John P. Linderman [Wed, 9 Apr 2008 08:32:31 +0000]
changes to perlsec.pod and call for removal of quicksort
From: "John P. Linderman" <jpl@research.att.com>
Message-Id: <
200804091232.m39CWVkY33722822@raptor.research.att.com>
p4raw-id: //depot/perl@33664
Jim Cromie [Tue, 18 Mar 2008 11:14:44 +0000]
Re: TODOs (was Re: summer of code mentor applications starting (and ending) next week)
Message-ID: <
47DFF884.80802@gmail.com>
[all except the "squeeze the optree" suggestion]
p4raw-id: //depot/perl@33663
Rafael Garcia-Suarez [Tue, 8 Apr 2008 15:31:46 +0000]
Remove old option leftover in docs
p4raw-id: //depot/perl@33661
Nicholas Clark [Mon, 7 Apr 2008 17:50:31 +0000]
So why didn't the build break for me? Fix problems spotted by Jerry
Hedden.
p4raw-id: //depot/perl@33659
Nicholas Clark [Mon, 7 Apr 2008 17:36:41 +0000]
Tidy up compiler warning spotted by Jerry Hedden.
p4raw-id: //depot/perl@33658
Nicholas Clark [Mon, 7 Apr 2008 14:45:33 +0000]
Split out S_refcounted_he_new_common() from
Perl_refcounted_he_new_common(), so that Perl_store_cop_label() can
call it without needing to create two temporary SVs. Use it in
newSTATEOP() and eliminate the two temporary SVs. Make
Perl_fetch_cop_label() more defensive by not assuming that the value
for ":" is always a PV. Remove its "compatibility" macro.
p4raw-id: //depot/perl@33657
Nicholas Clark [Mon, 7 Apr 2008 11:29:51 +0000]
Eliminate cop_label from struct cop by storing a label as the first
entry in the hints hash. Most statements don't have labels, so this
will save memory. Not sure how much.
p4raw-id: //depot/perl@33656
Nicholas Clark [Sun, 6 Apr 2008 21:53:57 +0000]
Change 33653 (inevitably) missed one cop_label, because I mistook it
for part of some other context-stack related struct.
p4raw-id: //depot/perl@33655
Nicholas Clark [Sun, 6 Apr 2008 20:27:34 +0000]
A macro used only once isn't a typing saving, and 3 macro definitions
never used are more obfuscation than clarification, so inline the only
use of SAVECOPLABEL_FREE(), and remove the unthreaded variant and
both SAVECOPLABEL()s. Exterminate! Exterminate! Exterminate!
p4raw-id: //depot/perl@33654
Nicholas Clark [Sun, 6 Apr 2008 20:11:52 +0000]
Replace all remaining accesses to COP's cop_label with CopLABEL().
p4raw-id: //depot/perl@33653
Nicholas Clark [Sun, 6 Apr 2008 19:32:03 +0000]
Add attempting Exterminate! cop_label to perltodo.
p4raw-id: //depot/perl@33652
Paul Fenwick [Mon, 31 Mar 2008 22:53:04 +0000]
Re: [PATCH][NUDGE] Documentation fix for perlfunc/shutdown
Message-ID: <
47F0D0A0.8010801@perltraining.com.au>
p4raw-id: //depot/perl@33651
Paul Fenwick [Sun, 6 Apr 2008 11:28:27 +0000]
perlfunc.pod: atan2(0,0) returns 0, not undef
Message-ID: <
47F8273B.6010504@perltraining.com.au>
p4raw-id: //depot/perl@33650
Nicholas Clark [Sun, 6 Apr 2008 09:29:24 +0000]
Missed updated a test description, as spotted by vincent.
p4raw-id: //depot/perl@33649
Nicholas Clark [Sat, 5 Apr 2008 13:03:18 +0000]
UCD 5.1.0
p4raw-id: //depot/perl@33648
Jarkko Hietaniemi [Sat, 5 Apr 2008 04:16:17 +0000]
Tru64: align fp policies of gcc and cc
Message-Id: <
200804050116.m351GH5P042062@kosh.hut.fi>
p4raw-id: //depot/perl@33647
Steve Hay [Fri, 4 Apr 2008 08:28:11 +0000]
Run overload.pl from regen.pl, and check overload.c as well as overload.h
p4raw-id: //depot/perl@33646
Jan Dubois [Thu, 3 Apr 2008 18:50:08 +0000]
Make PL_AMG_names and PL_AMG_namelens static
From: "Jan Dubois" <jand@activestate.com>
Message-ID: <
02c401c895f6$369e8130$a3db8390$@com>
p4raw-id: //depot/perl@33645
Jan Dubois [Thu, 3 Apr 2008 17:34:13 +0000]
PL_bincompat_opt should be exported on AIX and Windows
From: "Jan Dubois" <jand@activestate.com>
Message-ID: <
02c001c895eb$9bc3e920$d34bbb60$@com>
(with one tweak--it should be PL_bincompat_options!)
p4raw-id: //depot/perl@33644
Nicholas Clark [Thu, 3 Apr 2008 17:14:51 +0000]
Pack lib/CGI/t/upload_post_text.txt with uupacktool.pl. (It contains a
GIF, which clearly isn't text)
p4raw-id: //depot/perl@33643
Nicholas Clark [Thu, 3 Apr 2008 15:41:26 +0000]
Clean up some compiler warnings spotted by smokers.
p4raw-id: //depot/perl@33639
Nicholas Clark [Wed, 2 Apr 2008 21:03:33 +0000]
In Perl_sv_grow(), no need to do PERL_STRLEN_ROUNDUP() if we have
malloc_size() to get the true allocated space, as PERL_STRLEN_ROUNDUP()
might actually bump the request across an allocation size boundary.
p4raw-id: //depot/perl@33632
Jerry D. Hedden [Wed, 2 Apr 2008 13:45:26 +0000]
Re: Change 33598: [PATCH] Re: Smoke [5.11.0] 33456 PASS darwin 9.2.0 (macppcG5/1 cpu)
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510804021045s1f7ec22ftbf5b9e2f2c6c619e@mail.gmail.com>
p4raw-id: //depot/perl@33631
Nicholas Clark [Wed, 2 Apr 2008 18:23:58 +0000]
Fix bug in change 33379 - if we're using malloc_size() as well as being
careful to work with PERL_TRACK_MEMPOOL, we also have to be careful to
work *without* it.
p4raw-id: //depot/perl@33629
Nicholas Clark [Tue, 1 Apr 2008 19:59:54 +0000]
Define sv_insert() as a wrapper to sv_insert_flags(), and move
Perl_sv_insert() to mathoms.c
p4raw-id: //depot/perl@33627
Nicholas Clark [Tue, 1 Apr 2008 18:47:01 +0000]
Add the correct argument assertion macro to sv_insert_flags().
Don't want anything wrongly marked as non-NULL slipping through.
p4raw-id: //depot/perl@33626
Steve Hay [Tue, 1 Apr 2008 16:42:04 +0000]
Upgrade to Win32-0.35
p4raw-id: //depot/perl@33625
Rafael Garcia-Suarez [Tue, 1 Apr 2008 09:38:03 +0000]
Module::CoreList wasn't listing Config.pm everywhere.
p4raw-id: //depot/perl@33624
Nicholas Clark [Mon, 31 Mar 2008 23:08:45 +0000]
$VERSION++ for all the non-dual life modules outside ext/ that
Porting/cmpVERSION.pl reports differ from the 5.10.0 release.
All the rest of the differences would seem to be dual-life. :-(
p4raw-id: //depot/perl@33623
Nicholas Clark [Mon, 31 Mar 2008 22:33:36 +0000]
Fix bug in the code for --opened - it was not working on blead.
p4raw-id: //depot/perl@33622
Nicholas Clark [Mon, 31 Mar 2008 22:25:19 +0000]
$VERSION++ for all the non-dual life modules in ext/ that
Porting/cmpVERSION.pl reports differ from the 5.10.0 release.
p4raw-id: //depot/perl@33621
Vincent Pit [Mon, 31 Mar 2008 19:05:44 +0000]
Double magic with substr
Message-ID: <
47F119E8.5010106@profvince.com>
p4raw-id: //depot/perl@33618
Rafael Garcia-Suarez [Mon, 31 Mar 2008 15:21:42 +0000]
Config.pm needs to be listed in Module::CoreList, too
p4raw-id: //depot/perl@33612
Nicholas Clark [Sun, 30 Mar 2008 23:19:29 +0000]
Give DynaLoader a non-dev version number.
p4raw-id: //depot/perl@33609
Robin Barker [Wed, 19 Mar 2008 10:55:59 +0000]
add -v to regen.pl and friends
From: "Robin Barker" <Robin.Barker@npl.co.uk>
Message-ID: <
46A0F33545E63740BC7563DE59CA9C6D093AA6@exchsvr2.npl.ad.local>
p4raw-id: //depot/perl@33608
Nicholas Clark [Mon, 24 Mar 2008 16:35:09 +0000]
detecting Configure test goofs (Was Re: branch prediction hints)
Message-ID: <
20080324163509.GW79799@plum.flirble.org>
p4raw-id: //depot/perl@33606
Nicholas Clark [Sat, 29 Mar 2008 16:15:52 +0000]
GDBM_File isn't dual life, so no need for a dev module version number.
p4raw-id: //depot/perl@33601
Dominic Dunlop [Tue, 11 Mar 2008 21:54:04 +0000]
Re: Smoke [5.11.0] 33456 PASS darwin 9.2.0 (macppcG5/1 cpu)
Message-Id: <
3B7752C8-D5A2-452C-B3E0-C453FFCBCAFA@mac.com>
[but rename ******* to i_mallocmalloc.U, and then fix up all the files
that Porting/checkcfgvar.pl says need i_mallocmalloc declared]
p4raw-id: //depot/perl@33598
Nicholas Clark [Sat, 29 Mar 2008 11:08:07 +0000]
Add i_syspoll to all the pre-canned configs.
p4raw-id: //depot/perl@33597
Nicholas Clark [Sat, 29 Mar 2008 07:08:50 +0000]
Fix compiler warning about comparison of pointer types.
p4raw-id: //depot/perl@33596
Gerard Goossen [Wed, 26 Mar 2008 14:07:13 +0000]
do not use SVTYPEMASK to prevent cleaning of PL_fdpid and PL_strtab
Message-ID: <
20080326130713.GL4409@ostwald>
Date: Wed, 26 Mar 2008 14:07:13 +0100
p4raw-id: //depot/perl@33595
Dominic Dunlop [Fri, 14 Mar 2008 14:45:39 +0000]
Re: Tests failed on PPC64
Message-Id: <
53E6407E-B87C-4D6A-A6E7-D842BDF30292@mac.com>
Date: Fri, 14 Mar 2008 14:45:39 +0100
p4raw-id: //depot/perl@33594
Steve Peters [Fri, 28 Mar 2008 19:16:53 +0000]
Additional CGI.pm test files that got missed at some point.
p4raw-id: //depot/perl@33592
Craig A. Berry [Fri, 28 Mar 2008 19:06:35 +0000]
Make the nv_preserves... checking on VMS more robust (and more like
what Configure does). Problem using long int without long double
reported by Jeremy Begg.
p4raw-id: //depot/perl@33591
Craig A. Berry [Fri, 28 Mar 2008 19:01:13 +0000]
Make uudmap.h dependencies more explicit in the VMS build
(needed by MMS but not MMK).
p4raw-id: //depot/perl@33589
Gerard Goossen [Thu, 27 Mar 2008 13:55:31 +0000]
MAD dump xml escape regex
Message-ID: <
20080327125531.GN4409@ostwald>
p4raw-id: //depot/perl@33584
Paul Marquess [Thu, 27 Mar 2008 11:12:22 +0000]
DB_File 1.817
From: "Paul Marquess" <Paul.Marquess@ntlworld.com>
Message-ID: <
051d01c88ffb$6f662eb0$6901a8c0@myopwv.com>
p4raw-id: //depot/perl@33583
Nicholas Clark [Wed, 26 Mar 2008 21:05:20 +0000]
The offset for pos is stored as bytes, and converted to (Unicode)
character position when read, if needed. The code for setting pos
inside subst was incorrectly converting to character position before
storing the value. This code appears to have been buggy since it was
added in 2000 in change 7562.
p4raw-id: //depot/perl@33580
Nicholas Clark [Wed, 26 Mar 2008 18:59:03 +0000]
Don't call S_utf8_mg_pos_cache_update(), and hence don't even create
the magic for the UTF-8 cache, if the UTF-8 caching is disabled.
p4raw-id: //depot/perl@33579
Nicholas Clark [Wed, 26 Mar 2008 18:38:10 +0000]
Update makemeta to add the 3 official "resources" - homepage, license
and bugtracker.
p4raw-id: //depot/perl@33578
Nicholas Clark [Wed, 26 Mar 2008 18:37:15 +0000]
Re-run Porting/makemeta
p4raw-id: //depot/perl@33577
Nicholas Clark [Wed, 26 Mar 2008 16:50:06 +0000]
With change 33252, MM_Unix.pm isn't quite in sync with upstream any
more. Of the changes references in 33252, 6.44 accepted and improved
change 32879 (grep in void context), so only changes 32905 and 32964
are outstanding (ABSPERL with whitespace).
p4raw-id: //depot/perl@33575
Nicholas Clark [Wed, 26 Mar 2008 16:20:11 +0000]
Fix whitespace difference between blead and upstream in MM_Unix.pm
p4raw-id: //depot/perl@33573
Nicholas Clark [Wed, 26 Mar 2008 11:45:12 +0000]
Actually use the computed @path so as to write the awkwardly named test
files to the right place when in core. Without this the open fails
(silently as intended) and the tests are skipped on all OS.
p4raw-id: //depot/perl@33571
Leon Brocard [Wed, 26 Mar 2008 10:53:43 +0000]
Re: [PATCH] perlhack.pod update walkthrough
From: "Leon Brocard" <acme@astray.com>
Message-ID: <
a92222c80803260353k6cae53eieea909aed0a967b5@mail.gmail.com>
p4raw-id: //depot/perl@33570
chromatic [Wed, 26 Mar 2008 00:12:09 +0000]
Typo Fixes
Message-Id: <
200803260012.09967.chromatic@wgz.org>
p4raw-id: //depot/perl@33569
Steve Peters [Wed, 26 Mar 2008 03:18:01 +0000]
Upgrade to CGI-3.35.
p4raw-id: //depot/perl@33568
Nicholas Clark [Tue, 25 Mar 2008 16:55:03 +0000]
Patches for VMS by Craig Berry from
http://rt.cpan.org/Public/Bug/Display.html?id=34095
(1.50 has 3 failures on VMS)
p4raw-id: //depot/perl@33567
Steve Hay [Mon, 17 Mar 2008 14:36:54 +0000]
RE: [PATCH revised] Fix ExtUtils::Install under Cygwin
From: "Steve Hay" <SteveHay@planit.com>
Message-ID: <
1B32FF956ABF414C9BCE5E487A1497E70176BD61@ukmail02.planit.group>
"OK, so how about the attached. This fixes up -w for all compilers so
that it is symmetrical with chmod(), and adds a note to perltodo on
fixing POSIX::access() and chdir()."
The whole long thread started here:
http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2008-03/msg00056.html
p4raw-id: //depot/perl@33566
Jerry D. Hedden [Tue, 25 Mar 2008 11:51:00 +0000]
Re: Change 33556: [PATCH] borg parent.pm
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510803250851w52ea0c84n9876834d8e8b79e3@mail.gmail.com>
p4raw-id: //depot/perl@33565
Steve Peters [Tue, 25 Mar 2008 15:27:06 +0000]
Upgrade to CGI.pm-3.34. There are still a few differences, so adding
a version bump.
p4raw-id: //depot/perl@33564
Nicholas Clark [Tue, 25 Mar 2008 11:36:22 +0000]
Repeat change 33508 - change 33561 re-instated the mistake in change
33507.
p4raw-id: //depot/perl@33563
Jarkko Hietaniemi [Wed, 12 Mar 2008 21:43:53 +0000]
"646"
Message-ID: <
47D886D9.6060001@iki.fi>
Encode::Alias fix for encoding "646" (aka ASCII) on Solaris
p4raw-id: //depot/perl@33562
Steven Schubiger [Thu, 13 Mar 2008 17:58:00 +0000]
Re: [PATCH] sv.c: consting
Message-ID: <
20080313165800.GD31102@refcnt.homeunix.org>
p4raw-id: //depot/perl@33561
Rafael Garcia-Suarez [Tue, 25 Mar 2008 09:52:27 +0000]
Use sv_setpvs() like a few lines before since change #33557
p4raw-link: @33557 on //depot/perl:
502d9230ec570254fed51ae721c1da50944a5cbf
p4raw-id: //depot/perl@33560
pancho@pancho.name [Fri, 21 Mar 2008 03:31:04 +0000]
[perl #51964] Typo on regular expression at perlopentut manual page.
From: pancho@pancho.name (via RT) <perlbug-followup@perl.org>
Message-ID: <rt-3.6.HEAD-25460-
1206095463-311.51964-75-0@perl.org>
p4raw-id: //depot/perl@33559
David Nicol [Fri, 21 Mar 2008 12:56:12 +0000]
Re: local $@ has an unwanted side effect
From: "David Nicol" <davidnicol@gmail.com>
Message-ID: <
934f64a20803211056q5148027ava77af36f51c96418@mail.gmail.com>
(with Tim Bunce's amendments)
p4raw-id: //depot/perl@33558
Vincent Pit [Sat, 22 Mar 2008 13:37:42 +0000]
Re: [PATCH] Double warning with perl -we 'my $a; substr $a, 0, 10,
Message-ID: <
47E4FD96.6080304@profvince.com>
p4raw-id: //depot/perl@33557
Yitzchak Scott-Thoennes [Wed, 5 Mar 2008 17:19:32 +0000]
borg parent.pm
From: "Yitzchak Scott-Thoennes" <sthoenna@efn.org>
Message-ID: <57512.71.32.86.11.
1204766372.squirrel@webmail.efn.org>
Plus bump base.pm's version to a non-alpha number
p4raw-id: //depot/perl@33556
Niko Tyni [Mon, 24 Mar 2008 23:32:24 +0000]
IO::Socket::INET unnecessarily resolves "udp"
Message-Id: <
1206394344-30835-1-git-send-email-ntyni@debian.org>
p4raw-id: //depot/perl@33554
Vincent Pit [Mon, 24 Mar 2008 21:30:10 +0000]
apidoc mismatch for Perl_magic_clearhint
Message-ID: <
47E80F52.4030805@profvince.com>
p4raw-id: //depot/perl@33553
Steve Peters [Mon, 24 Mar 2008 17:55:29 +0000]
Upgrade to Module-Pluggable-3.8
p4raw-id: //depot/perl@33552
Nicholas Clark [Mon, 24 Mar 2008 16:30:26 +0000]
Give the test file for __builtin_expect() the correct name (try.c, not
builtin.c). Correct the user visible text, which was clearly a cut &
paste from __builtin_choose_expr. For both, also run the program as
part of the checks. Add a return value to the test program, and make
it return something useful (that the if statement worked).
p4raw-id: //depot/perl@33551
John E. Malmberg [Sun, 23 Mar 2008 14:29:25 +0000]
configure.com and vms.c fixes.
From: "John E. Malmberg" <wb8tyw@qsl.net>
Message-id: <
47E6AF95.1000704@qsl.net>
p4raw-id: //depot/perl@33550
H.Merijn Brand [Fri, 21 Mar 2008 10:28:42 +0000]
Final move from meta-3.0 to meta-3.5
p4raw-id: //depot/perl@33547
Vincent Pit [Thu, 20 Mar 2008 12:06:47 +0000]
Double warning with perl -we '\&$x'
From: "Vincent Pit" <perl@profvince.com>
Message-ID: <38561.147.210.17.175.
1206011207.squirrel@147.210.17.175>
p4raw-id: //depot/perl@33544
Nicholas Clark [Thu, 13 Mar 2008 12:47:51 +0000]
Re: Smoke [5.11.0] 33500 FAIL(F) MSWin32 WinXP/.Net SP2 (x86/2 cpu)
Message-ID: <
20080313124751.GX79799@plum.flirble.org>
Fixes test failures introduced by #33480 in non-fork()-enabled builds.
p4raw-id: //depot/perl@33543
Sérgio Durigan Júnior [Tue, 18 Mar 2008 15:35:54 +0000]
Fixing "make distclean" to remove "config.arch"
Message-Id: <
1205865354.16048.56.camel@miki>
p4raw-id: //depot/perl@33542
Steve Hay [Tue, 18 Mar 2008 08:41:13 +0000]
Silence some warnings introduced by #33507
p4raw-id: //depot/perl@33541
Nicholas Clark [Mon, 17 Mar 2008 00:17:26 +0000]
Drag autodoc.pl and overload.pl into the age of safer_open().
Thanks to the wisdom of london.pm, stuff the filename into the SCALAR
slot of the typeglob created in safer_open(), so that ...
Add safer_close(), that will die (with the filename) if the close
fails.
p4raw-id: //depot/perl@33539
Nicholas Clark [Sat, 15 Mar 2008 18:37:34 +0000]
Rename safer_rename() to rename_if_different(), to accurately describe
what it does. Use File::Compare rather than Digest::MD5, as the files
are small enough to simply read in. (File::Compare dates from 5.004)
Remove safer_rename_always(), which isn't used.
DRY by replacing the cargo-culted "open or die" with a new function
safer_open(), which uses Gensym (5.002) to create an anonymous file
handle, and opens and binmodes the file, or dies.
This necessitates replacing bareword file handles with lexicals in all
the callers.
Correct the names of files in close or die constructions.
p4raw-id: //depot/perl@33538
Jim Cromie [Tue, 11 Mar 2008 18:16:14 +0000]
Re: [patch] refine make regen to be more selective
Message-ID: <
47D720CE.7060004@gmail.com>
Date: Tue, 11 Mar 2008 18:16:14 -0600
p4raw-id: //depot/perl@33537
Nicholas Clark [Fri, 14 Mar 2008 17:33:17 +0000]
++$Pod::Html::VERSION
p4raw-id: //depot/perl@33534
Nicholas Clark [Fri, 14 Mar 2008 15:13:43 +0000]
Update the use of single quotes to be consistent with the advice in
http://www.cl.cam.ac.uk/~mgk25/ucs/quotes.html
p4raw-id: //depot/perl@33531
Nicholas Clark [Fri, 14 Mar 2008 00:12:20 +0000]
Fix a couple of minor differences between the Encode 2.14 CPAN tarball
and blead.
p4raw-id: //depot/perl@33523
Nicholas Clark [Thu, 13 Mar 2008 16:03:56 +0000]
Change 33507 had a const too far (Perl_mg_dup()'s mg is assigned to).
p4raw-id: //depot/perl@33508
Steven Schubiger [Wed, 12 Mar 2008 12:38:46 +0000]
Re: [PATCH] sv.c: consting
Message-ID: <
20080312113846.GB31102@refcnt.homeunix.org>
p4raw-id: //depot/perl@33507
Robert May [Thu, 13 Mar 2008 19:36:33 +0000]
[perl #51674] op/alarm.t hangs on 5.11.0 (Windows Vista only)
From: "Robert May" <rob@themayfamily.me.uk>
Message-ID: <
54bdc7510803130706q6523793ak1f3f1c6578fa97c1@mail.gmail.com>
p4raw-id: //depot/perl@33506
Nicholas Clark [Thu, 13 Mar 2008 13:02:05 +0000]
Upgrade to Module::Pluggable 3.7
p4raw-id: //depot/perl@33505
Nicholas Clark [Thu, 13 Mar 2008 13:01:44 +0000]
Correct two spelling mistakes, and reformat some comments.
p4raw-id: //depot/perl@33504
Rafael Garcia-Suarez [Thu, 13 Mar 2008 10:43:13 +0000]
Regexp::DESTROY was only added in 5.8.1
p4raw-id: //depot/perl@33503
Nicholas Clark [Wed, 12 Mar 2008 19:40:01 +0000]
Oops. Change 33499 forgot to add the new Makefile.PL to the MANIFEST.
p4raw-id: //depot/perl@33500
Nicholas Clark [Wed, 12 Mar 2008 19:32:09 +0000]
Move Module::Pluggable into ext/ as the next version has actions in its
Makefile.PL that we will need to run as part of a core build.
p4raw-id: //depot/perl@33499
Nicholas Clark [Wed, 12 Mar 2008 19:09:21 +0000]
Change 33492 did not spread the protection wide enough. There were
still two more races to be lost.
1: The close() could still happen after the (premature) mutex release
allowed another thread to dup() to that file descriptor.
2: The initial dup() could happen whilst another thread was in the
mutex protected region, and had temporarily closed the file
descriptor.
Race conditions remain with any other thread that actually does I/O
during the execution of the mutex protected region (as noted in a
comment), and dup() failure is not handled gracefully (also noted).
p4raw-id: //depot/perl@33498
blino@mandriva.com [Wed, 12 Mar 2008 02:59:45 +0000]
[perl #51636] segmentation fault with array ties
From: blino@mandriva.com (via RT) <perlbug-followup@perl.org>
Message-ID: <rt-3.6.HEAD-25460-
1205315984-377.51636-75-0@perl.org>
p4raw-id: //depot/perl@33495
Steve Peters [Wed, 12 Mar 2008 14:20:49 +0000]
Upgrade to Encode-2.24
p4raw-id: //depot/perl@33493
Nicholas Clark [Wed, 12 Mar 2008 12:23:55 +0000]
We need mutex protection in PerlIOStdio_close() for the duration of
holding our true love file handle open, to stop anything else
temporarily using it for a quick dup() fling, and then closing the
file handle underneath us.
I suspect that the lack of this protection was the cause of the threads
free.t and blocks.t failures on OS X on 5.8.x, where usefaststdio is
the default, and PerlIO is unable to "invalidate" the FILE *.
p4raw-id: //depot/perl@33492
Nicholas Clark [Wed, 12 Mar 2008 11:46:17 +0000]
Correct logic error in PerlIOStdio_close() - 0 is an acceptable value
from dup(), so it can't also be the "don't do anything later" value.
p4raw-id: //depot/perl@33491
Steven Schubiger [Tue, 11 Mar 2008 19:59:37 +0000]
Re: [PATCH] sv.c: consting
Message-ID: <
20080311185937.GA18713@refcnt.homeunix.org>
p4raw-id: //depot/perl@33489
Paul Fenwick [Wed, 12 Mar 2008 13:19:15 +0000]
Correct misleading example in perlsyn.pod (given/when/default)
Message-ID: <
47D73DA3.8050300@perltraining.com.au>
p4raw-id: //depot/perl@33488