-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathammintrin.h
254 lines (223 loc) · 10.9 KB
/
ammintrin.h
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
/****
* Copyright (C) 2007-2008 Advanced Micro Devices Inc. All rights reserved.
*
* The information and source code contained herein is the exclusive
* property of Advanced Micro Devices and may not be disclosed, examined
* or reproduced in whole or in part without explicit written authorization
* from the company.
*
* ammintrin.h - Definitions for AMD-specific intrinsics
*
****/
#pragma once
#ifndef __midl
#ifndef _INCLUDED_AMM
#define _INCLUDED_AMM
#if defined(_M_CEE_PURE)
#error ERROR: This file is not supported in the pure mode!
#else
#if defined __cplusplus
extern "C" { /* Intrinsics use C name-mangling. */
#endif
/*
* Vector integer comparison control macros
*/
#define _MM_PCOMCTRL_LT 0
#define _MM_PCOMCTRL_LE 1
#define _MM_PCOMCTRL_GT 2
#define _MM_PCOMCTRL_GE 3
#define _MM_PCOMCTRL_EQ 4
#define _MM_PCOMCTRL_NEQ 5
#define _MM_PCOMCTRL_FALSE 6
#define _MM_PCOMCTRL_TRUE 7
/*
* MACRO functions for vector integer comparisons
*/
#define _mm_comlt_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epu8(v1, v2) _mm_com_epu8(v1, v2, _MM_PCOMCTRL_TRUE)
#define _mm_comlt_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epu16(v1, v2) _mm_com_epu16(v1, v2, _MM_PCOMCTRL_TRUE)
#define _mm_comlt_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epu32(v1, v2) _mm_com_epu32(v1, v2, _MM_PCOMCTRL_TRUE)
#define _mm_comlt_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epu64(v1, v2) _mm_com_epu64(v1, v2, _MM_PCOMCTRL_TRUE)
#define _mm_comlt_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epi8(v1, v2) _mm_com_epi8(v1, v2, _MM_PCOMCTRL_TRUE)
#define _mm_comlt_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epi16(v1, v2) _mm_com_epi16(v1, v2, _MM_PCOMCTRL_TRUE)
#define _mm_comlt_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epi32(v1, v2) _mm_com_epi32(v1, v2, _MM_PCOMCTRL_TRUE)
#define _mm_comlt_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_LT)
#define _mm_comle_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_LE)
#define _mm_comgt_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_GT)
#define _mm_comge_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_GE)
#define _mm_comeq_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_EQ)
#define _mm_comneq_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_NEQ)
#define _mm_comfalse_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_FALSE)
#define _mm_comtrue_epi64(v1, v2) _mm_com_epi64(v1, v2, _MM_PCOMCTRL_TRUE)
/* SSE5 intrinsics */
/* Float/double multiply-accumulate */
__m128 _mm_macc_ps(__m128, __m128, __m128);
__m128d _mm_macc_pd(__m128d, __m128d, __m128d);
__m128 _mm_macc_ss(__m128, __m128, __m128);
__m128d _mm_macc_sd(__m128d, __m128d, __m128d);
__m128 _mm_maddsub_ps(__m128, __m128, __m128);
__m128d _mm_maddsub_pd(__m128d, __m128d, __m128d);
__m128 _mm_msubadd_ps(__m128, __m128, __m128);
__m128d _mm_msubadd_pd(__m128d, __m128d, __m128d);
__m128 _mm_msub_ps(__m128, __m128, __m128);
__m128d _mm_msub_pd(__m128d, __m128d, __m128d);
__m128 _mm_msub_ss(__m128, __m128, __m128);
__m128d _mm_msub_sd(__m128d, __m128d, __m128d);
__m128 _mm_nmacc_ps(__m128, __m128, __m128);
__m128d _mm_nmacc_pd(__m128d, __m128d, __m128d);
__m128 _mm_nmacc_ss(__m128, __m128, __m128);
__m128d _mm_nmacc_sd(__m128d, __m128d, __m128d);
__m128 _mm_nmsub_ps(__m128, __m128, __m128);
__m128d _mm_nmsub_pd(__m128d, __m128d, __m128d);
__m128 _mm_nmsub_ss(__m128, __m128, __m128);
__m128d _mm_nmsub_sd(__m128d, __m128d, __m128d);
/* Integer multiply-accumulate */
__m128i _mm_maccs_epi16(__m128i, __m128i, __m128i);
__m128i _mm_macc_epi16(__m128i, __m128i, __m128i);
__m128i _mm_maccsd_epi16(__m128i, __m128i, __m128i);
__m128i _mm_maccd_epi16(__m128i, __m128i, __m128i);
__m128i _mm_maccs_epi32(__m128i, __m128i, __m128i);
__m128i _mm_macc_epi32(__m128i, __m128i, __m128i);
__m128i _mm_maccslo_epi32(__m128i, __m128i, __m128i);
__m128i _mm_macclo_epi32(__m128i, __m128i, __m128i);
__m128i _mm_maccshi_epi32(__m128i, __m128i, __m128i);
__m128i _mm_macchi_epi32(__m128i, __m128i, __m128i);
__m128i _mm_maddsd_epi16(__m128i, __m128i, __m128i);
__m128i _mm_maddd_epi16(__m128i, __m128i, __m128i);
/* Horizontal add/subtract */
__m128i _mm_haddw_epi8(__m128i);
__m128i _mm_haddd_epi8(__m128i);
__m128i _mm_haddq_epi8(__m128i);
__m128i _mm_haddd_epi16(__m128i);
__m128i _mm_haddq_epi16(__m128i);
__m128i _mm_haddq_epi32(__m128i);
__m128i _mm_haddw_epu8(__m128i);
__m128i _mm_haddd_epu8(__m128i);
__m128i _mm_haddq_epu8(__m128i);
__m128i _mm_haddd_epu16(__m128i);
__m128i _mm_haddq_epu16(__m128i);
__m128i _mm_haddq_epu32(__m128i);
__m128i _mm_hsubw_epi8(__m128i);
__m128i _mm_hsubd_epi16(__m128i);
__m128i _mm_hsubq_epi32(__m128i);
/* Vector conditional moves */
__m128i _mm_cmov_si128(__m128i, __m128i, __m128i);
__m128i _mm_perm_epi8(__m128i, __m128i, __m128i);
/* Vector shifts and rotates */
__m128i _mm_rot_epi8(__m128i, __m128i);
__m128i _mm_rot_epi16(__m128i, __m128i);
__m128i _mm_rot_epi32(__m128i, __m128i);
__m128i _mm_rot_epi64(__m128i, __m128i);
__m128i _mm_roti_epi8(__m128i, int);
__m128i _mm_roti_epi16(__m128i, int);
__m128i _mm_roti_epi32(__m128i, int);
__m128i _mm_roti_epi64(__m128i, int);
__m128i _mm_shl_epi8(__m128i, __m128i);
__m128i _mm_shl_epi16(__m128i, __m128i);
__m128i _mm_shl_epi32(__m128i, __m128i);
__m128i _mm_shl_epi64(__m128i, __m128i);
__m128i _mm_sha_epi8(__m128i, __m128i);
__m128i _mm_sha_epi16(__m128i, __m128i);
__m128i _mm_sha_epi32(__m128i, __m128i);
__m128i _mm_sha_epi64(__m128i, __m128i);
/* Vector integer comparisons */
__m128i _mm_com_epu8(__m128i, __m128i, int);
__m128i _mm_com_epu16(__m128i, __m128i, int);
__m128i _mm_com_epu32(__m128i, __m128i, int);
__m128i _mm_com_epu64(__m128i, __m128i, int);
__m128i _mm_com_epi8(__m128i, __m128i, int);
__m128i _mm_com_epi16(__m128i, __m128i, int);
__m128i _mm_com_epi32(__m128i, __m128i, int);
__m128i _mm_com_epi64(__m128i, __m128i, int);
/* Precision control */
__m128 _mm_frcz_ps(__m128);
__m128d _mm_frcz_pd(__m128d);
__m128 _mm_frcz_ss(__m128, __m128);
__m128d _mm_frcz_sd(__m128d, __m128d);
/* Control values for permute2 intrinsics */
#define _MM_PERMUTE2_COPY 0 /* just copy the selected value */
/* Note that using the constant 1 would have the same effect as 0 */
#define _MM_PERMUTE2_ZEROIF1 2 /* zero selected value if src3 bit is 1 */
#define _MM_PERMUTE2_ZEROIF0 3 /* zero selected value if src3 bit is 3 */
/* Permutation */
__m128 _mm_permute2_ps(__m128, __m128, __m128i, int);
__m128d _mm_permute2_pd(__m128d, __m128d, __m128i, int);
/* YMM versions */
__m256 _mm256_macc_ps(__m256, __m256, __m256);
__m256d _mm256_macc_pd(__m256d, __m256d, __m256d);
__m256 _mm256_maddsub_ps(__m256, __m256, __m256);
__m256d _mm256_maddsub_pd(__m256d, __m256d, __m256d);
__m256 _mm256_msubadd_ps(__m256, __m256, __m256);
__m256d _mm256_msubadd_pd(__m256d, __m256d, __m256d);
__m256 _mm256_msub_ps(__m256, __m256, __m256);
__m256d _mm256_msub_pd(__m256d, __m256d, __m256d);
__m256 _mm256_nmacc_ps(__m256, __m256, __m256);
__m256d _mm256_nmacc_pd(__m256d, __m256d, __m256d);
__m256 _mm256_nmsub_ps(__m256, __m256, __m256);
__m256d _mm256_nmsub_pd(__m256d, __m256d, __m256d);
__m256i _mm256_cmov_si256(__m256i, __m256i, __m256i);
__m256 _mm256_frcz_ps(__m256);
__m256d _mm256_frcz_pd(__m256d);
__m256 _mm256_permute2_ps(__m256, __m256, __m256i, int);
__m256d _mm256_permute2_pd(__m256d, __m256d, __m256i, int);
/* LWP intrinsics */
void __llwpcb(void *);
void *__slwpcb();
void __lwpval32(unsigned int, unsigned int, unsigned int);
void __lwpval64(unsigned __int64, unsigned int, unsigned int);
unsigned char __lwpins32(unsigned int, unsigned int, unsigned int);
unsigned char __lwpins64(unsigned __int64, unsigned int, unsigned int);
#if defined __cplusplus
}; /* End "C" */
#endif /* __cplusplus */
#endif /* _M_CEE_PURE */
#endif /* INCLUDED_AMM */
#endif /* __midl */