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 08:07:28 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_r2cb.native -fma -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -name r2cb_10 -include rdft/scalar/r2cb.h */
cannam@167: 
cannam@167: /*
cannam@167:  * This function contains 34 FP additions, 20 FP multiplications,
cannam@167:  * (or, 14 additions, 0 multiplications, 20 fused multiply/add),
cannam@167:  * 26 stack variables, 5 constants, and 20 memory accesses
cannam@167:  */
cannam@167: #include "rdft/scalar/r2cb.h"
cannam@167: 
cannam@167: static void r2cb_10(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(KP1_902113032, +1.902113032590307144232878666758764286811397268);
cannam@167:      DK(KP1_118033988, +1.118033988749894848204586834365638117720309180);
cannam@167:      DK(KP500000000, +0.500000000000000000000000000000000000000000000);
cannam@167:      DK(KP618033988, +0.618033988749894848204586834365638117720309180);
cannam@167:      DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
cannam@167:      {
cannam@167: 	  INT i;
cannam@167: 	  for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) {
cannam@167: 	       E T3, Tb, Tn, Tu, Tk, Tv, Ta, Ts, Te, Tg, Ti, Tj;
cannam@167: 	       {
cannam@167: 		    E T1, T2, Tl, Tm;
cannam@167: 		    T1 = Cr[0];
cannam@167: 		    T2 = Cr[WS(csr, 5)];
cannam@167: 		    T3 = T1 - T2;
cannam@167: 		    Tb = T1 + T2;
cannam@167: 		    Tl = Ci[WS(csi, 2)];
cannam@167: 		    Tm = Ci[WS(csi, 3)];
cannam@167: 		    Tn = Tl - Tm;
cannam@167: 		    Tu = Tl + Tm;
cannam@167: 	       }
cannam@167: 	       Ti = Ci[WS(csi, 4)];
cannam@167: 	       Tj = Ci[WS(csi, 1)];
cannam@167: 	       Tk = Ti - Tj;
cannam@167: 	       Tv = Ti + Tj;
cannam@167: 	       {
cannam@167: 		    E T6, Tc, T9, Td;
cannam@167: 		    {
cannam@167: 			 E T4, T5, T7, T8;
cannam@167: 			 T4 = Cr[WS(csr, 2)];
cannam@167: 			 T5 = Cr[WS(csr, 3)];
cannam@167: 			 T6 = T4 - T5;
cannam@167: 			 Tc = T4 + T5;
cannam@167: 			 T7 = Cr[WS(csr, 4)];
cannam@167: 			 T8 = Cr[WS(csr, 1)];
cannam@167: 			 T9 = T7 - T8;
cannam@167: 			 Td = T7 + T8;
cannam@167: 		    }
cannam@167: 		    Ta = T6 + T9;
cannam@167: 		    Ts = T6 - T9;
cannam@167: 		    Te = Tc + Td;
cannam@167: 		    Tg = Tc - Td;
cannam@167: 	       }
cannam@167: 	       R1[WS(rs, 2)] = FMA(KP2_000000000, Ta, T3);
cannam@167: 	       R0[0] = FMA(KP2_000000000, Te, Tb);
cannam@167: 	       {
cannam@167: 		    E To, Tq, Th, Tp, Tf;
cannam@167: 		    To = FNMS(KP618033988, Tn, Tk);
cannam@167: 		    Tq = FMA(KP618033988, Tk, Tn);
cannam@167: 		    Tf = FNMS(KP500000000, Te, Tb);
cannam@167: 		    Th = FNMS(KP1_118033988, Tg, Tf);
cannam@167: 		    Tp = FMA(KP1_118033988, Tg, Tf);
cannam@167: 		    R0[WS(rs, 4)] = FNMS(KP1_902113032, To, Th);
cannam@167: 		    R0[WS(rs, 2)] = FMA(KP1_902113032, Tq, Tp);
cannam@167: 		    R0[WS(rs, 1)] = FMA(KP1_902113032, To, Th);
cannam@167: 		    R0[WS(rs, 3)] = FNMS(KP1_902113032, Tq, Tp);
cannam@167: 	       }
cannam@167: 	       {
cannam@167: 		    E Tw, Ty, Tt, Tx, Tr;
cannam@167: 		    Tw = FMA(KP618033988, Tv, Tu);
cannam@167: 		    Ty = FNMS(KP618033988, Tu, Tv);
cannam@167: 		    Tr = FNMS(KP500000000, Ta, T3);
cannam@167: 		    Tt = FMA(KP1_118033988, Ts, Tr);
cannam@167: 		    Tx = FNMS(KP1_118033988, Ts, Tr);
cannam@167: 		    R1[0] = FNMS(KP1_902113032, Tw, Tt);
cannam@167: 		    R1[WS(rs, 3)] = FMA(KP1_902113032, Ty, Tx);
cannam@167: 		    R1[WS(rs, 4)] = FMA(KP1_902113032, Tw, Tt);
cannam@167: 		    R1[WS(rs, 1)] = FNMS(KP1_902113032, Ty, Tx);
cannam@167: 	       }
cannam@167: 	  }
cannam@167:      }
cannam@167: }
cannam@167: 
cannam@167: static const kr2c_desc desc = { 10, "r2cb_10", {14, 0, 20, 0}, &GENUS };
cannam@167: 
cannam@167: void X(codelet_r2cb_10) (planner *p) {
cannam@167:      X(kr2c_register) (p, r2cb_10, &desc);
cannam@167: }
cannam@167: 
cannam@167: #else
cannam@167: 
cannam@167: /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -name r2cb_10 -include rdft/scalar/r2cb.h */
cannam@167: 
cannam@167: /*
cannam@167:  * This function contains 34 FP additions, 14 FP multiplications,
cannam@167:  * (or, 26 additions, 6 multiplications, 8 fused multiply/add),
cannam@167:  * 26 stack variables, 5 constants, and 20 memory accesses
cannam@167:  */
cannam@167: #include "rdft/scalar/r2cb.h"
cannam@167: 
cannam@167: static void r2cb_10(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(KP500000000, +0.500000000000000000000000000000000000000000000);
cannam@167:      DK(KP1_902113032, +1.902113032590307144232878666758764286811397268);
cannam@167:      DK(KP1_175570504, +1.175570504584946258337411909278145537195304875);
cannam@167:      DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
cannam@167:      DK(KP1_118033988, +1.118033988749894848204586834365638117720309180);
cannam@167:      {
cannam@167: 	  INT i;
cannam@167: 	  for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) {
cannam@167: 	       E T3, Tb, Tn, Tv, Tk, Tu, Ta, Ts, Te, Tg, Ti, Tj;
cannam@167: 	       {
cannam@167: 		    E T1, T2, Tl, Tm;
cannam@167: 		    T1 = Cr[0];
cannam@167: 		    T2 = Cr[WS(csr, 5)];
cannam@167: 		    T3 = T1 - T2;
cannam@167: 		    Tb = T1 + T2;
cannam@167: 		    Tl = Ci[WS(csi, 4)];
cannam@167: 		    Tm = Ci[WS(csi, 1)];
cannam@167: 		    Tn = Tl - Tm;
cannam@167: 		    Tv = Tl + Tm;
cannam@167: 	       }
cannam@167: 	       Ti = Ci[WS(csi, 2)];
cannam@167: 	       Tj = Ci[WS(csi, 3)];
cannam@167: 	       Tk = Ti - Tj;
cannam@167: 	       Tu = Ti + Tj;
cannam@167: 	       {
cannam@167: 		    E T6, Tc, T9, Td;
cannam@167: 		    {
cannam@167: 			 E T4, T5, T7, T8;
cannam@167: 			 T4 = Cr[WS(csr, 2)];
cannam@167: 			 T5 = Cr[WS(csr, 3)];
cannam@167: 			 T6 = T4 - T5;
cannam@167: 			 Tc = T4 + T5;
cannam@167: 			 T7 = Cr[WS(csr, 4)];
cannam@167: 			 T8 = Cr[WS(csr, 1)];
cannam@167: 			 T9 = T7 - T8;
cannam@167: 			 Td = T7 + T8;
cannam@167: 		    }
cannam@167: 		    Ta = T6 + T9;
cannam@167: 		    Ts = KP1_118033988 * (T6 - T9);
cannam@167: 		    Te = Tc + Td;
cannam@167: 		    Tg = KP1_118033988 * (Tc - Td);
cannam@167: 	       }
cannam@167: 	       R1[WS(rs, 2)] = FMA(KP2_000000000, Ta, T3);
cannam@167: 	       R0[0] = FMA(KP2_000000000, Te, Tb);
cannam@167: 	       {
cannam@167: 		    E To, Tq, Th, Tp, Tf;
cannam@167: 		    To = FNMS(KP1_902113032, Tn, KP1_175570504 * Tk);
cannam@167: 		    Tq = FMA(KP1_902113032, Tk, KP1_175570504 * Tn);
cannam@167: 		    Tf = FNMS(KP500000000, Te, Tb);
cannam@167: 		    Th = Tf - Tg;
cannam@167: 		    Tp = Tg + Tf;
cannam@167: 		    R0[WS(rs, 1)] = Th - To;
cannam@167: 		    R0[WS(rs, 2)] = Tp + Tq;
cannam@167: 		    R0[WS(rs, 4)] = Th + To;
cannam@167: 		    R0[WS(rs, 3)] = Tp - Tq;
cannam@167: 	       }
cannam@167: 	       {
cannam@167: 		    E Tw, Ty, Tt, Tx, Tr;
cannam@167: 		    Tw = FNMS(KP1_902113032, Tv, KP1_175570504 * Tu);
cannam@167: 		    Ty = FMA(KP1_902113032, Tu, KP1_175570504 * Tv);
cannam@167: 		    Tr = FNMS(KP500000000, Ta, T3);
cannam@167: 		    Tt = Tr - Ts;
cannam@167: 		    Tx = Ts + Tr;
cannam@167: 		    R1[WS(rs, 3)] = Tt - Tw;
cannam@167: 		    R1[WS(rs, 4)] = Tx + Ty;
cannam@167: 		    R1[WS(rs, 1)] = Tt + Tw;
cannam@167: 		    R1[0] = Tx - Ty;
cannam@167: 	       }
cannam@167: 	  }
cannam@167:      }
cannam@167: }
cannam@167: 
cannam@167: static const kr2c_desc desc = { 10, "r2cb_10", {26, 6, 8, 0}, &GENUS };
cannam@167: 
cannam@167: void X(codelet_r2cb_10) (planner *p) {
cannam@167:      X(kr2c_register) (p, r2cb_10, &desc);
cannam@167: }
cannam@167: 
cannam@167: #endif