[linux-2.6-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>
#include <linux/refcount.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 */
	refcount_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 (WARN_ON_ONCE(!wb->bdi)) {
		/*
		 * A driver bug might cause a file to be removed before bdi was
		 * initialized.
		 */
		return;
	}

	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>
e58dd0de	15	#include <linux/refcount.h>
66114cad TH	16
	17	struct page;
	18	struct device;
	19	struct dentry;
	20
	21	/*
	22	* Bits in bdi_writeback.state
	23	*/
	24	enum wb_state {
66114cad TH	25	WB_registered, /* bdi_register() was done */
66114cad TH	26	WB_writeback_running, /* Writeback is in progress */
d6c10f1f	27	WB_has_dirty_io, /* Dirty inodes on ->b_{dirty\|io\|more_io} */
aac8d41c	28	WB_start_all, /* nr_pages == 0 (all) work pending */
66114cad TH	29	};
66114cad TH	30
4aa9c692 TH	31	enum wb_congested_state {
	32	WB_async_congested, /* The async (write) queue is getting full */
	33	WB_sync_congested, /* The sync queue is getting full */
	34	};
	35
66114cad TH	36	typedef int (congested_fn)(void *, int);
	37
	38	enum wb_stat_item {
	39	WB_RECLAIMABLE,
	40	WB_WRITEBACK,
	41	WB_DIRTIED,
	42	WB_WRITTEN,
	43	NR_WB_STAT_ITEMS
	44	};
	45
	46	#define WB_STAT_BATCH (8*(1+ilog2(nr_cpu_ids)))
	47
85009b4f JA	48	/*
	49	* why some writeback work was initiated
	50	*/
	51	enum wb_reason {
	52	WB_REASON_BACKGROUND,
	53	WB_REASON_VMSCAN,
	54	WB_REASON_SYNC,
	55	WB_REASON_PERIODIC,
	56	WB_REASON_LAPTOP_TIMER,
	57	WB_REASON_FREE_MORE_MEM,
	58	WB_REASON_FS_FREE_SPACE,
	59	/*
	60	* There is no bdi forker thread any more and works are done
	61	* by emergency worker, however, this is TPs userland visible
	62	* and we'll be exposing exactly the same information,
	63	* so it has a mismatch name.
	64	*/
	65	WB_REASON_FORKER_THREAD,
	66
	67	WB_REASON_MAX,
	68	};
	69
52ebea74 TH	70	/*
	71	* For cgroup writeback, multiple wb's may map to the same blkcg. Those
	72	* wb's can operate mostly independently but should share the congested
	73	* state. To facilitate such sharing, the congested state is tracked using
	74	* the following struct which is created on demand, indexed by blkcg ID on
	75	* its bdi, and refcounted.
	76	*/
4aa9c692 TH	77	struct bdi_writeback_congested {
4aa9c692 TH	78	unsigned long state; /* WB_[a]sync_congested flags */
e58dd0de	79	refcount_t refcnt; /* nr of attached wb's and blkg */
52ebea74 TH	80
52ebea74 TH	81	#ifdef CONFIG_CGROUP_WRITEBACK
b7d680d7 JK	82	struct backing_dev_info __bdi; / the associated bdi, set to NULL
	83	* on bdi unregistration. For memcg-wb
	84	* internal use only! */
52ebea74 TH	85	int blkcg_id; /* ID of the associated blkcg */
	86	struct rb_node rb_node; /* on bdi->cgwb_congestion_tree */
	87	#endif
4aa9c692 TH	88	};
4aa9c692 TH	89
52ebea74 TH	90	/*
	91	* Each wb (bdi_writeback) can perform writeback operations, is measured
	92	* and throttled, independently. Without cgroup writeback, each bdi
	93	* (bdi_writeback) is served by its embedded bdi->wb.
	94	*
	95	* On the default hierarchy, blkcg implicitly enables memcg. This allows
	96	* using memcg's page ownership for attributing writeback IOs, and every
	97	* memcg - blkcg combination can be served by its own wb by assigning a
	98	* dedicated wb to each memcg, which enables isolation across different
	99	* cgroups and propagation of IO back pressure down from the IO layer upto
	100	* the tasks which are generating the dirty pages to be written back.
	101	*
	102	* A cgroup wb is indexed on its bdi by the ID of the associated memcg,
	103	* refcounted with the number of inodes attached to it, and pins the memcg
	104	* and the corresponding blkcg. As the corresponding blkcg for a memcg may
	105	* change as blkcg is disabled and enabled higher up in the hierarchy, a wb
	106	* is tested for blkcg after lookup and removed from index on mismatch so
	107	* that a new wb for the combination can be created.
	108	*/
66114cad TH	109	struct bdi_writeback {
	110	struct backing_dev_info bdi; / our parent bdi */
	111
	112	unsigned long state; /* Always use atomic bitops on this */
	113	unsigned long last_old_flush; /* last old data flush */
	114
	115	struct list_head b_dirty; /* dirty inodes */
	116	struct list_head b_io; /* parked for writeback */
	117	struct list_head b_more_io; /* parked for more writeback */
	118	struct list_head b_dirty_time; /* time stamps are dirty */
	119	spinlock_t list_lock; /* protects the b_* lists */
	120
	121	struct percpu_counter stat[NR_WB_STAT_ITEMS];
	122
4aa9c692 TH	123	struct bdi_writeback_congested *congested;
4aa9c692 TH	124
66114cad TH	125	unsigned long bw_time_stamp; /* last time write bw is updated */
	126	unsigned long dirtied_stamp;
	127	unsigned long written_stamp; /* pages written at bw_time_stamp */
	128	unsigned long write_bandwidth; /* the estimated write bandwidth */
95a46c65	129	unsigned long avg_write_bandwidth; /* further smoothed write bw, > 0 */
66114cad TH	130
	131	/*
	132	* The base dirty throttle rate, re-calculated on every 200ms.
	133	* All the bdi tasks' dirty rate will be curbed under it.
	134	* @dirty_ratelimit tracks the estimated @balanced_dirty_ratelimit
	135	* in small steps and is much more smooth/stable than the latter.
	136	*/
	137	unsigned long dirty_ratelimit;
	138	unsigned long balanced_dirty_ratelimit;
	139
	140	struct fprop_local_percpu completions;
	141	int dirty_exceeded;
85009b4f	142	enum wb_reason start_all_reason;
66114cad TH	143
	144	spinlock_t work_lock; /* protects work_list & dwork scheduling */
	145	struct list_head work_list;
	146	struct delayed_work dwork; /* work item used for writeback */
52ebea74	147
b57d74af JA	148	unsigned long dirty_sleep; /* last wait */
b57d74af JA	149
b817525a TH	150	struct list_head bdi_node; /* anchored at bdi->wb_list */
b817525a TH	151
52ebea74 TH	152	#ifdef CONFIG_CGROUP_WRITEBACK
52ebea74 TH	153	struct percpu_ref refcnt; /* used only for !root wb's */
841710aa	154	struct fprop_local_percpu memcg_completions;
52ebea74 TH	155	struct cgroup_subsys_state memcg_css; / the associated memcg */
	156	struct cgroup_subsys_state blkcg_css; / and blkcg */
	157	struct list_head memcg_node; /* anchored at memcg->cgwb_list */
	158	struct list_head blkcg_node; /* anchored at blkcg->cgwb_list */
	159
	160	union {
	161	struct work_struct release_work;
	162	struct rcu_head rcu;
	163	};
	164	#endif
66114cad TH	165	};
	166
	167	struct backing_dev_info {
	168	struct list_head bdi_list;
ea1754a0	169	unsigned long ra_pages; /* max readahead in PAGE_SIZE units */
9491ae4a	170	unsigned long io_pages; /* max allowed IO size */
66114cad TH	171	congested_fn congested_fn; / Function pointer if device is md/dm */
	172	void congested_data; / Pointer to aux data for congested func */
	173
fca39346	174	const char *name;
66114cad	175
d03f6cdc	176	struct kref refcnt; /* Reference counter for the structure */
8db378a5	177	unsigned int capabilities; /* Device capabilities */
66114cad TH	178	unsigned int min_ratio;
	179	unsigned int max_ratio, max_prop_frac;
	180
95a46c65 TH	181	/*
	182	* Sum of avg_write_bw of wbs with dirty inodes. > 0 if there are
	183	* any dirty wbs, which is depended upon by bdi_has_dirty().
	184	*/
	185	atomic_long_t tot_write_bandwidth;
766a9d6e	186
52ebea74	187	struct bdi_writeback wb; /* the root writeback info for this bdi */
b817525a	188	struct list_head wb_list; /* list of all wbs */
52ebea74 TH	189	#ifdef CONFIG_CGROUP_WRITEBACK
	190	struct radix_tree_root cgwb_tree; /* radix tree of active cgroup wbs */
	191	struct rb_root cgwb_congested_tree; /* their congested states */
3ee7e869	192	struct mutex cgwb_release_mutex; /* protect shutdown of wb structs */
a13f35e8 TH	193	#else
a13f35e8 TH	194	struct bdi_writeback_congested *wb_congested;
52ebea74	195	#endif
cc395d7f TH	196	wait_queue_head_t wb_waitq;
cc395d7f TH	197
66114cad	198	struct device *dev;
df08c32c	199	struct device *owner;
66114cad TH	200
	201	struct timer_list laptop_mode_wb_timer;
	202
	203	#ifdef CONFIG_DEBUG_FS
	204	struct dentry *debug_dir;
	205	struct dentry *debug_stats;
	206	#endif
	207	};
	208
	209	enum {
	210	BLK_RW_ASYNC = 0,
	211	BLK_RW_SYNC = 1,
	212	};
	213
ec8a6f26 TH	214	void clear_wb_congested(struct bdi_writeback_congested *congested, int sync);
	215	void set_wb_congested(struct bdi_writeback_congested *congested, int sync);
	216
	217	static inline void clear_bdi_congested(struct backing_dev_info *bdi, int sync)
	218	{
	219	clear_wb_congested(bdi->wb.congested, sync);
	220	}
	221
	222	static inline void set_bdi_congested(struct backing_dev_info *bdi, int sync)
	223	{
	224	set_wb_congested(bdi->wb.congested, sync);
	225	}
66114cad	226
2e898e4c GT	227	struct wb_lock_cookie {
	228	bool locked;
	229	unsigned long flags;
	230	};
	231
21c6321f TH	232	#ifdef CONFIG_CGROUP_WRITEBACK
	233
	234	/**
	235	* wb_tryget - try to increment a wb's refcount
	236	* @wb: bdi_writeback to get
	237	*/
	238	static inline bool wb_tryget(struct bdi_writeback *wb)
	239	{
	240	if (wb != &wb->bdi->wb)
	241	return percpu_ref_tryget(&wb->refcnt);
	242	return true;
	243	}
	244
	245	/**
	246	* wb_get - increment a wb's refcount
	247	* @wb: bdi_writeback to get
	248	*/
	249	static inline void wb_get(struct bdi_writeback *wb)
	250	{
	251	if (wb != &wb->bdi->wb)
	252	percpu_ref_get(&wb->refcnt);
	253	}
	254
	255	/**
	256	* wb_put - decrement a wb's refcount
	257	* @wb: bdi_writeback to put
	258	*/
	259	static inline void wb_put(struct bdi_writeback *wb)
	260	{
368686a9 AR	261	if (WARN_ON_ONCE(!wb->bdi)) {
	262	/*
	263	* A driver bug might cause a file to be removed before bdi was
	264	* initialized.
	265	*/
	266	return;
	267	}
	268
21c6321f TH	269	if (wb != &wb->bdi->wb)
	270	percpu_ref_put(&wb->refcnt);
	271	}
	272
e8a7abf5 TH	273	/**
	274	* wb_dying - is a wb dying?
	275	* @wb: bdi_writeback of interest
	276	*
	277	* Returns whether @wb is unlinked and being drained.
	278	*/
	279	static inline bool wb_dying(struct bdi_writeback *wb)
	280	{
	281	return percpu_ref_is_dying(&wb->refcnt);
	282	}
	283
21c6321f TH	284	#else /* CONFIG_CGROUP_WRITEBACK */
	285
	286	static inline bool wb_tryget(struct bdi_writeback *wb)
	287	{
	288	return true;
	289	}
	290
	291	static inline void wb_get(struct bdi_writeback *wb)
	292	{
	293	}
	294
	295	static inline void wb_put(struct bdi_writeback *wb)
	296	{
	297	}
	298
e8a7abf5 TH	299	static inline bool wb_dying(struct bdi_writeback *wb)
	300	{
	301	return false;
	302	}
	303
21c6321f TH	304	#endif /* CONFIG_CGROUP_WRITEBACK */
21c6321f TH	305
66114cad	306	#endif /* __LINUX_BACKING_DEV_DEFS_H */