tree-checker.c 20.1 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*
 * Copyright (C) Qu Wenruo 2017.  All rights reserved.
 */

/*
 * The module is used to catch unexpected/corrupted tree block data.
 * Such behavior can be caused either by a fuzzed image or bugs.
 *
 * The objective is to do leaf/node validation checks when tree block is read
 * from disk, and check *every* possible member, so other code won't
 * need to checking them again.
 *
 * Due to the potential and unwanted damage, every checker needs to be
 * carefully reviewed otherwise so it does not prevent mount of valid images.
 */

#include "ctree.h"
#include "tree-checker.h"
#include "disk-io.h"
#include "compression.h"
22
#include "volumes.h"
23

24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
/*
 * Error message should follow the following format:
 * corrupt <type>: <identifier>, <reason>[, <bad_value>]
 *
 * @type:	leaf or node
 * @identifier:	the necessary info to locate the leaf/node.
 * 		It's recommened to decode key.objecitd/offset if it's
 * 		meaningful.
 * @reason:	describe the error
 * @bad_value:	optional, it's recommened to output bad value and its
 *		expected value (range).
 *
 * Since comma is used to separate the components, only space is allowed
 * inside each component.
 */

/*
 * Append generic "corrupt leaf/node root=%llu block=%llu slot=%d: " to @fmt.
 * Allows callers to customize the output.
 */
__printf(4, 5)
45
__cold
46
static void generic_err(const struct btrfs_fs_info *fs_info,
47 48 49 50 51 52 53 54 55 56 57
			const struct extent_buffer *eb, int slot,
			const char *fmt, ...)
{
	struct va_format vaf;
	va_list args;

	va_start(args, fmt);

	vaf.fmt = fmt;
	vaf.va = &args;

58
	btrfs_crit(fs_info,
59 60
		"corrupt %s: root=%llu block=%llu slot=%d, %pV",
		btrfs_header_level(eb) == 0 ? "leaf" : "node",
61
		btrfs_header_owner(eb), btrfs_header_bytenr(eb), slot, &vaf);
62 63 64
	va_end(args);
}

65 66 67 68 69
/*
 * Customized reporter for extent data item, since its key objectid and
 * offset has its own meaning.
 */
__printf(4, 5)
70
__cold
71
static void file_extent_err(const struct btrfs_fs_info *fs_info,
72 73 74 75 76 77 78 79 80 81 82 83 84
			    const struct extent_buffer *eb, int slot,
			    const char *fmt, ...)
{
	struct btrfs_key key;
	struct va_format vaf;
	va_list args;

	btrfs_item_key_to_cpu(eb, &key, slot);
	va_start(args, fmt);

	vaf.fmt = fmt;
	vaf.va = &args;

85
	btrfs_crit(fs_info,
86
	"corrupt %s: root=%llu block=%llu slot=%d ino=%llu file_offset=%llu, %pV",
87 88 89
		btrfs_header_level(eb) == 0 ? "leaf" : "node",
		btrfs_header_owner(eb), btrfs_header_bytenr(eb), slot,
		key.objectid, key.offset, &vaf);
90 91 92 93 94 95 96
	va_end(args);
}

/*
 * Return 0 if the btrfs_file_extent_##name is aligned to @alignment
 * Else return 1
 */
97
#define CHECK_FE_ALIGNED(fs_info, leaf, slot, fi, name, alignment)	      \
98 99
({									      \
	if (!IS_ALIGNED(btrfs_file_extent_##name((leaf), (fi)), (alignment))) \
100
		file_extent_err((fs_info), (leaf), (slot),		      \
101 102 103 104 105 106
	"invalid %s for file extent, have %llu, should be aligned to %u",     \
			(#name), btrfs_file_extent_##name((leaf), (fi)),      \
			(alignment));					      \
	(!IS_ALIGNED(btrfs_file_extent_##name((leaf), (fi)), (alignment)));   \
})

107
static int check_extent_data_item(struct btrfs_fs_info *fs_info,
108 109 110 111
				  struct extent_buffer *leaf,
				  struct btrfs_key *key, int slot)
{
	struct btrfs_file_extent_item *fi;
112
	u32 sectorsize = fs_info->sectorsize;
113 114 115
	u32 item_size = btrfs_item_size_nr(leaf, slot);

	if (!IS_ALIGNED(key->offset, sectorsize)) {
116
		file_extent_err(fs_info, leaf, slot,
117 118
"unaligned file_offset for file extent, have %llu should be aligned to %u",
			key->offset, sectorsize);
119 120 121 122 123 124
		return -EUCLEAN;
	}

	fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);

	if (btrfs_file_extent_type(leaf, fi) > BTRFS_FILE_EXTENT_TYPES) {
125
		file_extent_err(fs_info, leaf, slot,
126 127 128
		"invalid type for file extent, have %u expect range [0, %u]",
			btrfs_file_extent_type(leaf, fi),
			BTRFS_FILE_EXTENT_TYPES);
129 130 131 132 133 134 135 136
		return -EUCLEAN;
	}

	/*
	 * Support for new compression/encrption must introduce incompat flag,
	 * and must be caught in open_ctree().
	 */
	if (btrfs_file_extent_compression(leaf, fi) > BTRFS_COMPRESS_TYPES) {
137
		file_extent_err(fs_info, leaf, slot,
138 139 140
	"invalid compression for file extent, have %u expect range [0, %u]",
			btrfs_file_extent_compression(leaf, fi),
			BTRFS_COMPRESS_TYPES);
141 142 143
		return -EUCLEAN;
	}
	if (btrfs_file_extent_encryption(leaf, fi)) {
144
		file_extent_err(fs_info, leaf, slot,
145 146
			"invalid encryption for file extent, have %u expect 0",
			btrfs_file_extent_encryption(leaf, fi));
147 148 149 150 151
		return -EUCLEAN;
	}
	if (btrfs_file_extent_type(leaf, fi) == BTRFS_FILE_EXTENT_INLINE) {
		/* Inline extent must have 0 as key offset */
		if (key->offset) {
152
			file_extent_err(fs_info, leaf, slot,
153 154
		"invalid file_offset for inline file extent, have %llu expect 0",
				key->offset);
155 156 157 158 159 160 161 162 163 164 165
			return -EUCLEAN;
		}

		/* Compressed inline extent has no on-disk size, skip it */
		if (btrfs_file_extent_compression(leaf, fi) !=
		    BTRFS_COMPRESS_NONE)
			return 0;

		/* Uncompressed inline extent size must match item size */
		if (item_size != BTRFS_FILE_EXTENT_INLINE_DATA_START +
		    btrfs_file_extent_ram_bytes(leaf, fi)) {
166
			file_extent_err(fs_info, leaf, slot,
167 168 169
	"invalid ram_bytes for uncompressed inline extent, have %u expect %llu",
				item_size, BTRFS_FILE_EXTENT_INLINE_DATA_START +
				btrfs_file_extent_ram_bytes(leaf, fi));
170 171 172 173 174 175 176
			return -EUCLEAN;
		}
		return 0;
	}

	/* Regular or preallocated extent has fixed item size */
	if (item_size != sizeof(*fi)) {
177
		file_extent_err(fs_info, leaf, slot,
178
	"invalid item size for reg/prealloc file extent, have %u expect %zu",
179
			item_size, sizeof(*fi));
180 181
		return -EUCLEAN;
	}
182 183 184 185 186
	if (CHECK_FE_ALIGNED(fs_info, leaf, slot, fi, ram_bytes, sectorsize) ||
	    CHECK_FE_ALIGNED(fs_info, leaf, slot, fi, disk_bytenr, sectorsize) ||
	    CHECK_FE_ALIGNED(fs_info, leaf, slot, fi, disk_num_bytes, sectorsize) ||
	    CHECK_FE_ALIGNED(fs_info, leaf, slot, fi, offset, sectorsize) ||
	    CHECK_FE_ALIGNED(fs_info, leaf, slot, fi, num_bytes, sectorsize))
187 188 189 190
		return -EUCLEAN;
	return 0;
}

191 192 193
static int check_csum_item(struct btrfs_fs_info *fs_info,
			   struct extent_buffer *leaf, struct btrfs_key *key,
			   int slot)
194
{
195 196
	u32 sectorsize = fs_info->sectorsize;
	u32 csumsize = btrfs_super_csum_size(fs_info->super_copy);
197 198

	if (key->objectid != BTRFS_EXTENT_CSUM_OBJECTID) {
199
		generic_err(fs_info, leaf, slot,
200 201
		"invalid key objectid for csum item, have %llu expect %llu",
			key->objectid, BTRFS_EXTENT_CSUM_OBJECTID);
202 203 204
		return -EUCLEAN;
	}
	if (!IS_ALIGNED(key->offset, sectorsize)) {
205
		generic_err(fs_info, leaf, slot,
206 207
	"unaligned key offset for csum item, have %llu should be aligned to %u",
			key->offset, sectorsize);
208 209 210
		return -EUCLEAN;
	}
	if (!IS_ALIGNED(btrfs_item_size_nr(leaf, slot), csumsize)) {
211
		generic_err(fs_info, leaf, slot,
212 213
	"unaligned item size for csum item, have %u should be aligned to %u",
			btrfs_item_size_nr(leaf, slot), csumsize);
214 215 216 217 218
		return -EUCLEAN;
	}
	return 0;
}

219 220 221 222 223
/*
 * Customized reported for dir_item, only important new info is key->objectid,
 * which represents inode number
 */
__printf(4, 5)
224
__cold
225
static void dir_item_err(const struct btrfs_fs_info *fs_info,
226 227 228 229 230 231 232 233 234 235 236 237 238
			 const struct extent_buffer *eb, int slot,
			 const char *fmt, ...)
{
	struct btrfs_key key;
	struct va_format vaf;
	va_list args;

	btrfs_item_key_to_cpu(eb, &key, slot);
	va_start(args, fmt);

	vaf.fmt = fmt;
	vaf.va = &args;

239
	btrfs_crit(fs_info,
240
	"corrupt %s: root=%llu block=%llu slot=%d ino=%llu, %pV",
241 242 243
		btrfs_header_level(eb) == 0 ? "leaf" : "node",
		btrfs_header_owner(eb), btrfs_header_bytenr(eb), slot,
		key.objectid, &vaf);
244 245 246
	va_end(args);
}

247
static int check_dir_item(struct btrfs_fs_info *fs_info,
248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
			  struct extent_buffer *leaf,
			  struct btrfs_key *key, int slot)
{
	struct btrfs_dir_item *di;
	u32 item_size = btrfs_item_size_nr(leaf, slot);
	u32 cur = 0;

	di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
	while (cur < item_size) {
		u32 name_len;
		u32 data_len;
		u32 max_name_len;
		u32 total_size;
		u32 name_hash;
		u8 dir_type;

		/* header itself should not cross item boundary */
		if (cur + sizeof(*di) > item_size) {
266
			dir_item_err(fs_info, leaf, slot,
267
		"dir item header crosses item boundary, have %zu boundary %u",
268 269 270 271 272 273 274
				cur + sizeof(*di), item_size);
			return -EUCLEAN;
		}

		/* dir type check */
		dir_type = btrfs_dir_type(leaf, di);
		if (dir_type >= BTRFS_FT_MAX) {
275
			dir_item_err(fs_info, leaf, slot,
276 277 278 279 280 281 282
			"invalid dir item type, have %u expect [0, %u)",
				dir_type, BTRFS_FT_MAX);
			return -EUCLEAN;
		}

		if (key->type == BTRFS_XATTR_ITEM_KEY &&
		    dir_type != BTRFS_FT_XATTR) {
283
			dir_item_err(fs_info, leaf, slot,
284 285 286 287 288 289
		"invalid dir item type for XATTR key, have %u expect %u",
				dir_type, BTRFS_FT_XATTR);
			return -EUCLEAN;
		}
		if (dir_type == BTRFS_FT_XATTR &&
		    key->type != BTRFS_XATTR_ITEM_KEY) {
290
			dir_item_err(fs_info, leaf, slot,
291 292 293 294 295 296 297 298 299 300 301 302
			"xattr dir type found for non-XATTR key");
			return -EUCLEAN;
		}
		if (dir_type == BTRFS_FT_XATTR)
			max_name_len = XATTR_NAME_MAX;
		else
			max_name_len = BTRFS_NAME_LEN;

		/* Name/data length check */
		name_len = btrfs_dir_name_len(leaf, di);
		data_len = btrfs_dir_data_len(leaf, di);
		if (name_len > max_name_len) {
303
			dir_item_err(fs_info, leaf, slot,
304 305 306 307
			"dir item name len too long, have %u max %u",
				name_len, max_name_len);
			return -EUCLEAN;
		}
308 309
		if (name_len + data_len > BTRFS_MAX_XATTR_SIZE(fs_info)) {
			dir_item_err(fs_info, leaf, slot,
310 311
			"dir item name and data len too long, have %u max %u",
				name_len + data_len,
312
				BTRFS_MAX_XATTR_SIZE(fs_info));
313 314 315 316
			return -EUCLEAN;
		}

		if (data_len && dir_type != BTRFS_FT_XATTR) {
317
			dir_item_err(fs_info, leaf, slot,
318 319 320 321 322 323 324 325 326
			"dir item with invalid data len, have %u expect 0",
				data_len);
			return -EUCLEAN;
		}

		total_size = sizeof(*di) + name_len + data_len;

		/* header and name/data should not cross item boundary */
		if (cur + total_size > item_size) {
327
			dir_item_err(fs_info, leaf, slot,
328 329 330 331 332 333 334 335 336 337 338
		"dir item data crosses item boundary, have %u boundary %u",
				cur + total_size, item_size);
			return -EUCLEAN;
		}

		/*
		 * Special check for XATTR/DIR_ITEM, as key->offset is name
		 * hash, should match its name
		 */
		if (key->type == BTRFS_DIR_ITEM_KEY ||
		    key->type == BTRFS_XATTR_ITEM_KEY) {
339 340
			char namebuf[max(BTRFS_NAME_LEN, XATTR_NAME_MAX)];

341 342 343 344
			read_extent_buffer(leaf, namebuf,
					(unsigned long)(di + 1), name_len);
			name_hash = btrfs_name_hash(namebuf, name_len);
			if (key->offset != name_hash) {
345
				dir_item_err(fs_info, leaf, slot,
346 347 348 349 350 351 352 353 354 355 356
		"name hash mismatch with key, have 0x%016x expect 0x%016llx",
					name_hash, key->offset);
				return -EUCLEAN;
			}
		}
		cur += total_size;
		di = (struct btrfs_dir_item *)((void *)di + total_size);
	}
	return 0;
}

357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391
__printf(4, 5)
__cold
static void block_group_err(const struct btrfs_fs_info *fs_info,
			    const struct extent_buffer *eb, int slot,
			    const char *fmt, ...)
{
	struct btrfs_key key;
	struct va_format vaf;
	va_list args;

	btrfs_item_key_to_cpu(eb, &key, slot);
	va_start(args, fmt);

	vaf.fmt = fmt;
	vaf.va = &args;

	btrfs_crit(fs_info,
	"corrupt %s: root=%llu block=%llu slot=%d bg_start=%llu bg_len=%llu, %pV",
		btrfs_header_level(eb) == 0 ? "leaf" : "node",
		btrfs_header_owner(eb), btrfs_header_bytenr(eb), slot,
		key.objectid, key.offset, &vaf);
	va_end(args);
}

static int check_block_group_item(struct btrfs_fs_info *fs_info,
				  struct extent_buffer *leaf,
				  struct btrfs_key *key, int slot)
{
	struct btrfs_block_group_item bgi;
	u32 item_size = btrfs_item_size_nr(leaf, slot);
	u64 flags;
	u64 type;

	/*
	 * Here we don't really care about alignment since extent allocator can
392
	 * handle it.  We care more about the size.
393
	 */
394
	if (key->offset == 0) {
395
		block_group_err(fs_info, leaf, slot,
396
				"invalid block group size 0");
397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
		return -EUCLEAN;
	}

	if (item_size != sizeof(bgi)) {
		block_group_err(fs_info, leaf, slot,
			"invalid item size, have %u expect %zu",
				item_size, sizeof(bgi));
		return -EUCLEAN;
	}

	read_extent_buffer(leaf, &bgi, btrfs_item_ptr_offset(leaf, slot),
			   sizeof(bgi));
	if (btrfs_block_group_chunk_objectid(&bgi) !=
	    BTRFS_FIRST_CHUNK_TREE_OBJECTID) {
		block_group_err(fs_info, leaf, slot,
		"invalid block group chunk objectid, have %llu expect %llu",
				btrfs_block_group_chunk_objectid(&bgi),
				BTRFS_FIRST_CHUNK_TREE_OBJECTID);
		return -EUCLEAN;
	}

	if (btrfs_block_group_used(&bgi) > key->offset) {
		block_group_err(fs_info, leaf, slot,
			"invalid block group used, have %llu expect [0, %llu)",
				btrfs_block_group_used(&bgi), key->offset);
		return -EUCLEAN;
	}

	flags = btrfs_block_group_flags(&bgi);
	if (hweight64(flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) > 1) {
		block_group_err(fs_info, leaf, slot,
"invalid profile flags, have 0x%llx (%lu bits set) expect no more than 1 bit set",
			flags & BTRFS_BLOCK_GROUP_PROFILE_MASK,
			hweight64(flags & BTRFS_BLOCK_GROUP_PROFILE_MASK));
		return -EUCLEAN;
	}

	type = flags & BTRFS_BLOCK_GROUP_TYPE_MASK;
	if (type != BTRFS_BLOCK_GROUP_DATA &&
	    type != BTRFS_BLOCK_GROUP_METADATA &&
	    type != BTRFS_BLOCK_GROUP_SYSTEM &&
	    type != (BTRFS_BLOCK_GROUP_METADATA |
			   BTRFS_BLOCK_GROUP_DATA)) {
		block_group_err(fs_info, leaf, slot,
441
"invalid type, have 0x%llx (%lu bits set) expect either 0x%llx, 0x%llx, 0x%llx or 0x%llx",
442 443 444 445 446 447 448 449 450
			type, hweight64(type),
			BTRFS_BLOCK_GROUP_DATA, BTRFS_BLOCK_GROUP_METADATA,
			BTRFS_BLOCK_GROUP_SYSTEM,
			BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA);
		return -EUCLEAN;
	}
	return 0;
}

451 452 453
/*
 * Common point to switch the item-specific validation.
 */
454
static int check_leaf_item(struct btrfs_fs_info *fs_info,
455 456 457 458 459 460 461
			   struct extent_buffer *leaf,
			   struct btrfs_key *key, int slot)
{
	int ret = 0;

	switch (key->type) {
	case BTRFS_EXTENT_DATA_KEY:
462
		ret = check_extent_data_item(fs_info, leaf, key, slot);
463 464
		break;
	case BTRFS_EXTENT_CSUM_KEY:
465
		ret = check_csum_item(fs_info, leaf, key, slot);
466
		break;
467 468 469
	case BTRFS_DIR_ITEM_KEY:
	case BTRFS_DIR_INDEX_KEY:
	case BTRFS_XATTR_ITEM_KEY:
470
		ret = check_dir_item(fs_info, leaf, key, slot);
471
		break;
472 473 474
	case BTRFS_BLOCK_GROUP_ITEM_KEY:
		ret = check_block_group_item(fs_info, leaf, key, slot);
		break;
475 476 477 478
	}
	return ret;
}

479
static int check_leaf(struct btrfs_fs_info *fs_info, struct extent_buffer *leaf,
480
		      bool check_item_data)
481 482 483 484 485 486 487
{
	/* No valid key type is 0, so all key should be larger than this key */
	struct btrfs_key prev_key = {0, 0, 0};
	struct btrfs_key key;
	u32 nritems = btrfs_header_nritems(leaf);
	int slot;

488 489 490 491 492 493 494
	if (btrfs_header_level(leaf) != 0) {
		generic_err(fs_info, leaf, 0,
			"invalid level for leaf, have %d expect 0",
			btrfs_header_level(leaf));
		return -EUCLEAN;
	}

495 496 497 498 499 500 501 502 503
	/*
	 * Extent buffers from a relocation tree have a owner field that
	 * corresponds to the subvolume tree they are based on. So just from an
	 * extent buffer alone we can not find out what is the id of the
	 * corresponding subvolume tree, so we can not figure out if the extent
	 * buffer corresponds to the root of the relocation tree or not. So
	 * skip this check for relocation trees.
	 */
	if (nritems == 0 && !btrfs_header_flag(leaf, BTRFS_HEADER_FLAG_RELOC)) {
504
		u64 owner = btrfs_header_owner(leaf);
505 506
		struct btrfs_root *check_root;

507 508 509 510 511 512 513 514 515 516 517 518 519
		/* These trees must never be empty */
		if (owner == BTRFS_ROOT_TREE_OBJECTID ||
		    owner == BTRFS_CHUNK_TREE_OBJECTID ||
		    owner == BTRFS_EXTENT_TREE_OBJECTID ||
		    owner == BTRFS_DEV_TREE_OBJECTID ||
		    owner == BTRFS_FS_TREE_OBJECTID ||
		    owner == BTRFS_DATA_RELOC_TREE_OBJECTID) {
			generic_err(fs_info, leaf, 0,
			"invalid root, root %llu must never be empty",
				    owner);
			return -EUCLEAN;
		}
		key.objectid = owner;
520 521 522 523 524 525 526 527 528 529 530 531 532 533
		key.type = BTRFS_ROOT_ITEM_KEY;
		key.offset = (u64)-1;

		check_root = btrfs_get_fs_root(fs_info, &key, false);
		/*
		 * The only reason we also check NULL here is that during
		 * open_ctree() some roots has not yet been set up.
		 */
		if (!IS_ERR_OR_NULL(check_root)) {
			struct extent_buffer *eb;

			eb = btrfs_root_node(check_root);
			/* if leaf is the root, then it's fine */
			if (leaf != eb) {
534
				generic_err(fs_info, leaf, 0,
535 536
		"invalid nritems, have %u should not be 0 for non-root leaf",
					nritems);
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
				free_extent_buffer(eb);
				return -EUCLEAN;
			}
			free_extent_buffer(eb);
		}
		return 0;
	}

	if (nritems == 0)
		return 0;

	/*
	 * Check the following things to make sure this is a good leaf, and
	 * leaf users won't need to bother with similar sanity checks:
	 *
	 * 1) key ordering
	 * 2) item offset and size
	 *    No overlap, no hole, all inside the leaf.
	 * 3) item content
	 *    If possible, do comprehensive sanity check.
	 *    NOTE: All checks must only rely on the item data itself.
	 */
	for (slot = 0; slot < nritems; slot++) {
		u32 item_end_expected;
		int ret;

		btrfs_item_key_to_cpu(leaf, &key, slot);

		/* Make sure the keys are in the right order */
		if (btrfs_comp_cpu_keys(&prev_key, &key) >= 0) {
567
			generic_err(fs_info, leaf, slot,
568 569 570 571
	"bad key order, prev (%llu %u %llu) current (%llu %u %llu)",
				prev_key.objectid, prev_key.type,
				prev_key.offset, key.objectid, key.type,
				key.offset);
572 573 574 575 576 577 578 579 580 581 582 583 584 585
			return -EUCLEAN;
		}

		/*
		 * 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 (slot == 0)
			item_end_expected = BTRFS_LEAF_DATA_SIZE(fs_info);
		else
			item_end_expected = btrfs_item_offset_nr(leaf,
								 slot - 1);
		if (btrfs_item_end_nr(leaf, slot) != item_end_expected) {
586
			generic_err(fs_info, leaf, slot,
587 588 589
				"unexpected item end, have %u expect %u",
				btrfs_item_end_nr(leaf, slot),
				item_end_expected);
590 591 592 593 594 595 596 597 598 599
			return -EUCLEAN;
		}

		/*
		 * Check to make sure that we don't point outside of the leaf,
		 * just in case all the items are consistent to each other, but
		 * all point outside of the leaf.
		 */
		if (btrfs_item_end_nr(leaf, slot) >
		    BTRFS_LEAF_DATA_SIZE(fs_info)) {
600
			generic_err(fs_info, leaf, slot,
601 602 603
			"slot end outside of leaf, have %u expect range [0, %u]",
				btrfs_item_end_nr(leaf, slot),
				BTRFS_LEAF_DATA_SIZE(fs_info));
604 605 606 607 608 609
			return -EUCLEAN;
		}

		/* Also check if the item pointer overlaps with btrfs item. */
		if (btrfs_item_nr_offset(slot) + sizeof(struct btrfs_item) >
		    btrfs_item_ptr_offset(leaf, slot)) {
610
			generic_err(fs_info, leaf, slot,
611 612 613 614
		"slot overlaps with its data, item end %lu data start %lu",
				btrfs_item_nr_offset(slot) +
				sizeof(struct btrfs_item),
				btrfs_item_ptr_offset(leaf, slot));
615 616 617
			return -EUCLEAN;
		}

618 619 620 621 622
		if (check_item_data) {
			/*
			 * Check if the item size and content meet other
			 * criteria
			 */
623
			ret = check_leaf_item(fs_info, leaf, &key, slot);
624 625 626
			if (ret < 0)
				return ret;
		}
627 628 629 630 631 632 633 634 635

		prev_key.objectid = key.objectid;
		prev_key.type = key.type;
		prev_key.offset = key.offset;
	}

	return 0;
}

636 637
int btrfs_check_leaf_full(struct btrfs_fs_info *fs_info,
			  struct extent_buffer *leaf)
638
{
639
	return check_leaf(fs_info, leaf, true);
640 641
}

642
int btrfs_check_leaf_relaxed(struct btrfs_fs_info *fs_info,
643 644
			     struct extent_buffer *leaf)
{
645
	return check_leaf(fs_info, leaf, false);
646 647
}

648
int btrfs_check_node(struct btrfs_fs_info *fs_info, struct extent_buffer *node)
649 650 651 652
{
	unsigned long nr = btrfs_header_nritems(node);
	struct btrfs_key key, next_key;
	int slot;
653
	int level = btrfs_header_level(node);
654 655 656
	u64 bytenr;
	int ret = 0;

657 658 659 660 661 662
	if (level <= 0 || level >= BTRFS_MAX_LEVEL) {
		generic_err(fs_info, node, 0,
			"invalid level for node, have %d expect [1, %d]",
			level, BTRFS_MAX_LEVEL - 1);
		return -EUCLEAN;
	}
663 664
	if (nr == 0 || nr > BTRFS_NODEPTRS_PER_BLOCK(fs_info)) {
		btrfs_crit(fs_info,
665
"corrupt node: root=%llu block=%llu, nritems too %s, have %lu expect range [1,%u]",
666
			   btrfs_header_owner(node), node->start,
667
			   nr == 0 ? "small" : "large", nr,
668
			   BTRFS_NODEPTRS_PER_BLOCK(fs_info));
669
		return -EUCLEAN;
670 671 672 673 674 675 676 677
	}

	for (slot = 0; slot < nr - 1; slot++) {
		bytenr = btrfs_node_blockptr(node, slot);
		btrfs_node_key_to_cpu(node, &key, slot);
		btrfs_node_key_to_cpu(node, &next_key, slot + 1);

		if (!bytenr) {
678
			generic_err(fs_info, node, slot,
679 680 681 682
				"invalid NULL node pointer");
			ret = -EUCLEAN;
			goto out;
		}
683 684
		if (!IS_ALIGNED(bytenr, fs_info->sectorsize)) {
			generic_err(fs_info, node, slot,
685
			"unaligned pointer, have %llu should be aligned to %u",
686
				bytenr, fs_info->sectorsize);
687
			ret = -EUCLEAN;
688 689 690 691
			goto out;
		}

		if (btrfs_comp_cpu_keys(&key, &next_key) >= 0) {
692
			generic_err(fs_info, node, slot,
693 694 695 696 697
	"bad key order, current (%llu %u %llu) next (%llu %u %llu)",
				key.objectid, key.type, key.offset,
				next_key.objectid, next_key.type,
				next_key.offset);
			ret = -EUCLEAN;
698 699 700 701 702 703
			goto out;
		}
	}
out:
	return ret;
}