[linux-2.6-block.git] / include / linux / ceph / osd_client.h

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

#include <linux/bitrev.h>
#include <linux/completion.h>
#include <linux/kref.h>
#include <linux/mempool.h>
#include <linux/rbtree.h>
#include <linux/refcount.h>

#include <linux/ceph/types.h>
#include <linux/ceph/osdmap.h>
#include <linux/ceph/messenger.h>
#include <linux/ceph/msgpool.h>
#include <linux/ceph/auth.h>
#include <linux/ceph/pagelist.h>

struct ceph_msg;
struct ceph_snap_context;
struct ceph_osd_request;
struct ceph_osd_client;

/*
 * completion callback for async writepages
 */
typedef void (*ceph_osdc_callback_t)(struct ceph_osd_request *);

#define CEPH_HOMELESS_OSD	-1

/* a given osd we're communicating with */
struct ceph_osd {
	refcount_t o_ref;
	struct ceph_osd_client *o_osdc;
	int o_osd;
	int o_incarnation;
	struct rb_node o_node;
	struct ceph_connection o_con;
	struct rb_root o_requests;
	struct rb_root o_linger_requests;
	struct rb_root o_backoff_mappings;
	struct rb_root o_backoffs_by_id;
	struct list_head o_osd_lru;
	struct ceph_auth_handshake o_auth;
	unsigned long lru_ttl;
	struct list_head o_keepalive_item;
	struct mutex lock;
};

#define CEPH_OSD_SLAB_OPS	2
#define CEPH_OSD_MAX_OPS	16

enum ceph_osd_data_type {
	CEPH_OSD_DATA_TYPE_NONE = 0,
	CEPH_OSD_DATA_TYPE_PAGES,
	CEPH_OSD_DATA_TYPE_PAGELIST,
#ifdef CONFIG_BLOCK
	CEPH_OSD_DATA_TYPE_BIO,
#endif /* CONFIG_BLOCK */
	CEPH_OSD_DATA_TYPE_BVECS,
};

struct ceph_osd_data {
	enum ceph_osd_data_type	type;
	union {
		struct {
			struct page	**pages;
			u64		length;
			u32		alignment;
			bool		pages_from_pool;
			bool		own_pages;
		};
		struct ceph_pagelist	*pagelist;
#ifdef CONFIG_BLOCK
		struct {
			struct ceph_bio_iter	bio_pos;
			u32			bio_length;
		};
#endif /* CONFIG_BLOCK */
		struct {
			struct ceph_bvec_iter	bvec_pos;
			u32			num_bvecs;
		};
	};
};

struct ceph_osd_req_op {
	u16 op;           /* CEPH_OSD_OP_* */
	u32 flags;        /* CEPH_OSD_OP_FLAG_* */
	u32 indata_len;   /* request */
	u32 outdata_len;  /* reply */
	s32 rval;

	union {
		struct ceph_osd_data raw_data_in;
		struct {
			u64 offset, length;
			u64 truncate_size;
			u32 truncate_seq;
			struct ceph_osd_data osd_data;
		} extent;
		struct {
			u32 name_len;
			u32 value_len;
			__u8 cmp_op;       /* CEPH_OSD_CMPXATTR_OP_* */
			__u8 cmp_mode;     /* CEPH_OSD_CMPXATTR_MODE_* */
			struct ceph_osd_data osd_data;
		} xattr;
		struct {
			const char *class_name;
			const char *method_name;
			struct ceph_osd_data request_info;
			struct ceph_osd_data request_data;
			struct ceph_osd_data response_data;
			__u8 class_len;
			__u8 method_len;
			u32 indata_len;
		} cls;
		struct {
			u64 cookie;
			__u8 op;           /* CEPH_OSD_WATCH_OP_ */
			u32 gen;
		} watch;
		struct {
			struct ceph_osd_data request_data;
		} notify_ack;
		struct {
			u64 cookie;
			struct ceph_osd_data request_data;
			struct ceph_osd_data response_data;
		} notify;
		struct {
			struct ceph_osd_data response_data;
		} list_watchers;
		struct {
			u64 expected_object_size;
			u64 expected_write_size;
		} alloc_hint;
	};
};

struct ceph_osd_request_target {
	struct ceph_object_id base_oid;
	struct ceph_object_locator base_oloc;
	struct ceph_object_id target_oid;
	struct ceph_object_locator target_oloc;

	struct ceph_pg pgid;               /* last raw pg we mapped to */
	struct ceph_spg spgid;             /* last actual spg we mapped to */
	u32 pg_num;
	u32 pg_num_mask;
	struct ceph_osds acting;
	struct ceph_osds up;
	int size;
	int min_size;
	bool sort_bitwise;
	bool recovery_deletes;

	unsigned int flags;                /* CEPH_OSD_FLAG_* */
	bool paused;

	u32 epoch;
	u32 last_force_resend;

	int osd;
};

/* an in-flight request */
struct ceph_osd_request {
	u64             r_tid;              /* unique for this client */
	struct rb_node  r_node;
	struct rb_node  r_mc_node;          /* map check */
	struct work_struct r_complete_work;
	struct ceph_osd *r_osd;

	struct ceph_osd_request_target r_t;
#define r_base_oid	r_t.base_oid
#define r_base_oloc	r_t.base_oloc
#define r_flags		r_t.flags

	struct ceph_msg  *r_request, *r_reply;
	u32               r_sent;      /* >0 if r_request is sending/sent */

	/* request osd ops array  */
	unsigned int		r_num_ops;

	int               r_result;

	struct ceph_osd_client *r_osdc;
	struct kref       r_kref;
	bool              r_mempool;
	struct completion r_completion;       /* private to osd_client.c */
	ceph_osdc_callback_t r_callback;
	struct list_head  r_unsafe_item;

	struct inode *r_inode;         	      /* for use by callbacks */
	void *r_priv;			      /* ditto */

	/* set by submitter */
	u64 r_snapid;                         /* for reads, CEPH_NOSNAP o/w */
	struct ceph_snap_context *r_snapc;    /* for writes */
	struct timespec r_mtime;              /* ditto */
	u64 r_data_offset;                    /* ditto */
	bool r_linger;                        /* don't resend on failure */

	/* internal */
	unsigned long r_stamp;                /* jiffies, send or check time */
	unsigned long r_start_stamp;          /* jiffies */
	int r_attempts;
	u32 r_map_dne_bound;

	struct ceph_osd_req_op r_ops[];
};

struct ceph_request_redirect {
	struct ceph_object_locator oloc;
};

/*
 * osd request identifier
 *
 * caller name + incarnation# + tid to unique identify this request
 */
struct ceph_osd_reqid {
	struct ceph_entity_name name;
	__le64 tid;
	__le32 inc;
} __packed;

struct ceph_blkin_trace_info {
	__le64 trace_id;
	__le64 span_id;
	__le64 parent_span_id;
} __packed;

typedef void (*rados_watchcb2_t)(void *arg, u64 notify_id, u64 cookie,
				 u64 notifier_id, void *data, size_t data_len);
typedef void (*rados_watcherrcb_t)(void *arg, u64 cookie, int err);

struct ceph_osd_linger_request {
	struct ceph_osd_client *osdc;
	u64 linger_id;
	bool committed;
	bool is_watch;                  /* watch or notify */

	struct ceph_osd *osd;
	struct ceph_osd_request *reg_req;
	struct ceph_osd_request *ping_req;
	unsigned long ping_sent;
	unsigned long watch_valid_thru;
	struct list_head pending_lworks;

	struct ceph_osd_request_target t;
	u32 map_dne_bound;

	struct timespec mtime;

	struct kref kref;
	struct mutex lock;
	struct rb_node node;            /* osd */
	struct rb_node osdc_node;       /* osdc */
	struct rb_node mc_node;         /* map check */
	struct list_head scan_item;

	struct completion reg_commit_wait;
	struct completion notify_finish_wait;
	int reg_commit_error;
	int notify_finish_error;
	int last_error;

	u32 register_gen;
	u64 notify_id;

	rados_watchcb2_t wcb;
	rados_watcherrcb_t errcb;
	void *data;

	struct page ***preply_pages;
	size_t *preply_len;
};

struct ceph_watch_item {
	struct ceph_entity_name name;
	u64 cookie;
	struct ceph_entity_addr addr;
};

struct ceph_spg_mapping {
	struct rb_node node;
	struct ceph_spg spgid;

	struct rb_root backoffs;
};

struct ceph_hobject_id {
	void *key;
	size_t key_len;
	void *oid;
	size_t oid_len;
	u64 snapid;
	u32 hash;
	u8 is_max;
	void *nspace;
	size_t nspace_len;
	s64 pool;

	/* cache */
	u32 hash_reverse_bits;
};

static inline void ceph_hoid_build_hash_cache(struct ceph_hobject_id *hoid)
{
	hoid->hash_reverse_bits = bitrev32(hoid->hash);
}

/*
 * PG-wide backoff: [begin, end)
 * per-object backoff: begin == end
 */
struct ceph_osd_backoff {
	struct rb_node spg_node;
	struct rb_node id_node;

	struct ceph_spg spgid;
	u64 id;
	struct ceph_hobject_id *begin;
	struct ceph_hobject_id *end;
};

#define CEPH_LINGER_ID_START	0xffff000000000000ULL

struct ceph_osd_client {
	struct ceph_client     *client;

	struct ceph_osdmap     *osdmap;       /* current map */
	struct rw_semaphore    lock;

	struct rb_root         osds;          /* osds */
	struct list_head       osd_lru;       /* idle osds */
	spinlock_t             osd_lru_lock;
	u32		       epoch_barrier;
	struct ceph_osd        homeless_osd;
	atomic64_t             last_tid;      /* tid of last request */
	u64                    last_linger_id;
	struct rb_root         linger_requests; /* lingering requests */
	struct rb_root         map_checks;
	struct rb_root         linger_map_checks;
	atomic_t               num_requests;
	atomic_t               num_homeless;
	bool                   abort_on_full; /* abort w/ ENOSPC when full */
	int                    abort_err;
	struct delayed_work    timeout_work;
	struct delayed_work    osds_timeout_work;
#ifdef CONFIG_DEBUG_FS
	struct dentry 	       *debugfs_file;
#endif

	mempool_t              *req_mempool;

	struct ceph_msgpool	msgpool_op;
	struct ceph_msgpool	msgpool_op_reply;

	struct workqueue_struct	*notify_wq;
	struct workqueue_struct	*completion_wq;
};

static inline bool ceph_osdmap_flag(struct ceph_osd_client *osdc, int flag)
{
	return osdc->osdmap->flags & flag;
}

extern int ceph_osdc_setup(void);
extern void ceph_osdc_cleanup(void);

extern int ceph_osdc_init(struct ceph_osd_client *osdc,
			  struct ceph_client *client);
extern void ceph_osdc_stop(struct ceph_osd_client *osdc);

extern void ceph_osdc_handle_reply(struct ceph_osd_client *osdc,
				   struct ceph_msg *msg);
extern void ceph_osdc_handle_map(struct ceph_osd_client *osdc,
				 struct ceph_msg *msg);
void ceph_osdc_update_epoch_barrier(struct ceph_osd_client *osdc, u32 eb);
void ceph_osdc_abort_requests(struct ceph_osd_client *osdc, int err);

extern void osd_req_op_init(struct ceph_osd_request *osd_req,
			    unsigned int which, u16 opcode, u32 flags);

extern void osd_req_op_raw_data_in_pages(struct ceph_osd_request *,
					unsigned int which,
					struct page **pages, u64 length,
					u32 alignment, bool pages_from_pool,
					bool own_pages);

extern void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
					unsigned int which, u16 opcode,
					u64 offset, u64 length,
					u64 truncate_size, u32 truncate_seq);
extern void osd_req_op_extent_update(struct ceph_osd_request *osd_req,
					unsigned int which, u64 length);
extern void osd_req_op_extent_dup_last(struct ceph_osd_request *osd_req,
				       unsigned int which, u64 offset_inc);

extern struct ceph_osd_data *osd_req_op_extent_osd_data(
					struct ceph_osd_request *osd_req,
					unsigned int which);

extern void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *,
					unsigned int which,
					struct page **pages, u64 length,
					u32 alignment, bool pages_from_pool,
					bool own_pages);
extern void osd_req_op_extent_osd_data_pagelist(struct ceph_osd_request *,
					unsigned int which,
					struct ceph_pagelist *pagelist);
#ifdef CONFIG_BLOCK
void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req,
				    unsigned int which,
				    struct ceph_bio_iter *bio_pos,
				    u32 bio_length);
#endif /* CONFIG_BLOCK */
void osd_req_op_extent_osd_data_bvecs(struct ceph_osd_request *osd_req,
				      unsigned int which,
				      struct bio_vec *bvecs, u32 num_bvecs,
				      u32 bytes);
void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req,
					 unsigned int which,
					 struct ceph_bvec_iter *bvec_pos);

extern void osd_req_op_cls_request_data_pagelist(struct ceph_osd_request *,
					unsigned int which,
					struct ceph_pagelist *pagelist);
extern void osd_req_op_cls_request_data_pages(struct ceph_osd_request *,
					unsigned int which,
					struct page **pages, u64 length,
					u32 alignment, bool pages_from_pool,
					bool own_pages);
void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req,
				       unsigned int which,
				       struct bio_vec *bvecs, u32 num_bvecs,
				       u32 bytes);
extern void osd_req_op_cls_response_data_pages(struct ceph_osd_request *,
					unsigned int which,
					struct page **pages, u64 length,
					u32 alignment, bool pages_from_pool,
					bool own_pages);
extern int osd_req_op_cls_init(struct ceph_osd_request *osd_req,
					unsigned int which, u16 opcode,
					const char *class, const char *method);
extern int osd_req_op_xattr_init(struct ceph_osd_request *osd_req, unsigned int which,
				 u16 opcode, const char *name, const void *value,
				 size_t size, u8 cmp_op, u8 cmp_mode);
extern void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
				       unsigned int which,
				       u64 expected_object_size,
				       u64 expected_write_size);

extern struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
					       struct ceph_snap_context *snapc,
					       unsigned int num_ops,
					       bool use_mempool,
					       gfp_t gfp_flags);
int ceph_osdc_alloc_messages(struct ceph_osd_request *req, gfp_t gfp);

extern struct ceph_osd_request *ceph_osdc_new_request(struct ceph_osd_client *,
				      struct ceph_file_layout *layout,
				      struct ceph_vino vino,
				      u64 offset, u64 *len,
				      unsigned int which, int num_ops,
				      int opcode, int flags,
				      struct ceph_snap_context *snapc,
				      u32 truncate_seq, u64 truncate_size,
				      bool use_mempool);

extern void ceph_osdc_get_request(struct ceph_osd_request *req);
extern void ceph_osdc_put_request(struct ceph_osd_request *req);

extern int ceph_osdc_start_request(struct ceph_osd_client *osdc,
				   struct ceph_osd_request *req,
				   bool nofail);
extern void ceph_osdc_cancel_request(struct ceph_osd_request *req);
extern int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
				  struct ceph_osd_request *req);
extern void ceph_osdc_sync(struct ceph_osd_client *osdc);

extern void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc);
void ceph_osdc_maybe_request_map(struct ceph_osd_client *osdc);

int ceph_osdc_call(struct ceph_osd_client *osdc,
		   struct ceph_object_id *oid,
		   struct ceph_object_locator *oloc,
		   const char *class, const char *method,
		   unsigned int flags,
		   struct page *req_page, size_t req_len,
		   struct page *resp_page, size_t *resp_len);

extern int ceph_osdc_readpages(struct ceph_osd_client *osdc,
			       struct ceph_vino vino,
			       struct ceph_file_layout *layout,
			       u64 off, u64 *plen,
			       u32 truncate_seq, u64 truncate_size,
			       struct page **pages, int nr_pages,
			       int page_align);

extern int ceph_osdc_writepages(struct ceph_osd_client *osdc,
				struct ceph_vino vino,
				struct ceph_file_layout *layout,
				struct ceph_snap_context *sc,
				u64 off, u64 len,
				u32 truncate_seq, u64 truncate_size,
				struct timespec *mtime,
				struct page **pages, int nr_pages);

/* watch/notify */
struct ceph_osd_linger_request *
ceph_osdc_watch(struct ceph_osd_client *osdc,
		struct ceph_object_id *oid,
		struct ceph_object_locator *oloc,
		rados_watchcb2_t wcb,
		rados_watcherrcb_t errcb,
		void *data);
int ceph_osdc_unwatch(struct ceph_osd_client *osdc,
		      struct ceph_osd_linger_request *lreq);

int ceph_osdc_notify_ack(struct ceph_osd_client *osdc,
			 struct ceph_object_id *oid,
			 struct ceph_object_locator *oloc,
			 u64 notify_id,
			 u64 cookie,
			 void *payload,
			 size_t payload_len);
int ceph_osdc_notify(struct ceph_osd_client *osdc,
		     struct ceph_object_id *oid,
		     struct ceph_object_locator *oloc,
		     void *payload,
		     size_t payload_len,
		     u32 timeout,
		     struct page ***preply_pages,
		     size_t *preply_len);
int ceph_osdc_watch_check(struct ceph_osd_client *osdc,
			  struct ceph_osd_linger_request *lreq);
int ceph_osdc_list_watchers(struct ceph_osd_client *osdc,
			    struct ceph_object_id *oid,
			    struct ceph_object_locator *oloc,
			    struct ceph_watch_item **watchers,
			    u32 *num_watchers);
#endif
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
f24e9980 SW	2	#ifndef _FS_CEPH_OSD_CLIENT_H
	3	#define _FS_CEPH_OSD_CLIENT_H
	4
a02a946d	5	#include <linux/bitrev.h>
f24e9980	6	#include <linux/completion.h>
415e49a9	7	#include <linux/kref.h>
f24e9980 SW	8	#include <linux/mempool.h>
f24e9980 SW	9	#include <linux/rbtree.h>
02113a0f	10	#include <linux/refcount.h>
f24e9980	11
6c4a1915 AE	12	#include <linux/ceph/types.h>
	13	#include <linux/ceph/osdmap.h>
	14	#include <linux/ceph/messenger.h>
b2aa5d0b	15	#include <linux/ceph/msgpool.h>
6c4a1915	16	#include <linux/ceph/auth.h>
c885837f	17	#include <linux/ceph/pagelist.h>
f24e9980 SW	18
	19	struct ceph_msg;
	20	struct ceph_snap_context;
	21	struct ceph_osd_request;
	22	struct ceph_osd_client;
	23
	24	/*
	25	* completion callback for async writepages
	26	*/
85e084fe	27	typedef void (ceph_osdc_callback_t)(struct ceph_osd_request );
f24e9980	28
63244fa1 ID	29	#define CEPH_HOMELESS_OSD -1
63244fa1 ID	30
f24e9980 SW	31	/* a given osd we're communicating with */
f24e9980 SW	32	struct ceph_osd {
02113a0f	33	refcount_t o_ref;
f24e9980 SW	34	struct ceph_osd_client *o_osdc;
	35	int o_osd;
	36	int o_incarnation;
	37	struct rb_node o_node;
	38	struct ceph_connection o_con;
5aea3dcd	39	struct rb_root o_requests;
922dab61	40	struct rb_root o_linger_requests;
a02a946d ID	41	struct rb_root o_backoff_mappings;
a02a946d ID	42	struct rb_root o_backoffs_by_id;
f5a2041b	43	struct list_head o_osd_lru;
6c4a1915	44	struct ceph_auth_handshake o_auth;
f5a2041b	45	unsigned long lru_ttl;
422d2cb8	46	struct list_head o_keepalive_item;
5aea3dcd	47	struct mutex lock;
f24e9980 SW	48	};
f24e9980 SW	49
3f1af42a ID	50	#define CEPH_OSD_SLAB_OPS 2
3f1af42a ID	51	#define CEPH_OSD_MAX_OPS 16
1b83bef2	52
2ac2b7a6	53	enum ceph_osd_data_type {
ec9123c5	54	CEPH_OSD_DATA_TYPE_NONE = 0,
2ac2b7a6	55	CEPH_OSD_DATA_TYPE_PAGES,
9a5e6d09	56	CEPH_OSD_DATA_TYPE_PAGELIST,
2ac2b7a6 AE	57	#ifdef CONFIG_BLOCK
	58	CEPH_OSD_DATA_TYPE_BIO,
	59	#endif /* CONFIG_BLOCK */
b9e281c2	60	CEPH_OSD_DATA_TYPE_BVECS,
2ac2b7a6 AE	61	};
2ac2b7a6 AE	62
2794a82a	63	struct ceph_osd_data {
2ac2b7a6 AE	64	enum ceph_osd_data_type type;
2ac2b7a6 AE	65	union {
2794a82a AE	66	struct {
2794a82a AE	67	struct page **pages;
e0c59487	68	u64 length;
2794a82a AE	69	u32 alignment;
	70	bool pages_from_pool;
	71	bool own_pages;
	72	};
9a5e6d09	73	struct ceph_pagelist *pagelist;
2794a82a	74	#ifdef CONFIG_BLOCK
fdce58cc	75	struct {
5359a17d ID	76	struct ceph_bio_iter bio_pos;
5359a17d ID	77	u32 bio_length;
fdce58cc	78	};
2794a82a	79	#endif /* CONFIG_BLOCK */
0010f705 ID	80	struct {
	81	struct ceph_bvec_iter bvec_pos;
	82	u32 num_bvecs;
	83	};
2794a82a AE	84	};
	85	};
	86
79528734 AE	87	struct ceph_osd_req_op {
79528734 AE	88	u16 op; /* CEPH_OSD_OP_* */
7b25bf5f	89	u32 flags; /* CEPH_OSD_OP_FLAG_* */
de2aa102	90	u32 indata_len; /* request */
7665d85b YZ	91	u32 outdata_len; /* reply */
	92	s32 rval;
	93
79528734	94	union {
49719778	95	struct ceph_osd_data raw_data_in;
79528734 AE	96	struct {
	97	u64 offset, length;
	98	u64 truncate_size;
	99	u32 truncate_seq;
5476492f	100	struct ceph_osd_data osd_data;
79528734	101	} extent;
d74b50be	102	struct {
d7d5a007 ID	103	u32 name_len;
d7d5a007 ID	104	u32 value_len;
d74b50be YZ	105	__u8 cmp_op; /* CEPH_OSD_CMPXATTR_OP_* */
	106	__u8 cmp_mode; /* CEPH_OSD_CMPXATTR_MODE_* */
	107	struct ceph_osd_data osd_data;
	108	} xattr;
79528734 AE	109	struct {
	110	const char *class_name;
	111	const char *method_name;
5476492f	112	struct ceph_osd_data request_info;
04017e29	113	struct ceph_osd_data request_data;
5476492f	114	struct ceph_osd_data response_data;
79528734 AE	115	__u8 class_len;
79528734 AE	116	__u8 method_len;
bb873b53	117	u32 indata_len;
79528734 AE	118	} cls;
	119	struct {
	120	u64 cookie;
922dab61 ID	121	__u8 op; /* CEPH_OSD_WATCH_OP_ */
922dab61 ID	122	u32 gen;
79528734	123	} watch;
922dab61 ID	124	struct {
	125	struct ceph_osd_data request_data;
	126	} notify_ack;
19079203 ID	127	struct {
	128	u64 cookie;
	129	struct ceph_osd_data request_data;
	130	struct ceph_osd_data response_data;
	131	} notify;
a4ed38d7 DF	132	struct {
	133	struct ceph_osd_data response_data;
	134	} list_watchers;
c647b8a8 ID	135	struct {
	136	u64 expected_object_size;
	137	u64 expected_write_size;
	138	} alloc_hint;
79528734 AE	139	};
	140	};
	141
63244fa1 ID	142	struct ceph_osd_request_target {
	143	struct ceph_object_id base_oid;
	144	struct ceph_object_locator base_oloc;
	145	struct ceph_object_id target_oid;
	146	struct ceph_object_locator target_oloc;
	147
dc98ff72 ID	148	struct ceph_pg pgid; /* last raw pg we mapped to */
dc98ff72 ID	149	struct ceph_spg spgid; /* last actual spg we mapped to */
63244fa1 ID	150	u32 pg_num;
	151	u32 pg_num_mask;
	152	struct ceph_osds acting;
	153	struct ceph_osds up;
	154	int size;
	155	int min_size;
	156	bool sort_bitwise;
ae78dd81	157	bool recovery_deletes;
63244fa1 ID	158
	159	unsigned int flags; /* CEPH_OSD_FLAG_* */
	160	bool paused;
	161
04c7d789	162	u32 epoch;
dc93e0e2 ID	163	u32 last_force_resend;
dc93e0e2 ID	164
63244fa1 ID	165	int osd;
	166	};
	167
f24e9980 SW	168	/* an in-flight request */
	169	struct ceph_osd_request {
	170	u64 r_tid; /* unique for this client */
	171	struct rb_node r_node;
4609245e	172	struct rb_node r_mc_node; /* map check */
88bc1922	173	struct work_struct r_complete_work;
f24e9980	174	struct ceph_osd *r_osd;
a66dd383 ID	175
	176	struct ceph_osd_request_target r_t;
	177	#define r_base_oid r_t.base_oid
	178	#define r_base_oloc r_t.base_oloc
	179	#define r_flags r_t.flags
f24e9980 SW	180
f24e9980 SW	181	struct ceph_msg r_request, r_reply;
f24e9980	182	u32 r_sent; /* >0 if r_request is sending/sent */
1b83bef2	183
79528734 AE	184	/* request osd ops array */
79528734 AE	185	unsigned int r_num_ops;
79528734	186
1b83bef2	187	int r_result;
f24e9980 SW	188
f24e9980 SW	189	struct ceph_osd_client *r_osdc;
415e49a9	190	struct kref r_kref;
f24e9980	191	bool r_mempool;
b18b9550	192	struct completion r_completion; /* private to osd_client.c */
26be8808	193	ceph_osdc_callback_t r_callback;
f24e9980 SW	194	struct list_head r_unsafe_item;
	195
	196	struct inode r_inode; / for use by callbacks */
3d14c5d2	197	void r_priv; / ditto */
f24e9980	198
bb873b53 ID	199	/* set by submitter */
	200	u64 r_snapid; /* for reads, CEPH_NOSNAP o/w */
	201	struct ceph_snap_context r_snapc; / for writes */
	202	struct timespec r_mtime; /* ditto */
	203	u64 r_data_offset; /* ditto */
922dab61	204	bool r_linger; /* don't resend on failure */
f24e9980	205
bb873b53 ID	206	/* internal */
bb873b53 ID	207	unsigned long r_stamp; /* jiffies, send or check time */
7cc5e38f	208	unsigned long r_start_stamp; /* jiffies */
bb873b53	209	int r_attempts;
4609245e	210	u32 r_map_dne_bound;
3f1af42a ID	211
3f1af42a ID	212	struct ceph_osd_req_op r_ops[];
f24e9980 SW	213	};
f24e9980 SW	214
205ee118 ID	215	struct ceph_request_redirect {
	216	struct ceph_object_locator oloc;
	217	};
	218
8cb441c0 ID	219	/*
	220	* osd request identifier
	221	*
	222	* caller name + incarnation# + tid to unique identify this request
	223	*/
	224	struct ceph_osd_reqid {
	225	struct ceph_entity_name name;
	226	__le64 tid;
	227	__le32 inc;
	228	} __packed;
	229
	230	struct ceph_blkin_trace_info {
	231	__le64 trace_id;
	232	__le64 span_id;
	233	__le64 parent_span_id;
	234	} __packed;
	235
922dab61 ID	236	typedef void (rados_watchcb2_t)(void arg, u64 notify_id, u64 cookie,
	237	u64 notifier_id, void *data, size_t data_len);
	238	typedef void (rados_watcherrcb_t)(void arg, u64 cookie, int err);
	239
	240	struct ceph_osd_linger_request {
a40c4f10	241	struct ceph_osd_client *osdc;
922dab61 ID	242	u64 linger_id;
922dab61 ID	243	bool committed;
19079203	244	bool is_watch; /* watch or notify */
922dab61 ID	245
	246	struct ceph_osd *osd;
	247	struct ceph_osd_request *reg_req;
	248	struct ceph_osd_request *ping_req;
	249	unsigned long ping_sent;
b07d3c4b ID	250	unsigned long watch_valid_thru;
b07d3c4b ID	251	struct list_head pending_lworks;
922dab61 ID	252
922dab61 ID	253	struct ceph_osd_request_target t;
4609245e	254	u32 map_dne_bound;
922dab61 ID	255
	256	struct timespec mtime;
	257
a40c4f10	258	struct kref kref;
922dab61 ID	259	struct mutex lock;
	260	struct rb_node node; /* osd */
	261	struct rb_node osdc_node; /* osdc */
4609245e	262	struct rb_node mc_node; /* map check */
922dab61 ID	263	struct list_head scan_item;
	264
	265	struct completion reg_commit_wait;
19079203	266	struct completion notify_finish_wait;
922dab61	267	int reg_commit_error;
19079203	268	int notify_finish_error;
922dab61 ID	269	int last_error;
	270
	271	u32 register_gen;
19079203	272	u64 notify_id;
a40c4f10	273
922dab61 ID	274	rados_watchcb2_t wcb;
	275	rados_watcherrcb_t errcb;
	276	void *data;
19079203 ID	277
	278	struct page ***preply_pages;
	279	size_t *preply_len;
a40c4f10 YS	280	};
a40c4f10 YS	281
a4ed38d7 DF	282	struct ceph_watch_item {
	283	struct ceph_entity_name name;
	284	u64 cookie;
	285	struct ceph_entity_addr addr;
	286	};
	287
a02a946d ID	288	struct ceph_spg_mapping {
	289	struct rb_node node;
	290	struct ceph_spg spgid;
	291
	292	struct rb_root backoffs;
	293	};
	294
	295	struct ceph_hobject_id {
	296	void *key;
	297	size_t key_len;
	298	void *oid;
	299	size_t oid_len;
	300	u64 snapid;
	301	u32 hash;
	302	u8 is_max;
	303	void *nspace;
	304	size_t nspace_len;
	305	s64 pool;
	306
	307	/* cache */
	308	u32 hash_reverse_bits;
	309	};
	310
	311	static inline void ceph_hoid_build_hash_cache(struct ceph_hobject_id *hoid)
	312	{
	313	hoid->hash_reverse_bits = bitrev32(hoid->hash);
	314	}
	315
	316	/*
	317	* PG-wide backoff: [begin, end)
	318	* per-object backoff: begin == end
	319	*/
	320	struct ceph_osd_backoff {
	321	struct rb_node spg_node;
	322	struct rb_node id_node;
	323
	324	struct ceph_spg spgid;
	325	u64 id;
	326	struct ceph_hobject_id *begin;
	327	struct ceph_hobject_id *end;
	328	};
	329
264048af ID	330	#define CEPH_LINGER_ID_START 0xffff000000000000ULL
264048af ID	331
f24e9980 SW	332	struct ceph_osd_client {
	333	struct ceph_client *client;
	334
	335	struct ceph_osdmap osdmap; / current map */
5aea3dcd	336	struct rw_semaphore lock;
f24e9980	337
f24e9980	338	struct rb_root osds; /* osds */
f5a2041b	339	struct list_head osd_lru; /* idle osds */
9dd2845c	340	spinlock_t osd_lru_lock;
58eb7932	341	u32 epoch_barrier;
5aea3dcd ID	342	struct ceph_osd homeless_osd;
5aea3dcd ID	343	atomic64_t last_tid; /* tid of last request */
922dab61 ID	344	u64 last_linger_id;
922dab61 ID	345	struct rb_root linger_requests; /* lingering requests */
4609245e ID	346	struct rb_root map_checks;
4609245e ID	347	struct rb_root linger_map_checks;
5aea3dcd ID	348	atomic_t num_requests;
5aea3dcd ID	349	atomic_t num_homeless;
c843d13c	350	bool abort_on_full; /* abort w/ ENOSPC when full */
66850df5	351	int abort_err;
f24e9980	352	struct delayed_work timeout_work;
f5a2041b	353	struct delayed_work osds_timeout_work;
039934b8	354	#ifdef CONFIG_DEBUG_FS
f24e9980	355	struct dentry *debugfs_file;
039934b8	356	#endif
f24e9980 SW	357
	358	mempool_t *req_mempool;
	359
0d59ab81	360	struct ceph_msgpool msgpool_op;
c16e7869	361	struct ceph_msgpool msgpool_op_reply;
a40c4f10	362
a40c4f10	363	struct workqueue_struct *notify_wq;
88bc1922	364	struct workqueue_struct *completion_wq;
f24e9980 SW	365	};
f24e9980 SW	366
b7ec35b3 ID	367	static inline bool ceph_osdmap_flag(struct ceph_osd_client *osdc, int flag)
	368	{
	369	return osdc->osdmap->flags & flag;
	370	}
	371
5522ae0b AE	372	extern int ceph_osdc_setup(void);
	373	extern void ceph_osdc_cleanup(void);
	374
f24e9980 SW	375	extern int ceph_osdc_init(struct ceph_osd_client *osdc,
	376	struct ceph_client *client);
	377	extern void ceph_osdc_stop(struct ceph_osd_client *osdc);
	378
	379	extern void ceph_osdc_handle_reply(struct ceph_osd_client *osdc,
	380	struct ceph_msg *msg);
	381	extern void ceph_osdc_handle_map(struct ceph_osd_client *osdc,
	382	struct ceph_msg *msg);
58eb7932	383	void ceph_osdc_update_epoch_barrier(struct ceph_osd_client *osdc, u32 eb);
66850df5	384	void ceph_osdc_abort_requests(struct ceph_osd_client *osdc, int err);
f24e9980	385
49719778	386	extern void osd_req_op_init(struct ceph_osd_request *osd_req,
144cba14	387	unsigned int which, u16 opcode, u32 flags);
49719778 AE	388
	389	extern void osd_req_op_raw_data_in_pages(struct ceph_osd_request *,
	390	unsigned int which,
	391	struct page **pages, u64 length,
	392	u32 alignment, bool pages_from_pool,
	393	bool own_pages);
	394
c99d2d4a AE	395	extern void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
c99d2d4a AE	396	unsigned int which, u16 opcode,
33803f33 AE	397	u64 offset, u64 length,
33803f33 AE	398	u64 truncate_size, u32 truncate_seq);
c99d2d4a AE	399	extern void osd_req_op_extent_update(struct ceph_osd_request *osd_req,
c99d2d4a AE	400	unsigned int which, u64 length);
2c63f49a YZ	401	extern void osd_req_op_extent_dup_last(struct ceph_osd_request *osd_req,
2c63f49a YZ	402	unsigned int which, u64 offset_inc);
a4ce40a9 AE	403
	404	extern struct ceph_osd_data *osd_req_op_extent_osd_data(
	405	struct ceph_osd_request *osd_req,
406e2c9f	406	unsigned int which);
a4ce40a9 AE	407
a4ce40a9 AE	408	extern void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *,
406e2c9f	409	unsigned int which,
a4ce40a9 AE	410	struct page **pages, u64 length,
	411	u32 alignment, bool pages_from_pool,
	412	bool own_pages);
	413	extern void osd_req_op_extent_osd_data_pagelist(struct ceph_osd_request *,
406e2c9f	414	unsigned int which,
a4ce40a9 AE	415	struct ceph_pagelist *pagelist);
a4ce40a9 AE	416	#ifdef CONFIG_BLOCK
5359a17d ID	417	void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req,
	418	unsigned int which,
	419	struct ceph_bio_iter *bio_pos,
	420	u32 bio_length);
a4ce40a9	421	#endif /* CONFIG_BLOCK */
0010f705 ID	422	void osd_req_op_extent_osd_data_bvecs(struct ceph_osd_request *osd_req,
	423	unsigned int which,
	424	struct bio_vec *bvecs, u32 num_bvecs,
	425	u32 bytes);
b9e281c2 ID	426	void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req,
	427	unsigned int which,
	428	struct ceph_bvec_iter *bvec_pos);
a4ce40a9	429
04017e29 AE	430	extern void osd_req_op_cls_request_data_pagelist(struct ceph_osd_request *,
	431	unsigned int which,
	432	struct ceph_pagelist *pagelist);
6c57b554 AE	433	extern void osd_req_op_cls_request_data_pages(struct ceph_osd_request *,
	434	unsigned int which,
	435	struct page **pages, u64 length,
	436	u32 alignment, bool pages_from_pool,
	437	bool own_pages);
b9e281c2 ID	438	void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req,
b9e281c2 ID	439	unsigned int which,
0010f705 ID	440	struct bio_vec *bvecs, u32 num_bvecs,
0010f705 ID	441	u32 bytes);
a4ce40a9	442	extern void osd_req_op_cls_response_data_pages(struct ceph_osd_request *,
c99d2d4a	443	unsigned int which,
a4ce40a9 AE	444	struct page **pages, u64 length,
	445	u32 alignment, bool pages_from_pool,
	446	bool own_pages);
fe943d50	447	extern int osd_req_op_cls_init(struct ceph_osd_request *osd_req,
c99d2d4a	448	unsigned int which, u16 opcode,
04017e29	449	const char class, const char method);
d74b50be YZ	450	extern int osd_req_op_xattr_init(struct ceph_osd_request *osd_req, unsigned int which,
	451	u16 opcode, const char name, const void value,
	452	size_t size, u8 cmp_op, u8 cmp_mode);
c647b8a8 ID	453	extern void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
	454	unsigned int which,
	455	u64 expected_object_size,
	456	u64 expected_write_size);
33803f33	457
3499e8a5	458	extern struct ceph_osd_request ceph_osdc_alloc_request(struct ceph_osd_client osdc,
3499e8a5	459	struct ceph_snap_context *snapc,
acead002	460	unsigned int num_ops,
3499e8a5	461	bool use_mempool,
54a54007	462	gfp_t gfp_flags);
13d1ad16	463	int ceph_osdc_alloc_messages(struct ceph_osd_request *req, gfp_t gfp);
3499e8a5	464
f24e9980 SW	465	extern struct ceph_osd_request ceph_osdc_new_request(struct ceph_osd_client ,
	466	struct ceph_file_layout *layout,
	467	struct ceph_vino vino,
acead002	468	u64 offset, u64 *len,
715e4cd4 YZ	469	unsigned int which, int num_ops,
715e4cd4 YZ	470	int opcode, int flags,
f24e9980	471	struct ceph_snap_context *snapc,
acead002	472	u32 truncate_seq, u64 truncate_size,
153e5167	473	bool use_mempool);
f24e9980	474
9e94af20 ID	475	extern void ceph_osdc_get_request(struct ceph_osd_request *req);
9e94af20 ID	476	extern void ceph_osdc_put_request(struct ceph_osd_request *req);
f24e9980 SW	477
	478	extern int ceph_osdc_start_request(struct ceph_osd_client *osdc,
	479	struct ceph_osd_request *req,
	480	bool nofail);
c9f9b93d	481	extern void ceph_osdc_cancel_request(struct ceph_osd_request *req);
f24e9980 SW	482	extern int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
	483	struct ceph_osd_request *req);
	484	extern void ceph_osdc_sync(struct ceph_osd_client *osdc);
	485
dd935f44	486	extern void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc);
7cca78c9	487	void ceph_osdc_maybe_request_map(struct ceph_osd_client *osdc);
dd935f44	488
428a7158 DF	489	int ceph_osdc_call(struct ceph_osd_client *osdc,
	490	struct ceph_object_id *oid,
	491	struct ceph_object_locator *oloc,
	492	const char class, const char method,
	493	unsigned int flags,
	494	struct page *req_page, size_t req_len,
	495	struct page resp_page, size_t resp_len);
	496
f24e9980 SW	497	extern int ceph_osdc_readpages(struct ceph_osd_client *osdc,
	498	struct ceph_vino vino,
	499	struct ceph_file_layout *layout,
	500	u64 off, u64 *plen,
	501	u32 truncate_seq, u64 truncate_size,
b7495fc2 SW	502	struct page **pages, int nr_pages,
b7495fc2 SW	503	int page_align);
f24e9980 SW	504
	505	extern int ceph_osdc_writepages(struct ceph_osd_client *osdc,
	506	struct ceph_vino vino,
	507	struct ceph_file_layout *layout,
	508	struct ceph_snap_context *sc,
	509	u64 off, u64 len,
	510	u32 truncate_seq, u64 truncate_size,
	511	struct timespec *mtime,
24808826	512	struct page **pages, int nr_pages);
f24e9980	513
922dab61 ID	514	/* watch/notify */
	515	struct ceph_osd_linger_request *
	516	ceph_osdc_watch(struct ceph_osd_client *osdc,
	517	struct ceph_object_id *oid,
	518	struct ceph_object_locator *oloc,
	519	rados_watchcb2_t wcb,
	520	rados_watcherrcb_t errcb,
	521	void *data);
	522	int ceph_osdc_unwatch(struct ceph_osd_client *osdc,
	523	struct ceph_osd_linger_request *lreq);
	524
	525	int ceph_osdc_notify_ack(struct ceph_osd_client *osdc,
	526	struct ceph_object_id *oid,
	527	struct ceph_object_locator *oloc,
	528	u64 notify_id,
	529	u64 cookie,
	530	void *payload,
	531	size_t payload_len);
19079203 ID	532	int ceph_osdc_notify(struct ceph_osd_client *osdc,
	533	struct ceph_object_id *oid,
	534	struct ceph_object_locator *oloc,
	535	void *payload,
	536	size_t payload_len,
	537	u32 timeout,
	538	struct page ***preply_pages,
	539	size_t *preply_len);
b07d3c4b ID	540	int ceph_osdc_watch_check(struct ceph_osd_client *osdc,
b07d3c4b ID	541	struct ceph_osd_linger_request *lreq);
a4ed38d7 DF	542	int ceph_osdc_list_watchers(struct ceph_osd_client *osdc,
	543	struct ceph_object_id *oid,
	544	struct ceph_object_locator *oloc,
	545	struct ceph_watch_item **watchers,
	546	u32 *num_watchers);
f24e9980 SW	547	#endif
f24e9980 SW	548