Mercurial > hg > sv-dependency-builds

annotate src/fftw-3.3.8/dft/simd/common/n2bv_14.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 08:05:11 EDT 2018 */
Chris@82 23
Chris@82 24 #include "dft/codelet-dft.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_notw_c.native -fma -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 14 -name n2bv_14 -with-ostride 2 -include dft/simd/n2b.h -store-multiple 2 */
Chris@82 29
Chris@82 30 /*
Chris@82 31 * This function contains 74 FP additions, 48 FP multiplications,
Chris@82 32 * (or, 32 additions, 6 multiplications, 42 fused multiply/add),
Chris@82 33 * 51 stack variables, 6 constants, and 35 memory accesses
Chris@82 34 */
Chris@82 35 #include "dft/simd/n2b.h"
Chris@82 36
Chris@82 37 static void n2bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@82 38 {
Chris@82 39 DVK(KP801937735, +0.801937735804838252472204639014890102331838324);
Chris@82 40 DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
Chris@82 41 DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
Chris@82 42 DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
Chris@82 43 DVK(KP692021471, +0.692021471630095869627814897002069140197260599);
Chris@82 44 DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
Chris@82 45 {
Chris@82 46 INT i;
Chris@82 47 const R *xi;
Chris@82 48 R *xo;
Chris@82 49 xi = ii;
Chris@82 50 xo = io;
Chris@82 51 for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) {
Chris@82 52 V T3, TH, Ts, TV, TW, Tt, Tu, TU, Ta, To, Th, Tp, TC, Tx, TK;
Chris@82 53 V TQ, TN, TR, T14, TZ, T1, T2;
Chris@82 54 T1 = LD(&(xi[0]), ivs, &(xi[0]));
Chris@82 55 T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Chris@82 56 T3 = VSUB(T1, T2);
Chris@82 57 TH = VADD(T1, T2);
Chris@82 58 {
Chris@82 59 V T6, TI, T9, TJ, Tn, TP, Tk, TO, Tg, TM, Td, TL;
Chris@82 60 {
Chris@82 61 V T4, T5, Ti, Tj;
Chris@82 62 T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Chris@82 63 T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Chris@82 64 T6 = VSUB(T4, T5);
Chris@82 65 TI = VADD(T4, T5);
Chris@82 66 {
Chris@82 67 V T7, T8, Tl, Tm;
Chris@82 68 T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Chris@82 69 T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Chris@82 70 T9 = VSUB(T7, T8);
Chris@82 71 TJ = VADD(T7, T8);
Chris@82 72 Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Chris@82 73 Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Chris@82 74 Tn = VSUB(Tl, Tm);
Chris@82 75 TP = VADD(Tl, Tm);
Chris@82 76 }
Chris@82 77 Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Chris@82 78 Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Chris@82 79 Tk = VSUB(Ti, Tj);
Chris@82 80 TO = VADD(Ti, Tj);
Chris@82 81 {
Chris@82 82 V Te, Tf, Tb, Tc;
Chris@82 83 Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Chris@82 84 Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Chris@82 85 Tg = VSUB(Te, Tf);
Chris@82 86 TM = VADD(Te, Tf);
Chris@82 87 Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Chris@82 88 Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Chris@82 89 Td = VSUB(Tb, Tc);
Chris@82 90 TL = VADD(Tb, Tc);
Chris@82 91 }
Chris@82 92 }
Chris@82 93 Ts = VSUB(Tk, Tn);
Chris@82 94 TV = VSUB(TP, TO);
Chris@82 95 TW = VSUB(TM, TL);
Chris@82 96 Tt = VSUB(Td, Tg);
Chris@82 97 Tu = VSUB(T6, T9);
Chris@82 98 TU = VSUB(TI, TJ);
Chris@82 99 Ta = VADD(T6, T9);
Chris@82 100 To = VADD(Tk, Tn);
Chris@82 101 Th = VADD(Td, Tg);
Chris@82 102 Tp = VFNMS(LDK(KP356895867), To, Th);
Chris@82 103 TC = VFNMS(LDK(KP356895867), Th, Ta);
Chris@82 104 Tx = VFNMS(LDK(KP356895867), Ta, To);
Chris@82 105 TK = VADD(TI, TJ);
Chris@82 106 TQ = VADD(TO, TP);
Chris@82 107 TN = VADD(TL, TM);
Chris@82 108 TR = VFNMS(LDK(KP356895867), TK, TQ);
Chris@82 109 T14 = VFNMS(LDK(KP356895867), TQ, TN);
Chris@82 110 TZ = VFNMS(LDK(KP356895867), TN, TK);
Chris@82 111 }
Chris@82 112 {
Chris@82 113 V T19, T1a, T1b, T1e, T1c, T1g, T1h;
Chris@82 114 T19 = VADD(T3, VADD(Ta, VADD(Th, To)));
Chris@82 115 STM2(&(xo[14]), T19, ovs, &(xo[2]));
Chris@82 116 T1a = VADD(TH, VADD(TK, VADD(TN, TQ)));
Chris@82 117 STM2(&(xo[0]), T1a, ovs, &(xo[0]));
Chris@82 118 {
Chris@82 119 V Tr, Tw, Tq, Tv;
Chris@82 120 Tq = VFNMS(LDK(KP692021471), Tp, Ta);
Chris@82 121 Tr = VFNMS(LDK(KP900968867), Tq, T3);
Chris@82 122 Tv = VFNMS(LDK(KP554958132), Tu, Tt);
Chris@82 123 Tw = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tv, Ts));
Chris@82 124 T1b = VFMAI(Tw, Tr);
Chris@82 125 STM2(&(xo[6]), T1b, ovs, &(xo[2]));
Chris@82 126 T1c = VFNMSI(Tw, Tr);
Chris@82 127 STM2(&(xo[22]), T1c, ovs, &(xo[2]));
Chris@82 128 }
Chris@82 129 {
Chris@82 130 V T16, T18, T15, T17, T1d;
Chris@82 131 T15 = VFNMS(LDK(KP692021471), T14, TK);
Chris@82 132 T16 = VFNMS(LDK(KP900968867), T15, TH);
Chris@82 133 T17 = VFMA(LDK(KP554958132), TU, TW);
Chris@82 134 T18 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T17, TV));
Chris@82 135 T1d = VFNMSI(T18, T16);
Chris@82 136 STM2(&(xo[20]), T1d, ovs, &(xo[0]));
Chris@82 137 STN2(&(xo[20]), T1d, T1c, ovs);
Chris@82 138 T1e = VFMAI(T18, T16);
Chris@82 139 STM2(&(xo[8]), T1e, ovs, &(xo[0]));
Chris@82 140 }
Chris@82 141 {
Chris@82 142 V Tz, TB, Ty, TA, T1f;
Chris@82 143 Ty = VFNMS(LDK(KP692021471), Tx, Th);
Chris@82 144 Tz = VFNMS(LDK(KP900968867), Ty, T3);
Chris@82 145 TA = VFMA(LDK(KP554958132), Tt, Ts);
Chris@82 146 TB = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TA, Tu));
Chris@82 147 T1f = VFNMSI(TB, Tz);
Chris@82 148 STM2(&(xo[10]), T1f, ovs, &(xo[2]));
Chris@82 149 STN2(&(xo[8]), T1e, T1f, ovs);
Chris@82 150 T1g = VFMAI(TB, Tz);
Chris@82 151 STM2(&(xo[18]), T1g, ovs, &(xo[2]));
Chris@82 152 }
Chris@82 153 {
Chris@82 154 V TT, TY, TS, TX, T1i;
Chris@82 155 TS = VFNMS(LDK(KP692021471), TR, TN);
Chris@82 156 TT = VFNMS(LDK(KP900968867), TS, TH);
Chris@82 157 TX = VFMA(LDK(KP554958132), TW, TV);
Chris@82 158 TY = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TX, TU));
Chris@82 159 T1h = VFNMSI(TY, TT);
Chris@82 160 STM2(&(xo[24]), T1h, ovs, &(xo[0]));
Chris@82 161 T1i = VFMAI(TY, TT);
Chris@82 162 STM2(&(xo[4]), T1i, ovs, &(xo[0]));
Chris@82 163 STN2(&(xo[4]), T1i, T1b, ovs);
Chris@82 164 }
Chris@82 165 {
Chris@82 166 V T11, T13, T10, T12, T1j, T1k;
Chris@82 167 T10 = VFNMS(LDK(KP692021471), TZ, TQ);
Chris@82 168 T11 = VFNMS(LDK(KP900968867), T10, TH);
Chris@82 169 T12 = VFNMS(LDK(KP554958132), TV, TU);
Chris@82 170 T13 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T12, TW));
Chris@82 171 T1j = VFNMSI(T13, T11);
Chris@82 172 STM2(&(xo[16]), T1j, ovs, &(xo[0]));
Chris@82 173 STN2(&(xo[16]), T1j, T1g, ovs);
Chris@82 174 T1k = VFMAI(T13, T11);
Chris@82 175 STM2(&(xo[12]), T1k, ovs, &(xo[0]));
Chris@82 176 STN2(&(xo[12]), T1k, T19, ovs);
Chris@82 177 }
Chris@82 178 {
Chris@82 179 V TE, TG, TD, TF, T1l, T1m;
Chris@82 180 TD = VFNMS(LDK(KP692021471), TC, To);
Chris@82 181 TE = VFNMS(LDK(KP900968867), TD, T3);
Chris@82 182 TF = VFMA(LDK(KP554958132), Ts, Tu);
Chris@82 183 TG = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TF, Tt));
Chris@82 184 T1l = VFMAI(TG, TE);
Chris@82 185 STM2(&(xo[2]), T1l, ovs, &(xo[2]));
Chris@82 186 STN2(&(xo[0]), T1a, T1l, ovs);
Chris@82 187 T1m = VFNMSI(TG, TE);
Chris@82 188 STM2(&(xo[26]), T1m, ovs, &(xo[2]));
Chris@82 189 STN2(&(xo[24]), T1h, T1m, ovs);
Chris@82 190 }
Chris@82 191 }
Chris@82 192 }
Chris@82 193 }
Chris@82 194 VLEAVE();
Chris@82 195 }
Chris@82 196
Chris@82 197 static const kdft_desc desc = { 14, XSIMD_STRING("n2bv_14"), {32, 6, 42, 0}, &GENUS, 0, 2, 0, 0 };
Chris@82 198
Chris@82 199 void XSIMD(codelet_n2bv_14) (planner *p) {
Chris@82 200 X(kdft_register) (p, n2bv_14, &desc);
Chris@82 201 }
Chris@82 202
Chris@82 203 #else
Chris@82 204
Chris@82 205 /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 14 -name n2bv_14 -with-ostride 2 -include dft/simd/n2b.h -store-multiple 2 */
Chris@82 206
Chris@82 207 /*
Chris@82 208 * This function contains 74 FP additions, 36 FP multiplications,
Chris@82 209 * (or, 50 additions, 12 multiplications, 24 fused multiply/add),
Chris@82 210 * 41 stack variables, 6 constants, and 35 memory accesses
Chris@82 211 */
Chris@82 212 #include "dft/simd/n2b.h"
Chris@82 213
Chris@82 214 static void n2bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@82 215 {
Chris@82 216 DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
Chris@82 217 DVK(KP222520933, +0.222520933956314404288902564496794759466355569);
Chris@82 218 DVK(KP623489801, +0.623489801858733530525004884004239810632274731);
Chris@82 219 DVK(KP781831482, +0.781831482468029808708444526674057750232334519);
Chris@82 220 DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
Chris@82 221 DVK(KP433883739, +0.433883739117558120475768332848358754609990728);
Chris@82 222 {
Chris@82 223 INT i;
Chris@82 224 const R *xi;
Chris@82 225 R *xo;
Chris@82 226 xi = ii;
Chris@82 227 xo = io;
Chris@82 228 for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) {
Chris@82 229 V Tp, Ty, Tl, TL, Tq, TE, T7, TJ, Ts, TB, Te, TK, Tr, TH, Tn;
Chris@82 230 V To;
Chris@82 231 Tn = LD(&(xi[0]), ivs, &(xi[0]));
Chris@82 232 To = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Chris@82 233 Tp = VSUB(Tn, To);
Chris@82 234 Ty = VADD(Tn, To);
Chris@82 235 {
Chris@82 236 V Th, TC, Tk, TD;
Chris@82 237 {
Chris@82 238 V Tf, Tg, Ti, Tj;
Chris@82 239 Tf = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Chris@82 240 Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Chris@82 241 Th = VSUB(Tf, Tg);
Chris@82 242 TC = VADD(Tf, Tg);
Chris@82 243 Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Chris@82 244 Tj = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Chris@82 245 Tk = VSUB(Ti, Tj);
Chris@82 246 TD = VADD(Ti, Tj);
Chris@82 247 }
Chris@82 248 Tl = VSUB(Th, Tk);
Chris@82 249 TL = VSUB(TD, TC);
Chris@82 250 Tq = VADD(Th, Tk);
Chris@82 251 TE = VADD(TC, TD);
Chris@82 252 }
Chris@82 253 {
Chris@82 254 V T3, Tz, T6, TA;
Chris@82 255 {
Chris@82 256 V T1, T2, T4, T5;
Chris@82 257 T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Chris@82 258 T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Chris@82 259 T3 = VSUB(T1, T2);
Chris@82 260 Tz = VADD(T1, T2);
Chris@82 261 T4 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Chris@82 262 T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Chris@82 263 T6 = VSUB(T4, T5);
Chris@82 264 TA = VADD(T4, T5);
Chris@82 265 }
Chris@82 266 T7 = VSUB(T3, T6);
Chris@82 267 TJ = VSUB(Tz, TA);
Chris@82 268 Ts = VADD(T3, T6);
Chris@82 269 TB = VADD(Tz, TA);
Chris@82 270 }
Chris@82 271 {
Chris@82 272 V Ta, TF, Td, TG;
Chris@82 273 {
Chris@82 274 V T8, T9, Tb, Tc;
Chris@82 275 T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Chris@82 276 T9 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Chris@82 277 Ta = VSUB(T8, T9);
Chris@82 278 TF = VADD(T8, T9);
Chris@82 279 Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Chris@82 280 Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Chris@82 281 Td = VSUB(Tb, Tc);
Chris@82 282 TG = VADD(Tb, Tc);
Chris@82 283 }
Chris@82 284 Te = VSUB(Ta, Td);
Chris@82 285 TK = VSUB(TG, TF);
Chris@82 286 Tr = VADD(Ta, Td);
Chris@82 287 TH = VADD(TF, TG);
Chris@82 288 }
Chris@82 289 {
Chris@82 290 V TR, TS, TU, TV;
Chris@82 291 TR = VADD(Tp, VADD(Ts, VADD(Tq, Tr)));
Chris@82 292 STM2(&(xo[14]), TR, ovs, &(xo[2]));
Chris@82 293 TS = VADD(Ty, VADD(TB, VADD(TE, TH)));
Chris@82 294 STM2(&(xo[0]), TS, ovs, &(xo[0]));
Chris@82 295 {
Chris@82 296 V TT, Tm, Tt, TQ, TP, TW;
Chris@82 297 Tm = VBYI(VFMA(LDK(KP433883739), T7, VFNMS(LDK(KP781831482), Tl, VMUL(LDK(KP974927912), Te))));
Chris@82 298 Tt = VFMA(LDK(KP623489801), Tq, VFNMS(LDK(KP222520933), Tr, VFNMS(LDK(KP900968867), Ts, Tp)));
Chris@82 299 TT = VADD(Tm, Tt);
Chris@82 300 STM2(&(xo[6]), TT, ovs, &(xo[2]));
Chris@82 301 TU = VSUB(Tt, Tm);
Chris@82 302 STM2(&(xo[22]), TU, ovs, &(xo[2]));
Chris@82 303 TQ = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK))));
Chris@82 304 TP = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TB, Ty)));
Chris@82 305 TV = VSUB(TP, TQ);
Chris@82 306 STM2(&(xo[24]), TV, ovs, &(xo[0]));
Chris@82 307 TW = VADD(TP, TQ);
Chris@82 308 STM2(&(xo[4]), TW, ovs, &(xo[0]));
Chris@82 309 STN2(&(xo[4]), TW, TT, ovs);
Chris@82 310 }
Chris@82 311 {
Chris@82 312 V T10, TM, TI, TZ;
Chris@82 313 {
Chris@82 314 V Tu, Tv, TX, TY;
Chris@82 315 Tu = VBYI(VFMA(LDK(KP781831482), T7, VFMA(LDK(KP974927912), Tl, VMUL(LDK(KP433883739), Te))));
Chris@82 316 Tv = VFMA(LDK(KP623489801), Ts, VFNMS(LDK(KP900968867), Tr, VFNMS(LDK(KP222520933), Tq, Tp)));
Chris@82 317 TX = VADD(Tu, Tv);
Chris@82 318 STM2(&(xo[2]), TX, ovs, &(xo[2]));
Chris@82 319 STN2(&(xo[0]), TS, TX, ovs);
Chris@82 320 TY = VSUB(Tv, Tu);
Chris@82 321 STM2(&(xo[26]), TY, ovs, &(xo[2]));
Chris@82 322 STN2(&(xo[24]), TV, TY, ovs);
Chris@82 323 }
Chris@82 324 TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ))));
Chris@82 325 TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TE, Ty)));
Chris@82 326 TZ = VSUB(TI, TM);
Chris@82 327 STM2(&(xo[12]), TZ, ovs, &(xo[0]));
Chris@82 328 STN2(&(xo[12]), TZ, TR, ovs);
Chris@82 329 T10 = VADD(TI, TM);
Chris@82 330 STM2(&(xo[16]), T10, ovs, &(xo[0]));
Chris@82 331 {
Chris@82 332 V T11, TO, TN, T12;
Chris@82 333 TO = VBYI(VFMA(LDK(KP433883739), TJ, VFNMS(LDK(KP974927912), TK, VMUL(LDK(KP781831482), TL))));
Chris@82 334 TN = VFMA(LDK(KP623489801), TE, VFNMS(LDK(KP222520933), TH, VFNMS(LDK(KP900968867), TB, Ty)));
Chris@82 335 T11 = VSUB(TN, TO);
Chris@82 336 STM2(&(xo[8]), T11, ovs, &(xo[0]));
Chris@82 337 T12 = VADD(TN, TO);
Chris@82 338 STM2(&(xo[20]), T12, ovs, &(xo[0]));
Chris@82 339 STN2(&(xo[20]), T12, TU, ovs);
Chris@82 340 {
Chris@82 341 V Tx, Tw, T13, T14;
Chris@82 342 Tx = VBYI(VFNMS(LDK(KP781831482), Te, VFNMS(LDK(KP433883739), Tl, VMUL(LDK(KP974927912), T7))));
Chris@82 343 Tw = VFMA(LDK(KP623489801), Tr, VFNMS(LDK(KP900968867), Tq, VFNMS(LDK(KP222520933), Ts, Tp)));
Chris@82 344 T13 = VSUB(Tw, Tx);
Chris@82 345 STM2(&(xo[10]), T13, ovs, &(xo[2]));
Chris@82 346 STN2(&(xo[8]), T11, T13, ovs);
Chris@82 347 T14 = VADD(Tx, Tw);
Chris@82 348 STM2(&(xo[18]), T14, ovs, &(xo[2]));
Chris@82 349 STN2(&(xo[16]), T10, T14, ovs);
Chris@82 350 }
Chris@82 351 }
Chris@82 352 }
Chris@82 353 }
Chris@82 354 }
Chris@82 355 }
Chris@82 356 VLEAVE();
Chris@82 357 }
Chris@82 358
Chris@82 359 static const kdft_desc desc = { 14, XSIMD_STRING("n2bv_14"), {50, 12, 24, 0}, &GENUS, 0, 2, 0, 0 };
Chris@82 360
Chris@82 361 void XSIMD(codelet_n2bv_14) (planner *p) {
Chris@82 362 X(kdft_register) (p, n2bv_14, &desc);
Chris@82 363 }
Chris@82 364
Chris@82 365 #endif