Re: [PATCH v2 3/8] resource: Introduce resource_assigned() for discerning active resources

[Jonathan Cameron <jonathan.cameron@huawei.com>]

On Wed, 12 Nov 2025 18:14:41 -0800
Dan Williams <dan.j.williams@intel.com> wrote:

> A PCI bridge resource lifecycle involves both a "request" and "assign"
> phase. At any point in time that resource may not yet be assigned, or may
> have failed to assign (because it does not fit).
> 
> There are multiple conventions to determine when assignment has not
> completed: IORESOURCE_UNSET, IORESOURCE_DISABLED, and checking whether the
> resource is parented.
> 
> In code paths that are known to not be racing assignment, e.g. post
> subsys_initcall(), the most reliable method to judge that a bridge resource
> is assigned is to check the resource is parented [1].
> 
> Introduce a resource_assigned() helper for this purpose.
> 
> Link: http://lore.kernel.org/2b9f7f7b-d6a4-be59-14d4-7b4ffccfe373@linux.intel.com [1]
> Suggested-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
> Cc: Bjorn Helgaas <bhelgaas@google.com>
> Signed-off-by: Dan Williams <dan.j.williams@intel.com>

Given you replied (and I'm happy to accept that reply) to the doc formatting comment
I had on v1.  This LGTM
Reviewed-by: Jonathan Cameron <jonathan.cameron@huawei.com>

Nice to have a follow up series applying this more widely if anyone has time.
Lots of them in pci/setup-bus.c for instance. 
One of those is doing an assignment check but maybe isn't good to change as it is

if (res->parent)
	return res->parent;

Others all look like low hanging fruit for the readability improvement this
brings.


> ---
>  include/linux/ioport.h | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/include/linux/ioport.h b/include/linux/ioport.h
> index e8b2d6aa4013..9afa30f9346f 100644
> --- a/include/linux/ioport.h
> +++ b/include/linux/ioport.h
> @@ -334,6 +334,15 @@ static inline bool resource_union(const struct resource *r1, const struct resour
>  	return true;
>  }
>  
> +/*
> + * Check if this resource is added to a resource tree or detached. Caller is
> + * responsible for not racing assignment.
> + */
> +static inline bool resource_assigned(struct resource *res)
> +{
> +	return res->parent;
> +}
> +
>  int find_resource_space(struct resource *root, struct resource *new,
>  			resource_size_t size, struct resource_constraint *constraint);
>