Mercurial > hg > sv-dependency-builds

annotate src/fftw-3.3.8/rdft/scalar/r2cf/r2cf_5.c @ 82:d0c2a83c1364

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add FFTW 3.3.8 source, and a Linux build
author Chris Cannam
date Tue, 19 Nov 2019 14:52:55 +0000
parents
children
rev   line source
Chris@82 1 /*
Chris@82 2 * Copyright (c) 2003, 2007-14 Matteo Frigo
Chris@82 3 * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
Chris@82 4 *
Chris@82 5 * This program is free software; you can redistribute it and/or modify
Chris@82 6 * it under the terms of the GNU General Public License as published by
Chris@82 7 * the Free Software Foundation; either version 2 of the License, or
Chris@82 8 * (at your option) any later version.
Chris@82 9 *
Chris@82 10 * This program is distributed in the hope that it will be useful,
Chris@82 11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
Chris@82 12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
Chris@82 13 * GNU General Public License for more details.
Chris@82 14 *
Chris@82 15 * You should have received a copy of the GNU General Public License
Chris@82 16 * along with this program; if not, write to the Free Software
Chris@82 17 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
Chris@82 18 *
Chris@82 19 */
Chris@82 20
Chris@82 21 /* This file was automatically generated --- DO NOT EDIT */
Chris@82 22 /* Generated on Thu May 24 08:06:26 EDT 2018 */
Chris@82 23
Chris@82 24 #include "rdft/codelet-rdft.h"
Chris@82 25
Chris@82 26 #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
Chris@82 27
Chris@82 28 /* Generated by: ../../../genfft/gen_r2cf.native -fma -compact -variables 4 -pipeline-latency 4 -n 5 -name r2cf_5 -include rdft/scalar/r2cf.h */
Chris@82 29
Chris@82 30 /*
Chris@82 31 * This function contains 12 FP additions, 7 FP multiplications,
Chris@82 32 * (or, 7 additions, 2 multiplications, 5 fused multiply/add),
Chris@82 33 * 17 stack variables, 4 constants, and 10 memory accesses
Chris@82 34 */
Chris@82 35 #include "rdft/scalar/r2cf.h"
Chris@82 36
Chris@82 37 static void r2cf_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
Chris@82 38 {
Chris@82 39 DK(KP559016994, +0.559016994374947424102293417182819058860154590);
Chris@82 40 DK(KP250000000, +0.250000000000000000000000000000000000000000000);
Chris@82 41 DK(KP618033988, +0.618033988749894848204586834365638117720309180);
Chris@82 42 DK(KP951056516, +0.951056516295153572116439333379382143405698634);
Chris@82 43 {
Chris@82 44 INT i;
Chris@82 45 for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) {
Chris@82 46 E T7, T8, T9, T3, Ta, T6, Tb, Tc;
Chris@82 47 T7 = R0[0];
Chris@82 48 {
Chris@82 49 E T1, T2, T4, T5;
Chris@82 50 T1 = R0[WS(rs, 2)];
Chris@82 51 T2 = R1[0];
Chris@82 52 T8 = T2 + T1;
Chris@82 53 T4 = R0[WS(rs, 1)];
Chris@82 54 T5 = R1[WS(rs, 1)];
Chris@82 55 T9 = T4 + T5;
Chris@82 56 T3 = T1 - T2;
Chris@82 57 Ta = T8 + T9;
Chris@82 58 T6 = T4 - T5;
Chris@82 59 }
Chris@82 60 Ci[WS(csi, 1)] = KP951056516 * (FNMS(KP618033988, T6, T3));
Chris@82 61 Cr[0] = T7 + Ta;
Chris@82 62 Ci[WS(csi, 2)] = KP951056516 * (FMA(KP618033988, T3, T6));
Chris@82 63 Tb = FNMS(KP250000000, Ta, T7);
Chris@82 64 Tc = T8 - T9;
Chris@82 65 Cr[WS(csr, 1)] = FMA(KP559016994, Tc, Tb);
Chris@82 66 Cr[WS(csr, 2)] = FNMS(KP559016994, Tc, Tb);
Chris@82 67 }
Chris@82 68 }
Chris@82 69 }
Chris@82 70
Chris@82 71 static const kr2c_desc desc = { 5, "r2cf_5", {7, 2, 5, 0}, &GENUS };
Chris@82 72
Chris@82 73 void X(codelet_r2cf_5) (planner *p) {
Chris@82 74 X(kr2c_register) (p, r2cf_5, &desc);
Chris@82 75 }
Chris@82 76
Chris@82 77 #else
Chris@82 78
Chris@82 79 /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 5 -name r2cf_5 -include rdft/scalar/r2cf.h */
Chris@82 80
Chris@82 81 /*
Chris@82 82 * This function contains 12 FP additions, 6 FP multiplications,
Chris@82 83 * (or, 9 additions, 3 multiplications, 3 fused multiply/add),
Chris@82 84 * 17 stack variables, 4 constants, and 10 memory accesses
Chris@82 85 */
Chris@82 86 #include "rdft/scalar/r2cf.h"
Chris@82 87
Chris@82 88 static void r2cf_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
Chris@82 89 {
Chris@82 90 DK(KP250000000, +0.250000000000000000000000000000000000000000000);
Chris@82 91 DK(KP559016994, +0.559016994374947424102293417182819058860154590);
Chris@82 92 DK(KP587785252, +0.587785252292473129168705954639072768597652438);
Chris@82 93 DK(KP951056516, +0.951056516295153572116439333379382143405698634);
Chris@82 94 {
Chris@82 95 INT i;
Chris@82 96 for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) {
Chris@82 97 E Ta, T7, T8, T3, Tb, T6, T9, Tc;
Chris@82 98 Ta = R0[0];
Chris@82 99 {
Chris@82 100 E T1, T2, T4, T5;
Chris@82 101 T1 = R0[WS(rs, 2)];
Chris@82 102 T2 = R1[0];
Chris@82 103 T7 = T2 + T1;
Chris@82 104 T4 = R0[WS(rs, 1)];
Chris@82 105 T5 = R1[WS(rs, 1)];
Chris@82 106 T8 = T4 + T5;
Chris@82 107 T3 = T1 - T2;
Chris@82 108 Tb = T7 + T8;
Chris@82 109 T6 = T4 - T5;
Chris@82 110 }
Chris@82 111 Ci[WS(csi, 1)] = FNMS(KP587785252, T6, KP951056516 * T3);
Chris@82 112 Cr[0] = Ta + Tb;
Chris@82 113 Ci[WS(csi, 2)] = FMA(KP587785252, T3, KP951056516 * T6);
Chris@82 114 T9 = KP559016994 * (T7 - T8);
Chris@82 115 Tc = FNMS(KP250000000, Tb, Ta);
Chris@82 116 Cr[WS(csr, 1)] = T9 + Tc;
Chris@82 117 Cr[WS(csr, 2)] = Tc - T9;
Chris@82 118 }
Chris@82 119 }
Chris@82 120 }
Chris@82 121
Chris@82 122 static const kr2c_desc desc = { 5, "r2cf_5", {9, 3, 3, 0}, &GENUS };
Chris@82 123
Chris@82 124 void X(codelet_r2cf_5) (planner *p) {
Chris@82 125 X(kr2c_register) (p, r2cf_5, &desc);
Chris@82 126 }
Chris@82 127
Chris@82 128 #endif