hb_12.c

最新的FFT程序

			 }
		    }
	       }
	  }
     }
     return W;
}

static const tw_instr twinstr[] = {
     {TW_FULL, 0, 12},
     {TW_NEXT, 1, 0}
};

static const hc2hc_desc desc = { 12, "hb_12", twinstr, &GENUS, {72, 22, 46, 0}, 0, 0, 0 };

void X(codelet_hb_12) (planner *p) {
     X(khc2hc_register) (p, hb_12, &desc);
}
#else				/* HAVE_FMA */

/* Generated by: ../../../genfft/gen_hc2hc -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -dif -name hb_12 -include hb.h */

/*
 * This function contains 118 FP additions, 60 FP multiplications,
 * (or, 88 additions, 30 multiplications, 30 fused multiply/add),
 * 39 stack variables, and 48 memory accesses
 */
/*
 * Generator Id's : 
 * $Id: algsimp.ml,v 1.8 2006-01-05 03:04:27 stevenj Exp $
 * $Id: fft.ml,v 1.4 2006-01-05 03:04:27 stevenj Exp $
 * $Id: gen_hc2hc.ml,v 1.15 2006-01-05 03:04:27 stevenj Exp $
 */

#include "hb.h"

static const R *hb_12(R *rio, R *iio, const R *W, stride ios, INT m, INT dist)
{
     DK(KP500000000, +0.500000000000000000000000000000000000000000000);
     DK(KP866025403, +0.866025403784438646763723170752936183471402627);
     INT i;
     for (i = m - 2; i > 0; i = i - 2, rio = rio + dist, iio = iio - dist, W = W + 22, MAKE_VOLATILE_STRIDE(ios)) {
	  E T5, Tt, T12, T1M, T1i, T1U, Tl, TM, T1c, T1Y, T1s, T1Q, Ta, Ty, T15;
	  E T1N, T1l, T1V, Tg, TH, T19, T1X, T1p, T1P;
	  {
	       E T1, Tp, T4, T1g, Ts, T11, T10, T1h;
	       T1 = rio[0];
	       Tp = iio[0];
	       {
		    E T2, T3, Tq, Tr;
		    T2 = rio[WS(ios, 4)];
		    T3 = iio[-WS(ios, 8)];
		    T4 = T2 + T3;
		    T1g = KP866025403 * (T2 - T3);
		    Tq = rio[WS(ios, 8)];
		    Tr = iio[-WS(ios, 4)];
		    Ts = Tq - Tr;
		    T11 = KP866025403 * (Tq + Tr);
	       }
	       T5 = T1 + T4;
	       Tt = Tp - Ts;
	       T10 = FNMS(KP500000000, T4, T1);
	       T12 = T10 - T11;
	       T1M = T10 + T11;
	       T1h = FMA(KP500000000, Ts, Tp);
	       T1i = T1g + T1h;
	       T1U = T1h - T1g;
	  }
	  {
	       E Th, TL, Tk, T1a, TK, T1r, T1b, T1q;
	       Th = iio[-WS(ios, 9)];
	       TL = rio[WS(ios, 9)];
	       {
		    E Ti, Tj, TI, TJ;
		    Ti = rio[WS(ios, 1)];
		    Tj = rio[WS(ios, 5)];
		    Tk = Ti + Tj;
		    T1a = KP866025403 * (Ti - Tj);
		    TI = iio[-WS(ios, 5)];
		    TJ = iio[-WS(ios, 1)];
		    TK = TI + TJ;
		    T1r = KP866025403 * (TI - TJ);
	       }
	       Tl = Th + Tk;
	       TM = TK - TL;
	       T1b = FMA(KP500000000, TK, TL);
	       T1c = T1a - T1b;
	       T1Y = T1a + T1b;
	       T1q = FNMS(KP500000000, Tk, Th);
	       T1s = T1q + T1r;
	       T1Q = T1q - T1r;
	  }
	  {
	       E T6, Tx, T9, T1j, Tw, T14, T13, T1k;
	       T6 = iio[-WS(ios, 6)];
	       Tx = rio[WS(ios, 6)];
	       {
		    E T7, T8, Tu, Tv;
		    T7 = iio[-WS(ios, 10)];
		    T8 = rio[WS(ios, 2)];
		    T9 = T7 + T8;
		    T1j = KP866025403 * (T7 - T8);
		    Tu = rio[WS(ios, 10)];
		    Tv = iio[-WS(ios, 2)];
		    Tw = Tu - Tv;
		    T14 = KP866025403 * (Tu + Tv);
	       }
	       Ta = T6 + T9;
	       Ty = Tw + Tx;
	       T13 = FNMS(KP500000000, T9, T6);
	       T15 = T13 + T14;
	       T1N = T13 - T14;
	       T1k = FMS(KP500000000, Tw, Tx);
	       T1l = T1j + T1k;
	       T1V = T1k - T1j;
	  }
	  {
	       E Tc, TD, Tf, T17, TG, T1o, T18, T1n;
	       Tc = rio[WS(ios, 3)];
	       TD = iio[-WS(ios, 3)];
	       {
		    E Td, Te, TE, TF;
		    Td = iio[-WS(ios, 7)];
		    Te = iio[-WS(ios, 11)];
		    Tf = Td + Te;
		    T17 = KP866025403 * (Td - Te);
		    TE = rio[WS(ios, 7)];
		    TF = rio[WS(ios, 11)];
		    TG = TE + TF;
		    T1o = KP866025403 * (TE - TF);
	       }
	       Tg = Tc + Tf;
	       TH = TD - TG;
	       T18 = FMA(KP500000000, TG, TD);
	       T19 = T17 + T18;
	       T1X = T18 - T17;
	       T1n = FNMS(KP500000000, Tf, Tc);
	       T1p = T1n + T1o;
	       T1P = T1n - T1o;
	  }
	  {
	       E Tb, Tm, TU, TW, TX, TY, TT, TV;
	       Tb = T5 + Ta;
	       Tm = Tg + Tl;
	       TU = Tb - Tm;
	       TW = Tt - Ty;
	       TX = TH + TM;
	       TY = TW - TX;
	       rio[0] = Tb + Tm;
	       iio[-WS(ios, 11)] = TW + TX;
	       TT = W[10];
	       TV = W[11];
	       rio[WS(ios, 6)] = FNMS(TV, TY, TT * TU);
	       iio[-WS(ios, 5)] = FMA(TV, TU, TT * TY);
	  }
	  {
	       E T28, T2g, T2c, T2e;
	       {
		    E T26, T27, T2a, T2b;
		    T26 = T1M - T1N;
		    T27 = T1X + T1Y;
		    T28 = T26 - T27;
		    T2g = T26 + T27;
		    T2a = T1U - T1V;
		    T2b = T1P - T1Q;
		    T2c = T2a + T2b;
		    T2e = T2a - T2b;
	       }
	       {
		    E T25, T29, T2d, T2f;
		    T25 = W[8];
		    T29 = W[9];
		    rio[WS(ios, 5)] = FNMS(T29, T2c, T25 * T28);
		    iio[-WS(ios, 6)] = FMA(T25, T2c, T29 * T28);
		    T2d = W[20];
		    T2f = W[21];
		    iio[0] = FMA(T2d, T2e, T2f * T2g);
		    rio[WS(ios, 11)] = FNMS(T2f, T2e, T2d * T2g);
	       }
	  }
	  {
	       E TA, TS, TO, TQ;
	       {
		    E To, Tz, TC, TN;
		    To = Tg - Tl;
		    Tz = Tt + Ty;
		    TA = To + Tz;
		    TS = Tz - To;
		    TC = T5 - Ta;
		    TN = TH - TM;
		    TO = TC - TN;
		    TQ = TC + TN;
	       }
	       {
		    E Tn, TB, TP, TR;
		    Tn = W[16];
		    TB = W[17];
		    iio[-WS(ios, 2)] = FMA(Tn, TA, TB * TO);
		    rio[WS(ios, 9)] = FNMS(TB, TA, Tn * TO);
		    TP = W[4];
		    TR = W[5];
		    rio[WS(ios, 3)] = FNMS(TR, TS, TP * TQ);
		    iio[-WS(ios, 8)] = FMA(TP, TS, TR * TQ);
	       }
	  }
	  {
	       E T1S, T22, T20, T24;
	       {
		    E T1O, T1R, T1W, T1Z;
		    T1O = T1M + T1N;
		    T1R = T1P + T1Q;
		    T1S = T1O - T1R;
		    T22 = T1O + T1R;
		    T1W = T1U + T1V;
		    T1Z = T1X - T1Y;
		    T20 = T1W - T1Z;
		    T24 = T1W + T1Z;
	       }
	       {
		    E T1L, T1T, T21, T23;
		    T1L = W[2];
		    T1T = W[3];
		    rio[WS(ios, 2)] = FNMS(T1T, T20, T1L * T1S);
		    iio[-WS(ios, 9)] = FMA(T1T, T1S, T1L * T20);
		    T21 = W[14];
		    T23 = W[15];
		    rio[WS(ios, 8)] = FNMS(T23, T24, T21 * T22);
		    iio[-WS(ios, 3)] = FMA(T23, T22, T21 * T24);
	       }
	  }
	  {
	       E T1C, T1I, T1G, T1K;
	       {
		    E T1A, T1B, T1E, T1F;
		    T1A = T12 + T15;
		    T1B = T1p + T1s;
		    T1C = T1A - T1B;
		    T1I = T1A + T1B;
		    T1E = T1i + T1l;
		    T1F = T19 + T1c;
		    T1G = T1E - T1F;
		    T1K = T1E + T1F;
	       }
	       {
		    E T1z, T1D, T1H, T1J;
		    T1z = W[18];
		    T1D = W[19];
		    rio[WS(ios, 10)] = FNMS(T1D, T1G, T1z * T1C);
		    iio[-WS(ios, 1)] = FMA(T1D, T1C, T1z * T1G);
		    T1H = W[6];
		    T1J = W[7];
		    rio[WS(ios, 4)] = FNMS(T1J, T1K, T1H * T1I);
		    iio[-WS(ios, 7)] = FMA(T1J, T1I, T1H * T1K);
	       }
	  }
	  {
	       E T1e, T1y, T1u, T1w;
	       {
		    E T16, T1d, T1m, T1t;
		    T16 = T12 - T15;
		    T1d = T19 - T1c;
		    T1e = T16 - T1d;
		    T1y = T16 + T1d;
		    T1m = T1i - T1l;
		    T1t = T1p - T1s;
		    T1u = T1m + T1t;
		    T1w = T1m - T1t;
	       }
	       {
		    E TZ, T1f, T1v, T1x;
		    TZ = W[0];
		    T1f = W[1];
		    rio[WS(ios, 1)] = FNMS(T1f, T1u, TZ * T1e);
		    iio[-WS(ios, 10)] = FMA(TZ, T1u, T1f * T1e);
		    T1v = W[12];
		    T1x = W[13];
		    iio[-WS(ios, 4)] = FMA(T1v, T1w, T1x * T1y);
		    rio[WS(ios, 7)] = FNMS(T1x, T1w, T1v * T1y);
	       }
	  }
     }
     return W;
}

static const tw_instr twinstr[] = {
     {TW_FULL, 0, 12},
     {TW_NEXT, 1, 0}
};

static const hc2hc_desc desc = { 12, "hb_12", twinstr, &GENUS, {88, 30, 30, 0}, 0, 0, 0 };

void X(codelet_hb_12) (planner *p) {
     X(khc2hc_register) (p, hb_12, &desc);
}
#endif				/* HAVE_FMA */