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 07:35:42 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 4 -name n1_4 -include n.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 16 FP additions, 0 FP multiplications,
Chris@10:  * (or, 16 additions, 0 multiplications, 0 fused multiply/add),
Chris@10:  * 13 stack variables, 0 constants, and 16 memory accesses
Chris@10:  */
Chris@10: #include "n.h"
Chris@10: 
Chris@10: static void n1_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@10: {
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(16, is), MAKE_VOLATILE_STRIDE(16, os)) {
Chris@10: 	       E T4, Tb, T3, Tf, T9, T5, Tc, Td;
Chris@10: 	       {
Chris@10: 		    E T1, T2, T7, T8;
Chris@10: 		    T1 = ri[0];
Chris@10: 		    T2 = ri[WS(is, 2)];
Chris@10: 		    T7 = ii[0];
Chris@10: 		    T8 = ii[WS(is, 2)];
Chris@10: 		    T4 = ri[WS(is, 1)];
Chris@10: 		    Tb = T1 - T2;
Chris@10: 		    T3 = T1 + T2;
Chris@10: 		    Tf = T7 + T8;
Chris@10: 		    T9 = T7 - T8;
Chris@10: 		    T5 = ri[WS(is, 3)];
Chris@10: 		    Tc = ii[WS(is, 1)];
Chris@10: 		    Td = ii[WS(is, 3)];
Chris@10: 	       }
Chris@10: 	       {
Chris@10: 		    E T6, Ta, Te, Tg;
Chris@10: 		    T6 = T4 + T5;
Chris@10: 		    Ta = T4 - T5;
Chris@10: 		    Te = Tc - Td;
Chris@10: 		    Tg = Tc + Td;
Chris@10: 		    io[WS(os, 3)] = Ta + T9;
Chris@10: 		    io[WS(os, 1)] = T9 - Ta;
Chris@10: 		    ro[0] = T3 + T6;
Chris@10: 		    ro[WS(os, 2)] = T3 - T6;
Chris@10: 		    io[0] = Tf + Tg;
Chris@10: 		    io[WS(os, 2)] = Tf - Tg;
Chris@10: 		    ro[WS(os, 3)] = Tb - Te;
Chris@10: 		    ro[WS(os, 1)] = Tb + Te;
Chris@10: 	       }
Chris@10: 	  }
Chris@10:      }
Chris@10: }
Chris@10: 
Chris@10: static const kdft_desc desc = { 4, "n1_4", {16, 0, 0, 0}, &GENUS, 0, 0, 0, 0 };
Chris@10: 
Chris@10: void X(codelet_n1_4) (planner *p) {
Chris@10:      X(kdft_register) (p, n1_4, &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 4 -name n1_4 -include n.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 16 FP additions, 0 FP multiplications,
Chris@10:  * (or, 16 additions, 0 multiplications, 0 fused multiply/add),
Chris@10:  * 13 stack variables, 0 constants, and 16 memory accesses
Chris@10:  */
Chris@10: #include "n.h"
Chris@10: 
Chris@10: static void n1_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
Chris@10: {
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(16, is), MAKE_VOLATILE_STRIDE(16, os)) {
Chris@10: 	       E T3, Tb, T9, Tf, T6, Ta, Te, Tg;
Chris@10: 	       {
Chris@10: 		    E T1, T2, T7, T8;
Chris@10: 		    T1 = ri[0];
Chris@10: 		    T2 = ri[WS(is, 2)];
Chris@10: 		    T3 = T1 + T2;
Chris@10: 		    Tb = T1 - T2;
Chris@10: 		    T7 = ii[0];
Chris@10: 		    T8 = ii[WS(is, 2)];
Chris@10: 		    T9 = T7 - T8;
Chris@10: 		    Tf = T7 + T8;
Chris@10: 	       }
Chris@10: 	       {
Chris@10: 		    E T4, T5, Tc, Td;
Chris@10: 		    T4 = ri[WS(is, 1)];
Chris@10: 		    T5 = ri[WS(is, 3)];
Chris@10: 		    T6 = T4 + T5;
Chris@10: 		    Ta = T4 - T5;
Chris@10: 		    Tc = ii[WS(is, 1)];
Chris@10: 		    Td = ii[WS(is, 3)];
Chris@10: 		    Te = Tc - Td;
Chris@10: 		    Tg = Tc + Td;
Chris@10: 	       }
Chris@10: 	       ro[WS(os, 2)] = T3 - T6;
Chris@10: 	       io[WS(os, 2)] = Tf - Tg;
Chris@10: 	       ro[0] = T3 + T6;
Chris@10: 	       io[0] = Tf + Tg;
Chris@10: 	       io[WS(os, 1)] = T9 - Ta;
Chris@10: 	       ro[WS(os, 1)] = Tb + Te;
Chris@10: 	       io[WS(os, 3)] = Ta + T9;
Chris@10: 	       ro[WS(os, 3)] = Tb - Te;
Chris@10: 	  }
Chris@10:      }
Chris@10: }
Chris@10: 
Chris@10: static const kdft_desc desc = { 4, "n1_4", {16, 0, 0, 0}, &GENUS, 0, 0, 0, 0 };
Chris@10: 
Chris@10: void X(codelet_n1_4) (planner *p) {
Chris@10:      X(kdft_register) (p, n1_4, &desc);
Chris@10: }
Chris@10: 
Chris@10: #endif				/* HAVE_FMA */