t1fv_5.c

快速fft变换

/*
 * Copyright (c) 2003, 2007-8 Matteo Frigo
 * Copyright (c) 2003, 2007-8 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 *
 */

/* This file was automatically generated --- DO NOT EDIT */
/* Generated on Sat Nov 15 20:47:31 EST 2008 */

#include "codelet-dft.h"

#ifdef HAVE_FMA

/* Generated by: ../../../genfft/gen_twiddle_c -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1fv_5 -include t1f.h */

/*
 * This function contains 20 FP additions, 19 FP multiplications,
 * (or, 11 additions, 10 multiplications, 9 fused multiply/add),
 * 26 stack variables, 4 constants, and 10 memory accesses
 */
#include "t1f.h"

static void t1fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     INT m;
     R *x;
     x = ri;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(rs)) {
	  V T1, T2, T9, T4, T7;
	  T1 = LD(&(x[0]), ms, &(x[0]));
	  T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
	  T9 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
	  T4 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
	  T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
	  {
	       V T3, Ta, T5, T8;
	       T3 = BYTWJ(&(W[0]), T2);
	       Ta = BYTWJ(&(W[TWVL * 4]), T9);
	       T5 = BYTWJ(&(W[TWVL * 6]), T4);
	       T8 = BYTWJ(&(W[TWVL * 2]), T7);
	       {
		    V T6, Tg, Tb, Th;
		    T6 = VADD(T3, T5);
		    Tg = VSUB(T3, T5);
		    Tb = VADD(T8, Ta);
		    Th = VSUB(T8, Ta);
		    {
			 V Te, Tc, Tk, Ti, Td, Tj, Tf;
			 Te = VSUB(T6, Tb);
			 Tc = VADD(T6, Tb);
			 Tk = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tg, Th));
			 Ti = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Th, Tg));
			 Td = VFNMS(LDK(KP250000000), Tc, T1);
			 ST(&(x[0]), VADD(T1, Tc), ms, &(x[0]));
			 Tj = VFNMS(LDK(KP559016994), Te, Td);
			 Tf = VFMA(LDK(KP559016994), Te, Td);
			 ST(&(x[WS(rs, 2)]), VFMAI(Tk, Tj), ms, &(x[0]));
			 ST(&(x[WS(rs, 3)]), VFNMSI(Tk, Tj), ms, &(x[WS(rs, 1)]));
			 ST(&(x[WS(rs, 4)]), VFMAI(Ti, Tf), ms, &(x[0]));
			 ST(&(x[WS(rs, 1)]), VFNMSI(Ti, Tf), ms, &(x[WS(rs, 1)]));
		    }
	       }
	  }
     }
}

static const tw_instr twinstr[] = {
     VTW(0, 1),
     VTW(0, 2),
     VTW(0, 3),
     VTW(0, 4),
     {TW_NEXT, VL, 0}
};

static const ct_desc desc = { 5, "t1fv_5", twinstr, &GENUS, {11, 10, 9, 0}, 0, 0, 0 };

void X(codelet_t1fv_5) (planner *p) {
     X(kdft_dit_register) (p, t1fv_5, &desc);
}
#else				/* HAVE_FMA */

/* Generated by: ../../../genfft/gen_twiddle_c -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1fv_5 -include t1f.h */

/*
 * This function contains 20 FP additions, 14 FP multiplications,
 * (or, 17 additions, 11 multiplications, 3 fused multiply/add),
 * 20 stack variables, 4 constants, and 10 memory accesses
 */
#include "t1f.h"

static void t1fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     INT m;
     R *x;
     x = ri;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(rs)) {
	  V Tc, Tg, Th, T5, Ta, Td;
	  Tc = LD(&(x[0]), ms, &(x[0]));
	  {
	       V T2, T9, T4, T7;
	       {
		    V T1, T8, T3, T6;
		    T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
		    T2 = BYTWJ(&(W[0]), T1);
		    T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
		    T9 = BYTWJ(&(W[TWVL * 4]), T8);
		    T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
		    T4 = BYTWJ(&(W[TWVL * 6]), T3);
		    T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
		    T7 = BYTWJ(&(W[TWVL * 2]), T6);
	       }
	       Tg = VSUB(T2, T4);
	       Th = VSUB(T7, T9);
	       T5 = VADD(T2, T4);
	       Ta = VADD(T7, T9);
	       Td = VADD(T5, Ta);
	  }
	  ST(&(x[0]), VADD(Tc, Td), ms, &(x[0]));
	  {
	       V Ti, Tj, Tf, Tk, Tb, Te;
	       Ti = VBYI(VFMA(LDK(KP951056516), Tg, VMUL(LDK(KP587785252), Th)));
	       Tj = VBYI(VFNMS(LDK(KP587785252), Tg, VMUL(LDK(KP951056516), Th)));
	       Tb = VMUL(LDK(KP559016994), VSUB(T5, Ta));
	       Te = VFNMS(LDK(KP250000000), Td, Tc);
	       Tf = VADD(Tb, Te);
	       Tk = VSUB(Te, Tb);
	       ST(&(x[WS(rs, 1)]), VSUB(Tf, Ti), ms, &(x[WS(rs, 1)]));
	       ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)]));
	       ST(&(x[WS(rs, 4)]), VADD(Ti, Tf), ms, &(x[0]));
	       ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0]));
	  }
     }
}

static const tw_instr twinstr[] = {
     VTW(0, 1),
     VTW(0, 2),
     VTW(0, 3),
     VTW(0, 4),
     {TW_NEXT, VL, 0}
};

static const ct_desc desc = { 5, "t1fv_5", twinstr, &GENUS, {17, 11, 3, 0}, 0, 0, 0 };

void X(codelet_t1fv_5) (planner *p) {
     X(kdft_dit_register) (p, t1fv_5, &desc);
}
#endif				/* HAVE_FMA */