Mercurial > hg > sv-dependency-builds

annotate src/fftw-3.3.5/rdft/scalar/r2cf/r2cf_11.c @ 56:af97cad61ff0

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add updated build of PortAudio for OSX
author Chris Cannam <cannam@all-day-breakfast.com>
date Tue, 03 Jan 2017 15:10:52 +0000
parents 2cd0e3b3e1fd
children
rev   line source
Chris@42 1 /*
Chris@42 2 * Copyright (c) 2003, 2007-14 Matteo Frigo
Chris@42 3 * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
Chris@42 4 *
Chris@42 5 * This program is free software; you can redistribute it and/or modify
Chris@42 6 * it under the terms of the GNU General Public License as published by
Chris@42 7 * the Free Software Foundation; either version 2 of the License, or
Chris@42 8 * (at your option) any later version.
Chris@42 9 *
Chris@42 10 * This program is distributed in the hope that it will be useful,
Chris@42 11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
Chris@42 12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
Chris@42 13 * GNU General Public License for more details.
Chris@42 14 *
Chris@42 15 * You should have received a copy of the GNU General Public License
Chris@42 16 * along with this program; if not, write to the Free Software
Chris@42 17 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
Chris@42 18 *
Chris@42 19 */
Chris@42 20
Chris@42 21 /* This file was automatically generated --- DO NOT EDIT */
Chris@42 22 /* Generated on Sat Jul 30 16:46:03 EDT 2016 */
Chris@42 23
Chris@42 24 #include "codelet-rdft.h"
Chris@42 25
Chris@42 26 #ifdef HAVE_FMA
Chris@42 27
Chris@42 28 /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 11 -name r2cf_11 -include r2cf.h */
Chris@42 29
Chris@42 30 /*
Chris@42 31 * This function contains 60 FP additions, 50 FP multiplications,
Chris@42 32 * (or, 15 additions, 5 multiplications, 45 fused multiply/add),
Chris@42 33 * 51 stack variables, 10 constants, and 22 memory accesses
Chris@42 34 */
Chris@42 35 #include "r2cf.h"
Chris@42 36
Chris@42 37 static void r2cf_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
Chris@42 38 {
Chris@42 39 DK(KP959492973, +0.959492973614497389890368057066327699062454848);
Chris@42 40 DK(KP876768831, +0.876768831002589333891339807079336796764054852);
Chris@42 41 DK(KP918985947, +0.918985947228994779780736114132655398124909697);
Chris@42 42 DK(KP989821441, +0.989821441880932732376092037776718787376519372);
Chris@42 43 DK(KP778434453, +0.778434453334651800608337670740821884709317477);
Chris@42 44 DK(KP830830026, +0.830830026003772851058548298459246407048009821);
Chris@42 45 DK(KP715370323, +0.715370323453429719112414662767260662417897278);
Chris@42 46 DK(KP634356270, +0.634356270682424498893150776899916060542806975);
Chris@42 47 DK(KP342584725, +0.342584725681637509502641509861112333758894680);
Chris@42 48 DK(KP521108558, +0.521108558113202722944698153526659300680427422);
Chris@42 49 {
Chris@42 50 INT i;
Chris@42 51 for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) {
Chris@42 52 E T1, Tg, TF, TB, TI, TL, Tz, TA;
Chris@42 53 {
Chris@42 54 E T4, TC, TE, T7, TD, Ta, TS, TG, TJ, Td, TP, TM, Ty, Tq, Th;
Chris@42 55 E Tt, Tl;
Chris@42 56 T1 = R0[0];
Chris@42 57 {
Chris@42 58 E Tb, Tc, Tx, Tp;
Chris@42 59 {
Chris@42 60 E T2, T3, Te, Tf;
Chris@42 61 T2 = R1[0];
Chris@42 62 T3 = R0[WS(rs, 5)];
Chris@42 63 Te = R1[WS(rs, 2)];
Chris@42 64 Tf = R0[WS(rs, 3)];
Chris@42 65 {
Chris@42 66 E T5, T6, T8, T9;
Chris@42 67 T5 = R0[WS(rs, 1)];
Chris@42 68 T4 = T2 + T3;
Chris@42 69 TC = T3 - T2;
Chris@42 70 Tg = Te + Tf;
Chris@42 71 TE = Tf - Te;
Chris@42 72 T6 = R1[WS(rs, 4)];
Chris@42 73 T8 = R1[WS(rs, 1)];
Chris@42 74 T9 = R0[WS(rs, 4)];
Chris@42 75 Tb = R0[WS(rs, 2)];
Chris@42 76 T7 = T5 + T6;
Chris@42 77 TD = T5 - T6;
Chris@42 78 Ta = T8 + T9;
Chris@42 79 TF = T9 - T8;
Chris@42 80 Tc = R1[WS(rs, 3)];
Chris@42 81 }
Chris@42 82 }
Chris@42 83 TS = FMA(KP521108558, TC, TD);
Chris@42 84 TG = FMA(KP521108558, TF, TE);
Chris@42 85 TJ = FMA(KP521108558, TE, TC);
Chris@42 86 Td = Tb + Tc;
Chris@42 87 TB = Tb - Tc;
Chris@42 88 Tx = FNMS(KP342584725, Ta, T7);
Chris@42 89 Tp = FNMS(KP342584725, T4, Ta);
Chris@42 90 TP = FNMS(KP521108558, TB, TF);
Chris@42 91 TM = FNMS(KP521108558, TD, TB);
Chris@42 92 Ty = FNMS(KP634356270, Tx, Td);
Chris@42 93 Tq = FNMS(KP634356270, Tp, Tg);
Chris@42 94 Th = FNMS(KP342584725, Tg, Td);
Chris@42 95 Tt = FNMS(KP342584725, Td, T4);
Chris@42 96 Tl = FNMS(KP342584725, T7, Tg);
Chris@42 97 }
Chris@42 98 {
Chris@42 99 E Tu, Ts, TN, Tv;
Chris@42 100 {
Chris@42 101 E Tm, TU, Tj, Ti, TT;
Chris@42 102 TT = FMA(KP715370323, TS, TF);
Chris@42 103 Ti = FNMS(KP634356270, Th, Ta);
Chris@42 104 Tu = FNMS(KP634356270, Tt, T7);
Chris@42 105 Tm = FNMS(KP634356270, Tl, T4);
Chris@42 106 TU = FMA(KP830830026, TT, TB);
Chris@42 107 Tj = FNMS(KP778434453, Ti, T7);
Chris@42 108 {
Chris@42 109 E Tk, TR, To, Tn, TQ, Tr;
Chris@42 110 TQ = FMA(KP715370323, TP, TC);
Chris@42 111 Tn = FNMS(KP778434453, Tm, Ta);
Chris@42 112 Ci[WS(csi, 5)] = KP989821441 * (FMA(KP918985947, TU, TE));
Chris@42 113 Tk = FNMS(KP876768831, Tj, T4);
Chris@42 114 TR = FNMS(KP830830026, TQ, TE);
Chris@42 115 To = FNMS(KP876768831, Tn, Td);
Chris@42 116 Tr = FNMS(KP778434453, Tq, Td);
Chris@42 117 Cr[WS(csr, 5)] = FNMS(KP959492973, Tk, T1);
Chris@42 118 Ci[WS(csi, 4)] = KP989821441 * (FNMS(KP918985947, TR, TD));
Chris@42 119 Cr[WS(csr, 4)] = FNMS(KP959492973, To, T1);
Chris@42 120 Ts = FNMS(KP876768831, Tr, T7);
Chris@42 121 }
Chris@42 122 }
Chris@42 123 TN = FNMS(KP715370323, TM, TE);
Chris@42 124 Tv = FNMS(KP778434453, Tu, Tg);
Chris@42 125 Cr[0] = T1 + T4 + T7 + Ta + Td + Tg;
Chris@42 126 Cr[WS(csr, 3)] = FNMS(KP959492973, Ts, T1);
Chris@42 127 {
Chris@42 128 E TO, Tw, TH, TK;
Chris@42 129 TO = FNMS(KP830830026, TN, TF);
Chris@42 130 Tw = FNMS(KP876768831, Tv, Ta);
Chris@42 131 TH = FMA(KP715370323, TG, TD);
Chris@42 132 TK = FNMS(KP715370323, TJ, TB);
Chris@42 133 Ci[WS(csi, 3)] = KP989821441 * (FNMS(KP918985947, TO, TC));
Chris@42 134 Cr[WS(csr, 2)] = FNMS(KP959492973, Tw, T1);
Chris@42 135 TI = FNMS(KP830830026, TH, TC);
Chris@42 136 TL = FMA(KP830830026, TK, TD);
Chris@42 137 Tz = FNMS(KP778434453, Ty, T4);
Chris@42 138 }
Chris@42 139 }
Chris@42 140 }
Chris@42 141 Ci[WS(csi, 2)] = KP989821441 * (FMA(KP918985947, TI, TB));
Chris@42 142 Ci[WS(csi, 1)] = KP989821441 * (FNMS(KP918985947, TL, TF));
Chris@42 143 TA = FNMS(KP876768831, Tz, Tg);
Chris@42 144 Cr[WS(csr, 1)] = FNMS(KP959492973, TA, T1);
Chris@42 145 }
Chris@42 146 }
Chris@42 147 }
Chris@42 148
Chris@42 149 static const kr2c_desc desc = { 11, "r2cf_11", {15, 5, 45, 0}, &GENUS };
Chris@42 150
Chris@42 151 void X(codelet_r2cf_11) (planner *p) {
Chris@42 152 X(kr2c_register) (p, r2cf_11, &desc);
Chris@42 153 }
Chris@42 154
Chris@42 155 #else /* HAVE_FMA */
Chris@42 156
Chris@42 157 /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 11 -name r2cf_11 -include r2cf.h */
Chris@42 158
Chris@42 159 /*
Chris@42 160 * This function contains 60 FP additions, 50 FP multiplications,
Chris@42 161 * (or, 20 additions, 10 multiplications, 40 fused multiply/add),
Chris@42 162 * 28 stack variables, 10 constants, and 22 memory accesses
Chris@42 163 */
Chris@42 164 #include "r2cf.h"
Chris@42 165
Chris@42 166 static void r2cf_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
Chris@42 167 {
Chris@42 168 DK(KP654860733, +0.654860733945285064056925072466293553183791199);
Chris@42 169 DK(KP142314838, +0.142314838273285140443792668616369668791051361);
Chris@42 170 DK(KP959492973, +0.959492973614497389890368057066327699062454848);
Chris@42 171 DK(KP415415013, +0.415415013001886425529274149229623203524004910);
Chris@42 172 DK(KP841253532, +0.841253532831181168861811648919367717513292498);
Chris@42 173 DK(KP989821441, +0.989821441880932732376092037776718787376519372);
Chris@42 174 DK(KP909631995, +0.909631995354518371411715383079028460060241051);
Chris@42 175 DK(KP281732556, +0.281732556841429697711417915346616899035777899);
Chris@42 176 DK(KP540640817, +0.540640817455597582107635954318691695431770608);
Chris@42 177 DK(KP755749574, +0.755749574354258283774035843972344420179717445);
Chris@42 178 {
Chris@42 179 INT i;
Chris@42 180 for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) {
Chris@42 181 E T1, T4, Tl, Tg, Th, Td, Ti, Ta, Tk, T7, Tj, Tb, Tc;
Chris@42 182 T1 = R0[0];
Chris@42 183 {
Chris@42 184 E T2, T3, Te, Tf;
Chris@42 185 T2 = R0[WS(rs, 1)];
Chris@42 186 T3 = R1[WS(rs, 4)];
Chris@42 187 T4 = T2 + T3;
Chris@42 188 Tl = T3 - T2;
Chris@42 189 Te = R1[0];
Chris@42 190 Tf = R0[WS(rs, 5)];
Chris@42 191 Tg = Te + Tf;
Chris@42 192 Th = Tf - Te;
Chris@42 193 }
Chris@42 194 Tb = R1[WS(rs, 1)];
Chris@42 195 Tc = R0[WS(rs, 4)];
Chris@42 196 Td = Tb + Tc;
Chris@42 197 Ti = Tc - Tb;
Chris@42 198 {
Chris@42 199 E T8, T9, T5, T6;
Chris@42 200 T8 = R1[WS(rs, 2)];
Chris@42 201 T9 = R0[WS(rs, 3)];
Chris@42 202 Ta = T8 + T9;
Chris@42 203 Tk = T9 - T8;
Chris@42 204 T5 = R0[WS(rs, 2)];
Chris@42 205 T6 = R1[WS(rs, 3)];
Chris@42 206 T7 = T5 + T6;
Chris@42 207 Tj = T6 - T5;
Chris@42 208 }
Chris@42 209 Ci[WS(csi, 4)] = FMA(KP755749574, Th, KP540640817 * Ti) + FNMS(KP909631995, Tk, KP281732556 * Tj) - (KP989821441 * Tl);
Chris@42 210 Cr[WS(csr, 4)] = FMA(KP841253532, Td, T1) + FNMS(KP959492973, T7, KP415415013 * Ta) + FNMA(KP142314838, T4, KP654860733 * Tg);
Chris@42 211 Ci[WS(csi, 2)] = FMA(KP909631995, Th, KP755749574 * Tl) + FNMA(KP540640817, Tk, KP989821441 * Tj) - (KP281732556 * Ti);
Chris@42 212 Ci[WS(csi, 5)] = FMA(KP281732556, Th, KP755749574 * Ti) + FNMS(KP909631995, Tj, KP989821441 * Tk) - (KP540640817 * Tl);
Chris@42 213 Ci[WS(csi, 1)] = FMA(KP540640817, Th, KP909631995 * Tl) + FMA(KP989821441, Ti, KP755749574 * Tj) + (KP281732556 * Tk);
Chris@42 214 Ci[WS(csi, 3)] = FMA(KP989821441, Th, KP540640817 * Tj) + FNMS(KP909631995, Ti, KP755749574 * Tk) - (KP281732556 * Tl);
Chris@42 215 Cr[WS(csr, 3)] = FMA(KP415415013, Td, T1) + FNMS(KP654860733, Ta, KP841253532 * T7) + FNMA(KP959492973, T4, KP142314838 * Tg);
Chris@42 216 Cr[WS(csr, 1)] = FMA(KP841253532, Tg, T1) + FNMS(KP959492973, Ta, KP415415013 * T4) + FNMA(KP654860733, T7, KP142314838 * Td);
Chris@42 217 Cr[0] = T1 + Tg + T4 + Td + T7 + Ta;
Chris@42 218 Cr[WS(csr, 2)] = FMA(KP415415013, Tg, T1) + FNMS(KP142314838, T7, KP841253532 * Ta) + FNMA(KP959492973, Td, KP654860733 * T4);
Chris@42 219 Cr[WS(csr, 5)] = FMA(KP841253532, T4, T1) + FNMS(KP142314838, Ta, KP415415013 * T7) + FNMA(KP654860733, Td, KP959492973 * Tg);
Chris@42 220 }
Chris@42 221 }
Chris@42 222 }
Chris@42 223
Chris@42 224 static const kr2c_desc desc = { 11, "r2cf_11", {20, 10, 40, 0}, &GENUS };
Chris@42 225
Chris@42 226 void X(codelet_r2cf_11) (planner *p) {
Chris@42 227 X(kr2c_register) (p, r2cf_11, &desc);
Chris@42 228 }
Chris@42 229
Chris@42 230 #endif /* HAVE_FMA */