volumes.h 16.4 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
/*
 * Copyright (C) 2007 Oracle.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#ifndef __BTRFS_VOLUMES_
#define __BTRFS_VOLUMES_
21

22
#include <linux/bio.h>
23
#include <linux/sort.h>
24
#include <linux/btrfs.h>
25
#include "async-thread.h"
26

27 28
extern struct mutex uuid_mutex;

29
#define BTRFS_STRIPE_LEN	SZ_64K
30

31
struct buffer_head;
32 33 34 35 36
struct btrfs_pending_bios {
	struct bio *head;
	struct bio *tail;
};

37 38 39 40 41 42 43 44 45 46 47 48 49
/*
 * Use sequence counter to get consistent device stat data on
 * 32-bit processors.
 */
#if BITS_PER_LONG==32 && defined(CONFIG_SMP)
#include <linux/seqlock.h>
#define __BTRFS_NEED_DEVICE_DATA_ORDERED
#define btrfs_device_data_ordered_init(device)	\
	seqcount_init(&device->data_seqcount)
#else
#define btrfs_device_data_ordered_init(device) do { } while (0)
#endif

50
#define BTRFS_DEV_STATE_WRITEABLE	(0)
51
#define BTRFS_DEV_STATE_IN_FS_METADATA	(1)
52
#define BTRFS_DEV_STATE_MISSING		(2)
53

54 55
struct btrfs_device {
	struct list_head dev_list;
56
	struct list_head dev_alloc_list;
Y
Yan Zheng 已提交
57
	struct btrfs_fs_devices *fs_devices;
58
	struct btrfs_fs_info *fs_info;
59

60 61 62 63 64 65
	struct rcu_string *name;

	u64 generation;

	spinlock_t io_lock ____cacheline_aligned;
	int running_pending;
66 67
	/* regular prio bios */
	struct btrfs_pending_bios pending_bios;
68
	/* sync bios */
69 70
	struct btrfs_pending_bios pending_sync_bios;

71 72 73 74 75
	struct block_device *bdev;

	/* the mode sent to blkdev_get */
	fmode_t mode;

76
	unsigned long dev_state;
77
	int is_tgtdev_for_dev_replace;
78
	blk_status_t last_flush_error;
79
	int flush_bio_sent;
80

81 82 83 84
#ifdef __BTRFS_NEED_DEVICE_DATA_ORDERED
	seqcount_t data_seqcount;
#endif

85 86 87
	/* the internal btrfs device id */
	u64 devid;

88
	/* size of the device in memory */
89 90
	u64 total_bytes;

91
	/* size of the device on disk */
92 93
	u64 disk_total_bytes;

94 95 96 97 98 99 100 101
	/* bytes used */
	u64 bytes_used;

	/* optimal io alignment for this device */
	u32 io_align;

	/* optimal io width for this device */
	u32 io_width;
102 103
	/* type and info about this device */
	u64 type;
104 105 106 107 108

	/* minimal io size for this device */
	u32 sector_size;

	/* physical drive uuid (or lvm uuid) */
109
	u8 uuid[BTRFS_UUID_SIZE];
110

111 112 113 114 115 116 117 118
	/*
	 * size of the device on the current transaction
	 *
	 * This variant is update when committing the transaction,
	 * and protected by device_list_mutex
	 */
	u64 commit_total_bytes;

119 120
	/* bytes used on the current transaction */
	u64 commit_bytes_used;
121 122 123 124 125 126 127
	/*
	 * used to manage the device which is resized
	 *
	 * It is protected by chunk_lock.
	 */
	struct list_head resized_list;

128 129 130 131
	/* for sending down flush barriers */
	struct bio *flush_bio;
	struct completion flush_wait;

A
Arne Jansen 已提交
132
	/* per-device scrub information */
133
	struct scrub_ctx *scrub_device;
A
Arne Jansen 已提交
134

135
	struct btrfs_work work;
136
	struct rcu_head rcu;
137 138 139 140 141 142 143 144

	/* readahead state */
	spinlock_t reada_lock;
	atomic_t reada_in_flight;
	u64 reada_next;
	struct reada_zone *reada_curr_zone;
	struct radix_tree_root reada_zones;
	struct radix_tree_root reada_extents;
C
Chris Mason 已提交
145

146 147
	/* disk I/O failure stats. For detailed description refer to
	 * enum btrfs_dev_stat_values in ioctl.h */
148
	int dev_stats_valid;
149 150 151

	/* Counter to record the change of device stats */
	atomic_t dev_stats_ccnt;
152
	atomic_t dev_stat_values[BTRFS_DEV_STAT_VALUES_MAX];
153 154
};

155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221
/*
 * If we read those variants at the context of their own lock, we needn't
 * use the following helpers, reading them directly is safe.
 */
#if BITS_PER_LONG==32 && defined(CONFIG_SMP)
#define BTRFS_DEVICE_GETSET_FUNCS(name)					\
static inline u64							\
btrfs_device_get_##name(const struct btrfs_device *dev)			\
{									\
	u64 size;							\
	unsigned int seq;						\
									\
	do {								\
		seq = read_seqcount_begin(&dev->data_seqcount);		\
		size = dev->name;					\
	} while (read_seqcount_retry(&dev->data_seqcount, seq));	\
	return size;							\
}									\
									\
static inline void							\
btrfs_device_set_##name(struct btrfs_device *dev, u64 size)		\
{									\
	preempt_disable();						\
	write_seqcount_begin(&dev->data_seqcount);			\
	dev->name = size;						\
	write_seqcount_end(&dev->data_seqcount);			\
	preempt_enable();						\
}
#elif BITS_PER_LONG==32 && defined(CONFIG_PREEMPT)
#define BTRFS_DEVICE_GETSET_FUNCS(name)					\
static inline u64							\
btrfs_device_get_##name(const struct btrfs_device *dev)			\
{									\
	u64 size;							\
									\
	preempt_disable();						\
	size = dev->name;						\
	preempt_enable();						\
	return size;							\
}									\
									\
static inline void							\
btrfs_device_set_##name(struct btrfs_device *dev, u64 size)		\
{									\
	preempt_disable();						\
	dev->name = size;						\
	preempt_enable();						\
}
#else
#define BTRFS_DEVICE_GETSET_FUNCS(name)					\
static inline u64							\
btrfs_device_get_##name(const struct btrfs_device *dev)			\
{									\
	return dev->name;						\
}									\
									\
static inline void							\
btrfs_device_set_##name(struct btrfs_device *dev, u64 size)		\
{									\
	dev->name = size;						\
}
#endif

BTRFS_DEVICE_GETSET_FUNCS(total_bytes);
BTRFS_DEVICE_GETSET_FUNCS(disk_total_bytes);
BTRFS_DEVICE_GETSET_FUNCS(bytes_used);

222 223 224 225
struct btrfs_fs_devices {
	u8 fsid[BTRFS_FSID_SIZE]; /* FS specific uuid */

	u64 num_devices;
226
	u64 open_devices;
Y
Yan Zheng 已提交
227
	u64 rw_devices;
228
	u64 missing_devices;
Y
Yan Zheng 已提交
229
	u64 total_rw_bytes;
J
Josef Bacik 已提交
230
	u64 total_devices;
231
	struct block_device *latest_bdev;
232 233 234

	/* all of the devices in the FS, protected by a mutex
	 * so we can safely walk it to write out the supers without
235 236 237
	 * worrying about add/remove by the multi-device code.
	 * Scrubbing super can kick off supers writing by holding
	 * this mutex lock.
238 239
	 */
	struct mutex device_list_mutex;
240
	struct list_head devices;
241

242
	struct list_head resized_devices;
243 244
	/* devices not currently being allocated */
	struct list_head alloc_list;
245
	struct list_head list;
Y
Yan Zheng 已提交
246 247 248 249 250

	struct btrfs_fs_devices *seed;
	int seeding;

	int opened;
C
Chris Mason 已提交
251 252 253 254 255

	/* set when we find or add a device that doesn't have the
	 * nonrot flag set
	 */
	int rotating;
256

257
	struct btrfs_fs_info *fs_info;
258
	/* sysfs kobjects */
259
	struct kobject fsid_kobj;
260 261
	struct kobject *device_dir_kobj;
	struct completion kobj_unregister;
262 263
};

264 265
#define BTRFS_BIO_INLINE_CSUM_SIZE	64

266 267 268 269 270 271 272 273 274
/*
 * we need the mirror number and stripe index to be passed around
 * the call chain while we are processing end_io (especially errors).
 * Really, what we need is a btrfs_bio structure that has this info
 * and is properly sized with its stripe array, but we're not there
 * quite yet.  We have our own btrfs bioset, and all of the bios
 * we allocate are actually btrfs_io_bios.  We'll cram as much of
 * struct btrfs_bio as we can into this over time.
 */
275
typedef void (btrfs_io_bio_end_io_t) (struct btrfs_io_bio *bio, int err);
276
struct btrfs_io_bio {
277 278 279
	unsigned int mirror_num;
	unsigned int stripe_index;
	u64 logical;
280 281 282 283
	u8 *csum;
	u8 csum_inline[BTRFS_BIO_INLINE_CSUM_SIZE];
	u8 *csum_allocated;
	btrfs_io_bio_end_io_t *end_io;
284
	struct bvec_iter iter;
285 286 287 288
	/*
	 * This member must come last, bio_alloc_bioset will allocate enough
	 * bytes for entire btrfs_io_bio but relies on bio being last.
	 */
289 290 291 292 293 294 295 296
	struct bio bio;
};

static inline struct btrfs_io_bio *btrfs_io_bio(struct bio *bio)
{
	return container_of(bio, struct btrfs_io_bio, bio);
}

297 298 299
struct btrfs_bio_stripe {
	struct btrfs_device *dev;
	u64 physical;
300
	u64 length; /* only used for discard mappings */
301 302
};

303 304 305 306
struct btrfs_bio;
typedef void (btrfs_bio_end_io_t) (struct btrfs_bio *bio, int err);

struct btrfs_bio {
307
	refcount_t refs;
308
	atomic_t stripes_pending;
309
	struct btrfs_fs_info *fs_info;
Z
Zhao Lei 已提交
310
	u64 map_type; /* get from map_lookup->type */
311
	bio_end_io_t *end_io;
312
	struct bio *orig_bio;
313
	unsigned long flags;
314
	void *private;
315 316
	atomic_t error;
	int max_errors;
317
	int num_stripes;
318
	int mirror_num;
319 320
	int num_tgtdevs;
	int *tgtdev_map;
321 322 323 324 325 326
	/*
	 * logical block numbers for the start of each stripe
	 * The last one or two are p/q.  These are sorted,
	 * so raid_map[0] is the start of our full stripe
	 */
	u64 *raid_map;
327 328 329
	struct btrfs_bio_stripe stripes[];
};

330 331 332 333
struct btrfs_device_info {
	struct btrfs_device *dev;
	u64 dev_offset;
	u64 max_avail;
334
	u64 total_avail;
335 336
};

337 338 339 340 341
struct btrfs_raid_attr {
	int sub_stripes;	/* sub_stripes info for map */
	int dev_stripes;	/* stripes per dev */
	int devs_max;		/* max devs to use */
	int devs_min;		/* min devs needed */
342
	int tolerated_failures; /* max tolerated fail devs */
343 344 345 346
	int devs_increment;	/* ndevs has to be a multiple of this */
	int ncopies;		/* how many copies to data has */
};

Z
Zhao Lei 已提交
347
extern const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES];
348
extern const int btrfs_raid_mindev_error[BTRFS_NR_RAID_TYPES];
Z
Zhao Lei 已提交
349 350
extern const u64 btrfs_raid_group[BTRFS_NR_RAID_TYPES];

351 352 353 354
struct map_lookup {
	u64 type;
	int io_align;
	int io_width;
355
	u64 stripe_len;
356 357 358 359 360
	int num_stripes;
	int sub_stripes;
	struct btrfs_bio_stripe stripes[];
};

A
Arne Jansen 已提交
361 362 363
#define map_lookup_size(n) (sizeof(struct map_lookup) + \
			    (sizeof(struct btrfs_bio_stripe) * (n)))

364
struct btrfs_balance_args;
365
struct btrfs_balance_progress;
366 367 368 369 370 371 372 373
struct btrfs_balance_control {
	struct btrfs_fs_info *fs_info;

	struct btrfs_balance_args data;
	struct btrfs_balance_args meta;
	struct btrfs_balance_args sys;

	u64 flags;
374 375

	struct btrfs_balance_progress stat;
376 377
};

378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398
enum btrfs_map_op {
	BTRFS_MAP_READ,
	BTRFS_MAP_WRITE,
	BTRFS_MAP_DISCARD,
	BTRFS_MAP_GET_READ_MIRRORS,
};

static inline enum btrfs_map_op btrfs_op(struct bio *bio)
{
	switch (bio_op(bio)) {
	case REQ_OP_DISCARD:
		return BTRFS_MAP_DISCARD;
	case REQ_OP_WRITE:
		return BTRFS_MAP_WRITE;
	default:
		WARN_ON_ONCE(1);
	case REQ_OP_READ:
		return BTRFS_MAP_READ;
	}
}

399 400
int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start,
				   u64 end, u64 *length);
401 402
void btrfs_get_bbio(struct btrfs_bio *bbio);
void btrfs_put_bbio(struct btrfs_bio *bbio);
403
int btrfs_map_block(struct btrfs_fs_info *fs_info, enum btrfs_map_op op,
404
		    u64 logical, u64 *length,
405
		    struct btrfs_bio **bbio_ret, int mirror_num);
406
int btrfs_map_sblock(struct btrfs_fs_info *fs_info, enum btrfs_map_op op,
407
		     u64 logical, u64 *length,
408
		     struct btrfs_bio **bbio_ret);
409
int btrfs_rmap_block(struct btrfs_fs_info *fs_info,
Y
Yan Zheng 已提交
410 411
		     u64 chunk_start, u64 physical, u64 devid,
		     u64 **logical, int *naddrs, int *stripe_len);
412
int btrfs_read_sys_array(struct btrfs_fs_info *fs_info);
413
int btrfs_read_chunk_tree(struct btrfs_fs_info *fs_info);
414
int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
415
		      struct btrfs_fs_info *fs_info, u64 type);
416 417
void btrfs_mapping_init(struct btrfs_mapping_tree *tree);
void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree);
418 419
blk_status_t btrfs_map_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
			   int mirror_num, int async_submit);
420
int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
421 422
		       fmode_t flags, void *holder);
int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
423 424
			  struct btrfs_fs_devices **fs_devices_ret);
int btrfs_close_devices(struct btrfs_fs_devices *fs_devices);
425
void btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices, int step);
426 427
void btrfs_assign_next_active_device(struct btrfs_fs_info *fs_info,
		struct btrfs_device *device, struct btrfs_device *this_dev);
428
int btrfs_find_device_missing_or_by_path(struct btrfs_fs_info *fs_info,
429
					 const char *device_path,
430
					 struct btrfs_device **device);
431
int btrfs_find_device_by_devspec(struct btrfs_fs_info *fs_info, u64 devid,
432
					 const char *devpath,
433
					 struct btrfs_device **device);
434 435 436
struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
					const u64 *devid,
					const u8 *uuid);
437
int btrfs_rm_device(struct btrfs_fs_info *fs_info,
438
		    const char *device_path, u64 devid);
439
void btrfs_cleanup_fs_uuids(void);
440
int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len);
441 442
int btrfs_grow_device(struct btrfs_trans_handle *trans,
		      struct btrfs_device *device, u64 new_size);
443
struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
Y
Yan Zheng 已提交
444
				       u8 *uuid, u8 *fsid);
445
int btrfs_shrink_device(struct btrfs_device *device, u64 new_size);
446
int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *path);
447
int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
448
				  const char *device_path,
449
				  struct btrfs_device *srcdev,
450
				  struct btrfs_device **device_out);
451 452
int btrfs_balance(struct btrfs_balance_control *bctl,
		  struct btrfs_ioctl_balance_args *bargs);
453
int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info);
454
int btrfs_recover_balance(struct btrfs_fs_info *fs_info);
455
int btrfs_pause_balance(struct btrfs_fs_info *fs_info);
456
int btrfs_cancel_balance(struct btrfs_fs_info *fs_info);
457
int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info);
458
int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info);
459
int btrfs_chunk_readonly(struct btrfs_fs_info *fs_info, u64 chunk_offset);
460 461 462
int find_free_dev_extent_start(struct btrfs_transaction *transaction,
			 struct btrfs_device *device, u64 num_bytes,
			 u64 search_start, u64 *start, u64 *max_avail);
463 464
int find_free_dev_extent(struct btrfs_trans_handle *trans,
			 struct btrfs_device *device, u64 num_bytes,
465
			 u64 *start, u64 *max_avail);
466
void btrfs_dev_stat_inc_and_print(struct btrfs_device *dev, int index);
467
int btrfs_get_dev_stats(struct btrfs_fs_info *fs_info,
468
			struct btrfs_ioctl_get_dev_stats *stats);
469
void btrfs_init_devices_late(struct btrfs_fs_info *fs_info);
470 471 472
int btrfs_init_dev_stats(struct btrfs_fs_info *fs_info);
int btrfs_run_dev_stats(struct btrfs_trans_handle *trans,
			struct btrfs_fs_info *fs_info);
473 474 475 476
void btrfs_rm_dev_replace_remove_srcdev(struct btrfs_fs_info *fs_info,
					struct btrfs_device *srcdev);
void btrfs_rm_dev_replace_free_srcdev(struct btrfs_fs_info *fs_info,
				      struct btrfs_device *srcdev);
477 478 479 480
void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
				      struct btrfs_device *tgtdev);
void btrfs_init_dev_replace_tgtdev_for_resume(struct btrfs_fs_info *fs_info,
					      struct btrfs_device *tgtdev);
481
void btrfs_scratch_superblocks(struct block_device *bdev, const char *device_path);
482
int btrfs_is_parity_mirror(struct btrfs_fs_info *fs_info,
483
			   u64 logical, u64 len);
484
unsigned long btrfs_full_stripe_len(struct btrfs_fs_info *fs_info,
D
David Woodhouse 已提交
485
				    u64 logical);
486
int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans,
487
				struct btrfs_fs_info *fs_info,
488
				u64 chunk_offset, u64 chunk_size);
489
int btrfs_remove_chunk(struct btrfs_trans_handle *trans,
490
		       struct btrfs_fs_info *fs_info, u64 chunk_offset);
491

492 493 494 495
static inline void btrfs_dev_stat_inc(struct btrfs_device *dev,
				      int index)
{
	atomic_inc(dev->dev_stat_values + index);
496 497 498 499 500 501
	/*
	 * This memory barrier orders stores updating statistics before stores
	 * updating dev_stats_ccnt.
	 *
	 * It pairs with smp_rmb() in btrfs_run_dev_stats().
	 */
502 503
	smp_mb__before_atomic();
	atomic_inc(&dev->dev_stats_ccnt);
504 505 506 507 508 509 510 511 512 513 514 515 516 517
}

static inline int btrfs_dev_stat_read(struct btrfs_device *dev,
				      int index)
{
	return atomic_read(dev->dev_stat_values + index);
}

static inline int btrfs_dev_stat_read_and_reset(struct btrfs_device *dev,
						int index)
{
	int ret;

	ret = atomic_xchg(dev->dev_stat_values + index, 0);
518 519 520 521 522 523 524
	/*
	 * atomic_xchg implies a full memory barriers as per atomic_t.txt:
	 * - RMW operations that have a return value are fully ordered;
	 *
	 * This implicit memory barriers is paired with the smp_rmb in
	 * btrfs_run_dev_stats
	 */
525
	atomic_inc(&dev->dev_stats_ccnt);
526 527 528 529 530 531 532
	return ret;
}

static inline void btrfs_dev_stat_set(struct btrfs_device *dev,
				      int index, unsigned long val)
{
	atomic_set(dev->dev_stat_values + index, val);
533 534 535 536 537 538
	/*
	 * This memory barrier orders stores updating statistics before stores
	 * updating dev_stats_ccnt.
	 *
	 * It pairs with smp_rmb() in btrfs_run_dev_stats().
	 */
539 540
	smp_mb__before_atomic();
	atomic_inc(&dev->dev_stats_ccnt);
541 542 543 544 545 546 547
}

static inline void btrfs_dev_stat_reset(struct btrfs_device *dev,
					int index)
{
	btrfs_dev_stat_set(dev, index, 0);
}
548 549

void btrfs_update_commit_device_size(struct btrfs_fs_info *fs_info);
550
void btrfs_update_commit_device_bytes_used(struct btrfs_fs_info *fs_info,
551
					struct btrfs_transaction *transaction);
552

553
struct list_head *btrfs_get_fs_uuids(void);
554 555
void btrfs_set_fs_info_ptr(struct btrfs_fs_info *fs_info);
void btrfs_reset_fs_info_ptr(struct btrfs_fs_info *fs_info);
556

557 558
bool btrfs_check_rw_degradable(struct btrfs_fs_info *fs_info);

559
#endif