• 设为首页
  • 点击收藏
  • 手机版
    手机扫一扫访问
    迪恩网络手机版
  • 关注官方公众号
    微信扫一扫关注
    公众号
登陆 注册

C++ LDW函数代码示例

原作者: [db:作者] 来自: [db:来源] 收藏 邀请

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20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，您的评价将有助于我们的系统推荐出更棒的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


示例16: t3bv_10

static void t3bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     {
	  INT m;
	  R *x;
	  x = 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(10, rs)) {
	       V T1, T7, Th, Tx, Tr, Td, Tp, T6, Tv, Tc, Te, Ti, Tl, T2, T3;
	       V T5;
	       T2 = LDW(&(W[0]));
	       T3 = LDW(&(W[TWVL * 2]));
	       T5 = LDW(&(W[TWVL * 4]));
	       T1 = LD(&(x[0]), ms, &(x[0]));
	       T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
	       {
		    V To, Tw, Tq, Tu, Ta, T4, Tt, Tk, Tb;
		    To = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
		    Tw = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
		    Tq = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
		    Tu = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
		    Ta = VZMULJ(T2, T3);
		    T4 = VZMUL(T2, T3);
		    Th = VZMULJ(T2, T5);
		    Tt = VZMULJ(T3, T5);
		    Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
		    Tx = VZMUL(T2, Tw);
		    Tr = VZMUL(T5, Tq);
		    Tk = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
		    Td = VZMULJ(Ta, T5);
		    Tp = VZMUL(T4, To);
		    T6 = VZMULJ(T4, T5);
		    Tv = VZMUL(Tt, Tu);
		    Tc = VZMUL(Ta, Tb);
		    Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
		    Ti = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
		    Tl = VZMUL(T3, Tk);
	       }
	       {
		    V TN, Ts, T8, Ty, TO, Tf, Tj;
		    TN = VADD(Tp, Tr);
		    Ts = VSUB(Tp, Tr);
		    T8 = VZMUL(T6, T7);
		    Ty = VSUB(Tv, Tx);
		    TO = VADD(Tv, Tx);
		    Tf = VZMUL(Td, Te);
		    Tj = VZMUL(Th, Ti);
		    {
			 V T9, TJ, TP, TU, Tz, TF, Tg, TK, Tm, TL;
			 T9 = VSUB(T1, T8);
			 TJ = VADD(T1, T8);
			 TP = VADD(TN, TO);
			 TU = VSUB(TN, TO);
			 Tz = VADD(Ts, Ty);
			 TF = VSUB(Ts, Ty);
			 Tg = VSUB(Tc, Tf);
			 TK = VADD(Tc, Tf);
			 Tm = VSUB(Tj, Tl);
			 TL = VADD(Tj, Tl);
			 {
			      V TM, TV, Tn, TE;
			      TM = VADD(TK, TL);
			      TV = VSUB(TK, TL);
			      Tn = VADD(Tg, Tm);
			      TE = VSUB(Tg, Tm);
			      {
				   V TW, TY, TS, TQ, TG, TI, TC, TA, TR, TB;
				   TW = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TV, TU));
				   TY = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TU, TV));
				   TS = VSUB(TM, TP);
				   TQ = VADD(TM, TP);
				   TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TF, TE));
				   TI = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TE, TF));
				   TC = VSUB(Tn, Tz);
				   TA = VADD(Tn, Tz);
				   ST(&(x[0]), VADD(TJ, TQ), ms, &(x[0]));
				   TR = VFNMS(LDK(KP250000000), TQ, TJ);
				   ST(&(x[WS(rs, 5)]), VADD(T9, TA), ms, &(x[WS(rs, 1)]));
				   TB = VFNMS(LDK(KP250000000), TA, T9);
				   {
					V TX, TT, TH, TD;
					TX = VFMA(LDK(KP559016994), TS, TR);
					TT = VFNMS(LDK(KP559016994), TS, TR);
					TH = VFNMS(LDK(KP559016994), TC, TB);
					TD = VFMA(LDK(KP559016994), TC, TB);
					ST(&(x[WS(rs, 8)]), VFMAI(TW, TT), ms, &(x[0]));
					ST(&(x[WS(rs, 2)]), VFNMSI(TW, TT), ms, &(x[0]));
					ST(&(x[WS(rs, 6)]), VFMAI(TY, TX), ms, &(x[0]));
					ST(&(x[WS(rs, 4)]), VFNMSI(TY, TX), ms, &(x[0]));
					ST(&(x[WS(rs, 9)]), VFNMSI(TG, TD), ms, &(x[WS(rs, 1)]));
					ST(&(x[WS(rs, 1)]), VFMAI(TG, TD), ms, &(x[WS(rs, 1)]));
					ST(&(x[WS(rs, 7)]), VFNMSI(TI, TH), ms, &(x[WS(rs, 1)]));
					ST(&(x[WS(rs, 3)]), VFMAI(TI, TH), ms, &(x[WS(rs, 1)]));
				   }
			      }
			 }
		    }
//.........这里部分代码省略.........
开发者ID:Enny1991，项目名称:ExApp，代码行数:101，代码来源:t3bv_10.c


示例17: t3fv_20

static void t3fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
     DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
     DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
     DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
     {
	  INT m;
	  R *x;
	  x = ri;
	  for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(20, rs)) {
	       V T1k, T1w, T1r, T1z, T1o, T1y, T1v, T1h;
	       {
		    V T2, T8, T3, Td;
		    T2 = LDW(&(W[0]));
		    T8 = LDW(&(W[TWVL * 2]));
		    T3 = LDW(&(W[TWVL * 4]));
		    Td = LDW(&(W[TWVL * 6]));
		    {
			 V T7, TM, T1F, T23, T1p, Tp, T1j, T27, T1P, T1I, T1i, T1L, T28, T1S, T1q;
			 V TE, T1n, T1d, T26, T2e;
			 {
			      V T1, TK, T5, TH;
			      T1 = LD(&(x[0]), ms, &(x[0]));
			      TK = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
			      T5 = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
			      TH = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
			      {
				   V TA, Tx, TU, T1O, T14, Th, T1G, T1R, T1b, T1J, To, Ts, TV, Tv, TO;
				   V TQ, TT, Ty, TB;
				   {
					V Tq, Tt, T17, T1a, Tk, Tn;
					{
					     V Tl, Ti, T15, T18, TZ, Tc, T6, Tb, Tf, T10, T12, TL;
					     {
						  V TJ, Ta, T9, T4;
						  Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
						  TA = VZMULJ(T2, T8);
						  T9 = VZMUL(T2, T8);
						  Tx = VZMUL(T8, T3);
						  Tl = VZMULJ(T8, T3);
						  T4 = VZMUL(T2, T3);
						  Tq = VZMULJ(T2, T3);
						  Tt = VZMULJ(T2, Td);
						  Ti = VZMULJ(T8, Td);
						  T15 = VZMULJ(TA, Td);
						  T18 = VZMULJ(TA, T3);
						  TU = VZMUL(TA, T3);
						  TJ = VZMULJ(T9, Td);
						  TZ = VZMUL(T9, T3);
						  Tc = VZMULJ(T9, T3);
						  T6 = VZMULJ(T4, T5);
						  Tb = VZMULJ(T9, Ta);
						  Tf = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
						  T10 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
						  T12 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
						  TL = VZMULJ(TJ, TK);
					     }
					     {
						  V T1D, T11, T13, T19, T1E, Tg, T16, TI, Te, Tj, Tm;
						  T16 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)]));
						  TI = VZMULJ(Tc, TH);
						  Te = VZMULJ(Tc, Td);
						  T7 = VSUB(T1, T6);
						  T1D = VADD(T1, T6);
						  T11 = VZMULJ(TZ, T10);
						  T13 = VZMULJ(T8, T12);
						  T19 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
						  T17 = VZMULJ(T15, T16);
						  TM = VSUB(TI, TL);
						  T1E = VADD(TI, TL);
						  Tg = VZMULJ(Te, Tf);
						  Tj = LD(&(x[WS(rs, 16)]), ms, &(x[0]));
						  Tm = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
						  T1O = VADD(T11, T13);
						  T14 = VSUB(T11, T13);
						  T1a = VZMULJ(T18, T19);
						  T1F = VSUB(T1D, T1E);
						  T23 = VADD(T1D, T1E);
						  Th = VSUB(Tb, Tg);
						  T1G = VADD(Tb, Tg);
						  Tk = VZMULJ(Ti, Tj);
						  Tn = VZMULJ(Tl, Tm);
					     }
					}
					{
					     V Tr, Tu, TN, TP, TS;
					     Tr = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
					     T1R = VADD(T17, T1a);
					     T1b = VSUB(T17, T1a);
					     Tu = LD(&(x[WS(rs, 18)]), ms, &(x[0]));
					     TN = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
					     TP = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)]));
					     TS = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
					     T1J = VADD(Tk, Tn);
					     To = VSUB(Tk, Tn);
					     Ts = VZMULJ(Tq, Tr);
					     TV = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
					     Tv = VZMULJ(Tt, Tu);
					     TO = VZMULJ(T3, TN);
//.........这里部分代码省略.........
开发者ID:barnex，项目名称:fftw，代码行数:101，代码来源:dft_simd_sse2_t3fv_20.c


示例18: hc2cfdftv_8

static void hc2cfdftv_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     DVK(KP353553390, +0.353553390593273762200422181052424519642417969);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     {
	  INT m;
	  for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 14)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(32, rs)) {
	       V Ta, TE, Tr, TF, Tl, TK, Tw, TG, T1, T6, T3, T8, T2, T7, T4;
	       V T9, T5, To, Tq, Tn, Tp, Tc, Th, Te, Tj, Td, Ti, Tf, Tk, Tb;
	       V Tg, Tt, Tv, Ts, Tu, Ty, Tz, Tm, Tx, TC, TD, TA, TB, TI, TO;
	       V TL, TP, TH, TJ, TM, TR, TN, TQ;
	       T1 = LD(&(Rp[0]), ms, &(Rp[0]));
	       T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0]));
	       T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
	       T3 = VCONJ(T2);
	       T7 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0]));
	       T8 = VCONJ(T7);
	       T4 = VADD(T1, T3);
	       T5 = LDW(&(W[TWVL * 6]));
	       T9 = VZMULJ(T5, VADD(T6, T8));
	       Ta = VADD(T4, T9);
	       TE = VMUL(LDK(KP500000000), VSUB(T4, T9));
	       Tn = LDW(&(W[0]));
	       To = VZMULIJ(Tn, VSUB(T3, T1));
	       Tp = LDW(&(W[TWVL * 8]));
	       Tq = VZMULIJ(Tp, VSUB(T8, T6));
	       Tr = VADD(To, Tq);
	       TF = VSUB(To, Tq);
	       Tc = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
	       Th = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)]));
	       Td = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
	       Te = VCONJ(Td);
	       Ti = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)]));
	       Tj = VCONJ(Ti);
	       Tb = LDW(&(W[TWVL * 2]));
	       Tf = VZMULJ(Tb, VADD(Tc, Te));
	       Tg = LDW(&(W[TWVL * 10]));
	       Tk = VZMULJ(Tg, VADD(Th, Tj));
	       Tl = VADD(Tf, Tk);
	       TK = VSUB(Tf, Tk);
	       Ts = LDW(&(W[TWVL * 4]));
	       Tt = VZMULIJ(Ts, VSUB(Te, Tc));
	       Tu = LDW(&(W[TWVL * 12]));
	       Tv = VZMULIJ(Tu, VSUB(Tj, Th));
	       Tw = VADD(Tt, Tv);
	       TG = VSUB(Tv, Tt);
	       Tm = VADD(Ta, Tl);
	       Tx = VADD(Tr, Tw);
	       Ty = VCONJ(VMUL(LDK(KP500000000), VSUB(Tm, Tx)));
	       Tz = VMUL(LDK(KP500000000), VADD(Tm, Tx));
	       ST(&(Rm[WS(rs, 3)]), Ty, -ms, &(Rm[WS(rs, 1)]));
	       ST(&(Rp[0]), Tz, ms, &(Rp[0]));
	       TA = VSUB(Ta, Tl);
	       TB = VBYI(VSUB(Tw, Tr));
	       TC = VCONJ(VMUL(LDK(KP500000000), VSUB(TA, TB)));
	       TD = VMUL(LDK(KP500000000), VADD(TA, TB));
	       ST(&(Rm[WS(rs, 1)]), TC, -ms, &(Rm[WS(rs, 1)]));
	       ST(&(Rp[WS(rs, 2)]), TD, ms, &(Rp[0]));
	       TH = VMUL(LDK(KP353553390), VADD(TF, TG));
	       TI = VADD(TE, TH);
	       TO = VSUB(TE, TH);
	       TJ = VMUL(LDK(KP707106781), VSUB(TG, TF));
	       TL = VMUL(LDK(KP500000000), VBYI(VSUB(TJ, TK)));
	       TP = VMUL(LDK(KP500000000), VBYI(VADD(TK, TJ)));
	       TM = VCONJ(VSUB(TI, TL));
	       ST(&(Rm[0]), TM, -ms, &(Rm[0]));
	       TR = VADD(TO, TP);
	       ST(&(Rp[WS(rs, 3)]), TR, ms, &(Rp[WS(rs, 1)]));
	       TN = VADD(TI, TL);
	       ST(&(Rp[WS(rs, 1)]), TN, ms, &(Rp[WS(rs, 1)]));
	       TQ = VCONJ(VSUB(TO, TP));
	       ST(&(Rm[WS(rs, 2)]), TQ, -ms, &(Rm[0]));
	  }
     }
     VLEAVE();
}
开发者ID:barnex，项目名称:fftw，代码行数:77，代码来源:rdft_simd_sse2_hc2cfdftv_8.c


示例19: t3bv_8

static void t3bv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
     {
	  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(8, rs)) {
	       V T2, T3, Tb, T1, T5, Tn, Tq, T8, Td, T4, Ta, Tp, Tg, Ti, T9;
	       T2 = LDW(&(W[0]));
	       T3 = LDW(&(W[TWVL * 2]));
	       Tb = LDW(&(W[TWVL * 4]));
	       T1 = LD(&(x[0]), ms, &(x[0]));
	       T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
	       Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
	       Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
	       T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
	       Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
	       T4 = VZMUL(T2, T3);
	       Ta = VZMULJ(T2, T3);
	       Tp = VZMULJ(T2, Tb);
	       Tg = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
	       Ti = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
	       T9 = VZMUL(T2, T8);
	       {
		    V T6, To, Tc, Tr, Th, Tj;
		    T6 = VZMUL(T4, T5);
		    To = VZMUL(Ta, Tn);
		    Tc = VZMULJ(Ta, Tb);
		    Tr = VZMUL(Tp, Tq);
		    Th = VZMUL(Tb, Tg);
		    Tj = VZMUL(T3, Ti);
		    {
			 V Tx, T7, Te, Ts, Ty, Tk, TB;
			 Tx = VADD(T1, T6);
			 T7 = VSUB(T1, T6);
			 Te = VZMUL(Tc, Td);
			 Ts = VSUB(To, Tr);
			 Ty = VADD(To, Tr);
			 Tk = VSUB(Th, Tj);
			 TB = VADD(Th, Tj);
			 {
			      V Tf, TA, Tz, TD;
			      Tf = VSUB(T9, Te);
			      TA = VADD(T9, Te);
			      Tz = VSUB(Tx, Ty);
			      TD = VADD(Tx, Ty);
			      {
				   V TC, TE, Tl, Tt;
				   TC = VSUB(TA, TB);
				   TE = VADD(TA, TB);
				   Tl = VADD(Tf, Tk);
				   Tt = VSUB(Tf, Tk);
				   {
					V Tu, Tw, Tm, Tv;
					ST(&(x[0]), VADD(TD, TE), ms, &(x[0]));
					ST(&(x[WS(rs, 4)]), VSUB(TD, TE), ms, &(x[0]));
					ST(&(x[WS(rs, 2)]), VFMAI(TC, Tz), ms, &(x[0]));
					ST(&(x[WS(rs, 6)]), VFNMSI(TC, Tz), ms, &(x[0]));
					Tu = VFNMS(LDK(KP707106781), Tt, Ts);
					Tw = VFMA(LDK(KP707106781), Tt, Ts);
					Tm = VFNMS(LDK(KP707106781), Tl, T7);
					Tv = VFMA(LDK(KP707106781), Tl, T7);
					ST(&(x[WS(rs, 1)]), VFMAI(Tw, Tv), ms, &(x[WS(rs, 1)]));
					ST(&(x[WS(rs, 7)]), VFNMSI(Tw, Tv), ms, &(x[WS(rs, 1)]));
					ST(&(x[WS(rs, 5)]), VFMAI(Tu, Tm), ms, &(x[WS(rs, 1)]));
					ST(&(x[WS(rs, 3)]), VFNMSI(Tu, Tm), ms, &(x[WS(rs, 1)]));
				   }
			      }
			 }
		    }
	       }
	  }
     }
     VLEAVE();
}
开发者ID:barnex，项目名称:fftw，代码行数:76，代码来源:dft_simd_sse2_t3bv_8.c


示例20: hc2cfdftv_12

static void hc2cfdftv_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
     DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
     DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
     INT m;
     for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 22)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(rs)) {
	  V T3, T7, TH, TE, Th, TC, Tq, T11, TU, Tx, Tb, Tz, Tu, Tw, Tp;
	  V Tl, T9, Ta, T8, Ty, Tn, To, Tm, TG, T1, T2, Tt, T5, T6, T4;
	  V Tv, Tj, Tk, Ti, TD, Tf, Tg, Te, TB, TT, TF, TR, Tr;
	  T1 = LD(&(Rp[0]), ms, &(Rp[0]));
	  T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
	  Tt = LDW(&(W[0]));
	  T5

鲜花
握手
雷人
路过
鸡蛋
该文章已有0人参与评论

请发表评论

全部评论

专题导读
More+
上一篇：
C++ LD_WORD函数代码示例发布时间：2022-05-30
下一篇：
C++ LDST_L_BIT函数代码示例发布时间：2022-05-30
热门推荐
More+
热门话题
More+
阅读排行榜

扫描微信二维码

查看手机版网站

随时了解更新最新资讯

139-2527-9053

在线客服（服务时间 9:00～18:00）

在线QQ客服
地址：深圳市南山区西丽大学城创智工业园
电邮：jeky_zhao#qq.com
移动电话：139-2527-9053

Powered by 互联科技 X3.4© 2001-2213 极客世界.|Sitemap