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 08:05:19 EDT 2018 */
cannam@167: 
cannam@167: #include "dft/codelet-dft.h"
cannam@167: 
cannam@167: #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
cannam@167: 
cannam@167: /* Generated by: ../../../genfft/gen_notw.native -fma -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n2sv_4 -with-ostride 1 -include dft/simd/n2s.h -store-multiple 4 */
cannam@167: 
cannam@167: /*
cannam@167:  * This function contains 16 FP additions, 0 FP multiplications,
cannam@167:  * (or, 16 additions, 0 multiplications, 0 fused multiply/add),
cannam@167:  * 17 stack variables, 0 constants, and 18 memory accesses
cannam@167:  */
cannam@167: #include "dft/simd/n2s.h"
cannam@167: 
cannam@167: static void n2sv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
cannam@167: {
cannam@167:      {
cannam@167: 	  INT i;
cannam@167: 	  for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) {
cannam@167: 	       V T3, Tb, T9, Tf, T6, Ta, Te, Tg;
cannam@167: 	       {
cannam@167: 		    V T1, T2, T7, T8;
cannam@167: 		    T1 = LD(&(ri[0]), ivs, &(ri[0]));
cannam@167: 		    T2 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0]));
cannam@167: 		    T3 = VADD(T1, T2);
cannam@167: 		    Tb = VSUB(T1, T2);
cannam@167: 		    T7 = LD(&(ii[0]), ivs, &(ii[0]));
cannam@167: 		    T8 = LD(&(ii[WS(is, 2)]), ivs, &(ii[0]));
cannam@167: 		    T9 = VSUB(T7, T8);
cannam@167: 		    Tf = VADD(T7, T8);
cannam@167: 	       }
cannam@167: 	       {
cannam@167: 		    V T4, T5, Tc, Td;
cannam@167: 		    T4 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)]));
cannam@167: 		    T5 = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)]));
cannam@167: 		    T6 = VADD(T4, T5);
cannam@167: 		    Ta = VSUB(T4, T5);
cannam@167: 		    Tc = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)]));
cannam@167: 		    Td = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)]));
cannam@167: 		    Te = VSUB(Tc, Td);
cannam@167: 		    Tg = VADD(Tc, Td);
cannam@167: 	       }
cannam@167: 	       {
cannam@167: 		    V Th, Ti, Tj, Tk;
cannam@167: 		    Th = VSUB(T3, T6);
cannam@167: 		    STM4(&(ro[2]), Th, ovs, &(ro[0]));
cannam@167: 		    Ti = VSUB(Tf, Tg);
cannam@167: 		    STM4(&(io[2]), Ti, ovs, &(io[0]));
cannam@167: 		    Tj = VADD(T3, T6);
cannam@167: 		    STM4(&(ro[0]), Tj, ovs, &(ro[0]));
cannam@167: 		    Tk = VADD(Tf, Tg);
cannam@167: 		    STM4(&(io[0]), Tk, ovs, &(io[0]));
cannam@167: 		    {
cannam@167: 			 V Tl, Tm, Tn, To;
cannam@167: 			 Tl = VSUB(T9, Ta);
cannam@167: 			 STM4(&(io[1]), Tl, ovs, &(io[1]));
cannam@167: 			 Tm = VADD(Tb, Te);
cannam@167: 			 STM4(&(ro[1]), Tm, ovs, &(ro[1]));
cannam@167: 			 Tn = VADD(Ta, T9);
cannam@167: 			 STM4(&(io[3]), Tn, ovs, &(io[1]));
cannam@167: 			 STN4(&(io[0]), Tk, Tl, Ti, Tn, ovs);
cannam@167: 			 To = VSUB(Tb, Te);
cannam@167: 			 STM4(&(ro[3]), To, ovs, &(ro[1]));
cannam@167: 			 STN4(&(ro[0]), Tj, Tm, Th, To, ovs);
cannam@167: 		    }
cannam@167: 	       }
cannam@167: 	  }
cannam@167:      }
cannam@167:      VLEAVE();
cannam@167: }
cannam@167: 
cannam@167: static const kdft_desc desc = { 4, XSIMD_STRING("n2sv_4"), {16, 0, 0, 0}, &GENUS, 0, 1, 0, 0 };
cannam@167: 
cannam@167: void XSIMD(codelet_n2sv_4) (planner *p) {
cannam@167:      X(kdft_register) (p, n2sv_4, &desc);
cannam@167: }
cannam@167: 
cannam@167: #else
cannam@167: 
cannam@167: /* Generated by: ../../../genfft/gen_notw.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n2sv_4 -with-ostride 1 -include dft/simd/n2s.h -store-multiple 4 */
cannam@167: 
cannam@167: /*
cannam@167:  * This function contains 16 FP additions, 0 FP multiplications,
cannam@167:  * (or, 16 additions, 0 multiplications, 0 fused multiply/add),
cannam@167:  * 17 stack variables, 0 constants, and 18 memory accesses
cannam@167:  */
cannam@167: #include "dft/simd/n2s.h"
cannam@167: 
cannam@167: static void n2sv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
cannam@167: {
cannam@167:      {
cannam@167: 	  INT i;
cannam@167: 	  for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) {
cannam@167: 	       V T3, Tb, T9, Tf, T6, Ta, Te, Tg;
cannam@167: 	       {
cannam@167: 		    V T1, T2, T7, T8;
cannam@167: 		    T1 = LD(&(ri[0]), ivs, &(ri[0]));
cannam@167: 		    T2 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0]));
cannam@167: 		    T3 = VADD(T1, T2);
cannam@167: 		    Tb = VSUB(T1, T2);
cannam@167: 		    T7 = LD(&(ii[0]), ivs, &(ii[0]));
cannam@167: 		    T8 = LD(&(ii[WS(is, 2)]), ivs, &(ii[0]));
cannam@167: 		    T9 = VSUB(T7, T8);
cannam@167: 		    Tf = VADD(T7, T8);
cannam@167: 	       }
cannam@167: 	       {
cannam@167: 		    V T4, T5, Tc, Td;
cannam@167: 		    T4 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)]));
cannam@167: 		    T5 = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)]));
cannam@167: 		    T6 = VADD(T4, T5);
cannam@167: 		    Ta = VSUB(T4, T5);
cannam@167: 		    Tc = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)]));
cannam@167: 		    Td = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)]));
cannam@167: 		    Te = VSUB(Tc, Td);
cannam@167: 		    Tg = VADD(Tc, Td);
cannam@167: 	       }
cannam@167: 	       {
cannam@167: 		    V Th, Ti, Tj, Tk;
cannam@167: 		    Th = VSUB(T3, T6);
cannam@167: 		    STM4(&(ro[2]), Th, ovs, &(ro[0]));
cannam@167: 		    Ti = VSUB(Tf, Tg);
cannam@167: 		    STM4(&(io[2]), Ti, ovs, &(io[0]));
cannam@167: 		    Tj = VADD(T3, T6);
cannam@167: 		    STM4(&(ro[0]), Tj, ovs, &(ro[0]));
cannam@167: 		    Tk = VADD(Tf, Tg);
cannam@167: 		    STM4(&(io[0]), Tk, ovs, &(io[0]));
cannam@167: 		    {
cannam@167: 			 V Tl, Tm, Tn, To;
cannam@167: 			 Tl = VSUB(T9, Ta);
cannam@167: 			 STM4(&(io[1]), Tl, ovs, &(io[1]));
cannam@167: 			 Tm = VADD(Tb, Te);
cannam@167: 			 STM4(&(ro[1]), Tm, ovs, &(ro[1]));
cannam@167: 			 Tn = VADD(Ta, T9);
cannam@167: 			 STM4(&(io[3]), Tn, ovs, &(io[1]));
cannam@167: 			 STN4(&(io[0]), Tk, Tl, Ti, Tn, ovs);
cannam@167: 			 To = VSUB(Tb, Te);
cannam@167: 			 STM4(&(ro[3]), To, ovs, &(ro[1]));
cannam@167: 			 STN4(&(ro[0]), Tj, Tm, Th, To, ovs);
cannam@167: 		    }
cannam@167: 	       }
cannam@167: 	  }
cannam@167:      }
cannam@167:      VLEAVE();
cannam@167: }
cannam@167: 
cannam@167: static const kdft_desc desc = { 4, XSIMD_STRING("n2sv_4"), {16, 0, 0, 0}, &GENUS, 0, 1, 0, 0 };
cannam@167: 
cannam@167: void XSIMD(codelet_n2sv_4) (planner *p) {
cannam@167:      X(kdft_register) (p, n2sv_4, &desc);
cannam@167: }
cannam@167: 
cannam@167: #endif