|
d@0
|
1 /*
|
|
d@0
|
2 * Copyright (c) 2003, 2007-8 Matteo Frigo
|
|
d@0
|
3 * Copyright (c) 2003, 2007-8 Massachusetts Institute of Technology
|
|
d@0
|
4 *
|
|
d@0
|
5 * This program is free software; you can redistribute it and/or modify
|
|
d@0
|
6 * it under the terms of the GNU General Public License as published by
|
|
d@0
|
7 * the Free Software Foundation; either version 2 of the License, or
|
|
d@0
|
8 * (at your option) any later version.
|
|
d@0
|
9 *
|
|
d@0
|
10 * This program is distributed in the hope that it will be useful,
|
|
d@0
|
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
d@0
|
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
d@0
|
13 * GNU General Public License for more details.
|
|
d@0
|
14 *
|
|
d@0
|
15 * You should have received a copy of the GNU General Public License
|
|
d@0
|
16 * along with this program; if not, write to the Free Software
|
|
d@0
|
17 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
|
d@0
|
18 *
|
|
d@0
|
19 */
|
|
d@0
|
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 2 -name X(spu_n2fv_2) */
|
|
d@0
|
21
|
|
d@0
|
22 /*
|
|
d@0
|
23 * This function contains 2 FP additions, 0 FP multiplications,
|
|
d@0
|
24 * (or, 2 additions, 0 multiplications, 0 fused multiply/add),
|
|
d@0
|
25 * 7 stack variables, 0 constants, and 5 memory accesses
|
|
d@0
|
26 */
|
|
d@0
|
27 #include "fftw-spu.h"
|
|
d@0
|
28
|
|
d@0
|
29 void X(spu_n2fv_2) (const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) {
|
|
d@0
|
30 INT i;
|
|
d@0
|
31 const R *xi;
|
|
d@0
|
32 R *xo;
|
|
d@0
|
33 xi = ri;
|
|
d@0
|
34 xo = ro;
|
|
d@0
|
35 for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
|
|
d@0
|
36 V T1, T2, T3, T4;
|
|
d@0
|
37 T1 = LD(&(xi[0]), ivs, &(xi[0]));
|
|
d@0
|
38 T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)]));
|
|
d@0
|
39 T3 = VSUB(T1, T2);
|
|
d@0
|
40 STM2(&(xo[2]), T3, ovs, &(xo[2]));
|
|
d@0
|
41 T4 = VADD(T1, T2);
|
|
d@0
|
42 STM2(&(xo[0]), T4, ovs, &(xo[0]));
|
|
d@0
|
43 STN2(&(xo[0]), T4, T3, ovs);
|
|
d@0
|
44 }
|
|
d@0
|
45 }
|
