On 2022-10-20 22:20, Yang Yingliang wrote:
> kset_register() is currently used in some places without calling
> kset_put() in error path, because the callers think it should be
> kset internal thing to do, but the driver core can not know what
> caller doing with that memory at times. The memory could be freed
> both in kset_put() and error path of caller, if it is called in
> kset_register().
>
> So make the function documentation more explicit about calling
> kset_put() in the error path of caller.
>
> Signed-off-by: Yang Yingliang <yangyingliang@xxxxxxxxxx>
> ---
> lib/kobject.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/lib/kobject.c b/lib/kobject.c
> index a0b2dbfcfa23..6da04353d974 100644
> --- a/lib/kobject.c
> +++ b/lib/kobject.c
> @@ -834,6 +834,9 @@ EXPORT_SYMBOL_GPL(kobj_sysfs_ops);
> /**
> * kset_register() - Initialize and add a kset.
> * @k: kset.
> + *
> + * If this function returns an error, kset_put() must be called to
> + * properly clean up the memory associated with the object.
> */
And I'd continue the sentence, with " ... with the object,
for instance the memory for the kset.kobj.name when kobj_set_name(&kset.kobj, format, ...)
was called before calling kset_register()."
This makes it clear what we want to make sure is freed, in case of an early error
from kset_register().
Regards,
Luben
> int kset_register(struct kset *k)
> {
