On Fri 23-10-15 15:06:59, Jeff Layton wrote:
> At the time that this code was originally written, call_srcu didn't
> exist, so this thread was required to ensure that we waited for that
> SRCU grace period to settle before finally freeing the object.
> 
> It does exist now however and we can much more efficiently use call_srcu
> to handle this. That also allows us to potentially use srcu_barrier
> to ensure that they are all of the callbacks have run before proceeding.
> In order to conserve space, we union the rcu_head with the g_list.
> 
> This will be necessary for nfsd which will allocate marks from a
> dedicated slabcache. We have to be able to ensure that all of the
> objects are destroyed before destroying the cache. That's fairly
> difficult to ensure with dedicated thread doing the destruction.

The patch looks good. Just one nit below:

> diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h
> index 533c4408529a..6b7e89f45aa4 100644
> --- a/include/linux/fsnotify_backend.h
> +++ b/include/linux/fsnotify_backend.h
> @@ -220,7 +220,10 @@ struct fsnotify_mark {
>  	/* List of marks by group->i_fsnotify_marks. Also reused for queueing
>  	 * mark into destroy_list when it's waiting for the end of SRCU period
>  	 * before it can be freed. [group->mark_mutex] */

Please update this comment to not speak about destroy_list. After that feel
free to add:

Reviewed-by: Jan Kara <jack@suse.com>

								Honza

> -	struct list_head g_list;
> +	union {
> +		struct list_head g_list;
> +		struct rcu_head g_rcu;
> +	};
>  	/* Protects inode / mnt pointers, flags, masks */
>  	spinlock_t lock;
>  	/* List of marks for inode / vfsmount [obj_lock] */


									Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR