shift)
return val << shift | val >> (64 - shift);
else
return val;
}
#endif
/*-------------------------------------------------
rotr_64 - circularly shift a 64-bit value right
by the specified number of bits (modulo 64)
-------------------------------------------------*/
#ifndef rotr_64
constexpr uint64_t rotr_64(uint64_t val, int shift)
{
shift &= 63;
if (shift)
return val >> shift | val << (64 - shift);
else
return val;
}
#endif
/***************************************************************************
INLINE TIMING FUNCTIONS
***************************************************************************/
/*-------------------------------------------------
get_profile_ticks - return a tick counter
from the processor that can be used for
profiling. It does not need to run at any
particular rate.
-------------------------------------------------*/
#ifndef get_profile_ticks
inline int64_t get_profile_ticks() noexcept
{
return osd_ticks();
}
#endif
#endif // MAME_OSD_EMINLINE_H
|