Re: [PATCH v1 09/27] xen/riscv: implement make_intc_domU_node()

[Jan Beulich <jbeulich@xxxxxxxx>]

On 10.03.2026 18:08, Oleksii Kurochko wrote:
> --- a/xen/arch/riscv/dom0less-build.c
> +++ b/xen/arch/riscv/dom0less-build.c
> @@ -3,6 +3,15 @@
>  #include <xen/fdt-kernel.h>
>  #include <xen/init.h>
>  
> +#include <asm/intc.h>
> +
> +int __init make_intc_domU_node(struct kernel_info *kinfo)
> +{
> +    intc_make_domu_dt_node(kinfo);
> +
> +    return 0;
> +}

Is this wrapper really needed? Can't what's intc_make_domu_dt_node() right
now become make_intc_domU_node()?

> @@ -41,6 +41,10 @@ struct intc_hw_operations {
>  
>      /* handle external interrupt */
>      void (*handle_interrupt)(struct cpu_user_regs *regs);
> +
> +    /* Create interrupt controller node for domain */
> +    int (*make_dom_dt_node)(const struct kernel_info *kinfo,
> +                            const struct dt_device_node *intc);

An __init-only hook is somewhat risky, just to mention it. In IOMMU code
besides struct iommu_ops we have struct iommu_init_ops, just to give an
example of where the same could have been the case.

> --- a/xen/arch/riscv/intc.c
> +++ b/xen/arch/riscv/intc.c
> @@ -67,3 +67,11 @@ void intc_route_irq_to_xen(struct irq_desc *desc, unsigned 
> int priority)
>      intc_set_irq_type(desc, desc->arch.type);
>      intc_set_irq_priority(desc, priority);
>  }
> +
> +int __init intc_make_domu_dt_node(const struct kernel_info *kinfo)
> +{
> +    if ( intc_hw_ops && intc_hw_ops->make_dom_dt_node )
> +        return intc_hw_ops->make_dom_dt_node(kinfo, intc_hw_ops->info->node);
> +
> +    return -ENOSYS;

How do you justify this choice of return value? This isn't even a hypercall
handler.

Jan