2
* Copyright (c) 2003, 2007-8 Matteo Frigo
3
* Copyright (c) 2003, 2007-8 Massachusetts Institute of Technology
5
* This program is free software; you can redistribute it and/or modify
6
* it under the terms of the GNU General Public License as published by
7
* the Free Software Foundation; either version 2 of the License, or
8
* (at your option) any later version.
10
* This program is distributed in the hope that it will be useful,
11
* but WITHOUT ANY WARRANTY; without even the implied warranty of
12
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
* GNU General Public License for more details.
15
* You should have received a copy of the GNU General Public License
16
* along with this program; if not, write to the Free Software
17
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
20
/* Generated by: ../../genfft/gen_notw_c -standalone -fma -reorder-insns -simd -compact -variables 100000 -with-ostride 2 -include fftw-spu.h -store-multiple 2 -n 14 -name X(spu_n2fv_14) */
23
* This function contains 74 FP additions, 48 FP multiplications,
24
* (or, 32 additions, 6 multiplications, 42 fused multiply/add),
25
* 93 stack variables, 6 constants, and 35 memory accesses
29
void X(spu_n2fv_14) (const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) {
30
DVK(KP801937735, +0.801937735804838252472204639014890102331838324);
31
DVK(KP974927912, +0.974927912181823607018131682993931217232785801);
32
DVK(KP554958132, +0.554958132087371191422194871006410481067288862);
33
DVK(KP900968867, +0.900968867902419126236102319507445051165919162);
34
DVK(KP692021471, +0.692021471630095869627814897002069140197260599);
35
DVK(KP356895867, +0.356895867892209443894399510021300583399127187);
41
for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
42
V TH, T3, Ta, Ts, TV, TN, TK, TW, Tt, To, Th, Tu, TU, TQ, Tp;
43
V TC, Tx, TR, T14, TZ, T1, T2, TI, T6, TJ, T9, TP, Tn, TO, Tk;
44
V TM, Tg, TL, Td, T4, T5, T7, T8, Tl, Tm, Ti, Tj, Te, Tf, Tb;
45
V Tc, T1a, T1b, T19, T1c, Tr, Tw, Tq, Tv, T16, T18, T15, T17, T1d, T1e;
46
V T1f, T1i, Tz, TB, Ty, TA, T1g, TT, TY, TS, TX, T1h, T1j, T11, T13;
47
V T10, T12, T1k, TE, TG, TD, TF, T1l, T1m;
48
T1 = LD(&(xi[0]), ivs, &(xi[0]));
49
T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)]));
52
T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0]));
53
T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)]));
56
T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0]));
57
T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)]));
60
Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0]));
61
Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
64
Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0]));
65
Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)]));
68
Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0]));
69
Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)]));
72
Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0]));
73
Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)]));
88
Tp = VFNMS(LDK(KP356895867), Ta, To);
89
TC = VFNMS(LDK(KP356895867), To, Th);
90
Tx = VFNMS(LDK(KP356895867), Th, Ta);
91
TR = VFNMS(LDK(KP356895867), TQ, TN);
92
T14 = VFNMS(LDK(KP356895867), TN, TK);
93
TZ = VFNMS(LDK(KP356895867), TK, TQ);
94
T19 = VADD(T3, VADD(Ta, VADD(Th, To)));
95
STM2(&(xo[14]), T19, ovs, &(xo[2]));
96
T1a = VADD(TH, VADD(TK, VADD(TN, TQ)));
97
STM2(&(xo[0]), T1a, ovs, &(xo[0]));
98
Tq = VFNMS(LDK(KP692021471), Tp, Th);
99
Tr = VFNMS(LDK(KP900968867), Tq, T3);
100
Tv = VFMA(LDK(KP554958132), Tu, Tt);
101
Tw = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tv, Ts));
102
T1b = VFNMSI(Tw, Tr);
103
STM2(&(xo[10]), T1b, ovs, &(xo[2]));
105
STM2(&(xo[18]), T1c, ovs, &(xo[2]));
106
T15 = VFNMS(LDK(KP692021471), T14, TQ);
107
T16 = VFNMS(LDK(KP900968867), T15, TH);
108
T17 = VFNMS(LDK(KP554958132), TU, TW);
109
T18 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T17, TV));
110
T1d = VFMAI(T18, T16);
111
STM2(&(xo[12]), T1d, ovs, &(xo[0]));
112
STN2(&(xo[12]), T1d, T19, ovs);
113
T1e = VFNMSI(T18, T16);
114
STM2(&(xo[16]), T1e, ovs, &(xo[0]));
115
STN2(&(xo[16]), T1e, T1c, ovs);
116
Ty = VFNMS(LDK(KP692021471), Tx, To);
117
Tz = VFNMS(LDK(KP900968867), Ty, T3);
118
TA = VFMA(LDK(KP554958132), Tt, Ts);
119
TB = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TA, Tu));
120
T1f = VFNMSI(TB, Tz);
121
STM2(&(xo[26]), T1f, ovs, &(xo[2]));
123
STM2(&(xo[2]), T1g, ovs, &(xo[2]));
124
STN2(&(xo[0]), T1a, T1g, ovs);
125
TS = VFNMS(LDK(KP692021471), TR, TK);
126
TT = VFNMS(LDK(KP900968867), TS, TH);
127
TX = VFMA(LDK(KP554958132), TW, TV);
128
TY = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TX, TU));
130
STM2(&(xo[8]), T1h, ovs, &(xo[0]));
131
STN2(&(xo[8]), T1h, T1b, ovs);
132
T1i = VFNMSI(TY, TT);
133
STM2(&(xo[20]), T1i, ovs, &(xo[0]));
134
T10 = VFNMS(LDK(KP692021471), TZ, TN);
135
T11 = VFNMS(LDK(KP900968867), T10, TH);
136
T12 = VFMA(LDK(KP554958132), TV, TU);
137
T13 = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), T12, TW));
138
T1j = VFMAI(T13, T11);
139
STM2(&(xo[4]), T1j, ovs, &(xo[0]));
140
T1k = VFNMSI(T13, T11);
141
STM2(&(xo[24]), T1k, ovs, &(xo[0]));
142
STN2(&(xo[24]), T1k, T1f, ovs);
143
TD = VFNMS(LDK(KP692021471), TC, Ta);
144
TE = VFNMS(LDK(KP900968867), TD, T3);
145
TF = VFNMS(LDK(KP554958132), Ts, Tu);
146
TG = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TF, Tt));
147
T1l = VFNMSI(TG, TE);
148
STM2(&(xo[22]), T1l, ovs, &(xo[2]));
149
STN2(&(xo[20]), T1i, T1l, ovs);
151
STM2(&(xo[6]), T1m, ovs, &(xo[2]));
152
STN2(&(xo[4]), T1j, T1m, ovs);
added
removed
cell/spu/spu_n2fv_14.spuc
