Merge tag 'threads-v5.8' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux

Pull thread updates from Christian Brauner:
 "We have been discussing using pidfds to attach to namespaces for quite
  a while and the patches have in one form or another already existed
  for about a year. But I wanted to wait to see how the general api
  would be received and adopted.

  This contains the changes to make it possible to use pidfds to attach
  to the namespaces of a process, i.e. they can be passed as the first
  argument to the setns() syscall.

  When only a single namespace type is specified the semantics are
  equivalent to passing an nsfd. That means setns(nsfd, CLONE_NEWNET)
  equals setns(pidfd, CLONE_NEWNET).

  However, when a pidfd is passed, multiple namespace flags can be
  specified in the second setns() argument and setns() will attach the
  caller to all the specified namespaces all at once or to none of them.

  Specifying 0 is not valid together with a pidfd. Here are just two
  obvious examples:

    setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
    setns(pidfd, CLONE_NEWUSER);

  Allowing to also attach subsets of namespaces supports various
  use-cases where callers setns to a subset of namespaces to retain
  privilege, perform an action and then re-attach another subset of
  namespaces.

  Apart from significantly reducing the number of syscalls needed to
  attach to all currently supported namespaces (eight "open+setns"
  sequences vs just a single "setns()"), this also allows atomic setns
  to a set of namespaces, i.e. either attaching to all namespaces
  succeeds or we fail without having changed anything.

  This is centered around a new internal struct nsset which holds all
  information necessary for a task to switch to a new set of namespaces
  atomically. Fwiw, with this change a pidfd becomes the only token
  needed to interact with a container. I'm expecting this to be
  picked-up by util-linux for nsenter rather soon.

  Associated with this change is a shiny new test-suite dedicated to
  setns() (for pidfds and nsfds alike)"

* tag 'threads-v5.8' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
  selftests/pidfd: add pidfd setns tests
  nsproxy: attach to namespaces via pidfds
  nsproxy: add struct nsset

fs/namespace.c

@@ -1786,6 +1786,11 @@ static struct mnt_namespace *to_mnt_ns(struct ns_common *ns)
 	return container_of(ns, struct mnt_namespace, ns);
 }
 
+struct ns_common *from_mnt_ns(struct mnt_namespace *mnt)
+{
+	return &mnt->ns;
+}
+
 static bool mnt_ns_loop(struct dentry *dentry)
 {
 	/* Could bind mounting the mount namespace inode cause a
@@ -4013,16 +4018,18 @@ static void mntns_put(struct ns_common *ns)
 	put_mnt_ns(to_mnt_ns(ns));
 }
 
-static int mntns_install(struct nsproxy *nsproxy, struct ns_common *ns)
+static int mntns_install(struct nsset *nsset, struct ns_common *ns)
 {
-	struct fs_struct *fs = current->fs;
+	struct nsproxy *nsproxy = nsset->nsproxy;
+	struct fs_struct *fs = nsset->fs;
 	struct mnt_namespace *mnt_ns = to_mnt_ns(ns), *old_mnt_ns;
+	struct user_namespace *user_ns = nsset->cred->user_ns;
 	struct path root;
 	int err;
 
 	if (!ns_capable(mnt_ns->user_ns, CAP_SYS_ADMIN) ||
-	    !ns_capable(current_user_ns(), CAP_SYS_CHROOT) ||
-	    !ns_capable(current_user_ns(), CAP_SYS_ADMIN))
+	    !ns_capable(user_ns, CAP_SYS_CHROOT) ||
+	    !ns_capable(user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 
 	if (is_anon_ns(mnt_ns))

fs/nsfs.c

@@ -229,6 +229,11 @@ int ns_get_name(char *buf, size_t size, struct task_struct *task,
 	return res;
 }
 
+bool proc_ns_file(const struct file *file)
+{
+	return file->f_op == &ns_file_operations;
+}
+
 struct file *proc_ns_fget(int fd)
 {
 	struct file *file;

include/linux/mnt_namespace.h

@@ -6,10 +6,12 @@
 struct mnt_namespace;
 struct fs_struct;
 struct user_namespace;
+struct ns_common;
 
 extern struct mnt_namespace *copy_mnt_ns(unsigned long, struct mnt_namespace *,
 		struct user_namespace *, struct fs_struct *);
 extern void put_mnt_ns(struct mnt_namespace *ns);
+extern struct ns_common *from_mnt_ns(struct mnt_namespace *);
 
 extern const struct file_operations proc_mounts_operations;
 extern const struct file_operations proc_mountinfo_operations;

include/linux/nsproxy.h

@@ -41,6 +41,30 @@ struct nsproxy {
 };
 extern struct nsproxy init_nsproxy;
 
+/*
+ * A structure to encompass all bits needed to install
+ * a partial or complete new set of namespaces.
+ *
+ * If a new user namespace is requested cred will
+ * point to a modifiable set of credentials. If a pointer
+ * to a modifiable set is needed nsset_cred() must be
+ * used and tested.
+ */
+struct nsset {
+	unsigned flags;
+	struct nsproxy *nsproxy;
+	struct fs_struct *fs;
+	const struct cred *cred;
+};
+
+static inline struct cred *nsset_cred(struct nsset *set)
+{
+	if (set->flags & CLONE_NEWUSER)
+		return (struct cred *)set->cred;
+
+	return NULL;
+}
+
 /*
  * the namespaces access rules are:
  *

include/linux/proc_fs.h

@@ -179,4 +179,6 @@ static inline struct pid_namespace *proc_pid_ns(const struct inode *inode)
 	return inode->i_sb->s_fs_info;
 }
 
+bool proc_ns_file(const struct file *file);
+
 #endif /* _LINUX_PROC_FS_H */

include/linux/proc_ns.h

@@ -8,7 +8,7 @@
 #include <linux/ns_common.h>
 
 struct pid_namespace;
-struct nsproxy;
+struct nsset;
 struct path;
 struct task_struct;
 struct inode;
@@ -19,7 +19,7 @@ struct proc_ns_operations {
 	int type;
 	struct ns_common *(*get)(struct task_struct *task);
 	void (*put)(struct ns_common *ns);
-	int (*install)(struct nsproxy *nsproxy, struct ns_common *ns);
+	int (*install)(struct nsset *nsset, struct ns_common *ns);
 	struct user_namespace *(*owner)(struct ns_common *ns);
 	struct ns_common *(*get_parent)(struct ns_common *ns);
 } __randomize_layout;

ipc/namespace.c

@@ -177,15 +177,14 @@ static void ipcns_put(struct ns_common *ns)
 	return put_ipc_ns(to_ipc_ns(ns));
 }
 
-static int ipcns_install(struct nsproxy *nsproxy, struct ns_common *new)
+static int ipcns_install(struct nsset *nsset, struct ns_common *new)
 {
+	struct nsproxy *nsproxy = nsset->nsproxy;
 	struct ipc_namespace *ns = to_ipc_ns(new);
 	if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN) ||
-	    !ns_capable(current_user_ns(), CAP_SYS_ADMIN))
+	    !ns_capable(nsset->cred->user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 
-	/* Ditch state from the old ipc namespace */
-	exit_sem(current);
 	put_ipc_ns(nsproxy->ipc_ns);
 	nsproxy->ipc_ns = get_ipc_ns(ns);
 	return 0;

kernel/cgroup/namespace.c

@@ -95,11 +95,12 @@ static inline struct cgroup_namespace *to_cg_ns(struct ns_common *ns)
 	return container_of(ns, struct cgroup_namespace, ns);
 }
 
-static int cgroupns_install(struct nsproxy *nsproxy, struct ns_common *ns)
+static int cgroupns_install(struct nsset *nsset, struct ns_common *ns)
 {
+	struct nsproxy *nsproxy = nsset->nsproxy;
 	struct cgroup_namespace *cgroup_ns = to_cg_ns(ns);
 
-	if (!ns_capable(current_user_ns(), CAP_SYS_ADMIN) ||
+	if (!ns_capable(nsset->cred->user_ns, CAP_SYS_ADMIN) ||
 	    !ns_capable(cgroup_ns->user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 

kernel/nsproxy.c

@@ -19,6 +19,8 @@
 #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>
@@ -257,37 +259,296 @@ void exit_task_namespaces(struct task_struct *p)
 	switch_task_namespaces(p, NULL);
 }
 
-SYSCALL_DEFINE2(setns, int, fd, int, nstype)
+static int check_setns_flags(unsigned long flags)
 {
-	struct task_struct *tsk = current;
-	struct nsproxy *new_nsproxy;
-	struct file *file;
-	struct ns_common *ns;
-	int err;
+	if (!flags || (flags & ~(CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC |
+				 CLONE_NEWNET | CLONE_NEWUSER | CLONE_NEWPID |
+				 CLONE_NEWCGROUP)))
+		return -EINVAL;
 
-	file = proc_ns_fget(fd);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
+#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
 
-	err = -EINVAL;
-	ns = get_proc_ns(file_inode(file));
-	if (nstype && (ns->ops->type != nstype))
-		goto out;
+	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);
 
-	new_nsproxy = create_new_namespaces(0, tsk, current_user_ns(), tsk->fs);
-	if (IS_ERR(new_nsproxy)) {
-		err = PTR_ERR(new_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;
 	}
 
-	err = ns->ops->install(new_nsproxy, ns);
-	if (err) {
-		free_nsproxy(new_nsproxy);
-		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
+
+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);
 	}
-	switch_task_namespaces(tsk, new_nsproxy);
 
-	perf_event_namespaces(tsk);
+#ifdef CONFIG_IPC_NS
+	if (flags & CLONE_NEWIPC)
+		exit_sem(me);
+#endif
+
+	/* transfer ownership */
+	switch_task_namespaces(me, nsset->nsproxy);
+	nsset->nsproxy = NULL;
+}
+
+SYSCALL_DEFINE2(setns, int, fd, int, flags)
+{
+	struct file *file;
+	struct ns_common *ns = NULL;
+	struct nsset nsset = {};
+	int err = 0;
+
+	file = fget(fd);
+	if (!file)
+		return -EBADF;
+
+	if (proc_ns_file(file)) {
+		ns = get_proc_ns(file_inode(file));
+		if (flags && (ns->ops->type != flags))
+			err = -EINVAL;
+		flags = ns->ops->type;
+	} else if (!IS_ERR(pidfd_pid(file))) {
+		err = check_setns_flags(flags);
+	} else {
+		err = -EBADF;
+	}
+	if (err)
+		goto out;
+
+	err = prepare_nsset(flags, &nsset);
+	if (err)
+		goto out;
+
+	if (proc_ns_file(file))
+		err = validate_ns(&nsset, ns);
+	else
+		err = validate_nsset(&nsset, file->private_data);
+	if (!err) {
+		commit_nsset(&nsset);
+		perf_event_namespaces(current);
+	}
+	put_nsset(&nsset);
 out:
 	fput(file);
 	return err;

kernel/pid_namespace.c

@@ -378,13 +378,14 @@ static void pidns_put(struct ns_common *ns)
 	put_pid_ns(to_pid_ns(ns));
 }
 
-static int pidns_install(struct nsproxy *nsproxy, struct ns_common *ns)
+static int pidns_install(struct nsset *nsset, struct ns_common *ns)
 {
+	struct nsproxy *nsproxy = nsset->nsproxy;
 	struct pid_namespace *active = task_active_pid_ns(current);
 	struct pid_namespace *ancestor, *new = to_pid_ns(ns);
 
 	if (!ns_capable(new->user_ns, CAP_SYS_ADMIN) ||
-	    !ns_capable(current_user_ns(), CAP_SYS_ADMIN))
+	    !ns_capable(nsset->cred->user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 
 	/*

kernel/time/namespace.c

@@ -280,8 +280,9 @@ static void timens_put(struct ns_common *ns)
 	put_time_ns(to_time_ns(ns));
 }
 
-static int timens_install(struct nsproxy *nsproxy, struct ns_common *new)
+static int timens_install(struct nsset *nsset, struct ns_common *new)
 {
+	struct nsproxy *nsproxy = nsset->nsproxy;
 	struct time_namespace *ns = to_time_ns(new);
 	int err;
 
@@ -289,7 +290,7 @@ static int timens_install(struct nsproxy *nsproxy, struct ns_common *new)
 		return -EUSERS;
 
 	if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN) ||
-	    !ns_capable(current_user_ns(), CAP_SYS_ADMIN))
+	    !ns_capable(nsset->cred->user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 
 	timens_set_vvar_page(current, ns);

kernel/user_namespace.c

@@ -1253,7 +1253,7 @@ static void userns_put(struct ns_common *ns)
 	put_user_ns(to_user_ns(ns));
 }
 
-static int userns_install(struct nsproxy *nsproxy, struct ns_common *ns)
+static int userns_install(struct nsset *nsset, struct ns_common *ns)
 {
 	struct user_namespace *user_ns = to_user_ns(ns);
 	struct cred *cred;
@@ -1274,14 +1274,14 @@ static int userns_install(struct nsproxy *nsproxy, struct ns_common *ns)
 	if (!ns_capable(user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 
-	cred = prepare_creds();
+	cred = nsset_cred(nsset);
 	if (!cred)
-		return -ENOMEM;
+		return -EINVAL;
 
 	put_user_ns(cred->user_ns);
 	set_cred_user_ns(cred, get_user_ns(user_ns));
 
-	return commit_creds(cred);
+	return 0;
 }
 
 struct ns_common *ns_get_owner(struct ns_common *ns)

kernel/utsname.c

@@ -140,12 +140,13 @@ static void utsns_put(struct ns_common *ns)
 	put_uts_ns(to_uts_ns(ns));
 }
 
-static int utsns_install(struct nsproxy *nsproxy, struct ns_common *new)
+static int utsns_install(struct nsset *nsset, struct ns_common *new)
 {
+	struct nsproxy *nsproxy = nsset->nsproxy;
 	struct uts_namespace *ns = to_uts_ns(new);
 
 	if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN) ||
-	    !ns_capable(current_user_ns(), CAP_SYS_ADMIN))
+	    !ns_capable(nsset->cred->user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 
 	get_uts_ns(ns);

net/core/net_namespace.c

@@ -1353,12 +1353,13 @@ static void netns_put(struct ns_common *ns)
 	put_net(to_net_ns(ns));
 }
 
-static int netns_install(struct nsproxy *nsproxy, struct ns_common *ns)
+static int netns_install(struct nsset *nsset, struct ns_common *ns)
 {
+	struct nsproxy *nsproxy = nsset->nsproxy;
 	struct net *net = to_net_ns(ns);
 
 	if (!ns_capable(net->user_ns, CAP_SYS_ADMIN) ||
-	    !ns_capable(current_user_ns(), CAP_SYS_ADMIN))
+	    !ns_capable(nsset->cred->user_ns, CAP_SYS_ADMIN))
 		return -EPERM;
 
 	put_net(nsproxy->net_ns);

tools/testing/selftests/pidfd/.gitignore

@@ -5,3 +5,4 @@ pidfd_test
 pidfd_wait
 pidfd_fdinfo_test
 pidfd_getfd_test
+pidfd_setns_test

tools/testing/selftests/pidfd/Makefile

 # SPDX-License-Identifier: GPL-2.0-only
 CFLAGS += -g -I../../../../usr/include/ -pthread
 
-TEST_GEN_PROGS := pidfd_test pidfd_fdinfo_test pidfd_open_test pidfd_poll_test pidfd_wait pidfd_getfd_test
+TEST_GEN_PROGS := pidfd_test pidfd_fdinfo_test pidfd_open_test \
+	pidfd_poll_test pidfd_wait pidfd_getfd_test pidfd_setns_test
 
 include ../lib.mk
 

tools/testing/selftests/pidfd/config

+CONFIG_UTS_NS=y
+CONFIG_IPC_NS=y
+CONFIG_USER_NS=y
+CONFIG_PID_NS=y
+CONFIG_NET_NS=y
+CONFIG_CGROUPS=y

tools/testing/selftests/pidfd/pidfd_setns_test.c

+// SPDX-License-Identifier: GPL-2.0
+
+#define _GNU_SOURCE
+#include <errno.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <linux/types.h>
+#include <sched.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <syscall.h>
+#include <sys/prctl.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <linux/kcmp.h>
+
+#include "pidfd.h"
+#include "../clone3/clone3_selftests.h"
+#include "../kselftest.h"
+#include "../kselftest_harness.h"
+
+enum {
+	PIDFD_NS_USER,
+	PIDFD_NS_MNT,
+	PIDFD_NS_PID,
+	PIDFD_NS_UTS,
+	PIDFD_NS_IPC,
+	PIDFD_NS_NET,
+	PIDFD_NS_CGROUP,
+	PIDFD_NS_PIDCLD,
+	PIDFD_NS_MAX
+};
+
+const struct ns_info {
+	const char *name;
+	int flag;
+} ns_info[] = {
+	[PIDFD_NS_USER]   = { "user",             CLONE_NEWUSER,   },
+	[PIDFD_NS_MNT]    = { "mnt",              CLONE_NEWNS,     },
+	[PIDFD_NS_PID]    = { "pid",              CLONE_NEWPID,    },
+	[PIDFD_NS_UTS]    = { "uts",              CLONE_NEWUTS,    },
+	[PIDFD_NS_IPC]    = { "ipc",              CLONE_NEWIPC,    },
+	[PIDFD_NS_NET]    = { "net",              CLONE_NEWNET,    },
+	[PIDFD_NS_CGROUP] = { "cgroup",           CLONE_NEWCGROUP, },
+	[PIDFD_NS_PIDCLD] = { "pid_for_children", 0,               },
+};
+
+FIXTURE(current_nsset)
+{
+	pid_t pid;
+	int pidfd;
+	int nsfds[PIDFD_NS_MAX];
+
+	pid_t child_pid_exited;
+	int child_pidfd_exited;
+
+	pid_t child_pid1;
+	int child_pidfd1;
+	int child_nsfds1[PIDFD_NS_MAX];
+
+	pid_t child_pid2;
+	int child_pidfd2;
+	int child_nsfds2[PIDFD_NS_MAX];
+};
+
+static int sys_waitid(int which, pid_t pid, int options)
+{
+	return syscall(__NR_waitid, which, pid, NULL, options, NULL);
+}
+
+pid_t create_child(int *pidfd, unsigned flags)
+{
+	struct clone_args args = {
+		.flags		= CLONE_PIDFD | flags,
+		.exit_signal	= SIGCHLD,
+		.pidfd		= ptr_to_u64(pidfd),
+	};
+
+	return sys_clone3(&args, sizeof(struct clone_args));
+}
+
+FIXTURE_SETUP(current_nsset)
+{
+	int i, proc_fd, ret;
+
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		self->nsfds[i]		= -EBADF;
+		self->child_nsfds1[i]	= -EBADF;
+		self->child_nsfds2[i]	= -EBADF;
+	}
+
+	proc_fd = open("/proc/self/ns", O_DIRECTORY | O_CLOEXEC);
+	ASSERT_GE(proc_fd, 0) {
+		TH_LOG("%m - Failed to open /proc/self/ns");
+	}
+
+	self->pid = getpid();
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+		self->nsfds[i] = openat(proc_fd, info->name, O_RDONLY | O_CLOEXEC);
+		if (self->nsfds[i] < 0) {
+			EXPECT_EQ(errno, ENOENT) {
+				TH_LOG("%m - Failed to open %s namespace for process %d",
+				       info->name, self->pid);
+			}
+		}
+	}
+
+	self->pidfd = sys_pidfd_open(self->pid, 0);
+	EXPECT_GT(self->pidfd, 0) {
+		TH_LOG("%m - Failed to open pidfd for process %d", self->pid);
+	}
+
+	/* Create task that exits right away. */
+	self->child_pid_exited = create_child(&self->child_pidfd_exited,
+					      CLONE_NEWUSER | CLONE_NEWNET);
+	EXPECT_GT(self->child_pid_exited, 0);
+
+	if (self->child_pid_exited == 0)
+		_exit(EXIT_SUCCESS);
+
+	ASSERT_EQ(sys_waitid(P_PID, self->child_pid_exited, WEXITED | WNOWAIT), 0);
+
+	self->pidfd = sys_pidfd_open(self->pid, 0);
+	EXPECT_GE(self->pidfd, 0) {
+		TH_LOG("%m - Failed to open pidfd for process %d", self->pid);
+	}
+
+	/* Create tasks that will be stopped. */
+	self->child_pid1 = create_child(&self->child_pidfd1,
+					CLONE_NEWUSER | CLONE_NEWNS |
+					CLONE_NEWCGROUP | CLONE_NEWIPC |
+					CLONE_NEWUTS | CLONE_NEWPID |
+					CLONE_NEWNET);
+	EXPECT_GE(self->child_pid1, 0);
+
+	if (self->child_pid1 == 0) {
+		pause();
+		_exit(EXIT_SUCCESS);
+	}
+
+	self->child_pid2 = create_child(&self->child_pidfd2,
+					CLONE_NEWUSER | CLONE_NEWNS |
+					CLONE_NEWCGROUP | CLONE_NEWIPC |
+					CLONE_NEWUTS | CLONE_NEWPID |
+					CLONE_NEWNET);
+	EXPECT_GE(self->child_pid2, 0);
+
+	if (self->child_pid2 == 0) {
+		pause();
+		_exit(EXIT_SUCCESS);
+	}
+
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		char p[100];
+
+		const struct ns_info *info = &ns_info[i];
+
+		self->nsfds[i] = openat(proc_fd, info->name, O_RDONLY | O_CLOEXEC);
+		if (self->nsfds[i] < 0) {
+			EXPECT_EQ(errno, ENOENT) {
+				TH_LOG("%m - Failed to open %s namespace for process %d",
+				       info->name, self->pid);
+			}
+		}
+
+		ret = snprintf(p, sizeof(p), "/proc/%d/ns/%s",
+			       self->child_pid1, info->name);
+		EXPECT_GT(ret, 0);
+		EXPECT_LT(ret, sizeof(p));
+
+		self->child_nsfds1[i] = open(p, O_RDONLY | O_CLOEXEC);
+		if (self->child_nsfds1[i] < 0) {
+			EXPECT_EQ(errno, ENOENT) {
+				TH_LOG("%m - Failed to open %s namespace for process %d",
+				       info->name, self->child_pid1);
+			}
+		}
+
+		ret = snprintf(p, sizeof(p), "/proc/%d/ns/%s",
+			       self->child_pid2, info->name);
+		EXPECT_GT(ret, 0);
+		EXPECT_LT(ret, sizeof(p));
+
+		self->child_nsfds2[i] = open(p, O_RDONLY | O_CLOEXEC);
+		if (self->child_nsfds2[i] < 0) {
+			EXPECT_EQ(errno, ENOENT) {
+				TH_LOG("%m - Failed to open %s namespace for process %d",
+				       info->name, self->child_pid1);
+			}
+		}
+	}
+
+	close(proc_fd);
+}
+
+FIXTURE_TEARDOWN(current_nsset)
+{
+	int i;
+
+	ASSERT_EQ(sys_pidfd_send_signal(self->child_pidfd1,
+					SIGKILL, NULL, 0), 0);
+	ASSERT_EQ(sys_pidfd_send_signal(self->child_pidfd2,
+					SIGKILL, NULL, 0), 0);
+
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		if (self->nsfds[i] >= 0)
+			close(self->nsfds[i]);
+		if (self->child_nsfds1[i] >= 0)
+			close(self->child_nsfds1[i]);
+		if (self->child_nsfds2[i] >= 0)
+			close(self->child_nsfds2[i]);
+	}
+
+	if (self->child_pidfd1 >= 0)
+		EXPECT_EQ(0, close(self->child_pidfd1));
+	if (self->child_pidfd2 >= 0)
+		EXPECT_EQ(0, close(self->child_pidfd2));
+	ASSERT_EQ(sys_waitid(P_PID, self->child_pid_exited, WEXITED), 0);
+	ASSERT_EQ(sys_waitid(P_PID, self->child_pid1, WEXITED), 0);
+	ASSERT_EQ(sys_waitid(P_PID, self->child_pid2, WEXITED), 0);
+}
+
+static int preserve_ns(const int pid, const char *ns)
+{
+	int ret;
+	char path[50];
+
+	ret = snprintf(path, sizeof(path), "/proc/%d/ns/%s", pid, ns);
+	if (ret < 0 || (size_t)ret >= sizeof(path))
+		return -EIO;
+
+	return open(path, O_RDONLY | O_CLOEXEC);
+}
+
+static int in_same_namespace(int ns_fd1, pid_t pid2, const char *ns)
+{
+	int ns_fd2 = -EBADF;
+	int ret = -1;
+	struct stat ns_st1, ns_st2;
+
+	ret = fstat(ns_fd1, &ns_st1);
+	if (ret < 0)
+		return -1;
+
+	ns_fd2 = preserve_ns(pid2, ns);
+	if (ns_fd2 < 0)
+		return -1;
+
+	ret = fstat(ns_fd2, &ns_st2);
+	close(ns_fd2);
+	if (ret < 0)
+		return -1;
+
+	/* processes are in the same namespace */
+	if ((ns_st1.st_dev == ns_st2.st_dev) &&
+	    (ns_st1.st_ino == ns_st2.st_ino))
+		return 1;
+
+	/* processes are in different namespaces */
+	return 0;
+}
+
+/* Test that we can't pass garbage to the kernel. */
+TEST_F(current_nsset, invalid_flags)
+{
+	ASSERT_NE(setns(self->pidfd, 0), 0);
+	EXPECT_EQ(errno, EINVAL);
+
+	ASSERT_NE(setns(self->pidfd, -1), 0);
+	EXPECT_EQ(errno, EINVAL);
+
+	ASSERT_NE(setns(self->pidfd, CLONE_VM), 0);
+	EXPECT_EQ(errno, EINVAL);
+
+	ASSERT_NE(setns(self->pidfd, CLONE_NEWUSER | CLONE_VM), 0);
+	EXPECT_EQ(errno, EINVAL);
+}
+
+/* Test that we can't attach to a task that has already exited. */
+TEST_F(current_nsset, pidfd_exited_child)
+{
+	int i;
+	pid_t pid;
+
+	ASSERT_NE(setns(self->child_pidfd_exited, CLONE_NEWUSER | CLONE_NEWNET),
+		  0);
+	EXPECT_EQ(errno, ESRCH);
+
+	pid = getpid();
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+		/* Verify that we haven't changed any namespaces. */
+		if (self->nsfds[i] >= 0)
+			ASSERT_EQ(in_same_namespace(self->nsfds[i], pid, info->name), 1);
+	}
+}
+
+TEST_F(current_nsset, pidfd_incremental_setns)
+{
+	int i;
+	pid_t pid;
+
+	pid = getpid();
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+		int nsfd;
+
+		if (self->child_nsfds1[i] < 0)
+			continue;
+
+		if (info->flag) {
+			ASSERT_EQ(setns(self->child_pidfd1, info->flag), 0) {
+				TH_LOG("%m - Failed to setns to %s namespace of %d via pidfd %d",
+				       info->name, self->child_pid1,
+				       self->child_pidfd1);
+			}
+		}
+
+		/* Verify that we have changed to the correct namespaces. */
+		if (info->flag == CLONE_NEWPID)
+			nsfd = self->nsfds[i];
+		else
+			nsfd = self->child_nsfds1[i];
+		ASSERT_EQ(in_same_namespace(nsfd, pid, info->name), 1) {
+			TH_LOG("setns failed to place us correctly into %s namespace of %d via pidfd %d",
+			       info->name, self->child_pid1,
+			       self->child_pidfd1);
+		}
+		TH_LOG("Managed to correctly setns to %s namespace of %d via pidfd %d",
+		       info->name, self->child_pid1, self->child_pidfd1);
+	}
+}
+
+TEST_F(current_nsset, nsfd_incremental_setns)
+{
+	int i;
+	pid_t pid;
+
+	pid = getpid();
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+		int nsfd;
+
+		if (self->child_nsfds1[i] < 0)
+			continue;
+
+		if (info->flag) {
+			ASSERT_EQ(setns(self->child_nsfds1[i], info->flag), 0) {
+				TH_LOG("%m - Failed to setns to %s namespace of %d via nsfd %d",
+				       info->name, self->child_pid1,
+				       self->child_nsfds1[i]);
+			}
+		}
+
+		/* Verify that we have changed to the correct namespaces. */
+		if (info->flag == CLONE_NEWPID)
+			nsfd = self->nsfds[i];
+		else
+			nsfd = self->child_nsfds1[i];
+		ASSERT_EQ(in_same_namespace(nsfd, pid, info->name), 1) {
+			TH_LOG("setns failed to place us correctly into %s namespace of %d via nsfd %d",
+			       info->name, self->child_pid1,
+			       self->child_nsfds1[i]);
+		}
+		TH_LOG("Managed to correctly setns to %s namespace of %d via nsfd %d",
+		       info->name, self->child_pid1, self->child_nsfds1[i]);
+	}
+}
+
+TEST_F(current_nsset, pidfd_one_shot_setns)
+{
+	unsigned flags = 0;
+	int i;
+	pid_t pid;
+
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+
+		if (self->child_nsfds1[i] < 0)
+			continue;
+
+		flags |= info->flag;
+		TH_LOG("Adding %s namespace of %d to list of namespaces to attach to",
+		       info->name, self->child_pid1);
+	}
+
+	ASSERT_EQ(setns(self->child_pidfd1, flags), 0) {
+		TH_LOG("%m - Failed to setns to namespaces of %d",
+		       self->child_pid1);
+	}
+
+	pid = getpid();
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+		int nsfd;
+
+		if (self->child_nsfds1[i] < 0)
+			continue;
+
+		/* Verify that we have changed to the correct namespaces. */
+		if (info->flag == CLONE_NEWPID)
+			nsfd = self->nsfds[i];
+		else
+			nsfd = self->child_nsfds1[i];
+		ASSERT_EQ(in_same_namespace(nsfd, pid, info->name), 1) {
+			TH_LOG("setns failed to place us correctly into %s namespace of %d",
+			       info->name, self->child_pid1);
+		}
+		TH_LOG("Managed to correctly setns to %s namespace of %d",
+		       info->name, self->child_pid1);
+	}
+}
+
+TEST_F(current_nsset, no_foul_play)
+{
+	unsigned flags = 0;
+	int i;
+
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+
+		if (self->child_nsfds1[i] < 0)
+			continue;
+
+		flags |= info->flag;
+		if (info->flag) /* No use logging pid_for_children. */
+			TH_LOG("Adding %s namespace of %d to list of namespaces to attach to",
+			       info->name, self->child_pid1);
+	}
+
+	ASSERT_EQ(setns(self->child_pidfd1, flags), 0) {
+		TH_LOG("%m - Failed to setns to namespaces of %d vid pidfd %d",
+		       self->child_pid1, self->child_pidfd1);
+	}
+
+	/*
+	 * Can't setns to a user namespace outside of our hierarchy since we
+	 * don't have caps in there and didn't create it. That means that under
+	 * no circumstances should we be able to setns to any of the other
+	 * ones since they aren't owned by our user namespace.
+	 */
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+
+		if (self->child_nsfds2[i] < 0 || !info->flag)
+			continue;
+
+		ASSERT_NE(setns(self->child_pidfd2, info->flag), 0) {
+			TH_LOG("Managed to setns to %s namespace of %d via pidfd %d",
+			       info->name, self->child_pid2,
+			       self->child_pidfd2);
+		}
+		TH_LOG("%m - Correctly failed to setns to %s namespace of %d via pidfd %d",
+		       info->name, self->child_pid2,
+		       self->child_pidfd2);
+
+		ASSERT_NE(setns(self->child_nsfds2[i], info->flag), 0) {
+			TH_LOG("Managed to setns to %s namespace of %d via nsfd %d",
+			       info->name, self->child_pid2,
+			       self->child_nsfds2[i]);
+		}
+		TH_LOG("%m - Correctly failed to setns to %s namespace of %d via nsfd %d",
+		       info->name, self->child_pid2,
+		       self->child_nsfds2[i]);
+	}
+}
+
+TEST_HARNESS_MAIN