Merge branch 'for-3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup

Pull cgroup updates from Tejun Heo:
 "A lot of activities on the cgroup front.  Most changes aren't visible
  to userland at all at this point and are laying foundation for the
  planned unified hierarchy.

   - The biggest change is decoupling the lifetime management of css
     (cgroup_subsys_state) from that of cgroup's.  Because controllers
     (cpu, memory, block and so on) will need to be dynamically enabled
     and disabled, css which is the association point between a cgroup
     and a controller may come and go dynamically across the lifetime of
     a cgroup.  Till now, css's were created when the associated cgroup
     was created and stayed till the cgroup got destroyed.

     Assumptions around this tight coupling permeated through cgroup
     core and controllers.  These assumptions are gradually removed,
     which consists bulk of patches, and css destruction path is
     completely decoupled from cgroup destruction path.  Note that
     decoupling of creation path is relatively easy on top of these
     changes and the patchset is pending for the next window.

   - cgroup has its own event mechanism cgroup.event_control, which is
     only used by memcg.  It is overly complex trying to achieve high
     flexibility whose benefits seem dubious at best.  Going forward,
     new events will simply generate file modified event and the
     existing mechanism is being made specific to memcg.  This pull
     request contains prepatory patches for such change.

   - Various fixes and cleanups"

Fixed up conflict in kernel/cgroup.c as per Tejun.

* 'for-3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup: (69 commits)
  cgroup: fix cgroup_css() invocation in css_from_id()
  cgroup: make cgroup_write_event_control() use css_from_dir() instead of __d_cgrp()
  cgroup: make cgroup_event hold onto cgroup_subsys_state instead of cgroup
  cgroup: implement CFTYPE_NO_PREFIX
  cgroup: make cgroup_css() take cgroup_subsys * instead and allow NULL subsys
  cgroup: rename cgroup_css_from_dir() to css_from_dir() and update its syntax
  cgroup: fix cgroup_write_event_control()
  cgroup: fix subsystem file accesses on the root cgroup
  cgroup: change cgroup_from_id() to css_from_id()
  cgroup: use css_get() in cgroup_create() to check CSS_ROOT
  cpuset: remove an unncessary forward declaration
  cgroup: RCU protect each cgroup_subsys_state release
  cgroup: move subsys file removal to kill_css()
  cgroup: factor out kill_css()
  cgroup: decouple cgroup_subsys_state destruction from cgroup destruction
  cgroup: replace cgroup->css_kill_cnt with ->nr_css
  cgroup: bounce cgroup_subsys_state ref kill confirmation to a work item
  cgroup: move cgroup->subsys[] assignment to online_css()
  cgroup: reorganize css init / exit paths
  cgroup: add __rcu modifier to cgroup->subsys[]
  ...

1 files changed, 919 insertions, 724 deletions

diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index e91963302c0d..e0aeb32415ff 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -81,7 +81,7 @@
  */
 #ifdef CONFIG_PROVE_RCU
 DEFINE_MUTEX(cgroup_mutex);
-EXPORT_SYMBOL_GPL(cgroup_mutex);	/* only for task_subsys_state_check() */
+EXPORT_SYMBOL_GPL(cgroup_mutex);	/* only for lockdep */
 #else
 static DEFINE_MUTEX(cgroup_mutex);
 #endif
@@ -117,6 +117,7 @@ struct cfent {
 	struct list_head		node;
 	struct dentry			*dentry;
 	struct cftype			*type;
+	struct cgroup_subsys_state	*css;
 
 	/* file xattrs */
 	struct simple_xattrs		xattrs;
@@ -159,9 +160,9 @@ struct css_id {
  */
 struct cgroup_event {
 	/*
-	 * Cgroup which the event belongs to.
+	 * css which the event belongs to.
 	 */
-	struct cgroup *cgrp;
+	struct cgroup_subsys_state *css;
 	/*
 	 * Control file which the event associated.
 	 */
@@ -215,10 +216,33 @@ static u64 cgroup_serial_nr_next = 1;
  */
 static int need_forkexit_callback __read_mostly;
 
-static void cgroup_offline_fn(struct work_struct *work);
+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 cgroup_subsys *subsys,
-			      struct cftype cfts[], bool is_add);
+static int cgroup_addrm_files(struct cgroup *cgrp, struct cftype cfts[],
+			      bool is_add);
+
+/**
+ * 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)
@@ -365,9 +389,11 @@ static struct cgrp_cset_link init_cgrp_cset_link;
 static int cgroup_init_idr(struct cgroup_subsys *ss,
 			   struct cgroup_subsys_state *css);
 
-/* 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 cgroup_iter_start() */
+/*
+ * 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;
 
@@ -392,10 +418,12 @@ static unsigned long css_set_hash(struct cgroup_subsys_state *css[])
 	return key;
 }
 
-/* We don't maintain the lists running through each css_set to its
- * task until after the first call to cgroup_iter_start(). This
- * reduces the fork()/exit() overhead for people who have cgroups
- * compiled into their kernel but not actually in use */
+/*
+ * 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)
@@ -464,7 +492,7 @@ static inline void put_css_set_taskexit(struct css_set *cset)
  * @new_cgrp: cgroup that's being entered by the task
  * @template: desired set of css pointers in css_set (pre-calculated)
  *
- * Returns true if "cg" matches "old_cg" except for the hierarchy
+ * 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,
@@ -555,7 +583,7 @@ static struct css_set *find_existing_css_set(struct css_set *old_cset,
 			/* Subsystem is in this hierarchy. So we want
 			 * the subsystem state from the new
 			 * cgroup */
-			template[i] = cgrp->subsys[i];
+			template[i] = cgroup_css(cgrp, ss);
 		} else {
 			/* Subsystem is not in this hierarchy, so we
 			 * don't want to change the subsystem state */
@@ -803,8 +831,7 @@ static struct cgroup *task_cgroup_from_root(struct task_struct *task,
 
 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, bool base_files,
-			       unsigned long subsys_mask);
+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;
 
@@ -813,8 +840,7 @@ static struct backing_dev_info cgroup_backing_dev_info = {
 	.capabilities	= BDI_CAP_NO_ACCT_AND_WRITEBACK,
 };
 
-static int alloc_css_id(struct cgroup_subsys *ss,
-			struct cgroup *parent, struct cgroup *child);
+static int alloc_css_id(struct cgroup_subsys_state *child_css);
 
 static struct inode *cgroup_new_inode(umode_t mode, struct super_block *sb)
 {
@@ -845,15 +871,8 @@ static struct cgroup_name *cgroup_alloc_name(struct dentry *dentry)
 static void cgroup_free_fn(struct work_struct *work)
 {
 	struct cgroup *cgrp = container_of(work, struct cgroup, destroy_work);
-	struct cgroup_subsys *ss;
 
 	mutex_lock(&cgroup_mutex);
-	/*
-	 * Release the subsystem state objects.
-	 */
-	for_each_root_subsys(cgrp->root, ss)
-		ss->css_free(cgrp);
-
 	cgrp->root->number_of_cgroups--;
 	mutex_unlock(&cgroup_mutex);
 
@@ -864,8 +883,6 @@ static void cgroup_free_fn(struct work_struct *work)
 	 */
 	dput(cgrp->parent->dentry);
 
-	ida_simple_remove(&cgrp->root->cgroup_ida, cgrp->id);
-
 	/*
 	 * Drop the active superblock reference that we took when we
 	 * created the cgroup. This will free cgrp->root, if we are
@@ -956,27 +973,22 @@ static void cgroup_rm_file(struct cgroup *cgrp, const struct cftype *cft)
 }
 
 /**
- * cgroup_clear_directory - selective removal of base and subsystem files
- * @dir: directory containing the files
- * @base_files: true if the base files should be removed
+ * 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_directory(struct dentry *dir, bool base_files,
-				   unsigned long subsys_mask)
+static void cgroup_clear_dir(struct cgroup *cgrp, unsigned long subsys_mask)
 {
-	struct cgroup *cgrp = __d_cgrp(dir);
 	struct cgroup_subsys *ss;
+	int i;
 
-	for_each_root_subsys(cgrp->root, ss) {
+	for_each_subsys(ss, i) {
 		struct cftype_set *set;
-		if (!test_bit(ss->subsys_id, &subsys_mask))
+
+		if (!test_bit(i, &subsys_mask))
 			continue;
 		list_for_each_entry(set, &ss->cftsets, node)
-			cgroup_addrm_files(cgrp, NULL, set->cfts, false);
-	}
-	if (base_files) {
-		while (!list_empty(&cgrp->files))
-			cgroup_rm_file(cgrp, NULL);
+			cgroup_addrm_files(cgrp, set->cfts, false);
 	}
 }
 
@@ -986,9 +998,6 @@ static void cgroup_clear_directory(struct dentry *dir, bool base_files,
 static void cgroup_d_remove_dir(struct dentry *dentry)
 {
 	struct dentry *parent;
-	struct cgroupfs_root *root = dentry->d_sb->s_fs_info;
-
-	cgroup_clear_directory(dentry, true, root->subsys_mask);
 
 	parent = dentry->d_parent;
 	spin_lock(&parent->d_lock);
@@ -1009,79 +1018,84 @@ static int rebind_subsystems(struct cgroupfs_root *root,
 {
 	struct cgroup *cgrp = &root->top_cgroup;
 	struct cgroup_subsys *ss;
-	int i;
+	unsigned long pinned = 0;
+	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) {
-		unsigned long bit = 1UL << i;
-
-		if (!(bit & added_mask))
+		if (!(added_mask & (1 << i)))
 			continue;
 
+		/* is the subsystem mounted elsewhere? */
 		if (ss->root != &cgroup_dummy_root) {
-			/* Subsystem isn't free */
-			return -EBUSY;
+			ret = -EBUSY;
+			goto out_put;
+		}
+
+		/* pin the module */
+		if (!try_module_get(ss->module)) {
+			ret = -ENOENT;
+			goto out_put;
 		}
+		pinned |= 1 << i;
 	}
 
-	/* Currently we don't handle adding/removing subsystems when
-	 * any child cgroups exist. This is theoretically supportable
-	 * but involves complex error handling, so it's being left until
-	 * later */
-	if (root->number_of_cgroups > 1)
-		return -EBUSY;
+	/* subsys could be missing if unloaded between parsing and here */
+	if (added_mask != pinned) {
+		ret = -ENOENT;
+		goto out_put;
+	}
+
+	ret = cgroup_populate_dir(cgrp, added_mask);
+	if (ret)
+		goto out_put;
+
+	/*
+	 * Nothing can fail from this point on.  Remove files for the
+	 * removed subsystems and rebind each subsystem.
+	 */
+	cgroup_clear_dir(cgrp, removed_mask);
 
-	/* Process each subsystem */
 	for_each_subsys(ss, i) {
 		unsigned long bit = 1UL << i;
 
 		if (bit & added_mask) {
 			/* We're binding this subsystem to this hierarchy */
-			BUG_ON(cgrp->subsys[i]);
-			BUG_ON(!cgroup_dummy_top->subsys[i]);
-			BUG_ON(cgroup_dummy_top->subsys[i]->cgroup != cgroup_dummy_top);
+			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;
 
-			cgrp->subsys[i] = cgroup_dummy_top->subsys[i];
-			cgrp->subsys[i]->cgroup = cgrp;
 			list_move(&ss->sibling, &root->subsys_list);
 			ss->root = root;
 			if (ss->bind)
-				ss->bind(cgrp);
+				ss->bind(cgroup_css(cgrp, ss));
 
 			/* refcount was already taken, and we're keeping it */
 			root->subsys_mask |= bit;
 		} else if (bit & removed_mask) {
 			/* We're removing this subsystem */
-			BUG_ON(cgrp->subsys[i] != cgroup_dummy_top->subsys[i]);
-			BUG_ON(cgrp->subsys[i]->cgroup != cgrp);
+			BUG_ON(cgroup_css(cgrp, ss) != cgroup_css(cgroup_dummy_top, ss));
+			BUG_ON(cgroup_css(cgrp, ss)->cgroup != cgrp);
 
 			if (ss->bind)
-				ss->bind(cgroup_dummy_top);
-			cgroup_dummy_top->subsys[i]->cgroup = cgroup_dummy_top;
-			cgrp->subsys[i] = NULL;
+				ss->bind(cgroup_css(cgroup_dummy_top, ss));
+
+			cgroup_css(cgroup_dummy_top, ss)->cgroup = cgroup_dummy_top;
+			RCU_INIT_POINTER(cgrp->subsys[i], NULL);
+
 			cgroup_subsys[i]->root = &cgroup_dummy_root;
 			list_move(&ss->sibling, &cgroup_dummy_root.subsys_list);
 
 			/* subsystem is now free - drop reference on module */
 			module_put(ss->module);
 			root->subsys_mask &= ~bit;
-		} else if (bit & root->subsys_mask) {
-			/* Subsystem state should already exist */
-			BUG_ON(!cgrp->subsys[i]);
-			/*
-			 * a refcount was taken, but we already had one, so
-			 * drop the extra reference.
-			 */
-			module_put(ss->module);
-#ifdef CONFIG_MODULE_UNLOAD
-			BUG_ON(ss->module && !module_refcount(ss->module));
-#endif
-		} else {
-			/* Subsystem state shouldn't exist */
-			BUG_ON(cgrp->subsys[i]);
 		}
 	}
 
@@ -1092,6 +1106,12 @@ static int rebind_subsystems(struct cgroupfs_root *root,
 	root->flags |= CGRP_ROOT_SUBSYS_BOUND;
 
 	return 0;
+
+out_put:
+	for_each_subsys(ss, i)
+		if (pinned & (1 << i))
+			module_put(ss->module);
+	return ret;
 }
 
 static int cgroup_show_options(struct seq_file *seq, struct dentry *dentry)
@@ -1142,7 +1162,6 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
 	char *token, *o = data;
 	bool all_ss = false, one_ss = false;
 	unsigned long mask = (unsigned long)-1;
-	bool module_pin_failed = false;
 	struct cgroup_subsys *ss;
 	int i;
 
@@ -1285,52 +1304,9 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
 	if (!opts->subsys_mask && !opts->name)
 		return -EINVAL;
 
-	/*
-	 * Grab references on all the modules we'll need, so the subsystems
-	 * don't dance around before rebind_subsystems attaches them. This may
-	 * take duplicate reference counts on a subsystem that's already used,
-	 * but rebind_subsystems handles this case.
-	 */
-	for_each_subsys(ss, i) {
-		if (!(opts->subsys_mask & (1UL << i)))
-			continue;
-		if (!try_module_get(cgroup_subsys[i]->module)) {
-			module_pin_failed = true;
-			break;
-		}
-	}
-	if (module_pin_failed) {
-		/*
-		 * oops, one of the modules was going away. this means that we
-		 * raced with a module_delete call, and to the user this is
-		 * essentially a "subsystem doesn't exist" case.
-		 */
-		for (i--; i >= 0; i--) {
-			/* drop refcounts only on the ones we took */
-			unsigned long bit = 1UL << i;
-
-			if (!(bit & opts->subsys_mask))
-				continue;
-			module_put(cgroup_subsys[i]->module);
-		}
-		return -ENOENT;
-	}
-
 	return 0;
 }
 
-static void drop_parsed_module_refcounts(unsigned long subsys_mask)
-{
-	struct cgroup_subsys *ss;
-	int i;
-
-	mutex_lock(&cgroup_mutex);
-	for_each_subsys(ss, i)
-		if (subsys_mask & (1UL << i))
-			module_put(cgroup_subsys[i]->module);
-	mutex_unlock(&cgroup_mutex);
-}
-
 static int cgroup_remount(struct super_block *sb, int *flags, char *data)
 {
 	int ret = 0;
@@ -1370,22 +1346,15 @@ static int cgroup_remount(struct super_block *sb, int *flags, char *data)
 		goto out_unlock;
 	}
 
-	/*
-	 * Clear out the files of subsystems that should be removed, do
-	 * this before rebind_subsystems, since rebind_subsystems may
-	 * change this hierarchy's subsys_list.
-	 */
-	cgroup_clear_directory(cgrp->dentry, false, removed_mask);
-
-	ret = rebind_subsystems(root, added_mask, removed_mask);
-	if (ret) {
-		/* rebind_subsystems failed, re-populate the removed files */
-		cgroup_populate_dir(cgrp, false, removed_mask);
+	/* remounting is not allowed for populated hierarchies */
+	if (root->number_of_cgroups > 1) {
+		ret = -EBUSY;
 		goto out_unlock;
 	}
 
-	/* re-populate subsystem files */
-	cgroup_populate_dir(cgrp, false, added_mask);
+	ret = rebind_subsystems(root, added_mask, removed_mask);
+	if (ret)
+		goto out_unlock;
 
 	if (opts.release_agent)
 		strcpy(root->release_agent_path, opts.release_agent);
@@ -1395,8 +1364,6 @@ static int cgroup_remount(struct super_block *sb, int *flags, char *data)
 	mutex_unlock(&cgroup_root_mutex);
 	mutex_unlock(&cgroup_mutex);
 	mutex_unlock(&cgrp->dentry->d_inode->i_mutex);
-	if (ret)
-		drop_parsed_module_refcounts(opts.subsys_mask);
 	return ret;
 }
 
@@ -1416,6 +1383,7 @@ static void init_cgroup_housekeeping(struct cgroup *cgrp)
 	INIT_LIST_HEAD(&cgrp->release_list);
 	INIT_LIST_HEAD(&cgrp->pidlists);
 	mutex_init(&cgrp->pidlist_mutex);
+	cgrp->dummy_css.cgroup = cgrp;
 	INIT_LIST_HEAD(&cgrp->event_list);
 	spin_lock_init(&cgrp->event_list_lock);
 	simple_xattrs_init(&cgrp->xattrs);
@@ -1431,6 +1399,7 @@ static void init_cgroup_root(struct cgroupfs_root *root)
 	cgrp->root = root;
 	RCU_INIT_POINTER(cgrp->name, &root_cgroup_name);
 	init_cgroup_housekeeping(cgrp);
+	idr_init(&root->cgroup_idr);
 }
 
 static int cgroup_init_root_id(struct cgroupfs_root *root, int start, int end)
@@ -1503,7 +1472,6 @@ static struct cgroupfs_root *cgroup_root_from_opts(struct cgroup_sb_opts *opts)
 	 */
 	root->subsys_mask = opts->subsys_mask;
 	root->flags = opts->flags;
-	ida_init(&root->cgroup_ida);
 	if (opts->release_agent)
 		strcpy(root->release_agent_path, opts->release_agent);
 	if (opts->name)
@@ -1519,7 +1487,7 @@ static void cgroup_free_root(struct cgroupfs_root *root)
 		/* hierarhcy ID shoulid already have been released */
 		WARN_ON_ONCE(root->hierarchy_id);
 
-		ida_destroy(&root->cgroup_ida);
+		idr_destroy(&root->cgroup_idr);
 		kfree(root);
 	}
 }
@@ -1584,7 +1552,9 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 	int ret = 0;
 	struct super_block *sb;
 	struct cgroupfs_root *new_root;
+	struct list_head tmp_links;
 	struct inode *inode;
+	const struct cred *cred;
 
 	/* First find the desired set of subsystems */
 	mutex_lock(&cgroup_mutex);
@@ -1600,7 +1570,7 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 	new_root = cgroup_root_from_opts(&opts);
 	if (IS_ERR(new_root)) {
 		ret = PTR_ERR(new_root);
-		goto drop_modules;
+		goto out_err;
 	}
 	opts.new_root = new_root;
 
@@ -1609,17 +1579,15 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 	if (IS_ERR(sb)) {
 		ret = PTR_ERR(sb);
 		cgroup_free_root(opts.new_root);
-		goto drop_modules;
+		goto out_err;
 	}
 
 	root = sb->s_fs_info;
 	BUG_ON(!root);
 	if (root == opts.new_root) {
 		/* We used the new root structure, so this is a new hierarchy */
-		struct list_head tmp_links;
 		struct cgroup *root_cgrp = &root->top_cgroup;
 		struct cgroupfs_root *existing_root;
-		const struct cred *cred;
 		int i;
 		struct css_set *cset;
 
@@ -1634,6 +1602,11 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 		mutex_lock(&cgroup_mutex);
 		mutex_lock(&cgroup_root_mutex);
 
+		root_cgrp->id = idr_alloc(&root->cgroup_idr, root_cgrp,
+					   0, 1, GFP_KERNEL);
+		if (root_cgrp->id < 0)
+			goto unlock_drop;
+
 		/* Check for name clashes with existing mounts */
 		ret = -EBUSY;
 		if (strlen(root->name))
@@ -1657,26 +1630,37 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 		if (ret)
 			goto unlock_drop;
 
+		sb->s_root->d_fsdata = root_cgrp;
+		root_cgrp->dentry = sb->s_root;
+
+		/*
+		 * We're inside get_sb() and will call lookup_one_len() to
+		 * create the root files, which doesn't work if SELinux is
+		 * in use.  The following cred dancing somehow works around
+		 * it.  See 2ce9738ba ("cgroupfs: use init_cred when
+		 * populating new cgroupfs mount") for more details.
+		 */
+		cred = override_creds(&init_cred);
+
+		ret = cgroup_addrm_files(root_cgrp, cgroup_base_files, true);
+		if (ret)
+			goto rm_base_files;
+
 		ret = rebind_subsystems(root, root->subsys_mask, 0);
-		if (ret == -EBUSY) {
-			free_cgrp_cset_links(&tmp_links);
-			goto unlock_drop;
-		}
+		if (ret)
+			goto rm_base_files;
+
+		revert_creds(cred);
+
 		/*
 		 * There must be no failure case after here, since rebinding
 		 * takes care of subsystems' refcounts, which are explicitly
 		 * dropped in the failure exit path.
 		 */
 
-		/* EBUSY should be the only error here */
-		BUG_ON(ret);
-
 		list_add(&root->root_list, &cgroup_roots);
 		cgroup_root_count++;
 
-		sb->s_root->d_fsdata = root_cgrp;
-		root->top_cgroup.dentry = sb->s_root;
-
 		/* Link the top cgroup in this hierarchy into all
 		 * the css_set objects */
 		write_lock(&css_set_lock);
@@ -1689,9 +1673,6 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 		BUG_ON(!list_empty(&root_cgrp->children));
 		BUG_ON(root->number_of_cgroups != 1);
 
-		cred = override_creds(&init_cred);
-		cgroup_populate_dir(root_cgrp, true, root->subsys_mask);
-		revert_creds(cred);
 		mutex_unlock(&cgroup_root_mutex);
 		mutex_unlock(&cgroup_mutex);
 		mutex_unlock(&inode->i_mutex);
@@ -1711,15 +1692,16 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 				pr_warning("cgroup: new mount options do not match the existing superblock, will be ignored\n");
 			}
 		}
-
-		/* no subsys rebinding, so refcounts don't change */
-		drop_parsed_module_refcounts(opts.subsys_mask);
 	}
 
 	kfree(opts.release_agent);
 	kfree(opts.name);
 	return dget(sb->s_root);
 
+ rm_base_files:
+	free_cgrp_cset_links(&tmp_links);
+	cgroup_addrm_files(&root->top_cgroup, cgroup_base_files, false);
+	revert_creds(cred);
  unlock_drop:
 	cgroup_exit_root_id(root);
 	mutex_unlock(&cgroup_root_mutex);
@@ -1727,8 +1709,6 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
 	mutex_unlock(&inode->i_mutex);
  drop_new_super:
 	deactivate_locked_super(sb);
- drop_modules:
-	drop_parsed_module_refcounts(opts.subsys_mask);
  out_err:
 	kfree(opts.release_agent);
 	kfree(opts.name);
@@ -1746,6 +1726,7 @@ static void cgroup_kill_sb(struct super_block *sb) {
 	BUG_ON(root->number_of_cgroups != 1);
 	BUG_ON(!list_empty(&cgrp->children));
 
+	mutex_lock(&cgrp->dentry->d_inode->i_mutex);
 	mutex_lock(&cgroup_mutex);
 	mutex_lock(&cgroup_root_mutex);
 
@@ -1778,6 +1759,7 @@ static void cgroup_kill_sb(struct super_block *sb) {
 
 	mutex_unlock(&cgroup_root_mutex);
 	mutex_unlock(&cgroup_mutex);
+	mutex_unlock(&cgrp->dentry->d_inode->i_mutex);
 
 	simple_xattrs_free(&cgrp->xattrs);
 
@@ -1889,7 +1871,7 @@ EXPORT_SYMBOL_GPL(task_cgroup_path);
 struct task_and_cgroup {
 	struct task_struct	*task;
 	struct cgroup		*cgrp;
-	struct css_set		*cg;
+	struct css_set		*cset;
 };
 
 struct cgroup_taskset {
@@ -1939,18 +1921,20 @@ struct task_struct *cgroup_taskset_next(struct cgroup_taskset *tset)
 EXPORT_SYMBOL_GPL(cgroup_taskset_next);
 
 /**
- * cgroup_taskset_cur_cgroup - return the matching cgroup for the current task
+ * cgroup_taskset_cur_css - return the matching css for the current task
  * @tset: taskset of interest
+ * @subsys_id: the ID of the target subsystem
  *
- * Return the cgroup for the current (last returned) task of @tset.  This
- * function must be preceded by either cgroup_taskset_first() or
- * cgroup_taskset_next().
+ * Return the css for the current (last returned) task of @tset for
+ * subsystem specified by @subsys_id.  This function must be preceded by
+ * either cgroup_taskset_first() or cgroup_taskset_next().
  */
-struct cgroup *cgroup_taskset_cur_cgroup(struct cgroup_taskset *tset)
+struct cgroup_subsys_state *cgroup_taskset_cur_css(struct cgroup_taskset *tset,
+						   int subsys_id)
 {
-	return tset->cur_cgrp;
+	return cgroup_css(tset->cur_cgrp, cgroup_subsys[subsys_id]);
 }
-EXPORT_SYMBOL_GPL(cgroup_taskset_cur_cgroup);
+EXPORT_SYMBOL_GPL(cgroup_taskset_cur_css);
 
 /**
  * cgroup_taskset_size - return the number of tasks in taskset
@@ -2089,8 +2073,10 @@ static int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk,
 	 * step 1: check that we can legitimately attach to the cgroup.
 	 */
 	for_each_root_subsys(root, ss) {
+		struct cgroup_subsys_state *css = cgroup_css(cgrp, ss);
+
 		if (ss->can_attach) {
-			retval = ss->can_attach(cgrp, &tset);
+			retval = ss->can_attach(css, &tset);
 			if (retval) {
 				failed_ss = ss;
 				goto out_cancel_attach;
@@ -2107,8 +2093,8 @@ static int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk,
 
 		tc = flex_array_get(group, i);
 		old_cset = task_css_set(tc->task);
-		tc->cg = find_css_set(old_cset, cgrp);
-		if (!tc->cg) {
+		tc->cset = find_css_set(old_cset, cgrp);
+		if (!tc->cset) {
 			retval = -ENOMEM;
 			goto out_put_css_set_refs;
 		}
@@ -2121,7 +2107,7 @@ static int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk,
 	 */
 	for (i = 0; i < group_size; i++) {
 		tc = flex_array_get(group, i);
-		cgroup_task_migrate(tc->cgrp, tc->task, tc->cg);
+		cgroup_task_migrate(tc->cgrp, tc->task, tc->cset);
 	}
 	/* nothing is sensitive to fork() after this point. */
 
@@ -2129,8 +2115,10 @@ static int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk,
 	 * step 4: do subsystem attach callbacks.
 	 */
 	for_each_root_subsys(root, ss) {
+		struct cgroup_subsys_state *css = cgroup_css(cgrp, ss);
+
 		if (ss->attach)
-			ss->attach(cgrp, &tset);
+			ss->attach(css, &tset);
 	}
 
 	/*
@@ -2141,18 +2129,20 @@ out_put_css_set_refs:
 	if (retval) {
 		for (i = 0; i < group_size; i++) {
 			tc = flex_array_get(group, i);
-			if (!tc->cg)
+			if (!tc->cset)
 				break;
-			put_css_set(tc->cg);
+			put_css_set(tc->cset);
 		}
 	}
 out_cancel_attach:
 	if (retval) {
 		for_each_root_subsys(root, ss) {
+			struct cgroup_subsys_state *css = cgroup_css(cgrp, ss);
+
 			if (ss == failed_ss)
 				break;
 			if (ss->cancel_attach)
-				ss->cancel_attach(cgrp, &tset);
+				ss->cancel_attach(css, &tset);
 		}
 	}
 out_free_group_list:
@@ -2253,9 +2243,9 @@ int cgroup_attach_task_all(struct task_struct *from, struct task_struct *tsk)
 
 	mutex_lock(&cgroup_mutex);
 	for_each_active_root(root) {
-		struct cgroup *from_cg = task_cgroup_from_root(from, root);
+		struct cgroup *from_cgrp = task_cgroup_from_root(from, root);
 
-		retval = cgroup_attach_task(from_cg, tsk, false);
+		retval = cgroup_attach_task(from_cgrp, tsk, false);
 		if (retval)
 			break;
 	}
@@ -2265,34 +2255,38 @@ int cgroup_attach_task_all(struct task_struct *from, struct task_struct *tsk)
 }
 EXPORT_SYMBOL_GPL(cgroup_attach_task_all);
 
-static int cgroup_tasks_write(struct cgroup *cgrp, struct cftype *cft, u64 pid)
+static int cgroup_tasks_write(struct cgroup_subsys_state *css,
+			      struct cftype *cft, u64 pid)
 {
-	return attach_task_by_pid(cgrp, pid, false);
+	return attach_task_by_pid(css->cgroup, pid, false);
 }
 
-static int cgroup_procs_write(struct cgroup *cgrp, struct cftype *cft, u64 tgid)
+static int cgroup_procs_write(struct cgroup_subsys_state *css,
+			      struct cftype *cft, u64 tgid)
 {
-	return attach_task_by_pid(cgrp, tgid, true);
+	return attach_task_by_pid(css->cgroup, tgid, true);
 }
 
-static int cgroup_release_agent_write(struct cgroup *cgrp, struct cftype *cft,
-				      const char *buffer)
+static int cgroup_release_agent_write(struct cgroup_subsys_state *css,
+				      struct cftype *cft, const char *buffer)
 {
-	BUILD_BUG_ON(sizeof(cgrp->root->release_agent_path) < PATH_MAX);
+	BUILD_BUG_ON(sizeof(css->cgroup->root->release_agent_path) < PATH_MAX);
 	if (strlen(buffer) >= PATH_MAX)
 		return -EINVAL;
-	if (!cgroup_lock_live_group(cgrp))
+	if (!cgroup_lock_live_group(css->cgroup))
 		return -ENODEV;
 	mutex_lock(&cgroup_root_mutex);
-	strcpy(cgrp->root->release_agent_path, buffer);
+	strcpy(css->cgroup->root->release_agent_path, buffer);
 	mutex_unlock(&cgroup_root_mutex);
 	mutex_unlock(&cgroup_mutex);
 	return 0;
 }
 
-static int cgroup_release_agent_show(struct cgroup *cgrp, struct cftype *cft,
-				     struct seq_file *seq)
+static int cgroup_release_agent_show(struct cgroup_subsys_state *css,
+				     struct cftype *cft, struct seq_file *seq)
 {
+	struct cgroup *cgrp = css->cgroup;
+
 	if (!cgroup_lock_live_group(cgrp))
 		return -ENODEV;
 	seq_puts(seq, cgrp->root->release_agent_path);
@@ -2301,20 +2295,20 @@ static int cgroup_release_agent_show(struct cgroup *cgrp, struct cftype *cft,
 	return 0;
 }
 
-static int cgroup_sane_behavior_show(struct cgroup *cgrp, struct cftype *cft,
-				     struct seq_file *seq)
+static int cgroup_sane_behavior_show(struct cgroup_subsys_state *css,
+				     struct cftype *cft, struct seq_file *seq)
 {
-	seq_printf(seq, "%d\n", cgroup_sane_behavior(cgrp));
+	seq_printf(seq, "%d\n", cgroup_sane_behavior(css->cgroup));
 	return 0;
 }
 
 /* A buffer size big enough for numbers or short strings */
 #define CGROUP_LOCAL_BUFFER_SIZE 64
 
-static ssize_t cgroup_write_X64(struct cgroup *cgrp, struct cftype *cft,
-				struct file *file,
-				const char __user *userbuf,
-				size_t nbytes, loff_t *unused_ppos)
+static ssize_t cgroup_write_X64(struct cgroup_subsys_state *css,
+				struct cftype *cft, struct file *file,
+				const char __user *userbuf, size_t nbytes,
+				loff_t *unused_ppos)
 {
 	char buffer[CGROUP_LOCAL_BUFFER_SIZE];
 	int retval = 0;
@@ -2332,22 +2326,22 @@ static ssize_t cgroup_write_X64(struct cgroup *cgrp, struct cftype *cft,
 		u64 val = simple_strtoull(strstrip(buffer), &end, 0);
 		if (*end)
 			return -EINVAL;
-		retval = cft->write_u64(cgrp, cft, val);
+		retval = cft->write_u64(css, cft, val);
 	} else {
 		s64 val = simple_strtoll(strstrip(buffer), &end, 0);
 		if (*end)
 			return -EINVAL;
-		retval = cft->write_s64(cgrp, cft, val);
+		retval = cft->write_s64(css, cft, val);
 	}
 	if (!retval)
 		retval = nbytes;
 	return retval;
 }
 
-static ssize_t cgroup_write_string(struct cgroup *cgrp, struct cftype *cft,
-				   struct file *file,
-				   const char __user *userbuf,
-				   size_t nbytes, loff_t *unused_ppos)
+static ssize_t cgroup_write_string(struct cgroup_subsys_state *css,
+				   struct cftype *cft, struct file *file,
+				   const char __user *userbuf, size_t nbytes,
+				   loff_t *unused_ppos)
 {
 	char local_buffer[CGROUP_LOCAL_BUFFER_SIZE];
 	int retval = 0;
@@ -2370,7 +2364,7 @@ static ssize_t cgroup_write_string(struct cgroup *cgrp, struct cftype *cft,
 	}
 
 	buffer[nbytes] = 0;     /* nul-terminate */
-	retval = cft->write_string(cgrp, cft, strstrip(buffer));
+	retval = cft->write_string(css, cft, strstrip(buffer));
 	if (!retval)
 		retval = nbytes;
 out:
@@ -2380,65 +2374,60 @@ out:
 }
 
 static ssize_t cgroup_file_write(struct file *file, const char __user *buf,
-						size_t nbytes, loff_t *ppos)
+				 size_t nbytes, loff_t *ppos)
 {
+	struct cfent *cfe = __d_cfe(file->f_dentry);
 	struct cftype *cft = __d_cft(file->f_dentry);
-	struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent);
+	struct cgroup_subsys_state *css = cfe->css;
 
-	if (cgroup_is_dead(cgrp))
-		return -ENODEV;
 	if (cft->write)
-		return cft->write(cgrp, cft, file, buf, nbytes, ppos);
+		return cft->write(css, cft, file, buf, nbytes, ppos);
 	if (cft->write_u64 || cft->write_s64)
-		return cgroup_write_X64(cgrp, cft, file, buf, nbytes, ppos);
+		return cgroup_write_X64(css, cft, file, buf, nbytes, ppos);
 	if (cft->write_string)
-		return cgroup_write_string(cgrp, cft, file, buf, nbytes, ppos);
+		return cgroup_write_string(css, cft, file, buf, nbytes, ppos);
 	if (cft->trigger) {
-		int ret = cft->trigger(cgrp, (unsigned int)cft->private);
+		int ret = cft->trigger(css, (unsigned int)cft->private);
 		return ret ? ret : nbytes;
 	}
 	return -EINVAL;
 }
 
-static ssize_t cgroup_read_u64(struct cgroup *cgrp, struct cftype *cft,
-			       struct file *file,
-			       char __user *buf, size_t nbytes,
-			       loff_t *ppos)
+static ssize_t cgroup_read_u64(struct cgroup_subsys_state *css,
+			       struct cftype *cft, struct file *file,
+			       char __user *buf, size_t nbytes, loff_t *ppos)
 {
 	char tmp[CGROUP_LOCAL_BUFFER_SIZE];
-	u64 val = cft->read_u64(cgrp, cft);
+	u64 val = cft->read_u64(css, cft);
 	int len = sprintf(tmp, "%llu\n", (unsigned long long) val);
 
 	return simple_read_from_buffer(buf, nbytes, ppos, tmp, len);
 }
 
-static ssize_t cgroup_read_s64(struct cgroup *cgrp, struct cftype *cft,
-			       struct file *file,
-			       char __user *buf, size_t nbytes,
-			       loff_t *ppos)
+static ssize_t cgroup_read_s64(struct cgroup_subsys_state *css,
+			       struct cftype *cft, struct file *file,
+			       char __user *buf, size_t nbytes, loff_t *ppos)
 {
 	char tmp[CGROUP_LOCAL_BUFFER_SIZE];
-	s64 val = cft->read_s64(cgrp, cft);
+	s64 val = cft->read_s64(css, cft);
 	int len = sprintf(tmp, "%lld\n", (long long) val);
 
 	return simple_read_from_buffer(buf, nbytes, ppos, tmp, len);
 }
 
 static ssize_t cgroup_file_read(struct file *file, char __user *buf,
-				   size_t nbytes, loff_t *ppos)
+				size_t nbytes, loff_t *ppos)
 {
+	struct cfent *cfe = __d_cfe(file->f_dentry);
 	struct cftype *cft = __d_cft(file->f_dentry);
-	struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent);
-
-	if (cgroup_is_dead(cgrp))
-		return -ENODEV;
+	struct cgroup_subsys_state *css = cfe->css;
 
 	if (cft->read)
-		return cft->read(cgrp, cft, file, buf, nbytes, ppos);
+		return cft->read(css, cft, file, buf, nbytes, ppos);
 	if (cft->read_u64)
-		return cgroup_read_u64(cgrp, cft, file, buf, nbytes, ppos);
+		return cgroup_read_u64(css, cft, file, buf, nbytes, ppos);
 	if (cft->read_s64)
-		return cgroup_read_s64(cgrp, cft, file, buf, nbytes, ppos);
+		return cgroup_read_s64(css, cft, file, buf, nbytes, ppos);
 	return -EINVAL;
 }
 
@@ -2447,11 +2436,6 @@ static ssize_t cgroup_file_read(struct file *file, char __user *buf,
  * supports string->u64 maps, but can be extended in future.
  */
 
-struct cgroup_seqfile_state {
-	struct cftype *cft;
-	struct cgroup *cgroup;
-};
-
 static int cgroup_map_add(struct cgroup_map_cb *cb, const char *key, u64 value)
 {
 	struct seq_file *sf = cb->state;
@@ -2460,69 +2444,86 @@ static int cgroup_map_add(struct cgroup_map_cb *cb, const char *key, u64 value)
 
 static int cgroup_seqfile_show(struct seq_file *m, void *arg)
 {
-	struct cgroup_seqfile_state *state = m->private;
-	struct cftype *cft = state->cft;
+	struct cfent *cfe = m->private;
+	struct cftype *cft = cfe->type;
+	struct cgroup_subsys_state *css = cfe->css;
+
 	if (cft->read_map) {
 		struct cgroup_map_cb cb = {
 			.fill = cgroup_map_add,
 			.state = m,
 		};
-		return cft->read_map(state->cgroup, cft, &cb);
+		return cft->read_map(css, cft, &cb);
 	}
-	return cft->read_seq_string(state->cgroup, cft, m);
-}
-
-static int cgroup_seqfile_release(struct inode *inode, struct file *file)
-{
-	struct seq_file *seq = file->private_data;
-	kfree(seq->priva