167 lines
4.8 KiB
C
167 lines
4.8 KiB
C
/*
|
|
* Copyright (c) 2003, 2007-14 Matteo Frigo
|
|
* Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
|
|
*
|
|
* This program is free software; you can redistribute it and/or modify
|
|
* it under the terms of the GNU General Public License as published by
|
|
* the Free Software Foundation; either version 2 of the License, or
|
|
* (at your option) any later version.
|
|
*
|
|
* This program is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
* GNU General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU General Public License
|
|
* along with this program; if not, write to the Free Software
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
*
|
|
*/
|
|
|
|
/* This file was automatically generated --- DO NOT EDIT */
|
|
/* Generated on Tue Sep 14 10:46:12 EDT 2021 */
|
|
|
|
#include "rdft/codelet-rdft.h"
|
|
|
|
#if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
|
|
|
|
/* Generated by: ../../../genfft/gen_hc2hc.native -fma -compact -variables 4 -pipeline-latency 4 -n 3 -dit -name hf_3 -include rdft/scalar/hf.h */
|
|
|
|
/*
|
|
* This function contains 16 FP additions, 14 FP multiplications,
|
|
* (or, 6 additions, 4 multiplications, 10 fused multiply/add),
|
|
* 15 stack variables, 2 constants, and 12 memory accesses
|
|
*/
|
|
#include "rdft/scalar/hf.h"
|
|
|
|
static void hf_3(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms)
|
|
{
|
|
DK(KP866025403, +0.866025403784438646763723170752936183471402627);
|
|
DK(KP500000000, +0.500000000000000000000000000000000000000000000);
|
|
{
|
|
INT m;
|
|
for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) {
|
|
E T1, Tl, T7, Th, Td, Tj;
|
|
T1 = cr[0];
|
|
Tl = ci[0];
|
|
{
|
|
E T3, T6, T4, Tg, T2, T5;
|
|
T3 = cr[WS(rs, 1)];
|
|
T6 = ci[WS(rs, 1)];
|
|
T2 = W[0];
|
|
T4 = T2 * T3;
|
|
Tg = T2 * T6;
|
|
T5 = W[1];
|
|
T7 = FMA(T5, T6, T4);
|
|
Th = FNMS(T5, T3, Tg);
|
|
}
|
|
{
|
|
E T9, Tc, Ta, Ti, T8, Tb;
|
|
T9 = cr[WS(rs, 2)];
|
|
Tc = ci[WS(rs, 2)];
|
|
T8 = W[2];
|
|
Ta = T8 * T9;
|
|
Ti = T8 * Tc;
|
|
Tb = W[3];
|
|
Td = FMA(Tb, Tc, Ta);
|
|
Tj = FNMS(Tb, T9, Ti);
|
|
}
|
|
{
|
|
E Tk, Te, Tf, To, Tm, Tn;
|
|
Tk = Th - Tj;
|
|
Te = T7 + Td;
|
|
Tf = FNMS(KP500000000, Te, T1);
|
|
cr[0] = T1 + Te;
|
|
ci[0] = FNMS(KP866025403, Tk, Tf);
|
|
cr[WS(rs, 1)] = FMA(KP866025403, Tk, Tf);
|
|
To = Td - T7;
|
|
Tm = Th + Tj;
|
|
Tn = FNMS(KP500000000, Tm, Tl);
|
|
cr[WS(rs, 2)] = FMS(KP866025403, To, Tn);
|
|
ci[WS(rs, 2)] = Tm + Tl;
|
|
ci[WS(rs, 1)] = FMA(KP866025403, To, Tn);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
static const tw_instr twinstr[] = {
|
|
{ TW_FULL, 1, 3 },
|
|
{ TW_NEXT, 1, 0 }
|
|
};
|
|
|
|
static const hc2hc_desc desc = { 3, "hf_3", twinstr, &GENUS, { 6, 4, 10, 0 } };
|
|
|
|
void X(codelet_hf_3) (planner *p) {
|
|
X(khc2hc_register) (p, hf_3, &desc);
|
|
}
|
|
#else
|
|
|
|
/* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 3 -dit -name hf_3 -include rdft/scalar/hf.h */
|
|
|
|
/*
|
|
* This function contains 16 FP additions, 12 FP multiplications,
|
|
* (or, 10 additions, 6 multiplications, 6 fused multiply/add),
|
|
* 15 stack variables, 2 constants, and 12 memory accesses
|
|
*/
|
|
#include "rdft/scalar/hf.h"
|
|
|
|
static void hf_3(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms)
|
|
{
|
|
DK(KP866025403, +0.866025403784438646763723170752936183471402627);
|
|
DK(KP500000000, +0.500000000000000000000000000000000000000000000);
|
|
{
|
|
INT m;
|
|
for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) {
|
|
E T1, Ti, T6, Te, Tb, Tf, Tc, Tj;
|
|
T1 = cr[0];
|
|
Ti = ci[0];
|
|
{
|
|
E T3, T5, T2, T4;
|
|
T3 = cr[WS(rs, 1)];
|
|
T5 = ci[WS(rs, 1)];
|
|
T2 = W[0];
|
|
T4 = W[1];
|
|
T6 = FMA(T2, T3, T4 * T5);
|
|
Te = FNMS(T4, T3, T2 * T5);
|
|
}
|
|
{
|
|
E T8, Ta, T7, T9;
|
|
T8 = cr[WS(rs, 2)];
|
|
Ta = ci[WS(rs, 2)];
|
|
T7 = W[2];
|
|
T9 = W[3];
|
|
Tb = FMA(T7, T8, T9 * Ta);
|
|
Tf = FNMS(T9, T8, T7 * Ta);
|
|
}
|
|
Tc = T6 + Tb;
|
|
Tj = Te + Tf;
|
|
{
|
|
E Td, Tg, Th, Tk;
|
|
cr[0] = T1 + Tc;
|
|
Td = FNMS(KP500000000, Tc, T1);
|
|
Tg = KP866025403 * (Te - Tf);
|
|
ci[0] = Td - Tg;
|
|
cr[WS(rs, 1)] = Td + Tg;
|
|
ci[WS(rs, 2)] = Tj + Ti;
|
|
Th = KP866025403 * (Tb - T6);
|
|
Tk = FNMS(KP500000000, Tj, Ti);
|
|
cr[WS(rs, 2)] = Th - Tk;
|
|
ci[WS(rs, 1)] = Th + Tk;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
static const tw_instr twinstr[] = {
|
|
{ TW_FULL, 1, 3 },
|
|
{ TW_NEXT, 1, 0 }
|
|
};
|
|
|
|
static const hc2hc_desc desc = { 3, "hf_3", twinstr, &GENUS, { 10, 6, 6, 0 } };
|
|
|
|
void X(codelet_hf_3) (planner *p) {
|
|
X(khc2hc_register) (p, hf_3, &desc);
|
|
}
|
|
#endif
|