Re: [PATCH v6 12/24] x86/resctrl: Add cpumask_any_housekeeping() for limbo/overflow

[Reinette Chatre]

Hi James,

On 9/14/2023 10:21 AM, James Morse wrote:
> The limbo and overflow code picks a CPU to use from the domain's list
> of online CPUs. Work is then scheduled on these CPUs to maintain
> the limbo list and any counters that may overflow.
> 
> cpumask_any() may pick a CPU that is marked nohz_full, which will
> either penalise the work that CPU was dedicated to, or delay the
> processing of limbo list or counters that may overflow. Perhaps
> indefinitely. Delaying the overflow handling will skew the bandwidth
> values calculated by mba_sc, which expects to be called once a second.
> 
> Add cpumask_any_housekeeping() as a replacement for cpumask_any()
> that prefers housekeeping CPUs. This helper will still return
> a nohz_full CPU if that is the only option. The CPU to use is
> re-evaluated each time the limbo/overflow work runs. This ensures
> the work will move off a nohz_full CPU once a housekeeping CPU is
> available.
> 
> Reviewed-by: Shaopeng Tan <tan.shaopeng@xxxxxxxxxxx>
> Tested-by: Shaopeng Tan <tan.shaopeng@xxxxxxxxxxx>
> Tested-By: Peter Newman <peternewman@xxxxxxxxxx>
> Signed-off-by: James Morse <james.morse@xxxxxxx>
> ---
> Changes since v3:
>  * typos fixed
> 
> Changes since v4:
>  * Made temporary variables unsigned
> 
> Changes since v5:
>  * Restructured cpumask_any_housekeeping() to avoid later churn.
> ---
>  arch/x86/kernel/cpu/resctrl/internal.h | 24 ++++++++++++++++++++++++
>  arch/x86/kernel/cpu/resctrl/monitor.c  | 17 ++++++++++++-----
>  2 files changed, 36 insertions(+), 5 deletions(-)
> 
> diff --git a/arch/x86/kernel/cpu/resctrl/internal.h b/arch/x86/kernel/cpu/resctrl/internal.h
> index f06d3d3e0808..37bb3de37a4a 100644
> --- a/arch/x86/kernel/cpu/resctrl/internal.h
> +++ b/arch/x86/kernel/cpu/resctrl/internal.h
> @@ -7,6 +7,7 @@
>  #include <linux/kernfs.h>
>  #include <linux/fs_context.h>
>  #include <linux/jump_label.h>
> +#include <linux/tick.h>
>  #include <asm/resctrl.h>
>  

Please maintain the empty line between groups of headers.


...

> diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c b/arch/x86/kernel/cpu/resctrl/monitor.c
> index 0bbed8c62d42..993837e46db1 100644
> --- a/arch/x86/kernel/cpu/resctrl/monitor.c
> +++ b/arch/x86/kernel/cpu/resctrl/monitor.c
> @@ -782,9 +782,9 @@ static void mbm_update(struct rdt_resource *r, struct rdt_domain *d,
>  void cqm_handle_limbo(struct work_struct *work)
>  {
>  	unsigned long delay = msecs_to_jiffies(CQM_LIMBOCHECK_INTERVAL);
> -	int cpu = smp_processor_id();
>  	struct rdt_resource *r;
>  	struct rdt_domain *d;
> +	int cpu;
>  
>  	mutex_lock(&rdtgroup_mutex);
>  
> @@ -793,8 +793,10 @@ void cqm_handle_limbo(struct work_struct *work)
>  
>  	__check_limbo(d, false);
>  
> -	if (has_busy_rmid(d))
> +	if (has_busy_rmid(d)) {
> +		cpu = cpumask_any_housekeeping(&d->cpu_mask);
>  		schedule_delayed_work_on(cpu, &d->cqm_limbo, delay);
> +	}
>  

ok - but if you do change the CPU the worker is running on then
I also expect d->cqm_work_cpu to be updated. Otherwise the offline
code will not be able to determine if the worker needs to move.

>  	mutex_unlock(&rdtgroup_mutex);
>  }
> @@ -804,7 +806,7 @@ void cqm_setup_limbo_handler(struct rdt_domain *dom, unsigned long delay_ms)
>  	unsigned long delay = msecs_to_jiffies(delay_ms);
>  	int cpu;
>  
> -	cpu = cpumask_any(&dom->cpu_mask);
> +	cpu = cpumask_any_housekeeping(&dom->cpu_mask);
>  	dom->cqm_work_cpu = cpu;
>  
>  	schedule_delayed_work_on(cpu, &dom->cqm_limbo, delay);
> @@ -814,10 +816,10 @@ void mbm_handle_overflow(struct work_struct *work)
>  {
>  	unsigned long delay = msecs_to_jiffies(MBM_OVERFLOW_INTERVAL);
>  	struct rdtgroup *prgrp, *crgrp;
> -	int cpu = smp_processor_id();
>  	struct list_head *head;
>  	struct rdt_resource *r;
>  	struct rdt_domain *d;
> +	int cpu;
>  
>  	mutex_lock(&rdtgroup_mutex);
>  
> @@ -838,6 +840,11 @@ void mbm_handle_overflow(struct work_struct *work)
>  			update_mba_bw(prgrp, d);
>  	}
>  
> +	/*
> +	 * Re-check for housekeeping CPUs. This allows the overflow handler to
> +	 * move off a nohz_full CPU quickly.
> +	 */
> +	cpu = cpumask_any_housekeeping(&d->cpu_mask);
>  	schedule_delayed_work_on(cpu, &d->mbm_over, delay);
>  

Similar to above I expect a change like this to
be accompanied by an update to d->mbm_work_cpu.

>  out_unlock:
> @@ -851,7 +858,7 @@ void mbm_setup_overflow_handler(struct rdt_domain *dom, unsigned long delay_ms)
>  
>  	if (!static_branch_likely(&rdt_mon_enable_key))
>  		return;
> -	cpu = cpumask_any(&dom->cpu_mask);
> +	cpu = cpumask_any_housekeeping(&dom->cpu_mask);
>  	dom->mbm_work_cpu = cpu;
>  	schedule_delayed_work_on(cpu, &dom->mbm_over, delay);
>  }


Reinette