[vfio-users] can't get my video card to connect to my vm

[Alex Williamson]

On Mon, Mar 7, 2016 at 10:34 AM, Stein van Broekhoven <stein at aapjeisbaas.nl>
wrote:

> Dear All,
>
> I can't get the video card which is using vfio to connect to my vm:
> lspci -nnk
> 06:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc.
> [AMD/ATI] Turks XT [Radeon HD 6670/7670] [1002:6758]
>         Subsystem: ASUSTeK Computer Inc. Device [1043:03ea]
>         Kernel driver in use: vfio-pci
>         Kernel modules: radeon, fglrx
> 06:00.1 Audio device [0403]: Advanced Micro Devices, Inc. [AMD/ATI]
> Turks/Whistler HDMI Audio [Radeon HD 6000 Series] [1002:aa90]
>         Subsystem: ASUSTeK Computer Inc. Device [1043:aa90]
>         Kernel driver in use: vfio-pci
>         Kernel modules: snd_hda_intel
>
> I start the vm from cli with qemu command with sudo
> The devices 06:00.0 and 06:00.1 are together in group 22
> I added /dev/vfio/22 to acl in qemu config
>

You should never need to touch this acl list, besides it's for libvirt
which you're not using if you're using the qemu command.


> When using vfio_iommu_type1.allow_unsafe_interrupts=1 on start vm I get
> the error:
> stdout:
> qemu-system-x86_64: -device
> vfio-pci,host=06:00.0,addr=09.0,multifunction=on: vfio: Error: Failed to
> setup INTx fd: Device or resource busy
> qemu-system-x86_64: -device
> vfio-pci,host=06:00.0,addr=09.0,multifunction=on: Device initialization
> failed
> dmesg:
> genirq: Flags mismatch irq 55. 00000080 (vfio-intx(0000:06:00.0)) vs.
> 00000800 (fglrx[1]@PCI:6:0:0)
>
> Without vfio_iommu_type1.allow_unsafe_interrupts=1 on start vm I get the
> error:
> stdout:
> vfio: failed to set iommu for container: Operation not permitted
> dmesg:
> vfio_iommu_type1_attach_group: No interrupt remapping support.  Use the
> module param "allow_unsafe_interrupts" to enable VFIO IOMMU support on this
> platform
>
> system:
> sabertooth 990FX v1.1
> AMD FX-8120
> main GPU: Radeon HD 5xxx
> VM GPU: Radeon HD 6xxx
> Manjaro (arch) kernel 4.1
>

looks like fglrx is still holding onto the IRQ and there's a flags
mismatch, you should prevent fglrx from claiming the GPU by using pci-stub
or vfio-pci to claim the device directly.  990FX should support interrupt
remapping, so it's likely not enabled in your kernel.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20160307/751d88ae/attachment.htm>