2
* Copyright (c) 2003, 2006 Matteo Frigo
3
* Copyright (c) 2003, 2006 Massachusetts Institute of Technology
5
* This program 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 2 of the License, or
8
* (at your option) any later version.
10
* This program 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.
15
* You should have received a copy of the GNU General Public License
16
* along with this program; if not, write to the Free Software
17
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
21
/* This file was automatically generated --- DO NOT EDIT */
22
/* Generated on Sun Jul 2 14:18:41 EDT 2006 */
24
#include "codelet-rdft.h"
28
/* Generated by: ../../../genfft/gen_r2hc -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 7 -name r2hc_7 -include r2hc.h */
31
* This function contains 24 FP additions, 18 FP multiplications,
32
* (or, 9 additions, 3 multiplications, 15 fused multiply/add),
33
* 25 stack variables, and 14 memory accesses
37
* $Id: algsimp.ml,v 1.9 2006-02-12 23:34:12 athena Exp $
38
* $Id: fft.ml,v 1.4 2006-01-05 03:04:27 stevenj Exp $
39
* $Id: gen_r2hc.ml,v 1.18 2006-02-12 23:34:12 athena Exp $
44
static void r2hc_7(const R *I, R *ro, R *io, stride is, stride ros, stride ios, INT v, INT ivs, INT ovs)
46
DK(KP900968867, +0.900968867902419126236102319507445051165919162);
47
DK(KP801937735, +0.801937735804838252472204639014890102331838324);
48
DK(KP974927912, +0.974927912181823607018131682993931217232785801);
49
DK(KP692021471, +0.692021471630095869627814897002069140197260599);
50
DK(KP554958132, +0.554958132087371191422194871006410481067288862);
51
DK(KP356895867, +0.356895867892209443894399510021300583399127187);
53
for (i = v; i > 0; i = i - 1, I = I + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(ros), MAKE_VOLATILE_STRIDE(ios)) {
56
E Th, T4, Ti, Ta, Tj, T7, Td, T5, T6, Tl, Tk;
73
Td = FNMS(KP356895867, T4, Ta);
74
Tl = FMA(KP554958132, Ti, Th);
75
Tk = FMA(KP554958132, Tj, Ti);
78
Tm = FNMS(KP554958132, Th, Tj);
79
ro[0] = T1 + T4 + T7 + Ta;
80
Tf = FNMS(KP356895867, T7, T4);
81
Tb = FNMS(KP356895867, Ta, T7);
82
Te = FNMS(KP692021471, Td, T7);
83
io[WS(ios, 2)] = KP974927912 * (FNMS(KP801937735, Tk, Th));
84
io[WS(ios, 3)] = KP974927912 * (FNMS(KP801937735, Tm, Ti));
85
Tg = FNMS(KP692021471, Tf, Ta);
86
Tc = FNMS(KP692021471, Tb, T4);
87
ro[WS(ros, 2)] = FNMS(KP900968867, Te, T1);
88
io[WS(ios, 1)] = KP974927912 * (FMA(KP801937735, Tl, Tj));
91
ro[WS(ros, 1)] = FNMS(KP900968867, Tg, T1);
92
ro[WS(ros, 3)] = FNMS(KP900968867, Tc, T1);
96
static const kr2hc_desc desc = { 7, "r2hc_7", {9, 3, 15, 0}, &GENUS, 0, 0, 0, 0, 0 };
98
void X(codelet_r2hc_7) (planner *p) {
99
X(kr2hc_register) (p, r2hc_7, &desc);
104
/* Generated by: ../../../genfft/gen_r2hc -compact -variables 4 -pipeline-latency 4 -n 7 -name r2hc_7 -include r2hc.h */
107
* This function contains 24 FP additions, 18 FP multiplications,
108
* (or, 12 additions, 6 multiplications, 12 fused multiply/add),
109
* 20 stack variables, and 14 memory accesses
113
* $Id: algsimp.ml,v 1.9 2006-02-12 23:34:12 athena Exp $
114
* $Id: fft.ml,v 1.4 2006-01-05 03:04:27 stevenj Exp $
115
* $Id: gen_r2hc.ml,v 1.18 2006-02-12 23:34:12 athena Exp $
120
static void r2hc_7(const R *I, R *ro, R *io, stride is, stride ros, stride ios, INT v, INT ivs, INT ovs)
122
DK(KP222520933, +0.222520933956314404288902564496794759466355569);
123
DK(KP900968867, +0.900968867902419126236102319507445051165919162);
124
DK(KP623489801, +0.623489801858733530525004884004239810632274731);
125
DK(KP433883739, +0.433883739117558120475768332848358754609990728);
126
DK(KP781831482, +0.781831482468029808708444526674057750232334519);
127
DK(KP974927912, +0.974927912181823607018131682993931217232785801);
129
for (i = v; i > 0; i = i - 1, I = I + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(ros), MAKE_VOLATILE_STRIDE(ios)) {
130
E T1, Ta, Tb, T4, Td, T7, Tc, T8, T9;
147
io[WS(ios, 2)] = FNMS(KP781831482, Tc, KP974927912 * Tb) - (KP433883739 * Td);
148
io[WS(ios, 1)] = FMA(KP781831482, Tb, KP974927912 * Td) + (KP433883739 * Tc);
149
ro[WS(ros, 2)] = FMA(KP623489801, T7, T1) + FNMA(KP900968867, T4, KP222520933 * Ta);
150
io[WS(ios, 3)] = FMA(KP433883739, Tb, KP974927912 * Tc) - (KP781831482 * Td);
151
ro[WS(ros, 3)] = FMA(KP623489801, T4, T1) + FNMA(KP222520933, T7, KP900968867 * Ta);
152
ro[WS(ros, 1)] = FMA(KP623489801, Ta, T1) + FNMA(KP900968867, T7, KP222520933 * T4);
153
ro[0] = T1 + Ta + T4 + T7;
157
static const kr2hc_desc desc = { 7, "r2hc_7", {12, 6, 12, 0}, &GENUS, 0, 0, 0, 0, 0 };
159
void X(codelet_r2hc_7) (planner *p) {
160
X(kr2hc_register) (p, r2hc_7, &desc);
163
#endif /* HAVE_FMA */