/* * 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 * */ /* Plans for handling vector transform loops. These are *just* the loops, and rely on child plans for the actual RDFT2s. They form a wrapper around solvers that don't have apply functions for non-null vectors. vrank-geq1-rdft2 plans also recursively handle the case of multi-dimensional vectors, obviating the need for most solvers to deal with this. We can also play games here, such as reordering the vector loops. Each vrank-geq1-rdft2 plan reduces the vector rank by 1, picking out a dimension determined by the vecloop_dim field of the solver. */ #include "rdft.h" typedef struct { solver super; int vecloop_dim; const int *buddies; int nbuddies; } S; typedef struct { plan_rdft2 super; plan *cld; INT vl; INT rvs, cvs; const S *solver; } P; static void apply(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; INT i, vl = ego->vl; INT rvs = ego->rvs, cvs = ego->cvs; rdft2apply cldapply = ((plan_rdft2 *) ego->cld)->apply; for (i = 0; i < vl; ++i) { cldapply(ego->cld, r0 + i * rvs, r1 + i * rvs, cr + i * cvs, ci + i * cvs); } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; p->print(p, "(rdft2-vrank>=1-x%D/%d%(%p%))", ego->vl, s->vecloop_dim, ego->cld); } static int pickdim(const S *ego, const tensor *vecsz, int oop, int *dp) { return X(pickdim)(ego->vecloop_dim, ego->buddies, ego->nbuddies, vecsz, oop, dp); } static int applicable0(const solver *ego_, const problem *p_, int *dp) { const S *ego = (const S *) ego_; const problem_rdft2 *p = (const problem_rdft2 *) p_; if (FINITE_RNK(p->vecsz->rnk) && p->vecsz->rnk > 0 && pickdim(ego, p->vecsz, p->r0 != p->cr, dp)) { if (p->r0 != p->cr) return 1; /* can always operate out-of-place */ return(X(rdft2_inplace_strides)(p, *dp)); } return 0; } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *)ego_; if (!applicable0(ego_, p_, dp)) return 0; /* fftw2 behavior */ if (NO_VRANK_SPLITSP(plnr) && (ego->vecloop_dim != ego->buddies[0])) return 0; if (NO_UGLYP(plnr)) { const problem_rdft2 *p = (const problem_rdft2 *) p_; iodim *d = p->vecsz->dims + *dp; /* Heuristic: if the transform is multi-dimensional, and the vector stride is less than the transform size, then we probably want to use a rank>=2 plan first in order to combine this vector with the transform-dimension vectors. */ if (p->sz->rnk > 1 && X(imin)(X(iabs)(d->is), X(iabs)(d->os)) < X(rdft2_tensor_max_index)(p->sz, p->kind) ) return 0; /* Heuristic: don't use a vrank-geq1 for rank-0 vrank-1 transforms, since this case is better handled by rank-0 solvers. */ if (p->sz->rnk == 0 && p->vecsz->rnk == 1) return 0; if (NO_NONTHREADEDP(plnr)) return 0; /* prefer threaded version */ } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft2 *p; P *pln; plan *cld; int vdim; iodim *d; INT rvs, cvs; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &vdim)) return (plan *) 0; p = (const problem_rdft2 *) p_; d = p->vecsz->dims + vdim; A(d->n > 1); /* or else, p->ri + d->is etc. are invalid */ X(rdft2_strides)(p->kind, d, &rvs, &cvs); cld = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)( X(tensor_copy)(p->sz), X(tensor_copy_except)(p->vecsz, vdim), TAINT(p->r0, rvs), TAINT(p->r1, rvs), TAINT(p->cr, cvs), TAINT(p->ci, cvs), p->kind)); if (!cld) return (plan *) 0; pln = MKPLAN_RDFT2(P, &padt, apply); pln->cld = cld; pln->vl = d->n; pln->rvs = rvs; pln->cvs = cvs; pln->solver = ego; X(ops_zero)(&pln->super.super.ops); pln->super.super.ops.other = 3.14159; /* magic to prefer codelet loops */ X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); if (p->sz->rnk != 1 || (p->sz->dims[0].n > 128)) pln->super.super.pcost = pln->vl * cld->pcost; return &(pln->super.super); } static solver *mksolver(int vecloop_dim, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->vecloop_dim = vecloop_dim; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(rdft2_vrank_geq1_register)(planner *p) { int i; /* FIXME: Should we try other vecloop_dim values? */ static const int buddies[] = { 1, -1 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); }