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 07:05:08 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_twiddle_c.native -fma -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1fuv_2 -include dft/simd/t1fu.h */
29
30 /*
31 * This function contains 3 FP additions, 2 FP multiplications,
32 * (or, 3 additions, 2 multiplications, 0 fused multiply/add),
33 * 5 stack variables, 0 constants, and 4 memory accesses
34 */
35 #include "dft/simd/t1fu.h"
36
t1fuv_2(R * ri,R * ii,const R * W,stride rs,INT mb,INT me,INT ms)37 static void t1fuv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
38 {
39 {
40 INT m;
41 R *x;
42 x = ri;
43 for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) {
44 V T1, T3, T2;
45 T1 = LD(&(x[0]), ms, &(x[0]));
46 T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
47 T3 = BYTWJ(&(W[0]), T2);
48 ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)]));
49 ST(&(x[0]), VADD(T1, T3), ms, &(x[0]));
50 }
51 }
52 VLEAVE();
53 }
54
55 static const tw_instr twinstr[] = {
56 VTW(0, 1),
57 { TW_NEXT, VL, 0 }
58 };
59
60 static const ct_desc desc = { 2, XSIMD_STRING("t1fuv_2"), twinstr, &GENUS, { 3, 2, 0, 0 }, 0, 0, 0 };
61
XSIMD(codelet_t1fuv_2)62 void XSIMD(codelet_t1fuv_2) (planner *p) {
63 X(kdft_dit_register) (p, t1fuv_2, &desc);
64 }
65 #else
66
67 /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1fuv_2 -include dft/simd/t1fu.h */
68
69 /*
70 * This function contains 3 FP additions, 2 FP multiplications,
71 * (or, 3 additions, 2 multiplications, 0 fused multiply/add),
72 * 5 stack variables, 0 constants, and 4 memory accesses
73 */
74 #include "dft/simd/t1fu.h"
75
t1fuv_2(R * ri,R * ii,const R * W,stride rs,INT mb,INT me,INT ms)76 static void t1fuv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
77 {
78 {
79 INT m;
80 R *x;
81 x = ri;
82 for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) {
83 V T1, T3, T2;
84 T1 = LD(&(x[0]), ms, &(x[0]));
85 T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
86 T3 = BYTWJ(&(W[0]), T2);
87 ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)]));
88 ST(&(x[0]), VADD(T1, T3), ms, &(x[0]));
89 }
90 }
91 VLEAVE();
92 }
93
94 static const tw_instr twinstr[] = {
95 VTW(0, 1),
96 { TW_NEXT, VL, 0 }
97 };
98
99 static const ct_desc desc = { 2, XSIMD_STRING("t1fuv_2"), twinstr, &GENUS, { 3, 2, 0, 0 }, 0, 0, 0 };
100
XSIMD(codelet_t1fuv_2)101 void XSIMD(codelet_t1fuv_2) (planner *p) {
102 X(kdft_dit_register) (p, t1fuv_2, &desc);
103 }
104 #endif
105