annotate src/fftw-3.3.8/dft/scalar/codelets/n1_3.c @ 83:ae30d91d2ffe

Replace these with versions built using an older toolset (so as to avoid ABI compatibilities when linking on Ubuntu 14.04 for packaging purposes)

/*
 * Copyright (c) 2003, 2007-14 Matteo Frigo
 * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
 *
 */

/* This file was automatically generated --- DO NOT EDIT */
/* Generated on Thu May 24 08:04:10 EDT 2018 */

#include "dft/codelet-dft.h"

#if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)

/* Generated by: ../../../genfft/gen_notw.native -fma -compact -variables 4 -pipeline-latency 4 -n 3 -name n1_3 -include dft/scalar/n.h */

/*
 * This function contains 12 FP additions, 6 FP multiplications,
 * (or, 6 additions, 0 multiplications, 6 fused multiply/add),
 * 15 stack variables, 2 constants, and 12 memory accesses
 */
#include "dft/scalar/n.h"

static void n1_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DK(KP500000000, +0.500000000000000000000000000000000000000000000);
     {
	  INT i;
	  for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) {
	       E T1, T9, T4, Tc, T8, Ta, T5, Tb;
	       T1 = ri[0];
	       T9 = ii[0];
	       {
		    E T2, T3, T6, T7;
		    T2 = ri[WS(is, 1)];
		    T3 = ri[WS(is, 2)];
		    T4 = T2 + T3;
		    Tc = T3 - T2;
		    T6 = ii[WS(is, 1)];
		    T7 = ii[WS(is, 2)];
		    T8 = T6 - T7;
		    Ta = T6 + T7;
	       }
	       ro[0] = T1 + T4;
	       io[0] = T9 + Ta;
	       T5 = FNMS(KP500000000, T4, T1);
	       ro[WS(os, 2)] = FNMS(KP866025403, T8, T5);
	       ro[WS(os, 1)] = FMA(KP866025403, T8, T5);
	       Tb = FNMS(KP500000000, Ta, T9);
	       io[WS(os, 1)] = FMA(KP866025403, Tc, Tb);
	       io[WS(os, 2)] = FNMS(KP866025403, Tc, Tb);
	  }
     }
}

static const kdft_desc desc = { 3, "n1_3", {6, 0, 6, 0}, &GENUS, 0, 0, 0, 0 };

void X(codelet_n1_3) (planner *p) {
     X(kdft_register) (p, n1_3, &desc);
}

#else

/* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 3 -name n1_3 -include dft/scalar/n.h */

/*
 * This function contains 12 FP additions, 4 FP multiplications,
 * (or, 10 additions, 2 multiplications, 2 fused multiply/add),
 * 15 stack variables, 2 constants, and 12 memory accesses
 */
#include "dft/scalar/n.h"

static void n1_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DK(KP500000000, +0.500000000000000000000000000000000000000000000);
     DK(KP866025403, +0.866025403784438646763723170752936183471402627);
     {
	  INT i;
	  for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) {
	       E T1, Ta, T4, T9, T8, Tb, T5, Tc;
	       T1 = ri[0];
	       Ta = ii[0];
	       {
		    E T2, T3, T6, T7;
		    T2 = ri[WS(is, 1)];
		    T3 = ri[WS(is, 2)];
		    T4 = T2 + T3;
		    T9 = KP866025403 * (T3 - T2);
		    T6 = ii[WS(is, 1)];
		    T7 = ii[WS(is, 2)];
		    T8 = KP866025403 * (T6 - T7);
		    Tb = T6 + T7;
	       }
	       ro[0] = T1 + T4;
	       io[0] = Ta + Tb;
	       T5 = FNMS(KP500000000, T4, T1);
	       ro[WS(os, 2)] = T5 - T8;
	       ro[WS(os, 1)] = T5 + T8;
	       Tc = FNMS(KP500000000, Tb, Ta);
	       io[WS(os, 1)] = T9 + Tc;
	       io[WS(os, 2)] = Tc - T9;
	  }
     }
}

static const kdft_desc desc = { 3, "n1_3", {10, 2, 2, 0}, &GENUS, 0, 0, 0, 0 };

void X(codelet_n1_3) (planner *p) {
     X(kdft_register) (p, n1_3, &desc);
}

#endif