summaryrefslogtreecommitdiffstatshomepage
path: root/src/osd/eigccarm.h
blob: 90004daf9d4e99c0fe848eb4fea41c13350dd3f6 (plain) (blame)
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
// license:BSD-3-Clause
// copyright-holders:Vas Crabb
/***************************************************************************

    eigccarm.h

    ARM/AArch64 inline implementations for GCC compilers. This code is
    automatically included if appropriate by eminline.h.

***************************************************************************/

#ifndef MAME_OSD_EIGCCARM_H
#define MAME_OSD_EIGCCARM_H


/***************************************************************************
    INLINE MATH FUNCTIONS
***************************************************************************/

/*-------------------------------------------------
    mul_32x32 - perform a signed 32 bit x 32 bit
    multiply and return the full 64 bit result
-------------------------------------------------*/

// GCC can do a good job of this.


/*-------------------------------------------------
    mulu_32x32 - perform an unsigned 32 bit x
    32 bit multiply and return the full 64 bit
    result
-------------------------------------------------*/

// GCC can do a good job of this


/*-------------------------------------------------
    mul_32x32_hi - perform a signed 32 bit x 32 bit
    multiply and return the upper 32 bits of the
    result
-------------------------------------------------*/

// GCC can do a good job of this


/*-------------------------------------------------
    mulu_32x32_hi - perform an unsigned 32 bit x
    32 bit multiply and return the upper 32 bits
    of the result
-------------------------------------------------*/

// GCC can do a good job of this


/*-------------------------------------------------
    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
-------------------------------------------------*/

#if !defined(__aarch64__)
#define mul_32x32_shift _mul_32x32_shift
inline int32_t ATTR_CONST ATTR_FORCE_INLINE
_mul_32x32_shift(int32_t val1, int32_t val2, uint8_t shift)
{
	uint32_t l, h;

	__asm__ (
		" smull  %[l], %[h], %[val1], %[val2] \n"
		: [l]      "=r" (l)
		, [h]      "=r" (h)
		: [val1]   "%r" (val1)
		, [val2]   "r"  (val2)
	);

	// Valid for (0 <= shift <= 31)
	return int32_t((l >> shift) | (h << (32 - 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
-------------------------------------------------*/

#if !defined(__aarch64__)
#define mulu_32x32_shift _mulu_32x32_shift
inline uint32_t ATTR_CONST ATTR_FORCE_INLINE
_mulu_32x32_shift(uint32_t val1, uint32_t val2, uint8_t shift)
{
	uint32_t l, h;

	__asm__ (
		" umull  %[l], %[h], %[val1], %[val2] \n"
		: [l]      "=r" (l)
		, [h]      "=r" (h)
		: [val1]   "%r" (val1)
		, [val2]   "r"  (val2)
	);

	// Valid for (0 <= shift <= 31)
	return (l >> shift) | (h << (32 - shift));
}
#endif


/*-------------------------------------------------
    div_64x32 - perform a signed 64 bit x 32 bit
    divide and return the 32 bit quotient
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    divu_64x32 - perform an unsigned 64 bit x 32 bit
    divide and return the 32 bit quotient
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    div_64x32_rem - perform a signed 64 bit x 32
    bit divide and return the 32 bit quotient and
    32 bit remainder
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    divu_64x32_rem - perform an unsigned 64 bit x
    32 bit divide and return the 32 bit quotient
    and 32 bit remainder
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    div_32x32_shift - perform a signed divide of
    two 32 bit values, shifting the first before
    division, and returning the 32 bit quotient
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    divu_32x32_shift - perform an unsigned divide of
    two 32 bit values, shifting the first before
    division, and returning the 32 bit quotient
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    mod_64x32 - perform a signed 64 bit x 32 bit
    divide and return the 32 bit remainder
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    modu_64x32 - perform an unsigned 64 bit x 32 bit
    divide and return the 32 bit remainder
-------------------------------------------------*/

// TBD


/*-------------------------------------------------
    recip_approx - compute an approximate floating
    point reciprocal
-------------------------------------------------*/

#if defined(__aarch64__)
#define recip_approx _recip_approx
inline float ATTR_CONST ATTR_FORCE_INLINE
_recip_approx(float value)
{
	float result;

	__asm__ (
		" frecpe  %s[result], %s[value] \n"
		: [result] "=w" (result)
		: [value]  "w"  (value)
	);

	return result;
}
#endif


/*-------------------------------------------------
    mul_64x64 - perform a signed 64 bit x 64 bit
    multiply and return the full 128 bit result
-------------------------------------------------*/

#if defined(__aarch64__)
#define mul_64x64 _mul_64x64
inline int64_t ATTR_FORCE_INLINE
_mul_64x64(int64_t a, int64_t b, int64_t &hi)
{
	__int128 const r(__int128(a) * b);
	hi = int64_t(uint64_t((unsigned __int128)r >> 64));
	return int64_t(uint64_t((unsigned __int128)r));
}
#endif


/*-------------------------------------------------
    mulu_64x64 - perform an unsigned 64 bit x 64
    bit multiply and return the full 128 bit result
-------------------------------------------------*/

#if defined(__aarch64__)
#define mulu_64x64 _mulu_64x64
inline uint64_t ATTR_FORCE_INLINE
_mulu_64x64(uint64_t a, uint64_t b, uint64_t &hi)
{
	unsigned __int128 const r((unsigned __int128)a * b);
	hi = uint64_t(r >> 64);
	return uint64_t(r);
}
#endif



/***************************************************************************
    INLINE BIT MANIPULATION FUNCTIONS
***************************************************************************/

/*-------------------------------------------------
    count_leading_zeros_32 - return the number of
    leading zero bits in a 32-bit value
-------------------------------------------------*/

#if defined(__aarch64__)
#define count_leading_zeros_32 _count_leading_zeros_32
inline uint8_t ATTR_CONST ATTR_FORCE_INLINE
_count_leading_zeros_32(uint32_t value)
{
	uint32_t result;

	__asm__ (
		" clz  %w[result], %w[value] \n"
		: [result] "=r" (result)
		: [value]  "r"  (value)
	);

	return uint8_t(result);
}
#endif


/*-------------------------------------------------
    count_leading_ones_32 - return the number of
    leading one bits in a 32-bit value
-------------------------------------------------*/

#if defined(__aarch64__)
#define count_leading_ones_32 _count_leading_ones_32
inline uint8_t ATTR_CONST ATTR_FORCE_INLINE
_count_leading_ones_32(uint32_t value)
{
	uint32_t result;

	__asm__ (
		" clz  %w[result], %w[value] \n"
		: [result] "=r" (result)
		: [value]  "r"  (~value)
	);

	return uint8_t(result);
}
#endif


/*-------------------------------------------------
    count_leading_zeros_64 - return the number of
    leading zero bits in a 64-bit value
-------------------------------------------------*/

#if defined(__aarch64__)
#define count_leading_zeros_64 _count_leading_zeros_64
inline uint8_t ATTR_CONST ATTR_FORCE_INLINE
_count_leading_zeros_64(uint64_t value)
{
	uint64_t result;

	__asm__ (
		" clz  %[result], %[value] \n"
		: [result] "=r" (result)
		: [value]  "r"  (value)
	);

	return uint8_t(result);
}
#endif


/*-------------------------------------------------
    count_leading_ones_64 - return the number of
    leading one bits in a 64-bit value
-------------------------------------------------*/

#if defined(__aarch64__)
#define count_leading_ones_64 _count_leading_ones_64
inline uint8_t ATTR_CONST ATTR_FORCE_INLINE
_count_leading_ones_64(uint64_t value)
{
	uint64_t result;

	__asm__ (
		" clz  %[result], %[value] \n"
		: [result] "=r" (result)
		: [value]  "r"  (~value)
	);

	return uint8_t(result);
}
#endif

#endif // MAME_OSD_EIGCCARM_H