3 Pumpkin - Notes on handling the Perl Patch Pumpkin And Porting Perl
7 There is no simple synopsis, yet.
11 This document attempts to begin to describe some of the considerations
12 involved in patching, porting, and maintaining perl.
14 This document is still under construction, and still subject to
15 significant changes. Still, I hope parts of it will be useful,
16 so I'm releasing it even though it's not done.
18 For the most part, it's a collection of anecdotal information that
19 already assumes some familiarity with the Perl sources. I really need
20 an introductory section that describes the organization of the sources
21 and all the various auxiliary files that are part of the distribution.
23 =head1 Where Do I Get Perl Sources and Related Material?
25 The Comprehensive Perl Archive Network (or CPAN) is the place to go.
26 There are many mirrors, but the easiest thing to use is probably
27 L<http://www.cpan.org/README.html> , which automatically points you to a
28 mirror site "close" to you.
30 =head2 Perl5-porters mailing list
32 The mailing list perl5-porters@perl.org
33 is the main group working with the development of perl. If you're
34 interested in all the latest developments, you should definitely
35 subscribe. The list is high volume, but generally has a
36 fairly low noise level.
38 To subscribe to perl5-porters, send an email to
40 perl5-porters-subscribe@perl.org
42 Archives of the list are held at:
44 https://lists.perl.org/list/perl5-porters.html
46 =head1 How are Perl Releases Numbered?
48 Beginning with v5.6.0, even versions will stand for maintenance releases
49 and odd versions for development releases, i.e., v5.6.x for maintenance
50 releases, and v5.7.x for development releases. Before v5.6.0, subversions
51 _01 through _49 were reserved for bug-fix maintenance releases, and
52 subversions _50 through _99 for unstable development versions.
54 For example, in v5.6.1, the revision number is 5, the version is 6,
55 and 1 is the subversion.
57 For compatibility with the older numbering scheme the composite floating
58 point version number continues to be available as the magic variable $],
59 and amounts to C<$revision + $version/1000 + $subversion/100000>. This
60 can still be used in comparisons.
62 print "You've got an old perl\n" if $] < 5.005_03;
64 In addition, the version is also available as a string in $^V.
66 print "You've got a new perl\n" if $^V and $^V ge v5.6.0;
68 You can also require particular version (or later) with:
72 or using the new syntax available only from v5.6 onward:
76 At some point in the future, we may need to decide what to call the
77 next big revision. In the .package file used by metaconfig to
78 generate Configure, there are two variables that might be relevant:
79 $baserev=5 and $package=perl5.
81 Perl releases produced by the members of perl5-porters are usually
82 available on CPAN in the F<src/5.0/maint> and F<src/5.0/devel>
85 =head2 Maintenance and Development Subversions
87 The first rule of maintenance work is "First, do no harm."
89 Trial releases of bug-fix maintenance releases are announced on
90 perl5-porters. Trial releases use the new subversion number (to avoid
91 testers installing it over the previous release) and include a 'local
92 patch' entry in F<patchlevel.h>. The distribution file contains the
93 string C<MAINT_TRIAL> to make clear that the file is not meant for
96 In general, the names of official distribution files for the public
97 always match the regular expression:
99 ^perl\d+\.(\d+)\.\d+(-MAINT_TRIAL_\d+)\.tar\.gz$
101 C<$1> in the pattern is always an even number for maintenance
102 versions, and odd for developer releases.
104 In the past, release managers sometimes invented naming conventions on the fly.
105 If you are releasing perl, before you invent a new name for any of the three
106 types of perl distributions, please inform the people from the CPAN who are
107 doing indexing and provide the trees of symlinks and the like. They will have
108 to know I<in advance> what you decide.
110 =head2 Why is it called the patch pumpkin?
112 Chip Salzenberg gets credit for that, with a nod to his cow orker,
113 David Croy. We had passed around various names (baton, token, hot
114 potato) but none caught on. Then, Chip asked:
118 Who has the patch pumpkin?
120 To explain: David Croy once told me once that at a previous job,
121 there was one tape drive and multiple systems that used it for backups.
122 But instead of some high-tech exclusion software, they used a low-tech
123 method to prevent multiple simultaneous backups: a stuffed pumpkin.
124 No one was allowed to make backups unless they had the "backup pumpkin".
130 =head1 Philosophical Issues in Patching and Porting Perl
132 There are no absolute rules, but there are some general guidelines I
133 have tried to follow as I apply patches to the perl sources.
134 (This section is still under construction.)
136 =head2 Solve problems as generally as possible
138 Never implement a specific restricted solution to a problem when you
139 can solve the same problem in a more general, flexible way.
141 For example, for dynamic loading to work on some SVR4 systems, we had
142 to build a shared libperl.so library. In order to build "FAT" binaries
143 on NeXT 4.0 systems, we had to build a special libperl library. Rather
144 than continuing to build a contorted nest of special cases, I
145 generalized the process of building libperl so that NeXT and SVR4 users
146 could still get their work done, but others could build a shared
147 libperl if they wanted to as well.
149 Contain your changes carefully. Assume nothing about other operating
150 systems, not even closely related ones. Your changes must not affect
153 Spy shamelessly on how similar patching or porting issues have been
156 If feasible, try to keep filenames 8.3-compliant to humor those poor
157 souls that get joy from running Perl under such dire limitations.
158 There's a script, F<check83.pl>, for keeping your nose 8.3-clean.
159 In a similar vein, do not create files or directories which differ only
160 in case (upper versus lower).
162 =head2 Seek consensus on major changes
164 If you are making big changes, don't do it in secret. Discuss the
165 ideas in advance on perl5-porters.
167 =head2 Keep the documentation up-to-date
169 If your changes may affect how users use perl, then check to be sure
170 that the documentation is in sync with your changes. Be sure to
171 check all the files F<pod/*.pod> and also the F<INSTALL> document.
173 Consider writing the appropriate documentation first and then
174 implementing your change to correspond to the documentation.
176 =head2 Avoid machine-specific #ifdef's
178 To the extent reasonable, try to avoid machine-specific #ifdef's in
179 the sources. Instead, use feature-specific #ifdef's. The reason is
180 that the machine-specific #ifdef's may not be valid across major
181 releases of the operating system. Further, the feature-specific tests
182 may help out folks on another platform who have the same problem.
184 =head2 Machine-specific files
190 If you have many machine-specific #defines or #includes, consider
191 creating an "osish.h" (F<os2ish.h>, F<vmsish.h>, and so on) and including
192 that in F<perl.h>. If you have several machine-specific files (function
193 emulations, function stubs, build utility wrappers) you may create a
194 separate subdirectory (vms, win32) and put the files in there.
195 Remember to update C<MANIFEST> when you add files.
197 If your system supports dynamic loading but none of the existing
198 methods at F<ext/DynaLoader/dl_*.xs> work for you, you must write
199 a new one. Study the existing ones to see what kind of interface
204 There are two kinds of hints: hints for building Perl and hints for
205 extensions. The former live in the C<hints> subdirectory, the latter
206 in C<ext/*/hints> subdirectories.
208 The top level hints are Bourne-shell scripts that set, modify and
209 unset appropriate Configure variables, based on the Configure command
210 line options and possibly existing config.sh and Policy.sh files from
211 previous Configure runs.
213 The extension hints are written in Perl (by the time they are used
214 miniperl has been built) and control the building of their respective
215 extensions. They can be used to for example manipulate compilation
218 =item build and installation Makefiles, scripts, and so forth
220 Sometimes you will also need to tweak the Perl build and installation
221 procedure itself, like for example F<Makefile.SH> and F<installperl>.
222 Tread very carefully, even more than usual. Contain your changes
227 Many of the tests in C<t> subdirectory assume machine-specific things
228 like existence of certain functions, something about filesystem
229 semantics, certain external utilities and their error messages. Use
230 the C<$^O> and the C<Config> module (which contains the results of the
231 Configure run, in effect the C<config.sh> converted to Perl) to either
232 skip (preferably not) or customize (preferable) the tests for your
237 Certain standard modules may need updating if your operating system
238 sports for example a native filesystem naming. You may want to update
239 some or all of the modules File::Basename, File::Spec, File::Path, and
240 File::Copy to become aware of your native filesystem syntax and
243 Remember to have a $VERSION in the modules. You can use the
244 F<Porting/checkVERSION.pl> script for checking this.
248 If your operating system comes from outside UNIX you almost certainly
249 will have differences in the available operating system functionality
250 (missing system calls, different semantics, whatever). Please
251 document these at F<pod/perlport.pod>. If your operating system is
252 the first B<not> to have a system call also update the list of
253 "portability-bewares" at the beginning of F<pod/perlfunc.pod>.
255 A file called F<README.youros> at the top level that explains things
256 like how to install perl at this platform, where to get any possibly
257 required additional software, and for example what test suite errors
258 to expect, is nice too. Such files are in the process of being written
259 in pod format and will eventually be renamed F<INSTALL.youros>.
261 You may also want to write a separate F<.pod> file for your operating
262 system to tell about existing mailing lists, os-specific modules,
263 documentation, whatever. Please name these along the lines of
264 F<perl>I<youros>.pod. [unfinished: where to put this file (the pod/
265 subdirectory, of course: but more importantly, which/what index files
270 =head2 Allow for lots of testing
272 We should never release a main version without testing it as a
275 =head2 Test popular applications and modules.
277 We should never release a main version without testing whether or not
278 it breaks various popular modules and applications. A partial list of
279 such things would include majordomo, metaconfig, apache, Tk, CGI,
280 libnet, and libwww, to name just a few. Of course it's quite possible
281 that some of those things will be just plain broken and need to be fixed,
282 but, in general, we ought to try to avoid breaking widely-installed
285 =head2 Automated generation of derivative files
287 The F<embed.h>, F<keywords.h>, F<opcode.h>, F<regcharclass.h>,
288 F<l1_char_class_tab.h>, and F<perltoc.pod> files
289 are all automatically generated by perl scripts. In general, don't
290 patch these directly; patch the data files instead.
292 Also F<Makefile> is automatically produced from F<Makefile.SH>.
293 In general, look out for all F<*.SH> files.
295 Finally, the sample files F<config.sh> and F<config_H> in the
296 F<Porting/> subdirectory are generated by the script F<Porting/mksample>.
298 =head3 Files generated by metaconfig
300 F<Configure>, F<config_h.SH> and F<Porting/Glossary> are generated by
301 B<metaconfig> (see below for more information on how to use this system)
302 and direct changes to these files should in general not be pushed to blead.
314 changes pre-approved by the metaconfig maintainers
318 Such changes should also be notified to the metaconfig maintainers by
319 creating an issue at <https://github.com/Perl/metaconfig/issues>.
321 Alternatively, do consider if the F<*ish.h> files or the hint files might
322 be a better place for your changes.
324 =head1 Working with metaconfig
326 Information about how to use metaconfig can be found in the F<README>
327 and F<README_U> files in the metaconfig repository containing Perl's
331 git clone https://github.com/Perl/metaconfig.git
332 # or using a registered github.com identity with ssh
333 git clone github.com:Perl/metaconfig.git
335 Since metaconfig is hard to change, running correction scripts after
336 this generation is sometimes needed. Configure gained complexity over
337 time, and the order in which config_h.SH is generated can cause havoc
338 when compiling perl. Therefor, you need to run Porting/config_h.pl
339 after that generation. All that and more is described in the README
340 files that come with the metaunits.
342 =head1 How to Make a Distribution
344 This section has now been expanded and moved into its own file,
345 F<Porting/release_managers_guide.pod>.
347 I've kept some of the subsections here for now, as they don't directly
348 relate to building a release any more, but still contain what might be
349 useful information - DAPM 7/2009.
353 If you are using metaconfig to regenerate Configure, then you should note
354 that metaconfig actually uses MANIFEST.new, so you want to be sure
355 MANIFEST.new is up-to-date too. I haven't found the MANIFEST/MANIFEST.new
356 distinction particularly useful, but that's probably because I still haven't
357 learned how to use the full suite of tools in the dist distribution.
362 This will build a config.sh and config.h. You can skip this if you haven't
363 changed Configure or config_h.SH at all. I use the following command
365 sh Configure -Dprefix=/opt/perl -Doptimize=-O -Dusethreads \
367 -Dcf_email='yourname@yourhost.yourplace.com' \
368 -Dperladmin='yourname@yourhost.yourplace.com' \
369 -Dmydomain='.yourplace.com' \
370 -Dmyhostname='yourhost' \
373 =head2 Update Porting/config.sh and Porting/config_H
376 This section needs revision. We're currently working on easing
377 the task of keeping the vms, win32, and plan9 config.sh info
378 up-to-date. The plan is to use keep up-to-date 'canned' config.sh
379 files in the appropriate subdirectories and then generate 'canned'
380 config.h files for vms, win32, etc. from the generic config.sh file.
381 This is to ease maintenance. When Configure gets updated, the parts
382 sometimes get scrambled around, and the changes in config_H can
383 sometimes be very hard to follow. config.sh, on the other hand, can
384 safely be sorted, so it's easy to track (typically very small) changes
385 to config.sh and then propagate them to a canned 'config.h' by any
386 number of means, including a perl script in win32/ or carrying
387 F<config.sh> and F<config_h.SH> to a Unix system and running sh
388 config_h.SH.) Vms uses F<configure.com> to generate its own F<config.sh>
389 and F<config.h>. If you want to add a new variable to F<config.sh> check
390 with vms folk how to add it to configure.com too.
393 The F<Porting/config.sh> and F<Porting/config_H> files are provided to
394 help those folks who can't run Configure. It is important to keep
395 them up-to-date. If you have changed F<config_h.SH>, those changes must
396 be reflected in config_H as well. (The name config_H was chosen to
397 distinguish the file from config.h even on case-insensitive file systems.)
398 Simply edit the existing config_H file; keep the first few explanatory
399 lines and then copy your new config.h below.
401 It may also be necessary to update win32/config.?c, and
402 F<plan9/config.plan9>, though you should be quite careful in doing so if
403 you are not familiar with those systems. You might want to issue your
404 patch with a promise to quickly issue a follow-up that handles those
407 =head2 make regen_perly
409 If F<perly.y> has been edited, it is necessary to run this target to rebuild
410 F<perly.h>, F<perly.act> and F<perly.tab>. In fact this target just runs the Perl
411 script F<regen_perly.pl>. Note that F<perly.c> is I<not> rebuilt; this is just a
412 plain static file now.
414 This target relies on you having Bison installed on your system. Running
415 the target will tell you if you haven't got the right version, and if so,
416 where to get the right one. Or if you prefer, you could hack
417 F<regen_perly.pl> to work with your version of Bison. The important things
418 are that the regexes can still extract out the right chunks of the Bison
419 output into F<perly.act> and F<perly.tab>, and that the contents of those two
420 files, plus F<perly.h>, are functionally equivalent to those produced by the
421 supported version of Bison.
423 Note that in the old days, you had to do C<make run_byacc> instead.
425 =head2 make regen_all
427 This target takes care of the regen_headers target.
428 (It used to also call the regen_pods target, but that has been eliminated.)
430 =head2 make regen_headers
432 The F<embed.h>, F<keywords.h>, and F<opcode.h> files are all automatically
433 generated by perl scripts. Since the user isn't guaranteed to have a
434 working perl, we can't require the user to generate them. Hence you have
435 to, if you're making a distribution.
437 I used to include rules like the following in the makefile:
439 # The following three header files are generated automatically
440 # The correct versions should be already supplied with the perl kit,
441 # in case you don't have perl or 'sh' available.
442 # The - is to ignore error return codes in case you have the source
443 # installed read-only or you don't have perl yet.
444 keywords.h: keywords.pl
445 @echo "Don't worry if this fails."
449 However, I got B<lots> of mail consisting of people worrying because the
450 command failed. I eventually decided that I would save myself time
451 and effort by manually running C<make regen_headers> myself rather
452 than answering all the questions and complaints about the failing
455 =head2 globvar.sym, and perlio.sym
457 Make sure these files are up-to-date. Read the comments in these
458 files and in F<perl_exp.SH> to see what to do.
460 =head2 Binary compatibility
462 If you do change F<embed.fnc> think carefully about
463 what you are doing. To the extent reasonable, we'd like to maintain
464 source and binary compatibility with older releases of perl. That way,
465 extensions built under one version of perl will continue to work with
466 new versions of perl.
468 Of course, some incompatible changes may well be necessary. I'm just
469 suggesting that we not make any such changes without thinking carefully
470 about them first. If possible, we should provide
471 backwards-compatibility stubs. There's a lot of XS code out there.
472 Let's not force people to keep changing it.
476 F<dist/Devel-PPPort/PPPort.pm> needs to be synchronized to include all
477 new macros added to .h files (normally F<perl.h> and F<XSUB.h>, but others
478 as well). Since chances are that when a new macro is added the
479 committer will forget to update F<PPPort.pm>, it's the best to diff for
480 changes in .h files when making a new release and making sure that
481 F<PPPort.pm> contains them all.
483 The Steering Council can delegate the synchronization responsibility to
484 anybody else, but the release process is the only place where we can make
485 sure that no new macros fell through the cracks.
490 The F<Porting/todo.pod> file contains a roughly-categorized unordered
491 list of aspects of Perl that could use enhancement, features that could
492 be added, areas that could be cleaned up, and so on. During your term
493 as pumpkin-holder, you will probably address some of these issues, and
494 perhaps identify others which, while you decide not to address them this
495 time around, may be tackled in the future. Update the file to reflect
496 the situation as it stands when you hand over the pumpkin.
498 You might like, early in your pumpkin-holding career, to see if you
499 can find champions for particular issues on the to-do list: an issue
500 owned is an issue more likely to be resolved.
502 There are also some more porting-specific L</Todo> items later in this
505 =head2 OS/2-specific updates
507 In the os2 directory is F<diff.configure>, a set of OS/2-specific
508 diffs against B<Configure>. If you make changes to Configure, you may
509 want to consider regenerating this diff file to save trouble for the
512 You can also consider the OS/2 diffs as reminders of portability
513 things that need to be fixed in Configure.
515 =head2 VMS-specific updates
517 The Perl revision number appears as "perl5" in F<configure.com>.
518 It is courteous to update that if necessary.
521 =head2 Making a new patch
523 I find the F<makepatch> utility quite handy for making patches.
524 You can obtain it from any CPAN archive under
525 L<https://www.cpan.org/authors/id/J/JV/JV/>. There are a couple
526 of differences between my version and the standard one. I have mine do
529 # Print a reassuring "End of Patch" note so people won't
530 # wonder if their mailer truncated patches.
531 print "\n\nEnd of Patch.\n";
533 at the end. That's because I used to get questions from people asking
534 if their mail was truncated.
536 It also writes Index: lines which include the new directory prefix
537 (change Index: print, approx line 294 or 310 depending on the version,
538 to read: print PATCH ("Index: $newdir$new\n");). That helps patches
539 work with more POSIX conformant patch programs.
541 Here's how I generate a new patch. I'll use the hypothetical
542 5.004_07 to 5.004_08 patch as an example.
544 # unpack perl5.004_07/
545 gzip -d -c perl5.004_07.tar.gz | tar -xof -
546 # unpack perl5.004_08/
547 gzip -d -c perl5.004_08.tar.gz | tar -xof -
548 makepatch perl5.004_07 perl5.004_08 > perl5.004_08.pat
550 Makepatch will automatically generate appropriate B<rm> commands to remove
551 deleted files. Unfortunately, it will not correctly set permissions
552 for newly created files, so you may have to do so manually. For example,
553 patch 5.003_04 created a new test F<t/op/gv.t> which needs to be executable,
554 so at the top of the patch, I inserted the following lines:
560 Now, of course, my patch is now wrong because makepatch didn't know I
561 was going to do that command, and it patched against /dev/null.
563 So, what I do is sort out all such shell commands that need to be in the
564 patch (including possible mv-ing of files, if needed) and put that in the
565 shell commands at the top of the patch. Next, I delete all the patch parts
566 of perl5.004_08.pat, leaving just the shell commands. Then, I do the
570 sh ../perl5.004_08.pat
572 makepatch perl5.004_07 perl5.004_08 >> perl5.004_08.pat
574 (Note the append to preserve my shell commands.)
575 Now, my patch will line up with what the end users are going to do.
577 =head2 Testing your patch
579 It seems obvious, but be sure to test your patch. That is, verify that
580 it produces exactly the same thing as your full distribution.
583 gzip -d -c perl5.004_07.tar.gz | tar -xf -
585 sh ../perl5.004_08.pat
586 patch -p1 -N < ../perl5.004_08.pat
588 gdiff -r perl5.004_07 perl5.004_08
590 where B<gdiff> is GNU diff. Other diff's may also do recursive checking.
594 Again, it's obvious, but you should test your new version as widely as you
595 can. You can be sure you'll hear about it quickly if your version doesn't
596 work on both ANSI and pre-ANSI compilers, and on common systems such as
597 SunOS 4.1.[34], Solaris, and Linux.
599 If your changes include conditional code, try to test the different
600 branches as thoroughly as you can. For example, if your system
601 supports dynamic loading, you can also test static loading with
605 You can also hand-tweak your config.h to try out different #ifdef
612 =item gcc -ansi -pedantic
614 Configure -Dgccansipedantic [ -Dcc=gcc ] will enable (via the cflags script,
615 not $Config{ccflags}) the gcc strict ANSI C flags -ansi and -pedantic for
616 the compilation of the core files on platforms where it knows it can
617 do so (like Linux, see cflags.SH for the full list), and on some
618 platforms only one (Solaris can do only -pedantic, not -ansi).
619 The flag -DPERL_GCC_PEDANTIC also gets added, since gcc does not add
620 any internal cpp flag to signify that -pedantic is being used, as it
621 does for -ansi (__STRICT_ANSI__).
623 Note that the -ansi and -pedantic are enabled only for version 3 (and
624 later) of gcc, since even gcc version 2.95.4 finds lots of seemingly
625 false "value computed not used" errors from Perl.
627 The -ansi and -pedantic are useful in catching at least the following
628 nonportable practices:
634 gcc-specific extensions
650 The -Dgccansipedantic should be used only when cleaning up the code,
651 not for production builds, since otherwise gcc cannot inline certain
656 =head1 Common Gotchas
660 =item Probably Prefer POSIX
662 It's often the case that you'll need to choose whether to do
663 something the BSD-ish way or the POSIX-ish way. It's usually not
664 a big problem when the two systems use different names for similar
665 functions, such as memcmp() and bcmp(). The perl.h header file
666 handles these by appropriate #defines, selecting the POSIX mem*()
667 functions if available, but falling back on the b*() functions, if
670 More serious is the case where some brilliant person decided to
671 use the same function name but give it a different meaning or
672 calling sequence :-). getpgrp() and setpgrp() come to mind.
673 These are a real problem on systems that aim for conformance to
674 one standard (e.g. POSIX), but still try to support the other way
675 of doing things (e.g. BSD). My general advice (still not really
676 implemented in the source) is to do something like the following.
677 Suppose there are two alternative versions, fooPOSIX() and
681 /* use fooPOSIX(); */
684 /* try to emulate fooPOSIX() with fooBSD();
685 perhaps with the following: */
686 # define fooPOSIX fooBSD
688 # /* Uh, oh. We have to supply our own. */
689 # define fooPOSIX Perl_fooPOSIX
693 =item Think positively
695 If you need to add an #ifdef test, it is usually easier to follow if you
696 think positively, e.g.
698 #ifdef HAS_NEATO_FEATURE
699 /* use neato feature */
701 /* use some fallback mechanism */
704 rather than the more impenetrable
706 #ifndef MISSING_NEATO_FEATURE
707 /* Not missing it, so we must have it, so use it */
709 /* Are missing it, so fall back on something else. */
712 Of course for this toy example, there's not much difference. But when
713 the #ifdef's start spanning a couple of screen fulls, and the #else's
714 are marked something like
716 #else /* !MISSING_NEATO_FEATURE */
718 I find it easy to get lost.
720 =item Providing Missing Functions -- Problem
722 Not all systems have all the neat functions you might want or need, so
723 you might decide to be helpful and provide an emulation. This is
724 sound in theory and very kind of you, but please be careful about what
725 you name the function. Let me use the C<pause()> function as an
728 Perl5.003 has the following in F<perl.h>
731 #define pause() sleep((32767<<16)+32767)
734 Configure sets HAS_PAUSE if the system has the pause() function, so
735 this #define only kicks in if the pause() function is missing.
738 Unfortunately, some systems apparently have a prototype for pause()
739 in F<unistd.h>, but don't actually have the function in the library.
740 (Or maybe they do have it in a library we're not using.)
742 Thus, the compiler sees something like
744 extern int pause(void);
746 #define pause() sleep((32767<<16)+32767)
748 and dies with an error message. (Some compilers don't mind this;
749 others apparently do.)
751 To work around this, 5.003_03 and later have the following in perl.h:
753 /* Some unistd.h's give a prototype for pause() even though
754 HAS_PAUSE ends up undefined. This causes the #define
755 below to be rejected by the compiler. Sigh.
760 # define Pause() sleep((32767<<16)+32767)
765 The curious reader may wonder why I didn't do the following in
771 sleep((32767<<16)+32767);
775 That is, since the function is missing, just provide it.
776 Then things would probably be been alright, it would seem.
778 Well, almost. It could be made to work. The problem arises from the
779 conflicting needs of dynamic loading and namespace protection.
781 For dynamic loading to work on AIX (and VMS) we need to provide a list
782 of symbols to be exported. This is done by the script F<perl_exp.SH>,
783 which reads F<embed.fnc>. Thus, the C<pause>
784 symbol would have to be added to F<embed.fnc> So far, so good.
786 On the other hand, one of the goals of Perl5 is to make it easy to
787 either extend or embed perl and link it with other libraries. This
788 means we have to be careful to keep the visible namespace "clean".
789 That is, we don't want perl's global variables to conflict with
790 those in the other application library. Although this work is still
791 in progress, the way it is currently done is via the F<embed.h> file.
792 This file is built from the F<embed.fnc> file,
793 since those files already list the globally visible symbols. If we
794 had added C<pause> to F<embed.fnc>, then F<embed.h> would contain the
797 #define pause Perl_pause
799 and calls to C<pause> in the perl sources would now point to
800 C<Perl_pause>. Now, when B<ld> is run to build the F<perl> executable,
801 it will go looking for C<perl_pause>, which probably won't exist in any
802 of the standard libraries. Thus the build of perl will fail.
804 Those systems where C<HAS_PAUSE> is not defined would be ok, however,
805 since they would get a C<Perl_pause> function in util.c. The rest of
806 the world would be in trouble.
808 And yes, this scenario has happened. On SCO, the function C<chsize>
809 is available. (I think it's in F<-lx>, the Xenix compatibility
810 library.) Since the perl4 days (and possibly before), Perl has
811 included a C<chsize> function that gets called something akin to
814 I32 chsize(fd, length)
820 #define chsize Perl_chsize
822 to F<embed.h>, the compile started failing on SCO systems.
824 The "fix" is to give the function a different name. The one
825 implemented in 5.003_05 isn't optimal, but here's what was done:
828 # ifdef my_chsize /* Probably #defined to Perl_my_chsize */
829 # undef my_chsize /* in embed.h */
831 # define my_chsize chsize
834 My explanatory comment in patch 5.003_05 said:
836 Undef and then re-define my_chsize from Perl_my_chsize to
837 just plain chsize if this system HAS_CHSIZE. This probably only
838 applies to SCO. This shows the perils of having internal
839 functions with the same name as external library functions :-).
841 Now, we can safely put C<my_chsize> in C<embed.fnc>, export it, and
842 hide it with F<embed.h>.
844 To be consistent with what I did for C<pause>, I probably should have
845 called the new function C<Chsize>, rather than C<my_chsize>.
846 However, the perl sources are quite inconsistent on this (Consider
847 New, Mymalloc, and Myremalloc, to name just a few.)
849 There is a problem with this fix, however, in that C<Perl_chsize>
850 was available as a F<libperl.a> library function in 5.003, but it
851 isn't available any more (as of 5.003_07). This means that we've
852 broken binary compatibility. This is not good.
854 =item Providing missing functions -- some ideas
856 We currently don't have a standard way of handling such missing
857 function names. Right now, I'm effectively thinking aloud about a
858 solution. Some day, I'll try to formally propose a solution.
860 Part of the problem is that we want to have some functions listed as
861 exported but not have their names mangled by embed.h or possibly
862 conflict with names in standard system headers. We actually already
863 have such a list at the end of F<perl_exp.SH> (though that list is
866 # extra globals not included above.
867 cat <<END >> perl.exp
891 This still needs much thought, but I'm inclined to think that one
892 possible solution is to prefix all such functions with C<perl_> in the
893 source and list them along with the other C<perl_*> functions in
896 Thus, for C<chsize>, we'd do something like the following:
900 # define perl_chsize chsize
903 then in some file (e.g. F<util.c> or F<doio.c>) do
906 I32 perl_chsize(fd, length)
907 /* implement the function here . . . */
910 Alternatively, we could just always use C<chsize> everywhere and move
911 C<chsize> from F<embed.fnc> to the end of F<perl_exp.SH>. That would
912 probably be fine as long as our C<chsize> function agreed with all the
913 C<chsize> function prototypes in the various systems we'll be using.
914 As long as the prototypes in actual use don't vary that much, this is
915 probably a good alternative. (As a counter-example, note how Configure
916 and perl have to go through hoops to find and use get Malloc_t and
917 Free_t for C<malloc> and C<free>.)
919 At the moment, this latter option is what I tend to prefer.
921 =item All the world's a VAX
923 Sorry, showing my age:-). Still, all the world is not BSD 4.[34],
924 SVR4, or POSIX. Be aware that SVR3-derived systems are still quite
925 common (do you have any idea how many systems run SCO?) If you don't
926 have a bunch of v7 manuals handy, the metaconfig units (by default
927 installed in F</usr/local/lib/dist/U>) are a good resource to look at
932 =head1 Miscellaneous Topics
936 Why does perl use a metaconfig-generated Configure script instead of an
937 autoconf-generated configure script?
939 Metaconfig and autoconf are two tools with very similar purposes.
940 Metaconfig is actually the older of the two, and was originally written
941 by Larry Wall, while autoconf is probably now used in a wider variety of
942 packages. The autoconf info file discusses the history of autoconf and
943 how it came to be. The curious reader is referred there for further
946 Overall, both tools are quite good, I think, and the choice of which one
947 to use could be argued either way. In March, 1994, when I was just
948 starting to work on Configure support for Perl5, I considered both
949 autoconf and metaconfig, and eventually decided to use metaconfig for the
954 =item Compatibility with Perl4
956 Perl4 used metaconfig, so many of the #ifdef's were already set up for
957 metaconfig. Of course metaconfig had evolved some since Perl4's days,
958 but not so much that it posed any serious problems.
960 =item Metaconfig worked for me
962 My system at the time was Interactive 2.2, an SVR3.2/386 derivative that
963 also had some POSIX support. Metaconfig-generated Configure scripts
964 worked fine for me on that system. On the other hand, autoconf-generated
965 scripts usually didn't. (They did come quite close, though, in some
966 cases.) At the time, I actually fetched a large number of GNU packages
967 and checked. Not a single one configured and compiled correctly
968 out-of-the-box with the system's cc compiler.
970 =item Configure can be interactive
972 With both autoconf and metaconfig, if the script works, everything is
973 fine. However, one of my main problems with autoconf-generated scripts
974 was that if it guessed wrong about something, it could be B<very> hard to
975 go back and fix it. For example, autoconf always insisted on passing the
976 -Xp flag to cc (to turn on POSIX behavior), even when that wasn't what I
977 wanted or needed for that package. There was no way short of editing the
978 configure script to turn this off. You couldn't just edit the resulting
979 Makefile at the end because the -Xp flag influenced a number of other
982 Metaconfig's Configure scripts, on the other hand, can be interactive.
983 Thus if Configure is guessing things incorrectly, you can go back and fix
984 them. This isn't as important now as it was when we were actively
985 developing Configure support for new features such as dynamic loading,
986 but it's still useful occasionally.
990 At the time, autoconf-generated scripts were covered under the GNU Public
991 License, and hence weren't suitable for inclusion with Perl, which has a
992 different licensing policy. (Autoconf's licensing has since changed.)
996 Metaconfig builds up Configure from a collection of discrete pieces
997 called "units". You can override the standard behavior by supplying your
998 own unit. With autoconf, you have to patch the standard files instead.
999 I find the metaconfig "unit" method easier to work with. Others
1000 may find metaconfig's units clumsy to work with.
1004 =head2 Why isn't there a directory to override Perl's library?
1006 Mainly because no one's gotten around to making one. Note that
1007 "making one" involves changing perl.c, Configure, config_h.SH (and
1008 associated files, see above), and I<documenting> it all in the
1011 Apparently, most folks who want to override one of the standard library
1012 files simply do it by overwriting the standard library files.
1016 In the perl.c sources, you'll find an undocumented APPLLIB_EXP
1017 variable, sort of like PRIVLIB_EXP and ARCHLIB_EXP (which are
1018 documented in config_h.SH). Here's what APPLLIB_EXP is for, from
1019 a mail message from Larry:
1021 The main intent of APPLLIB_EXP is for folks who want to send out a
1022 version of Perl embedded in their product. They would set the
1023 symbol to be the name of the library containing the files needed
1024 to run or to support their particular application. This works at
1025 the "override" level to make sure they get their own versions of
1026 any library code that they absolutely must have configuration
1029 As such, I don't see any conflict with a sysadmin using it for a
1030 override-ish sort of thing, when installing a generic Perl. It
1031 should probably have been named something to do with overriding
1032 though. Since it's undocumented we could still change it... :-)
1034 Given that it's already there, you can use it to override distribution modules.
1035 One way to do that is to add
1037 ccflags="$ccflags -DAPPLLIB_EXP=\"/my/override\""
1039 to your config.over file. (You have to be particularly careful to get the
1040 double quotes in. APPLLIB_EXP must be a valid C string. It might
1041 actually be easier to just #define it yourself in perl.c.)
1043 Then perl.c will put /my/override ahead of ARCHLIB and PRIVLIB. Perl will
1044 also search architecture-specific and version-specific subdirectories of
1047 =head2 Shared libperl.so location
1049 Why isn't the shared libperl.so installed in /usr/lib/ along
1050 with "all the other" shared libraries? Instead, it is installed
1051 in $archlib, which is typically something like
1053 /usr/local/lib/perl5/archname/5.00404
1055 and is architecture- and version-specific.
1057 The basic reason why a shared libperl.so gets put in $archlib is so that
1058 you can have more than one version of perl on the system at the same time,
1059 and have each refer to its own libperl.so.
1061 Three examples might help. All of these work now; none would work if you
1062 put libperl.so in /usr/lib.
1068 Suppose you want to have both threaded and non-threaded perl versions
1069 around. Configure will name both perl libraries "libperl.so" (so that
1070 you can link to them with -lperl). The perl binaries tell them apart
1071 by having looking in the appropriate $archlib directories.
1075 Suppose you have perl5.004_04 installed and you want to try to compile
1076 it again, perhaps with different options or after applying a patch.
1077 If you already have libperl.so installed in /usr/lib/, then it may be
1078 either difficult or impossible to get ld.so to find the new libperl.so
1079 that you're trying to build. If, instead, libperl.so is tucked away in
1080 $archlib, then you can always just change $archlib in the current perl
1081 you're trying to build so that ld.so won't find your old libperl.so.
1082 (The INSTALL file suggests you do this when building a debugging perl.)
1086 The shared perl library is not a "well-behaved" shared library with
1087 proper major and minor version numbers, so you can't necessarily
1088 have perl5.004_04 and perl5.004_05 installed simultaneously. Suppose
1089 perl5.004_04 were to install /usr/lib/libperl.so.4.4, and perl5.004_05
1090 were to install /usr/lib/libperl.so.4.5. Now, when you try to run
1091 perl5.004_04, ld.so might try to load libperl.so.4.5, since it has
1092 the right "major version" number. If this works at all, it almost
1093 certainly defeats the reason for keeping perl5.004_04 around. Worse,
1094 with development subversions, you certainly can't guarantee that
1095 libperl.so.4.4 and libperl.so.4.55 will be compatible.
1097 Anyway, all this leads to quite obscure failures that are sure to drive
1098 casual users crazy. Even experienced users will get confused :-). Upon
1099 reflection, I'd say leave libperl.so in $archlib.
1103 =head2 Indentation style
1105 Over the years Perl has become a mishmash of
1106 various indentation styles, but the original "Larry style" can
1107 probably be restored with (GNU) indent somewhat like this:
1109 indent -kr -nce -psl -sc
1111 A more ambitious solution would also specify a list of Perl specific
1112 types with -TSV -TAV -THV .. -TMAGIC -TPerlIO ... but that list would
1113 be quite ungainly. Also note that GNU indent also doesn't do aligning
1114 of consecutive assignments, which would truly wreck the layout in
1115 places like sv.c:Perl_sv_upgrade() or sv.c:Perl_clone_using().
1116 Similarly nicely aligned &&s, ||s and ==s would not be respected.
1118 =head1 Upload Your Work to CPAN
1120 You can upload your work to CPAN if you have a CPAN id. Check out
1121 L<http://www.cpan.org/modules/04pause.html> for information on
1122 _PAUSE_, the Perl Author's Upload Server.
1124 I typically upload both the patch file, e.g. F<perl5.004_08.pat.gz>
1125 and the full tar file, e.g. F<perl5.004_08.tar.gz>.
1127 If you want your patch to appear in the F<src/5.0/unsupported>
1128 directory on CPAN, send e-mail to the CPAN master librarian. (Check
1129 out L<http://www.cpan.org/CPAN.html> ).
1131 =head1 Help Save the World
1133 You should definitely announce your patch on the perl5-porters list.
1137 Here, in no particular order, are some Configure and build-related
1138 items that merit consideration. This list isn't exhaustive, it's just
1139 what I came up with off the top of my head.
1141 =head2 Adding missing library functions to Perl
1143 The perl Configure script automatically determines which headers and
1144 functions you have available on your system and arranges for them to be
1145 included in the compilation and linking process. Occasionally, when porting
1146 perl to an operating system for the first time, you may find that the
1147 operating system is missing a key function. While perl may still build
1148 without this function, no perl program will be able to reference the missing
1149 function. You may be able to write the missing function yourself, or you
1150 may be able to find the missing function in the distribution files for
1151 another software package. In this case, you need to instruct the perl
1152 configure-and-build process to use your function. Perform these steps.
1158 Code and test the function you wish to add. Test it carefully; you will
1159 have a much easier time debugging your code independently than when it is a
1164 Here is an implementation of the POSIX truncate function for an operating
1165 system (VOS) that does not supply one, but which does supply the ftruncate()
1168 /* Beginning of modification history */
1169 /* Written 02-01-02 by Nick Ing-Simmons (nick@ing-simmons.net) */
1170 /* End of modification history */
1172 /* VOS doesn't supply a truncate function, so we build one up
1173 from the available POSIX functions. */
1176 #include <sys/types.h>
1180 truncate(const char *path, off_t len)
1182 int fd = open(path,O_WRONLY);
1185 code = ftruncate(fd,len);
1191 Place this file into a subdirectory that has the same name as the operating
1192 system. This file is named perl/vos/vos.c
1196 If your operating system has a hints file (in perl/hints/XXX.sh for an
1197 operating system named XXX), then start with it. If your operating system
1198 has no hints file, then create one. You can use a hints file for a similar
1199 operating system, if one exists, as a template.
1203 Add lines like the following to your hints file. The first line
1204 (d_truncate="define") instructs Configure that the truncate() function
1205 exists. The second line (archobjs="vos.o") instructs the makefiles that the
1206 perl executable depends on the existence of a file named "vos.o". (Make
1207 will automatically look for "vos.c" and compile it with the same options as
1208 the perl source code). The final line ("test -h...") adds a symbolic link
1209 to the top-level directory so that make can find vos.c. Of course, you
1210 should use your own operating system name for the source file of extensions,
1213 # VOS does not have truncate() but we supply one in vos.c
1217 # Help gmake find vos.c
1218 test -h vos.c || ln -s vos/vos.c vos.c
1220 The hints file is a series of shell commands that are run in the top-level
1221 directory (the "perl" directory). Thus, these commands are simply executed
1222 by Configure at an appropriate place during its execution.
1226 At this point, you can run the Configure script and rebuild perl. Carefully
1227 test the newly-built perl to ensure that normal paths, and error paths,
1228 behave as you expect.
1232 =head2 Good ideas waiting for round tuits
1236 =item Configure -Dsrc=/blah/blah
1238 We should be able to emulate B<configure --srcdir>. Tom Tromey
1239 tromey@creche.cygnus.com has submitted some patches to
1240 the dist-users mailing list along these lines. They have been folded
1241 back into the main distribution, but various parts of the perl
1242 Configure/build/install process still assume src='.'.
1244 =item Hint file fixes
1246 Various hint files work around Configure problems. We ought to fix
1247 Configure so that most of them aren't needed.
1249 =item Hint file information
1251 Some of the hint file information (particularly dynamic loading stuff)
1252 ought to be fed back into the main metaconfig distribution.
1256 =head2 Probably good ideas waiting for round tuits
1260 =item GNU configure --options
1262 I've received sensible suggestions for --exec_prefix and other
1263 GNU configure --options. It's not always obvious exactly what is
1264 intended, but this merits investigation.
1266 =item Try gcc if cc fails
1268 Currently, we just give up.
1270 =item bypassing safe*alloc wrappers
1272 On some systems, it may be safe to call the system malloc directly
1273 without going through the util.c safe* layers. (Such systems would
1274 accept free(0), for example.) This might be a time-saver for systems
1275 that already have a good malloc. (Recent Linux libc's apparently have
1276 a nice malloc that is well-tuned for the system.)
1280 =head2 Vague possibilities
1284 =item gconvert replacement
1286 Maybe include a replacement function that doesn't lose data in rare
1287 cases of coercion between string and numerical values.
1289 =item Improve makedepend
1291 The current makedepend process is clunky and annoyingly slow, but it
1292 works for most folks. Alas, it assumes that there is a filename
1293 $firstmakefile that the B<make> command will try to use before it uses
1294 F<Makefile>. Such may not be the case for all B<make> commands,
1295 particularly those on non-Unix systems.
1297 Probably some variant of the BSD F<.depend> file will be useful.
1298 We ought to check how other packages do this, if they do it at all.
1299 We could probably pre-generate the dependencies (with the exception of
1300 malloc.o, which could probably be determined at F<Makefile.SH>
1303 =item GNU Makefile standard targets
1305 GNU software generally has standardized Makefile targets. Unless we
1306 have good reason to do otherwise, I see no reason not to support them.
1310 Somehow, straighten out, document, and implement lockf(), flock(),
1311 and/or fcntl() file locking. It's a mess. See $d_fcntl_can_lock
1312 in recent config.sh files though.
1316 =head2 Copyright Issues
1318 The following is based on the consensus of a couple of IPR lawyers,
1319 but it is of course not a legally binding statement, just a common
1326 Tacking on copyright statements is unnecessary to begin with because
1327 of the Berne convention. But assuming you want to go ahead...
1331 The right form of a copyright statement is
1333 Copyright (C) Year, Year, ... by Someone
1335 The (C) is not required everywhere but it doesn't hurt and in certain
1336 jurisdictions it is required, so let's leave it in. (Yes, it's true
1337 that in some jurisdictions the "(C)" is not legally binding, one should
1338 use the true ringed-C. But we don't have that character available for
1339 Perl's source code.)
1341 The years must be listed out separately. Year-Year is not correct.
1342 Only the years when the piece has changed 'significantly' may be added.
1346 One cannot give away one's copyright trivially. One can give one's
1347 copyright away by using public domain, but even that requires a little
1348 bit more than just saying 'this is in public domain'. (What it
1349 exactly requires depends on your jurisdiction.) But barring public
1350 domain, one cannot "transfer" one's copyright to another person or
1351 entity. In the context of software, it means that contributors cannot
1352 give away their copyright or "transfer" it to the "owner" of the software.
1354 Also remember that in many cases if you are employed by someone,
1355 your work may be copyrighted to your employer, even when you are
1356 contributing on your own time (this all depends on too many things
1357 to list here). But the bottom line is that you definitely can't give
1358 away a copyright you may not even have.
1360 What is possible, however, is that the software can simply state
1362 Copyright (C) Year, Year, ... by Someone and others
1364 and then list the "others" somewhere in the distribution.
1365 And this is exactly what Perl does. (The "somewhere" is
1366 AUTHORS and the Changes* files.)
1370 Split files, merged files, and generated files are problematic.
1371 The rule of thumb: in split files, copy the copyright years of
1372 the original file to all the new files; in merged files make
1373 an union of the copyright years of all the old files; in generated
1374 files propagate the copyright years of the generating file(s).
1378 The files of Perl source code distribution do carry a lot of
1379 copyrights, by various people. (There are many copyrights embedded in
1380 perl.c, for example.) The most straightforward thing for perl releasers to
1381 do is to simply update Larry's copyrights at the beginning of the
1382 *.[hcy], *.pl, and README files, and leave all other
1383 copyrights alone. Doing more than that requires quite a bit of tracking.
1389 Original author: Andy Dougherty doughera@lafayette.edu .
1390 Additions by Chip Salzenberg chip@perl.com, Tim Bunce and the perl5
1393 All opinions expressed herein are those of the authorZ<>(s).
1395 =head1 LAST MODIFIED
1397 2017-10-13 Dominic Hargreaves