[Ocfs2-devel] [PATCH 5/6] ocfs2: Fix memory leak on mlog_sys_init error path

Pekka Enberg penberg at cs.helsinki.fi
Fri Jul 23 10:01:33 PDT 2010


Patrick Pannuto wrote:
> This leak is the same as the bus path; mlog_kset.kobj.name is
> set, but kset_register fails, thus nothing will ever call
> kset_unregister and name will be leaked
> 
> Signed-off-by: Patrick Pannuto <ppannuto at codeaurora.org>
> ---
>  fs/ocfs2/cluster/masklog.c |    7 ++++++-
>  1 files changed, 6 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/ocfs2/cluster/masklog.c b/fs/ocfs2/cluster/masklog.c
> index c7fba39..de715d4 100644
> --- a/fs/ocfs2/cluster/masklog.c
> +++ b/fs/ocfs2/cluster/masklog.c
> @@ -154,6 +154,7 @@ static struct kset mlog_kset = {
>  int mlog_sys_init(struct kset *o2cb_kset)
>  {
>  	int i = 0;
> +	int ret;
>  
>  	while (mlog_attrs[i].attr.mode) {
>  		mlog_attr_ptrs[i] = &mlog_attrs[i].attr;
> @@ -163,7 +164,11 @@ int mlog_sys_init(struct kset *o2cb_kset)
>  
>  	kobject_set_name(&mlog_kset.kobj, "logmask");
>  	mlog_kset.kobj.kset = o2cb_kset;
> -	return kset_register(&mlog_kset);
> +
> +	ret = kset_register(&mlog_kset);
> +	if (ret)
> +		kfree(mlog_kset.kobj.name);

Aren't we supposed to use kobject_cleanup() here?

> +	return ret;
>  }
>  
>  void mlog_sys_shutdown(void)




More information about the Ocfs2-devel mailing list