[linux-block.git] / include / linux / backing-dev-defs.h

/* SPDX-License-Identifier: GPL-2.0 */
#ifndef __LINUX_BACKING_DEV_DEFS_H
#define __LINUX_BACKING_DEV_DEFS_H

#include <linux/list.h>
#include <linux/radix-tree.h>
#include <linux/rbtree.h>
#include <linux/spinlock.h>
#include <linux/percpu_counter.h>
#include <linux/percpu-refcount.h>
#include <linux/flex_proportions.h>
#include <linux/timer.h>
#include <linux/workqueue.h>
#include <linux/kref.h>

struct page;
struct device;
struct dentry;

/*
 * Bits in bdi_writeback.state
 */
enum wb_state {
	WB_registered,		/* bdi_register() was done */
	WB_writeback_running,	/* Writeback is in progress */
	WB_has_dirty_io,	/* Dirty inodes on ->b_{dirty|io|more_io} */
	WB_start_all,		/* nr_pages == 0 (all) work pending */
};

enum wb_congested_state {
	WB_async_congested,	/* The async (write) queue is getting full */
	WB_sync_congested,	/* The sync queue is getting full */
};

typedef int (congested_fn)(void *, int);

enum wb_stat_item {
	WB_RECLAIMABLE,
	WB_WRITEBACK,
	WB_DIRTIED,
	WB_WRITTEN,
	NR_WB_STAT_ITEMS
};

#define WB_STAT_BATCH (8*(1+ilog2(nr_cpu_ids)))

/*
 * why some writeback work was initiated
 */
enum wb_reason {
	WB_REASON_BACKGROUND,
	WB_REASON_VMSCAN,
	WB_REASON_SYNC,
	WB_REASON_PERIODIC,
	WB_REASON_LAPTOP_TIMER,
	WB_REASON_FREE_MORE_MEM,
	WB_REASON_FS_FREE_SPACE,
	/*
	 * There is no bdi forker thread any more and works are done
	 * by emergency worker, however, this is TPs userland visible
	 * and we'll be exposing exactly the same information,
	 * so it has a mismatch name.
	 */
	WB_REASON_FORKER_THREAD,

	WB_REASON_MAX,
};

/*
 * For cgroup writeback, multiple wb's may map to the same blkcg.  Those
 * wb's can operate mostly independently but should share the congested
 * state.  To facilitate such sharing, the congested state is tracked using
 * the following struct which is created on demand, indexed by blkcg ID on
 * its bdi, and refcounted.
 */
struct bdi_writeback_congested {
	unsigned long state;		/* WB_[a]sync_congested flags */
	atomic_t refcnt;		/* nr of attached wb's and blkg */

#ifdef CONFIG_CGROUP_WRITEBACK
	struct backing_dev_info *__bdi;	/* the associated bdi, set to NULL
					 * on bdi unregistration. For memcg-wb
					 * internal use only! */
	int blkcg_id;			/* ID of the associated blkcg */
	struct rb_node rb_node;		/* on bdi->cgwb_congestion_tree */
#endif
};

/*
 * Each wb (bdi_writeback) can perform writeback operations, is measured
 * and throttled, independently.  Without cgroup writeback, each bdi
 * (bdi_writeback) is served by its embedded bdi->wb.
 *
 * On the default hierarchy, blkcg implicitly enables memcg.  This allows
 * using memcg's page ownership for attributing writeback IOs, and every
 * memcg - blkcg combination can be served by its own wb by assigning a
 * dedicated wb to each memcg, which enables isolation across different
 * cgroups and propagation of IO back pressure down from the IO layer upto
 * the tasks which are generating the dirty pages to be written back.
 *
 * A cgroup wb is indexed on its bdi by the ID of the associated memcg,
 * refcounted with the number of inodes attached to it, and pins the memcg
 * and the corresponding blkcg.  As the corresponding blkcg for a memcg may
 * change as blkcg is disabled and enabled higher up in the hierarchy, a wb
 * is tested for blkcg after lookup and removed from index on mismatch so
 * that a new wb for the combination can be created.
 */
struct bdi_writeback {
	struct backing_dev_info *bdi;	/* our parent bdi */

	unsigned long state;		/* Always use atomic bitops on this */
	unsigned long last_old_flush;	/* last old data flush */

	struct list_head b_dirty;	/* dirty inodes */
	struct list_head b_io;		/* parked for writeback */
	struct list_head b_more_io;	/* parked for more writeback */
	struct list_head b_dirty_time;	/* time stamps are dirty */
	spinlock_t list_lock;		/* protects the b_* lists */

	struct percpu_counter stat[NR_WB_STAT_ITEMS];

	struct bdi_writeback_congested *congested;

	unsigned long bw_time_stamp;	/* last time write bw is updated */
	unsigned long dirtied_stamp;
	unsigned long written_stamp;	/* pages written at bw_time_stamp */
	unsigned long write_bandwidth;	/* the estimated write bandwidth */
	unsigned long avg_write_bandwidth; /* further smoothed write bw, > 0 */

	/*
	 * The base dirty throttle rate, re-calculated on every 200ms.
	 * All the bdi tasks' dirty rate will be curbed under it.
	 * @dirty_ratelimit tracks the estimated @balanced_dirty_ratelimit
	 * in small steps and is much more smooth/stable than the latter.
	 */
	unsigned long dirty_ratelimit;
	unsigned long balanced_dirty_ratelimit;

	struct fprop_local_percpu completions;
	int dirty_exceeded;
	enum wb_reason start_all_reason;

	spinlock_t work_lock;		/* protects work_list & dwork scheduling */
	struct list_head work_list;
	struct delayed_work dwork;	/* work item used for writeback */

	unsigned long dirty_sleep;	/* last wait */

	struct list_head bdi_node;	/* anchored at bdi->wb_list */

#ifdef CONFIG_CGROUP_WRITEBACK
	struct percpu_ref refcnt;	/* used only for !root wb's */
	struct fprop_local_percpu memcg_completions;
	struct cgroup_subsys_state *memcg_css; /* the associated memcg */
	struct cgroup_subsys_state *blkcg_css; /* and blkcg */
	struct list_head memcg_node;	/* anchored at memcg->cgwb_list */
	struct list_head blkcg_node;	/* anchored at blkcg->cgwb_list */

	union {
		struct work_struct release_work;
		struct rcu_head rcu;
	};
#endif
};

struct backing_dev_info {
	struct list_head bdi_list;
	unsigned long ra_pages;	/* max readahead in PAGE_SIZE units */
	unsigned long io_pages;	/* max allowed IO size */
	congested_fn *congested_fn; /* Function pointer if device is md/dm */
	void *congested_data;	/* Pointer to aux data for congested func */

	const char *name;

	struct kref refcnt;	/* Reference counter for the structure */
	unsigned int capabilities; /* Device capabilities */
	unsigned int min_ratio;
	unsigned int max_ratio, max_prop_frac;

	/*
	 * Sum of avg_write_bw of wbs with dirty inodes.  > 0 if there are
	 * any dirty wbs, which is depended upon by bdi_has_dirty().
	 */
	atomic_long_t tot_write_bandwidth;

	struct bdi_writeback wb;  /* the root writeback info for this bdi */
	struct list_head wb_list; /* list of all wbs */
#ifdef CONFIG_CGROUP_WRITEBACK
	struct radix_tree_root cgwb_tree; /* radix tree of active cgroup wbs */
	struct rb_root cgwb_congested_tree; /* their congested states */
	struct mutex cgwb_release_mutex;  /* protect shutdown of wb structs */
#else
	struct bdi_writeback_congested *wb_congested;
#endif
	wait_queue_head_t wb_waitq;

	struct device *dev;
	struct device *owner;

	struct timer_list laptop_mode_wb_timer;

#ifdef CONFIG_DEBUG_FS
	struct dentry *debug_dir;
	struct dentry *debug_stats;
#endif
};

enum {
	BLK_RW_ASYNC	= 0,
	BLK_RW_SYNC	= 1,
};

void clear_wb_congested(struct bdi_writeback_congested *congested, int sync);
void set_wb_congested(struct bdi_writeback_congested *congested, int sync);

static inline void clear_bdi_congested(struct backing_dev_info *bdi, int sync)
{
	clear_wb_congested(bdi->wb.congested, sync);
}

static inline void set_bdi_congested(struct backing_dev_info *bdi, int sync)
{
	set_wb_congested(bdi->wb.congested, sync);
}

struct wb_lock_cookie {
	bool locked;
	unsigned long flags;
};

#ifdef CONFIG_CGROUP_WRITEBACK

/**
 * wb_tryget - try to increment a wb's refcount
 * @wb: bdi_writeback to get
 */
static inline bool wb_tryget(struct bdi_writeback *wb)
{
	if (wb != &wb->bdi->wb)
		return percpu_ref_tryget(&wb->refcnt);
	return true;
}

/**
 * wb_get - increment a wb's refcount
 * @wb: bdi_writeback to get
 */
static inline void wb_get(struct bdi_writeback *wb)
{
	if (wb != &wb->bdi->wb)
		percpu_ref_get(&wb->refcnt);
}

/**
 * wb_put - decrement a wb's refcount
 * @wb: bdi_writeback to put
 */
static inline void wb_put(struct bdi_writeback *wb)
{
	if (wb != &wb->bdi->wb)
		percpu_ref_put(&wb->refcnt);
}

/**
 * wb_dying - is a wb dying?
 * @wb: bdi_writeback of interest
 *
 * Returns whether @wb is unlinked and being drained.
 */
static inline bool wb_dying(struct bdi_writeback *wb)
{
	return percpu_ref_is_dying(&wb->refcnt);
}

#else	/* CONFIG_CGROUP_WRITEBACK */

static inline bool wb_tryget(struct bdi_writeback *wb)
{
	return true;
}

static inline void wb_get(struct bdi_writeback *wb)
{
}

static inline void wb_put(struct bdi_writeback *wb)
{
}

static inline bool wb_dying(struct bdi_writeback *wb)
{
	return false;
}

#endif	/* CONFIG_CGROUP_WRITEBACK */

#endif	/* __LINUX_BACKING_DEV_DEFS_H */
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
66114cad TH	2	#ifndef __LINUX_BACKING_DEV_DEFS_H
	3	#define __LINUX_BACKING_DEV_DEFS_H
	4
	5	#include <linux/list.h>
52ebea74 TH	6	#include <linux/radix-tree.h>
52ebea74 TH	7	#include <linux/rbtree.h>
66114cad TH	8	#include <linux/spinlock.h>
66114cad TH	9	#include <linux/percpu_counter.h>
52ebea74	10	#include <linux/percpu-refcount.h>
66114cad TH	11	#include <linux/flex_proportions.h>
	12	#include <linux/timer.h>
	13	#include <linux/workqueue.h>
d03f6cdc	14	#include <linux/kref.h>
66114cad TH	15
	16	struct page;
	17	struct device;
	18	struct dentry;
	19
	20	/*
	21	* Bits in bdi_writeback.state
	22	*/
	23	enum wb_state {
66114cad TH	24	WB_registered, /* bdi_register() was done */
66114cad TH	25	WB_writeback_running, /* Writeback is in progress */
d6c10f1f	26	WB_has_dirty_io, /* Dirty inodes on ->b_{dirty\|io\|more_io} */
aac8d41c	27	WB_start_all, /* nr_pages == 0 (all) work pending */
66114cad TH	28	};
66114cad TH	29
4aa9c692 TH	30	enum wb_congested_state {
	31	WB_async_congested, /* The async (write) queue is getting full */
	32	WB_sync_congested, /* The sync queue is getting full */
	33	};
	34
66114cad TH	35	typedef int (congested_fn)(void *, int);
	36
	37	enum wb_stat_item {
	38	WB_RECLAIMABLE,
	39	WB_WRITEBACK,
	40	WB_DIRTIED,
	41	WB_WRITTEN,
	42	NR_WB_STAT_ITEMS
	43	};
	44
	45	#define WB_STAT_BATCH (8*(1+ilog2(nr_cpu_ids)))
	46
85009b4f JA	47	/*
	48	* why some writeback work was initiated
	49	*/
	50	enum wb_reason {
	51	WB_REASON_BACKGROUND,
	52	WB_REASON_VMSCAN,
	53	WB_REASON_SYNC,
	54	WB_REASON_PERIODIC,
	55	WB_REASON_LAPTOP_TIMER,
	56	WB_REASON_FREE_MORE_MEM,
	57	WB_REASON_FS_FREE_SPACE,
	58	/*
	59	* There is no bdi forker thread any more and works are done
	60	* by emergency worker, however, this is TPs userland visible
	61	* and we'll be exposing exactly the same information,
	62	* so it has a mismatch name.
	63	*/
	64	WB_REASON_FORKER_THREAD,
	65
	66	WB_REASON_MAX,
	67	};
	68
52ebea74 TH	69	/*
	70	* For cgroup writeback, multiple wb's may map to the same blkcg. Those
	71	* wb's can operate mostly independently but should share the congested
	72	* state. To facilitate such sharing, the congested state is tracked using
	73	* the following struct which is created on demand, indexed by blkcg ID on
	74	* its bdi, and refcounted.
	75	*/
4aa9c692 TH	76	struct bdi_writeback_congested {
4aa9c692 TH	77	unsigned long state; /* WB_[a]sync_congested flags */
a13f35e8	78	atomic_t refcnt; /* nr of attached wb's and blkg */
52ebea74 TH	79
52ebea74 TH	80	#ifdef CONFIG_CGROUP_WRITEBACK
b7d680d7 JK	81	struct backing_dev_info __bdi; / the associated bdi, set to NULL
	82	* on bdi unregistration. For memcg-wb
	83	* internal use only! */
52ebea74 TH	84	int blkcg_id; /* ID of the associated blkcg */
	85	struct rb_node rb_node; /* on bdi->cgwb_congestion_tree */
	86	#endif
4aa9c692 TH	87	};
4aa9c692 TH	88
52ebea74 TH	89	/*
	90	* Each wb (bdi_writeback) can perform writeback operations, is measured
	91	* and throttled, independently. Without cgroup writeback, each bdi
	92	* (bdi_writeback) is served by its embedded bdi->wb.
	93	*
	94	* On the default hierarchy, blkcg implicitly enables memcg. This allows
	95	* using memcg's page ownership for attributing writeback IOs, and every
	96	* memcg - blkcg combination can be served by its own wb by assigning a
	97	* dedicated wb to each memcg, which enables isolation across different
	98	* cgroups and propagation of IO back pressure down from the IO layer upto
	99	* the tasks which are generating the dirty pages to be written back.
	100	*
	101	* A cgroup wb is indexed on its bdi by the ID of the associated memcg,
	102	* refcounted with the number of inodes attached to it, and pins the memcg
	103	* and the corresponding blkcg. As the corresponding blkcg for a memcg may
	104	* change as blkcg is disabled and enabled higher up in the hierarchy, a wb
	105	* is tested for blkcg after lookup and removed from index on mismatch so
	106	* that a new wb for the combination can be created.
	107	*/
66114cad TH	108	struct bdi_writeback {
	109	struct backing_dev_info bdi; / our parent bdi */
	110
	111	unsigned long state; /* Always use atomic bitops on this */
	112	unsigned long last_old_flush; /* last old data flush */
	113
	114	struct list_head b_dirty; /* dirty inodes */
	115	struct list_head b_io; /* parked for writeback */
	116	struct list_head b_more_io; /* parked for more writeback */
	117	struct list_head b_dirty_time; /* time stamps are dirty */
	118	spinlock_t list_lock; /* protects the b_* lists */
	119
	120	struct percpu_counter stat[NR_WB_STAT_ITEMS];
	121
4aa9c692 TH	122	struct bdi_writeback_congested *congested;
4aa9c692 TH	123
66114cad TH	124	unsigned long bw_time_stamp; /* last time write bw is updated */
	125	unsigned long dirtied_stamp;
	126	unsigned long written_stamp; /* pages written at bw_time_stamp */
	127	unsigned long write_bandwidth; /* the estimated write bandwidth */
95a46c65	128	unsigned long avg_write_bandwidth; /* further smoothed write bw, > 0 */
66114cad TH	129
	130	/*
	131	* The base dirty throttle rate, re-calculated on every 200ms.
	132	* All the bdi tasks' dirty rate will be curbed under it.
	133	* @dirty_ratelimit tracks the estimated @balanced_dirty_ratelimit
	134	* in small steps and is much more smooth/stable than the latter.
	135	*/
	136	unsigned long dirty_ratelimit;
	137	unsigned long balanced_dirty_ratelimit;
	138
	139	struct fprop_local_percpu completions;
	140	int dirty_exceeded;
85009b4f	141	enum wb_reason start_all_reason;
66114cad TH	142
	143	spinlock_t work_lock; /* protects work_list & dwork scheduling */
	144	struct list_head work_list;
	145	struct delayed_work dwork; /* work item used for writeback */
52ebea74	146
b57d74af JA	147	unsigned long dirty_sleep; /* last wait */
b57d74af JA	148
b817525a TH	149	struct list_head bdi_node; /* anchored at bdi->wb_list */
b817525a TH	150
52ebea74 TH	151	#ifdef CONFIG_CGROUP_WRITEBACK
52ebea74 TH	152	struct percpu_ref refcnt; /* used only for !root wb's */
841710aa	153	struct fprop_local_percpu memcg_completions;
52ebea74 TH	154	struct cgroup_subsys_state memcg_css; / the associated memcg */
	155	struct cgroup_subsys_state blkcg_css; / and blkcg */
	156	struct list_head memcg_node; /* anchored at memcg->cgwb_list */
	157	struct list_head blkcg_node; /* anchored at blkcg->cgwb_list */
	158
	159	union {
	160	struct work_struct release_work;
	161	struct rcu_head rcu;
	162	};
	163	#endif
66114cad TH	164	};
	165
	166	struct backing_dev_info {
	167	struct list_head bdi_list;
ea1754a0	168	unsigned long ra_pages; /* max readahead in PAGE_SIZE units */
9491ae4a	169	unsigned long io_pages; /* max allowed IO size */
66114cad TH	170	congested_fn congested_fn; / Function pointer if device is md/dm */
	171	void congested_data; / Pointer to aux data for congested func */
	172
fca39346	173	const char *name;
66114cad	174
d03f6cdc	175	struct kref refcnt; /* Reference counter for the structure */
8db378a5	176	unsigned int capabilities; /* Device capabilities */
66114cad TH	177	unsigned int min_ratio;
	178	unsigned int max_ratio, max_prop_frac;
	179
95a46c65 TH	180	/*
	181	* Sum of avg_write_bw of wbs with dirty inodes. > 0 if there are
	182	* any dirty wbs, which is depended upon by bdi_has_dirty().
	183	*/
	184	atomic_long_t tot_write_bandwidth;
766a9d6e	185
52ebea74	186	struct bdi_writeback wb; /* the root writeback info for this bdi */
b817525a	187	struct list_head wb_list; /* list of all wbs */
52ebea74 TH	188	#ifdef CONFIG_CGROUP_WRITEBACK
	189	struct radix_tree_root cgwb_tree; /* radix tree of active cgroup wbs */
	190	struct rb_root cgwb_congested_tree; /* their congested states */
3ee7e869	191	struct mutex cgwb_release_mutex; /* protect shutdown of wb structs */
a13f35e8 TH	192	#else
a13f35e8 TH	193	struct bdi_writeback_congested *wb_congested;
52ebea74	194	#endif
cc395d7f TH	195	wait_queue_head_t wb_waitq;
cc395d7f TH	196
66114cad	197	struct device *dev;
df08c32c	198	struct device *owner;
66114cad TH	199
	200	struct timer_list laptop_mode_wb_timer;
	201
	202	#ifdef CONFIG_DEBUG_FS
	203	struct dentry *debug_dir;
	204	struct dentry *debug_stats;
	205	#endif
	206	};
	207
	208	enum {
	209	BLK_RW_ASYNC = 0,
	210	BLK_RW_SYNC = 1,
	211	};
	212
ec8a6f26 TH	213	void clear_wb_congested(struct bdi_writeback_congested *congested, int sync);
	214	void set_wb_congested(struct bdi_writeback_congested *congested, int sync);
	215
	216	static inline void clear_bdi_congested(struct backing_dev_info *bdi, int sync)
	217	{
	218	clear_wb_congested(bdi->wb.congested, sync);
	219	}
	220
	221	static inline void set_bdi_congested(struct backing_dev_info *bdi, int sync)
	222	{
	223	set_wb_congested(bdi->wb.congested, sync);
	224	}
66114cad	225
2e898e4c GT	226	struct wb_lock_cookie {
	227	bool locked;
	228	unsigned long flags;
	229	};
	230
21c6321f TH	231	#ifdef CONFIG_CGROUP_WRITEBACK
	232
	233	/**
	234	* wb_tryget - try to increment a wb's refcount
	235	* @wb: bdi_writeback to get
	236	*/
	237	static inline bool wb_tryget(struct bdi_writeback *wb)
	238	{
	239	if (wb != &wb->bdi->wb)
	240	return percpu_ref_tryget(&wb->refcnt);
	241	return true;
	242	}
	243
	244	/**
	245	* wb_get - increment a wb's refcount
	246	* @wb: bdi_writeback to get
	247	*/
	248	static inline void wb_get(struct bdi_writeback *wb)
	249	{
	250	if (wb != &wb->bdi->wb)
	251	percpu_ref_get(&wb->refcnt);
	252	}
	253
	254	/**
	255	* wb_put - decrement a wb's refcount
	256	* @wb: bdi_writeback to put
	257	*/
	258	static inline void wb_put(struct bdi_writeback *wb)
	259	{
	260	if (wb != &wb->bdi->wb)
	261	percpu_ref_put(&wb->refcnt);
	262	}
	263
e8a7abf5 TH	264	/**
	265	* wb_dying - is a wb dying?
	266	* @wb: bdi_writeback of interest
	267	*
	268	* Returns whether @wb is unlinked and being drained.
	269	*/
	270	static inline bool wb_dying(struct bdi_writeback *wb)
	271	{
	272	return percpu_ref_is_dying(&wb->refcnt);
	273	}
	274
21c6321f TH	275	#else /* CONFIG_CGROUP_WRITEBACK */
	276
	277	static inline bool wb_tryget(struct bdi_writeback *wb)
	278	{
	279	return true;
	280	}
	281
	282	static inline void wb_get(struct bdi_writeback *wb)
	283	{
	284	}
	285
	286	static inline void wb_put(struct bdi_writeback *wb)
	287	{
	288	}
	289
e8a7abf5 TH	290	static inline bool wb_dying(struct bdi_writeback *wb)
	291	{
	292	return false;
	293	}
	294
21c6321f TH	295	#endif /* CONFIG_CGROUP_WRITEBACK */
21c6321f TH	296
66114cad	297	#endif /* __LINUX_BACKING_DEV_DEFS_H */