This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Detype locale.c comment
[perl5.git] / pod / perlhack.pod
CommitLineData
e8cd7eae
GS
1=head1 NAME
2
3perlhack - How to hack at the Perl internals
4
5=head1 DESCRIPTION
6
7This document attempts to explain how Perl development takes place,
8and ends with some suggestions for people wanting to become bona fide
9porters.
10
11The perl5-porters mailing list is where the Perl standard distribution
12is maintained and developed. The list can get anywhere from 10 to 150
13messages a day, depending on the heatedness of the debate. Most days
14there are two or three patches, extensions, features, or bugs being
15discussed at a time.
16
f8e3975a 17A searchable archive of the list is at either:
e8cd7eae
GS
18
19 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/
20
f8e3975a
IP
21or
22
23 http://archive.develooper.com/perl5-porters@perl.org/
24
e8cd7eae
GS
25List subscribers (the porters themselves) come in several flavours.
26Some are quiet curious lurkers, who rarely pitch in and instead watch
27the ongoing development to ensure they're forewarned of new changes or
28features in Perl. Some are representatives of vendors, who are there
29to make sure that Perl continues to compile and work on their
30platforms. Some patch any reported bug that they know how to fix,
31some are actively patching their pet area (threads, Win32, the regexp
32engine), while others seem to do nothing but complain. In other
33words, it's your usual mix of technical people.
34
35Over this group of porters presides Larry Wall. He has the final word
f6c51b38
GS
36in what does and does not change in the Perl language. Various
37releases of Perl are shepherded by a ``pumpking'', a porter
38responsible for gathering patches, deciding on a patch-by-patch
39feature-by-feature basis what will and will not go into the release.
caf100c0
JL
40For instance, Gurusamy Sarathy was the pumpking for the 5.6 release of
41Perl, and Jarkko Hietaniemi is the pumpking for the 5.8 release, and
42Hugo van der Sanden will be the pumpking for the 5.10 release.
e8cd7eae
GS
43
44In addition, various people are pumpkings for different things. For
45instance, Andy Dougherty and Jarkko Hietaniemi share the I<Configure>
caf100c0 46pumpkin.
e8cd7eae
GS
47
48Larry sees Perl development along the lines of the US government:
49there's the Legislature (the porters), the Executive branch (the
50pumpkings), and the Supreme Court (Larry). The legislature can
51discuss and submit patches to the executive branch all they like, but
52the executive branch is free to veto them. Rarely, the Supreme Court
53will side with the executive branch over the legislature, or the
54legislature over the executive branch. Mostly, however, the
55legislature and the executive branch are supposed to get along and
56work out their differences without impeachment or court cases.
57
58You might sometimes see reference to Rule 1 and Rule 2. Larry's power
59as Supreme Court is expressed in The Rules:
60
61=over 4
62
63=item 1
64
65Larry is always by definition right about how Perl should behave.
66This means he has final veto power on the core functionality.
67
68=item 2
69
70Larry is allowed to change his mind about any matter at a later date,
71regardless of whether he previously invoked Rule 1.
72
73=back
74
75Got that? Larry is always right, even when he was wrong. It's rare
76to see either Rule exercised, but they are often alluded to.
77
78New features and extensions to the language are contentious, because
79the criteria used by the pumpkings, Larry, and other porters to decide
80which features should be implemented and incorporated are not codified
81in a few small design goals as with some other languages. Instead,
82the heuristics are flexible and often difficult to fathom. Here is
83one person's list, roughly in decreasing order of importance, of
84heuristics that new features have to be weighed against:
85
86=over 4
87
88=item Does concept match the general goals of Perl?
89
90These haven't been written anywhere in stone, but one approximation
91is:
92
93 1. Keep it fast, simple, and useful.
94 2. Keep features/concepts as orthogonal as possible.
95 3. No arbitrary limits (platforms, data sizes, cultures).
96 4. Keep it open and exciting to use/patch/advocate Perl everywhere.
97 5. Either assimilate new technologies, or build bridges to them.
98
99=item Where is the implementation?
100
101All the talk in the world is useless without an implementation. In
102almost every case, the person or people who argue for a new feature
103will be expected to be the ones who implement it. Porters capable
104of coding new features have their own agendas, and are not available
105to implement your (possibly good) idea.
106
107=item Backwards compatibility
108
109It's a cardinal sin to break existing Perl programs. New warnings are
110contentious--some say that a program that emits warnings is not
111broken, while others say it is. Adding keywords has the potential to
112break programs, changing the meaning of existing token sequences or
113functions might break programs.
114
115=item Could it be a module instead?
116
117Perl 5 has extension mechanisms, modules and XS, specifically to avoid
118the need to keep changing the Perl interpreter. You can write modules
119that export functions, you can give those functions prototypes so they
120can be called like built-in functions, you can even write XS code to
121mess with the runtime data structures of the Perl interpreter if you
122want to implement really complicated things. If it can be done in a
123module instead of in the core, it's highly unlikely to be added.
124
125=item Is the feature generic enough?
126
127Is this something that only the submitter wants added to the language,
128or would it be broadly useful? Sometimes, instead of adding a feature
129with a tight focus, the porters might decide to wait until someone
130implements the more generalized feature. For instance, instead of
131implementing a ``delayed evaluation'' feature, the porters are waiting
132for a macro system that would permit delayed evaluation and much more.
133
134=item Does it potentially introduce new bugs?
135
136Radical rewrites of large chunks of the Perl interpreter have the
137potential to introduce new bugs. The smaller and more localized the
138change, the better.
139
140=item Does it preclude other desirable features?
141
142A patch is likely to be rejected if it closes off future avenues of
143development. For instance, a patch that placed a true and final
144interpretation on prototypes is likely to be rejected because there
145are still options for the future of prototypes that haven't been
146addressed.
147
148=item Is the implementation robust?
149
150Good patches (tight code, complete, correct) stand more chance of
151going in. Sloppy or incorrect patches might be placed on the back
152burner until the pumpking has time to fix, or might be discarded
153altogether without further notice.
154
155=item Is the implementation generic enough to be portable?
156
157The worst patches make use of a system-specific features. It's highly
158unlikely that nonportable additions to the Perl language will be
159accepted.
160
a936dd3c
NC
161=item Is the implementation tested?
162
163Patches which change behaviour (fixing bugs or introducing new features)
164must include regression tests to verify that everything works as expected.
165Without tests provided by the original author, how can anyone else changing
166perl in the future be sure that they haven't unwittingly broken the behaviour
167the patch implements? And without tests, how can the patch's author be
9d077eaa 168confident that his/her hard work put into the patch won't be accidentally
a936dd3c
NC
169thrown away by someone in the future?
170
e8cd7eae
GS
171=item Is there enough documentation?
172
173Patches without documentation are probably ill-thought out or
174incomplete. Nothing can be added without documentation, so submitting
175a patch for the appropriate manpages as well as the source code is
a936dd3c 176always a good idea.
e8cd7eae
GS
177
178=item Is there another way to do it?
179
180Larry said ``Although the Perl Slogan is I<There's More Than One Way
181to Do It>, I hesitate to make 10 ways to do something''. This is a
182tricky heuristic to navigate, though--one man's essential addition is
183another man's pointless cruft.
184
185=item Does it create too much work?
186
187Work for the pumpking, work for Perl programmers, work for module
188authors, ... Perl is supposed to be easy.
189
f6c51b38
GS
190=item Patches speak louder than words
191
192Working code is always preferred to pie-in-the-sky ideas. A patch to
193add a feature stands a much higher chance of making it to the language
194than does a random feature request, no matter how fervently argued the
195request might be. This ties into ``Will it be useful?'', as the fact
196that someone took the time to make the patch demonstrates a strong
197desire for the feature.
198
e8cd7eae
GS
199=back
200
201If you're on the list, you might hear the word ``core'' bandied
202around. It refers to the standard distribution. ``Hacking on the
203core'' means you're changing the C source code to the Perl
204interpreter. ``A core module'' is one that ships with Perl.
205
a1f349fd
MB
206=head2 Keeping in sync
207
e8cd7eae 208The source code to the Perl interpreter, in its different versions, is
f224927c
JH
209kept in a repository managed by a revision control system ( which is
210currently the Perforce program, see http://perforce.com/ ). The
e8cd7eae
GS
211pumpkings and a few others have access to the repository to check in
212changes. Periodically the pumpking for the development version of Perl
213will release a new version, so the rest of the porters can see what's
2be4c08b
GS
214changed. The current state of the main trunk of repository, and patches
215that describe the individual changes that have happened since the last
216public release are available at this location:
217
218 ftp://ftp.linux.activestate.com/pub/staff/gsar/APC/
219
a1f349fd
MB
220If you are a member of the perl5-porters mailing list, it is a good
221thing to keep in touch with the most recent changes. If not only to
222verify if what you would have posted as a bug report isn't already
223solved in the most recent available perl development branch, also
224known as perl-current, bleading edge perl, bleedperl or bleadperl.
2be4c08b
GS
225
226Needless to say, the source code in perl-current is usually in a perpetual
227state of evolution. You should expect it to be very buggy. Do B<not> use
228it for any purpose other than testing and development.
e8cd7eae 229
3e148164
JH
230Keeping in sync with the most recent branch can be done in several ways,
231but the most convenient and reliable way is using B<rsync>, available at
232ftp://rsync.samba.org/pub/rsync/ . (You can also get the most recent
233branch by FTP.)
a1f349fd
MB
234
235If you choose to keep in sync using rsync, there are two approaches
3e148164 236to doing so:
a1f349fd
MB
237
238=over 4
239
240=item rsync'ing the source tree
241
3e148164 242Presuming you are in the directory where your perl source resides
a1f349fd
MB
243and you have rsync installed and available, you can `upgrade' to
244the bleadperl using:
245
246 # rsync -avz rsync://ftp.linux.activestate.com/perl-current/ .
247
248This takes care of updating every single item in the source tree to
249the latest applied patch level, creating files that are new (to your
250distribution) and setting date/time stamps of existing files to
251reflect the bleadperl status.
252
c6d0653e
LC
253Note that this will not delete any files that were in '.' before
254the rsync. Once you are sure that the rsync is running correctly,
255run it with the --delete and the --dry-run options like this:
256
257 # rsync -avz --delete --dry-run rsync://ftp.linux.activestate.com/perl-current/ .
258
259This will I<simulate> an rsync run that also deletes files not
260present in the bleadperl master copy. Observe the results from
261this run closely. If you are sure that the actual run would delete
262no files precious to you, you could remove the '--dry-run' option.
263
a1f349fd
MB
264You can than check what patch was the latest that was applied by
265looking in the file B<.patch>, which will show the number of the
266latest patch.
267
268If you have more than one machine to keep in sync, and not all of
269them have access to the WAN (so you are not able to rsync all the
270source trees to the real source), there are some ways to get around
271this problem.
272
273=over 4
274
275=item Using rsync over the LAN
276
277Set up a local rsync server which makes the rsynced source tree
3e148164 278available to the LAN and sync the other machines against this
a1f349fd
MB
279directory.
280
1577cd80 281From http://rsync.samba.org/README.html :
a1f349fd
MB
282
283 "Rsync uses rsh or ssh for communication. It does not need to be
284 setuid and requires no special privileges for installation. It
3958b146 285 does not require an inetd entry or a daemon. You must, however,
a1f349fd
MB
286 have a working rsh or ssh system. Using ssh is recommended for
287 its security features."
288
289=item Using pushing over the NFS
290
291Having the other systems mounted over the NFS, you can take an
3e148164
JH
292active pushing approach by checking the just updated tree against
293the other not-yet synced trees. An example would be
294
295 #!/usr/bin/perl -w
296
297 use strict;
298 use File::Copy;
299
300 my %MF = map {
301 m/(\S+)/;
302 $1 => [ (stat $1)[2, 7, 9] ]; # mode, size, mtime
303 } `cat MANIFEST`;
304
305 my %remote = map { $_ => "/$_/pro/3gl/CPAN/perl-5.7.1" } qw(host1 host2);
306
307 foreach my $host (keys %remote) {
308 unless (-d $remote{$host}) {
309 print STDERR "Cannot Xsync for host $host\n";
310 next;
311 }
312 foreach my $file (keys %MF) {
313 my $rfile = "$remote{$host}/$file";
314 my ($mode, $size, $mtime) = (stat $rfile)[2, 7, 9];
315 defined $size or ($mode, $size, $mtime) = (0, 0, 0);
316 $size == $MF{$file}[1] && $mtime == $MF{$file}[2] and next;
317 printf "%4s %-34s %8d %9d %8d %9d\n",
318 $host, $file, $MF{$file}[1], $MF{$file}[2], $size, $mtime;
319 unlink $rfile;
320 copy ($file, $rfile);
321 utime time, $MF{$file}[2], $rfile;
322 chmod $MF{$file}[0], $rfile;
323 }
324 }
325
326though this is not perfect. It could be improved with checking
a1f349fd
MB
327file checksums before updating. Not all NFS systems support
328reliable utime support (when used over the NFS).
329
330=back
331
332=item rsync'ing the patches
333
334The source tree is maintained by the pumpking who applies patches to
335the files in the tree. These patches are either created by the
336pumpking himself using C<diff -c> after updating the file manually or
337by applying patches sent in by posters on the perl5-porters list.
338These patches are also saved and rsync'able, so you can apply them
339yourself to the source files.
340
341Presuming you are in a directory where your patches reside, you can
3e148164 342get them in sync with
a1f349fd
MB
343
344 # rsync -avz rsync://ftp.linux.activestate.com/perl-current-diffs/ .
345
346This makes sure the latest available patch is downloaded to your
347patch directory.
348
3e148164 349It's then up to you to apply these patches, using something like
a1f349fd 350
df3477ff 351 # last=`ls -t *.gz | sed q`
a1f349fd
MB
352 # rsync -avz rsync://ftp.linux.activestate.com/perl-current-diffs/ .
353 # find . -name '*.gz' -newer $last -exec gzcat {} \; >blead.patch
354 # cd ../perl-current
355 # patch -p1 -N <../perl-current-diffs/blead.patch
356
357or, since this is only a hint towards how it works, use CPAN-patchaperl
358from Andreas K├Ânig to have better control over the patching process.
359
360=back
361
f7e1e956 362=head2 Why rsync the source tree
a1f349fd
MB
363
364=over 4
365
10f58044 366=item It's easier to rsync the source tree
a1f349fd
MB
367
368Since you don't have to apply the patches yourself, you are sure all
369files in the source tree are in the right state.
370
371=item It's more recent
372
373According to Gurusamy Sarathy:
374
375 "... The rsync mirror is automatic and syncs with the repository
376 every five minutes.
377
3e148164 378 "Updating the patch area still requires manual intervention
a1f349fd
MB
379 (with all the goofiness that implies, which you've noted) and
380 is typically on a daily cycle. Making this process automatic
381 is on my tuit list, but don't ask me when."
382
383=item It's more reliable
384
3e148164 385Well, since the patches are updated by hand, I don't have to say any
a1f349fd
MB
386more ... (see Sarathy's remark).
387
388=back
389
f7e1e956 390=head2 Why rsync the patches
a1f349fd
MB
391
392=over 4
393
10f58044 394=item It's easier to rsync the patches
a1f349fd
MB
395
396If you have more than one machine that you want to keep in track with
3e148164 397bleadperl, it's easier to rsync the patches only once and then apply
a1f349fd
MB
398them to all the source trees on the different machines.
399
400In case you try to keep in pace on 5 different machines, for which
401only one of them has access to the WAN, rsync'ing all the source
3e148164 402trees should than be done 5 times over the NFS. Having
a1f349fd 403rsync'ed the patches only once, I can apply them to all the source
3e148164 404trees automatically. Need you say more ;-)
a1f349fd
MB
405
406=item It's a good reference
407
408If you do not only like to have the most recent development branch,
409but also like to B<fix> bugs, or extend features, you want to dive
410into the sources. If you are a seasoned perl core diver, you don't
411need no manuals, tips, roadmaps, perlguts.pod or other aids to find
412your way around. But if you are a starter, the patches may help you
413in finding where you should start and how to change the bits that
414bug you.
415
416The file B<Changes> is updated on occasions the pumpking sees as his
417own little sync points. On those occasions, he releases a tar-ball of
418the current source tree (i.e. perl@7582.tar.gz), which will be an
419excellent point to start with when choosing to use the 'rsync the
420patches' scheme. Starting with perl@7582, which means a set of source
421files on which the latest applied patch is number 7582, you apply all
f18956b7 422succeeding patches available from then on (7583, 7584, ...).
a1f349fd
MB
423
424You can use the patches later as a kind of search archive.
425
426=over 4
427
428=item Finding a start point
429
430If you want to fix/change the behaviour of function/feature Foo, just
431scan the patches for patches that mention Foo either in the subject,
3e148164 432the comments, or the body of the fix. A good chance the patch shows
a1f349fd
MB
433you the files that are affected by that patch which are very likely
434to be the starting point of your journey into the guts of perl.
435
436=item Finding how to fix a bug
437
438If you've found I<where> the function/feature Foo misbehaves, but you
439don't know how to fix it (but you do know the change you want to
440make), you can, again, peruse the patches for similar changes and
441look how others apply the fix.
442
443=item Finding the source of misbehaviour
444
445When you keep in sync with bleadperl, the pumpking would love to
3958b146 446I<see> that the community efforts really work. So after each of his
a1f349fd
MB
447sync points, you are to 'make test' to check if everything is still
448in working order. If it is, you do 'make ok', which will send an OK
449report to perlbug@perl.org. (If you do not have access to a mailer
3e148164 450from the system you just finished successfully 'make test', you can
a1f349fd
MB
451do 'make okfile', which creates the file C<perl.ok>, which you can
452than take to your favourite mailer and mail yourself).
453
3958b146 454But of course, as always, things will not always lead to a success
a1f349fd
MB
455path, and one or more test do not pass the 'make test'. Before
456sending in a bug report (using 'make nok' or 'make nokfile'), check
457the mailing list if someone else has reported the bug already and if
458so, confirm it by replying to that message. If not, you might want to
459trace the source of that misbehaviour B<before> sending in the bug,
460which will help all the other porters in finding the solution.
461
3e148164
JH
462Here the saved patches come in very handy. You can check the list of
463patches to see which patch changed what file and what change caused
464the misbehaviour. If you note that in the bug report, it saves the
465one trying to solve it, looking for that point.
a1f349fd
MB
466
467=back
468
469If searching the patches is too bothersome, you might consider using
470perl's bugtron to find more information about discussions and
471ramblings on posted bugs.
472
3e148164
JH
473If you want to get the best of both worlds, rsync both the source
474tree for convenience, reliability and ease and rsync the patches
475for reference.
476
52315700
RF
477=back
478
479
480=head2 Perlbug remote interface
481
482=over 4
483
f224927c
JH
484There are three (3) remote administrative interfaces for modifying bug
485status, category, etc. In all cases an admin must be first registered
486with the Perlbug database by sending an email request to
487richard@perl.org or bugmongers@perl.org.
52315700 488
f224927c
JH
489The main requirement is the willingness to classify, (with the
490emphasis on closing where possible :), outstanding bugs. Further
491explanation can be garnered from the web at http://bugs.perl.org/ , or
492by asking on the admin mailing list at: bugmongers@perl.org
52315700
RF
493
494For more info on the web see
495
496 http://bugs.perl.org/perlbug.cgi?req=spec
497
52315700
RF
498=item 1 http://bugs.perl.org
499
500Login via the web, (remove B<admin/> if only browsing), where interested Cc's, tests, patches and change-ids, etc. may be assigned.
501
502 http://bugs.perl.org/admin/index.html
503
504
505=item 2 bugdb@perl.org
506
507Where the subject line is used for commands:
508
509 To: bugdb@perl.org
510 Subject: -a close bugid1 bugid2 aix install
511
512 To: bugdb@perl.org
513 Subject: -h
514
515
516=item 3 commands_and_bugdids@bugs.perl.org
517
518Where the address itself is the source for the commands:
519
520 To: close_bugid1_bugid2_aix@bugs.perl.org
521
522 To: help@bugs.perl.org
523
524
525=item notes, patches, tests
526
527For patches and tests, the message body is assigned to the appropriate bug/s and forwarded to p5p for their attention.
528
529 To: test_<bugid1>_aix_close@bugs.perl.org
530 Subject: this is a test for the (now closed) aix bug
531
532 Test is the body of the mail
533
534=back
535
a1f349fd
MB
536=head2 Submitting patches
537
f7e1e956
MS
538Always submit patches to I<perl5-porters@perl.org>. If you're
539patching a core module and there's an author listed, send the author a
540copy (see L<Patching a core module>). This lets other porters review
541your patch, which catches a surprising number of errors in patches.
542Either use the diff program (available in source code form from
f224927c 543ftp://ftp.gnu.org/pub/gnu/ , or use Johan Vromans' I<makepatch>
f7e1e956
MS
544(available from I<CPAN/authors/id/JV/>). Unified diffs are preferred,
545but context diffs are accepted. Do not send RCS-style diffs or diffs
546without context lines. More information is given in the
547I<Porting/patching.pod> file in the Perl source distribution. Please
548patch against the latest B<development> version (e.g., if you're
549fixing a bug in the 5.005 track, patch against the latest 5.005_5x
550version). Only patches that survive the heat of the development
551branch get applied to maintenance versions.
552
553Your patch should update the documentation and test suite. See
554L<Writing a test>.
e8cd7eae
GS
555
556To report a bug in Perl, use the program I<perlbug> which comes with
557Perl (if you can't get Perl to work, send mail to the address
f18956b7 558I<perlbug@perl.org> or I<perlbug@perl.com>). Reporting bugs through
e8cd7eae 559I<perlbug> feeds into the automated bug-tracking system, access to
f224927c 560which is provided through the web at http://bugs.perl.org/ . It
e8cd7eae
GS
561often pays to check the archives of the perl5-porters mailing list to
562see whether the bug you're reporting has been reported before, and if
563so whether it was considered a bug. See above for the location of
564the searchable archives.
565
f224927c 566The CPAN testers ( http://testers.cpan.org/ ) are a group of
ba139f7d
JH
567volunteers who test CPAN modules on a variety of platforms. Perl
568Smokers ( http://archives.develooper.com/daily-build@perl.org/ )
569automatically tests Perl source releases on platforms with various
570configurations. Both efforts welcome volunteers.
e8cd7eae 571
e8cd7eae
GS
572It's a good idea to read and lurk for a while before chipping in.
573That way you'll get to see the dynamic of the conversations, learn the
574personalities of the players, and hopefully be better prepared to make
575a useful contribution when do you speak up.
576
577If after all this you still think you want to join the perl5-porters
f6c51b38
GS
578mailing list, send mail to I<perl5-porters-subscribe@perl.org>. To
579unsubscribe, send mail to I<perl5-porters-unsubscribe@perl.org>.
e8cd7eae 580
a422fd2d
SC
581To hack on the Perl guts, you'll need to read the following things:
582
583=over 3
584
585=item L<perlguts>
586
587This is of paramount importance, since it's the documentation of what
588goes where in the Perl source. Read it over a couple of times and it
589might start to make sense - don't worry if it doesn't yet, because the
590best way to study it is to read it in conjunction with poking at Perl
591source, and we'll do that later on.
592
593You might also want to look at Gisle Aas's illustrated perlguts -
594there's no guarantee that this will be absolutely up-to-date with the
595latest documentation in the Perl core, but the fundamentals will be
1577cd80 596right. ( http://gisle.aas.no/perl/illguts/ )
a422fd2d
SC
597
598=item L<perlxstut> and L<perlxs>
599
600A working knowledge of XSUB programming is incredibly useful for core
601hacking; XSUBs use techniques drawn from the PP code, the portion of the
602guts that actually executes a Perl program. It's a lot gentler to learn
603those techniques from simple examples and explanation than from the core
604itself.
605
606=item L<perlapi>
607
608The documentation for the Perl API explains what some of the internal
609functions do, as well as the many macros used in the source.
610
611=item F<Porting/pumpkin.pod>
612
613This is a collection of words of wisdom for a Perl porter; some of it is
614only useful to the pumpkin holder, but most of it applies to anyone
615wanting to go about Perl development.
616
617=item The perl5-porters FAQ
618
619This is posted to perl5-porters at the beginning on every month, and
f224927c 620should be available from http://perlhacker.org/p5p-faq ; alternatively,
a422fd2d
SC
621you can get the FAQ emailed to you by sending mail to
622C<perl5-porters-faq@perl.org>. It contains hints on reading
623perl5-porters, information on how perl5-porters works and how Perl
624development in general works.
625
626=back
627
628=head2 Finding Your Way Around
629
630Perl maintenance can be split into a number of areas, and certain people
631(pumpkins) will have responsibility for each area. These areas sometimes
632correspond to files or directories in the source kit. Among the areas are:
633
634=over 3
635
636=item Core modules
637
638Modules shipped as part of the Perl core live in the F<lib/> and F<ext/>
639subdirectories: F<lib/> is for the pure-Perl modules, and F<ext/>
640contains the core XS modules.
641
f7e1e956
MS
642=item Tests
643
644There are tests for nearly all the modules, built-ins and major bits
645of functionality. Test files all have a .t suffix. Module tests live
646in the F<lib/> and F<ext/> directories next to the module being
647tested. Others live in F<t/>. See L<Writing a test>
648
a422fd2d
SC
649=item Documentation
650
651Documentation maintenance includes looking after everything in the
652F<pod/> directory, (as well as contributing new documentation) and
653the documentation to the modules in core.
654
655=item Configure
656
657The configure process is the way we make Perl portable across the
658myriad of operating systems it supports. Responsibility for the
659configure, build and installation process, as well as the overall
660portability of the core code rests with the configure pumpkin - others
661help out with individual operating systems.
662
663The files involved are the operating system directories, (F<win32/>,
664F<os2/>, F<vms/> and so on) the shell scripts which generate F<config.h>
665and F<Makefile>, as well as the metaconfig files which generate
666F<Configure>. (metaconfig isn't included in the core distribution.)
667
668=item Interpreter
669
670And of course, there's the core of the Perl interpreter itself. Let's
671have a look at that in a little more detail.
672
673=back
674
675Before we leave looking at the layout, though, don't forget that
676F<MANIFEST> contains not only the file names in the Perl distribution,
677but short descriptions of what's in them, too. For an overview of the
678important files, try this:
679
680 perl -lne 'print if /^[^\/]+\.[ch]\s+/' MANIFEST
681
682=head2 Elements of the interpreter
683
684The work of the interpreter has two main stages: compiling the code
685into the internal representation, or bytecode, and then executing it.
686L<perlguts/Compiled code> explains exactly how the compilation stage
687happens.
688
689Here is a short breakdown of perl's operation:
690
691=over 3
692
693=item Startup
694
695The action begins in F<perlmain.c>. (or F<miniperlmain.c> for miniperl)
696This is very high-level code, enough to fit on a single screen, and it
697resembles the code found in L<perlembed>; most of the real action takes
698place in F<perl.c>
699
700First, F<perlmain.c> allocates some memory and constructs a Perl
701interpreter:
702
703 1 PERL_SYS_INIT3(&argc,&argv,&env);
704 2
705 3 if (!PL_do_undump) {
706 4 my_perl = perl_alloc();
707 5 if (!my_perl)
708 6 exit(1);
709 7 perl_construct(my_perl);
710 8 PL_perl_destruct_level = 0;
711 9 }
712
713Line 1 is a macro, and its definition is dependent on your operating
714system. Line 3 references C<PL_do_undump>, a global variable - all
715global variables in Perl start with C<PL_>. This tells you whether the
716current running program was created with the C<-u> flag to perl and then
717F<undump>, which means it's going to be false in any sane context.
718
719Line 4 calls a function in F<perl.c> to allocate memory for a Perl
720interpreter. It's quite a simple function, and the guts of it looks like
721this:
722
723 my_perl = (PerlInterpreter*)PerlMem_malloc(sizeof(PerlInterpreter));
724
725Here you see an example of Perl's system abstraction, which we'll see
726later: C<PerlMem_malloc> is either your system's C<malloc>, or Perl's
727own C<malloc> as defined in F<malloc.c> if you selected that option at
728configure time.
729
730Next, in line 7, we construct the interpreter; this sets up all the
731special variables that Perl needs, the stacks, and so on.
732
733Now we pass Perl the command line options, and tell it to go:
734
735 exitstatus = perl_parse(my_perl, xs_init, argc, argv, (char **)NULL);
736 if (!exitstatus) {
737 exitstatus = perl_run(my_perl);
738 }
739
740
741C<perl_parse> is actually a wrapper around C<S_parse_body>, as defined
742in F<perl.c>, which processes the command line options, sets up any
743statically linked XS modules, opens the program and calls C<yyparse> to
744parse it.
745
746=item Parsing
747
748The aim of this stage is to take the Perl source, and turn it into an op
749tree. We'll see what one of those looks like later. Strictly speaking,
750there's three things going on here.
751
752C<yyparse>, the parser, lives in F<perly.c>, although you're better off
753reading the original YACC input in F<perly.y>. (Yes, Virginia, there
754B<is> a YACC grammar for Perl!) The job of the parser is to take your
755code and `understand' it, splitting it into sentences, deciding which
756operands go with which operators and so on.
757
758The parser is nobly assisted by the lexer, which chunks up your input
759into tokens, and decides what type of thing each token is: a variable
760name, an operator, a bareword, a subroutine, a core function, and so on.
761The main point of entry to the lexer is C<yylex>, and that and its
762associated routines can be found in F<toke.c>. Perl isn't much like
763other computer languages; it's highly context sensitive at times, it can
764be tricky to work out what sort of token something is, or where a token
765ends. As such, there's a lot of interplay between the tokeniser and the
766parser, which can get pretty frightening if you're not used to it.
767
768As the parser understands a Perl program, it builds up a tree of
769operations for the interpreter to perform during execution. The routines
770which construct and link together the various operations are to be found
771in F<op.c>, and will be examined later.
772
773=item Optimization
774
775Now the parsing stage is complete, and the finished tree represents
776the operations that the Perl interpreter needs to perform to execute our
777program. Next, Perl does a dry run over the tree looking for
778optimisations: constant expressions such as C<3 + 4> will be computed
779now, and the optimizer will also see if any multiple operations can be
780replaced with a single one. For instance, to fetch the variable C<$foo>,
781instead of grabbing the glob C<*foo> and looking at the scalar
782component, the optimizer fiddles the op tree to use a function which
783directly looks up the scalar in question. The main optimizer is C<peep>
784in F<op.c>, and many ops have their own optimizing functions.
785
786=item Running
787
788Now we're finally ready to go: we have compiled Perl byte code, and all
789that's left to do is run it. The actual execution is done by the
790C<runops_standard> function in F<run.c>; more specifically, it's done by
791these three innocent looking lines:
792
793 while ((PL_op = CALL_FPTR(PL_op->op_ppaddr)(aTHX))) {
794 PERL_ASYNC_CHECK();
795 }
796
797You may be more comfortable with the Perl version of that:
798
799 PERL_ASYNC_CHECK() while $Perl::op = &{$Perl::op->{function}};
800
801Well, maybe not. Anyway, each op contains a function pointer, which
802stipulates the function which will actually carry out the operation.
803This function will return the next op in the sequence - this allows for
804things like C<if> which choose the next op dynamically at run time.
805The C<PERL_ASYNC_CHECK> makes sure that things like signals interrupt
806execution if required.
807
808The actual functions called are known as PP code, and they're spread
809between four files: F<pp_hot.c> contains the `hot' code, which is most
810often used and highly optimized, F<pp_sys.c> contains all the
811system-specific functions, F<pp_ctl.c> contains the functions which
812implement control structures (C<if>, C<while> and the like) and F<pp.c>
813contains everything else. These are, if you like, the C code for Perl's
814built-in functions and operators.
815
816=back
817
818=head2 Internal Variable Types
819
820You should by now have had a look at L<perlguts>, which tells you about
821Perl's internal variable types: SVs, HVs, AVs and the rest. If not, do
822that now.
823
824These variables are used not only to represent Perl-space variables, but
825also any constants in the code, as well as some structures completely
826internal to Perl. The symbol table, for instance, is an ordinary Perl
827hash. Your code is represented by an SV as it's read into the parser;
828any program files you call are opened via ordinary Perl filehandles, and
829so on.
830
831The core L<Devel::Peek|Devel::Peek> module lets us examine SVs from a
832Perl program. Let's see, for instance, how Perl treats the constant
833C<"hello">.
834
835 % perl -MDevel::Peek -e 'Dump("hello")'
836 1 SV = PV(0xa041450) at 0xa04ecbc
837 2 REFCNT = 1
838 3 FLAGS = (POK,READONLY,pPOK)
839 4 PV = 0xa0484e0 "hello"\0
840 5 CUR = 5
841 6 LEN = 6
842
843Reading C<Devel::Peek> output takes a bit of practise, so let's go
844through it line by line.
845
846Line 1 tells us we're looking at an SV which lives at C<0xa04ecbc> in
847memory. SVs themselves are very simple structures, but they contain a
848pointer to a more complex structure. In this case, it's a PV, a
849structure which holds a string value, at location C<0xa041450>. Line 2
850is the reference count; there are no other references to this data, so
851it's 1.
852
853Line 3 are the flags for this SV - it's OK to use it as a PV, it's a
854read-only SV (because it's a constant) and the data is a PV internally.
855Next we've got the contents of the string, starting at location
856C<0xa0484e0>.
857
858Line 5 gives us the current length of the string - note that this does
859B<not> include the null terminator. Line 6 is not the length of the
860string, but the length of the currently allocated buffer; as the string
861grows, Perl automatically extends the available storage via a routine
862called C<SvGROW>.
863
864You can get at any of these quantities from C very easily; just add
865C<Sv> to the name of the field shown in the snippet, and you've got a
866macro which will return the value: C<SvCUR(sv)> returns the current
867length of the string, C<SvREFCOUNT(sv)> returns the reference count,
868C<SvPV(sv, len)> returns the string itself with its length, and so on.
869More macros to manipulate these properties can be found in L<perlguts>.
870
871Let's take an example of manipulating a PV, from C<sv_catpvn>, in F<sv.c>
872
873 1 void
874 2 Perl_sv_catpvn(pTHX_ register SV *sv, register const char *ptr, register STRLEN len)
875 3 {
876 4 STRLEN tlen;
877 5 char *junk;
878
879 6 junk = SvPV_force(sv, tlen);
880 7 SvGROW(sv, tlen + len + 1);
881 8 if (ptr == junk)
882 9 ptr = SvPVX(sv);
883 10 Move(ptr,SvPVX(sv)+tlen,len,char);
884 11 SvCUR(sv) += len;
885 12 *SvEND(sv) = '\0';
886 13 (void)SvPOK_only_UTF8(sv); /* validate pointer */
887 14 SvTAINT(sv);
888 15 }
889
890This is a function which adds a string, C<ptr>, of length C<len> onto
891the end of the PV stored in C<sv>. The first thing we do in line 6 is
892make sure that the SV B<has> a valid PV, by calling the C<SvPV_force>
893macro to force a PV. As a side effect, C<tlen> gets set to the current
894value of the PV, and the PV itself is returned to C<junk>.
895
b1866b2d 896In line 7, we make sure that the SV will have enough room to accommodate
a422fd2d
SC
897the old string, the new string and the null terminator. If C<LEN> isn't
898big enough, C<SvGROW> will reallocate space for us.
899
900Now, if C<junk> is the same as the string we're trying to add, we can
901grab the string directly from the SV; C<SvPVX> is the address of the PV
902in the SV.
903
904Line 10 does the actual catenation: the C<Move> macro moves a chunk of
905memory around: we move the string C<ptr> to the end of the PV - that's
906the start of the PV plus its current length. We're moving C<len> bytes
907of type C<char>. After doing so, we need to tell Perl we've extended the
908string, by altering C<CUR> to reflect the new length. C<SvEND> is a
909macro which gives us the end of the string, so that needs to be a
910C<"\0">.
911
912Line 13 manipulates the flags; since we've changed the PV, any IV or NV
913values will no longer be valid: if we have C<$a=10; $a.="6";> we don't
914want to use the old IV of 10. C<SvPOK_only_utf8> is a special UTF8-aware
915version of C<SvPOK_only>, a macro which turns off the IOK and NOK flags
916and turns on POK. The final C<SvTAINT> is a macro which launders tainted
917data if taint mode is turned on.
918
919AVs and HVs are more complicated, but SVs are by far the most common
920variable type being thrown around. Having seen something of how we
921manipulate these, let's go on and look at how the op tree is
922constructed.
923
924=head2 Op Trees
925
926First, what is the op tree, anyway? The op tree is the parsed
927representation of your program, as we saw in our section on parsing, and
928it's the sequence of operations that Perl goes through to execute your
929program, as we saw in L</Running>.
930
931An op is a fundamental operation that Perl can perform: all the built-in
932functions and operators are ops, and there are a series of ops which
933deal with concepts the interpreter needs internally - entering and
934leaving a block, ending a statement, fetching a variable, and so on.
935
936The op tree is connected in two ways: you can imagine that there are two
937"routes" through it, two orders in which you can traverse the tree.
938First, parse order reflects how the parser understood the code, and
939secondly, execution order tells perl what order to perform the
940operations in.
941
942The easiest way to examine the op tree is to stop Perl after it has
943finished parsing, and get it to dump out the tree. This is exactly what
944the compiler backends L<B::Terse|B::Terse> and L<B::Debug|B::Debug> do.
945
946Let's have a look at how Perl sees C<$a = $b + $c>:
947
948 % perl -MO=Terse -e '$a=$b+$c'
949 1 LISTOP (0x8179888) leave
950 2 OP (0x81798b0) enter
951 3 COP (0x8179850) nextstate
952 4 BINOP (0x8179828) sassign
953 5 BINOP (0x8179800) add [1]
954 6 UNOP (0x81796e0) null [15]
955 7 SVOP (0x80fafe0) gvsv GV (0x80fa4cc) *b
956 8 UNOP (0x81797e0) null [15]
957 9 SVOP (0x8179700) gvsv GV (0x80efeb0) *c
958 10 UNOP (0x816b4f0) null [15]
959 11 SVOP (0x816dcf0) gvsv GV (0x80fa460) *a
960
961Let's start in the middle, at line 4. This is a BINOP, a binary
962operator, which is at location C<0x8179828>. The specific operator in
963question is C<sassign> - scalar assignment - and you can find the code
964which implements it in the function C<pp_sassign> in F<pp_hot.c>. As a
965binary operator, it has two children: the add operator, providing the
966result of C<$b+$c>, is uppermost on line 5, and the left hand side is on
967line 10.
968
969Line 10 is the null op: this does exactly nothing. What is that doing
970there? If you see the null op, it's a sign that something has been
971optimized away after parsing. As we mentioned in L</Optimization>,
972the optimization stage sometimes converts two operations into one, for
973example when fetching a scalar variable. When this happens, instead of
974rewriting the op tree and cleaning up the dangling pointers, it's easier
975just to replace the redundant operation with the null op. Originally,
976the tree would have looked like this:
977
978 10 SVOP (0x816b4f0) rv2sv [15]
979 11 SVOP (0x816dcf0) gv GV (0x80fa460) *a
980
981That is, fetch the C<a> entry from the main symbol table, and then look
982at the scalar component of it: C<gvsv> (C<pp_gvsv> into F<pp_hot.c>)
983happens to do both these things.
984
985The right hand side, starting at line 5 is similar to what we've just
986seen: we have the C<add> op (C<pp_add> also in F<pp_hot.c>) add together
987two C<gvsv>s.
988
989Now, what's this about?
990
991 1 LISTOP (0x8179888) leave
992 2 OP (0x81798b0) enter
993 3 COP (0x8179850) nextstate
994
995C<enter> and C<leave> are scoping ops, and their job is to perform any
996housekeeping every time you enter and leave a block: lexical variables
997are tidied up, unreferenced variables are destroyed, and so on. Every
998program will have those first three lines: C<leave> is a list, and its
999children are all the statements in the block. Statements are delimited
1000by C<nextstate>, so a block is a collection of C<nextstate> ops, with
1001the ops to be performed for each statement being the children of
1002C<nextstate>. C<enter> is a single op which functions as a marker.
1003
1004That's how Perl parsed the program, from top to bottom:
1005
1006 Program
1007 |
1008 Statement
1009 |
1010 =
1011 / \
1012 / \
1013 $a +
1014 / \
1015 $b $c
1016
1017However, it's impossible to B<perform> the operations in this order:
1018you have to find the values of C<$b> and C<$c> before you add them
1019together, for instance. So, the other thread that runs through the op
1020tree is the execution order: each op has a field C<op_next> which points
1021to the next op to be run, so following these pointers tells us how perl
1022executes the code. We can traverse the tree in this order using
1023the C<exec> option to C<B::Terse>:
1024
1025 % perl -MO=Terse,exec -e '$a=$b+$c'
1026 1 OP (0x8179928) enter
1027 2 COP (0x81798c8) nextstate
1028 3 SVOP (0x81796c8) gvsv GV (0x80fa4d4) *b
1029 4 SVOP (0x8179798) gvsv GV (0x80efeb0) *c
1030 5 BINOP (0x8179878) add [1]
1031 6 SVOP (0x816dd38) gvsv GV (0x80fa468) *a
1032 7 BINOP (0x81798a0) sassign
1033 8 LISTOP (0x8179900) leave
1034
1035This probably makes more sense for a human: enter a block, start a
1036statement. Get the values of C<$b> and C<$c>, and add them together.
1037Find C<$a>, and assign one to the other. Then leave.
1038
1039The way Perl builds up these op trees in the parsing process can be
1040unravelled by examining F<perly.y>, the YACC grammar. Let's take the
1041piece we need to construct the tree for C<$a = $b + $c>
1042
1043 1 term : term ASSIGNOP term
1044 2 { $$ = newASSIGNOP(OPf_STACKED, $1, $2, $3); }
1045 3 | term ADDOP term
1046 4 { $$ = newBINOP($2, 0, scalar($1), scalar($3)); }
1047
1048If you're not used to reading BNF grammars, this is how it works: You're
1049fed certain things by the tokeniser, which generally end up in upper
1050case. Here, C<ADDOP>, is provided when the tokeniser sees C<+> in your
1051code. C<ASSIGNOP> is provided when C<=> is used for assigning. These are
1052`terminal symbols', because you can't get any simpler than them.
1053
1054The grammar, lines one and three of the snippet above, tells you how to
1055build up more complex forms. These complex forms, `non-terminal symbols'
1056are generally placed in lower case. C<term> here is a non-terminal
1057symbol, representing a single expression.
1058
1059The grammar gives you the following rule: you can make the thing on the
1060left of the colon if you see all the things on the right in sequence.
1061This is called a "reduction", and the aim of parsing is to completely
1062reduce the input. There are several different ways you can perform a
1063reduction, separated by vertical bars: so, C<term> followed by C<=>
1064followed by C<term> makes a C<term>, and C<term> followed by C<+>
1065followed by C<term> can also make a C<term>.
1066
1067So, if you see two terms with an C<=> or C<+>, between them, you can
1068turn them into a single expression. When you do this, you execute the
1069code in the block on the next line: if you see C<=>, you'll do the code
1070in line 2. If you see C<+>, you'll do the code in line 4. It's this code
1071which contributes to the op tree.
1072
1073 | term ADDOP term
1074 { $$ = newBINOP($2, 0, scalar($1), scalar($3)); }
1075
1076What this does is creates a new binary op, and feeds it a number of
1077variables. The variables refer to the tokens: C<$1> is the first token in
1078the input, C<$2> the second, and so on - think regular expression
1079backreferences. C<$$> is the op returned from this reduction. So, we
1080call C<newBINOP> to create a new binary operator. The first parameter to
1081C<newBINOP>, a function in F<op.c>, is the op type. It's an addition
1082operator, so we want the type to be C<ADDOP>. We could specify this
1083directly, but it's right there as the second token in the input, so we
1084use C<$2>. The second parameter is the op's flags: 0 means `nothing
1085special'. Then the things to add: the left and right hand side of our
1086expression, in scalar context.
1087
1088=head2 Stacks
1089
1090When perl executes something like C<addop>, how does it pass on its
1091results to the next op? The answer is, through the use of stacks. Perl
1092has a number of stacks to store things it's currently working on, and
1093we'll look at the three most important ones here.
1094
1095=over 3
1096
1097=item Argument stack
1098
1099Arguments are passed to PP code and returned from PP code using the
1100argument stack, C<ST>. The typical way to handle arguments is to pop
1101them off the stack, deal with them how you wish, and then push the result
1102back onto the stack. This is how, for instance, the cosine operator
1103works:
1104
1105 NV value;
1106 value = POPn;
1107 value = Perl_cos(value);
1108 XPUSHn(value);
1109
1110We'll see a more tricky example of this when we consider Perl's macros
1111below. C<POPn> gives you the NV (floating point value) of the top SV on
1112the stack: the C<$x> in C<cos($x)>. Then we compute the cosine, and push
1113the result back as an NV. The C<X> in C<XPUSHn> means that the stack
1114should be extended if necessary - it can't be necessary here, because we
1115know there's room for one more item on the stack, since we've just
1116removed one! The C<XPUSH*> macros at least guarantee safety.
1117
1118Alternatively, you can fiddle with the stack directly: C<SP> gives you
1119the first element in your portion of the stack, and C<TOP*> gives you
1120the top SV/IV/NV/etc. on the stack. So, for instance, to do unary
1121negation of an integer:
1122
1123 SETi(-TOPi);
1124
1125Just set the integer value of the top stack entry to its negation.
1126
1127Argument stack manipulation in the core is exactly the same as it is in
1128XSUBs - see L<perlxstut>, L<perlxs> and L<perlguts> for a longer
1129description of the macros used in stack manipulation.
1130
1131=item Mark stack
1132
1133I say `your portion of the stack' above because PP code doesn't
1134necessarily get the whole stack to itself: if your function calls
1135another function, you'll only want to expose the arguments aimed for the
1136called function, and not (necessarily) let it get at your own data. The
1137way we do this is to have a `virtual' bottom-of-stack, exposed to each
1138function. The mark stack keeps bookmarks to locations in the argument
1139stack usable by each function. For instance, when dealing with a tied
1140variable, (internally, something with `P' magic) Perl has to call
1141methods for accesses to the tied variables. However, we need to separate
1142the arguments exposed to the method to the argument exposed to the
1143original function - the store or fetch or whatever it may be. Here's how
1144the tied C<push> is implemented; see C<av_push> in F<av.c>:
1145
1146 1 PUSHMARK(SP);
1147 2 EXTEND(SP,2);
1148 3 PUSHs(SvTIED_obj((SV*)av, mg));
1149 4 PUSHs(val);
1150 5 PUTBACK;
1151 6 ENTER;
1152 7 call_method("PUSH", G_SCALAR|G_DISCARD);
1153 8 LEAVE;
1154 9 POPSTACK;
13a2d996 1155
a422fd2d
SC
1156The lines which concern the mark stack are the first, fifth and last
1157lines: they save away, restore and remove the current position of the
1158argument stack.
1159
1160Let's examine the whole implementation, for practice:
1161
1162 1 PUSHMARK(SP);
1163
1164Push the current state of the stack pointer onto the mark stack. This is
1165so that when we've finished adding items to the argument stack, Perl
1166knows how many things we've added recently.
1167
1168 2 EXTEND(SP,2);
1169 3 PUSHs(SvTIED_obj((SV*)av, mg));
1170 4 PUSHs(val);
1171
1172We're going to add two more items onto the argument stack: when you have
1173a tied array, the C<PUSH> subroutine receives the object and the value
1174to be pushed, and that's exactly what we have here - the tied object,
1175retrieved with C<SvTIED_obj>, and the value, the SV C<val>.
1176
1177 5 PUTBACK;
1178
1179Next we tell Perl to make the change to the global stack pointer: C<dSP>
1180only gave us a local copy, not a reference to the global.
1181
1182 6 ENTER;
1183 7 call_method("PUSH", G_SCALAR|G_DISCARD);
1184 8 LEAVE;
1185
1186C<ENTER> and C<LEAVE> localise a block of code - they make sure that all
1187variables are tidied up, everything that has been localised gets
1188its previous value returned, and so on. Think of them as the C<{> and
1189C<}> of a Perl block.
1190
1191To actually do the magic method call, we have to call a subroutine in
1192Perl space: C<call_method> takes care of that, and it's described in
1193L<perlcall>. We call the C<PUSH> method in scalar context, and we're
1194going to discard its return value.
1195
1196 9 POPSTACK;
1197
1198Finally, we remove the value we placed on the mark stack, since we
1199don't need it any more.
1200
1201=item Save stack
1202
1203C doesn't have a concept of local scope, so perl provides one. We've
1204seen that C<ENTER> and C<LEAVE> are used as scoping braces; the save
1205stack implements the C equivalent of, for example:
1206
1207 {
1208 local $foo = 42;
1209 ...
1210 }
1211
1212See L<perlguts/Localising Changes> for how to use the save stack.
1213
1214=back
1215
1216=head2 Millions of Macros
1217
1218One thing you'll notice about the Perl source is that it's full of
1219macros. Some have called the pervasive use of macros the hardest thing
1220to understand, others find it adds to clarity. Let's take an example,
1221the code which implements the addition operator:
1222
1223 1 PP(pp_add)
1224 2 {
39644a26 1225 3 dSP; dATARGET; tryAMAGICbin(add,opASSIGN);
a422fd2d
SC
1226 4 {
1227 5 dPOPTOPnnrl_ul;
1228 6 SETn( left + right );
1229 7 RETURN;
1230 8 }
1231 9 }
1232
1233Every line here (apart from the braces, of course) contains a macro. The
1234first line sets up the function declaration as Perl expects for PP code;
1235line 3 sets up variable declarations for the argument stack and the
1236target, the return value of the operation. Finally, it tries to see if
1237the addition operation is overloaded; if so, the appropriate subroutine
1238is called.
1239
1240Line 5 is another variable declaration - all variable declarations start
1241with C<d> - which pops from the top of the argument stack two NVs (hence
1242C<nn>) and puts them into the variables C<right> and C<left>, hence the
1243C<rl>. These are the two operands to the addition operator. Next, we
1244call C<SETn> to set the NV of the return value to the result of adding
1245the two values. This done, we return - the C<RETURN> macro makes sure
1246that our return value is properly handled, and we pass the next operator
1247to run back to the main run loop.
1248
1249Most of these macros are explained in L<perlapi>, and some of the more
1250important ones are explained in L<perlxs> as well. Pay special attention
1251to L<perlguts/Background and PERL_IMPLICIT_CONTEXT> for information on
1252the C<[pad]THX_?> macros.
1253
a422fd2d
SC
1254=head2 Poking at Perl
1255
1256To really poke around with Perl, you'll probably want to build Perl for
1257debugging, like this:
1258
1259 ./Configure -d -D optimize=-g
1260 make
1261
1262C<-g> is a flag to the C compiler to have it produce debugging
1263information which will allow us to step through a running program.
1264F<Configure> will also turn on the C<DEBUGGING> compilation symbol which
1265enables all the internal debugging code in Perl. There are a whole bunch
1266of things you can debug with this: L<perlrun> lists them all, and the
1267best way to find out about them is to play about with them. The most
1268useful options are probably
1269
1270 l Context (loop) stack processing
1271 t Trace execution
1272 o Method and overloading resolution
1273 c String/numeric conversions
1274
1275Some of the functionality of the debugging code can be achieved using XS
1276modules.
13a2d996 1277
a422fd2d
SC
1278 -Dr => use re 'debug'
1279 -Dx => use O 'Debug'
1280
1281=head2 Using a source-level debugger
1282
1283If the debugging output of C<-D> doesn't help you, it's time to step
1284through perl's execution with a source-level debugger.
1285
1286=over 3
1287
1288=item *
1289
1290We'll use C<gdb> for our examples here; the principles will apply to any
1291debugger, but check the manual of the one you're using.
1292
1293=back
1294
1295To fire up the debugger, type
1296
1297 gdb ./perl
1298
1299You'll want to do that in your Perl source tree so the debugger can read
1300the source code. You should see the copyright message, followed by the
1301prompt.
1302
1303 (gdb)
1304
1305C<help> will get you into the documentation, but here are the most
1306useful commands:
1307
1308=over 3
1309
1310=item run [args]
1311
1312Run the program with the given arguments.
1313
1314=item break function_name
1315
1316=item break source.c:xxx
1317
1318Tells the debugger that we'll want to pause execution when we reach
cea6626f 1319either the named function (but see L<perlguts/Internal Functions>!) or the given
a422fd2d
SC
1320line in the named source file.
1321
1322=item step
1323
1324Steps through the program a line at a time.
1325
1326=item next
1327
1328Steps through the program a line at a time, without descending into
1329functions.
1330
1331=item continue
1332
1333Run until the next breakpoint.
1334
1335=item finish
1336
1337Run until the end of the current function, then stop again.
1338
13a2d996 1339=item 'enter'
a422fd2d
SC
1340
1341Just pressing Enter will do the most recent operation again - it's a
1342blessing when stepping through miles of source code.
1343
1344=item print
1345
1346Execute the given C code and print its results. B<WARNING>: Perl makes
1347heavy use of macros, and F<gdb> is not aware of macros. You'll have to
1348substitute them yourself. So, for instance, you can't say
1349
1350 print SvPV_nolen(sv)
1351
1352but you have to say
1353
1354 print Perl_sv_2pv_nolen(sv)
1355
1356You may find it helpful to have a "macro dictionary", which you can
1357produce by saying C<cpp -dM perl.c | sort>. Even then, F<cpp> won't
1358recursively apply the macros for you.
1359
1360=back
1361
1362=head2 Dumping Perl Data Structures
1363
1364One way to get around this macro hell is to use the dumping functions in
1365F<dump.c>; these work a little like an internal
1366L<Devel::Peek|Devel::Peek>, but they also cover OPs and other structures
1367that you can't get at from Perl. Let's take an example. We'll use the
1368C<$a = $b + $c> we used before, but give it a bit of context:
1369C<$b = "6XXXX"; $c = 2.3;>. Where's a good place to stop and poke around?
1370
1371What about C<pp_add>, the function we examined earlier to implement the
1372C<+> operator:
1373
1374 (gdb) break Perl_pp_add
1375 Breakpoint 1 at 0x46249f: file pp_hot.c, line 309.
1376
cea6626f 1377Notice we use C<Perl_pp_add> and not C<pp_add> - see L<perlguts/Internal Functions>.
a422fd2d
SC
1378With the breakpoint in place, we can run our program:
1379
1380 (gdb) run -e '$b = "6XXXX"; $c = 2.3; $a = $b + $c'
1381
1382Lots of junk will go past as gdb reads in the relevant source files and
1383libraries, and then:
1384
1385 Breakpoint 1, Perl_pp_add () at pp_hot.c:309
39644a26 1386 309 dSP; dATARGET; tryAMAGICbin(add,opASSIGN);
a422fd2d
SC
1387 (gdb) step
1388 311 dPOPTOPnnrl_ul;
1389 (gdb)
1390
1391We looked at this bit of code before, and we said that C<dPOPTOPnnrl_ul>
1392arranges for two C<NV>s to be placed into C<left> and C<right> - let's
1393slightly expand it:
1394
1395 #define dPOPTOPnnrl_ul NV right = POPn; \
1396 SV *leftsv = TOPs; \
1397 NV left = USE_LEFT(leftsv) ? SvNV(leftsv) : 0.0
1398
1399C<POPn> takes the SV from the top of the stack and obtains its NV either
1400directly (if C<SvNOK> is set) or by calling the C<sv_2nv> function.
1401C<TOPs> takes the next SV from the top of the stack - yes, C<POPn> uses
1402C<TOPs> - but doesn't remove it. We then use C<SvNV> to get the NV from
1403C<leftsv> in the same way as before - yes, C<POPn> uses C<SvNV>.
1404
1405Since we don't have an NV for C<$b>, we'll have to use C<sv_2nv> to
1406convert it. If we step again, we'll find ourselves there:
1407
1408 Perl_sv_2nv (sv=0xa0675d0) at sv.c:1669
1409 1669 if (!sv)
1410 (gdb)
1411
1412We can now use C<Perl_sv_dump> to investigate the SV:
1413
1414 SV = PV(0xa057cc0) at 0xa0675d0
1415 REFCNT = 1
1416 FLAGS = (POK,pPOK)
1417 PV = 0xa06a510 "6XXXX"\0
1418 CUR = 5
1419 LEN = 6
1420 $1 = void
1421
1422We know we're going to get C<6> from this, so let's finish the
1423subroutine:
1424
1425 (gdb) finish
1426 Run till exit from #0 Perl_sv_2nv (sv=0xa0675d0) at sv.c:1671
1427 0x462669 in Perl_pp_add () at pp_hot.c:311
1428 311 dPOPTOPnnrl_ul;
1429
1430We can also dump out this op: the current op is always stored in
1431C<PL_op>, and we can dump it with C<Perl_op_dump>. This'll give us
1432similar output to L<B::Debug|B::Debug>.
1433
1434 {
1435 13 TYPE = add ===> 14
1436 TARG = 1
1437 FLAGS = (SCALAR,KIDS)
1438 {
1439 TYPE = null ===> (12)
1440 (was rv2sv)
1441 FLAGS = (SCALAR,KIDS)
1442 {
1443 11 TYPE = gvsv ===> 12
1444 FLAGS = (SCALAR)
1445 GV = main::b
1446 }
1447 }
1448
10f58044 1449# finish this later #
a422fd2d
SC
1450
1451=head2 Patching
1452
1453All right, we've now had a look at how to navigate the Perl sources and
1454some things you'll need to know when fiddling with them. Let's now get
1455on and create a simple patch. Here's something Larry suggested: if a
1456C<U> is the first active format during a C<pack>, (for example,
1457C<pack "U3C8", @stuff>) then the resulting string should be treated as
1458UTF8 encoded.
1459
1460How do we prepare to fix this up? First we locate the code in question -
1461the C<pack> happens at runtime, so it's going to be in one of the F<pp>
1462files. Sure enough, C<pp_pack> is in F<pp.c>. Since we're going to be
1463altering this file, let's copy it to F<pp.c~>.
1464
a6ec74c1
JH
1465[Well, it was in F<pp.c> when this tutorial was written. It has now been
1466split off with C<pp_unpack> to its own file, F<pp_pack.c>]
1467
a422fd2d
SC
1468Now let's look over C<pp_pack>: we take a pattern into C<pat>, and then
1469loop over the pattern, taking each format character in turn into
1470C<datum_type>. Then for each possible format character, we swallow up
1471the other arguments in the pattern (a field width, an asterisk, and so
1472on) and convert the next chunk input into the specified format, adding
1473it onto the output SV C<cat>.
1474
1475How do we know if the C<U> is the first format in the C<pat>? Well, if
1476we have a pointer to the start of C<pat> then, if we see a C<U> we can
1477test whether we're still at the start of the string. So, here's where
1478C<pat> is set up:
1479
1480 STRLEN fromlen;
1481 register char *pat = SvPVx(*++MARK, fromlen);
1482 register char *patend = pat + fromlen;
1483 register I32 len;
1484 I32 datumtype;
1485 SV *fromstr;
1486
1487We'll have another string pointer in there:
1488
1489 STRLEN fromlen;
1490 register char *pat = SvPVx(*++MARK, fromlen);
1491 register char *patend = pat + fromlen;
1492 + char *patcopy;
1493 register I32 len;
1494 I32 datumtype;
1495 SV *fromstr;
1496
1497And just before we start the loop, we'll set C<patcopy> to be the start
1498of C<pat>:
1499
1500 items = SP - MARK;
1501 MARK++;
1502 sv_setpvn(cat, "", 0);
1503 + patcopy = pat;
1504 while (pat < patend) {
1505
1506Now if we see a C<U> which was at the start of the string, we turn on
1507the UTF8 flag for the output SV, C<cat>:
1508
1509 + if (datumtype == 'U' && pat==patcopy+1)
1510 + SvUTF8_on(cat);
1511 if (datumtype == '#') {
1512 while (pat < patend && *pat != '\n')
1513 pat++;
1514
1515Remember that it has to be C<patcopy+1> because the first character of
1516the string is the C<U> which has been swallowed into C<datumtype!>
1517
1518Oops, we forgot one thing: what if there are spaces at the start of the
1519pattern? C<pack(" U*", @stuff)> will have C<U> as the first active
1520character, even though it's not the first thing in the pattern. In this
1521case, we have to advance C<patcopy> along with C<pat> when we see spaces:
1522
1523 if (isSPACE(datumtype))
1524 continue;
1525
1526needs to become
1527
1528 if (isSPACE(datumtype)) {
1529 patcopy++;
1530 continue;
1531 }
1532
1533OK. That's the C part done. Now we must do two additional things before
1534this patch is ready to go: we've changed the behaviour of Perl, and so
1535we must document that change. We must also provide some more regression
1536tests to make sure our patch works and doesn't create a bug somewhere
1537else along the line.
1538
b23b8711
MS
1539The regression tests for each operator live in F<t/op/>, and so we
1540make a copy of F<t/op/pack.t> to F<t/op/pack.t~>. Now we can add our
1541tests to the end. First, we'll test that the C<U> does indeed create
1542Unicode strings.
1543
1544t/op/pack.t has a sensible ok() function, but if it didn't we could
35c336e6 1545use the one from t/test.pl.
b23b8711 1546
35c336e6
MS
1547 require './test.pl';
1548 plan( tests => 159 );
b23b8711
MS
1549
1550so instead of this:
a422fd2d
SC
1551
1552 print 'not ' unless "1.20.300.4000" eq sprintf "%vd", pack("U*",1,20,300,4000);
1553 print "ok $test\n"; $test++;
1554
35c336e6
MS
1555we can write the more sensible (see L<Test::More> for a full
1556explanation of is() and other testing functions).
b23b8711 1557
35c336e6 1558 is( "1.20.300.4000", sprintf "%vd", pack("U*",1,20,300,4000),
812f5127 1559 "U* produces unicode" );
b23b8711 1560
a422fd2d
SC
1561Now we'll test that we got that space-at-the-beginning business right:
1562
35c336e6 1563 is( "1.20.300.4000", sprintf "%vd", pack(" U*",1,20,300,4000),
812f5127 1564 " with spaces at the beginning" );
a422fd2d
SC
1565
1566And finally we'll test that we don't make Unicode strings if C<U> is B<not>
1567the first active format:
1568
35c336e6 1569 isnt( v1.20.300.4000, sprintf "%vd", pack("C0U*",1,20,300,4000),
812f5127 1570 "U* not first isn't unicode" );
a422fd2d 1571
35c336e6
MS
1572Mustn't forget to change the number of tests which appears at the top,
1573or else the automated tester will get confused. This will either look
1574like this:
a422fd2d 1575
35c336e6
MS
1576 print "1..156\n";
1577
1578or this:
1579
1580 plan( tests => 156 );
a422fd2d
SC
1581
1582We now compile up Perl, and run it through the test suite. Our new
1583tests pass, hooray!
1584
1585Finally, the documentation. The job is never done until the paperwork is
1586over, so let's describe the change we've just made. The relevant place
1587is F<pod/perlfunc.pod>; again, we make a copy, and then we'll insert
1588this text in the description of C<pack>:
1589
1590 =item *
1591
1592 If the pattern begins with a C<U>, the resulting string will be treated
1593 as Unicode-encoded. You can force UTF8 encoding on in a string with an
1594 initial C<U0>, and the bytes that follow will be interpreted as Unicode
1595 characters. If you don't want this to happen, you can begin your pattern
1596 with C<C0> (or anything else) to force Perl not to UTF8 encode your
1597 string, and then follow this with a C<U*> somewhere in your pattern.
1598
1599All done. Now let's create the patch. F<Porting/patching.pod> tells us
1600that if we're making major changes, we should copy the entire directory
1601to somewhere safe before we begin fiddling, and then do
13a2d996 1602
a422fd2d
SC
1603 diff -ruN old new > patch
1604
1605However, we know which files we've changed, and we can simply do this:
1606
1607 diff -u pp.c~ pp.c > patch
1608 diff -u t/op/pack.t~ t/op/pack.t >> patch
1609 diff -u pod/perlfunc.pod~ pod/perlfunc.pod >> patch
1610
1611We end up with a patch looking a little like this:
1612
1613 --- pp.c~ Fri Jun 02 04:34:10 2000
1614 +++ pp.c Fri Jun 16 11:37:25 2000
1615 @@ -4375,6 +4375,7 @@
1616 register I32 items;
1617 STRLEN fromlen;
1618 register char *pat = SvPVx(*++MARK, fromlen);
1619 + char *patcopy;
1620 register char *patend = pat + fromlen;
1621 register I32 len;
1622 I32 datumtype;
1623 @@ -4405,6 +4406,7 @@
1624 ...
1625
1626And finally, we submit it, with our rationale, to perl5-porters. Job
1627done!
1628
f7e1e956
MS
1629=head2 Patching a core module
1630
1631This works just like patching anything else, with an extra
1632consideration. Many core modules also live on CPAN. If this is so,
1633patch the CPAN version instead of the core and send the patch off to
1634the module maintainer (with a copy to p5p). This will help the module
1635maintainer keep the CPAN version in sync with the core version without
1636constantly scanning p5p.
1637
acbe17fc
JP
1638=head2 Adding a new function to the core
1639
1640If, as part of a patch to fix a bug, or just because you have an
1641especially good idea, you decide to add a new function to the core,
1642discuss your ideas on p5p well before you start work. It may be that
1643someone else has already attempted to do what you are considering and
1644can give lots of good advice or even provide you with bits of code
1645that they already started (but never finished).
1646
1647You have to follow all of the advice given above for patching. It is
1648extremely important to test any addition thoroughly and add new tests
1649to explore all boundary conditions that your new function is expected
1650to handle. If your new function is used only by one module (e.g. toke),
1651then it should probably be named S_your_function (for static); on the
210b36aa 1652other hand, if you expect it to accessible from other functions in
acbe17fc
JP
1653Perl, you should name it Perl_your_function. See L<perlguts/Internal Functions>
1654for more details.
1655
1656The location of any new code is also an important consideration. Don't
1657just create a new top level .c file and put your code there; you would
1658have to make changes to Configure (so the Makefile is created properly),
1659as well as possibly lots of include files. This is strictly pumpking
1660business.
1661
1662It is better to add your function to one of the existing top level
1663source code files, but your choice is complicated by the nature of
1664the Perl distribution. Only the files that are marked as compiled
1665static are located in the perl executable. Everything else is located
1666in the shared library (or DLL if you are running under WIN32). So,
1667for example, if a function was only used by functions located in
1668toke.c, then your code can go in toke.c. If, however, you want to call
1669the function from universal.c, then you should put your code in another
1670location, for example util.c.
1671
1672In addition to writing your c-code, you will need to create an
1673appropriate entry in embed.pl describing your function, then run
1674'make regen_headers' to create the entries in the numerous header
1675files that perl needs to compile correctly. See L<perlguts/Internal Functions>
1676for information on the various options that you can set in embed.pl.
1677You will forget to do this a few (or many) times and you will get
1678warnings during the compilation phase. Make sure that you mention
1679this when you post your patch to P5P; the pumpking needs to know this.
1680
1681When you write your new code, please be conscious of existing code
884bad00 1682conventions used in the perl source files. See L<perlstyle> for
acbe17fc
JP
1683details. Although most of the guidelines discussed seem to focus on
1684Perl code, rather than c, they all apply (except when they don't ;).
1685See also I<Porting/patching.pod> file in the Perl source distribution
1686for lots of details about both formatting and submitting patches of
1687your changes.
1688
1689Lastly, TEST TEST TEST TEST TEST any code before posting to p5p.
1690Test on as many platforms as you can find. Test as many perl
1691Configure options as you can (e.g. MULTIPLICITY). If you have
1692profiling or memory tools, see L<EXTERNAL TOOLS FOR DEBUGGING PERL>
210b36aa 1693below for how to use them to further test your code. Remember that
acbe17fc
JP
1694most of the people on P5P are doing this on their own time and
1695don't have the time to debug your code.
f7e1e956
MS
1696
1697=head2 Writing a test
1698
1699Every module and built-in function has an associated test file (or
1700should...). If you add or change functionality, you have to write a
1701test. If you fix a bug, you have to write a test so that bug never
1702comes back. If you alter the docs, it would be nice to test what the
1703new documentation says.
1704
1705In short, if you submit a patch you probably also have to patch the
1706tests.
1707
1708For modules, the test file is right next to the module itself.
1709F<lib/strict.t> tests F<lib/strict.pm>. This is a recent innovation,
1710so there are some snags (and it would be wonderful for you to brush
1711them out), but it basically works that way. Everything else lives in
1712F<t/>.
1713
1714=over 3
1715
1716=item F<t/base/>
1717
1718Testing of the absolute basic functionality of Perl. Things like
1719C<if>, basic file reads and writes, simple regexes, etc. These are
1720run first in the test suite and if any of them fail, something is
1721I<really> broken.
1722
1723=item F<t/cmd/>
1724
1725These test the basic control structures, C<if/else>, C<while>,
35c336e6 1726subroutines, etc.
f7e1e956
MS
1727
1728=item F<t/comp/>
1729
1730Tests basic issues of how Perl parses and compiles itself.
1731
1732=item F<t/io/>
1733
1734Tests for built-in IO functions, including command line arguments.
1735
1736=item F<t/lib/>
1737
1738The old home for the module tests, you shouldn't put anything new in
1739here. There are still some bits and pieces hanging around in here
1740that need to be moved. Perhaps you could move them? Thanks!
1741
1742=item F<t/op/>
1743
1744Tests for perl's built in functions that don't fit into any of the
1745other directories.
1746
1747=item F<t/pod/>
1748
1749Tests for POD directives. There are still some tests for the Pod
1750modules hanging around in here that need to be moved out into F<lib/>.
1751
1752=item F<t/run/>
1753
1754Testing features of how perl actually runs, including exit codes and
1755handling of PERL* environment variables.
1756
1757=back
1758
1759The core uses the same testing style as the rest of Perl, a simple
1760"ok/not ok" run through Test::Harness, but there are a few special
1761considerations.
1762
35c336e6
MS
1763There are three ways to write a test in the core. Test::More,
1764t/test.pl and ad hoc C<print $test ? "ok 42\n" : "not ok 42\n">. The
1765decision of which to use depends on what part of the test suite you're
1766working on. This is a measure to prevent a high-level failure (such
1767as Config.pm breaking) from causing basic functionality tests to fail.
1768
1769=over 4
1770
1771=item t/base t/comp
1772
1773Since we don't know if require works, or even subroutines, use ad hoc
1774tests for these two. Step carefully to avoid using the feature being
1775tested.
1776
1777=item t/cmd t/run t/io t/op
1778
1779Now that basic require() and subroutines are tested, you can use the
1780t/test.pl library which emulates the important features of Test::More
1781while using a minimum of core features.
1782
1783You can also conditionally use certain libraries like Config, but be
1784sure to skip the test gracefully if it's not there.
1785
1786=item t/lib ext lib
1787
1788Now that the core of Perl is tested, Test::More can be used. You can
1789also use the full suite of core modules in the tests.
1790
1791=back
f7e1e956
MS
1792
1793When you say "make test" Perl uses the F<t/TEST> program to run the
1794test suite. All tests are run from the F<t/> directory, B<not> the
1795directory which contains the test. This causes some problems with the
1796tests in F<lib/>, so here's some opportunity for some patching.
1797
1798You must be triply conscious of cross-platform concerns. This usually
1799boils down to using File::Spec and avoiding things like C<fork()> and
1800C<system()> unless absolutely necessary.
1801
e018f8be
JH
1802=head2 Special Make Test Targets
1803
1804There are various special make targets that can be used to test Perl
1805slightly differently than the standard "test" target. Not all them
1806are expected to give a 100% success rate. Many of them have several
1807aliases.
1808
1809=over 4
1810
1811=item coretest
1812
1813Run F<perl> on all but F<lib/*> tests.
1814
1815=item test.deparse
1816
1817Run all the tests through the B::Deparse. Not all tests will succeed.
1818
1819=item minitest
1820
1821Run F<miniperl> on F<t/base>, F<t/comp>, F<t/cmd>, F<t/run>, F<t/io>,
1822F<t/op>, and F<t/uni> tests.
1823
1824=item test.third check.third utest.third ucheck.third
1825
1826(Only in Tru64) Run all the tests using the memory leak + naughty
1827memory access tool "Third Degree". The log files will be named
1828F<perl3.log.testname>.
1829
1830=item test.torture torturetest
1831
1832Run all the usual tests and some extra tests. As of Perl 5.8.0 the
1833only extra tests are Abigail's JAPHs, t/japh/abigail.t.
1834
1835You can also run the torture test with F<t/harness> by giving
1836C<-torture> argument to F<t/harness>.
1837
1838=item utest ucheck test.utf8 check.utf8
1839
1840Run all the tests with -Mutf8. Not all tests will succeed.
1841
1842=back
f7e1e956 1843
902b9dbf
MF
1844=head1 EXTERNAL TOOLS FOR DEBUGGING PERL
1845
1846Sometimes it helps to use external tools while debugging and
1847testing Perl. This section tries to guide you through using
1848some common testing and debugging tools with Perl. This is
1849meant as a guide to interfacing these tools with Perl, not
1850as any kind of guide to the use of the tools themselves.
1851
1852=head2 Rational Software's Purify
1853
1854Purify is a commercial tool that is helpful in identifying
1855memory overruns, wild pointers, memory leaks and other such
1856badness. Perl must be compiled in a specific way for
1857optimal testing with Purify. Purify is available under
1858Windows NT, Solaris, HP-UX, SGI, and Siemens Unix.
1859
1860The only currently known leaks happen when there are
1861compile-time errors within eval or require. (Fixing these
1862is non-trivial, unfortunately, but they must be fixed
1863eventually.)
1864
1865=head2 Purify on Unix
1866
1867On Unix, Purify creates a new Perl binary. To get the most
1868benefit out of Purify, you should create the perl to Purify
1869using:
1870
1871 sh Configure -Accflags=-DPURIFY -Doptimize='-g' \
1872 -Uusemymalloc -Dusemultiplicity
1873
1874where these arguments mean:
1875
1876=over 4
1877
1878=item -Accflags=-DPURIFY
1879
1880Disables Perl's arena memory allocation functions, as well as
1881forcing use of memory allocation functions derived from the
1882system malloc.
1883
1884=item -Doptimize='-g'
1885
1886Adds debugging information so that you see the exact source
1887statements where the problem occurs. Without this flag, all
1888you will see is the source filename of where the error occurred.
1889
1890=item -Uusemymalloc
1891
1892Disable Perl's malloc so that Purify can more closely monitor
1893allocations and leaks. Using Perl's malloc will make Purify
1894report most leaks in the "potential" leaks category.
1895
1896=item -Dusemultiplicity
1897
1898Enabling the multiplicity option allows perl to clean up
1899thoroughly when the interpreter shuts down, which reduces the
1900number of bogus leak reports from Purify.
1901
1902=back
1903
1904Once you've compiled a perl suitable for Purify'ing, then you
1905can just:
1906
1907 make pureperl
1908
1909which creates a binary named 'pureperl' that has been Purify'ed.
1910This binary is used in place of the standard 'perl' binary
1911when you want to debug Perl memory problems.
1912
1f56d61a
JH
1913To minimize the number of memory leak false alarms
1914(see L</PERL_DESTRUCT_LEVEL>), set environment variable
1915PERL_DESTRUCT_LEVEL to 2.
1916
1917 setenv PERL_DESTRUCT_LEVEL 2
1918
1919In Bourne-type shells:
1920
1921 PERL_DESTRUCT_LEVEL=2
1922 export PERL_DESTRUCT_LEVEL
1923
902b9dbf
MF
1924As an example, to show any memory leaks produced during the
1925standard Perl testset you would create and run the Purify'ed
1926perl as:
1927
1928 make pureperl
1929 cd t
1930 ../pureperl -I../lib harness
1931
1932which would run Perl on test.pl and report any memory problems.
1933
1934Purify outputs messages in "Viewer" windows by default. If
1935you don't have a windowing environment or if you simply
1936want the Purify output to unobtrusively go to a log file
1937instead of to the interactive window, use these following
1938options to output to the log file "perl.log":
1939
1940 setenv PURIFYOPTIONS "-chain-length=25 -windows=no \
1941 -log-file=perl.log -append-logfile=yes"
1942
1943If you plan to use the "Viewer" windows, then you only need this option:
1944
1945 setenv PURIFYOPTIONS "-chain-length=25"
1946
c406981e
JH
1947In Bourne-type shells:
1948
98631ff8
JL
1949 PURIFYOPTIONS="..."
1950 export PURIFYOPTIONS
c406981e
JH
1951
1952or if you have the "env" utility:
1953
98631ff8 1954 env PURIFYOPTIONS="..." ../pureperl ...
c406981e 1955
902b9dbf
MF
1956=head2 Purify on NT
1957
1958Purify on Windows NT instruments the Perl binary 'perl.exe'
1959on the fly. There are several options in the makefile you
1960should change to get the most use out of Purify:
1961
1962=over 4
1963
1964=item DEFINES
1965
1966You should add -DPURIFY to the DEFINES line so the DEFINES
1967line looks something like:
1968
1969 DEFINES = -DWIN32 -D_CONSOLE -DNO_STRICT $(CRYPT_FLAG) -DPURIFY=1
1970
1971to disable Perl's arena memory allocation functions, as
1972well as to force use of memory allocation functions derived
1973from the system malloc.
1974
1975=item USE_MULTI = define
1976
1977Enabling the multiplicity option allows perl to clean up
1978thoroughly when the interpreter shuts down, which reduces the
1979number of bogus leak reports from Purify.
1980
1981=item #PERL_MALLOC = define
1982
1983Disable Perl's malloc so that Purify can more closely monitor
1984allocations and leaks. Using Perl's malloc will make Purify
1985report most leaks in the "potential" leaks category.
1986
1987=item CFG = Debug
1988
1989Adds debugging information so that you see the exact source
1990statements where the problem occurs. Without this flag, all
1991you will see is the source filename of where the error occurred.
1992
1993=back
1994
1995As an example, to show any memory leaks produced during the
1996standard Perl testset you would create and run Purify as:
1997
1998 cd win32
1999 make
2000 cd ../t
2001 purify ../perl -I../lib harness
2002
2003which would instrument Perl in memory, run Perl on test.pl,
2004then finally report any memory problems.
2005
f134cc4e
JH
2006B<NOTE>: as of Perl 5.8.0, the ext/Encode/t/Unicode.t takes
2007extraordinarily long (hours?) to complete under Purify. It has been
2008theorized that it would eventually finish, but nobody has so far been
2009patient enough :-) (This same extreme slowdown has been seen also with
2010the Third Degree tool, so the said test must be doing something that
2011is quite unfriendly for memory debuggers.) It is suggested that you
2012simply kill away that testing process.
2013
2014=head2 Compaq's/Digital's/HP's Third Degree
09187cb1
JH
2015
2016Third Degree is a tool for memory leak detection and memory access checks.
2017It is one of the many tools in the ATOM toolkit. The toolkit is only
2018available on Tru64 (formerly known as Digital UNIX formerly known as
2019DEC OSF/1).
2020
2021When building Perl, you must first run Configure with -Doptimize=-g
2022and -Uusemymalloc flags, after that you can use the make targets
51a35ef1
JH
2023"perl.third" and "test.third". (What is required is that Perl must be
2024compiled using the C<-g> flag, you may need to re-Configure.)
09187cb1 2025
64cea5fd 2026The short story is that with "atom" you can instrument the Perl
83f0ef60 2027executable to create a new executable called F<perl.third>. When the
4ae3d70a 2028instrumented executable is run, it creates a log of dubious memory
83f0ef60 2029traffic in file called F<perl.3log>. See the manual pages of atom and
4ae3d70a
JH
2030third for more information. The most extensive Third Degree
2031documentation is available in the Compaq "Tru64 UNIX Programmer's
2032Guide", chapter "Debugging Programs with Third Degree".
64cea5fd 2033
9c54ecba 2034The "test.third" leaves a lot of files named F<foo_bar.3log> in the t/
64cea5fd
JH
2035subdirectory. There is a problem with these files: Third Degree is so
2036effective that it finds problems also in the system libraries.
9c54ecba
JH
2037Therefore you should used the Porting/thirdclean script to cleanup
2038the F<*.3log> files.
64cea5fd
JH
2039
2040There are also leaks that for given certain definition of a leak,
2041aren't. See L</PERL_DESTRUCT_LEVEL> for more information.
2042
2043=head2 PERL_DESTRUCT_LEVEL
2044
2045If you want to run any of the tests yourself manually using the
2046pureperl or perl.third executables, please note that by default
2047perl B<does not> explicitly cleanup all the memory it has allocated
2048(such as global memory arenas) but instead lets the exit() of
2049the whole program "take care" of such allocations, also known
2050as "global destruction of objects".
2051
2052There is a way to tell perl to do complete cleanup: set the
2053environment variable PERL_DESTRUCT_LEVEL to a non-zero value.
2054The t/TEST wrapper does set this to 2, and this is what you
2055need to do too, if you don't want to see the "global leaks":
1f56d61a 2056For example, for "third-degreed" Perl:
64cea5fd 2057
1f56d61a 2058 env PERL_DESTRUCT_LEVEL=2 ./perl.third -Ilib t/foo/bar.t
09187cb1 2059
414f2397
RGS
2060(Note: the mod_perl apache module uses also this environment variable
2061for its own purposes and extended its semantics. Refer to the mod_perl
2062documentation for more information.)
2063
51a35ef1
JH
2064=head2 Profiling
2065
2066Depending on your platform there are various of profiling Perl.
2067
2068There are two commonly used techniques of profiling executables:
10f58044 2069I<statistical time-sampling> and I<basic-block counting>.
51a35ef1
JH
2070
2071The first method takes periodically samples of the CPU program
2072counter, and since the program counter can be correlated with the code
2073generated for functions, we get a statistical view of in which
2074functions the program is spending its time. The caveats are that very
2075small/fast functions have lower probability of showing up in the
2076profile, and that periodically interrupting the program (this is
2077usually done rather frequently, in the scale of milliseconds) imposes
2078an additional overhead that may skew the results. The first problem
2079can be alleviated by running the code for longer (in general this is a
2080good idea for profiling), the second problem is usually kept in guard
2081by the profiling tools themselves.
2082
10f58044 2083The second method divides up the generated code into I<basic blocks>.
51a35ef1
JH
2084Basic blocks are sections of code that are entered only in the
2085beginning and exited only at the end. For example, a conditional jump
2086starts a basic block. Basic block profiling usually works by
10f58044 2087I<instrumenting> the code by adding I<enter basic block #nnnn>
51a35ef1
JH
2088book-keeping code to the generated code. During the execution of the
2089code the basic block counters are then updated appropriately. The
2090caveat is that the added extra code can skew the results: again, the
2091profiling tools usually try to factor their own effects out of the
2092results.
2093
83f0ef60
JH
2094=head2 Gprof Profiling
2095
51a35ef1
JH
2096gprof is a profiling tool available in many UNIX platforms,
2097it uses F<statistical time-sampling>.
83f0ef60
JH
2098
2099You can build a profiled version of perl called "perl.gprof" by
51a35ef1
JH
2100invoking the make target "perl.gprof" (What is required is that Perl
2101must be compiled using the C<-pg> flag, you may need to re-Configure).
2102Running the profiled version of Perl will create an output file called
2103F<gmon.out> is created which contains the profiling data collected
2104during the execution.
83f0ef60
JH
2105
2106The gprof tool can then display the collected data in various ways.
2107Usually gprof understands the following options:
2108
2109=over 4
2110
2111=item -a
2112
2113Suppress statically defined functions from the profile.
2114
2115=item -b
2116
2117Suppress the verbose descriptions in the profile.
2118
2119=item -e routine
2120
2121Exclude the given routine and its descendants from the profile.
2122
2123=item -f routine
2124
2125Display only the given routine and its descendants in the profile.
2126
2127=item -s
2128
2129Generate a summary file called F<gmon.sum> which then may be given
2130to subsequent gprof runs to accumulate data over several runs.
2131
2132=item -z
2133
2134Display routines that have zero usage.
2135
2136=back
2137
2138For more detailed explanation of the available commands and output
2139formats, see your own local documentation of gprof.
2140
51a35ef1
JH
2141=head2 GCC gcov Profiling
2142
10f58044 2143Starting from GCC 3.0 I<basic block profiling> is officially available
51a35ef1
JH
2144for the GNU CC.
2145
2146You can build a profiled version of perl called F<perl.gcov> by
2147invoking the make target "perl.gcov" (what is required that Perl must
2148be compiled using gcc with the flags C<-fprofile-arcs
2149-ftest-coverage>, you may need to re-Configure).
2150
2151Running the profiled version of Perl will cause profile output to be
2152generated. For each source file an accompanying ".da" file will be
2153created.
2154
2155To display the results you use the "gcov" utility (which should
2156be installed if you have gcc 3.0 or newer installed). F<gcov> is
2157run on source code files, like this
2158
2159 gcov sv.c
2160
2161which will cause F<sv.c.gcov> to be created. The F<.gcov> files
2162contain the source code annotated with relative frequencies of
2163execution indicated by "#" markers.
2164
2165Useful options of F<gcov> include C<-b> which will summarise the
2166basic block, branch, and function call coverage, and C<-c> which
2167instead of relative frequencies will use the actual counts. For
2168more information on the use of F<gcov> and basic block profiling
2169with gcc, see the latest GNU CC manual, as of GCC 3.0 see
2170
2171 http://gcc.gnu.org/onlinedocs/gcc-3.0/gcc.html
2172
2173and its section titled "8. gcov: a Test Coverage Program"
2174
2175 http://gcc.gnu.org/onlinedocs/gcc-3.0/gcc_8.html#SEC132
2176
4ae3d70a
JH
2177=head2 Pixie Profiling
2178
51a35ef1
JH
2179Pixie is a profiling tool available on IRIX and Tru64 (aka Digital
2180UNIX aka DEC OSF/1) platforms. Pixie does its profiling using
10f58044 2181I<basic-block counting>.
4ae3d70a 2182
83f0ef60 2183You can build a profiled version of perl called F<perl.pixie> by
51a35ef1
JH
2184invoking the make target "perl.pixie" (what is required is that Perl
2185must be compiled using the C<-g> flag, you may need to re-Configure).
2186
2187In Tru64 a file called F<perl.Addrs> will also be silently created,
2188this file contains the addresses of the basic blocks. Running the
2189profiled version of Perl will create a new file called "perl.Counts"
2190which contains the counts for the basic block for that particular
2191program execution.
4ae3d70a 2192
51a35ef1 2193To display the results you use the F<prof> utility. The exact
4ae3d70a
JH
2194incantation depends on your operating system, "prof perl.Counts" in
2195IRIX, and "prof -pixie -all -L. perl" in Tru64.
2196
6c41479b
JH
2197In IRIX the following prof options are available:
2198
2199=over 4
2200
2201=item -h
2202
2203Reports the most heavily used lines in descending order of use.
6e36760b 2204Useful for finding the hotspot lines.
6c41479b
JH
2205
2206=item -l
2207
2208Groups lines by procedure, with procedures sorted in descending order of use.
2209Within a procedure, lines are listed in source order.
6e36760b 2210Useful for finding the hotspots of procedures.
6c41479b
JH
2211
2212=back
2213
2214In Tru64 the following options are available:
2215
2216=over 4
2217
3958b146 2218=item -p[rocedures]
6c41479b 2219
3958b146 2220Procedures sorted in descending order by the number of cycles executed
6e36760b 2221in each procedure. Useful for finding the hotspot procedures.
6c41479b
JH
2222(This is the default option.)
2223
24000d2f 2224=item -h[eavy]
6c41479b 2225
6e36760b
JH
2226Lines sorted in descending order by the number of cycles executed in
2227each line. Useful for finding the hotspot lines.
6c41479b 2228
24000d2f 2229=item -i[nvocations]
6c41479b 2230
6e36760b
JH
2231The called procedures are sorted in descending order by number of calls
2232made to the procedures. Useful for finding the most used procedures.
6c41479b 2233
24000d2f 2234=item -l[ines]
6c41479b
JH
2235
2236Grouped by procedure, sorted by cycles executed per procedure.
6e36760b 2237Useful for finding the hotspots of procedures.
6c41479b
JH
2238
2239=item -testcoverage
2240
2241The compiler emitted code for these lines, but the code was unexecuted.
2242
24000d2f 2243=item -z[ero]
6c41479b
JH
2244
2245Unexecuted procedures.
2246
aa500c9e 2247=back
6c41479b
JH
2248
2249For further information, see your system's manual pages for pixie and prof.
4ae3d70a 2250
b8ddf6b3
SB
2251=head2 Miscellaneous tricks
2252
2253=over 4
2254
2255=item *
2256
cc177e1a 2257Those debugging perl with the DDD frontend over gdb may find the
b8ddf6b3
SB
2258following useful:
2259
2260You can extend the data conversion shortcuts menu, so for example you
2261can display an SV's IV value with one click, without doing any typing.
2262To do that simply edit ~/.ddd/init file and add after:
2263
2264 ! Display shortcuts.
2265 Ddd*gdbDisplayShortcuts: \
2266 /t () // Convert to Bin\n\
2267 /d () // Convert to Dec\n\
2268 /x () // Convert to Hex\n\
2269 /o () // Convert to Oct(\n\
2270
2271the following two lines:
2272
2273 ((XPV*) (())->sv_any )->xpv_pv // 2pvx\n\
2274 ((XPVIV*) (())->sv_any )->xiv_iv // 2ivx
2275
2276so now you can do ivx and pvx lookups or you can plug there the
2277sv_peek "conversion":
2278
2279 Perl_sv_peek(my_perl, (SV*)()) // sv_peek
2280
2281(The my_perl is for threaded builds.)
2282Just remember that every line, but the last one, should end with \n\
2283
2284Alternatively edit the init file interactively via:
22853rd mouse button -> New Display -> Edit Menu
2286
2287Note: you can define up to 20 conversion shortcuts in the gdb
2288section.
2289
9965345d
JH
2290=item *
2291
2292If you see in a debugger a memory area mysteriously full of 0xabababab,
2293you may be seeing the effect of the Poison() macro, see L<perlclib>.
2294
b8ddf6b3
SB
2295=back
2296
a422fd2d
SC
2297=head2 CONCLUSION
2298
2299We've had a brief look around the Perl source, an overview of the stages
2300F<perl> goes through when it's running your code, and how to use a
902b9dbf
MF
2301debugger to poke at the Perl guts. We took a very simple problem and
2302demonstrated how to solve it fully - with documentation, regression
2303tests, and finally a patch for submission to p5p. Finally, we talked
2304about how to use external tools to debug and test Perl.
a422fd2d
SC
2305
2306I'd now suggest you read over those references again, and then, as soon
2307as possible, get your hands dirty. The best way to learn is by doing,
2308so:
2309
2310=over 3
2311
2312=item *
2313
2314Subscribe to perl5-porters, follow the patches and try and understand
2315them; don't be afraid to ask if there's a portion you're not clear on -
2316who knows, you may unearth a bug in the patch...
2317
2318=item *
2319
2320Keep up to date with the bleeding edge Perl distributions and get
2321familiar with the changes. Try and get an idea of what areas people are
2322working on and the changes they're making.
2323
2324=item *
2325
3e148164 2326Do read the README associated with your operating system, e.g. README.aix
a1f349fd
MB
2327on the IBM AIX OS. Don't hesitate to supply patches to that README if
2328you find anything missing or changed over a new OS release.
2329
2330=item *
2331
a422fd2d
SC
2332Find an area of Perl that seems interesting to you, and see if you can
2333work out how it works. Scan through the source, and step over it in the
2334debugger. Play, poke, investigate, fiddle! You'll probably get to
2335understand not just your chosen area but a much wider range of F<perl>'s
2336activity as well, and probably sooner than you'd think.
2337
2338=back
2339
2340=over 3
2341
2342=item I<The Road goes ever on and on, down from the door where it began.>
2343
2344=back
2345
2346If you can do these things, you've started on the long road to Perl porting.
2347Thanks for wanting to help make Perl better - and happy hacking!
2348
e8cd7eae
GS
2349=head1 AUTHOR
2350
2351This document was written by Nathan Torkington, and is maintained by
2352the perl5-porters mailing list.
2353