cannam@167: /*
cannam@167:  * Copyright (c) 2003, 2007-14 Matteo Frigo
cannam@167:  * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
cannam@167:  *
cannam@167:  * This program is free software; you can redistribute it and/or modify
cannam@167:  * it under the terms of the GNU General Public License as published by
cannam@167:  * the Free Software Foundation; either version 2 of the License, or
cannam@167:  * (at your option) any later version.
cannam@167:  *
cannam@167:  * This program is distributed in the hope that it will be useful,
cannam@167:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
cannam@167:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
cannam@167:  * GNU General Public License for more details.
cannam@167:  *
cannam@167:  * You should have received a copy of the GNU General Public License
cannam@167:  * along with this program; if not, write to the Free Software
cannam@167:  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
cannam@167:  *
cannam@167:  */
cannam@167: 
cannam@167: /* This file was automatically generated --- DO NOT EDIT */
cannam@167: /* Generated on Thu May 24 08:06:26 EDT 2018 */
cannam@167: 
cannam@167: #include "rdft/codelet-rdft.h"
cannam@167: 
cannam@167: #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
cannam@167: 
cannam@167: /* Generated by: ../../../genfft/gen_r2cf.native -fma -compact -variables 4 -pipeline-latency 4 -n 11 -name r2cf_11 -include rdft/scalar/r2cf.h */
cannam@167: 
cannam@167: /*
cannam@167:  * This function contains 60 FP additions, 50 FP multiplications,
cannam@167:  * (or, 15 additions, 5 multiplications, 45 fused multiply/add),
cannam@167:  * 42 stack variables, 10 constants, and 22 memory accesses
cannam@167:  */
cannam@167: #include "rdft/scalar/r2cf.h"
cannam@167: 
cannam@167: static void r2cf_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
cannam@167: {
cannam@167:      DK(KP918985947, +0.918985947228994779780736114132655398124909697);
cannam@167:      DK(KP989821441, +0.989821441880932732376092037776718787376519372);
cannam@167:      DK(KP830830026, +0.830830026003772851058548298459246407048009821);
cannam@167:      DK(KP715370323, +0.715370323453429719112414662767260662417897278);
cannam@167:      DK(KP959492973, +0.959492973614497389890368057066327699062454848);
cannam@167:      DK(KP876768831, +0.876768831002589333891339807079336796764054852);
cannam@167:      DK(KP778434453, +0.778434453334651800608337670740821884709317477);
cannam@167:      DK(KP634356270, +0.634356270682424498893150776899916060542806975);
cannam@167:      DK(KP342584725, +0.342584725681637509502641509861112333758894680);
cannam@167:      DK(KP521108558, +0.521108558113202722944698153526659300680427422);
cannam@167:      {
cannam@167: 	  INT i;
cannam@167: 	  for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) {
cannam@167: 	       E T1, T4, TC, Tg, TE, T7, TD, Ta, TF, Td, TB, TG, TM, TS, TJ;
cannam@167: 	       E TP, Ty, Tq, Ti, Tu, Tm, T5, T6;
cannam@167: 	       T1 = R0[0];
cannam@167: 	       {
cannam@167: 		    E T2, T3, Te, Tf;
cannam@167: 		    T2 = R1[0];
cannam@167: 		    T3 = R0[WS(rs, 5)];
cannam@167: 		    T4 = T2 + T3;
cannam@167: 		    TC = T3 - T2;
cannam@167: 		    Te = R1[WS(rs, 2)];
cannam@167: 		    Tf = R0[WS(rs, 3)];
cannam@167: 		    Tg = Te + Tf;
cannam@167: 		    TE = Tf - Te;
cannam@167: 	       }
cannam@167: 	       T5 = R0[WS(rs, 1)];
cannam@167: 	       T6 = R1[WS(rs, 4)];
cannam@167: 	       T7 = T5 + T6;
cannam@167: 	       TD = T5 - T6;
cannam@167: 	       {
cannam@167: 		    E T8, T9, Tb, Tc;
cannam@167: 		    T8 = R1[WS(rs, 1)];
cannam@167: 		    T9 = R0[WS(rs, 4)];
cannam@167: 		    Ta = T8 + T9;
cannam@167: 		    TF = T9 - T8;
cannam@167: 		    Tb = R0[WS(rs, 2)];
cannam@167: 		    Tc = R1[WS(rs, 3)];
cannam@167: 		    Td = Tb + Tc;
cannam@167: 		    TB = Tb - Tc;
cannam@167: 	       }
cannam@167: 	       TG = FMA(KP521108558, TF, TE);
cannam@167: 	       TM = FNMS(KP521108558, TD, TB);
cannam@167: 	       TS = FMA(KP521108558, TC, TD);
cannam@167: 	       TJ = FMA(KP521108558, TE, TC);
cannam@167: 	       TP = FNMS(KP521108558, TB, TF);
cannam@167: 	       {
cannam@167: 		    E Tx, Tp, Th, Tt, Tl;
cannam@167: 		    Tx = FNMS(KP342584725, Ta, T7);
cannam@167: 		    Ty = FNMS(KP634356270, Tx, Td);
cannam@167: 		    Tp = FNMS(KP342584725, T4, Ta);
cannam@167: 		    Tq = FNMS(KP634356270, Tp, Tg);
cannam@167: 		    Th = FNMS(KP342584725, Tg, Td);
cannam@167: 		    Ti = FNMS(KP634356270, Th, Ta);
cannam@167: 		    Tt = FNMS(KP342584725, Td, T4);
cannam@167: 		    Tu = FNMS(KP634356270, Tt, T7);
cannam@167: 		    Tl = FNMS(KP342584725, T7, Tg);
cannam@167: 		    Tm = FNMS(KP634356270, Tl, T4);
cannam@167: 	       }
cannam@167: 	       {
cannam@167: 		    E To, Tn, TI, TH;
cannam@167: 		    {
cannam@167: 			 E Tk, Tj, TU, TT;
cannam@167: 			 Tj = FNMS(KP778434453, Ti, T7);
cannam@167: 			 Tk = FNMS(KP876768831, Tj, T4);
cannam@167: 			 Cr[WS(csr, 5)] = FNMS(KP959492973, Tk, T1);
cannam@167: 			 TT = FMA(KP715370323, TS, TF);
cannam@167: 			 TU = FMA(KP830830026, TT, TB);
cannam@167: 			 Ci[WS(csi, 5)] = KP989821441 * (FMA(KP918985947, TU, TE));
cannam@167: 		    }
cannam@167: 		    Tn = FNMS(KP778434453, Tm, Ta);
cannam@167: 		    To = FNMS(KP876768831, Tn, Td);
cannam@167: 		    Cr[WS(csr, 4)] = FNMS(KP959492973, To, T1);
cannam@167: 		    {
cannam@167: 			 E TR, TQ, Ts, Tr;
cannam@167: 			 TQ = FMA(KP715370323, TP, TC);
cannam@167: 			 TR = FNMS(KP830830026, TQ, TE);
cannam@167: 			 Ci[WS(csi, 4)] = KP989821441 * (FNMS(KP918985947, TR, TD));
cannam@167: 			 Tr = FNMS(KP778434453, Tq, Td);
cannam@167: 			 Ts = FNMS(KP876768831, Tr, T7);
cannam@167: 			 Cr[WS(csr, 3)] = FNMS(KP959492973, Ts, T1);
cannam@167: 		    }
cannam@167: 		    {
cannam@167: 			 E TO, TN, Tw, Tv;
cannam@167: 			 TN = FNMS(KP715370323, TM, TE);
cannam@167: 			 TO = FNMS(KP830830026, TN, TF);
cannam@167: 			 Ci[WS(csi, 3)] = KP989821441 * (FNMS(KP918985947, TO, TC));
cannam@167: 			 Tv = FNMS(KP778434453, Tu, Tg);
cannam@167: 			 Tw = FNMS(KP876768831, Tv, Ta);
cannam@167: 			 Cr[WS(csr, 2)] = FNMS(KP959492973, Tw, T1);
cannam@167: 			 Cr[0] = T1 + T4 + T7 + Ta + Td + Tg;
cannam@167: 		    }
cannam@167: 		    TH = FMA(KP715370323, TG, TD);
cannam@167: 		    TI = FNMS(KP830830026, TH, TC);
cannam@167: 		    Ci[WS(csi, 2)] = KP989821441 * (FMA(KP918985947, TI, TB));
cannam@167: 		    {
cannam@167: 			 E TL, TK, TA, Tz;
cannam@167: 			 TK = FNMS(KP715370323, TJ, TB);
cannam@167: 			 TL = FMA(KP830830026, TK, TD);
cannam@167: 			 Ci[WS(csi, 1)] = KP989821441 * (FNMS(KP918985947, TL, TF));
cannam@167: 			 Tz = FNMS(KP778434453, Ty, T4);
cannam@167: 			 TA = FNMS(KP876768831, Tz, Tg);
cannam@167: 			 Cr[WS(csr, 1)] = FNMS(KP959492973, TA, T1);
cannam@167: 		    }
cannam@167: 	       }
cannam@167: 	  }
cannam@167:      }
cannam@167: }
cannam@167: 
cannam@167: static const kr2c_desc desc = { 11, "r2cf_11", {15, 5, 45, 0}, &GENUS };
cannam@167: 
cannam@167: void X(codelet_r2cf_11) (planner *p) {
cannam@167:      X(kr2c_register) (p, r2cf_11, &desc);
cannam@167: }
cannam@167: 
cannam@167: #else
cannam@167: 
cannam@167: /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 11 -name r2cf_11 -include rdft/scalar/r2cf.h */
cannam@167: 
cannam@167: /*
cannam@167:  * This function contains 60 FP additions, 50 FP multiplications,
cannam@167:  * (or, 20 additions, 10 multiplications, 40 fused multiply/add),
cannam@167:  * 28 stack variables, 10 constants, and 22 memory accesses
cannam@167:  */
cannam@167: #include "rdft/scalar/r2cf.h"
cannam@167: 
cannam@167: static void r2cf_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
cannam@167: {
cannam@167:      DK(KP654860733, +0.654860733945285064056925072466293553183791199);
cannam@167:      DK(KP142314838, +0.142314838273285140443792668616369668791051361);
cannam@167:      DK(KP959492973, +0.959492973614497389890368057066327699062454848);
cannam@167:      DK(KP415415013, +0.415415013001886425529274149229623203524004910);
cannam@167:      DK(KP841253532, +0.841253532831181168861811648919367717513292498);
cannam@167:      DK(KP989821441, +0.989821441880932732376092037776718787376519372);
cannam@167:      DK(KP909631995, +0.909631995354518371411715383079028460060241051);
cannam@167:      DK(KP281732556, +0.281732556841429697711417915346616899035777899);
cannam@167:      DK(KP540640817, +0.540640817455597582107635954318691695431770608);
cannam@167:      DK(KP755749574, +0.755749574354258283774035843972344420179717445);
cannam@167:      {
cannam@167: 	  INT i;
cannam@167: 	  for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) {
cannam@167: 	       E T1, T4, Tl, Tg, Th, Td, Ti, Ta, Tk, T7, Tj, Tb, Tc;
cannam@167: 	       T1 = R0[0];
cannam@167: 	       {
cannam@167: 		    E T2, T3, Te, Tf;
cannam@167: 		    T2 = R0[WS(rs, 1)];
cannam@167: 		    T3 = R1[WS(rs, 4)];
cannam@167: 		    T4 = T2 + T3;
cannam@167: 		    Tl = T3 - T2;
cannam@167: 		    Te = R1[0];
cannam@167: 		    Tf = R0[WS(rs, 5)];
cannam@167: 		    Tg = Te + Tf;
cannam@167: 		    Th = Tf - Te;
cannam@167: 	       }
cannam@167: 	       Tb = R1[WS(rs, 1)];
cannam@167: 	       Tc = R0[WS(rs, 4)];
cannam@167: 	       Td = Tb + Tc;
cannam@167: 	       Ti = Tc - Tb;
cannam@167: 	       {
cannam@167: 		    E T8, T9, T5, T6;
cannam@167: 		    T8 = R1[WS(rs, 2)];
cannam@167: 		    T9 = R0[WS(rs, 3)];
cannam@167: 		    Ta = T8 + T9;
cannam@167: 		    Tk = T9 - T8;
cannam@167: 		    T5 = R0[WS(rs, 2)];
cannam@167: 		    T6 = R1[WS(rs, 3)];
cannam@167: 		    T7 = T5 + T6;
cannam@167: 		    Tj = T6 - T5;
cannam@167: 	       }
cannam@167: 	       Ci[WS(csi, 4)] = FMA(KP755749574, Th, KP540640817 * Ti) + FNMS(KP909631995, Tk, KP281732556 * Tj) - (KP989821441 * Tl);
cannam@167: 	       Cr[WS(csr, 4)] = FMA(KP841253532, Td, T1) + FNMS(KP959492973, T7, KP415415013 * Ta) + FNMA(KP142314838, T4, KP654860733 * Tg);
cannam@167: 	       Ci[WS(csi, 2)] = FMA(KP909631995, Th, KP755749574 * Tl) + FNMA(KP540640817, Tk, KP989821441 * Tj) - (KP281732556 * Ti);
cannam@167: 	       Ci[WS(csi, 5)] = FMA(KP281732556, Th, KP755749574 * Ti) + FNMS(KP909631995, Tj, KP989821441 * Tk) - (KP540640817 * Tl);
cannam@167: 	       Ci[WS(csi, 1)] = FMA(KP540640817, Th, KP909631995 * Tl) + FMA(KP989821441, Ti, KP755749574 * Tj) + (KP281732556 * Tk);
cannam@167: 	       Ci[WS(csi, 3)] = FMA(KP989821441, Th, KP540640817 * Tj) + FNMS(KP909631995, Ti, KP755749574 * Tk) - (KP281732556 * Tl);
cannam@167: 	       Cr[WS(csr, 3)] = FMA(KP415415013, Td, T1) + FNMS(KP654860733, Ta, KP841253532 * T7) + FNMA(KP959492973, T4, KP142314838 * Tg);
cannam@167: 	       Cr[WS(csr, 1)] = FMA(KP841253532, Tg, T1) + FNMS(KP959492973, Ta, KP415415013 * T4) + FNMA(KP654860733, T7, KP142314838 * Td);
cannam@167: 	       Cr[0] = T1 + Tg + T4 + Td + T7 + Ta;
cannam@167: 	       Cr[WS(csr, 2)] = FMA(KP415415013, Tg, T1) + FNMS(KP142314838, T7, KP841253532 * Ta) + FNMA(KP959492973, Td, KP654860733 * T4);
cannam@167: 	       Cr[WS(csr, 5)] = FMA(KP841253532, T4, T1) + FNMS(KP142314838, Ta, KP415415013 * T7) + FNMA(KP654860733, Td, KP959492973 * Tg);
cannam@167: 	  }
cannam@167:      }
cannam@167: }
cannam@167: 
cannam@167: static const kr2c_desc desc = { 11, "r2cf_11", {20, 10, 40, 0}, &GENUS };
cannam@167: 
cannam@167: void X(codelet_r2cf_11) (planner *p) {
cannam@167:      X(kr2c_register) (p, r2cf_11, &desc);
cannam@167: }
cannam@167: 
cannam@167: #endif