aboutsummaryrefslogtreecommitdiffstats
path: root/gcc-4.9/gcc/tree-loop-distribution.c
blob: 9db92dbf9bfd72c2ad9cec4757c5e32ec224ae77 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
/* Loop distribution.
   Copyright (C) 2006-2014 Free Software Foundation, Inc.
   Contributed by Georges-Andre Silber <Georges-Andre.Silber@ensmp.fr>
   and Sebastian Pop <sebastian.pop@amd.com>.

This file is part of GCC.

GCC is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 3, or (at your option) any
later version.

GCC is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
for more details.

You should have received a copy of the GNU General Public License
along with GCC; see the file COPYING3.  If not see
<http://www.gnu.org/licenses/>.  */

/* This pass performs loop distribution: for example, the loop

   |DO I = 2, N
   |    A(I) = B(I) + C
   |    D(I) = A(I-1)*E
   |ENDDO

   is transformed to

   |DOALL I = 2, N
   |   A(I) = B(I) + C
   |ENDDO
   |
   |DOALL I = 2, N
   |   D(I) = A(I-1)*E
   |ENDDO

   This pass uses an RDG, Reduced Dependence Graph built on top of the
   data dependence relations.  The RDG is then topologically sorted to
   obtain a map of information producers/consumers based on which it
   generates the new loops.  */

#include "config.h"
#include "system.h"
#include "coretypes.h"
#include "tree.h"
#include "basic-block.h"
#include "tree-ssa-alias.h"
#include "internal-fn.h"
#include "gimple-expr.h"
#include "is-a.h"
#include "gimple.h"
#include "gimple-iterator.h"
#include "gimplify-me.h"
#include "stor-layout.h"
#include "gimple-ssa.h"
#include "tree-cfg.h"
#include "tree-phinodes.h"
#include "ssa-iterators.h"
#include "stringpool.h"
#include "tree-ssanames.h"
#include "tree-ssa-loop-manip.h"
#include "tree-ssa-loop.h"
#include "tree-into-ssa.h"
#include "tree-ssa.h"
#include "cfgloop.h"
#include "tree-chrec.h"
#include "tree-data-ref.h"
#include "tree-scalar-evolution.h"
#include "tree-pass.h"
#include "gimple-pretty-print.h"
#include "tree-vectorizer.h"


/* A Reduced Dependence Graph (RDG) vertex representing a statement.  */
typedef struct rdg_vertex
{
  /* The statement represented by this vertex.  */
  gimple stmt;

  /* Vector of data-references in this statement.  */
  vec<data_reference_p> datarefs;

  /* True when the statement contains a write to memory.  */
  bool has_mem_write;

  /* True when the statement contains a read from memory.  */
  bool has_mem_reads;
} *rdg_vertex_p;

#define RDGV_STMT(V)     ((struct rdg_vertex *) ((V)->data))->stmt
#define RDGV_DATAREFS(V) ((struct rdg_vertex *) ((V)->data))->datarefs
#define RDGV_HAS_MEM_WRITE(V) ((struct rdg_vertex *) ((V)->data))->has_mem_write
#define RDGV_HAS_MEM_READS(V) ((struct rdg_vertex *) ((V)->data))->has_mem_reads
#define RDG_STMT(RDG, I) RDGV_STMT (&(RDG->vertices[I]))
#define RDG_DATAREFS(RDG, I) RDGV_DATAREFS (&(RDG->vertices[I]))
#define RDG_MEM_WRITE_STMT(RDG, I) RDGV_HAS_MEM_WRITE (&(RDG->vertices[I]))
#define RDG_MEM_READS_STMT(RDG, I) RDGV_HAS_MEM_READS (&(RDG->vertices[I]))

/* Data dependence type.  */

enum rdg_dep_type
{
  /* Read After Write (RAW).  */
  flow_dd = 'f',

  /* Control dependence (execute conditional on).  */
  control_dd = 'c'
};

/* Dependence information attached to an edge of the RDG.  */

typedef struct rdg_edge
{
  /* Type of the dependence.  */
  enum rdg_dep_type type;
} *rdg_edge_p;

#define RDGE_TYPE(E)        ((struct rdg_edge *) ((E)->data))->type

/* Dump vertex I in RDG to FILE.  */

static void
dump_rdg_vertex (FILE *file, struct graph *rdg, int i)
{
  struct vertex *v = &(rdg->vertices[i]);
  struct graph_edge *e;

  fprintf (file, "(vertex %d: (%s%s) (in:", i,
	   RDG_MEM_WRITE_STMT (rdg, i) ? "w" : "",
	   RDG_MEM_READS_STMT (rdg, i) ? "r" : "");

  if (v->pred)
    for (e = v->pred; e; e = e->pred_next)
      fprintf (file, " %d", e->src);

  fprintf (file, ") (out:");

  if (v->succ)
    for (e = v->succ; e; e = e->succ_next)
      fprintf (file, " %d", e->dest);

  fprintf (file, ")\n");
  print_gimple_stmt (file, RDGV_STMT (v), 0, TDF_VOPS|TDF_MEMSYMS);
  fprintf (file, ")\n");
}

/* Call dump_rdg_vertex on stderr.  */

DEBUG_FUNCTION void
debug_rdg_vertex (struct graph *rdg, int i)
{
  dump_rdg_vertex (stderr, rdg, i);
}

/* Dump the reduced dependence graph RDG to FILE.  */

static void
dump_rdg (FILE *file, struct graph *rdg)
{
  fprintf (file, "(rdg\n");
  for (int i = 0; i < rdg->n_vertices; i++)
    dump_rdg_vertex (file, rdg, i);
  fprintf (file, ")\n");
}

/* Call dump_rdg on stderr.  */

DEBUG_FUNCTION void
debug_rdg (struct graph *rdg)
{
  dump_rdg (stderr, rdg);
}

static void
dot_rdg_1 (FILE *file, struct graph *rdg)
{
  int i;
  pretty_printer buffer;
  pp_needs_newline (&buffer) = false;
  buffer.buffer->stream = file;

  fprintf (file, "digraph RDG {\n");

  for (i = 0; i < rdg->n_vertices; i++)
    {
      struct vertex *v = &(rdg->vertices[i]);
      struct graph_edge *e;

      fprintf (file, "%d [label=\"[%d] ", i, i);
      pp_gimple_stmt_1 (&buffer, RDGV_STMT (v), 0, TDF_SLIM);
      pp_flush (&buffer);
      fprintf (file, "\"]\n");

      /* Highlight reads from memory.  */
      if (RDG_MEM_READS_STMT (rdg, i))
       fprintf (file, "%d [style=filled, fillcolor=green]\n", i);

      /* Highlight stores to memory.  */
      if (RDG_MEM_WRITE_STMT (rdg, i))
       fprintf (file, "%d [style=filled, fillcolor=red]\n", i);

      if (v->succ)
       for (e = v->succ; e; e = e->succ_next)
         switch (RDGE_TYPE (e))
           {
           case flow_dd:
             /* These are the most common dependences: don't print these. */
             fprintf (file, "%d -> %d \n", i, e->dest);
             break;

	   case control_dd:
             fprintf (file, "%d -> %d [label=control] \n", i, e->dest);
             break;

           default:
             gcc_unreachable ();
           }
    }

  fprintf (file, "}\n\n");
}

/* Display the Reduced Dependence Graph using dotty.  */

DEBUG_FUNCTION void
dot_rdg (struct graph *rdg)
{
  /* When debugging, you may want to enable the following code.  */
#if 1
  FILE *file = popen ("dot -Tx11", "w");
  if (!file)
    return;
  dot_rdg_1 (file, rdg);
  fflush (file);
  close (fileno (file));
  pclose (file);
#else
  dot_rdg_1 (stderr, rdg);
#endif
}

/* Returns the index of STMT in RDG.  */

static int
rdg_vertex_for_stmt (struct graph *rdg ATTRIBUTE_UNUSED, gimple stmt)
{
  int index = gimple_uid (stmt);
  gcc_checking_assert (index == -1 || RDG_STMT (rdg, index) == stmt);
  return index;
}

/* Creates dependence edges in RDG for all the uses of DEF.  IDEF is
   the index of DEF in RDG.  */

static void
create_rdg_edges_for_scalar (struct graph *rdg, tree def, int idef)
{
  use_operand_p imm_use_p;
  imm_use_iterator iterator;

  FOR_EACH_IMM_USE_FAST (imm_use_p, iterator, def)
    {
      struct graph_edge *e;
      int use = rdg_vertex_for_stmt (rdg, USE_STMT (imm_use_p));

      if (use < 0)
	continue;

      e = add_edge (rdg, idef, use);
      e->data = XNEW (struct rdg_edge);
      RDGE_TYPE (e) = flow_dd;
    }
}

/* Creates an edge for the control dependences of BB to the vertex V.  */

static void
create_edge_for_control_dependence (struct graph *rdg, basic_block bb,
				    int v, control_dependences *cd)
{
  bitmap_iterator bi;
  unsigned edge_n;
  EXECUTE_IF_SET_IN_BITMAP (cd->get_edges_dependent_on (bb->index),
			    0, edge_n, bi)
    {
      basic_block cond_bb = cd->get_edge (edge_n)->src;
      gimple stmt = last_stmt (cond_bb);
      if (stmt && is_ctrl_stmt (stmt))
	{
	  struct graph_edge *e;
	  int c = rdg_vertex_for_stmt (rdg, stmt);
	  if (c < 0)
	    continue;

	  e = add_edge (rdg, c, v);
	  e->data = XNEW (struct rdg_edge);
	  RDGE_TYPE (e) = control_dd;
	}
    }
}

/* Creates the edges of the reduced dependence graph RDG.  */

static void
create_rdg_flow_edges (struct graph *rdg)
{
  int i;
  def_operand_p def_p;
  ssa_op_iter iter;

  for (i = 0; i < rdg->n_vertices; i++)
    FOR_EACH_PHI_OR_STMT_DEF (def_p, RDG_STMT (rdg, i),
			      iter, SSA_OP_DEF)
      create_rdg_edges_for_scalar (rdg, DEF_FROM_PTR (def_p), i);
}

/* Creates the edges of the reduced dependence graph RDG.  */

static void
create_rdg_cd_edges (struct graph *rdg, control_dependences *cd)
{
  int i;

  for (i = 0; i < rdg->n_vertices; i++)
    {
      gimple stmt = RDG_STMT (rdg, i);
      if (gimple_code (stmt) == GIMPLE_PHI)
	{
	  edge_iterator ei;
	  edge e;
	  FOR_EACH_EDGE (e, ei, gimple_bb (stmt)->preds)
	      create_edge_for_control_dependence (rdg, e->src, i, cd);
	}
      else
	create_edge_for_control_dependence (rdg, gimple_bb (stmt), i, cd);
    }
}

/* Build the vertices of the reduced dependence graph RDG.  Return false
   if that failed.  */

static bool
create_rdg_vertices (struct graph *rdg, vec<gimple> stmts, loop_p loop,
		     vec<data_reference_p> *datarefs)
{
  int i;
  gimple stmt;

  FOR_EACH_VEC_ELT (stmts, i, stmt)
    {
      struct vertex *v = &(rdg->vertices[i]);

      /* Record statement to vertex mapping.  */
      gimple_set_uid (stmt, i);

      v->data = XNEW (struct rdg_vertex);
      RDGV_STMT (v) = stmt;
      RDGV_DATAREFS (v).create (0);
      RDGV_HAS_MEM_WRITE (v) = false;
      RDGV_HAS_MEM_READS (v) = false;
      if (gimple_code (stmt) == GIMPLE_PHI)
	continue;

      unsigned drp = datarefs->length ();
      if (!find_data_references_in_stmt (loop, stmt, datarefs))
	return false;
      for (unsigned j = drp; j < datarefs->length (); ++j)
	{
	  data_reference_p dr = (*datarefs)[j];
	  if (DR_IS_READ (dr))
	    RDGV_HAS_MEM_READS (v) = true;
	  else
	    RDGV_HAS_MEM_WRITE (v) = true;
	  RDGV_DATAREFS (v).safe_push (dr);
	}
    }
  return true;
}

/* Initialize STMTS with all the statements of LOOP.  The order in
   which we discover statements is important as
   generate_loops_for_partition is using the same traversal for
   identifying statements in loop copies.  */

static void
stmts_from_loop (struct loop *loop, vec<gimple> *stmts)
{
  unsigned int i;
  basic_block *bbs = get_loop_body_in_dom_order (loop);

  for (i = 0; i < loop->num_nodes; i++)
    {
      basic_block bb = bbs[i];
      gimple_stmt_iterator bsi;
      gimple stmt;

      for (bsi = gsi_start_phis (bb); !gsi_end_p (bsi); gsi_next (&bsi))
	if (!virtual_operand_p (gimple_phi_result (gsi_stmt (bsi))))
	  stmts->safe_push (gsi_stmt (bsi));

      for (bsi = gsi_start_bb (bb); !gsi_end_p (bsi); gsi_next (&bsi))
	{
	  stmt = gsi_stmt (bsi);
	  if (gimple_code (stmt) != GIMPLE_LABEL && !is_gimple_debug (stmt))
	    stmts->safe_push (stmt);
	}
    }

  free (bbs);
}

/* Free the reduced dependence graph RDG.  */

static void
free_rdg (struct graph *rdg)
{
  int i;

  for (i = 0; i < rdg->n_vertices; i++)
    {
      struct vertex *v = &(rdg->vertices[i]);
      struct graph_edge *e;

      for (e = v->succ; e; e = e->succ_next)
	free (e->data);

      if (v->data)
	{
	  gimple_set_uid (RDGV_STMT (v), -1);
	  free_data_refs (RDGV_DATAREFS (v));
	  free (v->data);
	}
    }

  free_graph (rdg);
}

/* Build the Reduced Dependence Graph (RDG) with one vertex per
   statement of the loop nest LOOP_NEST, and one edge per data dependence or
   scalar dependence.  */

static struct graph *
build_rdg (vec<loop_p> loop_nest, control_dependences *cd)
{
  struct graph *rdg;
  vec<data_reference_p> datarefs;

  /* Create the RDG vertices from the stmts of the loop nest.  */
  auto_vec<gimple, 10> stmts;
  stmts_from_loop (loop_nest[0], &stmts);
  rdg = new_graph (stmts.length ());
  datarefs.create (10);
  if (!create_rdg_vertices (rdg, stmts, loop_nest[0], &datarefs))
    {
      datarefs.release ();
      free_rdg (rdg);
      return NULL;
    }
  stmts.release ();

  create_rdg_flow_edges (rdg);
  if (cd)
    create_rdg_cd_edges (rdg, cd);

  datarefs.release ();

  return rdg;
}



enum partition_kind {
    PKIND_NORMAL, PKIND_MEMSET, PKIND_MEMCPY
};

typedef struct partition_s
{
  bitmap stmts;
  bitmap loops;
  bool reduction_p;
  enum partition_kind kind;
  /* data-references a kind != PKIND_NORMAL partition is about.  */
  data_reference_p main_dr;
  data_reference_p secondary_dr;
  tree niter;
  bool plus_one;
} *partition_t;


/* Allocate and initialize a partition from BITMAP.  */

static partition_t
partition_alloc (bitmap stmts, bitmap loops)
{
  partition_t partition = XCNEW (struct partition_s);
  partition->stmts = stmts ? stmts : BITMAP_ALLOC (NULL);
  partition->loops = loops ? loops : BITMAP_ALLOC (NULL);
  partition->reduction_p = false;
  partition->kind = PKIND_NORMAL;
  return partition;
}

/* Free PARTITION.  */

static void
partition_free (partition_t partition)
{
  BITMAP_FREE (partition->stmts);
  BITMAP_FREE (partition->loops);
  free (partition);
}

/* Returns true if the partition can be generated as a builtin.  */

static bool
partition_builtin_p (partition_t partition)
{
  return partition->kind != PKIND_NORMAL;
}

/* Returns true if the partition contains a reduction.  */

static bool
partition_reduction_p (partition_t partition)
{
  return partition->reduction_p;
}

/* Merge PARTITION into the partition DEST.  */

static void
partition_merge_into (partition_t dest, partition_t partition)
{
  dest->kind = PKIND_NORMAL;
  bitmap_ior_into (dest->stmts, partition->stmts);
  if (partition_reduction_p (partition))
    dest->reduction_p = true;
}


/* Returns true when DEF is an SSA_NAME defined in LOOP and used after
   the LOOP.  */

static bool
ssa_name_has_uses_outside_loop_p (tree def, loop_p loop)
{
  imm_use_iterator imm_iter;
  use_operand_p use_p;

  FOR_EACH_IMM_USE_FAST (use_p, imm_iter, def)
    {
      gimple use_stmt = USE_STMT (use_p);
      if (!is_gimple_debug (use_stmt)
	  && loop != loop_containing_stmt (use_stmt))
	return true;
    }

  return false;
}

/* Returns true when STMT defines a scalar variable used after the
   loop LOOP.  */

static bool
stmt_has_scalar_dependences_outside_loop (loop_p loop, gimple stmt)
{
  def_operand_p def_p;
  ssa_op_iter op_iter;

  if (gimple_code (stmt) == GIMPLE_PHI)
    return ssa_name_has_uses_outside_loop_p (gimple_phi_result (stmt), loop);

  FOR_EACH_SSA_DEF_OPERAND (def_p, stmt, op_iter, SSA_OP_DEF)
    if (ssa_name_has_uses_outside_loop_p (DEF_FROM_PTR (def_p), loop))
      return true;

  return false;
}

/* Return a copy of LOOP placed before LOOP.  */

static struct loop *
copy_loop_before (struct loop *loop)
{
  struct loop *res;
  edge preheader = loop_preheader_edge (loop);

  initialize_original_copy_tables ();
  res = slpeel_tree_duplicate_loop_to_edge_cfg (loop, NULL, preheader);
  gcc_assert (res != NULL);
  free_original_copy_tables ();
  delete_update_ssa ();

  return res;
}

/* Creates an empty basic block after LOOP.  */

static void
create_bb_after_loop (struct loop *loop)
{
  edge exit = single_exit (loop);

  if (!exit)
    return;

  split_edge (exit);
}

/* Generate code for PARTITION from the code in LOOP.  The loop is
   copied when COPY_P is true.  All the statements not flagged in the
   PARTITION bitmap are removed from the loop or from its copy.  The
   statements are indexed in sequence inside a basic block, and the
   basic blocks of a loop are taken in dom order.  */

static void
generate_loops_for_partition (struct loop *loop, partition_t partition,
			      bool copy_p)
{
  unsigned i;
  gimple_stmt_iterator bsi;
  basic_block *bbs;

  if (copy_p)
    {
      loop = copy_loop_before (loop);
      gcc_assert (loop != NULL);
      create_preheader (loop, CP_SIMPLE_PREHEADERS);
      create_bb_after_loop (loop);
    }

  /* Remove stmts not in the PARTITION bitmap.  */
  bbs = get_loop_body_in_dom_order (loop);

  if (MAY_HAVE_DEBUG_STMTS)
    for (i = 0; i < loop->num_nodes; i++)
      {
	basic_block bb = bbs[i];

	for (bsi = gsi_start_phis (bb); !gsi_end_p (bsi); gsi_next (&bsi))
	  {
	    gimple phi = gsi_stmt (bsi);
	    if (!virtual_operand_p (gimple_phi_result (phi))
		&& !bitmap_bit_p (partition->stmts, gimple_uid (phi)))
	      reset_debug_uses (phi);
	  }

	for (bsi = gsi_start_bb (bb); !gsi_end_p (bsi); gsi_next (&bsi))
	  {
	    gimple stmt = gsi_stmt (bsi);
	    if (gimple_code (stmt) != GIMPLE_LABEL
		&& !is_gimple_debug (stmt)
		&& !bitmap_bit_p (partition->stmts, gimple_uid (stmt)))
	      reset_debug_uses (stmt);
	  }
      }

  for (i = 0; i < loop->num_nodes; i++)
    {
      basic_block bb = bbs[i];

      for (bsi = gsi_start_phis (bb); !gsi_end_p (bsi);)
	{
	  gimple phi = gsi_stmt (bsi);
	  if (!virtual_operand_p (gimple_phi_result (phi))
	      && !bitmap_bit_p (partition->stmts, gimple_uid (phi)))
	    remove_phi_node (&bsi, true);
	  else
	    gsi_next (&bsi);
	}

      for (bsi = gsi_start_bb (bb); !gsi_end_p (bsi);)
	{
	  gimple stmt = gsi_stmt (bsi);
	  if (gimple_code (stmt) != GIMPLE_LABEL
	      && !is_gimple_debug (stmt)
	      && !bitmap_bit_p (partition->stmts, gimple_uid (stmt)))
	    {
	      /* Choose an arbitrary path through the empty CFG part
		 that this unnecessary control stmt controls.  */
	      if (gimple_code (stmt) == GIMPLE_COND)
		{
		  gimple_cond_make_false (stmt);
		  update_stmt (stmt);
		}
	      else if (gimple_code (stmt) == GIMPLE_SWITCH)
		{
		  gimple_switch_set_index
		      (stmt, CASE_LOW (gimple_switch_label (stmt, 1)));
		  update_stmt (stmt);
		}
	      else
		{
		  unlink_stmt_vdef (stmt);
		  gsi_remove (&bsi, true);
		  release_defs (stmt);
		  continue;
		}
	    }
	  gsi_next (&bsi);
	}
    }

  free (bbs);
}

/* Build the size argument for a memory operation call.  */

static tree
build_size_arg_loc (location_t loc, data_reference_p dr, tree nb_iter,
		    bool plus_one)
{
  tree size = fold_convert_loc (loc, sizetype, nb_iter);
  if (plus_one)
    size = size_binop (PLUS_EXPR, size, size_one_node);
  size = fold_build2_loc (loc, MULT_EXPR, sizetype, size,
			  TYPE_SIZE_UNIT (TREE_TYPE (DR_REF (dr))));
  size = fold_convert_loc (loc, size_type_node, size);
  return size;
}

/* Build an address argument for a memory operation call.  */

static tree
build_addr_arg_loc (location_t loc, data_reference_p dr, tree nb_bytes)
{
  tree addr_base;

  addr_base = size_binop_loc (loc, PLUS_EXPR, DR_OFFSET (dr), DR_INIT (dr));
  addr_base = fold_convert_loc (loc, sizetype, addr_base);

  /* Test for a negative stride, iterating over every element.  */
  if (tree_int_cst_sgn (DR_STEP (dr)) == -1)
    {
      addr_base = size_binop_loc (loc, MINUS_EXPR, addr_base,
				  fold_convert_loc (loc, sizetype, nb_bytes));
      addr_base = size_binop_loc (loc, PLUS_EXPR, addr_base,
				  TYPE_SIZE_UNIT (TREE_TYPE (DR_REF (dr))));
    }

  return fold_build_pointer_plus_loc (loc, DR_BASE_ADDRESS (dr), addr_base);
}

/* If VAL memory representation contains the same value in all bytes,
   return that value, otherwise return -1.
   E.g. for 0x24242424 return 0x24, for IEEE double
   747708026454360457216.0 return 0x44, etc.  */

static int
const_with_all_bytes_same (tree val)
{
  unsigned char buf[64];
  int i, len;

  if (integer_zerop (val)
      || real_zerop (val)
      || (TREE_CODE (val) == CONSTRUCTOR
          && !TREE_CLOBBER_P (val)
          && CONSTRUCTOR_NELTS (val) == 0))
    return 0;

  if (CHAR_BIT != 8 || BITS_PER_UNIT != 8)
    return -1;

  len = native_encode_expr (val, buf, sizeof (buf));
  if (len == 0)
    return -1;
  for (i = 1; i < len; i++)
    if (buf[i] != buf[0])
      return -1;
  return buf[0];
}

/* Generate a call to memset for PARTITION in LOOP.  */

static void
generate_memset_builtin (struct loop *loop, partition_t partition)
{
  gimple_stmt_iterator gsi;
  gimple stmt, fn_call;
  tree mem, fn, nb_bytes;
  location_t loc;
  tree val;

  stmt = DR_STMT (partition->main_dr);
  loc = gimple_location (stmt);

  /* The new statements will be placed before LOOP.  */
  gsi = gsi_last_bb (loop_preheader_edge (loop)->src);

  nb_bytes = build_size_arg_loc (loc, partition->main_dr, partition->niter,
				 partition->plus_one);
  nb_bytes = force_gimple_operand_gsi (&gsi, nb_bytes, true, NULL_TREE,
				       false, GSI_CONTINUE_LINKING);
  mem = build_addr_arg_loc (loc, partition->main_dr, nb_bytes);
  mem = force_gimple_operand_gsi (&gsi, mem, true, NULL_TREE,
				  false, GSI_CONTINUE_LINKING);

  /* This exactly matches the pattern recognition in classify_partition.  */
  val = gimple_assign_rhs1 (stmt);
  /* Handle constants like 0x15151515 and similarly
     floating point constants etc. where all bytes are the same.  */
  int bytev = const_with_all_bytes_same (val);
  if (bytev != -1)
    val = build_int_cst (integer_type_node, bytev);
  else if (TREE_CODE (val) == INTEGER_CST)
    val = fold_convert (integer_type_node, val);
  else if (!useless_type_conversion_p (integer_type_node, TREE_TYPE (val)))
    {
      gimple cstmt;
      tree tem = make_ssa_name (integer_type_node, NULL);
      cstmt = gimple_build_assign_with_ops (NOP_EXPR, tem, val, NULL_TREE);
      gsi_insert_after (&gsi, cstmt, GSI_CONTINUE_LINKING);
      val = tem;
    }

  fn = build_fold_addr_expr (builtin_decl_implicit (BUILT_IN_MEMSET));
  fn_call = gimple_build_call (fn, 3, mem, val, nb_bytes);
  gsi_insert_after (&gsi, fn_call, GSI_CONTINUE_LINKING);

  if (dump_file && (dump_flags & TDF_DETAILS))
    {
      fprintf (dump_file, "generated memset");
      if (bytev == 0)
	fprintf (dump_file, " zero\n");
      else
	fprintf (dump_file, "\n");
    }
}

/* Generate a call to memcpy for PARTITION in LOOP.  */

static void
generate_memcpy_builtin (struct loop *loop, partition_t partition)
{
  gimple_stmt_iterator gsi;
  gimple stmt, fn_call;
  tree dest, src, fn, nb_bytes;
  location_t loc;
  enum built_in_function kind;

  stmt = DR_STMT (partition->main_dr);
  loc = gimple_location (stmt);

  /* The new statements will be placed before LOOP.  */
  gsi = gsi_last_bb (loop_preheader_edge (loop)->src);

  nb_bytes = build_size_arg_loc (loc, partition->main_dr, partition->niter,
				 partition->plus_one);
  nb_bytes = force_gimple_operand_gsi (&gsi, nb_bytes, true, NULL_TREE,
				       false, GSI_CONTINUE_LINKING);
  dest = build_addr_arg_loc (loc, partition->main_dr, nb_bytes);
  src = build_addr_arg_loc (loc, partition->secondary_dr, nb_bytes);
  if (ptr_derefs_may_alias_p (dest, src))
    kind = BUILT_IN_MEMMOVE;
  else
    kind = BUILT_IN_MEMCPY;

  dest = force_gimple_operand_gsi (&gsi, dest, true, NULL_TREE,
				   false, GSI_CONTINUE_LINKING);
  src = force_gimple_operand_gsi (&gsi, src, true, NULL_TREE,
				  false, GSI_CONTINUE_LINKING);
  fn = build_fold_addr_expr (builtin_decl_implicit (kind));
  fn_call = gimple_build_call (fn, 3, dest, src, nb_bytes);
  gsi_insert_after (&gsi, fn_call, GSI_CONTINUE_LINKING);

  if (dump_file && (dump_flags & TDF_DETAILS))
    {
      if (kind == BUILT_IN_MEMCPY)
	fprintf (dump_file, "generated memcpy\n");
      else
	fprintf (dump_file, "generated memmove\n");
    }
}

/* Remove and destroy the loop LOOP.  */

static void
destroy_loop (struct loop *loop)
{
  unsigned nbbs = loop->num_nodes;
  edge exit = single_exit (loop);
  basic_block src = loop_preheader_edge (loop)->src, dest = exit->dest;
  basic_block *bbs;
  unsigned i;

  bbs = get_loop_body_in_dom_order (loop);

  redirect_edge_pred (exit, src);
  exit->flags &= ~(EDGE_TRUE_VALUE|EDGE_FALSE_VALUE);
  exit->flags |= EDGE_FALLTHRU;
  cancel_loop_tree (loop);
  rescan_loop_exit (exit, false, true);

  for (i = 0; i < nbbs; i++)
    {
      /* We have made sure to not leave any dangling uses of SSA
         names defined in the loop.  With the exception of virtuals.
	 Make sure we replace all uses of virtual defs that will remain
	 outside of the loop with the bare symbol as delete_basic_block
	 will release them.  */
      gimple_stmt_iterator gsi;
      for (gsi = gsi_start_phis (bbs[i]); !gsi_end_p (gsi); gsi_next (&gsi))
	{
	  gimple phi = gsi_stmt (gsi);
	  if (virtual_operand_p (gimple_phi_result (phi)))
	    mark_virtual_phi_result_for_renaming (phi);
	}
      for (gsi = gsi_start_bb (bbs[i]); !gsi_end_p (gsi); gsi_next (&gsi))
	{
	  gimple stmt = gsi_stmt (gsi);
	  tree vdef = gimple_vdef (stmt);
	  if (vdef && TREE_CODE (vdef) == SSA_NAME)
	    mark_virtual_operand_for_renaming (vdef);
	}
      delete_basic_block (bbs[i]);
    }
  free (bbs);

  set_immediate_dominator (CDI_DOMINATORS, dest,
			   recompute_dominator (CDI_DOMINATORS, dest));
}

/* Generates code for PARTITION.  */

static void
generate_code_for_partition (struct loop *loop,
			     partition_t partition, bool copy_p)
{
  switch (partition->kind)
    {
    case PKIND_NORMAL:
      /* Reductions all have to be in the last partition.  */
      gcc_assert (!partition_reduction_p (partition)
		  || !copy_p);
      generate_loops_for_partition (loop, partition, copy_p);
      return;

    case PKIND_MEMSET:
      generate_memset_builtin (loop, partition);
      break;

    case PKIND_MEMCPY:
      generate_memcpy_builtin (loop, partition);
      break;

    default:
      gcc_unreachable ();
    }

  /* Common tail for partitions we turn into a call.  If this was the last
     partition for which we generate code, we have to destroy the loop.  */
  if (!copy_p)
    destroy_loop (loop);
}


/* Returns a partition with all the statements needed for computing
   the vertex V of the RDG, also including the loop exit conditions.  */

static partition_t
build_rdg_partition_for_vertex (struct graph *rdg, int v)
{
  partition_t partition = partition_alloc (NULL, NULL);
  auto_vec<int, 3> nodes;
  unsigned i;
  int x;

  graphds_dfs (rdg, &v, 1, &nodes, false, NULL);

  FOR_EACH_VEC_ELT (nodes, i, x)
    {
      bitmap_set_bit (partition->stmts, x);
      bitmap_set_bit (partition->loops,
		      loop_containing_stmt (RDG_STMT (rdg, x))->num);
    }

  return partition;
}

/* Classifies the builtin kind we can generate for PARTITION of RDG and LOOP.
   For the moment we detect only the memset zero pattern.  */

static void
classify_partition (loop_p loop, struct graph *rdg, partition_t partition)
{
  bitmap_iterator bi;
  unsigned i;
  tree nb_iter;
  data_reference_p single_load, single_store;
  bool volatiles_p = false;
  bool plus_one = false;

  partition->kind = PKIND_NORMAL;
  partition->main_dr = NULL;
  partition->secondary_dr = NULL;
  partition->niter = NULL_TREE;
  partition->plus_one = false;

  EXECUTE_IF_SET_IN_BITMAP (partition->stmts, 0, i, bi)
    {
      gimple stmt = RDG_STMT (rdg, i);

      if (gimple_has_volatile_ops (stmt))
	volatiles_p = true;

      /* If the stmt has uses outside of the loop mark it as reduction.  */
      if (stmt_has_scalar_dependences_outside_loop (loop, stmt))
	{
	  partition->reduction_p = true;
	  return;
	}
    }

  /* Perform general partition disqualification for builtins.  */
  if (volatiles_p
      || !flag_tree_loop_distribute_patterns)
    return;

  /* Detect memset and memcpy.  */
  single_load = NULL;
  single_store = NULL;
  EXECUTE_IF_SET_IN_BITMAP (partition->stmts, 0, i, bi)
    {
      gimple stmt = RDG_STMT (rdg, i);
      data_reference_p dr;
      unsigned j;

      if (gimple_code (stmt) == GIMPLE_PHI)
	continue;

      /* Any scalar stmts are ok.  */
      if (!gimple_vuse (stmt))
	continue;

      /* Otherwise just regular loads/stores.  */
      if (!gimple_assign_single_p (stmt))
	return;

      /* But exactly one store and/or load.  */
      for (j = 0; RDG_DATAREFS (rdg, i).iterate (j, &dr); ++j)
	{
	  if (DR_IS_READ (dr))
	    {
	      if (single_load != NULL)
		return;
	      single_load = dr;
	    }
	  else
	    {
	      if (single_store != NULL)
		return;
	      single_store = dr;
	    }
	}
    }

  if (!single_store)
    return;

  nb_iter = number_of_latch_executions (loop);
  if (!nb_iter || nb_iter == chrec_dont_know)
    return;
  if (dominated_by_p (CDI_DOMINATORS, single_exit (loop)->src,
		      gimple_bb (DR_STMT (single_store))))
    plus_one = true;

  if (single_store && !single_load)
    {
      gimple stmt = DR_STMT (single_store);
      tree rhs = gimple_assign_rhs1 (stmt);
      if (const_with_all_bytes_same (rhs) == -1
	  && (!INTEGRAL_TYPE_P (TREE_TYPE (rhs))
	      || (TYPE_MODE (TREE_TYPE (rhs))
		  != TYPE_MODE (unsigned_char_type_node))))
	return;
      if (TREE_CODE (rhs) == SSA_NAME
	  && !SSA_NAME_IS_DEFAULT_DEF (rhs)
	  && flow_bb_inside_loop_p (loop, gimple_bb (SSA_NAME_DEF_STMT (rhs))))
	return;
      if (!adjacent_dr_p (single_store)
	  || !dominated_by_p (CDI_DOMINATORS,
			      loop->latch, gimple_bb (stmt)))
	return;
      partition->kind = PKIND_MEMSET;
      partition->main_dr = single_store;
      partition->niter = nb_iter;
      partition->plus_one = plus_one;
    }
  else if (single_store && single_load)
    {
      gimple store = DR_STMT (single_store);
      gimple load = DR_STMT (single_load);
      /* Direct aggregate copy or via an SSA name temporary.  */
      if (load != store
	  && gimple_assign_lhs (load) != gimple_assign_rhs1 (store))
	return;
      if (!adjacent_dr_p (single_store)
	  || !adjacent_dr_p (single_load)
	  || !operand_equal_p (DR_STEP (single_store),
			       DR_STEP (single_load), 0)
	  || !dominated_by_p (CDI_DOMINATORS,
			      loop->latch, gimple_bb (store)))
	return;
      /* Now check that if there is a dependence this dependence is
         of a suitable form for memmove.  */
      vec<loop_p> loops = vNULL;
      ddr_p ddr;
      loops.safe_push (loop);
      ddr = initialize_data_dependence_relation (single_load, single_store,
						 loops);
      compute_affine_dependence (ddr, loop);
      if (DDR_ARE_DEPENDENT (ddr) == chrec_dont_know)
	{
	  free_dependence_relation (ddr);
	  loops.release ();
	  return;
	}
      if (DDR_ARE_DEPENDENT (ddr) != chrec_known)
	{
	  if (DDR_NUM_DIST_VECTS (ddr) == 0)
	    {
	      free_dependence_relation (ddr);
	      loops.release ();
	      return;
	    }
	  lambda_vector dist_v;
	  FOR_EACH_VEC_ELT (DDR_DIST_VECTS (ddr), i, dist_v)
	    {
	      int dist = dist_v[index_in_loop_nest (loop->num,
						    DDR_LOOP_NEST (ddr))];
	      if (dist > 0 && !DDR_REVERSED_P (ddr))
		{
		  free_dependence_relation (ddr);
		  loops.release ();
		  return;
		}
	    }
	}
      free_dependence_relation (ddr);
      loops.release ();
      partition->kind = PKIND_MEMCPY;
      partition->main_dr = single_store;
      partition->secondary_dr = single_load;
      partition->niter = nb_iter;
      partition->plus_one = plus_one;
    }
}

/* For a data reference REF, return the declaration of its base
   address or NULL_TREE if the base is not determined.  */

static tree
ref_base_address (data_reference_p dr)
{
  tree base_address = DR_BASE_ADDRESS (dr);
  if (base_address
      && TREE_CODE (base_address) == ADDR_EXPR)
    return TREE_OPERAND (base_address, 0);

  return base_address;
}

/* Returns true when PARTITION1 and PARTITION2 have similar memory
   accesses in RDG.  */

static bool
similar_memory_accesses (struct graph *rdg, partition_t partition1,
			 partition_t partition2)
{
  unsigned i, j, k, l;
  bitmap_iterator bi, bj;
  data_reference_p ref1, ref2;

  /* First check whether in the intersection of the two partitions are
     any loads or stores.  Common loads are the situation that happens
     most often.  */
  EXECUTE_IF_AND_IN_BITMAP (partition1->stmts, partition2->stmts, 0, i, bi)
    if (RDG_MEM_WRITE_STMT (rdg, i)
	|| RDG_MEM_READS_STMT (rdg, i))
      return true;

  /* Then check all data-references against each other.  */
  EXECUTE_IF_SET_IN_BITMAP (partition1->stmts, 0, i, bi)
    if (RDG_MEM_WRITE_STMT (rdg, i)
	|| RDG_MEM_READS_STMT (rdg, i))
      EXECUTE_IF_SET_IN_BITMAP (partition2->stmts, 0, j, bj)
	if (RDG_MEM_WRITE_STMT (rdg, j)
	    || RDG_MEM_READS_STMT (rdg, j))
	  {
	    FOR_EACH_VEC_ELT (RDG_DATAREFS (rdg, i), k, ref1)
	      {
		tree base1 = ref_base_address (ref1);
		if (base1)
		  FOR_EACH_VEC_ELT (RDG_DATAREFS (rdg, j), l, ref2)
		    if (base1 == ref_base_address (ref2))
		      return true;
	      }
	  }

  return false;
}

/* Aggregate several components into a useful partition that is
   registered in the PARTITIONS vector.  Partitions will be
   distributed in different loops.  */

static void
rdg_build_partitions (struct graph *rdg,
		      vec<gimple> starting_stmts,
		      vec<partition_t> *partitions)
{
  bitmap processed = BITMAP_ALLOC (NULL);
  int i;
  gimple stmt;

  FOR_EACH_VEC_ELT (starting_stmts, i, stmt)
    {
      int v = rdg_vertex_for_stmt (rdg, stmt);

      if (dump_file && (dump_flags & TDF_DETAILS))
	fprintf (dump_file,
		 "ldist asked to generate code for vertex %d\n", v);

      /* If the vertex is already contained in another partition so
         is the partition rooted at it.  */
      if (bitmap_bit_p (processed, v))
	continue;

      partition_t partition = build_rdg_partition_for_vertex (rdg, v);
      bitmap_ior_into (processed, partition->stmts);

      if (dump_file && (dump_flags & TDF_DETAILS))
	{
	  fprintf (dump_file, "ldist useful partition:\n");
	  dump_bitmap (dump_file, partition->stmts);
	}

      partitions->safe_push (partition);
    }

  /* All vertices should have been assigned to at least one partition now,
     other than vertices belonging to dead code.  */

  BITMAP_FREE (processed);
}

/* Dump to FILE the PARTITIONS.  */

static void
dump_rdg_partitions (FILE *file, vec<partition_t> partitions)
{
  int i;
  partition_t partition;

  FOR_EACH_VEC_ELT (partitions, i, partition)
    debug_bitmap_file (file, partition->stmts);
}

/* Debug PARTITIONS.  */
extern void debug_rdg_partitions (vec<partition_t> );

DEBUG_FUNCTION void
debug_rdg_partitions (vec<partition_t> partitions)
{
  dump_rdg_partitions (stderr, partitions);
}

/* Returns the number of read and write operations in the RDG.  */

static int
number_of_rw_in_rdg (struct graph *rdg)
{
  int i, res = 0;

  for (i = 0; i < rdg->n_vertices; i++)
    {
      if (RDG_MEM_WRITE_STMT (rdg, i))
	++res;

      if (RDG_MEM_READS_STMT (rdg, i))
	++res;
    }

  return res;
}

/* Returns the number of read and write operations in a PARTITION of
   the RDG.  */

static int
number_of_rw_in_partition (struct graph *rdg, partition_t partition)
{
  int res = 0;
  unsigned i;
  bitmap_iterator ii;

  EXECUTE_IF_SET_IN_BITMAP (partition->stmts, 0, i, ii)
    {
      if (RDG_MEM_WRITE_STMT (rdg, i))
	++res;

      if (RDG_MEM_READS_STMT (rdg, i))
	++res;
    }

  return res;
}

/* Returns true when one of the PARTITIONS contains all the read or
   write operations of RDG.  */

static bool
partition_contains_all_rw (struct graph *rdg,
			   vec<partition_t> partitions)
{
  int i;
  partition_t partition;
  int nrw = number_of_rw_in_rdg (rdg);

  FOR_EACH_VEC_ELT (partitions, i, partition)
    if (nrw == number_of_rw_in_partition (rdg, partition))
      return true;

  return false;
}

/* Compute partition dependence created by the data references in DRS1
   and DRS2 and modify and return DIR according to that.  */

static int
pg_add_dependence_edges (struct graph *rdg, vec<loop_p> loops, int dir,
			 vec<data_reference_p> drs1,
			 vec<data_reference_p> drs2)
{
  data_reference_p dr1, dr2;

  /* dependence direction - 0 is no dependence, -1 is back,
     1 is forth, 2 is both (we can stop then, merging will occur).  */
  for (int ii = 0; drs1.iterate (ii, &dr1); ++ii)
    for (int jj = 0; drs2.iterate (jj, &dr2); ++jj)
      {
	int this_dir = 1;
	ddr_p ddr;
	/* Re-shuffle data-refs to be in dominator order.  */
	if (rdg_vertex_for_stmt (rdg, DR_STMT (dr1))
	    > rdg_vertex_for_stmt (rdg, DR_STMT (dr2)))
	  {
	    data_reference_p tem = dr1;
	    dr1 = dr2;
	    dr2 = tem;
	    this_dir = -this_dir;
	  }
	ddr = initialize_data_dependence_relation (dr1, dr2, loops);
	compute_affine_dependence (ddr, loops[0]);
	if (DDR_ARE_DEPENDENT (ddr) == chrec_dont_know)
	  this_dir = 2;
	else if (DDR_ARE_DEPENDENT (ddr) == NULL_TREE)
	  {
	    if (DDR_REVERSED_P (ddr))
	      {
		data_reference_p tem = dr1;
		dr1 = dr2;
		dr2 = tem;
		this_dir = -this_dir;
	      }
	    /* Known dependences can still be unordered througout the
	       iteration space, see gcc.dg/tree-ssa/ldist-16.c.  */
	    if (DDR_NUM_DIST_VECTS (ddr) != 1)
	      this_dir = 2;
	    /* If the overlap is exact preserve stmt order.  */
	    else if (lambda_vector_zerop (DDR_DIST_VECT (ddr, 0), 1))
	      ;
	    else
	      {
		/* Else as the distance vector is lexicographic positive
		   swap the dependence direction.  */
		this_dir = -this_dir;
	      }
	  }
	else
	  this_dir = 0;
	free_dependence_relation (ddr);
	if (dir == 0)
	  dir = this_dir;
	else if (dir != this_dir)
	  return 2;
      }
  return dir;
}

/* Compare postorder number of the partition graph vertices V1 and V2.  */

static int
pgcmp (const void *v1_, const void *v2_)
{
  const vertex *v1 = (const vertex *)v1_;
  const vertex *v2 = (const vertex *)v2_;
  return v2->post - v1->post;
}

/* Distributes the code from LOOP in such a way that producer
   statements are placed before consumer statements.  Tries to separate
   only the statements from STMTS into separate loops.
   Returns the number of distributed loops.  */

static int
distribute_loop (struct loop *loop, vec<gimple> stmts,
		 control_dependences *cd, int *nb_calls)
{
  struct graph *rdg;
  partition_t partition;
  bool any_builtin;
  int i, nbp;
  graph *pg = NULL;
  int num_sccs = 1;

  *nb_calls = 0;
  auto_vec<loop_p, 3> loop_nest;
  if (!find_loop_nest (loop, &loop_nest))
    return 0;

  rdg = build_rdg (loop_nest, cd);
  if (!rdg)
    {
      if (dump_file && (dump_flags & TDF_DETAILS))
	fprintf (dump_file,
		 "Loop %d not distributed: failed to build the RDG.\n",
		 loop->num);

      return 0;
    }

  if (dump_file && (dump_flags & TDF_DETAILS))
    dump_rdg (dump_file, rdg);

  auto_vec<partition_t, 3> partitions;
  rdg_build_partitions (rdg, stmts, &partitions);

  any_builtin = false;
  FOR_EACH_VEC_ELT (partitions, i, partition)
    {
      classify_partition (loop, rdg, partition);
      any_builtin |= partition_builtin_p (partition);
    }

  /* If we are only distributing patterns but did not detect any,
     simply bail out.  */
  if (!flag_tree_loop_distribution
      && !any_builtin)
    {
      nbp = 0;
      goto ldist_done;
    }

  /* If we are only distributing patterns fuse all partitions that
     were not classified as builtins.  This also avoids chopping
     a loop into pieces, separated by builtin calls.  That is, we
     only want no or a single loop body remaining.  */
  partition_t into;
  if (!flag_tree_loop_distribution)
    {
      for (i = 0; partitions.iterate (i, &into); ++i)
	if (!partition_builtin_p (into))
	  break;
      for (++i; partitions.iterate (i, &partition); ++i)
	if (!partition_builtin_p (partition))
	  {
	    if (dump_file && (dump_flags & TDF_DETAILS))
	      {
		fprintf (dump_file, "fusing non-builtin partitions\n");
		dump_bitmap (dump_file, into->stmts);
		dump_bitmap (dump_file, partition->stmts);
	      }
	    partition_merge_into (into, partition);
	    partitions.unordered_remove (i);
	    partition_free (partition);
	    i--;
	  }
    }

  /* Due to limitations in the transform phase we have to fuse all
     reduction partitions into the last partition so the existing
     loop will contain all loop-closed PHI nodes.  */
  for (i = 0; partitions.iterate (i, &into); ++i)
    if (partition_reduction_p (into))
      break;
  for (i = i + 1; partitions.iterate (i, &partition); ++i)
    if (partition_reduction_p (partition))
      {
	if (dump_file && (dump_flags & TDF_DETAILS))
	  {
	    fprintf (dump_file, "fusing partitions\n");
	    dump_bitmap (dump_file, into->stmts);
	    dump_bitmap (dump_file, partition->stmts);
	    fprintf (dump_file, "because they have reductions\n");
	  }
	partition_merge_into (into, partition);
	partitions.unordered_remove (i);
	partition_free (partition);
	i--;
      }

  /* Apply our simple cost model - fuse partitions with similar
     memory accesses.  */
  for (i = 0; partitions.iterate (i, &into); ++i)
    {
      if (partition_builtin_p (into))
	continue;
      for (int j = i + 1;
	   partitions.iterate (j, &partition); ++j)
	{
	  if (!partition_builtin_p (partition)
	      && similar_memory_accesses (rdg, into, partition))
	    {
	      if (dump_file && (dump_flags & TDF_DETAILS))
		{
		  fprintf (dump_file, "fusing partitions\n");
		  dump_bitmap (dump_file, into->stmts);
		  dump_bitmap (dump_file, partition->stmts);
		  fprintf (dump_file, "because they have similar "
			   "memory accesses\n");
		}
	      partition_merge_into (into, partition);
	      partitions.unordered_remove (j);
	      partition_free (partition);
	      j--;
	    }
	}
    }

  /* Build the partition dependency graph.  */
  if (partitions.length () > 1)
    {
      pg = new_graph (partitions.length ());
      struct pgdata {
	  partition_t partition;
	  vec<data_reference_p> writes;
	  vec<data_reference_p> reads;
      };
#define PGDATA(i) ((pgdata *)(pg->vertices[i].data))
      for (i = 0; partitions.iterate (i, &partition); ++i)
	{
	  vertex *v = &pg->vertices[i];
	  pgdata *data = new pgdata;
	  data_reference_p dr;
	  /* FIXME - leaks.  */
	  v->data = data;
	  bitmap_iterator bi;
	  unsigned j;
	  data->partition = partition;
	  data->reads = vNULL;
	  data->writes = vNULL;
	  EXECUTE_IF_SET_IN_BITMAP (partition->stmts, 0, j, bi)
	    for (int k = 0; RDG_DATAREFS (rdg, j).iterate (k, &dr); ++k)
	      if (DR_IS_READ (dr))
		data->reads.safe_push (dr);
	      else
		data->writes.safe_push (dr);
	}
      partition_t partition1, partition2;
      for (i = 0; partitions.iterate (i, &partition1); ++i)
	for (int j = i + 1; partitions.iterate (j, &partition2); ++j)
	  {
	    /* dependence direction - 0 is no dependence, -1 is back,
	       1 is forth, 2 is both (we can stop then, merging will occur).  */
	    int dir = 0;
	    dir = pg_add_dependence_edges (rdg, loop_nest, dir,
					   PGDATA(i)->writes,
					   PGDATA(j)->reads);
	    if (dir != 2)
	      dir = pg_add_dependence_edges (rdg, loop_nest, dir,
					     PGDATA(i)->reads,
					     PGDATA(j)->writes);
	    if (dir != 2)
	      dir = pg_add_dependence_edges (rdg, loop_nest, dir,
					     PGDATA(i)->writes,
					     PGDATA(j)->writes);
	    if (dir == 1 || dir == 2)
	      add_edge (pg, i, j);
	    if (dir == -1 || dir == 2)
	      add_edge (pg, j, i);
	  }

      /* Add edges to the reduction partition (if any) to force it last.  */
      unsigned j;
      for (j = 0; partitions.iterate (j, &partition); ++j)
	if (partition_reduction_p (partition))
	  break;
      if (j < partitions.length ())
	{
	  for (unsigned i = 0; partitions.iterate (i, &partition); ++i)
	    if (i != j)
	      add_edge (pg, i, j);
	}

      /* Compute partitions we cannot separate and fuse them.  */
      num_sccs = graphds_scc (pg, NULL);
      for (i = 0; i < num_sccs; ++i)
	{
	  partition_t first;
	  int j;
	  for (j = 0; partitions.iterate (j, &first); ++j)
	    if (pg->vertices[j].component == i)
	      break;
	  for (j = j + 1; partitions.iterate (j, &partition); ++j)
	    if (pg->vertices[j].component == i)
	      {
		if (dump_file && (dump_flags & TDF_DETAILS))
		  {
		    fprintf (dump_file, "fusing partitions\n");
		    dump_bitmap (dump_file, first->stmts);
		    dump_bitmap (dump_file, partition->stmts);
		    fprintf (dump_file, "because they are in the same "
			     "dependence SCC\n");
		  }
		partition_merge_into (first, partition);
		partitions[j] = NULL;
		partition_free (partition);
		PGDATA (j)->partition = NULL;
	      }
	}

      /* Now order the remaining nodes in postorder.  */
      qsort (pg->vertices, pg->n_vertices, sizeof (vertex), pgcmp);
      partitions.truncate (0);
      for (i = 0; i < pg->n_vertices; ++i)
	{
	  pgdata *data = PGDATA (i);
	  if (data->partition)
	    partitions.safe_push (data->partition);
	  data->reads.release ();
	  data->writes.release ();
	  delete data;
	}
      gcc_assert (partitions.length () == (unsigned)num_sccs);
      free_graph (pg);
    }

  nbp = partitions.length ();
  if (nbp == 0
      || (nbp == 1 && !partition_builtin_p (partitions[0]))
      || (nbp > 1 && partition_contains_all_rw (rdg, partitions)))
    {
      nbp = 0;
      goto ldist_done;
    }

  if (dump_file && (dump_flags & TDF_DETAILS))
    dump_rdg_partitions (dump_file, partitions);

  FOR_EACH_VEC_ELT (partitions, i, partition)
    {
      if (partition_builtin_p (partition))
	(*nb_calls)++;
      generate_code_for_partition (loop, partition, i < nbp - 1);
    }

 ldist_done:

  FOR_EACH_VEC_ELT (partitions, i, partition)
    partition_free (partition);

  free_rdg (rdg);
  return nbp - *nb_calls;
}

/* Distribute all loops in the current function.  */

static unsigned int
tree_loop_distribution (void)
{
  struct loop *loop;
  bool changed = false;
  basic_block bb;
  control_dependences *cd = NULL;

  FOR_ALL_BB_FN (bb, cfun)
    {
      gimple_stmt_iterator gsi;
      for (gsi = gsi_start_phis (bb); !gsi_end_p (gsi); gsi_next (&gsi))
	gimple_set_uid (gsi_stmt (gsi), -1);
      for (gsi = gsi_start_bb (bb); !gsi_end_p (gsi); gsi_next (&gsi))
	gimple_set_uid (gsi_stmt (gsi), -1);
    }

  /* We can at the moment only distribute non-nested loops, thus restrict
     walking to innermost loops.  */
  FOR_EACH_LOOP (loop, LI_ONLY_INNERMOST)
    {
      auto_vec<gimple> work_list;
      basic_block *bbs;
      int num = loop->num;
      unsigned int i;

      /* If the loop doesn't have a single exit we will fail anyway,
	 so do that early.  */
      if (!single_exit (loop))
	continue;

      /* Only optimize hot loops.  */
      if (!optimize_loop_for_speed_p (loop))
	continue;

      /* Initialize the worklist with stmts we seed the partitions with.  */
      bbs = get_loop_body_in_dom_order (loop);
      for (i = 0; i < loop->num_nodes; ++i)
	{
	  gimple_stmt_iterator gsi;
	  for (gsi = gsi_start_phis (bbs[i]); !gsi_end_p (gsi); gsi_next (&gsi))
	    {
	      gimple phi = gsi_stmt (gsi);
	      if (virtual_operand_p (gimple_phi_result (phi)))
		continue;
	      /* Distribute stmts which have defs that are used outside of
	         the loop.  */
	      if (!stmt_has_scalar_dependences_outside_loop (loop, phi))
		continue;
	      work_list.safe_push (phi);
	    }
	  for (gsi = gsi_start_bb (bbs[i]); !gsi_end_p (gsi); gsi_next (&gsi))
	    {
	      gimple stmt = gsi_stmt (gsi);

	      /* If there is a stmt with side-effects bail out - we
	         cannot and should not distribute this loop.  */
	      if (gimple_has_side_effects (stmt))
		{
		  work_list.truncate (0);
		  goto out;
		}

	      /* Distribute stmts which have defs that are used outside of
	         the loop.  */
	      if (stmt_has_scalar_dependences_outside_loop (loop, stmt))
		;
	      /* Otherwise only distribute stores for now.  */
	      else if (!gimple_vdef (stmt))
		continue;

	      work_list.safe_push (stmt);
	    }
	}
out:
      free (bbs);

      int nb_generated_loops = 0;
      int nb_generated_calls = 0;
      location_t loc = find_loop_location (loop);
      if (work_list.length () > 0)
	{
	  if (!cd)
	    {
	      calculate_dominance_info (CDI_DOMINATORS);
	      calculate_dominance_info (CDI_POST_DOMINATORS);
	      cd = new control_dependences (create_edge_list ());
	      free_dominance_info (CDI_POST_DOMINATORS);
	    }
	  nb_generated_loops = distribute_loop (loop, work_list, cd,
						&nb_generated_calls);
	}

      if (nb_generated_loops + nb_generated_calls > 0)
	{
	  changed = true;
	  dump_printf_loc (MSG_OPTIMIZED_LOCATIONS,
			   loc, "Loop %d distributed: split to %d loops "
			   "and %d library calls.\n",
			   num, nb_generated_loops, nb_generated_calls);
	}
      else if (dump_file && (dump_flags & TDF_DETAILS))
	fprintf (dump_file, "Loop %d is the same.\n", num);
    }

  if (cd)
    delete cd;

  if (changed)
    {
      mark_virtual_operands_for_renaming (cfun);
      rewrite_into_loop_closed_ssa (NULL, TODO_update_ssa);
    }

#ifdef ENABLE_CHECKING
  verify_loop_structure ();
#endif

  return 0;
}

static bool
gate_tree_loop_distribution (void)
{
  return flag_tree_loop_distribution
    || flag_tree_loop_distribute_patterns;
}

namespace {

const pass_data pass_data_loop_distribution =
{
  GIMPLE_PASS, /* type */
  "ldist", /* name */
  OPTGROUP_LOOP, /* optinfo_flags */
  true, /* has_gate */
  true, /* has_execute */
  TV_TREE_LOOP_DISTRIBUTION, /* tv_id */
  ( PROP_cfg | PROP_ssa ), /* properties_required */
  0, /* properties_provided */
  0, /* properties_destroyed */
  0, /* todo_flags_start */
  TODO_verify_ssa, /* todo_flags_finish */
};

class pass_loop_distribution : public gimple_opt_pass
{
public:
  pass_loop_distribution (gcc::context *ctxt)
    : gimple_opt_pass (pass_data_loop_distribution, ctxt)
  {}

  /* opt_pass methods: */
  bool gate () { return gate_tree_loop_distribution (); }
  unsigned int execute () { return tree_loop_distribution (); }

}; // class pass_loop_distribution

} // anon namespace

gimple_opt_pass *
make_pass_loop_distribution (gcc::context *ctxt)
{
  return new pass_loop_distribution (ctxt);
}