[linux-2.6-block.git] / fs / btrfs / delayed-ref.h

/* SPDX-License-Identifier: GPL-2.0 */
/*
 * Copyright (C) 2008 Oracle.  All rights reserved.
 */

#ifndef BTRFS_DELAYED_REF_H
#define BTRFS_DELAYED_REF_H

#include <linux/refcount.h>

/* these are the possible values of struct btrfs_delayed_ref_node->action */
#define BTRFS_ADD_DELAYED_REF    1 /* add one backref to the tree */
#define BTRFS_DROP_DELAYED_REF   2 /* delete one backref from the tree */
#define BTRFS_ADD_DELAYED_EXTENT 3 /* record a full extent allocation */
#define BTRFS_UPDATE_DELAYED_HEAD 4 /* not changing ref count on head ref */

struct btrfs_delayed_ref_node {
	struct rb_node ref_node;
	/*
	 * If action is BTRFS_ADD_DELAYED_REF, also link this node to
	 * ref_head->ref_add_list, then we do not need to iterate the
	 * whole ref_head->ref_list to find BTRFS_ADD_DELAYED_REF nodes.
	 */
	struct list_head add_list;

	/* the starting bytenr of the extent */
	u64 bytenr;

	/* the size of the extent */
	u64 num_bytes;

	/* seq number to keep track of insertion order */
	u64 seq;

	/* ref count on this data structure */
	refcount_t refs;

	/*
	 * how many refs is this entry adding or deleting.  For
	 * head refs, this may be a negative number because it is keeping
	 * track of the total mods done to the reference count.
	 * For individual refs, this will always be a positive number
	 *
	 * It may be more than one, since it is possible for a single
	 * parent to have more than one ref on an extent
	 */
	int ref_mod;

	unsigned int action:8;
	unsigned int type:8;
	/* is this node still in the rbtree? */
	unsigned int is_head:1;
	unsigned int in_tree:1;
};

struct btrfs_delayed_extent_op {
	struct btrfs_disk_key key;
	u8 level;
	bool update_key;
	bool update_flags;
	bool is_data;
	u64 flags_to_set;
};

/*
 * the head refs are used to hold a lock on a given extent, which allows us
 * to make sure that only one process is running the delayed refs
 * at a time for a single extent.  They also store the sum of all the
 * reference count modifications we've queued up.
 */
struct btrfs_delayed_ref_head {
	u64 bytenr;
	u64 num_bytes;
	refcount_t refs;
	/*
	 * the mutex is held while running the refs, and it is also
	 * held when checking the sum of reference modifications.
	 */
	struct mutex mutex;

	spinlock_t lock;
	struct rb_root ref_tree;
	/* accumulate add BTRFS_ADD_DELAYED_REF nodes to this ref_add_list. */
	struct list_head ref_add_list;

	struct rb_node href_node;

	struct btrfs_delayed_extent_op *extent_op;

	/*
	 * This is used to track the final ref_mod from all the refs associated
	 * with this head ref, this is not adjusted as delayed refs are run,
	 * this is meant to track if we need to do the csum accounting or not.
	 */
	int total_ref_mod;

	/*
	 * This is the current outstanding mod references for this bytenr.  This
	 * is used with lookup_extent_info to get an accurate reference count
	 * for a bytenr, so it is adjusted as delayed refs are run so that any
	 * on disk reference count + ref_mod is accurate.
	 */
	int ref_mod;

	/*
	 * For qgroup reserved space freeing.
	 *
	 * ref_root and reserved will be recorded after
	 * BTRFS_ADD_DELAYED_EXTENT is called.
	 * And will be used to free reserved qgroup space at
	 * run_delayed_refs() time.
	 */
	u64 qgroup_ref_root;
	u64 qgroup_reserved;

	/*
	 * when a new extent is allocated, it is just reserved in memory
	 * The actual extent isn't inserted into the extent allocation tree
	 * until the delayed ref is processed.  must_insert_reserved is
	 * used to flag a delayed ref so the accounting can be updated
	 * when a full insert is done.
	 *
	 * It is possible the extent will be freed before it is ever
	 * inserted into the extent allocation tree.  In this case
	 * we need to update the in ram accounting to properly reflect
	 * the free has happened.
	 */
	unsigned int must_insert_reserved:1;
	unsigned int is_data:1;
	unsigned int is_system:1;
	unsigned int processing:1;
};

struct btrfs_delayed_tree_ref {
	struct btrfs_delayed_ref_node node;
	u64 root;
	u64 parent;
	int level;
};

struct btrfs_delayed_data_ref {
	struct btrfs_delayed_ref_node node;
	u64 root;
	u64 parent;
	u64 objectid;
	u64 offset;
};

struct btrfs_delayed_ref_root {
	/* head ref rbtree */
	struct rb_root href_root;

	/* dirty extent records */
	struct rb_root dirty_extent_root;

	/* this spin lock protects the rbtree and the entries inside */
	spinlock_t lock;

	/* how many delayed ref updates we've queued, used by the
	 * throttling code
	 */
	atomic_t num_entries;

	/* total number of head nodes in tree */
	unsigned long num_heads;

	/* total number of head nodes ready for processing */
	unsigned long num_heads_ready;

	u64 pending_csums;

	/*
	 * set when the tree is flushing before a transaction commit,
	 * used by the throttling code to decide if new updates need
	 * to be run right away
	 */
	int flushing;

	u64 run_delayed_start;

	/*
	 * To make qgroup to skip given root.
	 * This is for snapshot, as btrfs_qgroup_inherit() will manually
	 * modify counters for snapshot and its source, so we should skip
	 * the snapshot in new_root/old_roots or it will get calculated twice
	 */
	u64 qgroup_to_skip;
};

extern struct kmem_cache *btrfs_delayed_ref_head_cachep;
extern struct kmem_cache *btrfs_delayed_tree_ref_cachep;
extern struct kmem_cache *btrfs_delayed_data_ref_cachep;
extern struct kmem_cache *btrfs_delayed_extent_op_cachep;

int __init btrfs_delayed_ref_init(void);
void __cold btrfs_delayed_ref_exit(void);

static inline struct btrfs_delayed_extent_op *
btrfs_alloc_delayed_extent_op(void)
{
	return kmem_cache_alloc(btrfs_delayed_extent_op_cachep, GFP_NOFS);
}

static inline void
btrfs_free_delayed_extent_op(struct btrfs_delayed_extent_op *op)
{
	if (op)
		kmem_cache_free(btrfs_delayed_extent_op_cachep, op);
}

static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref)
{
	WARN_ON(refcount_read(&ref->refs) == 0);
	if (refcount_dec_and_test(&ref->refs)) {
		WARN_ON(ref->in_tree);
		switch (ref->type) {
		case BTRFS_TREE_BLOCK_REF_KEY:
		case BTRFS_SHARED_BLOCK_REF_KEY:
			kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref);
			break;
		case BTRFS_EXTENT_DATA_REF_KEY:
		case BTRFS_SHARED_DATA_REF_KEY:
			kmem_cache_free(btrfs_delayed_data_ref_cachep, ref);
			break;
		default:
			BUG();
		}
	}
}

static inline void btrfs_put_delayed_ref_head(struct btrfs_delayed_ref_head *head)
{
	if (refcount_dec_and_test(&head->refs))
		kmem_cache_free(btrfs_delayed_ref_head_cachep, head);
}

int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
			       struct btrfs_trans_handle *trans,
			       u64 bytenr, u64 num_bytes, u64 parent,
			       u64 ref_root, int level, int action,
			       struct btrfs_delayed_extent_op *extent_op,
			       int *old_ref_mod, int *new_ref_mod);
int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
			       struct btrfs_trans_handle *trans,
			       u64 bytenr, u64 num_bytes,
			       u64 parent, u64 ref_root,
			       u64 owner, u64 offset, u64 reserved, int action,
			       int *old_ref_mod, int *new_ref_mod);
int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
				struct btrfs_trans_handle *trans,
				u64 bytenr, u64 num_bytes,
				struct btrfs_delayed_extent_op *extent_op);
void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans,
			      struct btrfs_fs_info *fs_info,
			      struct btrfs_delayed_ref_root *delayed_refs,
			      struct btrfs_delayed_ref_head *head);

struct btrfs_delayed_ref_head *
btrfs_find_delayed_ref_head(struct btrfs_delayed_ref_root *delayed_refs,
			    u64 bytenr);
int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans,
			   struct btrfs_delayed_ref_head *head);
static inline void btrfs_delayed_ref_unlock(struct btrfs_delayed_ref_head *head)
{
	mutex_unlock(&head->mutex);
}


struct btrfs_delayed_ref_head *
btrfs_select_ref_head(struct btrfs_trans_handle *trans);

int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,
			    struct btrfs_delayed_ref_root *delayed_refs,
			    u64 seq);

/*
 * helper functions to cast a node into its container
 */
static inline struct btrfs_delayed_tree_ref *
btrfs_delayed_node_to_tree_ref(struct btrfs_delayed_ref_node *node)
{
	return container_of(node, struct btrfs_delayed_tree_ref, node);
}

static inline struct btrfs_delayed_data_ref *
btrfs_delayed_node_to_data_ref(struct btrfs_delayed_ref_node *node)
{
	return container_of(node, struct btrfs_delayed_data_ref, node);
}

#endif
Commit	Line	Data
9888c340	1	/* SPDX-License-Identifier: GPL-2.0 */
56bec294 CM	2	/*
56bec294 CM	3	* Copyright (C) 2008 Oracle. All rights reserved.
56bec294	4	*/
9888c340 DS	5
	6	#ifndef BTRFS_DELAYED_REF_H
	7	#define BTRFS_DELAYED_REF_H
56bec294	8
6df8cdf5 ER	9	#include <linux/refcount.h>
6df8cdf5 ER	10
44a075bd	11	/* these are the possible values of struct btrfs_delayed_ref_node->action */
56bec294 CM	12	#define BTRFS_ADD_DELAYED_REF 1 /* add one backref to the tree */
	13	#define BTRFS_DROP_DELAYED_REF 2 /* delete one backref from the tree */
	14	#define BTRFS_ADD_DELAYED_EXTENT 3 /* record a full extent allocation */
1a81af4d	15	#define BTRFS_UPDATE_DELAYED_HEAD 4 /* not changing ref count on head ref */
56bec294 CM	16
56bec294 CM	17	struct btrfs_delayed_ref_node {
0e0adbcf	18	struct rb_node ref_node;
1d57ee94 WX	19	/*
	20	* If action is BTRFS_ADD_DELAYED_REF, also link this node to
	21	* ref_head->ref_add_list, then we do not need to iterate the
	22	* whole ref_head->ref_list to find BTRFS_ADD_DELAYED_REF nodes.
	23	*/
	24	struct list_head add_list;
c6fc2454	25
56bec294 CM	26	/* the starting bytenr of the extent */
	27	u64 bytenr;
	28
56bec294 CM	29	/* the size of the extent */
	30	u64 num_bytes;
	31
00f04b88 AJ	32	/* seq number to keep track of insertion order */
	33	u64 seq;
	34
56bec294	35	/* ref count on this data structure */
6df8cdf5	36	refcount_t refs;
56bec294 CM	37
	38	/*
	39	* how many refs is this entry adding or deleting. For
	40	* head refs, this may be a negative number because it is keeping
	41	* track of the total mods done to the reference count.
	42	* For individual refs, this will always be a positive number
	43	*
	44	* It may be more than one, since it is possible for a single
	45	* parent to have more than one ref on an extent
	46	*/
	47	int ref_mod;
	48
5d4f98a2 YZ	49	unsigned int action:8;
5d4f98a2 YZ	50	unsigned int type:8;
56bec294	51	/* is this node still in the rbtree? */
5d4f98a2	52	unsigned int is_head:1;
56bec294 CM	53	unsigned int in_tree:1;
	54	};
	55
5d4f98a2 YZ	56	struct btrfs_delayed_extent_op {
5d4f98a2 YZ	57	struct btrfs_disk_key key;
35b3ad50 DS	58	u8 level;
	59	bool update_key;
	60	bool update_flags;
	61	bool is_data;
5d4f98a2	62	u64 flags_to_set;
5d4f98a2 YZ	63	};
5d4f98a2 YZ	64
56bec294 CM	65	/*
	66	* the head refs are used to hold a lock on a given extent, which allows us
	67	* to make sure that only one process is running the delayed refs
	68	* at a time for a single extent. They also store the sum of all the
	69	* reference count modifications we've queued up.
	70	*/
	71	struct btrfs_delayed_ref_head {
d278850e JB	72	u64 bytenr;
	73	u64 num_bytes;
	74	refcount_t refs;
56bec294 CM	75	/*
	76	* the mutex is held while running the refs, and it is also
	77	* held when checking the sum of reference modifications.
	78	*/
	79	struct mutex mutex;
	80
d7df2c79	81	spinlock_t lock;
0e0adbcf	82	struct rb_root ref_tree;
1d57ee94 WX	83	/* accumulate add BTRFS_ADD_DELAYED_REF nodes to this ref_add_list. */
1d57ee94 WX	84	struct list_head ref_add_list;
c3e69d58	85
c46effa6 LB	86	struct rb_node href_node;
c46effa6 LB	87
5d4f98a2	88	struct btrfs_delayed_extent_op *extent_op;
1262133b JB	89
	90	/*
	91	* This is used to track the final ref_mod from all the refs associated
	92	* with this head ref, this is not adjusted as delayed refs are run,
	93	* this is meant to track if we need to do the csum accounting or not.
	94	*/
	95	int total_ref_mod;
	96
d278850e JB	97	/*
	98	* This is the current outstanding mod references for this bytenr. This
	99	* is used with lookup_extent_info to get an accurate reference count
	100	* for a bytenr, so it is adjusted as delayed refs are run so that any
	101	* on disk reference count + ref_mod is accurate.
	102	*/
	103	int ref_mod;
	104
f64d5ca8 QW	105	/*
	106	* For qgroup reserved space freeing.
	107	*
	108	* ref_root and reserved will be recorded after
	109	* BTRFS_ADD_DELAYED_EXTENT is called.
	110	* And will be used to free reserved qgroup space at
	111	* run_delayed_refs() time.
	112	*/
	113	u64 qgroup_ref_root;
	114	u64 qgroup_reserved;
	115
56bec294 CM	116	/*
	117	* when a new extent is allocated, it is just reserved in memory
	118	* The actual extent isn't inserted into the extent allocation tree
	119	* until the delayed ref is processed. must_insert_reserved is
	120	* used to flag a delayed ref so the accounting can be updated
	121	* when a full insert is done.
	122	*
	123	* It is possible the extent will be freed before it is ever
	124	* inserted into the extent allocation tree. In this case
	125	* we need to update the in ram accounting to properly reflect
	126	* the free has happened.
	127	*/
	128	unsigned int must_insert_reserved:1;
5d4f98a2	129	unsigned int is_data:1;
5e388e95	130	unsigned int is_system:1;
d7df2c79	131	unsigned int processing:1;
56bec294 CM	132	};
56bec294 CM	133
5d4f98a2	134	struct btrfs_delayed_tree_ref {
56bec294	135	struct btrfs_delayed_ref_node node;
eebe063b AJ	136	u64 root;
eebe063b AJ	137	u64 parent;
5d4f98a2 YZ	138	int level;
5d4f98a2 YZ	139	};
56bec294	140
5d4f98a2 YZ	141	struct btrfs_delayed_data_ref {
5d4f98a2 YZ	142	struct btrfs_delayed_ref_node node;
eebe063b AJ	143	u64 root;
eebe063b AJ	144	u64 parent;
5d4f98a2 YZ	145	u64 objectid;
5d4f98a2 YZ	146	u64 offset;
56bec294 CM	147	};
	148
	149	struct btrfs_delayed_ref_root {
c46effa6 LB	150	/* head ref rbtree */
	151	struct rb_root href_root;
	152
3368d001 QW	153	/* dirty extent records */
	154	struct rb_root dirty_extent_root;
	155
56bec294 CM	156	/* this spin lock protects the rbtree and the entries inside */
	157	spinlock_t lock;
	158
	159	/* how many delayed ref updates we've queued, used by the
	160	* throttling code
	161	*/
d7df2c79	162	atomic_t num_entries;
56bec294	163
c3e69d58 CM	164	/* total number of head nodes in tree */
	165	unsigned long num_heads;
	166
	167	/* total number of head nodes ready for processing */
	168	unsigned long num_heads_ready;
	169
1262133b JB	170	u64 pending_csums;
1262133b JB	171
56bec294 CM	172	/*
	173	* set when the tree is flushing before a transaction commit,
	174	* used by the throttling code to decide if new updates need
	175	* to be run right away
	176	*/
	177	int flushing;
c3e69d58 CM	178
c3e69d58 CM	179	u64 run_delayed_start;
9086db86 QW	180
	181	/*
	182	* To make qgroup to skip given root.
01327610	183	* This is for snapshot, as btrfs_qgroup_inherit() will manually
9086db86 QW	184	* modify counters for snapshot and its source, so we should skip
	185	* the snapshot in new_root/old_roots or it will get calculated twice
	186	*/
	187	u64 qgroup_to_skip;
56bec294 CM	188	};
56bec294 CM	189
78a6184a MX	190	extern struct kmem_cache *btrfs_delayed_ref_head_cachep;
	191	extern struct kmem_cache *btrfs_delayed_tree_ref_cachep;
	192	extern struct kmem_cache *btrfs_delayed_data_ref_cachep;
	193	extern struct kmem_cache *btrfs_delayed_extent_op_cachep;
	194
f5c29bd9	195	int __init btrfs_delayed_ref_init(void);
e67c718b	196	void __cold btrfs_delayed_ref_exit(void);
78a6184a MX	197
	198	static inline struct btrfs_delayed_extent_op *
	199	btrfs_alloc_delayed_extent_op(void)
	200	{
	201	return kmem_cache_alloc(btrfs_delayed_extent_op_cachep, GFP_NOFS);
	202	}
	203
	204	static inline void
	205	btrfs_free_delayed_extent_op(struct btrfs_delayed_extent_op *op)
	206	{
	207	if (op)
	208	kmem_cache_free(btrfs_delayed_extent_op_cachep, op);
	209	}
	210
56bec294 CM	211	static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref)
56bec294 CM	212	{
6df8cdf5 ER	213	WARN_ON(refcount_read(&ref->refs) == 0);
6df8cdf5 ER	214	if (refcount_dec_and_test(&ref->refs)) {
56bec294	215	WARN_ON(ref->in_tree);
78a6184a MX	216	switch (ref->type) {
	217	case BTRFS_TREE_BLOCK_REF_KEY:
	218	case BTRFS_SHARED_BLOCK_REF_KEY:
	219	kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref);
	220	break;
	221	case BTRFS_EXTENT_DATA_REF_KEY:
	222	case BTRFS_SHARED_DATA_REF_KEY:
	223	kmem_cache_free(btrfs_delayed_data_ref_cachep, ref);
	224	break;
78a6184a MX	225	default:
	226	BUG();
	227	}
56bec294 CM	228	}
	229	}
	230
d278850e JB	231	static inline void btrfs_put_delayed_ref_head(struct btrfs_delayed_ref_head *head)
	232	{
	233	if (refcount_dec_and_test(&head->refs))
	234	kmem_cache_free(btrfs_delayed_ref_head_cachep, head);
	235	}
	236
66d7e7f0 AJ	237	int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
66d7e7f0 AJ	238	struct btrfs_trans_handle *trans,
5d4f98a2 YZ	239	u64 bytenr, u64 num_bytes, u64 parent,
5d4f98a2 YZ	240	u64 ref_root, int level, int action,
7be07912 OS	241	struct btrfs_delayed_extent_op *extent_op,
7be07912 OS	242	int old_ref_mod, int new_ref_mod);
66d7e7f0 AJ	243	int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
66d7e7f0 AJ	244	struct btrfs_trans_handle *trans,
5d4f98a2 YZ	245	u64 bytenr, u64 num_bytes,
5d4f98a2 YZ	246	u64 parent, u64 ref_root,
7be07912 OS	247	u64 owner, u64 offset, u64 reserved, int action,
7be07912 OS	248	int old_ref_mod, int new_ref_mod);
66d7e7f0 AJ	249	int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
66d7e7f0 AJ	250	struct btrfs_trans_handle *trans,
5d4f98a2 YZ	251	u64 bytenr, u64 num_bytes,
5d4f98a2 YZ	252	struct btrfs_delayed_extent_op *extent_op);
ae1e206b JB	253	void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans,
	254	struct btrfs_fs_info *fs_info,
	255	struct btrfs_delayed_ref_root *delayed_refs,
	256	struct btrfs_delayed_ref_head *head);
56bec294	257
1887be66	258	struct btrfs_delayed_ref_head *
f72ad18e LB	259	btrfs_find_delayed_ref_head(struct btrfs_delayed_ref_root *delayed_refs,
f72ad18e LB	260	u64 bytenr);
c3e69d58 CM	261	int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans,
c3e69d58 CM	262	struct btrfs_delayed_ref_head *head);
093486c4 MX	263	static inline void btrfs_delayed_ref_unlock(struct btrfs_delayed_ref_head *head)
	264	{
	265	mutex_unlock(&head->mutex);
	266	}
	267
d7df2c79 JB	268
	269	struct btrfs_delayed_ref_head *
	270	btrfs_select_ref_head(struct btrfs_trans_handle *trans);
00f04b88	271
097b8a7c JS	272	int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,
097b8a7c JS	273	struct btrfs_delayed_ref_root *delayed_refs,
00f04b88 AJ	274	u64 seq);
00f04b88 AJ	275
56bec294 CM	276	/*
	277	* helper functions to cast a node into its container
	278	*/
5d4f98a2 YZ	279	static inline struct btrfs_delayed_tree_ref *
5d4f98a2 YZ	280	btrfs_delayed_node_to_tree_ref(struct btrfs_delayed_ref_node *node)
56bec294	281	{
5d4f98a2 YZ	282	return container_of(node, struct btrfs_delayed_tree_ref, node);
5d4f98a2 YZ	283	}
56bec294	284
5d4f98a2 YZ	285	static inline struct btrfs_delayed_data_ref *
	286	btrfs_delayed_node_to_data_ref(struct btrfs_delayed_ref_node *node)
	287	{
5d4f98a2	288	return container_of(node, struct btrfs_delayed_data_ref, node);
56bec294	289	}
9888c340	290
56bec294	291	#endif