[PATCH] mnt_expire is protected by namespace_sem, no need for vfsmount_lock

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This commit is contained in:
Al Viro 2008-03-22 16:14:30 -04:00
parent c35038beca
commit 6758f953d0

View File

@ -262,10 +262,8 @@ static struct vfsmount *clone_mnt(struct vfsmount *old, struct dentry *root,
/* stick the duplicate mount on the same expiry list
* as the original if that was on one */
if (flag & CL_EXPIRE) {
spin_lock(&vfsmount_lock);
if (!list_empty(&old->mnt_expire))
list_add(&mnt->mnt_expire, &old->mnt_expire);
spin_unlock(&vfsmount_lock);
}
}
return mnt;
@ -1127,11 +1125,9 @@ static noinline int do_move_mount(struct nameidata *nd, char *old_name)
if (err)
goto out1;
spin_lock(&vfsmount_lock);
/* if the mount is moved, it should no longer be expire
* automatically */
list_del_init(&old_nd.path.mnt->mnt_expire);
spin_unlock(&vfsmount_lock);
out1:
mutex_unlock(&nd->path.dentry->d_inode->i_mutex);
out:
@ -1198,12 +1194,9 @@ int do_add_mount(struct vfsmount *newmnt, struct nameidata *nd,
if ((err = graft_tree(newmnt, nd)))
goto unlock;
if (fslist) {
/* add to the specified expiration list */
spin_lock(&vfsmount_lock);
if (fslist) /* add to the specified expiration list */
list_add_tail(&newmnt->mnt_expire, fslist);
spin_unlock(&vfsmount_lock);
}
up_write(&namespace_sem);
return 0;