[mirror_ubuntu-zesty-kernel.git] / fs / nsfs.c

#include <linux/mount.h>
#include <linux/file.h>
#include <linux/fs.h>
#include <linux/proc_ns.h>
#include <linux/magic.h>
#include <linux/ktime.h>
#include <linux/seq_file.h>
#include <linux/user_namespace.h>
#include <linux/nsfs.h>

static struct vfsmount *nsfs_mnt;

static long ns_ioctl(struct file *filp, unsigned int ioctl,
			unsigned long arg);
static const struct file_operations ns_file_operations = {
	.llseek		= no_llseek,
	.unlocked_ioctl = ns_ioctl,
};

static char *ns_dname(struct dentry *dentry, char *buffer, int buflen)
{
	struct inode *inode = d_inode(dentry);
	const struct proc_ns_operations *ns_ops = dentry->d_fsdata;

	return dynamic_dname(dentry, buffer, buflen, "%s:[%lu]",
		ns_ops->name, inode->i_ino);
}

static void ns_prune_dentry(struct dentry *dentry)
{
	struct inode *inode = d_inode(dentry);
	if (inode) {
		struct ns_common *ns = inode->i_private;
		atomic_long_set(&ns->stashed, 0);
	}
}

const struct dentry_operations ns_dentry_operations =
{
	.d_prune	= ns_prune_dentry,
	.d_delete	= always_delete_dentry,
	.d_dname	= ns_dname,
};

static void nsfs_evict(struct inode *inode)
{
	struct ns_common *ns = inode->i_private;
	clear_inode(inode);
	ns->ops->put(ns);
}

static void *__ns_get_path(struct path *path, struct ns_common *ns)
{
	struct vfsmount *mnt = mntget(nsfs_mnt);
	struct qstr qname = { .name = "", };
	struct dentry *dentry;
	struct inode *inode;
	unsigned long d;

	rcu_read_lock();
	d = atomic_long_read(&ns->stashed);
	if (!d)
		goto slow;
	dentry = (struct dentry *)d;
	if (!lockref_get_not_dead(&dentry->d_lockref))
		goto slow;
	rcu_read_unlock();
	ns->ops->put(ns);
got_it:
	path->mnt = mnt;
	path->dentry = dentry;
	return NULL;
slow:
	rcu_read_unlock();
	inode = new_inode_pseudo(mnt->mnt_sb);
	if (!inode) {
		ns->ops->put(ns);
		mntput(mnt);
		return ERR_PTR(-ENOMEM);
	}
	inode->i_ino = ns->inum;
	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	inode->i_flags |= S_IMMUTABLE;
	inode->i_mode = S_IFREG | S_IRUGO;
	inode->i_fop = &ns_file_operations;
	inode->i_private = ns;

	dentry = d_alloc_pseudo(mnt->mnt_sb, &qname);
	if (!dentry) {
		iput(inode);
		mntput(mnt);
		return ERR_PTR(-ENOMEM);
	}
	d_instantiate(dentry, inode);
	dentry->d_fsdata = (void *)ns->ops;
	d = atomic_long_cmpxchg(&ns->stashed, 0, (unsigned long)dentry);
	if (d) {
		d_delete(dentry);	/* make sure ->d_prune() does nothing */
		dput(dentry);
		mntput(mnt);
		cpu_relax();
		return ERR_PTR(-EAGAIN);
	}
	goto got_it;
}

void *ns_get_path(struct path *path, struct task_struct *task,
			const struct proc_ns_operations *ns_ops)
{
	struct ns_common *ns;
	void *ret;

again:
	ns = ns_ops->get(task);
	if (!ns)
		return ERR_PTR(-ENOENT);

	ret = __ns_get_path(path, ns);
	if (IS_ERR(ret) && PTR_ERR(ret) == -EAGAIN)
		goto again;
	return ret;
}

static int open_related_ns(struct ns_common *ns,
		   struct ns_common *(*get_ns)(struct ns_common *ns))
{
	struct path path = {};
	struct file *f;
	void *err;
	int fd;

	fd = get_unused_fd_flags(O_CLOEXEC);
	if (fd < 0)
		return fd;

	while (1) {
		struct ns_common *relative;

		relative = get_ns(ns);
		if (IS_ERR(relative)) {
			put_unused_fd(fd);
			return PTR_ERR(relative);
		}

		err = __ns_get_path(&path, relative);
		if (IS_ERR(err) && PTR_ERR(err) == -EAGAIN)
			continue;
		break;
	}
	if (IS_ERR(err)) {
		put_unused_fd(fd);
		return PTR_ERR(err);
	}

	f = dentry_open(&path, O_RDONLY, current_cred());
	path_put(&path);
	if (IS_ERR(f)) {
		put_unused_fd(fd);
		fd = PTR_ERR(f);
	} else
		fd_install(fd, f);

	return fd;
}

static long ns_ioctl(struct file *filp, unsigned int ioctl,
			unsigned long arg)
{
	struct ns_common *ns = get_proc_ns(file_inode(filp));

	switch (ioctl) {
	case NS_GET_USERNS:
		return open_related_ns(ns, ns_get_owner);
	case NS_GET_PARENT:
		if (!ns->ops->get_parent)
			return -EINVAL;
		return open_related_ns(ns, ns->ops->get_parent);
	default:
		return -ENOTTY;
	}
}

int ns_get_name(char *buf, size_t size, struct task_struct *task,
			const struct proc_ns_operations *ns_ops)
{
	struct ns_common *ns;
	int res = -ENOENT;
	ns = ns_ops->get(task);
	if (ns) {
		res = snprintf(buf, size, "%s:[%u]", ns_ops->name, ns->inum);
		ns_ops->put(ns);
	}
	return res;
}

struct file *proc_ns_fget(int fd)
{
	struct file *file;

	file = fget(fd);
	if (!file)
		return ERR_PTR(-EBADF);

	if (file->f_op != &ns_file_operations)
		goto out_invalid;

	return file;

out_invalid:
	fput(file);
	return ERR_PTR(-EINVAL);
}

static int nsfs_show_path(struct seq_file *seq, struct dentry *dentry)
{
	struct inode *inode = d_inode(dentry);
	const struct proc_ns_operations *ns_ops = dentry->d_fsdata;

	seq_printf(seq, "%s:[%lu]", ns_ops->name, inode->i_ino);
	return 0;
}

static const struct super_operations nsfs_ops = {
	.statfs = simple_statfs,
	.evict_inode = nsfs_evict,
	.show_path = nsfs_show_path,
};
static struct dentry *nsfs_mount(struct file_system_type *fs_type,
			int flags, const char *dev_name, void *data)
{
	return mount_pseudo(fs_type, "nsfs:", &nsfs_ops,
			&ns_dentry_operations, NSFS_MAGIC);
}
static struct file_system_type nsfs = {
	.name = "nsfs",
	.mount = nsfs_mount,
	.kill_sb = kill_anon_super,
};

void __init nsfs_init(void)
{
	nsfs_mnt = kern_mount(&nsfs);
	if (IS_ERR(nsfs_mnt))
		panic("can't set nsfs up\n");
	nsfs_mnt->mnt_sb->s_flags &= ~MS_NOUSER;
}
Commit	Line	Data
e149ed2b AV	1	#include <linux/mount.h>
	2	#include <linux/file.h>
	3	#include <linux/fs.h>
	4	#include <linux/proc_ns.h>
	5	#include <linux/magic.h>
	6	#include <linux/ktime.h>
75509fd8	7	#include <linux/seq_file.h>
6786741d AV	8	#include <linux/user_namespace.h>
6786741d AV	9	#include <linux/nsfs.h>
e149ed2b AV	10
	11	static struct vfsmount *nsfs_mnt;
	12
6786741d AV	13	static long ns_ioctl(struct file *filp, unsigned int ioctl,
6786741d AV	14	unsigned long arg);
e149ed2b AV	15	static const struct file_operations ns_file_operations = {
e149ed2b AV	16	.llseek = no_llseek,
6786741d	17	.unlocked_ioctl = ns_ioctl,
e149ed2b AV	18	};
	19
	20	static char ns_dname(struct dentry dentry, char *buffer, int buflen)
	21	{
75c3cfa8	22	struct inode *inode = d_inode(dentry);
e149ed2b AV	23	const struct proc_ns_operations *ns_ops = dentry->d_fsdata;
	24
	25	return dynamic_dname(dentry, buffer, buflen, "%s:[%lu]",
	26	ns_ops->name, inode->i_ino);
	27	}
	28
	29	static void ns_prune_dentry(struct dentry *dentry)
	30	{
75c3cfa8	31	struct inode *inode = d_inode(dentry);
e149ed2b AV	32	if (inode) {
	33	struct ns_common *ns = inode->i_private;
	34	atomic_long_set(&ns->stashed, 0);
	35	}
	36	}
	37
	38	const struct dentry_operations ns_dentry_operations =
	39	{
	40	.d_prune = ns_prune_dentry,
	41	.d_delete = always_delete_dentry,
	42	.d_dname = ns_dname,
	43	};
	44
	45	static void nsfs_evict(struct inode *inode)
	46	{
	47	struct ns_common *ns = inode->i_private;
	48	clear_inode(inode);
	49	ns->ops->put(ns);
	50	}
	51
6786741d	52	static void __ns_get_path(struct path path, struct ns_common *ns)
e149ed2b AV	53	{
	54	struct vfsmount *mnt = mntget(nsfs_mnt);
	55	struct qstr qname = { .name = "", };
	56	struct dentry *dentry;
	57	struct inode *inode;
e149ed2b AV	58	unsigned long d;
e149ed2b AV	59
e149ed2b AV	60	rcu_read_lock();
	61	d = atomic_long_read(&ns->stashed);
	62	if (!d)
	63	goto slow;
	64	dentry = (struct dentry *)d;
	65	if (!lockref_get_not_dead(&dentry->d_lockref))
	66	goto slow;
	67	rcu_read_unlock();
6786741d	68	ns->ops->put(ns);
e149ed2b AV	69	got_it:
	70	path->mnt = mnt;
	71	path->dentry = dentry;
	72	return NULL;
	73	slow:
	74	rcu_read_unlock();
	75	inode = new_inode_pseudo(mnt->mnt_sb);
	76	if (!inode) {
6786741d	77	ns->ops->put(ns);
e149ed2b AV	78	mntput(mnt);
	79	return ERR_PTR(-ENOMEM);
	80	}
	81	inode->i_ino = ns->inum;
	82	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	83	inode->i_flags \|= S_IMMUTABLE;
	84	inode->i_mode = S_IFREG \| S_IRUGO;
	85	inode->i_fop = &ns_file_operations;
	86	inode->i_private = ns;
	87
	88	dentry = d_alloc_pseudo(mnt->mnt_sb, &qname);
	89	if (!dentry) {
	90	iput(inode);
	91	mntput(mnt);
	92	return ERR_PTR(-ENOMEM);
	93	}
	94	d_instantiate(dentry, inode);
6786741d	95	dentry->d_fsdata = (void *)ns->ops;
e149ed2b AV	96	d = atomic_long_cmpxchg(&ns->stashed, 0, (unsigned long)dentry);
	97	if (d) {
	98	d_delete(dentry); /* make sure ->d_prune() does nothing */
	99	dput(dentry);
6786741d	100	mntput(mnt);
e149ed2b	101	cpu_relax();
6786741d	102	return ERR_PTR(-EAGAIN);
e149ed2b AV	103	}
	104	goto got_it;
	105	}
	106
6786741d AV	107	void ns_get_path(struct path path, struct task_struct *task,
	108	const struct proc_ns_operations *ns_ops)
	109	{
	110	struct ns_common *ns;
	111	void *ret;
	112
	113	again:
	114	ns = ns_ops->get(task);
	115	if (!ns)
	116	return ERR_PTR(-ENOENT);
	117
	118	ret = __ns_get_path(path, ns);
	119	if (IS_ERR(ret) && PTR_ERR(ret) == -EAGAIN)
	120	goto again;
	121	return ret;
	122	}
	123
	124	static int open_related_ns(struct ns_common *ns,
	125	struct ns_common (get_ns)(struct ns_common *ns))
	126	{
	127	struct path path = {};
	128	struct file *f;
	129	void *err;
	130	int fd;
	131
	132	fd = get_unused_fd_flags(O_CLOEXEC);
	133	if (fd < 0)
	134	return fd;
	135
	136	while (1) {
	137	struct ns_common *relative;
	138
	139	relative = get_ns(ns);
	140	if (IS_ERR(relative)) {
	141	put_unused_fd(fd);
	142	return PTR_ERR(relative);
	143	}
	144
	145	err = __ns_get_path(&path, relative);
	146	if (IS_ERR(err) && PTR_ERR(err) == -EAGAIN)
	147	continue;
	148	break;
	149	}
	150	if (IS_ERR(err)) {
	151	put_unused_fd(fd);
	152	return PTR_ERR(err);
	153	}
	154
	155	f = dentry_open(&path, O_RDONLY, current_cred());
	156	path_put(&path);
	157	if (IS_ERR(f)) {
	158	put_unused_fd(fd);
	159	fd = PTR_ERR(f);
	160	} else
	161	fd_install(fd, f);
	162
	163	return fd;
	164	}
	165
	166	static long ns_ioctl(struct file *filp, unsigned int ioctl,
	167	unsigned long arg)
	168	{
	169	struct ns_common *ns = get_proc_ns(file_inode(filp));
	170
171	switch (ioctl) {
172	case NS_GET_USERNS:
173	return open_related_ns(ns, ns_get_owner);
a7306ed8 AV	174	case NS_GET_PARENT:
	175	if (!ns->ops->get_parent)
	176	return -EINVAL;
	177	return open_related_ns(ns, ns->ops->get_parent);
6786741d AV	178	default:
	179	return -ENOTTY;
	180	}
	181	}
	182
e149ed2b AV	183	int ns_get_name(char buf, size_t size, struct task_struct task,
	184	const struct proc_ns_operations *ns_ops)
	185	{
	186	struct ns_common *ns;
	187	int res = -ENOENT;
	188	ns = ns_ops->get(task);
	189	if (ns) {
	190	res = snprintf(buf, size, "%s:[%u]", ns_ops->name, ns->inum);
	191	ns_ops->put(ns);
	192	}
	193	return res;
	194	}
	195
	196	struct file *proc_ns_fget(int fd)
	197	{
	198	struct file *file;
	199
	200	file = fget(fd);
	201	if (!file)
	202	return ERR_PTR(-EBADF);
	203
	204	if (file->f_op != &ns_file_operations)
	205	goto out_invalid;
	206
	207	return file;
	208
	209	out_invalid:
	210	fput(file);
	211	return ERR_PTR(-EINVAL);
	212	}
	213
75509fd8 EB	214	static int nsfs_show_path(struct seq_file seq, struct dentry dentry)
	215	{
	216	struct inode *inode = d_inode(dentry);
	217	const struct proc_ns_operations *ns_ops = dentry->d_fsdata;
	218
6798a8ca JP	219	seq_printf(seq, "%s:[%lu]", ns_ops->name, inode->i_ino);
6798a8ca JP	220	return 0;
75509fd8 EB	221	}
75509fd8 EB	222
e149ed2b AV	223	static const struct super_operations nsfs_ops = {
	224	.statfs = simple_statfs,
	225	.evict_inode = nsfs_evict,
75509fd8	226	.show_path = nsfs_show_path,
e149ed2b AV	227	};
	228	static struct dentry nsfs_mount(struct file_system_type fs_type,
	229	int flags, const char dev_name, void data)
	230	{
	231	return mount_pseudo(fs_type, "nsfs:", &nsfs_ops,
	232	&ns_dentry_operations, NSFS_MAGIC);
	233	}
	234	static struct file_system_type nsfs = {
	235	.name = "nsfs",
	236	.mount = nsfs_mount,
	237	.kill_sb = kill_anon_super,
	238	};
	239
	240	void __init nsfs_init(void)
	241	{
	242	nsfs_mnt = kern_mount(&nsfs);
	243	if (IS_ERR(nsfs_mnt))
	244	panic("can't set nsfs up\n");
	245	nsfs_mnt->mnt_sb->s_flags &= ~MS_NOUSER;
	246	}