Mercurial > hg > sv-dependency-builds

annotate src/fftw-3.3.8/rdft/scalar/r2cf/hc2cfdft2_8.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 08:07:16 EDT 2018 */
Chris@82 23
Chris@82 24 #include "rdft/codelet-rdft.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_hc2cdft.native -fma -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hc2cfdft2_8 -include rdft/scalar/hc2cf.h */
Chris@82 29
Chris@82 30 /*
Chris@82 31 * This function contains 90 FP additions, 66 FP multiplications,
Chris@82 32 * (or, 60 additions, 36 multiplications, 30 fused multiply/add),
Chris@82 33 * 45 stack variables, 2 constants, and 32 memory accesses
Chris@82 34 */
Chris@82 35 #include "rdft/scalar/hc2cf.h"
Chris@82 36
Chris@82 37 static void hc2cfdft2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
Chris@82 38 {
Chris@82 39 DK(KP707106781, +0.707106781186547524400844362104849039284835938);
Chris@82 40 DK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@82 41 {
Chris@82 42 INT m;
Chris@82 43 for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) {
Chris@82 44 E T1, T2, Th, Tj, T4, T5, T6, Tk, TB, Tq, Tw, Tc, TM, TQ;
Chris@82 45 {
Chris@82 46 E T3, Ti, Tp, Tb, TL, TP;
Chris@82 47 T1 = W[0];
Chris@82 48 T2 = W[2];
Chris@82 49 T3 = T1 * T2;
Chris@82 50 Th = W[4];
Chris@82 51 Ti = T1 * Th;
Chris@82 52 Tj = W[5];
Chris@82 53 Tp = T1 * Tj;
Chris@82 54 T4 = W[1];
Chris@82 55 T5 = W[3];
Chris@82 56 Tb = T1 * T5;
Chris@82 57 T6 = FMA(T4, T5, T3);
Chris@82 58 Tk = FMA(T4, Tj, Ti);
Chris@82 59 TB = FMA(T4, T2, Tb);
Chris@82 60 Tq = FNMS(T4, Th, Tp);
Chris@82 61 Tw = FNMS(T4, T5, T3);
Chris@82 62 TL = T6 * Th;
Chris@82 63 TP = T6 * Tj;
Chris@82 64 Tc = FNMS(T4, T2, Tb);
Chris@82 65 TM = FMA(Tc, Tj, TL);
Chris@82 66 TQ = FNMS(Tc, Th, TP);
Chris@82 67 }
Chris@82 68 {
Chris@82 69 E TI, T1a, TY, T1u, TF, T1s, TS, T1c, Tg, T1n, T13, T1f, Tu, T1p, T17;
Chris@82 70 E T1h;
Chris@82 71 {
Chris@82 72 E TG, TH, TX, TT, TU, TV, TW, T1t;
Chris@82 73 TG = Ip[0];
Chris@82 74 TH = Im[0];
Chris@82 75 TX = TG + TH;
Chris@82 76 TT = Rm[0];
Chris@82 77 TU = Rp[0];
Chris@82 78 TV = TT - TU;
Chris@82 79 TI = TG - TH;
Chris@82 80 T1a = TU + TT;
Chris@82 81 TW = T1 * TV;
Chris@82 82 TY = FNMS(T4, TX, TW);
Chris@82 83 T1t = T4 * TV;
Chris@82 84 T1u = FMA(T1, TX, T1t);
Chris@82 85 }
Chris@82 86 {
Chris@82 87 E Tz, TR, TE, TN;
Chris@82 88 {
Chris@82 89 E Tx, Ty, TC, TD;
Chris@82 90 Tx = Ip[WS(rs, 2)];
Chris@82 91 Ty = Im[WS(rs, 2)];
Chris@82 92 Tz = Tx - Ty;
Chris@82 93 TR = Tx + Ty;
Chris@82 94 TC = Rp[WS(rs, 2)];
Chris@82 95 TD = Rm[WS(rs, 2)];
Chris@82 96 TE = TC + TD;
Chris@82 97 TN = TD - TC;
Chris@82 98 }
Chris@82 99 {
Chris@82 100 E TA, T1r, TO, T1b;
Chris@82 101 TA = Tw * Tz;
Chris@82 102 TF = FNMS(TB, TE, TA);
Chris@82 103 T1r = TQ * TN;
Chris@82 104 T1s = FMA(TM, TR, T1r);
Chris@82 105 TO = TM * TN;
Chris@82 106 TS = FNMS(TQ, TR, TO);
Chris@82 107 T1b = Tw * TE;
Chris@82 108 T1c = FMA(TB, Tz, T1b);
Chris@82 109 }
Chris@82 110 }
Chris@82 111 {
Chris@82 112 E T9, T12, Tf, T10;
Chris@82 113 {
Chris@82 114 E T7, T8, Td, Te;
Chris@82 115 T7 = Ip[WS(rs, 1)];
Chris@82 116 T8 = Im[WS(rs, 1)];
Chris@82 117 T9 = T7 - T8;
Chris@82 118 T12 = T7 + T8;
Chris@82 119 Td = Rp[WS(rs, 1)];
Chris@82 120 Te = Rm[WS(rs, 1)];
Chris@82 121 Tf = Td + Te;
Chris@82 122 T10 = Td - Te;
Chris@82 123 }
Chris@82 124 {
Chris@82 125 E Ta, T1m, T11, T1e;
Chris@82 126 Ta = T6 * T9;
Chris@82 127 Tg = FNMS(Tc, Tf, Ta);
Chris@82 128 T1m = T2 * T12;
Chris@82 129 T1n = FNMS(T5, T10, T1m);
Chris@82 130 T11 = T2 * T10;
Chris@82 131 T13 = FMA(T5, T12, T11);
Chris@82 132 T1e = T6 * Tf;
Chris@82 133 T1f = FMA(Tc, T9, T1e);
Chris@82 134 }
Chris@82 135 }
Chris@82 136 {
Chris@82 137 E Tn, T16, Tt, T14;
Chris@82 138 {
Chris@82 139 E Tl, Tm, Tr, Ts;
Chris@82 140 Tl = Ip[WS(rs, 3)];
Chris@82 141 Tm = Im[WS(rs, 3)];
Chris@82 142 Tn = Tl - Tm;
Chris@82 143 T16 = Tl + Tm;
Chris@82 144 Tr = Rp[WS(rs, 3)];
Chris@82 145 Ts = Rm[WS(rs, 3)];
Chris@82 146 Tt = Tr + Ts;
Chris@82 147 T14 = Tr - Ts;
Chris@82 148 }
Chris@82 149 {
Chris@82 150 E To, T1o, T15, T1g;
Chris@82 151 To = Tk * Tn;
Chris@82 152 Tu = FNMS(Tq, Tt, To);
Chris@82 153 T1o = Th * T16;
Chris@82 154 T1p = FNMS(Tj, T14, T1o);
Chris@82 155 T15 = Th * T14;
Chris@82 156 T17 = FMA(Tj, T16, T15);
Chris@82 157 T1g = Tk * Tt;
Chris@82 158 T1h = FMA(Tq, Tn, T1g);
Chris@82 159 }
Chris@82 160 }
Chris@82 161 {
Chris@82 162 E TK, T1l, T1w, T1y, T19, T1k, T1j, T1x;
Chris@82 163 {
Chris@82 164 E Tv, TJ, T1q, T1v;
Chris@82 165 Tv = Tg + Tu;
Chris@82 166 TJ = TF + TI;
Chris@82 167 TK = Tv + TJ;
Chris@82 168 T1l = TJ - Tv;
Chris@82 169 T1q = T1n + T1p;
Chris@82 170 T1v = T1s + T1u;
Chris@82 171 T1w = T1q - T1v;
Chris@82 172 T1y = T1q + T1v;
Chris@82 173 }
Chris@82 174 {
Chris@82 175 E TZ, T18, T1d, T1i;
Chris@82 176 TZ = TS + TY;
Chris@82 177 T18 = T13 + T17;
Chris@82 178 T19 = TZ - T18;
Chris@82 179 T1k = T18 + TZ;
Chris@82 180 T1d = T1a + T1c;
Chris@82 181 T1i = T1f + T1h;
Chris@82 182 T1j = T1d - T1i;
Chris@82 183 T1x = T1d + T1i;
Chris@82 184 }
Chris@82 185 Ip[0] = KP500000000 * (TK + T19);
Chris@82 186 Rp[0] = KP500000000 * (T1x + T1y);
Chris@82 187 Im[WS(rs, 3)] = KP500000000 * (T19 - TK);
Chris@82 188 Rm[WS(rs, 3)] = KP500000000 * (T1x - T1y);
Chris@82 189 Rm[WS(rs, 1)] = KP500000000 * (T1j - T1k);
Chris@82 190 Im[WS(rs, 1)] = KP500000000 * (T1w - T1l);
Chris@82 191 Rp[WS(rs, 2)] = KP500000000 * (T1j + T1k);
Chris@82 192 Ip[WS(rs, 2)] = KP500000000 * (T1l + T1w);
Chris@82 193 }
Chris@82 194 {
Chris@82 195 E T1B, T1N, T1L, T1R, T1E, T1O, T1H, T1P;
Chris@82 196 {
Chris@82 197 E T1z, T1A, T1J, T1K;
Chris@82 198 T1z = TI - TF;
Chris@82 199 T1A = T1f - T1h;
Chris@82 200 T1B = T1z - T1A;
Chris@82 201 T1N = T1A + T1z;
Chris@82 202 T1J = T1a - T1c;
Chris@82 203 T1K = Tg - Tu;
Chris@82 204 T1L = T1J - T1K;
Chris@82 205 T1R = T1J + T1K;
Chris@82 206 }
Chris@82 207 {
Chris@82 208 E T1C, T1D, T1F, T1G;
Chris@82 209 T1C = T1p - T1n;
Chris@82 210 T1D = T13 - T17;
Chris@82 211 T1E = T1C + T1D;
Chris@82 212 T1O = T1C - T1D;
Chris@82 213 T1F = TY - TS;
Chris@82 214 T1G = T1u - T1s;
Chris@82 215 T1H = T1F - T1G;
Chris@82 216 T1P = T1F + T1G;
Chris@82 217 }
Chris@82 218 {
Chris@82 219 E T1I, T1S, T1M, T1Q;
Chris@82 220 T1I = T1E + T1H;
Chris@82 221 Ip[WS(rs, 1)] = KP500000000 * (FMA(KP707106781, T1I, T1B));
Chris@82 222 Im[WS(rs, 2)] = -(KP500000000 * (FNMS(KP707106781, T1I, T1B)));
Chris@82 223 T1S = T1O + T1P;
Chris@82 224 Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP707106781, T1S, T1R));
Chris@82 225 Rp[WS(rs, 1)] = KP500000000 * (FMA(KP707106781, T1S, T1R));
Chris@82 226 T1M = T1H - T1E;
Chris@82 227 Rm[0] = KP500000000 * (FNMS(KP707106781, T1M, T1L));
Chris@82 228 Rp[WS(rs, 3)] = KP500000000 * (FMA(KP707106781, T1M, T1L));
Chris@82 229 T1Q = T1O - T1P;
Chris@82 230 Ip[WS(rs, 3)] = KP500000000 * (FMA(KP707106781, T1Q, T1N));
Chris@82 231 Im[0] = -(KP500000000 * (FNMS(KP707106781, T1Q, T1N)));
Chris@82 232 }
Chris@82 233 }
Chris@82 234 }
Chris@82 235 }
Chris@82 236 }
Chris@82 237 }
Chris@82 238
Chris@82 239 static const tw_instr twinstr[] = {
Chris@82 240 {TW_CEXP, 1, 1},
Chris@82 241 {TW_CEXP, 1, 3},
Chris@82 242 {TW_CEXP, 1, 7},
Chris@82 243 {TW_NEXT, 1, 0}
Chris@82 244 };
Chris@82 245
Chris@82 246 static const hc2c_desc desc = { 8, "hc2cfdft2_8", twinstr, &GENUS, {60, 36, 30, 0} };
Chris@82 247
Chris@82 248 void X(codelet_hc2cfdft2_8) (planner *p) {
Chris@82 249 X(khc2c_register) (p, hc2cfdft2_8, &desc, HC2C_VIA_DFT);
Chris@82 250 }
Chris@82 251 #else
Chris@82 252
Chris@82 253 /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hc2cfdft2_8 -include rdft/scalar/hc2cf.h */
Chris@82 254
Chris@82 255 /*
Chris@82 256 * This function contains 90 FP additions, 56 FP multiplications,
Chris@82 257 * (or, 72 additions, 38 multiplications, 18 fused multiply/add),
Chris@82 258 * 51 stack variables, 2 constants, and 32 memory accesses
Chris@82 259 */
Chris@82 260 #include "rdft/scalar/hc2cf.h"
Chris@82 261
Chris@82 262 static void hc2cfdft2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
Chris@82 263 {
Chris@82 264 DK(KP353553390, +0.353553390593273762200422181052424519642417969);
Chris@82 265 DK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@82 266 {
Chris@82 267 INT m;
Chris@82 268 for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) {
Chris@82 269 E T1, T4, T2, T5, Tu, Ty, T7, Td, Ti, Tj, Tk, TP, To, TN;
Chris@82 270 {
Chris@82 271 E T3, Tc, T6, Tb;
Chris@82 272 T1 = W[0];
Chris@82 273 T4 = W[1];
Chris@82 274 T2 = W[2];
Chris@82 275 T5 = W[3];
Chris@82 276 T3 = T1 * T2;
Chris@82 277 Tc = T4 * T2;
Chris@82 278 T6 = T4 * T5;
Chris@82 279 Tb = T1 * T5;
Chris@82 280 Tu = T3 - T6;
Chris@82 281 Ty = Tb + Tc;
Chris@82 282 T7 = T3 + T6;
Chris@82 283 Td = Tb - Tc;
Chris@82 284 Ti = W[4];
Chris@82 285 Tj = W[5];
Chris@82 286 Tk = FMA(T1, Ti, T4 * Tj);
Chris@82 287 TP = FNMS(Td, Ti, T7 * Tj);
Chris@82 288 To = FNMS(T4, Ti, T1 * Tj);
Chris@82 289 TN = FMA(T7, Ti, Td * Tj);
Chris@82 290 }
Chris@82 291 {
Chris@82 292 E TF, T11, TC, T12, T1d, T1e, T1q, TM, TR, T1p, Th, Ts, T15, T14, T1a;
Chris@82 293 E T1b, T1m, TV, TY, T1n;
Chris@82 294 {
Chris@82 295 E TD, TE, TL, TI, TJ, TK, Tx, TQ, TB, TO;
Chris@82 296 TD = Ip[0];
Chris@82 297 TE = Im[0];
Chris@82 298 TL = TD + TE;
Chris@82 299 TI = Rm[0];
Chris@82 300 TJ = Rp[0];
Chris@82 301 TK = TI - TJ;
Chris@82 302 {
Chris@82 303 E Tv, Tw, Tz, TA;
Chris@82 304 Tv = Ip[WS(rs, 2)];
Chris@82 305 Tw = Im[WS(rs, 2)];
Chris@82 306 Tx = Tv - Tw;
Chris@82 307 TQ = Tv + Tw;
Chris@82 308 Tz = Rp[WS(rs, 2)];
Chris@82 309 TA = Rm[WS(rs, 2)];
Chris@82 310 TB = Tz + TA;
Chris@82 311 TO = Tz - TA;
Chris@82 312 }
Chris@82 313 TF = TD - TE;
Chris@82 314 T11 = TJ + TI;
Chris@82 315 TC = FNMS(Ty, TB, Tu * Tx);
Chris@82 316 T12 = FMA(Tu, TB, Ty * Tx);
Chris@82 317 T1d = FNMS(TP, TO, TN * TQ);
Chris@82 318 T1e = FMA(T4, TK, T1 * TL);
Chris@82 319 T1q = T1e - T1d;
Chris@82 320 TM = FNMS(T4, TL, T1 * TK);
Chris@82 321 TR = FMA(TN, TO, TP * TQ);
Chris@82 322 T1p = TR + TM;
Chris@82 323 }
Chris@82 324 {
Chris@82 325 E Ta, TU, Tg, TT, Tn, TX, Tr, TW;
Chris@82 326 {
Chris@82 327 E T8, T9, Te, Tf;
Chris@82 328 T8 = Ip[WS(rs, 1)];
Chris@82 329 T9 = Im[WS(rs, 1)];
Chris@82 330 Ta = T8 - T9;
Chris@82 331 TU = T8 + T9;
Chris@82 332 Te = Rp[WS(rs, 1)];
Chris@82 333 Tf = Rm[WS(rs, 1)];
Chris@82 334 Tg = Te + Tf;
Chris@82 335 TT = Te - Tf;
Chris@82 336 }
Chris@82 337 {
Chris@82 338 E Tl, Tm, Tp, Tq;
Chris@82 339 Tl = Ip[WS(rs, 3)];
Chris@82 340 Tm = Im[WS(rs, 3)];
Chris@82 341 Tn = Tl - Tm;
Chris@82 342 TX = Tl + Tm;
Chris@82 343 Tp = Rp[WS(rs, 3)];
Chris@82 344 Tq = Rm[WS(rs, 3)];
Chris@82 345 Tr = Tp + Tq;
Chris@82 346 TW = Tp - Tq;
Chris@82 347 }
Chris@82 348 Th = FNMS(Td, Tg, T7 * Ta);
Chris@82 349 Ts = FNMS(To, Tr, Tk * Tn);
Chris@82 350 T15 = FMA(Tk, Tr, To * Tn);
Chris@82 351 T14 = FMA(T7, Tg, Td * Ta);
Chris@82 352 T1a = FNMS(T5, TT, T2 * TU);
Chris@82 353 T1b = FNMS(Tj, TW, Ti * TX);
Chris@82 354 T1m = T1b - T1a;
Chris@82 355 TV = FMA(T2, TT, T5 * TU);
Chris@82 356 TY = FMA(Ti, TW, Tj * TX);
Chris@82 357 T1n = TV - TY;
Chris@82 358 }
Chris@82 359 {
Chris@82 360 E T1l, T1x, T1A, T1C, T1s, T1w, T1v, T1B;
Chris@82 361 {
Chris@82 362 E T1j, T1k, T1y, T1z;
Chris@82 363 T1j = TF - TC;
Chris@82 364 T1k = T14 - T15;
Chris@82 365 T1l = KP500000000 * (T1j - T1k);
Chris@82 366 T1x = KP500000000 * (T1k + T1j);
Chris@82 367 T1y = T1m - T1n;
Chris@82 368 T1z = T1p + T1q;
Chris@82 369 T1A = KP353553390 * (T1y - T1z);
Chris@82 370 T1C = KP353553390 * (T1y + T1z);
Chris@82 371 }
Chris@82 372 {
Chris@82 373 E T1o, T1r, T1t, T1u;
Chris@82 374 T1o = T1m + T1n;
Chris@82 375 T1r = T1p - T1q;
Chris@82 376 T1s = KP353553390 * (T1o + T1r);
Chris@82 377 T1w = KP353553390 * (T1r - T1o);
Chris@82 378 T1t = T11 - T12;
Chris@82 379 T1u = Th - Ts;
Chris@82 380 T1v = KP500000000 * (T1t - T1u);
Chris@82 381 T1B = KP500000000 * (T1t + T1u);
Chris@82 382 }
Chris@82 383 Ip[WS(rs, 1)] = T1l + T1s;
Chris@82 384 Rp[WS(rs, 1)] = T1B + T1C;
Chris@82 385 Im[WS(rs, 2)] = T1s - T1l;
Chris@82 386 Rm[WS(rs, 2)] = T1B - T1C;
Chris@82 387 Rm[0] = T1v - T1w;
Chris@82 388 Im[0] = T1A - T1x;
Chris@82 389 Rp[WS(rs, 3)] = T1v + T1w;
Chris@82 390 Ip[WS(rs, 3)] = T1x + T1A;
Chris@82 391 }
Chris@82 392 {
Chris@82 393 E TH, T19, T1g, T1i, T10, T18, T17, T1h;
Chris@82 394 {
Chris@82 395 E Tt, TG, T1c, T1f;
Chris@82 396 Tt = Th + Ts;
Chris@82 397 TG = TC + TF;
Chris@82 398 TH = Tt + TG;
Chris@82 399 T19 = TG - Tt;
Chris@82 400 T1c = T1a + T1b;
Chris@82 401 T1f = T1d + T1e;
Chris@82 402 T1g = T1c - T1f;
Chris@82 403 T1i = T1c + T1f;
Chris@82 404 }
Chris@82 405 {
Chris@82 406 E TS, TZ, T13, T16;
Chris@82 407 TS = TM - TR;
Chris@82 408 TZ = TV + TY;
Chris@82 409 T10 = TS - TZ;
Chris@82 410 T18 = TZ + TS;
Chris@82 411 T13 = T11 + T12;
Chris@82 412 T16 = T14 + T15;
Chris@82 413 T17 = T13 - T16;
Chris@82 414 T1h = T13 + T16;
Chris@82 415 }
Chris@82 416 Ip[0] = KP500000000 * (TH + T10);
Chris@82 417 Rp[0] = KP500000000 * (T1h + T1i);
Chris@82 418 Im[WS(rs, 3)] = KP500000000 * (T10 - TH);
Chris@82 419 Rm[WS(rs, 3)] = KP500000000 * (T1h - T1i);
Chris@82 420 Rm[WS(rs, 1)] = KP500000000 * (T17 - T18);
Chris@82 421 Im[WS(rs, 1)] = KP500000000 * (T1g - T19);
Chris@82 422 Rp[WS(rs, 2)] = KP500000000 * (T17 + T18);
Chris@82 423 Ip[WS(rs, 2)] = KP500000000 * (T19 + T1g);
Chris@82 424 }
Chris@82 425 }
Chris@82 426 }
Chris@82 427 }
Chris@82 428 }
Chris@82 429
Chris@82 430 static const tw_instr twinstr[] = {
Chris@82 431 {TW_CEXP, 1, 1},
Chris@82 432 {TW_CEXP, 1, 3},
Chris@82 433 {TW_CEXP, 1, 7},
Chris@82 434 {TW_NEXT, 1, 0}
Chris@82 435 };
Chris@82 436
Chris@82 437 static const hc2c_desc desc = { 8, "hc2cfdft2_8", twinstr, &GENUS, {72, 38, 18, 0} };
Chris@82 438
Chris@82 439 void X(codelet_hc2cfdft2_8) (planner *p) {
Chris@82 440 X(khc2c_register) (p, hc2cfdft2_8, &desc, HC2C_VIA_DFT);
Chris@82 441 }
Chris@82 442 #endif