[linux-2.6-block.git] / fs / nsfs.c

// SPDX-License-Identifier: GPL-2.0
#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>
#include <linux/uaccess.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 = nsfs_mnt;
	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 = mntget(mnt);
	path->dentry = dentry;
	return NULL;
slow:
	rcu_read_unlock();
	inode = new_inode_pseudo(mnt->mnt_sb);
	if (!inode) {
		ns->ops->put(ns);
		return ERR_PTR(-ENOMEM);
	}
	inode->i_ino = ns->inum;
	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
	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, &empty_name);
	if (!dentry) {
		iput(inode);
		return ERR_PTR(-ENOMEM);
	}
	d_instantiate(dentry, inode);
	dentry->d_flags |= DCACHE_RCUACCESS;
	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);
		cpu_relax();
		return ERR_PTR(-EAGAIN);
	}
	goto got_it;
}

void *ns_get_path_cb(struct path *path, ns_get_path_helper_t *ns_get_cb,
		     void *private_data)
{
	struct ns_common *ns;
	void *ret;

again:
	ns = ns_get_cb(private_data);
	if (!ns)
		return ERR_PTR(-ENOENT);

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

struct ns_get_path_task_args {
	const struct proc_ns_operations *ns_ops;
	struct task_struct *task;
};

static struct ns_common *ns_get_path_task(void *private_data)
{
	struct ns_get_path_task_args *args = private_data;

	return args->ns_ops->get(args->task);
}

void *ns_get_path(struct path *path, struct task_struct *task,
		  const struct proc_ns_operations *ns_ops)
{
	struct ns_get_path_task_args args = {
		.ns_ops	= ns_ops,
		.task	= task,
	};

	return ns_get_path_cb(path, ns_get_path_task, &args);
}

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 user_namespace *user_ns;
	struct ns_common *ns = get_proc_ns(file_inode(filp));
	uid_t __user *argp;
	uid_t uid;

	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);
	case NS_GET_NSTYPE:
		return ns->ops->type;
	case NS_GET_OWNER_UID:
		if (ns->ops->type != CLONE_NEWUSER)
			return -EINVAL;
		user_ns = container_of(ns, struct user_namespace, ns);
		argp = (uid_t __user *) arg;
		uid = from_kuid_munged(current_user_ns(), user_ns->owner);
		return put_user(uid, argp);
	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;
	const char *name;
	ns = ns_ops->get(task);
	if (ns) {
		name = ns_ops->real_ns_name ? : ns_ops->name;
		res = snprintf(buf, size, "%s:[%u]", 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 &= ~SB_NOUSER;
}
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
e149ed2b AV	2	#include <linux/mount.h>
	3	#include <linux/file.h>
	4	#include <linux/fs.h>
	5	#include <linux/proc_ns.h>
	6	#include <linux/magic.h>
	7	#include <linux/ktime.h>
75509fd8	8	#include <linux/seq_file.h>
6786741d AV	9	#include <linux/user_namespace.h>
6786741d AV	10	#include <linux/nsfs.h>
d95fa3c7	11	#include <linux/uaccess.h>
e149ed2b AV	12
	13	static struct vfsmount *nsfs_mnt;
	14
6786741d AV	15	static long ns_ioctl(struct file *filp, unsigned int ioctl,
6786741d AV	16	unsigned long arg);
e149ed2b AV	17	static const struct file_operations ns_file_operations = {
e149ed2b AV	18	.llseek = no_llseek,
6786741d	19	.unlocked_ioctl = ns_ioctl,
e149ed2b AV	20	};
	21
	22	static char ns_dname(struct dentry dentry, char *buffer, int buflen)
	23	{
75c3cfa8	24	struct inode *inode = d_inode(dentry);
e149ed2b AV	25	const struct proc_ns_operations *ns_ops = dentry->d_fsdata;
	26
	27	return dynamic_dname(dentry, buffer, buflen, "%s:[%lu]",
	28	ns_ops->name, inode->i_ino);
	29	}
	30
	31	static void ns_prune_dentry(struct dentry *dentry)
	32	{
75c3cfa8	33	struct inode *inode = d_inode(dentry);
e149ed2b AV	34	if (inode) {
	35	struct ns_common *ns = inode->i_private;
	36	atomic_long_set(&ns->stashed, 0);
	37	}
	38	}
	39
	40	const struct dentry_operations ns_dentry_operations =
	41	{
	42	.d_prune = ns_prune_dentry,
	43	.d_delete = always_delete_dentry,
	44	.d_dname = ns_dname,
	45	};
	46
	47	static void nsfs_evict(struct inode *inode)
	48	{
	49	struct ns_common *ns = inode->i_private;
	50	clear_inode(inode);
	51	ns->ops->put(ns);
	52	}
	53
6786741d	54	static void __ns_get_path(struct path path, struct ns_common *ns)
e149ed2b	55	{
213b067c	56	struct vfsmount *mnt = nsfs_mnt;
e149ed2b AV	57	struct dentry *dentry;
e149ed2b AV	58	struct inode *inode;
e149ed2b AV	59	unsigned long d;
e149ed2b AV	60
e149ed2b AV	61	rcu_read_lock();
	62	d = atomic_long_read(&ns->stashed);
	63	if (!d)
	64	goto slow;
	65	dentry = (struct dentry *)d;
	66	if (!lockref_get_not_dead(&dentry->d_lockref))
	67	goto slow;
	68	rcu_read_unlock();
6786741d	69	ns->ops->put(ns);
e149ed2b	70	got_it:
213b067c	71	path->mnt = mntget(mnt);
e149ed2b AV	72	path->dentry = dentry;
	73	return NULL;
	74	slow:
	75	rcu_read_unlock();
	76	inode = new_inode_pseudo(mnt->mnt_sb);
	77	if (!inode) {
6786741d	78	ns->ops->put(ns);
e149ed2b AV	79	return ERR_PTR(-ENOMEM);
	80	}
	81	inode->i_ino = ns->inum;
078cd827	82	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
e149ed2b AV	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
cdf01226	88	dentry = d_alloc_pseudo(mnt->mnt_sb, &empty_name);
e149ed2b AV	89	if (!dentry) {
e149ed2b AV	90	iput(inode);
e149ed2b AV	91	return ERR_PTR(-ENOMEM);
	92	}
	93	d_instantiate(dentry, inode);
073c516f	94	dentry->d_flags \|= DCACHE_RCUACCESS;
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);
	100	cpu_relax();
6786741d	101	return ERR_PTR(-EAGAIN);
e149ed2b AV	102	}
	103	goto got_it;
	104	}
	105
cdab6ba8 JK	106	void ns_get_path_cb(struct path path, ns_get_path_helper_t *ns_get_cb,
cdab6ba8 JK	107	void *private_data)
6786741d AV	108	{
	109	struct ns_common *ns;
	110	void *ret;
	111
	112	again:
cdab6ba8	113	ns = ns_get_cb(private_data);
6786741d AV	114	if (!ns)
	115	return ERR_PTR(-ENOENT);
	116
	117	ret = __ns_get_path(path, ns);
	118	if (IS_ERR(ret) && PTR_ERR(ret) == -EAGAIN)
	119	goto again;
	120	return ret;
	121	}
	122
cdab6ba8 JK	123	struct ns_get_path_task_args {
	124	const struct proc_ns_operations *ns_ops;
	125	struct task_struct *task;
	126	};
	127
	128	static struct ns_common ns_get_path_task(void private_data)
	129	{
	130	struct ns_get_path_task_args *args = private_data;
	131
	132	return args->ns_ops->get(args->task);
	133	}
	134
	135	void ns_get_path(struct path path, struct task_struct *task,
	136	const struct proc_ns_operations *ns_ops)
	137	{
	138	struct ns_get_path_task_args args = {
	139	.ns_ops = ns_ops,
	140	.task = task,
	141	};
	142
	143	return ns_get_path_cb(path, ns_get_path_task, &args);
	144	}
	145
c62cce2c	146	int open_related_ns(struct ns_common *ns,
6786741d AV	147	struct ns_common (get_ns)(struct ns_common *ns))
	148	{
	149	struct path path = {};
	150	struct file *f;
	151	void *err;
	152	int fd;
	153
	154	fd = get_unused_fd_flags(O_CLOEXEC);
	155	if (fd < 0)
	156	return fd;
	157
	158	while (1) {
	159	struct ns_common *relative;
	160
	161	relative = get_ns(ns);
	162	if (IS_ERR(relative)) {
	163	put_unused_fd(fd);
	164	return PTR_ERR(relative);
	165	}
	166
	167	err = __ns_get_path(&path, relative);
	168	if (IS_ERR(err) && PTR_ERR(err) == -EAGAIN)
	169	continue;
	170	break;
	171	}
	172	if (IS_ERR(err)) {
	173	put_unused_fd(fd);
	174	return PTR_ERR(err);
	175	}
	176
	177	f = dentry_open(&path, O_RDONLY, current_cred());
	178	path_put(&path);
	179	if (IS_ERR(f)) {
	180	put_unused_fd(fd);
	181	fd = PTR_ERR(f);
	182	} else
	183	fd_install(fd, f);
	184
	185	return fd;
	186	}
	187
	188	static long ns_ioctl(struct file *filp, unsigned int ioctl,
	189	unsigned long arg)
	190	{
d95fa3c7	191	struct user_namespace *user_ns;
6786741d	192	struct ns_common *ns = get_proc_ns(file_inode(filp));
d95fa3c7 MK	193	uid_t __user *argp;
d95fa3c7 MK	194	uid_t uid;
6786741d AV	195
	196	switch (ioctl) {
	197	case NS_GET_USERNS:
	198	return open_related_ns(ns, ns_get_owner);
a7306ed8 AV	199	case NS_GET_PARENT:
	200	if (!ns->ops->get_parent)
	201	return -EINVAL;
	202	return open_related_ns(ns, ns->ops->get_parent);
e5ff5ce6 MK	203	case NS_GET_NSTYPE:
e5ff5ce6 MK	204	return ns->ops->type;
d95fa3c7 MK	205	case NS_GET_OWNER_UID:
	206	if (ns->ops->type != CLONE_NEWUSER)
	207	return -EINVAL;
	208	user_ns = container_of(ns, struct user_namespace, ns);
	209	argp = (uid_t __user *) arg;
	210	uid = from_kuid_munged(current_user_ns(), user_ns->owner);
	211	return put_user(uid, argp);
6786741d AV	212	default:
	213	return -ENOTTY;
	214	}
	215	}
	216
e149ed2b AV	217	int ns_get_name(char buf, size_t size, struct task_struct task,
	218	const struct proc_ns_operations *ns_ops)
	219	{
	220	struct ns_common *ns;
	221	int res = -ENOENT;
25b14e92	222	const char *name;
e149ed2b AV	223	ns = ns_ops->get(task);
e149ed2b AV	224	if (ns) {
25b14e92 KT	225	name = ns_ops->real_ns_name ? : ns_ops->name;
25b14e92 KT	226	res = snprintf(buf, size, "%s:[%u]", name, ns->inum);
e149ed2b AV	227	ns_ops->put(ns);
	228	}
	229	return res;
	230	}
	231
	232	struct file *proc_ns_fget(int fd)
	233	{
	234	struct file *file;
	235
	236	file = fget(fd);
	237	if (!file)
	238	return ERR_PTR(-EBADF);
	239
	240	if (file->f_op != &ns_file_operations)
	241	goto out_invalid;
	242
	243	return file;
	244
	245	out_invalid:
	246	fput(file);
	247	return ERR_PTR(-EINVAL);
	248	}
	249
75509fd8 EB	250	static int nsfs_show_path(struct seq_file seq, struct dentry dentry)
	251	{
	252	struct inode *inode = d_inode(dentry);
	253	const struct proc_ns_operations *ns_ops = dentry->d_fsdata;
	254
6798a8ca JP	255	seq_printf(seq, "%s:[%lu]", ns_ops->name, inode->i_ino);
6798a8ca JP	256	return 0;
75509fd8 EB	257	}
75509fd8 EB	258
e149ed2b AV	259	static const struct super_operations nsfs_ops = {
	260	.statfs = simple_statfs,
	261	.evict_inode = nsfs_evict,
75509fd8	262	.show_path = nsfs_show_path,
e149ed2b AV	263	};
	264	static struct dentry nsfs_mount(struct file_system_type fs_type,
	265	int flags, const char dev_name, void data)
	266	{
	267	return mount_pseudo(fs_type, "nsfs:", &nsfs_ops,
	268	&ns_dentry_operations, NSFS_MAGIC);
	269	}
	270	static struct file_system_type nsfs = {
	271	.name = "nsfs",
	272	.mount = nsfs_mount,
	273	.kill_sb = kill_anon_super,
	274	};
	275
	276	void __init nsfs_init(void)
	277	{
	278	nsfs_mnt = kern_mount(&nsfs);
	279	if (IS_ERR(nsfs_mnt))
	280	panic("can't set nsfs up\n");
1751e8a6	281	nsfs_mnt->mnt_sb->s_flags &= ~SB_NOUSER;
e149ed2b	282	}