libceph.h 9.4 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
2 3 4
#ifndef _FS_CEPH_LIBCEPH_H
#define _FS_CEPH_LIBCEPH_H

5
#include <linux/ceph/ceph_debug.h>
6 7 8 9 10

#include <asm/unaligned.h>
#include <linux/backing-dev.h>
#include <linux/completion.h>
#include <linux/exportfs.h>
11
#include <linux/bug.h>
12 13 14 15 16 17
#include <linux/fs.h>
#include <linux/mempool.h>
#include <linux/pagemap.h>
#include <linux/wait.h>
#include <linux/writeback.h>
#include <linux/slab.h>
18
#include <linux/refcount.h>
19

20 21 22 23 24 25
#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>
26
#include <linux/ceph/string_table.h>
27 28 29 30 31 32 33 34

/*
 * 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 */
35
#define CEPH_OPT_NOMSGAUTH	  (1<<4) /* don't require msg signing feat */
C
Chaitanya Huilgol 已提交
36
#define CEPH_OPT_TCP_NODELAY	  (1<<5) /* TCP_NODELAY on TCP sockets */
37
#define CEPH_OPT_NOMSGSIGN	  (1<<6) /* don't sign msgs */
38

C
Chaitanya Huilgol 已提交
39
#define CEPH_OPT_DEFAULT   (CEPH_OPT_TCP_NODELAY)
40 41 42 43 44 45 46 47 48 49

#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;
50 51 52
	unsigned long mount_timeout;		/* jiffies */
	unsigned long osd_idle_ttl;		/* jiffies */
	unsigned long osd_keepalive_timeout;	/* jiffies */
53
	unsigned long osd_request_timeout;	/* jiffies */
54 55 56 57 58 59 60 61 62 63 64

	/*
	 * 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;
65
	struct ceph_crypto_key *key;
66 67 68 69 70
};

/*
 * defaults
 */
71 72 73
#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)
74
#define CEPH_OSD_REQUEST_TIMEOUT_DEFAULT 0  /* no timeout */
I
Ilya Dryomov 已提交
75

76
#define CEPH_MONC_HUNT_INTERVAL		msecs_to_jiffies(3 * 1000)
I
Ilya Dryomov 已提交
77 78
#define CEPH_MONC_PING_INTERVAL		msecs_to_jiffies(10 * 1000)
#define CEPH_MONC_PING_TIMEOUT		msecs_to_jiffies(30 * 1000)
79 80
#define CEPH_MONC_HUNT_BACKOFF		2
#define CEPH_MONC_HUNT_MAX_MULT		10
81 82

#define CEPH_MSG_MAX_FRONT_LEN	(16*1024*1024)
83
#define CEPH_MSG_MAX_MIDDLE_LEN	(16*1024*1024)
I
Ilya Dryomov 已提交
84 85 86 87 88 89 90

/*
 * Handle the largest possible rbd object in one message.
 * There is no limit on the size of cephfs objects, but it has to obey
 * rsize and wsize mount options anyway.
 */
#define CEPH_MSG_MAX_DATA_LEN	(32*1024*1024)
91 92 93 94 95 96 97 98 99 100 101 102

#define CEPH_AUTH_NAME_DEFAULT   "guest"

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

103 104 105 106 107
static inline unsigned long ceph_timeout_jiffies(unsigned long timeout)
{
	return timeout ?: MAX_SCHEDULE_TIMEOUT;
}

108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
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 *);

130 131
	u64 supported_features;
	u64 required_features;
132

133
	struct ceph_messenger msgr;   /* messenger instance */
134 135 136 137 138 139 140
	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;
141
	struct dentry *debugfs_options;
142 143 144
#endif
};

145
#define from_msgr(ms)	container_of(ms, struct ceph_client, msgr)
146 147 148 149 150 151 152 153 154 155 156 157 158 159 160


/*
 * 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 {
161
	refcount_t nref;
162
	u64 seq;
163
	u32 num_snaps;
164 165 166
	u64 snaps[];
};

167 168 169 170 171
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);
172 173 174 175 176 177 178

/*
 * 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)
{
179 180
	return ((off+len+PAGE_SIZE-1) >> PAGE_SHIFT) -
		(off >> PAGE_SHIFT);
181 182
}

183 184 185 186 187
#define RB_BYVAL(a)      (a)
#define RB_BYPTR(a)      (&(a))
#define RB_CMP3WAY(a, b) ((a) < (b) ? -1 : (a) > (b))

#define DEFINE_RB_INSDEL_FUNCS2(name, type, keyfld, cmpexp, keyexp, nodefld) \
I
Ilya Dryomov 已提交
188 189 190 191 192 193 194 195 196
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);		\
197
		int cmp;						\
I
Ilya Dryomov 已提交
198 199
									\
		parent = *n;						\
200 201
		cmp = cmpexp(keyexp(t->keyfld), keyexp(cur->keyfld));	\
		if (cmp < 0)						\
I
Ilya Dryomov 已提交
202
			n = &(*n)->rb_left;				\
203
		else if (cmp > 0)					\
I
Ilya Dryomov 已提交
204 205 206 207 208 209 210 211 212 213 214 215 216 217 218
			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);					\
}

219 220 221 222 223 224 225
/*
 * @lookup_param_type is a parameter and not constructed from (@type,
 * @keyfld) with typeof() because adding const is too unwieldy.
 */
#define DEFINE_RB_LOOKUP_FUNC2(name, type, keyfld, cmpexp, keyexp,	\
			       lookup_param_type, nodefld)		\
static type *lookup_##name(struct rb_root *root, lookup_param_type key)	\
I
Ilya Dryomov 已提交
226 227 228 229 230
{									\
	struct rb_node *n = root->rb_node;				\
									\
	while (n) {							\
		type *cur = rb_entry(n, type, nodefld);			\
231
		int cmp;						\
I
Ilya Dryomov 已提交
232
									\
233 234
		cmp = cmpexp(key, keyexp(cur->keyfld));			\
		if (cmp < 0)						\
I
Ilya Dryomov 已提交
235
			n = n->rb_left;					\
236
		else if (cmp > 0)					\
I
Ilya Dryomov 已提交
237 238 239 240 241 242 243 244
			n = n->rb_right;				\
		else							\
			return cur;					\
	}								\
									\
	return NULL;							\
}

245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261
#define DEFINE_RB_FUNCS2(name, type, keyfld, cmpexp, keyexp,		\
			 lookup_param_type, nodefld)			\
DEFINE_RB_INSDEL_FUNCS2(name, type, keyfld, cmpexp, keyexp, nodefld)	\
DEFINE_RB_LOOKUP_FUNC2(name, type, keyfld, cmpexp, keyexp,		\
		       lookup_param_type, nodefld)

/*
 * Shorthands for integer keys.
 */
#define DEFINE_RB_INSDEL_FUNCS(name, type, keyfld, nodefld)		\
DEFINE_RB_INSDEL_FUNCS2(name, type, keyfld, RB_CMP3WAY, RB_BYVAL, nodefld)

#define DEFINE_RB_LOOKUP_FUNC(name, type, keyfld, nodefld)		\
extern type __lookup_##name##_key;					\
DEFINE_RB_LOOKUP_FUNC2(name, type, keyfld, RB_CMP3WAY, RB_BYVAL,	\
		       typeof(__lookup_##name##_key.keyfld), nodefld)

I
Ilya Dryomov 已提交
262 263 264 265
#define DEFINE_RB_FUNCS(name, type, keyfld, nodefld)			\
DEFINE_RB_INSDEL_FUNCS(name, type, keyfld, nodefld)			\
DEFINE_RB_LOOKUP_FUNC(name, type, keyfld, nodefld)

266 267
extern struct kmem_cache *ceph_inode_cachep;
extern struct kmem_cache *ceph_cap_cachep;
268
extern struct kmem_cache *ceph_cap_flush_cachep;
269 270
extern struct kmem_cache *ceph_dentry_cachep;
extern struct kmem_cache *ceph_file_cachep;
C
Chengguang Xu 已提交
271
extern struct kmem_cache *ceph_dir_file_cachep;
272

273
/* ceph_common.c */
274 275
extern bool libceph_compatible(void *data);

276 277
extern const char *ceph_msg_type_name(int type);
extern int ceph_check_fsid(struct ceph_client *client, struct ceph_fsid *fsid);
278
extern void *ceph_kvmalloc(size_t size, gfp_t flags);
279

280
extern struct ceph_options *ceph_parse_options(char *options,
281 282 283
			      const char *dev_name, const char *dev_name_end,
			      int (*parse_extra_token)(char *c, void *private),
			      void *private);
284
int ceph_print_client_options(struct seq_file *m, struct ceph_client *client);
285 286 287
extern void ceph_destroy_options(struct ceph_options *opt);
extern int ceph_compare_options(struct ceph_options *new_opt,
				struct ceph_client *client);
288
struct ceph_client *ceph_create_client(struct ceph_options *opt, void *private);
289
struct ceph_entity_addr *ceph_client_addr(struct ceph_client *client);
290
u64 ceph_client_gid(struct ceph_client *client);
291 292 293 294 295 296 297 298
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);

299
extern struct page **ceph_get_direct_page_vector(const void __user *data,
300 301 302 303
						 int num_pages,
						 bool write_page);
extern void ceph_put_page_vector(struct page **pages, int num_pages,
				 bool dirty);
304 305
extern struct page **ceph_alloc_page_vector(int num_pages, gfp_t flags);
extern int ceph_copy_user_to_page_vector(struct page **pages,
306
					 const void __user *data,
307
					 loff_t off, size_t len);
308
extern void ceph_copy_to_page_vector(struct page **pages,
309
				    const void *data,
310
				    loff_t off, size_t len);
311
extern void ceph_copy_from_page_vector(struct page **pages,
312
				    void *data,
313 314 315 316 317
				    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 */