当前位置: 首页>>代码示例>>C++>>正文


C++ VADD函数代码示例

本文整理汇总了C++中VADD函数的典型用法代码示例。如果您正苦于以下问题:C++ VADD函数的具体用法?C++ VADD怎么用?C++ VADD使用的例子?那么, 这里精选的函数代码示例或许可以为您提供帮助。


在下文中一共展示了VADD函数的15个代码示例,这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞,您的评价将有助于系统推荐出更棒的C++代码示例。

示例1: n2fv_12

static void n2fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     {
	  INT i;
	  const R *xi;
	  R *xo;
	  xi = ri;
	  xo = ro;
	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) {
	       V T5, Ta, TG, TF, TB, Tt, Ti, Tm, TJ, TI, TA, Tp;
	       {
		    V T1, T6, T4, Tr, T9, Ts;
		    T1 = LD(&(xi[0]), ivs, &(xi[0]));
		    T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    {
			 V T2, T3, T7, T8;
			 T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
			 T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
			 T4 = VADD(T2, T3);
			 Tr = VSUB(T3, T2);
			 T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
			 T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
			 T9 = VADD(T7, T8);
			 Ts = VSUB(T8, T7);
		    }
		    T5 = VFNMS(LDK(KP500000000), T4, T1);
		    Ta = VFNMS(LDK(KP500000000), T9, T6);
		    TG = VADD(T6, T9);
		    TF = VADD(T1, T4);
		    TB = VADD(Tr, Ts);
		    Tt = VSUB(Tr, Ts);
	       }
	       {
		    V Tk, Tn, Te, Tl, Th, To;
		    Tk = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
		    Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
		    {
			 V Tc, Td, Tf, Tg;
			 Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
			 Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
			 Te = VSUB(Tc, Td);
			 Tl = VADD(Td, Tc);
			 Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
			 Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
			 Th = VSUB(Tf, Tg);
			 To = VADD(Tf, Tg);
		    }
		    Ti = VADD(Te, Th);
		    Tm = VFNMS(LDK(KP500000000), Tl, Tk);
		    TJ = VADD(Tn, To);
		    TI = VADD(Tk, Tl);
		    TA = VSUB(Te, Th);
		    Tp = VFNMS(LDK(KP500000000), To, Tn);
	       }
	       {
		    V TN, TO, TP, TQ, TT, TU;
		    {
			 V TH, TK, TL, TM;
			 TH = VSUB(TF, TG);
			 TK = VSUB(TI, TJ);
			 TN = VFNMSI(TK, TH);
			 STM2(&(xo[18]), TN, ovs, &(xo[2]));
			 TO = VFMAI(TK, TH);
			 STM2(&(xo[6]), TO, ovs, &(xo[2]));
			 TL = VADD(TF, TG);
			 TM = VADD(TI, TJ);
			 TP = VSUB(TL, TM);
			 STM2(&(xo[12]), TP, ovs, &(xo[0]));
			 TQ = VADD(TL, TM);
			 STM2(&(xo[0]), TQ, ovs, &(xo[0]));
		    }
		    {
			 V Tj, Tv, Tu, Tw, Tb, Tq, TR, TS;
			 Tb = VSUB(T5, Ta);
			 Tj = VFMA(LDK(KP866025403), Ti, Tb);
			 Tv = VFNMS(LDK(KP866025403), Ti, Tb);
			 Tq = VSUB(Tm, Tp);
			 Tu = VFNMS(LDK(KP866025403), Tt, Tq);
			 Tw = VFMA(LDK(KP866025403), Tt, Tq);
			 TR = VFNMSI(Tu, Tj);
			 STM2(&(xo[2]), TR, ovs, &(xo[2]));
			 STN2(&(xo[0]), TQ, TR, ovs);
			 TS = VFMAI(Tw, Tv);
			 STM2(&(xo[14]), TS, ovs, &(xo[2]));
			 STN2(&(xo[12]), TP, TS, ovs);
			 TT = VFMAI(Tu, Tj);
			 STM2(&(xo[22]), TT, ovs, &(xo[2]));
			 TU = VFNMSI(Tw, Tv);
			 STM2(&(xo[10]), TU, ovs, &(xo[2]));
		    }
		    {
			 V TC, TE, Tz, TD, Tx, Ty;
			 TC = VMUL(LDK(KP866025403), VSUB(TA, TB));
			 TE = VMUL(LDK(KP866025403), VADD(TB, TA));
			 Tx = VADD(T5, Ta);
			 Ty = VADD(Tm, Tp);
			 Tz = VSUB(Tx, Ty);
			 TD = VADD(Tx, Ty);
//.........这里部分代码省略.........
开发者ID:DougBurke,项目名称:sherpa,代码行数:101,代码来源:n2fv_12.c

示例2: t2bv_10

static void t2bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     INT m;
     R *x;
     x = ii;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(rs)) {
	  V Tu, TH, Tg, Tl, Tp, TD, TE, TJ, T5, Ta, To, TA, TB, TI, Tr;
	  V Tt, Ts;
	  Tr = LD(&(x[0]), ms, &(x[0]));
	  Ts = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
	  Tt = BYTW(&(W[TWVL * 8]), Ts);
	  Tu = VSUB(Tr, Tt);
	  TH = VADD(Tr, Tt);
	  {
	       V Td, Tk, Tf, Ti;
	       {
		    V Tc, Tj, Te, Th;
		    Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
		    Td = BYTW(&(W[TWVL * 6]), Tc);
		    Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
		    Tk = BYTW(&(W[0]), Tj);
		    Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
		    Tf = BYTW(&(W[TWVL * 16]), Te);
		    Th = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
		    Ti = BYTW(&(W[TWVL * 10]), Th);
	       }
	       Tg = VSUB(Td, Tf);
	       Tl = VSUB(Ti, Tk);
	       Tp = VADD(Tg, Tl);
	       TD = VADD(Td, Tf);
	       TE = VADD(Ti, Tk);
	       TJ = VADD(TD, TE);
	  }
	  {
	       V T2, T9, T4, T7;
	       {
		    V T1, T8, T3, T6;
		    T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
		    T2 = BYTW(&(W[TWVL * 2]), T1);
		    T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
		    T9 = BYTW(&(W[TWVL * 4]), T8);
		    T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
		    T4 = BYTW(&(W[TWVL * 12]), T3);
		    T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
		    T7 = BYTW(&(W[TWVL * 14]), T6);
	       }
	       T5 = VSUB(T2, T4);
	       Ta = VSUB(T7, T9);
	       To = VADD(T5, Ta);
	       TA = VADD(T2, T4);
	       TB = VADD(T7, T9);
	       TI = VADD(TA, TB);
	  }
	  {
	       V Tq, Tv, Tw, Tn, Tz, Tb, Tm, Ty, Tx;
	       Tq = VMUL(LDK(KP559016994), VSUB(To, Tp));
	       Tv = VADD(To, Tp);
	       Tw = VFNMS(LDK(KP250000000), Tv, Tu);
	       Tb = VSUB(T5, Ta);
	       Tm = VSUB(Tg, Tl);
	       Tn = VBYI(VFMA(LDK(KP951056516), Tb, VMUL(LDK(KP587785252), Tm)));
	       Tz = VBYI(VFNMS(LDK(KP951056516), Tm, VMUL(LDK(KP587785252), Tb)));
	       ST(&(x[WS(rs, 5)]), VADD(Tu, Tv), ms, &(x[WS(rs, 1)]));
	       Ty = VSUB(Tw, Tq);
	       ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)]));
	       ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)]));
	       Tx = VADD(Tq, Tw);
	       ST(&(x[WS(rs, 1)]), VADD(Tn, Tx), ms, &(x[WS(rs, 1)]));
	       ST(&(x[WS(rs, 9)]), VSUB(Tx, Tn), ms, &(x[WS(rs, 1)]));
	  }
	  {
	       V TM, TK, TL, TG, TP, TC, TF, TO, TN;
	       TM = VMUL(LDK(KP559016994), VSUB(TI, TJ));
	       TK = VADD(TI, TJ);
	       TL = VFNMS(LDK(KP250000000), TK, TH);
	       TC = VSUB(TA, TB);
	       TF = VSUB(TD, TE);
	       TG = VBYI(VFNMS(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC)));
	       TP = VBYI(VFMA(LDK(KP951056516), TC, VMUL(LDK(KP587785252), TF)));
	       ST(&(x[0]), VADD(TH, TK), ms, &(x[0]));
	       TO = VADD(TM, TL);
	       ST(&(x[WS(rs, 4)]), VSUB(TO, TP), ms, &(x[0]));
	       ST(&(x[WS(rs, 6)]), VADD(TP, TO), ms, &(x[0]));
	       TN = VSUB(TL, TM);
	       ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0]));
	       ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0]));
	  }
     }
}
开发者ID:8cH9azbsFifZ,项目名称:wspr,代码行数:93,代码来源:t2bv_10.c

示例3: DVK

static const R *t1fv_15(R *ri, R *ii, const R *W, stride ios, INT m, INT dist)
{
     DVK(KP216506350, +0.216506350946109661690930792688234045867850657);
     DVK(KP484122918, +0.484122918275927110647408174972799951354115213);
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP509036960, +0.509036960455127183450980863393907648510733164);
     DVK(KP823639103, +0.823639103546331925877420039278190003029660514);
     DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     INT i;
     R *x;
     x = ri;
     for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 28), MAKE_VOLATILE_STRIDE(ios)) {
	  V T1e, T7, TP, T12, T15, Tf, Tn, To, T1b, T1c, T1f, TQ, TR, TS, Tw;
	  V TE, TF, TT, TU, TV;
	  {
	       V T1, T5, T3, T4, T2, T6;
	       T1 = LD(&(x[0]), dist, &(x[0]));
	       T4 = LD(&(x[WS(ios, 10)]), dist, &(x[0]));
	       T5 = BYTWJ(&(W[TWVL * 18]), T4);
	       T2 = LD(&(x[WS(ios, 5)]), dist, &(x[WS(ios, 1)]));
	       T3 = BYTWJ(&(W[TWVL * 8]), T2);
	       T1e = VSUB(T5, T3);
	       T6 = VADD(T3, T5);
	       T7 = VADD(T1, T6);
	       TP = VFNMS(LDK(KP500000000), T6, T1);
	  }
	  {
	       V T9, Tq, Ty, Th, Te, T13, Tv, T10, TD, T11, Tm, T14;
	       {
		    V T8, Tp, Tx, Tg;
		    T8 = LD(&(x[WS(ios, 3)]), dist, &(x[WS(ios, 1)]));
		    T9 = BYTWJ(&(W[TWVL * 4]), T8);
		    Tp = LD(&(x[WS(ios, 6)]), dist, &(x[0]));
		    Tq = BYTWJ(&(W[TWVL * 10]), Tp);
		    Tx = LD(&(x[WS(ios, 9)]), dist, &(x[WS(ios, 1)]));
		    Ty = BYTWJ(&(W[TWVL * 16]), Tx);
		    Tg = LD(&(x[WS(ios, 12)]), dist, &(x[0]));
		    Th = BYTWJ(&(W[TWVL * 22]), Tg);
	       }
	       {
		    V Tb, Td, Ta, Tc;
		    Ta = LD(&(x[WS(ios, 8)]), dist, &(x[0]));
		    Tb = BYTWJ(&(W[TWVL * 14]), Ta);
		    Tc = LD(&(x[WS(ios, 13)]), dist, &(x[WS(ios, 1)]));
		    Td = BYTWJ(&(W[TWVL * 24]), Tc);
		    Te = VADD(Tb, Td);
		    T13 = VSUB(Td, Tb);
	       }
	       {
		    V Ts, Tu, Tr, Tt;
		    Tr = LD(&(x[WS(ios, 11)]), dist, &(x[WS(ios, 1)]));
		    Ts = BYTWJ(&(W[TWVL * 20]), Tr);
		    Tt = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)]));
		    Tu = BYTWJ(&(W[0]), Tt);
		    Tv = VADD(Ts, Tu);
		    T10 = VSUB(Tu, Ts);
	       }
	       {
		    V TA, TC, Tz, TB;
		    Tz = LD(&(x[WS(ios, 14)]), dist, &(x[0]));
		    TA = BYTWJ(&(W[TWVL * 26]), Tz);
		    TB = LD(&(x[WS(ios, 4)]), dist, &(x[0]));
		    TC = BYTWJ(&(W[TWVL * 6]), TB);
		    TD = VADD(TA, TC);
		    T11 = VSUB(TC, TA);
	       }
	       {
		    V Tj, Tl, Ti, Tk;
		    Ti = LD(&(x[WS(ios, 2)]), dist, &(x[0]));
		    Tj = BYTWJ(&(W[TWVL * 2]), Ti);
		    Tk = LD(&(x[WS(ios, 7)]), dist, &(x[WS(ios, 1)]));
		    Tl = BYTWJ(&(W[TWVL * 12]), Tk);
		    Tm = VADD(Tj, Tl);
		    T14 = VSUB(Tl, Tj);
	       }
	       T12 = VSUB(T10, T11);
	       T15 = VSUB(T13, T14);
	       Tf = VADD(T9, Te);
	       Tn = VADD(Th, Tm);
	       To = VADD(Tf, Tn);
	       T1b = VADD(T13, T14);
	       T1c = VADD(T10, T11);
	       T1f = VADD(T1b, T1c);
	       TQ = VFNMS(LDK(KP500000000), Te, T9);
	       TR = VFNMS(LDK(KP500000000), Tm, Th);
	       TS = VADD(TQ, TR);
	       Tw = VADD(Tq, Tv);
	       TE = VADD(Ty, TD);
	       TF = VADD(Tw, TE);
	       TT = VFNMS(LDK(KP500000000), Tv, Tq);
	       TU = VFNMS(LDK(KP500000000), TD, Ty);
	       TV = VADD(TT, TU);
	  }
	  {
	       V TI, TG, TH, TM, TO, TK, TL, TN, TJ;
	       TI = VMUL(LDK(KP559016994), VSUB(To, TF));
//.........这里部分代码省略.........
开发者ID:exic,项目名称:last.fm-dbus,代码行数:101,代码来源:t1fv_15.c

示例4: n2bv_16

static void n2bv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
     DVK(KP414213562, +0.414213562373095048801688724209698078569671875);
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     INT i;
     const R *xi;
     R *xo;
     xi = ii;
     xo = io;
     for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
	  V T7, Tu, TF, TB, T13, TL, TO, TX, TC, Te, TP, Th, TQ, Tk, TW;
	  V T16;
	  {
	       V TH, TU, Tz, Tf, TK, TV, TA, TM, Ta, TN, Td, Tg, Ti, Tj;
	       {
		    V T1, T2, T4, T5, To, Tp, Tr, Ts;
		    T1 = LD(&(xi[0]), ivs, &(xi[0]));
		    T2 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
		    T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
		    T5 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
		    To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
		    Tp = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
		    Tr = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
		    Ts = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    {
			 V T8, TI, Tq, TJ, Tt, T9, Tb, Tc, T3, T6;
			 T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
			 TH = VSUB(T1, T2);
			 T3 = VADD(T1, T2);
			 TU = VSUB(T4, T5);
			 T6 = VADD(T4, T5);
			 TI = VSUB(To, Tp);
			 Tq = VADD(To, Tp);
			 TJ = VSUB(Tr, Ts);
			 Tt = VADD(Tr, Ts);
			 T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
			 Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
			 Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
			 T7 = VSUB(T3, T6);
			 Tz = VADD(T3, T6);
			 Tf = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
			 TK = VADD(TI, TJ);
			 TV = VSUB(TI, TJ);
			 TA = VADD(Tq, Tt);
			 Tu = VSUB(Tq, Tt);
			 TM = VSUB(T8, T9);
			 Ta = VADD(T8, T9);
			 TN = VSUB(Tb, Tc);
			 Td = VADD(Tb, Tc);
			 Tg = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
			 Ti = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
			 Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
		    }
	       }
	       TF = VADD(Tz, TA);
	       TB = VSUB(Tz, TA);
	       T13 = VFNMS(LDK(KP707106781), TK, TH);
	       TL = VFMA(LDK(KP707106781), TK, TH);
	       TO = VFNMS(LDK(KP414213562), TN, TM);
	       TX = VFMA(LDK(KP414213562), TM, TN);
	       TC = VADD(Ta, Td);
	       Te = VSUB(Ta, Td);
	       TP = VSUB(Tf, Tg);
	       Th = VADD(Tf, Tg);
	       TQ = VSUB(Tj, Ti);
	       Tk = VADD(Ti, Tj);
	       TW = VFMA(LDK(KP707106781), TV, TU);
	       T16 = VFNMS(LDK(KP707106781), TV, TU);
	  }
	  {
	       V TY, TR, Tl, TD;
	       TY = VFMA(LDK(KP414213562), TP, TQ);
	       TR = VFNMS(LDK(KP414213562), TQ, TP);
	       Tl = VSUB(Th, Tk);
	       TD = VADD(Th, Tk);
	       {
		    V TS, T17, TZ, T14;
		    TS = VADD(TO, TR);
		    T17 = VSUB(TO, TR);
		    TZ = VSUB(TX, TY);
		    T14 = VADD(TX, TY);
		    {
			 V TE, TG, Tm, Tv;
			 TE = VSUB(TC, TD);
			 TG = VADD(TC, TD);
			 Tm = VADD(Te, Tl);
			 Tv = VSUB(Te, Tl);
			 {
			      V T18, T1a, TT, T11;
			      T18 = VFMA(LDK(KP923879532), T17, T16);
			      T1a = VFNMS(LDK(KP923879532), T17, T16);
			      TT = VFNMS(LDK(KP923879532), TS, TL);
			      T11 = VFMA(LDK(KP923879532), TS, TL);
			      {
				   V T15, T19, T10, T12;
				   T15 = VFNMS(LDK(KP923879532), T14, T13);
				   T19 = VFMA(LDK(KP923879532), T14, T13);
				   T10 = VFNMS(LDK(KP923879532), TZ, TW);
				   T12 = VFMA(LDK(KP923879532), TZ, TW);
//.........这里部分代码省略.........
开发者ID:bambang,项目名称:vsipl,代码行数:101,代码来源:n2bv_16.c

示例5: n2bv_14

static void n2bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
     DVK(KP222520933, +0.222520933956314404288902564496794759466355569);
     DVK(KP623489801, +0.623489801858733530525004884004239810632274731);
     DVK(KP781831482, +0.781831482468029808708444526674057750232334519);
     DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
     DVK(KP433883739, +0.433883739117558120475768332848358754609990728);
     {
	  INT i;
	  const R *xi;
	  R *xo;
	  xi = ii;
	  xo = io;
	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) {
	       V Tp, Ty, Tl, TL, Tq, TE, T7, TJ, Ts, TB, Te, TK, Tr, TH, Tn;
	       V To;
	       Tn = LD(&(xi[0]), ivs, &(xi[0]));
	       To = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
	       Tp = VSUB(Tn, To);
	       Ty = VADD(Tn, To);
	       {
		    V Th, TC, Tk, TD;
		    {
			 V Tf, Tg, Ti, Tj;
			 Tf = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
			 Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
			 Th = VSUB(Tf, Tg);
			 TC = VADD(Tf, Tg);
			 Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
			 Tj = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
			 Tk = VSUB(Ti, Tj);
			 TD = VADD(Ti, Tj);
		    }
		    Tl = VSUB(Th, Tk);
		    TL = VSUB(TD, TC);
		    Tq = VADD(Th, Tk);
		    TE = VADD(TC, TD);
	       }
	       {
		    V T3, Tz, T6, TA;
		    {
			 V T1, T2, T4, T5;
			 T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
			 T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
			 T3 = VSUB(T1, T2);
			 Tz = VADD(T1, T2);
			 T4 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
			 T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
			 T6 = VSUB(T4, T5);
			 TA = VADD(T4, T5);
		    }
		    T7 = VSUB(T3, T6);
		    TJ = VSUB(Tz, TA);
		    Ts = VADD(T3, T6);
		    TB = VADD(Tz, TA);
	       }
	       {
		    V Ta, TF, Td, TG;
		    {
			 V T8, T9, Tb, Tc;
			 T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
			 T9 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
			 Ta = VSUB(T8, T9);
			 TF = VADD(T8, T9);
			 Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
			 Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
			 Td = VSUB(Tb, Tc);
			 TG = VADD(Tb, Tc);
		    }
		    Te = VSUB(Ta, Td);
		    TK = VSUB(TG, TF);
		    Tr = VADD(Ta, Td);
		    TH = VADD(TF, TG);
	       }
	       {
		    V TR, TS, TU, TV;
		    TR = VADD(Tp, VADD(Ts, VADD(Tq, Tr)));
		    STM2(&(xo[14]), TR, ovs, &(xo[2]));
		    TS = VADD(Ty, VADD(TB, VADD(TE, TH)));
		    STM2(&(xo[0]), TS, ovs, &(xo[0]));
		    {
			 V TT, Tm, Tt, TQ, TP, TW;
			 Tm = VBYI(VFMA(LDK(KP433883739), T7, VFNMS(LDK(KP781831482), Tl, VMUL(LDK(KP974927912), Te))));
			 Tt = VFMA(LDK(KP623489801), Tq, VFNMS(LDK(KP222520933), Tr, VFNMS(LDK(KP900968867), Ts, Tp)));
			 TT = VADD(Tm, Tt);
			 STM2(&(xo[6]), TT, ovs, &(xo[2]));
			 TU = VSUB(Tt, Tm);
			 STM2(&(xo[22]), TU, ovs, &(xo[2]));
			 TQ = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK))));
			 TP = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TB, Ty)));
			 TV = VSUB(TP, TQ);
			 STM2(&(xo[24]), TV, ovs, &(xo[0]));
			 TW = VADD(TP, TQ);
			 STM2(&(xo[4]), TW, ovs, &(xo[0]));
			 STN2(&(xo[4]), TW, TT, ovs);
		    }
		    {
			 V T10, TM, TI, TZ;
			 {
//.........这里部分代码省略.........
开发者ID:DougBurke,项目名称:sherpa,代码行数:101,代码来源:n2bv_14.c

示例6: t2sv_8

static void t2sv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     {
	  INT m;
	  for (m = mb, W = W + (mb * 6); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 6), MAKE_VOLATILE_STRIDE(rs)) {
	       V T2, T5, T3, T6, T8, Tc, Tg, Ti, Tl, Tm, Tn, Tz, Tp, Tx;
	       {
		    V T4, Tb, T7, Ta;
		    T2 = LDW(&(W[0]));
		    T5 = LDW(&(W[TWVL * 1]));
		    T3 = LDW(&(W[TWVL * 2]));
		    T6 = LDW(&(W[TWVL * 3]));
		    T4 = VMUL(T2, T3);
		    Tb = VMUL(T5, T3);
		    T7 = VMUL(T5, T6);
		    Ta = VMUL(T2, T6);
		    T8 = VSUB(T4, T7);
		    Tc = VADD(Ta, Tb);
		    Tg = VADD(T4, T7);
		    Ti = VSUB(Ta, Tb);
		    Tl = LDW(&(W[TWVL * 4]));
		    Tm = LDW(&(W[TWVL * 5]));
		    Tn = VFMA(T2, Tl, VMUL(T5, Tm));
		    Tz = VFNMS(Ti, Tl, VMUL(Tg, Tm));
		    Tp = VFNMS(T5, Tl, VMUL(T2, Tm));
		    Tx = VFMA(Tg, Tl, VMUL(Ti, Tm));
	       }
	       {
		    V Tf, T1i, TL, T1d, TJ, T17, TV, TY, Ts, T1j, TO, T1a, TC, T16, TQ;
		    V TT;
		    {
			 V T1, T1c, Te, T1b, T9, Td;
			 T1 = LD(&(ri[0]), ms, &(ri[0]));
			 T1c = LD(&(ii[0]), ms, &(ii[0]));
			 T9 = LD(&(ri[WS(rs, 4)]), ms, &(ri[0]));
			 Td = LD(&(ii[WS(rs, 4)]), ms, &(ii[0]));
			 Te = VFMA(T8, T9, VMUL(Tc, Td));
			 T1b = VFNMS(Tc, T9, VMUL(T8, Td));
			 Tf = VADD(T1, Te);
			 T1i = VSUB(T1c, T1b);
			 TL = VSUB(T1, Te);
			 T1d = VADD(T1b, T1c);
		    }
		    {
			 V TF, TW, TI, TX;
			 {
			      V TD, TE, TG, TH;
			      TD = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)]));
			      TE = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)]));
			      TF = VFMA(Tl, TD, VMUL(Tm, TE));
			      TW = VFNMS(Tm, TD, VMUL(Tl, TE));
			      TG = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)]));
			      TH = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)]));
			      TI = VFMA(T3, TG, VMUL(T6, TH));
			      TX = VFNMS(T6, TG, VMUL(T3, TH));
			 }
			 TJ = VADD(TF, TI);
			 T17 = VADD(TW, TX);
			 TV = VSUB(TF, TI);
			 TY = VSUB(TW, TX);
		    }
		    {
			 V Tk, TM, Tr, TN;
			 {
			      V Th, Tj, To, Tq;
			      Th = LD(&(ri[WS(rs, 2)]), ms, &(ri[0]));
			      Tj = LD(&(ii[WS(rs, 2)]), ms, &(ii[0]));
			      Tk = VFMA(Tg, Th, VMUL(Ti, Tj));
			      TM = VFNMS(Ti, Th, VMUL(Tg, Tj));
			      To = LD(&(ri[WS(rs, 6)]), ms, &(ri[0]));
			      Tq = LD(&(ii[WS(rs, 6)]), ms, &(ii[0]));
			      Tr = VFMA(Tn, To, VMUL(Tp, Tq));
			      TN = VFNMS(Tp, To, VMUL(Tn, Tq));
			 }
			 Ts = VADD(Tk, Tr);
			 T1j = VSUB(Tk, Tr);
			 TO = VSUB(TM, TN);
			 T1a = VADD(TM, TN);
		    }
		    {
			 V Tw, TR, TB, TS;
			 {
			      V Tu, Tv, Ty, TA;
			      Tu = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)]));
			      Tv = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)]));
			      Tw = VFMA(T2, Tu, VMUL(T5, Tv));
			      TR = VFNMS(T5, Tu, VMUL(T2, Tv));
			      Ty = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)]));
			      TA = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)]));
			      TB = VFMA(Tx, Ty, VMUL(Tz, TA));
			      TS = VFNMS(Tz, Ty, VMUL(Tx, TA));
			 }
			 TC = VADD(Tw, TB);
			 T16 = VADD(TR, TS);
			 TQ = VSUB(Tw, TB);
			 TT = VSUB(TR, TS);
		    }
		    {
			 V Tt, TK, T1f, T1g;
//.........这里部分代码省略.........
开发者ID:dstuck,项目名称:tinker_integrated_PIMC,代码行数:101,代码来源:t2sv_8.c

示例7: t1bv_15

static void t1bv_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP216506350, +0.216506350946109661690930792688234045867850657);
     DVK(KP484122918, +0.484122918275927110647408174972799951354115213);
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP509036960, +0.509036960455127183450980863393907648510733164);
     DVK(KP823639103, +0.823639103546331925877420039278190003029660514);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     {
	  INT m;
	  R *x;
	  x = ii;
	  for (m = mb, W = W + (mb * ((TWVL / VL) * 28)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 28), MAKE_VOLATILE_STRIDE(rs)) {
	       V Ts, TV, T1f, TZ, T10, Tb, Tm, Tt, T1j, T1k, T1l, TI, TM, TR, Tz;
	       V TD, TQ, T1g, T1h, T1i;
	       {
		    V TT, Tr, Tp, Tq, To, TU;
		    TT = LD(&(x[0]), ms, &(x[0]));
		    Tq = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
		    Tr = BYTW(&(W[TWVL * 18]), Tq);
		    To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
		    Tp = BYTW(&(W[TWVL * 8]), To);
		    Ts = VSUB(Tp, Tr);
		    TU = VADD(Tp, Tr);
		    TV = VFNMS(LDK(KP500000000), TU, TT);
		    T1f = VADD(TT, TU);
	       }
	       {
		    V Tx, TG, TK, TB, T5, Ty, Tg, TH, Tl, TL, Ta, TC;
		    {
			 V Tw, TF, TJ, TA;
			 Tw = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			 Tx = BYTW(&(W[TWVL * 4]), Tw);
			 TF = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
			 TG = BYTW(&(W[TWVL * 10]), TF);
			 TJ = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
			 TK = BYTW(&(W[TWVL * 16]), TJ);
			 TA = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
			 TB = BYTW(&(W[TWVL * 22]), TA);
		    }
		    {
			 V T2, T4, T1, T3;
			 T1 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
			 T2 = BYTW(&(W[TWVL * 14]), T1);
			 T3 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
			 T4 = BYTW(&(W[TWVL * 24]), T3);
			 T5 = VSUB(T2, T4);
			 Ty = VADD(T2, T4);
		    }
		    {
			 V Td, Tf, Tc, Te;
			 Tc = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
			 Td = BYTW(&(W[TWVL * 20]), Tc);
			 Te = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
			 Tf = BYTW(&(W[0]), Te);
			 Tg = VSUB(Td, Tf);
			 TH = VADD(Td, Tf);
		    }
		    {
			 V Ti, Tk, Th, Tj;
			 Th = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
			 Ti = BYTW(&(W[TWVL * 26]), Th);
			 Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
			 Tk = BYTW(&(W[TWVL * 6]), Tj);
			 Tl = VSUB(Ti, Tk);
			 TL = VADD(Ti, Tk);
		    }
		    {
			 V T7, T9, T6, T8;
			 T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
			 T7 = BYTW(&(W[TWVL * 2]), T6);
			 T8 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
			 T9 = BYTW(&(W[TWVL * 12]), T8);
			 Ta = VSUB(T7, T9);
			 TC = VADD(T7, T9);
		    }
		    TZ = VSUB(T5, Ta);
		    T10 = VSUB(Tg, Tl);
		    Tb = VADD(T5, Ta);
		    Tm = VADD(Tg, Tl);
		    Tt = VADD(Tb, Tm);
		    T1j = VADD(TG, TH);
		    T1k = VADD(TK, TL);
		    T1l = VADD(T1j, T1k);
		    TI = VFNMS(LDK(KP500000000), TH, TG);
		    TM = VFNMS(LDK(KP500000000), TL, TK);
		    TR = VADD(TI, TM);
		    Tz = VFNMS(LDK(KP500000000), Ty, Tx);
		    TD = VFNMS(LDK(KP500000000), TC, TB);
		    TQ = VADD(Tz, TD);
		    T1g = VADD(Tx, Ty);
		    T1h = VADD(TB, TC);
		    T1i = VADD(T1g, T1h);
	       }
	       {
		    V T1o, T1m, T1n, T1s, T1t, T1q, T1r, T1u, T1p;
//.........这里部分代码省略.........
开发者ID:AlexanderB1290,项目名称:scafacos,代码行数:101,代码来源:t1bv_15.c

示例8: SHIP_cell_pos

void SHIP_cell_pos(struct actor_t* a, int i, int j, vec3f pos)
{
	random_ship_t* s = (random_ship_t*)a->child;
	VMOV3(pos, 1.0f*(i - (s->w >> 1)) , 1.0f*j, 0.0f);
	VADD(pos, a->pos, pos);
}
开发者ID:javisantana,项目名称:cubyshot,代码行数:6,代码来源:random_ship.cpp

示例9: DVK

static const R *t1bv_32(R *ri, R *ii, const R *W, stride ios, int m, int dist)
{
     DVK(KP195090322, +0.195090322016128267848284868477022240927691618);
     DVK(KP980785280, +0.980785280403230449126182236134239036973933731);
     DVK(KP555570233, +0.555570233019602224742830813948532874374937191);
     DVK(KP831469612, +0.831469612302545237078788377617905756738560812);
     DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
     DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     int i;
     R *x;
     x = ii;
     BEGIN_SIMD();
     for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 62)) {
	  V T4, T1D, T2P, T3h, Tf, T1y, T2K, T3i, TC, T1w, T2G, T3e, Tr, T1v, T2D;
	  V T3d, T1k, T20, T2y, T3a, T1r, T21, T2v, T39, TV, T1X, T2r, T37, T12, T1Y;
	  V T2o, T36;
	  {
	       V T1, T1C, T3, T1A, T1B, T2, T1z, T2N, T2O;
	       T1 = LD(&(x[0]), dist, &(x[0]));
	       T1B = LD(&(x[WS(ios, 24)]), dist, &(x[0]));
	       T1C = BYTW(&(W[TWVL * 46]), T1B);
	       T2 = LD(&(x[WS(ios, 16)]), dist, &(x[0]));
	       T3 = BYTW(&(W[TWVL * 30]), T2);
	       T1z = LD(&(x[WS(ios, 8)]), dist, &(x[0]));
	       T1A = BYTW(&(W[TWVL * 14]), T1z);
	       T4 = VSUB(T1, T3);
	       T1D = VSUB(T1A, T1C);
	       T2N = VADD(T1, T3);
	       T2O = VADD(T1A, T1C);
	       T2P = VSUB(T2N, T2O);
	       T3h = VADD(T2N, T2O);
	  }
	  {
	       V T6, Td, T8, Tb;
	       {
		    V T5, Tc, T7, Ta;
		    T5 = LD(&(x[WS(ios, 4)]), dist, &(x[0]));
		    T6 = BYTW(&(W[TWVL * 6]), T5);
		    Tc = LD(&(x[WS(ios, 12)]), dist, &(x[0]));
		    Td = BYTW(&(W[TWVL * 22]), Tc);
		    T7 = LD(&(x[WS(ios, 20)]), dist, &(x[0]));
		    T8 = BYTW(&(W[TWVL * 38]), T7);
		    Ta = LD(&(x[WS(ios, 28)]), dist, &(x[0]));
		    Tb = BYTW(&(W[TWVL * 54]), Ta);
	       }
	       {
		    V T9, Te, T2I, T2J;
		    T9 = VSUB(T6, T8);
		    Te = VSUB(Tb, Td);
		    Tf = VMUL(LDK(KP707106781), VADD(T9, Te));
		    T1y = VMUL(LDK(KP707106781), VSUB(T9, Te));
		    T2I = VADD(T6, T8);
		    T2J = VADD(Tb, Td);
		    T2K = VSUB(T2I, T2J);
		    T3i = VADD(T2I, T2J);
	       }
	  }
	  {
	       V Tt, TA, Tv, Ty;
	       {
		    V Ts, Tz, Tu, Tx;
		    Ts = LD(&(x[WS(ios, 6)]), dist, &(x[0]));
		    Tt = BYTW(&(W[TWVL * 10]), Ts);
		    Tz = LD(&(x[WS(ios, 14)]), dist, &(x[0]));
		    TA = BYTW(&(W[TWVL * 26]), Tz);
		    Tu = LD(&(x[WS(ios, 22)]), dist, &(x[0]));
		    Tv = BYTW(&(W[TWVL * 42]), Tu);
		    Tx = LD(&(x[WS(ios, 30)]), dist, &(x[0]));
		    Ty = BYTW(&(W[TWVL * 58]), Tx);
	       }
	       {
		    V Tw, TB, T2E, T2F;
		    Tw = VSUB(Tt, Tv);
		    TB = VSUB(Ty, TA);
		    TC = VFNMS(LDK(KP382683432), TB, VMUL(LDK(KP923879532), Tw));
		    T1w = VFMA(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw));
		    T2E = VADD(Ty, TA);
		    T2F = VADD(Tt, Tv);
		    T2G = VSUB(T2E, T2F);
		    T3e = VADD(T2E, T2F);
	       }
	  }
	  {
	       V Ti, Tp, Tk, Tn;
	       {
		    V Th, To, Tj, Tm;
		    Th = LD(&(x[WS(ios, 2)]), dist, &(x[0]));
		    Ti = BYTW(&(W[TWVL * 2]), Th);
		    To = LD(&(x[WS(ios, 26)]), dist, &(x[0]));
		    Tp = BYTW(&(W[TWVL * 50]), To);
		    Tj = LD(&(x[WS(ios, 18)]), dist, &(x[0]));
		    Tk = BYTW(&(W[TWVL * 34]), Tj);
		    Tm = LD(&(x[WS(ios, 10)]), dist, &(x[0]));
		    Tn = BYTW(&(W[TWVL * 18]), Tm);
	       }
	       {
		    V Tl, Tq, T2B, T2C;
		    Tl = VSUB(Ti, Tk);
		    Tq = VSUB(Tn, Tp);
//.........这里部分代码省略.........
开发者ID:abrahamneben,项目名称:orbcomm_beam_mapping,代码行数:101,代码来源:t1bv_32.c

示例10: n1fv_9

static void n1fv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
     DVK(KP826351822, +0.826351822333069651148283373230685203999624323);
     DVK(KP879385241, +0.879385241571816768108218554649462939872416269);
     DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
     DVK(KP666666666, +0.666666666666666666666666666666666666666666667);
     DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
     DVK(KP907603734, +0.907603734547952313649323976213898122064543220);
     DVK(KP420276625, +0.420276625461206169731530603237061658838781920);
     DVK(KP673648177, +0.673648177666930348851716626769314796000375677);
     DVK(KP898197570, +0.898197570222573798468955502359086394667167570);
     DVK(KP347296355, +0.347296355333860697703433253538629592000751354);
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP439692620, +0.439692620785908384054109277324731469936208134);
     DVK(KP203604859, +0.203604859554852403062088995281827210665664861);
     DVK(KP152703644, +0.152703644666139302296566746461370407999248646);
     DVK(KP586256827, +0.586256827714544512072145703099641959914944179);
     DVK(KP968908795, +0.968908795874236621082202410917456709164223497);
     DVK(KP726681596, +0.726681596905677465811651808188092531873167623);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     {
	  INT i;
	  const R *xi;
	  R *xo;
	  xi = ri;
	  xo = ro;
	  for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(18, is), MAKE_VOLATILE_STRIDE(18, os)) {
	       V T1, T2, T3, T6, Tb, T7, T8, Tc, Td, Tv, T4;
	       T1 = LD(&(xi[0]), ivs, &(xi[0]));
	       T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
	       T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
	       T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
	       Tb = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
	       T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
	       T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
	       Tc = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
	       Tv = VSUB(T3, T2);
	       T4 = VADD(T2, T3);
	       {
		    V Tl, T9, Tm, Te, Tj, T5;
		    Tl = VSUB(T7, T8);
		    T9 = VADD(T7, T8);
		    Tm = VSUB(Td, Tc);
		    Te = VADD(Tc, Td);
		    Tj = VFNMS(LDK(KP500000000), T4, T1);
		    T5 = VADD(T1, T4);
		    {
			 V Tn, Ta, Tk, Tf;
			 Tn = VFNMS(LDK(KP500000000), T9, T6);
			 Ta = VADD(T6, T9);
			 Tk = VFNMS(LDK(KP500000000), Te, Tb);
			 Tf = VADD(Tb, Te);
			 {
			      V Ty, TC, To, TB, Tx, Ts, Tg, Ti;
			      Ty = VFNMS(LDK(KP726681596), Tl, Tn);
			      TC = VFMA(LDK(KP968908795), Tn, Tl);
			      To = VFNMS(LDK(KP586256827), Tn, Tm);
			      TB = VFNMS(LDK(KP152703644), Tm, Tk);
			      Tx = VFMA(LDK(KP203604859), Tk, Tm);
			      Ts = VFNMS(LDK(KP439692620), Tl, Tk);
			      Tg = VADD(Ta, Tf);
			      Ti = VMUL(LDK(KP866025403), VSUB(Tf, Ta));
			      {
				   V Tz, TI, TF, TD, Tt, Th, Tq, Tp;
				   Tp = VFNMS(LDK(KP347296355), To, Tl);
				   Tz = VFMA(LDK(KP898197570), Ty, Tx);
				   TI = VFNMS(LDK(KP898197570), Ty, Tx);
				   TF = VFNMS(LDK(KP673648177), TC, TB);
				   TD = VFMA(LDK(KP673648177), TC, TB);
				   Tt = VFNMS(LDK(KP420276625), Ts, Tm);
				   ST(&(xo[0]), VADD(T5, Tg), ovs, &(xo[0]));
				   Th = VFNMS(LDK(KP500000000), Tg, T5);
				   Tq = VFNMS(LDK(KP907603734), Tp, Tk);
				   {
					V TA, TJ, TE, TG, Tu, Tr, TK, TH, Tw;
					TA = VFMA(LDK(KP852868531), Tz, Tj);
					TJ = VFMA(LDK(KP666666666), TD, TI);
					TE = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tv, TD));
					TG = VFNMS(LDK(KP500000000), Tz, TF);
					Tu = VFNMS(LDK(KP826351822), Tt, Tn);
					ST(&(xo[WS(os, 6)]), VFNMSI(Ti, Th), ovs, &(xo[0]));
					ST(&(xo[WS(os, 3)]), VFMAI(Ti, Th), ovs, &(xo[WS(os, 1)]));
					Tr = VFNMS(LDK(KP939692620), Tq, Tj);
					TK = VMUL(LDK(KP866025403), VFMA(LDK(KP852868531), TJ, Tv));
					ST(&(xo[WS(os, 8)]), VFMAI(TE, TA), ovs, &(xo[0]));
					ST(&(xo[WS(os, 1)]), VFNMSI(TE, TA), ovs, &(xo[WS(os, 1)]));
					TH = VFMA(LDK(KP852868531), TG, Tj);
					Tw = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tv, Tu));
					ST(&(xo[WS(os, 4)]), VFMAI(TK, TH), ovs, &(xo[0]));
					ST(&(xo[WS(os, 5)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)]));
					ST(&(xo[WS(os, 7)]), VFMAI(Tw, Tr), ovs, &(xo[WS(os, 1)]));
					ST(&(xo[WS(os, 2)]), VFNMSI(Tw, Tr), ovs, &(xo[0]));
				   }
			      }
			 }
		    }
	       }
	  }
//.........这里部分代码省略.........
开发者ID:SKA-ScienceDataProcessor,项目名称:FastImaging,代码行数:101,代码来源:n1fv_9.c

示例11: n1bv_11

static void n1bv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
     DVK(KP959492973, +0.959492973614497389890368057066327699062454848);
     DVK(KP654860733, +0.654860733945285064056925072466293553183791199);
     DVK(KP142314838, +0.142314838273285140443792668616369668791051361);
     DVK(KP415415013, +0.415415013001886425529274149229623203524004910);
     DVK(KP841253532, +0.841253532831181168861811648919367717513292498);
     DVK(KP540640817, +0.540640817455597582107635954318691695431770608);
     DVK(KP909631995, +0.909631995354518371411715383079028460060241051);
     DVK(KP989821441, +0.989821441880932732376092037776718787376519372);
     DVK(KP755749574, +0.755749574354258283774035843972344420179717445);
     DVK(KP281732556, +0.281732556841429697711417915346616899035777899);
     int i;
     const R *xi;
     R *xo;
     xi = ii;
     xo = io;
     BEGIN_SIMD();
     for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
	  V Th, T3, Tm, Tf, Ti, Tc, Tj, T9, Tk, T6, Tl, Ta, Tb, Ts, Tt;
	  Th = LD(&(xi[0]), ivs, &(xi[0]));
	  {
	       V T1, T2, Td, Te;
	       T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
	       T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
	       T3 = VSUB(T1, T2);
	       Tm = VADD(T1, T2);
	       Td = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
	       Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
	       Tf = VSUB(Td, Te);
	       Ti = VADD(Td, Te);
	  }
	  Ta = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
	  Tb = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
	  Tc = VSUB(Ta, Tb);
	  Tj = VADD(Ta, Tb);
	  {
	       V T7, T8, T4, T5;
	       T7 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
	       T9 = VSUB(T7, T8);
	       Tk = VADD(T7, T8);
	       T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
	       T5 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
	       T6 = VSUB(T4, T5);
	       Tl = VADD(T4, T5);
	  }
	  ST(&(xo[0]), VADD(Th, VADD(Tm, VADD(Ti, VADD(Tl, VADD(Tj, Tk))))), ovs, &(xo[0]));
	  {
	       V Tg, Tn, Tu, Tv;
	       Tg = VBYI(VFMA(LDK(KP281732556), T3, VFMA(LDK(KP755749574), T6, VFNMS(LDK(KP909631995), Tc, VFNMS(LDK(KP540640817), Tf, VMUL(LDK(KP989821441), T9))))));
	       Tn = VFMA(LDK(KP841253532), Ti, VFMA(LDK(KP415415013), Tj, VFNMS(LDK(KP142314838), Tk, VFNMS(LDK(KP654860733), Tl, VFNMS(LDK(KP959492973), Tm, Th)))));
	       ST(&(xo[WS(os, 5)]), VADD(Tg, Tn), ovs, &(xo[WS(os, 1)]));
	       ST(&(xo[WS(os, 6)]), VSUB(Tn, Tg), ovs, &(xo[0]));
	       Tu = VBYI(VFMA(LDK(KP755749574), T3, VFMA(LDK(KP540640817), T6, VFNMS(LDK(KP909631995), T9, VFNMS(LDK(KP989821441), Tf, VMUL(LDK(KP281732556), Tc))))));
	       Tv = VFMA(LDK(KP841253532), Tl, VFMA(LDK(KP415415013), Tk, VFNMS(LDK(KP959492973), Tj, VFNMS(LDK(KP142314838), Ti, VFNMS(LDK(KP654860733), Tm, Th)))));
	       ST(&(xo[WS(os, 4)]), VADD(Tu, Tv), ovs, &(xo[0]));
	       ST(&(xo[WS(os, 7)]), VSUB(Tv, Tu), ovs, &(xo[WS(os, 1)]));
	  }
	  Ts = VBYI(VFMA(LDK(KP909631995), T3, VFNMS(LDK(KP540640817), T9, VFNMS(LDK(KP989821441), Tc, VFNMS(LDK(KP281732556), T6, VMUL(LDK(KP755749574), Tf))))));
	  Tt = VFMA(LDK(KP415415013), Tm, VFMA(LDK(KP841253532), Tk, VFNMS(LDK(KP142314838), Tj, VFNMS(LDK(KP959492973), Tl, VFNMS(LDK(KP654860733), Ti, Th)))));
	  ST(&(xo[WS(os, 2)]), VADD(Ts, Tt), ovs, &(xo[0]));
	  ST(&(xo[WS(os, 9)]), VSUB(Tt, Ts), ovs, &(xo[WS(os, 1)]));
	  {
	       V Tq, Tr, To, Tp;
	       Tq = VBYI(VFMA(LDK(KP540640817), T3, VFMA(LDK(KP909631995), Tf, VFMA(LDK(KP989821441), T6, VFMA(LDK(KP755749574), Tc, VMUL(LDK(KP281732556), T9))))));
	       Tr = VFMA(LDK(KP841253532), Tm, VFMA(LDK(KP415415013), Ti, VFNMS(LDK(KP959492973), Tk, VFNMS(LDK(KP654860733), Tj, VFNMS(LDK(KP142314838), Tl, Th)))));
	       ST(&(xo[WS(os, 1)]), VADD(Tq, Tr), ovs, &(xo[WS(os, 1)]));
	       ST(&(xo[WS(os, 10)]), VSUB(Tr, Tq), ovs, &(xo[0]));
	       To = VBYI(VFMA(LDK(KP989821441), T3, VFMA(LDK(KP540640817), Tc, VFNMS(LDK(KP909631995), T6, VFNMS(LDK(KP281732556), Tf, VMUL(LDK(KP755749574), T9))))));
	       Tp = VFMA(LDK(KP415415013), Tl, VFMA(LDK(KP841253532), Tj, VFNMS(LDK(KP654860733), Tk, VFNMS(LDK(KP959492973), Ti, VFNMS(LDK(KP142314838), Tm, Th)))));
	       ST(&(xo[WS(os, 3)]), VADD(To, Tp), ovs, &(xo[WS(os, 1)]));
	       ST(&(xo[WS(os, 8)]), VSUB(Tp, To), ovs, &(xo[0]));
	  }
     }
     END_SIMD();
}
开发者ID:abrahamneben,项目名称:orbcomm_beam_mapping,代码行数:77,代码来源:n1bv_11.c

示例12: n2sv_8

static void n2sv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     {
	  INT i;
	  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(32, is), MAKE_VOLATILE_STRIDE(32, os)) {
	       V T3, Tn, Ti, TC, T6, TB, Tl, To, Td, TN, Tz, TH, Ta, TM, Tu;
	       V TG;
	       {
		    V T1, T2, Tj, Tk;
		    T1 = LD(&(ri[0]), ivs, &(ri[0]));
		    T2 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0]));
		    T3 = VADD(T1, T2);
		    Tn = VSUB(T1, T2);
		    {
			 V Tg, Th, T4, T5;
			 Tg = LD(&(ii[0]), ivs, &(ii[0]));
			 Th = LD(&(ii[WS(is, 4)]), ivs, &(ii[0]));
			 Ti = VADD(Tg, Th);
			 TC = VSUB(Tg, Th);
			 T4 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0]));
			 T5 = LD(&(ri[WS(is, 6)]), ivs, &(ri[0]));
			 T6 = VADD(T4, T5);
			 TB = VSUB(T4, T5);
		    }
		    Tj = LD(&(ii[WS(is, 2)]), ivs, &(ii[0]));
		    Tk = LD(&(ii[WS(is, 6)]), ivs, &(ii[0]));
		    Tl = VADD(Tj, Tk);
		    To = VSUB(Tj, Tk);
		    {
			 V Tb, Tc, Tv, Tw, Tx, Ty;
			 Tb = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)]));
			 Tc = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)]));
			 Tv = VSUB(Tb, Tc);
			 Tw = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)]));
			 Tx = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)]));
			 Ty = VSUB(Tw, Tx);
			 Td = VADD(Tb, Tc);
			 TN = VADD(Tw, Tx);
			 Tz = VSUB(Tv, Ty);
			 TH = VADD(Tv, Ty);
		    }
		    {
			 V T8, T9, Tq, Tr, Ts, Tt;
			 T8 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)]));
			 T9 = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)]));
			 Tq = VSUB(T8, T9);
			 Tr = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)]));
			 Ts = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)]));
			 Tt = VSUB(Tr, Ts);
			 Ta = VADD(T8, T9);
			 TM = VADD(Tr, Ts);
			 Tu = VADD(Tq, Tt);
			 TG = VSUB(Tt, Tq);
		    }
	       }
	       {
		    V TR, TS, TT, TU, TV, TW, TX, TY;
		    {
			 V T7, Te, TP, TQ;
			 T7 = VADD(T3, T6);
			 Te = VADD(Ta, Td);
			 TR = VSUB(T7, Te);
			 STM4(&(ro[4]), TR, ovs, &(ro[0]));
			 TS = VADD(T7, Te);
			 STM4(&(ro[0]), TS, ovs, &(ro[0]));
			 TP = VADD(Ti, Tl);
			 TQ = VADD(TM, TN);
			 TT = VSUB(TP, TQ);
			 STM4(&(io[4]), TT, ovs, &(io[0]));
			 TU = VADD(TP, TQ);
			 STM4(&(io[0]), TU, ovs, &(io[0]));
		    }
		    {
			 V Tf, Tm, TL, TO;
			 Tf = VSUB(Td, Ta);
			 Tm = VSUB(Ti, Tl);
			 TV = VADD(Tf, Tm);
			 STM4(&(io[2]), TV, ovs, &(io[0]));
			 TW = VSUB(Tm, Tf);
			 STM4(&(io[6]), TW, ovs, &(io[0]));
			 TL = VSUB(T3, T6);
			 TO = VSUB(TM, TN);
			 TX = VSUB(TL, TO);
			 STM4(&(ro[6]), TX, ovs, &(ro[0]));
			 TY = VADD(TL, TO);
			 STM4(&(ro[2]), TY, ovs, &(ro[0]));
		    }
		    {
			 V TZ, T10, T11, T12;
			 {
			      V Tp, TA, TJ, TK;
			      Tp = VADD(Tn, To);
			      TA = VADD(Tu, Tz);
			      TZ = VFNMS(LDK(KP707106781), TA, Tp);
			      STM4(&(ro[5]), TZ, ovs, &(ro[1]));
			      T10 = VFMA(LDK(KP707106781), TA, Tp);
			      STM4(&(ro[1]), T10, ovs, &(ro[1]));
			      TJ = VSUB(TC, TB);
			      TK = VADD(TG, TH);
//.........这里部分代码省略.........
开发者ID:DougBurke,项目名称:sherpa,代码行数:101,代码来源:n2sv_8.c

示例13: t1fv_20

static void t1fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     {
	  INT m;
	  R *x;
	  x = ri;
	  for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) {
	       V T4, Tx, T1m, T1K, TZ, T16, T17, T10, Tf, Tq, Tr, T1O, T1P, T1Q, T1w;
	       V T1z, T1A, TI, TT, TU, T1L, T1M, T1N, T1p, T1s, T1t, Ts, TV;
	       {
		    V T1, Tw, T3, Tu, Tv, T2, Tt, T1k, T1l;
		    T1 = LD(&(x[0]), ms, &(x[0]));
		    Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
		    Tw = BYTWJ(&(W[TWVL * 28]), Tv);
		    T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
		    T3 = BYTWJ(&(W[TWVL * 18]), T2);
		    Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
		    Tu = BYTWJ(&(W[TWVL * 8]), Tt);
		    T4 = VSUB(T1, T3);
		    Tx = VSUB(Tu, Tw);
		    T1k = VADD(T1, T3);
		    T1l = VADD(Tu, Tw);
		    T1m = VSUB(T1k, T1l);
		    T1K = VADD(T1k, T1l);
	       }
	       {
		    V T9, T1n, TN, T1v, TS, T1y, Te, T1q, Tk, T1u, TC, T1o, TH, T1r, Tp;
		    V T1x;
		    {
			 V T6, T8, T5, T7;
			 T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
			 T6 = BYTWJ(&(W[TWVL * 6]), T5);
			 T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
			 T8 = BYTWJ(&(W[TWVL * 26]), T7);
			 T9 = VSUB(T6, T8);
			 T1n = VADD(T6, T8);
		    }
		    {
			 V TK, TM, TJ, TL;
			 TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
			 TK = BYTWJ(&(W[TWVL * 24]), TJ);
			 TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			 TM = BYTWJ(&(W[TWVL * 4]), TL);
			 TN = VSUB(TK, TM);
			 T1v = VADD(TK, TM);
		    }
		    {
			 V TP, TR, TO, TQ;
			 TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)]));
			 TP = BYTWJ(&(W[TWVL * 32]), TO);
			 TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
			 TR = BYTWJ(&(W[TWVL * 12]), TQ);
			 TS = VSUB(TP, TR);
			 T1y = VADD(TP, TR);
		    }
		    {
			 V Tb, Td, Ta, Tc;
			 Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0]));
			 Tb = BYTWJ(&(W[TWVL * 30]), Ta);
			 Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
			 Td = BYTWJ(&(W[TWVL * 10]), Tc);
			 Te = VSUB(Tb, Td);
			 T1q = VADD(Tb, Td);
		    }
		    {
			 V Th, Tj, Tg, Ti;
			 Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
			 Th = BYTWJ(&(W[TWVL * 14]), Tg);
			 Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0]));
			 Tj = BYTWJ(&(W[TWVL * 34]), Ti);
			 Tk = VSUB(Th, Tj);
			 T1u = VADD(Th, Tj);
		    }
		    {
			 V Tz, TB, Ty, TA;
			 Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
			 Tz = BYTWJ(&(W[TWVL * 16]), Ty);
			 TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)]));
			 TB = BYTWJ(&(W[TWVL * 36]), TA);
			 TC = VSUB(Tz, TB);
			 T1o = VADD(Tz, TB);
		    }
		    {
			 V TE, TG, TD, TF;
			 TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
			 TE = BYTWJ(&(W[0]), TD);
			 TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
			 TG = BYTWJ(&(W[TWVL * 20]), TF);
			 TH = VSUB(TE, TG);
			 T1r = VADD(TE, TG);
		    }
		    {
			 V Tm, To, Tl, Tn;
			 Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
			 Tm = BYTWJ(&(W[TWVL * 22]), Tl);
			 Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
//.........这里部分代码省略.........
开发者ID:DougBurke,项目名称:sherpa,代码行数:101,代码来源:t1fv_20.c

示例14: t2sv_16


//.........这里部分代码省略.........
					T1u = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)]));
					T1j = VFNMS(Tm, T3, T1i);
					T1p = VFMA(Tm, T3, T1i);
					T1m = VFNMS(Tm, T6, T1e);
					T1f = VFMA(Tm, T6, T1e);
					T1D = VFNMS(TC, T3, T1C);
					T1O = VFMA(TC, T3, T1C);
					T1L = VFNMS(TC, T6, T1y);
					T1z = VFMA(TC, T6, T1y);
					T9 = VMUL(T7, T8);
					T3x = VMUL(T7, Tc);
					T1v = VMUL(TM, T1u);
					T1w = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)]));
					T1A = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)]));
					T1E = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)]));
				   }
				   {
					V T1x, T2x, T1F, T2z, T1N, T1P;
					{
					     V T1H, T1J, T1I, T2E;
					     {
						  V Td, T3y, T2w, T1B, T2y;
						  T1H = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)]));
						  T1J = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)]));
						  Td = VFMA(Tb, Tc, T9);
						  T3y = VFNMS(Tb, T8, T3x);
						  T1M = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)]));
						  T1x = VFMA(TO, T1w, T1v);
						  T2w = VMUL(TM, T1w);
						  T1B = VMUL(T1z, T1A);
						  T2y = VMUL(T1z, T1E);
						  T1I = VMUL(Tf, T1H);
						  T2E = VMUL(Tf, T1J);
						  Te = VADD(T1, Td);
						  T1U = VSUB(T1, Td);
						  T3A = VADD(T3y, T3z);
						  T3L = VSUB(T3z, T3y);
						  T2x = VFNMS(TO, T1u, T2w);
						  T1F = VFMA(T1D, T1E, T1B);
						  T2z = VFNMS(T1D, T1A, T2y);
						  T1N = VMUL(T1L, T1M);
						  T1P = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)]));
					     }
					     T1K = VFMA(Th, T1J, T1I);
					     T2F = VFNMS(Th, T1H, T2E);
					}
					Tj = LD(&(ri[WS(rs, 4)]), ms, &(ri[0]));
					Tn = LD(&(ii[WS(rs, 4)]), ms, &(ii[0]));
					T2D = VSUB(T1x, T1F);
					T1G = VADD(T1x, T1F);
					T3h = VADD(T2x, T2z);
					T2A = VSUB(T2x, T2z);
					T1Q = VFMA(T1O, T1P, T1N);
					T2G = VMUL(T1L, T1P);
					Tk = VMUL(Ti, Tj);
					T1V = VMUL(Ti, Tn);
					Tr = LD(&(ri[WS(rs, 12)]), ms, &(ri[0]));
					Tv = LD(&(ii[WS(rs, 12)]), ms, &(ii[0]));
				   }
			      }
			      {
				   V TE, T22, T15, T17, TK, T16, T2h, T24, T19, T1b;
				   {
					V To, T1W, TG, TJ, Tw, T1Y, TH, T23;
					{
					     V TA, TD, TB, T21, T2H, Ts, T1X;
开发者ID:Aegisub,项目名称:fftw3,代码行数:67,代码来源:t2sv_16.c

示例15: n2sv_16

static void n2sv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     DVK(KP414213562, +0.414213562373095048801688724209698078569671875);
     INT i;
     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(is), MAKE_VOLATILE_STRIDE(os)) {
	  V T2p, T2q, T2r, T2s, T2x, T2y, T2z, T2A, T1M, T1N, T1L, T1P, T2F, T2G, T2H;
	  V T2I, T1O, T1Q;
	  {
	       V T1l, T1H, T1R, T7, T1x, TN, TC, T25, T1E, T1b, T1Z, Tt, T2h, T22, T1D;
	       V T1g, T1n, TQ, T11, Ti, Te, T26, T1m, TT, T1S, TJ, TZ, T1V, TW, Tl;
	       V T12, T13;
	       {
		    V Tq, T1c, Tp, T20, T1a, Tr, T1d, T1e;
		    {
			 V T1, T2, Tw, Tx, T4, T5, Tz, TA;
			 T1 = LD(&(ri[0]), ivs, &(ri[0]));
			 T2 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0]));
			 Tw = LD(&(ii[0]), ivs, &(ii[0]));
			 Tx = LD(&(ii[WS(is, 8)]), ivs, &(ii[0]));
			 T4 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0]));
			 T5 = LD(&(ri[WS(is, 12)]), ivs, &(ri[0]));
			 Tz = LD(&(ii[WS(is, 4)]), ivs, &(ii[0]));
			 TA = LD(&(ii[WS(is, 12)]), ivs, &(ii[0]));
			 {
			      V Tn, TL, T3, T1k, Ty, T1j, T6, TM, TB, To, T18, T19;
			      Tn = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)]));
			      TL = VSUB(T1, T2);
			      T3 = VADD(T1, T2);
			      T1k = VSUB(Tw, Tx);
			      Ty = VADD(Tw, Tx);
			      T1j = VSUB(T4, T5);
			      T6 = VADD(T4, T5);
			      TM = VSUB(Tz, TA);
			      TB = VADD(Tz, TA);
			      To = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)]));
			      T18 = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)]));
			      T19 = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)]));
			      Tq = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)]));
			      T1l = VADD(T1j, T1k);
			      T1H = VSUB(T1k, T1j);
			      T1R = VSUB(T3, T6);
			      T7 = VADD(T3, T6);
			      T1x = VADD(TL, TM);
			      TN = VSUB(TL, TM);
			      TC = VADD(Ty, TB);
			      T25 = VSUB(Ty, TB);
			      T1c = VSUB(Tn, To);
			      Tp = VADD(Tn, To);
			      T20 = VADD(T18, T19);
			      T1a = VSUB(T18, T19);
			      Tr = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)]));
			      T1d = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)]));
			      T1e = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)]));
			 }
		    }
		    {
			 V Tb, Ta, TF, Tc, TG, TH, TP, TO;
			 {
			      V T8, T9, TD, TE;
			      T8 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0]));
			      T9 = LD(&(ri[WS(is, 10)]), ivs, &(ri[0]));
			      TD = LD(&(ii[WS(is, 2)]), ivs, &(ii[0]));
			      TE = LD(&(ii[WS(is, 10)]), ivs, &(ii[0]));
			      Tb = LD(&(ri[WS(is, 14)]), ivs, &(ri[0]));
			      {
				   V T17, Ts, T21, T1f;
				   T17 = VSUB(Tq, Tr);
				   Ts = VADD(Tq, Tr);
				   T21 = VADD(T1d, T1e);
				   T1f = VSUB(T1d, T1e);
				   TP = VSUB(T8, T9);
				   Ta = VADD(T8, T9);
				   TO = VSUB(TD, TE);
				   TF = VADD(TD, TE);
				   T1E = VSUB(T1a, T17);
				   T1b = VADD(T17, T1a);
				   T1Z = VSUB(Tp, Ts);
				   Tt = VADD(Tp, Ts);
				   T2h = VADD(T20, T21);
				   T22 = VSUB(T20, T21);
				   T1D = VADD(T1c, T1f);
				   T1g = VSUB(T1c, T1f);
				   Tc = LD(&(ri[WS(is, 6)]), ivs, &(ri[0]));
			      }
			      TG = LD(&(ii[WS(is, 14)]), ivs, &(ii[0]));
			      TH = LD(&(ii[WS(is, 6)]), ivs, &(ii[0]));
			 }
			 T1n = VADD(TP, TO);
			 TQ = VSUB(TO, TP);
			 {
			      V Tg, Th, TX, TR, Td, TS, TI, TY, Tj, Tk;
			      Tg = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)]));
			      Th = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)]));
			      TX = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)]));
			      TR = VSUB(Tb, Tc);
			      Td = VADD(Tb, Tc);
			      TS = VSUB(TG, TH);
			      TI = VADD(TG, TH);
//.........这里部分代码省略.........
开发者ID:exic,项目名称:last.fm-dbus,代码行数:101,代码来源:n2sv_16.c


注:本文中的VADD函数示例由纯净天空整理自Github/MSDocs等开源代码及文档管理平台,相关代码片段筛选自各路编程大神贡献的开源项目,源码版权归原作者所有,传播和使用请参考对应项目的License;未经允许,请勿转载。