This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
perlhack.pod
[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 36in what does and does not change in the Perl language. Various
b432a672
AL
37releases of Perl are shepherded by a "pumpking", a porter
38responsible for gathering patches, deciding on a patch-by-patch,
f6c51b38 39feature-by-feature basis what will and will not go into the release.
caf100c0 40For instance, Gurusamy Sarathy was the pumpking for the 5.6 release of
961f29c6 41Perl, and Jarkko Hietaniemi was the pumpking for the 5.8 release, and
1a88dbf8 42Rafael Garcia-Suarez holds the pumpking crown for the 5.10 release.
e8cd7eae
GS
43
44In addition, various people are pumpkings for different things. For
961f29c6
MB
45instance, Andy Dougherty and Jarkko Hietaniemi did a grand job as the
46I<Configure> pumpkin up till the 5.8 release. For the 5.10 release
47H.Merijn Brand took over.
e8cd7eae
GS
48
49Larry sees Perl development along the lines of the US government:
50there's the Legislature (the porters), the Executive branch (the
51pumpkings), and the Supreme Court (Larry). The legislature can
52discuss and submit patches to the executive branch all they like, but
53the executive branch is free to veto them. Rarely, the Supreme Court
54will side with the executive branch over the legislature, or the
55legislature over the executive branch. Mostly, however, the
56legislature and the executive branch are supposed to get along and
57work out their differences without impeachment or court cases.
58
59You might sometimes see reference to Rule 1 and Rule 2. Larry's power
60as Supreme Court is expressed in The Rules:
61
62=over 4
63
64=item 1
65
66Larry is always by definition right about how Perl should behave.
67This means he has final veto power on the core functionality.
68
69=item 2
70
71Larry is allowed to change his mind about any matter at a later date,
72regardless of whether he previously invoked Rule 1.
73
74=back
75
76Got that? Larry is always right, even when he was wrong. It's rare
77to see either Rule exercised, but they are often alluded to.
78
79New features and extensions to the language are contentious, because
80the criteria used by the pumpkings, Larry, and other porters to decide
81which features should be implemented and incorporated are not codified
82in a few small design goals as with some other languages. Instead,
83the heuristics are flexible and often difficult to fathom. Here is
84one person's list, roughly in decreasing order of importance, of
85heuristics that new features have to be weighed against:
86
87=over 4
88
89=item Does concept match the general goals of Perl?
90
91These haven't been written anywhere in stone, but one approximation
92is:
93
94 1. Keep it fast, simple, and useful.
95 2. Keep features/concepts as orthogonal as possible.
96 3. No arbitrary limits (platforms, data sizes, cultures).
97 4. Keep it open and exciting to use/patch/advocate Perl everywhere.
98 5. Either assimilate new technologies, or build bridges to them.
99
100=item Where is the implementation?
101
102All the talk in the world is useless without an implementation. In
103almost every case, the person or people who argue for a new feature
104will be expected to be the ones who implement it. Porters capable
105of coding new features have their own agendas, and are not available
106to implement your (possibly good) idea.
107
108=item Backwards compatibility
109
110It's a cardinal sin to break existing Perl programs. New warnings are
111contentious--some say that a program that emits warnings is not
112broken, while others say it is. Adding keywords has the potential to
113break programs, changing the meaning of existing token sequences or
114functions might break programs.
115
116=item Could it be a module instead?
117
118Perl 5 has extension mechanisms, modules and XS, specifically to avoid
119the need to keep changing the Perl interpreter. You can write modules
120that export functions, you can give those functions prototypes so they
121can be called like built-in functions, you can even write XS code to
122mess with the runtime data structures of the Perl interpreter if you
123want to implement really complicated things. If it can be done in a
124module instead of in the core, it's highly unlikely to be added.
125
126=item Is the feature generic enough?
127
128Is this something that only the submitter wants added to the language,
129or would it be broadly useful? Sometimes, instead of adding a feature
130with a tight focus, the porters might decide to wait until someone
131implements the more generalized feature. For instance, instead of
b432a672 132implementing a "delayed evaluation" feature, the porters are waiting
e8cd7eae
GS
133for a macro system that would permit delayed evaluation and much more.
134
135=item Does it potentially introduce new bugs?
136
137Radical rewrites of large chunks of the Perl interpreter have the
138potential to introduce new bugs. The smaller and more localized the
139change, the better.
140
141=item Does it preclude other desirable features?
142
143A patch is likely to be rejected if it closes off future avenues of
144development. For instance, a patch that placed a true and final
145interpretation on prototypes is likely to be rejected because there
146are still options for the future of prototypes that haven't been
147addressed.
148
149=item Is the implementation robust?
150
151Good patches (tight code, complete, correct) stand more chance of
152going in. Sloppy or incorrect patches might be placed on the back
153burner until the pumpking has time to fix, or might be discarded
154altogether without further notice.
155
156=item Is the implementation generic enough to be portable?
157
158The worst patches make use of a system-specific features. It's highly
159unlikely that nonportable additions to the Perl language will be
160accepted.
161
a936dd3c
NC
162=item Is the implementation tested?
163
164Patches which change behaviour (fixing bugs or introducing new features)
165must include regression tests to verify that everything works as expected.
166Without tests provided by the original author, how can anyone else changing
167perl in the future be sure that they haven't unwittingly broken the behaviour
168the patch implements? And without tests, how can the patch's author be
9d077eaa 169confident that his/her hard work put into the patch won't be accidentally
a936dd3c
NC
170thrown away by someone in the future?
171
e8cd7eae
GS
172=item Is there enough documentation?
173
174Patches without documentation are probably ill-thought out or
175incomplete. Nothing can be added without documentation, so submitting
176a patch for the appropriate manpages as well as the source code is
a936dd3c 177always a good idea.
e8cd7eae
GS
178
179=item Is there another way to do it?
180
b432a672
AL
181Larry said "Although the Perl Slogan is I<There's More Than One Way
182to Do It>, I hesitate to make 10 ways to do something". This is a
e8cd7eae
GS
183tricky heuristic to navigate, though--one man's essential addition is
184another man's pointless cruft.
185
186=item Does it create too much work?
187
188Work for the pumpking, work for Perl programmers, work for module
189authors, ... Perl is supposed to be easy.
190
f6c51b38
GS
191=item Patches speak louder than words
192
193Working code is always preferred to pie-in-the-sky ideas. A patch to
194add a feature stands a much higher chance of making it to the language
195than does a random feature request, no matter how fervently argued the
b432a672 196request might be. This ties into "Will it be useful?", as the fact
f6c51b38
GS
197that someone took the time to make the patch demonstrates a strong
198desire for the feature.
199
e8cd7eae
GS
200=back
201
b432a672
AL
202If you're on the list, you might hear the word "core" bandied
203around. It refers to the standard distribution. "Hacking on the
204core" means you're changing the C source code to the Perl
205interpreter. "A core module" is one that ships with Perl.
e8cd7eae 206
a1f349fd
MB
207=head2 Keeping in sync
208
e8cd7eae 209The source code to the Perl interpreter, in its different versions, is
f224927c
JH
210kept in a repository managed by a revision control system ( which is
211currently the Perforce program, see http://perforce.com/ ). The
e8cd7eae
GS
212pumpkings and a few others have access to the repository to check in
213changes. Periodically the pumpking for the development version of Perl
214will release a new version, so the rest of the porters can see what's
2be4c08b
GS
215changed. The current state of the main trunk of repository, and patches
216that describe the individual changes that have happened since the last
217public release are available at this location:
218
eb8afeef
SP
219 http://public.activestate.com/pub/apc/
220 ftp://public.activestate.com/pub/apc/
2be4c08b 221
0cfb3454
GS
222If you're looking for a particular change, or a change that affected
223a particular set of files, you may find the B<Perl Repository Browser>
224useful:
225
226 http://public.activestate.com/cgi-bin/perlbrowse
227
228You may also want to subscribe to the perl5-changes mailing list to
229receive a copy of each patch that gets submitted to the maintenance
230and development "branches" of the perl repository. See
231http://lists.perl.org/ for subscription information.
232
a1f349fd
MB
233If you are a member of the perl5-porters mailing list, it is a good
234thing to keep in touch with the most recent changes. If not only to
235verify if what you would have posted as a bug report isn't already
236solved in the most recent available perl development branch, also
237known as perl-current, bleading edge perl, bleedperl or bleadperl.
2be4c08b
GS
238
239Needless to say, the source code in perl-current is usually in a perpetual
240state of evolution. You should expect it to be very buggy. Do B<not> use
241it for any purpose other than testing and development.
e8cd7eae 242
3e148164
JH
243Keeping in sync with the most recent branch can be done in several ways,
244but the most convenient and reliable way is using B<rsync>, available at
245ftp://rsync.samba.org/pub/rsync/ . (You can also get the most recent
246branch by FTP.)
a1f349fd
MB
247
248If you choose to keep in sync using rsync, there are two approaches
3e148164 249to doing so:
a1f349fd
MB
250
251=over 4
252
253=item rsync'ing the source tree
254
3e148164 255Presuming you are in the directory where your perl source resides
b432a672 256and you have rsync installed and available, you can "upgrade" to
a1f349fd
MB
257the bleadperl using:
258
eb8afeef 259 # rsync -avz rsync://public.activestate.com/perl-current/ .
a1f349fd
MB
260
261This takes care of updating every single item in the source tree to
262the latest applied patch level, creating files that are new (to your
263distribution) and setting date/time stamps of existing files to
264reflect the bleadperl status.
265
c6d0653e
LC
266Note that this will not delete any files that were in '.' before
267the rsync. Once you are sure that the rsync is running correctly,
268run it with the --delete and the --dry-run options like this:
269
eb8afeef 270 # rsync -avz --delete --dry-run rsync://public.activestate.com/perl-current/ .
c6d0653e
LC
271
272This will I<simulate> an rsync run that also deletes files not
273present in the bleadperl master copy. Observe the results from
274this run closely. If you are sure that the actual run would delete
275no files precious to you, you could remove the '--dry-run' option.
276
a1f349fd
MB
277You can than check what patch was the latest that was applied by
278looking in the file B<.patch>, which will show the number of the
279latest patch.
280
281If you have more than one machine to keep in sync, and not all of
282them have access to the WAN (so you are not able to rsync all the
283source trees to the real source), there are some ways to get around
284this problem.
285
286=over 4
287
288=item Using rsync over the LAN
289
290Set up a local rsync server which makes the rsynced source tree
3e148164 291available to the LAN and sync the other machines against this
a1f349fd
MB
292directory.
293
1577cd80 294From http://rsync.samba.org/README.html :
a1f349fd
MB
295
296 "Rsync uses rsh or ssh for communication. It does not need to be
297 setuid and requires no special privileges for installation. It
3958b146 298 does not require an inetd entry or a daemon. You must, however,
a1f349fd
MB
299 have a working rsh or ssh system. Using ssh is recommended for
300 its security features."
301
302=item Using pushing over the NFS
303
304Having the other systems mounted over the NFS, you can take an
3e148164
JH
305active pushing approach by checking the just updated tree against
306the other not-yet synced trees. An example would be
307
308 #!/usr/bin/perl -w
309
310 use strict;
311 use File::Copy;
312
313 my %MF = map {
314 m/(\S+)/;
315 $1 => [ (stat $1)[2, 7, 9] ]; # mode, size, mtime
316 } `cat MANIFEST`;
317
318 my %remote = map { $_ => "/$_/pro/3gl/CPAN/perl-5.7.1" } qw(host1 host2);
319
320 foreach my $host (keys %remote) {
321 unless (-d $remote{$host}) {
322 print STDERR "Cannot Xsync for host $host\n";
323 next;
324 }
325 foreach my $file (keys %MF) {
326 my $rfile = "$remote{$host}/$file";
327 my ($mode, $size, $mtime) = (stat $rfile)[2, 7, 9];
328 defined $size or ($mode, $size, $mtime) = (0, 0, 0);
329 $size == $MF{$file}[1] && $mtime == $MF{$file}[2] and next;
330 printf "%4s %-34s %8d %9d %8d %9d\n",
331 $host, $file, $MF{$file}[1], $MF{$file}[2], $size, $mtime;
332 unlink $rfile;
333 copy ($file, $rfile);
334 utime time, $MF{$file}[2], $rfile;
335 chmod $MF{$file}[0], $rfile;
336 }
337 }
338
339though this is not perfect. It could be improved with checking
a1f349fd
MB
340file checksums before updating. Not all NFS systems support
341reliable utime support (when used over the NFS).
342
343=back
344
345=item rsync'ing the patches
346
347The source tree is maintained by the pumpking who applies patches to
348the files in the tree. These patches are either created by the
349pumpking himself using C<diff -c> after updating the file manually or
350by applying patches sent in by posters on the perl5-porters list.
351These patches are also saved and rsync'able, so you can apply them
352yourself to the source files.
353
354Presuming you are in a directory where your patches reside, you can
3e148164 355get them in sync with
a1f349fd 356
eb8afeef 357 # rsync -avz rsync://public.activestate.com/perl-current-diffs/ .
a1f349fd
MB
358
359This makes sure the latest available patch is downloaded to your
360patch directory.
361
3e148164 362It's then up to you to apply these patches, using something like
a1f349fd 363
7d3c2c28 364 # last="`cat ../perl-current/.patch`.gz"
eb8afeef 365 # rsync -avz rsync://public.activestate.com/perl-current-diffs/ .
a1f349fd
MB
366 # find . -name '*.gz' -newer $last -exec gzcat {} \; >blead.patch
367 # cd ../perl-current
368 # patch -p1 -N <../perl-current-diffs/blead.patch
369
370or, since this is only a hint towards how it works, use CPAN-patchaperl
371from Andreas K├Ânig to have better control over the patching process.
372
373=back
374
f7e1e956 375=head2 Why rsync the source tree
a1f349fd
MB
376
377=over 4
378
10f58044 379=item It's easier to rsync the source tree
a1f349fd
MB
380
381Since you don't have to apply the patches yourself, you are sure all
382files in the source tree are in the right state.
383
a1f349fd
MB
384=item It's more reliable
385
0cfb3454
GS
386While both the rsync-able source and patch areas are automatically
387updated every few minutes, keep in mind that applying patches may
388sometimes mean careful hand-holding, especially if your version of
389the C<patch> program does not understand how to deal with new files,
390files with 8-bit characters, or files without trailing newlines.
a1f349fd
MB
391
392=back
393
f7e1e956 394=head2 Why rsync the patches
a1f349fd
MB
395
396=over 4
397
10f58044 398=item It's easier to rsync the patches
a1f349fd
MB
399
400If you have more than one machine that you want to keep in track with
3e148164 401bleadperl, it's easier to rsync the patches only once and then apply
a1f349fd
MB
402them to all the source trees on the different machines.
403
404In case you try to keep in pace on 5 different machines, for which
405only one of them has access to the WAN, rsync'ing all the source
3e148164 406trees should than be done 5 times over the NFS. Having
a1f349fd 407rsync'ed the patches only once, I can apply them to all the source
3e148164 408trees automatically. Need you say more ;-)
a1f349fd
MB
409
410=item It's a good reference
411
412If you do not only like to have the most recent development branch,
413but also like to B<fix> bugs, or extend features, you want to dive
414into the sources. If you are a seasoned perl core diver, you don't
415need no manuals, tips, roadmaps, perlguts.pod or other aids to find
416your way around. But if you are a starter, the patches may help you
417in finding where you should start and how to change the bits that
418bug you.
419
420The file B<Changes> is updated on occasions the pumpking sees as his
421own little sync points. On those occasions, he releases a tar-ball of
422the current source tree (i.e. perl@7582.tar.gz), which will be an
423excellent point to start with when choosing to use the 'rsync the
424patches' scheme. Starting with perl@7582, which means a set of source
425files on which the latest applied patch is number 7582, you apply all
f18956b7 426succeeding patches available from then on (7583, 7584, ...).
a1f349fd
MB
427
428You can use the patches later as a kind of search archive.
429
430=over 4
431
432=item Finding a start point
433
434If you want to fix/change the behaviour of function/feature Foo, just
435scan the patches for patches that mention Foo either in the subject,
3e148164 436the comments, or the body of the fix. A good chance the patch shows
a1f349fd
MB
437you the files that are affected by that patch which are very likely
438to be the starting point of your journey into the guts of perl.
439
440=item Finding how to fix a bug
441
442If you've found I<where> the function/feature Foo misbehaves, but you
443don't know how to fix it (but you do know the change you want to
444make), you can, again, peruse the patches for similar changes and
445look how others apply the fix.
446
447=item Finding the source of misbehaviour
448
449When you keep in sync with bleadperl, the pumpking would love to
3958b146 450I<see> that the community efforts really work. So after each of his
a1f349fd
MB
451sync points, you are to 'make test' to check if everything is still
452in working order. If it is, you do 'make ok', which will send an OK
453report to perlbug@perl.org. (If you do not have access to a mailer
3e148164 454from the system you just finished successfully 'make test', you can
a1f349fd
MB
455do 'make okfile', which creates the file C<perl.ok>, which you can
456than take to your favourite mailer and mail yourself).
457
3958b146 458But of course, as always, things will not always lead to a success
a1f349fd
MB
459path, and one or more test do not pass the 'make test'. Before
460sending in a bug report (using 'make nok' or 'make nokfile'), check
461the mailing list if someone else has reported the bug already and if
462so, confirm it by replying to that message. If not, you might want to
463trace the source of that misbehaviour B<before> sending in the bug,
464which will help all the other porters in finding the solution.
465
3e148164
JH
466Here the saved patches come in very handy. You can check the list of
467patches to see which patch changed what file and what change caused
468the misbehaviour. If you note that in the bug report, it saves the
469one trying to solve it, looking for that point.
a1f349fd
MB
470
471=back
472
473If searching the patches is too bothersome, you might consider using
474perl's bugtron to find more information about discussions and
475ramblings on posted bugs.
476
3e148164
JH
477If you want to get the best of both worlds, rsync both the source
478tree for convenience, reliability and ease and rsync the patches
479for reference.
480
52315700
RF
481=back
482
fcc89a64
DR
483=head2 Working with the source
484
485Because you cannot use the Perforce client, you cannot easily generate
486diffs against the repository, nor will merges occur when you update
487via rsync. If you edit a file locally and then rsync against the
488latest source, changes made in the remote copy will I<overwrite> your
489local versions!
490
491The best way to deal with this is to maintain a tree of symlinks to
492the rsync'd source. Then, when you want to edit a file, you remove
493the symlink, copy the real file into the other tree, and edit it. You
494can then diff your edited file against the original to generate a
495patch, and you can safely update the original tree.
496
497Perl's F<Configure> script can generate this tree of symlinks for you.
498The following example assumes that you have used rsync to pull a copy
499of the Perl source into the F<perl-rsync> directory. In the directory
500above that one, you can execute the following commands:
501
502 mkdir perl-dev
503 cd perl-dev
504 ../perl-rsync/Configure -Dmksymlinks -Dusedevel -D"optimize=-g"
505
506This will start the Perl configuration process. After a few prompts,
507you should see something like this:
508
509 Symbolic links are supported.
510
511 Checking how to test for symbolic links...
512 Your builtin 'test -h' may be broken.
513 Trying external '/usr/bin/test -h'.
514 You can test for symbolic links with '/usr/bin/test -h'.
515
516 Creating the symbolic links...
517 (First creating the subdirectories...)
518 (Then creating the symlinks...)
519
520The specifics may vary based on your operating system, of course.
521After you see this, you can abort the F<Configure> script, and you
522will see that the directory you are in has a tree of symlinks to the
523F<perl-rsync> directories and files.
524
525If you plan to do a lot of work with the Perl source, here are some
526Bourne shell script functions that can make your life easier:
527
528 function edit {
529 if [ -L $1 ]; then
530 mv $1 $1.orig
531 cp $1.orig $1
532 vi $1
533 else
534 /bin/vi $1
535 fi
536 }
537
538 function unedit {
539 if [ -L $1.orig ]; then
540 rm $1
541 mv $1.orig $1
542 fi
543 }
544
545Replace "vi" with your favorite flavor of editor.
546
547Here is another function which will quickly generate a patch for the
548files which have been edited in your symlink tree:
549
550 mkpatchorig() {
551 local diffopts
552 for f in `find . -name '*.orig' | sed s,^\./,,`
553 do
554 case `echo $f | sed 's,.orig$,,;s,.*\.,,'` in
555 c) diffopts=-p ;;
556 pod) diffopts='-F^=' ;;
557 *) diffopts= ;;
558 esac
559 diff -du $diffopts $f `echo $f | sed 's,.orig$,,'`
560 done
561 }
562
563This function produces patches which include enough context to make
564your changes obvious. This makes it easier for the Perl pumpking(s)
565to review them when you send them to the perl5-porters list, and that
566means they're more likely to get applied.
567
568This function assumed a GNU diff, and may require some tweaking for
569other diff variants.
52315700 570
3fd28c4e 571=head2 Perlbug administration
52315700 572
3fd28c4e
RF
573There is a single remote administrative interface for modifying bug status,
574category, open issues etc. using the B<RT> I<bugtracker> system, maintained
575by I<Robert Spier>. Become an administrator, and close any bugs you can get
576your sticky mitts on:
52315700 577
3fd28c4e 578 http://rt.perl.org
52315700 579
3fd28c4e 580The bugtracker mechanism for B<perl5> bugs in particular is at:
52315700 581
3fd28c4e 582 http://bugs6.perl.org/perlbug
52315700 583
3fd28c4e 584To email the bug system administrators:
52315700 585
3fd28c4e 586 "perlbug-admin" <perlbug-admin@perl.org>
52315700 587
52315700 588
a1f349fd
MB
589=head2 Submitting patches
590
f7e1e956
MS
591Always submit patches to I<perl5-porters@perl.org>. If you're
592patching a core module and there's an author listed, send the author a
593copy (see L<Patching a core module>). This lets other porters review
594your patch, which catches a surprising number of errors in patches.
595Either use the diff program (available in source code form from
f224927c 596ftp://ftp.gnu.org/pub/gnu/ , or use Johan Vromans' I<makepatch>
f7e1e956
MS
597(available from I<CPAN/authors/id/JV/>). Unified diffs are preferred,
598but context diffs are accepted. Do not send RCS-style diffs or diffs
599without context lines. More information is given in the
600I<Porting/patching.pod> file in the Perl source distribution. Please
601patch against the latest B<development> version (e.g., if you're
602fixing a bug in the 5.005 track, patch against the latest 5.005_5x
603version). Only patches that survive the heat of the development
604branch get applied to maintenance versions.
605
606Your patch should update the documentation and test suite. See
607L<Writing a test>.
e8cd7eae
GS
608
609To report a bug in Perl, use the program I<perlbug> which comes with
610Perl (if you can't get Perl to work, send mail to the address
f18956b7 611I<perlbug@perl.org> or I<perlbug@perl.com>). Reporting bugs through
e8cd7eae 612I<perlbug> feeds into the automated bug-tracking system, access to
f224927c 613which is provided through the web at http://bugs.perl.org/ . It
e8cd7eae
GS
614often pays to check the archives of the perl5-porters mailing list to
615see whether the bug you're reporting has been reported before, and if
616so whether it was considered a bug. See above for the location of
617the searchable archives.
618
f224927c 619The CPAN testers ( http://testers.cpan.org/ ) are a group of
ba139f7d
JH
620volunteers who test CPAN modules on a variety of platforms. Perl
621Smokers ( http://archives.develooper.com/daily-build@perl.org/ )
622automatically tests Perl source releases on platforms with various
623configurations. Both efforts welcome volunteers.
e8cd7eae 624
e8cd7eae
GS
625It's a good idea to read and lurk for a while before chipping in.
626That way you'll get to see the dynamic of the conversations, learn the
627personalities of the players, and hopefully be better prepared to make
628a useful contribution when do you speak up.
629
630If after all this you still think you want to join the perl5-porters
f6c51b38
GS
631mailing list, send mail to I<perl5-porters-subscribe@perl.org>. To
632unsubscribe, send mail to I<perl5-porters-unsubscribe@perl.org>.
e8cd7eae 633
a422fd2d
SC
634To hack on the Perl guts, you'll need to read the following things:
635
636=over 3
637
638=item L<perlguts>
639
640This is of paramount importance, since it's the documentation of what
641goes where in the Perl source. Read it over a couple of times and it
642might start to make sense - don't worry if it doesn't yet, because the
643best way to study it is to read it in conjunction with poking at Perl
644source, and we'll do that later on.
645
646You might also want to look at Gisle Aas's illustrated perlguts -
647there's no guarantee that this will be absolutely up-to-date with the
648latest documentation in the Perl core, but the fundamentals will be
1577cd80 649right. ( http://gisle.aas.no/perl/illguts/ )
a422fd2d
SC
650
651=item L<perlxstut> and L<perlxs>
652
653A working knowledge of XSUB programming is incredibly useful for core
654hacking; XSUBs use techniques drawn from the PP code, the portion of the
655guts that actually executes a Perl program. It's a lot gentler to learn
656those techniques from simple examples and explanation than from the core
657itself.
658
659=item L<perlapi>
660
661The documentation for the Perl API explains what some of the internal
662functions do, as well as the many macros used in the source.
663
664=item F<Porting/pumpkin.pod>
665
666This is a collection of words of wisdom for a Perl porter; some of it is
667only useful to the pumpkin holder, but most of it applies to anyone
668wanting to go about Perl development.
669
670=item The perl5-porters FAQ
671
7d7d5695
RGS
672This should be available from http://simon-cozens.org/writings/p5p-faq ;
673alternatively, you can get the FAQ emailed to you by sending mail to
674C<perl5-porters-faq@perl.org>. It contains hints on reading perl5-porters,
675information on how perl5-porters works and how Perl development in general
676works.
a422fd2d
SC
677
678=back
679
680=head2 Finding Your Way Around
681
682Perl maintenance can be split into a number of areas, and certain people
683(pumpkins) will have responsibility for each area. These areas sometimes
684correspond to files or directories in the source kit. Among the areas are:
685
686=over 3
687
688=item Core modules
689
690Modules shipped as part of the Perl core live in the F<lib/> and F<ext/>
691subdirectories: F<lib/> is for the pure-Perl modules, and F<ext/>
692contains the core XS modules.
693
f7e1e956
MS
694=item Tests
695
696There are tests for nearly all the modules, built-ins and major bits
697of functionality. Test files all have a .t suffix. Module tests live
698in the F<lib/> and F<ext/> directories next to the module being
699tested. Others live in F<t/>. See L<Writing a test>
700
a422fd2d
SC
701=item Documentation
702
703Documentation maintenance includes looking after everything in the
704F<pod/> directory, (as well as contributing new documentation) and
705the documentation to the modules in core.
706
707=item Configure
708
709The configure process is the way we make Perl portable across the
710myriad of operating systems it supports. Responsibility for the
711configure, build and installation process, as well as the overall
712portability of the core code rests with the configure pumpkin - others
713help out with individual operating systems.
714
715The files involved are the operating system directories, (F<win32/>,
716F<os2/>, F<vms/> and so on) the shell scripts which generate F<config.h>
717and F<Makefile>, as well as the metaconfig files which generate
718F<Configure>. (metaconfig isn't included in the core distribution.)
719
720=item Interpreter
721
722And of course, there's the core of the Perl interpreter itself. Let's
723have a look at that in a little more detail.
724
725=back
726
727Before we leave looking at the layout, though, don't forget that
728F<MANIFEST> contains not only the file names in the Perl distribution,
729but short descriptions of what's in them, too. For an overview of the
730important files, try this:
731
732 perl -lne 'print if /^[^\/]+\.[ch]\s+/' MANIFEST
733
734=head2 Elements of the interpreter
735
736The work of the interpreter has two main stages: compiling the code
737into the internal representation, or bytecode, and then executing it.
738L<perlguts/Compiled code> explains exactly how the compilation stage
739happens.
740
741Here is a short breakdown of perl's operation:
742
743=over 3
744
745=item Startup
746
747The action begins in F<perlmain.c>. (or F<miniperlmain.c> for miniperl)
748This is very high-level code, enough to fit on a single screen, and it
749resembles the code found in L<perlembed>; most of the real action takes
750place in F<perl.c>
751
752First, F<perlmain.c> allocates some memory and constructs a Perl
753interpreter:
754
755 1 PERL_SYS_INIT3(&argc,&argv,&env);
756 2
757 3 if (!PL_do_undump) {
758 4 my_perl = perl_alloc();
759 5 if (!my_perl)
760 6 exit(1);
761 7 perl_construct(my_perl);
762 8 PL_perl_destruct_level = 0;
763 9 }
764
765Line 1 is a macro, and its definition is dependent on your operating
766system. Line 3 references C<PL_do_undump>, a global variable - all
767global variables in Perl start with C<PL_>. This tells you whether the
768current running program was created with the C<-u> flag to perl and then
769F<undump>, which means it's going to be false in any sane context.
770
771Line 4 calls a function in F<perl.c> to allocate memory for a Perl
772interpreter. It's quite a simple function, and the guts of it looks like
773this:
774
775 my_perl = (PerlInterpreter*)PerlMem_malloc(sizeof(PerlInterpreter));
776
777Here you see an example of Perl's system abstraction, which we'll see
778later: C<PerlMem_malloc> is either your system's C<malloc>, or Perl's
779own C<malloc> as defined in F<malloc.c> if you selected that option at
780configure time.
781
782Next, in line 7, we construct the interpreter; this sets up all the
783special variables that Perl needs, the stacks, and so on.
784
785Now we pass Perl the command line options, and tell it to go:
786
787 exitstatus = perl_parse(my_perl, xs_init, argc, argv, (char **)NULL);
788 if (!exitstatus) {
789 exitstatus = perl_run(my_perl);
790 }
791
792
793C<perl_parse> is actually a wrapper around C<S_parse_body>, as defined
794in F<perl.c>, which processes the command line options, sets up any
795statically linked XS modules, opens the program and calls C<yyparse> to
796parse it.
797
798=item Parsing
799
800The aim of this stage is to take the Perl source, and turn it into an op
801tree. We'll see what one of those looks like later. Strictly speaking,
802there's three things going on here.
803
804C<yyparse>, the parser, lives in F<perly.c>, although you're better off
805reading the original YACC input in F<perly.y>. (Yes, Virginia, there
806B<is> a YACC grammar for Perl!) The job of the parser is to take your
b432a672 807code and "understand" it, splitting it into sentences, deciding which
a422fd2d
SC
808operands go with which operators and so on.
809
810The parser is nobly assisted by the lexer, which chunks up your input
811into tokens, and decides what type of thing each token is: a variable
812name, an operator, a bareword, a subroutine, a core function, and so on.
813The main point of entry to the lexer is C<yylex>, and that and its
814associated routines can be found in F<toke.c>. Perl isn't much like
815other computer languages; it's highly context sensitive at times, it can
816be tricky to work out what sort of token something is, or where a token
817ends. As such, there's a lot of interplay between the tokeniser and the
818parser, which can get pretty frightening if you're not used to it.
819
820As the parser understands a Perl program, it builds up a tree of
821operations for the interpreter to perform during execution. The routines
822which construct and link together the various operations are to be found
823in F<op.c>, and will be examined later.
824
825=item Optimization
826
827Now the parsing stage is complete, and the finished tree represents
828the operations that the Perl interpreter needs to perform to execute our
829program. Next, Perl does a dry run over the tree looking for
830optimisations: constant expressions such as C<3 + 4> will be computed
831now, and the optimizer will also see if any multiple operations can be
832replaced with a single one. For instance, to fetch the variable C<$foo>,
833instead of grabbing the glob C<*foo> and looking at the scalar
834component, the optimizer fiddles the op tree to use a function which
835directly looks up the scalar in question. The main optimizer is C<peep>
836in F<op.c>, and many ops have their own optimizing functions.
837
838=item Running
839
840Now we're finally ready to go: we have compiled Perl byte code, and all
841that's left to do is run it. The actual execution is done by the
842C<runops_standard> function in F<run.c>; more specifically, it's done by
843these three innocent looking lines:
844
845 while ((PL_op = CALL_FPTR(PL_op->op_ppaddr)(aTHX))) {
846 PERL_ASYNC_CHECK();
847 }
848
849You may be more comfortable with the Perl version of that:
850
851 PERL_ASYNC_CHECK() while $Perl::op = &{$Perl::op->{function}};
852
853Well, maybe not. Anyway, each op contains a function pointer, which
854stipulates the function which will actually carry out the operation.
855This function will return the next op in the sequence - this allows for
856things like C<if> which choose the next op dynamically at run time.
857The C<PERL_ASYNC_CHECK> makes sure that things like signals interrupt
858execution if required.
859
860The actual functions called are known as PP code, and they're spread
b432a672 861between four files: F<pp_hot.c> contains the "hot" code, which is most
a422fd2d
SC
862often used and highly optimized, F<pp_sys.c> contains all the
863system-specific functions, F<pp_ctl.c> contains the functions which
864implement control structures (C<if>, C<while> and the like) and F<pp.c>
865contains everything else. These are, if you like, the C code for Perl's
866built-in functions and operators.
867
dfc98234
DM
868Note that each C<pp_> function is expected to return a pointer to the next
869op. Calls to perl subs (and eval blocks) are handled within the same
870runops loop, and do not consume extra space on the C stack. For example,
871C<pp_entersub> and C<pp_entertry> just push a C<CxSUB> or C<CxEVAL> block
872struct onto the context stack which contain the address of the op
873following the sub call or eval. They then return the first op of that sub
874or eval block, and so execution continues of that sub or block. Later, a
875C<pp_leavesub> or C<pp_leavetry> op pops the C<CxSUB> or C<CxEVAL>,
876retrieves the return op from it, and returns it.
877
878=item Exception handing
879
28a5cf3b 880Perl's exception handing (i.e. C<die> etc) is built on top of the low-level
dfc98234 881C<setjmp()>/C<longjmp()> C-library functions. These basically provide a
28a5cf3b 882way to capture the current PC and SP registers and later restore them; i.e.
dfc98234
DM
883a C<longjmp()> continues at the point in code where a previous C<setjmp()>
884was done, with anything further up on the C stack being lost. This is why
885code should always save values using C<SAVE_FOO> rather than in auto
886variables.
887
888The perl core wraps C<setjmp()> etc in the macros C<JMPENV_PUSH> and
889C<JMPENV_JUMP>. The basic rule of perl exceptions is that C<exit>, and
890C<die> (in the absence of C<eval>) perform a C<JMPENV_JUMP(2)>, while
891C<die> within C<eval> does a C<JMPENV_JUMP(3)>.
892
893At entry points to perl, such as C<perl_parse()>, C<perl_run()> and
894C<call_sv(cv, G_EVAL)> each does a C<JMPENV_PUSH>, then enter a runops
895loop or whatever, and handle possible exception returns. For a 2 return,
896final cleanup is performed, such as popping stacks and calling C<CHECK> or
897C<END> blocks. Amongst other things, this is how scope cleanup still
898occurs during an C<exit>.
899
900If a C<die> can find a C<CxEVAL> block on the context stack, then the
901stack is popped to that level and the return op in that block is assigned
902to C<PL_restartop>; then a C<JMPENV_JUMP(3)> is performed. This normally
903passes control back to the guard. In the case of C<perl_run> and
904C<call_sv>, a non-null C<PL_restartop> triggers re-entry to the runops
905loop. The is the normal way that C<die> or C<croak> is handled within an
906C<eval>.
907
908Sometimes ops are executed within an inner runops loop, such as tie, sort
909or overload code. In this case, something like
910
911 sub FETCH { eval { die } }
912
913would cause a longjmp right back to the guard in C<perl_run>, popping both
914runops loops, which is clearly incorrect. One way to avoid this is for the
915tie code to do a C<JMPENV_PUSH> before executing C<FETCH> in the inner
916runops loop, but for efficiency reasons, perl in fact just sets a flag,
917using C<CATCH_SET(TRUE)>. The C<pp_require>, C<pp_entereval> and
918C<pp_entertry> ops check this flag, and if true, they call C<docatch>,
919which does a C<JMPENV_PUSH> and starts a new runops level to execute the
920code, rather than doing it on the current loop.
921
922As a further optimisation, on exit from the eval block in the C<FETCH>,
923execution of the code following the block is still carried on in the inner
924loop. When an exception is raised, C<docatch> compares the C<JMPENV>
925level of the C<CxEVAL> with C<PL_top_env> and if they differ, just
926re-throws the exception. In this way any inner loops get popped.
927
928Here's an example.
929
930 1: eval { tie @a, 'A' };
931 2: sub A::TIEARRAY {
932 3: eval { die };
933 4: die;
934 5: }
935
936To run this code, C<perl_run> is called, which does a C<JMPENV_PUSH> then
937enters a runops loop. This loop executes the eval and tie ops on line 1,
938with the eval pushing a C<CxEVAL> onto the context stack.
939
940The C<pp_tie> does a C<CATCH_SET(TRUE)>, then starts a second runops loop
941to execute the body of C<TIEARRAY>. When it executes the entertry op on
942line 3, C<CATCH_GET> is true, so C<pp_entertry> calls C<docatch> which
943does a C<JMPENV_PUSH> and starts a third runops loop, which then executes
944the die op. At this point the C call stack looks like this:
945
946 Perl_pp_die
947 Perl_runops # third loop
948 S_docatch_body
949 S_docatch
950 Perl_pp_entertry
951 Perl_runops # second loop
952 S_call_body
953 Perl_call_sv
954 Perl_pp_tie
955 Perl_runops # first loop
956 S_run_body
957 perl_run
958 main
959
960and the context and data stacks, as shown by C<-Dstv>, look like:
961
962 STACK 0: MAIN
963 CX 0: BLOCK =>
964 CX 1: EVAL => AV() PV("A"\0)
965 retop=leave
966 STACK 1: MAGIC
967 CX 0: SUB =>
968 retop=(null)
969 CX 1: EVAL => *
970 retop=nextstate
971
972The die pops the first C<CxEVAL> off the context stack, sets
973C<PL_restartop> from it, does a C<JMPENV_JUMP(3)>, and control returns to
974the top C<docatch>. This then starts another third-level runops level,
975which executes the nextstate, pushmark and die ops on line 4. At the point
976that the second C<pp_die> is called, the C call stack looks exactly like
977that above, even though we are no longer within an inner eval; this is
978because of the optimization mentioned earlier. However, the context stack
979now looks like this, ie with the top CxEVAL popped:
980
981 STACK 0: MAIN
982 CX 0: BLOCK =>
983 CX 1: EVAL => AV() PV("A"\0)
984 retop=leave
985 STACK 1: MAGIC
986 CX 0: SUB =>
987 retop=(null)
988
989The die on line 4 pops the context stack back down to the CxEVAL, leaving
990it as:
991
992 STACK 0: MAIN
993 CX 0: BLOCK =>
994
995As usual, C<PL_restartop> is extracted from the C<CxEVAL>, and a
996C<JMPENV_JUMP(3)> done, which pops the C stack back to the docatch:
997
998 S_docatch
999 Perl_pp_entertry
1000 Perl_runops # second loop
1001 S_call_body
1002 Perl_call_sv
1003 Perl_pp_tie
1004 Perl_runops # first loop
1005 S_run_body
1006 perl_run
1007 main
1008
1009In this case, because the C<JMPENV> level recorded in the C<CxEVAL>
1010differs from the current one, C<docatch> just does a C<JMPENV_JUMP(3)>
1011and the C stack unwinds to:
1012
1013 perl_run
1014 main
1015
1016Because C<PL_restartop> is non-null, C<run_body> starts a new runops loop
1017and execution continues.
1018
a422fd2d
SC
1019=back
1020
1021=head2 Internal Variable Types
1022
1023You should by now have had a look at L<perlguts>, which tells you about
1024Perl's internal variable types: SVs, HVs, AVs and the rest. If not, do
1025that now.
1026
1027These variables are used not only to represent Perl-space variables, but
1028also any constants in the code, as well as some structures completely
1029internal to Perl. The symbol table, for instance, is an ordinary Perl
1030hash. Your code is represented by an SV as it's read into the parser;
1031any program files you call are opened via ordinary Perl filehandles, and
1032so on.
1033
1034The core L<Devel::Peek|Devel::Peek> module lets us examine SVs from a
1035Perl program. Let's see, for instance, how Perl treats the constant
1036C<"hello">.
1037
1038 % perl -MDevel::Peek -e 'Dump("hello")'
1039 1 SV = PV(0xa041450) at 0xa04ecbc
1040 2 REFCNT = 1
1041 3 FLAGS = (POK,READONLY,pPOK)
1042 4 PV = 0xa0484e0 "hello"\0
1043 5 CUR = 5
1044 6 LEN = 6
1045
1046Reading C<Devel::Peek> output takes a bit of practise, so let's go
1047through it line by line.
1048
1049Line 1 tells us we're looking at an SV which lives at C<0xa04ecbc> in
1050memory. SVs themselves are very simple structures, but they contain a
1051pointer to a more complex structure. In this case, it's a PV, a
1052structure which holds a string value, at location C<0xa041450>. Line 2
1053is the reference count; there are no other references to this data, so
1054it's 1.
1055
1056Line 3 are the flags for this SV - it's OK to use it as a PV, it's a
1057read-only SV (because it's a constant) and the data is a PV internally.
1058Next we've got the contents of the string, starting at location
1059C<0xa0484e0>.
1060
1061Line 5 gives us the current length of the string - note that this does
1062B<not> include the null terminator. Line 6 is not the length of the
1063string, but the length of the currently allocated buffer; as the string
1064grows, Perl automatically extends the available storage via a routine
1065called C<SvGROW>.
1066
1067You can get at any of these quantities from C very easily; just add
1068C<Sv> to the name of the field shown in the snippet, and you've got a
1069macro which will return the value: C<SvCUR(sv)> returns the current
1070length of the string, C<SvREFCOUNT(sv)> returns the reference count,
1071C<SvPV(sv, len)> returns the string itself with its length, and so on.
1072More macros to manipulate these properties can be found in L<perlguts>.
1073
1074Let's take an example of manipulating a PV, from C<sv_catpvn>, in F<sv.c>
1075
1076 1 void
1077 2 Perl_sv_catpvn(pTHX_ register SV *sv, register const char *ptr, register STRLEN len)
1078 3 {
1079 4 STRLEN tlen;
1080 5 char *junk;
1081
1082 6 junk = SvPV_force(sv, tlen);
1083 7 SvGROW(sv, tlen + len + 1);
1084 8 if (ptr == junk)
1085 9 ptr = SvPVX(sv);
1086 10 Move(ptr,SvPVX(sv)+tlen,len,char);
1087 11 SvCUR(sv) += len;
1088 12 *SvEND(sv) = '\0';
1089 13 (void)SvPOK_only_UTF8(sv); /* validate pointer */
1090 14 SvTAINT(sv);
1091 15 }
1092
1093This is a function which adds a string, C<ptr>, of length C<len> onto
1094the end of the PV stored in C<sv>. The first thing we do in line 6 is
1095make sure that the SV B<has> a valid PV, by calling the C<SvPV_force>
1096macro to force a PV. As a side effect, C<tlen> gets set to the current
1097value of the PV, and the PV itself is returned to C<junk>.
1098
b1866b2d 1099In line 7, we make sure that the SV will have enough room to accommodate
a422fd2d
SC
1100the old string, the new string and the null terminator. If C<LEN> isn't
1101big enough, C<SvGROW> will reallocate space for us.
1102
1103Now, if C<junk> is the same as the string we're trying to add, we can
1104grab the string directly from the SV; C<SvPVX> is the address of the PV
1105in the SV.
1106
1107Line 10 does the actual catenation: the C<Move> macro moves a chunk of
1108memory around: we move the string C<ptr> to the end of the PV - that's
1109the start of the PV plus its current length. We're moving C<len> bytes
1110of type C<char>. After doing so, we need to tell Perl we've extended the
1111string, by altering C<CUR> to reflect the new length. C<SvEND> is a
1112macro which gives us the end of the string, so that needs to be a
1113C<"\0">.
1114
1115Line 13 manipulates the flags; since we've changed the PV, any IV or NV
1116values will no longer be valid: if we have C<$a=10; $a.="6";> we don't
1e54db1a 1117want to use the old IV of 10. C<SvPOK_only_utf8> is a special UTF-8-aware
a422fd2d
SC
1118version of C<SvPOK_only>, a macro which turns off the IOK and NOK flags
1119and turns on POK. The final C<SvTAINT> is a macro which launders tainted
1120data if taint mode is turned on.
1121
1122AVs and HVs are more complicated, but SVs are by far the most common
1123variable type being thrown around. Having seen something of how we
1124manipulate these, let's go on and look at how the op tree is
1125constructed.
1126
1127=head2 Op Trees
1128
1129First, what is the op tree, anyway? The op tree is the parsed
1130representation of your program, as we saw in our section on parsing, and
1131it's the sequence of operations that Perl goes through to execute your
1132program, as we saw in L</Running>.
1133
1134An op is a fundamental operation that Perl can perform: all the built-in
1135functions and operators are ops, and there are a series of ops which
1136deal with concepts the interpreter needs internally - entering and
1137leaving a block, ending a statement, fetching a variable, and so on.
1138
1139The op tree is connected in two ways: you can imagine that there are two
1140"routes" through it, two orders in which you can traverse the tree.
1141First, parse order reflects how the parser understood the code, and
1142secondly, execution order tells perl what order to perform the
1143operations in.
1144
1145The easiest way to examine the op tree is to stop Perl after it has
1146finished parsing, and get it to dump out the tree. This is exactly what
7d7d5695
RGS
1147the compiler backends L<B::Terse|B::Terse>, L<B::Concise|B::Concise>
1148and L<B::Debug|B::Debug> do.
a422fd2d
SC
1149
1150Let's have a look at how Perl sees C<$a = $b + $c>:
1151
1152 % perl -MO=Terse -e '$a=$b+$c'
1153 1 LISTOP (0x8179888) leave
1154 2 OP (0x81798b0) enter
1155 3 COP (0x8179850) nextstate
1156 4 BINOP (0x8179828) sassign
1157 5 BINOP (0x8179800) add [1]
1158 6 UNOP (0x81796e0) null [15]
1159 7 SVOP (0x80fafe0) gvsv GV (0x80fa4cc) *b
1160 8 UNOP (0x81797e0) null [15]
1161 9 SVOP (0x8179700) gvsv GV (0x80efeb0) *c
1162 10 UNOP (0x816b4f0) null [15]
1163 11 SVOP (0x816dcf0) gvsv GV (0x80fa460) *a
1164
1165Let's start in the middle, at line 4. This is a BINOP, a binary
1166operator, which is at location C<0x8179828>. The specific operator in
1167question is C<sassign> - scalar assignment - and you can find the code
1168which implements it in the function C<pp_sassign> in F<pp_hot.c>. As a
1169binary operator, it has two children: the add operator, providing the
1170result of C<$b+$c>, is uppermost on line 5, and the left hand side is on
1171line 10.
1172
1173Line 10 is the null op: this does exactly nothing. What is that doing
1174there? If you see the null op, it's a sign that something has been
1175optimized away after parsing. As we mentioned in L</Optimization>,
1176the optimization stage sometimes converts two operations into one, for
1177example when fetching a scalar variable. When this happens, instead of
1178rewriting the op tree and cleaning up the dangling pointers, it's easier
1179just to replace the redundant operation with the null op. Originally,
1180the tree would have looked like this:
1181
1182 10 SVOP (0x816b4f0) rv2sv [15]
1183 11 SVOP (0x816dcf0) gv GV (0x80fa460) *a
1184
1185That is, fetch the C<a> entry from the main symbol table, and then look
1186at the scalar component of it: C<gvsv> (C<pp_gvsv> into F<pp_hot.c>)
1187happens to do both these things.
1188
1189The right hand side, starting at line 5 is similar to what we've just
1190seen: we have the C<add> op (C<pp_add> also in F<pp_hot.c>) add together
1191two C<gvsv>s.
1192
1193Now, what's this about?
1194
1195 1 LISTOP (0x8179888) leave
1196 2 OP (0x81798b0) enter
1197 3 COP (0x8179850) nextstate
1198
1199C<enter> and C<leave> are scoping ops, and their job is to perform any
1200housekeeping every time you enter and leave a block: lexical variables
1201are tidied up, unreferenced variables are destroyed, and so on. Every
1202program will have those first three lines: C<leave> is a list, and its
1203children are all the statements in the block. Statements are delimited
1204by C<nextstate>, so a block is a collection of C<nextstate> ops, with
1205the ops to be performed for each statement being the children of
1206C<nextstate>. C<enter> is a single op which functions as a marker.
1207
1208That's how Perl parsed the program, from top to bottom:
1209
1210 Program
1211 |
1212 Statement
1213 |
1214 =
1215 / \
1216 / \
1217 $a +
1218 / \
1219 $b $c
1220
1221However, it's impossible to B<perform> the operations in this order:
1222you have to find the values of C<$b> and C<$c> before you add them
1223together, for instance. So, the other thread that runs through the op
1224tree is the execution order: each op has a field C<op_next> which points
1225to the next op to be run, so following these pointers tells us how perl
1226executes the code. We can traverse the tree in this order using
1227the C<exec> option to C<B::Terse>:
1228
1229 % perl -MO=Terse,exec -e '$a=$b+$c'
1230 1 OP (0x8179928) enter
1231 2 COP (0x81798c8) nextstate
1232 3 SVOP (0x81796c8) gvsv GV (0x80fa4d4) *b
1233 4 SVOP (0x8179798) gvsv GV (0x80efeb0) *c
1234 5 BINOP (0x8179878) add [1]
1235 6 SVOP (0x816dd38) gvsv GV (0x80fa468) *a
1236 7 BINOP (0x81798a0) sassign
1237 8 LISTOP (0x8179900) leave
1238
1239This probably makes more sense for a human: enter a block, start a
1240statement. Get the values of C<$b> and C<$c>, and add them together.
1241Find C<$a>, and assign one to the other. Then leave.
1242
1243The way Perl builds up these op trees in the parsing process can be
1244unravelled by examining F<perly.y>, the YACC grammar. Let's take the
1245piece we need to construct the tree for C<$a = $b + $c>
1246
1247 1 term : term ASSIGNOP term
1248 2 { $$ = newASSIGNOP(OPf_STACKED, $1, $2, $3); }
1249 3 | term ADDOP term
1250 4 { $$ = newBINOP($2, 0, scalar($1), scalar($3)); }
1251
1252If you're not used to reading BNF grammars, this is how it works: You're
1253fed certain things by the tokeniser, which generally end up in upper
1254case. Here, C<ADDOP>, is provided when the tokeniser sees C<+> in your
1255code. C<ASSIGNOP> is provided when C<=> is used for assigning. These are
b432a672 1256"terminal symbols", because you can't get any simpler than them.
a422fd2d
SC
1257
1258The grammar, lines one and three of the snippet above, tells you how to
b432a672 1259build up more complex forms. These complex forms, "non-terminal symbols"
a422fd2d
SC
1260are generally placed in lower case. C<term> here is a non-terminal
1261symbol, representing a single expression.
1262
1263The grammar gives you the following rule: you can make the thing on the
1264left of the colon if you see all the things on the right in sequence.
1265This is called a "reduction", and the aim of parsing is to completely
1266reduce the input. There are several different ways you can perform a
1267reduction, separated by vertical bars: so, C<term> followed by C<=>
1268followed by C<term> makes a C<term>, and C<term> followed by C<+>
1269followed by C<term> can also make a C<term>.
1270
1271So, if you see two terms with an C<=> or C<+>, between them, you can
1272turn them into a single expression. When you do this, you execute the
1273code in the block on the next line: if you see C<=>, you'll do the code
1274in line 2. If you see C<+>, you'll do the code in line 4. It's this code
1275which contributes to the op tree.
1276
1277 | term ADDOP term
1278 { $$ = newBINOP($2, 0, scalar($1), scalar($3)); }
1279
1280What this does is creates a new binary op, and feeds it a number of
1281variables. The variables refer to the tokens: C<$1> is the first token in
1282the input, C<$2> the second, and so on - think regular expression
1283backreferences. C<$$> is the op returned from this reduction. So, we
1284call C<newBINOP> to create a new binary operator. The first parameter to
1285C<newBINOP>, a function in F<op.c>, is the op type. It's an addition
1286operator, so we want the type to be C<ADDOP>. We could specify this
1287directly, but it's right there as the second token in the input, so we
b432a672
AL
1288use C<$2>. The second parameter is the op's flags: 0 means "nothing
1289special". Then the things to add: the left and right hand side of our
a422fd2d
SC
1290expression, in scalar context.
1291
1292=head2 Stacks
1293
1294When perl executes something like C<addop>, how does it pass on its
1295results to the next op? The answer is, through the use of stacks. Perl
1296has a number of stacks to store things it's currently working on, and
1297we'll look at the three most important ones here.
1298
1299=over 3
1300
1301=item Argument stack
1302
1303Arguments are passed to PP code and returned from PP code using the
1304argument stack, C<ST>. The typical way to handle arguments is to pop
1305them off the stack, deal with them how you wish, and then push the result
1306back onto the stack. This is how, for instance, the cosine operator
1307works:
1308
1309 NV value;
1310 value = POPn;
1311 value = Perl_cos(value);
1312 XPUSHn(value);
1313
1314We'll see a more tricky example of this when we consider Perl's macros
1315below. C<POPn> gives you the NV (floating point value) of the top SV on
1316the stack: the C<$x> in C<cos($x)>. Then we compute the cosine, and push
1317the result back as an NV. The C<X> in C<XPUSHn> means that the stack
1318should be extended if necessary - it can't be necessary here, because we
1319know there's room for one more item on the stack, since we've just
1320removed one! The C<XPUSH*> macros at least guarantee safety.
1321
1322Alternatively, you can fiddle with the stack directly: C<SP> gives you
1323the first element in your portion of the stack, and C<TOP*> gives you
1324the top SV/IV/NV/etc. on the stack. So, for instance, to do unary
1325negation of an integer:
1326
1327 SETi(-TOPi);
1328
1329Just set the integer value of the top stack entry to its negation.
1330
1331Argument stack manipulation in the core is exactly the same as it is in
1332XSUBs - see L<perlxstut>, L<perlxs> and L<perlguts> for a longer
1333description of the macros used in stack manipulation.
1334
1335=item Mark stack
1336
b432a672 1337I say "your portion of the stack" above because PP code doesn't
a422fd2d
SC
1338necessarily get the whole stack to itself: if your function calls
1339another function, you'll only want to expose the arguments aimed for the
1340called function, and not (necessarily) let it get at your own data. The
b432a672 1341way we do this is to have a "virtual" bottom-of-stack, exposed to each
a422fd2d
SC
1342function. The mark stack keeps bookmarks to locations in the argument
1343stack usable by each function. For instance, when dealing with a tied
b432a672 1344variable, (internally, something with "P" magic) Perl has to call
a422fd2d
SC
1345methods for accesses to the tied variables. However, we need to separate
1346the arguments exposed to the method to the argument exposed to the
ed233832
DM
1347original function - the store or fetch or whatever it may be. Here's
1348roughly how the tied C<push> is implemented; see C<av_push> in F<av.c>:
a422fd2d
SC
1349
1350 1 PUSHMARK(SP);
1351 2 EXTEND(SP,2);
1352 3 PUSHs(SvTIED_obj((SV*)av, mg));
1353 4 PUSHs(val);
1354 5 PUTBACK;
1355 6 ENTER;
1356 7 call_method("PUSH", G_SCALAR|G_DISCARD);
1357 8 LEAVE;
13a2d996 1358
a422fd2d
SC
1359Let's examine the whole implementation, for practice:
1360
1361 1 PUSHMARK(SP);
1362
1363Push the current state of the stack pointer onto the mark stack. This is
1364so that when we've finished adding items to the argument stack, Perl
1365knows how many things we've added recently.
1366
1367 2 EXTEND(SP,2);
1368 3 PUSHs(SvTIED_obj((SV*)av, mg));
1369 4 PUSHs(val);
1370
1371We're going to add two more items onto the argument stack: when you have
1372a tied array, the C<PUSH> subroutine receives the object and the value
1373to be pushed, and that's exactly what we have here - the tied object,
1374retrieved with C<SvTIED_obj>, and the value, the SV C<val>.
1375
1376 5 PUTBACK;
1377
e89a6d4e
JD
1378Next we tell Perl to update the global stack pointer from our internal
1379variable: C<dSP> only gave us a local copy, not a reference to the global.
a422fd2d
SC
1380
1381 6 ENTER;
1382 7 call_method("PUSH", G_SCALAR|G_DISCARD);
1383 8 LEAVE;
1384
1385C<ENTER> and C<LEAVE> localise a block of code - they make sure that all
1386variables are tidied up, everything that has been localised gets
1387its previous value returned, and so on. Think of them as the C<{> and
1388C<}> of a Perl block.
1389
1390To actually do the magic method call, we have to call a subroutine in
1391Perl space: C<call_method> takes care of that, and it's described in
1392L<perlcall>. We call the C<PUSH> method in scalar context, and we're
e89a6d4e
JD
1393going to discard its return value. The call_method() function
1394removes the top element of the mark stack, so there is nothing for
1395the caller to clean up.
a422fd2d 1396
a422fd2d
SC
1397=item Save stack
1398
1399C doesn't have a concept of local scope, so perl provides one. We've
1400seen that C<ENTER> and C<LEAVE> are used as scoping braces; the save
1401stack implements the C equivalent of, for example:
1402
1403 {
1404 local $foo = 42;
1405 ...
1406 }
1407
1408See L<perlguts/Localising Changes> for how to use the save stack.
1409
1410=back
1411
1412=head2 Millions of Macros
1413
1414One thing you'll notice about the Perl source is that it's full of
1415macros. Some have called the pervasive use of macros the hardest thing
1416to understand, others find it adds to clarity. Let's take an example,
1417the code which implements the addition operator:
1418
1419 1 PP(pp_add)
1420 2 {
39644a26 1421 3 dSP; dATARGET; tryAMAGICbin(add,opASSIGN);
a422fd2d
SC
1422 4 {
1423 5 dPOPTOPnnrl_ul;
1424 6 SETn( left + right );
1425 7 RETURN;
1426 8 }
1427 9 }
1428
1429Every line here (apart from the braces, of course) contains a macro. The
1430first line sets up the function declaration as Perl expects for PP code;
1431line 3 sets up variable declarations for the argument stack and the
1432target, the return value of the operation. Finally, it tries to see if
1433the addition operation is overloaded; if so, the appropriate subroutine
1434is called.
1435
1436Line 5 is another variable declaration - all variable declarations start
1437with C<d> - which pops from the top of the argument stack two NVs (hence
1438C<nn>) and puts them into the variables C<right> and C<left>, hence the
1439C<rl>. These are the two operands to the addition operator. Next, we
1440call C<SETn> to set the NV of the return value to the result of adding
1441the two values. This done, we return - the C<RETURN> macro makes sure
1442that our return value is properly handled, and we pass the next operator
1443to run back to the main run loop.
1444
1445Most of these macros are explained in L<perlapi>, and some of the more
1446important ones are explained in L<perlxs> as well. Pay special attention
1447to L<perlguts/Background and PERL_IMPLICIT_CONTEXT> for information on
1448the C<[pad]THX_?> macros.
1449
52d59bef
JH
1450=head2 The .i Targets
1451
1452You can expand the macros in a F<foo.c> file by saying
1453
1454 make foo.i
1455
1456which will expand the macros using cpp. Don't be scared by the results.
1457
955fec6b
JH
1458=head1 SOURCE CODE STATIC ANALYSIS
1459
1460Various tools exist for analysing C source code B<statically>, as
1461opposed to B<dynamically>, that is, without executing the code.
1462It is possible to detect resource leaks, undefined behaviour, type
1463mismatches, portability problems, code paths that would cause illegal
1464memory accesses, and other similar problems by just parsing the C code
1465and looking at the resulting graph, what does it tell about the
1466execution and data flows. As a matter of fact, this is exactly
1467how C compilers know to give warnings about dubious code.
1468
1469=head2 lint, splint
1470
1471The good old C code quality inspector, C<lint>, is available in
1472several platforms, but please be aware that there are several
1473different implementations of it by different vendors, which means that
1474the flags are not identical across different platforms.
1475
1476There is a lint variant called C<splint> (Secure Programming Lint)
1477available from http://www.splint.org/ that should compile on any
1478Unix-like platform.
1479
1480There are C<lint> and <splint> targets in Makefile, but you may have
1481to diddle with the flags (see above).
1482
1483=head2 Coverity
1484
1485Coverity (http://www.coverity.com/) is a product similar to lint and
1486as a testbed for their product they periodically check several open
1487source projects, and they give out accounts to open source developers
1488to the defect databases.
1489
1490=head2 cpd (cut-and-paste detector)
1491
1492The cpd tool detects cut-and-paste coding. If one instance of the
1493cut-and-pasted code changes, all the other spots should probably be
1494changed, too. Therefore such code should probably be turned into a
1495subroutine or a macro.
1496
1497cpd (http://pmd.sourceforge.net/cpd.html) is part of the pmd project
1498(http://pmd.sourceforge.net/). pmd was originally written for static
1499analysis of Java code, but later the cpd part of it was extended to
1500parse also C and C++.
1501
1502Download the pmd-X.y.jar from the SourceForge site, and then run
1503it on source code thusly:
1504
1505 java -cp pmd-X.Y.jar net.sourceforge.pmd.cpd.CPD --minimum-tokens 100 --files /some/where/src --language c > cpd.txt
1506
1507You may run into memory limits, in which case you should use the -Xmx option:
1508
1509 java -Xmx512M ...
1510
1511=head2 gcc warnings
1512
1513Though much can be written about the inconsistency and coverage
1514problems of gcc warnings (like C<-Wall> not meaning "all the
1515warnings", or some common portability problems not being covered by
1516C<-Wall>, or C<-ansi> and C<-pedantic> both being a poorly defined
1517collection of warnings, and so forth), gcc is still a useful tool in
1518keeping our coding nose clean.
1519
1520The C<-Wall> is by default on.
1521
1522The C<-ansi> (and its sidekick, C<-pedantic>) would be nice to be
1523on always, but unfortunately they are not safe on all platforms,
1524they can for example cause fatal conflicts with the system headers
1525(Solaris being a prime example). The C<cflags> frontend selects
1526C<-ansi -pedantic> for the platforms where they are known to be safe.
1527
1528Starting from Perl 5.9.4 the following extra flags are added:
1529
1530=over 4
1531
1532=item *
1533
1534C<-Wendif-labels>
1535
1536=item *
1537
1538C<-Wextra>
1539
1540=item *
1541
1542C<-Wdeclaration-after-statement>
1543
1544=back
1545
1546The following flags would be nice to have but they would first need
1547their own Stygian stablemaster:
1548
1549=over 4
1550
1551=item *
1552
1553C<-Wpointer-arith>
1554
1555=item *
1556
1557C<-Wshadow>
1558
1559=item *
1560
1561C<-Wstrict-prototypes>
1562
1563=item *
1564
1565=back
1566
1567The C<-Wtraditional> is another example of the annoying tendency of
1568gcc to bundle a lot of warnings under one switch -- it would be
1569impossible to deploy in practice because it would complain a lot -- but
1570it does contain some warnings that would be beneficial to have available
1571on their own, such as the warning about string constants inside macros
1572containing the macro arguments: this behaved differently pre-ANSI
1573than it does in ANSI, and some C compilers are still in transition,
1574AIX being an example.
1575
1576=head2 Warnings of other C compilers
1577
1578Other C compilers (yes, there B<are> other C compilers than gcc) often
1579have their "strict ANSI" or "strict ANSI with some portability extensions"
1580modes on, like for example the Sun Workshop has its C<-Xa> mode on
1581(though implicitly), or the DEC (these days, HP...) has its C<-std1>
1582mode on.
1583
1584=head2 DEBUGGING
1585
1586You can compile a special debugging version of Perl, which allows you
1587to use the C<-D> option of Perl to tell more about what Perl is doing.
1588But sometimes there is no alternative than to dive in with a debugger,
1589either to see the stack trace of a core dump (very useful in a bug
1590report), or trying to figure out what went wrong before the core dump
1591happened, or how did we end up having wrong or unexpected results.
1592
a422fd2d
SC
1593=head2 Poking at Perl
1594
1595To really poke around with Perl, you'll probably want to build Perl for
1596debugging, like this:
1597
1598 ./Configure -d -D optimize=-g
1599 make
1600
1601C<-g> is a flag to the C compiler to have it produce debugging
955fec6b
JH
1602information which will allow us to step through a running program,
1603and to see in which C function we are at (without the debugging
1604information we might see only the numerical addresses of the functions,
1605which is not very helpful).
1606
a422fd2d
SC
1607F<Configure> will also turn on the C<DEBUGGING> compilation symbol which
1608enables all the internal debugging code in Perl. There are a whole bunch
1609of things you can debug with this: L<perlrun> lists them all, and the
1610best way to find out about them is to play about with them. The most
1611useful options are probably
1612
1613 l Context (loop) stack processing
1614 t Trace execution
1615 o Method and overloading resolution
1616 c String/numeric conversions
1617
1618Some of the functionality of the debugging code can be achieved using XS
1619modules.
13a2d996 1620
a422fd2d
SC
1621 -Dr => use re 'debug'
1622 -Dx => use O 'Debug'
1623
1624=head2 Using a source-level debugger
1625
1626If the debugging output of C<-D> doesn't help you, it's time to step
1627through perl's execution with a source-level debugger.
1628
1629=over 3
1630
1631=item *
1632
955fec6b
JH
1633We'll use C<gdb> for our examples here; the principles will apply to
1634any debugger (many vendors call their debugger C<dbx>), but check the
1635manual of the one you're using.
a422fd2d
SC
1636
1637=back
1638
1639To fire up the debugger, type
1640
1641 gdb ./perl
1642
955fec6b
JH
1643Or if you have a core dump:
1644
1645 gdb ./perl core
1646
a422fd2d
SC
1647You'll want to do that in your Perl source tree so the debugger can read
1648the source code. You should see the copyright message, followed by the
1649prompt.
1650
1651 (gdb)
1652
1653C<help> will get you into the documentation, but here are the most
1654useful commands:
1655
1656=over 3
1657
1658=item run [args]
1659
1660Run the program with the given arguments.
1661
1662=item break function_name
1663
1664=item break source.c:xxx
1665
1666Tells the debugger that we'll want to pause execution when we reach
cea6626f 1667either the named function (but see L<perlguts/Internal Functions>!) or the given
a422fd2d
SC
1668line in the named source file.
1669
1670=item step
1671
1672Steps through the program a line at a time.
1673
1674=item next
1675
1676Steps through the program a line at a time, without descending into
1677functions.
1678
1679=item continue
1680
1681Run until the next breakpoint.
1682
1683=item finish
1684
1685Run until the end of the current function, then stop again.
1686
13a2d996 1687=item 'enter'
a422fd2d
SC
1688
1689Just pressing Enter will do the most recent operation again - it's a
1690blessing when stepping through miles of source code.
1691
1692=item print
1693
1694Execute the given C code and print its results. B<WARNING>: Perl makes
52d59bef
JH
1695heavy use of macros, and F<gdb> does not necessarily support macros
1696(see later L</"gdb macro support">). You'll have to substitute them
1697yourself, or to invoke cpp on the source code files
1698(see L</"The .i Targets">)
1699So, for instance, you can't say
a422fd2d
SC
1700
1701 print SvPV_nolen(sv)
1702
1703but you have to say
1704
1705 print Perl_sv_2pv_nolen(sv)
1706
ffc145e8
RK
1707=back
1708
a422fd2d
SC
1709You may find it helpful to have a "macro dictionary", which you can
1710produce by saying C<cpp -dM perl.c | sort>. Even then, F<cpp> won't
52d59bef
JH
1711recursively apply those macros for you.
1712
1713=head2 gdb macro support
a422fd2d 1714
52d59bef 1715Recent versions of F<gdb> have fairly good macro support, but
ea031e66
RGS
1716in order to use it you'll need to compile perl with macro definitions
1717included in the debugging information. Using F<gcc> version 3.1, this
1718means configuring with C<-Doptimize=-g3>. Other compilers might use a
1719different switch (if they support debugging macros at all).
1720
a422fd2d
SC
1721=head2 Dumping Perl Data Structures
1722
1723One way to get around this macro hell is to use the dumping functions in
1724F<dump.c>; these work a little like an internal
1725L<Devel::Peek|Devel::Peek>, but they also cover OPs and other structures
1726that you can't get at from Perl. Let's take an example. We'll use the
1727C<$a = $b + $c> we used before, but give it a bit of context:
1728C<$b = "6XXXX"; $c = 2.3;>. Where's a good place to stop and poke around?
1729
1730What about C<pp_add>, the function we examined earlier to implement the
1731C<+> operator:
1732
1733 (gdb) break Perl_pp_add
1734 Breakpoint 1 at 0x46249f: file pp_hot.c, line 309.
1735
cea6626f 1736Notice we use C<Perl_pp_add> and not C<pp_add> - see L<perlguts/Internal Functions>.
a422fd2d
SC
1737With the breakpoint in place, we can run our program:
1738
1739 (gdb) run -e '$b = "6XXXX"; $c = 2.3; $a = $b + $c'
1740
1741Lots of junk will go past as gdb reads in the relevant source files and
1742libraries, and then:
1743
1744 Breakpoint 1, Perl_pp_add () at pp_hot.c:309
39644a26 1745 309 dSP; dATARGET; tryAMAGICbin(add,opASSIGN);
a422fd2d
SC
1746 (gdb) step
1747 311 dPOPTOPnnrl_ul;
1748 (gdb)
1749
1750We looked at this bit of code before, and we said that C<dPOPTOPnnrl_ul>
1751arranges for two C<NV>s to be placed into C<left> and C<right> - let's
1752slightly expand it:
1753
1754 #define dPOPTOPnnrl_ul NV right = POPn; \
1755 SV *leftsv = TOPs; \
1756 NV left = USE_LEFT(leftsv) ? SvNV(leftsv) : 0.0
1757
1758C<POPn> takes the SV from the top of the stack and obtains its NV either
1759directly (if C<SvNOK> is set) or by calling the C<sv_2nv> function.
1760C<TOPs> takes the next SV from the top of the stack - yes, C<POPn> uses
1761C<TOPs> - but doesn't remove it. We then use C<SvNV> to get the NV from
1762C<leftsv> in the same way as before - yes, C<POPn> uses C<SvNV>.
1763
1764Since we don't have an NV for C<$b>, we'll have to use C<sv_2nv> to
1765convert it. If we step again, we'll find ourselves there:
1766
1767 Perl_sv_2nv (sv=0xa0675d0) at sv.c:1669
1768 1669 if (!sv)
1769 (gdb)
1770
1771We can now use C<Perl_sv_dump> to investigate the SV:
1772
1773 SV = PV(0xa057cc0) at 0xa0675d0
1774 REFCNT = 1
1775 FLAGS = (POK,pPOK)
1776 PV = 0xa06a510 "6XXXX"\0
1777 CUR = 5
1778 LEN = 6
1779 $1 = void
1780
1781We know we're going to get C<6> from this, so let's finish the
1782subroutine:
1783
1784 (gdb) finish
1785 Run till exit from #0 Perl_sv_2nv (sv=0xa0675d0) at sv.c:1671
1786 0x462669 in Perl_pp_add () at pp_hot.c:311
1787 311 dPOPTOPnnrl_ul;
1788
1789We can also dump out this op: the current op is always stored in
1790C<PL_op>, and we can dump it with C<Perl_op_dump>. This'll give us
1791similar output to L<B::Debug|B::Debug>.
1792
1793 {
1794 13 TYPE = add ===> 14
1795 TARG = 1
1796 FLAGS = (SCALAR,KIDS)
1797 {
1798 TYPE = null ===> (12)
1799 (was rv2sv)
1800 FLAGS = (SCALAR,KIDS)
1801 {
1802 11 TYPE = gvsv ===> 12
1803 FLAGS = (SCALAR)
1804 GV = main::b
1805 }
1806 }
1807
10f58044 1808# finish this later #
a422fd2d
SC
1809
1810=head2 Patching
1811
1812All right, we've now had a look at how to navigate the Perl sources and
1813some things you'll need to know when fiddling with them. Let's now get
1814on and create a simple patch. Here's something Larry suggested: if a
1815C<U> is the first active format during a C<pack>, (for example,
1816C<pack "U3C8", @stuff>) then the resulting string should be treated as
1e54db1a 1817UTF-8 encoded.
a422fd2d
SC
1818
1819How do we prepare to fix this up? First we locate the code in question -
1820the C<pack> happens at runtime, so it's going to be in one of the F<pp>
1821files. Sure enough, C<pp_pack> is in F<pp.c>. Since we're going to be
1822altering this file, let's copy it to F<pp.c~>.
1823
a6ec74c1
JH
1824[Well, it was in F<pp.c> when this tutorial was written. It has now been
1825split off with C<pp_unpack> to its own file, F<pp_pack.c>]
1826
a422fd2d
SC
1827Now let's look over C<pp_pack>: we take a pattern into C<pat>, and then
1828loop over the pattern, taking each format character in turn into
1829C<datum_type>. Then for each possible format character, we swallow up
1830the other arguments in the pattern (a field width, an asterisk, and so
1831on) and convert the next chunk input into the specified format, adding
1832it onto the output SV C<cat>.
1833
1834How do we know if the C<U> is the first format in the C<pat>? Well, if
1835we have a pointer to the start of C<pat> then, if we see a C<U> we can
1836test whether we're still at the start of the string. So, here's where
1837C<pat> is set up:
1838
1839 STRLEN fromlen;
1840 register char *pat = SvPVx(*++MARK, fromlen);
1841 register char *patend = pat + fromlen;
1842 register I32 len;
1843 I32 datumtype;
1844 SV *fromstr;
1845
1846We'll have another string pointer in there:
1847
1848 STRLEN fromlen;
1849 register char *pat = SvPVx(*++MARK, fromlen);
1850 register char *patend = pat + fromlen;
1851 + char *patcopy;
1852 register I32 len;
1853 I32 datumtype;
1854 SV *fromstr;
1855
1856And just before we start the loop, we'll set C<patcopy> to be the start
1857of C<pat>:
1858
1859 items = SP - MARK;
1860 MARK++;
1861 sv_setpvn(cat, "", 0);
1862 + patcopy = pat;
1863 while (pat < patend) {
1864
1865Now if we see a C<U> which was at the start of the string, we turn on
1e54db1a 1866the C<UTF8> flag for the output SV, C<cat>:
a422fd2d
SC
1867
1868 + if (datumtype == 'U' && pat==patcopy+1)
1869 + SvUTF8_on(cat);
1870 if (datumtype == '#') {
1871 while (pat < patend && *pat != '\n')
1872 pat++;
1873
1874Remember that it has to be C<patcopy+1> because the first character of
1875the string is the C<U> which has been swallowed into C<datumtype!>
1876
1877Oops, we forgot one thing: what if there are spaces at the start of the
1878pattern? C<pack(" U*", @stuff)> will have C<U> as the first active
1879character, even though it's not the first thing in the pattern. In this
1880case, we have to advance C<patcopy> along with C<pat> when we see spaces:
1881
1882 if (isSPACE(datumtype))
1883 continue;
1884
1885needs to become
1886
1887 if (isSPACE(datumtype)) {
1888 patcopy++;
1889 continue;
1890 }
1891
1892OK. That's the C part done. Now we must do two additional things before
1893this patch is ready to go: we've changed the behaviour of Perl, and so
1894we must document that change. We must also provide some more regression
1895tests to make sure our patch works and doesn't create a bug somewhere
1896else along the line.
1897
b23b8711
MS
1898The regression tests for each operator live in F<t/op/>, and so we
1899make a copy of F<t/op/pack.t> to F<t/op/pack.t~>. Now we can add our
1900tests to the end. First, we'll test that the C<U> does indeed create
1901Unicode strings.
1902
1903t/op/pack.t has a sensible ok() function, but if it didn't we could
35c336e6 1904use the one from t/test.pl.
b23b8711 1905
35c336e6
MS
1906 require './test.pl';
1907 plan( tests => 159 );
b23b8711
MS
1908
1909so instead of this:
a422fd2d
SC
1910
1911 print 'not ' unless "1.20.300.4000" eq sprintf "%vd", pack("U*",1,20,300,4000);
1912 print "ok $test\n"; $test++;
1913
35c336e6
MS
1914we can write the more sensible (see L<Test::More> for a full
1915explanation of is() and other testing functions).
b23b8711 1916
35c336e6 1917 is( "1.20.300.4000", sprintf "%vd", pack("U*",1,20,300,4000),
812f5127 1918 "U* produces unicode" );
b23b8711 1919
a422fd2d
SC
1920Now we'll test that we got that space-at-the-beginning business right:
1921
35c336e6 1922 is( "1.20.300.4000", sprintf "%vd", pack(" U*",1,20,300,4000),
812f5127 1923 " with spaces at the beginning" );
a422fd2d
SC
1924
1925And finally we'll test that we don't make Unicode strings if C<U> is B<not>
1926the first active format:
1927
35c336e6 1928 isnt( v1.20.300.4000, sprintf "%vd", pack("C0U*",1,20,300,4000),
812f5127 1929 "U* not first isn't unicode" );
a422fd2d 1930
35c336e6
MS
1931Mustn't forget to change the number of tests which appears at the top,
1932or else the automated tester will get confused. This will either look
1933like this:
a422fd2d 1934
35c336e6
MS
1935 print "1..156\n";
1936
1937or this:
1938
1939 plan( tests => 156 );
a422fd2d
SC
1940
1941We now compile up Perl, and run it through the test suite. Our new
1942tests pass, hooray!
1943
1944Finally, the documentation. The job is never done until the paperwork is
1945over, so let's describe the change we've just made. The relevant place
1946is F<pod/perlfunc.pod>; again, we make a copy, and then we'll insert
1947this text in the description of C<pack>:
1948
1949 =item *
1950
1951 If the pattern begins with a C<U>, the resulting string will be treated
1e54db1a
JH
1952 as UTF-8-encoded Unicode. You can force UTF-8 encoding on in a string
1953 with an initial C<U0>, and the bytes that follow will be interpreted as
1954 Unicode characters. If you don't want this to happen, you can begin your
1955 pattern with C<C0> (or anything else) to force Perl not to UTF-8 encode your
a422fd2d
SC
1956 string, and then follow this with a C<U*> somewhere in your pattern.
1957
1958All done. Now let's create the patch. F<Porting/patching.pod> tells us
1959that if we're making major changes, we should copy the entire directory
1960to somewhere safe before we begin fiddling, and then do
13a2d996 1961
a422fd2d
SC
1962 diff -ruN old new > patch
1963
1964However, we know which files we've changed, and we can simply do this:
1965
1966 diff -u pp.c~ pp.c > patch
1967 diff -u t/op/pack.t~ t/op/pack.t >> patch
1968 diff -u pod/perlfunc.pod~ pod/perlfunc.pod >> patch
1969
1970We end up with a patch looking a little like this:
1971
1972 --- pp.c~ Fri Jun 02 04:34:10 2000
1973 +++ pp.c Fri Jun 16 11:37:25 2000
1974 @@ -4375,6 +4375,7 @@
1975 register I32 items;
1976 STRLEN fromlen;
1977 register char *pat = SvPVx(*++MARK, fromlen);
1978 + char *patcopy;
1979 register char *patend = pat + fromlen;
1980 register I32 len;
1981 I32 datumtype;
1982 @@ -4405,6 +4406,7 @@
1983 ...
1984
1985And finally, we submit it, with our rationale, to perl5-porters. Job
1986done!
1987
f7e1e956
MS
1988=head2 Patching a core module
1989
1990This works just like patching anything else, with an extra
1991consideration. Many core modules also live on CPAN. If this is so,
1992patch the CPAN version instead of the core and send the patch off to
1993the module maintainer (with a copy to p5p). This will help the module
1994maintainer keep the CPAN version in sync with the core version without
1995constantly scanning p5p.
1996
db300100
RGS
1997The list of maintainers of core modules is usefully documented in
1998F<Porting/Maintainers.pl>.
1999
acbe17fc
JP
2000=head2 Adding a new function to the core
2001
2002If, as part of a patch to fix a bug, or just because you have an
2003especially good idea, you decide to add a new function to the core,
2004discuss your ideas on p5p well before you start work. It may be that
2005someone else has already attempted to do what you are considering and
2006can give lots of good advice or even provide you with bits of code
2007that they already started (but never finished).
2008
2009You have to follow all of the advice given above for patching. It is
2010extremely important to test any addition thoroughly and add new tests
2011to explore all boundary conditions that your new function is expected
2012to handle. If your new function is used only by one module (e.g. toke),
2013then it should probably be named S_your_function (for static); on the
210b36aa 2014other hand, if you expect it to accessible from other functions in
acbe17fc
JP
2015Perl, you should name it Perl_your_function. See L<perlguts/Internal Functions>
2016for more details.
2017
2018The location of any new code is also an important consideration. Don't
2019just create a new top level .c file and put your code there; you would
2020have to make changes to Configure (so the Makefile is created properly),
2021as well as possibly lots of include files. This is strictly pumpking
2022business.
2023
2024It is better to add your function to one of the existing top level
2025source code files, but your choice is complicated by the nature of
2026the Perl distribution. Only the files that are marked as compiled
2027static are located in the perl executable. Everything else is located
2028in the shared library (or DLL if you are running under WIN32). So,
2029for example, if a function was only used by functions located in
2030toke.c, then your code can go in toke.c. If, however, you want to call
2031the function from universal.c, then you should put your code in another
2032location, for example util.c.
2033
2034In addition to writing your c-code, you will need to create an
2035appropriate entry in embed.pl describing your function, then run
2036'make regen_headers' to create the entries in the numerous header
2037files that perl needs to compile correctly. See L<perlguts/Internal Functions>
2038for information on the various options that you can set in embed.pl.
2039You will forget to do this a few (or many) times and you will get
2040warnings during the compilation phase. Make sure that you mention
2041this when you post your patch to P5P; the pumpking needs to know this.
2042
2043When you write your new code, please be conscious of existing code
884bad00 2044conventions used in the perl source files. See L<perlstyle> for
acbe17fc
JP
2045details. Although most of the guidelines discussed seem to focus on
2046Perl code, rather than c, they all apply (except when they don't ;).
2047See also I<Porting/patching.pod> file in the Perl source distribution
2048for lots of details about both formatting and submitting patches of
2049your changes.
2050
2051Lastly, TEST TEST TEST TEST TEST any code before posting to p5p.
2052Test on as many platforms as you can find. Test as many perl
2053Configure options as you can (e.g. MULTIPLICITY). If you have
2054profiling or memory tools, see L<EXTERNAL TOOLS FOR DEBUGGING PERL>
210b36aa 2055below for how to use them to further test your code. Remember that
acbe17fc
JP
2056most of the people on P5P are doing this on their own time and
2057don't have the time to debug your code.
f7e1e956
MS
2058
2059=head2 Writing a test
2060
2061Every module and built-in function has an associated test file (or
2062should...). If you add or change functionality, you have to write a
2063test. If you fix a bug, you have to write a test so that bug never
2064comes back. If you alter the docs, it would be nice to test what the
2065new documentation says.
2066
2067In short, if you submit a patch you probably also have to patch the
2068tests.
2069
2070For modules, the test file is right next to the module itself.
2071F<lib/strict.t> tests F<lib/strict.pm>. This is a recent innovation,
2072so there are some snags (and it would be wonderful for you to brush
2073them out), but it basically works that way. Everything else lives in
2074F<t/>.
2075
2076=over 3
2077
2078=item F<t/base/>
2079
2080Testing of the absolute basic functionality of Perl. Things like
2081C<if>, basic file reads and writes, simple regexes, etc. These are
2082run first in the test suite and if any of them fail, something is
2083I<really> broken.
2084
2085=item F<t/cmd/>
2086
2087These test the basic control structures, C<if/else>, C<while>,
35c336e6 2088subroutines, etc.
f7e1e956
MS
2089
2090=item F<t/comp/>
2091
2092Tests basic issues of how Perl parses and compiles itself.
2093
2094=item F<t/io/>
2095
2096Tests for built-in IO functions, including command line arguments.
2097
2098=item F<t/lib/>
2099
2100The old home for the module tests, you shouldn't put anything new in
2101here. There are still some bits and pieces hanging around in here
2102that need to be moved. Perhaps you could move them? Thanks!
2103
2104=item F<t/op/>
2105
2106Tests for perl's built in functions that don't fit into any of the
2107other directories.
2108
2109=item F<t/pod/>
2110
2111Tests for POD directives. There are still some tests for the Pod
2112modules hanging around in here that need to be moved out into F<lib/>.
2113
2114=item F<t/run/>
2115
2116Testing features of how perl actually runs, including exit codes and
2117handling of PERL* environment variables.
2118
244d9cb7
RGS
2119=item F<t/uni/>
2120
2121Tests for the core support of Unicode.
2122
2123=item F<t/win32/>
2124
2125Windows-specific tests.
2126
2127=item F<t/x2p>
2128
2129A test suite for the s2p converter.
2130
f7e1e956
MS
2131=back
2132
2133The core uses the same testing style as the rest of Perl, a simple
2134"ok/not ok" run through Test::Harness, but there are a few special
2135considerations.
2136
35c336e6
MS
2137There are three ways to write a test in the core. Test::More,
2138t/test.pl and ad hoc C<print $test ? "ok 42\n" : "not ok 42\n">. The
2139decision of which to use depends on what part of the test suite you're
2140working on. This is a measure to prevent a high-level failure (such
2141as Config.pm breaking) from causing basic functionality tests to fail.
2142
2143=over 4
2144
2145=item t/base t/comp
2146
2147Since we don't know if require works, or even subroutines, use ad hoc
2148tests for these two. Step carefully to avoid using the feature being
2149tested.
2150
2151=item t/cmd t/run t/io t/op
2152
2153Now that basic require() and subroutines are tested, you can use the
2154t/test.pl library which emulates the important features of Test::More
2155while using a minimum of core features.
2156
2157You can also conditionally use certain libraries like Config, but be
2158sure to skip the test gracefully if it's not there.
2159
2160=item t/lib ext lib
2161
2162Now that the core of Perl is tested, Test::More can be used. You can
2163also use the full suite of core modules in the tests.
2164
2165=back
f7e1e956
MS
2166
2167When you say "make test" Perl uses the F<t/TEST> program to run the
7205a85d
YO
2168test suite (except under Win32 where it uses F<t/harness> instead.)
2169All tests are run from the F<t/> directory, B<not> the directory
2170which contains the test. This causes some problems with the tests
2171in F<lib/>, so here's some opportunity for some patching.
f7e1e956
MS
2172
2173You must be triply conscious of cross-platform concerns. This usually
2174boils down to using File::Spec and avoiding things like C<fork()> and
2175C<system()> unless absolutely necessary.
2176
e018f8be
JH
2177=head2 Special Make Test Targets
2178
2179There are various special make targets that can be used to test Perl
2180slightly differently than the standard "test" target. Not all them
2181are expected to give a 100% success rate. Many of them have several
7205a85d
YO
2182aliases, and many of them are not available on certain operating
2183systems.
e018f8be
JH
2184
2185=over 4
2186
2187=item coretest
2188
7d7d5695 2189Run F<perl> on all core tests (F<t/*> and F<lib/[a-z]*> pragma tests).
e018f8be 2190
7205a85d
YO
2191(Not available on Win32)
2192
e018f8be
JH
2193=item test.deparse
2194
b26492ee
RGS
2195Run all the tests through B::Deparse. Not all tests will succeed.
2196
7205a85d
YO
2197(Not available on Win32)
2198
b26492ee
RGS
2199=item test.taintwarn
2200
2201Run all tests with the B<-t> command-line switch. Not all tests
2202are expected to succeed (until they're specifically fixed, of course).
e018f8be 2203
7205a85d
YO
2204(Not available on Win32)
2205
e018f8be
JH
2206=item minitest
2207
2208Run F<miniperl> on F<t/base>, F<t/comp>, F<t/cmd>, F<t/run>, F<t/io>,
2209F<t/op>, and F<t/uni> tests.
2210
7a834142
JH
2211=item test.valgrind check.valgrind utest.valgrind ucheck.valgrind
2212
2213(Only in Linux) Run all the tests using the memory leak + naughty
2214memory access tool "valgrind". The log files will be named
2215F<testname.valgrind>.
2216
e018f8be
JH
2217=item test.third check.third utest.third ucheck.third
2218
2219(Only in Tru64) Run all the tests using the memory leak + naughty
2220memory access tool "Third Degree". The log files will be named
60a57c1c 2221F<perl.3log.testname>.
e018f8be
JH
2222
2223=item test.torture torturetest
2224
2225Run all the usual tests and some extra tests. As of Perl 5.8.0 the
244d9cb7 2226only extra tests are Abigail's JAPHs, F<t/japh/abigail.t>.
e018f8be
JH
2227
2228You can also run the torture test with F<t/harness> by giving
2229C<-torture> argument to F<t/harness>.
2230
2231=item utest ucheck test.utf8 check.utf8
2232
2233Run all the tests with -Mutf8. Not all tests will succeed.
2234
7205a85d
YO
2235(Not available on Win32)
2236
cc0710ff
RGS
2237=item minitest.utf16 test.utf16
2238
2239Runs the tests with UTF-16 encoded scripts, encoded with different
2240versions of this encoding.
2241
2242C<make utest.utf16> runs the test suite with a combination of C<-utf8> and
2243C<-utf16> arguments to F<t/TEST>.
2244
7205a85d
YO
2245(Not available on Win32)
2246
244d9cb7
RGS
2247=item test_harness
2248
2249Run the test suite with the F<t/harness> controlling program, instead of
2250F<t/TEST>. F<t/harness> is more sophisticated, and uses the
2251L<Test::Harness> module, thus using this test target supposes that perl
2252mostly works. The main advantage for our purposes is that it prints a
00bf5cd9
RGS
2253detailed summary of failed tests at the end. Also, unlike F<t/TEST>, it
2254doesn't redirect stderr to stdout.
244d9cb7 2255
7205a85d
YO
2256Note that under Win32 F<t/harness> is always used instead of F<t/TEST>, so
2257there is no special "test_harness" target.
2258
2259Under Win32's "test" target you may use the TEST_SWITCHES and TEST_FILES
2260environment variables to control the behaviour of F<t/harness>. This means
2261you can say
2262
2263 nmake test TEST_FILES="op/*.t"
2264 nmake test TEST_SWITCHES="-torture" TEST_FILES="op/*.t"
2265
2266=item test-notty test_notty
2267
2268Sets PERL_SKIP_TTY_TEST to true before running normal test.
2269
244d9cb7
RGS
2270=back
2271
2272=head2 Running tests by hand
2273
2274You can run part of the test suite by hand by using one the following
2275commands from the F<t/> directory :
2276
2277 ./perl -I../lib TEST list-of-.t-files
2278
2279or
2280
2281 ./perl -I../lib harness list-of-.t-files
2282
2283(if you don't specify test scripts, the whole test suite will be run.)
2284
7205a85d
YO
2285=head3 Using t/harness for testing
2286
2287If you use C<harness> for testing you have several command line options
2288available to you. The arguments are as follows, and are in the order
2289that they must appear if used together.
2290
2291 harness -v -torture -re=pattern LIST OF FILES TO TEST
2292 harness -v -torture -re LIST OF PATTERNS TO MATCH
2293
2294If C<LIST OF FILES TO TEST> is omitted the file list is obtained from
2295the manifest. The file list may include shell wildcards which will be
2296expanded out.
2297
2298=over 4
2299
2300=item -v
2301
2302Run the tests under verbose mode so you can see what tests were run,
2303and debug outbut.
2304
2305=item -torture
2306
2307Run the torture tests as well as the normal set.
2308
2309=item -re=PATTERN
2310
2311Filter the file list so that all the test files run match PATTERN.
2312Note that this form is distinct from the B<-re LIST OF PATTERNS> form below
2313in that it allows the file list to be provided as well.
2314
2315=item -re LIST OF PATTERNS
2316
2317Filter the file list so that all the test files run match
2318/(LIST|OF|PATTERNS)/. Note that with this form the patterns
2319are joined by '|' and you cannot supply a list of files, instead
2320the test files are obtained from the MANIFEST.
2321
2322=back
2323
244d9cb7
RGS
2324You can run an individual test by a command similar to
2325
2326 ./perl -I../lib patho/to/foo.t
2327
2328except that the harnesses set up some environment variables that may
2329affect the execution of the test :
2330
2331=over 4
2332
2333=item PERL_CORE=1
2334
2335indicates that we're running this test part of the perl core test suite.
2336This is useful for modules that have a dual life on CPAN.
2337
2338=item PERL_DESTRUCT_LEVEL=2
2339
2340is set to 2 if it isn't set already (see L</PERL_DESTRUCT_LEVEL>)
2341
2342=item PERL
2343
2344(used only by F<t/TEST>) if set, overrides the path to the perl executable
2345that should be used to run the tests (the default being F<./perl>).
2346
2347=item PERL_SKIP_TTY_TEST
2348
2349if set, tells to skip the tests that need a terminal. It's actually set
2350automatically by the Makefile, but can also be forced artificially by
2351running 'make test_notty'.
2352
e018f8be 2353=back
f7e1e956 2354
902b9dbf
MF
2355=head1 EXTERNAL TOOLS FOR DEBUGGING PERL
2356
2357Sometimes it helps to use external tools while debugging and
2358testing Perl. This section tries to guide you through using
2359some common testing and debugging tools with Perl. This is
2360meant as a guide to interfacing these tools with Perl, not
2361as any kind of guide to the use of the tools themselves.
2362
a958818a
JH
2363B<NOTE 1>: Running under memory debuggers such as Purify, valgrind, or
2364Third Degree greatly slows down the execution: seconds become minutes,
2365minutes become hours. For example as of Perl 5.8.1, the
2366ext/Encode/t/Unicode.t takes extraordinarily long to complete under
2367e.g. Purify, Third Degree, and valgrind. Under valgrind it takes more
2368than six hours, even on a snappy computer-- the said test must be
2369doing something that is quite unfriendly for memory debuggers. If you
2370don't feel like waiting, that you can simply kill away the perl
2371process.
2372
2373B<NOTE 2>: To minimize the number of memory leak false alarms (see
2374L</PERL_DESTRUCT_LEVEL> for more information), you have to have
2375environment variable PERL_DESTRUCT_LEVEL set to 2. The F<TEST>
2376and harness scripts do that automatically. But if you are running
2377some of the tests manually-- for csh-like shells:
2378
2379 setenv PERL_DESTRUCT_LEVEL 2
2380
2381and for Bourne-type shells:
2382
2383 PERL_DESTRUCT_LEVEL=2
2384 export PERL_DESTRUCT_LEVEL
2385
2386or in UNIXy environments you can also use the C<env> command:
2387
2388 env PERL_DESTRUCT_LEVEL=2 valgrind ./perl -Ilib ...
a1b65709 2389
37c0adeb
JH
2390B<NOTE 3>: There are known memory leaks when there are compile-time
2391errors within eval or require, seeing C<S_doeval> in the call stack
2392is a good sign of these. Fixing these leaks is non-trivial,
2393unfortunately, but they must be fixed eventually.
2394
902b9dbf
MF
2395=head2 Rational Software's Purify
2396
2397Purify is a commercial tool that is helpful in identifying
2398memory overruns, wild pointers, memory leaks and other such
2399badness. Perl must be compiled in a specific way for
2400optimal testing with Purify. Purify is available under
2401Windows NT, Solaris, HP-UX, SGI, and Siemens Unix.
2402
902b9dbf
MF
2403=head2 Purify on Unix
2404
2405On Unix, Purify creates a new Perl binary. To get the most
2406benefit out of Purify, you should create the perl to Purify
2407using:
2408
2409 sh Configure -Accflags=-DPURIFY -Doptimize='-g' \
2410 -Uusemymalloc -Dusemultiplicity
2411
2412where these arguments mean:
2413
2414=over 4
2415
2416=item -Accflags=-DPURIFY
2417
2418Disables Perl's arena memory allocation functions, as well as
2419forcing use of memory allocation functions derived from the
2420system malloc.
2421
2422=item -Doptimize='-g'
2423
2424Adds debugging information so that you see the exact source
2425statements where the problem occurs. Without this flag, all
2426you will see is the source filename of where the error occurred.
2427
2428=item -Uusemymalloc
2429
2430Disable Perl's malloc so that Purify can more closely monitor
2431allocations and leaks. Using Perl's malloc will make Purify
2432report most leaks in the "potential" leaks category.
2433
2434=item -Dusemultiplicity
2435
2436Enabling the multiplicity option allows perl to clean up
2437thoroughly when the interpreter shuts down, which reduces the
2438number of bogus leak reports from Purify.
2439
2440=back
2441
2442Once you've compiled a perl suitable for Purify'ing, then you
2443can just:
2444
2445 make pureperl
2446
2447which creates a binary named 'pureperl' that has been Purify'ed.
2448This binary is used in place of the standard 'perl' binary
2449when you want to debug Perl memory problems.
2450
2451As an example, to show any memory leaks produced during the
2452standard Perl testset you would create and run the Purify'ed
2453perl as:
2454
2455 make pureperl
2456 cd t
2457 ../pureperl -I../lib harness
2458
2459which would run Perl on test.pl and report any memory problems.
2460
2461Purify outputs messages in "Viewer" windows by default. If
2462you don't have a windowing environment or if you simply
2463want the Purify output to unobtrusively go to a log file
2464instead of to the interactive window, use these following
2465options to output to the log file "perl.log":
2466
2467 setenv PURIFYOPTIONS "-chain-length=25 -windows=no \
2468 -log-file=perl.log -append-logfile=yes"
2469
2470If you plan to use the "Viewer" windows, then you only need this option:
2471
2472 setenv PURIFYOPTIONS "-chain-length=25"
2473
c406981e
JH
2474In Bourne-type shells:
2475
98631ff8
JL
2476 PURIFYOPTIONS="..."
2477 export PURIFYOPTIONS
c406981e
JH
2478
2479or if you have the "env" utility:
2480
98631ff8 2481 env PURIFYOPTIONS="..." ../pureperl ...
c406981e 2482
902b9dbf
MF
2483=head2 Purify on NT
2484
2485Purify on Windows NT instruments the Perl binary 'perl.exe'
2486on the fly. There are several options in the makefile you
2487should change to get the most use out of Purify:
2488
2489=over 4
2490
2491=item DEFINES
2492
2493You should add -DPURIFY to the DEFINES line so the DEFINES
2494line looks something like:
2495
2496 DEFINES = -DWIN32 -D_CONSOLE -DNO_STRICT $(CRYPT_FLAG) -DPURIFY=1
2497
2498to disable Perl's arena memory allocation functions, as
2499well as to force use of memory allocation functions derived
2500from the system malloc.
2501
2502=item USE_MULTI = define
2503
2504Enabling the multiplicity option allows perl to clean up
2505thoroughly when the interpreter shuts down, which reduces the
2506number of bogus leak reports from Purify.
2507
2508=item #PERL_MALLOC = define
2509
2510Disable Perl's malloc so that Purify can more closely monitor
2511allocations and leaks. Using Perl's malloc will make Purify
2512report most leaks in the "potential" leaks category.
2513
2514=item CFG = Debug
2515
2516Adds debugging information so that you see the exact source
2517statements where the problem occurs. Without this flag, all
2518you will see is the source filename of where the error occurred.
2519
2520=back
2521
2522As an example, to show any memory leaks produced during the
2523standard Perl testset you would create and run Purify as:
2524
2525 cd win32
2526 make
2527 cd ../t
2528 purify ../perl -I../lib harness
2529
2530which would instrument Perl in memory, run Perl on test.pl,
2531then finally report any memory problems.
2532
7a834142
JH
2533=head2 valgrind
2534
2535The excellent valgrind tool can be used to find out both memory leaks
2536and illegal memory accesses. As of August 2003 it unfortunately works
2537only on x86 (ELF) Linux. The special "test.valgrind" target can be used
d44161bf
MHM
2538to run the tests under valgrind. Found errors and memory leaks are
2539logged in files named F<test.valgrind>.
2540
2541As system libraries (most notably glibc) are also triggering errors,
2542valgrind allows to suppress such errors using suppression files. The
2543default suppression file that comes with valgrind already catches a lot
2544of them. Some additional suppressions are defined in F<t/perl.supp>.
7a834142
JH
2545
2546To get valgrind and for more information see
2547
2548 http://developer.kde.org/~sewardj/
2549
f134cc4e 2550=head2 Compaq's/Digital's/HP's Third Degree
09187cb1
JH
2551
2552Third Degree is a tool for memory leak detection and memory access checks.
2553It is one of the many tools in the ATOM toolkit. The toolkit is only
2554available on Tru64 (formerly known as Digital UNIX formerly known as
2555DEC OSF/1).
2556
2557When building Perl, you must first run Configure with -Doptimize=-g
2558and -Uusemymalloc flags, after that you can use the make targets
51a35ef1
JH
2559"perl.third" and "test.third". (What is required is that Perl must be
2560compiled using the C<-g> flag, you may need to re-Configure.)
09187cb1 2561
64cea5fd 2562The short story is that with "atom" you can instrument the Perl
83f0ef60 2563executable to create a new executable called F<perl.third>. When the
4ae3d70a 2564instrumented executable is run, it creates a log of dubious memory
83f0ef60 2565traffic in file called F<perl.3log>. See the manual pages of atom and
4ae3d70a
JH
2566third for more information. The most extensive Third Degree
2567documentation is available in the Compaq "Tru64 UNIX Programmer's
2568Guide", chapter "Debugging Programs with Third Degree".
64cea5fd 2569
9c54ecba 2570The "test.third" leaves a lot of files named F<foo_bar.3log> in the t/
64cea5fd
JH
2571subdirectory. There is a problem with these files: Third Degree is so
2572effective that it finds problems also in the system libraries.
9c54ecba
JH
2573Therefore you should used the Porting/thirdclean script to cleanup
2574the F<*.3log> files.
64cea5fd
JH
2575
2576There are also leaks that for given certain definition of a leak,
2577aren't. See L</PERL_DESTRUCT_LEVEL> for more information.
2578
2579=head2 PERL_DESTRUCT_LEVEL
2580
a958818a
JH
2581If you want to run any of the tests yourself manually using e.g.
2582valgrind, or the pureperl or perl.third executables, please note that
2583by default perl B<does not> explicitly cleanup all the memory it has
2584allocated (such as global memory arenas) but instead lets the exit()
2585of the whole program "take care" of such allocations, also known as
2586"global destruction of objects".
64cea5fd
JH
2587
2588There is a way to tell perl to do complete cleanup: set the
2589environment variable PERL_DESTRUCT_LEVEL to a non-zero value.
2590The t/TEST wrapper does set this to 2, and this is what you
2591need to do too, if you don't want to see the "global leaks":
1f56d61a 2592For example, for "third-degreed" Perl:
64cea5fd 2593
1f56d61a 2594 env PERL_DESTRUCT_LEVEL=2 ./perl.third -Ilib t/foo/bar.t
09187cb1 2595
414f2397
RGS
2596(Note: the mod_perl apache module uses also this environment variable
2597for its own purposes and extended its semantics. Refer to the mod_perl
287a822c
RGS
2598documentation for more information. Also, spawned threads do the
2599equivalent of setting this variable to the value 1.)
5a6c59ef
DM
2600
2601If, at the end of a run you get the message I<N scalars leaked>, you can
fd0854ff
DM
2602recompile with C<-DDEBUG_LEAKING_SCALARS>, which will cause the addresses
2603of all those leaked SVs to be dumped along with details as to where each
2604SV was originally allocated. This information is also displayed by
2605Devel::Peek. Note that the extra details recorded with each SV increases
2606memory usage, so it shouldn't be used in production environments. It also
2607converts C<new_SV()> from a macro into a real function, so you can use
2608your favourite debugger to discover where those pesky SVs were allocated.
414f2397 2609
46c6c7e2
JH
2610=head2 PERL_MEM_LOG
2611
9f653bb5 2612If compiled with C<-DPERL_MEM_LOG>, all Newx() and Renew() allocations
46c6c7e2
JH
2613and Safefree() in the Perl core go through logging functions, which is
2614handy for breakpoint setting. If also compiled with C<-DPERL_MEM_LOG_STDERR>,
e352bcff
JH
2615the allocations and frees are logged to STDERR (or more precisely, to the
2616file descriptor 2) in these logging functions, with the calling source code
2617file and line number (and C function name, if supported by the C compiler).
2618
2619This logging is somewhat similar to C<-Dm> but independent of C<-DDEBUGGING>,
2620and at a higher level (the C<-Dm> is directly at the point of C<malloc()>,
a015a77e 2621while the C<PERL_MEM_LOG> is at the level of C<New()>).
46c6c7e2 2622
51a35ef1
JH
2623=head2 Profiling
2624
2625Depending on your platform there are various of profiling Perl.
2626
2627There are two commonly used techniques of profiling executables:
10f58044 2628I<statistical time-sampling> and I<basic-block counting>.
51a35ef1
JH
2629
2630The first method takes periodically samples of the CPU program
2631counter, and since the program counter can be correlated with the code
2632generated for functions, we get a statistical view of in which
2633functions the program is spending its time. The caveats are that very
2634small/fast functions have lower probability of showing up in the
2635profile, and that periodically interrupting the program (this is
2636usually done rather frequently, in the scale of milliseconds) imposes
2637an additional overhead that may skew the results. The first problem
2638can be alleviated by running the code for longer (in general this is a
2639good idea for profiling), the second problem is usually kept in guard
2640by the profiling tools themselves.
2641
10f58044 2642The second method divides up the generated code into I<basic blocks>.
51a35ef1
JH
2643Basic blocks are sections of code that are entered only in the
2644beginning and exited only at the end. For example, a conditional jump
2645starts a basic block. Basic block profiling usually works by
10f58044 2646I<instrumenting> the code by adding I<enter basic block #nnnn>
51a35ef1
JH
2647book-keeping code to the generated code. During the execution of the
2648code the basic block counters are then updated appropriately. The
2649caveat is that the added extra code can skew the results: again, the
2650profiling tools usually try to factor their own effects out of the
2651results.
2652
83f0ef60
JH
2653=head2 Gprof Profiling
2654
51a35ef1
JH
2655gprof is a profiling tool available in many UNIX platforms,
2656it uses F<statistical time-sampling>.
83f0ef60
JH
2657
2658You can build a profiled version of perl called "perl.gprof" by
51a35ef1
JH
2659invoking the make target "perl.gprof" (What is required is that Perl
2660must be compiled using the C<-pg> flag, you may need to re-Configure).
2661Running the profiled version of Perl will create an output file called
2662F<gmon.out> is created which contains the profiling data collected
2663during the execution.
83f0ef60
JH
2664
2665The gprof tool can then display the collected data in various ways.
2666Usually gprof understands the following options:
2667
2668=over 4
2669
2670=item -a
2671
2672Suppress statically defined functions from the profile.
2673
2674=item -b
2675
2676Suppress the verbose descriptions in the profile.
2677
2678=item -e routine
2679
2680Exclude the given routine and its descendants from the profile.
2681
2682=item -f routine
2683
2684Display only the given routine and its descendants in the profile.
2685
2686=item -s
2687
2688Generate a summary file called F<gmon.sum> which then may be given
2689to subsequent gprof runs to accumulate data over several runs.
2690
2691=item -z
2692
2693Display routines that have zero usage.
2694
2695=back
2696
2697For more detailed explanation of the available commands and output
2698formats, see your own local documentation of gprof.
2699
51a35ef1
JH
2700=head2 GCC gcov Profiling
2701
10f58044 2702Starting from GCC 3.0 I<basic block profiling> is officially available
51a35ef1
JH
2703for the GNU CC.
2704
2705You can build a profiled version of perl called F<perl.gcov> by
2706invoking the make target "perl.gcov" (what is required that Perl must
2707be compiled using gcc with the flags C<-fprofile-arcs
2708-ftest-coverage>, you may need to re-Configure).
2709
2710Running the profiled version of Perl will cause profile output to be
2711generated. For each source file an accompanying ".da" file will be
2712created.
2713
2714To display the results you use the "gcov" utility (which should
2715be installed if you have gcc 3.0 or newer installed). F<gcov> is
2716run on source code files, like this
2717
2718 gcov sv.c
2719
2720which will cause F<sv.c.gcov> to be created. The F<.gcov> files
2721contain the source code annotated with relative frequencies of
2722execution indicated by "#" markers.
2723
2724Useful options of F<gcov> include C<-b> which will summarise the
2725basic block, branch, and function call coverage, and C<-c> which
2726instead of relative frequencies will use the actual counts. For
2727more information on the use of F<gcov> and basic block profiling
2728with gcc, see the latest GNU CC manual, as of GCC 3.0 see
2729
2730 http://gcc.gnu.org/onlinedocs/gcc-3.0/gcc.html
2731
2732and its section titled "8. gcov: a Test Coverage Program"
2733
2734 http://gcc.gnu.org/onlinedocs/gcc-3.0/gcc_8.html#SEC132
2735
4ae3d70a
JH
2736=head2 Pixie Profiling
2737
51a35ef1
JH
2738Pixie is a profiling tool available on IRIX and Tru64 (aka Digital
2739UNIX aka DEC OSF/1) platforms. Pixie does its profiling using
10f58044 2740I<basic-block counting>.
4ae3d70a 2741
83f0ef60 2742You can build a profiled version of perl called F<perl.pixie> by
51a35ef1
JH
2743invoking the make target "perl.pixie" (what is required is that Perl
2744must be compiled using the C<-g> flag, you may need to re-Configure).
2745
2746In Tru64 a file called F<perl.Addrs> will also be silently created,
2747this file contains the addresses of the basic blocks. Running the
2748profiled version of Perl will create a new file called "perl.Counts"
2749which contains the counts for the basic block for that particular
2750program execution.
4ae3d70a 2751
51a35ef1 2752To display the results you use the F<prof> utility. The exact
4ae3d70a
JH
2753incantation depends on your operating system, "prof perl.Counts" in
2754IRIX, and "prof -pixie -all -L. perl" in Tru64.
2755
6c41479b
JH
2756In IRIX the following prof options are available:
2757
2758=over 4
2759
2760=item -h
2761
2762Reports the most heavily used lines in descending order of use.
6e36760b 2763Useful for finding the hotspot lines.
6c41479b
JH
2764
2765=item -l
2766
2767Groups lines by procedure, with procedures sorted in descending order of use.
2768Within a procedure, lines are listed in source order.
6e36760b 2769Useful for finding the hotspots of procedures.
6c41479b
JH
2770
2771=back
2772
2773In Tru64 the following options are available:
2774
2775=over 4
2776
3958b146 2777=item -p[rocedures]
6c41479b 2778
3958b146 2779Procedures sorted in descending order by the number of cycles executed
6e36760b 2780in each procedure. Useful for finding the hotspot procedures.
6c41479b
JH
2781(This is the default option.)
2782
24000d2f 2783=item -h[eavy]
6c41479b 2784
6e36760b
JH
2785Lines sorted in descending order by the number of cycles executed in
2786each line. Useful for finding the hotspot lines.
6c41479b 2787
24000d2f 2788=item -i[nvocations]
6c41479b 2789
6e36760b
JH
2790The called procedures are sorted in descending order by number of calls
2791made to the procedures. Useful for finding the most used procedures.
6c41479b 2792
24000d2f 2793=item -l[ines]
6c41479b
JH
2794
2795Grouped by procedure, sorted by cycles executed per procedure.
6e36760b 2796Useful for finding the hotspots of procedures.
6c41479b
JH
2797
2798=item -testcoverage
2799
2800The compiler emitted code for these lines, but the code was unexecuted.
2801
24000d2f 2802=item -z[ero]
6c41479b
JH
2803
2804Unexecuted procedures.
2805
aa500c9e 2806=back
6c41479b
JH
2807
2808For further information, see your system's manual pages for pixie and prof.
4ae3d70a 2809
b8ddf6b3
SB
2810=head2 Miscellaneous tricks
2811
2812=over 4
2813
2814=item *
2815
cc177e1a 2816Those debugging perl with the DDD frontend over gdb may find the
b8ddf6b3
SB
2817following useful:
2818
2819You can extend the data conversion shortcuts menu, so for example you
2820can display an SV's IV value with one click, without doing any typing.
2821To do that simply edit ~/.ddd/init file and add after:
2822
2823 ! Display shortcuts.
2824 Ddd*gdbDisplayShortcuts: \
2825 /t () // Convert to Bin\n\
2826 /d () // Convert to Dec\n\
2827 /x () // Convert to Hex\n\
2828 /o () // Convert to Oct(\n\
2829
2830the following two lines:
2831
2832 ((XPV*) (())->sv_any )->xpv_pv // 2pvx\n\
2833 ((XPVIV*) (())->sv_any )->xiv_iv // 2ivx
2834
2835so now you can do ivx and pvx lookups or you can plug there the
2836sv_peek "conversion":
2837
2838 Perl_sv_peek(my_perl, (SV*)()) // sv_peek
2839
2840(The my_perl is for threaded builds.)
2841Just remember that every line, but the last one, should end with \n\
2842
2843Alternatively edit the init file interactively via:
28443rd mouse button -> New Display -> Edit Menu
2845
2846Note: you can define up to 20 conversion shortcuts in the gdb
2847section.
2848
9965345d
JH
2849=item *
2850
7e337ee0
JH
2851If you see in a debugger a memory area mysteriously full of 0xABABABAB
2852or 0xEFEFEFEF, you may be seeing the effect of the Poison() macros,
2853see L<perlclib>.
9965345d 2854
b8ddf6b3
SB
2855=back
2856
955fec6b 2857=head1 CONCLUSION
a422fd2d 2858
955fec6b
JH
2859We've had a brief look around the Perl source, how to maintain quality
2860of the source code, an overview of the stages F<perl> goes through
2861when it's running your code, how to use debuggers to poke at the Perl
2862guts, and finally how to analyse the execution of Perl. We took a very
2863simple problem and demonstrated how to solve it fully - with
2864documentation, regression tests, and finally a patch for submission to
2865p5p. Finally, we talked about how to use external tools to debug and
2866test Perl.
a422fd2d
SC
2867
2868I'd now suggest you read over those references again, and then, as soon
2869as possible, get your hands dirty. The best way to learn is by doing,
2870so:
2871
2872=over 3
2873
2874=item *
2875
2876Subscribe to perl5-porters, follow the patches and try and understand
2877them; don't be afraid to ask if there's a portion you're not clear on -
2878who knows, you may unearth a bug in the patch...
2879
2880=item *
2881
2882Keep up to date with the bleeding edge Perl distributions and get
2883familiar with the changes. Try and get an idea of what areas people are
2884working on and the changes they're making.
2885
2886=item *
2887
3e148164 2888Do read the README associated with your operating system, e.g. README.aix
a1f349fd
MB
2889on the IBM AIX OS. Don't hesitate to supply patches to that README if
2890you find anything missing or changed over a new OS release.
2891
2892=item *
2893
a422fd2d
SC
2894Find an area of Perl that seems interesting to you, and see if you can
2895work out how it works. Scan through the source, and step over it in the
2896debugger. Play, poke, investigate, fiddle! You'll probably get to
2897understand not just your chosen area but a much wider range of F<perl>'s
2898activity as well, and probably sooner than you'd think.
2899
2900=back
2901
2902=over 3
2903
2904=item I<The Road goes ever on and on, down from the door where it began.>
2905
2906=back
2907
2908If you can do these things, you've started on the long road to Perl porting.
2909Thanks for wanting to help make Perl better - and happy hacking!
2910
e8cd7eae
GS
2911=head1 AUTHOR
2912
2913This document was written by Nathan Torkington, and is maintained by
2914the perl5-porters mailing list.
2915