Ilya Zakharevich [Sun, 1 Dec 1996 11:07:19 +0000]
_10+ under OS/2
_10+ under OS/2 is almost clean: below is a minor patch to malloc.c
which makes compilation of x2p possible with -DEMERGENCY_SBRK in
ccflags, and the patch which reduces os2/diff.configure twice:
p5p-msgid: <
199612011107.GAA10805@monk.mps.ohio-state.edu>
Jeff McDougal [Tue, 3 Dec 1996 13:33:05 +0000]
BUG in hints/hpux.sh
Tom, I was not able to run the Configure script on my new HP D-server running HP-UX 10.01.
The problem was a single line in the hints/hpux.sh file:
xxcontext=`grep $(printf %#x $(getconf CPU_VERSION)) /usr/include/sys/unistd.h`
I commented this line out and inserted the following to get a clean execution under the
Bourne shell:
DECCPU=`getconf CPU_VERSION`
HEXCPU=`printf %#x ${DECCPU}`
xxcontext=`grep $HEXCPU /usr/include/sys/unistd.h`
I am very happy to have Perl 5.003 on my new development box.
Jeff McDougal
p5p-msgid: <
32A42C11.7FA2@cris.com>
Roderick Schertler [Fri, 29 Nov 1996 23:20:16 +0000]
hints/dgux.sh update
Here's an update for the dgux hints file.
p5p-msgid: <24178.
849309616@eeyore.ibcinc.com>
Ilya Zakharevich [Thu, 28 Nov 1996 11:50:58 +0000]
Re: 5.003_09: PADTMP fix
Nick Ing-Simmons writes:
> >Here is the better fix: We mark PADTMPs as STEALABLE when they are
> >created, and unset this if they are put on fake array @_:
> >
> >Nick, please check it with Tk, this fixes the test case you sent.
> >
> >A much simplified test case is included into op/misc.t.
>
> Seems to work here. My other testcases and Tk continue to work.
>
> I still don't really understand what PADTEMPs are - care to explain
> goal of the enhancement?
Chip, please apply this:
p5p-msgid: <
199611281150.GAA06884@monk.mps.ohio-state.edu>
Chip Salzenberg [Thu, 5 Dec 1996 11:02:11 +0000]
Less malloc in magic
Chip Salzenberg [Fri, 6 Dec 1996 06:56:00 +0000]
[shell changes from patch from perl5.003_10 to perl5.003_11]
Change from running these commands:
# new directories for new modules
test -d lib/Class || mkdir lib/Class
test -d lib/User || mkdir lib/User
# get rid of old extension
rm -rf ext/FileHandle
exit 0
Chip Salzenberg [Fri, 29 Nov 1996 17:31:00 +0000]
[differences between cumulative patch application and perl5.003_10]
Perl 5 Porters [Fri, 29 Nov 1996 17:31:00 +0000]
[inseparable changes from patch from perl5.003_09 to perl5.003_10]
CORE LANGUAGE CHANGES
Subject: Allow &{sub {...}} without warning
From: Chip Salzenberg <chip@atlantic.net>
Files: toke.c
Subject: Make parens optional on [gs]ethost and [gs]et{pw,gr} function
From: Chip Salzenberg <chip@atlantic.net>
Files: toke.c
Subject: Fix syntax error with "$x [0]" and "$x {y}" and "@x {y}"
From: Chip Salzenberg <chip@atlantic.net>
Files: toke.c
DOCUMENTATION
Subject: Improve documentation for sysread() and syswrite()
From: Chip Salzenberg <chip@atlantic.net>
Files: pod/perlfunc.pod
Subject: Document how to use $SIG{ALRM} and alarm()
Date: Tue, 26 Nov 1996 11:42:49 -0500
From: Roderick Schertler <roderick@ibcinc.com>
Files: pod/perlfunc.pod
Msg-ID: <5898.
849026569@eeyore.ibcinc.com>
(applied based on p5p patch as commit
5fa5e7dfc2abaaadd377c97cd1ebe78ea844da88)
OTHER CORE CHANGES
Subject: Hash key memory corruption fix and naming cleanup
From: Chip Salzenberg <chip@atlantic.net>
Files: hv.c hv.h perl.h
Subject: Undo broken perf. patch (PADTMP stealing)
From: Chip Salzenberg <chip@atlantic.net>
Files: sv.c
Subject: Make SV unstudied in sv_gets()
From: Chip Salzenberg <chip@atlantic.net>
Files: sv.c
Subject: Better support for UVs
From: Paul Marquess <pmarquess@bfsec.bt.co.uk>
Files: global.sym old_global.sym perl.h pp.c pp.h proto.h sv.c sv.h
Subject: Minor locale cleanups
From: Chip Salzenberg <chip@atlantic.net>
Files: t/lib/posix.t util.c
Accept "POSIX" locale as standard like "C". Reset locale to
'C' when testing strtod() in t/lib/posix.t.
Subject: Always taint result of sprintf() on float
From: Chip Salzenberg <chip@atlantic.net>
Files: doop.c
Subject: Fix spurious warning from bitwise string ops
From: Chip Salzenberg <chip@atlantic.net>
Files: doop.c
Subject: Eliminate warning on {,sys}read(,$newvar,)
From: Chip Salzenberg <chip@atlantic.net>
Files: doop.c pp_sys.c
Subject: Namespace cleanup
From: Chip Salzenberg <chip@atlantic.net>
Files: global.sym old_global.sym perl.h
Subject: Modify perl_exp.SH; create old_perl_exp.SH; document old_*
From: Chip Salzenberg <chip@atlantic.net>
Files: Configure INSTALL MANIFEST old_perl_exp.SH perl_exp.SH
PORTABILITY
Subject: Reliable signal patch
Date: Tue, 26 Nov 1996 05:40:50 -0500 (EST)
From: Kenneth Albanowski <kjahds@kjahds.com>
Files: global.sym mg.c old_global.sym perl.h pp_sys.c proto.h util.c
Msg-ID: <Pine.LNX.3.93.
961126053209.294J-100000@kjahds.com>
(applied based on p5p patch as commit
679728958e74b0ccd6d61567d84851f1ef994e1f)
Subject: Emulate missing flock() with either fcntl() or lockf()
From: Chip Salzenberg <chip@atlantic.net>
Files: pp_sys.c
Subject: 3_09: minor patches for OS/2
Date: Wed, 27 Nov 1996 03:30:05 -0500 (EST)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Files: doio.c global.sym malloc.c old_global.sym os2/Makefile.SHs os2/OS2/ExtAttr/Makefile.PL os2/OS2/PrfDB/Makefile.PL os2/OS2/Process/Makefile.PL os2/OS2/REXX/Makefile.PL os2/os2.c os2/os2ish.h perl.h
Subject: 3_09: minor patches
This patches mostly enable commpilation under OS/2, and fix malloc.c.
Enjoy,
p5p-msgid: <
199611270830.DAA04985@monk.mps.ohio-state.edu>
Subject: Re: 5.003_09 and QNX
Date: Wed, 27 Nov 96 13:36:06 est
From: Norton Allen <nort@bottesini.harvard.edu>
Files: Configure MANIFEST README.qnx hints/qnx.sh qnx/ar qnx/cpp t/TEST toke.c util.c x2p/proto.h
Msg-ID: <
9611271836.AA14460@bottesini.harvard.edu>
(applied based on p5p patch as commit
c5117498be098729dc2af28089bd130c88c8d42b)
Roderick Schertler [Tue, 26 Nov 1996 16:42:49 +0000]
Document how to use $SIG{ALRM} and alarm()
Subject: Re: reliable signal patch (was Re: Reliable signals)
On Tue, 26 Nov 1996 11:14:03 -0500 (EST), Kenneth Albanowski <kjahds@kjahds.com> said:
>
> Thus with restarting being the default, you can restart, if you want,
> by returning normally from the signal handler, or interrupt by dying.
Thanks, now I get it. This is even inferred in perlipc(1). Here's some
more explicit documentation.
p5p-msgid: <5898.
849026569@eeyore.ibcinc.com>
Chip Salzenberg [Thu, 28 Nov 1996 22:14:24 +0000]
Diagnostic cleanup
Norton Allen [Wed, 27 Nov 1996 18:36:06 +0000]
Re: 5.003_09 and QNX
[editor's note: original messages may not have been delivered:
the included message may be the only remaining trace.
see note on <
9611271732.AA17020@bottesini.harvard.edu>:
--- |3
[ There was a set of quite minor patches posted on 11/20; I got
them, but who knows who else did, the list being what it is...
I am also nearly ready to submit a QNX hints file together with
a couple auxilliary support files.]
...
]
Subject: QNX Port
Greetings,
QNX is now compiling 5.003_08 successfully and passing all
tests except for two which have been identified as library
errors and are in the process of being fixed by the vendor.
I'd like to submit a few small patches which makes this
work. The patches are accompanied by text describing their
motivation.
If I am moving too fast, please tell me slow down. If you have
suggestions for other/better approaches, please let me know.
I will follow up with a hints file and a (possibly more
controversial) Configure change.
-Norton Allen
Under QNX, I need to include <sys/wait.h> in util.c to prototype
waitpid(). This appears straightforward to me, but if it causes
trouble, it could be made OS-dependent.
This patch is against perl5.003_08
p5p-msgid: <
9611201749.AA11327@bottesini.harvard.edu>
private-msgid: <
9611271836.AA14460@bottesini.harvard.edu>
Nick Ing-Simmons [Sat, 23 Nov 1996 19:54:52 +0000]
AutoLoader::AUTOLOAD optimization
I notice that although *.al are right alongside autosplit.ix the
trick AutoLoader::import uses to speed looking for the latter
is not used to find the *.al files - anyone know why?
On NFS mounted file systems at least, all the failed stats/opens are expensive.
(Sadly I am testing this at home where everything is local...)
p5p-msgid: <
199611231954.TAA09921@ni-s.u-net.com>
Kenneth Albanowski [Tue, 26 Nov 1996 10:40:50 +0000]
Reliable signal patch
Subject: reliable signal patch (was Re: Reliable signals)
On Tue, 26 Nov 1996, Kenneth Albanowski wrote:
> > I found that 2 places in perl still use unreliable signals:
> > my_pclose() in util.c
> > pp_system() in pp_sys.c
> >
> > Is this working as designed, or an omission?
> I guess a new set of signal macros are needed, to reliably save and
> restore signal settings.
Here is a patch that accomplishes this. It replaces the rsignal
construction in mg.c with a generalized set of functions: rsignal,
rsignalsave, rsignalrestore, and rsignalstate.
Note that global.sym is patched to accomodate the new globally accessible
functions. They probably will be of use to XSUBs too.
This was diffed from a slightly messy 5.003_08, but should apply
acceptably over any 5.003_* series.
p5p-msgid: <Pine.LNX.3.93.
961126053209.294J-100000@kjahds.com>
Chip Salzenberg [Fri, 29 Nov 1996 17:31:00 +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, 29 Nov 1996 17:31:00 +0000]
Minor patch to debugger
(this is the same change as commit
9e0ff5ef23e91404cf40a8d53c9c0dcd28c6448f, but as applied)
Ilya Zakharevich [Fri, 29 Nov 1996 05:33:49 +0000]
Minor patch to debugger
This makes the new behaviour a little bit more transparent (now you
need q to quit), and avoids one malloc/free pair in the signal handler.
p5p-msgid: <
199611290533.AAA08053@monk.mps.ohio-state.edu>
Charles Bailey [Tue, 26 Nov 1996 22:28:23 +0000]
Re: updated patch on the sysread, syswrite for VMS
Subject: Re: updated patch on the sysread, syswrite
> Based on feedback I got, here is an updated patch. I would appreciate
> even more feedback as I added a t/op test that exercises the sysread,
> syswrite. Especially appreciated experiences on non-UNIX systems like
> OS/2, VMS, we'll see how portable my very UNIXish ideas of what those
> two calls should do are... to be applied on top of 3_08.
Looks OK under VMS. Appended is a patch to the test, which works around
the fact that file size as returned by -s isn't updated until a file is
closed.
p5p-msgid: <01ICB648K2XG001A1D@hmivax.humgen.upenn.edu>
Chip Salzenberg [Thu, 28 Nov 1996 22:14:29 +0000]
Save message when calling __DIE__ hook
Chip Salzenberg [Thu, 28 Nov 1996 22:14:29 +0000]
Don't call fcntl(fileno(rsfp)) if !rsfp
Chip Salzenberg [Thu, 28 Nov 1996 22:15:01 +0000]
Fix regex matching of chars with high bit set
Chip Salzenberg [Fri, 29 Nov 1996 17:31:00 +0000]
[shell changes from patch from perl5.003_09 to perl5.003_10]
Change from running these commands:
# We'll create new scripts, but patch won't automatically make them
# executable.
test -d qnx || mkdir qnx
touch qnx/ar qnx/cpp
chmod +x qnx/ar qnx/cpp
touch old_perl_exp.SH
chmod +x *.SH *.pl
exit 0
Chip Salzenberg [Tue, 26 Nov 1996 08:48:00 +0000]
[differences between cumulative patch application and perl5.003_09]
Perl 5 Porters [Tue, 26 Nov 1996 08:48:00 +0000]
[inseparable changes from patch from perl5.003_08 to perl5.003_09]
CORE LANGUAGE CHANGES
Subject: Lexical locales
From: Chip Salzenberg <chip@atlantic.net>
Files: too many to list
make effectiveness of locales depend on C<use locale>
Subject: Lexical scoping cleanup
From: Chip Salzenberg <chip@atlantic.net>
Files: many... but mostly perly.y and toke.c
tighten scoping of lexical variables, somewhat on the
new constructs and somewhat on the old
Subject: memory corruption / security bug in sysread,syswrite + patch
Date: Mon, 25 Nov 1996 21:46:31 +0200 (EET)
From: Jarkko Hietaniemi <jhi@cc.hut.fi>
Files: MANIFEST pod/perldiag.pod pod/perlfunc.pod pp_sys.c t/op/sysio.t
Msg-ID: <
199611251946.VAA30459@alpha.hut.fi>
(applied based on p5p patch as commit
d7090df90a9cb89c83787d916e40d92a616b146d)
DOCUMENTATION
Subject: perldiag documentation patch.
Date: Wed, 20 Nov 96 16:07:28 GMT
From: Paul Marquess <pmarquess@bfsec.bt.co.uk>
Files: pod/perldiag.pod
private-msgid: <
9611201607.AA12729@claudius.bfsec.bt.co.uk>
Subject: a missing perldiag entry
Date: Thu, 21 Nov 1996 15:24:02 -0500
From: Gurusamy Sarathy <gsar@engin.umich.edu>
Files: pod/perldiag.pod
private-msgid: <
199611212024.PAA15758@aatma.engin.umich.edu>
Subject: perlfunc patch
Date: Wed, 20 Nov 96 14:04:08 GMT
From: Paul Marquess <pmarquess@bfsec.bt.co.uk>
Files: pod/perlfunc.pod
Following on from the patch to make uc, lc etc default to $_ (as per
Camel II), here is a followup patch to perlfunc that documents the
change. I think I have documented all the other cases where $_
defaulting works as well.
p5p-msgid: <
9611201404.AA12477@claudius.bfsec.bt.co.uk>
OTHER CORE CHANGES
Subject: Properly prototype safe{malloc,calloc,realloc,free}.
From: Chip Salzenberg <chip@atlantic.net>
Files: proto.h
Subject: UnixWare 2.1 fix for perl5.003_08 - cope with fp->_cnt < -1, allow debugging
Date: Wed, 20 Nov 1996 14:27:06 +0100
From: John Hughes <john@AtlanTech.COM>
Files: sv.c
UnixWare 2.1 has no fp->_base so most of the debugging stuff in sv_gets just
core dumps.
Also, for some unknown reason fp->_cnt is sometimes < -1, screwing up the
initial SvGROW in svgets.
Appart from that its io is std.
p5p-msgid: <
01BBD6EE.E915C860@malvinas.AtlanTech.COM>
Subject: die -> croak
Date: Thu, 21 Nov 1996 16:11:21 -0500
From: Gurusamy Sarathy <gsar@engin.umich.edu>
Files: pp_ctl.c
private-msgid: <
199611212111.QAA17070@aatma.engin.umich.edu>
Subject: Cleanup of {,un}pack('w').
From: Chip Salzenberg <chip@atlantic.net>
Files: pp.c
Subject: Cleanups from Ilya.
From: Chip Salzenberg <chip@atlantic.net>
Files: gv.c malloc.c pod/perlguts.pod pp_ctl.c
Subject: Fix for unpack('w') on 64-bit systems.
From: Chip Salzenberg <chip@atlantic.net>
Files: pp.c
Subject: Re: LC_NUMERIC support is ready + performance
Date: Mon, 25 Nov 1996 22:08:27 -0500 (EST)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Files: sv.c
Chip Salzenberg writes:
>
> Having thought about the use of our own gcvt() and atof(), I've run
> away in horror. It's just too hairy.
>
> So I've implemented the only viable alternative I know of: Toggling
> LC_NUMERIC to/from "C" as needed.
>
> Patch follows.
>
> I think _09 is *very* close.
Since _09 is going to be alpha anyway, I reiterate my question:
Is there any reason to not include my hash/array performance
patches in _09?
Btw, here is the next performance patch. It makes PADTMP values
stealable too. I do not do by setting TEMP flags on them, since it
would be a very distributed patch, and it would break some places
which check for TEMP for some other reasons (yes, I checked ;-).
This patch decreases *twice* the memory usage of
perl -e '$a = "a" x 1e6; 1'
Enjoy,
p5p-msgid: <
199611260308.WAA02677@monk.mps.ohio-state.edu>
Subject: Hash key sharing improvements from Ilya.
From: Chip Salzenberg <chip@atlantic.net>
Files: hv.c hv.h proto.h
Subject: Mortal stack pre-allocation from Ilya.
From: Chip Salzenberg <chip@atlantic.net>
Files: pp.c pp.h pp_ctl.c pp_hot.c pp_sys.c
PORTABILITY
Subject: VMS patches post-5.003_08
Date: Fri, 22 Nov 1996 18:16:31 -0500 (EST)
From: Charles Bailey <bailey@hmivax.humgen.upenn.edu>
Files: lib/ExtUtils/MM_Unix.pm lib/ExtUtils/MM_VMS.pm lib/ExtUtils/MakeMaker.pm lib/File/Path.pm mg.c pp_ctl.c utils/h2xs.PL vms/config.vms vms/descrip.mms vms/gen_shrfls.pl vms/genconfig.pl vms/perlvms.pod vms/vms.c vms/vmsish.h
Here're diffs to bring a base 5.003_08 up to the current VMS working
sources. Nearly all of the changes are VMS-specific, and comprise
miscellaneous bugfixes accumulated since 5.003_07, rather than any
particular problem with 5.003_08. I'm posting them here since some
of the patches change core files, and I'd like to insure that I
haven't accidentally created problems for anyone else.
With these and a couple of of the small patches already send to p5p,
5.003_08 builds clean and passes all tests under VMS.
Thanks, Chip, for all the work.
p5p-msgid: <1996Nov22.181631.1603238@hmivax.humgen.upenn.edu>
Jarkko Hietaniemi [Mon, 25 Nov 1996 19:46:31 +0000]
memory corruption / security bug in sysread,syswrite + patch
Using a negative offset in sysread() gives interesting results.
I get either
assertion botched: OV_MAGIC(op, bucket) == MAGIC
zsh: 22828 abort perl xp1 < /etc/passwd
if the offset is 'mild' or
panic: realloc at xp2 line 1.
if the offset is 'wild'.
Using a negative offset in syswrite() opens up interesting vistas,
like, say, your stack :-)
A patch follows. 'Reasonably' small negative values are accepted,
they count from the end of the data. One possible point of debate:
should that be +1? That is, if the offset is negative, should that
mean 1) at the point 2) after the point? For sysread(), "after" might
sometimes be a better choice?
p5p-msgid: <
199611231705.TAA02671@alpha.hut.fi>
private-msgid: <
199611251946.VAA30459@alpha.hut.fi>
Chip Salzenberg [Sat, 23 Nov 1996 14:01:01 +0000]
Fix type mismatches in x2p's safe{alloc,realloc,free}.
Chip Salzenberg [Mon, 25 Nov 1996 12:56:45 +0000]
Update locale documentation.
Joseph S. Myers [Wed, 20 Nov 1996 23:54:41 +0000]
Patch for pod/perlpod.pod
This patch documents the E<lt> and E<gt> pod escapes, and adds a section
on common pod pitfalls (in particular the text that translators add around
L<> links, since these seem very commonly to be used incorrectly).
p5p-msgid: <Pine.LNX.3.95.
961120235016.6666A-100000@hammer.chu.cam.ac.uk>
private-msgid: <Pine.LNX.3.95.
961120235016.6666A-100000@hammer.chu.cam.ac.uk
Chip Salzenberg [Tue, 26 Nov 1996 08:48:00 +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, 26 Nov 1996 08:48:00 +0000]
Re: blib
(this is the same change as commit
be9cdb876bca65c372096fa7626a5a9897471b1d, but as applied)
Nick Ing-Simmons [Sat, 23 Nov 1996 09:17:40 +0000]
blib
Chip Salzenberg <chip@atlantic.net> writes:
>According to Karl Glazebrook:
>> How about putting blib.pm in 5.004? It's a darn useful utility and only
>> consumes a few bytes.
>
>Um, where/what is it?
It allows
perl -Mblib script
to be used to test script on pre-installed extension when things are still
in MakeMaker's 'blib' area.
Here is my current version. It now puts absolute paths in @INC to speed
Auto/Dyna loading and to allow script to chdir()
p5p-msgid: <
199611230917.JAA00471@ni-s.u-net.com>
Chip Salzenberg [Mon, 25 Nov 1996 12:56:59 +0000]
Socket test improvement from Ilya.
Achim Bohnet [Tue, 19 Nov 1996 18:54:23 +0000]
minimal timelocal.pl for _09
private-msgid: <
9611191854.AA19586@o09.rosat.mpe-garching.mpg.de>
Chip Salzenberg [Sat, 23 Nov 1996 14:01:27 +0000]
Fix bitwise op test; clean up a couple of others
Chip Salzenberg [Sat, 23 Nov 1996 14:01:14 +0000]
Update documentation and warning in I18N::Collate.
Chip Salzenberg [Sat, 23 Nov 1996 14:01:10 +0000]
Elide spurious space in db-hash.t
Jeff Okamoto [Mon, 25 Nov 1996 21:16:17 +0000]
Another HP "patch" that didn't make it (new hints file)
p5p-msgid: <
199611252116.AA245766577@hpcc123.corp.hp.com>
Jeff Okamoto [Tue, 26 Nov 1996 02:15:26 +0000]
HP patches didn't make it into _08
There are two fixes for HP-UX that didn't make it into the _08 release. I'd
like to ask that they be added into the _09 release. The first is the patch
to dl_hpux.xs that adds some debugging options. The second was a new hints
file. Here's the patch, and I'll send the hints file under separate cover.
p5p-msgid: <
199611252114.AA245686448@hpcc123.corp.hp.com>
private-msgid: <
199611260215.AA100414526@hpcc123.corp.hp.com>
Ilya Zakharevich [Sun, 24 Nov 1996 11:47:25 +0000]
5.003_08: OS/2-specific bugs/enhancements
This patch modifies the following files:
hints/os2.sh
New flags added.
os2/Changes
Documentation...
os2/Makefile.SHs
perl_init_i18nl10n is now in main list of symbols.
os2/OS2/PrfDB/PrfDB.pm
@ISA corrected.
os2/os2.c
reliable signals used, popen not used any more.
May work under OS/2 2.1 again.
README.os2
misprints and minor additions.
p5p-msgid: <
199611241147.GAA00490@monk.mps.ohio-state.edu>
Chip Salzenberg [Mon, 25 Nov 1996 12:56:20 +0000]
Add new symbols to old_global.sym, too.
Chip Salzenberg [Sat, 23 Nov 1996 14:02:42 +0000]
Patch for embed.pl when !EMBED && !MULTIPLICITY
Nick Ing-Simmons [Thu, 21 Nov 1996 15:47:46 +0000]
"static" call to UNIVERSAL::can
Is there some reason why 'can' does not allow a class name,
but (silently) insists on an instance?
I wanted to do this :
sub Construct
{
my $class = (caller(0))[0];
if ($class->can('Something'))
{
...
}
else
{
...
}
}
can just returns undef in this case, even if class has method in
question.
Anyone object to a patch?
p5p-msgid: <
199611211407.OAA14645@pluto>
private-msgid: <
199611211547.PAA15878@pluto>
Chip Salzenberg [Sat, 23 Nov 1996 14:01:08 +0000]
Configure fix for handling DynaLoader
Chip Salzenberg [Tue, 26 Nov 1996 08:48:00 +0000]
[shell changes from patch from perl5.003_08 to perl5.003_09]
Change from running these commands:
# We'll create a new test, but patch won't automatically make it
# executable.
touch t/op/sysio.t
chmod +x t/op/sysio.t
exit 0
Chip Salzenberg [Tue, 19 Nov 1996 02:16:00 +0000]
[differences between cumulative patch application and perl5.003_08]
Perl 5 Porters [Tue, 19 Nov 1996 02:16:00 +0000]
[inseparable changes from patch from perl5.003_07 to perl5.003_08]
CORE LANGUAGE CHANGES
Subject: Bitwise op sign rationalization
From: Chip Salzenberg <chip@atlantic.net>
Files: op.c opcode.pl pod/perlop.pod pod/perltoc.pod pp.c pp.h pp_hot.c proto.h sv.c t/op/bop.t
Make bitwise ops result in unsigned values, unless C<use
integer> is in effect. Includes initial support for UVs.
Subject: Defined scoping for C<my> in control structures
From: Chip Salzenberg <chip@atlantic.net>
Files: op.c perly.c perly.c.diff perly.h perly.y proto.h toke.c
Finally defines semantics of "my" in control expressions,
like the condition of "if" and "while". In all cases, scope
of a "my" var extends to the end of the entire control
structure. Also adds new construct "for my", which
automatically declares the control variable "my" and limits
its scope to the loop.
Subject: Fix ++/-- after int conversion (e.g. 'printf "%d"')
From: Chip Salzenberg <chip@atlantic.net>
Files: pp.c pp_hot.c sv.c
This patch makes Perl correctly ignore SvIVX() if either
NOK or POK is true, since SvIVX() may be a truncated or
overflowed version of the real value.
Subject: Make code match Camel II re: functions that use $_
From: Paul Marquess <pmarquess@bfsec.bt.co.uk>
Files: opcode.pl
Subject: Provide scalar context on left side of "->"
From: Chip Salzenberg <chip@atlantic.net>
Files: perly.c perly.y
Subject: Quote bearword package/handle FOO in "funcname FOO => 'bar'"
From: Chip Salzenberg <chip@atlantic.net>
Files: toke.c
OTHER CORE CHANGES
Subject: Warn on overflow of octal and hex integers
From: Chip Salzenberg <chip@atlantic.net>
Files: proto.h toke.c util.c
Subject: If -w active, warn for commas and hashes ('#') in qw()
From: Chip Salzenberg <chip@atlantic.net>
Files: toke.c
Subject: Fixes for pack('w')
From: Ulrich Pfeifer <pfeifer@charly.informatik.uni-dortmund.de>
Files: pp.c t/op/pack.t
Subject: More complete output from sv_dump()
From: Gurusamy Sarathy <gsar@engin.umich.edu>
Files: sv.c
Subject: Major '..' and debugger patches
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Files: lib/perl5db.pl op.c pp_ctl.c scope.c scope.h
Subject: Fix for formline()
From: Gurusamy Sarathy <gsar@engin.umich.edu>
Files: global.sym mg.c perl.h pod/perldiag.pod pp_ctl.c proto.h sv.c t/op/write.t
Subject: Fix stack botch in untie and binmode
From: Gurusamy Sarathy <gsar@engin.umich.edu>
Files: pp_sys.c
Subject: Complete EMBED, including symbols from interp.sym
From: Chip Salzenberg <chip@atlantic.net>
Files: MANIFEST embed.pl ext/DynaLoader/dlutils.c ext/SDBM_File/sdbm/sdbm.h global.sym handy.h malloc.c perl.h pp_sys.c proto.h regexec.c toke.c util.c x2p/Makefile.SH x2p/a2p.h x2p/handy.h x2p/util.h
New define EMBEDMYMALLOC makes embedding total by
avoiding "Mymalloc" etc.
Subject: Support old embedding for people who want it
From: Chip Salzenberg <chip@atlantic.net>
Files: MANIFEST Makefile.SH old_embed.pl old_global.sym
PORTABILITY
Subject: Miscellaneous VMS fixes
From: Charles Bailey <bailey@HMIVAX.HUMGEN.UPENN.EDU>
Files: lib/ExtUtils/Liblist.pm lib/ExtUtils/MM_VMS.pm lib/Math/Complex.pm lib/Time/Local.pm lib/timelocal.pl perl.h perl_exp.SH proto.h t/TEST t/io/read.t t/lib/findbin.t t/lib/getopt.t util.c utils/h2xs.PL vms/Makefile vms/config.vms vms/descrip.mms vms/ext/Stdio/Stdio.pm vms/ext/Stdio/Stdio.xs vms/perlvms.pod vms/test.com vms/vms.c
Subject: DJGPP patches (MS-DOS)
From: "Douglas E. Wegscheid" <wegscd@whirlpool.com>
Files: doio.c dosish.h ext/SDBM_File/sdbm/sdbm.c handy.h lib/AutoSplit.pm lib/Cwd.pm lib/File/Find.pm malloc.c perl.c perl.h pp_sys.c proto.h sv.c util.c
Subject: Patch to make Perl work under AmigaOS
From: "Norbert Pueschel" <pueschel@imsdd.meb.uni-bonn.de>
Files: MANIFEST hints/amigaos.sh installman lib/File/Basename.pm lib/File/Find.pm pod/pod2man.PL pp_sys.c util.c
Chip Salzenberg [Mon, 18 Nov 1996 06:22:52 +0000]
Fix a2p translation of '{print "a" "b" "c"}'
David Hammen [Mon, 18 Nov 1996 06:46:52 +0000]
Re: strtod / strtol patch for POSIX module
Chip Salzenberg [Mon, 18 Nov 1996 06:06:58 +0000]
Kludge for bareword handles
Add 'require IO::Handle' at beginning of FileHandle.pm
Stephen Zander [Mon, 18 Nov 1996 06:05:28 +0000]
Getopts::Std patch for hash support
Paul Marquess [Mon, 18 Nov 1996 06:06:27 +0000]
DB_File 1.05
Luther Huffman [Mon, 18 Nov 1996 06:06:37 +0000]
Plan 9 update
Tim Bunce [Mon, 18 Nov 1996 06:06:15 +0000]
Make C<no FOO> fail if C<unimport FOO> fails
Chip Salzenberg [Tue, 19 Nov 1996 02:16:00 +0000]
[shell changes from patch from perl5.003_07 to perl5.003_08]
Change from running these commands:
# We'll create a new script, but patch won't automatically make it
# executable. Also, remove an obsolete file.
touch old_embed.pl
chmod +x old_embed.pl
rm -f x2p/handy.h
exit 0
Andy Dougherty [Thu, 10 Oct 1996 06:54:32 +0000]
perl 5.003_07: [patch introduction and re-organisation]
# This is my patch perl5.003_07.pat to perl5.003_06
# The full description is below.
# Please execute the following commands before applying this patch.
# (You can feed this patch to 'sh' to do so.)
# Andy Dougherty <doughera@lafcol.lafayette.edu>
# We'll create a new test, but patch won't automatically make it
# executable.
touch t/lib/io_taint.t
chmod +x t/lib/io_taint.t
exit 0
This is patch perl5.003_07.pat to perl version 5.003_06.
This takes you from 5.003_06 to 5.003_07.
To apply this patch, run the above commands,
cd to your perl source directory and then type
patch -p1 -N < perl5.003_07.pat
The changes are described after each /^Index:/ line below. This is
designed so you can examine each change with a command such as
csplit -k perl5.003_07.pat '/^Index:/' '{99}'
Patch and enjoy,
Andy Dougherty doughera@lafcol.lafayette.edu
Dept. of Physics
Lafayette College, Easton PA 18042
Norton Allen [Wed, 9 Oct 1996 20:00:29 +0000]
Re: sh Configure?
Extract x2p/Makefile.SH and x2p/cflags.SH correctly down
in the x2p directory, even if $0 isn't set to the full
pathname of the file being extracted.
Norton Allen [Wed, 9 Oct 1996 20:00:29 +0000]
Re: sh Configure?
Extract x2p/Makefile.SH and x2p/cflags.SH correctly down
in the x2p directory, even if $0 isn't set to the full
pathname of the file being extracted.
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: util.c
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
uses my_syspopen, my_syspclose ifdef OS2. my_pclose is defined
as my_syspclose ifdef OS2 and can FORK (as OS2 does).
Gurusamy Sarathy [Wed, 9 Oct 1996 04:41:27 +0000]
more t/op/sort.t tests
Spider Boardman [Thu, 10 Oct 1996 05:09:59 +0000]
Re: 5.003_06 is available (results on ULTRIX)
fix t/lib/socket.t to treat TCP like the stream protocol it is
rather than expecting it behave rationally in all cases.
Randy J. Ray [Wed, 9 Oct 1996 02:24:48 +0000]
PATCH: untaint method for IO::Handle, 5.003_06 version
This is a re-post of my patch to Graham's IO library to add a method in
IO::Handle called "untaint", that sets the IOf_UNTAINT flag on an object
that is of or inherits from IO::Handle. With this flag set, data read from
said handle is not tainted, whether running under -T, suid or sgid.
This patch adds the method to IO.xs, adds documentation and warning to the
pod of IO/Handle.pm, creates a new test in t/lib called io_taint.t, and
adds mention of the new file to MANIFEST.
Test suite for the untaint method of class IO::Handle.
Ilya Zakharevich [Thu, 10 Oct 1996 02:30:38 +0000]
perl 5.003_07: t/lib/io_pipe.t
Date: Wed, 9 Oct 1996 22:30:38 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Better error message on dying.
Ilya Zakharevich [Thu, 10 Oct 1996 02:30:38 +0000]
perl 5.003_07: 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
Date: Wed, 9 Oct 1996 22:30:38 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
File mode under OS/2 is not what you expect. However, this has
nothing to do with databases, _and_ there is a test
for this in stat.t (which dutifully fails). There is
no point to consider this behaviour as a bug in
database code.
So OS/2 is special-cased in these tests.
Gurusamy Sarathy [Wed, 9 Oct 1996 03:54:47 +0000]
Re: Sorting lists of integers doesn't always work
>> > print sort (4,1,2,3);
>> >
>> > actually prints "4123", i.e. doesn't actually sort. Bug? Feature?
This broke between 5.001n and 5.002. There was a long winded thread
about sorting undefs in some order (rather than coredumping) around
the 5.002beta times (search for "bogorefs" in the subject-line on
p5p archive for details). Larry added in some code that presumes that
the private flags are set by the time qsort() is called:
Unfortunately, sv_2pv() does not set the POKp flag, so the above
code breaks! Here's a patch against 5.00306.
Chris Faylor [Wed, 9 Oct 1996 19:07:24 +0000]
perl 5.003_07: pp_sys.c
Date: Wed, 9 Oct 1996 19:07:24 GMT
From: Chris Faylor <cgf@bbc.com>
The problem is that SCO apparently needs to have a file opened
with write privileges for chsize to work correctly.
Ilya Zakharevich [Thu, 10 Oct 1996 02:30:38 +0000]
perl 5.003_07: pod/perlvar.pod
Date: Wed, 9 Oct 1996 22:30:38 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
$^E under OS/2.
Perl 5 Porters [Wed, 9 Oct 1996 01:38:42 +0000]
perl 5.003_07: pod/perltrap.pod
Clarified that warn() _always_ printed to STDERR, both in perl4
and perl5.
Ilya Zakharevich [Thu, 10 Oct 1996 02:30:38 +0000]
perl 5.003_07: pod/perlop.pod
Date: Wed, 9 Oct 1996 22:30:38 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Crossrefs corrected.
Perl 5 Porters [Thu, 10 Oct 1996 00:07:32 +0000]
Updated version with high bits intact.
Ulrich Pfeifer [Fri, 20 Sep 1996 11:17:14 +0000]
Re: Patch for ASN.1 compressed integer in pack/unpack
Ilya Zakharevich [Thu, 10 Oct 1996 02:30:38 +0000]
perl 5.003_07: pod/perldiag.pod
Date: Wed, 9 Oct 1996 22:30:38 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
mention that malloc in berkeley DB is broken, and PERL_BADFREE.
OS/2-specific messages added.
Perl 5 Porters [Wed, 9 Oct 1996 09:48:41 +0000]
Updated for 5.003_07
Perl 5 Porters [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: perl.c
Date: Wed, 9 Oct 1996 19:03:41 +0000
From: Tim Bunce <Tim.Bunce@ig.co.uk>
Subject: Infinte loop with perl_destruct_level and $SIG{__WARN__}
I've just started using purify on a perl with DBD::Oracle linked in
(the number of uninitialised memory reads in the Oracle libraries
is frightning!).
If perl_destruct_level and $SIG{__WARN__} are set then I see a range
of problems typified by this example and folowed by a core dump:
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Copywrite of OS/2 port now has \n\n.
Now deletes -e file (again!) if compilation is interrupted.
Ilya Zakharevich [Thu, 10 Oct 1996 02:29:44 +0000]
perl 5.003_07: os2/os2ish.h
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
sys/un.h is not very useful without Merlin toolkit.
updates for fork()ing.
Ilya Zakharevich [Thu, 10 Oct 1996 02:29:44 +0000]
perl 5.003_07: os2/os2.c
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
/bin/sh is translated to the configured value of location of sh.exe.
popen() used even if we can fork (as we do now).
builtins added for the sake of path manipulation.
Perl 5 Porters [Wed, 9 Oct 1996 09:23:00 +0000]
Updated.
Ilya Zakharevich [Thu, 10 Oct 1996 02:29:44 +0000]
perl 5.003_07: os2/OS2/ExtAttr/t/os2_ea.t
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Use `unlink' where appropriate.
Ilya Zakharevich [Thu, 10 Oct 1996 02:29:44 +0000]
perl 5.003_07: os2/Makefile.SHs
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Convoluted process to create chimera executables added.
aout_clean is done automatically on clean.
Perl 5 Porters [Thu, 10 Oct 1996 02:29:44 +0000]
perl 5.003_07: dev/null
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
sys/un.h is not very useful without Merlin toolkit.
updates for fork()ing.
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
added.
Chip Salzenberg [Thu, 10 Oct 1996 14:33:08 +0000]
Re: Group fix for 5.003_06
The group problems recently experienced are due to a small error
introduced in 5.003_06. This patch is required to fix the bug:
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: makedepend.SH
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
weed out perl_exp.SH, config_h.SH
(They have these funny names to avoid names like perl.exp.SH
with more than two '.' Such names are illegal on some systems.)
Perl 5 Porters [Wed, 9 Oct 1996 01:47:01 +0000]
Updated to version 2.4 to match Getopt::Long.
Brent B. Powers [Thu, 10 Oct 1996 18:22:05 +0000]
Re: patch for lib.pm
Ignore undefined entries.
Perl 5 Porters [Wed, 9 Oct 1996 01:46:20 +0000]
Update to version 2.4.
Gisle Aas [Fri, 20 Sep 1996 13:04:04 +0000]
Documentation patch to the FindBin module
Ilya Zakharevich [Thu, 10 Oct 1996 04:42:29 +0000]
Cleanup after new test
Below are patches for File::Copy (copying to filehandles was just
plain broken under OS/2 and VMS)
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: lib/ExtUtils/xsubpp
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Logic for processing RETVAL documented (at last!).
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: lib/ExtUtils/typemap
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
`bool' entry added.
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: lib/ExtUtils/MM_Unix.pm
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Made `use strict'-clean even in parts shadowed by Autoloading.
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: lib/Cwd.pm
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Use builtin methods if present under OS/2 (maybe should be
done outside of OS/2 too?).
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: installperl
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Restore timestamps under OS/2 (needed for binary install).
Perl 5 Porters [Wed, 9 Oct 1996 02:34:18 +0000]
Don't include <sys/select.h> along with <time.h>.
Ilya Zakharevich [Thu, 10 Oct 1996 02:29:44 +0000]
perl 5.003_07: hints/os2.sh
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Some optimization (speedup in loading GNU utilities with some
memory present - 32M should be quite enough).
Test for revision of EMX, and setting fork()ing appropriately.
libc was in .../st/... instead of mt.
README.os2 is installed as pod/perlos2.pod.
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: gv.c
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
Better error message for overload.
Perl 5 Porters [Tue, 8 Oct 1996 22:12:21 +0000]
perl 5.003_07: ext/SDBM_File/sdbm/sdbm.h perl.h
Change Myrealloc to Myremalloc to conform to 5.003's version.
I left in the Mycalloc since malloc.c now includes a calloc,
and we might need to hide it.
Randy J. Ray [Wed, 9 Oct 1996 02:24:48 +0000]
PATCH: untaint method for IO::Handle, 5.003_06 version
This is a re-post of my patch to Graham's IO library to add a method in
IO::Handle called "untaint", that sets the IOf_UNTAINT flag on an object
that is of or inherits from IO::Handle. With this flag set, data read from
said handle is not tainted, whether running under -T, suid or sgid.
This patch adds the method to IO.xs, adds documentation and warning to the
pod of IO/Handle.pm, creates a new test in t/lib called io_taint.t, and
adds mention of the new file to MANIFEST.
Document IO::Handle::untaint and give warning about the bad
things it can do.
Randy J. Ray [Wed, 9 Oct 1996 02:24:48 +0000]
PATCH: untaint method for IO::Handle, 5.003_06 version
This is a re-post of my patch to Graham's IO library to add a method in
IO::Handle called "untaint", that sets the IOf_UNTAINT flag on an object
that is of or inherits from IO::Handle. With this flag set, data read from
said handle is not tainted, whether running under -T, suid or sgid.
This patch adds the method to IO.xs, adds documentation and warning to the
pod of IO/Handle.pm, creates a new test in t/lib called io_taint.t, and
adds mention of the new file to MANIFEST.
Add method "untaint" into class IO::Handle
Perl 5 Porters [Thu, 10 Oct 1996 04:01:53 +0000]
Update SH_PATH comment.
Perl 5 Porters [Thu, 10 Oct 1996 01:08:41 +0000]
New version.
Ilya Zakharevich [Thu, 10 Oct 1996 02:32:22 +0000]
perl 5.003_07: Makefile.SH
Date: Wed, 9 Oct 1996 22:32:22 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
All the executable targets are moved into the same chunk with
shared library target, which is delegated to
$osname/Makefile.SHs if found.
config.h should depend on config_h.SH.
Remove mkmanifest target, since it will generate incorrectly
sorted MANIFEST file, I would imagine (I haven't checked).
Perl 5 Porters [Thu, 10 Oct 1996 02:29:44 +0000]
perl 5.003_07: MANIFEST
Date: Tue, 8 Oct 1996 22:24:48 -0400
From: "Randy J. Ray" <rjray@uswest.com>
Subject: PATCH: untaint method for IO::Handle, 5.003_06 version
This is a re-post of my patch to Graham's IO library to add a method in
IO::Handle called "untaint", that sets the IOf_UNTAINT flag on an object
that is of or inherits from IO::Handle. With this flag set, data read from
said handle is not tainted, whether running under -T, suid or sgid.
This patch adds the method to IO.xs, adds documentation and warning to the
pod of IO/Handle.pm, creates a new test in t/lib called io_taint.t, and
adds mention of the new file to MANIFEST.
Add mention of t/lib/io_taint.t
Date: Wed, 9 Oct 1996 22:29:44 -0400 (EDT)
From: Ilya Zakharevich <ilya@math.ohio-state.edu>
os2/Changes added.