[PATCH] namespaces: incorporate fs namespace into nsproxy

This moves the mount namespace into the nsproxy.  The mount namespace count
now refers to the number of nsproxies point to it, rather than the number of
tasks.  As a result, the unshare_namespace() function in kernel/fork.c no
longer checks whether it is being shared.

Signed-off-by: Serge Hallyn <serue@us.ibm.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Andrey Savochkin <saw@sw.ru>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

fs/namespace.c

@@ -133,7 +133,7 @@ struct vfsmount *lookup_mnt(struct vfsmount *mnt, struct dentry *dentry)
 
 static inline int check_mnt(struct vfsmount *mnt)
 {
-	return mnt->mnt_namespace == current->namespace;
+	return mnt->mnt_namespace == current->nsproxy->namespace;
 }
 
 static void touch_namespace(struct namespace *ns)
@@ -830,7 +830,7 @@ static int attach_recursive_mnt(struct vfsmount *source_mnt,
 	if (parent_nd) {
 		detach_mnt(source_mnt, parent_nd);
 		attach_mnt(source_mnt, nd);
-		touch_namespace(current->namespace);
+		touch_namespace(current->nsproxy->namespace);
 	} else {
 		mnt_set_mountpoint(dest_mnt, dest_dentry, source_mnt);
 		commit_tree(source_mnt);
@@ -1441,7 +1441,7 @@ dput_out:
  */
 struct namespace *dup_namespace(struct task_struct *tsk, struct fs_struct *fs)
 {
-	struct namespace *namespace = tsk->namespace;
+	struct namespace *namespace = tsk->nsproxy->namespace;
 	struct namespace *new_ns;
 	struct vfsmount *rootmnt = NULL, *pwdmnt = NULL, *altrootmnt = NULL;
 	struct vfsmount *p, *q;
@@ -1508,7 +1508,7 @@ struct namespace *dup_namespace(struct task_struct *tsk, struct fs_struct *fs)
 
 int copy_namespace(int flags, struct task_struct *tsk)
 {
-	struct namespace *namespace = tsk->namespace;
+	struct namespace *namespace = tsk->nsproxy->namespace;
 	struct namespace *new_ns;
 	int err = 0;
 
@@ -1531,7 +1531,7 @@ int copy_namespace(int flags, struct task_struct *tsk)
 		goto out;
 	}
 
-	tsk->namespace = new_ns;
+	tsk->nsproxy->namespace = new_ns;
 
 out:
 	put_namespace(namespace);
@@ -1754,7 +1754,7 @@ asmlinkage long sys_pivot_root(const char __user * new_root,
 	detach_mnt(user_nd.mnt, &root_parent);
 	attach_mnt(user_nd.mnt, &old_nd);     /* mount old root on put_old */
 	attach_mnt(new_nd.mnt, &root_parent); /* mount new_root on / */
-	touch_namespace(current->namespace);
+	touch_namespace(current->nsproxy->namespace);
 	spin_unlock(&vfsmount_lock);
 	chroot_fs_refs(&user_nd, &new_nd);
 	security_sb_post_pivotroot(&user_nd, &new_nd);
@@ -1780,7 +1780,6 @@ static void __init init_mount_tree(void)
 {
 	struct vfsmount *mnt;
 	struct namespace *namespace;
-	struct task_struct *g, *p;
 
 	mnt = do_kern_mount("rootfs", 0, "rootfs", NULL);
 	if (IS_ERR(mnt))
@@ -1796,13 +1795,8 @@ static void __init init_mount_tree(void)
 	namespace->root = mnt;
 	mnt->mnt_namespace = namespace;
 
-	init_task.namespace = namespace;
+	init_task.nsproxy->namespace = namespace;
-	read_lock(&tasklist_lock);
+	get_namespace(namespace);
-	do_each_thread(g, p) {
-		get_namespace(namespace);
-		p->namespace = namespace;
-	} while_each_thread(g, p);
-	read_unlock(&tasklist_lock);
 
 	set_fs_pwd(current->fs, namespace->root, namespace->root->mnt_root);
 	set_fs_root(current->fs, namespace->root, namespace->root->mnt_root);

fs/proc/base.c

@@ -71,6 +71,7 @@
 #include <linux/cpuset.h>
 #include <linux/audit.h>
 #include <linux/poll.h>
+#include <linux/nsproxy.h>
 #include "internal.h"
 
 /* NOTE:
@@ -473,7 +474,7 @@ static int mounts_open(struct inode *inode, struct file *file)
 
 	if (task) {
 		task_lock(task);
-		namespace = task->namespace;
+		namespace = task->nsproxy->namespace;
 		if (namespace)
 			get_namespace(namespace);
 		task_unlock(task);
@@ -544,7 +545,7 @@ static int mountstats_open(struct inode *inode, struct file *file)
 
 		if (task) {
 			task_lock(task);
-			namespace = task->namespace;
+			namespace = task->nsproxy->namespace;
 			if (namespace)
 				get_namespace(namespace);
 			task_unlock(task);

include/linux/init_task.h

@@ -72,6 +72,7 @@ extern struct nsproxy init_nsproxy;
 #define INIT_NSPROXY(nsproxy) {						\
 	.count		= ATOMIC_INIT(1),				\
 	.nslock		= SPIN_LOCK_UNLOCKED,				\
+	.namespace	= NULL,						\
 }
 
 #define INIT_SIGHAND(sighand) {						\

include/linux/namespace.h

@@ -4,6 +4,7 @@
 
 #include <linux/mount.h>
 #include <linux/sched.h>
+#include <linux/nsproxy.h>
 
 struct namespace {
 	atomic_t		count;
@@ -26,11 +27,8 @@ static inline void put_namespace(struct namespace *namespace)
 
 static inline void exit_namespace(struct task_struct *p)
 {
-	struct namespace *namespace = p->namespace;
+	struct namespace *namespace = p->nsproxy->namespace;
 	if (namespace) {
-		task_lock(p);
-		p->namespace = NULL;
-		task_unlock(p);
 		put_namespace(namespace);
 	}
 }

include/linux/nsproxy.h

@@ -4,6 +4,8 @@
 #include <linux/spinlock.h>
 #include <linux/sched.h>
 
+struct namespace;
+
 /*
  * A structure to contain pointers to all per-process
  * namespaces - fs (mount), uts, network, sysvipc, etc.
@@ -19,6 +21,7 @@
 struct nsproxy {
 	atomic_t count;
 	spinlock_t nslock;
+	struct namespace *namespace;
 };
 extern struct nsproxy init_nsproxy;
 

include/linux/sched.h

@@ -238,7 +238,6 @@ extern signed long schedule_timeout_interruptible(signed long timeout);
 extern signed long schedule_timeout_uninterruptible(signed long timeout);
 asmlinkage void schedule(void);
 
-struct namespace;
 struct nsproxy;
 
 /* Maximum number of active map areas.. This is a random (large) number */
@@ -897,8 +896,7 @@ struct task_struct {
 	struct fs_struct *fs;
 /* open file information */
 	struct files_struct *files;
-/* namespace */
+/* namespaces */
-	struct namespace *namespace;
 	struct nsproxy *nsproxy;
 /* signal handlers */
 	struct signal_struct *signal;

kernel/exit.c

@@ -399,11 +399,8 @@ void daemonize(const char *name, ...)
 	current->fs = fs;
 	atomic_inc(&fs->count);
 
-	exit_namespace(current);
 	exit_task_namespaces(current);
-	current->namespace = init_task.namespace;
 	current->nsproxy = init_task.nsproxy;
-	get_namespace(current->namespace);
 	get_task_namespaces(current);
 
  	exit_files(current);
@@ -923,7 +920,6 @@ fastcall NORET_TYPE void do_exit(long code)
 	exit_sem(tsk);
 	__exit_files(tsk);
 	__exit_fs(tsk);
-	exit_namespace(tsk);
 	exit_task_namespaces(tsk);
 	exit_thread();
 	cpuset_exit(tsk);

kernel/fork.c

@@ -1119,11 +1119,9 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 		goto bad_fork_cleanup_mm;
 	if ((retval = copy_namespaces(clone_flags, p)))
 		goto bad_fork_cleanup_keys;
-	if ((retval = copy_namespace(clone_flags, p)))
-		goto bad_fork_cleanup_namespaces;
 	retval = copy_thread(0, clone_flags, stack_start, stack_size, p, regs);
 	if (retval)
-		goto bad_fork_cleanup_namespace;
+		goto bad_fork_cleanup_namespaces;
 
 	p->set_child_tid = (clone_flags & CLONE_CHILD_SETTID) ? child_tidptr : NULL;
 	/*
@@ -1215,7 +1213,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 		spin_unlock(&current->sighand->siglock);
 		write_unlock_irq(&tasklist_lock);
 		retval = -ERESTARTNOINTR;
-		goto bad_fork_cleanup_namespace;
+		goto bad_fork_cleanup_namespaces;
 	}
 
 	if (clone_flags & CLONE_THREAD) {
@@ -1263,8 +1261,6 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 	proc_fork_connector(p);
 	return p;
 
-bad_fork_cleanup_namespace:
-	exit_namespace(p);
 bad_fork_cleanup_namespaces:
 	exit_task_namespaces(p);
 bad_fork_cleanup_keys:
@@ -1519,10 +1515,9 @@ static int unshare_fs(unsigned long unshare_flags, struct fs_struct **new_fsp)
  */
 static int unshare_namespace(unsigned long unshare_flags, struct namespace **new_nsp, struct fs_struct *new_fs)
 {
-	struct namespace *ns = current->namespace;
+	struct namespace *ns = current->nsproxy->namespace;
 
-	if ((unshare_flags & CLONE_NEWNS) &&
+	if ((unshare_flags & CLONE_NEWNS) && ns) {
-	    (ns && atomic_read(&ns->count) > 1)) {
 		if (!capable(CAP_SYS_ADMIN))
 			return -EPERM;
 
@@ -1655,8 +1650,8 @@ asmlinkage long sys_unshare(unsigned long unshare_flags)
 		}
 
 		if (new_ns) {
-			ns = current->namespace;
+			ns = current->nsproxy->namespace;
-			current->namespace = new_ns;
+			current->nsproxy->namespace = new_ns;
 			new_ns = ns;
 		}
 

kernel/nsproxy.c

@@ -13,6 +13,7 @@
 #include <linux/version.h>
 #include <linux/nsproxy.h>
 #include <linux/init_task.h>
+#include <linux/namespace.h>
 
 struct nsproxy init_nsproxy = INIT_NSPROXY(init_nsproxy);
 
@@ -55,6 +56,11 @@ struct nsproxy *dup_namespaces(struct nsproxy *orig)
 {
 	struct nsproxy *ns = clone_namespaces(orig);
 
+	if (ns) {
+		if (ns->namespace)
+			get_namespace(ns->namespace);
+	}
+
 	return ns;
 }
 
@@ -65,16 +71,40 @@ struct nsproxy *dup_namespaces(struct nsproxy *orig)
 int copy_namespaces(int flags, struct task_struct *tsk)
 {
 	struct nsproxy *old_ns = tsk->nsproxy;
+	struct nsproxy *new_ns;
+	int err = 0;
 
 	if (!old_ns)
 		return 0;
 
 	get_nsproxy(old_ns);
 
-	return 0;
+	if (!(flags & CLONE_NEWNS))
+		return 0;
+
+	new_ns = clone_namespaces(old_ns);
+	if (!new_ns) {
+		err = -ENOMEM;
+		goto out;
+	}
+
+	tsk->nsproxy = new_ns;
+
+	err = copy_namespace(flags, tsk);
+	if (err) {
+		tsk->nsproxy = old_ns;
+		put_nsproxy(new_ns);
+		goto out;
+	}
+
+out:
+	put_nsproxy(old_ns);
+	return err;
 }
 
 void free_nsproxy(struct nsproxy *ns)
 {
+		if (ns->namespace)
+			put_namespace(ns->namespace);
 		kfree(ns);
 }