static void t1bv_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 = ii;
for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(12, rs)) {
V TI, Ti, TA, T7, Tm, TE, Tw, Tk, Tf, TB, TU, TM;
{
V T9, TK, Tj, TL, Te;
{
V T1, T4, T2, Tp, Tt, Tr;
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]));
Tp = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
{
V T5, T3, Tq, Tu, Ts, 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 = BYTW(&(W[TWVL * 14]), T4);
T3 = BYTW(&(W[TWVL * 6]), T2);
Tq = BYTW(&(W[TWVL * 16]), Tp);
Tu = BYTW(&(W[TWVL * 8]), Tt);
Ts = BYTW(&(W[0]), Tr);
T9 = BYTW(&(W[TWVL * 10]), T8);
Td = BYTW(&(W[TWVL * 2]), Tc);
Tb = BYTW(&(W[TWVL * 18]), Ta);
{
V Th, T6, Tl, Tv;
Th = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
TK = VSUB(T3, T5);
T6 = VADD(T3, T5);
Tl = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
Tv = VADD(Ts, Tu);
TI = VSUB(Tu, Ts);
Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
TL = VSUB(Tb, Td);
Te = VADD(Tb, Td);
Ti = BYTW(&(W[TWVL * 4]), Th);
TA = VFNMS(LDK(KP500000000), T6, T1);
T7 = VADD(T1, T6);
Tm = BYTW(&(W[TWVL * 20]), Tl);
TE = VFNMS(LDK(KP500000000), Tv, Tq);
Tw = VADD(Tq, Tv);
}
}
}
Tk = BYTW(&(W[TWVL * 12]), Tj);
Tf = VADD(T9, Te);
TB = VFNMS(LDK(KP500000000), Te, T9);
TU = VSUB(TK, TL);
TM = VADD(TK, TL);
}
{
V Tn, TH, TC, TQ, Ty, Tg;
Tn = VADD(Tk, Tm);
TH = VSUB(Tk, Tm);
TC = VADD(TA, TB);
TQ = VSUB(TA, TB);
Ty = VADD(T7, Tf);
Tg = VSUB(T7, Tf);
{
V To, TD, TJ, TR;
To = VADD(Ti, Tn);
TD = VFNMS(LDK(KP500000000), Tn, Ti);
TJ = VSUB(TH, TI);
TR = VADD(TH, TI);
{
V TP, TN, TW, TS, TO, TG, TX, TV;
{
V Tz, Tx, TF, TT;
Tz = VADD(To, Tw);
Tx = VSUB(To, Tw);
TF = VADD(TD, TE);
TT = VSUB(TD, TE);
TP = VMUL(LDK(KP866025403), VADD(TM, TJ));
TN = VMUL(LDK(KP866025403), VSUB(TJ, TM));
TW = VFMA(LDK(KP866025403), TR, TQ);
TS = VFNMS(LDK(KP866025403), TR, TQ);
ST(&(x[WS(rs, 6)]), VSUB(Ty, Tz), ms, &(x[0]));
ST(&(x[0]), VADD(Ty, Tz), ms, &(x[0]));
ST(&(x[WS(rs, 9)]), VFMAI(Tx, Tg), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFNMSI(Tx, Tg), ms, &(x[WS(rs, 1)]));
TO = VADD(TC, TF);
TG = VSUB(TC, TF);
TX = VFNMS(LDK(KP866025403), TU, TT);
TV = VFMA(LDK(KP866025403), TU, TT);
}
ST(&(x[WS(rs, 8)]), VFNMSI(TP, TO), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(TN, TG), ms, &(x[0]));
ST(&(x[WS(rs, 10)]), VFNMSI(TN, TG), ms, &(x[0]));
ST(&(x[WS(rs, 5)]), VFMAI(TX, TW), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFNMSI(TX, TW), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 11)]), VFNMSI(TV, TS), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFMAI(TV, TS), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
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(KP342020143, +0.342020143325668733044099614682259580763083368);
DVK(KP813797681, +0.813797681349373692844693217248393223289101568);
DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
DVK(KP296198132, +0.296198132726023843175338011893050938967728390);
DVK(KP642787609, +0.642787609686539326322643409907263432907559884);
DVK(KP663413948, +0.663413948168938396205421319635891297216863310);
DVK(KP556670399, +0.556670399226419366452912952047023132968291906);
DVK(KP766044443, +0.766044443118978035202392650555416673935832457);
DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
DVK(KP150383733, +0.150383733180435296639271897612501926072238258);
DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
DVK(KP173648177, +0.173648177666930348851716626769314796000375677);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
{
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(18, is), MAKE_VOLATILE_STRIDE(18, os)) {
V T5, Ts, Tj, To, Tf, Tn, Tp, Tu, Tl, Ta, Tk, Tm, Tt;
{
V T1, T2, T3, T4;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T4 = VADD(T2, T3);
T5 = VADD(T1, T4);
Ts = VMUL(LDK(KP866025403), VSUB(T3, T2));
Tj = VFNMS(LDK(KP500000000), T4, T1);
}
{
V Tb, Te, Tc, Td;
Tb = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Te = VADD(Tc, Td);
To = VSUB(Td, Tc);
Tf = VADD(Tb, Te);
Tn = VFNMS(LDK(KP500000000), Te, Tb);
Tp = VFMA(LDK(KP173648177), Tn, VMUL(LDK(KP852868531), To));
Tu = VFNMS(LDK(KP984807753), Tn, VMUL(LDK(KP150383733), To));
}
{
V T6, T9, T7, T8;
T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T9 = VADD(T7, T8);
Tl = VSUB(T8, T7);
Ta = VADD(T6, T9);
Tk = VFNMS(LDK(KP500000000), T9, T6);
Tm = VFMA(LDK(KP766044443), Tk, VMUL(LDK(KP556670399), Tl));
Tt = VFNMS(LDK(KP642787609), Tk, VMUL(LDK(KP663413948), Tl));
}
{
V Ti, Tg, Th, Tz, TA;
Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Tf, Ta)));
Tg = VADD(Ta, Tf);
Th = VFNMS(LDK(KP500000000), Tg, T5);
ST(&(xo[0]), VADD(T5, Tg), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VADD(Th, Ti), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 6)]), VSUB(Th, Ti), ovs, &(xo[0]));
Tz = VFMA(LDK(KP173648177), Tk, VFNMS(LDK(KP296198132), To, VFNMS(LDK(KP939692620), Tn, VFNMS(LDK(KP852868531), Tl, Tj))));
TA = VBYI(VSUB(VFNMS(LDK(KP342020143), Tn, VFNMS(LDK(KP150383733), Tl, VFNMS(LDK(KP984807753), Tk, VMUL(LDK(KP813797681), To)))), Ts));
ST(&(xo[WS(os, 7)]), VSUB(Tz, TA), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VADD(Tz, TA), ovs, &(xo[0]));
{
V Tr, Tx, Tw, Ty, Tq, Tv;
Tq = VADD(Tm, Tp);
Tr = VADD(Tj, Tq);
Tx = VFMA(LDK(KP866025403), VSUB(Tt, Tu), VFNMS(LDK(KP500000000), Tq, Tj));
Tv = VADD(Tt, Tu);
Tw = VBYI(VADD(Ts, Tv));
Ty = VBYI(VADD(Ts, VFNMS(LDK(KP500000000), Tv, VMUL(LDK(KP866025403), VSUB(Tp, Tm)))));
ST(&(xo[WS(os, 8)]), VSUB(Tr, Tw), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VADD(Tx, Ty), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VADD(Tw, Tr), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 5)]), VSUB(Tx, Ty), ovs, &(xo[WS(os, 1)]));
}
}
}
}
VLEAVE();
}
static void n1bv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
DVK(KP959492973, +0.959492973614497389890368057066327699062454848);
DVK(KP654860733, +0.654860733945285064056925072466293553183791199);
DVK(KP142314838, +0.142314838273285140443792668616369668791051361);
DVK(KP415415013, +0.415415013001886425529274149229623203524004910);
DVK(KP841253532, +0.841253532831181168861811648919367717513292498);
DVK(KP540640817, +0.540640817455597582107635954318691695431770608);
DVK(KP909631995, +0.909631995354518371411715383079028460060241051);
DVK(KP989821441, +0.989821441880932732376092037776718787376519372);
DVK(KP755749574, +0.755749574354258283774035843972344420179717445);
DVK(KP281732556, +0.281732556841429697711417915346616899035777899);
int i;
const R *xi;
R *xo;
xi = ii;
xo = io;
BEGIN_SIMD();
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
V Th, T3, Tm, Tf, Ti, Tc, Tj, T9, Tk, T6, Tl, Ta, Tb, Ts, Tt;
Th = LD(&(xi[0]), ivs, &(xi[0]));
{
V T1, T2, Td, Te;
T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T3 = VSUB(T1, T2);
Tm = VADD(T1, T2);
Td = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tf = VSUB(Td, Te);
Ti = VADD(Td, Te);
}
Ta = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tb = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Tc = VSUB(Ta, Tb);
Tj = VADD(Ta, Tb);
{
V T7, T8, T4, T5;
T7 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T9 = VSUB(T7, T8);
Tk = VADD(T7, T8);
T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T5 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T6 = VSUB(T4, T5);
Tl = VADD(T4, T5);
}
ST(&(xo[0]), VADD(Th, VADD(Tm, VADD(Ti, VADD(Tl, VADD(Tj, Tk))))), ovs, &(xo[0]));
{
V Tg, Tn, Tu, Tv;
Tg = VBYI(VFMA(LDK(KP281732556), T3, VFMA(LDK(KP755749574), T6, VFNMS(LDK(KP909631995), Tc, VFNMS(LDK(KP540640817), Tf, VMUL(LDK(KP989821441), T9))))));
Tn = VFMA(LDK(KP841253532), Ti, VFMA(LDK(KP415415013), Tj, VFNMS(LDK(KP142314838), Tk, VFNMS(LDK(KP654860733), Tl, VFNMS(LDK(KP959492973), Tm, Th)))));
ST(&(xo[WS(os, 5)]), VADD(Tg, Tn), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 6)]), VSUB(Tn, Tg), ovs, &(xo[0]));
Tu = VBYI(VFMA(LDK(KP755749574), T3, VFMA(LDK(KP540640817), T6, VFNMS(LDK(KP909631995), T9, VFNMS(LDK(KP989821441), Tf, VMUL(LDK(KP281732556), Tc))))));
Tv = VFMA(LDK(KP841253532), Tl, VFMA(LDK(KP415415013), Tk, VFNMS(LDK(KP959492973), Tj, VFNMS(LDK(KP142314838), Ti, VFNMS(LDK(KP654860733), Tm, Th)))));
ST(&(xo[WS(os, 4)]), VADD(Tu, Tv), ovs, &(xo[0]));
ST(&(xo[WS(os, 7)]), VSUB(Tv, Tu), ovs, &(xo[WS(os, 1)]));
}
Ts = VBYI(VFMA(LDK(KP909631995), T3, VFNMS(LDK(KP540640817), T9, VFNMS(LDK(KP989821441), Tc, VFNMS(LDK(KP281732556), T6, VMUL(LDK(KP755749574), Tf))))));
Tt = VFMA(LDK(KP415415013), Tm, VFMA(LDK(KP841253532), Tk, VFNMS(LDK(KP142314838), Tj, VFNMS(LDK(KP959492973), Tl, VFNMS(LDK(KP654860733), Ti, Th)))));
ST(&(xo[WS(os, 2)]), VADD(Ts, Tt), ovs, &(xo[0]));
ST(&(xo[WS(os, 9)]), VSUB(Tt, Ts), ovs, &(xo[WS(os, 1)]));
{
V Tq, Tr, To, Tp;
Tq = VBYI(VFMA(LDK(KP540640817), T3, VFMA(LDK(KP909631995), Tf, VFMA(LDK(KP989821441), T6, VFMA(LDK(KP755749574), Tc, VMUL(LDK(KP281732556), T9))))));
Tr = VFMA(LDK(KP841253532), Tm, VFMA(LDK(KP415415013), Ti, VFNMS(LDK(KP959492973), Tk, VFNMS(LDK(KP654860733), Tj, VFNMS(LDK(KP142314838), Tl, Th)))));
ST(&(xo[WS(os, 1)]), VADD(Tq, Tr), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 10)]), VSUB(Tr, Tq), ovs, &(xo[0]));
To = VBYI(VFMA(LDK(KP989821441), T3, VFMA(LDK(KP540640817), Tc, VFNMS(LDK(KP909631995), T6, VFNMS(LDK(KP281732556), Tf, VMUL(LDK(KP755749574), T9))))));
Tp = VFMA(LDK(KP415415013), Tl, VFMA(LDK(KP841253532), Tj, VFNMS(LDK(KP654860733), Tk, VFNMS(LDK(KP959492973), Ti, VFNMS(LDK(KP142314838), Tm, Th)))));
ST(&(xo[WS(os, 3)]), VADD(To, Tp), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 8)]), VSUB(Tp, To), ovs, &(xo[0]));
}
}
END_SIMD();
}
static void n2bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
DVK(KP222520933, +0.222520933956314404288902564496794759466355569);
DVK(KP623489801, +0.623489801858733530525004884004239810632274731);
DVK(KP781831482, +0.781831482468029808708444526674057750232334519);
DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
DVK(KP433883739, +0.433883739117558120475768332848358754609990728);
int i;
const R *xi;
R *xo;
xi = ii;
xo = io;
BEGIN_SIMD();
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
V Tp, Ty, Tl, TL, Tq, TE, T7, TJ, Ts, TB, Te, TK, Tr, TH, Tn;
V To;
Tn = LD(&(xi[0]), ivs, &(xi[0]));
To = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Tp = VSUB(Tn, To);
Ty = VADD(Tn, To);
{
V Th, TC, Tk, TD;
{
V Tf, Tg, Ti, Tj;
Tf = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Th = VSUB(Tf, Tg);
TC = VADD(Tf, Tg);
Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tj = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Tk = VSUB(Ti, Tj);
TD = VADD(Ti, Tj);
}
Tl = VSUB(Th, Tk);
TL = VSUB(TD, TC);
Tq = VADD(Th, Tk);
TE = VADD(TC, TD);
}
{
V T3, Tz, T6, TA;
{
V T1, T2, T4, T5;
T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
Tz = VADD(T1, T2);
T4 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T6 = VSUB(T4, T5);
TA = VADD(T4, T5);
}
T7 = VSUB(T3, T6);
TJ = VSUB(Tz, TA);
Ts = VADD(T3, T6);
TB = VADD(Tz, TA);
}
{
V Ta, TF, Td, TG;
{
V T8, T9, Tb, Tc;
T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T9 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
Ta = VSUB(T8, T9);
TF = VADD(T8, T9);
Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Td = VSUB(Tb, Tc);
TG = VADD(Tb, Tc);
}
Te = VSUB(Ta, Td);
TK = VSUB(TG, TF);
Tr = VADD(Ta, Td);
TH = VADD(TF, TG);
}
ST(&(xo[14]), VADD(Tp, VADD(Ts, VADD(Tq, Tr))), ovs, &(xo[2]));
ST(&(xo[0]), VADD(Ty, VADD(TB, VADD(TE, TH))), ovs, &(xo[0]));
{
V Tm, Tt, TQ, TP;
Tm = VBYI(VFMA(LDK(KP433883739), T7, VFNMS(LDK(KP781831482), Tl, VMUL(LDK(KP974927912), Te))));
Tt = VFMA(LDK(KP623489801), Tq, VFNMS(LDK(KP222520933), Tr, VFNMS(LDK(KP900968867), Ts, Tp)));
ST(&(xo[6]), VADD(Tm, Tt), ovs, &(xo[2]));
ST(&(xo[22]), VSUB(Tt, Tm), ovs, &(xo[2]));
TQ = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK))));
TP = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TB, Ty)));
ST(&(xo[24]), VSUB(TP, TQ), ovs, &(xo[0]));
ST(&(xo[4]), VADD(TP, TQ), ovs, &(xo[0]));
}
{
V Tu, Tv, TM, TI;
Tu = VBYI(VFMA(LDK(KP781831482), T7, VFMA(LDK(KP974927912), Tl, VMUL(LDK(KP433883739), Te))));
Tv = VFMA(LDK(KP623489801), Ts, VFNMS(LDK(KP900968867), Tr, VFNMS(LDK(KP222520933), Tq, Tp)));
ST(&(xo[2]), VADD(Tu, Tv), ovs, &(xo[2]));
ST(&(xo[26]), VSUB(Tv, Tu), ovs, &(xo[2]));
TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ))));
TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TE, Ty)));
ST(&(xo[12]), VSUB(TI, TM), ovs, &(xo[0]));
ST(&(xo[16]), VADD(TI, TM), ovs, &(xo[0]));
}
{
V TO, TN, Tx, Tw;
TO = VBYI(VFMA(LDK(KP433883739), TJ, VFNMS(LDK(KP974927912), TK, VMUL(LDK(KP781831482), TL))));
TN = VFMA(LDK(KP623489801), TE, VFNMS(LDK(KP222520933), TH, VFNMS(LDK(KP900968867), TB, Ty)));
ST(&(xo[8]), VSUB(TN, TO), ovs, &(xo[0]));
ST(&(xo[20]), VADD(TN, TO), ovs, &(xo[0]));
Tx = VBYI(VFNMS(LDK(KP781831482), Te, VFNMS(LDK(KP433883739), Tl, VMUL(LDK(KP974927912), T7))));
Tw = VFMA(LDK(KP623489801), Tr, VFNMS(LDK(KP900968867), Tq, VFNMS(LDK(KP222520933), Ts, Tp)));
ST(&(xo[10]), VSUB(Tw, Tx), ovs, &(xo[2]));
ST(&(xo[18]), VADD(Tx, Tw), ovs, &(xo[2]));
}
}
END_SIMD();
}
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(14, is), MAKE_VOLATILE_STRIDE(14, os)) {
V T1, T2, T3, T8, T9, T5, T6;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
{
V Tg, T4, Te, Ta, Tf, T7;
Tg = VSUB(T2, T3);
T4 = VADD(T2, T3);
Te = VSUB(T8, T9);
Ta = VADD(T8, T9);
Tf = VSUB(T5, T6);
T7 = VADD(T5, T6);
{
V Tr, Tj, Tm, Th, To, Tb;
Tr = VFMA(LDK(KP554958132), Te, Tg);
Tj = VFNMS(LDK(KP356895867), T4, Ta);
Tm = VFMA(LDK(KP554958132), Tf, Te);
Th = VFNMS(LDK(KP554958132), Tg, Tf);
ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, Ta))), ovs, &(xo[0]));
To = VFNMS(LDK(KP356895867), T7, T4);
Tb = VFNMS(LDK(KP356895867), Ta, T7);
{
V Ts, Tk, Tn, Ti;
Ts = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tr, Tf));
Tk = VFNMS(LDK(KP692021471), Tj, T7);
Tn = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tm, Tg));
Ti = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Th, Te));
{
V Tp, Tc, Tl, Tq, Td;
Tp = VFNMS(LDK(KP692021471), To, Ta);
Tc = VFNMS(LDK(KP692021471), Tb, T4);
Tl = VFNMS(LDK(KP900968867), Tk, T1);
Tq = VFNMS(LDK(KP900968867), Tp, T1);
Td = VFNMS(LDK(KP900968867), Tc, T1);
ST(&(xo[WS(os, 5)]), VFNMSI(Tn, Tl), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VFMAI(Tn, Tl), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VFNMSI(Ts, Tq), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFMAI(Ts, Tq), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VFNMSI(Ti, Td), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(Ti, Td), ovs, &(xo[WS(os, 1)]));
}
}
}
}
}
}
VLEAVE();
}
static void m2bv_64_0(const R *xi, R *xo, stride is, int ivs, int ovs)
{
DVK(KP634393284, +0.634393284163645498215171613225493370675687095);
DVK(KP773010453, +0.773010453362736960810906609758469800971041293);
DVK(KP290284677, +0.290284677254462367636192375817395274691476278);
DVK(KP956940335, +0.956940335732208864935797886980269969482849206);
DVK(KP098017140, +0.098017140329560601994195563888641845861136673);
DVK(KP995184726, +0.995184726672196886244836953109479921575474869);
DVK(KP471396736, +0.471396736825997648556387625905254377657460319);
DVK(KP881921264, +0.881921264348355029712756863660388349508442621);
DVK(KP195090322, +0.195090322016128267848284868477022240927691618);
DVK(KP980785280, +0.980785280403230449126182236134239036973933731);
DVK(KP555570233, +0.555570233019602224742830813948532874374937191);
DVK(KP831469612, +0.831469612302545237078788377617905756738560812);
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
V T4p, T5u, Tb, T3A, T2q, T3v, T6G, T78, Tq, T3w, T6B, T79, T2l, T3B, T4w;
V T5r, TI, T2g, T6u, T74, T3q, T3D, T4E, T5o, TZ, T2h, T6x, T75, T3t, T3E;
V T4L, T5p, T23, T2N, T6m, T70, T6p, T71, T2c, T2O, T3i, T3Y, T5f, T5R, T5k;
V T5S, T3l, T3Z, T1s, T2K, T6f, T6X, T6i, T6Y, T1B, T2L, T3b, T3V, T4Y, T5O;
V T53, T5P, T3e, T3W;
{
V T3, T4n, T2p, T4o, T6, T5s, T9, T5t;
{
V T1, T2, T2n, T2o;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0]));
T3 = VSUB(T1, T2);
T4n = VADD(T1, T2);
T2n = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
T2o = LD(&(xi[WS(is, 48)]), ivs, &(xi[0]));
T2p = VSUB(T2n, T2o);
T4o = VADD(T2n, T2o);
}
{
V T4, T5, T7, T8;
T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0]));
T6 = VSUB(T4, T5);
T5s = VADD(T4, T5);
T7 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0]));
T9 = VSUB(T7, T8);
T5t = VADD(T7, T8);
}
T4p = VSUB(T4n, T4o);
T5u = VSUB(T5s, T5t);
{
V Ta, T2m, T6E, T6F;
Ta = VMUL(LDK(KP707106781), VADD(T6, T9));
Tb = VSUB(T3, Ta);
T3A = VADD(T3, Ta);
T2m = VMUL(LDK(KP707106781), VSUB(T6, T9));
T2q = VSUB(T2m, T2p);
T3v = VADD(T2p, T2m);
T6E = VADD(T4n, T4o);
T6F = VADD(T5s, T5t);
T6G = VSUB(T6E, T6F);
T78 = VADD(T6E, T6F);
}
}
{
V Te, T4q, To, T4t, Th, T4r, Tl, T4u;
{
V Tc, Td, Tm, Tn;
Tc = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Td = LD(&(xi[WS(is, 36)]), ivs, &(xi[0]));
Te = VSUB(Tc, Td);
T4q = VADD(Tc, Td);
Tm = LD(&(xi[WS(is, 60)]), ivs, &(xi[0]));
Tn = LD(&(xi[WS(is, 28)]), ivs, &(xi[0]));
To = VSUB(Tm, Tn);
T4t = VADD(Tm, Tn);
}
{
V Tf, Tg, Tj, Tk;
Tf = LD(&(xi[WS(is, 20)]), ivs, &(xi[0]));
Tg = LD(&(xi[WS(is, 52)]), ivs, &(xi[0]));
Th = VSUB(Tf, Tg);
T4r = VADD(Tf, Tg);
Tj = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Tk = LD(&(xi[WS(is, 44)]), ivs, &(xi[0]));
Tl = VSUB(Tj, Tk);
T4u = VADD(Tj, Tk);
}
{
V Ti, Tp, T6z, T6A;
Ti = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th));
Tp = VFNMS(LDK(KP382683432), To, VMUL(LDK(KP923879532), Tl));
Tq = VSUB(Ti, Tp);
T3w = VADD(Ti, Tp);
T6z = VADD(T4q, T4r);
T6A = VADD(T4t, T4u);
T6B = VSUB(T6z, T6A);
T79 = VADD(T6z, T6A);
}
{
V T2j, T2k, T4s, T4v;
T2j = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te));
T2k = VFMA(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl));
T2l = VSUB(T2j, T2k);
T3B = VADD(T2j, T2k);
T4s = VSUB(T4q, T4r);
T4v = VSUB(T4t, T4u);
T4w = VMUL(LDK(KP707106781), VADD(T4s, T4v));
T5r = VMUL(LDK(KP707106781), VSUB(T4s, T4v));
}
}
{
V TB, T4z, TF, T4y, Ty, T4C, TG, T4B;
{
V Tz, TA, TD, TE;
Tz = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
TA = LD(&(xi[WS(is, 50)]), ivs, &(xi[0]));
TB = VSUB(Tz, TA);
T4z = VADD(Tz, TA);
TD = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
TE = LD(&(xi[WS(is, 34)]), ivs, &(xi[0]));
TF = VSUB(TD, TE);
T4y = VADD(TD, TE);
{
V Ts, Tt, Tu, Tv, Tw, Tx;
Ts = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tt = LD(&(xi[WS(is, 42)]), ivs, &(xi[0]));
Tu = VSUB(Ts, Tt);
Tv = LD(&(xi[WS(is, 58)]), ivs, &(xi[0]));
Tw = LD(&(xi[WS(is, 26)]), ivs, &(xi[0]));
Tx = VSUB(Tv, Tw);
Ty = VMUL(LDK(KP707106781), VSUB(Tu, Tx));
T4C = VADD(Tv, Tw);
TG = VMUL(LDK(KP707106781), VADD(Tu, Tx));
T4B = VADD(Ts, Tt);
}
}
{
V TC, TH, T6s, T6t;
TC = VSUB(Ty, TB);
TH = VSUB(TF, TG);
TI = VFMA(LDK(KP831469612), TC, VMUL(LDK(KP555570233), TH));
T2g = VFNMS(LDK(KP555570233), TC, VMUL(LDK(KP831469612), TH));
T6s = VADD(T4y, T4z);
T6t = VADD(T4B, T4C);
T6u = VSUB(T6s, T6t);
T74 = VADD(T6s, T6t);
}
{
V T3o, T3p, T4A, T4D;
T3o = VADD(TB, Ty);
T3p = VADD(TF, TG);
T3q = VFMA(LDK(KP980785280), T3o, VMUL(LDK(KP195090322), T3p));
T3D = VFNMS(LDK(KP195090322), T3o, VMUL(LDK(KP980785280), T3p));
T4A = VSUB(T4y, T4z);
T4D = VSUB(T4B, T4C);
T4E = VFMA(LDK(KP382683432), T4A, VMUL(LDK(KP923879532), T4D));
T5o = VFNMS(LDK(KP382683432), T4D, VMUL(LDK(KP923879532), T4A));
}
}
{
V TS, T4J, TW, T4I, TP, T4G, TX, T4F;
{
V TQ, TR, TU, TV;
TQ = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
TR = LD(&(xi[WS(is, 46)]), ivs, &(xi[0]));
TS = VSUB(TQ, TR);
T4J = VADD(TQ, TR);
TU = LD(&(xi[WS(is, 62)]), ivs, &(xi[0]));
TV = LD(&(xi[WS(is, 30)]), ivs, &(xi[0]));
TW = VSUB(TU, TV);
T4I = VADD(TU, TV);
{
V TJ, TK, TL, TM, TN, TO;
TJ = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
TK = LD(&(xi[WS(is, 38)]), ivs, &(xi[0]));
TL = VSUB(TJ, TK);
TM = LD(&(xi[WS(is, 54)]), ivs, &(xi[0]));
TN = LD(&(xi[WS(is, 22)]), ivs, &(xi[0]));
TO = VSUB(TM, TN);
TP = VMUL(LDK(KP707106781), VSUB(TL, TO));
T4G = VADD(TM, TN);
TX = VMUL(LDK(KP707106781), VADD(TL, TO));
T4F = VADD(TJ, TK);
}
}
{
V TT, TY, T6v, T6w;
TT = VSUB(TP, TS);
TY = VSUB(TW, TX);
TZ = VFNMS(LDK(KP555570233), TY, VMUL(LDK(KP831469612), TT));
T2h = VFMA(LDK(KP555570233), TT, VMUL(LDK(KP831469612), TY));
T6v = VADD(T4I, T4J);
T6w = VADD(T4F, T4G);
T6x = VSUB(T6v, T6w);
T75 = VADD(T6v, T6w);
}
{
V T3r, T3s, T4H, T4K;
T3r = VADD(TS, TP);
T3s = VADD(TW, TX);
T3t = VFNMS(LDK(KP195090322), T3s, VMUL(LDK(KP980785280), T3r));
T3E = VFMA(LDK(KP195090322), T3r, VMUL(LDK(KP980785280), T3s));
T4H = VSUB(T4F, T4G);
T4K = VSUB(T4I, T4J);
T4L = VFNMS(LDK(KP382683432), T4K, VMUL(LDK(KP923879532), T4H));
T5p = VFMA(LDK(KP923879532), T4K, VMUL(LDK(KP382683432), T4H));
}
}
{
V T21, T5h, T26, T5g, T1Y, T5d, T27, T5c, T55, T56, T1J, T57, T29, T58, T59;
V T1Q, T5a, T2a;
{
V T1Z, T20, T24, T25;
T1Z = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
T20 = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)]));
T21 = VSUB(T1Z, T20);
T5h = VADD(T1Z, T20);
T24 = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)]));
T25 = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)]));
T26 = VSUB(T24, T25);
T5g = VADD(T24, T25);
}
{
V T1S, T1T, T1U, T1V, T1W, T1X;
T1S = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T1T = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)]));
T1U = VSUB(T1S, T1T);
T1V = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)]));
T1W = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)]));
T1X = VSUB(T1V, T1W);
T1Y = VMUL(LDK(KP707106781), VSUB(T1U, T1X));
T5d = VADD(T1V, T1W);
T27 = VMUL(LDK(KP707106781), VADD(T1U, T1X));
T5c = VADD(T1S, T1T);
}
{
V T1F, T1I, T1M, T1P;
{
V T1D, T1E, T1G, T1H;
T1D = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T1E = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)]));
T1F = VSUB(T1D, T1E);
T55 = VADD(T1D, T1E);
T1G = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
T1H = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)]));
T1I = VSUB(T1G, T1H);
T56 = VADD(T1G, T1H);
}
T1J = VFNMS(LDK(KP382683432), T1I, VMUL(LDK(KP923879532), T1F));
T57 = VSUB(T55, T56);
T29 = VFMA(LDK(KP382683432), T1F, VMUL(LDK(KP923879532), T1I));
{
V T1K, T1L, T1N, T1O;
T1K = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)]));
T1L = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)]));
T1M = VSUB(T1K, T1L);
T58 = VADD(T1K, T1L);
T1N = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
T1O = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)]));
T1P = VSUB(T1N, T1O);
T59 = VADD(T1N, T1O);
}
T1Q = VFMA(LDK(KP923879532), T1M, VMUL(LDK(KP382683432), T1P));
T5a = VSUB(T58, T59);
T2a = VFNMS(LDK(KP382683432), T1M, VMUL(LDK(KP923879532), T1P));
}
{
V T1R, T22, T6k, T6l;
T1R = VSUB(T1J, T1Q);
T22 = VSUB(T1Y, T21);
T23 = VSUB(T1R, T22);
T2N = VADD(T22, T1R);
T6k = VADD(T5g, T5h);
T6l = VADD(T5c, T5d);
T6m = VSUB(T6k, T6l);
T70 = VADD(T6k, T6l);
}
{
V T6n, T6o, T28, T2b;
T6n = VADD(T55, T56);
T6o = VADD(T58, T59);
T6p = VSUB(T6n, T6o);
T71 = VADD(T6n, T6o);
T28 = VSUB(T26, T27);
T2b = VSUB(T29, T2a);
T2c = VSUB(T28, T2b);
T2O = VADD(T28, T2b);
}
{
V T3g, T3h, T5b, T5e;
T3g = VADD(T26, T27);
T3h = VADD(T1J, T1Q);
T3i = VADD(T3g, T3h);
T3Y = VSUB(T3g, T3h);
T5b = VMUL(LDK(KP707106781), VSUB(T57, T5a));
T5e = VSUB(T5c, T5d);
T5f = VSUB(T5b, T5e);
T5R = VADD(T5e, T5b);
}
{
V T5i, T5j, T3j, T3k;
T5i = VSUB(T5g, T5h);
T5j = VMUL(LDK(KP707106781), VADD(T57, T5a));
T5k = VSUB(T5i, T5j);
T5S = VADD(T5i, T5j);
T3j = VADD(T21, T1Y);
T3k = VADD(T29, T2a);
T3l = VADD(T3j, T3k);
T3Z = VSUB(T3k, T3j);
}
}
{
V T1q, T50, T1v, T4Z, T1n, T4W, T1w, T4V, T4O, T4P, T18, T4Q, T1y, T4R, T4S;
V T1f, T4T, T1z;
{
V T1o, T1p, T1t, T1u;
T1o = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
T1p = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)]));
T1q = VSUB(T1o, T1p);
T50 = VADD(T1o, T1p);
T1t = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T1u = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)]));
T1v = VSUB(T1t, T1u);
T4Z = VADD(T1t, T1u);
}
{
V T1h, T1i, T1j, T1k, T1l, T1m;
T1h = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T1i = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)]));
T1j = VSUB(T1h, T1i);
T1k = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)]));
T1l = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)]));
T1m = VSUB(T1k, T1l);
T1n = VMUL(LDK(KP707106781), VSUB(T1j, T1m));
T4W = VADD(T1k, T1l);
T1w = VMUL(LDK(KP707106781), VADD(T1j, T1m));
T4V = VADD(T1h, T1i);
}
{
V T14, T17, T1b, T1e;
{
V T12, T13, T15, T16;
T12 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T13 = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)]));
T14 = VSUB(T12, T13);
T4O = VADD(T12, T13);
T15 = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)]));
T16 = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)]));
T17 = VSUB(T15, T16);
T4P = VADD(T15, T16);
}
T18 = VFNMS(LDK(KP382683432), T17, VMUL(LDK(KP923879532), T14));
T4Q = VSUB(T4O, T4P);
T1y = VFMA(LDK(KP382683432), T14, VMUL(LDK(KP923879532), T17));
{
V T19, T1a, T1c, T1d;
T19 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)]));
T1a = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)]));
T1b = VSUB(T19, T1a);
T4R = VADD(T19, T1a);
T1c = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
T1d = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)]));
T1e = VSUB(T1c, T1d);
T4S = VADD(T1c, T1d);
}
T1f = VFMA(LDK(KP923879532), T1b, VMUL(LDK(KP382683432), T1e));
T4T = VSUB(T4R, T4S);
T1z = VFNMS(LDK(KP382683432), T1b, VMUL(LDK(KP923879532), T1e));
}
{
V T1g, T1r, T6d, T6e;
T1g = VSUB(T18, T1f);
T1r = VSUB(T1n, T1q);
T1s = VSUB(T1g, T1r);
T2K = VADD(T1r, T1g);
T6d = VADD(T4Z, T50);
T6e = VADD(T4V, T4W);
T6f = VSUB(T6d, T6e);
T6X = VADD(T6d, T6e);
}
{
V T6g, T6h, T1x, T1A;
T6g = VADD(T4O, T4P);
T6h = VADD(T4R, T4S);
T6i = VSUB(T6g, T6h);
T6Y = VADD(T6g, T6h);
T1x = VSUB(T1v, T1w);
T1A = VSUB(T1y, T1z);
T1B = VSUB(T1x, T1A);
T2L = VADD(T1x, T1A);
}
{
V T39, T3a, T4U, T4X;
T39 = VADD(T1v, T1w);
T3a = VADD(T18, T1f);
T3b = VADD(T39, T3a);
T3V = VSUB(T39, T3a);
T4U = VMUL(LDK(KP707106781), VSUB(T4Q, T4T));
T4X = VSUB(T4V, T4W);
T4Y = VSUB(T4U, T4X);
T5O = VADD(T4X, T4U);
}
{
V T51, T52, T3c, T3d;
T51 = VSUB(T4Z, T50);
T52 = VMUL(LDK(KP707106781), VADD(T4Q, T4T));
T53 = VSUB(T51, T52);
T5P = VADD(T51, T52);
T3c = VADD(T1q, T1n);
T3d = VADD(T1y, T1z);
T3e = VADD(T3c, T3d);
T3W = VSUB(T3d, T3c);
}
}
{
V T7h, T7l, T7k, T7m;
{
V T7f, T7g, T7i, T7j;
T7f = VADD(T78, T79);
T7g = VADD(T74, T75);
T7h = VSUB(T7f, T7g);
T7l = VADD(T7f, T7g);
T7i = VADD(T6X, T6Y);
T7j = VADD(T70, T71);
T7k = VBYI(VSUB(T7i, T7j));
T7m = VADD(T7i, T7j);
}
ST(&(xo[96]), VSUB(T7h, T7k), ovs, &(xo[0]));
ST(&(xo[0]), VADD(T7l, T7m), ovs, &(xo[0]));
ST(&(xo[32]), VADD(T7h, T7k), ovs, &(xo[0]));
ST(&(xo[64]), VSUB(T7l, T7m), ovs, &(xo[0]));
}
{
V T76, T7a, T73, T7b, T6Z, T72;
T76 = VSUB(T74, T75);
T7a = VSUB(T78, T79);
T6Z = VSUB(T6X, T6Y);
T72 = VSUB(T70, T71);
T73 = VMUL(LDK(KP707106781), VSUB(T6Z, T72));
T7b = VMUL(LDK(KP707106781), VADD(T6Z, T72));
{
V T77, T7c, T7d, T7e;
T77 = VBYI(VSUB(T73, T76));
T7c = VSUB(T7a, T7b);
ST(&(xo[48]), VADD(T77, T7c), ovs, &(xo[0]));
ST(&(xo[80]), VSUB(T7c, T77), ovs, &(xo[0]));
T7d = VBYI(VADD(T76, T73));
T7e = VADD(T7a, T7b);
ST(&(xo[16]), VADD(T7d, T7e), ovs, &(xo[0]));
ST(&(xo[112]), VSUB(T7e, T7d), ovs, &(xo[0]));
}
}
{
V T6C, T6S, T6I, T6P, T6r, T6Q, T6L, T6T, T6y, T6H;
T6y = VMUL(LDK(KP707106781), VSUB(T6u, T6x));
T6C = VSUB(T6y, T6B);
T6S = VADD(T6B, T6y);
T6H = VMUL(LDK(KP707106781), VADD(T6u, T6x));
T6I = VSUB(T6G, T6H);
T6P = VADD(T6G, T6H);
{
V T6j, T6q, T6J, T6K;
T6j = VFNMS(LDK(KP382683432), T6i, VMUL(LDK(KP923879532), T6f));
T6q = VFMA(LDK(KP923879532), T6m, VMUL(LDK(KP382683432), T6p));
T6r = VSUB(T6j, T6q);
T6Q = VADD(T6j, T6q);
T6J = VFMA(LDK(KP382683432), T6f, VMUL(LDK(KP923879532), T6i));
T6K = VFNMS(LDK(KP382683432), T6m, VMUL(LDK(KP923879532), T6p));
T6L = VSUB(T6J, T6K);
T6T = VADD(T6J, T6K);
}
{
V T6D, T6M, T6V, T6W;
T6D = VBYI(VSUB(T6r, T6C));
T6M = VSUB(T6I, T6L);
ST(&(xo[40]), VADD(T6D, T6M), ovs, &(xo[0]));
ST(&(xo[88]), VSUB(T6M, T6D), ovs, &(xo[0]));
T6V = VSUB(T6P, T6Q);
T6W = VBYI(VSUB(T6T, T6S));
ST(&(xo[72]), VSUB(T6V, T6W), ovs, &(xo[0]));
ST(&(xo[56]), VADD(T6V, T6W), ovs, &(xo[0]));
}
{
V T6N, T6O, T6R, T6U;
T6N = VBYI(VADD(T6C, T6r));
T6O = VADD(T6I, T6L);
ST(&(xo[24]), VADD(T6N, T6O), ovs, &(xo[0]));
ST(&(xo[104]), VSUB(T6O, T6N), ovs, &(xo[0]));
T6R = VADD(T6P, T6Q);
T6U = VBYI(VADD(T6S, T6T));
ST(&(xo[120]), VSUB(T6R, T6U), ovs, &(xo[0]));
ST(&(xo[8]), VADD(T6R, T6U), ovs, &(xo[0]));
}
}
{
V T5N, T68, T61, T69, T5U, T65, T5Y, T66;
{
V T5L, T5M, T5Z, T60;
T5L = VADD(T4p, T4w);
T5M = VADD(T5o, T5p);
T5N = VSUB(T5L, T5M);
T68 = VADD(T5L, T5M);
T5Z = VFNMS(LDK(KP195090322), T5O, VMUL(LDK(KP980785280), T5P));
T60 = VFMA(LDK(KP195090322), T5R, VMUL(LDK(KP980785280), T5S));
T61 = VSUB(T5Z, T60);
T69 = VADD(T5Z, T60);
}
{
V T5Q, T5T, T5W, T5X;
T5Q = VFMA(LDK(KP980785280), T5O, VMUL(LDK(KP195090322), T5P));
T5T = VFNMS(LDK(KP195090322), T5S, VMUL(LDK(KP980785280), T5R));
T5U = VSUB(T5Q, T5T);
T65 = VADD(T5Q, T5T);
T5W = VADD(T4E, T4L);
T5X = VADD(T5u, T5r);
T5Y = VSUB(T5W, T5X);
T66 = VADD(T5X, T5W);
}
{
V T5V, T62, T6b, T6c;
T5V = VADD(T5N, T5U);
T62 = VBYI(VADD(T5Y, T61));
ST(&(xo[100]), VSUB(T5V, T62), ovs, &(xo[0]));
ST(&(xo[28]), VADD(T5V, T62), ovs, &(xo[0]));
T6b = VBYI(VADD(T66, T65));
T6c = VADD(T68, T69);
ST(&(xo[4]), VADD(T6b, T6c), ovs, &(xo[0]));
ST(&(xo[124]), VSUB(T6c, T6b), ovs, &(xo[0]));
}
{
V T63, T64, T67, T6a;
T63 = VSUB(T5N, T5U);
T64 = VBYI(VSUB(T61, T5Y));
ST(&(xo[92]), VSUB(T63, T64), ovs, &(xo[0]));
ST(&(xo[36]), VADD(T63, T64), ovs, &(xo[0]));
T67 = VBYI(VSUB(T65, T66));
T6a = VSUB(T68, T69);
ST(&(xo[60]), VADD(T67, T6a), ovs, &(xo[0]));
ST(&(xo[68]), VSUB(T6a, T67), ovs, &(xo[0]));
}
}
{
V T11, T2C, T2v, T2D, T2e, T2z, T2s, T2A;
{
V Tr, T10, T2t, T2u;
Tr = VSUB(Tb, Tq);
T10 = VSUB(TI, TZ);
T11 = VSUB(Tr, T10);
T2C = VADD(Tr, T10);
T2t = VFNMS(LDK(KP471396736), T1s, VMUL(LDK(KP881921264), T1B));
T2u = VFMA(LDK(KP471396736), T23, VMUL(LDK(KP881921264), T2c));
T2v = VSUB(T2t, T2u);
T2D = VADD(T2t, T2u);
}
{
V T1C, T2d, T2i, T2r;
T1C = VFMA(LDK(KP881921264), T1s, VMUL(LDK(KP471396736), T1B));
T2d = VFNMS(LDK(KP471396736), T2c, VMUL(LDK(KP881921264), T23));
T2e = VSUB(T1C, T2d);
T2z = VADD(T1C, T2d);
T2i = VSUB(T2g, T2h);
T2r = VSUB(T2l, T2q);
T2s = VSUB(T2i, T2r);
T2A = VADD(T2r, T2i);
}
{
V T2f, T2w, T2F, T2G;
T2f = VADD(T11, T2e);
T2w = VBYI(VADD(T2s, T2v));
ST(&(xo[106]), VSUB(T2f, T2w), ovs, &(xo[2]));
ST(&(xo[22]), VADD(T2f, T2w), ovs, &(xo[2]));
T2F = VBYI(VADD(T2A, T2z));
T2G = VADD(T2C, T2D);
ST(&(xo[10]), VADD(T2F, T2G), ovs, &(xo[2]));
ST(&(xo[118]), VSUB(T2G, T2F), ovs, &(xo[2]));
}
{
V T2x, T2y, T2B, T2E;
T2x = VSUB(T11, T2e);
T2y = VBYI(VSUB(T2v, T2s));
ST(&(xo[86]), VSUB(T2x, T2y), ovs, &(xo[2]));
ST(&(xo[42]), VADD(T2x, T2y), ovs, &(xo[2]));
T2B = VBYI(VSUB(T2z, T2A));
T2E = VSUB(T2C, T2D);
ST(&(xo[54]), VADD(T2B, T2E), ovs, &(xo[2]));
ST(&(xo[74]), VSUB(T2E, T2B), ovs, &(xo[2]));
}
}
{
V T3n, T3O, T3J, T3R, T3y, T3Q, T3G, T3N;
{
V T3f, T3m, T3H, T3I;
T3f = VFNMS(LDK(KP098017140), T3e, VMUL(LDK(KP995184726), T3b));
T3m = VFMA(LDK(KP995184726), T3i, VMUL(LDK(KP098017140), T3l));
T3n = VSUB(T3f, T3m);
T3O = VADD(T3f, T3m);
T3H = VFMA(LDK(KP098017140), T3b, VMUL(LDK(KP995184726), T3e));
T3I = VFNMS(LDK(KP098017140), T3i, VMUL(LDK(KP995184726), T3l));
T3J = VSUB(T3H, T3I);
T3R = VADD(T3H, T3I);
}
{
V T3u, T3x, T3C, T3F;
T3u = VADD(T3q, T3t);
T3x = VADD(T3v, T3w);
T3y = VSUB(T3u, T3x);
T3Q = VADD(T3x, T3u);
T3C = VADD(T3A, T3B);
T3F = VADD(T3D, T3E);
T3G = VSUB(T3C, T3F);
T3N = VADD(T3C, T3F);
}
{
V T3z, T3K, T3T, T3U;
T3z = VBYI(VSUB(T3n, T3y));
T3K = VSUB(T3G, T3J);
ST(&(xo[34]), VADD(T3z, T3K), ovs, &(xo[2]));
ST(&(xo[94]), VSUB(T3K, T3z), ovs, &(xo[2]));
T3T = VSUB(T3N, T3O);
T3U = VBYI(VSUB(T3R, T3Q));
ST(&(xo[66]), VSUB(T3T, T3U), ovs, &(xo[2]));
ST(&(xo[62]), VADD(T3T, T3U), ovs, &(xo[2]));
}
{
V T3L, T3M, T3P, T3S;
T3L = VBYI(VADD(T3y, T3n));
T3M = VADD(T3G, T3J);
ST(&(xo[30]), VADD(T3L, T3M), ovs, &(xo[2]));
ST(&(xo[98]), VSUB(T3M, T3L), ovs, &(xo[2]));
T3P = VADD(T3N, T3O);
T3S = VBYI(VADD(T3Q, T3R));
ST(&(xo[126]), VSUB(T3P, T3S), ovs, &(xo[2]));
ST(&(xo[2]), VADD(T3P, T3S), ovs, &(xo[2]));
}
}
{
V T4N, T5G, T5z, T5H, T5m, T5D, T5w, T5E;
{
V T4x, T4M, T5x, T5y;
T4x = VSUB(T4p, T4w);
T4M = VSUB(T4E, T4L);
T4N = VSUB(T4x, T4M);
T5G = VADD(T4x, T4M);
T5x = VFNMS(LDK(KP555570233), T4Y, VMUL(LDK(KP831469612), T53));
T5y = VFMA(LDK(KP555570233), T5f, VMUL(LDK(KP831469612), T5k));
T5z = VSUB(T5x, T5y);
T5H = VADD(T5x, T5y);
}
{
V T54, T5l, T5q, T5v;
T54 = VFMA(LDK(KP831469612), T4Y, VMUL(LDK(KP555570233), T53));
T5l = VFNMS(LDK(KP555570233), T5k, VMUL(LDK(KP831469612), T5f));
T5m = VSUB(T54, T5l);
T5D = VADD(T54, T5l);
T5q = VSUB(T5o, T5p);
T5v = VSUB(T5r, T5u);
T5w = VSUB(T5q, T5v);
T5E = VADD(T5v, T5q);
}
{
V T5n, T5A, T5J, T5K;
T5n = VADD(T4N, T5m);
T5A = VBYI(VADD(T5w, T5z));
ST(&(xo[108]), VSUB(T5n, T5A), ovs, &(xo[0]));
ST(&(xo[20]), VADD(T5n, T5A), ovs, &(xo[0]));
T5J = VBYI(VADD(T5E, T5D));
T5K = VADD(T5G, T5H);
ST(&(xo[12]), VADD(T5J, T5K), ovs, &(xo[0]));
ST(&(xo[116]), VSUB(T5K, T5J), ovs, &(xo[0]));
}
{
V T5B, T5C, T5F, T5I;
T5B = VSUB(T4N, T5m);
T5C = VBYI(VSUB(T5z, T5w));
ST(&(xo[84]), VSUB(T5B, T5C), ovs, &(xo[0]));
ST(&(xo[44]), VADD(T5B, T5C), ovs, &(xo[0]));
T5F = VBYI(VSUB(T5D, T5E));
T5I = VSUB(T5G, T5H);
ST(&(xo[52]), VADD(T5F, T5I), ovs, &(xo[0]));
ST(&(xo[76]), VSUB(T5I, T5F), ovs, &(xo[0]));
}
}
{
V T2J, T34, T2X, T35, T2Q, T31, T2U, T32;
{
V T2H, T2I, T2V, T2W;
T2H = VADD(Tb, Tq);
T2I = VADD(T2g, T2h);
T2J = VSUB(T2H, T2I);
T34 = VADD(T2H, T2I);
T2V = VFNMS(LDK(KP290284677), T2K, VMUL(LDK(KP956940335), T2L));
T2W = VFMA(LDK(KP290284677), T2N, VMUL(LDK(KP956940335), T2O));
T2X = VSUB(T2V, T2W);
T35 = VADD(T2V, T2W);
}
{
V T2M, T2P, T2S, T2T;
T2M = VFMA(LDK(KP956940335), T2K, VMUL(LDK(KP290284677), T2L));
T2P = VFNMS(LDK(KP290284677), T2O, VMUL(LDK(KP956940335), T2N));
T2Q = VSUB(T2M, T2P);
T31 = VADD(T2M, T2P);
T2S = VADD(TI, TZ);
T2T = VADD(T2q, T2l);
T2U = VSUB(T2S, T2T);
T32 = VADD(T2T, T2S);
}
{
V T2R, T2Y, T37, T38;
T2R = VADD(T2J, T2Q);
T2Y = VBYI(VADD(T2U, T2X));
ST(&(xo[102]), VSUB(T2R, T2Y), ovs, &(xo[2]));
ST(&(xo[26]), VADD(T2R, T2Y), ovs, &(xo[2]));
T37 = VBYI(VADD(T32, T31));
T38 = VADD(T34, T35);
ST(&(xo[6]), VADD(T37, T38), ovs, &(xo[2]));
ST(&(xo[122]), VSUB(T38, T37), ovs, &(xo[2]));
}
{
V T2Z, T30, T33, T36;
T2Z = VSUB(T2J, T2Q);
T30 = VBYI(VSUB(T2X, T2U));
ST(&(xo[90]), VSUB(T2Z, T30), ovs, &(xo[2]));
ST(&(xo[38]), VADD(T2Z, T30), ovs, &(xo[2]));
T33 = VBYI(VSUB(T31, T32));
T36 = VSUB(T34, T35);
ST(&(xo[58]), VADD(T33, T36), ovs, &(xo[2]));
ST(&(xo[70]), VSUB(T36, T33), ovs, &(xo[2]));
}
}
{
V T41, T4g, T4b, T4j, T44, T4i, T48, T4f;
{
V T3X, T40, T49, T4a;
T3X = VFNMS(LDK(KP634393284), T3W, VMUL(LDK(KP773010453), T3V));
T40 = VFMA(LDK(KP773010453), T3Y, VMUL(LDK(KP634393284), T3Z));
T41 = VSUB(T3X, T40);
T4g = VADD(T3X, T40);
T49 = VFMA(LDK(KP634393284), T3V, VMUL(LDK(KP773010453), T3W));
T4a = VFNMS(LDK(KP634393284), T3Y, VMUL(LDK(KP773010453), T3Z));
T4b = VSUB(T49, T4a);
T4j = VADD(T49, T4a);
}
{
V T42, T43, T46, T47;
T42 = VSUB(T3D, T3E);
T43 = VSUB(T3w, T3v);
T44 = VSUB(T42, T43);
T4i = VADD(T43, T42);
T46 = VSUB(T3A, T3B);
T47 = VSUB(T3q, T3t);
T48 = VSUB(T46, T47);
T4f = VADD(T46, T47);
}
{
V T45, T4c, T4l, T4m;
T45 = VBYI(VSUB(T41, T44));
T4c = VSUB(T48, T4b);
ST(&(xo[46]), VADD(T45, T4c), ovs, &(xo[2]));
ST(&(xo[82]), VSUB(T4c, T45), ovs, &(xo[2]));
T4l = VSUB(T4f, T4g);
T4m = VBYI(VSUB(T4j, T4i));
ST(&(xo[78]), VSUB(T4l, T4m), ovs, &(xo[2]));
ST(&(xo[50]), VADD(T4l, T4m), ovs, &(xo[2]));
}
{
V T4d, T4e, T4h, T4k;
T4d = VBYI(VADD(T44, T41));
T4e = VADD(T48, T4b);
ST(&(xo[18]), VADD(T4d, T4e), ovs, &(xo[2]));
ST(&(xo[110]), VSUB(T4e, T4d), ovs, &(xo[2]));
T4h = VADD(T4f, T4g);
T4k = VBYI(VADD(T4i, T4j));
ST(&(xo[114]), VSUB(T4h, T4k), ovs, &(xo[2]));
ST(&(xo[14]), VADD(T4h, T4k), ovs, &(xo[2]));
}
}
}
}
static void t1fuv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
{
INT m;
R *x;
x = ri;
for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(rs)) {
V Tr, TH, Tg, Tl, Tm, TA, TB, TJ, T5, Ta, Tb, TD, TE, TI, To;
V Tq, Tp;
To = LD(&(x[0]), ms, &(x[0]));
Tp = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tq = BYTWJ(&(W[TWVL * 8]), Tp);
Tr = VSUB(To, Tq);
TH = VADD(To, Tq);
{
V Td, Tk, Tf, Ti;
{
V Tc, Tj, Te, Th;
Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Td = BYTWJ(&(W[TWVL * 6]), Tc);
Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Tk = BYTWJ(&(W[0]), Tj);
Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
Tf = BYTWJ(&(W[TWVL * 16]), Te);
Th = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Ti = BYTWJ(&(W[TWVL * 10]), Th);
}
Tg = VSUB(Td, Tf);
Tl = VSUB(Ti, Tk);
Tm = VADD(Tg, Tl);
TA = VADD(Td, Tf);
TB = VADD(Ti, Tk);
TJ = VADD(TA, TB);
}
{
V T2, T9, T4, T7;
{
V T1, T8, T3, T6;
T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
T2 = BYTWJ(&(W[TWVL * 2]), T1);
T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T9 = BYTWJ(&(W[TWVL * 4]), T8);
T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T4 = BYTWJ(&(W[TWVL * 12]), T3);
T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
T7 = BYTWJ(&(W[TWVL * 14]), T6);
}
T5 = VSUB(T2, T4);
Ta = VSUB(T7, T9);
Tb = VADD(T5, Ta);
TD = VADD(T2, T4);
TE = VADD(T7, T9);
TI = VADD(TD, TE);
}
{
V Tn, Ts, Tt, Tx, Tz, Tv, Tw, Ty, Tu;
Tn = VMUL(LDK(KP559016994), VSUB(Tb, Tm));
Ts = VADD(Tb, Tm);
Tt = VFNMS(LDK(KP250000000), Ts, Tr);
Tv = VSUB(T5, Ta);
Tw = VSUB(Tg, Tl);
Tx = VBYI(VFMA(LDK(KP951056516), Tv, VMUL(LDK(KP587785252), Tw)));
Tz = VBYI(VFNMS(LDK(KP587785252), Tv, VMUL(LDK(KP951056516), Tw)));
ST(&(x[WS(rs, 5)]), VADD(Tr, Ts), ms, &(x[WS(rs, 1)]));
Ty = VSUB(Tt, Tn);
ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)]));
Tu = VADD(Tn, Tt);
ST(&(x[WS(rs, 1)]), VSUB(Tu, Tx), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 9)]), VADD(Tx, Tu), ms, &(x[WS(rs, 1)]));
}
{
V TM, TK, TL, TG, TO, TC, TF, TP, TN;
TM = VMUL(LDK(KP559016994), VSUB(TI, TJ));
TK = VADD(TI, TJ);
TL = VFNMS(LDK(KP250000000), TK, TH);
TC = VSUB(TA, TB);
TF = VSUB(TD, TE);
TG = VBYI(VFNMS(LDK(KP587785252), TF, VMUL(LDK(KP951056516), TC)));
TO = VBYI(VFMA(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC)));
ST(&(x[0]), VADD(TH, TK), ms, &(x[0]));
TP = VADD(TM, TL);
ST(&(x[WS(rs, 4)]), VADD(TO, TP), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VSUB(TP, TO), ms, &(x[0]));
TN = VSUB(TL, TM);
ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0]));
ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0]));
}
}
}
VLEAVE();
}
static void n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP904176221, +0.904176221990848204433795481776887926501523162);
DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
DVK(KP516520780, +0.516520780623489722840901288569017135705033622);
DVK(KP522026385, +0.522026385161275033714027226654165028300441940);
DVK(KP957805992, +0.957805992594665126462521754605754580515587217);
DVK(KP600477271, +0.600477271932665282925769253334763009352012849);
DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
DVK(KP769338817, +0.769338817572980603471413688209101117038278899);
DVK(KP859542535, +0.859542535098774820163672132761689612766401925);
DVK(KP581704778, +0.581704778510515730456870384989698884939833902);
DVK(KP853480001, +0.853480001859823990758994934970528322872359049);
DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
DVK(KP226109445, +0.226109445035782405468510155372505010481906348);
DVK(KP301479260, +0.301479260047709873958013540496673347309208464);
DVK(KP686558370, +0.686558370781754340655719594850823015421401653);
DVK(KP514918778, +0.514918778086315755491789696138117261566051239);
DVK(KP038632954, +0.038632954644348171955506895830342264440241080);
DVK(KP612264650, +0.612264650376756543746494474777125408779395514);
DVK(KP302775637, +0.302775637731994646559610633735247973125648287);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
{
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
V T1, T7, T2, Tg, Tf, TN, Th, Tq, Ta, Tj, T5, Tr, Tk;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
{
V Td, Te, T8, T9, T3, T4;
Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T4 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Tf = VADD(Td, Te);
TN = VSUB(Td, Te);
Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Tq = VSUB(T8, T9);
Ta = VADD(T8, T9);
Tj = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T5 = VADD(T3, T4);
Tr = VSUB(T4, T3);
Tk = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
}
{
V Tt, Ti, Ty, Tb, Ts, TQ, Tx, T6, Tu, Tl;
Tt = VSUB(Tg, Th);
Ti = VADD(Tg, Th);
Ty = VFMS(LDK(KP500000000), Ta, T7);
Tb = VADD(T7, Ta);
Ts = VSUB(Tq, Tr);
TQ = VADD(Tr, Tq);
Tx = VFNMS(LDK(KP500000000), T5, T2);
T6 = VADD(T2, T5);
Tu = VSUB(Tj, Tk);
Tl = VADD(Tj, Tk);
{
V TK, Tz, Tc, TX, Tv, TO, TL, Tm;
TK = VADD(Tx, Ty);
Tz = VSUB(Tx, Ty);
Tc = VADD(T6, Tb);
TX = VSUB(T6, Tb);
Tv = VSUB(Tt, Tu);
TO = VADD(Tt, Tu);
TL = VSUB(Ti, Tl);
Tm = VADD(Ti, Tl);
{
V TF, Tw, TP, TY, TT, TM, TA, Tn;
TF = VSUB(Ts, Tv);
Tw = VADD(Ts, Tv);
TP = VFNMS(LDK(KP500000000), TO, TN);
TY = VADD(TN, TO);
TT = VFNMS(LDK(KP866025403), TL, TK);
TM = VFMA(LDK(KP866025403), TL, TK);
TA = VFNMS(LDK(KP500000000), Tm, Tf);
Tn = VADD(Tf, Tm);
{
V T1f, T1n, TI, T18, T1k, T1c, TD, T17, T10, T1m, T16, T1e, TU, TR;
TU = VFNMS(LDK(KP866025403), TQ, TP);
TR = VFMA(LDK(KP866025403), TQ, TP);
{
V TZ, T15, TE, TB;
TZ = VFMA(LDK(KP302775637), TY, TX);
T15 = VFNMS(LDK(KP302775637), TX, TY);
TE = VSUB(Tz, TA);
TB = VADD(Tz, TA);
{
V TH, To, TV, T13;
TH = VSUB(Tc, Tn);
To = VADD(Tc, Tn);
TV = VFNMS(LDK(KP612264650), TU, TT);
T13 = VFMA(LDK(KP612264650), TT, TU);
{
V TS, T12, TG, T1b;
TS = VFNMS(LDK(KP038632954), TR, TM);
T12 = VFMA(LDK(KP038632954), TM, TR);
TG = VFNMS(LDK(KP514918778), TF, TE);
T1b = VFMA(LDK(KP686558370), TE, TF);
{
V TC, T1a, Tp, TW, T14;
TC = VFMA(LDK(KP301479260), TB, Tw);
T1a = VFNMS(LDK(KP226109445), Tw, TB);
Tp = VFNMS(LDK(KP083333333), To, T1);
ST(&(xo[0]), VADD(T1, To), ovs, &(xo[0]));
T1f = VFMA(LDK(KP853480001), TV, TS);
TW = VFNMS(LDK(KP853480001), TV, TS);
T1n = VFMA(LDK(KP853480001), T13, T12);
T14 = VFNMS(LDK(KP853480001), T13, T12);
TI = VFMA(LDK(KP581704778), TH, TG);
T18 = VFNMS(LDK(KP859542535), TG, TH);
T1k = VFMA(LDK(KP769338817), T1b, T1a);
T1c = VFNMS(LDK(KP769338817), T1b, T1a);
TD = VFMA(LDK(KP503537032), TC, Tp);
T17 = VFNMS(LDK(KP251768516), TC, Tp);
T10 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), TZ, TW));
T1m = VFNMS(LDK(KP522026385), TW, TZ);
T16 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), T15, T14));
T1e = VFNMS(LDK(KP522026385), T14, T15);
}
}
}
}
{
V T1o, T1q, T1g, T1i, T1d, T1h, T1l, T1p;
{
V T11, TJ, T19, T1j;
T11 = VFMA(LDK(KP516520780), TI, TD);
TJ = VFNMS(LDK(KP516520780), TI, TD);
T19 = VFMA(LDK(KP300462606), T18, T17);
T1j = VFNMS(LDK(KP300462606), T18, T17);
T1o = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1n, T1m));
T1q = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1n, T1m));
T1g = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1f, T1e));
T1i = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1f, T1e));
ST(&(xo[WS(os, 12)]), VFNMSI(T16, T11), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFMAI(T16, T11), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 8)]), VFMAI(T10, TJ), ovs, &(xo[0]));
ST(&(xo[WS(os, 5)]), VFNMSI(T10, TJ), ovs, &(xo[WS(os, 1)]));
T1d = VFNMS(LDK(KP503537032), T1c, T19);
T1h = VFMA(LDK(KP503537032), T1c, T19);
T1l = VFNMS(LDK(KP503537032), T1k, T1j);
T1p = VFMA(LDK(KP503537032), T1k, T1j);
}
ST(&(xo[WS(os, 9)]), VFMAI(T1g, T1d), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VFNMSI(T1g, T1d), ovs, &(xo[0]));
ST(&(xo[WS(os, 10)]), VFNMSI(T1i, T1h), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(T1i, T1h), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VFMAI(T1o, T1l), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 6)]), VFNMSI(T1o, T1l), ovs, &(xo[0]));
ST(&(xo[WS(os, 11)]), VFMAI(T1q, T1p), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VFNMSI(T1q, T1p), ovs, &(xo[0]));
}
}
}
}
}
}
}
VLEAVE();
}
void input(FILE *fd, void (*cb)(struct message *), struct message *p)
{
int c;
#define ST(X) do { p->status = (X); } while(0)
#define MK(X) do { p->X = p->p; } while(0)
#define RST() do { \
ST(AT_BOM); \
p->p = p->buffer; \
p->sz = p->argc = 0; \
p->org = p->cmd = NULL; \
} while (0)
#define ADDCH(X) do { \
*p->p++ = X; p->sz++; \
if (p->sz >= (sizeof p->buffer)-1) \
RST(); \
} while (0)
#define ADDARG(s) do { \
assert(p->argc < MAX_ARGS); \
p->argv[p->argc] = s; \
if (s) p->argc++; \
} while (0)
RST();
while ((c = fgetc(fd)) != EOF) {
switch (p->status) {
case AT_BOM:
switch (c) {
case ':': ST(IN_ORG); MK(org); break;
case ' ': break; /* ignored */
case '\r': ST(IN_EOL); break;
case '\n': RST(); break;
default: ST(IN_CMD); MK(cmd); ADDCH(c); break;
} break;
case IN_ORG:
switch(c) {
case ' ': ST(IN_SPC0); ADDCH(0); break;
case '\r': ST(IN_EOL); ADDCH(0); break;
case '\n': RST(); break;
case ':': /* NO BREAK HERE, valid char */
default: ADDCH(c); break;
} break;
case IN_SPC0:
switch(c) {
case ' ': break; /* ignore extra */
case '\r': ST(IN_EOL); break;
case '\n': RST(); break;
default: ST(IN_CMD); MK(cmd); ADDCH(c); break;
} break;
case IN_CMD:
switch(c) {
case ' ': ST(IN_SPC1); ADDCH(0); break;
case '\r': ST(IN_EOL); ADDCH(0); break;
case '\n': ADDARG(NULL); cb(p); RST(); break;
case ':': /* NO BREAK HERE, valid char */
default: ADDCH(c); break;
} break;
case IN_SPC1:
switch(c) {
case ' ': break; /* ignore */
case '\r': ST(IN_EOL); break;
case '\n': ADDARG(NULL); cb(p); RST(); break;
case ':': ST(IN_ARGN); ADDARG(p->p); break;
default: ST(IN_ARG); ADDARG(p->p); ADDCH(c); break;
} break;
case IN_ARG:
switch (c) {
case ' ': ST(IN_SPC1); ADDCH(0); break;
case '\r': ST(IN_EOL); ADDCH(0); break;
case '\n': ADDARG(NULL); cb(p); RST(); break;
case ':':
default: ADDCH(c); break;
} break;
case IN_ARGN:
switch (c) {
case '\r': ST(IN_EOL); ADDCH(0); break;
case '\n': ADDARG(NULL); cb(p); RST(); break;
default: ADDCH(c); break;
} break;
case IN_EOL:
switch (c) {
case '\r': break;
case '\n': ADDARG(NULL); cb(p); RST(); break;
case ':': RST(); ST(IN_ORG); break;
default: RST(); break;
} break;
} /* switch */
} /* while */
} /* input */
static void n1fv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP195090322, +0.195090322016128267848284868477022240927691618);
DVK(KP980785280, +0.980785280403230449126182236134239036973933731);
DVK(KP555570233, +0.555570233019602224742830813948532874374937191);
DVK(KP831469612, +0.831469612302545237078788377617905756738560812);
DVK(KP382683432, +0.382683432365089771728459984030398866761344562);
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) {
V T1T, T1W, T2K, T2x, T16, T1A, Tb, T1p, TT, T1v, TY, T1w, T27, T2a, T2b;
V T2H, T2O, TC, T1s, TH, T1t, T20, T23, T24, T2E, T2N, T2g, T2j, Tq, T1B;
V T19, T1q, T2A, T2L;
{
V T3, T1R, T15, T1S, T6, T1U, T9, T1V, T12, Ta;
{
V T1, T2, T13, T14;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
T3 = VSUB(T1, T2);
T1R = VADD(T1, T2);
T13 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T14 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0]));
T15 = VSUB(T13, T14);
T1S = VADD(T13, T14);
}
{
V T4, T5, T7, T8;
T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0]));
T6 = VSUB(T4, T5);
T1U = VADD(T4, T5);
T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T9 = VSUB(T7, T8);
T1V = VADD(T7, T8);
}
T1T = VADD(T1R, T1S);
T1W = VADD(T1U, T1V);
T2K = VSUB(T1V, T1U);
T2x = VSUB(T1R, T1S);
T12 = VMUL(LDK(KP707106781), VSUB(T9, T6));
T16 = VSUB(T12, T15);
T1A = VADD(T15, T12);
Ta = VMUL(LDK(KP707106781), VADD(T6, T9));
Tb = VADD(T3, Ta);
T1p = VSUB(T3, Ta);
}
{
V TL, T25, TX, T26, TO, T28, TR, T29;
{
V TJ, TK, TV, TW;
TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)]));
TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
TL = VSUB(TJ, TK);
T25 = VADD(TJ, TK);
TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
TW = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)]));
TX = VSUB(TV, TW);
T26 = VADD(TV, TW);
}
{
V TM, TN, TP, TQ;
TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
TO = VSUB(TM, TN);
T28 = VADD(TM, TN);
TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)]));
TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
TR = VSUB(TP, TQ);
T29 = VADD(TP, TQ);
}
{
V TS, TU, T2F, T2G;
TS = VMUL(LDK(KP707106781), VADD(TO, TR));
TT = VADD(TL, TS);
T1v = VSUB(TL, TS);
TU = VMUL(LDK(KP707106781), VSUB(TR, TO));
TY = VSUB(TU, TX);
T1w = VADD(TX, TU);
T27 = VADD(T25, T26);
T2a = VADD(T28, T29);
T2b = VSUB(T27, T2a);
T2F = VSUB(T25, T26);
T2G = VSUB(T29, T28);
T2H = VFNMS(LDK(KP382683432), T2G, VMUL(LDK(KP923879532), T2F));
T2O = VFMA(LDK(KP382683432), T2F, VMUL(LDK(KP923879532), T2G));
}
}
{
V Tu, T1Y, TG, T1Z, Tx, T21, TA, T22;
{
V Ts, Tt, TE, TF;
Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
Tu = VSUB(Ts, Tt);
T1Y = VADD(Ts, Tt);
TE = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
TF = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)]));
TG = VSUB(TE, TF);
T1Z = VADD(TE, TF);
}
{
V Tv, Tw, Ty, Tz;
Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)]));
Tx = VSUB(Tv, Tw);
T21 = VADD(Tv, Tw);
Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)]));
Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
TA = VSUB(Ty, Tz);
T22 = VADD(Ty, Tz);
}
{
V TB, TD, T2C, T2D;
TB = VMUL(LDK(KP707106781), VADD(Tx, TA));
TC = VADD(Tu, TB);
T1s = VSUB(Tu, TB);
TD = VMUL(LDK(KP707106781), VSUB(TA, Tx));
TH = VSUB(TD, TG);
T1t = VADD(TG, TD);
T20 = VADD(T1Y, T1Z);
T23 = VADD(T21, T22);
T24 = VSUB(T20, T23);
T2C = VSUB(T1Y, T1Z);
T2D = VSUB(T22, T21);
T2E = VFMA(LDK(KP923879532), T2C, VMUL(LDK(KP382683432), T2D));
T2N = VFNMS(LDK(KP382683432), T2C, VMUL(LDK(KP923879532), T2D));
}
}
{
V Te, T2h, To, T2f, Th, T2i, Tl, T2e, Ti, Tp;
{
V Tc, Td, Tm, Tn;
Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
Te = VSUB(Tc, Td);
T2h = VADD(Tc, Td);
Tm = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Tn = LD(&(xi[WS(is, 22)]), ivs, &(xi[0]));
To = VSUB(Tm, Tn);
T2f = VADD(Tm, Tn);
}
{
V Tf, Tg, Tj, Tk;
Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0]));
Th = VSUB(Tf, Tg);
T2i = VADD(Tf, Tg);
Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0]));
Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
Tl = VSUB(Tj, Tk);
T2e = VADD(Tj, Tk);
}
T2g = VADD(T2e, T2f);
T2j = VADD(T2h, T2i);
Ti = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te));
Tp = VFMA(LDK(KP923879532), Tl, VMUL(LDK(KP382683432), To));
Tq = VADD(Ti, Tp);
T1B = VSUB(Tp, Ti);
{
V T17, T18, T2y, T2z;
T17 = VFNMS(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl));
T18 = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th));
T19 = VSUB(T17, T18);
T1q = VADD(T18, T17);
T2y = VSUB(T2h, T2i);
T2z = VSUB(T2e, T2f);
T2A = VMUL(LDK(KP707106781), VADD(T2y, T2z));
T2L = VMUL(LDK(KP707106781), VSUB(T2z, T2y));
}
}
{
V T2d, T2n, T2m, T2o;
{
V T1X, T2c, T2k, T2l;
T1X = VSUB(T1T, T1W);
T2c = VMUL(LDK(KP707106781), VADD(T24, T2b));
T2d = VADD(T1X, T2c);
T2n = VSUB(T1X, T2c);
T2k = VSUB(T2g, T2j);
T2l = VMUL(LDK(KP707106781), VSUB(T2b, T24));
T2m = VBYI(VADD(T2k, T2l));
T2o = VBYI(VSUB(T2l, T2k));
}
ST(&(xo[WS(os, 28)]), VSUB(T2d, T2m), ovs, &(xo[0]));
ST(&(xo[WS(os, 12)]), VADD(T2n, T2o), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VADD(T2d, T2m), ovs, &(xo[0]));
ST(&(xo[WS(os, 20)]), VSUB(T2n, T2o), ovs, &(xo[0]));
}
{
V T2r, T2v, T2u, T2w;
{
V T2p, T2q, T2s, T2t;
T2p = VADD(T1T, T1W);
T2q = VADD(T2j, T2g);
T2r = VADD(T2p, T2q);
T2v = VSUB(T2p, T2q);
T2s = VADD(T20, T23);
T2t = VADD(T27, T2a);
T2u = VADD(T2s, T2t);
T2w = VBYI(VSUB(T2t, T2s));
}
ST(&(xo[WS(os, 16)]), VSUB(T2r, T2u), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VADD(T2v, T2w), ovs, &(xo[0]));
ST(&(xo[0]), VADD(T2r, T2u), ovs, &(xo[0]));
ST(&(xo[WS(os, 24)]), VSUB(T2v, T2w), ovs, &(xo[0]));
}
{
V T2V, T2Z, T2Y, T30;
{
V T2T, T2U, T2W, T2X;
T2T = VSUB(T2H, T2E);
T2U = VSUB(T2L, T2K);
T2V = VBYI(VSUB(T2T, T2U));
T2Z = VBYI(VADD(T2U, T2T));
T2W = VSUB(T2x, T2A);
T2X = VSUB(T2O, T2N);
T2Y = VSUB(T2W, T2X);
T30 = VADD(T2W, T2X);
}
ST(&(xo[WS(os, 10)]), VADD(T2V, T2Y), ovs, &(xo[0]));
ST(&(xo[WS(os, 26)]), VSUB(T30, T2Z), ovs, &(xo[0]));
ST(&(xo[WS(os, 22)]), VSUB(T2Y, T2V), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VADD(T2Z, T30), ovs, &(xo[0]));
}
{
V T2J, T2R, T2Q, T2S;
{
V T2B, T2I, T2M, T2P;
T2B = VADD(T2x, T2A);
T2I = VADD(T2E, T2H);
T2J = VADD(T2B, T2I);
T2R = VSUB(T2B, T2I);
T2M = VADD(T2K, T2L);
T2P = VADD(T2N, T2O);
T2Q = VBYI(VADD(T2M, T2P));
T2S = VBYI(VSUB(T2P, T2M));
}
ST(&(xo[WS(os, 30)]), VSUB(T2J, T2Q), ovs, &(xo[0]));
ST(&(xo[WS(os, 14)]), VADD(T2R, T2S), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VADD(T2J, T2Q), ovs, &(xo[0]));
ST(&(xo[WS(os, 18)]), VSUB(T2R, T2S), ovs, &(xo[0]));
}
{
V T1r, T1C, T1M, T1K, T1F, T1N, T1y, T1J;
T1r = VADD(T1p, T1q);
T1C = VADD(T1A, T1B);
T1M = VSUB(T1p, T1q);
T1K = VSUB(T1B, T1A);
{
V T1D, T1E, T1u, T1x;
T1D = VFNMS(LDK(KP555570233), T1s, VMUL(LDK(KP831469612), T1t));
T1E = VFMA(LDK(KP555570233), T1v, VMUL(LDK(KP831469612), T1w));
T1F = VADD(T1D, T1E);
T1N = VSUB(T1E, T1D);
T1u = VFMA(LDK(KP831469612), T1s, VMUL(LDK(KP555570233), T1t));
T1x = VFNMS(LDK(KP555570233), T1w, VMUL(LDK(KP831469612), T1v));
T1y = VADD(T1u, T1x);
T1J = VSUB(T1x, T1u);
}
{
V T1z, T1G, T1P, T1Q;
T1z = VADD(T1r, T1y);
T1G = VBYI(VADD(T1C, T1F));
ST(&(xo[WS(os, 29)]), VSUB(T1z, T1G), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VADD(T1z, T1G), ovs, &(xo[WS(os, 1)]));
T1P = VBYI(VADD(T1K, T1J));
T1Q = VADD(T1M, T1N);
ST(&(xo[WS(os, 5)]), VADD(T1P, T1Q), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 27)]), VSUB(T1Q, T1P), ovs, &(xo[WS(os, 1)]));
}
{
V T1H, T1I, T1L, T1O;
T1H = VSUB(T1r, T1y);
T1I = VBYI(VSUB(T1F, T1C));
ST(&(xo[WS(os, 19)]), VSUB(T1H, T1I), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 13)]), VADD(T1H, T1I), ovs, &(xo[WS(os, 1)]));
T1L = VBYI(VSUB(T1J, T1K));
T1O = VSUB(T1M, T1N);
ST(&(xo[WS(os, 11)]), VADD(T1L, T1O), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 21)]), VSUB(T1O, T1L), ovs, &(xo[WS(os, 1)]));
}
}
{
V Tr, T1a, T1k, T1i, T1d, T1l, T10, T1h;
Tr = VADD(Tb, Tq);
T1a = VADD(T16, T19);
T1k = VSUB(Tb, Tq);
T1i = VSUB(T19, T16);
{
V T1b, T1c, TI, TZ;
T1b = VFNMS(LDK(KP195090322), TC, VMUL(LDK(KP980785280), TH));
T1c = VFMA(LDK(KP195090322), TT, VMUL(LDK(KP980785280), TY));
T1d = VADD(T1b, T1c);
T1l = VSUB(T1c, T1b);
TI = VFMA(LDK(KP980785280), TC, VMUL(LDK(KP195090322), TH));
TZ = VFNMS(LDK(KP195090322), TY, VMUL(LDK(KP980785280), TT));
T10 = VADD(TI, TZ);
T1h = VSUB(TZ, TI);
}
{
V T11, T1e, T1n, T1o;
T11 = VADD(Tr, T10);
T1e = VBYI(VADD(T1a, T1d));
ST(&(xo[WS(os, 31)]), VSUB(T11, T1e), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 1)]), VADD(T11, T1e), ovs, &(xo[WS(os, 1)]));
T1n = VBYI(VADD(T1i, T1h));
T1o = VADD(T1k, T1l);
ST(&(xo[WS(os, 7)]), VADD(T1n, T1o), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 25)]), VSUB(T1o, T1n), ovs, &(xo[WS(os, 1)]));
}
{
V T1f, T1g, T1j, T1m;
T1f = VSUB(Tr, T10);
T1g = VBYI(VSUB(T1d, T1a));
ST(&(xo[WS(os, 17)]), VSUB(T1f, T1g), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 15)]), VADD(T1f, T1g), ovs, &(xo[WS(os, 1)]));
T1j = VBYI(VSUB(T1h, T1i));
T1m = VSUB(T1k, T1l);
ST(&(xo[WS(os, 9)]), VADD(T1j, T1m), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 23)]), VSUB(T1m, T1j), ovs, &(xo[WS(os, 1)]));
}
}
}
}
VLEAVE();
}
static void n1fv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP980785280, +0.980785280403230449126182236134239036973933731);
DVK(KP831469612, +0.831469612302545237078788377617905756738560812);
DVK(KP198912367, +0.198912367379658006911597622644676228597850501);
DVK(KP668178637, +0.668178637919298919997757686523080761552472251);
DVK(KP923879532, +0.923879532511286756128183189396788286822416626);
DVK(KP414213562, +0.414213562373095048801688724209698078569671875);
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
INT i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) {
V T1h, Tr, T1a, T1k, TI, T1b, T1L, T1P, T1I, T1G, T1O, T1Q, T1H, T1z, T1c;
V TZ;
{
V T2x, T1T, T2K, T1W, T1p, Tb, T1A, T16, Tu, TF, T2N, T2H, T2b, T2t, TY;
V T1w, TT, T1v, T20, T2C, Tj, Te, T2h, To, T2f, T23, T2D, TB, TG, Th;
V T2i, Tk;
{
V TL, TW, TP, TQ, T2F, T27, T28, TO;
{
V T1, T2, T12, T13, T4, T5, T7, T8;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
T12 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T13 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0]));
T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0]));
T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
{
V TM, T25, T26, TN;
{
V TJ, T3, T14, T1U, T6, T1V, T9, TK, TU, TV, T1R, T1S, Ta, T15;
TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)]));
T1R = VADD(T1, T2);
T3 = VSUB(T1, T2);
T1S = VADD(T12, T13);
T14 = VSUB(T12, T13);
T1U = VADD(T4, T5);
T6 = VSUB(T4, T5);
T1V = VADD(T7, T8);
T9 = VSUB(T7, T8);
TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
TU = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)]));
T2x = VSUB(T1R, T1S);
T1T = VADD(T1R, T1S);
TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T2K = VSUB(T1V, T1U);
T1W = VADD(T1U, T1V);
Ta = VADD(T6, T9);
T15 = VSUB(T9, T6);
T25 = VADD(TJ, TK);
TL = VSUB(TJ, TK);
T26 = VADD(TV, TU);
TW = VSUB(TU, TV);
TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)]));
T1p = VFNMS(LDK(KP707106781), Ta, T3);
Tb = VFMA(LDK(KP707106781), Ta, T3);
T1A = VFMA(LDK(KP707106781), T15, T14);
T16 = VFNMS(LDK(KP707106781), T15, T14);
TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
}
T2F = VSUB(T25, T26);
T27 = VADD(T25, T26);
T28 = VADD(TM, TN);
TO = VSUB(TM, TN);
}
}
{
V Ty, T21, Tx, Tz, T1Y, T1Z;
{
V Ts, Tt, TD, T29, TR, TE, Tv, Tw;
Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
TD = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T29 = VADD(TP, TQ);
TR = VSUB(TP, TQ);
TE = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)]));
Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)]));
Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)]));
T1Y = VADD(Ts, Tt);
Tu = VSUB(Ts, Tt);
{
V T2G, T2a, TX, TS;
T2G = VSUB(T29, T28);
T2a = VADD(T28, T29);
TX = VSUB(TR, TO);
TS = VADD(TO, TR);
T1Z = VADD(TD, TE);
TF = VSUB(TD, TE);
T21 = VADD(Tv, Tw);
Tx = VSUB(Tv, Tw);
T2N = 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);
T2h = VADD(Tc, Td);
To = VSUB(Tm, Tn);
T2f = VADD(Tn, Tm);
T23 = VADD(T21, T22);
T2D = VSUB(T21, T22);
TB = VADD(Tx, TA);
TG = VSUB(Tx, TA);
Th = VSUB(Tf, Tg);
T2i = 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, T2O, T2s, T2y, T2j, T17, Ti, T2e, Tl, T2c, T2l, T24;
T1X = VSUB(T1T, T1W);
T2p = VADD(T1T, T1W);
T2E = VFNMS(LDK(KP414213562), T2D, T2C);
T2O = 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(T2h, T2i);
T2j = VADD(T2h, T2i);
T17 = VFMA(LDK(KP414213562), Te, Th);
Ti = VFNMS(LDK(KP414213562), Th, Te);
T2e = VADD(Tj, Tk);
Tl = VSUB(Tj, Tk);
T2c = VADD(T24, T2b);
T2l = VSUB(T2b, T24);
{
V T2L, T2A, T2q, T2k;
T2P = VSUB(T2N, T2O);
T2U = VADD(T2O, T2N);
{
V T2z, T2g, T18, Tp;
T2z = VSUB(T2e, T2f);
T2g = VADD(T2e, T2f);
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 = VSUB(T2t, T2s);
T2u = VADD(T2s, T2t);
T2L = VSUB(T2z, T2y);
T2A = VADD(T2y, T2z);
T2q = VADD(T2j, T2g);
T2k = VSUB(T2g, T2j);
T1q = VADD(T17, T18);
T19 = VSUB(T17, T18);
T1B = VSUB(Tp, Ti);
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 = VSUB(T2p, T2q);
T2r = VADD(T2p, T2q);
T2o = VFMA(LDK(KP707106781), T2l, T2k);
T2m = VFNMS(LDK(KP707106781), T2l, T2k);
T2X = VSUB(T2H, T2E);
T2I = VADD(T2E, T2H);
}
}
{
V T2V, T2Z, T2Y, T30, T2R, T2J;
T2V = VFNMS(LDK(KP923879532), T2U, T2T);
T2Z = VFMA(LDK(KP923879532), T2U, T2T);
ST(&(xo[WS(os, 24)]), VFNMSI(T2w, T2v), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VFMAI(T2w, T2v), ovs, &(xo[0]));
ST(&(xo[0]), VADD(T2r, T2u), ovs, &(xo[0]));
ST(&(xo[WS(os, 16)]), VSUB(T2r, T2u), ovs, &(xo[0]));
ST(&(xo[WS(os, 28)]), VFNMSI(T2o, T2n), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VFMAI(T2o, T2n), ovs, &(xo[0]));
ST(&(xo[WS(os, 20)]), VFMAI(T2m, T2d), ovs, &(xo[0]));
ST(&(xo[WS(os, 12)]), VFNMSI(T2m, T2d), 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 = VFMA(LDK(KP923879532), T1B, T1A);
T1M = VFNMS(LDK(KP923879532), T1B, T1A);
ST(&(xo[WS(os, 6)]), VFNMSI(T30, T2Z), ovs, &(xo[0]));
ST(&(xo[WS(os, 26)]), VFMAI(T30, T2Z), ovs, &(xo[0]));
ST(&(xo[WS(os, 22)]), VFNMSI(T2Y, T2V), ovs, &(xo[0]));
ST(&(xo[WS(os, 10)]), VFMAI(T2Y, T2V), 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);
ST(&(xo[WS(os, 30)]), VFNMSI(T2S, T2R), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VFMAI(T2S, T2R), ovs, &(xo[0]));
ST(&(xo[WS(os, 18)]), VFMAI(T2Q, T2J), ovs, &(xo[0]));
ST(&(xo[WS(os, 14)]), VFNMSI(T2Q, T2J), ovs, &(xo[0]));
T1K = VADD(T1D, T1E);
T1F = VSUB(T1D, T1E);
T1N = VSUB(T1x, T1u);
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 = VFMA(LDK(KP831469612), T1N, T1M);
T1Q = VFNMS(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;
ST(&(xo[WS(os, 21)]), VFNMSI(T1O, T1L), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 11)]), VFMAI(T1O, T1L), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 27)]), VFMAI(T1Q, T1P), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 5)]), VFNMSI(T1Q, T1P), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VFMAI(T1I, T1H), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 29)]), VFNMSI(T1I, T1H), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 19)]), VFMAI(T1G, T1z), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 13)]), VFNMSI(T1G, T1z), ovs, &(xo[WS(os, 1)]));
T1d = VSUB(T1b, T1c);
T1i = VADD(T1b, T1c);
T10 = VADD(TI, TZ);
T1l = VSUB(TZ, TI);
{
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 = VFMA(LDK(KP980785280), T1l, T1k);
T1m = VFNMS(LDK(KP980785280), T1l, T1k);
T11 = VFNMS(LDK(KP980785280), T10, Tr);
T1f = VFMA(LDK(KP980785280), T10, Tr);
ST(&(xo[WS(os, 23)]), VFMAI(T1m, T1j), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VFNMSI(T1m, T1j), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 25)]), VFNMSI(T1o, T1n), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VFMAI(T1o, T1n), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 31)]), VFMAI(T1g, T1f), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 1)]), VFNMSI(T1g, T1f), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 15)]), VFMAI(T1e, T11), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 17)]), VFNMSI(T1e, T11), ovs, &(xo[WS(os, 1)]));
}
}
}
}
VLEAVE();
}
static void q1fv_4(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms)
{
INT m;
R *x;
x = ri;
for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(rs), MAKE_VOLATILE_STRIDE(vs)) {
V Tb, Tm, Tx, TI;
{
V Tc, T9, T3, TG, TA, TH, TD, Ta, T6, Td, Tn, To, Tq, Tr, Tf;
V Tg;
{
V T1, T2, Ty, Tz, TB, TC, T4, T5;
T1 = LD(&(x[0]), ms, &(x[0]));
T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Ty = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)]));
Tz = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)]));
TB = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)]));
TC = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)]));
T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Tc = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)]));
T9 = VADD(T1, T2);
T3 = VSUB(T1, T2);
TG = VADD(Ty, Tz);
TA = VSUB(Ty, Tz);
TH = VADD(TB, TC);
TD = VSUB(TB, TC);
Ta = VADD(T4, T5);
T6 = VSUB(T4, T5);
Td = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)]));
Tn = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)]));
To = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)]));
Tq = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)]));
Tr = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)]));
Tf = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)]));
Tg = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)]));
}
{
V Tk, Te, Tv, Tp, Tw, Ts, Tl, Th, T7, TE, Tu, TF;
ST(&(x[0]), VADD(T9, Ta), ms, &(x[0]));
Tk = VADD(Tc, Td);
Te = VSUB(Tc, Td);
Tv = VADD(Tn, To);
Tp = VSUB(Tn, To);
Tw = VADD(Tq, Tr);
Ts = VSUB(Tq, Tr);
Tl = VADD(Tf, Tg);
Th = VSUB(Tf, Tg);
ST(&(x[WS(rs, 3)]), VADD(TG, TH), ms, &(x[WS(rs, 1)]));
T7 = BYTWJ(&(W[0]), VFNMSI(T6, T3));
TE = BYTWJ(&(W[0]), VFNMSI(TD, TA));
{
V Tt, Ti, Tj, T8;
T8 = BYTWJ(&(W[TWVL * 4]), VFMAI(T6, T3));
ST(&(x[WS(rs, 2)]), VADD(Tv, Tw), ms, &(x[0]));
Tt = BYTWJ(&(W[0]), VFNMSI(Ts, Tp));
ST(&(x[WS(rs, 1)]), VADD(Tk, Tl), ms, &(x[WS(rs, 1)]));
Ti = BYTWJ(&(W[0]), VFNMSI(Th, Te));
Tj = BYTWJ(&(W[TWVL * 4]), VFMAI(Th, Te));
ST(&(x[WS(vs, 1)]), T7, ms, &(x[WS(vs, 1)]));
ST(&(x[WS(vs, 1) + WS(rs, 3)]), TE, ms, &(x[WS(vs, 1) + WS(rs, 1)]));
ST(&(x[WS(vs, 3)]), T8, ms, &(x[WS(vs, 3)]));
Tu = BYTWJ(&(W[TWVL * 4]), VFMAI(Ts, Tp));
ST(&(x[WS(vs, 1) + WS(rs, 2)]), Tt, ms, &(x[WS(vs, 1)]));
TF = BYTWJ(&(W[TWVL * 4]), VFMAI(TD, TA));
ST(&(x[WS(vs, 1) + WS(rs, 1)]), Ti, ms, &(x[WS(vs, 1) + WS(rs, 1)]));
ST(&(x[WS(vs, 3) + WS(rs, 1)]), Tj, ms, &(x[WS(vs, 3) + WS(rs, 1)]));
}
Tb = BYTWJ(&(W[TWVL * 2]), VSUB(T9, Ta));
Tm = BYTWJ(&(W[TWVL * 2]), VSUB(Tk, Tl));
Tx = BYTWJ(&(W[TWVL * 2]), VSUB(Tv, Tw));
ST(&(x[WS(vs, 3) + WS(rs, 2)]), Tu, ms, &(x[WS(vs, 3)]));
TI = BYTWJ(&(W[TWVL * 2]), VSUB(TG, TH));
ST(&(x[WS(vs, 3) + WS(rs, 3)]), TF, ms, &(x[WS(vs, 3) + WS(rs, 1)]));
}
}
ST(&(x[WS(vs, 2)]), Tb, ms, &(x[WS(vs, 2)]));
ST(&(x[WS(vs, 2) + WS(rs, 1)]), Tm, ms, &(x[WS(vs, 2) + WS(rs, 1)]));
ST(&(x[WS(vs, 2) + WS(rs, 2)]), Tx, ms, &(x[WS(vs, 2)]));
ST(&(x[WS(vs, 2) + WS(rs, 3)]), TI, ms, &(x[WS(vs, 2) + WS(rs, 1)]));
}
}
static void t1bv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
DVK(KP907603734, +0.907603734547952313649323976213898122064543220);
DVK(KP666666666, +0.666666666666666666666666666666666666666666667);
DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
DVK(KP879385241, +0.879385241571816768108218554649462939872416269);
DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
DVK(KP826351822, +0.826351822333069651148283373230685203999624323);
DVK(KP347296355, +0.347296355333860697703433253538629592000751354);
DVK(KP898197570, +0.898197570222573798468955502359086394667167570);
DVK(KP673648177, +0.673648177666930348851716626769314796000375677);
DVK(KP420276625, +0.420276625461206169731530603237061658838781920);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP586256827, +0.586256827714544512072145703099641959914944179);
DVK(KP968908795, +0.968908795874236621082202410917456709164223497);
DVK(KP726681596, +0.726681596905677465811651808188092531873167623);
DVK(KP439692620, +0.439692620785908384054109277324731469936208134);
DVK(KP203604859, +0.203604859554852403062088995281827210665664861);
DVK(KP152703644, +0.152703644666139302296566746461370407999248646);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
INT m;
R *x;
x = ii;
for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(rs)) {
V T1, T3, T5, T9, Tn, Tb, Td, Th, Tj, Tx, T6;
T1 = LD(&(x[0]), ms, &(x[0]));
{
V T2, T4, T8, Tm;
T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
T8 = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
{
V Ta, Tc, Tg, Ti;
Ta = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tc = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Ti = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
T3 = BYTW(&(W[TWVL * 4]), T2);
T5 = BYTW(&(W[TWVL * 10]), T4);
T9 = BYTW(&(W[TWVL * 2]), T8);
Tn = BYTW(&(W[0]), Tm);
Tb = BYTW(&(W[TWVL * 8]), Ta);
Td = BYTW(&(W[TWVL * 14]), Tc);
Th = BYTW(&(W[TWVL * 6]), Tg);
Tj = BYTW(&(W[TWVL * 12]), Ti);
}
}
Tx = VSUB(T3, T5);
T6 = VADD(T3, T5);
{
V Tl, Te, Tk, To, T7, TN;
Tl = VSUB(Td, Tb);
Te = VADD(Tb, Td);
Tk = VSUB(Th, Tj);
To = VADD(Th, Tj);
T7 = VFNMS(LDK(KP500000000), T6, T1);
TN = VADD(T1, T6);
{
V Tf, TP, Tp, TO;
Tf = VFNMS(LDK(KP500000000), Te, T9);
TP = VADD(T9, Te);
Tp = VFNMS(LDK(KP500000000), To, Tn);
TO = VADD(Tn, To);
{
V Tz, TC, Tu, TD, TA, Tq, TQ, TS;
Tz = VFNMS(LDK(KP152703644), Tl, Tf);
TC = VFMA(LDK(KP203604859), Tf, Tl);
Tu = VFNMS(LDK(KP439692620), Tk, Tf);
TD = VFNMS(LDK(KP726681596), Tk, Tp);
TA = VFMA(LDK(KP968908795), Tp, Tk);
Tq = VFNMS(LDK(KP586256827), Tp, Tl);
TQ = VADD(TO, TP);
TS = VMUL(LDK(KP866025403), VSUB(TO, TP));
{
V TI, TB, TH, TE, Tr, TR, Tw, Tv;
Tv = VFNMS(LDK(KP420276625), Tu, Tl);
TI = VFMA(LDK(KP673648177), TA, Tz);
TB = VFNMS(LDK(KP673648177), TA, Tz);
TH = VFNMS(LDK(KP898197570), TD, TC);
TE = VFMA(LDK(KP898197570), TD, TC);
Tr = VFNMS(LDK(KP347296355), Tq, Tk);
ST(&(x[0]), VADD(TQ, TN), ms, &(x[0]));
TR = VFNMS(LDK(KP500000000), TQ, TN);
Tw = VFNMS(LDK(KP826351822), Tv, Tp);
{
V TM, TL, TF, TJ, Ts, Ty, TG, TK, Tt;
TM = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tx, TI));
TL = VFMA(LDK(KP852868531), TE, T7);
TF = VFNMS(LDK(KP500000000), TE, TB);
TJ = VFMA(LDK(KP666666666), TI, TH);
Ts = VFNMS(LDK(KP907603734), Tr, Tf);
ST(&(x[WS(rs, 6)]), VFNMSI(TS, TR), ms, &(x[0]));
ST(&(x[WS(rs, 3)]), VFMAI(TS, TR), ms, &(x[WS(rs, 1)]));
Ty = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tx, Tw));
ST(&(x[WS(rs, 8)]), VFNMSI(TM, TL), ms, &(x[0]));
ST(&(x[WS(rs, 1)]), VFMAI(TM, TL), ms, &(x[WS(rs, 1)]));
TG = VFMA(LDK(KP852868531), TF, T7);
TK = VMUL(LDK(KP866025403), VFNMS(LDK(KP852868531), TJ, Tx));
Tt = VFNMS(LDK(KP939692620), Ts, T7);
ST(&(x[WS(rs, 5)]), VFNMSI(TK, TG), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 4)]), VFMAI(TK, TG), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(Ty, Tt), ms, &(x[0]));
ST(&(x[WS(rs, 7)]), VFNMSI(Ty, Tt), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
}
static void t1bv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP939692620, +0.939692620785908384054109277324731469936208134);
DVK(KP296198132, +0.296198132726023843175338011893050938967728390);
DVK(KP852868531, +0.852868531952443209628250963940074071936020296);
DVK(KP173648177, +0.173648177666930348851716626769314796000375677);
DVK(KP556670399, +0.556670399226419366452912952047023132968291906);
DVK(KP766044443, +0.766044443118978035202392650555416673935832457);
DVK(KP642787609, +0.642787609686539326322643409907263432907559884);
DVK(KP663413948, +0.663413948168938396205421319635891297216863310);
DVK(KP150383733, +0.150383733180435296639271897612501926072238258);
DVK(KP342020143, +0.342020143325668733044099614682259580763083368);
DVK(KP813797681, +0.813797681349373692844693217248393223289101568);
DVK(KP984807753, +0.984807753012208059366743024589523013670643252);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
INT m;
R *x;
x = ii;
for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(rs)) {
V T1, T6, Tu, Tg, Tf, TD, Tq, Tp, TE;
T1 = LD(&(x[0]), ms, &(x[0]));
{
V T3, T5, T2, T4;
T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
T3 = BYTW(&(W[TWVL * 4]), T2);
T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
T5 = BYTW(&(W[TWVL * 10]), T4);
T6 = VADD(T3, T5);
Tu = VMUL(LDK(KP866025403), VSUB(T3, T5));
}
{
V T9, Td, Tb, T8, Tc, Ta, Te;
T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
T9 = BYTW(&(W[0]), T8);
Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Td = BYTW(&(W[TWVL * 12]), Tc);
Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
Tb = BYTW(&(W[TWVL * 6]), Ta);
Tg = VSUB(Tb, Td);
Te = VADD(Tb, Td);
Tf = VFNMS(LDK(KP500000000), Te, T9);
TD = VADD(T9, Te);
}
{
V Tj, Tn, Tl, Ti, Tm, Tk, To;
Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tj = BYTW(&(W[TWVL * 2]), Ti);
Tm = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Tn = BYTW(&(W[TWVL * 14]), Tm);
Tk = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tl = BYTW(&(W[TWVL * 8]), Tk);
Tq = VSUB(Tl, Tn);
To = VADD(Tl, Tn);
Tp = VFNMS(LDK(KP500000000), To, Tj);
TE = VADD(Tj, To);
}
{
V TF, TG, TH, TI;
TF = VBYI(VMUL(LDK(KP866025403), VSUB(TD, TE)));
TG = VADD(T1, T6);
TH = VADD(TD, TE);
TI = VFNMS(LDK(KP500000000), TH, TG);
ST(&(x[WS(rs, 3)]), VADD(TF, TI), ms, &(x[WS(rs, 1)]));
ST(&(x[0]), VADD(TG, TH), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VSUB(TI, TF), ms, &(x[0]));
}
{
V TC, Tv, Tw, Tx, Th, Tr, Ts, T7, TB;
TC = VBYI(VSUB(VFMA(LDK(KP984807753), Tf, VFMA(LDK(KP813797681), Tq, VFNMS(LDK(KP150383733), Tg, VMUL(LDK(KP342020143), Tp)))), Tu));
Tv = VFMA(LDK(KP663413948), Tg, VMUL(LDK(KP642787609), Tf));
Tw = VFMA(LDK(KP150383733), Tq, VMUL(LDK(KP984807753), Tp));
Tx = VADD(Tv, Tw);
Th = VFNMS(LDK(KP556670399), Tg, VMUL(LDK(KP766044443), Tf));
Tr = VFNMS(LDK(KP852868531), Tq, VMUL(LDK(KP173648177), Tp));
Ts = VADD(Th, Tr);
T7 = VFNMS(LDK(KP500000000), T6, T1);
TB = VFMA(LDK(KP852868531), Tg, VFMA(LDK(KP173648177), Tf, VFMA(LDK(KP296198132), Tq, VFNMS(LDK(KP939692620), Tp, T7))));
ST(&(x[WS(rs, 7)]), VSUB(TB, TC), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 2)]), VADD(TB, TC), ms, &(x[0]));
{
V Tt, Ty, Tz, TA;
Tt = VADD(T7, Ts);
Ty = VBYI(VADD(Tu, Tx));
ST(&(x[WS(rs, 8)]), VSUB(Tt, Ty), ms, &(x[0]));
ST(&(x[WS(rs, 1)]), VADD(Tt, Ty), ms, &(x[WS(rs, 1)]));
Tz = VBYI(VADD(Tu, VFNMS(LDK(KP500000000), Tx, VMUL(LDK(KP866025403), VSUB(Th, Tr)))));
TA = VFMA(LDK(KP866025403), VSUB(Tw, Tv), VFNMS(LDK(KP500000000), Ts, T7));
ST(&(x[WS(rs, 4)]), VADD(Tz, TA), ms, &(x[0]));
ST(&(x[WS(rs, 5)]), VSUB(TA, Tz), ms, &(x[WS(rs, 1)]));
}
}
}
}
static void n1bv_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP792626838, +0.792626838241819413632131824093538848057784557);
DVK(KP876091699, +0.876091699473550838204498029706869638173524346);
DVK(KP803003575, +0.803003575438660414833440593570376004635464850);
DVK(KP617882369, +0.617882369114440893914546919006756321695042882);
DVK(KP242145790, +0.242145790282157779872542093866183953459003101);
DVK(KP968583161, +0.968583161128631119490168375464735813836012403);
DVK(KP999544308, +0.999544308746292983948881682379742149196758193);
DVK(KP683113946, +0.683113946453479238701949862233725244439656928);
DVK(KP559154169, +0.559154169276087864842202529084232643714075927);
DVK(KP904730450, +0.904730450839922351881287709692877908104763647);
DVK(KP829049696, +0.829049696159252993975487806364305442437946767);
DVK(KP831864738, +0.831864738706457140726048799369896829771167132);
DVK(KP916574801, +0.916574801383451584742370439148878693530976769);
DVK(KP894834959, +0.894834959464455102997960030820114611498661386);
DVK(KP809385824, +0.809385824416008241660603814668679683846476688);
DVK(KP447417479, +0.447417479732227551498980015410057305749330693);
DVK(KP860541664, +0.860541664367944677098261680920518816412804187);
DVK(KP897376177, +0.897376177523557693138608077137219684419427330);
DVK(KP876306680, +0.876306680043863587308115903922062583399064238);
DVK(KP681693190, +0.681693190061530575150324149145440022633095390);
DVK(KP560319534, +0.560319534973832390111614715371676131169633784);
DVK(KP855719849, +0.855719849902058969314654733608091555096772472);
DVK(KP237294955, +0.237294955877110315393888866460840817927895961);
DVK(KP949179823, +0.949179823508441261575555465843363271711583843);
DVK(KP904508497, +0.904508497187473712051146708591409529430077295);
DVK(KP997675361, +0.997675361079556513670859573984492383596555031);
DVK(KP262346850, +0.262346850930607871785420028382979691334784273);
DVK(KP763932022, +0.763932022500210303590826331268723764559381640);
DVK(KP992114701, +0.992114701314477831049793042785778521453036709);
DVK(KP690983005, +0.690983005625052575897706582817180941139845410);
DVK(KP952936919, +0.952936919628306576880750665357914584765951388);
DVK(KP998026728, +0.998026728428271561952336806863450553336905220);
DVK(KP570584518, +0.570584518783621657366766175430996792655723863);
DVK(KP669429328, +0.669429328479476605641803240971985825917022098);
DVK(KP923225144, +0.923225144846402650453449441572664695995209956);
DVK(KP906616052, +0.906616052148196230441134447086066874408359177);
DVK(KP956723877, +0.956723877038460305821989399535483155872969262);
DVK(KP522616830, +0.522616830205754336872861364785224694908468440);
DVK(KP945422727, +0.945422727388575946270360266328811958657216298);
DVK(KP912575812, +0.912575812670962425556968549836277086778922727);
DVK(KP982009705, +0.982009705009746369461829878184175962711969869);
DVK(KP921078979, +0.921078979742360627699756128143719920817673854);
DVK(KP734762448, +0.734762448793050413546343770063151342619912334);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP958953096, +0.958953096729998668045963838399037225970891871);
DVK(KP867381224, +0.867381224396525206773171885031575671309956167);
DVK(KP269969613, +0.269969613759572083574752974412347470060951301);
DVK(KP244189809, +0.244189809627953270309879511234821255780225091);
DVK(KP845997307, +0.845997307939530944175097360758058292389769300);
DVK(KP772036680, +0.772036680810363904029489473607579825330539880);
DVK(KP132830569, +0.132830569247582714407653942074819768844536507);
DVK(KP120146378, +0.120146378570687701782758537356596213647956445);
DVK(KP987388751, +0.987388751065621252324603216482382109400433949);
DVK(KP893101515, +0.893101515366181661711202267938416198338079437);
DVK(KP786782374, +0.786782374965295178365099601674911834788448471);
DVK(KP869845200, +0.869845200362138853122720822420327157933056305);
DVK(KP447533225, +0.447533225982656890041886979663652563063114397);
DVK(KP494780565, +0.494780565770515410344588413655324772219443730);
DVK(KP578046249, +0.578046249379945007321754579646815604023525655);
DVK(KP522847744, +0.522847744331509716623755382187077770911012542);
DVK(KP059835404, +0.059835404262124915169548397419498386427871950);
DVK(KP066152395, +0.066152395967733048213034281011006031460903353);
DVK(KP603558818, +0.603558818296015001454675132653458027918768137);
DVK(KP667278218, +0.667278218140296670899089292254759909713898805);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
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 T1g, T1k, T1I, T24, T2a, T1G, T1A, T1l, T1B, T1H, T1d;
{
V T2z, T1q, Ta, T9, T3n, Ty, Tl, T2O, T2W, T2l, T2s, TV, T1i, T1K, T1S;
V T3z, T3t, Tk, T3o, Tp, T2g, T2N, T2V, T2o, T2t, T1a, T1j, T1J, T1R, Tz;
V Tt, TA, Tw;
{
V T1, T5, T6, T2, T3;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0]));
{
V TH, TW, TK, TS, T10, T8, TN, TT, T17, TZ, T11;
TH = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
TW = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
{
V TI, TJ, TL, T7, T1p, T4, T1o, TM, TX, TY;
TI = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
TJ = LD(&(xi[WS(is, 22)]), ivs, &(xi[0]));
TL = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
T7 = VADD(T5, T6);
T1p = VSUB(T5, T6);
T4 = VADD(T2, T3);
T1o = VSUB(T2, T3);
TM = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
TX = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
TK = VADD(TI, TJ);
TS = VSUB(TI, TJ);
TY = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)]));
T10 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
T2z = VFNMS(LDK(KP618033988), T1o, T1p);
T1q = VFMA(LDK(KP618033988), T1p, T1o);
Ta = VSUB(T4, T7);
T8 = VADD(T4, T7);
TN = VADD(TL, TM);
TT = VSUB(TM, TL);
T17 = VSUB(TX, TY);
TZ = VADD(TX, TY);
T11 = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
}
{
V Tc, T2m, T19, Tn, To, Tr, Tj, T16, T2n, Ts, Tu, Tv;
{
V TU, T2j, TO, TQ, T12, T18;
Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T9 = VFNMS(LDK(KP250000000), T8, T1);
T3n = VADD(T1, T8);
TU = VFNMS(LDK(KP618033988), TT, TS);
T2j = VFMA(LDK(KP618033988), TS, TT);
TO = VADD(TK, TN);
TQ = VSUB(TN, TK);
T12 = VADD(T10, T11);
T18 = VSUB(T10, T11);
Ty = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
{
V T3r, T15, T13, Tf, Ti, T2k, TR, TP, T3s, T14;
{
V Td, Te, Tg, Th;
Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Te = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)]));
Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Th = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
TP = VFNMS(LDK(KP250000000), TO, TH);
T3r = VADD(TH, TO);
T2m = VFNMS(LDK(KP618033988), T17, T18);
T19 = VFMA(LDK(KP618033988), T18, T17);
T15 = VSUB(T12, TZ);
T13 = VADD(TZ, T12);
Tf = VADD(Td, Te);
Tn = VSUB(Td, Te);
To = VSUB(Th, Tg);
Ti = VADD(Tg, Th);
}
T2k = VFMA(LDK(KP559016994), TQ, TP);
TR = VFNMS(LDK(KP559016994), TQ, TP);
Tr = LD(&(xi[WS(is, 24)]), ivs, &(xi[0]));
T3s = VADD(TW, T13);
T14 = VFNMS(LDK(KP250000000), T13, TW);
Tj = VADD(Tf, Ti);
Tl = VSUB(Tf, Ti);
T2O = VFNMS(LDK(KP667278218), T2k, T2j);
T2W = VFMA(LDK(KP603558818), T2j, T2k);
T2l = VFMA(LDK(KP066152395), T2k, T2j);
T2s = VFNMS(LDK(KP059835404), T2j, T2k);
TV = VFNMS(LDK(KP522847744), TU, TR);
T1i = VFMA(LDK(KP578046249), TR, TU);
T1K = VFNMS(LDK(KP494780565), TR, TU);
T1S = VFMA(LDK(KP447533225), TU, TR);
T16 = VFNMS(LDK(KP559016994), T15, T14);
T2n = VFMA(LDK(KP559016994), T15, T14);
T3z = VSUB(T3r, T3s);
T3t = VADD(T3r, T3s);
Ts = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tu = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
Tv = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
}
}
Tk = VFNMS(LDK(KP250000000), Tj, Tc);
T3o = VADD(Tc, Tj);
Tp = VFNMS(LDK(KP618033988), To, Tn);
T2g = VFMA(LDK(KP618033988), Tn, To);
T2N = VFMA(LDK(KP066152395), T2n, T2m);
T2V = VFNMS(LDK(KP059835404), T2m, T2n);
T2o = VFMA(LDK(KP869845200), T2n, T2m);
T2t = VFNMS(LDK(KP786782374), T2m, T2n);
T1a = VFNMS(LDK(KP893101515), T19, T16);
T1j = VFMA(LDK(KP987388751), T16, T19);
T1J = VFNMS(LDK(KP120146378), T19, T16);
T1R = VFMA(LDK(KP132830569), T16, T19);
Tz = VADD(Ts, Tr);
Tt = VSUB(Tr, Ts);
TA = VADD(Tv, Tu);
Tw = VSUB(Tu, Tv);
}
}
}
{
V T2p, T2I, T2u, T2C, Tx, T2d, T2X, T34, T2P, T3b, T2b, Tb, T2Q, T2Z, T2h;
V T2w, Tq, T1e, T1M, T1U, TE, T2c, T3q, T3y;
T2p = VFNMS(LDK(KP772036680), T2o, T2l);
T2I = VFMA(LDK(KP772036680), T2o, T2l);
T2u = VFMA(LDK(KP772036680), T2t, T2s);
T2C = VFNMS(LDK(KP772036680), T2t, T2s);
{
V TD, TB, Tm, T2f, T3p, TC;
Tx = VFMA(LDK(KP618033988), Tw, Tt);
T2d = VFNMS(LDK(KP618033988), Tt, Tw);
TD = VSUB(Tz, TA);
TB = VADD(Tz, TA);
Tm = VFMA(LDK(KP559016994), Tl, Tk);
T2f = VFNMS(LDK(KP559016994), Tl, Tk);
T2X = VFMA(LDK(KP845997307), T2W, T2V);
T34 = VFNMS(LDK(KP845997307), T2W, T2V);
T2P = VFNMS(LDK(KP845997307), T2O, T2N);
T3b = VFMA(LDK(KP845997307), T2O, T2N);
T2b = VFNMS(LDK(KP559016994), Ta, T9);
Tb = VFMA(LDK(KP559016994), Ta, T9);
T3p = VADD(Ty, TB);
TC = VFMS(LDK(KP250000000), TB, Ty);
T2Q = VFNMS(LDK(KP522847744), T2g, T2f);
T2Z = VFMA(LDK(KP578046249), T2f, T2g);
T2h = VFMA(LDK(KP893101515), T2g, T2f);
T2w = VFNMS(LDK(KP987388751), T2f, T2g);
Tq = VFNMS(LDK(KP244189809), Tp, Tm);
T1e = VFMA(LDK(KP269969613), Tm, Tp);
T1M = VFMA(LDK(KP667278218), Tm, Tp);
T1U = VFNMS(LDK(KP603558818), Tp, Tm);
TE = VFNMS(LDK(KP559016994), TD, TC);
T2c = VFMA(LDK(KP559016994), TD, TC);
T3q = VADD(T3o, T3p);
T3y = VSUB(T3o, T3p);
}
{
V T1Z, T25, T1P, T22, T1X, TG, T1b, T28, T1t, T1y, T1x, T1E, T1Q, T1Y;
{
V T26, T1L, T1T, TF, T1f, T1W, T3m, T3g, T2M, T2G, T39, T3j, T21, T1O, T20;
V T27;
T26 = VFMA(LDK(KP867381224), T1K, T1J);
T1L = VFNMS(LDK(KP867381224), T1K, T1J);
T20 = VFNMS(LDK(KP958953096), T1S, T1R);
T1T = VFMA(LDK(KP958953096), T1S, T1R);
{
V T2R, T2Y, T2e, T2v, T1N, T1V;
T2R = VFNMS(LDK(KP494780565), T2c, T2d);
T2Y = VFMA(LDK(KP447533225), T2d, T2c);
T2e = VFMA(LDK(KP120146378), T2d, T2c);
T2v = VFNMS(LDK(KP132830569), T2c, T2d);
TF = VFNMS(LDK(KP667278218), TE, Tx);
T1f = VFMA(LDK(KP603558818), Tx, TE);
T1N = VFMA(LDK(KP869845200), TE, Tx);
T1V = VFNMS(LDK(KP786782374), Tx, TE);
{
V T3A, T3C, T3w, T3u;
T3A = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T3z, T3y));
T3C = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T3y, T3z));
T3w = VSUB(T3q, T3t);
T3u = VADD(T3q, T3t);
{
V T2B, T2x, T2H, T2i;
T2B = VFMA(LDK(KP734762448), T2w, T2v);
T2x = VFNMS(LDK(KP734762448), T2w, T2v);
T2H = VFNMS(LDK(KP734762448), T2h, T2e);
T2i = VFMA(LDK(KP734762448), T2h, T2e);
{
V T30, T35, T3c, T2S, T3v;
T30 = VFNMS(LDK(KP921078979), T2Z, T2Y);
T35 = VFMA(LDK(KP921078979), T2Z, T2Y);
T3c = VFMA(LDK(KP982009705), T2R, T2Q);
T2S = VFNMS(LDK(KP982009705), T2R, T2Q);
T1W = VFMA(LDK(KP912575812), T1V, T1U);
T1Z = VFNMS(LDK(KP912575812), T1V, T1U);
T1O = VFMA(LDK(KP912575812), T1N, T1M);
T25 = VFNMS(LDK(KP912575812), T1N, T1M);
ST(&(xo[0]), VADD(T3u, T3n), ovs, &(xo[0]));
T3v = VFNMS(LDK(KP250000000), T3u, T3n);
{
V T2y, T2J, T2q, T2D;
T2y = VFMA(LDK(KP945422727), T2x, T2u);
T2J = VFMA(LDK(KP522616830), T2x, T2I);
T2q = VFMA(LDK(KP956723877), T2p, T2i);
T2D = VFNMS(LDK(KP522616830), T2i, T2C);
{
V T3e, T31, T36, T2T;
T3e = VFMA(LDK(KP906616052), T30, T2X);
T31 = VFNMS(LDK(KP906616052), T30, T2X);
T36 = VFNMS(LDK(KP923225144), T2S, T2P);
T2T = VFMA(LDK(KP923225144), T2S, T2P);
{
V T3k, T3d, T3x, T3B;
T3k = VFNMS(LDK(KP669429328), T3b, T3c);
T3d = VFMA(LDK(KP570584518), T3c, T3b);
T3x = VFMA(LDK(KP559016994), T3w, T3v);
T3B = VFNMS(LDK(KP559016994), T3w, T3v);
{
V T2A, T2K, T2r, T2E;
T2A = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T2z, T2y));
T2K = VFNMS(LDK(KP690983005), T2J, T2u);
T2r = VFMA(LDK(KP992114701), T2q, T2b);
T2E = VFMA(LDK(KP763932022), T2D, T2p);
{
V T32, T3a, T37, T3h;
T32 = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T2z, T31));
T3a = VFMA(LDK(KP262346850), T31, T2z);
T37 = VFNMS(LDK(KP997675361), T36, T35);
T3h = VFNMS(LDK(KP904508497), T36, T34);
{
V T2U, T33, T3l, T3f;
T2U = VFMA(LDK(KP949179823), T2T, T2b);
T33 = VFNMS(LDK(KP237294955), T2T, T2b);
T3l = VFNMS(LDK(KP669429328), T3e, T3k);
T3f = VFMA(LDK(KP618033988), T3e, T3d);
ST(&(xo[WS(os, 20)]), VFNMSI(T3A, T3x), ovs, &(xo[0]));
ST(&(xo[WS(os, 5)]), VFMAI(T3A, T3x), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 15)]), VFMAI(T3C, T3B), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 10)]), VFNMSI(T3C, T3B), ovs, &(xo[0]));
{
V T2L, T2F, T38, T3i;
T2L = VFMA(LDK(KP855719849), T2K, T2H);
ST(&(xo[WS(os, 3)]), VFMAI(T2A, T2r), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 22)]), VFNMSI(T2A, T2r), ovs, &(xo[0]));
T2F = VFNMS(LDK(KP855719849), T2E, T2B);
T38 = VFMA(LDK(KP560319534), T37, T34);
T3i = VFNMS(LDK(KP681693190), T3h, T35);
ST(&(xo[WS(os, 2)]), VFMAI(T32, T2U), ovs, &(xo[0]));
ST(&(xo[WS(os, 23)]), VFNMSI(T32, T2U), ovs, &(xo[WS(os, 1)]));
T3m = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T3l, T3a));
T3g = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T3f, T3a));
T2M = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2L, T2z));
T2G = VFMA(LDK(KP897376177), T2F, T2b);
T39 = VFNMS(LDK(KP949179823), T38, T33);
T3j = VFNMS(LDK(KP860541664), T3i, T33);
T21 = VFMA(LDK(KP447417479), T1O, T20);
}
}
}
}
}
}
}
}
}
}
}
T1P = VFNMS(LDK(KP809385824), T1O, T1L);
ST(&(xo[WS(os, 17)]), VFNMSI(T2M, T2G), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 8)]), VFMAI(T2M, T2G), ovs, &(xo[0]));
ST(&(xo[WS(os, 13)]), VFMAI(T3g, T39), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 12)]), VFNMSI(T3g, T39), ovs, &(xo[0]));
ST(&(xo[WS(os, 18)]), VFMAI(T3m, T3j), ovs, &(xo[0]));
ST(&(xo[WS(os, 7)]), VFNMSI(T3m, T3j), ovs, &(xo[WS(os, 1)]));
T22 = VFMA(LDK(KP690983005), T21, T1L);
T27 = VFMA(LDK(KP447417479), T1W, T26);
T1X = VFMA(LDK(KP894834959), T1W, T1T);
{
V T1r, T1s, T1v, T1w;
T1r = VFNMS(LDK(KP916574801), T1f, T1e);
T1g = VFMA(LDK(KP916574801), T1f, T1e);
T1k = VFNMS(LDK(KP831864738), T1j, T1i);
T1s = VFMA(LDK(KP831864738), T1j, T1i);
T1v = VFNMS(LDK(KP829049696), TF, Tq);
TG = VFMA(LDK(KP829049696), TF, Tq);
T1b = VFMA(LDK(KP831864738), T1a, TV);
T1w = VFNMS(LDK(KP831864738), T1a, TV);
T28 = VFNMS(LDK(KP763932022), T27, T1T);
T1t = VFMA(LDK(KP904730450), T1s, T1r);
T1y = VFNMS(LDK(KP904730450), T1s, T1r);
T1x = VFMA(LDK(KP559154169), T1w, T1v);
T1E = VFNMS(LDK(KP683113946), T1v, T1w);
}
}
T1Q = VFNMS(LDK(KP992114701), T1P, Tb);
T1Y = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T1X, T1q));
{
V T1u, T1F, T1z, T1h, T1c, T23, T29;
T23 = VFNMS(LDK(KP999544308), T22, T1Z);
T29 = VFNMS(LDK(KP999544308), T28, T25);
T1I = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1t, T1q));
T1u = VFNMS(LDK(KP242145790), T1t, T1q);
T1F = VFMA(LDK(KP617882369), T1y, T1E);
T1z = VFMA(LDK(KP559016994), T1y, T1x);
T1h = VFNMS(LDK(KP904730450), T1b, TG);
T1c = VFMA(LDK(KP904730450), T1b, TG);
ST(&(xo[WS(os, 21)]), VFMAI(T1Y, T1Q), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VFNMSI(T1Y, T1Q), ovs, &(xo[0]));
T24 = VFNMS(LDK(KP803003575), T23, Tb);
T2a = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T29, T1q));
T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T1F, T1u));
T1A = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1z, T1u));
T1l = VFNMS(LDK(KP904730450), T1k, T1h);
T1B = VADD(T1g, T1h);
T1H = VFMA(LDK(KP968583161), T1c, Tb);
T1d = VFNMS(LDK(KP242145790), T1c, Tb);
}
}
}
}
ST(&(xo[WS(os, 16)]), VFMAI(T2a, T24), ovs, &(xo[0]));
ST(&(xo[WS(os, 9)]), VFNMSI(T2a, T24), ovs, &(xo[WS(os, 1)]));
{
V T1m, T1C, T1n, T1D;
T1m = VFNMS(LDK(KP618033988), T1l, T1g);
T1C = VFNMS(LDK(KP683113946), T1B, T1k);
ST(&(xo[WS(os, 24)]), VFNMSI(T1I, T1H), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFMAI(T1I, T1H), ovs, &(xo[WS(os, 1)]));
T1n = VFNMS(LDK(KP876091699), T1m, T1d);
T1D = VFMA(LDK(KP792626838), T1C, T1d);
ST(&(xo[WS(os, 19)]), VFNMSI(T1A, T1n), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 6)]), VFMAI(T1A, T1n), ovs, &(xo[0]));
ST(&(xo[WS(os, 14)]), VFNMSI(T1G, T1D), ovs, &(xo[0]));
ST(&(xo[WS(os, 11)]), VFMAI(T1G, T1D), ovs, &(xo[WS(os, 1)]));
}
}
}
static void t2fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
INT m;
R *x;
x = ri;
for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(rs)) {
V T4, Tq, Tm, Tr, T9, Tt, Te, Tu, T1, T3, T2;
T1 = LD(&(x[0]), ms, &(x[0]));
T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
T3 = BYTWJ(&(W[TWVL * 6]), T2);
T4 = VSUB(T1, T3);
Tq = VADD(T1, T3);
{
V Tj, Tl, Ti, Tk;
Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Tj = BYTWJ(&(W[TWVL * 2]), Ti);
Tk = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tl = BYTWJ(&(W[TWVL * 10]), Tk);
Tm = VSUB(Tj, Tl);
Tr = VADD(Tj, Tl);
}
{
V T6, T8, T5, T7;
T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
T6 = BYTWJ(&(W[0]), T5);
T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
T8 = BYTWJ(&(W[TWVL * 8]), T7);
T9 = VSUB(T6, T8);
Tt = VADD(T6, T8);
}
{
V Tb, Td, Ta, Tc;
Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Tb = BYTWJ(&(W[TWVL * 12]), Ta);
Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Td = BYTWJ(&(W[TWVL * 4]), Tc);
Te = VSUB(Tb, Td);
Tu = VADD(Tb, Td);
}
{
V Ts, Tv, Tw, Tx;
Ts = VADD(Tq, Tr);
Tv = VADD(Tt, Tu);
ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0]));
ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0]));
Tw = VSUB(Tq, Tr);
Tx = VBYI(VSUB(Tu, Tt));
ST(&(x[WS(rs, 6)]), VSUB(Tw, Tx), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VADD(Tw, Tx), ms, &(x[0]));
{
V Tg, To, Tn, Tp, Tf, Th;
Tf = VMUL(LDK(KP707106781), VADD(T9, Te));
Tg = VADD(T4, Tf);
To = VSUB(T4, Tf);
Th = VMUL(LDK(KP707106781), VSUB(Te, T9));
Tn = VBYI(VSUB(Th, Tm));
Tp = VBYI(VADD(Tm, Th));
ST(&(x[WS(rs, 7)]), VSUB(Tg, Tn), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VADD(Tg, Tn), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 5)]), VSUB(To, Tp), ms, &(x[WS(rs, 1)]));
}
}
}
}
static void n1bv_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP809016994, +0.809016994374947424102293417182819058860154590);
DVK(KP309016994, +0.309016994374947424102293417182819058860154590);
DVK(KP637423989, +0.637423989748689710176712811676016195434917298);
DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344);
DVK(KP125333233, +0.125333233564304245373118759816508793942918247);
DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418);
DVK(KP770513242, +0.770513242775789230803009636396177847271667672);
DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596);
DVK(KP992114701, +0.992114701314477831049793042785778521453036709);
DVK(KP250666467, +0.250666467128608490746237519633017587885836494);
DVK(KP851558583, +0.851558583130145297725004891488503407959946084);
DVK(KP904827052, +0.904827052466019527713668647932697593970413911);
DVK(KP425779291, +0.425779291565072648862502445744251703979973042);
DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822);
DVK(KP497379774, +0.497379774329709576484567492012895936835134813);
DVK(KP968583161, +0.968583161128631119490168375464735813836012403);
DVK(KP248689887, +0.248689887164854788242283746006447968417567406);
DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806);
DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483);
DVK(KP535826794, +0.535826794978996618271308767867639978063575346);
DVK(KP481753674, +0.481753674101715274987191502872129653528542010);
DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477);
DVK(KP844327925, +0.844327925502015078548558063966681505381659241);
DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691);
DVK(KP963507348, +0.963507348203430549974383005744259307057084020);
DVK(KP876306680, +0.876306680043863587308115903922062583399064238);
DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439);
DVK(KP062790519, +0.062790519529313376076178224565631133122484832);
DVK(KP684547105, +0.684547105928688673732283357621209269889519233);
DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328);
DVK(KP998026728, +0.998026728428271561952336806863450553336905220);
DVK(KP125581039, +0.125581039058626752152356449131262266244969664);
DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465);
DVK(KP728968627, +0.728968627421411523146730319055259111372571664);
DVK(KP293892626, +0.293892626146236564584352977319536384298826219);
DVK(KP475528258, +0.475528258147576786058219666689691071702849317);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
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 T1b, T2o, T1v, T1e, T2W, T2P, T2Q, T2U, T11, T27, TY, T26, T12, T2f, T1j;
V T28, TM, T24, TJ, T23, TN, T2e, T1i, T25, T2M, T2N, T2T, Tm, T1W, Tt;
V T1X, Tu, T20, Tw, T1Y, T7, T1U, Te, T1T, Tf, T21, Tx, T1V;
{
V T1c, T1a, T1t, T17, T1r;
T1c = LD(&(xi[0]), ivs, &(xi[0]));
{
V T18, T19, T15, T16;
T18 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T19 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)]));
T1a = VADD(T18, T19);
T1t = VSUB(T18, T19);
T15 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T16 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0]));
T17 = VADD(T15, T16);
T1r = VSUB(T15, T16);
}
{
V T2n, T1s, T1u, T1d;
T1b = VMUL(LDK(KP559016994), VSUB(T17, T1a));
T2n = VMUL(LDK(KP587785252), T1r);
T2o = VFNMS(LDK(KP951056516), T1t, T2n);
T1s = VMUL(LDK(KP951056516), T1r);
T1u = VMUL(LDK(KP587785252), T1t);
T1v = VADD(T1s, T1u);
T1d = VADD(T17, T1a);
T1e = VFNMS(LDK(KP250000000), T1d, T1c);
T2W = VADD(T1c, T1d);
}
}
{
V TG, TV, TF, TL, TH, TK, TU, T10, TW, TZ, TX, TI;
TG = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
TV = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
{
V Tz, TA, TB, TC, TD, TE;
Tz = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
TA = LD(&(xi[WS(is, 22)]), ivs, &(xi[0]));
TB = VADD(Tz, TA);
TC = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
TD = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)]));
TE = VADD(TC, TD);
TF = VMUL(LDK(KP559016994), VSUB(TB, TE));
TL = VSUB(TC, TD);
TH = VADD(TB, TE);
TK = VSUB(Tz, TA);
}
{
V TO, TP, TQ, TR, TS, TT;
TO = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
TP = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)]));
TQ = VADD(TO, TP);
TR = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
TS = LD(&(xi[WS(is, 18)]), ivs, &(xi[0]));
TT = VADD(TR, TS);
TU = VMUL(LDK(KP559016994), VSUB(TQ, TT));
T10 = VSUB(TR, TS);
TW = VADD(TQ, TT);
TZ = VSUB(TO, TP);
}
T2P = VADD(TG, TH);
T2Q = VADD(TV, TW);
T2U = VADD(T2P, T2Q);
T11 = VFMA(LDK(KP475528258), TZ, VMUL(LDK(KP293892626), T10));
T27 = VFNMS(LDK(KP475528258), T10, VMUL(LDK(KP293892626), TZ));
TX = VFNMS(LDK(KP250000000), TW, TV);
TY = VADD(TU, TX);
T26 = VSUB(TX, TU);
T12 = VFNMS(LDK(KP1_369094211), T11, VMUL(LDK(KP728968627), TY));
T2f = VFMA(LDK(KP125581039), T27, VMUL(LDK(KP998026728), T26));
T1j = VFMA(LDK(KP1_457937254), T11, VMUL(LDK(KP684547105), TY));
T28 = VFNMS(LDK(KP1_996053456), T27, VMUL(LDK(KP062790519), T26));
TM = VFMA(LDK(KP475528258), TK, VMUL(LDK(KP293892626), TL));
T24 = VFNMS(LDK(KP475528258), TL, VMUL(LDK(KP293892626), TK));
TI = VFNMS(LDK(KP250000000), TH, TG);
TJ = VADD(TF, TI);
T23 = VSUB(TI, TF);
TN = VFNMS(LDK(KP963507348), TM, VMUL(LDK(KP876306680), TJ));
T2e = VFMA(LDK(KP1_071653589), T24, VMUL(LDK(KP844327925), T23));
T1i = VFMA(LDK(KP1_752613360), TM, VMUL(LDK(KP481753674), TJ));
T25 = VFNMS(LDK(KP1_688655851), T24, VMUL(LDK(KP535826794), T23));
}
{
V Tb, Tq, T3, Tc, T6, Ta, Ti, Tr, Tl, Tp, Ts, Td;
Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tq = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
{
V T1, T2, T8, T4, T5, T9;
T1 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T2 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0]));
T8 = VADD(T1, T2);
T4 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)]));
T9 = VADD(T4, T5);
T3 = VSUB(T1, T2);
Tc = VADD(T8, T9);
T6 = VSUB(T4, T5);
Ta = VMUL(LDK(KP559016994), VSUB(T8, T9));
}
{
V Tg, Th, Tn, Tj, Tk, To;
Tg = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Th = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)]));
Tn = VADD(Tg, Th);
Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Tk = LD(&(xi[WS(is, 16)]), ivs, &(xi[0]));
To = VADD(Tj, Tk);
Ti = VSUB(Tg, Th);
Tr = VADD(Tn, To);
Tl = VSUB(Tj, Tk);
Tp = VMUL(LDK(KP559016994), VSUB(Tn, To));
}
T2M = VADD(Tq, Tr);
T2N = VADD(Tb, Tc);
T2T = VADD(T2M, T2N);
Tm = VFMA(LDK(KP475528258), Ti, VMUL(LDK(KP293892626), Tl));
T1W = VFNMS(LDK(KP475528258), Tl, VMUL(LDK(KP293892626), Ti));
Ts = VFNMS(LDK(KP250000000), Tr, Tq);
Tt = VADD(Tp, Ts);
T1X = VSUB(Ts, Tp);
Tu = VFMA(LDK(KP1_937166322), Tm, VMUL(LDK(KP248689887), Tt));
T20 = VFNMS(LDK(KP963507348), T1W, VMUL(LDK(KP876306680), T1X));
Tw = VFNMS(LDK(KP497379774), Tm, VMUL(LDK(KP968583161), Tt));
T1Y = VFMA(LDK(KP1_752613360), T1W, VMUL(LDK(KP481753674), T1X));
T7 = VFMA(LDK(KP475528258), T3, VMUL(LDK(KP293892626), T6));
T1U = VFNMS(LDK(KP475528258), T6, VMUL(LDK(KP293892626), T3));
Td = VFNMS(LDK(KP250000000), Tc, Tb);
Te = VADD(Ta, Td);
T1T = VSUB(Td, Ta);
Tf = VFMA(LDK(KP1_071653589), T7, VMUL(LDK(KP844327925), Te));
T21 = VFMA(LDK(KP1_809654104), T1U, VMUL(LDK(KP425779291), T1T));
Tx = VFNMS(LDK(KP1_688655851), T7, VMUL(LDK(KP535826794), Te));
T1V = VFNMS(LDK(KP851558583), T1U, VMUL(LDK(KP904827052), T1T));
}
{
V T2V, T2X, T2Y, T2S, T30, T2O, T2R, T31, T2Z;
T2V = VMUL(LDK(KP559016994), VSUB(T2T, T2U));
T2X = VADD(T2T, T2U);
T2Y = VFNMS(LDK(KP250000000), T2X, T2W);
T2O = VSUB(T2M, T2N);
T2R = VSUB(T2P, T2Q);
T2S = VBYI(VFMA(LDK(KP951056516), T2O, VMUL(LDK(KP587785252), T2R)));
T30 = VBYI(VFNMS(LDK(KP951056516), T2R, VMUL(LDK(KP587785252), T2O)));
ST(&(xo[0]), VADD(T2W, T2X), ovs, &(xo[0]));
T31 = VSUB(T2Y, T2V);
ST(&(xo[WS(os, 10)]), VADD(T30, T31), ovs, &(xo[0]));
ST(&(xo[WS(os, 15)]), VSUB(T31, T30), ovs, &(xo[WS(os, 1)]));
T2Z = VADD(T2V, T2Y);
ST(&(xo[WS(os, 5)]), VADD(T2S, T2Z), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 20)]), VSUB(T2Z, T2S), ovs, &(xo[0]));
}
{
V T1Z, T2i, T2j, T2g, T2w, T2x, T2y, T2G, T2H, T2I, T2D, T2E, T2F, T2z, T2A;
V T2B, T2p, T2m, T2q, T2b, T2c, T2a, T2d, T2h, T2r;
T1Z = VSUB(T1V, T1Y);
T2i = VADD(T20, T21);
T2j = VSUB(T25, T28);
T2g = VSUB(T2e, T2f);
T2w = VFMA(LDK(KP1_369094211), T1W, VMUL(LDK(KP728968627), T1X));
T2x = VFNMS(LDK(KP992114701), T1T, VMUL(LDK(KP250666467), T1U));
T2y = VADD(T2w, T2x);
T2G = VFNMS(LDK(KP125581039), T24, VMUL(LDK(KP998026728), T23));
T2H = VFMA(LDK(KP1_274847979), T27, VMUL(LDK(KP770513242), T26));
T2I = VADD(T2G, T2H);
T2D = VFNMS(LDK(KP1_457937254), T1W, VMUL(LDK(KP684547105), T1X));
T2E = VFMA(LDK(KP1_984229402), T1U, VMUL(LDK(KP125333233), T1T));
T2F = VADD(T2D, T2E);
T2z = VFMA(LDK(KP1_996053456), T24, VMUL(LDK(KP062790519), T23));
T2A = VFNMS(LDK(KP637423989), T26, VMUL(LDK(KP1_541026485), T27));
T2B = VADD(T2z, T2A);
{
V T2k, T2l, T22, T29;
T2k = VADD(T1Y, T1V);
T2l = VADD(T2e, T2f);
T2p = VADD(T2k, T2l);
T2m = VMUL(LDK(KP559016994), VSUB(T2k, T2l));
T2q = VFNMS(LDK(KP250000000), T2p, T2o);
T2b = VSUB(T1e, T1b);
T22 = VSUB(T20, T21);
T29 = VADD(T25, T28);
T2c = VADD(T22, T29);
T2a = VMUL(LDK(KP559016994), VSUB(T22, T29));
T2d = VFNMS(LDK(KP250000000), T2c, T2b);
}
{
V T2u, T2v, T2C, T2J;
T2u = VADD(T2b, T2c);
T2v = VBYI(VADD(T2o, T2p));
ST(&(xo[WS(os, 23)]), VSUB(T2u, T2v), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VADD(T2u, T2v), ovs, &(xo[0]));
T2C = VADD(T2b, VADD(T2y, T2B));
T2J = VBYI(VSUB(VADD(T2F, T2I), T2o));
ST(&(xo[WS(os, 22)]), VSUB(T2C, T2J), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VADD(T2C, T2J), ovs, &(xo[WS(os, 1)]));
}
T2h = VFMA(LDK(KP951056516), T1Z, VADD(T2a, VFNMS(LDK(KP587785252), T2g, T2d)));
T2r = VBYI(VADD(VFMA(LDK(KP951056516), T2i, VMUL(LDK(KP587785252), T2j)), VADD(T2m, T2q)));
ST(&(xo[WS(os, 18)]), VSUB(T2h, T2r), ovs, &(xo[0]));
ST(&(xo[WS(os, 7)]), VADD(T2h, T2r), ovs, &(xo[WS(os, 1)]));
{
V T2s, T2t, T2K, T2L;
T2s = VFMA(LDK(KP587785252), T1Z, VFMA(LDK(KP951056516), T2g, VSUB(T2d, T2a)));
T2t = VBYI(VADD(VFNMS(LDK(KP951056516), T2j, VMUL(LDK(KP587785252), T2i)), VSUB(T2q, T2m)));
ST(&(xo[WS(os, 13)]), VSUB(T2s, T2t), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 12)]), VADD(T2s, T2t), ovs, &(xo[0]));
T2K = VBYI(VSUB(VFMA(LDK(KP951056516), VSUB(T2w, T2x), VFMA(LDK(KP309016994), T2F, VFNMS(LDK(KP809016994), T2I, VMUL(LDK(KP587785252), VSUB(T2z, T2A))))), T2o));
T2L = VFMA(LDK(KP309016994), T2y, VFMA(LDK(KP951056516), VSUB(T2E, T2D), VFMA(LDK(KP587785252), VSUB(T2H, T2G), VFNMS(LDK(KP809016994), T2B, T2b))));
ST(&(xo[WS(os, 8)]), VADD(T2K, T2L), ovs, &(xo[0]));
ST(&(xo[WS(os, 17)]), VSUB(T2L, T2K), ovs, &(xo[WS(os, 1)]));
}
}
{
V Tv, T1m, T1n, T1k, T1D, T1E, T1F, T1N, T1O, T1P, T1K, T1L, T1M, T1G, T1H;
V T1I, T1w, T1q, T1x, T1f, T1g, T14, T1h, T1l, T1y;
Tv = VSUB(Tf, Tu);
T1m = VSUB(Tw, Tx);
T1n = VSUB(TN, T12);
T1k = VSUB(T1i, T1j);
T1D = VFMA(LDK(KP1_688655851), Tm, VMUL(LDK(KP535826794), Tt));
T1E = VFMA(LDK(KP1_541026485), T7, VMUL(LDK(KP637423989), Te));
T1F = VSUB(T1D, T1E);
T1N = VFMA(LDK(KP851558583), TM, VMUL(LDK(KP904827052), TJ));
T1O = VFMA(LDK(KP1_984229402), T11, VMUL(LDK(KP125333233), TY));
T1P = VADD(T1N, T1O);
T1K = VFNMS(LDK(KP1_071653589), Tm, VMUL(LDK(KP844327925), Tt));
T1L = VFNMS(LDK(KP770513242), Te, VMUL(LDK(KP1_274847979), T7));
T1M = VADD(T1K, T1L);
T1G = VFNMS(LDK(KP425779291), TJ, VMUL(LDK(KP1_809654104), TM));
T1H = VFNMS(LDK(KP992114701), TY, VMUL(LDK(KP250666467), T11));
T1I = VADD(T1G, T1H);
{
V T1o, T1p, Ty, T13;
T1o = VADD(Tu, Tf);
T1p = VADD(T1i, T1j);
T1w = VADD(T1o, T1p);
T1q = VMUL(LDK(KP559016994), VSUB(T1o, T1p));
T1x = VFNMS(LDK(KP250000000), T1w, T1v);
T1f = VADD(T1b, T1e);
Ty = VADD(Tw, Tx);
T13 = VADD(TN, T12);
T1g = VADD(Ty, T13);
T14 = VMUL(LDK(KP559016994), VSUB(Ty, T13));
T1h = VFNMS(LDK(KP250000000), T1g, T1f);
}
{
V T1B, T1C, T1J, T1Q;
T1B = VADD(T1f, T1g);
T1C = VBYI(VADD(T1v, T1w));
ST(&(xo[WS(os, 24)]), VSUB(T1B, T1C), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VADD(T1B, T1C), ovs, &(xo[WS(os, 1)]));
T1J = VADD(T1f, VADD(T1F, T1I));
T1Q = VBYI(VSUB(VADD(T1M, T1P), T1v));
ST(&(xo[WS(os, 21)]), VSUB(T1J, T1Q), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VADD(T1J, T1Q), ovs, &(xo[0]));
}
T1l = VFMA(LDK(KP951056516), Tv, VADD(T14, VFNMS(LDK(KP587785252), T1k, T1h)));
T1y = VBYI(VADD(VFMA(LDK(KP951056516), T1m, VMUL(LDK(KP587785252), T1n)), VADD(T1q, T1x)));
ST(&(xo[WS(os, 19)]), VSUB(T1l, T1y), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 6)]), VADD(T1l, T1y), ovs, &(xo[0]));
{
V T1z, T1A, T1R, T1S;
T1z = VFMA(LDK(KP587785252), Tv, VFMA(LDK(KP951056516), T1k, VSUB(T1h, T14)));
T1A = VBYI(VADD(VFNMS(LDK(KP951056516), T1n, VMUL(LDK(KP587785252), T1m)), VSUB(T1x, T1q)));
ST(&(xo[WS(os, 14)]), VSUB(T1z, T1A), ovs, &(xo[0]));
ST(&(xo[WS(os, 11)]), VADD(T1z, T1A), ovs, &(xo[WS(os, 1)]));
T1R = VBYI(VSUB(VFMA(LDK(KP309016994), T1M, VFMA(LDK(KP951056516), VADD(T1D, T1E), VFNMS(LDK(KP809016994), T1P, VMUL(LDK(KP587785252), VSUB(T1G, T1H))))), T1v));
T1S = VFMA(LDK(KP951056516), VSUB(T1L, T1K), VFMA(LDK(KP309016994), T1F, VFMA(LDK(KP587785252), VSUB(T1O, T1N), VFNMS(LDK(KP809016994), T1I, T1f))));
ST(&(xo[WS(os, 9)]), VADD(T1R, T1S), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 16)]), VSUB(T1S, T1R), ovs, &(xo[0]));
}
}
}
}
static void n1fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(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);
ST(&(xo[0]), VADD(TL, TM), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VSUB(TL, TM), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(TK, TH), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VFMAI(TE, TD), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VFNMSI(TE, TD), ovs, &(xo[0]));
ST(&(xo[WS(os, 10)]), VFNMSI(TC, Tz), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VFMAI(TC, Tz), ovs, &(xo[0]));
ST(&(xo[WS(os, 5)]), VFNMSI(Tw, Tv), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VFMAI(Tw, Tv), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 11)]), VFMAI(Tu, Tj), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 1)]), VFNMSI(Tu, Tj), ovs, &(xo[WS(os, 1)]));
}
}
}
}
}
}
}
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(KP500000000, +0.500000000000000000000000000000000000000000000);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
int i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
BEGIN_SIMD();
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
V T5, Ta, TJ, Ty, Tq, Tp, Tg, Tl, TI, TA, Tz, Tu;
{
V T1, T6, T4, Tw, T9, Tx;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
{
V T2, T3, T7, T8;
T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T4 = VADD(T2, T3);
Tw = VSUB(T3, T2);
T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T9 = VADD(T7, T8);
Tx = VSUB(T8, T7);
}
T5 = VADD(T1, T4);
Ta = VADD(T6, T9);
TJ = VADD(Tw, Tx);
Ty = VMUL(LDK(KP866025403), VSUB(Tw, Tx));
Tq = VFNMS(LDK(KP500000000), T9, T6);
Tp = VFNMS(LDK(KP500000000), T4, T1);
}
{
V Tc, Th, Tf, Ts, Tk, Tt;
Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
{
V Td, Te, Ti, Tj;
Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
Tf = VADD(Td, Te);
Ts = VSUB(Te, Td);
Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tk = VADD(Ti, Tj);
Tt = VSUB(Tj, Ti);
}
Tg = VADD(Tc, Tf);
Tl = VADD(Th, Tk);
TI = VADD(Ts, Tt);
TA = VFNMS(LDK(KP500000000), Tk, Th);
Tz = VFNMS(LDK(KP500000000), Tf, Tc);
Tu = VMUL(LDK(KP866025403), VSUB(Ts, Tt));
}
{
V Tb, Tm, Tn, To;
Tb = VSUB(T5, Ta);
Tm = VBYI(VSUB(Tg, Tl));
ST(&(xo[18]), VSUB(Tb, Tm), ovs, &(xo[2]));
ST(&(xo[6]), VADD(Tb, Tm), ovs, &(xo[2]));
Tn = VADD(T5, Ta);
To = VADD(Tg, Tl);
ST(&(xo[12]), VSUB(Tn, To), ovs, &(xo[0]));
ST(&(xo[0]), VADD(Tn, To), ovs, &(xo[0]));
}
{
V Tv, TE, TC, TD, Tr, TB;
Tr = VSUB(Tp, Tq);
Tv = VSUB(Tr, Tu);
TE = VADD(Tr, Tu);
TB = VSUB(Tz, TA);
TC = VBYI(VADD(Ty, TB));
TD = VBYI(VSUB(Ty, TB));
ST(&(xo[10]), VSUB(Tv, TC), ovs, &(xo[2]));
ST(&(xo[22]), VSUB(TE, TD), ovs, &(xo[2]));
ST(&(xo[14]), VADD(TC, Tv), ovs, &(xo[2]));
ST(&(xo[2]), VADD(TD, TE), ovs, &(xo[2]));
}
{
V TK, TM, TH, TL, TF, TG;
TK = VBYI(VMUL(LDK(KP866025403), VSUB(TI, TJ)));
TM = VBYI(VMUL(LDK(KP866025403), VADD(TJ, TI)));
TF = VADD(Tp, Tq);
TG = VADD(Tz, TA);
TH = VSUB(TF, TG);
TL = VADD(TF, TG);
ST(&(xo[20]), VSUB(TH, TK), ovs, &(xo[0]));
ST(&(xo[8]), VADD(TL, TM), ovs, &(xo[0]));
ST(&(xo[4]), VADD(TH, TK), ovs, &(xo[0]));
ST(&(xo[16]), VSUB(TL, TM), ovs, &(xo[0]));
}
}
END_SIMD();
}
longdouble strtold_dm(const char *p,char **endp)
{
longdouble ldval;
int exp;
long long msdec,lsdec;
unsigned long msscale;
char dot,sign;
int pow;
int ndigits;
const char *pinit = p;
#if __DMC__
static char infinity[] = "infinity";
static char nans[] = "nans";
#endif
unsigned int old_cw;
#if _WIN32 && __DMC__
fenv_t flagp;
fegetenv(&flagp); /* Store all exceptions, and current status word */
if (_8087)
{
// disable exceptions from occurring, set max precision, and round to nearest
#if __DMC__
__asm
{
fstcw word ptr old_cw
mov EAX,old_cw
mov ECX,EAX
and EAX,0xf0c0
or EAX,033fh
mov old_cw,EAX
fldcw word ptr old_cw
mov old_cw,ECX
}
#else
old_cw = _control87(_MCW_EM | _PC_64 | _RC_NEAR,
_MCW_EM | _MCW_PC | _MCW_RC);
#endif
}
#endif
while (isspace(*p))
p++;
sign = 0; /* indicating + */
switch (*p)
{ case '-':
sign++;
/* FALL-THROUGH */
case '+':
p++;
}
ldval = 0.0;
dot = 0; /* if decimal point has been seen */
exp = 0;
msdec = lsdec = 0;
msscale = 1;
ndigits = 0;
#if __DMC__
switch (*p)
{ case 'i':
case 'I':
if (memicmp(p,infinity,8) == 0)
{ p += 8;
goto L4;
}
if (memicmp(p,infinity,3) == 0) /* is it "inf"? */
{ p += 3;
L4:
ldval = HUGE_VAL;
goto L3;
}
break;
case 'n':
case 'N':
if (memicmp(p,nans,4) == 0) /* "nans"? */
{ p += 4;
ldval = NANS;
goto L5;
}
if (memicmp(p,nans,3) == 0) /* "nan"? */
{ p += 3;
ldval = NAN;
L5:
if (*p == '(') /* if (n-char-sequence) */
goto Lerr; /* invalid input */
goto L3;
}
}
#endif
if (*p == '0' && (p[1] == 'x' || p[1] == 'X'))
{ int guard = 0;
int anydigits = 0;
p += 2;
while (1)
{ int i = *p;
while (isxdigit(i))
{
anydigits = 1;
i = isalpha(i) ? ((i & ~0x20) - ('A' - 10)) : i - '0';
if (ndigits < 16)
{
msdec = msdec * 16 + i;
if (msdec)
ndigits++;
}
else if (ndigits == 16)
{
while (msdec >= 0)
{
exp--;
msdec <<= 1;
i <<= 1;
if (i & 0x10)
msdec |= 1;
}
guard = i << 4;
ndigits++;
exp += 4;
}
else
{
guard |= i;
exp += 4;
}
exp -= dot;
i = *++p;
}
#if _WIN32 && __DMC__
if (i == *__locale_decpoint && !dot)
#else
if (i == '.' && !dot)
#endif
{ p++;
dot = 4;
}
else
break;
}
// Round up if (guard && (sticky || odd))
if (guard & 0x80 && (guard & 0x7F || msdec & 1))
{
msdec++;
if (msdec == 0) // overflow
{ msdec = 0x8000000000000000LL;
exp++;
}
}
if (anydigits == 0) // if error (no digits seen)
goto Lerr;
if (*p == 'p' || *p == 'P')
{
char sexp;
int e;
sexp = 0;
switch (*++p)
{ case '-': sexp++;
case '+': p++;
}
ndigits = 0;
e = 0;
while (isdigit(*p))
{
if (e < 0x7FFFFFFF / 10 - 10) // prevent integer overflow
{
e = e * 10 + *p - '0';
}
p++;
ndigits = 1;
}
exp += (sexp) ? -e : e;
if (!ndigits) // if no digits in exponent
goto Lerr;
if (msdec)
{
#if __DMC__
// The 8087 has no instruction to load an
// unsigned long long
if (msdec < 0)
{
*(long long *)&ldval = msdec;
((unsigned short *)&ldval)[4] = 0x3FFF + 63;
}
else
{ // But does for a signed one
__asm
{
fild qword ptr msdec
fstp tbyte ptr ldval
}
}
#else
int e2 = 0x3FFF + 63;
// left justify mantissa
while (msdec >= 0)
{ msdec <<= 1;
e2--;
}
// Stuff mantissa directly into long double
union U
{
longdouble ld;
struct S
{
long long mantissa;
unsigned short exp;
} s;
} u;
u.s.mantissa = msdec;
u.s.exp = e2;
ldval = u.ld;
#endif
#if 0
if (0)
{ int i;
printf("msdec = x%llx, ldval = %Lg\n", msdec, ldval);
for (i = 0; i < 5; i++)
printf("%04x ",((unsigned short *)&ldval)[i]);
printf("\n");
printf("%llx\n",ldval);
}
#endif
// Exponent is power of 2, not power of 10
#if _WIN32 && __DMC__
__asm
{
fild dword ptr exp
fld tbyte ptr ldval
fscale // ST(0) = ST(0) * (2**ST(1))
fstp ST(1)
fstp tbyte ptr ldval
}
#else
ldval = ldexpl(ldval,exp);
#endif
}
goto L6;
}
static void t1fuv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP618033988, +0.618033988749894848204586834365638117720309180);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
{
INT m;
R *x;
x = ri;
for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(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 n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, int v, int ivs, int ovs)
{
DVK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
DVK(KP083333333, +0.083333333333333333333333333333333333333333333);
DVK(KP075902986, +0.075902986037193865983102897245103540356428373);
DVK(KP251768516, +0.251768516431883313623436926934233488546674281);
DVK(KP132983124, +0.132983124607418643793760531921092974399165133);
DVK(KP258260390, +0.258260390311744861420450644284508567852516811);
DVK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
DVK(KP300238635, +0.300238635966332641462884626667381504676006424);
DVK(KP011599105, +0.011599105605768290721655456654083252189827041);
DVK(KP156891391, +0.156891391051584611046832726756003269660212636);
DVK(KP256247671, +0.256247671582936600958684654061725059144125175);
DVK(KP174138601, +0.174138601152135905005660794929264742616964676);
DVK(KP575140729, +0.575140729474003121368385547455453388461001608);
DVK(KP503537032, +0.503537032863766627246873853868466977093348562);
DVK(KP113854479, +0.113854479055790798974654345867655310534642560);
DVK(KP265966249, +0.265966249214837287587521063842185948798330267);
DVK(KP387390585, +0.387390585467617292130675966426762851778775217);
DVK(KP300462606, +0.300462606288665774426601772289207995520941381);
DVK(KP866025403, +0.866025403784438646763723170752936183471402627);
DVK(KP500000000, +0.500000000000000000000000000000000000000000000);
int i;
const R *xi;
R *xo;
xi = ri;
xo = ro;
BEGIN_SIMD();
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs)) {
V TW, Tb, Tm, Tu, TC, TR, TX, TK, TU, Tz, TB, TN, TT;
TW = LD(&(xi[0]), ivs, &(xi[0]));
{
V T3, TH, Tl, Tw, Tp, Tg, Tv, To, T6, Tr, T9, Ts, Ta, TI, T1;
V T2, Tq, Tt;
T1 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
TH = VADD(T1, T2);
{
V Th, Ti, Tj, Tk;
Th = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Tj = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
Tk = VADD(Ti, Tj);
Tl = VADD(Th, Tk);
Tw = VSUB(Ti, Tj);
Tp = VFNMS(LDK(KP500000000), Tk, Th);
}
{
V Tc, Td, Te, Tf;
Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Td = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Tf = VADD(Td, Te);
Tg = VADD(Tc, Tf);
Tv = VSUB(Td, Te);
To = VFNMS(LDK(KP500000000), Tf, Tc);
}
{
V T4, T5, T7, T8;
T4 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
T5 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
T6 = VSUB(T4, T5);
Tr = VADD(T4, T5);
T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T9 = VSUB(T7, T8);
Ts = VADD(T7, T8);
}
Ta = VADD(T6, T9);
TI = VADD(Tr, Ts);
Tb = VADD(T3, Ta);
Tm = VSUB(Tg, Tl);
Tq = VSUB(To, Tp);
Tt = VMUL(LDK(KP866025403), VSUB(Tr, Ts));
Tu = VADD(Tq, Tt);
TC = VSUB(Tq, Tt);
{
V TP, TQ, TG, TJ;
TP = VADD(Tg, Tl);
TQ = VADD(TH, TI);
TR = VMUL(LDK(KP300462606), VSUB(TP, TQ));
TX = VADD(TP, TQ);
TG = VADD(To, Tp);
TJ = VFNMS(LDK(KP500000000), TI, TH);
TK = VSUB(TG, TJ);
TU = VADD(TG, TJ);
}
{
V Tx, Ty, TL, TM;
Tx = VMUL(LDK(KP866025403), VSUB(Tv, Tw));
Ty = VFNMS(LDK(KP500000000), Ta, T3);
Tz = VSUB(Tx, Ty);
TB = VADD(Tx, Ty);
TL = VADD(Tv, Tw);
TM = VSUB(T6, T9);
TN = VSUB(TL, TM);
TT = VADD(TL, TM);
}
}
ST(&(xo[0]), VADD(TW, TX), ovs, &(xo[0]));
{
V T19, T1n, T14, T13, T1f, T1k, Tn, TE, T1e, T1j, TS, T1m, TZ, T1c, TA;
V TD;
{
V T17, T18, T11, T12;
T17 = VFMA(LDK(KP387390585), TN, VMUL(LDK(KP265966249), TK));
T18 = VFNMS(LDK(KP503537032), TU, VMUL(LDK(KP113854479), TT));
T19 = VSUB(T17, T18);
T1n = VADD(T17, T18);
T14 = VFMA(LDK(KP575140729), Tm, VMUL(LDK(KP174138601), Tb));
T11 = VFNMS(LDK(KP156891391), TB, VMUL(LDK(KP256247671), TC));
T12 = VFMA(LDK(KP011599105), Tz, VMUL(LDK(KP300238635), Tu));
T13 = VSUB(T11, T12);
T1f = VADD(T14, T13);
T1k = VMUL(LDK(KP1_732050807), VADD(T11, T12));
}
Tn = VFNMS(LDK(KP174138601), Tm, VMUL(LDK(KP575140729), Tb));
TA = VFNMS(LDK(KP300238635), Tz, VMUL(LDK(KP011599105), Tu));
TD = VFMA(LDK(KP256247671), TB, VMUL(LDK(KP156891391), TC));
TE = VSUB(TA, TD);
T1e = VMUL(LDK(KP1_732050807), VADD(TD, TA));
T1j = VSUB(Tn, TE);
{
V TO, T1b, TV, TY, T1a;
TO = VFNMS(LDK(KP132983124), TN, VMUL(LDK(KP258260390), TK));
T1b = VSUB(TR, TO);
TV = VFMA(LDK(KP251768516), TT, VMUL(LDK(KP075902986), TU));
TY = VFNMS(LDK(KP083333333), TX, TW);
T1a = VSUB(TY, TV);
TS = VFMA(LDK(KP2_000000000), TO, TR);
T1m = VADD(T1b, T1a);
TZ = VFMA(LDK(KP2_000000000), TV, TY);
T1c = VSUB(T1a, T1b);
}
{
V TF, T10, T1l, T1o;
TF = VBYI(VFMA(LDK(KP2_000000000), TE, Tn));
T10 = VADD(TS, TZ);
ST(&(xo[WS(os, 1)]), VADD(TF, T10), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 12)]), VSUB(T10, TF), ovs, &(xo[0]));
{
V T15, T16, T1p, T1q;
T15 = VBYI(VFMS(LDK(KP2_000000000), T13, T14));
T16 = VSUB(TZ, TS);
ST(&(xo[WS(os, 5)]), VADD(T15, T16), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 8)]), VSUB(T16, T15), ovs, &(xo[0]));
T1p = VADD(T1n, T1m);
T1q = VBYI(VADD(T1j, T1k));
ST(&(xo[WS(os, 4)]), VSUB(T1p, T1q), ovs, &(xo[0]));
ST(&(xo[WS(os, 9)]), VADD(T1q, T1p), ovs, &(xo[WS(os, 1)]));
}
T1l = VBYI(VSUB(T1j, T1k));
T1o = VSUB(T1m, T1n);
ST(&(xo[WS(os, 3)]), VADD(T1l, T1o), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 10)]), VSUB(T1o, T1l), ovs, &(xo[0]));
{
V T1h, T1i, T1d, T1g;
T1h = VBYI(VSUB(T1e, T1f));
T1i = VSUB(T1c, T19);
ST(&(xo[WS(os, 6)]), VADD(T1h, T1i), ovs, &(xo[0]));
ST(&(xo[WS(os, 7)]), VSUB(T1i, T1h), ovs, &(xo[WS(os, 1)]));
T1d = VADD(T19, T1c);
T1g = VBYI(VADD(T1e, T1f));
ST(&(xo[WS(os, 2)]), VSUB(T1d, T1g), ovs, &(xo[0]));
ST(&(xo[WS(os, 11)]), VADD(T1g, T1d), ovs, &(xo[WS(os, 1)]));
}
}
}
}
END_SIMD();
}
static void n1bv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP959492973, +0.959492973614497389890368057066327699062454848);
DVK(KP876768831, +0.876768831002589333891339807079336796764054852);
DVK(KP918985947, +0.918985947228994779780736114132655398124909697);
DVK(KP989821441, +0.989821441880932732376092037776718787376519372);
DVK(KP778434453, +0.778434453334651800608337670740821884709317477);
DVK(KP830830026, +0.830830026003772851058548298459246407048009821);
DVK(KP372785597, +0.372785597771792209609773152906148328659002598);
DVK(KP634356270, +0.634356270682424498893150776899916060542806975);
DVK(KP715370323, +0.715370323453429719112414662767260662417897278);
DVK(KP342584725, +0.342584725681637509502641509861112333758894680);
DVK(KP521108558, +0.521108558113202722944698153526659300680427422);
{
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, Tb, T4, Tq, Tg, Tm, T7, Tp, Ta, To, Tc, T11;
T1 = LD(&(xi[0]), ivs, &(xi[0]));
{
V T2, T3, Te, Tf;
T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tf = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
{
V T5, T6, T8, T9;
T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T6 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T9 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T4 = VADD(T2, T3);
Tq = VSUB(T2, T3);
Tg = VADD(Te, Tf);
Tm = VSUB(Te, Tf);
T7 = VADD(T5, T6);
Tp = VSUB(T5, T6);
Ta = VADD(T8, T9);
To = VSUB(T8, T9);
Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
}
}
T11 = VFMA(LDK(KP521108558), Tm, Tq);
{
V TA, TS, TE, TW, Td, Tn, Ts, Tw, Tr, Tv, TT, TF;
Tr = VFNMS(LDK(KP521108558), Tq, Tp);
Tv = VFNMS(LDK(KP342584725), T7, Tg);
TA = VFMA(LDK(KP715370323), To, Tq);
TS = VFMA(LDK(KP521108558), To, Tm);
TE = VFNMS(LDK(KP342584725), T4, Ta);
TW = VFNMS(LDK(KP342584725), Ta, T7);
Td = VADD(Tb, Tc);
Tn = VSUB(Tb, Tc);
Ts = VFNMS(LDK(KP715370323), Tr, To);
Tw = VFNMS(LDK(KP634356270), Tv, T4);
TT = VFNMS(LDK(KP715370323), TS, Tp);
TF = VFNMS(LDK(KP634356270), TE, Tg);
{
V Tu, TV, TD, TL, T14, TP, TZ, Tj, Tz, TI, TB, TJ, TM;
TB = VFMA(LDK(KP372785597), Tn, TA);
TJ = VFNMS(LDK(KP521108558), Tp, Tn);
{
V T12, TN, TX, Th;
T12 = VFMA(LDK(KP715370323), T11, Tn);
ST(&(xo[0]), VADD(Tg, VADD(Td, VADD(Ta, VADD(T7, VADD(T4, T1))))), ovs, &(xo[0]));
TN = VFNMS(LDK(KP342584725), Td, T4);
TX = VFNMS(LDK(KP634356270), TW, Td);
Th = VFNMS(LDK(KP342584725), Tg, Td);
{
V Tt, Tx, TU, TG;
Tt = VFNMS(LDK(KP830830026), Ts, Tn);
Tx = VFNMS(LDK(KP778434453), Tw, Ta);
TU = VFMA(LDK(KP830830026), TT, Tq);
TG = VFNMS(LDK(KP778434453), TF, Td);
{
V TC, TK, T13, TO;
TC = VFNMS(LDK(KP830830026), TB, Tm);
TK = VFMA(LDK(KP715370323), TJ, Tm);
T13 = VFMA(LDK(KP830830026), T12, Tp);
TO = VFNMS(LDK(KP634356270), TN, T7);
{
V TY, Ti, Ty, TH;
TY = VFNMS(LDK(KP778434453), TX, T4);
Ti = VFNMS(LDK(KP634356270), Th, Ta);
Tu = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), Tt, Tm));
Ty = VFNMS(LDK(KP876768831), Tx, Td);
TV = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), TU, Tn));
TH = VFNMS(LDK(KP876768831), TG, T7);
TD = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), TC, Tp));
TL = VFNMS(LDK(KP830830026), TK, To);
T14 = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), T13, To));
TP = VFNMS(LDK(KP778434453), TO, Tg);
TZ = VFNMS(LDK(KP876768831), TY, Tg);
Tj = VFNMS(LDK(KP778434453), Ti, T7);
Tz = VFNMS(LDK(KP959492973), Ty, T1);
TI = VFNMS(LDK(KP959492973), TH, T1);
}
}
}
}
TM = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), TL, Tq));
{
V TQ, T10, Tk, TR, Tl;
TQ = VFNMS(LDK(KP876768831), TP, Ta);
T10 = VFNMS(LDK(KP959492973), TZ, T1);
Tk = VFNMS(LDK(KP876768831), Tj, T4);
ST(&(xo[WS(os, 7)]), VFMAI(TD, Tz), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 4)]), VFNMSI(TD, Tz), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VFNMSI(TM, TI), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(TM, TI), ovs, &(xo[WS(os, 1)]));
TR = VFNMS(LDK(KP959492973), TQ, T1);
ST(&(xo[WS(os, 10)]), VFNMSI(T14, T10), ovs, &(xo[0]));
ST(&(xo[WS(os, 1)]), VFMAI(T14, T10), ovs, &(xo[WS(os, 1)]));
Tl = VFNMS(LDK(KP959492973), Tk, T1);
ST(&(xo[WS(os, 9)]), VFMAI(TV, TR), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 2)]), VFNMSI(TV, TR), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VFNMSI(Tu, Tl), ovs, &(xo[0]));
ST(&(xo[WS(os, 5)]), VFMAI(Tu, Tl), ovs, &(xo[WS(os, 1)]));
}
}
}
}
}
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(8, rs)) {
V T2, T3, Ta, T4, Tb, Tc, Tp;
T2 = LDW(&(W[0]));
T3 = LDW(&(W[TWVL * 2]));
Ta = VZMULJ(T2, T3);
T4 = VZMUL(T2, T3);
Tb = LDW(&(W[TWVL * 4]));
Tc = VZMULJ(Ta, Tb);
Tp = VZMULJ(T2, Tb);
{
V T7, Tx, Ts, Ty, Tf, TA, Tk, TB, T1, T6, T5;
T1 = LD(&(x[0]), ms, &(x[0]));
T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
T6 = VZMULJ(T4, T5);
T7 = VSUB(T1, T6);
Tx = VADD(T1, T6);
{
V To, Tr, Tn, Tq;
Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
To = VZMULJ(Ta, Tn);
Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Tr = VZMULJ(Tp, Tq);
Ts = VSUB(To, Tr);
Ty = VADD(To, Tr);
}
{
V T9, Te, T8, Td;
T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
T9 = VZMULJ(T2, T8);
Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Te = VZMULJ(Tc, Td);
Tf = VSUB(T9, Te);
TA = VADD(T9, Te);
}
{
V Th, Tj, Tg, Ti;
Tg = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Th = VZMULJ(Tb, Tg);
Ti = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Tj = VZMULJ(T3, Ti);
Tk = VSUB(Th, Tj);
TB = VADD(Th, Tj);
}
{
V Tz, TC, TD, TE;
Tz = VADD(Tx, Ty);
TC = VADD(TA, TB);
ST(&(x[WS(rs, 4)]), VSUB(Tz, TC), ms, &(x[0]));
ST(&(x[0]), VADD(Tz, TC), ms, &(x[0]));
TD = VSUB(Tx, Ty);
TE = VSUB(TB, TA);
ST(&(x[WS(rs, 6)]), VFNMSI(TE, TD), ms, &(x[0]));
ST(&(x[WS(rs, 2)]), VFMAI(TE, TD), ms, &(x[0]));
{
V Tm, Tv, Tu, Tw, Tl, Tt;
Tl = VADD(Tf, Tk);
Tm = VFMA(LDK(KP707106781), Tl, T7);
Tv = VFNMS(LDK(KP707106781), Tl, T7);
Tt = VSUB(Tk, Tf);
Tu = VFNMS(LDK(KP707106781), Tt, Ts);
Tw = VFMA(LDK(KP707106781), Tt, Ts);
ST(&(x[WS(rs, 1)]), VFNMSI(Tu, Tm), 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, 5)]), VFNMSI(Tw, Tv), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
VLEAVE();
}
static void t1fv_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) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) {
V T4, Tx, T1m, T1K, TZ, T16, T17, T10, Tf, Tq, Tr, T1O, T1P, T1Q, T1w;
V T1z, T1A, TI, TT, TU, T1L, T1M, T1N, T1p, T1s, T1t, Ts, TV;
{
V T1, Tw, T3, Tu, Tv, T2, Tt, T1k, T1l;
T1 = LD(&(x[0]), ms, &(x[0]));
Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)]));
Tw = BYTWJ(&(W[TWVL * 28]), Tv);
T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
T3 = BYTWJ(&(W[TWVL * 18]), T2);
Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tu = BYTWJ(&(W[TWVL * 8]), Tt);
T4 = VSUB(T1, T3);
Tx = VSUB(Tu, Tw);
T1k = VADD(T1, T3);
T1l = VADD(Tu, Tw);
T1m = VSUB(T1k, T1l);
T1K = VADD(T1k, T1l);
}
{
V T9, T1n, TN, T1v, TS, T1y, Te, T1q, Tk, T1u, TC, T1o, TH, T1r, Tp;
V T1x;
{
V T6, T8, T5, T7;
T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
T6 = BYTWJ(&(W[TWVL * 6]), T5);
T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0]));
T8 = BYTWJ(&(W[TWVL * 26]), T7);
T9 = VSUB(T6, T8);
T1n = VADD(T6, T8);
}
{
V TK, TM, TJ, TL;
TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)]));
TK = BYTWJ(&(W[TWVL * 24]), TJ);
TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
TM = BYTWJ(&(W[TWVL * 4]), TL);
TN = VSUB(TK, TM);
T1v = VADD(TK, TM);
}
{
V TP, TR, TO, TQ;
TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)]));
TP = BYTWJ(&(W[TWVL * 32]), TO);
TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
TR = BYTWJ(&(W[TWVL * 12]), TQ);
TS = VSUB(TP, TR);
T1y = VADD(TP, TR);
}
{
V Tb, Td, Ta, Tc;
Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0]));
Tb = BYTWJ(&(W[TWVL * 30]), Ta);
Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
Td = BYTWJ(&(W[TWVL * 10]), Tc);
Te = VSUB(Tb, Td);
T1q = VADD(Tb, Td);
}
{
V Th, Tj, Tg, Ti;
Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
Th = BYTWJ(&(W[TWVL * 14]), Tg);
Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0]));
Tj = BYTWJ(&(W[TWVL * 34]), Ti);
Tk = VSUB(Th, Tj);
T1u = VADD(Th, Tj);
}
{
V Tz, TB, Ty, TA;
Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
Tz = BYTWJ(&(W[TWVL * 16]), Ty);
TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)]));
TB = BYTWJ(&(W[TWVL * 36]), TA);
TC = VSUB(Tz, TB);
T1o = VADD(Tz, TB);
}
{
V TE, TG, TD, TF;
TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
TE = BYTWJ(&(W[0]), TD);
TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
TG = BYTWJ(&(W[TWVL * 20]), TF);
TH = VSUB(TE, TG);
T1r = VADD(TE, TG);
}
{
V Tm, To, Tl, Tn;
Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0]));
Tm = BYTWJ(&(W[TWVL * 22]), Tl);
Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
To = BYTWJ(&(W[TWVL * 2]), Tn);
Tp = VSUB(Tm, To);
T1x = VADD(Tm, To);
}
TZ = VSUB(TH, TC);
T16 = VSUB(T9, Te);
T17 = VSUB(Tk, Tp);
T10 = VSUB(TS, TN);
Tf = VADD(T9, Te);
Tq = VADD(Tk, Tp);
Tr = VADD(Tf, Tq);
T1O = VADD(T1u, T1v);
T1P = VADD(T1x, T1y);
T1Q = VADD(T1O, T1P);
T1w = VSUB(T1u, T1v);
T1z = VSUB(T1x, T1y);
T1A = VADD(T1w, T1z);
TI = VADD(TC, TH);
TT = VADD(TN, TS);
TU = VADD(TI, TT);
T1L = VADD(T1n, T1o);
T1M = VADD(T1q, T1r);
T1N = VADD(T1L, T1M);
T1p = VSUB(T1n, T1o);
T1s = VSUB(T1q, T1r);
T1t = VADD(T1p, T1s);
}
Ts = VADD(T4, Tr);
TV = VADD(Tx, TU);
ST(&(x[WS(rs, 5)]), VFNMSI(TV, Ts), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 15)]), VFMAI(TV, Ts), ms, &(x[WS(rs, 1)]));
{
V T1T, T1R, T1S, T1X, T1Z, T1V, T1W, T1Y, T1U;
T1T = VSUB(T1N, T1Q);
T1R = VADD(T1N, T1Q);
T1S = VFNMS(LDK(KP250000000), T1R, T1K);
T1V = VSUB(T1L, T1M);
T1W = VSUB(T1O, T1P);
T1X = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1W, T1V));
T1Z = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1V, T1W));
ST(&(x[0]), VADD(T1K, T1R), ms, &(x[0]));
T1Y = VFNMS(LDK(KP559016994), T1T, T1S);
ST(&(x[WS(rs, 8)]), VFNMSI(T1Z, T1Y), ms, &(x[0]));
ST(&(x[WS(rs, 12)]), VFMAI(T1Z, T1Y), ms, &(x[0]));
T1U = VFMA(LDK(KP559016994), T1T, T1S);
ST(&(x[WS(rs, 4)]), VFMAI(T1X, T1U), ms, &(x[0]));
ST(&(x[WS(rs, 16)]), VFNMSI(T1X, T1U), ms, &(x[0]));
}
{
V T1D, T1B, T1C, T1H, T1J, T1F, T1G, T1I, T1E;
T1D = VSUB(T1t, T1A);
T1B = VADD(T1t, T1A);
T1C = VFNMS(LDK(KP250000000), T1B, T1m);
T1F = VSUB(T1w, T1z);
T1G = VSUB(T1p, T1s);
T1H = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1G, T1F));
T1J = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1F, T1G));
ST(&(x[WS(rs, 10)]), VADD(T1m, T1B), ms, &(x[0]));
T1I = VFMA(LDK(KP559016994), T1D, T1C);
ST(&(x[WS(rs, 6)]), VFNMSI(T1J, T1I), ms, &(x[0]));
ST(&(x[WS(rs, 14)]), VFMAI(T1J, T1I), ms, &(x[0]));
T1E = VFNMS(LDK(KP559016994), T1D, T1C);
ST(&(x[WS(rs, 2)]), VFMAI(T1H, T1E), ms, &(x[0]));
ST(&(x[WS(rs, 18)]), VFNMSI(T1H, T1E), ms, &(x[0]));
}
{
V T11, T18, T1g, T1d, T15, T1f, TY, T1c;
T11 = VFMA(LDK(KP618033988), T10, TZ);
T18 = VFMA(LDK(KP618033988), T17, T16);
T1g = VFNMS(LDK(KP618033988), T16, T17);
T1d = VFNMS(LDK(KP618033988), TZ, T10);
{
V T13, T14, TW, TX;
T13 = VFNMS(LDK(KP250000000), TU, Tx);
T14 = VSUB(TT, TI);
T15 = VFNMS(LDK(KP559016994), T14, T13);
T1f = VFMA(LDK(KP559016994), T14, T13);
TW = VFNMS(LDK(KP250000000), Tr, T4);
TX = VSUB(Tf, Tq);
TY = VFMA(LDK(KP559016994), TX, TW);
T1c = VFNMS(LDK(KP559016994), TX, TW);
}
{
V T12, T19, T1i, T1j;
T12 = VFMA(LDK(KP951056516), T11, TY);
T19 = VFMA(LDK(KP951056516), T18, T15);
ST(&(x[WS(rs, 1)]), VFNMSI(T19, T12), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 19)]), VFMAI(T19, T12), ms, &(x[WS(rs, 1)]));
T1i = VFMA(LDK(KP951056516), T1d, T1c);
T1j = VFMA(LDK(KP951056516), T1g, T1f);
ST(&(x[WS(rs, 13)]), VFNMSI(T1j, T1i), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(T1j, T1i), ms, &(x[WS(rs, 1)]));
}
{
V T1a, T1b, T1e, T1h;
T1a = VFNMS(LDK(KP951056516), T11, TY);
T1b = VFNMS(LDK(KP951056516), T18, T15);
ST(&(x[WS(rs, 9)]), VFNMSI(T1b, T1a), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 11)]), VFMAI(T1b, T1a), ms, &(x[WS(rs, 1)]));
T1e = VFNMS(LDK(KP951056516), T1d, T1c);
T1h = VFNMS(LDK(KP951056516), T1g, T1f);
ST(&(x[WS(rs, 17)]), VFNMSI(T1h, T1e), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFMAI(T1h, T1e), ms, &(x[WS(rs, 1)]));
}
}
}
}
VLEAVE();
}
static void n1bv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
{
DVK(KP250000000, +0.250000000000000000000000000000000000000000000);
DVK(KP559016994, +0.559016994374947424102293417182819058860154590);
DVK(KP587785252, +0.587785252292473129168705954639072768597652438);
DVK(KP951056516, +0.951056516295153572116439333379382143405698634);
{
INT i;
const R *xi;
R *xo;
xi = ii;
xo = io;
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) {
V Tl, Ty, T7, Te, Tw, Tt, Tz, TA, TB, Tg, Th, Tm, Tj, Tk;
Tj = LD(&(xi[0]), ivs, &(xi[0]));
Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
Tl = VSUB(Tj, Tk);
Ty = VADD(Tj, Tk);
{
V T3, Tr, Td, Tv, T6, Ts, Ta, Tu;
{
V T1, T2, Tb, Tc;
T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
T3 = VSUB(T1, T2);
Tr = VADD(T1, T2);
Tb = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
Td = VSUB(Tb, Tc);
Tv = VADD(Tb, Tc);
}
{
V T4, T5, T8, T9;
T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
T6 = VSUB(T4, T5);
Ts = VADD(T4, T5);
T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
Ta = VSUB(T8, T9);
Tu = VADD(T8, T9);
}
T7 = VSUB(T3, T6);
Te = VSUB(Ta, Td);
Tw = VSUB(Tu, Tv);
Tt = VSUB(Tr, Ts);
Tz = VADD(Tr, Ts);
TA = VADD(Tu, Tv);
TB = VADD(Tz, TA);
Tg = VADD(T3, T6);
Th = VADD(Ta, Td);
Tm = VADD(Tg, Th);
}
ST(&(xo[WS(os, 5)]), VADD(Tl, Tm), ovs, &(xo[WS(os, 1)]));
ST(&(xo[0]), VADD(Ty, TB), ovs, &(xo[0]));
{
V Tf, Tq, To, Tp, Ti, Tn;
Tf = VBYI(VFMA(LDK(KP951056516), T7, VMUL(LDK(KP587785252), Te)));
Tq = VBYI(VFNMS(LDK(KP951056516), Te, VMUL(LDK(KP587785252), T7)));
Ti = VMUL(LDK(KP559016994), VSUB(Tg, Th));
Tn = VFNMS(LDK(KP250000000), Tm, Tl);
To = VADD(Ti, Tn);
Tp = VSUB(Tn, Ti);
ST(&(xo[WS(os, 1)]), VADD(Tf, To), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VADD(Tq, Tp), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VSUB(To, Tf), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 3)]), VSUB(Tp, Tq), ovs, &(xo[WS(os, 1)]));
}
{
V Tx, TG, TE, TF, TC, TD;
Tx = VBYI(VFNMS(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt)));
TG = VBYI(VFMA(LDK(KP951056516), Tt, VMUL(LDK(KP587785252), Tw)));
TC = VFNMS(LDK(KP250000000), TB, Ty);
TD = VMUL(LDK(KP559016994), VSUB(Tz, TA));
TE = VSUB(TC, TD);
TF = VADD(TD, TC);
ST(&(xo[WS(os, 2)]), VADD(Tx, TE), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VADD(TG, TF), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VSUB(TE, Tx), ovs, &(xo[0]));
ST(&(xo[WS(os, 4)]), VSUB(TF, TG), ovs, &(xo[0]));
}
}
}
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 n1bv_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(20, is), MAKE_VOLATILE_STRIDE(20, 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, Tz, Tj;
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));
Tz = VFNMS(LDK(KP250000000), Ty, Tr);
ST(&(xo[0]), VADD(Tr, Ty), ovs, &(xo[0]));
Tj = VFNMS(LDK(KP250000000), Ti, T3);
ST(&(xo[WS(os, 5)]), VADD(T3, Ti), ovs, &(xo[WS(os, 1)]));
{
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);
ST(&(xo[WS(os, 4)]), VFNMSI(TG, TF), ovs, &(xo[0]));
ST(&(xo[WS(os, 6)]), VFMAI(TG, TF), ovs, &(xo[0]));
ST(&(xo[WS(os, 8)]), VFMAI(TE, TB), ovs, &(xo[0]));
ST(&(xo[WS(os, 2)]), VFNMSI(TE, TB), ovs, &(xo[0]));
ST(&(xo[WS(os, 3)]), VFMAI(Tq, Tp), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 7)]), VFNMSI(Tq, Tp), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 9)]), VFNMSI(To, Tl), ovs, &(xo[WS(os, 1)]));
ST(&(xo[WS(os, 1)]), VFMAI(To, Tl), ovs, &(xo[WS(os, 1)]));
}
}
}
}
}
}
VLEAVE();
}
static void t1fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms)
{
DVK(KP707106781, +0.707106781186547524400844362104849039284835938);
{
INT m;
R *x;
x = ri;
for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(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 = BYTWJ(&(W[TWVL * 6]), T2);
Ti = BYTWJ(&(W[TWVL * 2]), Th);
Tk = BYTWJ(&(W[TWVL * 10]), Tj);
T6 = BYTWJ(&(W[0]), T5);
T8 = BYTWJ(&(W[TWVL * 8]), T7);
Tb = BYTWJ(&(W[TWVL * 12]), Ta);
Td = BYTWJ(&(W[TWVL * 4]), Tc);
{
V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts;
Tq = VADD(T1, T3);
T4 = VSUB(T1, T3);
Tr = VADD(Ti, Tk);
Tl = VSUB(Ti, Tk);
Tt = VADD(T6, T8);
T9 = VSUB(T6, T8);
Tu = VADD(Tb, Td);
Te = VSUB(Tb, Td);
Tw = VSUB(Tq, Tr);
Ts = VADD(Tq, Tr);
{
V Tx, Tv, Tm, Tf;
Tx = VSUB(Tu, Tt);
Tv = VADD(Tt, Tu);
Tm = VSUB(Te, T9);
Tf = VADD(T9, Te);
{
V Tp, Tn, To, Tg;
ST(&(x[WS(rs, 2)]), VFMAI(Tx, Tw), ms, &(x[0]));
ST(&(x[WS(rs, 6)]), VFNMSI(Tx, Tw), ms, &(x[0]));
ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0]));
Tp = VFMA(LDK(KP707106781), Tm, Tl);
Tn = VFNMS(LDK(KP707106781), Tm, Tl);
To = VFNMS(LDK(KP707106781), Tf, T4);
Tg = VFMA(LDK(KP707106781), Tf, T4);
ST(&(x[WS(rs, 5)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 3)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)]));
}
}
}
}
}
}
VLEAVE();
}
static void t1bv_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 = ii;
for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(12, rs)) {
V T1, Tt, T6, T7, TB, Tq, TC, TD, T9, Tu, Te, Tf, Tx, Tl, Ty;
V Tz;
{
V T5, T3, T4, T2;
T1 = LD(&(x[0]), ms, &(x[0]));
T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0]));
T5 = BYTW(&(W[TWVL * 14]), T4);
T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0]));
T3 = BYTW(&(W[TWVL * 6]), T2);
Tt = VSUB(T3, T5);
T6 = VADD(T3, T5);
T7 = VFNMS(LDK(KP500000000), T6, T1);
}
{
V Tn, Tp, Tm, TA, To;
Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)]));
Tn = BYTW(&(W[0]), Tm);
TA = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)]));
TB = BYTW(&(W[TWVL * 16]), TA);
To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)]));
Tp = BYTW(&(W[TWVL * 8]), To);
Tq = VSUB(Tn, Tp);
TC = VADD(Tn, Tp);
TD = VFNMS(LDK(KP500000000), TC, TB);
}
{
V Td, Tb, T8, Tc, Ta;
T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0]));
T9 = BYTW(&(W[TWVL * 10]), T8);
Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0]));
Td = BYTW(&(W[TWVL * 2]), Tc);
Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0]));
Tb = BYTW(&(W[TWVL * 18]), Ta);
Tu = VSUB(Tb, Td);
Te = VADD(Tb, Td);
Tf = VFNMS(LDK(KP500000000), Te, T9);
}
{
V Ti, Tk, Th, Tw, Tj;
Th = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)]));
Ti = BYTW(&(W[TWVL * 12]), Th);
Tw = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)]));
Tx = BYTW(&(W[TWVL * 4]), Tw);
Tj = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)]));
Tk = BYTW(&(W[TWVL * 20]), Tj);
Tl = VSUB(Ti, Tk);
Ty = VADD(Ti, Tk);
Tz = VFNMS(LDK(KP500000000), Ty, Tx);
}
{
V Ts, TG, TF, TH;
{
V Tg, Tr, Tv, TE;
Tg = VSUB(T7, Tf);
Tr = VMUL(LDK(KP866025403), VSUB(Tl, Tq));
Ts = VSUB(Tg, Tr);
TG = VADD(Tg, Tr);
Tv = VMUL(LDK(KP866025403), VSUB(Tt, Tu));
TE = VSUB(Tz, TD);
TF = VBYI(VADD(Tv, TE));
TH = VBYI(VSUB(TE, Tv));
}
ST(&(x[WS(rs, 11)]), VSUB(Ts, TF), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 5)]), VADD(TG, TH), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 1)]), VADD(Ts, TF), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 7)]), VSUB(TG, TH), ms, &(x[WS(rs, 1)]));
}
{
V TS, TW, TV, TX;
{
V TQ, TR, TT, TU;
TQ = VADD(T1, T6);
TR = VADD(T9, Te);
TS = VSUB(TQ, TR);
TW = VADD(TQ, TR);
TT = VADD(Tx, Ty);
TU = VADD(TB, TC);
TV = VBYI(VSUB(TT, TU));
TX = VADD(TT, TU);
}
ST(&(x[WS(rs, 3)]), VSUB(TS, TV), ms, &(x[WS(rs, 1)]));
ST(&(x[0]), VADD(TW, TX), ms, &(x[0]));
ST(&(x[WS(rs, 9)]), VADD(TS, TV), ms, &(x[WS(rs, 1)]));
ST(&(x[WS(rs, 6)]), VSUB(TW, TX), ms, &(x[0]));
}
{
V TK, TO, TN, TP;
{
V TI, TJ, TL, TM;
TI = VADD(Tl, Tq);
TJ = VADD(Tt, Tu);
TK = VBYI(VMUL(LDK(KP866025403), VSUB(TI, TJ)));
TO = VBYI(VMUL(LDK(KP866025403), VADD(TJ, TI)));
TL = VADD(T7, Tf);
TM = VADD(Tz, TD);
TN = VSUB(TL, TM);
TP = VADD(TL, TM);
}
ST(&(x[WS(rs, 2)]), VADD(TK, TN), ms, &(x[0]));
ST(&(x[WS(rs, 8)]), VSUB(TP, TO), ms, &(x[0]));
ST(&(x[WS(rs, 10)]), VSUB(TN, TK), ms, &(x[0]));
ST(&(x[WS(rs, 4)]), VADD(TO, TP), ms, &(x[0]));
}
}
}
VLEAVE();
}
