RE: [PATCH v4 03/17] xen/arm: implement node distance helpers for Arm

[Henry Wang <Henry.Wang@xxxxxxx>]

Hi Jan,

> -----Original Message-----
> From: Jan Beulich <jbeulich@xxxxxxxx>
> Subject: Re: [PATCH v4 03/17] xen/arm: implement node distance helpers for
> Arm
> 
> On 25.04.2023 11:31, Henry Wang wrote:
> >> -----Original Message-----
> >> From: Jan Beulich <jbeulich@xxxxxxxx>
> >>
> >> On 25.04.2023 09:56, Henry Wang wrote:
> >>> +    /* NUMA defines NUMA_NO_DISTANCE as unreachable and 0-9 are
> >> undefined */
> >>> +    if ( distance >= NUMA_NO_DISTANCE || distance <=
> >> NUMA_DISTANCE_UDF_MAX ||
> >>> +         (from == to && distance != NUMA_LOCAL_DISTANCE) )
> >>> +    {
> >>> +        printk(KERN_WARNING
> >>> +               "NUMA: invalid distance: from=%"PRIu8" to=%"PRIu8"
> >> distance=%"PRIu32"\n",
> >>> +               from, to, distance);
> >>> +        return;
> >>> +    }
> >>
> >> I appreciate the checking that node-local references are
> >> NUMA_LOCAL_DISTANCE,
> >> but if they're wrongly passed into here, shouldn't the resulting array 
> >> still
> >> have NUMA_LOCAL_DISTANCE on its diagonal, at least as far as present
> nodes
> >> go?
> >
> > Apologies in advance to ask more specific details from you as I am not sure
> > if I can correctly understand the "if they're wrongly passed into here" 
> > case.
> Do you
> > mean we are always guaranteed that if from == to, the distance will always
> be
> > NUMA_LOCAL_DISTANCE so the (from == to && distance !=
> NUMA_LOCAL_DISTANCE)
> > check is redundant and can be removed? Thanks.
> 
> It's a little odd that you ask me: It is your code which insists on
> NUMA_LOCAL_DISTANCE when from == to. If you insist on your caller to pass
> only this one value, then I think it is only logical to also set the
> respective table entries to that value (rather than leaving them at
> NUMA_NO_DISTANCE).

I think I understand what do you mean now. I was only checking this specific
patch so didn't make the connection with the latter patches.

Kind regards,
Henry

> 
> Jan