Mercurial > hg > sv-dependency-builds

annotate src/fftw-3.3.5/dft/simd/common/n1fv_13.c @ 84:08ae793730bd

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add null config files
author Chris Cannam
date Mon, 02 Mar 2020 14:03:47 +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 16:38:41 EDT 2016 */
Chris@42 23
Chris@42 24 #include "codelet-dft.h"
Chris@42 25
Chris@42 26 #ifdef HAVE_FMA
Chris@42 27
Chris@42 28 /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 13 -name n1fv_13 -include n1f.h */
Chris@42 29
Chris@42 30 /*
Chris@42 31 * This function contains 88 FP additions, 63 FP multiplications,
Chris@42 32 * (or, 31 additions, 6 multiplications, 57 fused multiply/add),
Chris@42 33 * 96 stack variables, 23 constants, and 26 memory accesses
Chris@42 34 */
Chris@42 35 #include "n1f.h"
Chris@42 36
Chris@42 37 static void n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@42 38 {
Chris@42 39 DVK(KP904176221, +0.904176221990848204433795481776887926501523162);
Chris@42 40 DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
Chris@42 41 DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
Chris@42 42 DVK(KP516520780, +0.516520780623489722840901288569017135705033622);
Chris@42 43 DVK(KP522026385, +0.522026385161275033714027226654165028300441940);
Chris@42 44 DVK(KP957805992, +0.957805992594665126462521754605754580515587217);
Chris@42 45 DVK(KP600477271, +0.600477271932665282925769253334763009352012849);
Chris@42 46 DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
Chris@42 47 DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
Chris@42 48 DVK(KP769338817, +0.769338817572980603471413688209101117038278899);
Chris@42 49 DVK(KP859542535, +0.859542535098774820163672132761689612766401925);
Chris@42 50 DVK(KP581704778, +0.581704778510515730456870384989698884939833902);
Chris@42 51 DVK(KP853480001, +0.853480001859823990758994934970528322872359049);
Chris@42 52 DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
Chris@42 53 DVK(KP226109445, +0.226109445035782405468510155372505010481906348);
Chris@42 54 DVK(KP301479260, +0.301479260047709873958013540496673347309208464);
Chris@42 55 DVK(KP686558370, +0.686558370781754340655719594850823015421401653);
Chris@42 56 DVK(KP514918778, +0.514918778086315755491789696138117261566051239);
Chris@42 57 DVK(KP038632954, +0.038632954644348171955506895830342264440241080);
Chris@42 58 DVK(KP612264650, +0.612264650376756543746494474777125408779395514);
Chris@42 59 DVK(KP302775637, +0.302775637731994646559610633735247973125648287);
Chris@42 60 DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
Chris@42 61 DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@42 62 {
Chris@42 63 INT i;
Chris@42 64 const R *xi;
Chris@42 65 R *xo;
Chris@42 66 xi = ri;
Chris@42 67 xo = ro;
Chris@42 68 for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(26, is), MAKE_VOLATILE_STRIDE(26, os)) {
Chris@42 69 V T1, T7, T2, Tg, Tf, TN, Th, Tq, Ta, Tj, T5, Tr, Tk;
Chris@42 70 T1 = LD(&(xi[0]), ivs, &(xi[0]));
Chris@42 71 {
Chris@42 72 V Td, Te, T8, T9, T3, T4;
Chris@42 73 Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Chris@42 74 Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Chris@42 75 T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Chris@42 76 T8 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Chris@42 77 T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Chris@42 78 T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Chris@42 79 T3 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Chris@42 80 T4 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Chris@42 81 Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Chris@42 82 Tf = VADD(Td, Te);
Chris@42 83 TN = VSUB(Td, Te);
Chris@42 84 Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Chris@42 85 Tq = VSUB(T8, T9);
Chris@42 86 Ta = VADD(T8, T9);
Chris@42 87 Tj = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Chris@42 88 T5 = VADD(T3, T4);
Chris@42 89 Tr = VSUB(T4, T3);
Chris@42 90 Tk = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Chris@42 91 }
Chris@42 92 {
Chris@42 93 V Tt, Ti, Ty, Tb, Ts, TQ, Tx, T6, Tu, Tl;
Chris@42 94 Tt = VSUB(Tg, Th);
Chris@42 95 Ti = VADD(Tg, Th);
Chris@42 96 Ty = VFMS(LDK(KP500000000), Ta, T7);
Chris@42 97 Tb = VADD(T7, Ta);
Chris@42 98 Ts = VSUB(Tq, Tr);
Chris@42 99 TQ = VADD(Tr, Tq);
Chris@42 100 Tx = VFNMS(LDK(KP500000000), T5, T2);
Chris@42 101 T6 = VADD(T2, T5);
Chris@42 102 Tu = VSUB(Tj, Tk);
Chris@42 103 Tl = VADD(Tj, Tk);
Chris@42 104 {
Chris@42 105 V TK, Tz, Tc, TX, Tv, TO, TL, Tm;
Chris@42 106 TK = VADD(Tx, Ty);
Chris@42 107 Tz = VSUB(Tx, Ty);
Chris@42 108 Tc = VADD(T6, Tb);
Chris@42 109 TX = VSUB(T6, Tb);
Chris@42 110 Tv = VSUB(Tt, Tu);
Chris@42 111 TO = VADD(Tt, Tu);
Chris@42 112 TL = VSUB(Ti, Tl);
Chris@42 113 Tm = VADD(Ti, Tl);
Chris@42 114 {
Chris@42 115 V TF, Tw, TP, TY, TT, TM, TA, Tn;
Chris@42 116 TF = VSUB(Ts, Tv);
Chris@42 117 Tw = VADD(Ts, Tv);
Chris@42 118 TP = VFNMS(LDK(KP500000000), TO, TN);
Chris@42 119 TY = VADD(TN, TO);
Chris@42 120 TT = VFNMS(LDK(KP866025403), TL, TK);
Chris@42 121 TM = VFMA(LDK(KP866025403), TL, TK);
Chris@42 122 TA = VFNMS(LDK(KP500000000), Tm, Tf);
Chris@42 123 Tn = VADD(Tf, Tm);
Chris@42 124 {
Chris@42 125 V T1f, T1n, TI, T18, T1k, T1c, TD, T17, T10, T1m, T16, T1e, TU, TR;
Chris@42 126 TU = VFNMS(LDK(KP866025403), TQ, TP);
Chris@42 127 TR = VFMA(LDK(KP866025403), TQ, TP);
Chris@42 128 {
Chris@42 129 V TZ, T15, TE, TB;
Chris@42 130 TZ = VFMA(LDK(KP302775637), TY, TX);
Chris@42 131 T15 = VFNMS(LDK(KP302775637), TX, TY);
Chris@42 132 TE = VSUB(Tz, TA);
Chris@42 133 TB = VADD(Tz, TA);
Chris@42 134 {
Chris@42 135 V TH, To, TV, T13;
Chris@42 136 TH = VSUB(Tc, Tn);
Chris@42 137 To = VADD(Tc, Tn);
Chris@42 138 TV = VFNMS(LDK(KP612264650), TU, TT);
Chris@42 139 T13 = VFMA(LDK(KP612264650), TT, TU);
Chris@42 140 {
Chris@42 141 V TS, T12, TG, T1b;
Chris@42 142 TS = VFNMS(LDK(KP038632954), TR, TM);
Chris@42 143 T12 = VFMA(LDK(KP038632954), TM, TR);
Chris@42 144 TG = VFNMS(LDK(KP514918778), TF, TE);
Chris@42 145 T1b = VFMA(LDK(KP686558370), TE, TF);
Chris@42 146 {
Chris@42 147 V TC, T1a, Tp, TW, T14;
Chris@42 148 TC = VFMA(LDK(KP301479260), TB, Tw);
Chris@42 149 T1a = VFNMS(LDK(KP226109445), Tw, TB);
Chris@42 150 Tp = VFNMS(LDK(KP083333333), To, T1);
Chris@42 151 ST(&(xo[0]), VADD(T1, To), ovs, &(xo[0]));
Chris@42 152 T1f = VFMA(LDK(KP853480001), TV, TS);
Chris@42 153 TW = VFNMS(LDK(KP853480001), TV, TS);
Chris@42 154 T1n = VFMA(LDK(KP853480001), T13, T12);
Chris@42 155 T14 = VFNMS(LDK(KP853480001), T13, T12);
Chris@42 156 TI = VFMA(LDK(KP581704778), TH, TG);
Chris@42 157 T18 = VFNMS(LDK(KP859542535), TG, TH);
Chris@42 158 T1k = VFMA(LDK(KP769338817), T1b, T1a);
Chris@42 159 T1c = VFNMS(LDK(KP769338817), T1b, T1a);
Chris@42 160 TD = VFMA(LDK(KP503537032), TC, Tp);
Chris@42 161 T17 = VFNMS(LDK(KP251768516), TC, Tp);
Chris@42 162 T10 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), TZ, TW));
Chris@42 163 T1m = VFNMS(LDK(KP522026385), TW, TZ);
Chris@42 164 T16 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), T15, T14));
Chris@42 165 T1e = VFNMS(LDK(KP522026385), T14, T15);
Chris@42 166 }
Chris@42 167 }
Chris@42 168 }
Chris@42 169 }
Chris@42 170 {
Chris@42 171 V T1o, T1q, T1g, T1i, T1d, T1h, T1l, T1p;
Chris@42 172 {
Chris@42 173 V T11, TJ, T19, T1j;
Chris@42 174 T11 = VFMA(LDK(KP516520780), TI, TD);
Chris@42 175 TJ = VFNMS(LDK(KP516520780), TI, TD);
Chris@42 176 T19 = VFMA(LDK(KP300462606), T18, T17);
Chris@42 177 T1j = VFNMS(LDK(KP300462606), T18, T17);
Chris@42 178 T1o = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1n, T1m));
Chris@42 179 T1q = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1n, T1m));
Chris@42 180 T1g = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1f, T1e));
Chris@42 181 T1i = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1f, T1e));
Chris@42 182 ST(&(xo[WS(os, 12)]), VFNMSI(T16, T11), ovs, &(xo[0]));
Chris@42 183 ST(&(xo[WS(os, 1)]), VFMAI(T16, T11), ovs, &(xo[WS(os, 1)]));
Chris@42 184 ST(&(xo[WS(os, 8)]), VFMAI(T10, TJ), ovs, &(xo[0]));
Chris@42 185 ST(&(xo[WS(os, 5)]), VFNMSI(T10, TJ), ovs, &(xo[WS(os, 1)]));
Chris@42 186 T1d = VFNMS(LDK(KP503537032), T1c, T19);
Chris@42 187 T1h = VFMA(LDK(KP503537032), T1c, T19);
Chris@42 188 T1l = VFNMS(LDK(KP503537032), T1k, T1j);
Chris@42 189 T1p = VFMA(LDK(KP503537032), T1k, T1j);
Chris@42 190 }
Chris@42 191 ST(&(xo[WS(os, 9)]), VFMAI(T1g, T1d), ovs, &(xo[WS(os, 1)]));
Chris@42 192 ST(&(xo[WS(os, 4)]), VFNMSI(T1g, T1d), ovs, &(xo[0]));
Chris@42 193 ST(&(xo[WS(os, 10)]), VFNMSI(T1i, T1h), ovs, &(xo[0]));
Chris@42 194 ST(&(xo[WS(os, 3)]), VFMAI(T1i, T1h), ovs, &(xo[WS(os, 1)]));
Chris@42 195 ST(&(xo[WS(os, 7)]), VFMAI(T1o, T1l), ovs, &(xo[WS(os, 1)]));
Chris@42 196 ST(&(xo[WS(os, 6)]), VFNMSI(T1o, T1l), ovs, &(xo[0]));
Chris@42 197 ST(&(xo[WS(os, 11)]), VFMAI(T1q, T1p), ovs, &(xo[WS(os, 1)]));
Chris@42 198 ST(&(xo[WS(os, 2)]), VFNMSI(T1q, T1p), ovs, &(xo[0]));
Chris@42 199 }
Chris@42 200 }
Chris@42 201 }
Chris@42 202 }
Chris@42 203 }
Chris@42 204 }
Chris@42 205 }
Chris@42 206 VLEAVE();
Chris@42 207 }
Chris@42 208
Chris@42 209 static const kdft_desc desc = { 13, XSIMD_STRING("n1fv_13"), {31, 6, 57, 0}, &GENUS, 0, 0, 0, 0 };
Chris@42 210
Chris@42 211 void XSIMD(codelet_n1fv_13) (planner *p) {
Chris@42 212 X(kdft_register) (p, n1fv_13, &desc);
Chris@42 213 }
Chris@42 214
Chris@42 215 #else /* HAVE_FMA */
Chris@42 216
Chris@42 217 /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 13 -name n1fv_13 -include n1f.h */
Chris@42 218
Chris@42 219 /*
Chris@42 220 * This function contains 88 FP additions, 34 FP multiplications,
Chris@42 221 * (or, 69 additions, 15 multiplications, 19 fused multiply/add),
Chris@42 222 * 60 stack variables, 20 constants, and 26 memory accesses
Chris@42 223 */
Chris@42 224 #include "n1f.h"
Chris@42 225
Chris@42 226 static void n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@42 227 {
Chris@42 228 DVK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
Chris@42 229 DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
Chris@42 230 DVK(KP075902986, +0.075902986037193865983102897245103540356428373);
Chris@42 231 DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
Chris@42 232 DVK(KP132983124, +0.132983124607418643793760531921092974399165133);
Chris@42 233 DVK(KP258260390, +0.258260390311744861420450644284508567852516811);
Chris@42 234 DVK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
Chris@42 235 DVK(KP300238635, +0.300238635966332641462884626667381504676006424);
Chris@42 236 DVK(KP011599105, +0.011599105605768290721655456654083252189827041);
Chris@42 237 DVK(KP156891391, +0.156891391051584611046832726756003269660212636);
Chris@42 238 DVK(KP256247671, +0.256247671582936600958684654061725059144125175);
Chris@42 239 DVK(KP174138601, +0.174138601152135905005660794929264742616964676);
Chris@42 240 DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
Chris@42 241 DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
Chris@42 242 DVK(KP113854479, +0.113854479055790798974654345867655310534642560);
Chris@42 243 DVK(KP265966249, +0.265966249214837287587521063842185948798330267);
Chris@42 244 DVK(KP387390585, +0.387390585467617292130675966426762851778775217);
Chris@42 245 DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
Chris@42 246 DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
Chris@42 247 DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@42 248 {
Chris@42 249 INT i;
Chris@42 250 const R *xi;
Chris@42 251 R *xo;
Chris@42 252 xi = ri;
Chris@42 253 xo = ro;
Chris@42 254 for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(26, is), MAKE_VOLATILE_STRIDE(26, os)) {
Chris@42 255 V TW, Tb, Tm, Tu, TC, TR, TX, TK, TU, Tz, TB, TN, TT;
Chris@42 256 TW = LD(&(xi[0]), ivs, &(xi[0]));
Chris@42 257 {
Chris@42 258 V T3, TH, Tl, Tw, Tp, Tg, Tv, To, T6, Tr, T9, Ts, Ta, TI, T1;
Chris@42 259 V T2, Tq, Tt;
Chris@42 260 T1 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Chris@42 261 T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Chris@42 262 T3 = VSUB(T1, T2);
Chris@42 263 TH = VADD(T1, T2);
Chris@42 264 {
Chris@42 265 V Th, Ti, Tj, Tk;
Chris@42 266 Th = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Chris@42 267 Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Chris@42 268 Tj = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Chris@42 269 Tk = VADD(Ti, Tj);
Chris@42 270 Tl = VADD(Th, Tk);
Chris@42 271 Tw = VSUB(Ti, Tj);
Chris@42 272 Tp = VFNMS(LDK(KP500000000), Tk, Th);
Chris@42 273 }
Chris@42 274 {
Chris@42 275 V Tc, Td, Te, Tf;
Chris@42 276 Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Chris@42 277 Td = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Chris@42 278 Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Chris@42 279 Tf = VADD(Td, Te);
Chris@42 280 Tg = VADD(Tc, Tf);
Chris@42 281 Tv = VSUB(Td, Te);
Chris@42 282 To = VFNMS(LDK(KP500000000), Tf, Tc);
Chris@42 283 }
Chris@42 284 {
Chris@42 285 V T4, T5, T7, T8;
Chris@42 286 T4 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Chris@42 287 T5 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Chris@42 288 T6 = VSUB(T4, T5);
Chris@42 289 Tr = VADD(T4, T5);
Chris@42 290 T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Chris@42 291 T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Chris@42 292 T9 = VSUB(T7, T8);
Chris@42 293 Ts = VADD(T7, T8);
Chris@42 294 }
Chris@42 295 Ta = VADD(T6, T9);
Chris@42 296 TI = VADD(Tr, Ts);
Chris@42 297 Tb = VADD(T3, Ta);
Chris@42 298 Tm = VSUB(Tg, Tl);
Chris@42 299 Tq = VSUB(To, Tp);
Chris@42 300 Tt = VMUL(LDK(KP866025403), VSUB(Tr, Ts));
Chris@42 301 Tu = VADD(Tq, Tt);
Chris@42 302 TC = VSUB(Tq, Tt);
Chris@42 303 {
Chris@42 304 V TP, TQ, TG, TJ;
Chris@42 305 TP = VADD(Tg, Tl);
Chris@42 306 TQ = VADD(TH, TI);
Chris@42 307 TR = VMUL(LDK(KP300462606), VSUB(TP, TQ));
Chris@42 308 TX = VADD(TP, TQ);
Chris@42 309 TG = VADD(To, Tp);
Chris@42 310 TJ = VFNMS(LDK(KP500000000), TI, TH);
Chris@42 311 TK = VSUB(TG, TJ);
Chris@42 312 TU = VADD(TG, TJ);
Chris@42 313 }
Chris@42 314 {
Chris@42 315 V Tx, Ty, TL, TM;
Chris@42 316 Tx = VMUL(LDK(KP866025403), VSUB(Tv, Tw));
Chris@42 317 Ty = VFNMS(LDK(KP500000000), Ta, T3);
Chris@42 318 Tz = VSUB(Tx, Ty);
Chris@42 319 TB = VADD(Tx, Ty);
Chris@42 320 TL = VADD(Tv, Tw);
Chris@42 321 TM = VSUB(T6, T9);
Chris@42 322 TN = VSUB(TL, TM);
Chris@42 323 TT = VADD(TL, TM);
Chris@42 324 }
Chris@42 325 }
Chris@42 326 ST(&(xo[0]), VADD(TW, TX), ovs, &(xo[0]));
Chris@42 327 {
Chris@42 328 V T19, T1n, T14, T13, T1f, T1k, Tn, TE, T1e, T1j, TS, T1m, TZ, T1c, TA;
Chris@42 329 V TD;
Chris@42 330 {
Chris@42 331 V T17, T18, T11, T12;
Chris@42 332 T17 = VFMA(LDK(KP387390585), TN, VMUL(LDK(KP265966249), TK));
Chris@42 333 T18 = VFNMS(LDK(KP503537032), TU, VMUL(LDK(KP113854479), TT));
Chris@42 334 T19 = VSUB(T17, T18);
Chris@42 335 T1n = VADD(T17, T18);
Chris@42 336 T14 = VFMA(LDK(KP575140729), Tm, VMUL(LDK(KP174138601), Tb));
Chris@42 337 T11 = VFNMS(LDK(KP156891391), TB, VMUL(LDK(KP256247671), TC));
Chris@42 338 T12 = VFMA(LDK(KP011599105), Tz, VMUL(LDK(KP300238635), Tu));
Chris@42 339 T13 = VSUB(T11, T12);
Chris@42 340 T1f = VADD(T14, T13);
Chris@42 341 T1k = VMUL(LDK(KP1_732050807), VADD(T11, T12));
Chris@42 342 }
Chris@42 343 Tn = VFNMS(LDK(KP174138601), Tm, VMUL(LDK(KP575140729), Tb));
Chris@42 344 TA = VFNMS(LDK(KP300238635), Tz, VMUL(LDK(KP011599105), Tu));
Chris@42 345 TD = VFMA(LDK(KP256247671), TB, VMUL(LDK(KP156891391), TC));
Chris@42 346 TE = VSUB(TA, TD);
Chris@42 347 T1e = VMUL(LDK(KP1_732050807), VADD(TD, TA));
Chris@42 348 T1j = VSUB(Tn, TE);
Chris@42 349 {
Chris@42 350 V TO, T1b, TV, TY, T1a;
Chris@42 351 TO = VFNMS(LDK(KP132983124), TN, VMUL(LDK(KP258260390), TK));
Chris@42 352 T1b = VSUB(TR, TO);
Chris@42 353 TV = VFMA(LDK(KP251768516), TT, VMUL(LDK(KP075902986), TU));
Chris@42 354 TY = VFNMS(LDK(KP083333333), TX, TW);
Chris@42 355 T1a = VSUB(TY, TV);
Chris@42 356 TS = VFMA(LDK(KP2_000000000), TO, TR);
Chris@42 357 T1m = VADD(T1b, T1a);
Chris@42 358 TZ = VFMA(LDK(KP2_000000000), TV, TY);
Chris@42 359 T1c = VSUB(T1a, T1b);
Chris@42 360 }
Chris@42 361 {
Chris@42 362 V TF, T10, T1l, T1o;
Chris@42 363 TF = VBYI(VFMA(LDK(KP2_000000000), TE, Tn));
Chris@42 364 T10 = VADD(TS, TZ);
Chris@42 365 ST(&(xo[WS(os, 1)]), VADD(TF, T10), ovs, &(xo[WS(os, 1)]));
Chris@42 366 ST(&(xo[WS(os, 12)]), VSUB(T10, TF), ovs, &(xo[0]));
Chris@42 367 {
Chris@42 368 V T15, T16, T1p, T1q;
Chris@42 369 T15 = VBYI(VFMS(LDK(KP2_000000000), T13, T14));
Chris@42 370 T16 = VSUB(TZ, TS);
Chris@42 371 ST(&(xo[WS(os, 5)]), VADD(T15, T16), ovs, &(xo[WS(os, 1)]));
Chris@42 372 ST(&(xo[WS(os, 8)]), VSUB(T16, T15), ovs, &(xo[0]));
Chris@42 373 T1p = VADD(T1n, T1m);
Chris@42 374 T1q = VBYI(VADD(T1j, T1k));
Chris@42 375 ST(&(xo[WS(os, 4)]), VSUB(T1p, T1q), ovs, &(xo[0]));
Chris@42 376 ST(&(xo[WS(os, 9)]), VADD(T1q, T1p), ovs, &(xo[WS(os, 1)]));
Chris@42 377 }
Chris@42 378 T1l = VBYI(VSUB(T1j, T1k));
Chris@42 379 T1o = VSUB(T1m, T1n);
Chris@42 380 ST(&(xo[WS(os, 3)]), VADD(T1l, T1o), ovs, &(xo[WS(os, 1)]));
Chris@42 381 ST(&(xo[WS(os, 10)]), VSUB(T1o, T1l), ovs, &(xo[0]));
Chris@42 382 {
Chris@42 383 V T1h, T1i, T1d, T1g;
Chris@42 384 T1h = VBYI(VSUB(T1e, T1f));
Chris@42 385 T1i = VSUB(T1c, T19);
Chris@42 386 ST(&(xo[WS(os, 6)]), VADD(T1h, T1i), ovs, &(xo[0]));
Chris@42 387 ST(&(xo[WS(os, 7)]), VSUB(T1i, T1h), ovs, &(xo[WS(os, 1)]));
Chris@42 388 T1d = VADD(T19, T1c);
Chris@42 389 T1g = VBYI(VADD(T1e, T1f));
Chris@42 390 ST(&(xo[WS(os, 2)]), VSUB(T1d, T1g), ovs, &(xo[0]));
Chris@42 391 ST(&(xo[WS(os, 11)]), VADD(T1g, T1d), ovs, &(xo[WS(os, 1)]));
Chris@42 392 }
Chris@42 393 }
Chris@42 394 }
Chris@42 395 }
Chris@42 396 }
Chris@42 397 VLEAVE();
Chris@42 398 }
Chris@42 399
Chris@42 400 static const kdft_desc desc = { 13, XSIMD_STRING("n1fv_13"), {69, 15, 19, 0}, &GENUS, 0, 0, 0, 0 };
Chris@42 401
Chris@42 402 void XSIMD(codelet_n1fv_13) (planner *p) {
Chris@42 403 X(kdft_register) (p, n1fv_13, &desc);
Chris@42 404 }
Chris@42 405
Chris@42 406 #endif /* HAVE_FMA */