File: perlhacktips.html

package info (click to toggle)
perl-doc-html 5.26.0-4
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 39,400 kB
  • sloc: xml: 36; makefile: 7
file content (1657 lines) | stat: -rw-r--r-- 110,662 bytes parent folder | download
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
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
"http://www.w3.org/TR/html4/loose.dtd">
<html>
<head>
  <title>perlhacktips - perldoc.perl.org</title>
  <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
  <meta http-equiv="Content-Language" content="en-gb">
  <link rel="search" type="application/opensearchdescription+xml" title="Search perldoc.perl.org" href="/static/search.xml"/>
  <link href="static/css-20100830.css" rel="stylesheet" rev="stylesheet" type="text/css" media="screen">
  <link href="static/exploreperl.css" rel="stylesheet" rev="stylesheet" type="text/css">
</head>

<body onLoad="perldoc.startup();" onPageShow="if (event.persisted) perldoc.startup();">
    <div id="page">
      
      <div id="header">
	<div id="homepage_link">
	  <a href="index.html"></a>
	</div>
	<div id="strapline">
	  Perl Programming Documentation
	</div>
	<div id="download_link" class="download">
	  <a href="http://www.perl.org/get.html">Download Perl</a>
	</div>
	<div id="explore_link" class="download">
	  <a id="explore_anchor" href="#">Explore</a>
	</div>
      </div>
      
      <div id="body">
        <div id="left_column">
          <div class="side_group">
            
	    <div class="side_panel doc_panel">
              <p>Manual</p>
              <ul>
                <li><a href="index-overview.html">Overview</a>
                <li><a href="index-tutorials.html">Tutorials</a>
                <li><a href="index-faq.html">FAQs</a>
                <li><a href="index-history.html">History / Changes</a>
                <li><a href="index-licence.html">License</a>
              </ul>
            </div>
            <div class="side_panel doc_panel">
              <p>Reference</p>
              <ul>
                <li><a href="index-language.html">Language</a>
                <li><a href="index-functions.html">Functions</a>
                <li><a href="perlop.html">Operators</a>
                <li><a href="perlvar.html">Special Variables</a>
                <li><a href="index-pragmas.html">Pragmas</a>
                <li><a href="index-utilities.html">Utilities</a>
                <li><a href="index-internals.html">Internals</a>
                <li><a href="index-platforms.html">Platform Specific</a>
              </ul>
            </div>
            <div class="side_panel doc_panel">
              <p>Modules</p>
              <ul>
		<li>
		
                
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		
                  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		
                  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		    
		  
		
                  
		
                  
		
                  
		    
		  
		
                  
		
                  
		
		
                    <a href="index-modules-A.html">A</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-B.html">B</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-C.html">C</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-D.html">D</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-E.html">E</a>
                    
                      
                        <li>
                      
                    
                
                    <a href="index-modules-F.html">F</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-G.html">G</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-H.html">H</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-I.html">I</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-L.html">L</a>
                    
                      
                        <li>
                      
                    
                
                    <a href="index-modules-M.html">M</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-N.html">N</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-O.html">O</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-P.html">P</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-S.html">S</a>
                    
                      
                        <li>
                      
                    
                
                    <a href="index-modules-T.html">T</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-U.html">U</a>
                    
                      
                        &bull;
                      
                    
                
                    <a href="index-modules-X.html">X</a>
                    
                
              </ul>
            </div>
            
	      <div class="side_panel doc_panel">
		<p>Tools</p>
		<ul>
		  <li><a href="preferences.html">Preferences</a>
		</ul>
	      </div>
            
          </div>
        </div>
        <div id="centre_column">
          <div id="content_header">
            <div id="title_bar">
              <div id="page_name">
                <h1>perlhacktips</h1>
              </div>
              <div id="perl_version">
                Perl 5 version 26.0 documentation
              </div>
              <div class="page_links" id="page_links_top">
                <a href="#" onClick="toolbar.goToTop();return false;">Go to top</a>
		
              </div>
	      <div class="page_links" id="page_links_bottom">
		
                  <a href="#" id="page_index_toggle">Show page index</a> &bull;
		
                <a href="#" id="recent_pages_toggle">Show recent pages</a>		
	      </div>
	      <div id="search_form">
		<form action="search.html" method="GET" id="search">
		  <input type="text" name="q" id="search_box" alt="Search">
		</form>
	      </div>
            </div>
            <div id="breadcrumbs">
                
    <a href="index.html">Home</a> &gt;
    
      
        <a href="index-internals.html">Internals and C language interface</a> &gt;
      
    
    perlhacktips
  

            </div>
          </div>
          <div id="content_body">
	    <!--[if lt IE 7]>
 <div class="noscript">
   <p>
     <strong>It looks like you're using Internet Explorer 6. This is a very old
     browser which does not offer full support for modern websites.</strong>
   </p>
   <p>
     Unfortunately this means that this website will not work on
     your computer.
   </p>
   <p>
     Don't miss out though! To view the site (and get a better experience from
     many other websites), simply upgrade to
     <a href="http://www.microsoft.com/windows/Internet-explorer/default.aspx">Internet
Explorer 8</a>
     or download an alternative browser such as
     <a href="http://www.mozilla.com/en-US/firefox/firefox.html">Firefox</a>,
     <a href="http://www.apple.com/safari/download/">Safari</a>, or
     <a href="http://www.google.co.uk/chrome">Google Chrome</a>.
   </p>
   <p>
     All of these browsers are free. If you're using a PC at work, you may
     need to contact your IT administrator.
   </p>
 </div>
<![endif]-->
	    <noscript>
	      <div class="noscript">
	      <p>
                <strong>Please note: Many features of this site require JavaScript. You appear to have JavaScript disabled,
	        or are running a non-JavaScript capable web browser.</strong>
	      </p>
	      <p>
		To get the best experience, please enable JavaScript or download a modern web browser such as <a href="http://www.microsoft.com/windows/Internet-explorer/default.aspx">Internet Explorer 8</a>, <a href="http://www.mozilla.com/en-US/firefox/firefox.html">Firefox</a>, <a href="http://www.apple.com/safari/download/">Safari</a>, or <a href="http://www.google.co.uk/chrome">Google Chrome</a>.
              </p>
	      </div>
	    </noscript>

	    <div id="recent_pages" class="hud_container">
	      <div id="recent_pages_header" class="hud_header">
		<div id="recent_pages_close" class="hud_close"><a href="#" onClick="recentPages.hide();return false;"></a></div>
		<div id="recent_pages_title" class="hud_title"><span class="hud_span_top">Recently read</span></div>
		<div id="recent_pages_topright" class="hud_topright"></div>
	      </div>
	      <div id="recent_pages_content" class="hud_content">
	      </div>
	      <div id="recent_pages_footer" class="hud_footer">
		<div id="recent_pages_bottomleft" class="hud_bottomleft"></div>
		<div id="recent_pages_bottom" class="hud_bottom"><span class="hud_span_bottom"></span></div>
		<div id="recent_pages_resize" class="hud_resize"></div>
	      </div>
	    </div>
  
	    <div id="from_search"></div>
            <h1>perlhacktips</h1>


  <!--    -->
<ul><li><a href="#NAME">NAME</a><li><a href="#DESCRIPTION">DESCRIPTION</a><li><a href="#COMMON-PROBLEMS">COMMON PROBLEMS</a><ul><li><a href="#Perl-environment-problems">Perl environment problems</a><li><a href="#Portability-problems">Portability problems</a><li><a href="#Problematic-System-Interfaces">Problematic System Interfaces</a><li><a href="#Security-problems">Security problems</a></ul><li><a href="#DEBUGGING">DEBUGGING</a><ul><li><a href="#Poking-at-Perl">Poking at Perl</a><li><a href="#Using-a-source-level-debugger">Using a source-level debugger</a><li><a href="#gdb-macro-support">gdb macro support</a><li><a href="#Dumping-Perl-Data-Structures">Dumping Perl Data Structures</a><li><a href="#Using-gdb-to-look-at-specific-parts-of-a-program">Using gdb to look at specific parts of a program</a><li><a href="#Using-gdb-to-look-at-what-the-parser%2flexer-are-doing">Using gdb to look at what the parser/lexer are doing</a></ul><li><a href="#SOURCE-CODE-STATIC-ANALYSIS">SOURCE CODE STATIC ANALYSIS</a><ul><li><a href="#lint">lint</a><li><a href="#Coverity">Coverity</a><li><a href="#HP-UX-cadvise-(Code-Advisor)">HP-UX cadvise (Code Advisor)</a><li><a href="#cpd-(cut-and-paste-detector)">cpd (cut-and-paste detector)</a><li><a href="#gcc-warnings">gcc warnings</a><li><a href="#Warnings-of-other-C-compilers">Warnings of other C compilers</a></ul><li><a href="#MEMORY-DEBUGGERS">MEMORY DEBUGGERS</a><ul><li><a href="#valgrind">valgrind</a><li><a href="#AddressSanitizer">AddressSanitizer</a></ul><li><a href="#PROFILING">PROFILING</a><ul><li><a href="#Gprof-Profiling">Gprof Profiling</a><li><a href="#GCC-gcov-Profiling">GCC gcov Profiling</a></ul><li><a href="#MISCELLANEOUS-TRICKS">MISCELLANEOUS TRICKS</a><ul><li><a href="#PERL_DESTRUCT_LEVEL">PERL_DESTRUCT_LEVEL</a><li><a href="#PERL_MEM_LOG">PERL_MEM_LOG</a><li><a href="#DDD-over-gdb">DDD over gdb</a><li><a href="#C-backtrace">C backtrace</a><li><a href="#Poison">Poison</a><li><a href="#Read-only-optrees">Read-only optrees</a><li><a href="#When-is-a-bool-not-a-bool%3f">When is a bool not a bool?</a><li><a href="#The-.i-Targets">The .i Targets</a></ul><li><a href="#AUTHOR">AUTHOR</a></ul><a name="NAME"></a><h1>NAME</h1>
<p>perlhacktips - Tips for Perl core C code hacking</p>
<a name="DESCRIPTION"></a><h1>DESCRIPTION</h1>
<p>This document will help you learn the best way to go about hacking on
the Perl core C code.  It covers common problems, debugging, profiling,
and more.</p>
<p>If you haven't read <a href="perlhack.html">perlhack</a> and <a href="perlhacktut.html">perlhacktut</a> yet, you might want
to do that first.</p>
<a name="COMMON-PROBLEMS"></a><h1>COMMON PROBLEMS</h1>
<p>Perl source plays by ANSI C89 rules: no C99 (or C++) extensions.  In
some cases we have to take pre-ANSI requirements into consideration.
You don't care about some particular platform having broken Perl? I
hear there is still a strong demand for J2EE programmers.</p>
<a name="Perl-environment-problems"></a><h2>Perl environment problems</h2>
<ul>
<li>
<p>Not compiling with threading</p>
<p>Compiling with threading (-Duseithreads) completely rewrites the
function prototypes of Perl.  You better try your changes with that.
Related to this is the difference between "Perl_-less" and "Perl_-ly"
APIs, for example:</p>
<pre class="verbatim"><ol><li>  <span class="i">Perl_sv_setiv</span><span class="s">(</span><span class="w">aTHX_</span> ...<span class="s">)</span><span class="sc">;</span></li><li>  <span class="i">sv_setiv</span><span class="s">(</span>...<span class="s">)</span><span class="sc">;</span></li></ol></pre><p>The first one explicitly passes in the context, which is needed for
e.g. threaded builds.  The second one does that implicitly; do not get
them mixed.  If you are not passing in a aTHX_, you will need to do a
dTHX (or a dVAR) as the first thing in the function.</p>
<p>See <a href="perlguts.html#How-multiple-interpreters-and-concurrency-are-supported">How multiple interpreters and concurrency are supported in perlguts</a> for further discussion about context.</p>
</li>
<li>
<p>Not compiling with -DDEBUGGING</p>
<p>The DEBUGGING define exposes more code to the compiler, therefore more
ways for things to go wrong.  You should try it.</p>
</li>
<li>
<p>Introducing (non-read-only) globals</p>
<p>Do not introduce any modifiable globals, truly global or file static.
They are bad form and complicate multithreading and other forms of
concurrency.  The right way is to introduce them as new interpreter
variables, see <i>intrpvar.h</i> (at the very end for binary
compatibility).</p>
<p>Introducing read-only (const) globals is okay, as long as you verify
with e.g. <code class="inline"><span class="w">nm</span> <span class="w">libperl</span>.<span class="w">a</span>|<span class="w">egrep</span> -<span class="w">v</span> <span class="q">' [TURtr] '</span></code>
 (if your <code class="inline"><span class="w">nm</span></code>
 has
BSD-style output) that the data you added really is read-only.  (If it
is, it shouldn't show up in the output of that command.)</p>
<p>If you want to have static strings, make them constant:</p>
<pre class="verbatim"><ol><li>  <span class="w">static</span> <span class="w">const</span> <span class="w">char</span> <span class="w">etc</span><span class="s">[</span><span class="s">]</span> = <span class="q">&quot;...&quot;</span><span class="sc">;</span></li></ol></pre><p>If you want to have arrays of constant strings, note carefully the
right combination of <code class="inline"><span class="w">const</span></code>
s:</p>
<pre class="verbatim"><ol><li>    <span class="w">static</span> <span class="w">const</span> <span class="w">char</span> * <span class="w">const</span> <span class="w">yippee</span><span class="s">[</span><span class="s">]</span> =</li><li>        <span class="s">{</span><span class="q">&quot;hi&quot;</span><span class="cm">,</span> <span class="q">&quot;ho&quot;</span><span class="cm">,</span> <span class="q">&quot;silver&quot;</span><span class="s">}</span><span class="sc">;</span></li></ol></pre><p>There is a way to completely hide any modifiable globals (they are all
moved to heap), the compilation setting
<code class="inline">-<span class="w">DPERL_GLOBAL_STRUCT_PRIVATE</span></code>
.  It is not normally used, but can be
used for testing, read more about it in <a href="perlguts.html#Background-and-PERL_IMPLICIT_CONTEXT">Background and PERL_IMPLICIT_CONTEXT in perlguts</a>.</p>
</li>
<li>
<p>Not exporting your new function</p>
<p>Some platforms (Win32, AIX, VMS, OS/2, to name a few) require any
function that is part of the public API (the shared Perl library) to be
explicitly marked as exported.  See the discussion about <i>embed.pl</i> in
<a href="perlguts.html">perlguts</a>.</p>
</li>
<li>
<p>Exporting your new function</p>
<p>The new shiny result of either genuine new functionality or your
arduous refactoring is now ready and correctly exported.  So what could
possibly go wrong?</p>
<p>Maybe simply that your function did not need to be exported in the
first place.  Perl has a long and not so glorious history of exporting
functions that it should not have.</p>
<p>If the function is used only inside one source code file, make it
static.  See the discussion about <i>embed.pl</i> in <a href="perlguts.html">perlguts</a>.</p>
<p>If the function is used across several files, but intended only for
Perl's internal use (and this should be the common case), do not export
it to the public API.  See the discussion about <i>embed.pl</i> in
<a href="perlguts.html">perlguts</a>.</p>
</li>
</ul>
<a name="Portability-problems"></a><h2>Portability problems</h2>
<p>The following are common causes of compilation and/or execution
failures, not common to Perl as such.  The C FAQ is good bedtime
reading.  Please test your changes with as many C compilers and
platforms as possible; we will, anyway, and it's nice to save oneself
from public embarrassment.</p>
<p>If using gcc, you can add the <code class="inline">-<span class="w">std</span>=<span class="w">c89</span></code>
 option which will hopefully
catch most of these unportabilities.  (However it might also catch
incompatibilities in your system's header files.)</p>
<p>Use the Configure <code class="inline">-<span class="w">Dgccansipedantic</span></code>
 flag to enable the gcc <code class="inline">-<span class="w">ansi</span>
-<span class="w">pedantic</span></code>
 flags which enforce stricter ANSI rules.</p>
<p>If using the <code class="inline"><span class="w">gcc</span> -<span class="w">Wall</span></code>
 note that not all the possible warnings (like
<code class="inline">-<span class="w">Wuninitialized</span></code>
) are given unless you also compile with <code class="inline">-O</code>
.</p>
<p>Note that if using gcc, starting from Perl 5.9.5 the Perl core source
code files (the ones at the top level of the source code distribution,
but not e.g. the extensions under ext/) are automatically compiled with
as many as possible of the <code class="inline">-<span class="w">std</span>=<span class="w">c89</span></code>
, <code class="inline">-<span class="w">ansi</span></code>
, <code class="inline">-<span class="w">pedantic</span></code>
, and a
selection of <code class="inline">-W</code>
 flags (see cflags.SH).</p>
<p>Also study <a href="perlport.html">perlport</a> carefully to avoid any bad assumptions about the
operating system, filesystems, character set, and so forth.</p>
<p>You may once in a while try a "make microperl" to see whether we can
still compile Perl with just the bare minimum of interfaces.  (See
README.micro.)</p>
<p>Do not assume an operating system indicates a certain compiler.</p>
<ul>
<li>
<p>Casting pointers to integers or casting integers to pointers</p>
<pre class="verbatim"><ol><li>    void castaway(U8* p)</li><li>    {</li><li>      IV i = p;</li></ol></pre><p>or</p>
<pre class="verbatim"><ol><li>    void castaway(U8* p)</li><li>    {</li><li>      IV i = (IV)p;</li></ol></pre><p>Both are bad, and broken, and unportable.  Use the PTR2IV() macro that
does it right.  (Likewise, there are PTR2UV(), PTR2NV(), INT2PTR(), and
NUM2PTR().)</p>
</li>
<li>
<p>Casting between function pointers and data pointers</p>
<p>Technically speaking casting between function pointers and data
pointers is unportable and undefined, but practically speaking it seems
to work, but you should use the FPTR2DPTR() and DPTR2FPTR() macros.
Sometimes you can also play games with unions.</p>
</li>
<li>
<p>Assuming sizeof(int) == sizeof(long)</p>
<p>There are platforms where longs are 64 bits, and platforms where ints
are 64 bits, and while we are out to shock you, even platforms where
shorts are 64 bits.  This is all legal according to the C standard.  (In
other words, "long long" is not a portable way to specify 64 bits, and
"long long" is not even guaranteed to be any wider than "long".)</p>
<p>Instead, use the definitions IV, UV, IVSIZE, I32SIZE, and so forth.
Avoid things like I32 because they are <b>not</b> guaranteed to be
<i>exactly</i> 32 bits, they are <i>at least</i> 32 bits, nor are they
guaranteed to be <b>int</b> or <b>long</b>.  If you really explicitly need
64-bit variables, use I64 and U64, but only if guarded by HAS_QUAD.</p>
</li>
<li>
<p>Assuming one can dereference any type of pointer for any type of data</p>
<pre class="verbatim"><ol><li>  <span class="w">char</span> *<span class="w">p</span> = ...<span class="sc">;</span></li><li>  <span class="w">long</span> <span class="w">pony</span> = <span class="i">*(</span><span class="w">long</span> *<span class="p">)</span><span class="w">p</span><span class="sc">;</span>    <span class="q">/* BAD */</span></li></ol></pre><p>Many platforms, quite rightly so, will give you a core dump instead of
a pony if the p happens not to be correctly aligned.</p>
</li>
<li>
<p>Lvalue casts</p>
<pre class="verbatim"><ol><li>  <span class="s">(</span><a class="l_k" href="functions/int.html">int</a><span class="s">)</span>*<span class="w">p</span> = ...<span class="sc">;</span>    <span class="q">/* BAD */</span></li></ol></pre><p>Simply not portable.  Get your lvalue to be of the right type, or maybe
use temporary variables, or dirty tricks with unions.</p>
</li>
<li>
<p>Assume <b>anything</b> about structs (especially the ones you don't
control, like the ones coming from the system headers)</p>
<ul>
<li>
<p>That a certain field exists in a struct</p>
</li>
<li>
<p>That no other fields exist besides the ones you know of</p>
</li>
<li>
<p>That a field is of certain signedness, sizeof, or type</p>
</li>
<li>
<p>That the fields are in a certain order</p>
<ul>
<li>
<p>While C guarantees the ordering specified in the struct definition,
between different platforms the definitions might differ</p>
</li>
</ul>
</li>
<li>
<p>That the sizeof(struct) or the alignments are the same everywhere</p>
<ul>
<li>
<p>There might be padding bytes between the fields to align the fields -
the bytes can be anything</p>
</li>
<li>
<p>Structs are required to be aligned to the maximum alignment required by
the fields - which for native types is for usually equivalent to
sizeof() of the field</p>
</li>
</ul>
</li>
</ul>
</li>
<li>
<p>Assuming the character set is ASCIIish</p>
<p>Perl can compile and run under EBCDIC platforms.  See <a href="perlebcdic.html">perlebcdic</a>.
This is transparent for the most part, but because the character sets
differ, you shouldn't use numeric (decimal, octal, nor hex) constants
to refer to characters.  You can safely say <code class="inline"><span class="q">&#39;A&#39;</span></code>
, but not <code class="inline"><span class="n">0x41</span></code>
.
You can safely say <code class="inline"><span class="q">&#39;\n&#39;</span></code>
, but not <code class="inline">\<span class="n">012</span></code>
.  However, you can use
macros defined in <i>utf8.h</i> to specify any code point portably.
<code class="inline"><span class="i">LATIN1_TO_NATIVE</span><span class="s">(</span><span class="n">0xDF</span><span class="s">)</span></code>
 is going to be the code point that means
LATIN SMALL LETTER SHARP S on whatever platform you are running on (on
ASCII platforms it compiles without adding any extra code, so there is
zero performance hit on those).  The acceptable inputs to
<code class="inline"><span class="w">LATIN1_TO_NATIVE</span></code>
 are from <code class="inline"><span class="n">0x00</span></code>
 through <code class="inline"><span class="n">0xFF</span></code>
.  If your input
isn't guaranteed to be in that range, use <code class="inline"><span class="w">UNICODE_TO_NATIVE</span></code>
 instead.
<code class="inline"><span class="w">NATIVE_TO_LATIN1</span></code>
 and <code class="inline"><span class="w">NATIVE_TO_UNICODE</span></code>
 translate the opposite
direction.</p>
<p>If you need the string representation of a character that doesn't have a
mnemonic name in C, you should add it to the list in
<i>regen/unicode_constants.pl</i>, and have Perl create <code class="inline"><span class="c">#define</span></code>
's for you,
based on the current platform.</p>
<p>Note that the <code class="inline">is<i>FOO</i></code> and <code class="inline">to<i>FOO</i></code> macros in <i>handy.h</i> work
properly on native code points and strings.</p>
<p>Also, the range 'A' - 'Z' in ASCII is an unbroken sequence of 26 upper
case alphabetic characters.  That is not true in EBCDIC.  Nor for 'a' to
'z'.  But '0' - '9' is an unbroken range in both systems.  Don't assume
anything about other ranges.  (Note that special handling of ranges in
regular expression patterns and transliterations makes it appear to Perl
code that the aforementioned ranges are all unbroken.)</p>
<p>Many of the comments in the existing code ignore the possibility of
EBCDIC, and may be wrong therefore, even if the code works.  This is
actually a tribute to the successful transparent insertion of being
able to handle EBCDIC without having to change pre-existing code.</p>
<p>UTF-8 and UTF-EBCDIC are two different encodings used to represent
Unicode code points as sequences of bytes.  Macros  with the same names
(but different definitions) in <i>utf8.h</i> and <i>utfebcdic.h</i> are used to
allow the calling code to think that there is only one such encoding.
This is almost always referred to as <code class="inline"><span class="w">utf8</span></code>
, but it means the EBCDIC
version as well.  Again, comments in the code may well be wrong even if
the code itself is right.  For example, the concept of UTF-8 <code class="inline"><span class="w">invariant</span>
<span class="w">characters</span></code>
 differs between ASCII and EBCDIC.  On ASCII platforms, only
characters that do not have the high-order bit set (i.e.  whose ordinals
are strict ASCII, 0 - 127) are invariant, and the documentation and
comments in the code may assume that, often referring to something
like, say, <code class="inline"><span class="w">hibit</span></code>
.  The situation differs and is not so simple on
EBCDIC machines, but as long as the code itself uses the
<code class="inline"><span class="i">NATIVE_IS_INVARIANT</span><span class="s">(</span><span class="s">)</span></code>
 macro appropriately, it works, even if the
comments are wrong.</p>
<p>As noted in <a href="perlhack.html#TESTING">TESTING in perlhack</a>, when writing test scripts, the file
<i>t/charset_tools.pl</i> contains some helpful functions for writing tests
valid on both ASCII and EBCDIC platforms.  Sometimes, though, a test
can't use a function and it's inconvenient to have different test
versions depending on the platform.  There are 20 code points that are
the same in all 4 character sets currently recognized by Perl (the 3
EBCDIC code pages plus ISO 8859-1 (ASCII/Latin1)).  These can be used in
such tests, though there is a small possibility that Perl will become
available in yet another character set, breaking your test.  All but one
of these code points are C0 control characters.  The most significant
controls that are the same are <code class="inline">\<span class="n">0</span></code>
, <code class="inline">\<span class="w">r</span></code>
, and <code class="inline">\<span class="i">N</span><span class="s">{</span><span class="w">VT</span><span class="s">}</span></code>
 (also
specifiable as <code class="inline">\<span class="w">cK</span></code>
, <code class="inline">\<span class="w">x0B</span></code>
, <code class="inline">\<span class="i">N</span><span class="s">{</span><span class="w">U</span>+<span class="n">0</span><span class="w">B</span><span class="s">}</span></code>
, or <code class="inline">\<span class="n">013</span></code>
).  The single
non-control is U+00B6 PILCROW SIGN.  The controls that are the same have
the same bit pattern in all 4 character sets, regardless of the UTF8ness
of the string containing them.  The bit pattern for U+B6 is the same in
all 4 for non-UTF8 strings, but differs in each when its containing
string is UTF-8 encoded.  The only other code points that have some sort
of sameness across all 4 character sets are the pair 0xDC and 0xFC.
Together these represent upper- and lowercase LATIN LETTER U WITH
DIAERESIS, but which is upper and which is lower may be reversed: 0xDC
is the capital in Latin1 and 0xFC is the small letter, while 0xFC is the
capital in EBCDIC and 0xDC is the small one.  This factoid may be
exploited in writing case insensitive tests that are the same across all
4 character sets.</p>
</li>
<li>
<p>Assuming the character set is just ASCII</p>
<p>ASCII is a 7 bit encoding, but bytes have 8 bits in them.  The 128 extra
characters have different meanings depending on the locale.  Absent a
locale, currently these extra characters are generally considered to be
unassigned, and this has presented some problems.  This has being
changed starting in 5.12 so that these characters can be considered to
be Latin-1 (ISO-8859-1).</p>
</li>
<li>
<p>Mixing #define and #ifdef</p>
<pre class="verbatim"><ol><li>  <span class="c">#define BURGLE(x) ... \</span></li><li>  <span class="c">#ifdef BURGLE_OLD_STYLE        /* BAD */</span></li><li>  ... <a class="l_k" href="functions/do.html">do</a> <span class="w">it</span> <span class="w">the</span> <span class="w">old</span> <span class="w">way</span> ... \</li><li>  <span class="c">#else</span></li><li>  ... <a class="l_k" href="functions/do.html">do</a> <span class="w">it</span> <span class="w">the</span> <span class="w">new</span> <span class="w">way</span> ... \</li><li>  <span class="c">#endif</span></li></ol></pre><p>You cannot portably "stack" cpp directives.  For example in the above
you need two separate BURGLE() #defines, one for each #ifdef branch.</p>
</li>
<li>
<p>Adding non-comment stuff after #endif or #else</p>
<pre class="verbatim"><ol><li>  <span class="c">#ifdef SNOSH</span></li><li>  ...</li><li>  <span class="c">#else !SNOSH    /* BAD */</span></li><li>  ...</li><li>  <span class="c">#endif SNOSH    /* BAD */</span></li></ol></pre><p>The #endif and #else cannot portably have anything non-comment after
them.  If you want to document what is going (which is a good idea
especially if the branches are long), use (C) comments:</p>
<pre class="verbatim"><ol><li>  <span class="c">#ifdef SNOSH</span></li><li>  ...</li><li>  <span class="c">#else /* !SNOSH */</span></li><li>  ...</li><li>  <span class="c">#endif /* SNOSH */</span></li></ol></pre><p>The gcc option <code class="inline">-<span class="w">Wendif</span>-<span class="w">labels</span></code>
 warns about the bad variant (by
default on starting from Perl 5.9.4).</p>
</li>
<li>
<p>Having a comma after the last element of an enum list</p>
<pre class="verbatim"><ol><li>  <span class="w">enum</span> <span class="w">color</span> <span class="s">{</span></li><li>    <span class="w">CERULEAN</span><span class="cm">,</span></li><li>    <span class="w">CHARTREUSE</span><span class="cm">,</span></li><li>    <span class="w">CINNABAR</span><span class="cm">,</span>     <span class="q">/* BAD */</span></li><li>  <span class="s">}</span><span class="sc">;</span></li></ol></pre><p>is not portable.  Leave out the last comma.</p>
<p>Also note that whether enums are implicitly morphable to ints varies
between compilers, you might need to (int).</p>
</li>
<li>
<p>Using //-comments</p>
<pre class="verbatim"><ol><li>  <span class="q">//</span> <span class="w">This</span> <span class="w">function</span> <span class="w">bamfoodles</span> <span class="w">the</span> <span class="w">zorklator</span>.   <span class="q">/* BAD */</span></li></ol></pre><p>That is C99 or C++.  Perl is C89.  Using the //-comments is silently
allowed by many C compilers but cranking up the ANSI C89 strictness
(which we like to do) causes the compilation to fail.</p>
</li>
<li>
<p>Mixing declarations and code</p>
<pre class="verbatim"><ol><li>  void zorklator()</li><li>  {</li><li>    int n = 3;</li><li>    set_zorkmids(n);    /* BAD */</li><li>    int q = 4;</li></ol></pre><p>That is C99 or C++.  Some C compilers allow that, but you shouldn't.</p>
<p>The gcc option <code class="inline">-<span class="w">Wdeclaration</span>-<span class="w">after</span>-<span class="w">statements</span></code>
 scans for such
problems (by default on starting from Perl 5.9.4).</p>
</li>
<li>
<p>Introducing variables inside for()</p>
<pre class="verbatim"><ol><li>  for(int i = ...; ...; ...) {    /* BAD */</li></ol></pre><p>That is C99 or C++.  While it would indeed be awfully nice to have that
also in C89, to limit the scope of the loop variable, alas, we cannot.</p>
</li>
<li>
<p>Mixing signed char pointers with unsigned char pointers</p>
<pre class="verbatim"><ol><li>  int foo(char *s) { ... }</li><li>  ...</li><li>  unsigned char *t = ...; /* Or U8* t = ... */</li><li>  foo(t);   /* BAD */</li></ol></pre><p>While this is legal practice, it is certainly dubious, and downright
fatal in at least one platform: for example VMS cc considers this a
fatal error.  One cause for people often making this mistake is that a
"naked char" and therefore dereferencing a "naked char pointer" have an
undefined signedness: it depends on the compiler and the flags of the
compiler and the underlying platform whether the result is signed or
unsigned.  For this very same reason using a 'char' as an array index is
bad.</p>
</li>
<li>
<p>Macros that have string constants and their arguments as substrings of
the string constants</p>
<pre class="verbatim"><ol><li>  <span class="c">#define FOO(n) printf(&quot;number = %d\n&quot;, n)    /* BAD */</span></li><li>  <span class="i">FOO</span><span class="s">(</span><span class="n">10</span><span class="s">)</span><span class="sc">;</span></li></ol></pre><p>Pre-ANSI semantics for that was equivalent to</p>
<pre class="verbatim"><ol><li>  <a class="l_k" href="functions/printf.html">printf</a><span class="s">(</span><span class="q">&quot;10umber = %d\10&quot;</span><span class="s">)</span><span class="sc">;</span></li></ol></pre><p>which is probably not what you were expecting.  Unfortunately at least
one reasonably common and modern C compiler does "real backward
compatibility" here, in AIX that is what still happens even though the
rest of the AIX compiler is very happily C89.</p>
</li>
<li>
<p>Using printf formats for non-basic C types</p>
<pre class="verbatim"><ol><li>   <span class="w">IV</span> <span class="w">i</span> = ...<span class="sc">;</span></li><li>   <a class="l_k" href="functions/printf.html">printf</a><span class="s">(</span><span class="q">&quot;i = %d\n&quot;</span><span class="cm">,</span> <span class="w">i</span><span class="s">)</span><span class="sc">;</span>    <span class="q">/* BAD */</span></li></ol></pre><p>While this might by accident work in some platform (where IV happens to
be an <code class="inline"><a class="l_k" href="functions/int.html">int</a></code>), in general it cannot.  IV might be something larger.  Even
worse the situation is with more specific types (defined by Perl's
configuration step in <i>config.h</i>):</p>
<pre class="verbatim"><ol><li>   <span class="w">Uid_t</span> <span class="w">who</span> = ...<span class="sc">;</span></li><li>   <a class="l_k" href="functions/printf.html">printf</a><span class="s">(</span><span class="q">&quot;who = %d\n&quot;</span><span class="cm">,</span> <span class="w">who</span><span class="s">)</span><span class="sc">;</span>    <span class="q">/* BAD */</span></li></ol></pre><p>The problem here is that Uid_t might be not only not <code class="inline"><a class="l_k" href="functions/int.html">int</a></code>-wide but it
might also be unsigned, in which case large uids would be printed as
negative values.</p>
<p>There is no simple solution to this because of printf()'s limited
intelligence, but for many types the right format is available as with
either 'f' or '_f' suffix, for example:</p>
<pre class="verbatim"><ol><li>   IVdf /* IV in decimal */</li><li>   UVxf /* UV is hexadecimal */</li><li></li><li>   printf("i = %"IVdf"\n", i); /* The IVdf is a string constant. */</li><li></li><li>   Uid_t_f /* Uid_t in decimal */</li><li></li><li>   printf("who = %"Uid_t_f"\n", who);</li></ol></pre><p>Or you can try casting to a "wide enough" type:</p>
<pre class="verbatim"><ol><li>   printf("i = %"IVdf"\n", (IV)something_very_small_and_signed);</li></ol></pre><p>See <a href="perlguts.html#Formatted-Printing-of-Size_t-and-SSize_t">Formatted Printing of Size_t and SSize_t in perlguts</a> for how to
print those.</p>
<p>Also remember that the <code class="inline"><span class="i">%p</span></code>
 format really does require a void pointer:</p>
<pre class="verbatim"><ol><li>   U8* p = ...;</li><li>   printf("p = %p\n", (void*)p);</li></ol></pre><p>The gcc option <code class="inline">-<span class="w">Wformat</span></code>
 scans for such problems.</p>
</li>
<li>
<p>Blindly using variadic macros</p>
<p>gcc has had them for a while with its own syntax, and C99 brought them
with a standardized syntax.  Don't use the former, and use the latter
only if the HAS_C99_VARIADIC_MACROS is defined.</p>
</li>
<li>
<p>Blindly passing va_list</p>
<p>Not all platforms support passing va_list to further varargs (stdarg)
functions.  The right thing to do is to copy the va_list using the
Perl_va_copy() if the NEED_VA_COPY is defined.</p>
</li>
<li>
<p>Using gcc statement expressions</p>
<pre class="verbatim"><ol><li>   <span class="w">val</span> = <span class="s">(</span><span class="s">{</span>...<span class="sc">;</span>...<span class="sc">;</span>...<span class="s">}</span><span class="s">)</span><span class="sc">;</span>    <span class="q">/* BAD */</span></li></ol></pre><p>While a nice extension, it's not portable.  The Perl code does
admittedly use them if available to gain some extra speed (essentially
as a funky form of inlining), but you shouldn't.</p>
</li>
<li>
<p>Binding together several statements in a macro</p>
<p>Use the macros STMT_START and STMT_END.</p>
<pre class="verbatim"><ol><li>   <span class="i">STMT_START</span> <span class="s">{</span></li><li>      ...</li><li>   <span class="s">}</span> <span class="w">STMT_END</span></li></ol></pre></li>
<li>
<p>Testing for operating systems or versions when should be testing for
features</p>
<pre class="verbatim"><ol><li>  <span class="c">#ifdef __FOONIX__    /* BAD */</span></li><li>  <span class="w">foo</span> = <span class="i">quux</span><span class="s">(</span><span class="s">)</span><span class="sc">;</span></li><li>  <span class="c">#endif</span></li></ol></pre><p>Unless you know with 100% certainty that quux() is only ever available
for the "Foonix" operating system <b>and</b> that is available <b>and</b>
correctly working for <b>all</b> past, present, <b>and</b> future versions of
"Foonix", the above is very wrong.  This is more correct (though still
not perfect, because the below is a compile-time check):</p>
<pre class="verbatim"><ol><li>  <span class="c">#ifdef HAS_QUUX</span></li><li>  <span class="w">foo</span> = <span class="i">quux</span><span class="s">(</span><span class="s">)</span><span class="sc">;</span></li><li>  <span class="c">#endif</span></li></ol></pre><p>How does the HAS_QUUX become defined where it needs to be?  Well, if
Foonix happens to be Unixy enough to be able to run the Configure
script, and Configure has been taught about detecting and testing
quux(), the HAS_QUUX will be correctly defined.  In other platforms, the
corresponding configuration step will hopefully do the same.</p>
<p>In a pinch, if you cannot wait for Configure to be educated, or if you
have a good hunch of where quux() might be available, you can
temporarily try the following:</p>
<pre class="verbatim"><ol><li>  <span class="c">#if (defined(__FOONIX__) || defined(__BARNIX__))</span></li><li>  <span class="c"># define HAS_QUUX</span></li><li>  <span class="c">#endif</span></li><li></li><li>  ...</li><li></li><li>  <span class="c">#ifdef HAS_QUUX</span></li><li>  <span class="w">foo</span> = <span class="i">quux</span><span class="s">(</span><span class="s">)</span><span class="sc">;</span></li><li>  <span class="c">#endif</span></li></ol></pre><p>But in any case, try to keep the features and operating systems
separate.</p>
<p>A good resource on the predefined macros for various operating
systems, compilers, and so forth is
<a href="http://sourceforge.net/p/predef/wiki/Home/">http://sourceforge.net/p/predef/wiki/Home/</a></p>
</li>
<li>
<p>Assuming the contents of static memory pointed to by the return values
of Perl wrappers for C library functions doesn't change.  Many C library
functions return pointers to static storage that can be overwritten by
subsequent calls to the same or related functions.  Perl has
light-weight wrappers for some of these functions, and which don't make
copies of the static memory.  A good example is the interface to the
environment variables that are in effect for the program.  Perl has
<code class="inline"><span class="w">PerlEnv_getenv</span></code>
 to get values from the environment.  But the return is
a pointer to static memory in the C library.  If you are using the value
to immediately test for something, that's fine, but if you save the
value and expect it to be unchanged by later processing, you would be
wrong, but perhaps you wouldn't know it because different C library
implementations behave differently, and the one on the platform you're
testing on might work for your situation.  But on some platforms, a
subsequent call to <code class="inline"><span class="w">PerlEnv_getenv</span></code>
 or related function WILL overwrite
the memory that your first call points to.  This has led to some
hard-to-debug problems.  Do a <a href="perlapi.html#savepv">savepv in perlapi</a> to make a copy, thus
avoiding these problems.  You will have to free the copy when you're
done to avoid memory leaks.  If you don't have control over when it gets
freed, you'll need to make the copy in a mortal scalar, like so:</p>
<pre class="verbatim"><ol><li> <a class="l_k" href="functions/if.html">if</a> <span class="s">(</span><span class="s">(</span><span class="q">s = PerlEnv_getenv(&quot;foo&quot;) ==</span> <span class="w">NULL</span><span class="s">)</span> <span class="s">{</span></li><li>    ... <span class="q">/* handle NULL case */</span></li><li> <span class="s">}</span></li><li> <a class="l_k" href="functions/else.html">else</a> <span class="s">{</span></li><li>     <span class="q">s = SvPVX(sv_2mortal(newSVpv(s, 0)));</span></li><li> <span class="q"> }</span></li></ol></pre><p>The above example works only if <code class="inline"><span class="q">&quot;s&quot;</span></code>
 is <code class="inline"><span class="w">NUL</span></code>
-terminated; otherwise
you have to pass its length to <code class="inline"><span class="w">newSVpv</span></code>
.</p>
</li>
</ul>
<a name="Problematic-System-Interfaces"></a><h2>Problematic System Interfaces</h2>
<ul>
<li>
<p>malloc(0), realloc(0), calloc(0, 0) are non-portable.  To be portable
allocate at least one byte.  (In general you should rarely need to work
at this low level, but instead use the various malloc wrappers.)</p>
</li>
<li>
<p>snprintf() - the return type is unportable.  Use my_snprintf() instead.</p>
</li>
</ul>
<a name="Security-problems"></a><h2>Security problems</h2>
<p>Last but not least, here are various tips for safer coding.
See also <a href="perlclib.html">perlclib</a> for libc/stdio replacements one should use.</p>
<ul>
<li>
<p>Do not use gets()</p>
<p>Or we will publicly ridicule you.  Seriously.</p>
</li>
<li>
<p>Do not use tmpfile()</p>
<p>Use mkstemp() instead.</p>
</li>
<li>
<p>Do not use strcpy() or strcat() or strncpy() or strncat()</p>
<p>Use my_strlcpy() and my_strlcat() instead: they either use the native
implementation, or Perl's own implementation (borrowed from the public
domain implementation of INN).</p>
</li>
<li>
<p>Do not use sprintf() or vsprintf()</p>
<p>If you really want just plain byte strings, use my_snprintf() and
my_vsnprintf() instead, which will try to use snprintf() and
vsnprintf() if those safer APIs are available.  If you want something
fancier than a plain byte string, use
<a href="perlapi.html#form">Perl_form ()</a> or SVs and
<a href="perlapi.html#sv_catpvf">Perl_sv_catpvf() </a>.</p>
<p>Note that glibc <code class="inline"><a class="l_k" href="functions/printf.html">printf()</a></code>, <code class="inline"><a class="l_k" href="functions/sprintf.html">sprintf()</a></code>, etc. are buggy before glibc
version 2.17.  They won't allow a <code class="inline"><span class="i">%.</span><span class="q">s</span></code>
 format with a precision to
create a string that isn't valid UTF-8 if the current underlying locale
of the program is UTF-8.  What happens is that the <code class="inline"><span class="i">%s</span></code>
 and its operand are
simply skipped without any notice.
<a href="https://sourceware.org/bugzilla/show_bug.cgi?id=6530">https://sourceware.org/bugzilla/show_bug.cgi?id=6530</a>.</p>
</li>
<li>
<p>Do not use atoi()</p>
<p>Use grok_atoUV() instead.  atoi() has ill-defined behavior on overflows,
and cannot be used for incremental parsing.  It is also affected by locale,
which is bad.</p>
</li>
<li>
<p>Do not use strtol() or strtoul()</p>
<p>Use grok_atoUV() instead.  strtol() or strtoul() (or their IV/UV-friendly
macro disguises, Strtol() and Strtoul(), or Atol() and Atoul() are
affected by locale, which is bad.</p>
</li>
</ul>
<a name="DEBUGGING"></a><h1>DEBUGGING</h1>
<p>You can compile a special debugging version of Perl, which allows you
to use the <code class="inline">-<span class="w">D</span></code>
 option of Perl to tell more about what Perl is doing.
But sometimes there is no alternative than to dive in with a debugger,
either to see the stack trace of a core dump (very useful in a bug
report), or trying to figure out what went wrong before the core dump
happened, or how did we end up having wrong or unexpected results.</p>
<a name="Poking-at-Perl"></a><h2>Poking at Perl</h2>
<p>To really poke around with Perl, you'll probably want to build Perl for
debugging, like this:</p>
<pre class="verbatim"><ol><li>    .<span class="q">/Configure -d -DDEBUGGING</span></li><li>    <span class="q">    make</span></li></ol></pre><p><code class="inline">-<span class="w">DDEBUGGING</span></code>
 turns on the C compiler's <code class="inline">-g</code>
 flag to have it produce
debugging information which will allow us to step through a running
program, and to see in which C function we are at (without the debugging
information we might see only the numerical addresses of the functions,
which is not very helpful). It will also turn on the <code class="inline"><span class="w">DEBUGGING</span></code>

compilation symbol which enables all the internal debugging code in Perl.
There are a whole bunch of things you can debug with this: <a href="perlrun.html">perlrun</a>
lists them all, and the best way to find out about them is to play about
with them.  The most useful options are probably</p>
<pre class="verbatim"><ol><li>    <span class="w">l</span>  <span class="w">Context</span> <span class="s">(</span><span class="w">loop</span><span class="s">)</span> <span class="w">stack</span> <span class="w">processing</span></li><li>    <span class="q">s  Stack snapshots (with v, displays all stacks)</span></li><li>    <span class="q">    t  Trace execution</span></li><li>    <span class="q">    o  Method and overloading resolution</span></li><li>    <span class="q">    c  String/numeric conversions</span></li></ol></pre><p>For example</p>
<pre class="verbatim"><ol><li>    <span class="i">$ perl</span> -<span class="w">Dst</span> -e <span class="q">&#39;$a + 1&#39;</span></li><li>    ....</li><li>    <span class="s">(</span>-e<span class="co">:</span><span class="n">1</span><span class="s">)</span>	<span class="i">gvsv</span><span class="s">(</span><span class="w">main::a</span><span class="s">)</span></li><li>        <span class="cm">=&gt;</span>  <span class="i">UNDEF</span></li><li>    <span class="s">(</span>-e<span class="co">:</span><span class="n">1</span><span class="s">)</span>	<span class="i">const</span><span class="s">(</span><span class="i">IV</span><span class="s">(</span><span class="n">1</span><span class="s">)</span><span class="s">)</span></li><li>        <span class="cm">=&gt;</span>  <span class="w">UNDEF</span>  <span class="i">IV</span><span class="s">(</span><span class="n">1</span><span class="s">)</span></li><li>    <span class="s">(</span>-e<span class="co">:</span><span class="n">1</span><span class="s">)</span>	<span class="w">add</span></li><li>        <span class="cm">=&gt;</span>  <span class="i">NV</span><span class="s">(</span><span class="n">1</span><span class="s">)</span></li></ol></pre><p>Some of the functionality of the debugging code can be achieved with a
non-debugging perl by using XS modules:</p>
<pre class="verbatim"><ol><li>    -<span class="w">Dr</span> <span class="cm">=&gt;</span> <a class="l_k" href="functions/use.html">use</a> <span class="w">re</span> <span class="q">&#39;debug&#39;</span></li><li>    -<span class="w">Dx</span> <span class="cm">=&gt;</span> <a class="l_k" href="functions/use.html">use</a> <span class="w">O</span> <span class="q">&#39;Debug&#39;</span></li></ol></pre><a name="Using-a-source-level-debugger"></a><h2>Using a source-level debugger</h2>
<p>If the debugging output of <code class="inline">-<span class="w">D</span></code>
 doesn't help you, it's time to step
through perl's execution with a source-level debugger.</p>
<ul>
<li>
<p>We'll use <code class="inline"><span class="w">gdb</span></code>
 for our examples here; the principles will apply to
any debugger (many vendors call their debugger <code class="inline"><span class="w">dbx</span></code>
), but check the
manual of the one you're using.</p>
</li>
</ul>
<p>To fire up the debugger, type</p>
<pre class="verbatim"><ol><li>    gdb ./perl</li></ol></pre><p>Or if you have a core dump:</p>
<pre class="verbatim"><ol><li>    gdb ./perl core</li></ol></pre><p>You'll want to do that in your Perl source tree so the debugger can
read the source code.  You should see the copyright message, followed by
the prompt.</p>
<pre class="verbatim"><ol><li>    <span class="s">(</span><span class="w">gdb</span><span class="s">)</span></li></ol></pre><p><code class="inline"><span class="w">help</span></code>
 will get you into the documentation, but here are the most
useful commands:</p>
<ul>
<li><a name="*-run-%5bargs%5d"></a><b>run [args]</b>
<p>Run the program with the given arguments.</p>
</li>
<li><a name="*-break-function_name"></a><b>break function_name</b>
</li>
<li><a name="*-break-source.c%3axxx"></a><b>break source.c:xxx</b>
<p>Tells the debugger that we'll want to pause execution when we reach
either the named function (but see <a href="perlguts.html#Internal-Functions">Internal Functions in perlguts</a>!) or
the given line in the named source file.</p>
</li>
<li><a name="*-step"></a><b>step</b>
<p>Steps through the program a line at a time.</p>
</li>
<li><a name="*-next"></a><b>next</b>
<p>Steps through the program a line at a time, without descending into
functions.</p>
</li>
<li><a name="*-continue"></a><b>continue</b>
<p>Run until the next breakpoint.</p>
</li>
<li><a name="*-finish"></a><b>finish</b>
<p>Run until the end of the current function, then stop again.</p>
</li>
<li><a name="*-'enter'"></a><b>'enter'</b>
<p>Just pressing Enter will do the most recent operation again - it's a
blessing when stepping through miles of source code.</p>
</li>
<li><a name="*-ptype"></a><b>ptype</b>
<p>Prints the C definition of the argument given.</p>
<pre class="verbatim"><ol><li>  <span class="s">(</span><span class="w">gdb</span><span class="s">)</span> <span class="w">ptype</span> <span class="w">PL_op</span></li><li>  <span class="w">type</span> = <span class="w">struct</span> <span class="i">op</span> <span class="s">{</span></li><li>      <span class="w">OP</span> *<span class="w">op_next</span><span class="sc">;</span></li><li>      <span class="w">OP</span> *<span class="w">op_sibparent</span><span class="sc">;</span></li><li>      <span class="w">OP</span> *<span class="s">(</span><span class="i">*op_ppaddr</span><span class="s">)</span><span class="s">(</span><span class="w">void</span><span class="s">)</span><span class="sc">;</span></li><li>      <span class="w">PADOFFSET</span> <span class="w">op_targ</span><span class="sc">;</span></li><li>      <span class="w">unsigned</span> <a class="l_k" href="functions/int.html">int</a> <span class="w">op_type</span> <span class="co">:</span> <span class="n">9</span><span class="sc">;</span></li><li>      <span class="w">unsigned</span> <a class="l_k" href="functions/int.html">int</a> <span class="w">op_opt</span> <span class="co">:</span> <span class="n">1</span><span class="sc">;</span></li><li>      <span class="w">unsigned</span> <a class="l_k" href="functions/int.html">int</a> <span class="w">op_slabbed</span> <span class="co">:</span> <span class="n">1</span><span class="sc">;</span></li><li>      <span class="w">unsigned</span> <a class="l_k" href="functions/int.html">int</a> <span class="w">op_savefree</span> <span class="co">:</span> <span class="n">1</span><span class="sc">;</span></li><li>      <span class="w">unsigned</span> <a class="l_k" href="functions/int.html">int</a> <span class="w">op_static</span> <span class="co">:</span> <span class="n">1</span><span class="sc">;</span></li><li>      <span class="w">unsigned</span> <a class="l_k" href="functions/int.html">int</a> <span class="w">op_folded</span> <span class="co">:</span> <span class="n">1</span><span class="sc">;</span></li><li>      <span class="w">unsigned</span> <a class="l_k" href="functions/int.html">int</a> <span class="w">op_spare</span> <span class="co">:</span> <span class="n">2</span><span class="sc">;</span></li><li>      <span class="w">U8</span> <span class="w">op_flags</span><span class="sc">;</span></li><li>      <span class="w">U8</span> <span class="w">op_private</span><span class="sc">;</span></li><li>  <span class="s">}</span> <span class="i">*</span></li></ol></pre></li>
<li><a name="*-print"></a><b>print</b>
<p>Execute the given C code and print its results.  <b>WARNING</b>: Perl makes
heavy use of macros, and <i>gdb</i> does not necessarily support macros
(see later <a href="#gdb-macro-support">gdb macro support</a>).  You'll have to substitute them
yourself, or to invoke cpp on the source code files (see <a href="#The-.i-Targets">The .i Targets</a>) So, for instance, you can't say</p>
<pre class="verbatim"><ol><li>    <a class="l_k" href="functions/print.html">print</a> <span class="i">SvPV_nolen</span><span class="s">(</span><span class="w">sv</span><span class="s">)</span></li></ol></pre><p>but you have to say</p>
<pre class="verbatim"><ol><li>    <a class="l_k" href="functions/print.html">print</a> <span class="i">Perl_sv_2pv_nolen</span><span class="s">(</span><span class="w">sv</span><span class="s">)</span></li></ol></pre></li>
</ul>
<p>You may find it helpful to have a "macro dictionary", which you can
produce by saying <code class="inline"><span class="w">cpp</span> -<span class="w">dM</span> <span class="w">perl</span>.<span class="w">c</span> | <a class="l_k" href="functions/sort.html">sort</a></code>
.  Even then, <i>cpp</i> won't
recursively apply those macros for you.</p>
<a name="gdb-macro-support"></a><h2>gdb macro support</h2>
<p>Recent versions of <i>gdb</i> have fairly good macro support, but in order
to use it you'll need to compile perl with macro definitions included
in the debugging information.  Using <i>gcc</i> version 3.1, this means
configuring with <code class="inline">-<span class="w">Doptimize</span>=-<span class="w">g3</span></code>
.  Other compilers might use a
different switch (if they support debugging macros at all).</p>
<a name="Dumping-Perl-Data-Structures"></a><h2>Dumping Perl Data Structures</h2>
<p>One way to get around this macro hell is to use the dumping functions
in <i>dump.c</i>; these work a little like an internal
<a href="Devel/Peek.html">Devel::Peek</a>, but they also cover OPs and other
structures that you can't get at from Perl.  Let's take an example.
We'll use the <code class="inline"><span class="i">$a</span> = <span class="i">$b</span> + <span class="i">$c</span></code>
 we used before, but give it a bit of
context: <code class="inline"><span class="i">$b</span> = <span class="q">&quot;6XXXX&quot;</span><span class="sc">;</span> <span class="i">$c</span> = <span class="n">2.3</span><span class="sc">;</span></code>
.  Where's a good place to stop and
poke around?</p>
<p>What about <code class="inline"><span class="w">pp_add</span></code>
, the function we examined earlier to implement the
<code class="inline">+</code>
 operator:</p>
<pre class="verbatim"><ol><li>    (gdb) break Perl_pp_add</li><li>    Breakpoint 1 at 0x46249f: file pp_hot.c, line 309.</li></ol></pre><p>Notice we use <code class="inline"><span class="w">Perl_pp_add</span></code>
 and not <code class="inline"><span class="w">pp_add</span></code>
 - see
<a href="perlguts.html#Internal-Functions">Internal Functions in perlguts</a>.  With the breakpoint in place, we can
run our program:</p>
<pre class="verbatim"><ol><li>    (gdb) run -e '$b = "6XXXX"; $c = 2.3; $a = $b + $c'</li></ol></pre><p>Lots of junk will go past as gdb reads in the relevant source files and
libraries, and then:</p>
<pre class="verbatim"><ol><li>    Breakpoint 1, Perl_pp_add () at pp_hot.c:309</li><li>    309         dSP; dATARGET; tryAMAGICbin(add,opASSIGN);</li><li>    (gdb) step</li><li>    311           dPOPTOPnnrl_ul;</li><li>    (gdb)</li></ol></pre><p>We looked at this bit of code before, and we said that
<code class="inline"><span class="w">dPOPTOPnnrl_ul</span></code>
 arranges for two <code class="inline"><span class="w">NV</span></code>
s to be placed into <code class="inline"><span class="w">left</span></code>
 and
<code class="inline"><span class="w">right</span></code>
 - let's slightly expand it:</p>
<pre class="verbatim"><ol><li> <span class="c">#define dPOPTOPnnrl_ul  NV right = POPn; \</span></li><li>                         <span class="w">SV</span> *<span class="w">leftsv</span> = <span class="w">TOPs</span><span class="sc">;</span> \</li><li>                         <span class="w">NV</span> <span class="w">left</span> = <span class="i">USE_LEFT</span><span class="s">(</span><span class="w">leftsv</span><span class="s">)</span> ? <span class="i">SvNV</span><span class="s">(</span><span class="w">leftsv</span><span class="s">)</span> <span class="co">:</span> <span class="n">0.0</span></li></ol></pre><p><code class="inline"><span class="w">POPn</span></code>
 takes the SV from the top of the stack and obtains its NV
either directly (if <code class="inline"><span class="w">SvNOK</span></code>
 is set) or by calling the <code class="inline"><span class="w">sv_2nv</span></code>

function.  <code class="inline"><span class="w">TOPs</span></code>
 takes the next SV from the top of the stack - yes,
<code class="inline"><span class="w">POPn</span></code>
 uses <code class="inline"><span class="w">TOPs</span></code>
 - but doesn't remove it.  We then use <code class="inline"><span class="w">SvNV</span></code>
 to
get the NV from <code class="inline"><span class="w">leftsv</span></code>
 in the same way as before - yes, <code class="inline"><span class="w">POPn</span></code>
 uses
<code class="inline"><span class="w">SvNV</span></code>
.</p>
<p>Since we don't have an NV for <code class="inline"><span class="i">$b</span></code>
, we'll have to use <code class="inline"><span class="w">sv_2nv</span></code>
 to
convert it.  If we step again, we'll find ourselves there:</p>
<pre class="verbatim"><ol><li>    <span class="s">(</span><span class="w">gdb</span><span class="s">)</span> <span class="w">step</span></li><li>    <span class="w">Perl_sv_2nv</span> <span class="s">(</span><span class="w">sv</span>=<span class="n">0xa0675d0</span><span class="s">)</span> <span class="w">at</span> <span class="w">sv</span>.<span class="w">c</span><span class="co">:</span><span class="n">1669</span></li><li>    <span class="n">1669</span>        <a class="l_k" href="functions/if.html">if</a> <span class="s">(</span>!<span class="w">sv</span><span class="s">)</span></li><li>    <span class="s">(</span><span class="w">gdb</span><span class="s">)</span></li></ol></pre><p>We can now use <code class="inline"><span class="w">Perl_sv_dump</span></code>
 to investigate the SV:</p>
<pre class="verbatim"><ol><li>    <span class="s">(</span><span class="w">gdb</span><span class="s">)</span> <a class="l_k" href="functions/print.html">print</a> <span class="i">Perl_sv_dump</span><span class="s">(</span><span class="w">sv</span><span class="s">)</span></li><li>    <span class="w">SV</span> = <span class="i">PV</span><span class="s">(</span><span class="n">0xa057cc0</span><span class="s">)</span> <span class="w">at</span> <span class="n">0xa0675d0</span></li><li>    <span class="w">REFCNT</span> = <span class="n">1</span></li><li>    <span class="w">FLAGS</span> = <span class="s">(</span><span class="w">POK</span><span class="cm">,</span><span class="w">pPOK</span><span class="s">)</span></li><li>    <span class="w">PV</span> = <span class="n">0xa06a510</span> <span class="q">&quot;6XXXX&quot;</span>\<span class="n">0</span></li><li>    <span class="w">CUR</span> = <span class="n">5</span></li><li>    <span class="w">LEN</span> = <span class="n">6</span></li><li>    <span class="i">$1</span> = <span class="w">void</span></li></ol></pre><p>We know we're going to get <code class="inline"><span class="n">6</span></code>
 from this, so let's finish the
subroutine:</p>
<pre class="verbatim"><ol><li>    (gdb) finish</li><li>    Run till exit from #0  Perl_sv_2nv (sv=0xa0675d0) at sv.c:1671</li><li>    0x462669 in Perl_pp_add () at pp_hot.c:311</li><li>    311           dPOPTOPnnrl_ul;</li></ol></pre><p>We can also dump out this op: the current op is always stored in
<code class="inline"><span class="w">PL_op</span></code>
, and we can dump it with <code class="inline"><span class="w">Perl_op_dump</span></code>
.  This'll give us
similar output to <a href="B/Debug.html">B::Debug</a>.</p>
<pre class="verbatim"><ol><li>    <span class="s">(</span><span class="w">gdb</span><span class="s">)</span> <a class="l_k" href="functions/print.html">print</a> <span class="i">Perl_op_dump</span><span class="s">(</span><span class="w">PL_op</span><span class="s">)</span></li><li>    <span class="s">{</span></li><li>    <span class="n">13</span>  <span class="w">TYPE</span> = <span class="w">add</span>  ==<span class="cm">=&gt;</span> <span class="n">14</span></li><li>        <span class="w">TARG</span> = <span class="n">1</span></li><li>        <span class="w">FLAGS</span> = <span class="s">(</span><span class="w">SCALAR</span><span class="cm">,</span><span class="w">KIDS</span><span class="s">)</span></li><li>        <span class="s">{</span></li><li>            <span class="w">TYPE</span> = <span class="w">null</span>  ==<span class="cm">=&gt;</span> <span class="s">(</span><span class="n">12</span><span class="s">)</span></li><li>              <span class="s">(</span><span class="w">was</span> <span class="w">rv2sv</span><span class="s">)</span></li><li>            <span class="w">FLAGS</span> = <span class="s">(</span><span class="w">SCALAR</span><span class="cm">,</span><span class="w">KIDS</span><span class="s">)</span></li><li>            <span class="s">{</span></li><li>    <span class="n">11</span>          <span class="w">TYPE</span> = <span class="w">gvsv</span>  ==<span class="cm">=&gt;</span> <span class="n">12</span></li><li>                <span class="w">FLAGS</span> = <span class="s">(</span><span class="w">SCALAR</span><span class="s">)</span></li><li>                <span class="w">GV</span> = <span class="w">main::b</span></li><li>            <span class="s">}</span></li><li>        <span class="s">}</span></li></ol></pre><p># finish this later #</p>
<a name="Using-gdb-to-look-at-specific-parts-of-a-program"></a><h2>Using gdb to look at specific parts of a program</h2>
<p>With the example above, you knew to look for <code class="inline"><span class="w">Perl_pp_add</span></code>
, but what if
there were multiple calls to it all over the place, or you didn't know what
the op was you were looking for?</p>
<p>One way to do this is to inject a rare call somewhere near what you're looking
for.  For example, you could add <code class="inline"><a class="l_k" href="functions/study.html">study</a></code> before your method:</p>
<pre class="verbatim"><ol><li>    <a class="l_k" href="functions/study.html">study</a><span class="sc">;</span></li></ol></pre><p>And in gdb do:</p>
<pre class="verbatim"><ol><li>    <span class="s">(</span><span class="w">gdb</span><span class="s">)</span> <a class="l_k" href="functions/break.html">break</a> <span class="w">Perl_pp_study</span></li></ol></pre><p>And then step until you hit what you're
looking for.  This works well in a loop
if you want to only break at certain iterations:</p>
<pre class="verbatim"><ol><li>    <a class="l_k" href="functions/for.html">for</a> <a class="l_k" href="functions/my.html">my</a> <span class="i">$c</span> <span class="s">(</span><span class="n">1</span>..<span class="n">100</span><span class="s">)</span> <span class="s">{</span></li><li>        <a class="l_k" href="functions/study.html">study</a> <a class="l_k" href="functions/if.html">if</a> <span class="i">$c</span> == <span class="n">50</span><span class="sc">;</span></li><li>    <span class="s">}</span></li></ol></pre><a name="Using-gdb-to-look-at-what-the-parser%2flexer-are-doing"></a><h2>Using gdb to look at what the parser/lexer are doing</h2>
<p>If you want to see what perl is doing when parsing/lexing your code, you can
use <code class="inline">BEGIN <span class="s">{</span><span class="s">}</span></code>
:</p>
<pre class="verbatim"><ol><li>    <a class="l_k" href="functions/print.html">print</a> <span class="q">&quot;Before\n&quot;</span><span class="sc">;</span></li><li>    <a class="l_k" href="functions/BEGIN.html">BEGIN</a> <span class="s">{</span> <a class="l_k" href="functions/study.html">study</a><span class="sc">;</span> <span class="s">}</span></li><li>    <a class="l_k" href="functions/print.html">print</a> <span class="q">&quot;After\n&quot;</span><span class="sc">;</span></li></ol></pre><p>And in gdb:</p>
<pre class="verbatim"><ol><li>    <span class="s">(</span><span class="w">gdb</span><span class="s">)</span> <a class="l_k" href="functions/break.html">break</a> <span class="w">Perl_pp_study</span></li></ol></pre><p>If you want to see what the parser/lexer is doing inside of <code class="inline">if</code>
 blocks and
the like you need to be a little trickier:</p>
<pre class="verbatim"><ol><li>    <a class="l_k" href="functions/if.html">if</a> <span class="s">(</span><span class="i">$a</span> &amp;&amp; <span class="i">$b</span> &amp;&amp; <a class="l_k" href="functions/do.html">do</a> <span class="s">{</span> <a class="l_k" href="functions/BEGIN.html">BEGIN</a> <span class="s">{</span> <a class="l_k" href="functions/study.html">study</a> <span class="s">}</span> <span class="n">1</span> <span class="s">}</span> &amp;&amp; <span class="i">$c</span><span class="s">)</span> <span class="s">{</span> ... <span class="s">}</span></li></ol></pre><a name="SOURCE-CODE-STATIC-ANALYSIS"></a><h1>SOURCE CODE STATIC ANALYSIS</h1>
<p>Various tools exist for analysing C source code <b>statically</b>, as
opposed to <b>dynamically</b>, that is, without executing the code.  It is
possible to detect resource leaks, undefined behaviour, type
mismatches, portability problems, code paths that would cause illegal
memory accesses, and other similar problems by just parsing the C code
and looking at the resulting graph, what does it tell about the
execution and data flows.  As a matter of fact, this is exactly how C
compilers know to give warnings about dubious code.</p>
<a name="lint"></a><h2>lint</h2>
<p>The good old C code quality inspector, <code class="inline"><span class="w">lint</span></code>
, is available in several
platforms, but please be aware that there are several different
implementations of it by different vendors, which means that the flags
are not identical across different platforms.</p>
<p>There is a <code class="inline"><span class="w">lint</span></code>
 target in Makefile, but you may have to
diddle with the flags (see above).</p>
<a name="Coverity"></a><h2>Coverity</h2>
<p>Coverity (<a href="http://www.coverity.com/">http://www.coverity.com/</a>) is a product similar to lint and as
a testbed for their product they periodically check several open source
projects, and they give out accounts to open source developers to the
defect databases.</p>
<p>There is Coverity setup for the perl5 project:
<a href="https://scan.coverity.com/projects/perl5">https://scan.coverity.com/projects/perl5</a></p>
<a name="HP-UX-cadvise-(Code-Advisor)"></a><h2>HP-UX cadvise (Code Advisor)</h2>
<p>HP has a C/C++ static analyzer product for HP-UX caller Code Advisor.
(Link not given here because the URL is horribly long and seems horribly
unstable; use the search engine of your choice to find it.)  The use of
the <code class="inline"><span class="w">cadvise_cc</span></code>
 recipe with <code class="inline"><span class="w">Configure</span> ... -<span class="w">Dcc</span>=.<span class="q">/cadvise_cc</span></code>

(see cadvise "User Guide") is recommended; as is the use of <code class="inline">+<span class="w">wall</span></code>
.</p>
<a name="cpd-(cut-and-paste-detector)"></a><h2>cpd (cut-and-paste detector)</h2>
<p>The cpd tool detects cut-and-paste coding.  If one instance of the
cut-and-pasted code changes, all the other spots should probably be
changed, too.  Therefore such code should probably be turned into a
subroutine or a macro.</p>
<p>cpd (<a href="http://pmd.sourceforge.net/cpd.html">http://pmd.sourceforge.net/cpd.html</a>) is part of the pmd project
(<a href="http://pmd.sourceforge.net/">http://pmd.sourceforge.net/</a>).  pmd was originally written for static
analysis of Java code, but later the cpd part of it was extended to
parse also C and C++.</p>
<p>Download the pmd-bin-X.Y.zip () from the SourceForge site, extract the
pmd-X.Y.jar from it, and then run that on source code thusly:</p>
<pre class="verbatim"><ol><li>  <span class="w">java</span> -<span class="w">cp</span> <span class="w">pmd</span><a class="l_k" href="functions/-X.html">-X</a>.<span class="w">Y</span>.<span class="w">jar</span> <span class="w">net</span>.<span class="w">sourceforge</span>.<span class="w">pmd</span>.<span class="w">cpd</span>.<span class="w">CPD</span> \</li><li>   --<span class="w">minimum</span>-<span class="w">tokens</span> <span class="n">100</span> --<span class="w">files</span> /<span class="w">some</span>/<span class="w">where</span>/<span class="w">src</span> --<span class="w">language</span> <span class="w">c</span> &gt; <span class="w">cpd</span>.<span class="w">txt</span></li></ol></pre><p>You may run into memory limits, in which case you should use the -Xmx
option:</p>
<pre class="verbatim"><ol><li>  <span class="w">java</span> -<span class="w">Xmx512M</span> ...</li></ol></pre><a name="gcc-warnings"></a><h2>gcc warnings</h2>
<p>Though much can be written about the inconsistency and coverage
problems of gcc warnings (like <code class="inline">-<span class="w">Wall</span></code>
 not meaning "all the warnings",
or some common portability problems not being covered by <code class="inline">-<span class="w">Wall</span></code>
, or
<code class="inline">-<span class="w">ansi</span></code>
 and <code class="inline">-<span class="w">pedantic</span></code>
 both being a poorly defined collection of
warnings, and so forth), gcc is still a useful tool in keeping our
coding nose clean.</p>
<p>The <code class="inline">-<span class="w">Wall</span></code>
 is by default on.</p>
<p>The <code class="inline">-<span class="w">ansi</span></code>
 (and its sidekick, <code class="inline">-<span class="w">pedantic</span></code>
) would be nice to be on
always, but unfortunately they are not safe on all platforms, they can
for example cause fatal conflicts with the system headers (Solaris
being a prime example).  If Configure <code class="inline">-<span class="w">Dgccansipedantic</span></code>
 is used, the
<code class="inline"><span class="w">cflags</span></code>
 frontend selects <code class="inline">-<span class="w">ansi</span> -<span class="w">pedantic</span></code>
 for the platforms where
they are known to be safe.</p>
<p>Starting from Perl 5.9.4 the following extra flags are added:</p>
<ul>
<li>
<p><code class="inline">-<span class="w">Wendif</span>-<span class="w">labels</span></code>
</p>
</li>
<li>
<p><code class="inline">-<span class="w">Wextra</span></code>
</p>
</li>
<li>
<p><code class="inline">-<span class="w">Wdeclaration</span>-<span class="w">after</span>-<span class="w">statement</span></code>
</p>
</li>
</ul>
<p>The following flags would be nice to have but they would first need
their own Augean stablemaster:</p>
<ul>
<li>
<p><code class="inline">-<span class="w">Wpointer</span>-<span class="w">arith</span></code>
</p>
</li>
<li>
<p><code class="inline">-<span class="w">Wshadow</span></code>
</p>
</li>
<li>
<p><code class="inline">-<span class="w">Wstrict</span>-<span class="w">prototypes</span></code>
</p>
</li>
</ul>
<p>The <code class="inline">-<span class="w">Wtraditional</span></code>
 is another example of the annoying tendency of gcc
to bundle a lot of warnings under one switch (it would be impossible to
deploy in practice because it would complain a lot) but it does contain
some warnings that would be beneficial to have available on their own,
such as the warning about string constants inside macros containing the
macro arguments: this behaved differently pre-ANSI than it does in
ANSI, and some C compilers are still in transition, AIX being an
example.</p>
<a name="Warnings-of-other-C-compilers"></a><h2>Warnings of other C compilers</h2>
<p>Other C compilers (yes, there <b>are</b> other C compilers than gcc) often
have their "strict ANSI" or "strict ANSI with some portability
extensions" modes on, like for example the Sun Workshop has its <code class="inline">-<span class="w">Xa</span></code>

mode on (though implicitly), or the DEC (these days, HP...) has its
<code class="inline">-<span class="w">std1</span></code>
 mode on.</p>
<a name="MEMORY-DEBUGGERS"></a><h1>MEMORY DEBUGGERS</h1>
<p><b>NOTE 1</b>: Running under older memory debuggers such as Purify,
valgrind or Third Degree greatly slows down the execution: seconds
become minutes, minutes become hours.  For example as of Perl 5.8.1, the
ext/Encode/t/Unicode.t takes extraordinarily long to complete under
e.g. Purify, Third Degree, and valgrind.  Under valgrind it takes more
than six hours, even on a snappy computer.  The said test must be doing
something that is quite unfriendly for memory debuggers.  If you don't
feel like waiting, that you can simply kill away the perl process.
Roughly valgrind slows down execution by factor 10, AddressSanitizer by
factor 2.</p>
<p><b>NOTE 2</b>: To minimize the number of memory leak false alarms (see
<a href="#PERL_DESTRUCT_LEVEL">PERL_DESTRUCT_LEVEL</a> for more information), you have to set the
environment variable PERL_DESTRUCT_LEVEL to 2.  For example, like this:</p>
<pre class="verbatim"><ol><li>    env PERL_DESTRUCT_LEVEL=2 valgrind ./perl -Ilib ...</li></ol></pre><p><b>NOTE 3</b>: There are known memory leaks when there are compile-time
errors within eval or require, seeing <code class="inline"><span class="w">S_doeval</span></code>
 in the call stack is
a good sign of these.  Fixing these leaks is non-trivial, unfortunately,
but they must be fixed eventually.</p>
<p><b>NOTE 4</b>: <a href="DynaLoader.html">DynaLoader</a> will not clean up after itself completely
unless Perl is built with the Configure option
<code class="inline">-<span class="w">Accflags</span>=-<span class="w">DDL_UNLOAD_ALL_AT_EXIT</span></code>
.</p>
<a name="valgrind"></a><h2>valgrind</h2>
<p>The valgrind tool can be used to find out both memory leaks and illegal
heap memory accesses.  As of version 3.3.0, Valgrind only supports Linux
on x86, x86-64 and PowerPC and Darwin (OS X) on x86 and x86-64.  The
special "test.valgrind" target can be used to run the tests under
valgrind.  Found errors and memory leaks are logged in files named
<i>testfile.valgrind</i> and by default output is displayed inline.</p>
<p>Example usage:</p>
<pre class="verbatim"><ol><li>    <span class="w">make</span> <span class="w">test</span>.<span class="w">valgrind</span></li></ol></pre><p>Since valgrind adds significant overhead, tests will take much longer to
run.  The valgrind tests support being run in parallel to help with this:</p>
<pre class="verbatim"><ol><li>    <span class="w">TEST_JOBS</span>=<span class="n">9</span> <span class="w">make</span> <span class="w">test</span>.<span class="w">valgrind</span></li></ol></pre><p>Note that the above two invocations will be very verbose as reachable
memory and leak-checking is enabled by default.  If you want to just see
pure errors, try:</p>
<pre class="verbatim"><ol><li>    <span class="w">VG_OPTS</span>=<span class="q">&#39;-q --leak-check=no --show-reachable=no&#39;</span> <span class="w">TEST_JOBS</span>=<span class="n">9</span> \</li><li>        <span class="w">make</span> <span class="w">test</span>.<span class="w">valgrind</span></li></ol></pre><p>Valgrind also provides a cachegrind tool, invoked on perl as:</p>
<pre class="verbatim"><ol><li>    <span class="w">VG_OPTS</span>=--<span class="w">tool</span>=<span class="w">cachegrind</span> <span class="w">make</span> <span class="w">test</span>.<span class="w">valgrind</span></li></ol></pre><p>As system libraries (most notably glibc) are also triggering errors,
valgrind allows to suppress such errors using suppression files.  The
default suppression file that comes with valgrind already catches a lot
of them.  Some additional suppressions are defined in <i>t/perl.supp</i>.</p>
<p>To get valgrind and for more information see</p>
<pre class="verbatim"><ol><li>    <span class="j">http:</span><span class="q">//</span><span class="w">valgrind</span>.<span class="w">org</span>/</li></ol></pre><a name="AddressSanitizer"></a><h2>AddressSanitizer</h2>
<p>AddressSanitizer is a clang and gcc extension, included in clang since
v3.1 and gcc since v4.8.  It checks illegal heap pointers, global
pointers, stack pointers and use after free errors, and is fast enough
that you can easily compile your debugging or optimized perl with it.
It does not check memory leaks though.  AddressSanitizer is available
for Linux, Mac OS X and soon on Windows.</p>
<p>To build perl with AddressSanitizer, your Configure invocation should
look like:</p>
<pre class="verbatim"><ol><li>    <span class="w">sh</span> <span class="w">Configure</span> -<span class="w">des</span> -<span class="w">Dcc</span>=<span class="w">clang</span> \</li><li>       -<span class="w">Accflags</span>=-<span class="w">faddress</span>-<span class="w">sanitizer</span> -<span class="w">Aldflags</span>=-<span class="w">faddress</span>-<span class="w">sanitizer</span> \</li><li>       -<span class="w">Alddlflags</span>=-<span class="w">shared</span>\ -<span class="w">faddress</span>-<span class="w">sanitizer</span></li></ol></pre><p>where these arguments mean:</p>
<ul>
<li><a name="*--Dcc%3dclang"></a><b>-Dcc=clang</b>
<p>This should be replaced by the full path to your clang executable if it
is not in your path.</p>
</li>
<li><a name="*--Accflags%3d-faddress-sanitizer"></a><b>-Accflags=-faddress-sanitizer</b>
<p>Compile perl and extensions sources with AddressSanitizer.</p>
</li>
<li><a name="*--Aldflags%3d-faddress-sanitizer"></a><b>-Aldflags=-faddress-sanitizer</b>
<p>Link the perl executable with AddressSanitizer.</p>
</li>
<li><a name="*--Alddlflags%3d-shared%5c--faddress-sanitizer"></a><b>-Alddlflags=-shared\ -faddress-sanitizer</b>
<p>Link dynamic extensions with AddressSanitizer.  You must manually
specify <code class="inline">-<span class="w">shared</span></code>
 because using <code class="inline">-<span class="w">Alddlflags</span>=-<span class="w">shared</span></code>
 will prevent
Configure from setting a default value for <code class="inline"><span class="w">lddlflags</span></code>
, which usually
contains <code class="inline">-<span class="w">shared</span></code>
 (at least on Linux).</p>
</li>
</ul>
<p>See also
<a href="http://code.google.com/p/address-sanitizer/wiki/AddressSanitizer">http://code.google.com/p/address-sanitizer/wiki/AddressSanitizer</a>.</p>
<a name="PROFILING"></a><h1>PROFILING</h1>
<p>Depending on your platform there are various ways of profiling Perl.</p>
<p>There are two commonly used techniques of profiling executables:
<i>statistical time-sampling</i> and <i>basic-block counting</i>.</p>
<p>The first method takes periodically samples of the CPU program counter,
and since the program counter can be correlated with the code generated
for functions, we get a statistical view of in which functions the
program is spending its time.  The caveats are that very small/fast
functions have lower probability of showing up in the profile, and that
periodically interrupting the program (this is usually done rather
frequently, in the scale of milliseconds) imposes an additional
overhead that may skew the results.  The first problem can be alleviated
by running the code for longer (in general this is a good idea for
profiling), the second problem is usually kept in guard by the
profiling tools themselves.</p>
<p>The second method divides up the generated code into <i>basic blocks</i>.
Basic blocks are sections of code that are entered only in the
beginning and exited only at the end.  For example, a conditional jump
starts a basic block.  Basic block profiling usually works by
<i>instrumenting</i> the code by adding <i>enter basic block #nnnn</i>
book-keeping code to the generated code.  During the execution of the
code the basic block counters are then updated appropriately.  The
caveat is that the added extra code can skew the results: again, the
profiling tools usually try to factor their own effects out of the
results.</p>
<a name="Gprof-Profiling"></a><h2>Gprof Profiling</h2>
<p><i>gprof</i> is a profiling tool available in many Unix platforms which
uses <i>statistical time-sampling</i>.  You can build a profiled version of
<i>perl</i> by compiling using gcc with the flag <code class="inline">-<span class="w">pg</span></code>
.  Either edit
<i>config.sh</i> or re-run <i>Configure</i>.  Running the profiled version of
Perl will create an output file called <i>gmon.out</i> which contains the
profiling data collected during the execution.</p>
<p>quick hint:</p>
<pre class="verbatim"><ol><li>    <span class="i">$ sh</span> <span class="w">Configure</span> -<span class="w">des</span> -<span class="w">Dusedevel</span> -<span class="w">Accflags</span>=<span class="q">&#39;-pg&#39;</span> \</li><li>        -<span class="w">Aldflags</span>=<span class="q">&#39;-pg&#39;</span> -<span class="w">Alddlflags</span>=<span class="q">&#39;-pg -shared&#39;</span> \</li><li>        &amp;&amp; <span class="w">make</span> <span class="w">perl</span></li><li>    <span class="i">$ .</span>/<span class="w">perl</span> ... <span class="c"># creates gmon.out in current directory</span></li><li>    <span class="i">$ gprof</span> .<span class="q">/perl &gt; out</span></li><li>    <span class="q">    $ less out</span></li></ol></pre><p>(you probably need to add <code class="inline">-<span class="w">shared</span></code>
 to the &lt;-Alddlflags&gt; line until RT
#118199 is resolved)</p>
<p>The <i>gprof</i> tool can then display the collected data in various ways.
Usually <i>gprof</i> understands the following options:</p>
<ul>
<li><a name="*--a"></a><b>-a</b>
<p>Suppress statically defined functions from the profile.</p>
</li>
<li><a name="*--b"></a><b>-b</b>
<p>Suppress the verbose descriptions in the profile.</p>
</li>
<li><a name="*--e-routine"></a><b>-e routine</b>
<p>Exclude the given routine and its descendants from the profile.</p>
</li>
<li><a name="*--f-routine"></a><b>-f routine</b>
<p>Display only the given routine and its descendants in the profile.</p>
</li>
<li><a name="*--s"></a><b>-s</b>
<p>Generate a summary file called <i>gmon.sum</i> which then may be given to
subsequent gprof runs to accumulate data over several runs.</p>
</li>
<li><a name="*--z"></a><b>-z</b>
<p>Display routines that have zero usage.</p>
</li>
</ul>
<p>For more detailed explanation of the available commands and output
formats, see your own local documentation of <i>gprof</i>.</p>
<a name="GCC-gcov-Profiling"></a><h2>GCC gcov Profiling</h2>
<p><i>basic block profiling</i> is officially available in gcc 3.0 and later.
You can build a profiled version of <i>perl</i> by compiling using gcc with
the flags <code class="inline">-<span class="w">fprofile</span>-<span class="w">arcs</span> -<span class="w">ftest</span>-<span class="w">coverage</span></code>
.  Either edit <i>config.sh</i>
or re-run <i>Configure</i>.</p>
<p>quick hint:</p>
<pre class="verbatim"><ol><li>    <span class="i">$ sh</span> <span class="w">Configure</span> -<span class="w">des</span> -<span class="w">Dusedevel</span> -<span class="w">Doptimize</span>=<span class="q">&#39;-g&#39;</span> \</li><li>        -<span class="w">Accflags</span>=<span class="q">&#39;-fprofile-arcs -ftest-coverage&#39;</span> \</li><li>        -<span class="w">Aldflags</span>=<span class="q">&#39;-fprofile-arcs -ftest-coverage&#39;</span> \</li><li>        -<span class="w">Alddlflags</span>=<span class="q">&#39;-fprofile-arcs -ftest-coverage -shared&#39;</span> \</li><li>        &amp;&amp; <span class="w">make</span> <span class="w">perl</span></li><li>    <span class="i">$ rm</span> -f <span class="w">regexec</span>.<span class="w">c</span>.<span class="w">gcov</span> <span class="w">regexec</span>.<span class="w">gcda</span></li><li>    <span class="i">$ .</span>/<span class="w">perl</span> ...</li><li>    <span class="i">$ gcov</span> <span class="w">regexec</span>.<span class="w">c</span></li><li>    <span class="i">$ less</span> <span class="w">regexec</span>.<span class="w">c</span>.<span class="w">gcov</span></li></ol></pre><p>(you probably need to add <code class="inline">-<span class="w">shared</span></code>
 to the &lt;-Alddlflags&gt; line until RT
#118199 is resolved)</p>
<p>Running the profiled version of Perl will cause profile output to be
generated.  For each source file an accompanying <i>.gcda</i> file will be
created.</p>
<p>To display the results you use the <i>gcov</i> utility (which should be
installed if you have gcc 3.0 or newer installed).  <i>gcov</i> is run on
source code files, like this</p>
<pre class="verbatim"><ol><li>    <span class="w">gcov</span> <span class="w">sv</span>.<span class="w">c</span></li></ol></pre><p>which will cause <i>sv.c.gcov</i> to be created.  The <i>.gcov</i> files contain
the source code annotated with relative frequencies of execution
indicated by "#" markers.  If you want to generate <i>.gcov</i> files for
all profiled object files, you can run something like this:</p>
<pre class="verbatim"><ol><li>    <a class="l_k" href="functions/for.html">for</a> <span class="w">file</span> <span class="w">in</span> <span class="q">`find . -name \*.gcno`</span></li><li>    <a class="l_k" href="functions/do.html">do</a> <span class="w">sh</span> -c <span class="q">&quot;cd `dirname $file` &amp;&amp; gcov `basename $file .gcno`&quot;</span></li><li>    <span class="w">done</span></li></ol></pre><p>Useful options of <i>gcov</i> include <code class="inline">-b</code>
 which will summarise the basic
block, branch, and function call coverage, and <code class="inline">-c</code>
 which instead of
relative frequencies will use the actual counts.  For more information
on the use of <i>gcov</i> and basic block profiling with gcc, see the
latest GNU CC manual.  As of gcc 4.8, this is at
<a href="http://gcc.gnu.org/onlinedocs/gcc/Gcov-Intro.html#Gcov-Intro">http://gcc.gnu.org/onlinedocs/gcc/Gcov-Intro.html#Gcov-Intro</a></p>
<a name="MISCELLANEOUS-TRICKS"></a><h1>MISCELLANEOUS TRICKS</h1>
<a name="PERL_DESTRUCT_LEVEL"></a><h2>PERL_DESTRUCT_LEVEL</h2>
<p>If you want to run any of the tests yourself manually using e.g.
valgrind, please note that by default perl <b>does not</b> explicitly
cleanup all the memory it has allocated (such as global memory arenas)
but instead lets the exit() of the whole program "take care" of such
allocations, also known as "global destruction of objects".</p>
<p>There is a way to tell perl to do complete cleanup: set the environment
variable PERL_DESTRUCT_LEVEL to a non-zero value.  The t/TEST wrapper
does set this to 2, and this is what you need to do too, if you don't
want to see the "global leaks": For example, for running under valgrind</p>
<pre class="verbatim"><ol><li>    <span class="w">env</span> <span class="w">PERL_DESTRUCT_LEVEL</span>=<span class="n">2</span> <span class="w">valgrind</span> .<span class="q">/perl -Ilib t/</span><span class="w">foo</span>/<span class="w">bar</span>.<span class="w">t</span></li></ol></pre><p>(Note: the mod_perl apache module uses also this environment variable
for its own purposes and extended its semantics.  Refer to the mod_perl
documentation for more information.  Also, spawned threads do the
equivalent of setting this variable to the value 1.)</p>
<p>If, at the end of a run you get the message <i>N scalars leaked</i>, you
can recompile with <code class="inline">-<span class="w">DDEBUG_LEAKING_SCALARS</span></code>
,
(<code class="inline"><span class="w">Configure</span> -<span class="w">Accflags</span>=-<span class="w">DDEBUG_LEAKING_SCALARS</span></code>
), which will cause the
addresses of all those leaked SVs to be dumped along with details as to
where each SV was originally allocated.  This information is also
displayed by Devel::Peek.  Note that the extra details recorded with
each SV increases memory usage, so it shouldn't be used in production
environments.  It also converts <code class="inline"><span class="i">new_SV</span><span class="s">(</span><span class="s">)</span></code>
 from a macro into a real
function, so you can use your favourite debugger to discover where
those pesky SVs were allocated.</p>
<p>If you see that you're leaking memory at runtime, but neither valgrind
nor <code class="inline">-<span class="w">DDEBUG_LEAKING_SCALARS</span></code>
 will find anything, you're probably
leaking SVs that are still reachable and will be properly cleaned up
during destruction of the interpreter.  In such cases, using the <code class="inline">-<span class="w">Dm</span></code>

switch can point you to the source of the leak.  If the executable was
built with <code class="inline">-<span class="w">DDEBUG_LEAKING_SCALARS</span></code>
, <code class="inline">-<span class="w">Dm</span></code>
 will output SV
allocations in addition to memory allocations.  Each SV allocation has a
distinct serial number that will be written on creation and destruction
of the SV.  So if you're executing the leaking code in a loop, you need
to look for SVs that are created, but never destroyed between each
cycle.  If such an SV is found, set a conditional breakpoint within
<code class="inline"><span class="i">new_SV</span><span class="s">(</span><span class="s">)</span></code>
 and make it break only when <code class="inline"><span class="w">PL_sv_serial</span></code>
 is equal to the
serial number of the leaking SV.  Then you will catch the interpreter in
exactly the state where the leaking SV is allocated, which is
sufficient in many cases to find the source of the leak.</p>
<p>As <code class="inline">-<span class="w">Dm</span></code>
 is using the PerlIO layer for output, it will by itself
allocate quite a bunch of SVs, which are hidden to avoid recursion.  You
can bypass the PerlIO layer if you use the SV logging provided by
<code class="inline">-<span class="w">DPERL_MEM_LOG</span></code>
 instead.</p>
<a name="PERL_MEM_LOG"></a><h2>PERL_MEM_LOG</h2>
<p>If compiled with <code class="inline">-<span class="w">DPERL_MEM_LOG</span></code>
 (<code class="inline">-<span class="w">Accflags</span>=-<span class="w">DPERL_MEM_LOG</span></code>
), both
memory and SV allocations go through logging functions, which is
handy for breakpoint setting.</p>
<p>Unless <code class="inline">-<span class="w">DPERL_MEM_LOG_NOIMPL</span></code>
 (<code class="inline">-<span class="w">Accflags</span>=-<span class="w">DPERL_MEM_LOG_NOIMPL</span></code>
) is
also compiled, the logging functions read $ENV{PERL_MEM_LOG} to
determine whether to log the event, and if so how:</p>
<pre class="verbatim"><ol><li>    <span class="i">$ENV</span>{<span class="w">PERL_MEM_LOG</span>} =~ <span class="q">/m/</span>           <span class="w">Log</span> <span class="w">all</span> <span class="w">memory</span> <span class="w">ops</span></li><li>    <span class="i">$ENV</span>{<span class="w">PERL_MEM_LOG</span>} =~ <span class="q">/s/</span>           <span class="w">Log</span> <span class="w">all</span> <span class="w">SV</span> <span class="w">ops</span></li><li>    <span class="i">$ENV</span>{<span class="w">PERL_MEM_LOG</span>} =~ <span class="q">/t/</span>           <span class="w">include</span> <span class="w">timestamp</span> <span class="w">in</span> <span class="w">Log</span></li><li>    <span class="i">$ENV</span>{<span class="w">PERL_MEM_LOG</span>} =~ <span class="q">/^(\d+)/</span>      <a class="l_k" href="functions/write.html">write</a> <span class="w">to</span> <span class="w">FD</span> <a class="l_k" href="functions/given.html">given</a> <span class="s">(</span><span class="w">default</span> <span class="w">is</span> <span class="n">2</span><span class="s">)</span></li></ol></pre><p>Memory logging is somewhat similar to <code class="inline">-<span class="w">Dm</span></code>
 but is independent of
<code class="inline">-<span class="w">DDEBUGGING</span></code>
, and at a higher level; all uses of Newx(), Renew(), and
Safefree() are logged with the caller's source code file and line
number (and C function name, if supported by the C compiler).  In
contrast, <code class="inline">-<span class="w">Dm</span></code>
 is directly at the point of <code class="inline"><span class="i">malloc</span><span class="s">(</span><span class="s">)</span></code>
.  SV logging is
similar.</p>
<p>Since the logging doesn't use PerlIO, all SV allocations are logged and
no extra SV allocations are introduced by enabling the logging.  If
compiled with <code class="inline">-<span class="w">DDEBUG_LEAKING_SCALARS</span></code>
, the serial number for each SV
allocation is also logged.</p>
<a name="DDD-over-gdb"></a><h2>DDD over gdb</h2>
<p>Those debugging perl with the DDD frontend over gdb may find the
following useful:</p>
<p>You can extend the data conversion shortcuts menu, so for example you
can display an SV's IV value with one click, without doing any typing.
To do that simply edit ~/.ddd/init file and add after:</p>
<pre class="verbatim"><ol><li>  ! Display shortcuts.</li><li>  Ddd*gdbDisplayShortcuts: \</li><li>  /t ()   // Convert to Bin\n\</li><li>  /d ()   // Convert to Dec\n\</li><li>  /x ()   // Convert to Hex\n\</li><li>  /o ()   // Convert to Oct(\n\</li></ol></pre><p>the following two lines:</p>
<pre class="verbatim"><ol><li>  ((XPV*) (())-&gt;sv_any )-&gt;xpv_pv  // 2pvx\n\</li><li>  ((XPVIV*) (())-&gt;sv_any )-&gt;xiv_iv // 2ivx</li></ol></pre><p>so now you can do ivx and pvx lookups or you can plug there the sv_peek
"conversion":</p>
<pre class="verbatim"><ol><li>  <span class="i">Perl_sv_peek</span><span class="s">(</span><span class="w">my_perl</span><span class="cm">,</span> <span class="s">(</span><span class="w">SV</span>*<span class="s">)</span><span class="s">(</span><span class="s">)</span><span class="s">)</span> // <span class="w">sv_peek</span></li></ol></pre><p>(The my_perl is for threaded builds.)  Just remember that every line,
but the last one, should end with \n\</p>
<p>Alternatively edit the init file interactively via: 3rd mouse button -&gt;
New Display -&gt; Edit Menu</p>
<p>Note: you can define up to 20 conversion shortcuts in the gdb section.</p>
<a name="C-backtrace"></a><h2>C backtrace</h2>
<p>On some platforms Perl supports retrieving the C level backtrace
(similar to what symbolic debuggers like gdb do).</p>
<p>The backtrace returns the stack trace of the C call frames,
with the symbol names (function names), the object names (like "perl"),
and if it can, also the source code locations (<a href="file:line">file:line</a>).</p>
<p>The supported platforms are Linux, and OS X (some *BSD might
work at least partly, but they have not yet been tested).</p>
<p>This feature hasn't been tested with multiple threads, but it will
only show the backtrace of the thread doing the backtracing.</p>
<p>The feature needs to be enabled with <code class="inline"><span class="w">Configure</span> -<span class="w">Dusecbacktrace</span></code>
.</p>
<p>The <code class="inline">-<span class="w">Dusecbacktrace</span></code>
 also enables keeping the debug information when
compiling/linking (often: <code class="inline">-g</code>
).  Many compilers/linkers do support
having both optimization and keeping the debug information.  The debug
information is needed for the symbol names and the source locations.</p>
<p>Static functions might not be visible for the backtrace.</p>
<p>Source code locations, even if available, can often be missing or
misleading if the compiler has e.g. inlined code.  Optimizer can
make matching the source code and the object code quite challenging.</p>
<ul>
<li><a name="Linux"></a><b>Linux</b>
<p>You <b>must</b> have the BFD (-lbfd) library installed, otherwise <code class="inline"><span class="w">perl</span></code>
 will
fail to link.  The BFD is usually distributed as part of the GNU binutils.</p>
<p>Summary: <code class="inline"><span class="w">Configure</span> ... -<span class="w">Dusecbacktrace</span></code>

and you need <code class="inline">-<span class="w">lbfd</span></code>
.</p>
</li>
<li><a name="OS-X"></a><b>OS X</b>
<p>The source code locations are supported <b>only</b> if you have
the Developer Tools installed.  (BFD is <b>not</b> needed.)</p>
<p>Summary: <code class="inline"><span class="w">Configure</span> ... -<span class="w">Dusecbacktrace</span></code>

and installing the Developer Tools would be good.</p>
</li>
</ul>
<p>Optionally, for trying out the feature, you may want to enable
automatic dumping of the backtrace just before a warning or croak (die)
message is emitted, by adding <code class="inline">-<span class="w">Accflags</span>=-<span class="w">DUSE_C_BACKTRACE_ON_ERROR</span></code>

for Configure.</p>
<p>Unless the above additional feature is enabled, nothing about the
backtrace functionality is visible, except for the Perl/XS level.</p>
<p>Furthermore, even if you have enabled this feature to be compiled,
you need to enable it in runtime with an environment variable:
<code class="inline"><span class="w">PERL_C_BACKTRACE_ON_ERROR</span>=<span class="n">10</span></code>
.  It must be an integer higher
than zero, telling the desired frame count.</p>
<p>Retrieving the backtrace from Perl level (using for example an XS
extension) would be much less exciting than one would hope: normally
you would see <code class="inline"><span class="w">runops</span></code>
, <code class="inline"><span class="w">entersub</span></code>
, and not much else.  This API is
intended to be called <b>from within</b> the Perl implementation, not from
Perl level execution.</p>
<p>The C API for the backtrace is as follows:</p>
<ul>
<li><a name="get_c_backtrace"></a><b>get_c_backtrace</b>
</li>
<li><a name="free_c_backtrace"></a><b>free_c_backtrace</b>
</li>
<li><a name="get_c_backtrace_dump"></a><b>get_c_backtrace_dump</b>
</li>
<li><a name="dump_c_backtrace"></a><b>dump_c_backtrace</b>
</li>
</ul>
<a name="Poison"></a><h2>Poison</h2>
<p>If you see in a debugger a memory area mysteriously full of 0xABABABAB
or 0xEFEFEFEF, you may be seeing the effect of the Poison() macros, see
<a href="perlclib.html">perlclib</a>.</p>
<a name="Read-only-optrees"></a><h2>Read-only optrees</h2>
<p>Under ithreads the optree is read only.  If you want to enforce this, to
check for write accesses from buggy code, compile with
<code class="inline">-<span class="w">Accflags</span>=-<span class="w">DPERL_DEBUG_READONLY_OPS</span></code>

to enable code that allocates op memory
via <code class="inline"><span class="w">mmap</span></code>
, and sets it read-only when it is attached to a subroutine.
Any write access to an op results in a <code class="inline"><span class="w">SIGBUS</span></code>
 and abort.</p>
<p>This code is intended for development only, and may not be portable
even to all Unix variants.  Also, it is an 80% solution, in that it
isn't able to make all ops read only.  Specifically it does not apply to
op slabs belonging to <code class="inline">BEGIN</code>
 blocks.</p>
<p>However, as an 80% solution it is still effective, as it has caught
bugs in the past.</p>
<a name="When-is-a-bool-not-a-bool%3f"></a><h2>When is a bool not a bool?</h2>
<p>On pre-C99 compilers, <code class="inline"><span class="w">bool</span></code>
 is defined as equivalent to <code class="inline"><span class="w">char</span></code>
.
Consequently assignment of any larger type to a <code class="inline"><span class="w">bool</span></code>
 is unsafe and may be
truncated.  The <code class="inline"><span class="w">cBOOL</span></code>
 macro exists to cast it correctly; you may also find
that using it is shorter and clearer than writing out the equivalent
conditional expression longhand.</p>
<p>On those platforms and compilers where <code class="inline"><span class="w">bool</span></code>
 really is a boolean (C++,
C99), it is easy to forget the cast.  You can force <code class="inline"><span class="w">bool</span></code>
 to be a <code class="inline"><span class="w">char</span></code>

by compiling with <code class="inline">-<span class="w">Accflags</span>=-<span class="w">DPERL_BOOL_AS_CHAR</span></code>
.  You may also wish to
run <code class="inline"><span class="w">Configure</span></code>
 with something like</p>
<pre class="verbatim"><ol><li>    -<span class="w">Accflags</span>=<span class="q">&#39;-Wconversion -Wno-sign-conversion -Wno-shorten-64-to-32&#39;</span></li></ol></pre><p>or your compiler's equivalent to make it easier to spot any unsafe truncations
that show up.</p>
<p>The <code class="inline"><span class="w">TRUE</span></code>
 and <code class="inline"><span class="w">FALSE</span></code>
 macros are available for situations where using them
would clarify intent. (But they always just mean the same as the integers 1 and
0 regardless, so using them isn't compulsory.)</p>
<a name="The-.i-Targets"></a><h2>The .i Targets</h2>
<p>You can expand the macros in a <i>foo.c</i> file by saying</p>
<pre class="verbatim"><ol><li>    <span class="w">make</span> <span class="w">foo</span>.<span class="w">i</span></li></ol></pre><p>which will expand the macros using cpp.  Don't be scared by the
results.</p>
<a name="AUTHOR"></a><h1>AUTHOR</h1>
<p>This document was originally written by Nathan Torkington, and is
maintained by the perl5-porters mailing list.</p>




  <div id="page_index" class="hud_container">
    <div id="page_index_header" class="hud_header">
      <div id="page_index_close" class="hud_close"><a href="#" onClick="pageIndex.hide();return false;"></a></div>
      <div id="page_index_title" class="hud_title"><span class="hud_span_top">Page index</span></div>
      <div id="page_index_topright" class="hud_topright"></div>
    </div>
    <div id="page_index_content" class="hud_content">
      <ul><li><a href="#NAME">NAME</a><li><a href="#DESCRIPTION">DESCRIPTION</a><li><a href="#COMMON-PROBLEMS">COMMON PROBLEMS</a><ul><li><a href="#Perl-environment-problems">Perl environment problems</a><li><a href="#Portability-problems">Portability problems</a><li><a href="#Problematic-System-Interfaces">Problematic System Interfaces</a><li><a href="#Security-problems">Security problems</a></ul><li><a href="#DEBUGGING">DEBUGGING</a><ul><li><a href="#Poking-at-Perl">Poking at Perl</a><li><a href="#Using-a-source-level-debugger">Using a source-level debugger</a><li><a href="#gdb-macro-support">gdb macro support</a><li><a href="#Dumping-Perl-Data-Structures">Dumping Perl Data Structures</a><li><a href="#Using-gdb-to-look-at-specific-parts-of-a-program">Using gdb to look at specific parts of a program</a><li><a href="#Using-gdb-to-look-at-what-the-parser%2flexer-are-doing">Using gdb to look at what the parser/lexer are doing</a></ul><li><a href="#SOURCE-CODE-STATIC-ANALYSIS">SOURCE CODE STATIC ANALYSIS</a><ul><li><a href="#lint">lint</a><li><a href="#Coverity">Coverity</a><li><a href="#HP-UX-cadvise-(Code-Advisor)">HP-UX cadvise (Code Advisor)</a><li><a href="#cpd-(cut-and-paste-detector)">cpd (cut-and-paste detector)</a><li><a href="#gcc-warnings">gcc warnings</a><li><a href="#Warnings-of-other-C-compilers">Warnings of other C compilers</a></ul><li><a href="#MEMORY-DEBUGGERS">MEMORY DEBUGGERS</a><ul><li><a href="#valgrind">valgrind</a><li><a href="#AddressSanitizer">AddressSanitizer</a></ul><li><a href="#PROFILING">PROFILING</a><ul><li><a href="#Gprof-Profiling">Gprof Profiling</a><li><a href="#GCC-gcov-Profiling">GCC gcov Profiling</a></ul><li><a href="#MISCELLANEOUS-TRICKS">MISCELLANEOUS TRICKS</a><ul><li><a href="#PERL_DESTRUCT_LEVEL">PERL_DESTRUCT_LEVEL</a><li><a href="#PERL_MEM_LOG">PERL_MEM_LOG</a><li><a href="#DDD-over-gdb">DDD over gdb</a><li><a href="#C-backtrace">C backtrace</a><li><a href="#Poison">Poison</a><li><a href="#Read-only-optrees">Read-only optrees</a><li><a href="#When-is-a-bool-not-a-bool%3f">When is a bool not a bool?</a><li><a href="#The-.i-Targets">The .i Targets</a></ul><li><a href="#AUTHOR">AUTHOR</a></ul>
    </div>
    <div id="page_index_footer" class="hud_footer">
      <div id="page_index_bottomleft" class="hud_bottomleft"></div>
      <div id="page_index_bottom" class="hud_bottom"><span class="hud_span_bottom"></span></div>
      <div id="page_index_resize" class="hud_resize"></div>
    </div>
  </div>


	    &nbsp;
          </div>
          <div id="content_footer">
          </div>
        </div>
        <div class="clear"></div>
      </div>
      
    <div id="footer">
      <div id="footer_content">
        <div id="footer_strapline">
          perldoc.perl.org - Official documentation for the Perl programming language
        </div>
        <div id="footer_links">
          <div id="address">
            <p class="name">Contact details</p>
            <p class="address">
	      Site maintained by <a href="mailto:jj@jonallen.info">Jon Allen (JJ)</a><br>
	    </p>
            <p class="contact">
              Documentation maintained by the <a href="http://lists.cpan.org/showlist.cgi?name=perl5-porters">Perl 5 Porters</a>
            </p>
          </div>
          <ul class="f1">
            <li>Manual
              <ul class="f2">
                <li><a href="index-overview.html">Overview</a>
                <li><a href="index-tutorials.html">Tutorials</a>
                <li><a href="index-faq.html">FAQs</a>
                <li><a href="index-history.html">Changes</a>
              </ul>
            <li>Reference
              <ul class="f2">
                <li><a href="index-language.html">Language</a>
                <li><a href="index-functions.html">Functions</a>
                <li><a href="perlop.html">Operators</a>
                <li><a href="perlvar.html">Variables</a>
              </ul>
            <li>Modules
              <ul class="f2">
                <li><a href="index-modules-A.html">Modules</a>
                <li><a href="index-pragmas.html">Pragmas</a>
                <li><a href="index-utilities.html">Utilities</a>
              </ul>
            <li>Misc
              <ul class="f2">
                <li><a href="index-licence.html">License</a>
                <li><a href="index-internals.html">Internals</a>
                <li><a href="index-platforms.html">Platforms</a>
              </ul>          </ul>
          <div class="clear"></div>
        </div>
      </div>
      <div id="footer_end">
      </div>
    </div>
      
    </div>
      <script language="JavaScript" type="text/javascript" src="static/exploreperl.js"></script>
      <script language="JavaScript" src="static/combined-20100403.js" type="text/javascript"></script>
<script language="JavaScript" type="text/javascript">
  perldoc.setPath(0);
  perldoc.pageName    = 'perlhacktips';
  perldoc.pageAddress = 'perlhacktips.html';
  perldoc.contentPage = 1;
  explorePerl.render();
  explorePerl.addEvents('explore_anchor');
</script>
    
  </body>
</html>