Commit | Line | Data |
e8cd7eae |
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 | |
17 | A searchable archive of the list is at: |
18 | |
19 | http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/ |
20 | |
21 | The list is also archived under the usenet group name |
22 | C<perl.porters-gw> at: |
23 | |
24 | http://www.deja.com/ |
25 | |
26 | List subscribers (the porters themselves) come in several flavours. |
27 | Some are quiet curious lurkers, who rarely pitch in and instead watch |
28 | the ongoing development to ensure they're forewarned of new changes or |
29 | features in Perl. Some are representatives of vendors, who are there |
30 | to make sure that Perl continues to compile and work on their |
31 | platforms. Some patch any reported bug that they know how to fix, |
32 | some are actively patching their pet area (threads, Win32, the regexp |
33 | engine), while others seem to do nothing but complain. In other |
34 | words, it's your usual mix of technical people. |
35 | |
36 | Over this group of porters presides Larry Wall. He has the final word |
f6c51b38 |
37 | in what does and does not change in the Perl language. Various |
38 | releases of Perl are shepherded by a ``pumpking'', a porter |
39 | responsible for gathering patches, deciding on a patch-by-patch |
40 | feature-by-feature basis what will and will not go into the release. |
41 | For instance, Gurusamy Sarathy is the pumpking for the 5.6 release of |
42 | Perl. |
e8cd7eae |
43 | |
44 | In addition, various people are pumpkings for different things. For |
45 | instance, Andy Dougherty and Jarkko Hietaniemi share the I<Configure> |
46 | pumpkin, and Tom Christiansen is the documentation pumpking. |
47 | |
48 | Larry sees Perl development along the lines of the US government: |
49 | there's the Legislature (the porters), the Executive branch (the |
50 | pumpkings), and the Supreme Court (Larry). The legislature can |
51 | discuss and submit patches to the executive branch all they like, but |
52 | the executive branch is free to veto them. Rarely, the Supreme Court |
53 | will side with the executive branch over the legislature, or the |
54 | legislature over the executive branch. Mostly, however, the |
55 | legislature and the executive branch are supposed to get along and |
56 | work out their differences without impeachment or court cases. |
57 | |
58 | You might sometimes see reference to Rule 1 and Rule 2. Larry's power |
59 | as Supreme Court is expressed in The Rules: |
60 | |
61 | =over 4 |
62 | |
63 | =item 1 |
64 | |
65 | Larry is always by definition right about how Perl should behave. |
66 | This means he has final veto power on the core functionality. |
67 | |
68 | =item 2 |
69 | |
70 | Larry is allowed to change his mind about any matter at a later date, |
71 | regardless of whether he previously invoked Rule 1. |
72 | |
73 | =back |
74 | |
75 | Got that? Larry is always right, even when he was wrong. It's rare |
76 | to see either Rule exercised, but they are often alluded to. |
77 | |
78 | New features and extensions to the language are contentious, because |
79 | the criteria used by the pumpkings, Larry, and other porters to decide |
80 | which features should be implemented and incorporated are not codified |
81 | in a few small design goals as with some other languages. Instead, |
82 | the heuristics are flexible and often difficult to fathom. Here is |
83 | one person's list, roughly in decreasing order of importance, of |
84 | heuristics that new features have to be weighed against: |
85 | |
86 | =over 4 |
87 | |
88 | =item Does concept match the general goals of Perl? |
89 | |
90 | These haven't been written anywhere in stone, but one approximation |
91 | is: |
92 | |
93 | 1. Keep it fast, simple, and useful. |
94 | 2. Keep features/concepts as orthogonal as possible. |
95 | 3. No arbitrary limits (platforms, data sizes, cultures). |
96 | 4. Keep it open and exciting to use/patch/advocate Perl everywhere. |
97 | 5. Either assimilate new technologies, or build bridges to them. |
98 | |
99 | =item Where is the implementation? |
100 | |
101 | All the talk in the world is useless without an implementation. In |
102 | almost every case, the person or people who argue for a new feature |
103 | will be expected to be the ones who implement it. Porters capable |
104 | of coding new features have their own agendas, and are not available |
105 | to implement your (possibly good) idea. |
106 | |
107 | =item Backwards compatibility |
108 | |
109 | It's a cardinal sin to break existing Perl programs. New warnings are |
110 | contentious--some say that a program that emits warnings is not |
111 | broken, while others say it is. Adding keywords has the potential to |
112 | break programs, changing the meaning of existing token sequences or |
113 | functions might break programs. |
114 | |
115 | =item Could it be a module instead? |
116 | |
117 | Perl 5 has extension mechanisms, modules and XS, specifically to avoid |
118 | the need to keep changing the Perl interpreter. You can write modules |
119 | that export functions, you can give those functions prototypes so they |
120 | can be called like built-in functions, you can even write XS code to |
121 | mess with the runtime data structures of the Perl interpreter if you |
122 | want to implement really complicated things. If it can be done in a |
123 | module instead of in the core, it's highly unlikely to be added. |
124 | |
125 | =item Is the feature generic enough? |
126 | |
127 | Is this something that only the submitter wants added to the language, |
128 | or would it be broadly useful? Sometimes, instead of adding a feature |
129 | with a tight focus, the porters might decide to wait until someone |
130 | implements the more generalized feature. For instance, instead of |
131 | implementing a ``delayed evaluation'' feature, the porters are waiting |
132 | for a macro system that would permit delayed evaluation and much more. |
133 | |
134 | =item Does it potentially introduce new bugs? |
135 | |
136 | Radical rewrites of large chunks of the Perl interpreter have the |
137 | potential to introduce new bugs. The smaller and more localized the |
138 | change, the better. |
139 | |
140 | =item Does it preclude other desirable features? |
141 | |
142 | A patch is likely to be rejected if it closes off future avenues of |
143 | development. For instance, a patch that placed a true and final |
144 | interpretation on prototypes is likely to be rejected because there |
145 | are still options for the future of prototypes that haven't been |
146 | addressed. |
147 | |
148 | =item Is the implementation robust? |
149 | |
150 | Good patches (tight code, complete, correct) stand more chance of |
151 | going in. Sloppy or incorrect patches might be placed on the back |
152 | burner until the pumpking has time to fix, or might be discarded |
153 | altogether without further notice. |
154 | |
155 | =item Is the implementation generic enough to be portable? |
156 | |
157 | The worst patches make use of a system-specific features. It's highly |
158 | unlikely that nonportable additions to the Perl language will be |
159 | accepted. |
160 | |
161 | =item Is there enough documentation? |
162 | |
163 | Patches without documentation are probably ill-thought out or |
164 | incomplete. Nothing can be added without documentation, so submitting |
165 | a patch for the appropriate manpages as well as the source code is |
166 | always a good idea. If appropriate, patches should add to the test |
167 | suite as well. |
168 | |
169 | =item Is there another way to do it? |
170 | |
171 | Larry said ``Although the Perl Slogan is I<There's More Than One Way |
172 | to Do It>, I hesitate to make 10 ways to do something''. This is a |
173 | tricky heuristic to navigate, though--one man's essential addition is |
174 | another man's pointless cruft. |
175 | |
176 | =item Does it create too much work? |
177 | |
178 | Work for the pumpking, work for Perl programmers, work for module |
179 | authors, ... Perl is supposed to be easy. |
180 | |
f6c51b38 |
181 | =item Patches speak louder than words |
182 | |
183 | Working code is always preferred to pie-in-the-sky ideas. A patch to |
184 | add a feature stands a much higher chance of making it to the language |
185 | than does a random feature request, no matter how fervently argued the |
186 | request might be. This ties into ``Will it be useful?'', as the fact |
187 | that someone took the time to make the patch demonstrates a strong |
188 | desire for the feature. |
189 | |
e8cd7eae |
190 | =back |
191 | |
192 | If you're on the list, you might hear the word ``core'' bandied |
193 | around. It refers to the standard distribution. ``Hacking on the |
194 | core'' means you're changing the C source code to the Perl |
195 | interpreter. ``A core module'' is one that ships with Perl. |
196 | |
197 | The source code to the Perl interpreter, in its different versions, is |
198 | kept in a repository managed by a revision control system (which is |
199 | currently the Perforce program, see http://perforce.com/). The |
200 | pumpkings and a few others have access to the repository to check in |
201 | changes. Periodically the pumpking for the development version of Perl |
202 | will release a new version, so the rest of the porters can see what's |
2be4c08b |
203 | changed. The current state of the main trunk of repository, and patches |
204 | that describe the individual changes that have happened since the last |
205 | public release are available at this location: |
206 | |
207 | ftp://ftp.linux.activestate.com/pub/staff/gsar/APC/ |
208 | |
209 | Selective parts are also visible via the rsync protocol. To get all |
210 | the individual changes to the mainline since the last development |
211 | release, use the following command: |
212 | |
6c7df0a8 |
213 | rsync -avz rsync://ftp.linux.activestate.com/perl-diffs perl-diffs |
2be4c08b |
214 | |
215 | Use this to get the latest source tree in full: |
216 | |
6c7df0a8 |
217 | rsync -avz rsync://ftp.linux.activestate.com/perl-current perl-current |
2be4c08b |
218 | |
219 | Needless to say, the source code in perl-current is usually in a perpetual |
220 | state of evolution. You should expect it to be very buggy. Do B<not> use |
221 | it for any purpose other than testing and development. |
e8cd7eae |
222 | |
223 | Always submit patches to I<perl5-porters@perl.org>. This lets other |
224 | porters review your patch, which catches a surprising number of errors |
f6c51b38 |
225 | in patches. Either use the diff program (available in source code |
226 | form from I<ftp://ftp.gnu.org/pub/gnu/>), or use Johan Vromans' |
227 | I<makepatch> (available from I<CPAN/authors/id/JV/>). Unified diffs |
228 | are preferred, but context diffs are accepted. Do not send RCS-style |
229 | diffs or diffs without context lines. More information is given in |
230 | the I<Porting/patching.pod> file in the Perl source distribution. |
231 | Please patch against the latest B<development> version (e.g., if |
232 | you're fixing a bug in the 5.005 track, patch against the latest |
233 | 5.005_5x version). Only patches that survive the heat of the |
234 | development branch get applied to maintenance versions. |
e8cd7eae |
235 | |
236 | Your patch should update the documentation and test suite. |
237 | |
238 | To report a bug in Perl, use the program I<perlbug> which comes with |
239 | Perl (if you can't get Perl to work, send mail to the address |
240 | I<perlbug@perl.com> or I<perlbug@perl.org>). Reporting bugs through |
241 | I<perlbug> feeds into the automated bug-tracking system, access to |
242 | which is provided through the web at I<http://bugs.perl.org/>. It |
243 | often pays to check the archives of the perl5-porters mailing list to |
244 | see whether the bug you're reporting has been reported before, and if |
245 | so whether it was considered a bug. See above for the location of |
246 | the searchable archives. |
247 | |
248 | The CPAN testers (I<http://testers.cpan.org/>) are a group of |
249 | volunteers who test CPAN modules on a variety of platforms. Perl Labs |
f6c51b38 |
250 | (I<http://labs.perl.org/>) automatically tests Perl source releases on |
251 | platforms and gives feedback to the CPAN testers mailing list. Both |
252 | efforts welcome volunteers. |
e8cd7eae |
253 | |
e8cd7eae |
254 | It's a good idea to read and lurk for a while before chipping in. |
255 | That way you'll get to see the dynamic of the conversations, learn the |
256 | personalities of the players, and hopefully be better prepared to make |
257 | a useful contribution when do you speak up. |
258 | |
259 | If after all this you still think you want to join the perl5-porters |
f6c51b38 |
260 | mailing list, send mail to I<perl5-porters-subscribe@perl.org>. To |
261 | unsubscribe, send mail to I<perl5-porters-unsubscribe@perl.org>. |
e8cd7eae |
262 | |
a422fd2d |
263 | To hack on the Perl guts, you'll need to read the following things: |
264 | |
265 | =over 3 |
266 | |
267 | =item L<perlguts> |
268 | |
269 | This is of paramount importance, since it's the documentation of what |
270 | goes where in the Perl source. Read it over a couple of times and it |
271 | might start to make sense - don't worry if it doesn't yet, because the |
272 | best way to study it is to read it in conjunction with poking at Perl |
273 | source, and we'll do that later on. |
274 | |
275 | You might also want to look at Gisle Aas's illustrated perlguts - |
276 | there's no guarantee that this will be absolutely up-to-date with the |
277 | latest documentation in the Perl core, but the fundamentals will be |
278 | right. (http://gisle.aas.no/perl/illguts/) |
279 | |
280 | =item L<perlxstut> and L<perlxs> |
281 | |
282 | A working knowledge of XSUB programming is incredibly useful for core |
283 | hacking; XSUBs use techniques drawn from the PP code, the portion of the |
284 | guts that actually executes a Perl program. It's a lot gentler to learn |
285 | those techniques from simple examples and explanation than from the core |
286 | itself. |
287 | |
288 | =item L<perlapi> |
289 | |
290 | The documentation for the Perl API explains what some of the internal |
291 | functions do, as well as the many macros used in the source. |
292 | |
293 | =item F<Porting/pumpkin.pod> |
294 | |
295 | This is a collection of words of wisdom for a Perl porter; some of it is |
296 | only useful to the pumpkin holder, but most of it applies to anyone |
297 | wanting to go about Perl development. |
298 | |
299 | =item The perl5-porters FAQ |
300 | |
301 | This is posted to perl5-porters at the beginning on every month, and |
302 | should be available from http://perlhacker.org/p5p-faq; alternatively, |
303 | you can get the FAQ emailed to you by sending mail to |
304 | C<perl5-porters-faq@perl.org>. It contains hints on reading |
305 | perl5-porters, information on how perl5-porters works and how Perl |
306 | development in general works. |
307 | |
308 | =back |
309 | |
310 | =head2 Finding Your Way Around |
311 | |
312 | Perl maintenance can be split into a number of areas, and certain people |
313 | (pumpkins) will have responsibility for each area. These areas sometimes |
314 | correspond to files or directories in the source kit. Among the areas are: |
315 | |
316 | =over 3 |
317 | |
318 | =item Core modules |
319 | |
320 | Modules shipped as part of the Perl core live in the F<lib/> and F<ext/> |
321 | subdirectories: F<lib/> is for the pure-Perl modules, and F<ext/> |
322 | contains the core XS modules. |
323 | |
324 | =item Documentation |
325 | |
326 | Documentation maintenance includes looking after everything in the |
327 | F<pod/> directory, (as well as contributing new documentation) and |
328 | the documentation to the modules in core. |
329 | |
330 | =item Configure |
331 | |
332 | The configure process is the way we make Perl portable across the |
333 | myriad of operating systems it supports. Responsibility for the |
334 | configure, build and installation process, as well as the overall |
335 | portability of the core code rests with the configure pumpkin - others |
336 | help out with individual operating systems. |
337 | |
338 | The files involved are the operating system directories, (F<win32/>, |
339 | F<os2/>, F<vms/> and so on) the shell scripts which generate F<config.h> |
340 | and F<Makefile>, as well as the metaconfig files which generate |
341 | F<Configure>. (metaconfig isn't included in the core distribution.) |
342 | |
343 | =item Interpreter |
344 | |
345 | And of course, there's the core of the Perl interpreter itself. Let's |
346 | have a look at that in a little more detail. |
347 | |
348 | =back |
349 | |
350 | Before we leave looking at the layout, though, don't forget that |
351 | F<MANIFEST> contains not only the file names in the Perl distribution, |
352 | but short descriptions of what's in them, too. For an overview of the |
353 | important files, try this: |
354 | |
355 | perl -lne 'print if /^[^\/]+\.[ch]\s+/' MANIFEST |
356 | |
357 | =head2 Elements of the interpreter |
358 | |
359 | The work of the interpreter has two main stages: compiling the code |
360 | into the internal representation, or bytecode, and then executing it. |
361 | L<perlguts/Compiled code> explains exactly how the compilation stage |
362 | happens. |
363 | |
364 | Here is a short breakdown of perl's operation: |
365 | |
366 | =over 3 |
367 | |
368 | =item Startup |
369 | |
370 | The action begins in F<perlmain.c>. (or F<miniperlmain.c> for miniperl) |
371 | This is very high-level code, enough to fit on a single screen, and it |
372 | resembles the code found in L<perlembed>; most of the real action takes |
373 | place in F<perl.c> |
374 | |
375 | First, F<perlmain.c> allocates some memory and constructs a Perl |
376 | interpreter: |
377 | |
378 | 1 PERL_SYS_INIT3(&argc,&argv,&env); |
379 | 2 |
380 | 3 if (!PL_do_undump) { |
381 | 4 my_perl = perl_alloc(); |
382 | 5 if (!my_perl) |
383 | 6 exit(1); |
384 | 7 perl_construct(my_perl); |
385 | 8 PL_perl_destruct_level = 0; |
386 | 9 } |
387 | |
388 | Line 1 is a macro, and its definition is dependent on your operating |
389 | system. Line 3 references C<PL_do_undump>, a global variable - all |
390 | global variables in Perl start with C<PL_>. This tells you whether the |
391 | current running program was created with the C<-u> flag to perl and then |
392 | F<undump>, which means it's going to be false in any sane context. |
393 | |
394 | Line 4 calls a function in F<perl.c> to allocate memory for a Perl |
395 | interpreter. It's quite a simple function, and the guts of it looks like |
396 | this: |
397 | |
398 | my_perl = (PerlInterpreter*)PerlMem_malloc(sizeof(PerlInterpreter)); |
399 | |
400 | Here you see an example of Perl's system abstraction, which we'll see |
401 | later: C<PerlMem_malloc> is either your system's C<malloc>, or Perl's |
402 | own C<malloc> as defined in F<malloc.c> if you selected that option at |
403 | configure time. |
404 | |
405 | Next, in line 7, we construct the interpreter; this sets up all the |
406 | special variables that Perl needs, the stacks, and so on. |
407 | |
408 | Now we pass Perl the command line options, and tell it to go: |
409 | |
410 | exitstatus = perl_parse(my_perl, xs_init, argc, argv, (char **)NULL); |
411 | if (!exitstatus) { |
412 | exitstatus = perl_run(my_perl); |
413 | } |
414 | |
415 | |
416 | C<perl_parse> is actually a wrapper around C<S_parse_body>, as defined |
417 | in F<perl.c>, which processes the command line options, sets up any |
418 | statically linked XS modules, opens the program and calls C<yyparse> to |
419 | parse it. |
420 | |
421 | =item Parsing |
422 | |
423 | The aim of this stage is to take the Perl source, and turn it into an op |
424 | tree. We'll see what one of those looks like later. Strictly speaking, |
425 | there's three things going on here. |
426 | |
427 | C<yyparse>, the parser, lives in F<perly.c>, although you're better off |
428 | reading the original YACC input in F<perly.y>. (Yes, Virginia, there |
429 | B<is> a YACC grammar for Perl!) The job of the parser is to take your |
430 | code and `understand' it, splitting it into sentences, deciding which |
431 | operands go with which operators and so on. |
432 | |
433 | The parser is nobly assisted by the lexer, which chunks up your input |
434 | into tokens, and decides what type of thing each token is: a variable |
435 | name, an operator, a bareword, a subroutine, a core function, and so on. |
436 | The main point of entry to the lexer is C<yylex>, and that and its |
437 | associated routines can be found in F<toke.c>. Perl isn't much like |
438 | other computer languages; it's highly context sensitive at times, it can |
439 | be tricky to work out what sort of token something is, or where a token |
440 | ends. As such, there's a lot of interplay between the tokeniser and the |
441 | parser, which can get pretty frightening if you're not used to it. |
442 | |
443 | As the parser understands a Perl program, it builds up a tree of |
444 | operations for the interpreter to perform during execution. The routines |
445 | which construct and link together the various operations are to be found |
446 | in F<op.c>, and will be examined later. |
447 | |
448 | =item Optimization |
449 | |
450 | Now the parsing stage is complete, and the finished tree represents |
451 | the operations that the Perl interpreter needs to perform to execute our |
452 | program. Next, Perl does a dry run over the tree looking for |
453 | optimisations: constant expressions such as C<3 + 4> will be computed |
454 | now, and the optimizer will also see if any multiple operations can be |
455 | replaced with a single one. For instance, to fetch the variable C<$foo>, |
456 | instead of grabbing the glob C<*foo> and looking at the scalar |
457 | component, the optimizer fiddles the op tree to use a function which |
458 | directly looks up the scalar in question. The main optimizer is C<peep> |
459 | in F<op.c>, and many ops have their own optimizing functions. |
460 | |
461 | =item Running |
462 | |
463 | Now we're finally ready to go: we have compiled Perl byte code, and all |
464 | that's left to do is run it. The actual execution is done by the |
465 | C<runops_standard> function in F<run.c>; more specifically, it's done by |
466 | these three innocent looking lines: |
467 | |
468 | while ((PL_op = CALL_FPTR(PL_op->op_ppaddr)(aTHX))) { |
469 | PERL_ASYNC_CHECK(); |
470 | } |
471 | |
472 | You may be more comfortable with the Perl version of that: |
473 | |
474 | PERL_ASYNC_CHECK() while $Perl::op = &{$Perl::op->{function}}; |
475 | |
476 | Well, maybe not. Anyway, each op contains a function pointer, which |
477 | stipulates the function which will actually carry out the operation. |
478 | This function will return the next op in the sequence - this allows for |
479 | things like C<if> which choose the next op dynamically at run time. |
480 | The C<PERL_ASYNC_CHECK> makes sure that things like signals interrupt |
481 | execution if required. |
482 | |
483 | The actual functions called are known as PP code, and they're spread |
484 | between four files: F<pp_hot.c> contains the `hot' code, which is most |
485 | often used and highly optimized, F<pp_sys.c> contains all the |
486 | system-specific functions, F<pp_ctl.c> contains the functions which |
487 | implement control structures (C<if>, C<while> and the like) and F<pp.c> |
488 | contains everything else. These are, if you like, the C code for Perl's |
489 | built-in functions and operators. |
490 | |
491 | =back |
492 | |
493 | =head2 Internal Variable Types |
494 | |
495 | You should by now have had a look at L<perlguts>, which tells you about |
496 | Perl's internal variable types: SVs, HVs, AVs and the rest. If not, do |
497 | that now. |
498 | |
499 | These variables are used not only to represent Perl-space variables, but |
500 | also any constants in the code, as well as some structures completely |
501 | internal to Perl. The symbol table, for instance, is an ordinary Perl |
502 | hash. Your code is represented by an SV as it's read into the parser; |
503 | any program files you call are opened via ordinary Perl filehandles, and |
504 | so on. |
505 | |
506 | The core L<Devel::Peek|Devel::Peek> module lets us examine SVs from a |
507 | Perl program. Let's see, for instance, how Perl treats the constant |
508 | C<"hello">. |
509 | |
510 | % perl -MDevel::Peek -e 'Dump("hello")' |
511 | 1 SV = PV(0xa041450) at 0xa04ecbc |
512 | 2 REFCNT = 1 |
513 | 3 FLAGS = (POK,READONLY,pPOK) |
514 | 4 PV = 0xa0484e0 "hello"\0 |
515 | 5 CUR = 5 |
516 | 6 LEN = 6 |
517 | |
518 | Reading C<Devel::Peek> output takes a bit of practise, so let's go |
519 | through it line by line. |
520 | |
521 | Line 1 tells us we're looking at an SV which lives at C<0xa04ecbc> in |
522 | memory. SVs themselves are very simple structures, but they contain a |
523 | pointer to a more complex structure. In this case, it's a PV, a |
524 | structure which holds a string value, at location C<0xa041450>. Line 2 |
525 | is the reference count; there are no other references to this data, so |
526 | it's 1. |
527 | |
528 | Line 3 are the flags for this SV - it's OK to use it as a PV, it's a |
529 | read-only SV (because it's a constant) and the data is a PV internally. |
530 | Next we've got the contents of the string, starting at location |
531 | C<0xa0484e0>. |
532 | |
533 | Line 5 gives us the current length of the string - note that this does |
534 | B<not> include the null terminator. Line 6 is not the length of the |
535 | string, but the length of the currently allocated buffer; as the string |
536 | grows, Perl automatically extends the available storage via a routine |
537 | called C<SvGROW>. |
538 | |
539 | You can get at any of these quantities from C very easily; just add |
540 | C<Sv> to the name of the field shown in the snippet, and you've got a |
541 | macro which will return the value: C<SvCUR(sv)> returns the current |
542 | length of the string, C<SvREFCOUNT(sv)> returns the reference count, |
543 | C<SvPV(sv, len)> returns the string itself with its length, and so on. |
544 | More macros to manipulate these properties can be found in L<perlguts>. |
545 | |
546 | Let's take an example of manipulating a PV, from C<sv_catpvn>, in F<sv.c> |
547 | |
548 | 1 void |
549 | 2 Perl_sv_catpvn(pTHX_ register SV *sv, register const char *ptr, register STRLEN len) |
550 | 3 { |
551 | 4 STRLEN tlen; |
552 | 5 char *junk; |
553 | |
554 | 6 junk = SvPV_force(sv, tlen); |
555 | 7 SvGROW(sv, tlen + len + 1); |
556 | 8 if (ptr == junk) |
557 | 9 ptr = SvPVX(sv); |
558 | 10 Move(ptr,SvPVX(sv)+tlen,len,char); |
559 | 11 SvCUR(sv) += len; |
560 | 12 *SvEND(sv) = '\0'; |
561 | 13 (void)SvPOK_only_UTF8(sv); /* validate pointer */ |
562 | 14 SvTAINT(sv); |
563 | 15 } |
564 | |
565 | This is a function which adds a string, C<ptr>, of length C<len> onto |
566 | the end of the PV stored in C<sv>. The first thing we do in line 6 is |
567 | make sure that the SV B<has> a valid PV, by calling the C<SvPV_force> |
568 | macro to force a PV. As a side effect, C<tlen> gets set to the current |
569 | value of the PV, and the PV itself is returned to C<junk>. |
570 | |
b1866b2d |
571 | In line 7, we make sure that the SV will have enough room to accommodate |
a422fd2d |
572 | the old string, the new string and the null terminator. If C<LEN> isn't |
573 | big enough, C<SvGROW> will reallocate space for us. |
574 | |
575 | Now, if C<junk> is the same as the string we're trying to add, we can |
576 | grab the string directly from the SV; C<SvPVX> is the address of the PV |
577 | in the SV. |
578 | |
579 | Line 10 does the actual catenation: the C<Move> macro moves a chunk of |
580 | memory around: we move the string C<ptr> to the end of the PV - that's |
581 | the start of the PV plus its current length. We're moving C<len> bytes |
582 | of type C<char>. After doing so, we need to tell Perl we've extended the |
583 | string, by altering C<CUR> to reflect the new length. C<SvEND> is a |
584 | macro which gives us the end of the string, so that needs to be a |
585 | C<"\0">. |
586 | |
587 | Line 13 manipulates the flags; since we've changed the PV, any IV or NV |
588 | values will no longer be valid: if we have C<$a=10; $a.="6";> we don't |
589 | want to use the old IV of 10. C<SvPOK_only_utf8> is a special UTF8-aware |
590 | version of C<SvPOK_only>, a macro which turns off the IOK and NOK flags |
591 | and turns on POK. The final C<SvTAINT> is a macro which launders tainted |
592 | data if taint mode is turned on. |
593 | |
594 | AVs and HVs are more complicated, but SVs are by far the most common |
595 | variable type being thrown around. Having seen something of how we |
596 | manipulate these, let's go on and look at how the op tree is |
597 | constructed. |
598 | |
599 | =head2 Op Trees |
600 | |
601 | First, what is the op tree, anyway? The op tree is the parsed |
602 | representation of your program, as we saw in our section on parsing, and |
603 | it's the sequence of operations that Perl goes through to execute your |
604 | program, as we saw in L</Running>. |
605 | |
606 | An op is a fundamental operation that Perl can perform: all the built-in |
607 | functions and operators are ops, and there are a series of ops which |
608 | deal with concepts the interpreter needs internally - entering and |
609 | leaving a block, ending a statement, fetching a variable, and so on. |
610 | |
611 | The op tree is connected in two ways: you can imagine that there are two |
612 | "routes" through it, two orders in which you can traverse the tree. |
613 | First, parse order reflects how the parser understood the code, and |
614 | secondly, execution order tells perl what order to perform the |
615 | operations in. |
616 | |
617 | The easiest way to examine the op tree is to stop Perl after it has |
618 | finished parsing, and get it to dump out the tree. This is exactly what |
619 | the compiler backends L<B::Terse|B::Terse> and L<B::Debug|B::Debug> do. |
620 | |
621 | Let's have a look at how Perl sees C<$a = $b + $c>: |
622 | |
623 | % perl -MO=Terse -e '$a=$b+$c' |
624 | 1 LISTOP (0x8179888) leave |
625 | 2 OP (0x81798b0) enter |
626 | 3 COP (0x8179850) nextstate |
627 | 4 BINOP (0x8179828) sassign |
628 | 5 BINOP (0x8179800) add [1] |
629 | 6 UNOP (0x81796e0) null [15] |
630 | 7 SVOP (0x80fafe0) gvsv GV (0x80fa4cc) *b |
631 | 8 UNOP (0x81797e0) null [15] |
632 | 9 SVOP (0x8179700) gvsv GV (0x80efeb0) *c |
633 | 10 UNOP (0x816b4f0) null [15] |
634 | 11 SVOP (0x816dcf0) gvsv GV (0x80fa460) *a |
635 | |
636 | Let's start in the middle, at line 4. This is a BINOP, a binary |
637 | operator, which is at location C<0x8179828>. The specific operator in |
638 | question is C<sassign> - scalar assignment - and you can find the code |
639 | which implements it in the function C<pp_sassign> in F<pp_hot.c>. As a |
640 | binary operator, it has two children: the add operator, providing the |
641 | result of C<$b+$c>, is uppermost on line 5, and the left hand side is on |
642 | line 10. |
643 | |
644 | Line 10 is the null op: this does exactly nothing. What is that doing |
645 | there? If you see the null op, it's a sign that something has been |
646 | optimized away after parsing. As we mentioned in L</Optimization>, |
647 | the optimization stage sometimes converts two operations into one, for |
648 | example when fetching a scalar variable. When this happens, instead of |
649 | rewriting the op tree and cleaning up the dangling pointers, it's easier |
650 | just to replace the redundant operation with the null op. Originally, |
651 | the tree would have looked like this: |
652 | |
653 | 10 SVOP (0x816b4f0) rv2sv [15] |
654 | 11 SVOP (0x816dcf0) gv GV (0x80fa460) *a |
655 | |
656 | That is, fetch the C<a> entry from the main symbol table, and then look |
657 | at the scalar component of it: C<gvsv> (C<pp_gvsv> into F<pp_hot.c>) |
658 | happens to do both these things. |
659 | |
660 | The right hand side, starting at line 5 is similar to what we've just |
661 | seen: we have the C<add> op (C<pp_add> also in F<pp_hot.c>) add together |
662 | two C<gvsv>s. |
663 | |
664 | Now, what's this about? |
665 | |
666 | 1 LISTOP (0x8179888) leave |
667 | 2 OP (0x81798b0) enter |
668 | 3 COP (0x8179850) nextstate |
669 | |
670 | C<enter> and C<leave> are scoping ops, and their job is to perform any |
671 | housekeeping every time you enter and leave a block: lexical variables |
672 | are tidied up, unreferenced variables are destroyed, and so on. Every |
673 | program will have those first three lines: C<leave> is a list, and its |
674 | children are all the statements in the block. Statements are delimited |
675 | by C<nextstate>, so a block is a collection of C<nextstate> ops, with |
676 | the ops to be performed for each statement being the children of |
677 | C<nextstate>. C<enter> is a single op which functions as a marker. |
678 | |
679 | That's how Perl parsed the program, from top to bottom: |
680 | |
681 | Program |
682 | | |
683 | Statement |
684 | | |
685 | = |
686 | / \ |
687 | / \ |
688 | $a + |
689 | / \ |
690 | $b $c |
691 | |
692 | However, it's impossible to B<perform> the operations in this order: |
693 | you have to find the values of C<$b> and C<$c> before you add them |
694 | together, for instance. So, the other thread that runs through the op |
695 | tree is the execution order: each op has a field C<op_next> which points |
696 | to the next op to be run, so following these pointers tells us how perl |
697 | executes the code. We can traverse the tree in this order using |
698 | the C<exec> option to C<B::Terse>: |
699 | |
700 | % perl -MO=Terse,exec -e '$a=$b+$c' |
701 | 1 OP (0x8179928) enter |
702 | 2 COP (0x81798c8) nextstate |
703 | 3 SVOP (0x81796c8) gvsv GV (0x80fa4d4) *b |
704 | 4 SVOP (0x8179798) gvsv GV (0x80efeb0) *c |
705 | 5 BINOP (0x8179878) add [1] |
706 | 6 SVOP (0x816dd38) gvsv GV (0x80fa468) *a |
707 | 7 BINOP (0x81798a0) sassign |
708 | 8 LISTOP (0x8179900) leave |
709 | |
710 | This probably makes more sense for a human: enter a block, start a |
711 | statement. Get the values of C<$b> and C<$c>, and add them together. |
712 | Find C<$a>, and assign one to the other. Then leave. |
713 | |
714 | The way Perl builds up these op trees in the parsing process can be |
715 | unravelled by examining F<perly.y>, the YACC grammar. Let's take the |
716 | piece we need to construct the tree for C<$a = $b + $c> |
717 | |
718 | 1 term : term ASSIGNOP term |
719 | 2 { $$ = newASSIGNOP(OPf_STACKED, $1, $2, $3); } |
720 | 3 | term ADDOP term |
721 | 4 { $$ = newBINOP($2, 0, scalar($1), scalar($3)); } |
722 | |
723 | If you're not used to reading BNF grammars, this is how it works: You're |
724 | fed certain things by the tokeniser, which generally end up in upper |
725 | case. Here, C<ADDOP>, is provided when the tokeniser sees C<+> in your |
726 | code. C<ASSIGNOP> is provided when C<=> is used for assigning. These are |
727 | `terminal symbols', because you can't get any simpler than them. |
728 | |
729 | The grammar, lines one and three of the snippet above, tells you how to |
730 | build up more complex forms. These complex forms, `non-terminal symbols' |
731 | are generally placed in lower case. C<term> here is a non-terminal |
732 | symbol, representing a single expression. |
733 | |
734 | The grammar gives you the following rule: you can make the thing on the |
735 | left of the colon if you see all the things on the right in sequence. |
736 | This is called a "reduction", and the aim of parsing is to completely |
737 | reduce the input. There are several different ways you can perform a |
738 | reduction, separated by vertical bars: so, C<term> followed by C<=> |
739 | followed by C<term> makes a C<term>, and C<term> followed by C<+> |
740 | followed by C<term> can also make a C<term>. |
741 | |
742 | So, if you see two terms with an C<=> or C<+>, between them, you can |
743 | turn them into a single expression. When you do this, you execute the |
744 | code in the block on the next line: if you see C<=>, you'll do the code |
745 | in line 2. If you see C<+>, you'll do the code in line 4. It's this code |
746 | which contributes to the op tree. |
747 | |
748 | | term ADDOP term |
749 | { $$ = newBINOP($2, 0, scalar($1), scalar($3)); } |
750 | |
751 | What this does is creates a new binary op, and feeds it a number of |
752 | variables. The variables refer to the tokens: C<$1> is the first token in |
753 | the input, C<$2> the second, and so on - think regular expression |
754 | backreferences. C<$$> is the op returned from this reduction. So, we |
755 | call C<newBINOP> to create a new binary operator. The first parameter to |
756 | C<newBINOP>, a function in F<op.c>, is the op type. It's an addition |
757 | operator, so we want the type to be C<ADDOP>. We could specify this |
758 | directly, but it's right there as the second token in the input, so we |
759 | use C<$2>. The second parameter is the op's flags: 0 means `nothing |
760 | special'. Then the things to add: the left and right hand side of our |
761 | expression, in scalar context. |
762 | |
763 | =head2 Stacks |
764 | |
765 | When perl executes something like C<addop>, how does it pass on its |
766 | results to the next op? The answer is, through the use of stacks. Perl |
767 | has a number of stacks to store things it's currently working on, and |
768 | we'll look at the three most important ones here. |
769 | |
770 | =over 3 |
771 | |
772 | =item Argument stack |
773 | |
774 | Arguments are passed to PP code and returned from PP code using the |
775 | argument stack, C<ST>. The typical way to handle arguments is to pop |
776 | them off the stack, deal with them how you wish, and then push the result |
777 | back onto the stack. This is how, for instance, the cosine operator |
778 | works: |
779 | |
780 | NV value; |
781 | value = POPn; |
782 | value = Perl_cos(value); |
783 | XPUSHn(value); |
784 | |
785 | We'll see a more tricky example of this when we consider Perl's macros |
786 | below. C<POPn> gives you the NV (floating point value) of the top SV on |
787 | the stack: the C<$x> in C<cos($x)>. Then we compute the cosine, and push |
788 | the result back as an NV. The C<X> in C<XPUSHn> means that the stack |
789 | should be extended if necessary - it can't be necessary here, because we |
790 | know there's room for one more item on the stack, since we've just |
791 | removed one! The C<XPUSH*> macros at least guarantee safety. |
792 | |
793 | Alternatively, you can fiddle with the stack directly: C<SP> gives you |
794 | the first element in your portion of the stack, and C<TOP*> gives you |
795 | the top SV/IV/NV/etc. on the stack. So, for instance, to do unary |
796 | negation of an integer: |
797 | |
798 | SETi(-TOPi); |
799 | |
800 | Just set the integer value of the top stack entry to its negation. |
801 | |
802 | Argument stack manipulation in the core is exactly the same as it is in |
803 | XSUBs - see L<perlxstut>, L<perlxs> and L<perlguts> for a longer |
804 | description of the macros used in stack manipulation. |
805 | |
806 | =item Mark stack |
807 | |
808 | I say `your portion of the stack' above because PP code doesn't |
809 | necessarily get the whole stack to itself: if your function calls |
810 | another function, you'll only want to expose the arguments aimed for the |
811 | called function, and not (necessarily) let it get at your own data. The |
812 | way we do this is to have a `virtual' bottom-of-stack, exposed to each |
813 | function. The mark stack keeps bookmarks to locations in the argument |
814 | stack usable by each function. For instance, when dealing with a tied |
815 | variable, (internally, something with `P' magic) Perl has to call |
816 | methods for accesses to the tied variables. However, we need to separate |
817 | the arguments exposed to the method to the argument exposed to the |
818 | original function - the store or fetch or whatever it may be. Here's how |
819 | the tied C<push> is implemented; see C<av_push> in F<av.c>: |
820 | |
821 | 1 PUSHMARK(SP); |
822 | 2 EXTEND(SP,2); |
823 | 3 PUSHs(SvTIED_obj((SV*)av, mg)); |
824 | 4 PUSHs(val); |
825 | 5 PUTBACK; |
826 | 6 ENTER; |
827 | 7 call_method("PUSH", G_SCALAR|G_DISCARD); |
828 | 8 LEAVE; |
829 | 9 POPSTACK; |
830 | |
831 | The lines which concern the mark stack are the first, fifth and last |
832 | lines: they save away, restore and remove the current position of the |
833 | argument stack. |
834 | |
835 | Let's examine the whole implementation, for practice: |
836 | |
837 | 1 PUSHMARK(SP); |
838 | |
839 | Push the current state of the stack pointer onto the mark stack. This is |
840 | so that when we've finished adding items to the argument stack, Perl |
841 | knows how many things we've added recently. |
842 | |
843 | 2 EXTEND(SP,2); |
844 | 3 PUSHs(SvTIED_obj((SV*)av, mg)); |
845 | 4 PUSHs(val); |
846 | |
847 | We're going to add two more items onto the argument stack: when you have |
848 | a tied array, the C<PUSH> subroutine receives the object and the value |
849 | to be pushed, and that's exactly what we have here - the tied object, |
850 | retrieved with C<SvTIED_obj>, and the value, the SV C<val>. |
851 | |
852 | 5 PUTBACK; |
853 | |
854 | Next we tell Perl to make the change to the global stack pointer: C<dSP> |
855 | only gave us a local copy, not a reference to the global. |
856 | |
857 | 6 ENTER; |
858 | 7 call_method("PUSH", G_SCALAR|G_DISCARD); |
859 | 8 LEAVE; |
860 | |
861 | C<ENTER> and C<LEAVE> localise a block of code - they make sure that all |
862 | variables are tidied up, everything that has been localised gets |
863 | its previous value returned, and so on. Think of them as the C<{> and |
864 | C<}> of a Perl block. |
865 | |
866 | To actually do the magic method call, we have to call a subroutine in |
867 | Perl space: C<call_method> takes care of that, and it's described in |
868 | L<perlcall>. We call the C<PUSH> method in scalar context, and we're |
869 | going to discard its return value. |
870 | |
871 | 9 POPSTACK; |
872 | |
873 | Finally, we remove the value we placed on the mark stack, since we |
874 | don't need it any more. |
875 | |
876 | =item Save stack |
877 | |
878 | C doesn't have a concept of local scope, so perl provides one. We've |
879 | seen that C<ENTER> and C<LEAVE> are used as scoping braces; the save |
880 | stack implements the C equivalent of, for example: |
881 | |
882 | { |
883 | local $foo = 42; |
884 | ... |
885 | } |
886 | |
887 | See L<perlguts/Localising Changes> for how to use the save stack. |
888 | |
889 | =back |
890 | |
891 | =head2 Millions of Macros |
892 | |
893 | One thing you'll notice about the Perl source is that it's full of |
894 | macros. Some have called the pervasive use of macros the hardest thing |
895 | to understand, others find it adds to clarity. Let's take an example, |
896 | the code which implements the addition operator: |
897 | |
898 | 1 PP(pp_add) |
899 | 2 { |
900 | 3 djSP; dATARGET; tryAMAGICbin(add,opASSIGN); |
901 | 4 { |
902 | 5 dPOPTOPnnrl_ul; |
903 | 6 SETn( left + right ); |
904 | 7 RETURN; |
905 | 8 } |
906 | 9 } |
907 | |
908 | Every line here (apart from the braces, of course) contains a macro. The |
909 | first line sets up the function declaration as Perl expects for PP code; |
910 | line 3 sets up variable declarations for the argument stack and the |
911 | target, the return value of the operation. Finally, it tries to see if |
912 | the addition operation is overloaded; if so, the appropriate subroutine |
913 | is called. |
914 | |
915 | Line 5 is another variable declaration - all variable declarations start |
916 | with C<d> - which pops from the top of the argument stack two NVs (hence |
917 | C<nn>) and puts them into the variables C<right> and C<left>, hence the |
918 | C<rl>. These are the two operands to the addition operator. Next, we |
919 | call C<SETn> to set the NV of the return value to the result of adding |
920 | the two values. This done, we return - the C<RETURN> macro makes sure |
921 | that our return value is properly handled, and we pass the next operator |
922 | to run back to the main run loop. |
923 | |
924 | Most of these macros are explained in L<perlapi>, and some of the more |
925 | important ones are explained in L<perlxs> as well. Pay special attention |
926 | to L<perlguts/Background and PERL_IMPLICIT_CONTEXT> for information on |
927 | the C<[pad]THX_?> macros. |
928 | |
929 | |
930 | =head2 Poking at Perl |
931 | |
932 | To really poke around with Perl, you'll probably want to build Perl for |
933 | debugging, like this: |
934 | |
935 | ./Configure -d -D optimize=-g |
936 | make |
937 | |
938 | C<-g> is a flag to the C compiler to have it produce debugging |
939 | information which will allow us to step through a running program. |
940 | F<Configure> will also turn on the C<DEBUGGING> compilation symbol which |
941 | enables all the internal debugging code in Perl. There are a whole bunch |
942 | of things you can debug with this: L<perlrun> lists them all, and the |
943 | best way to find out about them is to play about with them. The most |
944 | useful options are probably |
945 | |
946 | l Context (loop) stack processing |
947 | t Trace execution |
948 | o Method and overloading resolution |
949 | c String/numeric conversions |
950 | |
951 | Some of the functionality of the debugging code can be achieved using XS |
952 | modules. |
953 | |
954 | -Dr => use re 'debug' |
955 | -Dx => use O 'Debug' |
956 | |
957 | =head2 Using a source-level debugger |
958 | |
959 | If the debugging output of C<-D> doesn't help you, it's time to step |
960 | through perl's execution with a source-level debugger. |
961 | |
962 | =over 3 |
963 | |
964 | =item * |
965 | |
966 | We'll use C<gdb> for our examples here; the principles will apply to any |
967 | debugger, but check the manual of the one you're using. |
968 | |
969 | =back |
970 | |
971 | To fire up the debugger, type |
972 | |
973 | gdb ./perl |
974 | |
975 | You'll want to do that in your Perl source tree so the debugger can read |
976 | the source code. You should see the copyright message, followed by the |
977 | prompt. |
978 | |
979 | (gdb) |
980 | |
981 | C<help> will get you into the documentation, but here are the most |
982 | useful commands: |
983 | |
984 | =over 3 |
985 | |
986 | =item run [args] |
987 | |
988 | Run the program with the given arguments. |
989 | |
990 | =item break function_name |
991 | |
992 | =item break source.c:xxx |
993 | |
994 | Tells the debugger that we'll want to pause execution when we reach |
995 | either the named function (but see L</Function names>!) or the given |
996 | line in the named source file. |
997 | |
998 | =item step |
999 | |
1000 | Steps through the program a line at a time. |
1001 | |
1002 | =item next |
1003 | |
1004 | Steps through the program a line at a time, without descending into |
1005 | functions. |
1006 | |
1007 | =item continue |
1008 | |
1009 | Run until the next breakpoint. |
1010 | |
1011 | =item finish |
1012 | |
1013 | Run until the end of the current function, then stop again. |
1014 | |
1015 | =item |
1016 | |
1017 | Just pressing Enter will do the most recent operation again - it's a |
1018 | blessing when stepping through miles of source code. |
1019 | |
1020 | =item print |
1021 | |
1022 | Execute the given C code and print its results. B<WARNING>: Perl makes |
1023 | heavy use of macros, and F<gdb> is not aware of macros. You'll have to |
1024 | substitute them yourself. So, for instance, you can't say |
1025 | |
1026 | print SvPV_nolen(sv) |
1027 | |
1028 | but you have to say |
1029 | |
1030 | print Perl_sv_2pv_nolen(sv) |
1031 | |
1032 | You may find it helpful to have a "macro dictionary", which you can |
1033 | produce by saying C<cpp -dM perl.c | sort>. Even then, F<cpp> won't |
1034 | recursively apply the macros for you. |
1035 | |
1036 | =back |
1037 | |
1038 | =head2 Dumping Perl Data Structures |
1039 | |
1040 | One way to get around this macro hell is to use the dumping functions in |
1041 | F<dump.c>; these work a little like an internal |
1042 | L<Devel::Peek|Devel::Peek>, but they also cover OPs and other structures |
1043 | that you can't get at from Perl. Let's take an example. We'll use the |
1044 | C<$a = $b + $c> we used before, but give it a bit of context: |
1045 | C<$b = "6XXXX"; $c = 2.3;>. Where's a good place to stop and poke around? |
1046 | |
1047 | What about C<pp_add>, the function we examined earlier to implement the |
1048 | C<+> operator: |
1049 | |
1050 | (gdb) break Perl_pp_add |
1051 | Breakpoint 1 at 0x46249f: file pp_hot.c, line 309. |
1052 | |
1053 | Notice we use C<Perl_pp_add> and not C<pp_add> - see L<perlguts/Function Names>. |
1054 | With the breakpoint in place, we can run our program: |
1055 | |
1056 | (gdb) run -e '$b = "6XXXX"; $c = 2.3; $a = $b + $c' |
1057 | |
1058 | Lots of junk will go past as gdb reads in the relevant source files and |
1059 | libraries, and then: |
1060 | |
1061 | Breakpoint 1, Perl_pp_add () at pp_hot.c:309 |
1062 | 309 djSP; dATARGET; tryAMAGICbin(add,opASSIGN); |
1063 | (gdb) step |
1064 | 311 dPOPTOPnnrl_ul; |
1065 | (gdb) |
1066 | |
1067 | We looked at this bit of code before, and we said that C<dPOPTOPnnrl_ul> |
1068 | arranges for two C<NV>s to be placed into C<left> and C<right> - let's |
1069 | slightly expand it: |
1070 | |
1071 | #define dPOPTOPnnrl_ul NV right = POPn; \ |
1072 | SV *leftsv = TOPs; \ |
1073 | NV left = USE_LEFT(leftsv) ? SvNV(leftsv) : 0.0 |
1074 | |
1075 | C<POPn> takes the SV from the top of the stack and obtains its NV either |
1076 | directly (if C<SvNOK> is set) or by calling the C<sv_2nv> function. |
1077 | C<TOPs> takes the next SV from the top of the stack - yes, C<POPn> uses |
1078 | C<TOPs> - but doesn't remove it. We then use C<SvNV> to get the NV from |
1079 | C<leftsv> in the same way as before - yes, C<POPn> uses C<SvNV>. |
1080 | |
1081 | Since we don't have an NV for C<$b>, we'll have to use C<sv_2nv> to |
1082 | convert it. If we step again, we'll find ourselves there: |
1083 | |
1084 | Perl_sv_2nv (sv=0xa0675d0) at sv.c:1669 |
1085 | 1669 if (!sv) |
1086 | (gdb) |
1087 | |
1088 | We can now use C<Perl_sv_dump> to investigate the SV: |
1089 | |
1090 | SV = PV(0xa057cc0) at 0xa0675d0 |
1091 | REFCNT = 1 |
1092 | FLAGS = (POK,pPOK) |
1093 | PV = 0xa06a510 "6XXXX"\0 |
1094 | CUR = 5 |
1095 | LEN = 6 |
1096 | $1 = void |
1097 | |
1098 | We know we're going to get C<6> from this, so let's finish the |
1099 | subroutine: |
1100 | |
1101 | (gdb) finish |
1102 | Run till exit from #0 Perl_sv_2nv (sv=0xa0675d0) at sv.c:1671 |
1103 | 0x462669 in Perl_pp_add () at pp_hot.c:311 |
1104 | 311 dPOPTOPnnrl_ul; |
1105 | |
1106 | We can also dump out this op: the current op is always stored in |
1107 | C<PL_op>, and we can dump it with C<Perl_op_dump>. This'll give us |
1108 | similar output to L<B::Debug|B::Debug>. |
1109 | |
1110 | { |
1111 | 13 TYPE = add ===> 14 |
1112 | TARG = 1 |
1113 | FLAGS = (SCALAR,KIDS) |
1114 | { |
1115 | TYPE = null ===> (12) |
1116 | (was rv2sv) |
1117 | FLAGS = (SCALAR,KIDS) |
1118 | { |
1119 | 11 TYPE = gvsv ===> 12 |
1120 | FLAGS = (SCALAR) |
1121 | GV = main::b |
1122 | } |
1123 | } |
1124 | |
1125 | < finish this later > |
1126 | |
1127 | =head2 Patching |
1128 | |
1129 | All right, we've now had a look at how to navigate the Perl sources and |
1130 | some things you'll need to know when fiddling with them. Let's now get |
1131 | on and create a simple patch. Here's something Larry suggested: if a |
1132 | C<U> is the first active format during a C<pack>, (for example, |
1133 | C<pack "U3C8", @stuff>) then the resulting string should be treated as |
1134 | UTF8 encoded. |
1135 | |
1136 | How do we prepare to fix this up? First we locate the code in question - |
1137 | the C<pack> happens at runtime, so it's going to be in one of the F<pp> |
1138 | files. Sure enough, C<pp_pack> is in F<pp.c>. Since we're going to be |
1139 | altering this file, let's copy it to F<pp.c~>. |
1140 | |
1141 | Now let's look over C<pp_pack>: we take a pattern into C<pat>, and then |
1142 | loop over the pattern, taking each format character in turn into |
1143 | C<datum_type>. Then for each possible format character, we swallow up |
1144 | the other arguments in the pattern (a field width, an asterisk, and so |
1145 | on) and convert the next chunk input into the specified format, adding |
1146 | it onto the output SV C<cat>. |
1147 | |
1148 | How do we know if the C<U> is the first format in the C<pat>? Well, if |
1149 | we have a pointer to the start of C<pat> then, if we see a C<U> we can |
1150 | test whether we're still at the start of the string. So, here's where |
1151 | C<pat> is set up: |
1152 | |
1153 | STRLEN fromlen; |
1154 | register char *pat = SvPVx(*++MARK, fromlen); |
1155 | register char *patend = pat + fromlen; |
1156 | register I32 len; |
1157 | I32 datumtype; |
1158 | SV *fromstr; |
1159 | |
1160 | We'll have another string pointer in there: |
1161 | |
1162 | STRLEN fromlen; |
1163 | register char *pat = SvPVx(*++MARK, fromlen); |
1164 | register char *patend = pat + fromlen; |
1165 | + char *patcopy; |
1166 | register I32 len; |
1167 | I32 datumtype; |
1168 | SV *fromstr; |
1169 | |
1170 | And just before we start the loop, we'll set C<patcopy> to be the start |
1171 | of C<pat>: |
1172 | |
1173 | items = SP - MARK; |
1174 | MARK++; |
1175 | sv_setpvn(cat, "", 0); |
1176 | + patcopy = pat; |
1177 | while (pat < patend) { |
1178 | |
1179 | Now if we see a C<U> which was at the start of the string, we turn on |
1180 | the UTF8 flag for the output SV, C<cat>: |
1181 | |
1182 | + if (datumtype == 'U' && pat==patcopy+1) |
1183 | + SvUTF8_on(cat); |
1184 | if (datumtype == '#') { |
1185 | while (pat < patend && *pat != '\n') |
1186 | pat++; |
1187 | |
1188 | Remember that it has to be C<patcopy+1> because the first character of |
1189 | the string is the C<U> which has been swallowed into C<datumtype!> |
1190 | |
1191 | Oops, we forgot one thing: what if there are spaces at the start of the |
1192 | pattern? C<pack(" U*", @stuff)> will have C<U> as the first active |
1193 | character, even though it's not the first thing in the pattern. In this |
1194 | case, we have to advance C<patcopy> along with C<pat> when we see spaces: |
1195 | |
1196 | if (isSPACE(datumtype)) |
1197 | continue; |
1198 | |
1199 | needs to become |
1200 | |
1201 | if (isSPACE(datumtype)) { |
1202 | patcopy++; |
1203 | continue; |
1204 | } |
1205 | |
1206 | OK. That's the C part done. Now we must do two additional things before |
1207 | this patch is ready to go: we've changed the behaviour of Perl, and so |
1208 | we must document that change. We must also provide some more regression |
1209 | tests to make sure our patch works and doesn't create a bug somewhere |
1210 | else along the line. |
1211 | |
1212 | The regression tests for each operator live in F<t/op/>, and so we make |
1213 | a copy of F<t/op/pack.t> to F<t/op/pack.t~>. Now we can add our tests |
1214 | to the end. First, we'll test that the C<U> does indeed create Unicode |
1215 | strings: |
1216 | |
1217 | print 'not ' unless "1.20.300.4000" eq sprintf "%vd", pack("U*",1,20,300,4000); |
1218 | print "ok $test\n"; $test++; |
1219 | |
1220 | Now we'll test that we got that space-at-the-beginning business right: |
1221 | |
1222 | print 'not ' unless "1.20.300.4000" eq |
1223 | sprintf "%vd", pack(" U*",1,20,300,4000); |
1224 | print "ok $test\n"; $test++; |
1225 | |
1226 | And finally we'll test that we don't make Unicode strings if C<U> is B<not> |
1227 | the first active format: |
1228 | |
1229 | print 'not ' unless v1.20.300.4000 ne |
1230 | sprintf "%vd", pack("C0U*",1,20,300,4000); |
1231 | print "ok $test\n"; $test++; |
1232 | |
b1866b2d |
1233 | Mustn't forget to change the number of tests which appears at the top, or |
a422fd2d |
1234 | else the automated tester will get confused: |
1235 | |
1236 | -print "1..156\n"; |
1237 | +print "1..159\n"; |
1238 | |
1239 | We now compile up Perl, and run it through the test suite. Our new |
1240 | tests pass, hooray! |
1241 | |
1242 | Finally, the documentation. The job is never done until the paperwork is |
1243 | over, so let's describe the change we've just made. The relevant place |
1244 | is F<pod/perlfunc.pod>; again, we make a copy, and then we'll insert |
1245 | this text in the description of C<pack>: |
1246 | |
1247 | =item * |
1248 | |
1249 | If the pattern begins with a C<U>, the resulting string will be treated |
1250 | as Unicode-encoded. You can force UTF8 encoding on in a string with an |
1251 | initial C<U0>, and the bytes that follow will be interpreted as Unicode |
1252 | characters. If you don't want this to happen, you can begin your pattern |
1253 | with C<C0> (or anything else) to force Perl not to UTF8 encode your |
1254 | string, and then follow this with a C<U*> somewhere in your pattern. |
1255 | |
1256 | All done. Now let's create the patch. F<Porting/patching.pod> tells us |
1257 | that if we're making major changes, we should copy the entire directory |
1258 | to somewhere safe before we begin fiddling, and then do |
1259 | |
1260 | diff -ruN old new > patch |
1261 | |
1262 | However, we know which files we've changed, and we can simply do this: |
1263 | |
1264 | diff -u pp.c~ pp.c > patch |
1265 | diff -u t/op/pack.t~ t/op/pack.t >> patch |
1266 | diff -u pod/perlfunc.pod~ pod/perlfunc.pod >> patch |
1267 | |
1268 | We end up with a patch looking a little like this: |
1269 | |
1270 | --- pp.c~ Fri Jun 02 04:34:10 2000 |
1271 | +++ pp.c Fri Jun 16 11:37:25 2000 |
1272 | @@ -4375,6 +4375,7 @@ |
1273 | register I32 items; |
1274 | STRLEN fromlen; |
1275 | register char *pat = SvPVx(*++MARK, fromlen); |
1276 | + char *patcopy; |
1277 | register char *patend = pat + fromlen; |
1278 | register I32 len; |
1279 | I32 datumtype; |
1280 | @@ -4405,6 +4406,7 @@ |
1281 | ... |
1282 | |
1283 | And finally, we submit it, with our rationale, to perl5-porters. Job |
1284 | done! |
1285 | |
902b9dbf |
1286 | =head1 EXTERNAL TOOLS FOR DEBUGGING PERL |
1287 | |
1288 | Sometimes it helps to use external tools while debugging and |
1289 | testing Perl. This section tries to guide you through using |
1290 | some common testing and debugging tools with Perl. This is |
1291 | meant as a guide to interfacing these tools with Perl, not |
1292 | as any kind of guide to the use of the tools themselves. |
1293 | |
1294 | =head2 Rational Software's Purify |
1295 | |
1296 | Purify is a commercial tool that is helpful in identifying |
1297 | memory overruns, wild pointers, memory leaks and other such |
1298 | badness. Perl must be compiled in a specific way for |
1299 | optimal testing with Purify. Purify is available under |
1300 | Windows NT, Solaris, HP-UX, SGI, and Siemens Unix. |
1301 | |
1302 | The only currently known leaks happen when there are |
1303 | compile-time errors within eval or require. (Fixing these |
1304 | is non-trivial, unfortunately, but they must be fixed |
1305 | eventually.) |
1306 | |
1307 | =head2 Purify on Unix |
1308 | |
1309 | On Unix, Purify creates a new Perl binary. To get the most |
1310 | benefit out of Purify, you should create the perl to Purify |
1311 | using: |
1312 | |
1313 | sh Configure -Accflags=-DPURIFY -Doptimize='-g' \ |
1314 | -Uusemymalloc -Dusemultiplicity |
1315 | |
1316 | where these arguments mean: |
1317 | |
1318 | =over 4 |
1319 | |
1320 | =item -Accflags=-DPURIFY |
1321 | |
1322 | Disables Perl's arena memory allocation functions, as well as |
1323 | forcing use of memory allocation functions derived from the |
1324 | system malloc. |
1325 | |
1326 | =item -Doptimize='-g' |
1327 | |
1328 | Adds debugging information so that you see the exact source |
1329 | statements where the problem occurs. Without this flag, all |
1330 | you will see is the source filename of where the error occurred. |
1331 | |
1332 | =item -Uusemymalloc |
1333 | |
1334 | Disable Perl's malloc so that Purify can more closely monitor |
1335 | allocations and leaks. Using Perl's malloc will make Purify |
1336 | report most leaks in the "potential" leaks category. |
1337 | |
1338 | =item -Dusemultiplicity |
1339 | |
1340 | Enabling the multiplicity option allows perl to clean up |
1341 | thoroughly when the interpreter shuts down, which reduces the |
1342 | number of bogus leak reports from Purify. |
1343 | |
1344 | =back |
1345 | |
1346 | Once you've compiled a perl suitable for Purify'ing, then you |
1347 | can just: |
1348 | |
1349 | make pureperl |
1350 | |
1351 | which creates a binary named 'pureperl' that has been Purify'ed. |
1352 | This binary is used in place of the standard 'perl' binary |
1353 | when you want to debug Perl memory problems. |
1354 | |
1355 | As an example, to show any memory leaks produced during the |
1356 | standard Perl testset you would create and run the Purify'ed |
1357 | perl as: |
1358 | |
1359 | make pureperl |
1360 | cd t |
1361 | ../pureperl -I../lib harness |
1362 | |
1363 | which would run Perl on test.pl and report any memory problems. |
1364 | |
1365 | Purify outputs messages in "Viewer" windows by default. If |
1366 | you don't have a windowing environment or if you simply |
1367 | want the Purify output to unobtrusively go to a log file |
1368 | instead of to the interactive window, use these following |
1369 | options to output to the log file "perl.log": |
1370 | |
1371 | setenv PURIFYOPTIONS "-chain-length=25 -windows=no \ |
1372 | -log-file=perl.log -append-logfile=yes" |
1373 | |
1374 | If you plan to use the "Viewer" windows, then you only need this option: |
1375 | |
1376 | setenv PURIFYOPTIONS "-chain-length=25" |
1377 | |
1378 | =head2 Purify on NT |
1379 | |
1380 | Purify on Windows NT instruments the Perl binary 'perl.exe' |
1381 | on the fly. There are several options in the makefile you |
1382 | should change to get the most use out of Purify: |
1383 | |
1384 | =over 4 |
1385 | |
1386 | =item DEFINES |
1387 | |
1388 | You should add -DPURIFY to the DEFINES line so the DEFINES |
1389 | line looks something like: |
1390 | |
1391 | DEFINES = -DWIN32 -D_CONSOLE -DNO_STRICT $(CRYPT_FLAG) -DPURIFY=1 |
1392 | |
1393 | to disable Perl's arena memory allocation functions, as |
1394 | well as to force use of memory allocation functions derived |
1395 | from the system malloc. |
1396 | |
1397 | =item USE_MULTI = define |
1398 | |
1399 | Enabling the multiplicity option allows perl to clean up |
1400 | thoroughly when the interpreter shuts down, which reduces the |
1401 | number of bogus leak reports from Purify. |
1402 | |
1403 | =item #PERL_MALLOC = define |
1404 | |
1405 | Disable Perl's malloc so that Purify can more closely monitor |
1406 | allocations and leaks. Using Perl's malloc will make Purify |
1407 | report most leaks in the "potential" leaks category. |
1408 | |
1409 | =item CFG = Debug |
1410 | |
1411 | Adds debugging information so that you see the exact source |
1412 | statements where the problem occurs. Without this flag, all |
1413 | you will see is the source filename of where the error occurred. |
1414 | |
1415 | =back |
1416 | |
1417 | As an example, to show any memory leaks produced during the |
1418 | standard Perl testset you would create and run Purify as: |
1419 | |
1420 | cd win32 |
1421 | make |
1422 | cd ../t |
1423 | purify ../perl -I../lib harness |
1424 | |
1425 | which would instrument Perl in memory, run Perl on test.pl, |
1426 | then finally report any memory problems. |
1427 | |
a422fd2d |
1428 | =head2 CONCLUSION |
1429 | |
1430 | We've had a brief look around the Perl source, an overview of the stages |
1431 | F<perl> goes through when it's running your code, and how to use a |
902b9dbf |
1432 | debugger to poke at the Perl guts. We took a very simple problem and |
1433 | demonstrated how to solve it fully - with documentation, regression |
1434 | tests, and finally a patch for submission to p5p. Finally, we talked |
1435 | about how to use external tools to debug and test Perl. |
a422fd2d |
1436 | |
1437 | I'd now suggest you read over those references again, and then, as soon |
1438 | as possible, get your hands dirty. The best way to learn is by doing, |
1439 | so: |
1440 | |
1441 | =over 3 |
1442 | |
1443 | =item * |
1444 | |
1445 | Subscribe to perl5-porters, follow the patches and try and understand |
1446 | them; don't be afraid to ask if there's a portion you're not clear on - |
1447 | who knows, you may unearth a bug in the patch... |
1448 | |
1449 | =item * |
1450 | |
1451 | Keep up to date with the bleeding edge Perl distributions and get |
1452 | familiar with the changes. Try and get an idea of what areas people are |
1453 | working on and the changes they're making. |
1454 | |
1455 | =item * |
1456 | |
1457 | Find an area of Perl that seems interesting to you, and see if you can |
1458 | work out how it works. Scan through the source, and step over it in the |
1459 | debugger. Play, poke, investigate, fiddle! You'll probably get to |
1460 | understand not just your chosen area but a much wider range of F<perl>'s |
1461 | activity as well, and probably sooner than you'd think. |
1462 | |
1463 | =back |
1464 | |
1465 | =over 3 |
1466 | |
1467 | =item I<The Road goes ever on and on, down from the door where it began.> |
1468 | |
1469 | =back |
1470 | |
1471 | If you can do these things, you've started on the long road to Perl porting. |
1472 | Thanks for wanting to help make Perl better - and happy hacking! |
1473 | |
e8cd7eae |
1474 | =head1 AUTHOR |
1475 | |
1476 | This document was written by Nathan Torkington, and is maintained by |
1477 | the perl5-porters mailing list. |
1478 | |