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
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
|
/* Inlining decision heuristics.
Copyright (C) 2003-2018 Free Software Foundation, Inc.
Contributed by Jan Hubicka
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/>. */
/* Inlining decision heuristics
The implementation of inliner is organized as follows:
inlining heuristics limits
can_inline_edge_p allow to check that particular inlining is allowed
by the limits specified by user (allowed function growth, growth and so
on).
Functions are inlined when it is obvious the result is profitable (such
as functions called once or when inlining reduce code size).
In addition to that we perform inlining of small functions and recursive
inlining.
inlining heuristics
The inliner itself is split into two passes:
pass_early_inlining
Simple local inlining pass inlining callees into current function.
This pass makes no use of whole unit analysis and thus it can do only
very simple decisions based on local properties.
The strength of the pass is that it is run in topological order
(reverse postorder) on the callgraph. Functions are converted into SSA
form just before this pass and optimized subsequently. As a result, the
callees of the function seen by the early inliner was already optimized
and results of early inlining adds a lot of optimization opportunities
for the local optimization.
The pass handle the obvious inlining decisions within the compilation
unit - inlining auto inline functions, inlining for size and
flattening.
main strength of the pass is the ability to eliminate abstraction
penalty in C++ code (via combination of inlining and early
optimization) and thus improve quality of analysis done by real IPA
optimizers.
Because of lack of whole unit knowledge, the pass can not really make
good code size/performance tradeoffs. It however does very simple
speculative inlining allowing code size to grow by
EARLY_INLINING_INSNS when callee is leaf function. In this case the
optimizations performed later are very likely to eliminate the cost.
pass_ipa_inline
This is the real inliner able to handle inlining with whole program
knowledge. It performs following steps:
1) inlining of small functions. This is implemented by greedy
algorithm ordering all inlinable cgraph edges by their badness and
inlining them in this order as long as inline limits allows doing so.
This heuristics is not very good on inlining recursive calls. Recursive
calls can be inlined with results similar to loop unrolling. To do so,
special purpose recursive inliner is executed on function when
recursive edge is met as viable candidate.
2) Unreachable functions are removed from callgraph. Inlining leads
to devirtualization and other modification of callgraph so functions
may become unreachable during the process. Also functions declared as
extern inline or virtual functions are removed, since after inlining
we no longer need the offline bodies.
3) Functions called once and not exported from the unit are inlined.
This should almost always lead to reduction of code size by eliminating
the need for offline copy of the function. */
#include "config.h"
#include "system.h"
#include "coretypes.h"
#include "backend.h"
#include "target.h"
#include "rtl.h"
#include "tree.h"
#include "gimple.h"
#include "alloc-pool.h"
#include "tree-pass.h"
#include "gimple-ssa.h"
#include "cgraph.h"
#include "lto-streamer.h"
#include "trans-mem.h"
#include "calls.h"
#include "tree-inline.h"
#include "params.h"
#include "profile.h"
#include "symbol-summary.h"
#include "tree-vrp.h"
#include "ipa-prop.h"
#include "ipa-fnsummary.h"
#include "ipa-inline.h"
#include "ipa-utils.h"
#include "sreal.h"
#include "auto-profile.h"
#include "builtins.h"
#include "fibonacci_heap.h"
#include "stringpool.h"
#include "attribs.h"
#include "asan.h"
typedef fibonacci_heap <sreal, cgraph_edge> edge_heap_t;
typedef fibonacci_node <sreal, cgraph_edge> edge_heap_node_t;
/* Statistics we collect about inlining algorithm. */
static int overall_size;
static profile_count max_count;
static profile_count spec_rem;
/* Return false when inlining edge E would lead to violating
limits on function unit growth or stack usage growth.
The relative function body growth limit is present generally
to avoid problems with non-linear behavior of the compiler.
To allow inlining huge functions into tiny wrapper, the limit
is always based on the bigger of the two functions considered.
For stack growth limits we always base the growth in stack usage
of the callers. We want to prevent applications from segfaulting
on stack overflow when functions with huge stack frames gets
inlined. */
static bool
caller_growth_limits (struct cgraph_edge *e)
{
struct cgraph_node *to = e->caller;
struct cgraph_node *what = e->callee->ultimate_alias_target ();
int newsize;
int limit = 0;
HOST_WIDE_INT stack_size_limit = 0, inlined_stack;
ipa_fn_summary *info, *what_info, *outer_info = ipa_fn_summaries->get (to);
/* Look for function e->caller is inlined to. While doing
so work out the largest function body on the way. As
described above, we want to base our function growth
limits based on that. Not on the self size of the
outer function, not on the self size of inline code
we immediately inline to. This is the most relaxed
interpretation of the rule "do not grow large functions
too much in order to prevent compiler from exploding". */
while (true)
{
info = ipa_fn_summaries->get (to);
if (limit < info->self_size)
limit = info->self_size;
if (stack_size_limit < info->estimated_self_stack_size)
stack_size_limit = info->estimated_self_stack_size;
if (to->global.inlined_to)
to = to->callers->caller;
else
break;
}
what_info = ipa_fn_summaries->get (what);
if (limit < what_info->self_size)
limit = what_info->self_size;
limit += limit * PARAM_VALUE (PARAM_LARGE_FUNCTION_GROWTH) / 100;
/* Check the size after inlining against the function limits. But allow
the function to shrink if it went over the limits by forced inlining. */
newsize = estimate_size_after_inlining (to, e);
if (newsize >= info->size
&& newsize > PARAM_VALUE (PARAM_LARGE_FUNCTION_INSNS)
&& newsize > limit)
{
e->inline_failed = CIF_LARGE_FUNCTION_GROWTH_LIMIT;
return false;
}
if (!what_info->estimated_stack_size)
return true;
/* FIXME: Stack size limit often prevents inlining in Fortran programs
due to large i/o datastructures used by the Fortran front-end.
We ought to ignore this limit when we know that the edge is executed
on every invocation of the caller (i.e. its call statement dominates
exit block). We do not track this information, yet. */
stack_size_limit += ((gcov_type)stack_size_limit
* PARAM_VALUE (PARAM_STACK_FRAME_GROWTH) / 100);
inlined_stack = (outer_info->stack_frame_offset
+ outer_info->estimated_self_stack_size
+ what_info->estimated_stack_size);
/* Check new stack consumption with stack consumption at the place
stack is used. */
if (inlined_stack > stack_size_limit
/* If function already has large stack usage from sibling
inline call, we can inline, too.
This bit overoptimistically assume that we are good at stack
packing. */
&& inlined_stack > info->estimated_stack_size
&& inlined_stack > PARAM_VALUE (PARAM_LARGE_STACK_FRAME))
{
e->inline_failed = CIF_LARGE_STACK_FRAME_GROWTH_LIMIT;
return false;
}
return true;
}
/* Dump info about why inlining has failed. */
static void
report_inline_failed_reason (struct cgraph_edge *e)
{
if (dump_file)
{
fprintf (dump_file, " not inlinable: %s -> %s, %s\n",
e->caller->dump_name (),
e->callee->dump_name (),
cgraph_inline_failed_string (e->inline_failed));
if ((e->inline_failed == CIF_TARGET_OPTION_MISMATCH
|| e->inline_failed == CIF_OPTIMIZATION_MISMATCH)
&& e->caller->lto_file_data
&& e->callee->ultimate_alias_target ()->lto_file_data)
{
fprintf (dump_file, " LTO objects: %s, %s\n",
e->caller->lto_file_data->file_name,
e->callee->ultimate_alias_target ()->lto_file_data->file_name);
}
if (e->inline_failed == CIF_TARGET_OPTION_MISMATCH)
cl_target_option_print_diff
(dump_file, 2, target_opts_for_fn (e->caller->decl),
target_opts_for_fn (e->callee->ultimate_alias_target ()->decl));
if (e->inline_failed == CIF_OPTIMIZATION_MISMATCH)
cl_optimization_print_diff
(dump_file, 2, opts_for_fn (e->caller->decl),
opts_for_fn (e->callee->ultimate_alias_target ()->decl));
}
}
/* Decide whether sanitizer-related attributes allow inlining. */
static bool
sanitize_attrs_match_for_inline_p (const_tree caller, const_tree callee)
{
if (!caller || !callee)
return true;
return ((sanitize_flags_p (SANITIZE_ADDRESS, caller)
== sanitize_flags_p (SANITIZE_ADDRESS, callee))
&& (sanitize_flags_p (SANITIZE_POINTER_COMPARE, caller)
== sanitize_flags_p (SANITIZE_POINTER_COMPARE, callee))
&& (sanitize_flags_p (SANITIZE_POINTER_SUBTRACT, caller)
== sanitize_flags_p (SANITIZE_POINTER_SUBTRACT, callee)));
}
/* Used for flags where it is safe to inline when caller's value is
grater than callee's. */
#define check_maybe_up(flag) \
(opts_for_fn (caller->decl)->x_##flag \
!= opts_for_fn (callee->decl)->x_##flag \
&& (!always_inline \
|| opts_for_fn (caller->decl)->x_##flag \
< opts_for_fn (callee->decl)->x_##flag))
/* Used for flags where it is safe to inline when caller's value is
smaller than callee's. */
#define check_maybe_down(flag) \
(opts_for_fn (caller->decl)->x_##flag \
!= opts_for_fn (callee->decl)->x_##flag \
&& (!always_inline \
|| opts_for_fn (caller->decl)->x_##flag \
> opts_for_fn (callee->decl)->x_##flag))
/* Used for flags where exact match is needed for correctness. */
#define check_match(flag) \
(opts_for_fn (caller->decl)->x_##flag \
!= opts_for_fn (callee->decl)->x_##flag)
/* Decide if we can inline the edge and possibly update
inline_failed reason.
We check whether inlining is possible at all and whether
caller growth limits allow doing so.
if REPORT is true, output reason to the dump file. */
static bool
can_inline_edge_p (struct cgraph_edge *e, bool report,
bool early = false)
{
gcc_checking_assert (e->inline_failed);
if (cgraph_inline_failed_type (e->inline_failed) == CIF_FINAL_ERROR)
{
if (report)
report_inline_failed_reason (e);
return false;
}
bool inlinable = true;
enum availability avail;
cgraph_node *caller = e->caller->global.inlined_to
? e->caller->global.inlined_to : e->caller;
cgraph_node *callee = e->callee->ultimate_alias_target (&avail, caller);
if (!callee->definition)
{
e->inline_failed = CIF_BODY_NOT_AVAILABLE;
inlinable = false;
}
if (!early && (!opt_for_fn (callee->decl, optimize)
|| !opt_for_fn (caller->decl, optimize)))
{
e->inline_failed = CIF_FUNCTION_NOT_OPTIMIZED;
inlinable = false;
}
else if (callee->calls_comdat_local)
{
e->inline_failed = CIF_USES_COMDAT_LOCAL;
inlinable = false;
}
else if (avail <= AVAIL_INTERPOSABLE)
{
e->inline_failed = CIF_OVERWRITABLE;
inlinable = false;
}
/* All edges with call_stmt_cannot_inline_p should have inline_failed
initialized to one of FINAL_ERROR reasons. */
else if (e->call_stmt_cannot_inline_p)
gcc_unreachable ();
/* Don't inline if the functions have different EH personalities. */
else if (DECL_FUNCTION_PERSONALITY (caller->decl)
&& DECL_FUNCTION_PERSONALITY (callee->decl)
&& (DECL_FUNCTION_PERSONALITY (caller->decl)
!= DECL_FUNCTION_PERSONALITY (callee->decl)))
{
e->inline_failed = CIF_EH_PERSONALITY;
inlinable = false;
}
/* TM pure functions should not be inlined into non-TM_pure
functions. */
else if (is_tm_pure (callee->decl) && !is_tm_pure (caller->decl))
{
e->inline_failed = CIF_UNSPECIFIED;
inlinable = false;
}
/* Check compatibility of target optimization options. */
else if (!targetm.target_option.can_inline_p (caller->decl,
callee->decl))
{
e->inline_failed = CIF_TARGET_OPTION_MISMATCH;
inlinable = false;
}
else if (!ipa_fn_summaries->get (callee)->inlinable)
{
e->inline_failed = CIF_FUNCTION_NOT_INLINABLE;
inlinable = false;
}
/* Don't inline a function with mismatched sanitization attributes. */
else if (!sanitize_attrs_match_for_inline_p (caller->decl, callee->decl))
{
e->inline_failed = CIF_ATTRIBUTE_MISMATCH;
inlinable = false;
}
if (!inlinable && report)
report_inline_failed_reason (e);
return inlinable;
}
/* Decide if we can inline the edge and possibly update
inline_failed reason.
We check whether inlining is possible at all and whether
caller growth limits allow doing so.
if REPORT is true, output reason to the dump file.
if DISREGARD_LIMITS is true, ignore size limits. */
static bool
can_inline_edge_by_limits_p (struct cgraph_edge *e, bool report,
bool disregard_limits = false, bool early = false)
{
gcc_checking_assert (e->inline_failed);
if (cgraph_inline_failed_type (e->inline_failed) == CIF_FINAL_ERROR)
{
if (report)
report_inline_failed_reason (e);
return false;
}
bool inlinable = true;
enum availability avail;
cgraph_node *caller = e->caller->global.inlined_to
? e->caller->global.inlined_to : e->caller;
cgraph_node *callee = e->callee->ultimate_alias_target (&avail, caller);
tree caller_tree = DECL_FUNCTION_SPECIFIC_OPTIMIZATION (caller->decl);
tree callee_tree
= callee ? DECL_FUNCTION_SPECIFIC_OPTIMIZATION (callee->decl) : NULL;
/* Check if caller growth allows the inlining. */
if (!DECL_DISREGARD_INLINE_LIMITS (callee->decl)
&& !disregard_limits
&& !lookup_attribute ("flatten",
DECL_ATTRIBUTES (caller->decl))
&& !caller_growth_limits (e))
inlinable = false;
/* Don't inline a function with a higher optimization level than the
caller. FIXME: this is really just tip of iceberg of handling
optimization attribute. */
else if (caller_tree != callee_tree)
{
bool always_inline =
(DECL_DISREGARD_INLINE_LIMITS (callee->decl)
&& lookup_attribute ("always_inline",
DECL_ATTRIBUTES (callee->decl)));
ipa_fn_summary *caller_info = ipa_fn_summaries->get (caller);
ipa_fn_summary *callee_info = ipa_fn_summaries->get (callee);
/* Until GCC 4.9 we did not check the semantics alterning flags
bellow and inline across optimization boundry.
Enabling checks bellow breaks several packages by refusing
to inline library always_inline functions. See PR65873.
Disable the check for early inlining for now until better solution
is found. */
if (always_inline && early)
;
/* There are some options that change IL semantics which means
we cannot inline in these cases for correctness reason.
Not even for always_inline declared functions. */
else if (check_match (flag_wrapv)
|| check_match (flag_trapv)
|| check_match (flag_pcc_struct_return)
/* When caller or callee does FP math, be sure FP codegen flags
compatible. */
|| ((caller_info->fp_expressions && callee_info->fp_expressions)
&& (check_maybe_up (flag_rounding_math)
|| check_maybe_up (flag_trapping_math)
|| check_maybe_down (flag_unsafe_math_optimizations)
|| check_maybe_down (flag_finite_math_only)
|| check_maybe_up (flag_signaling_nans)
|| check_maybe_down (flag_cx_limited_range)
|| check_maybe_up (flag_signed_zeros)
|| check_maybe_down (flag_associative_math)
|| check_maybe_down (flag_reciprocal_math)
|| check_maybe_down (flag_fp_int_builtin_inexact)
/* Strictly speaking only when the callee contains function
calls that may end up setting errno. */
|| check_maybe_up (flag_errno_math)))
/* We do not want to make code compiled with exceptions to be
brought into a non-EH function unless we know that the callee
does not throw.
This is tracked by DECL_FUNCTION_PERSONALITY. */
|| (check_maybe_up (flag_non_call_exceptions)
&& DECL_FUNCTION_PERSONALITY (callee->decl))
|| (check_maybe_up (flag_exceptions)
&& DECL_FUNCTION_PERSONALITY (callee->decl))
/* When devirtualization is diabled for callee, it is not safe
to inline it as we possibly mangled the type info.
Allow early inlining of always inlines. */
|| (!early && check_maybe_down (flag_devirtualize)))
{
e->inline_failed = CIF_OPTIMIZATION_MISMATCH;
inlinable = false;
}
/* gcc.dg/pr43564.c. Apply user-forced inline even at -O0. */
else if (always_inline)
;
/* When user added an attribute to the callee honor it. */
else if (lookup_attribute ("optimize", DECL_ATTRIBUTES (callee->decl))
&& opts_for_fn (caller->decl) != opts_for_fn (callee->decl))
{
e->inline_failed = CIF_OPTIMIZATION_MISMATCH;
inlinable = false;
}
/* If explicit optimize attribute are not used, the mismatch is caused
by different command line options used to build different units.
Do not care about COMDAT functions - those are intended to be
optimized with the optimization flags of module they are used in.
Also do not care about mixing up size/speed optimization when
DECL_DISREGARD_INLINE_LIMITS is set. */
else if ((callee->merged_comdat
&& !lookup_attribute ("optimize",
DECL_ATTRIBUTES (caller->decl)))
|| DECL_DISREGARD_INLINE_LIMITS (callee->decl))
;
/* If mismatch is caused by merging two LTO units with different
optimizationflags we want to be bit nicer. However never inline
if one of functions is not optimized at all. */
else if (!opt_for_fn (callee->decl, optimize)
|| !opt_for_fn (caller->decl, optimize))
{
e->inline_failed = CIF_OPTIMIZATION_MISMATCH;
inlinable = false;
}
/* If callee is optimized for size and caller is not, allow inlining if
code shrinks or we are in MAX_INLINE_INSNS_SINGLE limit and callee
is inline (and thus likely an unified comdat). This will allow caller
to run faster. */
else if (opt_for_fn (callee->decl, optimize_size)
> opt_for_fn (caller->decl, optimize_size))
{
int growth = estimate_edge_growth (e);
if (growth > 0
&& (!DECL_DECLARED_INLINE_P (callee->decl)
&& growth >= MAX (MAX_INLINE_INSNS_SINGLE,
MAX_INLINE_INSNS_AUTO)))
{
e->inline_failed = CIF_OPTIMIZATION_MISMATCH;
inlinable = false;
}
}
/* If callee is more aggressively optimized for performance than caller,
we generally want to inline only cheap (runtime wise) functions. */
else if (opt_for_fn (callee->decl, optimize_size)
< opt_for_fn (caller->decl, optimize_size)
|| (opt_for_fn (callee->decl, optimize)
> opt_for_fn (caller->decl, optimize)))
{
if (estimate_edge_time (e)
>= 20 + ipa_call_summaries->get (e)->call_stmt_time)
{
e->inline_failed = CIF_OPTIMIZATION_MISMATCH;
inlinable = false;
}
}
}
if (!inlinable && report)
report_inline_failed_reason (e);
return inlinable;
}
/* Return true if the edge E is inlinable during early inlining. */
static bool
can_early_inline_edge_p (struct cgraph_edge *e)
{
struct cgraph_node *callee = e->callee->ultimate_alias_target ();
/* Early inliner might get called at WPA stage when IPA pass adds new
function. In this case we can not really do any of early inlining
because function bodies are missing. */
if (cgraph_inline_failed_type (e->inline_failed) == CIF_FINAL_ERROR)
return false;
if (!gimple_has_body_p (callee->decl))
{
e->inline_failed = CIF_BODY_NOT_AVAILABLE;
return false;
}
/* In early inliner some of callees may not be in SSA form yet
(i.e. the callgraph is cyclic and we did not process
the callee by early inliner, yet). We don't have CIF code for this
case; later we will re-do the decision in the real inliner. */
if (!gimple_in_ssa_p (DECL_STRUCT_FUNCTION (e->caller->decl))
|| !gimple_in_ssa_p (DECL_STRUCT_FUNCTION (callee->decl)))
{
if (dump_file)
fprintf (dump_file, " edge not inlinable: not in SSA form\n");
return false;
}
if (!can_inline_edge_p (e, true, true)
|| !can_inline_edge_by_limits_p (e, true, false, true))
return false;
return true;
}
/* Return number of calls in N. Ignore cheap builtins. */
static int
num_calls (struct cgraph_node *n)
{
struct cgraph_edge *e;
int num = 0;
for (e = n->callees; e; e = e->next_callee)
if (!is_inexpensive_builtin (e->callee->decl))
num++;
return num;
}
/* Return true if we are interested in inlining small function. */
static bool
want_early_inline_function_p (struct cgraph_edge *e)
{
bool want_inline = true;
struct cgraph_node *callee = e->callee->ultimate_alias_target ();
if (DECL_DISREGARD_INLINE_LIMITS (callee->decl))
;
/* For AutoFDO, we need to make sure that before profile summary, all
hot paths' IR look exactly the same as profiled binary. As a result,
in einliner, we will disregard size limit and inline those callsites
that are:
* inlined in the profiled binary, and
* the cloned callee has enough samples to be considered "hot". */
else if (flag_auto_profile && afdo_callsite_hot_enough_for_early_inline (e))
;
else if (!DECL_DECLARED_INLINE_P (callee->decl)
&& !opt_for_fn (e->caller->decl, flag_inline_small_functions))
{
e->inline_failed = CIF_FUNCTION_NOT_INLINE_CANDIDATE;
report_inline_failed_reason (e);
want_inline = false;
}
else
{
int growth = estimate_edge_growth (e);
int n;
if (growth <= 0)
;
else if (!e->maybe_hot_p ()
&& growth > 0)
{
if (dump_file)
fprintf (dump_file, " will not early inline: %s->%s, "
"call is cold and code would grow by %i\n",
e->caller->dump_name (),
callee->dump_name (),
growth);
want_inline = false;
}
else if (growth > PARAM_VALUE (PARAM_EARLY_INLINING_INSNS))
{
if (dump_file)
fprintf (dump_file, " will not early inline: %s->%s, "
"growth %i exceeds --param early-inlining-insns\n",
e->caller->dump_name (),
callee->dump_name (),
growth);
want_inline = false;
}
else if ((n = num_calls (callee)) != 0
&& growth * (n + 1) > PARAM_VALUE (PARAM_EARLY_INLINING_INSNS))
{
if (dump_file)
fprintf (dump_file, " will not early inline: %s->%s, "
"growth %i exceeds --param early-inlining-insns "
"divided by number of calls\n",
e->caller->dump_name (),
callee->dump_name (),
growth);
want_inline = false;
}
}
return want_inline;
}
/* Compute time of the edge->caller + edge->callee execution when inlining
does not happen. */
inline sreal
compute_uninlined_call_time (struct cgraph_edge *edge,
sreal uninlined_call_time)
{
cgraph_node *caller = (edge->caller->global.inlined_to
? edge->caller->global.inlined_to
: edge->caller);
sreal freq = edge->sreal_frequency ();
if (freq > 0)
uninlined_call_time *= freq;
else
uninlined_call_time = uninlined_call_time >> 11;
sreal caller_time = ipa_fn_summaries->get (caller)->time;
return uninlined_call_time + caller_time;
}
/* Same as compute_uinlined_call_time but compute time when inlining
does happen. */
inline sreal
compute_inlined_call_time (struct cgraph_edge *edge,
sreal time)
{
cgraph_node *caller = (edge->caller->global.inlined_to
? edge->caller->global.inlined_to
: edge->caller);
sreal caller_time = ipa_fn_summaries->get (caller)->time;
sreal freq = edge->sreal_frequency ();
if (freq > 0)
time *= freq;
else
time = time >> 11;
/* This calculation should match one in ipa-inline-analysis.c
(estimate_edge_size_and_time). */
time -= (sreal)ipa_call_summaries->get (edge)->call_stmt_time * freq;
time += caller_time;
if (time <= 0)
time = ((sreal) 1) >> 8;
gcc_checking_assert (time >= 0);
return time;
}
/* Return true if the speedup for inlining E is bigger than
PARAM_MAX_INLINE_MIN_SPEEDUP. */
static bool
big_speedup_p (struct cgraph_edge *e)
{
sreal unspec_time;
sreal spec_time = estimate_edge_time (e, &unspec_time);
sreal time = compute_uninlined_call_time (e, unspec_time);
sreal inlined_time = compute_inlined_call_time (e, spec_time);
if ((time - inlined_time) * 100
> (sreal) (time * PARAM_VALUE (PARAM_INLINE_MIN_SPEEDUP)))
return true;
return false;
}
/* Return true if we are interested in inlining small function.
When REPORT is true, report reason to dump file. */
static bool
want_inline_small_function_p (struct cgraph_edge *e, bool report)
{
bool want_inline = true;
struct cgraph_node *callee = e->callee->ultimate_alias_target ();
/* Allow this function to be called before can_inline_edge_p,
since it's usually cheaper. */
if (cgraph_inline_failed_type (e->inline_failed) == CIF_FINAL_ERROR)
want_inline = false;
else if (DECL_DISREGARD_INLINE_LIMITS (callee->decl))
;
else if (!DECL_DECLARED_INLINE_P (callee->decl)
&& !opt_for_fn (e->caller->decl, flag_inline_small_functions))
{
e->inline_failed = CIF_FUNCTION_NOT_INLINE_CANDIDATE;
want_inline = false;
}
/* Do fast and conservative check if the function can be good
inline candidate. At the moment we allow inline hints to
promote non-inline functions to inline and we increase
MAX_INLINE_INSNS_SINGLE 16-fold for inline functions. */
else if ((!DECL_DECLARED_INLINE_P (callee->decl)
&& (!e->count.ipa ().initialized_p () || !e->maybe_hot_p ()))
&& ipa_fn_summaries->get (callee)->min_size
- ipa_call_summaries->get (e)->call_stmt_size
> MAX (MAX_INLINE_INSNS_SINGLE, MAX_INLINE_INSNS_AUTO))
{
e->inline_failed = CIF_MAX_INLINE_INSNS_AUTO_LIMIT;
want_inline = false;
}
else if ((DECL_DECLARED_INLINE_P (callee->decl)
|| e->count.ipa ().nonzero_p ())
&& ipa_fn_summaries->get (callee)->min_size
- ipa_call_summaries->get (e)->call_stmt_size
> 16 * MAX_INLINE_INSNS_SINGLE)
{
e->inline_failed = (DECL_DECLARED_INLINE_P (callee->decl)
? CIF_MAX_INLINE_INSNS_SINGLE_LIMIT
: CIF_MAX_INLINE_INSNS_AUTO_LIMIT);
want_inline = false;
}
else
{
int growth = estimate_edge_growth (e);
ipa_hints hints = estimate_edge_hints (e);
bool big_speedup = big_speedup_p (e);
if (growth <= 0)
;
/* Apply MAX_INLINE_INSNS_SINGLE limit. Do not do so when
hints suggests that inlining given function is very profitable. */
else if (DECL_DECLARED_INLINE_P (callee->decl)
&& growth >= MAX_INLINE_INSNS_SINGLE
&& ((!big_speedup
&& !(hints & (INLINE_HINT_indirect_call
| INLINE_HINT_known_hot
| INLINE_HINT_loop_iterations
| INLINE_HINT_array_index
| INLINE_HINT_loop_stride)))
|| growth >= MAX_INLINE_INSNS_SINGLE * 16))
{
e->inline_failed = CIF_MAX_INLINE_INSNS_SINGLE_LIMIT;
want_inline = false;
}
else if (!DECL_DECLARED_INLINE_P (callee->decl)
&& !opt_for_fn (e->caller->decl, flag_inline_functions))
{
/* growth_likely_positive is expensive, always test it last. */
if (growth >= MAX_INLINE_INSNS_SINGLE
|| growth_likely_positive (callee, growth))
{
e->inline_failed = CIF_NOT_DECLARED_INLINED;
want_inline = false;
}
}
/* Apply MAX_INLINE_INSNS_AUTO limit for functions not declared inline
Upgrade it to MAX_INLINE_INSNS_SINGLE when hints suggests that
inlining given function is very profitable. */
else if (!DECL_DECLARED_INLINE_P (callee->decl)
&& !big_speedup
&& !(hints & INLINE_HINT_known_hot)
&& growth >= ((hints & (INLINE_HINT_indirect_call
| INLINE_HINT_loop_iterations
| INLINE_HINT_array_index
| INLINE_HINT_loop_stride))
? MAX (MAX_INLINE_INSNS_AUTO,
MAX_INLINE_INSNS_SINGLE)
: MAX_INLINE_INSNS_AUTO))
{
/* growth_likely_positive is expensive, always test it last. */
if (growth >= MAX_INLINE_INSNS_SINGLE
|| growth_likely_positive (callee, growth))
{
e->inline_failed = CIF_MAX_INLINE_INSNS_AUTO_LIMIT;
want_inline = false;
}
}
/* If call is cold, do not inline when function body would grow. */
else if (!e->maybe_hot_p ()
&& (growth >= MAX_INLINE_INSNS_SINGLE
|| growth_likely_positive (callee, growth)))
{
e->inline_failed = CIF_UNLIKELY_CALL;
want_inline = false;
}
}
if (!want_inline && report)
report_inline_failed_reason (e);
return want_inline;
}
/* EDGE is self recursive edge.
We hand two cases - when function A is inlining into itself
or when function A is being inlined into another inliner copy of function
A within function B.
In first case OUTER_NODE points to the toplevel copy of A, while
in the second case OUTER_NODE points to the outermost copy of A in B.
In both cases we want to be extra selective since
inlining the call will just introduce new recursive calls to appear. */
static bool
want_inline_self_recursive_call_p (struct cgraph_edge *edge,
struct cgraph_node *outer_node,
bool peeling,
int depth)
{
char const *reason = NULL;
bool want_inline = true;
sreal caller_freq = 1;
int max_depth = PARAM_VALUE (PARAM_MAX_INLINE_RECURSIVE_DEPTH_AUTO);
if (DECL_DECLARED_INLINE_P (edge->caller->decl))
max_depth = PARAM_VALUE (PARAM_MAX_INLINE_RECURSIVE_DEPTH);
if (!edge->maybe_hot_p ())
{
reason = "recursive call is cold";
want_inline = false;
}
else if (depth > max_depth)
{
reason = "--param max-inline-recursive-depth exceeded.";
want_inline = false;
}
else if (outer_node->global.inlined_to
&& (caller_freq = outer_node->callers->sreal_frequency ()) == 0)
{
reason = "caller frequency is 0";
want_inline = false;
}
if (!want_inline)
;
/* Inlining of self recursive function into copy of itself within other
function is transformation similar to loop peeling.
Peeling is profitable if we can inline enough copies to make probability
of actual call to the self recursive function very small. Be sure that
the probability of recursion is small.
We ensure that the frequency of recursing is at most 1 - (1/max_depth).
This way the expected number of recursion is at most max_depth. */
else if (peeling)
{
sreal max_prob = (sreal)1 - ((sreal)1 / (sreal)max_depth);
int i;
for (i = 1; i < depth; i++)
max_prob = max_prob * max_prob;
if (edge->sreal_frequency () >= max_prob * caller_freq)
{
reason = "frequency of recursive call is too large";
want_inline = false;
}
}
/* Recursive inlining, i.e. equivalent of unrolling, is profitable if
recursion depth is large. We reduce function call overhead and increase
chances that things fit in hardware return predictor.
Recursive inlining might however increase cost of stack frame setup
actually slowing down functions whose recursion tree is wide rather than
deep.
Deciding reliably on when to do recursive inlining without profile feedback
is tricky. For now we disable recursive inlining when probability of self
recursion is low.
Recursive inlining of self recursive call within loop also results in
large loop depths that generally optimize badly. We may want to throttle
down inlining in those cases. In particular this seems to happen in one
of libstdc++ rb tree methods. */
else
{
if (edge->sreal_frequency () * 100
<= caller_freq
* PARAM_VALUE (PARAM_MIN_INLINE_RECURSIVE_PROBABILITY))
{
reason = "frequency of recursive call is too small";
want_inline = false;
}
}
if (!want_inline && dump_file)
fprintf (dump_file, " not inlining recursively: %s\n", reason);
return want_inline;
}
/* Return true when NODE has uninlinable caller;
set HAS_HOT_CALL if it has hot call.
Worker for cgraph_for_node_and_aliases. */
static bool
check_callers (struct cgraph_node *node, void *has_hot_call)
{
struct cgraph_edge *e;
for (e = node->callers; e; e = e->next_caller)
{
if (!opt_for_fn (e->caller->decl, flag_inline_functions_called_once)
|| !opt_for_fn (e->caller->decl, optimize))
return true;
if (!can_inline_edge_p (e, true))
return true;
if (e->recursive_p ())
return true;
if (!can_inline_edge_by_limits_p (e, true))
return true;
if (!(*(bool *)has_hot_call) && e->maybe_hot_p ())
*(bool *)has_hot_call = true;
}
return false;
}
/* If NODE has a caller, return true. */
static bool
has_caller_p (struct cgraph_node *node, void *data ATTRIBUTE_UNUSED)
{
if (node->callers)
return true;
return false;
}
/* Decide if inlining NODE would reduce unit size by eliminating
the offline copy of function.
When COLD is true the cold calls are considered, too. */
static bool
want_inline_function_to_all_callers_p (struct cgraph_node *node, bool cold)
{
bool has_hot_call = false;
/* Aliases gets inlined along with the function they alias. */
if (node->alias)
return false;
/* Already inlined? */
if (node->global.inlined_to)
return false;
/* Does it have callers? */
if (!node->call_for_symbol_and_aliases (has_caller_p, NULL, true))
return false;
/* Inlining into all callers would increase size? */
if (estimate_growth (node) > 0)
return false;
/* All inlines must be possible. */
if (node->call_for_symbol_and_aliases (check_callers, &has_hot_call,
true))
return false;
if (!cold && !has_hot_call)
return false;
return true;
}
/* A cost model driving the inlining heuristics in a way so the edges with
smallest badness are inlined first. After each inlining is performed
the costs of all caller edges of nodes affected are recomputed so the
metrics may accurately depend on values such as number of inlinable callers
of the function or function body size. */
static sreal
edge_badness (struct cgraph_edge *edge, bool dump)
{
sreal badness;
int growth;
sreal edge_time, unspec_edge_time;
struct cgraph_node *callee = edge->callee->ultimate_alias_target ();
struct ipa_fn_summary *callee_info = ipa_fn_summaries->get (callee);
ipa_hints hints;
cgraph_node *caller = (edge->caller->global.inlined_to
? edge->caller->global.inlined_to
: edge->caller);
growth = estimate_edge_growth (edge);
edge_time = estimate_edge_time (edge, &unspec_edge_time);
hints = estimate_edge_hints (edge);
gcc_checking_assert (edge_time >= 0);
/* Check that inlined time is better, but tolerate some roundoff issues.
FIXME: When callee profile drops to 0 we account calls more. This
should be fixed by never doing that. */
gcc_checking_assert ((edge_time * 100
- callee_info->time * 101).to_int () <= 0
|| callee->count.ipa ().initialized_p ());
gcc_checking_assert (growth <= callee_info->size);
if (dump)
{
fprintf (dump_file, " Badness calculation for %s -> %s\n",
edge->caller->dump_name (),
edge->callee->dump_name ());
fprintf (dump_file, " size growth %i, time %f unspec %f ",
growth,
edge_time.to_double (),
unspec_edge_time.to_double ());
ipa_dump_hints (dump_file, hints);
if (big_speedup_p (edge))
fprintf (dump_file, " big_speedup");
fprintf (dump_file, "\n");
}
/* Always prefer inlining saving code size. */
if (growth <= 0)
{
badness = (sreal) (-SREAL_MIN_SIG + growth) << (SREAL_MAX_EXP / 256);
if (dump)
fprintf (dump_file, " %f: Growth %d <= 0\n", badness.to_double (),
growth);
}
/* Inlining into EXTERNAL functions is not going to change anything unless
they are themselves inlined. */
else if (DECL_EXTERNAL (caller->decl))
{
if (dump)
fprintf (dump_file, " max: function is external\n");
return sreal::max ();
}
/* When profile is available. Compute badness as:
time_saved * caller_count
goodness = -------------------------------------------------
growth_of_caller * overall_growth * combined_size
badness = - goodness
Again use negative value to make calls with profile appear hotter
then calls without.
*/
else if (opt_for_fn (caller->decl, flag_guess_branch_prob)
|| caller->count.ipa ().nonzero_p ())
{
sreal numerator, denominator;
int overall_growth;
sreal inlined_time = compute_inlined_call_time (edge, edge_time);
numerator = (compute_uninlined_call_time (edge, unspec_edge_time)
- inlined_time);
if (numerator <= 0)
numerator = ((sreal) 1 >> 8);
if (caller->count.ipa ().nonzero_p ())
numerator *= caller->count.ipa ().to_gcov_type ();
else if (caller->count.ipa ().initialized_p ())
numerator = numerator >> 11;
denominator = growth;
overall_growth = callee_info->growth;
/* Look for inliner wrappers of the form:
inline_caller ()
{
do_fast_job...
if (need_more_work)
noninline_callee ();
}
Withhout panilizing this case, we usually inline noninline_callee
into the inline_caller because overall_growth is small preventing
further inlining of inline_caller.
Penalize only callgraph edges to functions with small overall
growth ...
*/
if (growth > overall_growth
/* ... and having only one caller which is not inlined ... */
&& callee_info->single_caller
&& !edge->caller->global.inlined_to
/* ... and edges executed only conditionally ... */
&& edge->sreal_frequency () < 1
/* ... consider case where callee is not inline but caller is ... */
&& ((!DECL_DECLARED_INLINE_P (edge->callee->decl)
&& DECL_DECLARED_INLINE_P (caller->decl))
/* ... or when early optimizers decided to split and edge
frequency still indicates splitting is a win ... */
|| (callee->split_part && !caller->split_part
&& edge->sreal_frequency () * 100
< PARAM_VALUE
(PARAM_PARTIAL_INLINING_ENTRY_PROBABILITY)
/* ... and do not overwrite user specified hints. */
&& (!DECL_DECLARED_INLINE_P (edge->callee->decl)
|| DECL_DECLARED_INLINE_P (caller->decl)))))
{
struct ipa_fn_summary *caller_info = ipa_fn_summaries->get (caller);
int caller_growth = caller_info->growth;
/* Only apply the penalty when caller looks like inline candidate,
and it is not called once and. */
if (!caller_info->single_caller && overall_growth < caller_growth
&& caller_info->inlinable
&& caller_info->size
< (DECL_DECLARED_INLINE_P (caller->decl)
? MAX_INLINE_INSNS_SINGLE : MAX_INLINE_INSNS_AUTO))
{
if (dump)
fprintf (dump_file,
" Wrapper penalty. Increasing growth %i to %i\n",
overall_growth, caller_growth);
overall_growth = caller_growth;
}
}
if (overall_growth > 0)
{
/* Strongly preffer functions with few callers that can be inlined
fully. The square root here leads to smaller binaries at average.
Watch however for extreme cases and return to linear function
when growth is large. */
if (overall_growth < 256)
overall_growth *= overall_growth;
else
overall_growth += 256 * 256 - 256;
denominator *= overall_growth;
}
denominator *= inlined_time;
badness = - numerator / denominator;
if (dump)
{
fprintf (dump_file,
" %f: guessed profile. frequency %f, count %" PRId64
" caller count %" PRId64
" time w/o inlining %f, time with inlining %f"
" overall growth %i (current) %i (original)"
" %i (compensated)\n",
badness.to_double (),
edge->sreal_frequency ().to_double (),
edge->count.ipa ().initialized_p () ? edge->count.ipa ().to_gcov_type () : -1,
caller->count.ipa ().initialized_p () ? caller->count.ipa ().to_gcov_type () : -1,
compute_uninlined_call_time (edge,
unspec_edge_time).to_double (),
inlined_time.to_double (),
estimate_growth (callee),
callee_info->growth, overall_growth);
}
}
/* When function local profile is not available or it does not give
useful information (ie frequency is zero), base the cost on
loop nest and overall size growth, so we optimize for overall number
of functions fully inlined in program. */
else
{
int nest = MIN (ipa_call_summaries->get (edge)->loop_depth, 8);
badness = growth;
/* Decrease badness if call is nested. */
if (badness > 0)
badness = badness >> nest;
else
badness = badness << nest;
if (dump)
fprintf (dump_file, " %f: no profile. nest %i\n",
badness.to_double (), nest);
}
gcc_checking_assert (badness != 0);
if (edge->recursive_p ())
badness = badness.shift (badness > 0 ? 4 : -4);
if ((hints & (INLINE_HINT_indirect_call
| INLINE_HINT_loop_iterations
| INLINE_HINT_array_index
| INLINE_HINT_loop_stride))
|| callee_info->growth <= 0)
badness = badness.shift (badness > 0 ? -2 : 2);
if (hints & (INLINE_HINT_same_scc))
badness = badness.shift (badness > 0 ? 3 : -3);
else if (hints & (INLINE_HINT_in_scc))
badness = badness.shift (badness > 0 ? 2 : -2);
else if (hints & (INLINE_HINT_cross_module))
badness = badness.shift (badness > 0 ? 1 : -1);
if (DECL_DISREGARD_INLINE_LIMITS (callee->decl))
badness = badness.shift (badness > 0 ? -4 : 4);
else if ((hints & INLINE_HINT_declared_inline))
badness = badness.shift (badness > 0 ? -3 : 3);
if (dump)
fprintf (dump_file, " Adjusted by hints %f\n", badness.to_double ());
return badness;
}
/* Recompute badness of EDGE and update its key in HEAP if needed. */
static inline void
update_edge_key (edge_heap_t *heap, struct cgraph_edge *edge)
{
sreal badness = edge_badness (edge, false);
if (edge->aux)
{
edge_heap_node_t *n = (edge_heap_node_t *) edge->aux;
gcc_checking_assert (n->get_data () == edge);
/* fibonacci_heap::replace_key does busy updating of the
heap that is unnecesarily expensive.
We do lazy increases: after extracting minimum if the key
turns out to be out of date, it is re-inserted into heap
with correct value. */
if (badness < n->get_key ())
{
if (dump_file && (dump_flags & TDF_DETAILS))
{
fprintf (dump_file,
" decreasing badness %s -> %s, %f to %f\n",
edge->caller->dump_name (),
edge->callee->dump_name (),
n->get_key ().to_double (),
badness.to_double ());
}
heap->decrease_key (n, badness);
}
}
else
{
if (dump_file && (dump_flags & TDF_DETAILS))
{
fprintf (dump_file,
" enqueuing call %s -> %s, badness %f\n",
edge->caller->dump_name (),
edge->callee->dump_name (),
badness.to_double ());
}
edge->aux = heap->insert (badness, edge);
}
}
/* NODE was inlined.
All caller edges needs to be resetted because
size estimates change. Similarly callees needs reset
because better context may be known. */
static void
reset_edge_caches (struct cgraph_node *node)
{
struct cgraph_edge *edge;
struct cgraph_edge *e = node->callees;
struct cgraph_node *where = node;
struct ipa_ref *ref;
if (where->global.inlined_to)
where = where->global.inlined_to;
for (edge = where->callers; edge; edge = edge->next_caller)
if (edge->inline_failed)
reset_edge_growth_cache (edge);
FOR_EACH_ALIAS (where, ref)
reset_edge_caches (dyn_cast <cgraph_node *> (ref->referring));
if (!e)
return;
while (true)
if (!e->inline_failed && e->callee->callees)
e = e->callee->callees;
else
{
if (e->inline_failed)
reset_edge_growth_cache (e);
if (e->next_callee)
e = e->next_callee;
else
{
do
{
if (e->caller == node)
return;
e = e->caller->callers;
}
while (!e->next_callee);
e = e->next_callee;
}
}
}
/* Recompute HEAP nodes for each of caller of NODE.
UPDATED_NODES track nodes we already visited, to avoid redundant work.
When CHECK_INLINABLITY_FOR is set, re-check for specified edge that
it is inlinable. Otherwise check all edges. */
static void
update_caller_keys (edge_heap_t *heap, struct cgraph_node *node,
bitmap updated_nodes,
struct cgraph_edge *check_inlinablity_for)
{
struct cgraph_edge *edge;
struct ipa_ref *ref;
if ((!node->alias && !ipa_fn_summaries->get (node)->inlinable)
|| node->global.inlined_to)
return;
if (!bitmap_set_bit (updated_nodes, node->uid))
return;
FOR_EACH_ALIAS (node, ref)
{
struct cgraph_node *alias = dyn_cast <cgraph_node *> (ref->referring);
update_caller_keys (heap, alias, updated_nodes, check_inlinablity_for);
}
for (edge = node->callers; edge; edge = edge->next_caller)
if (edge->inline_failed)
{
if (!check_inlinablity_for
|| check_inlinablity_for == edge)
{
if (can_inline_edge_p (edge, false)
&& want_inline_small_function_p (edge, false)
&& can_inline_edge_by_limits_p (edge, false))
update_edge_key (heap, edge);
else if (edge->aux)
{
report_inline_failed_reason (edge);
heap->delete_node ((edge_heap_node_t *) edge->aux);
edge->aux = NULL;
}
}
else if (edge->aux)
update_edge_key (heap, edge);
}
}
/* Recompute HEAP nodes for each uninlined call in NODE.
This is used when we know that edge badnesses are going only to increase
(we introduced new call site) and thus all we need is to insert newly
created edges into heap. */
static void
update_callee_keys (edge_heap_t *heap, struct cgraph_node *node,
bitmap updated_nodes)
{
struct cgraph_edge *e = node->callees;
if (!e)
return;
while (true)
if (!e->inline_failed && e->callee->callees)
e = e->callee->callees;
else
{
enum availability avail;
struct cgraph_node *callee;
/* We do not reset callee growth cache here. Since we added a new call,
growth chould have just increased and consequentely badness metric
don't need updating. */
if (e->inline_failed
&& (callee = e->callee->ultimate_alias_target (&avail, e->caller))
&& ipa_fn_summaries->get (callee)->inlinable
&& avail >= AVAIL_AVAILABLE
&& !bitmap_bit_p (updated_nodes, callee->uid))
{
if (can_inline_edge_p (e, false)
&& want_inline_small_function_p (e, false)
&& can_inline_edge_by_limits_p (e, false))
update_edge_key (heap, e);
else if (e->aux)
{
report_inline_failed_reason (e);
heap->delete_node ((edge_heap_node_t *) e->aux);
e->aux = NULL;
}
}
if (e->next_callee)
e = e->next_callee;
else
{
do
{
if (e->caller == node)
return;
e = e->caller->callers;
}
while (!e->next_callee);
e = e->next_callee;
}
}
}
/* Enqueue all recursive calls from NODE into priority queue depending on
how likely we want to recursively inline the call. */
static void
lookup_recursive_calls (struct cgraph_node *node, struct cgraph_node *where,
edge_heap_t *heap)
{
struct cgraph_edge *e;
enum availability avail;
for (e = where->callees; e; e = e->next_callee)
if (e->callee == node
|| (e->callee->ultimate_alias_target (&avail, e->caller) == node
&& avail > AVAIL_INTERPOSABLE))
heap->insert (-e->sreal_frequency (), e);
for (e = where->callees; e; e = e->next_callee)
if (!e->inline_failed)
lookup_recursive_calls (node, e->callee, heap);
}
/* Decide on recursive inlining: in the case function has recursive calls,
inline until body size reaches given argument. If any new indirect edges
are discovered in the process, add them to *NEW_EDGES, unless NEW_EDGES
is NULL. */
static bool
recursive_inlining (struct cgraph_edge *edge,
vec<cgraph_edge *> *new_edges)
{
int limit = PARAM_VALUE (PARAM_MAX_INLINE_INSNS_RECURSIVE_AUTO);
edge_heap_t heap (sreal::min ());
struct cgraph_node *node;
struct cgraph_edge *e;
struct cgraph_node *master_clone = NULL, *next;
int depth = 0;
int n = 0;
node = edge->caller;
if (node->global.inlined_to)
node = node->global.inlined_to;
if (DECL_DECLARED_INLINE_P (node->decl))
limit = PARAM_VALUE (PARAM_MAX_INLINE_INSNS_RECURSIVE);
/* Make sure that function is small enough to be considered for inlining. */
if (estimate_size_after_inlining (node, edge) >= limit)
return false;
lookup_recursive_calls (node, node, &heap);
if (heap.empty ())
return false;
if (dump_file)
fprintf (dump_file,
" Performing recursive inlining on %s\n",
node->name ());
/* Do the inlining and update list of recursive call during process. */
while (!heap.empty ())
{
struct cgraph_edge *curr = heap.extract_min ();
struct cgraph_node *cnode, *dest = curr->callee;
if (!can_inline_edge_p (curr, true)
|| can_inline_edge_by_limits_p (curr, true))
continue;
/* MASTER_CLONE is produced in the case we already started modified
the function. Be sure to redirect edge to the original body before
estimating growths otherwise we will be seeing growths after inlining
the already modified body. */
if (master_clone)
{
curr->redirect_callee (master_clone);
reset_edge_growth_cache (curr);
}
if (estimate_size_after_inlining (node, curr) > limit)
{
curr->redirect_callee (dest);
reset_edge_growth_cache (curr);
break;
}
depth = 1;
for (cnode = curr->caller;
cnode->global.inlined_to; cnode = cnode->callers->caller)
if (node->decl
== curr->callee->ultimate_alias_target ()->decl)
depth++;
if (!want_inline_self_recursive_call_p (curr, node, false, depth))
{
curr->redirect_callee (dest);
reset_edge_growth_cache (curr);
continue;
}
if (dump_file)
{
fprintf (dump_file,
" Inlining call of depth %i", depth);
if (node->count.nonzero_p ())
{
fprintf (dump_file, " called approx. %.2f times per call",
(double)curr->count.to_gcov_type ()
/ node->count.to_gcov_type ());
}
fprintf (dump_file, "\n");
}
if (!master_clone)
{
/* We need original clone to copy around. */
master_clone = node->create_clone (node->decl, node->count,
false, vNULL, true, NULL, NULL);
for (e = master_clone->callees; e; e = e->next_callee)
if (!e->inline_failed)
clone_inlined_nodes (e, true, false, NULL);
curr->redirect_callee (master_clone);
reset_edge_growth_cache (curr);
}
inline_call (curr, false, new_edges, &overall_size, true);
lookup_recursive_calls (node, curr->callee, &heap);
n++;
}
if (!heap.empty () && dump_file)
fprintf (dump_file, " Recursive inlining growth limit met.\n");
if (!master_clone)
return false;
if (dump_file)
fprintf (dump_file,
"\n Inlined %i times, "
"body grown from size %i to %i, time %f to %f\n", n,
ipa_fn_summaries->get (master_clone)->size,
ipa_fn_summaries->get (node)->size,
ipa_fn_summaries->get (master_clone)->time.to_double (),
ipa_fn_summaries->get (node)->time.to_double ());
/* Remove master clone we used for inlining. We rely that clones inlined
into master clone gets queued just before master clone so we don't
need recursion. */
for (node = symtab->first_function (); node != master_clone;
node = next)
{
next = symtab->next_function (node);
if (node->global.inlined_to == master_clone)
node->remove ();
}
master_clone->remove ();
return true;
}
/* Given whole compilation unit estimate of INSNS, compute how large we can
allow the unit to grow. */
static int
compute_max_insns (int insns)
{
int max_insns = insns;
if (max_insns < PARAM_VALUE (PARAM_LARGE_UNIT_INSNS))
max_insns = PARAM_VALUE (PARAM_LARGE_UNIT_INSNS);
return ((int64_t) max_insns
* (100 + PARAM_VALUE (PARAM_INLINE_UNIT_GROWTH)) / 100);
}
/* Compute badness of all edges in NEW_EDGES and add them to the HEAP. */
static void
add_new_edges_to_heap (edge_heap_t *heap, vec<cgraph_edge *> new_edges)
{
while (new_edges.length () > 0)
{
struct cgraph_edge *edge = new_edges.pop ();
gcc_assert (!edge->aux);
if (edge->inline_failed
&& can_inline_edge_p (edge, true)
&& want_inline_small_function_p (edge, true)
&& can_inline_edge_by_limits_p (edge, true))
edge->aux = heap->insert (edge_badness (edge, false), edge);
}
}
/* Remove EDGE from the fibheap. */
static void
heap_edge_removal_hook (struct cgraph_edge *e, void *data)
{
if (e->aux)
{
((edge_heap_t *)data)->delete_node ((edge_heap_node_t *)e->aux);
e->aux = NULL;
}
}
/* Return true if speculation of edge E seems useful.
If ANTICIPATE_INLINING is true, be conservative and hope that E
may get inlined. */
bool
speculation_useful_p (struct cgraph_edge *e, bool anticipate_inlining)
{
enum availability avail;
struct cgraph_node *target = e->callee->ultimate_alias_target (&avail,
e->caller);
struct cgraph_edge *direct, *indirect;
struct ipa_ref *ref;
gcc_assert (e->speculative && !e->indirect_unknown_callee);
if (!e->maybe_hot_p ())
return false;
/* See if IP optimizations found something potentially useful about the
function. For now we look only for CONST/PURE flags. Almost everything
else we propagate is useless. */
if (avail >= AVAIL_AVAILABLE)
{
int ecf_flags = flags_from_decl_or_type (target->decl);
if (ecf_flags & ECF_CONST)
{
e->speculative_call_info (direct, indirect, ref);
if (!(indirect->indirect_info->ecf_flags & ECF_CONST))
return true;
}
else if (ecf_flags & ECF_PURE)
{
e->speculative_call_info (direct, indirect, ref);
if (!(indirect->indirect_info->ecf_flags & ECF_PURE))
return true;
}
}
/* If we did not managed to inline the function nor redirect
to an ipa-cp clone (that are seen by having local flag set),
it is probably pointless to inline it unless hardware is missing
indirect call predictor. */
if (!anticipate_inlining && e->inline_failed && !target->local.local)
return false;
/* For overwritable targets there is not much to do. */
if (e->inline_failed
&& (!can_inline_edge_p (e, false)
|| !can_inline_edge_by_limits_p (e, false, true)))
return false;
/* OK, speculation seems interesting. */
return true;
}
/* We know that EDGE is not going to be inlined.
See if we can remove speculation. */
static void
resolve_noninline_speculation (edge_heap_t *edge_heap, struct cgraph_edge *edge)
{
if (edge->speculative && !speculation_useful_p (edge, false))
{
struct cgraph_node *node = edge->caller;
struct cgraph_node *where = node->global.inlined_to
? node->global.inlined_to : node;
auto_bitmap updated_nodes;
if (edge->count.ipa ().initialized_p ())
spec_rem += edge->count.ipa ();
edge->resolve_speculation ();
reset_edge_caches (where);
ipa_update_overall_fn_summary (where);
update_caller_keys (edge_heap, where,
updated_nodes, NULL);
update_callee_keys (edge_heap, where,
updated_nodes);
}
}
/* Return true if NODE should be accounted for overall size estimate.
Skip all nodes optimized for size so we can measure the growth of hot
part of program no matter of the padding. */
bool
inline_account_function_p (struct cgraph_node *node)
{
return (!DECL_EXTERNAL (node->decl)
&& !opt_for_fn (node->decl, optimize_size)
&& node->frequency != NODE_FREQUENCY_UNLIKELY_EXECUTED);
}
/* Count number of callers of NODE and store it into DATA (that
points to int. Worker for cgraph_for_node_and_aliases. */
static bool
sum_callers (struct cgraph_node *node, void *data)
{
struct cgraph_edge *e;
int *num_calls = (int *)data;
for (e = node->callers; e; e = e->next_caller)
(*num_calls)++;
return false;
}
/* We use greedy algorithm for inlining of small functions:
All inline candidates are put into prioritized heap ordered in
increasing badness.
The inlining of small functions is bounded by unit growth parameters. */
static void
inline_small_functions (void)
{
struct cgraph_node *node;
struct cgraph_edge *edge;
edge_heap_t edge_heap (sreal::min ());
auto_bitmap updated_nodes;
int min_size, max_size;
auto_vec<cgraph_edge *> new_indirect_edges;
int initial_size = 0;
struct cgraph_node **order = XCNEWVEC (cgraph_node *, symtab->cgraph_count);
struct cgraph_edge_hook_list *edge_removal_hook_holder;
new_indirect_edges.create (8);
edge_removal_hook_holder
= symtab->add_edge_removal_hook (&heap_edge_removal_hook, &edge_heap);
/* Compute overall unit size and other global parameters used by badness
metrics. */
max_count = profile_count::uninitialized ();
ipa_reduced_postorder (order, true, true, NULL);
free (order);
FOR_EACH_DEFINED_FUNCTION (node)
if (!node->global.inlined_to)
{
if (!node->alias && node->analyzed
&& (node->has_gimple_body_p () || node->thunk.thunk_p)
&& opt_for_fn (node->decl, optimize))
{
struct ipa_fn_summary *info = ipa_fn_summaries->get (node);
struct ipa_dfs_info *dfs = (struct ipa_dfs_info *) node->aux;
/* Do not account external functions, they will be optimized out
if not inlined. Also only count the non-cold portion of program. */
if (inline_account_function_p (node))
initial_size += info->size;
info->growth = estimate_growth (node);
int num_calls = 0;
node->call_for_symbol_and_aliases (sum_callers, &num_calls,
true);
if (num_calls == 1)
info->single_caller = true;
if (dfs && dfs->next_cycle)
{
struct cgraph_node *n2;
int id = dfs->scc_no + 1;
for (n2 = node; n2;
n2 = ((struct ipa_dfs_info *) n2->aux)->next_cycle)
if (opt_for_fn (n2->decl, optimize))
{
struct ipa_fn_summary *info2 = ipa_fn_summaries->get (n2);
if (info2->scc_no)
break;
info2->scc_no = id;
}
}
}
for (edge = node->callers; edge; edge = edge->next_caller)
max_count = max_count.max (edge->count.ipa ());
}
ipa_free_postorder_info ();
initialize_growth_caches ();
if (dump_file)
fprintf (dump_file,
"\nDeciding on inlining of small functions. Starting with size %i.\n",
initial_size);
overall_size = initial_size;
max_size = compute_max_insns (overall_size);
min_size = overall_size;
/* Populate the heap with all edges we might inline. */
FOR_EACH_DEFINED_FUNCTION (node)
{
bool update = false;
struct cgraph_edge *next = NULL;
bool has_speculative = false;
if (!opt_for_fn (node->decl, optimize))
continue;
if (dump_file)
fprintf (dump_file, "Enqueueing calls in %s.\n", node->dump_name ());
for (edge = node->callees; edge; edge = next)
{
next = edge->next_callee;
if (edge->inline_failed
&& !edge->aux
&& can_inline_edge_p (edge, true)
&& want_inline_small_function_p (edge, true)
&& can_inline_edge_by_limits_p (edge, true)
&& edge->inline_failed)
{
gcc_assert (!edge->aux);
update_edge_key (&edge_heap, edge);
}
if (edge->speculative)
has_speculative = true;
}
if (has_speculative)
for (edge = node->callees; edge; edge = next)
if (edge->speculative && !speculation_useful_p (edge,
edge->aux != NULL))
{
edge->resolve_speculation ();
update = true;
}
if (update)
{
struct cgraph_node *where = node->global.inlined_to
? node->global.inlined_to : node;
ipa_update_overall_fn_summary (where);
reset_edge_caches (where);
update_caller_keys (&edge_heap, where,
updated_nodes, NULL);
update_callee_keys (&edge_heap, where,
updated_nodes);
bitmap_clear (updated_nodes);
}
}
gcc_assert (in_lto_p
|| !(max_count > 0)
|| (profile_info && flag_branch_probabilities));
while (!edge_heap.empty ())
{
int old_size = overall_size;
struct cgraph_node *where, *callee;
sreal badness = edge_heap.min_key ();
sreal current_badness;
int growth;
edge = edge_heap.extract_min ();
gcc_assert (edge->aux);
edge->aux = NULL;
if (!edge->inline_failed || !edge->callee->analyzed)
continue;
#if CHECKING_P
/* Be sure that caches are maintained consistent.
This check is affected by scaling roundoff errors when compiling for
IPA this we skip it in that case. */
if (!edge->callee->count.ipa_p ()
&& (!max_count.initialized_p () || !max_count.nonzero_p ()))
{
sreal cached_badness = edge_badness (edge, false);
int old_size_est = estimate_edge_size (edge);
sreal old_time_est = estimate_edge_time (edge);
int old_hints_est = estimate_edge_hints (edge);
reset_edge_growth_cache (edge);
gcc_assert (old_size_est == estimate_edge_size (edge));
gcc_assert (old_time_est == estimate_edge_time (edge));
/* FIXME:
gcc_assert (old_hints_est == estimate_edge_hints (edge));
fails with profile feedback because some hints depends on
maybe_hot_edge_p predicate and because callee gets inlined to other
calls, the edge may become cold.
This ought to be fixed by computing relative probabilities
for given invocation but that will be better done once whole
code is converted to sreals. Disable for now and revert to "wrong"
value so enable/disable checking paths agree. */
edge_growth_cache[edge->uid].hints = old_hints_est + 1;
/* When updating the edge costs, we only decrease badness in the keys.
Increases of badness are handled lazilly; when we see key with out
of date value on it, we re-insert it now. */
current_badness = edge_badness (edge, false);
gcc_assert (cached_badness == current_badness);
gcc_assert (current_badness >= badness);
}
else
current_badness = edge_badness (edge, false);
#else
current_badness = edge_badness (edge, false);
#endif
if (current_badness != badness)
{
if (edge_heap.min () && current_badness > edge_heap.min_key ())
{
edge->aux = edge_heap.insert (current_badness, edge);
continue;
}
else
badness = current_badness;
}
if (!can_inline_edge_p (edge, true)
|| !can_inline_edge_by_limits_p (edge, true))
{
resolve_noninline_speculation (&edge_heap, edge);
continue;
}
callee = edge->callee->ultimate_alias_target ();
growth = estimate_edge_growth (edge);
if (dump_file)
{
fprintf (dump_file,
"\nConsidering %s with %i size\n",
callee->dump_name (),
ipa_fn_summaries->get (callee)->size);
fprintf (dump_file,
" to be inlined into %s in %s:%i\n"
" Estimated badness is %f, frequency %.2f.\n",
edge->caller->dump_name (),
edge->call_stmt
&& (LOCATION_LOCUS (gimple_location ((const gimple *)
edge->call_stmt))
> BUILTINS_LOCATION)
? gimple_filename ((const gimple *) edge->call_stmt)
: "unknown",
edge->call_stmt
? gimple_lineno ((const gimple *) edge->call_stmt)
: -1,
badness.to_double (),
edge->sreal_frequency ().to_double ());
if (edge->count.ipa ().initialized_p ())
{
fprintf (dump_file, " Called ");
edge->count.ipa ().dump (dump_file);
fprintf (dump_file, " times\n");
}
if (dump_flags & TDF_DETAILS)
edge_badness (edge, true);
}
if (overall_size + growth > max_size
&& !DECL_DISREGARD_INLINE_LIMITS (callee->decl))
{
edge->inline_failed = CIF_INLINE_UNIT_GROWTH_LIMIT;
report_inline_failed_reason (edge);
resolve_noninline_speculation (&edge_heap, edge);
continue;
}
if (!want_inline_small_function_p (edge, true))
{
resolve_noninline_speculation (&edge_heap, edge);
continue;
}
/* Heuristics for inlining small functions work poorly for
recursive calls where we do effects similar to loop unrolling.
When inlining such edge seems profitable, leave decision on
specific inliner. */
if (edge->recursive_p ())
{
where = edge->caller;
if (where->global.inlined_to)
where = where->global.inlined_to;
if (!recursive_inlining (edge,
opt_for_fn (edge->caller->decl,
flag_indirect_inlining)
? &new_indirect_edges : NULL))
{
edge->inline_failed = CIF_RECURSIVE_INLINING;
resolve_noninline_speculation (&edge_heap, edge);
continue;
}
reset_edge_caches (where);
/* Recursive inliner inlines all recursive calls of the function
at once. Consequently we need to update all callee keys. */
if (opt_for_fn (edge->caller->decl, flag_indirect_inlining))
add_new_edges_to_heap (&edge_heap, new_indirect_edges);
update_callee_keys (&edge_heap, where, updated_nodes);
bitmap_clear (updated_nodes);
}
else
{
struct cgraph_node *outer_node = NULL;
int depth = 0;
/* Consider the case where self recursive function A is inlined
into B. This is desired optimization in some cases, since it
leads to effect similar of loop peeling and we might completely
optimize out the recursive call. However we must be extra
selective. */
where = edge->caller;
while (where->global.inlined_to)
{
if (where->decl == callee->decl)
outer_node = where, depth++;
where = where->callers->caller;
}
if (outer_node
&& !want_inline_self_recursive_call_p (edge, outer_node,
true, depth))
{
edge->inline_failed
= (DECL_DISREGARD_INLINE_LIMITS (edge->callee->decl)
? CIF_RECURSIVE_INLINING : CIF_UNSPECIFIED);
resolve_noninline_speculation (&edge_heap, edge);
continue;
}
else if (depth && dump_file)
fprintf (dump_file, " Peeling recursion with depth %i\n", depth);
gcc_checking_assert (!callee->global.inlined_to);
inline_call (edge, true, &new_indirect_edges, &overall_size, true);
add_new_edges_to_heap (&edge_heap, new_indirect_edges);
reset_edge_caches (edge->callee);
update_callee_keys (&edge_heap, where, updated_nodes);
}
where = edge->caller;
if (where->global.inlined_to)
where = where->global.inlined_to;
/* Our profitability metric can depend on local properties
such as number of inlinable calls and size of the function body.
After inlining these properties might change for the function we
inlined into (since it's body size changed) and for the functions
called by function we inlined (since number of it inlinable callers
might change). */
update_caller_keys (&edge_heap, where, updated_nodes, NULL);
/* Offline copy count has possibly changed, recompute if profile is
available. */
struct cgraph_node *n = cgraph_node::get (edge->callee->decl);
if (n != edge->callee && n->analyzed && n->count.ipa ().initialized_p ())
update_callee_keys (&edge_heap, n, updated_nodes);
bitmap_clear (updated_nodes);
if (dump_file)
{
fprintf (dump_file,
" Inlined %s into %s which now has time %f and size %i, "
"net change of %+i.\n",
xstrdup_for_dump (edge->callee->name ()),
xstrdup_for_dump (edge->caller->name ()),
ipa_fn_summaries->get (edge->caller)->time.to_double (),
ipa_fn_summaries->get (edge->caller)->size,
overall_size - old_size);
}
if (min_size > overall_size)
{
min_size = overall_size;
max_size = compute_max_insns (min_size);
if (dump_file)
fprintf (dump_file, "New minimal size reached: %i\n", min_size);
}
}
free_growth_caches ();
if (dump_file)
fprintf (dump_file,
"Unit growth for small function inlining: %i->%i (%i%%)\n",
initial_size, overall_size,
initial_size ? overall_size * 100 / (initial_size) - 100: 0);
symtab->remove_edge_removal_hook (edge_removal_hook_holder);
}
/* Flatten NODE. Performed both during early inlining and
at IPA inlining time. */
static void
flatten_function (struct cgraph_node *node, bool early)
{
struct cgraph_edge *e;
/* We shouldn't be called recursively when we are being processed. */
gcc_assert (node->aux == NULL);
node->aux = (void *) node;
for (e = node->callees; e; e = e->next_callee)
{
struct cgraph_node *orig_callee;
struct cgraph_node *callee = e->callee->ultimate_alias_target ();
/* We've hit cycle? It is time to give up. */
if (callee->aux)
{
if (dump_file)
fprintf (dump_file,
"Not inlining %s into %s to avoid cycle.\n",
xstrdup_for_dump (callee->name ()),
xstrdup_for_dump (e->caller->name ()));
if (cgraph_inline_failed_type (e->inline_failed) != CIF_FINAL_ERROR)
e->inline_failed = CIF_RECURSIVE_INLINING;
continue;
}
/* When the edge is already inlined, we just need to recurse into
it in order to fully flatten the leaves. */
if (!e->inline_failed)
{
flatten_function (callee, early);
continue;
}
/* Flatten attribute needs to be processed during late inlining. For
extra code quality we however do flattening during early optimization,
too. */
if (!early
? !can_inline_edge_p (e, true)
&& !can_inline_edge_by_limits_p (e, true)
: !can_early_inline_edge_p (e))
continue;
if (e->recursive_p ())
{
if (dump_file)
fprintf (dump_file, "Not inlining: recursive call.\n");
continue;
}
if (gimple_in_ssa_p (DECL_STRUCT_FUNCTION (node->decl))
!= gimple_in_ssa_p (DECL_STRUCT_FUNCTION (callee->decl)))
{
if (dump_file)
fprintf (dump_file, "Not inlining: SSA form does not match.\n");
continue;
}
/* Inline the edge and flatten the inline clone. Avoid
recursing through the original node if the node was cloned. */
if (dump_file)
fprintf (dump_file, " Inlining %s into %s.\n",
xstrdup_for_dump (callee->name ()),
xstrdup_for_dump (e->caller->name ()));
orig_callee = callee;
inline_call (e, true, NULL, NULL, false);
if (e->callee != orig_callee)
orig_callee->aux = (void *) node;
flatten_function (e->callee, early);
if (e->callee != orig_callee)
orig_callee->aux = NULL;
}
node->aux = NULL;
if (!node->global.inlined_to)
ipa_update_overall_fn_summary (node);
}
/* Inline NODE to all callers. Worker for cgraph_for_node_and_aliases.
DATA points to number of calls originally found so we avoid infinite
recursion. */
static bool
inline_to_all_callers_1 (struct cgraph_node *node, void *data,
hash_set<cgraph_node *> *callers)
{
int *num_calls = (int *)data;
bool callee_removed = false;
while (node->callers && !node->global.inlined_to)
{
struct cgraph_node *caller = node->callers->caller;
if (!can_inline_edge_p (node->callers, true)
|| !can_inline_edge_by_limits_p (node->callers, true)
|| node->callers->recursive_p ())
{
if (dump_file)
fprintf (dump_file, "Uninlinable call found; giving up.\n");
*num_calls = 0;
return false;
}
if (dump_file)
{
fprintf (dump_file,
"\nInlining %s size %i.\n",
node->name (),
ipa_fn_summaries->get (node)->size);
fprintf (dump_file,
" Called once from %s %i insns.\n",
node->callers->caller->name (),
ipa_fn_summaries->get (node->callers->caller)->size);
}
/* Remember which callers we inlined to, delaying updating the
overall summary. */
callers->add (node->callers->caller);
inline_call (node->callers, true, NULL, NULL, false, &callee_removed);
if (dump_file)
fprintf (dump_file,
" Inlined into %s which now has %i size\n",
caller->name (),
ipa_fn_summaries->get (caller)->size);
if (!(*num_calls)--)
{
if (dump_file)
fprintf (dump_file, "New calls found; giving up.\n");
return callee_removed;
}
if (callee_removed)
return true;
}
return false;
}
/* Wrapper around inline_to_all_callers_1 doing delayed overall summary
update. */
static bool
inline_to_all_callers (struct cgraph_node *node, void *data)
{
hash_set<cgraph_node *> callers;
bool res = inline_to_all_callers_1 (node, data, &callers);
/* Perform the delayed update of the overall summary of all callers
processed. This avoids quadratic behavior in the cases where
we have a lot of calls to the same function. */
for (hash_set<cgraph_node *>::iterator i = callers.begin ();
i != callers.end (); ++i)
ipa_update_overall_fn_summary (*i);
return res;
}
/* Output overall time estimate. */
static void
dump_overall_stats (void)
{
sreal sum_weighted = 0, sum = 0;
struct cgraph_node *node;
FOR_EACH_DEFINED_FUNCTION (node)
if (!node->global.inlined_to
&& !node->alias)
{
sreal time = ipa_fn_summaries->get (node)->time;
sum += time;
if (node->count.ipa ().initialized_p ())
sum_weighted += time * node->count.ipa ().to_gcov_type ();
}
fprintf (dump_file, "Overall time estimate: "
"%f weighted by profile: "
"%f\n", sum.to_double (), sum_weighted.to_double ());
}
/* Output some useful stats about inlining. */
static void
dump_inline_stats (void)
{
int64_t inlined_cnt = 0, inlined_indir_cnt = 0;
int64_t inlined_virt_cnt = 0, inlined_virt_indir_cnt = 0;
int64_t noninlined_cnt = 0, noninlined_indir_cnt = 0;
int64_t noninlined_virt_cnt = 0, noninlined_virt_indir_cnt = 0;
int64_t inlined_speculative = 0, inlined_speculative_ply = 0;
int64_t indirect_poly_cnt = 0, indirect_cnt = 0;
int64_t reason[CIF_N_REASONS][2];
sreal reason_freq[CIF_N_REASONS];
int i;
struct cgraph_node *node;
memset (reason, 0, sizeof (reason));
for (i=0; i < CIF_N_REASONS; i++)
reason_freq[i] = 0;
FOR_EACH_DEFINED_FUNCTION (node)
{
struct cgraph_edge *e;
for (e = node->callees; e; e = e->next_callee)
{
if (e->inline_failed)
{
if (e->count.ipa ().initialized_p ())
reason[(int) e->inline_failed][0] += e->count.ipa ().to_gcov_type ();
reason_freq[(int) e->inline_failed] += e->sreal_frequency ();
reason[(int) e->inline_failed][1] ++;
if (DECL_VIRTUAL_P (e->callee->decl)
&& e->count.ipa ().initialized_p ())
{
if (e->indirect_inlining_edge)
noninlined_virt_indir_cnt += e->count.ipa ().to_gcov_type ();
else
noninlined_virt_cnt += e->count.ipa ().to_gcov_type ();
}
else if (e->count.ipa ().initialized_p ())
{
if (e->indirect_inlining_edge)
noninlined_indir_cnt += e->count.ipa ().to_gcov_type ();
else
noninlined_cnt += e->count.ipa ().to_gcov_type ();
}
}
else if (e->count.ipa ().initialized_p ())
{
if (e->speculative)
{
if (DECL_VIRTUAL_P (e->callee->decl))
inlined_speculative_ply += e->count.ipa ().to_gcov_type ();
else
inlined_speculative += e->count.ipa ().to_gcov_type ();
}
else if (DECL_VIRTUAL_P (e->callee->decl))
{
if (e->indirect_inlining_edge)
inlined_virt_indir_cnt += e->count.ipa ().to_gcov_type ();
else
inlined_virt_cnt += e->count.ipa ().to_gcov_type ();
}
else
{
if (e->indirect_inlining_edge)
inlined_indir_cnt += e->count.ipa ().to_gcov_type ();
else
inlined_cnt += e->count.ipa ().to_gcov_type ();
}
}
}
for (e = node->indirect_calls; e; e = e->next_callee)
if (e->indirect_info->polymorphic
& e->count.ipa ().initialized_p ())
indirect_poly_cnt += e->count.ipa ().to_gcov_type ();
else if (e->count.ipa ().initialized_p ())
indirect_cnt += e->count.ipa ().to_gcov_type ();
}
if (max_count.initialized_p ())
{
fprintf (dump_file,
"Inlined %" PRId64 " + speculative "
"%" PRId64 " + speculative polymorphic "
"%" PRId64 " + previously indirect "
"%" PRId64 " + virtual "
"%" PRId64 " + virtual and previously indirect "
"%" PRId64 "\n" "Not inlined "
"%" PRId64 " + previously indirect "
"%" PRId64 " + virtual "
"%" PRId64 " + virtual and previously indirect "
"%" PRId64 " + stil indirect "
"%" PRId64 " + still indirect polymorphic "
"%" PRId64 "\n", inlined_cnt,
inlined_speculative, inlined_speculative_ply,
inlined_indir_cnt, inlined_virt_cnt, inlined_virt_indir_cnt,
noninlined_cnt, noninlined_indir_cnt, noninlined_virt_cnt,
noninlined_virt_indir_cnt, indirect_cnt, indirect_poly_cnt);
fprintf (dump_file, "Removed speculations ");
spec_rem.dump (dump_file);
fprintf (dump_file, "\n");
}
dump_overall_stats ();
fprintf (dump_file, "\nWhy inlining failed?\n");
for (i = 0; i < CIF_N_REASONS; i++)
if (reason[i][1])
fprintf (dump_file, "%-50s: %8i calls, %8f freq, %" PRId64" count\n",
cgraph_inline_failed_string ((cgraph_inline_failed_t) i),
(int) reason[i][1], reason_freq[i].to_double (), reason[i][0]);
}
/* Called when node is removed. */
static void
flatten_remove_node_hook (struct cgraph_node *node, void *data)
{
if (lookup_attribute ("flatten", DECL_ATTRIBUTES (node->decl)) == NULL)
return;
hash_set<struct cgraph_node *> *removed
= (hash_set<struct cgraph_node *> *) data;
removed->add (node);
}
/* Decide on the inlining. We do so in the topological order to avoid
expenses on updating data structures. */
static unsigned int
ipa_inline (void)
{
struct cgraph_node *node;
int nnodes;
struct cgraph_node **order;
int i, j;
int cold;
bool remove_functions = false;
order = XCNEWVEC (struct cgraph_node *, symtab->cgraph_count);
if (dump_file)
ipa_dump_fn_summaries (dump_file);
nnodes = ipa_reverse_postorder (order);
spec_rem = profile_count::zero ();
FOR_EACH_FUNCTION (node)
{
node->aux = 0;
/* Recompute the default reasons for inlining because they may have
changed during merging. */
if (in_lto_p)
{
for (cgraph_edge *e = node->callees; e; e = e->next_callee)
{
gcc_assert (e->inline_failed);
initialize_inline_failed (e);
}
for (cgraph_edge *e = node->indirect_calls; e; e = e->next_callee)
initialize_inline_failed (e);
}
}
if (dump_file)
fprintf (dump_file, "\nFlattening functions:\n");
/* First shrink order array, so that it only contains nodes with
flatten attribute. */
for (i = nnodes - 1, j = i; i >= 0; i--)
{
node = order[i];
if (lookup_attribute ("flatten",
DECL_ATTRIBUTES (node->decl)) != NULL)
order[j--] = order[i];
}
/* After the above loop, order[j + 1] ... order[nnodes - 1] contain
nodes with flatten attribute. If there is more than one such
node, we need to register a node removal hook, as flatten_function
could remove other nodes with flatten attribute. See PR82801. */
struct cgraph_node_hook_list *node_removal_hook_holder = NULL;
hash_set<struct cgraph_node *> *flatten_removed_nodes = NULL;
if (j < nnodes - 2)
{
flatten_removed_nodes = new hash_set<struct cgraph_node *>;
node_removal_hook_holder
= symtab->add_cgraph_removal_hook (&flatten_remove_node_hook,
flatten_removed_nodes);
}
/* In the first pass handle functions to be flattened. Do this with
a priority so none of our later choices will make this impossible. */
for (i = nnodes - 1; i > j; i--)
{
node = order[i];
if (flatten_removed_nodes
&& flatten_removed_nodes->contains (node))
continue;
/* Handle nodes to be flattened.
Ideally when processing callees we stop inlining at the
entry of cycles, possibly cloning that entry point and
try to flatten itself turning it into a self-recursive
function. */
if (dump_file)
fprintf (dump_file, "Flattening %s\n", node->name ());
flatten_function (node, false);
}
if (j < nnodes - 2)
{
symtab->remove_cgraph_removal_hook (node_removal_hook_holder);
delete flatten_removed_nodes;
}
free (order);
if (dump_file)
dump_overall_stats ();
inline_small_functions ();
gcc_assert (symtab->state == IPA_SSA);
symtab->state = IPA_SSA_AFTER_INLINING;
/* Do first after-inlining removal. We want to remove all "stale" extern
inline functions and virtual functions so we really know what is called
once. */
symtab->remove_unreachable_nodes (dump_file);
/* Inline functions with a property that after inlining into all callers the
code size will shrink because the out-of-line copy is eliminated.
We do this regardless on the callee size as long as function growth limits
are met. */
if (dump_file)
fprintf (dump_file,
"\nDeciding on functions to be inlined into all callers and "
"removing useless speculations:\n");
/* Inlining one function called once has good chance of preventing
inlining other function into the same callee. Ideally we should
work in priority order, but probably inlining hot functions first
is good cut without the extra pain of maintaining the queue.
??? this is not really fitting the bill perfectly: inlining function
into callee often leads to better optimization of callee due to
increased context for optimization.
For example if main() function calls a function that outputs help
and then function that does the main optmization, we should inline
the second with priority even if both calls are cold by themselves.
We probably want to implement new predicate replacing our use of
maybe_hot_edge interpreted as maybe_hot_edge || callee is known
to be hot. */
for (cold = 0; cold <= 1; cold ++)
{
FOR_EACH_DEFINED_FUNCTION (node)
{
struct cgraph_edge *edge, *next;
bool update=false;
if (!opt_for_fn (node->decl, optimize)
|| !opt_for_fn (node->decl, flag_inline_functions_called_once))
continue;
for (edge = node->callees; edge; edge = next)
{
next = edge->next_callee;
if (edge->speculative && !speculation_useful_p (edge, false))
{
if (edge->count.ipa ().initialized_p ())
spec_rem += edge->count.ipa ();
edge->resolve_speculation ();
update = true;
remove_functions = true;
}
}
if (update)
{
struct cgraph_node *where = node->global.inlined_to
? node->global.inlined_to : node;
reset_edge_caches (where);
ipa_update_overall_fn_summary (where);
}
if (want_inline_function_to_all_callers_p (node, cold))
{
int num_calls = 0;
node->call_for_symbol_and_aliases (sum_callers, &num_calls,
true);
while (node->call_for_symbol_and_aliases
(inline_to_all_callers, &num_calls, true))
;
remove_functions = true;
}
}
}
/* Free ipa-prop structures if they are no longer needed. */
ipa_free_all_structures_after_iinln ();
if (dump_file)
{
fprintf (dump_file,
"\nInlined %i calls, eliminated %i functions\n\n",
ncalls_inlined, nfunctions_inlined);
dump_inline_stats ();
}
if (dump_file)
ipa_dump_fn_summaries (dump_file);
return remove_functions ? TODO_remove_functions : 0;
}
/* Inline always-inline function calls in NODE. */
static bool
inline_always_inline_functions (struct cgraph_node *node)
{
struct cgraph_edge *e;
bool inlined = false;
for (e = node->callees; e; e = e->next_callee)
{
struct cgraph_node *callee = e->callee->ultimate_alias_target ();
if (!DECL_DISREGARD_INLINE_LIMITS (callee->decl))
continue;
if (e->recursive_p ())
{
if (dump_file)
fprintf (dump_file, " Not inlining recursive call to %s.\n",
e->callee->name ());
e->inline_failed = CIF_RECURSIVE_INLINING;
continue;
}
if (!can_early_inline_edge_p (e))
{
/* Set inlined to true if the callee is marked "always_inline" but
is not inlinable. This will allow flagging an error later in
expand_call_inline in tree-inline.c. */
if (lookup_attribute ("always_inline",
DECL_ATTRIBUTES (callee->decl)) != NULL)
inlined = true;
continue;
}
if (dump_file)
fprintf (dump_file, " Inlining %s into %s (always_inline).\n",
xstrdup_for_dump (e->callee->name ()),
xstrdup_for_dump (e->caller->name ()));
inline_call (e, true, NULL, NULL, false);
inlined = true;
}
if (inlined)
ipa_update_overall_fn_summary (node);
return inlined;
}
/* Decide on the inlining. We do so in the topological order to avoid
expenses on updating data structures. */
static bool
early_inline_small_functions (struct cgraph_node *node)
{
struct cgraph_edge *e;
bool inlined = false;
for (e = node->callees; e; e = e->next_callee)
{
struct cgraph_node *callee = e->callee->ultimate_alias_target ();
if (!ipa_fn_summaries->get (callee)->inlinable
|| !e->inline_failed)
continue;
/* Do not consider functions not declared inline. */
if (!DECL_DECLARED_INLINE_P (callee->decl)
&& !opt_for_fn (node->decl, flag_inline_small_functions)
&& !opt_for_fn (node->decl, flag_inline_functions))
continue;
if (dump_file)
fprintf (dump_file, "Considering inline candidate %s.\n",
callee->name ());
if (!can_early_inline_edge_p (e))
continue;
if (e->recursive_p ())
{
if (dump_file)
fprintf (dump_file, " Not inlining: recursive call.\n");
continue;
}
if (!want_early_inline_function_p (e))
continue;
if (dump_file)
fprintf (dump_file, " Inlining %s into %s.\n",
xstrdup_for_dump (callee->name ()),
xstrdup_for_dump (e->caller->name ()));
inline_call (e, true, NULL, NULL, false);
inlined = true;
}
if (inlined)
ipa_update_overall_fn_summary (node);
return inlined;
}
unsigned int
early_inliner (function *fun)
{
struct cgraph_node *node = cgraph_node::get (current_function_decl);
struct cgraph_edge *edge;
unsigned int todo = 0;
int iterations = 0;
bool inlined = false;
if (seen_error ())
return 0;
/* Do nothing if datastructures for ipa-inliner are already computed. This
happens when some pass decides to construct new function and
cgraph_add_new_function calls lowering passes and early optimization on
it. This may confuse ourself when early inliner decide to inline call to
function clone, because function clones don't have parameter list in
ipa-prop matching their signature. */
if (ipa_node_params_sum)
return 0;
if (flag_checking)
node->verify ();
node->remove_all_references ();
/* Even when not optimizing or not inlining inline always-inline
functions. */
inlined = inline_always_inline_functions (node);
if (!optimize
|| flag_no_inline
|| !flag_early_inlining
/* Never inline regular functions into always-inline functions
during incremental inlining. This sucks as functions calling
always inline functions will get less optimized, but at the
same time inlining of functions calling always inline
function into an always inline function might introduce
cycles of edges to be always inlined in the callgraph.
We might want to be smarter and just avoid this type of inlining. */
|| (DECL_DISREGARD_INLINE_LIMITS (node->decl)
&& lookup_attribute ("always_inline",
DECL_ATTRIBUTES (node->decl))))
;
else if (lookup_attribute ("flatten",
DECL_ATTRIBUTES (node->decl)) != NULL)
{
/* When the function is marked to be flattened, recursively inline
all calls in it. */
if (dump_file)
fprintf (dump_file,
"Flattening %s\n", node->name ());
flatten_function (node, true);
inlined = true;
}
else
{
/* If some always_inline functions was inlined, apply the changes.
This way we will not account always inline into growth limits and
moreover we will inline calls from always inlines that we skipped
previously because of conditional above. */
if (inlined)
{
timevar_push (TV_INTEGRATION);
todo |= optimize_inline_calls (current_function_decl);
/* optimize_inline_calls call above might have introduced new
statements that don't have inline parameters computed. */
for (edge = node->callees; edge; edge = edge->next_callee)
{
struct ipa_call_summary *es = ipa_call_summaries->get (edge);
es->call_stmt_size
= estimate_num_insns (edge->call_stmt, &eni_size_weights);
es->call_stmt_time
= estimate_num_insns (edge->call_stmt, &eni_time_weights);
}
ipa_update_overall_fn_summary (node);
inlined = false;
timevar_pop (TV_INTEGRATION);
}
/* We iterate incremental inlining to get trivial cases of indirect
inlining. */
while (iterations < PARAM_VALUE (PARAM_EARLY_INLINER_MAX_ITERATIONS)
&& early_inline_small_functions (node))
{
timevar_push (TV_INTEGRATION);
todo |= optimize_inline_calls (current_function_decl);
/* Technically we ought to recompute inline parameters so the new
iteration of early inliner works as expected. We however have
values approximately right and thus we only need to update edge
info that might be cleared out for newly discovered edges. */
for (edge = node->callees; edge; edge = edge->next_callee)
{
/* We have no summary for new bound store calls yet. */
struct ipa_call_summary *es = ipa_call_summaries->get (edge);
es->call_stmt_size
= estimate_num_insns (edge->call_stmt, &eni_size_weights);
es->call_stmt_time
= estimate_num_insns (edge->call_stmt, &eni_time_weights);
if (edge->callee->decl
&& !gimple_check_call_matching_types (
edge->call_stmt, edge->callee->decl, false))
{
edge->inline_failed = CIF_MISMATCHED_ARGUMENTS;
edge->call_stmt_cannot_inline_p = true;
}
}
if (iterations < PARAM_VALUE (PARAM_EARLY_INLINER_MAX_ITERATIONS) - 1)
ipa_update_overall_fn_summary (node);
timevar_pop (TV_INTEGRATION);
iterations++;
inlined = false;
}
if (dump_file)
fprintf (dump_file, "Iterations: %i\n", iterations);
}
if (inlined)
{
timevar_push (TV_INTEGRATION);
todo |= optimize_inline_calls (current_function_decl);
timevar_pop (TV_INTEGRATION);
}
fun->always_inline_functions_inlined = true;
return todo;
}
/* Do inlining of small functions. Doing so early helps profiling and other
passes to be somewhat more effective and avoids some code duplication in
later real inlining pass for testcases with very many function calls. */
namespace {
const pass_data pass_data_early_inline =
{
GIMPLE_PASS, /* type */
"einline", /* name */
OPTGROUP_INLINE, /* optinfo_flags */
TV_EARLY_INLINING, /* tv_id */
PROP_ssa, /* properties_required */
0, /* properties_provided */
0, /* properties_destroyed */
0, /* todo_flags_start */
0, /* todo_flags_finish */
};
class pass_early_inline : public gimple_opt_pass
{
public:
pass_early_inline (gcc::context *ctxt)
: gimple_opt_pass (pass_data_early_inline, ctxt)
{}
/* opt_pass methods: */
virtual unsigned int execute (function *);
}; // class pass_early_inline
unsigned int
pass_early_inline::execute (function *fun)
{
return early_inliner (fun);
}
} // anon namespace
gimple_opt_pass *
make_pass_early_inline (gcc::context *ctxt)
{
return new pass_early_inline (ctxt);
}
namespace {
const pass_data pass_data_ipa_inline =
{
IPA_PASS, /* type */
"inline", /* name */
OPTGROUP_INLINE, /* optinfo_flags */
TV_IPA_INLINING, /* tv_id */
0, /* properties_required */
0, /* properties_provided */
0, /* properties_destroyed */
0, /* todo_flags_start */
( TODO_dump_symtab ), /* todo_flags_finish */
};
class pass_ipa_inline : public ipa_opt_pass_d
{
public:
pass_ipa_inline (gcc::context *ctxt)
: ipa_opt_pass_d (pass_data_ipa_inline, ctxt,
NULL, /* generate_summary */
NULL, /* write_summary */
NULL, /* read_summary */
NULL, /* write_optimization_summary */
NULL, /* read_optimization_summary */
NULL, /* stmt_fixup */
0, /* function_transform_todo_flags_start */
inline_transform, /* function_transform */
NULL) /* variable_transform */
{}
/* opt_pass methods: */
virtual unsigned int execute (function *) { return ipa_inline (); }
}; // class pass_ipa_inline
} // anon namespace
ipa_opt_pass_d *
make_pass_ipa_inline (gcc::context *ctxt)
{
return new pass_ipa_inline (ctxt);
}
|