Ilya Zakharevich [Thu, 27 Feb 1997 06:53:51 +0000]
pp_undef was not always freeing memory
As I found, pp_undef in
$a = 'aaa';
$a = 0;
undef $a;
was not reclaiming the storage, as it does in
$a = 'aaa';
undef $a;
The patch is below fixes this shortcoming (tests ok, perl rebuilds
itself OK),
Enjoy,
p5p-msgid: <
199702270707.CAA13978@monk.mps.ohio-state.edu>
private-msgid: <
199702270653.BAA13949@monk.mps.ohio-state.edu>
Gurusamy Sarathy [Tue, 25 Feb 1997 18:24:02 +0000]
Clean up and document API for hashes
On Tue, 25 Feb 1997 09:01:57 EST, Chip Salzenberg wrote:
>I've had 99.44% good reports about _90. I'd love to go public
>immediately, but Sarathy's perl_call fix is just too valuable to pass
>up, and there's a taint test in the works. So please don't tell
>anyone that _90 is the first public beta; that will likely be _91.
FWIW, the Beta Zvoid is alive and kicking on Linux 2.0.28.
You might want to put this in too. (The patches to hv.h are mostly
to an unused macro that will be useful for extension writers).
p5p-msgid: <
199702251824.NAA14859@aatma.engin.umich.edu>
Gurusamy Sarathy [Tue, 25 Feb 1997 22:00:49 +0000]
_90 op/tie.t gets "Segmentation fault" with Sarathy's perl_call patch
On Tue, 25 Feb 1997 14:41:30 EST, Chip Salzenberg wrote:
>According to Gurusamy Sarathy:
>> On Tue, 25 Feb 1997 10:07:02 PST, David Dyck wrote:
>> >summary:
>> > op/tie.t gets "Segmentation fault" with Sarathy's perl_call patch
>>
>> Try this patch.
>
>That's got it. Now pureperl doesn't complain about op/tie.t.
>
Good. This here patch sets up the flag in a few more places (to avoid
redundant sigjmp()s which was why t/op/tie.t triggered execution
of docatch()). This patch fixes that flag leak.
p5p-msgid: <
199702252200.RAA16853@aatma.engin.umich.edu>
Gurusamy Sarathy [Tue, 25 Feb 1997 19:25:39 +0000]
_90 op/tie.t gets "Segmentation fault" with Sarathy's perl_call patch
On Tue, 25 Feb 1997 10:07:02 PST, David Dyck wrote:
>
>summary:
> op/tie.t gets "Segmentation fault" with Sarathy's perl_call patch
>
Oops. Thanks for that report (Evil how the segfault doesn't show
up under `make test`. Evil.)
Try this patch. (I was being risque with the runlevel pretense...)
p5p-msgid: <
199702251925.OAA15498@aatma.engin.umich.edu>
Gurusamy Sarathy [Tue, 25 Feb 1997 07:25:56 +0000]
Fix perl_call_*() when !G_EVAL
On Mon, 24 Feb 1997 15:19:17 EST, Gurusamy Sarathy wrote:
>On Mon, 24 Feb 1997 12:53:57 GMT, Tim Bunce wrote:
>>> From: Tom Christiansen <tchrist@jhereg.perl.com>
>>> >Dprof "works".
>>> Then how come it's not in the core? :-(
>>I'd certainly like it to be there for 5.004.
>
>I'd agree, except there's this bug in perl_call_*() that makes
>it fail to run this fully:
>
> % perl -d:DProf -e 'sub T { eval { die "burp" } } T(); print "zip\n"'
> %
Ok, here's a patch for the perl_call_*() problems with error traps,
meant for 5.004 (hope I didn't miss the boat!).
This is a subset of the functionality contained in Michael Schroeder's
stack-of-stacks patch. The patch itself if simple: code that calls
runops() without explicitly setting up a jmp_buf sets a flag that
indicates doeval() is responsible for catching any longjmp()s
locally. The three places that call doeval() then call setjmp()
based on this flag.
This patch is binary compatible and minimal (as opposed to the
stack-of-stacks patch which has other issues involved, making it
more complicated). There's a testsuite with 9 tests (3_28 fails all
but one).
p5p-msgid: <
199702250725.CAA09192@aatma.engin.umich.edu>
Tom Phoenix (with help from M.J.T. Guy [Mon, 24 Feb 1997 16:36:06 +0000]
Updates to perlfunc.pod
Signed-off-by: Tom Phoenix (with help from M.J.T. Guy <tomphoenix@unknown>
Andy Dougherty [Fri, 28 Feb 1997 16:35:49 +0000]
Warn about intrusive sfio behavior
private-msgid: <Pine.SOL.3.95q.
970228112136.24038G-100000@fractal.lafayette.
Doug MacEachern [Thu, 27 Feb 1997 16:25:04 +0000]
Allow explicit '-lperl' in link arguments
private-msgid: <
199702271625.LAA25402@postman.osf.org>
Chip Salzenberg [Tue, 25 Feb 1997 01:12:02 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Chip Salzenberg [Tue, 25 Feb 1997 01:12:02 +0000]
Don't warn on use of CCFLAGS
(this is the same change as commit
69d91587b2e24056be123e153ca28f4e82fa8dff, but as applied)
Andreas Koenig [Tue, 25 Feb 1997 10:38:43 +0000]
Don't warn on use of CCFLAGS
Subject: CCFLAGS (Was: This message did not make it ...)
>>>>> Raphael Manfredi writes:
> Making ODBM_File (dynamic)
> Processing hints file hints/hpux.pl
> 'CCFLAGS' is not a known MakeMaker parameter name.
> error. Upon closer inspection, it looks like the hint file is
> doing:
> $self->{CCFLAGS} = $Config{ccflags} . ' -DDBM_BUG_DUPLICATE_FREE' ;
> Why the 'CCFLAGS' key is unknown is beyond me. It would be nice
> to find the roots of this problem before we go to beta.
Finally, here it is. Unfortunately, I cannot remember how it was lost
from the list.
p5p-msgid: <
199702251038.LAA13123@anna.in-berlin.de>
Chip Salzenberg [Tue, 25 Feb 1997 01:12:02 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Chip Salzenberg [Tue, 25 Feb 1997 01:12:02 +0000]
Follow up on elimination of $` $& $' in libraries
(this is the same change as commit
efa45da9e2f446e4e79f819d2fa3a099e54425fa, but as applied)
M.J.T. Guy [Fri, 28 Feb 1997 13:59:42 +0000]
Follow up on elimination of $` $& $' in libraries
Ilya Zakharevich <ilya@math.ohio-state.edu> wrote
> $' $` $& eliminated from all the known *.pm.
Close, but you overlooked the 'use English' in diagnostics.pm.
See additional patch below.
p5p-msgid: <E0w0Sqc-00046E-00@ursa.cus.cam.ac.uk>
Ilya Zakharevich [Tue, 25 Feb 1997 19:37:07 +0000]
Make IPC::Open3 work without fork()
private-msgid: <
199702251937.OAA10718@monk.mps.ohio-state.edu>
Chip Salzenberg [Tue, 25 Feb 1997 01:58:14 +0000]
Don't require() in a signal handler
Chip Salzenberg [Tue, 25 Feb 1997 01:12:02 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Chip Salzenberg [Tue, 25 Feb 1997 01:12:02 +0000]
Update OS/2 Configure diff
(this is the same change as commit
ef05b0d7eb79c109aa2e843e7ab5d5abad251c55, but as applied)
Ilya Zakharevich [Tue, 25 Feb 1997 19:06:23 +0000]
Update OS/2 Configure diff
Chip Salzenberg writes:
>
> (I missed the $randbits case in _90; it is fixed in _91.)
The broken $randbits test (no $ldflags) is just ignored under OS/2
anyway, so this should help.
p5p-msgid: <
199702251906.OAA10608@monk.mps.ohio-state.edu>
Gurusamy Sarathy [Thu, 27 Feb 1997 18:02:30 +0000]
Silence bogus typo warning on $DB::postponed
private-msgid: <
199702271802.NAA12505@aatma.engin.umich.edu>
Ilya Zakharevich [Thu, 27 Feb 1997 19:24:36 +0000]
Fix SEGV when debugging with foreach() lvalue patch
p5p-msgid: <
199702271924.OAA14557@monk.mps.ohio-state.edu>
Chip Salzenberg [Fri, 21 Feb 1997 14:41:53 +0000]
[differences between cumulative patch application and perl-5.003_90]
Perl 5 Porters [Fri, 21 Feb 1997 14:41:53 +0000]
[inseparable changes from match from perl5.003_28 to perl-5.003_90]
BUILD PROCESS
Subject: Re: ccdlflags don't quite work
Date: Mon, 24 Feb 1997 16:07:07 -0500 (EST)
From: Andy Dougherty <doughera@fractal.phys.lafayette.edu>
Files: Configure
Msg-ID: <Pine.SOL.3.95q.
970224160630.5700E-100000@fractal.lafayette.e
(applied based on p5p patch as commit
bc75e68b7b49a5ac07dc839aec7ca05eba51175f)
Subject: Use $ccflags, $ldflags, $libs when determining $randbits
From: Chip Salzenberg <chip@perl.com>
Files: Configure
Subject: 'installperl -v' doesn't do enough
Date: Mon, 24 Feb 1997 08:42:59 -0500
From: Spider Boardman <spider@Orb.Nashua.NH.US>
Files: installperl
Msg-ID: <
199702241342.IAA25945@Orb.Nashua.NH.US>
(applied based on p5p patch as commit
7287d43f85a3731984c0e0ef1f20e1515b4140f2)
Subject: installperl breaks running system (for a while)
Date: Mon, 24 Feb 1997 09:12:11 -0500
From: Spider Boardman <spider@Orb.Nashua.NH.US>
Files: installperl
Msg-ID: <
199702241412.JAA11829@Orb.Nashua.NH.US>
(applied based on p5p patch as commit
ec2a3d87ba26a9586ad981051a949403027c147a)
CORE PORTABILITY
Subject: Digital UNIX and 3_28
Date: Sun, 23 Feb 1997 16:27:19 +0200 (EET)
From: Jarkko Hietaniemi <jhi@iki.fi>
Files: Configure MANIFEST ext/NDBM_File/hints/dec_osf.pl ext/ODBM_File/hints/dec_osf.pl hints/dec_osf.sh
Msg-ID: <
199702231427.QAA13807@alpha.hut.fi>
(applied based on p5p patch as commit
58c0852f6968d0b4520a03e22a56226185ab78c6)
Subject: AmigaOS patches to 5.003_28
Date: Sat, 22 Feb 1997 18:08:02 +0100
From: "Norbert Pueschel" <pueschel@imsdd.meb.uni-bonn.de>
Files: README.amiga hints/amigaos.sh t/io/fs.t t/lib/anydbm.t t/lib/db-btree.t t/lib/db-hash.t t/lib/db-recno.t t/lib/gdbm.t t/lib/ndbm.t t/lib/odbm.t t/lib/sdbm.t t/op/magic.t t/op/stat.t
Msg-ID: <
77724759@Armageddon.meb.uni-bonn.de>
(applied based on p5p patch as commit
b8e6d11c134e93a7795379ceb62b7f950607c667)
TESTS
Subject: Don't use <*> where readdir() will do
From: Chip Salzenberg <chip@perl.com>
Files: t/op/stat.t
Subject: Allow for $^X to be 'miniperl'
Date: Sun, 23 Feb 1997 16:22:45 +0100
From: Dominic Dunlop <domo@slipper.ip.lu>
Files: t/op/magic.t
private-msgid: <v
03020903af360f31aced@[194.51.248.65]>
Spider Boardman [Mon, 24 Feb 1997 14:12:11 +0000]
installperl breaks running system (for a while)
"installperl" causes scripts with "/usr/bin/perl" or equivalent to
fail during the process of installation, which can take a while with
slow disks and/or slow CPUs. It should install the libraries and such
*before* linking "$installbin/perl" to the version-dependent executable.
p5p-msgid: <
199702241412.JAA11829@Orb.Nashua.NH.US>
Chip Salzenberg [Fri, 21 Feb 1997 14:41:53 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Spider Boardman [Mon, 24 Feb 1997 13:42:59 +0000]
'installperl -v' doesn't do enough
"installperl -v" doesn't install enough for even "perl5.00328 -V" to
work. It should install the version-dependent portions of the "lib"
directory as well as the perl executable.
p5p-msgid: <
199702241342.IAA25945@Orb.Nashua.NH.US>
Chip Salzenberg [Fri, 21 Feb 1997 14:41:53 +0000]
Document one-argument limitation with #! line
(this is the same change as commit
3c5180aa4845c30d7891bc90e491cc6f9b6ecd85, but as applied)
Tom Phoenix [Mon, 24 Feb 1997 02:41:02 +0000]
Document one-argument limitation with #! line
On Sun, 23 Feb 1997, Tom Christiansen wrote:
> If it is important to you that your bug be fixed, do monitor the
> Changes file in any development releases since the time you submitted
> the bug, and encourage the maintainers with kind words (but never any
> flames!). Feel free to resend your bug report if the next released
> version of perl comes out and your bug is still present.
That's good. And how about adding something like this?
p5p-msgid: <Pine.GSO.3.95q.
970223151026.18178P-100000@kelly.teleport.com>
private-msgid: <Pine.GSO.3.95q.
970223182745.15989A-100000@kelly.teleport.com
Andy Dougherty [Mon, 24 Feb 1997 21:07:07 +0000]
Re: ccdlflags don't quite work
On Mon, 24 Feb 1997, Matthias Urlichs wrote:
> Two problems:
>
> # How will the perl executable find the installed shared $libperl?
> # Add $xxx to ccdlflags.
> ... and it does this every time I reconfigure Perl without first saying
> "rm config.sh", so the list grows longer and longer and... Ugh. Probably
> the best solution is to add a Configure variable which does nothing but
> remember if we have added $xxx yet.
I think this will fix the problem.
p5p-msgid: <Pine.SOL.3.95q.
970224160630.5700E-100000@fractal.lafayette.edu>
private-msgid: <Pine.SOL.3.95q.
970224160630.5700E-100000@fractal.lafayette.e
Jarkko Hietaniemi [Fri, 21 Feb 1997 15:38:06 +0000]
Fix typo in 'Tolkien quotation typo' fix
Norbert Pueschel [Sat, 22 Feb 1997 17:08:02 +0000]
AmigaOS patches to 5.003_28
Here are some patches for AmigaOS, mainly to switch of unwanted tests.
With these patches, perl 5.003_28 passes all tests under AmigaOS.
You should check if there are other hint files that need the new
dont_use_nlink variable added.
p5p-msgid: <
77724759@Armageddon.meb.uni-bonn.de>
Andy Dougherty [Mon, 24 Feb 1997 21:06:02 +0000]
pumpkin-1.9.pod
private-msgid: <Pine.SOL.3.95q.
970224155702.5700D-100000@fractal.lafayette.e
Jarkko Hietaniemi [Sun, 23 Feb 1997 14:27:19 +0000]
Digital UNIX and 3_28
Please try this on 3_28 in as many Digital UNIX / DEC OSF/1 boxes you
can and report the results back to me. I have already 100% positive
make test results from
Digital UNIX 4.0 cc static/shared, gcc static/shared
Digital UNIX 4.0B cc static/shared, gcc static/shared
DEC OSF/1 3.2C cc static/shared, gcc static/shared
so you can test some other os / compiler combinations first.
p5p-msgid: <
199702231427.QAA13807@alpha.hut.fi>
Chip Salzenberg [Fri, 21 Feb 1997 14:41:53 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Chip Salzenberg [Fri, 21 Feb 1997 14:41:53 +0000]
Add documentation and '-h' option to perlbug
This was the summary description that Chip wrote for the two changes:
4fc194b2052979cb8d28873cf62c0ce84cd214f5
3f4fc4634a226330cf98eab6871f26f45d52e425
Gurusamy Sarathy [Mon, 24 Feb 1997 20:09:51 +0000]
perlbug.pod for 5.004
On Mon, 24 Feb 1997 16:10:10 GMT, "M.J.T. Guy" wrote:
>You wrote to perl5-porters:
>> I hope I got Kenneth right as the AUTHOR of perlbug.
>
>Doesn't he deserve the E<lt> treatment, like the rest of you?
>
>>+ Kenneth Albanowski <kjahds@kjahds.com>, subsequently I<doc>tored by
Sure. That, and a few pod2man noises fixed below. Goes on top
of the previous patch.
p5p-msgid: <
199702242009.PAA02849@aatma.engin.umich.edu>
Gurusamy Sarathy [Mon, 24 Feb 1997 22:29:30 +0000]
Re: Hash key created by subroutine call? (fwd)
On Mon, 24 Feb 1997 06:20:14 MST, Tom Christiansen wrote:
>Considering how frequently this gets asked, how about
>putting it in perlsub?
>
>------- start of forwarded message -------
>From: Rick Smith <ricks@sd.znet.com>
>Newsgroups: comp.lang.perl.misc
>Subject: Hash key created by subroutine call?
>Date: Sat, 22 Feb 1997 21:47:52 -0800
I seem to be in a documental state today.
p5p-msgid: <
199702242229.RAA04395@aatma.engin.umich.edu>
Gurusamy Sarathy [Sun, 23 Feb 1997 11:00:00 +0000]
perlbug.pod for 5.004
On Sun, 23 Feb 1997 15:06:24 MST, Tom Christiansen wrote:
>Maybe this should talk more about the perlbug program, or be podfully
>included in that script?
>
>--tom
I've made some additions, along with a merge into utils/perlbug.PL.
(I didn't see Tom(rootbeer)'s suggestion before making the patch,
so that's not in.)
I hope I got Kenneth right as the AUTHOR of perlbug.
Do we need a `ln -s utils/perlbug pod/perlbug.pod` too?
p5p-msgid: <
199702240854.DAA27128@aatma.engin.umich.edu>
Andy Dougherty [Mon, 24 Feb 1997 22:09:09 +0000]
Post-28 INSTALL updates
Here are some more updates to the INSTALL file. Specifically, I revised
the malloc section a bit to give more of an overall perspective (and,
admittedly, less detail).
p5p-msgid: <Pine.SOL.3.95q.
970224170713.5700H-100000@fractal.lafayette.edu>
private-msgid: <Pine.SOL.3.95q.
970224170713.5700H-100000@fractal.lafayette.e
Tom Phoenix [Tue, 31 Dec 1996 17:59:20 +0000]
More thoroughly test rand() and srand()
Chip Salzenberg [Tue, 19 Nov 1996 00:07:00 +0000]
C<use vars> didn't work until 5.002
Tim Bunce [Mon, 24 Feb 1997 16:05:17 +0000]
Trivial patch to make ExtUtils::Install more -w clean
p5p-msgid: <
9702241605.AA17436@toad.ig.co.uk>
Roderick Schertler [Sat, 22 Feb 1997 18:03:08 +0000]
Pod::Text fixes
Here are some bug fixes for Pod::Text.
Is it a bug that text given to =for and =begin doesn't get indented? It
seems like a bug to me, but I imagine one could argue the other way so I
wanted to test opinion before fixing it.
p5p-msgid: <350.
856634588@eeyore.ibcinc.com>
Roderick Schertler [Sun, 23 Feb 1997 19:29:57 +0000]
Fix fd leak in IO::Pipe
Subject: Re: Another goof with IO::Pipe?
On Fri, 21 Feb 1997 18:51:10 -0500 (EST), Ilya Zakharevich <ilya@math.ohio-state.edu> said:
>
> IO::Pipe basically does:
> $fh = $rw ? $me->reader() : $me->writer(); # close the other end
> bless $io, "IO::Handle";
> $io->fdopen($fh, $mode);
>
> In the corresponding place IPC::Open3 adds:
>
> $fh->close;
>
> Which one is correct?
The ->close is necessary, the IO::Pipe code is leaving two copies of the
pipe in the child. It dup()s the pipe onto STDIN/STDOUT, but doesn't
then close the pipe.
Here's an fd scan of a child whose parent did ->reader:
FD Type Mode UID Inode File System FS Type rdev
0 char special 0620 roderick 686 13054,42068 dg/ux 23, 8
1 FIFO special 0000 roderick 23 900, 0 dg/ux --
2 char special 0620 roderick 686 13054,42068 dg/ux 23, 8
4 FIFO special 0000 roderick 23 900, 0 dg/ux --
fd 4 is original write side of the pipe (the read side was fd 3) which
should have been closed by this point.
p5p-msgid: <pzn2sv722y.fsf@eeyore.ibcinc.com>
Ilya Zakharevich [Fri, 21 Feb 1997 14:53:46 +0000]
Don't clobber $1 et al in debugger's DB::sub()
Chip Salzenberg [Fri, 21 Feb 1997 14:59:31 +0000]
Don't assume that sizeof(int) >= sizeof(void*)
Chip Salzenberg [Thu, 20 Feb 1997 16:52:53 +0000]
Update VMS version
Stephen Zander [Mon, 24 Feb 1997 21:24:54 +0000]
Hints for DC/OSx
Subject: Re: Perl 5.003_28 is available
Chip Salzenberg wrote:
> This is our sixth (yes, _sixth_) beta candidate. If I could give one
> piece of advice to any pumpkin holder, it would be:
Better late than never...
With the patch to hints/dcosx below 5.003_28 passes avarything
except test 9 of op/pack. This seems to be a problem with
0x
FFFFFFFF being treated as unsigned while
unpack("I",pack("I",0x
FFFFFFFF) is signed.
./perl -e 'print 0x
FFFFFFFF, "\n", unpack("I",pack("I",0xFFFFFFFF)),"\n";'
4294967295
-1
I'm assuming that gcvt is doing something strange. Any suggestions/experiences?
p5p-msgid: <
199702242124.NAA03796@wsuse5.mckesson.com>
Spider Boardman [Sat, 22 Feb 1997 09:51:48 +0000]
Ultrix hints
Subject: Re: Perl 5.003_28 is available
The hint file for ULTRIX is out of date with respect to the size
of the Perl_yylex() function. I found this after enabling -O2,
which works in recent versions of the compiler. Also, despite
the inability of the standard dist unit to detect the fact, the
ULTRIX dirent struct does too have a d_namelen field.
Patch for that attached.
Also, of course, as expected, ULTRIX still fails t/io/fs.t test
#25, since [f]truncate can't be used to extend a file on that OS.
All other tests are fine.
p5p-msgid: <
199702220951.EAA08156@Orb.Nashua.NH.US>
Chip Salzenberg [Wed, 19 Feb 1997 19:04:03 +0000]
Automatically call srand() before rand() if user didn't
Chip Salzenberg [Fri, 21 Feb 1997 16:41:00 +0000]
[differences between cumulative patch application and perl5.003_28]
Perl 5 Porters [Fri, 21 Feb 1997 16:41:00 +0000]
[inseparable changes from patch from perl5.003_27 to perl5.003_28]
CORE LANGUAGE CHANGES
Subject: Don't let C<sub foo;> undefine &foo
From: Chip Salzenberg <chip@perl.com>
Files: op.c
Subject: Make code, doc agree on $ENV{PATH} and `cmd`
From: Chip Salzenberg <chip@perl.com>
Files: pod/perlsec.pod pp_sys.c
Subject: Optimize keys() and values() in void context
From: Chip Salzenberg <chip@perl.com>
Files: doop.c op.c
CORE PORTABILITY
Subject: VMS patches post _27
Date: Thu, 20 Feb 1997 01:58:46 -0500 (EST)
From: Charles Bailey <bailey@HMIVAX.HUMGEN.UPENN.EDU>
Files: MANIFEST dosish.h hv.c lib/ExtUtils/MM_VMS.pm lib/ExtUtils/xsubpp perl.c perlsdio.h pod/perldelta.pod pod/perlvar.pod t/op/closure.t unixish.h vms/Makefile vms/descrip.mms vms/ext/filespec.t vms/genconfig.pl vms/vms.c vms/vmsish.h
private-msgid: <01IFMEMPN1IU0057E2@hmivax.humgen.upenn.edu>
Subject: Re: OS/2 patch for _27
Date: Thu, 20 Feb 1997 19:24:16 -0500 (EST)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Files: INSTALL README.os2 lib/Test/Harness.pm os2/Changes os2/OS2/PrfDB/t/os2_prfdb.t os2/os2.c os2/os2ish.h os2/perl2cmd.pl perl.c pod/perldelta.pod t/TEST t/harness t/op/magic.t
Msg-ID: <
199702210024.TAA03174@monk.mps.ohio-state.edu>
(applied based on p5p patch as commit
833d3f255ed68b969f062cec63d33f853ed9237c)
DOCUMENTATION
Subject: INSTALL updates since _26
Date: Tue, 18 Feb 1997 16:00:08 -0500 (EST)
From: Andy Dougherty <doughera@fractal.phys.lafayette.edu>
Files: INSTALL
Msg-ID: <Pine.SOL.3.95q.
970218155815.2014F-100000@fractal.lafayette.e
(applied based on p5p patch as commit
a8247d96fd6167a3b920e63aedee5592cd6e29a7)
Subject: Document "$$0" change
From: Chip Salzenberg <chip@perl.com>
Files: pod/perldelta.pod
Subject: Don't recommend impossible //o for C<$x =~ $y>
From: Chip Salzenberg <chip@perl.com>
Files: pod/perlop.pod
Subject: Correct doc that claimed that <FH> was never false
From: Chip Salzenberg <chip@perl.com>
Files: pod/perldelta.pod pod/perlop.pod
Subject: Document C<$?> vs. $SIG{CHLD}
From: Ulrich Pfeifer <pfeifer@charly.informatik.uni-dortmund.de>
Files: pod/perlvar.pod
Subject: Add pumpkin.pod
From: Chip Salzenberg <chip@perl.com>
Files: MANIFEST Porting/pumpkin.pod
Subject: Don't say "associat*ve arr*y"
From: Chip Salzenberg <chip@perl.com>
Files: MANIFEST gv.h hv.c lib/Env.pm lib/overload.pm opcode.pl pod/perl.pod pod/perldelta.pod pod/perldiag.pod pod/perlfunc.pod pod/perlguts.pod pod/perlmod.pod pod/perltie.pod pod/perltoc.pod pod/perltrap.pod x2p/a2p.pod
OTHER CORE CHANGES
Subject: Fix a typo
From: Chip Salzenberg <chip@perl.com>
Files: pp_sys.c
Subject: Fix perl_call_sv(..., G_NOARGS)
From: Chip Salzenberg <chip@perl.com>
Files: perl.c
Subject: Fix SIGSEGV when cloning sub with complex expression
From: Chip Salzenberg <chip@perl.com>
Files: op.c
Andy Dougherty [Tue, 18 Feb 1997 21:00:08 +0000]
INSTALL updates since _26
Here are my proposed updates to the INSTALL file. Comments are welcome.
(I'd particularly welcome an AIX bos.* blurb from someone who understand
those things.)
It's not all that much, really; just some sections have been moved around
into what I hope is a more logical order.
p5p-msgid: <Pine.SOL.3.95q.
970218155815.2014F-100000@fractal.lafayette.edu>
private-msgid: <Pine.SOL.3.95q.
970218155815.2014F-100000@fractal.lafayette.e
Hugo van der Sanden [Tue, 18 Feb 1997 11:05:59 +0000]
Add test for grep() and wantarray
Subject: Re: subscript when wantarray
In <
199702172207.RAA20755@rio.atlantic.net>, Chip Salzenberg writes:
:First, here's a patch that makes the first case print "2"
[...]
:Second, the reason that it prints "2" instead of "3" is that
:C<(undef)[0]> evaluates to the empty list in an array context,
:but to C<undef> in a scalar context. This is on purpose.
Yep, I expected that.
Patch worked fine here; below is an extra test using the same code as my
original example.
p5p-msgid: <
199702181105.LAA17895@tyree.iii.co.uk>
Ilya Zakharevich [Fri, 21 Feb 1997 00:24:16 +0000]
Re: OS/2 patch for _27
The bulk of this patch is README.os2 which has a lot of formatting
stripped (it is at the end). The rest (340 lines of -u) is quite simple.
INSTALL
t/harness documented.
lib/Test/Harness.pm
Can test files with -T on hash-bang line.
Gets more info from tests which coredump.
Will not try to load Devel::CoreStack on each test.
os2/Changes
Updated.
os2/OS2/PrfDB/t/os2_prfdb.t
File closed.
os2/os2.c
Poor-man's setup of environment if DLL run from non-conforming EXE.
os2/os2ish.h
Poor-man's setup of environment if DLL run from non-conforming EXE.
ALTERNATE_SHEBANG defined.
os2/perl2cmd.pl
Updated for ALTERNATE_SHEBANG.
perl.c
If ALTERNATE_SHEBANG defined, just ignore -S on the starting
line. (It might be already used by OS to start perl.)
pod/perldelta.pod
Malloc's PERL_DEBUG_MSTATS, EMERGENCY_SBR, PACK_MALLOC,
TWO_POT_OPTIMIZE documented.
t/harness
Does not require -I../lib any more.
t/op/magic.t
Now works under OS/2 with both tests.
t/TEST
Adds an advice to use
perl harness
to get better granularity.
README.os2
A lot of markup stripped (in the hope that smart converters
will add it back), minor updates.
p5p-msgid: <
199702210024.TAA03174@monk.mps.ohio-state.edu>
Chip Salzenberg [Mon, 17 Feb 1997 22:29:53 +0000]
Tweak tests to notice $dont_use_nlink
Chip Salzenberg [Mon, 17 Feb 1997 22:22:49 +0000]
Don't run locale test if -DNO_LOCALE
Chip Salzenberg [Wed, 19 Feb 1997 22:43:28 +0000]
Include 'study' in regexp.t
Graham Barr [Tue, 18 Feb 1997 07:48:40 +0000]
prototype error in File::stat
The prototype for lstat in File::stat is wrong. It stats that is expects
a GLOB. It should be expecting a scalar, the same as File::stat::stat.
p5p-msgid: <
199702180748.HAA14151@ultra-boy>
Chip Salzenberg [Tue, 18 Feb 1997 00:25:45 +0000]
Remove redundant clearerr() from IO::Seekable
Chip Salzenberg [Fri, 21 Feb 1997 16:41:00 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Chip Salzenberg [Fri, 21 Feb 1997 16:41:00 +0000]
Avoid $` $& $' in libraries
(this is the same change as commit
2724d5068a405436d3f2cd00aeb8f7b460b24fec, but as applied)
Ilya Zakharevich [Fri, 21 Feb 1997 02:07:30 +0000]
Avoid $` $& $' in libraries
Subject: Pessimal $` $& $' in libraries
Corrected:
lib/diagnostics.pm
lib/Getopt/Long.pm
lib/Pod/Text.pm
os2/OS2/REXX/REXX.pm
$' $` $& eliminated from all the known *.pm.
Btw, by this I proudly introduce usage of $+ in standard perl modules ;-).
p5p-msgid: <
199702210207.VAA03560@monk.mps.ohio-state.edu>
Ilya Zakharevich [Fri, 21 Feb 1997 07:37:59 +0000]
Debugger patch
private-msgid: <
199702210737.CAA03951@monk.mps.ohio-state.edu>
Chip Salzenberg [Tue, 18 Feb 1997 01:29:35 +0000]
Fix the Tolkein quote
Ilya Zakharevich [Fri, 21 Feb 1997 02:44:13 +0000]
Minor update to malloc.c
This change of defines may satisfy Nick in his future life, when he
needs an effecient malloc() with DEBUGGING perl again:
p5p-msgid: <
199702210244.VAA03676@monk.mps.ohio-state.edu>
Chip Salzenberg [Tue, 18 Feb 1997 00:20:07 +0000]
Undo signal patch -- it broke die() in signal
Keith Neufeld [Tue, 18 Feb 1997 04:49:16 +0000]
No version of AIX has working setre[ug]id()
Jarkko Hietaniemi [Tue, 18 Feb 1997 02:25:14 +0000]
New hints for Digital UNIX
Ilya Zakharevich [Tue, 18 Feb 1997 21:22:53 +0000]
When overloading, don't throw away nomethod's value
Chip Salzenberg [Tue, 18 Feb 1997 21:08:02 +0000]
Turn off 'expression tainted' flag at end of runops()
Chip Salzenberg [Wed, 19 Feb 1997 21:58:01 +0000]
Don't taint $x in C<$x = ($tainted =~ /(\w+)/)>
Chip Salzenberg [Fri, 21 Feb 1997 16:41:00 +0000]
[shell changes from patch from perl5.003_27 to perl5.003_28]
Change from running these commands:
# VMS no longer uses 'make' to build Perl
rm -f vms/Makefile
# ready to patch
exit 0
Chip Salzenberg [Tue, 18 Feb 1997 01:22:00 +0000]
[differences between cumulative patch application and perl5.003_27]
Perl 5 Porters [Tue, 18 Feb 1997 01:22:00 +0000]
[inseparable changes from patch from perl5.003_26 to perl5.003_27]
BUILD PROCESS
Subject: Fix eval "" in Configure
Date: Fri, 14 Feb 1997 13:09:53 -0500
From: John L. Allen <allen@gateway.grumman.com>
Files: Configure
Subject: Re: Configure problem on IRIX - me dumb
p5p-msgid: <
9702141809.AA17001@gateway.grumman.com>
Subject: Don't link with -lsfio if sfio is not requested
From: Chip Salzenberg <chip@perl.com>
Files: Configure
Subject: perl5.003_26 Configure change "win" for AIX 4
Date: Fri, 14 Feb 1997 13:59:02 -0600 (CST)
From: Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu>
Files: Configure
p5p-msgid: <Pine.OSF.3.95.
970214135751.32654A-100000@dogbert.cc.ndsu.NoDak.edu>
private-msgid: <Pine.OSF.3.95.
970214135751.32654A-100000@dogbert.cc.ndsu.NoD
CORE LANGUAGE CHANGES
Subject: Better looks_like_number() function [sv.c]
Date: Fri, 14 Feb 1997 18:08:52 +0100
From: Gisle Aas <aas@bergen.sn.no>
Files: sv.c
Msg-ID: <
199702141708.SAA17546@bergen.sn.no>
(applied based on p5p patch as commit
8dbaa58ee2aba7cc22d84199a674c58bbf108b46)
Subject: Remove redundant functions UNIVERSAL::{class,is_instance}
Date: 14 Feb 1997 15:52:21 +0000
From: Gisle Aas <aas@bergen.sn.no>
Files: pod/perldelta.pod pod/perlobj.pod t/op/universal.t universal.c
Msg-ID: <hwwsbpeq2.fsf@bergen.sn.no>
(applied based on p5p patch as commit
77bb9b23081b62119e8fbe9f5655b8802e4537ae)
Subject: Allow C<setpgrp $$>
Date: 16 Feb 1997 23:19:12 -0500
From: Roderick Schertler <roderick@gate.net>
Files: pp_sys.c
Msg-ID: <pzraigyshr.fsf@eeyore.ibcinc.com>
(applied based on p5p patch as commit
3d2573a84a1aa655d5da58c57b3fc20e04d40f9f)
Subject: Fix syntax error on C<&$1>
From: Chip Salzenberg <chip@perl.com>
Files: toke.c
Subject: Fix grep() with refs in array context
From: Chip Salzenberg <chip@perl.com>
Files: pp.c
CORE PORTABILITY
Subject: Eliminate $^S; add C<use vmsish qw(status exit time)>
Date: Mon, 17 Feb 1997 02:45:26 -0500 (EST)
From: Charles Bailey <bailey@HMIVAX.HUMGEN.UPENN.EDU>
Files: MANIFEST gv.c lib/English.pm lib/ExtUtils/MM_VMS.pm lib/ExtUtils/Mksymlists.pm lib/ExtUtils/xsubpp mg.c op.c perl.c perl.h pod/perldelta.pod pod/perlmod.pod pod/perlvar.pod pp_ctl.c pp_sys.c utils/perldoc.PL vms/Makefile vms/config.vms vms/descrip.mms vms/ext/Stdio/Stdio.pm vms/ext/Stdio/Stdio.xs vms/ext/XSSymSet.pm vms/ext/vmsish.pm vms/vms.c vms/vmsish.h win32/makedef.pl
private-msgid: <01IFI9CFKL0S004R2V@hmivax.humgen.upenn.edu>
LIBRARY AND EXTENSIONS
Subject: Remove Fatal.pm
From: Chip Salzenberg <chip@perl.com>
Files: MANIFEST lib/Fatal.pm pod/perldelta.pod pod/perlmod.pod pod/roffitall t/lib/fatal.t
Subject: Refresh MakeMaker to 5.40
From: Andy Dougherty <doughera@lafcol.lafayette.edu>
Files: lib/ExtUtils/Liblist.pm lib/ExtUtils/MM_Unix.pm lib/ExtUtils/MakeMaker.pm lib/ExtUtils/Mksymlists.pm
OTHER CORE CHANGES
Subject: Fix core dump when embedding
From: Chip Salzenberg <chip@perl.com>
Files: perl.c
Subject: Re: Fragile signals
Date: Thu, 13 Feb 1997 01:44:39 -0500 (EST)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Files: mg.c
Msg-ID: <
199702130644.BAA07572@monk.mps.ohio-state.edu>
(applied based on p5p patch as commit
09df8c7df7dfc9853902f1fdd8a6d95f53be66fc)
Subject: Make format strings correspond exactly to parameters
Date: 13 Feb 1997 17:24:31 -0500
From: Roderick Schertler <roderick@gate.net>
Files: doio.c ext/DB_File/DB_File.xs ext/Opcode/Opcode.xs gv.c op.c perl.c pp_ctl.c pp_sys.c regcomp.c toke.c
Msg-ID: <pz7mkc1h0g.fsf@eeyore.ibcinc.com>
(applied based on p5p patch as commit
bf81aadd817bdea29720b072eef945df2da8463b)
Subject: Don't try to attach 'o' magic to read-only values
From: Chip Salzenberg <chip@perl.com>
Files: sv.c
Subject: Fix carriage-return message
From: Chip Salzenberg <chip@perl.com>
Files: toke.c
Subject: In <=>, test for equality first
From: Chip Salzenberg <chip@perl.com>
Files: pp.c
Subject: Don't mark sv_{true,false} PADTMP
From: Chip Salzenberg <chip@perl.com>
Files: op.c
Roderick Schertler [Thu, 13 Feb 1997 22:24:31 +0000]
Make format strings correspond exactly to parameters
Subject: Re: Hereis weirdness in 5.003_26
On Thu, 13 Feb 1997 15:13:18 -0500 (EST), Chip Salzenberg <chip@atlantic.net> said:
> According to Ilya Zakharevich:
>>
>> Why was not it catched by gcc prototypes?
>
> I don't use gcc. Does anyone who uses gcc compile with -Wproto ?
-Wformat, you mean. I wasn't previously, but I will in the future. It
turned up a few bugs (some universal, some which would trigger only
where I32 != int or the like). I think -Wformat should be added to
CFLAGS automatically when appropriate.
p5p-msgid: <pz7mkc1h0g.fsf@eeyore.ibcinc.com>
Chip Salzenberg [Mon, 17 Feb 1997 05:58:57 +0000]
Fix formatting in perldiag
Ilya Zakharevich [Thu, 13 Feb 1997 06:44:39 +0000]
Re: Fragile signals
Ilya Zakharevich writes:
>
> Some messages of mine get stuck in the chain:
>
> Nico => Debian => floor
>
> I repost them now:
>
> As you may remember, the following program
>
> #!/usr/bin/perl
> sub foo { ++$a }
> $SIG{ALRM} = sub { $run = 0 ; return };
> while (1) {
> $a = 0;
> $run = 1;
> alarm 1;
> foo while $run;
> print $i++, ": $a\n";
> }
> die "Not reached";
> __END__
>
> will not call malloc/realloc/free in sighandler, but nevertheless it
> would not run more than approx. 25 iterations. This shows that 4% of
> subroutine-calling frame is not signal-safe.
>
> The following patch makes it run approx. twice as long (on _17) (thus
Well, the following patch makes it run 4 times as long. The only
failure mode I have seen was silent walking off runops().
Here is the statistics (number of iterations before failure):
Before:
23, 24, 14, 22, 19, 13, 30, 11
(4 core dumps, 2 attempts to free unref scalar, 1 panic:
leave-scope, 1 silent)
After:
42, 936, 63, 131, 280.
(all silent)
It looks like handling of internal stacks needs a lot of
improvement. What this patch does is only a symptomatic treatment, but
it may go into 004 nevertheless (especially if one can see why op
becomes 0 at some moment). I tried saving/restoring op in the handler,
but this does not help...
p5p-msgid: <
199702130644.BAA07572@monk.mps.ohio-state.edu>
Tim Bunce [Mon, 10 Feb 1997 19:00:59 +0000]
Update To-Do list
Subject: Re: Perl Futures (was: Tentative 5.005 and 5.006 schedules)
p5p-msgid: <
9702101900.AA25293@toad.ig.co.uk>
Roderick Schertler [Mon, 17 Feb 1997 04:19:12 +0000]
Allow C<setpgrp $$>
Subject: Re: Perl question, re: POSIX setpgrp
On Fri, 14 Feb 1997 16:31:53 GMT, Chris Vo <chrisv@on.bell.ca> said:
>
>>> POSIX setpgrp can't take an argument at ./check_ntp line 21.
>
> where line 21 reads as:
> setpgrp (0, $$);
Replace that with just
setpgrp;
This does the same thing and will work on all systems.
I think there's a bug here. Perl is expecting a pgrp arg of 0 to mean
$$, but it doesn't allow an explicit $$. Even this seems a little odd,
as on neither system I've got available at the moment is it mentioned
that a 0 pgrp means getpid() (they both mention that a 0 pid means
getpid(), though).
p5p-msgid: <pzraigyshr.fsf@eeyore.ibcinc.com>
Ulrich Pfeifer [Mon, 10 Feb 1997 17:38:45 +0000]
pod2man: missing '-' in name section shouldn't be fatal
Subject: pod2man: missing '-' in name section is fatal
p5p-msgid: <yfmzpxcimsa.fsf@ls6.informatik.uni-dortmund.de>
Gisle Aas [Fri, 14 Feb 1997 15:52:21 +0000]
Remove redundant functions UNIVERSAL::{class,is_instance}
Nick Ing-Simmons <nik@tiuk.ti.com> writes:
> Loose them!
p5p-msgid: <hwwsbpeq2.fsf@bergen.sn.no>
Chip Salzenberg [Mon, 17 Feb 1997 05:59:16 +0000]
Remove non-portable locale tests
Gisle Aas [Fri, 14 Feb 1997 17:08:52 +0000]
Better looks_like_number() function [sv.c]
The old looks_like_number() function was too forgiving. It classified
strings like these as numbers:
"+"
"+e+"
"e3"
"."
With this patch perl -lwe 'print int(".")' prints:
Argument "." isn't numeric in int at -e line 1.
0
as it should and perl still passes all its tests.
p5p-msgid: <
199702141708.SAA17546@bergen.sn.no>
Andreas Koenig [Mon, 17 Feb 1997 05:59:16 +0000]
Refresh Test::Harness to 1.15
Andreas Koenig [Mon, 17 Feb 1997 05:59:13 +0000]
Refresh CPAN.pm to 1.21
Chip Salzenberg [Mon, 17 Feb 1997 05:58:59 +0000]
Update os2/diff.configure
Chip Salzenberg [Mon, 17 Feb 1997 05:58:59 +0000]
Update Plan 9, Win32, VMS configs with $shortsize and $longsize
Chip Salzenberg [Tue, 18 Feb 1997 01:22:00 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Chip Salzenberg [Tue, 18 Feb 1997 01:22:00 +0000]
Irix6.4 (with 7.1 compilers)
(this is the same change as commit
57d51fe74c853a68e4abef82668e025a797ee752, but as applied)
John Stoffel [Thu, 13 Feb 1997 02:38:51 +0000]
Irix6.4 (with 7.1 compilers)
Subject: Irix6.4 (with 7.1 compilers) passes 100% with patch
>>>>> "John" == John Stoffel <jfs@fluent.com> writes:
John> I had to apply the following patch to the Irix_6_4.sh hints file to
John> get it to compile under Irix 6.4 with the version 7.1 compilers
John> installed. Thanks to Hallvard Furuseth for pointing out that the
John> pp_sys_cflags line _has_ to be in the left most column. This leads me
John> to believe that the following hints files will have problems:
Ooops! I messed up on the patch for Irix_6_2.sh, so here's the
entire patch again to fix the problem as reported under Irix 6.[24].
I don't have a 6.3 machine to test with right now, since I have to
install the compilers first. More when I get a chance.
p5p-msgid: <
199702130238.VAA24468@jfs.Fluent.COM>
Chip Salzenberg [Tue, 18 Feb 1997 01:22:00 +0000]
[dummy merge]
This merge exists so that the p5p version of the patch and the applied
version are both in the history
Chip Salzenberg [Tue, 18 Feb 1997 01:22:00 +0000]
Digital UNIX hints
(this is the same change as commit
bcbb37e68c8af7f356dd1bf2b25fd24f46baffe3, but as applied)
Jarkko Hietaniemi [Sat, 15 Feb 1997 19:06:33 +0000]
Digital UNIX hints
Subject: Re: Digital UNIX style patch
p5p-msgid: <
199702181749.TAA08255@alpha.hut.fi>
private-msgid: <
199702151906.VAA22999@alpha.hut.fi>
Chip Salzenberg [Mon, 17 Feb 1997 05:59:02 +0000]
Eliminate FP exceptions under SCO 5
Chip Salzenberg [Mon, 10 Feb 1997 22:11:59 +0000]
Fix sub call through magic var (e.g. C<&$1>)
Chip Salzenberg [Tue, 18 Feb 1997 01:22:00 +0000]
[shell changes from patch from perl5.003_26 to perl5.003_27]
Change from running these commands:
# this never worked anyway
rm -f lib/Fatal.pm
rm -f t/lib/fatal.t
# ready to patch
exit 0
Perl 5 Porters [Mon, 10 Feb 1997 19:29:00 +0000]
[inseparable changes from patch from perl5.003_25 to perl5.003_26]
CORE LANGUAGE CHANGES
Subject: Make \r in script an error (per Larry)
From: Chip Salzenberg <chip@perl.com>
Files: pod/perldiag.pod toke.c
CORE PORTABILITY
Subject: VMS patches post _25
Date: Fri, 07 Feb 1997 01:56:12 -0500 (EST)
From: Charles Bailey <bailey@HMIVAX.HUMGEN.UPENN.EDU>
Files: Porting/Glossary lib/ExtUtils/Liblist.pm lib/ExtUtils/MM_VMS.pm lib/ExtUtils/xsubpp perl.c vms/Makefile vms/config.vms vms/descrip.mms vms/genconfig.pl vms/perlvms.pod vms/vms.c vms/vmsish.h x2p/a2p.c
private-msgid: <01IF48W3P39W0050BD@hmivax.humgen.upenn.edu>
LIBRARY AND EXTENSIONS
Subject: Make diagnostics module strip formatting directives
From: Chip Salzenberg <chip@perl.com>
Files: lib/diagnostics.pm pod/perldiag.pod
OTHER CORE CHANGES
Subject: Fix (yet another) Tk closure problem
From: Chip Salzenberg <chip@perl.com>
Files: op.c perl.c pp_ctl.c
Subject: Fix value of C<foreach>
From: Chip Salzenberg <chip@perl.com>
Files: cop.h pp_ctl.c
Subject: Refine 'runaway string' heuristic
From: Chip Salzenberg <chip@perl.com>
Files: toke.c
Subject: Fix core dump on C<print "a", last> in eval
From: Chip Salzenberg <chip@perl.com>
Files: pp_ctl.c
Paul Marquess [Tue, 4 Feb 1997 09:41:53 +0000]
Describe interation of untie and DESTROY
Signed-off-by: Paul Marquess <pmarquess@bfsec.bt.co.uk>
John L. Allen [Thu, 6 Feb 1997 15:07:28 +0000]
return *FH pod patch
p5p-msgid: <
9702061507.AA04474@gateway.grumman.com>