[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>
#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_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_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_FOREIGN_FLUSH,

	WB_REASON_MAX,
};

struct wb_completion {
	atomic_t		cnt;
	wait_queue_head_t	*waitq;
};

#define __WB_COMPLETION_INIT(_waitq)	\
	(struct wb_completion){ .cnt = ATOMIC_INIT(1), .waitq = (_waitq) }

/*
 * If one wants to wait for one or more wb_writeback_works, each work's
 * ->done should be set to a wb_completion defined using the following
 * macro.  Once all work items are issued with wb_queue_work(), the caller
 * can wait for the completion of all using wb_wait_for_completion().  Work
 * items which are waited upon aren't freed automatically on completion.
 */
#define WB_COMPLETION_INIT(bdi)		__WB_COMPLETION_INIT(&(bdi)->wb_waitq)

#define DEFINE_WB_COMPLETION(cmpl, bdi)	\
	struct wb_completion cmpl = WB_COMPLETION_INIT(bdi)

/*
 * 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.
 *
 * Each bdi_writeback that is not embedded into the backing_dev_info must hold
 * a reference to the parent backing_dev_info.  See cgwb_create() for details.
 */
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 */

	atomic_t writeback_inodes;	/* number of inodes under writeback */
	struct percpu_counter stat[NR_WB_STAT_ITEMS];

	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 */
	struct delayed_work bw_dwork;	/* work item used for bandwidth estimate */

	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 */
	struct list_head b_attached;	/* attached inodes, protected by list_lock */
	struct list_head offline_node;	/* anchored at offline_cgwbs */

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

struct backing_dev_info {
	u64 id;
	struct rb_node rb_node; /* keyed by ->id */
	struct list_head bdi_list;
	unsigned long ra_pages;	/* max readahead in PAGE_SIZE units */
	unsigned long io_pages;	/* max allowed IO size */

	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;
	/*
	 * Jiffies when last process was dirty throttled on this bdi. Used by
	 * blk-wbt.
	 */
	unsigned long last_bdp_sleep;

	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 mutex cgwb_release_mutex;  /* protect shutdown of wb structs */
	struct rw_semaphore wb_switch_rwsem; /* no cgwb switch while syncing */
#endif
	wait_queue_head_t wb_waitq;

	struct device *dev;
	char dev_name[64];
	struct device *owner;

	struct timer_list laptop_mode_wb_timer;

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

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
 * @nr: number of references to put
 */
static inline void wb_put_many(struct bdi_writeback *wb, unsigned long nr)
{
	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_many(&wb->refcnt, nr);
}

/**
 * wb_put - decrement a wb's refcount
 * @wb: bdi_writeback to put
 */
static inline void wb_put(struct bdi_writeback *wb)
{
	wb_put_many(wb, 1);
}

/**
 * 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 void wb_put_many(struct bdi_writeback *wb, unsigned long nr)
{
}

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
66114cad TH	31	enum wb_stat_item {
	32	WB_RECLAIMABLE,
	33	WB_WRITEBACK,
	34	WB_DIRTIED,
	35	WB_WRITTEN,
	36	NR_WB_STAT_ITEMS
	37	};
	38
	39	#define WB_STAT_BATCH (8*(1+ilog2(nr_cpu_ids)))
	40
85009b4f JA	41	/*
	42	* why some writeback work was initiated
	43	*/
	44	enum wb_reason {
	45	WB_REASON_BACKGROUND,
	46	WB_REASON_VMSCAN,
	47	WB_REASON_SYNC,
	48	WB_REASON_PERIODIC,
	49	WB_REASON_LAPTOP_TIMER,
85009b4f JA	50	WB_REASON_FS_FREE_SPACE,
	51	/*
	52	* There is no bdi forker thread any more and works are done
	53	* by emergency worker, however, this is TPs userland visible
	54	* and we'll be exposing exactly the same information,
	55	* so it has a mismatch name.
	56	*/
	57	WB_REASON_FORKER_THREAD,
97b27821	58	WB_REASON_FOREIGN_FLUSH,
85009b4f JA	59
	60	WB_REASON_MAX,
	61	};
	62
5b9cce4c TH	63	struct wb_completion {
	64	atomic_t cnt;
	65	wait_queue_head_t *waitq;
	66	};
	67
	68	#define __WB_COMPLETION_INIT(_waitq) \
	69	(struct wb_completion){ .cnt = ATOMIC_INIT(1), .waitq = (_waitq) }
	70
	71	/*
	72	* If one wants to wait for one or more wb_writeback_works, each work's
	73	* ->done should be set to a wb_completion defined using the following
	74	* macro. Once all work items are issued with wb_queue_work(), the caller
	75	* can wait for the completion of all using wb_wait_for_completion(). Work
	76	* items which are waited upon aren't freed automatically on completion.
	77	*/
	78	#define WB_COMPLETION_INIT(bdi) __WB_COMPLETION_INIT(&(bdi)->wb_waitq)
	79
	80	#define DEFINE_WB_COMPLETION(cmpl, bdi) \
	81	struct wb_completion cmpl = WB_COMPLETION_INIT(bdi)
	82
52ebea74 TH	83	/*
	84	* Each wb (bdi_writeback) can perform writeback operations, is measured
	85	* and throttled, independently. Without cgroup writeback, each bdi
	86	* (bdi_writeback) is served by its embedded bdi->wb.
	87	*
	88	* On the default hierarchy, blkcg implicitly enables memcg. This allows
	89	* using memcg's page ownership for attributing writeback IOs, and every
	90	* memcg - blkcg combination can be served by its own wb by assigning a
	91	* dedicated wb to each memcg, which enables isolation across different
	92	* cgroups and propagation of IO back pressure down from the IO layer upto
	93	* the tasks which are generating the dirty pages to be written back.
	94	*
	95	* A cgroup wb is indexed on its bdi by the ID of the associated memcg,
	96	* refcounted with the number of inodes attached to it, and pins the memcg
	97	* and the corresponding blkcg. As the corresponding blkcg for a memcg may
	98	* change as blkcg is disabled and enabled higher up in the hierarchy, a wb
	99	* is tested for blkcg after lookup and removed from index on mismatch so
	100	* that a new wb for the combination can be created.
efee1713 CH	101	*
	102	* Each bdi_writeback that is not embedded into the backing_dev_info must hold
	103	* a reference to the parent backing_dev_info. See cgwb_create() for details.
52ebea74	104	*/
66114cad TH	105	struct bdi_writeback {
	106	struct backing_dev_info bdi; / our parent bdi */
	107
	108	unsigned long state; /* Always use atomic bitops on this */
	109	unsigned long last_old_flush; /* last old data flush */
	110
	111	struct list_head b_dirty; /* dirty inodes */
	112	struct list_head b_io; /* parked for writeback */
	113	struct list_head b_more_io; /* parked for more writeback */
	114	struct list_head b_dirty_time; /* time stamps are dirty */
	115	spinlock_t list_lock; /* protects the b_* lists */
	116
633a2abb	117	atomic_t writeback_inodes; /* number of inodes under writeback */
66114cad TH	118	struct percpu_counter stat[NR_WB_STAT_ITEMS];
	119
	120	unsigned long bw_time_stamp; /* last time write bw is updated */
	121	unsigned long dirtied_stamp;
	122	unsigned long written_stamp; /* pages written at bw_time_stamp */
	123	unsigned long write_bandwidth; /* the estimated write bandwidth */
95a46c65	124	unsigned long avg_write_bandwidth; /* further smoothed write bw, > 0 */
66114cad TH	125
	126	/*
	127	* The base dirty throttle rate, re-calculated on every 200ms.
	128	* All the bdi tasks' dirty rate will be curbed under it.
	129	* @dirty_ratelimit tracks the estimated @balanced_dirty_ratelimit
	130	* in small steps and is much more smooth/stable than the latter.
	131	*/
	132	unsigned long dirty_ratelimit;
	133	unsigned long balanced_dirty_ratelimit;
	134
	135	struct fprop_local_percpu completions;
	136	int dirty_exceeded;
85009b4f	137	enum wb_reason start_all_reason;
66114cad TH	138
	139	spinlock_t work_lock; /* protects work_list & dwork scheduling */
	140	struct list_head work_list;
	141	struct delayed_work dwork; /* work item used for writeback */
45a2966f	142	struct delayed_work bw_dwork; /* work item used for bandwidth estimate */
52ebea74	143
b817525a TH	144	struct list_head bdi_node; /* anchored at bdi->wb_list */
b817525a TH	145
52ebea74 TH	146	#ifdef CONFIG_CGROUP_WRITEBACK
52ebea74 TH	147	struct percpu_ref refcnt; /* used only for !root wb's */
841710aa	148	struct fprop_local_percpu memcg_completions;
52ebea74 TH	149	struct cgroup_subsys_state memcg_css; / the associated memcg */
	150	struct cgroup_subsys_state blkcg_css; / and blkcg */
	151	struct list_head memcg_node; /* anchored at memcg->cgwb_list */
	152	struct list_head blkcg_node; /* anchored at blkcg->cgwb_list */
f3b6a6df	153	struct list_head b_attached; /* attached inodes, protected by list_lock */
c22d70a1	154	struct list_head offline_node; /* anchored at offline_cgwbs */
52ebea74 TH	155
	156	union {
	157	struct work_struct release_work;
	158	struct rcu_head rcu;
	159	};
	160	#endif
66114cad TH	161	};
	162
	163	struct backing_dev_info {
34f8fe50 TH	164	u64 id;
34f8fe50 TH	165	struct rb_node rb_node; /* keyed by ->id */
66114cad	166	struct list_head bdi_list;
ea1754a0	167	unsigned long ra_pages; /* max readahead in PAGE_SIZE units */
9491ae4a	168	unsigned long io_pages; /* max allowed IO size */
66114cad	169
d03f6cdc	170	struct kref refcnt; /* Reference counter for the structure */
8db378a5	171	unsigned int capabilities; /* Device capabilities */
66114cad TH	172	unsigned int min_ratio;
	173	unsigned int max_ratio, max_prop_frac;
	174
95a46c65 TH	175	/*
	176	* Sum of avg_write_bw of wbs with dirty inodes. > 0 if there are
	177	* any dirty wbs, which is depended upon by bdi_has_dirty().
	178	*/
	179	atomic_long_t tot_write_bandwidth;
f814bdda JK	180	/*
	181	* Jiffies when last process was dirty throttled on this bdi. Used by
	182	* blk-wbt.
	183	*/
	184	unsigned long last_bdp_sleep;
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
52ebea74 TH	189	struct radix_tree_root cgwb_tree; /* radix tree of active cgroup wbs */
3ee7e869	190	struct mutex cgwb_release_mutex; /* protect shutdown of wb structs */
7fc5854f	191	struct rw_semaphore wb_switch_rwsem; /* no cgwb switch while syncing */
52ebea74	192	#endif
cc395d7f TH	193	wait_queue_head_t wb_waitq;
cc395d7f TH	194
66114cad	195	struct device *dev;
6bd87eec	196	char dev_name[64];
df08c32c	197	struct device *owner;
66114cad TH	198
	199	struct timer_list laptop_mode_wb_timer;
	200
	201	#ifdef CONFIG_DEBUG_FS
	202	struct dentry *debug_dir;
66114cad TH	203	#endif
	204	};
	205
2e898e4c GT	206	struct wb_lock_cookie {
	207	bool locked;
	208	unsigned long flags;
	209	};
	210
21c6321f TH	211	#ifdef CONFIG_CGROUP_WRITEBACK
	212
	213	/**
	214	* wb_tryget - try to increment a wb's refcount
	215	* @wb: bdi_writeback to get
	216	*/
	217	static inline bool wb_tryget(struct bdi_writeback *wb)
	218	{
	219	if (wb != &wb->bdi->wb)
	220	return percpu_ref_tryget(&wb->refcnt);
	221	return true;
	222	}
	223
	224	/**
	225	* wb_get - increment a wb's refcount
	226	* @wb: bdi_writeback to get
	227	*/
	228	static inline void wb_get(struct bdi_writeback *wb)
	229	{
	230	if (wb != &wb->bdi->wb)
	231	percpu_ref_get(&wb->refcnt);
	232	}
	233
	234	/**
	235	* wb_put - decrement a wb's refcount
	236	* @wb: bdi_writeback to put
f5fbe6b7	237	* @nr: number of references to put
21c6321f	238	*/
f5fbe6b7	239	static inline void wb_put_many(struct bdi_writeback *wb, unsigned long nr)
21c6321f	240	{
368686a9 AR	241	if (WARN_ON_ONCE(!wb->bdi)) {
	242	/*
	243	* A driver bug might cause a file to be removed before bdi was
	244	* initialized.
	245	*/
	246	return;
	247	}
	248
21c6321f	249	if (wb != &wb->bdi->wb)
f5fbe6b7 RG	250	percpu_ref_put_many(&wb->refcnt, nr);
	251	}
	252
	253	/**
	254	* wb_put - decrement a wb's refcount
	255	* @wb: bdi_writeback to put
	256	*/
	257	static inline void wb_put(struct bdi_writeback *wb)
	258	{
	259	wb_put_many(wb, 1);
21c6321f TH	260	}
21c6321f TH	261
e8a7abf5 TH	262	/**
	263	* wb_dying - is a wb dying?
	264	* @wb: bdi_writeback of interest
	265	*
	266	* Returns whether @wb is unlinked and being drained.
	267	*/
	268	static inline bool wb_dying(struct bdi_writeback *wb)
	269	{
	270	return percpu_ref_is_dying(&wb->refcnt);
	271	}
	272
21c6321f TH	273	#else /* CONFIG_CGROUP_WRITEBACK */
	274
	275	static inline bool wb_tryget(struct bdi_writeback *wb)
	276	{
	277	return true;
	278	}
	279
	280	static inline void wb_get(struct bdi_writeback *wb)
	281	{
	282	}
	283
	284	static inline void wb_put(struct bdi_writeback *wb)
	285	{
	286	}
	287
f5fbe6b7 RG	288	static inline void wb_put_many(struct bdi_writeback *wb, unsigned long nr)
	289	{
	290	}
	291
e8a7abf5 TH	292	static inline bool wb_dying(struct bdi_writeback *wb)
	293	{
	294	return false;
	295	}
	296
21c6321f TH	297	#endif /* CONFIG_CGROUP_WRITEBACK */
21c6321f TH	298
66114cad	299	#endif /* __LINUX_BACKING_DEV_DEFS_H */