2 * Copyright (C) 1995-2008 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 Dead node elimination and Procedure Inlining.
23 * @author Michael Beck, Goetz Lindenmaier
32 #include "irgraph_t.h"
35 #include "iroptimize.h"
52 #include "irbackedge_t.h"
58 #include "analyze_irg_args.h"
59 #include "iredges_t.h"
63 #include "iropt_dbg.h"
66 DEBUG_ONLY(static firm_dbg_module_t *dbg;)
68 /*------------------------------------------------------------------*/
69 /* Routines for dead node elimination / copying garbage collection */
71 /*------------------------------------------------------------------*/
74 * Remember the new node in the old node by using a field all nodes have.
76 #define set_new_node(oldn, newn) set_irn_link(oldn, newn)
79 * Get this new node, before the old node is forgotten.
81 #define get_new_node(oldn) get_irn_link(oldn)
84 * Check if a new node was set.
86 #define has_new_node(n) (get_new_node(n) != NULL)
89 * We use the block_visited flag to mark that we have computed the
90 * number of useful predecessors for this block.
91 * Further we encode the new arity in this flag in the old blocks.
92 * Remembering the arity is useful, as it saves a lot of pointer
93 * accesses. This function is called for all Phi and Block nodes
97 compute_new_arity(ir_node *b)
99 int i, res, irn_arity;
102 irg_v = get_irg_block_visited(current_ir_graph);
103 block_v = get_Block_block_visited(b);
104 if (block_v >= irg_v) {
105 /* we computed the number of preds for this block and saved it in the
107 return block_v - irg_v;
109 /* compute the number of good predecessors */
110 res = irn_arity = get_irn_arity(b);
111 for (i = 0; i < irn_arity; i++)
112 if (is_Bad(get_irn_n(b, i))) res--;
113 /* save it in the flag. */
114 set_Block_block_visited(b, irg_v + res);
120 * Copies the node to the new obstack. The Ins of the new node point to
121 * the predecessors on the old obstack. For block/phi nodes not all
122 * predecessors might be copied. n->link points to the new node.
123 * For Phi and Block nodes the function allocates in-arrays with an arity
124 * only for useful predecessors. The arity is determined by counting
125 * the non-bad predecessors of the block.
127 * @param n The node to be copied
128 * @param env if non-NULL, the node number attribute will be copied to the new node
130 * Note: Also used for loop unrolling.
132 static void copy_node(ir_node *n, void *env)
136 ir_op *op = get_irn_op(n);
140 /* node copied already */
142 } else if (op == op_Block) {
144 new_arity = compute_new_arity(n);
145 n->attr.block.graph_arr = NULL;
147 block = get_nodes_block(n);
149 new_arity = compute_new_arity(block);
151 new_arity = get_irn_arity(n);
154 nn = new_ir_node(get_irn_dbg_info(n),
161 /* Copy the attributes. These might point to additional data. If this
162 was allocated on the old obstack the pointers now are dangling. This
163 frees e.g. the memory of the graph_arr allocated in new_immBlock. */
164 if (op == op_Block) {
165 /* we cannot allow blocks WITHOUT macroblock input */
166 set_Block_MacroBlock(nn, get_Block_MacroBlock(n));
168 copy_node_attr(n, nn);
171 /* for easier debugging, we want to copy the node numbers too */
172 nn->node_nr = n->node_nr;
176 hook_dead_node_elim_subst(current_ir_graph, n, nn);
180 * Copies new predecessors of old node to new node remembered in link.
181 * Spare the Bad predecessors of Phi and Block nodes.
183 static void copy_preds(ir_node *n, void *env)
189 nn = get_new_node(n);
192 /* copy the macro block header */
193 ir_node *mbh = get_Block_MacroBlock(n);
196 /* this block is a macroblock header */
197 set_Block_MacroBlock(nn, nn);
199 /* get the macro block header */
200 ir_node *nmbh = get_new_node(mbh);
201 assert(nmbh != NULL);
202 set_Block_MacroBlock(nn, nmbh);
205 /* Don't copy Bad nodes. */
207 irn_arity = get_irn_arity(n);
208 for (i = 0; i < irn_arity; i++) {
209 if (! is_Bad(get_irn_n(n, i))) {
210 ir_node *pred = get_irn_n(n, i);
211 set_irn_n(nn, j, get_new_node(pred));
215 /* repair the block visited flag from above misuse. Repair it in both
216 graphs so that the old one can still be used. */
217 set_Block_block_visited(nn, 0);
218 set_Block_block_visited(n, 0);
219 /* Local optimization could not merge two subsequent blocks if
220 in array contained Bads. Now it's possible.
221 We don't call optimize_in_place as it requires
222 that the fields in ir_graph are set properly. */
223 if (!has_Block_entity(nn) &&
224 get_opt_control_flow_straightening() &&
225 get_Block_n_cfgpreds(nn) == 1 &&
226 is_Jmp(get_Block_cfgpred(nn, 0))) {
227 ir_node *old = get_nodes_block(get_Block_cfgpred(nn, 0));
229 /* Jmp jumps into the block it is in -- deal self cycle. */
230 assert(is_Bad(get_new_node(get_irg_bad(current_ir_graph))));
231 exchange(nn, get_new_node(get_irg_bad(current_ir_graph)));
236 } else if (is_Phi(n) && get_irn_arity(n) > 0) {
237 /* Don't copy node if corresponding predecessor in block is Bad.
238 The Block itself should not be Bad. */
239 block = get_nodes_block(n);
240 set_nodes_block(nn, get_new_node(block));
242 irn_arity = get_irn_arity(n);
243 for (i = 0; i < irn_arity; i++) {
244 if (! is_Bad(get_irn_n(block, i))) {
245 ir_node *pred = get_irn_n(n, i);
246 set_irn_n(nn, j, get_new_node(pred));
247 /*if (is_backedge(n, i)) set_backedge(nn, j);*/
251 /* If the pre walker reached this Phi after the post walker visited the
252 block block_visited is > 0. */
253 set_Block_block_visited(get_nodes_block(n), 0);
254 /* Compacting the Phi's ins might generate Phis with only one
256 if (get_irn_arity(nn) == 1)
257 exchange(nn, get_irn_n(nn, 0));
259 irn_arity = get_irn_arity(n);
260 for (i = -1; i < irn_arity; i++)
261 set_irn_n(nn, i, get_new_node(get_irn_n(n, i)));
263 /* Now the new node is complete. We can add it to the hash table for CSE.
264 @@@ inlining aborts if we identify End. Why? */
266 add_identities(current_ir_graph->value_table, nn);
270 * Copies the graph recursively, compacts the keep-alives of the end node.
272 * @param irg the graph to be copied
273 * @param copy_node_nr If non-zero, the node number will be copied
275 static void copy_graph(ir_graph *irg, int copy_node_nr)
277 ir_node *oe, *ne, *ob, *nb, *om, *nm; /* old end, new end, old bad, new bad, old NoMem, new NoMem */
278 ir_node *ka; /* keep alive */
282 /* Some nodes must be copied by hand, sigh */
283 vfl = get_irg_visited(irg);
284 set_irg_visited(irg, vfl + 1);
286 oe = get_irg_end(irg);
287 mark_irn_visited(oe);
288 /* copy the end node by hand, allocate dynamic in array! */
289 ne = new_ir_node(get_irn_dbg_info(oe),
296 /* Copy the attributes. Well, there might be some in the future... */
297 copy_node_attr(oe, ne);
298 set_new_node(oe, ne);
300 /* copy the Bad node */
301 ob = get_irg_bad(irg);
302 mark_irn_visited(ob);
303 nb = new_ir_node(get_irn_dbg_info(ob),
310 copy_node_attr(ob, nb);
311 set_new_node(ob, nb);
313 /* copy the NoMem node */
314 om = get_irg_no_mem(irg);
315 mark_irn_visited(om);
316 nm = new_ir_node(get_irn_dbg_info(om),
323 copy_node_attr(om, nm);
324 set_new_node(om, nm);
326 /* copy the live nodes */
327 set_irg_visited(irg, vfl);
328 irg_walk(get_nodes_block(oe), copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
330 /* Note: from yet, the visited flag of the graph is equal to vfl + 1 */
332 /* visit the anchors as well */
333 for (i = get_irg_n_anchors(irg) - 1; i >= 0; --i) {
334 ir_node *n = get_irg_anchor(irg, i);
336 if (n && (get_irn_visited(n) <= vfl)) {
337 set_irg_visited(irg, vfl);
338 irg_walk(n, copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
342 /* copy_preds for the end node ... */
343 set_nodes_block(ne, get_new_node(get_nodes_block(oe)));
345 /*- ... and now the keep alives. -*/
346 /* First pick the not marked block nodes and walk them. We must pick these
347 first as else we will oversee blocks reachable from Phis. */
348 irn_arity = get_End_n_keepalives(oe);
349 for (i = 0; i < irn_arity; i++) {
350 ka = get_End_keepalive(oe, i);
352 if (get_irn_visited(ka) <= vfl) {
353 /* We must keep the block alive and copy everything reachable */
354 set_irg_visited(irg, vfl);
355 irg_walk(ka, copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
357 add_End_keepalive(ne, get_new_node(ka));
361 /* Now pick other nodes. Here we will keep all! */
362 irn_arity = get_End_n_keepalives(oe);
363 for (i = 0; i < irn_arity; i++) {
364 ka = get_End_keepalive(oe, i);
366 if (get_irn_visited(ka) <= vfl) {
367 /* We didn't copy the node yet. */
368 set_irg_visited(irg, vfl);
369 irg_walk(ka, copy_node, copy_preds, INT_TO_PTR(copy_node_nr));
371 add_End_keepalive(ne, get_new_node(ka));
375 /* start block sometimes only reached after keep alives */
376 set_nodes_block(nb, get_new_node(get_nodes_block(ob)));
377 set_nodes_block(nm, get_new_node(get_nodes_block(om)));
381 * Copies the graph reachable from current_ir_graph->end to the obstack
382 * in current_ir_graph and fixes the environment.
383 * Then fixes the fields in current_ir_graph containing nodes of the
386 * @param copy_node_nr If non-zero, the node number will be copied
389 copy_graph_env(int copy_node_nr)
391 ir_graph *irg = current_ir_graph;
392 ir_node *old_end, *new_anchor;
395 /* remove end_except and end_reg nodes */
396 old_end = get_irg_end(irg);
397 set_irg_end_except (irg, old_end);
398 set_irg_end_reg (irg, old_end);
400 /* Not all nodes remembered in irg might be reachable
401 from the end node. Assure their link is set to NULL, so that
402 we can test whether new nodes have been computed. */
403 for (i = get_irg_n_anchors(irg) - 1; i >= 0; --i) {
404 ir_node *n = get_irg_anchor(irg, i);
406 set_new_node(n, NULL);
408 /* we use the block walk flag for removing Bads from Blocks ins. */
409 inc_irg_block_visited(irg);
412 copy_graph(irg, copy_node_nr);
415 old_end = get_irg_end(irg);
416 new_anchor = new_Anchor(irg);
418 for (i = get_irg_n_anchors(irg) - 1; i >= 0; --i) {
419 ir_node *n = get_irg_anchor(irg, i);
421 set_irn_n(new_anchor, i, get_new_node(n));
424 irg->anchor = new_anchor;
426 /* ensure the new anchor is placed in the endblock */
427 set_nodes_block(new_anchor, get_irg_end_block(irg));
431 * Copies all reachable nodes to a new obstack. Removes bad inputs
432 * from block nodes and the corresponding inputs from Phi nodes.
433 * Merges single exit blocks with single entry blocks and removes
435 * Adds all new nodes to a new hash table for CSE. Does not
436 * perform CSE, so the hash table might contain common subexpressions.
438 void dead_node_elimination(ir_graph *irg)
441 #ifdef INTERPROCEDURAL_VIEW
442 int rem_ipview = get_interprocedural_view();
444 struct obstack *graveyard_obst = NULL;
445 struct obstack *rebirth_obst = NULL;
447 edges_deactivate(irg);
449 /* inform statistics that we started a dead-node elimination run */
450 hook_dead_node_elim(irg, 1);
452 /* Remember external state of current_ir_graph. */
453 rem = current_ir_graph;
454 current_ir_graph = irg;
455 #ifdef INTERPROCEDURAL_VIEW
456 set_interprocedural_view(0);
459 assert(get_irg_phase_state(irg) != phase_building);
461 /* Handle graph state */
462 free_callee_info(irg);
466 /* @@@ so far we loose loops when copying */
467 free_loop_information(irg);
469 set_irg_doms_inconsistent(irg);
471 /* A quiet place, where the old obstack can rest in peace,
472 until it will be cremated. */
473 graveyard_obst = irg->obst;
475 /* A new obstack, where the reachable nodes will be copied to. */
476 rebirth_obst = XMALLOC(struct obstack);
477 irg->obst = rebirth_obst;
478 obstack_init(irg->obst);
479 irg->last_node_idx = 0;
481 /* We also need a new value table for CSE */
482 del_identities(irg->value_table);
483 irg->value_table = new_identities();
485 /* Copy the graph from the old to the new obstack */
486 copy_graph_env(/*copy_node_nr=*/1);
488 /* Free memory from old unoptimized obstack */
489 obstack_free(graveyard_obst, 0); /* First empty the obstack ... */
490 xfree(graveyard_obst); /* ... then free it. */
492 /* inform statistics that the run is over */
493 hook_dead_node_elim(irg, 0);
495 current_ir_graph = rem;
496 #ifdef INTERPROCEDURAL_VIEW
497 set_interprocedural_view(rem_ipview);
501 ir_graph_pass_t *dead_node_elimination_pass(const char *name)
503 return def_graph_pass(name ? name : "dce", dead_node_elimination);
507 * Relink bad predecessors of a block and store the old in array to the
508 * link field. This function is called by relink_bad_predecessors().
509 * The array of link field starts with the block operand at position 0.
510 * If block has bad predecessors, create a new in array without bad preds.
511 * Otherwise let in array untouched.
513 static void relink_bad_block_predecessors(ir_node *n, void *env)
515 ir_node **new_in, *irn;
516 int i, new_irn_n, old_irn_arity, new_irn_arity = 0;
519 /* if link field of block is NULL, look for bad predecessors otherwise
520 this is already done */
521 if (is_Block(n) && get_irn_link(n) == NULL) {
522 /* save old predecessors in link field (position 0 is the block operand)*/
523 set_irn_link(n, get_irn_in(n));
525 /* count predecessors without bad nodes */
526 old_irn_arity = get_irn_arity(n);
527 for (i = 0; i < old_irn_arity; i++)
528 if (!is_Bad(get_irn_n(n, i)))
531 /* arity changing: set new predecessors without bad nodes */
532 if (new_irn_arity < old_irn_arity) {
533 /* Get new predecessor array. We do not resize the array, as we must
534 keep the old one to update Phis. */
535 new_in = NEW_ARR_D(ir_node *, current_ir_graph->obst, (new_irn_arity+1));
537 /* set new predecessors in array */
540 for (i = 0; i < old_irn_arity; i++) {
541 irn = get_irn_n(n, i);
543 new_in[new_irn_n] = irn;
544 is_backedge(n, i) ? set_backedge(n, new_irn_n-1) : set_not_backedge(n, new_irn_n-1);
548 /* ARR_SETLEN(int, n->attr.block.backedge, new_irn_arity); */
549 ARR_SHRINKLEN(n->attr.block.backedge, new_irn_arity);
551 } /* ir node has bad predecessors */
552 } /* Block is not relinked */
556 * Relinks Bad predecessors from Blocks and Phis called by walker
557 * remove_bad_predecesors(). If n is a Block, call
558 * relink_bad_block_redecessors(). If n is a Phi-node, call also the relinking
559 * function of Phi's Block. If this block has bad predecessors, relink preds
562 static void relink_bad_predecessors(ir_node *n, void *env)
564 ir_node *block, **old_in;
565 int i, old_irn_arity, new_irn_arity;
567 /* relink bad predecessors of a block */
569 relink_bad_block_predecessors(n, env);
571 /* If Phi node relink its block and its predecessors */
573 /* Relink predecessors of phi's block */
574 block = get_nodes_block(n);
575 if (get_irn_link(block) == NULL)
576 relink_bad_block_predecessors(block, env);
578 old_in = (ir_node **)get_irn_link(block); /* Of Phi's Block */
579 old_irn_arity = ARR_LEN(old_in);
581 /* Relink Phi predecessors if count of predecessors changed */
582 if (old_irn_arity != ARR_LEN(get_irn_in(block))) {
583 /* set new predecessors in array
584 n->in[0] remains the same block */
586 for (i = 1; i < old_irn_arity; i++)
587 if (!is_Bad(old_in[i])) {
588 n->in[new_irn_arity] = n->in[i];
589 is_backedge(n, i) ? set_backedge(n, new_irn_arity) : set_not_backedge(n, new_irn_arity);
593 ARR_SETLEN(ir_node *, n->in, new_irn_arity);
594 ARR_SETLEN(int, n->attr.phi.u.backedge, new_irn_arity);
596 } /* n is a Phi node */
600 * Removes Bad Bad predecessors from Blocks and the corresponding
601 * inputs to Phi nodes as in dead_node_elimination but without
603 * On walking up set the link field to NULL, on walking down call
604 * relink_bad_predecessors() (This function stores the old in array
605 * to the link field and sets a new in array if arity of predecessors
608 void remove_bad_predecessors(ir_graph *irg)
610 panic("Fix backedge handling first");
611 irg_walk_graph(irg, firm_clear_link, relink_bad_predecessors, NULL);
618 __)|_| | \_/ | \_/(/_ |_/\__|__
620 The following stuff implements a facility that automatically patches
621 registered ir_node pointers to the new node when a dead node elimination occurs.
624 struct _survive_dce_t {
628 hook_entry_t dead_node_elim;
629 hook_entry_t dead_node_elim_subst;
632 typedef struct _survive_dce_list_t {
633 struct _survive_dce_list_t *next;
635 } survive_dce_list_t;
637 static void dead_node_hook(void *context, ir_graph *irg, int start)
639 survive_dce_t *sd = context;
642 /* Create a new map before the dead node elimination is performed. */
644 sd->new_places = pmap_create_ex(pmap_count(sd->places));
646 /* Patch back all nodes if dead node elimination is over and something is to be done. */
647 pmap_destroy(sd->places);
648 sd->places = sd->new_places;
649 sd->new_places = NULL;
654 * Hook called when dead node elimination replaces old by nw.
656 static void dead_node_subst_hook(void *context, ir_graph *irg, ir_node *old, ir_node *nw)
658 survive_dce_t *sd = context;
659 survive_dce_list_t *list = pmap_get(sd->places, old);
662 /* If the node is to be patched back, write the new address to all registered locations. */
664 survive_dce_list_t *p;
666 for (p = list; p; p = p->next)
669 pmap_insert(sd->new_places, nw, list);
674 * Make a new Survive DCE environment.
676 survive_dce_t *new_survive_dce(void)
678 survive_dce_t *res = XMALLOC(survive_dce_t);
679 obstack_init(&res->obst);
680 res->places = pmap_create();
681 res->new_places = NULL;
683 res->dead_node_elim.hook._hook_dead_node_elim = dead_node_hook;
684 res->dead_node_elim.context = res;
685 res->dead_node_elim.next = NULL;
687 res->dead_node_elim_subst.hook._hook_dead_node_elim_subst = dead_node_subst_hook;
688 res->dead_node_elim_subst.context = res;
689 res->dead_node_elim_subst.next = NULL;
691 register_hook(hook_dead_node_elim, &res->dead_node_elim);
692 register_hook(hook_dead_node_elim_subst, &res->dead_node_elim_subst);
697 * Free a Survive DCE environment.
699 void free_survive_dce(survive_dce_t *sd)
701 obstack_free(&sd->obst, NULL);
702 pmap_destroy(sd->places);
703 unregister_hook(hook_dead_node_elim, &sd->dead_node_elim);
704 unregister_hook(hook_dead_node_elim_subst, &sd->dead_node_elim_subst);
709 * Register a node pointer to be patched upon DCE.
710 * When DCE occurs, the node pointer specified by @p place will be
711 * patched to the new address of the node it is pointing to.
713 * @param sd The Survive DCE environment.
714 * @param place The address of the node pointer.
716 void survive_dce_register_irn(survive_dce_t *sd, ir_node **place)
718 if (*place != NULL) {
719 ir_node *irn = *place;
720 survive_dce_list_t *curr = pmap_get(sd->places, irn);
721 survive_dce_list_t *nw = OALLOC(&sd->obst, survive_dce_list_t);
726 pmap_insert(sd->places, irn, nw);
730 /*--------------------------------------------------------------------*/
731 /* Functionality for inlining */
732 /*--------------------------------------------------------------------*/
735 * Copy node for inlineing. Updates attributes that change when
736 * inlineing but not for dead node elimination.
738 * Copies the node by calling copy_node() and then updates the entity if
739 * it's a local one. env must be a pointer of the frame type of the
740 * inlined procedure. The new entities must be in the link field of
743 static void copy_node_inline(ir_node *n, void *env)
746 ir_type *frame_tp = (ir_type *)env;
750 nn = get_new_node(n);
752 /* use copied entities from the new frame */
753 if (get_entity_owner(get_Sel_entity(n)) == frame_tp) {
754 set_Sel_entity(nn, get_entity_link(get_Sel_entity(n)));
756 } else if (is_Block(n)) {
757 nn = get_new_node(n);
758 nn->attr.block.irg.irg = current_ir_graph;
763 * Copies new predecessors of old node and move constants to
766 static void copy_preds_inline(ir_node *n, void *env)
771 nn = skip_Id(get_new_node(n));
772 if (is_irn_constlike(nn)) {
773 /* move Constants into the start block */
774 set_nodes_block(nn, get_irg_start_block(current_ir_graph));
776 n = identify_remember(current_ir_graph->value_table, nn);
785 * Walker: checks if P_value_arg_base is used.
787 static void find_addr(ir_node *node, void *env)
789 int *allow_inline = env;
791 ir_graph *irg = current_ir_graph;
792 if (get_Sel_ptr(node) == get_irg_frame(irg)) {
793 /* access to frame */
794 ir_entity *ent = get_Sel_entity(node);
795 if (get_entity_owner(ent) != get_irg_frame_type(irg)) {
796 /* access to value_type */
800 } else if (is_Alloc(node) && get_Alloc_where(node) == stack_alloc) {
802 * Refuse to inline alloca call unless user explicitly forced so as this
803 * may change program's memory overhead drastically when the function
804 * using alloca is called in loop. In GCC present in SPEC2000 inlining
805 * into schedule_block cause it to require 2GB of ram instead of 256MB.
807 * Sorrily this is true with our implementation also.
808 * Moreover, we cannot differentiate between alloca() and VLA yet, so this
809 * disables inlining of functions using VLA (with are completely save).
812 * - add a flag to the Alloc node for "real" alloca() calls
813 * - add a new Stack-Restore node at the end of a function using alloca()
820 * Check if we can inline a given call.
821 * Currently, we cannot inline two cases:
822 * - call with compound arguments
823 * - graphs that take the address of a parameter
825 * check these conditions here
827 static int can_inline(ir_node *call, ir_graph *called_graph)
829 ir_type *call_type = get_Call_type(call);
830 int params, ress, i, res;
831 assert(is_Method_type(call_type));
833 params = get_method_n_params(call_type);
834 ress = get_method_n_ress(call_type);
836 /* check parameters for compound arguments */
837 for (i = 0; i < params; ++i) {
838 ir_type *p_type = get_method_param_type(call_type, i);
840 if (is_compound_type(p_type))
844 /* check results for compound arguments */
845 for (i = 0; i < ress; ++i) {
846 ir_type *r_type = get_method_res_type(call_type, i);
848 if (is_compound_type(r_type))
853 irg_walk_graph(called_graph, find_addr, NULL, &res);
859 exc_handler, /**< There is a handler. */
860 exc_no_handler /**< Exception handling not represented. */
863 /* Inlines a method at the given call site. */
864 int inline_method(ir_node *call, ir_graph *called_graph)
867 ir_node *post_call, *post_bl;
868 ir_node *in[pn_Start_max];
869 ir_node *end, *end_bl, *block;
874 int arity, n_ret, n_exc, n_res, i, n, j, rem_opt, irn_arity, n_params;
876 enum exc_mode exc_handling;
877 ir_type *called_frame, *curr_frame, *mtp, *ctp;
880 irg_inline_property prop = get_irg_inline_property(called_graph);
881 unsigned long visited;
883 if (prop == irg_inline_forbidden)
886 ent = get_irg_entity(called_graph);
888 mtp = get_entity_type(ent);
889 ctp = get_Call_type(call);
890 n_params = get_method_n_params(mtp);
891 n_res = get_method_n_ress(mtp);
892 if (n_params > get_method_n_params(ctp)) {
893 /* this is a bad feature of C: without a prototype, we can
894 * call a function with less parameters than needed. Currently
895 * we don't support this, although we could use Unknown than. */
898 if (n_res != get_method_n_ress(ctp)) {
902 /* Argh, compiling C has some bad consequences:
903 * It is implementation dependent what happens in that case.
904 * We support inlining, if the bitsize of the types matches AND
905 * the same arithmetic is used. */
906 for (i = n_params - 1; i >= 0; --i) {
907 ir_type *param_tp = get_method_param_type(mtp, i);
908 ir_type *arg_tp = get_method_param_type(ctp, i);
910 if (param_tp != arg_tp) {
911 ir_mode *pmode = get_type_mode(param_tp);
912 ir_mode *amode = get_type_mode(arg_tp);
914 if (pmode == NULL || amode == NULL)
916 if (get_mode_size_bits(pmode) != get_mode_size_bits(amode))
918 if (get_mode_arithmetic(pmode) != get_mode_arithmetic(amode))
920 /* otherwise we can simply "reinterpret" the bits */
923 for (i = n_res - 1; i >= 0; --i) {
924 ir_type *decl_res_tp = get_method_res_type(mtp, i);
925 ir_type *used_res_tp = get_method_res_type(ctp, i);
927 if (decl_res_tp != used_res_tp) {
928 ir_mode *decl_mode = get_type_mode(decl_res_tp);
929 ir_mode *used_mode = get_type_mode(used_res_tp);
930 if (decl_mode == NULL || used_mode == NULL)
932 if (get_mode_size_bits(decl_mode) != get_mode_size_bits(used_mode))
934 if (get_mode_arithmetic(decl_mode) != get_mode_arithmetic(used_mode))
936 /* otherwise we can "reinterpret" the bits */
940 irg = get_irn_irg(call);
943 * We cannot inline a recursive call. The graph must be copied before
944 * the call the inline_method() using create_irg_copy().
946 if (called_graph == irg)
950 * currently, we cannot inline two cases:
951 * - call with compound arguments
952 * - graphs that take the address of a parameter
954 if (! can_inline(call, called_graph))
957 rem = current_ir_graph;
958 current_ir_graph = irg;
960 DB((dbg, LEVEL_1, "Inlining %+F(%+F) into %+F\n", call, called_graph, irg));
962 /* -- Turn off optimizations, this can cause problems when allocating new nodes. -- */
963 rem_opt = get_opt_optimize();
966 /* Handle graph state */
967 assert(get_irg_phase_state(irg) != phase_building);
968 assert(get_irg_pinned(irg) == op_pin_state_pinned);
969 assert(get_irg_pinned(called_graph) == op_pin_state_pinned);
970 set_irg_outs_inconsistent(irg);
971 set_irg_extblk_inconsistent(irg);
972 set_irg_doms_inconsistent(irg);
973 set_irg_loopinfo_inconsistent(irg);
974 set_irg_callee_info_state(irg, irg_callee_info_inconsistent);
975 set_irg_entity_usage_state(irg, ir_entity_usage_not_computed);
977 /* -- Check preconditions -- */
978 assert(is_Call(call));
980 /* here we know we WILL inline, so inform the statistics */
981 hook_inline(call, called_graph);
983 /* -- Decide how to handle exception control flow: Is there a handler
984 for the Call node, or do we branch directly to End on an exception?
986 0 There is a handler.
987 2 Exception handling not represented in Firm. -- */
989 ir_node *Xproj = NULL;
991 for (proj = get_irn_link(call); proj; proj = get_irn_link(proj)) {
992 long proj_nr = get_Proj_proj(proj);
993 if (proj_nr == pn_Call_X_except) Xproj = proj;
995 exc_handling = Xproj != NULL ? exc_handler : exc_no_handler;
998 /* create the argument tuple */
999 NEW_ARR_A(ir_type *, args_in, n_params);
1001 block = get_nodes_block(call);
1002 for (i = n_params - 1; i >= 0; --i) {
1003 ir_node *arg = get_Call_param(call, i);
1004 ir_type *param_tp = get_method_param_type(mtp, i);
1005 ir_mode *mode = get_type_mode(param_tp);
1007 if (mode != get_irn_mode(arg)) {
1008 arg = new_r_Conv(block, arg, mode);
1014 the procedure and later replaces the Start node of the called graph.
1015 Post_call is the old Call node and collects the results of the called
1016 graph. Both will end up being a tuple. -- */
1017 post_bl = get_nodes_block(call);
1018 set_irg_current_block(irg, post_bl);
1019 /* XxMxPxPxPxT of Start + parameter of Call */
1020 in[pn_Start_X_initial_exec] = new_Jmp();
1021 in[pn_Start_M] = get_Call_mem(call);
1022 in[pn_Start_P_frame_base] = get_irg_frame(irg);
1023 in[pn_Start_P_tls] = get_irg_tls(irg);
1024 in[pn_Start_T_args] = new_Tuple(n_params, args_in);
1025 pre_call = new_Tuple(pn_Start_max, in);
1029 The new block gets the ins of the old block, pre_call and all its
1030 predecessors and all Phi nodes. -- */
1031 part_block(pre_call);
1033 /* -- Prepare state for dead node elimination -- */
1034 /* Visited flags in calling irg must be >= flag in called irg.
1035 Else walker and arity computation will not work. */
1036 if (get_irg_visited(irg) <= get_irg_visited(called_graph))
1037 set_irg_visited(irg, get_irg_visited(called_graph) + 1);
1038 if (get_irg_block_visited(irg) < get_irg_block_visited(called_graph))
1039 set_irg_block_visited(irg, get_irg_block_visited(called_graph));
1040 visited = get_irg_visited(irg);
1042 /* Set pre_call as new Start node in link field of the start node of
1043 calling graph and pre_calls block as new block for the start block
1045 Further mark these nodes so that they are not visited by the
1047 set_irn_link(get_irg_start(called_graph), pre_call);
1048 set_irn_visited(get_irg_start(called_graph), visited);
1049 set_irn_link(get_irg_start_block(called_graph), get_nodes_block(pre_call));
1050 set_irn_visited(get_irg_start_block(called_graph), visited);
1052 set_irn_link(get_irg_bad(called_graph), get_irg_bad(current_ir_graph));
1053 set_irn_visited(get_irg_bad(called_graph), visited);
1055 set_irn_link(get_irg_no_mem(called_graph), get_irg_no_mem(current_ir_graph));
1056 set_irn_visited(get_irg_no_mem(called_graph), visited);
1058 /* Initialize for compaction of in arrays */
1059 inc_irg_block_visited(irg);
1061 /* -- Replicate local entities of the called_graph -- */
1062 /* copy the entities. */
1063 irp_reserve_resources(irp, IR_RESOURCE_ENTITY_LINK);
1064 called_frame = get_irg_frame_type(called_graph);
1065 curr_frame = get_irg_frame_type(irg);
1066 for (i = 0, n = get_class_n_members(called_frame); i < n; ++i) {
1067 ir_entity *new_ent, *old_ent;
1068 old_ent = get_class_member(called_frame, i);
1069 new_ent = copy_entity_own(old_ent, curr_frame);
1070 set_entity_link(old_ent, new_ent);
1073 /* visited is > than that of called graph. With this trick visited will
1074 remain unchanged so that an outer walker, e.g., searching the call nodes
1075 to inline, calling this inline will not visit the inlined nodes. */
1076 set_irg_visited(irg, get_irg_visited(irg)-1);
1078 /* -- Performing dead node elimination inlines the graph -- */
1079 /* Copies the nodes to the obstack of current_ir_graph. Updates links to new
1081 irg_walk(get_irg_end(called_graph), copy_node_inline, copy_preds_inline,
1082 get_irg_frame_type(called_graph));
1084 irp_free_resources(irp, IR_RESOURCE_ENTITY_LINK);
1086 /* Repair called_graph */
1087 set_irg_visited(called_graph, get_irg_visited(irg));
1088 set_irg_block_visited(called_graph, get_irg_block_visited(irg));
1089 set_Block_block_visited(get_irg_start_block(called_graph), 0);
1091 /* -- Merge the end of the inlined procedure with the call site -- */
1092 /* We will turn the old Call node into a Tuple with the following
1095 0: Phi of all Memories of Return statements.
1096 1: Jmp from new Block that merges the control flow from all exception
1097 predecessors of the old end block.
1098 2: Tuple of all arguments.
1099 3: Phi of Exception memories.
1100 In case the old Call directly branches to End on an exception we don't
1101 need the block merging all exceptions nor the Phi of the exception
1105 /* -- Precompute some values -- */
1106 end_bl = get_new_node(get_irg_end_block(called_graph));
1107 end = get_new_node(get_irg_end(called_graph));
1108 arity = get_Block_n_cfgpreds(end_bl); /* arity = n_exc + n_ret */
1109 n_res = get_method_n_ress(get_Call_type(call));
1111 res_pred = XMALLOCN(ir_node*, n_res);
1112 cf_pred = XMALLOCN(ir_node*, arity);
1114 set_irg_current_block(irg, post_bl); /* just to make sure */
1116 /* -- archive keepalives -- */
1117 irn_arity = get_irn_arity(end);
1118 for (i = 0; i < irn_arity; i++) {
1119 ir_node *ka = get_End_keepalive(end, i);
1121 add_End_keepalive(get_irg_end(irg), ka);
1124 /* The new end node will die. We need not free as the in array is on the obstack:
1125 copy_node() only generated 'D' arrays. */
1127 /* -- Replace Return nodes by Jump nodes. -- */
1129 for (i = 0; i < arity; i++) {
1131 ret = get_Block_cfgpred(end_bl, i);
1132 if (is_Return(ret)) {
1133 cf_pred[n_ret] = new_r_Jmp(get_nodes_block(ret));
1137 set_irn_in(post_bl, n_ret, cf_pred);
1139 /* -- Build a Tuple for all results of the method.
1140 Add Phi node if there was more than one Return. -- */
1141 turn_into_tuple(post_call, pn_Call_max);
1142 /* First the Memory-Phi */
1144 for (i = 0; i < arity; i++) {
1145 ret = get_Block_cfgpred(end_bl, i);
1146 if (is_Return(ret)) {
1147 cf_pred[n_mem_phi++] = get_Return_mem(ret);
1149 /* memory output for some exceptions is directly connected to End */
1151 cf_pred[n_mem_phi++] = new_r_Proj(get_nodes_block(ret), ret, mode_M, 3);
1152 } else if (is_fragile_op(ret)) {
1153 /* We rely that all cfops have the memory output at the same position. */
1154 cf_pred[n_mem_phi++] = new_r_Proj(get_nodes_block(ret), ret, mode_M, 0);
1155 } else if (is_Raise(ret)) {
1156 cf_pred[n_mem_phi++] = new_r_Proj(get_nodes_block(ret), ret, mode_M, 1);
1159 phi = new_Phi(n_mem_phi, cf_pred, mode_M);
1160 set_Tuple_pred(call, pn_Call_M, phi);
1161 /* Conserve Phi-list for further inlinings -- but might be optimized */
1162 if (get_nodes_block(phi) == post_bl) {
1163 set_irn_link(phi, get_irn_link(post_bl));
1164 set_irn_link(post_bl, phi);
1166 /* Now the real results */
1168 for (j = 0; j < n_res; j++) {
1169 ir_type *res_type = get_method_res_type(ctp, j);
1170 ir_mode *res_mode = get_type_mode(res_type);
1172 for (i = 0; i < arity; i++) {
1173 ret = get_Block_cfgpred(end_bl, i);
1174 if (is_Return(ret)) {
1175 ir_node *res = get_Return_res(ret, j);
1176 if (get_irn_mode(res) != res_mode) {
1177 ir_node *block = get_nodes_block(res);
1178 res = new_r_Conv(block, res, res_mode);
1180 cf_pred[n_ret] = res;
1185 phi = new_Phi(n_ret, cf_pred, get_irn_mode(cf_pred[0]));
1189 /* Conserve Phi-list for further inlinings -- but might be optimized */
1190 if (get_nodes_block(phi) == post_bl) {
1191 set_Phi_next(phi, get_Block_phis(post_bl));
1192 set_Block_phis(post_bl, phi);
1195 set_Tuple_pred(call, pn_Call_T_result, new_Tuple(n_res, res_pred));
1197 set_Tuple_pred(call, pn_Call_T_result, new_Bad());
1199 /* handle the regular call */
1200 set_Tuple_pred(call, pn_Call_X_regular, new_Jmp());
1202 /* For now, we cannot inline calls with value_base */
1203 set_Tuple_pred(call, pn_Call_P_value_res_base, new_Bad());
1205 /* Finally the exception control flow.
1206 We have two possible situations:
1207 First if the Call branches to an exception handler:
1208 We need to add a Phi node to
1209 collect the memory containing the exception objects. Further we need
1210 to add another block to get a correct representation of this Phi. To
1211 this block we add a Jmp that resolves into the X output of the Call
1212 when the Call is turned into a tuple.
1213 Second: There is no exception edge. Just add all inlined exception
1214 branches to the End node.
1216 if (exc_handling == exc_handler) {
1218 for (i = 0; i < arity; i++) {
1220 ret = get_Block_cfgpred(end_bl, i);
1221 irn = skip_Proj(ret);
1222 if (is_fragile_op(irn) || is_Raise(irn)) {
1223 cf_pred[n_exc] = ret;
1230 set_Tuple_pred(call, pn_Call_X_except, cf_pred[0]);
1232 ir_node *block = new_Block(n_exc, cf_pred);
1233 set_cur_block(block);
1234 set_Tuple_pred(call, pn_Call_X_except, new_Jmp());
1237 set_Tuple_pred(call, pn_Call_X_except, new_Bad());
1240 ir_node *main_end_bl;
1241 int main_end_bl_arity;
1242 ir_node **end_preds;
1244 /* assert(exc_handling == 1 || no exceptions. ) */
1246 for (i = 0; i < arity; i++) {
1247 ir_node *ret = get_Block_cfgpred(end_bl, i);
1248 ir_node *irn = skip_Proj(ret);
1250 if (is_fragile_op(irn) || is_Raise(irn)) {
1251 cf_pred[n_exc] = ret;
1255 main_end_bl = get_irg_end_block(irg);
1256 main_end_bl_arity = get_irn_arity(main_end_bl);
1257 end_preds = XMALLOCN(ir_node*, n_exc + main_end_bl_arity);
1259 for (i = 0; i < main_end_bl_arity; ++i)
1260 end_preds[i] = get_irn_n(main_end_bl, i);
1261 for (i = 0; i < n_exc; ++i)
1262 end_preds[main_end_bl_arity + i] = cf_pred[i];
1263 set_irn_in(main_end_bl, n_exc + main_end_bl_arity, end_preds);
1264 set_Tuple_pred(call, pn_Call_X_except, new_Bad());
1270 /* -- Turn CSE back on. -- */
1271 set_optimize(rem_opt);
1272 current_ir_graph = rem;
1277 /********************************************************************/
1278 /* Apply inlining to small methods. */
1279 /********************************************************************/
1281 static struct obstack temp_obst;
1283 /** Represents a possible inlinable call in a graph. */
1284 typedef struct _call_entry {
1285 ir_node *call; /**< The Call node. */
1286 ir_graph *callee; /**< The callee IR-graph. */
1287 list_head list; /**< List head for linking the next one. */
1288 int loop_depth; /**< The loop depth of this call. */
1289 int benefice; /**< The calculated benefice of this call. */
1290 unsigned local_adr:1; /**< Set if this call gets an address of a local variable. */
1291 unsigned all_const:1; /**< Set if this call has only constant parameters. */
1295 * environment for inlining small irgs
1297 typedef struct _inline_env_t {
1298 struct obstack obst; /**< An obstack where call_entries are allocated on. */
1299 list_head calls; /**< The call entry list. */
1303 * Returns the irg called from a Call node. If the irg is not
1304 * known, NULL is returned.
1306 * @param call the call node
1308 static ir_graph *get_call_called_irg(ir_node *call)
1312 addr = get_Call_ptr(call);
1313 if (is_Global(addr)) {
1314 ir_entity *ent = get_Global_entity(addr);
1315 /* we don't know which function gets finally bound to a weak symbol */
1316 if (get_entity_linkage(ent) & IR_LINKAGE_WEAK)
1319 return get_entity_irg(ent);
1326 * Walker: Collect all calls to known graphs inside a graph.
1328 static void collect_calls(ir_node *call, void *env)
1331 if (is_Call(call)) {
1332 ir_graph *called_irg = get_call_called_irg(call);
1334 if (called_irg != NULL) {
1335 /* The Call node calls a locally defined method. Remember to inline. */
1336 inline_env_t *ienv = env;
1337 call_entry *entry = OALLOC(&ienv->obst, call_entry);
1339 entry->callee = called_irg;
1340 entry->loop_depth = 0;
1341 entry->benefice = 0;
1342 entry->local_adr = 0;
1343 entry->all_const = 0;
1345 list_add_tail(&entry->list, &ienv->calls);
1351 * Inlines all small methods at call sites where the called address comes
1352 * from a Const node that references the entity representing the called
1354 * The size argument is a rough measure for the code size of the method:
1355 * Methods where the obstack containing the firm graph is smaller than
1358 void inline_small_irgs(ir_graph *irg, int size)
1360 ir_graph *rem = current_ir_graph;
1364 current_ir_graph = irg;
1365 /* Handle graph state */
1366 assert(get_irg_phase_state(irg) != phase_building);
1367 free_callee_info(irg);
1369 /* Find Call nodes to inline.
1370 (We can not inline during a walk of the graph, as inlining the same
1371 method several times changes the visited flag of the walked graph:
1372 after the first inlining visited of the callee equals visited of
1373 the caller. With the next inlining both are increased.) */
1374 obstack_init(&env.obst);
1375 INIT_LIST_HEAD(&env.calls);
1376 irg_walk_graph(irg, NULL, collect_calls, &env);
1378 if (! list_empty(&env.calls)) {
1379 /* There are calls to inline */
1380 ir_reserve_resources(irg, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1381 collect_phiprojs(irg);
1383 list_for_each_entry(call_entry, entry, &env.calls, list) {
1384 ir_graph *callee = entry->callee;
1385 irg_inline_property prop = get_irg_inline_property(callee);
1387 if (prop == irg_inline_forbidden) {
1391 if (prop >= irg_inline_forced ||
1392 _obstack_memory_used(callee->obst) - (int)obstack_room(callee->obst) < size) {
1393 inline_method(entry->call, callee);
1396 ir_free_resources(irg, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1398 obstack_free(&env.obst, NULL);
1399 current_ir_graph = rem;
1402 struct inline_small_irgs_pass_t {
1403 ir_graph_pass_t pass;
1408 * Wrapper to run inline_small_irgs() as a pass.
1410 static int inline_small_irgs_wrapper(ir_graph *irg, void *context)
1412 struct inline_small_irgs_pass_t *pass = context;
1414 inline_small_irgs(irg, pass->size);
1418 /* create a pass for inline_small_irgs() */
1419 ir_graph_pass_t *inline_small_irgs_pass(const char *name, int size)
1421 struct inline_small_irgs_pass_t *pass =
1422 XMALLOCZ(struct inline_small_irgs_pass_t);
1425 return def_graph_pass_constructor(
1426 &pass->pass, name ? name : "inline_small_irgs", inline_small_irgs_wrapper);
1430 * Environment for inlining irgs.
1433 list_head calls; /**< List of of all call nodes in this graph. */
1434 unsigned *local_weights; /**< Once allocated, the beneficial weight for transmitting local addresses. */
1435 unsigned n_nodes; /**< Number of nodes in graph except Id, Tuple, Proj, Start, End. */
1436 unsigned n_blocks; /**< Number of Blocks in graph without Start and End block. */
1437 unsigned n_nodes_orig; /**< for statistics */
1438 unsigned n_call_nodes; /**< Number of Call nodes in the graph. */
1439 unsigned n_call_nodes_orig; /**< for statistics */
1440 unsigned n_callers; /**< Number of known graphs that call this graphs. */
1441 unsigned n_callers_orig; /**< for statistics */
1442 unsigned got_inline:1; /**< Set, if at least one call inside this graph was inlined. */
1443 unsigned recursive:1; /**< Set, if this function is self recursive. */
1447 * Allocate a new environment for inlining.
1449 static inline_irg_env *alloc_inline_irg_env(void)
1451 inline_irg_env *env = OALLOC(&temp_obst, inline_irg_env);
1452 INIT_LIST_HEAD(&env->calls);
1453 env->local_weights = NULL;
1454 env->n_nodes = -2; /* do not count count Start, End */
1455 env->n_blocks = -2; /* do not count count Start, End Block */
1456 env->n_nodes_orig = -2; /* do not count Start, End */
1457 env->n_call_nodes = 0;
1458 env->n_call_nodes_orig = 0;
1460 env->n_callers_orig = 0;
1461 env->got_inline = 0;
1466 typedef struct walker_env {
1467 inline_irg_env *x; /**< the inline environment */
1468 char ignore_runtime; /**< the ignore runtime flag */
1469 char ignore_callers; /**< if set, do change callers data */
1473 * post-walker: collect all calls in the inline-environment
1474 * of a graph and sum some statistics.
1476 static void collect_calls2(ir_node *call, void *ctx)
1479 inline_irg_env *x = env->x;
1480 ir_opcode code = get_irn_opcode(call);
1484 /* count meaningful nodes in irg */
1485 if (code != iro_Proj && code != iro_Tuple && code != iro_Sync) {
1486 if (code != iro_Block) {
1494 if (code != iro_Call) return;
1496 /* check, if it's a runtime call */
1497 if (env->ignore_runtime) {
1498 ir_node *symc = get_Call_ptr(call);
1500 if (is_Global(symc)) {
1501 ir_entity *ent = get_Global_entity(symc);
1503 if (get_entity_additional_properties(ent) & mtp_property_runtime)
1508 /* collect all call nodes */
1510 ++x->n_call_nodes_orig;
1512 callee = get_call_called_irg(call);
1513 if (callee != NULL) {
1514 if (! env->ignore_callers) {
1515 inline_irg_env *callee_env = get_irg_link(callee);
1516 /* count all static callers */
1517 ++callee_env->n_callers;
1518 ++callee_env->n_callers_orig;
1520 if (callee == current_ir_graph)
1523 /* link it in the list of possible inlinable entries */
1524 entry = OALLOC(&temp_obst, call_entry);
1526 entry->callee = callee;
1527 entry->loop_depth = get_irn_loop(get_nodes_block(call))->depth;
1528 entry->benefice = 0;
1529 entry->local_adr = 0;
1530 entry->all_const = 0;
1532 list_add_tail(&entry->list, &x->calls);
1537 * Returns TRUE if the number of callers is 0 in the irg's environment,
1538 * hence this irg is a leave.
1540 inline static int is_leave(ir_graph *irg)
1542 inline_irg_env *env = get_irg_link(irg);
1543 return env->n_call_nodes == 0;
1547 * Returns TRUE if the number of nodes in the callee is
1548 * smaller then size in the irg's environment.
1550 inline static int is_smaller(ir_graph *callee, unsigned size)
1552 inline_irg_env *env = get_irg_link(callee);
1553 return env->n_nodes < size;
1557 * Duplicate a call entry.
1559 * @param entry the original entry to duplicate
1560 * @param new_call the new call node
1561 * @param loop_depth_delta
1562 * delta value for the loop depth
1564 static call_entry *duplicate_call_entry(const call_entry *entry,
1565 ir_node *new_call, int loop_depth_delta) {
1566 call_entry *nentry = OALLOC(&temp_obst, call_entry);
1567 nentry->call = new_call;
1568 nentry->callee = entry->callee;
1569 nentry->benefice = entry->benefice;
1570 nentry->loop_depth = entry->loop_depth + loop_depth_delta;
1571 nentry->local_adr = entry->local_adr;
1572 nentry->all_const = entry->all_const;
1578 * Append all call nodes of the source environment to the nodes of in the destination
1581 * @param dst destination environment
1582 * @param src source environment
1583 * @param loop_depth the loop depth of the call that is replaced by the src list
1585 static void append_call_list(inline_irg_env *dst, inline_irg_env *src, int loop_depth)
1587 call_entry *entry, *nentry;
1589 /* Note that the src list points to Call nodes in the inlined graph, but
1590 we need Call nodes in our graph. Luckily the inliner leaves this information
1591 in the link field. */
1592 list_for_each_entry(call_entry, entry, &src->calls, list) {
1593 nentry = duplicate_call_entry(entry, get_irn_link(entry->call), loop_depth);
1594 list_add_tail(&nentry->list, &dst->calls);
1596 dst->n_call_nodes += src->n_call_nodes;
1597 dst->n_nodes += src->n_nodes;
1601 * Inlines small leave methods at call sites where the called address comes
1602 * from a Const node that references the entity representing the called
1604 * The size argument is a rough measure for the code size of the method:
1605 * Methods where the obstack containing the firm graph is smaller than
1608 void inline_leave_functions(unsigned maxsize, unsigned leavesize,
1609 unsigned size, int ignore_runtime)
1611 inline_irg_env *env;
1617 call_entry *entry, *next;
1618 const call_entry *centry;
1619 pmap *copied_graphs;
1620 pmap_entry *pm_entry;
1622 rem = current_ir_graph;
1623 obstack_init(&temp_obst);
1625 /* a map for the copied graphs, used to inline recursive calls */
1626 copied_graphs = pmap_create();
1628 /* extend all irgs by a temporary data structure for inlining. */
1629 n_irgs = get_irp_n_irgs();
1630 for (i = 0; i < n_irgs; ++i)
1631 set_irg_link(get_irp_irg(i), alloc_inline_irg_env());
1633 /* Pre-compute information in temporary data structure. */
1634 wenv.ignore_runtime = ignore_runtime;
1635 wenv.ignore_callers = 0;
1636 for (i = 0; i < n_irgs; ++i) {
1637 ir_graph *irg = get_irp_irg(i);
1639 assert(get_irg_phase_state(irg) != phase_building);
1640 free_callee_info(irg);
1642 assure_cf_loop(irg);
1643 wenv.x = get_irg_link(irg);
1644 irg_walk_graph(irg, NULL, collect_calls2, &wenv);
1647 /* -- and now inline. -- */
1649 /* Inline leaves recursively -- we might construct new leaves. */
1653 for (i = 0; i < n_irgs; ++i) {
1655 int phiproj_computed = 0;
1657 current_ir_graph = get_irp_irg(i);
1658 env = get_irg_link(current_ir_graph);
1660 ir_reserve_resources(current_ir_graph, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1661 list_for_each_entry_safe(call_entry, entry, next, &env->calls, list) {
1663 irg_inline_property prop;
1665 if (env->n_nodes > maxsize)
1669 callee = entry->callee;
1671 prop = get_irg_inline_property(callee);
1672 if (prop == irg_inline_forbidden) {
1676 if (is_leave(callee) && (
1677 is_smaller(callee, leavesize) || prop >= irg_inline_forced)) {
1678 if (!phiproj_computed) {
1679 phiproj_computed = 1;
1680 collect_phiprojs(current_ir_graph);
1682 did_inline = inline_method(call, callee);
1685 inline_irg_env *callee_env = get_irg_link(callee);
1687 /* call was inlined, Phi/Projs for current graph must be recomputed */
1688 phiproj_computed = 0;
1690 /* Do some statistics */
1691 env->got_inline = 1;
1692 --env->n_call_nodes;
1693 env->n_nodes += callee_env->n_nodes;
1694 --callee_env->n_callers;
1696 /* remove this call from the list */
1697 list_del(&entry->list);
1702 ir_free_resources(current_ir_graph, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1704 } while (did_inline);
1706 /* inline other small functions. */
1707 for (i = 0; i < n_irgs; ++i) {
1709 int phiproj_computed = 0;
1711 current_ir_graph = get_irp_irg(i);
1712 env = get_irg_link(current_ir_graph);
1714 ir_reserve_resources(current_ir_graph, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1716 /* note that the list of possible calls is updated during the process */
1717 list_for_each_entry_safe(call_entry, entry, next, &env->calls, list) {
1718 irg_inline_property prop;
1723 callee = entry->callee;
1725 prop = get_irg_inline_property(callee);
1726 if (prop == irg_inline_forbidden) {
1730 e = pmap_find(copied_graphs, callee);
1733 * Remap callee if we have a copy.
1734 * FIXME: Should we do this only for recursive Calls ?
1739 if (prop >= irg_inline_forced ||
1740 (is_smaller(callee, size) && env->n_nodes < maxsize) /* small function */) {
1741 if (current_ir_graph == callee) {
1743 * Recursive call: we cannot directly inline because we cannot walk
1744 * the graph and change it. So we have to make a copy of the graph
1748 inline_irg_env *callee_env;
1751 ir_free_resources(current_ir_graph, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1754 * No copy yet, create one.
1755 * Note that recursive methods are never leaves, so it is sufficient
1756 * to test this condition here.
1758 copy = create_irg_copy(callee);
1760 /* create_irg_copy() destroys the Proj links, recompute them */
1761 phiproj_computed = 0;
1763 ir_reserve_resources(current_ir_graph, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1765 /* allocate new environment */
1766 callee_env = alloc_inline_irg_env();
1767 set_irg_link(copy, callee_env);
1769 assure_cf_loop(copy);
1770 wenv.x = callee_env;
1771 wenv.ignore_callers = 1;
1772 irg_walk_graph(copy, NULL, collect_calls2, &wenv);
1775 * Enter the entity of the original graph. This is needed
1776 * for inline_method(). However, note that ent->irg still points
1777 * to callee, NOT to copy.
1779 set_irg_entity(copy, get_irg_entity(callee));
1781 pmap_insert(copied_graphs, callee, copy);
1784 /* we have only one caller: the original graph */
1785 callee_env->n_callers = 1;
1786 callee_env->n_callers_orig = 1;
1788 if (! phiproj_computed) {
1789 phiproj_computed = 1;
1790 collect_phiprojs(current_ir_graph);
1792 did_inline = inline_method(call, callee);
1794 inline_irg_env *callee_env = (inline_irg_env *)get_irg_link(callee);
1796 /* call was inlined, Phi/Projs for current graph must be recomputed */
1797 phiproj_computed = 0;
1799 /* callee was inline. Append it's call list. */
1800 env->got_inline = 1;
1801 --env->n_call_nodes;
1802 append_call_list(env, callee_env, entry->loop_depth);
1803 --callee_env->n_callers;
1805 /* after we have inlined callee, all called methods inside callee
1806 are now called once more */
1807 list_for_each_entry(call_entry, centry, &callee_env->calls, list) {
1808 inline_irg_env *penv = get_irg_link(centry->callee);
1812 /* remove this call from the list */
1813 list_del(&entry->list);
1818 ir_free_resources(current_ir_graph, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
1821 for (i = 0; i < n_irgs; ++i) {
1822 irg = get_irp_irg(i);
1823 env = get_irg_link(irg);
1825 if (env->got_inline) {
1826 optimize_graph_df(irg);
1829 if (env->got_inline || (env->n_callers_orig != env->n_callers)) {
1830 DB((dbg, LEVEL_1, "Nodes:%3d ->%3d, calls:%3d ->%3d, callers:%3d ->%3d, -- %s\n",
1831 env->n_nodes_orig, env->n_nodes, env->n_call_nodes_orig, env->n_call_nodes,
1832 env->n_callers_orig, env->n_callers,
1833 get_entity_name(get_irg_entity(irg))));
1837 /* kill the copied graphs: we don't need them anymore */
1838 foreach_pmap(copied_graphs, pm_entry) {
1839 ir_graph *copy = pm_entry->value;
1841 /* reset the entity, otherwise it will be deleted in the next step ... */
1842 set_irg_entity(copy, NULL);
1843 free_ir_graph(copy);
1845 pmap_destroy(copied_graphs);
1847 obstack_free(&temp_obst, NULL);
1848 current_ir_graph = rem;
1851 struct inline_leave_functions_pass_t {
1852 ir_prog_pass_t pass;
1860 * Wrapper to run inline_leave_functions() as a ir_prog pass.
1862 static int inline_leave_functions_wrapper(ir_prog *irp, void *context)
1864 struct inline_leave_functions_pass_t *pass = context;
1867 inline_leave_functions(
1868 pass->maxsize, pass->leavesize,
1869 pass->size, pass->ignore_runtime);
1873 /* create a pass for inline_leave_functions() */
1874 ir_prog_pass_t *inline_leave_functions_pass(
1875 const char *name, unsigned maxsize, unsigned leavesize,
1876 unsigned size, int ignore_runtime) {
1877 struct inline_leave_functions_pass_t *pass =
1878 XMALLOCZ(struct inline_leave_functions_pass_t);
1880 pass->maxsize = maxsize;
1881 pass->leavesize = leavesize;
1883 pass->ignore_runtime = ignore_runtime;
1885 return def_prog_pass_constructor(
1887 name ? name : "inline_leave_functions",
1888 inline_leave_functions_wrapper);
1892 * Calculate the parameter weights for transmitting the address of a local variable.
1894 static unsigned calc_method_local_weight(ir_node *arg)
1897 unsigned v, weight = 0;
1899 for (i = get_irn_n_outs(arg) - 1; i >= 0; --i) {
1900 ir_node *succ = get_irn_out(arg, i);
1902 switch (get_irn_opcode(succ)) {
1905 /* Loads and Store can be removed */
1909 /* check if all args are constant */
1910 for (j = get_Sel_n_indexs(succ) - 1; j >= 0; --j) {
1911 ir_node *idx = get_Sel_index(succ, j);
1912 if (! is_Const(idx))
1915 /* Check users on this Sel. Note: if a 0 is returned here, there was
1916 some unsupported node. */
1917 v = calc_method_local_weight(succ);
1920 /* we can kill one Sel with constant indexes, this is cheap */
1924 /* when looking backward we might find Id nodes */
1925 weight += calc_method_local_weight(succ);
1928 /* unoptimized tuple */
1929 for (j = get_Tuple_n_preds(succ) - 1; j >= 0; --j) {
1930 ir_node *pred = get_Tuple_pred(succ, j);
1932 /* look for Proj(j) */
1933 for (k = get_irn_n_outs(succ) - 1; k >= 0; --k) {
1934 ir_node *succ_succ = get_irn_out(succ, k);
1935 if (is_Proj(succ_succ)) {
1936 if (get_Proj_proj(succ_succ) == j) {
1938 weight += calc_method_local_weight(succ_succ);
1941 /* this should NOT happen */
1949 /* any other node: unsupported yet or bad. */
1957 * Calculate the parameter weights for transmitting the address of a local variable.
1959 static void analyze_irg_local_weights(inline_irg_env *env, ir_graph *irg)
1961 ir_entity *ent = get_irg_entity(irg);
1963 int nparams, i, proj_nr;
1964 ir_node *irg_args, *arg;
1966 mtp = get_entity_type(ent);
1967 nparams = get_method_n_params(mtp);
1969 /* allocate a new array. currently used as 'analysed' flag */
1970 env->local_weights = NEW_ARR_D(unsigned, &temp_obst, nparams);
1972 /* If the method haven't parameters we have nothing to do. */
1976 assure_irg_outs(irg);
1977 irg_args = get_irg_args(irg);
1978 for (i = get_irn_n_outs(irg_args) - 1; i >= 0; --i) {
1979 arg = get_irn_out(irg_args, i);
1980 proj_nr = get_Proj_proj(arg);
1981 env->local_weights[proj_nr] = calc_method_local_weight(arg);
1986 * Calculate the benefice for transmitting an local variable address.
1987 * After inlining, the local variable might be transformed into a
1988 * SSA variable by scalar_replacement().
1990 static unsigned get_method_local_adress_weight(ir_graph *callee, int pos)
1992 inline_irg_env *env = get_irg_link(callee);
1994 if (env->local_weights != NULL) {
1995 if (pos < ARR_LEN(env->local_weights))
1996 return env->local_weights[pos];
2000 analyze_irg_local_weights(env, callee);
2002 if (pos < ARR_LEN(env->local_weights))
2003 return env->local_weights[pos];
2008 * Calculate a benefice value for inlining the given call.
2010 * @param call the call node we have to inspect
2011 * @param callee the called graph
2013 static int calc_inline_benefice(call_entry *entry, ir_graph *callee)
2015 ir_node *call = entry->call;
2016 ir_entity *ent = get_irg_entity(callee);
2020 int i, n_params, all_const;
2022 irg_inline_property prop;
2024 inline_irg_env *callee_env;
2026 prop = get_irg_inline_property(callee);
2027 if (prop == irg_inline_forbidden) {
2028 DB((dbg, LEVEL_2, "In %+F Call to %+F: inlining forbidden\n",
2030 return entry->benefice = INT_MIN;
2033 if (get_irg_additional_properties(callee) & mtp_property_noreturn) {
2034 DB((dbg, LEVEL_2, "In %+F Call to %+F: not inlining noreturn or weak\n",
2036 return entry->benefice = INT_MIN;
2039 /* costs for every passed parameter */
2040 n_params = get_Call_n_params(call);
2041 mtp = get_entity_type(ent);
2042 cc = get_method_calling_convention(mtp);
2043 if (cc & cc_reg_param) {
2044 /* register parameter, smaller costs for register parameters */
2045 int max_regs = cc & ~cc_bits;
2047 if (max_regs < n_params)
2048 weight += max_regs * 2 + (n_params - max_regs) * 5;
2050 weight += n_params * 2;
2052 /* parameters are passed an stack */
2053 weight += 5 * n_params;
2056 /* constant parameters improve the benefice */
2057 frame_ptr = get_irg_frame(current_ir_graph);
2059 for (i = 0; i < n_params; ++i) {
2060 ir_node *param = get_Call_param(call, i);
2062 if (is_Const(param)) {
2063 weight += get_method_param_weight(ent, i);
2066 if (is_SymConst(param))
2067 weight += get_method_param_weight(ent, i);
2068 else if (is_Sel(param) && get_Sel_ptr(param) == frame_ptr) {
2070 * An address of a local variable is transmitted. After
2071 * inlining, scalar_replacement might be able to remove the
2072 * local variable, so honor this.
2074 v = get_method_local_adress_weight(callee, i);
2077 entry->local_adr = 1;
2081 entry->all_const = all_const;
2083 callee_env = get_irg_link(callee);
2084 if (callee_env->n_callers == 1 &&
2085 callee != current_ir_graph &&
2086 !entity_is_externally_visible(ent)) {
2090 /* give a bonus for functions with one block */
2091 if (callee_env->n_blocks == 1)
2092 weight = weight * 3 / 2;
2094 /* and one for small non-recursive functions: we want them to be inlined in mostly every case */
2095 if (callee_env->n_nodes < 30 && !callee_env->recursive)
2098 /* and finally for leaves: they do not increase the register pressure
2099 because of callee safe registers */
2100 if (callee_env->n_call_nodes == 0)
2103 /** it's important to inline inner loops first */
2104 if (entry->loop_depth > 30)
2105 weight += 30 * 1024;
2107 weight += entry->loop_depth * 1024;
2110 * All arguments constant is probably a good sign, give an extra bonus
2115 return entry->benefice = weight;
2118 static ir_graph **irgs;
2119 static int last_irg;
2122 * Callgraph walker, collect all visited graphs.
2124 static void callgraph_walker(ir_graph *irg, void *data)
2127 irgs[last_irg++] = irg;
2131 * Creates an inline order for all graphs.
2133 * @return the list of graphs.
2135 static ir_graph **create_irg_list(void)
2137 ir_entity **free_methods;
2139 int n_irgs = get_irp_n_irgs();
2141 cgana(&arr_len, &free_methods);
2142 xfree(free_methods);
2144 compute_callgraph();
2147 irgs = XMALLOCNZ(ir_graph*, n_irgs);
2149 callgraph_walk(NULL, callgraph_walker, NULL);
2150 assert(n_irgs == last_irg);
2156 * Push a call onto the priority list if its benefice is big enough.
2158 * @param pqueue the priority queue of calls
2159 * @param call the call entry
2160 * @param inlien_threshold
2161 * the threshold value
2163 static void maybe_push_call(pqueue_t *pqueue, call_entry *call,
2164 int inline_threshold)
2166 ir_graph *callee = call->callee;
2167 irg_inline_property prop = get_irg_inline_property(callee);
2168 int benefice = calc_inline_benefice(call, callee);
2170 DB((dbg, LEVEL_2, "In %+F Call %+F to %+F has benefice %d\n",
2171 get_irn_irg(call->call), call->call, callee, benefice));
2173 if (prop < irg_inline_forced && benefice < inline_threshold) {
2177 pqueue_put(pqueue, call, benefice);
2181 * Try to inline calls into a graph.
2183 * @param irg the graph into which we inline
2184 * @param maxsize do NOT inline if the size of irg gets
2185 * bigger than this amount
2186 * @param inline_threshold
2187 * threshold value for inline decision
2188 * @param copied_graphs
2189 * map containing copied of recursive graphs
2191 static void inline_into(ir_graph *irg, unsigned maxsize,
2192 int inline_threshold, pmap *copied_graphs)
2194 int phiproj_computed = 0;
2195 inline_irg_env *env = get_irg_link(irg);
2196 call_entry *curr_call;
2200 if (env->n_call_nodes == 0)
2203 if (env->n_nodes > maxsize) {
2204 DB((dbg, LEVEL_2, "%+F: too big (%d)\n", irg, env->n_nodes));
2208 current_ir_graph = irg;
2209 ir_reserve_resources(irg, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
2211 /* put irgs into the pqueue */
2212 pqueue = new_pqueue();
2214 list_for_each_entry(call_entry, curr_call, &env->calls, list) {
2215 assert(is_Call(curr_call->call));
2216 maybe_push_call(pqueue, curr_call, inline_threshold);
2219 /* note that the list of possible calls is updated during the process */
2220 while (!pqueue_empty(pqueue)) {
2222 call_entry *curr_call = pqueue_pop_front(pqueue);
2223 ir_graph *callee = curr_call->callee;
2224 ir_node *call_node = curr_call->call;
2225 inline_irg_env *callee_env = get_irg_link(callee);
2226 irg_inline_property prop = get_irg_inline_property(callee);
2228 const call_entry *centry;
2231 if ((prop < irg_inline_forced) && env->n_nodes + callee_env->n_nodes > maxsize) {
2232 DB((dbg, LEVEL_2, "%+F: too big (%d) + %+F (%d)\n", irg,
2233 env->n_nodes, callee, callee_env->n_nodes));
2237 e = pmap_find(copied_graphs, callee);
2239 int benefice = curr_call->benefice;
2241 * Reduce the weight for recursive function IFF not all arguments are const.
2242 * inlining recursive functions is rarely good.
2244 if (!curr_call->all_const)
2246 if (benefice < inline_threshold)
2250 * Remap callee if we have a copy.
2253 callee_env = get_irg_link(callee);
2256 if (current_ir_graph == callee) {
2258 * Recursive call: we cannot directly inline because we cannot
2259 * walk the graph and change it. So we have to make a copy of
2262 int benefice = curr_call->benefice;
2266 * Reduce the weight for recursive function IFF not all arguments are const.
2267 * inlining recursive functions is rarely good.
2269 if (!curr_call->all_const)
2271 if (benefice < inline_threshold)
2274 ir_free_resources(irg, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
2277 * No copy yet, create one.
2278 * Note that recursive methods are never leaves, so it is
2279 * sufficient to test this condition here.
2281 copy = create_irg_copy(callee);
2283 /* create_irg_copy() destroys the Proj links, recompute them */
2284 phiproj_computed = 0;
2286 ir_reserve_resources(irg, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
2288 /* allocate a new environment */
2289 callee_env = alloc_inline_irg_env();
2290 set_irg_link(copy, callee_env);
2292 assure_cf_loop(copy);
2293 wenv.x = callee_env;
2294 wenv.ignore_callers = 1;
2295 irg_walk_graph(copy, NULL, collect_calls2, &wenv);
2298 * Enter the entity of the original graph. This is needed
2299 * for inline_method(). However, note that ent->irg still points
2300 * to callee, NOT to copy.
2302 set_irg_entity(copy, get_irg_entity(callee));
2304 pmap_insert(copied_graphs, callee, copy);
2307 /* we have only one caller: the original graph */
2308 callee_env->n_callers = 1;
2309 callee_env->n_callers_orig = 1;
2311 if (! phiproj_computed) {
2312 phiproj_computed = 1;
2313 collect_phiprojs(current_ir_graph);
2315 did_inline = inline_method(call_node, callee);
2319 /* call was inlined, Phi/Projs for current graph must be recomputed */
2320 phiproj_computed = 0;
2322 /* remove it from the caller list */
2323 list_del(&curr_call->list);
2325 /* callee was inline. Append it's call list. */
2326 env->got_inline = 1;
2327 --env->n_call_nodes;
2329 /* we just generate a bunch of new calls */
2330 loop_depth = curr_call->loop_depth;
2331 list_for_each_entry(call_entry, centry, &callee_env->calls, list) {
2332 inline_irg_env *penv = get_irg_link(centry->callee);
2334 call_entry *new_entry;
2336 /* after we have inlined callee, all called methods inside
2337 * callee are now called once more */
2340 /* Note that the src list points to Call nodes in the inlined graph,
2341 * but we need Call nodes in our graph. Luckily the inliner leaves
2342 * this information in the link field. */
2343 new_call = get_irn_link(centry->call);
2344 assert(is_Call(new_call));
2346 new_entry = duplicate_call_entry(centry, new_call, loop_depth);
2347 list_add_tail(&new_entry->list, &env->calls);
2348 maybe_push_call(pqueue, new_entry, inline_threshold);
2351 env->n_call_nodes += callee_env->n_call_nodes;
2352 env->n_nodes += callee_env->n_nodes;
2353 --callee_env->n_callers;
2355 ir_free_resources(irg, IR_RESOURCE_IRN_LINK|IR_RESOURCE_PHI_LIST);
2360 * Heuristic inliner. Calculates a benefice value for every call and inlines
2361 * those calls with a value higher than the threshold.
2363 void inline_functions(unsigned maxsize, int inline_threshold,
2364 opt_ptr after_inline_opt)
2366 inline_irg_env *env;
2370 pmap *copied_graphs;
2371 pmap_entry *pm_entry;
2374 rem = current_ir_graph;
2375 obstack_init(&temp_obst);
2377 irgs = create_irg_list();
2379 /* a map for the copied graphs, used to inline recursive calls */
2380 copied_graphs = pmap_create();
2382 /* extend all irgs by a temporary data structure for inlining. */
2383 n_irgs = get_irp_n_irgs();
2384 for (i = 0; i < n_irgs; ++i)
2385 set_irg_link(irgs[i], alloc_inline_irg_env());
2387 /* Pre-compute information in temporary data structure. */
2388 wenv.ignore_runtime = 0;
2389 wenv.ignore_callers = 0;
2390 for (i = 0; i < n_irgs; ++i) {
2391 ir_graph *irg = irgs[i];
2393 free_callee_info(irg);
2395 wenv.x = get_irg_link(irg);
2396 assure_cf_loop(irg);
2397 irg_walk_graph(irg, NULL, collect_calls2, &wenv);
2400 /* -- and now inline. -- */
2401 for (i = 0; i < n_irgs; ++i) {
2402 ir_graph *irg = irgs[i];
2404 inline_into(irg, maxsize, inline_threshold, copied_graphs);
2407 for (i = 0; i < n_irgs; ++i) {
2408 ir_graph *irg = irgs[i];
2410 env = get_irg_link(irg);
2411 if (env->got_inline && after_inline_opt != NULL) {
2412 /* this irg got calls inlined: optimize it */
2413 after_inline_opt(irg);
2415 if (env->got_inline || (env->n_callers_orig != env->n_callers)) {
2416 DB((dbg, LEVEL_1, "Nodes:%3d ->%3d, calls:%3d ->%3d, callers:%3d ->%3d, -- %s\n",
2417 env->n_nodes_orig, env->n_nodes, env->n_call_nodes_orig, env->n_call_nodes,
2418 env->n_callers_orig, env->n_callers,
2419 get_entity_name(get_irg_entity(irg))));
2423 /* kill the copied graphs: we don't need them anymore */
2424 foreach_pmap(copied_graphs, pm_entry) {
2425 ir_graph *copy = pm_entry->value;
2427 /* reset the entity, otherwise it will be deleted in the next step ... */
2428 set_irg_entity(copy, NULL);
2429 free_ir_graph(copy);
2431 pmap_destroy(copied_graphs);
2435 obstack_free(&temp_obst, NULL);
2436 current_ir_graph = rem;
2439 struct inline_functions_pass_t {
2440 ir_prog_pass_t pass;
2442 int inline_threshold;
2443 opt_ptr after_inline_opt;
2447 * Wrapper to run inline_functions() as a ir_prog pass.
2449 static int inline_functions_wrapper(ir_prog *irp, void *context)
2451 struct inline_functions_pass_t *pass = context;
2454 inline_functions(pass->maxsize, pass->inline_threshold,
2455 pass->after_inline_opt);
2459 /* create a ir_prog pass for inline_functions */
2460 ir_prog_pass_t *inline_functions_pass(
2461 const char *name, unsigned maxsize, int inline_threshold,
2462 opt_ptr after_inline_opt) {
2463 struct inline_functions_pass_t *pass =
2464 XMALLOCZ(struct inline_functions_pass_t);
2466 pass->maxsize = maxsize;
2467 pass->inline_threshold = inline_threshold;
2468 pass->after_inline_opt = after_inline_opt;
2470 return def_prog_pass_constructor(
2471 &pass->pass, name ? name : "inline_functions",
2472 inline_functions_wrapper);
2475 void firm_init_inline(void)
2477 FIRM_DBG_REGISTER(dbg, "firm.opt.inline");