Hey Pasi,
Ive downloaded 2.6.33.4 from kernel.org. The XEN stuff is compiled as static
and so is raid,i have created a initrd image by issueing the following (Debian
OS)
Mkinitramfs -o /boot/initrd.img-2.6.33.4 2.6.33.4
When booting i get the same error
[ 0.898795] md: raid0 personality registered for level 0
[ 0.898814] md: raid1 personality registered for level 1
[ 0.898832] md: raid10 personality registered for level 10
[ 0.898850] md: raid6 personality registered for level 6
[ 0.898867] md: raid5 personality registered for level 5
[ 0.898884] md: raid4 personality registered for level 4
[ 0.898902] md: multipath personality registered for level -4
[ 0.898920] md: faulty personality registered for level -5
[ 0.899014] device-mapper: uevent: version 1.0.3
[ 0.899545] device-mapper: ioctl: 4.17.0-ioctl (2010-03-05) initialised:
dm-devel@xxxxxxxxxx
[ 0.899921] device-mapper: multipath: version 1.1.1 loaded
[ 0.899946] device-mapper: multipath round-robin: version 1.0.0 loaded
[ 0.900432] cpuidle: using governor ladder
[ 0.900454] cpuidle: using governor menu
[ 0.900488] No iBFT detected.
[ 0.900586] Netfilter messages via NETLINK v0.30.
[ 0.900662] nf_conntrack version 0.5.0 (1938 buckets, 7752 max)
[ 0.900808] CONFIG_NF_CT_ACCT is deprecated and will be removed soon. Please
use
[ 0.900831] nf_conntrack.acct=1 kernel parameter, acct=1 nf_conntrack module
option or
[ 0.900852] sysctl net.netfilter.nf_conntrack_acct=1 to enable it.
[ 0.901215] ctnetlink v0.93: registering with nfnetlink.
[ 0.901324] xt_time: kernel timezone is -0000
[ 0.902209] ip_tables: (C) 2000-2006 Netfilter Core Team
[ 0.902297] ClusterIP Version 0.8 loaded successfully
[ 0.902355] arp_tables: (C) 2002 David S. Miller
[ 0.902410] TCP cubic registered
[ 0.902423] NET: Registered protocol family 17
[ 0.902463] Using IPI No-Shortcut mode
[ 0.903337] registered taskstats version 1
[ 1.000079] XENBUS: Device with no driver: device/console/0
[ 1.000131] drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
[ 1.000229] Freeing unused kernel memory: 424k freed
Loading, please wait...
Begin: Loading essential drivers ... done.
Begin: Running /scripts/init-premount ... FATAL: Error inserting fan
(/lib/modules/2.6.33.4/kernel/drivers/acpi/fan.ko): No such device
FATAL: Error inserting thermal
(/lib/modules/2.6.33.4/kernel/drivers/acpi/thermal.ko): No such device
done.
Begin: Mounting root file system ... Begin: Running /scripts/local-top ...
Begin: LoadingFailure: failed to load module raid1.
done.
Begin: Assembling all MD arrays ... [ 1.991528] md: md0 stopped.
mdadm: no devices found for /dev/md0
[ 2.004392] md: md1 stopped.
mdadm: no devices found for /dev/md1
[ 2.021788] md: md2 stopped.
mdadm: no devices found for /dev/md2
[ 2.047173] md: md3 stopped.
mdadm: no devices found for /dev/md3
Failure: failed to assemble all arrays.
done.
Begin: Waiting for udev to process events ... done.
done.
Begin: Waiting for root file system ...
Im not sure if i have missed something out, any ideas?
Cheers
Ian
-----Original Message-----
From: Pasi Kärkkäinen [mailto:pasik@xxxxxx]
Sent: 14 May 2010 21:18
To: Ian Tobin
Cc: xen-users@xxxxxxxxxxxxxxxxxxx
Subject: Re: [Xen-users] pv_ops Dom0 kernel for DomU?
On Fri, May 14, 2010 at 09:05:02PM +0100, Ian Tobin wrote:
> Oh right, thats new to me. Ive never had to do that before, is that new to
> the pv_ops?
>
It's not really pvops specific.. it's more about the fact that nowadays all/most
drivers are built as modules - so you need to have an initrd image to load them.
If you compiled all the xen* stuff as static/built-in then you wouldn't need an
initrd..
> Many thanks for the info, you learn something new everyday :)
>
Np.
-- Pasi
> Ian
>
>
> -----Original Message-----
> From: Pasi Kärkkäinen [mailto:pasik@xxxxxx]
> Sent: 14 May 2010 20:50
> To: Ian Tobin
> Cc: xen-users@xxxxxxxxxxxxxxxxxxx
> Subject: Re: [Xen-users] pv_ops Dom0 kernel for DomU?
>
> On Fri, May 14, 2010 at 08:45:03PM +0100, Ian Tobin wrote:
> > Ah ok, what would be the procedure to do that?
> >
> > Thanks
> >
>
> Usually it's best to generate the initrd image in the guest,
> and use the distro provided domU kernel.
>
> If you install the guest using the guest distro installer it'll do it for you
> automatically.
>
> You can also do it from dom0.. example for centos/rhel/fedora:
>
> mkinitrd -v -f --with-module=xenblk --with-module=xennet
> /boot/initrd-domU-version.img <kernelversion>
>
> -- Pasi
>
> > Ian
> >
> >
> >
> > -----Original Message-----
> > From: Pasi Kärkkäinen [mailto:pasik@xxxxxx]
> > Sent: 14 May 2010 20:32
> > To: Ian Tobin
> > Cc: xen-users@xxxxxxxxxxxxxxxxxxx
> > Subject: Re: [Xen-users] pv_ops Dom0 kernel for DomU?
> >
> > On Fri, May 14, 2010 at 08:27:50PM +0100, Ian Tobin wrote:
> > > Ok strange. I have set the DomU config to use the kernel i use for Dom0
> > > and when booting it gets so far to mount the root filesystem and then
> > > complains that it cant load the raid1 module which then says it can't
> > > load md0, md1, md2 and md3 which are my Dom0 software raid disks.
> > >
> > > When i use the 2.6.8.18 kernel the domU boots fine.
> > >
> > > I have in the config to use the initrd.img-2.6.32.12 ram disk but it
> > > doesn't work, am i missing something obvious?
> > >
> >
> > Are you using the dom0 initrd for domU?
> > If you are, you shouldn't.
> >
> > domU needs to have its own initrd that loads drivers for the virtual
> > hardware the Xen guest has..
> >
> > -- Pasi
> >
> > > Thanks
> > >
> > > Ian
> > >
> > >
> > > -----Original Message-----
> > > From: Pasi Kärkkäinen [mailto:pasik@xxxxxx]
> > > Sent: 14 May 2010 19:48
> > > To: Ian Tobin
> > > Cc: xen-users@xxxxxxxxxxxxxxxxxxx
> > > Subject: Re: [Xen-users] pv_ops Dom0 kernel for DomU?
> > >
> > > On Fri, May 14, 2010 at 03:08:57PM +0100, Ian Tobin wrote:
> > > > Hi,
> > > >
> > > >
> > > >
> > > > Just wondered if we can use the pv_ops kernel in Jermeys tree for
> > > > linux PV
> > > > domU or do we need to get that one from kernel.org ?
> > > >
> > >
> > > Both will work.
> > >
> > > upstream kernel.org kernels do have basic Xen PV domU support
> > > out-of-the-box.
> > > Jeremy's tree has some additional features, like pcifront/pciback etc,
> > > that are not yet upstream.
> > >
> > > -- Pasi
> > >
> > > >
> > > >
> > > > Thanks
> > > >
> > > >
> > > >
> > > > Ian
> > > >
> > > >
> > > >
> > > >
> > >
> > > > _______________________________________________
> > > > Xen-users mailing list
> > > > Xen-users@xxxxxxxxxxxxxxxxxxx
> > > > http://lists.xensource.com/xen-users
> > >
> > >
> >
> >
>
>
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
Home