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 07:36:52 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_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 13 -name n1fv_13 -include n1f.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 88 FP additions, 63 FP multiplications,
Chris@10:  * (or, 31 additions, 6 multiplications, 57 fused multiply/add),
Chris@10:  * 96 stack variables, 23 constants, and 26 memory accesses
Chris@10:  */
Chris@10: #include "n1f.h"
Chris@10: 
Chris@10: 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@10: {
Chris@10:      DVK(KP904176221, +0.904176221990848204433795481776887926501523162);
Chris@10:      DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
Chris@10:      DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
Chris@10:      DVK(KP516520780, +0.516520780623489722840901288569017135705033622);
Chris@10:      DVK(KP522026385, +0.522026385161275033714027226654165028300441940);
Chris@10:      DVK(KP957805992, +0.957805992594665126462521754605754580515587217);
Chris@10:      DVK(KP600477271, +0.600477271932665282925769253334763009352012849);
Chris@10:      DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
Chris@10:      DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
Chris@10:      DVK(KP769338817, +0.769338817572980603471413688209101117038278899);
Chris@10:      DVK(KP859542535, +0.859542535098774820163672132761689612766401925);
Chris@10:      DVK(KP581704778, +0.581704778510515730456870384989698884939833902);
Chris@10:      DVK(KP853480001, +0.853480001859823990758994934970528322872359049);
Chris@10:      DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
Chris@10:      DVK(KP226109445, +0.226109445035782405468510155372505010481906348);
Chris@10:      DVK(KP301479260, +0.301479260047709873958013540496673347309208464);
Chris@10:      DVK(KP686558370, +0.686558370781754340655719594850823015421401653);
Chris@10:      DVK(KP514918778, +0.514918778086315755491789696138117261566051239);
Chris@10:      DVK(KP038632954, +0.038632954644348171955506895830342264440241080);
Chris@10:      DVK(KP612264650, +0.612264650376756543746494474777125408779395514);
Chris@10:      DVK(KP302775637, +0.302775637731994646559610633735247973125648287);
Chris@10:      DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
Chris@10:      DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@10:      {
Chris@10: 	  INT i;
Chris@10: 	  const R *xi;
Chris@10: 	  R *xo;
Chris@10: 	  xi = ri;
Chris@10: 	  xo = ro;
Chris@10: 	  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@10: 	       V T1, T7, T2, Tg, Tf, TN, Th, Tq, Ta, Tj, T5, Tr, Tk;
Chris@10: 	       T1 = LD(&(xi[0]), ivs, &(xi[0]));
Chris@10: 	       {
Chris@10: 		    V Td, Te, T8, T9, T3, T4;
Chris@10: 		    Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Chris@10: 		    Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 		    T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Chris@10: 		    T8 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Chris@10: 		    T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Chris@10: 		    T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 		    T3 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 		    T4 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 		    Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 		    Tf = VADD(Td, Te);
Chris@10: 		    TN = VSUB(Td, Te);
Chris@10: 		    Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Chris@10: 		    Tq = VSUB(T8, T9);
Chris@10: 		    Ta = VADD(T8, T9);
Chris@10: 		    Tj = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 		    T5 = VADD(T3, T4);
Chris@10: 		    Tr = VSUB(T4, T3);
Chris@10: 		    Tk = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Chris@10: 	       }
Chris@10: 	       {
Chris@10: 		    V Tt, Ti, Ty, Tb, Ts, TQ, Tx, T6, Tu, Tl;
Chris@10: 		    Tt = VSUB(Tg, Th);
Chris@10: 		    Ti = VADD(Tg, Th);
Chris@10: 		    Ty = VFMS(LDK(KP500000000), Ta, T7);
Chris@10: 		    Tb = VADD(T7, Ta);
Chris@10: 		    Ts = VSUB(Tq, Tr);
Chris@10: 		    TQ = VADD(Tr, Tq);
Chris@10: 		    Tx = VFNMS(LDK(KP500000000), T5, T2);
Chris@10: 		    T6 = VADD(T2, T5);
Chris@10: 		    Tu = VSUB(Tj, Tk);
Chris@10: 		    Tl = VADD(Tj, Tk);
Chris@10: 		    {
Chris@10: 			 V TK, Tz, Tc, TX, Tv, TO, TL, Tm;
Chris@10: 			 TK = VADD(Tx, Ty);
Chris@10: 			 Tz = VSUB(Tx, Ty);
Chris@10: 			 Tc = VADD(T6, Tb);
Chris@10: 			 TX = VSUB(T6, Tb);
Chris@10: 			 Tv = VSUB(Tt, Tu);
Chris@10: 			 TO = VADD(Tt, Tu);
Chris@10: 			 TL = VSUB(Ti, Tl);
Chris@10: 			 Tm = VADD(Ti, Tl);
Chris@10: 			 {
Chris@10: 			      V TF, Tw, TP, TY, TT, TM, TA, Tn;
Chris@10: 			      TF = VSUB(Ts, Tv);
Chris@10: 			      Tw = VADD(Ts, Tv);
Chris@10: 			      TP = VFNMS(LDK(KP500000000), TO, TN);
Chris@10: 			      TY = VADD(TN, TO);
Chris@10: 			      TT = VFNMS(LDK(KP866025403), TL, TK);
Chris@10: 			      TM = VFMA(LDK(KP866025403), TL, TK);
Chris@10: 			      TA = VFNMS(LDK(KP500000000), Tm, Tf);
Chris@10: 			      Tn = VADD(Tf, Tm);
Chris@10: 			      {
Chris@10: 				   V T1f, T1n, TI, T18, T1k, T1c, TD, T17, T10, T1m, T16, T1e, TU, TR;
Chris@10: 				   TU = VFNMS(LDK(KP866025403), TQ, TP);
Chris@10: 				   TR = VFMA(LDK(KP866025403), TQ, TP);
Chris@10: 				   {
Chris@10: 					V TZ, T15, TE, TB;
Chris@10: 					TZ = VFMA(LDK(KP302775637), TY, TX);
Chris@10: 					T15 = VFNMS(LDK(KP302775637), TX, TY);
Chris@10: 					TE = VSUB(Tz, TA);
Chris@10: 					TB = VADD(Tz, TA);
Chris@10: 					{
Chris@10: 					     V TH, To, TV, T13;
Chris@10: 					     TH = VSUB(Tc, Tn);
Chris@10: 					     To = VADD(Tc, Tn);
Chris@10: 					     TV = VFNMS(LDK(KP612264650), TU, TT);
Chris@10: 					     T13 = VFMA(LDK(KP612264650), TT, TU);
Chris@10: 					     {
Chris@10: 						  V TS, T12, TG, T1b;
Chris@10: 						  TS = VFNMS(LDK(KP038632954), TR, TM);
Chris@10: 						  T12 = VFMA(LDK(KP038632954), TM, TR);
Chris@10: 						  TG = VFNMS(LDK(KP514918778), TF, TE);
Chris@10: 						  T1b = VFMA(LDK(KP686558370), TE, TF);
Chris@10: 						  {
Chris@10: 						       V TC, T1a, Tp, TW, T14;
Chris@10: 						       TC = VFMA(LDK(KP301479260), TB, Tw);
Chris@10: 						       T1a = VFNMS(LDK(KP226109445), Tw, TB);
Chris@10: 						       Tp = VFNMS(LDK(KP083333333), To, T1);
Chris@10: 						       ST(&(xo[0]), VADD(T1, To), ovs, &(xo[0]));
Chris@10: 						       T1f = VFMA(LDK(KP853480001), TV, TS);
Chris@10: 						       TW = VFNMS(LDK(KP853480001), TV, TS);
Chris@10: 						       T1n = VFMA(LDK(KP853480001), T13, T12);
Chris@10: 						       T14 = VFNMS(LDK(KP853480001), T13, T12);
Chris@10: 						       TI = VFMA(LDK(KP581704778), TH, TG);
Chris@10: 						       T18 = VFNMS(LDK(KP859542535), TG, TH);
Chris@10: 						       T1k = VFMA(LDK(KP769338817), T1b, T1a);
Chris@10: 						       T1c = VFNMS(LDK(KP769338817), T1b, T1a);
Chris@10: 						       TD = VFMA(LDK(KP503537032), TC, Tp);
Chris@10: 						       T17 = VFNMS(LDK(KP251768516), TC, Tp);
Chris@10: 						       T10 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), TZ, TW));
Chris@10: 						       T1m = VFNMS(LDK(KP522026385), TW, TZ);
Chris@10: 						       T16 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), T15, T14));
Chris@10: 						       T1e = VFNMS(LDK(KP522026385), T14, T15);
Chris@10: 						  }
Chris@10: 					     }
Chris@10: 					}
Chris@10: 				   }
Chris@10: 				   {
Chris@10: 					V T1o, T1q, T1g, T1i, T1d, T1h, T1l, T1p;
Chris@10: 					{
Chris@10: 					     V T11, TJ, T19, T1j;
Chris@10: 					     T11 = VFMA(LDK(KP516520780), TI, TD);
Chris@10: 					     TJ = VFNMS(LDK(KP516520780), TI, TD);
Chris@10: 					     T19 = VFMA(LDK(KP300462606), T18, T17);
Chris@10: 					     T1j = VFNMS(LDK(KP300462606), T18, T17);
Chris@10: 					     T1o = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1n, T1m));
Chris@10: 					     T1q = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1n, T1m));
Chris@10: 					     T1g = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1f, T1e));
Chris@10: 					     T1i = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1f, T1e));
Chris@10: 					     ST(&(xo[WS(os, 12)]), VFNMSI(T16, T11), ovs, &(xo[0]));
Chris@10: 					     ST(&(xo[WS(os, 1)]), VFMAI(T16, T11), ovs, &(xo[WS(os, 1)]));
Chris@10: 					     ST(&(xo[WS(os, 8)]), VFMAI(T10, TJ), ovs, &(xo[0]));
Chris@10: 					     ST(&(xo[WS(os, 5)]), VFNMSI(T10, TJ), ovs, &(xo[WS(os, 1)]));
Chris@10: 					     T1d = VFNMS(LDK(KP503537032), T1c, T19);
Chris@10: 					     T1h = VFMA(LDK(KP503537032), T1c, T19);
Chris@10: 					     T1l = VFNMS(LDK(KP503537032), T1k, T1j);
Chris@10: 					     T1p = VFMA(LDK(KP503537032), T1k, T1j);
Chris@10: 					}
Chris@10: 					ST(&(xo[WS(os, 9)]), VFMAI(T1g, T1d), ovs, &(xo[WS(os, 1)]));
Chris@10: 					ST(&(xo[WS(os, 4)]), VFNMSI(T1g, T1d), ovs, &(xo[0]));
Chris@10: 					ST(&(xo[WS(os, 10)]), VFNMSI(T1i, T1h), ovs, &(xo[0]));
Chris@10: 					ST(&(xo[WS(os, 3)]), VFMAI(T1i, T1h), ovs, &(xo[WS(os, 1)]));
Chris@10: 					ST(&(xo[WS(os, 7)]), VFMAI(T1o, T1l), ovs, &(xo[WS(os, 1)]));
Chris@10: 					ST(&(xo[WS(os, 6)]), VFNMSI(T1o, T1l), ovs, &(xo[0]));
Chris@10: 					ST(&(xo[WS(os, 11)]), VFMAI(T1q, T1p), ovs, &(xo[WS(os, 1)]));
Chris@10: 					ST(&(xo[WS(os, 2)]), VFNMSI(T1q, T1p), ovs, &(xo[0]));
Chris@10: 				   }
Chris@10: 			      }
Chris@10: 			 }
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	  }
Chris@10:      }
Chris@10:      VLEAVE();
Chris@10: }
Chris@10: 
Chris@10: static const kdft_desc desc = { 13, XSIMD_STRING("n1fv_13"), {31, 6, 57, 0}, &GENUS, 0, 0, 0, 0 };
Chris@10: 
Chris@10: void XSIMD(codelet_n1fv_13) (planner *p) {
Chris@10:      X(kdft_register) (p, n1fv_13, &desc);
Chris@10: }
Chris@10: 
Chris@10: #else				/* HAVE_FMA */
Chris@10: 
Chris@10: /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 13 -name n1fv_13 -include n1f.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 88 FP additions, 34 FP multiplications,
Chris@10:  * (or, 69 additions, 15 multiplications, 19 fused multiply/add),
Chris@10:  * 60 stack variables, 20 constants, and 26 memory accesses
Chris@10:  */
Chris@10: #include "n1f.h"
Chris@10: 
Chris@10: 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@10: {
Chris@10:      DVK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
Chris@10:      DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
Chris@10:      DVK(KP075902986, +0.075902986037193865983102897245103540356428373);
Chris@10:      DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
Chris@10:      DVK(KP132983124, +0.132983124607418643793760531921092974399165133);
Chris@10:      DVK(KP258260390, +0.258260390311744861420450644284508567852516811);
Chris@10:      DVK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
Chris@10:      DVK(KP300238635, +0.300238635966332641462884626667381504676006424);
Chris@10:      DVK(KP011599105, +0.011599105605768290721655456654083252189827041);
Chris@10:      DVK(KP156891391, +0.156891391051584611046832726756003269660212636);
Chris@10:      DVK(KP256247671, +0.256247671582936600958684654061725059144125175);
Chris@10:      DVK(KP174138601, +0.174138601152135905005660794929264742616964676);
Chris@10:      DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
Chris@10:      DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
Chris@10:      DVK(KP113854479, +0.113854479055790798974654345867655310534642560);
Chris@10:      DVK(KP265966249, +0.265966249214837287587521063842185948798330267);
Chris@10:      DVK(KP387390585, +0.387390585467617292130675966426762851778775217);
Chris@10:      DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
Chris@10:      DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
Chris@10:      DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@10:      {
Chris@10: 	  INT i;
Chris@10: 	  const R *xi;
Chris@10: 	  R *xo;
Chris@10: 	  xi = ri;
Chris@10: 	  xo = ro;
Chris@10: 	  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@10: 	       V TW, Tb, Tm, Tu, TC, TR, TX, TK, TU, Tz, TB, TN, TT;
Chris@10: 	       TW = LD(&(xi[0]), ivs, &(xi[0]));
Chris@10: 	       {
Chris@10: 		    V T3, TH, Tl, Tw, Tp, Tg, Tv, To, T6, Tr, T9, Ts, Ta, TI, T1;
Chris@10: 		    V T2, Tq, Tt;
Chris@10: 		    T1 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Chris@10: 		    T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 		    T3 = VSUB(T1, T2);
Chris@10: 		    TH = VADD(T1, T2);
Chris@10: 		    {
Chris@10: 			 V Th, Ti, Tj, Tk;
Chris@10: 			 Th = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Chris@10: 			 Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Chris@10: 			 Tj = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Chris@10: 			 Tk = VADD(Ti, Tj);
Chris@10: 			 Tl = VADD(Th, Tk);
Chris@10: 			 Tw = VSUB(Ti, Tj);
Chris@10: 			 Tp = VFNMS(LDK(KP500000000), Tk, Th);
Chris@10: 		    }
Chris@10: 		    {
Chris@10: 			 V Tc, Td, Te, Tf;
Chris@10: 			 Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 			 Td = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 			 Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 			 Tf = VADD(Td, Te);
Chris@10: 			 Tg = VADD(Tc, Tf);
Chris@10: 			 Tv = VSUB(Td, Te);
Chris@10: 			 To = VFNMS(LDK(KP500000000), Tf, Tc);
Chris@10: 		    }
Chris@10: 		    {
Chris@10: 			 V T4, T5, T7, T8;
Chris@10: 			 T4 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 			 T5 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Chris@10: 			 T6 = VSUB(T4, T5);
Chris@10: 			 Tr = VADD(T4, T5);
Chris@10: 			 T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Chris@10: 			 T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Chris@10: 			 T9 = VSUB(T7, T8);
Chris@10: 			 Ts = VADD(T7, T8);
Chris@10: 		    }
Chris@10: 		    Ta = VADD(T6, T9);
Chris@10: 		    TI = VADD(Tr, Ts);
Chris@10: 		    Tb = VADD(T3, Ta);
Chris@10: 		    Tm = VSUB(Tg, Tl);
Chris@10: 		    Tq = VSUB(To, Tp);
Chris@10: 		    Tt = VMUL(LDK(KP866025403), VSUB(Tr, Ts));
Chris@10: 		    Tu = VADD(Tq, Tt);
Chris@10: 		    TC = VSUB(Tq, Tt);
Chris@10: 		    {
Chris@10: 			 V TP, TQ, TG, TJ;
Chris@10: 			 TP = VADD(Tg, Tl);
Chris@10: 			 TQ = VADD(TH, TI);
Chris@10: 			 TR = VMUL(LDK(KP300462606), VSUB(TP, TQ));
Chris@10: 			 TX = VADD(TP, TQ);
Chris@10: 			 TG = VADD(To, Tp);
Chris@10: 			 TJ = VFNMS(LDK(KP500000000), TI, TH);
Chris@10: 			 TK = VSUB(TG, TJ);
Chris@10: 			 TU = VADD(TG, TJ);
Chris@10: 		    }
Chris@10: 		    {
Chris@10: 			 V Tx, Ty, TL, TM;
Chris@10: 			 Tx = VMUL(LDK(KP866025403), VSUB(Tv, Tw));
Chris@10: 			 Ty = VFNMS(LDK(KP500000000), Ta, T3);
Chris@10: 			 Tz = VSUB(Tx, Ty);
Chris@10: 			 TB = VADD(Tx, Ty);
Chris@10: 			 TL = VADD(Tv, Tw);
Chris@10: 			 TM = VSUB(T6, T9);
Chris@10: 			 TN = VSUB(TL, TM);
Chris@10: 			 TT = VADD(TL, TM);
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	       ST(&(xo[0]), VADD(TW, TX), ovs, &(xo[0]));
Chris@10: 	       {
Chris@10: 		    V T19, T1n, T14, T13, T1f, T1k, Tn, TE, T1e, T1j, TS, T1m, TZ, T1c, TA;
Chris@10: 		    V TD;
Chris@10: 		    {
Chris@10: 			 V T17, T18, T11, T12;
Chris@10: 			 T17 = VFMA(LDK(KP387390585), TN, VMUL(LDK(KP265966249), TK));
Chris@10: 			 T18 = VFNMS(LDK(KP503537032), TU, VMUL(LDK(KP113854479), TT));
Chris@10: 			 T19 = VSUB(T17, T18);
Chris@10: 			 T1n = VADD(T17, T18);
Chris@10: 			 T14 = VFMA(LDK(KP575140729), Tm, VMUL(LDK(KP174138601), Tb));
Chris@10: 			 T11 = VFNMS(LDK(KP156891391), TB, VMUL(LDK(KP256247671), TC));
Chris@10: 			 T12 = VFMA(LDK(KP011599105), Tz, VMUL(LDK(KP300238635), Tu));
Chris@10: 			 T13 = VSUB(T11, T12);
Chris@10: 			 T1f = VADD(T14, T13);
Chris@10: 			 T1k = VMUL(LDK(KP1_732050807), VADD(T11, T12));
Chris@10: 		    }
Chris@10: 		    Tn = VFNMS(LDK(KP174138601), Tm, VMUL(LDK(KP575140729), Tb));
Chris@10: 		    TA = VFNMS(LDK(KP300238635), Tz, VMUL(LDK(KP011599105), Tu));
Chris@10: 		    TD = VFMA(LDK(KP256247671), TB, VMUL(LDK(KP156891391), TC));
Chris@10: 		    TE = VSUB(TA, TD);
Chris@10: 		    T1e = VMUL(LDK(KP1_732050807), VADD(TD, TA));
Chris@10: 		    T1j = VSUB(Tn, TE);
Chris@10: 		    {
Chris@10: 			 V TO, T1b, TV, TY, T1a;
Chris@10: 			 TO = VFNMS(LDK(KP132983124), TN, VMUL(LDK(KP258260390), TK));
Chris@10: 			 T1b = VSUB(TR, TO);
Chris@10: 			 TV = VFMA(LDK(KP251768516), TT, VMUL(LDK(KP075902986), TU));
Chris@10: 			 TY = VFNMS(LDK(KP083333333), TX, TW);
Chris@10: 			 T1a = VSUB(TY, TV);
Chris@10: 			 TS = VFMA(LDK(KP2_000000000), TO, TR);
Chris@10: 			 T1m = VADD(T1b, T1a);
Chris@10: 			 TZ = VFMA(LDK(KP2_000000000), TV, TY);
Chris@10: 			 T1c = VSUB(T1a, T1b);
Chris@10: 		    }
Chris@10: 		    {
Chris@10: 			 V TF, T10, T1l, T1o;
Chris@10: 			 TF = VBYI(VFMA(LDK(KP2_000000000), TE, Tn));
Chris@10: 			 T10 = VADD(TS, TZ);
Chris@10: 			 ST(&(xo[WS(os, 1)]), VADD(TF, T10), ovs, &(xo[WS(os, 1)]));
Chris@10: 			 ST(&(xo[WS(os, 12)]), VSUB(T10, TF), ovs, &(xo[0]));
Chris@10: 			 {
Chris@10: 			      V T15, T16, T1p, T1q;
Chris@10: 			      T15 = VBYI(VFMS(LDK(KP2_000000000), T13, T14));
Chris@10: 			      T16 = VSUB(TZ, TS);
Chris@10: 			      ST(&(xo[WS(os, 5)]), VADD(T15, T16), ovs, &(xo[WS(os, 1)]));
Chris@10: 			      ST(&(xo[WS(os, 8)]), VSUB(T16, T15), ovs, &(xo[0]));
Chris@10: 			      T1p = VADD(T1n, T1m);
Chris@10: 			      T1q = VBYI(VADD(T1j, T1k));
Chris@10: 			      ST(&(xo[WS(os, 4)]), VSUB(T1p, T1q), ovs, &(xo[0]));
Chris@10: 			      ST(&(xo[WS(os, 9)]), VADD(T1q, T1p), ovs, &(xo[WS(os, 1)]));
Chris@10: 			 }
Chris@10: 			 T1l = VBYI(VSUB(T1j, T1k));
Chris@10: 			 T1o = VSUB(T1m, T1n);
Chris@10: 			 ST(&(xo[WS(os, 3)]), VADD(T1l, T1o), ovs, &(xo[WS(os, 1)]));
Chris@10: 			 ST(&(xo[WS(os, 10)]), VSUB(T1o, T1l), ovs, &(xo[0]));
Chris@10: 			 {
Chris@10: 			      V T1h, T1i, T1d, T1g;
Chris@10: 			      T1h = VBYI(VSUB(T1e, T1f));
Chris@10: 			      T1i = VSUB(T1c, T19);
Chris@10: 			      ST(&(xo[WS(os, 6)]), VADD(T1h, T1i), ovs, &(xo[0]));
Chris@10: 			      ST(&(xo[WS(os, 7)]), VSUB(T1i, T1h), ovs, &(xo[WS(os, 1)]));
Chris@10: 			      T1d = VADD(T19, T1c);
Chris@10: 			      T1g = VBYI(VADD(T1e, T1f));
Chris@10: 			      ST(&(xo[WS(os, 2)]), VSUB(T1d, T1g), ovs, &(xo[0]));
Chris@10: 			      ST(&(xo[WS(os, 11)]), VADD(T1g, T1d), ovs, &(xo[WS(os, 1)]));
Chris@10: 			 }
Chris@10: 		    }
Chris@10: 	       }
Chris@10: 	  }
Chris@10:      }
Chris@10:      VLEAVE();
Chris@10: }
Chris@10: 
Chris@10: static const kdft_desc desc = { 13, XSIMD_STRING("n1fv_13"), {69, 15, 19, 0}, &GENUS, 0, 0, 0, 0 };
Chris@10: 
Chris@10: void XSIMD(codelet_n1fv_13) (planner *p) {
Chris@10:      X(kdft_register) (p, n1fv_13, &desc);
Chris@10: }
Chris@10: 
Chris@10: #endif				/* HAVE_FMA */