[deliverable/linux.git] / fs / ceph / mds_client.h

#ifndef _FS_CEPH_MDS_CLIENT_H
#define _FS_CEPH_MDS_CLIENT_H

#include <linux/completion.h>
#include <linux/kref.h>
#include <linux/list.h>
#include <linux/mutex.h>
#include <linux/rbtree.h>
#include <linux/spinlock.h>

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

/*
 * Some lock dependencies:
 *
 * session->s_mutex
 *         mdsc->mutex
 *
 *         mdsc->snap_rwsem
 *
 *         ci->i_ceph_lock
 *                 mdsc->snap_flush_lock
 *                 mdsc->cap_delay_lock
 *
 */

struct ceph_fs_client;
struct ceph_cap;

/*
 * parsed info about a single inode.  pointers are into the encoded
 * on-wire structures within the mds reply message payload.
 */
struct ceph_mds_reply_info_in {
	struct ceph_mds_reply_inode *in;
	struct ceph_dir_layout dir_layout;
	u32 symlink_len;
	char *symlink;
	u32 xattr_len;
	char *xattr_data;
	u64 inline_version;
	u32 inline_len;
	char *inline_data;
};

/*
 * parsed info about an mds reply, including information about
 * either: 1) the target inode and/or its parent directory and dentry,
 * and directory contents (for readdir results), or
 * 2) the file range lock info (for fcntl F_GETLK results).
 */
struct ceph_mds_reply_info_parsed {
	struct ceph_mds_reply_head    *head;

	/* trace */
	struct ceph_mds_reply_info_in diri, targeti;
	struct ceph_mds_reply_dirfrag *dirfrag;
	char                          *dname;
	u32                           dname_len;
	struct ceph_mds_reply_lease   *dlease;

	/* extra */
	union {
		/* for fcntl F_GETLK results */
		struct ceph_filelock *filelock_reply;

		/* for readdir results */
		struct {
			struct ceph_mds_reply_dirfrag *dir_dir;
			size_t			      dir_buf_size;
			int                           dir_nr;
			char                          **dir_dname;
			u32                           *dir_dname_len;
			struct ceph_mds_reply_lease   **dir_dlease;
			struct ceph_mds_reply_info_in *dir_in;
			u8                            dir_complete, dir_end;
		};

		/* for create results */
		struct {
			bool has_create_ino;
			u64 ino;
		};
	};

	/* encoded blob describing snapshot contexts for certain
	   operations (e.g., open) */
	void *snapblob;
	int snapblob_len;
};


/*
 * cap releases are batched and sent to the MDS en masse.
 */
#define CEPH_CAPS_PER_RELEASE ((PAGE_CACHE_SIZE -			\
				sizeof(struct ceph_mds_cap_release)) /	\
			       sizeof(struct ceph_mds_cap_item))


/*
 * state associated with each MDS<->client session
 */
enum {
	CEPH_MDS_SESSION_NEW = 1,
	CEPH_MDS_SESSION_OPENING = 2,
	CEPH_MDS_SESSION_OPEN = 3,
	CEPH_MDS_SESSION_HUNG = 4,
	CEPH_MDS_SESSION_CLOSING = 5,
	CEPH_MDS_SESSION_RESTARTING = 6,
	CEPH_MDS_SESSION_RECONNECTING = 7,
};

struct ceph_mds_session {
	struct ceph_mds_client *s_mdsc;
	int               s_mds;
	int               s_state;
	unsigned long     s_ttl;      /* time until mds kills us */
	u64               s_seq;      /* incoming msg seq # */
	struct mutex      s_mutex;    /* serialize session messages */

	struct ceph_connection s_con;

	struct ceph_auth_handshake s_auth;

	/* protected by s_gen_ttl_lock */
	spinlock_t        s_gen_ttl_lock;
	u32               s_cap_gen;  /* inc each time we get mds stale msg */
	unsigned long     s_cap_ttl;  /* when session caps expire */

	/* protected by s_cap_lock */
	spinlock_t        s_cap_lock;
	struct list_head  s_caps;     /* all caps issued by this session */
	int               s_nr_caps, s_trim_caps;
	int               s_num_cap_releases;
	int		  s_cap_reconnect;
	int		  s_readonly;
	struct list_head  s_cap_releases; /* waiting cap_release messages */
	struct list_head  s_cap_releases_done; /* ready to send */
	struct ceph_cap  *s_cap_iterator;

	/* protected by mutex */
	struct list_head  s_cap_flushing;     /* inodes w/ flushing caps */
	struct list_head  s_cap_snaps_flushing;
	unsigned long     s_renew_requested; /* last time we sent a renew req */
	u64               s_renew_seq;

	atomic_t          s_ref;
	struct list_head  s_waiting;  /* waiting requests */
	struct list_head  s_unsafe;   /* unsafe requests */
};

/*
 * modes of choosing which MDS to send a request to
 */
enum {
	USE_ANY_MDS,
	USE_RANDOM_MDS,
	USE_AUTH_MDS,   /* prefer authoritative mds for this metadata item */
};

struct ceph_mds_request;
struct ceph_mds_client;

/*
 * request completion callback
 */
typedef void (*ceph_mds_request_callback_t) (struct ceph_mds_client *mdsc,
					     struct ceph_mds_request *req);
/*
 * wait for request completion callback
 */
typedef int (*ceph_mds_request_wait_callback_t) (struct ceph_mds_client *mdsc,
						 struct ceph_mds_request *req);

/*
 * an in-flight mds request
 */
struct ceph_mds_request {
	u64 r_tid;                   /* transaction id */
	struct rb_node r_node;
	struct ceph_mds_client *r_mdsc;

	int r_op;                    /* mds op code */

	/* operation on what? */
	struct inode *r_inode;              /* arg1 */
	struct dentry *r_dentry;            /* arg1 */
	struct dentry *r_old_dentry;        /* arg2: rename from or link from */
	struct inode *r_old_dentry_dir;     /* arg2: old dentry's parent dir */
	char *r_path1, *r_path2;
	struct ceph_vino r_ino1, r_ino2;

	struct inode *r_locked_dir; /* dir (if any) i_mutex locked by vfs */
	struct inode *r_target_inode;       /* resulting inode */

	struct mutex r_fill_mutex;

	union ceph_mds_request_args r_args;
	int r_fmode;        /* file mode, if expecting cap */
	kuid_t r_uid;
	kgid_t r_gid;
	struct timespec r_stamp;

	/* for choosing which mds to send this request to */
	int r_direct_mode;
	u32 r_direct_hash;      /* choose dir frag based on this dentry hash */
	bool r_direct_is_hash;  /* true if r_direct_hash is valid */

	/* data payload is used for xattr ops */
	struct ceph_pagelist *r_pagelist;

	/* what caps shall we drop? */
	int r_inode_drop, r_inode_unless;
	int r_dentry_drop, r_dentry_unless;
	int r_old_dentry_drop, r_old_dentry_unless;
	struct inode *r_old_inode;
	int r_old_inode_drop, r_old_inode_unless;

	struct ceph_msg  *r_request;  /* original request */
	int r_request_release_offset;
	struct ceph_msg  *r_reply;
	struct ceph_mds_reply_info_parsed r_reply_info;
	struct page *r_locked_page;
	int r_err;
	bool r_aborted;

	unsigned long r_timeout;  /* optional.  jiffies */
	unsigned long r_started;  /* start time to measure timeout against */
	unsigned long r_request_started; /* start time for mds request only,
					    used to measure lease durations */

	/* link unsafe requests to parent directory, for fsync */
	struct inode	*r_unsafe_dir;
	struct list_head r_unsafe_dir_item;

	struct ceph_mds_session *r_session;

	int               r_attempts;   /* resend attempts */
	int               r_num_fwd;    /* number of forward attempts */
	int               r_resend_mds; /* mds to resend to next, if any*/
	u32               r_sent_on_mseq; /* cap mseq request was sent at*/

	struct kref       r_kref;
	struct list_head  r_wait;
	struct completion r_completion;
	struct completion r_safe_completion;
	ceph_mds_request_callback_t r_callback;
	ceph_mds_request_wait_callback_t r_wait_for_completion;
	struct list_head  r_unsafe_item;  /* per-session unsafe list item */
	bool		  r_got_unsafe, r_got_safe, r_got_result;

	bool              r_did_prepopulate;
	u32               r_readdir_offset;

	struct ceph_cap_reservation r_caps_reservation;
	int r_num_caps;
};

/*
 * mds client state
 */
struct ceph_mds_client {
	struct ceph_fs_client  *fsc;
	struct mutex            mutex;         /* all nested structures */

	struct ceph_mdsmap      *mdsmap;
	struct completion       safe_umount_waiters;
	wait_queue_head_t       session_close_wq;
	struct list_head        waiting_for_map;

	struct ceph_mds_session **sessions;    /* NULL for mds if no session */
	int                     max_sessions;  /* len of s_mds_sessions */
	int                     stopping;      /* true if shutting down */

	/*
	 * snap_rwsem will cover cap linkage into snaprealms, and
	 * realm snap contexts.  (later, we can do per-realm snap
	 * contexts locks..)  the empty list contains realms with no
	 * references (implying they contain no inodes with caps) that
	 * should be destroyed.
	 */
	struct rw_semaphore     snap_rwsem;
	struct rb_root          snap_realms;
	struct list_head        snap_empty;
	spinlock_t              snap_empty_lock;  /* protect snap_empty */

	u64                    last_tid;      /* most recent mds request */
	struct rb_root         request_tree;  /* pending mds requests */
	struct delayed_work    delayed_work;  /* delayed work */
	unsigned long    last_renew_caps;  /* last time we renewed our caps */
	struct list_head cap_delay_list;   /* caps with delayed release */
	spinlock_t       cap_delay_lock;   /* protects cap_delay_list */
	struct list_head snap_flush_list;  /* cap_snaps ready to flush */
	spinlock_t       snap_flush_lock;

	u64               cap_flush_seq;
	struct list_head  cap_dirty;        /* inodes with dirty caps */
	struct list_head  cap_dirty_migrating; /* ...that are migration... */
	int               num_cap_flushing; /* # caps we are flushing */
	spinlock_t        cap_dirty_lock;   /* protects above items */
	wait_queue_head_t cap_flushing_wq;

	/*
	 * Cap reservations
	 *
	 * Maintain a global pool of preallocated struct ceph_caps, referenced
	 * by struct ceph_caps_reservations.  This ensures that we preallocate
	 * memory needed to successfully process an MDS response.  (If an MDS
	 * sends us cap information and we fail to process it, we will have
	 * problems due to the client and MDS being out of sync.)
	 *
	 * Reservations are 'owned' by a ceph_cap_reservation context.
	 */
	spinlock_t	caps_list_lock;
	struct		list_head caps_list; /* unused (reserved or
						unreserved) */
	int		caps_total_count;    /* total caps allocated */
	int		caps_use_count;      /* in use */
	int		caps_reserve_count;  /* unused, reserved */
	int		caps_avail_count;    /* unused, unreserved */
	int		caps_min_count;      /* keep at least this many
						(unreserved) */
	spinlock_t	  dentry_lru_lock;
	struct list_head  dentry_lru;
	int		  num_dentry;
};

extern const char *ceph_mds_op_name(int op);

extern struct ceph_mds_session *
__ceph_lookup_mds_session(struct ceph_mds_client *, int mds);

static inline struct ceph_mds_session *
ceph_get_mds_session(struct ceph_mds_session *s)
{
	atomic_inc(&s->s_ref);
	return s;
}

extern const char *ceph_session_state_name(int s);

extern void ceph_put_mds_session(struct ceph_mds_session *s);

extern int ceph_send_msg_mds(struct ceph_mds_client *mdsc,
			     struct ceph_msg *msg, int mds);

extern int ceph_mdsc_init(struct ceph_fs_client *fsc);
extern void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc);
extern void ceph_mdsc_destroy(struct ceph_fs_client *fsc);

extern void ceph_mdsc_sync(struct ceph_mds_client *mdsc);

extern void ceph_mdsc_lease_release(struct ceph_mds_client *mdsc,
				    struct inode *inode,
				    struct dentry *dn);

extern void ceph_invalidate_dir_request(struct ceph_mds_request *req);
extern int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
					   struct inode *dir);
extern struct ceph_mds_request *
ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode);
extern void ceph_mdsc_submit_request(struct ceph_mds_client *mdsc,
				     struct ceph_mds_request *req);
extern int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
				struct inode *dir,
				struct ceph_mds_request *req);
static inline void ceph_mdsc_get_request(struct ceph_mds_request *req)
{
	kref_get(&req->r_kref);
}
extern void ceph_mdsc_release_request(struct kref *kref);
static inline void ceph_mdsc_put_request(struct ceph_mds_request *req)
{
	kref_put(&req->r_kref, ceph_mdsc_release_request);
}

extern int ceph_add_cap_releases(struct ceph_mds_client *mdsc,
				 struct ceph_mds_session *session);
extern void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
				   struct ceph_mds_session *session);

extern void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc);

extern char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base,
				  int stop_on_nosnap);

extern void __ceph_mdsc_drop_dentry_lease(struct dentry *dentry);
extern void ceph_mdsc_lease_send_msg(struct ceph_mds_session *session,
				     struct inode *inode,
				     struct dentry *dentry, char action,
				     u32 seq);

extern void ceph_mdsc_handle_map(struct ceph_mds_client *mdsc,
				 struct ceph_msg *msg);

extern struct ceph_mds_session *
ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target);
extern void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
					  struct ceph_mds_session *session);

#endif
Commit	Line	Data
2f2dc053 SW	1	#ifndef _FS_CEPH_MDS_CLIENT_H
	2	#define _FS_CEPH_MDS_CLIENT_H
	3
	4	#include <linux/completion.h>
153c8e6b	5	#include <linux/kref.h>
2f2dc053 SW	6	#include <linux/list.h>
2f2dc053 SW	7	#include <linux/mutex.h>
44ca18f2	8	#include <linux/rbtree.h>
2f2dc053 SW	9	#include <linux/spinlock.h>
2f2dc053 SW	10
3d14c5d2 YS	11	#include <linux/ceph/types.h>
	12	#include <linux/ceph/messenger.h>
	13	#include <linux/ceph/mdsmap.h>
6c4a1915	14	#include <linux/ceph/auth.h>
2f2dc053 SW	15
	16	/*
	17	* Some lock dependencies:
	18	*
	19	* session->s_mutex
	20	* mdsc->mutex
	21	*
	22	* mdsc->snap_rwsem
	23	*
be655596	24	* ci->i_ceph_lock
2f2dc053 SW	25	* mdsc->snap_flush_lock
	26	* mdsc->cap_delay_lock
	27	*
	28	*/
	29
3d14c5d2	30	struct ceph_fs_client;
2f2dc053 SW	31	struct ceph_cap;
	32
	33	/*
	34	* parsed info about a single inode. pointers are into the encoded
	35	* on-wire structures within the mds reply message payload.
	36	*/
	37	struct ceph_mds_reply_info_in {
	38	struct ceph_mds_reply_inode *in;
14303d20	39	struct ceph_dir_layout dir_layout;
2f2dc053 SW	40	u32 symlink_len;
	41	char *symlink;
	42	u32 xattr_len;
	43	char *xattr_data;
fb01d1f8 YZ	44	u64 inline_version;
	45	u32 inline_len;
	46	char *inline_data;
2f2dc053 SW	47	};
	48
	49	/*
25933abd HS	50	* parsed info about an mds reply, including information about
	51	* either: 1) the target inode and/or its parent directory and dentry,
	52	* and directory contents (for readdir results), or
	53	* 2) the file range lock info (for fcntl F_GETLK results).
2f2dc053 SW	54	*/
	55	struct ceph_mds_reply_info_parsed {
	56	struct ceph_mds_reply_head *head;
	57
25933abd	58	/* trace */
2f2dc053 SW	59	struct ceph_mds_reply_info_in diri, targeti;
	60	struct ceph_mds_reply_dirfrag *dirfrag;
	61	char *dname;
	62	u32 dname_len;
	63	struct ceph_mds_reply_lease *dlease;
	64
25933abd HS	65	/* extra */
	66	union {
	67	/* for fcntl F_GETLK results */
	68	struct ceph_filelock *filelock_reply;
	69
	70	/* for readdir results */
	71	struct {
	72	struct ceph_mds_reply_dirfrag *dir_dir;
54008399	73	size_t dir_buf_size;
25933abd HS	74	int dir_nr;
	75	char **dir_dname;
	76	u32 *dir_dname_len;
	77	struct ceph_mds_reply_lease **dir_dlease;
	78	struct ceph_mds_reply_info_in *dir_in;
	79	u8 dir_complete, dir_end;
	80	};
6e8575fa SL	81
	82	/* for create results */
	83	struct {
	84	bool has_create_ino;
	85	u64 ino;
	86	};
25933abd	87	};
2f2dc053 SW	88
	89	/* encoded blob describing snapshot contexts for certain
	90	operations (e.g., open) */
	91	void *snapblob;
	92	int snapblob_len;
	93	};
	94
	95
	96	/*
	97	* cap releases are batched and sent to the MDS en masse.
	98	*/
	99	#define CEPH_CAPS_PER_RELEASE ((PAGE_CACHE_SIZE - \
	100	sizeof(struct ceph_mds_cap_release)) / \
	101	sizeof(struct ceph_mds_cap_item))
	102
	103
	104	/*
	105	* state associated with each MDS<->client session
	106	*/
	107	enum {
	108	CEPH_MDS_SESSION_NEW = 1,
	109	CEPH_MDS_SESSION_OPENING = 2,
	110	CEPH_MDS_SESSION_OPEN = 3,
	111	CEPH_MDS_SESSION_HUNG = 4,
	112	CEPH_MDS_SESSION_CLOSING = 5,
	113	CEPH_MDS_SESSION_RESTARTING = 6,
	114	CEPH_MDS_SESSION_RECONNECTING = 7,
	115	};
	116
	117	struct ceph_mds_session {
	118	struct ceph_mds_client *s_mdsc;
	119	int s_mds;
	120	int s_state;
	121	unsigned long s_ttl; /* time until mds kills us */
	122	u64 s_seq; /* incoming msg seq # */
	123	struct mutex s_mutex; /* serialize session messages */
	124
	125	struct ceph_connection s_con;
	126
6c4a1915	127	struct ceph_auth_handshake s_auth;
4e7a5dcd	128
d8fb02ab AE	129	/* protected by s_gen_ttl_lock */
d8fb02ab AE	130	spinlock_t s_gen_ttl_lock;
2f2dc053 SW	131	u32 s_cap_gen; /* inc each time we get mds stale msg */
2f2dc053 SW	132	unsigned long s_cap_ttl; /* when session caps expire */
d8fb02ab AE	133
	134	/* protected by s_cap_lock */
	135	spinlock_t s_cap_lock;
2f2dc053 SW	136	struct list_head s_caps; /* all caps issued by this session */
	137	int s_nr_caps, s_trim_caps;
	138	int s_num_cap_releases;
99a9c273	139	int s_cap_reconnect;
03f4fcb0	140	int s_readonly;
2f2dc053 SW	141	struct list_head s_cap_releases; /* waiting cap_release messages */
2f2dc053 SW	142	struct list_head s_cap_releases_done; /* ready to send */
7c1332b8	143	struct ceph_cap *s_cap_iterator;
2f2dc053 SW	144
	145	/* protected by mutex */
	146	struct list_head s_cap_flushing; /* inodes w/ flushing caps */
	147	struct list_head s_cap_snaps_flushing;
	148	unsigned long s_renew_requested; /* last time we sent a renew req */
	149	u64 s_renew_seq;
	150
	151	atomic_t s_ref;
	152	struct list_head s_waiting; /* waiting requests */
	153	struct list_head s_unsafe; /* unsafe requests */
	154	};
	155
	156	/*
	157	* modes of choosing which MDS to send a request to
	158	*/
	159	enum {
	160	USE_ANY_MDS,
	161	USE_RANDOM_MDS,
	162	USE_AUTH_MDS, /* prefer authoritative mds for this metadata item */
	163	};
	164
	165	struct ceph_mds_request;
	166	struct ceph_mds_client;
	167
	168	/*
	169	* request completion callback
	170	*/
	171	typedef void (ceph_mds_request_callback_t) (struct ceph_mds_client mdsc,
	172	struct ceph_mds_request *req);
9280be24 YZ	173	/*
	174	* wait for request completion callback
	175	*/
	176	typedef int (ceph_mds_request_wait_callback_t) (struct ceph_mds_client mdsc,
	177	struct ceph_mds_request *req);
2f2dc053 SW	178
	179	/*
	180	* an in-flight mds request
	181	*/
	182	struct ceph_mds_request {
	183	u64 r_tid; /* transaction id */
44ca18f2	184	struct rb_node r_node;
37151668	185	struct ceph_mds_client *r_mdsc;
2f2dc053 SW	186
2f2dc053 SW	187	int r_op; /* mds op code */
2f2dc053 SW	188
	189	/* operation on what? */
	190	struct inode r_inode; / arg1 */
	191	struct dentry r_dentry; / arg1 */
	192	struct dentry r_old_dentry; / arg2: rename from or link from */
41b02e1f	193	struct inode r_old_dentry_dir; / arg2: old dentry's parent dir */
2f2dc053 SW	194	char r_path1, r_path2;
	195	struct ceph_vino r_ino1, r_ino2;
	196
	197	struct inode r_locked_dir; / dir (if any) i_mutex locked by vfs */
	198	struct inode r_target_inode; / resulting inode */
	199
b4556396 SW	200	struct mutex r_fill_mutex;
b4556396 SW	201
2f2dc053 SW	202	union ceph_mds_request_args r_args;
2f2dc053 SW	203	int r_fmode; /* file mode, if expecting cap */
ff3d0046 EB	204	kuid_t r_uid;
ff3d0046 EB	205	kgid_t r_gid;
b8e69066	206	struct timespec r_stamp;
2f2dc053 SW	207
	208	/* for choosing which mds to send this request to */
	209	int r_direct_mode;
	210	u32 r_direct_hash; /* choose dir frag based on this dentry hash */
	211	bool r_direct_is_hash; /* true if r_direct_hash is valid */
	212
	213	/* data payload is used for xattr ops */
25e6bae3	214	struct ceph_pagelist *r_pagelist;
2f2dc053 SW	215
	216	/* what caps shall we drop? */
	217	int r_inode_drop, r_inode_unless;
	218	int r_dentry_drop, r_dentry_unless;
	219	int r_old_dentry_drop, r_old_dentry_unless;
	220	struct inode *r_old_inode;
	221	int r_old_inode_drop, r_old_inode_unless;
	222
	223	struct ceph_msg r_request; / original request */
e979cf50	224	int r_request_release_offset;
2f2dc053 SW	225	struct ceph_msg *r_reply;
2f2dc053 SW	226	struct ceph_mds_reply_info_parsed r_reply_info;
01deead0	227	struct page *r_locked_page;
2f2dc053	228	int r_err;
5b1daecd	229	bool r_aborted;
2f2dc053 SW	230
	231	unsigned long r_timeout; /* optional. jiffies */
	232	unsigned long r_started; /* start time to measure timeout against */
	233	unsigned long r_request_started; /* start time for mds request only,
	234	used to measure lease durations */
	235
	236	/* link unsafe requests to parent directory, for fsync */
	237	struct inode *r_unsafe_dir;
	238	struct list_head r_unsafe_dir_item;
	239
	240	struct ceph_mds_session *r_session;
	241
	242	int r_attempts; /* resend attempts */
	243	int r_num_fwd; /* number of forward attempts */
2f2dc053	244	int r_resend_mds; /* mds to resend to next, if any*/
e55b71f8	245	u32 r_sent_on_mseq; /* cap mseq request was sent at*/
2f2dc053	246
153c8e6b	247	struct kref r_kref;
2f2dc053 SW	248	struct list_head r_wait;
	249	struct completion r_completion;
	250	struct completion r_safe_completion;
	251	ceph_mds_request_callback_t r_callback;
9280be24	252	ceph_mds_request_wait_callback_t r_wait_for_completion;
2f2dc053	253	struct list_head r_unsafe_item; /* per-session unsafe list item */
e1518c7c	254	bool r_got_unsafe, r_got_safe, r_got_result;
2f2dc053 SW	255
	256	bool r_did_prepopulate;
	257	u32 r_readdir_offset;
	258
	259	struct ceph_cap_reservation r_caps_reservation;
	260	int r_num_caps;
	261	};
	262
	263	/*
	264	* mds client state
	265	*/
	266	struct ceph_mds_client {
3d14c5d2	267	struct ceph_fs_client *fsc;
2f2dc053 SW	268	struct mutex mutex; /* all nested structures */
	269
	270	struct ceph_mdsmap *mdsmap;
f3c60c59 SW	271	struct completion safe_umount_waiters;
f3c60c59 SW	272	wait_queue_head_t session_close_wq;
2f2dc053 SW	273	struct list_head waiting_for_map;
	274
	275	struct ceph_mds_session *sessions; / NULL for mds if no session */
	276	int max_sessions; /* len of s_mds_sessions */
	277	int stopping; /* true if shutting down */
	278
	279	/*
	280	* snap_rwsem will cover cap linkage into snaprealms, and
	281	* realm snap contexts. (later, we can do per-realm snap
	282	* contexts locks..) the empty list contains realms with no
	283	* references (implying they contain no inodes with caps) that
	284	* should be destroyed.
	285	*/
	286	struct rw_semaphore snap_rwsem;
a105f00c	287	struct rb_root snap_realms;
2f2dc053 SW	288	struct list_head snap_empty;
	289	spinlock_t snap_empty_lock; /* protect snap_empty */
	290
	291	u64 last_tid; /* most recent mds request */
44ca18f2	292	struct rb_root request_tree; /* pending mds requests */
2f2dc053 SW	293	struct delayed_work delayed_work; /* delayed work */
	294	unsigned long last_renew_caps; /* last time we renewed our caps */
	295	struct list_head cap_delay_list; /* caps with delayed release */
	296	spinlock_t cap_delay_lock; /* protects cap_delay_list */
	297	struct list_head snap_flush_list; /* cap_snaps ready to flush */
	298	spinlock_t snap_flush_lock;
	299
	300	u64 cap_flush_seq;
	301	struct list_head cap_dirty; /* inodes with dirty caps */
db354052	302	struct list_head cap_dirty_migrating; /* ...that are migration... */
2f2dc053 SW	303	int num_cap_flushing; /* # caps we are flushing */
	304	spinlock_t cap_dirty_lock; /* protects above items */
	305	wait_queue_head_t cap_flushing_wq;
	306
37151668 YS	307	/*
	308	* Cap reservations
	309	*
	310	* Maintain a global pool of preallocated struct ceph_caps, referenced
	311	* by struct ceph_caps_reservations. This ensures that we preallocate
	312	* memory needed to successfully process an MDS response. (If an MDS
	313	* sends us cap information and we fail to process it, we will have
	314	* problems due to the client and MDS being out of sync.)
	315	*
	316	* Reservations are 'owned' by a ceph_cap_reservation context.
	317	*/
	318	spinlock_t caps_list_lock;
	319	struct list_head caps_list; /* unused (reserved or
	320	unreserved) */
	321	int caps_total_count; /* total caps allocated */
	322	int caps_use_count; /* in use */
	323	int caps_reserve_count; /* unused, reserved */
	324	int caps_avail_count; /* unused, unreserved */
	325	int caps_min_count; /* keep at least this many
	326	(unreserved) */
2f2dc053 SW	327	spinlock_t dentry_lru_lock;
	328	struct list_head dentry_lru;
	329	int num_dentry;
	330	};
	331
	332	extern const char *ceph_mds_op_name(int op);
	333
	334	extern struct ceph_mds_session *
	335	__ceph_lookup_mds_session(struct ceph_mds_client *, int mds);
	336
	337	static inline struct ceph_mds_session *
	338	ceph_get_mds_session(struct ceph_mds_session *s)
	339	{
	340	atomic_inc(&s->s_ref);
	341	return s;
	342	}
	343
a687ecaf JS	344	extern const char *ceph_session_state_name(int s);
a687ecaf JS	345
2f2dc053 SW	346	extern void ceph_put_mds_session(struct ceph_mds_session *s);
	347
	348	extern int ceph_send_msg_mds(struct ceph_mds_client *mdsc,
	349	struct ceph_msg *msg, int mds);
	350
3d14c5d2	351	extern int ceph_mdsc_init(struct ceph_fs_client *fsc);
2f2dc053	352	extern void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc);
3d14c5d2	353	extern void ceph_mdsc_destroy(struct ceph_fs_client *fsc);
2f2dc053 SW	354
	355	extern void ceph_mdsc_sync(struct ceph_mds_client *mdsc);
	356
	357	extern void ceph_mdsc_lease_release(struct ceph_mds_client *mdsc,
	358	struct inode *inode,
2f90b852	359	struct dentry *dn);
2f2dc053	360
167c9e35	361	extern void ceph_invalidate_dir_request(struct ceph_mds_request *req);
54008399 YZ	362	extern int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
54008399 YZ	363	struct inode *dir);
2f2dc053 SW	364	extern struct ceph_mds_request *
	365	ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode);
	366	extern void ceph_mdsc_submit_request(struct ceph_mds_client *mdsc,
	367	struct ceph_mds_request *req);
	368	extern int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
	369	struct inode *dir,
	370	struct ceph_mds_request *req);
	371	static inline void ceph_mdsc_get_request(struct ceph_mds_request *req)
	372	{
153c8e6b SW	373	kref_get(&req->r_kref);
	374	}
	375	extern void ceph_mdsc_release_request(struct kref *kref);
	376	static inline void ceph_mdsc_put_request(struct ceph_mds_request *req)
	377	{
	378	kref_put(&req->r_kref, ceph_mdsc_release_request);
2f2dc053	379	}
2f2dc053	380
2b2300d6	381	extern int ceph_add_cap_releases(struct ceph_mds_client *mdsc,
ee6b272b	382	struct ceph_mds_session *session);
3d7ded4d SW	383	extern void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
	384	struct ceph_mds_session *session);
	385
2f2dc053 SW	386	extern void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc);
	387
	388	extern char ceph_mdsc_build_path(struct dentry dentry, int plen, u64 base,
	389	int stop_on_nosnap);
	390
	391	extern void __ceph_mdsc_drop_dentry_lease(struct dentry *dentry);
	392	extern void ceph_mdsc_lease_send_msg(struct ceph_mds_session *session,
	393	struct inode *inode,
	394	struct dentry *dentry, char action,
	395	u32 seq);
	396
	397	extern void ceph_mdsc_handle_map(struct ceph_mds_client *mdsc,
	398	struct ceph_msg *msg);
	399
5d72d13c YZ	400	extern struct ceph_mds_session *
5d72d13c YZ	401	ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target);
154f42c2 SW	402	extern void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
	403	struct ceph_mds_session *session);
	404
2f2dc053	405	#endif