From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 5 Feb 2010 07:21:06 +0000 (-0500)
Subject: get rid of useless vfsmount_lock use in put_mnt_ns()
X-Git-Tag: v2.6.34-rc1~205^2~14
X-Git-Url: http://bbs.cooldavid.org/git/?a=commitdiff_plain;h=d498b25a4f877be535246c4e5b6ee5890781a477;p=net-next-2.6.git

get rid of useless vfsmount_lock use in put_mnt_ns()

It hadn't been needed since we'd sanitized the logics in
mark_mounts_for_expiry() (which, in turn, used to be a
rudiment of bad old times when namespace_sem was per-ns).

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---

diff --git a/fs/namespace.c b/fs/namespace.c
index 970fe79d786..b0b15cc2117 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -2344,17 +2344,13 @@ void __init mnt_init(void)
 
 void put_mnt_ns(struct mnt_namespace *ns)
 {
-	struct vfsmount *root;
 	LIST_HEAD(umount_list);
 
-	if (!atomic_dec_and_lock(&ns->count, &vfsmount_lock))
+	if (!atomic_dec_and_test(&ns->count))
 		return;
-	root = ns->root;
-	ns->root = NULL;
-	spin_unlock(&vfsmount_lock);
 	down_write(&namespace_sem);
 	spin_lock(&vfsmount_lock);
-	umount_tree(root, 0, &umount_list);
+	umount_tree(ns->root, 0, &umount_list);
 	spin_unlock(&vfsmount_lock);
 	up_write(&namespace_sem);
 	release_mounts(&umount_list);