Re: [Xen-users] kernel BUG at /usr/src/linux-stable/drivers/net/xen-netback/netback.c:540!

[Hans de Bruin <jmdebruin@xxxxxxxxx>]

On 12/19/2014 11:06 AM, Ian Campbell wrote:
> On Thu, 2014-12-18 at 22:36 +0100, Hans de Bruin wrote:
> > On 12/17/2014 11:06 PM, Hans de Bruin wrote:
> > > Hi
> > >
> > > I just opgraded my test-host and guest to 3.14.27. Which probable was
> > > not a good idea. A nfsroot guest triggers a kernel BUG in the host when
> > > is starts its mediatomb deamon. Before the upgrade the host and guest
> > > were running 3.14.23:
> >
> > After some testing it looks like the guest is to blame, It does not
> > matter if the host is running .23 or .27, but if the guest boots .27 the
> > BUG happens and always when mediatomb is starting.
> >
> > The above is after start-up of the host and the first boot of the guest.
> > When a .23 guest is successfully started, it can be shut down and
> > restarted with a .27 kernel.
> >
> > Is there anything I should test or check before I start mailing to the
> > dev-list?
> Please check later kernels for netback fixes, I have a feeling this
> issue was fixed but I don't recall the details.
I choose 3.14 because its is the latest long term stable. the .27 has a 
xen-netfront fix (http://lwn.net/Articles/626580/):
Seth Forshee (1):
      xen-netfront: Remove BUGs on paged skb data which crosses
      a page boundary

It could have broken my upgrade (along with all the other fixes since 
.23). I will try 3.17.7. The 3.18 series have a bisected bug that 
spontaneously unmount my nfs-volumes.

--
Hans


_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users