On Tue, Jan 28, 2025 at 09:42:26AM +0100, Sebastian Andrzej Siewior wrote:
> Using RCU lifetime rules to access kernfs_node::name can avoid the
> trouble kernfs_rename_lock in kernfs_name() and kernfs_path_from_node()
> if the fs was created with KERNFS_ROOT_INVARIANT_PARENT. This is useful
> as it allows to implement kernfs_path_from_node() only with RCU
> protection and avoiding kernfs_rename_lock. The lock is only required if
> the __parent node can be changed and the function requires an unchanged
> hierarchy while it iterates from the node to its parent.
A short mention of how avoiding kernfs_rename_lock matters would be great -
ie. where did this show up?
> diff --git a/arch/x86/kernel/cpu/resctrl/internal.h b/arch/x86/kernel/cpu/resctrl/internal.h
> index 955999aecfca9..cc83dbc70a8ef 100644
> --- a/arch/x86/kernel/cpu/resctrl/internal.h
> +++ b/arch/x86/kernel/cpu/resctrl/internal.h
> @@ -505,6 +505,11 @@ int parse_bw(struct rdt_parse_data *data, struct resctrl_schema *s,
>
> extern struct mutex rdtgroup_mutex;
>
> +static inline const char *rdt_kn_get_name(const struct kernfs_node *kn)
> +{
> + return rcu_dereference_check(kn->name, lockdep_is_held(&rdtgroup_mutex));
> +}
Maybe rdt_kn_name()?
Other than those,
Acked-by: Tejun Heo <tj@xxxxxxxxxx>
Thanks.
--
tejun
