Re: [PATCH 1/6] dax/hmem, e820, resource: Defer Soft Reserved registration until hmem is ready

[Koralahalli Channabasappa, Smita]

Hi Boris,

On 9/9/2025 9:12 AM, Borislav Petkov wrote:
> On Fri, Aug 22, 2025 at 03:41:57AM +0000, Smita Koralahalli wrote:
> > diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c
> > index c3acbd26408b..aef1ff2cabda 100644
> > --- a/arch/x86/kernel/e820.c
> > +++ b/arch/x86/kernel/e820.c
> > @@ -1153,7 +1153,7 @@ void __init e820__reserve_resources_late(void)
> >   	res = e820_res;
> >   	for (i = 0; i < e820_table->nr_entries; i++) {
> >   		if (!res->parent && res->end)
> > -			insert_resource_expand_to_fit(&iomem_resource, res);
> > +			insert_resource_late(res);
> >   		res++;
> >   	}
>
> Btw, this doesn't even build and cover letter doesn't say what it applies
> ontop so I applied it on my pile of tip/master.
>
> kernel/resource.c: In function ‘region_intersects_soft_reserve’:
> kernel/resource.c:694:36: error: ‘soft_reserve_resource’ undeclared (first use in this function); did you mean ‘devm_release_resource’?
>    694 |         ret = __region_intersects(&soft_reserve_resource, start, size, flags,
>        |                                    ^~~~~~~~~~~~~~~~~~~~~
>        |                                    devm_release_resource
> kernel/resource.c:694:36: note: each undeclared identifier is reported only once for each function it appears in
> make[3]: *** [scripts/Makefile.build:287: kernel/resource.o] Error 1
> make[2]: *** [scripts/Makefile.build:556: kernel] Error 2
> make[2]: *** Waiting for unfinished jobs....
> make[1]: *** [/mnt/kernel/kernel/2nd/linux/Makefile:2011: .] Error 2
> make: *** [Makefile:248: __sub-make] Error 2

Apologies for the delay.

This was based on mainline. I have rebased the series onto the latest 
mainline and sent out a new revision and noted it in the cover letter.

https://lore.kernel.org/all/20250930044757.214798-2-Smita.KoralahalliChannabasappa@xxxxxxx/

> Also, I'd do this resource insertion a bit differently:
>
> insert_resource_expand_to_fit(struct resource *new)
> {
> 	struct resource *root = &iomem_resource;
>
> 	if (new->desc == IORES_DESC_SOFT_RESERVED)
> 		root = &soft_reserve_resource;
>
> 	return __insert_resource_expand_to_fit(root, new);
> }
>
> and rename the current insert_resource_expand_to_fit() to the __ variant.

I have made these changes as well.

Thanks
Smita

> It looks like you want to intercept all callers of
> insert_resource_expand_to_fit() instead of defining a separate set which works
> on the soft-reserve thing.
>
> Oh well, the resource code is yucky already.