nsz Git - libfirm/blob - ir/be/bespillbelady3.c

   1 /*
   2  * Copyright (C) 1995-2008 University of Karlsruhe.  All right reserved.
   3  *
   4  * This file is part of libFirm.
   5  *
   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.
  10  *
  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.
  14  *
  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
  17  * PURPOSE.
  18  */
  19
  20 /**
  21  * @file
  22  * @brief       MIN-algorithm with some twists (aka belady spiller v3)
  23  * @author      Matthias Braun
  24  * @date        15.01.2008
  25  * @version     $Id$
  26  *
  27  * TODO:
  28  *   - handle phis correctly, decide whether we should spill them ~ok
  29  *   - merge multiple start worksets of blocks ~ok
  30  *   - how to and when to do the tentative phase...
  31  */
  32 #include "config.h"
  33
  34 #include <stdbool.h>
  35
  36 #include "debug.h"
  37 #include "list.h"
  38 #include "pdeq.h"
  39
  40 #include "irnode_t.h"
  41 #include "irprintf.h"
  42 #include "iredges_t.h"
  43 #include "irloop_t.h"
  44 #include "irgwalk.h"
  45 #include "execfreq.h"
  46
  47 #include "bemodule.h"
  48 #include "bespill.h"
  49 #include "beutil.h"
  50 #include "bespilloptions.h"
  51 #include "besched_t.h"
  52 #include "be_t.h"
  53
  54 #ifndef NDEBUG
  55 #define EXPENSIVE_CHECKS
  56 #endif
  57
  58 DEBUG_ONLY(static firm_dbg_module_t *dbg = NULL;)
  59
  60 typedef struct loop_edge_t loop_edge_t;
  61 struct loop_edge_t {
  62         ir_node     *block;
  63         int          pos;
  64         loop_edge_t *next;
  65 };
  66
  67 typedef struct loop_info_t loop_info_t;
  68 struct loop_info_t {
  69         ir_loop     *loop;
  70         loop_edge_t *entry_edges;
  71         loop_edge_t *exit_edges;
  72         size_t       max_register_pressure;
  73 };
  74
  75 typedef struct worklist_entry_t worklist_entry_t;
  76 struct worklist_entry_t {
  77         struct list_head  head;
  78         ir_node          *value;
  79         ir_node          *reload_point;
  80         ir_loop          *unused_livethrough_loop;
  81 };
  82
  83 typedef struct worklist_t worklist_t;
  84 struct worklist_t {
  85         struct list_head  live_values;
  86         size_t            n_live_values;
  87         ir_visited_t      visited;
  88 };
  89
  90 typedef struct block_info_t block_info_t;
  91 struct block_info_t {
  92         worklist_t *start_worklist;
  93         worklist_t *end_worklist;
  94 };
  95
  96 /** The register class we currently run on. */
  97 static const arch_register_class_t *cls;
  98 static struct obstack               obst;
  99 static spill_env_t                 *senv;
 100 static size_t                       n_regs;
 101 static size_t                       max_register_pressure;
 102 static bool                         tentative_mode;
 103 static bool                         should_have_reached_fixpoint;
 104 static bool                         do_push_unused_livethroughs;
 105 /** Execution frequency for the current graph. */
 106 static ir_exec_freq                *exec_freq;
 107 static ir_visited_t                 worklist_visited;
 108
 109 static worklist_t *new_worklist(void)
 110 {
 111         worklist_t *worklist = obstack_alloc(&obst, sizeof(worklist[0]));
 112         memset(worklist, 0, sizeof(worklist[0]));
 113
 114         INIT_LIST_HEAD(&worklist->live_values);
 115         worklist->n_live_values    = 0;
 116
 117         return worklist;
 118 }
 119
 120 static void mark_irn_not_visited(ir_node *node)
 121 {
 122         set_irn_visited(node, get_irg_visited(current_ir_graph) - 1);
 123 }
 124
 125 static bool worklist_contains(const ir_node *node)
 126 {
 127         return irn_visited(node);
 128 }
 129
 130 static block_info_t *get_block_info(ir_node *block)
 131 {
 132         block_info_t *info = get_irn_link(block);
 133         if (info != NULL)
 134                 return info;
 135
 136         info = obstack_alloc(&obst, sizeof(info[0]));
 137         memset(info, 0, sizeof(info[0]));
 138         set_irn_link(block, info);
 139         return info;
 140 }
 141
 142 static void deactivate_worklist(const worklist_t *worklist)
 143 {
 144         struct list_head *entry;
 145
 146         list_for_each(entry, &worklist->live_values) {
 147                 worklist_entry_t *wl_entry
 148                         = list_entry(entry, worklist_entry_t, head);
 149                 assert(worklist_contains(wl_entry->value));
 150                 mark_irn_not_visited(wl_entry->value);
 151                 set_irn_link(wl_entry->value, NULL);
 152         }
 153 }
 154
 155 static void activate_worklist(const worklist_t *worklist)
 156 {
 157         struct list_head *entry;
 158
 159         list_for_each(entry, &worklist->live_values) {
 160                 worklist_entry_t *wl_entry
 161                         = list_entry(entry, worklist_entry_t, head);
 162                 assert(!worklist_contains(wl_entry->value));
 163                 mark_irn_visited(wl_entry->value);
 164                 set_irn_link(wl_entry->value, wl_entry);
 165         }
 166 }
 167
 168 /**
 169  * duplicate a worklist and directly activates it
 170  */
 171 static void fill_and_activate_worklist(worklist_t *new_worklist,
 172                 const worklist_t *worklist, ir_node *block, ir_node *succ_block,
 173                 int succ_pos)
 174 {
 175         ir_node          *reload_point  = NULL;
 176         size_t            n_live_values = 0;
 177         struct list_head *entry;
 178
 179         if (succ_block != NULL &&
 180                         (get_Block_n_cfgpreds(succ_block) > 1
 181                          || get_irn_n_edges_kind(block, EDGE_KIND_BLOCK) > 1)) {
 182                 reload_point = be_get_end_of_block_insertion_point(block);
 183         }
 184
 185         list_for_each(entry, &worklist->live_values) {
 186                 worklist_entry_t *wl_entry  = list_entry(entry, worklist_entry_t, head);
 187                 ir_node          *value     = wl_entry->value;
 188                 worklist_entry_t *new_entry;
 189
 190                 if (new_worklist->n_live_values >= n_regs)
 191                         break;
 192
 193                 if (is_Phi(value) && get_nodes_block(value) == succ_block) {
 194                         value = get_Phi_pred(value, succ_pos);
 195
 196                         /* can happen for unknown phi preds */
 197                         if (!arch_irn_consider_in_reg_alloc(cls, value))
 198                                 continue;
 199                 }
 200
 201                 if (irn_visited_else_mark(value))
 202                         continue;
 203
 204                 new_entry = obstack_alloc(&obst, sizeof(new_entry[0]));
 205                 memset(new_entry, 0, sizeof(new_entry[0]));
 206
 207                 new_entry->value = value;
 208                 if (reload_point != NULL) {
 209                         new_entry->reload_point = reload_point;
 210                 } else {
 211                         new_entry->reload_point = wl_entry->reload_point;
 212                 }
 213
 214                 list_add_tail(&new_entry->head, &new_worklist->live_values);
 215                 ++n_live_values;
 216
 217                 set_irn_link(value, new_entry);
 218                 new_worklist->n_live_values++;
 219         }
 220 }
 221
 222 static worklist_t *duplicate_worklist(const worklist_t *worklist)
 223 {
 224         worklist_t       *new_worklist;
 225         struct list_head *entry;
 226
 227         new_worklist = obstack_alloc(&obst, sizeof(new_worklist[0]));
 228         memset(new_worklist, 0, sizeof(new_worklist[0]));
 229         INIT_LIST_HEAD(&new_worklist->live_values);
 230         new_worklist->n_live_values = worklist->n_live_values;
 231
 232         list_for_each(entry, &worklist->live_values) {
 233                 worklist_entry_t *wl_entry  = list_entry(entry, worklist_entry_t, head);
 234                 worklist_entry_t *new_entry
 235                         = obstack_alloc(&obst, sizeof(new_entry[0]));
 236
 237                 memcpy(new_entry, wl_entry, sizeof(new_entry[0]));
 238                 list_add_tail(&new_entry->head, &new_worklist->live_values);
 239         }
 240
 241         return new_worklist;
 242 }
 243
 244 #ifdef DEBUG_libfirm
 245 static void print_worklist(const worklist_t *worklist, int level)
 246 {
 247         struct list_head *entry;
 248
 249         DB((dbg, level, "%d values: ", worklist->n_live_values));
 250         list_for_each(entry, &worklist->live_values) {
 251                 worklist_entry_t *wl_entry
 252                         = list_entry(entry, worklist_entry_t, head);
 253
 254                 DB((dbg, level, "%+F ", wl_entry->value));
 255         }
 256 }
 257 #endif
 258
 259
 260 /**
 261  * Clear the link fields of a loop and all
 262  * its child loops.
 263  *
 264  * @param loop  the root loop
 265  */
 266 static void clear_loop_info(ir_loop *loop)
 267 {
 268         int n_elements = get_loop_n_elements(loop);
 269         int i;
 270
 271         loop->link = NULL;
 272         for (i = 0; i < n_elements; ++i) {
 273                 loop_element element = get_loop_element(loop, i);
 274                 if (*element.kind != k_ir_loop)
 275                         continue;
 276
 277                 clear_loop_info(element.son);
 278         }
 279 }
 280
 281 static loop_info_t *get_loop_info(ir_loop *loop)
 282 {
 283         loop_info_t *info = loop->link;
 284         if (info != NULL)
 285                 return info;
 286
 287         info = obstack_alloc(&obst, sizeof(info[0]));
 288         memset(info, 0, sizeof(info[0]));
 289         info->loop = loop;
 290         loop->link = info;
 291         return info;
 292 }
 293
 294 /**
 295  * constructs loop in+out edges when called in a block walker
 296  */
 297 static void construct_loop_edges(ir_node *block, void *data)
 298 {
 299         int      n_cfgpreds = get_Block_n_cfgpreds(block);
 300         ir_loop *loop       = get_irn_loop(block);
 301         int      i;
 302
 303         (void) data;
 304
 305         for (i = 0; i < n_cfgpreds; ++i) {
 306                 ir_node     *cfgpred_block = get_Block_cfgpred_block(block, i);
 307                 ir_loop     *cfgpred_loop  = get_irn_loop(cfgpred_block);
 308                 loop_edge_t *edge;
 309                 bool        is_exit_edge;
 310                 ir_loop     *l, *goal;
 311
 312                 if (cfgpred_loop == loop)
 313                         continue;
 314
 315                 /* critical edges are splitted, so we can't jump from 1 loop
 316                  * directly into another without going out of it */
 317                 assert(get_loop_depth(cfgpred_loop) != get_loop_depth(loop));
 318
 319                 /* edge out of loop */
 320                 if (get_loop_depth(cfgpred_loop) > get_loop_depth(loop)) {
 321                         is_exit_edge = true;
 322                         l            = cfgpred_loop;
 323                         goal         = loop;
 324                 } else {
 325                         is_exit_edge = false;
 326                         l            = loop;
 327                         goal         = cfgpred_loop;
 328                 }
 329
 330                 /* this might be a jump out/in multiple loops */
 331                 do {
 332                         loop_info_t *l_info = get_loop_info(l);
 333
 334                         edge = obstack_alloc(&obst, sizeof(edge[0]));
 335                         memset(edge, 0, sizeof(edge[0]));
 336                         edge->block = block;
 337                         edge->pos   = i;
 338
 339                         if (is_exit_edge) {
 340                                 edge->next         = l_info->exit_edges;
 341                                 l_info->exit_edges = edge;
 342                                 assert(get_loop_depth(loop) < get_loop_depth(l));
 343                         } else {
 344                                 edge->next          = l_info->entry_edges;
 345                                 l_info->entry_edges = edge;
 346                                 assert(get_loop_depth(cfgpred_loop) < get_loop_depth(l));
 347                         }
 348                         l = get_loop_outer_loop(l);
 349                 } while (l != goal);
 350         }
 351 }
 352
 353
 354
 355
 356 static void place_reload(worklist_entry_t *entry)
 357 {
 358         if (tentative_mode)
 359                 return;
 360
 361         DB((dbg, LEVEL_1, "reload of %+F before %+F\n", entry->value,
 362                 entry->reload_point));
 363         assert(entry->reload_point != NULL);
 364         be_add_reload(senv, entry->value, entry->reload_point, cls, 1);
 365         entry->reload_point = NULL;
 366 }
 367
 368 /**
 369  * makes sure the worklist contains not more than n_regs - room_needed entries
 370  */
 371 static void make_room(worklist_t *worklist, size_t room_needed)
 372 {
 373         int               i;
 374         int               spills_needed;
 375         struct list_head *entry;
 376
 377         spills_needed = worklist->n_live_values + room_needed - n_regs;
 378         if (spills_needed <= 0)
 379                 return;
 380
 381         entry = worklist->live_values.next;
 382         for(i = spills_needed; i > 0; --i) {
 383                 struct list_head *next = entry->next;
 384                 worklist_entry_t *wl_entry
 385                         = list_entry(entry, worklist_entry_t, head);
 386
 387                 assert(worklist_contains(wl_entry->value));
 388                 mark_irn_not_visited(wl_entry->value);
 389                 place_reload(wl_entry);
 390                 list_del(entry);
 391
 392                 entry = next;
 393         }
 394         worklist->n_live_values -= spills_needed;
 395 }
 396
 397 /**
 398  * a value was used, so bring it to the back of the worklist (which might
 399  * result in a spill of another value).
 400  */
 401 static void val_used(worklist_t *worklist, ir_node *value, ir_node *sched_point)
 402 {
 403         /* already in the worklist? move around, otherwise add at back */
 404         worklist_entry_t *entry = get_irn_link(value);
 405
 406         assert(arch_irn_consider_in_reg_alloc(cls, value));
 407
 408         if (worklist_contains(value)) {
 409                 assert(entry != NULL);
 410
 411                 list_del(&entry->head);
 412         } else {
 413                 if (entry == NULL) {
 414                         entry        = obstack_alloc(&obst, sizeof(entry[0]));
 415                         memset(entry, 0, sizeof(entry[0]));
 416
 417                         entry->value = value;
 418                         set_irn_link(value, entry);
 419                 }
 420
 421                 ++worklist->n_live_values;
 422                 mark_irn_visited(value);
 423         }
 424
 425         entry->reload_point = sched_point;
 426         list_add_tail(&entry->head, &worklist->live_values);
 427 }
 428
 429 static void worklist_remove(worklist_t *worklist, ir_node *value)
 430 {
 431         worklist_entry_t *entry = get_irn_link(value);
 432         assert(entry != NULL);
 433         list_del(&entry->head);
 434         --worklist->n_live_values;
 435
 436         assert(worklist_contains(value));
 437         mark_irn_not_visited(value);
 438 }
 439
 440 static void update_max_pressure(worklist_t *worklist)
 441 {
 442         if (worklist->n_live_values > max_register_pressure)
 443                 max_register_pressure = worklist->n_live_values;
 444 }
 445
 446 static void do_spilling(ir_node *block, worklist_t *worklist)
 447 {
 448         ir_node *node;
 449
 450         assert(worklist != NULL);
 451
 452         sched_foreach_reverse(block, node) {
 453                 int    i, arity;
 454                 size_t n_defs = 0;
 455
 456                 DB((dbg, LEVEL_2, "\t%+F... ", node));
 457                 update_max_pressure(worklist);
 458
 459                 if (is_Phi(node)) {
 460                         ir_node *node2;
 461                         /* TODO: if we have some free registers, then we could decide to
 462                          * not spill some phis (but not for phis where at least 1 input is
 463                          * themselfes) */
 464
 465                         /* we have to spill all phis that are not live */
 466                         sched_foreach_reverse_from(node, node2) {
 467                                 assert(is_Phi(node2));
 468
 469                                 if (worklist_contains(node2))
 470                                         continue;
 471                                 if (!arch_irn_consider_in_reg_alloc(cls, node2))
 472                                         continue;
 473
 474                                 if (!tentative_mode)
 475                                         be_spill_phi(senv, node2);
 476                         }
 477                         DB((dbg, LEVEL_2, "\n"));
 478                         break;
 479                 }
 480
 481                 /* remove values defined by this instruction from the workset. Values
 482                  * defined but not in the workset need free registers */
 483                 if (get_irn_mode(node) == mode_T) {
 484                         const ir_edge_t *edge;
 485
 486                         foreach_out_edge(node, edge) {
 487                                 ir_node *proj = get_edge_src_irn(edge);
 488                                 if (!arch_irn_consider_in_reg_alloc(cls, proj))
 489                                         continue;
 490                                 if (worklist_contains(proj)) {
 491                                         worklist_remove(worklist, proj);
 492                                 } else {
 493                                         ++n_defs;
 494                                 }
 495                         }
 496                 } else if (arch_irn_consider_in_reg_alloc(cls, node)) {
 497                         if (worklist_contains(node)) {
 498                                 worklist_remove(worklist, node);
 499                         } else {
 500                                 n_defs = 1;
 501                         }
 502                 }
 503
 504                 /* make sure we have enough free registers for the spills */
 505                 make_room(worklist, n_defs);
 506
 507                 /* put all values used by the instruction into the workset */
 508                 arity = get_irn_arity(node);
 509                 for(i = 0; i < arity; ++i) {
 510                         ir_node *use = get_irn_n(node, i);
 511
 512                         if (!arch_irn_consider_in_reg_alloc(cls, use))
 513                                 continue;
 514
 515                         val_used(worklist, use, node);
 516                 }
 517
 518                 /* we might have too many values in the worklist now and need to spill
 519                  * some */
 520                 make_room(worklist, 0);
 521
 522 #ifdef DEBUG_libfirm
 523                 print_worklist(worklist, LEVEL_2);
 524                 DB((dbg, LEVEL_2, "\n"));
 525 #endif
 526         }
 527
 528         update_max_pressure(worklist);
 529
 530 #ifdef DEBUG_libfirm
 531         DB((dbg, LEVEL_1, "worklist at begin of %+F:", block));
 532         print_worklist(worklist, LEVEL_1);
 533         DB((dbg, LEVEL_1, "\n"));
 534 #endif
 535 }
 536
 537 static bool worklists_equal(const worklist_t *wl1, const worklist_t *wl2)
 538 {
 539         const struct list_head *i1 = &wl1->live_values;
 540         const struct list_head *i2 = &wl2->live_values;
 541
 542         for ( ; i1 != &wl1->live_values && i2 != &wl2->live_values;
 543                         i1 = i1->next, i2 = i2->next) {
 544                 worklist_entry_t *entry1 = list_entry(i1, worklist_entry_t, head);
 545                 worklist_entry_t *entry2 = list_entry(i2, worklist_entry_t, head);
 546
 547                 if (entry1->value != entry2->value)
 548                         return false;
 549         }
 550         /* both lists at the end */
 551         if (i1 != &wl1->live_values || i2 != &wl2->live_values)
 552                 return false;
 553
 554         return true;
 555 }
 556
 557 static bool fill_start_worklist(worklist_t *new_worklist, ir_node *block)
 558 {
 559         double           best_execfreq   = -1;
 560         worklist_t      *best_worklist   = NULL;
 561         ir_node         *best_succ_block = NULL;
 562         int              best_pos;
 563         const ir_edge_t *edge;
 564
 565         /* construct worklist */
 566         foreach_block_succ(block, edge) {
 567                 ir_node      *succ_block = get_edge_src_irn(edge);
 568                 double       execfreq    = get_block_execfreq(exec_freq, succ_block);
 569                 block_info_t *block_info;
 570                 worklist_t   *succ_worklist;
 571
 572                 if (execfreq < best_execfreq)
 573                         continue;
 574
 575                 block_info    = get_block_info(succ_block);
 576                 succ_worklist = block_info->start_worklist;
 577
 578                 if (succ_worklist == NULL || succ_worklist->visited >= worklist_visited)
 579                         continue;
 580
 581                 best_execfreq   = execfreq;
 582                 best_worklist   = succ_worklist;
 583                 best_succ_block = succ_block;
 584                 best_pos        = get_edge_src_pos(edge);
 585         }
 586
 587         if (best_worklist == NULL)
 588                 return false;
 589         best_worklist->visited = worklist_visited;
 590
 591         fill_and_activate_worklist(new_worklist, best_worklist, block,
 592                         best_succ_block, best_pos);
 593         return true;
 594 }
 595
 596 static worklist_t *construct_start_worklist(ir_node *block)
 597 {
 598         worklist_t *worklist = new_worklist();
 599
 600         ++worklist_visited;
 601
 602         while(fill_start_worklist(worklist, block)) {
 603                 if (worklist->n_live_values >= n_regs)
 604                         break;
 605         }
 606
 607         return worklist;
 608 }
 609
 610 static void process_block(ir_node *block, void *env)
 611 {
 612         block_info_t    *block_info;
 613         worklist_t      *worklist;
 614         int              n_preds;
 615
 616         (void) env;
 617         DB((dbg, LEVEL_1, "Processing %+F\n", block));
 618
 619         worklist   = construct_start_worklist(block);
 620         block_info = get_block_info(block);
 621
 622 #ifdef DEBUG_libfirm
 623         DB((dbg, LEVEL_1, "worklist at end of %+F:", block));
 624         print_worklist(worklist, LEVEL_1);
 625         DB((dbg, LEVEL_1, "\n"));
 626
 627         if (should_have_reached_fixpoint) {
 628                 assert(worklists_equal(block_info->end_worklist, worklist));
 629         }
 630 #endif
 631         block_info->end_worklist = duplicate_worklist(worklist);
 632
 633         do_spilling(block, worklist);
 634         deactivate_worklist(worklist);
 635
 636 #ifdef DEBUG_libfirm
 637         if (should_have_reached_fixpoint) {
 638                 assert(worklists_equal(block_info->start_worklist, worklist));
 639         }
 640 #endif
 641         block_info->start_worklist = worklist;
 642
 643         /* we shouldn't have any live values left at the start block */
 644         n_preds = get_Block_n_cfgpreds(block);
 645         //assert(n_preds != 0 || worklist->n_live_values == 0);
 646 }
 647
 648 typedef union block_or_loop_t block_or_loop_t;
 649 union block_or_loop_t {
 650         ir_node *block;
 651         ir_loop *loop;
 652 };
 653
 654 static block_or_loop_t *loop_blocks;
 655 static ir_loop         *current_loop;
 656
 657 static void find_blocks(ir_node *block);
 658
 659 static void find_in_loop(ir_loop *loop, ir_node *entry)
 660 {
 661         loop_info_t     *loop_info = get_loop_info(loop);
 662         block_or_loop_t block_or_loop;
 663         loop_edge_t     *edge;
 664
 665         /* simply mark 1 block in the loop to indicate that the loop was already
 666          * processed */
 667         ir_node *some_block = loop_info->entry_edges->block;
 668         if (Block_block_visited(some_block))
 669                 return;
 670
 671         block_or_loop.loop = loop;
 672         ARR_APP1(block_or_loop_t, loop_blocks, block_or_loop);
 673
 674 #ifndef NDEBUG
 675         {
 676                 /* we should be 1 of the entry blocks */
 677                 loop_edge_t *edge  = loop_info->entry_edges;
 678                 bool         found = false;
 679                 for ( ; edge != NULL; edge = edge->next) {
 680                         if (edge->block == entry) {
 681                                 found = true;
 682                                 break;
 683                         }
 684                 }
 685                 assert(found);
 686         }
 687 #else
 688         (void) entry;
 689 #endif
 690
 691         /* check all loop successors */
 692         for (edge = loop_info->exit_edges; edge != NULL; edge = edge->next) {
 693                 ir_node *succ      = edge->block;
 694                 ir_loop *succ_loop = get_irn_loop(succ);
 695
 696                 if (succ_loop == current_loop) {
 697                         find_blocks(succ);
 698                 } else {
 699                         assert(get_loop_depth(succ_loop) < get_loop_depth(current_loop));
 700                 }
 701         }
 702 }
 703
 704 static void find_blocks(ir_node *block)
 705 {
 706         const ir_edge_t *edge;
 707         block_or_loop_t block_or_loop;
 708
 709         if (Block_block_visited(block))
 710                 return;
 711
 712         block_or_loop.block = block;
 713
 714         ARR_APP1(block_or_loop_t, loop_blocks, block_or_loop);
 715         mark_Block_block_visited(block);
 716
 717         foreach_block_succ(block, edge) {
 718                 ir_node *succ = get_edge_src_irn(edge);
 719
 720                 /* is it part of the current loop? */
 721                 ir_loop *loop = get_irn_loop(succ);
 722                 if (loop != current_loop) {
 723                         /* a sub-loop? */
 724                         if (get_loop_depth(loop) > get_loop_depth(current_loop)) {
 725                                 find_in_loop(loop, succ);
 726                         } else {
 727                                 /* parent loop: we're not interested in the block */
 728                         }
 729                 } else {
 730                         find_blocks(succ);
 731                 }
 732         }
 733 }
 734
 735 /**
 736  * append an entry to a worklist. WARNING: The entry must not already be in the
 737  * worklist.
 738  */
 739 static void worklist_append(worklist_t *worklist, ir_node *value,
 740                             ir_node *reload_point,
 741                                                         ir_loop *unused_livethrough_loop)
 742 {
 743         worklist_entry_t *entry     = obstack_alloc(&obst, sizeof(entry[0]));
 744         memset(entry, 0, sizeof(entry[0]));
 745
 746 #ifdef EXPENSIVE_CHECKS
 747         {
 748                 struct list_head *entry;
 749                 list_for_each(entry, &worklist->live_values) {
 750                         worklist_entry_t *wl_entry
 751                                 = list_entry(entry, worklist_entry_t, head);
 752                         assert(wl_entry->value != value);
 753                 }
 754         }
 755 #endif
 756
 757         entry->value                   = value;
 758         entry->reload_point            = reload_point;
 759         entry->unused_livethrough_loop = unused_livethrough_loop;
 760         list_add_tail(&entry->head, &worklist->live_values);
 761         ++worklist->n_live_values;
 762         assert(worklist->n_live_values <= n_regs);
 763 }
 764
 765 static void push_unused_livethrough(loop_info_t *loop_info, ir_node *value)
 766 {
 767         loop_edge_t *edge;
 768         ++worklist_visited;
 769
 770         /* add the value to all loop exit and entry blocks */
 771         for (edge = loop_info->exit_edges; edge != NULL; edge = edge->next) {
 772                 ir_node            *block
 773                         = get_Block_cfgpred_block(edge->block, edge->pos);
 774                 const block_info_t *info     = get_block_info(block);
 775                 worklist_t         *worklist = info->end_worklist;
 776                 ir_node            *reload_point = NULL;
 777
 778                 if (worklist->visited >= worklist_visited)
 779                         continue;
 780                 worklist->visited = worklist_visited;
 781
 782                 /* TODO: we need a smarter mechanism here, that makes the reloader place
 783                  * reload nodes on all loop exits... */
 784
 785                 worklist_append(worklist, value, reload_point, loop_info->loop);
 786         }
 787         edge = loop_info->entry_edges;
 788         for ( ; edge != NULL; edge = edge->next) {
 789                 ir_node            *entry_block = edge->block;
 790                 const block_info_t *info        = get_block_info(entry_block);
 791                 worklist_t         *worklist    = info->start_worklist;
 792                 ir_node            *pred_block;
 793                 ir_node            *reload_point;
 794
 795                 if (worklist->visited >= worklist_visited)
 796                         continue;
 797                 worklist->visited = worklist_visited;
 798
 799                 pred_block   = get_Block_cfgpred_block(entry_block, edge->pos);
 800                 reload_point = be_get_end_of_block_insertion_point(pred_block);
 801
 802                 worklist_append(worklist, value, reload_point, loop_info->loop);
 803         }
 804
 805         set_irn_link(value, NULL);
 806         ++loop_info->max_register_pressure;
 807 }
 808
 809 static void push_unused_livethroughs(loop_info_t *loop_info)
 810 {
 811         loop_edge_t *edge;
 812
 813         /* we can only push unused livethroughs if register pressure inside the loop
 814          * was low enough */
 815         if (loop_info->max_register_pressure >= n_regs)
 816                 return;
 817
 818         /* find unused livethroughs: register pressure in the loop was low enough
 819          * which means that we had no spills which implies that at every point in
 820          * the loop all*/
 821         for (edge = loop_info->exit_edges; edge != NULL; edge = edge->next) {
 822                 ir_node            *block          = edge->block;
 823                 const block_info_t *info           = get_block_info(block);
 824                 worklist_t         *start_worklist = info->start_worklist;
 825                 ir_node            *exit_block;
 826                 const block_info_t *exit_info;
 827                 worklist_t         *end_worklist;
 828                 struct list_head   *entry;
 829
 830                 if (start_worklist == NULL)
 831                         continue;
 832
 833                 exit_block   = get_Block_cfgpred_block(edge->block, edge->pos);
 834                 exit_info    = get_block_info(exit_block);
 835                 end_worklist = exit_info->end_worklist;
 836
 837                 activate_worklist(end_worklist);
 838                 /* all values contained in the start_worklist, which are not available
 839                  * in the end_worklist, must be unused livethroughs */
 840
 841                 list_for_each(entry, &start_worklist->live_values) {
 842                         worklist_entry_t *wl_entry
 843                                 = list_entry(entry, worklist_entry_t, head);
 844                         ir_node *value = wl_entry->value;
 845
 846                         if (loop_info->max_register_pressure >= n_regs)
 847                                 break;
 848
 849
 850                         if (!worklist_contains(value)) {
 851                                 /* add it to all loop exits */
 852                                 DB((dbg, LEVEL_2, "Found unused livethrough: %+F (regpressure in loop %d)\n", value, loop_info->max_register_pressure));
 853                                 push_unused_livethrough(loop_info, value);
 854                                 /* the value should now be part of end_worklist, so mark it */
 855                                 mark_irn_visited(value);
 856                         }
 857                 }
 858
 859                 deactivate_worklist(end_worklist);
 860         }
 861 }
 862
 863 static void process_block_or_loop(const block_or_loop_t block_or_loop)
 864 {
 865         if (block_or_loop.loop->kind == k_ir_loop) {
 866                 loop_info_t *loop_info = get_loop_info(block_or_loop.loop);
 867
 868                 if (do_push_unused_livethroughs)
 869                         push_unused_livethroughs(loop_info);
 870
 871                 if (loop_info->max_register_pressure > max_register_pressure)
 872                         max_register_pressure = loop_info->max_register_pressure;
 873
 874                 return;
 875         }
 876         process_block(block_or_loop.block, NULL);
 877 }
 878
 879 static void process_loop(ir_loop *loop)
 880 {
 881         int         n_elements = get_loop_n_elements(loop);
 882         int         i, len;
 883         loop_info_t *loop_info;
 884         loop_edge_t *edge;
 885         ir_node     *some_block;
 886
 887         /* first handle all sub-loops */
 888         for (i = 0; i < n_elements; ++i) {
 889                 loop_element element = get_loop_element(loop, i);
 890                 if (*element.kind != k_ir_loop)
 891                         continue;
 892
 893                 process_loop(element.son);
 894         }
 895
 896         /* create a postorder of the blocks */
 897         loop_info = get_loop_info(loop);
 898         edge      = loop_info->entry_edges;
 899         if (edge != NULL) {
 900                 some_block = edge->block;
 901         } else {
 902                 assert(loop == get_irg_loop(current_ir_graph));
 903                 some_block = get_irg_start_block(current_ir_graph);
 904         }
 905
 906         loop_blocks  = NEW_ARR_F(block_or_loop_t,0);
 907         current_loop = loop;
 908
 909         ir_reserve_resources(current_ir_graph, IR_RESOURCE_BLOCK_VISITED);
 910         inc_irg_block_visited(current_ir_graph);
 911         find_blocks(some_block);
 912         /* for endless loops the end-block might be unreachable */
 913         if (loop == get_irg_loop(current_ir_graph)) {
 914                 find_blocks(get_irg_end_block(current_ir_graph));
 915         }
 916         ir_free_resources(current_ir_graph, IR_RESOURCE_BLOCK_VISITED);
 917
 918         DB((dbg, LEVEL_3, "Block List for loop %p\n", loop));
 919         len = ARR_LEN(loop_blocks);
 920         for (i = 0; i < len; ++i) {
 921                 block_or_loop_t block_or_loop = loop_blocks[i];
 922                 if (block_or_loop.loop->kind == k_ir_loop) {
 923                         DB((dbg, LEVEL_3, " L-%p", block_or_loop.loop));
 924                 } else {
 925                         DB((dbg, LEVEL_3, " B-%+F", block_or_loop.block));
 926                 }
 927         }
 928         DB((dbg, LEVEL_3, "\n"));
 929
 930         max_register_pressure = 0;
 931
 932         /* run1: tentative phase */
 933         tentative_mode = true;
 934         for (i = len-1; i >= 0; --i) {
 935                 process_block_or_loop(loop_blocks[i]);
 936         }
 937
 938         /* run2: tentative phase - should reach fixpoint */
 939         tentative_mode = true;
 940         for (i = len-1; i >= 0; --i) {
 941                 process_block_or_loop(loop_blocks[i]);
 942         }
 943
 944 #ifndef NDEBUG
 945         /* run3: tentative phase - check fixpoint */
 946         tentative_mode               = true;
 947         should_have_reached_fixpoint = true;
 948         for (i = len-1; i >= 0; --i) {
 949                 process_block_or_loop(loop_blocks[i]);
 950         }
 951         should_have_reached_fixpoint = false;
 952 #endif
 953
 954         /* run4: add spills/reloads */
 955         tentative_mode              = false;
 956         do_push_unused_livethroughs = true;
 957         for (i = len-1; i >= 0; --i) {
 958                 process_block_or_loop(loop_blocks[i]);
 959         }
 960         do_push_unused_livethroughs = false;
 961
 962         loop_info->max_register_pressure = max_register_pressure;
 963         DB((dbg, LEVEL_2, "Regpressure in loop %p: %u\n", loop,
 964                                 (unsigned) max_register_pressure));
 965
 966         DEL_ARR_F(loop_blocks);
 967 }
 968
 969 static void fix_block_borders(ir_node *block, void *data)
 970 {
 971         block_info_t *block_info     = get_block_info(block);
 972         worklist_t   *start_worklist = block_info->start_worklist;
 973         int           n_cfgpreds     = get_Block_n_cfgpreds(block);
 974         ir_loop      *loop           = get_irn_loop(block);
 975         int           i;
 976
 977         (void) data;
 978         assert(start_worklist != NULL);
 979
 980         for (i = 0; i < n_cfgpreds; ++i) {
 981                 ir_node      *pred_block      = get_Block_cfgpred_block(block, i);
 982                 block_info_t *pred_block_info = get_block_info(pred_block);
 983                 worklist_t   *end_worklist    = pred_block_info->end_worklist;
 984                 ir_loop      *pred_loop       = get_irn_loop(pred_block);
 985                 bool          is_loop_entry   = false;
 986                 struct list_head *entry;
 987
 988                 assert(end_worklist != NULL);
 989
 990                 if (get_loop_depth(pred_loop) < get_loop_depth(loop)) {
 991                         is_loop_entry = true;
 992                 }
 993
 994                 /* reload missing values */
 995                 activate_worklist(end_worklist);
 996
 997                 list_for_each(entry, &start_worklist->live_values) {
 998                         worklist_entry_t *wl_entry
 999                                 = list_entry(entry, worklist_entry_t, head);
1000                         ir_node          *value = wl_entry->value;
1001
1002                         if (is_Phi(value) && get_nodes_block(value) == block) {
1003                                 value = get_irn_n(value, i);
1004
1005                                 /* we might have unknowns as argument for the phi */
1006                                 if (!arch_irn_consider_in_reg_alloc(cls, value))
1007                                         continue;
1008                         }
1009
1010                         if (worklist_contains(value))
1011                                 continue;
1012                         if (wl_entry->unused_livethrough_loop != NULL && !is_loop_entry)
1013                                 continue;
1014
1015                         be_add_reload_on_edge(senv, value, block, i, cls, 1);
1016                 }
1017
1018                 deactivate_worklist(end_worklist);
1019         }
1020 }
1021
1022 /**
1023  * Run the spill algorithm.
1024  *
1025  * @param birg  the graph to run on
1026  * @param ncls  the register class to run on
1027  */
1028 static void be_spill_belady3(be_irg_t *birg, const arch_register_class_t *ncls)
1029 {
1030         ir_graph *irg = be_get_birg_irg(birg);
1031
1032         cls    = ncls;
1033         n_regs = cls->n_regs - be_put_ignore_regs(birg, cls, NULL);
1034
1035         /* shortcut for register classes with ignore regs only */
1036         if (n_regs == 0)
1037                 return;
1038
1039         worklist_visited = 0;
1040         exec_freq        = be_get_birg_exec_freq(birg);
1041
1042         be_clear_links(irg);
1043         ir_reserve_resources(irg, IR_RESOURCE_IRN_VISITED | IR_RESOURCE_IRN_LINK
1044                         | IR_RESOURCE_LOOP_LINK);
1045         inc_irg_visited(irg);
1046
1047         obstack_init(&obst);
1048         senv = be_new_spill_env(birg);
1049
1050         assure_cf_loop(irg);
1051         clear_loop_info(get_irg_loop(irg));
1052         irg_block_walk_graph(irg, construct_loop_edges, NULL, NULL);
1053
1054         process_loop(get_irg_loop(irg));
1055
1056         irg_block_walk_graph(irg, fix_block_borders, NULL, NULL);
1057
1058         ir_free_resources(irg, IR_RESOURCE_IRN_VISITED | IR_RESOURCE_IRN_LINK
1059                         | IR_RESOURCE_LOOP_LINK);
1060
1061         be_insert_spills_reloads(senv);
1062
1063         obstack_free(&obst, NULL);
1064
1065         /* clean up */
1066         be_delete_spill_env(senv);
1067 }
1068
1069 void be_init_spillbelady3(void)
1070 {
1071         static be_spiller_t belady3_spiller = {
1072                 be_spill_belady3
1073         };
1074
1075         be_register_spiller("belady3", &belady3_spiller);
1076         FIRM_DBG_REGISTER(dbg, "firm.be.spill.belady3");
1077 }
1078
1079 BE_REGISTER_MODULE_CONSTRUCTOR(be_init_spillbelady3);