195 lines
6.3 KiB
C
195 lines
6.3 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 10:46:24 EDT 2021 */
|
||
|
|
|
||
|
|
#include "rdft/codelet-rdft.h"
|
||
|
|
|
||
|
|
#if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
|
||
|
|
|
||
|
|
/* Generated by: ../../../genfft/gen_r2cf.native -fma -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cfII_10 -dft-II -include rdft/scalar/r2cfII.h */
|
||
|
|
|
||
|
|
/*
|
||
|
|
* This function contains 32 FP additions, 18 FP multiplications,
|
||
|
|
* (or, 14 additions, 0 multiplications, 18 fused multiply/add),
|
||
|
|
* 21 stack variables, 4 constants, and 20 memory accesses
|
||
|
|
*/
|
||
|
|
#include "rdft/scalar/r2cfII.h"
|
||
|
|
|
||
|
|
static void r2cfII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
|
||
|
|
{
|
||
|
|
DK(KP951056516, +0.951056516295153572116439333379382143405698634);
|
||
|
|
DK(KP559016994, +0.559016994374947424102293417182819058860154590);
|
||
|
|
DK(KP250000000, +0.250000000000000000000000000000000000000000000);
|
||
|
|
DK(KP618033988, +0.618033988749894848204586834365638117720309180);
|
||
|
|
{
|
||
|
|
INT i;
|
||
|
|
for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) {
|
||
|
|
E T1, To, T8, Tt, Ta, Ts, Te, Tq, Th, Tn;
|
||
|
|
T1 = R0[0];
|
||
|
|
To = R1[WS(rs, 2)];
|
||
|
|
{
|
||
|
|
E T2, T3, T4, T5, T6, T7;
|
||
|
|
T2 = R0[WS(rs, 2)];
|
||
|
|
T3 = R0[WS(rs, 3)];
|
||
|
|
T4 = T2 - T3;
|
||
|
|
T5 = R0[WS(rs, 4)];
|
||
|
|
T6 = R0[WS(rs, 1)];
|
||
|
|
T7 = T5 - T6;
|
||
|
|
T8 = T4 + T7;
|
||
|
|
Tt = T5 + T6;
|
||
|
|
Ta = T4 - T7;
|
||
|
|
Ts = T2 + T3;
|
||
|
|
}
|
||
|
|
{
|
||
|
|
E Tc, Td, Tm, Tf, Tg, Tl;
|
||
|
|
Tc = R1[0];
|
||
|
|
Td = R1[WS(rs, 4)];
|
||
|
|
Tm = Tc + Td;
|
||
|
|
Tf = R1[WS(rs, 1)];
|
||
|
|
Tg = R1[WS(rs, 3)];
|
||
|
|
Tl = Tf + Tg;
|
||
|
|
Te = Tc - Td;
|
||
|
|
Tq = Tm + Tl;
|
||
|
|
Th = Tf - Tg;
|
||
|
|
Tn = Tl - Tm;
|
||
|
|
}
|
||
|
|
Cr[WS(csr, 2)] = T1 + T8;
|
||
|
|
Ci[WS(csi, 2)] = Tn - To;
|
||
|
|
{
|
||
|
|
E Ti, Tk, Tb, Tj, T9;
|
||
|
|
Ti = FMA(KP618033988, Th, Te);
|
||
|
|
Tk = FNMS(KP618033988, Te, Th);
|
||
|
|
T9 = FNMS(KP250000000, T8, T1);
|
||
|
|
Tb = FMA(KP559016994, Ta, T9);
|
||
|
|
Tj = FNMS(KP559016994, Ta, T9);
|
||
|
|
Cr[WS(csr, 4)] = FNMS(KP951056516, Ti, Tb);
|
||
|
|
Cr[WS(csr, 3)] = FMA(KP951056516, Tk, Tj);
|
||
|
|
Cr[0] = FMA(KP951056516, Ti, Tb);
|
||
|
|
Cr[WS(csr, 1)] = FNMS(KP951056516, Tk, Tj);
|
||
|
|
}
|
||
|
|
{
|
||
|
|
E Tu, Tw, Tr, Tv, Tp;
|
||
|
|
Tu = FMA(KP618033988, Tt, Ts);
|
||
|
|
Tw = FNMS(KP618033988, Ts, Tt);
|
||
|
|
Tp = FMA(KP250000000, Tn, To);
|
||
|
|
Tr = FMA(KP559016994, Tq, Tp);
|
||
|
|
Tv = FNMS(KP559016994, Tq, Tp);
|
||
|
|
Ci[0] = -(FMA(KP951056516, Tu, Tr));
|
||
|
|
Ci[WS(csi, 3)] = FMA(KP951056516, Tw, Tv);
|
||
|
|
Ci[WS(csi, 4)] = FMS(KP951056516, Tu, Tr);
|
||
|
|
Ci[WS(csi, 1)] = FNMS(KP951056516, Tw, Tv);
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
|
||
|
|
static const kr2c_desc desc = { 10, "r2cfII_10", { 14, 0, 18, 0 }, &GENUS };
|
||
|
|
|
||
|
|
void X(codelet_r2cfII_10) (planner *p) { X(kr2c_register) (p, r2cfII_10, &desc);
|
||
|
|
}
|
||
|
|
|
||
|
|
#else
|
||
|
|
|
||
|
|
/* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cfII_10 -dft-II -include rdft/scalar/r2cfII.h */
|
||
|
|
|
||
|
|
/*
|
||
|
|
* This function contains 32 FP additions, 12 FP multiplications,
|
||
|
|
* (or, 26 additions, 6 multiplications, 6 fused multiply/add),
|
||
|
|
* 21 stack variables, 4 constants, and 20 memory accesses
|
||
|
|
*/
|
||
|
|
#include "rdft/scalar/r2cfII.h"
|
||
|
|
|
||
|
|
static void r2cfII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
|
||
|
|
{
|
||
|
|
DK(KP250000000, +0.250000000000000000000000000000000000000000000);
|
||
|
|
DK(KP587785252, +0.587785252292473129168705954639072768597652438);
|
||
|
|
DK(KP951056516, +0.951056516295153572116439333379382143405698634);
|
||
|
|
DK(KP559016994, +0.559016994374947424102293417182819058860154590);
|
||
|
|
{
|
||
|
|
INT i;
|
||
|
|
for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) {
|
||
|
|
E T1, To, T8, Tq, T9, Tp, Te, Ts, Th, Tn;
|
||
|
|
T1 = R0[0];
|
||
|
|
To = R1[WS(rs, 2)];
|
||
|
|
{
|
||
|
|
E T2, T3, T4, T5, T6, T7;
|
||
|
|
T2 = R0[WS(rs, 2)];
|
||
|
|
T3 = R0[WS(rs, 3)];
|
||
|
|
T4 = T2 - T3;
|
||
|
|
T5 = R0[WS(rs, 4)];
|
||
|
|
T6 = R0[WS(rs, 1)];
|
||
|
|
T7 = T5 - T6;
|
||
|
|
T8 = T4 + T7;
|
||
|
|
Tq = T5 + T6;
|
||
|
|
T9 = KP559016994 * (T4 - T7);
|
||
|
|
Tp = T2 + T3;
|
||
|
|
}
|
||
|
|
{
|
||
|
|
E Tc, Td, Tm, Tf, Tg, Tl;
|
||
|
|
Tc = R1[0];
|
||
|
|
Td = R1[WS(rs, 4)];
|
||
|
|
Tm = Tc + Td;
|
||
|
|
Tf = R1[WS(rs, 1)];
|
||
|
|
Tg = R1[WS(rs, 3)];
|
||
|
|
Tl = Tf + Tg;
|
||
|
|
Te = Tc - Td;
|
||
|
|
Ts = KP559016994 * (Tm + Tl);
|
||
|
|
Th = Tf - Tg;
|
||
|
|
Tn = Tl - Tm;
|
||
|
|
}
|
||
|
|
Cr[WS(csr, 2)] = T1 + T8;
|
||
|
|
Ci[WS(csi, 2)] = Tn - To;
|
||
|
|
{
|
||
|
|
E Ti, Tk, Tb, Tj, Ta;
|
||
|
|
Ti = FMA(KP951056516, Te, KP587785252 * Th);
|
||
|
|
Tk = FNMS(KP587785252, Te, KP951056516 * Th);
|
||
|
|
Ta = FNMS(KP250000000, T8, T1);
|
||
|
|
Tb = T9 + Ta;
|
||
|
|
Tj = Ta - T9;
|
||
|
|
Cr[WS(csr, 4)] = Tb - Ti;
|
||
|
|
Cr[WS(csr, 3)] = Tj + Tk;
|
||
|
|
Cr[0] = Tb + Ti;
|
||
|
|
Cr[WS(csr, 1)] = Tj - Tk;
|
||
|
|
}
|
||
|
|
{
|
||
|
|
E Tr, Tw, Tu, Tv, Tt;
|
||
|
|
Tr = FMA(KP951056516, Tp, KP587785252 * Tq);
|
||
|
|
Tw = FNMS(KP587785252, Tp, KP951056516 * Tq);
|
||
|
|
Tt = FMA(KP250000000, Tn, To);
|
||
|
|
Tu = Ts + Tt;
|
||
|
|
Tv = Tt - Ts;
|
||
|
|
Ci[0] = -(Tr + Tu);
|
||
|
|
Ci[WS(csi, 3)] = Tw + Tv;
|
||
|
|
Ci[WS(csi, 4)] = Tr - Tu;
|
||
|
|
Ci[WS(csi, 1)] = Tv - Tw;
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
|
||
|
|
static const kr2c_desc desc = { 10, "r2cfII_10", { 26, 6, 6, 0 }, &GENUS };
|
||
|
|
|
||
|
|
void X(codelet_r2cfII_10) (planner *p) { X(kr2c_register) (p, r2cfII_10, &desc);
|
||
|
|
}
|
||
|
|
|
||
|
|
#endif
|