当前位置: 首页>>代码示例>>C++>>正文


C++ LD函数代码示例

本文整理汇总了C++中LD函数的典型用法代码示例。如果您正苦于以下问题:C++ LD函数的具体用法?C++ LD怎么用?C++ LD使用的例子?那么恭喜您, 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LD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: q1fv_4

static void q1fv_4(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms)
{
     {
	  INT m;
	  R *x;
	  x = ri;
	  for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, vs)) {
	       V Tb, Tm, Tx, TI;
	       {
		    V Tc, T9, T3, TG, TA, TH, TD, Ta, T6, Td, Tn, To, Tq, Tr, Tf;
		    V Tg;
		    {
			 V T1, T2, Ty, Tz, TB, TC, T4, T5;
			 T1 = LD(&(x[0]), ms, &(x[0]));
			 T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
			 Ty = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)]));
			 Tz = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)]));
			 TB = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)]));
			 TC = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)]));
			 T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
			 T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			 Tc = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)]));
			 T9 = VADD(T1, T2);
			 T3 = VSUB(T1, T2);
			 TG = VADD(Ty, Tz);
			 TA = VSUB(Ty, Tz);
			 TH = VADD(TB, TC);
			 TD = VSUB(TB, TC);
			 Ta = VADD(T4, T5);
			 T6 = VSUB(T4, T5);
			 Td = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)]));
			 Tn = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)]));
			 To = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)]));
			 Tq = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)]));
			 Tr = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)]));
			 Tf = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)]));
			 Tg = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)]));
		    }
		    {
			 V Tk, Te, Tv, Tp, Tw, Ts, Tl, Th, T7, TE, Tu, TF;
			 ST(&(x[0]), VADD(T9, Ta), ms, &(x[0]));
			 Tk = VADD(Tc, Td);
			 Te = VSUB(Tc, Td);
			 Tv = VADD(Tn, To);
			 Tp = VSUB(Tn, To);
			 Tw = VADD(Tq, Tr);
			 Ts = VSUB(Tq, Tr);
			 Tl = VADD(Tf, Tg);
			 Th = VSUB(Tf, Tg);
			 ST(&(x[WS(rs, 3)]), VADD(TG, TH), ms, &(x[WS(rs, 1)]));
			 T7 = BYTWJ(&(W[0]), VFNMSI(T6, T3));
			 TE = BYTWJ(&(W[0]), VFNMSI(TD, TA));
			 {
			      V Tt, Ti, Tj, T8;
			      T8 = BYTWJ(&(W[TWVL * 4]), VFMAI(T6, T3));
			      ST(&(x[WS(rs, 2)]), VADD(Tv, Tw), ms, &(x[0]));
			      Tt = BYTWJ(&(W[0]), VFNMSI(Ts, Tp));
			      ST(&(x[WS(rs, 1)]), VADD(Tk, Tl), ms, &(x[WS(rs, 1)]));
			      Ti = BYTWJ(&(W[0]), VFNMSI(Th, Te));
			      Tj = BYTWJ(&(W[TWVL * 4]), VFMAI(Th, Te));
			      ST(&(x[WS(vs, 1)]), T7, ms, &(x[WS(vs, 1)]));
			      ST(&(x[WS(vs, 1) + WS(rs, 3)]), TE, ms, &(x[WS(vs, 1) + WS(rs, 1)]));
			      ST(&(x[WS(vs, 3)]), T8, ms, &(x[WS(vs, 3)]));
			      Tu = BYTWJ(&(W[TWVL * 4]), VFMAI(Ts, Tp));
			      ST(&(x[WS(vs, 1) + WS(rs, 2)]), Tt, ms, &(x[WS(vs, 1)]));
			      TF = BYTWJ(&(W[TWVL * 4]), VFMAI(TD, TA));
			      ST(&(x[WS(vs, 1) + WS(rs, 1)]), Ti, ms, &(x[WS(vs, 1) + WS(rs, 1)]));
			      ST(&(x[WS(vs, 3) + WS(rs, 1)]), Tj, ms, &(x[WS(vs, 3) + WS(rs, 1)]));
			 }
			 Tb = BYTWJ(&(W[TWVL * 2]), VSUB(T9, Ta));
			 Tm = BYTWJ(&(W[TWVL * 2]), VSUB(Tk, Tl));
			 Tx = BYTWJ(&(W[TWVL * 2]), VSUB(Tv, Tw));
			 ST(&(x[WS(vs, 3) + WS(rs, 2)]), Tu, ms, &(x[WS(vs, 3)]));
			 TI = BYTWJ(&(W[TWVL * 2]), VSUB(TG, TH));
			 ST(&(x[WS(vs, 3) + WS(rs, 3)]), TF, ms, &(x[WS(vs, 3) + WS(rs, 1)]));
		    }
	       }
	       ST(&(x[WS(vs, 2)]), Tb, ms, &(x[WS(vs, 2)]));
	       ST(&(x[WS(vs, 2) + WS(rs, 1)]), Tm, ms, &(x[WS(vs, 2) + WS(rs, 1)]));
	       ST(&(x[WS(vs, 2) + WS(rs, 2)]), Tx, ms, &(x[WS(vs, 2)]));
	       ST(&(x[WS(vs, 2) + WS(rs, 3)]), TI, ms, &(x[WS(vs, 2) + WS(rs, 1)]));
	  }
     }
     VLEAVE();
}
开发者ID:SKA-ScienceDataProcessor,项目名称:FastImaging,代码行数:85,代码来源:q1fv_4.c

示例2: n2bv_20

static void n2bv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     {
	  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(40, is), MAKE_VOLATILE_STRIDE(40, os)) {
	       V T1H, T1I, TS, TA, TN, TV, T1M, T1N, T1O, T1P, T1R, T1S, TK, TU, TR;
	       V Tl;
	       {
		    V T3, TE, T1r, T13, Ta, TL, Tz, TG, Ts, TF, Th, TM, T1u, T1C, T1n;
		    V T1a, T1m, T1h, T1x, T1D, Tk, Ti;
		    {
			 V T1, T2, TC, TD;
			 T1 = LD(&(xi[0]), ivs, &(xi[0]));
			 T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
			 TC = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
			 TD = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
			 {
			      V T14, T6, T1c, Tv, Tm, T1f, Ty, T17, T9, Tn, Tp, T1b, Td, Tq, Te;
			      V Tf, T15, To;
			      {
				   V Tw, Tx, T7, T8, Tb, Tc;
				   {
					V T4, T5, Tt, Tu, T11, T12;
					T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
					T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
					Tt = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
					Tu = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
					Tw = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
					T3 = VSUB(T1, T2);
					T11 = VADD(T1, T2);
					TE = VSUB(TC, TD);
					T12 = VADD(TC, TD);
					T14 = VADD(T4, T5);
					T6 = VSUB(T4, T5);
					T1c = VADD(Tt, Tu);
					Tv = VSUB(Tt, Tu);
					Tx = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
					T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
					T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
					T1r = VADD(T11, T12);
					T13 = VSUB(T11, T12);
				   }
				   Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
				   Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
				   Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
				   T1f = VADD(Tw, Tx);
				   Ty = VSUB(Tw, Tx);
				   T17 = VADD(T7, T8);
				   T9 = VSUB(T7, T8);
				   Tn = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
				   Tp = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
				   T1b = VADD(Tb, Tc);
				   Td = VSUB(Tb, Tc);
				   Tq = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
				   Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
				   Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
			      }
			      Ta = VADD(T6, T9);
			      TL = VSUB(T6, T9);
			      T15 = VADD(Tm, Tn);
			      To = VSUB(Tm, Tn);
			      Tz = VSUB(Tv, Ty);
			      TG = VADD(Tv, Ty);
			      {
				   V T1d, T1v, T18, Tr, T1e, Tg, T16, T1s;
				   T1d = VSUB(T1b, T1c);
				   T1v = VADD(T1b, T1c);
				   T18 = VADD(Tp, Tq);
				   Tr = VSUB(Tp, Tq);
				   T1e = VADD(Te, Tf);
				   Tg = VSUB(Te, Tf);
				   T16 = VSUB(T14, T15);
				   T1s = VADD(T14, T15);
				   {
					V T1t, T19, T1w, T1g;
					T1t = VADD(T17, T18);
					T19 = VSUB(T17, T18);
					Ts = VSUB(To, Tr);
					TF = VADD(To, Tr);
					T1w = VADD(T1e, T1f);
					T1g = VSUB(T1e, T1f);
					Th = VADD(Td, Tg);
					TM = VSUB(Td, Tg);
					T1u = VADD(T1s, T1t);
					T1C = VSUB(T1s, T1t);
					T1n = VSUB(T16, T19);
					T1a = VADD(T16, T19);
					T1m = VSUB(T1d, T1g);
					T1h = VADD(T1d, T1g);
					T1x = VADD(T1v, T1w);
					T1D = VSUB(T1v, T1w);
				   }
//.........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:n2bv_20.c

示例3: n1fv_10

static void n1fv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     {
	  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(20, is), MAKE_VOLATILE_STRIDE(20, os)) {
	       V Tb, Tr, T3, Ts, T6, Tw, Tg, Tt, T9, Tc, T1, T2;
	       T1 = LD(&(xi[0]), ivs, &(xi[0]));
	       T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       {
		    V T4, T5, Te, Tf, T7, T8;
		    T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
		    T5 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
		    Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
		    T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
		    T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
		    Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
		    Tr = VADD(T1, T2);
		    T3 = VSUB(T1, T2);
		    Ts = VADD(T4, T5);
		    T6 = VSUB(T4, T5);
		    Tw = VADD(Te, Tf);
		    Tg = VSUB(Te, Tf);
		    Tt = VADD(T7, T8);
		    T9 = VSUB(T7, T8);
		    Tc = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
	       }
	       {
		    V TD, Tu, Tm, Ta, Td, Tv;
		    TD = VSUB(Ts, Tt);
		    Tu = VADD(Ts, Tt);
		    Tm = VSUB(T6, T9);
		    Ta = VADD(T6, T9);
		    Td = VSUB(Tb, Tc);
		    Tv = VADD(Tb, Tc);
		    {
			 V TC, Tx, Tn, Th;
			 TC = VSUB(Tv, Tw);
			 Tx = VADD(Tv, Tw);
			 Tn = VSUB(Td, Tg);
			 Th = VADD(Td, Tg);
			 {
			      V Ty, TA, TE, TG, Ti, Tk, To, Tq, Tz, Tj;
			      Ty = VADD(Tu, Tx);
			      TA = VSUB(Tu, Tx);
			      TE = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TD, TC));
			      TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TC, TD));
			      Ti = VADD(Ta, Th);
			      Tk = VSUB(Ta, Th);
			      To = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tn, Tm));
			      Tq = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tm, Tn));
			      Tz = VFNMS(LDK(KP250000000), Ty, Tr);
			      ST(&(xo[0]), VADD(Tr, Ty), ovs, &(xo[0]));
			      Tj = VFNMS(LDK(KP250000000), Ti, T3);
			      ST(&(xo[WS(os, 5)]), VADD(T3, Ti), ovs, &(xo[WS(os, 1)]));
			      {
				   V TB, TF, Tl, Tp;
				   TB = VFNMS(LDK(KP559016994), TA, Tz);
				   TF = VFMA(LDK(KP559016994), TA, Tz);
				   Tl = VFMA(LDK(KP559016994), Tk, Tj);
				   Tp = VFNMS(LDK(KP559016994), Tk, Tj);
				   ST(&(xo[WS(os, 4)]), VFMAI(TG, TF), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 6)]), VFNMSI(TG, TF), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 8)]), VFNMSI(TE, TB), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 2)]), VFMAI(TE, TB), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 3)]), VFNMSI(Tq, Tp), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 7)]), VFMAI(Tq, Tp), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 9)]), VFMAI(To, Tl), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 1)]), VFNMSI(To, Tl), ovs, &(xo[WS(os, 1)]));
			      }
			 }
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:barnex,项目名称:fftw,代码行数:85,代码来源:dft_simd_sse2_n1fv_10.c

示例4: n1bv_12

static void n1bv_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 = ii;
        xo = io;
        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 T1, T6, Tc, Th, Td, Te, Ti, Tz, T4, TA, T9, Tj, Tf, Tw;
            {
                V T2, T3, T7, T8;
                T1 = LD(&(xi[0]), ivs, &(xi[0]));
                T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
                T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
                T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
                T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
                T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
                Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
                Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
                Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
                Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
                Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
                Tz = VSUB(T2, T3);
                T4 = VADD(T2, T3);
                TA = VSUB(T7, T8);
                T9 = VADD(T7, T8);
                Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
            }
            Tf = VADD(Td, Te);
            Tw = VSUB(Td, Te);
            {
                V T5, Tp, TJ, TB, Ta, Tq, Tk, Tx, Tg, Ts;
                T5 = VADD(T1, T4);
                Tp = VFNMS(LDK(KP500000000), T4, T1);
                TJ = VSUB(Tz, TA);
                TB = VADD(Tz, TA);
                Ta = VADD(T6, T9);
                Tq = VFNMS(LDK(KP500000000), T9, T6);
                Tk = VADD(Ti, Tj);
                Tx = VSUB(Tj, Ti);
                Tg = VADD(Tc, Tf);
                Ts = VFNMS(LDK(KP500000000), Tf, Tc);
                {
                    V Tr, TF, Tb, Tn, TG, Ty, Tl, Tt;
                    Tr = VADD(Tp, Tq);
                    TF = VSUB(Tp, Tq);
                    Tb = VSUB(T5, Ta);
                    Tn = VADD(T5, Ta);
                    TG = VADD(Tw, Tx);
                    Ty = VSUB(Tw, Tx);
                    Tl = VADD(Th, Tk);
                    Tt = VFNMS(LDK(KP500000000), Tk, Th);
                    {
                        V TC, TE, TH, TL, Tu, TI, Tm, To;
                        TC = VMUL(LDK(KP866025403), VSUB(Ty, TB));
                        TE = VMUL(LDK(KP866025403), VADD(TB, Ty));
                        TH = VFNMS(LDK(KP866025403), TG, TF);
                        TL = VFMA(LDK(KP866025403), TG, TF);
                        Tu = VADD(Ts, Tt);
                        TI = VSUB(Ts, Tt);
                        Tm = VSUB(Tg, Tl);
                        To = VADD(Tg, Tl);
                        {
                            V TK, TM, Tv, TD;
                            TK = VFMA(LDK(KP866025403), TJ, TI);
                            TM = VFNMS(LDK(KP866025403), TJ, TI);
                            Tv = VSUB(Tr, Tu);
                            TD = VADD(Tr, Tu);
                            ST(&(xo[0]), VADD(Tn, To), ovs, &(xo[0]));
                            ST(&(xo[WS(os, 6)]), VSUB(Tn, To), ovs, &(xo[0]));
                            ST(&(xo[WS(os, 9)]), VFMAI(Tm, Tb), ovs, &(xo[WS(os, 1)]));
                            ST(&(xo[WS(os, 3)]), VFNMSI(Tm, Tb), ovs, &(xo[WS(os, 1)]));
                            ST(&(xo[WS(os, 5)]), VFMAI(TM, TL), ovs, &(xo[WS(os, 1)]));
                            ST(&(xo[WS(os, 7)]), VFNMSI(TM, TL), ovs, &(xo[WS(os, 1)]));
                            ST(&(xo[WS(os, 11)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)]));
                            ST(&(xo[WS(os, 1)]), VFMAI(TK, TH), ovs, &(xo[WS(os, 1)]));
                            ST(&(xo[WS(os, 8)]), VFNMSI(TE, TD), ovs, &(xo[0]));
                            ST(&(xo[WS(os, 4)]), VFMAI(TE, TD), ovs, &(xo[0]));
                            ST(&(xo[WS(os, 2)]), VFMAI(TC, Tv), ovs, &(xo[0]));
                            ST(&(xo[WS(os, 10)]), VFNMSI(TC, Tv), ovs, &(xo[0]));
                        }
                    }
                }
            }
        }
    }
    VLEAVE();
}
开发者ID:dpl0,项目名称:bioinformatics,代码行数:91,代码来源:n1bv_12.c

示例5: t1bv_12

static void t1bv_12(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     INT m;
     R *x;
     x = ii;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(rs)) {
	  V TI, Ti, TA, T7, Tm, TE, Tw, Tk, Tf, TB, TU, TM;
	  {
	       V T9, TK, Tj, TL, Te;
	       {
		    V T1, T4, T2, Tp, Tt, Tr;
		    T1 = LD(&(x[0]), ms, &(x[0]));
		    T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
		    T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
		    Tp = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
		    Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
		    Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
		    {
			 V T5, T3, Tq, Tu, Ts, Td, Tb, T8, Tc, Ta;
			 T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
			 Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
			 Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
			 T5 = BYTW(&(W[TWVL * 14]), T4);
			 T3 = BYTW(&(W[TWVL * 6]), T2);
			 Tq = BYTW(&(W[TWVL * 16]), Tp);
			 Tu = BYTW(&(W[TWVL * 8]), Tt);
			 Ts = BYTW(&(W[0]), Tr);
			 T9 = BYTW(&(W[TWVL * 10]), T8);
			 Td = BYTW(&(W[TWVL * 2]), Tc);
			 Tb = BYTW(&(W[TWVL * 18]), Ta);
			 {
			      V Th, T6, Tl, Tv;
			      Th = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			      TK = VSUB(T3, T5);
			      T6 = VADD(T3, T5);
			      Tl = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
			      Tv = VADD(Ts, Tu);
			      TI = VSUB(Tu, Ts);
			      Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
			      TL = VSUB(Tb, Td);
			      Te = VADD(Tb, Td);
			      Ti = BYTW(&(W[TWVL * 4]), Th);
			      TA = VFNMS(LDK(KP500000000), T6, T1);
			      T7 = VADD(T1, T6);
			      Tm = BYTW(&(W[TWVL * 20]), Tl);
			      TE = VFNMS(LDK(KP500000000), Tv, Tq);
			      Tw = VADD(Tq, Tv);
			 }
		    }
	       }
	       Tk = BYTW(&(W[TWVL * 12]), Tj);
	       Tf = VADD(T9, Te);
	       TB = VFNMS(LDK(KP500000000), Te, T9);
	       TU = VSUB(TK, TL);
	       TM = VADD(TK, TL);
	  }
	  {
	       V Tn, TH, TC, TQ, Ty, Tg;
	       Tn = VADD(Tk, Tm);
	       TH = VSUB(Tk, Tm);
	       TC = VADD(TA, TB);
	       TQ = VSUB(TA, TB);
	       Ty = VADD(T7, Tf);
	       Tg = VSUB(T7, Tf);
	       {
		    V To, TD, TJ, TR;
		    To = VADD(Ti, Tn);
		    TD = VFNMS(LDK(KP500000000), Tn, Ti);
		    TJ = VSUB(TH, TI);
		    TR = VADD(TH, TI);
		    {
			 V TP, TN, TW, TS, TO, TG, TX, TV;
			 {
			      V Tz, Tx, TF, TT;
			      Tz = VADD(To, Tw);
			      Tx = VSUB(To, Tw);
			      TF = VADD(TD, TE);
			      TT = VSUB(TD, TE);
			      TP = VMUL(LDK(KP866025403), VADD(TM, TJ));
			      TN = VMUL(LDK(KP866025403), VSUB(TJ, TM));
			      TW = VFMA(LDK(KP866025403), TR, TQ);
			      TS = VFNMS(LDK(KP866025403), TR, TQ);
			      ST(&(x[WS(rs, 6)]), VSUB(Ty, Tz), ms, &(x[0]));
			      ST(&(x[0]), VADD(Ty, Tz), ms, &(x[0]));
			      ST(&(x[WS(rs, 9)]), VFMAI(Tx, Tg), ms, &(x[WS(rs, 1)]));
			      ST(&(x[WS(rs, 3)]), VFNMSI(Tx, Tg), ms, &(x[WS(rs, 1)]));
			      TO = VADD(TC, TF);
			      TG = VSUB(TC, TF);
			      TX = VFNMS(LDK(KP866025403), TU, TT);
			      TV = VFMA(LDK(KP866025403), TU, TT);
			 }
			 ST(&(x[WS(rs, 8)]), VFNMSI(TP, TO), ms, &(x[0]));
			 ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0]));
			 ST(&(x[WS(rs, 2)]), VFMAI(TN, TG), ms, &(x[0]));
			 ST(&(x[WS(rs, 10)]), VFNMSI(TN, TG), ms, &(x[0]));
			 ST(&(x[WS(rs, 5)]), VFMAI(TX, TW), ms, &(x[WS(rs, 1)]));
			 ST(&(x[WS(rs, 7)]), VFNMSI(TX, TW), ms, &(x[WS(rs, 1)]));
			 ST(&(x[WS(rs, 11)]), VFNMSI(TV, TS), ms, &(x[WS(rs, 1)]));
//.........这里部分代码省略.........
开发者ID:phillipstanleymarbell,项目名称:sunflower-simulator,代码行数:101,代码来源:t1bv_12.c

示例6: t3bv_10

static void t3bv_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) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs)) {
	  V T1, T2, T3, Ti, T6, T7, TA, Tb, To;
	  T1 = LDW(&(W[0]));
	  T2 = LDW(&(W[TWVL * 2]));
	  T3 = VZMULJ(T1, T2);
	  Ti = VZMUL(T1, T2);
	  T6 = LDW(&(W[TWVL * 4]));
	  T7 = VZMULJ(T3, T6);
	  TA = VZMULJ(Ti, T6);
	  Tb = VZMULJ(T1, T6);
	  To = VZMULJ(T2, T6);
	  {
	       V TD, TQ, Tn, Tt, Tx, TM, TN, TS, Ta, Tg, Tw, TJ, TK, TR, Tz;
	       V TC, TB;
	       Tz = LD(&(x[0]), ms, &(x[0]));
	       TB = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
	       TC = VZMUL(TA, TB);
	       TD = VSUB(Tz, TC);
	       TQ = VADD(Tz, TC);
	       {
		    V Tk, Ts, Tm, Tq;
		    {
			 V Tj, Tr, Tl, Tp;
			 Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
			 Tk = VZMUL(Ti, Tj);
			 Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
			 Ts = VZMUL(T1, Tr);
			 Tl = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
			 Tm = VZMUL(T6, Tl);
			 Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
			 Tq = VZMUL(To, Tp);
		    }
		    Tn = VSUB(Tk, Tm);
		    Tt = VSUB(Tq, Ts);
		    Tx = VADD(Tn, Tt);
		    TM = VADD(Tk, Tm);
		    TN = VADD(Tq, Ts);
		    TS = VADD(TM, TN);
	       }
	       {
		    V T5, Tf, T9, Td;
		    {
			 V T4, Te, T8, Tc;
			 T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
			 T5 = VZMUL(T3, T4);
			 Te = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			 Tf = VZMUL(T2, Te);
			 T8 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
			 T9 = VZMUL(T7, T8);
			 Tc = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
			 Td = VZMUL(Tb, Tc);
		    }
		    Ta = VSUB(T5, T9);
		    Tg = VSUB(Td, Tf);
		    Tw = VADD(Ta, Tg);
		    TJ = VADD(T5, T9);
		    TK = VADD(Td, Tf);
		    TR = VADD(TJ, TK);
	       }
	       {
		    V Ty, TE, TF, Tv, TI, Th, Tu, TH, TG;
		    Ty = VMUL(LDK(KP559016994), VSUB(Tw, Tx));
		    TE = VADD(Tw, Tx);
		    TF = VFNMS(LDK(KP250000000), TE, TD);
		    Th = VSUB(Ta, Tg);
		    Tu = VSUB(Tn, Tt);
		    Tv = VBYI(VFMA(LDK(KP951056516), Th, VMUL(LDK(KP587785252), Tu)));
		    TI = VBYI(VFNMS(LDK(KP951056516), Tu, VMUL(LDK(KP587785252), Th)));
		    ST(&(x[WS(rs, 5)]), VADD(TD, TE), ms, &(x[WS(rs, 1)]));
		    TH = VSUB(TF, Ty);
		    ST(&(x[WS(rs, 3)]), VSUB(TH, TI), ms, &(x[WS(rs, 1)]));
		    ST(&(x[WS(rs, 7)]), VADD(TI, TH), ms, &(x[WS(rs, 1)]));
		    TG = VADD(Ty, TF);
		    ST(&(x[WS(rs, 1)]), VADD(Tv, TG), ms, &(x[WS(rs, 1)]));
		    ST(&(x[WS(rs, 9)]), VSUB(TG, Tv), ms, &(x[WS(rs, 1)]));
	       }
	       {
		    V TV, TT, TU, TP, TY, TL, TO, TX, TW;
		    TV = VMUL(LDK(KP559016994), VSUB(TR, TS));
		    TT = VADD(TR, TS);
		    TU = VFNMS(LDK(KP250000000), TT, TQ);
		    TL = VSUB(TJ, TK);
		    TO = VSUB(TM, TN);
		    TP = VBYI(VFNMS(LDK(KP951056516), TO, VMUL(LDK(KP587785252), TL)));
		    TY = VBYI(VFMA(LDK(KP951056516), TL, VMUL(LDK(KP587785252), TO)));
		    ST(&(x[0]), VADD(TQ, TT), ms, &(x[0]));
		    TX = VADD(TV, TU);
		    ST(&(x[WS(rs, 4)]), VSUB(TX, TY), ms, &(x[0]));
		    ST(&(x[WS(rs, 6)]), VADD(TY, TX), ms, &(x[0]));
		    TW = VSUB(TU, TV);
		    ST(&(x[WS(rs, 2)]), VADD(TP, TW), ms, &(x[0]));
//.........这里部分代码省略.........
开发者ID:BackupTheBerlios,项目名称:openvsipl,代码行数:101,代码来源:t3bv_10.c

示例7: t1fuv_7

static void t1fuv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
     DVK(KP801937735, +0.801937735804838252472204639014890102331838324);
     DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
     DVK(KP692021471, +0.692021471630095869627814897002069140197260599);
     DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
     DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
     INT m;
     R *x;
     x = ri;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(rs)) {
	  V T1, T2, T4, Te, Tc, T9, T7;
	  T1 = LD(&(x[0]), ms, &(x[0]));
	  T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
	  T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
	  Te = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
	  Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
	  T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
	  T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
	  {
	       V T3, T5, Tf, Td, Ta, T8;
	       T3 = BYTWJ(&(W[0]), T2);
	       T5 = BYTWJ(&(W[TWVL * 10]), T4);
	       Tf = BYTWJ(&(W[TWVL * 6]), Te);
	       Td = BYTWJ(&(W[TWVL * 4]), Tc);
	       Ta = BYTWJ(&(W[TWVL * 8]), T9);
	       T8 = BYTWJ(&(W[TWVL * 2]), T7);
	       {
		    V T6, Tk, Tg, Tl, Tb, Tm;
		    T6 = VADD(T3, T5);
		    Tk = VSUB(T5, T3);
		    Tg = VADD(Td, Tf);
		    Tl = VSUB(Tf, Td);
		    Tb = VADD(T8, Ta);
		    Tm = VSUB(Ta, T8);
		    {
			 V Th, Ts, Tp, Tu, Tn, Tx, Ti, Tt;
			 Th = VFNMS(LDK(KP356895867), T6, Tg);
			 Ts = VFMA(LDK(KP554958132), Tl, Tk);
			 ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0]));
			 Tp = VFNMS(LDK(KP356895867), Tb, T6);
			 Tu = VFNMS(LDK(KP356895867), Tg, Tb);
			 Tn = VFMA(LDK(KP554958132), Tm, Tl);
			 Tx = VFNMS(LDK(KP554958132), Tk, Tm);
			 Ti = VFNMS(LDK(KP692021471), Th, Tb);
			 Tt = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Ts, Tm));
			 {
			      V Tq, Tv, To, Ty, Tj, Tr, Tw;
			      Tq = VFNMS(LDK(KP692021471), Tp, Tg);
			      Tv = VFNMS(LDK(KP692021471), Tu, T6);
			      To = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tn, Tk));
			      Ty = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tx, Tl));
			      Tj = VFNMS(LDK(KP900968867), Ti, T1);
			      Tr = VFNMS(LDK(KP900968867), Tq, T1);
			      Tw = VFNMS(LDK(KP900968867), Tv, T1);
			      ST(&(x[WS(rs, 2)]), VFMAI(To, Tj), ms, &(x[0]));
			      ST(&(x[WS(rs, 5)]), VFNMSI(To, Tj), ms, &(x[WS(rs, 1)]));
			      ST(&(x[WS(rs, 1)]), VFMAI(Tt, Tr), ms, &(x[WS(rs, 1)]));
			      ST(&(x[WS(rs, 6)]), VFNMSI(Tt, Tr), ms, &(x[0]));
			      ST(&(x[WS(rs, 3)]), VFMAI(Ty, Tw), ms, &(x[WS(rs, 1)]));
			      ST(&(x[WS(rs, 4)]), VFNMSI(Ty, Tw), ms, &(x[0]));
			 }
		    }
	       }
	  }
     }
}
开发者ID:BackupTheBerlios,项目名称:openvsipl,代码行数:68,代码来源:t1fuv_7.c

示例8: 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(is), MAKE_VOLATILE_STRIDE(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]));
			      }
			 }
		    }
	       }
	  }
     }
}
开发者ID:phillipstanleymarbell,项目名称:sunflower-simulator,代码行数:100,代码来源:n1fv_9.c

示例9: INC

RETf INC( float &x, const __m128 y ) { __m128 t=ADD(LD(x),y); return STR(x,t); }
开发者ID:Claycau,项目名称:bigbrother,代码行数:1,代码来源:sse.hpp

示例10: t1sv_4

static void t1sv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     {
	  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 T1, Tv, T3, T6, T5, Ta, Td, Tc, Tg, Tj, Tt, T4, Tf, Ti, Tn;
	       V Tb, T2, T9;
	       T1 = LD(&(ri[0]), ms, &(ri[0]));
	       Tv = LD(&(ii[0]), ms, &(ii[0]));
	       T3 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0]));
	       T6 = LD(&(ii[WS(rs, 2)]), ms, &(ii[0]));
	       T2 = LDW(&(W[TWVL * 2]));
	       T5 = LDW(&(W[TWVL * 3]));
	       Ta = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)]));
	       Td = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)]));
	       T9 = LDW(&(W[0]));
	       Tc = LDW(&(W[TWVL * 1]));
	       Tg = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)]));
	       Tj = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)]));
	       Tt = VMUL(T2, T6);
	       T4 = VMUL(T2, T3);
	       Tf = LDW(&(W[TWVL * 4]));
	       Ti = LDW(&(W[TWVL * 5]));
	       Tn = VMUL(T9, Td);
	       Tb = VMUL(T9, Ta);
	       {
		    V Tu, T7, Tp, Th, To, Te;
		    Tu = VFNMS(T5, T3, Tt);
		    T7 = VFMA(T5, T6, T4);
		    Tp = VMUL(Tf, Tj);
		    Th = VMUL(Tf, Tg);
		    To = VFNMS(Tc, Ta, Tn);
		    Te = VFMA(Tc, Td, Tb);
		    {
			 V Tw, Tx, T8, Tm, Tq, Tk;
			 Tw = VADD(Tu, Tv);
			 Tx = VSUB(Tv, Tu);
			 T8 = VADD(T1, T7);
			 Tm = VSUB(T1, T7);
			 Tq = VFNMS(Ti, Tg, Tp);
			 Tk = VFMA(Ti, Tj, Th);
			 {
			      V Ts, Tr, Tl, Ty;
			      Ts = VADD(To, Tq);
			      Tr = VSUB(To, Tq);
			      Tl = VADD(Te, Tk);
			      Ty = VSUB(Te, Tk);
			      ST(&(ri[WS(rs, 1)]), VADD(Tm, Tr), ms, &(ri[WS(rs, 1)]));
			      ST(&(ri[WS(rs, 3)]), VSUB(Tm, Tr), ms, &(ri[WS(rs, 1)]));
			      ST(&(ii[WS(rs, 2)]), VSUB(Tw, Ts), ms, &(ii[0]));
			      ST(&(ii[0]), VADD(Ts, Tw), ms, &(ii[0]));
			      ST(&(ii[WS(rs, 3)]), VADD(Ty, Tx), ms, &(ii[WS(rs, 1)]));
			      ST(&(ii[WS(rs, 1)]), VSUB(Tx, Ty), ms, &(ii[WS(rs, 1)]));
			      ST(&(ri[0]), VADD(T8, Tl), ms, &(ri[0]));
			      ST(&(ri[WS(rs, 2)]), VSUB(T8, Tl), ms, &(ri[0]));
			 }
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:dstuck,项目名称:tinker_integrated_PIMC,代码行数:62,代码来源:t1sv_4.c

示例11: STDU

SprxPatch emulator_api_patches[] =
{
	// Read umd patches
	{ psp_read, STDU(SP, 0xFF90, SP), &condition_psp_iso },
	{ psp_read+4, MFLR(R0), &condition_psp_iso },
	{ psp_read+8, STD(R0, 0x80, SP), &condition_psp_iso },
	{ psp_read+0x0C, MR(R8, R7), &condition_psp_iso },
	{ psp_read+0x10, MR(R7, R6), &condition_psp_iso },
	{ psp_read+0x14, MR(R6, R5), &condition_psp_iso },
	{ psp_read+0x18, MR(R5, R4), &condition_psp_iso },
	{ psp_read+0x1C, MR(R4, R3), &condition_psp_iso },
	{ psp_read+0x20, LI(R3, SYSCALL8_OPCODE_READ_PSP_UMD), &condition_psp_iso },
	{ psp_read+0x24, LI(R11, 8), &condition_psp_iso },
	{ psp_read+0x28, SC, &condition_psp_iso },
	{ psp_read+0x2C, LD(R0, 0x80, SP), &condition_psp_iso },
	{ psp_read+0x30, MTLR(R0), &condition_psp_iso },
	{ psp_read+0x34, ADDI(SP, SP, 0x70), &condition_psp_iso },
	{ psp_read+0x38, BLR, &condition_psp_iso },

	// Read header patches
	{ psp_read+0x3C, STDU(SP, 0xFF90, SP), &condition_psp_iso },
	{ psp_read+0x40, MFLR(R0), &condition_psp_iso },
	{ psp_read+0x44, STD(R0, 0x80, SP), &condition_psp_iso },
	{ psp_read+0x48, MR(R7, R6), &condition_psp_iso },
	{ psp_read+0x4C, MR(R6, R5), &condition_psp_iso },
	{ psp_read+0x50, MR(R5, R4), &condition_psp_iso },
	{ psp_read+0x54, MR(R4, R3), &condition_psp_iso },
	{ psp_read+0x58, LI(R3, SYSCALL8_OPCODE_READ_PSP_HEADER), &condition_psp_iso },
	{ psp_read+0x5C, LI(R11, 8), &condition_psp_iso },
	{ psp_read+0x60, SC, &condition_psp_iso },
开发者ID:Joonie86,项目名称:COBRA-7.3,代码行数:30,代码来源:modulespatch.c

示例12: t2sv_4

static void t2sv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     {
	  INT m;
	  for (m = mb, W = W + (mb * 4); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 4), MAKE_VOLATILE_STRIDE(8, rs)) {
	       V T2, T6, T3, T5, T1, Tx, T8, Tc, Tf, Ta, T4, Th, Tj, Tl;
	       T2 = LDW(&(W[0]));
	       T6 = LDW(&(W[TWVL * 3]));
	       T3 = LDW(&(W[TWVL * 2]));
	       T5 = LDW(&(W[TWVL * 1]));
	       T1 = LD(&(ri[0]), ms, &(ri[0]));
	       Tx = LD(&(ii[0]), ms, &(ii[0]));
	       T8 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0]));
	       Tc = LD(&(ii[WS(rs, 2)]), ms, &(ii[0]));
	       Tf = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)]));
	       Ta = VMUL(T2, T6);
	       T4 = VMUL(T2, T3);
	       Th = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)]));
	       Tj = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)]));
	       Tl = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)]));
	       {
		    V Tg, Tb, T7, Tp, Tk, Tr, Ti;
		    Tg = VMUL(T2, Tf);
		    Tb = VFNMS(T5, T3, Ta);
		    T7 = VFMA(T5, T6, T4);
		    Tp = VMUL(T2, Th);
		    Tk = VMUL(T3, Tj);
		    Tr = VMUL(T3, Tl);
		    Ti = VFMA(T5, Th, Tg);
		    {
			 V Tv, T9, Tq, Tm, Ts, Tw, Td;
			 Tv = VMUL(T7, Tc);
			 T9 = VMUL(T7, T8);
			 Tq = VFNMS(T5, Tf, Tp);
			 Tm = VFMA(T6, Tl, Tk);
			 Ts = VFNMS(T6, Tj, Tr);
			 Tw = VFNMS(Tb, T8, Tv);
			 Td = VFMA(Tb, Tc, T9);
			 {
			      V Tn, TA, Tu, Tt;
			      Tn = VADD(Ti, Tm);
			      TA = VSUB(Ti, Tm);
			      Tu = VADD(Tq, Ts);
			      Tt = VSUB(Tq, Ts);
			      {
				   V Ty, Tz, Te, To;
				   Ty = VADD(Tw, Tx);
				   Tz = VSUB(Tx, Tw);
				   Te = VADD(T1, Td);
				   To = VSUB(T1, Td);
				   ST(&(ii[WS(rs, 3)]), VADD(TA, Tz), ms, &(ii[WS(rs, 1)]));
				   ST(&(ii[WS(rs, 1)]), VSUB(Tz, TA), ms, &(ii[WS(rs, 1)]));
				   ST(&(ii[WS(rs, 2)]), VSUB(Ty, Tu), ms, &(ii[0]));
				   ST(&(ii[0]), VADD(Tu, Ty), ms, &(ii[0]));
				   ST(&(ri[WS(rs, 1)]), VADD(To, Tt), ms, &(ri[WS(rs, 1)]));
				   ST(&(ri[WS(rs, 3)]), VSUB(To, Tt), ms, &(ri[WS(rs, 1)]));
				   ST(&(ri[0]), VADD(Te, Tn), ms, &(ri[0]));
				   ST(&(ri[WS(rs, 2)]), VSUB(Te, Tn), ms, &(ri[0]));
			      }
			 }
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:SKA-ScienceDataProcessor,项目名称:FastImaging,代码行数:66,代码来源:t2sv_4.c

示例13: n2fv_13

static void n2fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
     DVK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
     DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
     DVK(KP075902986, +0.075902986037193865983102897245103540356428373);
     DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
     DVK(KP132983124, +0.132983124607418643793760531921092974399165133);
     DVK(KP258260390, +0.258260390311744861420450644284508567852516811);
     DVK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
     DVK(KP300238635, +0.300238635966332641462884626667381504676006424);
     DVK(KP011599105, +0.011599105605768290721655456654083252189827041);
     DVK(KP156891391, +0.156891391051584611046832726756003269660212636);
     DVK(KP256247671, +0.256247671582936600958684654061725059144125175);
     DVK(KP174138601, +0.174138601152135905005660794929264742616964676);
     DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
     DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
     DVK(KP113854479, +0.113854479055790798974654345867655310534642560);
     DVK(KP265966249, +0.265966249214837287587521063842185948798330267);
     DVK(KP387390585, +0.387390585467617292130675966426762851778775217);
     DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     int i;
     const R *xi;
     R *xo;
     xi = ri;
     xo = ro;
     BEGIN_SIMD();
     for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
	  V TW, Tb, Tm, Tu, TC, TR, TX, TK, TU, Tz, TB, TN, TT;
	  TW = LD(&(xi[0]), ivs, &(xi[0]));
	  {
	       V T3, TH, Tl, Tw, Tp, Tg, Tv, To, T6, Tr, T9, Ts, Ta, TI, T1;
	       V T2, Tq, Tt;
	       T1 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
	       T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       T3 = VSUB(T1, T2);
	       TH = VADD(T1, T2);
	       {
		    V Th, Ti, Tj, Tk;
		    Th = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
		    Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
		    Tj = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
		    Tk = VADD(Ti, Tj);
		    Tl = VADD(Th, Tk);
		    Tw = VSUB(Ti, Tj);
		    Tp = VFNMS(LDK(KP500000000), Tk, Th);
	       }
	       {
		    V Tc, Td, Te, Tf;
		    Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
		    Td = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
		    Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
		    Tf = VADD(Td, Te);
		    Tg = VADD(Tc, Tf);
		    Tv = VSUB(Td, Te);
		    To = VFNMS(LDK(KP500000000), Tf, Tc);
	       }
	       {
		    V T4, T5, T7, T8;
		    T4 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
		    T5 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
		    T6 = VSUB(T4, T5);
		    Tr = VADD(T4, T5);
		    T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
		    T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
		    T9 = VSUB(T7, T8);
		    Ts = VADD(T7, T8);
	       }
	       Ta = VADD(T6, T9);
	       TI = VADD(Tr, Ts);
	       Tb = VADD(T3, Ta);
	       Tm = VSUB(Tg, Tl);
	       Tq = VSUB(To, Tp);
	       Tt = VMUL(LDK(KP866025403), VSUB(Tr, Ts));
	       Tu = VADD(Tq, Tt);
	       TC = VSUB(Tq, Tt);
	       {
		    V TP, TQ, TG, TJ;
		    TP = VADD(Tg, Tl);
		    TQ = VADD(TH, TI);
		    TR = VMUL(LDK(KP300462606), VSUB(TP, TQ));
		    TX = VADD(TP, TQ);
		    TG = VADD(To, Tp);
		    TJ = VFNMS(LDK(KP500000000), TI, TH);
		    TK = VSUB(TG, TJ);
		    TU = VADD(TG, TJ);
	       }
	       {
		    V Tx, Ty, TL, TM;
		    Tx = VMUL(LDK(KP866025403), VSUB(Tv, Tw));
		    Ty = VFNMS(LDK(KP500000000), Ta, T3);
		    Tz = VSUB(Tx, Ty);
		    TB = VADD(Tx, Ty);
		    TL = VADD(Tv, Tw);
		    TM = VSUB(T6, T9);
		    TN = VSUB(TL, TM);
		    TT = VADD(TL, TM);
	       }
	  }
//.........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:n2fv_13.c

示例14: DEC

RETf DEC( float &x, const __m128 y ) { __m128 t=SUB(LD(x),y); return STR(x,t); }
开发者ID:Claycau,项目名称:bigbrother,代码行数:1,代码来源:sse.hpp

示例15: n1bv_7

static void n1bv_7(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(KP692021471, +0.692021471630095869627814897002069140197260599);
     DVK(KP801937735, +0.801937735804838252472204639014890102331838324);
     DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
     DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
     DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
     {
	  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 T1, T2, T3, T8, T9, T5, T6;
	       T1 = LD(&(xi[0]), ivs, &(xi[0]));
	       T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
	       T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
	       T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
	       T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
	       T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
	       T6 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
	       {
		    V Tg, T4, Te, Ta, Tf, T7;
		    Tg = VSUB(T2, T3);
		    T4 = VADD(T2, T3);
		    Te = VSUB(T8, T9);
		    Ta = VADD(T8, T9);
		    Tf = VSUB(T5, T6);
		    T7 = VADD(T5, T6);
		    {
			 V Tr, Tj, Tm, Th, To, Tb;
			 Tr = VFMA(LDK(KP554958132), Te, Tg);
			 Tj = VFNMS(LDK(KP356895867), T4, Ta);
			 Tm = VFMA(LDK(KP554958132), Tf, Te);
			 Th = VFNMS(LDK(KP554958132), Tg, Tf);
			 ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, Ta))), ovs, &(xo[0]));
			 To = VFNMS(LDK(KP356895867), T7, T4);
			 Tb = VFNMS(LDK(KP356895867), Ta, T7);
			 {
			      V Ts, Tk, Tn, Ti;
			      Ts = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tr, Tf));
			      Tk = VFNMS(LDK(KP692021471), Tj, T7);
			      Tn = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tm, Tg));
			      Ti = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Th, Te));
			      {
				   V Tp, Tc, Tl, Tq, Td;
				   Tp = VFNMS(LDK(KP692021471), To, Ta);
				   Tc = VFNMS(LDK(KP692021471), Tb, T4);
				   Tl = VFNMS(LDK(KP900968867), Tk, T1);
				   Tq = VFNMS(LDK(KP900968867), Tp, T1);
				   Td = VFNMS(LDK(KP900968867), Tc, T1);
				   ST(&(xo[WS(os, 5)]), VFNMSI(Tn, Tl), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 2)]), VFMAI(Tn, Tl), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 6)]), VFNMSI(Ts, Tq), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 1)]), VFMAI(Ts, Tq), ovs, &(xo[WS(os, 1)]));
				   ST(&(xo[WS(os, 4)]), VFNMSI(Ti, Td), ovs, &(xo[0]));
				   ST(&(xo[WS(os, 3)]), VFMAI(Ti, Td), ovs, &(xo[WS(os, 1)]));
			      }
			 }
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:AlexanderB1290,项目名称:scafacos,代码行数:67,代码来源:n1bv_7.c


注:本文中的LD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。