Re: [PATCH 5/6] virt: steal_governor: Warn on unexpected preferred mask states

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] virt: steal_governor: unreachable WARN_ON() in decrease_preferred_cpus()
--

commit 735c9f0cd33ee22cf920d749e3d1bde33a6cf851
Author: Yury Norov <ynorov@nvidia.com>

virt: steal_governor: Warn on unexpected preferred mask states

This commit adds WARN_ON() checks in decrease_preferred_cpus() and
steal_governor_loop() to warn if cpu_preferred_mask enters an unexpected
state.

> diff --git a/drivers/virt/steal_governor.c b/drivers/virt/steal_governor.c
> index bc3936e69eed9..5679b7ddc078f 100644
> --- a/drivers/virt/steal_governor.c
> +++ b/drivers/virt/steal_governor.c
> @@ -145,7 +145,9 @@ static void decrease_preferred_cpus(void)
>  	/* Always leave first housekeeping core as preferred. */
>  	first_hk_core = topology_sibling_cpumask(cpu);
>  	cpu = cpumask_last(cpu_preferred_mask);
> -	if (cpu >= nr_cpu_ids)
> +
> +	/* We can't end up with no preferred CPUs under normal conditions. */
> +	if (WARN_ON(cpu >= nr_cpu_ids))
>  		return;

[Severity: Low]
Can this WARN_ON() in decrease_preferred_cpus() ever be reached?

Just above this check in drivers/virt/steal_governor.c,
decrease_preferred_cpus() already checks whether the intersection of the
housekeeping cpumask and cpu_preferred_mask is empty while holding
cpus_read_lock() and cpu_preferred_mutex:

	cpu = cpumask_first_and(housekeeping_cpumask(HK_TYPE_KERNEL_NOISE),
				cpu_preferred_mask);
	if (cpu >= nr_cpu_ids)
		return;

If cpu_preferred_mask is empty on entry, cpumask_first_and() returns
nr_cpu_ids and decrease_preferred_cpus() returns early before reaching
cpumask_last(cpu_preferred_mask). Otherwise, at least one bit in
cpu_preferred_mask is set, so cpumask_last(cpu_preferred_mask) is guaranteed
to be less than nr_cpu_ids.

Should this check be removed, or should the warning be placed on the earlier
check if an empty preferred mask is what this is meant to catch?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261101192203.6979-1-ynorov@nvidia.com?part=5