furnace/extern/fftw/rdft/direct2.c

172 lines
4.3 KiB
C
Raw Normal View History

/*
* 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
*
*/
/* direct RDFT2 R2HC/HC2R solver, if we have a codelet */
#include "rdft/rdft.h"
typedef struct {
solver super;
const kr2c_desc *desc;
kr2c k;
} S;
typedef struct {
plan_rdft2 super;
stride rs, cs;
INT vl;
INT ivs, ovs;
kr2c k;
const S *slv;
INT ilast;
} P;
static void apply(const plan *ego_, R *r0, R *r1, R *cr, R *ci)
{
const P *ego = (const P *) ego_;
ASSERT_ALIGNED_DOUBLE;
ego->k(r0, r1, cr, ci,
ego->rs, ego->cs, ego->cs,
ego->vl, ego->ivs, ego->ovs);
}
static void apply_r2hc(const plan *ego_, R *r0, R *r1, R *cr, R *ci)
{
const P *ego = (const P *) ego_;
INT i, vl = ego->vl, ovs = ego->ovs;
ASSERT_ALIGNED_DOUBLE;
ego->k(r0, r1, cr, ci,
ego->rs, ego->cs, ego->cs,
vl, ego->ivs, ovs);
for (i = 0; i < vl; ++i, ci += ovs)
ci[0] = ci[ego->ilast] = 0;
}
static void destroy(plan *ego_)
{
P *ego = (P *) ego_;
X(stride_destroy)(ego->rs);
X(stride_destroy)(ego->cs);
}
static void print(const plan *ego_, printer *p)
{
const P *ego = (const P *) ego_;
const S *s = ego->slv;
p->print(p, "(rdft2-%s-direct-%D%v \"%s\")",
X(rdft_kind_str)(s->desc->genus->kind), s->desc->n,
ego->vl, s->desc->nam);
}
static int applicable(const solver *ego_, const problem *p_)
{
const S *ego = (const S *) ego_;
const kr2c_desc *desc = ego->desc;
const problem_rdft2 *p = (const problem_rdft2 *) p_;
INT vl;
INT ivs, ovs;
return (
1
&& p->sz->rnk == 1
&& p->vecsz->rnk <= 1
&& p->sz->dims[0].n == desc->n
&& p->kind == desc->genus->kind
/* check strides etc */
&& X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs)
&& (0
/* can operate out-of-place */
|| p->r0 != p->cr
/*
* can compute one transform in-place, no matter
* what the strides are.
*/
|| p->vecsz->rnk == 0
/* can operate in-place as long as strides are the same */
|| X(rdft2_inplace_strides)(p, RNK_MINFTY)
)
);
}
static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr)
{
const S *ego = (const S *) ego_;
P *pln;
const problem_rdft2 *p;
iodim *d;
int r2hc_kindp;
static const plan_adt padt = {
X(rdft2_solve), X(null_awake), print, destroy
};
UNUSED(plnr);
if (!applicable(ego_, p_))
return (plan *)0;
p = (const problem_rdft2 *) p_;
r2hc_kindp = R2HC_KINDP(p->kind);
A(r2hc_kindp || HC2R_KINDP(p->kind));
pln = MKPLAN_RDFT2(P, &padt, p->kind == R2HC ? apply_r2hc : apply);
d = p->sz->dims;
pln->k = ego->k;
pln->rs = X(mkstride)(d->n, r2hc_kindp ? d->is : d->os);
pln->cs = X(mkstride)(d->n, r2hc_kindp ? d->os : d->is);
X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs);
/* Nyquist freq., if any */
pln->ilast = (d->n % 2) ? 0 : (d->n/2) * d->os;
pln->slv = ego;
X(ops_zero)(&pln->super.super.ops);
X(ops_madd2)(pln->vl / ego->desc->genus->vl,
&ego->desc->ops,
&pln->super.super.ops);
if (p->kind == R2HC)
pln->super.super.ops.other += 2 * pln->vl; /* + 2 stores */
pln->super.super.could_prune_now_p = 1;
return &(pln->super.super);
}
/* constructor */
solver *X(mksolver_rdft2_direct)(kr2c k, const kr2c_desc *desc)
{
static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 };
S *slv = MKSOLVER(S, &sadt);
slv->k = k;
slv->desc = desc;
return &(slv->super);
}