Problem getting x86_64 dom0 to boot on a FC4 machine

Problem getting x86_64 dom0 to boot on a FC4 machine

[Greg Brackley]

I'm struggling to get Xen to boot on a FC4 Opteron box. I've included the 
tail of the boot log below [1].  I suspect that the problems relates to the 
software RAID 1 root and boot partitions, and how it relates to the initrd 
image. The RAID volumes fail to mount.

I compiled the Xen snapshot from the 23rd September, and the only change 
I've made is to enable the Silicon image SATA controller module. My grub 
config is included below [4].

In an attempt to get things going, I have moved the boot and root RAID 
partition disks onto the onboard silicon image SATA controller [2]. I also 
seem to be having problems with the Marvell mv_sata driver, but am hopeful 
that solving the other boot issues may overcome these problems.

I created an initrd image using the Fedora mkinitrd script. The /init script 
that is generated is included below [3]. Was using the Fedora mkinitrd 
script with a vanilla kernel a good idea? Is it a udev problem? Any help 
would be greatly appreciated. I'd be very keen to run the x86_64 version on 
Xen on the machine.

Greg

--



[1] Xen console boot log

Loading raid1.ko module
md: raid1 personality registered as nr 3
Loading xor.ko module
raid5: automatically using best checksumming function: generic_sse
   generic_sse:  4025.200 MB/sec
raid5: using function: generic_sse (4025.200 MB/sec)
Loading raid5.ko module
md: raid5 personality registered as nr 4
raidautorun: RAID_AUTORUN failed: 19
raidautorun: RAID_AUTORUN failed: 19
raidautorun: RAID_AUTORUN failed: 19
Creating root device
Mounting root filesystem
EXT3-fs: unable to read superblock
mount: error 22 mounting ext3
Switching to new root
ERROR opening /dev/console!!!!: 2
error dup2'ing fd of 0 to 0
error dup2'ing fd of 0 to 1
error dup2'ing fd of 0 to 2
unmounting old /proc
unmounting old /sys
switchroot: mount failed: 22
Kernel panic - not syncing: Attempted to kill init!


[2] Machine Config
Tyan S2882
Single Opteron
512M RAM
Marvell SATA Controller (Highpoint 1820A)
RAID 1 /boot
RAID 1 /
RAID 5 for non-system data mounted on /raid


[3] /init script extraced from the initrd image
#!/bin/nash

mount -t proc /proc /proc
setquiet
echo Mounted /proc filesystem
echo Mounting sysfs
mount -t sysfs /sys /sys
echo Creating /dev
mount -o mode=0755 -t tmpfs /dev /dev
mknod /dev/console c 5 1
mknod /dev/null c 1 3
mknod /dev/zero c 1 5
mkdir /dev/pts
mkdir /dev/shm
echo Starting udev
/sbin/udevstart
echo -n "/sbin/hotplug" > /proc/sys/kernel/hotplug
echo "Loading sata_sil.ko module"
insmod /lib/sata_sil.ko
echo "Loading md.ko module"
insmod /lib/md.ko
echo "Loading raid1.ko module"
insmod /lib/raid1.ko
echo "Loading xor.ko module"
insmod /lib/xor.ko
echo "Loading raid5.ko module"
insmod /lib/raid5.ko
/sbin/udevstart
raidautorun /dev/md0
raidautorun /dev/md1
raidautorun /dev/md2
echo Creating root device
mkrootdev /dev/root
echo Mounting root filesystem
mount -o defaults --ro -t ext3 /dev/root /sysroot
echo Switching to new root
switchroot --movedev /sysroot

[4] grub

title Xen (2.6.12xen0) serial
        root (hd0,0)
        kernel /xen.gz dom0_mem=192000 com1=38400,8n1
        module /vmlinuz-2.6.12-xen0  ro root=/dev/md2  console=ttyS0
        module /initrd-2.6.12-xen0.img

Re: Problem getting x86_64 dom0 to boot on a FC4 machine

[Andrew Theurer]

This is a problem with udev on FC4.  I just ran in to the same issue. We 
would get this every 3rd or 4th reboot.  I simply updated everything 
with 'yum update" and I have not had this happen again.  Also, I have 
seen this on my FC4 domUs, so you probably want to refresh that image as 
well.

-Andrew

Greg Brackley wrote:

>I'm struggling to get Xen to boot on a FC4 Opteron box. I've included the 
>tail of the boot log below [1].  I suspect that the problems relates to the 
>software RAID 1 root and boot partitions, and how it relates to the initrd 
>image. The RAID volumes fail to mount.
>
>I compiled the Xen snapshot from the 23rd September, and the only change 
>I've made is to enable the Silicon image SATA controller module. My grub 
>config is included below [4].
>
>In an attempt to get things going, I have moved the boot and root RAID 
>partition disks onto the onboard silicon image SATA controller [2]. I also 
>seem to be having problems with the Marvell mv_sata driver, but am hopeful 
>that solving the other boot issues may overcome these problems.
>
>I created an initrd image using the Fedora mkinitrd script. The /init script 
>that is generated is included below [3]. Was using the Fedora mkinitrd 
>script with a vanilla kernel a good idea? Is it a udev problem? Any help 
>would be greatly appreciated. I'd be very keen to run the x86_64 version on 
>Xen on the machine.
>
>Greg
>
>--
>
>
>
>[1] Xen console boot log
>
>Loading raid1.ko module
>md: raid1 personality registered as nr 3
>Loading xor.ko module
>raid5: automatically using best checksumming function: generic_sse
>   generic_sse:  4025.200 MB/sec
>raid5: using function: generic_sse (4025.200 MB/sec)
>Loading raid5.ko module
>md: raid5 personality registered as nr 4
>raidautorun: RAID_AUTORUN failed: 19
>raidautorun: RAID_AUTORUN failed: 19
>raidautorun: RAID_AUTORUN failed: 19
>Creating root device
>Mounting root filesystem
>EXT3-fs: unable to read superblock
>mount: error 22 mounting ext3
>Switching to new root
>ERROR opening /dev/console!!!!: 2
>error dup2'ing fd of 0 to 0
>error dup2'ing fd of 0 to 1
>error dup2'ing fd of 0 to 2
>unmounting old /proc
>unmounting old /sys
>switchroot: mount failed: 22
>Kernel panic - not syncing: Attempted to kill init!
>
>
>[2] Machine Config
>Tyan S2882
>Single Opteron
>512M RAM
>Marvell SATA Controller (Highpoint 1820A)
>RAID 1 /boot
>RAID 1 /
>RAID 5 for non-system data mounted on /raid
>
>
>[3] /init script extraced from the initrd image
>#!/bin/nash
>
>mount -t proc /proc /proc
>setquiet
>echo Mounted /proc filesystem
>echo Mounting sysfs
>mount -t sysfs /sys /sys
>echo Creating /dev
>mount -o mode=0755 -t tmpfs /dev /dev
>mknod /dev/console c 5 1
>mknod /dev/null c 1 3
>mknod /dev/zero c 1 5
>mkdir /dev/pts
>mkdir /dev/shm
>echo Starting udev
>/sbin/udevstart
>echo -n "/sbin/hotplug" > /proc/sys/kernel/hotplug
>echo "Loading sata_sil.ko module"
>insmod /lib/sata_sil.ko
>echo "Loading md.ko module"
>insmod /lib/md.ko
>echo "Loading raid1.ko module"
>insmod /lib/raid1.ko
>echo "Loading xor.ko module"
>insmod /lib/xor.ko
>echo "Loading raid5.ko module"
>insmod /lib/raid5.ko
>/sbin/udevstart
>raidautorun /dev/md0
>raidautorun /dev/md1
>raidautorun /dev/md2
>echo Creating root device
>mkrootdev /dev/root
>echo Mounting root filesystem
>mount -o defaults --ro -t ext3 /dev/root /sysroot
>echo Switching to new root
>switchroot --movedev /sysroot
>
>[4] grub
>
>title Xen (2.6.12xen0) serial
>        root (hd0,0)
>        kernel /xen.gz dom0_mem=192000 com1=38400,8n1
>        module /vmlinuz-2.6.12-xen0  ro root=/dev/md2  console=ttyS0
>        module /initrd-2.6.12-xen0.img
>
>
>
>
>_______________________________________________
>Xen-devel mailing list
>Xen-devel@lists.xensource.com
>http://lists.xensource.com/xen-devel
>
>  
>

Re: Problem getting x86_64 dom0 to boot on a FC4 machine

[Greg Brackley]

"Andrew Theurer" <habanero@us.ibm.com> wrote in message 
news:433E998D.5090004@us.ibm.com...
> with 'yum update" and I have not had this happen again.  Also, I have

Thanks for that idea. I had already done a yum update, so this didn't help.

Out of band Larry Hovind recommended compiling the RAID support into the 
kernel (c.f. as modules). I did this for all multiple device support 
(RAID/LVM), and then rebuilt the initrd (specifying that the modules were 
now built-in). I can now get the privileged domain 0 booting. I don't 
understand what limitation it has overcome, but it does get it up and going.

Still need to get a Marvell SATA driver going, and overcome a 
dma/bonding/vlan/ethernet kernel oops.

Thanks,

Greg :-)