aboutsummaryrefslogtreecommitdiff
path: root/fs/namespace.c
diff options
context:
space:
mode:
authorAl Viro <viro@zeniv.linux.org.uk>2011-11-25 00:46:35 -0500
committerAl Viro <viro@zeniv.linux.org.uk>2012-01-03 22:57:09 -0500
commit143c8c91cee7efdd732ec5f61b3471fc46192f20 (patch)
treedd3e9a9b8b88f1876a258fe1bf274532aeb72a62 /fs/namespace.c
parent900148dcac6bc93ca688d64a7f9a9f8d706e0d1c (diff)
vfs: mnt_ns moved to struct mount
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Diffstat (limited to 'fs/namespace.c')
-rw-r--r--fs/namespace.c45
1 files changed, 23 insertions, 22 deletions
diff --git a/fs/namespace.c b/fs/namespace.c
index 3e95cc26dda..4cdb7f69861 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -505,7 +505,7 @@ struct vfsmount *lookup_mnt(struct path *path)
}
}
-static inline int check_mnt(struct vfsmount *mnt)
+static inline int check_mnt(struct mount *mnt)
{
return mnt->mnt_ns == current->nsproxy->mnt_ns;
}
@@ -614,13 +614,13 @@ static void commit_tree(struct mount *mnt)
struct mount *parent = mnt->mnt_parent;
struct mount *m;
LIST_HEAD(head);
- struct mnt_namespace *n = parent->mnt.mnt_ns;
+ struct mnt_namespace *n = parent->mnt_ns;
BUG_ON(parent == mnt);
list_add_tail(&head, &mnt->mnt.mnt_list);
list_for_each_entry(m, &head, mnt.mnt_list) {
- m->mnt.mnt_ns = n;
+ m->mnt_ns = n;
__mnt_make_longterm(m);
}
@@ -1234,8 +1234,8 @@ void umount_tree(struct mount *mnt, int propagate, struct list_head *kill)
list_for_each_entry(p, &tmp_list, mnt_hash) {
list_del_init(&p->mnt_expire);
list_del_init(&p->mnt.mnt_list);
- __touch_mnt_namespace(p->mnt.mnt_ns);
- p->mnt.mnt_ns = NULL;
+ __touch_mnt_namespace(p->mnt_ns);
+ p->mnt_ns = NULL;
__mnt_make_shortterm(p);
list_del_init(&p->mnt_child);
if (mnt_has_parent(p)) {
@@ -1367,7 +1367,7 @@ SYSCALL_DEFINE2(umount, char __user *, name, int, flags)
retval = -EINVAL;
if (path.dentry != path.mnt->mnt_root)
goto dput_and_out;
- if (!check_mnt(path.mnt))
+ if (!check_mnt(mnt))
goto dput_and_out;
retval = -EPERM;
@@ -1619,7 +1619,7 @@ static int attach_recursive_mnt(struct mount *source_mnt,
if (parent_path) {
detach_mnt(source_mnt, parent_path);
attach_mnt(source_mnt, path);
- touch_mnt_namespace(parent_path->mnt->mnt_ns);
+ touch_mnt_namespace(source_mnt->mnt_ns);
} else {
mnt_set_mountpoint(dest_mnt, dest_dentry, source_mnt);
commit_tree(source_mnt);
@@ -1765,7 +1765,7 @@ static int do_loopback(struct path *path, char *old_name,
if (IS_MNT_UNBINDABLE(old_path.mnt))
goto out2;
- if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt))
+ if (!check_mnt(real_mount(path->mnt)) || !check_mnt(old))
goto out2;
err = -ENOMEM;
@@ -1818,11 +1818,12 @@ static int do_remount(struct path *path, int flags, int mnt_flags,
{
int err;
struct super_block *sb = path->mnt->mnt_sb;
+ struct mount *mnt = real_mount(path->mnt);
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
- if (!check_mnt(path->mnt))
+ if (!check_mnt(mnt))
return -EINVAL;
if (path->dentry != path->mnt->mnt_root)
@@ -1839,14 +1840,14 @@ static int do_remount(struct path *path, int flags, int mnt_flags,
err = do_remount_sb(sb, flags, data, 0);
if (!err) {
br_write_lock(vfsmount_lock);
- mnt_flags |= path->mnt->mnt_flags & MNT_PROPAGATION_MASK;
- path->mnt->mnt_flags = mnt_flags;
+ mnt_flags |= mnt->mnt.mnt_flags & MNT_PROPAGATION_MASK;
+ mnt->mnt.mnt_flags = mnt_flags;
br_write_unlock(vfsmount_lock);
}
up_write(&sb->s_umount);
if (!err) {
br_write_lock(vfsmount_lock);
- touch_mnt_namespace(path->mnt->mnt_ns);
+ touch_mnt_namespace(mnt->mnt_ns);
br_write_unlock(vfsmount_lock);
}
return err;
@@ -1880,8 +1881,10 @@ static int do_move_mount(struct path *path, char *old_name)
if (err < 0)
goto out;
+ old = real_mount(old_path.mnt);
+
err = -EINVAL;
- if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt))
+ if (!check_mnt(real_mount(path->mnt)) || !check_mnt(old))
goto out1;
if (d_unlinked(path->dentry))
@@ -1891,8 +1894,6 @@ static int do_move_mount(struct path *path, char *old_name)
if (old_path.dentry != old_path.mnt->mnt_root)
goto out1;
- old = real_mount(old_path.mnt);
-
if (!mnt_has_parent(old))
goto out1;
@@ -1984,7 +1985,7 @@ static int do_add_mount(struct mount *newmnt, struct path *path, int mnt_flags)
return err;
err = -EINVAL;
- if (!(mnt_flags & MNT_SHRINKABLE) && !check_mnt(path->mnt))
+ if (!(mnt_flags & MNT_SHRINKABLE) && !check_mnt(real_mount(path->mnt)))
goto unlock;
/* Refuse the same filesystem on the same mount point */
@@ -2112,7 +2113,7 @@ void mark_mounts_for_expiry(struct list_head *mounts)
}
while (!list_empty(&graveyard)) {
mnt = list_first_entry(&graveyard, struct mount, mnt_expire);
- touch_mnt_namespace(mnt->mnt.mnt_ns);
+ touch_mnt_namespace(mnt->mnt_ns);
umount_tree(mnt, 1, &umounts);
}
br_write_unlock(vfsmount_lock);
@@ -2185,7 +2186,7 @@ static void shrink_submounts(struct mount *mnt, struct list_head *umounts)
while (!list_empty(&graveyard)) {
m = list_first_entry(&graveyard, struct mount,
mnt_expire);
- touch_mnt_namespace(m->mnt.mnt_ns);
+ touch_mnt_namespace(m->mnt_ns);
umount_tree(m, 1, umounts);
}
}
@@ -2423,7 +2424,7 @@ static struct mnt_namespace *dup_mnt_ns(struct mnt_namespace *mnt_ns,
p = real_mount(mnt_ns->root);
q = new;
while (p) {
- q->mnt.mnt_ns = new_ns;
+ q->mnt_ns = new_ns;
__mnt_make_longterm(q);
if (fs) {
if (&p->mnt == fs->root.mnt) {
@@ -2479,7 +2480,7 @@ static struct mnt_namespace *create_mnt_ns(struct vfsmount *mnt)
new_ns = alloc_mnt_ns();
if (!IS_ERR(new_ns)) {
- mnt->mnt_ns = new_ns;
+ real_mount(mnt)->mnt_ns = new_ns;
__mnt_make_longterm(real_mount(mnt));
new_ns->root = mnt;
list_add(&new_ns->list, &new_ns->root->mnt_list);
@@ -2644,7 +2645,7 @@ SYSCALL_DEFINE2(pivot_root, const char __user *, new_root,
IS_MNT_SHARED(&new_mnt->mnt_parent->mnt) ||
IS_MNT_SHARED(&root_mnt->mnt_parent->mnt))
goto out4;
- if (!check_mnt(root.mnt) || !check_mnt(new.mnt))
+ if (!check_mnt(root_mnt) || !check_mnt(new_mnt))
goto out4;
error = -ENOENT;
if (d_unlinked(new.dentry))
@@ -2793,5 +2794,5 @@ EXPORT_SYMBOL(kern_unmount);
bool our_mnt(struct vfsmount *mnt)
{
- return check_mnt(mnt);
+ return check_mnt(real_mount(mnt));
}