fft.h
4.57 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
//
// fft.h: this file is part of the SL program suite.
//
// Copyright (C) 2020 daiteq.
//
//
#ifndef SL_BENCHMARKS_FFT_FFT_H
# define SL_BENCHMARKS_FFT_FFT_H
#define FT float
#if (defined(PACKED) && (!defined(GCC)))
typedef float cpx_t __attribute__((ext_vector_type(2)));
#else
typedef struct { FT x; FT y; } cpx_t;
#endif
#if defined(PACKED)
#if defined(GCC)
#define opcode_FADDPS(A, B, RES) \
__asm__ __volatile__ ( \
"faddps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "If" (A), "If" (B) \
) \
#define opcode_FSUBPS(A, B, RES) \
__asm__ __volatile__ ( \
"fsubps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "If" (A), "If" (B) \
) \
#define opcode_FMULPS(A, B, RES) \
__asm__ __volatile__ ( \
"fmulps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "If" (A), "If" (B) \
) \
#endif
#define opcode_FADDRPS(A, B, RES) \
__asm__ __volatile__ ( \
"faddrps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "f" (A), "f" (B) \
) \
#define opcode_FSUBRPS(A, B, RES) \
__asm__ __volatile__ ( \
"fsubrps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "f" (A), "f" (B) \
) \
#define opcode_FMULXPS(A, B, RES) \
__asm__ __volatile__ ( \
"fmulxps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "f" (A), "f" (B) \
) \
#define opcode_FSUBADDRPS(A, B, RES) \
__asm__ __volatile__ ( \
"fsubaddrps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "If" (A), "If" (B) \
) \
#define opcode_FMULRPS(A, B, RES) \
__asm__ __volatile__ ( \
"fmulrps\t%1, %2, %0 \n" \
: "=f" (RES) \
: "f" (A), "f" (B) \
) \
#endif
#if defined (PACKED)
#if defined(GCC)
#define dcadd(a,b,z) \
opcode_FADDPS(*((double*)&a),*((double*)&b),*((double*)&z));
#define dcsub(a,b,z) \
opcode_FSUBPS(*((double*)&a),*((double*)&b),*((double*)&z));
// 1. C = FMULP(A,B) = (ar*br, ai*bi)
// 2. D = FMULXP(A,B) = (ar*bi, ai*br)
// 3. H = FSUBADDRP(C,D) = (ar*br-ai*bi, ar*bi+ai*br)
#define dcmul(a,b,z) \
do { \
cpx_t tc,td; \
opcode_FMULPS(*((double*)&a),*((double*)&b),*((double*)&tc)); \
opcode_FMULXPS(*((double*)&a),*((double*)&b),*((double*)&td)); \
opcode_FSUBADDRPS(*((double*)&tc),*((double*)&td),*((double*)&z)); \
} while (0)
#define dcabss(a,z) \
do { \
cpx_t tc, td, th; \
th.x=FZERO; th.y=FZERO; \
opcode_FMULPS(*((double*)&a),*((double*)&a),*((double*)&tc)); \
opcode_FADDRPS(*((double*)&tc),*((double*)&th),*((double*)&td)); \
z=td.x; \
} while (0)
#else
#define dcadd(a,b,z) \
z=a+b;
#define dcsub(a,b,z) \
z=a-b;
// 1. C = FMULP(A,B) = (ar*br, ai*bi)
// 2. D = FMULXP(A,B) = (ar*bi, ai*br)
// 3. H = FSUBADDRP(C,D) = (ar*br-ai*bi, ar*bi+ai*br)
#define dcmul(a,b,z) \
do { \
cpx_t tc,td; \
tc=a*b; \
opcode_FMULXPS(a,b,td); \
opcode_FSUBADDRPS(tc,td,z); \
} while (0)
#define dcabss(a,z) \
do { \
cpx_t tc, td, th; \
th.x=FZERO; th.y=FZERO; \
tc=a*a; \
opcode_FADDRPS(tc,th,td); \
z=td.x; \
} while (0)
#endif
#else
#define dcadd(a,b,z) \
z.x=a.x+b.x; \
z.y=a.y+b.y;
#define dcsub(a,b,z) \
z.x=a.x-b.x; \
z.y=a.y-b.y;
#define dcmul(a,b,z) \
do { \
cpx_t th; \
th.x=a.x*b.x-a.y*b.y; \
th.y=a.y*b.x+a.x*b.y; \
z=th; \
} while (0)
#define dcabss(a,z) \
z=a.x*a.x+a.y*a.y;
#endif
/* low-level FFT (for benchmarks) */
static
void FFT_1(unsigned long M, cpx_t*restrict, unsigned long, const void*, unsigned);
#define STRINGIFY_(N) # N
#define STRINGIFY(N) STRINGIFY_(N)
#define MAKENAME_(N, SZ) tables/fft_table ## N ## _ ## SZ ## _data.h
#define MAKENAME(N, SZ) MAKENAME_(N, SZ)
#endif // ! SL_BENCHMARKS_FFT_FFT_H