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 Thu Dec 10 07:04:56 EST 2020 */
23
24 #include "dft/codelet-dft.h"
25
26 #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
27
28 /* Generated by: ../../../genfft/gen_notw_c.native -fma -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 2 -name n2bv_2 -with-ostride 2 -include dft/simd/n2b.h -store-multiple 2 */
29
30 /*
31 * This function contains 2 FP additions, 0 FP multiplications,
32 * (or, 2 additions, 0 multiplications, 0 fused multiply/add),
33 * 7 stack variables, 0 constants, and 5 memory accesses
34 */
35 #include "dft/simd/n2b.h"
36
n2bv_2(const R * ri,const R * ii,R * ro,R * io,stride is,stride os,INT v,INT ivs,INT ovs)37 static void n2bv_2(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 const R *xi;
42 R *xo;
43 xi = ii;
44 xo = io;
45 for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) {
46 V T1, T2, T3, T4;
47 T1 = LD(&(xi[0]), ivs, &(xi[0]));
48 T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
49 T3 = VSUB(T1, T2);
50 STM2(&(xo[2]), T3, ovs, &(xo[2]));
51 T4 = VADD(T1, T2);
52 STM2(&(xo[0]), T4, ovs, &(xo[0]));
53 STN2(&(xo[0]), T4, T3, ovs);
54 }
55 }
56 VLEAVE();
57 }
58
59 static const kdft_desc desc = { 2, XSIMD_STRING("n2bv_2"), { 2, 0, 0, 0 }, &GENUS, 0, 2, 0, 0 };
60
XSIMD(codelet_n2bv_2)61 void XSIMD(codelet_n2bv_2) (planner *p) { X(kdft_register) (p, n2bv_2, &desc);
62 }
63
64 #else
65
66 /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 2 -name n2bv_2 -with-ostride 2 -include dft/simd/n2b.h -store-multiple 2 */
67
68 /*
69 * This function contains 2 FP additions, 0 FP multiplications,
70 * (or, 2 additions, 0 multiplications, 0 fused multiply/add),
71 * 7 stack variables, 0 constants, and 5 memory accesses
72 */
73 #include "dft/simd/n2b.h"
74
n2bv_2(const R * ri,const R * ii,R * ro,R * io,stride is,stride os,INT v,INT ivs,INT ovs)75 static void n2bv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
76 {
77 {
78 INT i;
79 const R *xi;
80 R *xo;
81 xi = ii;
82 xo = io;
83 for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) {
84 V T1, T2, T3, T4;
85 T1 = LD(&(xi[0]), ivs, &(xi[0]));
86 T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
87 T3 = VSUB(T1, T2);
88 STM2(&(xo[2]), T3, ovs, &(xo[2]));
89 T4 = VADD(T1, T2);
90 STM2(&(xo[0]), T4, ovs, &(xo[0]));
91 STN2(&(xo[0]), T4, T3, ovs);
92 }
93 }
94 VLEAVE();
95 }
96
97 static const kdft_desc desc = { 2, XSIMD_STRING("n2bv_2"), { 2, 0, 0, 0 }, &GENUS, 0, 2, 0, 0 };
98
XSIMD(codelet_n2bv_2)99 void XSIMD(codelet_n2bv_2) (planner *p) { X(kdft_register) (p, n2bv_2, &desc);
100 }
101
102 #endif
103