Bo Lindbergh [Thu, 14 Dec 2006 13:34:06 +0000]
perl5db.pl Mac OS X fix
Message-Id: <
658AFA03-517E-4068-93CA-06D4A93B5CA5@hagernas.com>
p4raw-id: //depot/perl@29555
Nicholas Clark [Thu, 14 Dec 2006 11:23:16 +0000]
Turn a tab into 2 spaces to make more room for comments about SV flags.
Sorry about damaging the change history - dig through this change with
http://public.activestate.com/cgi-bin/perlbrowse?show_blame=Show+Annotated+File&filename=sv.h&rev=&change=29552
p4raw-id: //depot/perl@29554
Steve Hay [Thu, 14 Dec 2006 09:37:45 +0000]
Silence a VC compiler warning
p4raw-id: //depot/perl@29553
Nicholas Clark [Wed, 13 Dec 2006 20:56:24 +0000]
Swap the order of PERL_FBM_PREVIOUS_{L,H}_OFFSET_FROM_TABLE on little
endian machines to hint to the compiler that a 16 bit read could work.
p4raw-id: //depot/perl@29552
Nicholas Clark [Wed, 13 Dec 2006 19:55:08 +0000]
Restore the two comments describing BmRARE and BmPREVIOUS that I'd
inadvertently deleted. Explain the setting and testing of
SVpbm_VALID.
p4raw-id: //depot/perl@29551
Ilya Zakharevich [Tue, 12 Dec 2006 23:28:25 +0000]
Text mode wrongly set on pipe file descriptors
Message-ID: <
20061213072825.GA26300@powdermilk.math.berkeley.edu>
p4raw-id: //depot/perl@29550
Dave Mitchell [Wed, 13 Dec 2006 17:24:49 +0000]
make -Dpv display the parser stack before reduction
(rather than afterwards)
p4raw-id: //depot/perl@29549
Dave Mitchell [Wed, 13 Dec 2006 17:16:22 +0000]
misc MAD coredump fixes and parser leak fixes
- fix MAD coredump in tr///
- fix mad coredump in multi-line string literals
- kill some MAD uninit value warnings
- don't allow assignment to $n in perly.y
- make op_dump handle op_latefree flags
p4raw-id: //depot/perl@29548
Dave Mitchell [Wed, 13 Dec 2006 12:47:35 +0000]
fix double free introduced by #29543 (spotted by Nicholas)
p4raw-id: //depot/perl@29547
Anton Berezin [Tue, 12 Dec 2006 19:59:38 +0000]
t/op/groups.t: prevent failure on modern FreeBSDs
Message-ID: <
20061212185938.GB49084@heechee.tobez.org>
p4raw-id: //depot/perl@29546
Steve Hay [Wed, 13 Dec 2006 10:37:11 +0000]
Do the same as #29514 for Win32
p4raw-id: //depot/perl@29545
Nicholas Clark [Wed, 13 Dec 2006 08:35:43 +0000]
Eliminate PVBM. Store fast Boyer-Moore tables in PVGV.
Add the placeholder for new type, temporarily named BIND, for binding
and aliasing in 6 on 5.
p4raw-id: //depot/perl@29544
Dave Mitchell [Wed, 13 Dec 2006 01:47:34 +0000]
fix parser leaks caused by croaking while shifting or reducing
e.g. these no longer leak:
eval q[my $x; local $x] while 1;
eval q[$s = sub <> {}] while 1;
p4raw-id: //depot/perl@29543
Dave Mitchell [Wed, 13 Dec 2006 00:16:18 +0000]
fixup some ival/opval type changes from perly.y/madly.y merger
p4raw-id: //depot/perl@29542
Nicholas Clark [Tue, 12 Dec 2006 23:02:56 +0000]
In theory Perl_magic_setbm() should clear SvTAIL() too. In practice,
I don't think that it matters as this routine is never actualy called
because nothing exposes PVBMs to the world so that the world can assign
to them.
p4raw-id: //depot/perl@29541
Nicholas Clark [Tue, 12 Dec 2006 22:30:29 +0000]
Correct the dumping logic for the various bits of flag bending
associated with change 27313 or therabouts.
p4raw-id: //depot/perl@29540
Nicholas Clark [Tue, 12 Dec 2006 22:03:21 +0000]
Oops. As of some point one stopped being able to assign integers and
floating point values to typeglobs.
p4raw-id: //depot/perl@29539
Nicholas Clark [Tue, 12 Dec 2006 21:54:22 +0000]
Correct some assumptions about PVBM table offset in B.xs
p4raw-id: //depot/perl@29538
Nicholas Clark [Tue, 12 Dec 2006 21:31:10 +0000]
Assert that SvPAD_TYPED_on(), SvPAD_OUR_on() and SvPAD_STATE_on()
are never called on PVGVs.
p4raw-id: //depot/perl@29537
Nicholas Clark [Tue, 12 Dec 2006 21:00:37 +0000]
Assert that PVGVs are never SvVALID() in the PVBM sense.
p4raw-id: //depot/perl@29536
Nicholas Clark [Tue, 12 Dec 2006 19:54:05 +0000]
pads haven't actually contained any PVGVs since change 27313.
So correct the docs and assert this in the code.
p4raw-id: //depot/perl@29535
Nicholas Clark [Tue, 12 Dec 2006 18:52:55 +0000]
Shrink Perl_sv_peek() and Perl_do_sv_dump() by using type to name
lookup tables.
p4raw-id: //depot/perl@29534
Steve Hay [Tue, 12 Dec 2006 18:02:19 +0000]
Restore (differently) the setting of $Config{osvers} on Win32
that was removed by #29509
p4raw-id: //depot/perl@29533
Steve Hay [Tue, 12 Dec 2006 17:49:51 +0000]
Silence VC warnings about possible operator precedence problems
and signed/unsigned mismatches
p4raw-id: //depot/perl@29532
Jarkko Hietaniemi [Tue, 12 Dec 2006 15:18:53 +0000]
mg.c: #ifdef only the different bits
Message-Id: <
20061212131853.C2CBA43A67@anubis.hut.fi>
p4raw-id: //depot/perl@29531
H.Merijn Brand [Tue, 12 Dec 2006 11:36:41 +0000]
The overly-picky AIX xlc compiler doesn't like the style
from patch #29446
p4raw-id: //depot/perl@29530
Rafael Garcia-Suarez [Tue, 12 Dec 2006 11:07:27 +0000]
Skip IO::Compress::Zlib in perlivp, since that's a distribution name,
not a proper module name.
p4raw-id: //depot/perl@29529
Jan Dubois [Mon, 11 Dec 2006 13:09:09 +0000]
get ext/Win32/Win32.xs to compile on cygwin
Date: Mon, 11 Dec 2006 13:09:09 -0800
Message-ID: <isfrn2tbfleo1svmmi5kh673f4uh69am11@4ax.com>
and:
Date: Mon, 11 Dec 2006 15:08:53 -0800
Message-ID: <kvorn2h0d14jo3khmmj2a9svsp2mplrodv@4ax.com>
p4raw-id: //depot/perl@29528
Nicholas Clark [Tue, 12 Dec 2006 00:07:48 +0000]
By moving xbm_previous into the SvPVX, we save a U16 from struct xpvbm,
which will likely equate to 4 bytes with padding. (But we add 2 bytes
to the SvPVX buffer...). However, the real win is that xpvbm is now
identical to xpvmg, and all active FBM compiled strings are SvVALID()
true, so now we can eliminate xpvbm, freeing a type for 6 on 5 proxies.
p4raw-id: //depot/perl@29527
Nicholas Clark [Mon, 11 Dec 2006 23:44:30 +0000]
Fix allocation bug in change 29524.
p4raw-id: //depot/perl@29526
Nicholas Clark [Mon, 11 Dec 2006 22:52:46 +0000]
Add BmPREVIOUS_set(), and use it in the one place that BmPREVIOUS is
modified.
p4raw-id: //depot/perl@29525
Nicholas Clark [Mon, 11 Dec 2006 22:18:53 +0000]
Move PERL_FBM_TABLE_OFFSET and PERL_FBM_FLAGS_OFFSET_FROM_TABLE to sv.h
Stow BmRARE in the SvPVX, and so delete xbm_rare.
Can you see what it is yet?
p4raw-id: //depot/perl@29524
Nicholas Clark [Mon, 11 Dec 2006 21:52:12 +0000]
Fix typo in change 29518 which broke non-debugging builds.
(Here Mark, you can have the spare ; back again)
p4raw-id: //depot/perl@29523
Nicholas Clark [Mon, 11 Dec 2006 21:21:09 +0000]
Convert the -1 to a more explicit PERL_FBM_FLAGS_OFFSET_FROM_TABLE.
p4raw-id: //depot/perl@29522
Nicholas Clark [Mon, 11 Dec 2006 21:15:06 +0000]
Assert that BmRARE() BmPREVIOUS() and BmUSEFUL() are only called on
valid PVBMs.
p4raw-id: //depot/perl@29521
Nicholas Clark [Mon, 11 Dec 2006 20:55:20 +0000]
Rename FBM_TABLE_OFFSET to PERL_FBM_TABLE_OFFSET prior to moving it
to a header.
p4raw-id: //depot/perl@29520
Rafael Garcia-Suarez [Mon, 11 Dec 2006 18:33:21 +0000]
Don't load the Win32 module on non-win32 platforms
p4raw-id: //depot/perl@29519
Nicholas Clark [Mon, 11 Dec 2006 18:26:31 +0000]
Change SvTAIL() to check that both SVpbm_TAIL|SVpbm_VALID are true.
SVpbm_VALID is the same bit value is SVf_IVisUV, which means that
PVBMs can't actually ever be IOK. Therefore move BmUSEFUL() into the
IV union, and save one I32 per PVBM.
p4raw-id: //depot/perl@29518
Rafael Garcia-Suarez [Mon, 11 Dec 2006 17:09:17 +0000]
Changes and README are not needed for Switch.pm
p4raw-id: //depot/perl@29516
Rafael Garcia-Suarez [Mon, 11 Dec 2006 16:53:58 +0000]
As pointed out by Valentin Guignon, there is most
probably one \n too many in the end-of-pod marker.
p4raw-id: //depot/perl@29515
Rafael Garcia-Suarez [Mon, 11 Dec 2006 16:49:06 +0000]
Configure needs to know about the extensions IO/Compress/Base
and IO/Compress/Zlib, so they get built.
p4raw-id: //depot/perl@29514
Rafael Garcia-Suarez [Mon, 11 Dec 2006 16:16:46 +0000]
Move files from ext/Compress/IO to ext/IO/Compress
p4raw-id: //depot/perl@29513
Bo Lindbergh [Sat, 9 Dec 2006 12:17:53 +0000]
Re: [PATCH] Deparse.pm bugfix
Message-Id: <
A4BDE74B-DB3A-41C0-B2BE-FCEE0E15AB54@hagernas.com>
p4raw-id: //depot/perl@29512
Jan Dubois [Sun, 10 Dec 2006 11:47:58 +0000]
re: [PATCH] Move Win32::* functions from win32/win32.c to ext/Win32/Win32.xs
Message-ID: <9moon29qtn5c23t9fer229ealsu06bkf11@4ax.com>
p4raw-id: //depot/perl@29511
Jan Dubois [Fri, 8 Dec 2006 19:07:06 +0000]
Second patch from:
Subject: [PATCH] Move Win32::* functions from win32/win32.c to ext/Win32/Win32.xs
Message-ID: <lc9kn2tb0p5sdd4q69rbc7067r4imar59r@4ax.com>
p4raw-id: //depot/perl@29510
Jan Dubois [Fri, 8 Dec 2006 19:07:06 +0000]
First patch from:
Subject: [PATCH] Move Win32::* functions from win32/win32.c to ext/Win32/Win32.xs
Message-ID: <lc9kn2tb0p5sdd4q69rbc7067r4imar59r@4ax.com>
p4raw-id: //depot/perl@29509
Dave Mitchell [Mon, 11 Dec 2006 11:41:10 +0000]
remove extraneous debugging code introduced by #29504
p4raw-id: //depot/perl@29508
Jerrad Pierce [Sun, 10 Dec 2006 14:38:18 +0000]
Fatal v1.03
Message-Id: <
200612101938.kBAJcI9D008026@all-night-tool.mit.edu>
Allow to make fatal built-ins in CORE::GLOBAL
p4raw-id: //depot/perl@29507
Marcus Holland-Moritz [Sun, 10 Dec 2006 22:32:32 +0000]
Re: [perl #41065] Out of memory!, while extending scalar
Message-ID: <
20061210223232.0f3a5318@r2d2>
p4raw-id: //depot/perl@29506
Rafael Garcia-Suarez [Mon, 11 Dec 2006 09:41:45 +0000]
perlhack.pod nit by Jarkko
p4raw-id: //depot/perl@29505
Dave Mitchell [Mon, 11 Dec 2006 00:48:06 +0000]
fix eval qw(BEGIN{die}) style leaks (second attempt).
Repeat of change #28319 (backed out by change #28720), this time
freeing ops using the right pad. Also backs out the remempad
parser addition from change #29501; instead a new auxiliary
paser stack is added, which records the current value of
PL_comppad for every state.
p4raw-link: @29501 on //depot/perl:
2af555bf3f2b3ca8e114df3f5f680d40bd24d6bf
p4raw-link: @28720 on //depot/perl:
c86b7e916b443ee192c5638ad9d077ad4e244713
p4raw-link: @28319 on //depot/perl:
eb7d7d25d2f780edcbedc124a5bdca0d53ad8687
p4raw-id: //depot/perl@29504
Nicholas Clark [Sun, 10 Dec 2006 23:05:25 +0000]
Remove code duplication in S_to_utf8_substr() and S_to_byte_substr()
by taking advantage of how anchored_* and float_* are stored in arrays
to use a loop.
p4raw-id: //depot/perl@29503
Nicholas Clark [Sun, 10 Dec 2006 22:31:56 +0000]
Downgrading a fixed or floating substring of a pattern whilst matching
a studied string seems to get to a "should not happen" [bug #41067]
It seems that Perl_regexec_flags() assumes that if the pre-downgraded
substring is FBM compiled, then the downgraded version will be too,
hence changing the downgrade and upgrade routines to FBM compile seems
to be a correct fix.
p4raw-id: //depot/perl@29502
Dave Mitchell [Sun, 10 Dec 2006 18:50:05 +0000]
#28315 could crash when freeing ops with different pads
Add hook to parser to record current PL_comppad, then use this
when popping ops off the parser stack after parser error
p4raw-id: //depot/perl@29501
Dave Mitchell [Sun, 10 Dec 2006 17:18:00 +0000]
parser: expand yy_is_opval[] to include all value types
and rename to yy_type_tab[]. Then use this table to improve stack
dumping with -Dpv
p4raw-id: //depot/perl@29500
Graham Barr [Sun, 10 Dec 2006 16:28:34 +0000]
Restore change 27486 accidently removed (also added to 1.19 before release)
p4raw-id: //depot/perl@29499
Graham Barr [Sun, 10 Dec 2006 16:17:14 +0000]
Upgrade to Scalar-List-Utils-1.19
p4raw-id: //depot/perl@29498
Ilya Zakharevich [Sat, 9 Dec 2006 16:33:41 +0000]
Re: [PATCH 5.8.8] Build on OS/2 - missing part
Message-ID: <
20061210003341.GA14200@powdermilk.math.berkeley.edu>
p4raw-id: //depot/perl@29497
Nicholas Clark [Sun, 10 Dec 2006 01:13:09 +0000]
Swap SVpbm_TAIL and SVpbm_VALID.
p4raw-id: //depot/perl@29496
Steve Peters [Sat, 9 Dec 2006 22:44:27 +0000]
Upgrade to Time-HiRes-1.9703
p4raw-id: //depot/perl@29495
Nicholas Clark [Sat, 9 Dec 2006 21:19:45 +0000]
PVBMs mustn't enter anything that turns SVIVisUV_on().
p4raw-id: //depot/perl@29494
Nicholas Clark [Sat, 9 Dec 2006 21:17:37 +0000]
Exterminate! 2 pieces of superfluous code related to UV setting.
p4raw-id: //depot/perl@29493
Nicholas Clark [Fri, 8 Dec 2006 23:49:51 +0000]
In Perl_fbm_instr(), no need to calculate the address of the table if
we're going to return anyway.
p4raw-id: //depot/perl@29492
Nicholas Clark [Fri, 8 Dec 2006 23:22:31 +0000]
assert that SvTAIL() isn't being called on PVAVs or PVHVs
p4raw-id: //depot/perl@29491
Nicholas Clark [Fri, 8 Dec 2006 22:56:54 +0000]
Wrap some macro arguments in () to avoid precedence problems.
p4raw-id: //depot/perl@29490
Nicholas Clark [Fri, 8 Dec 2006 22:56:08 +0000]
Perl_screaminstr() must be called with littlestr as a PVBM.
p4raw-id: //depot/perl@29489
Dave Mitchell [Fri, 8 Dec 2006 22:20:41 +0000]
add missing OP_REFCNT_LOCK/UNLOCKs and document it
p4raw-id: //depot/perl@29488
Nicholas Clark [Fri, 8 Dec 2006 22:11:56 +0000]
Assert that BmRARE, BmUSEFUL and BmPREVIOUS are only called on SVtPVBM
p4raw-id: //depot/perl@29487
Nicholas Clark [Fri, 8 Dec 2006 21:09:57 +0000]
No need to upgrade if all we're going to do is return.
p4raw-id: //depot/perl@29486
Steve Hay [Thu, 7 Dec 2006 12:34:10 +0000]
Make use of $Config{d_pseudofork} in a couple of core tests
Can't really use it in ext/IO/t/*.t and lib/Test/Simple/t/fork.t
because they are dual-lived modules and must work with older perls
that don't have d_pseudofork
p4raw-id: //depot/perl@29485
Jan Dubois [Wed, 6 Dec 2006 16:31:42 +0000]
Move Win32 from win32/ext/Win32 to ext/Win32
Message-ID: <1unen2lsrq4pnnt06933h96pqj1d8luh3b@4ax.com>
p4raw-id: //depot/perl@29483
Jerry D. Hedden [Wed, 6 Dec 2006 11:12:13 +0000]
Perl_save_destructor is not a mathom
From: "Jerry D. Hedden" <jdhedden@yahoo.com>
Message-ID: <
20061206191213.35036.qmail@web30206.mail.mud.yahoo.com>
p4raw-id: //depot/perl@29482
Jan Dubois [Wed, 6 Dec 2006 16:26:38 +0000]
Makefile.PL changes to compiler Win32.xs using cygwin
Message-ID: <5nnen2p2h4lhremj0qf1tmpkl0asvk87hr@4ax.com>
p4raw-id: //depot/perl@29481
Steve Hay [Wed, 6 Dec 2006 17:28:03 +0000]
Better fix than #29479 for exporting refcounted_he_chain_2hv
p4raw-id: //depot/perl@29480
Yves Orton [Wed, 6 Dec 2006 17:52:08 +0000]
Re: [PATCH] User pragmas now accessible from B
Message-ID: <
9b18b3110612060852s7cfbba66id5aae25e5bb20ae0@mail.gmail.com>
p4raw-id: //depot/perl@29479
Steve Hay [Wed, 6 Dec 2006 14:19:55 +0000]
Upgrade to Time-HiRes 1.9702
p4raw-id: //depot/perl@29478
Jarkko Hietaniemi [Tue, 5 Dec 2006 16:08:33 +0000]
Re: When should PERL_SYS_TERM() be called? [was: Re: [PATCH] Re: [PATCH] Re: [PATCH] abstract mempool header testing]
Message-ID: <
4575DFD1.1060108@iki.fi>
Fixes a problem spotted by Jan Dubois:
The PerlMemShared pool is only shared between the interpreters that
can share data structures (interpreters created by perl_clone(), which
mean pseudo-fork, and threads.xs). The pool is not shared between
interpreters that are created separately by perl_alloc().
[...]
I guess this means PL_perlio_fd_refcnt needs to be allocated and
freed by standard malloc() and not go through any abstraction.
p4raw-id: //depot/perl@29477
Rafael Garcia-Suarez [Wed, 6 Dec 2006 13:26:18 +0000]
save_I8 is not a mathom.
p4raw-id: //depot/perl@29476
Joshua ben Jore [Tue, 5 Dec 2006 13:18:21 +0000]
User pragmas now accessible from B
From: "Joshua ben Jore" <twists@gmail.com>
Message-ID: <
dc5c751d0612051318n2e2f08bfh185fab6323a980b@mail.gmail.com>
p4raw-id: //depot/perl@29475
Craig A. Berry [Wed, 6 Dec 2006 05:29:50 +0000]
Avoid my_regdupe type mismatch warning after #29430
p4raw-id: //depot/perl@29474
Craig A. Berry [Wed, 6 Dec 2006 04:48:35 +0000]
d_pseudofork for configure.com
p4raw-id: //depot/perl@29473
Yves Orton [Tue, 5 Dec 2006 22:53:48 +0000]
Re: [PATCH] Further tweaks to make it easier to create regexp engine plug ins.
Message-ID: <
9b18b3110612051353k7745792byabfdf98e18eb48c9@mail.gmail.com>
Fixes a minor bug in change 29468.
p4raw-id: //depot/perl@29472
Steve Peters [Tue, 5 Dec 2006 21:00:38 +0000]
Silence a warning from Module::CoreList that occurs when the module version is
something like X.XX_XX.
p4raw-id: //depot/perl@29471
Steve Hay [Tue, 5 Dec 2006 16:41:55 +0000]
Make the check for PERL_IMPLICIT_SYS better
p4raw-id: //depot/perl@29470
Steve Hay [Tue, 5 Dec 2006 16:06:31 +0000]
Extend d_pseudofork to Win32-land and set it when appropriate
(namely, when we have USE_ITHREADS and PERL_IMPLICIT_SYSTEM)
p4raw-id: //depot/perl@29469
Yves Orton [Tue, 5 Dec 2006 16:13:36 +0000]
Further tweaks to make it easier to create regexp engine plug ins.
Message-ID: <
9b18b3110612050713g77cac516x46fb5baac99b47c9@mail.gmail.com>
(with tweaks)
p4raw-id: //depot/perl@29468
H.Merijn Brand [Tue, 5 Dec 2006 14:55:14 +0000]
Introduction of d_pseudofork
p4raw-id: //depot/perl@29467
Jarkko Hietaniemi [Mon, 4 Dec 2006 22:53:03 +0000]
Re: When should PERL_SYS_TERM() be called? [was: Re: [PATCH] Re: [PATCH] Re: [PATCH] abstract mempool header testing]
Message-ID: <
4574ED1F.40508@iki.fi>
Re-instates #29424 (previously reverted by #29451), now fixed to work
with PERL_IMPLICIT_SYS, thanks to Jan Dubois. Also adds PERLIO_TERM to
the Symbian port.
p4raw-id: //depot/perl@29465
Steve Hay [Tue, 5 Dec 2006 09:38:14 +0000]
Skip Perl_regdupe_internal rather than Perl_regdupe
when building on Win32 without ithreads (see #29430)
p4raw-id: //depot/perl@29464
Nicholas Clark [Mon, 4 Dec 2006 22:50:09 +0000]
Move Perl_get_db_sub() from pp_hot.c to util.c
p4raw-id: //depot/perl@29463
H.Merijn Brand [Mon, 4 Dec 2006 18:09:20 +0000]
Revert #29454
p4raw-id: //depot/perl@29462
Dave Mitchell [Mon, 4 Dec 2006 18:08:31 +0000]
fix some casting issues with perly.y / madly.y merger
p4raw-id: //depot/perl@29461
Rafael Garcia-Suarez [Mon, 4 Dec 2006 17:50:51 +0000]
Better version of last patch, by Yves Orton.
Actually the regexp engine structure only needs
one compilation function hook.
p4raw-id: //depot/perl@29459
Rafael Garcia-Suarez [Mon, 4 Dec 2006 17:22:02 +0000]
The new regexp compilation function must be added to the engine structure.
p4raw-id: //depot/perl@29458
Rafael Garcia-Suarez [Mon, 4 Dec 2006 16:24:09 +0000]
Actually submit previous change.
p4raw-id: //depot/perl@29457
Rafael Garcia-Suarez [Mon, 4 Dec 2006 16:15:54 +0000]
Split Perl_pregcomp in two functions. Patch by Audrey Tang,
to facilitate pluging in other regexp engines.
p4raw-id: //depot/perl@29456
Dave Mitchell [Mon, 4 Dec 2006 15:38:05 +0000]
merge perly.y and madlu.y
p4raw-id: //depot/perl@29455
Rafael Garcia-Suarez [Mon, 4 Dec 2006 15:48:12 +0000]
Re: install misses Compress/IO/{Base,Zlib}
From: "Rafael Garcia-Suarez" <rgarciasuarez@gmail.com>
Message-ID: <
b77c1dce0612040648j411247c5i27a305c86cbac401@mail.gmail.com>
p4raw-id: //depot/perl@29454
Jos I. Boumans [Fri, 1 Dec 2006 12:24:08 +0000]
Add File::Fetch to the core
From: "Jos Boumans" <kane@dwim.org>
Message-ID: <22195.80.127.35.68.
1164972248.squirrel@webmail.xs4all.nl>
p4raw-id: //depot/perl@29452
Steve Hay [Mon, 4 Dec 2006 14:31:08 +0000]
Change #29424 may well be correct in itself, but currently doesn't
work when perl is built with PERL_IMPLICIT_SYS because PERLIO_TERM
uses PerlMemShared_free() which, in that case, involves the perlhost
that has already been freed itself by perl_free(), which is called
before PERL_SYS_TERM().
The order of perl_destruct(), perl_free(), PERL_SYS_TERM() cannot
be changed (it is advertised in perlembed for a start), so just
revert #29424 and the appropriate parts of #29442 for now.
Perhaps PL_perlio_fd_refcnt needs allocating differently, or else
the perlhost needs freeing later (after PerlIO)?
p4raw-id: //depot/perl@29451