Mercurial > hg > sv-dependency-builds

comparison src/fftw-3.3.3/dft/scalar/codelets/n1_4.c @ 95:89f5e221ed7b

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add FFTW3
author Chris Cannam <cannam@all-day-breakfast.com>
date Wed, 20 Mar 2013 15:35:50 +0000
parents
children
comparison
equal deleted inserted replaced
94:d278df1123f9 95:89f5e221ed7b
1 /*
2 * Copyright (c) 2003, 2007-11 Matteo Frigo
3 * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology
4 *
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.
9 *
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.
14 *
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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
18 *
19 */
20
21 /* This file was automatically generated --- DO NOT EDIT */
22 /* Generated on Sun Nov 25 07:35:42 EST 2012 */
23
24 #include "codelet-dft.h"
25
26 #ifdef HAVE_FMA
27
28 /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -name n1_4 -include n.h */
29
30 /*
31 * This function contains 16 FP additions, 0 FP multiplications,
32 * (or, 16 additions, 0 multiplications, 0 fused multiply/add),
33 * 13 stack variables, 0 constants, and 16 memory accesses
34 */
35 #include "n.h"
36
37 static void n1_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
38 {
39 {
40 INT i;
41 for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) {
42 E T4, Tb, T3, Tf, T9, T5, Tc, Td;
43 {
44 E T1, T2, T7, T8;
45 T1 = ri[0];
46 T2 = ri[WS(is, 2)];
47 T7 = ii[0];
48 T8 = ii[WS(is, 2)];
49 T4 = ri[WS(is, 1)];
50 Tb = T1 - T2;
51 T3 = T1 + T2;
52 Tf = T7 + T8;
53 T9 = T7 - T8;
54 T5 = ri[WS(is, 3)];
55 Tc = ii[WS(is, 1)];
56 Td = ii[WS(is, 3)];
57 }
58 {
59 E T6, Ta, Te, Tg;
60 T6 = T4 + T5;
61 Ta = T4 - T5;
62 Te = Tc - Td;
63 Tg = Tc + Td;
64 io[WS(os, 3)] = Ta + T9;
65 io[WS(os, 1)] = T9 - Ta;
66 ro[0] = T3 + T6;
67 ro[WS(os, 2)] = T3 - T6;
68 io[0] = Tf + Tg;
69 io[WS(os, 2)] = Tf - Tg;
70 ro[WS(os, 3)] = Tb - Te;
71 ro[WS(os, 1)] = Tb + Te;
72 }
73 }
74 }
75 }
76
77 static const kdft_desc desc = { 4, "n1_4", {16, 0, 0, 0}, &GENUS, 0, 0, 0, 0 };
78
79 void X(codelet_n1_4) (planner *p) {
80 X(kdft_register) (p, n1_4, &desc);
81 }
82
83 #else /* HAVE_FMA */
84
85 /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 4 -name n1_4 -include n.h */
86
87 /*
88 * This function contains 16 FP additions, 0 FP multiplications,
89 * (or, 16 additions, 0 multiplications, 0 fused multiply/add),
90 * 13 stack variables, 0 constants, and 16 memory accesses
91 */
92 #include "n.h"
93
94 static void n1_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
95 {
96 {
97 INT i;
98 for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) {
99 E T3, Tb, T9, Tf, T6, Ta, Te, Tg;
100 {
101 E T1, T2, T7, T8;
102 T1 = ri[0];
103 T2 = ri[WS(is, 2)];
104 T3 = T1 + T2;
105 Tb = T1 - T2;
106 T7 = ii[0];
107 T8 = ii[WS(is, 2)];
108 T9 = T7 - T8;
109 Tf = T7 + T8;
110 }
111 {
112 E T4, T5, Tc, Td;
113 T4 = ri[WS(is, 1)];
114 T5 = ri[WS(is, 3)];
115 T6 = T4 + T5;
116 Ta = T4 - T5;
117 Tc = ii[WS(is, 1)];
118 Td = ii[WS(is, 3)];
119 Te = Tc - Td;
120 Tg = Tc + Td;
121 }
122 ro[WS(os, 2)] = T3 - T6;
123 io[WS(os, 2)] = Tf - Tg;
124 ro[0] = T3 + T6;
125 io[0] = Tf + Tg;
126 io[WS(os, 1)] = T9 - Ta;
127 ro[WS(os, 1)] = Tb + Te;
128 io[WS(os, 3)] = Ta + T9;
129 ro[WS(os, 3)] = Tb - Te;
130 }
131 }
132 }
133
134 static const kdft_desc desc = { 4, "n1_4", {16, 0, 0, 0}, &GENUS, 0, 0, 0, 0 };
135
136 void X(codelet_n1_4) (planner *p) {
137 X(kdft_register) (p, n1_4, &desc);
138 }
139
140 #endif /* HAVE_FMA */