Re: [PATCH v2] x86,sched: allow topologies where NUMA nodes share an LLC

[Tim Chen]

On 03/22/2018 01:49 PM, Alison Schofield wrote:
>
> +	 */
> +	if (!topology_same_node(c, o) &&
> +	    (c->x86_vendor == X86_VENDOR_INTEL &&
> +	     c->x86_model == INTEL_FAM6_SKYLAKE_X)) {
> +		/* Use NUMA instead of coregroups for scheduling: */
> +		x86_has_numa_in_package = true;

x86_has_numa_in_package will only be set true for SKYLAKE in the above? 

This boolean probably should be set for (!topology_same_node(c, o) && match_die(c, o)) and not
dependent on cpu family.  Only the return value should depend on cpu family.

Tim


> +
> +		/*
> +		 * Return value doesn't actually matter because we
> +		 * are throwing away coregroups for scheduling anyway.
> +		 * Return false to bypass topology broken bug messages
> +		 * and fixups in sched_domain().
> +		 */
> +		return false;
> +	}
> +
> +	return topology_sane(c, o, "llc");
>  }
>  
>  /*
> @@ -454,12 +492,6 @@ static struct sched_domain_topology_level x86_topology[] = {
>  	{ NULL, },
>  };
>  
> -/*
> - * Set if a package/die has multiple NUMA nodes inside.
> - * AMD Magny-Cours and Intel Cluster-on-Die have this.
> - */
> -static bool x86_has_numa_in_package;
> -
>  void set_cpu_sibling_map(int cpu)
>  {
>  	bool has_smt = smp_num_siblings > 1;
>