[mirror_ubuntu-artful-kernel.git] / include / linux / ceph / osd_client.h

#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 */
};

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 bio	*bio;		/* list of bios */
			size_t		bio_length;	/* total in list */
		};
#endif /* CONFIG_BLOCK */
	};
};

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 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 */
	bool r_abort_on_full;		      /* return ENOSPC when full */

	/* 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;
	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;
};

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);

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
extern void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *,
					unsigned int which,
					struct bio *bio, size_t bio_length);
#endif /* CONFIG_BLOCK */

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);
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 void 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
f24e9980 SW	1	#ifndef _FS_CEPH_OSD_CLIENT_H
	2	#define _FS_CEPH_OSD_CLIENT_H
	3
a02a946d	4	#include <linux/bitrev.h>
f24e9980	5	#include <linux/completion.h>
415e49a9	6	#include <linux/kref.h>
f24e9980 SW	7	#include <linux/mempool.h>
f24e9980 SW	8	#include <linux/rbtree.h>
02113a0f	9	#include <linux/refcount.h>
f24e9980	10
6c4a1915 AE	11	#include <linux/ceph/types.h>
	12	#include <linux/ceph/osdmap.h>
	13	#include <linux/ceph/messenger.h>
b2aa5d0b	14	#include <linux/ceph/msgpool.h>
6c4a1915	15	#include <linux/ceph/auth.h>
c885837f	16	#include <linux/ceph/pagelist.h>
f24e9980 SW	17
	18	struct ceph_msg;
	19	struct ceph_snap_context;
	20	struct ceph_osd_request;
	21	struct ceph_osd_client;
	22
	23	/*
	24	* completion callback for async writepages
	25	*/
85e084fe	26	typedef void (ceph_osdc_callback_t)(struct ceph_osd_request );
f24e9980	27
63244fa1 ID	28	#define CEPH_HOMELESS_OSD -1
63244fa1 ID	29
f24e9980 SW	30	/* a given osd we're communicating with */
f24e9980 SW	31	struct ceph_osd {
02113a0f	32	refcount_t o_ref;
f24e9980 SW	33	struct ceph_osd_client *o_osdc;
	34	int o_osd;
	35	int o_incarnation;
	36	struct rb_node o_node;
	37	struct ceph_connection o_con;
5aea3dcd	38	struct rb_root o_requests;
922dab61	39	struct rb_root o_linger_requests;
a02a946d ID	40	struct rb_root o_backoff_mappings;
a02a946d ID	41	struct rb_root o_backoffs_by_id;
f5a2041b	42	struct list_head o_osd_lru;
6c4a1915	43	struct ceph_auth_handshake o_auth;
f5a2041b	44	unsigned long lru_ttl;
422d2cb8	45	struct list_head o_keepalive_item;
5aea3dcd	46	struct mutex lock;
f24e9980 SW	47	};
f24e9980 SW	48
3f1af42a ID	49	#define CEPH_OSD_SLAB_OPS 2
3f1af42a ID	50	#define CEPH_OSD_MAX_OPS 16
1b83bef2	51
2ac2b7a6	52	enum ceph_osd_data_type {
ec9123c5	53	CEPH_OSD_DATA_TYPE_NONE = 0,
2ac2b7a6	54	CEPH_OSD_DATA_TYPE_PAGES,
9a5e6d09	55	CEPH_OSD_DATA_TYPE_PAGELIST,
2ac2b7a6 AE	56	#ifdef CONFIG_BLOCK
	57	CEPH_OSD_DATA_TYPE_BIO,
	58	#endif /* CONFIG_BLOCK */
	59	};
	60
2794a82a	61	struct ceph_osd_data {
2ac2b7a6 AE	62	enum ceph_osd_data_type type;
2ac2b7a6 AE	63	union {
2794a82a AE	64	struct {
2794a82a AE	65	struct page **pages;
e0c59487	66	u64 length;
2794a82a AE	67	u32 alignment;
	68	bool pages_from_pool;
	69	bool own_pages;
	70	};
9a5e6d09	71	struct ceph_pagelist *pagelist;
2794a82a	72	#ifdef CONFIG_BLOCK
fdce58cc AE	73	struct {
	74	struct bio bio; / list of bios */
	75	size_t bio_length; /* total in list */
	76	};
2794a82a AE	77	#endif /* CONFIG_BLOCK */
	78	};
	79	};
	80
79528734 AE	81	struct ceph_osd_req_op {
79528734 AE	82	u16 op; /* CEPH_OSD_OP_* */
7b25bf5f	83	u32 flags; /* CEPH_OSD_OP_FLAG_* */
de2aa102	84	u32 indata_len; /* request */
7665d85b YZ	85	u32 outdata_len; /* reply */
	86	s32 rval;
	87
79528734	88	union {
49719778	89	struct ceph_osd_data raw_data_in;
79528734 AE	90	struct {
	91	u64 offset, length;
	92	u64 truncate_size;
	93	u32 truncate_seq;
5476492f	94	struct ceph_osd_data osd_data;
79528734	95	} extent;
d74b50be	96	struct {
d7d5a007 ID	97	u32 name_len;
d7d5a007 ID	98	u32 value_len;
d74b50be YZ	99	__u8 cmp_op; /* CEPH_OSD_CMPXATTR_OP_* */
	100	__u8 cmp_mode; /* CEPH_OSD_CMPXATTR_MODE_* */
	101	struct ceph_osd_data osd_data;
	102	} xattr;
79528734 AE	103	struct {
	104	const char *class_name;
	105	const char *method_name;
5476492f	106	struct ceph_osd_data request_info;
04017e29	107	struct ceph_osd_data request_data;
5476492f	108	struct ceph_osd_data response_data;
79528734 AE	109	__u8 class_len;
79528734 AE	110	__u8 method_len;
bb873b53	111	u32 indata_len;
79528734 AE	112	} cls;
	113	struct {
	114	u64 cookie;
922dab61 ID	115	__u8 op; /* CEPH_OSD_WATCH_OP_ */
922dab61 ID	116	u32 gen;
79528734	117	} watch;
922dab61 ID	118	struct {
	119	struct ceph_osd_data request_data;
	120	} notify_ack;
19079203 ID	121	struct {
	122	u64 cookie;
	123	struct ceph_osd_data request_data;
	124	struct ceph_osd_data response_data;
	125	} notify;
a4ed38d7 DF	126	struct {
	127	struct ceph_osd_data response_data;
	128	} list_watchers;
c647b8a8 ID	129	struct {
	130	u64 expected_object_size;
	131	u64 expected_write_size;
	132	} alloc_hint;
79528734 AE	133	};
	134	};
	135
63244fa1 ID	136	struct ceph_osd_request_target {
	137	struct ceph_object_id base_oid;
	138	struct ceph_object_locator base_oloc;
	139	struct ceph_object_id target_oid;
	140	struct ceph_object_locator target_oloc;
	141
dc98ff72 ID	142	struct ceph_pg pgid; /* last raw pg we mapped to */
dc98ff72 ID	143	struct ceph_spg spgid; /* last actual spg we mapped to */
63244fa1 ID	144	u32 pg_num;
	145	u32 pg_num_mask;
	146	struct ceph_osds acting;
	147	struct ceph_osds up;
	148	int size;
	149	int min_size;
	150	bool sort_bitwise;
ae78dd81	151	bool recovery_deletes;
63244fa1 ID	152
	153	unsigned int flags; /* CEPH_OSD_FLAG_* */
	154	bool paused;
	155
04c7d789	156	u32 epoch;
dc93e0e2 ID	157	u32 last_force_resend;
dc93e0e2 ID	158
63244fa1 ID	159	int osd;
	160	};
	161
f24e9980 SW	162	/* an in-flight request */
	163	struct ceph_osd_request {
	164	u64 r_tid; /* unique for this client */
	165	struct rb_node r_node;
4609245e	166	struct rb_node r_mc_node; /* map check */
f24e9980	167	struct ceph_osd *r_osd;
a66dd383 ID	168
	169	struct ceph_osd_request_target r_t;
	170	#define r_base_oid r_t.base_oid
	171	#define r_base_oloc r_t.base_oloc
	172	#define r_flags r_t.flags
f24e9980 SW	173
f24e9980 SW	174	struct ceph_msg r_request, r_reply;
f24e9980	175	u32 r_sent; /* >0 if r_request is sending/sent */
1b83bef2	176
79528734 AE	177	/* request osd ops array */
79528734 AE	178	unsigned int r_num_ops;
79528734	179
1b83bef2	180	int r_result;
f24e9980 SW	181
f24e9980 SW	182	struct ceph_osd_client *r_osdc;
415e49a9	183	struct kref r_kref;
f24e9980	184	bool r_mempool;
b18b9550	185	struct completion r_completion; /* private to osd_client.c */
26be8808	186	ceph_osdc_callback_t r_callback;
f24e9980 SW	187	struct list_head r_unsafe_item;
	188
	189	struct inode r_inode; / for use by callbacks */
3d14c5d2	190	void r_priv; / ditto */
f24e9980	191
bb873b53 ID	192	/* set by submitter */
	193	u64 r_snapid; /* for reads, CEPH_NOSNAP o/w */
	194	struct ceph_snap_context r_snapc; / for writes */
	195	struct timespec r_mtime; /* ditto */
	196	u64 r_data_offset; /* ditto */
922dab61	197	bool r_linger; /* don't resend on failure */
a1f4020a	198	bool r_abort_on_full; /* return ENOSPC when full */
f24e9980	199
bb873b53 ID	200	/* internal */
bb873b53 ID	201	unsigned long r_stamp; /* jiffies, send or check time */
7cc5e38f	202	unsigned long r_start_stamp; /* jiffies */
bb873b53	203	int r_attempts;
4609245e	204	u32 r_map_dne_bound;
3f1af42a ID	205
3f1af42a ID	206	struct ceph_osd_req_op r_ops[];
f24e9980 SW	207	};
f24e9980 SW	208
205ee118 ID	209	struct ceph_request_redirect {
	210	struct ceph_object_locator oloc;
	211	};
	212
8cb441c0 ID	213	/*
	214	* osd request identifier
	215	*
	216	* caller name + incarnation# + tid to unique identify this request
	217	*/
	218	struct ceph_osd_reqid {
	219	struct ceph_entity_name name;
	220	__le64 tid;
	221	__le32 inc;
	222	} __packed;
	223
	224	struct ceph_blkin_trace_info {
	225	__le64 trace_id;
	226	__le64 span_id;
	227	__le64 parent_span_id;
	228	} __packed;
	229
922dab61 ID	230	typedef void (rados_watchcb2_t)(void arg, u64 notify_id, u64 cookie,
	231	u64 notifier_id, void *data, size_t data_len);
	232	typedef void (rados_watcherrcb_t)(void arg, u64 cookie, int err);
	233
	234	struct ceph_osd_linger_request {
a40c4f10	235	struct ceph_osd_client *osdc;
922dab61 ID	236	u64 linger_id;
922dab61 ID	237	bool committed;
19079203	238	bool is_watch; /* watch or notify */
922dab61 ID	239
	240	struct ceph_osd *osd;
	241	struct ceph_osd_request *reg_req;
	242	struct ceph_osd_request *ping_req;
	243	unsigned long ping_sent;
b07d3c4b ID	244	unsigned long watch_valid_thru;
b07d3c4b ID	245	struct list_head pending_lworks;
922dab61 ID	246
922dab61 ID	247	struct ceph_osd_request_target t;
4609245e	248	u32 map_dne_bound;
922dab61 ID	249
	250	struct timespec mtime;
	251
a40c4f10	252	struct kref kref;
922dab61 ID	253	struct mutex lock;
	254	struct rb_node node; /* osd */
	255	struct rb_node osdc_node; /* osdc */
4609245e	256	struct rb_node mc_node; /* map check */
922dab61 ID	257	struct list_head scan_item;
	258
	259	struct completion reg_commit_wait;
19079203	260	struct completion notify_finish_wait;
922dab61	261	int reg_commit_error;
19079203	262	int notify_finish_error;
922dab61 ID	263	int last_error;
	264
	265	u32 register_gen;
19079203	266	u64 notify_id;
a40c4f10	267
922dab61 ID	268	rados_watchcb2_t wcb;
	269	rados_watcherrcb_t errcb;
	270	void *data;
19079203 ID	271
	272	struct page ***preply_pages;
	273	size_t *preply_len;
a40c4f10 YS	274	};
a40c4f10 YS	275
a4ed38d7 DF	276	struct ceph_watch_item {
	277	struct ceph_entity_name name;
	278	u64 cookie;
	279	struct ceph_entity_addr addr;
	280	};
	281
a02a946d ID	282	struct ceph_spg_mapping {
	283	struct rb_node node;
	284	struct ceph_spg spgid;
	285
	286	struct rb_root backoffs;
	287	};
	288
	289	struct ceph_hobject_id {
	290	void *key;
	291	size_t key_len;
	292	void *oid;
	293	size_t oid_len;
	294	u64 snapid;
	295	u32 hash;
	296	u8 is_max;
	297	void *nspace;
	298	size_t nspace_len;
	299	s64 pool;
	300
	301	/* cache */
	302	u32 hash_reverse_bits;
	303	};
	304
	305	static inline void ceph_hoid_build_hash_cache(struct ceph_hobject_id *hoid)
	306	{
	307	hoid->hash_reverse_bits = bitrev32(hoid->hash);
	308	}
	309
	310	/*
	311	* PG-wide backoff: [begin, end)
	312	* per-object backoff: begin == end
	313	*/
	314	struct ceph_osd_backoff {
	315	struct rb_node spg_node;
	316	struct rb_node id_node;
	317
	318	struct ceph_spg spgid;
	319	u64 id;
	320	struct ceph_hobject_id *begin;
	321	struct ceph_hobject_id *end;
	322	};
	323
264048af ID	324	#define CEPH_LINGER_ID_START 0xffff000000000000ULL
264048af ID	325
f24e9980 SW	326	struct ceph_osd_client {
	327	struct ceph_client *client;
	328
	329	struct ceph_osdmap osdmap; / current map */
5aea3dcd	330	struct rw_semaphore lock;
f24e9980	331
f24e9980	332	struct rb_root osds; /* osds */
f5a2041b	333	struct list_head osd_lru; /* idle osds */
9dd2845c	334	spinlock_t osd_lru_lock;
58eb7932	335	u32 epoch_barrier;
5aea3dcd ID	336	struct ceph_osd homeless_osd;
5aea3dcd ID	337	atomic64_t last_tid; /* tid of last request */
922dab61 ID	338	u64 last_linger_id;
922dab61 ID	339	struct rb_root linger_requests; /* lingering requests */
4609245e ID	340	struct rb_root map_checks;
4609245e ID	341	struct rb_root linger_map_checks;
5aea3dcd ID	342	atomic_t num_requests;
5aea3dcd ID	343	atomic_t num_homeless;
f24e9980	344	struct delayed_work timeout_work;
f5a2041b	345	struct delayed_work osds_timeout_work;
039934b8	346	#ifdef CONFIG_DEBUG_FS
f24e9980	347	struct dentry *debugfs_file;
039934b8	348	#endif
f24e9980 SW	349
	350	mempool_t *req_mempool;
	351
0d59ab81	352	struct ceph_msgpool msgpool_op;
c16e7869	353	struct ceph_msgpool msgpool_op_reply;
a40c4f10	354
a40c4f10	355	struct workqueue_struct *notify_wq;
f24e9980 SW	356	};
f24e9980 SW	357
b7ec35b3 ID	358	static inline bool ceph_osdmap_flag(struct ceph_osd_client *osdc, int flag)
	359	{
	360	return osdc->osdmap->flags & flag;
	361	}
	362
5522ae0b AE	363	extern int ceph_osdc_setup(void);
	364	extern void ceph_osdc_cleanup(void);
	365
f24e9980 SW	366	extern int ceph_osdc_init(struct ceph_osd_client *osdc,
	367	struct ceph_client *client);
	368	extern void ceph_osdc_stop(struct ceph_osd_client *osdc);
	369
	370	extern void ceph_osdc_handle_reply(struct ceph_osd_client *osdc,
	371	struct ceph_msg *msg);
	372	extern void ceph_osdc_handle_map(struct ceph_osd_client *osdc,
	373	struct ceph_msg *msg);
58eb7932	374	void ceph_osdc_update_epoch_barrier(struct ceph_osd_client *osdc, u32 eb);
f24e9980	375
49719778	376	extern void osd_req_op_init(struct ceph_osd_request *osd_req,
144cba14	377	unsigned int which, u16 opcode, u32 flags);
49719778 AE	378
	379	extern void osd_req_op_raw_data_in_pages(struct ceph_osd_request *,
	380	unsigned int which,
	381	struct page **pages, u64 length,
	382	u32 alignment, bool pages_from_pool,
	383	bool own_pages);
	384
c99d2d4a AE	385	extern void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
c99d2d4a AE	386	unsigned int which, u16 opcode,
33803f33 AE	387	u64 offset, u64 length,
33803f33 AE	388	u64 truncate_size, u32 truncate_seq);
c99d2d4a AE	389	extern void osd_req_op_extent_update(struct ceph_osd_request *osd_req,
c99d2d4a AE	390	unsigned int which, u64 length);
2c63f49a YZ	391	extern void osd_req_op_extent_dup_last(struct ceph_osd_request *osd_req,
2c63f49a YZ	392	unsigned int which, u64 offset_inc);
a4ce40a9 AE	393
	394	extern struct ceph_osd_data *osd_req_op_extent_osd_data(
	395	struct ceph_osd_request *osd_req,
406e2c9f	396	unsigned int which);
a4ce40a9 AE	397
a4ce40a9 AE	398	extern void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *,
406e2c9f	399	unsigned int which,
a4ce40a9 AE	400	struct page **pages, u64 length,
	401	u32 alignment, bool pages_from_pool,
	402	bool own_pages);
	403	extern void osd_req_op_extent_osd_data_pagelist(struct ceph_osd_request *,
406e2c9f	404	unsigned int which,
a4ce40a9 AE	405	struct ceph_pagelist *pagelist);
	406	#ifdef CONFIG_BLOCK
	407	extern void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *,
406e2c9f	408	unsigned int which,
a4ce40a9 AE	409	struct bio *bio, size_t bio_length);
	410	#endif /* CONFIG_BLOCK */
	411
04017e29 AE	412	extern void osd_req_op_cls_request_data_pagelist(struct ceph_osd_request *,
	413	unsigned int which,
	414	struct ceph_pagelist *pagelist);
6c57b554 AE	415	extern void osd_req_op_cls_request_data_pages(struct ceph_osd_request *,
	416	unsigned int which,
	417	struct page **pages, u64 length,
	418	u32 alignment, bool pages_from_pool,
	419	bool own_pages);
a4ce40a9	420	extern void osd_req_op_cls_response_data_pages(struct ceph_osd_request *,
c99d2d4a	421	unsigned int which,
a4ce40a9 AE	422	struct page **pages, u64 length,
	423	u32 alignment, bool pages_from_pool,
	424	bool own_pages);
c99d2d4a AE	425	extern void osd_req_op_cls_init(struct ceph_osd_request *osd_req,
c99d2d4a AE	426	unsigned int which, u16 opcode,
04017e29	427	const char class, const char method);
d74b50be YZ	428	extern int osd_req_op_xattr_init(struct ceph_osd_request *osd_req, unsigned int which,
	429	u16 opcode, const char name, const void value,
	430	size_t size, u8 cmp_op, u8 cmp_mode);
c647b8a8 ID	431	extern void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
	432	unsigned int which,
	433	u64 expected_object_size,
	434	u64 expected_write_size);
33803f33	435
3499e8a5	436	extern struct ceph_osd_request ceph_osdc_alloc_request(struct ceph_osd_client osdc,
3499e8a5	437	struct ceph_snap_context *snapc,
acead002	438	unsigned int num_ops,
3499e8a5	439	bool use_mempool,
54a54007	440	gfp_t gfp_flags);
13d1ad16	441	int ceph_osdc_alloc_messages(struct ceph_osd_request *req, gfp_t gfp);
3499e8a5	442
f24e9980 SW	443	extern struct ceph_osd_request ceph_osdc_new_request(struct ceph_osd_client ,
	444	struct ceph_file_layout *layout,
	445	struct ceph_vino vino,
acead002	446	u64 offset, u64 *len,
715e4cd4 YZ	447	unsigned int which, int num_ops,
715e4cd4 YZ	448	int opcode, int flags,
f24e9980	449	struct ceph_snap_context *snapc,
acead002	450	u32 truncate_seq, u64 truncate_size,
153e5167	451	bool use_mempool);
f24e9980	452
9e94af20 ID	453	extern void ceph_osdc_get_request(struct ceph_osd_request *req);
9e94af20 ID	454	extern void ceph_osdc_put_request(struct ceph_osd_request *req);
f24e9980 SW	455
	456	extern int ceph_osdc_start_request(struct ceph_osd_client *osdc,
	457	struct ceph_osd_request *req,
	458	bool nofail);
c9f9b93d	459	extern void ceph_osdc_cancel_request(struct ceph_osd_request *req);
f24e9980 SW	460	extern int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
	461	struct ceph_osd_request *req);
	462	extern void ceph_osdc_sync(struct ceph_osd_client *osdc);
	463
dd935f44	464	extern void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc);
7cca78c9	465	void ceph_osdc_maybe_request_map(struct ceph_osd_client *osdc);
dd935f44	466
428a7158 DF	467	int ceph_osdc_call(struct ceph_osd_client *osdc,
	468	struct ceph_object_id *oid,
	469	struct ceph_object_locator *oloc,
	470	const char class, const char method,
	471	unsigned int flags,
	472	struct page *req_page, size_t req_len,
	473	struct page resp_page, size_t resp_len);
	474
f24e9980 SW	475	extern int ceph_osdc_readpages(struct ceph_osd_client *osdc,
	476	struct ceph_vino vino,
	477	struct ceph_file_layout *layout,
	478	u64 off, u64 *plen,
	479	u32 truncate_seq, u64 truncate_size,
b7495fc2 SW	480	struct page **pages, int nr_pages,
b7495fc2 SW	481	int page_align);
f24e9980 SW	482
	483	extern int ceph_osdc_writepages(struct ceph_osd_client *osdc,
	484	struct ceph_vino vino,
	485	struct ceph_file_layout *layout,
	486	struct ceph_snap_context *sc,
	487	u64 off, u64 len,
	488	u32 truncate_seq, u64 truncate_size,
	489	struct timespec *mtime,
24808826	490	struct page **pages, int nr_pages);
f24e9980	491
922dab61 ID	492	/* watch/notify */
	493	struct ceph_osd_linger_request *
	494	ceph_osdc_watch(struct ceph_osd_client *osdc,
	495	struct ceph_object_id *oid,
	496	struct ceph_object_locator *oloc,
	497	rados_watchcb2_t wcb,
	498	rados_watcherrcb_t errcb,
	499	void *data);
	500	int ceph_osdc_unwatch(struct ceph_osd_client *osdc,
	501	struct ceph_osd_linger_request *lreq);
	502
	503	int ceph_osdc_notify_ack(struct ceph_osd_client *osdc,
	504	struct ceph_object_id *oid,
	505	struct ceph_object_locator *oloc,
	506	u64 notify_id,
	507	u64 cookie,
	508	void *payload,
	509	size_t payload_len);
19079203 ID	510	int ceph_osdc_notify(struct ceph_osd_client *osdc,
	511	struct ceph_object_id *oid,
	512	struct ceph_object_locator *oloc,
	513	void *payload,
	514	size_t payload_len,
	515	u32 timeout,
	516	struct page ***preply_pages,
	517	size_t *preply_len);
b07d3c4b ID	518	int ceph_osdc_watch_check(struct ceph_osd_client *osdc,
b07d3c4b ID	519	struct ceph_osd_linger_request *lreq);
a4ed38d7 DF	520	int ceph_osdc_list_watchers(struct ceph_osd_client *osdc,
	521	struct ceph_object_id *oid,
	522	struct ceph_object_locator *oloc,
	523	struct ceph_watch_item **watchers,
	524	u32 *num_watchers);
f24e9980 SW	525	#endif
f24e9980 SW	526