[thirdparty/kernel/linux.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 */
};

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
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 */
	60	};
	61
2794a82a	62	struct ceph_osd_data {
2ac2b7a6 AE	63	enum ceph_osd_data_type type;
2ac2b7a6 AE	64	union {
2794a82a AE	65	struct {
2794a82a AE	66	struct page **pages;
e0c59487	67	u64 length;
2794a82a AE	68	u32 alignment;
	69	bool pages_from_pool;
	70	bool own_pages;
	71	};
9a5e6d09	72	struct ceph_pagelist *pagelist;
2794a82a	73	#ifdef CONFIG_BLOCK
fdce58cc AE	74	struct {
	75	struct bio bio; / list of bios */
	76	size_t bio_length; /* total in list */
	77	};
2794a82a AE	78	#endif /* CONFIG_BLOCK */
	79	};
	80	};
	81
79528734 AE	82	struct ceph_osd_req_op {
79528734 AE	83	u16 op; /* CEPH_OSD_OP_* */
7b25bf5f	84	u32 flags; /* CEPH_OSD_OP_FLAG_* */
de2aa102	85	u32 indata_len; /* request */
7665d85b YZ	86	u32 outdata_len; /* reply */
	87	s32 rval;
	88
79528734	89	union {
49719778	90	struct ceph_osd_data raw_data_in;
79528734 AE	91	struct {
	92	u64 offset, length;
	93	u64 truncate_size;
	94	u32 truncate_seq;
5476492f	95	struct ceph_osd_data osd_data;
79528734	96	} extent;
d74b50be	97	struct {
d7d5a007 ID	98	u32 name_len;
d7d5a007 ID	99	u32 value_len;
d74b50be YZ	100	__u8 cmp_op; /* CEPH_OSD_CMPXATTR_OP_* */
	101	__u8 cmp_mode; /* CEPH_OSD_CMPXATTR_MODE_* */
	102	struct ceph_osd_data osd_data;
	103	} xattr;
79528734 AE	104	struct {
	105	const char *class_name;
	106	const char *method_name;
5476492f	107	struct ceph_osd_data request_info;
04017e29	108	struct ceph_osd_data request_data;
5476492f	109	struct ceph_osd_data response_data;
79528734 AE	110	__u8 class_len;
79528734 AE	111	__u8 method_len;
bb873b53	112	u32 indata_len;
79528734 AE	113	} cls;
	114	struct {
	115	u64 cookie;
922dab61 ID	116	__u8 op; /* CEPH_OSD_WATCH_OP_ */
922dab61 ID	117	u32 gen;
79528734	118	} watch;
922dab61 ID	119	struct {
	120	struct ceph_osd_data request_data;
	121	} notify_ack;
19079203 ID	122	struct {
	123	u64 cookie;
	124	struct ceph_osd_data request_data;
	125	struct ceph_osd_data response_data;
	126	} notify;
a4ed38d7 DF	127	struct {
	128	struct ceph_osd_data response_data;
	129	} list_watchers;
c647b8a8 ID	130	struct {
	131	u64 expected_object_size;
	132	u64 expected_write_size;
	133	} alloc_hint;
79528734 AE	134	};
	135	};
	136
63244fa1 ID	137	struct ceph_osd_request_target {
	138	struct ceph_object_id base_oid;
	139	struct ceph_object_locator base_oloc;
	140	struct ceph_object_id target_oid;
	141	struct ceph_object_locator target_oloc;
	142
dc98ff72 ID	143	struct ceph_pg pgid; /* last raw pg we mapped to */
dc98ff72 ID	144	struct ceph_spg spgid; /* last actual spg we mapped to */
63244fa1 ID	145	u32 pg_num;
	146	u32 pg_num_mask;
	147	struct ceph_osds acting;
	148	struct ceph_osds up;
	149	int size;
	150	int min_size;
	151	bool sort_bitwise;
ae78dd81	152	bool recovery_deletes;
63244fa1 ID	153
	154	unsigned int flags; /* CEPH_OSD_FLAG_* */
	155	bool paused;
	156
04c7d789	157	u32 epoch;
dc93e0e2 ID	158	u32 last_force_resend;
dc93e0e2 ID	159
63244fa1 ID	160	int osd;
	161	};
	162
f24e9980 SW	163	/* an in-flight request */
	164	struct ceph_osd_request {
	165	u64 r_tid; /* unique for this client */
	166	struct rb_node r_node;
4609245e	167	struct rb_node r_mc_node; /* map check */
f24e9980	168	struct ceph_osd *r_osd;
a66dd383 ID	169
	170	struct ceph_osd_request_target r_t;
	171	#define r_base_oid r_t.base_oid
	172	#define r_base_oloc r_t.base_oloc
	173	#define r_flags r_t.flags
f24e9980 SW	174
f24e9980 SW	175	struct ceph_msg r_request, r_reply;
f24e9980	176	u32 r_sent; /* >0 if r_request is sending/sent */
1b83bef2	177
79528734 AE	178	/* request osd ops array */
79528734 AE	179	unsigned int r_num_ops;
79528734	180
1b83bef2	181	int r_result;
f24e9980 SW	182
f24e9980 SW	183	struct ceph_osd_client *r_osdc;
415e49a9	184	struct kref r_kref;
f24e9980	185	bool r_mempool;
b18b9550	186	struct completion r_completion; /* private to osd_client.c */
26be8808	187	ceph_osdc_callback_t r_callback;
f24e9980 SW	188	struct list_head r_unsafe_item;
	189
	190	struct inode r_inode; / for use by callbacks */
3d14c5d2	191	void r_priv; / ditto */
f24e9980	192
bb873b53 ID	193	/* set by submitter */
	194	u64 r_snapid; /* for reads, CEPH_NOSNAP o/w */
	195	struct ceph_snap_context r_snapc; / for writes */
	196	struct timespec r_mtime; /* ditto */
	197	u64 r_data_offset; /* ditto */
922dab61	198	bool r_linger; /* don't resend on failure */
a1f4020a	199	bool r_abort_on_full; /* return ENOSPC when full */
f24e9980	200
bb873b53 ID	201	/* internal */
bb873b53 ID	202	unsigned long r_stamp; /* jiffies, send or check time */
7cc5e38f	203	unsigned long r_start_stamp; /* jiffies */
bb873b53	204	int r_attempts;
4609245e	205	u32 r_map_dne_bound;
3f1af42a ID	206
3f1af42a ID	207	struct ceph_osd_req_op r_ops[];
f24e9980 SW	208	};
f24e9980 SW	209
205ee118 ID	210	struct ceph_request_redirect {
	211	struct ceph_object_locator oloc;
	212	};
	213
8cb441c0 ID	214	/*
	215	* osd request identifier
	216	*
	217	* caller name + incarnation# + tid to unique identify this request
	218	*/
	219	struct ceph_osd_reqid {
	220	struct ceph_entity_name name;
	221	__le64 tid;
	222	__le32 inc;
	223	} __packed;
	224
	225	struct ceph_blkin_trace_info {
	226	__le64 trace_id;
	227	__le64 span_id;
	228	__le64 parent_span_id;
	229	} __packed;
	230
922dab61 ID	231	typedef void (rados_watchcb2_t)(void arg, u64 notify_id, u64 cookie,
	232	u64 notifier_id, void *data, size_t data_len);
	233	typedef void (rados_watcherrcb_t)(void arg, u64 cookie, int err);
	234
	235	struct ceph_osd_linger_request {
a40c4f10	236	struct ceph_osd_client *osdc;
922dab61 ID	237	u64 linger_id;
922dab61 ID	238	bool committed;
19079203	239	bool is_watch; /* watch or notify */
922dab61 ID	240
	241	struct ceph_osd *osd;
	242	struct ceph_osd_request *reg_req;
	243	struct ceph_osd_request *ping_req;
	244	unsigned long ping_sent;
b07d3c4b ID	245	unsigned long watch_valid_thru;
b07d3c4b ID	246	struct list_head pending_lworks;
922dab61 ID	247
922dab61 ID	248	struct ceph_osd_request_target t;
4609245e	249	u32 map_dne_bound;
922dab61 ID	250
	251	struct timespec mtime;
	252
a40c4f10	253	struct kref kref;
922dab61 ID	254	struct mutex lock;
	255	struct rb_node node; /* osd */
	256	struct rb_node osdc_node; /* osdc */
4609245e	257	struct rb_node mc_node; /* map check */
922dab61 ID	258	struct list_head scan_item;
	259
	260	struct completion reg_commit_wait;
19079203	261	struct completion notify_finish_wait;
922dab61	262	int reg_commit_error;
19079203	263	int notify_finish_error;
922dab61 ID	264	int last_error;
	265
	266	u32 register_gen;
19079203	267	u64 notify_id;
a40c4f10	268
922dab61 ID	269	rados_watchcb2_t wcb;
	270	rados_watcherrcb_t errcb;
	271	void *data;
19079203 ID	272
	273	struct page ***preply_pages;
	274	size_t *preply_len;
a40c4f10 YS	275	};
a40c4f10 YS	276
a4ed38d7 DF	277	struct ceph_watch_item {
	278	struct ceph_entity_name name;
	279	u64 cookie;
	280	struct ceph_entity_addr addr;
	281	};
	282
a02a946d ID	283	struct ceph_spg_mapping {
	284	struct rb_node node;
	285	struct ceph_spg spgid;
	286
	287	struct rb_root backoffs;
	288	};
	289
	290	struct ceph_hobject_id {
	291	void *key;
	292	size_t key_len;
	293	void *oid;
	294	size_t oid_len;
	295	u64 snapid;
	296	u32 hash;
	297	u8 is_max;
	298	void *nspace;
	299	size_t nspace_len;
	300	s64 pool;
	301
	302	/* cache */
	303	u32 hash_reverse_bits;
	304	};
	305
	306	static inline void ceph_hoid_build_hash_cache(struct ceph_hobject_id *hoid)
	307	{
	308	hoid->hash_reverse_bits = bitrev32(hoid->hash);
	309	}
	310
	311	/*
	312	* PG-wide backoff: [begin, end)
	313	* per-object backoff: begin == end
	314	*/
	315	struct ceph_osd_backoff {
	316	struct rb_node spg_node;
	317	struct rb_node id_node;
	318
	319	struct ceph_spg spgid;
	320	u64 id;
	321	struct ceph_hobject_id *begin;
	322	struct ceph_hobject_id *end;
	323	};
	324
264048af ID	325	#define CEPH_LINGER_ID_START 0xffff000000000000ULL
264048af ID	326
f24e9980 SW	327	struct ceph_osd_client {
	328	struct ceph_client *client;
	329
	330	struct ceph_osdmap osdmap; / current map */
5aea3dcd	331	struct rw_semaphore lock;
f24e9980	332
f24e9980	333	struct rb_root osds; /* osds */
f5a2041b	334	struct list_head osd_lru; /* idle osds */
9dd2845c	335	spinlock_t osd_lru_lock;
58eb7932	336	u32 epoch_barrier;
5aea3dcd ID	337	struct ceph_osd homeless_osd;
5aea3dcd ID	338	atomic64_t last_tid; /* tid of last request */
922dab61 ID	339	u64 last_linger_id;
922dab61 ID	340	struct rb_root linger_requests; /* lingering requests */
4609245e ID	341	struct rb_root map_checks;
4609245e ID	342	struct rb_root linger_map_checks;
5aea3dcd ID	343	atomic_t num_requests;
5aea3dcd ID	344	atomic_t num_homeless;
f24e9980	345	struct delayed_work timeout_work;
f5a2041b	346	struct delayed_work osds_timeout_work;
039934b8	347	#ifdef CONFIG_DEBUG_FS
f24e9980	348	struct dentry *debugfs_file;
039934b8	349	#endif
f24e9980 SW	350
	351	mempool_t *req_mempool;
	352
0d59ab81	353	struct ceph_msgpool msgpool_op;
c16e7869	354	struct ceph_msgpool msgpool_op_reply;
a40c4f10	355
a40c4f10	356	struct workqueue_struct *notify_wq;
f24e9980 SW	357	};
f24e9980 SW	358
b7ec35b3 ID	359	static inline bool ceph_osdmap_flag(struct ceph_osd_client *osdc, int flag)
	360	{
	361	return osdc->osdmap->flags & flag;
	362	}
	363
5522ae0b AE	364	extern int ceph_osdc_setup(void);
	365	extern void ceph_osdc_cleanup(void);
	366
f24e9980 SW	367	extern int ceph_osdc_init(struct ceph_osd_client *osdc,
	368	struct ceph_client *client);
	369	extern void ceph_osdc_stop(struct ceph_osd_client *osdc);
	370
	371	extern void ceph_osdc_handle_reply(struct ceph_osd_client *osdc,
	372	struct ceph_msg *msg);
	373	extern void ceph_osdc_handle_map(struct ceph_osd_client *osdc,
	374	struct ceph_msg *msg);
58eb7932	375	void ceph_osdc_update_epoch_barrier(struct ceph_osd_client *osdc, u32 eb);
f24e9980	376
49719778	377	extern void osd_req_op_init(struct ceph_osd_request *osd_req,
144cba14	378	unsigned int which, u16 opcode, u32 flags);
49719778 AE	379
	380	extern void osd_req_op_raw_data_in_pages(struct ceph_osd_request *,
	381	unsigned int which,
	382	struct page **pages, u64 length,
	383	u32 alignment, bool pages_from_pool,
	384	bool own_pages);
	385
c99d2d4a AE	386	extern void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
c99d2d4a AE	387	unsigned int which, u16 opcode,
33803f33 AE	388	u64 offset, u64 length,
33803f33 AE	389	u64 truncate_size, u32 truncate_seq);
c99d2d4a AE	390	extern void osd_req_op_extent_update(struct ceph_osd_request *osd_req,
c99d2d4a AE	391	unsigned int which, u64 length);
2c63f49a YZ	392	extern void osd_req_op_extent_dup_last(struct ceph_osd_request *osd_req,
2c63f49a YZ	393	unsigned int which, u64 offset_inc);
a4ce40a9 AE	394
	395	extern struct ceph_osd_data *osd_req_op_extent_osd_data(
	396	struct ceph_osd_request *osd_req,
406e2c9f	397	unsigned int which);
a4ce40a9 AE	398
a4ce40a9 AE	399	extern void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *,
406e2c9f	400	unsigned int which,
a4ce40a9 AE	401	struct page **pages, u64 length,
	402	u32 alignment, bool pages_from_pool,
	403	bool own_pages);
	404	extern void osd_req_op_extent_osd_data_pagelist(struct ceph_osd_request *,
406e2c9f	405	unsigned int which,
a4ce40a9 AE	406	struct ceph_pagelist *pagelist);
	407	#ifdef CONFIG_BLOCK
	408	extern void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *,
406e2c9f	409	unsigned int which,
a4ce40a9 AE	410	struct bio *bio, size_t bio_length);
	411	#endif /* CONFIG_BLOCK */
	412
04017e29 AE	413	extern void osd_req_op_cls_request_data_pagelist(struct ceph_osd_request *,
	414	unsigned int which,
	415	struct ceph_pagelist *pagelist);
6c57b554 AE	416	extern void osd_req_op_cls_request_data_pages(struct ceph_osd_request *,
	417	unsigned int which,
	418	struct page **pages, u64 length,
	419	u32 alignment, bool pages_from_pool,
	420	bool own_pages);
a4ce40a9	421	extern void osd_req_op_cls_response_data_pages(struct ceph_osd_request *,
c99d2d4a	422	unsigned int which,
a4ce40a9 AE	423	struct page **pages, u64 length,
	424	u32 alignment, bool pages_from_pool,
	425	bool own_pages);
c99d2d4a AE	426	extern void osd_req_op_cls_init(struct ceph_osd_request *osd_req,
c99d2d4a AE	427	unsigned int which, u16 opcode,
04017e29	428	const char class, const char method);
d74b50be YZ	429	extern int osd_req_op_xattr_init(struct ceph_osd_request *osd_req, unsigned int which,
	430	u16 opcode, const char name, const void value,
	431	size_t size, u8 cmp_op, u8 cmp_mode);
c647b8a8 ID	432	extern void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
	433	unsigned int which,
	434	u64 expected_object_size,
	435	u64 expected_write_size);
33803f33	436
3499e8a5	437	extern struct ceph_osd_request ceph_osdc_alloc_request(struct ceph_osd_client osdc,
3499e8a5	438	struct ceph_snap_context *snapc,
acead002	439	unsigned int num_ops,
3499e8a5	440	bool use_mempool,
54a54007	441	gfp_t gfp_flags);
13d1ad16	442	int ceph_osdc_alloc_messages(struct ceph_osd_request *req, gfp_t gfp);
3499e8a5	443
f24e9980 SW	444	extern struct ceph_osd_request ceph_osdc_new_request(struct ceph_osd_client ,
	445	struct ceph_file_layout *layout,
	446	struct ceph_vino vino,
acead002	447	u64 offset, u64 *len,
715e4cd4 YZ	448	unsigned int which, int num_ops,
715e4cd4 YZ	449	int opcode, int flags,
f24e9980	450	struct ceph_snap_context *snapc,
acead002	451	u32 truncate_seq, u64 truncate_size,
153e5167	452	bool use_mempool);
f24e9980	453
9e94af20 ID	454	extern void ceph_osdc_get_request(struct ceph_osd_request *req);
9e94af20 ID	455	extern void ceph_osdc_put_request(struct ceph_osd_request *req);
f24e9980 SW	456
	457	extern int ceph_osdc_start_request(struct ceph_osd_client *osdc,
	458	struct ceph_osd_request *req,
	459	bool nofail);
c9f9b93d	460	extern void ceph_osdc_cancel_request(struct ceph_osd_request *req);
f24e9980 SW	461	extern int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
	462	struct ceph_osd_request *req);
	463	extern void ceph_osdc_sync(struct ceph_osd_client *osdc);
	464
dd935f44	465	extern void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc);
7cca78c9	466	void ceph_osdc_maybe_request_map(struct ceph_osd_client *osdc);
dd935f44	467
428a7158 DF	468	int ceph_osdc_call(struct ceph_osd_client *osdc,
	469	struct ceph_object_id *oid,
	470	struct ceph_object_locator *oloc,
	471	const char class, const char method,
	472	unsigned int flags,
	473	struct page *req_page, size_t req_len,
	474	struct page resp_page, size_t resp_len);
	475
f24e9980 SW	476	extern int ceph_osdc_readpages(struct ceph_osd_client *osdc,
	477	struct ceph_vino vino,
	478	struct ceph_file_layout *layout,
	479	u64 off, u64 *plen,
	480	u32 truncate_seq, u64 truncate_size,
b7495fc2 SW	481	struct page **pages, int nr_pages,
b7495fc2 SW	482	int page_align);
f24e9980 SW	483
	484	extern int ceph_osdc_writepages(struct ceph_osd_client *osdc,
	485	struct ceph_vino vino,
	486	struct ceph_file_layout *layout,
	487	struct ceph_snap_context *sc,
	488	u64 off, u64 len,
	489	u32 truncate_seq, u64 truncate_size,
	490	struct timespec *mtime,
24808826	491	struct page **pages, int nr_pages);
f24e9980	492
922dab61 ID	493	/* watch/notify */
	494	struct ceph_osd_linger_request *
	495	ceph_osdc_watch(struct ceph_osd_client *osdc,
	496	struct ceph_object_id *oid,
	497	struct ceph_object_locator *oloc,
	498	rados_watchcb2_t wcb,
	499	rados_watcherrcb_t errcb,
	500	void *data);
	501	int ceph_osdc_unwatch(struct ceph_osd_client *osdc,
	502	struct ceph_osd_linger_request *lreq);
	503
	504	int ceph_osdc_notify_ack(struct ceph_osd_client *osdc,
	505	struct ceph_object_id *oid,
	506	struct ceph_object_locator *oloc,
	507	u64 notify_id,
	508	u64 cookie,
	509	void *payload,
	510	size_t payload_len);
19079203 ID	511	int ceph_osdc_notify(struct ceph_osd_client *osdc,
	512	struct ceph_object_id *oid,
	513	struct ceph_object_locator *oloc,
	514	void *payload,
	515	size_t payload_len,
	516	u32 timeout,
	517	struct page ***preply_pages,
	518	size_t *preply_len);
b07d3c4b ID	519	int ceph_osdc_watch_check(struct ceph_osd_client *osdc,
b07d3c4b ID	520	struct ceph_osd_linger_request *lreq);
a4ed38d7 DF	521	int ceph_osdc_list_watchers(struct ceph_osd_client *osdc,
	522	struct ceph_object_id *oid,
	523	struct ceph_object_locator *oloc,
	524	struct ceph_watch_item **watchers,
	525	u32 *num_watchers);
f24e9980 SW	526	#endif
f24e9980 SW	527