[linux-2.6-block.git] / kernel / nsproxy.c

// SPDX-License-Identifier: GPL-2.0-only
/*
 *  Copyright (C) 2006 IBM Corporation
 *
 *  Author: Serge Hallyn <serue@us.ibm.com>
 *
 *  Jun 2006 - namespaces support
 *             OpenVZ, SWsoft Inc.
 *             Pavel Emelianov <xemul@openvz.org>
 */

#include <linux/slab.h>
#include <linux/export.h>
#include <linux/nsproxy.h>
#include <linux/init_task.h>
#include <linux/mnt_namespace.h>
#include <linux/utsname.h>
#include <linux/pid_namespace.h>
#include <net/net_namespace.h>
#include <linux/ipc_namespace.h>
#include <linux/time_namespace.h>
#include <linux/fs_struct.h>
#include <linux/proc_fs.h>
#include <linux/proc_ns.h>
#include <linux/file.h>
#include <linux/syscalls.h>
#include <linux/cgroup.h>
#include <linux/perf_event.h>

static struct kmem_cache *nsproxy_cachep;

struct nsproxy init_nsproxy = {
	.count			= REFCOUNT_INIT(1),
	.uts_ns			= &init_uts_ns,
#if defined(CONFIG_POSIX_MQUEUE) || defined(CONFIG_SYSVIPC)
	.ipc_ns			= &init_ipc_ns,
#endif
	.mnt_ns			= NULL,
	.pid_ns_for_children	= &init_pid_ns,
#ifdef CONFIG_NET
	.net_ns			= &init_net,
#endif
#ifdef CONFIG_CGROUPS
	.cgroup_ns		= &init_cgroup_ns,
#endif
#ifdef CONFIG_TIME_NS
	.time_ns		= &init_time_ns,
	.time_ns_for_children	= &init_time_ns,
#endif
};

static inline struct nsproxy *create_nsproxy(void)
{
	struct nsproxy *nsproxy;

	nsproxy = kmem_cache_alloc(nsproxy_cachep, GFP_KERNEL);
	if (nsproxy)
		refcount_set(&nsproxy->count, 1);
	return nsproxy;
}

/*
 * Create new nsproxy and all of its the associated namespaces.
 * Return the newly created nsproxy.  Do not attach this to the task,
 * leave it to the caller to do proper locking and attach it to task.
 */
static struct nsproxy *create_new_namespaces(unsigned long flags,
	struct task_struct *tsk, struct user_namespace *user_ns,
	struct fs_struct *new_fs)
{
	struct nsproxy *new_nsp;
	int err;

	new_nsp = create_nsproxy();
	if (!new_nsp)
		return ERR_PTR(-ENOMEM);

	new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, user_ns, new_fs);
	if (IS_ERR(new_nsp->mnt_ns)) {
		err = PTR_ERR(new_nsp->mnt_ns);
		goto out_ns;
	}

	new_nsp->uts_ns = copy_utsname(flags, user_ns, tsk->nsproxy->uts_ns);
	if (IS_ERR(new_nsp->uts_ns)) {
		err = PTR_ERR(new_nsp->uts_ns);
		goto out_uts;
	}

	new_nsp->ipc_ns = copy_ipcs(flags, user_ns, tsk->nsproxy->ipc_ns);
	if (IS_ERR(new_nsp->ipc_ns)) {
		err = PTR_ERR(new_nsp->ipc_ns);
		goto out_ipc;
	}

	new_nsp->pid_ns_for_children =
		copy_pid_ns(flags, user_ns, tsk->nsproxy->pid_ns_for_children);
	if (IS_ERR(new_nsp->pid_ns_for_children)) {
		err = PTR_ERR(new_nsp->pid_ns_for_children);
		goto out_pid;
	}

	new_nsp->cgroup_ns = copy_cgroup_ns(flags, user_ns,
					    tsk->nsproxy->cgroup_ns);
	if (IS_ERR(new_nsp->cgroup_ns)) {
		err = PTR_ERR(new_nsp->cgroup_ns);
		goto out_cgroup;
	}

	new_nsp->net_ns = copy_net_ns(flags, user_ns, tsk->nsproxy->net_ns);
	if (IS_ERR(new_nsp->net_ns)) {
		err = PTR_ERR(new_nsp->net_ns);
		goto out_net;
	}

	new_nsp->time_ns_for_children = copy_time_ns(flags, user_ns,
					tsk->nsproxy->time_ns_for_children);
	if (IS_ERR(new_nsp->time_ns_for_children)) {
		err = PTR_ERR(new_nsp->time_ns_for_children);
		goto out_time;
	}
	new_nsp->time_ns = get_time_ns(tsk->nsproxy->time_ns);

	return new_nsp;

out_time:
	put_net(new_nsp->net_ns);
out_net:
	put_cgroup_ns(new_nsp->cgroup_ns);
out_cgroup:
	if (new_nsp->pid_ns_for_children)
		put_pid_ns(new_nsp->pid_ns_for_children);
out_pid:
	if (new_nsp->ipc_ns)
		put_ipc_ns(new_nsp->ipc_ns);
out_ipc:
	if (new_nsp->uts_ns)
		put_uts_ns(new_nsp->uts_ns);
out_uts:
	if (new_nsp->mnt_ns)
		put_mnt_ns(new_nsp->mnt_ns);
out_ns:
	kmem_cache_free(nsproxy_cachep, new_nsp);
	return ERR_PTR(err);
}

/*
 * called from clone.  This now handles copy for nsproxy and all
 * namespaces therein.
 */
int copy_namespaces(unsigned long flags, struct task_struct *tsk)
{
	struct nsproxy *old_ns = tsk->nsproxy;
	struct user_namespace *user_ns = task_cred_xxx(tsk, user_ns);
	struct nsproxy *new_ns;

	if (likely(!(flags & (CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC |
			      CLONE_NEWPID | CLONE_NEWNET |
			      CLONE_NEWCGROUP | CLONE_NEWTIME)))) {
		if ((flags & CLONE_VM) ||
		    likely(old_ns->time_ns_for_children == old_ns->time_ns)) {
			get_nsproxy(old_ns);
			return 0;
		}
	} else if (!ns_capable(user_ns, CAP_SYS_ADMIN))
		return -EPERM;

	/*
	 * CLONE_NEWIPC must detach from the undolist: after switching
	 * to a new ipc namespace, the semaphore arrays from the old
	 * namespace are unreachable.  In clone parlance, CLONE_SYSVSEM
	 * means share undolist with parent, so we must forbid using
	 * it along with CLONE_NEWIPC.
	 */
	if ((flags & (CLONE_NEWIPC | CLONE_SYSVSEM)) ==
		(CLONE_NEWIPC | CLONE_SYSVSEM))
		return -EINVAL;

	new_ns = create_new_namespaces(flags, tsk, user_ns, tsk->fs);
	if (IS_ERR(new_ns))
		return  PTR_ERR(new_ns);

	if ((flags & CLONE_VM) == 0)
		timens_on_fork(new_ns, tsk);

	tsk->nsproxy = new_ns;
	return 0;
}

void free_nsproxy(struct nsproxy *ns)
{
	if (ns->mnt_ns)
		put_mnt_ns(ns->mnt_ns);
	if (ns->uts_ns)
		put_uts_ns(ns->uts_ns);
	if (ns->ipc_ns)
		put_ipc_ns(ns->ipc_ns);
	if (ns->pid_ns_for_children)
		put_pid_ns(ns->pid_ns_for_children);
	if (ns->time_ns)
		put_time_ns(ns->time_ns);
	if (ns->time_ns_for_children)
		put_time_ns(ns->time_ns_for_children);
	put_cgroup_ns(ns->cgroup_ns);
	put_net(ns->net_ns);
	kmem_cache_free(nsproxy_cachep, ns);
}

/*
 * Called from unshare. Unshare all the namespaces part of nsproxy.
 * On success, returns the new nsproxy.
 */
int unshare_nsproxy_namespaces(unsigned long unshare_flags,
	struct nsproxy **new_nsp, struct cred *new_cred, struct fs_struct *new_fs)
{
	struct user_namespace *user_ns;
	int err = 0;

	if (!(unshare_flags & (CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC |
			       CLONE_NEWNET | CLONE_NEWPID | CLONE_NEWCGROUP |
			       CLONE_NEWTIME)))
		return 0;

	user_ns = new_cred ? new_cred->user_ns : current_user_ns();
	if (!ns_capable(user_ns, CAP_SYS_ADMIN))
		return -EPERM;

	*new_nsp = create_new_namespaces(unshare_flags, current, user_ns,
					 new_fs ? new_fs : current->fs);
	if (IS_ERR(*new_nsp)) {
		err = PTR_ERR(*new_nsp);
		goto out;
	}

out:
	return err;
}

void switch_task_namespaces(struct task_struct *p, struct nsproxy *new)
{
	struct nsproxy *ns;

	might_sleep();

	task_lock(p);
	ns = p->nsproxy;
	p->nsproxy = new;
	task_unlock(p);

	if (ns)
		put_nsproxy(ns);
}

void exit_task_namespaces(struct task_struct *p)
{
	switch_task_namespaces(p, NULL);
}

int exec_task_namespaces(void)
{
	struct task_struct *tsk = current;
	struct nsproxy *new;

	if (tsk->nsproxy->time_ns_for_children == tsk->nsproxy->time_ns)
		return 0;

	new = create_new_namespaces(0, tsk, current_user_ns(), tsk->fs);
	if (IS_ERR(new))
		return PTR_ERR(new);

	timens_on_fork(new, tsk);
	switch_task_namespaces(tsk, new);
	return 0;
}

static int check_setns_flags(unsigned long flags)
{
	if (!flags || (flags & ~(CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC |
				 CLONE_NEWNET | CLONE_NEWTIME | CLONE_NEWUSER |
				 CLONE_NEWPID | CLONE_NEWCGROUP)))
		return -EINVAL;

#ifndef CONFIG_USER_NS
	if (flags & CLONE_NEWUSER)
		return -EINVAL;
#endif
#ifndef CONFIG_PID_NS
	if (flags & CLONE_NEWPID)
		return -EINVAL;
#endif
#ifndef CONFIG_UTS_NS
	if (flags & CLONE_NEWUTS)
		return -EINVAL;
#endif
#ifndef CONFIG_IPC_NS
	if (flags & CLONE_NEWIPC)
		return -EINVAL;
#endif
#ifndef CONFIG_CGROUPS
	if (flags & CLONE_NEWCGROUP)
		return -EINVAL;
#endif
#ifndef CONFIG_NET_NS
	if (flags & CLONE_NEWNET)
		return -EINVAL;
#endif
#ifndef CONFIG_TIME_NS
	if (flags & CLONE_NEWTIME)
		return -EINVAL;
#endif

	return 0;
}

static void put_nsset(struct nsset *nsset)
{
	unsigned flags = nsset->flags;

	if (flags & CLONE_NEWUSER)
		put_cred(nsset_cred(nsset));
	/*
	 * We only created a temporary copy if we attached to more than just
	 * the mount namespace.
	 */
	if (nsset->fs && (flags & CLONE_NEWNS) && (flags & ~CLONE_NEWNS))
		free_fs_struct(nsset->fs);
	if (nsset->nsproxy)
		free_nsproxy(nsset->nsproxy);
}

static int prepare_nsset(unsigned flags, struct nsset *nsset)
{
	struct task_struct *me = current;

	nsset->nsproxy = create_new_namespaces(0, me, current_user_ns(), me->fs);
	if (IS_ERR(nsset->nsproxy))
		return PTR_ERR(nsset->nsproxy);

	if (flags & CLONE_NEWUSER)
		nsset->cred = prepare_creds();
	else
		nsset->cred = current_cred();
	if (!nsset->cred)
		goto out;

	/* Only create a temporary copy of fs_struct if we really need to. */
	if (flags == CLONE_NEWNS) {
		nsset->fs = me->fs;
	} else if (flags & CLONE_NEWNS) {
		nsset->fs = copy_fs_struct(me->fs);
		if (!nsset->fs)
			goto out;
	}

	nsset->flags = flags;
	return 0;

out:
	put_nsset(nsset);
	return -ENOMEM;
}

static inline int validate_ns(struct nsset *nsset, struct ns_common *ns)
{
	return ns->ops->install(nsset, ns);
}

/*
 * This is the inverse operation to unshare().
 * Ordering is equivalent to the standard ordering used everywhere else
 * during unshare and process creation. The switch to the new set of
 * namespaces occurs at the point of no return after installation of
 * all requested namespaces was successful in commit_nsset().
 */
static int validate_nsset(struct nsset *nsset, struct pid *pid)
{
	int ret = 0;
	unsigned flags = nsset->flags;
	struct user_namespace *user_ns = NULL;
	struct pid_namespace *pid_ns = NULL;
	struct nsproxy *nsp;
	struct task_struct *tsk;

	/* Take a "snapshot" of the target task's namespaces. */
	rcu_read_lock();
	tsk = pid_task(pid, PIDTYPE_PID);
	if (!tsk) {
		rcu_read_unlock();
		return -ESRCH;
	}

	if (!ptrace_may_access(tsk, PTRACE_MODE_READ_REALCREDS)) {
		rcu_read_unlock();
		return -EPERM;
	}

	task_lock(tsk);
	nsp = tsk->nsproxy;
	if (nsp)
		get_nsproxy(nsp);
	task_unlock(tsk);
	if (!nsp) {
		rcu_read_unlock();
		return -ESRCH;
	}

#ifdef CONFIG_PID_NS
	if (flags & CLONE_NEWPID) {
		pid_ns = task_active_pid_ns(tsk);
		if (unlikely(!pid_ns)) {
			rcu_read_unlock();
			ret = -ESRCH;
			goto out;
		}
		get_pid_ns(pid_ns);
	}
#endif

#ifdef CONFIG_USER_NS
	if (flags & CLONE_NEWUSER)
		user_ns = get_user_ns(__task_cred(tsk)->user_ns);
#endif
	rcu_read_unlock();

	/*
	 * Install requested namespaces. The caller will have
	 * verified earlier that the requested namespaces are
	 * supported on this kernel. We don't report errors here
	 * if a namespace is requested that isn't supported.
	 */
#ifdef CONFIG_USER_NS
	if (flags & CLONE_NEWUSER) {
		ret = validate_ns(nsset, &user_ns->ns);
		if (ret)
			goto out;
	}
#endif

	if (flags & CLONE_NEWNS) {
		ret = validate_ns(nsset, from_mnt_ns(nsp->mnt_ns));
		if (ret)
			goto out;
	}

#ifdef CONFIG_UTS_NS
	if (flags & CLONE_NEWUTS) {
		ret = validate_ns(nsset, &nsp->uts_ns->ns);
		if (ret)
			goto out;
	}
#endif

#ifdef CONFIG_IPC_NS
	if (flags & CLONE_NEWIPC) {
		ret = validate_ns(nsset, &nsp->ipc_ns->ns);
		if (ret)
			goto out;
	}
#endif

#ifdef CONFIG_PID_NS
	if (flags & CLONE_NEWPID) {
		ret = validate_ns(nsset, &pid_ns->ns);
		if (ret)
			goto out;
	}
#endif

#ifdef CONFIG_CGROUPS
	if (flags & CLONE_NEWCGROUP) {
		ret = validate_ns(nsset, &nsp->cgroup_ns->ns);
		if (ret)
			goto out;
	}
#endif

#ifdef CONFIG_NET_NS
	if (flags & CLONE_NEWNET) {
		ret = validate_ns(nsset, &nsp->net_ns->ns);
		if (ret)
			goto out;
	}
#endif

#ifdef CONFIG_TIME_NS
	if (flags & CLONE_NEWTIME) {
		ret = validate_ns(nsset, &nsp->time_ns->ns);
		if (ret)
			goto out;
	}
#endif

out:
	if (pid_ns)
		put_pid_ns(pid_ns);
	if (nsp)
		put_nsproxy(nsp);
	put_user_ns(user_ns);

	return ret;
}

/*
 * This is the point of no return. There are just a few namespaces
 * that do some actual work here and it's sufficiently minimal that
 * a separate ns_common operation seems unnecessary for now.
 * Unshare is doing the same thing. If we'll end up needing to do
 * more in a given namespace or a helper here is ultimately not
 * exported anymore a simple commit handler for each namespace
 * should be added to ns_common.
 */
static void commit_nsset(struct nsset *nsset)
{
	unsigned flags = nsset->flags;
	struct task_struct *me = current;

#ifdef CONFIG_USER_NS
	if (flags & CLONE_NEWUSER) {
		/* transfer ownership */
		commit_creds(nsset_cred(nsset));
		nsset->cred = NULL;
	}
#endif

	/* We only need to commit if we have used a temporary fs_struct. */
	if ((flags & CLONE_NEWNS) && (flags & ~CLONE_NEWNS)) {
		set_fs_root(me->fs, &nsset->fs->root);
		set_fs_pwd(me->fs, &nsset->fs->pwd);
	}

#ifdef CONFIG_IPC_NS
	if (flags & CLONE_NEWIPC)
		exit_sem(me);
#endif

#ifdef CONFIG_TIME_NS
	if (flags & CLONE_NEWTIME)
		timens_commit(me, nsset->nsproxy->time_ns);
#endif

	/* transfer ownership */
	switch_task_namespaces(me, nsset->nsproxy);
	nsset->nsproxy = NULL;
}

SYSCALL_DEFINE2(setns, int, fd, int, flags)
{
	struct fd f = fdget(fd);
	struct ns_common *ns = NULL;
	struct nsset nsset = {};
	int err = 0;

	if (!f.file)
		return -EBADF;

	if (proc_ns_file(f.file)) {
		ns = get_proc_ns(file_inode(f.file));
		if (flags && (ns->ops->type != flags))
			err = -EINVAL;
		flags = ns->ops->type;
	} else if (!IS_ERR(pidfd_pid(f.file))) {
		err = check_setns_flags(flags);
	} else {
		err = -EINVAL;
	}
	if (err)
		goto out;

	err = prepare_nsset(flags, &nsset);
	if (err)
		goto out;

	if (proc_ns_file(f.file))
		err = validate_ns(&nsset, ns);
	else
		err = validate_nsset(&nsset, pidfd_pid(f.file));
	if (!err) {
		commit_nsset(&nsset);
		perf_event_namespaces(current);
	}
	put_nsset(&nsset);
out:
	fdput(f);
	return err;
}

int __init nsproxy_cache_init(void)
{
	nsproxy_cachep = KMEM_CACHE(nsproxy, SLAB_PANIC|SLAB_ACCOUNT);
	return 0;
}
Commit	Line	Data
b886d83c	1	// SPDX-License-Identifier: GPL-2.0-only
ab516013 SH	2	/*
	3	* Copyright (C) 2006 IBM Corporation
	4	*
	5	* Author: Serge Hallyn <serue@us.ibm.com>
	6	*
25b21cb2 KK	7	* Jun 2006 - namespaces support
	8	* OpenVZ, SWsoft Inc.
	9	* Pavel Emelianov <xemul@openvz.org>
ab516013 SH	10	*/
ab516013 SH	11
5a0e3ad6	12	#include <linux/slab.h>
9984de1a	13	#include <linux/export.h>
ab516013	14	#include <linux/nsproxy.h>
0437eb59	15	#include <linux/init_task.h>
6b3286ed	16	#include <linux/mnt_namespace.h>
4865ecf1	17	#include <linux/utsname.h>
9a575a92	18	#include <linux/pid_namespace.h>
9dd776b6	19	#include <net/net_namespace.h>
ae5e1b22	20	#include <linux/ipc_namespace.h>
769071ac	21	#include <linux/time_namespace.h>
f2a8d52e	22	#include <linux/fs_struct.h>
303cc571	23	#include <linux/proc_fs.h>
0bb80f24	24	#include <linux/proc_ns.h>
0663c6f8 EB	25	#include <linux/file.h>
0663c6f8 EB	26	#include <linux/syscalls.h>
a79a908f	27	#include <linux/cgroup.h>
e4222673	28	#include <linux/perf_event.h>
0437eb59	29
98c0d07c CLG	30	static struct kmem_cache *nsproxy_cachep;
98c0d07c CLG	31
8467005d	32	struct nsproxy init_nsproxy = {
2ddd3cac	33	.count = REFCOUNT_INIT(1),
c2b1df2e	34	.uts_ns = &init_uts_ns,
8467005d	35	#if defined(CONFIG_POSIX_MQUEUE) \|\| defined(CONFIG_SYSVIPC)
c2b1df2e	36	.ipc_ns = &init_ipc_ns,
8467005d	37	#endif
c2b1df2e AL	38	.mnt_ns = NULL,
c2b1df2e AL	39	.pid_ns_for_children = &init_pid_ns,
8467005d	40	#ifdef CONFIG_NET
c2b1df2e	41	.net_ns = &init_net,
8467005d	42	#endif
a79a908f AK	43	#ifdef CONFIG_CGROUPS
	44	.cgroup_ns = &init_cgroup_ns,
	45	#endif
769071ac AV	46	#ifdef CONFIG_TIME_NS
	47	.time_ns = &init_time_ns,
	48	.time_ns_for_children = &init_time_ns,
	49	#endif
8467005d	50	};
ab516013	51
90af90d7	52	static inline struct nsproxy *create_nsproxy(void)
ab516013	53	{
90af90d7	54	struct nsproxy *nsproxy;
ab516013	55
90af90d7 AD	56	nsproxy = kmem_cache_alloc(nsproxy_cachep, GFP_KERNEL);
90af90d7 AD	57	if (nsproxy)
2ddd3cac	58	refcount_set(&nsproxy->count, 1);
90af90d7	59	return nsproxy;
ab516013 SH	60	}
	61
	62	/*
e3222c4e BP	63	* Create new nsproxy and all of its the associated namespaces.
	64	* Return the newly created nsproxy. Do not attach this to the task,
	65	* leave it to the caller to do proper locking and attach it to task.
ab516013	66	*/
213dd266	67	static struct nsproxy *create_new_namespaces(unsigned long flags,
bcf58e72 EB	68	struct task_struct tsk, struct user_namespace user_ns,
bcf58e72 EB	69	struct fs_struct *new_fs)
ab516013	70	{
e3222c4e	71	struct nsproxy *new_nsp;
467e9f4b	72	int err;
ab516013	73
90af90d7	74	new_nsp = create_nsproxy();
e3222c4e BP	75	if (!new_nsp)
e3222c4e BP	76	return ERR_PTR(-ENOMEM);
1651e14e	77
bcf58e72	78	new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, user_ns, new_fs);
467e9f4b CLG	79	if (IS_ERR(new_nsp->mnt_ns)) {
467e9f4b CLG	80	err = PTR_ERR(new_nsp->mnt_ns);
e3222c4e	81	goto out_ns;
467e9f4b	82	}
e3222c4e	83
bcf58e72	84	new_nsp->uts_ns = copy_utsname(flags, user_ns, tsk->nsproxy->uts_ns);
467e9f4b CLG	85	if (IS_ERR(new_nsp->uts_ns)) {
467e9f4b CLG	86	err = PTR_ERR(new_nsp->uts_ns);
e3222c4e	87	goto out_uts;
467e9f4b	88	}
e3222c4e	89
bcf58e72	90	new_nsp->ipc_ns = copy_ipcs(flags, user_ns, tsk->nsproxy->ipc_ns);
467e9f4b CLG	91	if (IS_ERR(new_nsp->ipc_ns)) {
467e9f4b CLG	92	err = PTR_ERR(new_nsp->ipc_ns);
e3222c4e	93	goto out_ipc;
467e9f4b	94	}
e3222c4e	95
c2b1df2e AL	96	new_nsp->pid_ns_for_children =
	97	copy_pid_ns(flags, user_ns, tsk->nsproxy->pid_ns_for_children);
	98	if (IS_ERR(new_nsp->pid_ns_for_children)) {
	99	err = PTR_ERR(new_nsp->pid_ns_for_children);
e3222c4e	100	goto out_pid;
467e9f4b	101	}
e3222c4e	102
a79a908f AK	103	new_nsp->cgroup_ns = copy_cgroup_ns(flags, user_ns,
	104	tsk->nsproxy->cgroup_ns);
	105	if (IS_ERR(new_nsp->cgroup_ns)) {
	106	err = PTR_ERR(new_nsp->cgroup_ns);
	107	goto out_cgroup;
	108	}
	109
bcf58e72	110	new_nsp->net_ns = copy_net_ns(flags, user_ns, tsk->nsproxy->net_ns);
9dd776b6 EB	111	if (IS_ERR(new_nsp->net_ns)) {
	112	err = PTR_ERR(new_nsp->net_ns);
	113	goto out_net;
	114	}
	115
769071ac AV	116	new_nsp->time_ns_for_children = copy_time_ns(flags, user_ns,
	117	tsk->nsproxy->time_ns_for_children);
	118	if (IS_ERR(new_nsp->time_ns_for_children)) {
	119	err = PTR_ERR(new_nsp->time_ns_for_children);
	120	goto out_time;
	121	}
	122	new_nsp->time_ns = get_time_ns(tsk->nsproxy->time_ns);
	123
e3222c4e BP	124	return new_nsp;
e3222c4e BP	125
769071ac AV	126	out_time:
769071ac AV	127	put_net(new_nsp->net_ns);
9dd776b6	128	out_net:
a79a908f AK	129	put_cgroup_ns(new_nsp->cgroup_ns);
a79a908f AK	130	out_cgroup:
c2b1df2e AL	131	if (new_nsp->pid_ns_for_children)
c2b1df2e AL	132	put_pid_ns(new_nsp->pid_ns_for_children);
e3222c4e BP	133	out_pid:
	134	if (new_nsp->ipc_ns)
	135	put_ipc_ns(new_nsp->ipc_ns);
	136	out_ipc:
	137	if (new_nsp->uts_ns)
	138	put_uts_ns(new_nsp->uts_ns);
	139	out_uts:
	140	if (new_nsp->mnt_ns)
	141	put_mnt_ns(new_nsp->mnt_ns);
	142	out_ns:
98c0d07c	143	kmem_cache_free(nsproxy_cachep, new_nsp);
467e9f4b	144	return ERR_PTR(err);
ab516013 SH	145	}
	146
	147	/*
	148	* called from clone. This now handles copy for nsproxy and all
	149	* namespaces therein.
	150	*/
213dd266	151	int copy_namespaces(unsigned long flags, struct task_struct *tsk)
ab516013 SH	152	{
ab516013 SH	153	struct nsproxy *old_ns = tsk->nsproxy;
b33c77ef	154	struct user_namespace *user_ns = task_cred_xxx(tsk, user_ns);
1651e14e	155	struct nsproxy *new_ns;
ab516013	156
dbef0c1c	157	if (likely(!(flags & (CLONE_NEWNS \| CLONE_NEWUTS \| CLONE_NEWIPC \|
a79a908f	158	CLONE_NEWPID \| CLONE_NEWNET \|
769071ac	159	CLONE_NEWCGROUP \| CLONE_NEWTIME)))) {
2b5f9dad AV	160	if ((flags & CLONE_VM) \|\|
2b5f9dad AV	161	likely(old_ns->time_ns_for_children == old_ns->time_ns)) {
769071ac AV	162	get_nsproxy(old_ns);
	163	return 0;
	164	}
	165	} else if (!ns_capable(user_ns, CAP_SYS_ADMIN))
dbef0c1c EB	166	return -EPERM;
dbef0c1c EB	167
02fdb36a SH	168	/*
	169	* CLONE_NEWIPC must detach from the undolist: after switching
	170	* to a new ipc namespace, the semaphore arrays from the old
	171	* namespace are unreachable. In clone parlance, CLONE_SYSVSEM
	172	* means share undolist with parent, so we must forbid using
	173	* it along with CLONE_NEWIPC.
	174	*/
21e85194	175	if ((flags & (CLONE_NEWIPC \| CLONE_SYSVSEM)) ==
aabe19b8	176	(CLONE_NEWIPC \| CLONE_SYSVSEM))
dbef0c1c	177	return -EINVAL;
02fdb36a	178
d7d48f62	179	new_ns = create_new_namespaces(flags, tsk, user_ns, tsk->fs);
dbef0c1c EB	180	if (IS_ERR(new_ns))
dbef0c1c EB	181	return PTR_ERR(new_ns);
9a575a92	182
2b5f9dad AV	183	if ((flags & CLONE_VM) == 0)
2b5f9dad AV	184	timens_on_fork(new_ns, tsk);
769071ac	185
e3222c4e	186	tsk->nsproxy = new_ns;
dbef0c1c	187	return 0;
ab516013 SH	188	}
	189
	190	void free_nsproxy(struct nsproxy *ns)
	191	{
9a575a92 CLG	192	if (ns->mnt_ns)
	193	put_mnt_ns(ns->mnt_ns);
	194	if (ns->uts_ns)
	195	put_uts_ns(ns->uts_ns);
	196	if (ns->ipc_ns)
	197	put_ipc_ns(ns->ipc_ns);
c2b1df2e AL	198	if (ns->pid_ns_for_children)
c2b1df2e AL	199	put_pid_ns(ns->pid_ns_for_children);
769071ac AV	200	if (ns->time_ns)
	201	put_time_ns(ns->time_ns);
	202	if (ns->time_ns_for_children)
	203	put_time_ns(ns->time_ns_for_children);
a79a908f	204	put_cgroup_ns(ns->cgroup_ns);
9dd776b6	205	put_net(ns->net_ns);
98c0d07c	206	kmem_cache_free(nsproxy_cachep, ns);
ab516013	207	}
e3222c4e BP	208
	209	/*
	210	* Called from unshare. Unshare all the namespaces part of nsproxy.
4e71e474	211	* On success, returns the new nsproxy.
e3222c4e BP	212	*/
e3222c4e BP	213	int unshare_nsproxy_namespaces(unsigned long unshare_flags,
b2e0d987	214	struct nsproxy *new_nsp, struct cred new_cred, struct fs_struct *new_fs)
e3222c4e	215	{
bcf58e72	216	struct user_namespace *user_ns;
e3222c4e BP	217	int err = 0;
e3222c4e BP	218
77ec739d	219	if (!(unshare_flags & (CLONE_NEWNS \| CLONE_NEWUTS \| CLONE_NEWIPC \|
769071ac AV	220	CLONE_NEWNET \| CLONE_NEWPID \| CLONE_NEWCGROUP \|
769071ac AV	221	CLONE_NEWTIME)))
e3222c4e BP	222	return 0;
e3222c4e BP	223
b2e0d987 EB	224	user_ns = new_cred ? new_cred->user_ns : current_user_ns();
b2e0d987 EB	225	if (!ns_capable(user_ns, CAP_SYS_ADMIN))
e3222c4e BP	226	return -EPERM;
e3222c4e BP	227
bcf58e72	228	*new_nsp = create_new_namespaces(unshare_flags, current, user_ns,
b2e0d987	229	new_fs ? new_fs : current->fs);
858d72ea	230	if (IS_ERR(*new_nsp)) {
e3222c4e	231	err = PTR_ERR(*new_nsp);
858d72ea SH	232	goto out;
	233	}
	234
858d72ea	235	out:
e3222c4e BP	236	return err;
e3222c4e BP	237	}
98c0d07c	238
cf7b708c PE	239	void switch_task_namespaces(struct task_struct p, struct nsproxy new)
	240	{
	241	struct nsproxy *ns;
	242
	243	might_sleep();
	244
728dba3a	245	task_lock(p);
cf7b708c	246	ns = p->nsproxy;
728dba3a EB	247	p->nsproxy = new;
728dba3a EB	248	task_unlock(p);
cf7b708c	249
aabe19b8 HS	250	if (ns)
aabe19b8 HS	251	put_nsproxy(ns);
cf7b708c PE	252	}
	253
	254	void exit_task_namespaces(struct task_struct *p)
	255	{
	256	switch_task_namespaces(p, NULL);
	257	}
	258
2b5f9dad AV	259	int exec_task_namespaces(void)
	260	{
	261	struct task_struct *tsk = current;
	262	struct nsproxy *new;
	263
	264	if (tsk->nsproxy->time_ns_for_children == tsk->nsproxy->time_ns)
	265	return 0;
	266
	267	new = create_new_namespaces(0, tsk, current_user_ns(), tsk->fs);
	268	if (IS_ERR(new))
	269	return PTR_ERR(new);
	270
	271	timens_on_fork(new, tsk);
	272	switch_task_namespaces(tsk, new);
	273	return 0;
	274	}
	275
303cc571 CB	276	static int check_setns_flags(unsigned long flags)
	277	{
	278	if (!flags \|\| (flags & ~(CLONE_NEWNS \| CLONE_NEWUTS \| CLONE_NEWIPC \|
76c12881 CB	279	CLONE_NEWNET \| CLONE_NEWTIME \| CLONE_NEWUSER \|
76c12881 CB	280	CLONE_NEWPID \| CLONE_NEWCGROUP)))
303cc571 CB	281	return -EINVAL;
	282
	283	#ifndef CONFIG_USER_NS
	284	if (flags & CLONE_NEWUSER)
	285	return -EINVAL;
	286	#endif
	287	#ifndef CONFIG_PID_NS
	288	if (flags & CLONE_NEWPID)
	289	return -EINVAL;
	290	#endif
	291	#ifndef CONFIG_UTS_NS
	292	if (flags & CLONE_NEWUTS)
	293	return -EINVAL;
	294	#endif
	295	#ifndef CONFIG_IPC_NS
	296	if (flags & CLONE_NEWIPC)
	297	return -EINVAL;
	298	#endif
	299	#ifndef CONFIG_CGROUPS
	300	if (flags & CLONE_NEWCGROUP)
	301	return -EINVAL;
	302	#endif
	303	#ifndef CONFIG_NET_NS
	304	if (flags & CLONE_NEWNET)
	305	return -EINVAL;
	306	#endif
76c12881 CB	307	#ifndef CONFIG_TIME_NS
	308	if (flags & CLONE_NEWTIME)
	309	return -EINVAL;
	310	#endif
303cc571 CB	311
	312	return 0;
	313	}
	314
f2a8d52e CB	315	static void put_nsset(struct nsset *nsset)
	316	{
	317	unsigned flags = nsset->flags;
	318
	319	if (flags & CLONE_NEWUSER)
	320	put_cred(nsset_cred(nsset));
303cc571 CB	321	/*
	322	* We only created a temporary copy if we attached to more than just
	323	* the mount namespace.
	324	*/
	325	if (nsset->fs && (flags & CLONE_NEWNS) && (flags & ~CLONE_NEWNS))
	326	free_fs_struct(nsset->fs);
f2a8d52e CB	327	if (nsset->nsproxy)
	328	free_nsproxy(nsset->nsproxy);
	329	}
	330
303cc571	331	static int prepare_nsset(unsigned flags, struct nsset *nsset)
f2a8d52e CB	332	{
	333	struct task_struct *me = current;
	334
	335	nsset->nsproxy = create_new_namespaces(0, me, current_user_ns(), me->fs);
	336	if (IS_ERR(nsset->nsproxy))
	337	return PTR_ERR(nsset->nsproxy);
	338
303cc571	339	if (flags & CLONE_NEWUSER)
f2a8d52e CB	340	nsset->cred = prepare_creds();
	341	else
	342	nsset->cred = current_cred();
	343	if (!nsset->cred)
	344	goto out;
	345
303cc571 CB	346	/* Only create a temporary copy of fs_struct if we really need to. */
303cc571 CB	347	if (flags == CLONE_NEWNS) {
f2a8d52e	348	nsset->fs = me->fs;
303cc571 CB	349	} else if (flags & CLONE_NEWNS) {
	350	nsset->fs = copy_fs_struct(me->fs);
	351	if (!nsset->fs)
	352	goto out;
	353	}
f2a8d52e	354
303cc571	355	nsset->flags = flags;
f2a8d52e CB	356	return 0;
	357
	358	out:
	359	put_nsset(nsset);
	360	return -ENOMEM;
	361	}
	362
303cc571 CB	363	static inline int validate_ns(struct nsset nsset, struct ns_common ns)
	364	{
	365	return ns->ops->install(nsset, ns);
	366	}
	367
	368	/*
	369	* This is the inverse operation to unshare().
	370	* Ordering is equivalent to the standard ordering used everywhere else
	371	* during unshare and process creation. The switch to the new set of
	372	* namespaces occurs at the point of no return after installation of
	373	* all requested namespaces was successful in commit_nsset().
	374	*/
	375	static int validate_nsset(struct nsset nsset, struct pid pid)
	376	{
	377	int ret = 0;
	378	unsigned flags = nsset->flags;
	379	struct user_namespace *user_ns = NULL;
	380	struct pid_namespace *pid_ns = NULL;
	381	struct nsproxy *nsp;
	382	struct task_struct *tsk;
	383
	384	/* Take a "snapshot" of the target task's namespaces. */
	385	rcu_read_lock();
	386	tsk = pid_task(pid, PIDTYPE_PID);
	387	if (!tsk) {
	388	rcu_read_unlock();
	389	return -ESRCH;
	390	}
	391
	392	if (!ptrace_may_access(tsk, PTRACE_MODE_READ_REALCREDS)) {
	393	rcu_read_unlock();
	394	return -EPERM;
	395	}
	396
	397	task_lock(tsk);
	398	nsp = tsk->nsproxy;
	399	if (nsp)
	400	get_nsproxy(nsp);
	401	task_unlock(tsk);
	402	if (!nsp) {
	403	rcu_read_unlock();
	404	return -ESRCH;
	405	}
	406
	407	#ifdef CONFIG_PID_NS
	408	if (flags & CLONE_NEWPID) {
	409	pid_ns = task_active_pid_ns(tsk);
	410	if (unlikely(!pid_ns)) {
	411	rcu_read_unlock();
	412	ret = -ESRCH;
	413	goto out;
	414	}
	415	get_pid_ns(pid_ns);
	416	}
	417	#endif
	418
	419	#ifdef CONFIG_USER_NS
	420	if (flags & CLONE_NEWUSER)
	421	user_ns = get_user_ns(__task_cred(tsk)->user_ns);
	422	#endif
	423	rcu_read_unlock();
	424
	425	/*
	426	* Install requested namespaces. The caller will have
427	* verified earlier that the requested namespaces are
428	* supported on this kernel. We don't report errors here
429	* if a namespace is requested that isn't supported.
430	*/
431	#ifdef CONFIG_USER_NS
432	if (flags & CLONE_NEWUSER) {
433	ret = validate_ns(nsset, &user_ns->ns);
434	if (ret)
435	goto out;
436	}
437	#endif
438
439	if (flags & CLONE_NEWNS) {
440	ret = validate_ns(nsset, from_mnt_ns(nsp->mnt_ns));
441	if (ret)
442	goto out;
443	}
444
445	#ifdef CONFIG_UTS_NS
446	if (flags & CLONE_NEWUTS) {
447	ret = validate_ns(nsset, &nsp->uts_ns->ns);
448	if (ret)
449	goto out;
450	}
451	#endif
452
453	#ifdef CONFIG_IPC_NS
454	if (flags & CLONE_NEWIPC) {
455	ret = validate_ns(nsset, &nsp->ipc_ns->ns);
456	if (ret)
457	goto out;
458	}
459	#endif
460
461	#ifdef CONFIG_PID_NS
462	if (flags & CLONE_NEWPID) {
463	ret = validate_ns(nsset, &pid_ns->ns);
464	if (ret)
465	goto out;
466	}
467	#endif
468
469	#ifdef CONFIG_CGROUPS
470	if (flags & CLONE_NEWCGROUP) {
471	ret = validate_ns(nsset, &nsp->cgroup_ns->ns);
472	if (ret)
473	goto out;
474	}
475	#endif
476
477	#ifdef CONFIG_NET_NS
478	if (flags & CLONE_NEWNET) {
479	ret = validate_ns(nsset, &nsp->net_ns->ns);
480	if (ret)
481	goto out;
482	}
483	#endif
484
76c12881 CB	485	#ifdef CONFIG_TIME_NS
	486	if (flags & CLONE_NEWTIME) {
	487	ret = validate_ns(nsset, &nsp->time_ns->ns);
	488	if (ret)
	489	goto out;
	490	}
	491	#endif
	492
303cc571 CB	493	out:
	494	if (pid_ns)
	495	put_pid_ns(pid_ns);
	496	if (nsp)
	497	put_nsproxy(nsp);
	498	put_user_ns(user_ns);
	499
	500	return ret;
	501	}
	502
f2a8d52e CB	503	/*
	504	* This is the point of no return. There are just a few namespaces
	505	* that do some actual work here and it's sufficiently minimal that
	506	* a separate ns_common operation seems unnecessary for now.
	507	* Unshare is doing the same thing. If we'll end up needing to do
	508	* more in a given namespace or a helper here is ultimately not
	509	* exported anymore a simple commit handler for each namespace
	510	* should be added to ns_common.
	511	*/
	512	static void commit_nsset(struct nsset *nsset)
	513	{
	514	unsigned flags = nsset->flags;
	515	struct task_struct *me = current;
	516
	517	#ifdef CONFIG_USER_NS
	518	if (flags & CLONE_NEWUSER) {
	519	/* transfer ownership */
	520	commit_creds(nsset_cred(nsset));
	521	nsset->cred = NULL;
	522	}
	523	#endif
	524
303cc571 CB	525	/* We only need to commit if we have used a temporary fs_struct. */
	526	if ((flags & CLONE_NEWNS) && (flags & ~CLONE_NEWNS)) {
	527	set_fs_root(me->fs, &nsset->fs->root);
	528	set_fs_pwd(me->fs, &nsset->fs->pwd);
	529	}
	530
f2a8d52e CB	531	#ifdef CONFIG_IPC_NS
	532	if (flags & CLONE_NEWIPC)
	533	exit_sem(me);
	534	#endif
	535
76c12881 CB	536	#ifdef CONFIG_TIME_NS
	537	if (flags & CLONE_NEWTIME)
	538	timens_commit(me, nsset->nsproxy->time_ns);
	539	#endif
	540
f2a8d52e CB	541	/* transfer ownership */
	542	switch_task_namespaces(me, nsset->nsproxy);
	543	nsset->nsproxy = NULL;
	544	}
	545
303cc571	546	SYSCALL_DEFINE2(setns, int, fd, int, flags)
0663c6f8	547	{
281117cc	548	struct fd f = fdget(fd);
303cc571	549	struct ns_common *ns = NULL;
f2a8d52e	550	struct nsset nsset = {};
303cc571	551	int err = 0;
0663c6f8	552
281117cc	553	if (!f.file)
303cc571 CB	554	return -EBADF;
303cc571 CB	555
281117cc AV	556	if (proc_ns_file(f.file)) {
281117cc AV	557	ns = get_proc_ns(file_inode(f.file));
303cc571 CB	558	if (flags && (ns->ops->type != flags))
	559	err = -EINVAL;
	560	flags = ns->ops->type;
281117cc	561	} else if (!IS_ERR(pidfd_pid(f.file))) {
303cc571 CB	562	err = check_setns_flags(flags);
303cc571 CB	563	} else {
e571d4ee	564	err = -EINVAL;
303cc571 CB	565	}
303cc571 CB	566	if (err)
0663c6f8 EB	567	goto out;
0663c6f8 EB	568
303cc571	569	err = prepare_nsset(flags, &nsset);
f2a8d52e	570	if (err)
0663c6f8	571	goto out;
0663c6f8	572
281117cc	573	if (proc_ns_file(f.file))
303cc571 CB	574	err = validate_ns(&nsset, ns);
303cc571 CB	575	else
cb12fd8e	576	err = validate_nsset(&nsset, pidfd_pid(f.file));
f2a8d52e CB	577	if (!err) {
	578	commit_nsset(&nsset);
	579	perf_event_namespaces(current);
0663c6f8	580	}
f2a8d52e	581	put_nsset(&nsset);
0663c6f8	582	out:
281117cc	583	fdput(f);
0663c6f8 EB	584	return err;
	585	}
	586
66577193	587	int __init nsproxy_cache_init(void)
98c0d07c	588	{
30acd0bd	589	nsproxy_cachep = KMEM_CACHE(nsproxy, SLAB_PANIC\|SLAB_ACCOUNT);
98c0d07c CLG	590	return 0;
98c0d07c CLG	591	}