+/*
+ * Copyright (C) 1995-2008 University of Karlsruhe. All right reserved.
+ *
+ * This file is part of libFirm.
+ *
+ * This file may be distributed and/or modified under the terms of the
+ * GNU General Public License version 2 as published by the Free Software
+ * Foundation and appearing in the file LICENSE.GPL included in the
+ * packaging of this file.
+ *
+ * Licensees holding valid libFirm Professional Edition licenses may use
+ * this file in accordance with the libFirm Commercial License.
+ * Agreement provided with the Software.
+ *
+ * This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
+ * WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE.
+ */
+
/**
- * Author: Daniel Grund
- * Date: 28.02.2006
- * Copyright: (c) Universitaet Karlsruhe
- * Licence: This file protected by GPL - GNU GENERAL PUBLIC LICENSE.
+ * @file
+ * @brief ILP based copy minimization.
+ * @author Daniel Grund
+ * @date 28.02.2006
+ * @version $Id$
*
* ILP formalization using G=(V, E, Q):
- * - 1 class of variables: equal color vars
+ * - 2 class of variables: coloring vars x_ic and equal color vars y_ij
* - Path constraints
- * - Clique path constraints
+ * - Clique-star constraints
*
*
* \min \sum_{ (i,j) \in Q } w_ij y_ij
*
- * y_ij = 1 (i,j) \in E
+ * \sum_c x_nc = 1 n \in N, c \in C
*
- * \sum_c y_nc = |C| - 1 n \in N, c \in C
+ * x_nc = 0 n \in N, c \not\in C(n)
*
- * y_nc = 1 n \in N, c \not\in C(n)
+ * \sum x_nc <= 1 x_nc \in Clique \in AllCliques, c \in C
*
* \sum_{e \in p} y_e >= 1 p \in P path constraints
*
- * \sum_{e \in cp} y_e >= |cp| - 1 cp \in CP clique-path constraints
+ * \sum_{e \in cs} y_e >= |cs| - 1 cs \in CP clique-star constraints
*
- * y_ij \in N, w_ij \in R^+
+ * x_nc, y_ij \in N, w_ij \in R^+
*/
-
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif /* HAVE_CONFIG_H */
+#include "firm_config.h"
+
#ifdef WITH_ILP
+#include "bitset.h"
+#include "raw_bitset.h"
+#include "pdeq.h"
+
+#include "irtools.h"
+#include "irgwalk.h"
#include "becopyilp_t.h"
#include "beifg_t.h"
-#include "irtools.h"
+#include "besched_t.h"
#define DEBUG_LVL 1
typedef struct _local_env_t {
- firm_dbg_module_t *dbg;
double time_limit;
int first_x_var, last_x_var;
+ int n_colors;
+ bitset_t *normal_colors;
pmap *nr_2_irn;
+ DEBUG_ONLY(firm_dbg_module_t *dbg;)
} local_env_t;
static void build_coloring_cstr(ilp_env_t *ienv) {
- be_ifg_t *ifg = ienv->co->cenv->ifg;
- void *iter = be_ifg_nodes_iter_alloca(ifg);
+ be_ifg_t *ifg = ienv->co->cenv->ifg;
+ void *iter = be_ifg_nodes_iter_alloca(ifg);
bitset_t *colors;
ir_node *irn;
char buf[16];
be_ifg_foreach_node(ifg, iter, irn)
if (!sr_is_removed(ienv->sr, irn)) {
- int col, cst_idx;
- arch_register_req_t req;
- int curr_node_color = get_irn_col(ienv->co, irn);
- int node_nr = (int)get_irn_node_nr(irn);
+ bitset_pos_t col;
+ int cst_idx;
+ const arch_register_req_t *req;
+ int curr_node_color = get_irn_col(irn);
+ int node_nr = (int)get_irn_idx(irn);
local_env_t *lenv = ienv->env;
pmap_insert(lenv->nr_2_irn, INT_TO_PTR(node_nr), irn);
- arch_get_register_req(ienv->co->aenv, &req, irn, -1);
+ req = arch_get_register_req(irn, -1);
+
+ bitset_clear_all(colors);
/* get assignable colors */
- if (arch_register_req_is(&req, limited))
- req.limited(req.limited_env, colors);
- else
- arch_put_non_ignore_regs(ienv->co->aenv, req.cls, colors);
+ if (arch_register_req_is(req, limited)) {
+ rbitset_copy_to_bitset(req->limited, colors);
+ } else {
+ bitset_copy(colors, lenv->normal_colors);
+ }
/* add the coloring constraint */
cst_idx = lpp_add_cst(ienv->lp, NULL, lpp_equal, 1.0);
bitset_foreach(colors, col) {
int var_idx = lpp_add_var(ienv->lp, name_cdd(buf, 'x', node_nr, col), lpp_binary, 0.0);
- lpp_set_start_value(ienv->lp, var_idx, (col == curr_node_color) ? 1.0 : 0.0);
+ lpp_set_start_value(ienv->lp, var_idx, (col == (unsigned) curr_node_color) ? 1.0 : 0.0);
lpp_set_factor_fast(ienv->lp, cst_idx, var_idx, 1);
lenv->last_x_var = var_idx;
}
static void build_interference_cstr(ilp_env_t *ienv) {
- lpp_t *lpp = ienv->lp;
- be_ifg_t *ifg = ienv->co->cenv->ifg;
- int n_colors = arch_register_class_n_regs(ienv->co->cls);
+ lpp_t *lpp = ienv->lp;
+ local_env_t *lenv = ienv->env;
+ be_ifg_t *ifg = ienv->co->cenv->ifg;
+ int n_colors = lenv->n_colors;
int i, col;
void *iter = be_ifg_cliques_iter_alloca(ifg);
- ir_node *clique = alloca(sizeof(*clique) * n_colors);
+ ir_node **clique = alloca(sizeof(*clique) * n_colors);
int size;
char buf[16];
/* for each maximal clique */
- be_ifg_foreach_clique(ifg, iter, &clique, &size) {
+ be_ifg_foreach_clique(ifg, iter, clique, &size) {
+ int realsize = 0;
- if (size < 2)
+ for (i=0; i<size; ++i)
+ if (!sr_is_removed(ienv->sr, clique[i]))
+ ++realsize;
+
+ if (realsize < 2)
continue;
/* for all colors */
int cst_idx = lpp_add_cst(lpp, NULL, lpp_less, 1.0);
/* for each member of this clique */
- for (i=0; i<size, ++i) {
+ for (i=0; i<size; ++i) {
ir_node *irn = clique[i];
if (!sr_is_removed(ienv->sr, irn)) {
- int var_idx = lpp_get_var_idx(lpp, name_cdd(buf, 'x', (int)get_irn_node_nr(irn), col));
+ int var_idx = lpp_get_var_idx(lpp, name_cdd(buf, 'x', (int)get_irn_idx(irn), col));
lpp_set_factor_fast(lpp, cst_idx, var_idx, 1);
}
}
}
}
+
+/**
+ * TODO: Remove the dependency of the opt-units data structure
+ * by walking over all affinity edges. Graph structure
+ * does not provide this walker, yet.
+ */
static void build_affinity_cstr(ilp_env_t *ienv) {
+ local_env_t *lenv = ienv->env;
+ int n_colors = lenv->n_colors;
unit_t *curr;
- int n_colors = arch_register_class_n_regs(ienv->co->cls);
/* for all optimization units */
list_for_each_entry(unit_t, curr, &ienv->co->units, units) {
int root_col, arg_col;
root = curr->nodes[0];
- root_nr = (int) get_irn_node_nr(root);
- root_col = get_irn_col(ienv->co, root);
+ root_nr = (int) get_irn_idx(root);
+ root_col = get_irn_col(root);
for (i = 1; i < curr->node_count; ++i) {
arg = curr->nodes[i];
- arg_nr = (int) get_irn_node_nr(arg);
- arg_col = get_irn_col(ienv->co, arg);
+ arg_nr = (int) get_irn_idx(arg);
+ arg_col = get_irn_col(arg);
/* add a new affinity variable */
y_idx = lpp_add_var(ienv->lp, name_cdd_sorted(buf, 'y', root_nr, arg_nr), lpp_binary, curr->costs[i]);
lpp_set_factor_fast(ienv->lp, cst_idx, root_idx, 1.0);
lpp_set_factor_fast(ienv->lp, cst_idx, arg_idx, -1.0);
- lpp_set_factor_fast(ienv->lp, cst_idx, root_idx, -1.0);
+ lpp_set_factor_fast(ienv->lp, cst_idx, y_idx, -1.0);
}
}
}
}
-static void build_path_cstr(ilp_env_t *ienv) {
+/**
+ * Helping stuff for build_clique_star_cstr
+ */
+typedef struct _edge_t {
+ ir_node *n1, *n2;
+} edge_t;
+
+static int compare_edge_t(const void *k1, const void *k2, size_t size) {
+ const edge_t *e1 = k1;
+ const edge_t *e2 = k2;
+ (void) size;
+
+ return ! (e1->n1 == e2->n1 && e1->n2 == e2->n2);
+}
+
+#define HASH_EDGE(e) (hash_irn((e)->n1) ^ hash_irn((e)->n2))
+
+static INLINE edge_t *add_edge(set *edges, ir_node *n1, ir_node *n2, int *counter) {
+ edge_t new_edge;
+
+ if (PTR_TO_INT(n1) < PTR_TO_INT(n2)) {
+ new_edge.n1 = n1;
+ new_edge.n2 = n2;
+ } else {
+ new_edge.n1 = n2;
+ new_edge.n2 = n1;
+ }
+ (*counter)++;
+ return set_insert(edges, &new_edge, sizeof(new_edge), HASH_EDGE(&new_edge));
+}
+
+static INLINE edge_t *find_edge(set *edges, ir_node *n1, ir_node *n2) {
+ edge_t new_edge;
+
+ if (PTR_TO_INT(n1) < PTR_TO_INT(n2)) {
+ new_edge.n1 = n1;
+ new_edge.n2 = n2;
+ } else {
+ new_edge.n1 = n2;
+ new_edge.n2 = n1;
+ }
+ return set_find(edges, &new_edge, sizeof(new_edge), HASH_EDGE(&new_edge));
+}
+
+static INLINE void remove_edge(set *edges, ir_node *n1, ir_node *n2, int *counter) {
+ edge_t new_edge, *e;
+
+ if (PTR_TO_INT(n1) < PTR_TO_INT(n2)) {
+ new_edge.n1 = n1;
+ new_edge.n2 = n2;
+ } else {
+ new_edge.n1 = n2;
+ new_edge.n2 = n1;
+ }
+ e = set_find(edges, &new_edge, sizeof(new_edge), HASH_EDGE(&new_edge));
+ if (e) {
+ e->n1 = NULL;
+ e->n2 = NULL;
+ (*counter)--;
+ }
+}
+
+#define pset_foreach(pset, irn) for(irn=pset_first(pset); irn; irn=pset_next(pset))
+
+/**
+ * Search for an interference clique and an external node
+ * with affinity edges to all nodes of the clique.
+ * At most 1 node of the clique can be colored equally with the external node.
+ */
+static void build_clique_star_cstr(ilp_env_t *ienv) {
+ affinity_node_t *aff;
+
+ /* for each node with affinity edges */
+ co_gs_foreach_aff_node(ienv->co, aff) {
+ struct obstack ob;
+ neighb_t *nbr;
+ const ir_node *center = aff->irn;
+ ir_node **nodes;
+ set *edges;
+ int i, o, n_nodes, n_edges;
+
+ if (arch_irn_is(aff->irn, ignore))
+ continue;
+
+ obstack_init(&ob);
+ edges = new_set(compare_edge_t, 8);
+
+ /* get all affinity neighbours */
+ n_nodes = 0;
+ co_gs_foreach_neighb(aff, nbr) {
+ if (!arch_irn_is(nbr->irn, ignore)) {
+ obstack_ptr_grow(&ob, nbr->irn);
+ ++n_nodes;
+ }
+ }
+ nodes = obstack_finish(&ob);
+
+ /* get all interference edges between these */
+ n_edges = 0;
+ for (i=0; i<n_nodes; ++i)
+ for (o=0; o<i; ++o)
+ if (be_ifg_connected(ienv->co->cenv->ifg, nodes[i], nodes[o]))
+ add_edge(edges, nodes[i], nodes[o], &n_edges);
+
+ /* cover all these interference edges with maximal cliques */
+ while (n_edges) {
+ edge_t *e;
+ pset *clique = pset_new_ptr(8);
+ int growed;
+
+ /* get 2 starting nodes to form a clique */
+ for (e=set_first(edges); !e->n1; e=set_next(edges))
+ /*nothing*/ ;
+
+ /* we could be stepped out of the loop before the set iterated to the end */
+ set_break(edges);
+
+ pset_insert_ptr(clique, e->n1);
+ pset_insert_ptr(clique, e->n2);
+ remove_edge(edges, e->n1, e->n2, &n_edges);
+
+ /* while the clique is growing */
+ do {
+ growed = 0;
+
+ /* search for a candidate to extend the clique */
+ for (i=0; i<n_nodes; ++i) {
+ ir_node *member, *cand = nodes[i];
+ int is_cand;
+
+ /* if its already in the clique try the next */
+ if (pset_find_ptr(clique, cand))
+ continue;
+
+ /* are there all necessary interferences? */
+ is_cand = 1;
+ pset_foreach(clique, member) {
+ if (!find_edge(edges, cand, member)) {
+ is_cand = 0;
+ pset_break(clique);
+ break;
+ }
+ }
+
+ /* now we know if we have a clique extender */
+ if (is_cand) {
+ /* first remove all covered edges */
+ pset_foreach(clique, member)
+ remove_edge(edges, cand, member, &n_edges);
+
+ /* insert into clique */
+ pset_insert_ptr(clique, cand);
+ growed = 1;
+ break;
+ }
+ }
+ } while (growed);
+
+ /* now the clique is maximal. Finally add the constraint */
+ {
+ ir_node *member;
+ int var_idx, cst_idx, center_nr, member_nr;
+ char buf[16];
+
+ cst_idx = lpp_add_cst(ienv->lp, NULL, lpp_greater, pset_count(clique)-1);
+ center_nr = get_irn_idx(center);
+
+ pset_foreach(clique, member) {
+ member_nr = get_irn_idx(member);
+ var_idx = lpp_get_var_idx(ienv->lp, name_cdd_sorted(buf, 'y', center_nr, member_nr));
+ lpp_set_factor_fast(ienv->lp, cst_idx, var_idx, 1.0);
+ }
+ }
+ del_pset(clique);
+ }
+
+ del_set(edges);
+ obstack_free(&ob, NULL);
+ }
}
-static void build_clique_path_cstr(ilp_env_t *ienv) {
+static void extend_path(ilp_env_t *ienv, pdeq *path, const ir_node *irn) {
+ be_ifg_t *ifg = ienv->co->cenv->ifg;
+ int i, len;
+ ir_node **curr_path;
+ affinity_node_t *aff;
+ neighb_t *nbr;
+
+ /* do not walk backwards or in circles */
+ if (pdeq_contains(path, irn))
+ return;
+
+ if (arch_irn_is(irn, ignore))
+ return;
+
+ /* insert the new irn */
+ pdeq_putr(path, irn);
+
+
+
+ /* check for forbidden interferences */
+ len = pdeq_len(path);
+ curr_path = alloca(len * sizeof(*curr_path));
+ pdeq_copyl(path, (const void **)curr_path);
+
+ for (i=1; i<len; ++i)
+ if (be_ifg_connected(ifg, irn, curr_path[i]))
+ goto end;
+
+
+
+ /* check for terminating interference */
+ if (be_ifg_connected(ifg, irn, curr_path[0])) {
+
+ /* One node is not a path. */
+ /* And a path of length 2 is covered by a clique star constraint. */
+ if (len > 2) {
+ /* finally build the constraint */
+ int cst_idx = lpp_add_cst(ienv->lp, NULL, lpp_greater, 1.0);
+ for (i=1; i<len; ++i) {
+ char buf[16];
+ int nr_1 = get_irn_idx(curr_path[i-1]);
+ int nr_2 = get_irn_idx(curr_path[i]);
+ int var_idx = lpp_get_var_idx(ienv->lp, name_cdd_sorted(buf, 'y', nr_1, nr_2));
+ lpp_set_factor_fast(ienv->lp, cst_idx, var_idx, 1.0);
+ }
+ }
+
+ /* this path cannot be extended anymore */
+ goto end;
+ }
+
+
+
+ /* recursively extend the path */
+ aff = get_affinity_info(ienv->co, irn);
+ co_gs_foreach_neighb(aff, nbr)
+ extend_path(ienv, path, nbr->irn);
+
+
+end:
+ /* remove the irn */
+ pdeq_getr(path);
+
+}
+
+/**
+ * Search a path of affinity edges, whose ends are connected
+ * by an interference edge and there are no other interference
+ * edges in between.
+ * Then at least one of these affinity edges must break.
+ */
+static void build_path_cstr(ilp_env_t *ienv) {
+ affinity_node_t *aff_info;
+
+ /* for each node with affinity edges */
+ co_gs_foreach_aff_node(ienv->co, aff_info) {
+ pdeq *path = new_pdeq();
+
+ extend_path(ienv, path, aff_info->irn);
+
+ del_pdeq(path);
+ }
}
static void ilp2_build(ilp_env_t *ienv) {
build_coloring_cstr(ienv);
build_interference_cstr(ienv);
build_affinity_cstr(ienv);
+ build_clique_star_cstr(ienv);
build_path_cstr(ienv);
- build_clique_path_cstr(ienv);
lower_bound = co_get_lower_bound(ienv->co) - co_get_inevit_copy_costs(ienv->co);
lpp_set_bound(ienv->lp, lower_bound);
static void ilp2_apply(ilp_env_t *ienv) {
local_env_t *lenv = ienv->env;
- double sol[];
- lpp_sol_state_t state;
- int count;
-
- count = lenv->last_x_var - lenv->first_x_var + 1;
- sol = xmalloc(count * sizeof(sol[0]));
- state = lpp_get_solution(ienv->lp, sol, lenv->first_x_var, lenv->last_x_var);
- if (state != lpp_optimal) {
- printf("WARNING %s: Solution state is not 'optimal': %d\n", ienv->co->name, state);
- assert(state >= lpp_feasible && "The solution should at least be feasible!");
- }
+ int i;
+
+ /* first check if there was sth. to optimize */
+ if (lenv->first_x_var >= 0) {
+ int count = lenv->last_x_var - lenv->first_x_var + 1;
+ double *sol = XMALLOCN(double, count);
+ lpp_sol_state_t state = lpp_get_solution(ienv->lp, sol, lenv->first_x_var, lenv->last_x_var);
- for (i=0; i<count; ++i) {
- char c;
- int nodenr, color;
- char var_name[16];
+ if (state != lpp_optimal) {
+ printf("WARNING %s: Solution state is not 'optimal': %d\n", ienv->co->name, state);
+ assert(state >= lpp_feasible && "The solution should at least be feasible!");
+ }
+
+ for (i=0; i<count; ++i) {
+ int nodenr, color;
+ char var_name[16];
- if (sol[i] > 1-EPSILON) { /* split variable name into components */
- lpp_get_var_name(ienv->lp, lenv->first_x_var+i, var_name, sizeof(var_name));
+ if (sol[i] > 1-EPSILON) { /* split variable name into components */
+ lpp_get_var_name(ienv->lp, lenv->first_x_var+i, var_name, sizeof(var_name));
- if (sscanf(var_name, "x_%d_%d", &nodenr, &color) == 2) {
- ir_node *irn = pmap_get(lenv->nr_2_irn, INT_TO_PTR(nodenr));
- assert(irn && "This node number must be present in the map");
+ if (sscanf(var_name, "x_%d_%d", &nodenr, &color) == 2) {
+ ir_node *irn = pmap_get(lenv->nr_2_irn, INT_TO_PTR(nodenr));
+ assert(irn && "This node number must be present in the map");
- set_irn_col(ienv->co, irn, color);
- } else
- assert(0 && "This should be a x-var");
+ set_irn_col(ienv->co, irn, color);
+ } else
+ assert(0 && "This should be a x-var");
+ }
}
+
+ xfree(sol);
}
#ifdef COPYOPT_STAT
#endif
}
-int co_solve_ilp2(copy_opt_t *co, double time_limit) {
+int co_solve_ilp2(copy_opt_t *co) {
lpp_sol_state_t sol_state;
ilp_env_t *ienv;
local_env_t my;
- my.time_limit = time_limit;
+ ASSERT_OU_AVAIL(co); //See build_clique_st
+ ASSERT_GS_AVAIL(co);
+
+ my.time_limit = 0;
my.first_x_var = -1;
my.last_x_var = -1;
my.nr_2_irn = pmap_create();
- my.dbg = firm_dbg_register("ir.be.coilp2");
- firm_dbg_set_mask(my.dbg, DEBUG_LVL);
+ FIRM_DBG_REGISTER(my.dbg, "firm.be.coilp2");
+
+ my.normal_colors = bitset_alloca(arch_register_class_n_regs(co->cls));
+ bitset_clear_all(my.normal_colors);
+ arch_put_non_ignore_regs(co->cls, my.normal_colors);
+ my.n_colors = bitset_popcnt(my.normal_colors);
ienv = new_ilp_env(co, ilp2_build, ilp2_apply, &my);
#else /* WITH_ILP */
-static void only_that_you_can_compile_without_WITH_ILP_defined(void) {
+static INLINE void only_that_you_can_compile_without_WITH_ILP_defined(void) {
}
#endif /* WITH_ILP */
projects / libfirm / blobdiff