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

#ifndef _FS_CEPH_OSDMAP_H
#define _FS_CEPH_OSDMAP_H

#include <linux/rbtree.h>
#include <linux/ceph/types.h>
#include <linux/ceph/decode.h>
#include <linux/ceph/ceph_fs.h>
#include <linux/crush/crush.h>

/*
 * The osd map describes the current membership of the osd cluster and
 * specifies the mapping of objects to placement groups and placement
 * groups to (sets of) osds.  That is, it completely specifies the
 * (desired) distribution of all data objects in the system at some
 * point in time.
 *
 * Each map version is identified by an epoch, which increases monotonically.
 *
 * The map can be updated either via an incremental map (diff) describing
 * the change between two successive epochs, or as a fully encoded map.
 */
struct ceph_pg {
	uint64_t pool;
	uint32_t seed;
};

#define CEPH_POOL_FLAG_HASHPSPOOL  1

struct ceph_pg_pool_info {
	struct rb_node node;
	s64 id;
	u8 type;
	u8 size;
	u8 crush_ruleset;
	u8 object_hash;
	u32 pg_num, pgp_num;
	int pg_num_mask, pgp_num_mask;
	s64 read_tier;
	s64 write_tier; /* wins for read+write ops */
	u64 flags;
	char *name;
};

static inline bool ceph_can_shift_osds(struct ceph_pg_pool_info *pool)
{
	switch (pool->type) {
	case CEPH_POOL_TYPE_REP:
		return true;
	case CEPH_POOL_TYPE_EC:
		return false;
	default:
		BUG_ON(1);
	}
}

struct ceph_object_locator {
	s64 pool;
};

/*
 * Maximum supported by kernel client object name length
 *
 * (probably outdated: must be >= RBD_MAX_MD_NAME_LEN -- currently 100)
 */
#define CEPH_MAX_OID_NAME_LEN 100

/*
 * 51-char inline_name is long enough for all cephfs and all but one
 * rbd requests: <imgname> in "<imgname>.rbd"/"rbd_id.<imgname>" can be
 * arbitrarily long (~PAGE_SIZE).  It's done once during rbd map; all
 * other rbd requests fit into inline_name.
 *
 * Makes ceph_object_id 64 bytes on 64-bit.
 */
#define CEPH_OID_INLINE_LEN 52

/*
 * Both inline and external buffers have space for a NUL-terminator,
 * which is carried around.  It's not required though - RADOS object
 * names don't have to be NUL-terminated and may contain NULs.
 */
struct ceph_object_id {
	char *name;
	char inline_name[CEPH_OID_INLINE_LEN];
	int name_len;
};

static inline void ceph_oid_init(struct ceph_object_id *oid)
{
	oid->name = oid->inline_name;
	oid->name_len = 0;
}

static inline bool ceph_oid_empty(const struct ceph_object_id *oid)
{
	return oid->name == oid->inline_name && !oid->name_len;
}

void ceph_oid_copy(struct ceph_object_id *dest,
		   const struct ceph_object_id *src);
__printf(2, 3)
void ceph_oid_printf(struct ceph_object_id *oid, const char *fmt, ...);
__printf(3, 4)
int ceph_oid_aprintf(struct ceph_object_id *oid, gfp_t gfp,
		     const char *fmt, ...);
void ceph_oid_destroy(struct ceph_object_id *oid);

struct ceph_pg_mapping {
	struct rb_node node;
	struct ceph_pg pgid;

	union {
		struct {
			int len;
			int osds[];
		} pg_temp;
		struct {
			int osd;
		} primary_temp;
	};
};

struct ceph_osdmap {
	struct ceph_fsid fsid;
	u32 epoch;
	u32 mkfs_epoch;
	struct ceph_timespec created, modified;

	u32 flags;         /* CEPH_OSDMAP_* */

	u32 max_osd;       /* size of osd_state, _offload, _addr arrays */
	u8 *osd_state;     /* CEPH_OSD_* */
	u32 *osd_weight;   /* 0 = failed, 0x10000 = 100% normal */
	struct ceph_entity_addr *osd_addr;

	struct rb_root pg_temp;
	struct rb_root primary_temp;

	u32 *osd_primary_affinity;

	struct rb_root pg_pools;
	u32 pool_max;

	/* the CRUSH map specifies the mapping of placement groups to
	 * the list of osds that store+replicate them. */
	struct crush_map *crush;

	struct mutex crush_scratch_mutex;
	int crush_scratch_ary[CEPH_PG_MAX_SIZE * 3];
};

static inline int ceph_osd_exists(struct ceph_osdmap *map, int osd)
{
	return osd >= 0 && osd < map->max_osd &&
	       (map->osd_state[osd] & CEPH_OSD_EXISTS);
}

static inline int ceph_osd_is_up(struct ceph_osdmap *map, int osd)
{
	return ceph_osd_exists(map, osd) &&
	       (map->osd_state[osd] & CEPH_OSD_UP);
}

static inline int ceph_osd_is_down(struct ceph_osdmap *map, int osd)
{
	return !ceph_osd_is_up(map, osd);
}

static inline bool ceph_osdmap_flag(struct ceph_osdmap *map, int flag)
{
	return map && (map->flags & flag);
}

extern char *ceph_osdmap_state_str(char *str, int len, int state);
extern u32 ceph_get_primary_affinity(struct ceph_osdmap *map, int osd);

static inline struct ceph_entity_addr *ceph_osd_addr(struct ceph_osdmap *map,
						     int osd)
{
	if (osd >= map->max_osd)
		return NULL;
	return &map->osd_addr[osd];
}

static inline int ceph_decode_pgid(void **p, void *end, struct ceph_pg *pgid)
{
	__u8 version;

	if (!ceph_has_room(p, end, 1 + 8 + 4 + 4)) {
		pr_warn("incomplete pg encoding\n");
		return -EINVAL;
	}
	version = ceph_decode_8(p);
	if (version > 1) {
		pr_warn("do not understand pg encoding %d > 1\n",
			(int)version);
		return -EINVAL;
	}

	pgid->pool = ceph_decode_64(p);
	pgid->seed = ceph_decode_32(p);
	*p += 4;	/* skip deprecated preferred value */

	return 0;
}

extern struct ceph_osdmap *ceph_osdmap_decode(void **p, void *end);
extern struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
					    struct ceph_osdmap *map,
					    struct ceph_messenger *msgr);
extern void ceph_osdmap_destroy(struct ceph_osdmap *map);

/* calculate mapping of a file extent to an object */
extern int ceph_calc_file_object_mapping(struct ceph_file_layout *layout,
					 u64 off, u64 len,
					 u64 *bno, u64 *oxoff, u64 *oxlen);

/* calculate mapping of object to a placement group */
extern int ceph_oloc_oid_to_pg(struct ceph_osdmap *osdmap,
			       struct ceph_object_locator *oloc,
			       struct ceph_object_id *oid,
			       struct ceph_pg *pg_out);

extern int ceph_calc_pg_acting(struct ceph_osdmap *osdmap,
			       struct ceph_pg pgid,
			       int *osds, int *primary);
extern int ceph_calc_pg_primary(struct ceph_osdmap *osdmap,
				struct ceph_pg pgid);

extern struct ceph_pg_pool_info *ceph_pg_pool_by_id(struct ceph_osdmap *map,
						    u64 id);

extern const char *ceph_pg_pool_name_by_id(struct ceph_osdmap *map, u64 id);
extern int ceph_pg_poolid_by_name(struct ceph_osdmap *map, const char *name);

#endif
Commit	Line	Data
f24e9980 SW	1	#ifndef _FS_CEPH_OSDMAP_H
	2	#define _FS_CEPH_OSDMAP_H
	3
	4	#include <linux/rbtree.h>
a1ce3928	5	#include <linux/ceph/types.h>
ef4859d6	6	#include <linux/ceph/decode.h>
a1ce3928	7	#include <linux/ceph/ceph_fs.h>
3d14c5d2	8	#include <linux/crush/crush.h>
f24e9980 SW	9
	10	/*
	11	* The osd map describes the current membership of the osd cluster and
	12	* specifies the mapping of objects to placement groups and placement
	13	* groups to (sets of) osds. That is, it completely specifies the
	14	* (desired) distribution of all data objects in the system at some
	15	* point in time.
	16	*
	17	* Each map version is identified by an epoch, which increases monotonically.
	18	*
	19	* The map can be updated either via an incremental map (diff) describing
	20	* the change between two successive epochs, or as a fully encoded map.
	21	*/
5b191d99 SW	22	struct ceph_pg {
	23	uint64_t pool;
	24	uint32_t seed;
	25	};
	26
83ca14fd SW	27	#define CEPH_POOL_FLAG_HASHPSPOOL 1
83ca14fd SW	28
f24e9980	29	struct ceph_pg_pool_info {
4fc51be8	30	struct rb_node node;
4f6a7e5e SW	31	s64 id;
	32	u8 type;
	33	u8 size;
	34	u8 crush_ruleset;
	35	u8 object_hash;
	36	u32 pg_num, pgp_num;
	37	int pg_num_mask, pgp_num_mask;
17a13e40 ID	38	s64 read_tier;
17a13e40 ID	39	s64 write_tier; /* wins for read+write ops */
4f6a7e5e	40	u64 flags;
2844a76a	41	char *name;
f24e9980 SW	42	};
f24e9980 SW	43
2abebdbc ID	44	static inline bool ceph_can_shift_osds(struct ceph_pg_pool_info *pool)
	45	{
	46	switch (pool->type) {
	47	case CEPH_POOL_TYPE_REP:
	48	return true;
	49	case CEPH_POOL_TYPE_EC:
	50	return false;
	51	default:
	52	BUG_ON(1);
	53	}
	54	}
	55
4f6a7e5e	56	struct ceph_object_locator {
22116525	57	s64 pool;
4f6a7e5e SW	58	};
4f6a7e5e SW	59
4295f221 ID	60	/*
	61	* Maximum supported by kernel client object name length
	62	*
	63	* (probably outdated: must be >= RBD_MAX_MD_NAME_LEN -- currently 100)
	64	*/
	65	#define CEPH_MAX_OID_NAME_LEN 100
	66
d30291b9 ID	67	/*
	68	* 51-char inline_name is long enough for all cephfs and all but one
	69	* rbd requests: <imgname> in "<imgname>.rbd"/"rbd_id.<imgname>" can be
	70	* arbitrarily long (~PAGE_SIZE). It's done once during rbd map; all
	71	* other rbd requests fit into inline_name.
	72	*
	73	* Makes ceph_object_id 64 bytes on 64-bit.
	74	*/
	75	#define CEPH_OID_INLINE_LEN 52
	76
	77	/*
	78	* Both inline and external buffers have space for a NUL-terminator,
	79	* which is carried around. It's not required though - RADOS object
	80	* names don't have to be NUL-terminated and may contain NULs.
	81	*/
4295f221	82	struct ceph_object_id {
d30291b9 ID	83	char *name;
d30291b9 ID	84	char inline_name[CEPH_OID_INLINE_LEN];
4295f221 ID	85	int name_len;
	86	};
	87
d30291b9 ID	88	static inline void ceph_oid_init(struct ceph_object_id *oid)
	89	{
	90	oid->name = oid->inline_name;
	91	oid->name_len = 0;
	92	}
	93
	94	static inline bool ceph_oid_empty(const struct ceph_object_id *oid)
	95	{
	96	return oid->name == oid->inline_name && !oid->name_len;
	97	}
	98
	99	void ceph_oid_copy(struct ceph_object_id *dest,
	100	const struct ceph_object_id *src);
	101	__printf(2, 3)
	102	void ceph_oid_printf(struct ceph_object_id oid, const char fmt, ...);
	103	__printf(3, 4)
	104	int ceph_oid_aprintf(struct ceph_object_id *oid, gfp_t gfp,
	105	const char *fmt, ...);
	106	void ceph_oid_destroy(struct ceph_object_id *oid);
	107
f24e9980 SW	108	struct ceph_pg_mapping {
f24e9980 SW	109	struct rb_node node;
5b191d99	110	struct ceph_pg pgid;
35a935d7 ID	111
	112	union {
	113	struct {
	114	int len;
	115	int osds[];
	116	} pg_temp;
9686f94c ID	117	struct {
	118	int osd;
	119	} primary_temp;
35a935d7	120	};
f24e9980 SW	121	};
	122
	123	struct ceph_osdmap {
	124	struct ceph_fsid fsid;
	125	u32 epoch;
	126	u32 mkfs_epoch;
	127	struct ceph_timespec created, modified;
	128
	129	u32 flags; /* CEPH_OSDMAP_* */
	130
	131	u32 max_osd; /* size of osd_state, _offload, _addr arrays */
	132	u8 osd_state; / CEPH_OSD_* */
	133	u32 osd_weight; / 0 = failed, 0x10000 = 100% normal */
	134	struct ceph_entity_addr *osd_addr;
	135
	136	struct rb_root pg_temp;
9686f94c ID	137	struct rb_root primary_temp;
9686f94c ID	138
2cfa34f2 ID	139	u32 *osd_primary_affinity;
2cfa34f2 ID	140
4fc51be8 SW	141	struct rb_root pg_pools;
4fc51be8 SW	142	u32 pool_max;
f24e9980 SW	143
	144	/* the CRUSH map specifies the mapping of placement groups to
	145	* the list of osds that store+replicate them. */
	146	struct crush_map *crush;
9d521470 ID	147
	148	struct mutex crush_scratch_mutex;
	149	int crush_scratch_ary[CEPH_PG_MAX_SIZE * 3];
f24e9980 SW	150	};
f24e9980 SW	151
246138fa ID	152	static inline int ceph_osd_exists(struct ceph_osdmap *map, int osd)
	153	{
	154	return osd >= 0 && osd < map->max_osd &&
	155	(map->osd_state[osd] & CEPH_OSD_EXISTS);
	156	}
	157
f24e9980 SW	158	static inline int ceph_osd_is_up(struct ceph_osdmap *map, int osd)
f24e9980 SW	159	{
246138fa ID	160	return ceph_osd_exists(map, osd) &&
	161	(map->osd_state[osd] & CEPH_OSD_UP);
	162	}
	163
	164	static inline int ceph_osd_is_down(struct ceph_osdmap *map, int osd)
	165	{
	166	return !ceph_osd_is_up(map, osd);
f24e9980 SW	167	}
	168
	169	static inline bool ceph_osdmap_flag(struct ceph_osdmap *map, int flag)
	170	{
	171	return map && (map->flags & flag);
	172	}
	173
	174	extern char ceph_osdmap_state_str(char str, int len, int state);
2cfa34f2	175	extern u32 ceph_get_primary_affinity(struct ceph_osdmap *map, int osd);
f24e9980 SW	176
	177	static inline struct ceph_entity_addr ceph_osd_addr(struct ceph_osdmap map,
	178	int osd)
	179	{
	180	if (osd >= map->max_osd)
	181	return NULL;
	182	return &map->osd_addr[osd];
	183	}
	184
ef4859d6 AE	185	static inline int ceph_decode_pgid(void *p, void end, struct ceph_pg *pgid)
	186	{
	187	__u8 version;
	188
	189	if (!ceph_has_room(p, end, 1 + 8 + 4 + 4)) {
3ef650d3	190	pr_warn("incomplete pg encoding\n");
ef4859d6 AE	191	return -EINVAL;
	192	}
	193	version = ceph_decode_8(p);
	194	if (version > 1) {
3ef650d3	195	pr_warn("do not understand pg encoding %d > 1\n",
ef4859d6 AE	196	(int)version);
	197	return -EINVAL;
	198	}
	199
	200	pgid->pool = ceph_decode_64(p);
	201	pgid->seed = ceph_decode_32(p);
	202	p += 4; / skip deprecated preferred value */
	203
	204	return 0;
	205	}
	206
a2505d63	207	extern struct ceph_osdmap ceph_osdmap_decode(void p, void end);
f24e9980 SW	208	extern struct ceph_osdmap osdmap_apply_incremental(void p, void end,
	209	struct ceph_osdmap *map,
	210	struct ceph_messenger *msgr);
	211	extern void ceph_osdmap_destroy(struct ceph_osdmap *map);
	212
	213	/* calculate mapping of a file extent to an object */
d63b77f4	214	extern int ceph_calc_file_object_mapping(struct ceph_file_layout *layout,
e8afad65	215	u64 off, u64 len,
d63b77f4	216	u64 bno, u64 oxoff, u64 *oxlen);
f24e9980 SW	217
f24e9980 SW	218	/* calculate mapping of object to a placement group */
7c13cb64 ID	219	extern int ceph_oloc_oid_to_pg(struct ceph_osdmap *osdmap,
	220	struct ceph_object_locator *oloc,
	221	struct ceph_object_id *oid,
	222	struct ceph_pg *pg_out);
	223
12979354	224	extern int ceph_calc_pg_acting(struct ceph_osdmap *osdmap,
5b191d99	225	struct ceph_pg pgid,
8008ab10	226	int osds, int primary);
51042122	227	extern int ceph_calc_pg_primary(struct ceph_osdmap *osdmap,
5b191d99	228	struct ceph_pg pgid);
f24e9980	229
ce7f6a27 ID	230	extern struct ceph_pg_pool_info ceph_pg_pool_by_id(struct ceph_osdmap map,
	231	u64 id);
	232
72afc71f	233	extern const char ceph_pg_pool_name_by_id(struct ceph_osdmap map, u64 id);
7669a2c9 YS	234	extern int ceph_pg_poolid_by_name(struct ceph_osdmap map, const char name);
7669a2c9 YS	235
f24e9980	236	#endif