This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Zefram [Tue, 19 Dec 2017 17:12:33 +0000 (17:12 +0000)]
remove broken PerlLIO_mkstemp() definition
This was only defined in non-PERL_IMPLICIT_SYS mode, and not used
anywhere.
Zefram [Tue, 19 Dec 2017 17:11:25 +0000 (17:11 +0000)]
factor out remaining fcntl F_SETFD calls
Zefram [Tue, 19 Dec 2017 16:53:39 +0000 (16:53 +0000)]
set FD_CLOEXEC atomically in easy cases
In many places where a file descriptor is being opened, open it with
FD_CLOEXEC already set if possible. This commit covers the easy cases,
where the file descriptor arises without the use of PerlIO, pp_open,
or my_popen.
Zefram [Tue, 19 Dec 2017 09:38:37 +0000 (09:38 +0000)]
*_cloexec() I/O functions
New functions PerlLIO_dup_cloexec(), PerlLIO_dup2_cloexec(),
PerlLIO_open_cloexec(), PerlLIO_open3_cloexec(), PerlProc_pipe_cloexec(),
PerlSock_socket_cloexec(), PerlSock_accept_cloexec(), and
PerlSock_socketpair_cloexec() each do the same thing as their
"_cloexec"-less counterpart, but return with the FD_CLOEXEC flag set on
each new file descriptor. They set the flag atomically as part of the
file descriptor creation syscall where possible, but will fall back to
setting it separately from creation where necessary.
In all cases, setting the flag atomically depends not only on the correct
syscall interface being defined, but on it being actually implemented
in the runtime kernel. Each function will experiment to see whether
the atomic flag setting actually works, and is prepared for the flag to
cause EINVAL or ENOSYS or to be ignored.
Zefram [Tue, 19 Dec 2017 10:35:58 +0000 (10:35 +0000)]
configure probes for accept4(), dup3(), pipe2()
These will shortly be used to implement I/O operations that create file
descriptors with the FD_CLOEXEC flag set atomically.
Zefram [Fri, 22 Dec 2017 10:07:47 +0000 (10:07 +0000)]
set PERL_EXIT_DESTRUCT_END in all embeddings
The new tests in commit
503bc07b4b9e34ed04a725b2bc8faec1ae0f3be2
showed up platform differences in whether END blocks get run.
Set PERL_EXIT_DESTRUCT_END in all embeddings to make this consistent.
Zefram [Fri, 22 Dec 2017 07:54:38 +0000 (07:54 +0000)]
test more about exit from special blocks
Commit
0301e899536a22752f40481d8a1d141b7a7dda82 introduced some tests
about ways of exiting from special blocks. Make those tests more
thorough, checking which kinds of special blocks execute after exiting
from a special block. This tests, for example, that exiting from a BEGIN
block prevents INIT blocks from running, which when applied to exit(0)
is how that bugfix commit broke Module::Install [perl #132577].
Zefram [Fri, 22 Dec 2017 05:32:41 +0000 (05:32 +0000)]
preserve numericness of system() args on Win32
On Windows there's a nasty variation in the meaning of arguments
to Perl's system(), in which a numeric first argument isn't used as
part of the command to run, but instead selects between two different
operations to perform with the command (whether to wait for the command
to complete or not). Therefore the reduction of argument scalars to
their operative values in the parent process, which was added in commit
64def2aeaeb63f92dadc6dfa33486c1d7b311963, needs to preserve numericness
of arguments on Windows. Fixes [perl #132633].
Chris 'BinGOs' Williams [Thu, 21 Dec 2017 08:59:41 +0000 (08:59 +0000)]
Cogito ergo sum
Zefram [Thu, 21 Dec 2017 05:40:49 +0000 (05:40 +0000)]
fix up faulty perl embeddings
Some platform-specific embeddings of perl were misusing the
return values from perl_parse() and perl_run(), in some cases
causing failure due to exit(0) combined with the recent changes
in commit
0301e899536a22752f40481d8a1d141b7a7dda82. Commit
d4a50999a5525c2681d59cae5fcd94f94ff897fd partially fixed a Windows
embedding. More fully fix that, along with NetWare and OS/2. Even in
embeddings with correct logic, stop using a variable named "exitstatus"
to hold the result of perl_parse() or perl_run(), to avoid misleading
people who copy the code.
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 23:49:23 +0000 (23:49 +0000)]
Update Module-CoreList for v5.27.8
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 23:45:41 +0000 (23:45 +0000)]
Bump the perl version in various places for v5.27.8
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 23:28:57 +0000 (23:28 +0000)]
New perldelta for v5.27.8
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 23:21:53 +0000 (23:21 +0000)]
Include epigraph for v5.27.7
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 20:44:26 +0000 (20:44 +0000)]
add new release to perlhist
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 20:17:39 +0000 (20:17 +0000)]
Finalise perldelta
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 19:49:33 +0000 (19:49 +0000)]
Update Module::CoreList for v5.27.7
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 19:32:45 +0000 (19:32 +0000)]
Start finalisation of perldelta
Chris 'BinGOs' Williams [Wed, 20 Dec 2017 19:21:30 +0000 (19:21 +0000)]
Resolve issue with Win32 caused by
0301e8995
From #p5p
[18:12] < xenu> is it just me or is it impossible to build perl using visual c++?
[18:13] < xenu> it fails with 'NMAKE : fatal error U1077: '..\perl.exe' : return code '0x100'"
[18:14] < pink_mist> I'm pretty sure it ought to be possible
[18:14] < xenu> the funny part is that perl.exe built by msvc *always* exits with 255 exit code
[18:14] < xenu> even if you run it like this ..\perl.exe -e "print 1"
[18:14] < xenu> it will print '1' and terminate with exit code 255
[18:42] < xenu> Zefram:
0301e899536a22752f40481d8a1d141b7a7dda82 broke win32
[18:43] < xenu> basically the problem is than PerlRun in win32/perllib.c uses return value of perl_run() as a exit code
[19:08] < BinGOs> xenu: does 'exitstatus = perl_destruct(my_perl);' fix the issue?
[19:20] < xenu> BinGOs: yep
[19:20] < BinGOs> okay cool thanks for confirming.
[19:20] < BinGOs> I'll commit that now.
Karl Williamson [Fri, 9 Jun 2017 13:27:31 +0000 (07:27 -0600)]
t/re/reg_mesg.t: White-space only
Indent a newly-formed block
Karl Williamson [Fri, 9 Jun 2017 04:05:15 +0000 (22:05 -0600)]
t/re/reg_mesg.t: Test that death actually happens.
Until now, patterns that are supposed to die weren't actually being
checked for that. Instead the message that was raised was checked
against the expected value, and if they matched, it assumed that death
accompanied it. But this isn't necessarily true. Also, some entries
were given an empty expectation, which was wrong. So this makes sure of
that. Those empty patterns matched anything at all, so things went
undetected.
To be sure of death, this commit appends a statement with a lingering
effect to the string eval of what's supposed to die. If the effect
happens, it didn't die.
Karl Williamson [Tue, 19 Dec 2017 18:43:46 +0000 (11:43 -0700)]
reg_mesg.t: Move some tests to proper places
This was detected in [perl #131522]. Some tests were in the death
section, but don't die. The next commit will add tests that actually
work to detect death.
David Mitchell [Tue, 19 Dec 2017 15:40:41 +0000 (15:40 +0000)]
[MERGE] s/// return value fixups
David Mitchell [Tue, 19 Dec 2017 15:28:06 +0000 (15:28 +0000)]
s/// in boolean context: simplify return value
Normally s/// returns a count of the number of iterations, but
as an optimisation, in boolean context it returns PL_sv_yes/PL_sv_zero
instead. But in the places where it decides which immortal var to return,
the number of iterations is always > 0, so PL_sv_zero never gets returned.
So skip testing whether iters > 0 and always just return PL_sv_yes.
(In non-boolean scalar context, it still returns the iteration count as
before.)
David Mitchell [Tue, 19 Dec 2017 14:52:49 +0000 (14:52 +0000)]
avoid tainting boolean return value of s///
RT #132385
s/// normally returns an integer count, but sometimes for efficiency it
will return a boolean instead (PL_sv_yes/PL_sv_zero).
In these cases, don't try to taint the return value, since it will die
with 'Modification of a read-only value'.
David Mitchell [Tue, 19 Dec 2017 12:09:15 +0000 (12:09 +0000)]
s///: return boolean in not-in-place branch
A while back, s/// (and other ops) were optimised to return
PL_sv_yes/PL_sv_zero rather than an iteration count in boolean context.
This optimisation was missed in one place in pp_subst(): the 'can modify
in place' branch was done, but the other branch was missed.
This commit fixes that.
Karen Etheridge [Tue, 19 Dec 2017 03:01:26 +0000 (19:01 -0800)]
minor punctuation improvement
Karl Williamson [Tue, 19 Dec 2017 02:04:56 +0000 (19:04 -0700)]
Forbid literal \v in /[]/ under re strict, except under /x
Vertical space in a bracketed character class is likely to be an error;
either a typo or omission of /x. It will match literally, but be
displayed like /x is in effect.
For now, just add this check to re 'strict', which is experimental,
designed to catch things the current regex parser doesn't.
Karl Williamson [Tue, 19 Dec 2017 02:03:26 +0000 (19:03 -0700)]
perldiag: Fix obsolete text
Since 5.26.0, code points must fit in a signed integer, not unsigned.
This commit should be considered for backporting to a future 5.26.x
release
Karl Williamson [Tue, 19 Dec 2017 02:02:37 +0000 (19:02 -0700)]
reg_mesg.t: Fix comment
Karl Williamson [Tue, 19 Dec 2017 02:01:37 +0000 (19:01 -0700)]
reg_mesg.t: Fix typo
James E Keenan [Mon, 18 Dec 2017 22:02:55 +0000 (17:02 -0500)]
Demonstrate how File::Spec->catfile('.', @other_args) works.
For: RT # 131296
Karl Williamson [Fri, 8 Dec 2017 05:10:52 +0000 (22:10 -0700)]
Emulate C99 int_fast8_t, etc typedefs
These are useful if you know you have a variable with a restricted range,
and so could fit in a byte or 16 bits, but speed is more important.
These C99 typedefs allows you to specify the minimal size you need, and
allows the compiler to substitute a wider type if it is faster.
This commit adds typedefs spelled the same as the C99 ones, but upper
cased. On non-C99 compilers, it just uses 'int' behind the scenes,
which should be safe.
These are currently restricted to core to be sure these aren't a bad
idea before they are made public.
Karl Williamson [Mon, 18 Dec 2017 19:49:55 +0000 (12:49 -0700)]
mktables: Use already set variable
The value for this variable is already known; use that instead of
rederiving it.
Karl Williamson [Sat, 16 Dec 2017 06:41:51 +0000 (23:41 -0700)]
Change number to mnemonic
This is in preparation for future commits that will use it in multiple
places
Father Chrysostomos [Mon, 18 Dec 2017 17:37:08 +0000 (09:37 -0800)]
perlxs: Document bitwise calling conventions
Chris 'BinGOs' Williams [Mon, 18 Dec 2017 13:31:43 +0000 (13:31 +0000)]
Customisations for Digest-SHA tests
Chris 'BinGOs' Williams [Mon, 18 Dec 2017 13:15:34 +0000 (13:15 +0000)]
Update Digest-SHA to CPAN version 6.00
[DELTA]
6.00 Fri Dec 8 22:44:44 MST 2017
- added "tag" option (BSD-style checksums) to shasum
-- consistent with GNU sha1sum, sha224sum, etc.
-- thanks to Christopher Tubbs for suggestion
- modified SHA.pm to use XSLoader
-- falls back to DynaLoader if necessary
- decoupled Digest::SHA and Digest::SHA::PurePerl
-- scripts/tests in each module now fully independent
-- Digest::SHA::PurePerl to be phased out eventually
-- past versions will remain available
David Mitchell [Mon, 18 Dec 2017 11:50:03 +0000 (11:50 +0000)]
enable PERL_OP_PARENT always
Remove support for PERL_NO_OP_PARENT, which was a way of turning off the
(on by default) PERL_OP_PARENT define.
See http://nntp.perl.org/group/perl.perl5.porters/248072.
Zefram [Mon, 18 Dec 2017 03:21:05 +0000 (03:21 +0000)]
align version numbers in the PathTools distro
Zefram [Mon, 18 Dec 2017 03:18:39 +0000 (03:18 +0000)]
remove support for running PathTools on MacOS
MacOS Classic hasn't been a supported platform since 5.14. Note that
the code for processing MacOS-style paths remains, and can be run on
supported platforms. [perl #130818]
Zefram [Mon, 18 Dec 2017 03:12:04 +0000 (03:12 +0000)]
fix pod markup in warnings.pm
Commit
c4583f59133164b3f392c31e9b9573276ec17e74 introduced a pod error.
Zefram [Mon, 18 Dec 2017 03:08:31 +0000 (03:08 +0000)]
remove support for running PathTools on EPOC
EPOC hasn't been a supported platform since 5.18. Note that the code
for processing EPOC-style paths remains, and can be run on supported
platforms. [perl #130818]
Father Chrysostomos [Mon, 18 Dec 2017 02:43:55 +0000 (18:43 -0800)]
perldelta: Avoid ‘deprecated’ for non-deprecated module
Saying the module is deprecated (in core) can easily be misunder-
stood. The module itself is certainly not deprecated.
Father Chrysostomos [Mon, 18 Dec 2017 02:00:33 +0000 (18:00 -0800)]
perldelta for #132468 warnings functions
Father Chrysostomos [Mon, 18 Dec 2017 01:59:33 +0000 (17:59 -0800)]
perldelta: Alphabetise modules
Father Chrysostomos [Mon, 18 Dec 2017 01:56:52 +0000 (17:56 -0800)]
[perl #132468] At _at_level warnings functions
Father Chrysostomos [Mon, 11 Dec 2017 21:38:24 +0000 (13:38 -0800)]
pp_ctl.c: Add -D output for inward goto
Zefram [Sun, 17 Dec 2017 11:02:23 +0000 (11:02 +0000)]
merge branch zefram/dumb_match
Karl Williamson [Fri, 10 Nov 2017 01:31:15 +0000 (18:31 -0700)]
Unicode::UCD: max code point is now IV_MAX
Return the correct value when asked.
Karl Williamson [Sat, 16 Dec 2017 19:13:01 +0000 (12:13 -0700)]
regcomp.sym: Clarify regnode comment
Karl Williamson [Sat, 16 Dec 2017 18:53:15 +0000 (11:53 -0700)]
regen/mk_invlists.pl: reformulate a branch and outdent
This turns a test whose one branch dies into a postfix type, which means
there is no need for an else, and that code can be outdented.
Karl Williamson [Sat, 16 Dec 2017 18:48:12 +0000 (11:48 -0700)]
regen/mk_invlists.pl: Outdent code
This was unnecessarily indented. This commit also removes an
unnecessary {...} block, and outdents the interior as well.
No other changes are made
Karl Williamson [Sat, 16 Dec 2017 18:37:21 +0000 (11:37 -0700)]
regen/mk_invlists.pl: Fix typo
Karl Williamson [Sun, 17 Dec 2017 00:05:50 +0000 (17:05 -0700)]
perlre: '#' is a metacharacter under /x
Zefram [Sat, 16 Dec 2017 22:46:33 +0000 (22:46 +0000)]
make dNOOP usable outside function in C++
Commit
7db66e12883f0832ca80164b723768b848187bda changed dNOOP in C++ from
an external function declaration (which was running into inconsistent
linkage) to a no-op statement. The no-op statement is only valid where
statements are valid, i.e., inside a function, so this doesn't actually
work everywhere that a declaration is expected.
Switch dNOOP to be an incomplete struct declaration, regardless of
language. This can certainly be used both inside and outside functions,
is valid to repeat, and does not have the linkage problem of a function
declaration. It should also be as unlikely to attract an "unused"
warning as an external function declaration.
Zefram [Sat, 16 Dec 2017 09:30:16 +0000 (09:30 +0000)]
typo fix in doc
Zefram [Sat, 16 Dec 2017 08:34:25 +0000 (08:34 +0000)]
show modern usage for {XS,Dyna}Loader
Fixes [perl #132247].
Zefram [Sat, 16 Dec 2017 07:12:48 +0000 (07:12 +0000)]
semicolon-friendly diagnostic control
New macros {GCC,CLANG}_DIAG_{IGNORE,RESTORE}_{DECL,STMT}, which take a
following semicolon. It is necessary to use the _DECL or _STMT version
as appropriate to the context. Fixes [perl #130726].
Zefram [Sat, 16 Dec 2017 05:33:20 +0000 (05:33 +0000)]
perform system() arg processing before fork
A lot of things can happen when stringifying an argument list: side
effects, warnings, exceptions. In the case of system(), these effects
should happen in the context of the parent process. The stringification
can also depend on which process it happens in, as in the case of
$$, and in that case it should also happen in the parent process.
Therefore reduce the argument scalars to strings first thing in pp_system.
Fixes [perl #121105].
Zefram [Sat, 16 Dec 2017 04:47:08 +0000 (04:47 +0000)]
update and clarify "die" and "warn" doc
They didn't fully describe the handling of reference operands, and
there were other errors in their description of the treatment of the
operand list. Fixes [perl #121372].
Zefram [Sat, 16 Dec 2017 02:17:08 +0000 (02:17 +0000)]
retry gdbm_open() upon block size error
It is reported that gdbm_open() with default block size can fail on
filesystems with a non-power-of-two block size. In that event, retry
opening, forcing a power-of-two block size. Fixes [perl #119623].
Zefram [Sat, 16 Dec 2017 01:48:22 +0000 (01:48 +0000)]
add ticket number to perldelta entry
Zefram [Sat, 16 Dec 2017 01:44:32 +0000 (01:44 +0000)]
widen size-type variables in pack/unpack
Most size-type variables in pp_pack.c were of type I32, with a
smattering of other types. Use SSize_t in place of I32, and generally
use size_t-width variables as appropriate. Fixes [perl #119367].
Marco Fontani [Fri, 15 Dec 2017 17:16:31 +0000 (10:16 -0700)]
perlfunc: remove stray plus from start of line
Looks to be an artifact of applying my original patch, which was
amended as part of
e135ff695231a81e2a70a739e8d813525432fd4d, whereby
the text was slightly amended.
Zefram [Fri, 15 Dec 2017 09:51:33 +0000 (09:51 +0000)]
document GDBM_File's delete vs each conflict
Fixes [perl #117449].
Zefram [Fri, 15 Dec 2017 07:00:11 +0000 (07:00 +0000)]
rewrite "ref" documentation for clarity
Zefram [Fri, 15 Dec 2017 05:12:32 +0000 (05:12 +0000)]
clean up directories of split pod files
Fixes [perl #114316].
Zefram [Fri, 15 Dec 2017 02:14:25 +0000 (02:14 +0000)]
Zefram [Fri, 15 Dec 2017 02:12:42 +0000 (02:12 +0000)]
Zefram [Fri, 15 Dec 2017 02:11:13 +0000 (02:11 +0000)]
Zefram [Fri, 15 Dec 2017 02:09:15 +0000 (02:09 +0000)]
Zefram [Fri, 15 Dec 2017 02:06:54 +0000 (02:06 +0000)]
Zefram [Fri, 15 Dec 2017 02:03:33 +0000 (02:03 +0000)]
Zefram [Fri, 15 Dec 2017 02:00:05 +0000 (02:00 +0000)]
Zefram [Fri, 15 Dec 2017 01:58:26 +0000 (01:58 +0000)]
Zefram [Fri, 15 Dec 2017 01:55:17 +0000 (01:55 +0000)]
Zefram [Fri, 15 Dec 2017 01:53:56 +0000 (01:53 +0000)]
Zefram [Fri, 15 Dec 2017 01:50:55 +0000 (01:50 +0000)]
Zefram [Fri, 15 Dec 2017 01:48:41 +0000 (01:48 +0000)]
Zefram [Fri, 15 Dec 2017 01:46:30 +0000 (01:46 +0000)]
Zefram [Fri, 15 Dec 2017 01:44:39 +0000 (01:44 +0000)]
Zefram [Fri, 15 Dec 2017 01:43:34 +0000 (01:43 +0000)]
Zefram [Fri, 15 Dec 2017 01:39:19 +0000 (01:39 +0000)]
Zefram [Fri, 15 Dec 2017 01:32:39 +0000 (01:32 +0000)]
Zefram [Fri, 15 Dec 2017 01:30:38 +0000 (01:30 +0000)]
Zefram [Fri, 15 Dec 2017 01:29:23 +0000 (01:29 +0000)]
Zefram [Fri, 15 Dec 2017 01:27:12 +0000 (01:27 +0000)]
Zefram [Fri, 15 Dec 2017 01:24:36 +0000 (01:24 +0000)]
Zefram [Fri, 15 Dec 2017 01:20:03 +0000 (01:20 +0000)]
Zefram [Fri, 15 Dec 2017 01:16:14 +0000 (01:16 +0000)]
Zefram [Fri, 15 Dec 2017 01:14:01 +0000 (01:14 +0000)]
Zefram [Fri, 15 Dec 2017 01:11:47 +0000 (01:11 +0000)]
Zefram [Fri, 15 Dec 2017 01:09:48 +0000 (01:09 +0000)]
Zefram [Fri, 15 Dec 2017 01:08:38 +0000 (01:08 +0000)]
Zefram [Fri, 15 Dec 2017 01:07:08 +0000 (01:07 +0000)]
Zefram [Fri, 15 Dec 2017 01:04:32 +0000 (01:04 +0000)]
Zefram [Fri, 15 Dec 2017 01:02:39 +0000 (01:02 +0000)]
Zefram [Fri, 15 Dec 2017 00:59:13 +0000 (00:59 +0000)]
Zefram [Fri, 15 Dec 2017 00:56:11 +0000 (00:56 +0000)]