[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-API] XCP 1.6 and GPU passthrough

Rob Hoes <Rob.Hoes@xxxxxxxxxx> a écrit :

Is your VM a PV guest?

I'm not sure. I installed an ubuntu 12.04 from the iso ubuntu-12.04.1-server-amd64.iso. I guess that the linux kernel in this distribution has some Xen drivers and is a aware that it is running in a virtualized environment. So I guess that it is a PV guest but I'm not 100% sure.

To do this, try the following command to see if your card was detected by XCP:

[root@xcp ~]# xe gpu-group-list uuid ( RO) : cd98c84c-73c6-39aa-4ae2-4844f5718210 name-label ( RW): Group of Matrox Electronics Systems Ltd. MGA G200eW WPCM450 GPUs name-description ( RW): uuid ( RO) : 773b9dfb-4663-9b0a-1122-da80fcb08c2a name-label ( RW): Group of NVIDIA Corporation GF100 [Tesla M2070Q] GPUs
   name-description ( RW): When I start my VM I get the following error:

The following table provides per-host reasons for why the VM could not be started:

xcp    : Cannot start here [Api_errors.Server_error("VM_REQUIRES_IOMMU", _)]

There were no hosts available to complete the specified operation.

So if I understand correctly, that means that my machine cannot do IOMMU?
I checked in XEN logs and I found:

(XEN) VMX: Supported advanced features:
(XEN)  - APIC MMIO access virtualisation
(XEN)  - APIC TPR shadow
(XEN)  - Extended Page Tables (EPT)
(XEN)  - Virtual-Processor Identifiers (VPID)
(XEN)  - Virtual NMI
(XEN)  - MSR direct-access bitmap
(XEN)  - Unrestricted Guest
(XEN) HVM: ASIDs enabled.
(XEN) HVM: VMX enabled
(XEN) HVM: Hardware Assisted Paging (HAP) detected
(XEN) HVM: HAP page sizes: 4kB, 2MB, 1GB

I also checked on ark.intel for my xeon E5640 and it is said that VT-d is supported. I suspect that my BIOS cannot enable it :(


PS: Sorry Rob for the duplicated email, I forgot to send it to the ML the first time.

Xen-api mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.