ovl: Cleanup redundant blank lines

This patch removes redundant blanks lines in overlayfs.

---
 fs/overlayfs/copy_up.c |    1 -
 fs/overlayfs/inode.c   |    1 -
 fs/overlayfs/super.c   |    1 -
 3 files changed, 0 insertions(+), 3 deletions(-)

index ea10a87..66d161b 100644
@@ -191,7 +191,6 @@ int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat)
 		ovl_set_timestamps(upperdentry, stat);
 
 	return err;
-
 }
 
 static int ovl_copy_up_locked(struct dentry *workdir, struct dentry *upperdir,

index 07d74b2..8ef1a03 100644
@@ -430,5 +430,4 @@ struct inode *ovl_new_inode(struct super_block *sb, umode_t mode,
 	}
 
 	return inode;
-
 }

index f16d318..bdd1195 100644
@@ -98,7 +98,6 @@ void ovl_path_upper(struct dentry *dentry, struct path *path)
 
 enum ovl_path_type ovl_path_real(struct dentry *dentry, struct path *path)
 {
-
 	enum ovl_path_type type = ovl_path_type(dentry);
 
 	if (type == OVL_PATH_LOWER)

1.6.0.2

Hi Miklos,

Will you apply my previous patch "ovl: Use macros to present ovl_xattr"? These two patches(previous one and this one) are based on branch *overlayfs-current* and conflict with each other.

I don't know which branch I should use, overlayfs-current or overlayfs-next. How could I join the development?

By the way, do you think we can make use of "kmem_cache" to manage "struct ovl_entry"? I think it will speedup operations like *ls*. I saw your work about multi-layer support. The size of "struct ovl_entry" will be variable. In addition, some free paths of ovl_entry rely on RCU. So maybe it's not easy.

Thanks~!

Hu

If it's a bugfix, then do it against -current. If it's a cleanup patch, like the one you did, do it against -next.

With the multi-layer support in overlayfs-next "struct ovl_entry" becomes variable sized, so we can't use kmem_cache for it.

Thanks,
Miklos