2 * Copyright (C) 1995-2007 University of Karlsruhe. All right reserved.
4 * This file is part of libFirm.
6 * This file may be distributed and/or modified under the terms of the
7 * GNU General Public License version 2 as published by the Free Software
8 * Foundation and appearing in the file LICENSE.GPL included in the
9 * packaging of this file.
11 * Licensees holding valid libFirm Professional Edition licenses may use
12 * this file in accordance with the libFirm Commercial License.
13 * Agreement provided with the Software.
15 * This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
16 * WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR
22 * @brief Optimizations for a whole ir graph, i.e., a procedure.
23 * @author Christian Schaefer, Goetz Lindenmaier, Sebastian Felis,
34 #include "irgraph_t.h"
47 #include "pdeq.h" /* Fuer code placement */
52 #include "irbackedge_t.h"
59 #include "iredges_t.h"
62 /*------------------------------------------------------------------*/
63 /* apply optimizations of iropt to all nodes. */
64 /*------------------------------------------------------------------*/
67 * A wrapper around optimize_inplace_2() to be called from a walker.
69 static void optimize_in_place_wrapper (ir_node *n, void *env) {
70 ir_node *optimized = optimize_in_place_2(n);
71 if (optimized != n) exchange (n, optimized);
75 * Do local optimizations for a node.
77 * @param n the IR-node where to start. Typically the End node
80 * @note current_ir_graph must be set
82 static INLINE void do_local_optimize(ir_node *n) {
83 /* Handle graph state */
84 assert(get_irg_phase_state(current_ir_graph) != phase_building);
86 if (get_opt_global_cse())
87 set_irg_pinned(current_ir_graph, op_pin_state_floats);
88 set_irg_outs_inconsistent(current_ir_graph);
89 set_irg_doms_inconsistent(current_ir_graph);
90 set_irg_loopinfo_inconsistent(current_ir_graph);
92 /* Clean the value_table in irg for the CSE. */
93 del_identities(current_ir_graph->value_table);
94 current_ir_graph->value_table = new_identities();
96 /* walk over the graph */
97 irg_walk(n, firm_clear_link, optimize_in_place_wrapper, NULL);
100 /* Applies local optimizations (see iropt.h) to all nodes reachable from node n */
101 void local_optimize_node(ir_node *n) {
102 ir_graph *rem = current_ir_graph;
103 current_ir_graph = get_irn_irg(n);
105 do_local_optimize(n);
107 current_ir_graph = rem;
111 * Block-Walker: uses dominance depth to mark dead blocks.
113 static void kill_dead_blocks(ir_node *block, void *env) {
114 if (get_Block_dom_depth(block) < 0) {
116 * Note that the new dominance code correctly handles
117 * the End block, i.e. it is always reachable from Start
119 set_Block_dead(block);
123 /* Applies local optimizations (see iropt.h) to all nodes reachable from node n. */
124 void local_optimize_graph(ir_graph *irg) {
125 ir_graph *rem = current_ir_graph;
126 current_ir_graph = irg;
128 if (get_irg_dom_state(irg) == dom_consistent)
129 irg_block_walk_graph(irg, NULL, kill_dead_blocks, NULL);
131 do_local_optimize(get_irg_end(irg));
133 current_ir_graph = rem;
137 * Enqueue all users of a node to a wait queue.
138 * Handles mode_T nodes.
140 static void enqueue_users(ir_node *n, pdeq *waitq) {
141 const ir_edge_t *edge;
143 foreach_out_edge(n, edge) {
144 ir_node *succ = get_edge_src_irn(edge);
146 if (get_irn_link(succ) != waitq) {
147 pdeq_putr(waitq, succ);
148 set_irn_link(succ, waitq);
150 if (get_irn_mode(succ) == mode_T) {
151 /* A mode_T node has Proj's. Because most optimizations
152 run on the Proj's we have to enqueue them also. */
153 enqueue_users(succ, waitq);
159 * Data flow optimization walker.
160 * Optimizes all nodes and enqueue it's users
163 static void opt_walker(ir_node *n, void *env) {
167 optimized = optimize_in_place_2(n);
168 set_irn_link(optimized, NULL);
170 if (optimized != n) {
171 enqueue_users(n, waitq);
172 exchange(n, optimized);
176 /* Applies local optimizations to all nodes in the graph until fixpoint. */
177 void optimize_graph_df(ir_graph *irg) {
178 pdeq *waitq = new_pdeq();
179 int state = edges_activated(irg);
180 ir_graph *rem = current_ir_graph;
184 current_ir_graph = irg;
189 if (get_opt_global_cse())
190 set_irg_pinned(current_ir_graph, op_pin_state_floats);
192 /* Clean the value_table in irg for the CSE. */
193 del_identities(irg->value_table);
194 irg->value_table = new_identities();
196 if (get_irg_dom_state(irg) == dom_consistent)
197 irg_block_walk_graph(irg, NULL, kill_dead_blocks, NULL);
199 /* invalidate info */
200 set_irg_outs_inconsistent(irg);
201 set_irg_doms_inconsistent(irg);
202 set_irg_loopinfo_inconsistent(irg);
204 set_using_irn_link(irg);
206 /* walk over the graph, but don't touch keep-alives */
207 irg_walk(get_irg_end_block(irg), NULL, opt_walker, waitq);
209 end = get_irg_end(irg);
211 /* optimize keep-alives by removing superfluous ones */
212 for (i = get_End_n_keepalives(end) - 1; i >= 0; --i) {
213 ir_node *ka = get_End_keepalive(end, i);
215 if (irn_visited(ka) && !is_irn_keep(ka)) {
216 /* this node can be regularly visited, no need to keep it */
217 set_End_keepalive(end, i, get_irg_bad(irg));
220 /* now walk again and visit all not yet visited nodes */
221 set_irg_visited(current_ir_graph, get_irg_visited(irg) - 1);
222 irg_walk(get_irg_end(irg), NULL, opt_walker, waitq);
224 /* finish the wait queue */
225 while (! pdeq_empty(waitq)) {
226 ir_node *n = pdeq_getl(waitq);
228 opt_walker(n, waitq);
233 clear_using_irn_link(irg);
236 edges_deactivate(irg);
238 current_ir_graph = rem;
242 /*------------------------------------------------------------------*/
243 /* Routines for dead node elimination / copying garbage collection */
244 /* of the obstack. */
245 /*------------------------------------------------------------------*/
248 * Remember the new node in the old node by using a field all nodes have.
250 #define set_new_node(oldn, newn) set_irn_link(oldn, newn)
253 * Get this new node, before the old node is forgotten.
255 #define get_new_node(oldn) get_irn_link(oldn)
258 * Check if a new node was set.
260 #define has_new_node(n) (get_new_node(n) != NULL)
263 * We use the block_visited flag to mark that we have computed the
264 * number of useful predecessors for this block.
265 * Further we encode the new arity in this flag in the old blocks.
266 * Remembering the arity is useful, as it saves a lot of pointer
267 * accesses. This function is called for all Phi and Block nodes
271 compute_new_arity(ir_node *b) {
272 int i, res, irn_arity;
275 irg_v = get_irg_block_visited(current_ir_graph);
276 block_v = get_Block_block_visited(b);
277 if (block_v >= irg_v) {
278 /* we computed the number of preds for this block and saved it in the
280 return block_v - irg_v;
282 /* compute the number of good predecessors */
283 res = irn_arity = get_irn_arity(b);
284 for (i = 0; i < irn_arity; i++)
285 if (get_irn_opcode(get_irn_n(b, i)) == iro_Bad) res--;
286 /* save it in the flag. */
287 set_Block_block_visited(b, irg_v + res);
293 * Copies the node to the new obstack. The Ins of the new node point to
294 * the predecessors on the old obstack. For block/phi nodes not all
295 * predecessors might be copied. n->link points to the new node.
296 * For Phi and Block nodes the function allocates in-arrays with an arity
297 * only for useful predecessors. The arity is determined by counting
298 * the non-bad predecessors of the block.
300 * @param n The node to be copied
301 * @param env if non-NULL, the node number attribute will be copied to the new node
303 * Note: Also used for loop unrolling.
305 static void copy_node(ir_node *n, void *env) {
308 ir_op *op = get_irn_op(n);
310 /* The end node looses it's flexible in array. This doesn't matter,
311 as dead node elimination builds End by hand, inlineing doesn't use
313 /* assert(op == op_End || ((_ARR_DESCR(n->in))->cookie != ARR_F_MAGIC)); */
316 /* node copied already */
318 } else if (op == op_Block) {
320 new_arity = compute_new_arity(n);
321 n->attr.block.graph_arr = NULL;
323 block = get_nodes_block(n);
325 new_arity = compute_new_arity(block);
327 new_arity = get_irn_arity(n);
330 nn = new_ir_node(get_irn_dbg_info(n),
337 /* Copy the attributes. These might point to additional data. If this
338 was allocated on the old obstack the pointers now are dangling. This
339 frees e.g. the memory of the graph_arr allocated in new_immBlock. */
340 copy_node_attr(n, nn);
344 int copy_node_nr = env != NULL;
346 /* for easier debugging, we want to copy the node numbers too */
347 nn->node_nr = n->node_nr;
353 hook_dead_node_elim_subst(current_ir_graph, n, nn);
357 * Copies new predecessors of old node to new node remembered in link.
358 * Spare the Bad predecessors of Phi and Block nodes.
361 copy_preds(ir_node *n, void *env) {
365 nn = get_new_node(n);
368 /* Don't copy Bad nodes. */
370 irn_arity = get_irn_arity(n);
371 for (i = 0; i < irn_arity; i++) {
372 if (! is_Bad(get_irn_n(n, i))) {
373 set_irn_n (nn, j, get_new_node(get_irn_n(n, i)));
374 /*if (is_backedge(n, i)) set_backedge(nn, j);*/
378 /* repair the block visited flag from above misuse. Repair it in both
379 graphs so that the old one can still be used. */
380 set_Block_block_visited(nn, 0);
381 set_Block_block_visited(n, 0);
382 /* Local optimization could not merge two subsequent blocks if
383 in array contained Bads. Now it's possible.
384 We don't call optimize_in_place as it requires
385 that the fields in ir_graph are set properly. */
386 if ((get_opt_control_flow_straightening()) &&
387 (get_Block_n_cfgpreds(nn) == 1) &&
388 (get_irn_op(get_Block_cfgpred(nn, 0)) == op_Jmp)) {
389 ir_node *old = get_nodes_block(get_Block_cfgpred(nn, 0));
391 /* Jmp jumps into the block it is in -- deal self cycle. */
392 assert(is_Bad(get_new_node(get_irg_bad(current_ir_graph))));
393 exchange(nn, get_new_node(get_irg_bad(current_ir_graph)));
398 } else if (get_irn_op(n) == op_Phi) {
399 /* Don't copy node if corresponding predecessor in block is Bad.
400 The Block itself should not be Bad. */
401 block = get_nodes_block(n);
402 set_irn_n(nn, -1, get_new_node(block));
404 irn_arity = get_irn_arity(n);
405 for (i = 0; i < irn_arity; i++) {
406 if (! is_Bad(get_irn_n(block, i))) {
407 set_irn_n(nn, j, get_new_node(get_irn_n(n, i)));
408 /*if (is_backedge(n, i)) set_backedge(nn, j);*/
412 /* If the pre walker reached this Phi after the post walker visited the
413 block block_visited is > 0. */
414 set_Block_block_visited(get_nodes_block(n), 0);
415 /* Compacting the Phi's ins might generate Phis with only one
417 if (get_irn_arity(nn) == 1)
418 exchange(nn, get_irn_n(nn, 0));
420 irn_arity = get_irn_arity(n);
421 for (i = -1; i < irn_arity; i++)
422 set_irn_n (nn, i, get_new_node(get_irn_n(n, i)));
424 /* Now the new node is complete. We can add it to the hash table for CSE.
425 @@@ inlining aborts if we identify End. Why? */
426 if (get_irn_op(nn) != op_End)
427 add_identities(current_ir_graph->value_table, nn);
431 * Copies the graph recursively, compacts the keep-alives of the end node.
433 * @param irg the graph to be copied
434 * @param copy_node_nr If non-zero, the node number will be copied
436 static void copy_graph(ir_graph *irg, int copy_node_nr) {
437 ir_node *oe, *ne, *ob, *nb, *om, *nm; /* old end, new end, old bad, new bad, old NoMem, new NoMem */
438 ir_node *ka; /* keep alive */
442 /* Some nodes must be copied by hand, sigh */
443 vfl = get_irg_visited(irg);
444 set_irg_visited(irg, vfl + 1);
446 oe = get_irg_end(irg);
447 mark_irn_visited(oe);
448 /* copy the end node by hand, allocate dynamic in array! */
449 ne = new_ir_node(get_irn_dbg_info(oe),
456 /* Copy the attributes. Well, there might be some in the future... */
457 copy_node_attr(oe, ne);
458 set_new_node(oe, ne);
460 /* copy the Bad node */
461 ob = get_irg_bad(irg);
462 mark_irn_visited(ob);
463 nb = new_ir_node(get_irn_dbg_info(ob),
470 copy_node_attr(ob, nb);
471 set_new_node(ob, nb);
473 /* copy the NoMem node */
474 om = get_irg_no_mem(irg);
475 mark_irn_visited(om);
476 nm = new_ir_node(get_irn_dbg_info(om),
483 copy_node_attr(om, nm);
484 set_new_node(om, nm);
486 /* copy the live nodes */
487 set_irg_visited(irg, vfl);
488 irg_walk(get_nodes_block(oe), copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
490 /* Note: from yet, the visited flag of the graph is equal to vfl + 1 */
492 /* visit the anchors as well */
493 for (i = anchor_max - 1; i >= 0; --i) {
494 ir_node *n = irg->anchors[i];
496 if (n && (get_irn_visited(n) <= vfl)) {
497 set_irg_visited(irg, vfl);
498 irg_walk(n, copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
502 /* copy_preds for the end node ... */
503 set_nodes_block(ne, get_new_node(get_nodes_block(oe)));
505 /*- ... and now the keep alives. -*/
506 /* First pick the not marked block nodes and walk them. We must pick these
507 first as else we will oversee blocks reachable from Phis. */
508 irn_arity = get_End_n_keepalives(oe);
509 for (i = 0; i < irn_arity; i++) {
510 ka = get_End_keepalive(oe, i);
512 if (get_irn_visited(ka) <= vfl) {
513 /* We must keep the block alive and copy everything reachable */
514 set_irg_visited(irg, vfl);
515 irg_walk(ka, copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
517 add_End_keepalive(ne, get_new_node(ka));
521 /* Now pick other nodes. Here we will keep all! */
522 irn_arity = get_End_n_keepalives(oe);
523 for (i = 0; i < irn_arity; i++) {
524 ka = get_End_keepalive(oe, i);
526 if (get_irn_visited(ka) <= vfl) {
527 /* We didn't copy the node yet. */
528 set_irg_visited(irg, vfl);
529 irg_walk(ka, copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
531 add_End_keepalive(ne, get_new_node(ka));
535 /* start block sometimes only reached after keep alives */
536 set_nodes_block(nb, get_new_node(get_nodes_block(ob)));
537 set_nodes_block(nm, get_new_node(get_nodes_block(om)));
541 * Copies the graph reachable from current_ir_graph->end to the obstack
542 * in current_ir_graph and fixes the environment.
543 * Then fixes the fields in current_ir_graph containing nodes of the
546 * @param copy_node_nr If non-zero, the node number will be copied
549 copy_graph_env(int copy_node_nr) {
550 ir_graph *irg = current_ir_graph;
551 ir_node *old_end, *n;
554 /* remove end_except and end_reg nodes */
555 old_end = get_irg_end(irg);
556 set_irg_end_except (irg, old_end);
557 set_irg_end_reg (irg, old_end);
559 /* Not all nodes remembered in irg might be reachable
560 from the end node. Assure their link is set to NULL, so that
561 we can test whether new nodes have been computed. */
562 for (i = anchor_max - 1; i >= 0; --i) {
564 set_new_node(irg->anchors[i], NULL);
566 /* we use the block walk flag for removing Bads from Blocks ins. */
567 inc_irg_block_visited(irg);
570 copy_graph(irg, copy_node_nr);
572 /* fix the fields in irg */
573 old_end = get_irg_end(irg);
574 for (i = anchor_max - 1; i >= 0; --i) {
577 irg->anchors[i] = get_new_node(n);
583 * Copies all reachable nodes to a new obstack. Removes bad inputs
584 * from block nodes and the corresponding inputs from Phi nodes.
585 * Merges single exit blocks with single entry blocks and removes
587 * Adds all new nodes to a new hash table for CSE. Does not
588 * perform CSE, so the hash table might contain common subexpressions.
591 dead_node_elimination(ir_graph *irg) {
592 if (get_opt_optimize() && get_opt_dead_node_elimination()) {
594 int rem_ipview = get_interprocedural_view();
595 struct obstack *graveyard_obst = NULL;
596 struct obstack *rebirth_obst = NULL;
597 assert(! edges_activated(irg) && "dead node elimination requires disabled edges");
599 /* inform statistics that we started a dead-node elimination run */
600 hook_dead_node_elim(irg, 1);
602 /* Remember external state of current_ir_graph. */
603 rem = current_ir_graph;
604 current_ir_graph = irg;
605 set_interprocedural_view(0);
607 assert(get_irg_phase_state(irg) != phase_building);
609 /* Handle graph state */
610 free_callee_info(irg);
614 /* @@@ so far we loose loops when copying */
615 free_loop_information(irg);
617 set_irg_doms_inconsistent(irg);
619 /* A quiet place, where the old obstack can rest in peace,
620 until it will be cremated. */
621 graveyard_obst = irg->obst;
623 /* A new obstack, where the reachable nodes will be copied to. */
624 rebirth_obst = xmalloc(sizeof(*rebirth_obst));
625 irg->obst = rebirth_obst;
626 obstack_init(irg->obst);
627 irg->last_node_idx = 0;
629 /* We also need a new value table for CSE */
630 del_identities(irg->value_table);
631 irg->value_table = new_identities();
633 /* Copy the graph from the old to the new obstack */
634 copy_graph_env(/*copy_node_nr=*/1);
636 /* Free memory from old unoptimized obstack */
637 obstack_free(graveyard_obst, 0); /* First empty the obstack ... */
638 xfree (graveyard_obst); /* ... then free it. */
640 /* inform statistics that the run is over */
641 hook_dead_node_elim(irg, 0);
643 current_ir_graph = rem;
644 set_interprocedural_view(rem_ipview);
649 * Relink bad predecessors of a block and store the old in array to the
650 * link field. This function is called by relink_bad_predecessors().
651 * The array of link field starts with the block operand at position 0.
652 * If block has bad predecessors, create a new in array without bad preds.
653 * Otherwise let in array untouched.
655 static void relink_bad_block_predecessors(ir_node *n, void *env) {
656 ir_node **new_in, *irn;
657 int i, new_irn_n, old_irn_arity, new_irn_arity = 0;
659 /* if link field of block is NULL, look for bad predecessors otherwise
660 this is already done */
661 if (get_irn_op(n) == op_Block &&
662 get_irn_link(n) == NULL) {
664 /* save old predecessors in link field (position 0 is the block operand)*/
665 set_irn_link(n, get_irn_in(n));
667 /* count predecessors without bad nodes */
668 old_irn_arity = get_irn_arity(n);
669 for (i = 0; i < old_irn_arity; i++)
670 if (!is_Bad(get_irn_n(n, i))) new_irn_arity++;
672 /* arity changing: set new predecessors without bad nodes */
673 if (new_irn_arity < old_irn_arity) {
674 /* Get new predecessor array. We do not resize the array, as we must
675 keep the old one to update Phis. */
676 new_in = NEW_ARR_D (ir_node *, current_ir_graph->obst, (new_irn_arity+1));
678 /* set new predecessors in array */
681 for (i = 0; i < old_irn_arity; i++) {
682 irn = get_irn_n(n, i);
684 new_in[new_irn_n] = irn;
685 is_backedge(n, i) ? set_backedge(n, new_irn_n-1) : set_not_backedge(n, new_irn_n-1);
689 /* ARR_SETLEN(int, n->attr.block.backedge, new_irn_arity); */
690 ARR_SHRINKLEN(n->attr.block.backedge, new_irn_arity);
692 } /* ir node has bad predecessors */
693 } /* Block is not relinked */
697 * Relinks Bad predecessors from Blocks and Phis called by walker
698 * remove_bad_predecesors(). If n is a Block, call
699 * relink_bad_block_redecessors(). If n is a Phi-node, call also the relinking
700 * function of Phi's Block. If this block has bad predecessors, relink preds
703 static void relink_bad_predecessors(ir_node *n, void *env) {
704 ir_node *block, **old_in;
705 int i, old_irn_arity, new_irn_arity;
707 /* relink bad predecessors of a block */
708 if (get_irn_op(n) == op_Block)
709 relink_bad_block_predecessors(n, env);
711 /* If Phi node relink its block and its predecessors */
712 if (get_irn_op(n) == op_Phi) {
714 /* Relink predecessors of phi's block */
715 block = get_nodes_block(n);
716 if (get_irn_link(block) == NULL)
717 relink_bad_block_predecessors(block, env);
719 old_in = (ir_node **)get_irn_link(block); /* Of Phi's Block */
720 old_irn_arity = ARR_LEN(old_in);
722 /* Relink Phi predecessors if count of predecessors changed */
723 if (old_irn_arity != ARR_LEN(get_irn_in(block))) {
724 /* set new predecessors in array
725 n->in[0] remains the same block */
727 for(i = 1; i < old_irn_arity; i++)
728 if (!is_Bad((ir_node *)old_in[i])) {
729 n->in[new_irn_arity] = n->in[i];
730 is_backedge(n, i) ? set_backedge(n, new_irn_arity) : set_not_backedge(n, new_irn_arity);
734 ARR_SETLEN(ir_node *, n->in, new_irn_arity);
735 ARR_SETLEN(int, n->attr.phi_backedge, new_irn_arity);
737 } /* n is a Phi node */
741 * Removes Bad Bad predecessors from Blocks and the corresponding
742 * inputs to Phi nodes as in dead_node_elimination but without
744 * On walking up set the link field to NULL, on walking down call
745 * relink_bad_predecessors() (This function stores the old in array
746 * to the link field and sets a new in array if arity of predecessors
749 void remove_bad_predecessors(ir_graph *irg) {
750 irg_walk_graph(irg, firm_clear_link, relink_bad_predecessors, NULL);
757 __)|_| | \_/ | \_/(/_ |_/\__|__
759 The following stuff implements a facility that automatically patches
760 registered ir_node pointers to the new node when a dead node elimination occurs.
763 struct _survive_dce_t {
767 hook_entry_t dead_node_elim;
768 hook_entry_t dead_node_elim_subst;
771 typedef struct _survive_dce_list_t {
772 struct _survive_dce_list_t *next;
774 } survive_dce_list_t;
776 static void dead_node_hook(void *context, ir_graph *irg, int start) {
777 survive_dce_t *sd = context;
779 /* Create a new map before the dead node elimination is performed. */
781 sd->new_places = pmap_create_ex(pmap_count(sd->places));
783 /* Patch back all nodes if dead node elimination is over and something is to be done. */
784 pmap_destroy(sd->places);
785 sd->places = sd->new_places;
786 sd->new_places = NULL;
791 * Hook called when dead node elimination replaces old by nw.
793 static void dead_node_subst_hook(void *context, ir_graph *irg, ir_node *old, ir_node *nw) {
794 survive_dce_t *sd = context;
795 survive_dce_list_t *list = pmap_get(sd->places, old);
797 /* If the node is to be patched back, write the new address to all registered locations. */
799 survive_dce_list_t *p;
801 for (p = list; p; p = p->next)
804 pmap_insert(sd->new_places, nw, list);
809 * Make a new Survive DCE environment.
811 survive_dce_t *new_survive_dce(void) {
812 survive_dce_t *res = xmalloc(sizeof(res[0]));
813 obstack_init(&res->obst);
814 res->places = pmap_create();
815 res->new_places = NULL;
817 res->dead_node_elim.hook._hook_dead_node_elim = dead_node_hook;
818 res->dead_node_elim.context = res;
819 res->dead_node_elim.next = NULL;
821 res->dead_node_elim_subst.hook._hook_dead_node_elim_subst = dead_node_subst_hook;
822 res->dead_node_elim_subst.context = res;
823 res->dead_node_elim_subst.next = NULL;
825 #ifndef FIRM_ENABLE_HOOKS
826 assert(0 && "need hooks enabled");
829 register_hook(hook_dead_node_elim, &res->dead_node_elim);
830 register_hook(hook_dead_node_elim_subst, &res->dead_node_elim_subst);
835 * Free a Survive DCE environment.
837 void free_survive_dce(survive_dce_t *sd) {
838 obstack_free(&sd->obst, NULL);
839 pmap_destroy(sd->places);
840 unregister_hook(hook_dead_node_elim, &sd->dead_node_elim);
841 unregister_hook(hook_dead_node_elim_subst, &sd->dead_node_elim_subst);
846 * Register a node pointer to be patched upon DCE.
847 * When DCE occurs, the node pointer specified by @p place will be
848 * patched to the new address of the node it is pointing to.
850 * @param sd The Survive DCE environment.
851 * @param place The address of the node pointer.
853 void survive_dce_register_irn(survive_dce_t *sd, ir_node **place) {
854 if (*place != NULL) {
855 ir_node *irn = *place;
856 survive_dce_list_t *curr = pmap_get(sd->places, irn);
857 survive_dce_list_t *nw = obstack_alloc(&sd->obst, sizeof(nw[0]));
862 pmap_insert(sd->places, irn, nw);
866 /*--------------------------------------------------------------------*/
867 /* Functionality for inlining */
868 /*--------------------------------------------------------------------*/
871 * Copy node for inlineing. Updates attributes that change when
872 * inlineing but not for dead node elimination.
874 * Copies the node by calling copy_node() and then updates the entity if
875 * it's a local one. env must be a pointer of the frame type of the
876 * inlined procedure. The new entities must be in the link field of
880 copy_node_inline(ir_node *n, void *env) {
882 ir_type *frame_tp = (ir_type *)env;
885 if (get_irn_op(n) == op_Sel) {
886 nn = get_new_node (n);
888 if (get_entity_owner(get_Sel_entity(n)) == frame_tp) {
889 set_Sel_entity(nn, get_entity_link(get_Sel_entity(n)));
891 } else if (get_irn_op(n) == op_Block) {
892 nn = get_new_node (n);
893 nn->attr.block.irg = current_ir_graph;
898 * Walker: checks if P_value_arg_base is used.
900 static void find_addr(ir_node *node, void *env) {
901 int *allow_inline = env;
902 if (is_Proj(node) && get_irn_op(get_Proj_pred(node)) == op_Start) {
903 if (get_Proj_proj(node) == pn_Start_P_value_arg_base)
909 * Check if we can inline a given call.
910 * Currently, we cannot inline two cases:
911 * - call with compound arguments
912 * - graphs that take the address of a parameter
914 * check these conditions here
916 static int can_inline(ir_node *call, ir_graph *called_graph) {
917 ir_type *call_type = get_Call_type(call);
918 int params, ress, i, res;
919 assert(is_Method_type(call_type));
921 params = get_method_n_params(call_type);
922 ress = get_method_n_ress(call_type);
924 /* check parameters for compound arguments */
925 for (i = 0; i < params; ++i) {
926 ir_type *p_type = get_method_param_type(call_type, i);
928 if (is_compound_type(p_type))
932 /* check results for compound arguments */
933 for (i = 0; i < ress; ++i) {
934 ir_type *r_type = get_method_res_type(call_type, i);
936 if (is_compound_type(r_type))
941 irg_walk_graph(called_graph, find_addr, NULL, &res);
947 exc_handler = 0, /**< There is a handler. */
948 exc_to_end = 1, /**< Branches to End. */
949 exc_no_handler = 2 /**< Exception handling not represented. */
952 /* Inlines a method at the given call site. */
953 int inline_method(ir_node *call, ir_graph *called_graph) {
955 ir_node *post_call, *post_bl;
956 ir_node *in[pn_Start_max];
957 ir_node *end, *end_bl;
961 int arity, n_ret, n_exc, n_res, i, j, rem_opt, irn_arity;
962 enum exc_mode exc_handling;
963 ir_type *called_frame;
964 irg_inline_property prop = get_irg_inline_property(called_graph);
966 if ( (prop < irg_inline_forced) &&
967 (!get_opt_optimize() || !get_opt_inline() || (prop == irg_inline_forbidden))) return 0;
969 /* Do not inline variadic functions. */
970 if (get_method_variadicity(get_entity_type(get_irg_entity(called_graph))) == variadicity_variadic)
973 assert(get_method_n_params(get_entity_type(get_irg_entity(called_graph))) ==
974 get_method_n_params(get_Call_type(call)));
977 * currently, we cannot inline two cases:
978 * - call with compound arguments
979 * - graphs that take the address of a parameter
981 if (! can_inline(call, called_graph))
984 /* -- Turn off optimizations, this can cause problems when allocating new nodes. -- */
985 rem_opt = get_opt_optimize();
988 /* Handle graph state */
989 assert(get_irg_phase_state(current_ir_graph) != phase_building);
990 assert(get_irg_pinned(current_ir_graph) == op_pin_state_pinned);
991 assert(get_irg_pinned(called_graph) == op_pin_state_pinned);
992 set_irg_outs_inconsistent(current_ir_graph);
993 set_irg_extblk_inconsistent(current_ir_graph);
994 set_irg_doms_inconsistent(current_ir_graph);
995 set_irg_loopinfo_inconsistent(current_ir_graph);
996 set_irg_callee_info_state(current_ir_graph, irg_callee_info_inconsistent);
998 /* -- Check preconditions -- */
999 assert(is_Call(call));
1000 /* @@@ does not work for InterfaceIII.java after cgana
1001 assert(get_Call_type(call) == get_entity_type(get_irg_entity(called_graph)));
1002 assert(smaller_type(get_entity_type(get_irg_entity(called_graph)),
1003 get_Call_type(call)));
1005 if (called_graph == current_ir_graph) {
1006 set_optimize(rem_opt);
1010 /* here we know we WILL inline, so inform the statistics */
1011 hook_inline(call, called_graph);
1013 /* -- Decide how to handle exception control flow: Is there a handler
1014 for the Call node, or do we branch directly to End on an exception?
1016 0 There is a handler.
1018 2 Exception handling not represented in Firm. -- */
1020 ir_node *proj, *Mproj = NULL, *Xproj = NULL;
1021 for (proj = get_irn_link(call); proj; proj = get_irn_link(proj)) {
1022 long proj_nr = get_Proj_proj(proj);
1023 if (proj_nr == pn_Call_X_except) Xproj = proj;
1024 if (proj_nr == pn_Call_M_except) Mproj = proj;
1026 if (Mproj) { assert(Xproj); exc_handling = exc_handler; } /* Mproj */
1027 else if (Xproj) { exc_handling = exc_to_end; } /* !Mproj && Xproj */
1028 else { exc_handling = exc_no_handler; } /* !Mproj && !Xproj */
1032 the procedure and later replaces the Start node of the called graph.
1033 Post_call is the old Call node and collects the results of the called
1034 graph. Both will end up being a tuple. -- */
1035 post_bl = get_nodes_block(call);
1036 set_irg_current_block(current_ir_graph, post_bl);
1037 /* XxMxPxPxPxT of Start + parameter of Call */
1038 in[pn_Start_X_initial_exec] = new_Jmp();
1039 in[pn_Start_M] = get_Call_mem(call);
1040 in[pn_Start_P_frame_base] = get_irg_frame(current_ir_graph);
1041 in[pn_Start_P_globals] = get_irg_globals(current_ir_graph);
1042 in[pn_Start_P_tls] = get_irg_tls(current_ir_graph);
1043 in[pn_Start_T_args] = new_Tuple(get_Call_n_params(call), get_Call_param_arr(call));
1044 /* in[pn_Start_P_value_arg_base] = ??? */
1045 assert(pn_Start_P_value_arg_base == pn_Start_max - 1 && "pn_Start_P_value_arg_base not supported, fix");
1046 pre_call = new_Tuple(pn_Start_max - 1, in);
1050 The new block gets the ins of the old block, pre_call and all its
1051 predecessors and all Phi nodes. -- */
1052 part_block(pre_call);
1054 /* -- Prepare state for dead node elimination -- */
1055 /* Visited flags in calling irg must be >= flag in called irg.
1056 Else walker and arity computation will not work. */
1057 if (get_irg_visited(current_ir_graph) <= get_irg_visited(called_graph))
1058 set_irg_visited(current_ir_graph, get_irg_visited(called_graph)+1);
1059 if (get_irg_block_visited(current_ir_graph)< get_irg_block_visited(called_graph))
1060 set_irg_block_visited(current_ir_graph, get_irg_block_visited(called_graph));
1061 /* Set pre_call as new Start node in link field of the start node of
1062 calling graph and pre_calls block as new block for the start block
1064 Further mark these nodes so that they are not visited by the
1066 set_irn_link(get_irg_start(called_graph), pre_call);
1067 set_irn_visited(get_irg_start(called_graph), get_irg_visited(current_ir_graph));
1068 set_irn_link(get_irg_start_block(called_graph), get_nodes_block(pre_call));
1069 set_irn_visited(get_irg_start_block(called_graph), get_irg_visited(current_ir_graph));
1070 set_irn_link(get_irg_bad(called_graph), get_irg_bad(current_ir_graph));
1071 set_irn_visited(get_irg_bad(called_graph), get_irg_visited(current_ir_graph));
1073 /* Initialize for compaction of in arrays */
1074 inc_irg_block_visited(current_ir_graph);
1076 /* -- Replicate local entities of the called_graph -- */
1077 /* copy the entities. */
1078 called_frame = get_irg_frame_type(called_graph);
1079 for (i = 0; i < get_class_n_members(called_frame); i++) {
1080 ir_entity *new_ent, *old_ent;
1081 old_ent = get_class_member(called_frame, i);
1082 new_ent = copy_entity_own(old_ent, get_cur_frame_type());
1083 set_entity_link(old_ent, new_ent);
1086 /* visited is > than that of called graph. With this trick visited will
1087 remain unchanged so that an outer walker, e.g., searching the call nodes
1088 to inline, calling this inline will not visit the inlined nodes. */
1089 set_irg_visited(current_ir_graph, get_irg_visited(current_ir_graph)-1);
1091 /* -- Performing dead node elimination inlines the graph -- */
1092 /* Copies the nodes to the obstack of current_ir_graph. Updates links to new
1094 irg_walk(get_irg_end(called_graph), copy_node_inline, copy_preds,
1095 get_irg_frame_type(called_graph));
1097 /* Repair called_graph */
1098 set_irg_visited(called_graph, get_irg_visited(current_ir_graph));
1099 set_irg_block_visited(called_graph, get_irg_block_visited(current_ir_graph));
1100 set_Block_block_visited(get_irg_start_block(called_graph), 0);
1102 /* -- Merge the end of the inlined procedure with the call site -- */
1103 /* We will turn the old Call node into a Tuple with the following
1106 0: Phi of all Memories of Return statements.
1107 1: Jmp from new Block that merges the control flow from all exception
1108 predecessors of the old end block.
1109 2: Tuple of all arguments.
1110 3: Phi of Exception memories.
1111 In case the old Call directly branches to End on an exception we don't
1112 need the block merging all exceptions nor the Phi of the exception
1116 /* -- Precompute some values -- */
1117 end_bl = get_new_node(get_irg_end_block(called_graph));
1118 end = get_new_node(get_irg_end(called_graph));
1119 arity = get_irn_arity(end_bl); /* arity = n_exc + n_ret */
1120 n_res = get_method_n_ress(get_Call_type(call));
1122 res_pred = xmalloc(n_res * sizeof(*res_pred));
1123 cf_pred = xmalloc(arity * sizeof(*res_pred));
1125 set_irg_current_block(current_ir_graph, post_bl); /* just to make sure */
1127 /* -- archive keepalives -- */
1128 irn_arity = get_irn_arity(end);
1129 for (i = 0; i < irn_arity; i++) {
1130 ir_node *ka = get_End_keepalive(end, i);
1132 add_End_keepalive(get_irg_end(current_ir_graph), ka);
1135 /* The new end node will die. We need not free as the in array is on the obstack:
1136 copy_node() only generated 'D' arrays. */
1138 /* -- Replace Return nodes by Jump nodes. -- */
1140 for (i = 0; i < arity; i++) {
1142 ret = get_irn_n(end_bl, i);
1143 if (is_Return(ret)) {
1144 cf_pred[n_ret] = new_r_Jmp(current_ir_graph, get_nodes_block(ret));
1148 set_irn_in(post_bl, n_ret, cf_pred);
1150 /* -- Build a Tuple for all results of the method.
1151 Add Phi node if there was more than one Return. -- */
1152 turn_into_tuple(post_call, pn_Call_max);
1153 /* First the Memory-Phi */
1155 for (i = 0; i < arity; i++) {
1156 ret = get_irn_n(end_bl, i);
1157 if (is_Return(ret)) {
1158 cf_pred[n_ret] = get_Return_mem(ret);
1162 phi = new_Phi(n_ret, cf_pred, mode_M);
1163 set_Tuple_pred(call, pn_Call_M_regular, phi);
1164 /* Conserve Phi-list for further inlinings -- but might be optimized */
1165 if (get_nodes_block(phi) == post_bl) {
1166 set_irn_link(phi, get_irn_link(post_bl));
1167 set_irn_link(post_bl, phi);
1169 /* Now the real results */
1171 for (j = 0; j < n_res; j++) {
1173 for (i = 0; i < arity; i++) {
1174 ret = get_irn_n(end_bl, i);
1175 if (get_irn_op(ret) == op_Return) {
1176 cf_pred[n_ret] = get_Return_res(ret, j);
1181 phi = new_Phi(n_ret, cf_pred, get_irn_mode(cf_pred[0]));
1185 /* Conserve Phi-list for further inlinings -- but might be optimized */
1186 if (get_nodes_block(phi) == post_bl) {
1187 set_irn_link(phi, get_irn_link(post_bl));
1188 set_irn_link(post_bl, phi);
1191 set_Tuple_pred(call, pn_Call_T_result, new_Tuple(n_res, res_pred));
1193 set_Tuple_pred(call, pn_Call_T_result, new_Bad());
1196 /* For now, we cannot inline calls with value_base */
1197 set_Tuple_pred(call, pn_Call_P_value_res_base, new_Bad());
1199 /* Finally the exception control flow.
1200 We have two (three) possible situations:
1201 First if the Call branches to an exception handler: We need to add a Phi node to
1202 collect the memory containing the exception objects. Further we need
1203 to add another block to get a correct representation of this Phi. To
1204 this block we add a Jmp that resolves into the X output of the Call
1205 when the Call is turned into a tuple.
1206 Second the Call branches to End, the exception is not handled. Just
1207 add all inlined exception branches to the End node.
1208 Third: there is no Exception edge at all. Handle as case two. */
1209 if (exc_handling == exc_handler) {
1211 for (i = 0; i < arity; i++) {
1213 ret = get_irn_n(end_bl, i);
1214 irn = skip_Proj(ret);
1215 if (is_fragile_op(irn) || (get_irn_op(irn) == op_Raise)) {
1216 cf_pred[n_exc] = ret;
1221 new_Block(n_exc, cf_pred); /* watch it: current_block is changed! */
1222 set_Tuple_pred(call, pn_Call_X_except, new_Jmp());
1223 /* The Phi for the memories with the exception objects */
1225 for (i = 0; i < arity; i++) {
1227 ret = skip_Proj(get_irn_n(end_bl, i));
1229 cf_pred[n_exc] = new_r_Proj(current_ir_graph, get_nodes_block(ret), ret, mode_M, 3);
1231 } else if (is_fragile_op(ret)) {
1232 /* We rely that all cfops have the memory output at the same position. */
1233 cf_pred[n_exc] = new_r_Proj(current_ir_graph, get_nodes_block(ret), ret, mode_M, 0);
1235 } else if (get_irn_op(ret) == op_Raise) {
1236 cf_pred[n_exc] = new_r_Proj(current_ir_graph, get_nodes_block(ret), ret, mode_M, 1);
1240 set_Tuple_pred(call, pn_Call_M_except, new_Phi(n_exc, cf_pred, mode_M));
1242 set_Tuple_pred(call, pn_Call_X_except, new_Bad());
1243 set_Tuple_pred(call, pn_Call_M_except, new_Bad());
1245 set_Tuple_pred(call, pn_Call_X_regular, new_Bad());
1247 ir_node *main_end_bl;
1248 int main_end_bl_arity;
1249 ir_node **end_preds;
1251 /* assert(exc_handling == 1 || no exceptions. ) */
1253 for (i = 0; i < arity; i++) {
1254 ir_node *ret = get_irn_n(end_bl, i);
1255 ir_node *irn = skip_Proj(ret);
1257 if (is_fragile_op(irn) || (get_irn_op(irn) == op_Raise)) {
1258 cf_pred[n_exc] = ret;
1262 main_end_bl = get_irg_end_block(current_ir_graph);
1263 main_end_bl_arity = get_irn_arity(main_end_bl);
1264 end_preds = xmalloc((n_exc + main_end_bl_arity) * sizeof(*end_preds));
1266 for (i = 0; i < main_end_bl_arity; ++i)
1267 end_preds[i] = get_irn_n(main_end_bl, i);
1268 for (i = 0; i < n_exc; ++i)
1269 end_preds[main_end_bl_arity + i] = cf_pred[i];
1270 set_irn_in(main_end_bl, n_exc + main_end_bl_arity, end_preds);
1271 set_Tuple_pred(call, pn_Call_X_regular, new_Bad());
1272 set_Tuple_pred(call, pn_Call_X_except, new_Bad());
1273 set_Tuple_pred(call, pn_Call_M_except, new_Bad());
1279 /* -- Turn CSE back on. -- */
1280 set_optimize(rem_opt);
1285 /********************************************************************/
1286 /* Apply inlineing to small methods. */
1287 /********************************************************************/
1289 /** Represents a possible inlinable call in a graph. */
1290 typedef struct _call_entry call_entry;
1291 struct _call_entry {
1292 ir_node *call; /**< the Call */
1293 ir_graph *callee; /**< the callee called here */
1294 call_entry *next; /**< for linking the next one */
1298 * environment for inlining small irgs
1300 typedef struct _inline_env_t {
1301 struct obstack obst; /**< an obstack where call_entries are allocated on. */
1302 call_entry *head; /**< the head of the call entry list */
1303 call_entry *tail; /**< the tail of the call entry list */
1307 * Returns the irg called from a Call node. If the irg is not
1308 * known, NULL is returned.
1310 static ir_graph *get_call_called_irg(ir_node *call) {
1312 ir_graph *called_irg = NULL;
1314 addr = get_Call_ptr(call);
1315 if (is_SymConst(addr) && get_SymConst_kind(addr) == symconst_addr_ent) {
1316 called_irg = get_entity_irg(get_SymConst_entity(addr));
1323 * Walker: Collect all calls to known graphs inside a graph.
1325 static void collect_calls(ir_node *call, void *env) {
1326 if (is_Call(call)) {
1327 ir_graph *called_irg = get_call_called_irg(call);
1329 /* The Call node calls a locally defined method. Remember to inline. */
1330 inline_env_t *ienv = env;
1331 call_entry *entry = obstack_alloc(&ienv->obst, sizeof(*entry));
1333 entry->callee = called_irg;
1336 if (ienv->tail == NULL)
1339 ienv->tail->next = entry;
1346 * Inlines all small methods at call sites where the called address comes
1347 * from a Const node that references the entity representing the called
1349 * The size argument is a rough measure for the code size of the method:
1350 * Methods where the obstack containing the firm graph is smaller than
1353 void inline_small_irgs(ir_graph *irg, int size) {
1354 ir_graph *rem = current_ir_graph;
1357 DEBUG_ONLY(firm_dbg_module_t *dbg;)
1359 if (!(get_opt_optimize() && get_opt_inline())) return;
1361 FIRM_DBG_REGISTER(dbg, "firm.opt.inline");
1363 current_ir_graph = irg;
1364 /* Handle graph state */
1365 assert(get_irg_phase_state(irg) != phase_building);
1366 free_callee_info(irg);
1368 /* Find Call nodes to inline.
1369 (We can not inline during a walk of the graph, as inlineing the same
1370 method several times changes the visited flag of the walked graph:
1371 after the first inlineing visited of the callee equals visited of
1372 the caller. With the next inlineing both are increased.) */
1373 obstack_init(&env.obst);
1374 env.head = env.tail = NULL;
1375 irg_walk_graph(irg, NULL, collect_calls, &env);
1377 if (env.head != NULL) {
1378 /* There are calls to inline */
1379 collect_phiprojs(irg);
1380 for (entry = env.head; entry != NULL; entry = entry->next) {
1381 ir_graph *callee = entry->callee;
1382 if (((_obstack_memory_used(callee->obst) - (int)obstack_room(callee->obst)) < size) ||
1383 (get_irg_inline_property(callee) >= irg_inline_forced)) {
1384 inline_method(entry->call, callee);
1388 obstack_free(&env.obst, NULL);
1389 current_ir_graph = rem;
1393 * Environment for inlining irgs.
1396 int n_nodes; /**< Number of nodes in graph except Id, Tuple, Proj, Start, End. */
1397 int n_nodes_orig; /**< for statistics */
1398 call_entry *call_head; /**< The head of the list of all call nodes in this graph. */
1399 call_entry *call_tail; /**< The tail of the list of all call nodes in this graph .*/
1400 int n_call_nodes; /**< Number of Call nodes in the graph. */
1401 int n_call_nodes_orig; /**< for statistics */
1402 int n_callers; /**< Number of known graphs that call this graphs. */
1403 int n_callers_orig; /**< for statistics */
1404 int got_inline; /**< Set, if at leat one call inside this graph was inlined. */
1408 * Allocate a new environment for inlining.
1410 static inline_irg_env *alloc_inline_irg_env(struct obstack *obst) {
1411 inline_irg_env *env = obstack_alloc(obst, sizeof(*env));
1412 env->n_nodes = -2; /* do not count count Start, End */
1413 env->n_nodes_orig = -2; /* do not count Start, End */
1414 env->call_head = NULL;
1415 env->call_tail = NULL;
1416 env->n_call_nodes = 0;
1417 env->n_call_nodes_orig = 0;
1419 env->n_callers_orig = 0;
1420 env->got_inline = 0;
1424 typedef struct walker_env {
1425 struct obstack *obst; /**< the obstack for allocations. */
1426 inline_irg_env *x; /**< the inline environment */
1427 int ignore_runtime; /**< the ignore runtime flag */
1431 * post-walker: collect all calls in the inline-environment
1432 * of a graph and sum some statistics.
1434 static void collect_calls2(ir_node *call, void *ctx) {
1436 inline_irg_env *x = env->x;
1437 ir_op *op = get_irn_op(call);
1441 /* count meaningful nodes in irg */
1442 if (op != op_Proj && op != op_Tuple && op != op_Sync) {
1447 if (op != op_Call) return;
1449 /* check, if it's a runtime call */
1450 if (env->ignore_runtime) {
1451 ir_node *symc = get_Call_ptr(call);
1453 if (is_SymConst(symc) && get_SymConst_kind(symc) == symconst_addr_ent) {
1454 ir_entity *ent = get_SymConst_entity(symc);
1456 if (get_entity_additional_properties(ent) & mtp_property_runtime)
1461 /* collect all call nodes */
1463 ++x->n_call_nodes_orig;
1465 callee = get_call_called_irg(call);
1467 inline_irg_env *callee_env = get_irg_link(callee);
1468 /* count all static callers */
1469 ++callee_env->n_callers;
1470 ++callee_env->n_callers_orig;
1472 /* link it in the list of possible inlinable entries */
1473 entry = obstack_alloc(env->obst, sizeof(*entry));
1475 entry->callee = callee;
1477 if (x->call_tail == NULL)
1478 x->call_head = entry;
1480 x->call_tail->next = entry;
1481 x->call_tail = entry;
1486 * Returns TRUE if the number of callers in 0 in the irg's environment,
1487 * hence this irg is a leave.
1489 INLINE static int is_leave(ir_graph *irg) {
1490 inline_irg_env *env = get_irg_link(irg);
1491 return env->n_call_nodes == 0;
1495 * Returns TRUE if the number of callers is smaller size in the irg's environment.
1497 INLINE static int is_smaller(ir_graph *callee, int size) {
1498 inline_irg_env *env = get_irg_link(callee);
1499 return env->n_nodes < size;
1503 * Append the nodes of the list src to the nodes of the list in environment dst.
1505 static void append_call_list(struct obstack *obst, inline_irg_env *dst, call_entry *src) {
1506 call_entry *entry, *nentry;
1508 /* Note that the src list points to Call nodes in the inlined graph, but
1509 we need Call nodes in our graph. Luckily the inliner leaves this information
1510 in the link field. */
1511 for (entry = src; entry != NULL; entry = entry->next) {
1512 nentry = obstack_alloc(obst, sizeof(*nentry));
1513 nentry->call = get_irn_link(entry->call);
1514 nentry->callee = entry->callee;
1515 nentry->next = NULL;
1516 dst->call_tail->next = nentry;
1517 dst->call_tail = nentry;
1522 * Inlines small leave methods at call sites where the called address comes
1523 * from a Const node that references the entity representing the called
1525 * The size argument is a rough measure for the code size of the method:
1526 * Methods where the obstack containing the firm graph is smaller than
1529 void inline_leave_functions(int maxsize, int leavesize, int size, int ignore_runtime) {
1530 inline_irg_env *env;
1536 call_entry *entry, *tail;
1537 const call_entry *centry;
1538 struct obstack obst;
1539 DEBUG_ONLY(firm_dbg_module_t *dbg;)
1541 if (!(get_opt_optimize() && get_opt_inline())) return;
1543 FIRM_DBG_REGISTER(dbg, "firm.opt.inline");
1544 rem = current_ir_graph;
1545 obstack_init(&obst);
1547 /* extend all irgs by a temporary data structure for inlining. */
1548 n_irgs = get_irp_n_irgs();
1549 for (i = 0; i < n_irgs; ++i)
1550 set_irg_link(get_irp_irg(i), alloc_inline_irg_env(&obst));
1552 /* Precompute information in temporary data structure. */
1554 wenv.ignore_runtime = ignore_runtime;
1555 for (i = 0; i < n_irgs; ++i) {
1556 ir_graph *irg = get_irp_irg(i);
1558 assert(get_irg_phase_state(irg) != phase_building);
1559 free_callee_info(irg);
1561 wenv.x = get_irg_link(irg);
1562 irg_walk_graph(irg, NULL, collect_calls2, &wenv);
1565 /* -- and now inline. -- */
1567 /* Inline leaves recursively -- we might construct new leaves. */
1571 for (i = 0; i < n_irgs; ++i) {
1573 int phiproj_computed = 0;
1575 current_ir_graph = get_irp_irg(i);
1576 env = (inline_irg_env *)get_irg_link(current_ir_graph);
1579 for (entry = env->call_head; entry != NULL; entry = entry->next) {
1582 if (env->n_nodes > maxsize) break;
1585 callee = entry->callee;
1587 if (is_leave(callee) && is_smaller(callee, leavesize)) {
1588 if (!phiproj_computed) {
1589 phiproj_computed = 1;
1590 collect_phiprojs(current_ir_graph);
1592 did_inline = inline_method(call, callee);
1595 /* Do some statistics */
1596 inline_irg_env *callee_env = (inline_irg_env *)get_irg_link(callee);
1598 env->got_inline = 1;
1599 --env->n_call_nodes;
1600 env->n_nodes += callee_env->n_nodes;
1601 --callee_env->n_callers;
1603 /* remove this call from the list */
1605 tail->next = entry->next;
1607 env->call_head = entry->next;
1613 env->call_tail = tail;
1615 } while (did_inline);
1617 /* inline other small functions. */
1618 for (i = 0; i < n_irgs; ++i) {
1620 int phiproj_computed = 0;
1622 current_ir_graph = get_irp_irg(i);
1623 env = (inline_irg_env *)get_irg_link(current_ir_graph);
1625 /* note that the list of possible calls is updated during the process */
1627 for (entry = env->call_head; entry != NULL; entry = entry->next) {
1631 callee = entry->callee;
1633 if (((is_smaller(callee, size) && (env->n_nodes < maxsize)) || /* small function */
1634 (get_irg_inline_property(callee) >= irg_inline_forced))) {
1635 if (!phiproj_computed) {
1636 phiproj_computed = 1;
1637 collect_phiprojs(current_ir_graph);
1639 if (inline_method(call, callee)) {
1640 inline_irg_env *callee_env = (inline_irg_env *)get_irg_link(callee);
1642 /* callee was inline. Append it's call list. */
1643 env->got_inline = 1;
1644 --env->n_call_nodes;
1645 append_call_list(&obst, env, callee_env->call_head);
1646 env->n_call_nodes += callee_env->n_call_nodes;
1647 env->n_nodes += callee_env->n_nodes;
1648 --callee_env->n_callers;
1650 /* after we have inlined callee, all called methods inside callee
1651 are now called once more */
1652 for (centry = callee_env->call_head; centry != NULL; centry = centry->next) {
1653 inline_irg_env *penv = get_irg_link(centry->callee);
1657 /* remove this call from the list */
1659 tail->next = entry->next;
1661 env->call_head = entry->next;
1667 env->call_tail = tail;
1670 for (i = 0; i < n_irgs; ++i) {
1671 irg = get_irp_irg(i);
1672 env = (inline_irg_env *)get_irg_link(irg);
1674 if (env->got_inline) {
1675 /* this irg got calls inlined */
1676 set_irg_outs_inconsistent(irg);
1677 set_irg_doms_inconsistent(irg);
1679 optimize_graph_df(irg);
1682 if (env->got_inline || (env->n_callers_orig != env->n_callers))
1683 DB((dbg, SET_LEVEL_1, "Nodes:%3d ->%3d, calls:%3d ->%3d, callers:%3d ->%3d, -- %s\n",
1684 env->n_nodes_orig, env->n_nodes, env->n_call_nodes_orig, env->n_call_nodes,
1685 env->n_callers_orig, env->n_callers,
1686 get_entity_name(get_irg_entity(irg))));
1689 obstack_free(&obst, NULL);
1690 current_ir_graph = rem;
1693 /*******************************************************************/
1694 /* Code Placement. Pins all floating nodes to a block where they */
1695 /* will be executed only if needed. */
1696 /*******************************************************************/
1699 * Returns non-zero, is a block is not reachable from Start.
1701 * @param block the block to test
1704 is_Block_unreachable(ir_node *block) {
1705 return is_Block_dead(block) || get_Block_dom_depth(block) < 0;
1709 * Find the earliest correct block for node n. --- Place n into the
1710 * same Block as its dominance-deepest Input.
1712 * We have to avoid calls to get_nodes_block() here
1713 * because the graph is floating.
1715 * move_out_of_loops() expects that place_floats_early() have placed
1716 * all "living" nodes into a living block. That's why we must
1717 * move nodes in dead block with "live" successors into a valid
1719 * We move them just into the same block as it's successor (or
1720 * in case of a Phi into the effective use block). For Phi successors,
1721 * this may still be a dead block, but then there is no real use, as
1722 * the control flow will be dead later.
1724 * @param n the node to be placed
1725 * @param worklist a worklist, predecessors of non-floating nodes are placed here
1728 place_floats_early(ir_node *n, waitq *worklist) {
1731 /* we must not run into an infinite loop */
1732 assert(irn_not_visited(n));
1733 mark_irn_visited(n);
1735 /* Place floating nodes. */
1736 if (get_irn_pinned(n) == op_pin_state_floats) {
1737 ir_node *curr_block = get_irn_n(n, -1);
1738 int in_dead_block = is_Block_unreachable(curr_block);
1740 ir_node *b = NULL; /* The block to place this node in */
1742 assert(is_no_Block(n));
1744 if (is_irn_start_block_placed(n)) {
1745 /* These nodes will not be placed by the loop below. */
1746 b = get_irg_start_block(current_ir_graph);
1750 /* find the block for this node. */
1751 irn_arity = get_irn_arity(n);
1752 for (i = 0; i < irn_arity; i++) {
1753 ir_node *pred = get_irn_n(n, i);
1754 ir_node *pred_block;
1756 if ((irn_not_visited(pred))
1757 && (get_irn_pinned(pred) == op_pin_state_floats)) {
1760 * If the current node is NOT in a dead block, but one of its
1761 * predecessors is, we must move the predecessor to a live block.
1762 * Such thing can happen, if global CSE chose a node from a dead block.
1763 * We move it simply to our block.
1764 * Note that neither Phi nor End nodes are floating, so we don't
1765 * need to handle them here.
1767 if (! in_dead_block) {
1768 if (get_irn_pinned(pred) == op_pin_state_floats &&
1769 is_Block_unreachable(get_irn_n(pred, -1)))
1770 set_nodes_block(pred, curr_block);
1772 place_floats_early(pred, worklist);
1776 * A node in the Bad block must stay in the bad block,
1777 * so don't compute a new block for it.
1782 /* Because all loops contain at least one op_pin_state_pinned node, now all
1783 our inputs are either op_pin_state_pinned or place_early() has already
1784 been finished on them. We do not have any unfinished inputs! */
1785 pred_block = get_irn_n(pred, -1);
1786 if ((!is_Block_dead(pred_block)) &&
1787 (get_Block_dom_depth(pred_block) > depth)) {
1789 depth = get_Block_dom_depth(pred_block);
1791 /* Avoid that the node is placed in the Start block */
1792 if ((depth == 1) && (get_Block_dom_depth(get_irn_n(n, -1)) > 1)
1793 && get_irg_phase_state(current_ir_graph) != phase_backend) {
1794 b = get_Block_cfg_out(get_irg_start_block(current_ir_graph), 0);
1795 assert(b != get_irg_start_block(current_ir_graph));
1800 set_nodes_block(n, b);
1804 * Add predecessors of non floating nodes and non-floating predecessors
1805 * of floating nodes to worklist and fix their blocks if the are in dead block.
1807 irn_arity = get_irn_arity(n);
1809 if (get_irn_op(n) == op_End) {
1811 * Simplest case: End node. Predecessors are keep-alives,
1812 * no need to move out of dead block.
1814 for (i = -1; i < irn_arity; ++i) {
1815 ir_node *pred = get_irn_n(n, i);
1816 if (irn_not_visited(pred))
1817 waitq_put(worklist, pred);
1819 } else if (is_Block(n)) {
1821 * Blocks: Predecessors are control flow, no need to move
1822 * them out of dead block.
1824 for (i = irn_arity - 1; i >= 0; --i) {
1825 ir_node *pred = get_irn_n(n, i);
1826 if (irn_not_visited(pred))
1827 waitq_put(worklist, pred);
1829 } else if (is_Phi(n)) {
1831 ir_node *curr_block = get_irn_n(n, -1);
1832 int in_dead_block = is_Block_unreachable(curr_block);
1835 * Phi nodes: move nodes from dead blocks into the effective use
1836 * of the Phi-input if the Phi is not in a bad block.
1838 pred = get_irn_n(n, -1);
1839 if (irn_not_visited(pred))
1840 waitq_put(worklist, pred);
1842 for (i = irn_arity - 1; i >= 0; --i) {
1843 ir_node *pred = get_irn_n(n, i);
1845 if (irn_not_visited(pred)) {
1846 if (! in_dead_block &&
1847 get_irn_pinned(pred) == op_pin_state_floats &&
1848 is_Block_unreachable(get_irn_n(pred, -1))) {
1849 set_nodes_block(pred, get_Block_cfgpred_block(curr_block, i));
1851 waitq_put(worklist, pred);
1856 ir_node *curr_block = get_irn_n(n, -1);
1857 int in_dead_block = is_Block_unreachable(curr_block);
1860 * All other nodes: move nodes from dead blocks into the same block.
1862 pred = get_irn_n(n, -1);
1863 if (irn_not_visited(pred))
1864 waitq_put(worklist, pred);
1866 for (i = irn_arity - 1; i >= 0; --i) {
1867 ir_node *pred = get_irn_n(n, i);
1869 if (irn_not_visited(pred)) {
1870 if (! in_dead_block &&
1871 get_irn_pinned(pred) == op_pin_state_floats &&
1872 is_Block_unreachable(get_irn_n(pred, -1))) {
1873 set_nodes_block(pred, curr_block);
1875 waitq_put(worklist, pred);
1882 * Floating nodes form subgraphs that begin at nodes as Const, Load,
1883 * Start, Call and that end at op_pin_state_pinned nodes as Store, Call. Place_early
1884 * places all floating nodes reachable from its argument through floating
1885 * nodes and adds all beginnings at op_pin_state_pinned nodes to the worklist.
1887 * @param worklist a worklist, used for the algorithm, empty on in/output
1889 static void place_early(waitq *worklist) {
1891 inc_irg_visited(current_ir_graph);
1893 /* this inits the worklist */
1894 place_floats_early(get_irg_end(current_ir_graph), worklist);
1896 /* Work the content of the worklist. */
1897 while (!waitq_empty(worklist)) {
1898 ir_node *n = waitq_get(worklist);
1899 if (irn_not_visited(n))
1900 place_floats_early(n, worklist);
1903 set_irg_outs_inconsistent(current_ir_graph);
1904 set_irg_pinned(current_ir_graph, op_pin_state_pinned);
1908 * Compute the deepest common ancestor of block and dca.
1910 static ir_node *calc_dca(ir_node *dca, ir_node *block) {
1913 /* we do not want to place nodes in dead blocks */
1914 if (is_Block_dead(block))
1917 /* We found a first legal placement. */
1918 if (!dca) return block;
1920 /* Find a placement that is dominates both, dca and block. */
1921 while (get_Block_dom_depth(block) > get_Block_dom_depth(dca))
1922 block = get_Block_idom(block);
1924 while (get_Block_dom_depth(dca) > get_Block_dom_depth(block)) {
1925 dca = get_Block_idom(dca);
1928 while (block != dca) {
1929 block = get_Block_idom(block); dca = get_Block_idom(dca);
1935 /** Deepest common dominance ancestor of DCA and CONSUMER of PRODUCER.
1936 * I.e., DCA is the block where we might place PRODUCER.
1937 * A data flow edge points from producer to consumer.
1940 consumer_dom_dca(ir_node *dca, ir_node *consumer, ir_node *producer) {
1941 ir_node *block = NULL;
1943 /* Compute the latest block into which we can place a node so that it is
1945 if (get_irn_op(consumer) == op_Phi) {
1946 /* our consumer is a Phi-node, the effective use is in all those
1947 blocks through which the Phi-node reaches producer */
1949 ir_node *phi_block = get_nodes_block(consumer);
1950 irn_arity = get_irn_arity(consumer);
1952 for (i = 0; i < irn_arity; i++) {
1953 if (get_irn_n(consumer, i) == producer) {
1954 ir_node *new_block = get_nodes_block(get_Block_cfgpred(phi_block, i));
1956 if (! is_Block_unreachable(new_block))
1957 block = calc_dca(block, new_block);
1962 block = get_irn_n(producer, -1);
1964 assert(is_no_Block(consumer));
1965 block = get_nodes_block(consumer);
1968 /* Compute the deepest common ancestor of block and dca. */
1969 return calc_dca(dca, block);
1972 /* FIXME: the name clashes here with the function from ana/field_temperature.c
1974 static INLINE int get_irn_loop_depth(ir_node *n) {
1975 return get_loop_depth(get_irn_loop(n));
1979 * Move n to a block with less loop depth than it's current block. The
1980 * new block must be dominated by early.
1982 * @param n the node that should be moved
1983 * @param early the earliest block we can n move to
1985 static void move_out_of_loops(ir_node *n, ir_node *early) {
1986 ir_node *best, *dca;
1990 /* Find the region deepest in the dominator tree dominating
1991 dca with the least loop nesting depth, but still dominated
1992 by our early placement. */
1993 dca = get_nodes_block(n);
1996 while (dca != early) {
1997 dca = get_Block_idom(dca);
1998 if (!dca || is_Bad(dca)) break; /* may be Bad if not reachable from Start */
1999 if (get_irn_loop_depth(dca) < get_irn_loop_depth(best)) {
2003 if (best != get_nodes_block(n)) {
2005 printf("Moving out of loop: "); DDMN(n);
2006 printf(" Outermost block: "); DDMN(early);
2007 printf(" Best block: "); DDMN(best);
2008 printf(" Innermost block: "); DDMN(get_nodes_block(n));
2010 set_nodes_block(n, best);
2015 * Find the latest legal block for N and place N into the
2016 * `optimal' Block between the latest and earliest legal block.
2017 * The `optimal' block is the dominance-deepest block of those
2018 * with the least loop-nesting-depth. This places N out of as many
2019 * loops as possible and then makes it as control dependent as
2022 * @param n the node to be placed
2023 * @param worklist a worklist, all successors of non-floating nodes are
2026 static void place_floats_late(ir_node *n, pdeq *worklist) {
2030 assert(irn_not_visited(n)); /* no multiple placement */
2032 mark_irn_visited(n);
2034 /* no need to place block nodes, control nodes are already placed. */
2035 if ((get_irn_op(n) != op_Block) &&
2037 (get_irn_mode(n) != mode_X)) {
2038 /* Remember the early_blk placement of this block to move it
2039 out of loop no further than the early_blk placement. */
2040 early_blk = get_irn_n(n, -1);
2043 * BEWARE: Here we also get code, that is live, but
2044 * was in a dead block. If the node is life, but because
2045 * of CSE in a dead block, we still might need it.
2048 /* Assure that our users are all placed, except the Phi-nodes.
2049 --- Each data flow cycle contains at least one Phi-node. We
2050 have to break the `user has to be placed before the
2051 producer' dependence cycle and the Phi-nodes are the
2052 place to do so, because we need to base our placement on the
2053 final region of our users, which is OK with Phi-nodes, as they
2054 are op_pin_state_pinned, and they never have to be placed after a
2055 producer of one of their inputs in the same block anyway. */
2056 for (i = get_irn_n_outs(n) - 1; i >= 0; --i) {
2057 ir_node *succ = get_irn_out(n, i);
2058 if (irn_not_visited(succ) && (get_irn_op(succ) != op_Phi))
2059 place_floats_late(succ, worklist);
2062 if (! is_Block_dead(early_blk)) {
2063 /* do only move things that where not dead */
2064 ir_op *op = get_irn_op(n);
2066 /* We have to determine the final block of this node... except for
2067 constants and Projs */
2068 if ((get_irn_pinned(n) == op_pin_state_floats) &&
2070 (op != op_SymConst) &&
2073 ir_node *dca = NULL; /* deepest common ancestor in the
2074 dominator tree of all nodes'
2075 blocks depending on us; our final
2076 placement has to dominate DCA. */
2077 for (i = get_irn_n_outs(n) - 1; i >= 0; --i) {
2078 ir_node *succ = get_irn_out(n, i);
2081 if (get_irn_op(succ) == op_End) {
2083 * This consumer is the End node, a keep alive edge.
2084 * This is not a real consumer, so we ignore it
2089 /* ignore if succ is in dead code */
2090 succ_blk = get_irn_n(succ, -1);
2091 if (is_Block_unreachable(succ_blk))
2093 dca = consumer_dom_dca(dca, succ, n);
2096 set_nodes_block(n, dca);
2097 move_out_of_loops(n, early_blk);
2103 /* Add successors of all non-floating nodes on list. (Those of floating
2104 nodes are placed already and therefore are marked.) */
2105 for (i = 0; i < get_irn_n_outs(n); i++) {
2106 ir_node *succ = get_irn_out(n, i);
2107 if (irn_not_visited(get_irn_out(n, i))) {
2108 pdeq_putr(worklist, succ);
2114 * Place floating nodes on the given worklist as late as possible using
2115 * the dominance tree.
2117 * @param worklist the worklist containing the nodes to place
2119 static void place_late(waitq *worklist) {
2121 inc_irg_visited(current_ir_graph);
2123 /* This fills the worklist initially. */
2124 place_floats_late(get_irg_start_block(current_ir_graph), worklist);
2126 /* And now empty the worklist again... */
2127 while (!waitq_empty(worklist)) {
2128 ir_node *n = waitq_get(worklist);
2129 if (irn_not_visited(n))
2130 place_floats_late(n, worklist);
2134 /* Code Placement. */
2135 void place_code(ir_graph *irg) {
2137 ir_graph *rem = current_ir_graph;
2139 current_ir_graph = irg;
2141 /* Handle graph state */
2142 assert(get_irg_phase_state(irg) != phase_building);
2145 if (1 || get_irg_loopinfo_state(irg) != loopinfo_consistent) {
2146 free_loop_information(irg);
2147 construct_backedges(irg);
2150 /* Place all floating nodes as early as possible. This guarantees
2151 a legal code placement. */
2152 worklist = new_waitq();
2153 place_early(worklist);
2155 /* place_early() invalidates the outs, place_late needs them. */
2156 compute_irg_outs(irg);
2158 /* Now move the nodes down in the dominator tree. This reduces the
2159 unnecessary executions of the node. */
2160 place_late(worklist);
2162 set_irg_outs_inconsistent(current_ir_graph);
2163 set_irg_loopinfo_inconsistent(current_ir_graph);
2164 del_waitq(worklist);
2165 current_ir_graph = rem;
2169 * Called by walker of remove_critical_cf_edges().
2171 * Place an empty block to an edge between a blocks of multiple
2172 * predecessors and a block of multiple successors.
2175 * @param env Environment of walker. The changed field.
2177 static void walk_critical_cf_edges(ir_node *n, void *env) {
2179 ir_node *pre, *block, *jmp;
2181 ir_graph *irg = get_irn_irg(n);
2183 /* Block has multiple predecessors */
2184 arity = get_irn_arity(n);
2186 if (n == get_irg_end_block(irg))
2187 return; /* No use to add a block here. */
2189 for (i = 0; i < arity; ++i) {
2192 pre = get_irn_n(n, i);
2193 cfop = get_irn_op(skip_Proj(pre));
2194 /* Predecessor has multiple successors. Insert new control flow edge but
2195 ignore exception edges. */
2196 if (! is_op_fragile(cfop) && is_op_forking(cfop)) {
2197 /* set predecessor of new block */
2198 block = new_r_Block(irg, 1, &pre);
2199 /* insert new jmp node to new block */
2200 jmp = new_r_Jmp(irg, block);
2201 /* set successor of new block */
2202 set_irn_n(n, i, jmp);
2204 } /* predecessor has multiple successors */
2205 } /* for all predecessors */
2206 } /* n is a multi-entry block */
2209 void remove_critical_cf_edges(ir_graph *irg) {
2212 irg_block_walk_graph(irg, NULL, walk_critical_cf_edges, &changed);
2214 /* control flow changed */
2215 set_irg_outs_inconsistent(irg);
2216 set_irg_extblk_inconsistent(irg);
2217 set_irg_doms_inconsistent(irg);
2218 set_irg_loopinfo_inconsistent(irg);