|
0
|
1 /* Copyright (C) 2007, 2008, 2009 Free Software Foundation, Inc.
|
|
|
2
|
|
|
3 This file is part of GCC.
|
|
|
4
|
|
|
5 GCC is free software; you can redistribute it and/or modify
|
|
|
6 it under the terms of the GNU General Public License as published by
|
|
|
7 the Free Software Foundation; either version 3, or (at your option)
|
|
|
8 any later version.
|
|
|
9
|
|
|
10 GCC is distributed in the hope that it will be useful,
|
|
|
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
13 GNU General Public License for more details.
|
|
|
14
|
|
|
15 Under Section 7 of GPL version 3, you are granted additional
|
|
|
16 permissions described in the GCC Runtime Library Exception, version
|
|
|
17 3.1, as published by the Free Software Foundation.
|
|
|
18
|
|
|
19 You should have received a copy of the GNU General Public License and
|
|
|
20 a copy of the GCC Runtime Library Exception along with this program;
|
|
|
21 see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
|
|
|
22 <http://www.gnu.org/licenses/>. */
|
|
|
23
|
|
|
24 /* Implemented from the specification included in the AMD Programmers
|
|
|
25 Manual Update, version 2.x */
|
|
|
26
|
|
|
27 #ifndef _AMMINTRIN_H_INCLUDED
|
|
|
28 #define _AMMINTRIN_H_INCLUDED
|
|
|
29
|
|
|
30 #ifndef __SSE4A__
|
|
|
31 # error "SSE4A instruction set not enabled"
|
|
|
32 #else
|
|
|
33
|
|
|
34 /* We need definitions from the SSE3, SSE2 and SSE header files*/
|
|
|
35 #include <pmmintrin.h>
|
|
|
36
|
|
|
37 extern __inline void __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
|
|
38 _mm_stream_sd (double * __P, __m128d __Y)
|
|
|
39 {
|
|
|
40 __builtin_ia32_movntsd (__P, (__v2df) __Y);
|
|
|
41 }
|
|
|
42
|
|
|
43 extern __inline void __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
|
|
44 _mm_stream_ss (float * __P, __m128 __Y)
|
|
|
45 {
|
|
|
46 __builtin_ia32_movntss (__P, (__v4sf) __Y);
|
|
|
47 }
|
|
|
48
|
|
|
49 extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
|
|
50 _mm_extract_si64 (__m128i __X, __m128i __Y)
|
|
|
51 {
|
|
|
52 return (__m128i) __builtin_ia32_extrq ((__v2di) __X, (__v16qi) __Y);
|
|
|
53 }
|
|
|
54
|
|
|
55 #ifdef __OPTIMIZE__
|
|
|
56 extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
|
|
57 _mm_extracti_si64 (__m128i __X, unsigned const int __I, unsigned const int __L)
|
|
|
58 {
|
|
|
59 return (__m128i) __builtin_ia32_extrqi ((__v2di) __X, __I, __L);
|
|
|
60 }
|
|
|
61 #else
|
|
|
62 #define _mm_extracti_si64(X, I, L) \
|
|
|
63 ((__m128i) __builtin_ia32_extrqi ((__v2di)(__m128i)(X), \
|
|
|
64 (unsigned int)(I), (unsigned int)(L)))
|
|
|
65 #endif
|
|
|
66
|
|
|
67 extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
|
|
68 _mm_insert_si64 (__m128i __X,__m128i __Y)
|
|
|
69 {
|
|
|
70 return (__m128i) __builtin_ia32_insertq ((__v2di)__X, (__v2di)__Y);
|
|
|
71 }
|
|
|
72
|
|
|
73 #ifdef __OPTIMIZE__
|
|
|
74 extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
|
|
75 _mm_inserti_si64(__m128i __X, __m128i __Y, unsigned const int __I, unsigned const int __L)
|
|
|
76 {
|
|
|
77 return (__m128i) __builtin_ia32_insertqi ((__v2di)__X, (__v2di)__Y, __I, __L);
|
|
|
78 }
|
|
|
79 #else
|
|
|
80 #define _mm_inserti_si64(X, Y, I, L) \
|
|
|
81 ((__m128i) __builtin_ia32_insertqi ((__v2di)(__m128i)(X), \
|
|
|
82 (__v2di)(__m128i)(Y), \
|
|
|
83 (unsigned int)(I), (unsigned int)(L)))
|
|
|
84 #endif
|
|
|
85
|
|
|
86 #endif /* __SSE4A__ */
|
|
|
87
|
|
|
88 #endif /* _AMMINTRIN_H_INCLUDED */
|
