Re: [PATCH v2 1/6] cxl/region: Validate interleave selector bits

[Robert Richter <rrichter@amd.com>]

On 26.06.26 13:37:22, Robert Richter wrote:
> On 11.06.26 10:47:25, Alison Schofield wrote:

> > @@ -2067,6 +2062,11 @@ static int cxl_region_attach(struct cxl_region *cxlr,
> >  		return -ENXIO;
> >  	}
> >  
> > +	/* Keep this patch bisectable until position arithmetic is updated */
> > +	if (cxlrd->cxlsd.cxld.interleave_granularity >
> > +	    p->interleave_granularity)
> > +		return -ENXIO;
> > +
> 
> Hmm, that looks odd, even if it's only temporary. Could you explain
> why that is needed here.

Ok, I missed the point that the iterator stops at !is_cxl_root(iter)
and thus does not consider the cfmws interleaving configs of the root
decoders here, which are not part of the parent selector bit mask
then.

That creates a 4th selector bit mask for the root decoders to keep
track of. With that added I think all the stride handling in patch #3
to calculate the endpoint positions can be simplified.

-Robert