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

#ifndef _FS_CEPH_OSD_CLIENT_H
#define _FS_CEPH_OSD_CLIENT_H

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

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

/* 
 * Maximum object name size 
 * (must be at least as big as RBD_MAX_MD_NAME_LEN -- currently 100) 
 */
#define MAX_OBJ_NAME_SIZE 100

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

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

/* a given osd we're communicating with */
struct ceph_osd {
	atomic_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 list_head o_requests;
	struct list_head o_linger_requests;
	struct list_head o_osd_lru;
	struct ceph_auth_handshake o_auth;
	unsigned long lru_ttl;
	int o_marked_for_keepalive;
	struct list_head o_keepalive_item;
};


#define CEPH_OSD_MAX_OP 10

enum ceph_osd_data_type {
	CEPH_OSD_DATA_TYPE_NONE,
	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 bio       	*bio;
#endif /* CONFIG_BLOCK */
	};
};

/* an in-flight request */
struct ceph_osd_request {
	u64             r_tid;              /* unique for this client */
	struct rb_node  r_node;
	struct list_head r_req_lru_item;
	struct list_head r_osd_item;
	struct list_head r_linger_item;
	struct list_head r_linger_osd;
	struct ceph_osd *r_osd;
	struct ceph_pg   r_pgid;
	int              r_pg_osds[CEPH_PG_MAX_SIZE];
	int              r_num_pg_osds;

	struct ceph_connection *r_con_filling_msg;

	struct ceph_msg  *r_request, *r_reply;
	int               r_flags;     /* any additional flags for the osd */
	u32               r_sent;      /* >0 if r_request is sending/sent */
	int               r_num_ops;

	/* encoded message content */
	struct ceph_osd_op *r_request_ops;
	/* these are updated on each send */
	__le32           *r_request_osdmap_epoch;
	__le32           *r_request_flags;
	__le64           *r_request_pool;
	void             *r_request_pgid;
	__le32           *r_request_attempts;
	struct ceph_eversion *r_request_reassert_version;

	int               r_result;
	int               r_reply_op_len[CEPH_OSD_MAX_OP];
	s32               r_reply_op_result[CEPH_OSD_MAX_OP];
	int               r_got_reply;
	int		  r_linger;
	int		  r_completed;

	struct ceph_osd_client *r_osdc;
	struct kref       r_kref;
	bool              r_mempool;
	struct completion r_completion, r_safe_completion;
	ceph_osdc_callback_t r_callback, r_safe_callback;
	struct ceph_eversion r_reassert_version;
	struct list_head  r_unsafe_item;

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

	char              r_oid[MAX_OBJ_NAME_SIZE];          /* object name */
	int               r_oid_len;
	u64               r_snapid;
	unsigned long     r_stamp;            /* send OR check time */

	struct ceph_file_layout r_file_layout;
	struct ceph_snap_context *r_snapc;    /* snap context for writes */

	struct ceph_osd_data r_data_in;
	struct ceph_osd_data r_data_out;
};

struct ceph_osd_event {
	u64 cookie;
	int one_shot;
	struct ceph_osd_client *osdc;
	void (*cb)(u64, u64, u8, void *);
	void *data;
	struct rb_node node;
	struct list_head osd_node;
	struct kref kref;
};

struct ceph_osd_event_work {
	struct work_struct work;
	struct ceph_osd_event *event;
        u64 ver;
        u64 notify_id;
        u8 opcode;
};

struct ceph_osd_client {
	struct ceph_client     *client;

	struct ceph_osdmap     *osdmap;       /* current map */
	struct rw_semaphore    map_sem;
	struct completion      map_waiters;
	u64                    last_requested_map;

	struct mutex           request_mutex;
	struct rb_root         osds;          /* osds */
	struct list_head       osd_lru;       /* idle osds */
	u64                    timeout_tid;   /* tid of timeout triggering rq */
	u64                    last_tid;      /* tid of last request */
	struct rb_root         requests;      /* pending requests */
	struct list_head       req_lru;	      /* in-flight lru */
	struct list_head       req_unsent;    /* unsent/need-resend queue */
	struct list_head       req_notarget;  /* map to no osd */
	struct list_head       req_linger;    /* lingering requests */
	int                    num_requests;
	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;

	spinlock_t		event_lock;
	struct rb_root		event_tree;
	u64			event_count;

	struct workqueue_struct	*notify_wq;
};

struct ceph_osd_req_op {
	u16 op;           /* CEPH_OSD_OP_* */
	u32 payload_len;
	union {
		struct {
			u64 offset, length;
			u64 truncate_size;
			u32 truncate_seq;
		} extent;
		struct {
			const char *name;
			const void *val;
			u32 name_len;
			u32 value_len;
			__u8 cmp_op;       /* CEPH_OSD_CMPXATTR_OP_* */
			__u8 cmp_mode;     /* CEPH_OSD_CMPXATTR_MODE_* */
		} xattr;
		struct {
			const char *class_name;
			const char *method_name;
			const void *indata;
			u32 indata_len;
			__u8 class_len;
			__u8 method_len;
			__u8 argc;
		} cls;
		struct {
			u64 cookie;
			u64 count;
		} pgls;
	        struct {
		        u64 snapid;
	        } snap;
		struct {
			u64 cookie;
			u64 ver;
			u32 prot_ver;
			u32 timeout;
			__u8 flag;
		} watch;
	};
};

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);

extern struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
					       struct ceph_snap_context *snapc,
					       unsigned int num_op,
					       bool use_mempool,
					       gfp_t gfp_flags);

extern void ceph_osdc_build_request(struct ceph_osd_request *req, u64 off,
				    unsigned int num_op,
				    struct ceph_osd_req_op *src_ops,
				    struct ceph_snap_context *snapc,
				    u64 snap_id,
				    struct timespec *mtime);

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, int op, int flags,
				      struct ceph_snap_context *snapc,
				      int do_sync, u32 truncate_seq,
				      u64 truncate_size,
				      struct timespec *mtime,
				      bool use_mempool);

extern void ceph_osdc_set_request_linger(struct ceph_osd_client *osdc,
					 struct ceph_osd_request *req);
extern void ceph_osdc_unregister_linger_request(struct ceph_osd_client *osdc,
						struct ceph_osd_request *req);

static inline void ceph_osdc_get_request(struct ceph_osd_request *req)
{
	kref_get(&req->r_kref);
}
extern void ceph_osdc_release_request(struct kref *kref);
static inline void ceph_osdc_put_request(struct ceph_osd_request *req)
{
	kref_put(&req->r_kref, ceph_osdc_release_request);
}

extern int ceph_osdc_start_request(struct ceph_osd_client *osdc,
				   struct ceph_osd_request *req,
				   bool nofail);
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 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 events */
extern int ceph_osdc_create_event(struct ceph_osd_client *osdc,
				  void (*event_cb)(u64, u64, u8, void *),
				  void *data, struct ceph_osd_event **pevent);
extern void ceph_osdc_cancel_event(struct ceph_osd_event *event);
extern void ceph_osdc_put_event(struct ceph_osd_event *event);
#endif
Commit	Line	Data
f24e9980 SW	1	#ifndef _FS_CEPH_OSD_CLIENT_H
	2	#define _FS_CEPH_OSD_CLIENT_H
	3
	4	#include <linux/completion.h>
415e49a9	5	#include <linux/kref.h>
f24e9980 SW	6	#include <linux/mempool.h>
	7	#include <linux/rbtree.h>
	8
6c4a1915 AE	9	#include <linux/ceph/types.h>
	10	#include <linux/ceph/osdmap.h>
	11	#include <linux/ceph/messenger.h>
	12	#include <linux/ceph/auth.h>
c885837f	13	#include <linux/ceph/pagelist.h>
f24e9980	14
224736d9 SP	15	/*
	16	* Maximum object name size
	17	* (must be at least as big as RBD_MAX_MD_NAME_LEN -- currently 100)
	18	*/
	19	#define MAX_OBJ_NAME_SIZE 100
	20
f24e9980 SW	21	struct ceph_msg;
	22	struct ceph_snap_context;
	23	struct ceph_osd_request;
	24	struct ceph_osd_client;
4e7a5dcd	25	struct ceph_authorizer;
f24e9980 SW	26
	27	/*
	28	* completion callback for async writepages
	29	*/
	30	typedef void (ceph_osdc_callback_t)(struct ceph_osd_request ,
	31	struct ceph_msg *);
	32
	33	/* a given osd we're communicating with */
	34	struct ceph_osd {
	35	atomic_t o_ref;
	36	struct ceph_osd_client *o_osdc;
	37	int o_osd;
	38	int o_incarnation;
	39	struct rb_node o_node;
	40	struct ceph_connection o_con;
	41	struct list_head o_requests;
a40c4f10	42	struct list_head o_linger_requests;
f5a2041b	43	struct list_head o_osd_lru;
6c4a1915	44	struct ceph_auth_handshake o_auth;
f5a2041b	45	unsigned long lru_ttl;
422d2cb8 YS	46	int o_marked_for_keepalive;
422d2cb8 YS	47	struct list_head o_keepalive_item;
f24e9980 SW	48	};
f24e9980 SW	49
1b83bef2 SW	50
	51	#define CEPH_OSD_MAX_OP 10
	52
2ac2b7a6 AE	53	enum ceph_osd_data_type {
	54	CEPH_OSD_DATA_TYPE_NONE,
	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
9a5e6d09	74	struct bio *bio;
2794a82a AE	75	#endif /* CONFIG_BLOCK */
	76	};
	77	};
	78
f24e9980 SW	79	/* an in-flight request */
	80	struct ceph_osd_request {
	81	u64 r_tid; /* unique for this client */
	82	struct rb_node r_node;
422d2cb8	83	struct list_head r_req_lru_item;
f24e9980	84	struct list_head r_osd_item;
a40c4f10 YS	85	struct list_head r_linger_item;
a40c4f10 YS	86	struct list_head r_linger_osd;
f24e9980	87	struct ceph_osd *r_osd;
5b191d99	88	struct ceph_pg r_pgid;
d85b7056 SW	89	int r_pg_osds[CEPH_PG_MAX_SIZE];
d85b7056 SW	90	int r_num_pg_osds;
f24e9980	91
0d59ab81	92	struct ceph_connection *r_con_filling_msg;
350b1c32	93
f24e9980	94	struct ceph_msg r_request, r_reply;
f24e9980 SW	95	int r_flags; /* any additional flags for the osd */
f24e9980 SW	96	u32 r_sent; /* >0 if r_request is sending/sent */
1b83bef2 SW	97	int r_num_ops;
	98
	99	/* encoded message content */
	100	struct ceph_osd_op *r_request_ops;
	101	/* these are updated on each send */
	102	__le32 *r_request_osdmap_epoch;
	103	__le32 *r_request_flags;
	104	__le64 *r_request_pool;
	105	void *r_request_pgid;
	106	__le32 *r_request_attempts;
	107	struct ceph_eversion *r_request_reassert_version;
	108
	109	int r_result;
	110	int r_reply_op_len[CEPH_OSD_MAX_OP];
	111	s32 r_reply_op_result[CEPH_OSD_MAX_OP];
350b1c32	112	int r_got_reply;
a40c4f10	113	int r_linger;
0d5af164	114	int r_completed;
f24e9980 SW	115
f24e9980 SW	116	struct ceph_osd_client *r_osdc;
415e49a9	117	struct kref r_kref;
f24e9980 SW	118	bool r_mempool;
	119	struct completion r_completion, r_safe_completion;
	120	ceph_osdc_callback_t r_callback, r_safe_callback;
	121	struct ceph_eversion r_reassert_version;
	122	struct list_head r_unsafe_item;
	123
	124	struct inode r_inode; / for use by callbacks */
3d14c5d2	125	void r_priv; / ditto */
f24e9980	126
224736d9	127	char r_oid[MAX_OBJ_NAME_SIZE]; /* object name */
f24e9980	128	int r_oid_len;
2169aea6	129	u64 r_snapid;
3dd72fc0	130	unsigned long r_stamp; /* send OR check time */
f24e9980 SW	131
	132	struct ceph_file_layout r_file_layout;
	133	struct ceph_snap_context r_snapc; / snap context for writes */
68b4476b	134
0fff87ec AE	135	struct ceph_osd_data r_data_in;
0fff87ec AE	136	struct ceph_osd_data r_data_out;
f24e9980 SW	137	};
f24e9980 SW	138
a40c4f10 YS	139	struct ceph_osd_event {
	140	u64 cookie;
	141	int one_shot;
	142	struct ceph_osd_client *osdc;
	143	void (cb)(u64, u64, u8, void );
	144	void *data;
	145	struct rb_node node;
	146	struct list_head osd_node;
	147	struct kref kref;
a40c4f10 YS	148	};
	149
	150	struct ceph_osd_event_work {
	151	struct work_struct work;
	152	struct ceph_osd_event *event;
	153	u64 ver;
	154	u64 notify_id;
	155	u8 opcode;
	156	};
	157
f24e9980 SW	158	struct ceph_osd_client {
	159	struct ceph_client *client;
	160
	161	struct ceph_osdmap osdmap; / current map */
	162	struct rw_semaphore map_sem;
	163	struct completion map_waiters;
	164	u64 last_requested_map;
	165
	166	struct mutex request_mutex;
	167	struct rb_root osds; /* osds */
f5a2041b	168	struct list_head osd_lru; /* idle osds */
f24e9980 SW	169	u64 timeout_tid; /* tid of timeout triggering rq */
	170	u64 last_tid; /* tid of last request */
	171	struct rb_root requests; /* pending requests */
6f6c7006 SW	172	struct list_head req_lru; /* in-flight lru */
	173	struct list_head req_unsent; /* unsent/need-resend queue */
	174	struct list_head req_notarget; /* map to no osd */
a40c4f10	175	struct list_head req_linger; /* lingering requests */
f24e9980 SW	176	int num_requests;
f24e9980 SW	177	struct delayed_work timeout_work;
f5a2041b	178	struct delayed_work osds_timeout_work;
039934b8	179	#ifdef CONFIG_DEBUG_FS
f24e9980	180	struct dentry *debugfs_file;
039934b8	181	#endif
f24e9980 SW	182
	183	mempool_t *req_mempool;
	184
0d59ab81	185	struct ceph_msgpool msgpool_op;
c16e7869	186	struct ceph_msgpool msgpool_op_reply;
a40c4f10 YS	187
	188	spinlock_t event_lock;
	189	struct rb_root event_tree;
	190	u64 event_count;
	191
	192	struct workqueue_struct *notify_wq;
f24e9980 SW	193	};
f24e9980 SW	194
68b4476b YS	195	struct ceph_osd_req_op {
68b4476b YS	196	u16 op; /* CEPH_OSD_OP_* */
e7e319a9	197	u32 payload_len;
68b4476b YS	198	union {
	199	struct {
	200	u64 offset, length;
	201	u64 truncate_size;
	202	u32 truncate_seq;
	203	} extent;
	204	struct {
	205	const char *name;
2a24d1f4	206	const void *val;
e7e319a9	207	u32 name_len;
68b4476b YS	208	u32 value_len;
	209	__u8 cmp_op; /* CEPH_OSD_CMPXATTR_OP_* */
	210	__u8 cmp_mode; /* CEPH_OSD_CMPXATTR_MODE_* */
	211	} xattr;
	212	struct {
ae1533b6	213	const char *class_name;
ae1533b6	214	const char *method_name;
2a24d1f4	215	const void *indata;
68b4476b	216	u32 indata_len;
e7e319a9 AE	217	__u8 class_len;
	218	__u8 method_len;
	219	__u8 argc;
68b4476b YS	220	} cls;
68b4476b YS	221	struct {
e7e319a9 AE	222	u64 cookie;
e7e319a9 AE	223	u64 count;
68b4476b	224	} pgls;
ae1533b6 YS	225	struct {
	226	u64 snapid;
	227	} snap;
a40c4f10 YS	228	struct {
	229	u64 cookie;
	230	u64 ver;
a40c4f10 YS	231	u32 prot_ver;
a40c4f10 YS	232	u32 timeout;
e7e319a9	233	__u8 flag;
a40c4f10	234	} watch;
68b4476b	235	};
68b4476b YS	236	};
68b4476b YS	237
f24e9980 SW	238	extern int ceph_osdc_init(struct ceph_osd_client *osdc,
	239	struct ceph_client *client);
	240	extern void ceph_osdc_stop(struct ceph_osd_client *osdc);
	241
	242	extern void ceph_osdc_handle_reply(struct ceph_osd_client *osdc,
	243	struct ceph_msg *msg);
	244	extern void ceph_osdc_handle_map(struct ceph_osd_client *osdc,
	245	struct ceph_msg *msg);
	246
3499e8a5	247	extern struct ceph_osd_request ceph_osdc_alloc_request(struct ceph_osd_client osdc,
3499e8a5	248	struct ceph_snap_context *snapc,
ae7ca4a3	249	unsigned int num_op,
3499e8a5	250	bool use_mempool,
54a54007	251	gfp_t gfp_flags);
3499e8a5	252
175face2	253	extern void ceph_osdc_build_request(struct ceph_osd_request *req, u64 off,
ae7ca4a3	254	unsigned int num_op,
68b4476b YS	255	struct ceph_osd_req_op *src_ops,
68b4476b YS	256	struct ceph_snap_context *snapc,
4d6b250b	257	u64 snap_id,
af77f26c	258	struct timespec *mtime);
3499e8a5	259
f24e9980 SW	260	extern struct ceph_osd_request ceph_osdc_new_request(struct ceph_osd_client ,
	261	struct ceph_file_layout *layout,
	262	struct ceph_vino vino,
	263	u64 offset, u64 *len, int op, int flags,
	264	struct ceph_snap_context *snapc,
	265	int do_sync, u32 truncate_seq,
	266	u64 truncate_size,
	267	struct timespec *mtime,
153e5167	268	bool use_mempool);
f24e9980	269
a40c4f10 YS	270	extern void ceph_osdc_set_request_linger(struct ceph_osd_client *osdc,
	271	struct ceph_osd_request *req);
	272	extern void ceph_osdc_unregister_linger_request(struct ceph_osd_client *osdc,
	273	struct ceph_osd_request *req);
	274
f24e9980 SW	275	static inline void ceph_osdc_get_request(struct ceph_osd_request *req)
f24e9980 SW	276	{
415e49a9 SW	277	kref_get(&req->r_kref);
	278	}
	279	extern void ceph_osdc_release_request(struct kref *kref);
	280	static inline void ceph_osdc_put_request(struct ceph_osd_request *req)
	281	{
	282	kref_put(&req->r_kref, ceph_osdc_release_request);
f24e9980	283	}
f24e9980 SW	284
	285	extern int ceph_osdc_start_request(struct ceph_osd_client *osdc,
	286	struct ceph_osd_request *req,
	287	bool nofail);
	288	extern int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
	289	struct ceph_osd_request *req);
	290	extern void ceph_osdc_sync(struct ceph_osd_client *osdc);
	291
	292	extern int ceph_osdc_readpages(struct ceph_osd_client *osdc,
	293	struct ceph_vino vino,
	294	struct ceph_file_layout *layout,
	295	u64 off, u64 *plen,
	296	u32 truncate_seq, u64 truncate_size,
b7495fc2 SW	297	struct page **pages, int nr_pages,
b7495fc2 SW	298	int page_align);
f24e9980 SW	299
	300	extern int ceph_osdc_writepages(struct ceph_osd_client *osdc,
	301	struct ceph_vino vino,
	302	struct ceph_file_layout *layout,
	303	struct ceph_snap_context *sc,
	304	u64 off, u64 len,
	305	u32 truncate_seq, u64 truncate_size,
	306	struct timespec *mtime,
24808826	307	struct page **pages, int nr_pages);
f24e9980	308
a40c4f10 YS	309	/* watch/notify events */
	310	extern int ceph_osdc_create_event(struct ceph_osd_client *osdc,
	311	void (event_cb)(u64, u64, u8, void ),
3c663bbd	312	void data, struct ceph_osd_event *pevent);
a40c4f10	313	extern void ceph_osdc_cancel_event(struct ceph_osd_event *event);
a40c4f10	314	extern void ceph_osdc_put_event(struct ceph_osd_event *event);
f24e9980 SW	315	#endif
f24e9980 SW	316