File: t1bv_4.c

package info (click to toggle)
fftw3 3.2.2-1
  • links: PTS, VCS
  • area: main
  • in suites: squeeze
  • size: 16,824 kB
  • ctags: 8,846
  • sloc: ansic: 246,029; sh: 10,285; ml: 5,470; perl: 1,423; makefile: 935; fortran: 110; asm: 26
file content (128 lines) | stat: -rw-r--r-- 4,254 bytes parent folder | download
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
/*
 * Copyright (c) 2003, 2007-8 Matteo Frigo
 * Copyright (c) 2003, 2007-8 Massachusetts Institute of Technology
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 *
 */

/* This file was automatically generated --- DO NOT EDIT */
/* Generated on Sun Jul 12 06:42:29 EDT 2009 */

#include "codelet-dft.h"

#ifdef HAVE_FMA

/* Generated by: ../../../genfft/gen_twiddle_c -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1bv_4 -include t1b.h -sign 1 */

/*
 * This function contains 11 FP additions, 8 FP multiplications,
 * (or, 9 additions, 6 multiplications, 2 fused multiply/add),
 * 13 stack variables, 0 constants, and 8 memory accesses
 */
#include "t1b.h"

static void t1bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     INT m;
     R *x;
     x = ii;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs)) {
	  V T1, T7, T2, T5, T8, T3, T6;
	  T1 = LD(&(x[0]), ms, &(x[0]));
	  T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
	  T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
	  T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
	  T8 = BYTW(&(W[TWVL * 4]), T7);
	  T3 = BYTW(&(W[TWVL * 2]), T2);
	  T6 = BYTW(&(W[0]), T5);
	  {
	       V Ta, T4, Tb, T9;
	       Ta = VADD(T1, T3);
	       T4 = VSUB(T1, T3);
	       Tb = VADD(T6, T8);
	       T9 = VSUB(T6, T8);
	       ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0]));
	       ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0]));
	       ST(&(x[WS(rs, 1)]), VFMAI(T9, T4), ms, &(x[WS(rs, 1)]));
	       ST(&(x[WS(rs, 3)]), VFNMSI(T9, T4), ms, &(x[WS(rs, 1)]));
	  }
     }
}

static const tw_instr twinstr[] = {
     VTW(0, 1),
     VTW(0, 2),
     VTW(0, 3),
     {TW_NEXT, VL, 0}
};

static const ct_desc desc = { 4, "t1bv_4", twinstr, &GENUS, {9, 6, 2, 0}, 0, 0, 0 };

void X(codelet_t1bv_4) (planner *p) {
     X(kdft_dit_register) (p, t1bv_4, &desc);
}
#else				/* HAVE_FMA */

/* Generated by: ../../../genfft/gen_twiddle_c -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1bv_4 -include t1b.h -sign 1 */

/*
 * This function contains 11 FP additions, 6 FP multiplications,
 * (or, 11 additions, 6 multiplications, 0 fused multiply/add),
 * 13 stack variables, 0 constants, and 8 memory accesses
 */
#include "t1b.h"

static void t1bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     INT m;
     R *x;
     x = ii;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs)) {
	  V T1, T8, T3, T6, T7, T2, T5;
	  T1 = LD(&(x[0]), ms, &(x[0]));
	  T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
	  T8 = BYTW(&(W[TWVL * 4]), T7);
	  T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
	  T3 = BYTW(&(W[TWVL * 2]), T2);
	  T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
	  T6 = BYTW(&(W[0]), T5);
	  {
	       V T4, T9, Ta, Tb;
	       T4 = VSUB(T1, T3);
	       T9 = VBYI(VSUB(T6, T8));
	       ST(&(x[WS(rs, 3)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)]));
	       ST(&(x[WS(rs, 1)]), VADD(T4, T9), ms, &(x[WS(rs, 1)]));
	       Ta = VADD(T1, T3);
	       Tb = VADD(T6, T8);
	       ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0]));
	       ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0]));
	  }
     }
}

static const tw_instr twinstr[] = {
     VTW(0, 1),
     VTW(0, 2),
     VTW(0, 3),
     {TW_NEXT, VL, 0}
};

static const ct_desc desc = { 4, "t1bv_4", twinstr, &GENUS, {11, 6, 0, 0}, 0, 0, 0 };

void X(codelet_t1bv_4) (planner *p) {
     X(kdft_dit_register) (p, t1bv_4, &desc);
}
#endif				/* HAVE_FMA */