volumes.h 16.3 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
2 3 4 5
/*
 * Copyright (C) 2007 Oracle.  All rights reserved.
 */

6 7
#ifndef BTRFS_VOLUMES_H
#define BTRFS_VOLUMES_H
8

9
#include <linux/bio.h>
10
#include <linux/sort.h>
11
#include <linux/btrfs.h>
12
#include "async-thread.h"
13

14 15
extern struct mutex uuid_mutex;

16
#define BTRFS_STRIPE_LEN	SZ_64K
17

18
struct buffer_head;
19 20 21 22 23
struct btrfs_pending_bios {
	struct bio *head;
	struct bio *tail;
};

24 25 26 27 28 29 30 31 32 33 34 35 36
/*
 * 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

37
#define BTRFS_DEV_STATE_WRITEABLE	(0)
38
#define BTRFS_DEV_STATE_IN_FS_METADATA	(1)
39
#define BTRFS_DEV_STATE_MISSING		(2)
40
#define BTRFS_DEV_STATE_REPLACE_TGT	(3)
41
#define BTRFS_DEV_STATE_FLUSH_SENT	(4)
42

43 44
struct btrfs_device {
	struct list_head dev_list;
45
	struct list_head dev_alloc_list;
Y
Yan Zheng 已提交
46
	struct btrfs_fs_devices *fs_devices;
47
	struct btrfs_fs_info *fs_info;
48

49 50 51 52 53 54
	struct rcu_string *name;

	u64 generation;

	spinlock_t io_lock ____cacheline_aligned;
	int running_pending;
55 56
	/* regular prio bios */
	struct btrfs_pending_bios pending_bios;
57
	/* sync bios */
58 59
	struct btrfs_pending_bios pending_sync_bios;

60 61 62 63 64
	struct block_device *bdev;

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

65
	unsigned long dev_state;
66
	blk_status_t last_flush_error;
67
	int flush_bio_sent;
68

69 70 71 72
#ifdef __BTRFS_NEED_DEVICE_DATA_ORDERED
	seqcount_t data_seqcount;
#endif

73 74 75
	/* the internal btrfs device id */
	u64 devid;

76
	/* size of the device in memory */
77 78
	u64 total_bytes;

79
	/* size of the device on disk */
80 81
	u64 disk_total_bytes;

82 83 84 85 86 87 88 89
	/* bytes used */
	u64 bytes_used;

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

	/* optimal io width for this device */
	u32 io_width;
90 91
	/* type and info about this device */
	u64 type;
92 93 94 95 96

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

	/* physical drive uuid (or lvm uuid) */
97
	u8 uuid[BTRFS_UUID_SIZE];
98

99 100 101 102 103 104 105 106
	/*
	 * 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;

107 108
	/* bytes used on the current transaction */
	u64 commit_bytes_used;
109 110 111 112 113 114 115
	/*
	 * used to manage the device which is resized
	 *
	 * It is protected by chunk_lock.
	 */
	struct list_head resized_list;

116 117 118 119
	/* for sending down flush barriers */
	struct bio *flush_bio;
	struct completion flush_wait;

A
Arne Jansen 已提交
120
	/* per-device scrub information */
121
	struct scrub_ctx *scrub_ctx;
A
Arne Jansen 已提交
122

123
	struct btrfs_work work;
124
	struct rcu_head rcu;
125 126 127 128 129 130 131

	/* readahead state */
	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
132

133 134
	/* disk I/O failure stats. For detailed description refer to
	 * enum btrfs_dev_stat_values in ioctl.h */
135
	int dev_stats_valid;
136 137 138

	/* Counter to record the change of device stats */
	atomic_t dev_stats_ccnt;
139
	atomic_t dev_stat_values[BTRFS_DEV_STAT_VALUES_MAX];
140 141
};

142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208
/*
 * 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);

209 210
struct btrfs_fs_devices {
	u8 fsid[BTRFS_FSID_SIZE]; /* FS specific uuid */
211
	struct list_head fs_list;
212 213

	u64 num_devices;
214
	u64 open_devices;
Y
Yan Zheng 已提交
215
	u64 rw_devices;
216
	u64 missing_devices;
Y
Yan Zheng 已提交
217
	u64 total_rw_bytes;
J
Josef Bacik 已提交
218
	u64 total_devices;
219
	struct block_device *latest_bdev;
220 221 222

	/* all of the devices in the FS, protected by a mutex
	 * so we can safely walk it to write out the supers without
223 224 225
	 * worrying about add/remove by the multi-device code.
	 * Scrubbing super can kick off supers writing by holding
	 * this mutex lock.
226 227
	 */
	struct mutex device_list_mutex;
228
	struct list_head devices;
229

230
	struct list_head resized_devices;
231 232
	/* devices not currently being allocated */
	struct list_head alloc_list;
Y
Yan Zheng 已提交
233 234 235 236 237

	struct btrfs_fs_devices *seed;
	int seeding;

	int opened;
C
Chris Mason 已提交
238 239 240 241 242

	/* set when we find or add a device that doesn't have the
	 * nonrot flag set
	 */
	int rotating;
243

244
	struct btrfs_fs_info *fs_info;
245
	/* sysfs kobjects */
246
	struct kobject fsid_kobj;
247 248
	struct kobject *device_dir_kobj;
	struct completion kobj_unregister;
249 250
};

251 252
#define BTRFS_BIO_INLINE_CSUM_SIZE	64

253 254 255 256 257 258 259 260 261
/*
 * 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.
 */
262
typedef void (btrfs_io_bio_end_io_t) (struct btrfs_io_bio *bio, int err);
263
struct btrfs_io_bio {
264 265 266
	unsigned int mirror_num;
	unsigned int stripe_index;
	u64 logical;
267 268 269 270
	u8 *csum;
	u8 csum_inline[BTRFS_BIO_INLINE_CSUM_SIZE];
	u8 *csum_allocated;
	btrfs_io_bio_end_io_t *end_io;
271
	struct bvec_iter iter;
272 273 274 275
	/*
	 * This member must come last, bio_alloc_bioset will allocate enough
	 * bytes for entire btrfs_io_bio but relies on bio being last.
	 */
276 277 278 279 280 281 282 283
	struct bio bio;
};

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

284 285 286
struct btrfs_bio_stripe {
	struct btrfs_device *dev;
	u64 physical;
287
	u64 length; /* only used for discard mappings */
288 289
};

290 291 292 293
struct btrfs_bio;
typedef void (btrfs_bio_end_io_t) (struct btrfs_bio *bio, int err);

struct btrfs_bio {
294
	refcount_t refs;
295
	atomic_t stripes_pending;
296
	struct btrfs_fs_info *fs_info;
Z
Zhao Lei 已提交
297
	u64 map_type; /* get from map_lookup->type */
298
	bio_end_io_t *end_io;
299
	struct bio *orig_bio;
300
	unsigned long flags;
301
	void *private;
302 303
	atomic_t error;
	int max_errors;
304
	int num_stripes;
305
	int mirror_num;
306 307
	int num_tgtdevs;
	int *tgtdev_map;
308 309 310 311 312 313
	/*
	 * 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;
314 315 316
	struct btrfs_bio_stripe stripes[];
};

317 318 319 320
struct btrfs_device_info {
	struct btrfs_device *dev;
	u64 dev_offset;
	u64 max_avail;
321
	u64 total_avail;
322 323
};

324 325 326 327 328
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 */
329
	int tolerated_failures; /* max tolerated fail devs */
330 331
	int devs_increment;	/* ndevs has to be a multiple of this */
	int ncopies;		/* how many copies to data has */
332
	int mindev_error;	/* error code if min devs requisite is unmet */
333
	const char raid_name[8]; /* name of the raid */
334
	u64 bg_flag;		/* block group flag of the raid */
335 336
};

Z
Zhao Lei 已提交
337 338
extern const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES];

339 340 341 342
struct map_lookup {
	u64 type;
	int io_align;
	int io_width;
343
	u64 stripe_len;
344 345 346 347 348
	int num_stripes;
	int sub_stripes;
	struct btrfs_bio_stripe stripes[];
};

A
Arne Jansen 已提交
349 350 351
#define map_lookup_size(n) (sizeof(struct map_lookup) + \
			    (sizeof(struct btrfs_bio_stripe) * (n)))

352
struct btrfs_balance_args;
353
struct btrfs_balance_progress;
354 355 356 357 358 359
struct btrfs_balance_control {
	struct btrfs_balance_args data;
	struct btrfs_balance_args meta;
	struct btrfs_balance_args sys;

	u64 flags;
360 361

	struct btrfs_balance_progress stat;
362 363
};

364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384
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;
	}
}

385 386
int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start,
				   u64 end, u64 *length);
387 388
void btrfs_get_bbio(struct btrfs_bio *bbio);
void btrfs_put_bbio(struct btrfs_bio *bbio);
389
int btrfs_map_block(struct btrfs_fs_info *fs_info, enum btrfs_map_op op,
390
		    u64 logical, u64 *length,
391
		    struct btrfs_bio **bbio_ret, int mirror_num);
392
int btrfs_map_sblock(struct btrfs_fs_info *fs_info, enum btrfs_map_op op,
393
		     u64 logical, u64 *length,
394
		     struct btrfs_bio **bbio_ret);
395 396
int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
		     u64 physical, u64 **logical, int *naddrs, int *stripe_len);
397
int btrfs_read_sys_array(struct btrfs_fs_info *fs_info);
398
int btrfs_read_chunk_tree(struct btrfs_fs_info *fs_info);
399
int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
400
		      struct btrfs_fs_info *fs_info, u64 type);
401 402
void btrfs_mapping_init(struct btrfs_mapping_tree *tree);
void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree);
403 404
blk_status_t btrfs_map_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
			   int mirror_num, int async_submit);
405
int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
406 407
		       fmode_t flags, void *holder);
int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
408 409
			  struct btrfs_fs_devices **fs_devices_ret);
int btrfs_close_devices(struct btrfs_fs_devices *fs_devices);
410
void btrfs_free_extra_devids(struct btrfs_fs_devices *fs_devices, int step);
411 412
void btrfs_assign_next_active_device(struct btrfs_fs_info *fs_info,
		struct btrfs_device *device, struct btrfs_device *this_dev);
413
int btrfs_find_device_missing_or_by_path(struct btrfs_fs_info *fs_info,
414
					 const char *device_path,
415
					 struct btrfs_device **device);
416
int btrfs_find_device_by_devspec(struct btrfs_fs_info *fs_info, u64 devid,
417
					 const char *devpath,
418
					 struct btrfs_device **device);
419 420 421
struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
					const u64 *devid,
					const u8 *uuid);
422
void btrfs_free_device(struct btrfs_device *device);
423
int btrfs_rm_device(struct btrfs_fs_info *fs_info,
424
		    const char *device_path, u64 devid);
425
void __exit btrfs_cleanup_fs_uuids(void);
426
int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len);
427 428
int btrfs_grow_device(struct btrfs_trans_handle *trans,
		      struct btrfs_device *device, u64 new_size);
429
struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
Y
Yan Zheng 已提交
430
				       u8 *uuid, u8 *fsid);
431
int btrfs_shrink_device(struct btrfs_device *device, u64 new_size);
432
int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *path);
433 434
int btrfs_balance(struct btrfs_fs_info *fs_info,
		  struct btrfs_balance_control *bctl,
435
		  struct btrfs_ioctl_balance_args *bargs);
436
int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info);
437
int btrfs_recover_balance(struct btrfs_fs_info *fs_info);
438
int btrfs_pause_balance(struct btrfs_fs_info *fs_info);
439
int btrfs_cancel_balance(struct btrfs_fs_info *fs_info);
440
int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info);
441
int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info);
442
int btrfs_chunk_readonly(struct btrfs_fs_info *fs_info, u64 chunk_offset);
443 444 445
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);
446 447
int find_free_dev_extent(struct btrfs_trans_handle *trans,
			 struct btrfs_device *device, u64 num_bytes,
448
			 u64 *start, u64 *max_avail);
449
void btrfs_dev_stat_inc_and_print(struct btrfs_device *dev, int index);
450
int btrfs_get_dev_stats(struct btrfs_fs_info *fs_info,
451
			struct btrfs_ioctl_get_dev_stats *stats);
452
void btrfs_init_devices_late(struct btrfs_fs_info *fs_info);
453 454 455
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);
456 457 458 459
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);
460 461
void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
				      struct btrfs_device *tgtdev);
462
void btrfs_scratch_superblocks(struct block_device *bdev, const char *device_path);
463
int btrfs_is_parity_mirror(struct btrfs_fs_info *fs_info,
464
			   u64 logical, u64 len);
465
unsigned long btrfs_full_stripe_len(struct btrfs_fs_info *fs_info,
D
David Woodhouse 已提交
466
				    u64 logical);
467
int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans,
468
				struct btrfs_fs_info *fs_info,
469
				u64 chunk_offset, u64 chunk_size);
470
int btrfs_remove_chunk(struct btrfs_trans_handle *trans,
471
		       struct btrfs_fs_info *fs_info, u64 chunk_offset);
472

473 474 475 476
static inline void btrfs_dev_stat_inc(struct btrfs_device *dev,
				      int index)
{
	atomic_inc(dev->dev_stat_values + index);
477 478 479 480 481 482
	/*
	 * This memory barrier orders stores updating statistics before stores
	 * updating dev_stats_ccnt.
	 *
	 * It pairs with smp_rmb() in btrfs_run_dev_stats().
	 */
483 484
	smp_mb__before_atomic();
	atomic_inc(&dev->dev_stats_ccnt);
485 486 487 488 489 490 491 492 493 494 495 496 497 498
}

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);
499 500 501 502 503 504 505
	/*
	 * 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
	 */
506
	atomic_inc(&dev->dev_stats_ccnt);
507 508 509 510 511 512 513
	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);
514 515 516 517 518 519
	/*
	 * This memory barrier orders stores updating statistics before stores
	 * updating dev_stats_ccnt.
	 *
	 * It pairs with smp_rmb() in btrfs_run_dev_stats().
	 */
520 521
	smp_mb__before_atomic();
	atomic_inc(&dev->dev_stats_ccnt);
522 523 524 525 526 527 528
}

static inline void btrfs_dev_stat_reset(struct btrfs_device *dev,
					int index)
{
	btrfs_dev_stat_set(dev, index, 0);
}
529

530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551
/*
 * Convert block group flags (BTRFS_BLOCK_GROUP_*) to btrfs_raid_types, which
 * can be used as index to access btrfs_raid_array[].
 */
static inline enum btrfs_raid_types btrfs_bg_flags_to_raid_index(u64 flags)
{
	if (flags & BTRFS_BLOCK_GROUP_RAID10)
		return BTRFS_RAID_RAID10;
	else if (flags & BTRFS_BLOCK_GROUP_RAID1)
		return BTRFS_RAID_RAID1;
	else if (flags & BTRFS_BLOCK_GROUP_DUP)
		return BTRFS_RAID_DUP;
	else if (flags & BTRFS_BLOCK_GROUP_RAID0)
		return BTRFS_RAID_RAID0;
	else if (flags & BTRFS_BLOCK_GROUP_RAID5)
		return BTRFS_RAID_RAID5;
	else if (flags & BTRFS_BLOCK_GROUP_RAID6)
		return BTRFS_RAID_RAID6;

	return BTRFS_RAID_SINGLE; /* BTRFS_BLOCK_GROUP_SINGLE */
}

552 553
const char *get_raid_name(enum btrfs_raid_types type);

554
void btrfs_update_commit_device_size(struct btrfs_fs_info *fs_info);
555
void btrfs_update_commit_device_bytes_used(struct btrfs_transaction *trans);
556

557
struct list_head *btrfs_get_fs_uuids(void);
558 559
void btrfs_set_fs_info_ptr(struct btrfs_fs_info *fs_info);
void btrfs_reset_fs_info_ptr(struct btrfs_fs_info *fs_info);
560 561
bool btrfs_check_rw_degradable(struct btrfs_fs_info *fs_info,
					struct btrfs_device *failing_dev);
562

563
#endif