cgroup.c

/*
 *  Generic process-grouping system.
 *
 *  Based originally on the cpuset system, extracted by Paul Menage
 *  Copyright (C) 2006 Google, Inc
 *
 *  Notifications support
 *  Copyright (C) 2009 Nokia Corporation
 *  Author: Kirill A. Shutemov
 *
 *  Copyright notices from the original cpuset code:
 *  --------------------------------------------------
 *  Copyright (C) 2003 BULL SA.
 *  Copyright (C) 2004-2006 Silicon Graphics, Inc.
 *
 *  Portions derived from Patrick Mochel's sysfs code.
 *  sysfs is Copyright (c) 2001-3 Patrick Mochel
 *
 *  2003-10-10 Written by Simon Derr.
 *  2003-10-22 Updates by Stephen Hemminger.
 *  2004 May-July Rework by Paul Jackson.
 *  ---------------------------------------------------
 *
 *  This file is subject to the terms and conditions of the GNU General Public
 *  License.  See the file COPYING in the main directory of the Linux
 *  distribution for more details.
 */

#include <linux/cgroup.h>
#include <linux/cred.h>
#include <linux/ctype.h>
#include <linux/errno.h>
#include <linux/init_task.h>
#include <linux/kernel.h>
#include <linux/list.h>
#include <linux/mm.h>
#include <linux/mutex.h>
#include <linux/mount.h>
#include <linux/pagemap.h>
#include <linux/proc_fs.h>
#include <linux/rcupdate.h>
#include <linux/sched.h>
#include <linux/backing-dev.h>
#include <linux/slab.h>
#include <linux/magic.h>
#include <linux/spinlock.h>
#include <linux/string.h>
#include <linux/sort.h>
#include <linux/kmod.h>
#include <linux/delayacct.h>
#include <linux/cgroupstats.h>
#include <linux/hashtable.h>
#include <linux/namei.h>
#include <linux/pid_namespace.h>
#include <linux/idr.h>
#include <linux/vmalloc.h> /* TODO: replace with more sophisticated array */
#include <linux/flex_array.h> /* used in cgroup_attach_task */
#include <linux/kthread.h>

#include <linux/atomic.h>

/*
 * pidlists linger the following amount before being destroyed.  The goal
 * is avoiding frequent destruction in the middle of consecutive read calls
 * Expiring in the middle is a performance problem not a correctness one.
 * 1 sec should be enough.
 */
#define CGROUP_PIDLIST_DESTROY_DELAY	HZ

/*
 * cgroup_mutex is the master lock.  Any modification to cgroup or its
 * hierarchy must be performed while holding it.
 *
 * cgroup_root_mutex nests inside cgroup_mutex and should be held to modify
 * cgroupfs_root of any cgroup hierarchy - subsys list, flags,
 * release_agent_path and so on.  Modifying requires both cgroup_mutex and
 * cgroup_root_mutex.  Readers can acquire either of the two.  This is to
 * break the following locking order cycle.
 *
 *  A. cgroup_mutex -> cred_guard_mutex -> s_type->i_mutex_key -> namespace_sem
 *  B. namespace_sem -> cgroup_mutex
 *
 * B happens only through cgroup_show_options() and using cgroup_root_mutex
 * breaks it.
 */
#ifdef CONFIG_PROVE_RCU
DEFINE_MUTEX(cgroup_mutex);
EXPORT_SYMBOL_GPL(cgroup_mutex);	/* only for lockdep */
#else
static DEFINE_MUTEX(cgroup_mutex);
#endif

static DEFINE_MUTEX(cgroup_root_mutex);

#define cgroup_assert_mutex_or_rcu_locked()				\
	rcu_lockdep_assert(rcu_read_lock_held() ||			\
			   lockdep_is_held(&cgroup_mutex),		\
			   "cgroup_mutex or RCU read lock required");

#ifdef CONFIG_LOCKDEP
#define cgroup_assert_mutex_or_root_locked()				\
	WARN_ON_ONCE(debug_locks && (!lockdep_is_held(&cgroup_mutex) &&	\
				     !lockdep_is_held(&cgroup_root_mutex)))
#else
#define cgroup_assert_mutex_or_root_locked()	do { } while (0)
#endif

/*
 * cgroup destruction makes heavy use of work items and there can be a lot
 * of concurrent destructions.  Use a separate workqueue so that cgroup
 * destruction work items don't end up filling up max_active of system_wq
 * which may lead to deadlock.
 */
static struct workqueue_struct *cgroup_destroy_wq;

/*
 * pidlist destructions need to be flushed on cgroup destruction.  Use a
 * separate workqueue as flush domain.
 */
static struct workqueue_struct *cgroup_pidlist_destroy_wq;

/* generate an array of cgroup subsystem pointers */
#define SUBSYS(_x) [_x ## _cgrp_id] = &_x ## _cgrp_subsys,
static struct cgroup_subsys *cgroup_subsys[] = {
#include <linux/cgroup_subsys.h>
};
#undef SUBSYS

/* array of cgroup subsystem names */
#define SUBSYS(_x) [_x ## _cgrp_id] = #_x,
static const char *cgroup_subsys_name[] = {
#include <linux/cgroup_subsys.h>
};
#undef SUBSYS

/*
 * The dummy hierarchy, reserved for the subsystems that are otherwise
 * unattached - it never has more than a single cgroup, and all tasks are
 * part of that cgroup.
 */
static struct cgroupfs_root cgroup_dummy_root;

/* dummy_top is a shorthand for the dummy hierarchy's top cgroup */
static struct cgroup * const cgroup_dummy_top = &cgroup_dummy_root.top_cgroup;

/* The list of hierarchy roots */

static LIST_HEAD(cgroup_roots);
static int cgroup_root_count;

/*
 * Hierarchy ID allocation and mapping.  It follows the same exclusion
 * rules as other root ops - both cgroup_mutex and cgroup_root_mutex for
 * writes, either for reads.
 */
static DEFINE_IDR(cgroup_hierarchy_idr);

static struct cgroup_name root_cgroup_name = { .name = "/" };

/*
 * Assign a monotonically increasing serial number to cgroups.  It
 * guarantees cgroups with bigger numbers are newer than those with smaller
 * numbers.  Also, as cgroups are always appended to the parent's
 * ->children list, it guarantees that sibling cgroups are always sorted in
 * the ascending serial number order on the list.  Protected by
 * cgroup_mutex.
 */
static u64 cgroup_serial_nr_next = 1;

/* This flag indicates whether tasks in the fork and exit paths should
 * check for fork/exit handlers to call. This avoids us having to do
 * extra work in the fork/exit path if none of the subsystems need to
 * be called.
 */
static int need_forkexit_callback __read_mostly;

static struct cftype cgroup_base_files[];

static void cgroup_destroy_css_killed(struct cgroup *cgrp);
static int cgroup_destroy_locked(struct cgroup *cgrp);
static int cgroup_addrm_files(struct cgroup *cgrp, struct cftype cfts[],
			      bool is_add);
static int cgroup_file_release(struct inode *inode, struct file *file);
static void cgroup_pidlist_destroy_all(struct cgroup *cgrp);

/**
 * cgroup_css - obtain a cgroup's css for the specified subsystem
 * @cgrp: the cgroup of interest
 * @ss: the subsystem of interest (%NULL returns the dummy_css)
 *
 * Return @cgrp's css (cgroup_subsys_state) associated with @ss.  This
 * function must be called either under cgroup_mutex or rcu_read_lock() and
 * the caller is responsible for pinning the returned css if it wants to
 * keep accessing it outside the said locks.  This function may return
 * %NULL if @cgrp doesn't have @subsys_id enabled.
 */
static struct cgroup_subsys_state *cgroup_css(struct cgroup *cgrp,
					      struct cgroup_subsys *ss)
{
	if (ss)
		return rcu_dereference_check(cgrp->subsys[ss->subsys_id],
					     lockdep_is_held(&cgroup_mutex));
	else
		return &cgrp->dummy_css;
}

/* convenient tests for these bits */
static inline bool cgroup_is_dead(const struct cgroup *cgrp)
{
	return test_bit(CGRP_DEAD, &cgrp->flags);
}

/**
 * cgroup_is_descendant - test ancestry
 * @cgrp: the cgroup to be tested
 * @ancestor: possible ancestor of @cgrp
 *
 * Test whether @cgrp is a descendant of @ancestor.  It also returns %true
 * if @cgrp == @ancestor.  This function is safe to call as long as @cgrp
 * and @ancestor are accessible.
 */
bool cgroup_is_descendant(struct cgroup *cgrp, struct cgroup *ancestor)
{
	while (cgrp) {
		if (cgrp == ancestor)
			return true;
		cgrp = cgrp->parent;
	}
	return false;
}
EXPORT_SYMBOL_GPL(cgroup_is_descendant);

static int cgroup_is_releasable(const struct cgroup *cgrp)
{
	const int bits =
		(1 << CGRP_RELEASABLE) |
		(1 << CGRP_NOTIFY_ON_RELEASE);
	return (cgrp->flags & bits) == bits;
}

static int notify_on_release(const struct cgroup *cgrp)
{
	return test_bit(CGRP_NOTIFY_ON_RELEASE, &cgrp->flags);
}

/**
 * for_each_css - iterate all css's of a cgroup
 * @css: the iteration cursor
 * @ssid: the index of the subsystem, CGROUP_SUBSYS_COUNT after reaching the end
 * @cgrp: the target cgroup to iterate css's of
 *
 * Should be called under cgroup_mutex.
 */
#define for_each_css(css, ssid, cgrp)					\
	for ((ssid) = 0; (ssid) < CGROUP_SUBSYS_COUNT; (ssid)++)	\
		if (!((css) = rcu_dereference_check(			\
				(cgrp)->subsys[(ssid)],			\
				lockdep_is_held(&cgroup_mutex)))) { }	\
		else

/**
 * for_each_subsys - iterate all enabled cgroup subsystems
 * @ss: the iteration cursor
 * @ssid: the index of @ss, CGROUP_SUBSYS_COUNT after reaching the end
 */
#define for_each_subsys(ss, ssid)					\
	for ((ssid) = 0; (ssid) < CGROUP_SUBSYS_COUNT &&		\
	     (((ss) = cgroup_subsys[ssid]) || true); (ssid)++)

/* iterate across the active hierarchies */
#define for_each_active_root(root)					\
	list_for_each_entry((root), &cgroup_roots, root_list)

static inline struct cgroup *__d_cgrp(struct dentry *dentry)
{
	return dentry->d_fsdata;
}

static inline struct cfent *__d_cfe(struct dentry *dentry)
{
	return dentry->d_fsdata;
}

static inline struct cftype *__d_cft(struct dentry *dentry)
{
	return __d_cfe(dentry)->type;
}

/**
 * cgroup_lock_live_group - take cgroup_mutex and check that cgrp is alive.
 * @cgrp: the cgroup to be checked for liveness
 *
 * On success, returns true; the mutex should be later unlocked.  On
 * failure returns false with no lock held.
 */
static bool cgroup_lock_live_group(struct cgroup *cgrp)
{
	mutex_lock(&cgroup_mutex);
	if (cgroup_is_dead(cgrp)) {
		mutex_unlock(&cgroup_mutex);
		return false;
	}
	return true;
}

/* the list of cgroups eligible for automatic release. Protected by
 * release_list_lock */
static LIST_HEAD(release_list);
static DEFINE_RAW_SPINLOCK(release_list_lock);
static void cgroup_release_agent(struct work_struct *work);
static DECLARE_WORK(release_agent_work, cgroup_release_agent);
static void check_for_release(struct cgroup *cgrp);

/*
 * A cgroup can be associated with multiple css_sets as different tasks may
 * belong to different cgroups on different hierarchies.  In the other
 * direction, a css_set is naturally associated with multiple cgroups.
 * This M:N relationship is represented by the following link structure
 * which exists for each association and allows traversing the associations
 * from both sides.
 */
struct cgrp_cset_link {
	/* the cgroup and css_set this link associates */
	struct cgroup		*cgrp;
	struct css_set		*cset;

	/* list of cgrp_cset_links anchored at cgrp->cset_links */
	struct list_head	cset_link;

	/* list of cgrp_cset_links anchored at css_set->cgrp_links */
	struct list_head	cgrp_link;
};

/* The default css_set - used by init and its children prior to any
 * hierarchies being mounted. It contains a pointer to the root state
 * for each subsystem. Also used to anchor the list of css_sets. Not
 * reference-counted, to improve performance when child cgroups
 * haven't been created.
 */

static struct css_set init_css_set;
static struct cgrp_cset_link init_cgrp_cset_link;

/*
 * css_set_lock protects the list of css_set objects, and the chain of
 * tasks off each css_set.  Nests outside task->alloc_lock due to
 * css_task_iter_start().
 */
static DEFINE_RWLOCK(css_set_lock);
static int css_set_count;

/*
 * hash table for cgroup groups. This improves the performance to find
 * an existing css_set. This hash doesn't (currently) take into
 * account cgroups in empty hierarchies.
 */
#define CSS_SET_HASH_BITS	7
static DEFINE_HASHTABLE(css_set_table, CSS_SET_HASH_BITS);

static unsigned long css_set_hash(struct cgroup_subsys_state *css[])
{
	unsigned long key = 0UL;
	struct cgroup_subsys *ss;
	int i;

	for_each_subsys(ss, i)
		key += (unsigned long)css[i];
	key = (key >> 16) ^ key;

	return key;
}

/*
 * We don't maintain the lists running through each css_set to its task
 * until after the first call to css_task_iter_start().  This reduces the
 * fork()/exit() overhead for people who have cgroups compiled into their
 * kernel but not actually in use.
 */
static int use_task_css_set_links __read_mostly;

static void __put_css_set(struct css_set *cset, int taskexit)
{
	struct cgrp_cset_link *link, *tmp_link;

	/*
	 * Ensure that the refcount doesn't hit zero while any readers
	 * can see it. Similar to atomic_dec_and_lock(), but for an
	 * rwlock
	 */
	if (atomic_add_unless(&cset->refcount, -1, 1))
		return;
	write_lock(&css_set_lock);
	if (!atomic_dec_and_test(&cset->refcount)) {
		write_unlock(&css_set_lock);
		return;
	}

	/* This css_set is dead. unlink it and release cgroup refcounts */
	hash_del(&cset->hlist);
	css_set_count--;

	list_for_each_entry_safe(link, tmp_link, &cset->cgrp_links, cgrp_link) {
		struct cgroup *cgrp = link->cgrp;

		list_del(&link->cset_link);
		list_del(&link->cgrp_link);

		/* @cgrp can't go away while we're holding css_set_lock */
		if (list_empty(&cgrp->cset_links) && notify_on_release(cgrp)) {
			if (taskexit)
				set_bit(CGRP_RELEASABLE, &cgrp->flags);
			check_for_release(cgrp);
		}

		kfree(link);
	}

	write_unlock(&css_set_lock);
	kfree_rcu(cset, rcu_head);
}

/*
 * refcounted get/put for css_set objects
 */
static inline void get_css_set(struct css_set *cset)
{
	atomic_inc(&cset->refcount);
}

static inline void put_css_set(struct css_set *cset)
{
	__put_css_set(cset, 0);
}

static inline void put_css_set_taskexit(struct css_set *cset)
{
	__put_css_set(cset, 1);
}

/**
 * compare_css_sets - helper function for find_existing_css_set().
 * @cset: candidate css_set being tested
 * @old_cset: existing css_set for a task
 * @new_cgrp: cgroup that's being entered by the task
 * @template: desired set of css pointers in css_set (pre-calculated)
 *
 * Returns true if "cset" matches "old_cset" except for the hierarchy
 * which "new_cgrp" belongs to, for which it should match "new_cgrp".
 */
static bool compare_css_sets(struct css_set *cset,
			     struct css_set *old_cset,
			     struct cgroup *new_cgrp,
			     struct cgroup_subsys_state *template[])
{
	struct list_head *l1, *l2;

	if (memcmp(template, cset->subsys, sizeof(cset->subsys))) {
		/* Not all subsystems matched */
		return false;
	}

	/*
	 * Compare cgroup pointers in order to distinguish between
	 * different cgroups in heirarchies with no subsystems. We
	 * could get by with just this check alone (and skip the
	 * memcmp above) but on most setups the memcmp check will
	 * avoid the need for this more expensive check on almost all
	 * candidates.
	 */

	l1 = &cset->cgrp_links;
	l2 = &old_cset->cgrp_links;
	while (1) {
		struct cgrp_cset_link *link1, *link2;
		struct cgroup *cgrp1, *cgrp2;

		l1 = l1->next;
		l2 = l2->next;
		/* See if we reached the end - both lists are equal length. */
		if (l1 == &cset->cgrp_links) {
			BUG_ON(l2 != &old_cset->cgrp_links);
			break;
		} else {
			BUG_ON(l2 == &old_cset->cgrp_links);
		}
		/* Locate the cgroups associated with these links. */
		link1 = list_entry(l1, struct cgrp_cset_link, cgrp_link);
		link2 = list_entry(l2, struct cgrp_cset_link, cgrp_link);
		cgrp1 = link1->cgrp;
		cgrp2 = link2->cgrp;
		/* Hierarchies should be linked in the same order. */
		BUG_ON(cgrp1->root != cgrp2->root);

		/*
		 * If this hierarchy is the hierarchy of the cgroup
		 * that's changing, then we need to check that this
		 * css_set points to the new cgroup; if it's any other
		 * hierarchy, then this css_set should point to the
		 * same cgroup as the old css_set.
		 */
		if (cgrp1->root == new_cgrp->root) {
			if (cgrp1 != new_cgrp)
				return false;
		} else {
			if (cgrp1 != cgrp2)
				return false;
		}
	}
	return true;
}

/**
 * find_existing_css_set - init css array and find the matching css_set
 * @old_cset: the css_set that we're using before the cgroup transition
 * @cgrp: the cgroup that we're moving into
 * @template: out param for the new set of csses, should be clear on entry
 */
static struct css_set *find_existing_css_set(struct css_set *old_cset,
					struct cgroup *cgrp,
					struct cgroup_subsys_state *template[])
{
	struct cgroupfs_root *root = cgrp->root;
	struct cgroup_subsys *ss;
	struct css_set *cset;
	unsigned long key;
	int i;

	/*
	 * Build the set of subsystem state objects that we want to see in the
	 * new css_set. while subsystems can change globally, the entries here
	 * won't change, so no need for locking.
	 */
	for_each_subsys(ss, i) {
		if (root->subsys_mask & (1UL << i)) {
			/* Subsystem is in this hierarchy. So we want
			 * the subsystem state from the new
			 * cgroup */
			template[i] = cgroup_css(cgrp, ss);
		} else {
			/* Subsystem is not in this hierarchy, so we
			 * don't want to change the subsystem state */
			template[i] = old_cset->subsys[i];
		}
	}

	key = css_set_hash(template);
	hash_for_each_possible(css_set_table, cset, hlist, key) {
		if (!compare_css_sets(cset, old_cset, cgrp, template))
			continue;

		/* This css_set matches what we need */
		return cset;
	}

	/* No existing cgroup group matched */
	return NULL;
}

static void free_cgrp_cset_links(struct list_head *links_to_free)
{
	struct cgrp_cset_link *link, *tmp_link;

	list_for_each_entry_safe(link, tmp_link, links_to_free, cset_link) {
		list_del(&link->cset_link);
		kfree(link);
	}
}

/**
 * allocate_cgrp_cset_links - allocate cgrp_cset_links
 * @count: the number of links to allocate
 * @tmp_links: list_head the allocated links are put on
 *
 * Allocate @count cgrp_cset_link structures and chain them on @tmp_links
 * through ->cset_link.  Returns 0 on success or -errno.
 */
static int allocate_cgrp_cset_links(int count, struct list_head *tmp_links)
{
	struct cgrp_cset_link *link;
	int i;

	INIT_LIST_HEAD(tmp_links);

	for (i = 0; i < count; i++) {
		link = kzalloc(sizeof(*link), GFP_KERNEL);
		if (!link) {
			free_cgrp_cset_links(tmp_links);
			return -ENOMEM;
		}
		list_add(&link->cset_link, tmp_links);
	}
	return 0;
}

/**
 * link_css_set - a helper function to link a css_set to a cgroup
 * @tmp_links: cgrp_cset_link objects allocated by allocate_cgrp_cset_links()
 * @cset: the css_set to be linked
 * @cgrp: the destination cgroup
 */
static void link_css_set(struct list_head *tmp_links, struct css_set *cset,
			 struct cgroup *cgrp)
{
	struct cgrp_cset_link *link;

	BUG_ON(list_empty(tmp_links));
	link = list_first_entry(tmp_links, struct cgrp_cset_link, cset_link);
	link->cset = cset;
	link->cgrp = cgrp;
	list_move(&link->cset_link, &cgrp->cset_links);
	/*
	 * Always add links to the tail of the list so that the list
	 * is sorted by order of hierarchy creation
	 */
	list_add_tail(&link->cgrp_link, &cset->cgrp_links);
}

/**
 * find_css_set - return a new css_set with one cgroup updated
 * @old_cset: the baseline css_set
 * @cgrp: the cgroup to be updated
 *
 * Return a new css_set that's equivalent to @old_cset, but with @cgrp
 * substituted into the appropriate hierarchy.
 */
static struct css_set *find_css_set(struct css_set *old_cset,
				    struct cgroup *cgrp)
{
	struct cgroup_subsys_state *template[CGROUP_SUBSYS_COUNT] = { };
	struct css_set *cset;
	struct list_head tmp_links;
	struct cgrp_cset_link *link;
	unsigned long key;

	lockdep_assert_held(&cgroup_mutex);

	/* First see if we already have a cgroup group that matches
	 * the desired set */
	read_lock(&css_set_lock);
	cset = find_existing_css_set(old_cset, cgrp, template);
	if (cset)
		get_css_set(cset);
	read_unlock(&css_set_lock);

	if (cset)
		return cset;

	cset = kzalloc(sizeof(*cset), GFP_KERNEL);
	if (!cset)
		return NULL;

	/* Allocate all the cgrp_cset_link objects that we'll need */
	if (allocate_cgrp_cset_links(cgroup_root_count, &tmp_links) < 0) {
		kfree(cset);
		return NULL;
	}

	atomic_set(&cset->refcount, 1);
	INIT_LIST_HEAD(&cset->cgrp_links);
	INIT_LIST_HEAD(&cset->tasks);
	INIT_HLIST_NODE(&cset->hlist);

	/* Copy the set of subsystem state objects generated in
	 * find_existing_css_set() */
	memcpy(cset->subsys, template, sizeof(cset->subsys));

	write_lock(&css_set_lock);
	/* Add reference counts and links from the new css_set. */
	list_for_each_entry(link, &old_cset->cgrp_links, cgrp_link) {
		struct cgroup *c = link->cgrp;

		if (c->root == cgrp->root)
			c = cgrp;
		link_css_set(&tmp_links, cset, c);
	}

	BUG_ON(!list_empty(&tmp_links));

	css_set_count++;

	/* Add this cgroup group to the hash table */
	key = css_set_hash(cset->subsys);
	hash_add(css_set_table, &cset->hlist, key);

	write_unlock(&css_set_lock);

	return cset;
}

/*
 * Return the cgroup for "task" from the given hierarchy. Must be
 * called with cgroup_mutex held.
 */
static struct cgroup *task_cgroup_from_root(struct task_struct *task,
					    struct cgroupfs_root *root)
{
	struct css_set *cset;
	struct cgroup *res = NULL;

	BUG_ON(!mutex_is_locked(&cgroup_mutex));
	read_lock(&css_set_lock);
	/*
	 * No need to lock the task - since we hold cgroup_mutex the
	 * task can't change groups, so the only thing that can happen
	 * is that it exits and its css is set back to init_css_set.
	 */
	cset = task_css_set(task);
	if (cset == &init_css_set) {
		res = &root->top_cgroup;
	} else {
		struct cgrp_cset_link *link;

		list_for_each_entry(link, &cset->cgrp_links, cgrp_link) {
			struct cgroup *c = link->cgrp;

			if (c->root == root) {
				res = c;
				break;
			}
		}
	}
	read_unlock(&css_set_lock);
	BUG_ON(!res);
	return res;
}

/*
 * There is one global cgroup mutex. We also require taking
 * task_lock() when dereferencing a task's cgroup subsys pointers.
 * See "The task_lock() exception", at the end of this comment.
 *
 * A task must hold cgroup_mutex to modify cgroups.
 *
 * Any task can increment and decrement the count field without lock.
 * So in general, code holding cgroup_mutex can't rely on the count
 * field not changing.  However, if the count goes to zero, then only
 * cgroup_attach_task() can increment it again.  Because a count of zero
 * means that no tasks are currently attached, therefore there is no
 * way a task attached to that cgroup can fork (the other way to
 * increment the count).  So code holding cgroup_mutex can safely
 * assume that if the count is zero, it will stay zero. Similarly, if
 * a task holds cgroup_mutex on a cgroup with zero count, it
 * knows that the cgroup won't be removed, as cgroup_rmdir()
 * needs that mutex.
 *
 * The fork and exit callbacks cgroup_fork() and cgroup_exit(), don't
 * (usually) take cgroup_mutex.  These are the two most performance
 * critical pieces of code here.  The exception occurs on cgroup_exit(),
 * when a task in a notify_on_release cgroup exits.  Then cgroup_mutex
 * is taken, and if the cgroup count is zero, a usermode call made
 * to the release agent with the name of the cgroup (path relative to
 * the root of cgroup file system) as the argument.
 *
 * A cgroup can only be deleted if both its 'count' of using tasks
 * is zero, and its list of 'children' cgroups is empty.  Since all
 * tasks in the system use _some_ cgroup, and since there is always at
 * least one task in the system (init, pid == 1), therefore, top_cgroup
 * always has either children cgroups and/or using tasks.  So we don't
 * need a special hack to ensure that top_cgroup cannot be deleted.
 *
 *	The task_lock() exception
 *
 * The need for this exception arises from the action of
 * cgroup_attach_task(), which overwrites one task's cgroup pointer with
 * another.  It does so using cgroup_mutex, however there are
 * several performance critical places that need to reference
 * task->cgroup without the expense of grabbing a system global
 * mutex.  Therefore except as noted below, when dereferencing or, as
 * in cgroup_attach_task(), modifying a task's cgroup pointer we use
 * task_lock(), which acts on a spinlock (task->alloc_lock) already in
 * the task_struct routinely used for such matters.
 *
 * P.S.  One more locking exception.  RCU is used to guard the
 * update of a tasks cgroup pointer by cgroup_attach_task()
 */

/*
 * A couple of forward declarations required, due to cyclic reference loop:
 * cgroup_mkdir -> cgroup_create -> cgroup_populate_dir ->
 * cgroup_add_file -> cgroup_create_file -> cgroup_dir_inode_operations
 * -> cgroup_mkdir.
 */

static int cgroup_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode);
static int cgroup_rmdir(struct inode *unused_dir, struct dentry *dentry);
static int cgroup_populate_dir(struct cgroup *cgrp, unsigned long subsys_mask);
static const struct inode_operations cgroup_dir_inode_operations;
static const struct file_operations proc_cgroupstats_operations;

static struct backing_dev_info cgroup_backing_dev_info = {
	.name		= "cgroup",
	.capabilities	= BDI_CAP_NO_ACCT_AND_WRITEBACK,
};

static struct inode *cgroup_new_inode(umode_t mode, struct super_block *sb)
{
	struct inode *inode = new_inode(sb);

	if (inode) {
		inode->i_ino = get_next_ino();
		inode->i_mode = mode;
		inode->i_uid = current_fsuid();
		inode->i_gid = current_fsgid();
		inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
		inode->i_mapping->backing_dev_info = &cgroup_backing_dev_info;
	}
	return inode;
}

static struct cgroup_name *cgroup_alloc_name(struct dentry *dentry)
{
	struct cgroup_name *name;

	name = kmalloc(sizeof(*name) + dentry->d_name.len + 1, GFP_KERNEL);
	if (!name)
		return NULL;
	strcpy(name->name, dentry->d_name.name);
	return name;
}

static void cgroup_free_fn(struct work_struct *work)
{
	struct cgroup *cgrp = container_of(work, struct cgroup, destroy_work);

	mutex_lock(&cgroup_mutex);
	cgrp->root->number_of_cgroups--;
	mutex_unlock(&cgroup_mutex);

	/*
	 * We get a ref to the parent's dentry, and put the ref when
	 * this cgroup is being freed, so it's guaranteed that the
	 * parent won't be destroyed before its children.
	 */
	dput(cgrp->parent->dentry);

	/*
	 * Drop the active superblock reference that we took when we
	 * created the cgroup. This will free cgrp->root, if we are
	 * holding the last reference to @sb.
	 */
	deactivate_super(cgrp->root->sb);

	cgroup_pidlist_destroy_all(cgrp);

	simple_xattrs_free(&cgrp->xattrs);

	kfree(rcu_dereference_raw(cgrp->name));
	kfree(cgrp);
}

static void cgroup_free_rcu(struct rcu_head *head)
{
	struct cgroup *cgrp = container_of(head, struct cgroup, rcu_head);

	INIT_WORK(&cgrp->destroy_work, cgroup_free_fn);
	queue_work(cgroup_destroy_wq, &cgrp->destroy_work);
}

static void cgroup_diput(struct dentry *dentry, struct inode *inode)
{
	/* is dentry a directory ? if so, kfree() associated cgroup */
	if (S_ISDIR(inode->i_mode)) {
		struct cgroup *cgrp = dentry->d_fsdata;

		BUG_ON(!(cgroup_is_dead(cgrp)));

		/*
		 * XXX: cgrp->id is only used to look up css's.  As cgroup
		 * and css's lifetimes will be decoupled, it should be made
		 * per-subsystem and moved to css->id so that lookups are
		 * successful until the target css is released.
		 */
		idr_remove(&cgrp->root->cgroup_idr, cgrp->id);
		cgrp->id = -1;

		call_rcu(&cgrp->rcu_head, cgroup_free_rcu);
	} else {
		struct cfent *cfe = __d_cfe(dentry);
		struct cgroup *cgrp = dentry->d_parent->d_fsdata;

		WARN_ONCE(!list_empty(&cfe->node) &&
			  cgrp != &cgrp->root->top_cgroup,
			  "cfe still linked for %s\n", cfe->type->name);
		simple_xattrs_free(&cfe->xattrs);
		kfree(cfe);
	}
	iput(inode);
}

static void remove_dir(struct dentry *d)
{
	struct dentry *parent = dget(d->d_parent);

	d_delete(d);
	simple_rmdir(parent->d_inode, d);
	dput(parent);
}

static void cgroup_rm_file(struct cgroup *cgrp, const struct cftype *cft)
{
	struct cfent *cfe;

	lockdep_assert_held(&cgrp->dentry->d_inode->i_mutex);
	lockdep_assert_held(&cgroup_mutex);

	/*
	 * If we're doing cleanup due to failure of cgroup_create(),
	 * the corresponding @cfe may not exist.
	 */
	list_for_each_entry(cfe, &cgrp->files, node) {
		struct dentry *d = cfe->dentry;

		if (cft && cfe->type != cft)
			continue;

		dget(d);
		d_delete(d);
		simple_unlink(cgrp->dentry->d_inode, d);
		list_del_init(&cfe->node);
		dput(d);

		break;
	}
}

/**
 * cgroup_clear_dir - remove subsys files in a cgroup directory
 * @cgrp: target cgroup
 * @subsys_mask: mask of the subsystem ids whose files should be removed
 */
static void cgroup_clear_dir(struct cgroup *cgrp, unsigned long subsys_mask)
{
	struct cgroup_subsys *ss;
	int i;

	for_each_subsys(ss, i) {
		struct cftype_set *set;

		if (!test_bit(i, &subsys_mask))
			continue;
		list_for_each_entry(set, &ss->cftsets, node)
			cgroup_addrm_files(cgrp, set->cfts, false);
	}
}

/*
 * NOTE : the dentry must have been dget()'ed
 */
static void cgroup_d_remove_dir(struct dentry *dentry)
{
	struct dentry *parent;

	parent = dentry->d_parent;
	spin_lock(&parent->d_lock);
	spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED);
	list_del_init(&dentry->d_u.d_child);
	spin_unlock(&dentry->d_lock);
	spin_unlock(&parent->d_lock);
	remove_dir(dentry);
}

static int rebind_subsystems(struct cgroupfs_root *root,
			     unsigned long added_mask, unsigned removed_mask)
{
	struct cgroup *cgrp = &root->top_cgroup;
	struct cgroup_subsys *ss;
	int i, ret;

	BUG_ON(!mutex_is_locked(&cgroup_mutex));
	BUG_ON(!mutex_is_locked(&cgroup_root_mutex));

	/* Check that any added subsystems are currently free */
	for_each_subsys(ss, i)
		if ((added_mask & (1 << i)) && ss->root != &cgroup_dummy_root)
			return -EBUSY;

	ret = cgroup_populate_dir(cgrp, added_mask);
	if (ret)
		return ret;

	/*
	 * Nothing can fail from this point on.  Remove files for the
	 * removed subsystems and rebind each subsystem.
	 */
	cgroup_clear_dir(cgrp, removed_mask);

	for_each_subsys(ss, i) {
		unsigned long bit = 1UL << i;

		if (bit & added_mask) {
			/* We're binding this subsystem to this hierarchy */
			BUG_ON(cgroup_css(cgrp, ss));
			BUG_ON(!cgroup_css(cgroup_dummy_top, ss));
			BUG_ON(cgroup_css(cgroup_dummy_top, ss)->cgroup != cgroup_dummy_top);

			rcu_assign_pointer(cgrp->subsys[i],
					   cgroup_css(cgroup_dummy_top, ss));
			cgroup_css(cgrp, ss)->cgroup = cgrp;

			ss->root = root;