当前位置: 首页>>代码示例>>C++>>正文

C++ LDW函数代码示例

本文整理汇总了C++中LDW函数的典型用法代码示例。如果您正苦于以下问题：C++ LDW函数的具体用法？C++ LDW怎么用？C++ LDW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LDW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: t3fv_4

static void t3fv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     INT m;
     R *x;
     x = ri;
     for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(rs)) {
	  V T2, T3, T4;
	  T2 = LDW(&(W[0]));
	  T3 = LDW(&(W[TWVL * 2]));
	  T4 = VZMULJ(T2, T3);
	  {
	       V T1, Tb, T6, T9, Ta, T5, T8;
	       T1 = LD(&(x[0]), ms, &(x[0]));
	       Ta = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
	       Tb = VZMULJ(T3, Ta);
	       T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
	       T6 = VZMULJ(T4, T5);
	       T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
	       T9 = VZMULJ(T2, T8);
	       {
		    V T7, Tc, Td, Te;
		    T7 = VSUB(T1, T6);
		    Tc = VBYI(VSUB(T9, Tb));
		    ST(&(x[WS(rs, 1)]), VSUB(T7, Tc), ms, &(x[WS(rs, 1)]));
		    ST(&(x[WS(rs, 3)]), VADD(T7, Tc), ms, &(x[WS(rs, 1)]));
		    Td = VADD(T1, T6);
		    Te = VADD(T9, Tb);
		    ST(&(x[WS(rs, 2)]), VSUB(Td, Te), ms, &(x[0]));
		    ST(&(x[0]), VADD(Td, Te), ms, &(x[0]));
	       }
	  }
     }
}
开发者ID:BackupTheBerlios，项目名称:openvsipl，代码行数:33，代码来源:t3fv_4.c

示例2: hc2cfdftv_4

static void hc2cfdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     {
	  INT m;
	  for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(16, rs)) {
	       V T1, T2, Tb, T5, T6, T4, T9, T3, Tc, T7, Ta, Tg, T8, Td, Th;
	       V Tf, Te, Ti, Tj;
	       T1 = LD(&(Rp[0]), ms, &(Rp[0]));
	       T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
	       Tb = LDW(&(W[0]));
	       T5 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
	       T6 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
	       T4 = LDW(&(W[TWVL * 2]));
	       T9 = LDW(&(W[TWVL * 4]));
	       T3 = VFMACONJ(T2, T1);
	       Tc = VZMULIJ(Tb, VFNMSCONJ(T2, T1));
	       T7 = VZMULJ(T4, VFMACONJ(T6, T5));
	       Ta = VZMULIJ(T9, VFNMSCONJ(T6, T5));
	       Tg = VADD(T3, T7);
	       T8 = VSUB(T3, T7);
	       Td = VSUB(Ta, Tc);
	       Th = VADD(Tc, Ta);
	       Tf = VCONJ(VMUL(LDK(KP500000000), VFMAI(Td, T8)));
	       Te = VMUL(LDK(KP500000000), VFNMSI(Td, T8));
	       Ti = VMUL(LDK(KP500000000), VSUB(Tg, Th));
	       Tj = VCONJ(VMUL(LDK(KP500000000), VADD(Th, Tg)));
	       ST(&(Rm[0]), Tf, -ms, &(Rm[0]));
	       ST(&(Rp[WS(rs, 1)]), Te, ms, &(Rp[WS(rs, 1)]));
	       ST(&(Rp[0]), Ti, ms, &(Rp[0]));
	       ST(&(Rm[WS(rs, 1)]), Tj, -ms, &(Rm[WS(rs, 1)]));
	  }
     }
     VLEAVE();
}
开发者ID:LiZimo，项目名称:interactiveCart，代码行数:35，代码来源:hc2cfdftv_4.c

示例3: t1sv_2

static void t1sv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     {
	  INT m;
	  for (m = mb, W = W + (mb * 2); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 2), MAKE_VOLATILE_STRIDE(4, rs)) {
	       V T1, T8, T6, T7;
	       T1 = LD(&(ri[0]), ms, &(ri[0]));
	       T8 = LD(&(ii[0]), ms, &(ii[0]));
	       {
		    V T3, T5, T2, T4;
		    T3 = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)]));
		    T5 = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)]));
		    T2 = LDW(&(W[0]));
		    T4 = LDW(&(W[TWVL * 1]));
		    T6 = VFMA(T2, T3, VMUL(T4, T5));
		    T7 = VFNMS(T4, T3, VMUL(T2, T5));
	       }
	       ST(&(ri[WS(rs, 1)]), VSUB(T1, T6), ms, &(ri[WS(rs, 1)]));
	       ST(&(ii[WS(rs, 1)]), VSUB(T8, T7), ms, &(ii[WS(rs, 1)]));
	       ST(&(ri[0]), VADD(T1, T6), ms, &(ri[0]));
	       ST(&(ii[0]), VADD(T7, T8), ms, &(ii[0]));
	  }
     }
     VLEAVE();
}
开发者ID:SKA-ScienceDataProcessor，项目名称:FastImaging，代码行数:25，代码来源:t1sv_2.c

示例4: hc2cfdftv_4

static void hc2cfdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     INT m;
     for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs)) {
	  V T4, Tc, T9, Te, T1, T3, T2, Tb, T6, T8, T7, T5, Td, Tg, Th;
	  V Ta, Tf, Tk, Tl, Ti, Tj;
	  T1 = LD(&(Rp[0]), ms, &(Rp[0]));
	  T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
	  T3 = VCONJ(T2);
	  T4 = VADD(T1, T3);
	  Tb = LDW(&(W[0]));
	  Tc = VZMULIJ(Tb, VSUB(T3, T1));
	  T6 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
	  T7 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
	  T8 = VCONJ(T7);
	  T5 = LDW(&(W[TWVL * 2]));
	  T9 = VZMULJ(T5, VADD(T6, T8));
	  Td = LDW(&(W[TWVL * 4]));
	  Te = VZMULIJ(Td, VSUB(T8, T6));
	  Ta = VSUB(T4, T9);
	  Tf = VBYI(VSUB(Tc, Te));
	  Tg = VMUL(LDK(KP500000000), VSUB(Ta, Tf));
	  Th = VCONJ(VMUL(LDK(KP500000000), VADD(Ta, Tf)));
	  ST(&(Rp[WS(rs, 1)]), Tg, ms, &(Rp[WS(rs, 1)]));
	  ST(&(Rm[0]), Th, -ms, &(Rm[0]));
	  Ti = VADD(T4, T9);
	  Tj = VADD(Tc, Te);
	  Tk = VCONJ(VMUL(LDK(KP500000000), VSUB(Ti, Tj)));
	  Tl = VMUL(LDK(KP500000000), VADD(Ti, Tj));
	  ST(&(Rm[WS(rs, 1)]), Tk, -ms, &(Rm[WS(rs, 1)]));
	  ST(&(Rp[0]), Tl, ms, &(Rp[0]));
     }
}
开发者ID:BackupTheBerlios，项目名称:openvsipl，代码行数:34，代码来源:hc2cfdftv_4.c

示例5: hc2cbdftv_4

static void hc2cbdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     {
	  INT m;
	  for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs)) {
	       V T2, T3, T5, T6, Tf, T1, T9, Ta, T4, Tb, T7, Tc, Th, T8, Tg;
	       V Te, Td, Ti, Tj;
	       T2 = LD(&(Rp[0]), ms, &(Rp[0]));
	       T3 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
	       T5 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
	       T6 = LD(&(Rm[0]), -ms, &(Rm[0]));
	       Tf = LDW(&(W[0]));
	       T1 = LDW(&(W[TWVL * 4]));
	       T9 = LDW(&(W[TWVL * 2]));
	       Ta = VFMACONJ(T3, T2);
	       T4 = VFNMSCONJ(T3, T2);
	       Tb = VFMACONJ(T6, T5);
	       T7 = VFNMSCONJ(T6, T5);
	       Tc = VZMUL(T9, VSUB(Ta, Tb));
	       Th = VADD(Ta, Tb);
	       T8 = VZMULI(T1, VFNMSI(T7, T4));
	       Tg = VZMULI(Tf, VFMAI(T7, T4));
	       Te = VCONJ(VSUB(Tc, T8));
	       Td = VADD(T8, Tc);
	       Ti = VADD(Tg, Th);
	       Tj = VCONJ(VSUB(Th, Tg));
	       ST(&(Rm[WS(rs, 1)]), Te, -ms, &(Rm[WS(rs, 1)]));
	       ST(&(Rp[WS(rs, 1)]), Td, ms, &(Rp[WS(rs, 1)]));
	       ST(&(Rp[0]), Ti, ms, &(Rp[0]));
	       ST(&(Rm[0]), Tj, -ms, &(Rm[0]));
	  }
     }
     VLEAVE();
}
开发者ID:BGCX261，项目名称:zpr-rozpoznawanie-dzwieku-git，代码行数:34，代码来源:hc2cbdftv_4.c

示例6: LD

static const R *t1sv_2(R *ri, R *ii, const R *W, stride ios, INT m, INT dist)
{
     INT i;
     for (i = m; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * dist), ii = ii + ((2 * VL) * dist), W = W + ((2 * VL) * 2), MAKE_VOLATILE_STRIDE(ios)) {
	  V T1, Ta, T3, T6, T2, T5;
	  T1 = LD(&(ri[0]), dist, &(ri[0]));
	  Ta = LD(&(ii[0]), dist, &(ii[0]));
	  T3 = LD(&(ri[WS(ios, 1)]), dist, &(ri[WS(ios, 1)]));
	  T6 = LD(&(ii[WS(ios, 1)]), dist, &(ii[WS(ios, 1)]));
	  T2 = LDW(&(W[0]));
	  T5 = LDW(&(W[TWVL * 1]));
	  {
	       V T8, T4, T9, T7;
	       T8 = VMUL(T2, T6);
	       T4 = VMUL(T2, T3);
	       T9 = VFNMS(T5, T3, T8);
	       T7 = VFMA(T5, T6, T4);
	       ST(&(ii[0]), VADD(T9, Ta), dist, &(ii[0]));
	       ST(&(ii[WS(ios, 1)]), VSUB(Ta, T9), dist, &(ii[WS(ios, 1)]));
	       ST(&(ri[0]), VADD(T1, T7), dist, &(ri[0]));
	       ST(&(ri[WS(ios, 1)]), VSUB(T1, T7), dist, &(ri[WS(ios, 1)]));
	  }
     }
     return W;
}
开发者ID:exic，项目名称:last.fm-dbus，代码行数:25，代码来源:t1sv_2.c

示例7: hc2cbdftv_4

static void hc2cbdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     INT m;
     for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs)) {
	  V T5, Tc, T9, Td, T2, T4, T3, T6, T8, T7, Tj, Ti, Th, Tk, Tl;
	  V Ta, Te, T1, Tb, Tf, Tg;
	  T2 = LD(&(Rp[0]), ms, &(Rp[0]));
	  T3 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
	  T4 = VCONJ(T3);
	  T5 = VSUB(T2, T4);
	  Tc = VADD(T2, T4);
	  T6 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
	  T7 = LD(&(Rm[0]), -ms, &(Rm[0]));
	  T8 = VCONJ(T7);
	  T9 = VBYI(VSUB(T6, T8));
	  Td = VADD(T6, T8);
	  Tj = VADD(Tc, Td);
	  Th = LDW(&(W[0]));
	  Ti = VZMULI(Th, VADD(T5, T9));
	  Tk = VADD(Ti, Tj);
	  ST(&(Rp[0]), Tk, ms, &(Rp[0]));
	  Tl = VCONJ(VSUB(Tj, Ti));
	  ST(&(Rm[0]), Tl, -ms, &(Rm[0]));
	  T1 = LDW(&(W[TWVL * 4]));
	  Ta = VZMULI(T1, VSUB(T5, T9));
	  Tb = LDW(&(W[TWVL * 2]));
	  Te = VZMUL(Tb, VSUB(Tc, Td));
	  Tf = VADD(Ta, Te);
	  ST(&(Rp[WS(rs, 1)]), Tf, ms, &(Rp[WS(rs, 1)]));
	  Tg = VCONJ(VSUB(Te, Ta));
	  ST(&(Rm[WS(rs, 1)]), Tg, -ms, &(Rm[WS(rs, 1)]));
     }
}
开发者ID:BackupTheBerlios，项目名称:openvsipl，代码行数:33，代码来源:hc2cbdftv_4.c

示例8: hc2cfdftv_6

static void hc2cfdftv_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     {
	  INT m;
	  for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 10)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(24, rs)) {
	       V Ta, Tu, Tn, Tw, Ti, Tv, T1, T8, Tg, Tf, T7, T3, Te, T6, T2;
	       V T4, T9, T5, Tk, Tm, Tj, Tl, Tc, Th, Tb, Td, Tr, Tp, Tq, To;
	       V Tt, Ts, TA, Ty, Tz, Tx, TC, TB;
	       T1 = LD(&(Rp[0]), ms, &(Rp[0]));
	       T8 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
	       Tg = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0]));
	       Te = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0]));
	       Tf = VCONJ(Te);
	       T6 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
	       T7 = VCONJ(T6);
	       T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
	       T3 = VCONJ(T2);
	       T4 = VADD(T1, T3);
	       T5 = LDW(&(W[TWVL * 4]));
	       T9 = VZMULIJ(T5, VSUB(T7, T8));
	       Ta = VADD(T4, T9);
	       Tu = VSUB(T4, T9);
	       Tj = LDW(&(W[0]));
	       Tk = VZMULIJ(Tj, VSUB(T3, T1));
	       Tl = LDW(&(W[TWVL * 6]));
	       Tm = VZMULJ(Tl, VADD(Tf, Tg));
	       Tn = VADD(Tk, Tm);
	       Tw = VSUB(Tm, Tk);
	       Tb = LDW(&(W[TWVL * 2]));
	       Tc = VZMULJ(Tb, VADD(T7, T8));
	       Td = LDW(&(W[TWVL * 8]));
	       Th = VZMULIJ(Td, VSUB(Tf, Tg));
	       Ti = VADD(Tc, Th);
	       Tv = VSUB(Tc, Th);
	       Tr = VMUL(LDK(KP500000000), VBYI(VMUL(LDK(KP866025403), VSUB(Tn, Ti))));
	       To = VADD(Ti, Tn);
	       Tp = VMUL(LDK(KP500000000), VADD(Ta, To));
	       Tq = VFNMS(LDK(KP250000000), To, VMUL(LDK(KP500000000), Ta));
	       ST(&(Rp[0]), Tp, ms, &(Rp[0]));
	       Tt = VCONJ(VADD(Tq, Tr));
	       ST(&(Rm[WS(rs, 1)]), Tt, -ms, &(Rm[WS(rs, 1)]));
	       Ts = VSUB(Tq, Tr);
	       ST(&(Rp[WS(rs, 2)]), Ts, ms, &(Rp[0]));
	       TA = VMUL(LDK(KP500000000), VBYI(VMUL(LDK(KP866025403), VSUB(Tw, Tv))));
	       Tx = VADD(Tv, Tw);
	       Ty = VCONJ(VMUL(LDK(KP500000000), VADD(Tu, Tx)));
	       Tz = VFNMS(LDK(KP250000000), Tx, VMUL(LDK(KP500000000), Tu));
	       ST(&(Rm[WS(rs, 2)]), Ty, -ms, &(Rm[0]));
	       TC = VADD(Tz, TA);
	       ST(&(Rp[WS(rs, 1)]), TC, ms, &(Rp[WS(rs, 1)]));
	       TB = VCONJ(VSUB(Tz, TA));
	       ST(&(Rm[0]), TB, -ms, &(Rm[0]));
	  }
     }
     VLEAVE();
}
开发者ID:LiZimo，项目名称:interactiveCart，代码行数:59，代码来源:hc2cfdftv_6.c

示例9: 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, T4, T3, T5, T6, T8;
	       T2 = LDW(&(W[0]));
	       T4 = LDW(&(W[TWVL * 1]));
	       T3 = LDW(&(W[TWVL * 2]));
	       T5 = LDW(&(W[TWVL * 3]));
	       T6 = VFMA(T2, T3, VMUL(T4, T5));
	       T8 = VFNMS(T4, T3, VMUL(T2, T5));
	       {
		    V T1, Tp, Ta, To, Te, Tk, Th, Tl, T7, T9;
		    T1 = LD(&(ri[0]), ms, &(ri[0]));
		    Tp = LD(&(ii[0]), ms, &(ii[0]));
		    T7 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0]));
		    T9 = LD(&(ii[WS(rs, 2)]), ms, &(ii[0]));
		    Ta = VFMA(T6, T7, VMUL(T8, T9));
		    To = VFNMS(T8, T7, VMUL(T6, T9));
		    {
			 V Tc, Td, Tf, Tg;
			 Tc = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)]));
			 Td = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)]));
			 Te = VFMA(T2, Tc, VMUL(T4, Td));
			 Tk = VFNMS(T4, Tc, VMUL(T2, Td));
			 Tf = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)]));
			 Tg = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)]));
			 Th = VFMA(T3, Tf, VMUL(T5, Tg));
			 Tl = VFNMS(T5, Tf, VMUL(T3, Tg));
		    }
		    {
			 V Tb, Ti, Tn, Tq;
			 Tb = VADD(T1, Ta);
			 Ti = VADD(Te, Th);
			 ST(&(ri[WS(rs, 2)]), VSUB(Tb, Ti), ms, &(ri[0]));
			 ST(&(ri[0]), VADD(Tb, Ti), ms, &(ri[0]));
			 Tn = VADD(Tk, Tl);
			 Tq = VADD(To, Tp);
			 ST(&(ii[0]), VADD(Tn, Tq), ms, &(ii[0]));
			 ST(&(ii[WS(rs, 2)]), VSUB(Tq, Tn), ms, &(ii[0]));
		    }
		    {
			 V Tj, Tm, Tr, Ts;
			 Tj = VSUB(T1, Ta);
			 Tm = VSUB(Tk, Tl);
			 ST(&(ri[WS(rs, 3)]), VSUB(Tj, Tm), ms, &(ri[WS(rs, 1)]));
			 ST(&(ri[WS(rs, 1)]), VADD(Tj, Tm), ms, &(ri[WS(rs, 1)]));
			 Tr = VSUB(Tp, To);
			 Ts = VSUB(Te, Th);
			 ST(&(ii[WS(rs, 1)]), VSUB(Tr, Ts), ms, &(ii[WS(rs, 1)]));
			 ST(&(ii[WS(rs, 3)]), VADD(Ts, Tr), ms, &(ii[WS(rs, 1)]));
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:SKA-ScienceDataProcessor，项目名称:FastImaging，代码行数:58，代码来源:t2sv_4.c

示例10: t3fv_5

static void t3fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     {
	  INT m;
	  R *x;
	  x = ri;
	  for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(5, rs)) {
	       V T1, T4, T5, T9;
	       T1 = LDW(&(W[0]));
	       T4 = LDW(&(W[TWVL * 2]));
	       T5 = VZMUL(T1, T4);
	       T9 = VZMULJ(T1, T4);
	       {
		    V Tg, Tk, Tl, T8, Te, Th;
		    Tg = LD(&(x[0]), ms, &(x[0]));
		    {
			 V T3, Td, T7, Tb;
			 {
			      V T2, Tc, T6, Ta;
			      T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
			      T3 = VZMULJ(T1, T2);
			      Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
			      Td = VZMULJ(T4, Tc);
			      T6 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
			      T7 = VZMULJ(T5, T6);
			      Ta = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
			      Tb = VZMULJ(T9, Ta);
			 }
			 Tk = VSUB(T3, T7);
			 Tl = VSUB(Tb, Td);
			 T8 = VADD(T3, T7);
			 Te = VADD(Tb, Td);
			 Th = VADD(T8, Te);
		    }
		    ST(&(x[0]), VADD(Tg, Th), ms, &(x[0]));
		    {
			 V Tm, Tn, Tj, To, Tf, Ti;
			 Tm = VBYI(VFMA(LDK(KP951056516), Tk, VMUL(LDK(KP587785252), Tl)));
			 Tn = VBYI(VFNMS(LDK(KP587785252), Tk, VMUL(LDK(KP951056516), Tl)));
			 Tf = VMUL(LDK(KP559016994), VSUB(T8, Te));
			 Ti = VFNMS(LDK(KP250000000), Th, Tg);
			 Tj = VADD(Tf, Ti);
			 To = VSUB(Ti, Tf);
			 ST(&(x[WS(rs, 1)]), VSUB(Tj, Tm), ms, &(x[WS(rs, 1)]));
			 ST(&(x[WS(rs, 3)]), VSUB(To, Tn), ms, &(x[WS(rs, 1)]));
			 ST(&(x[WS(rs, 4)]), VADD(Tm, Tj), ms, &(x[0]));
			 ST(&(x[WS(rs, 2)]), VADD(Tn, To), ms, &(x[0]));
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:Aegisub，项目名称:fftw3，代码行数:57，代码来源:t3fv_5.c

示例11: hc2cbdftv_6

static void hc2cbdftv_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     INT m;
     for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 10)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(rs)) {
	  V T5, Th, Te, Ts, Tk, Tm, T2, T4, T3, T6, Tc, T8, Tb, T7, Ta;
	  V T9, Td, Ti, Tj, TA, Tf, Tn, Tv, Tt, Tz, T1, Tl, Tg, Tu, Tr;
	  V Tq, Ty, To, Tp, TC, TB, Tx, Tw;
	  T2 = LD(&(Rp[0]), ms, &(Rp[0]));
	  T3 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0]));
	  T4 = VCONJ(T3);
	  T5 = VSUB(T2, T4);
	  Th = VADD(T2, T4);
	  T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0]));
	  Tc = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
	  T7 = LD(&(Rm[0]), -ms, &(Rm[0]));
	  T8 = VCONJ(T7);
	  Ta = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
	  Tb = VCONJ(Ta);
	  T9 = VSUB(T6, T8);
	  Td = VSUB(Tb, Tc);
	  Te = VADD(T9, Td);
	  Ts = VBYI(VMUL(LDK(KP866025403), VSUB(T9, Td)));
	  Ti = VADD(T6, T8);
	  Tj = VADD(Tb, Tc);
	  Tk = VADD(Ti, Tj);
	  Tm = VBYI(VMUL(LDK(KP866025403), VSUB(Ti, Tj)));
	  TA = VADD(Th, Tk);
	  T1 = LDW(&(W[TWVL * 4]));
	  Tf = VZMULI(T1, VADD(T5, Te));
	  Tl = VFNMS(LDK(KP500000000), Tk, Th);
	  Tg = LDW(&(W[TWVL * 2]));
	  Tn = VZMUL(Tg, VSUB(Tl, Tm));
	  Tu = LDW(&(W[TWVL * 6]));
	  Tv = VZMUL(Tu, VADD(Tm, Tl));
	  Tr = VFNMS(LDK(KP500000000), Te, T5);
	  Tq = LDW(&(W[TWVL * 8]));
	  Tt = VZMULI(Tq, VSUB(Tr, Ts));
	  Ty = LDW(&(W[0]));
	  Tz = VZMULI(Ty, VADD(Ts, Tr));
	  To = VADD(Tf, Tn);
	  ST(&(Rp[WS(rs, 1)]), To, ms, &(Rp[WS(rs, 1)]));
	  Tp = VCONJ(VSUB(Tn, Tf));
	  ST(&(Rm[WS(rs, 1)]), Tp, -ms, &(Rm[WS(rs, 1)]));
	  TC = VCONJ(VSUB(TA, Tz));
	  ST(&(Rm[0]), TC, -ms, &(Rm[0]));
	  TB = VADD(Tz, TA);
	  ST(&(Rp[0]), TB, ms, &(Rp[0]));
	  Tx = VCONJ(VSUB(Tv, Tt));
	  ST(&(Rm[WS(rs, 2)]), Tx, -ms, &(Rm[0]));
	  Tw = VADD(Tt, Tv);
	  ST(&(Rp[WS(rs, 2)]), Tw, ms, &(Rp[0]));
     }
}
开发者ID:BackupTheBerlios，项目名称:openvsipl，代码行数:55，代码来源:hc2cbdftv_6.c

示例12: LD

static const R *t1sv_4(R *ri, R *ii, const R *W, stride ios, INT m, INT dist)
{
     INT i;
     for (i = m; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * dist), ii = ii + ((2 * VL) * dist), W = W + ((2 * VL) * 6), MAKE_VOLATILE_STRIDE(ios)) {
	  V T1, Tv, T3, T6, T5, Ta, Td, Tc, Tg, Tj, Tt, T4, Tf, Ti, Tn;
	  V Tb, T2, T9;
	  T1 = LD(&(ri[0]), dist, &(ri[0]));
	  Tv = LD(&(ii[0]), dist, &(ii[0]));
	  T3 = LD(&(ri[WS(ios, 2)]), dist, &(ri[0]));
	  T6 = LD(&(ii[WS(ios, 2)]), dist, &(ii[0]));
	  T2 = LDW(&(W[TWVL * 2]));
	  T5 = LDW(&(W[TWVL * 3]));
	  Ta = LD(&(ri[WS(ios, 1)]), dist, &(ri[WS(ios, 1)]));
	  Td = LD(&(ii[WS(ios, 1)]), dist, &(ii[WS(ios, 1)]));
	  T9 = LDW(&(W[0]));
	  Tc = LDW(&(W[TWVL * 1]));
	  Tg = LD(&(ri[WS(ios, 3)]), dist, &(ri[WS(ios, 1)]));
	  Tj = LD(&(ii[WS(ios, 3)]), dist, &(ii[WS(ios, 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(ios, 1)]), VADD(Tm, Tr), dist, &(ri[WS(ios, 1)]));
			 ST(&(ri[WS(ios, 3)]), VSUB(Tm, Tr), dist, &(ri[WS(ios, 1)]));
			 ST(&(ii[WS(ios, 2)]), VSUB(Tw, Ts), dist, &(ii[0]));
			 ST(&(ii[0]), VADD(Ts, Tw), dist, &(ii[0]));
			 ST(&(ii[WS(ios, 3)]), VADD(Ty, Tx), dist, &(ii[WS(ios, 1)]));
			 ST(&(ii[WS(ios, 1)]), VSUB(Tx, Ty), dist, &(ii[WS(ios, 1)]));
			 ST(&(ri[0]), VADD(T8, Tl), dist, &(ri[0]));
			 ST(&(ri[WS(ios, 2)]), VSUB(T8, Tl), dist, &(ri[0]));
		    }
	       }
	  }
     }
     return W;
}
开发者ID:exic，项目名称:last.fm-dbus，代码行数:60，代码来源:t1sv_4.c

示例13: t1sv_2

static void t1sv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     INT m;
     for (m = mb, W = W + (mb * 2); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 2), MAKE_VOLATILE_STRIDE(rs)) {
	  V T1, Ta, T3, T6, T2, T5;
	  T1 = LD(&(ri[0]), ms, &(ri[0]));
	  Ta = LD(&(ii[0]), ms, &(ii[0]));
	  T3 = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)]));
	  T6 = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)]));
	  T2 = LDW(&(W[0]));
	  T5 = LDW(&(W[TWVL * 1]));
	  {
	       V T8, T4, T9, T7;
	       T8 = VMUL(T2, T6);
	       T4 = VMUL(T2, T3);
	       T9 = VFNMS(T5, T3, T8);
	       T7 = VFMA(T5, T6, T4);
	       ST(&(ii[0]), VADD(T9, Ta), ms, &(ii[0]));
	       ST(&(ii[WS(rs, 1)]), VSUB(Ta, T9), ms, &(ii[WS(rs, 1)]));
	       ST(&(ri[0]), VADD(T1, T7), ms, &(ri[0]));
	       ST(&(ri[WS(rs, 1)]), VSUB(T1, T7), ms, &(ri[WS(rs, 1)]));
	  }
     }
}
开发者ID:phillipstanleymarbell，项目名称:sunflower-simulator，代码行数:24，代码来源:t1sv_2.c

示例14: hc2cfdftv_2

static void hc2cfdftv_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     INT m;
     for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 2)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(rs)) {
	  V T1, T2, T4, T3, T5, T7, T6;
	  T1 = LD(&(Rp[0]), ms, &(Rp[0]));
	  T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
	  T4 = LDW(&(W[0]));
	  T3 = VFMACONJ(T2, T1);
	  T5 = VZMULIJ(T4, VFNMSCONJ(T2, T1));
	  T7 = VCONJ(VMUL(LDK(KP500000000), VADD(T3, T5)));
	  T6 = VMUL(LDK(KP500000000), VSUB(T3, T5));
	  ST(&(Rm[0]), T7, -ms, &(Rm[0]));
	  ST(&(Rp[0]), T6, ms, &(Rp[0]));
     }
}
开发者ID:8cH9azbsFifZ，项目名称:wspr，代码行数:17，代码来源:hc2cfdftv_2.c

示例15: hc2cbdftv_2

static void hc2cbdftv_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     {
	  INT m;
	  for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 2)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(8, rs)) {
	       V T2, T3, T1, T5, T4, T7, T6;
	       T2 = LD(&(Rp[0]), ms, &(Rp[0]));
	       T3 = LD(&(Rm[0]), -ms, &(Rm[0]));
	       T1 = LDW(&(W[0]));
	       T5 = VFMACONJ(T3, T2);
	       T4 = VZMULI(T1, VFNMSCONJ(T3, T2));
	       T7 = VCONJ(VSUB(T5, T4));
	       T6 = VADD(T4, T5);
	       ST(&(Rm[0]), T7, -ms, &(Rm[0]));
	       ST(&(Rp[0]), T6, ms, &(Rp[0]));
	  }
     }
     VLEAVE();
}
开发者ID:Aegisub，项目名称:fftw3，代码行数:19，代码来源:hc2cbdftv_2.c


注：本文中的LDW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。