这篇教程C++ BYTWJ函数代码示例写得很实用，希望能帮到您。
本文整理汇总了C++中BYTWJ函数的典型用法代码示例。如果您正苦于以下问题：C++ BYTWJ函数的具体用法？C++ BYTWJ怎么用？C++ BYTWJ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BYTWJ函数的24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: DVKstatic const R *t1fv_3(R *ri, R *ii, const R *W, stride ios, int m, int dist){ DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); int i; R *x; x = ri; BEGIN_SIMD(); for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 4)) { V T1, T3, T5, T6, T2, T4, T7, T8; T1 = LD(&(x[0]), dist, &(x[0])); T2 = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)])); T3 = BYTWJ(&(W[0]), T2); T4 = LD(&(x[WS(ios, 2)]), dist, &(x[0])); T5 = BYTWJ(&(W[TWVL * 2]), T4); T6 = VADD(T3, T5); ST(&(x[0]), VADD(T1, T6), dist, &(x[0])); T7 = VFNMS(LDK(KP500000000), T6, T1); T8 = VBYI(VMUL(LDK(KP866025403), VSUB(T5, T3))); ST(&(x[WS(ios, 2)]), VSUB(T7, T8), dist, &(x[0])); ST(&(x[WS(ios, 1)]), VADD(T7, T8), dist, &(x[WS(ios, 1)])); } END_SIMD(); return W;}
开发者ID:OS2World，项目名称:MM-SOUND-PM123，代码行数:25，
示例2: t1fuv_4static void t1fuv_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) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T3, T6, T7, T2, T5; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 4]), T7); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 2]), T2); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); { V T4, T9, Ta, Tb; T4 = VSUB(T1, T3); T9 = VBYI(VSUB(T6, T8)); ST(&(x[WS(rs, 1)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T4, T9), ms, &(x[WS(rs, 1)])); Ta = VADD(T1, T3); Tb = VADD(T6, T8); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); } } } VLEAVE();}
开发者ID:bohrasd，项目名称:windowsrtdev，代码行数:30，
示例3: t1fuv_3static void t1fuv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms){ DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); 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 T1, T2, T4; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, T8, T6, T7; T3 = BYTWJ(&(W[0]), T2); T5 = BYTWJ(&(W[TWVL * 2]), T4); T8 = VMUL(LDK(KP866025403), VSUB(T5, T3)); T6 = VADD(T3, T5); T7 = VFNMS(LDK(KP500000000), T6, T1); ST(&(x[0]), VADD(T1, T6), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T8, T7), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(T8, T7), ms, &(x[0])); } }}
开发者ID:8cH9azbsFifZ，项目名称:wspr，代码行数:25，
示例4: DVKstatic const R *t1fv_6(R *ri, R *ii, const R *W, stride ios, int m, int dist){ DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); int i; R *x; x = ri; BEGIN_SIMD(); for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 10)) { V T4, Ti, Te, Tk, T9, Tj, T1, T3, T2; T1 = LD(&(x[0]), dist, &(x[0])); T2 = LD(&(x[WS(ios, 3)]), dist, &(x[WS(ios, 1)])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T4 = VSUB(T1, T3); Ti = VADD(T1, T3); { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(ios, 4)]), dist, &(x[0])); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Tc = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)])); Td = BYTWJ(&(W[0]), Tc); Te = VSUB(Tb, Td); Tk = VADD(Tb, Td); } { V T6, T8, T5, T7; T5 = LD(&(x[WS(ios, 2)]), dist, &(x[0])); T6 = BYTWJ(&(W[TWVL * 2]), T5); T7 = LD(&(x[WS(ios, 5)]), dist, &(x[WS(ios, 1)])); T8 = BYTWJ(&(W[TWVL * 8]), T7); T9 = VSUB(T6, T8); Tj = VADD(T6, T8); } { V Th, Tf, Tg, Tn, Tl, Tm; Th = VBYI(VMUL(LDK(KP866025403), VSUB(Te, T9))); Tf = VADD(T9, Te); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(ios, 3)]), VADD(T4, Tf), dist, &(x[WS(ios, 1)])); ST(&(x[WS(ios, 1)]), VADD(Tg, Th), dist, &(x[WS(ios, 1)])); ST(&(x[WS(ios, 5)]), VSUB(Tg, Th), dist, &(x[WS(ios, 1)])); Tn = VBYI(VMUL(LDK(KP866025403), VSUB(Tk, Tj))); Tl = VADD(Tj, Tk); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[0]), VADD(Ti, Tl), dist, &(x[0])); ST(&(x[WS(ios, 4)]), VADD(Tm, Tn), dist, &(x[0])); ST(&(x[WS(ios, 2)]), VSUB(Tm, Tn), dist, &(x[0])); } } END_SIMD(); return W;}
开发者ID:abrahamneben，项目名称:orbcomm_beam_mapping，代码行数:52，
示例5: t1fuv_6static void t1fuv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms){ DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(rs)) { V T1, T2, Ta, Tc, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T3, Tb, Td, T6, T8; T3 = BYTWJ(&(W[TWVL * 4]), T2); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Td = BYTWJ(&(W[0]), Tc); T6 = BYTWJ(&(W[TWVL * 2]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); { V Ti, T4, Tk, Te, Tj, T9; Ti = VADD(T1, T3); T4 = VSUB(T1, T3); Tk = VADD(Tb, Td); Te = VSUB(Tb, Td); Tj = VADD(T6, T8); T9 = VSUB(T6, T8); { V Tl, Tn, Tf, Th, Tm, Tg; Tl = VADD(Tj, Tk); Tn = VMUL(LDK(KP866025403), VSUB(Tk, Tj)); Tf = VADD(T9, Te); Th = VMUL(LDK(KP866025403), VSUB(Te, T9)); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(rs, 3)]), VADD(T4, Tf), ms, &(x[WS(rs, 1)])); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(rs, 2)]), VFNMSI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(Th, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Th, Tg), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE();}
开发者ID:dstuck，项目名称:tinker_integrated_PIMC，代码行数:52，
示例6: t1fv_5static void t1fv_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) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V Tc, Tg, Th, T5, Ta, Td; Tc = LD(&(x[0]), ms, &(x[0])); { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[0]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T4 = BYTWJ(&(W[TWVL * 6]), T3); T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 2]), T6); } Tg = VSUB(T2, T4); Th = VSUB(T7, T9); T5 = VADD(T2, T4); Ta = VADD(T7, T9); Td = VADD(T5, Ta); } ST(&(x[0]), VADD(Tc, Td), ms, &(x[0])); { V Ti, Tj, Tf, Tk, Tb, Te; Ti = VBYI(VFMA(LDK(KP951056516), Tg, VMUL(LDK(KP587785252), Th))); Tj = VBYI(VFNMS(LDK(KP587785252), Tg, VMUL(LDK(KP951056516), Th))); Tb = VMUL(LDK(KP559016994), VSUB(T5, Ta)); Te = VFNMS(LDK(KP250000000), Td, Tc); Tf = VADD(Tb, Te); Tk = VSUB(Te, Tb); ST(&(x[WS(rs, 1)]), VSUB(Tf, Ti), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0])); } } } VLEAVE();}
开发者ID:SKA-ScienceDataProcessor，项目名称:FastImaging，代码行数:50，
示例7: DVKstatic const R *t1fv_6(R *ri, R *ii, const R *W, stride ios, INT m, INT dist){ DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); INT i; R *x; x = ri; for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(ios)) { V T1, T2, Ta, Tc, T5, T7; T1 = LD(&(x[0]), dist, &(x[0])); T2 = LD(&(x[WS(ios, 3)]), dist, &(x[WS(ios, 1)])); Ta = LD(&(x[WS(ios, 4)]), dist, &(x[0])); Tc = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)])); T5 = LD(&(x[WS(ios, 2)]), dist, &(x[0])); T7 = LD(&(x[WS(ios, 5)]), dist, &(x[WS(ios, 1)])); { V T3, Tb, Td, T6, T8; T3 = BYTWJ(&(W[TWVL * 4]), T2); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Td = BYTWJ(&(W[0]), Tc); T6 = BYTWJ(&(W[TWVL * 2]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); { V Ti, T4, Tk, Te, Tj, T9; Ti = VADD(T1, T3); T4 = VSUB(T1, T3); Tk = VADD(Tb, Td); Te = VSUB(Tb, Td); Tj = VADD(T6, T8); T9 = VSUB(T6, T8); { V Tl, Tn, Tf, Th, Tm, Tg; Tl = VADD(Tj, Tk); Tn = VMUL(LDK(KP866025403), VSUB(Tk, Tj)); Tf = VADD(T9, Te); Th = VMUL(LDK(KP866025403), VSUB(Te, T9)); ST(&(x[0]), VADD(Ti, Tl), dist, &(x[0])); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(ios, 3)]), VADD(T4, Tf), dist, &(x[WS(ios, 1)])); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(ios, 2)]), VFNMSI(Tn, Tm), dist, &(x[0])); ST(&(x[WS(ios, 4)]), VFMAI(Tn, Tm), dist, &(x[0])); ST(&(x[WS(ios, 5)]), VFNMSI(Th, Tg), dist, &(x[WS(ios, 1)])); ST(&(x[WS(ios, 1)]), VFMAI(Th, Tg), dist, &(x[WS(ios, 1)])); } } } } return W;}
开发者ID:exic，项目名称:last.fm-dbus，代码行数:50，
示例8: q1fv_2static void q1fv_2(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) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(rs), MAKE_VOLATILE_STRIDE(vs)) { V T1, T2, T3, T4, T5, T6; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), VSUB(T1, T2)); T4 = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T5 = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T6 = BYTWJ(&(W[0]), VSUB(T4, T5)); ST(&(x[WS(vs, 1)]), T3, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), T6, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[0]), VADD(T1, T2), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T4, T5), ms, &(x[WS(rs, 1)])); }}
开发者ID:phillipstanleymarbell，项目名称:sunflower-simulator，代码行数:19，
示例9: LDstatic const R *q1fv_2(R *ri, R *ii, const R *W, stride is, stride vs, INT m, INT dist){ INT i; R *x; x = ri; for (i = 0; i < m; i = i + VL, x = x + (VL * dist), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(vs)) { V T1, T2, T3, T4, T5, T6; T1 = LD(&(x[0]), dist, &(x[0])); T2 = LD(&(x[WS(is, 1)]), dist, &(x[WS(is, 1)])); T3 = BYTWJ(&(W[0]), VSUB(T1, T2)); T4 = LD(&(x[WS(vs, 1)]), dist, &(x[WS(vs, 1)])); T5 = LD(&(x[WS(vs, 1) + WS(is, 1)]), dist, &(x[WS(vs, 1) + WS(is, 1)])); T6 = BYTWJ(&(W[0]), VSUB(T4, T5)); ST(&(x[WS(vs, 1)]), T3, dist, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(is, 1)]), T6, dist, &(x[WS(vs, 1) + WS(is, 1)])); ST(&(x[0]), VADD(T1, T2), dist, &(x[0])); ST(&(x[WS(is, 1)]), VADD(T4, T5), dist, &(x[WS(is, 1)])); } return W;}
开发者ID:exic，项目名称:last.fm-dbus，代码行数:20，
示例10: t1fuv_7static void t1fuv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms){ DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); INT 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, Tg, Tj, T6, Ti, Tb, Tk, Tp, To; T1 = LD(&(x[0]), ms, &(x[0])); { V Td, Tf, Tc, Te; Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 4]), Tc); Te = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tf = BYTWJ(&(W[TWVL * 6]), Te); Tg = VADD(Td, Tf); Tj = VSUB(Tf, Td); } { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); Ti = VSUB(T5, T3); } { V T8, Ta, T7, T9; T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 2]), T7); T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = BYTWJ(&(W[TWVL * 8]), T9); Tb = VADD(T8, Ta); Tk = VSUB(Ta, T8); } ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0])); Tp = VBYI(VFMA(LDK(KP433883739), Ti, VFNMS(LDK(KP781831482), Tk, VMUL(LDK(KP974927912), Tj)))); To = VFMA(LDK(KP623489801), Tb, VFNMS(LDK(KP222520933), Tg, VFNMS(LDK(KP900968867), T6, T1))); ST(&(x[WS(rs, 4)]), VSUB(To, Tp), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); { V Tl, Th, Tn, Tm; Tl = VBYI(VFNMS(LDK(KP781831482), Tj, VFNMS(LDK(KP433883739), Tk, VMUL(LDK(KP974927912), Ti)))); Th = VFMA(LDK(KP623489801), Tg, VFNMS(LDK(KP900968867), Tb, VFNMS(LDK(KP222520933), T6, T1))); ST(&(x[WS(rs, 5)]), VSUB(Th, Tl), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(Th, Tl), ms, &(x[0])); Tn = VBYI(VFMA(LDK(KP781831482), Ti, VFMA(LDK(KP974927912), Tk, VMUL(LDK(KP433883739), Tj)))); Tm = VFMA(LDK(KP623489801), T6, VFNMS(LDK(KP900968867), Tg, VFNMS(LDK(KP222520933), Tb, T1))); ST(&(x[WS(rs, 6)]), VSUB(Tm, Tn), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(Tm, Tn), ms, &(x[WS(rs, 1)])); } }}
开发者ID:BackupTheBerlios，项目名称:openvsipl，代码行数:59，
示例11: BEGIN_SIMDstatic const R *q1fv_2(R *ri, R *ii, const R *W, stride is, stride vs, int m, int dist){ int i; R *x; x = ri; BEGIN_SIMD(); for (i = 0; i < m; i = i + VL, x = x + (VL * dist), W = W + (TWVL * 2)) { V T1, T2, T3, T4, T5, T6; T1 = LD(&(x[0]), dist, &(x[0])); T2 = LD(&(x[WS(is, 1)]), dist, &(x[WS(is, 1)])); T3 = BYTWJ(&(W[0]), VSUB(T1, T2)); T4 = LD(&(x[WS(vs, 1)]), dist, &(x[WS(vs, 1)])); T5 = LD(&(x[WS(vs, 1) + WS(is, 1)]), dist, &(x[WS(vs, 1) + WS(is, 1)])); T6 = BYTWJ(&(W[0]), VSUB(T4, T5)); ST(&(x[WS(vs, 1)]), T3, dist, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(is, 1)]), T6, dist, &(x[WS(vs, 1) + WS(is, 1)])); ST(&(x[0]), VADD(T1, T2), dist, &(x[0])); ST(&(x[WS(is, 1)]), VADD(T4, T5), dist, &(x[WS(is, 1)])); } END_SIMD(); return W;}
开发者ID:abrahamneben，项目名称:orbcomm_beam_mapping，代码行数:22，
示例12: LDstatic const R *t2fv_2(R *ri, R *ii, const R *W, stride ios, INT m, INT dist){ INT i; R *x; x = ri; for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(ios)) { V T1, T3, T2; T1 = LD(&(x[0]), dist, &(x[0])); T2 = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[WS(ios, 1)]), VSUB(T1, T3), dist, &(x[WS(ios, 1)])); ST(&(x[0]), VADD(T1, T3), dist, &(x[0])); } return W;}
开发者ID:exic，项目名称:last.fm-dbus，代码行数:15，
示例13: t1fuv_2static void t1fuv_2(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) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); } } VLEAVE();}
开发者ID:barnex，项目名称:fftw，代码行数:17，
示例14: BEGIN_SIMDstatic const R *t1fv_2(R *ri, R *ii, const R *W, stride ios, int m, int dist){ int i; R *x; x = ri; BEGIN_SIMD(); for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 2)) { V T1, T3, T2; T1 = LD(&(x[0]), dist, &(x[0])); T2 = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[WS(ios, 1)]), VSUB(T1, T3), dist, &(x[WS(ios, 1)])); ST(&(x[0]), VADD(T1, T3), dist, &(x[0])); } END_SIMD(); return W;}
开发者ID:abrahamneben，项目名称:orbcomm_beam_mapping，代码行数:17，
示例15: t1fuv_9static void t1fuv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms){ DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP296198132, +0.296198132726023843175338011893050938967728390); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP173648177, +0.173648177666930348851716626769314796000375677); DVK(KP556670399, +0.556670399226419366452912952047023132968291906); DVK(KP766044443, +0.766044443118978035202392650555416673935832457); DVK(KP642787609, +0.642787609686539326322643409907263432907559884); DVK(KP663413948, +0.663413948168938396205421319635891297216863310); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP150383733, +0.150383733180435296639271897612501926072238258); DVK(KP342020143, +0.342020143325668733044099614682259580763083368); DVK(KP813797681, +0.813797681349373692844693217248393223289101568); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(rs)) { V T1, T6, TA, Tt, Tf, Ts, Tw, Tn, Tv; T1 = LD(&(x[0]), ms, &(x[0])); { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); TA = VMUL(LDK(KP866025403), VSUB(T5, T3)); } { V T9, Td, Tb, T8, Tc, Ta, Te; T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[0]), T8); Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 12]), Tc); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Tt = VSUB(Td, Tb); Te = VADD(Tb, Td); Tf = VADD(T9, Te); Ts = VFNMS(LDK(KP500000000), Te, T9); } { V Th, Tl, Tj, Tg, Tk, Ti, Tm; Tg = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 2]), Tg); Tk = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tl = BYTWJ(&(W[TWVL * 14]), Tk); Ti = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 8]), Ti); Tw = VSUB(Tl, Tj); Tm = VADD(Tj, Tl); Tn = VADD(Th, Tm); Tv = VFNMS(LDK(KP500000000), Tm, Th); } { V Tq, T7, To, Tp; Tq = VBYI(VMUL(LDK(KP866025403), VSUB(Tn, Tf))); T7 = VADD(T1, T6); To = VADD(Tf, Tn); Tp = VFNMS(LDK(KP500000000), To, T7); ST(&(x[0]), VADD(T7, To), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(Tp, Tq), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(Tp, Tq), ms, &(x[0])); } { V TI, TB, TC, TD, Tu, Tx, Ty, Tr, TH; TI = VBYI(VSUB(VFNMS(LDK(KP342020143), Tv, VFNMS(LDK(KP150383733), Tt, VFNMS(LDK(KP984807753), Ts, VMUL(LDK(KP813797681), Tw)))), TA)); TB = VFNMS(LDK(KP642787609), Ts, VMUL(LDK(KP663413948), Tt)); TC = VFNMS(LDK(KP984807753), Tv, VMUL(LDK(KP150383733), Tw)); TD = VADD(TB, TC); Tu = VFMA(LDK(KP766044443), Ts, VMUL(LDK(KP556670399), Tt)); Tx = VFMA(LDK(KP173648177), Tv, VMUL(LDK(KP852868531), Tw)); Ty = VADD(Tu, Tx); Tr = VFNMS(LDK(KP500000000), T6, T1); TH = VFMA(LDK(KP173648177), Ts, VFNMS(LDK(KP296198132), Tw, VFNMS(LDK(KP939692620), Tv, VFNMS(LDK(KP852868531), Tt, Tr)))); ST(&(x[WS(rs, 7)]), VSUB(TH, TI), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(TH, TI), ms, &(x[0])); { V Tz, TE, TF, TG; Tz = VADD(Tr, Ty); TE = VBYI(VADD(TA, TD)); ST(&(x[WS(rs, 8)]), VSUB(Tz, TE), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(TE, Tz), ms, &(x[WS(rs, 1)])); TF = VFMA(LDK(KP866025403), VSUB(TB, TC), VFNMS(LDK(KP500000000), Ty, Tr)); TG = VBYI(VADD(TA, VFNMS(LDK(KP500000000), TD, VMUL(LDK(KP866025403), VSUB(Tx, Tu))))); ST(&(x[WS(rs, 5)]), VSUB(TF, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(TF, TG), ms, &(x[0])); } } } } VLEAVE();}
开发者ID:dstuck，项目名称:tinker_integrated_PIMC，代码行数:97，
示例16: t1fuv_8static void t1fuv_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 = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(rs)) { V T4, Tq, Tm, Tr, T9, Tt, Te, Tu, T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 6]), T2); T4 = VSUB(T1, T3); Tq = VADD(T1, T3); { V Tj, Tl, Ti, Tk; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tl = BYTWJ(&(W[TWVL * 10]), Tk); Tm = VSUB(Tj, Tl); Tr = VADD(Tj, Tl); } { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 8]), T7); T9 = VSUB(T6, T8); Tt = VADD(T6, T8); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 4]), Tc); Te = VSUB(Tb, Td); Tu = VADD(Tb, Td); } { V Ts, Tv, Tw, Tx; Ts = VADD(Tq, Tr); Tv = VADD(Tt, Tu); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); Tw = VSUB(Tq, Tr); Tx = VBYI(VSUB(Tu, Tt)); ST(&(x[WS(rs, 6)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tw, Tx), ms, &(x[0])); { V Tg, To, Tn, Tp, Tf, Th; Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VADD(T4, Tf); To = VSUB(T4, Tf); Th = VMUL(LDK(KP707106781), VSUB(Te, T9)); Tn = VBYI(VSUB(Th, Tm)); Tp = VBYI(VADD(Tm, Th)); ST(&(x[WS(rs, 7)]), VSUB(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(To, Tp), ms, &(x[WS(rs, 1)])); } } }}
开发者ID:8cH9azbsFifZ，项目名称:wspr，代码行数:66，
示例17: t1fv_12static void t1fv_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 = ri; 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(12, rs)) { V T1, TC, T6, T7, Ty, Tq, Tz, TA, T9, TD, Te, Tf, Tu, Tl, Tv; V Tw; { V T5, T3, T4, T2; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 14]), T4); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 6]), T2); TC = VSUB(T5, T3); T6 = VADD(T3, T5); T7 = VFNMS(LDK(KP500000000), T6, T1); } { V Tn, Tp, Tm, Tx, To; Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tn = BYTWJ(&(W[0]), Tm); Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Ty = BYTWJ(&(W[TWVL * 16]), Tx); To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tp = BYTWJ(&(W[TWVL * 8]), To); Tq = VSUB(Tn, Tp); Tz = VADD(Tn, Tp); TA = VFNMS(LDK(KP500000000), Tz, Ty); } { V Td, Tb, T8, Tc, Ta; T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T9 = BYTWJ(&(W[TWVL * 10]), T8); Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 2]), Tc); Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 18]), Ta); TD = VSUB(Td, Tb); Te = VADD(Tb, Td); Tf = VFNMS(LDK(KP500000000), Te, T9); } { V Ti, Tk, Th, Tt, Tj; Th = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Ti = BYTWJ(&(W[TWVL * 20]), Th); Tt = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 4]), Tt); Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[TWVL * 12]), Tj); Tl = VSUB(Ti, Tk); Tv = VADD(Tk, Ti); Tw = VFNMS(LDK(KP500000000), Tv, Tu); } { V Ts, TG, TF, TH; { V Tg, Tr, TB, TE; Tg = VSUB(T7, Tf); Tr = VADD(Tl, Tq); Ts = VFMA(LDK(KP866025403), Tr, Tg); TG = VFNMS(LDK(KP866025403), Tr, Tg); TB = VSUB(Tw, TA); TE = VSUB(TC, TD); TF = VFNMS(LDK(KP866025403), TE, TB); TH = VFMA(LDK(KP866025403), TE, TB); } ST(&(x[WS(rs, 1)]), VFNMSI(TF, Ts), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(TH, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(TF, Ts), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(TH, TG), ms, &(x[WS(rs, 1)])); } { V TS, TW, TV, TX; { V TQ, TR, TT, TU; TQ = VADD(T1, T6); TR = VADD(T9, Te); TS = VSUB(TQ, TR); TW = VADD(TQ, TR); TT = VADD(Tu, Tv); TU = VADD(Ty, Tz); TV = VSUB(TT, TU); TX = VADD(TT, TU); } ST(&(x[WS(rs, 9)]), VFNMSI(TV, TS), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(TW, TX), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(TV, TS), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(TW, TX), ms, &(x[0])); } { V TK, TO, TN, TP; { V TI, TJ, TL, TM; TI = VADD(T7, Tf); TJ = VADD(Tw, TA);//.........这里部分代码省略.........
开发者ID:DougBurke，项目名称:sherpa，代码行数:101，
示例18: t1fv_15static void t1fv_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms){ DVK(KP216506350, +0.216506350946109661690930792688234045867850657); DVK(KP484122918, +0.484122918275927110647408174972799951354115213); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP509036960, +0.509036960455127183450980863393907648510733164); DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 28)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 28), MAKE_VOLATILE_STRIDE(rs)) { V T1e, T7, TP, T12, T15, Tf, Tn, To, T1b, T1c, T1f, TQ, TR, TS, Tw; V TE, TF, TT, TU, TV; { V T1, T5, T3, T4, T2, T6; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 18]), T4); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 8]), T2); T1e = VSUB(T5, T3); T6 = VADD(T3, T5); T7 = VADD(T1, T6); TP = VFNMS(LDK(KP500000000), T6, T1); } { V T9, Tq, Ty, Th, Te, T13, Tv, T10, TD, T11, Tm, T14; { V T8, Tp, Tx, Tg; T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tq = BYTWJ(&(W[TWVL * 10]), Tp); Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Ty = BYTWJ(&(W[TWVL * 16]), Tx); Tg = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 22]), Tg); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 14]), Ta); Tc = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 24]), Tc); Te = VADD(Tb, Td); T13 = VSUB(Td, Tb); } { V Ts, Tu, Tr, Tt; Tr = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Ts = BYTWJ(&(W[TWVL * 20]), Tr); Tt = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[0]), Tt); Tv = VADD(Ts, Tu); T10 = VSUB(Tu, Ts); } { V TA, TC, Tz, TB; Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 26]), Tz); TB = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TC = BYTWJ(&(W[TWVL * 6]), TB); TD = VADD(TA, TC); T11 = VSUB(TC, TA); } { V Tj, Tl, Ti, Tk; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tl = BYTWJ(&(W[TWVL * 12]), Tk); Tm = VADD(Tj, Tl); T14 = VSUB(Tl, Tj); } T12 = VSUB(T10, T11); T15 = VSUB(T13, T14); Tf = VADD(T9, Te); Tn = VADD(Th, Tm); To = VADD(Tf, Tn); T1b = VADD(T13, T14); T1c = VADD(T10, T11); T1f = VADD(T1b, T1c); TQ = VFNMS(LDK(KP500000000), Te, T9); TR = VFNMS(LDK(KP500000000), Tm, Th); TS = VADD(TQ, TR); Tw = VADD(Tq, Tv); TE = VADD(Ty, TD); TF = VADD(Tw, TE); TT = VFNMS(LDK(KP500000000), Tv, Tq); TU = VFNMS(LDK(KP500000000), TD, Ty); TV = VADD(TT, TU); } { V TI, TG, TH, TM, TO, TK, TL, TN, TJ; TI = VMUL(LDK(KP559016994), VSUB(To, TF));//.........这里部分代码省略.........
开发者ID:bambang，项目名称:vsipl，代码行数:101，
示例19: t1fv_16static void t1fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms){ DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TO, Ta, TJ, TP, T14, Tq, T1i, T10, T1b, T1l, T13, T1c, TR, Tl, T15; V Tv; { V Tc, TW, T4, T19, T9, TD, TI, Tj, TZ, T1a, Te, Th, Tn, Tr, Tu; V Tp; { V T1, T2, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tz, TG, TB, TE; Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TE = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V Ti, TY, TX, Td, Tg, Tm, Tt, To; { V T3, T6, T8, TA, TH, TC, TF, Tb; Tb = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 14]), T2); T6 = BYTWJ(&(W[TWVL * 6]), T5); T8 = BYTWJ(&(W[TWVL * 22]), T7); TA = BYTWJ(&(W[TWVL * 26]), Tz); TH = BYTWJ(&(W[TWVL * 18]), TG); TC = BYTWJ(&(W[TWVL * 10]), TB); TF = BYTWJ(&(W[TWVL * 2]), TE); Tc = BYTWJ(&(W[0]), Tb); TW = VSUB(T1, T3); T4 = VADD(T1, T3); T19 = VSUB(T6, T8); T9 = VADD(T6, T8); Ti = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TD = VADD(TA, TC); TY = VSUB(TA, TC); TI = VADD(TF, TH); TX = VSUB(TF, TH); } Td = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tm = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 24]), Ti); Tt = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); To = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TZ = VADD(TX, TY); T1a = VSUB(TY, TX); Te = BYTWJ(&(W[TWVL * 16]), Td); Th = BYTWJ(&(W[TWVL * 8]), Tg); Tn = BYTWJ(&(W[TWVL * 28]), Tm); Tr = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 20]), Tt); Tp = BYTWJ(&(W[TWVL * 12]), To); } } } { V Tf, T11, Tk, T12, Ts; TO = VADD(T4, T9); Ta = VSUB(T4, T9); TJ = VSUB(TD, TI); TP = VADD(TI, TD); Tf = VADD(Tc, Te); T11 = VSUB(Tc, Te); Tk = VADD(Th, Tj); T12 = VSUB(Th, Tj); Ts = BYTWJ(&(W[TWVL * 4]), Tr); T14 = VSUB(Tn, Tp); Tq = VADD(Tn, Tp); T1i = VFNMS(LDK(KP707106781), TZ, TW); T10 = VFMA(LDK(KP707106781), TZ, TW); T1b = VFNMS(LDK(KP707106781), T1a, T19); T1l = VFMA(LDK(KP707106781), T1a, T19); T13 = VFNMS(LDK(KP414213562), T12, T11); T1c = VFMA(LDK(KP414213562), T11, T12); TR = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); T15 = VSUB(Tu, Ts); Tv = VADD(Ts, Tu); } } { V T1d, T16, TS, Tw, TU, TQ; T1d = VFMA(LDK(KP414213562), T14, T15); T16 = VFNMS(LDK(KP414213562), T15, T14); TS = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); TU = VSUB(TO, TP); TQ = VADD(TO, TP);//.........这里部分代码省略.........
开发者ID:Aegisub，项目名称:fftw3，代码行数:101，
示例20: t2fv_32static void t2fv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms){ DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(rs)) { V T26, T25, T1Z, T22, T1W, T2a, T2k, T2g; { V T4, T1z, T2o, T32, T2r, T3f, Tf, T1A, T34, T2L, T1D, TC, T33, T2O, T1C; V Tr, T2C, T3a, T2F, T3b, T1r, T21, T1k, T20, TQ, TM, TS, TL, T2t, TJ; V T10, T2u; { V Tt, T9, T2p, Te, T2q, TA, Tu, Tx; { V T1, T1x, T2, T1v; T1 = LD(&(x[0]), ms, &(x[0])); T1x = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1v = LD(&(x[WS(rs, 8)]), ms, &(x[0])); { V T5, Tc, T7, Ta, T2m, T2n; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); { V T1y, T3, T1w, T6, Td, T8, Tb, Ts, Tz; Ts = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T1y = BYTWJ(&(W[TWVL * 46]), T1x); T3 = BYTWJ(&(W[TWVL * 30]), T2); T1w = BYTWJ(&(W[TWVL * 14]), T1v); T6 = BYTWJ(&(W[TWVL * 6]), T5); Td = BYTWJ(&(W[TWVL * 22]), Tc); T8 = BYTWJ(&(W[TWVL * 38]), T7); Tb = BYTWJ(&(W[TWVL * 54]), Ta); Tt = BYTWJ(&(W[TWVL * 58]), Ts); Tz = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T4 = VSUB(T1, T3); T2m = VADD(T1, T3); T1z = VSUB(T1w, T1y); T2n = VADD(T1w, T1y); T9 = VSUB(T6, T8); T2p = VADD(T6, T8); Te = VSUB(Tb, Td); T2q = VADD(Tb, Td); TA = BYTWJ(&(W[TWVL * 10]), Tz); } Tu = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2o = VADD(T2m, T2n); T32 = VSUB(T2m, T2n); Tx = LD(&(x[WS(rs, 22)]), ms, &(x[0])); } } { V Tv, To, Ty, Ti, Tj, Tm, Th; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2r = VADD(T2p, T2q); T3f = VSUB(T2q, T2p); Tf = VADD(T9, Te); T1A = VSUB(Te, T9); Tv = BYTWJ(&(W[TWVL * 26]), Tu); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Ty = BYTWJ(&(W[TWVL * 42]), Tx); Ti = BYTWJ(&(W[TWVL * 2]), Th); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); { V T1f, T1h, T1a, T1c, T18, T2A, T2B, T1p; { V T15, T17, T1o, T1m; { V Tw, T2J, Tp, T2K, TB, Tk, Tn, T1n, T14, T16; T14 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T16 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tt, Tv); T2J = VADD(Tt, Tv); Tp = BYTWJ(&(W[TWVL * 50]), To); T2K = VADD(TA, Ty); TB = VSUB(Ty, TA); Tk = BYTWJ(&(W[TWVL * 34]), Tj); Tn = BYTWJ(&(W[TWVL * 18]), Tm); T15 = BYTWJ(&(W[TWVL * 60]), T14); T17 = BYTWJ(&(W[TWVL * 28]), T16); T1n = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); { V T2M, Tl, T2N, Tq, T1l; T1l = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T34 = VSUB(T2J, T2K); T2L = VADD(T2J, T2K); T1D = VFMA(LDK(KP414213562), Tw, TB); TC = VFNMS(LDK(KP414213562), TB, Tw); T2M = VADD(Ti, Tk);//.........这里部分代码省略.........
开发者ID:AlexanderB1290，项目名称:scafacos，代码行数:101，
示例21: t1fv_10static void t1fv_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 = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Tr, TH, Tg, Tl, Tm, TA, TB, TJ, T5, Ta, Tb, TD, TE, TI, To; V Tq, Tp; To = LD(&(x[0]), ms, &(x[0])); Tp = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tq = BYTWJ(&(W[TWVL * 8]), Tp); Tr = VSUB(To, Tq); TH = VADD(To, Tq); { V Td, Tk, Tf, Ti; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 6]), Tc); Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[0]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTWJ(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 10]), Th); } Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VADD(Tg, Tl); TA = VADD(Td, Tf); TB = VADD(Ti, Tk); TJ = VADD(TA, TB); } { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTWJ(&(W[TWVL * 2]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTWJ(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 14]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VADD(T5, Ta); TD = VADD(T2, T4); TE = VADD(T7, T9); TI = VADD(TD, TE); } { V Tn, Ts, Tt, Tx, Tz, Tv, Tw, Ty, Tu; Tn = VMUL(LDK(KP559016994), VSUB(Tb, Tm)); Ts = VADD(Tb, Tm); Tt = VFNMS(LDK(KP250000000), Ts, Tr); Tv = VSUB(T5, Ta); Tw = VSUB(Tg, Tl); Tx = VBYI(VFMA(LDK(KP951056516), Tv, VMUL(LDK(KP587785252), Tw))); Tz = VBYI(VFNMS(LDK(KP587785252), Tv, VMUL(LDK(KP951056516), Tw))); ST(&(x[WS(rs, 5)]), VADD(Tr, Ts), ms, &(x[WS(rs, 1)])); Ty = VSUB(Tt, Tn); ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)])); Tu = VADD(Tn, Tt); ST(&(x[WS(rs, 1)]), VSUB(Tu, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(Tx, Tu), ms, &(x[WS(rs, 1)])); } { V TM, TK, TL, TG, TO, TC, TF, TP, TN; TM = VMUL(LDK(KP559016994), VSUB(TI, TJ)); TK = VADD(TI, TJ); TL = VFNMS(LDK(KP250000000), TK, TH); TC = VSUB(TA, TB); TF = VSUB(TD, TE); TG = VBYI(VFNMS(LDK(KP587785252), TF, VMUL(LDK(KP951056516), TC))); TO = VBYI(VFMA(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC))); ST(&(x[0]), VADD(TH, TK), ms, &(x[0])); TP = VADD(TM, TL); ST(&(x[WS(rs, 4)]), VADD(TO, TP), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VSUB(TP, TO), ms, &(x[0])); TN = VSUB(TL, TM); ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0])); } } } VLEAVE();}
开发者ID:Aegisub，项目名称:fftw3，代码行数:96，
示例22: t2fv_8static void t2fv_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 = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(rs)) { V T1, T2, Th, Tj, T5, T7, Ta, Tc; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Ti, Tk, T6, T8, Tb, Td; T3 = BYTWJ(&(W[TWVL * 6]), T2); Ti = BYTWJ(&(W[TWVL * 2]), Th); Tk = BYTWJ(&(W[TWVL * 10]), Tj); T6 = BYTWJ(&(W[0]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Td = BYTWJ(&(W[TWVL * 4]), Tc); { V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts; Tq = VADD(T1, T3); T4 = VSUB(T1, T3); Tr = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); Tt = VADD(T6, T8); T9 = VSUB(T6, T8); Tu = VADD(Tb, Td); Te = VSUB(Tb, Td); Tw = VSUB(Tq, Tr); Ts = VADD(Tq, Tr); { V Tx, Tv, Tm, Tf; Tx = VSUB(Tu, Tt); Tv = VADD(Tt, Tu); Tm = VSUB(Te, T9); Tf = VADD(T9, Te); { V Tp, Tn, To, Tg; ST(&(x[WS(rs, 2)]), VFMAI(Tx, Tw), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Tx, Tw), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); Tp = VFMA(LDK(KP707106781), Tm, Tl); Tn = VFNMS(LDK(KP707106781), Tm, Tl); To = VFNMS(LDK(KP707106781), Tf, T4); Tg = VFMA(LDK(KP707106781), Tf, T4); ST(&(x[WS(rs, 5)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE();}
开发者ID:dstuck，项目名称:tinker_integrated_PIMC，代码行数:66，
示例23: q1fv_4static 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:barnex，项目名称:fftw，代码行数:85，
示例24: t1fuv_7static 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(7, 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])); } } } } } } VLEAVE();}
开发者ID:Aegisub，项目名称:fftw3，代码行数:71，
注：本文中的BYTWJ函数示例整理自Github/MSDocs等源码及文档管理平台，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。
C++ BZ2_bzBuffToBuffDecompress函数代码示例
C++ BYTW函数代码示例 |
