Commit | Line | Data |
---|---|---|
e8cd7eae GS |
1 | =head1 NAME |
2 | ||
3 | perlhack - How to hack at the Perl internals | |
4 | ||
5 | =head1 DESCRIPTION | |
6 | ||
7 | This document attempts to explain how Perl development takes place, | |
8 | and ends with some suggestions for people wanting to become bona fide | |
9 | porters. | |
10 | ||
11 | The perl5-porters mailing list is where the Perl standard distribution | |
12 | is maintained and developed. The list can get anywhere from 10 to 150 | |
13 | messages a day, depending on the heatedness of the debate. Most days | |
14 | there are two or three patches, extensions, features, or bugs being | |
15 | discussed at a time. | |
16 | ||
f8e3975a | 17 | A 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 |
21 | or |
22 | ||
23 | http://archive.develooper.com/perl5-porters@perl.org/ | |
24 | ||
e8cd7eae GS |
25 | List subscribers (the porters themselves) come in several flavours. |
26 | Some are quiet curious lurkers, who rarely pitch in and instead watch | |
27 | the ongoing development to ensure they're forewarned of new changes or | |
28 | features in Perl. Some are representatives of vendors, who are there | |
29 | to make sure that Perl continues to compile and work on their | |
30 | platforms. Some patch any reported bug that they know how to fix, | |
31 | some are actively patching their pet area (threads, Win32, the regexp | |
32 | engine), while others seem to do nothing but complain. In other | |
33 | words, it's your usual mix of technical people. | |
34 | ||
35 | Over this group of porters presides Larry Wall. He has the final word | |
f6c51b38 | 36 | in what does and does not change in the Perl language. Various |
b432a672 AL |
37 | releases of Perl are shepherded by a "pumpking", a porter |
38 | responsible for gathering patches, deciding on a patch-by-patch, | |
f6c51b38 | 39 | feature-by-feature basis what will and will not go into the release. |
caf100c0 | 40 | For instance, Gurusamy Sarathy was the pumpking for the 5.6 release of |
961f29c6 | 41 | Perl, and Jarkko Hietaniemi was the pumpking for the 5.8 release, and |
1a88dbf8 | 42 | Rafael Garcia-Suarez holds the pumpking crown for the 5.10 release. |
e8cd7eae GS |
43 | |
44 | In addition, various people are pumpkings for different things. For | |
961f29c6 MB |
45 | instance, Andy Dougherty and Jarkko Hietaniemi did a grand job as the |
46 | I<Configure> pumpkin up till the 5.8 release. For the 5.10 release | |
47 | H.Merijn Brand took over. | |
e8cd7eae GS |
48 | |
49 | Larry sees Perl development along the lines of the US government: | |
50 | there's the Legislature (the porters), the Executive branch (the | |
51 | pumpkings), and the Supreme Court (Larry). The legislature can | |
52 | discuss and submit patches to the executive branch all they like, but | |
53 | the executive branch is free to veto them. Rarely, the Supreme Court | |
54 | will side with the executive branch over the legislature, or the | |
55 | legislature over the executive branch. Mostly, however, the | |
56 | legislature and the executive branch are supposed to get along and | |
57 | work out their differences without impeachment or court cases. | |
58 | ||
59 | You might sometimes see reference to Rule 1 and Rule 2. Larry's power | |
60 | as Supreme Court is expressed in The Rules: | |
61 | ||
62 | =over 4 | |
63 | ||
64 | =item 1 | |
65 | ||
66 | Larry is always by definition right about how Perl should behave. | |
67 | This means he has final veto power on the core functionality. | |
68 | ||
69 | =item 2 | |
70 | ||
71 | Larry is allowed to change his mind about any matter at a later date, | |
72 | regardless of whether he previously invoked Rule 1. | |
73 | ||
74 | =back | |
75 | ||
76 | Got that? Larry is always right, even when he was wrong. It's rare | |
77 | to see either Rule exercised, but they are often alluded to. | |
78 | ||
79 | New features and extensions to the language are contentious, because | |
80 | the criteria used by the pumpkings, Larry, and other porters to decide | |
81 | which features should be implemented and incorporated are not codified | |
82 | in a few small design goals as with some other languages. Instead, | |
83 | the heuristics are flexible and often difficult to fathom. Here is | |
84 | one person's list, roughly in decreasing order of importance, of | |
85 | heuristics that new features have to be weighed against: | |
86 | ||
87 | =over 4 | |
88 | ||
89 | =item Does concept match the general goals of Perl? | |
90 | ||
91 | These haven't been written anywhere in stone, but one approximation | |
92 | is: | |
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 | ||
102 | All the talk in the world is useless without an implementation. In | |
103 | almost every case, the person or people who argue for a new feature | |
104 | will be expected to be the ones who implement it. Porters capable | |
105 | of coding new features have their own agendas, and are not available | |
106 | to implement your (possibly good) idea. | |
107 | ||
108 | =item Backwards compatibility | |
109 | ||
110 | It's a cardinal sin to break existing Perl programs. New warnings are | |
111 | contentious--some say that a program that emits warnings is not | |
112 | broken, while others say it is. Adding keywords has the potential to | |
113 | break programs, changing the meaning of existing token sequences or | |
114 | functions might break programs. | |
115 | ||
116 | =item Could it be a module instead? | |
117 | ||
118 | Perl 5 has extension mechanisms, modules and XS, specifically to avoid | |
119 | the need to keep changing the Perl interpreter. You can write modules | |
120 | that export functions, you can give those functions prototypes so they | |
121 | can be called like built-in functions, you can even write XS code to | |
122 | mess with the runtime data structures of the Perl interpreter if you | |
123 | want to implement really complicated things. If it can be done in a | |
124 | module instead of in the core, it's highly unlikely to be added. | |
125 | ||
126 | =item Is the feature generic enough? | |
127 | ||
128 | Is this something that only the submitter wants added to the language, | |
129 | or would it be broadly useful? Sometimes, instead of adding a feature | |
130 | with a tight focus, the porters might decide to wait until someone | |
131 | implements the more generalized feature. For instance, instead of | |
b432a672 | 132 | implementing a "delayed evaluation" feature, the porters are waiting |
e8cd7eae GS |
133 | for a macro system that would permit delayed evaluation and much more. |
134 | ||
135 | =item Does it potentially introduce new bugs? | |
136 | ||
137 | Radical rewrites of large chunks of the Perl interpreter have the | |
138 | potential to introduce new bugs. The smaller and more localized the | |
139 | change, the better. | |
140 | ||
141 | =item Does it preclude other desirable features? | |
142 | ||
143 | A patch is likely to be rejected if it closes off future avenues of | |
144 | development. For instance, a patch that placed a true and final | |
145 | interpretation on prototypes is likely to be rejected because there | |
146 | are still options for the future of prototypes that haven't been | |
147 | addressed. | |
148 | ||
149 | =item Is the implementation robust? | |
150 | ||
151 | Good patches (tight code, complete, correct) stand more chance of | |
152 | going in. Sloppy or incorrect patches might be placed on the back | |
153 | burner until the pumpking has time to fix, or might be discarded | |
154 | altogether without further notice. | |
155 | ||
156 | =item Is the implementation generic enough to be portable? | |
157 | ||
158 | The worst patches make use of a system-specific features. It's highly | |
159 | unlikely that nonportable additions to the Perl language will be | |
160 | accepted. | |
161 | ||
a936dd3c NC |
162 | =item Is the implementation tested? |
163 | ||
164 | Patches which change behaviour (fixing bugs or introducing new features) | |
165 | must include regression tests to verify that everything works as expected. | |
166 | Without tests provided by the original author, how can anyone else changing | |
167 | perl in the future be sure that they haven't unwittingly broken the behaviour | |
168 | the patch implements? And without tests, how can the patch's author be | |
9d077eaa | 169 | confident that his/her hard work put into the patch won't be accidentally |
a936dd3c NC |
170 | thrown away by someone in the future? |
171 | ||
e8cd7eae GS |
172 | =item Is there enough documentation? |
173 | ||
174 | Patches without documentation are probably ill-thought out or | |
175 | incomplete. Nothing can be added without documentation, so submitting | |
176 | a patch for the appropriate manpages as well as the source code is | |
a936dd3c | 177 | always a good idea. |
e8cd7eae GS |
178 | |
179 | =item Is there another way to do it? | |
180 | ||
b432a672 AL |
181 | Larry said "Although the Perl Slogan is I<There's More Than One Way |
182 | to Do It>, I hesitate to make 10 ways to do something". This is a | |
e8cd7eae GS |
183 | tricky heuristic to navigate, though--one man's essential addition is |
184 | another man's pointless cruft. | |
185 | ||
186 | =item Does it create too much work? | |
187 | ||
188 | Work for the pumpking, work for Perl programmers, work for module | |
189 | authors, ... Perl is supposed to be easy. | |
190 | ||
f6c51b38 GS |
191 | =item Patches speak louder than words |
192 | ||
193 | Working code is always preferred to pie-in-the-sky ideas. A patch to | |
194 | add a feature stands a much higher chance of making it to the language | |
195 | than does a random feature request, no matter how fervently argued the | |
b432a672 | 196 | request might be. This ties into "Will it be useful?", as the fact |
f6c51b38 GS |
197 | that someone took the time to make the patch demonstrates a strong |
198 | desire for the feature. | |
199 | ||
e8cd7eae GS |
200 | =back |
201 | ||
b432a672 AL |
202 | If you're on the list, you might hear the word "core" bandied |
203 | around. It refers to the standard distribution. "Hacking on the | |
204 | core" means you're changing the C source code to the Perl | |
205 | interpreter. "A core module" is one that ships with Perl. | |
e8cd7eae | 206 | |
a1f349fd MB |
207 | =head2 Keeping in sync |
208 | ||
e8cd7eae | 209 | The source code to the Perl interpreter, in its different versions, is |
f224927c JH |
210 | kept in a repository managed by a revision control system ( which is |
211 | currently the Perforce program, see http://perforce.com/ ). The | |
e8cd7eae GS |
212 | pumpkings and a few others have access to the repository to check in |
213 | changes. Periodically the pumpking for the development version of Perl | |
214 | will release a new version, so the rest of the porters can see what's | |
2be4c08b GS |
215 | changed. The current state of the main trunk of repository, and patches |
216 | that describe the individual changes that have happened since the last | |
217 | public 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 |
222 | If you're looking for a particular change, or a change that affected |
223 | a particular set of files, you may find the B<Perl Repository Browser> | |
224 | useful: | |
225 | ||
226 | http://public.activestate.com/cgi-bin/perlbrowse | |
227 | ||
228 | You may also want to subscribe to the perl5-changes mailing list to | |
229 | receive a copy of each patch that gets submitted to the maintenance | |
230 | and development "branches" of the perl repository. See | |
231 | http://lists.perl.org/ for subscription information. | |
232 | ||
a1f349fd MB |
233 | If you are a member of the perl5-porters mailing list, it is a good |
234 | thing to keep in touch with the most recent changes. If not only to | |
235 | verify if what you would have posted as a bug report isn't already | |
236 | solved in the most recent available perl development branch, also | |
237 | known as perl-current, bleading edge perl, bleedperl or bleadperl. | |
2be4c08b GS |
238 | |
239 | Needless to say, the source code in perl-current is usually in a perpetual | |
240 | state of evolution. You should expect it to be very buggy. Do B<not> use | |
241 | it for any purpose other than testing and development. | |
e8cd7eae | 242 | |
3e148164 JH |
243 | Keeping in sync with the most recent branch can be done in several ways, |
244 | but the most convenient and reliable way is using B<rsync>, available at | |
245 | ftp://rsync.samba.org/pub/rsync/ . (You can also get the most recent | |
246 | branch by FTP.) | |
a1f349fd MB |
247 | |
248 | If you choose to keep in sync using rsync, there are two approaches | |
3e148164 | 249 | to doing so: |
a1f349fd MB |
250 | |
251 | =over 4 | |
252 | ||
253 | =item rsync'ing the source tree | |
254 | ||
3e148164 | 255 | Presuming you are in the directory where your perl source resides |
b432a672 | 256 | and you have rsync installed and available, you can "upgrade" to |
a1f349fd MB |
257 | the bleadperl using: |
258 | ||
eb8afeef | 259 | # rsync -avz rsync://public.activestate.com/perl-current/ . |
a1f349fd MB |
260 | |
261 | This takes care of updating every single item in the source tree to | |
262 | the latest applied patch level, creating files that are new (to your | |
263 | distribution) and setting date/time stamps of existing files to | |
264 | reflect the bleadperl status. | |
265 | ||
c6d0653e LC |
266 | Note that this will not delete any files that were in '.' before |
267 | the rsync. Once you are sure that the rsync is running correctly, | |
268 | run 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 | |
272 | This will I<simulate> an rsync run that also deletes files not | |
273 | present in the bleadperl master copy. Observe the results from | |
274 | this run closely. If you are sure that the actual run would delete | |
275 | no files precious to you, you could remove the '--dry-run' option. | |
276 | ||
a1f349fd MB |
277 | You can than check what patch was the latest that was applied by |
278 | looking in the file B<.patch>, which will show the number of the | |
279 | latest patch. | |
280 | ||
281 | If you have more than one machine to keep in sync, and not all of | |
282 | them have access to the WAN (so you are not able to rsync all the | |
283 | source trees to the real source), there are some ways to get around | |
284 | this problem. | |
285 | ||
286 | =over 4 | |
287 | ||
288 | =item Using rsync over the LAN | |
289 | ||
290 | Set up a local rsync server which makes the rsynced source tree | |
3e148164 | 291 | available to the LAN and sync the other machines against this |
a1f349fd MB |
292 | directory. |
293 | ||
1577cd80 | 294 | From 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 | ||
304 | Having the other systems mounted over the NFS, you can take an | |
3e148164 JH |
305 | active pushing approach by checking the just updated tree against |
306 | the 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 | ||
339 | though this is not perfect. It could be improved with checking | |
a1f349fd MB |
340 | file checksums before updating. Not all NFS systems support |
341 | reliable utime support (when used over the NFS). | |
342 | ||
343 | =back | |
344 | ||
345 | =item rsync'ing the patches | |
346 | ||
347 | The source tree is maintained by the pumpking who applies patches to | |
348 | the files in the tree. These patches are either created by the | |
349 | pumpking himself using C<diff -c> after updating the file manually or | |
350 | by applying patches sent in by posters on the perl5-porters list. | |
351 | These patches are also saved and rsync'able, so you can apply them | |
352 | yourself to the source files. | |
353 | ||
354 | Presuming you are in a directory where your patches reside, you can | |
3e148164 | 355 | get them in sync with |
a1f349fd | 356 | |
eb8afeef | 357 | # rsync -avz rsync://public.activestate.com/perl-current-diffs/ . |
a1f349fd MB |
358 | |
359 | This makes sure the latest available patch is downloaded to your | |
360 | patch directory. | |
361 | ||
3e148164 | 362 | It'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 | ||
370 | or, since this is only a hint towards how it works, use CPAN-patchaperl | |
371 | from Andreas König 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 | |
381 | Since you don't have to apply the patches yourself, you are sure all | |
382 | files in the source tree are in the right state. | |
383 | ||
a1f349fd MB |
384 | =item It's more reliable |
385 | ||
0cfb3454 GS |
386 | While both the rsync-able source and patch areas are automatically |
387 | updated every few minutes, keep in mind that applying patches may | |
388 | sometimes mean careful hand-holding, especially if your version of | |
389 | the C<patch> program does not understand how to deal with new files, | |
390 | files 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 | |
400 | If you have more than one machine that you want to keep in track with | |
3e148164 | 401 | bleadperl, it's easier to rsync the patches only once and then apply |
a1f349fd MB |
402 | them to all the source trees on the different machines. |
403 | ||
404 | In case you try to keep in pace on 5 different machines, for which | |
405 | only one of them has access to the WAN, rsync'ing all the source | |
3e148164 | 406 | trees should than be done 5 times over the NFS. Having |
a1f349fd | 407 | rsync'ed the patches only once, I can apply them to all the source |
3e148164 | 408 | trees automatically. Need you say more ;-) |
a1f349fd MB |
409 | |
410 | =item It's a good reference | |
411 | ||
412 | If you do not only like to have the most recent development branch, | |
413 | but also like to B<fix> bugs, or extend features, you want to dive | |
414 | into the sources. If you are a seasoned perl core diver, you don't | |
415 | need no manuals, tips, roadmaps, perlguts.pod or other aids to find | |
416 | your way around. But if you are a starter, the patches may help you | |
417 | in finding where you should start and how to change the bits that | |
418 | bug you. | |
419 | ||
420 | The file B<Changes> is updated on occasions the pumpking sees as his | |
421 | own little sync points. On those occasions, he releases a tar-ball of | |
422 | the current source tree (i.e. perl@7582.tar.gz), which will be an | |
423 | excellent point to start with when choosing to use the 'rsync the | |
424 | patches' scheme. Starting with perl@7582, which means a set of source | |
425 | files on which the latest applied patch is number 7582, you apply all | |
f18956b7 | 426 | succeeding patches available from then on (7583, 7584, ...). |
a1f349fd MB |
427 | |
428 | You can use the patches later as a kind of search archive. | |
429 | ||
430 | =over 4 | |
431 | ||
432 | =item Finding a start point | |
433 | ||
434 | If you want to fix/change the behaviour of function/feature Foo, just | |
435 | scan the patches for patches that mention Foo either in the subject, | |
3e148164 | 436 | the comments, or the body of the fix. A good chance the patch shows |
a1f349fd MB |
437 | you the files that are affected by that patch which are very likely |
438 | to be the starting point of your journey into the guts of perl. | |
439 | ||
440 | =item Finding how to fix a bug | |
441 | ||
442 | If you've found I<where> the function/feature Foo misbehaves, but you | |
443 | don't know how to fix it (but you do know the change you want to | |
444 | make), you can, again, peruse the patches for similar changes and | |
445 | look how others apply the fix. | |
446 | ||
447 | =item Finding the source of misbehaviour | |
448 | ||
449 | When you keep in sync with bleadperl, the pumpking would love to | |
3958b146 | 450 | I<see> that the community efforts really work. So after each of his |
a1f349fd MB |
451 | sync points, you are to 'make test' to check if everything is still |
452 | in working order. If it is, you do 'make ok', which will send an OK | |
902821cc | 453 | report to I<perlbug@perl.org>. (If you do not have access to a mailer |
3e148164 | 454 | from the system you just finished successfully 'make test', you can |
a1f349fd MB |
455 | do 'make okfile', which creates the file C<perl.ok>, which you can |
456 | than take to your favourite mailer and mail yourself). | |
457 | ||
3958b146 | 458 | But of course, as always, things will not always lead to a success |
a1f349fd MB |
459 | path, and one or more test do not pass the 'make test'. Before |
460 | sending in a bug report (using 'make nok' or 'make nokfile'), check | |
461 | the mailing list if someone else has reported the bug already and if | |
462 | so, confirm it by replying to that message. If not, you might want to | |
463 | trace the source of that misbehaviour B<before> sending in the bug, | |
464 | which will help all the other porters in finding the solution. | |
465 | ||
3e148164 JH |
466 | Here the saved patches come in very handy. You can check the list of |
467 | patches to see which patch changed what file and what change caused | |
468 | the misbehaviour. If you note that in the bug report, it saves the | |
469 | one trying to solve it, looking for that point. | |
a1f349fd MB |
470 | |
471 | =back | |
472 | ||
473 | If searching the patches is too bothersome, you might consider using | |
474 | perl's bugtron to find more information about discussions and | |
475 | ramblings on posted bugs. | |
476 | ||
3e148164 JH |
477 | If you want to get the best of both worlds, rsync both the source |
478 | tree for convenience, reliability and ease and rsync the patches | |
479 | for reference. | |
480 | ||
52315700 RF |
481 | =back |
482 | ||
fcc89a64 DR |
483 | =head2 Working with the source |
484 | ||
485 | Because you cannot use the Perforce client, you cannot easily generate | |
486 | diffs against the repository, nor will merges occur when you update | |
487 | via rsync. If you edit a file locally and then rsync against the | |
488 | latest source, changes made in the remote copy will I<overwrite> your | |
489 | local versions! | |
490 | ||
491 | The best way to deal with this is to maintain a tree of symlinks to | |
492 | the rsync'd source. Then, when you want to edit a file, you remove | |
493 | the symlink, copy the real file into the other tree, and edit it. You | |
494 | can then diff your edited file against the original to generate a | |
495 | patch, and you can safely update the original tree. | |
496 | ||
497 | Perl's F<Configure> script can generate this tree of symlinks for you. | |
498 | The following example assumes that you have used rsync to pull a copy | |
499 | of the Perl source into the F<perl-rsync> directory. In the directory | |
500 | above 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 | ||
506 | This will start the Perl configuration process. After a few prompts, | |
507 | you 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 | ||
520 | The specifics may vary based on your operating system, of course. | |
521 | After you see this, you can abort the F<Configure> script, and you | |
522 | will see that the directory you are in has a tree of symlinks to the | |
523 | F<perl-rsync> directories and files. | |
524 | ||
525 | If you plan to do a lot of work with the Perl source, here are some | |
526 | Bourne 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 | ||
545 | Replace "vi" with your favorite flavor of editor. | |
546 | ||
547 | Here is another function which will quickly generate a patch for the | |
548 | files 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 | ||
563 | This function produces patches which include enough context to make | |
564 | your changes obvious. This makes it easier for the Perl pumpking(s) | |
565 | to review them when you send them to the perl5-porters list, and that | |
566 | means they're more likely to get applied. | |
567 | ||
568 | This function assumed a GNU diff, and may require some tweaking for | |
569 | other diff variants. | |
52315700 | 570 | |
3fd28c4e | 571 | =head2 Perlbug administration |
52315700 | 572 | |
902821cc RGS |
573 | There is a single remote administrative interface for modifying bug status, |
574 | category, open issues etc. using the B<RT> bugtracker system, maintained | |
575 | by Robert Spier. Become an administrator, and close any bugs you can get | |
3fd28c4e | 576 | your sticky mitts on: |
52315700 | 577 | |
902821cc | 578 | http://rt.perl.org/rt3/ |
52315700 | 579 | |
3fd28c4e | 580 | To email the bug system administrators: |
52315700 | 581 | |
3fd28c4e | 582 | "perlbug-admin" <perlbug-admin@perl.org> |
52315700 | 583 | |
a1f349fd MB |
584 | =head2 Submitting patches |
585 | ||
f7e1e956 MS |
586 | Always submit patches to I<perl5-porters@perl.org>. If you're |
587 | patching a core module and there's an author listed, send the author a | |
588 | copy (see L<Patching a core module>). This lets other porters review | |
589 | your patch, which catches a surprising number of errors in patches. | |
590 | Either use the diff program (available in source code form from | |
f224927c | 591 | ftp://ftp.gnu.org/pub/gnu/ , or use Johan Vromans' I<makepatch> |
f7e1e956 MS |
592 | (available from I<CPAN/authors/id/JV/>). Unified diffs are preferred, |
593 | but context diffs are accepted. Do not send RCS-style diffs or diffs | |
594 | without context lines. More information is given in the | |
595 | I<Porting/patching.pod> file in the Perl source distribution. Please | |
824d470b RGS |
596 | patch against the latest B<development> version. (e.g., even if you're |
597 | fixing a bug in the 5.8 track, patch against the latest B<development> | |
598 | version rsynced from rsync://public.activestate.com/perl-current/ ) | |
599 | ||
600 | If changes are accepted, they are applied to the development branch. Then | |
601 | the 5.8 pumpking decides which of those patches is to be backported to the | |
602 | maint branch. Only patches that survive the heat of the development | |
f7e1e956 MS |
603 | branch get applied to maintenance versions. |
604 | ||
824d470b | 605 | Your patch should also update the documentation and test suite. See |
f7e1e956 | 606 | L<Writing a test>. |
e8cd7eae | 607 | |
824d470b RGS |
608 | Patching documentation also follows the same order: if accepted, a patch |
609 | is first applied to B<development>, and if relevant then it's backported | |
610 | to B<maintenance>. (With an exception for some patches that document | |
611 | behaviour that only appears in the maintenance branch, but which has | |
612 | changed in the development version.) | |
613 | ||
e8cd7eae GS |
614 | To report a bug in Perl, use the program I<perlbug> which comes with |
615 | Perl (if you can't get Perl to work, send mail to the address | |
f18956b7 | 616 | I<perlbug@perl.org> or I<perlbug@perl.com>). Reporting bugs through |
e8cd7eae | 617 | I<perlbug> feeds into the automated bug-tracking system, access to |
902821cc | 618 | which is provided through the web at http://rt.perl.org/rt3/ . It |
e8cd7eae GS |
619 | often pays to check the archives of the perl5-porters mailing list to |
620 | see whether the bug you're reporting has been reported before, and if | |
621 | so whether it was considered a bug. See above for the location of | |
622 | the searchable archives. | |
623 | ||
f224927c | 624 | The CPAN testers ( http://testers.cpan.org/ ) are a group of |
ba139f7d | 625 | volunteers who test CPAN modules on a variety of platforms. Perl |
902821cc RGS |
626 | Smokers ( http://www.nntp.perl.org/group/perl.daily-build ) |
627 | automatically test Perl source releases on platforms with various | |
ba139f7d | 628 | configurations. Both efforts welcome volunteers. |
e8cd7eae | 629 | |
e8cd7eae GS |
630 | It's a good idea to read and lurk for a while before chipping in. |
631 | That way you'll get to see the dynamic of the conversations, learn the | |
632 | personalities of the players, and hopefully be better prepared to make | |
633 | a useful contribution when do you speak up. | |
634 | ||
635 | If after all this you still think you want to join the perl5-porters | |
f6c51b38 GS |
636 | mailing list, send mail to I<perl5-porters-subscribe@perl.org>. To |
637 | unsubscribe, send mail to I<perl5-porters-unsubscribe@perl.org>. | |
e8cd7eae | 638 | |
a422fd2d SC |
639 | To hack on the Perl guts, you'll need to read the following things: |
640 | ||
641 | =over 3 | |
642 | ||
643 | =item L<perlguts> | |
644 | ||
645 | This is of paramount importance, since it's the documentation of what | |
646 | goes where in the Perl source. Read it over a couple of times and it | |
647 | might start to make sense - don't worry if it doesn't yet, because the | |
648 | best way to study it is to read it in conjunction with poking at Perl | |
649 | source, and we'll do that later on. | |
650 | ||
651 | You might also want to look at Gisle Aas's illustrated perlguts - | |
652 | there's no guarantee that this will be absolutely up-to-date with the | |
653 | latest documentation in the Perl core, but the fundamentals will be | |
1577cd80 | 654 | right. ( http://gisle.aas.no/perl/illguts/ ) |
a422fd2d SC |
655 | |
656 | =item L<perlxstut> and L<perlxs> | |
657 | ||
658 | A working knowledge of XSUB programming is incredibly useful for core | |
659 | hacking; XSUBs use techniques drawn from the PP code, the portion of the | |
660 | guts that actually executes a Perl program. It's a lot gentler to learn | |
661 | those techniques from simple examples and explanation than from the core | |
662 | itself. | |
663 | ||
664 | =item L<perlapi> | |
665 | ||
666 | The documentation for the Perl API explains what some of the internal | |
667 | functions do, as well as the many macros used in the source. | |
668 | ||
669 | =item F<Porting/pumpkin.pod> | |
670 | ||
671 | This is a collection of words of wisdom for a Perl porter; some of it is | |
672 | only useful to the pumpkin holder, but most of it applies to anyone | |
673 | wanting to go about Perl development. | |
674 | ||
675 | =item The perl5-porters FAQ | |
676 | ||
902821cc RGS |
677 | This should be available from http://dev.perl.org/perl5/docs/p5p-faq.html . |
678 | It contains hints on reading perl5-porters, information on how | |
679 | perl5-porters works and how Perl development in general works. | |
a422fd2d SC |
680 | |
681 | =back | |
682 | ||
683 | =head2 Finding Your Way Around | |
684 | ||
685 | Perl maintenance can be split into a number of areas, and certain people | |
686 | (pumpkins) will have responsibility for each area. These areas sometimes | |
687 | correspond to files or directories in the source kit. Among the areas are: | |
688 | ||
689 | =over 3 | |
690 | ||
691 | =item Core modules | |
692 | ||
693 | Modules shipped as part of the Perl core live in the F<lib/> and F<ext/> | |
694 | subdirectories: F<lib/> is for the pure-Perl modules, and F<ext/> | |
695 | contains the core XS modules. | |
696 | ||
f7e1e956 MS |
697 | =item Tests |
698 | ||
699 | There are tests for nearly all the modules, built-ins and major bits | |
700 | of functionality. Test files all have a .t suffix. Module tests live | |
701 | in the F<lib/> and F<ext/> directories next to the module being | |
702 | tested. Others live in F<t/>. See L<Writing a test> | |
703 | ||
a422fd2d SC |
704 | =item Documentation |
705 | ||
706 | Documentation maintenance includes looking after everything in the | |
707 | F<pod/> directory, (as well as contributing new documentation) and | |
708 | the documentation to the modules in core. | |
709 | ||
710 | =item Configure | |
711 | ||
712 | The configure process is the way we make Perl portable across the | |
713 | myriad of operating systems it supports. Responsibility for the | |
714 | configure, build and installation process, as well as the overall | |
715 | portability of the core code rests with the configure pumpkin - others | |
716 | help out with individual operating systems. | |
717 | ||
718 | The files involved are the operating system directories, (F<win32/>, | |
719 | F<os2/>, F<vms/> and so on) the shell scripts which generate F<config.h> | |
720 | and F<Makefile>, as well as the metaconfig files which generate | |
721 | F<Configure>. (metaconfig isn't included in the core distribution.) | |
722 | ||
723 | =item Interpreter | |
724 | ||
725 | And of course, there's the core of the Perl interpreter itself. Let's | |
726 | have a look at that in a little more detail. | |
727 | ||
728 | =back | |
729 | ||
730 | Before we leave looking at the layout, though, don't forget that | |
731 | F<MANIFEST> contains not only the file names in the Perl distribution, | |
732 | but short descriptions of what's in them, too. For an overview of the | |
733 | important files, try this: | |
734 | ||
735 | perl -lne 'print if /^[^\/]+\.[ch]\s+/' MANIFEST | |
736 | ||
737 | =head2 Elements of the interpreter | |
738 | ||
739 | The work of the interpreter has two main stages: compiling the code | |
740 | into the internal representation, or bytecode, and then executing it. | |
741 | L<perlguts/Compiled code> explains exactly how the compilation stage | |
742 | happens. | |
743 | ||
744 | Here is a short breakdown of perl's operation: | |
745 | ||
746 | =over 3 | |
747 | ||
748 | =item Startup | |
749 | ||
750 | The action begins in F<perlmain.c>. (or F<miniperlmain.c> for miniperl) | |
751 | This is very high-level code, enough to fit on a single screen, and it | |
752 | resembles the code found in L<perlembed>; most of the real action takes | |
753 | place in F<perl.c> | |
754 | ||
755 | First, F<perlmain.c> allocates some memory and constructs a Perl | |
756 | interpreter: | |
757 | ||
758 | 1 PERL_SYS_INIT3(&argc,&argv,&env); | |
759 | 2 | |
760 | 3 if (!PL_do_undump) { | |
761 | 4 my_perl = perl_alloc(); | |
762 | 5 if (!my_perl) | |
763 | 6 exit(1); | |
764 | 7 perl_construct(my_perl); | |
765 | 8 PL_perl_destruct_level = 0; | |
766 | 9 } | |
767 | ||
768 | Line 1 is a macro, and its definition is dependent on your operating | |
769 | system. Line 3 references C<PL_do_undump>, a global variable - all | |
770 | global variables in Perl start with C<PL_>. This tells you whether the | |
771 | current running program was created with the C<-u> flag to perl and then | |
772 | F<undump>, which means it's going to be false in any sane context. | |
773 | ||
774 | Line 4 calls a function in F<perl.c> to allocate memory for a Perl | |
775 | interpreter. It's quite a simple function, and the guts of it looks like | |
776 | this: | |
777 | ||
778 | my_perl = (PerlInterpreter*)PerlMem_malloc(sizeof(PerlInterpreter)); | |
779 | ||
780 | Here you see an example of Perl's system abstraction, which we'll see | |
781 | later: C<PerlMem_malloc> is either your system's C<malloc>, or Perl's | |
782 | own C<malloc> as defined in F<malloc.c> if you selected that option at | |
783 | configure time. | |
784 | ||
785 | Next, in line 7, we construct the interpreter; this sets up all the | |
786 | special variables that Perl needs, the stacks, and so on. | |
787 | ||
788 | Now we pass Perl the command line options, and tell it to go: | |
789 | ||
790 | exitstatus = perl_parse(my_perl, xs_init, argc, argv, (char **)NULL); | |
791 | if (!exitstatus) { | |
792 | exitstatus = perl_run(my_perl); | |
793 | } | |
794 | ||
795 | ||
796 | C<perl_parse> is actually a wrapper around C<S_parse_body>, as defined | |
797 | in F<perl.c>, which processes the command line options, sets up any | |
798 | statically linked XS modules, opens the program and calls C<yyparse> to | |
799 | parse it. | |
800 | ||
801 | =item Parsing | |
802 | ||
803 | The aim of this stage is to take the Perl source, and turn it into an op | |
804 | tree. We'll see what one of those looks like later. Strictly speaking, | |
805 | there's three things going on here. | |
806 | ||
807 | C<yyparse>, the parser, lives in F<perly.c>, although you're better off | |
808 | reading the original YACC input in F<perly.y>. (Yes, Virginia, there | |
809 | B<is> a YACC grammar for Perl!) The job of the parser is to take your | |
b432a672 | 810 | code and "understand" it, splitting it into sentences, deciding which |
a422fd2d SC |
811 | operands go with which operators and so on. |
812 | ||
813 | The parser is nobly assisted by the lexer, which chunks up your input | |
814 | into tokens, and decides what type of thing each token is: a variable | |
815 | name, an operator, a bareword, a subroutine, a core function, and so on. | |
816 | The main point of entry to the lexer is C<yylex>, and that and its | |
817 | associated routines can be found in F<toke.c>. Perl isn't much like | |
818 | other computer languages; it's highly context sensitive at times, it can | |
819 | be tricky to work out what sort of token something is, or where a token | |
820 | ends. As such, there's a lot of interplay between the tokeniser and the | |
821 | parser, which can get pretty frightening if you're not used to it. | |
822 | ||
823 | As the parser understands a Perl program, it builds up a tree of | |
824 | operations for the interpreter to perform during execution. The routines | |
825 | which construct and link together the various operations are to be found | |
826 | in F<op.c>, and will be examined later. | |
827 | ||
828 | =item Optimization | |
829 | ||
830 | Now the parsing stage is complete, and the finished tree represents | |
831 | the operations that the Perl interpreter needs to perform to execute our | |
832 | program. Next, Perl does a dry run over the tree looking for | |
833 | optimisations: constant expressions such as C<3 + 4> will be computed | |
834 | now, and the optimizer will also see if any multiple operations can be | |
835 | replaced with a single one. For instance, to fetch the variable C<$foo>, | |
836 | instead of grabbing the glob C<*foo> and looking at the scalar | |
837 | component, the optimizer fiddles the op tree to use a function which | |
838 | directly looks up the scalar in question. The main optimizer is C<peep> | |
839 | in F<op.c>, and many ops have their own optimizing functions. | |
840 | ||
841 | =item Running | |
842 | ||
843 | Now we're finally ready to go: we have compiled Perl byte code, and all | |
844 | that's left to do is run it. The actual execution is done by the | |
845 | C<runops_standard> function in F<run.c>; more specifically, it's done by | |
846 | these three innocent looking lines: | |
847 | ||
848 | while ((PL_op = CALL_FPTR(PL_op->op_ppaddr)(aTHX))) { | |
849 | PERL_ASYNC_CHECK(); | |
850 | } | |
851 | ||
852 | You may be more comfortable with the Perl version of that: | |
853 | ||
854 | PERL_ASYNC_CHECK() while $Perl::op = &{$Perl::op->{function}}; | |
855 | ||
856 | Well, maybe not. Anyway, each op contains a function pointer, which | |
857 | stipulates the function which will actually carry out the operation. | |
858 | This function will return the next op in the sequence - this allows for | |
859 | things like C<if> which choose the next op dynamically at run time. | |
860 | The C<PERL_ASYNC_CHECK> makes sure that things like signals interrupt | |
861 | execution if required. | |
862 | ||
863 | The actual functions called are known as PP code, and they're spread | |
b432a672 | 864 | between four files: F<pp_hot.c> contains the "hot" code, which is most |
a422fd2d SC |
865 | often used and highly optimized, F<pp_sys.c> contains all the |
866 | system-specific functions, F<pp_ctl.c> contains the functions which | |
867 | implement control structures (C<if>, C<while> and the like) and F<pp.c> | |
868 | contains everything else. These are, if you like, the C code for Perl's | |
869 | built-in functions and operators. | |
870 | ||
dfc98234 DM |
871 | Note that each C<pp_> function is expected to return a pointer to the next |
872 | op. Calls to perl subs (and eval blocks) are handled within the same | |
873 | runops loop, and do not consume extra space on the C stack. For example, | |
874 | C<pp_entersub> and C<pp_entertry> just push a C<CxSUB> or C<CxEVAL> block | |
875 | struct onto the context stack which contain the address of the op | |
876 | following the sub call or eval. They then return the first op of that sub | |
877 | or eval block, and so execution continues of that sub or block. Later, a | |
878 | C<pp_leavesub> or C<pp_leavetry> op pops the C<CxSUB> or C<CxEVAL>, | |
879 | retrieves the return op from it, and returns it. | |
880 | ||
881 | =item Exception handing | |
882 | ||
28a5cf3b | 883 | Perl's exception handing (i.e. C<die> etc) is built on top of the low-level |
dfc98234 | 884 | C<setjmp()>/C<longjmp()> C-library functions. These basically provide a |
28a5cf3b | 885 | way to capture the current PC and SP registers and later restore them; i.e. |
dfc98234 DM |
886 | a C<longjmp()> continues at the point in code where a previous C<setjmp()> |
887 | was done, with anything further up on the C stack being lost. This is why | |
888 | code should always save values using C<SAVE_FOO> rather than in auto | |
889 | variables. | |
890 | ||
891 | The perl core wraps C<setjmp()> etc in the macros C<JMPENV_PUSH> and | |
892 | C<JMPENV_JUMP>. The basic rule of perl exceptions is that C<exit>, and | |
893 | C<die> (in the absence of C<eval>) perform a C<JMPENV_JUMP(2)>, while | |
894 | C<die> within C<eval> does a C<JMPENV_JUMP(3)>. | |
895 | ||
896 | At entry points to perl, such as C<perl_parse()>, C<perl_run()> and | |
897 | C<call_sv(cv, G_EVAL)> each does a C<JMPENV_PUSH>, then enter a runops | |
898 | loop or whatever, and handle possible exception returns. For a 2 return, | |
899 | final cleanup is performed, such as popping stacks and calling C<CHECK> or | |
900 | C<END> blocks. Amongst other things, this is how scope cleanup still | |
901 | occurs during an C<exit>. | |
902 | ||
903 | If a C<die> can find a C<CxEVAL> block on the context stack, then the | |
904 | stack is popped to that level and the return op in that block is assigned | |
905 | to C<PL_restartop>; then a C<JMPENV_JUMP(3)> is performed. This normally | |
906 | passes control back to the guard. In the case of C<perl_run> and | |
907 | C<call_sv>, a non-null C<PL_restartop> triggers re-entry to the runops | |
908 | loop. The is the normal way that C<die> or C<croak> is handled within an | |
909 | C<eval>. | |
910 | ||
911 | Sometimes ops are executed within an inner runops loop, such as tie, sort | |
912 | or overload code. In this case, something like | |
913 | ||
914 | sub FETCH { eval { die } } | |
915 | ||
916 | would cause a longjmp right back to the guard in C<perl_run>, popping both | |
917 | runops loops, which is clearly incorrect. One way to avoid this is for the | |
918 | tie code to do a C<JMPENV_PUSH> before executing C<FETCH> in the inner | |
919 | runops loop, but for efficiency reasons, perl in fact just sets a flag, | |
920 | using C<CATCH_SET(TRUE)>. The C<pp_require>, C<pp_entereval> and | |
921 | C<pp_entertry> ops check this flag, and if true, they call C<docatch>, | |
922 | which does a C<JMPENV_PUSH> and starts a new runops level to execute the | |
923 | code, rather than doing it on the current loop. | |
924 | ||
925 | As a further optimisation, on exit from the eval block in the C<FETCH>, | |
926 | execution of the code following the block is still carried on in the inner | |
927 | loop. When an exception is raised, C<docatch> compares the C<JMPENV> | |
928 | level of the C<CxEVAL> with C<PL_top_env> and if they differ, just | |
929 | re-throws the exception. In this way any inner loops get popped. | |
930 | ||
931 | Here's an example. | |
932 | ||
933 | 1: eval { tie @a, 'A' }; | |
934 | 2: sub A::TIEARRAY { | |
935 | 3: eval { die }; | |
936 | 4: die; | |
937 | 5: } | |
938 | ||
939 | To run this code, C<perl_run> is called, which does a C<JMPENV_PUSH> then | |
940 | enters a runops loop. This loop executes the eval and tie ops on line 1, | |
941 | with the eval pushing a C<CxEVAL> onto the context stack. | |
942 | ||
943 | The C<pp_tie> does a C<CATCH_SET(TRUE)>, then starts a second runops loop | |
944 | to execute the body of C<TIEARRAY>. When it executes the entertry op on | |
945 | line 3, C<CATCH_GET> is true, so C<pp_entertry> calls C<docatch> which | |
946 | does a C<JMPENV_PUSH> and starts a third runops loop, which then executes | |
947 | the die op. At this point the C call stack looks like this: | |
948 | ||
949 | Perl_pp_die | |
950 | Perl_runops # third loop | |
951 | S_docatch_body | |
952 | S_docatch | |
953 | Perl_pp_entertry | |
954 | Perl_runops # second loop | |
955 | S_call_body | |
956 | Perl_call_sv | |
957 | Perl_pp_tie | |
958 | Perl_runops # first loop | |
959 | S_run_body | |
960 | perl_run | |
961 | main | |
962 | ||
963 | and the context and data stacks, as shown by C<-Dstv>, look like: | |
964 | ||
965 | STACK 0: MAIN | |
966 | CX 0: BLOCK => | |
967 | CX 1: EVAL => AV() PV("A"\0) | |
968 | retop=leave | |
969 | STACK 1: MAGIC | |
970 | CX 0: SUB => | |
971 | retop=(null) | |
972 | CX 1: EVAL => * | |
973 | retop=nextstate | |
974 | ||
975 | The die pops the first C<CxEVAL> off the context stack, sets | |
976 | C<PL_restartop> from it, does a C<JMPENV_JUMP(3)>, and control returns to | |
977 | the top C<docatch>. This then starts another third-level runops level, | |
978 | which executes the nextstate, pushmark and die ops on line 4. At the point | |
979 | that the second C<pp_die> is called, the C call stack looks exactly like | |
980 | that above, even though we are no longer within an inner eval; this is | |
981 | because of the optimization mentioned earlier. However, the context stack | |
982 | now looks like this, ie with the top CxEVAL popped: | |
983 | ||
984 | STACK 0: MAIN | |
985 | CX 0: BLOCK => | |
986 | CX 1: EVAL => AV() PV("A"\0) | |
987 | retop=leave | |
988 | STACK 1: MAGIC | |
989 | CX 0: SUB => | |
990 | retop=(null) | |
991 | ||
992 | The die on line 4 pops the context stack back down to the CxEVAL, leaving | |
993 | it as: | |
994 | ||
995 | STACK 0: MAIN | |
996 | CX 0: BLOCK => | |
997 | ||
998 | As usual, C<PL_restartop> is extracted from the C<CxEVAL>, and a | |
999 | C<JMPENV_JUMP(3)> done, which pops the C stack back to the docatch: | |
1000 | ||
1001 | S_docatch | |
1002 | Perl_pp_entertry | |
1003 | Perl_runops # second loop | |
1004 | S_call_body | |
1005 | Perl_call_sv | |
1006 | Perl_pp_tie | |
1007 | Perl_runops # first loop | |
1008 | S_run_body | |
1009 | perl_run | |
1010 | main | |
1011 | ||
1012 | In this case, because the C<JMPENV> level recorded in the C<CxEVAL> | |
1013 | differs from the current one, C<docatch> just does a C<JMPENV_JUMP(3)> | |
1014 | and the C stack unwinds to: | |
1015 | ||
1016 | perl_run | |
1017 | main | |
1018 | ||
1019 | Because C<PL_restartop> is non-null, C<run_body> starts a new runops loop | |
1020 | and execution continues. | |
1021 | ||
a422fd2d SC |
1022 | =back |
1023 | ||
1024 | =head2 Internal Variable Types | |
1025 | ||
1026 | You should by now have had a look at L<perlguts>, which tells you about | |
1027 | Perl's internal variable types: SVs, HVs, AVs and the rest. If not, do | |
1028 | that now. | |
1029 | ||
1030 | These variables are used not only to represent Perl-space variables, but | |
1031 | also any constants in the code, as well as some structures completely | |
1032 | internal to Perl. The symbol table, for instance, is an ordinary Perl | |
1033 | hash. Your code is represented by an SV as it's read into the parser; | |
1034 | any program files you call are opened via ordinary Perl filehandles, and | |
1035 | so on. | |
1036 | ||
1037 | The core L<Devel::Peek|Devel::Peek> module lets us examine SVs from a | |
1038 | Perl program. Let's see, for instance, how Perl treats the constant | |
1039 | C<"hello">. | |
1040 | ||
1041 | % perl -MDevel::Peek -e 'Dump("hello")' | |
1042 | 1 SV = PV(0xa041450) at 0xa04ecbc | |
1043 | 2 REFCNT = 1 | |
1044 | 3 FLAGS = (POK,READONLY,pPOK) | |
1045 | 4 PV = 0xa0484e0 "hello"\0 | |
1046 | 5 CUR = 5 | |
1047 | 6 LEN = 6 | |
1048 | ||
1049 | Reading C<Devel::Peek> output takes a bit of practise, so let's go | |
1050 | through it line by line. | |
1051 | ||
1052 | Line 1 tells us we're looking at an SV which lives at C<0xa04ecbc> in | |
1053 | memory. SVs themselves are very simple structures, but they contain a | |
1054 | pointer to a more complex structure. In this case, it's a PV, a | |
1055 | structure which holds a string value, at location C<0xa041450>. Line 2 | |
1056 | is the reference count; there are no other references to this data, so | |
1057 | it's 1. | |
1058 | ||
1059 | Line 3 are the flags for this SV - it's OK to use it as a PV, it's a | |
1060 | read-only SV (because it's a constant) and the data is a PV internally. | |
1061 | Next we've got the contents of the string, starting at location | |
1062 | C<0xa0484e0>. | |
1063 | ||
1064 | Line 5 gives us the current length of the string - note that this does | |
1065 | B<not> include the null terminator. Line 6 is not the length of the | |
1066 | string, but the length of the currently allocated buffer; as the string | |
1067 | grows, Perl automatically extends the available storage via a routine | |
1068 | called C<SvGROW>. | |
1069 | ||
1070 | You can get at any of these quantities from C very easily; just add | |
1071 | C<Sv> to the name of the field shown in the snippet, and you've got a | |
1072 | macro which will return the value: C<SvCUR(sv)> returns the current | |
1073 | length of the string, C<SvREFCOUNT(sv)> returns the reference count, | |
1074 | C<SvPV(sv, len)> returns the string itself with its length, and so on. | |
1075 | More macros to manipulate these properties can be found in L<perlguts>. | |
1076 | ||
1077 | Let's take an example of manipulating a PV, from C<sv_catpvn>, in F<sv.c> | |
1078 | ||
1079 | 1 void | |
1080 | 2 Perl_sv_catpvn(pTHX_ register SV *sv, register const char *ptr, register STRLEN len) | |
1081 | 3 { | |
1082 | 4 STRLEN tlen; | |
1083 | 5 char *junk; | |
1084 | ||
1085 | 6 junk = SvPV_force(sv, tlen); | |
1086 | 7 SvGROW(sv, tlen + len + 1); | |
1087 | 8 if (ptr == junk) | |
1088 | 9 ptr = SvPVX(sv); | |
1089 | 10 Move(ptr,SvPVX(sv)+tlen,len,char); | |
1090 | 11 SvCUR(sv) += len; | |
1091 | 12 *SvEND(sv) = '\0'; | |
1092 | 13 (void)SvPOK_only_UTF8(sv); /* validate pointer */ | |
1093 | 14 SvTAINT(sv); | |
1094 | 15 } | |
1095 | ||
1096 | This is a function which adds a string, C<ptr>, of length C<len> onto | |
1097 | the end of the PV stored in C<sv>. The first thing we do in line 6 is | |
1098 | make sure that the SV B<has> a valid PV, by calling the C<SvPV_force> | |
1099 | macro to force a PV. As a side effect, C<tlen> gets set to the current | |
1100 | value of the PV, and the PV itself is returned to C<junk>. | |
1101 | ||
b1866b2d | 1102 | In line 7, we make sure that the SV will have enough room to accommodate |
a422fd2d SC |
1103 | the old string, the new string and the null terminator. If C<LEN> isn't |
1104 | big enough, C<SvGROW> will reallocate space for us. | |
1105 | ||
1106 | Now, if C<junk> is the same as the string we're trying to add, we can | |
1107 | grab the string directly from the SV; C<SvPVX> is the address of the PV | |
1108 | in the SV. | |
1109 | ||
1110 | Line 10 does the actual catenation: the C<Move> macro moves a chunk of | |
1111 | memory around: we move the string C<ptr> to the end of the PV - that's | |
1112 | the start of the PV plus its current length. We're moving C<len> bytes | |
1113 | of type C<char>. After doing so, we need to tell Perl we've extended the | |
1114 | string, by altering C<CUR> to reflect the new length. C<SvEND> is a | |
1115 | macro which gives us the end of the string, so that needs to be a | |
1116 | C<"\0">. | |
1117 | ||
1118 | Line 13 manipulates the flags; since we've changed the PV, any IV or NV | |
1119 | values will no longer be valid: if we have C<$a=10; $a.="6";> we don't | |
1e54db1a | 1120 | want to use the old IV of 10. C<SvPOK_only_utf8> is a special UTF-8-aware |
a422fd2d SC |
1121 | version of C<SvPOK_only>, a macro which turns off the IOK and NOK flags |
1122 | and turns on POK. The final C<SvTAINT> is a macro which launders tainted | |
1123 | data if taint mode is turned on. | |
1124 | ||
1125 | AVs and HVs are more complicated, but SVs are by far the most common | |
1126 | variable type being thrown around. Having seen something of how we | |
1127 | manipulate these, let's go on and look at how the op tree is | |
1128 | constructed. | |
1129 | ||
1130 | =head2 Op Trees | |
1131 | ||
1132 | First, what is the op tree, anyway? The op tree is the parsed | |
1133 | representation of your program, as we saw in our section on parsing, and | |
1134 | it's the sequence of operations that Perl goes through to execute your | |
1135 | program, as we saw in L</Running>. | |
1136 | ||
1137 | An op is a fundamental operation that Perl can perform: all the built-in | |
1138 | functions and operators are ops, and there are a series of ops which | |
1139 | deal with concepts the interpreter needs internally - entering and | |
1140 | leaving a block, ending a statement, fetching a variable, and so on. | |
1141 | ||
1142 | The op tree is connected in two ways: you can imagine that there are two | |
1143 | "routes" through it, two orders in which you can traverse the tree. | |
1144 | First, parse order reflects how the parser understood the code, and | |
1145 | secondly, execution order tells perl what order to perform the | |
1146 | operations in. | |
1147 | ||
1148 | The easiest way to examine the op tree is to stop Perl after it has | |
1149 | finished parsing, and get it to dump out the tree. This is exactly what | |
7d7d5695 RGS |
1150 | the compiler backends L<B::Terse|B::Terse>, L<B::Concise|B::Concise> |
1151 | and L<B::Debug|B::Debug> do. | |
a422fd2d SC |
1152 | |
1153 | Let's have a look at how Perl sees C<$a = $b + $c>: | |
1154 | ||
1155 | % perl -MO=Terse -e '$a=$b+$c' | |
1156 | 1 LISTOP (0x8179888) leave | |
1157 | 2 OP (0x81798b0) enter | |
1158 | 3 COP (0x8179850) nextstate | |
1159 | 4 BINOP (0x8179828) sassign | |
1160 | 5 BINOP (0x8179800) add [1] | |
1161 | 6 UNOP (0x81796e0) null [15] | |
1162 | 7 SVOP (0x80fafe0) gvsv GV (0x80fa4cc) *b | |
1163 | 8 UNOP (0x81797e0) null [15] | |
1164 | 9 SVOP (0x8179700) gvsv GV (0x80efeb0) *c | |
1165 | 10 UNOP (0x816b4f0) null [15] | |
1166 | 11 SVOP (0x816dcf0) gvsv GV (0x80fa460) *a | |
1167 | ||
1168 | Let's start in the middle, at line 4. This is a BINOP, a binary | |
1169 | operator, which is at location C<0x8179828>. The specific operator in | |
1170 | question is C<sassign> - scalar assignment - and you can find the code | |
1171 | which implements it in the function C<pp_sassign> in F<pp_hot.c>. As a | |
1172 | binary operator, it has two children: the add operator, providing the | |
1173 | result of C<$b+$c>, is uppermost on line 5, and the left hand side is on | |
1174 | line 10. | |
1175 | ||
1176 | Line 10 is the null op: this does exactly nothing. What is that doing | |
1177 | there? If you see the null op, it's a sign that something has been | |
1178 | optimized away after parsing. As we mentioned in L</Optimization>, | |
1179 | the optimization stage sometimes converts two operations into one, for | |
1180 | example when fetching a scalar variable. When this happens, instead of | |
1181 | rewriting the op tree and cleaning up the dangling pointers, it's easier | |
1182 | just to replace the redundant operation with the null op. Originally, | |
1183 | the tree would have looked like this: | |
1184 | ||
1185 | 10 SVOP (0x816b4f0) rv2sv [15] | |
1186 | 11 SVOP (0x816dcf0) gv GV (0x80fa460) *a | |
1187 | ||
1188 | That is, fetch the C<a> entry from the main symbol table, and then look | |
1189 | at the scalar component of it: C<gvsv> (C<pp_gvsv> into F<pp_hot.c>) | |
1190 | happens to do both these things. | |
1191 | ||
1192 | The right hand side, starting at line 5 is similar to what we've just | |
1193 | seen: we have the C<add> op (C<pp_add> also in F<pp_hot.c>) add together | |
1194 | two C<gvsv>s. | |
1195 | ||
1196 | Now, what's this about? | |
1197 | ||
1198 | 1 LISTOP (0x8179888) leave | |
1199 | 2 OP (0x81798b0) enter | |
1200 | 3 COP (0x8179850) nextstate | |
1201 | ||
1202 | C<enter> and C<leave> are scoping ops, and their job is to perform any | |
1203 | housekeeping every time you enter and leave a block: lexical variables | |
1204 | are tidied up, unreferenced variables are destroyed, and so on. Every | |
1205 | program will have those first three lines: C<leave> is a list, and its | |
1206 | children are all the statements in the block. Statements are delimited | |
1207 | by C<nextstate>, so a block is a collection of C<nextstate> ops, with | |
1208 | the ops to be performed for each statement being the children of | |
1209 | C<nextstate>. C<enter> is a single op which functions as a marker. | |
1210 | ||
1211 | That's how Perl parsed the program, from top to bottom: | |
1212 | ||
1213 | Program | |
1214 | | | |
1215 | Statement | |
1216 | | | |
1217 | = | |
1218 | / \ | |
1219 | / \ | |
1220 | $a + | |
1221 | / \ | |
1222 | $b $c | |
1223 | ||
1224 | However, it's impossible to B<perform> the operations in this order: | |
1225 | you have to find the values of C<$b> and C<$c> before you add them | |
1226 | together, for instance. So, the other thread that runs through the op | |
1227 | tree is the execution order: each op has a field C<op_next> which points | |
1228 | to the next op to be run, so following these pointers tells us how perl | |
1229 | executes the code. We can traverse the tree in this order using | |
1230 | the C<exec> option to C<B::Terse>: | |
1231 | ||
1232 | % perl -MO=Terse,exec -e '$a=$b+$c' | |
1233 | 1 OP (0x8179928) enter | |
1234 | 2 COP (0x81798c8) nextstate | |
1235 | 3 SVOP (0x81796c8) gvsv GV (0x80fa4d4) *b | |
1236 | 4 SVOP (0x8179798) gvsv GV (0x80efeb0) *c | |
1237 | 5 BINOP (0x8179878) add [1] | |
1238 | 6 SVOP (0x816dd38) gvsv GV (0x80fa468) *a | |
1239 | 7 BINOP (0x81798a0) sassign | |
1240 | 8 LISTOP (0x8179900) leave | |
1241 | ||
1242 | This probably makes more sense for a human: enter a block, start a | |
1243 | statement. Get the values of C<$b> and C<$c>, and add them together. | |
1244 | Find C<$a>, and assign one to the other. Then leave. | |
1245 | ||
1246 | The way Perl builds up these op trees in the parsing process can be | |
1247 | unravelled by examining F<perly.y>, the YACC grammar. Let's take the | |
1248 | piece we need to construct the tree for C<$a = $b + $c> | |
1249 | ||
1250 | 1 term : term ASSIGNOP term | |
1251 | 2 { $$ = newASSIGNOP(OPf_STACKED, $1, $2, $3); } | |
1252 | 3 | term ADDOP term | |
1253 | 4 { $$ = newBINOP($2, 0, scalar($1), scalar($3)); } | |
1254 | ||
1255 | If you're not used to reading BNF grammars, this is how it works: You're | |
1256 | fed certain things by the tokeniser, which generally end up in upper | |
1257 | case. Here, C<ADDOP>, is provided when the tokeniser sees C<+> in your | |
1258 | code. C<ASSIGNOP> is provided when C<=> is used for assigning. These are | |
b432a672 | 1259 | "terminal symbols", because you can't get any simpler than them. |
a422fd2d SC |
1260 | |
1261 | The grammar, lines one and three of the snippet above, tells you how to | |
b432a672 | 1262 | build up more complex forms. These complex forms, "non-terminal symbols" |
a422fd2d SC |
1263 | are generally placed in lower case. C<term> here is a non-terminal |
1264 | symbol, representing a single expression. | |
1265 | ||
1266 | The grammar gives you the following rule: you can make the thing on the | |
1267 | left of the colon if you see all the things on the right in sequence. | |
1268 | This is called a "reduction", and the aim of parsing is to completely | |
1269 | reduce the input. There are several different ways you can perform a | |
1270 | reduction, separated by vertical bars: so, C<term> followed by C<=> | |
1271 | followed by C<term> makes a C<term>, and C<term> followed by C<+> | |
1272 | followed by C<term> can also make a C<term>. | |
1273 | ||
1274 | So, if you see two terms with an C<=> or C<+>, between them, you can | |
1275 | turn them into a single expression. When you do this, you execute the | |
1276 | code in the block on the next line: if you see C<=>, you'll do the code | |
1277 | in line 2. If you see C<+>, you'll do the code in line 4. It's this code | |
1278 | which contributes to the op tree. | |
1279 | ||
1280 | | term ADDOP term | |
1281 | { $$ = newBINOP($2, 0, scalar($1), scalar($3)); } | |
1282 | ||
1283 | What this does is creates a new binary op, and feeds it a number of | |
1284 | variables. The variables refer to the tokens: C<$1> is the first token in | |
1285 | the input, C<$2> the second, and so on - think regular expression | |
1286 | backreferences. C<$$> is the op returned from this reduction. So, we | |
1287 | call C<newBINOP> to create a new binary operator. The first parameter to | |
1288 | C<newBINOP>, a function in F<op.c>, is the op type. It's an addition | |
1289 | operator, so we want the type to be C<ADDOP>. We could specify this | |
1290 | directly, but it's right there as the second token in the input, so we | |
b432a672 AL |
1291 | use C<$2>. The second parameter is the op's flags: 0 means "nothing |
1292 | special". Then the things to add: the left and right hand side of our | |
a422fd2d SC |
1293 | expression, in scalar context. |
1294 | ||
1295 | =head2 Stacks | |
1296 | ||
1297 | When perl executes something like C<addop>, how does it pass on its | |
1298 | results to the next op? The answer is, through the use of stacks. Perl | |
1299 | has a number of stacks to store things it's currently working on, and | |
1300 | we'll look at the three most important ones here. | |
1301 | ||
1302 | =over 3 | |
1303 | ||
1304 | =item Argument stack | |
1305 | ||
1306 | Arguments are passed to PP code and returned from PP code using the | |
1307 | argument stack, C<ST>. The typical way to handle arguments is to pop | |
1308 | them off the stack, deal with them how you wish, and then push the result | |
1309 | back onto the stack. This is how, for instance, the cosine operator | |
1310 | works: | |
1311 | ||
1312 | NV value; | |
1313 | value = POPn; | |
1314 | value = Perl_cos(value); | |
1315 | XPUSHn(value); | |
1316 | ||
1317 | We'll see a more tricky example of this when we consider Perl's macros | |
1318 | below. C<POPn> gives you the NV (floating point value) of the top SV on | |
1319 | the stack: the C<$x> in C<cos($x)>. Then we compute the cosine, and push | |
1320 | the result back as an NV. The C<X> in C<XPUSHn> means that the stack | |
1321 | should be extended if necessary - it can't be necessary here, because we | |
1322 | know there's room for one more item on the stack, since we've just | |
1323 | removed one! The C<XPUSH*> macros at least guarantee safety. | |
1324 | ||
1325 | Alternatively, you can fiddle with the stack directly: C<SP> gives you | |
1326 | the first element in your portion of the stack, and C<TOP*> gives you | |
1327 | the top SV/IV/NV/etc. on the stack. So, for instance, to do unary | |
1328 | negation of an integer: | |
1329 | ||
1330 | SETi(-TOPi); | |
1331 | ||
1332 | Just set the integer value of the top stack entry to its negation. | |
1333 | ||
1334 | Argument stack manipulation in the core is exactly the same as it is in | |
1335 | XSUBs - see L<perlxstut>, L<perlxs> and L<perlguts> for a longer | |
1336 | description of the macros used in stack manipulation. | |
1337 | ||
1338 | =item Mark stack | |
1339 | ||
b432a672 | 1340 | I say "your portion of the stack" above because PP code doesn't |
a422fd2d SC |
1341 | necessarily get the whole stack to itself: if your function calls |
1342 | another function, you'll only want to expose the arguments aimed for the | |
1343 | called function, and not (necessarily) let it get at your own data. The | |
b432a672 | 1344 | way we do this is to have a "virtual" bottom-of-stack, exposed to each |
a422fd2d SC |
1345 | function. The mark stack keeps bookmarks to locations in the argument |
1346 | stack usable by each function. For instance, when dealing with a tied | |
b432a672 | 1347 | variable, (internally, something with "P" magic) Perl has to call |
a422fd2d SC |
1348 | methods for accesses to the tied variables. However, we need to separate |
1349 | the arguments exposed to the method to the argument exposed to the | |
ed233832 DM |
1350 | original function - the store or fetch or whatever it may be. Here's |
1351 | roughly how the tied C<push> is implemented; see C<av_push> in F<av.c>: | |
a422fd2d SC |
1352 | |
1353 | 1 PUSHMARK(SP); | |
1354 | 2 EXTEND(SP,2); | |
1355 | 3 PUSHs(SvTIED_obj((SV*)av, mg)); | |
1356 | 4 PUSHs(val); | |
1357 | 5 PUTBACK; | |
1358 | 6 ENTER; | |
1359 | 7 call_method("PUSH", G_SCALAR|G_DISCARD); | |
1360 | 8 LEAVE; | |
13a2d996 | 1361 | |
a422fd2d SC |
1362 | Let's examine the whole implementation, for practice: |
1363 | ||
1364 | 1 PUSHMARK(SP); | |
1365 | ||
1366 | Push the current state of the stack pointer onto the mark stack. This is | |
1367 | so that when we've finished adding items to the argument stack, Perl | |
1368 | knows how many things we've added recently. | |
1369 | ||
1370 | 2 EXTEND(SP,2); | |
1371 | 3 PUSHs(SvTIED_obj((SV*)av, mg)); | |
1372 | 4 PUSHs(val); | |
1373 | ||
1374 | We're going to add two more items onto the argument stack: when you have | |
1375 | a tied array, the C<PUSH> subroutine receives the object and the value | |
1376 | to be pushed, and that's exactly what we have here - the tied object, | |
1377 | retrieved with C<SvTIED_obj>, and the value, the SV C<val>. | |
1378 | ||
1379 | 5 PUTBACK; | |
1380 | ||
e89a6d4e JD |
1381 | Next we tell Perl to update the global stack pointer from our internal |
1382 | variable: C<dSP> only gave us a local copy, not a reference to the global. | |
a422fd2d SC |
1383 | |
1384 | 6 ENTER; | |
1385 | 7 call_method("PUSH", G_SCALAR|G_DISCARD); | |
1386 | 8 LEAVE; | |
1387 | ||
1388 | C<ENTER> and C<LEAVE> localise a block of code - they make sure that all | |
1389 | variables are tidied up, everything that has been localised gets | |
1390 | its previous value returned, and so on. Think of them as the C<{> and | |
1391 | C<}> of a Perl block. | |
1392 | ||
1393 | To actually do the magic method call, we have to call a subroutine in | |
1394 | Perl space: C<call_method> takes care of that, and it's described in | |
1395 | L<perlcall>. We call the C<PUSH> method in scalar context, and we're | |
e89a6d4e JD |
1396 | going to discard its return value. The call_method() function |
1397 | removes the top element of the mark stack, so there is nothing for | |
1398 | the caller to clean up. | |
a422fd2d | 1399 | |
a422fd2d SC |
1400 | =item Save stack |
1401 | ||
1402 | C doesn't have a concept of local scope, so perl provides one. We've | |
1403 | seen that C<ENTER> and C<LEAVE> are used as scoping braces; the save | |
1404 | stack implements the C equivalent of, for example: | |
1405 | ||
1406 | { | |
1407 | local $foo = 42; | |
1408 | ... | |
1409 | } | |
1410 | ||
1411 | See L<perlguts/Localising Changes> for how to use the save stack. | |
1412 | ||
1413 | =back | |
1414 | ||
1415 | =head2 Millions of Macros | |
1416 | ||
1417 | One thing you'll notice about the Perl source is that it's full of | |
1418 | macros. Some have called the pervasive use of macros the hardest thing | |
1419 | to understand, others find it adds to clarity. Let's take an example, | |
1420 | the code which implements the addition operator: | |
1421 | ||
1422 | 1 PP(pp_add) | |
1423 | 2 { | |
39644a26 | 1424 | 3 dSP; dATARGET; tryAMAGICbin(add,opASSIGN); |
a422fd2d SC |
1425 | 4 { |
1426 | 5 dPOPTOPnnrl_ul; | |
1427 | 6 SETn( left + right ); | |
1428 | 7 RETURN; | |
1429 | 8 } | |
1430 | 9 } | |
1431 | ||
1432 | Every line here (apart from the braces, of course) contains a macro. The | |
1433 | first line sets up the function declaration as Perl expects for PP code; | |
1434 | line 3 sets up variable declarations for the argument stack and the | |
1435 | target, the return value of the operation. Finally, it tries to see if | |
1436 | the addition operation is overloaded; if so, the appropriate subroutine | |
1437 | is called. | |
1438 | ||
1439 | Line 5 is another variable declaration - all variable declarations start | |
1440 | with C<d> - which pops from the top of the argument stack two NVs (hence | |
1441 | C<nn>) and puts them into the variables C<right> and C<left>, hence the | |
1442 | C<rl>. These are the two operands to the addition operator. Next, we | |
1443 | call C<SETn> to set the NV of the return value to the result of adding | |
1444 | the two values. This done, we return - the C<RETURN> macro makes sure | |
1445 | that our return value is properly handled, and we pass the next operator | |
1446 | to run back to the main run loop. | |
1447 | ||
1448 | Most of these macros are explained in L<perlapi>, and some of the more | |
1449 | important ones are explained in L<perlxs> as well. Pay special attention | |
1450 | to L<perlguts/Background and PERL_IMPLICIT_CONTEXT> for information on | |
1451 | the C<[pad]THX_?> macros. | |
1452 | ||
52d59bef JH |
1453 | =head2 The .i Targets |
1454 | ||
1455 | You can expand the macros in a F<foo.c> file by saying | |
1456 | ||
1457 | make foo.i | |
1458 | ||
1459 | which will expand the macros using cpp. Don't be scared by the results. | |
1460 | ||
955fec6b JH |
1461 | =head1 SOURCE CODE STATIC ANALYSIS |
1462 | ||
1463 | Various tools exist for analysing C source code B<statically>, as | |
1464 | opposed to B<dynamically>, that is, without executing the code. | |
1465 | It is possible to detect resource leaks, undefined behaviour, type | |
1466 | mismatches, portability problems, code paths that would cause illegal | |
1467 | memory accesses, and other similar problems by just parsing the C code | |
1468 | and looking at the resulting graph, what does it tell about the | |
1469 | execution and data flows. As a matter of fact, this is exactly | |
1470 | how C compilers know to give warnings about dubious code. | |
1471 | ||
1472 | =head2 lint, splint | |
1473 | ||
1474 | The good old C code quality inspector, C<lint>, is available in | |
1475 | several platforms, but please be aware that there are several | |
1476 | different implementations of it by different vendors, which means that | |
1477 | the flags are not identical across different platforms. | |
1478 | ||
1479 | There is a lint variant called C<splint> (Secure Programming Lint) | |
1480 | available from http://www.splint.org/ that should compile on any | |
1481 | Unix-like platform. | |
1482 | ||
1483 | There are C<lint> and <splint> targets in Makefile, but you may have | |
1484 | to diddle with the flags (see above). | |
1485 | ||
1486 | =head2 Coverity | |
1487 | ||
1488 | Coverity (http://www.coverity.com/) is a product similar to lint and | |
1489 | as a testbed for their product they periodically check several open | |
1490 | source projects, and they give out accounts to open source developers | |
1491 | to the defect databases. | |
1492 | ||
1493 | =head2 cpd (cut-and-paste detector) | |
1494 | ||
1495 | The cpd tool detects cut-and-paste coding. If one instance of the | |
1496 | cut-and-pasted code changes, all the other spots should probably be | |
1497 | changed, too. Therefore such code should probably be turned into a | |
1498 | subroutine or a macro. | |
1499 | ||
1500 | cpd (http://pmd.sourceforge.net/cpd.html) is part of the pmd project | |
1501 | (http://pmd.sourceforge.net/). pmd was originally written for static | |
1502 | analysis of Java code, but later the cpd part of it was extended to | |
1503 | parse also C and C++. | |
1504 | ||
1505 | Download the pmd-X.y.jar from the SourceForge site, and then run | |
1506 | it on source code thusly: | |
1507 | ||
1508 | java -cp pmd-X.Y.jar net.sourceforge.pmd.cpd.CPD --minimum-tokens 100 --files /some/where/src --language c > cpd.txt | |
1509 | ||
1510 | You may run into memory limits, in which case you should use the -Xmx option: | |
1511 | ||
1512 | java -Xmx512M ... | |
1513 | ||
1514 | =head2 gcc warnings | |
1515 | ||
1516 | Though much can be written about the inconsistency and coverage | |
1517 | problems of gcc warnings (like C<-Wall> not meaning "all the | |
1518 | warnings", or some common portability problems not being covered by | |
1519 | C<-Wall>, or C<-ansi> and C<-pedantic> both being a poorly defined | |
1520 | collection of warnings, and so forth), gcc is still a useful tool in | |
1521 | keeping our coding nose clean. | |
1522 | ||
1523 | The C<-Wall> is by default on. | |
1524 | ||
a8e98a71 JH |
1525 | The C<-ansi> (and its sidekick, C<-pedantic>) would be nice to be on |
1526 | always, but unfortunately they are not safe on all platforms, they can | |
1527 | for example cause fatal conflicts with the system headers (Solaris | |
1528 | being a prime example). If Configure C<-Dgccansipedantic> is used, | |
1529 | the C<cflags> frontend selects C<-ansi -pedantic> for the platforms | |
1530 | where they are known to be safe. | |
955fec6b JH |
1531 | |
1532 | Starting from Perl 5.9.4 the following extra flags are added: | |
1533 | ||
1534 | =over 4 | |
1535 | ||
1536 | =item * | |
1537 | ||
1538 | C<-Wendif-labels> | |
1539 | ||
1540 | =item * | |
1541 | ||
1542 | C<-Wextra> | |
1543 | ||
1544 | =item * | |
1545 | ||
1546 | C<-Wdeclaration-after-statement> | |
1547 | ||
1548 | =back | |
1549 | ||
1550 | The following flags would be nice to have but they would first need | |
1551 | their own Stygian stablemaster: | |
1552 | ||
1553 | =over 4 | |
1554 | ||
1555 | =item * | |
1556 | ||
1557 | C<-Wpointer-arith> | |
1558 | ||
1559 | =item * | |
1560 | ||
1561 | C<-Wshadow> | |
1562 | ||
1563 | =item * | |
1564 | ||
1565 | C<-Wstrict-prototypes> | |
1566 | ||
955fec6b JH |
1567 | =back |
1568 | ||
1569 | The C<-Wtraditional> is another example of the annoying tendency of | |
1570 | gcc to bundle a lot of warnings under one switch -- it would be | |
1571 | impossible to deploy in practice because it would complain a lot -- but | |
1572 | it does contain some warnings that would be beneficial to have available | |
1573 | on their own, such as the warning about string constants inside macros | |
1574 | containing the macro arguments: this behaved differently pre-ANSI | |
1575 | than it does in ANSI, and some C compilers are still in transition, | |
1576 | AIX being an example. | |
1577 | ||
1578 | =head2 Warnings of other C compilers | |
1579 | ||
1580 | Other C compilers (yes, there B<are> other C compilers than gcc) often | |
1581 | have their "strict ANSI" or "strict ANSI with some portability extensions" | |
1582 | modes on, like for example the Sun Workshop has its C<-Xa> mode on | |
1583 | (though implicitly), or the DEC (these days, HP...) has its C<-std1> | |
1584 | mode on. | |
1585 | ||
1586 | =head2 DEBUGGING | |
1587 | ||
1588 | You can compile a special debugging version of Perl, which allows you | |
1589 | to use the C<-D> option of Perl to tell more about what Perl is doing. | |
1590 | But sometimes there is no alternative than to dive in with a debugger, | |
1591 | either to see the stack trace of a core dump (very useful in a bug | |
1592 | report), or trying to figure out what went wrong before the core dump | |
1593 | happened, or how did we end up having wrong or unexpected results. | |
1594 | ||
a422fd2d SC |
1595 | =head2 Poking at Perl |
1596 | ||
1597 | To really poke around with Perl, you'll probably want to build Perl for | |
1598 | debugging, like this: | |
1599 | ||
1600 | ./Configure -d -D optimize=-g | |
1601 | make | |
1602 | ||
1603 | C<-g> is a flag to the C compiler to have it produce debugging | |
955fec6b JH |
1604 | information which will allow us to step through a running program, |
1605 | and to see in which C function we are at (without the debugging | |
1606 | information we might see only the numerical addresses of the functions, | |
1607 | which is not very helpful). | |
1608 | ||
a422fd2d SC |
1609 | F<Configure> will also turn on the C<DEBUGGING> compilation symbol which |
1610 | enables all the internal debugging code in Perl. There are a whole bunch | |
1611 | of things you can debug with this: L<perlrun> lists them all, and the | |
1612 | best way to find out about them is to play about with them. The most | |
1613 | useful options are probably | |
1614 | ||
1615 | l Context (loop) stack processing | |
1616 | t Trace execution | |
1617 | o Method and overloading resolution | |
1618 | c String/numeric conversions | |
1619 | ||
1620 | Some of the functionality of the debugging code can be achieved using XS | |
1621 | modules. | |
13a2d996 | 1622 | |
a422fd2d SC |
1623 | -Dr => use re 'debug' |
1624 | -Dx => use O 'Debug' | |
1625 | ||
1626 | =head2 Using a source-level debugger | |
1627 | ||
1628 | If the debugging output of C<-D> doesn't help you, it's time to step | |
1629 | through perl's execution with a source-level debugger. | |
1630 | ||
1631 | =over 3 | |
1632 | ||
1633 | =item * | |
1634 | ||
955fec6b JH |
1635 | We'll use C<gdb> for our examples here; the principles will apply to |
1636 | any debugger (many vendors call their debugger C<dbx>), but check the | |
1637 | manual of the one you're using. | |
a422fd2d SC |
1638 | |
1639 | =back | |
1640 | ||
1641 | To fire up the debugger, type | |
1642 | ||
1643 | gdb ./perl | |
1644 | ||
955fec6b JH |
1645 | Or if you have a core dump: |
1646 | ||
1647 | gdb ./perl core | |
1648 | ||
a422fd2d SC |
1649 | You'll want to do that in your Perl source tree so the debugger can read |
1650 | the source code. You should see the copyright message, followed by the | |
1651 | prompt. | |
1652 | ||
1653 | (gdb) | |
1654 | ||
1655 | C<help> will get you into the documentation, but here are the most | |
1656 | useful commands: | |
1657 | ||
1658 | =over 3 | |
1659 | ||
1660 | =item run [args] | |
1661 | ||
1662 | Run the program with the given arguments. | |
1663 | ||
1664 | =item break function_name | |
1665 | ||
1666 | =item break source.c:xxx | |
1667 | ||
1668 | Tells the debugger that we'll want to pause execution when we reach | |
cea6626f | 1669 | either the named function (but see L<perlguts/Internal Functions>!) or the given |
a422fd2d SC |
1670 | line in the named source file. |
1671 | ||
1672 | =item step | |
1673 | ||
1674 | Steps through the program a line at a time. | |
1675 | ||
1676 | =item next | |
1677 | ||
1678 | Steps through the program a line at a time, without descending into | |
1679 | functions. | |
1680 | ||
1681 | =item continue | |
1682 | ||
1683 | Run until the next breakpoint. | |
1684 | ||
1685 | =item finish | |
1686 | ||
1687 | Run until the end of the current function, then stop again. | |
1688 | ||
13a2d996 | 1689 | =item 'enter' |
a422fd2d SC |
1690 | |
1691 | Just pressing Enter will do the most recent operation again - it's a | |
1692 | blessing when stepping through miles of source code. | |
1693 | ||
1694 | =item print | |
1695 | ||
1696 | Execute the given C code and print its results. B<WARNING>: Perl makes | |
52d59bef JH |
1697 | heavy use of macros, and F<gdb> does not necessarily support macros |
1698 | (see later L</"gdb macro support">). You'll have to substitute them | |
1699 | yourself, or to invoke cpp on the source code files | |
1700 | (see L</"The .i Targets">) | |
1701 | So, for instance, you can't say | |
a422fd2d SC |
1702 | |
1703 | print SvPV_nolen(sv) | |
1704 | ||
1705 | but you have to say | |
1706 | ||
1707 | print Perl_sv_2pv_nolen(sv) | |
1708 | ||
ffc145e8 RK |
1709 | =back |
1710 | ||
a422fd2d SC |
1711 | You may find it helpful to have a "macro dictionary", which you can |
1712 | produce by saying C<cpp -dM perl.c | sort>. Even then, F<cpp> won't | |
52d59bef JH |
1713 | recursively apply those macros for you. |
1714 | ||
1715 | =head2 gdb macro support | |
a422fd2d | 1716 | |
52d59bef | 1717 | Recent versions of F<gdb> have fairly good macro support, but |
ea031e66 RGS |
1718 | in order to use it you'll need to compile perl with macro definitions |
1719 | included in the debugging information. Using F<gcc> version 3.1, this | |
1720 | means configuring with C<-Doptimize=-g3>. Other compilers might use a | |
1721 | different switch (if they support debugging macros at all). | |
1722 | ||
a422fd2d SC |
1723 | =head2 Dumping Perl Data Structures |
1724 | ||
1725 | One way to get around this macro hell is to use the dumping functions in | |
1726 | F<dump.c>; these work a little like an internal | |
1727 | L<Devel::Peek|Devel::Peek>, but they also cover OPs and other structures | |
1728 | that you can't get at from Perl. Let's take an example. We'll use the | |
1729 | C<$a = $b + $c> we used before, but give it a bit of context: | |
1730 | C<$b = "6XXXX"; $c = 2.3;>. Where's a good place to stop and poke around? | |
1731 | ||
1732 | What about C<pp_add>, the function we examined earlier to implement the | |
1733 | C<+> operator: | |
1734 | ||
1735 | (gdb) break Perl_pp_add | |
1736 | Breakpoint 1 at 0x46249f: file pp_hot.c, line 309. | |
1737 | ||
cea6626f | 1738 | Notice we use C<Perl_pp_add> and not C<pp_add> - see L<perlguts/Internal Functions>. |
a422fd2d SC |
1739 | With the breakpoint in place, we can run our program: |
1740 | ||
1741 | (gdb) run -e '$b = "6XXXX"; $c = 2.3; $a = $b + $c' | |
1742 | ||
1743 | Lots of junk will go past as gdb reads in the relevant source files and | |
1744 | libraries, and then: | |
1745 | ||
1746 | Breakpoint 1, Perl_pp_add () at pp_hot.c:309 | |
39644a26 | 1747 | 309 dSP; dATARGET; tryAMAGICbin(add,opASSIGN); |
a422fd2d SC |
1748 | (gdb) step |
1749 | 311 dPOPTOPnnrl_ul; | |
1750 | (gdb) | |
1751 | ||
1752 | We looked at this bit of code before, and we said that C<dPOPTOPnnrl_ul> | |
1753 | arranges for two C<NV>s to be placed into C<left> and C<right> - let's | |
1754 | slightly expand it: | |
1755 | ||
1756 | #define dPOPTOPnnrl_ul NV right = POPn; \ | |
1757 | SV *leftsv = TOPs; \ | |
1758 | NV left = USE_LEFT(leftsv) ? SvNV(leftsv) : 0.0 | |
1759 | ||
1760 | C<POPn> takes the SV from the top of the stack and obtains its NV either | |
1761 | directly (if C<SvNOK> is set) or by calling the C<sv_2nv> function. | |
1762 | C<TOPs> takes the next SV from the top of the stack - yes, C<POPn> uses | |
1763 | C<TOPs> - but doesn't remove it. We then use C<SvNV> to get the NV from | |
1764 | C<leftsv> in the same way as before - yes, C<POPn> uses C<SvNV>. | |
1765 | ||
1766 | Since we don't have an NV for C<$b>, we'll have to use C<sv_2nv> to | |
1767 | convert it. If we step again, we'll find ourselves there: | |
1768 | ||
1769 | Perl_sv_2nv (sv=0xa0675d0) at sv.c:1669 | |
1770 | 1669 if (!sv) | |
1771 | (gdb) | |
1772 | ||
1773 | We can now use C<Perl_sv_dump> to investigate the SV: | |
1774 | ||
1775 | SV = PV(0xa057cc0) at 0xa0675d0 | |
1776 | REFCNT = 1 | |
1777 | FLAGS = (POK,pPOK) | |
1778 | PV = 0xa06a510 "6XXXX"\0 | |
1779 | CUR = 5 | |
1780 | LEN = 6 | |
1781 | $1 = void | |
1782 | ||
1783 | We know we're going to get C<6> from this, so let's finish the | |
1784 | subroutine: | |
1785 | ||
1786 | (gdb) finish | |
1787 | Run till exit from #0 Perl_sv_2nv (sv=0xa0675d0) at sv.c:1671 | |
1788 | 0x462669 in Perl_pp_add () at pp_hot.c:311 | |
1789 | 311 dPOPTOPnnrl_ul; | |
1790 | ||
1791 | We can also dump out this op: the current op is always stored in | |
1792 | C<PL_op>, and we can dump it with C<Perl_op_dump>. This'll give us | |
1793 | similar output to L<B::Debug|B::Debug>. | |
1794 | ||
1795 | { | |
1796 | 13 TYPE = add ===> 14 | |
1797 | TARG = 1 | |
1798 | FLAGS = (SCALAR,KIDS) | |
1799 | { | |
1800 | TYPE = null ===> (12) | |
1801 | (was rv2sv) | |
1802 | FLAGS = (SCALAR,KIDS) | |
1803 | { | |
1804 | 11 TYPE = gvsv ===> 12 | |
1805 | FLAGS = (SCALAR) | |
1806 | GV = main::b | |
1807 | } | |
1808 | } | |
1809 | ||
10f58044 | 1810 | # finish this later # |
a422fd2d SC |
1811 | |
1812 | =head2 Patching | |
1813 | ||
1814 | All right, we've now had a look at how to navigate the Perl sources and | |
1815 | some things you'll need to know when fiddling with them. Let's now get | |
1816 | on and create a simple patch. Here's something Larry suggested: if a | |
1817 | C<U> is the first active format during a C<pack>, (for example, | |
1818 | C<pack "U3C8", @stuff>) then the resulting string should be treated as | |
1e54db1a | 1819 | UTF-8 encoded. |
a422fd2d SC |
1820 | |
1821 | How do we prepare to fix this up? First we locate the code in question - | |
1822 | the C<pack> happens at runtime, so it's going to be in one of the F<pp> | |
1823 | files. Sure enough, C<pp_pack> is in F<pp.c>. Since we're going to be | |
1824 | altering this file, let's copy it to F<pp.c~>. | |
1825 | ||
a6ec74c1 JH |
1826 | [Well, it was in F<pp.c> when this tutorial was written. It has now been |
1827 | split off with C<pp_unpack> to its own file, F<pp_pack.c>] | |
1828 | ||
a422fd2d SC |
1829 | Now let's look over C<pp_pack>: we take a pattern into C<pat>, and then |
1830 | loop over the pattern, taking each format character in turn into | |
1831 | C<datum_type>. Then for each possible format character, we swallow up | |
1832 | the other arguments in the pattern (a field width, an asterisk, and so | |
1833 | on) and convert the next chunk input into the specified format, adding | |
1834 | it onto the output SV C<cat>. | |
1835 | ||
1836 | How do we know if the C<U> is the first format in the C<pat>? Well, if | |
1837 | we have a pointer to the start of C<pat> then, if we see a C<U> we can | |
1838 | test whether we're still at the start of the string. So, here's where | |
1839 | C<pat> is set up: | |
1840 | ||
1841 | STRLEN fromlen; | |
1842 | register char *pat = SvPVx(*++MARK, fromlen); | |
1843 | register char *patend = pat + fromlen; | |
1844 | register I32 len; | |
1845 | I32 datumtype; | |
1846 | SV *fromstr; | |
1847 | ||
1848 | We'll have another string pointer in there: | |
1849 | ||
1850 | STRLEN fromlen; | |
1851 | register char *pat = SvPVx(*++MARK, fromlen); | |
1852 | register char *patend = pat + fromlen; | |
1853 | + char *patcopy; | |
1854 | register I32 len; | |
1855 | I32 datumtype; | |
1856 | SV *fromstr; | |
1857 | ||
1858 | And just before we start the loop, we'll set C<patcopy> to be the start | |
1859 | of C<pat>: | |
1860 | ||
1861 | items = SP - MARK; | |
1862 | MARK++; | |
1863 | sv_setpvn(cat, "", 0); | |
1864 | + patcopy = pat; | |
1865 | while (pat < patend) { | |
1866 | ||
1867 | Now if we see a C<U> which was at the start of the string, we turn on | |
1e54db1a | 1868 | the C<UTF8> flag for the output SV, C<cat>: |
a422fd2d SC |
1869 | |
1870 | + if (datumtype == 'U' && pat==patcopy+1) | |
1871 | + SvUTF8_on(cat); | |
1872 | if (datumtype == '#') { | |
1873 | while (pat < patend && *pat != '\n') | |
1874 | pat++; | |
1875 | ||
1876 | Remember that it has to be C<patcopy+1> because the first character of | |
1877 | the string is the C<U> which has been swallowed into C<datumtype!> | |
1878 | ||
1879 | Oops, we forgot one thing: what if there are spaces at the start of the | |
1880 | pattern? C<pack(" U*", @stuff)> will have C<U> as the first active | |
1881 | character, even though it's not the first thing in the pattern. In this | |
1882 | case, we have to advance C<patcopy> along with C<pat> when we see spaces: | |
1883 | ||
1884 | if (isSPACE(datumtype)) | |
1885 | continue; | |
1886 | ||
1887 | needs to become | |
1888 | ||
1889 | if (isSPACE(datumtype)) { | |
1890 | patcopy++; | |
1891 | continue; | |
1892 | } | |
1893 | ||
1894 | OK. That's the C part done. Now we must do two additional things before | |
1895 | this patch is ready to go: we've changed the behaviour of Perl, and so | |
1896 | we must document that change. We must also provide some more regression | |
1897 | tests to make sure our patch works and doesn't create a bug somewhere | |
1898 | else along the line. | |
1899 | ||
b23b8711 MS |
1900 | The regression tests for each operator live in F<t/op/>, and so we |
1901 | make a copy of F<t/op/pack.t> to F<t/op/pack.t~>. Now we can add our | |
1902 | tests to the end. First, we'll test that the C<U> does indeed create | |
1903 | Unicode strings. | |
1904 | ||
1905 | t/op/pack.t has a sensible ok() function, but if it didn't we could | |
35c336e6 | 1906 | use the one from t/test.pl. |
b23b8711 | 1907 | |
35c336e6 MS |
1908 | require './test.pl'; |
1909 | plan( tests => 159 ); | |
b23b8711 MS |
1910 | |
1911 | so instead of this: | |
a422fd2d SC |
1912 | |
1913 | print 'not ' unless "1.20.300.4000" eq sprintf "%vd", pack("U*",1,20,300,4000); | |
1914 | print "ok $test\n"; $test++; | |
1915 | ||
35c336e6 MS |
1916 | we can write the more sensible (see L<Test::More> for a full |
1917 | explanation of is() and other testing functions). | |
b23b8711 | 1918 | |
35c336e6 | 1919 | is( "1.20.300.4000", sprintf "%vd", pack("U*",1,20,300,4000), |
812f5127 | 1920 | "U* produces unicode" ); |
b23b8711 | 1921 | |
a422fd2d SC |
1922 | Now we'll test that we got that space-at-the-beginning business right: |
1923 | ||
35c336e6 | 1924 | is( "1.20.300.4000", sprintf "%vd", pack(" U*",1,20,300,4000), |
812f5127 | 1925 | " with spaces at the beginning" ); |
a422fd2d SC |
1926 | |
1927 | And finally we'll test that we don't make Unicode strings if C<U> is B<not> | |
1928 | the first active format: | |
1929 | ||
35c336e6 | 1930 | isnt( v1.20.300.4000, sprintf "%vd", pack("C0U*",1,20,300,4000), |
812f5127 | 1931 | "U* not first isn't unicode" ); |
a422fd2d | 1932 | |
35c336e6 MS |
1933 | Mustn't forget to change the number of tests which appears at the top, |
1934 | or else the automated tester will get confused. This will either look | |
1935 | like this: | |
a422fd2d | 1936 | |
35c336e6 MS |
1937 | print "1..156\n"; |
1938 | ||
1939 | or this: | |
1940 | ||
1941 | plan( tests => 156 ); | |
a422fd2d SC |
1942 | |
1943 | We now compile up Perl, and run it through the test suite. Our new | |
1944 | tests pass, hooray! | |
1945 | ||
1946 | Finally, the documentation. The job is never done until the paperwork is | |
1947 | over, so let's describe the change we've just made. The relevant place | |
1948 | is F<pod/perlfunc.pod>; again, we make a copy, and then we'll insert | |
1949 | this text in the description of C<pack>: | |
1950 | ||
1951 | =item * | |
1952 | ||
1953 | If the pattern begins with a C<U>, the resulting string will be treated | |
1e54db1a JH |
1954 | as UTF-8-encoded Unicode. You can force UTF-8 encoding on in a string |
1955 | with an initial C<U0>, and the bytes that follow will be interpreted as | |
1956 | Unicode characters. If you don't want this to happen, you can begin your | |
1957 | pattern with C<C0> (or anything else) to force Perl not to UTF-8 encode your | |
a422fd2d SC |
1958 | string, and then follow this with a C<U*> somewhere in your pattern. |
1959 | ||
1960 | All done. Now let's create the patch. F<Porting/patching.pod> tells us | |
1961 | that if we're making major changes, we should copy the entire directory | |
1962 | to somewhere safe before we begin fiddling, and then do | |
13a2d996 | 1963 | |
a422fd2d SC |
1964 | diff -ruN old new > patch |
1965 | ||
1966 | However, we know which files we've changed, and we can simply do this: | |
1967 | ||
1968 | diff -u pp.c~ pp.c > patch | |
1969 | diff -u t/op/pack.t~ t/op/pack.t >> patch | |
1970 | diff -u pod/perlfunc.pod~ pod/perlfunc.pod >> patch | |
1971 | ||
1972 | We end up with a patch looking a little like this: | |
1973 | ||
1974 | --- pp.c~ Fri Jun 02 04:34:10 2000 | |
1975 | +++ pp.c Fri Jun 16 11:37:25 2000 | |
1976 | @@ -4375,6 +4375,7 @@ | |
1977 | register I32 items; | |
1978 | STRLEN fromlen; | |
1979 | register char *pat = SvPVx(*++MARK, fromlen); | |
1980 | + char *patcopy; | |
1981 | register char *patend = pat + fromlen; | |
1982 | register I32 len; | |
1983 | I32 datumtype; | |
1984 | @@ -4405,6 +4406,7 @@ | |
1985 | ... | |
1986 | ||
1987 | And finally, we submit it, with our rationale, to perl5-porters. Job | |
1988 | done! | |
1989 | ||
f7e1e956 MS |
1990 | =head2 Patching a core module |
1991 | ||
1992 | This works just like patching anything else, with an extra | |
1993 | consideration. Many core modules also live on CPAN. If this is so, | |
1994 | patch the CPAN version instead of the core and send the patch off to | |
1995 | the module maintainer (with a copy to p5p). This will help the module | |
1996 | maintainer keep the CPAN version in sync with the core version without | |
1997 | constantly scanning p5p. | |
1998 | ||
db300100 RGS |
1999 | The list of maintainers of core modules is usefully documented in |
2000 | F<Porting/Maintainers.pl>. | |
2001 | ||
acbe17fc JP |
2002 | =head2 Adding a new function to the core |
2003 | ||
2004 | If, as part of a patch to fix a bug, or just because you have an | |
2005 | especially good idea, you decide to add a new function to the core, | |
2006 | discuss your ideas on p5p well before you start work. It may be that | |
2007 | someone else has already attempted to do what you are considering and | |
2008 | can give lots of good advice or even provide you with bits of code | |
2009 | that they already started (but never finished). | |
2010 | ||
2011 | You have to follow all of the advice given above for patching. It is | |
2012 | extremely important to test any addition thoroughly and add new tests | |
2013 | to explore all boundary conditions that your new function is expected | |
2014 | to handle. If your new function is used only by one module (e.g. toke), | |
2015 | then it should probably be named S_your_function (for static); on the | |
210b36aa | 2016 | other hand, if you expect it to accessible from other functions in |
acbe17fc JP |
2017 | Perl, you should name it Perl_your_function. See L<perlguts/Internal Functions> |
2018 | for more details. | |
2019 | ||
2020 | The location of any new code is also an important consideration. Don't | |
2021 | just create a new top level .c file and put your code there; you would | |
2022 | have to make changes to Configure (so the Makefile is created properly), | |
2023 | as well as possibly lots of include files. This is strictly pumpking | |
2024 | business. | |
2025 | ||
2026 | It is better to add your function to one of the existing top level | |
2027 | source code files, but your choice is complicated by the nature of | |
2028 | the Perl distribution. Only the files that are marked as compiled | |
2029 | static are located in the perl executable. Everything else is located | |
2030 | in the shared library (or DLL if you are running under WIN32). So, | |
2031 | for example, if a function was only used by functions located in | |
2032 | toke.c, then your code can go in toke.c. If, however, you want to call | |
2033 | the function from universal.c, then you should put your code in another | |
2034 | location, for example util.c. | |
2035 | ||
2036 | In addition to writing your c-code, you will need to create an | |
2037 | appropriate entry in embed.pl describing your function, then run | |
2038 | 'make regen_headers' to create the entries in the numerous header | |
2039 | files that perl needs to compile correctly. See L<perlguts/Internal Functions> | |
2040 | for information on the various options that you can set in embed.pl. | |
2041 | You will forget to do this a few (or many) times and you will get | |
2042 | warnings during the compilation phase. Make sure that you mention | |
2043 | this when you post your patch to P5P; the pumpking needs to know this. | |
2044 | ||
2045 | When you write your new code, please be conscious of existing code | |
884bad00 | 2046 | conventions used in the perl source files. See L<perlstyle> for |
acbe17fc JP |
2047 | details. Although most of the guidelines discussed seem to focus on |
2048 | Perl code, rather than c, they all apply (except when they don't ;). | |
2049 | See also I<Porting/patching.pod> file in the Perl source distribution | |
2050 | for lots of details about both formatting and submitting patches of | |
2051 | your changes. | |
2052 | ||
2053 | Lastly, TEST TEST TEST TEST TEST any code before posting to p5p. | |
2054 | Test on as many platforms as you can find. Test as many perl | |
2055 | Configure options as you can (e.g. MULTIPLICITY). If you have | |
2056 | profiling or memory tools, see L<EXTERNAL TOOLS FOR DEBUGGING PERL> | |
210b36aa | 2057 | below for how to use them to further test your code. Remember that |
acbe17fc JP |
2058 | most of the people on P5P are doing this on their own time and |
2059 | don't have the time to debug your code. | |
f7e1e956 MS |
2060 | |
2061 | =head2 Writing a test | |
2062 | ||
2063 | Every module and built-in function has an associated test file (or | |
2064 | should...). If you add or change functionality, you have to write a | |
2065 | test. If you fix a bug, you have to write a test so that bug never | |
2066 | comes back. If you alter the docs, it would be nice to test what the | |
2067 | new documentation says. | |
2068 | ||
2069 | In short, if you submit a patch you probably also have to patch the | |
2070 | tests. | |
2071 | ||
2072 | For modules, the test file is right next to the module itself. | |
2073 | F<lib/strict.t> tests F<lib/strict.pm>. This is a recent innovation, | |
2074 | so there are some snags (and it would be wonderful for you to brush | |
2075 | them out), but it basically works that way. Everything else lives in | |
2076 | F<t/>. | |
2077 | ||
2078 | =over 3 | |
2079 | ||
2080 | =item F<t/base/> | |
2081 | ||
2082 | Testing of the absolute basic functionality of Perl. Things like | |
2083 | C<if>, basic file reads and writes, simple regexes, etc. These are | |
2084 | run first in the test suite and if any of them fail, something is | |
2085 | I<really> broken. | |
2086 | ||
2087 | =item F<t/cmd/> | |
2088 | ||
2089 | These test the basic control structures, C<if/else>, C<while>, | |
35c336e6 | 2090 | subroutines, etc. |
f7e1e956 MS |
2091 | |
2092 | =item F<t/comp/> | |
2093 | ||
2094 | Tests basic issues of how Perl parses and compiles itself. | |
2095 | ||
2096 | =item F<t/io/> | |
2097 | ||
2098 | Tests for built-in IO functions, including command line arguments. | |
2099 | ||
2100 | =item F<t/lib/> | |
2101 | ||
2102 | The old home for the module tests, you shouldn't put anything new in | |
2103 | here. There are still some bits and pieces hanging around in here | |
2104 | that need to be moved. Perhaps you could move them? Thanks! | |
2105 | ||
2106 | =item F<t/op/> | |
2107 | ||
2108 | Tests for perl's built in functions that don't fit into any of the | |
2109 | other directories. | |
2110 | ||
2111 | =item F<t/pod/> | |
2112 | ||
2113 | Tests for POD directives. There are still some tests for the Pod | |
2114 | modules hanging around in here that need to be moved out into F<lib/>. | |
2115 | ||
2116 | =item F<t/run/> | |
2117 | ||
2118 | Testing features of how perl actually runs, including exit codes and | |
2119 | handling of PERL* environment variables. | |
2120 | ||
244d9cb7 RGS |
2121 | =item F<t/uni/> |
2122 | ||
2123 | Tests for the core support of Unicode. | |
2124 | ||
2125 | =item F<t/win32/> | |
2126 | ||
2127 | Windows-specific tests. | |
2128 | ||
2129 | =item F<t/x2p> | |
2130 | ||
2131 | A test suite for the s2p converter. | |
2132 | ||
f7e1e956 MS |
2133 | =back |
2134 | ||
2135 | The core uses the same testing style as the rest of Perl, a simple | |
2136 | "ok/not ok" run through Test::Harness, but there are a few special | |
2137 | considerations. | |
2138 | ||
35c336e6 MS |
2139 | There are three ways to write a test in the core. Test::More, |
2140 | t/test.pl and ad hoc C<print $test ? "ok 42\n" : "not ok 42\n">. The | |
2141 | decision of which to use depends on what part of the test suite you're | |
2142 | working on. This is a measure to prevent a high-level failure (such | |
2143 | as Config.pm breaking) from causing basic functionality tests to fail. | |
2144 | ||
2145 | =over 4 | |
2146 | ||
2147 | =item t/base t/comp | |
2148 | ||
2149 | Since we don't know if require works, or even subroutines, use ad hoc | |
2150 | tests for these two. Step carefully to avoid using the feature being | |
2151 | tested. | |
2152 | ||
2153 | =item t/cmd t/run t/io t/op | |
2154 | ||
2155 | Now that basic require() and subroutines are tested, you can use the | |
2156 | t/test.pl library which emulates the important features of Test::More | |
2157 | while using a minimum of core features. | |
2158 | ||
2159 | You can also conditionally use certain libraries like Config, but be | |
2160 | sure to skip the test gracefully if it's not there. | |
2161 | ||
2162 | =item t/lib ext lib | |
2163 | ||
2164 | Now that the core of Perl is tested, Test::More can be used. You can | |
2165 | also use the full suite of core modules in the tests. | |
2166 | ||
2167 | =back | |
f7e1e956 MS |
2168 | |
2169 | When you say "make test" Perl uses the F<t/TEST> program to run the | |
7205a85d YO |
2170 | test suite (except under Win32 where it uses F<t/harness> instead.) |
2171 | All tests are run from the F<t/> directory, B<not> the directory | |
2172 | which contains the test. This causes some problems with the tests | |
2173 | in F<lib/>, so here's some opportunity for some patching. | |
f7e1e956 MS |
2174 | |
2175 | You must be triply conscious of cross-platform concerns. This usually | |
2176 | boils down to using File::Spec and avoiding things like C<fork()> and | |
2177 | C<system()> unless absolutely necessary. | |
2178 | ||
e018f8be JH |
2179 | =head2 Special Make Test Targets |
2180 | ||
2181 | There are various special make targets that can be used to test Perl | |
2182 | slightly differently than the standard "test" target. Not all them | |
2183 | are expected to give a 100% success rate. Many of them have several | |
7205a85d YO |
2184 | aliases, and many of them are not available on certain operating |
2185 | systems. | |
e018f8be JH |
2186 | |
2187 | =over 4 | |
2188 | ||
2189 | =item coretest | |
2190 | ||
7d7d5695 | 2191 | Run F<perl> on all core tests (F<t/*> and F<lib/[a-z]*> pragma tests). |
e018f8be | 2192 | |
7205a85d YO |
2193 | (Not available on Win32) |
2194 | ||
e018f8be JH |
2195 | =item test.deparse |
2196 | ||
b26492ee RGS |
2197 | Run all the tests through B::Deparse. Not all tests will succeed. |
2198 | ||
7205a85d YO |
2199 | (Not available on Win32) |
2200 | ||
b26492ee RGS |
2201 | =item test.taintwarn |
2202 | ||
2203 | Run all tests with the B<-t> command-line switch. Not all tests | |
2204 | are expected to succeed (until they're specifically fixed, of course). | |
e018f8be | 2205 | |
7205a85d YO |
2206 | (Not available on Win32) |
2207 | ||
e018f8be JH |
2208 | =item minitest |
2209 | ||
2210 | Run F<miniperl> on F<t/base>, F<t/comp>, F<t/cmd>, F<t/run>, F<t/io>, | |
2211 | F<t/op>, and F<t/uni> tests. | |
2212 | ||
7a834142 JH |
2213 | =item test.valgrind check.valgrind utest.valgrind ucheck.valgrind |
2214 | ||
2215 | (Only in Linux) Run all the tests using the memory leak + naughty | |
2216 | memory access tool "valgrind". The log files will be named | |
2217 | F<testname.valgrind>. | |
2218 | ||
e018f8be JH |
2219 | =item test.third check.third utest.third ucheck.third |
2220 | ||
2221 | (Only in Tru64) Run all the tests using the memory leak + naughty | |
2222 | memory access tool "Third Degree". The log files will be named | |
60a57c1c | 2223 | F<perl.3log.testname>. |
e018f8be JH |
2224 | |
2225 | =item test.torture torturetest | |
2226 | ||
2227 | Run all the usual tests and some extra tests. As of Perl 5.8.0 the | |
244d9cb7 | 2228 | only extra tests are Abigail's JAPHs, F<t/japh/abigail.t>. |
e018f8be JH |
2229 | |
2230 | You can also run the torture test with F<t/harness> by giving | |
2231 | C<-torture> argument to F<t/harness>. | |
2232 | ||
2233 | =item utest ucheck test.utf8 check.utf8 | |
2234 | ||
2235 | Run all the tests with -Mutf8. Not all tests will succeed. | |
2236 | ||
7205a85d YO |
2237 | (Not available on Win32) |
2238 | ||
cc0710ff RGS |
2239 | =item minitest.utf16 test.utf16 |
2240 | ||
2241 | Runs the tests with UTF-16 encoded scripts, encoded with different | |
2242 | versions of this encoding. | |
2243 | ||
2244 | C<make utest.utf16> runs the test suite with a combination of C<-utf8> and | |
2245 | C<-utf16> arguments to F<t/TEST>. | |
2246 | ||
7205a85d YO |
2247 | (Not available on Win32) |
2248 | ||
244d9cb7 RGS |
2249 | =item test_harness |
2250 | ||
2251 | Run the test suite with the F<t/harness> controlling program, instead of | |
2252 | F<t/TEST>. F<t/harness> is more sophisticated, and uses the | |
2253 | L<Test::Harness> module, thus using this test target supposes that perl | |
2254 | mostly works. The main advantage for our purposes is that it prints a | |
00bf5cd9 RGS |
2255 | detailed summary of failed tests at the end. Also, unlike F<t/TEST>, it |
2256 | doesn't redirect stderr to stdout. | |
244d9cb7 | 2257 | |
7205a85d YO |
2258 | Note that under Win32 F<t/harness> is always used instead of F<t/TEST>, so |
2259 | there is no special "test_harness" target. | |
2260 | ||
2261 | Under Win32's "test" target you may use the TEST_SWITCHES and TEST_FILES | |
2262 | environment variables to control the behaviour of F<t/harness>. This means | |
2263 | you can say | |
2264 | ||
2265 | nmake test TEST_FILES="op/*.t" | |
2266 | nmake test TEST_SWITCHES="-torture" TEST_FILES="op/*.t" | |
2267 | ||
2268 | =item test-notty test_notty | |
2269 | ||
2270 | Sets PERL_SKIP_TTY_TEST to true before running normal test. | |
2271 | ||
244d9cb7 RGS |
2272 | =back |
2273 | ||
2274 | =head2 Running tests by hand | |
2275 | ||
2276 | You can run part of the test suite by hand by using one the following | |
2277 | commands from the F<t/> directory : | |
2278 | ||
2279 | ./perl -I../lib TEST list-of-.t-files | |
2280 | ||
2281 | or | |
2282 | ||
2283 | ./perl -I../lib harness list-of-.t-files | |
2284 | ||
2285 | (if you don't specify test scripts, the whole test suite will be run.) | |
2286 | ||
7205a85d YO |
2287 | =head3 Using t/harness for testing |
2288 | ||
2289 | If you use C<harness> for testing you have several command line options | |
2290 | available to you. The arguments are as follows, and are in the order | |
2291 | that they must appear if used together. | |
2292 | ||
2293 | harness -v -torture -re=pattern LIST OF FILES TO TEST | |
2294 | harness -v -torture -re LIST OF PATTERNS TO MATCH | |
2295 | ||
2296 | If C<LIST OF FILES TO TEST> is omitted the file list is obtained from | |
2297 | the manifest. The file list may include shell wildcards which will be | |
2298 | expanded out. | |
2299 | ||
2300 | =over 4 | |
2301 | ||
2302 | =item -v | |
2303 | ||
2304 | Run the tests under verbose mode so you can see what tests were run, | |
2305 | and debug outbut. | |
2306 | ||
2307 | =item -torture | |
2308 | ||
2309 | Run the torture tests as well as the normal set. | |
2310 | ||
2311 | =item -re=PATTERN | |
2312 | ||
2313 | Filter the file list so that all the test files run match PATTERN. | |
2314 | Note that this form is distinct from the B<-re LIST OF PATTERNS> form below | |
2315 | in that it allows the file list to be provided as well. | |
2316 | ||
2317 | =item -re LIST OF PATTERNS | |
2318 | ||
2319 | Filter the file list so that all the test files run match | |
2320 | /(LIST|OF|PATTERNS)/. Note that with this form the patterns | |
2321 | are joined by '|' and you cannot supply a list of files, instead | |
2322 | the test files are obtained from the MANIFEST. | |
2323 | ||
2324 | =back | |
2325 | ||
244d9cb7 RGS |
2326 | You can run an individual test by a command similar to |
2327 | ||
2328 | ./perl -I../lib patho/to/foo.t | |
2329 | ||
2330 | except that the harnesses set up some environment variables that may | |
2331 | affect the execution of the test : | |
2332 | ||
2333 | =over 4 | |
2334 | ||
2335 | =item PERL_CORE=1 | |
2336 | ||
2337 | indicates that we're running this test part of the perl core test suite. | |
2338 | This is useful for modules that have a dual life on CPAN. | |
2339 | ||
2340 | =item PERL_DESTRUCT_LEVEL=2 | |
2341 | ||
2342 | is set to 2 if it isn't set already (see L</PERL_DESTRUCT_LEVEL>) | |
2343 | ||
2344 | =item PERL | |
2345 | ||
2346 | (used only by F<t/TEST>) if set, overrides the path to the perl executable | |
2347 | that should be used to run the tests (the default being F<./perl>). | |
2348 | ||
2349 | =item PERL_SKIP_TTY_TEST | |
2350 | ||
2351 | if set, tells to skip the tests that need a terminal. It's actually set | |
2352 | automatically by the Makefile, but can also be forced artificially by | |
2353 | running 'make test_notty'. | |
2354 | ||
e018f8be | 2355 | =back |
f7e1e956 | 2356 | |
d7889f52 JH |
2357 | =head2 Common problems when patching Perl source code |
2358 | ||
2359 | Perl source plays by ANSI C89 rules: no C99 (or C++) extensions. In | |
2360 | some cases we have to take pre-ANSI requirements into consideration. | |
2361 | You don't care about some particular platform having broken Perl? | |
2362 | I hear there is still a strong demand for J2EE programmers. | |
2363 | ||
2364 | =head2 Perl environment problems | |
2365 | ||
2366 | =over 4 | |
2367 | ||
2368 | =item * | |
2369 | ||
2370 | Not compiling with threading | |
2371 | ||
2372 | Compiling with threading (-Duseithreads) completely rewrites | |
2373 | the function prototypes of Perl. You better try your changes | |
0bec6c03 | 2374 | with that. Related to this is the difference between "Perl_-less" |
d7889f52 JH |
2375 | and "Perl_-ly" APIs, for example: |
2376 | ||
2377 | Perl_sv_setiv(aTHX_ ...); | |
2378 | sv_setiv(...); | |
2379 | ||
ee9468a2 RGS |
2380 | The first one explicitly passes in the context, which is needed for e.g. |
2381 | threaded builds. The second one does that implicitly; do not get them | |
def4ed7d JH |
2382 | mixed. If you are not passing in a aTHX_, you will need to do a dTHX |
2383 | (or a dVAR) as the first thing in the function. | |
d7889f52 JH |
2384 | |
2385 | See L<perlguts/"How multiple interpreters and concurrency are supported"> | |
2386 | for further discussion about context. | |
2387 | ||
2388 | =item * | |
2389 | ||
2390 | Not compiling with -DDEBUGGING | |
2391 | ||
2392 | The DEBUGGING define exposes more code to the compiler, | |
0bec6c03 | 2393 | therefore more ways for things to go wrong. You should try it. |
d7889f52 JH |
2394 | |
2395 | =item * | |
2396 | ||
ee9468a2 RGS |
2397 | Introducing (non-read-only) globals |
2398 | ||
2399 | Do not introduce any modifiable globals, truly global or file static. | |
bc028b6b JH |
2400 | They are bad form and complicate multithreading and other forms of |
2401 | concurrency. The right way is to introduce them as new interpreter | |
2402 | variables, see F<intrpvar.h> (at the very end for binary compatibility). | |
ee9468a2 RGS |
2403 | |
2404 | Introducing read-only (const) globals is okay, as long as you verify | |
2405 | with e.g. C<nm libperl.a|egrep -v ' [TURtr] '> (if your C<nm> has | |
2406 | BSD-style output) that the data you added really is read-only. | |
2407 | (If it is, it shouldn't show up in the output of that command.) | |
2408 | ||
2409 | If you want to have static strings, make them constant: | |
2410 | ||
2411 | static const char etc[] = "..."; | |
2412 | ||
bc028b6b | 2413 | If you want to have arrays of constant strings, note carefully |
ee9468a2 RGS |
2414 | the right combination of C<const>s: |
2415 | ||
2416 | static const char * const yippee[] = | |
2417 | {"hi", "ho", "silver"}; | |
2418 | ||
bc028b6b JH |
2419 | There is a way to completely hide any modifiable globals (they are all |
2420 | moved to heap), the compilation setting C<-DPERL_GLOBAL_STRUCT_PRIVATE>. | |
2421 | It is not normally used, but can be used for testing, read more | |
def4ed7d | 2422 | about it in L<perlguts/"Background and PERL_IMPLICIT_CONTEXT">. |
bc028b6b | 2423 | |
ee9468a2 RGS |
2424 | =item * |
2425 | ||
d7889f52 JH |
2426 | Not exporting your new function |
2427 | ||
2428 | Some platforms (Win32, AIX, VMS, OS/2, to name a few) require any | |
2429 | function that is part of the public API (the shared Perl library) | |
2430 | to be explicitly marked as exported. See the discussion about | |
2431 | F<embed.pl> in L<perlguts>. | |
2432 | ||
2433 | =item * | |
2434 | ||
2435 | Exporting your new function | |
2436 | ||
2437 | The new shiny result of either genuine new functionality or your | |
2438 | arduous refactoring is now ready and correctly exported. So what | |
def4ed7d | 2439 | could possibly go wrong? |
d7889f52 JH |
2440 | |
2441 | Maybe simply that your function did not need to be exported in the | |
2442 | first place. Perl has a long and not so glorious history of exporting | |
2443 | functions that it should not have. | |
2444 | ||
2445 | If the function is used only inside one source code file, make it | |
2446 | static. See the discussion about F<embed.pl> in L<perlguts>. | |
2447 | ||
2448 | If the function is used across several files, but intended only for | |
2449 | Perl's internal use (and this should be the common case), do not | |
2450 | export it to the public API. See the discussion about F<embed.pl> | |
2451 | in L<perlguts>. | |
2452 | ||
2453 | =back | |
2454 | ||
5b38b9cd | 2455 | =head2 Portability problems |
d7889f52 JH |
2456 | |
2457 | The following are common causes of compilation and/or execution | |
2458 | failures, not common to Perl as such. The C FAQ is good bedtime | |
0bec6c03 JH |
2459 | reading. Please test your changes with as many C compilers and |
2460 | platforms as possible -- we will, anyway, and it's nice to save | |
2461 | oneself from public embarrassment. | |
2462 | ||
9aaf14db RGS |
2463 | If using gcc, you can add the C<-std=c89> option which will hopefully |
2464 | catch most of these unportabilities. (However it might also catch | |
2465 | incompatibilities in your system's header files.) | |
d1307786 | 2466 | |
a8e98a71 JH |
2467 | Use the Configure C<-Dgccansipedantic> flag to enable the gcc |
2468 | C<-ansi -pedantic> flags which enforce stricter ANSI rules. | |
2469 | ||
def4ed7d JH |
2470 | If using the C<gcc -Wall> note that not all the possible warnings |
2471 | (like C<-Wunitialized>) are given unless you also compile with C<-O>. | |
2472 | ||
2473 | Note that if using gcc, starting from Perl 5.9.5 the Perl core source | |
2474 | code files (the ones at the top level of the source code distribution, | |
2475 | but not e.g. the extensions under ext/) are automatically compiled | |
2476 | with as many as possible of the C<-std=c89>, C<-ansi>, C<-pedantic>, | |
2477 | and a selection of C<-W> flags (see cflags.SH). | |
27565cb6 | 2478 | |
0bec6c03 | 2479 | Also study L<perlport> carefully to avoid any bad assumptions |
def4ed7d | 2480 | about the operating system, filesystems, and so forth. |
0bec6c03 | 2481 | |
606fd33d | 2482 | You may once in a while try a "make microperl" to see whether we |
63796a85 | 2483 | can still compile Perl with just the bare minimum of interfaces. |
606fd33d | 2484 | (See README.micro.) |
ee9468a2 | 2485 | |
0bec6c03 | 2486 | Do not assume an operating system indicates a certain compiler. |
d7889f52 JH |
2487 | |
2488 | =over 4 | |
2489 | ||
2490 | =item * | |
2491 | ||
2492 | Casting pointers to integers or casting integers to pointers | |
2493 | ||
2494 | void castaway(U8* p) | |
2495 | { | |
2496 | IV i = p; | |
2497 | ||
2498 | or | |
2499 | ||
2500 | void castaway(U8* p) | |
2501 | { | |
2502 | IV i = (IV)p; | |
2503 | ||
ee9468a2 | 2504 | Both are bad, and broken, and unportable. Use the PTR2IV() |
d7889f52 JH |
2505 | macro that does it right. (Likewise, there are PTR2UV(), PTR2NV(), |
2506 | INT2PTR(), and NUM2PTR().) | |
2507 | ||
2508 | =item * | |
2509 | ||
0bec6c03 JH |
2510 | Casting between data function pointers and data pointers |
2511 | ||
d7889f52 JH |
2512 | Technically speaking casting between function pointers and data |
2513 | pointers is unportable and undefined, but practically speaking | |
2514 | it seems to work, but you should use the FPTR2DPTR() and DPTR2FPTR() | |
0bec6c03 | 2515 | macros. Sometimes you can also play games with unions. |
d7889f52 JH |
2516 | |
2517 | =item * | |
2518 | ||
2519 | Assuming sizeof(int) == sizeof(long) | |
2520 | ||
2521 | There are platforms where longs are 64 bits, and platforms where ints | |
2522 | are 64 bits, and while we are out to shock you, even platforms where | |
2523 | shorts are 64 bits. This is all legal according to the C standard. | |
2524 | (In other words, "long long" is not a portable way to specify 64 bits, | |
2525 | and "long long" is not even guaranteed to be any wider than "long".) | |
63796a85 JH |
2526 | |
2527 | Instead, use the definitions IV, UV, IVSIZE, I32SIZE, and so forth. | |
2528 | Avoid things like I32 because they are B<not> guaranteed to be | |
2529 | I<exactly> 32 bits, they are I<at least> 32 bits, nor are they | |
2530 | guaranteed to be B<int> or B<long>. If you really explicitly need | |
2531 | 64-bit variables, use I64 and U64, but only if guarded by HAS_QUAD. | |
d7889f52 JH |
2532 | |
2533 | =item * | |
2534 | ||
2535 | Assuming one can dereference any type of pointer for any type of data | |
2536 | ||
2537 | char *p = ...; | |
def4ed7d | 2538 | long pony = *p; /* BAD */ |
d7889f52 JH |
2539 | |
2540 | Many platforms, quite rightly so, will give you a core dump instead | |
2541 | of a pony if the p happens not be correctly aligned. | |
2542 | ||
2543 | =item * | |
2544 | ||
2545 | Lvalue casts | |
2546 | ||
def4ed7d | 2547 | (int)*p = ...; /* BAD */ |
d7889f52 JH |
2548 | |
2549 | Simply not portable. Get your lvalue to be of the right type, | |
27565cb6 JH |
2550 | or maybe use temporary variables, or dirty tricks with unions. |
2551 | ||
2552 | =item * | |
2553 | ||
606fd33d JH |
2554 | Assume B<anything> about structs (especially the ones you |
2555 | don't control, like the ones coming from the system headers) | |
27565cb6 JH |
2556 | |
2557 | =over 8 | |
2558 | ||
2559 | =item * | |
2560 | ||
2561 | That a certain field exists in a struct | |
2562 | ||
2563 | =item * | |
2564 | ||
902821cc | 2565 | That no other fields exist besides the ones you know of |
27565cb6 JH |
2566 | |
2567 | =item * | |
2568 | ||
606fd33d | 2569 | That a field is of certain signedness, sizeof, or type |
27565cb6 JH |
2570 | |
2571 | =item * | |
2572 | ||
2573 | That the fields are in a certain order | |
2574 | ||
606fd33d JH |
2575 | =over 8 |
2576 | ||
27565cb6 JH |
2577 | =item * |
2578 | ||
606fd33d JH |
2579 | While C guarantees the ordering specified in the struct definition, |
2580 | between different platforms the definitions might differ | |
2581 | ||
2582 | =back | |
27565cb6 JH |
2583 | |
2584 | =item * | |
2585 | ||
606fd33d JH |
2586 | That the sizeof(struct) or the alignments are the same everywhere |
2587 | ||
2588 | =over 8 | |
27565cb6 JH |
2589 | |
2590 | =item * | |
2591 | ||
606fd33d JH |
2592 | There might be padding bytes between the fields to align the fields - |
2593 | the bytes can be anything | |
2594 | ||
2595 | =item * | |
2596 | ||
2597 | Structs are required to be aligned to the maximum alignment required | |
2598 | by the fields - which for native types is for usually equivalent to | |
2599 | sizeof() of the field | |
2600 | ||
2601 | =back | |
27565cb6 JH |
2602 | |
2603 | =back | |
d7889f52 JH |
2604 | |
2605 | =item * | |
2606 | ||
0bec6c03 JH |
2607 | Mixing #define and #ifdef |
2608 | ||
2609 | #define BURGLE(x) ... \ | |
def4ed7d | 2610 | #ifdef BURGLE_OLD_STYLE /* BAD */ |
0bec6c03 JH |
2611 | ... do it the old way ... \ |
2612 | #else | |
2613 | ... do it the new way ... \ | |
2614 | #endif | |
2615 | ||
ee9468a2 RGS |
2616 | You cannot portably "stack" cpp directives. For example in the above |
2617 | you need two separate BURGLE() #defines, one for each #ifdef branch. | |
2618 | ||
2619 | =item * | |
2620 | ||
2621 | Adding stuff after #endif or #else | |
2622 | ||
2623 | #ifdef SNOSH | |
2624 | ... | |
def4ed7d | 2625 | #else !SNOSH /* BAD */ |
ee9468a2 | 2626 | ... |
def4ed7d | 2627 | #endif SNOSH /* BAD */ |
ee9468a2 | 2628 | |
def4ed7d JH |
2629 | The #endif and #else cannot portably have anything non-comment after |
2630 | them. If you want to document what is going (which is a good idea | |
2631 | especially if the branches are long), use (C) comments: | |
ee9468a2 RGS |
2632 | |
2633 | #ifdef SNOSH | |
2634 | ... | |
2635 | #else /* !SNOSH */ | |
2636 | ... | |
2637 | #endif /* SNOSH */ | |
2638 | ||
2639 | The gcc option C<-Wendif-labels> warns about the bad variant | |
2640 | (by default on starting from Perl 5.9.4). | |
0bec6c03 JH |
2641 | |
2642 | =item * | |
2643 | ||
27565cb6 JH |
2644 | Having a comma after the last element of an enum list |
2645 | ||
2646 | enum color { | |
2647 | CERULEAN, | |
2648 | CHARTREUSE, | |
def4ed7d | 2649 | CINNABAR, /* BAD */ |
27565cb6 JH |
2650 | }; |
2651 | ||
2652 | is not portable. Leave out the last comma. | |
2653 | ||
2654 | Also note that whether enums are implicitly morphable to ints | |
2655 | varies between compilers, you might need to (int). | |
2656 | ||
2657 | =item * | |
2658 | ||
d7889f52 JH |
2659 | Using //-comments |
2660 | ||
def4ed7d | 2661 | // This function bamfoodles the zorklator. /* BAD */ |
d7889f52 JH |
2662 | |
2663 | That is C99 or C++. Perl is C89. Using the //-comments is silently | |
0bec6c03 JH |
2664 | allowed by many C compilers but cranking up the ANSI C89 strictness |
2665 | (which we like to do) causes the compilation to fail. | |
d7889f52 JH |
2666 | |
2667 | =item * | |
2668 | ||
2669 | Mixing declarations and code | |
2670 | ||
2671 | void zorklator() | |
2672 | { | |
2673 | int n = 3; | |
def4ed7d | 2674 | set_zorkmids(n); /* BAD */ |
d7889f52 JH |
2675 | int q = 4; |
2676 | ||
0bec6c03 JH |
2677 | That is C99 or C++. Some C compilers allow that, but you shouldn't. |
2678 | ||
63796a85 JH |
2679 | The gcc option C<-Wdeclaration-after-statements> scans for such problems |
2680 | (by default on starting from Perl 5.9.4). | |
2681 | ||
0bec6c03 JH |
2682 | =item * |
2683 | ||
2684 | Introducing variables inside for() | |
2685 | ||
def4ed7d | 2686 | for(int i = ...; ...; ...) { /* BAD */ |
0bec6c03 JH |
2687 | |
2688 | That is C99 or C++. While it would indeed be awfully nice to have that | |
2689 | also in C89, to limit the scope of the loop variable, alas, we cannot. | |
d7889f52 JH |
2690 | |
2691 | =item * | |
2692 | ||
2693 | Mixing signed char pointers with unsigned char pointers | |
2694 | ||
2695 | int foo(char *s) { ... } | |
2696 | ... | |
2697 | unsigned char *t = ...; /* Or U8* t = ... */ | |
def4ed7d | 2698 | foo(t); /* BAD */ |
d7889f52 JH |
2699 | |
2700 | While this is legal practice, it is certainly dubious, and downright | |
2701 | fatal in at least one platform: for example VMS cc considers this a | |
def4ed7d JH |
2702 | fatal error. One cause for people often making this mistake is that a |
2703 | "naked char" and therefore dereferencing a "naked char pointer" have | |
2704 | an undefined signedness: it depends on the compiler and the flags of | |
2705 | the compiler and the underlying platform whether the result is signed | |
2706 | or unsigned. For this very same reason using a 'char' as an array | |
2707 | index is bad. | |
d7889f52 JH |
2708 | |
2709 | =item * | |
2710 | ||
2711 | Macros that have string constants and their arguments as substrings of | |
2712 | the string constants | |
2713 | ||
def4ed7d | 2714 | #define FOO(n) printf("number = %d\n", n) /* BAD */ |
d7889f52 JH |
2715 | FOO(10); |
2716 | ||
2717 | Pre-ANSI semantics for that was equivalent to | |
2718 | ||
2719 | printf("10umber = %d\10"); | |
2720 | ||
0bec6c03 JH |
2721 | which is probably not what you were expecting. Unfortunately at least |
2722 | one reasonably common and modern C compiler does "real backward | |
63796a85 | 2723 | compatibility" here, in AIX that is what still happens even though the |
0bec6c03 JH |
2724 | rest of the AIX compiler is very happily C89. |
2725 | ||
2726 | =item * | |
2727 | ||
ee9468a2 RGS |
2728 | Using printf formats for non-basic C types |
2729 | ||
2730 | IV i = ...; | |
def4ed7d | 2731 | printf("i = %d\n", i); /* BAD */ |
ee9468a2 RGS |
2732 | |
2733 | While this might by accident work in some platform (where IV happens | |
2734 | to be an C<int>), in general it cannot. IV might be something larger. | |
2735 | Even worse the situation is with more specific types (defined by Perl's | |
2736 | configuration step in F<config.h>): | |
2737 | ||
2738 | Uid_t who = ...; | |
def4ed7d | 2739 | printf("who = %d\n", who); /* BAD */ |
ee9468a2 RGS |
2740 | |
2741 | The problem here is that Uid_t might be not only not C<int>-wide | |
2742 | but it might also be unsigned, in which case large uids would be | |
2743 | printed as negative values. | |
2744 | ||
2745 | There is no simple solution to this because of printf()'s limited | |
2746 | intelligence, but for many types the right format is available as | |
2747 | with either 'f' or '_f' suffix, for example: | |
2748 | ||
2749 | IVdf /* IV in decimal */ | |
2750 | UVxf /* UV is hexadecimal */ | |
2751 | ||
2752 | printf("i = %"IVdf"\n", i); /* The IVdf is a string constant. */ | |
2753 | ||
2754 | Uid_t_f /* Uid_t in decimal */ | |
2755 | ||
2756 | printf("who = %"Uid_t_f"\n", who); | |
2757 | ||
63796a85 JH |
2758 | Or you can try casting to a "wide enough" type: |
2759 | ||
2760 | printf("i = %"IVdf"\n", (IV)something_very_small_and_signed); | |
2761 | ||
2762 | Also remember that the C<%p> format really does require a void pointer: | |
2763 | ||
2764 | U8* p = ...; | |
2765 | printf("p = %p\n", (void*)p); | |
2766 | ||
ee9468a2 RGS |
2767 | The gcc option C<-Wformat> scans for such problems. |
2768 | ||
2769 | =item * | |
2770 | ||
0bec6c03 JH |
2771 | Blindly using variadic macros |
2772 | ||
63796a85 JH |
2773 | gcc has had them for a while with its own syntax, and C99 brought |
2774 | them with a standardized syntax. Don't use the former, and use | |
2775 | the latter only if the HAS_C99_VARIADIC_MACROS is defined. | |
0bec6c03 JH |
2776 | |
2777 | =item * | |
2778 | ||
2779 | Blindly passing va_list | |
2780 | ||
2781 | Not all platforms support passing va_list to further varargs (stdarg) | |
2782 | functions. The right thing to do is to copy the va_list using the | |
2783 | Perl_va_copy() if the NEED_VA_COPY is defined. | |
d7889f52 | 2784 | |
ee9468a2 RGS |
2785 | =item * |
2786 | ||
e5afc1ae | 2787 | Using gcc statement expressions |
63796a85 | 2788 | |
def4ed7d | 2789 | val = ({...;...;...}); /* BAD */ |
63796a85 | 2790 | |
def4ed7d JH |
2791 | While a nice extension, it's not portable. The Perl code does |
2792 | admittedly use them if available to gain some extra speed | |
2793 | (essentially as a funky form of inlining), but you shouldn't. | |
63796a85 JH |
2794 | |
2795 | =item * | |
2796 | ||
2797 | Binding together several statements | |
2798 | ||
2799 | Use the macros STMT_START and STMT_END. | |
2800 | ||
2801 | STMT_START { | |
2802 | ... | |
2803 | } STMT_END | |
2804 | ||
2805 | =item * | |
2806 | ||
ee9468a2 RGS |
2807 | Testing for operating systems or versions when should be testing for features |
2808 | ||
def4ed7d | 2809 | #ifdef __FOONIX__ /* BAD */ |
ee9468a2 RGS |
2810 | foo = quux(); |
2811 | #endif | |
2812 | ||
2813 | Unless you know with 100% certainty that quux() is only ever available | |
2814 | for the "Foonix" operating system B<and> that is available B<and> | |
2815 | correctly working for B<all> past, present, B<and> future versions of | |
2816 | "Foonix", the above is very wrong. This is more correct (though still | |
2817 | not perfect, because the below is a compile-time check): | |
2818 | ||
2819 | #ifdef HAS_QUUX | |
2820 | foo = quux(); | |
2821 | #endif | |
2822 | ||
def4ed7d | 2823 | How does the HAS_QUUX become defined where it needs to be? Well, if |
ee9468a2 RGS |
2824 | Foonix happens to be UNIXy enought to be able to run the Configure |
2825 | script, and Configure has been taught about detecting and testing | |
2826 | quux(), the HAS_QUUX will be correctly defined. In other platforms, | |
2827 | the corresponding configuration step will hopefully do the same. | |
2828 | ||
2829 | In a pinch, if you cannot wait for Configure to be educated, | |
2830 | or if you have a good hunch of where quux() might be available, | |
2831 | you can temporarily try the following: | |
2832 | ||
2833 | #if (defined(__FOONIX__) || defined(__BARNIX__)) | |
2834 | # define HAS_QUUX | |
2835 | #endif | |
2836 | ||
2837 | ... | |
2838 | ||
2839 | #ifdef HAS_QUUX | |
2840 | foo = quux(); | |
2841 | #endif | |
2842 | ||
2843 | But in any case, try to keep the features and operating systems separate. | |
2844 | ||
d7889f52 JH |
2845 | =back |
2846 | ||
2847 | =head2 Security problems | |
2848 | ||
2849 | Last but not least, here are various tips for safer coding. | |
2850 | ||
2851 | =over 4 | |
2852 | ||
2853 | =item * | |
2854 | ||
2855 | Do not use gets() | |
2856 | ||
2857 | Or we will publicly ridicule you. Seriously. | |
2858 | ||
2859 | =item * | |
2860 | ||
d1307786 | 2861 | Do not use strcpy() or strcat() or strncpy() or strncat() |
d7889f52 | 2862 | |
d1307786 JH |
2863 | Use my_strlcpy() and my_strlcat() instead: they either use the native |
2864 | implementation, or Perl's own implementation (borrowed from the public | |
2865 | domain implementation of INN). | |
d7889f52 JH |
2866 | |
2867 | =item * | |
2868 | ||
2869 | Do not use sprintf() or vsprintf() | |
2870 | ||
0bec6c03 | 2871 | If you really want just plain byte strings, use my_snprintf() |
64d9b66b | 2872 | and my_vsnprintf() instead, which will try to use snprintf() and |
0bec6c03 JH |
2873 | vsnprintf() if those safer APIs are available. If you want something |
2874 | fancier than a plain byte string, use SVs and Perl_sv_catpvf(). | |
d7889f52 JH |
2875 | |
2876 | =back | |
2877 | ||
902b9dbf MLF |
2878 | =head1 EXTERNAL TOOLS FOR DEBUGGING PERL |
2879 | ||
2880 | Sometimes it helps to use external tools while debugging and | |
2881 | testing Perl. This section tries to guide you through using | |
2882 | some common testing and debugging tools with Perl. This is | |
2883 | meant as a guide to interfacing these tools with Perl, not | |
2884 | as any kind of guide to the use of the tools themselves. | |
2885 | ||
a958818a JH |
2886 | B<NOTE 1>: Running under memory debuggers such as Purify, valgrind, or |
2887 | Third Degree greatly slows down the execution: seconds become minutes, | |
2888 | minutes become hours. For example as of Perl 5.8.1, the | |
2889 | ext/Encode/t/Unicode.t takes extraordinarily long to complete under | |
2890 | e.g. Purify, Third Degree, and valgrind. Under valgrind it takes more | |
2891 | than six hours, even on a snappy computer-- the said test must be | |
2892 | doing something that is quite unfriendly for memory debuggers. If you | |
2893 | don't feel like waiting, that you can simply kill away the perl | |
2894 | process. | |
2895 | ||
2896 | B<NOTE 2>: To minimize the number of memory leak false alarms (see | |
2897 | L</PERL_DESTRUCT_LEVEL> for more information), you have to have | |
2898 | environment variable PERL_DESTRUCT_LEVEL set to 2. The F<TEST> | |
2899 | and harness scripts do that automatically. But if you are running | |
2900 | some of the tests manually-- for csh-like shells: | |
2901 | ||
2902 | setenv PERL_DESTRUCT_LEVEL 2 | |
2903 | ||
2904 | and for Bourne-type shells: | |
2905 | ||
2906 | PERL_DESTRUCT_LEVEL=2 | |
2907 | export PERL_DESTRUCT_LEVEL | |
2908 | ||
2909 | or in UNIXy environments you can also use the C<env> command: | |
2910 | ||
2911 | env PERL_DESTRUCT_LEVEL=2 valgrind ./perl -Ilib ... | |
a1b65709 | 2912 | |
37c0adeb JH |
2913 | B<NOTE 3>: There are known memory leaks when there are compile-time |
2914 | errors within eval or require, seeing C<S_doeval> in the call stack | |
2915 | is a good sign of these. Fixing these leaks is non-trivial, | |
2916 | unfortunately, but they must be fixed eventually. | |
2917 | ||
902b9dbf MLF |
2918 | =head2 Rational Software's Purify |
2919 | ||
2920 | Purify is a commercial tool that is helpful in identifying | |
2921 | memory overruns, wild pointers, memory leaks and other such | |
2922 | badness. Perl must be compiled in a specific way for | |
2923 | optimal testing with Purify. Purify is available under | |
2924 | Windows NT, Solaris, HP-UX, SGI, and Siemens Unix. | |
2925 | ||
902b9dbf MLF |
2926 | =head2 Purify on Unix |
2927 | ||
2928 | On Unix, Purify creates a new Perl binary. To get the most | |
2929 | benefit out of Purify, you should create the perl to Purify | |
2930 | using: | |
2931 | ||
2932 | sh Configure -Accflags=-DPURIFY -Doptimize='-g' \ | |
2933 | -Uusemymalloc -Dusemultiplicity | |
2934 | ||
2935 | where these arguments mean: | |
2936 | ||
2937 | =over 4 | |
2938 | ||
2939 | =item -Accflags=-DPURIFY | |
2940 | ||
2941 | Disables Perl's arena memory allocation functions, as well as | |
2942 | forcing use of memory allocation functions derived from the | |
2943 | system malloc. | |
2944 | ||
2945 | =item -Doptimize='-g' | |
2946 | ||
2947 | Adds debugging information so that you see the exact source | |
2948 | statements where the problem occurs. Without this flag, all | |
2949 | you will see is the source filename of where the error occurred. | |
2950 | ||
2951 | =item -Uusemymalloc | |
2952 | ||
2953 | Disable Perl's malloc so that Purify can more closely monitor | |
2954 | allocations and leaks. Using Perl's malloc will make Purify | |
2955 | report most leaks in the "potential" leaks category. | |
2956 | ||
2957 | =item -Dusemultiplicity | |
2958 | ||
2959 | Enabling the multiplicity option allows perl to clean up | |
2960 | thoroughly when the interpreter shuts down, which reduces the | |
2961 | number of bogus leak reports from Purify. | |
2962 | ||
2963 | =back | |
2964 | ||
2965 | Once you've compiled a perl suitable for Purify'ing, then you | |
2966 | can just: | |
2967 | ||
2968 | make pureperl | |
2969 | ||
2970 | which creates a binary named 'pureperl' that has been Purify'ed. | |
2971 | This binary is used in place of the standard 'perl' binary | |
2972 | when you want to debug Perl memory problems. | |
2973 | ||
2974 | As an example, to show any memory leaks produced during the | |
2975 | standard Perl testset you would create and run the Purify'ed | |
2976 | perl as: | |
2977 | ||
2978 | make pureperl | |
2979 | cd t | |
2980 | ../pureperl -I../lib harness | |
2981 | ||
2982 | which would run Perl on test.pl and report any memory problems. | |
2983 | ||
2984 | Purify outputs messages in "Viewer" windows by default. If | |
2985 | you don't have a windowing environment or if you simply | |
2986 | want the Purify output to unobtrusively go to a log file | |
2987 | instead of to the interactive window, use these following | |
2988 | options to output to the log file "perl.log": | |
2989 | ||
2990 | setenv PURIFYOPTIONS "-chain-length=25 -windows=no \ | |
2991 | -log-file=perl.log -append-logfile=yes" | |
2992 | ||
2993 | If you plan to use the "Viewer" windows, then you only need this option: | |
2994 | ||
2995 | setenv PURIFYOPTIONS "-chain-length=25" | |
2996 | ||
c406981e JH |
2997 | In Bourne-type shells: |
2998 | ||
98631ff8 JL |
2999 | PURIFYOPTIONS="..." |
3000 | export PURIFYOPTIONS | |
c406981e JH |
3001 | |
3002 | or if you have the "env" utility: | |
3003 | ||
98631ff8 | 3004 | env PURIFYOPTIONS="..." ../pureperl ... |
c406981e | 3005 | |
902b9dbf MLF |
3006 | =head2 Purify on NT |
3007 | ||
3008 | Purify on Windows NT instruments the Perl binary 'perl.exe' | |
3009 | on the fly. There are several options in the makefile you | |
3010 | should change to get the most use out of Purify: | |
3011 | ||
3012 | =over 4 | |
3013 | ||
3014 | =item DEFINES | |
3015 | ||
3016 | You should add -DPURIFY to the DEFINES line so the DEFINES | |
3017 | line looks something like: | |
3018 | ||
3019 | DEFINES = -DWIN32 -D_CONSOLE -DNO_STRICT $(CRYPT_FLAG) -DPURIFY=1 | |
3020 | ||
3021 | to disable Perl's arena memory allocation functions, as | |
3022 | well as to force use of memory allocation functions derived | |
3023 | from the system malloc. | |
3024 | ||
3025 | =item USE_MULTI = define | |
3026 | ||
3027 | Enabling the multiplicity option allows perl to clean up | |
3028 | thoroughly when the interpreter shuts down, which reduces the | |
3029 | number of bogus leak reports from Purify. | |
3030 | ||
3031 | =item #PERL_MALLOC = define | |
3032 | ||
3033 | Disable Perl's malloc so that Purify can more closely monitor | |
3034 | allocations and leaks. Using Perl's malloc will make Purify | |
3035 | report most leaks in the "potential" leaks category. | |
3036 | ||
3037 | =item CFG = Debug | |
3038 | ||
3039 | Adds debugging information so that you see the exact source | |
3040 | statements where the problem occurs. Without this flag, all | |
3041 | you will see is the source filename of where the error occurred. | |
3042 | ||
3043 | =back | |
3044 | ||
3045 | As an example, to show any memory leaks produced during the | |
3046 | standard Perl testset you would create and run Purify as: | |
3047 | ||
3048 | cd win32 | |
3049 | make | |
3050 | cd ../t | |
3051 | purify ../perl -I../lib harness | |
3052 | ||
3053 | which would instrument Perl in memory, run Perl on test.pl, | |
3054 | then finally report any memory problems. | |
3055 | ||
7a834142 JH |
3056 | =head2 valgrind |
3057 | ||
3058 | The excellent valgrind tool can be used to find out both memory leaks | |
3059 | and illegal memory accesses. As of August 2003 it unfortunately works | |
3060 | only on x86 (ELF) Linux. The special "test.valgrind" target can be used | |
d44161bf MHM |
3061 | to run the tests under valgrind. Found errors and memory leaks are |
3062 | logged in files named F<test.valgrind>. | |
3063 | ||
3064 | As system libraries (most notably glibc) are also triggering errors, | |
3065 | valgrind allows to suppress such errors using suppression files. The | |
3066 | default suppression file that comes with valgrind already catches a lot | |
3067 | of them. Some additional suppressions are defined in F<t/perl.supp>. | |
7a834142 JH |
3068 | |
3069 | To get valgrind and for more information see | |
3070 | ||
3071 | http://developer.kde.org/~sewardj/ | |
3072 | ||
f134cc4e | 3073 | =head2 Compaq's/Digital's/HP's Third Degree |
09187cb1 JH |
3074 | |
3075 | Third Degree is a tool for memory leak detection and memory access checks. | |
3076 | It is one of the many tools in the ATOM toolkit. The toolkit is only | |
3077 | available on Tru64 (formerly known as Digital UNIX formerly known as | |
3078 | DEC OSF/1). | |
3079 | ||
3080 | When building Perl, you must first run Configure with -Doptimize=-g | |
3081 | and -Uusemymalloc flags, after that you can use the make targets | |
51a35ef1 JH |
3082 | "perl.third" and "test.third". (What is required is that Perl must be |
3083 | compiled using the C<-g> flag, you may need to re-Configure.) | |
09187cb1 | 3084 | |
64cea5fd | 3085 | The short story is that with "atom" you can instrument the Perl |
83f0ef60 | 3086 | executable to create a new executable called F<perl.third>. When the |
4ae3d70a | 3087 | instrumented executable is run, it creates a log of dubious memory |
83f0ef60 | 3088 | traffic in file called F<perl.3log>. See the manual pages of atom and |
4ae3d70a JH |
3089 | third for more information. The most extensive Third Degree |
3090 | documentation is available in the Compaq "Tru64 UNIX Programmer's | |
3091 | Guide", chapter "Debugging Programs with Third Degree". | |
64cea5fd | 3092 | |
9c54ecba | 3093 | The "test.third" leaves a lot of files named F<foo_bar.3log> in the t/ |
64cea5fd JH |
3094 | subdirectory. There is a problem with these files: Third Degree is so |
3095 | effective that it finds problems also in the system libraries. | |
9c54ecba JH |
3096 | Therefore you should used the Porting/thirdclean script to cleanup |
3097 | the F<*.3log> files. | |
64cea5fd JH |
3098 | |
3099 | There are also leaks that for given certain definition of a leak, | |
3100 | aren't. See L</PERL_DESTRUCT_LEVEL> for more information. | |
3101 | ||
3102 | =head2 PERL_DESTRUCT_LEVEL | |
3103 | ||
a958818a JH |
3104 | If you want to run any of the tests yourself manually using e.g. |
3105 | valgrind, or the pureperl or perl.third executables, please note that | |
3106 | by default perl B<does not> explicitly cleanup all the memory it has | |
3107 | allocated (such as global memory arenas) but instead lets the exit() | |
3108 | of the whole program "take care" of such allocations, also known as | |
3109 | "global destruction of objects". | |
64cea5fd JH |
3110 | |
3111 | There is a way to tell perl to do complete cleanup: set the | |
3112 | environment variable PERL_DESTRUCT_LEVEL to a non-zero value. | |
3113 | The t/TEST wrapper does set this to 2, and this is what you | |
3114 | need to do too, if you don't want to see the "global leaks": | |
1f56d61a | 3115 | For example, for "third-degreed" Perl: |
64cea5fd | 3116 | |
1f56d61a | 3117 | env PERL_DESTRUCT_LEVEL=2 ./perl.third -Ilib t/foo/bar.t |
09187cb1 | 3118 | |
414f2397 RGS |
3119 | (Note: the mod_perl apache module uses also this environment variable |
3120 | for its own purposes and extended its semantics. Refer to the mod_perl | |
287a822c RGS |
3121 | documentation for more information. Also, spawned threads do the |
3122 | equivalent of setting this variable to the value 1.) | |
5a6c59ef DM |
3123 | |
3124 | If, at the end of a run you get the message I<N scalars leaked>, you can | |
fd0854ff DM |
3125 | recompile with C<-DDEBUG_LEAKING_SCALARS>, which will cause the addresses |
3126 | of all those leaked SVs to be dumped along with details as to where each | |
3127 | SV was originally allocated. This information is also displayed by | |
3128 | Devel::Peek. Note that the extra details recorded with each SV increases | |
3129 | memory usage, so it shouldn't be used in production environments. It also | |
3130 | converts C<new_SV()> from a macro into a real function, so you can use | |
3131 | your favourite debugger to discover where those pesky SVs were allocated. | |
414f2397 | 3132 | |
46c6c7e2 JH |
3133 | =head2 PERL_MEM_LOG |
3134 | ||
9f653bb5 | 3135 | If compiled with C<-DPERL_MEM_LOG>, all Newx() and Renew() allocations |
46c6c7e2 JH |
3136 | and Safefree() in the Perl core go through logging functions, which is |
3137 | handy for breakpoint setting. If also compiled with C<-DPERL_MEM_LOG_STDERR>, | |
e352bcff JH |
3138 | the allocations and frees are logged to STDERR (or more precisely, to the |
3139 | file descriptor 2) in these logging functions, with the calling source code | |
3140 | file and line number (and C function name, if supported by the C compiler). | |
3141 | ||
3142 | This logging is somewhat similar to C<-Dm> but independent of C<-DDEBUGGING>, | |
3143 | and at a higher level (the C<-Dm> is directly at the point of C<malloc()>, | |
a015a77e | 3144 | while the C<PERL_MEM_LOG> is at the level of C<New()>). |
46c6c7e2 | 3145 | |
51a35ef1 JH |
3146 | =head2 Profiling |
3147 | ||
3148 | Depending on your platform there are various of profiling Perl. | |
3149 | ||
3150 | There are two commonly used techniques of profiling executables: | |
10f58044 | 3151 | I<statistical time-sampling> and I<basic-block counting>. |
51a35ef1 JH |
3152 | |
3153 | The first method takes periodically samples of the CPU program | |
3154 | counter, and since the program counter can be correlated with the code | |
3155 | generated for functions, we get a statistical view of in which | |
3156 | functions the program is spending its time. The caveats are that very | |
3157 | small/fast functions have lower probability of showing up in the | |
3158 | profile, and that periodically interrupting the program (this is | |
3159 | usually done rather frequently, in the scale of milliseconds) imposes | |
3160 | an additional overhead that may skew the results. The first problem | |
3161 | can be alleviated by running the code for longer (in general this is a | |
3162 | good idea for profiling), the second problem is usually kept in guard | |
3163 | by the profiling tools themselves. | |
3164 | ||
10f58044 | 3165 | The second method divides up the generated code into I<basic blocks>. |
51a35ef1 JH |
3166 | Basic blocks are sections of code that are entered only in the |
3167 | beginning and exited only at the end. For example, a conditional jump | |
3168 | starts a basic block. Basic block profiling usually works by | |
10f58044 | 3169 | I<instrumenting> the code by adding I<enter basic block #nnnn> |
51a35ef1 JH |
3170 | book-keeping code to the generated code. During the execution of the |
3171 | code the basic block counters are then updated appropriately. The | |
3172 | caveat is that the added extra code can skew the results: again, the | |
3173 | profiling tools usually try to factor their own effects out of the | |
3174 | results. | |
3175 | ||
83f0ef60 JH |
3176 | =head2 Gprof Profiling |
3177 | ||
51a35ef1 JH |
3178 | gprof is a profiling tool available in many UNIX platforms, |
3179 | it uses F<statistical time-sampling>. | |
83f0ef60 JH |
3180 | |
3181 | You can build a profiled version of perl called "perl.gprof" by | |
51a35ef1 JH |
3182 | invoking the make target "perl.gprof" (What is required is that Perl |
3183 | must be compiled using the C<-pg> flag, you may need to re-Configure). | |
3184 | Running the profiled version of Perl will create an output file called | |
3185 | F<gmon.out> is created which contains the profiling data collected | |
3186 | during the execution. | |
83f0ef60 JH |
3187 | |
3188 | The gprof tool can then display the collected data in various ways. | |
3189 | Usually gprof understands the following options: | |
3190 | ||
3191 | =over 4 | |
3192 | ||
3193 | =item -a | |
3194 | ||
3195 | Suppress statically defined functions from the profile. | |
3196 | ||
3197 | =item -b | |
3198 | ||
3199 | Suppress the verbose descriptions in the profile. | |
3200 | ||
3201 | =item -e routine | |
3202 | ||
3203 | Exclude the given routine and its descendants from the profile. | |
3204 | ||
3205 | =item -f routine | |
3206 | ||
3207 | Display only the given routine and its descendants in the profile. | |
3208 | ||
3209 | =item -s | |
3210 | ||
3211 | Generate a summary file called F<gmon.sum> which then may be given | |
3212 | to subsequent gprof runs to accumulate data over several runs. | |
3213 | ||
3214 | =item -z | |
3215 | ||
3216 | Display routines that have zero usage. | |
3217 | ||
3218 | =back | |
3219 | ||
3220 | For more detailed explanation of the available commands and output | |
3221 | formats, see your own local documentation of gprof. | |
3222 | ||
51a35ef1 JH |
3223 | =head2 GCC gcov Profiling |
3224 | ||
10f58044 | 3225 | Starting from GCC 3.0 I<basic block profiling> is officially available |
51a35ef1 JH |
3226 | for the GNU CC. |
3227 | ||
3228 | You can build a profiled version of perl called F<perl.gcov> by | |
3229 | invoking the make target "perl.gcov" (what is required that Perl must | |
3230 | be compiled using gcc with the flags C<-fprofile-arcs | |
3231 | -ftest-coverage>, you may need to re-Configure). | |
3232 | ||
3233 | Running the profiled version of Perl will cause profile output to be | |
3234 | generated. For each source file an accompanying ".da" file will be | |
3235 | created. | |
3236 | ||
3237 | To display the results you use the "gcov" utility (which should | |
3238 | be installed if you have gcc 3.0 or newer installed). F<gcov> is | |
3239 | run on source code files, like this | |
3240 | ||
3241 | gcov sv.c | |
3242 | ||
3243 | which will cause F<sv.c.gcov> to be created. The F<.gcov> files | |
3244 | contain the source code annotated with relative frequencies of | |
3245 | execution indicated by "#" markers. | |
3246 | ||
3247 | Useful options of F<gcov> include C<-b> which will summarise the | |
3248 | basic block, branch, and function call coverage, and C<-c> which | |
3249 | instead of relative frequencies will use the actual counts. For | |
3250 | more information on the use of F<gcov> and basic block profiling | |
3251 | with gcc, see the latest GNU CC manual, as of GCC 3.0 see | |
3252 | ||
3253 | http://gcc.gnu.org/onlinedocs/gcc-3.0/gcc.html | |
3254 | ||
3255 | and its section titled "8. gcov: a Test Coverage Program" | |
3256 | ||
3257 | http://gcc.gnu.org/onlinedocs/gcc-3.0/gcc_8.html#SEC132 | |
3258 | ||
4ae3d70a JH |
3259 | =head2 Pixie Profiling |
3260 | ||
51a35ef1 JH |
3261 | Pixie is a profiling tool available on IRIX and Tru64 (aka Digital |
3262 | UNIX aka DEC OSF/1) platforms. Pixie does its profiling using | |
10f58044 | 3263 | I<basic-block counting>. |
4ae3d70a | 3264 | |
83f0ef60 | 3265 | You can build a profiled version of perl called F<perl.pixie> by |
51a35ef1 JH |
3266 | invoking the make target "perl.pixie" (what is required is that Perl |
3267 | must be compiled using the C<-g> flag, you may need to re-Configure). | |
3268 | ||
3269 | In Tru64 a file called F<perl.Addrs> will also be silently created, | |
3270 | this file contains the addresses of the basic blocks. Running the | |
3271 | profiled version of Perl will create a new file called "perl.Counts" | |
3272 | which contains the counts for the basic block for that particular | |
3273 | program execution. | |
4ae3d70a | 3274 | |
51a35ef1 | 3275 | To display the results you use the F<prof> utility. The exact |
4ae3d70a JH |
3276 | incantation depends on your operating system, "prof perl.Counts" in |
3277 | IRIX, and "prof -pixie -all -L. perl" in Tru64. | |
3278 | ||
6c41479b JH |
3279 | In IRIX the following prof options are available: |
3280 | ||
3281 | =over 4 | |
3282 | ||
3283 | =item -h | |
3284 | ||
3285 | Reports the most heavily used lines in descending order of use. | |
6e36760b | 3286 | Useful for finding the hotspot lines. |
6c41479b JH |
3287 | |
3288 | =item -l | |
3289 | ||
3290 | Groups lines by procedure, with procedures sorted in descending order of use. | |
3291 | Within a procedure, lines are listed in source order. | |
6e36760b | 3292 | Useful for finding the hotspots of procedures. |
6c41479b JH |
3293 | |
3294 | =back | |
3295 | ||
3296 | In Tru64 the following options are available: | |
3297 | ||
3298 | =over 4 | |
3299 | ||
3958b146 | 3300 | =item -p[rocedures] |
6c41479b | 3301 | |
3958b146 | 3302 | Procedures sorted in descending order by the number of cycles executed |
6e36760b | 3303 | in each procedure. Useful for finding the hotspot procedures. |
6c41479b JH |
3304 | (This is the default option.) |
3305 | ||
24000d2f | 3306 | =item -h[eavy] |
6c41479b | 3307 | |
6e36760b JH |
3308 | Lines sorted in descending order by the number of cycles executed in |
3309 | each line. Useful for finding the hotspot lines. | |
6c41479b | 3310 | |
24000d2f | 3311 | =item -i[nvocations] |
6c41479b | 3312 | |
6e36760b JH |
3313 | The called procedures are sorted in descending order by number of calls |
3314 | made to the procedures. Useful for finding the most used procedures. | |
6c41479b | 3315 | |
24000d2f | 3316 | =item -l[ines] |
6c41479b JH |
3317 | |
3318 | Grouped by procedure, sorted by cycles executed per procedure. | |
6e36760b | 3319 | Useful for finding the hotspots of procedures. |
6c41479b JH |
3320 | |
3321 | =item -testcoverage | |
3322 | ||
3323 | The compiler emitted code for these lines, but the code was unexecuted. | |
3324 | ||
24000d2f | 3325 | =item -z[ero] |
6c41479b JH |
3326 | |
3327 | Unexecuted procedures. | |
3328 | ||
aa500c9e | 3329 | =back |
6c41479b JH |
3330 | |
3331 | For further information, see your system's manual pages for pixie and prof. | |
4ae3d70a | 3332 | |
b8ddf6b3 SB |
3333 | =head2 Miscellaneous tricks |
3334 | ||
3335 | =over 4 | |
3336 | ||
3337 | =item * | |
3338 | ||
cc177e1a | 3339 | Those debugging perl with the DDD frontend over gdb may find the |
b8ddf6b3 SB |
3340 | following useful: |
3341 | ||
3342 | You can extend the data conversion shortcuts menu, so for example you | |
3343 | can display an SV's IV value with one click, without doing any typing. | |
3344 | To do that simply edit ~/.ddd/init file and add after: | |
3345 | ||
3346 | ! Display shortcuts. | |
3347 | Ddd*gdbDisplayShortcuts: \ | |
3348 | /t () // Convert to Bin\n\ | |
3349 | /d () // Convert to Dec\n\ | |
3350 | /x () // Convert to Hex\n\ | |
3351 | /o () // Convert to Oct(\n\ | |
3352 | ||
3353 | the following two lines: | |
3354 | ||
3355 | ((XPV*) (())->sv_any )->xpv_pv // 2pvx\n\ | |
3356 | ((XPVIV*) (())->sv_any )->xiv_iv // 2ivx | |
3357 | ||
3358 | so now you can do ivx and pvx lookups or you can plug there the | |
3359 | sv_peek "conversion": | |
3360 | ||
3361 | Perl_sv_peek(my_perl, (SV*)()) // sv_peek | |
3362 | ||
3363 | (The my_perl is for threaded builds.) | |
3364 | Just remember that every line, but the last one, should end with \n\ | |
3365 | ||
3366 | Alternatively edit the init file interactively via: | |
3367 | 3rd mouse button -> New Display -> Edit Menu | |
3368 | ||
3369 | Note: you can define up to 20 conversion shortcuts in the gdb | |
3370 | section. | |
3371 | ||
9965345d JH |
3372 | =item * |
3373 | ||
7e337ee0 JH |
3374 | If you see in a debugger a memory area mysteriously full of 0xABABABAB |
3375 | or 0xEFEFEFEF, you may be seeing the effect of the Poison() macros, | |
3376 | see L<perlclib>. | |
9965345d | 3377 | |
b8ddf6b3 SB |
3378 | =back |
3379 | ||
955fec6b | 3380 | =head1 CONCLUSION |
a422fd2d | 3381 | |
955fec6b JH |
3382 | We've had a brief look around the Perl source, how to maintain quality |
3383 | of the source code, an overview of the stages F<perl> goes through | |
3384 | when it's running your code, how to use debuggers to poke at the Perl | |
3385 | guts, and finally how to analyse the execution of Perl. We took a very | |
3386 | simple problem and demonstrated how to solve it fully - with | |
3387 | documentation, regression tests, and finally a patch for submission to | |
3388 | p5p. Finally, we talked about how to use external tools to debug and | |
3389 | test Perl. | |
a422fd2d SC |
3390 | |
3391 | I'd now suggest you read over those references again, and then, as soon | |
3392 | as possible, get your hands dirty. The best way to learn is by doing, | |
3393 | so: | |
3394 | ||
3395 | =over 3 | |
3396 | ||
3397 | =item * | |
3398 | ||
3399 | Subscribe to perl5-porters, follow the patches and try and understand | |
3400 | them; don't be afraid to ask if there's a portion you're not clear on - | |
3401 | who knows, you may unearth a bug in the patch... | |
3402 | ||
3403 | =item * | |
3404 | ||
3405 | Keep up to date with the bleeding edge Perl distributions and get | |
3406 | familiar with the changes. Try and get an idea of what areas people are | |
3407 | working on and the changes they're making. | |
3408 | ||
3409 | =item * | |
3410 | ||
3e148164 | 3411 | Do read the README associated with your operating system, e.g. README.aix |
a1f349fd MB |
3412 | on the IBM AIX OS. Don't hesitate to supply patches to that README if |
3413 | you find anything missing or changed over a new OS release. | |
3414 | ||
3415 | =item * | |
3416 | ||
a422fd2d SC |
3417 | Find an area of Perl that seems interesting to you, and see if you can |
3418 | work out how it works. Scan through the source, and step over it in the | |
3419 | debugger. Play, poke, investigate, fiddle! You'll probably get to | |
3420 | understand not just your chosen area but a much wider range of F<perl>'s | |
3421 | activity as well, and probably sooner than you'd think. | |
3422 | ||
3423 | =back | |
3424 | ||
3425 | =over 3 | |
3426 | ||
3427 | =item I<The Road goes ever on and on, down from the door where it began.> | |
3428 | ||
3429 | =back | |
3430 | ||
64d9b66b | 3431 | If you can do these things, you've started on the long road to Perl porting. |
a422fd2d SC |
3432 | Thanks for wanting to help make Perl better - and happy hacking! |
3433 | ||
e8cd7eae GS |
3434 | =head1 AUTHOR |
3435 | ||
3436 | This document was written by Nathan Torkington, and is maintained by | |
3437 | the perl5-porters mailing list. |