To check out this repository please hg clone the following URL, or open the URL using EasyMercurial or your preferred Mercurial client.
The primary repository for this project is hosted at https://github.com/sonic-visualiser/sv-dependency-builds .
This repository is a read-only copy which is updated automatically every hour.
root / src / fftw-3.3.8 / dft / scalar / codelets / n1_11.c @ 167:bd3cc4d1df30
History | View | Annotate | Download (14.8 KB)
| 1 |
/*
|
|---|---|
| 2 |
* Copyright (c) 2003, 2007-14 Matteo Frigo
|
| 3 |
* Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
|
| 4 |
*
|
| 5 |
* This program is free software; you can redistribute it and/or modify
|
| 6 |
* it under the terms of the GNU General Public License as published by
|
| 7 |
* the Free Software Foundation; either version 2 of the License, or
|
| 8 |
* (at your option) any later version.
|
| 9 |
*
|
| 10 |
* This program is distributed in the hope that it will be useful,
|
| 11 |
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
| 12 |
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
| 13 |
* GNU General Public License for more details.
|
| 14 |
*
|
| 15 |
* You should have received a copy of the GNU General Public License
|
| 16 |
* along with this program; if not, write to the Free Software
|
| 17 |
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
| 18 |
*
|
| 19 |
*/
|
| 20 |
|
| 21 |
/* This file was automatically generated --- DO NOT EDIT */
|
| 22 |
/* Generated on Thu May 24 08:04:10 EDT 2018 */
|
| 23 |
|
| 24 |
#include "dft/codelet-dft.h" |
| 25 |
|
| 26 |
#if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
|
| 27 |
|
| 28 |
/* Generated by: ../../../genfft/gen_notw.native -fma -compact -variables 4 -pipeline-latency 4 -n 11 -name n1_11 -include dft/scalar/n.h */
|
| 29 |
|
| 30 |
/*
|
| 31 |
* This function contains 140 FP additions, 110 FP multiplications,
|
| 32 |
* (or, 30 additions, 0 multiplications, 110 fused multiply/add),
|
| 33 |
* 62 stack variables, 10 constants, and 44 memory accesses
|
| 34 |
*/
|
| 35 |
#include "dft/scalar/n.h" |
| 36 |
|
| 37 |
static void n1_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) |
| 38 |
{
|
| 39 |
DK(KP989821441, +0.989821441880932732376092037776718787376519372); |
| 40 |
DK(KP959492973, +0.959492973614497389890368057066327699062454848); |
| 41 |
DK(KP918985947, +0.918985947228994779780736114132655398124909697); |
| 42 |
DK(KP830830026, +0.830830026003772851058548298459246407048009821); |
| 43 |
DK(KP876768831, +0.876768831002589333891339807079336796764054852); |
| 44 |
DK(KP778434453, +0.778434453334651800608337670740821884709317477); |
| 45 |
DK(KP715370323, +0.715370323453429719112414662767260662417897278); |
| 46 |
DK(KP521108558, +0.521108558113202722944698153526659300680427422); |
| 47 |
DK(KP634356270, +0.634356270682424498893150776899916060542806975); |
| 48 |
DK(KP342584725, +0.342584725681637509502641509861112333758894680); |
| 49 |
{
|
| 50 |
INT i; |
| 51 |
for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(44, is), MAKE_VOLATILE_STRIDE(44, os)) { |
| 52 |
E T1, T1f, T4, T1u, Tg, T1q, T7, T1t, Ta, T1s, Td, T1r, Ti, TP, T26; |
| 53 |
E TG, T1X, T1O, T1w, TY, T1F, T17, To, T1i, TA, T1k, Tr, T1h, Tu, T1j; |
| 54 |
E Tx, T1g, TC, TU, T21, TL, T1S, T1J, T1m, T13, T1A, T1c; |
| 55 |
T1 = ri[0];
|
| 56 |
T1f = ii[0];
|
| 57 |
{
|
| 58 |
E T5, T6, Tp, Tq; |
| 59 |
{
|
| 60 |
E T2, T3, Te, Tf; |
| 61 |
T2 = ri[WS(is, 1)];
|
| 62 |
T3 = ri[WS(is, 10)];
|
| 63 |
T4 = T2 + T3; |
| 64 |
T1u = T3 - T2; |
| 65 |
Te = ri[WS(is, 5)];
|
| 66 |
Tf = ri[WS(is, 6)];
|
| 67 |
Tg = Te + Tf; |
| 68 |
T1q = Tf - Te; |
| 69 |
} |
| 70 |
T5 = ri[WS(is, 2)];
|
| 71 |
T6 = ri[WS(is, 9)];
|
| 72 |
T7 = T5 + T6; |
| 73 |
T1t = T6 - T5; |
| 74 |
{
|
| 75 |
E T8, T9, Tb, Tc; |
| 76 |
T8 = ri[WS(is, 3)];
|
| 77 |
T9 = ri[WS(is, 8)];
|
| 78 |
Ta = T8 + T9; |
| 79 |
T1s = T9 - T8; |
| 80 |
Tb = ri[WS(is, 4)];
|
| 81 |
Tc = ri[WS(is, 7)];
|
| 82 |
Td = Tb + Tc; |
| 83 |
T1r = Tc - Tb; |
| 84 |
} |
| 85 |
{
|
| 86 |
E Th, TO, T25, TF, T1W; |
| 87 |
Th = FNMS(KP342584725, Ta, T7); |
| 88 |
Ti = FNMS(KP634356270, Th, Td); |
| 89 |
TO = FNMS(KP342584725, T4, Ta); |
| 90 |
TP = FNMS(KP634356270, TO, Tg); |
| 91 |
T25 = FMA(KP521108558, T1q, T1u); |
| 92 |
T26 = FMA(KP715370323, T25, T1r); |
| 93 |
TF = FNMS(KP342584725, Td, T4); |
| 94 |
TG = FNMS(KP634356270, TF, T7); |
| 95 |
T1W = FMA(KP521108558, T1s, T1q); |
| 96 |
T1X = FNMS(KP715370323, T1W, T1t); |
| 97 |
} |
| 98 |
{
|
| 99 |
E T1N, T1v, TX, T1E, T16; |
| 100 |
T1N = FNMS(KP521108558, T1t, T1r); |
| 101 |
T1O = FMA(KP715370323, T1N, T1q); |
| 102 |
T1v = FNMS(KP521108558, T1u, T1t); |
| 103 |
T1w = FNMS(KP715370323, T1v, T1s); |
| 104 |
TX = FNMS(KP342584725, T7, Tg); |
| 105 |
TY = FNMS(KP634356270, TX, T4); |
| 106 |
T1E = FMA(KP521108558, T1r, T1s); |
| 107 |
T1F = FMA(KP715370323, T1E, T1u); |
| 108 |
T16 = FNMS(KP342584725, Tg, Td); |
| 109 |
T17 = FNMS(KP634356270, T16, Ta); |
| 110 |
} |
| 111 |
{
|
| 112 |
E Tm, Tn, Ty, Tz; |
| 113 |
Tm = ii[WS(is, 3)];
|
| 114 |
Tn = ii[WS(is, 8)];
|
| 115 |
To = Tm - Tn; |
| 116 |
T1i = Tm + Tn; |
| 117 |
Ty = ii[WS(is, 5)];
|
| 118 |
Tz = ii[WS(is, 6)];
|
| 119 |
TA = Ty - Tz; |
| 120 |
T1k = Ty + Tz; |
| 121 |
} |
| 122 |
Tp = ii[WS(is, 2)];
|
| 123 |
Tq = ii[WS(is, 9)];
|
| 124 |
Tr = Tp - Tq; |
| 125 |
T1h = Tp + Tq; |
| 126 |
{
|
| 127 |
E Ts, Tt, Tv, Tw; |
| 128 |
Ts = ii[WS(is, 4)];
|
| 129 |
Tt = ii[WS(is, 7)];
|
| 130 |
Tu = Ts - Tt; |
| 131 |
T1j = Ts + Tt; |
| 132 |
Tv = ii[WS(is, 1)];
|
| 133 |
Tw = ii[WS(is, 10)];
|
| 134 |
Tx = Tv - Tw; |
| 135 |
T1g = Tv + Tw; |
| 136 |
} |
| 137 |
{
|
| 138 |
E TB, TT, T20, TK, T1R; |
| 139 |
TB = FMA(KP521108558, TA, Tx); |
| 140 |
TC = FMA(KP715370323, TB, Tu); |
| 141 |
TT = FNMS(KP521108558, Tr, Tu); |
| 142 |
TU = FMA(KP715370323, TT, TA); |
| 143 |
T20 = FNMS(KP342584725, T1i, T1h); |
| 144 |
T21 = FNMS(KP634356270, T20, T1j); |
| 145 |
TK = FMA(KP521108558, To, TA); |
| 146 |
TL = FNMS(KP715370323, TK, Tr); |
| 147 |
T1R = FNMS(KP342584725, T1j, T1g); |
| 148 |
T1S = FNMS(KP634356270, T1R, T1h); |
| 149 |
} |
| 150 |
{
|
| 151 |
E T1I, T1l, T12, T1z, T1b; |
| 152 |
T1I = FNMS(KP342584725, T1g, T1i); |
| 153 |
T1J = FNMS(KP634356270, T1I, T1k); |
| 154 |
T1l = FNMS(KP342584725, T1k, T1j); |
| 155 |
T1m = FNMS(KP634356270, T1l, T1i); |
| 156 |
T12 = FMA(KP521108558, Tu, To); |
| 157 |
T13 = FMA(KP715370323, T12, Tx); |
| 158 |
T1z = FNMS(KP342584725, T1h, T1k); |
| 159 |
T1A = FNMS(KP634356270, T1z, T1g); |
| 160 |
T1b = FNMS(KP521108558, Tx, Tr); |
| 161 |
T1c = FNMS(KP715370323, T1b, To); |
| 162 |
} |
| 163 |
} |
| 164 |
ro[0] = T1 + T4 + T7 + Ta + Td + Tg;
|
| 165 |
io[0] = T1f + T1g + T1h + T1i + T1j + T1k;
|
| 166 |
{
|
| 167 |
E Tk, TE, Tj, TD, Tl; |
| 168 |
Tj = FNMS(KP778434453, Ti, T4); |
| 169 |
Tk = FNMS(KP876768831, Tj, Tg); |
| 170 |
TD = FMA(KP830830026, TC, Tr); |
| 171 |
TE = FMA(KP918985947, TD, To); |
| 172 |
Tl = FNMS(KP959492973, Tk, T1); |
| 173 |
ro[WS(os, 10)] = FNMS(KP989821441, TE, Tl);
|
| 174 |
ro[WS(os, 1)] = FMA(KP989821441, TE, Tl);
|
| 175 |
} |
| 176 |
{
|
| 177 |
E T23, T28, T22, T27, T24; |
| 178 |
T22 = FNMS(KP778434453, T21, T1g); |
| 179 |
T23 = FNMS(KP876768831, T22, T1k); |
| 180 |
T27 = FMA(KP830830026, T26, T1t); |
| 181 |
T28 = FMA(KP918985947, T27, T1s); |
| 182 |
T24 = FNMS(KP959492973, T23, T1f); |
| 183 |
io[WS(os, 1)] = FMA(KP989821441, T28, T24);
|
| 184 |
io[WS(os, 10)] = FNMS(KP989821441, T28, T24);
|
| 185 |
} |
| 186 |
{
|
| 187 |
E T1U, T1Z, T1T, T1Y, T1V; |
| 188 |
T1T = FNMS(KP778434453, T1S, T1k); |
| 189 |
T1U = FNMS(KP876768831, T1T, T1i); |
| 190 |
T1Y = FMA(KP830830026, T1X, T1u); |
| 191 |
T1Z = FNMS(KP918985947, T1Y, T1r); |
| 192 |
T1V = FNMS(KP959492973, T1U, T1f); |
| 193 |
io[WS(os, 2)] = FNMS(KP989821441, T1Z, T1V);
|
| 194 |
io[WS(os, 9)] = FMA(KP989821441, T1Z, T1V);
|
| 195 |
} |
| 196 |
{
|
| 197 |
E TI, TN, TH, TM, TJ; |
| 198 |
TH = FNMS(KP778434453, TG, Tg); |
| 199 |
TI = FNMS(KP876768831, TH, Ta); |
| 200 |
TM = FMA(KP830830026, TL, Tx); |
| 201 |
TN = FNMS(KP918985947, TM, Tu); |
| 202 |
TJ = FNMS(KP959492973, TI, T1); |
| 203 |
ro[WS(os, 2)] = FNMS(KP989821441, TN, TJ);
|
| 204 |
ro[WS(os, 9)] = FMA(KP989821441, TN, TJ);
|
| 205 |
} |
| 206 |
{
|
| 207 |
E TR, TW, TQ, TV, TS; |
| 208 |
TQ = FNMS(KP778434453, TP, Td); |
| 209 |
TR = FNMS(KP876768831, TQ, T7); |
| 210 |
TV = FNMS(KP830830026, TU, To); |
| 211 |
TW = FNMS(KP918985947, TV, Tx); |
| 212 |
TS = FNMS(KP959492973, TR, T1); |
| 213 |
ro[WS(os, 8)] = FNMS(KP989821441, TW, TS);
|
| 214 |
ro[WS(os, 3)] = FMA(KP989821441, TW, TS);
|
| 215 |
} |
| 216 |
{
|
| 217 |
E T1L, T1Q, T1K, T1P, T1M; |
| 218 |
T1K = FNMS(KP778434453, T1J, T1j); |
| 219 |
T1L = FNMS(KP876768831, T1K, T1h); |
| 220 |
T1P = FNMS(KP830830026, T1O, T1s); |
| 221 |
T1Q = FNMS(KP918985947, T1P, T1u); |
| 222 |
T1M = FNMS(KP959492973, T1L, T1f); |
| 223 |
io[WS(os, 3)] = FMA(KP989821441, T1Q, T1M);
|
| 224 |
io[WS(os, 8)] = FNMS(KP989821441, T1Q, T1M);
|
| 225 |
} |
| 226 |
{
|
| 227 |
E T10, T15, TZ, T14, T11; |
| 228 |
TZ = FNMS(KP778434453, TY, Ta); |
| 229 |
T10 = FNMS(KP876768831, TZ, Td); |
| 230 |
T14 = FNMS(KP830830026, T13, TA); |
| 231 |
T15 = FMA(KP918985947, T14, Tr); |
| 232 |
T11 = FNMS(KP959492973, T10, T1); |
| 233 |
ro[WS(os, 4)] = FNMS(KP989821441, T15, T11);
|
| 234 |
ro[WS(os, 7)] = FMA(KP989821441, T15, T11);
|
| 235 |
} |
| 236 |
{
|
| 237 |
E T1C, T1H, T1B, T1G, T1D; |
| 238 |
T1B = FNMS(KP778434453, T1A, T1i); |
| 239 |
T1C = FNMS(KP876768831, T1B, T1j); |
| 240 |
T1G = FNMS(KP830830026, T1F, T1q); |
| 241 |
T1H = FMA(KP918985947, T1G, T1t); |
| 242 |
T1D = FNMS(KP959492973, T1C, T1f); |
| 243 |
io[WS(os, 4)] = FNMS(KP989821441, T1H, T1D);
|
| 244 |
io[WS(os, 7)] = FMA(KP989821441, T1H, T1D);
|
| 245 |
} |
| 246 |
{
|
| 247 |
E T1o, T1y, T1n, T1x, T1p; |
| 248 |
T1n = FNMS(KP778434453, T1m, T1h); |
| 249 |
T1o = FNMS(KP876768831, T1n, T1g); |
| 250 |
T1x = FNMS(KP830830026, T1w, T1r); |
| 251 |
T1y = FNMS(KP918985947, T1x, T1q); |
| 252 |
T1p = FNMS(KP959492973, T1o, T1f); |
| 253 |
io[WS(os, 5)] = FMA(KP989821441, T1y, T1p);
|
| 254 |
io[WS(os, 6)] = FNMS(KP989821441, T1y, T1p);
|
| 255 |
} |
| 256 |
{
|
| 257 |
E T19, T1e, T18, T1d, T1a; |
| 258 |
T18 = FNMS(KP778434453, T17, T7); |
| 259 |
T19 = FNMS(KP876768831, T18, T4); |
| 260 |
T1d = FNMS(KP830830026, T1c, Tu); |
| 261 |
T1e = FNMS(KP918985947, T1d, TA); |
| 262 |
T1a = FNMS(KP959492973, T19, T1); |
| 263 |
ro[WS(os, 6)] = FNMS(KP989821441, T1e, T1a);
|
| 264 |
ro[WS(os, 5)] = FMA(KP989821441, T1e, T1a);
|
| 265 |
} |
| 266 |
} |
| 267 |
} |
| 268 |
} |
| 269 |
|
| 270 |
static const kdft_desc desc = { 11, "n1_11", {30, 0, 110, 0}, &GENUS, 0, 0, 0, 0 }; |
| 271 |
|
| 272 |
void X(codelet_n1_11) (planner *p) {
|
| 273 |
X(kdft_register) (p, n1_11, &desc); |
| 274 |
} |
| 275 |
|
| 276 |
#else
|
| 277 |
|
| 278 |
/* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 11 -name n1_11 -include dft/scalar/n.h */
|
| 279 |
|
| 280 |
/*
|
| 281 |
* This function contains 140 FP additions, 100 FP multiplications,
|
| 282 |
* (or, 60 additions, 20 multiplications, 80 fused multiply/add),
|
| 283 |
* 41 stack variables, 10 constants, and 44 memory accesses
|
| 284 |
*/
|
| 285 |
#include "dft/scalar/n.h" |
| 286 |
|
| 287 |
static void n1_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) |
| 288 |
{
|
| 289 |
DK(KP654860733, +0.654860733945285064056925072466293553183791199); |
| 290 |
DK(KP142314838, +0.142314838273285140443792668616369668791051361); |
| 291 |
DK(KP959492973, +0.959492973614497389890368057066327699062454848); |
| 292 |
DK(KP415415013, +0.415415013001886425529274149229623203524004910); |
| 293 |
DK(KP841253532, +0.841253532831181168861811648919367717513292498); |
| 294 |
DK(KP989821441, +0.989821441880932732376092037776718787376519372); |
| 295 |
DK(KP909631995, +0.909631995354518371411715383079028460060241051); |
| 296 |
DK(KP281732556, +0.281732556841429697711417915346616899035777899); |
| 297 |
DK(KP540640817, +0.540640817455597582107635954318691695431770608); |
| 298 |
DK(KP755749574, +0.755749574354258283774035843972344420179717445); |
| 299 |
{
|
| 300 |
INT i; |
| 301 |
for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(44, is), MAKE_VOLATILE_STRIDE(44, os)) { |
| 302 |
E T1, TM, T4, TG, Tk, TR, Tw, TN, T7, TK, Ta, TH, Tn, TQ, Td; |
| 303 |
E TJ, Tq, TO, Tt, TP, Tg, TI; |
| 304 |
{
|
| 305 |
E T2, T3, Ti, Tj; |
| 306 |
T1 = ri[0];
|
| 307 |
TM = ii[0];
|
| 308 |
T2 = ri[WS(is, 1)];
|
| 309 |
T3 = ri[WS(is, 10)];
|
| 310 |
T4 = T2 + T3; |
| 311 |
TG = T3 - T2; |
| 312 |
Ti = ii[WS(is, 1)];
|
| 313 |
Tj = ii[WS(is, 10)];
|
| 314 |
Tk = Ti - Tj; |
| 315 |
TR = Ti + Tj; |
| 316 |
{
|
| 317 |
E Tu, Tv, T5, T6; |
| 318 |
Tu = ii[WS(is, 2)];
|
| 319 |
Tv = ii[WS(is, 9)];
|
| 320 |
Tw = Tu - Tv; |
| 321 |
TN = Tu + Tv; |
| 322 |
T5 = ri[WS(is, 2)];
|
| 323 |
T6 = ri[WS(is, 9)];
|
| 324 |
T7 = T5 + T6; |
| 325 |
TK = T6 - T5; |
| 326 |
} |
| 327 |
} |
| 328 |
{
|
| 329 |
E T8, T9, To, Tp; |
| 330 |
T8 = ri[WS(is, 3)];
|
| 331 |
T9 = ri[WS(is, 8)];
|
| 332 |
Ta = T8 + T9; |
| 333 |
TH = T9 - T8; |
| 334 |
{
|
| 335 |
E Tl, Tm, Tb, Tc; |
| 336 |
Tl = ii[WS(is, 3)];
|
| 337 |
Tm = ii[WS(is, 8)];
|
| 338 |
Tn = Tl - Tm; |
| 339 |
TQ = Tl + Tm; |
| 340 |
Tb = ri[WS(is, 4)];
|
| 341 |
Tc = ri[WS(is, 7)];
|
| 342 |
Td = Tb + Tc; |
| 343 |
TJ = Tc - Tb; |
| 344 |
} |
| 345 |
To = ii[WS(is, 4)];
|
| 346 |
Tp = ii[WS(is, 7)];
|
| 347 |
Tq = To - Tp; |
| 348 |
TO = To + Tp; |
| 349 |
{
|
| 350 |
E Tr, Ts, Te, Tf; |
| 351 |
Tr = ii[WS(is, 5)];
|
| 352 |
Ts = ii[WS(is, 6)];
|
| 353 |
Tt = Tr - Ts; |
| 354 |
TP = Tr + Ts; |
| 355 |
Te = ri[WS(is, 5)];
|
| 356 |
Tf = ri[WS(is, 6)];
|
| 357 |
Tg = Te + Tf; |
| 358 |
TI = Tf - Te; |
| 359 |
} |
| 360 |
} |
| 361 |
{
|
| 362 |
E Tx, Th, TZ, T10; |
| 363 |
ro[0] = T1 + T4 + T7 + Ta + Td + Tg;
|
| 364 |
io[0] = TM + TR + TN + TQ + TO + TP;
|
| 365 |
Tx = FMA(KP755749574, Tk, KP540640817 * Tn) + FNMS(KP909631995, Tt, KP281732556 * Tq) - (KP989821441 * Tw); |
| 366 |
Th = FMA(KP841253532, Ta, T1) + FNMS(KP959492973, Td, KP415415013 * Tg) + FNMA(KP142314838, T7, KP654860733 * T4); |
| 367 |
ro[WS(os, 7)] = Th - Tx;
|
| 368 |
ro[WS(os, 4)] = Th + Tx;
|
| 369 |
TZ = FMA(KP755749574, TG, KP540640817 * TH) + FNMS(KP909631995, TI, KP281732556 * TJ) - (KP989821441 * TK); |
| 370 |
T10 = FMA(KP841253532, TQ, TM) + FNMS(KP959492973, TO, KP415415013 * TP) + FNMA(KP142314838, TN, KP654860733 * TR); |
| 371 |
io[WS(os, 4)] = TZ + T10;
|
| 372 |
io[WS(os, 7)] = T10 - TZ;
|
| 373 |
{
|
| 374 |
E TX, TY, Tz, Ty; |
| 375 |
TX = FMA(KP909631995, TG, KP755749574 * TK) + FNMA(KP540640817, TI, KP989821441 * TJ) - (KP281732556 * TH); |
| 376 |
TY = FMA(KP415415013, TR, TM) + FNMS(KP142314838, TO, KP841253532 * TP) + FNMA(KP959492973, TQ, KP654860733 * TN); |
| 377 |
io[WS(os, 2)] = TX + TY;
|
| 378 |
io[WS(os, 9)] = TY - TX;
|
| 379 |
Tz = FMA(KP909631995, Tk, KP755749574 * Tw) + FNMA(KP540640817, Tt, KP989821441 * Tq) - (KP281732556 * Tn); |
| 380 |
Ty = FMA(KP415415013, T4, T1) + FNMS(KP142314838, Td, KP841253532 * Tg) + FNMA(KP959492973, Ta, KP654860733 * T7); |
| 381 |
ro[WS(os, 9)] = Ty - Tz;
|
| 382 |
ro[WS(os, 2)] = Ty + Tz;
|
| 383 |
} |
| 384 |
} |
| 385 |
{
|
| 386 |
E TB, TA, TT, TU; |
| 387 |
TB = FMA(KP540640817, Tk, KP909631995 * Tw) + FMA(KP989821441, Tn, KP755749574 * Tq) + (KP281732556 * Tt); |
| 388 |
TA = FMA(KP841253532, T4, T1) + FNMS(KP959492973, Tg, KP415415013 * T7) + FNMA(KP654860733, Td, KP142314838 * Ta); |
| 389 |
ro[WS(os, 10)] = TA - TB;
|
| 390 |
ro[WS(os, 1)] = TA + TB;
|
| 391 |
{
|
| 392 |
E TV, TW, TD, TC; |
| 393 |
TV = FMA(KP540640817, TG, KP909631995 * TK) + FMA(KP989821441, TH, KP755749574 * TJ) + (KP281732556 * TI); |
| 394 |
TW = FMA(KP841253532, TR, TM) + FNMS(KP959492973, TP, KP415415013 * TN) + FNMA(KP654860733, TO, KP142314838 * TQ); |
| 395 |
io[WS(os, 1)] = TV + TW;
|
| 396 |
io[WS(os, 10)] = TW - TV;
|
| 397 |
TD = FMA(KP989821441, Tk, KP540640817 * Tq) + FNMS(KP909631995, Tn, KP755749574 * Tt) - (KP281732556 * Tw); |
| 398 |
TC = FMA(KP415415013, Ta, T1) + FNMS(KP654860733, Tg, KP841253532 * Td) + FNMA(KP959492973, T7, KP142314838 * T4); |
| 399 |
ro[WS(os, 8)] = TC - TD;
|
| 400 |
ro[WS(os, 3)] = TC + TD;
|
| 401 |
} |
| 402 |
TT = FMA(KP989821441, TG, KP540640817 * TJ) + FNMS(KP909631995, TH, KP755749574 * TI) - (KP281732556 * TK); |
| 403 |
TU = FMA(KP415415013, TQ, TM) + FNMS(KP654860733, TP, KP841253532 * TO) + FNMA(KP959492973, TN, KP142314838 * TR); |
| 404 |
io[WS(os, 3)] = TT + TU;
|
| 405 |
io[WS(os, 8)] = TU - TT;
|
| 406 |
{
|
| 407 |
E TL, TS, TF, TE; |
| 408 |
TL = FMA(KP281732556, TG, KP755749574 * TH) + FNMS(KP909631995, TJ, KP989821441 * TI) - (KP540640817 * TK); |
| 409 |
TS = FMA(KP841253532, TN, TM) + FNMS(KP142314838, TP, KP415415013 * TO) + FNMA(KP654860733, TQ, KP959492973 * TR); |
| 410 |
io[WS(os, 5)] = TL + TS;
|
| 411 |
io[WS(os, 6)] = TS - TL;
|
| 412 |
TF = FMA(KP281732556, Tk, KP755749574 * Tn) + FNMS(KP909631995, Tq, KP989821441 * Tt) - (KP540640817 * Tw); |
| 413 |
TE = FMA(KP841253532, T7, T1) + FNMS(KP142314838, Tg, KP415415013 * Td) + FNMA(KP654860733, Ta, KP959492973 * T4); |
| 414 |
ro[WS(os, 6)] = TE - TF;
|
| 415 |
ro[WS(os, 5)] = TE + TF;
|
| 416 |
} |
| 417 |
} |
| 418 |
} |
| 419 |
} |
| 420 |
} |
| 421 |
|
| 422 |
static const kdft_desc desc = { 11, "n1_11", {60, 20, 80, 0}, &GENUS, 0, 0, 0, 0 }; |
| 423 |
|
| 424 |
void X(codelet_n1_11) (planner *p) {
|
| 425 |
X(kdft_register) (p, n1_11, &desc); |
| 426 |
} |
| 427 |
|
| 428 |
#endif
|