Re: [PATCH v10 02/28] x86/resctrl: Move L3 initialization into new helper function

["Luck, Tony" <tony.luck@intel.com>]

On Thu, Sep 18, 2025 at 02:49:05PM -0700, Reinette Chatre wrote:
> Hi Tony,
> 
> On 9/12/25 3:10 PM, Tony Luck wrote:
> > All resctrl monitor events are associated with the L3 resource, but
> > this is about to change.
> 
> Please see Boris's feedback about changelogs in [1]. To address that,
> please rework the changelogs to not have so much copy&pasted context 
> in patches. 

I understand that Boris doesn't want to see large amounts of text copied
from the cover letter into each patch. But there is still a need to meet the
"Describe your problem" requirement for a good commit message.

Several of the prepatory patches in this series make changes to expand the
capabilities of fs/resctrl to handle monitor events from resources other
than RDT_RESOURCE_L3.

Is a single short sentence repeated in several patches "too much"?

> 
> > 
> > To prepare for additional types of monitoring domains, move open coded L3
> > resource monitoring domain initialization from domain_add_cpu_mon() into
> > a new helper l3_mon_domain_setup() called by domain_add_cpu_mon().
> > 
> > Signed-off-by: Tony Luck <tony.luck@intel.com>
> > ---
> >  arch/x86/kernel/cpu/resctrl/core.c | 57 +++++++++++++++++-------------
> >  1 file changed, 32 insertions(+), 25 deletions(-)
> > 
> > diff --git a/arch/x86/kernel/cpu/resctrl/core.c b/arch/x86/kernel/cpu/resctrl/core.c
> > index 8be2619db2e7..055df4d406d0 100644
> > --- a/arch/x86/kernel/cpu/resctrl/core.c
> > +++ b/arch/x86/kernel/cpu/resctrl/core.c
> > @@ -496,37 +496,13 @@ static void domain_add_cpu_ctrl(int cpu, struct rdt_resource *r)
> >  	}
> >  }
> >  
> > -static void domain_add_cpu_mon(int cpu, struct rdt_resource *r)
> > +static void l3_mon_domain_setup(int cpu, int id, struct rdt_resource *r, struct list_head *add_pos)
> >  {
> > -	int id = get_domain_id_from_scope(cpu, r->mon_scope);
> > -	struct list_head *add_pos = NULL;
> >  	struct rdt_hw_mon_domain *hw_dom;
> > -	struct rdt_domain_hdr *hdr;
> >  	struct rdt_mon_domain *d;
> >  	struct cacheinfo *ci;
> >  	int err;
> >  
> > -	lockdep_assert_held(&domain_list_lock);
> > -
> > -	if (id < 0) {
> > -		pr_warn_once("Can't find monitor domain id for CPU:%d scope:%d for resource %s\n",
> > -			     cpu, r->mon_scope, r->name);
> > -		return;
> > -	}
> > -
> > -	hdr = resctrl_find_domain(&r->mon_domains, id, &add_pos);
> > -	if (hdr) {
> > -		if (!domain_header_is_valid(hdr, RESCTRL_MON_DOMAIN, r->rid))
> > -			return;
> > -		d = container_of(hdr, struct rdt_mon_domain, hdr);
> > -
> > -		cpumask_set_cpu(cpu, &d->hdr.cpu_mask);
> > -		/* Update the mbm_assign_mode state for the CPU if supported */
> > -		if (r->mon.mbm_cntr_assignable)
> > -			resctrl_arch_mbm_cntr_assign_set_one(r);
> 
> Rebase error? Note the assignable counter initialization done on CPU online ...

Right. This code was lost in the rebase.
> 
> > -		return;
> > -	}
> > -
> >  	hw_dom = kzalloc_node(sizeof(*hw_dom), GFP_KERNEL, cpu_to_node(cpu));
> >  	if (!hw_dom)
> >  		return;
> > @@ -565,6 +541,37 @@ static void domain_add_cpu_mon(int cpu, struct rdt_resource *r)
> >  	}
> >  }
> >  
> > +static void domain_add_cpu_mon(int cpu, struct rdt_resource *r)
> > +{
> > +	int id = get_domain_id_from_scope(cpu, r->mon_scope);
> > +	struct list_head *add_pos = NULL;
> > +	struct rdt_domain_hdr *hdr;
> > +
> > +	lockdep_assert_held(&domain_list_lock);
> > +
> > +	if (id < 0) {
> > +		pr_warn_once("Can't find monitor domain id for CPU:%d scope:%d for resource %s\n",
> > +			     cpu, r->mon_scope, r->name);
> > +		return;
> > +	}
> > +
> > +	hdr = resctrl_find_domain(&r->mon_domains, id, &add_pos);
> > +	if (hdr) {
> > +		cpumask_set_cpu(cpu, &hdr->cpu_mask);
> > +
> 
> ... assignable counter initialization no longer done on CPU online.
> 
> Looking closer the l3_mon_domain_setup() also now contains assignable counter
> initialization that is gated by a RDT_RESOURCE_L3 check. Considering the flow
> I think it may thus be simpler and consistent to not return here
> but instead have the additional initialization done in resource specific
> areas below. 

Yes.
> 
> > +		return;
> > +	}
> > +
> > +	switch (r->rid) {
> > +	case RDT_RESOURCE_L3:
> 
> Something like:
> 		if (hdr) {
> 			/* do resource specific CPU initialization here */
> 			return;
> 		}

In this specific case the resource specific operation needs to happen
on every CPU (it updates the per-CPU MSR_IA32_L3_QOS_EXT_CFG). So I
think the code fragment needs to be:

	switch (r->rid) {
	case RDT_RESOURCE_L3:
		/* Update the mbm_assign_mode state for the CPU if supported */
		if (r->mon.mbm_cntr_assignable)
			resctrl_arch_mbm_cntr_assign_set_one(r);
		if (!hdr)
			l3_mon_domain_setup(cpu, id, r, add_pos);
		break;

> 
> > +		l3_mon_domain_setup(cpu, id, r, add_pos);
> > +		break;
> > +	default:
> > +		pr_warn_once("Unknown resource rid=%d\n", r->rid);
> > +		break;
> > +	}
> > +}
> > +
> >  static void domain_add_cpu(int cpu, struct rdt_resource *r)
> >  {
> >  	if (r->alloc_capable)
> 
> Reinette

-Tony