This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Porting/todo.pod: Add note to ‘repack the optree’
[perl5.git] / Porting / todo.pod
CommitLineData
7711098a
GS
1=head1 NAME
2
c3143508 3todo - Perl TO-DO list
7711098a
GS
4
5=head1 DESCRIPTION
e50bb9a1 6
049aabcb 7This is a list of wishes for Perl. The most up to date version of this file
c3143508 8is at L<http://perl5.git.perl.org/perl.git/blob_plain/HEAD:/Porting/todo.pod>
049aabcb
NC
9
10The tasks we think are smaller or easier are listed first. Anyone is welcome
11to work on any of these, but it's a good idea to first contact
12I<perl5-porters@perl.org> to avoid duplication of effort, and to learn from
13any previous attempts. By all means contact a pumpking privately first if you
14prefer.
e50bb9a1 15
0bdfc961
NC
16Whilst patches to make the list shorter are most welcome, ideas to add to
17the list are also encouraged. Check the perl5-porters archives for past
b4af8972
RB
18ideas, and any discussion about them. One set of archives may be found at
19L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/>
938c8732 20
617eabfa
NC
21What can we offer you in return? Fame, fortune, and everlasting glory? Maybe
22not, but if your patch is incorporated, then we'll add your name to the
23F<AUTHORS> file, which ships in the official distribution. How many other
24programming languages offer you 1 line of immortality?
938c8732 25
0bdfc961 26=head1 Tasks that only need Perl knowledge
e50bb9a1 27
aa384da9
KW
28=head2 Classify bug tickets by type
29
1e5abc58 30Known bugs in Perl are tracked by L<https://rt.perl.org/> (which also
aa384da9 31includes Perl 6). A summary can be found at
1e5abc58 32L<https://rt.perl.org/NoAuth/perl5/Overview.html>.
aa384da9
KW
33It shows bugs classified by "type". However, the type of many of the
34bugs is "unknown". This greatly lowers the chances of them getting
35fixed, as the number of open bugs is overwhelming -- too many to wade
36through for someone to try to find the bugs in the parts of
37Perl that s/he knows well enough to try to fix. This task involves
38going through these bugs and classifying them into one or more types.
39
40=head2 Ongoing: investigate new bug reports
41
42When a bug report is filed, it would be very helpful to have someone do
43a quick investigation to see if it is a real problem, and to reply to
44the poster about it, asking for example code that reproduces the
45problem. Such code should be added to the test suite as TODO tests, and
46the ticket should be classified by type. To get started on this task,
47look at the tickets that are marked as "New Issues" in
1e5abc58 48L<https://rt.perl.org/NoAuth/perl5/Overview.html>.
aa384da9 49
de2b17d8
NC
50=head2 Migrate t/ from custom TAP generation
51
52Many tests below F<t/> still generate TAP by "hand", rather than using library
96090e4f 53functions. As explained in L<perlhack/TESTING>, tests in F<t/> are
de2b17d8
NC
54written in a particular way to test that more complex constructions actually
55work before using them routinely. Hence they don't use C<Test::More>, but
56instead there is an intentionally simpler library, F<t/test.pl>. However,
57quite a few tests in F<t/> have not been refactored to use it. Refactoring
58any of these tests, one at a time, is a useful thing TODO.
59
f3227d3b
JK
60The subdirectories F<base>, F<cmd>, F<comp> and F<opbasic>, that contain the
61most basic tests, should be excluded from this task.
0d8e5a42 62
0be987a2
NC
63=head2 Automate perldelta generation
64
65The perldelta file accompanying each release summaries the major changes.
66It's mostly manually generated currently, but some of that could be
67automated with a bit of perl, specifically the generation of
68
69=over
70
71=item Modules and Pragmata
72
73=item New Documentation
74
75=item New Tests
76
77=back
78
79See F<Porting/how_to_write_a_perldelta.pod> for details.
80
0bdfc961 81=head2 Make Schwern poorer
e50bb9a1 82
613bd4f7 83We should have tests for everything. When all the core's modules are tested,
0bdfc961
NC
84Schwern has promised to donate to $500 to TPF. We may need volunteers to
85hold him upside down and shake vigorously in order to actually extract the
86cash.
3958b146 87
831efc5a
JK
88=head2 Write descriptions for all tests
89
90Many individual tests in the test suite lack descriptions (or names, or labels
91-- call them what you will). Many files completely lack descriptions, meaning
92that the only output you get is the test numbers. If all tests had
93descriptions, understanding what the tests are testing and why they sometimes
94fail would both get a whole lot easier.
95
0bdfc961 96=head2 Improve the coverage of the core tests
e50bb9a1 97
e1020413 98Use Devel::Cover to ascertain the core modules' test coverage, then add
02f21748 99tests that are currently missing.
30222c0f 100
0bdfc961 101=head2 test B
e50bb9a1 102
0bdfc961 103A full test suite for the B module would be nice.
e50bb9a1 104
0bdfc961 105=head2 A decent benchmark
e50bb9a1 106
617eabfa 107C<perlbench> seems impervious to any recent changes made to the perl core. It
0bdfc961
NC
108would be useful to have a reasonable general benchmarking suite that roughly
109represented what current perl programs do, and measurably reported whether
110tweaks to the core improve, degrade or don't really affect performance, to
111guide people attempting to optimise the guts of perl. Gisle would welcome
4e1c9055
NC
112new tests for perlbench. Steffen Schwingon would welcome help with
113L<Benchmark::Perl::Formance>
6168cf99 114
0bdfc961 115=head2 fix tainting bugs
6168cf99 116
82046965
FC
117Fix the bugs revealed by running the test suite with the C<-t> switch.
118Setting the TEST_ARGS environment variable to C<-taintwarn> will accomplish
119this.
e50bb9a1 120
0bdfc961 121=head2 Dual life everything
e50bb9a1 122
0bdfc961
NC
123As part of the "dists" plan, anything that doesn't belong in the smallest perl
124distribution needs to be dual lifed. Anything else can be too. Figure out what
125changes would be needed to package that module and its tests up for CPAN, and
126do so. Test it with older perl releases, and fix the problems you find.
e50bb9a1 127
a393eb28
RGS
128To make a minimal perl distribution, it's useful to look at
129F<t/lib/commonsense.t>.
130
0bdfc961 131=head2 POSIX memory footprint
e50bb9a1 132
0bdfc961
NC
133Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
134various times worked to cut it down. There is probably still fat to cut out -
135for example POSIX passes Exporter some very memory hungry data structures.
e50bb9a1 136
8c422da5
NC
137=head2 makedef.pl and conditional compilation
138
139The script F<makedef.pl> that generates the list of exported symbols on
140platforms which need this. Functions are declared in F<embed.fnc>, variables
141in F<intrpvar.h>. Quite a few of the functions and variables are conditionally
142declared there, using C<#ifdef>. However, F<makedef.pl> doesn't understand the
143C macros, so the rules about which symbols are present when is duplicated in
144the Perl code. Writing things twice is bad, m'kay. It would be good to teach
145F<.pl> to understand the conditional compilation, and hence remove the
146duplication, and the mistakes it has caused.
e50bb9a1 147
801de10e
NC
148=head2 use strict; and AutoLoad
149
150Currently if you write
151
152 package Whack;
153 use AutoLoader 'AUTOLOAD';
154 use strict;
155 1;
156 __END__
157 sub bloop {
158 print join (' ', No, strict, here), "!\n";
159 }
160
161then C<use strict;> isn't in force within the autoloaded subroutines. It would
162be more consistent (and less surprising) to arrange for all lexical pragmas
163in force at the __END__ block to be in force within each autoloaded subroutine.
164
773b3597
RGS
165There's a similar problem with SelfLoader.
166
91d0cbf6
NC
167=head2 profile installman
168
169The F<installman> script is slow. All it is doing text processing, which we're
170told is something Perl is good at. So it would be nice to know what it is doing
171that is taking so much CPU, and where possible address it.
172
c69ca1d4 173=head2 enable lexical enabling/disabling of individual warnings
a9ed9b74
JV
174
175Currently, warnings can only be enabled or disabled by category. There
176are times when it would be useful to quash a single warning, not a
177whole category.
91d0cbf6 178
85234543
KW
179=head2 document diagnostics
180
181Many diagnostic messages are not currently documented. The list is at the end
182of t/porting/diag.t.
183
aa384da9
KW
184=head2 Write TODO tests for open bugs
185
186Sometimes bugs get fixed as a side effect of something else, and
187the bug remains open because no one realizes that it has been fixed.
188Ideally, every open bug should have a TODO test in the core test suite.
189
0bdfc961 190=head1 Tasks that need a little sysadmin-type knowledge
e50bb9a1 191
0bdfc961
NC
192Or if you prefer, tasks that you would learn from, and broaden your skills
193base...
e50bb9a1 194
cd793d32 195=head2 make HTML install work
e50bb9a1 196
78b489b0 197There is an C<install.html> target in the Makefile. It's marked as
adebf063
NC
198"experimental". It would be good to get this tested, make it work reliably, and
199remove the "experimental" tag. This would include
200
201=over 4
202
203=item 1
204
205Checking that cross linking between various parts of the documentation works.
206In particular that links work between the modules (files with POD in F<lib/>)
207and the core documentation (files in F<pod/>)
208
209=item 2
210
78b489b0
NC
211Improving the code that split C<perlfunc> into chunks, preferably with
212general case code added to L<Pod::Functions> that could be used elsewhere.
213
617eabfa
NC
214Challenges here are correctly identifying the groups of functions that go
215together, and making the right named external cross-links point to the right
78b489b0
NC
216page. Currently this works reasonably well in the general case, and correctly
217parses two or more C<=items> giving the different parameter lists for the
218same function, such used by C<substr>. However it fails completely where
219I<different> functions are listed as a sequence of C<=items> but share the
220same description. All the functions from C<getpwnam> to C<endprotoent> have
221individual stub pages, with only the page for C<endservent> holding the
222description common to all. Likewise C<q>, C<qq> and C<qw> have stub pages,
223instead of sharing the body of C<qx>.
224
225Note also the current code isn't ideal with the two forms of C<select>, mushing
226them both into one F<select.html> with the two descriptions run together.
227Fixing this may well be a special case.
adebf063
NC
228
229=back
3a89a73c 230
0bdfc961
NC
231=head2 compressed man pages
232
233Be able to install them. This would probably need a configure test to see how
234the system does compressed man pages (same directory/different directory?
235same filename/different filename), as well as tweaking the F<installman> script
236to compress as necessary.
237
30222c0f
NC
238=head2 Add a code coverage target to the Makefile
239
240Make it easy for anyone to run Devel::Cover on the core's tests. The steps
241to do this manually are roughly
242
243=over 4
244
245=item *
246
247do a normal C<Configure>, but include Devel::Cover as a module to install
f11a3063 248(see L<INSTALL> for how to do this)
30222c0f
NC
249
250=item *
251
252 make perl
253
254=item *
255
f185f654 256 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
30222c0f
NC
257
258=item *
259
260Process the resulting Devel::Cover database
261
262=back
263
264This just give you the coverage of the F<.pm>s. To also get the C level
265coverage you need to
266
267=over 4
268
269=item *
270
271Additionally tell C<Configure> to use the appropriate C compiler flags for
272C<gcov>
273
274=item *
275
276 make perl.gcov
277
278(instead of C<make perl>)
279
280=item *
281
282After running the tests run C<gcov> to generate all the F<.gcov> files.
283(Including down in the subdirectories of F<ext/>
284
285=item *
286
287(From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
288to get their stats into the cover_db directory.
289
290=item *
291
292Then process the Devel::Cover database
293
294=back
295
296It would be good to add a single switch to C<Configure> to specify that you
297wanted to perform perl level coverage, and another to specify C level
298coverage, and have C<Configure> and the F<Makefile> do all the right things
299automatically.
300
02f21748 301=head2 Make Config.pm cope with differences between built and installed perl
0bdfc961
NC
302
303Quite often vendors ship a perl binary compiled with their (pay-for)
304compilers. People install a free compiler, such as gcc. To work out how to
305build extensions, Perl interrogates C<%Config>, so in this situation
306C<%Config> describes compilers that aren't there, and extension building
307fails. This forces people into choosing between re-compiling perl themselves
308using the compiler they have, or only using modules that the vendor ships.
309
310It would be good to find a way teach C<Config.pm> about the installation setup,
311possibly involving probing at install time or later, so that the C<%Config> in
312a binary distribution better describes the installed machine, when the
313installed machine differs from the build machine in some significant way.
314
728f4ecd
NC
315=head2 linker specification files
316
317Some platforms mandate that you provide a list of a shared library's external
318symbols to the linker, so the core already has the infrastructure in place to
4e1c9055
NC
319do this for generating shared perl libraries. Florian Ragwitz has been working
320to offer this for the GNU toolchain, to allow Unix users to test that the
728f4ecd 321export list is correct, and to build a perl that does not pollute the global
32d539f5 322namespace with private symbols, and will fail in the same way as msvc or mingw
4e1c9055 323builds or when using PERL_DL_NONLAZY=1. See the branch smoke-me/rafl/ld_export
728f4ecd 324
a229ae3b
RGS
325=head2 Cross-compile support
326
4e1c9055
NC
327We get requests for "how to cross compile Perl". The vast majority of these
328seem to be for a couple of scenarios:
329
330=over 4
331
332=item *
333
334Platforms that could build natively using F<./Configure> (I<e.g.> Linux or
335NetBSD on MIPS or ARM) but people want to use a beefier machine (and on the
336same OS) to build more easily.
337
338=item *
339
340Platforms that can't build natively, but no (significant) porting changes
341are needed to our current source code. Prime example of this is Android.
342
343=back
344
345There are several scripts and tools for cross-compiling perl for other
346platforms. However, these are somewhat inconsistent and scattered across the
347codebase, none are documented well, none are clearly flexible enough to
c5fb089a 348be confident that they can support any TARGET/HOST platform pair other than
4e1c9055
NC
349that which they were developed on, and it's not clear how bitrotted they are.
350
351For example, C<Configure> understands C<-Dusecrosscompile> option. This option
a229ae3b 352arranges for building C<miniperl> for TARGET machine, so this C<miniperl> is
4e1c9055
NC
353assumed then to be copied to TARGET machine and used as a replacement of
354full C<perl> executable. This code is almost 10 years old. Meanwhile, the
355F<Cross/> directory contains two different approaches for cross compiling to
356ARM Linux targets, relying on hand curated F<config.sh> files, but that code
357is getting on for 5 years old, and requires insider knowledge of perl's
358build system to draft a F<config.sh> for a new platform.
359
c5fb089a 360Jess Robinson has submitted a grant to TPF to work on cleaning this up.
0bdfc961 361
98fca0e8
NC
362=head2 Split "linker" from "compiler"
363
364Right now, Configure probes for two commands, and sets two variables:
365
366=over 4
367
b91dd380 368=item * C<cc> (in F<cc.U>)
98fca0e8
NC
369
370This variable holds the name of a command to execute a C compiler which
371can resolve multiple global references that happen to have the same
372name. Usual values are F<cc> and F<gcc>.
373Fervent ANSI compilers may be called F<c89>. AIX has F<xlc>.
374
b91dd380 375=item * C<ld> (in F<dlsrc.U>)
98fca0e8
NC
376
377This variable indicates the program to be used to link
378libraries for dynamic loading. On some systems, it is F<ld>.
379On ELF systems, it should be C<$cc>. Mostly, we'll try to respect
380the hint file setting.
381
382=back
383
8d159ec1
NC
384There is an implicit historical assumption from around Perl5.000alpha
385something, that C<$cc> is also the correct command for linking object files
386together to make an executable. This may be true on Unix, but it's not true
387on other platforms, and there are a maze of work arounds in other places (such
388as F<Makefile.SH>) to cope with this.
98fca0e8
NC
389
390Ideally, we should create a new variable to hold the name of the executable
391linker program, probe for it in F<Configure>, and centralise all the special
392case logic there or in hints files.
393
394A small bikeshed issue remains - what to call it, given that C<$ld> is already
8d159ec1
NC
395taken (arguably for the wrong thing now, but on SunOS 4.1 it is the command
396for creating dynamically-loadable modules) and C<$link> could be confused with
397the Unix command line executable of the same name, which does something
398completely different. Andy Dougherty makes the counter argument "In parrot, I
399tried to call the command used to link object files and libraries into an
400executable F<link>, since that's what my vaguely-remembered DOS and VMS
401experience suggested. I don't think any real confusion has ensued, so it's
402probably a reasonable name for perl5 to use."
98fca0e8
NC
403
404"Alas, I've always worried that introducing it would make things worse,
405since now the module building utilities would have to look for
406C<$Config{link}> and institute a fall-back plan if it weren't found."
8d159ec1
NC
407Although I can see that as confusing, given that C<$Config{d_link}> is true
408when (hard) links are available.
98fca0e8 409
75585ce3
SP
410=head2 Configure Windows using PowerShell
411
412Currently, Windows uses hard-coded config files based to build the
413config.h for compiling Perl. Makefiles are also hard-coded and need to be
414hand edited prior to building Perl. While this makes it easy to create a perl.exe
415that works across multiple Windows versions, being able to accurately
416configure a perl.exe for a specific Windows versions and VS C++ would be
417a nice enhancement. With PowerShell available on Windows XP and up, this
418may now be possible. Step 1 might be to investigate whether this is possible
419and use this to clean up our current makefile situation. Step 2 would be to
420see if there would be a way to use our existing metaconfig units to configure a
421Windows Perl or whether we go in a separate direction and make it so. Of
422course, we all know what step 3 is.
423
0bdfc961
NC
424=head1 Tasks that need a little C knowledge
425
426These tasks would need a little C knowledge, but don't need any specific
427background or experience with XS, or how the Perl interpreter works
428
3d826b29
NC
429=head2 Weed out needless PERL_UNUSED_ARG
430
431The C code uses the macro C<PERL_UNUSED_ARG> to stop compilers warning about
432unused arguments. Often the arguments can't be removed, as there is an
433external constraint that determines the prototype of the function, so this
434approach is valid. However, there are some cases where C<PERL_UNUSED_ARG>
435could be removed. Specifically
436
437=over 4
438
439=item *
440
441The prototypes of (nearly all) static functions can be changed
442
443=item *
444
445Unused arguments generated by short cut macros are wasteful - the short cut
446macro used can be changed.
447
448=back
449
bcbaa2d5
RGS
450=head2 -Duse32bit*
451
452Natively 64-bit systems need neither -Duse64bitint nor -Duse64bitall.
453On these systems, it might be the default compilation mode, and there
454is currently no guarantee that passing no use64bitall option to the
455Configure process will build a 32bit perl. Implementing -Duse32bit*
1e2ea6fa 456options would be nice for perl 5.25.2.
bcbaa2d5 457
fee0a0f7 458=head2 Profile Perl - am I hot or not?
62403a3c 459
fee0a0f7
NC
460The Perl source code is stable enough that it makes sense to profile it,
461identify and optimise the hotspots. It would be good to measure the
462performance of the Perl interpreter using free tools such as cachegrind,
463gprof, and dtrace, and work to reduce the bottlenecks they reveal.
464
465As part of this, the idea of F<pp_hot.c> is that it contains the I<hot> ops,
466the ops that are most commonly used. The idea is that by grouping them, their
467object code will be adjacent in the executable, so they have a greater chance
468of already being in the CPU cache (or swapped in) due to being near another op
469already in use.
62403a3c
NC
470
471Except that it's not clear if these really are the most commonly used ops. So
fee0a0f7
NC
472as part of exercising your skills with coverage and profiling tools you might
473want to determine what ops I<really> are the most commonly used. And in turn
474suggest evictions and promotions to achieve a better F<pp_hot.c>.
62403a3c 475
91d0cbf6
NC
476One piece of Perl code that might make a good testbed is F<installman>.
477
a229ae3b 478=head2 Improve win32/wince.c
0bdfc961 479
a229ae3b 480Currently, numerous functions look virtually, if not completely,
c23989d1 481identical in both F<win32/wince.c> and F<win32/win32.c> files, which can't
6d71adcd
NC
482be good.
483
c5b31784
SH
484=head2 Use secure CRT functions when building with VC8 on Win32
485
486Visual C++ 2005 (VC++ 8.x) deprecated a number of CRT functions on the basis
487that they were "unsafe" and introduced differently named secure versions of
488them as replacements, e.g. instead of writing
489
490 FILE* f = fopen(__FILE__, "r");
491
492one should now write
493
494 FILE* f;
495 errno_t err = fopen_s(&f, __FILE__, "r");
496
497Currently, the warnings about these deprecations have been disabled by adding
498-D_CRT_SECURE_NO_DEPRECATE to the CFLAGS. It would be nice to remove that
499warning suppressant and actually make use of the new secure CRT functions.
500
501There is also a similar issue with POSIX CRT function names like fileno having
502been deprecated in favour of ISO C++ conformant names like _fileno. These
26a6faa8 503warnings are also currently suppressed by adding -D_CRT_NONSTDC_NO_DEPRECATE. It
c5b31784
SH
504might be nice to do as Microsoft suggest here too, although, unlike the secure
505functions issue, there is presumably little or no benefit in this case.
506
038ae9a4
SH
507=head2 Fix POSIX::access() and chdir() on Win32
508
509These functions currently take no account of DACLs and therefore do not behave
510correctly in situations where access is restricted by DACLs (as opposed to the
511read-only attribute).
512
513Furthermore, POSIX::access() behaves differently for directories having the
514read-only attribute set depending on what CRT library is being used. For
515example, the _access() function in the VC6 and VC7 CRTs (wrongly) claim that
516such directories are not writable, whereas in fact all directories are writable
517unless access is denied by DACLs. (In the case of directories, the read-only
518attribute actually only means that the directory cannot be deleted.) This CRT
519bug is fixed in the VC8 and VC9 CRTs (but, of course, the directory may still
520not actually be writable if access is indeed denied by DACLs).
521
522For the chdir() issue, see ActiveState bug #74552:
b4af8972 523L<http://bugs.activestate.com/show_bug.cgi?id=74552>
038ae9a4
SH
524
525Therefore, DACLs should be checked both for consistency across CRTs and for
526the correct answer.
527
528(Note that perl's -w operator should not be modified to check DACLs. It has
529been written so that it reflects the state of the read-only attribute, even
530for directories (whatever CRT is being used), for symmetry with chmod().)
531
8964cfe0
NC
532=head2 Arenas for GPs? For MAGIC?
533
534C<struct gp> and C<struct magic> are both currently allocated by C<malloc>.
535It might be a speed or memory saving to change to using arenas. Or it might
536not. It would need some suitable benchmarking first. In particular, C<GP>s
537can probably be changed with minimal compatibility impact (probably nothing
538outside of the core, or even outside of F<gv.c> allocates them), but they
539probably aren't allocated/deallocated often enough for a speed saving. Whereas
540C<MAGIC> is allocated/deallocated more often, but in turn, is also something
541more externally visible, so changing the rules here may bite external code.
542
3880c8ec
NC
543=head2 Shared arenas
544
545Several SV body structs are now the same size, notably PVMG and PVGV, PVAV and
546PVHV, and PVCV and PVFM. It should be possible to allocate and return same
547sized bodies from the same actual arena, rather than maintaining one arena for
548each. This could save 4-6K per thread, of memory no longer tied up in the
549not-yet-allocated part of an arena.
550
8964cfe0 551
6d71adcd
NC
552=head1 Tasks that need a knowledge of XS
553
554These tasks would need C knowledge, and roughly the level of knowledge of
555the perl API that comes from writing modules that use XS to interface to
556C.
557
e851c105
DG
558=head2 Write an XS cookbook
559
560Create pod/perlxscookbook.pod with short, task-focused 'recipes' in XS that
561demonstrate common tasks and good practices. (Some of these might be
562extracted from perlguts.) The target audience should be XS novices, who need
563more examples than perlguts but something less overwhelming than perlapi.
564Recipes should provide "one pretty good way to do it" instead of TIMTOWTDI.
565
5b7d14ff
DG
566Rather than focusing on interfacing Perl to C libraries, such a cookbook
567should probably focus on how to optimize Perl routines by re-writing them
568in XS. This will likely be more motivating to those who mostly work in
569Perl but are looking to take the next step into XS.
570
571Deconstructing and explaining some simpler XS modules could be one way to
572bootstrap a cookbook. (List::Util? Class::XSAccessor? Tree::Ternary_XS?)
573Another option could be deconstructing the implementation of some simpler
574functions in op.c.
575
0b162fb0 576=head2 Document how XSUBs can use C<cv_set_call_checker> to inline themselves as OPs
05fb4e20
NC
577
578For a simple XSUB, often the subroutine dispatch takes more time than the
0b162fb0
NC
579XSUB itself. v5.14.0 now allows XSUBs to register a function which will be
580called when the parser is finished building an C<entersub> op which calls
581them.
582
583Registration is done with C<Perl_cv_set_call_checker>, is documented at the
584API level in L<perlapi>, and L<perl5140delta/Custom per-subroutine check hooks>
585notes that it can be used to inline a subroutine, by replacing it with a
586custom op. However there is no further detail of the code needed to do this.
587It would be useful to add one or more annotated examples of how to create
588XSUBs that inline.
589
590This should provide a measurable speed up to simple XSUBs inside
05fb4e20
NC
591tight loops. Initially one would have to write the OP alternative
592implementation by hand, but it's likely that this should be reasonably
593straightforward for the type of XSUB that would benefit the most. Longer
594term, once the run-time implementation is proven, it should be possible to
595progressively update ExtUtils::ParseXS to generate OP implementations for
596some XSUBs.
597
1ab33739
FC
598=head2 Document how XS modules can install lexical subs
599
600There is an example in XS::APItest (look for C<lexical_import> in
601F<ext/XS-APItest/APItest.xs>). The documentation could be based on it.
602
318bf708
NC
603=head2 Remove the use of SVs as temporaries in dump.c
604
605F<dump.c> contains debugging routines to dump out the contains of perl data
606structures, such as C<SV>s, C<AV>s and C<HV>s. Currently, the dumping code
607B<uses> C<SV>s for its temporary buffers, which was a logical initial
608implementation choice, as they provide ready made memory handling.
609
610However, they also lead to a lot of confusion when it happens that what you're
611trying to debug is seen by the code in F<dump.c>, correctly or incorrectly, as
612a temporary scalar it can use for a temporary buffer. It's also not possible
613to dump scalars before the interpreter is properly set up, such as during
614ithreads cloning. It would be good to progressively replace the use of scalars
615as string accumulation buffers with something much simpler, directly allocated
616by C<malloc>. The F<dump.c> code is (or should be) only producing 7 bit
617US-ASCII, so output character sets are not an issue.
618
619Producing and proving an internal simple buffer allocation would make it easier
620to re-write the internals of the PerlIO subsystem to avoid using C<SV>s for
621B<its> buffers, use of which can cause problems similar to those of F<dump.c>,
622at similar times.
623
5d96f598
NC
624=head2 safely supporting POSIX SA_SIGINFO
625
626Some years ago Jarkko supplied patches to provide support for the POSIX
627SA_SIGINFO feature in Perl, passing the extra data to the Perl signal handler.
628
629Unfortunately, it only works with "unsafe" signals, because under safe
630signals, by the time Perl gets to run the signal handler, the extra
631information has been lost. Moreover, it's not easy to store it somewhere,
632as you can't call mutexs, or do anything else fancy, from inside a signal
633handler.
634
635So it strikes me that we could provide safe SA_SIGINFO support
636
637=over 4
638
639=item 1
640
641Provide global variables for two file descriptors
642
643=item 2
644
645When the first request is made via C<sigaction> for C<SA_SIGINFO>, create a
646pipe, store the reader in one, the writer in the other
647
648=item 3
649
650In the "safe" signal handler (C<Perl_csighandler()>/C<S_raise_signal()>), if
651the C<siginfo_t> pointer non-C<NULL>, and the writer file handle is open,
652
653=over 8
654
655=item 1
656
657serialise signal number, C<struct siginfo_t> (or at least the parts we care
658about) into a small auto char buff
659
660=item 2
661
662C<write()> that (non-blocking) to the writer fd
663
664=over 12
665
666=item 1
667
668if it writes 100%, flag the signal in a counter of "signals on the pipe" akin
669to the current per-signal-number counts
670
671=item 2
672
673if it writes 0%, assume the pipe is full. Flag the data as lost?
674
675=item 3
676
677if it writes partially, croak a panic, as your OS is broken.
678
679=back
680
681=back
682
683=item 4
684
685in the regular C<PERL_ASYNC_CHECK()> processing, if there are "signals on
686the pipe", read the data out, deserialise, build the Perl structures on
687the stack (code in C<Perl_sighandler()>, the "unsafe" handler), and call as
688usual.
689
690=back
691
692I think that this gets us decent C<SA_SIGINFO> support, without the current risk
693of running Perl code inside the signal handler context. (With all the dangers
694of things like C<malloc> corruption that that currently offers us)
695
696For more information see the thread starting with this message:
b4af8972 697L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2008-03/msg00305.html>
5d96f598 698
6d71adcd
NC
699=head2 autovivification
700
701Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
702
703This task is incremental - even a little bit of work on it will help.
704
705=head2 Unicode in Filenames
706
707chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
708opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
709system, truncate, unlink, utime, -X. All these could potentially accept
710Unicode filenames either as input or output (and in the case of system
711and qx Unicode in general, as input or output to/from the shell).
712Whether a filesystem - an operating system pair understands Unicode in
713filenames varies.
714
715Known combinations that have some level of understanding include
716Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
717OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
718create Unicode filenames, what forms of Unicode are accepted and used
719(UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
720and so on, varies. Finding the right level of interfacing to Perl
721requires some thought. Remember that an OS does not implicate a
722filesystem.
723
724(The Windows -C command flag "wide API support" has been at least
725temporarily retired in 5.8.1, and the -C has been repurposed, see
726L<perlrun>.)
727
87a942b1
JH
728Most probably the right way to do this would be this:
729L</"Virtualize operating system access">.
730
6d71adcd
NC
731=head2 Unicode in %ENV
732
733Currently the %ENV entries are always byte strings.
87a942b1 734See L</"Virtualize operating system access">.
6d71adcd 735
799c141b
SH
736(See RT ticket #113536 for information on Win32's handling of %ENV,
737which was fixed to work with native ANSI codepage characters in the
738environment, but still doesn't work with other characters outside of
739that codepage present in the environment.)
740
1f2e7916
JD
741=head2 Unicode and glob()
742
743Currently glob patterns and filenames returned from File::Glob::glob()
87a942b1 744are always byte strings. See L</"Virtualize operating system access">.
1f2e7916 745
6d71adcd
NC
746=head2 use less 'memory'
747
748Investigate trade offs to switch out perl's choices on memory usage.
749Particularly perl should be able to give memory back.
750
751This task is incremental - even a little bit of work on it will help.
752
753=head2 Re-implement C<:unique> in a way that is actually thread-safe
754
755The old implementation made bad assumptions on several levels. A good 90%
756solution might be just to make C<:unique> work to share the string buffer
757of SvPVs. That way large constant strings can be shared between ithreads,
758such as the configuration information in F<Config>.
759
760=head2 Make tainting consistent
761
762Tainting would be easier to use if it didn't take documented shortcuts and
763allow taint to "leak" everywhere within an expression.
764
765=head2 readpipe(LIST)
766
767system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
768running a shell. readpipe() (the function behind qx//) could be similarly
3b17061e
FC
769extended. Note that changing readpipe() itself may not be the solution, as
770it currently has unary precedence, and allowing a list would change the
771precedence.
6d71adcd 772
6d71adcd
NC
773=head2 Audit the code for destruction ordering assumptions
774
775Change 25773 notes
776
f185f654
KW
777 /* Need to check SvMAGICAL, as during global destruction it may be that
778 AvARYLEN(av) has been freed before av, and hence the SvANY() pointer
779 is now part of the linked list of SV heads, rather than pointing to
780 the original body. */
781 /* FIXME - audit the code for other bugs like this one. */
6d71adcd
NC
782
783adding the C<SvMAGICAL> check to
784
785 if (AvARYLEN(av) && SvMAGICAL(AvARYLEN(av))) {
786 MAGIC *mg = mg_find (AvARYLEN(av), PERL_MAGIC_arylen);
787
788Go through the core and look for similar assumptions that SVs have particular
789types, as all bets are off during global destruction.
790
749904bf
JH
791=head2 Extend PerlIO and PerlIO::Scalar
792
793PerlIO::Scalar doesn't know how to truncate(). Implementing this
794would require extending the PerlIO vtable.
795
796Similarly the PerlIO vtable doesn't know about formats (write()), or
797about stat(), or chmod()/chown(), utime(), or flock().
798
799(For PerlIO::Scalar it's hard to see what e.g. mode bits or ownership
800would mean.)
801
802PerlIO doesn't do directories or symlinks, either: mkdir(), rmdir(),
803opendir(), closedir(), seekdir(), rewinddir(), glob(); symlink(),
804readlink().
805
94da6c29
JH
806See also L</"Virtualize operating system access">.
807
d6c1e11f
JH
808=head2 Organize error messages
809
810Perl's diagnostics (error messages, see L<perldiag>) could use
a8d0aeb9 811reorganizing and formalizing so that each error message has its
d6c1e11f
JH
812stable-for-all-eternity unique id, categorized by severity, type, and
813subsystem. (The error messages would be listed in a datafile outside
c4bd451b
CB
814of the Perl source code, and the source code would only refer to the
815messages by the id.) This clean-up and regularizing should apply
d6c1e11f
JH
816for all croak() messages.
817
818This would enable all sorts of things: easier translation/localization
819of the messages (though please do keep in mind the caveats of
820L<Locale::Maketext> about too straightforward approaches to
821translation), filtering by severity, and instead of grepping for a
822particular error message one could look for a stable error id. (Of
823course, changing the error messages by default would break all the
824existing software depending on some particular error message...)
825
826This kind of functionality is known as I<message catalogs>. Look for
827inspiration for example in the catgets() system, possibly even use it
828if available-- but B<only> if available, all platforms will B<not>
de96509d 829have catgets().
d6c1e11f
JH
830
831For the really pure at heart, consider extending this item to cover
47a9c258 832also the warning messages (see L<warnings>, F<regen/warnings.pl>).
3236f110 833
0bdfc961 834=head1 Tasks that need a knowledge of the interpreter
3298bd4d 835
0bdfc961
NC
836These tasks would need C knowledge, and knowledge of how the interpreter works,
837or a willingness to learn.
3298bd4d 838
10517af5
JD
839=head2 forbid labels with keyword names
840
841Currently C<goto keyword> "computes" the label value:
842
843 $ perl -e 'goto print'
844 Can't find label 1 at -e line 1.
845
343c8006
JD
846It is controversial if the right way to avoid the confusion is to forbid
847labels with keyword names, or if it would be better to always treat
848bareword expressions after a "goto" as a label and never as a keyword.
10517af5 849
de6375e3
RGS
850=head2 truncate() prototype
851
852The prototype of truncate() is currently C<$$>. It should probably
f3fccad6 853be C<*$> instead. (This is changed in F<regen/opcodes>.)
de6375e3 854
565590b5
NC
855=head2 error reporting of [$a ; $b]
856
857Using C<;> inside brackets is a syntax error, and we don't propose to change
858that by giving it any meaning. However, it's not reported very helpfully:
859
860 $ perl -e '$a = [$b; $c];'
861 syntax error at -e line 1, near "$b;"
862 syntax error at -e line 1, near "$c]"
863 Execution of -e aborted due to compilation errors.
864
865It should be possible to hook into the tokeniser or the lexer, so that when a
866C<;> is parsed where it is not legal as a statement terminator (ie inside
867C<{}> used as a hashref, C<[]> or C<()>) it issues an error something like
868I<';' isn't legal inside an expression - if you need multiple statements use a
869do {...} block>. See the thread starting at
b4af8972 870L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2008-09/msg00573.html>
565590b5 871
e053a921
RS
872=head2 strict as warnings
873
b8b41556 874See L<http://markmail.org/message/vbrupaslr3bybmvk>, where Joshua ben Jore
e053a921
RS
875writes: I've been of the opinion that everything strict.pm does ought to be
876able to considered just warnings that have been promoted to 'FATAL'.
877
718140ec
NC
878=head2 lexicals used only once
879
880This warns:
881
882 $ perl -we '$pie = 42'
883 Name "main::pie" used only once: possible typo at -e line 1.
884
885This does not:
886
887 $ perl -we 'my $pie = 42'
888
889Logically all lexicals used only once should warn, if the user asks for
d6f4ea2e
SP
890warnings. An unworked RT ticket (#5087) has been open for almost seven
891years for this discrepancy.
718140ec 892
a3d15f9a
RGS
893=head2 UTF-8 revamp
894
85c006b6
KW
895The handling of Unicode is unclean in many places. In the regex engine
896there are especially many problems. The swash data structure could be
897replaced my something better. Inversion lists and maps are likely
898candidates. The whole Unicode database could be placed in-core for a
899huge speed-up. Only minimal work was done on the optimizer when utf8
900was added, with the result that the synthetic start class often will
901fail to narrow down the possible choices when given non-Latin1 input.
4e1c9055 902Karl Williamson has been working on this - talk to him.
a3d15f9a 903
636e63cb
NC
904=head2 state variable initialization in list context
905
906Currently this is illegal:
907
908 state ($a, $b) = foo();
909
a2874905 910In Perl 6, C<state ($a) = foo();> and C<(state $a) = foo();> have different
a8d0aeb9 911semantics, which is tricky to implement in Perl 5 as currently they produce
a2874905 912the same opcode trees. The Perl 6 design is firm, so it would be good to
a8d0aeb9 913implement the necessary code in Perl 5. There are comments in
a2874905
NC
914C<Perl_newASSIGNOP()> that show the code paths taken by various assignment
915constructions involving state variables.
636e63cb 916
a393eb28
RGS
917=head2 A does() built-in
918
919Like ref(), only useful. It would call the C<DOES> method on objects; it
920would also tell whether something can be dereferenced as an
921array/hash/etc., or used as a regexp, etc.
922L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-03/msg00481.html>
923
924=head2 Tied filehandles and write() don't mix
925
926There is no method on tied filehandles to allow them to be called back by
927formats.
4fedb12c 928
53967bb9
RGS
929=head2 Propagate compilation hints to the debugger
930
931Currently a debugger started with -dE on the command-line doesn't see the
932features enabled by -E. More generally hints (C<$^H> and C<%^H>) aren't
933propagated to the debugger. Probably it would be a good thing to propagate
934hints from the innermost non-C<DB::> scope: this would make code eval'ed
935in the debugger see the features (and strictures, etc.) currently in
936scope.
937
d10fc472 938=head2 Attach/detach debugger from running program
1626a787 939
cd793d32
NC
940The old perltodo notes "With C<gdb>, you can attach the debugger to a running
941program if you pass the process ID. It would be good to do this with the Perl
0bdfc961
NC
942debugger on a running Perl program, although I'm not sure how it would be
943done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
1626a787 944
c5fb089a 945=head2 regexp optimizer optional
0bdfc961 946
c5fb089a
DS
947The regexp optimizer is not optional. It should be configurable to be optional
948and to allow its performance to be measured and its bugs to be easily
949demonstrated.
0bdfc961 950
ef36c6a7
RGS
951=head2 C</w> regex modifier
952
953That flag would enable to match whole words, and also to interpolate
954arrays as alternations. With it, C</P/w> would be roughly equivalent to:
955
956 do { local $"='|'; /\b(?:P)\b/ }
957
b4af8972
RB
958See
959L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2007-01/msg00400.html>
ef36c6a7
RGS
960for the discussion.
961
0bdfc961
NC
962=head2 optional optimizer
963
964Make the peephole optimizer optional. Currently it performs two tasks as
965it walks the optree - genuine peephole optimisations, and necessary fixups of
966ops. It would be good to find an efficient way to switch out the
967optimisations whilst keeping the fixups.
968
969=head2 You WANT *how* many
970
971Currently contexts are void, scalar and list. split has a special mechanism in
972place to pass in the number of return values wanted. It would be useful to
973have a general mechanism for this, backwards compatible and little speed hit.
974This would allow proposals such as short circuiting sort to be implemented
975as a module on CPAN.
976
de535794 977=head2 Self-ties
2810d901 978
de535794 979Self-ties are currently illegal because they caused too many segfaults. Maybe
a8d0aeb9 980the causes of these could be tracked down and self-ties on all types
de535794 981reinstated.
0bdfc961
NC
982
983=head2 Optimize away @_
984
985The old perltodo notes "Look at the "reification" code in C<av.c>".
986
87a942b1
JH
987=head2 Virtualize operating system access
988
989Implement a set of "vtables" that virtualizes operating system access
9fe0b8be
RS
990(chdir(), chmod(), dbmopen(), getenv(), glob(), link(), mkdir(), open(),
991opendir(), readdir(), rename(), rmdir(), stat(), sysopen(), uname(),
992unlink(), etc.) At the very least these interfaces should take SVs as
993"name" arguments instead of bare char pointers; probably the most
994flexible and extensible way would be for the Perl-facing interfaces to
995accept HVs. The system needs to be per-operating-system and
996per-file-system hookable/filterable, preferably both from XS and Perl
997level (L<perlport/"Files and Filesystems"> is good reading at this
998point, in fact, all of L<perlport> is.)
87a942b1 999
e1a3d5d1
JH
1000This has actually already been implemented (but only for Win32),
1001take a look at F<iperlsys.h> and F<win32/perlhost.h>. While all Win32
1002variants go through a set of "vtables" for operating system access,
e1020413 1003non-Win32 systems currently go straight for the POSIX/Unix-style
e1a3d5d1
JH
1004system/library call. Similar system as for Win32 should be
1005implemented for all platforms. The existing Win32 implementation
1006probably does not need to survive alongside this proposed new
1007implementation, the approaches could be merged.
87a942b1
JH
1008
1009What would this give us? One often-asked-for feature this would
94da6c29
JH
1010enable is using Unicode for filenames, and other "names" like %ENV,
1011usernames, hostnames, and so forth.
1012(See L<perlunicode/"When Unicode Does Not Happen">.)
1013
1014But this kind of virtualization would also allow for things like
1015virtual filesystems, virtual networks, and "sandboxes" (though as long
1016as dynamic loading of random object code is allowed, not very safe
1017sandboxes since external code of course know not of Perl's vtables).
1018An example of a smaller "sandbox" is that this feature can be used to
1019implement per-thread working directories: Win32 already does this.
1020
1021See also L</"Extend PerlIO and PerlIO::Scalar">.
87a942b1 1022
52960e22
JC
1023=head2 repack the optree
1024
af4a745c
FC
1025B<Note:> This entry was written in reference to the I<old> slab allocator,
1026removed in commit 7aef8e5bd14.
1027
52960e22 1028Repacking the optree after execution order is determined could allow
057163d7 1029removal of NULL ops, and optimal ordering of OPs with respect to cache-line
2723c0fb 1030filling. I think that
057163d7
NC
1031the best way to do this is to make it an optional step just before the
1032completed optree is attached to anything else, and to use the slab allocator
2723c0fb
FC
1033unchanged--but allocate a single slab the right size, avoiding partial
1034slabs--, so that freeing ops is identical whether or not this step runs.
057163d7
NC
1035Note that the slab allocator allocates ops downwards in memory, so one would
1036have to actually "allocate" the ops in reverse-execution order to get them
1037contiguous in memory in execution order.
1038
b4af8972
RB
1039See
1040L<http://www.nntp.perl.org/group/perl.perl5.porters/2007/12/msg131975.html>
057163d7
NC
1041
1042Note that running this copy, and then freeing all the old location ops would
1043cause their slabs to be freed, which would eliminate possible memory wastage if
1044the previous suggestion is implemented, and we swap slabs more frequently.
52960e22 1045
12e06b6f
NC
1046=head2 eliminate incorrect line numbers in warnings
1047
1048This code
1049
1050 use warnings;
1051 my $undef;
f703fc96 1052
12e06b6f
NC
1053 if ($undef == 3) {
1054 } elsif ($undef == 0) {
1055 }
1056
18a16cc5 1057used to produce this output:
12e06b6f
NC
1058
1059 Use of uninitialized value in numeric eq (==) at wrong.pl line 4.
1060 Use of uninitialized value in numeric eq (==) at wrong.pl line 4.
1061
18a16cc5
NC
1062where the line of the second warning was misreported - it should be line 5.
1063Rafael fixed this - the problem arose because there was no nextstate OP
1064between the execution of the C<if> and the C<elsif>, hence C<PL_curcop> still
1065reports that the currently executing line is line 4. The solution was to inject
1066a nextstate OPs for each C<elsif>, although it turned out that the nextstate
1067OP needed to be a nulled OP, rather than a live nextstate OP, else other line
1068numbers became misreported. (Jenga!)
12e06b6f
NC
1069
1070The problem is more general than C<elsif> (although the C<elsif> case is the
1071most common and the most confusing). Ideally this code
1072
1073 use warnings;
1074 my $undef;
f703fc96 1075
12e06b6f
NC
1076 my $a = $undef + 1;
1077 my $b
1078 = $undef
1079 + 1;
1080
1081would produce this output
1082
f185f654
KW
1083 Use of uninitialized value $undef in addition (+) at wrong.pl line 4.
1084 Use of uninitialized value $undef in addition (+) at wrong.pl line 7.
12e06b6f
NC
1085
1086(rather than lines 4 and 5), but this would seem to require every OP to carry
1087(at least) line number information.
1088
1089What might work is to have an optional line number in memory just before the
1090BASEOP structure, with a flag bit in the op to say whether it's present.
1091Initially during compile every OP would carry its line number. Then add a late
c5fb089a 1092pass to the optimizer (potentially combined with L</repack the optree>) which
12e06b6f
NC
1093looks at the two ops on every edge of the graph of the execution path. If
1094the line number changes, flags the destination OP with this information.
1095Once all paths are traced, replace every op with the flag with a
1096nextstate-light op (that just updates C<PL_curcop>), which in turn then passes
1097control on to the true op. All ops would then be replaced by variants that
1098do not store the line number. (Which, logically, why it would work best in
1099conjunction with L</repack the optree>, as that is already copying/reallocating
1100all the OPs)
1101
18a16cc5
NC
1102(Although I should note that we're not certain that doing this for the general
1103case is worth it)
1104
52960e22
JC
1105=head2 optimize tail-calls
1106
1107Tail-calls present an opportunity for broadly applicable optimization;
1108anywhere that C<< return foo(...) >> is called, the outer return can
1109be replaced by a goto, and foo will return directly to the outer
1110caller, saving (conservatively) 25% of perl's call&return cost, which
1111is relatively higher than in C. The scheme language is known to do
1112this heavily. B::Concise provides good insight into where this
1113optimization is possible, ie anywhere entersub,leavesub op-sequence
1114occurs.
1115
1116 perl -MO=Concise,-exec,a,b,-main -e 'sub a{ 1 }; sub b {a()}; b(2)'
1117
1118Bottom line on this is probably a new pp_tailcall function which
1119combines the code in pp_entersub, pp_leavesub. This should probably
1120be done 1st in XS, and using B::Generate to patch the new OP into the
1121optrees.
1122
e12cb30b 1123=head2 Add C<0odddd>
0c397127
KW
1124
1125It has been proposed that octal constants be specifiable through the syntax
1126C<0oddddd>, parallel to the existing construct to specify hex constants
1127C<0xddddd>
1128
bf7d9bd8
AC
1129=head2 Revisit the regex super-linear cache code
1130
1131Perl executes regexes using the traditional backtracking algorithm, which
1132makes it possible to implement a variety of powerful pattern-matching
1133features (like embedded code blocks), at the cost of taking exponential time
1134to run on some pathological patterns. The exponential-time problem is
1135mitigated by the I<super-linear cache>, which detects when we're processing
1136such a pathological pattern, and does some additional bookkeeping to avoid
1137much of the work. However, that code has bit-rotted a little; some patterns
1138don't make as much use of it as they should. The proposal is to analyse
1139where the current cache code has problems, and extend it to cover those cases.
1140
1141See also
1142L<http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2013-01/msg00339.html>
1143
0bdfc961
NC
1144=head1 Big projects
1145
1146Tasks that will get your name mentioned in the description of the "Highlights
1e2ea6fa 1147of 5.25.2"
0bdfc961
NC
1148
1149=head2 make ithreads more robust
1150
45a81a90 1151Generally make ithreads more robust.
0bdfc961
NC
1152
1153This task is incremental - even a little bit of work on it will help, and
1154will be greatly appreciated.
1155
07577ec1
FC
1156One bit would be to determine how to clone directory handles on systems
1157without a C<fchdir> function (in sv.c:Perl_dirp_dup).
6c047da7 1158
59c7f7d5
RGS
1159Fix Perl_sv_dup, et al so that threads can return objects.
1160
44a7a252
JV
1161=head1 Tasks for microperl
1162
1163
1164[ Each and every one of these may be obsolete, but they were listed
1165 in the old Todo.micro file]
1166
44a7a252
JV
1167=head2 do away with fork/exec/wait?
1168
1169(system, popen should be enough?)
1170
1171=head2 some of the uconfig.sh really needs to be probed (using cc) in buildtime:
1172
1173(uConfigure? :-) native datatype widths and endianness come to mind
1174