Of course, sorry for the delay.
Update: I did eventually get the guest to boot using Xen's domain configs, just not yet using libvirt. I specified hdtype='ahci' in the config. FWIW, I also converted the guest's disk image to raw format to fix an unrelated issue (Xen reported no boot device found).
In a different attempt, I also hand crafted a libvirt XML config, taking pieces from the working XML config on KVM and combining it with Xen's config. This was unsuccessful, the disk kept showing up as lsi or ata0.
Thanks, Micah
Here is the relevant debug output, taken as a newly created vm starts up:
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (guest:429) XML fetched from libvirt object:
<domain type='xen' id='68'>
<name>testvm</name>
<uuid>6d5aec78-d898-4d20-9580-332e967417a1</uuid>
<memory unit='KiB'>524288</memory>
<currentMemory unit='KiB'>524288</currentMemory>
<vcpu placement='static'>1</vcpu>
<os>
<type arch='x86_64' machine='xenfv'>hvm</type>
<loader type='rom'>/usr/lib/xen/boot/hvmloader</loader>
<boot dev='hd'/>
<bootmenu enable='yes'/>
</os>
<features>
<acpi/>
<apic/>
<pae/>
</features>
<clock offset='utc'/>
<on_poweroff>destroy</on_poweroff>
<on_reboot>restart</on_reboot>
<on_crash>restart</on_crash>
<devices>
<emulator>/usr/lib/xen/bin/qemu-system-i386</emulator>
<disk type='file' device='disk'>
<driver name='qemu' type='qcow2'/>
<source file='/home/micah/testvm.qcow2'/>
<backingStore/>
<target dev='sda' bus='sata'/>
<address type='drive' controller='0' bus='0' target='0' unit='0'/>
</disk>
<controller type='sata' index='0'/>
<interface type='bridge'>
<mac address='00:16:3e:65:b6:01'/>
<source bridge='virbr0'/>
<target dev='vif68.0-emu'/>
<model type='e1000'/>
</interface>
<serial type='pty'>
<target port='0'/>
</serial>
<console type='pty'>
<target type='serial' port='0'/>
</console>
<input type='mouse' bus='ps2'/>
<input type='keyboard' bus='ps2'/>
<graphics type='vnc' port='5900' autoport='yes' keymap='en-us'/>
<video>
<model type='cirrus' vram='8192' heads='1' primary='yes'/>
</video>
</devices>
</domain>
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (create:2278) Install completed
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (connection:1050) domain=testvm status=Running added
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (connection:1044) New domain=testvm requested, but it's already tracked.
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (create:205) Closing new vm wizard
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (details:645) Closing VM details: <vmmDomainVirtinst object at 0x7f57b8147c80 (virtManager+domain+vmmDomainVirtinst at 0x563dcb29fc00)>
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (create:2215) User closed customize window, closing wizard
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (details:609) Showing VM details: <vmmDomain object at 0x7f57b80c1cd0 (virtManager+domain+vmmDomain at 0x563dcdac86c0)>
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (engine:401) window counter incremented to 2
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (console:663) Starting connect process for proto=vnc trans=ssh connhost=xendev0 connuser=root connport= gaddr=127.0.0.1 gport=5900 gtlsport=None gsocket=None
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (sshtunnels:253) Pre-generated ssh command for ginfo: ssh -l root xendev0 sh -c 'nc -q 2>&1 | grep "requires an argument" >/dev/null;if [ $? -eq 0 ] ; then CMD="nc -q 0 127.0.0.1 5900";else CMD="nc 127.0.0.1 5900";fi;eval "$CMD";'
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (sshtunnels:277) Generated tunnel fd=21 for viewer
[Sat, 18 Jun 2016 10:37:56 virt-manager 6712] DEBUG (sshtunnels:202) Opened tunnel PID=6761 ERRFD=19
[Sat, 18 Jun 2016 10:37:58 virt-manager 6712] DEBUG (console:781) Viewer connected
On Mon, Jun 13, 2016 at 7:42 AM, Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx> wrote:
On Fri, Jun 10, 2016 at 08:19:11AM -0600, John Busch wrote:
> Hello,
>
> I have a Linux 3.2 VM panicking when trying to boot (HVM) on Xen.
>
> I think this is related to the virtual storage controller that is presented
> to the VM.
>
> The qcow2 image runs fine on KVM, but on Xen panics during boot with "VFS:
> Cannot open root device "(null)" or unknown-block(0,0)".
Can you run 'virt-manager' with --debug to see what kind of XML it constructs pls?
Thanks.
> -->
> Both the KVM and Xen hosts are running stock Fedora 23, and virt-manager
> was used to configure both VMs.
>
> To troubleshoot, I cycled the guest image through every permutation of disk
> type (IDE, SCSI, SATA) using virt-manager on both KVM and Xen. Please see
> below for a difference between Xen and KVM.
>
> ---
>
> On Xen, the guest's boot menu reports the following for each disk type:
>
> SATA disk:
> lsi 00:02.0 1:0 Drive QEMU QEMU HARDDISK 2.5+
> Guest panics
>
> SCSI disk:
> lsi 00:02.0 1:0 Drive QEMU QEMU HARDDISK 2.5+
> Guest panics
>
> IDE disk:
> ata0-0: QEMU HARDDISK ATA-7 Hard-Disk
> Guest panics
>
> ---
>
> On KVM, the guest's boot menu reports the following for each disk type:
>
> SATA disk:
> ACHI/0: QEMU HARDDISK ATA-7 Hard-Disk
> ***Guest works!
>
> SCSI disk:
> lsi 00:02.0 1:0 Drive QEMU QEMU HARDDISK 2.5+
> Guest panics
>
> IDE disk:
> ata0-0: QEMU HARDDISK ATA-7 Hard-Disk
> Guest panics
>
> ---
>
> Please note the difference in the SATA disk between Xen and KVM. As far as
> I know, both are using the same version of QEMU. Does anyone know how to
> get the SATA disk to show up as an ACHI disk (instead of lsi) using
> virt-manager/libvirt on Xen?
>
> Thanks in advance!
>
> ~Micah
> xen mailing list
> xen@xxxxxxxxxxxxxxxxxxxxxxx
> https://lists.fedoraproject.org/admin/lists/xen@xxxxxxxxxxxxxxxxxxxxxxx
-- xen mailing list xen@xxxxxxxxxxxxxxxxxxxxxxx https://lists.fedoraproject.org/admin/lists/xen@xxxxxxxxxxxxxxxxxxxxxxx
