本文整理汇总了C++中LDK函数的典型用法代码示例。如果您正苦于以下问题：C++ LDK函数的具体用法？C++ LDK怎么用？C++ LDK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LDK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: t1fuv_10
static void t1fuv_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 = 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 Td, TA, T4, Ta, Tk, TE, Tp, TF, TB, T9, T1, T2, Tb;
T1 = LD(&(x[0]), ms, &(x[0]));
T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
{
V Tg, Tn, Ti, Tl;
Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tn = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Ti = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
{
V T6, T8, T5, Tc;
T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
{
V T3, Th, To, Tj, Tm, T7;
T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T3 = BYTWJ(&(W[TWVL * 8]), T2);
Th = BYTWJ(&(W[TWVL * 6]), Tg);
To = BYTWJ(&(W[0]), Tn);
Tj = BYTWJ(&(W[TWVL * 16]), Ti);
Tm = BYTWJ(&(W[TWVL * 10]), Tl);
T6 = BYTWJ(&(W[TWVL * 2]), T5);
Td = BYTWJ(&(W[TWVL * 4]), Tc);
T8 = BYTWJ(&(W[TWVL * 12]), T7);
TA = VADD(T1, T3);
T4 = VSUB(T1, T3);
Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Tk = VSUB(Th, Tj);
TE = VADD(Th, Tj);
Tp = VSUB(Tm, To);
TF = VADD(Tm, To);
}
TB = VADD(T6, T8);
T9 = VSUB(T6, T8);
}
}
Tb = BYTWJ(&(W[TWVL * 14]), Ta);
{
V TL, TG, Tw, Tq, TC, Te;
TL = VSUB(TE, TF);
TG = VADD(TE, TF);
Tw = VSUB(Tk, Tp);
Tq = VADD(Tk, Tp);
TC = VADD(Tb, Td);
Te = VSUB(Tb, Td);
{
V TM, TD, Tv, Tf;
TM = VSUB(TB, TC);
TD = VADD(TB, TC);
Tv = VSUB(T9, Te);
Tf = VADD(T9, Te);
{
V TP, TN, TH, TJ, Tz, Tx, Tr, Tt, TI, Ts;
TP = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TL, TM));
TN = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TM, TL));
TH = VADD(TD, TG);
TJ = VSUB(TD, TG);
Tz = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tv, Tw));
Tx = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tv));
Tr = VADD(Tf, Tq);
Tt = VSUB(Tf, Tq);
ST(&(x[0]), VADD(TA, TH), ms, &(x[0]));
TI = VFNMS(LDK(KP250000000), TH, TA);
ST(&(x[WS(rs, 5)]), VADD(T4, Tr), ms, &(x[WS(rs, 1)]));
Ts = VFNMS(LDK(KP250000000), Tr, T4);
{
V TK, TO, Tu, Ty;
TK = VFNMS(LDK(KP559016994), TJ, TI);
TO = VFMA(LDK(KP559016994), TJ, TI);
Tu = VFMA(LDK(KP559016994), Tt, Ts);
Ty = VFNMS(LDK(KP559016994), Tt, Ts);
ST(&(x[WS(rs, 8)]), VFNMSI(TN, TK), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(TN, TK), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(TP, TO), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0]));
ST(&(x[WS(rs, 9)]), VFMAI(Tx, Tu), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFNMSI(Tx, Tu), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(Tz, Ty), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFNMSI(Tz, Ty), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
VLEAVE();
}示例2: n1fv_16
static void n1fv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) {
V Tp, T13, Tu, TN, Tm, T14, Tv, TY, T7, T17, Ty, TT, Te, T16, Tx;
V TQ;
{
V Tn, To, TM, Ts, Tt, TL;
Tn = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
To = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
TM = VADD(Tn, To);
Ts = LD(&(xi[0]), ivs, &(xi[0]));
Tt = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
TL = VADD(Ts, Tt);
Tp = VSUB(Tn, To);
T13 = VADD(TL, TM);
Tu = VSUB(Ts, Tt);
TN = VSUB(TL, TM);
}
{
V Ti, TW, Tl, TX;
{
V Tg, Th, Tj, Tk;
Tg = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Ti = VSUB(Tg, Th);
TW = VADD(Tg, Th);
Tj = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Tk = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tl = VSUB(Tj, Tk);
TX = VADD(Tj, Tk);
}
Tm = VMUL(LDK(KP707106781), VSUB(Ti, Tl));
T14 = VADD(TX, TW);
Tv = VMUL(LDK(KP707106781), VADD(Tl, Ti));
TY = VSUB(TW, TX);
}
{
V T3, TR, T6, TS;
{
V T1, T2, T4, T5;
T1 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
TR = VADD(T1, T2);
T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T5 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
T6 = VSUB(T4, T5);
TS = VADD(T4, T5);
}
T7 = VFNMS(LDK(KP923879532), T6, VMUL(LDK(KP382683432), T3));
T17 = VADD(TR, TS);
Ty = VFMA(LDK(KP923879532), T3, VMUL(LDK(KP382683432), T6));
TT = VSUB(TR, TS);
}
{
V Ta, TO, Td, TP;
{
V T8, T9, Tb, Tc;
T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Ta = VSUB(T8, T9);
TO = VADD(T8, T9);
Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Td = VSUB(Tb, Tc);
TP = VADD(Tb, Tc);
}
Te = VFMA(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), Td));
T16 = VADD(TO, TP);
Tx = VFNMS(LDK(KP382683432), Td, VMUL(LDK(KP923879532), Ta));
TQ = VSUB(TO, TP);
}
{
V T15, T18, T19, T1a;
T15 = VADD(T13, T14);
T18 = VADD(T16, T17);
ST(&(xo[WS(os, 8)]), VSUB(T15, T18), ovs, &(xo[0]));
ST(&(xo[0]), VADD(T15, T18), ovs, &(xo[0]));
T19 = VSUB(T13, T14);
T1a = VBYI(VSUB(T17, T16));
ST(&(xo[WS(os, 12)]), VSUB(T19, T1a), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VADD(T19, T1a), ovs, &(xo[0]));
}
{
V TV, T11, T10, T12, TU, TZ;
TU = VMUL(LDK(KP707106781), VADD(TQ, TT));
TV = VADD(TN, TU);
T11 = VSUB(TN, TU);
TZ = VMUL(LDK(KP707106781), VSUB(TT, TQ));
T10 = VBYI(VADD(TY, TZ));
T12 = VBYI(VSUB(TZ, TY));
//.........这里部分代码省略.........
示例3: n1bv_14
static void n1bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
DVK(KP222520933, +0.222520933956314404288902564496794759466355569);
DVK(KP623489801, +0.623489801858733530525004884004239810632274731);
DVK(KP781831482, +0.781831482468029808708444526674057750232334519);
DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
DVK(KP433883739, +0.433883739117558120475768332848358754609990728);
{
INT i;
const R *xi;
R *xo;
xi = ii;
xo = io;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V Tp, Ty, Tl, TL, Tq, TE, T7, TJ, Ts, TB, Te, TK, Tr, TH, Tn;
V To;
Tn = LD(&(xi[0]), ivs, &(xi[0]));
To = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Tp = VSUB(Tn, To);
Ty = VADD(Tn, To);
{
V Th, TC, Tk, TD;
{
V Tf, Tg, Ti, Tj;
Tf = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Th = VSUB(Tf, Tg);
TC = VADD(Tf, Tg);
Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tj = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tk = VSUB(Ti, Tj);
TD = VADD(Ti, Tj);
}
Tl = VSUB(Th, Tk);
TL = VSUB(TD, TC);
Tq = VADD(Th, Tk);
TE = VADD(TC, TD);
}
{
V T3, Tz, T6, TA;
{
V T1, T2, T4, T5;
T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
Tz = VADD(T1, T2);
T4 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T6 = VSUB(T4, T5);
TA = VADD(T4, T5);
}
T7 = VSUB(T3, T6);
TJ = VSUB(Tz, TA);
Ts = VADD(T3, T6);
TB = VADD(Tz, TA);
}
{
V Ta, TF, Td, TG;
{
V T8, T9, Tb, Tc;
T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T9 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Ta = VSUB(T8, T9);
TF = VADD(T8, T9);
Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Td = VSUB(Tb, Tc);
TG = VADD(Tb, Tc);
}
Te = VSUB(Ta, Td);
TK = VSUB(TG, TF);
Tr = VADD(Ta, Td);
TH = VADD(TF, TG);
}
ST(&(xo[WS(os, 7)]), VADD(Tp, VADD(Ts, VADD(Tq, Tr))), ovs, &(xo[WS(os, 1)]));
ST(&(xo[0]), VADD(Ty, VADD(TB, VADD(TE, TH))), ovs, &(xo[0]));
{
V Tm, Tt, TQ, TP;
Tm = VBYI(VFMA(LDK(KP433883739), T7, VFNMS(LDK(KP781831482), Tl, VMUL(LDK(KP974927912), Te))));
Tt = VFMA(LDK(KP623489801), Tq, VFNMS(LDK(KP222520933), Tr, VFNMS(LDK(KP900968867), Ts, Tp)));
ST(&(xo[WS(os, 3)]), VADD(Tm, Tt), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 11)]), VSUB(Tt, Tm), ovs, &(xo[WS(os, 1)]));
TQ = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK))));
TP = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TB, Ty)));
ST(&(xo[WS(os, 12)]), VSUB(TP, TQ), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VADD(TP, TQ), ovs, &(xo[0]));
}
{
V Tu, Tv, TM, TI;
Tu = VBYI(VFMA(LDK(KP781831482), T7, VFMA(LDK(KP974927912), Tl, VMUL(LDK(KP433883739), Te))));
Tv = VFMA(LDK(KP623489801), Ts, VFNMS(LDK(KP900968867), Tr, VFNMS(LDK(KP222520933), Tq, Tp)));
ST(&(xo[WS(os, 1)]), VADD(Tu, Tv), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 13)]), VSUB(Tv, Tu), ovs, &(xo[WS(os, 1)]));
TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ))));
TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TE, Ty)));
ST(&(xo[WS(os, 6)]), VSUB(TI, TM), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VADD(TI, TM), ovs, &(xo[0]));
}
{
//.........这里部分代码省略.........
示例4: n1fv_10
static void n1fv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V Ti, Ty, Tm, Tn, Tw, Tt, Tz, TA, TB, T7, Te, Tj, Tg, Th;
Tg = LD(&(xi[0]), ivs, &(xi[0]));
Th = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Ti = VSUB(Tg, Th);
Ty = VADD(Tg, Th);
{
V T3, Tu, Td, Ts, T6, Tv, Ta, Tr;
{
V T1, T2, Tb, Tc;
T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
Tu = VADD(T1, T2);
Tb = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Td = VSUB(Tb, Tc);
Ts = VADD(Tb, Tc);
}
{
V T4, T5, T8, T9;
T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T6 = VSUB(T4, T5);
Tv = VADD(T4, T5);
T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Ta = VSUB(T8, T9);
Tr = VADD(T8, T9);
}
Tm = VSUB(T3, T6);
Tn = VSUB(Ta, Td);
Tw = VSUB(Tu, Tv);
Tt = VSUB(Tr, Ts);
Tz = VADD(Tu, Tv);
TA = VADD(Tr, Ts);
TB = VADD(Tz, TA);
T7 = VADD(T3, T6);
Te = VADD(Ta, Td);
Tj = VADD(T7, Te);
}
ST(&(xo[WS(os, 5)]), VADD(Ti, Tj), ovs, &(xo[WS(os, 1)]));
ST(&(xo[0]), VADD(Ty, TB), ovs, &(xo[0]));
{
V To, Tq, Tl, Tp, Tf, Tk;
To = VBYI(VFMA(LDK(KP951056516), Tm, VMUL(LDK(KP587785252), Tn)));
Tq = VBYI(VFNMS(LDK(KP587785252), Tm, VMUL(LDK(KP951056516), Tn)));
Tf = VMUL(LDK(KP559016994), VSUB(T7, Te));
Tk = VFNMS(LDK(KP250000000), Tj, Ti);
Tl = VADD(Tf, Tk);
Tp = VSUB(Tk, Tf);
ST(&(xo[WS(os, 1)]), VSUB(Tl, To), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VADD(Tq, Tp), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VADD(To, Tl), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VSUB(Tp, Tq), ovs, &(xo[WS(os, 1)]));
}
{
V Tx, TF, TE, TG, TC, TD;
Tx = VBYI(VFNMS(LDK(KP587785252), Tw, VMUL(LDK(KP951056516), Tt)));
TF = VBYI(VFMA(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt)));
TC = VFNMS(LDK(KP250000000), TB, Ty);
TD = VMUL(LDK(KP559016994), VSUB(Tz, TA));
TE = VSUB(TC, TD);
TG = VADD(TD, TC);
ST(&(xo[WS(os, 2)]), VADD(Tx, TE), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VSUB(TG, TF), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VSUB(TE, Tx), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VADD(TF, TG), ovs, &(xo[0]));
}
}
}示例5: t1bv_16
static void t1bv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
INT m;
R *x;
x = ii;
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(rs)) {
V TJ, T1b, TD, T1c, T17, T18, Ty, TK, T10, T11, T12, Tb, TM, T13, T14;
V T15, Tm, TN, TG, TI, TH;
TG = LD(&(x[0]), ms, &(x[0]));
TH = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
TI = BYTW(&(W[TWVL * 14]), TH);
TJ = VSUB(TG, TI);
T1b = VADD(TG, TI);
{
V TA, TC, Tz, TB;
Tz = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
TA = BYTW(&(W[TWVL * 6]), Tz);
TB = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
TC = BYTW(&(W[TWVL * 22]), TB);
TD = VSUB(TA, TC);
T1c = VADD(TA, TC);
}
{
V Tp, Tw, Tr, Tu, Ts, Tx;
{
V To, Tv, Tq, Tt;
To = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tp = BYTW(&(W[TWVL * 2]), To);
Tv = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tw = BYTW(&(W[TWVL * 10]), Tv);
Tq = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
Tr = BYTW(&(W[TWVL * 18]), Tq);
Tt = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
Tu = BYTW(&(W[TWVL * 26]), Tt);
}
T17 = VADD(Tp, Tr);
T18 = VADD(Tu, Tw);
Ts = VSUB(Tp, Tr);
Tx = VSUB(Tu, Tw);
Ty = VMUL(LDK(KP707106781), VSUB(Ts, Tx));
TK = VMUL(LDK(KP707106781), VADD(Ts, Tx));
}
{
V T2, T9, T4, T7, T5, Ta;
{
V T1, T8, T3, T6;
T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
T2 = BYTW(&(W[0]), T1);
T8 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
T9 = BYTW(&(W[TWVL * 24]), T8);
T3 = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
T4 = BYTW(&(W[TWVL * 16]), T3);
T6 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
T7 = BYTW(&(W[TWVL * 8]), T6);
}
T10 = VADD(T2, T4);
T11 = VADD(T7, T9);
T12 = VSUB(T10, T11);
T5 = VSUB(T2, T4);
Ta = VSUB(T7, T9);
Tb = VFNMS(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), T5));
TM = VFMA(LDK(KP382683432), T5, VMUL(LDK(KP923879532), Ta));
}
{
V Td, Tk, Tf, Ti, Tg, Tl;
{
V Tc, Tj, Te, Th;
Tc = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
Td = BYTW(&(W[TWVL * 28]), Tc);
Tj = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
Tk = BYTW(&(W[TWVL * 20]), Tj);
Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Tf = BYTW(&(W[TWVL * 12]), Te);
Th = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Ti = BYTW(&(W[TWVL * 4]), Th);
}
T13 = VADD(Td, Tf);
T14 = VADD(Ti, Tk);
T15 = VSUB(T13, T14);
Tg = VSUB(Td, Tf);
Tl = VSUB(Ti, Tk);
Tm = VFMA(LDK(KP923879532), Tg, VMUL(LDK(KP382683432), Tl));
TN = VFNMS(LDK(KP382683432), Tg, VMUL(LDK(KP923879532), Tl));
}
{
V T1a, T1g, T1f, T1h;
{
V T16, T19, T1d, T1e;
T16 = VMUL(LDK(KP707106781), VSUB(T12, T15));
T19 = VSUB(T17, T18);
T1a = VBYI(VSUB(T16, T19));
T1g = VBYI(VADD(T19, T16));
T1d = VSUB(T1b, T1c);
T1e = VMUL(LDK(KP707106781), VADD(T12, T15));
T1f = VSUB(T1d, T1e);
T1h = VADD(T1d, T1e);
//.........这里部分代码省略.........
示例6: hc2cbdftv_16
static void hc2cbdftv_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
INT m;
for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 30)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(64, rs)) {
V Tf, T16, TZ, T1C, TI, T1a, TV, T1D, T1F, T1G, Ty, T19, TC, T17, TS;
V T10;
{
V T2, TD, T4, TF, Tc, Tb, Td, T6, T8, T9, T3, TE, Ta, T7, T5;
V Te, TX, TY, TG, TH, TT, TU, Tj, TM, Tw, TQ, Tn, TN, Ts, TP;
V Tg, Ti, Th, Tt, Tv, Tu, Tk, Tm, Tl, Tr, Tq, Tp, To, Tx, TA;
V TB, TO, TR;
T2 = LD(&(Rp[0]), ms, &(Rp[0]));
TD = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0]));
T3 = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)]));
T4 = VCONJ(T3);
TE = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)]));
TF = VCONJ(TE);
Tc = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0]));
Ta = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
Tb = VCONJ(Ta);
Td = VSUB(Tb, Tc);
T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0]));
T7 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)]));
T8 = VCONJ(T7);
T9 = VSUB(T6, T8);
T5 = VSUB(T2, T4);
Te = VMUL(LDK(KP707106781), VADD(T9, Td));
Tf = VADD(T5, Te);
T16 = VSUB(T5, Te);
TX = VADD(T2, T4);
TY = VADD(TD, TF);
TZ = VSUB(TX, TY);
T1C = VADD(TX, TY);
TG = VSUB(TD, TF);
TH = VMUL(LDK(KP707106781), VSUB(T9, Td));
TI = VADD(TG, TH);
T1a = VSUB(TH, TG);
TT = VADD(T6, T8);
TU = VADD(Tb, Tc);
TV = VSUB(TT, TU);
T1D = VADD(TT, TU);
Tg = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
Th = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0]));
Ti = VCONJ(Th);
Tj = VSUB(Tg, Ti);
TM = VADD(Tg, Ti);
Tt = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)]));
Tu = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0]));
Tv = VCONJ(Tu);
Tw = VSUB(Tt, Tv);
TQ = VADD(Tt, Tv);
Tk = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)]));
Tl = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0]));
Tm = VCONJ(Tl);
Tn = VSUB(Tk, Tm);
TN = VADD(Tk, Tm);
Tr = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)]));
Tp = LD(&(Rm[0]), -ms, &(Rm[0]));
Tq = VCONJ(Tp);
Ts = VSUB(Tq, Tr);
TP = VADD(Tq, Tr);
T1F = VADD(TM, TN);
T1G = VADD(TP, TQ);
To = VFNMS(LDK(KP382683432), Tn, VMUL(LDK(KP923879532), Tj));
Tx = VFMA(LDK(KP923879532), Ts, VMUL(LDK(KP382683432), Tw));
Ty = VADD(To, Tx);
T19 = VSUB(To, Tx);
TA = VFMA(LDK(KP382683432), Tj, VMUL(LDK(KP923879532), Tn));
TB = VFNMS(LDK(KP382683432), Ts, VMUL(LDK(KP923879532), Tw));
TC = VADD(TA, TB);
T17 = VSUB(TA, TB);
TO = VSUB(TM, TN);
TR = VSUB(TP, TQ);
TS = VMUL(LDK(KP707106781), VSUB(TO, TR));
T10 = VMUL(LDK(KP707106781), VADD(TO, TR));
}
{
V T21, T1W, T1u, T20, T1I, T1O, TK, T1S, T12, T1e, T1k, T1A, T1o, T1w, T1c;
V T1M, T1U, T1V, T1T, T1s, T1t, T1r, T1Z, T1E, T1H, T1B, T1N, Tz, TJ, T1;
V T1R, TW, T11, TL, T1d, T1i, T1j, T1h, T1z, T1m, T1n, T1l, T1v, T18, T1b;
V T15, T1L, T13, T1g, T1X, T23, T14, T1f, T1Y, T22, T1p, T1y, T1J, T1Q, T1q;
V T1x, T1K, T1P;
T1U = VADD(T1C, T1D);
T1V = VADD(T1F, T1G);
T21 = VADD(T1U, T1V);
T1T = LDW(&(W[TWVL * 14]));
T1W = VZMUL(T1T, VSUB(T1U, T1V));
T1s = VADD(Tf, Ty);
T1t = VBYI(VADD(TI, TC));
T1r = LDW(&(W[TWVL * 28]));
T1u = VZMULI(T1r, VSUB(T1s, T1t));
T1Z = LDW(&(W[0]));
T20 = VZMULI(T1Z, VADD(T1s, T1t));
T1E = VSUB(T1C, T1D);
T1H = VBYI(VSUB(T1F, T1G));
T1B = LDW(&(W[TWVL * 22]));
//.........这里部分代码省略.........
示例7: n1fv_9
static void n1fv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP342020143, +0.342020143325668733044099614682259580763083368);
DVK(KP813797681, +0.813797681349373692844693217248393223289101568);
DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
DVK(KP296198132, +0.296198132726023843175338011893050938967728390);
DVK(KP642787609, +0.642787609686539326322643409907263432907559884);
DVK(KP663413948, +0.663413948168938396205421319635891297216863310);
DVK(KP556670399, +0.556670399226419366452912952047023132968291906);
DVK(KP766044443, +0.766044443118978035202392650555416673935832457);
DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
DVK(KP150383733, +0.150383733180435296639271897612501926072238258);
DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
DVK(KP173648177, +0.173648177666930348851716626769314796000375677);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T5, Ts, Tj, To, Tf, Tn, Tp, Tu, Tl, Ta, Tk, Tm, Tt;
{
V T1, T2, T3, T4;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T4 = VADD(T2, T3);
T5 = VADD(T1, T4);
Ts = VMUL(LDK(KP866025403), VSUB(T3, T2));
Tj = VFNMS(LDK(KP500000000), T4, T1);
}
{
V Tb, Te, Tc, Td;
Tb = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Te = VADD(Tc, Td);
To = VSUB(Td, Tc);
Tf = VADD(Tb, Te);
Tn = VFNMS(LDK(KP500000000), Te, Tb);
Tp = VFMA(LDK(KP173648177), Tn, VMUL(LDK(KP852868531), To));
Tu = VFNMS(LDK(KP984807753), Tn, VMUL(LDK(KP150383733), To));
}
{
V T6, T9, T7, T8;
T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T9 = VADD(T7, T8);
Tl = VSUB(T8, T7);
Ta = VADD(T6, T9);
Tk = VFNMS(LDK(KP500000000), T9, T6);
Tm = VFMA(LDK(KP766044443), Tk, VMUL(LDK(KP556670399), Tl));
Tt = VFNMS(LDK(KP642787609), Tk, VMUL(LDK(KP663413948), Tl));
}
{
V Ti, Tg, Th, Tz, TA;
Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Tf, Ta)));
Tg = VADD(Ta, Tf);
Th = VFNMS(LDK(KP500000000), Tg, T5);
ST(&(xo[0]), VADD(T5, Tg), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VADD(Th, Ti), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 6)]), VSUB(Th, Ti), ovs, &(xo[0]));
Tz = VFMA(LDK(KP173648177), Tk, VFNMS(LDK(KP296198132), To, VFNMS(LDK(KP939692620), Tn, VFNMS(LDK(KP852868531), Tl, Tj))));
TA = VBYI(VSUB(VFNMS(LDK(KP342020143), Tn, VFNMS(LDK(KP150383733), Tl, VFNMS(LDK(KP984807753), Tk, VMUL(LDK(KP813797681), To)))), Ts));
ST(&(xo[WS(os, 7)]), VSUB(Tz, TA), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VADD(Tz, TA), ovs, &(xo[0]));
{
V Tr, Tx, Tw, Ty, Tq, Tv;
Tq = VADD(Tm, Tp);
Tr = VADD(Tj, Tq);
Tx = VFMA(LDK(KP866025403), VSUB(Tt, Tu), VFNMS(LDK(KP500000000), Tq, Tj));
Tv = VADD(Tt, Tu);
Tw = VBYI(VADD(Ts, Tv));
Ty = VBYI(VADD(Ts, VFNMS(LDK(KP500000000), Tv, VMUL(LDK(KP866025403), VSUB(Tp, Tm)))));
ST(&(xo[WS(os, 8)]), VSUB(Tr, Tw), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VADD(Tx, Ty), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VADD(Tw, Tr), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 5)]), VSUB(Tx, Ty), ovs, &(xo[WS(os, 1)]));
}
}
}
}示例8: n1bv_12
static void n1bv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
INT i;
const R *xi;
R *xo;
xi = ii;
xo = io;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T5, Ta, TG, TF, Ty, Tm, Ti, Tp, TJ, TI, Tx, Ts;
{
V T1, T6, T4, Tk, T9, Tl;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
{
V T2, T3, T7, T8;
T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T4 = VADD(T2, T3);
Tk = VSUB(T2, T3);
T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T9 = VADD(T7, T8);
Tl = VSUB(T7, T8);
}
T5 = VFNMS(LDK(KP500000000), T4, T1);
Ta = VFNMS(LDK(KP500000000), T9, T6);
TG = VADD(T6, T9);
TF = VADD(T1, T4);
Ty = VADD(Tk, Tl);
Tm = VMUL(LDK(KP866025403), VSUB(Tk, Tl));
}
{
V Tn, Tq, Te, To, Th, Tr;
Tn = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tq = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
{
V Tc, Td, Tf, Tg;
Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Td = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Te = VSUB(Tc, Td);
To = VADD(Tc, Td);
Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Th = VSUB(Tf, Tg);
Tr = VADD(Tf, Tg);
}
Ti = VMUL(LDK(KP866025403), VSUB(Te, Th));
Tp = VFNMS(LDK(KP500000000), To, Tn);
TJ = VADD(Tq, Tr);
TI = VADD(Tn, To);
Tx = VADD(Te, Th);
Ts = VFNMS(LDK(KP500000000), Tr, Tq);
}
{
V TH, TK, TL, TM;
TH = VSUB(TF, TG);
TK = VBYI(VSUB(TI, TJ));
ST(&(xo[WS(os, 3)]), VSUB(TH, TK), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VADD(TH, TK), ovs, &(xo[WS(os, 1)]));
TL = VADD(TF, TG);
TM = VADD(TI, TJ);
ST(&(xo[WS(os, 6)]), VSUB(TL, TM), ovs, &(xo[0]));
ST(&(xo[0]), VADD(TL, TM), ovs, &(xo[0]));
}
{
V Tj, Tv, Tu, Tw, Tb, Tt;
Tb = VSUB(T5, Ta);
Tj = VSUB(Tb, Ti);
Tv = VADD(Tb, Ti);
Tt = VSUB(Tp, Ts);
Tu = VBYI(VADD(Tm, Tt));
Tw = VBYI(VSUB(Tt, Tm));
ST(&(xo[WS(os, 11)]), VSUB(Tj, Tu), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 5)]), VADD(Tv, Tw), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 1)]), VADD(Tj, Tu), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VSUB(Tv, Tw), ovs, &(xo[WS(os, 1)]));
}
{
V Tz, TD, TC, TE, TA, TB;
Tz = VBYI(VMUL(LDK(KP866025403), VSUB(Tx, Ty)));
TD = VBYI(VMUL(LDK(KP866025403), VADD(Ty, Tx)));
TA = VADD(T5, Ta);
TB = VADD(Tp, Ts);
TC = VSUB(TA, TB);
TE = VADD(TA, TB);
ST(&(xo[WS(os, 2)]), VADD(Tz, TC), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VSUB(TE, TD), ovs, &(xo[0]));
ST(&(xo[WS(os, 10)]), VSUB(TC, Tz), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VADD(TD, TE), ovs, &(xo[0]));
}
}
}示例9: t2bv_8
static void t2bv_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) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(rs)) {
V Tl, Tq, Tg, Tr, T5, Tt, Ta, Tu, Ti, Tk, Tj;
Ti = LD(&(x[0]), ms, &(x[0]));
Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tk = BYTW(&(W[TWVL * 6]), Tj);
Tl = VSUB(Ti, Tk);
Tq = VADD(Ti, Tk);
{
V Td, Tf, Tc, Te;
Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Td = BYTW(&(W[TWVL * 2]), Tc);
Te = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tf = BYTW(&(W[TWVL * 10]), Te);
Tg = VSUB(Td, Tf);
Tr = VADD(Td, Tf);
}
{
V T2, T4, T1, T3;
T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
T2 = BYTW(&(W[0]), T1);
T3 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
T4 = BYTW(&(W[TWVL * 8]), T3);
T5 = VSUB(T2, T4);
Tt = VADD(T2, T4);
}
{
V T7, T9, T6, T8;
T6 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T7 = BYTW(&(W[TWVL * 12]), T6);
T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T9 = BYTW(&(W[TWVL * 4]), T8);
Ta = VSUB(T7, T9);
Tu = VADD(T7, T9);
}
{
V Ts, Tv, Tw, Tx;
Ts = VSUB(Tq, Tr);
Tv = VBYI(VSUB(Tt, Tu));
ST(&(x[WS(rs, 6)]), VSUB(Ts, Tv), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VADD(Ts, Tv), ms, &(x[0]));
Tw = VADD(Tq, Tr);
Tx = VADD(Tt, Tu);
ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0]));
ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0]));
{
V Th, To, Tn, Tp, Tb, Tm;
Tb = VMUL(LDK(KP707106781), VSUB(T5, Ta));
Th = VBYI(VSUB(Tb, Tg));
To = VBYI(VADD(Tg, Tb));
Tm = VMUL(LDK(KP707106781), VADD(T5, Ta));
Tn = VSUB(Tl, Tm);
Tp = VADD(Tl, Tm);
ST(&(x[WS(rs, 3)]), VADD(Th, Tn), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VSUB(Tp, To), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 5)]), VSUB(Tn, Th), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VADD(To, Tp), ms, &(x[WS(rs, 1)]));
}
}
}
}示例10: t1fuv_8
static 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)]));
}
}
}
}示例11: DVK
static const R *t3fv_16(R *ri, R *ii, const R *W, stride ios, INT m, INT dist)
{
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP414213562, +0.414213562373095048801688724209698078569671875);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
INT i;
R *x;
x = ri;
for (i = m; i > 0; i = i - VL, x = x + (VL * dist), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(ios)) {
V T13, Tg, TY, T14, T1A, T1q, T1f, T1x, T1r, T1i, Tt, T16, TB, T1j, T1k;
V TH;
{
V T2, T8, Tu, T3;
T2 = LDW(&(W[0]));
T8 = LDW(&(W[TWVL * 2]));
Tu = LDW(&(W[TWVL * 6]));
T3 = LDW(&(W[TWVL * 4]));
{
V Ty, T1o, Tf, T1b, T7, Tr, TR, TX, T1g, Tl, To, Tw, TG, Tz, T1p;
V T1e, TC;
{
V T1, T5, Ta, Td;
T1 = LD(&(x[0]), dist, &(x[0]));
T5 = LD(&(x[WS(ios, 8)]), dist, &(x[0]));
Ta = LD(&(x[WS(ios, 4)]), dist, &(x[0]));
Td = LD(&(x[WS(ios, 12)]), dist, &(x[0]));
{
V Tx, TO, TE, Tb, Tm, Tp, TN, Te, T6, TW, TP, TS;
{
V TM, T9, TL, Tc, TU, T4, TV;
TM = LD(&(x[WS(ios, 14)]), dist, &(x[0]));
Tx = VZMULJ(T2, T8);
T9 = VZMUL(T2, T8);
TL = VZMULJ(T2, Tu);
TO = VZMULJ(T8, T3);
Tc = VZMUL(T8, T3);
TU = VZMUL(T2, T3);
T4 = VZMULJ(T2, T3);
TV = LD(&(x[WS(ios, 10)]), dist, &(x[0]));
TE = VZMUL(Tx, T3);
Ty = VZMULJ(Tx, T3);
Tb = VZMULJ(T9, Ta);
Tm = VZMULJ(T9, T3);
Tp = VZMUL(T9, T3);
TN = VZMULJ(TL, TM);
Te = VZMULJ(Tc, Td);
T6 = VZMULJ(T4, T5);
TW = VZMULJ(TU, TV);
}
TP = LD(&(x[WS(ios, 6)]), dist, &(x[0]));
TS = LD(&(x[WS(ios, 2)]), dist, &(x[0]));
{
V TQ, TT, Ti, Tk, Tn, Th, Tq, Tj;
Th = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)]));
Tq = LD(&(x[WS(ios, 13)]), dist, &(x[WS(ios, 1)]));
Tj = LD(&(x[WS(ios, 9)]), dist, &(x[WS(ios, 1)]));
T1o = VSUB(Tb, Te);
Tf = VADD(Tb, Te);
T1b = VSUB(T1, T6);
T7 = VADD(T1, T6);
TQ = VZMULJ(TO, TP);
TT = VZMULJ(Tx, TS);
Ti = VZMULJ(T2, Th);
Tr = VZMULJ(Tp, Tq);
Tk = VZMULJ(T3, Tj);
Tn = LD(&(x[WS(ios, 5)]), dist, &(x[WS(ios, 1)]));
{
V T1d, T1c, Tv, TF;
Tv = LD(&(x[WS(ios, 15)]), dist, &(x[WS(ios, 1)]));
TF = LD(&(x[WS(ios, 11)]), dist, &(x[WS(ios, 1)]));
T1d = VSUB(TN, TQ);
TR = VADD(TN, TQ);
T1c = VSUB(TT, TW);
TX = VADD(TT, TW);
T1g = VSUB(Ti, Tk);
Tl = VADD(Ti, Tk);
To = VZMULJ(Tm, Tn);
Tw = VZMULJ(Tu, Tv);
TG = VZMULJ(TE, TF);
Tz = LD(&(x[WS(ios, 7)]), dist, &(x[WS(ios, 1)]));
T1p = VSUB(T1d, T1c);
T1e = VADD(T1c, T1d);
TC = LD(&(x[WS(ios, 3)]), dist, &(x[WS(ios, 1)]));
}
}
}
}
{
V T1h, Ts, TA, TD;
T13 = VADD(T7, Tf);
Tg = VSUB(T7, Tf);
T1h = VSUB(To, Tr);
Ts = VADD(To, Tr);
TY = VSUB(TR, TX);
T14 = VADD(TX, TR);
TA = VZMULJ(Ty, Tz);
T1A = VFMA(LDK(KP707106781), T1p, T1o);
T1q = VFNMS(LDK(KP707106781), T1p, T1o);
T1f = VFMA(LDK(KP707106781), T1e, T1b);
T1x = VFNMS(LDK(KP707106781), T1e, T1b);
//.........这里部分代码省略.........
示例12: t3fv_10
static void t3fv_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 = 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(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 = VZMULJ(T2, Tw);
Tr = VZMULJ(T5, Tq);
Tk = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Td = VZMULJ(Ta, T5);
Tp = VZMULJ(T4, To);
T6 = VZMULJ(T4, T5);
Tv = VZMULJ(Tt, Tu);
Tc = VZMULJ(Ta, Tb);
Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Ti = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Tl = VZMULJ(T3, Tk);
}
{
V TN, Ts, T8, Ty, TO, Tf, Tj;
TN = VADD(Tp, Tr);
Ts = VSUB(Tp, Tr);
T8 = VZMULJ(T6, T7);
Ty = VSUB(Tv, Tx);
TO = VADD(Tv, Tx);
Tf = VZMULJ(Td, Te);
Tj = VZMULJ(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)]), VFNMSI(TW, TT), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(TW, TT), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(TY, TX), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VFMAI(TY, TX), ms, &(x[0]));
ST(&(x[WS(rs, 9)]), VFMAI(TG, TD), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFNMSI(TG, TD), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(TI, TH), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFNMSI(TI, TH), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
//.........这里部分代码省略.........
示例13: m1fv_32_0
static void m1fv_32_0(const R *xi, R *xo, stride is, stride os, int ivs, int ovs)
{
DVK(KP195090322, +0.195090322016128267848284868477022240927691618);
DVK(KP980785280, +0.980785280403230449126182236134239036973933731);
DVK(KP555570233, +0.555570233019602224742830813948532874374937191);
DVK(KP831469612, +0.831469612302545237078788377617905756738560812);
DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
V T1T, T1W, T2K, T2x, T16, T1A, Tb, T1p, TT, T1v, TY, T1w, T27, T2a, T2b;
V T2H, T2O, TC, T1s, TH, T1t, T20, T23, T24, T2E, T2N, T2g, T2j, Tq, T1B;
V T19, T1q, T2A, T2L;
{
V T3, T1R, T15, T1S, T6, T1U, T9, T1V, T12, Ta;
{
V T1, T2, T13, T14;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
T3 = VSUB(T1, T2);
T1R = VADD(T1, T2);
T13 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T14 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0]));
T15 = VSUB(T13, T14);
T1S = VADD(T13, T14);
}
{
V T4, T5, T7, T8;
T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0]));
T6 = VSUB(T4, T5);
T1U = VADD(T4, T5);
T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T9 = VSUB(T7, T8);
T1V = VADD(T7, T8);
}
T1T = VADD(T1R, T1S);
T1W = VADD(T1U, T1V);
T2K = VSUB(T1V, T1U);
T2x = VSUB(T1R, T1S);
T12 = VMUL(LDK(KP707106781), VSUB(T9, T6));
T16 = VSUB(T12, T15);
T1A = VADD(T15, T12);
Ta = VMUL(LDK(KP707106781), VADD(T6, T9));
Tb = VADD(T3, Ta);
T1p = VSUB(T3, Ta);
}
{
V TL, T25, TX, T26, TO, T28, TR, T29;
{
V TJ, TK, TV, TW;
TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)]));
TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
TL = VSUB(TJ, TK);
T25 = VADD(TJ, TK);
TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
TW = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)]));
TX = VSUB(TV, TW);
T26 = VADD(TV, TW);
}
{
V TM, TN, TP, TQ;
TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
TO = VSUB(TM, TN);
T28 = VADD(TM, TN);
TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)]));
TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
TR = VSUB(TP, TQ);
T29 = VADD(TP, TQ);
}
{
V TS, TU, T2F, T2G;
TS = VMUL(LDK(KP707106781), VADD(TO, TR));
TT = VADD(TL, TS);
T1v = VSUB(TL, TS);
TU = VMUL(LDK(KP707106781), VSUB(TR, TO));
TY = VSUB(TU, TX);
T1w = VADD(TX, TU);
T27 = VADD(T25, T26);
T2a = VADD(T28, T29);
T2b = VSUB(T27, T2a);
T2F = VSUB(T25, T26);
T2G = VSUB(T29, T28);
T2H = VFNMS(LDK(KP382683432), T2G, VMUL(LDK(KP923879532), T2F));
T2O = VFMA(LDK(KP382683432), T2F, VMUL(LDK(KP923879532), T2G));
}
}
{
V Tu, T1Y, TG, T1Z, Tx, T21, TA, T22;
{
V Ts, Tt, TE, TF;
Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
Tu = VSUB(Ts, Tt);
T1Y = VADD(Ts, Tt);
TE = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
TF = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)]));
TG = VSUB(TE, TF);
//.........这里部分代码省略.........
示例14: n2bv_8
static void n2bv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
INT i;
const R *xi;
R *xo;
xi = ii;
xo = io;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T1, T2, Tc, Td, T4, T5, T7, T8;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
{
V T3, Tj, Te, Tk, T6, Tm, T9, Tn, Tp, Tl;
T3 = VSUB(T1, T2);
Tj = VADD(T1, T2);
Te = VSUB(Tc, Td);
Tk = VADD(Tc, Td);
T6 = VSUB(T4, T5);
Tm = VADD(T4, T5);
T9 = VSUB(T7, T8);
Tn = VADD(T7, T8);
Tp = VADD(Tj, Tk);
Tl = VSUB(Tj, Tk);
{
V Tq, To, Ta, Tf;
Tq = VADD(Tm, Tn);
To = VSUB(Tm, Tn);
Ta = VADD(T6, T9);
Tf = VSUB(T6, T9);
{
V Tr, Ts, Tt, Tu, Tg, Ti, Tb, Th;
Tr = VFMAI(To, Tl);
STM2(&(xo[4]), Tr, ovs, &(xo[0]));
Ts = VFNMSI(To, Tl);
STM2(&(xo[12]), Ts, ovs, &(xo[0]));
Tt = VADD(Tp, Tq);
STM2(&(xo[0]), Tt, ovs, &(xo[0]));
Tu = VSUB(Tp, Tq);
STM2(&(xo[8]), Tu, ovs, &(xo[0]));
Tg = VFNMS(LDK(KP707106781), Tf, Te);
Ti = VFMA(LDK(KP707106781), Tf, Te);
Tb = VFNMS(LDK(KP707106781), Ta, T3);
Th = VFMA(LDK(KP707106781), Ta, T3);
{
V Tv, Tw, Tx, Ty;
Tv = VFNMSI(Ti, Th);
STM2(&(xo[14]), Tv, ovs, &(xo[2]));
STN2(&(xo[12]), Ts, Tv, ovs);
Tw = VFMAI(Ti, Th);
STM2(&(xo[2]), Tw, ovs, &(xo[2]));
STN2(&(xo[0]), Tt, Tw, ovs);
Tx = VFMAI(Tg, Tb);
STM2(&(xo[10]), Tx, ovs, &(xo[2]));
STN2(&(xo[8]), Tu, Tx, ovs);
Ty = VFNMSI(Tg, Tb);
STM2(&(xo[6]), Ty, ovs, &(xo[2]));
STN2(&(xo[4]), Tr, Ty, ovs);
}
}
}
}
}
}示例15: t1fv_12
static 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 Tq, Ti, T7, TQ, Tu, TA, TU, Tk, TR, Tf, TE, TM;
{
V T9, TC, Tj, TD, Te;
{
V T1, T4, T2, Tm, Tx, To;
T1 = LD(&(x[0]), ms, &(x[0]));
T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
{
V T5, T3, Tn, Ty, Tp, Td, Tb, T8, Tc, Ta;
T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
T5 = BYTWJ(&(W[TWVL * 14]), T4);
T3 = BYTWJ(&(W[TWVL * 6]), T2);
Tn = BYTWJ(&(W[0]), Tm);
Ty = BYTWJ(&(W[TWVL * 16]), Tx);
Tp = BYTWJ(&(W[TWVL * 8]), To);
T9 = BYTWJ(&(W[TWVL * 10]), T8);
Td = BYTWJ(&(W[TWVL * 2]), Tc);
Tb = BYTWJ(&(W[TWVL * 18]), Ta);
{
V Th, T6, Tt, Tz;
Th = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
TC = VSUB(T5, T3);
T6 = VADD(T3, T5);
Tt = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Tz = VADD(Tn, Tp);
Tq = VSUB(Tn, Tp);
Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
TD = VSUB(Td, Tb);
Te = VADD(Tb, Td);
Ti = BYTWJ(&(W[TWVL * 20]), Th);
T7 = VFNMS(LDK(KP500000000), T6, T1);
TQ = VADD(T1, T6);
Tu = BYTWJ(&(W[TWVL * 4]), Tt);
TA = VFNMS(LDK(KP500000000), Tz, Ty);
TU = VADD(Ty, Tz);
}
}
}
Tk = BYTWJ(&(W[TWVL * 12]), Tj);
TR = VADD(T9, Te);
Tf = VFNMS(LDK(KP500000000), Te, T9);
TE = VSUB(TC, TD);
TM = VADD(TC, TD);
}
{
V Tv, Tl, TI, Tg, TW, TS;
Tv = VADD(Tk, Ti);
Tl = VSUB(Ti, Tk);
TI = VADD(T7, Tf);
Tg = VSUB(T7, Tf);
TW = VADD(TQ, TR);
TS = VSUB(TQ, TR);
{
V TT, Tw, TL, Tr;
TT = VADD(Tu, Tv);
Tw = VFNMS(LDK(KP500000000), Tv, Tu);
TL = VSUB(Tl, Tq);
Tr = VADD(Tl, Tq);
{
V TP, TN, TG, Ts, TO, TK, TH, TF;
{
V TX, TV, TJ, TB;
TX = VADD(TT, TU);
TV = VSUB(TT, TU);
TJ = VADD(Tw, TA);
TB = VSUB(Tw, TA);
TP = VMUL(LDK(KP866025403), VADD(TM, TL));
TN = VMUL(LDK(KP866025403), VSUB(TL, TM));
TG = VFNMS(LDK(KP866025403), Tr, Tg);
Ts = VFMA(LDK(KP866025403), Tr, Tg);
ST(&(x[WS(rs, 6)]), VSUB(TW, TX), ms, &(x[0]));
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, 9)]), VFNMSI(TV, TS), ms, &(x[WS(rs, 1)]));
TO = VADD(TI, TJ);
TK = VSUB(TI, TJ);
TH = VFMA(LDK(KP866025403), TE, TB);
TF = VFNMS(LDK(KP866025403), TE, TB);
}
ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0]));
ST(&(x[WS(rs, 8)]), VFNMSI(TP, TO), ms, &(x[0]));
ST(&(x[WS(rs, 10)]), VFNMSI(TN, TK), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(TN, TK), ms, &(x[0]));
ST(&(x[WS(rs, 5)]), VFNMSI(TH, TG), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(TH, TG), ms, &(x[WS(rs, 1)]));
//.........这里部分代码省略.........