Re: [PATCH 4/7] x86/resctrl: Add code to setup monitoring at L3 or NODE scope.

[James Morse]

Hi Tony,

On 26/01/2023 18:41, Tony Luck wrote:
> When Sub-NUMA cluster is enabled (snc_ways > 1) use the RDT_RESOURCE_NODE
> instead of RDT_RESOURCE_L3 for all monitoring operations.
> 
> The mon_scale and num_rmid values from CPUID(0xf,0x1),(EBX,ECX) must be
> scaled down by the number of Sub-NUMA Clusters.
> 
> A subsequent change will detect sub-NUMA cluster mode and set
> "snc_ways". For now set to one (meaning each L3 cache spans one
> node).

(I'm looking at decoupling "monitoring is always on RDT_RESOURCE_L3" as a separate thing
 to enabling SNC ... just in case my comments seem strange!)


> diff --git a/arch/x86/kernel/cpu/resctrl/core.c b/arch/x86/kernel/cpu/resctrl/core.c
> index 19be6fe42ef3..53b2ab37af2f 100644
> --- a/arch/x86/kernel/cpu/resctrl/core.c
> +++ b/arch/x86/kernel/cpu/resctrl/core.c
> @@ -786,7 +791,13 @@ static __init bool get_rdt_alloc_resources(void)
>  
>  static __init bool get_rdt_mon_resources(void)
>  {
> -	struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_L3].r_resctrl;
> +	struct rdt_resource *r;
> +
> +	/* When SNC enabled, monitor functions at node instead of L3 cache scope */
> +	if (snc_ways > 1)
> +		r = &rdt_resources_all[RDT_RESOURCE_NODE].r_resctrl;
> +	else
> +		r = &rdt_resources_all[RDT_RESOURCE_L3].r_resctrl;

Could this be hidden in a helper with some name like resctrl_arch_get_mbm_resource()?
You have the same pattern again in rdt_get_tree(). If this gets more complex in the
future, it means its outside the filesystem code, and all in one place.


Thanks,

James


>  	if (rdt_cpu_has(X86_FEATURE_CQM_OCCUP_LLC))
>  		rdt_mon_features |= (1 << QOS_L3_OCCUP_EVENT_ID);

> diff --git a/arch/x86/kernel/cpu/resctrl/rdtgroup.c b/arch/x86/kernel/cpu/resctrl/rdtgroup.c
> index a6ba3080e5db..a0dc64a70d01 100644
> --- a/arch/x86/kernel/cpu/resctrl/rdtgroup.c
> +++ b/arch/x86/kernel/cpu/resctrl/rdtgroup.c
> @@ -2238,7 +2238,10 @@ static int rdt_get_tree(struct fs_context *fc)
>  		static_branch_enable_cpuslocked(&rdt_enable_key);
>  
>  	if (is_mbm_enabled()) {
> -		r = &rdt_resources_all[RDT_RESOURCE_L3].r_resctrl;
> +		if (snc_ways > 1)
> +			r = &rdt_resources_all[RDT_RESOURCE_NODE].r_resctrl;
> +		else
> +			r = &rdt_resources_all[RDT_RESOURCE_L3].r_resctrl;
>  		list_for_each_entry(dom, &r->domains, list)
>  			mbm_setup_overflow_handler(dom, MBM_OVERFLOW_INTERVAL);
>  	}