Mercurial > hg > sv-dependency-builds

annotate src/fftw-3.3.5/rdft/scalar/r2cb/r2cb_13.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 16:49:27 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_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 13 -name r2cb_13 -include r2cb.h */
Chris@42 29
Chris@42 30 /*
Chris@42 31 * This function contains 76 FP additions, 58 FP multiplications,
Chris@42 32 * (or, 18 additions, 0 multiplications, 58 fused multiply/add),
Chris@42 33 * 76 stack variables, 26 constants, and 26 memory accesses
Chris@42 34 */
Chris@42 35 #include "r2cb.h"
Chris@42 36
Chris@42 37 static void r2cb_13(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(KP968287244, +0.968287244361984016049539446938120421179794516);
Chris@42 40 DK(KP875502302, +0.875502302409147941146295545768755143177842006);
Chris@42 41 DK(KP1_150281458, +1.150281458948006242736771094910906776922003215);
Chris@42 42 DK(KP1_040057143, +1.040057143777729238234261000998465604986476278);
Chris@42 43 DK(KP1_200954543, +1.200954543865330565851538506669526018704025697);
Chris@42 44 DK(KP769338817, +0.769338817572980603471413688209101117038278899);
Chris@42 45 DK(KP600925212, +0.600925212577331548853203544578415991041882762);
Chris@42 46 DK(KP1_033041561, +1.033041561246979445681802577138034271410067244);
Chris@42 47 DK(KP1_007074065, +1.007074065727533254493747707736933954186697125);
Chris@42 48 DK(KP503537032, +0.503537032863766627246873853868466977093348562);
Chris@42 49 DK(KP581704778, +0.581704778510515730456870384989698884939833902);
Chris@42 50 DK(KP859542535, +0.859542535098774820163672132761689612766401925);
Chris@42 51 DK(KP166666666, +0.166666666666666666666666666666666666666666667);
Chris@42 52 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
Chris@42 53 DK(KP301479260, +0.301479260047709873958013540496673347309208464);
Chris@42 54 DK(KP226109445, +0.226109445035782405468510155372505010481906348);
Chris@42 55 DK(KP686558370, +0.686558370781754340655719594850823015421401653);
Chris@42 56 DK(KP514918778, +0.514918778086315755491789696138117261566051239);
Chris@42 57 DK(KP957805992, +0.957805992594665126462521754605754580515587217);
Chris@42 58 DK(KP522026385, +0.522026385161275033714027226654165028300441940);
Chris@42 59 DK(KP853480001, +0.853480001859823990758994934970528322872359049);
Chris@42 60 DK(KP038632954, +0.038632954644348171955506895830342264440241080);
Chris@42 61 DK(KP612264650, +0.612264650376756543746494474777125408779395514);
Chris@42 62 DK(KP302775637, +0.302775637731994646559610633735247973125648287);
Chris@42 63 DK(KP866025403, +0.866025403784438646763723170752936183471402627);
Chris@42 64 DK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@42 65 {
Chris@42 66 INT i;
Chris@42 67 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) {
Chris@42 68 E TW, T14, TS, TO, T18, T1e, TY, TX, TQ, Tq, TP, Tl, T1d, Tr;
Chris@42 69 {
Chris@42 70 E T1, TN, T16, TJ, TV, TG, TU, Tf, T2, T3, Tb, Ti, T4;
Chris@42 71 {
Chris@42 72 E Ts, TB, Tx, Ty, Tv, TE, Tt, Tu, Tz, TC;
Chris@42 73 Ts = Ci[WS(csi, 5)];
Chris@42 74 Tt = Ci[WS(csi, 2)];
Chris@42 75 Tu = Ci[WS(csi, 6)];
Chris@42 76 TB = Ci[WS(csi, 1)];
Chris@42 77 Tx = Ci[WS(csi, 3)];
Chris@42 78 Ty = Ci[WS(csi, 4)];
Chris@42 79 Tv = Tt + Tu;
Chris@42 80 TE = Tu - Tt;
Chris@42 81 T1 = Cr[0];
Chris@42 82 Tz = Tx + Ty;
Chris@42 83 TC = Tx - Ty;
Chris@42 84 {
Chris@42 85 E TL, Tw, T7, Ta;
Chris@42 86 TL = Ts + Tv;
Chris@42 87 Tw = FNMS(KP500000000, Tv, Ts);
Chris@42 88 T7 = Cr[WS(csr, 5)];
Chris@42 89 {
Chris@42 90 E TD, TM, TA, TH;
Chris@42 91 TD = FNMS(KP500000000, TC, TB);
Chris@42 92 TM = TB + TC;
Chris@42 93 TA = FMA(KP866025403, Tz, Tw);
Chris@42 94 TH = FNMS(KP866025403, Tz, Tw);
Chris@42 95 TN = FMA(KP302775637, TM, TL);
Chris@42 96 T16 = FNMS(KP302775637, TL, TM);
Chris@42 97 {
Chris@42 98 E TF, TI, T8, T9;
Chris@42 99 TF = FMA(KP866025403, TE, TD);
Chris@42 100 TI = FNMS(KP866025403, TE, TD);
Chris@42 101 T8 = Cr[WS(csr, 2)];
Chris@42 102 T9 = Cr[WS(csr, 6)];
Chris@42 103 TJ = FNMS(KP612264650, TI, TH);
Chris@42 104 TV = FMA(KP612264650, TH, TI);
Chris@42 105 TG = FNMS(KP038632954, TF, TA);
Chris@42 106 TU = FMA(KP038632954, TA, TF);
Chris@42 107 Tf = T8 - T9;
Chris@42 108 Ta = T8 + T9;
Chris@42 109 }
Chris@42 110 }
Chris@42 111 T2 = Cr[WS(csr, 1)];
Chris@42 112 T3 = Cr[WS(csr, 3)];
Chris@42 113 Tb = T7 + Ta;
Chris@42 114 Ti = FMS(KP500000000, Ta, T7);
Chris@42 115 T4 = Cr[WS(csr, 4)];
Chris@42 116 }
Chris@42 117 }
Chris@42 118 {
Chris@42 119 E T17, TK, T5, Te, Tk, Td;
Chris@42 120 TW = FMA(KP853480001, TV, TU);
Chris@42 121 T17 = FNMS(KP853480001, TV, TU);
Chris@42 122 TK = FNMS(KP853480001, TJ, TG);
Chris@42 123 T14 = FMA(KP853480001, TJ, TG);
Chris@42 124 T5 = T3 + T4;
Chris@42 125 Te = T3 - T4;
Chris@42 126 {
Chris@42 127 E Tn, Tg, Th, T6;
Chris@42 128 TS = FNMS(KP522026385, TK, TN);
Chris@42 129 TO = FMA(KP957805992, TN, TK);
Chris@42 130 Tn = Te - Tf;
Chris@42 131 Tg = Te + Tf;
Chris@42 132 Th = FNMS(KP500000000, T5, T2);
Chris@42 133 T6 = T2 + T5;
Chris@42 134 T18 = FNMS(KP522026385, T17, T16);
Chris@42 135 T1e = FMA(KP957805992, T16, T17);
Chris@42 136 {
Chris@42 137 E Tm, Tj, Tc, Tp, To;
Chris@42 138 Tm = Th + Ti;
Chris@42 139 Tj = Th - Ti;
Chris@42 140 Tc = T6 + Tb;
Chris@42 141 Tp = T6 - Tb;
Chris@42 142 To = FNMS(KP514918778, Tn, Tm);
Chris@42 143 TY = FMA(KP686558370, Tm, Tn);
Chris@42 144 TX = FNMS(KP226109445, Tg, Tj);
Chris@42 145 Tk = FMA(KP301479260, Tj, Tg);
Chris@42 146 R0[0] = FMA(KP2_000000000, Tc, T1);
Chris@42 147 Td = FNMS(KP166666666, Tc, T1);
Chris@42 148 TQ = FNMS(KP859542535, To, Tp);
Chris@42 149 Tq = FMA(KP581704778, Tp, To);
Chris@42 150 }
Chris@42 151 }
Chris@42 152 TP = FNMS(KP503537032, Tk, Td);
Chris@42 153 Tl = FMA(KP1_007074065, Tk, Td);
Chris@42 154 }
Chris@42 155 }
Chris@42 156 T1d = FNMS(KP1_033041561, Tq, Tl);
Chris@42 157 Tr = FMA(KP1_033041561, Tq, Tl);
Chris@42 158 {
Chris@42 159 E T13, TR, T19, TZ;
Chris@42 160 T13 = FNMS(KP600925212, TQ, TP);
Chris@42 161 TR = FMA(KP600925212, TQ, TP);
Chris@42 162 T19 = FMA(KP769338817, TY, TX);
Chris@42 163 TZ = FNMS(KP769338817, TY, TX);
Chris@42 164 R0[WS(rs, 4)] = FMA(KP1_200954543, T1e, T1d);
Chris@42 165 R1[WS(rs, 2)] = FNMS(KP1_200954543, T1e, T1d);
Chris@42 166 R0[WS(rs, 6)] = FMA(KP1_200954543, TO, Tr);
Chris@42 167 R1[0] = FNMS(KP1_200954543, TO, Tr);
Chris@42 168 {
Chris@42 169 E T1b, T15, T11, TT;
Chris@42 170 T1b = FNMS(KP1_040057143, T14, T13);
Chris@42 171 T15 = FMA(KP1_040057143, T14, T13);
Chris@42 172 T11 = FMA(KP1_150281458, TS, TR);
Chris@42 173 TT = FNMS(KP1_150281458, TS, TR);
Chris@42 174 {
Chris@42 175 E T1c, T1a, T12, T10;
Chris@42 176 T1c = FMA(KP875502302, T19, T18);
Chris@42 177 T1a = FNMS(KP875502302, T19, T18);
Chris@42 178 T12 = FMA(KP968287244, TZ, TW);
Chris@42 179 T10 = FNMS(KP968287244, TZ, TW);
Chris@42 180 R1[WS(rs, 5)] = FMA(KP1_150281458, T1c, T1b);
Chris@42 181 R0[WS(rs, 3)] = FNMS(KP1_150281458, T1c, T1b);
Chris@42 182 R1[WS(rs, 3)] = FMA(KP1_150281458, T1a, T15);
Chris@42 183 R0[WS(rs, 1)] = FNMS(KP1_150281458, T1a, T15);
Chris@42 184 R0[WS(rs, 5)] = FMA(KP1_040057143, T12, T11);
Chris@42 185 R0[WS(rs, 2)] = FNMS(KP1_040057143, T12, T11);
Chris@42 186 R1[WS(rs, 4)] = FMA(KP1_040057143, T10, TT);
Chris@42 187 R1[WS(rs, 1)] = FNMS(KP1_040057143, T10, TT);
Chris@42 188 }
Chris@42 189 }
Chris@42 190 }
Chris@42 191 }
Chris@42 192 }
Chris@42 193 }
Chris@42 194
Chris@42 195 static const kr2c_desc desc = { 13, "r2cb_13", {18, 0, 58, 0}, &GENUS };
Chris@42 196
Chris@42 197 void X(codelet_r2cb_13) (planner *p) {
Chris@42 198 X(kr2c_register) (p, r2cb_13, &desc);
Chris@42 199 }
Chris@42 200
Chris@42 201 #else /* HAVE_FMA */
Chris@42 202
Chris@42 203 /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 13 -name r2cb_13 -include r2cb.h */
Chris@42 204
Chris@42 205 /*
Chris@42 206 * This function contains 76 FP additions, 35 FP multiplications,
Chris@42 207 * (or, 56 additions, 15 multiplications, 20 fused multiply/add),
Chris@42 208 * 56 stack variables, 19 constants, and 26 memory accesses
Chris@42 209 */
Chris@42 210 #include "r2cb.h"
Chris@42 211
Chris@42 212 static void r2cb_13(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
Chris@42 213 {
Chris@42 214 DK(KP1_007074065, +1.007074065727533254493747707736933954186697125);
Chris@42 215 DK(KP227708958, +0.227708958111581597949308691735310621069285120);
Chris@42 216 DK(KP531932498, +0.531932498429674575175042127684371897596660533);
Chris@42 217 DK(KP774781170, +0.774781170935234584261351932853525703557550433);
Chris@42 218 DK(KP265966249, +0.265966249214837287587521063842185948798330267);
Chris@42 219 DK(KP516520780, +0.516520780623489722840901288569017135705033622);
Chris@42 220 DK(KP151805972, +0.151805972074387731966205794490207080712856746);
Chris@42 221 DK(KP503537032, +0.503537032863766627246873853868466977093348562);
Chris@42 222 DK(KP166666666, +0.166666666666666666666666666666666666666666667);
Chris@42 223 DK(KP600925212, +0.600925212577331548853203544578415991041882762);
Chris@42 224 DK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@42 225 DK(KP256247671, +0.256247671582936600958684654061725059144125175);
Chris@42 226 DK(KP156891391, +0.156891391051584611046832726756003269660212636);
Chris@42 227 DK(KP348277202, +0.348277202304271810011321589858529485233929352);
Chris@42 228 DK(KP1_150281458, +1.150281458948006242736771094910906776922003215);
Chris@42 229 DK(KP300238635, +0.300238635966332641462884626667381504676006424);
Chris@42 230 DK(KP011599105, +0.011599105605768290721655456654083252189827041);
Chris@42 231 DK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
Chris@42 232 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
Chris@42 233 {
Chris@42 234 INT i;
Chris@42 235 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) {
Chris@42 236 E TG, TS, TR, T15, TJ, TT, T1, Tm, Tc, Td, Tg, Tj, Tk, Tn, To;
Chris@42 237 E Tp;
Chris@42 238 {
Chris@42 239 E Ts, Tv, Tw, TE, TC, TB, Tz, TD, TA, TF;
Chris@42 240 {
Chris@42 241 E Tt, Tu, Tx, Ty;
Chris@42 242 Ts = Ci[WS(csi, 1)];
Chris@42 243 Tt = Ci[WS(csi, 3)];
Chris@42 244 Tu = Ci[WS(csi, 4)];
Chris@42 245 Tv = Tt - Tu;
Chris@42 246 Tw = FMS(KP2_000000000, Ts, Tv);
Chris@42 247 TE = KP1_732050807 * (Tt + Tu);
Chris@42 248 TC = Ci[WS(csi, 5)];
Chris@42 249 Tx = Ci[WS(csi, 6)];
Chris@42 250 Ty = Ci[WS(csi, 2)];
Chris@42 251 TB = Tx + Ty;
Chris@42 252 Tz = KP1_732050807 * (Tx - Ty);
Chris@42 253 TD = FNMS(KP2_000000000, TC, TB);
Chris@42 254 }
Chris@42 255 TA = Tw + Tz;
Chris@42 256 TF = TD - TE;
Chris@42 257 TG = FMA(KP011599105, TA, KP300238635 * TF);
Chris@42 258 TS = FNMS(KP011599105, TF, KP300238635 * TA);
Chris@42 259 {
Chris@42 260 E TP, TQ, TH, TI;
Chris@42 261 TP = Ts + Tv;
Chris@42 262 TQ = TB + TC;
Chris@42 263 TR = FNMS(KP348277202, TQ, KP1_150281458 * TP);
Chris@42 264 T15 = FMA(KP348277202, TP, KP1_150281458 * TQ);
Chris@42 265 TH = Tw - Tz;
Chris@42 266 TI = TE + TD;
Chris@42 267 TJ = FMA(KP156891391, TH, KP256247671 * TI);
Chris@42 268 TT = FNMS(KP256247671, TH, KP156891391 * TI);
Chris@42 269 }
Chris@42 270 }
Chris@42 271 {
Chris@42 272 E Tb, Ti, Tf, T6, Th, Te;
Chris@42 273 T1 = Cr[0];
Chris@42 274 {
Chris@42 275 E T7, T8, T9, Ta;
Chris@42 276 T7 = Cr[WS(csr, 5)];
Chris@42 277 T8 = Cr[WS(csr, 2)];
Chris@42 278 T9 = Cr[WS(csr, 6)];
Chris@42 279 Ta = T8 + T9;
Chris@42 280 Tb = T7 + Ta;
Chris@42 281 Ti = FNMS(KP500000000, Ta, T7);
Chris@42 282 Tf = T8 - T9;
Chris@42 283 }
Chris@42 284 {
Chris@42 285 E T2, T3, T4, T5;
Chris@42 286 T2 = Cr[WS(csr, 1)];
Chris@42 287 T3 = Cr[WS(csr, 3)];
Chris@42 288 T4 = Cr[WS(csr, 4)];
Chris@42 289 T5 = T3 + T4;
Chris@42 290 T6 = T2 + T5;
Chris@42 291 Th = FNMS(KP500000000, T5, T2);
Chris@42 292 Te = T3 - T4;
Chris@42 293 }
Chris@42 294 Tm = KP600925212 * (T6 - Tb);
Chris@42 295 Tc = T6 + Tb;
Chris@42 296 Td = FNMS(KP166666666, Tc, T1);
Chris@42 297 Tg = Te + Tf;
Chris@42 298 Tj = Th + Ti;
Chris@42 299 Tk = FMA(KP503537032, Tg, KP151805972 * Tj);
Chris@42 300 Tn = Th - Ti;
Chris@42 301 To = Te - Tf;
Chris@42 302 Tp = FNMS(KP265966249, To, KP516520780 * Tn);
Chris@42 303 }
Chris@42 304 R0[0] = FMA(KP2_000000000, Tc, T1);
Chris@42 305 {
Chris@42 306 E TK, T1b, TV, T12, T16, T18, TO, T1a, Tr, T17, T11, T13;
Chris@42 307 {
Chris@42 308 E TU, T14, TM, TN;
Chris@42 309 TK = KP1_732050807 * (TG + TJ);
Chris@42 310 T1b = KP1_732050807 * (TS - TT);
Chris@42 311 TU = TS + TT;
Chris@42 312 TV = TR - TU;
Chris@42 313 T12 = FMA(KP2_000000000, TU, TR);
Chris@42 314 T14 = TG - TJ;
Chris@42 315 T16 = FMS(KP2_000000000, T14, T15);
Chris@42 316 T18 = T14 + T15;
Chris@42 317 TM = FMA(KP774781170, To, KP531932498 * Tn);
Chris@42 318 TN = FNMS(KP1_007074065, Tj, KP227708958 * Tg);
Chris@42 319 TO = TM - TN;
Chris@42 320 T1a = TM + TN;
Chris@42 321 {
Chris@42 322 E Tl, Tq, TZ, T10;
Chris@42 323 Tl = Td - Tk;
Chris@42 324 Tq = Tm - Tp;
Chris@42 325 Tr = Tl - Tq;
Chris@42 326 T17 = Tq + Tl;
Chris@42 327 TZ = FMA(KP2_000000000, Tk, Td);
Chris@42 328 T10 = FMA(KP2_000000000, Tp, Tm);
Chris@42 329 T11 = TZ - T10;
Chris@42 330 T13 = T10 + TZ;
Chris@42 331 }
Chris@42 332 }
Chris@42 333 R1[WS(rs, 2)] = T11 - T12;
Chris@42 334 R0[WS(rs, 6)] = T13 - T16;
Chris@42 335 R1[0] = T13 + T16;
Chris@42 336 R0[WS(rs, 4)] = T11 + T12;
Chris@42 337 {
Chris@42 338 E TL, TW, T19, T1c;
Chris@42 339 TL = Tr - TK;
Chris@42 340 TW = TO - TV;
Chris@42 341 R1[WS(rs, 3)] = TL - TW;
Chris@42 342 R0[WS(rs, 1)] = TL + TW;
Chris@42 343 T19 = T17 - T18;
Chris@42 344 T1c = T1a + T1b;
Chris@42 345 R1[WS(rs, 1)] = T19 - T1c;
Chris@42 346 R1[WS(rs, 4)] = T1c + T19;
Chris@42 347 }
Chris@42 348 {
Chris@42 349 E T1d, T1e, TX, TY;
Chris@42 350 T1d = T1a - T1b;
Chris@42 351 T1e = T17 + T18;
Chris@42 352 R0[WS(rs, 2)] = T1d + T1e;
Chris@42 353 R0[WS(rs, 5)] = T1e - T1d;
Chris@42 354 TX = Tr + TK;
Chris@42 355 TY = TO + TV;
Chris@42 356 R0[WS(rs, 3)] = TX - TY;
Chris@42 357 R1[WS(rs, 5)] = TX + TY;
Chris@42 358 }
Chris@42 359 }
Chris@42 360 }
Chris@42 361 }
Chris@42 362 }
Chris@42 363
Chris@42 364 static const kr2c_desc desc = { 13, "r2cb_13", {56, 15, 20, 0}, &GENUS };
Chris@42 365
Chris@42 366 void X(codelet_r2cb_13) (planner *p) {
Chris@42 367 X(kr2c_register) (p, r2cb_13, &desc);
Chris@42 368 }
Chris@42 369
Chris@42 370 #endif /* HAVE_FMA */