disk-io.c 119.3 KB
Newer Older
C
Chris Mason 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
/*
 * 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.
 */

C
Chris Mason 已提交
19
#include <linux/fs.h>
20
#include <linux/blkdev.h>
C
Chris Mason 已提交
21
#include <linux/scatterlist.h>
C
Chris Mason 已提交
22
#include <linux/swap.h>
23
#include <linux/radix-tree.h>
C
Chris Mason 已提交
24
#include <linux/writeback.h>
C
Chris Mason 已提交
25
#include <linux/buffer_head.h>
26
#include <linux/workqueue.h>
27
#include <linux/kthread.h>
C
Chris Mason 已提交
28
#include <linux/freezer.h>
29
#include <linux/slab.h>
30
#include <linux/migrate.h>
31
#include <linux/ratelimit.h>
32
#include <linux/uuid.h>
S
Stefan Behrens 已提交
33
#include <linux/semaphore.h>
34
#include <asm/unaligned.h>
35 36
#include "ctree.h"
#include "disk-io.h"
37
#include "hash.h"
38
#include "transaction.h"
39
#include "btrfs_inode.h"
40
#include "volumes.h"
41
#include "print-tree.h"
42
#include "locking.h"
43
#include "tree-log.h"
44
#include "free-space-cache.h"
45
#include "inode-map.h"
46
#include "check-integrity.h"
47
#include "rcu-string.h"
48
#include "dev-replace.h"
D
David Woodhouse 已提交
49
#include "raid56.h"
50
#include "sysfs.h"
J
Josef Bacik 已提交
51
#include "qgroup.h"
52

53 54 55 56
#ifdef CONFIG_X86
#include <asm/cpufeature.h>
#endif

57
static const struct extent_io_ops btree_extent_io_ops;
58
static void end_workqueue_fn(struct btrfs_work *work);
59
static void free_fs_root(struct btrfs_root *root);
60
static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
L
liubo 已提交
61
				    int read_only);
62
static void btrfs_destroy_ordered_extents(struct btrfs_root *root);
L
liubo 已提交
63 64
static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
				      struct btrfs_root *root);
65
static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root);
L
liubo 已提交
66 67 68 69 70
static int btrfs_destroy_marked_extents(struct btrfs_root *root,
					struct extent_io_tree *dirty_pages,
					int mark);
static int btrfs_destroy_pinned_extent(struct btrfs_root *root,
				       struct extent_io_tree *pinned_extents);
71 72
static int btrfs_cleanup_transaction(struct btrfs_root *root);
static void btrfs_error_commit_super(struct btrfs_root *root);
73

C
Chris Mason 已提交
74
/*
75 76
 * btrfs_end_io_wq structs are used to do processing in task context when an IO
 * is complete.  This is used during reads to verify checksums, and it is used
C
Chris Mason 已提交
77 78
 * by writes to insert metadata for new file extents after IO is complete.
 */
79
struct btrfs_end_io_wq {
80 81 82 83 84
	struct bio *bio;
	bio_end_io_t *end_io;
	void *private;
	struct btrfs_fs_info *info;
	int error;
85
	enum btrfs_wq_endio_type metadata;
86
	struct list_head list;
87
	struct btrfs_work work;
88
};
89

90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
static struct kmem_cache *btrfs_end_io_wq_cache;

int __init btrfs_end_io_wq_init(void)
{
	btrfs_end_io_wq_cache = kmem_cache_create("btrfs_end_io_wq",
					sizeof(struct btrfs_end_io_wq),
					0,
					SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD,
					NULL);
	if (!btrfs_end_io_wq_cache)
		return -ENOMEM;
	return 0;
}

void btrfs_end_io_wq_exit(void)
{
	if (btrfs_end_io_wq_cache)
		kmem_cache_destroy(btrfs_end_io_wq_cache);
}

C
Chris Mason 已提交
110 111 112 113 114
/*
 * async submit bios are used to offload expensive checksumming
 * onto the worker threads.  They checksum file and metadata bios
 * just before they are sent down the IO stack.
 */
115 116 117 118
struct async_submit_bio {
	struct inode *inode;
	struct bio *bio;
	struct list_head list;
C
Chris Mason 已提交
119 120
	extent_submit_bio_hook_t *submit_bio_start;
	extent_submit_bio_hook_t *submit_bio_done;
121 122
	int rw;
	int mirror_num;
C
Chris Mason 已提交
123
	unsigned long bio_flags;
124 125 126 127 128
	/*
	 * bio_offset is optional, can be used if the pages in the bio
	 * can't tell us where in the file the bio should go
	 */
	u64 bio_offset;
129
	struct btrfs_work work;
130
	int error;
131 132
};

133 134 135 136 137 138 139 140 141 142
/*
 * Lockdep class keys for extent_buffer->lock's in this root.  For a given
 * eb, the lockdep key is determined by the btrfs_root it belongs to and
 * the level the eb occupies in the tree.
 *
 * Different roots are used for different purposes and may nest inside each
 * other and they require separate keysets.  As lockdep keys should be
 * static, assign keysets according to the purpose of the root as indicated
 * by btrfs_root->objectid.  This ensures that all special purpose roots
 * have separate keysets.
143
 *
144 145 146
 * Lock-nesting across peer nodes is always done with the immediate parent
 * node locked thus preventing deadlock.  As lockdep doesn't know this, use
 * subclass to avoid triggering lockdep warning in such cases.
147
 *
148 149 150
 * The key is set by the readpage_end_io_hook after the buffer has passed
 * csum validation but before the pages are unlocked.  It is also set by
 * btrfs_init_new_buffer on freshly allocated blocks.
151
 *
152 153 154
 * We also add a check to make sure the highest level of the tree is the
 * same as our lockdep setup here.  If BTRFS_MAX_LEVEL changes, this code
 * needs update as well.
155 156 157 158 159
 */
#ifdef CONFIG_DEBUG_LOCK_ALLOC
# if BTRFS_MAX_LEVEL != 8
#  error
# endif
160 161 162 163 164 165 166 167 168 169 170 171 172

static struct btrfs_lockdep_keyset {
	u64			id;		/* root objectid */
	const char		*name_stem;	/* lock name stem */
	char			names[BTRFS_MAX_LEVEL + 1][20];
	struct lock_class_key	keys[BTRFS_MAX_LEVEL + 1];
} btrfs_lockdep_keysets[] = {
	{ .id = BTRFS_ROOT_TREE_OBJECTID,	.name_stem = "root"	},
	{ .id = BTRFS_EXTENT_TREE_OBJECTID,	.name_stem = "extent"	},
	{ .id = BTRFS_CHUNK_TREE_OBJECTID,	.name_stem = "chunk"	},
	{ .id = BTRFS_DEV_TREE_OBJECTID,	.name_stem = "dev"	},
	{ .id = BTRFS_FS_TREE_OBJECTID,		.name_stem = "fs"	},
	{ .id = BTRFS_CSUM_TREE_OBJECTID,	.name_stem = "csum"	},
173
	{ .id = BTRFS_QUOTA_TREE_OBJECTID,	.name_stem = "quota"	},
174 175 176
	{ .id = BTRFS_TREE_LOG_OBJECTID,	.name_stem = "log"	},
	{ .id = BTRFS_TREE_RELOC_OBJECTID,	.name_stem = "treloc"	},
	{ .id = BTRFS_DATA_RELOC_TREE_OBJECTID,	.name_stem = "dreloc"	},
177
	{ .id = BTRFS_UUID_TREE_OBJECTID,	.name_stem = "uuid"	},
178
	{ .id = 0,				.name_stem = "tree"	},
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

void __init btrfs_init_lockdep(void)
{
	int i, j;

	/* initialize lockdep class names */
	for (i = 0; i < ARRAY_SIZE(btrfs_lockdep_keysets); i++) {
		struct btrfs_lockdep_keyset *ks = &btrfs_lockdep_keysets[i];

		for (j = 0; j < ARRAY_SIZE(ks->names); j++)
			snprintf(ks->names[j], sizeof(ks->names[j]),
				 "btrfs-%s-%02d", ks->name_stem, j);
	}
}

void btrfs_set_buffer_lockdep_class(u64 objectid, struct extent_buffer *eb,
				    int level)
{
	struct btrfs_lockdep_keyset *ks;

	BUG_ON(level >= ARRAY_SIZE(ks->keys));

	/* find the matching keyset, id 0 is the default entry */
	for (ks = btrfs_lockdep_keysets; ks->id; ks++)
		if (ks->id == objectid)
			break;

	lockdep_set_class_and_name(&eb->lock,
				   &ks->keys[level], ks->names[level]);
}

211 212
#endif

C
Chris Mason 已提交
213 214 215 216
/*
 * extents on the btree inode are pretty simple, there's one extent
 * that covers the entire device
 */
217
static struct extent_map *btree_get_extent(struct inode *inode,
218
		struct page *page, size_t pg_offset, u64 start, u64 len,
219
		int create)
220
{
221 222 223 224
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	struct extent_map *em;
	int ret;

225
	read_lock(&em_tree->lock);
226
	em = lookup_extent_mapping(em_tree, start, len);
227 228 229
	if (em) {
		em->bdev =
			BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
230
		read_unlock(&em_tree->lock);
231
		goto out;
232
	}
233
	read_unlock(&em_tree->lock);
234

235
	em = alloc_extent_map();
236 237 238 239 240
	if (!em) {
		em = ERR_PTR(-ENOMEM);
		goto out;
	}
	em->start = 0;
241
	em->len = (u64)-1;
C
Chris Mason 已提交
242
	em->block_len = (u64)-1;
243
	em->block_start = 0;
244
	em->bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
245

246
	write_lock(&em_tree->lock);
J
Josef Bacik 已提交
247
	ret = add_extent_mapping(em_tree, em, 0);
248 249
	if (ret == -EEXIST) {
		free_extent_map(em);
250
		em = lookup_extent_mapping(em_tree, start, len);
251
		if (!em)
252
			em = ERR_PTR(-EIO);
253
	} else if (ret) {
254
		free_extent_map(em);
255
		em = ERR_PTR(ret);
256
	}
257
	write_unlock(&em_tree->lock);
258

259 260
out:
	return em;
261 262
}

263
u32 btrfs_csum_data(char *data, u32 seed, size_t len)
264
{
265
	return btrfs_crc32c(seed, data, len);
266 267 268 269
}

void btrfs_csum_final(u32 crc, char *result)
{
270
	put_unaligned_le32(~crc, result);
271 272
}

C
Chris Mason 已提交
273 274 275 276
/*
 * compute the csum for a btree block, and either verify it or write it
 * into the csum field of the block.
 */
277 278
static int csum_tree_block(struct btrfs_fs_info *fs_info,
			   struct extent_buffer *buf,
279 280
			   int verify)
{
281
	u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
282
	char *result = NULL;
283 284 285 286 287 288 289 290
	unsigned long len;
	unsigned long cur_len;
	unsigned long offset = BTRFS_CSUM_SIZE;
	char *kaddr;
	unsigned long map_start;
	unsigned long map_len;
	int err;
	u32 crc = ~(u32)0;
291
	unsigned long inline_result;
292 293

	len = buf->len - offset;
C
Chris Mason 已提交
294
	while (len > 0) {
295
		err = map_private_extent_buffer(buf, offset, 32,
296
					&kaddr, &map_start, &map_len);
C
Chris Mason 已提交
297
		if (err)
298 299
			return 1;
		cur_len = min(len, map_len - (offset - map_start));
300
		crc = btrfs_csum_data(kaddr + offset - map_start,
301 302 303 304
				      crc, cur_len);
		len -= cur_len;
		offset += cur_len;
	}
305
	if (csum_size > sizeof(inline_result)) {
306
		result = kzalloc(csum_size, GFP_NOFS);
307 308 309 310 311 312
		if (!result)
			return 1;
	} else {
		result = (char *)&inline_result;
	}

313 314 315
	btrfs_csum_final(crc, result);

	if (verify) {
316
		if (memcmp_extent_buffer(buf, result, 0, csum_size)) {
317 318
			u32 val;
			u32 found = 0;
319
			memcpy(&found, result, csum_size);
320

321
			read_extent_buffer(buf, &val, 0, csum_size);
322
			printk_ratelimited(KERN_WARNING
323 324
				"BTRFS: %s checksum verify failed on %llu wanted %X found %X "
				"level %d\n",
325
				fs_info->sb->s_id, buf->start,
326
				val, found, btrfs_header_level(buf));
327 328
			if (result != (char *)&inline_result)
				kfree(result);
329 330 331
			return 1;
		}
	} else {
332
		write_extent_buffer(buf, result, 0, csum_size);
333
	}
334 335
	if (result != (char *)&inline_result)
		kfree(result);
336 337 338
	return 0;
}

C
Chris Mason 已提交
339 340 341 342 343 344
/*
 * we can't consider a given block up to date unless the transid of the
 * block matches the transid in the parent node's pointer.  This is how we
 * detect blocks that either didn't get written at all or got written
 * in the wrong place.
 */
345
static int verify_parent_transid(struct extent_io_tree *io_tree,
346 347
				 struct extent_buffer *eb, u64 parent_transid,
				 int atomic)
348
{
349
	struct extent_state *cached_state = NULL;
350
	int ret;
351
	bool need_lock = (current->journal_info == BTRFS_SEND_TRANS_STUB);
352 353 354 355

	if (!parent_transid || btrfs_header_generation(eb) == parent_transid)
		return 0;

356 357 358
	if (atomic)
		return -EAGAIN;

359 360 361 362 363
	if (need_lock) {
		btrfs_tree_read_lock(eb);
		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
	}

364
	lock_extent_bits(io_tree, eb->start, eb->start + eb->len - 1,
365
			 0, &cached_state);
366
	if (extent_buffer_uptodate(eb) &&
367 368 369 370
	    btrfs_header_generation(eb) == parent_transid) {
		ret = 0;
		goto out;
	}
371 372
	printk_ratelimited(KERN_ERR
	    "BTRFS (device %s): parent transid verify failed on %llu wanted %llu found %llu\n",
373 374
			eb->fs_info->sb->s_id, eb->start,
			parent_transid, btrfs_header_generation(eb));
375
	ret = 1;
376 377 378 379 380 381 382 383 384 385 386

	/*
	 * Things reading via commit roots that don't have normal protection,
	 * like send, can have a really old block in cache that may point at a
	 * block that has been free'd and re-allocated.  So don't clear uptodate
	 * if we find an eb that is under IO (dirty/writeback) because we could
	 * end up reading in the stale data and then writing it back out and
	 * making everybody very sad.
	 */
	if (!extent_buffer_under_io(eb))
		clear_extent_buffer_uptodate(eb);
C
Chris Mason 已提交
387
out:
388 389
	unlock_extent_cached(io_tree, eb->start, eb->start + eb->len - 1,
			     &cached_state, GFP_NOFS);
390 391
	if (need_lock)
		btrfs_tree_read_unlock_blocking(eb);
392 393 394
	return ret;
}

D
David Sterba 已提交
395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424
/*
 * Return 0 if the superblock checksum type matches the checksum value of that
 * algorithm. Pass the raw disk superblock data.
 */
static int btrfs_check_super_csum(char *raw_disk_sb)
{
	struct btrfs_super_block *disk_sb =
		(struct btrfs_super_block *)raw_disk_sb;
	u16 csum_type = btrfs_super_csum_type(disk_sb);
	int ret = 0;

	if (csum_type == BTRFS_CSUM_TYPE_CRC32) {
		u32 crc = ~(u32)0;
		const int csum_size = sizeof(crc);
		char result[csum_size];

		/*
		 * The super_block structure does not span the whole
		 * BTRFS_SUPER_INFO_SIZE range, we expect that the unused space
		 * is filled with zeros and is included in the checkum.
		 */
		crc = btrfs_csum_data(raw_disk_sb + BTRFS_CSUM_SIZE,
				crc, BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE);
		btrfs_csum_final(crc, result);

		if (memcmp(raw_disk_sb, result, csum_size))
			ret = 1;
	}

	if (csum_type >= ARRAY_SIZE(btrfs_csum_sizes)) {
425
		printk(KERN_ERR "BTRFS: unsupported checksum algorithm %u\n",
D
David Sterba 已提交
426 427 428 429 430 431 432
				csum_type);
		ret = 1;
	}

	return ret;
}

C
Chris Mason 已提交
433 434 435 436
/*
 * helper to read a given tree block, doing retries as required when
 * the checksums don't match and we have alternate mirrors to try.
 */
437 438
static int btree_read_extent_buffer_pages(struct btrfs_root *root,
					  struct extent_buffer *eb,
439
					  u64 start, u64 parent_transid)
440 441
{
	struct extent_io_tree *io_tree;
442
	int failed = 0;
443 444 445
	int ret;
	int num_copies = 0;
	int mirror_num = 0;
446
	int failed_mirror = 0;
447

448
	clear_bit(EXTENT_BUFFER_CORRUPT, &eb->bflags);
449 450
	io_tree = &BTRFS_I(root->fs_info->btree_inode)->io_tree;
	while (1) {
451 452
		ret = read_extent_buffer_pages(io_tree, eb, start,
					       WAIT_COMPLETE,
453
					       btree_get_extent, mirror_num);
454 455
		if (!ret) {
			if (!verify_parent_transid(io_tree, eb,
456
						   parent_transid, 0))
457 458 459 460
				break;
			else
				ret = -EIO;
		}
C
Chris Mason 已提交
461

462 463 464 465 466 467
		/*
		 * This buffer's crc is fine, but its contents are corrupted, so
		 * there is no reason to read the other copies, they won't be
		 * any less wrong.
		 */
		if (test_bit(EXTENT_BUFFER_CORRUPT, &eb->bflags))
468 469
			break;

470
		num_copies = btrfs_num_copies(root->fs_info,
471
					      eb->start, eb->len);
C
Chris Mason 已提交
472
		if (num_copies == 1)
473
			break;
C
Chris Mason 已提交
474

475 476 477 478 479
		if (!failed_mirror) {
			failed = 1;
			failed_mirror = eb->read_mirror;
		}

480
		mirror_num++;
481 482 483
		if (mirror_num == failed_mirror)
			mirror_num++;

C
Chris Mason 已提交
484
		if (mirror_num > num_copies)
485
			break;
486
	}
487

488
	if (failed && !ret && failed_mirror)
489 490 491
		repair_eb_io_failure(root, eb, failed_mirror);

	return ret;
492
}
493

C
Chris Mason 已提交
494
/*
C
Chris Mason 已提交
495 496
 * checksum a dirty tree block before IO.  This has extra checks to make sure
 * we only fill in the checksum field in the first page of a multi-page block
C
Chris Mason 已提交
497
 */
C
Chris Mason 已提交
498

499
static int csum_dirty_buffer(struct btrfs_fs_info *fs_info, struct page *page)
500
{
M
Miao Xie 已提交
501
	u64 start = page_offset(page);
502 503
	u64 found_start;
	struct extent_buffer *eb;
504

J
Josef Bacik 已提交
505 506 507
	eb = (struct extent_buffer *)page->private;
	if (page != eb->pages[0])
		return 0;
508
	found_start = btrfs_header_bytenr(eb);
509
	if (WARN_ON(found_start != start || !PageUptodate(page)))
J
Josef Bacik 已提交
510
		return 0;
511
	csum_tree_block(fs_info, eb, 0);
512 513 514
	return 0;
}

515
static int check_tree_block_fsid(struct btrfs_fs_info *fs_info,
Y
Yan Zheng 已提交
516 517
				 struct extent_buffer *eb)
{
518
	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
Y
Yan Zheng 已提交
519 520 521
	u8 fsid[BTRFS_UUID_SIZE];
	int ret = 1;

522
	read_extent_buffer(eb, fsid, btrfs_header_fsid(), BTRFS_FSID_SIZE);
Y
Yan Zheng 已提交
523 524 525 526 527 528 529 530 531 532
	while (fs_devices) {
		if (!memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE)) {
			ret = 0;
			break;
		}
		fs_devices = fs_devices->seed;
	}
	return ret;
}

533
#define CORRUPT(reason, eb, root, slot)				\
534 535
	btrfs_crit(root->fs_info, "corrupt leaf, %s: block=%llu,"	\
		   "root=%llu, slot=%d", reason,			\
536
	       btrfs_header_bytenr(eb),	root->objectid, slot)
537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598

static noinline int check_leaf(struct btrfs_root *root,
			       struct extent_buffer *leaf)
{
	struct btrfs_key key;
	struct btrfs_key leaf_key;
	u32 nritems = btrfs_header_nritems(leaf);
	int slot;

	if (nritems == 0)
		return 0;

	/* Check the 0 item */
	if (btrfs_item_offset_nr(leaf, 0) + btrfs_item_size_nr(leaf, 0) !=
	    BTRFS_LEAF_DATA_SIZE(root)) {
		CORRUPT("invalid item offset size pair", leaf, root, 0);
		return -EIO;
	}

	/*
	 * Check to make sure each items keys are in the correct order and their
	 * offsets make sense.  We only have to loop through nritems-1 because
	 * we check the current slot against the next slot, which verifies the
	 * next slot's offset+size makes sense and that the current's slot
	 * offset is correct.
	 */
	for (slot = 0; slot < nritems - 1; slot++) {
		btrfs_item_key_to_cpu(leaf, &leaf_key, slot);
		btrfs_item_key_to_cpu(leaf, &key, slot + 1);

		/* Make sure the keys are in the right order */
		if (btrfs_comp_cpu_keys(&leaf_key, &key) >= 0) {
			CORRUPT("bad key order", leaf, root, slot);
			return -EIO;
		}

		/*
		 * Make sure the offset and ends are right, remember that the
		 * item data starts at the end of the leaf and grows towards the
		 * front.
		 */
		if (btrfs_item_offset_nr(leaf, slot) !=
			btrfs_item_end_nr(leaf, slot + 1)) {
			CORRUPT("slot offset bad", leaf, root, slot);
			return -EIO;
		}

		/*
		 * Check to make sure that we don't point outside of the leaf,
		 * just incase all the items are consistent to eachother, but
		 * all point outside of the leaf.
		 */
		if (btrfs_item_end_nr(leaf, slot) >
		    BTRFS_LEAF_DATA_SIZE(root)) {
			CORRUPT("slot end outside of leaf", leaf, root, slot);
			return -EIO;
		}
	}

	return 0;
}

599 600 601
static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
				      u64 phy_offset, struct page *page,
				      u64 start, u64 end, int mirror)
602 603 604 605 606
{
	u64 found_start;
	int found_level;
	struct extent_buffer *eb;
	struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
607
	int ret = 0;
608
	int reads_done;
609 610 611

	if (!page->private)
		goto out;
C
Chris Mason 已提交
612

J
Josef Bacik 已提交
613
	eb = (struct extent_buffer *)page->private;
C
Chris Mason 已提交
614

615 616 617 618 619 620
	/* the pending IO might have been the only thing that kept this buffer
	 * in memory.  Make sure we have a ref for all this other checks
	 */
	extent_buffer_get(eb);

	reads_done = atomic_dec_and_test(&eb->io_pages);
621 622
	if (!reads_done)
		goto err;
623

624
	eb->read_mirror = mirror;
625
	if (test_bit(EXTENT_BUFFER_READ_ERR, &eb->bflags)) {
626 627 628 629
		ret = -EIO;
		goto err;
	}

630
	found_start = btrfs_header_bytenr(eb);
631
	if (found_start != eb->start) {
632
		printk_ratelimited(KERN_ERR "BTRFS (device %s): bad tree block start "
C
Chris Mason 已提交
633
			       "%llu %llu\n",
634
			       eb->fs_info->sb->s_id, found_start, eb->start);
635
		ret = -EIO;
636 637
		goto err;
	}
638
	if (check_tree_block_fsid(root->fs_info, eb)) {
639
		printk_ratelimited(KERN_ERR "BTRFS (device %s): bad fsid on block %llu\n",
640
			       eb->fs_info->sb->s_id, eb->start);
641 642 643
		ret = -EIO;
		goto err;
	}
644
	found_level = btrfs_header_level(eb);
645
	if (found_level >= BTRFS_MAX_LEVEL) {
646
		btrfs_err(root->fs_info, "bad tree block level %d",
647 648 649 650
			   (int)btrfs_header_level(eb));
		ret = -EIO;
		goto err;
	}
651

652 653
	btrfs_set_buffer_lockdep_class(btrfs_header_owner(eb),
				       eb, found_level);
654

655
	ret = csum_tree_block(root->fs_info, eb, 1);
656
	if (ret) {
657
		ret = -EIO;
658 659 660 661 662 663 664 665 666 667 668 669
		goto err;
	}

	/*
	 * If this is a leaf block and it is corrupt, set the corrupt bit so
	 * that we don't try and read the other copies of this block, just
	 * return -EIO.
	 */
	if (found_level == 0 && check_leaf(root, eb)) {
		set_bit(EXTENT_BUFFER_CORRUPT, &eb->bflags);
		ret = -EIO;
	}
670

671 672
	if (!ret)
		set_extent_buffer_uptodate(eb);
673
err:
674 675
	if (reads_done &&
	    test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags))
A
Arne Jansen 已提交
676 677
		btree_readahead_hook(root, eb, eb->start, ret);

D
David Woodhouse 已提交
678 679 680 681 682 683 684
	if (ret) {
		/*
		 * our io error hook is going to dec the io pages
		 * again, we have to make sure it has something
		 * to decrement
		 */
		atomic_inc(&eb->io_pages);
685
		clear_extent_buffer_uptodate(eb);
D
David Woodhouse 已提交
686
	}
687
	free_extent_buffer(eb);
688
out:
689
	return ret;
690 691
}

692
static int btree_io_failed_hook(struct page *page, int failed_mirror)
A
Arne Jansen 已提交
693 694 695 696
{
	struct extent_buffer *eb;
	struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;

J
Josef Bacik 已提交
697
	eb = (struct extent_buffer *)page->private;
698
	set_bit(EXTENT_BUFFER_READ_ERR, &eb->bflags);
699
	eb->read_mirror = failed_mirror;
D
David Woodhouse 已提交
700
	atomic_dec(&eb->io_pages);
701
	if (test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags))
A
Arne Jansen 已提交
702 703 704 705
		btree_readahead_hook(root, eb, eb->start, -EIO);
	return -EIO;	/* we fixed nothing */
}

706
static void end_workqueue_bio(struct bio *bio)
707
{
708
	struct btrfs_end_io_wq *end_io_wq = bio->bi_private;
709
	struct btrfs_fs_info *fs_info;
710 711
	struct btrfs_workqueue *wq;
	btrfs_work_func_t func;
712 713

	fs_info = end_io_wq->info;
714
	end_io_wq->error = bio->bi_error;
715

716
	if (bio->bi_rw & REQ_WRITE) {
717 718 719 720 721 722 723 724 725 726 727 728 729
		if (end_io_wq->metadata == BTRFS_WQ_ENDIO_METADATA) {
			wq = fs_info->endio_meta_write_workers;
			func = btrfs_endio_meta_write_helper;
		} else if (end_io_wq->metadata == BTRFS_WQ_ENDIO_FREE_SPACE) {
			wq = fs_info->endio_freespace_worker;
			func = btrfs_freespace_write_helper;
		} else if (end_io_wq->metadata == BTRFS_WQ_ENDIO_RAID56) {
			wq = fs_info->endio_raid56_workers;
			func = btrfs_endio_raid56_helper;
		} else {
			wq = fs_info->endio_write_workers;
			func = btrfs_endio_write_helper;
		}
730
	} else {
731 732 733 734 735
		if (unlikely(end_io_wq->metadata ==
			     BTRFS_WQ_ENDIO_DIO_REPAIR)) {
			wq = fs_info->endio_repair_workers;
			func = btrfs_endio_repair_helper;
		} else if (end_io_wq->metadata == BTRFS_WQ_ENDIO_RAID56) {
736 737 738 739 740 741 742 743 744
			wq = fs_info->endio_raid56_workers;
			func = btrfs_endio_raid56_helper;
		} else if (end_io_wq->metadata) {
			wq = fs_info->endio_meta_workers;
			func = btrfs_endio_meta_helper;
		} else {
			wq = fs_info->endio_workers;
			func = btrfs_endio_helper;
		}
745
	}
746 747 748

	btrfs_init_work(&end_io_wq->work, func, end_workqueue_fn, NULL, NULL);
	btrfs_queue_work(wq, &end_io_wq->work);
749 750
}

751
int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
752
			enum btrfs_wq_endio_type metadata)
753
{
754
	struct btrfs_end_io_wq *end_io_wq;
755

756
	end_io_wq = kmem_cache_alloc(btrfs_end_io_wq_cache, GFP_NOFS);
757 758 759 760 761
	if (!end_io_wq)
		return -ENOMEM;

	end_io_wq->private = bio->bi_private;
	end_io_wq->end_io = bio->bi_end_io;
762
	end_io_wq->info = info;
763 764
	end_io_wq->error = 0;
	end_io_wq->bio = bio;
765
	end_io_wq->metadata = metadata;
766 767 768

	bio->bi_private = end_io_wq;
	bio->bi_end_io = end_workqueue_bio;
769 770 771
	return 0;
}

772
unsigned long btrfs_async_submit_limit(struct btrfs_fs_info *info)
773
{
774
	unsigned long limit = min_t(unsigned long,
775
				    info->thread_pool_size,
776 777 778
				    info->fs_devices->open_devices);
	return 256 * limit;
}
779

C
Chris Mason 已提交
780 781 782
static void run_one_async_start(struct btrfs_work *work)
{
	struct async_submit_bio *async;
783
	int ret;
C
Chris Mason 已提交
784 785

	async = container_of(work, struct  async_submit_bio, work);
786 787 788 789 790
	ret = async->submit_bio_start(async->inode, async->rw, async->bio,
				      async->mirror_num, async->bio_flags,
				      async->bio_offset);
	if (ret)
		async->error = ret;
C
Chris Mason 已提交
791 792 793
}

static void run_one_async_done(struct btrfs_work *work)
794 795 796
{
	struct btrfs_fs_info *fs_info;
	struct async_submit_bio *async;
797
	int limit;
798 799 800

	async = container_of(work, struct  async_submit_bio, work);
	fs_info = BTRFS_I(async->inode)->root->fs_info;
801

802
	limit = btrfs_async_submit_limit(fs_info);
803 804
	limit = limit * 2 / 3;

805
	if (atomic_dec_return(&fs_info->nr_async_submits) < limit &&
806
	    waitqueue_active(&fs_info->async_submit_wait))
807 808
		wake_up(&fs_info->async_submit_wait);

809 810
	/* If an error occured we just want to clean up the bio and move on */
	if (async->error) {
811 812
		async->bio->bi_error = async->error;
		bio_endio(async->bio);
813 814 815
		return;
	}

C
Chris Mason 已提交
816
	async->submit_bio_done(async->inode, async->rw, async->bio,
817 818
			       async->mirror_num, async->bio_flags,
			       async->bio_offset);
C
Chris Mason 已提交
819 820 821 822 823 824 825
}

static void run_one_async_free(struct btrfs_work *work)
{
	struct async_submit_bio *async;

	async = container_of(work, struct  async_submit_bio, work);
826 827 828
	kfree(async);
}

829 830
int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
			int rw, struct bio *bio, int mirror_num,
C
Chris Mason 已提交
831
			unsigned long bio_flags,
832
			u64 bio_offset,
C
Chris Mason 已提交
833 834
			extent_submit_bio_hook_t *submit_bio_start,
			extent_submit_bio_hook_t *submit_bio_done)
835 836 837 838 839 840 841 842 843 844 845
{
	struct async_submit_bio *async;

	async = kmalloc(sizeof(*async), GFP_NOFS);
	if (!async)
		return -ENOMEM;

	async->inode = inode;
	async->rw = rw;
	async->bio = bio;
	async->mirror_num = mirror_num;
C
Chris Mason 已提交
846 847 848
	async->submit_bio_start = submit_bio_start;
	async->submit_bio_done = submit_bio_done;

849
	btrfs_init_work(&async->work, btrfs_worker_helper, run_one_async_start,
850
			run_one_async_done, run_one_async_free);
C
Chris Mason 已提交
851

C
Chris Mason 已提交
852
	async->bio_flags = bio_flags;
853
	async->bio_offset = bio_offset;
854

855 856
	async->error = 0;

857
	atomic_inc(&fs_info->nr_async_submits);
858

859
	if (rw & REQ_SYNC)
860
		btrfs_set_work_high_priority(&async->work);
861

862
	btrfs_queue_work(fs_info->workers, &async->work);
863

C
Chris Mason 已提交
864
	while (atomic_read(&fs_info->async_submit_draining) &&
865 866 867 868 869
	      atomic_read(&fs_info->nr_async_submits)) {
		wait_event(fs_info->async_submit_wait,
			   (atomic_read(&fs_info->nr_async_submits) == 0));
	}

870 871 872
	return 0;
}

873 874
static int btree_csum_one_bio(struct bio *bio)
{
875
	struct bio_vec *bvec;
876
	struct btrfs_root *root;
877
	int i, ret = 0;
878

879
	bio_for_each_segment_all(bvec, bio, i) {
880
		root = BTRFS_I(bvec->bv_page->mapping->host)->root;
881
		ret = csum_dirty_buffer(root->fs_info, bvec->bv_page);
882 883
		if (ret)
			break;
884
	}
885

886
	return ret;
887 888
}

C
Chris Mason 已提交
889 890
static int __btree_submit_bio_start(struct inode *inode, int rw,
				    struct bio *bio, int mirror_num,
891 892
				    unsigned long bio_flags,
				    u64 bio_offset)
893
{
894 895
	/*
	 * when we're called for a write, we're already in the async
896
	 * submission context.  Just jump into btrfs_map_bio
897
	 */
898
	return btree_csum_one_bio(bio);
C
Chris Mason 已提交
899
}
900

C
Chris Mason 已提交
901
static int __btree_submit_bio_done(struct inode *inode, int rw, struct bio *bio,
902 903
				 int mirror_num, unsigned long bio_flags,
				 u64 bio_offset)
C
Chris Mason 已提交
904
{
905 906
	int ret;

907
	/*
C
Chris Mason 已提交
908 909
	 * when we're called for a write, we're already in the async
	 * submission context.  Just jump into btrfs_map_bio
910
	 */
911
	ret = btrfs_map_bio(BTRFS_I(inode)->root, rw, bio, mirror_num, 1);
912 913 914 915
	if (ret) {
		bio->bi_error = ret;
		bio_endio(bio);
	}
916
	return ret;
917 918
}

919 920 921 922 923 924 925 926 927 928 929
static int check_async_write(struct inode *inode, unsigned long bio_flags)
{
	if (bio_flags & EXTENT_BIO_TREE_LOG)
		return 0;
#ifdef CONFIG_X86
	if (cpu_has_xmm4_2)
		return 0;
#endif
	return 1;
}

930
static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
931 932
				 int mirror_num, unsigned long bio_flags,
				 u64 bio_offset)
933
{
934
	int async = check_async_write(inode, bio_flags);
935 936
	int ret;

937
	if (!(rw & REQ_WRITE)) {
C
Chris Mason 已提交
938 939 940 941
		/*
		 * called for a read, do the setup so that checksum validation
		 * can happen in the async kernel threads
		 */
942
		ret = btrfs_bio_wq_end_io(BTRFS_I(inode)->root->fs_info,
943
					  bio, BTRFS_WQ_ENDIO_METADATA);
944
		if (ret)
945 946 947
			goto out_w_error;
		ret = btrfs_map_bio(BTRFS_I(inode)->root, rw, bio,
				    mirror_num, 0);
948 949 950
	} else if (!async) {
		ret = btree_csum_one_bio(bio);
		if (ret)
951 952 953 954 955 956 957 958 959 960 961 962 963
			goto out_w_error;
		ret = btrfs_map_bio(BTRFS_I(inode)->root, rw, bio,
				    mirror_num, 0);
	} else {
		/*
		 * kthread helpers are used to submit writes so that
		 * checksumming can happen in parallel across all CPUs
		 */
		ret = btrfs_wq_submit_bio(BTRFS_I(inode)->root->fs_info,
					  inode, rw, bio, mirror_num, 0,
					  bio_offset,
					  __btree_submit_bio_start,
					  __btree_submit_bio_done);
964
	}
965

966 967 968 969
	if (ret)
		goto out_w_error;
	return 0;

970
out_w_error:
971 972
	bio->bi_error = ret;
	bio_endio(bio);
973
	return ret;
974 975
}

J
Jan Beulich 已提交
976
#ifdef CONFIG_MIGRATION
977
static int btree_migratepage(struct address_space *mapping,
978 979
			struct page *newpage, struct page *page,
			enum migrate_mode mode)
980 981 982 983 984 985 986 987 988 989 990 991 992 993
{
	/*
	 * we can't safely write a btree page from here,
	 * we haven't done the locking hook
	 */
	if (PageDirty(page))
		return -EAGAIN;
	/*
	 * Buffers may be managed in a filesystem specific way.
	 * We must have no buffers or drop them.
	 */
	if (page_has_private(page) &&
	    !try_to_release_page(page, GFP_KERNEL))
		return -EAGAIN;
994
	return migrate_page(mapping, newpage, page, mode);
995
}
J
Jan Beulich 已提交
996
#endif
997

998 999 1000 1001

static int btree_writepages(struct address_space *mapping,
			    struct writeback_control *wbc)
{
1002 1003 1004
	struct btrfs_fs_info *fs_info;
	int ret;

1005
	if (wbc->sync_mode == WB_SYNC_NONE) {
1006 1007 1008 1009

		if (wbc->for_kupdate)
			return 0;

1010
		fs_info = BTRFS_I(mapping->host)->root->fs_info;
1011
		/* this is a bit racy, but that's ok */
1012 1013 1014
		ret = percpu_counter_compare(&fs_info->dirty_metadata_bytes,
					     BTRFS_DIRTY_METADATA_THRESH);
		if (ret < 0)
1015 1016
			return 0;
	}
1017
	return btree_write_cache_pages(mapping, wbc);
1018 1019
}

1020
static int btree_readpage(struct file *file, struct page *page)
1021
{
1022 1023
	struct extent_io_tree *tree;
	tree = &BTRFS_I(page->mapping->host)->io_tree;
1024
	return extent_read_full_page(tree, page, btree_get_extent, 0);
1025
}
C
Chris Mason 已提交
1026

1027
static int btree_releasepage(struct page *page, gfp_t gfp_flags)
1028
{
1029
	if (PageWriteback(page) || PageDirty(page))
C
Chris Mason 已提交
1030
		return 0;
1031

1032
	return try_release_extent_buffer(page);
1033 1034
}

1035 1036
static void btree_invalidatepage(struct page *page, unsigned int offset,
				 unsigned int length)
1037
{
1038 1039
	struct extent_io_tree *tree;
	tree = &BTRFS_I(page->mapping->host)->io_tree;
1040 1041
	extent_invalidatepage(tree, page, offset);
	btree_releasepage(page, GFP_NOFS);
1042
	if (PagePrivate(page)) {
1043 1044 1045
		btrfs_warn(BTRFS_I(page->mapping->host)->root->fs_info,
			   "page private not zero on page %llu",
			   (unsigned long long)page_offset(page));
1046 1047 1048 1049
		ClearPagePrivate(page);
		set_page_private(page, 0);
		page_cache_release(page);
	}
1050 1051
}

1052 1053
static int btree_set_page_dirty(struct page *page)
{
1054
#ifdef DEBUG
1055 1056 1057 1058 1059 1060 1061 1062
	struct extent_buffer *eb;

	BUG_ON(!PagePrivate(page));
	eb = (struct extent_buffer *)page->private;
	BUG_ON(!eb);
	BUG_ON(!test_bit(EXTENT_BUFFER_DIRTY, &eb->bflags));
	BUG_ON(!atomic_read(&eb->refs));
	btrfs_assert_tree_locked(eb);
1063
#endif
1064 1065 1066
	return __set_page_dirty_nobuffers(page);
}

1067
static const struct address_space_operations btree_aops = {
1068
	.readpage	= btree_readpage,
1069
	.writepages	= btree_writepages,
1070 1071
	.releasepage	= btree_releasepage,
	.invalidatepage = btree_invalidatepage,
1072
#ifdef CONFIG_MIGRATION
1073
	.migratepage	= btree_migratepage,
1074
#endif
1075
	.set_page_dirty = btree_set_page_dirty,
1076 1077
};

1078
void readahead_tree_block(struct btrfs_root *root, u64 bytenr)
C
Chris Mason 已提交
1079
{
1080 1081
	struct extent_buffer *buf = NULL;
	struct inode *btree_inode = root->fs_info->btree_inode;
C
Chris Mason 已提交
1082

1083
	buf = btrfs_find_create_tree_block(root, bytenr);
1084
	if (!buf)
1085
		return;
1086
	read_extent_buffer_pages(&BTRFS_I(btree_inode)->io_tree,
1087
				 buf, 0, WAIT_NONE, btree_get_extent, 0);
1088
	free_extent_buffer(buf);
C
Chris Mason 已提交
1089 1090
}

1091
int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr,
1092 1093 1094 1095 1096 1097 1098
			 int mirror_num, struct extent_buffer **eb)
{
	struct extent_buffer *buf = NULL;
	struct inode *btree_inode = root->fs_info->btree_inode;
	struct extent_io_tree *io_tree = &BTRFS_I(btree_inode)->io_tree;
	int ret;

1099
	buf = btrfs_find_create_tree_block(root, bytenr);
1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114
	if (!buf)
		return 0;

	set_bit(EXTENT_BUFFER_READAHEAD, &buf->bflags);

	ret = read_extent_buffer_pages(io_tree, buf, 0, WAIT_PAGE_LOCK,
				       btree_get_extent, mirror_num);
	if (ret) {
		free_extent_buffer(buf);
		return ret;
	}

	if (test_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags)) {
		free_extent_buffer(buf);
		return -EIO;
1115
	} else if (extent_buffer_uptodate(buf)) {
1116 1117 1118 1119 1120 1121 1122
		*eb = buf;
	} else {
		free_extent_buffer(buf);
	}
	return 0;
}

1123
struct extent_buffer *btrfs_find_tree_block(struct btrfs_fs_info *fs_info,
1124
					    u64 bytenr)
1125
{
1126
	return find_extent_buffer(fs_info, bytenr);
1127 1128 1129
}

struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root,
1130
						 u64 bytenr)
1131
{
1132
	if (btrfs_test_is_dummy_root(root))
1133 1134
		return alloc_test_extent_buffer(root->fs_info, bytenr);
	return alloc_extent_buffer(root->fs_info, bytenr);
1135 1136 1137
}


1138 1139
int btrfs_write_tree_block(struct extent_buffer *buf)
{
1140
	return filemap_fdatawrite_range(buf->pages[0]->mapping, buf->start,
1141
					buf->start + buf->len - 1);
1142 1143 1144 1145
}

int btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
{
1146
	return filemap_fdatawait_range(buf->pages[0]->mapping,
1147
				       buf->start, buf->start + buf->len - 1);
1148 1149
}

1150
struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
1151
				      u64 parent_transid)
1152 1153 1154 1155
{
	struct extent_buffer *buf = NULL;
	int ret;

1156
	buf = btrfs_find_create_tree_block(root, bytenr);
1157
	if (!buf)
1158
		return ERR_PTR(-ENOMEM);
1159

1160
	ret = btree_read_extent_buffer_pages(root, buf, 0, parent_transid);
1161 1162
	if (ret) {
		free_extent_buffer(buf);
1163
		return ERR_PTR(ret);
1164
	}
1165
	return buf;
1166

1167 1168
}

1169 1170
void clean_tree_block(struct btrfs_trans_handle *trans,
		      struct btrfs_fs_info *fs_info,
1171
		      struct extent_buffer *buf)
1172
{
1173
	if (btrfs_header_generation(buf) ==
1174
	    fs_info->running_transaction->transid) {
1175
		btrfs_assert_tree_locked(buf);
1176

1177
		if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &buf->bflags)) {
1178 1179 1180
			__percpu_counter_add(&fs_info->dirty_metadata_bytes,
					     -buf->len,
					     fs_info->dirty_metadata_batch);
1181 1182 1183 1184
			/* ugh, clear_extent_buffer_dirty needs to lock the page */
			btrfs_set_lock_blocking(buf);
			clear_extent_buffer_dirty(buf);
		}
1185
	}
1186 1187
}

1188 1189 1190 1191 1192 1193 1194 1195 1196
static struct btrfs_subvolume_writers *btrfs_alloc_subvolume_writers(void)
{
	struct btrfs_subvolume_writers *writers;
	int ret;

	writers = kmalloc(sizeof(*writers), GFP_NOFS);
	if (!writers)
		return ERR_PTR(-ENOMEM);

1197
	ret = percpu_counter_init(&writers->counter, 0, GFP_KERNEL);
1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213
	if (ret < 0) {
		kfree(writers);
		return ERR_PTR(ret);
	}

	init_waitqueue_head(&writers->wait);
	return writers;
}

static void
btrfs_free_subvolume_writers(struct btrfs_subvolume_writers *writers)
{
	percpu_counter_destroy(&writers->counter);
	kfree(writers);
}

1214 1215
static void __setup_root(u32 nodesize, u32 sectorsize, u32 stripesize,
			 struct btrfs_root *root, struct btrfs_fs_info *fs_info,
1216
			 u64 objectid)
1217
{
C
Chris Mason 已提交
1218
	root->node = NULL;
1219
	root->commit_root = NULL;
1220 1221
	root->sectorsize = sectorsize;
	root->nodesize = nodesize;
1222
	root->stripesize = stripesize;
1223
	root->state = 0;
1224
	root->orphan_cleanup_state = 0;
1225

1226 1227
	root->objectid = objectid;
	root->last_trans = 0;
1228
	root->highest_objectid = 0;
1229
	root->nr_delalloc_inodes = 0;
1230
	root->nr_ordered_extents = 0;
1231
	root->name = NULL;
1232
	root->inode_tree = RB_ROOT;
1233
	INIT_RADIX_TREE(&root->delayed_nodes_tree, GFP_ATOMIC);
1234
	root->block_rsv = NULL;
1235
	root->orphan_block_rsv = NULL;
1236 1237

	INIT_LIST_HEAD(&root->dirty_list);
1238
	INIT_LIST_HEAD(&root->root_list);
1239 1240
	INIT_LIST_HEAD(&root->delalloc_inodes);
	INIT_LIST_HEAD(&root->delalloc_root);
1241 1242
	INIT_LIST_HEAD(&root->ordered_extents);
	INIT_LIST_HEAD(&root->ordered_root);
1243 1244
	INIT_LIST_HEAD(&root->logged_list[0]);
	INIT_LIST_HEAD(&root->logged_list[1]);
1245
	spin_lock_init(&root->orphan_lock);
1246
	spin_lock_init(&root->inode_lock);
1247
	spin_lock_init(&root->delalloc_lock);
1248
	spin_lock_init(&root->ordered_extent_lock);
1249
	spin_lock_init(&root->accounting_lock);
1250 1251
	spin_lock_init(&root->log_extents_lock[0]);
	spin_lock_init(&root->log_extents_lock[1]);
1252
	mutex_init(&root->objectid_mutex);
1253
	mutex_init(&root->log_mutex);
1254
	mutex_init(&root->ordered_extent_mutex);
1255
	mutex_init(&root->delalloc_mutex);
Y
Yan Zheng 已提交
1256 1257 1258
	init_waitqueue_head(&root->log_writer_wait);
	init_waitqueue_head(&root->log_commit_wait[0]);
	init_waitqueue_head(&root->log_commit_wait[1]);
1259 1260
	INIT_LIST_HEAD(&root->log_ctxs[0]);
	INIT_LIST_HEAD(&root->log_ctxs[1]);
Y
Yan Zheng 已提交
1261 1262 1263
	atomic_set(&root->log_commit[0], 0);
	atomic_set(&root->log_commit[1], 0);
	atomic_set(&root->log_writers, 0);
M
Miao Xie 已提交
1264
	atomic_set(&root->log_batch, 0);
1265
	atomic_set(&root->orphan_inodes, 0);
1266
	atomic_set(&root->refs, 1);
1267
	atomic_set(&root->will_be_snapshoted, 0);
Y
Yan Zheng 已提交
1268
	root->log_transid = 0;
1269
	root->log_transid_committed = -1;
1270
	root->last_log_commit = 0;
1271 1272 1273
	if (fs_info)
		extent_io_tree_init(&root->dirty_log_pages,
				     fs_info->btree_inode->i_mapping);
C
Chris Mason 已提交
1274

1275 1276
	memset(&root->root_key, 0, sizeof(root->root_key));
	memset(&root->root_item, 0, sizeof(root->root_item));
1277
	memset(&root->defrag_progress, 0, sizeof(root->defrag_progress));
1278 1279 1280 1281
	if (fs_info)
		root->defrag_trans_start = fs_info->generation;
	else
		root->defrag_trans_start = 0;
1282
	root->root_key.objectid = objectid;
1283
	root->anon_dev = 0;
1284

1285
	spin_lock_init(&root->root_item_lock);
1286 1287
}

1288
static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info)
A
Al Viro 已提交
1289 1290 1291 1292 1293 1294 1295
{
	struct btrfs_root *root = kzalloc(sizeof(*root), GFP_NOFS);
	if (root)
		root->fs_info = fs_info;
	return root;
}

1296 1297 1298 1299 1300 1301 1302 1303 1304
#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
/* Should only be used by the testing infrastructure */
struct btrfs_root *btrfs_alloc_dummy_root(void)
{
	struct btrfs_root *root;

	root = btrfs_alloc_root(NULL);
	if (!root)
		return ERR_PTR(-ENOMEM);
1305
	__setup_root(4096, 4096, 4096, root, NULL, 1);
1306
	set_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state);
1307
	root->alloc_bytenr = 0;
1308 1309 1310 1311 1312

	return root;
}
#endif

1313 1314 1315 1316 1317 1318 1319 1320 1321
struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
				     struct btrfs_fs_info *fs_info,
				     u64 objectid)
{
	struct extent_buffer *leaf;
	struct btrfs_root *tree_root = fs_info->tree_root;
	struct btrfs_root *root;
	struct btrfs_key key;
	int ret = 0;
1322
	uuid_le uuid;
1323 1324 1325 1326 1327

	root = btrfs_alloc_root(fs_info);
	if (!root)
		return ERR_PTR(-ENOMEM);

1328 1329
	__setup_root(tree_root->nodesize, tree_root->sectorsize,
		tree_root->stripesize, root, fs_info, objectid);
1330 1331 1332 1333
	root->root_key.objectid = objectid;
	root->root_key.type = BTRFS_ROOT_ITEM_KEY;
	root->root_key.offset = 0;

1334
	leaf = btrfs_alloc_tree_block(trans, root, 0, objectid, NULL, 0, 0, 0);
1335 1336
	if (IS_ERR(leaf)) {
		ret = PTR_ERR(leaf);
1337
		leaf = NULL;
1338 1339 1340 1341 1342 1343 1344 1345 1346 1347
		goto fail;
	}

	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
	btrfs_set_header_bytenr(leaf, leaf->start);
	btrfs_set_header_generation(leaf, trans->transid);
	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
	btrfs_set_header_owner(leaf, objectid);
	root->node = leaf;

1348
	write_extent_buffer(leaf, fs_info->fsid, btrfs_header_fsid(),
1349 1350
			    BTRFS_FSID_SIZE);
	write_extent_buffer(leaf, fs_info->chunk_tree_uuid,
1351
			    btrfs_header_chunk_tree_uuid(leaf),
1352 1353 1354 1355
			    BTRFS_UUID_SIZE);
	btrfs_mark_buffer_dirty(leaf);

	root->commit_root = btrfs_root_node(root);
1356
	set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state);
1357 1358 1359 1360 1361 1362 1363 1364 1365 1366

	root->root_item.flags = 0;
	root->root_item.byte_limit = 0;
	btrfs_set_root_bytenr(&root->root_item, leaf->start);
	btrfs_set_root_generation(&root->root_item, trans->transid);
	btrfs_set_root_level(&root->root_item, 0);
	btrfs_set_root_refs(&root->root_item, 1);
	btrfs_set_root_used(&root->root_item, leaf->len);
	btrfs_set_root_last_snapshot(&root->root_item, 0);
	btrfs_set_root_dirid(&root->root_item, 0);
1367 1368
	uuid_le_gen(&uuid);
	memcpy(root->root_item.uuid, uuid.b, BTRFS_UUID_SIZE);
1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379
	root->root_item.drop_level = 0;

	key.objectid = objectid;
	key.type = BTRFS_ROOT_ITEM_KEY;
	key.offset = 0;
	ret = btrfs_insert_root(trans, tree_root, &key, &root->root_item);
	if (ret)
		goto fail;

	btrfs_tree_unlock(leaf);

1380 1381
	return root;

1382
fail:
1383 1384
	if (leaf) {
		btrfs_tree_unlock(leaf);
1385
		free_extent_buffer(root->commit_root);
1386 1387 1388
		free_extent_buffer(leaf);
	}
	kfree(root);
1389

1390
	return ERR_PTR(ret);
1391 1392
}

Y
Yan Zheng 已提交
1393 1394
static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
					 struct btrfs_fs_info *fs_info)
1395 1396 1397
{
	struct btrfs_root *root;
	struct btrfs_root *tree_root = fs_info->tree_root;
Y
Yan Zheng 已提交
1398
	struct extent_buffer *leaf;
1399

A
Al Viro 已提交
1400
	root = btrfs_alloc_root(fs_info);
1401
	if (!root)
Y
Yan Zheng 已提交
1402
		return ERR_PTR(-ENOMEM);
1403

1404 1405 1406
	__setup_root(tree_root->nodesize, tree_root->sectorsize,
		     tree_root->stripesize, root, fs_info,
		     BTRFS_TREE_LOG_OBJECTID);
1407 1408 1409 1410

	root->root_key.objectid = BTRFS_TREE_LOG_OBJECTID;
	root->root_key.type = BTRFS_ROOT_ITEM_KEY;
	root->root_key.offset = BTRFS_TREE_LOG_OBJECTID;
1411

Y
Yan Zheng 已提交
1412
	/*
1413 1414
	 * DON'T set REF_COWS for log trees
	 *
Y
Yan Zheng 已提交
1415 1416 1417 1418 1419
	 * log trees do not get reference counted because they go away
	 * before a real commit is actually done.  They do store pointers
	 * to file data extents, and those reference counts still get
	 * updated (along with back refs to the log tree).
	 */
1420

1421 1422
	leaf = btrfs_alloc_tree_block(trans, root, 0, BTRFS_TREE_LOG_OBJECTID,
			NULL, 0, 0, 0);
Y
Yan Zheng 已提交
1423 1424 1425 1426
	if (IS_ERR(leaf)) {
		kfree(root);
		return ERR_CAST(leaf);
	}
1427

1428 1429 1430 1431 1432
	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
	btrfs_set_header_bytenr(leaf, leaf->start);
	btrfs_set_header_generation(leaf, trans->transid);
	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
	btrfs_set_header_owner(leaf, BTRFS_TREE_LOG_OBJECTID);
Y
Yan Zheng 已提交
1433
	root->node = leaf;
1434 1435

	write_extent_buffer(root->node, root->fs_info->fsid,
1436
			    btrfs_header_fsid(), BTRFS_FSID_SIZE);
1437 1438
	btrfs_mark_buffer_dirty(root->node);
	btrfs_tree_unlock(root->node);
Y
Yan Zheng 已提交
1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468
	return root;
}

int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
			     struct btrfs_fs_info *fs_info)
{
	struct btrfs_root *log_root;

	log_root = alloc_log_tree(trans, fs_info);
	if (IS_ERR(log_root))
		return PTR_ERR(log_root);
	WARN_ON(fs_info->log_root_tree);
	fs_info->log_root_tree = log_root;
	return 0;
}

int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
		       struct btrfs_root *root)
{
	struct btrfs_root *log_root;
	struct btrfs_inode_item *inode_item;

	log_root = alloc_log_tree(trans, root->fs_info);
	if (IS_ERR(log_root))
		return PTR_ERR(log_root);

	log_root->last_trans = trans->transid;
	log_root->root_key.offset = root->root_key.objectid;

	inode_item = &log_root->root_item.inode;
1469 1470 1471
	btrfs_set_stack_inode_generation(inode_item, 1);
	btrfs_set_stack_inode_size(inode_item, 3);
	btrfs_set_stack_inode_nlink(inode_item, 1);
1472
	btrfs_set_stack_inode_nbytes(inode_item, root->nodesize);
1473
	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
Y
Yan Zheng 已提交
1474

1475
	btrfs_set_root_node(&log_root->root_item, log_root->node);
Y
Yan Zheng 已提交
1476 1477 1478 1479

	WARN_ON(root->log_root);
	root->log_root = log_root;
	root->log_transid = 0;
1480
	root->log_transid_committed = -1;
1481
	root->last_log_commit = 0;
1482 1483 1484
	return 0;
}

1485 1486
static struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
					       struct btrfs_key *key)
1487 1488 1489
{
	struct btrfs_root *root;
	struct btrfs_fs_info *fs_info = tree_root->fs_info;
1490
	struct btrfs_path *path;
1491
	u64 generation;
1492
	int ret;
1493

1494 1495
	path = btrfs_alloc_path();
	if (!path)
1496
		return ERR_PTR(-ENOMEM);
1497 1498 1499 1500 1501

	root = btrfs_alloc_root(fs_info);
	if (!root) {
		ret = -ENOMEM;
		goto alloc_fail;
1502 1503
	}

1504 1505
	__setup_root(tree_root->nodesize, tree_root->sectorsize,
		tree_root->stripesize, root, fs_info, key->objectid);
1506

1507 1508
	ret = btrfs_find_root(tree_root, key, path,
			      &root->root_item, &root->root_key);
1509
	if (ret) {
1510 1511
		if (ret > 0)
			ret = -ENOENT;
1512
		goto find_fail;
1513
	}
1514

1515
	generation = btrfs_root_generation(&root->root_item);
1516
	root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item),
1517
				     generation);
1518 1519
	if (IS_ERR(root->node)) {
		ret = PTR_ERR(root->node);
1520 1521 1522
		goto find_fail;
	} else if (!btrfs_buffer_uptodate(root->node, generation, 0)) {
		ret = -EIO;
1523 1524
		free_extent_buffer(root->node);
		goto find_fail;
1525
	}
1526
	root->commit_root = btrfs_root_node(root);
1527
out:
1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547
	btrfs_free_path(path);
	return root;

find_fail:
	kfree(root);
alloc_fail:
	root = ERR_PTR(ret);
	goto out;
}

struct btrfs_root *btrfs_read_fs_root(struct btrfs_root *tree_root,
				      struct btrfs_key *location)
{
	struct btrfs_root *root;

	root = btrfs_read_tree_root(tree_root, location);
	if (IS_ERR(root))
		return root;

	if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
1548
		set_bit(BTRFS_ROOT_REF_COWS, &root->state);
1549 1550
		btrfs_check_and_init_root_item(&root->root_item);
	}
1551

1552 1553 1554
	return root;
}

1555 1556 1557
int btrfs_init_fs_root(struct btrfs_root *root)
{
	int ret;
1558
	struct btrfs_subvolume_writers *writers;
1559 1560 1561 1562 1563 1564 1565 1566 1567

	root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS);
	root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned),
					GFP_NOFS);
	if (!root->free_ino_pinned || !root->free_ino_ctl) {
		ret = -ENOMEM;
		goto fail;
	}

1568 1569 1570 1571 1572 1573 1574
	writers = btrfs_alloc_subvolume_writers();
	if (IS_ERR(writers)) {
		ret = PTR_ERR(writers);
		goto fail;
	}
	root->subv_writers = writers;

1575
	btrfs_init_free_ino_ctl(root);
1576 1577
	spin_lock_init(&root->ino_cache_lock);
	init_waitqueue_head(&root->ino_cache_wait);
1578 1579 1580

	ret = get_anon_bdev(&root->anon_dev);
	if (ret)
1581
		goto free_writers;
1582
	return 0;
1583 1584 1585

free_writers:
	btrfs_free_subvolume_writers(root->subv_writers);
1586 1587 1588 1589 1590 1591
fail:
	kfree(root->free_ino_ctl);
	kfree(root->free_ino_pinned);
	return ret;
}

1592 1593
static struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
					       u64 root_id)
1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617
{
	struct btrfs_root *root;

	spin_lock(&fs_info->fs_roots_radix_lock);
	root = radix_tree_lookup(&fs_info->fs_roots_radix,
				 (unsigned long)root_id);
	spin_unlock(&fs_info->fs_roots_radix_lock);
	return root;
}

int btrfs_insert_fs_root(struct btrfs_fs_info *fs_info,
			 struct btrfs_root *root)
{
	int ret;

	ret = radix_tree_preload(GFP_NOFS & ~__GFP_HIGHMEM);
	if (ret)
		return ret;

	spin_lock(&fs_info->fs_roots_radix_lock);
	ret = radix_tree_insert(&fs_info->fs_roots_radix,
				(unsigned long)root->root_key.objectid,
				root);
	if (ret == 0)
1618
		set_bit(BTRFS_ROOT_IN_RADIX, &root->state);
1619 1620 1621 1622 1623 1624
	spin_unlock(&fs_info->fs_roots_radix_lock);
	radix_tree_preload_end();

	return ret;
}

1625 1626 1627
struct btrfs_root *btrfs_get_fs_root(struct btrfs_fs_info *fs_info,
				     struct btrfs_key *location,
				     bool check_ref)
1628 1629
{
	struct btrfs_root *root;
1630
	struct btrfs_path *path;
1631
	struct btrfs_key key;
1632 1633
	int ret;

1634 1635 1636 1637
	if (location->objectid == BTRFS_ROOT_TREE_OBJECTID)
		return fs_info->tree_root;
	if (location->objectid == BTRFS_EXTENT_TREE_OBJECTID)
		return fs_info->extent_root;
1638 1639 1640 1641
	if (location->objectid == BTRFS_CHUNK_TREE_OBJECTID)
		return fs_info->chunk_root;
	if (location->objectid == BTRFS_DEV_TREE_OBJECTID)
		return fs_info->dev_root;
1642 1643
	if (location->objectid == BTRFS_CSUM_TREE_OBJECTID)
		return fs_info->csum_root;
1644 1645 1646
	if (location->objectid == BTRFS_QUOTA_TREE_OBJECTID)
		return fs_info->quota_root ? fs_info->quota_root :
					     ERR_PTR(-ENOENT);
1647 1648 1649
	if (location->objectid == BTRFS_UUID_TREE_OBJECTID)
		return fs_info->uuid_root ? fs_info->uuid_root :
					    ERR_PTR(-ENOENT);
1650
again:
1651
	root = btrfs_lookup_fs_root(fs_info, location->objectid);
1652
	if (root) {
1653
		if (check_ref && btrfs_root_refs(&root->root_item) == 0)
1654
			return ERR_PTR(-ENOENT);
1655
		return root;
1656
	}
1657

1658
	root = btrfs_read_fs_root(fs_info->tree_root, location);
1659 1660
	if (IS_ERR(root))
		return root;
1661

1662
	if (check_ref && btrfs_root_refs(&root->root_item) == 0) {
1663
		ret = -ENOENT;
1664
		goto fail;
1665
	}
1666

1667
	ret = btrfs_init_fs_root(root);
1668 1669
	if (ret)
		goto fail;
1670

1671 1672 1673 1674 1675
	path = btrfs_alloc_path();
	if (!path) {
		ret = -ENOMEM;
		goto fail;
	}
1676 1677 1678 1679 1680
	key.objectid = BTRFS_ORPHAN_OBJECTID;
	key.type = BTRFS_ORPHAN_ITEM_KEY;
	key.offset = location->objectid;

	ret = btrfs_search_slot(NULL, fs_info->tree_root, &key, path, 0, 0);
1681
	btrfs_free_path(path);
1682 1683 1684
	if (ret < 0)
		goto fail;
	if (ret == 0)
1685
		set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state);
1686

1687
	ret = btrfs_insert_fs_root(fs_info, root);
1688
	if (ret) {
1689 1690 1691 1692 1693
		if (ret == -EEXIST) {
			free_fs_root(root);
			goto again;
		}
		goto fail;
1694
	}
1695
	return root;
1696 1697 1698
fail:
	free_fs_root(root);
	return ERR_PTR(ret);
1699 1700
}

C
Chris Mason 已提交
1701 1702 1703 1704 1705 1706
static int btrfs_congested_fn(void *congested_data, int bdi_bits)
{
	struct btrfs_fs_info *info = (struct btrfs_fs_info *)congested_data;
	int ret = 0;
	struct btrfs_device *device;
	struct backing_dev_info *bdi;
C
Chris Mason 已提交
1707

1708 1709
	rcu_read_lock();
	list_for_each_entry_rcu(device, &info->fs_devices->devices, dev_list) {
1710 1711
		if (!device->bdev)
			continue;
C
Chris Mason 已提交
1712
		bdi = blk_get_backing_dev_info(device->bdev);
1713
		if (bdi_congested(bdi, bdi_bits)) {
C
Chris Mason 已提交
1714 1715 1716 1717
			ret = 1;
			break;
		}
	}
1718
	rcu_read_unlock();
C
Chris Mason 已提交
1719 1720 1721 1722 1723
	return ret;
}

static int setup_bdi(struct btrfs_fs_info *info, struct backing_dev_info *bdi)
{
1724 1725
	int err;

1726
	err = bdi_setup_and_register(bdi, "btrfs");
1727 1728 1729
	if (err)
		return err;

1730
	bdi->ra_pages = VM_MAX_READAHEAD * 1024 / PAGE_CACHE_SIZE;
C
Chris Mason 已提交
1731 1732
	bdi->congested_fn	= btrfs_congested_fn;
	bdi->congested_data	= info;
1733
	bdi->capabilities |= BDI_CAP_CGROUP_WRITEBACK;
C
Chris Mason 已提交
1734 1735 1736
	return 0;
}

1737 1738 1739 1740 1741
/*
 * called by the kthread helper functions to finally call the bio end_io
 * functions.  This is where read checksum verification actually happens
 */
static void end_workqueue_fn(struct btrfs_work *work)
1742 1743
{
	struct bio *bio;
1744
	struct btrfs_end_io_wq *end_io_wq;
1745

1746
	end_io_wq = container_of(work, struct btrfs_end_io_wq, work);
1747
	bio = end_io_wq->bio;
1748

1749
	bio->bi_error = end_io_wq->error;
1750 1751
	bio->bi_private = end_io_wq->private;
	bio->bi_end_io = end_io_wq->end_io;
1752
	kmem_cache_free(btrfs_end_io_wq_cache, end_io_wq);
1753
	bio_endio(bio);
1754 1755
}

1756 1757 1758
static int cleaner_kthread(void *arg)
{
	struct btrfs_root *root = arg;
1759
	int again;
1760
	struct btrfs_trans_handle *trans;
1761 1762

	do {
1763
		again = 0;
1764

1765
		/* Make the cleaner go to sleep early. */
1766
		if (btrfs_need_cleaner_sleep(root))
1767 1768 1769 1770 1771
			goto sleep;

		if (!mutex_trylock(&root->fs_info->cleaner_mutex))
			goto sleep;

1772 1773 1774 1775
		/*
		 * Avoid the problem that we change the status of the fs
		 * during the above check and trylock.
		 */
1776
		if (btrfs_need_cleaner_sleep(root)) {
1777 1778
			mutex_unlock(&root->fs_info->cleaner_mutex);
			goto sleep;
1779
		}
1780

1781 1782 1783 1784 1785
		btrfs_run_delayed_iputs(root);
		again = btrfs_clean_one_deleted_snapshot(root);
		mutex_unlock(&root->fs_info->cleaner_mutex);

		/*
1786 1787
		 * The defragger has dealt with the R/O remount and umount,
		 * needn't do anything special here.
1788 1789
		 */
		btrfs_run_defrag_inodes(root->fs_info);
1790 1791 1792 1793 1794 1795 1796 1797 1798 1799

		/*
		 * Acquires fs_info->delete_unused_bgs_mutex to avoid racing
		 * with relocation (btrfs_relocate_chunk) and relocation
		 * acquires fs_info->cleaner_mutex (btrfs_relocate_block_group)
		 * after acquiring fs_info->delete_unused_bgs_mutex. So we
		 * can't hold, nor need to, fs_info->cleaner_mutex when deleting
		 * unused block groups.
		 */
		btrfs_delete_unused_bgs(root->fs_info);
1800
sleep:
D
David Sterba 已提交
1801
		if (!try_to_freeze() && !again) {
1802
			set_current_state(TASK_INTERRUPTIBLE);
1803 1804
			if (!kthread_should_stop())
				schedule();
1805 1806 1807
			__set_current_state(TASK_RUNNING);
		}
	} while (!kthread_should_stop());
1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835

	/*
	 * Transaction kthread is stopped before us and wakes us up.
	 * However we might have started a new transaction and COWed some
	 * tree blocks when deleting unused block groups for example. So
	 * make sure we commit the transaction we started to have a clean
	 * shutdown when evicting the btree inode - if it has dirty pages
	 * when we do the final iput() on it, eviction will trigger a
	 * writeback for it which will fail with null pointer dereferences
	 * since work queues and other resources were already released and
	 * destroyed by the time the iput/eviction/writeback is made.
	 */
	trans = btrfs_attach_transaction(root);
	if (IS_ERR(trans)) {
		if (PTR_ERR(trans) != -ENOENT)
			btrfs_err(root->fs_info,
				  "cleaner transaction attach returned %ld",
				  PTR_ERR(trans));
	} else {
		int ret;

		ret = btrfs_commit_transaction(trans, root);
		if (ret)
			btrfs_err(root->fs_info,
				  "cleaner open transaction commit returned %d",
				  ret);
	}

1836 1837 1838 1839 1840 1841 1842 1843
	return 0;
}

static int transaction_kthread(void *arg)
{
	struct btrfs_root *root = arg;
	struct btrfs_trans_handle *trans;
	struct btrfs_transaction *cur;
1844
	u64 transid;
1845 1846
	unsigned long now;
	unsigned long delay;
1847
	bool cannot_commit;
1848 1849

	do {
1850
		cannot_commit = false;
1851
		delay = HZ * root->fs_info->commit_interval;
1852 1853
		mutex_lock(&root->fs_info->transaction_kthread_mutex);

J
Josef Bacik 已提交
1854
		spin_lock(&root->fs_info->trans_lock);
1855 1856
		cur = root->fs_info->running_transaction;
		if (!cur) {
J
Josef Bacik 已提交
1857
			spin_unlock(&root->fs_info->trans_lock);
1858 1859
			goto sleep;
		}
Y
Yan Zheng 已提交
1860

1861
		now = get_seconds();
1862
		if (cur->state < TRANS_STATE_BLOCKED &&
1863 1864
		    (now < cur->start_time ||
		     now - cur->start_time < root->fs_info->commit_interval)) {
J
Josef Bacik 已提交
1865
			spin_unlock(&root->fs_info->trans_lock);
1866 1867 1868
			delay = HZ * 5;
			goto sleep;
		}
1869
		transid = cur->transid;
J
Josef Bacik 已提交
1870
		spin_unlock(&root->fs_info->trans_lock);
1871

1872
		/* If the file system is aborted, this will always fail. */
1873
		trans = btrfs_attach_transaction(root);
1874
		if (IS_ERR(trans)) {
1875 1876
			if (PTR_ERR(trans) != -ENOENT)
				cannot_commit = true;
1877
			goto sleep;
1878
		}
1879
		if (transid == trans->transid) {
1880
			btrfs_commit_transaction(trans, root);
1881 1882 1883
		} else {
			btrfs_end_transaction(trans, root);
		}
1884 1885 1886 1887
sleep:
		wake_up_process(root->fs_info->cleaner_kthread);
		mutex_unlock(&root->fs_info->transaction_kthread_mutex);

J
Josef Bacik 已提交
1888 1889 1890
		if (unlikely(test_bit(BTRFS_FS_STATE_ERROR,
				      &root->fs_info->fs_state)))
			btrfs_cleanup_transaction(root);
1891
		if (!try_to_freeze()) {
1892
			set_current_state(TASK_INTERRUPTIBLE);
1893
			if (!kthread_should_stop() &&
1894 1895
			    (!btrfs_transaction_blocked(root->fs_info) ||
			     cannot_commit))
1896
				schedule_timeout(delay);
1897 1898 1899 1900 1901 1902
			__set_current_state(TASK_RUNNING);
		}
	} while (!kthread_should_stop());
	return 0;
}

C
Chris Mason 已提交
1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
/*
 * this will find the highest generation in the array of
 * root backups.  The index of the highest array is returned,
 * or -1 if we can't find anything.
 *
 * We check to make sure the array is valid by comparing the
 * generation of the latest  root in the array with the generation
 * in the super block.  If they don't match we pitch it.
 */
static int find_newest_super_backup(struct btrfs_fs_info *info, u64 newest_gen)
{
	u64 cur;
	int newest_index = -1;
	struct btrfs_root_backup *root_backup;
	int i;

	for (i = 0; i < BTRFS_NUM_BACKUP_ROOTS; i++) {
		root_backup = info->super_copy->super_roots + i;
		cur = btrfs_backup_tree_root_gen(root_backup);
		if (cur == newest_gen)
			newest_index = i;
	}

	/* check to see if we actually wrapped around */
	if (newest_index == BTRFS_NUM_BACKUP_ROOTS - 1) {
		root_backup = info->super_copy->super_roots;
		cur = btrfs_backup_tree_root_gen(root_backup);
		if (cur == newest_gen)
			newest_index = 0;
	}
	return newest_index;
}


/*
 * find the oldest backup so we know where to store new entries
 * in the backup array.  This will set the backup_root_index
 * field in the fs_info struct
 */
static void find_oldest_super_backup(struct btrfs_fs_info *info,
				     u64 newest_gen)
{
	int newest_index = -1;

	newest_index = find_newest_super_backup(info, newest_gen);
	/* if there was garbage in there, just move along */
	if (newest_index == -1) {
		info->backup_root_index = 0;
	} else {
		info->backup_root_index = (newest_index + 1) % BTRFS_NUM_BACKUP_ROOTS;
	}
}

/*
 * copy all the root pointers into the super backup array.
 * this will bump the backup pointer by one when it is
 * done
 */
static void backup_super_roots(struct btrfs_fs_info *info)
{
	int next_backup;
	struct btrfs_root_backup *root_backup;
	int last_backup;

	next_backup = info->backup_root_index;
	last_backup = (next_backup + BTRFS_NUM_BACKUP_ROOTS - 1) %
		BTRFS_NUM_BACKUP_ROOTS;

	/*
	 * just overwrite the last backup if we're at the same generation
	 * this happens only at umount
	 */
	root_backup = info->super_for_commit->super_roots + last_backup;
	if (btrfs_backup_tree_root_gen(root_backup) ==
	    btrfs_header_generation(info->tree_root->node))
		next_backup = last_backup;

	root_backup = info->super_for_commit->super_roots + next_backup;

	/*
	 * make sure all of our padding and empty slots get zero filled
	 * regardless of which ones we use today
	 */
	memset(root_backup, 0, sizeof(*root_backup));

	info->backup_root_index = (next_backup + 1) % BTRFS_NUM_BACKUP_ROOTS;

	btrfs_set_backup_tree_root(root_backup, info->tree_root->node->start);
	btrfs_set_backup_tree_root_gen(root_backup,
			       btrfs_header_generation(info->tree_root->node));

	btrfs_set_backup_tree_root_level(root_backup,
			       btrfs_header_level(info->tree_root->node));

	btrfs_set_backup_chunk_root(root_backup, info->chunk_root->node->start);
	btrfs_set_backup_chunk_root_gen(root_backup,
			       btrfs_header_generation(info->chunk_root->node));
	btrfs_set_backup_chunk_root_level(root_backup,
			       btrfs_header_level(info->chunk_root->node));

	btrfs_set_backup_extent_root(root_backup, info->extent_root->node->start);
	btrfs_set_backup_extent_root_gen(root_backup,
			       btrfs_header_generation(info->extent_root->node));
	btrfs_set_backup_extent_root_level(root_backup,
			       btrfs_header_level(info->extent_root->node));

2009 2010 2011 2012 2013 2014 2015 2016
	/*
	 * we might commit during log recovery, which happens before we set
	 * the fs_root.  Make sure it is valid before we fill it in.
	 */
	if (info->fs_root && info->fs_root->node) {
		btrfs_set_backup_fs_root(root_backup,
					 info->fs_root->node->start);
		btrfs_set_backup_fs_root_gen(root_backup,
C
Chris Mason 已提交
2017
			       btrfs_header_generation(info->fs_root->node));
2018
		btrfs_set_backup_fs_root_level(root_backup,
C
Chris Mason 已提交
2019
			       btrfs_header_level(info->fs_root->node));
2020
	}
C
Chris Mason 已提交
2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101

	btrfs_set_backup_dev_root(root_backup, info->dev_root->node->start);
	btrfs_set_backup_dev_root_gen(root_backup,
			       btrfs_header_generation(info->dev_root->node));
	btrfs_set_backup_dev_root_level(root_backup,
				       btrfs_header_level(info->dev_root->node));

	btrfs_set_backup_csum_root(root_backup, info->csum_root->node->start);
	btrfs_set_backup_csum_root_gen(root_backup,
			       btrfs_header_generation(info->csum_root->node));
	btrfs_set_backup_csum_root_level(root_backup,
			       btrfs_header_level(info->csum_root->node));

	btrfs_set_backup_total_bytes(root_backup,
			     btrfs_super_total_bytes(info->super_copy));
	btrfs_set_backup_bytes_used(root_backup,
			     btrfs_super_bytes_used(info->super_copy));
	btrfs_set_backup_num_devices(root_backup,
			     btrfs_super_num_devices(info->super_copy));

	/*
	 * if we don't copy this out to the super_copy, it won't get remembered
	 * for the next commit
	 */
	memcpy(&info->super_copy->super_roots,
	       &info->super_for_commit->super_roots,
	       sizeof(*root_backup) * BTRFS_NUM_BACKUP_ROOTS);
}

/*
 * this copies info out of the root backup array and back into
 * the in-memory super block.  It is meant to help iterate through
 * the array, so you send it the number of backups you've already
 * tried and the last backup index you used.
 *
 * this returns -1 when it has tried all the backups
 */
static noinline int next_root_backup(struct btrfs_fs_info *info,
				     struct btrfs_super_block *super,
				     int *num_backups_tried, int *backup_index)
{
	struct btrfs_root_backup *root_backup;
	int newest = *backup_index;

	if (*num_backups_tried == 0) {
		u64 gen = btrfs_super_generation(super);

		newest = find_newest_super_backup(info, gen);
		if (newest == -1)
			return -1;

		*backup_index = newest;
		*num_backups_tried = 1;
	} else if (*num_backups_tried == BTRFS_NUM_BACKUP_ROOTS) {
		/* we've tried all the backups, all done */
		return -1;
	} else {
		/* jump to the next oldest backup */
		newest = (*backup_index + BTRFS_NUM_BACKUP_ROOTS - 1) %
			BTRFS_NUM_BACKUP_ROOTS;
		*backup_index = newest;
		*num_backups_tried += 1;
	}
	root_backup = super->super_roots + newest;

	btrfs_set_super_generation(super,
				   btrfs_backup_tree_root_gen(root_backup));
	btrfs_set_super_root(super, btrfs_backup_tree_root(root_backup));
	btrfs_set_super_root_level(super,
				   btrfs_backup_tree_root_level(root_backup));
	btrfs_set_super_bytes_used(super, btrfs_backup_bytes_used(root_backup));

	/*
	 * fixme: the total bytes and num_devices need to match or we should
	 * need a fsck
	 */
	btrfs_set_super_total_bytes(super, btrfs_backup_total_bytes(root_backup));
	btrfs_set_super_num_devices(super, btrfs_backup_num_devices(root_backup));
	return 0;
}

L
Liu Bo 已提交
2102 2103 2104
/* helper to cleanup workers */
static void btrfs_stop_all_workers(struct btrfs_fs_info *fs_info)
{
2105
	btrfs_destroy_workqueue(fs_info->fixup_workers);
2106
	btrfs_destroy_workqueue(fs_info->delalloc_workers);
2107
	btrfs_destroy_workqueue(fs_info->workers);
2108 2109 2110
	btrfs_destroy_workqueue(fs_info->endio_workers);
	btrfs_destroy_workqueue(fs_info->endio_meta_workers);
	btrfs_destroy_workqueue(fs_info->endio_raid56_workers);
2111
	btrfs_destroy_workqueue(fs_info->endio_repair_workers);
2112
	btrfs_destroy_workqueue(fs_info->rmw_workers);
2113 2114 2115
	btrfs_destroy_workqueue(fs_info->endio_meta_write_workers);
	btrfs_destroy_workqueue(fs_info->endio_write_workers);
	btrfs_destroy_workqueue(fs_info->endio_freespace_worker);
2116
	btrfs_destroy_workqueue(fs_info->submit_workers);
2117
	btrfs_destroy_workqueue(fs_info->delayed_workers);
2118
	btrfs_destroy_workqueue(fs_info->caching_workers);
2119
	btrfs_destroy_workqueue(fs_info->readahead_workers);
2120
	btrfs_destroy_workqueue(fs_info->flush_workers);
2121
	btrfs_destroy_workqueue(fs_info->qgroup_rescan_workers);
C
Chris Mason 已提交
2122
	btrfs_destroy_workqueue(fs_info->extent_workers);
L
Liu Bo 已提交
2123 2124
}

2125 2126 2127 2128 2129 2130 2131 2132 2133 2134
static void free_root_extent_buffers(struct btrfs_root *root)
{
	if (root) {
		free_extent_buffer(root->node);
		free_extent_buffer(root->commit_root);
		root->node = NULL;
		root->commit_root = NULL;
	}
}

C
Chris Mason 已提交
2135 2136 2137
/* helper to cleanup tree roots */
static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
{
2138
	free_root_extent_buffers(info->tree_root);
2139

2140 2141 2142 2143 2144 2145 2146
	free_root_extent_buffers(info->dev_root);
	free_root_extent_buffers(info->extent_root);
	free_root_extent_buffers(info->csum_root);
	free_root_extent_buffers(info->quota_root);
	free_root_extent_buffers(info->uuid_root);
	if (chunk_root)
		free_root_extent_buffers(info->chunk_root);
C
Chris Mason 已提交
2147 2148
}

2149
void btrfs_free_fs_roots(struct btrfs_fs_info *fs_info)
2150 2151 2152 2153 2154 2155 2156 2157 2158 2159
{
	int ret;
	struct btrfs_root *gang[8];
	int i;

	while (!list_empty(&fs_info->dead_roots)) {
		gang[0] = list_entry(fs_info->dead_roots.next,
				     struct btrfs_root, root_list);
		list_del(&gang[0]->root_list);

2160
		if (test_bit(BTRFS_ROOT_IN_RADIX, &gang[0]->state)) {
2161
			btrfs_drop_and_free_fs_root(fs_info, gang[0]);
2162 2163 2164
		} else {
			free_extent_buffer(gang[0]->node);
			free_extent_buffer(gang[0]->commit_root);
2165
			btrfs_put_fs_root(gang[0]);
2166 2167 2168 2169 2170 2171 2172 2173 2174 2175
		}
	}

	while (1) {
		ret = radix_tree_gang_lookup(&fs_info->fs_roots_radix,
					     (void **)gang, 0,
					     ARRAY_SIZE(gang));
		if (!ret)
			break;
		for (i = 0; i < ret; i++)
2176
			btrfs_drop_and_free_fs_root(fs_info, gang[i]);
2177
	}
2178 2179 2180 2181 2182 2183

	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) {
		btrfs_free_log_root_tree(NULL, fs_info);
		btrfs_destroy_pinned_extent(fs_info->tree_root,
					    fs_info->pinned_extents);
	}
2184
}
C
Chris Mason 已提交
2185

2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196
static void btrfs_init_scrub(struct btrfs_fs_info *fs_info)
{
	mutex_init(&fs_info->scrub_lock);
	atomic_set(&fs_info->scrubs_running, 0);
	atomic_set(&fs_info->scrub_pause_req, 0);
	atomic_set(&fs_info->scrubs_paused, 0);
	atomic_set(&fs_info->scrub_cancel_req, 0);
	init_waitqueue_head(&fs_info->scrub_pause_wait);
	fs_info->scrub_workers_refcnt = 0;
}

2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207
static void btrfs_init_balance(struct btrfs_fs_info *fs_info)
{
	spin_lock_init(&fs_info->balance_lock);
	mutex_init(&fs_info->balance_mutex);
	atomic_set(&fs_info->balance_running, 0);
	atomic_set(&fs_info->balance_pause_req, 0);
	atomic_set(&fs_info->balance_cancel_req, 0);
	fs_info->balance_ctl = NULL;
	init_waitqueue_head(&fs_info->balance_wait_q);
}

2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236
static void btrfs_init_btree_inode(struct btrfs_fs_info *fs_info,
				   struct btrfs_root *tree_root)
{
	fs_info->btree_inode->i_ino = BTRFS_BTREE_INODE_OBJECTID;
	set_nlink(fs_info->btree_inode, 1);
	/*
	 * we set the i_size on the btree inode to the max possible int.
	 * the real end of the address space is determined by all of
	 * the devices in the system
	 */
	fs_info->btree_inode->i_size = OFFSET_MAX;
	fs_info->btree_inode->i_mapping->a_ops = &btree_aops;

	RB_CLEAR_NODE(&BTRFS_I(fs_info->btree_inode)->rb_node);
	extent_io_tree_init(&BTRFS_I(fs_info->btree_inode)->io_tree,
			     fs_info->btree_inode->i_mapping);
	BTRFS_I(fs_info->btree_inode)->io_tree.track_uptodate = 0;
	extent_map_tree_init(&BTRFS_I(fs_info->btree_inode)->extent_tree);

	BTRFS_I(fs_info->btree_inode)->io_tree.ops = &btree_extent_io_ops;

	BTRFS_I(fs_info->btree_inode)->root = tree_root;
	memset(&BTRFS_I(fs_info->btree_inode)->location, 0,
	       sizeof(struct btrfs_key));
	set_bit(BTRFS_INODE_DUMMY,
		&BTRFS_I(fs_info->btree_inode)->runtime_flags);
	btrfs_insert_inode_hash(fs_info->btree_inode);
}

2237 2238 2239 2240 2241 2242 2243 2244 2245 2246
static void btrfs_init_dev_replace_locks(struct btrfs_fs_info *fs_info)
{
	fs_info->dev_replace.lock_owner = 0;
	atomic_set(&fs_info->dev_replace.nesting_level, 0);
	mutex_init(&fs_info->dev_replace.lock_finishing_cancel_unmount);
	mutex_init(&fs_info->dev_replace.lock_management_lock);
	mutex_init(&fs_info->dev_replace.lock);
	init_waitqueue_head(&fs_info->replace_wait);
}

2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260
static void btrfs_init_qgroup(struct btrfs_fs_info *fs_info)
{
	spin_lock_init(&fs_info->qgroup_lock);
	mutex_init(&fs_info->qgroup_ioctl_lock);
	fs_info->qgroup_tree = RB_ROOT;
	fs_info->qgroup_op_tree = RB_ROOT;
	INIT_LIST_HEAD(&fs_info->dirty_qgroups);
	fs_info->qgroup_seq = 1;
	fs_info->quota_enabled = 0;
	fs_info->pending_quota_state = 0;
	fs_info->qgroup_ulist = NULL;
	mutex_init(&fs_info->qgroup_rescan_lock);
}

2261 2262 2263 2264
static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info,
		struct btrfs_fs_devices *fs_devices)
{
	int max_active = fs_info->thread_pool_size;
2265
	unsigned int flags = WQ_MEM_RECLAIM | WQ_FREEZABLE | WQ_UNBOUND;
2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340

	fs_info->workers =
		btrfs_alloc_workqueue("worker", flags | WQ_HIGHPRI,
				      max_active, 16);

	fs_info->delalloc_workers =
		btrfs_alloc_workqueue("delalloc", flags, max_active, 2);

	fs_info->flush_workers =
		btrfs_alloc_workqueue("flush_delalloc", flags, max_active, 0);

	fs_info->caching_workers =
		btrfs_alloc_workqueue("cache", flags, max_active, 0);

	/*
	 * a higher idle thresh on the submit workers makes it much more
	 * likely that bios will be send down in a sane order to the
	 * devices
	 */
	fs_info->submit_workers =
		btrfs_alloc_workqueue("submit", flags,
				      min_t(u64, fs_devices->num_devices,
					    max_active), 64);

	fs_info->fixup_workers =
		btrfs_alloc_workqueue("fixup", flags, 1, 0);

	/*
	 * endios are largely parallel and should have a very
	 * low idle thresh
	 */
	fs_info->endio_workers =
		btrfs_alloc_workqueue("endio", flags, max_active, 4);
	fs_info->endio_meta_workers =
		btrfs_alloc_workqueue("endio-meta", flags, max_active, 4);
	fs_info->endio_meta_write_workers =
		btrfs_alloc_workqueue("endio-meta-write", flags, max_active, 2);
	fs_info->endio_raid56_workers =
		btrfs_alloc_workqueue("endio-raid56", flags, max_active, 4);
	fs_info->endio_repair_workers =
		btrfs_alloc_workqueue("endio-repair", flags, 1, 0);
	fs_info->rmw_workers =
		btrfs_alloc_workqueue("rmw", flags, max_active, 2);
	fs_info->endio_write_workers =
		btrfs_alloc_workqueue("endio-write", flags, max_active, 2);
	fs_info->endio_freespace_worker =
		btrfs_alloc_workqueue("freespace-write", flags, max_active, 0);
	fs_info->delayed_workers =
		btrfs_alloc_workqueue("delayed-meta", flags, max_active, 0);
	fs_info->readahead_workers =
		btrfs_alloc_workqueue("readahead", flags, max_active, 2);
	fs_info->qgroup_rescan_workers =
		btrfs_alloc_workqueue("qgroup-rescan", flags, 1, 0);
	fs_info->extent_workers =
		btrfs_alloc_workqueue("extent-refs", flags,
				      min_t(u64, fs_devices->num_devices,
					    max_active), 8);

	if (!(fs_info->workers && fs_info->delalloc_workers &&
	      fs_info->submit_workers && fs_info->flush_workers &&
	      fs_info->endio_workers && fs_info->endio_meta_workers &&
	      fs_info->endio_meta_write_workers &&
	      fs_info->endio_repair_workers &&
	      fs_info->endio_write_workers && fs_info->endio_raid56_workers &&
	      fs_info->endio_freespace_worker && fs_info->rmw_workers &&
	      fs_info->caching_workers && fs_info->readahead_workers &&
	      fs_info->fixup_workers && fs_info->delayed_workers &&
	      fs_info->extent_workers &&
	      fs_info->qgroup_rescan_workers)) {
		return -ENOMEM;
	}

	return 0;
}

2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365
static int btrfs_replay_log(struct btrfs_fs_info *fs_info,
			    struct btrfs_fs_devices *fs_devices)
{
	int ret;
	struct btrfs_root *tree_root = fs_info->tree_root;
	struct btrfs_root *log_tree_root;
	struct btrfs_super_block *disk_super = fs_info->super_copy;
	u64 bytenr = btrfs_super_log_root(disk_super);

	if (fs_devices->rw_devices == 0) {
		printk(KERN_WARNING "BTRFS: log replay required "
		       "on RO media\n");
		return -EIO;
	}

	log_tree_root = btrfs_alloc_root(fs_info);
	if (!log_tree_root)
		return -ENOMEM;

	__setup_root(tree_root->nodesize, tree_root->sectorsize,
			tree_root->stripesize, log_tree_root, fs_info,
			BTRFS_TREE_LOG_OBJECTID);

	log_tree_root->node = read_tree_block(tree_root, bytenr,
			fs_info->generation + 1);
2366 2367
	if (IS_ERR(log_tree_root->node)) {
		printk(KERN_ERR "BTRFS: failed to read log tree\n");
2368
		ret = PTR_ERR(log_tree_root->node);
2369
		kfree(log_tree_root);
2370
		return ret;
2371
	} else if (!extent_buffer_uptodate(log_tree_root->node)) {
2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395
		printk(KERN_ERR "BTRFS: failed to read log tree\n");
		free_extent_buffer(log_tree_root->node);
		kfree(log_tree_root);
		return -EIO;
	}
	/* returns with log_tree_root freed on success */
	ret = btrfs_recover_log_trees(log_tree_root);
	if (ret) {
		btrfs_error(tree_root->fs_info, ret,
			    "Failed to recover log tree");
		free_extent_buffer(log_tree_root->node);
		kfree(log_tree_root);
		return ret;
	}

	if (fs_info->sb->s_flags & MS_RDONLY) {
		ret = btrfs_commit_super(tree_root);
		if (ret)
			return ret;
	}

	return 0;
}

2396 2397 2398
static int btrfs_read_roots(struct btrfs_fs_info *fs_info,
			    struct btrfs_root *tree_root)
{
2399
	struct btrfs_root *root;
2400 2401 2402 2403 2404 2405 2406
	struct btrfs_key location;
	int ret;

	location.objectid = BTRFS_EXTENT_TREE_OBJECTID;
	location.type = BTRFS_ROOT_ITEM_KEY;
	location.offset = 0;

2407 2408 2409 2410 2411
	root = btrfs_read_tree_root(tree_root, &location);
	if (IS_ERR(root))
		return PTR_ERR(root);
	set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state);
	fs_info->extent_root = root;
2412 2413

	location.objectid = BTRFS_DEV_TREE_OBJECTID;
2414 2415 2416 2417 2418
	root = btrfs_read_tree_root(tree_root, &location);
	if (IS_ERR(root))
		return PTR_ERR(root);
	set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state);
	fs_info->dev_root = root;
2419 2420 2421
	btrfs_init_devices_late(fs_info);

	location.objectid = BTRFS_CSUM_TREE_OBJECTID;
2422 2423 2424 2425 2426
	root = btrfs_read_tree_root(tree_root, &location);
	if (IS_ERR(root))
		return PTR_ERR(root);
	set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state);
	fs_info->csum_root = root;
2427 2428

	location.objectid = BTRFS_QUOTA_TREE_OBJECTID;
2429 2430 2431
	root = btrfs_read_tree_root(tree_root, &location);
	if (!IS_ERR(root)) {
		set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state);
2432 2433
		fs_info->quota_enabled = 1;
		fs_info->pending_quota_state = 1;
2434
		fs_info->quota_root = root;
2435 2436 2437
	}

	location.objectid = BTRFS_UUID_TREE_OBJECTID;
2438 2439 2440
	root = btrfs_read_tree_root(tree_root, &location);
	if (IS_ERR(root)) {
		ret = PTR_ERR(root);
2441 2442 2443
		if (ret != -ENOENT)
			return ret;
	} else {
2444 2445
		set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state);
		fs_info->uuid_root = root;
2446 2447 2448 2449 2450
	}

	return 0;
}

A
Al Viro 已提交
2451 2452 2453
int open_ctree(struct super_block *sb,
	       struct btrfs_fs_devices *fs_devices,
	       char *options)
2454
{
2455 2456
	u32 sectorsize;
	u32 nodesize;
2457
	u32 stripesize;
2458
	u64 generation;
2459
	u64 features;
2460
	struct btrfs_key location;
2461
	struct buffer_head *bh;
2462
	struct btrfs_super_block *disk_super;
2463
	struct btrfs_fs_info *fs_info = btrfs_sb(sb);
2464
	struct btrfs_root *tree_root;
2465
	struct btrfs_root *chunk_root;
2466
	int ret;
2467
	int err = -EINVAL;
C
Chris Mason 已提交
2468 2469
	int num_backups_tried = 0;
	int backup_index = 0;
2470
	int max_active;
2471

2472
	tree_root = fs_info->tree_root = btrfs_alloc_root(fs_info);
A
Al Viro 已提交
2473
	chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info);
2474
	if (!tree_root || !chunk_root) {
C
Chris Mason 已提交
2475 2476 2477
		err = -ENOMEM;
		goto fail;
	}
2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490

	ret = init_srcu_struct(&fs_info->subvol_srcu);
	if (ret) {
		err = ret;
		goto fail;
	}

	ret = setup_bdi(fs_info, &fs_info->bdi);
	if (ret) {
		err = ret;
		goto fail_srcu;
	}

2491
	ret = percpu_counter_init(&fs_info->dirty_metadata_bytes, 0, GFP_KERNEL);
2492 2493 2494 2495 2496 2497 2498
	if (ret) {
		err = ret;
		goto fail_bdi;
	}
	fs_info->dirty_metadata_batch = PAGE_CACHE_SIZE *
					(1 + ilog2(nr_cpu_ids));

2499
	ret = percpu_counter_init(&fs_info->delalloc_bytes, 0, GFP_KERNEL);
2500 2501 2502 2503 2504
	if (ret) {
		err = ret;
		goto fail_dirty_metadata_bytes;
	}

2505
	ret = percpu_counter_init(&fs_info->bio_counter, 0, GFP_KERNEL);
2506 2507 2508 2509 2510
	if (ret) {
		err = ret;
		goto fail_delalloc_bytes;
	}

2511 2512 2513
	fs_info->btree_inode = new_inode(sb);
	if (!fs_info->btree_inode) {
		err = -ENOMEM;
2514
		goto fail_bio_counter;
2515 2516
	}

2517
	mapping_set_gfp_mask(fs_info->btree_inode->i_mapping, GFP_NOFS);
2518

2519
	INIT_RADIX_TREE(&fs_info->fs_roots_radix, GFP_ATOMIC);
2520
	INIT_RADIX_TREE(&fs_info->buffer_radix, GFP_ATOMIC);
C
Chris Mason 已提交
2521
	INIT_LIST_HEAD(&fs_info->trans_list);
2522
	INIT_LIST_HEAD(&fs_info->dead_roots);
Y
Yan, Zheng 已提交
2523
	INIT_LIST_HEAD(&fs_info->delayed_iputs);
2524
	INIT_LIST_HEAD(&fs_info->delalloc_roots);
2525
	INIT_LIST_HEAD(&fs_info->caching_block_groups);
2526
	spin_lock_init(&fs_info->delalloc_root_lock);
J
Josef Bacik 已提交
2527
	spin_lock_init(&fs_info->trans_lock);
2528
	spin_lock_init(&fs_info->fs_roots_radix_lock);
Y
Yan, Zheng 已提交
2529
	spin_lock_init(&fs_info->delayed_iput_lock);
C
Chris Mason 已提交
2530
	spin_lock_init(&fs_info->defrag_inodes_lock);
2531
	spin_lock_init(&fs_info->free_chunk_lock);
J
Jan Schmidt 已提交
2532
	spin_lock_init(&fs_info->tree_mod_seq_lock);
2533
	spin_lock_init(&fs_info->super_lock);
J
Josef Bacik 已提交
2534
	spin_lock_init(&fs_info->qgroup_op_lock);
2535
	spin_lock_init(&fs_info->buffer_lock);
2536
	spin_lock_init(&fs_info->unused_bgs_lock);
J
Jan Schmidt 已提交
2537
	rwlock_init(&fs_info->tree_mod_log_lock);
2538
	mutex_init(&fs_info->unused_bg_unpin_mutex);
2539
	mutex_init(&fs_info->delete_unused_bgs_mutex);
C
Chris Mason 已提交
2540
	mutex_init(&fs_info->reloc_mutex);
2541
	mutex_init(&fs_info->delalloc_root_mutex);
2542
	seqlock_init(&fs_info->profiles_lock);
2543
	init_rwsem(&fs_info->delayed_iput_sem);
2544

2545
	INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
2546
	INIT_LIST_HEAD(&fs_info->space_info);
J
Jan Schmidt 已提交
2547
	INIT_LIST_HEAD(&fs_info->tree_mod_seq_list);
2548
	INIT_LIST_HEAD(&fs_info->unused_bgs);
2549
	btrfs_mapping_init(&fs_info->mapping_tree);
2550 2551 2552 2553 2554 2555 2556 2557 2558
	btrfs_init_block_rsv(&fs_info->global_block_rsv,
			     BTRFS_BLOCK_RSV_GLOBAL);
	btrfs_init_block_rsv(&fs_info->delalloc_block_rsv,
			     BTRFS_BLOCK_RSV_DELALLOC);
	btrfs_init_block_rsv(&fs_info->trans_block_rsv, BTRFS_BLOCK_RSV_TRANS);
	btrfs_init_block_rsv(&fs_info->chunk_block_rsv, BTRFS_BLOCK_RSV_CHUNK);
	btrfs_init_block_rsv(&fs_info->empty_block_rsv, BTRFS_BLOCK_RSV_EMPTY);
	btrfs_init_block_rsv(&fs_info->delayed_block_rsv,
			     BTRFS_BLOCK_RSV_DELOPS);
2559
	atomic_set(&fs_info->nr_async_submits, 0);
2560
	atomic_set(&fs_info->async_delalloc_pages, 0);
2561
	atomic_set(&fs_info->async_submit_draining, 0);
2562
	atomic_set(&fs_info->nr_async_bios, 0);
C
Chris Mason 已提交
2563
	atomic_set(&fs_info->defrag_running, 0);
J
Josef Bacik 已提交
2564
	atomic_set(&fs_info->qgroup_op_seq, 0);
2565
	atomic64_set(&fs_info->tree_mod_seq, 0);
C
Chris Mason 已提交
2566
	fs_info->sb = sb;
2567
	fs_info->max_inline = BTRFS_DEFAULT_MAX_INLINE;
J
Josef Bacik 已提交
2568
	fs_info->metadata_ratio = 0;
C
Chris Mason 已提交
2569
	fs_info->defrag_inodes = RB_ROOT;
2570
	fs_info->free_chunk_space = 0;
J
Jan Schmidt 已提交
2571
	fs_info->tree_mod_log = RB_ROOT;
2572
	fs_info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
2573
	fs_info->avg_delayed_ref_runtime = NSEC_PER_SEC >> 6; /* div by 64 */
2574 2575 2576
	/* readahead state */
	INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT);
	spin_lock_init(&fs_info->reada_lock);
C
Chris Mason 已提交
2577

2578 2579
	fs_info->thread_pool_size = min_t(unsigned long,
					  num_online_cpus() + 2, 8);
2580

2581 2582
	INIT_LIST_HEAD(&fs_info->ordered_roots);
	spin_lock_init(&fs_info->ordered_root_lock);
2583 2584 2585 2586 2587 2588 2589
	fs_info->delayed_root = kmalloc(sizeof(struct btrfs_delayed_root),
					GFP_NOFS);
	if (!fs_info->delayed_root) {
		err = -ENOMEM;
		goto fail_iput;
	}
	btrfs_init_delayed_root(fs_info->delayed_root);
2590

2591
	btrfs_init_scrub(fs_info);
2592 2593 2594
#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
	fs_info->check_integrity_print_mask = 0;
#endif
2595
	btrfs_init_balance(fs_info);
2596
	btrfs_init_async_reclaim_work(&fs_info->async_reclaim_work);
A
Arne Jansen 已提交
2597

2598 2599
	sb->s_blocksize = 4096;
	sb->s_blocksize_bits = blksize_bits(4096);
J
Jens Axboe 已提交
2600
	sb->s_bdi = &fs_info->bdi;
2601

2602
	btrfs_init_btree_inode(fs_info, tree_root);
2603

J
Josef Bacik 已提交
2604
	spin_lock_init(&fs_info->block_group_cache_lock);
2605
	fs_info->block_group_cache_tree = RB_ROOT;
2606
	fs_info->first_logical_byte = (u64)-1;
J
Josef Bacik 已提交
2607

2608
	extent_io_tree_init(&fs_info->freed_extents[0],
2609
			     fs_info->btree_inode->i_mapping);
2610
	extent_io_tree_init(&fs_info->freed_extents[1],
2611
			     fs_info->btree_inode->i_mapping);
2612
	fs_info->pinned_extents = &fs_info->freed_extents[0];
2613
	fs_info->do_barriers = 1;
2614

C
Chris Mason 已提交
2615

2616
	mutex_init(&fs_info->ordered_operations_mutex);
2617
	mutex_init(&fs_info->tree_log_mutex);
2618
	mutex_init(&fs_info->chunk_mutex);
2619 2620
	mutex_init(&fs_info->transaction_kthread_mutex);
	mutex_init(&fs_info->cleaner_mutex);
2621
	mutex_init(&fs_info->volume_mutex);
2622
	mutex_init(&fs_info->ro_block_group_mutex);
2623
	init_rwsem(&fs_info->commit_root_sem);
2624
	init_rwsem(&fs_info->cleanup_work_sem);
2625
	init_rwsem(&fs_info->subvol_sem);
S
Stefan Behrens 已提交
2626
	sema_init(&fs_info->uuid_tree_rescan_sem, 1);
2627

2628
	btrfs_init_dev_replace_locks(fs_info);
2629
	btrfs_init_qgroup(fs_info);
2630

2631 2632 2633
	btrfs_init_free_cluster(&fs_info->meta_alloc_cluster);
	btrfs_init_free_cluster(&fs_info->data_alloc_cluster);

2634
	init_waitqueue_head(&fs_info->transaction_throttle);
2635
	init_waitqueue_head(&fs_info->transaction_wait);
S
Sage Weil 已提交
2636
	init_waitqueue_head(&fs_info->transaction_blocked_wait);
2637
	init_waitqueue_head(&fs_info->async_submit_wait);
2638

2639 2640
	INIT_LIST_HEAD(&fs_info->pinned_chunks);

D
David Woodhouse 已提交
2641 2642
	ret = btrfs_alloc_stripe_hash_table(fs_info);
	if (ret) {
2643
		err = ret;
D
David Woodhouse 已提交
2644 2645 2646
		goto fail_alloc;
	}

2647
	__setup_root(4096, 4096, 4096, tree_root,
C
Chris Mason 已提交
2648
		     fs_info, BTRFS_ROOT_TREE_OBJECTID);
2649

2650
	invalidate_bdev(fs_devices->latest_bdev);
D
David Sterba 已提交
2651 2652 2653 2654

	/*
	 * Read super block and check the signature bytes only
	 */
Y
Yan Zheng 已提交
2655
	bh = btrfs_read_dev_super(fs_devices->latest_bdev);
2656 2657
	if (!bh) {
		err = -EINVAL;
2658
		goto fail_alloc;
2659
	}
C
Chris Mason 已提交
2660

D
David Sterba 已提交
2661 2662 2663 2664 2665
	/*
	 * We want to check superblock checksum, the type is stored inside.
	 * Pass the whole disk block of size BTRFS_SUPER_INFO_SIZE (4k).
	 */
	if (btrfs_check_super_csum(bh->b_data)) {
2666
		printk(KERN_ERR "BTRFS: superblock checksum mismatch\n");
D
David Sterba 已提交
2667 2668 2669 2670 2671 2672 2673 2674 2675
		err = -EINVAL;
		goto fail_alloc;
	}

	/*
	 * super_copy is zeroed at allocation time and we never touch the
	 * following bytes up to INFO_SIZE, the checksum is calculated from
	 * the whole block of INFO_SIZE
	 */
2676 2677 2678
	memcpy(fs_info->super_copy, bh->b_data, sizeof(*fs_info->super_copy));
	memcpy(fs_info->super_for_commit, fs_info->super_copy,
	       sizeof(*fs_info->super_for_commit));
2679
	brelse(bh);
2680

2681
	memcpy(fs_info->fsid, fs_info->super_copy->fsid, BTRFS_FSID_SIZE);
2682

D
David Sterba 已提交
2683 2684
	ret = btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY);
	if (ret) {
2685
		printk(KERN_ERR "BTRFS: superblock contains fatal errors\n");
D
David Sterba 已提交
2686 2687 2688 2689
		err = -EINVAL;
		goto fail_alloc;
	}

2690
	disk_super = fs_info->super_copy;
2691
	if (!btrfs_super_root(disk_super))
2692
		goto fail_alloc;
2693

L
liubo 已提交
2694
	/* check FS state, whether FS is broken. */
2695 2696
	if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_ERROR)
		set_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state);
L
liubo 已提交
2697

C
Chris Mason 已提交
2698 2699 2700 2701 2702 2703 2704
	/*
	 * run through our array of backup supers and setup
	 * our ring pointer to the oldest one
	 */
	generation = btrfs_super_generation(disk_super);
	find_oldest_super_backup(fs_info, generation);

2705 2706 2707 2708 2709 2710
	/*
	 * In the long term, we'll store the compression type in the super
	 * block, and it'll be used for per file compression control.
	 */
	fs_info->compress_type = BTRFS_COMPRESS_ZLIB;

Y
Yan Zheng 已提交
2711 2712 2713
	ret = btrfs_parse_options(tree_root, options);
	if (ret) {
		err = ret;
2714
		goto fail_alloc;
Y
Yan Zheng 已提交
2715
	}
2716

2717 2718 2719 2720 2721
	features = btrfs_super_incompat_flags(disk_super) &
		~BTRFS_FEATURE_INCOMPAT_SUPP;
	if (features) {
		printk(KERN_ERR "BTRFS: couldn't mount because of "
		       "unsupported optional features (%Lx).\n",
2722
		       features);
2723
		err = -EINVAL;
2724
		goto fail_alloc;
2725 2726
	}

2727 2728 2729 2730
	/*
	 * Leafsize and nodesize were always equal, this is only a sanity check.
	 */
	if (le32_to_cpu(disk_super->__unused_leafsize) !=
2731 2732 2733 2734
	    btrfs_super_nodesize(disk_super)) {
		printk(KERN_ERR "BTRFS: couldn't mount because metadata "
		       "blocksizes don't match.  node %d leaf %d\n",
		       btrfs_super_nodesize(disk_super),
2735
		       le32_to_cpu(disk_super->__unused_leafsize));
2736 2737 2738
		err = -EINVAL;
		goto fail_alloc;
	}
2739
	if (btrfs_super_nodesize(disk_super) > BTRFS_MAX_METADATA_BLOCKSIZE) {
2740 2741
		printk(KERN_ERR "BTRFS: couldn't mount because metadata "
		       "blocksize (%d) was too large\n",
2742
		       btrfs_super_nodesize(disk_super));
2743 2744 2745 2746
		err = -EINVAL;
		goto fail_alloc;
	}

2747
	features = btrfs_super_incompat_flags(disk_super);
L
Li Zefan 已提交
2748
	features |= BTRFS_FEATURE_INCOMPAT_MIXED_BACKREF;
L
Li Zefan 已提交
2749
	if (tree_root->fs_info->compress_type == BTRFS_COMPRESS_LZO)
L
Li Zefan 已提交
2750
		features |= BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO;
2751

2752
	if (features & BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA)
2753
		printk(KERN_INFO "BTRFS: has skinny extents\n");
2754

2755 2756 2757 2758
	/*
	 * flag our filesystem as having big metadata blocks if
	 * they are bigger than the page size
	 */
2759
	if (btrfs_super_nodesize(disk_super) > PAGE_CACHE_SIZE) {
2760
		if (!(features & BTRFS_FEATURE_INCOMPAT_BIG_METADATA))
2761
			printk(KERN_INFO "BTRFS: flagging fs with big metadata feature\n");
2762 2763 2764
		features |= BTRFS_FEATURE_INCOMPAT_BIG_METADATA;
	}

2765 2766 2767
	nodesize = btrfs_super_nodesize(disk_super);
	sectorsize = btrfs_super_sectorsize(disk_super);
	stripesize = btrfs_super_stripesize(disk_super);
2768
	fs_info->dirty_metadata_batch = nodesize * (1 + ilog2(nr_cpu_ids));
2769
	fs_info->delalloc_batch = sectorsize * 512 * (1 + ilog2(nr_cpu_ids));
2770 2771 2772 2773 2774 2775

	/*
	 * mixed block groups end up with duplicate but slightly offset
	 * extent buffers for the same range.  It leads to corruptions
	 */
	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
2776
	    (sectorsize != nodesize)) {
2777
		printk(KERN_ERR "BTRFS: unequal leaf/node/sector sizes "
2778 2779 2780 2781 2782
				"are not allowed for mixed block groups on %s\n",
				sb->s_id);
		goto fail_alloc;
	}

2783 2784 2785 2786
	/*
	 * Needn't use the lock because there is no other task which will
	 * update the flag.
	 */
L
Li Zefan 已提交
2787
	btrfs_set_super_incompat_flags(disk_super, features);
2788

2789 2790 2791 2792 2793
	features = btrfs_super_compat_ro_flags(disk_super) &
		~BTRFS_FEATURE_COMPAT_RO_SUPP;
	if (!(sb->s_flags & MS_RDONLY) && features) {
		printk(KERN_ERR "BTRFS: couldn't mount RDWR because of "
		       "unsupported option features (%Lx).\n",
2794
		       features);
2795
		err = -EINVAL;
2796
		goto fail_alloc;
2797
	}
2798

2799
	max_active = fs_info->thread_pool_size;
2800

2801 2802 2803
	ret = btrfs_init_workqueues(fs_info, fs_devices);
	if (ret) {
		err = ret;
2804 2805
		goto fail_sb_buffer;
	}
2806

2807
	fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super);
C
Chris Mason 已提交
2808 2809
	fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages,
				    4 * 1024 * 1024 / PAGE_CACHE_SIZE);
2810

2811 2812
	tree_root->nodesize = nodesize;
	tree_root->sectorsize = sectorsize;
2813
	tree_root->stripesize = stripesize;
2814 2815 2816

	sb->s_blocksize = sectorsize;
	sb->s_blocksize_bits = blksize_bits(sectorsize);
2817

2818
	if (btrfs_super_magic(disk_super) != BTRFS_MAGIC) {
2819
		printk(KERN_ERR "BTRFS: valid FS not found on %s\n", sb->s_id);
C
Chris Mason 已提交
2820 2821
		goto fail_sb_buffer;
	}
2822

2823
	if (sectorsize != PAGE_SIZE) {
2824
		printk(KERN_ERR "BTRFS: incompatible sector size (%lu) "
2825
		       "found on %s\n", (unsigned long)sectorsize, sb->s_id);
2826 2827 2828
		goto fail_sb_buffer;
	}

2829
	mutex_lock(&fs_info->chunk_mutex);
Y
Yan Zheng 已提交
2830
	ret = btrfs_read_sys_array(tree_root);
2831
	mutex_unlock(&fs_info->chunk_mutex);
2832
	if (ret) {
2833
		printk(KERN_ERR "BTRFS: failed to read the system "
C
Chris Mason 已提交
2834
		       "array on %s\n", sb->s_id);
2835
		goto fail_sb_buffer;
2836
	}
2837

2838
	generation = btrfs_super_chunk_root_generation(disk_super);
2839

2840 2841
	__setup_root(nodesize, sectorsize, stripesize, chunk_root,
		     fs_info, BTRFS_CHUNK_TREE_OBJECTID);
2842 2843 2844

	chunk_root->node = read_tree_block(chunk_root,
					   btrfs_super_chunk_root(disk_super),
2845
					   generation);
2846 2847
	if (IS_ERR(chunk_root->node) ||
	    !extent_buffer_uptodate(chunk_root->node)) {
2848
		printk(KERN_ERR "BTRFS: failed to read chunk root on %s\n",
2849
		       sb->s_id);
2850
		chunk_root->node = NULL;
C
Chris Mason 已提交
2851
		goto fail_tree_roots;
2852
	}
2853 2854
	btrfs_set_root_node(&chunk_root->root_item, chunk_root->node);
	chunk_root->commit_root = btrfs_root_node(chunk_root);
2855

2856
	read_extent_buffer(chunk_root->node, fs_info->chunk_tree_uuid,
2857
	   btrfs_header_chunk_tree_uuid(chunk_root->node), BTRFS_UUID_SIZE);
2858

2859
	ret = btrfs_read_chunk_tree(chunk_root);
Y
Yan Zheng 已提交
2860
	if (ret) {
2861
		printk(KERN_ERR "BTRFS: failed to read chunk tree on %s\n",
C
Chris Mason 已提交
2862
		       sb->s_id);
C
Chris Mason 已提交
2863
		goto fail_tree_roots;
Y
Yan Zheng 已提交
2864
	}
2865

2866 2867 2868 2869
	/*
	 * keep the device that is marked to be the target device for the
	 * dev_replace procedure
	 */
2870
	btrfs_close_extra_devices(fs_devices, 0);
2871

2872
	if (!fs_devices->latest_bdev) {
2873
		printk(KERN_ERR "BTRFS: failed to read devices on %s\n",
2874 2875 2876 2877
		       sb->s_id);
		goto fail_tree_roots;
	}

C
Chris Mason 已提交
2878
retry_root_backup:
2879
	generation = btrfs_super_generation(disk_super);
2880

C
Chris Mason 已提交
2881
	tree_root->node = read_tree_block(tree_root,
2882
					  btrfs_super_root(disk_super),
2883
					  generation);
2884 2885
	if (IS_ERR(tree_root->node) ||
	    !extent_buffer_uptodate(tree_root->node)) {
2886
		printk(KERN_WARNING "BTRFS: failed to read tree root on %s\n",
2887
		       sb->s_id);
2888
		tree_root->node = NULL;
C
Chris Mason 已提交
2889
		goto recovery_tree_root;
2890
	}
C
Chris Mason 已提交
2891

2892 2893
	btrfs_set_root_node(&tree_root->root_item, tree_root->node);
	tree_root->commit_root = btrfs_root_node(tree_root);
2894
	btrfs_set_root_refs(&tree_root->root_item, 1);
2895

2896 2897
	ret = btrfs_read_roots(fs_info, tree_root);
	if (ret)
C
Chris Mason 已提交
2898
		goto recovery_tree_root;
2899

2900 2901 2902
	fs_info->generation = generation;
	fs_info->last_trans_committed = generation;

2903 2904
	ret = btrfs_recover_balance(fs_info);
	if (ret) {
2905
		printk(KERN_ERR "BTRFS: failed to recover balance\n");
2906 2907 2908
		goto fail_block_groups;
	}

2909 2910
	ret = btrfs_init_dev_stats(fs_info);
	if (ret) {
2911
		printk(KERN_ERR "BTRFS: failed to init dev_stats: %d\n",
2912 2913 2914 2915
		       ret);
		goto fail_block_groups;
	}

2916 2917
	ret = btrfs_init_dev_replace(fs_info);
	if (ret) {
2918
		pr_err("BTRFS: failed to init dev_replace: %d\n", ret);
2919 2920 2921
		goto fail_block_groups;
	}

2922
	btrfs_close_extra_devices(fs_devices, 1);
2923

2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935
	ret = btrfs_sysfs_add_fsid(fs_devices, NULL);
	if (ret) {
		pr_err("BTRFS: failed to init sysfs fsid interface: %d\n", ret);
		goto fail_block_groups;
	}

	ret = btrfs_sysfs_add_device(fs_devices);
	if (ret) {
		pr_err("BTRFS: failed to init sysfs device interface: %d\n", ret);
		goto fail_fsdev_sysfs;
	}

2936
	ret = btrfs_sysfs_add_mounted(fs_info);
2937
	if (ret) {
2938
		pr_err("BTRFS: failed to init sysfs interface: %d\n", ret);
2939
		goto fail_fsdev_sysfs;
2940 2941 2942 2943
	}

	ret = btrfs_init_space_info(fs_info);
	if (ret) {
2944
		printk(KERN_ERR "BTRFS: Failed to initial space info: %d\n", ret);
2945
		goto fail_sysfs;
2946 2947
	}

2948
	ret = btrfs_read_block_groups(fs_info->extent_root);
2949
	if (ret) {
2950
		printk(KERN_ERR "BTRFS: Failed to read block groups: %d\n", ret);
2951
		goto fail_sysfs;
2952
	}
2953 2954
	fs_info->num_tolerated_disk_barrier_failures =
		btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
2955 2956 2957
	if (fs_info->fs_devices->missing_devices >
	     fs_info->num_tolerated_disk_barrier_failures &&
	    !(sb->s_flags & MS_RDONLY)) {
2958 2959 2960
		pr_warn("BTRFS: missing devices(%llu) exceeds the limit(%d), writeable mount is not allowed\n",
			fs_info->fs_devices->missing_devices,
			fs_info->num_tolerated_disk_barrier_failures);
2961
		goto fail_sysfs;
2962
	}
C
Chris Mason 已提交
2963

2964 2965
	fs_info->cleaner_kthread = kthread_run(cleaner_kthread, tree_root,
					       "btrfs-cleaner");
2966
	if (IS_ERR(fs_info->cleaner_kthread))
2967
		goto fail_sysfs;
2968 2969 2970 2971

	fs_info->transaction_kthread = kthread_run(transaction_kthread,
						   tree_root,
						   "btrfs-transaction");
2972
	if (IS_ERR(fs_info->transaction_kthread))
2973
		goto fail_cleaner;
2974

C
Chris Mason 已提交
2975 2976 2977
	if (!btrfs_test_opt(tree_root, SSD) &&
	    !btrfs_test_opt(tree_root, NOSSD) &&
	    !fs_info->fs_devices->rotating) {
2978
		printk(KERN_INFO "BTRFS: detected SSD devices, enabling SSD "
C
Chris Mason 已提交
2979 2980 2981 2982
		       "mode\n");
		btrfs_set_opt(fs_info->mount_opt, SSD);
	}

2983 2984 2985 2986 2987
	/*
	 * Mount does not set all options immediatelly, we can do it now and do
	 * not have to wait for transaction commit
	 */
	btrfs_apply_pending_changes(fs_info);
2988

2989 2990 2991 2992 2993 2994 2995 2996
#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
	if (btrfs_test_opt(tree_root, CHECK_INTEGRITY)) {
		ret = btrfsic_mount(tree_root, fs_devices,
				    btrfs_test_opt(tree_root,
					CHECK_INTEGRITY_INCLUDING_EXTENT_DATA) ?
				    1 : 0,
				    fs_info->check_integrity_print_mask);
		if (ret)
2997
			printk(KERN_WARNING "BTRFS: failed to initialize"
2998 2999 3000
			       " integrity check module %s\n", sb->s_id);
	}
#endif
3001 3002 3003
	ret = btrfs_read_qgroup_config(fs_info);
	if (ret)
		goto fail_trans_kthread;
3004

L
liubo 已提交
3005
	/* do not make disk changes in broken FS */
3006
	if (btrfs_super_log_root(disk_super) != 0) {
3007
		ret = btrfs_replay_log(fs_info, fs_devices);
3008
		if (ret) {
3009
			err = ret;
3010
			goto fail_qgroup;
3011
		}
3012
	}
Z
Zheng Yan 已提交
3013

3014
	ret = btrfs_find_orphan_roots(tree_root);
3015
	if (ret)
3016
		goto fail_qgroup;
3017

3018
	if (!(sb->s_flags & MS_RDONLY)) {
3019
		ret = btrfs_cleanup_fs_roots(fs_info);
3020
		if (ret)
3021
			goto fail_qgroup;
3022

3023
		mutex_lock(&fs_info->cleaner_mutex);
3024
		ret = btrfs_recover_relocation(tree_root);
3025
		mutex_unlock(&fs_info->cleaner_mutex);
3026 3027
		if (ret < 0) {
			printk(KERN_WARNING
3028
			       "BTRFS: failed to recover relocation\n");
3029
			err = -EINVAL;
3030
			goto fail_qgroup;
3031
		}
3032
	}
Z
Zheng Yan 已提交
3033

3034 3035
	location.objectid = BTRFS_FS_TREE_OBJECTID;
	location.type = BTRFS_ROOT_ITEM_KEY;
3036
	location.offset = 0;
3037 3038

	fs_info->fs_root = btrfs_read_fs_root_no_name(fs_info, &location);
3039 3040
	if (IS_ERR(fs_info->fs_root)) {
		err = PTR_ERR(fs_info->fs_root);
3041
		goto fail_qgroup;
3042
	}
C
Chris Mason 已提交
3043

3044 3045
	if (sb->s_flags & MS_RDONLY)
		return 0;
I
Ilya Dryomov 已提交
3046

3047 3048 3049
	down_read(&fs_info->cleanup_work_sem);
	if ((ret = btrfs_orphan_cleanup(fs_info->fs_root)) ||
	    (ret = btrfs_orphan_cleanup(fs_info->tree_root))) {
3050
		up_read(&fs_info->cleanup_work_sem);
3051 3052 3053 3054
		close_ctree(tree_root);
		return ret;
	}
	up_read(&fs_info->cleanup_work_sem);
I
Ilya Dryomov 已提交
3055

3056 3057
	ret = btrfs_resume_balance_async(fs_info);
	if (ret) {
3058
		printk(KERN_WARNING "BTRFS: failed to resume balance\n");
3059 3060
		close_ctree(tree_root);
		return ret;
3061 3062
	}

3063 3064
	ret = btrfs_resume_dev_replace_async(fs_info);
	if (ret) {
3065
		pr_warn("BTRFS: failed to resume dev_replace\n");
3066 3067 3068 3069
		close_ctree(tree_root);
		return ret;
	}

3070 3071
	btrfs_qgroup_rescan_resume(fs_info);

3072
	if (!fs_info->uuid_root) {
3073
		pr_info("BTRFS: creating UUID tree\n");
3074 3075
		ret = btrfs_create_uuid_tree(fs_info);
		if (ret) {
3076
			pr_warn("BTRFS: failed to create the UUID tree %d\n",
3077 3078 3079 3080
				ret);
			close_ctree(tree_root);
			return ret;
		}
3081 3082 3083
	} else if (btrfs_test_opt(tree_root, RESCAN_UUID_TREE) ||
		   fs_info->generation !=
				btrfs_super_uuid_tree_generation(disk_super)) {
3084
		pr_info("BTRFS: checking UUID tree\n");
3085 3086
		ret = btrfs_check_uuid_tree(fs_info);
		if (ret) {
3087
			pr_warn("BTRFS: failed to check the UUID tree %d\n",
3088 3089 3090 3091 3092 3093
				ret);
			close_ctree(tree_root);
			return ret;
		}
	} else {
		fs_info->update_uuid_tree_gen = 1;
3094 3095
	}

3096 3097
	fs_info->open = 1;

A
Al Viro 已提交
3098
	return 0;
C
Chris Mason 已提交
3099

3100 3101
fail_qgroup:
	btrfs_free_qgroup_config(fs_info);
3102 3103
fail_trans_kthread:
	kthread_stop(fs_info->transaction_kthread);
J
Josef Bacik 已提交
3104
	btrfs_cleanup_transaction(fs_info->tree_root);
3105
	btrfs_free_fs_roots(fs_info);
3106
fail_cleaner:
3107
	kthread_stop(fs_info->cleaner_kthread);
3108 3109 3110 3111 3112 3113 3114

	/*
	 * make sure we're done with the btree inode before we stop our
	 * kthreads
	 */
	filemap_write_and_wait(fs_info->btree_inode->i_mapping);

3115
fail_sysfs:
3116
	btrfs_sysfs_remove_mounted(fs_info);
3117

3118 3119 3120
fail_fsdev_sysfs:
	btrfs_sysfs_remove_fsid(fs_info->fs_devices);

3121
fail_block_groups:
J
Josef Bacik 已提交
3122
	btrfs_put_block_group_cache(fs_info);
3123
	btrfs_free_block_groups(fs_info);
C
Chris Mason 已提交
3124 3125 3126

fail_tree_roots:
	free_root_pointers(fs_info, 1);
3127
	invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
C
Chris Mason 已提交
3128

C
Chris Mason 已提交
3129
fail_sb_buffer:
L
Liu Bo 已提交
3130
	btrfs_stop_all_workers(fs_info);
3131
fail_alloc:
3132
fail_iput:
3133 3134
	btrfs_mapping_tree_free(&fs_info->mapping_tree);

3135
	iput(fs_info->btree_inode);
3136 3137
fail_bio_counter:
	percpu_counter_destroy(&fs_info->bio_counter);
3138 3139
fail_delalloc_bytes:
	percpu_counter_destroy(&fs_info->delalloc_bytes);
3140 3141
fail_dirty_metadata_bytes:
	percpu_counter_destroy(&fs_info->dirty_metadata_bytes);
3142
fail_bdi:
3143
	bdi_destroy(&fs_info->bdi);
3144 3145
fail_srcu:
	cleanup_srcu_struct(&fs_info->subvol_srcu);
3146
fail:
D
David Woodhouse 已提交
3147
	btrfs_free_stripe_hash_table(fs_info);
3148
	btrfs_close_devices(fs_info->fs_devices);
A
Al Viro 已提交
3149
	return err;
C
Chris Mason 已提交
3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167

recovery_tree_root:
	if (!btrfs_test_opt(tree_root, RECOVERY))
		goto fail_tree_roots;

	free_root_pointers(fs_info, 0);

	/* don't use the log in recovery mode, it won't be valid */
	btrfs_set_super_log_root(disk_super, 0);

	/* we can't trust the free space cache either */
	btrfs_set_opt(fs_info->mount_opt, CLEAR_CACHE);

	ret = next_root_backup(fs_info, fs_info->super_copy,
			       &num_backups_tried, &backup_index);
	if (ret == -1)
		goto fail_block_groups;
	goto retry_root_backup;
3168 3169
}

3170 3171 3172 3173 3174
static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate)
{
	if (uptodate) {
		set_buffer_uptodate(bh);
	} else {
3175 3176 3177
		struct btrfs_device *device = (struct btrfs_device *)
			bh->b_private;

3178
		printk_ratelimited_in_rcu(KERN_WARNING "BTRFS: lost page write due to "
3179 3180
					  "I/O error on %s\n",
					  rcu_str_deref(device->name));
3181 3182 3183
		/* note, we dont' set_buffer_write_io_error because we have
		 * our own ways of dealing with the IO errors
		 */
3184
		clear_buffer_uptodate(bh);
3185
		btrfs_dev_stat_inc_and_print(device, BTRFS_DEV_STAT_WRITE_ERRS);
3186 3187 3188 3189 3190
	}
	unlock_buffer(bh);
	put_bh(bh);
}

Y
Yan Zheng 已提交
3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206
struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
{
	struct buffer_head *bh;
	struct buffer_head *latest = NULL;
	struct btrfs_super_block *super;
	int i;
	u64 transid = 0;
	u64 bytenr;

	/* we would like to check all the supers, but that would make
	 * a btrfs mount succeed after a mkfs from a different FS.
	 * So, we need to add a special mount option to scan for
	 * later supers, using BTRFS_SUPER_MIRROR_MAX instead
	 */
	for (i = 0; i < 1; i++) {
		bytenr = btrfs_sb_offset(i);
3207 3208
		if (bytenr + BTRFS_SUPER_INFO_SIZE >=
					i_size_read(bdev->bd_inode))
Y
Yan Zheng 已提交
3209
			break;
3210 3211
		bh = __bread(bdev, bytenr / 4096,
					BTRFS_SUPER_INFO_SIZE);
Y
Yan Zheng 已提交
3212 3213 3214 3215 3216
		if (!bh)
			continue;

		super = (struct btrfs_super_block *)bh->b_data;
		if (btrfs_super_bytenr(super) != bytenr ||
3217
		    btrfs_super_magic(super) != BTRFS_MAGIC) {
Y
Yan Zheng 已提交
3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232
			brelse(bh);
			continue;
		}

		if (!latest || btrfs_super_generation(super) > transid) {
			brelse(latest);
			latest = bh;
			transid = btrfs_super_generation(super);
		} else {
			brelse(bh);
		}
	}
	return latest;
}

3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243
/*
 * this should be called twice, once with wait == 0 and
 * once with wait == 1.  When wait == 0 is done, all the buffer heads
 * we write are pinned.
 *
 * They are released when wait == 1 is done.
 * max_mirrors must be the same for both runs, and it indicates how
 * many supers on this one device should be written.
 *
 * max_mirrors == 0 means to write them all.
 */
Y
Yan Zheng 已提交
3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259
static int write_dev_supers(struct btrfs_device *device,
			    struct btrfs_super_block *sb,
			    int do_barriers, int wait, int max_mirrors)
{
	struct buffer_head *bh;
	int i;
	int ret;
	int errors = 0;
	u32 crc;
	u64 bytenr;

	if (max_mirrors == 0)
		max_mirrors = BTRFS_SUPER_MIRROR_MAX;

	for (i = 0; i < max_mirrors; i++) {
		bytenr = btrfs_sb_offset(i);
3260 3261
		if (bytenr + BTRFS_SUPER_INFO_SIZE >=
		    device->commit_total_bytes)
Y
Yan Zheng 已提交
3262 3263 3264 3265 3266
			break;

		if (wait) {
			bh = __find_get_block(device->bdev, bytenr / 4096,
					      BTRFS_SUPER_INFO_SIZE);
3267 3268 3269 3270
			if (!bh) {
				errors++;
				continue;
			}
Y
Yan Zheng 已提交
3271
			wait_on_buffer(bh);
3272 3273 3274 3275 3276 3277 3278 3279 3280
			if (!buffer_uptodate(bh))
				errors++;

			/* drop our reference */
			brelse(bh);

			/* drop the reference from the wait == 0 run */
			brelse(bh);
			continue;
Y
Yan Zheng 已提交
3281 3282 3283 3284
		} else {
			btrfs_set_super_bytenr(sb, bytenr);

			crc = ~(u32)0;
3285
			crc = btrfs_csum_data((char *)sb +
Y
Yan Zheng 已提交
3286 3287 3288 3289 3290
					      BTRFS_CSUM_SIZE, crc,
					      BTRFS_SUPER_INFO_SIZE -
					      BTRFS_CSUM_SIZE);
			btrfs_csum_final(crc, sb->csum);

3291 3292 3293 3294
			/*
			 * one reference for us, and we leave it for the
			 * caller
			 */
Y
Yan Zheng 已提交
3295 3296
			bh = __getblk(device->bdev, bytenr / 4096,
				      BTRFS_SUPER_INFO_SIZE);
3297
			if (!bh) {
3298
				printk(KERN_ERR "BTRFS: couldn't get super "
3299 3300 3301 3302 3303
				       "buffer head for bytenr %Lu\n", bytenr);
				errors++;
				continue;
			}

Y
Yan Zheng 已提交
3304 3305
			memcpy(bh->b_data, sb, BTRFS_SUPER_INFO_SIZE);

3306
			/* one reference for submit_bh */
Y
Yan Zheng 已提交
3307
			get_bh(bh);
3308 3309

			set_buffer_uptodate(bh);
Y
Yan Zheng 已提交
3310 3311
			lock_buffer(bh);
			bh->b_end_io = btrfs_end_buffer_write_sync;
3312
			bh->b_private = device;
Y
Yan Zheng 已提交
3313 3314
		}

C
Chris Mason 已提交
3315 3316 3317 3318
		/*
		 * we fua the first super.  The others we allow
		 * to go down lazy.
		 */
3319 3320 3321 3322
		if (i == 0)
			ret = btrfsic_submit_bh(WRITE_FUA, bh);
		else
			ret = btrfsic_submit_bh(WRITE_SYNC, bh);
3323
		if (ret)
Y
Yan Zheng 已提交
3324 3325 3326 3327 3328
			errors++;
	}
	return errors < i ? 0 : -1;
}

C
Chris Mason 已提交
3329 3330 3331 3332
/*
 * endio for the write_dev_flush, this will wake anyone waiting
 * for the barrier when it is done
 */
3333
static void btrfs_end_empty_barrier(struct bio *bio)
C
Chris Mason 已提交
3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361
{
	if (bio->bi_private)
		complete(bio->bi_private);
	bio_put(bio);
}

/*
 * trigger flushes for one the devices.  If you pass wait == 0, the flushes are
 * sent down.  With wait == 1, it waits for the previous flush.
 *
 * any device where the flush fails with eopnotsupp are flagged as not-barrier
 * capable
 */
static int write_dev_flush(struct btrfs_device *device, int wait)
{
	struct bio *bio;
	int ret = 0;

	if (device->nobarriers)
		return 0;

	if (wait) {
		bio = device->flush_bio;
		if (!bio)
			return 0;

		wait_for_completion(&device->flush_wait);

3362 3363
		if (bio->bi_error) {
			ret = bio->bi_error;
3364 3365
			btrfs_dev_stat_inc_and_print(device,
				BTRFS_DEV_STAT_FLUSH_ERRS);
C
Chris Mason 已提交
3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378
		}

		/* drop the reference from the wait == 0 run */
		bio_put(bio);
		device->flush_bio = NULL;

		return ret;
	}

	/*
	 * one reference for us, and we leave it for the
	 * caller
	 */
3379
	device->flush_bio = NULL;
3380
	bio = btrfs_io_bio_alloc(GFP_NOFS, 0);
C
Chris Mason 已提交
3381 3382 3383 3384 3385 3386 3387 3388 3389 3390
	if (!bio)
		return -ENOMEM;

	bio->bi_end_io = btrfs_end_empty_barrier;
	bio->bi_bdev = device->bdev;
	init_completion(&device->flush_wait);
	bio->bi_private = &device->flush_wait;
	device->flush_bio = bio;

	bio_get(bio);
3391
	btrfsic_submit_bio(WRITE_FLUSH, bio);
C
Chris Mason 已提交
3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403

	return 0;
}

/*
 * send an empty flush down to each device in parallel,
 * then wait for them
 */
static int barrier_all_devices(struct btrfs_fs_info *info)
{
	struct list_head *head;
	struct btrfs_device *dev;
3404 3405
	int errors_send = 0;
	int errors_wait = 0;
C
Chris Mason 已提交
3406 3407 3408 3409 3410
	int ret;

	/* send down all the barriers */
	head = &info->fs_devices->devices;
	list_for_each_entry_rcu(dev, head, dev_list) {
3411 3412
		if (dev->missing)
			continue;
C
Chris Mason 已提交
3413
		if (!dev->bdev) {
3414
			errors_send++;
C
Chris Mason 已提交
3415 3416 3417 3418 3419 3420 3421
			continue;
		}
		if (!dev->in_fs_metadata || !dev->writeable)
			continue;

		ret = write_dev_flush(dev, 0);
		if (ret)
3422
			errors_send++;
C
Chris Mason 已提交
3423 3424 3425 3426
	}

	/* wait for all the barriers */
	list_for_each_entry_rcu(dev, head, dev_list) {
3427 3428
		if (dev->missing)
			continue;
C
Chris Mason 已提交
3429
		if (!dev->bdev) {
3430
			errors_wait++;
C
Chris Mason 已提交
3431 3432 3433 3434 3435 3436 3437
			continue;
		}
		if (!dev->in_fs_metadata || !dev->writeable)
			continue;

		ret = write_dev_flush(dev, 1);
		if (ret)
3438
			errors_wait++;
C
Chris Mason 已提交
3439
	}
3440 3441
	if (errors_send > info->num_tolerated_disk_barrier_failures ||
	    errors_wait > info->num_tolerated_disk_barrier_failures)
C
Chris Mason 已提交
3442 3443 3444 3445
		return -EIO;
	return 0;
}

3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465
int btrfs_get_num_tolerated_disk_barrier_failures(u64 flags)
{
	if ((flags & (BTRFS_BLOCK_GROUP_DUP |
		      BTRFS_BLOCK_GROUP_RAID0 |
		      BTRFS_AVAIL_ALLOC_BIT_SINGLE)) ||
	    ((flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0))
		return 0;

	if (flags & (BTRFS_BLOCK_GROUP_RAID1 |
		     BTRFS_BLOCK_GROUP_RAID5 |
		     BTRFS_BLOCK_GROUP_RAID10))
		return 1;

	if (flags & BTRFS_BLOCK_GROUP_RAID6)
		return 2;

	pr_warn("BTRFS: unknown raid type: %llu\n", flags);
	return 0;
}

3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479
int btrfs_calc_num_tolerated_disk_barrier_failures(
	struct btrfs_fs_info *fs_info)
{
	struct btrfs_ioctl_space_info space;
	struct btrfs_space_info *sinfo;
	u64 types[] = {BTRFS_BLOCK_GROUP_DATA,
		       BTRFS_BLOCK_GROUP_SYSTEM,
		       BTRFS_BLOCK_GROUP_METADATA,
		       BTRFS_BLOCK_GROUP_DATA | BTRFS_BLOCK_GROUP_METADATA};
	int i;
	int c;
	int num_tolerated_disk_barrier_failures =
		(int)fs_info->fs_devices->num_devices;

3480
	for (i = 0; i < ARRAY_SIZE(types); i++) {
3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497
		struct btrfs_space_info *tmp;

		sinfo = NULL;
		rcu_read_lock();
		list_for_each_entry_rcu(tmp, &fs_info->space_info, list) {
			if (tmp->flags == types[i]) {
				sinfo = tmp;
				break;
			}
		}
		rcu_read_unlock();

		if (!sinfo)
			continue;

		down_read(&sinfo->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
3498 3499 3500 3501 3502 3503 3504 3505 3506 3507
			u64 flags;

			if (list_empty(&sinfo->block_groups[c]))
				continue;

			btrfs_get_block_group_info(&sinfo->block_groups[c],
						   &space);
			if (space.total_bytes == 0 || space.used_bytes == 0)
				continue;
			flags = space.flags;
3508 3509 3510 3511 3512

			num_tolerated_disk_barrier_failures = min(
				num_tolerated_disk_barrier_failures,
				btrfs_get_num_tolerated_disk_barrier_failures(
					flags));
3513 3514 3515 3516 3517 3518 3519
		}
		up_read(&sinfo->groups_sem);
	}

	return num_tolerated_disk_barrier_failures;
}

3520
static int write_all_supers(struct btrfs_root *root, int max_mirrors)
3521
{
3522
	struct list_head *head;
3523
	struct btrfs_device *dev;
3524
	struct btrfs_super_block *sb;
3525 3526 3527
	struct btrfs_dev_item *dev_item;
	int ret;
	int do_barriers;
3528 3529
	int max_errors;
	int total_errors = 0;
3530
	u64 flags;
3531 3532

	do_barriers = !btrfs_test_opt(root, NOBARRIER);
C
Chris Mason 已提交
3533
	backup_super_roots(root->fs_info);
3534

3535
	sb = root->fs_info->super_for_commit;
3536
	dev_item = &sb->dev_item;
3537

3538
	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
3539
	head = &root->fs_info->fs_devices->devices;
3540
	max_errors = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
C
Chris Mason 已提交
3541

3542 3543 3544 3545 3546 3547 3548 3549 3550 3551
	if (do_barriers) {
		ret = barrier_all_devices(root->fs_info);
		if (ret) {
			mutex_unlock(
				&root->fs_info->fs_devices->device_list_mutex);
			btrfs_error(root->fs_info, ret,
				    "errors while submitting device barriers.");
			return ret;
		}
	}
C
Chris Mason 已提交
3552

3553
	list_for_each_entry_rcu(dev, head, dev_list) {
3554 3555 3556 3557
		if (!dev->bdev) {
			total_errors++;
			continue;
		}
Y
Yan Zheng 已提交
3558
		if (!dev->in_fs_metadata || !dev->writeable)
3559 3560
			continue;

Y
Yan Zheng 已提交
3561
		btrfs_set_stack_device_generation(dev_item, 0);
3562 3563
		btrfs_set_stack_device_type(dev_item, dev->type);
		btrfs_set_stack_device_id(dev_item, dev->devid);
3564
		btrfs_set_stack_device_total_bytes(dev_item,
3565
						   dev->commit_total_bytes);
3566 3567
		btrfs_set_stack_device_bytes_used(dev_item,
						  dev->commit_bytes_used);
3568 3569 3570 3571
		btrfs_set_stack_device_io_align(dev_item, dev->io_align);
		btrfs_set_stack_device_io_width(dev_item, dev->io_width);
		btrfs_set_stack_device_sector_size(dev_item, dev->sector_size);
		memcpy(dev_item->uuid, dev->uuid, BTRFS_UUID_SIZE);
Y
Yan Zheng 已提交
3572
		memcpy(dev_item->fsid, dev->fs_devices->fsid, BTRFS_UUID_SIZE);
Y
Yan Zheng 已提交
3573

3574 3575 3576
		flags = btrfs_super_flags(sb);
		btrfs_set_super_flags(sb, flags | BTRFS_HEADER_FLAG_WRITTEN);

Y
Yan Zheng 已提交
3577
		ret = write_dev_supers(dev, sb, do_barriers, 0, max_mirrors);
3578 3579
		if (ret)
			total_errors++;
3580
	}
3581
	if (total_errors > max_errors) {
3582
		btrfs_err(root->fs_info, "%d errors while writing supers",
C
Chris Mason 已提交
3583
		       total_errors);
3584
		mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
3585

3586 3587 3588 3589
		/* FUA is masked off if unsupported and can't be the reason */
		btrfs_error(root->fs_info, -EIO,
			    "%d errors while writing supers", total_errors);
		return -EIO;
3590
	}
3591

Y
Yan Zheng 已提交
3592
	total_errors = 0;
3593
	list_for_each_entry_rcu(dev, head, dev_list) {
3594 3595
		if (!dev->bdev)
			continue;
Y
Yan Zheng 已提交
3596
		if (!dev->in_fs_metadata || !dev->writeable)
3597 3598
			continue;

Y
Yan Zheng 已提交
3599 3600 3601
		ret = write_dev_supers(dev, sb, do_barriers, 1, max_mirrors);
		if (ret)
			total_errors++;
3602
	}
3603
	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
3604
	if (total_errors > max_errors) {
3605 3606 3607
		btrfs_error(root->fs_info, -EIO,
			    "%d errors while writing supers", total_errors);
		return -EIO;
3608
	}
3609 3610 3611
	return 0;
}

Y
Yan Zheng 已提交
3612 3613
int write_ctree_super(struct btrfs_trans_handle *trans,
		      struct btrfs_root *root, int max_mirrors)
3614
{
3615
	return write_all_supers(root, max_mirrors);
C
Chris Mason 已提交
3616 3617
}

3618 3619 3620
/* Drop a fs root from the radix tree and free it. */
void btrfs_drop_and_free_fs_root(struct btrfs_fs_info *fs_info,
				  struct btrfs_root *root)
C
Chris Mason 已提交
3621
{
3622
	spin_lock(&fs_info->fs_roots_radix_lock);
C
Chris Mason 已提交
3623 3624
	radix_tree_delete(&fs_info->fs_roots_radix,
			  (unsigned long)root->root_key.objectid);
3625
	spin_unlock(&fs_info->fs_roots_radix_lock);
3626 3627 3628 3629

	if (btrfs_root_refs(&root->root_item) == 0)
		synchronize_srcu(&fs_info->subvol_srcu);

3630
	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state))
L
Liu Bo 已提交
3631 3632
		btrfs_free_log(NULL, root);

3633 3634 3635 3636
	if (root->free_ino_pinned)
		__btrfs_remove_free_space_cache(root->free_ino_pinned);
	if (root->free_ino_ctl)
		__btrfs_remove_free_space_cache(root->free_ino_ctl);
3637 3638 3639 3640 3641
	free_fs_root(root);
}

static void free_fs_root(struct btrfs_root *root)
{
3642
	iput(root->ino_cache_inode);
3643
	WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
3644 3645
	btrfs_free_block_rsv(root, root->orphan_block_rsv);
	root->orphan_block_rsv = NULL;
3646 3647
	if (root->anon_dev)
		free_anon_bdev(root->anon_dev);
3648 3649
	if (root->subv_writers)
		btrfs_free_subvolume_writers(root->subv_writers);
3650 3651
	free_extent_buffer(root->node);
	free_extent_buffer(root->commit_root);
3652 3653
	kfree(root->free_ino_ctl);
	kfree(root->free_ino_pinned);
C
Chris Mason 已提交
3654
	kfree(root->name);
3655
	btrfs_put_fs_root(root);
C
Chris Mason 已提交
3656 3657
}

3658 3659 3660
void btrfs_free_fs_root(struct btrfs_root *root)
{
	free_fs_root(root);
C
Chris Mason 已提交
3661 3662
}

Y
Yan Zheng 已提交
3663
int btrfs_cleanup_fs_roots(struct btrfs_fs_info *fs_info)
C
Chris Mason 已提交
3664
{
Y
Yan Zheng 已提交
3665 3666
	u64 root_objectid = 0;
	struct btrfs_root *gang[8];
3667 3668 3669 3670
	int i = 0;
	int err = 0;
	unsigned int ret = 0;
	int index;
3671

Y
Yan Zheng 已提交
3672
	while (1) {
3673
		index = srcu_read_lock(&fs_info->subvol_srcu);
Y
Yan Zheng 已提交
3674 3675 3676
		ret = radix_tree_gang_lookup(&fs_info->fs_roots_radix,
					     (void **)gang, root_objectid,
					     ARRAY_SIZE(gang));
3677 3678
		if (!ret) {
			srcu_read_unlock(&fs_info->subvol_srcu, index);
Y
Yan Zheng 已提交
3679
			break;
3680
		}
3681
		root_objectid = gang[ret - 1]->root_key.objectid + 1;
3682

Y
Yan Zheng 已提交
3683
		for (i = 0; i < ret; i++) {
3684 3685 3686 3687 3688 3689 3690 3691 3692
			/* Avoid to grab roots in dead_roots */
			if (btrfs_root_refs(&gang[i]->root_item) == 0) {
				gang[i] = NULL;
				continue;
			}
			/* grab all the search result for later use */
			gang[i] = btrfs_grab_fs_root(gang[i]);
		}
		srcu_read_unlock(&fs_info->subvol_srcu, index);
3693

3694 3695 3696
		for (i = 0; i < ret; i++) {
			if (!gang[i])
				continue;
Y
Yan Zheng 已提交
3697
			root_objectid = gang[i]->root_key.objectid;
3698 3699
			err = btrfs_orphan_cleanup(gang[i]);
			if (err)
3700 3701
				break;
			btrfs_put_fs_root(gang[i]);
Y
Yan Zheng 已提交
3702 3703 3704
		}
		root_objectid++;
	}
3705 3706 3707 3708 3709 3710 3711

	/* release the uncleaned roots due to error */
	for (; i < ret; i++) {
		if (gang[i])
			btrfs_put_fs_root(gang[i]);
	}
	return err;
Y
Yan Zheng 已提交
3712
}
3713

Y
Yan Zheng 已提交
3714 3715 3716
int btrfs_commit_super(struct btrfs_root *root)
{
	struct btrfs_trans_handle *trans;
3717

Y
Yan Zheng 已提交
3718
	mutex_lock(&root->fs_info->cleaner_mutex);
Y
Yan, Zheng 已提交
3719
	btrfs_run_delayed_iputs(root);
Y
Yan Zheng 已提交
3720
	mutex_unlock(&root->fs_info->cleaner_mutex);
D
David Sterba 已提交
3721
	wake_up_process(root->fs_info->cleaner_kthread);
3722 3723 3724 3725 3726

	/* wait until ongoing cleanup work done */
	down_write(&root->fs_info->cleanup_work_sem);
	up_write(&root->fs_info->cleanup_work_sem);

3727
	trans = btrfs_join_transaction(root);
3728 3729
	if (IS_ERR(trans))
		return PTR_ERR(trans);
3730
	return btrfs_commit_transaction(trans, root);
Y
Yan Zheng 已提交
3731 3732
}

3733
void close_ctree(struct btrfs_root *root)
Y
Yan Zheng 已提交
3734 3735 3736 3737 3738 3739 3740
{
	struct btrfs_fs_info *fs_info = root->fs_info;
	int ret;

	fs_info->closing = 1;
	smp_mb();

S
Stefan Behrens 已提交
3741 3742 3743 3744 3745
	/* wait for the uuid_scan task to finish */
	down(&fs_info->uuid_tree_rescan_sem);
	/* avoid complains from lockdep et al., set sem back to initial state */
	up(&fs_info->uuid_tree_rescan_sem);

3746
	/* pause restriper - we want to resume on mount */
3747
	btrfs_pause_balance(fs_info);
3748

3749 3750
	btrfs_dev_replace_suspend_for_unmount(fs_info);

3751
	btrfs_scrub_cancel(fs_info);
C
Chris Mason 已提交
3752 3753 3754 3755 3756 3757

	/* wait for any defraggers to finish */
	wait_event(fs_info->transaction_wait,
		   (atomic_read(&fs_info->defrag_running) == 0));

	/* clear out the rbtree of defraggable inodes */
3758
	btrfs_cleanup_defrag_inodes(fs_info);
C
Chris Mason 已提交
3759

3760 3761
	cancel_work_sync(&fs_info->async_reclaim_work);

Y
Yan Zheng 已提交
3762
	if (!(fs_info->sb->s_flags & MS_RDONLY)) {
3763 3764 3765 3766 3767 3768 3769
		/*
		 * If the cleaner thread is stopped and there are
		 * block groups queued for removal, the deletion will be
		 * skipped when we quit the cleaner thread.
		 */
		btrfs_delete_unused_bgs(root->fs_info);

L
liubo 已提交
3770 3771
		ret = btrfs_commit_super(root);
		if (ret)
3772
			btrfs_err(fs_info, "commit super ret %d", ret);
L
liubo 已提交
3773 3774
	}

3775
	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state))
3776
		btrfs_error_commit_super(root);
3777

A
Al Viro 已提交
3778 3779
	kthread_stop(fs_info->transaction_kthread);
	kthread_stop(fs_info->cleaner_kthread);
3780

3781 3782 3783
	fs_info->closing = 2;
	smp_mb();

3784
	btrfs_free_qgroup_config(fs_info);
3785

3786
	if (percpu_counter_sum(&fs_info->delalloc_bytes)) {
3787
		btrfs_info(fs_info, "at unmount delalloc count %lld",
3788
		       percpu_counter_sum(&fs_info->delalloc_bytes));
C
Chris Mason 已提交
3789
	}
3790

3791
	btrfs_sysfs_remove_mounted(fs_info);
3792
	btrfs_sysfs_remove_fsid(fs_info->fs_devices);
3793

3794
	btrfs_free_fs_roots(fs_info);
3795

3796 3797
	btrfs_put_block_group_cache(fs_info);

3798 3799
	btrfs_free_block_groups(fs_info);

3800 3801 3802 3803 3804
	/*
	 * we must make sure there is not any read request to
	 * submit after we stopping all workers.
	 */
	invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
3805 3806
	btrfs_stop_all_workers(fs_info);

3807
	fs_info->open = 0;
3808
	free_root_pointers(fs_info, 1);
3809

3810
	iput(fs_info->btree_inode);
3811

3812 3813 3814 3815 3816
#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
	if (btrfs_test_opt(root, CHECK_INTEGRITY))
		btrfsic_unmount(root, fs_info->fs_devices);
#endif

3817
	btrfs_close_devices(fs_info->fs_devices);
3818
	btrfs_mapping_tree_free(&fs_info->mapping_tree);
3819

3820
	percpu_counter_destroy(&fs_info->dirty_metadata_bytes);
3821
	percpu_counter_destroy(&fs_info->delalloc_bytes);
3822
	percpu_counter_destroy(&fs_info->bio_counter);
C
Chris Mason 已提交
3823
	bdi_destroy(&fs_info->bdi);
3824
	cleanup_srcu_struct(&fs_info->subvol_srcu);
3825

D
David Woodhouse 已提交
3826 3827
	btrfs_free_stripe_hash_table(fs_info);

3828
	__btrfs_free_block_rsv(root->orphan_block_rsv);
3829
	root->orphan_block_rsv = NULL;
3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840

	lock_chunks(root);
	while (!list_empty(&fs_info->pinned_chunks)) {
		struct extent_map *em;

		em = list_first_entry(&fs_info->pinned_chunks,
				      struct extent_map, list);
		list_del_init(&em->list);
		free_extent_map(em);
	}
	unlock_chunks(root);
3841 3842
}

3843 3844
int btrfs_buffer_uptodate(struct extent_buffer *buf, u64 parent_transid,
			  int atomic)
3845
{
3846
	int ret;
3847
	struct inode *btree_inode = buf->pages[0]->mapping->host;
3848

3849
	ret = extent_buffer_uptodate(buf);
3850 3851 3852 3853
	if (!ret)
		return ret;

	ret = verify_parent_transid(&BTRFS_I(btree_inode)->io_tree, buf,
3854 3855 3856
				    parent_transid, atomic);
	if (ret == -EAGAIN)
		return ret;
3857
	return !ret;
3858 3859 3860
}

int btrfs_set_buffer_uptodate(struct extent_buffer *buf)
C
Chris Mason 已提交
3861
{
3862
	return set_extent_buffer_uptodate(buf);
3863
}
3864

3865 3866
void btrfs_mark_buffer_dirty(struct extent_buffer *buf)
{
3867
	struct btrfs_root *root;
3868
	u64 transid = btrfs_header_generation(buf);
3869
	int was_dirty;
3870

3871 3872 3873 3874 3875 3876 3877 3878 3879 3880
#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
	/*
	 * This is a fast path so only do this check if we have sanity tests
	 * enabled.  Normal people shouldn't be marking dummy buffers as dirty
	 * outside of the sanity tests.
	 */
	if (unlikely(test_bit(EXTENT_BUFFER_DUMMY, &buf->bflags)))
		return;
#endif
	root = BTRFS_I(buf->pages[0]->mapping->host)->root;
3881
	btrfs_assert_tree_locked(buf);
J
Julia Lawall 已提交
3882 3883
	if (transid != root->fs_info->generation)
		WARN(1, KERN_CRIT "btrfs transid mismatch buffer %llu, "
C
Chris Mason 已提交
3884
		       "found %llu running %llu\n",
3885
			buf->start, transid, root->fs_info->generation);
3886
	was_dirty = set_extent_buffer_dirty(buf);
3887 3888 3889 3890
	if (!was_dirty)
		__percpu_counter_add(&root->fs_info->dirty_metadata_bytes,
				     buf->len,
				     root->fs_info->dirty_metadata_batch);
3891 3892 3893 3894 3895 3896
#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
	if (btrfs_header_level(buf) == 0 && check_leaf(root, buf)) {
		btrfs_print_leaf(root, buf);
		ASSERT(0);
	}
#endif
3897 3898
}

3899 3900
static void __btrfs_btree_balance_dirty(struct btrfs_root *root,
					int flush_delayed)
3901 3902 3903 3904 3905
{
	/*
	 * looks as though older kernels can get into trouble with
	 * this code, they end up stuck in balance_dirty_pages forever
	 */
3906
	int ret;
3907 3908 3909 3910

	if (current->flags & PF_MEMALLOC)
		return;

3911 3912
	if (flush_delayed)
		btrfs_balance_delayed_items(root);
3913

3914 3915 3916
	ret = percpu_counter_compare(&root->fs_info->dirty_metadata_bytes,
				     BTRFS_DIRTY_METADATA_THRESH);
	if (ret > 0) {
3917 3918
		balance_dirty_pages_ratelimited(
				   root->fs_info->btree_inode->i_mapping);
3919 3920 3921 3922
	}
	return;
}

3923
void btrfs_btree_balance_dirty(struct btrfs_root *root)
C
Chris Mason 已提交
3924
{
3925 3926
	__btrfs_btree_balance_dirty(root, 1);
}
3927

3928 3929 3930
void btrfs_btree_balance_dirty_nodelay(struct btrfs_root *root)
{
	__btrfs_btree_balance_dirty(root, 0);
C
Chris Mason 已提交
3931
}
3932

3933
int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid)
3934
{
3935
	struct btrfs_root *root = BTRFS_I(buf->pages[0]->mapping->host)->root;
3936
	return btree_read_extent_buffer_pages(root, buf, 0, parent_transid);
3937
}
3938

3939
static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
L
liubo 已提交
3940 3941
			      int read_only)
{
D
David Sterba 已提交
3942 3943 3944
	struct btrfs_super_block *sb = fs_info->super_copy;
	int ret = 0;

3945 3946 3947
	if (btrfs_super_root_level(sb) >= BTRFS_MAX_LEVEL) {
		printk(KERN_ERR "BTRFS: tree_root level too big: %d >= %d\n",
				btrfs_super_root_level(sb), BTRFS_MAX_LEVEL);
D
David Sterba 已提交
3948 3949
		ret = -EINVAL;
	}
3950 3951 3952
	if (btrfs_super_chunk_root_level(sb) >= BTRFS_MAX_LEVEL) {
		printk(KERN_ERR "BTRFS: chunk_root level too big: %d >= %d\n",
				btrfs_super_chunk_root_level(sb), BTRFS_MAX_LEVEL);
D
David Sterba 已提交
3953 3954
		ret = -EINVAL;
	}
3955 3956 3957
	if (btrfs_super_log_root_level(sb) >= BTRFS_MAX_LEVEL) {
		printk(KERN_ERR "BTRFS: log_root level too big: %d >= %d\n",
				btrfs_super_log_root_level(sb), BTRFS_MAX_LEVEL);
D
David Sterba 已提交
3958 3959 3960
		ret = -EINVAL;
	}

D
David Sterba 已提交
3961
	/*
D
David Sterba 已提交
3962 3963
	 * The common minimum, we don't know if we can trust the nodesize/sectorsize
	 * items yet, they'll be verified later. Issue just a warning.
D
David Sterba 已提交
3964
	 */
3965
	if (!IS_ALIGNED(btrfs_super_root(sb), 4096))
D
David Sterba 已提交
3966
		printk(KERN_WARNING "BTRFS: tree_root block unaligned: %llu\n",
3967
				btrfs_super_root(sb));
3968
	if (!IS_ALIGNED(btrfs_super_chunk_root(sb), 4096))
3969 3970
		printk(KERN_WARNING "BTRFS: chunk_root block unaligned: %llu\n",
				btrfs_super_chunk_root(sb));
3971
	if (!IS_ALIGNED(btrfs_super_log_root(sb), 4096))
3972
		printk(KERN_WARNING "BTRFS: log_root block unaligned: %llu\n",
3973
				btrfs_super_log_root(sb));
D
David Sterba 已提交
3974

3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989
	/*
	 * Check the lower bound, the alignment and other constraints are
	 * checked later.
	 */
	if (btrfs_super_nodesize(sb) < 4096) {
		printk(KERN_ERR "BTRFS: nodesize too small: %u < 4096\n",
				btrfs_super_nodesize(sb));
		ret = -EINVAL;
	}
	if (btrfs_super_sectorsize(sb) < 4096) {
		printk(KERN_ERR "BTRFS: sectorsize too small: %u < 4096\n",
				btrfs_super_sectorsize(sb));
		ret = -EINVAL;
	}

D
David Sterba 已提交
3990 3991 3992 3993 3994 3995 3996 3997 3998 3999
	if (memcmp(fs_info->fsid, sb->dev_item.fsid, BTRFS_UUID_SIZE) != 0) {
		printk(KERN_ERR "BTRFS: dev_item UUID does not match fsid: %pU != %pU\n",
				fs_info->fsid, sb->dev_item.fsid);
		ret = -EINVAL;
	}

	/*
	 * Hint to catch really bogus numbers, bitflips or so, more exact checks are
	 * done later
	 */
4000
	if (btrfs_super_num_devices(sb) > (1UL << 31))
D
David Sterba 已提交
4001
		printk(KERN_WARNING "BTRFS: suspicious number of devices: %llu\n",
4002
				btrfs_super_num_devices(sb));
4003 4004 4005 4006
	if (btrfs_super_num_devices(sb) == 0) {
		printk(KERN_ERR "BTRFS: number of devices is 0\n");
		ret = -EINVAL;
	}
D
David Sterba 已提交
4007

4008
	if (btrfs_super_bytenr(sb) != BTRFS_SUPER_INFO_OFFSET) {
D
David Sterba 已提交
4009
		printk(KERN_ERR "BTRFS: super offset mismatch %llu != %u\n",
4010
				btrfs_super_bytenr(sb), BTRFS_SUPER_INFO_OFFSET);
D
David Sterba 已提交
4011 4012 4013
		ret = -EINVAL;
	}

4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025
	/*
	 * Obvious sys_chunk_array corruptions, it must hold at least one key
	 * and one chunk
	 */
	if (btrfs_super_sys_array_size(sb) > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) {
		printk(KERN_ERR "BTRFS: system chunk array too big %u > %u\n",
				btrfs_super_sys_array_size(sb),
				BTRFS_SYSTEM_CHUNK_ARRAY_SIZE);
		ret = -EINVAL;
	}
	if (btrfs_super_sys_array_size(sb) < sizeof(struct btrfs_disk_key)
			+ sizeof(struct btrfs_chunk)) {
4026
		printk(KERN_ERR "BTRFS: system chunk array too small %u < %zu\n",
4027 4028 4029 4030 4031 4032
				btrfs_super_sys_array_size(sb),
				sizeof(struct btrfs_disk_key)
				+ sizeof(struct btrfs_chunk));
		ret = -EINVAL;
	}

D
David Sterba 已提交
4033 4034 4035 4036
	/*
	 * The generation is a global counter, we'll trust it more than the others
	 * but it's still possible that it's the one that's wrong.
	 */
4037
	if (btrfs_super_generation(sb) < btrfs_super_chunk_root_generation(sb))
D
David Sterba 已提交
4038 4039
		printk(KERN_WARNING
			"BTRFS: suspicious: generation < chunk_root_generation: %llu < %llu\n",
4040 4041 4042
			btrfs_super_generation(sb), btrfs_super_chunk_root_generation(sb));
	if (btrfs_super_generation(sb) < btrfs_super_cache_generation(sb)
	    && btrfs_super_cache_generation(sb) != (u64)-1)
D
David Sterba 已提交
4043 4044
		printk(KERN_WARNING
			"BTRFS: suspicious: generation < cache_generation: %llu < %llu\n",
4045
			btrfs_super_generation(sb), btrfs_super_cache_generation(sb));
D
David Sterba 已提交
4046 4047

	return ret;
L
liubo 已提交
4048 4049
}

4050
static void btrfs_error_commit_super(struct btrfs_root *root)
L
liubo 已提交
4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062
{
	mutex_lock(&root->fs_info->cleaner_mutex);
	btrfs_run_delayed_iputs(root);
	mutex_unlock(&root->fs_info->cleaner_mutex);

	down_write(&root->fs_info->cleanup_work_sem);
	up_write(&root->fs_info->cleanup_work_sem);

	/* cleanup FS via transaction */
	btrfs_cleanup_transaction(root);
}

4063
static void btrfs_destroy_ordered_extents(struct btrfs_root *root)
L
liubo 已提交
4064 4065 4066
{
	struct btrfs_ordered_extent *ordered;

4067
	spin_lock(&root->ordered_extent_lock);
4068 4069 4070 4071
	/*
	 * This will just short circuit the ordered completion stuff which will
	 * make sure the ordered extent gets properly cleaned up.
	 */
4072
	list_for_each_entry(ordered, &root->ordered_extents,
4073 4074
			    root_extent_list)
		set_bit(BTRFS_ORDERED_IOERR, &ordered->flags);
4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089
	spin_unlock(&root->ordered_extent_lock);
}

static void btrfs_destroy_all_ordered_extents(struct btrfs_fs_info *fs_info)
{
	struct btrfs_root *root;
	struct list_head splice;

	INIT_LIST_HEAD(&splice);

	spin_lock(&fs_info->ordered_root_lock);
	list_splice_init(&fs_info->ordered_roots, &splice);
	while (!list_empty(&splice)) {
		root = list_first_entry(&splice, struct btrfs_root,
					ordered_root);
4090 4091
		list_move_tail(&root->ordered_root,
			       &fs_info->ordered_roots);
4092

4093
		spin_unlock(&fs_info->ordered_root_lock);
4094 4095
		btrfs_destroy_ordered_extents(root);

4096 4097
		cond_resched();
		spin_lock(&fs_info->ordered_root_lock);
4098 4099
	}
	spin_unlock(&fs_info->ordered_root_lock);
L
liubo 已提交
4100 4101
}

4102 4103
static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
				      struct btrfs_root *root)
L
liubo 已提交
4104 4105 4106 4107 4108 4109 4110 4111 4112
{
	struct rb_node *node;
	struct btrfs_delayed_ref_root *delayed_refs;
	struct btrfs_delayed_ref_node *ref;
	int ret = 0;

	delayed_refs = &trans->delayed_refs;

	spin_lock(&delayed_refs->lock);
4113
	if (atomic_read(&delayed_refs->num_entries) == 0) {
4114
		spin_unlock(&delayed_refs->lock);
4115
		btrfs_info(root->fs_info, "delayed_refs has NO entry");
L
liubo 已提交
4116 4117 4118
		return ret;
	}

4119 4120
	while ((node = rb_first(&delayed_refs->href_root)) != NULL) {
		struct btrfs_delayed_ref_head *head;
4121
		struct btrfs_delayed_ref_node *tmp;
4122
		bool pin_bytes = false;
L
liubo 已提交
4123

4124 4125 4126 4127 4128
		head = rb_entry(node, struct btrfs_delayed_ref_head,
				href_node);
		if (!mutex_trylock(&head->mutex)) {
			atomic_inc(&head->node.refs);
			spin_unlock(&delayed_refs->lock);
4129

4130
			mutex_lock(&head->mutex);
4131
			mutex_unlock(&head->mutex);
4132 4133 4134 4135 4136
			btrfs_put_delayed_ref(&head->node);
			spin_lock(&delayed_refs->lock);
			continue;
		}
		spin_lock(&head->lock);
4137 4138
		list_for_each_entry_safe_reverse(ref, tmp, &head->ref_list,
						 list) {
4139
			ref->in_tree = 0;
4140
			list_del(&ref->list);
4141 4142
			atomic_dec(&delayed_refs->num_entries);
			btrfs_put_delayed_ref(ref);
4143
		}
4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155
		if (head->must_insert_reserved)
			pin_bytes = true;
		btrfs_free_delayed_extent_op(head->extent_op);
		delayed_refs->num_heads--;
		if (head->processing == 0)
			delayed_refs->num_heads_ready--;
		atomic_dec(&delayed_refs->num_entries);
		head->node.in_tree = 0;
		rb_erase(&head->href_node, &delayed_refs->href_root);
		spin_unlock(&head->lock);
		spin_unlock(&delayed_refs->lock);
		mutex_unlock(&head->mutex);
L
liubo 已提交
4156

4157 4158 4159 4160
		if (pin_bytes)
			btrfs_pin_extent(root, head->node.bytenr,
					 head->node.num_bytes, 1);
		btrfs_put_delayed_ref(&head->node);
L
liubo 已提交
4161 4162 4163 4164 4165 4166 4167 4168 4169
		cond_resched();
		spin_lock(&delayed_refs->lock);
	}

	spin_unlock(&delayed_refs->lock);

	return ret;
}

4170
static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root)
L
liubo 已提交
4171 4172 4173 4174 4175 4176
{
	struct btrfs_inode *btrfs_inode;
	struct list_head splice;

	INIT_LIST_HEAD(&splice);

4177 4178
	spin_lock(&root->delalloc_lock);
	list_splice_init(&root->delalloc_inodes, &splice);
L
liubo 已提交
4179 4180

	while (!list_empty(&splice)) {
4181 4182
		btrfs_inode = list_first_entry(&splice, struct btrfs_inode,
					       delalloc_inodes);
L
liubo 已提交
4183 4184

		list_del_init(&btrfs_inode->delalloc_inodes);
4185 4186
		clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
			  &btrfs_inode->runtime_flags);
4187
		spin_unlock(&root->delalloc_lock);
L
liubo 已提交
4188 4189

		btrfs_invalidate_inodes(btrfs_inode->root);
4190

4191
		spin_lock(&root->delalloc_lock);
L
liubo 已提交
4192 4193
	}

4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219
	spin_unlock(&root->delalloc_lock);
}

static void btrfs_destroy_all_delalloc_inodes(struct btrfs_fs_info *fs_info)
{
	struct btrfs_root *root;
	struct list_head splice;

	INIT_LIST_HEAD(&splice);

	spin_lock(&fs_info->delalloc_root_lock);
	list_splice_init(&fs_info->delalloc_roots, &splice);
	while (!list_empty(&splice)) {
		root = list_first_entry(&splice, struct btrfs_root,
					 delalloc_root);
		list_del_init(&root->delalloc_root);
		root = btrfs_grab_fs_root(root);
		BUG_ON(!root);
		spin_unlock(&fs_info->delalloc_root_lock);

		btrfs_destroy_delalloc_inodes(root);
		btrfs_put_fs_root(root);

		spin_lock(&fs_info->delalloc_root_lock);
	}
	spin_unlock(&fs_info->delalloc_root_lock);
L
liubo 已提交
4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232
}

static int btrfs_destroy_marked_extents(struct btrfs_root *root,
					struct extent_io_tree *dirty_pages,
					int mark)
{
	int ret;
	struct extent_buffer *eb;
	u64 start = 0;
	u64 end;

	while (1) {
		ret = find_first_extent_bit(dirty_pages, start, &start, &end,
4233
					    mark, NULL);
L
liubo 已提交
4234 4235 4236 4237 4238
		if (ret)
			break;

		clear_extent_bits(dirty_pages, start, end, mark, GFP_NOFS);
		while (start <= end) {
4239
			eb = btrfs_find_tree_block(root->fs_info, start);
4240
			start += root->nodesize;
4241
			if (!eb)
L
liubo 已提交
4242
				continue;
4243
			wait_on_extent_buffer_writeback(eb);
L
liubo 已提交
4244

4245 4246 4247 4248
			if (test_and_clear_bit(EXTENT_BUFFER_DIRTY,
					       &eb->bflags))
				clear_extent_buffer_dirty(eb);
			free_extent_buffer_stale(eb);
L
liubo 已提交
4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261
		}
	}

	return ret;
}

static int btrfs_destroy_pinned_extent(struct btrfs_root *root,
				       struct extent_io_tree *pinned_extents)
{
	struct extent_io_tree *unpin;
	u64 start;
	u64 end;
	int ret;
4262
	bool loop = true;
L
liubo 已提交
4263 4264

	unpin = pinned_extents;
4265
again:
L
liubo 已提交
4266 4267
	while (1) {
		ret = find_first_extent_bit(unpin, 0, &start, &end,
4268
					    EXTENT_DIRTY, NULL);
L
liubo 已提交
4269 4270 4271 4272 4273 4274 4275 4276
		if (ret)
			break;

		clear_extent_dirty(unpin, start, end, GFP_NOFS);
		btrfs_error_unpin_extent_range(root, start, end);
		cond_resched();
	}

4277 4278 4279 4280 4281 4282 4283 4284 4285
	if (loop) {
		if (unpin == &root->fs_info->freed_extents[0])
			unpin = &root->fs_info->freed_extents[1];
		else
			unpin = &root->fs_info->freed_extents[0];
		loop = false;
		goto again;
	}

L
liubo 已提交
4286 4287 4288
	return 0;
}

4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307
static void btrfs_free_pending_ordered(struct btrfs_transaction *cur_trans,
				       struct btrfs_fs_info *fs_info)
{
	struct btrfs_ordered_extent *ordered;

	spin_lock(&fs_info->trans_lock);
	while (!list_empty(&cur_trans->pending_ordered)) {
		ordered = list_first_entry(&cur_trans->pending_ordered,
					   struct btrfs_ordered_extent,
					   trans_list);
		list_del_init(&ordered->trans_list);
		spin_unlock(&fs_info->trans_lock);

		btrfs_put_ordered_extent(ordered);
		spin_lock(&fs_info->trans_lock);
	}
	spin_unlock(&fs_info->trans_lock);
}

4308 4309 4310 4311 4312
void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
				   struct btrfs_root *root)
{
	btrfs_destroy_delayed_refs(cur_trans, root);

4313
	cur_trans->state = TRANS_STATE_COMMIT_START;
4314
	wake_up(&root->fs_info->transaction_blocked_wait);
4315

4316
	cur_trans->state = TRANS_STATE_UNBLOCKED;
4317
	wake_up(&root->fs_info->transaction_wait);
4318

4319
	btrfs_free_pending_ordered(cur_trans, root->fs_info);
4320 4321
	btrfs_destroy_delayed_inodes(root);
	btrfs_assert_delayed_root_empty(root);
4322 4323 4324

	btrfs_destroy_marked_extents(root, &cur_trans->dirty_pages,
				     EXTENT_DIRTY);
4325 4326
	btrfs_destroy_pinned_extent(root,
				    root->fs_info->pinned_extents);
4327

4328 4329 4330
	cur_trans->state =TRANS_STATE_COMPLETED;
	wake_up(&cur_trans->commit_wait);

4331 4332 4333 4334 4335 4336
	/*
	memset(cur_trans, 0, sizeof(*cur_trans));
	kmem_cache_free(btrfs_transaction_cachep, cur_trans);
	*/
}

4337
static int btrfs_cleanup_transaction(struct btrfs_root *root)
L
liubo 已提交
4338 4339 4340 4341 4342
{
	struct btrfs_transaction *t;

	mutex_lock(&root->fs_info->transaction_kthread_mutex);

J
Josef Bacik 已提交
4343
	spin_lock(&root->fs_info->trans_lock);
4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367
	while (!list_empty(&root->fs_info->trans_list)) {
		t = list_first_entry(&root->fs_info->trans_list,
				     struct btrfs_transaction, list);
		if (t->state >= TRANS_STATE_COMMIT_START) {
			atomic_inc(&t->use_count);
			spin_unlock(&root->fs_info->trans_lock);
			btrfs_wait_for_commit(root, t->transid);
			btrfs_put_transaction(t);
			spin_lock(&root->fs_info->trans_lock);
			continue;
		}
		if (t == root->fs_info->running_transaction) {
			t->state = TRANS_STATE_COMMIT_DOING;
			spin_unlock(&root->fs_info->trans_lock);
			/*
			 * We wait for 0 num_writers since we don't hold a trans
			 * handle open currently for this transaction.
			 */
			wait_event(t->writer_wait,
				   atomic_read(&t->num_writers) == 0);
		} else {
			spin_unlock(&root->fs_info->trans_lock);
		}
		btrfs_cleanup_one_transaction(t, root);
4368

4369 4370 4371
		spin_lock(&root->fs_info->trans_lock);
		if (t == root->fs_info->running_transaction)
			root->fs_info->running_transaction = NULL;
L
liubo 已提交
4372
		list_del_init(&t->list);
4373
		spin_unlock(&root->fs_info->trans_lock);
L
liubo 已提交
4374

4375 4376 4377 4378 4379 4380 4381 4382 4383 4384
		btrfs_put_transaction(t);
		trace_btrfs_transaction_commit(root);
		spin_lock(&root->fs_info->trans_lock);
	}
	spin_unlock(&root->fs_info->trans_lock);
	btrfs_destroy_all_ordered_extents(root->fs_info);
	btrfs_destroy_delayed_inodes(root);
	btrfs_assert_delayed_root_empty(root);
	btrfs_destroy_pinned_extent(root, root->fs_info->pinned_extents);
	btrfs_destroy_all_delalloc_inodes(root->fs_info);
L
liubo 已提交
4385 4386 4387 4388 4389
	mutex_unlock(&root->fs_info->transaction_kthread_mutex);

	return 0;
}

4390
static const struct extent_io_ops btree_extent_io_ops = {
4391
	.readpage_end_io_hook = btree_readpage_end_io_hook,
A
Arne Jansen 已提交
4392
	.readpage_io_failed_hook = btree_io_failed_hook,
4393
	.submit_bio_hook = btree_submit_bio_hook,
4394 4395
	/* note we're sharing with inode.c for the merge bio hook */
	.merge_bio_hook = btrfs_merge_bio_hook,
4396
};