[deliverable/linux.git] / include / linux / ceph / libceph.h

#ifndef _FS_CEPH_LIBCEPH_H
#define _FS_CEPH_LIBCEPH_H

#include <linux/ceph/ceph_debug.h>

#include <asm/unaligned.h>
#include <linux/backing-dev.h>
#include <linux/completion.h>
#include <linux/exportfs.h>
#include <linux/bug.h>
#include <linux/fs.h>
#include <linux/mempool.h>
#include <linux/pagemap.h>
#include <linux/wait.h>
#include <linux/writeback.h>
#include <linux/slab.h>

#include <linux/ceph/types.h>
#include <linux/ceph/messenger.h>
#include <linux/ceph/msgpool.h>
#include <linux/ceph/mon_client.h>
#include <linux/ceph/osd_client.h>
#include <linux/ceph/ceph_fs.h>

/*
 * mount options
 */
#define CEPH_OPT_FSID             (1<<0)
#define CEPH_OPT_NOSHARE          (1<<1) /* don't share client with other sbs */
#define CEPH_OPT_MYIP             (1<<2) /* specified my ip */
#define CEPH_OPT_NOCRC            (1<<3) /* no data crc on writes */
#define CEPH_OPT_NOMSGAUTH	  (1<<4) /* don't require msg signing feat */
#define CEPH_OPT_TCP_NODELAY	  (1<<5) /* TCP_NODELAY on TCP sockets */
#define CEPH_OPT_NOMSGSIGN	  (1<<6) /* don't sign msgs */

#define CEPH_OPT_DEFAULT   (CEPH_OPT_TCP_NODELAY)

#define ceph_set_opt(client, opt) \
	(client)->options->flags |= CEPH_OPT_##opt;
#define ceph_test_opt(client, opt) \
	(!!((client)->options->flags & CEPH_OPT_##opt))

struct ceph_options {
	int flags;
	struct ceph_fsid fsid;
	struct ceph_entity_addr my_addr;
	unsigned long mount_timeout;		/* jiffies */
	unsigned long osd_idle_ttl;		/* jiffies */
	unsigned long osd_keepalive_timeout;	/* jiffies */

	/*
	 * any type that can't be simply compared or doesn't need need
	 * to be compared should go beyond this point,
	 * ceph_compare_options() should be updated accordingly
	 */

	struct ceph_entity_addr *mon_addr; /* should be the first
					      pointer type of args */
	int num_mon;
	char *name;
	struct ceph_crypto_key *key;
};

/*
 * defaults
 */
#define CEPH_MOUNT_TIMEOUT_DEFAULT	msecs_to_jiffies(60 * 1000)
#define CEPH_OSD_KEEPALIVE_DEFAULT	msecs_to_jiffies(5 * 1000)
#define CEPH_OSD_IDLE_TTL_DEFAULT	msecs_to_jiffies(60 * 1000)

#define CEPH_MONC_HUNT_INTERVAL		msecs_to_jiffies(3 * 1000)
#define CEPH_MONC_PING_INTERVAL		msecs_to_jiffies(10 * 1000)
#define CEPH_MONC_PING_TIMEOUT		msecs_to_jiffies(30 * 1000)
#define CEPH_MONC_HUNT_BACKOFF		2
#define CEPH_MONC_HUNT_MAX_MULT		10

#define CEPH_MSG_MAX_FRONT_LEN	(16*1024*1024)
#define CEPH_MSG_MAX_MIDDLE_LEN	(16*1024*1024)
#define CEPH_MSG_MAX_DATA_LEN	(16*1024*1024)

#define CEPH_AUTH_NAME_DEFAULT   "guest"

/*
 * Delay telling the MDS we no longer want caps, in case we reopen
 * the file.  Delay a minimum amount of time, even if we send a cap
 * message for some other reason.  Otherwise, take the oppotunity to
 * update the mds to avoid sending another message later.
 */
#define CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT      5  /* cap release delay */
#define CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT     60  /* cap release delay */

#define CEPH_CAP_RELEASE_SAFETY_DEFAULT        (CEPH_CAPS_PER_RELEASE * 4)

/* mount state */
enum {
	CEPH_MOUNT_MOUNTING,
	CEPH_MOUNT_MOUNTED,
	CEPH_MOUNT_UNMOUNTING,
	CEPH_MOUNT_UNMOUNTED,
	CEPH_MOUNT_SHUTDOWN,
};

static inline unsigned long ceph_timeout_jiffies(unsigned long timeout)
{
	return timeout ?: MAX_SCHEDULE_TIMEOUT;
}

struct ceph_mds_client;

/*
 * per client state
 *
 * possibly shared by multiple mount points, if they are
 * mounting the same ceph filesystem/cluster.
 */
struct ceph_client {
	struct ceph_fsid fsid;
	bool have_fsid;

	void *private;

	struct ceph_options *options;

	struct mutex mount_mutex;      /* serialize mount attempts */
	wait_queue_head_t auth_wq;
	int auth_err;

	int (*extra_mon_dispatch)(struct ceph_client *, struct ceph_msg *);

	u64 supported_features;
	u64 required_features;

	struct ceph_messenger msgr;   /* messenger instance */
	struct ceph_mon_client monc;
	struct ceph_osd_client osdc;

#ifdef CONFIG_DEBUG_FS
	struct dentry *debugfs_dir;
	struct dentry *debugfs_monmap;
	struct dentry *debugfs_osdmap;
	struct dentry *debugfs_options;
#endif
};

#define from_msgr(ms)	container_of(ms, struct ceph_client, msgr)


/*
 * snapshots
 */

/*
 * A "snap context" is the set of existing snapshots when we
 * write data.  It is used by the OSD to guide its COW behavior.
 *
 * The ceph_snap_context is refcounted, and attached to each dirty
 * page, indicating which context the dirty data belonged when it was
 * dirtied.
 */
struct ceph_snap_context {
	atomic_t nref;
	u64 seq;
	u32 num_snaps;
	u64 snaps[];
};

extern struct ceph_snap_context *ceph_create_snap_context(u32 snap_count,
					gfp_t gfp_flags);
extern struct ceph_snap_context *ceph_get_snap_context(
					struct ceph_snap_context *sc);
extern void ceph_put_snap_context(struct ceph_snap_context *sc);

/*
 * calculate the number of pages a given length and offset map onto,
 * if we align the data.
 */
static inline int calc_pages_for(u64 off, u64 len)
{
	return ((off+len+PAGE_SIZE-1) >> PAGE_SHIFT) -
		(off >> PAGE_SHIFT);
}

/*
 * These are not meant to be generic - an integer key is assumed.
 */
#define DEFINE_RB_INSDEL_FUNCS(name, type, keyfld, nodefld)		\
static void insert_##name(struct rb_root *root, type *t)		\
{									\
	struct rb_node **n = &root->rb_node;				\
	struct rb_node *parent = NULL;					\
									\
	BUG_ON(!RB_EMPTY_NODE(&t->nodefld));				\
									\
	while (*n) {							\
		type *cur = rb_entry(*n, type, nodefld);		\
									\
		parent = *n;						\
		if (t->keyfld < cur->keyfld)				\
			n = &(*n)->rb_left;				\
		else if (t->keyfld > cur->keyfld)			\
			n = &(*n)->rb_right;				\
		else							\
			BUG();						\
	}								\
									\
	rb_link_node(&t->nodefld, parent, n);				\
	rb_insert_color(&t->nodefld, root);				\
}									\
static void erase_##name(struct rb_root *root, type *t)			\
{									\
	BUG_ON(RB_EMPTY_NODE(&t->nodefld));				\
	rb_erase(&t->nodefld, root);					\
	RB_CLEAR_NODE(&t->nodefld);					\
}

#define DEFINE_RB_LOOKUP_FUNC(name, type, keyfld, nodefld)		\
static type *lookup_##name(struct rb_root *root,			\
			   typeof(((type *)0)->keyfld) key)		\
{									\
	struct rb_node *n = root->rb_node;				\
									\
	while (n) {							\
		type *cur = rb_entry(n, type, nodefld);			\
									\
		if (key < cur->keyfld)					\
			n = n->rb_left;					\
		else if (key > cur->keyfld)				\
			n = n->rb_right;				\
		else							\
			return cur;					\
	}								\
									\
	return NULL;							\
}

#define DEFINE_RB_FUNCS(name, type, keyfld, nodefld)			\
DEFINE_RB_INSDEL_FUNCS(name, type, keyfld, nodefld)			\
DEFINE_RB_LOOKUP_FUNC(name, type, keyfld, nodefld)

extern struct kmem_cache *ceph_inode_cachep;
extern struct kmem_cache *ceph_cap_cachep;
extern struct kmem_cache *ceph_cap_flush_cachep;
extern struct kmem_cache *ceph_dentry_cachep;
extern struct kmem_cache *ceph_file_cachep;

/* ceph_common.c */
extern bool libceph_compatible(void *data);

extern const char *ceph_msg_type_name(int type);
extern int ceph_check_fsid(struct ceph_client *client, struct ceph_fsid *fsid);
extern void *ceph_kvmalloc(size_t size, gfp_t flags);

extern struct ceph_options *ceph_parse_options(char *options,
			      const char *dev_name, const char *dev_name_end,
			      int (*parse_extra_token)(char *c, void *private),
			      void *private);
int ceph_print_client_options(struct seq_file *m, struct ceph_client *client);
extern void ceph_destroy_options(struct ceph_options *opt);
extern int ceph_compare_options(struct ceph_options *new_opt,
				struct ceph_client *client);
extern struct ceph_client *ceph_create_client(struct ceph_options *opt,
					      void *private,
					      u64 supported_features,
					      u64 required_features);
extern u64 ceph_client_id(struct ceph_client *client);
extern void ceph_destroy_client(struct ceph_client *client);
extern int __ceph_open_session(struct ceph_client *client,
			       unsigned long started);
extern int ceph_open_session(struct ceph_client *client);

/* pagevec.c */
extern void ceph_release_page_vector(struct page **pages, int num_pages);

extern struct page **ceph_get_direct_page_vector(const void __user *data,
						 int num_pages,
						 bool write_page);
extern void ceph_put_page_vector(struct page **pages, int num_pages,
				 bool dirty);
extern struct page **ceph_alloc_page_vector(int num_pages, gfp_t flags);
extern int ceph_copy_user_to_page_vector(struct page **pages,
					 const void __user *data,
					 loff_t off, size_t len);
extern void ceph_copy_to_page_vector(struct page **pages,
				    const void *data,
				    loff_t off, size_t len);
extern void ceph_copy_from_page_vector(struct page **pages,
				    void *data,
				    loff_t off, size_t len);
extern void ceph_zero_page_vector_range(int off, int len, struct page **pages);


#endif /* _FS_CEPH_SUPER_H */
Commit	Line	Data
3d14c5d2 YS	1	#ifndef _FS_CEPH_LIBCEPH_H
	2	#define _FS_CEPH_LIBCEPH_H
	3
a1ce3928	4	#include <linux/ceph/ceph_debug.h>
3d14c5d2 YS	5
	6	#include <asm/unaligned.h>
	7	#include <linux/backing-dev.h>
	8	#include <linux/completion.h>
	9	#include <linux/exportfs.h>
187f1882	10	#include <linux/bug.h>
3d14c5d2 YS	11	#include <linux/fs.h>
	12	#include <linux/mempool.h>
	13	#include <linux/pagemap.h>
	14	#include <linux/wait.h>
	15	#include <linux/writeback.h>
	16	#include <linux/slab.h>
	17
a1ce3928 DH	18	#include <linux/ceph/types.h>
	19	#include <linux/ceph/messenger.h>
	20	#include <linux/ceph/msgpool.h>
	21	#include <linux/ceph/mon_client.h>
	22	#include <linux/ceph/osd_client.h>
	23	#include <linux/ceph/ceph_fs.h>
3d14c5d2	24
3d14c5d2 YS	25	/*
	26	* mount options
	27	*/
	28	#define CEPH_OPT_FSID (1<<0)
	29	#define CEPH_OPT_NOSHARE (1<<1) /* don't share client with other sbs */
	30	#define CEPH_OPT_MYIP (1<<2) /* specified my ip */
	31	#define CEPH_OPT_NOCRC (1<<3) /* no data crc on writes */
a51983e4	32	#define CEPH_OPT_NOMSGAUTH (1<<4) /* don't require msg signing feat */
ba988f87	33	#define CEPH_OPT_TCP_NODELAY (1<<5) /* TCP_NODELAY on TCP sockets */
a51983e4	34	#define CEPH_OPT_NOMSGSIGN (1<<6) /* don't sign msgs */
3d14c5d2	35
ba988f87	36	#define CEPH_OPT_DEFAULT (CEPH_OPT_TCP_NODELAY)
3d14c5d2 YS	37
	38	#define ceph_set_opt(client, opt) \
	39	(client)->options->flags \|= CEPH_OPT_##opt;
	40	#define ceph_test_opt(client, opt) \
	41	(!!((client)->options->flags & CEPH_OPT_##opt))
	42
	43	struct ceph_options {
	44	int flags;
	45	struct ceph_fsid fsid;
	46	struct ceph_entity_addr my_addr;
a319bf56 ID	47	unsigned long mount_timeout; /* jiffies */
	48	unsigned long osd_idle_ttl; /* jiffies */
	49	unsigned long osd_keepalive_timeout; /* jiffies */
3d14c5d2 YS	50
	51	/*
	52	* any type that can't be simply compared or doesn't need need
	53	* to be compared should go beyond this point,
	54	* ceph_compare_options() should be updated accordingly
	55	*/
	56
	57	struct ceph_entity_addr mon_addr; / should be the first
	58	pointer type of args */
	59	int num_mon;
	60	char *name;
8323c3aa	61	struct ceph_crypto_key *key;
3d14c5d2 YS	62	};
	63
	64	/*
	65	* defaults
	66	*/
a319bf56 ID	67	#define CEPH_MOUNT_TIMEOUT_DEFAULT msecs_to_jiffies(60 * 1000)
	68	#define CEPH_OSD_KEEPALIVE_DEFAULT msecs_to_jiffies(5 * 1000)
	69	#define CEPH_OSD_IDLE_TTL_DEFAULT msecs_to_jiffies(60 * 1000)
58d81b12	70
168b9090	71	#define CEPH_MONC_HUNT_INTERVAL msecs_to_jiffies(3 * 1000)
58d81b12 ID	72	#define CEPH_MONC_PING_INTERVAL msecs_to_jiffies(10 * 1000)
58d81b12 ID	73	#define CEPH_MONC_PING_TIMEOUT msecs_to_jiffies(30 * 1000)
168b9090 ID	74	#define CEPH_MONC_HUNT_BACKOFF 2
168b9090 ID	75	#define CEPH_MONC_HUNT_MAX_MULT 10
3d14c5d2 YS	76
3d14c5d2 YS	77	#define CEPH_MSG_MAX_FRONT_LEN (1610241024)
7b11ba37	78	#define CEPH_MSG_MAX_MIDDLE_LEN (1610241024)
3d14c5d2 YS	79	#define CEPH_MSG_MAX_DATA_LEN (1610241024)
	80
	81	#define CEPH_AUTH_NAME_DEFAULT "guest"
	82
	83	/*
	84	* Delay telling the MDS we no longer want caps, in case we reopen
	85	* the file. Delay a minimum amount of time, even if we send a cap
	86	* message for some other reason. Otherwise, take the oppotunity to
	87	* update the mds to avoid sending another message later.
	88	*/
	89	#define CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT 5 /* cap release delay */
	90	#define CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT 60 /* cap release delay */
	91
	92	#define CEPH_CAP_RELEASE_SAFETY_DEFAULT (CEPH_CAPS_PER_RELEASE * 4)
	93
	94	/* mount state */
	95	enum {
	96	CEPH_MOUNT_MOUNTING,
	97	CEPH_MOUNT_MOUNTED,
	98	CEPH_MOUNT_UNMOUNTING,
	99	CEPH_MOUNT_UNMOUNTED,
	100	CEPH_MOUNT_SHUTDOWN,
	101	};
	102
a319bf56 ID	103	static inline unsigned long ceph_timeout_jiffies(unsigned long timeout)
	104	{
	105	return timeout ?: MAX_SCHEDULE_TIMEOUT;
	106	}
	107
3d14c5d2 YS	108	struct ceph_mds_client;
	109
	110	/*
	111	* per client state
	112	*
	113	* possibly shared by multiple mount points, if they are
	114	* mounting the same ceph filesystem/cluster.
	115	*/
	116	struct ceph_client {
	117	struct ceph_fsid fsid;
	118	bool have_fsid;
	119
	120	void *private;
	121
	122	struct ceph_options *options;
	123
	124	struct mutex mount_mutex; /* serialize mount attempts */
	125	wait_queue_head_t auth_wq;
	126	int auth_err;
	127
	128	int (extra_mon_dispatch)(struct ceph_client , struct ceph_msg *);
	129
12b4629a ID	130	u64 supported_features;
12b4629a ID	131	u64 required_features;
3d14c5d2	132
15d9882c	133	struct ceph_messenger msgr; /* messenger instance */
3d14c5d2 YS	134	struct ceph_mon_client monc;
	135	struct ceph_osd_client osdc;
	136
	137	#ifdef CONFIG_DEBUG_FS
	138	struct dentry *debugfs_dir;
	139	struct dentry *debugfs_monmap;
	140	struct dentry *debugfs_osdmap;
5cf7bd30	141	struct dentry *debugfs_options;
3d14c5d2 YS	142	#endif
	143	};
	144
859bff51	145	#define from_msgr(ms) container_of(ms, struct ceph_client, msgr)
3d14c5d2 YS	146
	147
	148	/*
	149	* snapshots
	150	*/
	151
	152	/*
	153	* A "snap context" is the set of existing snapshots when we
	154	* write data. It is used by the OSD to guide its COW behavior.
	155	*
	156	* The ceph_snap_context is refcounted, and attached to each dirty
	157	* page, indicating which context the dirty data belonged when it was
	158	* dirtied.
	159	*/
	160	struct ceph_snap_context {
	161	atomic_t nref;
	162	u64 seq;
aa711ee3	163	u32 num_snaps;
3d14c5d2 YS	164	u64 snaps[];
	165	};
	166
4f0dcb10 AE	167	extern struct ceph_snap_context *ceph_create_snap_context(u32 snap_count,
	168	gfp_t gfp_flags);
	169	extern struct ceph_snap_context *ceph_get_snap_context(
	170	struct ceph_snap_context *sc);
	171	extern void ceph_put_snap_context(struct ceph_snap_context *sc);
3d14c5d2 YS	172
	173	/*
	174	* calculate the number of pages a given length and offset map onto,
	175	* if we align the data.
	176	*/
	177	static inline int calc_pages_for(u64 off, u64 len)
	178	{
09cbfeaf KS	179	return ((off+len+PAGE_SIZE-1) >> PAGE_SHIFT) -
09cbfeaf KS	180	(off >> PAGE_SHIFT);
3d14c5d2 YS	181	}
3d14c5d2 YS	182
fcd00b68 ID	183	/*
	184	* These are not meant to be generic - an integer key is assumed.
	185	*/
	186	#define DEFINE_RB_INSDEL_FUNCS(name, type, keyfld, nodefld) \
	187	static void insert_##name(struct rb_root root, type t) \
	188	{ \
	189	struct rb_node **n = &root->rb_node; \
	190	struct rb_node *parent = NULL; \
	191	\
	192	BUG_ON(!RB_EMPTY_NODE(&t->nodefld)); \
	193	\
	194	while (*n) { \
	195	type cur = rb_entry(n, type, nodefld); \
	196	\
	197	parent = *n; \
	198	if (t->keyfld < cur->keyfld) \
	199	n = &(*n)->rb_left; \
	200	else if (t->keyfld > cur->keyfld) \
	201	n = &(*n)->rb_right; \
	202	else \
	203	BUG(); \
	204	} \
	205	\
	206	rb_link_node(&t->nodefld, parent, n); \
	207	rb_insert_color(&t->nodefld, root); \
	208	} \
	209	static void erase_##name(struct rb_root root, type t) \
	210	{ \
	211	BUG_ON(RB_EMPTY_NODE(&t->nodefld)); \
	212	rb_erase(&t->nodefld, root); \
	213	RB_CLEAR_NODE(&t->nodefld); \
	214	}
	215
	216	#define DEFINE_RB_LOOKUP_FUNC(name, type, keyfld, nodefld) \
	217	static type lookup_##name(struct rb_root root, \
	218	typeof(((type *)0)->keyfld) key) \
	219	{ \
	220	struct rb_node *n = root->rb_node; \
	221	\
	222	while (n) { \
	223	type *cur = rb_entry(n, type, nodefld); \
	224	\
	225	if (key < cur->keyfld) \
	226	n = n->rb_left; \
	227	else if (key > cur->keyfld) \
	228	n = n->rb_right; \
	229	else \
	230	return cur; \
	231	} \
	232	\
	233	return NULL; \
	234	}
	235
	236	#define DEFINE_RB_FUNCS(name, type, keyfld, nodefld) \
	237	DEFINE_RB_INSDEL_FUNCS(name, type, keyfld, nodefld) \
	238	DEFINE_RB_LOOKUP_FUNC(name, type, keyfld, nodefld)
	239
eeb0bed5 ID	240	extern struct kmem_cache *ceph_inode_cachep;
eeb0bed5 ID	241	extern struct kmem_cache *ceph_cap_cachep;
f66fd9f0	242	extern struct kmem_cache *ceph_cap_flush_cachep;
eeb0bed5 ID	243	extern struct kmem_cache *ceph_dentry_cachep;
	244	extern struct kmem_cache *ceph_file_cachep;
	245
3d14c5d2	246	/* ceph_common.c */
72fe25e3 AE	247	extern bool libceph_compatible(void *data);
72fe25e3 AE	248
3d14c5d2 YS	249	extern const char *ceph_msg_type_name(int type);
3d14c5d2 YS	250	extern int ceph_check_fsid(struct ceph_client client, struct ceph_fsid fsid);
eeb0bed5	251	extern void *ceph_kvmalloc(size_t size, gfp_t flags);
3d14c5d2	252
ee57741c	253	extern struct ceph_options ceph_parse_options(char options,
3d14c5d2 YS	254	const char dev_name, const char dev_name_end,
	255	int (parse_extra_token)(char c, void *private),
	256	void *private);
ff40f9ae	257	int ceph_print_client_options(struct seq_file m, struct ceph_client client);
3d14c5d2 YS	258	extern void ceph_destroy_options(struct ceph_options *opt);
	259	extern int ceph_compare_options(struct ceph_options *new_opt,
	260	struct ceph_client *client);
	261	extern struct ceph_client ceph_create_client(struct ceph_options opt,
6ab00d46	262	void *private,
12b4629a ID	263	u64 supported_features,
12b4629a ID	264	u64 required_features);
3d14c5d2 YS	265	extern u64 ceph_client_id(struct ceph_client *client);
	266	extern void ceph_destroy_client(struct ceph_client *client);
	267	extern int __ceph_open_session(struct ceph_client *client,
	268	unsigned long started);
	269	extern int ceph_open_session(struct ceph_client *client);
	270
	271	/* pagevec.c */
	272	extern void ceph_release_page_vector(struct page **pages, int num_pages);
	273
b324814e	274	extern struct page *ceph_get_direct_page_vector(const void __user data,
b6aa5901 HC	275	int num_pages,
	276	bool write_page);
	277	extern void ceph_put_page_vector(struct page **pages, int num_pages,
	278	bool dirty);
3d14c5d2 YS	279	extern struct page **ceph_alloc_page_vector(int num_pages, gfp_t flags);
3d14c5d2 YS	280	extern int ceph_copy_user_to_page_vector(struct page **pages,
b324814e	281	const void __user *data,
3d14c5d2	282	loff_t off, size_t len);
903bb32e	283	extern void ceph_copy_to_page_vector(struct page **pages,
b324814e	284	const void *data,
3d14c5d2	285	loff_t off, size_t len);
903bb32e	286	extern void ceph_copy_from_page_vector(struct page **pages,
b324814e	287	void *data,
3d14c5d2	288	loff_t off, size_t len);
3d14c5d2 YS	289	extern void ceph_zero_page_vector_range(int off, int len, struct page **pages);
	290
	291
	292	#endif /* _FS_CEPH_SUPER_H */