[PATCH] tick/nohz: Expose housekeeping CPUs in sysfs

[PATCH] tick/nohz: Expose housekeeping CPUs in sysfs

[Aaron Tomlin]

Expose the current system-defined list of housekeeping CPUs in a new
sysfs file: /sys/devices/system/cpu/housekeeping.

This provides userspace performance tuning tools and resource managers
with a canonical, reliable method to accurately identify the cores
responsible for essential kernel maintenance workloads (RCU, timer
callbacks, and unbound workqueues). Currently, tooling must manually
calculate the housekeeping set by parsing complex kernel boot parameters
(like isolcpus= and nohz_full=) and system topology, which is prone to
error. This dedicated file simplifies the configuration of low-latency
workloads.

Signed-off-by: Aaron Tomlin <atomlin@atomlin.com>
---
 drivers/base/cpu.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
index f694b4b24d6c..c6511658f6cd 100644
--- a/drivers/base/cpu.c
+++ b/drivers/base/cpu.c
@@ -300,6 +300,19 @@ static ssize_t print_cpus_isolated(struct device *dev,
 }
 static DEVICE_ATTR(isolated, 0444, print_cpus_isolated, NULL);
 
+static ssize_t print_cpus_hk(struct device *dev,
+			     struct device_attribute *attr, char *buf)
+{
+	const struct cpumask *hk_mask;
+
+	hk_mask = housekeeping_cpumask(HK_TYPE_KERNEL_NOISE);
+
+	return housekeeping_enabled(HK_TYPE_KERNEL_NOISE) ?
+		sysfs_emit(buf, "%*pbl\n", cpumask_pr_args(hk_mask)) :
+		sysfs_emit(buf, "\n");
+}
+static DEVICE_ATTR(housekeeping, 0444, print_cpus_hk, NULL);
+
 #ifdef CONFIG_NO_HZ_FULL
 static ssize_t print_cpus_nohz_full(struct device *dev,
 				    struct device_attribute *attr, char *buf)
@@ -507,6 +520,7 @@ static struct attribute *cpu_root_attrs[] = {
 	&dev_attr_offline.attr,
 	&dev_attr_enabled.attr,
 	&dev_attr_isolated.attr,
+	&dev_attr_housekeeping.attr,
 #ifdef CONFIG_NO_HZ_FULL
 	&dev_attr_nohz_full.attr,
 #endif
-- 
2.49.0

Re: [PATCH] tick/nohz: Expose housekeeping CPUs in sysfs

[Greg KH]

On Sun, Oct 05, 2025 at 08:58:53PM -0400, Aaron Tomlin wrote:
> Expose the current system-defined list of housekeeping CPUs in a new
> sysfs file: /sys/devices/system/cpu/housekeeping.

But you did not document that new file in Documentation/ABI so it's a
bit hard to review if this is really correct or not.

> This provides userspace performance tuning tools and resource managers
> with a canonical, reliable method to accurately identify the cores
> responsible for essential kernel maintenance workloads (RCU, timer
> callbacks, and unbound workqueues). Currently, tooling must manually
> calculate the housekeeping set by parsing complex kernel boot parameters
> (like isolcpus= and nohz_full=) and system topology, which is prone to
> error. This dedicated file simplifies the configuration of low-latency
> workloads.
> 
> Signed-off-by: Aaron Tomlin <atomlin@atomlin.com>
> ---
>  drivers/base/cpu.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
> 
> diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
> index f694b4b24d6c..c6511658f6cd 100644
> --- a/drivers/base/cpu.c
> +++ b/drivers/base/cpu.c
> @@ -300,6 +300,19 @@ static ssize_t print_cpus_isolated(struct device *dev,
>  }
>  static DEVICE_ATTR(isolated, 0444, print_cpus_isolated, NULL);
>  
> +static ssize_t print_cpus_hk(struct device *dev,
> +			     struct device_attribute *attr, char *buf)
> +{
> +	const struct cpumask *hk_mask;
> +
> +	hk_mask = housekeeping_cpumask(HK_TYPE_KERNEL_NOISE);
> +
> +	return housekeeping_enabled(HK_TYPE_KERNEL_NOISE) ?
> +		sysfs_emit(buf, "%*pbl\n", cpumask_pr_args(hk_mask)) :
> +		sysfs_emit(buf, "\n");

Please write out if statements, it makes it easier over time to
maintain.

> +static DEVICE_ATTR(housekeeping, 0444, print_cpus_hk, NULL);

DEVICE_ATTR_RO()?

thanks,

greg k-h