On Thu, Sep 25, 2025 at 20:07:05 +0200, Andrea Bolognani via Devel wrote:
> These checks enforce some expectations that were, until now,
> documented solely through comments or not spelled out at all.
> 
> Signed-off-by: Andrea Bolognani <[email protected]>
> Reviewed-by: Michal Privoznik <[email protected]>
> ---
>  src/qemu/qemu_postparse.c | 26 +++++++++++++++++++++++---
>  1 file changed, 23 insertions(+), 3 deletions(-)
> 
> diff --git a/src/qemu/qemu_postparse.c b/src/qemu/qemu_postparse.c
> index d5708fbff9..cf08e6e07c 100644
> --- a/src/qemu/qemu_postparse.c
> +++ b/src/qemu/qemu_postparse.c
> @@ -1384,6 +1384,29 @@ qemuDomainDefAddDefaultDevices(virQEMUDriver *driver,
>          break;
>      }
>  
> +    /* Sanity check. If the machine type does not support PCI, asking
> +     * for PCI(e) root to be added is an obvious mistake */
> +    if ((addPCIRoot ||
> +         addPCIeRoot) &&
> +        !qemuDomainSupportsPCI(def)) {
> +        virReportError(VIR_ERR_INTERNAL_ERROR,
> +                       _("Machine type '%1$s' wants PCI but PCI is not 
> supported"),
> +                       def->os.machine);
> +        return -1;
> +    }
> +
> +    /* Sanity check. If the machine type supports PCI, we need to reflect
> +     * that fact in the XML or other parts of the machine handling code
> +     * might misbehave */
> +    if (qemuDomainSupportsPCI(def) &&
> +        !addPCIRoot &&
> +        !addPCIeRoot) {
> +        virReportError(VIR_ERR_INTERNAL_ERROR,
> +                       _("Machine type '%1$s' supports PCI but no PCI 
> controller added"),
> +                       def->os.machine);
> +        return -1;
> +    }

I still don't really like that we add sanity checks into (effectively)
the XML parser, but for now I don' thave any better suggestion.

Please don't add any other.

Reviewed-by: Peter Krempa <[email protected]>

Reply via email to