Mercurial > hg > js-dsp-test

comparison fft/fftw/fftw-3.3.4/rdft/scalar/r2cb/r2cb_6.c @ 19:26056e866c29

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add FFTW to comparison table
author Chris Cannam
date Tue, 06 Oct 2015 13:08:39 +0100
parents
children
comparison
equal deleted inserted replaced
18:8db794ca3e0b 19:26056e866c29
1 /*
2 * Copyright (c) 2003, 2007-14 Matteo Frigo
3 * Copyright (c) 2003, 2007-14 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 Tue Mar 4 13:50:24 EST 2014 */
23
24 #include "codelet-rdft.h"
25
26 #ifdef HAVE_FMA
27
28 /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cb_6 -include r2cb.h */
29
30 /*
31 * This function contains 14 FP additions, 6 FP multiplications,
32 * (or, 8 additions, 0 multiplications, 6 fused multiply/add),
33 * 13 stack variables, 2 constants, and 12 memory accesses
34 */
35 #include "r2cb.h"
36
37 static void r2cb_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
38 {
39 DK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
40 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
41 {
42 INT i;
43 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) {
44 E T4, T7, T3, Te, Tc, T5;
45 {
46 E T1, T2, Ta, Tb;
47 T1 = Cr[0];
48 T2 = Cr[WS(csr, 3)];
49 Ta = Ci[WS(csi, 2)];
50 Tb = Ci[WS(csi, 1)];
51 T4 = Cr[WS(csr, 2)];
52 T7 = T1 - T2;
53 T3 = T1 + T2;
54 Te = Ta + Tb;
55 Tc = Ta - Tb;
56 T5 = Cr[WS(csr, 1)];
57 }
58 {
59 E T6, T8, Td, T9;
60 T6 = T4 + T5;
61 T8 = T5 - T4;
62 Td = T7 + T8;
63 R1[WS(rs, 1)] = FNMS(KP2_000000000, T8, T7);
64 T9 = T3 - T6;
65 R0[0] = FMA(KP2_000000000, T6, T3);
66 R1[WS(rs, 2)] = FMA(KP1_732050807, Te, Td);
67 R1[0] = FNMS(KP1_732050807, Te, Td);
68 R0[WS(rs, 1)] = FMA(KP1_732050807, Tc, T9);
69 R0[WS(rs, 2)] = FNMS(KP1_732050807, Tc, T9);
70 }
71 }
72 }
73 }
74
75 static const kr2c_desc desc = { 6, "r2cb_6", {8, 0, 6, 0}, &GENUS };
76
77 void X(codelet_r2cb_6) (planner *p) {
78 X(kr2c_register) (p, r2cb_6, &desc);
79 }
80
81 #else /* HAVE_FMA */
82
83 /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cb_6 -include r2cb.h */
84
85 /*
86 * This function contains 14 FP additions, 4 FP multiplications,
87 * (or, 12 additions, 2 multiplications, 2 fused multiply/add),
88 * 17 stack variables, 2 constants, and 12 memory accesses
89 */
90 #include "r2cb.h"
91
92 static void r2cb_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
93 {
94 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
95 DK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
96 {
97 INT i;
98 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) {
99 E T3, T7, Tc, Te, T6, T8, T1, T2, T9, Td;
100 T1 = Cr[0];
101 T2 = Cr[WS(csr, 3)];
102 T3 = T1 - T2;
103 T7 = T1 + T2;
104 {
105 E Ta, Tb, T4, T5;
106 Ta = Ci[WS(csi, 2)];
107 Tb = Ci[WS(csi, 1)];
108 Tc = KP1_732050807 * (Ta - Tb);
109 Te = KP1_732050807 * (Ta + Tb);
110 T4 = Cr[WS(csr, 2)];
111 T5 = Cr[WS(csr, 1)];
112 T6 = T4 - T5;
113 T8 = T4 + T5;
114 }
115 R1[WS(rs, 1)] = FMA(KP2_000000000, T6, T3);
116 R0[0] = FMA(KP2_000000000, T8, T7);
117 T9 = T7 - T8;
118 R0[WS(rs, 2)] = T9 - Tc;
119 R0[WS(rs, 1)] = T9 + Tc;
120 Td = T3 - T6;
121 R1[0] = Td - Te;
122 R1[WS(rs, 2)] = Td + Te;
123 }
124 }
125 }
126
127 static const kr2c_desc desc = { 6, "r2cb_6", {12, 2, 2, 0}, &GENUS };
128
129 void X(codelet_r2cb_6) (planner *p) {
130 X(kr2c_register) (p, r2cb_6, &desc);
131 }
132
133 #endif /* HAVE_FMA */