Skip Menu |

This queue is for tickets about the PAR CPAN distribution.

Report information
The Basics
Id: 19038
Status: resolved
Priority: 0/
Queue: PAR

People
Owner: Nobody in particular
Requestors: dmacks [...] netspace.org
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: (no value)



Subject: Build failure is libperl is static
Date: Tue, 2 May 2006 16:40:36 -0400
To: bug-par [...] rt.cpan.org
From: Daniel Macks <dmacks [...] netspace.org>
If $dynperl is false, myldr/Makefile.PL does not write rules for generating sha1.c to Makefile. However, mktmpdir.h still has an #include for sha1.c, so compiling fails if libperl is static. Reproducible on Mac OS X (darwin kernel) 10.3 using locally-compiled perl-5.8.6 and on 10.4 (powerpc) using perl-5.8.1 and perl-5.8.4. No problem when using perl-5.8.1 on 10.3 or perl-5.8.6 on 10.4, because these are supplied by Apple with dynamic libperl. dan -- Daniel Macks dmacks@netspace.org http://www.netspace.org/~dmacks
On Di. 02. Mai 2006, 16:41:06, dmacks@netspace.org wrote: Show quoted text
> If $dynperl is false, myldr/Makefile.PL does not write rules for > generating sha1.c to Makefile. However, mktmpdir.h still has an > #include for sha1.c, so compiling fails if libperl is static. > > Reproducible on Mac OS X (darwin kernel) 10.3 using locally-compiled > perl-5.8.6 and on 10.4 (powerpc) using perl-5.8.1 and perl-5.8.4. No > problem when using perl-5.8.1 on 10.3 or perl-5.8.6 on 10.4, because > these are supplied by Apple with dynamic libperl. > > dan >
Hi dan, sorry for the late reply, but could you verify that this is still the case in the newer versions of PAR? (Particularily the most current release, 0.959.) I just looked at it and it *should* make sha1 alright in the current SVN trunk. There have been no changes to that part since 0.959. Thanks! Steffen
From: ema_zep [...] libero.it
On Sun Nov 12 11:22:46 2006, SMUELLER wrote: Show quoted text
> On Di. 02. Mai 2006, 16:41:06, dmacks@netspace.org wrote:
> > If $dynperl is false, myldr/Makefile.PL does not write rules for > > generating sha1.c to Makefile. However, mktmpdir.h still has an > > #include for sha1.c, so compiling fails if libperl is static. > > > > Reproducible on Mac OS X (darwin kernel) 10.3 using locally-compiled > > perl-5.8.6 and on 10.4 (powerpc) using perl-5.8.1 and perl-5.8.4. No > > problem when using perl-5.8.1 on 10.3 or perl-5.8.6 on 10.4, because > > these are supplied by Apple with dynamic libperl. > > > > dan > >
> > > > Hi dan, > > sorry for the late reply, but could you verify that this is still the > case in the newer versions of PAR? (Particularily the most current > release, 0.959.) I just looked at it and it *should* make sha1 alright > in the current SVN trunk. There have been no changes to that part since > 0.959. > > Thanks! > > Steffen
Hi Steffen. I can confirm that on a locally-compiled perl with static libperl, there is still a problem also with PAR 0.959 (Mac OS X 10.4.8, Intel, perl-5.8.8). Tests output attached. Cheers, Emanuele.

Message body is not shown because it is too large.

On Tue Nov 14 05:47:56 2006, emazep wrote: Show quoted text
> I can confirm that on a locally-compiled perl with static libperl, there > is still a problem also with PAR 0.959 (Mac OS X 10.4.8, Intel,
perl-5.8.8). Show quoted text
> > Tests output attached.
Hi Emanuele, I have since understood a little more of this problem. Still, it's unfixed. I have been discussing the issue on the PAR mailing list in the past few days. So far, there is no clear fix. You are welcome to join the discussion. I'll update this ticket once we fixed the bug. You can access the PAR mailing list using a newsreader and nntp.perl.org or find the start of the discussion at: http://beta.nntp.perl.org/group/perl.par/2006/11/msg2654.html or: http://www.nntp.perl.org/group/perl.par/2676 Steffen
CC: par [...] perl.org
Subject: Re: [rt.cpan.org #19038] Build failure is libperl is static
Date: Tue, 14 Nov 2006 11:32:37 -0500
To: Steffen M??ller via RT <bug-PAR [...] rt.cpan.org>
From: Daniel Macks <dmacks [...] netspace.org>
On Sun, Nov 12, 2006 at 11:22:47AM -0500, Steffen M??ller via RT wrote: Show quoted text
> > <URL: http://rt.cpan.org/Ticket/Display.html?id=19038 > > > On Di. 02. Mai 2006, 16:41:06, dmacks@netspace.org wrote:
> > If $dynperl is false, myldr/Makefile.PL does not write rules for > > generating sha1.c to Makefile. However, mktmpdir.h still has an > > #include for sha1.c, so compiling fails if libperl is static. > > > > Reproducible on Mac OS X (darwin kernel) 10.3 using locally-compiled > > perl-5.8.6 and on 10.4 (powerpc) using perl-5.8.1 and perl-5.8.4. No > > problem when using perl-5.8.1 on 10.3 or perl-5.8.6 on 10.4, because > > these are supplied by Apple with dynamic libperl.
> > sorry for the late reply, but could you verify that this is still the > case in the newer versions of PAR? (Particularily the most current > release, 0.959.) I just looked at it and it *should* make sha1 alright > in the current SVN trunk. There have been no changes to that part since > 0.959.
0.959 appears fixed with respect to the dynamic/static build problem, so this problem ticket can be closed IMO. dan -- Daniel Macks dmacks@netspace.org http://www.netspace.org/~dmacks
Marking as resolved. The problem Emanuele reported is valid, but most likely unrelated. Steffen
From: ema_zep [...] libero.it
On Thu Nov 16 11:46:29 2006, SMUELLER wrote: Show quoted text
> Marking as resolved. The problem Emanuele reported is valid, but most > likely unrelated.
Yes, it was definitely a different problem (though it emerged only with a static libperl as well, so I mentioned it as somewhat related). Anyway I'm glad to say that PAR 0.960 solves the problem reported by me (by skipping the parl-generation tests?). Thank you Steffen! Cheers, Emanuele.
On Wed Nov 22 21:46:51 2006, emazep wrote: Show quoted text
> [...]
Sorry, my comment made RT automatically change the ticket status from "Resolved" to "Open", while it should be Resolved of course. Emanuele.
On Wed Nov 22 21:46:51 2006, emazep wrote: Show quoted text
> Yes, it was definitely a different problem (though it emerged only with > a static libperl as well, so I mentioned it as somewhat related). > > Anyway I'm glad to say that PAR 0.960 solves the problem reported by me > (by skipping the parl-generation tests?).
Trust me, *I* am glad that the problem seems solved. I have been banging my head against this issue for some time. Anyhow, I fixed the parl-generation. The parl-generation tests are still in place. Show quoted text
> Thank you Steffen!
You are most welcome! All the best, Steffen