Jarkko Hietaniemi [Sun, 2 Feb 2003 15:09:39 +0000 (15:09 +0000)]
Plan 9: Supply also Russ Cox' original config.h for 5.8.0.
p4raw-id: //depot/perl@18629
Jarkko Hietaniemi [Sun, 2 Feb 2003 14:46:26 +0000 (14:46 +0000)]
Plan 9: Update config.plan9 based on Russ Cox' work.
p4raw-id: //depot/perl@18628
Jarkko Hietaniemi [Sun, 2 Feb 2003 14:24:00 +0000 (14:24 +0000)]
Plan 9: Update the list of object files.
p4raw-id: //depot/perl@18627
Jarkko Hietaniemi [Sun, 2 Feb 2003 14:03:16 +0000 (14:03 +0000)]
Plan 9: No Configure.
p4raw-id: //depot/perl@18626
Jarkko Hietaniemi [Sun, 2 Feb 2003 13:29:02 +0000 (13:29 +0000)]
Plan 9: doesn't really do Configure (undo #18585/#18586)
p4raw-id: //depot/perl@18624
Jarkko Hietaniemi [Sun, 2 Feb 2003 12:49:13 +0000 (12:49 +0000)]
minitest wasn't working. (The $^X /miniperl/ test in
io/crlf.t didn't work in Tru64, $^X was "./perl", used
the same 'eval q/use .../' trick everywhere.)
p4raw-id: //depot/perl@18622
Jarkko Hietaniemi [Sun, 2 Feb 2003 12:23:28 +0000 (12:23 +0000)]
Fix and test for [perl #15549 Empty \Q\E not permitted]
(test.pl-ise quotemeta.t, too)
p4raw-id: //depot/perl@18621
Artur Bergman [Sat, 1 Feb 2003 21:04:36 +0000 (21:04 +0000)]
Fix bug #15837, temporary from storable call hadn't gotten
freed yet and was cloned but without a real (AvREAL)
reference to it. @_ doesn't refcount!
p4raw-id: //depot/perl@18618
Jarkko Hietaniemi [Sat, 1 Feb 2003 19:43:09 +0000 (19:43 +0000)]
Integrate from perlio:
[ 18591]
PERL_IMPLICIT_SYS does NOT vector PerlIO_xxx but PerlSIO_xxx, so
correct XSUB.h
[ 18611]
Michael Schroeder's fix for re-try if stdio ops after
interrupts. (Calls to PERL_ASYNC_CHECK added by NI-S).
[ 18612]
PerlIO friendly version of speed up of $/ = undef case.
[ 18613]
fflush() before invalidating stdio's fileno.
p4raw-link: @18613 on //depot/perlio:
bbfd922f5bb0c97d0accf18f588b887418c94e6b
p4raw-link: @18612 on //depot/perlio:
e468d35bbed7491e8b221d65181da6200264b6a0
p4raw-link: @18611 on //depot/perlio:
4d94824190f1d3f4307c94794486b277afd6e971
p4raw-link: @18591 on //depot/perlio:
f9415d2377705b4cde579db69c15a29f1b9bb9f6
p4raw-id: //depot/perl@18616
Artur Bergman [Sat, 1 Feb 2003 12:01:01 +0000 (12:01 +0000)]
Don't free thread before using it, breaks platforms that
trust that I really want to free what I free, (like HPUX
and AIX).
p4raw-id: //depot/perl@18615
Nick Ing-Simmons [Sat, 1 Feb 2003 13:03:52 +0000 (13:03 +0000)]
Re: bleadperl: PerlIOStdio_invalidate_fileno breaks fclose semantics
Message-Id: <
20030201130352.1872.12@bactrian.ni-s.u-net.com>
p4raw-id: //depot/perl@18614
Nick Ing-Simmons [Sat, 1 Feb 2003 11:56:47 +0000 (11:56 +0000)]
fflush() before invalidating stdio's fileno.
p4raw-id: //depot/perlio@18613
Nick Ing-Simmons [Sat, 1 Feb 2003 11:37:30 +0000 (11:37 +0000)]
PerlIO friendly version of speed up of $/ = undef case.
p4raw-id: //depot/perlio@18612
Nick Ing-Simmons [Sat, 1 Feb 2003 09:26:31 +0000 (09:26 +0000)]
Michael Schroeder's fix for re-try if stdio ops after
interrupts. (Calls to PERL_ASYNC_CHECK added by NI-S).
p4raw-id: //depot/perlio@18611
Nick Ing-Simmons [Sat, 1 Feb 2003 09:19:20 +0000 (09:19 +0000)]
Integrate mainline
p4raw-id: //depot/perlio@18610
Jarkko Hietaniemi [Fri, 31 Jan 2003 08:58:40 +0000 (08:58 +0000)]
Go even further (back) than change #18608 and revert
to the 5.8.0 code, since otherwise split.t #40 fails.
The four new tests of pat.t should keep us more honest.
p4raw-link: @18608 on //depot/perl:
5dec093fd414e8adffdf65a1d75b0b0d7a2938c3
p4raw-id: //depot/perl@18609
Adrian M. Enache [Fri, 31 Jan 2003 09:20:59 +0000 (11:20 +0200)]
[FIX] Re: UTF-8 failures (surprise!)
Message-ID: <
20030131072059.GB6045@ratsnest.hole>
p4raw-id: //depot/perl@18608
Abhijit Menon-Sen [Fri, 31 Jan 2003 02:29:56 +0000 (02:29 +0000)]
Fix typo in version check ($[ to $]).
p4raw-id: //depot/perl@18607
Jose Auguste-Etienne [Thu, 30 Jan 2003 12:40:53 +0000 (09:40 -0300)]
Upper<->Lower case mistake in perlunicode.pod ?
Message-ID: <OF8E7C1B73.
22FEA7CE-ON03256CBE.
00528AB8@Cgss-Guyane.fr>
p4raw-id: //depot/perl@18606
Beckett Richard-qswi266 [Thu, 30 Jan 2003 11:43:17 +0000 (11:43 +0000)]
[perl #20616] code example in the FAQ that isn't strict-friendly.
Subject: [perl #20616] Tie::IxHash doc error.
From: Beckett Richard-qswi266 (via RT) <perlbug-followup@perl.org>
Message-Id: <rt-20616-49904.16.
891027432675@bugs6.perl.org>
p4raw-id: //depot/perl@18605
Rafael Garcia-Suarez [Thu, 30 Jan 2003 20:52:05 +0000 (20:52 +0000)]
FAQ sync.
p4raw-id: //depot/perl@18604
Adrian M. Enache [Thu, 30 Jan 2003 10:48:09 +0000 (12:48 +0200)]
[FIX] Re: UTF-8 failures (surprise!)
Message-ID: <
20030130084809.GA16902@ratsnest.hole>
Fix for half of the failures, with an expanded test.
p4raw-id: //depot/perl@18603
Todd C. Miller [Wed, 29 Jan 2003 18:38:37 +0000 (18:38 +0000)]
Updates to the openbsd hints
Subject: [perl #20606] [no subject]
From: "Todd C. Miller" (via RT) <perlbug-followup@perl.org>
Message-Id: <rt-20606-49879.4.
78874436094841@bugs6.perl.org>
p4raw-id: //depot/perl@18602
Jarkko Hietaniemi [Wed, 29 Jan 2003 05:58:36 +0000 (05:58 +0000)]
More of the #18600.
p4raw-id: //depot/perl@18601
Jarkko Hietaniemi [Wed, 29 Jan 2003 05:37:12 +0000 (05:37 +0000)]
Retract the %_/SVf change (part of #18456) for PROTOTYPE (CV)
dumping (but now use SvPV_nolen). (This change made an empty
prototype to show up as "_" under -Uuseperlio.)
p4raw-id: //depot/perl@18600
Paul Johnson [Tue, 28 Jan 2003 14:13:09 +0000 (15:13 +0100)]
Fix a very small memory leak in op_free, by reserving the
case op_seq == (U16)-1 for the compiler backend
Subject: Re: Freeing code
From: "Paul Johnson" <paul@pjcj.net>
Message-ID: <18918.193.134.254.145.
1043759589.squirrel@wesley.pjcj.net>
p4raw-id: //depot/perl@18599
Rafael Garcia-Suarez [Tue, 28 Jan 2003 20:43:02 +0000 (20:43 +0000)]
Integrate from maint-5.8:
Change 18450 :
Tiny output tweak.
Change 18429 :
AIX gcc (2.9) threaded build tweak (without the _THREAD_SAFE
things like drand48_data are undefined).
Change 18388 :
List the PERL_MAGIC_utf8 ('w') flag.
p4raw-id: //depot/perl@18598
p4raw-integrated: from //depot/maint-5.8/perl@18597 'copy in'
hints/aix.sh (@18286..) t/op/pat.t (@18290..) pod/perlguts.pod
(@18387..)
Slaven Rezic [Tue, 14 Jan 2003 20:58:54 +0000 (21:58 +0100)]
Re: [perl #20274] perlbug requires dynamic loading
Message-ID: <87n0m3v6yp.fsf@vran.herceg.de>
p4raw-id: //depot/perl@18597
Dave Mitchell [Sun, 26 Jan 2003 17:42:42 +0000 (17:42 +0000)]
Introduce the PUSHSUB_DB macro, equivalent to PUSHSUB, but
without the lval context tweak. Used in OP_DBSTATE.
Subject: Re: [PATCH] Re: change #9754: 5 x slowdown for perl -d
Message-ID: <
20030126174242.H16182@fdgroup.com>
p4raw-id: //depot/perl@18596
Nicholas Clark [Sat, 25 Jan 2003 00:25:50 +0000 (00:25 +0000)]
make Copy On Write work once again
Message-ID: <
20030125002550.GG278@Bagpuss.unfortu.net>
p4raw-id: //depot/perl@18595
Jarkko Hietaniemi [Mon, 27 Jan 2003 20:01:44 +0000 (20:01 +0000)]
Update the pod list for Plan 9.
p4raw-id: //depot/perl@18594
Artur Bergman [Sun, 26 Jan 2003 22:37:21 +0000 (22:37 +0000)]
When we copy things out of a hash and increment the
refcount of the shared scalar, we really must protect
that refcount increase with a lock, or else it might not
happen. Fixes segfault when objects would prematurely
be killed.
p4raw-id: //depot/perl@18593
Jarkko Hietaniemi [Sun, 26 Jan 2003 21:54:36 +0000 (21:54 +0000)]
Missing (STRLEN)casts.
p4raw-id: //depot/perl@18592
Nick Ing-Simmons [Sun, 26 Jan 2003 17:46:19 +0000 (17:46 +0000)]
PERL_IMPLICIT_SYS does NOT vector PerlIO_xxx but PerlSIO_xxx, so
correct XSUB.h
p4raw-id: //depot/perlio@18591
Dave Mitchell [Fri, 24 Jan 2003 18:44:19 +0000 (18:44 +0000)]
OPf_SPECIAL on OP_DBSTATE now indicates a breakpoint.
Subject: Re: [PATCH] Re: change #9754: 5 x slowdown for perl -d
Message-ID: <
20030124184419.F11179@fdgroup.com>
p4raw-id: //depot/perl@18589
Jarkko Hietaniemi [Sun, 26 Jan 2003 12:18:10 +0000 (12:18 +0000)]
Lest people get worried about not having PerlIO.
p4raw-id: //depot/perl@18588
Jarkko Hietaniemi [Sun, 26 Jan 2003 09:19:01 +0000 (09:19 +0000)]
Plan 9 doesn't do nm, from Russ Cox.
p4raw-id: //depot/perl@18586
Jarkko Hietaniemi [Sun, 26 Jan 2003 09:11:23 +0000 (09:11 +0000)]
Plan 9 doesn't really have the symlink API, from Russ Cox.
p4raw-id: //depot/perl@18584
Jarkko Hietaniemi [Sun, 26 Jan 2003 09:08:36 +0000 (09:08 +0000)]
Plan 9 compiler can't return quads from the ?: (condop).
(There may be more spots like this, so a Configure test
maybe in order, but this works for now.) From Russ Cox.
p4raw-id: //depot/perl@18583
Slaven Rezic [Thu, 23 Jan 2003 15:48:52 +0000 (16:48 +0100)]
Re: truncate using a globref
Message-Id: <
200301231448.h0NEmqnu022591@vran.herceg.de>
p4raw-id: //depot/perl@18581
Adrian M. Enache [Fri, 24 Jan 2003 06:23:54 +0000 (08:23 +0200)]
patch to speed up Perl's slurp mode
Message-Id: <
20030124042354.GA30362@ratsnest.hole>
p4raw-id: //depot/perl@18580
Adrian M. Enache [Sun, 26 Jan 2003 04:55:48 +0000 (06:55 +0200)]
Re: [perl #20408] SV's SvPVX get freed even when its SvLEN is zero
Message-Id: <
20030126025548.GB812@ratsnest.hole>
p4raw-id: //depot/perl@18579
Craig A. Berry [Thu, 23 Jan 2003 16:24:46 +0000 (10:24 -0600)]
const declaration fixup
From: "Craig A. Berry" <craigberry@mac.com>
Message-Id: <
3E306BAE.4030807@mac.com>
p4raw-id: //depot/perl@18578
Dave Mitchell [Thu, 23 Jan 2003 20:56:47 +0000 (20:56 +0000)]
move pad.c warnings to different test file
Message-ID: <
20030123205647.C8696@fdgroup.com>
p4raw-id: //depot/perl@18577
Nick Ing-Simmons [Thu, 23 Jan 2003 07:26:43 +0000 (07:26 +0000)]
Fix Solaris 32-bit invalidate case.
p4raw-id: //depot/perl@18573
Rafael Garcia-Suarez [Wed, 22 Jan 2003 21:40:21 +0000 (21:40 +0000)]
Integrate change 18448 from maint-5.8 :
Subject: Partial Data::Dumper patch
From: Richard Clamp <richardc@unixbeard.net>
Date: Sat, 4 Jan 2003 00:52:47 +0000
Message-ID: <
20030104005247.GA27685@mirth.demon.co.uk>
Subject: Re: Partial Data::Dumper patch
From: Slaven Rezic <slaven@rezic.de>
Date: 04 Jan 2003 02:48:08 +0100
Message-ID: <87u1gp4rdz.fsf@vran.herceg.de>
p4raw-id: //depot/perl@18572
p4raw-integrated: from //depot/maint-5.8/perl@18571 'copy in'
ext/Data/Dumper/t/dumper.t (@17645..)
p4raw-integrated: from //depot/maint-5.8/perl@18448 'merge in'
ext/Data/Dumper/Dumper.xs (@17645..)
Dave Mitchell [Thu, 16 Jan 2003 17:12:37 +0000 (17:12 +0000)]
Re: difference between my and our before introduction
Message-ID: <
20030116171237.C420@fdgroup.com>
(with a small test tweak)
p4raw-id: //depot/perl@18571
Nick Ing-Simmons [Wed, 22 Jan 2003 21:03:43 +0000 (21:03 +0000)]
Integrate mainline
p4raw-id: //depot/perlio@18570
Rafael Garcia-Suarez [Wed, 22 Jan 2003 21:00:59 +0000 (21:00 +0000)]
Fix [perl #19860] by adding a POSIX::fsync() stub.
p4raw-id: //depot/perl@18569
Jarkko Hietaniemi [Wed, 22 Jan 2003 17:43:45 +0000 (17:43 +0000)]
Integrate from perlio:
[ 18560]
Fixes for open.pm which attempts to load layers:
1. C equivalent of local $SIG{__WARN__} = sub {}
while loading layers to supress warnings lib/open.t does
not want.
2. The loading scheme does not recurse now so look for
new symptom of bad layer which is that a good module
fails to load (as we cannot open any files).
NOTE: In my opinion open.pm should probably die on bad layer
spec rather than just (maybe) warning and then allowing opens
to fail.
p4raw-link: @18560 on //depot/perlio:
c9bca74aca217023baf0f921dcffaaa072a83cf3
p4raw-id: //depot/perl@18561
Nick Ing-Simmons [Wed, 22 Jan 2003 17:19:41 +0000 (17:19 +0000)]
Fixes for open.pm which attempts to load layers:
1. C equivalent of local $SIG{__WARN__} = sub {}
while loading layers to supress warnings lib/open.t does
not want.
2. The loading scheme does not recurse now so look for
new symptom of bad layer which is that a good module
fails to load (as we cannot open any files).
NOTE: In my opinion open.pm should probably die on bad layer
spec rather than just (maybe) warning and then allowing opens
to fail.
p4raw-id: //depot/perlio@18560
Jarkko Hietaniemi [Wed, 22 Jan 2003 13:19:31 +0000 (13:19 +0000)]
Upgrade to Encode 1.86.
p4raw-id: //depot/perl@18559
Dave Mitchell [Sun, 19 Jan 2003 17:22:05 +0000 (17:22 +0000)]
Re: slight tweak to -DR flag - display whether on tmps stack
Message-ID: <
20030119172204.D24444@fdgroup.com>
p4raw-id: //depot/perl@18558
Nicholas Clark [Tue, 21 Jan 2003 22:27:21 +0000 (22:27 +0000)]
Re: [perl #9394] Re: [ID
20020525.002] coredump/ bad free warning in blead with SIGWARN
Message-ID: <
20030121222720.GG293@Bagpuss.unfortu.net>
p4raw-id: //depot/perl@18557
Robin Barker [Mon, 20 Jan 2003 15:26:21 +0000 (15:26 +0000)]
(was RE: [PATCH] %_ (was Re: [PATCH] operation on `PL_na' may be undefined))
Message-ID: <
533D273D4014D411AB1D00062938C4D904046299@hotel.npl.co.uk>
p4raw-id: //depot/perl@18556
Jarkko Hietaniemi [Wed, 22 Jan 2003 12:57:20 +0000 (12:57 +0000)]
Integrate from perlio:
[ 18539]
Try the invalidate_fileno() idea for non-leaky stdio sockets.
[ 18540]
stdio does NOT set errno==EBADF on invalidated fileno
[ 18547]
Tweak APItest for new semantics of printf.
Now printf() is always real printf() and not #define'd to
PerlIO_stdoutf() so PerlIO_flush() when STDOUT is redirected
back to original file does not affect stdout unless we are
using :stdio as implementation. So things printf()ed XS side
are not written to file (and turn up at end of test).
So add an XS function which does fflush(stdout), modify .pm
file to export it and .t file to call it.
[ 18548]
Add defined(__osf__) case to invalidate_fileno (for Tru64).
[ 18549]
Report layer name (without arg) on fail, and attempt to
autoload layer name. (Fix for [perl #20460]).
[ 18550]
#if defined(__irix__) (for stdio invalidate) from Jarkko.
p4raw-link: @18550 on //depot/perlio:
9837d3731bea1e0d3aaed58a46127574f76ffe53
p4raw-link: @18549 on //depot/perlio:
c7732655732824f25d1659f5ebb5d4c3538d32e8
p4raw-link: @18548 on //depot/perlio:
9110aad070a9791bfefe4a0d1718537633723305
p4raw-link: @18547 on //depot/perlio:
9d9116831508c6c9cd81c32834701af14b7647a4
p4raw-link: @18540 on //depot/perlio:
e85294730ea95da49196ebe8f63ce0db11ebd231
p4raw-link: @18539 on //depot/perlio:
0d7a53989a672378609205d638df0119542bcf7e
p4raw-id: //depot/perl@18555
Dave Mitchell [Sun, 19 Jan 2003 16:43:54 +0000 (16:43 +0000)]
Re: [perl #15439] unreferenced scalar due to double DESTROY
Message-ID: <
20030119164353.B24444@fdgroup.com>
p4raw-id: //depot/perl@18554
Martti Rahkila [Sat, 18 Jan 2003 00:25:39 +0000 (02:25 +0200)]
Term::Complete problem + fix (Was: Re: muttprofile + perl 5.8)
Message-ID: <
20030118002539.A30585@wurlitzer.hut.fi>
(chose the 'even safer' version)
p4raw-id: //depot/perl@18553
Nicholas Clark [Fri, 17 Jan 2003 19:40:45 +0000 (19:40 +0000)]
assert PV isn't RV
Message-ID: <
20030117194044.GB297@Bagpuss.unfortu.net>
p4raw-id: //depot/perl@18552
Jarkko Hietaniemi [Wed, 22 Jan 2003 12:28:11 +0000 (12:28 +0000)]
Integrate from maint-5.8:
[ 18375]
Better version of change #18370; use the new 'E' flag
so that cxinc is exposed (as cpp) only to List::Util
(well, any extension) (well, anyone who does -DPERL_EXT),
no need to export it globally.
p4raw-link: @18375 on //depot/maint-5.8/perl:
c8ec9884b4dd9ffdd0a7d1514952a74c55bfab3e
p4raw-link: @18370 on //depot/maint-5.8/perl:
94115b5677e78fbc3321ad88fedfaa3b6bc804bf
p4raw-id: //depot/perl@18551
p4raw-integrated: from //depot/maint-5.8/perl@18545 'copy in'
ext/List/Util/Makefile.PL (@17645..) 'merge in' embed.fnc
embed.h global.sym (@18370..)
Nick Ing-Simmons [Wed, 22 Jan 2003 12:27:48 +0000 (12:27 +0000)]
#if defined(__irix__) (for stdio invalidate) from Jarkko.
p4raw-id: //depot/perlio@18550
Nick Ing-Simmons [Wed, 22 Jan 2003 09:06:23 +0000 (09:06 +0000)]
Report layer name (without arg) on fail, and attempt to
autoload layer name. (Fix for [perl #20460]).
p4raw-id: //depot/perlio@18549
Nick Ing-Simmons [Wed, 22 Jan 2003 08:47:24 +0000 (08:47 +0000)]
Add defined(__osf__) case to invalidate_fileno (for Tru64).
p4raw-id: //depot/perlio@18548
Nick Ing-Simmons [Wed, 22 Jan 2003 08:32:59 +0000 (08:32 +0000)]
Tweak APItest for new semantics of printf.
Now printf() is always real printf() and not #define'd to
PerlIO_stdoutf() so PerlIO_flush() when STDOUT is redirected
back to original file does not affect stdout unless we are
using :stdio as implementation. So things printf()ed XS side
are not written to file (and turn up at end of test).
So add an XS function which does fflush(stdout), modify .pm
file to export it and .t file to call it.
p4raw-id: //depot/perlio@18547
Nick Ing-Simmons [Wed, 22 Jan 2003 08:26:14 +0000 (08:26 +0000)]
Integrate mainline
p4raw-id: //depot/perlio@18546
Richard Soderberg [Wed, 22 Jan 2003 02:08:13 +0000 (21:08 -0500)]
AUTHORS correction
Message-Id: <
45B4D3AE-2DD8-11D7-8CF9-
000393801F84@crystalflame.net>
p4raw-id: //depot/perl@18545
Jarkko Hietaniemi [Wed, 22 Jan 2003 03:45:01 +0000 (03:45 +0000)]
Continuation of #18542.
p4raw-id: //depot/perl@18544
Jarkko Hietaniemi [Wed, 22 Jan 2003 03:21:23 +0000 (03:21 +0000)]
Update the MD5 checksums.
p4raw-id: //depot/perl@18543
Hugo van der Sanden [Wed, 22 Jan 2003 01:37:52 +0000 (01:37 +0000)]
#18530 misses a couple of pTHX definitions
p4raw-id: //depot/perl@18542
Jarkko Hietaniemi [Tue, 21 Jan 2003 22:08:51 +0000 (22:08 +0000)]
Try to make including patchlevel.h a bit more
robust (use <> instead of "") and in case of
failure (PERL_VERSION didn't get defined) try
to fail (by trying to #include an "obviously"
named file that should not exist) (didn't use
#error since IIRC older cpps do not have it)
(Tries to address [perl #20276].)
p4raw-id: //depot/perl@18541
Nick Ing-Simmons [Tue, 21 Jan 2003 20:53:00 +0000 (20:53 +0000)]
stdio does NOT set errno==EBADF on invalidated fileno
p4raw-id: //depot/perlio@18540
Nick Ing-Simmons [Tue, 21 Jan 2003 20:22:55 +0000 (20:22 +0000)]
Try the invalidate_fileno() idea for non-leaky stdio sockets.
p4raw-id: //depot/perlio@18539
Yitzchak Scott-Thoennes [Mon, 30 Dec 2002 18:01:16 +0000 (10:01 -0800)]
Re: FreeBSD 5 hints patch
Message-ID: <spPE+gzkg+6F092yn@efn.org>
p4raw-id: //depot/perl@18538
Jarkko Hietaniemi [Tue, 21 Jan 2003 14:18:40 +0000 (14:18 +0000)]
Allow explicit -C enable/disable by -C:1/-C:0
(suggested by Peter Haworth).
p4raw-id: //depot/perl@18536
Chip Turner [Thu, 2 Jan 2003 11:35:57 +0000 (06:35 -0500)]
PATCH: perldoc in MAINT branch (probably unstable too)
Message-ID: <vzksmwb8q6q.fsf@minbar.devel.redhat.com>
p4raw-id: //depot/perl@18534
Hugo van der Sanden [Thu, 2 Jan 2003 14:33:49 +0000 (14:33 +0000)]
Re: [perl #17757] s///g fails when using English & study in 5.8.0
Message-Id: <
200301021433.h02EXno03562@crypt.compulink.co.uk>
p4raw-id: //depot/perl@18533
Martien Verbruggen [Tue, 31 Dec 2002 16:42:28 +0000 (03:42 +1100)]
Re: [perl #19236] perlsyn: implicit localisation in while()
Message-Id: <slrnb12bi4.4tt.mgjv@martien.heliotrope.home>
p4raw-id: //depot/perl@18532
Hugo van der Sanden [Tue, 21 Jan 2003 01:52:49 +0000 (01:52 +0000)]
integrate #18366-18370 from maint-5.8:
Implement the CPP namespace protection from [perl #8610];
unless a function has an 'A' flag, it's defined only if
PERL_CORE. (Why do we have so many many functions with an
'A' flag, is another matter.) (The number of #ifdef PERL_COREs
could be cut down further by having a state variable and
inserting #ifdefs and #endifs when the state flips.)
Caught by change #18366; thanks to the re extension we have some
functions that are file static but yet part of "the public API"...
Maybe a new class of exportability (a new embed.fnc flag) is
required for this case.
Continue on change #18367; implement the said flag, 'E'.
(Some nm output studying was needed to get the right set
of re symbols defined.)
In AIX (strict symbol exports) the cxinc is needed for List::Util.
p4raw-link: @18367 on //depot/maint-5.8/perl:
6f21c313c98cc4764e09fbbce2c8792576c90d7d
p4raw-link: @18366 on //depot/maint-5.8/perl:
704f60d9cfa8683df64f005e872b37aea8f9d60e
p4raw-id: //depot/perl@18531
Hugo van der Sanden [Tue, 21 Jan 2003 01:37:03 +0000 (01:37 +0000)]
integrate (by hand) #18353 and #18359 from maint-5.8:
Introduce a cache for UTF-8 data: length and byte<->char offset
mapping are stored in a new type of magic. Speeds up length(),
substr(), index(), rindex(), pos(), and some parts of s///.
The speedup varies a lot (on the usual suspects: what is the
access pattern of the data, compiler, CPU), but should be at
least one order of magnitude, and getting to the same magnitude
as byte string speeds, and in some cases (length on unchanged data)
even reaching the byte string speed. On the other hand, in some
cases (index) the byte speed is still faster by a factor of five
or so, but the bottleneck there does not seem to be any more
the byte<->char offset mapping (instead, the fbm_instr() speed).
There is one cache slot for the length, and only two for the
byte<->char offset mapping (the first one for the start->offset,
and the second for the offset->offset+length, when talking
in substr() terms).
Code this hairy is bound to have hairy trolls hiding under it.
[...]
A small tweak on top of #18353: don't display mg_len bytes of
mg_ptr for PERL_MAGIC_utf8 because that's not what's there.
p4raw-id: //depot/perl@18530
Hugo van der Sanden [Tue, 21 Jan 2003 00:44:20 +0000 (00:44 +0000)]
integrate #18349 from maint-5.8:
At least partially address [perl #10000] by speeding
up both the ASCII case (by about 2-3%) and the UTF-8 case
(by about 45%). The major trick is to avoid hitting the
costly S_reginclass(). (Even before this patch the speedup
since 5.8.0 was about 40-50%.) After this the UTF-8 case is
still about 30-60% slower than the ASCII case. (Note that
I'm unable to reproduce the 10-fold speed difference of the
original bug report; I can see a factor of 2 or 3, but no more.)
p4raw-id: //depot/perl@18529
p4raw-integrated: from //depot/maint-5.8/perl@18528 'merge in'
regexec.c (@18347..)
Dave Mitchell [Sun, 29 Dec 2002 15:09:31 +0000 (15:09 +0000)]
silence warning
Message-ID: <
20021229150931.C2325@fdgroup.com>
p4raw-id: //depot/perl@18528
H.Merijn Brand [Wed, 18 Dec 2002 08:30:32 +0000 (09:30 +0100)]
perl version reporting
From: "H.Merijn Brand" <h.m.brand@hccnet.nl>
Message-Id: <
20021218081957.ABD8.H.M.BRAND@hccnet.nl>
p4raw-id: //depot/perl@18527
Blair Zajac [Wed, 8 Jan 2003 12:28:35 +0000 (04:28 -0800)]
Re: When Storable and Devel::DProf mix, core dump [perl #19385]
Message-ID: <
3E1C89F3.2B4759C@orcaware.com>
p4raw-id: //depot/perl@18526
Jarkko Hietaniemi [Mon, 20 Jan 2003 17:51:10 +0000 (17:51 +0000)]
Integrate from perlio:
[ 18520]
Move #define printf out of perl.h into appropriate
PerlIO files.
p4raw-link: @18520 on //depot/perlio:
f2b1176c042dd0d792a0f916f759bd848210d4d8
p4raw-id: //depot/perl@18524
Jarkko Hietaniemi [Mon, 20 Jan 2003 17:48:51 +0000 (17:48 +0000)]
Upgrade to Digest 1.02.
p4raw-id: //depot/perl@18523
Adrian M. Enache [Mon, 20 Jan 2003 01:54:07 +0000 (01:54 +0000)]
[perl #20408] SV's SvPVX get freed even when its SvLEN is zero
From: "enache@rdslink.ro (via RT)" <perlbug-followup@perl.org>
Message-Id: <rt-20408-49009.18.
2567818321236@bugs6.perl.org>
p4raw-id: //depot/perl@18522
Jarkko Hietaniemi [Mon, 20 Jan 2003 17:37:35 +0000 (17:37 +0000)]
Retract #18154 (integrate of perlio #18507), not enough
threadsafety.
p4raw-id: //depot/perl@18521
Nick Ing-Simmons [Mon, 20 Jan 2003 15:09:08 +0000 (15:09 +0000)]
Move #define printf out of perl.h into appropriate
PerlIO files.
p4raw-id: //depot/perlio@18520
Jarkko Hietaniemi [Mon, 20 Jan 2003 12:25:43 +0000 (12:25 +0000)]
Encode MANIFEST update.
p4raw-id: //depot/perl@18519
Nick Ing-Simmons [Mon, 20 Jan 2003 09:37:52 +0000 (09:37 +0000)]
Integrate mainline
p4raw-id: //depot/perlio@18518
Nick Ing-Simmons [Mon, 20 Jan 2003 08:34:10 +0000 (08:34 +0000)]
Integrate #18405 from maint-5.8 which fixes whitespac
issues in open(FH,"&...")
p4raw-id: //depot/perl@18517
p4raw-integrated: from //depot/maint-5.8/perl@18516 'merge in' doio.c
(@18405..)
Jarkko Hietaniemi [Sun, 19 Jan 2003 13:48:10 +0000 (13:48 +0000)]
microperl update.
p4raw-id: //depot/perl@18516
Jarkko Hietaniemi [Sat, 18 Jan 2003 18:36:00 +0000 (18:36 +0000)]
Integrate from perlio:
[ 18507]
Thread-protection for dup/fclose/dup2 scheme of stdio leak
avoidance.
p4raw-link: @18507 on //depot/perlio:
997ec9b4c5029a8aa56f484e7b625da92e3e0521
p4raw-id: //depot/perl@18514
Jarkko Hietaniemi [Sat, 18 Jan 2003 18:07:01 +0000 (18:07 +0000)]
Slight pod reformatting based on [perl #20284].
p4raw-id: //depot/perl@18513
Jarkko Hietaniemi [Fri, 17 Jan 2003 21:00:15 +0000 (21:00 +0000)]
EPOC perl address updates.
p4raw-id: //depot/perl@18512
Jarkko Hietaniemi [Fri, 17 Jan 2003 20:23:42 +0000 (20:23 +0000)]
Upgrade to Digest 1.01.
p4raw-id: //depot/perl@18511
Jarkko Hietaniemi [Fri, 17 Jan 2003 19:53:37 +0000 (19:53 +0000)]
Typo fixes from jmc@prioris.mini.pw.edu.pl (from openbsd.bugs)
p4raw-id: //depot/perl@18510
Jarkko Hietaniemi [Fri, 17 Jan 2003 19:45:11 +0000 (19:45 +0000)]
Upgrade to Digest::MD5 2.22.
p4raw-id: //depot/perl@18509
Nick Ing-Simmons [Fri, 17 Jan 2003 15:10:34 +0000 (15:10 +0000)]
Re: 18457 on cygwin/
20030113
Message-Id: <
20030117151034.2651.3@bactrian.elixent.com>
p4raw-id: //depot/perl@18508
Nick Ing-Simmons [Fri, 17 Jan 2003 14:05:27 +0000 (14:05 +0000)]
Thread-protection for dup/fclose/dup2 scheme of stdio leak
avoidance.
p4raw-id: //depot/perlio@18507