On 07/03/2018 09:37 AM, Yalan Zhang wrote:
> Hi,
>
> I have a guest enabled vIOMMU, but on the guest there are several
> devices in the same iommu group.
> Could someone help to check if I missed something?
> Thank you very much!
>
> 1. guest xml:
> # virsh edit q
> ...
>
> hvm
> type='pflash'>/usr/share/OVMF/OVMF_CODE.secboot.fd
> /var/lib/libvirt/qemu/nvram/q_VARS.fd
>
> ...
>
> ...
>
>
>
>
>
> ...
>
> ...
>
>
>
> function='0x5'/>
>
>
>
>
>
> function='0x0'/>
>
>
>
>
>
> ...
> 2. guest has 'intel_iommu=on' enabled in kernel cmdline, then reboot guest
>
> 3. log in guest to check:
> # dmesg | grep -i DMAR
> [ 0.00] ACPI: DMAR 7d83f000 00050 (v01 BOCHS BXPCDMAR
> 0001 BXPC 0001)
> [ 0.00] DMAR: IOMMU enabled
> [ 0.155178] DMAR: Host address width 39
> [ 0.155180] DMAR: DRHD base: 0x00fed9 flags: 0x1
> [ 0.155221] DMAR: dmar0: reg_base_addr fed9 ver 1:0 cap
> 12008c22260286 ecap f00f5e
> [ 0.155228] DMAR: ATSR flags: 0x1
> [ 0.155231] DMAR-IR: IOAPIC id 0 under DRHD base 0xfed9 IOMMU 0
> [ 0.155232] DMAR-IR: Queued invalidation will be enabled to support
> x2apic and Intr-remapping.
> [ 0.156843] DMAR-IR: Enabled IRQ remapping in x2apic mode
> [ 2.112369] DMAR: No RMRR found
> [ 2.112505] DMAR: dmar0: Using Queued invalidation
> [ 2.112669] DMAR: Setting RMRR:
> [ 2.112671] DMAR: Prepare 0-16MiB unity mapping for LPC
> [ 2.112820] DMAR: Setting identity map for device :00:1f.0 [0x0 -
> 0xff]
> [ 2.211577] DMAR: Intel(R) Virtualization Technology for Directed I/O
> ===> This is expected
>
> # dmesg | grep -i iommu |grep device
> [ 2.212267] iommu: Adding device :00:00.0 to group 0
> [ 2.212287] iommu: Adding device :00:01.0 to group 1
> [ 2.212372] iommu: Adding device :00:02.0 to group 2
> [ 2.212392] iommu: Adding device :00:02.1 to group 2
> [ 2.212411] iommu: Adding device :00:02.2 to group 2
> [ 2.212444] iommu: Adding device :00:02.3 to group 2
> [ 2.212464] iommu: Adding device :00:02.4 to group 2
> [ 2.212482] iommu: Adding device :00:02.5 to group 2
> [ 2.212520] iommu: Adding device :00:1d.0 to group 3
> [ 2.212533] iommu: Adding device :00:1d.1 to group 3
> [ 2.212541] iommu: Adding device :00:1d.2 to group 3
> [ 2.212550] iommu: Adding device :00:1d.7 to group 3
> [ 2.212567] iommu: Adding device :00:1f.0 to group 4
> [ 2.212576] iommu: Adding device :00:1f.2 to group 4
> [ 2.212585] iommu: Adding device :00:1f.3 to group 4
> [ 2.212599] iommu: Adding device :01:00.0 to group 2
> [ 2.212605] iommu: Adding device :02:01.0 to group 2
> [ 2.212621] iommu: Adding device :04:00.0 to group 2
> [ 2.212634] iommu: Adding device :05:00.0 to group 2
> [ 2.212646] iommu: Adding device :06:00.0 to group 2
> [ 2.212657] iommu: Adding device :07:00.0 to group 2
> > several devices in the same iommu group
>
> # virsh nodedev-dumpxml pci__07_00_0
>
> pci__07_00_0
> /sys/devices/pci:00/:00:02.5/:07:00.0
> pci__00_02_5
>
> e1000
>
>
> 0
> 7
> 0
> 0
> 82540EM Gigabit Ethernet Controller
> Intel Corporation
>
>
>
>
>
>
>
>
>
>
>
>
>
>
All devices plugged into the same slot on pcie-root are in the same
iommu group, and all devices are in the same iommu group as the
pcie-root-port they are plugged into. Since libvirt groups 8
pcie-root-ports onto a single pcie-root slot (to conserve the relatively
limited number of slots), all the devices plugged into those 8
pcie-root-ports are in the same iommu group.
This can be solved by putting the pcie-root-ports *of the devices you
want to assign to L2 guests* onto pcie-root slots by themselves, i.e.
instead of addressing it at 00:02.5, address it at 00:10.0 (and don't
put any other pcie-root-ports at 00:10.x).
I had thought there was an open BZ about this, but can't find one
anywhere. At first glance it would seem that, since the pcie-root-ports
and iommu controller are all emulated devices, it would be simple to
just report all the pcie-root-ports as being on different iommu groups
regardless of their placement on pcie-root, but I believe when this was
brought up doubt was expressed over whether or not qemu truly can
guarantee that the devices are fully isolated from each other.
>
>
>
> Thus, can not attach the device to L2 guest:
> # cat hostdev.xml
>
>
>
>
>
> # virsh attach-device rhel hostdev.xml
> error: Failed to attach device from hostdev.xml
> error: internal error: unable to execute QEMU command 'device_add': vfio
> error: :07:00.0: group 2 is not viable
>
>
> ---
> Best Regards,
> Yalan Zhang
> IRC: yalzh