Any suggestions from perl SIG?

*** Bug 438931 has been marked as a duplicate of this bug. ***

Hi,

Here I found more informations about this issue (with patch):

http://use.perl.org/~Shlomi+Fish/journal/35423
http://www.nntp.perl.org/group/perl.module-authors/2008/01/msg6187.html

Hope this will help you a bit.

The patch from previous comment is used in upstream and also in Mandriva.

Pushing 2.54.  Please give it a test.

Running spamassasin gives this:
Subject: Cron <root@vmware165> /usr/share/spamassassin/sa-update.cron 2>&1 | tee
-a /var/log/sa-update.log

Constant subroutine Net::DNS::Resolver::Base::AF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at
/usr/lib/perl5/vendor_perl/5.10.0/i386-linux-thread-multi/Net/DNS/Resolver/Base.pm
line 65
Prototype mismatch: sub Net::DNS::Resolver::Base::AF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at
/usr/lib/perl5/vendor_perl/5.10.0/i386-linux-thread-multi/Net/DNS/Resolver/Base.pm
line 65
Constant subroutine Net::DNS::Resolver::Base::PF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at
/usr/lib/perl5/vendor_perl/5.10.0/i386-linux-thread-multi/Net/DNS/Resolver/Base.pm
line 65
Prototype mismatch: sub Net::DNS::Resolver::Base::PF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at
/usr/lib/perl5/vendor_perl/5.10.0/i386-linux-thread-multi/Net/DNS/Resolver/Base.pm
line 65
error: GPG validation failed!
The update downloaded successfully, but the GPG signature verification
failed.
channel: GPG validation failed, channel failed

It looks like partial fix.

Okay after reboot of machine, it works without errors.

# rpm -qa perl-IO-Socket*
perl-IO-Socket-INET6-2.54-1.fc9.noarch

I still have this problem. My details are in Bug# 438931.

Did you restart the service after upgrading to perl-IO-Socket-INET6-2.54-1.fc9?
 Comment #7 indicates that this is fixed but requires a service restart.

I have rebooted the machine!

My problem (as indicated in bug# 438931) is with a cron task - not exactly a
service. MRTG is failing with the duplication. Would you like me to provide an
updated copy of the email that's sent to me every 5 mins?

I confirm neither SA nor Amavisd-new are fixed, with or without restarts or reboots

Warren, going to get this fixed for F9?

If this is a simple redefinition of a Perl function, can't you simply _REMOVE_
one of the definitions?

I'm keen to see this fixed too... I'm getting 288 emails a day about it!

(In reply to comment #13)
> If this is a simple redefinition of a Perl function, can't you simply _REMOVE_
> one of the definitions?
> 
> I'm keen to see this fixed too... I'm getting 288 emails a day about it!

Unfortunately, the software that uses this function is very likely to be used in
cron jobs, so a lot of the users affected will be in your case. So it's not a
good idea to let it slip in F9

OK, I see the problem: Net::DNS has the same issues as IO::Socket::INET6 does.
I've fixed it up in perl-Net-DNS-0.63-3.fc9, and spamassassin and spamd run
silent for me now.

Can you guys update to this new perl-Net-DNS package and test? Thanks.

http://koji.fedoraproject.org/koji/buildinfo?buildID=45721

seems to fix it there, thanks!

*** Bug 442884 has been marked as a duplicate of this bug. ***

Running with:
mrtg-2.15.1-8.fc9.i386
perl-5.10.0-20.fc9.i386
perl-IO-Socket-INET6-2.54-1.fc9.noarch
perl-Net-DNS-0.63-3.fc9.i386
perl-SNMP_Session-1.08-4.fc9.noarch

! this is not yet resolved for mrtg as suggested above - 
The two machines have had these packages for a week and been rebooted numerous
times. I can't reopen this bug - should I reopen 442884 or {437681} to track the
mrtg problems ?

Please open a bug on mrtg, with the following:

1. A log of the exact error messages.
2. A reproducer (the simpler, the better)

Thanks,

~spot

mrtg-2.15.1-8.fc9.x86_64
perl-IO-Socket-INET6-2.54-1.fc9.noarch
perl-SNMP_Session-1.08-4.fc9.noarch

*Message*:

Constant subroutine SNMP_Session::AF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 140
Prototype mismatch: sub SNMP_Session::AF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 140
Constant subroutine SNMP_Session::PF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 140
Prototype mismatch: sub SNMP_Session::PF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 140
Constant subroutine SNMPv1_Session::AF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 594
Prototype mismatch: sub SNMPv1_Session::AF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 594
Constant subroutine SNMPv1_Session::PF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 594
Prototype mismatch: sub SNMPv1_Session::PF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/../lib64/mrtg2/SNMP_Session.pm line 594
Constant subroutine main::AF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/mrtg line 101
Prototype mismatch: sub main::AF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/mrtg line 101
Constant subroutine main::PF_INET6 redefined at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/mrtg line 101
Prototype mismatch: sub main::PF_INET6 () vs none at
/usr/lib/perl5/5.10.0/Exporter.pm line 66.
 at /usr/bin/mrtg line 101

I'll re-open my bug (Bug# 438931) then to see if that helps get this solved faster.

It is quite annoying.

(In reply to comment #19)
> Please open a bug on mrtg, with the following:
> 
> 1. A log of the exact error messages.
> 2. A reproducer (the simpler, the better)
Actually, my previously dupes bug 438931 has an easy reproducer...

(In reply to comment #22)
> Actually, my previously dupes bug 438931 has an easy reproducer...
oops, tab error, bug 442884.

I'm pretty sure I have this fixed now, please look at 442884.

I can confirm it appears fixed in 2.16.1-1 (in that MRTG no longer emails me
with complaints every 5 mins).

Cheers.

Is the patch in perl-Net-DNS really needed? Because it creates new bug 445314,
which is occur only with out patch.

I just want to warn you, I will not be able to fix this myself.

Couldn't we cooperate with upstream on this issue? There must be something wrong
in perl-IO-Socket-INET6.

(#26) Emphatically NOT.

There seems to have been a rash of changes to IO::Socket::INET6 recently.

Suggest backtracking to previous stable version 2.51 (Oct 2004) which I think
might make the problem go away.

Once F9 release is out of the way, the problem can be examined in more detail
and under less pressure.

Bug reports should be filed in CPAN RT for both of the examples above, so that
the maintainer is aware.

I can file this bug in upstream CPAN RT, but I don't think we should revert to a
four year old version, as it almost certainly won't resolve this bug (this is
caused by perl 5.10.0).

Reassigning to perl-Net-DNS, as that is the place where all of the patching has
taken place.

Upstream bug filed: http://rt.cpan.org/Public/Bug/Display.html?id=35752

Removing from F9 blocker, as we aren't holding the release for this.

(#31) Net::DNS is not implicated in this problem. Adding a bug to it is deeply
objectionable.

Bugs are like moles under your lawn - to do any good, you have to kill them, not
just fill up the holes!

(In reply to comment #34)
> (#31) Net::DNS is not implicated in this problem. Adding a bug to it is deeply
> objectionable.

Feel free to deeply object, however, it is the correct thing to do. The bug
comes from how IO::Socket::INET6 is used in Net::DNS (and other places), not
with IO::Socket::INET6 itself. There is some background information about this
bug in comment #3.

It may be possible to resolve this issue with fundamental changes to how
IO::Socket::INET6 is structured, but that is analogous to insisting that the
lawn be replaced with concrete to solve the mole problem.

(#35)
Concreting the lawn was your solution, forcing the moles to surface in the next
guy's garden.

Breaking Net::DNS IPv6 functionality is definitely not the correct thing to do.

Failing to comply with GPL clause 2(a) is not the correct way to do it.

Even if your change had not broken anything, the bug would still remain, same as
it was 6 weeks ago.

I suspect that you don't understand the GPL as well as you think, nor this bug.
Nevertheless, I'm working with upstream on a proper solution.

2.
    a) You must cause the modified files to carry prominent notices
    stating that you changed the files and the date of any change.

Not much scope for misunderstanding there.

Still a long way from understanding this bug.

I was aware.

*** Bug 445314 has been marked as a duplicate of this bug. ***

There was patch from Net::DNS upstream, which resolve all mentioned issues for
me. I push it to updates, please test and let us know.

CPAN Ticket http://rt.cpan.org/Public/Bug/Display.html?id=35752

perl-Net-DNS-0.63-4.fc9 has been submitted as an update for Fedora 9

Confirmed, this is a much better fix.

perl-Net-DNS-0.63-4.fc9 has been pushed to the Fedora 9 testing repository.  If problems still persist, please make note of it in this bug report.
 If you want to test the update, you can install it with 
 su -c 'yum --enablerepo=updates-testing update perl-Net-DNS'.  You can provide feedback for this update here: http://admin.fedoraproject.org/updates/F9/FEDORA-2008-3676

Changing version to '9' as part of upcoming Fedora 9 GA.
More information and reason for this action is here:
http://fedoraproject.org/wiki/BugZappers/HouseKeeping

perl-Net-DNS-0.63-4.fc9 has been pushed to the Fedora 9 stable repository.  If problems still persist, please make note of it in this bug report.