Re: [PATCH 10/10] hw/arm/virt: Fix devicetree warnings about the virtio-iommu node

[Peter Maydell <peter.maydell@linaro.org>]

On Wed, 24 Aug 2022 at 16:51, Jean-Philippe Brucker
<jean-philippe@linaro.org> wrote:
>
> dt-validate and dtc throw a few warnings when parsing the virtio-iommu
> node:
>
>   pcie@10000000: virtio_iommu@16:compatible: ['virtio,pci-iommu'] does not contain items matching the given schema
>   pcie@10000000: Unevaluated properties are not allowed (... 'virtio_iommu@16' were unexpected)
>   From schema: linux/Documentation/devicetree/bindings/pci/host-generic-pci.yaml
>   pcie@10000000: virtio_iommu@16:compatible: ['virtio,pci-iommu'] does not contain items matching the given schema
>   From schema: dtschema/schemas/pci/pci-bus.yaml
>
>   Warning (pci_device_reg): /pcie@10000000/virtio_iommu@16: PCI unit address format error, expected "2,0"
>
> The compatible property for a PCI child node should follow the rules
> from "PCI Bus Binding to: IEEE Std 1275-1994". It should contain the
> Vendor ID and Device ID (or class code).
>
> The unit-name should be "device,function".
>
> Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
> ---
> Note that this doesn't follow
> linux/Documentation/devicetree/bindings/virtio/iommu.txt, I'll update
> that document when converting it to yaml, hopefully this Linux cycle.
> The "virtio,pci-iommu" compatible string is not actually used by any
> driver and only QEMU implements it, so we can get rid of it.

I'm not sure you can just change the compat string like that,
unless you can guarantee that nobody anywhere has ever
looked for it in a dtb. Also, "virtio,pci-iommu" is much
clearer than "pci1af4,1057"...

-- PMM