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 07:39:45 EST 2012 */
Chris@10: 
Chris@10: #include "codelet-rdft.h"
Chris@10: 
Chris@10: #ifdef HAVE_FMA
Chris@10: 
Chris@10: /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -name r2cf_6 -include r2cf.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 14 FP additions, 4 FP multiplications,
Chris@10:  * (or, 12 additions, 2 multiplications, 2 fused multiply/add),
Chris@10:  * 13 stack variables, 2 constants, and 12 memory accesses
Chris@10:  */
Chris@10: #include "r2cf.h"
Chris@10: 
Chris@10: static void r2cf_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
Chris@10: {
Chris@10:      DK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@10:      DK(KP866025403, +0.866025403784438646763723170752936183471402627);
Chris@10:      {
Chris@10: 	  INT i;
Chris@10: 	  for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) {
Chris@10: 	       E T4, Td, T3, Tc, T9, T5;
Chris@10: 	       {
Chris@10: 		    E T1, T2, T7, T8;
Chris@10: 		    T1 = R0[0];
Chris@10: 		    T2 = R1[WS(rs, 1)];
Chris@10: 		    T7 = R0[WS(rs, 2)];
Chris@10: 		    T8 = R1[0];
Chris@10: 		    T4 = R0[WS(rs, 1)];
Chris@10: 		    Td = T1 + T2;
Chris@10: 		    T3 = T1 - T2;
Chris@10: 		    Tc = T7 + T8;
Chris@10: 		    T9 = T7 - T8;
Chris@10: 		    T5 = R1[WS(rs, 2)];
Chris@10: 	       }
Chris@10: 	       {
Chris@10: 		    E T6, Tb, Te, Ta;
Chris@10: 		    T6 = T4 - T5;
Chris@10: 		    Tb = T4 + T5;
Chris@10: 		    Te = Tb + Tc;
Chris@10: 		    Ci[WS(csi, 2)] = KP866025403 * (Tb - Tc);
Chris@10: 		    Ta = T6 + T9;
Chris@10: 		    Ci[WS(csi, 1)] = KP866025403 * (T9 - T6);
Chris@10: 		    Cr[0] = Td + Te;
Chris@10: 		    Cr[WS(csr, 2)] = FNMS(KP500000000, Te, Td);
Chris@10: 		    Cr[WS(csr, 3)] = T3 + Ta;
Chris@10: 		    Cr[WS(csr, 1)] = FNMS(KP500000000, Ta, T3);
Chris@10: 	       }
Chris@10: 	  }
Chris@10:      }
Chris@10: }
Chris@10: 
Chris@10: static const kr2c_desc desc = { 6, "r2cf_6", {12, 2, 2, 0}, &GENUS };
Chris@10: 
Chris@10: void X(codelet_r2cf_6) (planner *p) {
Chris@10:      X(kr2c_register) (p, r2cf_6, &desc);
Chris@10: }
Chris@10: 
Chris@10: #else				/* HAVE_FMA */
Chris@10: 
Chris@10: /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 6 -name r2cf_6 -include r2cf.h */
Chris@10: 
Chris@10: /*
Chris@10:  * This function contains 14 FP additions, 4 FP multiplications,
Chris@10:  * (or, 12 additions, 2 multiplications, 2 fused multiply/add),
Chris@10:  * 17 stack variables, 2 constants, and 12 memory accesses
Chris@10:  */
Chris@10: #include "r2cf.h"
Chris@10: 
Chris@10: static void r2cf_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
Chris@10: {
Chris@10:      DK(KP500000000, +0.500000000000000000000000000000000000000000000);
Chris@10:      DK(KP866025403, +0.866025403784438646763723170752936183471402627);
Chris@10:      {
Chris@10: 	  INT i;
Chris@10: 	  for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) {
Chris@10: 	       E T3, Td, T9, Tc, T6, Tb, T1, T2, Ta, Te;
Chris@10: 	       T1 = R0[0];
Chris@10: 	       T2 = R1[WS(rs, 1)];
Chris@10: 	       T3 = T1 - T2;
Chris@10: 	       Td = T1 + T2;
Chris@10: 	       {
Chris@10: 		    E T7, T8, T4, T5;
Chris@10: 		    T7 = R0[WS(rs, 2)];
Chris@10: 		    T8 = R1[0];
Chris@10: 		    T9 = T7 - T8;
Chris@10: 		    Tc = T7 + T8;
Chris@10: 		    T4 = R0[WS(rs, 1)];
Chris@10: 		    T5 = R1[WS(rs, 2)];
Chris@10: 		    T6 = T4 - T5;
Chris@10: 		    Tb = T4 + T5;
Chris@10: 	       }
Chris@10: 	       Ci[WS(csi, 1)] = KP866025403 * (T9 - T6);
Chris@10: 	       Ta = T6 + T9;
Chris@10: 	       Cr[WS(csr, 1)] = FNMS(KP500000000, Ta, T3);
Chris@10: 	       Cr[WS(csr, 3)] = T3 + Ta;
Chris@10: 	       Ci[WS(csi, 2)] = KP866025403 * (Tb - Tc);
Chris@10: 	       Te = Tb + Tc;
Chris@10: 	       Cr[WS(csr, 2)] = FNMS(KP500000000, Te, Td);
Chris@10: 	       Cr[0] = Td + Te;
Chris@10: 	  }
Chris@10:      }
Chris@10: }
Chris@10: 
Chris@10: static const kr2c_desc desc = { 6, "r2cf_6", {12, 2, 2, 0}, &GENUS };
Chris@10: 
Chris@10: void X(codelet_r2cf_6) (planner *p) {
Chris@10:      X(kr2c_register) (p, r2cf_6, &desc);
Chris@10: }
Chris@10: 
Chris@10: #endif				/* HAVE_FMA */