Justin,

Problem still exists for qeth layer3 devices. If qeth layer2 devices are
needed only, qeth can now be used without ipv6:
Starting with 2.6.25 the qeth driver is splitted into 3 modules: a core
module qeth, a layer2 module qeth_l2 and a layer3 module qeth_l3. The
ipv6-dependency is isolated in the layer3 module qeth_l3. Module qeth_l3
will only be loaded, if qeth layer3 devices are activated.

Regards, Ursula Braun, IBM Germany, Linux on System z development

-------- Forwarded Message --------
> From: Ursula Braun1 <ursula.br...@de.ibm.com>
> To: ubr...@linux.vnet.ibm.com
> Subject: Fw: qeth ipv6 dependency
> Date: Mon, 31 Aug 2009 14:01:07 +0200
>
> From:
> Justin Payne <jpa...@redhat.com>
> To:
> LINUX-390@VM.MARIST.EDU
> Date:
> 27.08.2009 15:55
> Subject:
> Re: qeth ipv6 dependency
>
>
> ______________________________________________________________________
>
>
>
> Was an alternative solution to this ever found?
>
>
> On 07/02/2007 02:47 AM, Ursula Braun1 wrote:
> > Brad,
> >
> > since qeth supports layer3 devices, it is difficult to make qeth
> > independent of ipv6. The best solution would be a kernel rebuild
> without
> > ipv6 configured.
> > There exists the SLES9 solution solving Novell Bugzilla 128561 you
> are
> > pointing to. Unfortunately it is more a quick hack and thus more a
> > circumvention than a real solution for the problem. Thus up to now
> there
> > have been no efforts to port it to upstream or to another
> distribution.
> > Nevertheless we are aware of this problem and we are going to
> investigate
> > into alternative solutions.
> >
> > Best regards,    Ursula Braun
> >
> >
> >               Brad Hinson
> >               <bhin...@redhat.c
> >               om>
> To
> >               Sent by: Linux on         LINUX-390@VM.MARIST.EDU
> >               390 Port
> cc
> >               <linux-...@vm.mar
> >               IST.EDU>
> Subject
> >                                         qeth ipv6 dependency
> >
> >               29.06.2007 21:42
> >
> >
> >               Please respond to
> >               Linux on 390 Port
> >               <linux-...@vm.mar
> >                   IST.EDU>
> >
> >
> >
> >
> >
> >
> > A customer has a security policy that only allows ipv4, so they
> can't
> > assign an ipv6 address to eth0, and can't have an ipv6 over ipv4
> tunnel.
> > Ideally, they'd like to remove the ipv6 kernel module altogether.
> >
> > In RHEL 4 and SLES 9, they were able to do this.  In RHEL 5 and SLES
> 10,
> > they're finding that dependencies don't allow them to do this.  It
> > appears there are too many hard linkages between qeth and ipv6.
>  When
> > they alias ipv6 off in modprobe.conf, they get messages like:
> >
> > qeth: Unknown symbol in6_dev_finish_destroy
> > qeth: Unknown symbol addrconf_lock
> > qeth: Unknown symbol unregister_inet6addr_notifier
> > qeth: Unknown symbol ndisc_mc_map
> > qeth: Unknown symbol register_inet6addr_notifier
> > qeth: Unknown symbol in6_dev_finish_destroy
> > qeth: Unknown symbol addrconf_lock
> > qeth: Unknown symbol unregister_inet6addr_notifier
> > qeth: Unknown symbol ndisc_mc_map
> > qeth: Unknown symbol register_inet6addr_notifier
> > qeth: Unknown symbol in6_dev_finish_destroy
> > qeth: Unknown symbol addrconf_lock
> > qeth: Unknown symbol unregister_inet6addr_notifier
> > qeth: Unknown symbol ndisc_mc_map
> > qeth: Unknown symbol register_inet6addr_notifier
> >
> > I found this link, so it seems someone has asked this question
> before:
> >
> >
> http://www.ibm.com/developerworks/linux/linux390/linux-2.6.5-s390-32-april2004.html
> >
> >
> > Specifically, Problem-ID 19880:
> >
> > Description: qeth: customer wants qeth driver to be loaded and
> running
> > without ipv6 module loaded prior to. Symptom: qeth module built with
> > ipv6 support will not load when ipv6 module is not loaded.
> >
> > Problem: qeth driver is using basic ipv6 functions which are needed
> to
> > register IPv6 IP addresses. These functions are not available when
> ipv6
> > module is not loaded. Solution: Use symbol_put/symbol_get and get
> > function addresses when available and call them. Use wrapper
> functions
> > then in qeth.
> >
> > ---------
> >
> > So the question is:  Does anyone know when this change happened in
> the
> > qeth code?  i.e. What was the decision/feature in the module that
> > necessitated this dependency?  Was the patch in the link above ever
> > proposed upstream?
> >
> > Thanks,
> > -Brad
> >

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to