Chris@10: /*
Chris@10:  * Copyright (c) 2003, 2007-11 Matteo Frigo
Chris@10:  * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology
Chris@10:  *
Chris@10:  * This program is free software; you can redistribute it and/or modify
Chris@10:  * it under the terms of the GNU General Public License as published by
Chris@10:  * the Free Software Foundation; either version 2 of the License, or
Chris@10:  * (at your option) any later version.
Chris@10:  *
Chris@10:  * This program is distributed in the hope that it will be useful,
Chris@10:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
Chris@10:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
Chris@10:  * GNU General Public License for more details.
Chris@10:  *
Chris@10:  * You should have received a copy of the GNU General Public License
Chris@10:  * along with this program; if not, write to the Free Software
Chris@10:  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
Chris@10:  *
Chris@10:  */
Chris@10: 
Chris@10: /* This file was automatically generated --- DO NOT EDIT */
Chris@10: /* Generated on Sun Nov 25 07:35:43 EST 2012 */
Chris@10: 
Chris@10: #include "codelet-dft.h"
Chris@10: 
Chris@10: #ifdef HAVE_FMA
Chris@10: 
Chris@10: /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 11 -name n1_11 -include n.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 140 FP additions, 110 FP multiplications,
Chris@10:  * (or, 30 additions, 0 multiplications, 110 fused multiply/add),
Chris@10:  * 84 stack variables, 10 constants, and 44 memory accesses
Chris@10:  */
Chris@10: #include "n.h"
Chris@10: 
Chris@10: static void n1_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@10: {
Chris@10:      DK(KP989821441, +0.989821441880932732376092037776718787376519372);
Chris@10:      DK(KP959492973, +0.959492973614497389890368057066327699062454848);
Chris@10:      DK(KP918985947, +0.918985947228994779780736114132655398124909697);
Chris@10:      DK(KP876768831, +0.876768831002589333891339807079336796764054852);
Chris@10:      DK(KP830830026, +0.830830026003772851058548298459246407048009821);
Chris@10:      DK(KP778434453, +0.778434453334651800608337670740821884709317477);
Chris@10:      DK(KP715370323, +0.715370323453429719112414662767260662417897278);
Chris@10:      DK(KP634356270, +0.634356270682424498893150776899916060542806975);
Chris@10:      DK(KP342584725, +0.342584725681637509502641509861112333758894680);
Chris@10:      DK(KP521108558, +0.521108558113202722944698153526659300680427422);
Chris@10:      {
Chris@10: 	  INT i;
Chris@10: 	  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)) {
Chris@10: 	       E T1, TA, T1p, T1y, T19, T1d, T1a, T1e;
Chris@10: 	       {
Chris@10: 		    E T1f, T1u, T4, T1q, Tg, T1t, T7, T1s, Ta, Td, T1r, TP, T1X, T26, Ti;
Chris@10: 		    E TG, T1O, T1w, TY, T1F, T17, To, T1i, T1k, T1h, Tr, T1j, Tu, T1g, Tx;
Chris@10: 		    E T21, TU, TL, TC, T1S, T1J, T1m, T12, T1z, T1b;
Chris@10: 		    T1 = ri[0];
Chris@10: 		    T1f = ii[0];
Chris@10: 		    {
Chris@10: 			 E T1E, T16, Tb, Tc, Tv, Tw;
Chris@10: 			 {
Chris@10: 			      E T2, T3, Te, Tf;
Chris@10: 			      T2 = ri[WS(is, 1)];
Chris@10: 			      T3 = ri[WS(is, 10)];
Chris@10: 			      Te = ri[WS(is, 5)];
Chris@10: 			      Tf = ri[WS(is, 6)];
Chris@10: 			      {
Chris@10: 				   E T5, T6, T8, T9;
Chris@10: 				   T5 = ri[WS(is, 2)];
Chris@10: 				   T1u = T3 - T2;
Chris@10: 				   T4 = T2 + T3;
Chris@10: 				   T1q = Tf - Te;
Chris@10: 				   Tg = Te + Tf;
Chris@10: 				   T6 = ri[WS(is, 9)];
Chris@10: 				   T8 = ri[WS(is, 3)];
Chris@10: 				   T9 = ri[WS(is, 8)];
Chris@10: 				   Tb = ri[WS(is, 4)];
Chris@10: 				   T1t = T6 - T5;
Chris@10: 				   T7 = T5 + T6;
Chris@10: 				   T1s = T9 - T8;
Chris@10: 				   Ta = T8 + T9;
Chris@10: 				   Tc = ri[WS(is, 7)];
Chris@10: 			      }
Chris@10: 			 }
Chris@10: 			 {
Chris@10: 			      E T25, Th, T1W, TO;
Chris@10: 			      T25 = FMA(KP521108558, T1q, T1u);
Chris@10: 			      T1W = FMA(KP521108558, T1s, T1q);
Chris@10: 			      TO = FNMS(KP342584725, T4, Ta);
Chris@10: 			      Th = FNMS(KP342584725, Ta, T7);
Chris@10: 			      Td = Tb + Tc;
Chris@10: 			      T1r = Tc - Tb;
Chris@10: 			      TP = FNMS(KP634356270, TO, Tg);
Chris@10: 			      T1X = FNMS(KP715370323, T1W, T1t);
Chris@10: 			      T26 = FMA(KP715370323, T25, T1r);
Chris@10: 			      {
Chris@10: 				   E TF, T1N, T1v, TX;
Chris@10: 				   TF = FNMS(KP342584725, Td, T4);
Chris@10: 				   Ti = FNMS(KP634356270, Th, Td);
Chris@10: 				   T1N = FNMS(KP521108558, T1t, T1r);
Chris@10: 				   T1v = FNMS(KP521108558, T1u, T1t);
Chris@10: 				   TG = FNMS(KP634356270, TF, T7);
Chris@10: 				   TX = FNMS(KP342584725, T7, Tg);
Chris@10: 				   T1O = FMA(KP715370323, T1N, T1q);
Chris@10: 				   T1w = FNMS(KP715370323, T1v, T1s);
Chris@10: 				   T1E = FMA(KP521108558, T1r, T1s);
Chris@10: 				   TY = FNMS(KP634356270, TX, T4);
Chris@10: 				   T16 = FNMS(KP342584725, Tg, Td);
Chris@10: 			      }
Chris@10: 			 }
Chris@10: 			 {
Chris@10: 			      E Ty, Tz, Tm, Tn;
Chris@10: 			      Tm = ii[WS(is, 3)];
Chris@10: 			      T1F = FMA(KP715370323, T1E, T1u);
Chris@10: 			      Tn = ii[WS(is, 8)];
Chris@10: 			      T17 = FNMS(KP634356270, T16, Ta);
Chris@10: 			      Ty = ii[WS(is, 5)];
Chris@10: 			      Tz = ii[WS(is, 6)];
Chris@10: 			      To = Tm - Tn;
Chris@10: 			      T1i = Tm + Tn;
Chris@10: 			      {
Chris@10: 				   E Tp, Tq, Ts, Tt;
Chris@10: 				   Tp = ii[WS(is, 2)];
Chris@10: 				   T1k = Ty + Tz;
Chris@10: 				   TA = Ty - Tz;
Chris@10: 				   Tq = ii[WS(is, 9)];
Chris@10: 				   Ts = ii[WS(is, 4)];
Chris@10: 				   Tt = ii[WS(is, 7)];
Chris@10: 				   Tv = ii[WS(is, 1)];
Chris@10: 				   T1h = Tp + Tq;
Chris@10: 				   Tr = Tp - Tq;
Chris@10: 				   T1j = Ts + Tt;
Chris@10: 				   Tu = Ts - Tt;
Chris@10: 				   Tw = ii[WS(is, 10)];
Chris@10: 			      }
Chris@10: 			 }
Chris@10: 			 {
Chris@10: 			      E TB, T1R, T20, TK, TT, T1I, T1l;
Chris@10: 			      T20 = FNMS(KP342584725, T1i, T1h);
Chris@10: 			      TK = FMA(KP521108558, To, TA);
Chris@10: 			      TT = FNMS(KP521108558, Tr, Tu);
Chris@10: 			      T1g = Tv + Tw;
Chris@10: 			      Tx = Tv - Tw;
Chris@10: 			      T21 = FNMS(KP634356270, T20, T1j);
Chris@10: 			      TU = FMA(KP715370323, TT, TA);
Chris@10: 			      TL = FNMS(KP715370323, TK, Tr);
Chris@10: 			      TB = FMA(KP521108558, TA, Tx);
Chris@10: 			      T1R = FNMS(KP342584725, T1j, T1g);
Chris@10: 			      T1I = FNMS(KP342584725, T1g, T1i);
Chris@10: 			      T1l = FNMS(KP342584725, T1k, T1j);
Chris@10: 			      TC = FMA(KP715370323, TB, Tu);
Chris@10: 			      T1S = FNMS(KP634356270, T1R, T1h);
Chris@10: 			      T1J = FNMS(KP634356270, T1I, T1k);
Chris@10: 			      T1m = FNMS(KP634356270, T1l, T1i);
Chris@10: 			      T12 = FMA(KP521108558, Tu, To);
Chris@10: 			      T1z = FNMS(KP342584725, T1h, T1k);
Chris@10: 			      T1b = FNMS(KP521108558, Tx, Tr);
Chris@10: 			 }
Chris@10: 		    }
Chris@10: 		    {
Chris@10: 			 E T13, T1A, T1c, T1Z, T1V, TH, TM, Tj, TD;
Chris@10: 			 ro[0] = T1 + T4 + T7 + Ta + Td + Tg;
Chris@10: 			 T13 = FMA(KP715370323, T12, Tx);
Chris@10: 			 T1A = FNMS(KP634356270, T1z, T1g);
Chris@10: 			 T1c = FNMS(KP715370323, T1b, To);
Chris@10: 			 io[0] = T1f + T1g + T1h + T1i + T1j + T1k;
Chris@10: 			 Tj = FNMS(KP778434453, Ti, T4);
Chris@10: 			 TD = FMA(KP830830026, TC, Tr);
Chris@10: 			 {
Chris@10: 			      E TE, T23, T28, Tl, Tk, T22, T27;
Chris@10: 			      T22 = FNMS(KP778434453, T21, T1g);
Chris@10: 			      T27 = FMA(KP830830026, T26, T1t);
Chris@10: 			      Tk = FNMS(KP876768831, Tj, Tg);
Chris@10: 			      TE = FMA(KP918985947, TD, To);
Chris@10: 			      T23 = FNMS(KP876768831, T22, T1k);
Chris@10: 			      T28 = FMA(KP918985947, T27, T1s);
Chris@10: 			      Tl = FNMS(KP959492973, Tk, T1);
Chris@10: 			      {
Chris@10: 				   E T1U, T1T, T24, T1Y;
Chris@10: 				   T1T = FNMS(KP778434453, T1S, T1k);
Chris@10: 				   T24 = FNMS(KP959492973, T23, T1f);
Chris@10: 				   T1Y = FMA(KP830830026, T1X, T1u);
Chris@10: 				   ro[WS(os, 1)] = FMA(KP989821441, TE, Tl);
Chris@10: 				   ro[WS(os, 10)] = FNMS(KP989821441, TE, Tl);
Chris@10: 				   T1U = FNMS(KP876768831, T1T, T1i);
Chris@10: 				   io[WS(os, 10)] = FNMS(KP989821441, T28, T24);
Chris@10: 				   io[WS(os, 1)] = FMA(KP989821441, T28, T24);
Chris@10: 				   T1Z = FNMS(KP918985947, T1Y, T1r);
Chris@10: 				   T1V = FNMS(KP959492973, T1U, T1f);
Chris@10: 			      }
Chris@10: 			      TH = FNMS(KP778434453, TG, Tg);
Chris@10: 			      TM = FMA(KP830830026, TL, Tx);
Chris@10: 			 }
Chris@10: 			 {
Chris@10: 			      E T1M, TZ, T14, T1Q;
Chris@10: 			      {
Chris@10: 				   E TN, TR, TV, TJ, TI, TQ, T1P;
Chris@10: 				   TQ = FNMS(KP778434453, TP, Td);
Chris@10: 				   io[WS(os, 9)] = FMA(KP989821441, T1Z, T1V);
Chris@10: 				   io[WS(os, 2)] = FNMS(KP989821441, T1Z, T1V);
Chris@10: 				   TI = FNMS(KP876768831, TH, Ta);
Chris@10: 				   TN = FNMS(KP918985947, TM, Tu);
Chris@10: 				   TR = FNMS(KP876768831, TQ, T7);
Chris@10: 				   TV = FNMS(KP830830026, TU, To);
Chris@10: 				   TJ = FNMS(KP959492973, TI, T1);
Chris@10: 				   {
Chris@10: 					E T1L, TS, TW, T1K;
Chris@10: 					T1K = FNMS(KP778434453, T1J, T1j);
Chris@10: 					TS = FNMS(KP959492973, TR, T1);
Chris@10: 					TW = FNMS(KP918985947, TV, Tx);
Chris@10: 					ro[WS(os, 9)] = FMA(KP989821441, TN, TJ);
Chris@10: 					ro[WS(os, 2)] = FNMS(KP989821441, TN, TJ);
Chris@10: 					T1L = FNMS(KP876768831, T1K, T1h);
Chris@10: 					ro[WS(os, 3)] = FMA(KP989821441, TW, TS);
Chris@10: 					ro[WS(os, 8)] = FNMS(KP989821441, TW, TS);
Chris@10: 					T1P = FNMS(KP830830026, T1O, T1s);
Chris@10: 					T1M = FNMS(KP959492973, T1L, T1f);
Chris@10: 				   }
Chris@10: 				   TZ = FNMS(KP778434453, TY, Ta);
Chris@10: 				   T14 = FNMS(KP830830026, T13, TA);
Chris@10: 				   T1Q = FNMS(KP918985947, T1P, T1u);
Chris@10: 			      }
Chris@10: 			      {
Chris@10: 				   E T15, T11, T1C, T1G, T1B, T10;
Chris@10: 				   T1B = FNMS(KP778434453, T1A, T1i);
Chris@10: 				   T10 = FNMS(KP876768831, TZ, Td);
Chris@10: 				   T15 = FMA(KP918985947, T14, Tr);
Chris@10: 				   io[WS(os, 8)] = FNMS(KP989821441, T1Q, T1M);
Chris@10: 				   io[WS(os, 3)] = FMA(KP989821441, T1Q, T1M);
Chris@10: 				   T11 = FNMS(KP959492973, T10, T1);
Chris@10: 				   T1C = FNMS(KP876768831, T1B, T1j);
Chris@10: 				   T1G = FNMS(KP830830026, T1F, T1q);
Chris@10: 				   {
Chris@10: 					E T1D, T1H, T1o, T1x, T1n, T18;
Chris@10: 					T1n = FNMS(KP778434453, T1m, T1h);
Chris@10: 					ro[WS(os, 7)] = FMA(KP989821441, T15, T11);
Chris@10: 					ro[WS(os, 4)] = FNMS(KP989821441, T15, T11);
Chris@10: 					T1D = FNMS(KP959492973, T1C, T1f);
Chris@10: 					T1H = FMA(KP918985947, T1G, T1t);
Chris@10: 					T1o = FNMS(KP876768831, T1n, T1g);
Chris@10: 					T1x = FNMS(KP830830026, T1w, T1r);
Chris@10: 					T18 = FNMS(KP778434453, T17, T7);
Chris@10: 					io[WS(os, 7)] = FMA(KP989821441, T1H, T1D);
Chris@10: 					io[WS(os, 4)] = FNMS(KP989821441, T1H, T1D);
Chris@10: 					T1p = FNMS(KP959492973, T1o, T1f);
Chris@10: 					T1y = FNMS(KP918985947, T1x, T1q);
Chris@10: 					T19 = FNMS(KP876768831, T18, T4);
Chris@10: 					T1d = FNMS(KP830830026, T1c, Tu);
Chris@10: 				   }
Chris@10: 			      }
Chris@10: 			 }
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	       io[WS(os, 6)] = FNMS(KP989821441, T1y, T1p);
Chris@10: 	       io[WS(os, 5)] = FMA(KP989821441, T1y, T1p);
Chris@10: 	       T1a = FNMS(KP959492973, T19, T1);
Chris@10: 	       T1e = FNMS(KP918985947, T1d, TA);
Chris@10: 	       ro[WS(os, 5)] = FMA(KP989821441, T1e, T1a);
Chris@10: 	       ro[WS(os, 6)] = FNMS(KP989821441, T1e, T1a);
Chris@10: 	  }
Chris@10:      }
Chris@10: }
Chris@10: 
Chris@10: static const kdft_desc desc = { 11, "n1_11", {30, 0, 110, 0}, &GENUS, 0, 0, 0, 0 };
Chris@10: 
Chris@10: void X(codelet_n1_11) (planner *p) {
Chris@10:      X(kdft_register) (p, n1_11, &desc);
Chris@10: }
Chris@10: 
Chris@10: #else				/* HAVE_FMA */
Chris@10: 
Chris@10: /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 11 -name n1_11 -include n.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 140 FP additions, 100 FP multiplications,
Chris@10:  * (or, 60 additions, 20 multiplications, 80 fused multiply/add),
Chris@10:  * 41 stack variables, 10 constants, and 44 memory accesses
Chris@10:  */
Chris@10: #include "n.h"
Chris@10: 
Chris@10: static void n1_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@10: {
Chris@10:      DK(KP654860733, +0.654860733945285064056925072466293553183791199);
Chris@10:      DK(KP142314838, +0.142314838273285140443792668616369668791051361);
Chris@10:      DK(KP959492973, +0.959492973614497389890368057066327699062454848);
Chris@10:      DK(KP415415013, +0.415415013001886425529274149229623203524004910);
Chris@10:      DK(KP841253532, +0.841253532831181168861811648919367717513292498);
Chris@10:      DK(KP989821441, +0.989821441880932732376092037776718787376519372);
Chris@10:      DK(KP909631995, +0.909631995354518371411715383079028460060241051);
Chris@10:      DK(KP281732556, +0.281732556841429697711417915346616899035777899);
Chris@10:      DK(KP540640817, +0.540640817455597582107635954318691695431770608);
Chris@10:      DK(KP755749574, +0.755749574354258283774035843972344420179717445);
Chris@10:      {
Chris@10: 	  INT i;
Chris@10: 	  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)) {
Chris@10: 	       E T1, TM, T4, TG, Tk, TR, Tw, TN, T7, TK, Ta, TH, Tn, TQ, Td;
Chris@10: 	       E TJ, Tq, TO, Tt, TP, Tg, TI;
Chris@10: 	       {
Chris@10: 		    E T2, T3, Ti, Tj;
Chris@10: 		    T1 = ri[0];
Chris@10: 		    TM = ii[0];
Chris@10: 		    T2 = ri[WS(is, 1)];
Chris@10: 		    T3 = ri[WS(is, 10)];
Chris@10: 		    T4 = T2 + T3;
Chris@10: 		    TG = T3 - T2;
Chris@10: 		    Ti = ii[WS(is, 1)];
Chris@10: 		    Tj = ii[WS(is, 10)];
Chris@10: 		    Tk = Ti - Tj;
Chris@10: 		    TR = Ti + Tj;
Chris@10: 		    {
Chris@10: 			 E Tu, Tv, T5, T6;
Chris@10: 			 Tu = ii[WS(is, 2)];
Chris@10: 			 Tv = ii[WS(is, 9)];
Chris@10: 			 Tw = Tu - Tv;
Chris@10: 			 TN = Tu + Tv;
Chris@10: 			 T5 = ri[WS(is, 2)];
Chris@10: 			 T6 = ri[WS(is, 9)];
Chris@10: 			 T7 = T5 + T6;
Chris@10: 			 TK = T6 - T5;
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	       {
Chris@10: 		    E T8, T9, To, Tp;
Chris@10: 		    T8 = ri[WS(is, 3)];
Chris@10: 		    T9 = ri[WS(is, 8)];
Chris@10: 		    Ta = T8 + T9;
Chris@10: 		    TH = T9 - T8;
Chris@10: 		    {
Chris@10: 			 E Tl, Tm, Tb, Tc;
Chris@10: 			 Tl = ii[WS(is, 3)];
Chris@10: 			 Tm = ii[WS(is, 8)];
Chris@10: 			 Tn = Tl - Tm;
Chris@10: 			 TQ = Tl + Tm;
Chris@10: 			 Tb = ri[WS(is, 4)];
Chris@10: 			 Tc = ri[WS(is, 7)];
Chris@10: 			 Td = Tb + Tc;
Chris@10: 			 TJ = Tc - Tb;
Chris@10: 		    }
Chris@10: 		    To = ii[WS(is, 4)];
Chris@10: 		    Tp = ii[WS(is, 7)];
Chris@10: 		    Tq = To - Tp;
Chris@10: 		    TO = To + Tp;
Chris@10: 		    {
Chris@10: 			 E Tr, Ts, Te, Tf;
Chris@10: 			 Tr = ii[WS(is, 5)];
Chris@10: 			 Ts = ii[WS(is, 6)];
Chris@10: 			 Tt = Tr - Ts;
Chris@10: 			 TP = Tr + Ts;
Chris@10: 			 Te = ri[WS(is, 5)];
Chris@10: 			 Tf = ri[WS(is, 6)];
Chris@10: 			 Tg = Te + Tf;
Chris@10: 			 TI = Tf - Te;
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	       {
Chris@10: 		    E Tx, Th, TZ, T10;
Chris@10: 		    ro[0] = T1 + T4 + T7 + Ta + Td + Tg;
Chris@10: 		    io[0] = TM + TR + TN + TQ + TO + TP;
Chris@10: 		    Tx = FMA(KP755749574, Tk, KP540640817 * Tn) + FNMS(KP909631995, Tt, KP281732556 * Tq) - (KP989821441 * Tw);
Chris@10: 		    Th = FMA(KP841253532, Ta, T1) + FNMS(KP959492973, Td, KP415415013 * Tg) + FNMA(KP142314838, T7, KP654860733 * T4);
Chris@10: 		    ro[WS(os, 7)] = Th - Tx;
Chris@10: 		    ro[WS(os, 4)] = Th + Tx;
Chris@10: 		    TZ = FMA(KP755749574, TG, KP540640817 * TH) + FNMS(KP909631995, TI, KP281732556 * TJ) - (KP989821441 * TK);
Chris@10: 		    T10 = FMA(KP841253532, TQ, TM) + FNMS(KP959492973, TO, KP415415013 * TP) + FNMA(KP142314838, TN, KP654860733 * TR);
Chris@10: 		    io[WS(os, 4)] = TZ + T10;
Chris@10: 		    io[WS(os, 7)] = T10 - TZ;
Chris@10: 		    {
Chris@10: 			 E TX, TY, Tz, Ty;
Chris@10: 			 TX = FMA(KP909631995, TG, KP755749574 * TK) + FNMA(KP540640817, TI, KP989821441 * TJ) - (KP281732556 * TH);
Chris@10: 			 TY = FMA(KP415415013, TR, TM) + FNMS(KP142314838, TO, KP841253532 * TP) + FNMA(KP959492973, TQ, KP654860733 * TN);
Chris@10: 			 io[WS(os, 2)] = TX + TY;
Chris@10: 			 io[WS(os, 9)] = TY - TX;
Chris@10: 			 Tz = FMA(KP909631995, Tk, KP755749574 * Tw) + FNMA(KP540640817, Tt, KP989821441 * Tq) - (KP281732556 * Tn);
Chris@10: 			 Ty = FMA(KP415415013, T4, T1) + FNMS(KP142314838, Td, KP841253532 * Tg) + FNMA(KP959492973, Ta, KP654860733 * T7);
Chris@10: 			 ro[WS(os, 9)] = Ty - Tz;
Chris@10: 			 ro[WS(os, 2)] = Ty + Tz;
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	       {
Chris@10: 		    E TB, TA, TT, TU;
Chris@10: 		    TB = FMA(KP540640817, Tk, KP909631995 * Tw) + FMA(KP989821441, Tn, KP755749574 * Tq) + (KP281732556 * Tt);
Chris@10: 		    TA = FMA(KP841253532, T4, T1) + FNMS(KP959492973, Tg, KP415415013 * T7) + FNMA(KP654860733, Td, KP142314838 * Ta);
Chris@10: 		    ro[WS(os, 10)] = TA - TB;
Chris@10: 		    ro[WS(os, 1)] = TA + TB;
Chris@10: 		    {
Chris@10: 			 E TV, TW, TD, TC;
Chris@10: 			 TV = FMA(KP540640817, TG, KP909631995 * TK) + FMA(KP989821441, TH, KP755749574 * TJ) + (KP281732556 * TI);
Chris@10: 			 TW = FMA(KP841253532, TR, TM) + FNMS(KP959492973, TP, KP415415013 * TN) + FNMA(KP654860733, TO, KP142314838 * TQ);
Chris@10: 			 io[WS(os, 1)] = TV + TW;
Chris@10: 			 io[WS(os, 10)] = TW - TV;
Chris@10: 			 TD = FMA(KP989821441, Tk, KP540640817 * Tq) + FNMS(KP909631995, Tn, KP755749574 * Tt) - (KP281732556 * Tw);
Chris@10: 			 TC = FMA(KP415415013, Ta, T1) + FNMS(KP654860733, Tg, KP841253532 * Td) + FNMA(KP959492973, T7, KP142314838 * T4);
Chris@10: 			 ro[WS(os, 8)] = TC - TD;
Chris@10: 			 ro[WS(os, 3)] = TC + TD;
Chris@10: 		    }
Chris@10: 		    TT = FMA(KP989821441, TG, KP540640817 * TJ) + FNMS(KP909631995, TH, KP755749574 * TI) - (KP281732556 * TK);
Chris@10: 		    TU = FMA(KP415415013, TQ, TM) + FNMS(KP654860733, TP, KP841253532 * TO) + FNMA(KP959492973, TN, KP142314838 * TR);
Chris@10: 		    io[WS(os, 3)] = TT + TU;
Chris@10: 		    io[WS(os, 8)] = TU - TT;
Chris@10: 		    {
Chris@10: 			 E TL, TS, TF, TE;
Chris@10: 			 TL = FMA(KP281732556, TG, KP755749574 * TH) + FNMS(KP909631995, TJ, KP989821441 * TI) - (KP540640817 * TK);
Chris@10: 			 TS = FMA(KP841253532, TN, TM) + FNMS(KP142314838, TP, KP415415013 * TO) + FNMA(KP654860733, TQ, KP959492973 * TR);
Chris@10: 			 io[WS(os, 5)] = TL + TS;
Chris@10: 			 io[WS(os, 6)] = TS - TL;
Chris@10: 			 TF = FMA(KP281732556, Tk, KP755749574 * Tn) + FNMS(KP909631995, Tq, KP989821441 * Tt) - (KP540640817 * Tw);
Chris@10: 			 TE = FMA(KP841253532, T7, T1) + FNMS(KP142314838, Tg, KP415415013 * Td) + FNMA(KP654860733, Ta, KP959492973 * T4);
Chris@10: 			 ro[WS(os, 6)] = TE - TF;
Chris@10: 			 ro[WS(os, 5)] = TE + TF;
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	  }
Chris@10:      }
Chris@10: }
Chris@10: 
Chris@10: static const kdft_desc desc = { 11, "n1_11", {60, 20, 80, 0}, &GENUS, 0, 0, 0, 0 };
Chris@10: 
Chris@10: void X(codelet_n1_11) (planner *p) {
Chris@10:      X(kdft_register) (p, n1_11, &desc);
Chris@10: }
Chris@10: 
Chris@10: #endif				/* HAVE_FMA */