[linux-2.6-block.git] / block / blk-cgroup.h

#ifndef _BLK_CGROUP_H
#define _BLK_CGROUP_H
/*
 * Common Block IO controller cgroup interface
 *
 * Based on ideas and code from CFQ, CFS and BFQ:
 * Copyright (C) 2003 Jens Axboe <axboe@kernel.dk>
 *
 * Copyright (C) 2008 Fabio Checconi <fabio@gandalf.sssup.it>
 *		      Paolo Valente <paolo.valente@unimore.it>
 *
 * Copyright (C) 2009 Vivek Goyal <vgoyal@redhat.com>
 * 	              Nauman Rafique <nauman@google.com>
 */

#include <linux/cgroup.h>
#include <linux/u64_stats_sync.h>
#include <linux/seq_file.h>
#include <linux/radix-tree.h>
#include <linux/blkdev.h>

/* Max limits for throttle policy */
#define THROTL_IOPS_MAX		UINT_MAX

/* CFQ specific, out here for blkcg->cfq_weight */
#define CFQ_WEIGHT_MIN		10
#define CFQ_WEIGHT_MAX		1000
#define CFQ_WEIGHT_DEFAULT	500

#ifdef CONFIG_BLK_CGROUP

enum blkg_rwstat_type {
	BLKG_RWSTAT_READ,
	BLKG_RWSTAT_WRITE,
	BLKG_RWSTAT_SYNC,
	BLKG_RWSTAT_ASYNC,

	BLKG_RWSTAT_NR,
	BLKG_RWSTAT_TOTAL = BLKG_RWSTAT_NR,
};

struct blkcg_gq;

struct blkcg {
	struct cgroup_subsys_state	css;
	spinlock_t			lock;

	struct radix_tree_root		blkg_tree;
	struct blkcg_gq			*blkg_hint;
	struct hlist_head		blkg_list;

	/* for policies to test whether associated blkcg has changed */
	uint64_t			id;

	/* TODO: per-policy storage in blkcg */
	unsigned int			cfq_weight;	/* belongs to cfq */
	unsigned int			cfq_leaf_weight;
};

struct blkg_stat {
	struct u64_stats_sync		syncp;
	uint64_t			cnt;
};

struct blkg_rwstat {
	struct u64_stats_sync		syncp;
	uint64_t			cnt[BLKG_RWSTAT_NR];
};

/*
 * A blkcg_gq (blkg) is association between a block cgroup (blkcg) and a
 * request_queue (q).  This is used by blkcg policies which need to track
 * information per blkcg - q pair.
 *
 * There can be multiple active blkcg policies and each has its private
 * data on each blkg, the size of which is determined by
 * blkcg_policy->pd_size.  blkcg core allocates and frees such areas
 * together with blkg and invokes pd_init/exit_fn() methods.
 *
 * Such private data must embed struct blkg_policy_data (pd) at the
 * beginning and pd_size can't be smaller than pd.
 */
struct blkg_policy_data {
	/* the blkg and policy id this per-policy data belongs to */
	struct blkcg_gq			*blkg;
	int				plid;

	/* used during policy activation */
	struct list_head		alloc_node;
};

/* association between a blk cgroup and a request queue */
struct blkcg_gq {
	/* Pointer to the associated request_queue */
	struct request_queue		*q;
	struct list_head		q_node;
	struct hlist_node		blkcg_node;
	struct blkcg			*blkcg;

	/* all non-root blkcg_gq's are guaranteed to have access to parent */
	struct blkcg_gq			*parent;

	/* request allocation list for this blkcg-q pair */
	struct request_list		rl;

	/* reference count */
	int				refcnt;

	/* is this blkg online? protected by both blkcg and q locks */
	bool				online;

	struct blkg_policy_data		*pd[BLKCG_MAX_POLS];

	struct rcu_head			rcu_head;
};

typedef void (blkcg_pol_init_pd_fn)(struct blkcg_gq *blkg);
typedef void (blkcg_pol_online_pd_fn)(struct blkcg_gq *blkg);
typedef void (blkcg_pol_offline_pd_fn)(struct blkcg_gq *blkg);
typedef void (blkcg_pol_exit_pd_fn)(struct blkcg_gq *blkg);
typedef void (blkcg_pol_reset_pd_stats_fn)(struct blkcg_gq *blkg);

struct blkcg_policy {
	int				plid;
	/* policy specific private data size */
	size_t				pd_size;
	/* cgroup files for the policy */
	struct cftype			*cftypes;

	/* operations */
	blkcg_pol_init_pd_fn		*pd_init_fn;
	blkcg_pol_online_pd_fn		*pd_online_fn;
	blkcg_pol_offline_pd_fn		*pd_offline_fn;
	blkcg_pol_exit_pd_fn		*pd_exit_fn;
	blkcg_pol_reset_pd_stats_fn	*pd_reset_stats_fn;
};

extern struct blkcg blkcg_root;

struct blkcg_gq *blkg_lookup(struct blkcg *blkcg, struct request_queue *q);
struct blkcg_gq *blkg_lookup_create(struct blkcg *blkcg,
				    struct request_queue *q);
int blkcg_init_queue(struct request_queue *q);
void blkcg_drain_queue(struct request_queue *q);
void blkcg_exit_queue(struct request_queue *q);

/* Blkio controller policy registration */
int blkcg_policy_register(struct blkcg_policy *pol);
void blkcg_policy_unregister(struct blkcg_policy *pol);
int blkcg_activate_policy(struct request_queue *q,
			  const struct blkcg_policy *pol);
void blkcg_deactivate_policy(struct request_queue *q,
			     const struct blkcg_policy *pol);

void blkcg_print_blkgs(struct seq_file *sf, struct blkcg *blkcg,
		       u64 (*prfill)(struct seq_file *,
				     struct blkg_policy_data *, int),
		       const struct blkcg_policy *pol, int data,
		       bool show_total);
u64 __blkg_prfill_u64(struct seq_file *sf, struct blkg_policy_data *pd, u64 v);
u64 __blkg_prfill_rwstat(struct seq_file *sf, struct blkg_policy_data *pd,
			 const struct blkg_rwstat *rwstat);
u64 blkg_prfill_stat(struct seq_file *sf, struct blkg_policy_data *pd, int off);
u64 blkg_prfill_rwstat(struct seq_file *sf, struct blkg_policy_data *pd,
		       int off);

u64 blkg_stat_recursive_sum(struct blkg_policy_data *pd, int off);
struct blkg_rwstat blkg_rwstat_recursive_sum(struct blkg_policy_data *pd,
					     int off);

struct blkg_conf_ctx {
	struct gendisk			*disk;
	struct blkcg_gq			*blkg;
	u64				v;
};

int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol,
		   const char *input, struct blkg_conf_ctx *ctx);
void blkg_conf_finish(struct blkg_conf_ctx *ctx);


static inline struct blkcg *css_to_blkcg(struct cgroup_subsys_state *css)
{
	return css ? container_of(css, struct blkcg, css) : NULL;
}

static inline struct blkcg *task_blkcg(struct task_struct *tsk)
{
	return css_to_blkcg(task_css(tsk, blkio_subsys_id));
}

static inline struct blkcg *bio_blkcg(struct bio *bio)
{
	if (bio && bio->bi_css)
		return css_to_blkcg(bio->bi_css);
	return task_blkcg(current);
}

/**
 * blkcg_parent - get the parent of a blkcg
 * @blkcg: blkcg of interest
 *
 * Return the parent blkcg of @blkcg.  Can be called anytime.
 */
static inline struct blkcg *blkcg_parent(struct blkcg *blkcg)
{
	return css_to_blkcg(css_parent(&blkcg->css));
}

/**
 * blkg_to_pdata - get policy private data
 * @blkg: blkg of interest
 * @pol: policy of interest
 *
 * Return pointer to private data associated with the @blkg-@pol pair.
 */
static inline struct blkg_policy_data *blkg_to_pd(struct blkcg_gq *blkg,
						  struct blkcg_policy *pol)
{
	return blkg ? blkg->pd[pol->plid] : NULL;
}

/**
 * pdata_to_blkg - get blkg associated with policy private data
 * @pd: policy private data of interest
 *
 * @pd is policy private data.  Determine the blkg it's associated with.
 */
static inline struct blkcg_gq *pd_to_blkg(struct blkg_policy_data *pd)
{
	return pd ? pd->blkg : NULL;
}

/**
 * blkg_path - format cgroup path of blkg
 * @blkg: blkg of interest
 * @buf: target buffer
 * @buflen: target buffer length
 *
 * Format the path of the cgroup of @blkg into @buf.
 */
static inline int blkg_path(struct blkcg_gq *blkg, char *buf, int buflen)
{
	int ret;

	ret = cgroup_path(blkg->blkcg->css.cgroup, buf, buflen);
	if (ret)
		strncpy(buf, "<unavailable>", buflen);
	return ret;
}

/**
 * blkg_get - get a blkg reference
 * @blkg: blkg to get
 *
 * The caller should be holding queue_lock and an existing reference.
 */
static inline void blkg_get(struct blkcg_gq *blkg)
{
	lockdep_assert_held(blkg->q->queue_lock);
	WARN_ON_ONCE(!blkg->refcnt);
	blkg->refcnt++;
}

void __blkg_release_rcu(struct rcu_head *rcu);

/**
 * blkg_put - put a blkg reference
 * @blkg: blkg to put
 *
 * The caller should be holding queue_lock.
 */
static inline void blkg_put(struct blkcg_gq *blkg)
{
	lockdep_assert_held(blkg->q->queue_lock);
	WARN_ON_ONCE(blkg->refcnt <= 0);
	if (!--blkg->refcnt)
		call_rcu(&blkg->rcu_head, __blkg_release_rcu);
}

struct blkcg_gq *__blkg_lookup(struct blkcg *blkcg, struct request_queue *q,
			       bool update_hint);

/**
 * blkg_for_each_descendant_pre - pre-order walk of a blkg's descendants
 * @d_blkg: loop cursor pointing to the current descendant
 * @pos_css: used for iteration
 * @p_blkg: target blkg to walk descendants of
 *
 * Walk @c_blkg through the descendants of @p_blkg.  Must be used with RCU
 * read locked.  If called under either blkcg or queue lock, the iteration
 * is guaranteed to include all and only online blkgs.  The caller may
 * update @pos_css by calling css_rightmost_descendant() to skip subtree.
 * @p_blkg is included in the iteration and the first node to be visited.
 */
#define blkg_for_each_descendant_pre(d_blkg, pos_css, p_blkg)		\
	css_for_each_descendant_pre((pos_css), &(p_blkg)->blkcg->css)	\
		if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css),	\
					      (p_blkg)->q, false)))

/**
 * blkg_for_each_descendant_post - post-order walk of a blkg's descendants
 * @d_blkg: loop cursor pointing to the current descendant
 * @pos_css: used for iteration
 * @p_blkg: target blkg to walk descendants of
 *
 * Similar to blkg_for_each_descendant_pre() but performs post-order
 * traversal instead.  Synchronization rules are the same.  @p_blkg is
 * included in the iteration and the last node to be visited.
 */
#define blkg_for_each_descendant_post(d_blkg, pos_css, p_blkg)		\
	css_for_each_descendant_post((pos_css), &(p_blkg)->blkcg->css)	\
		if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css),	\
					      (p_blkg)->q, false)))

/**
 * blk_get_rl - get request_list to use
 * @q: request_queue of interest
 * @bio: bio which will be attached to the allocated request (may be %NULL)
 *
 * The caller wants to allocate a request from @q to use for @bio.  Find
 * the request_list to use and obtain a reference on it.  Should be called
 * under queue_lock.  This function is guaranteed to return non-%NULL
 * request_list.
 */
static inline struct request_list *blk_get_rl(struct request_queue *q,
					      struct bio *bio)
{
	struct blkcg *blkcg;
	struct blkcg_gq *blkg;

	rcu_read_lock();

	blkcg = bio_blkcg(bio);

	/* bypass blkg lookup and use @q->root_rl directly for root */
	if (blkcg == &blkcg_root)
		goto root_rl;

	/*
	 * Try to use blkg->rl.  blkg lookup may fail under memory pressure
	 * or if either the blkcg or queue is going away.  Fall back to
	 * root_rl in such cases.
	 */
	blkg = blkg_lookup_create(blkcg, q);
	if (unlikely(IS_ERR(blkg)))
		goto root_rl;

	blkg_get(blkg);
	rcu_read_unlock();
	return &blkg->rl;
root_rl:
	rcu_read_unlock();
	return &q->root_rl;
}

/**
 * blk_put_rl - put request_list
 * @rl: request_list to put
 *
 * Put the reference acquired by blk_get_rl().  Should be called under
 * queue_lock.
 */
static inline void blk_put_rl(struct request_list *rl)
{
	/* root_rl may not have blkg set */
	if (rl->blkg && rl->blkg->blkcg != &blkcg_root)
		blkg_put(rl->blkg);
}

/**
 * blk_rq_set_rl - associate a request with a request_list
 * @rq: request of interest
 * @rl: target request_list
 *
 * Associate @rq with @rl so that accounting and freeing can know the
 * request_list @rq came from.
 */
static inline void blk_rq_set_rl(struct request *rq, struct request_list *rl)
{
	rq->rl = rl;
}

/**
 * blk_rq_rl - return the request_list a request came from
 * @rq: request of interest
 *
 * Return the request_list @rq is allocated from.
 */
static inline struct request_list *blk_rq_rl(struct request *rq)
{
	return rq->rl;
}

struct request_list *__blk_queue_next_rl(struct request_list *rl,
					 struct request_queue *q);
/**
 * blk_queue_for_each_rl - iterate through all request_lists of a request_queue
 *
 * Should be used under queue_lock.
 */
#define blk_queue_for_each_rl(rl, q)	\
	for ((rl) = &(q)->root_rl; (rl); (rl) = __blk_queue_next_rl((rl), (q)))

static inline void blkg_stat_init(struct blkg_stat *stat)
{
	u64_stats_init(&stat->syncp);
}

/**
 * blkg_stat_add - add a value to a blkg_stat
 * @stat: target blkg_stat
 * @val: value to add
 *
 * Add @val to @stat.  The caller is responsible for synchronizing calls to
 * this function.
 */
static inline void blkg_stat_add(struct blkg_stat *stat, uint64_t val)
{
	u64_stats_update_begin(&stat->syncp);
	stat->cnt += val;
	u64_stats_update_end(&stat->syncp);
}

/**
 * blkg_stat_read - read the current value of a blkg_stat
 * @stat: blkg_stat to read
 *
 * Read the current value of @stat.  This function can be called without
 * synchroniztion and takes care of u64 atomicity.
 */
static inline uint64_t blkg_stat_read(struct blkg_stat *stat)
{
	unsigned int start;
	uint64_t v;

	do {
		start = u64_stats_fetch_begin_bh(&stat->syncp);
		v = stat->cnt;
	} while (u64_stats_fetch_retry_bh(&stat->syncp, start));

	return v;
}

/**
 * blkg_stat_reset - reset a blkg_stat
 * @stat: blkg_stat to reset
 */
static inline void blkg_stat_reset(struct blkg_stat *stat)
{
	stat->cnt = 0;
}

/**
 * blkg_stat_merge - merge a blkg_stat into another
 * @to: the destination blkg_stat
 * @from: the source
 *
 * Add @from's count to @to.
 */
static inline void blkg_stat_merge(struct blkg_stat *to, struct blkg_stat *from)
{
	blkg_stat_add(to, blkg_stat_read(from));
}

static inline void blkg_rwstat_init(struct blkg_rwstat *rwstat)
{
	u64_stats_init(&rwstat->syncp);
}

/**
 * blkg_rwstat_add - add a value to a blkg_rwstat
 * @rwstat: target blkg_rwstat
 * @rw: mask of REQ_{WRITE|SYNC}
 * @val: value to add
 *
 * Add @val to @rwstat.  The counters are chosen according to @rw.  The
 * caller is responsible for synchronizing calls to this function.
 */
static inline void blkg_rwstat_add(struct blkg_rwstat *rwstat,
				   int rw, uint64_t val)
{
	u64_stats_update_begin(&rwstat->syncp);

	if (rw & REQ_WRITE)
		rwstat->cnt[BLKG_RWSTAT_WRITE] += val;
	else
		rwstat->cnt[BLKG_RWSTAT_READ] += val;
	if (rw & REQ_SYNC)
		rwstat->cnt[BLKG_RWSTAT_SYNC] += val;
	else
		rwstat->cnt[BLKG_RWSTAT_ASYNC] += val;

	u64_stats_update_end(&rwstat->syncp);
}

/**
 * blkg_rwstat_read - read the current values of a blkg_rwstat
 * @rwstat: blkg_rwstat to read
 *
 * Read the current snapshot of @rwstat and return it as the return value.
 * This function can be called without synchronization and takes care of
 * u64 atomicity.
 */
static inline struct blkg_rwstat blkg_rwstat_read(struct blkg_rwstat *rwstat)
{
	unsigned int start;
	struct blkg_rwstat tmp;

	do {
		start = u64_stats_fetch_begin_bh(&rwstat->syncp);
		tmp = *rwstat;
	} while (u64_stats_fetch_retry_bh(&rwstat->syncp, start));

	return tmp;
}

/**
 * blkg_rwstat_total - read the total count of a blkg_rwstat
 * @rwstat: blkg_rwstat to read
 *
 * Return the total count of @rwstat regardless of the IO direction.  This
 * function can be called without synchronization and takes care of u64
 * atomicity.
 */
static inline uint64_t blkg_rwstat_total(struct blkg_rwstat *rwstat)
{
	struct blkg_rwstat tmp = blkg_rwstat_read(rwstat);

	return tmp.cnt[BLKG_RWSTAT_READ] + tmp.cnt[BLKG_RWSTAT_WRITE];
}

/**
 * blkg_rwstat_reset - reset a blkg_rwstat
 * @rwstat: blkg_rwstat to reset
 */
static inline void blkg_rwstat_reset(struct blkg_rwstat *rwstat)
{
	memset(rwstat->cnt, 0, sizeof(rwstat->cnt));
}

/**
 * blkg_rwstat_merge - merge a blkg_rwstat into another
 * @to: the destination blkg_rwstat
 * @from: the source
 *
 * Add @from's counts to @to.
 */
static inline void blkg_rwstat_merge(struct blkg_rwstat *to,
				     struct blkg_rwstat *from)
{
	struct blkg_rwstat v = blkg_rwstat_read(from);
	int i;

	u64_stats_update_begin(&to->syncp);
	for (i = 0; i < BLKG_RWSTAT_NR; i++)
		to->cnt[i] += v.cnt[i];
	u64_stats_update_end(&to->syncp);
}

#else	/* CONFIG_BLK_CGROUP */

struct cgroup;
struct blkcg;

struct blkg_policy_data {
};

struct blkcg_gq {
};

struct blkcg_policy {
};

static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg, void *key) { return NULL; }
static inline int blkcg_init_queue(struct request_queue *q) { return 0; }
static inline void blkcg_drain_queue(struct request_queue *q) { }
static inline void blkcg_exit_queue(struct request_queue *q) { }
static inline int blkcg_policy_register(struct blkcg_policy *pol) { return 0; }
static inline void blkcg_policy_unregister(struct blkcg_policy *pol) { }
static inline int blkcg_activate_policy(struct request_queue *q,
					const struct blkcg_policy *pol) { return 0; }
static inline void blkcg_deactivate_policy(struct request_queue *q,
					   const struct blkcg_policy *pol) { }

static inline struct blkcg *bio_blkcg(struct bio *bio) { return NULL; }

static inline struct blkg_policy_data *blkg_to_pd(struct blkcg_gq *blkg,
						  struct blkcg_policy *pol) { return NULL; }
static inline struct blkcg_gq *pd_to_blkg(struct blkg_policy_data *pd) { return NULL; }
static inline char *blkg_path(struct blkcg_gq *blkg) { return NULL; }
static inline void blkg_get(struct blkcg_gq *blkg) { }
static inline void blkg_put(struct blkcg_gq *blkg) { }

static inline struct request_list *blk_get_rl(struct request_queue *q,
					      struct bio *bio) { return &q->root_rl; }
static inline void blk_put_rl(struct request_list *rl) { }
static inline void blk_rq_set_rl(struct request *rq, struct request_list *rl) { }
static inline struct request_list *blk_rq_rl(struct request *rq) { return &rq->q->root_rl; }

#define blk_queue_for_each_rl(rl, q)	\
	for ((rl) = &(q)->root_rl; (rl); (rl) = NULL)

#endif	/* CONFIG_BLK_CGROUP */
#endif	/* _BLK_CGROUP_H */
Commit	Line	Data
31e4c28d VG	1	#ifndef _BLK_CGROUP_H
	2	#define _BLK_CGROUP_H
	3	/*
	4	* Common Block IO controller cgroup interface
	5	*
	6	* Based on ideas and code from CFQ, CFS and BFQ:
	7	* Copyright (C) 2003 Jens Axboe <axboe@kernel.dk>
	8	*
	9	* Copyright (C) 2008 Fabio Checconi <fabio@gandalf.sssup.it>
	10	* Paolo Valente <paolo.valente@unimore.it>
	11	*
	12	* Copyright (C) 2009 Vivek Goyal <vgoyal@redhat.com>
	13	* Nauman Rafique <nauman@google.com>
	14	*/
	15
	16	#include <linux/cgroup.h>
575969a0	17	#include <linux/u64_stats_sync.h>
829fdb50	18	#include <linux/seq_file.h>
a637120e	19	#include <linux/radix-tree.h>
a051661c	20	#include <linux/blkdev.h>
31e4c28d	21
9355aede VG	22	/* Max limits for throttle policy */
	23	#define THROTL_IOPS_MAX UINT_MAX
	24
3381cb8d TH	25	/* CFQ specific, out here for blkcg->cfq_weight */
	26	#define CFQ_WEIGHT_MIN 10
	27	#define CFQ_WEIGHT_MAX 1000
	28	#define CFQ_WEIGHT_DEFAULT 500
	29
f48ec1d7 TH	30	#ifdef CONFIG_BLK_CGROUP
f48ec1d7 TH	31
edcb0722 TH	32	enum blkg_rwstat_type {
	33	BLKG_RWSTAT_READ,
	34	BLKG_RWSTAT_WRITE,
	35	BLKG_RWSTAT_SYNC,
	36	BLKG_RWSTAT_ASYNC,
	37
	38	BLKG_RWSTAT_NR,
	39	BLKG_RWSTAT_TOTAL = BLKG_RWSTAT_NR,
303a3acb DS	40	};
303a3acb DS	41
a637120e TH	42	struct blkcg_gq;
a637120e TH	43
3c798398	44	struct blkcg {
36558c8a TH	45	struct cgroup_subsys_state css;
36558c8a TH	46	spinlock_t lock;
a637120e TH	47
	48	struct radix_tree_root blkg_tree;
	49	struct blkcg_gq *blkg_hint;
36558c8a	50	struct hlist_head blkg_list;
9a9e8a26 TH	51
9a9e8a26 TH	52	/* for policies to test whether associated blkcg has changed */
36558c8a	53	uint64_t id;
3381cb8d	54
3c798398	55	/* TODO: per-policy storage in blkcg */
36558c8a	56	unsigned int cfq_weight; /* belongs to cfq */
e71357e1	57	unsigned int cfq_leaf_weight;
31e4c28d VG	58	};
31e4c28d VG	59
edcb0722 TH	60	struct blkg_stat {
	61	struct u64_stats_sync syncp;
	62	uint64_t cnt;
	63	};
	64
	65	struct blkg_rwstat {
	66	struct u64_stats_sync syncp;
	67	uint64_t cnt[BLKG_RWSTAT_NR];
	68	};
	69
f95a04af TH	70	/*
	71	* A blkcg_gq (blkg) is association between a block cgroup (blkcg) and a
	72	* request_queue (q). This is used by blkcg policies which need to track
	73	* information per blkcg - q pair.
	74	*
	75	* There can be multiple active blkcg policies and each has its private
	76	* data on each blkg, the size of which is determined by
	77	* blkcg_policy->pd_size. blkcg core allocates and frees such areas
	78	* together with blkg and invokes pd_init/exit_fn() methods.
	79	*
	80	* Such private data must embed struct blkg_policy_data (pd) at the
	81	* beginning and pd_size can't be smaller than pd.
	82	*/
0381411e	83	struct blkg_policy_data {
b276a876	84	/* the blkg and policy id this per-policy data belongs to */
3c798398	85	struct blkcg_gq *blkg;
b276a876	86	int plid;
0381411e	87
a2b1693b	88	/* used during policy activation */
36558c8a	89	struct list_head alloc_node;
0381411e TH	90	};
0381411e TH	91
3c798398 TH	92	/* association between a blk cgroup and a request queue */
3c798398 TH	93	struct blkcg_gq {
c875f4d0	94	/* Pointer to the associated request_queue */
36558c8a TH	95	struct request_queue *q;
	96	struct list_head q_node;
	97	struct hlist_node blkcg_node;
3c798398	98	struct blkcg *blkcg;
3c547865 TH	99
	100	/* all non-root blkcg_gq's are guaranteed to have access to parent */
	101	struct blkcg_gq *parent;
	102
a051661c TH	103	/* request allocation list for this blkcg-q pair */
a051661c TH	104	struct request_list rl;
3c547865	105
1adaf3dd	106	/* reference count */
36558c8a	107	int refcnt;
22084190	108
f427d909 TH	109	/* is this blkg online? protected by both blkcg and q locks */
	110	bool online;
	111
36558c8a	112	struct blkg_policy_data *pd[BLKCG_MAX_POLS];
1adaf3dd	113
36558c8a	114	struct rcu_head rcu_head;
31e4c28d VG	115	};
31e4c28d VG	116
3c798398	117	typedef void (blkcg_pol_init_pd_fn)(struct blkcg_gq *blkg);
f427d909 TH	118	typedef void (blkcg_pol_online_pd_fn)(struct blkcg_gq *blkg);
f427d909 TH	119	typedef void (blkcg_pol_offline_pd_fn)(struct blkcg_gq *blkg);
3c798398 TH	120	typedef void (blkcg_pol_exit_pd_fn)(struct blkcg_gq *blkg);
3c798398 TH	121	typedef void (blkcg_pol_reset_pd_stats_fn)(struct blkcg_gq *blkg);
3e252066	122
3c798398	123	struct blkcg_policy {
36558c8a TH	124	int plid;
36558c8a TH	125	/* policy specific private data size */
f95a04af	126	size_t pd_size;
36558c8a TH	127	/* cgroup files for the policy */
36558c8a TH	128	struct cftype *cftypes;
f9fcc2d3 TH	129
	130	/* operations */
	131	blkcg_pol_init_pd_fn *pd_init_fn;
f427d909 TH	132	blkcg_pol_online_pd_fn *pd_online_fn;
f427d909 TH	133	blkcg_pol_offline_pd_fn *pd_offline_fn;
f9fcc2d3 TH	134	blkcg_pol_exit_pd_fn *pd_exit_fn;
f9fcc2d3 TH	135	blkcg_pol_reset_pd_stats_fn *pd_reset_stats_fn;
3e252066 VG	136	};
3e252066 VG	137
3c798398	138	extern struct blkcg blkcg_root;
36558c8a	139
3c798398 TH	140	struct blkcg_gq blkg_lookup(struct blkcg blkcg, struct request_queue *q);
	141	struct blkcg_gq blkg_lookup_create(struct blkcg blkcg,
	142	struct request_queue *q);
36558c8a TH	143	int blkcg_init_queue(struct request_queue *q);
	144	void blkcg_drain_queue(struct request_queue *q);
	145	void blkcg_exit_queue(struct request_queue *q);
5efd6113	146
3e252066	147	/* Blkio controller policy registration */
3c798398 TH	148	int blkcg_policy_register(struct blkcg_policy *pol);
3c798398 TH	149	void blkcg_policy_unregister(struct blkcg_policy *pol);
36558c8a	150	int blkcg_activate_policy(struct request_queue *q,
3c798398	151	const struct blkcg_policy *pol);
36558c8a	152	void blkcg_deactivate_policy(struct request_queue *q,
3c798398	153	const struct blkcg_policy *pol);
3e252066	154
3c798398	155	void blkcg_print_blkgs(struct seq_file sf, struct blkcg blkcg,
f95a04af TH	156	u64 (prfill)(struct seq_file ,
f95a04af TH	157	struct blkg_policy_data *, int),
3c798398	158	const struct blkcg_policy *pol, int data,
ec399347	159	bool show_total);
f95a04af TH	160	u64 __blkg_prfill_u64(struct seq_file sf, struct blkg_policy_data pd, u64 v);
f95a04af TH	161	u64 __blkg_prfill_rwstat(struct seq_file sf, struct blkg_policy_data pd,
829fdb50	162	const struct blkg_rwstat *rwstat);
f95a04af TH	163	u64 blkg_prfill_stat(struct seq_file sf, struct blkg_policy_data pd, int off);
	164	u64 blkg_prfill_rwstat(struct seq_file sf, struct blkg_policy_data pd,
	165	int off);
829fdb50	166
16b3de66 TH	167	u64 blkg_stat_recursive_sum(struct blkg_policy_data *pd, int off);
	168	struct blkg_rwstat blkg_rwstat_recursive_sum(struct blkg_policy_data *pd,
	169	int off);
	170
829fdb50	171	struct blkg_conf_ctx {
36558c8a	172	struct gendisk *disk;
3c798398	173	struct blkcg_gq *blkg;
36558c8a	174	u64 v;
829fdb50 TH	175	};
829fdb50 TH	176
3c798398 TH	177	int blkg_conf_prep(struct blkcg blkcg, const struct blkcg_policy pol,
3c798398 TH	178	const char input, struct blkg_conf_ctx ctx);
829fdb50 TH	179	void blkg_conf_finish(struct blkg_conf_ctx *ctx);
	180
	181
a7c6d554 TH	182	static inline struct blkcg css_to_blkcg(struct cgroup_subsys_state css)
	183	{
	184	return css ? container_of(css, struct blkcg, css) : NULL;
	185	}
	186
b1208b56 TH	187	static inline struct blkcg task_blkcg(struct task_struct tsk)
b1208b56 TH	188	{
a7c6d554	189	return css_to_blkcg(task_css(tsk, blkio_subsys_id));
b1208b56 TH	190	}
	191
	192	static inline struct blkcg bio_blkcg(struct bio bio)
	193	{
	194	if (bio && bio->bi_css)
a7c6d554	195	return css_to_blkcg(bio->bi_css);
b1208b56 TH	196	return task_blkcg(current);
	197	}
	198
3c547865 TH	199	/**
	200	* blkcg_parent - get the parent of a blkcg
	201	* @blkcg: blkcg of interest
	202	*
	203	* Return the parent blkcg of @blkcg. Can be called anytime.
	204	*/
	205	static inline struct blkcg blkcg_parent(struct blkcg blkcg)
	206	{
63876986	207	return css_to_blkcg(css_parent(&blkcg->css));
3c547865 TH	208	}
3c547865 TH	209
0381411e TH	210	/**
	211	* blkg_to_pdata - get policy private data
	212	* @blkg: blkg of interest
	213	* @pol: policy of interest
	214	*
	215	* Return pointer to private data associated with the @blkg-@pol pair.
	216	*/
f95a04af TH	217	static inline struct blkg_policy_data blkg_to_pd(struct blkcg_gq blkg,
f95a04af TH	218	struct blkcg_policy *pol)
0381411e	219	{
f95a04af	220	return blkg ? blkg->pd[pol->plid] : NULL;
0381411e TH	221	}
	222
	223	/**
	224	* pdata_to_blkg - get blkg associated with policy private data
f95a04af	225	* @pd: policy private data of interest
0381411e	226	*
f95a04af	227	* @pd is policy private data. Determine the blkg it's associated with.
0381411e	228	*/
f95a04af	229	static inline struct blkcg_gq pd_to_blkg(struct blkg_policy_data pd)
0381411e	230	{
f95a04af	231	return pd ? pd->blkg : NULL;
0381411e TH	232	}
0381411e TH	233
54e7ed12 TH	234	/**
	235	* blkg_path - format cgroup path of blkg
	236	* @blkg: blkg of interest
	237	* @buf: target buffer
	238	* @buflen: target buffer length
	239	*
	240	* Format the path of the cgroup of @blkg into @buf.
	241	*/
3c798398	242	static inline int blkg_path(struct blkcg_gq blkg, char buf, int buflen)
afc24d49	243	{
54e7ed12 TH	244	int ret;
54e7ed12 TH	245
54e7ed12	246	ret = cgroup_path(blkg->blkcg->css.cgroup, buf, buflen);
54e7ed12 TH	247	if (ret)
	248	strncpy(buf, "<unavailable>", buflen);
	249	return ret;
afc24d49 VG	250	}
afc24d49 VG	251
1adaf3dd TH	252	/**
	253	* blkg_get - get a blkg reference
	254	* @blkg: blkg to get
	255	*
	256	* The caller should be holding queue_lock and an existing reference.
	257	*/
3c798398	258	static inline void blkg_get(struct blkcg_gq *blkg)
1adaf3dd TH	259	{
	260	lockdep_assert_held(blkg->q->queue_lock);
	261	WARN_ON_ONCE(!blkg->refcnt);
	262	blkg->refcnt++;
	263	}
	264
2a4fd070	265	void __blkg_release_rcu(struct rcu_head *rcu);
1adaf3dd TH	266
	267	/**
	268	* blkg_put - put a blkg reference
	269	* @blkg: blkg to put
	270	*
	271	* The caller should be holding queue_lock.
	272	*/
3c798398	273	static inline void blkg_put(struct blkcg_gq *blkg)
1adaf3dd TH	274	{
	275	lockdep_assert_held(blkg->q->queue_lock);
	276	WARN_ON_ONCE(blkg->refcnt <= 0);
	277	if (!--blkg->refcnt)
2a4fd070	278	call_rcu(&blkg->rcu_head, __blkg_release_rcu);
1adaf3dd TH	279	}
1adaf3dd TH	280
dd4a4ffc TH	281	struct blkcg_gq __blkg_lookup(struct blkcg blkcg, struct request_queue *q,
	282	bool update_hint);
	283
	284	/**
	285	* blkg_for_each_descendant_pre - pre-order walk of a blkg's descendants
	286	* @d_blkg: loop cursor pointing to the current descendant
492eb21b	287	* @pos_css: used for iteration
dd4a4ffc TH	288	* @p_blkg: target blkg to walk descendants of
	289	*
	290	* Walk @c_blkg through the descendants of @p_blkg. Must be used with RCU
	291	* read locked. If called under either blkcg or queue lock, the iteration
	292	* is guaranteed to include all and only online blkgs. The caller may
492eb21b	293	* update @pos_css by calling css_rightmost_descendant() to skip subtree.
bd8815a6	294	* @p_blkg is included in the iteration and the first node to be visited.
dd4a4ffc	295	*/
492eb21b TH	296	#define blkg_for_each_descendant_pre(d_blkg, pos_css, p_blkg) \
	297	css_for_each_descendant_pre((pos_css), &(p_blkg)->blkcg->css) \
	298	if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css), \
dd4a4ffc TH	299	(p_blkg)->q, false)))
dd4a4ffc TH	300
aa539cb3 TH	301	/**
	302	* blkg_for_each_descendant_post - post-order walk of a blkg's descendants
	303	* @d_blkg: loop cursor pointing to the current descendant
492eb21b	304	* @pos_css: used for iteration
aa539cb3 TH	305	* @p_blkg: target blkg to walk descendants of
	306	*
	307	* Similar to blkg_for_each_descendant_pre() but performs post-order
bd8815a6 TH	308	* traversal instead. Synchronization rules are the same. @p_blkg is
bd8815a6 TH	309	* included in the iteration and the last node to be visited.
aa539cb3	310	*/
492eb21b TH	311	#define blkg_for_each_descendant_post(d_blkg, pos_css, p_blkg) \
	312	css_for_each_descendant_post((pos_css), &(p_blkg)->blkcg->css) \
	313	if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css), \
aa539cb3 TH	314	(p_blkg)->q, false)))
aa539cb3 TH	315
a051661c TH	316	/**
	317	* blk_get_rl - get request_list to use
	318	* @q: request_queue of interest
	319	* @bio: bio which will be attached to the allocated request (may be %NULL)
	320	*
	321	* The caller wants to allocate a request from @q to use for @bio. Find
	322	* the request_list to use and obtain a reference on it. Should be called
	323	* under queue_lock. This function is guaranteed to return non-%NULL
	324	* request_list.
	325	*/
	326	static inline struct request_list blk_get_rl(struct request_queue q,
	327	struct bio *bio)
	328	{
	329	struct blkcg *blkcg;
	330	struct blkcg_gq *blkg;
	331
	332	rcu_read_lock();
	333
	334	blkcg = bio_blkcg(bio);
	335
	336	/* bypass blkg lookup and use @q->root_rl directly for root */
	337	if (blkcg == &blkcg_root)
	338	goto root_rl;
	339
	340	/*
	341	* Try to use blkg->rl. blkg lookup may fail under memory pressure
	342	* or if either the blkcg or queue is going away. Fall back to
	343	* root_rl in such cases.
	344	*/
	345	blkg = blkg_lookup_create(blkcg, q);
	346	if (unlikely(IS_ERR(blkg)))
	347	goto root_rl;
	348
	349	blkg_get(blkg);
	350	rcu_read_unlock();
	351	return &blkg->rl;
	352	root_rl:
	353	rcu_read_unlock();
	354	return &q->root_rl;
	355	}
	356
	357	/**
	358	* blk_put_rl - put request_list
	359	* @rl: request_list to put
	360	*
	361	* Put the reference acquired by blk_get_rl(). Should be called under
	362	* queue_lock.
	363	*/
	364	static inline void blk_put_rl(struct request_list *rl)
	365	{
	366	/* root_rl may not have blkg set */
	367	if (rl->blkg && rl->blkg->blkcg != &blkcg_root)
	368	blkg_put(rl->blkg);
	369	}
	370
	371	/**
	372	* blk_rq_set_rl - associate a request with a request_list
	373	* @rq: request of interest
	374	* @rl: target request_list
	375	*
	376	* Associate @rq with @rl so that accounting and freeing can know the
	377	* request_list @rq came from.
	378	*/
	379	static inline void blk_rq_set_rl(struct request rq, struct request_list rl)
380	{
381	rq->rl = rl;
382	}
383
384	/**
385	* blk_rq_rl - return the request_list a request came from
386	* @rq: request of interest
387	*
388	* Return the request_list @rq is allocated from.
389	*/
390	static inline struct request_list blk_rq_rl(struct request rq)
391	{
392	return rq->rl;
393	}
394
395	struct request_list __blk_queue_next_rl(struct request_list rl,
396	struct request_queue *q);
397	/**
398	* blk_queue_for_each_rl - iterate through all request_lists of a request_queue
399	*
400	* Should be used under queue_lock.
401	*/
402	#define blk_queue_for_each_rl(rl, q) \
403	for ((rl) = &(q)->root_rl; (rl); (rl) = __blk_queue_next_rl((rl), (q)))
404
90d3839b PZ	405	static inline void blkg_stat_init(struct blkg_stat *stat)
	406	{
	407	u64_stats_init(&stat->syncp);
	408	}
	409
edcb0722 TH	410	/**
	411	* blkg_stat_add - add a value to a blkg_stat
	412	* @stat: target blkg_stat
	413	* @val: value to add
	414	*
	415	* Add @val to @stat. The caller is responsible for synchronizing calls to
	416	* this function.
	417	*/
	418	static inline void blkg_stat_add(struct blkg_stat *stat, uint64_t val)
	419	{
	420	u64_stats_update_begin(&stat->syncp);
	421	stat->cnt += val;
	422	u64_stats_update_end(&stat->syncp);
	423	}
	424
	425	/**
	426	* blkg_stat_read - read the current value of a blkg_stat
	427	* @stat: blkg_stat to read
	428	*
	429	* Read the current value of @stat. This function can be called without
	430	* synchroniztion and takes care of u64 atomicity.
	431	*/
	432	static inline uint64_t blkg_stat_read(struct blkg_stat *stat)
	433	{
	434	unsigned int start;
	435	uint64_t v;
	436
	437	do {
2c575026	438	start = u64_stats_fetch_begin_bh(&stat->syncp);
edcb0722	439	v = stat->cnt;
2c575026	440	} while (u64_stats_fetch_retry_bh(&stat->syncp, start));
edcb0722 TH	441
	442	return v;
	443	}
	444
	445	/**
	446	* blkg_stat_reset - reset a blkg_stat
	447	* @stat: blkg_stat to reset
	448	*/
	449	static inline void blkg_stat_reset(struct blkg_stat *stat)
	450	{
	451	stat->cnt = 0;
	452	}
	453
16b3de66 TH	454	/**
	455	* blkg_stat_merge - merge a blkg_stat into another
	456	* @to: the destination blkg_stat
	457	* @from: the source
	458	*
	459	* Add @from's count to @to.
	460	*/
	461	static inline void blkg_stat_merge(struct blkg_stat to, struct blkg_stat from)
	462	{
	463	blkg_stat_add(to, blkg_stat_read(from));
	464	}
	465
90d3839b PZ	466	static inline void blkg_rwstat_init(struct blkg_rwstat *rwstat)
	467	{
	468	u64_stats_init(&rwstat->syncp);
	469	}
	470
edcb0722 TH	471	/**
	472	* blkg_rwstat_add - add a value to a blkg_rwstat
	473	* @rwstat: target blkg_rwstat
	474	* @rw: mask of REQ_{WRITE\|SYNC}
	475	* @val: value to add
	476	*
	477	* Add @val to @rwstat. The counters are chosen according to @rw. The
	478	* caller is responsible for synchronizing calls to this function.
	479	*/
	480	static inline void blkg_rwstat_add(struct blkg_rwstat *rwstat,
	481	int rw, uint64_t val)
	482	{
	483	u64_stats_update_begin(&rwstat->syncp);
	484
	485	if (rw & REQ_WRITE)
	486	rwstat->cnt[BLKG_RWSTAT_WRITE] += val;
	487	else
	488	rwstat->cnt[BLKG_RWSTAT_READ] += val;
	489	if (rw & REQ_SYNC)
	490	rwstat->cnt[BLKG_RWSTAT_SYNC] += val;
	491	else
	492	rwstat->cnt[BLKG_RWSTAT_ASYNC] += val;
	493
	494	u64_stats_update_end(&rwstat->syncp);
	495	}
	496
	497	/**
	498	* blkg_rwstat_read - read the current values of a blkg_rwstat
	499	* @rwstat: blkg_rwstat to read
	500	*
	501	* Read the current snapshot of @rwstat and return it as the return value.
	502	* This function can be called without synchronization and takes care of
	503	* u64 atomicity.
	504	*/
c94bed89	505	static inline struct blkg_rwstat blkg_rwstat_read(struct blkg_rwstat *rwstat)
edcb0722 TH	506	{
	507	unsigned int start;
	508	struct blkg_rwstat tmp;
	509
	510	do {
2c575026	511	start = u64_stats_fetch_begin_bh(&rwstat->syncp);
edcb0722	512	tmp = *rwstat;
2c575026	513	} while (u64_stats_fetch_retry_bh(&rwstat->syncp, start));
edcb0722 TH	514
	515	return tmp;
	516	}
	517
	518	/**
4d5e80a7	519	* blkg_rwstat_total - read the total count of a blkg_rwstat
edcb0722 TH	520	* @rwstat: blkg_rwstat to read
	521	*
	522	* Return the total count of @rwstat regardless of the IO direction. This
	523	* function can be called without synchronization and takes care of u64
	524	* atomicity.
	525	*/
4d5e80a7	526	static inline uint64_t blkg_rwstat_total(struct blkg_rwstat *rwstat)
edcb0722 TH	527	{
	528	struct blkg_rwstat tmp = blkg_rwstat_read(rwstat);
	529
	530	return tmp.cnt[BLKG_RWSTAT_READ] + tmp.cnt[BLKG_RWSTAT_WRITE];
	531	}
	532
	533	/**
	534	* blkg_rwstat_reset - reset a blkg_rwstat
	535	* @rwstat: blkg_rwstat to reset
	536	*/
	537	static inline void blkg_rwstat_reset(struct blkg_rwstat *rwstat)
	538	{
	539	memset(rwstat->cnt, 0, sizeof(rwstat->cnt));
	540	}
	541
16b3de66 TH	542	/**
	543	* blkg_rwstat_merge - merge a blkg_rwstat into another
	544	* @to: the destination blkg_rwstat
	545	* @from: the source
	546	*
	547	* Add @from's counts to @to.
	548	*/
	549	static inline void blkg_rwstat_merge(struct blkg_rwstat *to,
	550	struct blkg_rwstat *from)
	551	{
	552	struct blkg_rwstat v = blkg_rwstat_read(from);
	553	int i;
	554
	555	u64_stats_update_begin(&to->syncp);
	556	for (i = 0; i < BLKG_RWSTAT_NR; i++)
	557	to->cnt[i] += v.cnt[i];
	558	u64_stats_update_end(&to->syncp);
	559	}
	560
36558c8a TH	561	#else /* CONFIG_BLK_CGROUP */
	562
	563	struct cgroup;
b1208b56	564	struct blkcg;
2f5ea477	565
f95a04af TH	566	struct blkg_policy_data {
	567	};
	568
3c798398	569	struct blkcg_gq {
2f5ea477 JA	570	};
2f5ea477 JA	571
3c798398	572	struct blkcg_policy {
3e252066 VG	573	};
3e252066 VG	574
3c798398	575	static inline struct blkcg_gq blkg_lookup(struct blkcg blkcg, void *key) { return NULL; }
5efd6113 TH	576	static inline int blkcg_init_queue(struct request_queue *q) { return 0; }
	577	static inline void blkcg_drain_queue(struct request_queue *q) { }
	578	static inline void blkcg_exit_queue(struct request_queue *q) { }
3c798398 TH	579	static inline int blkcg_policy_register(struct blkcg_policy *pol) { return 0; }
3c798398 TH	580	static inline void blkcg_policy_unregister(struct blkcg_policy *pol) { }
a2b1693b	581	static inline int blkcg_activate_policy(struct request_queue *q,
3c798398	582	const struct blkcg_policy *pol) { return 0; }
a2b1693b	583	static inline void blkcg_deactivate_policy(struct request_queue *q,
3c798398 TH	584	const struct blkcg_policy *pol) { }
3c798398 TH	585
b1208b56	586	static inline struct blkcg bio_blkcg(struct bio bio) { return NULL; }
a051661c	587
f95a04af TH	588	static inline struct blkg_policy_data blkg_to_pd(struct blkcg_gq blkg,
	589	struct blkcg_policy *pol) { return NULL; }
	590	static inline struct blkcg_gq pd_to_blkg(struct blkg_policy_data pd) { return NULL; }
3c798398 TH	591	static inline char blkg_path(struct blkcg_gq blkg) { return NULL; }
	592	static inline void blkg_get(struct blkcg_gq *blkg) { }
	593	static inline void blkg_put(struct blkcg_gq *blkg) { }
afc24d49	594
a051661c TH	595	static inline struct request_list blk_get_rl(struct request_queue q,
	596	struct bio *bio) { return &q->root_rl; }
	597	static inline void blk_put_rl(struct request_list *rl) { }
	598	static inline void blk_rq_set_rl(struct request rq, struct request_list rl) { }
	599	static inline struct request_list blk_rq_rl(struct request rq) { return &rq->q->root_rl; }
	600
	601	#define blk_queue_for_each_rl(rl, q) \
	602	for ((rl) = &(q)->root_rl; (rl); (rl) = NULL)
	603
36558c8a TH	604	#endif /* CONFIG_BLK_CGROUP */
36558c8a TH	605	#endif /* _BLK_CGROUP_H */