Mercurial > hg > js-dsp-test
comparison fft/fftw/fftw-3.3.4/rdft/scalar/r2cf/r2cfII_10.c @ 19:26056e866c29
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 13:49:18 EST 2014 */ | |
| 23 | |
| 24 #include "codelet-rdft.h" | |
| 25 | |
| 26 #ifdef HAVE_FMA | |
| 27 | |
| 28 /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cfII_10 -dft-II -include r2cfII.h */ | |
| 29 | |
| 30 /* | |
| 31 * This function contains 32 FP additions, 18 FP multiplications, | |
| 32 * (or, 14 additions, 0 multiplications, 18 fused multiply/add), | |
| 33 * 37 stack variables, 4 constants, and 20 memory accesses | |
| 34 */ | |
| 35 #include "r2cfII.h" | |
| 36 | |
| 37 static void r2cfII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) | |
| 38 { | |
| 39 DK(KP951056516, +0.951056516295153572116439333379382143405698634); | |
| 40 DK(KP559016994, +0.559016994374947424102293417182819058860154590); | |
| 41 DK(KP250000000, +0.250000000000000000000000000000000000000000000); | |
| 42 DK(KP618033988, +0.618033988749894848204586834365638117720309180); | |
| 43 { | |
| 44 INT i; | |
| 45 for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { | |
| 46 E Tq, Ti, Tk, Tu, Tw, Tp, Tb, Tj, Tr, Tv; | |
| 47 { | |
| 48 E T1, To, Ts, Tt, T8, Ta, Te, Tm, Tl, Th, Tn, T9; | |
| 49 T1 = R0[0]; | |
| 50 To = R1[WS(rs, 2)]; | |
| 51 { | |
| 52 E T2, T3, T5, T6; | |
| 53 T2 = R0[WS(rs, 2)]; | |
| 54 T3 = R0[WS(rs, 3)]; | |
| 55 T5 = R0[WS(rs, 4)]; | |
| 56 T6 = R0[WS(rs, 1)]; | |
| 57 { | |
| 58 E Tc, T4, T7, Td, Tf, Tg; | |
| 59 Tc = R1[0]; | |
| 60 Ts = T2 + T3; | |
| 61 T4 = T2 - T3; | |
| 62 Tt = T5 + T6; | |
| 63 T7 = T5 - T6; | |
| 64 Td = R1[WS(rs, 4)]; | |
| 65 Tf = R1[WS(rs, 1)]; | |
| 66 Tg = R1[WS(rs, 3)]; | |
| 67 T8 = T4 + T7; | |
| 68 Ta = T4 - T7; | |
| 69 Te = Tc - Td; | |
| 70 Tm = Tc + Td; | |
| 71 Tl = Tf + Tg; | |
| 72 Th = Tf - Tg; | |
| 73 } | |
| 74 } | |
| 75 Cr[WS(csr, 2)] = T1 + T8; | |
| 76 Tn = Tl - Tm; | |
| 77 Tq = Tm + Tl; | |
| 78 Ti = FMA(KP618033988, Th, Te); | |
| 79 Tk = FNMS(KP618033988, Te, Th); | |
| 80 Ci[WS(csi, 2)] = Tn - To; | |
| 81 T9 = FNMS(KP250000000, T8, T1); | |
| 82 Tu = FMA(KP618033988, Tt, Ts); | |
| 83 Tw = FNMS(KP618033988, Ts, Tt); | |
| 84 Tp = FMA(KP250000000, Tn, To); | |
| 85 Tb = FMA(KP559016994, Ta, T9); | |
| 86 Tj = FNMS(KP559016994, Ta, T9); | |
| 87 } | |
| 88 Tr = FMA(KP559016994, Tq, Tp); | |
| 89 Tv = FNMS(KP559016994, Tq, Tp); | |
| 90 Cr[WS(csr, 1)] = FNMS(KP951056516, Tk, Tj); | |
| 91 Cr[WS(csr, 3)] = FMA(KP951056516, Tk, Tj); | |
| 92 Cr[0] = FMA(KP951056516, Ti, Tb); | |
| 93 Cr[WS(csr, 4)] = FNMS(KP951056516, Ti, Tb); | |
| 94 Ci[WS(csi, 1)] = FNMS(KP951056516, Tw, Tv); | |
| 95 Ci[WS(csi, 3)] = FMA(KP951056516, Tw, Tv); | |
| 96 Ci[WS(csi, 4)] = FMS(KP951056516, Tu, Tr); | |
| 97 Ci[0] = -(FMA(KP951056516, Tu, Tr)); | |
| 98 } | |
| 99 } | |
| 100 } | |
| 101 | |
| 102 static const kr2c_desc desc = { 10, "r2cfII_10", {14, 0, 18, 0}, &GENUS }; | |
| 103 | |
| 104 void X(codelet_r2cfII_10) (planner *p) { | |
| 105 X(kr2c_register) (p, r2cfII_10, &desc); | |
| 106 } | |
| 107 | |
| 108 #else /* HAVE_FMA */ | |
| 109 | |
| 110 /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cfII_10 -dft-II -include r2cfII.h */ | |
| 111 | |
| 112 /* | |
| 113 * This function contains 32 FP additions, 12 FP multiplications, | |
| 114 * (or, 26 additions, 6 multiplications, 6 fused multiply/add), | |
| 115 * 21 stack variables, 4 constants, and 20 memory accesses | |
| 116 */ | |
| 117 #include "r2cfII.h" | |
| 118 | |
| 119 static void r2cfII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) | |
| 120 { | |
| 121 DK(KP250000000, +0.250000000000000000000000000000000000000000000); | |
| 122 DK(KP587785252, +0.587785252292473129168705954639072768597652438); | |
| 123 DK(KP951056516, +0.951056516295153572116439333379382143405698634); | |
| 124 DK(KP559016994, +0.559016994374947424102293417182819058860154590); | |
| 125 { | |
| 126 INT i; | |
| 127 for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { | |
| 128 E T1, To, T8, Tq, T9, Tp, Te, Ts, Th, Tn; | |
| 129 T1 = R0[0]; | |
| 130 To = R1[WS(rs, 2)]; | |
| 131 { | |
| 132 E T2, T3, T4, T5, T6, T7; | |
| 133 T2 = R0[WS(rs, 2)]; | |
| 134 T3 = R0[WS(rs, 3)]; | |
| 135 T4 = T2 - T3; | |
| 136 T5 = R0[WS(rs, 4)]; | |
| 137 T6 = R0[WS(rs, 1)]; | |
| 138 T7 = T5 - T6; | |
| 139 T8 = T4 + T7; | |
| 140 Tq = T5 + T6; | |
| 141 T9 = KP559016994 * (T4 - T7); | |
| 142 Tp = T2 + T3; | |
| 143 } | |
| 144 { | |
| 145 E Tc, Td, Tm, Tf, Tg, Tl; | |
| 146 Tc = R1[0]; | |
| 147 Td = R1[WS(rs, 4)]; | |
| 148 Tm = Tc + Td; | |
| 149 Tf = R1[WS(rs, 1)]; | |
| 150 Tg = R1[WS(rs, 3)]; | |
| 151 Tl = Tf + Tg; | |
| 152 Te = Tc - Td; | |
| 153 Ts = KP559016994 * (Tm + Tl); | |
| 154 Th = Tf - Tg; | |
| 155 Tn = Tl - Tm; | |
| 156 } | |
| 157 Cr[WS(csr, 2)] = T1 + T8; | |
| 158 Ci[WS(csi, 2)] = Tn - To; | |
| 159 { | |
| 160 E Ti, Tk, Tb, Tj, Ta; | |
| 161 Ti = FMA(KP951056516, Te, KP587785252 * Th); | |
| 162 Tk = FNMS(KP587785252, Te, KP951056516 * Th); | |
| 163 Ta = FNMS(KP250000000, T8, T1); | |
| 164 Tb = T9 + Ta; | |
| 165 Tj = Ta - T9; | |
| 166 Cr[WS(csr, 4)] = Tb - Ti; | |
| 167 Cr[WS(csr, 3)] = Tj + Tk; | |
| 168 Cr[0] = Tb + Ti; | |
| 169 Cr[WS(csr, 1)] = Tj - Tk; | |
| 170 } | |
| 171 { | |
| 172 E Tr, Tw, Tu, Tv, Tt; | |
| 173 Tr = FMA(KP951056516, Tp, KP587785252 * Tq); | |
| 174 Tw = FNMS(KP587785252, Tp, KP951056516 * Tq); | |
| 175 Tt = FMA(KP250000000, Tn, To); | |
| 176 Tu = Ts + Tt; | |
| 177 Tv = Tt - Ts; | |
| 178 Ci[0] = -(Tr + Tu); | |
| 179 Ci[WS(csi, 3)] = Tw + Tv; | |
| 180 Ci[WS(csi, 4)] = Tr - Tu; | |
| 181 Ci[WS(csi, 1)] = Tv - Tw; | |
| 182 } | |
| 183 } | |
| 184 } | |
| 185 } | |
| 186 | |
| 187 static const kr2c_desc desc = { 10, "r2cfII_10", {26, 6, 6, 0}, &GENUS }; | |
| 188 | |
| 189 void X(codelet_r2cfII_10) (planner *p) { | |
| 190 X(kr2c_register) (p, r2cfII_10, &desc); | |
| 191 } | |
| 192 | |
| 193 #endif /* HAVE_FMA */ |