static const R *t2fv_8(R *ri, R *ii, const R *W, stride ios, INT m, INT dist)
{
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 * 14), MAKE_VOLATILE_STRIDE(ios)) {
V T1, T2, Th, Tj, T5, T7, Ta, Tc;
T1 = LD(&(x[0]), dist, &(x[0]));
T2 = LD(&(x[WS(ios, 4)]), dist, &(x[0]));
Th = LD(&(x[WS(ios, 2)]), dist, &(x[0]));
Tj = LD(&(x[WS(ios, 6)]), dist, &(x[0]));
T5 = LD(&(x[WS(ios, 1)]), dist, &(x[WS(ios, 1)]));
T7 = LD(&(x[WS(ios, 5)]), dist, &(x[WS(ios, 1)]));
Ta = LD(&(x[WS(ios, 7)]), dist, &(x[WS(ios, 1)]));
Tc = LD(&(x[WS(ios, 3)]), dist, &(x[WS(ios, 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(ios, 2)]), VFMAI(Tx, Tw), dist, &(x[0]));
ST(&(x[WS(ios, 6)]), VFNMSI(Tx, Tw), dist, &(x[0]));
ST(&(x[0]), VADD(Ts, Tv), dist, &(x[0]));
ST(&(x[WS(ios, 4)]), VSUB(Ts, Tv), dist, &(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(ios, 5)]), VFNMSI(Tp, To), dist, &(x[WS(ios, 1)]));
ST(&(x[WS(ios, 3)]), VFMAI(Tp, To), dist, &(x[WS(ios, 1)]));
ST(&(x[WS(ios, 7)]), VFMAI(Tn, Tg), dist, &(x[WS(ios, 1)]));
ST(&(x[WS(ios, 1)]), VFNMSI(Tn, Tg), dist, &(x[WS(ios, 1)]));
}
}
}
}
}
return W;
}
static void n1fv_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(KP801937735, +0.801937735804838252472204639014890102331838324);
DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
DVK(KP692021471, +0.692021471630095869627814897002069140197260599);
DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
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 TH, T3, TP, Tn, Ta, Ts, TW, TK, TO, Tk, TM, Tg, TL, Td, T1;
V T2;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
{
V Ti, TI, T6, TJ, T9, Tj, Te, Tf, Tb, Tc;
{
V T4, T5, T7, T8, Tl, Tm;
T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
TH = VADD(T1, T2);
T3 = VSUB(T1, T2);
TI = VADD(T4, T5);
T6 = VSUB(T4, T5);
TJ = VADD(T7, T8);
T9 = VSUB(T7, T8);
TP = VADD(Tl, Tm);
Tn = VSUB(Tl, Tm);
Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
}
Ta = VADD(T6, T9);
Ts = VSUB(T9, T6);
TW = VSUB(TJ, TI);
TK = VADD(TI, TJ);
TO = VADD(Ti, Tj);
Tk = VSUB(Ti, Tj);
TM = VADD(Te, Tf);
Tg = VSUB(Te, Tf);
TL = VADD(Tb, Tc);
Td = VSUB(Tb, Tc);
}
{
V T18, TB, T13, TY, TG, Tw, T11, Tr, T16, TT, Tz, TE, TU, TQ;
TU = VSUB(TO, TP);
TQ = VADD(TO, TP);
{
V Tt, To, TV, TN;
Tt = VSUB(Tn, Tk);
To = VADD(Tk, Tn);
TV = VSUB(TL, TM);
TN = VADD(TL, TM);
{
V Tu, Th, TZ, T17;
Tu = VSUB(Tg, Td);
Th = VADD(Td, Tg);
TZ = VFNMS(LDK(KP356895867), TK, TQ);
T17 = VFNMS(LDK(KP554958132), TU, TW);
{
V Tp, TA, T14, TR;
Tp = VFNMS(LDK(KP356895867), Ta, To);
TA = VFMA(LDK(KP554958132), Tt, Ts);
ST(&(xo[0]), VADD(TH, VADD(TK, VADD(TN, TQ))), ovs, &(xo[0]));
T14 = VFNMS(LDK(KP356895867), TN, TK);
TR = VFNMS(LDK(KP356895867), TQ, TN);
{
V T12, TX, Tx, TC;
T12 = VFMA(LDK(KP554958132), TV, TU);
TX = VFMA(LDK(KP554958132), TW, TV);
ST(&(xo[WS(os, 7)]), VADD(T3, VADD(Ta, VADD(Th, To))), ovs, &(xo[WS(os, 1)]));
Tx = VFNMS(LDK(KP356895867), Th, Ta);
TC = VFNMS(LDK(KP356895867), To, Th);
{
V TF, Tv, T10, Tq;
TF = VFNMS(LDK(KP554958132), Ts, Tu);
Tv = VFMA(LDK(KP554958132), Tu, Tt);
T10 = VFNMS(LDK(KP692021471), TZ, TN);
T18 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T17, TV));
Tq = VFNMS(LDK(KP692021471), Tp, Th);
TB = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TA, Tu));
{
V T15, TS, Ty, TD;
T15 = VFNMS(LDK(KP692021471), T14, TQ);
TS = VFNMS(LDK(KP692021471), TR, TK);
T13 = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), T12, TW));
TY = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TX, TU));
Ty = VFNMS(LDK(KP692021471), Tx, To);
TD = VFNMS(LDK(KP692021471), TC, Ta);
TG = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TF, Tt));
Tw = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tv, Ts));
T11 = VFNMS(LDK(KP900968867), T10, TH);
Tr = VFNMS(LDK(KP900968867), Tq, T3);
T16 = VFNMS(LDK(KP900968867), T15, TH);
TT = VFNMS(LDK(KP900968867), TS, TH);
Tz = VFNMS(LDK(KP900968867), Ty, T3);
TE = VFNMS(LDK(KP900968867), TD, T3);
}
}
}
}
}
}
ST(&(xo[WS(os, 12)]), VFNMSI(T13, T11), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VFMAI(T13, T11), ovs, &(xo[0]));
ST(&(xo[WS(os, 9)]), VFMAI(Tw, Tr), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 5)]), VFNMSI(Tw, Tr), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 8)]), VFNMSI(T18, T16), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VFMAI(T18, T16), ovs, &(xo[0]));
ST(&(xo[WS(os, 10)]), VFNMSI(TY, TT), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VFMAI(TY, TT), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFMAI(TB, Tz), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 13)]), VFNMSI(TB, Tz), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VFMAI(TG, TE), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 11)]), VFNMSI(TG, TE), ovs, &(xo[WS(os, 1)]));
}
}
}
static void t1fv_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();
}
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(16, is), MAKE_VOLATILE_STRIDE(16, 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();
}
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(18, is), MAKE_VOLATILE_STRIDE(18, 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]));
}
}
}
}
}
}
}
VLEAVE();
}
static void t1fuv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
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 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(9, rs)) {
V T1, T3, T5, T9, Th, Tb, Td, Tj, Tl, TD, T6;
T1 = LD(&(x[0]), ms, &(x[0]));
{
V T2, T4, T8, Tg;
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, 1)]), ms, &(x[WS(rs, 1)]));
Tg = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
{
V Ta, Tc, Ti, Tk;
Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Ti = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tk = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
T3 = BYTWJ(&(W[TWVL * 4]), T2);
T5 = BYTWJ(&(W[TWVL * 10]), T4);
T9 = BYTWJ(&(W[0]), T8);
Th = BYTWJ(&(W[TWVL * 2]), Tg);
Tb = BYTWJ(&(W[TWVL * 6]), Ta);
Td = BYTWJ(&(W[TWVL * 12]), Tc);
Tj = BYTWJ(&(W[TWVL * 8]), Ti);
Tl = BYTWJ(&(W[TWVL * 14]), Tk);
}
}
TD = VSUB(T5, T3);
T6 = VADD(T3, T5);
{
V Tt, Te, Tu, Tm, Tr, T7;
Tt = VSUB(Tb, Td);
Te = VADD(Tb, Td);
Tu = VSUB(Tl, Tj);
Tm = VADD(Tj, Tl);
Tr = VFNMS(LDK(KP500000000), T6, T1);
T7 = VADD(T1, T6);
{
V Tv, Tf, Ts, Tn;
Tv = VFNMS(LDK(KP500000000), Te, T9);
Tf = VADD(T9, Te);
Ts = VFNMS(LDK(KP500000000), Tm, Th);
Tn = VADD(Th, Tm);
{
V TG, TK, Tw, TJ, TF, TA, To, Tq;
TG = VFNMS(LDK(KP726681596), Tt, Tv);
TK = VFMA(LDK(KP968908795), Tv, Tt);
Tw = VFNMS(LDK(KP586256827), Tv, Tu);
TJ = VFNMS(LDK(KP152703644), Tu, Ts);
TF = VFMA(LDK(KP203604859), Ts, Tu);
TA = VFNMS(LDK(KP439692620), Tt, Ts);
To = VADD(Tf, Tn);
Tq = VMUL(LDK(KP866025403), VSUB(Tn, Tf));
{
V TQ, TH, TL, TN, TB, Tp, Ty, TI, Tx;
Tx = VFNMS(LDK(KP347296355), Tw, Tt);
TQ = VFNMS(LDK(KP898197570), TG, TF);
TH = VFMA(LDK(KP898197570), TG, TF);
TL = VFMA(LDK(KP673648177), TK, TJ);
TN = VFNMS(LDK(KP673648177), TK, TJ);
TB = VFNMS(LDK(KP420276625), TA, Tu);
ST(&(x[0]), VADD(T7, To), ms, &(x[0]));
Tp = VFNMS(LDK(KP500000000), To, T7);
Ty = VFNMS(LDK(KP907603734), Tx, Ts);
TI = VFMA(LDK(KP852868531), TH, Tr);
{
V TO, TR, TM, TC, Tz, TP, TS, TE;
TO = VFNMS(LDK(KP500000000), TH, TN);
TR = VFMA(LDK(KP666666666), TL, TQ);
TM = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), TD, TL));
TC = VFNMS(LDK(KP826351822), TB, Tv);
ST(&(x[WS(rs, 6)]), VFNMSI(Tq, Tp), ms, &(x[0]));
ST(&(x[WS(rs, 3)]), VFMAI(Tq, Tp), ms, &(x[WS(rs, 1)]));
Tz = VFNMS(LDK(KP939692620), Ty, Tr);
TP = VFMA(LDK(KP852868531), TO, Tr);
TS = VMUL(LDK(KP866025403), VFMA(LDK(KP852868531), TR, TD));
ST(&(x[WS(rs, 8)]), VFMAI(TM, TI), ms, &(x[0]));
ST(&(x[WS(rs, 1)]), VFNMSI(TM, TI), ms, &(x[WS(rs, 1)]));
TE = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), TD, TC));
ST(&(x[WS(rs, 4)]), VFMAI(TS, TP), ms, &(x[0]));
ST(&(x[WS(rs, 5)]), VFNMSI(TS, TP), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(TE, Tz), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 2)]), VFNMSI(TE, Tz), ms, &(x[0]));
}
}
}
}
}
}
}
VLEAVE();
}
static void hc2cfdftv_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
{
INT m;
for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 18)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(40, rs)) {
V T5, T6, Tw, Tr, Tc, Tj, Tl, Tm, Tk, Ts, Tg, Ty, T3, T4, T1;
V T2, Tv, Tq, Ta, Tb, T9, Ti, Te, Tf, Td, Tx, Tn, Tt, Th, TQ;
V TT, Tz, T7, TR, To, Tu, TU;
T1 = LD(&(Rp[0]), ms, &(Rp[0]));
T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
Tv = LDW(&(W[0]));
T5 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0]));
T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0]));
Tq = LDW(&(W[TWVL * 6]));
Ta = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
Tb = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
T9 = LDW(&(W[TWVL * 2]));
Ti = LDW(&(W[TWVL * 4]));
Tw = VZMULIJ(Tv, VFNMSCONJ(T2, T1));
Te = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)]));
Tf = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)]));
Tr = VZMULJ(Tq, VFMACONJ(T6, T5));
Td = LDW(&(W[TWVL * 12]));
Tx = LDW(&(W[TWVL * 10]));
Tc = VZMULJ(T9, VFMACONJ(Tb, Ta));
Tj = VZMULIJ(Ti, VFNMSCONJ(Tb, Ta));
Tl = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0]));
Tm = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0]));
Tk = LDW(&(W[TWVL * 14]));
Ts = LDW(&(W[TWVL * 16]));
Tg = VZMULIJ(Td, VFNMSCONJ(Tf, Te));
Ty = VZMULJ(Tx, VFMACONJ(Tf, Te));
T3 = VFMACONJ(T2, T1);
T4 = LDW(&(W[TWVL * 8]));
Tn = VZMULJ(Tk, VFMACONJ(Tm, Tl));
Tt = VZMULIJ(Ts, VFNMSCONJ(Tm, Tl));
Th = VSUB(Tc, Tg);
TQ = VADD(Tc, Tg);
TT = VADD(Tw, Ty);
Tz = VSUB(Tw, Ty);
T7 = VZMULIJ(T4, VFNMSCONJ(T6, T5));
TR = VADD(Tj, Tn);
To = VSUB(Tj, Tn);
Tu = VSUB(Tr, Tt);
TU = VADD(Tr, Tt);
{
V TP, T8, TS, T11, Tp, TH, TA, TG, TV, T12, TE, TB, TM, TI, TZ;
V TW, T17, T13, TD, TC, TY, TX, TL, TF, T10, T16, TN, TO, TK, TJ;
V T18, T19, T15, T14;
TP = VADD(T3, T7);
T8 = VSUB(T3, T7);
TS = VADD(TQ, TR);
T11 = VSUB(TQ, TR);
Tp = VSUB(Th, To);
TH = VADD(Th, To);
TA = VSUB(Tu, Tz);
TG = VADD(Tz, Tu);
TV = VADD(TT, TU);
T12 = VSUB(TU, TT);
TE = VSUB(Tp, TA);
TB = VADD(Tp, TA);
TM = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TG, TH));
TI = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TH, TG));
TZ = VSUB(TS, TV);
TW = VADD(TS, TV);
T17 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T11, T12));
T13 = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T12, T11));
TD = VFNMS(LDK(KP250000000), TB, T8);
TC = VMUL(LDK(KP500000000), VADD(T8, TB));
TY = VFNMS(LDK(KP250000000), TW, TP);
TX = VCONJ(VMUL(LDK(KP500000000), VADD(TP, TW)));
TL = VFMA(LDK(KP559016994), TE, TD);
TF = VFNMS(LDK(KP559016994), TE, TD);
ST(&(Rp[0]), TC, ms, &(Rp[0]));
T10 = VFMA(LDK(KP559016994), TZ, TY);
T16 = VFNMS(LDK(KP559016994), TZ, TY);
ST(&(Rm[WS(rs, 4)]), TX, -ms, &(Rm[0]));
TN = VCONJ(VMUL(LDK(KP500000000), VFNMSI(TM, TL)));
TO = VMUL(LDK(KP500000000), VFMAI(TM, TL));
TK = VMUL(LDK(KP500000000), VFMAI(TI, TF));
TJ = VCONJ(VMUL(LDK(KP500000000), VFNMSI(TI, TF)));
T18 = VMUL(LDK(KP500000000), VFNMSI(T17, T16));
T19 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T17, T16)));
T15 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T13, T10)));
T14 = VMUL(LDK(KP500000000), VFNMSI(T13, T10));
ST(&(Rm[WS(rs, 3)]), TN, -ms, &(Rm[WS(rs, 1)]));
ST(&(Rp[WS(rs, 4)]), TO, ms, &(Rp[0]));
ST(&(Rp[WS(rs, 2)]), TK, ms, &(Rp[0]));
ST(&(Rm[WS(rs, 1)]), TJ, -ms, &(Rm[WS(rs, 1)]));
ST(&(Rp[WS(rs, 3)]), T18, ms, &(Rp[WS(rs, 1)]));
ST(&(Rm[WS(rs, 2)]), T19, -ms, &(Rm[0]));
ST(&(Rm[0]), T15, -ms, &(Rm[0]));
ST(&(Rp[WS(rs, 1)]), T14, ms, &(Rp[WS(rs, 1)]));
}
}
}
VLEAVE();
}
static void n2fv_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 = 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, T6, Tk, Tn, Tc, Td, Tf, Tr, T4, Ts, T9, Tg, Te, Tl;
{
V T2, T3, T7, T8;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Tk = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tr = VSUB(T3, T2);
T4 = VADD(T2, T3);
Ts = VSUB(T8, T7);
T9 = VADD(T7, T8);
Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
}
Te = VSUB(Tc, Td);
Tl = VADD(Td, Tc);
{
V T5, TF, TB, Tt, Ta, TG, Th, To, Tm, TI;
T5 = VFNMS(LDK(KP500000000), T4, T1);
TF = VADD(T1, T4);
TB = VADD(Tr, Ts);
Tt = VSUB(Tr, Ts);
Ta = VFNMS(LDK(KP500000000), T9, T6);
TG = VADD(T6, T9);
Th = VSUB(Tf, Tg);
To = VADD(Tf, Tg);
Tm = VFNMS(LDK(KP500000000), Tl, Tk);
TI = VADD(Tk, Tl);
{
V TH, TL, Tb, Tx, TJ, Tp, Ti, TA;
TH = VSUB(TF, TG);
TL = VADD(TF, TG);
Tb = VSUB(T5, Ta);
Tx = VADD(T5, Ta);
TJ = VADD(Tn, To);
Tp = VFNMS(LDK(KP500000000), To, Tn);
Ti = VADD(Te, Th);
TA = VSUB(Te, Th);
{
V Tq, Ty, TK, TM;
Tq = VSUB(Tm, Tp);
Ty = VADD(Tm, Tp);
TK = VSUB(TI, TJ);
TM = VADD(TI, TJ);
{
V TC, TE, Tj, Tv;
TC = VMUL(LDK(KP866025403), VSUB(TA, TB));
TE = VMUL(LDK(KP866025403), VADD(TB, TA));
Tj = VFMA(LDK(KP866025403), Ti, Tb);
Tv = VFNMS(LDK(KP866025403), Ti, Tb);
{
V Tz, TD, Tu, Tw;
Tz = VSUB(Tx, Ty);
TD = VADD(Tx, Ty);
Tu = VFNMS(LDK(KP866025403), Tt, Tq);
Tw = VFMA(LDK(KP866025403), Tt, Tq);
{
V TN, TO, TP, TQ;
TN = VADD(TL, TM);
STM2(&(xo[0]), TN, ovs, &(xo[0]));
TO = VSUB(TL, TM);
STM2(&(xo[12]), TO, ovs, &(xo[0]));
TP = VFMAI(TK, TH);
STM2(&(xo[6]), TP, ovs, &(xo[2]));
TQ = VFNMSI(TK, TH);
STM2(&(xo[18]), TQ, ovs, &(xo[2]));
{
V TR, TS, TT, TU;
TR = VFMAI(TE, TD);
STM2(&(xo[8]), TR, ovs, &(xo[0]));
TS = VFNMSI(TE, TD);
STM2(&(xo[16]), TS, ovs, &(xo[0]));
STN2(&(xo[16]), TS, TQ, ovs);
TT = VFNMSI(TC, Tz);
STM2(&(xo[20]), TT, ovs, &(xo[0]));
TU = VFMAI(TC, Tz);
STM2(&(xo[4]), TU, ovs, &(xo[0]));
STN2(&(xo[4]), TU, TP, ovs);
{
V TV, TW, TX, TY;
TV = VFNMSI(Tw, Tv);
STM2(&(xo[10]), TV, ovs, &(xo[2]));
STN2(&(xo[8]), TR, TV, ovs);
TW = VFMAI(Tw, Tv);
STM2(&(xo[14]), TW, ovs, &(xo[2]));
STN2(&(xo[12]), TO, TW, ovs);
TX = VFMAI(Tu, Tj);
STM2(&(xo[22]), TX, ovs, &(xo[2]));
STN2(&(xo[20]), TT, TX, ovs);
TY = VFNMSI(Tu, Tj);
STM2(&(xo[2]), TY, ovs, &(xo[2]));
STN2(&(xo[0]), TN, TY, ovs);
}
}
}
}
}
}
}
}
}
}
static void t1bv_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 = ii;
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, 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 = BYTW(&(W[0]), T2);
T5 = BYTW(&(W[TWVL * 10]), T4);
Tf = BYTW(&(W[TWVL * 6]), Te);
Td = BYTW(&(W[TWVL * 4]), Tc);
Ta = BYTW(&(W[TWVL * 8]), T9);
T8 = BYTW(&(W[TWVL * 2]), T7);
{
V T6, Tm, Tg, Tk, Tb, Tl;
T6 = VADD(T3, T5);
Tm = VSUB(T3, T5);
Tg = VADD(Td, Tf);
Tk = VSUB(Td, Tf);
Tb = VADD(T8, Ta);
Tl = VSUB(T8, Ta);
{
V Tp, Tx, Tu, Th, Ts, Tn, Tq, Ty;
Tp = VFNMS(LDK(KP356895867), T6, Tg);
Tx = VFMA(LDK(KP554958132), Tk, Tm);
ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0]));
Tu = VFNMS(LDK(KP356895867), Tb, T6);
Th = VFNMS(LDK(KP356895867), Tg, Tb);
Ts = VFMA(LDK(KP554958132), Tl, Tk);
Tn = VFNMS(LDK(KP554958132), Tm, Tl);
Tq = VFNMS(LDK(KP692021471), Tp, Tb);
Ty = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tx, Tl));
{
V Tv, Ti, Tt, To, Tr, Tw, Tj;
Tv = VFNMS(LDK(KP692021471), Tu, Tg);
Ti = VFNMS(LDK(KP692021471), Th, T6);
Tt = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Ts, Tm));
To = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tn, Tk));
Tr = VFNMS(LDK(KP900968867), Tq, T1);
Tw = VFNMS(LDK(KP900968867), Tv, T1);
Tj = VFNMS(LDK(KP900968867), Ti, T1);
ST(&(x[WS(rs, 5)]), VFNMSI(Tt, Tr), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 2)]), VFMAI(Tt, Tr), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(Ty, Tw), ms, &(x[0]));
ST(&(x[WS(rs, 1)]), VFMAI(Ty, Tw), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 4)]), VFNMSI(To, Tj), ms, &(x[0]));
ST(&(x[WS(rs, 3)]), VFMAI(To, Tj), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
static void n2bv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
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 Tb, Tr, T3, Ts, T6, Tw, Tg, Tt, T9, Tc, T1, T2;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
{
V T4, T5, Te, Tf, T7, T8;
T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tr = VADD(T1, T2);
T3 = VSUB(T1, T2);
Ts = VADD(T4, T5);
T6 = VSUB(T4, T5);
Tw = VADD(Te, Tf);
Tg = VSUB(Te, Tf);
Tt = VADD(T7, T8);
T9 = VSUB(T7, T8);
Tc = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
}
{
V TD, Tu, Tm, Ta, Td, Tv;
TD = VSUB(Ts, Tt);
Tu = VADD(Ts, Tt);
Tm = VSUB(T6, T9);
Ta = VADD(T6, T9);
Td = VSUB(Tb, Tc);
Tv = VADD(Tb, Tc);
{
V TC, Tx, Tn, Th;
TC = VSUB(Tv, Tw);
Tx = VADD(Tv, Tw);
Tn = VSUB(Td, Tg);
Th = VADD(Td, Tg);
{
V Ty, TA, TE, TG, Ti, Tk, To, Tq;
Ty = VADD(Tu, Tx);
TA = VSUB(Tu, Tx);
TE = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TD, TC));
TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TC, TD));
Ti = VADD(Ta, Th);
Tk = VSUB(Ta, Th);
To = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tn, Tm));
Tq = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tm, Tn));
{
V Tz, TH, Tj, TI;
Tz = VFNMS(LDK(KP250000000), Ty, Tr);
TH = VADD(Tr, Ty);
STM2(&(xo[0]), TH, ovs, &(xo[0]));
Tj = VFNMS(LDK(KP250000000), Ti, T3);
TI = VADD(T3, Ti);
STM2(&(xo[10]), TI, ovs, &(xo[2]));
{
V TB, TF, Tl, Tp;
TB = VFNMS(LDK(KP559016994), TA, Tz);
TF = VFMA(LDK(KP559016994), TA, Tz);
Tl = VFMA(LDK(KP559016994), Tk, Tj);
Tp = VFNMS(LDK(KP559016994), Tk, Tj);
{
V TJ, TK, TL, TM;
TJ = VFNMSI(TG, TF);
STM2(&(xo[8]), TJ, ovs, &(xo[0]));
STN2(&(xo[8]), TJ, TI, ovs);
TK = VFMAI(TG, TF);
STM2(&(xo[12]), TK, ovs, &(xo[0]));
TL = VFMAI(TE, TB);
STM2(&(xo[16]), TL, ovs, &(xo[0]));
TM = VFNMSI(TE, TB);
STM2(&(xo[4]), TM, ovs, &(xo[0]));
{
V TN, TO, TP, TQ;
TN = VFMAI(Tq, Tp);
STM2(&(xo[6]), TN, ovs, &(xo[2]));
STN2(&(xo[4]), TM, TN, ovs);
TO = VFNMSI(Tq, Tp);
STM2(&(xo[14]), TO, ovs, &(xo[2]));
STN2(&(xo[12]), TK, TO, ovs);
TP = VFNMSI(To, Tl);
STM2(&(xo[18]), TP, ovs, &(xo[2]));
STN2(&(xo[16]), TL, TP, ovs);
TQ = VFMAI(To, Tl);
STM2(&(xo[2]), TQ, ovs, &(xo[2]));
STN2(&(xo[0]), TH, TQ, ovs);
}
}
}
}
}
}
}
}
}
static void n2fv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
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 T1H, T1I, TU, TI, TP, TX, T1M, T1N, T1O, T1P, T1R, T1S, TM, TW, TT;
V TF;
{
V T3, Tm, T1r, T13, Ta, TN, TH, TA, TG, Tt, Th, TO, T1u, T1C, T1n;
V T1a, T1m, T1h, T1x, T1D, TE, Ti;
{
V T1, T2, Tk, Tl;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tl = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
{
V T14, T6, T1c, Tw, Tn, T1f, Tz, T17, T9, To, Tq, T1b, Td, Tr, Te;
V Tf, T15, Tp;
{
V Tx, Ty, T7, T8, Tb, Tc;
{
V T4, T5, Tu, Tv, T11, T12;
T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
Tu = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Tv = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tx = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
T11 = VADD(T1, T2);
Tm = VSUB(Tk, Tl);
T12 = VADD(Tk, Tl);
T14 = VADD(T4, T5);
T6 = VSUB(T4, T5);
T1c = VADD(Tu, Tv);
Tw = VSUB(Tu, Tv);
Ty = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T1r = VADD(T11, T12);
T13 = VSUB(T11, T12);
}
Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T1f = VADD(Tx, Ty);
Tz = VSUB(Tx, Ty);
T17 = VADD(T7, T8);
T9 = VSUB(T7, T8);
To = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
Tq = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T1b = VADD(Tb, Tc);
Td = VSUB(Tb, Tc);
Tr = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
}
Ta = VADD(T6, T9);
TN = VSUB(T6, T9);
T15 = VADD(Tn, To);
Tp = VSUB(Tn, To);
TH = VSUB(Tz, Tw);
TA = VADD(Tw, Tz);
{
V T1d, T1v, T18, Ts, T1e, Tg, T16, T1s;
T1d = VSUB(T1b, T1c);
T1v = VADD(T1b, T1c);
T18 = VADD(Tq, Tr);
Ts = VSUB(Tq, Tr);
T1e = VADD(Te, Tf);
Tg = VSUB(Te, Tf);
T16 = VSUB(T14, T15);
T1s = VADD(T14, T15);
{
V T1t, T19, T1w, T1g;
T1t = VADD(T17, T18);
T19 = VSUB(T17, T18);
TG = VSUB(Ts, Tp);
Tt = VADD(Tp, Ts);
T1w = VADD(T1e, T1f);
T1g = VSUB(T1e, T1f);
Th = VADD(Td, Tg);
TO = 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);
}
}
}
}
TE = VSUB(Ta, Th);
Ti = VADD(Ta, Th);
{
V TL, T1k, T1A, Tj, TD, T1E, T1G, TK, TC, T1j, T1z, T1i, T1y, TB;
TL = VSUB(TA, Tt);
TB = VADD(Tt, TA);
T1i = VADD(T1a, T1h);
T1k = VSUB(T1a, T1h);
T1y = VADD(T1u, T1x);
T1A = VSUB(T1u, T1x);
Tj = VADD(T3, Ti);
TD = VFNMS(LDK(KP250000000), Ti, T3);
T1E = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1D, T1C));
T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1C, T1D));
TK = VFNMS(LDK(KP250000000), TB, Tm);
TC = VADD(Tm, TB);
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;
TU = VFNMS(LDK(KP618033988), TG, TH);
TI = VFMA(LDK(KP618033988), TH, TG);
TP = VFMA(LDK(KP618033988), TO, TN);
TX = VFNMS(LDK(KP618033988), TN, TO);
T1J = VFMAI(TC, Tj);
STM2(&(xo[30]), T1J, ovs, &(xo[2]));
T1K = VFNMSI(TC, Tj);
STM2(&(xo[10]), 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 = VFMAI(T1q, T1p);
STM2(&(xo[28]), T1L, ovs, &(xo[0]));
STN2(&(xo[28]), T1L, T1J, ovs);
T1M = VFNMSI(T1q, T1p);
STM2(&(xo[12]), T1M, ovs, &(xo[0]));
T1N = VFNMSI(T1o, T1l);
STM2(&(xo[36]), T1N, ovs, &(xo[0]));
T1O = VFMAI(T1o, T1l);
STM2(&(xo[4]), T1O, ovs, &(xo[0]));
T1P = VFNMSI(T1E, T1B);
STM2(&(xo[32]), T1P, ovs, &(xo[0]));
T1Q = VFMAI(T1E, T1B);
STM2(&(xo[8]), T1Q, ovs, &(xo[0]));
STN2(&(xo[8]), T1Q, T1K, ovs);
T1R = VFMAI(T1G, T1F);
STM2(&(xo[24]), T1R, ovs, &(xo[0]));
T1S = VFNMSI(T1G, T1F);
STM2(&(xo[16]), T1S, ovs, &(xo[0]));
TM = VFNMS(LDK(KP559016994), TL, TK);
TW = VFMA(LDK(KP559016994), TL, TK);
TT = VFNMS(LDK(KP559016994), TE, TD);
TF = VFMA(LDK(KP559016994), TE, TD);
}
}
}
{
V T10, TY, TQ, TS, TJ, TR, TZ, TV;
T10 = VFMA(LDK(KP951056516), TX, TW);
TY = VFNMS(LDK(KP951056516), TX, TW);
TQ = VFMA(LDK(KP951056516), TP, TM);
TS = VFNMS(LDK(KP951056516), TP, TM);
TJ = VFMA(LDK(KP951056516), TI, TF);
TR = VFNMS(LDK(KP951056516), TI, TF);
TZ = VFMA(LDK(KP951056516), TU, TT);
TV = VFNMS(LDK(KP951056516), TU, TT);
{
V T1T, T1U, T1V, T1W;
T1T = VFMAI(TS, TR);
STM2(&(xo[22]), T1T, ovs, &(xo[2]));
STN2(&(xo[20]), T1I, T1T, ovs);
T1U = VFNMSI(TS, TR);
STM2(&(xo[18]), T1U, ovs, &(xo[2]));
STN2(&(xo[16]), T1S, T1U, ovs);
T1V = VFMAI(TQ, TJ);
STM2(&(xo[38]), T1V, ovs, &(xo[2]));
STN2(&(xo[36]), T1N, T1V, ovs);
T1W = VFNMSI(TQ, TJ);
STM2(&(xo[2]), T1W, ovs, &(xo[2]));
STN2(&(xo[0]), T1H, T1W, ovs);
{
V T1X, T1Y, T1Z, T20;
T1X = VFMAI(TY, TV);
STM2(&(xo[6]), T1X, ovs, &(xo[2]));
STN2(&(xo[4]), T1O, T1X, ovs);
T1Y = VFNMSI(TY, TV);
STM2(&(xo[34]), T1Y, ovs, &(xo[2]));
STN2(&(xo[32]), T1P, T1Y, ovs);
T1Z = VFMAI(T10, TZ);
STM2(&(xo[14]), T1Z, ovs, &(xo[2]));
STN2(&(xo[12]), T1M, T1Z, ovs);
T20 = VFNMSI(T10, TZ);
STM2(&(xo[26]), T20, ovs, &(xo[2]));
STN2(&(xo[24]), T1R, T20, ovs);
}
}
}
}
}
static void n1bv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
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 TS, TA, TN, TV, TK, TU, TR, Tl;
{
V T3, TE, T1r, T13, Ta, TL, Tz, TG, Ts, TF, Th, TM, T1u, T1C, T1n;
V T1a, T1m, T1h, T1x, T1D, Tk, Ti;
{
V T1, T2, TC, TD;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
TC = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
TD = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
{
V T14, T6, T1c, Tv, Tm, T1f, Ty, T17, T9, Tn, Tp, T1b, Td, Tq, Te;
V Tf, T15, To;
{
V Tw, Tx, T7, T8, Tb, Tc;
{
V T4, T5, Tt, Tu, T11, T12;
T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
Tt = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Tu = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tw = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
T11 = VADD(T1, T2);
TE = VSUB(TC, TD);
T12 = VADD(TC, TD);
T14 = VADD(T4, T5);
T6 = VSUB(T4, T5);
T1c = VADD(Tt, Tu);
Tv = VSUB(Tt, Tu);
Tx = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T1r = VADD(T11, T12);
T13 = VSUB(T11, T12);
}
Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T1f = VADD(Tw, Tx);
Ty = VSUB(Tw, Tx);
T17 = VADD(T7, T8);
T9 = VSUB(T7, T8);
Tn = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
Tp = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T1b = VADD(Tb, Tc);
Td = VSUB(Tb, Tc);
Tq = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
}
Ta = VADD(T6, T9);
TL = VSUB(T6, T9);
T15 = VADD(Tm, Tn);
To = VSUB(Tm, Tn);
Tz = VSUB(Tv, Ty);
TG = VADD(Tv, Ty);
{
V T1d, T1v, T18, Tr, T1e, Tg, T16, T1s;
T1d = VSUB(T1b, T1c);
T1v = VADD(T1b, T1c);
T18 = VADD(Tp, Tq);
Tr = VSUB(Tp, Tq);
T1e = VADD(Te, Tf);
Tg = VSUB(Te, Tf);
T16 = VSUB(T14, T15);
T1s = VADD(T14, T15);
{
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);
ST(&(xo[0]), VADD(T1r, T1y), ovs, &(xo[0]));
T1z = VFNMS(LDK(KP250000000), T1y, T1r);
ST(&(xo[WS(os, 10)]), VADD(T13, T1i), ovs, &(xo[0]));
{
V T1p, T1l, T1o, T1q, T1F, T1B;
TS = VFNMS(LDK(KP618033988), Ts, Tz);
TA = VFMA(LDK(KP618033988), Tz, Ts);
TN = VFMA(LDK(KP618033988), TM, TL);
TV = VFNMS(LDK(KP618033988), TL, TM);
ST(&(xo[WS(os, 5)]), VFMAI(T10, TZ), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 15)]), VFNMSI(T10, TZ), ovs, &(xo[WS(os, 1)]));
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);
ST(&(xo[WS(os, 14)]), VFNMSI(T1q, T1p), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VFMAI(T1q, T1p), ovs, &(xo[0]));
ST(&(xo[WS(os, 18)]), VFMAI(T1o, T1l), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VFNMSI(T1o, T1l), ovs, &(xo[0]));
ST(&(xo[WS(os, 16)]), VFMAI(T1E, T1B), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VFNMSI(T1E, T1B), ovs, &(xo[0]));
ST(&(xo[WS(os, 12)]), VFNMSI(T1G, T1F), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VFMAI(T1G, T1F), ovs, &(xo[0]));
TK = VFMA(LDK(KP559016994), TJ, TI);
TU = VFNMS(LDK(KP559016994), TJ, TI);
TR = VFNMS(LDK(KP559016994), Tk, Tj);
Tl = VFMA(LDK(KP559016994), Tk, Tj);
}
}
}
{
V TY, TW, TO, TQ, TB, TP, TX, TT;
TY = VFMA(LDK(KP951056516), TV, TU);
TW = VFNMS(LDK(KP951056516), TV, TU);
TO = VFMA(LDK(KP951056516), TN, TK);
TQ = VFNMS(LDK(KP951056516), TN, TK);
TB = VFNMS(LDK(KP951056516), TA, Tl);
TP = VFMA(LDK(KP951056516), TA, Tl);
TX = VFNMS(LDK(KP951056516), TS, TR);
TT = VFMA(LDK(KP951056516), TS, TR);
ST(&(xo[WS(os, 9)]), VFMAI(TQ, TP), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 11)]), VFNMSI(TQ, TP), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 1)]), VFMAI(TO, TB), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 19)]), VFNMSI(TO, TB), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 17)]), VFMAI(TW, TT), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VFNMSI(TW, TT), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 13)]), VFMAI(TY, TX), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VFNMSI(TY, TX), ovs, &(xo[WS(os, 1)]));
}
}
}
static void hc2cfdftv_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
INT m;
for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 22)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(rs)) {
V T3, T7, TH, TE, Th, TC, Tq, T11, TU, Tx, Tb, Tz, Tu, Tw, Tp;
V Tl, T9, Ta, T8, Ty, Tn, To, Tm, TG, T1, T2, Tt, T5, T6, T4;
V Tv, Tj, Tk, Ti, TD, Tf, Tg, Te, TB, TT, TF, TR, Tr;
T1 = LD(&(Rp[0]), ms, &(Rp[0]));
T2 = LD(&(Rm[0]), -ms, &(Rm[0]));
Tt = LDW(&(W[0]));
T5 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0]));
T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0]));
T4 = LDW(&(W[TWVL * 6]));
Tv = LDW(&(W[TWVL * 8]));
Tn = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)]));
To = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)]));
T3 = VFMACONJ(T2, T1);
Tu = VZMULIJ(Tt, VFNMSCONJ(T2, T1));
Tm = LDW(&(W[TWVL * 2]));
TG = LDW(&(W[TWVL * 4]));
T7 = VZMULJ(T4, VFMACONJ(T6, T5));
Tw = VZMULIJ(Tv, VFNMSCONJ(T6, T5));
Tj = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)]));
Tk = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)]));
Ti = LDW(&(W[TWVL * 18]));
TD = LDW(&(W[TWVL * 20]));
Tp = VZMULJ(Tm, VFMACONJ(To, Tn));
TH = VZMULIJ(TG, VFNMSCONJ(To, Tn));
Tf = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)]));
Tg = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)]));
Te = LDW(&(W[TWVL * 10]));
TB = LDW(&(W[TWVL * 12]));
Tl = VZMULJ(Ti, VFMACONJ(Tk, Tj));
TE = VZMULIJ(TD, VFNMSCONJ(Tk, Tj));
T9 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0]));
Ta = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0]));
T8 = LDW(&(W[TWVL * 14]));
Ty = LDW(&(W[TWVL * 16]));
Th = VZMULJ(Te, VFMACONJ(Tg, Tf));
TC = VZMULIJ(TB, VFNMSCONJ(Tg, Tf));
Tq = VADD(Tl, Tp);
T11 = VSUB(Tp, Tl);
TU = VSUB(Tu, Tw);
Tx = VADD(Tu, Tw);
Tb = VZMULJ(T8, VFMACONJ(Ta, T9));
Tz = VZMULIJ(Ty, VFNMSCONJ(Ta, T9));
TT = VSUB(TC, TE);
TF = VADD(TC, TE);
TR = VFNMS(LDK(KP500000000), Tq, Th);
Tr = VADD(Th, Tq);
{
V TX, TA, T1d, TV, TY, TI, T1e, T12, TQ, Td, T10, Tc, T1a, TN, TJ;
V T1j, T1f, T1b, TS, TM, Ts, T17, T13, TZ, T1i, T1c, T16, TW, TP, TO;
V TL, TK, T1k, T1l, T1h, T1g, T18, T19, T15, T14;
T10 = VSUB(Tb, T7);
Tc = VADD(T7, Tb);
TX = VFNMS(LDK(KP500000000), Tx, Tz);
TA = VADD(Tx, Tz);
T1d = VADD(TU, TT);
TV = VSUB(TT, TU);
TY = VFNMS(LDK(KP500000000), TF, TH);
TI = VADD(TF, TH);
T1e = VADD(T10, T11);
T12 = VSUB(T10, T11);
TQ = VFNMS(LDK(KP500000000), Tc, T3);
Td = VADD(T3, Tc);
T1a = VADD(TX, TY);
TZ = VSUB(TX, TY);
TN = VADD(TA, TI);
TJ = VSUB(TA, TI);
T1j = VMUL(LDK(KP866025403), VADD(T1d, T1e));
T1f = VMUL(LDK(KP866025403), VSUB(T1d, T1e));
T1b = VADD(TQ, TR);
TS = VSUB(TQ, TR);
TM = VADD(Td, Tr);
Ts = VSUB(Td, Tr);
T17 = VFMA(LDK(KP866025403), T12, TZ);
T13 = VFNMS(LDK(KP866025403), T12, TZ);
T1i = VSUB(T1b, T1a);
T1c = VADD(T1a, T1b);
T16 = VFNMS(LDK(KP866025403), TV, TS);
TW = VFMA(LDK(KP866025403), TV, TS);
TP = VCONJ(VMUL(LDK(KP500000000), VADD(TN, TM)));
TO = VMUL(LDK(KP500000000), VSUB(TM, TN));
TL = VCONJ(VMUL(LDK(KP500000000), VFNMSI(TJ, Ts)));
TK = VMUL(LDK(KP500000000), VFMAI(TJ, Ts));
T1k = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1j, T1i)));
T1l = VMUL(LDK(KP500000000), VFMAI(T1j, T1i));
T1h = VMUL(LDK(KP500000000), VFMAI(T1f, T1c));
T1g = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1f, T1c)));
T18 = VMUL(LDK(KP500000000), VFNMSI(T17, T16));
T19 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T17, T16)));
T15 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T13, TW)));
T14 = VMUL(LDK(KP500000000), VFNMSI(T13, TW));
ST(&(Rm[WS(rs, 5)]), TP, -ms, &(Rm[WS(rs, 1)]));
ST(&(Rp[0]), TO, ms, &(Rp[0]));
ST(&(Rm[WS(rs, 2)]), TL, -ms, &(Rm[0]));
ST(&(Rp[WS(rs, 3)]), TK, ms, &(Rp[WS(rs, 1)]));
ST(&(Rm[WS(rs, 3)]), T1k, -ms, &(Rm[WS(rs, 1)]));
ST(&(Rp[WS(rs, 4)]), T1l, ms, &(Rp[0]));
ST(&(Rp[WS(rs, 2)]), T1h, ms, &(Rp[0]));
ST(&(Rm[WS(rs, 1)]), T1g, -ms, &(Rm[WS(rs, 1)]));
ST(&(Rp[WS(rs, 5)]), T18, ms, &(Rp[WS(rs, 1)]));
ST(&(Rm[WS(rs, 4)]), T19, -ms, &(Rm[0]));
ST(&(Rm[0]), T15, -ms, &(Rm[0]));
ST(&(Rp[WS(rs, 1)]), T14, ms, &(Rp[WS(rs, 1)]));
}
}
}
static void n2bv_6(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 = 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, T4, T5, T7, T8;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
{
V T3, Td, T6, Te, T9, Tf;
T3 = VSUB(T1, T2);
Td = VADD(T1, T2);
T6 = VSUB(T4, T5);
Te = VADD(T4, T5);
T9 = VSUB(T7, T8);
Tf = VADD(T7, T8);
{
V Tg, Ti, Ta, Tc;
Tg = VADD(Te, Tf);
Ti = VMUL(LDK(KP866025403), VSUB(Te, Tf));
Ta = VADD(T6, T9);
Tc = VMUL(LDK(KP866025403), VSUB(T6, T9));
{
V Th, Tj, Tb, Tk;
Th = VFNMS(LDK(KP500000000), Tg, Td);
Tj = VADD(Td, Tg);
STM2(&(xo[0]), Tj, ovs, &(xo[0]));
Tb = VFNMS(LDK(KP500000000), Ta, T3);
Tk = VADD(T3, Ta);
STM2(&(xo[6]), Tk, ovs, &(xo[2]));
{
V Tl, Tm, Tn, To;
Tl = VFMAI(Ti, Th);
STM2(&(xo[8]), Tl, ovs, &(xo[0]));
Tm = VFNMSI(Ti, Th);
STM2(&(xo[4]), Tm, ovs, &(xo[0]));
STN2(&(xo[4]), Tm, Tk, ovs);
Tn = VFNMSI(Tc, Tb);
STM2(&(xo[10]), Tn, ovs, &(xo[2]));
STN2(&(xo[8]), Tl, Tn, ovs);
To = VFMAI(Tc, Tb);
STM2(&(xo[2]), To, ovs, &(xo[2]));
STN2(&(xo[0]), Tj, To, ovs);
}
}
}
}
}
}
VLEAVE();
}
static void t3fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
{
INT m;
R *x;
x = ri;
for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(20, rs)) {
V T2, T8, T9, TA, T3, Tc, T4, TZ, T18, Tl, Tq, Tx, TU, Td, Te;
V T15, Ti, Tt, TJ;
T2 = LDW(&(W[0]));
T8 = LDW(&(W[TWVL * 2]));
T9 = VZMUL(T2, T8);
TA = VZMULJ(T2, T8);
T3 = LDW(&(W[TWVL * 4]));
Tc = VZMULJ(T9, T3);
T4 = VZMUL(T2, T3);
TZ = VZMUL(T9, T3);
T18 = VZMULJ(TA, T3);
Tl = VZMULJ(T8, T3);
Tq = VZMULJ(T2, T3);
Tx = VZMUL(T8, T3);
TU = VZMUL(TA, T3);
Td = LDW(&(W[TWVL * 6]));
Te = VZMULJ(Tc, Td);
T15 = VZMULJ(TA, Td);
Ti = VZMULJ(T8, Td);
Tt = VZMULJ(T2, Td);
TJ = VZMULJ(T9, Td);
{
V T7, TM, T1F, T23, T1i, T1p, T1q, T1j, Tp, TE, TF, T27, T28, T29, T1P;
V T1S, T1T, TY, T1c, T1d, T24, T25, T26, T1I, T1L, T1M, TG, T1e;
{
V T1, TL, T6, TI, TK, T5, TH, T1D, T1E;
T1 = LD(&(x[0]), ms, &(x[0]));
TK = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
TL = VZMULJ(TJ, TK);
T5 = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
T6 = VZMULJ(T4, T5);
TH = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
TI = VZMULJ(Tc, TH);
T7 = VSUB(T1, T6);
TM = VSUB(TI, TL);
T1D = VADD(T1, T6);
T1E = VADD(TI, TL);
T1F = VSUB(T1D, T1E);
T23 = VADD(T1D, T1E);
}
{
V Th, T1G, T14, T1O, T1b, T1R, To, T1J, Tw, T1N, TR, T1H, TX, T1K, TD;
V T1Q;
{
V Tb, Tg, Ta, Tf;
Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tb = VZMULJ(T9, Ta);
Tf = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
Tg = VZMULJ(Te, Tf);
Th = VSUB(Tb, Tg);
T1G = VADD(Tb, Tg);
}
{
V T11, T13, T10, T12;
T10 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
T11 = VZMULJ(TZ, T10);
T12 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T13 = VZMULJ(T8, T12);
T14 = VSUB(T11, T13);
T1O = VADD(T11, T13);
}
{
V T17, T1a, T16, T19;
T16 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)]));
T17 = VZMULJ(T15, T16);
T19 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T1a = VZMULJ(T18, T19);
T1b = VSUB(T17, T1a);
T1R = VADD(T17, T1a);
}
{
V Tk, Tn, Tj, Tm;
Tj = LD(&(x[WS(rs, 16)]), ms, &(x[0]));
Tk = VZMULJ(Ti, Tj);
Tm = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tn = VZMULJ(Tl, Tm);
To = VSUB(Tk, Tn);
T1J = VADD(Tk, Tn);
}
{
V Ts, Tv, Tr, Tu;
Tr = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Ts = VZMULJ(Tq, Tr);
Tu = LD(&(x[WS(rs, 18)]), ms, &(x[0]));
Tv = VZMULJ(Tt, Tu);
Tw = VSUB(Ts, Tv);
T1N = VADD(Ts, Tv);
}
{
V TO, TQ, TN, TP;
TN = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
TO = VZMULJ(T3, TN);
TP = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)]));
TQ = VZMULJ(Td, TP);
TR = VSUB(TO, TQ);
T1H = VADD(TO, TQ);
}
{
V TT, TW, TS, TV;
TS = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
TT = VZMULJ(T2, TS);
TV = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
TW = VZMULJ(TU, TV);
TX = VSUB(TT, TW);
T1K = VADD(TT, TW);
}
{
V Tz, TC, Ty, TB;
Ty = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
Tz = VZMULJ(Tx, Ty);
TB = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
TC = VZMULJ(TA, TB);
TD = VSUB(Tz, TC);
T1Q = VADD(Tz, TC);
}
T1i = VSUB(TX, TR);
T1p = VSUB(Th, To);
T1q = VSUB(Tw, TD);
T1j = VSUB(T1b, T14);
Tp = VADD(Th, To);
TE = VADD(Tw, TD);
TF = VADD(Tp, TE);
T27 = VADD(T1N, T1O);
T28 = VADD(T1Q, T1R);
T29 = VADD(T27, T28);
T1P = VSUB(T1N, T1O);
T1S = VSUB(T1Q, T1R);
T1T = VADD(T1P, T1S);
TY = VADD(TR, TX);
T1c = VADD(T14, T1b);
T1d = VADD(TY, T1c);
T24 = VADD(T1G, T1H);
T25 = VADD(T1J, T1K);
T26 = VADD(T24, T25);
T1I = VSUB(T1G, T1H);
T1L = VSUB(T1J, T1K);
T1M = VADD(T1I, T1L);
}
TG = VADD(T7, TF);
T1e = VADD(TM, T1d);
ST(&(x[WS(rs, 5)]), VFNMSI(T1e, TG), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 15)]), VFMAI(T1e, TG), ms, &(x[WS(rs, 1)]));
{
V T2c, T2a, T2b, T2g, T2i, T2e, T2f, T2h, T2d;
T2c = VSUB(T26, T29);
T2a = VADD(T26, T29);
T2b = VFNMS(LDK(KP250000000), T2a, T23);
T2e = VSUB(T24, T25);
T2f = VSUB(T27, T28);
T2g = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T2f, T2e));
T2i = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T2e, T2f));
ST(&(x[0]), VADD(T23, T2a), ms, &(x[0]));
T2h = VFNMS(LDK(KP559016994), T2c, T2b);
ST(&(x[WS(rs, 8)]), VFNMSI(T2i, T2h), ms, &(x[0]));
ST(&(x[WS(rs, 12)]), VFMAI(T2i, T2h), ms, &(x[0]));
T2d = VFMA(LDK(KP559016994), T2c, T2b);
ST(&(x[WS(rs, 4)]), VFMAI(T2g, T2d), ms, &(x[0]));
ST(&(x[WS(rs, 16)]), VFNMSI(T2g, T2d), ms, &(x[0]));
}
{
V T1W, T1U, T1V, T20, T22, T1Y, T1Z, T21, T1X;
T1W = VSUB(T1M, T1T);
T1U = VADD(T1M, T1T);
T1V = VFNMS(LDK(KP250000000), T1U, T1F);
T1Y = VSUB(T1P, T1S);
T1Z = VSUB(T1I, T1L);
T20 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1Z, T1Y));
T22 = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1Y, T1Z));
ST(&(x[WS(rs, 10)]), VADD(T1F, T1U), ms, &(x[0]));
T21 = VFMA(LDK(KP559016994), T1W, T1V);
ST(&(x[WS(rs, 6)]), VFNMSI(T22, T21), ms, &(x[0]));
ST(&(x[WS(rs, 14)]), VFMAI(T22, T21), ms, &(x[0]));
T1X = VFNMS(LDK(KP559016994), T1W, T1V);
ST(&(x[WS(rs, 2)]), VFMAI(T20, T1X), ms, &(x[0]));
ST(&(x[WS(rs, 18)]), VFNMSI(T20, T1X), ms, &(x[0]));
}
{
V T1k, T1r, T1z, T1w, T1o, T1y, T1h, T1v;
T1k = VFMA(LDK(KP618033988), T1j, T1i);
T1r = VFMA(LDK(KP618033988), T1q, T1p);
T1z = VFNMS(LDK(KP618033988), T1p, T1q);
T1w = VFNMS(LDK(KP618033988), T1i, T1j);
{
V T1m, T1n, T1f, T1g;
T1m = VFNMS(LDK(KP250000000), T1d, TM);
T1n = VSUB(T1c, TY);
T1o = VFNMS(LDK(KP559016994), T1n, T1m);
T1y = VFMA(LDK(KP559016994), T1n, T1m);
T1f = VFNMS(LDK(KP250000000), TF, T7);
T1g = VSUB(Tp, TE);
T1h = VFMA(LDK(KP559016994), T1g, T1f);
T1v = VFNMS(LDK(KP559016994), T1g, T1f);
}
{
V T1l, T1s, T1B, T1C;
T1l = VFMA(LDK(KP951056516), T1k, T1h);
T1s = VFMA(LDK(KP951056516), T1r, T1o);
ST(&(x[WS(rs, 1)]), VFNMSI(T1s, T1l), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 19)]), VFMAI(T1s, T1l), ms, &(x[WS(rs, 1)]));
T1B = VFMA(LDK(KP951056516), T1w, T1v);
T1C = VFMA(LDK(KP951056516), T1z, T1y);
ST(&(x[WS(rs, 13)]), VFNMSI(T1C, T1B), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(T1C, T1B), ms, &(x[WS(rs, 1)]));
}
{
V T1t, T1u, T1x, T1A;
T1t = VFNMS(LDK(KP951056516), T1k, T1h);
T1u = VFNMS(LDK(KP951056516), T1r, T1o);
ST(&(x[WS(rs, 9)]), VFNMSI(T1u, T1t), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 11)]), VFMAI(T1u, T1t), ms, &(x[WS(rs, 1)]));
T1x = VFNMS(LDK(KP951056516), T1w, T1v);
T1A = VFNMS(LDK(KP951056516), T1z, T1y);
ST(&(x[WS(rs, 17)]), VFNMSI(T1A, T1x), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFMAI(T1A, T1x), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
VLEAVE();
}
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)]));
}
}
}
}
}
}
static void t1bv_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP823639103, +0.823639103546331925877420039278190003029660514);
DVK(KP910592997, +0.910592997310029334643087372129977886038870291);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
{
INT m;
R *x;
x = ii;
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(15, rs)) {
V Tq, Ty, Th, TV, TK, Ts, T1f, T7, Tu, TA, TC, Tj, Tk, T1g, Tf;
{
V T1, T4, T2, T9, Te;
T1 = LD(&(x[0]), ms, &(x[0]));
T4 = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
{
V T8, Tp, Tx, Tg;
T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
Tg = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
{
V Tb, Td, Tr, T6, Tt, Tz, TB, Ti;
{
V T5, T3, Ta, Tc;
Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Tc = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
T5 = BYTW(&(W[TWVL * 18]), T4);
T3 = BYTW(&(W[TWVL * 8]), T2);
T9 = BYTW(&(W[TWVL * 4]), T8);
Tq = BYTW(&(W[TWVL * 10]), Tp);
Ty = BYTW(&(W[TWVL * 16]), Tx);
Th = BYTW(&(W[TWVL * 22]), Tg);
Tb = BYTW(&(W[TWVL * 14]), Ta);
Td = BYTW(&(W[TWVL * 24]), Tc);
Tr = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
TV = VSUB(T3, T5);
T6 = VADD(T3, T5);
Tt = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
}
Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
TB = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Te = VADD(Tb, Td);
TK = VSUB(Tb, Td);
Ts = BYTW(&(W[TWVL * 20]), Tr);
T1f = VADD(T1, T6);
T7 = VFNMS(LDK(KP500000000), T6, T1);
Tu = BYTW(&(W[0]), Tt);
TA = BYTW(&(W[TWVL * 26]), Tz);
TC = BYTW(&(W[TWVL * 6]), TB);
Tj = BYTW(&(W[TWVL * 2]), Ti);
Tk = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
}
}
T1g = VADD(T9, Te);
Tf = VFNMS(LDK(KP500000000), Te, T9);
}
{
V Tv, TN, TD, TO, Tl;
Tv = VADD(Ts, Tu);
TN = VSUB(Ts, Tu);
TD = VADD(TA, TC);
TO = VSUB(TA, TC);
Tl = BYTW(&(W[TWVL * 12]), Tk);
{
V Tw, T1j, TX, TP, TE, T1k, TL, Tm;
Tw = VFNMS(LDK(KP500000000), Tv, Tq);
T1j = VADD(Tq, Tv);
TX = VADD(TN, TO);
TP = VSUB(TN, TO);
TE = VFNMS(LDK(KP500000000), TD, Ty);
T1k = VADD(Ty, TD);
TL = VSUB(Tj, Tl);
Tm = VADD(Tj, Tl);
{
V TT, TF, T1q, T1l, TW, TM, T1h, Tn;
TT = VSUB(Tw, TE);
TF = VADD(Tw, TE);
T1q = VSUB(T1j, T1k);
T1l = VADD(T1j, T1k);
TW = VADD(TK, TL);
TM = VSUB(TK, TL);
T1h = VADD(Th, Tm);
Tn = VFNMS(LDK(KP500000000), Tm, Th);
{
V T10, TY, T16, TQ, T1r, T1i, TS, To, TZ, T1e;
T10 = VSUB(TW, TX);
TY = VADD(TW, TX);
T16 = VFNMS(LDK(KP618033988), TM, TP);
TQ = VFMA(LDK(KP618033988), TP, TM);
T1r = VSUB(T1g, T1h);
T1i = VADD(T1g, T1h);
TS = VSUB(Tf, Tn);
To = VADD(Tf, Tn);
TZ = VFNMS(LDK(KP250000000), TY, TV);
T1e = VMUL(LDK(KP866025403), VADD(TV, TY));
{
V T1u, T1s, T1o, T18, TU, TG, TI, T19, T11, T1n, T1m;
T1u = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1q, T1r));
T1s = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1r, T1q));
T1m = VADD(T1i, T1l);
T1o = VSUB(T1i, T1l);
T18 = VFNMS(LDK(KP618033988), TS, TT);
TU = VFMA(LDK(KP618033988), TT, TS);
TG = VADD(To, TF);
TI = VSUB(To, TF);
T19 = VFNMS(LDK(KP559016994), T10, TZ);
T11 = VFMA(LDK(KP559016994), T10, TZ);
ST(&(x[0]), VADD(T1f, T1m), ms, &(x[0]));
T1n = VFNMS(LDK(KP250000000), T1m, T1f);
{
V T1a, T1c, T14, T12, T1p, T1t, T15, TJ, T1d, TH;
T1d = VADD(T7, TG);
TH = VFNMS(LDK(KP250000000), TG, T7);
T1a = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T19, T18));
T1c = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T19, T18));
T14 = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T11, TU));
T12 = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T11, TU));
T1p = VFNMS(LDK(KP559016994), T1o, T1n);
T1t = VFMA(LDK(KP559016994), T1o, T1n);
ST(&(x[WS(rs, 10)]), VFMAI(T1e, T1d), ms, &(x[0]));
ST(&(x[WS(rs, 5)]), VFNMSI(T1e, T1d), ms, &(x[WS(rs, 1)]));
T15 = VFNMS(LDK(KP559016994), TI, TH);
TJ = VFMA(LDK(KP559016994), TI, TH);
{
V T17, T1b, T13, TR;
ST(&(x[WS(rs, 12)]), VFNMSI(T1s, T1p), ms, &(x[0]));
ST(&(x[WS(rs, 3)]), VFMAI(T1s, T1p), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 9)]), VFNMSI(T1u, T1t), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 6)]), VFMAI(T1u, T1t), ms, &(x[0]));
T17 = VFNMS(LDK(KP823639103), T16, T15);
T1b = VFMA(LDK(KP823639103), T16, T15);
T13 = VFMA(LDK(KP823639103), TQ, TJ);
TR = VFNMS(LDK(KP823639103), TQ, TJ);
ST(&(x[WS(rs, 13)]), VFMAI(T1a, T17), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 2)]), VFNMSI(T1a, T17), ms, &(x[0]));
ST(&(x[WS(rs, 8)]), VFMAI(T1c, T1b), ms, &(x[0]));
ST(&(x[WS(rs, 7)]), VFNMSI(T1c, T1b), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 11)]), VFMAI(T14, T13), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 4)]), VFNMSI(T14, T13), ms, &(x[0]));
ST(&(x[WS(rs, 14)]), VFNMSI(T12, TR), ms, &(x[0]));
ST(&(x[WS(rs, 1)]), VFMAI(T12, TR), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
}
}
}
VLEAVE();
}
static void t3bv_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 = ii;
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(rs)) {
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, TQ, TX, T1g, Tl, To, Tw, TG, Tz, T1p;
V T1e, TC;
{
V T1, T5, Ta, Td;
T1 = LD(&(x[0]), ms, &(x[0]));
T5 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Td = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
{
V TR, TN, TM, TE, Tb, Tp, Tm, Te, T6, TW, TO, TS;
{
V TL, Tx, T9, TU, Tc, T4, TV;
TL = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tx = VZMULJ(T2, T8);
T9 = VZMUL(T2, T8);
TR = VZMULJ(T2, Tu);
TU = VZMULJ(T8, T3);
Tc = VZMUL(T8, T3);
T4 = VZMULJ(T2, T3);
TN = VZMUL(T2, T3);
TV = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
TM = VZMUL(Tx, TL);
Ty = VZMULJ(Tx, T3);
TE = VZMUL(Tx, T3);
Tb = VZMUL(T9, Ta);
Tp = VZMUL(T9, T3);
Tm = VZMULJ(T9, T3);
Te = VZMUL(Tc, Td);
T6 = VZMUL(T4, T5);
TW = VZMUL(TU, TV);
}
TO = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
TS = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
{
V TP, TT, Ti, Tk, Tn, Th, Tq, Tj;
Th = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Tq = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
Tj = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
T1o = VSUB(Tb, Te);
Tf = VADD(Tb, Te);
T1b = VSUB(T1, T6);
T7 = VADD(T1, T6);
TP = VZMUL(TN, TO);
TT = VZMUL(TR, TS);
Ti = VZMUL(T2, Th);
Tr = VZMUL(Tp, Tq);
Tk = VZMUL(T3, Tj);
Tn = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
{
V T1c, T1d, Tv, TF;
Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
T1c = VSUB(TM, TP);
TQ = VADD(TM, TP);
T1d = VSUB(TT, TW);
TX = VADD(TT, TW);
T1g = VSUB(Ti, Tk);
Tl = VADD(Ti, Tk);
To = VZMUL(Tm, Tn);
Tw = VZMUL(Tu, Tv);
TG = VZMUL(TE, TF);
Tz = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T1p = VSUB(T1c, T1d);
T1e = VADD(T1c, T1d);
TC = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
}
}
}
}
{
V T1h, Ts, TA, TD;
T13 = VADD(T7, Tf);
Tg = VSUB(T7, Tf);
T1h = VSUB(To, Tr);
Ts = VADD(To, Tr);
TY = VSUB(TQ, TX);
T14 = VADD(TQ, TX);
TA = VZMUL(Ty, Tz);
T1A = VFNMS(LDK(KP707106781), T1p, T1o);
T1q = VFMA(LDK(KP707106781), T1p, T1o);
T1f = VFMA(LDK(KP707106781), T1e, T1b);
T1x = VFNMS(LDK(KP707106781), T1e, T1b);
TD = VZMUL(T8, TC);
T1r = VFMA(LDK(KP414213562), T1g, T1h);
T1i = VFNMS(LDK(KP414213562), T1h, T1g);
Tt = VSUB(Tl, Ts);
T16 = VADD(Tl, Ts);
TB = VADD(Tw, TA);
T1j = VSUB(Tw, TA);
T1k = VSUB(TG, TD);
TH = VADD(TD, TG);
}
}
}
{
V T15, T19, T1l, T1s, TI, T17;
T15 = VSUB(T13, T14);
T19 = VADD(T13, T14);
T1l = VFNMS(LDK(KP414213562), T1k, T1j);
T1s = VFMA(LDK(KP414213562), T1j, T1k);
TI = VSUB(TB, TH);
T17 = VADD(TB, TH);
{
V T1y, T1t, T1B, T1m;
T1y = VADD(T1r, T1s);
T1t = VSUB(T1r, T1s);
T1B = VSUB(T1i, T1l);
T1m = VADD(T1i, T1l);
{
V T18, T1a, TJ, TZ;
T18 = VSUB(T16, T17);
T1a = VADD(T16, T17);
TJ = VADD(Tt, TI);
TZ = VSUB(Tt, TI);
{
V T1u, T1w, T1z, T1D;
T1u = VFNMS(LDK(KP923879532), T1t, T1q);
T1w = VFMA(LDK(KP923879532), T1t, T1q);
T1z = VFNMS(LDK(KP923879532), T1y, T1x);
T1D = VFMA(LDK(KP923879532), T1y, T1x);
{
V T1n, T1v, T1C, T1E;
T1n = VFNMS(LDK(KP923879532), T1m, T1f);
T1v = VFMA(LDK(KP923879532), T1m, T1f);
T1C = VFMA(LDK(KP923879532), T1B, T1A);
T1E = VFNMS(LDK(KP923879532), T1B, T1A);
ST(&(x[WS(rs, 8)]), VSUB(T19, T1a), ms, &(x[0]));
ST(&(x[0]), VADD(T19, T1a), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VFMAI(T18, T15), ms, &(x[0]));
ST(&(x[WS(rs, 12)]), VFNMSI(T18, T15), ms, &(x[0]));
{
V T10, T12, TK, T11;
T10 = VFNMS(LDK(KP707106781), TZ, TY);
T12 = VFMA(LDK(KP707106781), TZ, TY);
TK = VFNMS(LDK(KP707106781), TJ, Tg);
T11 = VFMA(LDK(KP707106781), TJ, Tg);
ST(&(x[WS(rs, 15)]), VFNMSI(T1w, T1v), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFMAI(T1w, T1v), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 9)]), VFMAI(T1u, T1n), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFNMSI(T1u, T1n), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFNMSI(T1E, T1D), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 13)]), VFMAI(T1E, T1D), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 11)]), VFNMSI(T1C, T1z), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 5)]), VFMAI(T1C, T1z), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 2)]), VFMAI(T12, T11), ms, &(x[0]));
ST(&(x[WS(rs, 14)]), VFNMSI(T12, T11), ms, &(x[0]));
ST(&(x[WS(rs, 10)]), VFMAI(T10, TK), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(T10, TK), ms, &(x[0]));
}
}
}
}
}
}
}
}
VLEAVE();
}
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(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)]));
ST(&(x[WS(rs, 11)]), VFMAI(TF, Ts), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFNMSI(TF, Ts), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
static void n2bv_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 = ii;
xo = io;
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 T31, T32, T33, T34, T35, T36, T37, T38, T39, T3a, T3b, T3c, T1h, Tr, T3d;
V T3e, T3f, T3g, T1a, T1k, TI, T1b, T1L, T1P, T1I, T1G, T1O, T1Q, T1H, T1z;
V T1c, TZ;
{
V T2x, T1T, T2K, T1W, T1p, Tb, T1A, T16, Tu, TF, T2O, T2H, T2b, T2t, TY;
V T1w, TT, T1v, T20, T2C, Tj, Te, T2e, To, T2i, T23, T2D, TB, TG, Th;
V T2f, 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(T1U, T1V);
T1W = VADD(T1U, T1V);
Ta = VADD(T6, T9);
T15 = VSUB(T6, T9);
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 = VFNMS(LDK(KP707106781), T15, T14);
T16 = VFMA(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);
T2O = VFMA(LDK(KP414213562), T2F, T2G);
T2H = VFNMS(LDK(KP414213562), T2G, T2F);
T2b = VSUB(T27, T2a);
T2t = VADD(T27, T2a);
TY = VFMA(LDK(KP707106781), TX, TW);
T1w = VFNMS(LDK(KP707106781), TX, TW);
TT = VFMA(LDK(KP707106781), TS, TL);
T1v = VFNMS(LDK(KP707106781), TS, TL);
Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
}
}
T20 = VADD(T1Y, T1Z);
T2C = VSUB(T1Y, T1Z);
{
V Tc, Td, Tm, Tn;
Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
Tm = LD(&(xi[WS(is, 22)]), ivs, &(xi[0]));
Tn = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
{
V Tf, TA, T22, Tg;
Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
TA = VSUB(Ty, Tz);
T22 = VADD(Ty, Tz);
Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0]));
Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0]));
Te = VSUB(Tc, Td);
T2e = VADD(Tc, Td);
To = VSUB(Tm, Tn);
T2i = VADD(Tn, Tm);
T23 = VADD(T21, T22);
T2D = VSUB(T21, T22);
TB = VADD(Tx, TA);
TG = VSUB(Tx, TA);
Th = VSUB(Tf, Tg);
T2f = VADD(Tf, Tg);
Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
}
}
}
}
{
V T1t, TH, T1s, TC, T2P, T2U, T2n, T2d, T2w, T2u, T1q, T19, T1B, Tq, T2W;
V T2M, T2B, T2T, T2v, T2r, T2o, T2m, T2X, T2I;
{
V T1X, T2p, T2E, T2N, T2s, T2y, T2g, T17, Ti, T2h, Tl, T2c, T2l, T24;
T1X = VSUB(T1T, T1W);
T2p = VADD(T1T, T1W);
T2E = VFNMS(LDK(KP414213562), T2D, T2C);
T2N = VFMA(LDK(KP414213562), T2C, T2D);
T2s = VADD(T20, T23);
T24 = VSUB(T20, T23);
T1t = VFNMS(LDK(KP707106781), TG, TF);
TH = VFMA(LDK(KP707106781), TG, TF);
T1s = VFNMS(LDK(KP707106781), TB, Tu);
TC = VFMA(LDK(KP707106781), TB, Tu);
T2y = VSUB(T2e, T2f);
T2g = VADD(T2e, T2f);
T17 = VFMA(LDK(KP414213562), Te, Th);
Ti = VFNMS(LDK(KP414213562), Th, Te);
T2h = VADD(Tj, Tk);
Tl = VSUB(Tj, Tk);
T2c = VADD(T24, T2b);
T2l = VSUB(T24, T2b);
{
V T2L, T2A, T2q, T2k;
T2P = VSUB(T2N, T2O);
T2U = VADD(T2N, T2O);
{
V T2z, T2j, T18, Tp;
T2z = VSUB(T2h, T2i);
T2j = VADD(T2h, T2i);
T18 = VFMA(LDK(KP414213562), Tl, To);
Tp = VFNMS(LDK(KP414213562), To, Tl);
T2n = VFMA(LDK(KP707106781), T2c, T1X);
T2d = VFNMS(LDK(KP707106781), T2c, T1X);
T2w = VADD(T2s, T2t);
T2u = VSUB(T2s, T2t);
T2L = VSUB(T2y, T2z);
T2A = VADD(T2y, T2z);
T2q = VADD(T2g, T2j);
T2k = VSUB(T2g, T2j);
T1q = VADD(T17, T18);
T19 = VSUB(T17, T18);
T1B = VSUB(Ti, Tp);
Tq = VADD(Ti, Tp);
}
T2W = VFNMS(LDK(KP707106781), T2L, T2K);
T2M = VFMA(LDK(KP707106781), T2L, T2K);
T2B = VFMA(LDK(KP707106781), T2A, T2x);
T2T = VFNMS(LDK(KP707106781), T2A, T2x);
T2v = VADD(T2p, T2q);
T2r = VSUB(T2p, T2q);
T2o = VFMA(LDK(KP707106781), T2l, T2k);
T2m = VFNMS(LDK(KP707106781), T2l, T2k);
T2X = VSUB(T2E, T2H);
T2I = VADD(T2E, T2H);
}
}
{
V T2V, T2Z, T2Y, T30, T2R, T2J;
T2V = VFNMS(LDK(KP923879532), T2U, T2T);
T2Z = VFMA(LDK(KP923879532), T2U, T2T);
T31 = VSUB(T2v, T2w);
STM2(&(xo[32]), T31, ovs, &(xo[0]));
T32 = VADD(T2v, T2w);
STM2(&(xo[0]), T32, ovs, &(xo[0]));
T33 = VFMAI(T2u, T2r);
STM2(&(xo[16]), T33, ovs, &(xo[0]));
T34 = VFNMSI(T2u, T2r);
STM2(&(xo[48]), T34, ovs, &(xo[0]));
T35 = VFMAI(T2o, T2n);
STM2(&(xo[8]), T35, ovs, &(xo[0]));
T36 = VFNMSI(T2o, T2n);
STM2(&(xo[56]), T36, ovs, &(xo[0]));
T37 = VFMAI(T2m, T2d);
STM2(&(xo[40]), T37, ovs, &(xo[0]));
T38 = VFNMSI(T2m, T2d);
STM2(&(xo[24]), T38, ovs, &(xo[0]));
T2Y = VFMA(LDK(KP923879532), T2X, T2W);
T30 = VFNMS(LDK(KP923879532), T2X, T2W);
T2R = VFMA(LDK(KP923879532), T2I, T2B);
T2J = VFNMS(LDK(KP923879532), T2I, T2B);
{
V T1J, T1r, T1C, T1M, T2S, T2Q, T1u, T1D, T1E, T1x;
T1J = VFNMS(LDK(KP923879532), T1q, T1p);
T1r = VFMA(LDK(KP923879532), T1q, T1p);
T1C = VFNMS(LDK(KP923879532), T1B, T1A);
T1M = VFMA(LDK(KP923879532), T1B, T1A);
T39 = VFNMSI(T30, T2Z);
STM2(&(xo[12]), T39, ovs, &(xo[0]));
T3a = VFMAI(T30, T2Z);
STM2(&(xo[52]), T3a, ovs, &(xo[0]));
T3b = VFNMSI(T2Y, T2V);
STM2(&(xo[44]), T3b, ovs, &(xo[0]));
T3c = VFMAI(T2Y, T2V);
STM2(&(xo[20]), T3c, ovs, &(xo[0]));
T2S = VFMA(LDK(KP923879532), T2P, T2M);
T2Q = VFNMS(LDK(KP923879532), T2P, T2M);
T1u = VFMA(LDK(KP668178637), T1t, T1s);
T1D = VFNMS(LDK(KP668178637), T1s, T1t);
T1E = VFNMS(LDK(KP668178637), T1v, T1w);
T1x = VFMA(LDK(KP668178637), T1w, T1v);
{
V T1K, T1F, T1N, T1y;
T1h = VFNMS(LDK(KP923879532), Tq, Tb);
Tr = VFMA(LDK(KP923879532), Tq, Tb);
T3d = VFNMSI(T2S, T2R);
STM2(&(xo[60]), T3d, ovs, &(xo[0]));
T3e = VFMAI(T2S, T2R);
STM2(&(xo[4]), T3e, ovs, &(xo[0]));
T3f = VFMAI(T2Q, T2J);
STM2(&(xo[36]), T3f, ovs, &(xo[0]));
T3g = VFNMSI(T2Q, T2J);
STM2(&(xo[28]), T3g, ovs, &(xo[0]));
T1K = VADD(T1D, T1E);
T1F = VSUB(T1D, T1E);
T1N = VSUB(T1u, T1x);
T1y = VADD(T1u, T1x);
T1a = VFMA(LDK(KP923879532), T19, T16);
T1k = VFNMS(LDK(KP923879532), T19, T16);
TI = VFNMS(LDK(KP198912367), TH, TC);
T1b = VFMA(LDK(KP198912367), TC, TH);
T1L = VFMA(LDK(KP831469612), T1K, T1J);
T1P = VFNMS(LDK(KP831469612), T1K, T1J);
T1I = VFMA(LDK(KP831469612), T1F, T1C);
T1G = VFNMS(LDK(KP831469612), T1F, T1C);
T1O = VFNMS(LDK(KP831469612), T1N, T1M);
T1Q = VFMA(LDK(KP831469612), T1N, T1M);
T1H = VFMA(LDK(KP831469612), T1y, T1r);
T1z = VFNMS(LDK(KP831469612), T1y, T1r);
T1c = VFMA(LDK(KP198912367), TT, TY);
TZ = VFNMS(LDK(KP198912367), TY, TT);
}
}
}
}
}
{
V T1d, T1i, T10, T1l;
{
V T3h, T3i, T3j, T3k;
T3h = VFMAI(T1O, T1L);
STM2(&(xo[42]), T3h, ovs, &(xo[2]));
STN2(&(xo[40]), T37, T3h, ovs);
T3i = VFNMSI(T1O, T1L);
STM2(&(xo[22]), T3i, ovs, &(xo[2]));
STN2(&(xo[20]), T3c, T3i, ovs);
T3j = VFNMSI(T1Q, T1P);
STM2(&(xo[54]), T3j, ovs, &(xo[2]));
STN2(&(xo[52]), T3a, T3j, ovs);
T3k = VFMAI(T1Q, T1P);
STM2(&(xo[10]), T3k, ovs, &(xo[2]));
STN2(&(xo[8]), T35, T3k, ovs);
{
V T3l, T3m, T3n, T3o;
T3l = VFMAI(T1I, T1H);
STM2(&(xo[58]), T3l, ovs, &(xo[2]));
STN2(&(xo[56]), T36, T3l, ovs);
T3m = VFNMSI(T1I, T1H);
STM2(&(xo[6]), T3m, ovs, &(xo[2]));
STN2(&(xo[4]), T3e, T3m, ovs);
T3n = VFMAI(T1G, T1z);
STM2(&(xo[26]), T3n, ovs, &(xo[2]));
STN2(&(xo[24]), T38, T3n, ovs);
T3o = VFNMSI(T1G, T1z);
STM2(&(xo[38]), T3o, ovs, &(xo[2]));
STN2(&(xo[36]), T3f, T3o, ovs);
T1d = VSUB(T1b, T1c);
T1i = VADD(T1b, T1c);
T10 = VADD(TI, TZ);
T1l = VSUB(TI, TZ);
}
}
{
V T1n, T1j, T1e, T1g, T1o, T1m, T11, T1f;
T1n = VFMA(LDK(KP980785280), T1i, T1h);
T1j = VFNMS(LDK(KP980785280), T1i, T1h);
T1e = VFNMS(LDK(KP980785280), T1d, T1a);
T1g = VFMA(LDK(KP980785280), T1d, T1a);
T1o = VFNMS(LDK(KP980785280), T1l, T1k);
T1m = VFMA(LDK(KP980785280), T1l, T1k);
T11 = VFNMS(LDK(KP980785280), T10, Tr);
T1f = VFMA(LDK(KP980785280), T10, Tr);
{
V T3p, T3q, T3r, T3s;
T3p = VFNMSI(T1m, T1j);
STM2(&(xo[46]), T3p, ovs, &(xo[2]));
STN2(&(xo[44]), T3b, T3p, ovs);
T3q = VFMAI(T1m, T1j);
STM2(&(xo[18]), T3q, ovs, &(xo[2]));
STN2(&(xo[16]), T33, T3q, ovs);
T3r = VFMAI(T1o, T1n);
STM2(&(xo[50]), T3r, ovs, &(xo[2]));
STN2(&(xo[48]), T34, T3r, ovs);
T3s = VFNMSI(T1o, T1n);
STM2(&(xo[14]), T3s, ovs, &(xo[2]));
STN2(&(xo[12]), T39, T3s, ovs);
{
V T3t, T3u, T3v, T3w;
T3t = VFMAI(T1g, T1f);
STM2(&(xo[2]), T3t, ovs, &(xo[2]));
STN2(&(xo[0]), T32, T3t, ovs);
T3u = VFNMSI(T1g, T1f);
STM2(&(xo[62]), T3u, ovs, &(xo[2]));
STN2(&(xo[60]), T3d, T3u, ovs);
T3v = VFMAI(T1e, T11);
STM2(&(xo[34]), T3v, ovs, &(xo[2]));
STN2(&(xo[32]), T31, T3v, ovs);
T3w = VFNMSI(T1e, T11);
STM2(&(xo[30]), T3w, ovs, &(xo[2]));
STN2(&(xo[28]), T3g, T3w, ovs);
}
}
}
}
}
}
VLEAVE();
}
static void n2fv_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(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(is), MAKE_VOLATILE_STRIDE(os)) {
V T7, Tu, TF, TB, T13, TL, TO, TX, TC, Te, TP, Th, TQ, Tk, TW;
V T16;
{
V TH, TU, Tz, Tf, TK, TV, TA, TM, Ta, TN, Td, Tg, Ti, Tj;
{
V T1, T2, T4, T5, To, Tp, Tr, Ts;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
To = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
Tp = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Tr = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Ts = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
{
V T8, TJ, Tq, TI, Tt, T9, Tb, Tc, T3, T6;
T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
TH = VSUB(T1, T2);
T3 = VADD(T1, T2);
TU = VSUB(T4, T5);
T6 = VADD(T4, T5);
TJ = VSUB(To, Tp);
Tq = VADD(To, Tp);
TI = VSUB(Tr, Ts);
Tt = VADD(Tr, Ts);
T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
T7 = VSUB(T3, T6);
Tz = VADD(T3, T6);
Tf = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
TK = VADD(TI, TJ);
TV = VSUB(TJ, TI);
TA = VADD(Tt, Tq);
Tu = VSUB(Tq, Tt);
TM = VSUB(T8, T9);
Ta = VADD(T8, T9);
TN = VSUB(Tb, Tc);
Td = VADD(Tb, Tc);
Tg = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Ti = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
}
}
TF = VSUB(Tz, TA);
TB = VADD(Tz, TA);
T13 = VFNMS(LDK(KP707106781), TK, TH);
TL = VFMA(LDK(KP707106781), TK, TH);
TO = VFNMS(LDK(KP414213562), TN, TM);
TX = VFMA(LDK(KP414213562), TM, TN);
TC = VADD(Ta, Td);
Te = VSUB(Ta, Td);
TP = VSUB(Tf, Tg);
Th = VADD(Tf, Tg);
TQ = VSUB(Tj, Ti);
Tk = VADD(Ti, Tj);
TW = VFNMS(LDK(KP707106781), TV, TU);
T16 = VFMA(LDK(KP707106781), TV, TU);
}
{
V TY, TR, Tl, TD;
TY = VFMA(LDK(KP414213562), TP, TQ);
TR = VFNMS(LDK(KP414213562), TQ, TP);
Tl = VSUB(Th, Tk);
TD = VADD(Th, Tk);
{
V TS, T17, TZ, T14;
TS = VADD(TO, TR);
T17 = VSUB(TR, TO);
TZ = VSUB(TX, TY);
T14 = VADD(TX, TY);
{
V TE, TG, Tm, Tv;
TE = VADD(TC, TD);
TG = VSUB(TD, TC);
Tm = VADD(Te, Tl);
Tv = VSUB(Tl, Te);
{
V T18, T1a, TT, T11;
T18 = VFNMS(LDK(KP923879532), T17, T16);
T1a = VFMA(LDK(KP923879532), T17, T16);
TT = VFNMS(LDK(KP923879532), TS, TL);
T11 = VFMA(LDK(KP923879532), TS, TL);
{
V T15, T19, T10, T12;
T15 = VFNMS(LDK(KP923879532), T14, T13);
T19 = VFMA(LDK(KP923879532), T14, T13);
T10 = VFNMS(LDK(KP923879532), TZ, TW);
T12 = VFMA(LDK(KP923879532), TZ, TW);
{
V T1b, T1c, T1d, T1e;
T1b = VFMAI(TG, TF);
STM2(&(xo[8]), T1b, ovs, &(xo[0]));
T1c = VFNMSI(TG, TF);
STM2(&(xo[24]), T1c, ovs, &(xo[0]));
T1d = VADD(TB, TE);
STM2(&(xo[0]), T1d, ovs, &(xo[0]));
T1e = VSUB(TB, TE);
STM2(&(xo[16]), T1e, ovs, &(xo[0]));
{
V Tw, Ty, Tn, Tx;
Tw = VFNMS(LDK(KP707106781), Tv, Tu);
Ty = VFMA(LDK(KP707106781), Tv, Tu);
Tn = VFNMS(LDK(KP707106781), Tm, T7);
Tx = VFMA(LDK(KP707106781), Tm, T7);
{
V T1f, T1g, T1h, T1i;
T1f = VFMAI(T1a, T19);
STM2(&(xo[6]), T1f, ovs, &(xo[2]));
T1g = VFNMSI(T1a, T19);
STM2(&(xo[26]), T1g, ovs, &(xo[2]));
STN2(&(xo[24]), T1c, T1g, ovs);
T1h = VFMAI(T18, T15);
STM2(&(xo[22]), T1h, ovs, &(xo[2]));
T1i = VFNMSI(T18, T15);
STM2(&(xo[10]), T1i, ovs, &(xo[2]));
STN2(&(xo[8]), T1b, T1i, ovs);
{
V T1j, T1k, T1l, T1m;
T1j = VFNMSI(T12, T11);
STM2(&(xo[2]), T1j, ovs, &(xo[2]));
STN2(&(xo[0]), T1d, T1j, ovs);
T1k = VFMAI(T12, T11);
STM2(&(xo[30]), T1k, ovs, &(xo[2]));
T1l = VFMAI(T10, TT);
STM2(&(xo[14]), T1l, ovs, &(xo[2]));
T1m = VFNMSI(T10, TT);
STM2(&(xo[18]), T1m, ovs, &(xo[2]));
STN2(&(xo[16]), T1e, T1m, ovs);
{
V T1n, T1o, T1p, T1q;
T1n = VFNMSI(Ty, Tx);
STM2(&(xo[28]), T1n, ovs, &(xo[0]));
STN2(&(xo[28]), T1n, T1k, ovs);
T1o = VFMAI(Ty, Tx);
STM2(&(xo[4]), T1o, ovs, &(xo[0]));
STN2(&(xo[4]), T1o, T1f, ovs);
T1p = VFMAI(Tw, Tn);
STM2(&(xo[20]), T1p, ovs, &(xo[0]));
STN2(&(xo[20]), T1p, T1h, ovs);
T1q = VFNMSI(Tw, Tn);
STM2(&(xo[12]), T1q, ovs, &(xo[0]));
STN2(&(xo[12]), T1q, T1l, ovs);
}
}
}
}
}
}
}
}
}
}
}
}
VLEAVE();
}
static void t3fv_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) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs)) {
V T2, T3, Tb, T1, T5, Tn, Tq, T8, Td, T4, Ta, Tp, Tg, Ti, T9;
T2 = LDW(&(W[0]));
T3 = LDW(&(W[TWVL * 2]));
Tb = LDW(&(W[TWVL * 4]));
T1 = LD(&(x[0]), ms, &(x[0]));
T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
T4 = VZMUL(T2, T3);
Ta = VZMULJ(T2, T3);
Tp = VZMULJ(T2, Tb);
Tg = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Ti = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T9 = VZMULJ(T2, T8);
{
V T6, To, Tc, Tr, Th, Tj;
T6 = VZMULJ(T4, T5);
To = VZMULJ(Ta, Tn);
Tc = VZMULJ(Ta, Tb);
Tr = VZMULJ(Tp, Tq);
Th = VZMULJ(Tb, Tg);
Tj = VZMULJ(T3, Ti);
{
V Tx, T7, Te, Ts, Ty, Tk, TB;
Tx = VADD(T1, T6);
T7 = VSUB(T1, T6);
Te = VZMULJ(Tc, Td);
Ts = VSUB(To, Tr);
Ty = VADD(To, Tr);
Tk = VSUB(Th, Tj);
TB = VADD(Th, Tj);
{
V Tf, TA, Tz, TD;
Tf = VSUB(T9, Te);
TA = VADD(T9, Te);
Tz = VADD(Tx, Ty);
TD = VSUB(Tx, Ty);
{
V TC, TE, Tl, Tt;
TC = VADD(TA, TB);
TE = VSUB(TB, TA);
Tl = VADD(Tf, Tk);
Tt = VSUB(Tk, Tf);
{
V Tu, Tw, Tm, Tv;
ST(&(x[WS(rs, 2)]), VFMAI(TE, TD), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(TE, TD), ms, &(x[0]));
ST(&(x[0]), VADD(Tz, TC), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VSUB(Tz, TC), ms, &(x[0]));
Tu = VFNMS(LDK(KP707106781), Tt, Ts);
Tw = VFMA(LDK(KP707106781), Tt, Ts);
Tm = VFMA(LDK(KP707106781), Tl, T7);
Tv = VFNMS(LDK(KP707106781), Tl, T7);
ST(&(x[WS(rs, 5)]), VFNMSI(Tw, Tv), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFMAI(Tw, Tv), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(Tu, Tm), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFNMSI(Tu, Tm), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
}
static void t1buv_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(8, 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 = BYTW(&(W[TWVL * 6]), T2);
Ti = BYTW(&(W[TWVL * 2]), Th);
Tk = BYTW(&(W[TWVL * 10]), Tj);
T6 = BYTW(&(W[0]), T5);
T8 = BYTW(&(W[TWVL * 8]), T7);
Tb = BYTW(&(W[TWVL * 12]), Ta);
Td = BYTW(&(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 = VADD(Tq, Tr);
Ts = VSUB(Tq, Tr);
{
V Tx, Tv, Tm, Tf;
Tx = VADD(Tt, Tu);
Tv = VSUB(Tt, Tu);
Tm = VSUB(T9, Te);
Tf = VADD(T9, Te);
{
V Tp, Tn, To, Tg;
ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(Tv, Ts), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(Tv, Ts), ms, &(x[0]));
Tp = VFMA(LDK(KP707106781), Tm, Tl);
Tn = VFNMS(LDK(KP707106781), Tm, Tl);
To = VFMA(LDK(KP707106781), Tf, T4);
Tg = VFNMS(LDK(KP707106781), Tf, T4);
ST(&(x[WS(rs, 1)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 5)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
VLEAVE();
}
static void t3fv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
{
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(10, rs)) {
V T2, T3, T4, Ta, T5, T6, Tt, Td, Th;
T2 = LDW(&(W[0]));
T3 = LDW(&(W[TWVL * 2]));
T4 = VZMUL(T2, T3);
Ta = VZMULJ(T2, T3);
T5 = LDW(&(W[TWVL * 4]));
T6 = VZMULJ(T4, T5);
Tt = VZMULJ(T3, T5);
Td = VZMULJ(Ta, T5);
Th = VZMULJ(T2, T5);
{
V T9, TJ, Ts, Ty, Tz, TN, TO, TP, Tg, Tm, Tn, TK, TL, TM, T1;
V T8, T7;
T1 = LD(&(x[0]), ms, &(x[0]));
T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
T8 = VZMULJ(T6, T7);
T9 = VSUB(T1, T8);
TJ = VADD(T1, T8);
{
V Tp, Tx, Tr, Tv;
{
V To, Tw, Tq, Tu;
To = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tp = VZMULJ(T4, To);
Tw = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Tx = VZMULJ(T2, Tw);
Tq = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
Tr = VZMULJ(T5, Tq);
Tu = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tv = VZMULJ(Tt, Tu);
}
Ts = VSUB(Tp, Tr);
Ty = VSUB(Tv, Tx);
Tz = VADD(Ts, Ty);
TN = VADD(Tp, Tr);
TO = VADD(Tv, Tx);
TP = VADD(TN, TO);
}
{
V Tc, Tl, Tf, Tj;
{
V Tb, Tk, Te, Ti;
Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tc = VZMULJ(Ta, Tb);
Tk = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Tl = VZMULJ(T3, Tk);
Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Tf = VZMULJ(Td, Te);
Ti = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Tj = VZMULJ(Th, Ti);
}
Tg = VSUB(Tc, Tf);
Tm = VSUB(Tj, Tl);
Tn = VADD(Tg, Tm);
TK = VADD(Tc, Tf);
TL = VADD(Tj, Tl);
TM = VADD(TK, TL);
}
{
V TC, TA, TB, TG, TI, TE, TF, TH, TD;
TC = VSUB(Tn, Tz);
TA = VADD(Tn, Tz);
TB = VFNMS(LDK(KP250000000), TA, T9);
TE = VSUB(Tg, Tm);
TF = VSUB(Ts, Ty);
TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TF, TE));
TI = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TE, TF));
ST(&(x[WS(rs, 5)]), VADD(T9, TA), ms, &(x[WS(rs, 1)]));
TH = VFNMS(LDK(KP559016994), TC, TB);
ST(&(x[WS(rs, 3)]), VFNMSI(TI, TH), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(TI, TH), ms, &(x[WS(rs, 1)]));
TD = VFMA(LDK(KP559016994), TC, TB);
ST(&(x[WS(rs, 1)]), VFNMSI(TG, TD), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 9)]), VFMAI(TG, TD), ms, &(x[WS(rs, 1)]));
}
{
V TS, TQ, TR, TW, TY, TU, TV, TX, TT;
TS = VSUB(TM, TP);
TQ = VADD(TM, TP);
TR = VFNMS(LDK(KP250000000), TQ, TJ);
TU = VSUB(TN, TO);
TV = VSUB(TK, TL);
TW = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TV, TU));
TY = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TU, TV));
ST(&(x[0]), VADD(TJ, TQ), ms, &(x[0]));
TX = VFMA(LDK(KP559016994), TS, TR);
ST(&(x[WS(rs, 4)]), VFMAI(TY, TX), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(TY, TX), ms, &(x[0]));
TT = VFNMS(LDK(KP559016994), TS, TR);
ST(&(x[WS(rs, 2)]), VFMAI(TW, TT), ms, &(x[0]));
ST(&(x[WS(rs, 8)]), VFNMSI(TW, TT), ms, &(x[0]));
}
}
}
}
VLEAVE();
}
static void n1bv_15(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP823639103, +0.823639103546331925877420039278190003029660514);
DVK(KP910592997, +0.910592997310029334643087372129977886038870291);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
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 Tb, TH, Tw, TA, Th, T11, T5, Ti, T12, Ta, Tx, Te, Tq, T16, Tj;
V T1, T2, T3;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
{
V T6, T7, T8, Tm, Tn, To;
T6 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tn = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
To = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
{
V T4, Tc, T9, Td, Tp;
Tb = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T4 = VADD(T2, T3);
TH = VSUB(T2, T3);
Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Tw = VSUB(T7, T8);
T9 = VADD(T7, T8);
Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Tp = VADD(Tn, To);
TA = VSUB(Tn, To);
Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T11 = VADD(T1, T4);
T5 = VFNMS(LDK(KP500000000), T4, T1);
Ti = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
T12 = VADD(T6, T9);
Ta = VFNMS(LDK(KP500000000), T9, T6);
Tx = VSUB(Tc, Td);
Te = VADD(Tc, Td);
Tq = VFNMS(LDK(KP500000000), Tp, Tm);
T16 = VADD(Tm, Tp);
Tj = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
}
}
{
V TI, Ty, T13, Tf, Tz, Tk;
TI = VADD(Tw, Tx);
Ty = VSUB(Tw, Tx);
T13 = VADD(Tb, Te);
Tf = VFNMS(LDK(KP500000000), Te, Tb);
Tz = VSUB(Ti, Tj);
Tk = VADD(Ti, Tj);
{
V T1d, T14, Tg, TE, TJ, TB, T15, Tl;
T1d = VSUB(T12, T13);
T14 = VADD(T12, T13);
Tg = VADD(Ta, Tf);
TE = VSUB(Ta, Tf);
TJ = VADD(Tz, TA);
TB = VSUB(Tz, TA);
T15 = VADD(Th, Tk);
Tl = VFNMS(LDK(KP500000000), Tk, Th);
{
V TM, TK, TS, TC, T1c, T17, Tr, TF, TL, T10;
TM = VSUB(TI, TJ);
TK = VADD(TI, TJ);
TS = VFNMS(LDK(KP618033988), Ty, TB);
TC = VFMA(LDK(KP618033988), TB, Ty);
T1c = VSUB(T15, T16);
T17 = VADD(T15, T16);
Tr = VADD(Tl, Tq);
TF = VSUB(Tl, Tq);
TL = VFNMS(LDK(KP250000000), TK, TH);
T10 = VMUL(LDK(KP866025403), VADD(TH, TK));
{
V T1g, T1e, T1a, Tu, Ts, TU, TG, TV, TN, T19, T18, Tt, TZ;
T1g = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1c, T1d));
T1e = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1d, T1c));
T18 = VADD(T14, T17);
T1a = VSUB(T14, T17);
Tu = VSUB(Tg, Tr);
Ts = VADD(Tg, Tr);
TU = VFNMS(LDK(KP618033988), TE, TF);
TG = VFMA(LDK(KP618033988), TF, TE);
TV = VFNMS(LDK(KP559016994), TM, TL);
TN = VFMA(LDK(KP559016994), TM, TL);
ST(&(xo[0]), VADD(T11, T18), ovs, &(xo[0]));
T19 = VFNMS(LDK(KP250000000), T18, T11);
Tt = VFNMS(LDK(KP250000000), Ts, T5);
TZ = VADD(T5, Ts);
{
V TW, TY, TQ, TO, T1b, T1f, TR, Tv;
TW = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), TV, TU));
TY = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), TV, TU));
TQ = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), TN, TG));
TO = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), TN, TG));
T1b = VFNMS(LDK(KP559016994), T1a, T19);
T1f = VFMA(LDK(KP559016994), T1a, T19);
TR = VFNMS(LDK(KP559016994), Tu, Tt);
Tv = VFMA(LDK(KP559016994), Tu, Tt);
ST(&(xo[WS(os, 10)]), VFMAI(T10, TZ), ovs, &(xo[0]));
ST(&(xo[WS(os, 5)]), VFNMSI(T10, TZ), ovs, &(xo[WS(os, 1)]));
{
V TT, TX, TP, TD;
ST(&(xo[WS(os, 12)]), VFNMSI(T1e, T1b), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(T1e, T1b), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VFNMSI(T1g, T1f), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 6)]), VFMAI(T1g, T1f), ovs, &(xo[0]));
TT = VFNMS(LDK(KP823639103), TS, TR);
TX = VFMA(LDK(KP823639103), TS, TR);
TP = VFMA(LDK(KP823639103), TC, Tv);
TD = VFNMS(LDK(KP823639103), TC, Tv);
ST(&(xo[WS(os, 13)]), VFMAI(TW, TT), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VFNMSI(TW, TT), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VFMAI(TY, TX), ovs, &(xo[0]));
ST(&(xo[WS(os, 7)]), VFNMSI(TY, TX), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 11)]), VFMAI(TQ, TP), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VFNMSI(TQ, TP), ovs, &(xo[0]));
ST(&(xo[WS(os, 14)]), VFNMSI(TO, TD), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFMAI(TO, TD), ovs, &(xo[WS(os, 1)]));
}
}
}
}
}
}
}
}
VLEAVE();
}
static void t1fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
{
INT m;
R *x;
x = ri;
for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) {
V T4, Tx, T1m, T1K, T1y, Tk, Tf, T16, T10, TT, T1O, T1w, T1L, T1p, T1M;
V T1s, TZ, TI, T1x, Tp;
{
V T1, Tv, T2, Tt;
T1 = LD(&(x[0]), ms, &(x[0]));
Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
{
V T9, T1n, TN, T1v, TS, Te, T1q, T1u, TE, TG, Tm, T1o, TC, Tn, T1r;
V TH, To;
{
V TP, TR, Ta, Tc;
{
V T5, T7, TJ, TL, T1k, T1l;
T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
{
V Tw, T3, Tu, T6, T8, TK, TM, TO, TQ;
TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)]));
Tw = BYTWJ(&(W[TWVL * 28]), Tv);
T3 = BYTWJ(&(W[TWVL * 18]), T2);
Tu = BYTWJ(&(W[TWVL * 8]), Tt);
T6 = BYTWJ(&(W[TWVL * 6]), T5);
T8 = BYTWJ(&(W[TWVL * 26]), T7);
TK = BYTWJ(&(W[TWVL * 24]), TJ);
TM = BYTWJ(&(W[TWVL * 4]), TL);
TP = BYTWJ(&(W[TWVL * 32]), TO);
TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T4 = VSUB(T1, T3);
T1k = VADD(T1, T3);
Tx = VSUB(Tu, Tw);
T1l = VADD(Tu, Tw);
T9 = VSUB(T6, T8);
T1n = VADD(T6, T8);
TN = VSUB(TK, TM);
T1v = VADD(TK, TM);
TR = BYTWJ(&(W[TWVL * 12]), TQ);
}
Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0]));
T1m = VSUB(T1k, T1l);
T1K = VADD(T1k, T1l);
Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
}
{
V Tb, TA, Td, Th, Tj, Tz, Tg, Ti, Ty;
Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0]));
Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
TS = VSUB(TP, TR);
T1y = VADD(TP, TR);
Tb = BYTWJ(&(W[TWVL * 30]), Ta);
TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)]));
Td = BYTWJ(&(W[TWVL * 10]), Tc);
Th = BYTWJ(&(W[TWVL * 14]), Tg);
Tj = BYTWJ(&(W[TWVL * 34]), Ti);
Tz = BYTWJ(&(W[TWVL * 16]), Ty);
{
V TD, TF, TB, Tl;
TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
TB = BYTWJ(&(W[TWVL * 36]), TA);
Te = VSUB(Tb, Td);
T1q = VADD(Tb, Td);
Tk = VSUB(Th, Tj);
T1u = VADD(Th, Tj);
TE = BYTWJ(&(W[0]), TD);
TG = BYTWJ(&(W[TWVL * 20]), TF);
Tm = BYTWJ(&(W[TWVL * 22]), Tl);
T1o = VADD(Tz, TB);
TC = VSUB(Tz, TB);
Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
}
}
}
Tf = VADD(T9, Te);
T16 = VSUB(T9, Te);
T10 = VSUB(TS, TN);
TT = VADD(TN, TS);
T1r = VADD(TE, TG);
TH = VSUB(TE, TG);
T1O = VADD(T1u, T1v);
T1w = VSUB(T1u, T1v);
To = BYTWJ(&(W[TWVL * 2]), Tn);
T1L = VADD(T1n, T1o);
T1p = VSUB(T1n, T1o);
T1M = VADD(T1q, T1r);
T1s = VSUB(T1q, T1r);
TZ = VSUB(TH, TC);
TI = VADD(TC, TH);
T1x = VADD(Tm, To);
Tp = VSUB(Tm, To);
}
}
{
V T1V, T1N, T14, T1d, T11, T1G, T1t, T1z, T1P, Tq, T17, T13, TV, TU;
T1V = VSUB(T1L, T1M);
T1N = VADD(T1L, T1M);
T14 = VSUB(TT, TI);
TU = VADD(TI, TT);
T1d = VFNMS(LDK(KP618033988), TZ, T10);
T11 = VFMA(LDK(KP618033988), T10, TZ);
T1G = VSUB(T1p, T1s);
T1t = VADD(T1p, T1s);
T1z = VSUB(T1x, T1y);
T1P = VADD(T1x, T1y);
Tq = VADD(Tk, Tp);
T17 = VSUB(Tk, Tp);
T13 = VFNMS(LDK(KP250000000), TU, Tx);
TV = VADD(Tx, TU);
{
V T1J, T1H, T1D, T1Z, T1X, T1T, T1h, T1j, T1b, T19, T1C, T1S, T1c, TY, T1F;
V T1A;
T1F = VSUB(T1w, T1z);
T1A = VADD(T1w, T1z);
{
V T1W, T1Q, TX, Tr;
T1W = VSUB(T1O, T1P);
T1Q = VADD(T1O, T1P);
TX = VSUB(Tf, Tq);
Tr = VADD(Tf, Tq);
{
V T1g, T18, T1f, T15;
T1g = VFNMS(LDK(KP618033988), T16, T17);
T18 = VFMA(LDK(KP618033988), T17, T16);
T1f = VFMA(LDK(KP559016994), T14, T13);
T15 = VFNMS(LDK(KP559016994), T14, T13);
T1J = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1F, T1G));
T1H = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1G, T1F));
{
V T1B, T1R, TW, Ts;
T1B = VADD(T1t, T1A);
T1D = VSUB(T1t, T1A);
T1Z = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1V, T1W));
T1X = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1W, T1V));
T1R = VADD(T1N, T1Q);
T1T = VSUB(T1N, T1Q);
TW = VFNMS(LDK(KP250000000), Tr, T4);
Ts = VADD(T4, Tr);
T1h = VFNMS(LDK(KP951056516), T1g, T1f);
T1j = VFMA(LDK(KP951056516), T1g, T1f);
T1b = VFNMS(LDK(KP951056516), T18, T15);
T19 = VFMA(LDK(KP951056516), T18, T15);
ST(&(x[WS(rs, 10)]), VADD(T1m, T1B), ms, &(x[0]));
T1C = VFNMS(LDK(KP250000000), T1B, T1m);
ST(&(x[0]), VADD(T1K, T1R), ms, &(x[0]));
T1S = VFNMS(LDK(KP250000000), T1R, T1K);
T1c = VFNMS(LDK(KP559016994), TX, TW);
TY = VFMA(LDK(KP559016994), TX, TW);
ST(&(x[WS(rs, 15)]), VFMAI(TV, Ts), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 5)]), VFNMSI(TV, Ts), ms, &(x[WS(rs, 1)]));
}
}
}
{
V T1E, T1I, T1U, T1Y;
T1E = VFNMS(LDK(KP559016994), T1D, T1C);
T1I = VFMA(LDK(KP559016994), T1D, T1C);
T1U = VFMA(LDK(KP559016994), T1T, T1S);
T1Y = VFNMS(LDK(KP559016994), T1T, T1S);
{
V T1e, T1i, T1a, T12;
T1e = VFNMS(LDK(KP951056516), T1d, T1c);
T1i = VFMA(LDK(KP951056516), T1d, T1c);
T1a = VFNMS(LDK(KP951056516), T11, TY);
T12 = VFMA(LDK(KP951056516), T11, TY);
ST(&(x[WS(rs, 18)]), VFNMSI(T1H, T1E), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(T1H, T1E), ms, &(x[0]));
ST(&(x[WS(rs, 14)]), VFMAI(T1J, T1I), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(T1J, T1I), ms, &(x[0]));
ST(&(x[WS(rs, 16)]), VFNMSI(T1X, T1U), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VFMAI(T1X, T1U), ms, &(x[0]));
ST(&(x[WS(rs, 12)]), VFMAI(T1Z, T1Y), ms, &(x[0]));
ST(&(x[WS(rs, 8)]), VFNMSI(T1Z, T1Y), ms, &(x[0]));
ST(&(x[WS(rs, 3)]), VFMAI(T1h, T1e), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 17)]), VFNMSI(T1h, T1e), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(T1j, T1i), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 13)]), VFNMSI(T1j, T1i), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 11)]), VFMAI(T1b, T1a), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 9)]), VFNMSI(T1b, T1a), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 19)]), VFMAI(T19, T12), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFNMSI(T19, T12), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
VLEAVE();
}
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(32, 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]));
}
}
VLEAVE();
}
