本文整理汇总了C++中VFNMS函数的典型用法代码示例。如果您正苦于以下问题：C++ VFNMS函数的具体用法？C++ VFNMS怎么用？C++ VFNMS使用的例子？那么, 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VFNMS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: n1bv_10
static void n1bv_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 = 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 Tl, Ty, T7, Te, Tw, Tt, Tz, TA, TB, Tg, Th, Tm, Tj, Tk;
Tj = LD(&(xi[0]), ivs, &(xi[0]));
Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tl = VSUB(Tj, Tk);
Ty = VADD(Tj, Tk);
{
V T3, Tr, Td, Tv, T6, Ts, Ta, Tu;
{
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);
Tr = 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);
Tv = 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);
Ts = 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);
Tu = VADD(T8, T9);
}
T7 = VSUB(T3, T6);
Te = VSUB(Ta, Td);
Tw = VSUB(Tu, Tv);
Tt = VSUB(Tr, Ts);
Tz = VADD(Tr, Ts);
TA = VADD(Tu, Tv);
TB = VADD(Tz, TA);
Tg = VADD(T3, T6);
Th = VADD(Ta, Td);
Tm = VADD(Tg, Th);
}
ST(&(xo[WS(os, 5)]), VADD(Tl, Tm), ovs, &(xo[WS(os, 1)]));
ST(&(xo[0]), VADD(Ty, TB), ovs, &(xo[0]));
{
V Tf, Tq, To, Tp, Ti, Tn;
Tf = VBYI(VFMA(LDK(KP951056516), T7, VMUL(LDK(KP587785252), Te)));
Tq = VBYI(VFNMS(LDK(KP951056516), Te, VMUL(LDK(KP587785252), T7)));
Ti = VMUL(LDK(KP559016994), VSUB(Tg, Th));
Tn = VFNMS(LDK(KP250000000), Tm, Tl);
To = VADD(Ti, Tn);
Tp = VSUB(Tn, Ti);
ST(&(xo[WS(os, 1)]), VADD(Tf, To), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VADD(Tq, Tp), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VSUB(To, Tf), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VSUB(Tp, Tq), ovs, &(xo[WS(os, 1)]));
}
{
V Tx, TG, TE, TF, TC, TD;
Tx = VBYI(VFNMS(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt)));
TG = VBYI(VFMA(LDK(KP951056516), Tt, VMUL(LDK(KP587785252), Tw)));
TC = VFNMS(LDK(KP250000000), TB, Ty);
TD = VMUL(LDK(KP559016994), VSUB(Tz, TA));
TE = VSUB(TC, TD);
TF = VADD(TD, TC);
ST(&(xo[WS(os, 2)]), VADD(Tx, TE), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VADD(TG, TF), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VSUB(TE, Tx), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VSUB(TF, TG), ovs, &(xo[0]));
}
}
}示例2: t1fuv_10
static void t1fuv_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();
}示例3: n1fv_12
static void n1fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
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, Ta, TJ, Ty, Tq, Tp, Tg, Tl, TI, TA, Tz, Tu;
{
V T1, T6, T4, Tw, T9, Tx;
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);
Tw = VSUB(T3, T2);
T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T9 = VADD(T7, T8);
Tx = VSUB(T8, T7);
}
T5 = VADD(T1, T4);
Ta = VADD(T6, T9);
TJ = VADD(Tw, Tx);
Ty = VMUL(LDK(KP866025403), VSUB(Tw, Tx));
Tq = VFNMS(LDK(KP500000000), T9, T6);
Tp = VFNMS(LDK(KP500000000), T4, T1);
}
{
V Tc, Th, Tf, Ts, Tk, Tt;
Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
{
V Td, Te, Ti, Tj;
Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Tf = VADD(Td, Te);
Ts = VSUB(Te, Td);
Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tk = VADD(Ti, Tj);
Tt = VSUB(Tj, Ti);
}
Tg = VADD(Tc, Tf);
Tl = VADD(Th, Tk);
TI = VADD(Ts, Tt);
TA = VFNMS(LDK(KP500000000), Tk, Th);
Tz = VFNMS(LDK(KP500000000), Tf, Tc);
Tu = VMUL(LDK(KP866025403), VSUB(Ts, Tt));
}
{
V Tb, Tm, Tn, To;
Tb = VSUB(T5, Ta);
Tm = VBYI(VSUB(Tg, Tl));
ST(&(xo[WS(os, 9)]), VSUB(Tb, Tm), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VADD(Tb, Tm), ovs, &(xo[WS(os, 1)]));
Tn = VADD(T5, Ta);
To = VADD(Tg, Tl);
ST(&(xo[WS(os, 6)]), VSUB(Tn, To), ovs, &(xo[0]));
ST(&(xo[0]), VADD(Tn, To), ovs, &(xo[0]));
}
{
V Tv, TE, TC, TD, Tr, TB;
Tr = VSUB(Tp, Tq);
Tv = VSUB(Tr, Tu);
TE = VADD(Tr, Tu);
TB = VSUB(Tz, TA);
TC = VBYI(VADD(Ty, TB));
TD = VBYI(VSUB(Ty, TB));
ST(&(xo[WS(os, 5)]), VSUB(Tv, TC), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 11)]), VSUB(TE, TD), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VADD(TC, Tv), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 1)]), VADD(TD, TE), ovs, &(xo[WS(os, 1)]));
}
{
V TK, TM, TH, TL, TF, TG;
TK = VBYI(VMUL(LDK(KP866025403), VSUB(TI, TJ)));
TM = VBYI(VMUL(LDK(KP866025403), VADD(TJ, TI)));
TF = VADD(Tp, Tq);
TG = VADD(Tz, TA);
TH = VSUB(TF, TG);
TL = VADD(TF, TG);
ST(&(xo[WS(os, 10)]), VSUB(TH, TK), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VADD(TL, TM), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VADD(TH, TK), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VSUB(TL, TM), ovs, &(xo[0]));
}
}
}示例4: n1fv_13
static void n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP904176221, +0.904176221990848204433795481776887926501523162);
DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
DVK(KP516520780, +0.516520780623489722840901288569017135705033622);
DVK(KP522026385, +0.522026385161275033714027226654165028300441940);
DVK(KP957805992, +0.957805992594665126462521754605754580515587217);
DVK(KP600477271, +0.600477271932665282925769253334763009352012849);
DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
DVK(KP769338817, +0.769338817572980603471413688209101117038278899);
DVK(KP859542535, +0.859542535098774820163672132761689612766401925);
DVK(KP581704778, +0.581704778510515730456870384989698884939833902);
DVK(KP853480001, +0.853480001859823990758994934970528322872359049);
DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
DVK(KP226109445, +0.226109445035782405468510155372505010481906348);
DVK(KP301479260, +0.301479260047709873958013540496673347309208464);
DVK(KP686558370, +0.686558370781754340655719594850823015421401653);
DVK(KP514918778, +0.514918778086315755491789696138117261566051239);
DVK(KP038632954, +0.038632954644348171955506895830342264440241080);
DVK(KP612264650, +0.612264650376756543746494474777125408779395514);
DVK(KP302775637, +0.302775637731994646559610633735247973125648287);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
{
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T1, T7, T2, Tg, Tf, TN, Th, Tq, Ta, Tj, T5, Tr, Tk;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
{
V Td, Te, T8, T9, T3, T4;
Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T4 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Tf = VADD(Td, Te);
TN = VSUB(Td, Te);
Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Tq = VSUB(T8, T9);
Ta = VADD(T8, T9);
Tj = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T5 = VADD(T3, T4);
Tr = VSUB(T4, T3);
Tk = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
}
{
V Tt, Ti, Ty, Tb, Ts, TQ, Tx, T6, Tu, Tl;
Tt = VSUB(Tg, Th);
Ti = VADD(Tg, Th);
Ty = VFMS(LDK(KP500000000), Ta, T7);
Tb = VADD(T7, Ta);
Ts = VSUB(Tq, Tr);
TQ = VADD(Tr, Tq);
Tx = VFNMS(LDK(KP500000000), T5, T2);
T6 = VADD(T2, T5);
Tu = VSUB(Tj, Tk);
Tl = VADD(Tj, Tk);
{
V TK, Tz, Tc, TX, Tv, TO, TL, Tm;
TK = VADD(Tx, Ty);
Tz = VSUB(Tx, Ty);
Tc = VADD(T6, Tb);
TX = VSUB(T6, Tb);
Tv = VSUB(Tt, Tu);
TO = VADD(Tt, Tu);
TL = VSUB(Ti, Tl);
Tm = VADD(Ti, Tl);
{
V TF, Tw, TP, TY, TT, TM, TA, Tn;
TF = VSUB(Ts, Tv);
Tw = VADD(Ts, Tv);
TP = VFNMS(LDK(KP500000000), TO, TN);
TY = VADD(TN, TO);
TT = VFNMS(LDK(KP866025403), TL, TK);
TM = VFMA(LDK(KP866025403), TL, TK);
TA = VFNMS(LDK(KP500000000), Tm, Tf);
Tn = VADD(Tf, Tm);
{
V T1f, T1n, TI, T18, T1k, T1c, TD, T17, T10, T1m, T16, T1e, TU, TR;
TU = VFNMS(LDK(KP866025403), TQ, TP);
TR = VFMA(LDK(KP866025403), TQ, TP);
{
V TZ, T15, TE, TB;
TZ = VFMA(LDK(KP302775637), TY, TX);
T15 = VFNMS(LDK(KP302775637), TX, TY);
TE = VSUB(Tz, TA);
TB = VADD(Tz, TA);
{
V TH, To, TV, T13;
TH = VSUB(Tc, Tn);
//.........这里部分代码省略.........
示例5: n1fv_9
static void n1fv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
DVK(KP826351822, +0.826351822333069651148283373230685203999624323);
DVK(KP879385241, +0.879385241571816768108218554649462939872416269);
DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
DVK(KP666666666, +0.666666666666666666666666666666666666666666667);
DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
DVK(KP907603734, +0.907603734547952313649323976213898122064543220);
DVK(KP420276625, +0.420276625461206169731530603237061658838781920);
DVK(KP673648177, +0.673648177666930348851716626769314796000375677);
DVK(KP898197570, +0.898197570222573798468955502359086394667167570);
DVK(KP347296355, +0.347296355333860697703433253538629592000751354);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP439692620, +0.439692620785908384054109277324731469936208134);
DVK(KP203604859, +0.203604859554852403062088995281827210665664861);
DVK(KP152703644, +0.152703644666139302296566746461370407999248646);
DVK(KP586256827, +0.586256827714544512072145703099641959914944179);
DVK(KP968908795, +0.968908795874236621082202410917456709164223497);
DVK(KP726681596, +0.726681596905677465811651808188092531873167623);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T1, T2, T3, T6, Tb, T7, T8, Tc, Td, Tv, T4;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tb = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Tc = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Tv = VSUB(T3, T2);
T4 = VADD(T2, T3);
{
V Tl, T9, Tm, Te, Tj, T5;
Tl = VSUB(T7, T8);
T9 = VADD(T7, T8);
Tm = VSUB(Td, Tc);
Te = VADD(Tc, Td);
Tj = VFNMS(LDK(KP500000000), T4, T1);
T5 = VADD(T1, T4);
{
V Tn, Ta, Tk, Tf;
Tn = VFNMS(LDK(KP500000000), T9, T6);
Ta = VADD(T6, T9);
Tk = VFNMS(LDK(KP500000000), Te, Tb);
Tf = VADD(Tb, Te);
{
V Ty, TC, To, TB, Tx, Ts, Tg, Ti;
Ty = VFNMS(LDK(KP726681596), Tl, Tn);
TC = VFMA(LDK(KP968908795), Tn, Tl);
To = VFNMS(LDK(KP586256827), Tn, Tm);
TB = VFNMS(LDK(KP152703644), Tm, Tk);
Tx = VFMA(LDK(KP203604859), Tk, Tm);
Ts = VFNMS(LDK(KP439692620), Tl, Tk);
Tg = VADD(Ta, Tf);
Ti = VMUL(LDK(KP866025403), VSUB(Tf, Ta));
{
V Tz, TI, TF, TD, Tt, Th, Tq, Tp;
Tp = VFNMS(LDK(KP347296355), To, Tl);
Tz = VFMA(LDK(KP898197570), Ty, Tx);
TI = VFNMS(LDK(KP898197570), Ty, Tx);
TF = VFNMS(LDK(KP673648177), TC, TB);
TD = VFMA(LDK(KP673648177), TC, TB);
Tt = VFNMS(LDK(KP420276625), Ts, Tm);
ST(&(xo[0]), VADD(T5, Tg), ovs, &(xo[0]));
Th = VFNMS(LDK(KP500000000), Tg, T5);
Tq = VFNMS(LDK(KP907603734), Tp, Tk);
{
V TA, TJ, TE, TG, Tu, Tr, TK, TH, Tw;
TA = VFMA(LDK(KP852868531), Tz, Tj);
TJ = VFMA(LDK(KP666666666), TD, TI);
TE = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tv, TD));
TG = VFNMS(LDK(KP500000000), Tz, TF);
Tu = VFNMS(LDK(KP826351822), Tt, Tn);
ST(&(xo[WS(os, 6)]), VFNMSI(Ti, Th), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(Ti, Th), ovs, &(xo[WS(os, 1)]));
Tr = VFNMS(LDK(KP939692620), Tq, Tj);
TK = VMUL(LDK(KP866025403), VFMA(LDK(KP852868531), TJ, Tv));
ST(&(xo[WS(os, 8)]), VFMAI(TE, TA), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFNMSI(TE, TA), ovs, &(xo[WS(os, 1)]));
TH = VFMA(LDK(KP852868531), TG, Tj);
Tw = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tv, Tu));
ST(&(xo[WS(os, 4)]), VFMAI(TK, TH), ovs, &(xo[0]));
ST(&(xo[WS(os, 5)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VFMAI(Tw, Tr), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VFNMSI(Tw, Tr), ovs, &(xo[0]));
}
}
}
}
}
}
}示例6: t1bv_9
static void t1bv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
DVK(KP907603734, +0.907603734547952313649323976213898122064543220);
DVK(KP666666666, +0.666666666666666666666666666666666666666666667);
DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
DVK(KP879385241, +0.879385241571816768108218554649462939872416269);
DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
DVK(KP826351822, +0.826351822333069651148283373230685203999624323);
DVK(KP347296355, +0.347296355333860697703433253538629592000751354);
DVK(KP898197570, +0.898197570222573798468955502359086394667167570);
DVK(KP673648177, +0.673648177666930348851716626769314796000375677);
DVK(KP420276625, +0.420276625461206169731530603237061658838781920);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP586256827, +0.586256827714544512072145703099641959914944179);
DVK(KP968908795, +0.968908795874236621082202410917456709164223497);
DVK(KP726681596, +0.726681596905677465811651808188092531873167623);
DVK(KP439692620, +0.439692620785908384054109277324731469936208134);
DVK(KP203604859, +0.203604859554852403062088995281827210665664861);
DVK(KP152703644, +0.152703644666139302296566746461370407999248646);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
INT m;
R *x;
x = ii;
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, T3, T5, T9, Tn, Tb, Td, Th, Tj, Tx, T6;
T1 = LD(&(x[0]), ms, &(x[0]));
{
V T2, T4, T8, Tm;
T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
T8 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
{
V Ta, Tc, Tg, Ti;
Ta = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tc = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Ti = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T3 = BYTW(&(W[TWVL * 4]), T2);
T5 = BYTW(&(W[TWVL * 10]), T4);
T9 = BYTW(&(W[TWVL * 2]), T8);
Tn = BYTW(&(W[0]), Tm);
Tb = BYTW(&(W[TWVL * 8]), Ta);
Td = BYTW(&(W[TWVL * 14]), Tc);
Th = BYTW(&(W[TWVL * 6]), Tg);
Tj = BYTW(&(W[TWVL * 12]), Ti);
}
}
Tx = VSUB(T3, T5);
T6 = VADD(T3, T5);
{
V Tl, Te, Tk, To, T7, TN;
Tl = VSUB(Td, Tb);
Te = VADD(Tb, Td);
Tk = VSUB(Th, Tj);
To = VADD(Th, Tj);
T7 = VFNMS(LDK(KP500000000), T6, T1);
TN = VADD(T1, T6);
{
V Tf, TP, Tp, TO;
Tf = VFNMS(LDK(KP500000000), Te, T9);
TP = VADD(T9, Te);
Tp = VFNMS(LDK(KP500000000), To, Tn);
TO = VADD(Tn, To);
{
V Tz, TC, Tu, TD, TA, Tq, TQ, TS;
Tz = VFNMS(LDK(KP152703644), Tl, Tf);
TC = VFMA(LDK(KP203604859), Tf, Tl);
Tu = VFNMS(LDK(KP439692620), Tk, Tf);
TD = VFNMS(LDK(KP726681596), Tk, Tp);
TA = VFMA(LDK(KP968908795), Tp, Tk);
Tq = VFNMS(LDK(KP586256827), Tp, Tl);
TQ = VADD(TO, TP);
TS = VMUL(LDK(KP866025403), VSUB(TO, TP));
{
V TI, TB, TH, TE, Tr, TR, Tw, Tv;
Tv = VFNMS(LDK(KP420276625), Tu, Tl);
TI = VFMA(LDK(KP673648177), TA, Tz);
TB = VFNMS(LDK(KP673648177), TA, Tz);
TH = VFNMS(LDK(KP898197570), TD, TC);
TE = VFMA(LDK(KP898197570), TD, TC);
Tr = VFNMS(LDK(KP347296355), Tq, Tk);
ST(&(x[0]), VADD(TQ, TN), ms, &(x[0]));
TR = VFNMS(LDK(KP500000000), TQ, TN);
Tw = VFNMS(LDK(KP826351822), Tv, Tp);
{
V TM, TL, TF, TJ, Ts, Ty, TG, TK, Tt;
TM = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tx, TI));
TL = VFMA(LDK(KP852868531), TE, T7);
TF = VFNMS(LDK(KP500000000), TE, TB);
TJ = VFMA(LDK(KP666666666), TI, TH);
Ts = VFNMS(LDK(KP907603734), Tr, Tf);
ST(&(x[WS(rs, 6)]), VFNMSI(TS, TR), ms, &(x[0]));
ST(&(x[WS(rs, 3)]), VFMAI(TS, TR), ms, &(x[WS(rs, 1)]));
Ty = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tx, Tw));
ST(&(x[WS(rs, 8)]), VFNMSI(TM, TL), ms, &(x[0]));
ST(&(x[WS(rs, 1)]), VFMAI(TM, TL), ms, &(x[WS(rs, 1)]));
TG = VFMA(LDK(KP852868531), TF, T7);
TK = VMUL(LDK(KP866025403), VFNMS(LDK(KP852868531), TJ, Tx));
//.........这里部分代码省略.........
示例7: n1fv_32
static void n1fv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP980785280, +0.980785280403230449126182236134239036973933731);
DVK(KP831469612, +0.831469612302545237078788377617905756738560812);
DVK(KP198912367, +0.198912367379658006911597622644676228597850501);
DVK(KP668178637, +0.668178637919298919997757686523080761552472251);
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP414213562, +0.414213562373095048801688724209698078569671875);
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(64, is), MAKE_VOLATILE_STRIDE(64, os)) {
V T1h, Tr, T1a, T1k, TI, T1b, T1L, T1P, T1I, T1G, T1O, T1Q, T1H, T1z, T1c;
V TZ;
{
V T2x, T1T, T2K, T1W, T1p, Tb, T1A, T16, Tu, TF, T2N, T2H, T2b, T2t, TY;
V T1w, TT, T1v, T20, T2C, Tj, Te, T2h, To, T2f, T23, T2D, TB, TG, Th;
V T2i, Tk;
{
V TL, TW, TP, TQ, T2F, T27, T28, TO;
{
V T1, T2, T12, T13, T4, T5, T7, T8;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
T12 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T13 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0]));
T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0]));
T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
{
V TM, T25, T26, TN;
{
V TJ, T3, T14, T1U, T6, T1V, T9, TK, TU, TV, T1R, T1S, Ta, T15;
TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)]));
T1R = VADD(T1, T2);
T3 = VSUB(T1, T2);
T1S = VADD(T12, T13);
T14 = VSUB(T12, T13);
T1U = VADD(T4, T5);
T6 = VSUB(T4, T5);
T1V = VADD(T7, T8);
T9 = VSUB(T7, T8);
TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
TU = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)]));
T2x = VSUB(T1R, T1S);
T1T = VADD(T1R, T1S);
TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T2K = VSUB(T1V, T1U);
T1W = VADD(T1U, T1V);
Ta = VADD(T6, T9);
T15 = VSUB(T9, T6);
T25 = VADD(TJ, TK);
TL = VSUB(TJ, TK);
T26 = VADD(TV, TU);
TW = VSUB(TU, TV);
TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)]));
T1p = VFNMS(LDK(KP707106781), Ta, T3);
Tb = VFMA(LDK(KP707106781), Ta, T3);
T1A = VFMA(LDK(KP707106781), T15, T14);
T16 = VFNMS(LDK(KP707106781), T15, T14);
TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
}
T2F = VSUB(T25, T26);
T27 = VADD(T25, T26);
T28 = VADD(TM, TN);
TO = VSUB(TM, TN);
}
}
{
V Ty, T21, Tx, Tz, T1Y, T1Z;
{
V Ts, Tt, TD, T29, TR, TE, Tv, Tw;
Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
TD = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T29 = VADD(TP, TQ);
TR = VSUB(TP, TQ);
TE = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)]));
Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)]));
Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)]));
T1Y = VADD(Ts, Tt);
Tu = VSUB(Ts, Tt);
{
V T2G, T2a, TX, TS;
T2G = VSUB(T29, T28);
T2a = VADD(T28, T29);
TX = VSUB(TR, TO);
TS = VADD(TO, TR);
T1Z = VADD(TD, TE);
TF = VSUB(TD, TE);
T21 = VADD(Tv, Tw);
Tx = VSUB(Tv, Tw);
//.........这里部分代码省略.........
示例8: n2fv_8
static void n2fv_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 = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T1, T2, 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 = VSUB(Tj, Tk);
Tl = VADD(Tj, Tk);
{
V Tq, To, Ta, Tf;
Tq = VSUB(Tn, Tm);
To = VADD(Tm, Tn);
Ta = VADD(T6, T9);
Tf = VSUB(T9, T6);
{
V Tr, Ts, Tt, Tu, Tg, Ti, Tb, Th;
Tr = VADD(Tl, To);
STM2(&(xo[0]), Tr, ovs, &(xo[0]));
Ts = VSUB(Tl, To);
STM2(&(xo[8]), Ts, ovs, &(xo[0]));
Tt = VFMAI(Tq, Tp);
STM2(&(xo[4]), Tt, ovs, &(xo[0]));
Tu = VFNMSI(Tq, Tp);
STM2(&(xo[12]), Tu, ovs, &(xo[0]));
Tg = VFNMS(LDK(KP707106781), Tf, Te);
Ti = VFMA(LDK(KP707106781), Tf, Te);
Tb = VFMA(LDK(KP707106781), Ta, T3);
Th = VFNMS(LDK(KP707106781), Ta, T3);
{
V Tv, Tw, Tx, Ty;
Tv = VFMAI(Ti, Th);
STM2(&(xo[6]), Tv, ovs, &(xo[2]));
STN2(&(xo[4]), Tt, Tv, ovs);
Tw = VFNMSI(Ti, Th);
STM2(&(xo[10]), Tw, ovs, &(xo[2]));
STN2(&(xo[8]), Ts, Tw, ovs);
Tx = VFMAI(Tg, Tb);
STM2(&(xo[14]), Tx, ovs, &(xo[2]));
STN2(&(xo[12]), Tu, Tx, ovs);
Ty = VFNMSI(Tg, Tb);
STM2(&(xo[2]), Ty, ovs, &(xo[2]));
STN2(&(xo[0]), Tr, Ty, ovs);
}
}
}
}
}
}
VLEAVE();
}示例9: 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;
BEGIN_SIMD();
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
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]));
}
}
END_SIMD();
}示例10: t2fv_16
static void t2fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
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 TJ, T10, TD, T11, T1b, T1c, Ty, TK, T16, T17, T18, Tb, TN, T13, T14;
V T15, Tm, TM, TG, TI, TH;
TG = LD(&(x[0]), ms, &(x[0]));
TH = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
TI = BYTWJ(&(W[TWVL * 14]), TH);
TJ = VSUB(TG, TI);
T10 = VADD(TG, TI);
{
V TA, TC, Tz, TB;
Tz = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
TA = BYTWJ(&(W[TWVL * 6]), Tz);
TB = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
TC = BYTWJ(&(W[TWVL * 22]), TB);
TD = VSUB(TA, TC);
T11 = VADD(TA, TC);
}
{
V Tp, Tw, Tr, Tu, Ts, Tx;
{
V To, Tv, Tq, Tt;
To = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
Tp = BYTWJ(&(W[TWVL * 26]), To);
Tv = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
Tw = BYTWJ(&(W[TWVL * 18]), Tv);
Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tr = BYTWJ(&(W[TWVL * 10]), Tq);
Tt = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tu = BYTWJ(&(W[TWVL * 2]), Tt);
}
T1b = VADD(Tp, Tr);
T1c = VADD(Tu, Tw);
Ts = VSUB(Tp, Tr);
Tx = VSUB(Tu, Tw);
Ty = VMUL(LDK(KP707106781), VSUB(Ts, Tx));
TK = VMUL(LDK(KP707106781), VADD(Tx, Ts));
}
{
V T2, T9, T4, T7, T5, Ta;
{
V T1, T8, T3, T6;
T1 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
T2 = BYTWJ(&(W[TWVL * 28]), T1);
T8 = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
T9 = BYTWJ(&(W[TWVL * 20]), T8);
T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T4 = BYTWJ(&(W[TWVL * 12]), T3);
T6 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T7 = BYTWJ(&(W[TWVL * 4]), T6);
}
T16 = VADD(T2, T4);
T17 = VADD(T7, T9);
T18 = VSUB(T16, T17);
T5 = VSUB(T2, T4);
Ta = VSUB(T7, T9);
Tb = VFNMS(LDK(KP923879532), Ta, VMUL(LDK(KP382683432), T5));
TN = VFMA(LDK(KP923879532), T5, VMUL(LDK(KP382683432), Ta));
}
{
V Td, Tk, Tf, Ti, Tg, Tl;
{
V Tc, Tj, Te, Th;
Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Td = BYTWJ(&(W[0]), Tc);
Tj = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
Tk = BYTWJ(&(W[TWVL * 24]), Tj);
Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
Tf = BYTWJ(&(W[TWVL * 16]), Te);
Th = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Ti = BYTWJ(&(W[TWVL * 8]), Th);
}
T13 = VADD(Td, Tf);
T14 = VADD(Ti, Tk);
T15 = VSUB(T13, T14);
Tg = VSUB(Td, Tf);
Tl = VSUB(Ti, Tk);
Tm = VFMA(LDK(KP382683432), Tg, VMUL(LDK(KP923879532), Tl));
TM = VFNMS(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tg));
}
{
V T1a, T1g, T1f, T1h;
{
V T12, T19, T1d, T1e;
T12 = VSUB(T10, T11);
T19 = VMUL(LDK(KP707106781), VADD(T15, T18));
T1a = VADD(T12, T19);
T1g = VSUB(T12, T19);
T1d = VSUB(T1b, T1c);
T1e = VMUL(LDK(KP707106781), VSUB(T18, T15));
T1f = VBYI(VADD(T1d, T1e));
T1h = VBYI(VSUB(T1e, T1d));
//.........这里部分代码省略.........
示例11: hc2cbdftv_8
static void hc2cbdftv_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
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(rs)) {
V TJ, T4, Tf, TB, TD, TE, Tm, T1, Tj, TF, Tp, Tb, Tg, Tt, Tx;
V T2, T3, Td, Te, T5, T6, T8, T9, Tn, T7, To, Ta, Tk, Tl, TG;
V TL, Tq, Tc, Tu, Th, Tv, Ty, Tw, TC, Ti, TK, TA, Tz, TI, TH;
V Ts, Tr, TN, TM;
T2 = LD(&(Rp[0]), ms, &(Rp[0]));
T3 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)]));
Td = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0]));
Te = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
T5 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0]));
T8 = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)]));
T9 = LD(&(Rm[0]), -ms, &(Rm[0]));
TJ = LDW(&(W[0]));
Tk = VFMACONJ(T3, T2);
T4 = VFNMSCONJ(T3, T2);
Tl = VFMACONJ(Te, Td);
Tf = VFNMSCONJ(Te, Td);
Tn = VFMACONJ(T6, T5);
T7 = VFNMSCONJ(T6, T5);
To = VFMACONJ(T9, T8);
Ta = VFMSCONJ(T9, T8);
TB = LDW(&(W[TWVL * 8]));
TD = LDW(&(W[TWVL * 6]));
TE = VADD(Tk, Tl);
Tm = VSUB(Tk, Tl);
T1 = LDW(&(W[TWVL * 12]));
Tj = LDW(&(W[TWVL * 10]));
TF = VADD(Tn, To);
Tp = VSUB(Tn, To);
Tb = VADD(T7, Ta);
Tg = VSUB(T7, Ta);
Tt = LDW(&(W[TWVL * 4]));
Tx = LDW(&(W[TWVL * 2]));
TG = VZMUL(TD, VSUB(TE, TF));
TL = VADD(TE, TF);
Tq = VZMUL(Tj, VFNMSI(Tp, Tm));
Tc = VFMA(LDK(KP707106781), Tb, T4);
Tu = VFNMS(LDK(KP707106781), Tb, T4);
Th = VFMA(LDK(KP707106781), Tg, Tf);
Tv = VFNMS(LDK(KP707106781), Tg, Tf);
Ty = VZMUL(Tx, VFMAI(Tp, Tm));
Tw = VZMULI(Tt, VFNMSI(Tv, Tu));
TC = VZMULI(TB, VFMAI(Tv, Tu));
Ti = VZMULI(T1, VFNMSI(Th, Tc));
TK = VZMULI(TJ, VFMAI(Th, Tc));
TA = VCONJ(VSUB(Ty, Tw));
Tz = VADD(Tw, Ty);
TI = VCONJ(VSUB(TG, TC));
TH = VADD(TC, TG);
Ts = VCONJ(VSUB(Tq, Ti));
Tr = VADD(Ti, Tq);
TN = VCONJ(VSUB(TL, TK));
TM = VADD(TK, TL);
ST(&(Rm[WS(rs, 1)]), TA, -ms, &(Rm[WS(rs, 1)]));
ST(&(Rp[WS(rs, 1)]), Tz, ms, &(Rp[WS(rs, 1)]));
ST(&(Rm[WS(rs, 2)]), TI, -ms, &(Rm[0]));
ST(&(Rp[WS(rs, 2)]), TH, ms, &(Rp[0]));
ST(&(Rm[WS(rs, 3)]), Ts, -ms, &(Rm[WS(rs, 1)]));
ST(&(Rp[WS(rs, 3)]), Tr, ms, &(Rp[WS(rs, 1)]));
ST(&(Rm[0]), TN, -ms, &(Rm[0]));
ST(&(Rp[0]), TM, ms, &(Rp[0]));
}
}示例12: n2bv_20
//.........这里部分代码省略.........
{
V T1t, T19, T1w, T1g;
T1t = VADD(T17, T18);
T19 = VSUB(T17, T18);
Ts = VSUB(To, Tr);
TF = VADD(To, Tr);
T1w = VADD(T1e, T1f);
T1g = VSUB(T1e, T1f);
Th = VADD(Td, Tg);
TM = VSUB(Td, Tg);
T1u = VADD(T1s, T1t);
T1C = VSUB(T1s, T1t);
T1n = VSUB(T16, T19);
T1a = VADD(T16, T19);
T1m = VSUB(T1d, T1g);
T1h = VADD(T1d, T1g);
T1x = VADD(T1v, T1w);
T1D = VSUB(T1v, T1w);
}
}
}
}
Tk = VSUB(Ta, Th);
Ti = VADD(Ta, Th);
{
V TJ, T1k, T1A, TZ, Tj, T1E, T1G, TI, T10, T1j, T1z, T1i, T1y, TH;
TJ = VSUB(TF, TG);
TH = VADD(TF, TG);
T1i = VADD(T1a, T1h);
T1k = VSUB(T1a, T1h);
T1y = VADD(T1u, T1x);
T1A = VSUB(T1u, T1x);
TZ = VADD(T3, Ti);
Tj = VFNMS(LDK(KP250000000), Ti, T3);
T1E = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1D, T1C));
T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1C, T1D));
TI = VFNMS(LDK(KP250000000), TH, TE);
T10 = VADD(TE, TH);
T1j = VFNMS(LDK(KP250000000), T1i, T13);
T1H = VADD(T1r, T1y);
STM2(&(xo[0]), T1H, ovs, &(xo[0]));
T1z = VFNMS(LDK(KP250000000), T1y, T1r);
T1I = VADD(T13, T1i);
STM2(&(xo[20]), T1I, ovs, &(xo[0]));
{
V T1J, T1K, T1p, T1l, T1o, T1q, T1F, T1B, T1L, T1Q;
TS = VFNMS(LDK(KP618033988), Ts, Tz);
TA = VFMA(LDK(KP618033988), Tz, Ts);
TN = VFMA(LDK(KP618033988), TM, TL);
TV = VFNMS(LDK(KP618033988), TL, TM);
T1J = VFMAI(T10, TZ);
STM2(&(xo[10]), T1J, ovs, &(xo[2]));
T1K = VFNMSI(T10, TZ);
STM2(&(xo[30]), T1K, ovs, &(xo[2]));
T1p = VFMA(LDK(KP559016994), T1k, T1j);
T1l = VFNMS(LDK(KP559016994), T1k, T1j);
T1o = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1n, T1m));
T1q = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1m, T1n));
T1F = VFNMS(LDK(KP559016994), T1A, T1z);
T1B = VFMA(LDK(KP559016994), T1A, T1z);
T1L = VFNMSI(T1q, T1p);
STM2(&(xo[28]), T1L, ovs, &(xo[0]));
STN2(&(xo[28]), T1L, T1K, ovs);
T1M = VFMAI(T1q, T1p);
STM2(&(xo[12]), T1M, ovs, &(xo[0]));
T1N = VFMAI(T1o, T1l);示例13: n2sv_8
static void n2sv_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;
for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) {
V TF, TJ, TD, TR, TS, TT, TU, TV, TW, TE, TX, TY, TK, TI, TZ;
V T10, T11, T12;
{
V Tb, Tn, T3, TC, Ti, TB, T6, To, Tl, Tc, Tw, Tx, T8, T9, Tr;
V Ts;
{
V T1, T2, Tg, Th, T4, T5, Tj, Tk;
T1 = LD(&(ri[0]), ivs, &(ri[0]));
T2 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0]));
Tg = LD(&(ii[0]), ivs, &(ii[0]));
Th = LD(&(ii[WS(is, 4)]), ivs, &(ii[0]));
T4 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0]));
T5 = LD(&(ri[WS(is, 6)]), ivs, &(ri[0]));
Tj = LD(&(ii[WS(is, 2)]), ivs, &(ii[0]));
Tk = LD(&(ii[WS(is, 6)]), ivs, &(ii[0]));
Tb = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)]));
Tn = VSUB(T1, T2);
T3 = VADD(T1, T2);
TC = VSUB(Tg, Th);
Ti = VADD(Tg, Th);
TB = VSUB(T4, T5);
T6 = VADD(T4, T5);
To = VSUB(Tj, Tk);
Tl = VADD(Tj, Tk);
Tc = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)]));
Tw = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)]));
Tx = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)]));
T8 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)]));
T9 = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)]));
Tr = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)]));
Ts = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)]));
}
{
V TL, T7, TP, Tm, Tz, TH, Te, Tf, TO, TQ, TG, Tu, Tp, TA;
{
V Td, Tv, TN, Ty, Ta, Tq, TM, Tt;
TL = VSUB(T3, T6);
T7 = VADD(T3, T6);
Td = VADD(Tb, Tc);
Tv = VSUB(Tb, Tc);
TN = VADD(Tw, Tx);
Ty = VSUB(Tw, Tx);
Ta = VADD(T8, T9);
Tq = VSUB(T8, T9);
TM = VADD(Tr, Ts);
Tt = VSUB(Tr, Ts);
TP = VADD(Ti, Tl);
Tm = VSUB(Ti, Tl);
Tz = VSUB(Tv, Ty);
TH = VADD(Tv, Ty);
Te = VADD(Ta, Td);
Tf = VSUB(Td, Ta);
TO = VSUB(TM, TN);
TQ = VADD(TM, TN);
TG = VSUB(Tt, Tq);
Tu = VADD(Tq, Tt);
}
TF = VSUB(Tn, To);
Tp = VADD(Tn, To);
TJ = VSUB(TC, TB);
TD = VADD(TB, TC);
TR = VSUB(Tm, Tf);
STM4(&(io[6]), TR, ovs, &(io[0]));
TS = VADD(Tf, Tm);
STM4(&(io[2]), TS, ovs, &(io[0]));
TT = VADD(T7, Te);
STM4(&(ro[0]), TT, ovs, &(ro[0]));
TU = VSUB(T7, Te);
STM4(&(ro[4]), TU, ovs, &(ro[0]));
TV = VADD(TP, TQ);
STM4(&(io[0]), TV, ovs, &(io[0]));
TW = VSUB(TP, TQ);
STM4(&(io[4]), TW, ovs, &(io[0]));
TE = VSUB(Tz, Tu);
TA = VADD(Tu, Tz);
TX = VADD(TL, TO);
STM4(&(ro[2]), TX, ovs, &(ro[0]));
TY = VSUB(TL, TO);
STM4(&(ro[6]), TY, ovs, &(ro[0]));
TK = VADD(TG, TH);
TI = VSUB(TG, TH);
TZ = VFMA(LDK(KP707106781), TA, Tp);
STM4(&(ro[1]), TZ, ovs, &(ro[1]));
T10 = VFNMS(LDK(KP707106781), TA, Tp);
STM4(&(ro[5]), T10, ovs, &(ro[1]));
}
}
T11 = VFMA(LDK(KP707106781), TK, TJ);
STM4(&(io[1]), T11, ovs, &(io[1]));
T12 = VFNMS(LDK(KP707106781), TK, TJ);
STM4(&(io[5]), T12, ovs, &(io[1]));
{
V T13, T14, T15, T16;
T13 = VFMA(LDK(KP707106781), TE, TD);
//.........这里部分代码省略.........
示例14: t1fuv_7
static void t1fuv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
DVK(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)]));
}
}
}
VLEAVE();
}示例15: n1bv_7
static void n1bv_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
DVK(KP692021471, +0.692021471630095869627814897002069140197260599);
DVK(KP801937735, +0.801937735804838252472204639014890102331838324);
DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
{
INT i;
const R *xi;
R *xo;
xi = ii;
xo = io;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T1, T2, T3, T8, T9, T5, T6;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
{
V Tg, T4, Te, Ta, Tf, T7;
Tg = VSUB(T2, T3);
T4 = VADD(T2, T3);
Te = VSUB(T8, T9);
Ta = VADD(T8, T9);
Tf = VSUB(T5, T6);
T7 = VADD(T5, T6);
{
V Tr, Tj, Tm, Th, To, Tb;
Tr = VFMA(LDK(KP554958132), Te, Tg);
Tj = VFNMS(LDK(KP356895867), T4, Ta);
Tm = VFMA(LDK(KP554958132), Tf, Te);
Th = VFNMS(LDK(KP554958132), Tg, Tf);
ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, Ta))), ovs, &(xo[0]));
To = VFNMS(LDK(KP356895867), T7, T4);
Tb = VFNMS(LDK(KP356895867), Ta, T7);
{
V Ts, Tk, Tn, Ti;
Ts = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tr, Tf));
Tk = VFNMS(LDK(KP692021471), Tj, T7);
Tn = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tm, Tg));
Ti = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Th, Te));
{
V Tp, Tc, Tl, Tq, Td;
Tp = VFNMS(LDK(KP692021471), To, Ta);
Tc = VFNMS(LDK(KP692021471), Tb, T4);
Tl = VFNMS(LDK(KP900968867), Tk, T1);
Tq = VFNMS(LDK(KP900968867), Tp, T1);
Td = VFNMS(LDK(KP900968867), Tc, T1);
ST(&(xo[WS(os, 5)]), VFNMSI(Tn, Tl), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VFMAI(Tn, Tl), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VFNMSI(Ts, Tq), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFMAI(Ts, Tq), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VFNMSI(Ti, Td), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(Ti, Td), ovs, &(xo[WS(os, 1)]));
}
}
}
}
}
}
VLEAVE();
}