Description of problem: Constant subroutine IO::Socket::INET6::AF_INET6 redefined at /usr/lib/perl5/5.10.0/Exporter.pm line 66. at /usr/lib/perl5/vendor_perl/5.10.0/IO/Socket/INET6.pm line 16 Prototype mismatch: sub IO::Socket::INET6::AF_INET6 () vs none at /usr/lib/perl5/5.10.0/Exporter.pm line 66. at /usr/lib/perl5/vendor_perl/5.10.0/IO/Socket/INET6.pm line 16 Constant subroutine IO::Socket::INET6::PF_INET6 redefined at /usr/lib/perl5/5.10.0/Exporter.pm line 66. at /usr/lib/perl5/vendor_perl/5.10.0/IO/Socket/INET6.pm line 16 Prototype mismatch: sub IO::Socket::INET6::PF_INET6 () vs none at /usr/lib/perl5/5.10.0/Exporter.pm line 66. at /usr/lib/perl5/vendor_perl/5.10.0/IO/Socket/INET6.pm line 16 Constant subroutine Net::DNS::Resolver::Base::AF_INET6 redefined at /usr/lib/perl5/5.10.0/Exporter.pm line 66. at /usr/lib64/perl5/vendor_perl/5.10.0/x86_64-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/lib64/perl5/vendor_perl/5.10.0/x86_64-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/lib64/perl5/vendor_perl/5.10.0/x86_64-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/lib64/perl5/vendor_perl/5.10.0/x86_64-linux-thread-multi/Net/DNS/Resolver/Base.pm line 65 Version-Release number of selected component (if applicable): perl-IO-Socket-INET6-2.51-5.fc9.noarch perl-5.10.0-15.fc9.x86_64 How reproducible: Triggered every 15min by the spamassassin cron. As SA is usually used periodically like this, this is pretty much a blocker that will hit a lot of people
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.