memcontrol.c

	 * No LRU may hold pages because all pages are UNEVICTABLE or
	 * memcg is too small and all pages are not on LRU. In that case,
	 * we use curret node.
	 */
	if (unlikely(node == MAX_NUMNODES))
		node = numa_node_id();

	memcg->last_scanned_node = node;
	return node;
}

/*
 * Check all nodes whether it contains reclaimable pages or not.
 * For quick scan, we make use of scan_nodes. This will allow us to skip
 * unused nodes. But scan_nodes is lazily updated and may not cotain
 * enough new information. We need to do double check.
 */
static bool mem_cgroup_reclaimable(struct mem_cgroup *memcg, bool noswap)
{
	int nid;

	/*
	 * quick check...making use of scan_node.
	 * We can skip unused nodes.
	 */
	if (!nodes_empty(memcg->scan_nodes)) {
		for (nid = first_node(memcg->scan_nodes);
		     nid < MAX_NUMNODES;
		     nid = next_node(nid, memcg->scan_nodes)) {

			if (test_mem_cgroup_node_reclaimable(memcg, nid, noswap))
				return true;
		}
	}
	/*
	 * Check rest of nodes.
	 */
	for_each_node_state(nid, N_MEMORY) {
		if (node_isset(nid, memcg->scan_nodes))
			continue;
		if (test_mem_cgroup_node_reclaimable(memcg, nid, noswap))
			return true;
	}
	return false;
}

#else
int mem_cgroup_select_victim_node(struct mem_cgroup *memcg)
{
	return 0;
}

static bool mem_cgroup_reclaimable(struct mem_cgroup *memcg, bool noswap)
{
	return test_mem_cgroup_node_reclaimable(memcg, 0, noswap);
}
#endif

static int mem_cgroup_soft_reclaim(struct mem_cgroup *root_memcg,
				   struct zone *zone,
				   gfp_t gfp_mask,
				   unsigned long *total_scanned)
{
	struct mem_cgroup *victim = NULL;
	int total = 0;
	int loop = 0;
	unsigned long excess;
	unsigned long nr_scanned;
	struct mem_cgroup_reclaim_cookie reclaim = {
		.zone = zone,
		.priority = 0,
	};

	excess = res_counter_soft_limit_excess(&root_memcg->res) >> PAGE_SHIFT;

	while (1) {
		victim = mem_cgroup_iter(root_memcg, victim, &reclaim);
		if (!victim) {
			loop++;
			if (loop >= 2) {
				/*
				 * If we have not been able to reclaim
				 * anything, it might because there are
				 * no reclaimable pages under this hierarchy
				 */
				if (!total)
					break;
				/*
				 * We want to do more targeted reclaim.
				 * excess >> 2 is not to excessive so as to
				 * reclaim too much, nor too less that we keep
				 * coming back to reclaim from this cgroup
				 */
				if (total >= (excess >> 2) ||
					(loop > MEM_CGROUP_MAX_RECLAIM_LOOPS))
					break;
			}
			continue;
		}
		if (!mem_cgroup_reclaimable(victim, false))
			continue;
		total += mem_cgroup_shrink_node_zone(victim, gfp_mask, false,
						     zone, &nr_scanned);
		*total_scanned += nr_scanned;
		if (!res_counter_soft_limit_excess(&root_memcg->res))
			break;
	}
	mem_cgroup_iter_break(root_memcg, victim);
	return total;
}

#ifdef CONFIG_LOCKDEP
static struct lockdep_map memcg_oom_lock_dep_map = {
	.name = "memcg_oom_lock",
};
#endif

static DEFINE_SPINLOCK(memcg_oom_lock);

/*
 * Check OOM-Killer is already running under our hierarchy.
 * If someone is running, return false.
 */
static bool mem_cgroup_oom_trylock(struct mem_cgroup *memcg)
{
	struct mem_cgroup *iter, *failed = NULL;

	spin_lock(&memcg_oom_lock);

	for_each_mem_cgroup_tree(iter, memcg) {
		if (iter->oom_lock) {
			/*
			 * this subtree of our hierarchy is already locked
			 * so we cannot give a lock.
			 */
			failed = iter;
			mem_cgroup_iter_break(memcg, iter);
			break;
		} else
			iter->oom_lock = true;
	}

	if (failed) {
		/*
		 * OK, we failed to lock the whole subtree so we have
		 * to clean up what we set up to the failing subtree
		 */
		for_each_mem_cgroup_tree(iter, memcg) {
			if (iter == failed) {
				mem_cgroup_iter_break(memcg, iter);
				break;
			}
			iter->oom_lock = false;
		}
	} else
		mutex_acquire(&memcg_oom_lock_dep_map, 0, 1, _RET_IP_);

	spin_unlock(&memcg_oom_lock);

	return !failed;
}

static void mem_cgroup_oom_unlock(struct mem_cgroup *memcg)
{
	struct mem_cgroup *iter;

	spin_lock(&memcg_oom_lock);
	mutex_release(&memcg_oom_lock_dep_map, 1, _RET_IP_);
	for_each_mem_cgroup_tree(iter, memcg)
		iter->oom_lock = false;
	spin_unlock(&memcg_oom_lock);
}

static void mem_cgroup_mark_under_oom(struct mem_cgroup *memcg)
{
	struct mem_cgroup *iter;

	for_each_mem_cgroup_tree(iter, memcg)
		atomic_inc(&iter->under_oom);
}

static void mem_cgroup_unmark_under_oom(struct mem_cgroup *memcg)
{
	struct mem_cgroup *iter;

	/*
	 * When a new child is created while the hierarchy is under oom,
	 * mem_cgroup_oom_lock() may not be called. We have to use
	 * atomic_add_unless() here.
	 */
	for_each_mem_cgroup_tree(iter, memcg)
		atomic_add_unless(&iter->under_oom, -1, 0);
}

static DECLARE_WAIT_QUEUE_HEAD(memcg_oom_waitq);

struct oom_wait_info {
	struct mem_cgroup *memcg;
	wait_queue_t	wait;
};

static int memcg_oom_wake_function(wait_queue_t *wait,
	unsigned mode, int sync, void *arg)
{
	struct mem_cgroup *wake_memcg = (struct mem_cgroup *)arg;
	struct mem_cgroup *oom_wait_memcg;
	struct oom_wait_info *oom_wait_info;

	oom_wait_info = container_of(wait, struct oom_wait_info, wait);
	oom_wait_memcg = oom_wait_info->memcg;

	/*
	 * Both of oom_wait_info->memcg and wake_memcg are stable under us.
	 * Then we can use css_is_ancestor without taking care of RCU.
	 */
	if (!mem_cgroup_same_or_subtree(oom_wait_memcg, wake_memcg)
		&& !mem_cgroup_same_or_subtree(wake_memcg, oom_wait_memcg))
		return 0;
	return autoremove_wake_function(wait, mode, sync, arg);
}

static void memcg_wakeup_oom(struct mem_cgroup *memcg)
{
	atomic_inc(&memcg->oom_wakeups);
	/* for filtering, pass "memcg" as argument. */
	__wake_up(&memcg_oom_waitq, TASK_NORMAL, 0, memcg);
}

static void memcg_oom_recover(struct mem_cgroup *memcg)
{
	if (memcg && atomic_read(&memcg->under_oom))
		memcg_wakeup_oom(memcg);
}

static void mem_cgroup_oom(struct mem_cgroup *memcg, gfp_t mask, int order)
{
	if (!current->memcg_oom.may_oom)
		return;
	/*
	 * We are in the middle of the charge context here, so we
	 * don't want to block when potentially sitting on a callstack
	 * that holds all kinds of filesystem and mm locks.
	 *
	 * Also, the caller may handle a failed allocation gracefully
	 * (like optional page cache readahead) and so an OOM killer
	 * invocation might not even be necessary.
	 *
	 * That's why we don't do anything here except remember the
	 * OOM context and then deal with it at the end of the page
	 * fault when the stack is unwound, the locks are released,
	 * and when we know whether the fault was overall successful.
	 */
	css_get(&memcg->css);
	current->memcg_oom.memcg = memcg;
	current->memcg_oom.gfp_mask = mask;
	current->memcg_oom.order = order;
}

/**
 * mem_cgroup_oom_synchronize - complete memcg OOM handling
 * @handle: actually kill/wait or just clean up the OOM state
 *
 * This has to be called at the end of a page fault if the memcg OOM
 * handler was enabled.
 *
 * Memcg supports userspace OOM handling where failed allocations must
 * sleep on a waitqueue until the userspace task resolves the
 * situation.  Sleeping directly in the charge context with all kinds
 * of locks held is not a good idea, instead we remember an OOM state
 * in the task and mem_cgroup_oom_synchronize() has to be called at
 * the end of the page fault to complete the OOM handling.
 *
 * Returns %true if an ongoing memcg OOM situation was detected and
 * completed, %false otherwise.
 */
bool mem_cgroup_oom_synchronize(bool handle)
{
	struct mem_cgroup *memcg = current->memcg_oom.memcg;
	struct oom_wait_info owait;
	bool locked;

	/* OOM is global, do not handle */
	if (!memcg)
		return false;

	if (!handle)
		goto cleanup;

	owait.memcg = memcg;
	owait.wait.flags = 0;
	owait.wait.func = memcg_oom_wake_function;
	owait.wait.private = current;
	INIT_LIST_HEAD(&owait.wait.task_list);

	prepare_to_wait(&memcg_oom_waitq, &owait.wait, TASK_KILLABLE);
	mem_cgroup_mark_under_oom(memcg);

	locked = mem_cgroup_oom_trylock(memcg);

	if (locked)
		mem_cgroup_oom_notify(memcg);

	if (locked && !memcg->oom_kill_disable) {
		mem_cgroup_unmark_under_oom(memcg);
		finish_wait(&memcg_oom_waitq, &owait.wait);
		mem_cgroup_out_of_memory(memcg, current->memcg_oom.gfp_mask,
					 current->memcg_oom.order);
	} else {
		schedule();
		mem_cgroup_unmark_under_oom(memcg);
		finish_wait(&memcg_oom_waitq, &owait.wait);
	}

	if (locked) {
		mem_cgroup_oom_unlock(memcg);
		/*
		 * There is no guarantee that an OOM-lock contender
		 * sees the wakeups triggered by the OOM kill
		 * uncharges.  Wake any sleepers explicitely.
		 */
		memcg_oom_recover(memcg);
	}
cleanup:
	current->memcg_oom.memcg = NULL;
	css_put(&memcg->css);
	return true;
}

/*
 * Currently used to update mapped file statistics, but the routine can be
 * generalized to update other statistics as well.
 *
 * Notes: Race condition
 *
 * We usually use page_cgroup_lock() for accessing page_cgroup member but
 * it tends to be costly. But considering some conditions, we doesn't need
 * to do so _always_.
 *
 * Considering "charge", lock_page_cgroup() is not required because all
 * file-stat operations happen after a page is attached to radix-tree. There
 * are no race with "charge".
 *
 * Considering "uncharge", we know that memcg doesn't clear pc->mem_cgroup
 * at "uncharge" intentionally. So, we always see valid pc->mem_cgroup even
 * if there are race with "uncharge". Statistics itself is properly handled
 * by flags.
 *
 * Considering "move", this is an only case we see a race. To make the race
 * small, we check mm->moving_account and detect there are possibility of race
 * If there is, we take a lock.
 */

void __mem_cgroup_begin_update_page_stat(struct page *page,
				bool *locked, unsigned long *flags)
{
	struct mem_cgroup *memcg;
	struct page_cgroup *pc;

	pc = lookup_page_cgroup(page);
again:
	memcg = pc->mem_cgroup;
	if (unlikely(!memcg || !PageCgroupUsed(pc)))
		return;
	/*
	 * If this memory cgroup is not under account moving, we don't
	 * need to take move_lock_mem_cgroup(). Because we already hold
	 * rcu_read_lock(), any calls to move_account will be delayed until
	 * rcu_read_unlock() if mem_cgroup_stolen() == true.
	 */
	if (!mem_cgroup_stolen(memcg))
		return;

	move_lock_mem_cgroup(memcg, flags);
	if (memcg != pc->mem_cgroup || !PageCgroupUsed(pc)) {
		move_unlock_mem_cgroup(memcg, flags);
		goto again;
	}
	*locked = true;
}

void __mem_cgroup_end_update_page_stat(struct page *page, unsigned long *flags)
{
	struct page_cgroup *pc = lookup_page_cgroup(page);

	/*
	 * It's guaranteed that pc->mem_cgroup never changes while
	 * lock is held because a routine modifies pc->mem_cgroup
	 * should take move_lock_mem_cgroup().
	 */
	move_unlock_mem_cgroup(pc->mem_cgroup, flags);
}

void mem_cgroup_update_page_stat(struct page *page,
				 enum mem_cgroup_stat_index idx, int val)
{
	struct mem_cgroup *memcg;
	struct page_cgroup *pc = lookup_page_cgroup(page);
	unsigned long uninitialized_var(flags);

	if (mem_cgroup_disabled())
		return;

	VM_BUG_ON(!rcu_read_lock_held());
	memcg = pc->mem_cgroup;
	if (unlikely(!memcg || !PageCgroupUsed(pc)))
		return;

	this_cpu_add(memcg->stat->count[idx], val);
}

/*
 * size of first charge trial. "32" comes from vmscan.c's magic value.
 * TODO: maybe necessary to use big numbers in big irons.
 */
#define CHARGE_BATCH	32U
struct memcg_stock_pcp {
	struct mem_cgroup *cached; /* this never be root cgroup */
	unsigned int nr_pages;
	struct work_struct work;
	unsigned long flags;
#define FLUSHING_CACHED_CHARGE	0
};
static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock);
static DEFINE_MUTEX(percpu_charge_mutex);

/**
 * consume_stock: Try to consume stocked charge on this cpu.
 * @memcg: memcg to consume from.
 * @nr_pages: how many pages to charge.
 *
 * The charges will only happen if @memcg matches the current cpu's memcg
 * stock, and at least @nr_pages are available in that stock.  Failure to
 * service an allocation will refill the stock.
 *
 * returns true if successful, false otherwise.
 */
static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
{
	struct memcg_stock_pcp *stock;
	bool ret = true;

	if (nr_pages > CHARGE_BATCH)
		return false;

	stock = &get_cpu_var(memcg_stock);
	if (memcg == stock->cached && stock->nr_pages >= nr_pages)
		stock->nr_pages -= nr_pages;
	else /* need to call res_counter_charge */
		ret = false;
	put_cpu_var(memcg_stock);
	return ret;
}

/*
 * Returns stocks cached in percpu to res_counter and reset cached information.
 */
static void drain_stock(struct memcg_stock_pcp *stock)
{
	struct mem_cgroup *old = stock->cached;

	if (stock->nr_pages) {
		unsigned long bytes = stock->nr_pages * PAGE_SIZE;

		res_counter_uncharge(&old->res, bytes);
		if (do_swap_account)
			res_counter_uncharge(&old->memsw, bytes);
		stock->nr_pages = 0;
	}
	stock->cached = NULL;
}

/*
 * This must be called under preempt disabled or must be called by
 * a thread which is pinned to local cpu.
 */
static void drain_local_stock(struct work_struct *dummy)
{
	struct memcg_stock_pcp *stock = &__get_cpu_var(memcg_stock);
	drain_stock(stock);
	clear_bit(FLUSHING_CACHED_CHARGE, &stock->flags);
}

static void __init memcg_stock_init(void)
{
	int cpu;

	for_each_possible_cpu(cpu) {
		struct memcg_stock_pcp *stock =
					&per_cpu(memcg_stock, cpu);
		INIT_WORK(&stock->work, drain_local_stock);
	}
}

/*
 * Cache charges(val) which is from res_counter, to local per_cpu area.
 * This will be consumed by consume_stock() function, later.
 */
static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
{
	struct memcg_stock_pcp *stock = &get_cpu_var(memcg_stock);

	if (stock->cached != memcg) { /* reset if necessary */
		drain_stock(stock);
		stock->cached = memcg;
	}
	stock->nr_pages += nr_pages;
	put_cpu_var(memcg_stock);
}

/*
 * Drains all per-CPU charge caches for given root_memcg resp. subtree
 * of the hierarchy under it. sync flag says whether we should block
 * until the work is done.
 */
static void drain_all_stock(struct mem_cgroup *root_memcg, bool sync)
{
	int cpu, curcpu;

	/* Notify other cpus that system-wide "drain" is running */
	get_online_cpus();
	curcpu = get_cpu();
	for_each_online_cpu(cpu) {
		struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu);
		struct mem_cgroup *memcg;

		memcg = stock->cached;
		if (!memcg || !stock->nr_pages)
			continue;
		if (!mem_cgroup_same_or_subtree(root_memcg, memcg))
			continue;
		if (!test_and_set_bit(FLUSHING_CACHED_CHARGE, &stock->flags)) {
			if (cpu == curcpu)
				drain_local_stock(&stock->work);
			else
				schedule_work_on(cpu, &stock->work);
		}
	}
	put_cpu();

	if (!sync)
		goto out;

	for_each_online_cpu(cpu) {
		struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu);
		if (test_bit(FLUSHING_CACHED_CHARGE, &stock->flags))
			flush_work(&stock->work);
	}
out:
	put_online_cpus();
}

/*
 * Tries to drain stocked charges in other cpus. This function is asynchronous
 * and just put a work per cpu for draining localy on each cpu. Caller can
 * expects some charges will be back to res_counter later but cannot wait for
 * it.
 */
static void drain_all_stock_async(struct mem_cgroup *root_memcg)
{
	/*
	 * If someone calls draining, avoid adding more kworker runs.
	 */
	if (!mutex_trylock(&percpu_charge_mutex))
		return;
	drain_all_stock(root_memcg, false);
	mutex_unlock(&percpu_charge_mutex);
}

/* This is a synchronous drain interface. */
static void drain_all_stock_sync(struct mem_cgroup *root_memcg)
{
	/* called when force_empty is called */
	mutex_lock(&percpu_charge_mutex);
	drain_all_stock(root_memcg, true);
	mutex_unlock(&percpu_charge_mutex);
}

/*
 * This function drains percpu counter value from DEAD cpu and
 * move it to local cpu. Note that this function can be preempted.
 */
static void mem_cgroup_drain_pcp_counter(struct mem_cgroup *memcg, int cpu)
{
	int i;

	spin_lock(&memcg->pcp_counter_lock);
	for (i = 0; i < MEM_CGROUP_STAT_NSTATS; i++) {
		long x = per_cpu(memcg->stat->count[i], cpu);

		per_cpu(memcg->stat->count[i], cpu) = 0;
		memcg->nocpu_base.count[i] += x;
	}
	for (i = 0; i < MEM_CGROUP_EVENTS_NSTATS; i++) {
		unsigned long x = per_cpu(memcg->stat->events[i], cpu);

		per_cpu(memcg->stat->events[i], cpu) = 0;
		memcg->nocpu_base.events[i] += x;
	}
	spin_unlock(&memcg->pcp_counter_lock);
}

static int memcg_cpu_hotplug_callback(struct notifier_block *nb,
					unsigned long action,
					void *hcpu)
{
	int cpu = (unsigned long)hcpu;
	struct memcg_stock_pcp *stock;
	struct mem_cgroup *iter;

	if (action == CPU_ONLINE)
		return NOTIFY_OK;

	if (action != CPU_DEAD && action != CPU_DEAD_FROZEN)
		return NOTIFY_OK;

	for_each_mem_cgroup(iter)
		mem_cgroup_drain_pcp_counter(iter, cpu);

	stock = &per_cpu(memcg_stock, cpu);
	drain_stock(stock);
	return NOTIFY_OK;
}


/* See __mem_cgroup_try_charge() for details */
enum {
	CHARGE_OK,		/* success */
	CHARGE_RETRY,		/* need to retry but retry is not bad */
	CHARGE_NOMEM,		/* we can't do more. return -ENOMEM */
	CHARGE_WOULDBLOCK,	/* GFP_WAIT wasn't set and no enough res. */
};

static int mem_cgroup_do_charge(struct mem_cgroup *memcg, gfp_t gfp_mask,
				unsigned int nr_pages, unsigned int min_pages,
				bool invoke_oom)
{
	unsigned long csize = nr_pages * PAGE_SIZE;
	struct mem_cgroup *mem_over_limit;
	struct res_counter *fail_res;
	unsigned long flags = 0;
	int ret;

	ret = res_counter_charge(&memcg->res, csize, &fail_res);

	if (likely(!ret)) {
		if (!do_swap_account)
			return CHARGE_OK;
		ret = res_counter_charge(&memcg->memsw, csize, &fail_res);
		if (likely(!ret))
			return CHARGE_OK;

		res_counter_uncharge(&memcg->res, csize);
		mem_over_limit = mem_cgroup_from_res_counter(fail_res, memsw);
		flags |= MEM_CGROUP_RECLAIM_NOSWAP;
	} else
		mem_over_limit = mem_cgroup_from_res_counter(fail_res, res);
	/*
	 * Never reclaim on behalf of optional batching, retry with a
	 * single page instead.
	 */
	if (nr_pages > min_pages)
		return CHARGE_RETRY;

	if (!(gfp_mask & __GFP_WAIT))
		return CHARGE_WOULDBLOCK;

	if (gfp_mask & __GFP_NORETRY)
		return CHARGE_NOMEM;

	ret = mem_cgroup_reclaim(mem_over_limit, gfp_mask, flags);
	if (mem_cgroup_margin(mem_over_limit) >= nr_pages)
		return CHARGE_RETRY;
	/*
	 * Even though the limit is exceeded at this point, reclaim
	 * may have been able to free some pages.  Retry the charge
	 * before killing the task.
	 *
	 * Only for regular pages, though: huge pages are rather
	 * unlikely to succeed so close to the limit, and we fall back
	 * to regular pages anyway in case of failure.
	 */
	if (nr_pages <= (1 << PAGE_ALLOC_COSTLY_ORDER) && ret)
		return CHARGE_RETRY;

	/*
	 * At task move, charge accounts can be doubly counted. So, it's
	 * better to wait until the end of task_move if something is going on.
	 */
	if (mem_cgroup_wait_acct_move(mem_over_limit))
		return CHARGE_RETRY;

	if (invoke_oom)
		mem_cgroup_oom(mem_over_limit, gfp_mask, get_order(csize));

	return CHARGE_NOMEM;
}

/*
 * __mem_cgroup_try_charge() does
 * 1. detect memcg to be charged against from passed *mm and *ptr,
 * 2. update res_counter
 * 3. call memory reclaim if necessary.
 *
 * In some special case, if the task is fatal, fatal_signal_pending() or
 * has TIF_MEMDIE, this function returns -EINTR while writing root_mem_cgroup
 * to *ptr. There are two reasons for this. 1: fatal threads should quit as soon
 * as possible without any hazards. 2: all pages should have a valid
 * pc->mem_cgroup. If mm is NULL and the caller doesn't pass a valid memcg
 * pointer, that is treated as a charge to root_mem_cgroup.
 *
 * So __mem_cgroup_try_charge() will return
 *  0       ...  on success, filling *ptr with a valid memcg pointer.
 *  -ENOMEM ...  charge failure because of resource limits.
 *  -EINTR  ...  if thread is fatal. *ptr is filled with root_mem_cgroup.
 *
 * Unlike the exported interface, an "oom" parameter is added. if oom==true,
 * the oom-killer can be invoked.
 */
static int __mem_cgroup_try_charge(struct mm_struct *mm,
				   gfp_t gfp_mask,
				   unsigned int nr_pages,
				   struct mem_cgroup **ptr,
				   bool oom)
{
	unsigned int batch = max(CHARGE_BATCH, nr_pages);
	int nr_oom_retries = MEM_CGROUP_RECLAIM_RETRIES;
	struct mem_cgroup *memcg = NULL;
	int ret;

	/*
	 * Unlike gloval-vm's OOM-kill, we're not in memory shortage
	 * in system level. So, allow to go ahead dying process in addition to
	 * MEMDIE process.
	 */
	if (unlikely(test_thread_flag(TIF_MEMDIE)
		     || fatal_signal_pending(current)))
		goto bypass;

	if (unlikely(task_in_memcg_oom(current)))
		goto nomem;

	if (gfp_mask & __GFP_NOFAIL)
		oom = false;

	/*
	 * We always charge the cgroup the mm_struct belongs to.
	 * The mm_struct's mem_cgroup changes on task migration if the
	 * thread group leader migrates. It's possible that mm is not
	 * set, if so charge the root memcg (happens for pagecache usage).
	 */
	if (!*ptr && !mm)
		*ptr = root_mem_cgroup;
again:
	if (*ptr) { /* css should be a valid one */
		memcg = *ptr;
		if (mem_cgroup_is_root(memcg))
			goto done;
		if (consume_stock(memcg, nr_pages))
			goto done;
		css_get(&memcg->css);
	} else {
		struct task_struct *p;

		rcu_read_lock();
		p = rcu_dereference(mm->owner);
		/*
		 * Because we don't have task_lock(), "p" can exit.
		 * In that case, "memcg" can point to root or p can be NULL with
		 * race with swapoff. Then, we have small risk of mis-accouning.
		 * But such kind of mis-account by race always happens because
		 * we don't have cgroup_mutex(). It's overkill and we allo that
		 * small race, here.
		 * (*) swapoff at el will charge against mm-struct not against
		 * task-struct. So, mm->owner can be NULL.
		 */
		memcg = mem_cgroup_from_task(p);
		if (!memcg)
			memcg = root_mem_cgroup;
		if (mem_cgroup_is_root(memcg)) {
			rcu_read_unlock();
			goto done;
		}
		if (consume_stock(memcg, nr_pages)) {
			/*
			 * It seems dagerous to access memcg without css_get().
			 * But considering how consume_stok works, it's not
			 * necessary. If consume_stock success, some charges
			 * from this memcg are cached on this cpu. So, we
			 * don't need to call css_get()/css_tryget() before
			 * calling consume_stock().
			 */
			rcu_read_unlock();
			goto done;
		}
		/* after here, we may be blocked. we need to get refcnt */
		if (!css_tryget(&memcg->css)) {
			rcu_read_unlock();
			goto again;
		}
		rcu_read_unlock();
	}

	do {
		bool invoke_oom = oom && !nr_oom_retries;

		/* If killed, bypass charge */
		if (fatal_signal_pending(current)) {
			css_put(&memcg->css);
			goto bypass;
		}

		ret = mem_cgroup_do_charge(memcg, gfp_mask, batch,
					   nr_pages, invoke_oom);
		switch (ret) {
		case CHARGE_OK:
			break;
		case CHARGE_RETRY: /* not in OOM situation but retry */
			batch = nr_pages;
			css_put(&memcg->css);
			memcg = NULL;
			goto again;
		case CHARGE_WOULDBLOCK: /* !__GFP_WAIT */
			css_put(&memcg->css);
			goto nomem;
		case CHARGE_NOMEM: /* OOM routine works */
			if (!oom || invoke_oom) {
				css_put(&memcg->css);
				goto nomem;
			}
			nr_oom_retries--;
			break;
		}
	} while (ret != CHARGE_OK);

	if (batch > nr_pages)
		refill_stock(memcg, batch - nr_pages);
	css_put(&memcg->css);
done:
	*ptr = memcg;
	return 0;
nomem:
	if (!(gfp_mask & __GFP_NOFAIL)) {
		*ptr = NULL;
		return -ENOMEM;
	}
bypass:
	*ptr = root_mem_cgroup;
	return -EINTR;
}

/*
 * Somemtimes we have to undo a charge we got by try_charge().
 * This function is for that and do uncharge, put css's refcnt.
 * gotten by try_charge().
 */
static void __mem_cgroup_cancel_charge(struct mem_cgroup *memcg,
				       unsigned int nr_pages)
{
	if (!mem_cgroup_is_root(memcg)) {
		unsigned long bytes = nr_pages * PAGE_SIZE;

		res_counter_uncharge(&memcg->res, bytes);
		if (do_swap_account)
			res_counter_uncharge(&memcg->memsw, bytes);
	}
}

/*
 * Cancel chrages in this cgroup....doesn't propagate to parent cgroup.
 * This is useful when moving usage to parent cgroup.
 */
static void __mem_cgroup_cancel_local_charge(struct mem_cgroup *memcg,
					unsigned int nr_pages)
{
	unsigned long bytes = nr_pages * PAGE_SIZE;

	if (mem_cgroup_is_root(memcg))
		return;

	res_counter_uncharge_until(&memcg->res, memcg->res.parent, bytes);
	if (do_swap_account)
		res_counter_uncharge_until(&memcg->memsw,
						memcg->memsw.parent, bytes);
}

/*
 * A helper function to get mem_cgroup from ID. must be called under
 * rcu_read_lock().  The caller is responsible for calling css_tryget if
 * the mem_cgroup is used for charging. (dropping refcnt from swap can be
 * called against removed memcg.)
 */
static struct mem_cgroup *mem_cgroup_lookup(unsigned short id)
{
	/* ID 0 is unused ID */
	if (!id)
		return NULL;
	return mem_cgroup_from_id(id);
}

struct mem_cgroup *try_get_mem_cgroup_from_page(struct page *page)
{
	struct mem_cgroup *memcg = NULL;
	struct page_cgroup *pc;
	unsigned short id;
	swp_entry_t ent;

	VM_BUG_ON_PAGE(!PageLocked(page), page);

	pc = lookup_page_cgroup(page);
	lock_page_cgroup(pc);
	if (PageCgroupUsed(pc)) {
		memcg = pc->mem_cgroup;
		if (memcg && !css_tryget(&memcg->css))
			memcg = NULL;
	} else if (PageSwapCache(page)) {
		ent.val = page_private(page);
		id = lookup_swap_cgroup_id(ent);
		rcu_read_lock();
		memcg = mem_cgroup_lookup(id);
		if (memcg && !css_tryget(&memcg->css))
			memcg = NULL;
		rcu_read_unlock();
	}
	unlock_page_cgroup(pc);
	return memcg;
}

static void __mem_cgroup_commit_charge(struct mem_cgroup *memcg,
				       struct page *page,
				       unsigned int nr_pages,
				       enum charge_type ctype,
				       bool lrucare)
{
	struct page_cgroup *pc = lookup_page_cgroup(page);
	struct zone *uninitialized_var(zone);
	struct lruvec *lruvec;
	bool was_on_lru = false;
	bool anon;

	lock_page_cgroup(pc);
	VM_BUG_ON_PAGE(PageCgroupUsed(pc), page);
	/*
	 * we don't need page_cgroup_lock about tail pages, becase they are not
	 * accessed by any other context at this point.
	 */

	/*
	 * In some cases, SwapCache and FUSE(splice_buf->radixtree), the page
	 * may already be on some other mem_cgroup's LRU.  Take care of it.
	 */
	if (lrucare) {
		zone = page_zone(page);
		spin_lock_irq(&zone->lru_lock);
		if (PageLRU(page)) {
			lruvec = mem_cgroup_zone_lruvec(zone, pc->mem_cgroup);
			ClearPageLRU(page);
			del_page_from_lru_list(page, lruvec, page_lru(page));
			was_on_lru = true;
		}
	}

	pc->mem_cgroup = memcg;
	/*
	 * We access a page_cgroup asynchronously without lock_page_cgroup().
	 * Especially when a page_cgroup is taken from a page, pc->mem_cgroup
	 * is accessed after testing USED bit. To make pc->mem_cgroup visible
	 * before USED bit, we need memory barrier here.
	 * See mem_cgroup_add_lru_list(), etc.
	 */
	smp_wmb();
	SetPageCgroupUsed(pc);

	if (lrucare) {
		if (was_on_lru) {
			lruvec = mem_cgroup_zone_lruvec(zone, pc->mem_cgroup);
			VM_BUG_ON_PAGE(PageLRU(page), page);
			SetPageLRU(page);
			add_page_to_lru_list(page, lruvec, page_lru(page));
		}
		spin_unlock_irq(&zone->lru_lock);
	}

	if (ctype == MEM_CGROUP_CHARGE_TYPE_ANON)
		anon = true;
	else
		anon = false;

	mem_cgroup_charge_statistics(memcg, page, anon, nr_pages);
	unlock_page_cgroup(pc);

	/*
	 * "charge_statistics" updated event counter. Then, check it.
	 * Insert ancestor (and ancestor's ancestors), to softlimit RB-tree.
	 * if they exceeds softlimit.
	 */
	memcg_check_events(memcg, page);
}

static DEFINE_MUTEX(set_limit_mutex);