Re: [PATCH 5/6] dax/hmem: Reintroduce Soft Reserved ranges back into the iomem tree

[Koralahalli Channabasappa, Smita]

Hi Jonathan,

On 9/10/2025 6:41 AM, Jonathan Cameron wrote:
> On Fri, 22 Aug 2025 03:42:01 +0000
> Smita Koralahalli <Smita.KoralahalliChannabasappa@xxxxxxx> wrote:
>
> > Reworked from a patch by Alison Schofield <alison.schofield@xxxxxxxxx>
> >
> > Reintroduce Soft Reserved range into the iomem_resource tree for dax_hmem
> > to consume.
> >
> > This restores visibility in /proc/iomem for ranges actively in use, while
> > avoiding the early-boot conflicts that occurred when Soft Reserved was
> > published into iomem before CXL window and region discovery.
> >
> > Link: https://lore.kernel.org/linux-cxl/29312c0765224ae76862d59a17748c8188fb95f1.1692638817.git.alison.schofield@xxxxxxxxx/
> > Co-developed-by: Alison Schofield <alison.schofield@xxxxxxxxx>
> > Signed-off-by: Alison Schofield <alison.schofield@xxxxxxxxx>
> > Signed-off-by: Smita Koralahalli <Smita.KoralahalliChannabasappa@xxxxxxx>
> A few trivial things inline. Not are important enough to need a change though.
>
> Reviewed-by: Jonathan Cameron <jonathan.cameron@xxxxxxxxxx>
>
> > ---
> >   drivers/dax/hmem/hmem.c | 38 ++++++++++++++++++++++++++++++++++++++
> >   1 file changed, 38 insertions(+)
> >
> > diff --git a/drivers/dax/hmem/hmem.c b/drivers/dax/hmem/hmem.c
> > index 90978518e5f4..24a6e7e3d916 100644
> > --- a/drivers/dax/hmem/hmem.c
> > +++ b/drivers/dax/hmem/hmem.c
> > @@ -93,6 +93,40 @@ static void process_defer_work(struct work_struct *_work)
> >   	walk_hmem_resources(&pdev->dev, handle_deferred_cxl);
> >   }
> >   
> > +static void remove_soft_reserved(void *data)
> > +{
> > +	struct resource *r = data;
> > +
> > +	remove_resource(r);
>
> Type doesn't really help us here so why not skip the local variable.
> 	remove_resource(data);
> 	kfree(data);
>
> Though I'd rename data to r.
>
> > +	kfree(r);

Fixed in v2.

> > +}
> > +
> > +static int add_soft_reserve_into_iomem(struct device *host,
> > +				       const struct resource *res)
> > +{
> > +	struct resource *soft = kzalloc(sizeof(*soft), GFP_KERNEL);
> > +	int rc;
> > +
> > +	if (!soft)
> > +		return -ENOMEM;
> > +
> > +	*soft = DEFINE_RES_NAMED_DESC(res->start, (res->end - res->start + 1),
> > +				      "Soft Reserved", IORESOURCE_MEM,
> > +				      IORES_DESC_SOFT_RESERVED);
> > +
> > +	rc = insert_resource(&iomem_resource, soft);
> > +	if (rc) {
> > +		kfree(soft);
>
> Could use __free() magic here and steal the pointer when you setup the
> devm action below.  Only a small simplification in this case, so up to
> you.

Sure, I have incorporated.

> > +		return rc;
> > +	}
> > +
> > +	rc = devm_add_action_or_reset(host, remove_soft_reserved, soft);
> > +	if (rc)
> > +		return rc;
> > +
> > +	return 0;
>
> Trivial:
>
> 	return dev_add_action_or_reset(host...)

Included.

Thanks,
Smita

> > +}