extent_io.h 17.7 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
2 3 4

#ifndef BTRFS_EXTENT_IO_H
#define BTRFS_EXTENT_IO_H
5 6

#include <linux/rbtree.h>
7
#include <linux/refcount.h>
8
#include "ulist.h"
9 10

/* bits for the extent state */
11
#define EXTENT_DIRTY		(1U << 0)
N
Nikolay Borisov 已提交
12 13 14 15 16 17 18 19 20 21 22 23 24 25
#define EXTENT_UPTODATE		(1U << 1)
#define EXTENT_LOCKED		(1U << 2)
#define EXTENT_NEW		(1U << 3)
#define EXTENT_DELALLOC		(1U << 4)
#define EXTENT_DEFRAG		(1U << 5)
#define EXTENT_BOUNDARY		(1U << 6)
#define EXTENT_NODATASUM	(1U << 7)
#define EXTENT_CLEAR_META_RESV	(1U << 8)
#define EXTENT_NEED_WAIT	(1U << 9)
#define EXTENT_DAMAGED		(1U << 10)
#define EXTENT_NORESERVE	(1U << 11)
#define EXTENT_QGROUP_RESERVED	(1U << 12)
#define EXTENT_CLEAR_DATA_RESV	(1U << 13)
#define EXTENT_DELALLOC_NEW	(1U << 14)
26 27
#define EXTENT_DO_ACCOUNTING    (EXTENT_CLEAR_META_RESV | \
				 EXTENT_CLEAR_DATA_RESV)
28
#define EXTENT_CTLBITS		(EXTENT_DO_ACCOUNTING)
29

30 31 32 33
/*
 * flags for bio submission. The high bits indicate the compression
 * type for this bio
 */
C
Chris Mason 已提交
34
#define EXTENT_BIO_COMPRESSED 1
35
#define EXTENT_BIO_FLAG_SHIFT 16
C
Chris Mason 已提交
36

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
enum {
	EXTENT_BUFFER_UPTODATE,
	EXTENT_BUFFER_DIRTY,
	EXTENT_BUFFER_CORRUPT,
	/* this got triggered by readahead */
	EXTENT_BUFFER_READAHEAD,
	EXTENT_BUFFER_TREE_REF,
	EXTENT_BUFFER_STALE,
	EXTENT_BUFFER_WRITEBACK,
	/* read IO error */
	EXTENT_BUFFER_READ_ERR,
	EXTENT_BUFFER_UNMAPPED,
	EXTENT_BUFFER_IN_TREE,
	/* write IO error */
	EXTENT_BUFFER_WRITE_ERR,
};
53

54
/* these are flags for __process_pages_contig */
55 56 57 58 59
#define PAGE_UNLOCK		(1 << 0)
#define PAGE_CLEAR_DIRTY	(1 << 1)
#define PAGE_SET_WRITEBACK	(1 << 2)
#define PAGE_END_WRITEBACK	(1 << 3)
#define PAGE_SET_PRIVATE2	(1 << 4)
60
#define PAGE_SET_ERROR		(1 << 5)
61
#define PAGE_LOCK		(1 << 6)
62

63 64 65 66 67 68
/*
 * page->private values.  Every page that is controlled by the extent
 * map has page->private set to one.
 */
#define EXTENT_PAGE_PRIVATE 1

69 70 71 72 73 74 75 76 77 78 79 80 81 82
/*
 * The extent buffer bitmap operations are done with byte granularity instead of
 * word granularity for two reasons:
 * 1. The bitmaps must be little-endian on disk.
 * 2. Bitmap items are not guaranteed to be aligned to a word and therefore a
 *    single word in a bitmap may straddle two pages in the extent buffer.
 */
#define BIT_BYTE(nr) ((nr) / BITS_PER_BYTE)
#define BYTE_MASK ((1 << BITS_PER_BYTE) - 1)
#define BITMAP_FIRST_BYTE_MASK(start) \
	((BYTE_MASK << ((start) & (BITS_PER_BYTE - 1))) & BYTE_MASK)
#define BITMAP_LAST_BYTE_MASK(nbits) \
	(BYTE_MASK >> (-(nbits) & (BITS_PER_BYTE - 1)))

83
struct extent_state;
84
struct btrfs_root;
85
struct btrfs_inode;
86
struct btrfs_io_bio;
87
struct io_failure_record;
88

89
typedef	blk_status_t (extent_submit_bio_hook_t)(void *private_data, struct bio *bio,
90 91
				       int mirror_num, unsigned long bio_flags,
				       u64 bio_offset);
92 93

typedef blk_status_t (extent_submit_bio_start_t)(void *private_data,
94
		struct bio *bio, u64 bio_offset);
95

96
struct extent_io_ops {
97
	/*
98
	 * The following callbacks must be always defined, the function
99 100
	 * pointer will be called unconditionally.
	 */
101
	extent_submit_bio_hook_t *submit_bio_hook;
102 103 104
	int (*readpage_end_io_hook)(struct btrfs_io_bio *io_bio, u64 phy_offset,
				    struct page *page, u64 start, u64 end,
				    int mirror);
105 106
};

107 108 109 110 111 112 113 114 115 116 117
enum {
	IO_TREE_FS_INFO_FREED_EXTENTS0,
	IO_TREE_FS_INFO_FREED_EXTENTS1,
	IO_TREE_INODE_IO,
	IO_TREE_INODE_IO_FAILURE,
	IO_TREE_RELOC_BLOCKS,
	IO_TREE_TRANS_DIRTY_PAGES,
	IO_TREE_ROOT_DIRTY_LOG_PAGES,
	IO_TREE_SELFTEST,
};

118 119
struct extent_io_tree {
	struct rb_root state;
120
	struct btrfs_fs_info *fs_info;
121
	void *private_data;
122
	u64 dirty_bytes;
123
	bool track_uptodate;
124 125 126 127

	/* Who owns this io tree, should be one of IO_TREE_* */
	u8 owner;

128
	spinlock_t lock;
129
	const struct extent_io_ops *ops;
130 131 132 133 134 135
};

struct extent_state {
	u64 start;
	u64 end; /* inclusive */
	struct rb_node rb_node;
J
Josef Bacik 已提交
136 137

	/* ADD NEW ELEMENTS AFTER THIS */
138
	wait_queue_head_t wq;
139
	refcount_t refs;
140
	unsigned state;
141

142
	struct io_failure_record *failrec;
143

144
#ifdef CONFIG_BTRFS_DEBUG
145
	struct list_head leak_list;
146
#endif
147 148
};

149
#define INLINE_EXTENT_BUFFER_PAGES 16
150
#define MAX_INLINE_EXTENT_BUFFER_SIZE (INLINE_EXTENT_BUFFER_PAGES * PAGE_SIZE)
151 152 153
struct extent_buffer {
	u64 start;
	unsigned long len;
154
	unsigned long bflags;
155
	struct btrfs_fs_info *fs_info;
156
	spinlock_t refs_lock;
157
	atomic_t refs;
158
	atomic_t io_pages;
159
	int read_mirror;
160
	struct rcu_head rcu_head;
161
	pid_t lock_owner;
162

163 164 165 166 167 168 169
	/* count of read lock holders on the extent buffer */
	atomic_t write_locks;
	atomic_t read_locks;
	atomic_t blocking_writers;
	atomic_t blocking_readers;
	atomic_t spinning_readers;
	atomic_t spinning_writers;
170 171 172
	short lock_nested;
	/* >= 0 if eb belongs to a log tree, -1 otherwise */
	short log_index;
173 174 175 176 177 178 179 180

	/* protects write locks */
	rwlock_t lock;

	/* readers use lock_wq while they wait for the write
	 * lock holders to unlock
	 */
	wait_queue_head_t write_lock_wq;
181

182 183
	/* writers use read_lock_wq while they wait for readers
	 * to unlock
184
	 */
185
	wait_queue_head_t read_lock_wq;
186
	struct page *pages[INLINE_EXTENT_BUFFER_PAGES];
187 188 189
#ifdef CONFIG_BTRFS_DEBUG
	struct list_head leak_list;
#endif
190 191
};

192 193 194 195 196
/*
 * Structure to record how many bytes and which ranges are set/cleared
 */
struct extent_changeset {
	/* How many bytes are set/cleared in this operation */
197
	unsigned int bytes_changed;
198 199

	/* Changed ranges */
200
	struct ulist range_changed;
201 202
};

203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
static inline void extent_changeset_init(struct extent_changeset *changeset)
{
	changeset->bytes_changed = 0;
	ulist_init(&changeset->range_changed);
}

static inline struct extent_changeset *extent_changeset_alloc(void)
{
	struct extent_changeset *ret;

	ret = kmalloc(sizeof(*ret), GFP_KERNEL);
	if (!ret)
		return NULL;

	extent_changeset_init(ret);
	return ret;
}

static inline void extent_changeset_release(struct extent_changeset *changeset)
{
	if (!changeset)
		return;
	changeset->bytes_changed = 0;
	ulist_release(&changeset->range_changed);
}

static inline void extent_changeset_free(struct extent_changeset *changeset)
{
	if (!changeset)
		return;
	extent_changeset_release(changeset);
	kfree(changeset);
}

237 238 239 240 241 242 243 244 245 246 247
static inline void extent_set_compress_type(unsigned long *bio_flags,
					    int compress_type)
{
	*bio_flags |= compress_type << EXTENT_BIO_FLAG_SHIFT;
}

static inline int extent_compress_type(unsigned long bio_flags)
{
	return bio_flags >> EXTENT_BIO_FLAG_SHIFT;
}

248 249
struct extent_map_tree;

250
typedef struct extent_map *(get_extent_t)(struct btrfs_inode *inode,
251
					  struct page *page,
252
					  size_t pg_offset,
253 254 255
					  u64 start, u64 len,
					  int create);

256
void extent_io_tree_init(struct btrfs_fs_info *fs_info,
257 258
			 struct extent_io_tree *tree, unsigned int owner,
			 void *private_data);
259
int try_release_extent_mapping(struct page *page, gfp_t mask);
260
int try_release_extent_buffer(struct page *page);
261
int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
262
		     struct extent_state **cached);
263 264 265 266 267 268

static inline int lock_extent(struct extent_io_tree *tree, u64 start, u64 end)
{
	return lock_extent_bits(tree, start, end, NULL);
}

269
int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end);
270
int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
271
			  get_extent_t *get_extent, int mirror_num);
272
int __init extent_io_init(void);
273
void __cold extent_io_exit(void);
274 275 276

u64 count_range_bits(struct extent_io_tree *tree,
		     u64 *start, u64 search_end,
277
		     u64 max_bytes, unsigned bits, int contig);
278

279
void free_extent_state(struct extent_state *state);
280
int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end,
281
		   unsigned bits, int filled,
282
		   struct extent_state *cached_state);
283
int clear_record_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
284
		unsigned bits, struct extent_changeset *changeset);
285
int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
286
		     unsigned bits, int wake, int delete,
287
		     struct extent_state **cached);
288 289 290 291
int __clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
		     unsigned bits, int wake, int delete,
		     struct extent_state **cached, gfp_t mask,
		     struct extent_changeset *changeset);
292

293 294
static inline int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end)
{
295
	return clear_extent_bit(tree, start, end, EXTENT_LOCKED, 1, 0, NULL);
296 297 298
}

static inline int unlock_extent_cached(struct extent_io_tree *tree, u64 start,
299
		u64 end, struct extent_state **cached)
300
{
301
	return __clear_extent_bit(tree, start, end, EXTENT_LOCKED, 1, 0, cached,
302
				GFP_NOFS, NULL);
303 304
}

305 306
static inline int unlock_extent_cached_atomic(struct extent_io_tree *tree,
		u64 start, u64 end, struct extent_state **cached)
307
{
308 309
	return __clear_extent_bit(tree, start, end, EXTENT_LOCKED, 1, 0, cached,
				GFP_ATOMIC, NULL);
310 311 312
}

static inline int clear_extent_bits(struct extent_io_tree *tree, u64 start,
313
		u64 end, unsigned bits)
314 315 316 317 318 319
{
	int wake = 0;

	if (bits & EXTENT_LOCKED)
		wake = 1;

320
	return clear_extent_bit(tree, start, end, bits, wake, 0, NULL);
321 322
}

323
int set_record_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
324
			   unsigned bits, struct extent_changeset *changeset);
325
int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
326
		   unsigned bits, u64 *failed_start,
327
		   struct extent_state **cached_state, gfp_t mask);
328 329

static inline int set_extent_bits(struct extent_io_tree *tree, u64 start,
330
		u64 end, unsigned bits)
331
{
332
	return set_extent_bit(tree, start, end, bits, NULL, NULL, GFP_NOFS);
333 334
}

335
static inline int clear_extent_uptodate(struct extent_io_tree *tree, u64 start,
336
		u64 end, struct extent_state **cached_state)
337
{
338
	return __clear_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, 0,
339
				cached_state, GFP_NOFS, NULL);
340
}
341 342 343 344 345 346 347 348

static inline int set_extent_dirty(struct extent_io_tree *tree, u64 start,
		u64 end, gfp_t mask)
{
	return set_extent_bit(tree, start, end, EXTENT_DIRTY, NULL,
			      NULL, mask);
}

349
static inline int clear_extent_dirty(struct extent_io_tree *tree, u64 start,
350
				     u64 end, struct extent_state **cached)
351 352 353
{
	return clear_extent_bit(tree, start, end,
				EXTENT_DIRTY | EXTENT_DELALLOC |
354
				EXTENT_DO_ACCOUNTING, 0, 0, cached);
355 356
}

J
Josef Bacik 已提交
357
int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
358
		       unsigned bits, unsigned clear_bits,
359
		       struct extent_state **cached_state);
360 361

static inline int set_extent_delalloc(struct extent_io_tree *tree, u64 start,
362 363
				      u64 end, unsigned int extra_bits,
				      struct extent_state **cached_state)
364 365
{
	return set_extent_bit(tree, start, end,
366
			      EXTENT_DELALLOC | EXTENT_UPTODATE | extra_bits,
367
			      NULL, cached_state, GFP_NOFS);
368 369 370
}

static inline int set_extent_defrag(struct extent_io_tree *tree, u64 start,
371
		u64 end, struct extent_state **cached_state)
372 373 374
{
	return set_extent_bit(tree, start, end,
			      EXTENT_DELALLOC | EXTENT_UPTODATE | EXTENT_DEFRAG,
375
			      NULL, cached_state, GFP_NOFS);
376 377 378
}

static inline int set_extent_new(struct extent_io_tree *tree, u64 start,
379
		u64 end)
380
{
381 382
	return set_extent_bit(tree, start, end, EXTENT_NEW, NULL, NULL,
			GFP_NOFS);
383 384 385 386 387 388 389 390 391
}

static inline int set_extent_uptodate(struct extent_io_tree *tree, u64 start,
		u64 end, struct extent_state **cached_state, gfp_t mask)
{
	return set_extent_bit(tree, start, end, EXTENT_UPTODATE, NULL,
			      cached_state, mask);
}

392
int find_first_extent_bit(struct extent_io_tree *tree, u64 start,
393
			  u64 *start_ret, u64 *end_ret, unsigned bits,
394
			  struct extent_state **cached_state);
395 396
int extent_invalidatepage(struct extent_io_tree *tree,
			  struct page *page, unsigned long offset);
397
int extent_write_full_page(struct page *page, struct writeback_control *wbc);
398
int extent_write_locked_range(struct inode *inode, u64 start, u64 end,
399
			      int mode);
400
int extent_writepages(struct address_space *mapping,
401
		      struct writeback_control *wbc);
402 403
int btree_write_cache_pages(struct address_space *mapping,
			    struct writeback_control *wbc);
404 405
int extent_readpages(struct address_space *mapping, struct list_head *pages,
		     unsigned nr_pages);
Y
Yehuda Sadeh 已提交
406
int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
407
		__u64 start, __u64 len);
408 409
void set_page_extent_mapped(struct page *page);

410
struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info,
411
					  u64 start);
412 413
struct extent_buffer *__alloc_dummy_extent_buffer(struct btrfs_fs_info *fs_info,
						  u64 start, unsigned long len);
414
struct extent_buffer *alloc_dummy_extent_buffer(struct btrfs_fs_info *fs_info,
415
						u64 start);
416
struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src);
417
struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info,
418
					 u64 start);
419
void free_extent_buffer(struct extent_buffer *eb);
420
void free_extent_buffer_stale(struct extent_buffer *eb);
421 422 423
#define WAIT_NONE	0
#define WAIT_COMPLETE	1
#define WAIT_PAGE_LOCK	2
424
int read_extent_buffer_pages(struct extent_io_tree *tree,
425
			     struct extent_buffer *eb, int wait,
426
			     int mirror_num);
427
void wait_on_extent_buffer_writeback(struct extent_buffer *eb);
428

429
static inline int num_extent_pages(const struct extent_buffer *eb)
430
{
431 432
	return (round_up(eb->start + eb->len, PAGE_SIZE) >> PAGE_SHIFT) -
	       (eb->start >> PAGE_SHIFT);
433 434
}

435 436 437 438 439
static inline void extent_buffer_get(struct extent_buffer *eb)
{
	atomic_inc(&eb->refs);
}

440 441 442 443 444
static inline int extent_buffer_uptodate(struct extent_buffer *eb)
{
	return test_bit(EXTENT_BUFFER_UPTODATE, &eb->bflags);
}

445 446 447
int memcmp_extent_buffer(const struct extent_buffer *eb, const void *ptrv,
			 unsigned long start, unsigned long len);
void read_extent_buffer(const struct extent_buffer *eb, void *dst,
448 449
			unsigned long start,
			unsigned long len);
450 451
int read_extent_buffer_to_user(const struct extent_buffer *eb,
			       void __user *dst, unsigned long start,
452
			       unsigned long len);
453 454 455
void write_extent_buffer_fsid(struct extent_buffer *eb, const void *src);
void write_extent_buffer_chunk_tree_uuid(struct extent_buffer *eb,
		const void *src);
456 457
void write_extent_buffer(struct extent_buffer *eb, const void *src,
			 unsigned long start, unsigned long len);
458 459
void copy_extent_buffer_full(struct extent_buffer *dst,
			     struct extent_buffer *src);
460 461 462 463 464 465 466
void copy_extent_buffer(struct extent_buffer *dst, struct extent_buffer *src,
			unsigned long dst_offset, unsigned long src_offset,
			unsigned long len);
void memcpy_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
			   unsigned long src_offset, unsigned long len);
void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
			   unsigned long src_offset, unsigned long len);
467 468
void memzero_extent_buffer(struct extent_buffer *eb, unsigned long start,
			   unsigned long len);
469 470 471 472 473 474
int extent_buffer_test_bit(struct extent_buffer *eb, unsigned long start,
			   unsigned long pos);
void extent_buffer_bitmap_set(struct extent_buffer *eb, unsigned long start,
			      unsigned long pos, unsigned long len);
void extent_buffer_bitmap_clear(struct extent_buffer *eb, unsigned long start,
				unsigned long pos, unsigned long len);
475
void clear_extent_buffer_dirty(struct extent_buffer *eb);
476
bool set_extent_buffer_dirty(struct extent_buffer *eb);
477
void set_extent_buffer_uptodate(struct extent_buffer *eb);
478
void clear_extent_buffer_uptodate(struct extent_buffer *eb);
479
int extent_buffer_under_io(struct extent_buffer *eb);
480 481 482 483
int map_private_extent_buffer(const struct extent_buffer *eb,
			      unsigned long offset, unsigned long min_len,
			      char **map, unsigned long *map_start,
			      unsigned long *map_len);
484
void extent_range_clear_dirty_for_io(struct inode *inode, u64 start, u64 end);
485
void extent_range_redirty_for_io(struct inode *inode, u64 start, u64 end);
486
void extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end,
487
				 u64 delalloc_end, struct page *locked_page,
488
				 unsigned bits_to_clear,
489
				 unsigned long page_ops);
490
struct bio *btrfs_bio_alloc(struct block_device *bdev, u64 first_byte);
491
struct bio *btrfs_io_bio_alloc(unsigned int nr_iovecs);
492
struct bio *btrfs_bio_clone(struct bio *bio);
493
struct bio *btrfs_bio_clone_partial(struct bio *orig, int offset, int size);
494

495
struct btrfs_fs_info;
496
struct btrfs_inode;
497

498 499 500
int repair_io_failure(struct btrfs_fs_info *fs_info, u64 ino, u64 start,
		      u64 length, u64 logical, struct page *page,
		      unsigned int pg_offset, int mirror_num);
501 502 503 504
int clean_io_failure(struct btrfs_fs_info *fs_info,
		     struct extent_io_tree *failure_tree,
		     struct extent_io_tree *io_tree, u64 start,
		     struct page *page, u64 ino, unsigned int pg_offset);
505
void end_extent_writepage(struct page *page, int err, u64 start, u64 end);
506 507
int repair_eb_io_failure(struct btrfs_fs_info *fs_info,
			 struct extent_buffer *eb, int mirror_num);
508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527

/*
 * When IO fails, either with EIO or csum verification fails, we
 * try other mirrors that might have a good copy of the data.  This
 * io_failure_record is used to record state as we go through all the
 * mirrors.  If another mirror has good data, the page is set up to date
 * and things continue.  If a good mirror can't be found, the original
 * bio end_io callback is called to indicate things have failed.
 */
struct io_failure_record {
	struct page *page;
	u64 start;
	u64 len;
	u64 logical;
	unsigned long bio_flags;
	int this_mirror;
	int failed_mirror;
	int in_validation;
};

528

529 530
void btrfs_free_io_failure_record(struct btrfs_inode *inode, u64 start,
		u64 end);
531 532
int btrfs_get_io_failure_record(struct inode *inode, u64 start, u64 end,
				struct io_failure_record **failrec_ret);
533
bool btrfs_check_repairable(struct inode *inode, unsigned failed_bio_pages,
534
			    struct io_failure_record *failrec, int fail_mirror);
535 536 537
struct bio *btrfs_create_repair_bio(struct inode *inode, struct bio *failed_bio,
				    struct io_failure_record *failrec,
				    struct page *page, int pg_offset, int icsum,
538
				    bio_end_io_t *endio_func, void *data);
539 540 541
int free_io_failure(struct extent_io_tree *failure_tree,
		    struct extent_io_tree *io_tree,
		    struct io_failure_record *rec);
542
#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
543
bool find_lock_delalloc_range(struct inode *inode, struct extent_io_tree *tree,
544 545
			     struct page *locked_page, u64 *start,
			     u64 *end);
546
#endif
547
struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info,
548
					       u64 start);
549

550
#endif