check-integrity.c 102.7 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
/*
 * Copyright (C) STRATO AG 2011.  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.
 */

/*
 * This module can be used to catch cases when the btrfs kernel
 * code executes write requests to the disk that bring the file
 * system in an inconsistent state. In such a state, a power-loss
 * or kernel panic event would cause that the data on disk is
 * lost or at least damaged.
 *
 * Code is added that examines all block write requests during
 * runtime (including writes of the super block). Three rules
 * are verified and an error is printed on violation of the
 * rules:
 * 1. It is not allowed to write a disk block which is
 *    currently referenced by the super block (either directly
 *    or indirectly).
 * 2. When a super block is written, it is verified that all
 *    referenced (directly or indirectly) blocks fulfill the
 *    following requirements:
 *    2a. All referenced blocks have either been present when
 *        the file system was mounted, (i.e., they have been
 *        referenced by the super block) or they have been
 *        written since then and the write completion callback
40 41 42
 *        was called and no write error was indicated and a
 *        FLUSH request to the device where these blocks are
 *        located was received and completed.
43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96
 *    2b. All referenced blocks need to have a generation
 *        number which is equal to the parent's number.
 *
 * One issue that was found using this module was that the log
 * tree on disk became temporarily corrupted because disk blocks
 * that had been in use for the log tree had been freed and
 * reused too early, while being referenced by the written super
 * block.
 *
 * The search term in the kernel log that can be used to filter
 * on the existence of detected integrity issues is
 * "btrfs: attempt".
 *
 * The integrity check is enabled via mount options. These
 * mount options are only supported if the integrity check
 * tool is compiled by defining BTRFS_FS_CHECK_INTEGRITY.
 *
 * Example #1, apply integrity checks to all metadata:
 * mount /dev/sdb1 /mnt -o check_int
 *
 * Example #2, apply integrity checks to all metadata and
 * to data extents:
 * mount /dev/sdb1 /mnt -o check_int_data
 *
 * Example #3, apply integrity checks to all metadata and dump
 * the tree that the super block references to kernel messages
 * each time after a super block was written:
 * mount /dev/sdb1 /mnt -o check_int,check_int_print_mask=263
 *
 * If the integrity check tool is included and activated in
 * the mount options, plenty of kernel memory is used, and
 * plenty of additional CPU cycles are spent. Enabling this
 * functionality is not intended for normal use. In most
 * cases, unless you are a btrfs developer who needs to verify
 * the integrity of (super)-block write requests, do not
 * enable the config option BTRFS_FS_CHECK_INTEGRITY to
 * include and compile the integrity check tool.
 */

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/buffer_head.h>
#include <linux/mutex.h>
#include <linux/crc32c.h>
#include <linux/genhd.h>
#include <linux/blkdev.h>
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
#include "extent_io.h"
#include "volumes.h"
#include "print-tree.h"
#include "locking.h"
#include "check-integrity.h"
97
#include "rcu-string.h"
98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139

#define BTRFSIC_BLOCK_HASHTABLE_SIZE 0x10000
#define BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE 0x10000
#define BTRFSIC_DEV2STATE_HASHTABLE_SIZE 0x100
#define BTRFSIC_BLOCK_MAGIC_NUMBER 0x14491051
#define BTRFSIC_BLOCK_LINK_MAGIC_NUMBER 0x11070807
#define BTRFSIC_DEV2STATE_MAGIC_NUMBER 0x20111530
#define BTRFSIC_BLOCK_STACK_FRAME_MAGIC_NUMBER 20111300
#define BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL (200 - 6)	/* in characters,
							 * excluding " [...]" */
#define BTRFSIC_GENERATION_UNKNOWN ((u64)-1)

/*
 * The definition of the bitmask fields for the print_mask.
 * They are specified with the mount option check_integrity_print_mask.
 */
#define BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE			0x00000001
#define BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION		0x00000002
#define BTRFSIC_PRINT_MASK_TREE_AFTER_SB_WRITE			0x00000004
#define BTRFSIC_PRINT_MASK_TREE_BEFORE_SB_WRITE			0x00000008
#define BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH			0x00000010
#define BTRFSIC_PRINT_MASK_END_IO_BIO_BH			0x00000020
#define BTRFSIC_PRINT_MASK_VERBOSE				0x00000040
#define BTRFSIC_PRINT_MASK_VERY_VERBOSE				0x00000080
#define BTRFSIC_PRINT_MASK_INITIAL_TREE				0x00000100
#define BTRFSIC_PRINT_MASK_INITIAL_ALL_TREES			0x00000200
#define BTRFSIC_PRINT_MASK_INITIAL_DATABASE			0x00000400
#define BTRFSIC_PRINT_MASK_NUM_COPIES				0x00000800
#define BTRFSIC_PRINT_MASK_TREE_WITH_ALL_MIRRORS		0x00001000

struct btrfsic_dev_state;
struct btrfsic_state;

struct btrfsic_block {
	u32 magic_num;		/* only used for debug purposes */
	unsigned int is_metadata:1;	/* if it is meta-data, not data-data */
	unsigned int is_superblock:1;	/* if it is one of the superblocks */
	unsigned int is_iodone:1;	/* if is done by lower subsystem */
	unsigned int iodone_w_error:1;	/* error was indicated to endio */
	unsigned int never_written:1;	/* block was added because it was
					 * referenced, not because it was
					 * written */
140
	unsigned int mirror_num;	/* large enough to hold
141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212
					 * BTRFS_SUPER_MIRROR_MAX */
	struct btrfsic_dev_state *dev_state;
	u64 dev_bytenr;		/* key, physical byte num on disk */
	u64 logical_bytenr;	/* logical byte num on disk */
	u64 generation;
	struct btrfs_disk_key disk_key;	/* extra info to print in case of
					 * issues, will not always be correct */
	struct list_head collision_resolving_node;	/* list node */
	struct list_head all_blocks_node;	/* list node */

	/* the following two lists contain block_link items */
	struct list_head ref_to_list;	/* list */
	struct list_head ref_from_list;	/* list */
	struct btrfsic_block *next_in_same_bio;
	void *orig_bio_bh_private;
	union {
		bio_end_io_t *bio;
		bh_end_io_t *bh;
	} orig_bio_bh_end_io;
	int submit_bio_bh_rw;
	u64 flush_gen; /* only valid if !never_written */
};

/*
 * Elements of this type are allocated dynamically and required because
 * each block object can refer to and can be ref from multiple blocks.
 * The key to lookup them in the hashtable is the dev_bytenr of
 * the block ref to plus the one from the block refered from.
 * The fact that they are searchable via a hashtable and that a
 * ref_cnt is maintained is not required for the btrfs integrity
 * check algorithm itself, it is only used to make the output more
 * beautiful in case that an error is detected (an error is defined
 * as a write operation to a block while that block is still referenced).
 */
struct btrfsic_block_link {
	u32 magic_num;		/* only used for debug purposes */
	u32 ref_cnt;
	struct list_head node_ref_to;	/* list node */
	struct list_head node_ref_from;	/* list node */
	struct list_head collision_resolving_node;	/* list node */
	struct btrfsic_block *block_ref_to;
	struct btrfsic_block *block_ref_from;
	u64 parent_generation;
};

struct btrfsic_dev_state {
	u32 magic_num;		/* only used for debug purposes */
	struct block_device *bdev;
	struct btrfsic_state *state;
	struct list_head collision_resolving_node;	/* list node */
	struct btrfsic_block dummy_block_for_bio_bh_flush;
	u64 last_flush_gen;
	char name[BDEVNAME_SIZE];
};

struct btrfsic_block_hashtable {
	struct list_head table[BTRFSIC_BLOCK_HASHTABLE_SIZE];
};

struct btrfsic_block_link_hashtable {
	struct list_head table[BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE];
};

struct btrfsic_dev_state_hashtable {
	struct list_head table[BTRFSIC_DEV2STATE_HASHTABLE_SIZE];
};

struct btrfsic_block_data_ctx {
	u64 start;		/* virtual bytenr */
	u64 dev_bytenr;		/* physical bytenr on device */
	u32 len;
	struct btrfsic_dev_state *dev;
213 214 215
	char **datav;
	struct page **pagev;
	void *mem_to_free;
216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246
};

/* This structure is used to implement recursion without occupying
 * any stack space, refer to btrfsic_process_metablock() */
struct btrfsic_stack_frame {
	u32 magic;
	u32 nr;
	int error;
	int i;
	int limit_nesting;
	int num_copies;
	int mirror_num;
	struct btrfsic_block *block;
	struct btrfsic_block_data_ctx *block_ctx;
	struct btrfsic_block *next_block;
	struct btrfsic_block_data_ctx next_block_ctx;
	struct btrfs_header *hdr;
	struct btrfsic_stack_frame *prev;
};

/* Some state per mounted filesystem */
struct btrfsic_state {
	u32 print_mask;
	int include_extent_data;
	int csum_size;
	struct list_head all_blocks_list;
	struct btrfsic_block_hashtable block_hashtable;
	struct btrfsic_block_link_hashtable block_link_hashtable;
	struct btrfs_root *root;
	u64 max_superblock_generation;
	struct btrfsic_block *latest_superblock;
247 248
	u32 metablock_size;
	u32 datablock_size;
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296
};

static void btrfsic_block_init(struct btrfsic_block *b);
static struct btrfsic_block *btrfsic_block_alloc(void);
static void btrfsic_block_free(struct btrfsic_block *b);
static void btrfsic_block_link_init(struct btrfsic_block_link *n);
static struct btrfsic_block_link *btrfsic_block_link_alloc(void);
static void btrfsic_block_link_free(struct btrfsic_block_link *n);
static void btrfsic_dev_state_init(struct btrfsic_dev_state *ds);
static struct btrfsic_dev_state *btrfsic_dev_state_alloc(void);
static void btrfsic_dev_state_free(struct btrfsic_dev_state *ds);
static void btrfsic_block_hashtable_init(struct btrfsic_block_hashtable *h);
static void btrfsic_block_hashtable_add(struct btrfsic_block *b,
					struct btrfsic_block_hashtable *h);
static void btrfsic_block_hashtable_remove(struct btrfsic_block *b);
static struct btrfsic_block *btrfsic_block_hashtable_lookup(
		struct block_device *bdev,
		u64 dev_bytenr,
		struct btrfsic_block_hashtable *h);
static void btrfsic_block_link_hashtable_init(
		struct btrfsic_block_link_hashtable *h);
static void btrfsic_block_link_hashtable_add(
		struct btrfsic_block_link *l,
		struct btrfsic_block_link_hashtable *h);
static void btrfsic_block_link_hashtable_remove(struct btrfsic_block_link *l);
static struct btrfsic_block_link *btrfsic_block_link_hashtable_lookup(
		struct block_device *bdev_ref_to,
		u64 dev_bytenr_ref_to,
		struct block_device *bdev_ref_from,
		u64 dev_bytenr_ref_from,
		struct btrfsic_block_link_hashtable *h);
static void btrfsic_dev_state_hashtable_init(
		struct btrfsic_dev_state_hashtable *h);
static void btrfsic_dev_state_hashtable_add(
		struct btrfsic_dev_state *ds,
		struct btrfsic_dev_state_hashtable *h);
static void btrfsic_dev_state_hashtable_remove(struct btrfsic_dev_state *ds);
static struct btrfsic_dev_state *btrfsic_dev_state_hashtable_lookup(
		struct block_device *bdev,
		struct btrfsic_dev_state_hashtable *h);
static struct btrfsic_stack_frame *btrfsic_stack_frame_alloc(void);
static void btrfsic_stack_frame_free(struct btrfsic_stack_frame *sf);
static int btrfsic_process_superblock(struct btrfsic_state *state,
				      struct btrfs_fs_devices *fs_devices);
static int btrfsic_process_metablock(struct btrfsic_state *state,
				     struct btrfsic_block *block,
				     struct btrfsic_block_data_ctx *block_ctx,
				     int limit_nesting, int force_iodone_flag);
297 298 299
static void btrfsic_read_from_block_data(
	struct btrfsic_block_data_ctx *block_ctx,
	void *dst, u32 offset, size_t len);
300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
static int btrfsic_create_link_to_next_block(
		struct btrfsic_state *state,
		struct btrfsic_block *block,
		struct btrfsic_block_data_ctx
		*block_ctx, u64 next_bytenr,
		int limit_nesting,
		struct btrfsic_block_data_ctx *next_block_ctx,
		struct btrfsic_block **next_blockp,
		int force_iodone_flag,
		int *num_copiesp, int *mirror_nump,
		struct btrfs_disk_key *disk_key,
		u64 parent_generation);
static int btrfsic_handle_extent_data(struct btrfsic_state *state,
				      struct btrfsic_block *block,
				      struct btrfsic_block_data_ctx *block_ctx,
				      u32 item_offset, int force_iodone_flag);
static int btrfsic_map_block(struct btrfsic_state *state, u64 bytenr, u32 len,
			     struct btrfsic_block_data_ctx *block_ctx_out,
			     int mirror_num);
static int btrfsic_map_superblock(struct btrfsic_state *state, u64 bytenr,
				  u32 len, struct block_device *bdev,
				  struct btrfsic_block_data_ctx *block_ctx_out);
static void btrfsic_release_block_ctx(struct btrfsic_block_data_ctx *block_ctx);
static int btrfsic_read_block(struct btrfsic_state *state,
			      struct btrfsic_block_data_ctx *block_ctx);
static void btrfsic_dump_database(struct btrfsic_state *state);
326
static void btrfsic_complete_bio_end_io(struct bio *bio, int err);
327
static int btrfsic_test_for_metadata(struct btrfsic_state *state,
328
				     char **datav, unsigned int num_pages);
329
static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
330 331 332
					  u64 dev_bytenr, char **mapped_datav,
					  unsigned int num_pages,
					  struct bio *bio, int *bio_is_patched,
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383
					  struct buffer_head *bh,
					  int submit_bio_bh_rw);
static int btrfsic_process_written_superblock(
		struct btrfsic_state *state,
		struct btrfsic_block *const block,
		struct btrfs_super_block *const super_hdr);
static void btrfsic_bio_end_io(struct bio *bp, int bio_error_status);
static void btrfsic_bh_end_io(struct buffer_head *bh, int uptodate);
static int btrfsic_is_block_ref_by_superblock(const struct btrfsic_state *state,
					      const struct btrfsic_block *block,
					      int recursion_level);
static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state,
					struct btrfsic_block *const block,
					int recursion_level);
static void btrfsic_print_add_link(const struct btrfsic_state *state,
				   const struct btrfsic_block_link *l);
static void btrfsic_print_rem_link(const struct btrfsic_state *state,
				   const struct btrfsic_block_link *l);
static char btrfsic_get_block_type(const struct btrfsic_state *state,
				   const struct btrfsic_block *block);
static void btrfsic_dump_tree(const struct btrfsic_state *state);
static void btrfsic_dump_tree_sub(const struct btrfsic_state *state,
				  const struct btrfsic_block *block,
				  int indent_level);
static struct btrfsic_block_link *btrfsic_block_link_lookup_or_add(
		struct btrfsic_state *state,
		struct btrfsic_block_data_ctx *next_block_ctx,
		struct btrfsic_block *next_block,
		struct btrfsic_block *from_block,
		u64 parent_generation);
static struct btrfsic_block *btrfsic_block_lookup_or_add(
		struct btrfsic_state *state,
		struct btrfsic_block_data_ctx *block_ctx,
		const char *additional_string,
		int is_metadata,
		int is_iodone,
		int never_written,
		int mirror_num,
		int *was_created);
static int btrfsic_process_superblock_dev_mirror(
		struct btrfsic_state *state,
		struct btrfsic_dev_state *dev_state,
		struct btrfs_device *device,
		int superblock_mirror_num,
		struct btrfsic_dev_state **selected_dev_state,
		struct btrfs_super_block *selected_super);
static struct btrfsic_dev_state *btrfsic_dev_state_lookup(
		struct block_device *bdev);
static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
					   u64 bytenr,
					   struct btrfsic_dev_state *dev_state,
384
					   u64 dev_bytenr);
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651

static struct mutex btrfsic_mutex;
static int btrfsic_is_initialized;
static struct btrfsic_dev_state_hashtable btrfsic_dev_state_hashtable;


static void btrfsic_block_init(struct btrfsic_block *b)
{
	b->magic_num = BTRFSIC_BLOCK_MAGIC_NUMBER;
	b->dev_state = NULL;
	b->dev_bytenr = 0;
	b->logical_bytenr = 0;
	b->generation = BTRFSIC_GENERATION_UNKNOWN;
	b->disk_key.objectid = 0;
	b->disk_key.type = 0;
	b->disk_key.offset = 0;
	b->is_metadata = 0;
	b->is_superblock = 0;
	b->is_iodone = 0;
	b->iodone_w_error = 0;
	b->never_written = 0;
	b->mirror_num = 0;
	b->next_in_same_bio = NULL;
	b->orig_bio_bh_private = NULL;
	b->orig_bio_bh_end_io.bio = NULL;
	INIT_LIST_HEAD(&b->collision_resolving_node);
	INIT_LIST_HEAD(&b->all_blocks_node);
	INIT_LIST_HEAD(&b->ref_to_list);
	INIT_LIST_HEAD(&b->ref_from_list);
	b->submit_bio_bh_rw = 0;
	b->flush_gen = 0;
}

static struct btrfsic_block *btrfsic_block_alloc(void)
{
	struct btrfsic_block *b;

	b = kzalloc(sizeof(*b), GFP_NOFS);
	if (NULL != b)
		btrfsic_block_init(b);

	return b;
}

static void btrfsic_block_free(struct btrfsic_block *b)
{
	BUG_ON(!(NULL == b || BTRFSIC_BLOCK_MAGIC_NUMBER == b->magic_num));
	kfree(b);
}

static void btrfsic_block_link_init(struct btrfsic_block_link *l)
{
	l->magic_num = BTRFSIC_BLOCK_LINK_MAGIC_NUMBER;
	l->ref_cnt = 1;
	INIT_LIST_HEAD(&l->node_ref_to);
	INIT_LIST_HEAD(&l->node_ref_from);
	INIT_LIST_HEAD(&l->collision_resolving_node);
	l->block_ref_to = NULL;
	l->block_ref_from = NULL;
}

static struct btrfsic_block_link *btrfsic_block_link_alloc(void)
{
	struct btrfsic_block_link *l;

	l = kzalloc(sizeof(*l), GFP_NOFS);
	if (NULL != l)
		btrfsic_block_link_init(l);

	return l;
}

static void btrfsic_block_link_free(struct btrfsic_block_link *l)
{
	BUG_ON(!(NULL == l || BTRFSIC_BLOCK_LINK_MAGIC_NUMBER == l->magic_num));
	kfree(l);
}

static void btrfsic_dev_state_init(struct btrfsic_dev_state *ds)
{
	ds->magic_num = BTRFSIC_DEV2STATE_MAGIC_NUMBER;
	ds->bdev = NULL;
	ds->state = NULL;
	ds->name[0] = '\0';
	INIT_LIST_HEAD(&ds->collision_resolving_node);
	ds->last_flush_gen = 0;
	btrfsic_block_init(&ds->dummy_block_for_bio_bh_flush);
	ds->dummy_block_for_bio_bh_flush.is_iodone = 1;
	ds->dummy_block_for_bio_bh_flush.dev_state = ds;
}

static struct btrfsic_dev_state *btrfsic_dev_state_alloc(void)
{
	struct btrfsic_dev_state *ds;

	ds = kzalloc(sizeof(*ds), GFP_NOFS);
	if (NULL != ds)
		btrfsic_dev_state_init(ds);

	return ds;
}

static void btrfsic_dev_state_free(struct btrfsic_dev_state *ds)
{
	BUG_ON(!(NULL == ds ||
		 BTRFSIC_DEV2STATE_MAGIC_NUMBER == ds->magic_num));
	kfree(ds);
}

static void btrfsic_block_hashtable_init(struct btrfsic_block_hashtable *h)
{
	int i;

	for (i = 0; i < BTRFSIC_BLOCK_HASHTABLE_SIZE; i++)
		INIT_LIST_HEAD(h->table + i);
}

static void btrfsic_block_hashtable_add(struct btrfsic_block *b,
					struct btrfsic_block_hashtable *h)
{
	const unsigned int hashval =
	    (((unsigned int)(b->dev_bytenr >> 16)) ^
	     ((unsigned int)((uintptr_t)b->dev_state->bdev))) &
	     (BTRFSIC_BLOCK_HASHTABLE_SIZE - 1);

	list_add(&b->collision_resolving_node, h->table + hashval);
}

static void btrfsic_block_hashtable_remove(struct btrfsic_block *b)
{
	list_del(&b->collision_resolving_node);
}

static struct btrfsic_block *btrfsic_block_hashtable_lookup(
		struct block_device *bdev,
		u64 dev_bytenr,
		struct btrfsic_block_hashtable *h)
{
	const unsigned int hashval =
	    (((unsigned int)(dev_bytenr >> 16)) ^
	     ((unsigned int)((uintptr_t)bdev))) &
	     (BTRFSIC_BLOCK_HASHTABLE_SIZE - 1);
	struct list_head *elem;

	list_for_each(elem, h->table + hashval) {
		struct btrfsic_block *const b =
		    list_entry(elem, struct btrfsic_block,
			       collision_resolving_node);

		if (b->dev_state->bdev == bdev && b->dev_bytenr == dev_bytenr)
			return b;
	}

	return NULL;
}

static void btrfsic_block_link_hashtable_init(
		struct btrfsic_block_link_hashtable *h)
{
	int i;

	for (i = 0; i < BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE; i++)
		INIT_LIST_HEAD(h->table + i);
}

static void btrfsic_block_link_hashtable_add(
		struct btrfsic_block_link *l,
		struct btrfsic_block_link_hashtable *h)
{
	const unsigned int hashval =
	    (((unsigned int)(l->block_ref_to->dev_bytenr >> 16)) ^
	     ((unsigned int)(l->block_ref_from->dev_bytenr >> 16)) ^
	     ((unsigned int)((uintptr_t)l->block_ref_to->dev_state->bdev)) ^
	     ((unsigned int)((uintptr_t)l->block_ref_from->dev_state->bdev)))
	     & (BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE - 1);

	BUG_ON(NULL == l->block_ref_to);
	BUG_ON(NULL == l->block_ref_from);
	list_add(&l->collision_resolving_node, h->table + hashval);
}

static void btrfsic_block_link_hashtable_remove(struct btrfsic_block_link *l)
{
	list_del(&l->collision_resolving_node);
}

static struct btrfsic_block_link *btrfsic_block_link_hashtable_lookup(
		struct block_device *bdev_ref_to,
		u64 dev_bytenr_ref_to,
		struct block_device *bdev_ref_from,
		u64 dev_bytenr_ref_from,
		struct btrfsic_block_link_hashtable *h)
{
	const unsigned int hashval =
	    (((unsigned int)(dev_bytenr_ref_to >> 16)) ^
	     ((unsigned int)(dev_bytenr_ref_from >> 16)) ^
	     ((unsigned int)((uintptr_t)bdev_ref_to)) ^
	     ((unsigned int)((uintptr_t)bdev_ref_from))) &
	     (BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE - 1);
	struct list_head *elem;

	list_for_each(elem, h->table + hashval) {
		struct btrfsic_block_link *const l =
		    list_entry(elem, struct btrfsic_block_link,
			       collision_resolving_node);

		BUG_ON(NULL == l->block_ref_to);
		BUG_ON(NULL == l->block_ref_from);
		if (l->block_ref_to->dev_state->bdev == bdev_ref_to &&
		    l->block_ref_to->dev_bytenr == dev_bytenr_ref_to &&
		    l->block_ref_from->dev_state->bdev == bdev_ref_from &&
		    l->block_ref_from->dev_bytenr == dev_bytenr_ref_from)
			return l;
	}

	return NULL;
}

static void btrfsic_dev_state_hashtable_init(
		struct btrfsic_dev_state_hashtable *h)
{
	int i;

	for (i = 0; i < BTRFSIC_DEV2STATE_HASHTABLE_SIZE; i++)
		INIT_LIST_HEAD(h->table + i);
}

static void btrfsic_dev_state_hashtable_add(
		struct btrfsic_dev_state *ds,
		struct btrfsic_dev_state_hashtable *h)
{
	const unsigned int hashval =
	    (((unsigned int)((uintptr_t)ds->bdev)) &
	     (BTRFSIC_DEV2STATE_HASHTABLE_SIZE - 1));

	list_add(&ds->collision_resolving_node, h->table + hashval);
}

static void btrfsic_dev_state_hashtable_remove(struct btrfsic_dev_state *ds)
{
	list_del(&ds->collision_resolving_node);
}

static struct btrfsic_dev_state *btrfsic_dev_state_hashtable_lookup(
		struct block_device *bdev,
		struct btrfsic_dev_state_hashtable *h)
{
	const unsigned int hashval =
	    (((unsigned int)((uintptr_t)bdev)) &
	     (BTRFSIC_DEV2STATE_HASHTABLE_SIZE - 1));
	struct list_head *elem;

	list_for_each(elem, h->table + hashval) {
		struct btrfsic_dev_state *const ds =
		    list_entry(elem, struct btrfsic_dev_state,
			       collision_resolving_node);

		if (ds->bdev == bdev)
			return ds;
	}

	return NULL;
}

static int btrfsic_process_superblock(struct btrfsic_state *state,
				      struct btrfs_fs_devices *fs_devices)
{
652
	int ret = 0;
653 654 655 656 657 658 659
	struct btrfs_super_block *selected_super;
	struct list_head *dev_head = &fs_devices->devices;
	struct btrfs_device *device;
	struct btrfsic_dev_state *selected_dev_state = NULL;
	int pass;

	BUG_ON(NULL == state);
660
	selected_super = kzalloc(sizeof(*selected_super), GFP_NOFS);
661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
	if (NULL == selected_super) {
		printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
		return -1;
	}

	list_for_each_entry(device, dev_head, dev_list) {
		int i;
		struct btrfsic_dev_state *dev_state;

		if (!device->bdev || !device->name)
			continue;

		dev_state = btrfsic_dev_state_lookup(device->bdev);
		BUG_ON(NULL == dev_state);
		for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
			ret = btrfsic_process_superblock_dev_mirror(
					state, dev_state, device, i,
					&selected_dev_state, selected_super);
			if (0 != ret && 0 == i) {
				kfree(selected_super);
				return ret;
			}
		}
	}

	if (NULL == state->latest_superblock) {
		printk(KERN_INFO "btrfsic: no superblock found!\n");
		kfree(selected_super);
		return -1;
	}

	state->csum_size = btrfs_super_csum_size(selected_super);

	for (pass = 0; pass < 3; pass++) {
		int num_copies;
		int mirror_num;
		u64 next_bytenr;

		switch (pass) {
		case 0:
			next_bytenr = btrfs_super_root(selected_super);
			if (state->print_mask &
			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
				printk(KERN_INFO "root@%llu\n",
				       (unsigned long long)next_bytenr);
			break;
		case 1:
			next_bytenr = btrfs_super_chunk_root(selected_super);
			if (state->print_mask &
			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
				printk(KERN_INFO "chunk@%llu\n",
				       (unsigned long long)next_bytenr);
			break;
		case 2:
			next_bytenr = btrfs_super_log_root(selected_super);
			if (0 == next_bytenr)
				continue;
			if (state->print_mask &
			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
				printk(KERN_INFO "log@%llu\n",
				       (unsigned long long)next_bytenr);
			break;
		}

		num_copies =
726
		    btrfs_num_copies(state->root->fs_info,
727
				     next_bytenr, state->metablock_size);
728 729 730 731 732 733 734 735 736
		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
			       (unsigned long long)next_bytenr, num_copies);

		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
			struct btrfsic_block *next_block;
			struct btrfsic_block_data_ctx tmp_next_block_ctx;
			struct btrfsic_block_link *l;

737 738
			ret = btrfsic_map_block(state, next_bytenr,
						state->metablock_size,
739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766
						&tmp_next_block_ctx,
						mirror_num);
			if (ret) {
				printk(KERN_INFO "btrfsic:"
				       " btrfsic_map_block(root @%llu,"
				       " mirror %d) failed!\n",
				       (unsigned long long)next_bytenr,
				       mirror_num);
				kfree(selected_super);
				return -1;
			}

			next_block = btrfsic_block_hashtable_lookup(
					tmp_next_block_ctx.dev->bdev,
					tmp_next_block_ctx.dev_bytenr,
					&state->block_hashtable);
			BUG_ON(NULL == next_block);

			l = btrfsic_block_link_hashtable_lookup(
					tmp_next_block_ctx.dev->bdev,
					tmp_next_block_ctx.dev_bytenr,
					state->latest_superblock->dev_state->
					bdev,
					state->latest_superblock->dev_bytenr,
					&state->block_link_hashtable);
			BUG_ON(NULL == l);

			ret = btrfsic_read_block(state, &tmp_next_block_ctx);
767
			if (ret < (int)PAGE_CACHE_SIZE) {
768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805
				printk(KERN_INFO
				       "btrfsic: read @logical %llu failed!\n",
				       (unsigned long long)
				       tmp_next_block_ctx.start);
				btrfsic_release_block_ctx(&tmp_next_block_ctx);
				kfree(selected_super);
				return -1;
			}

			ret = btrfsic_process_metablock(state,
							next_block,
							&tmp_next_block_ctx,
							BTRFS_MAX_LEVEL + 3, 1);
			btrfsic_release_block_ctx(&tmp_next_block_ctx);
		}
	}

	kfree(selected_super);
	return ret;
}

static int btrfsic_process_superblock_dev_mirror(
		struct btrfsic_state *state,
		struct btrfsic_dev_state *dev_state,
		struct btrfs_device *device,
		int superblock_mirror_num,
		struct btrfsic_dev_state **selected_dev_state,
		struct btrfs_super_block *selected_super)
{
	struct btrfs_super_block *super_tmp;
	u64 dev_bytenr;
	struct buffer_head *bh;
	struct btrfsic_block *superblock_tmp;
	int pass;
	struct block_device *const superblock_bdev = device->bdev;

	/* super block bytenr is always the unmapped device bytenr */
	dev_bytenr = btrfs_sb_offset(superblock_mirror_num);
806 807 808 809
	if (dev_bytenr + BTRFS_SUPER_INFO_SIZE > device->total_bytes)
		return -1;
	bh = __bread(superblock_bdev, dev_bytenr / 4096,
		     BTRFS_SUPER_INFO_SIZE);
810 811 812 813 814 815
	if (NULL == bh)
		return -1;
	super_tmp = (struct btrfs_super_block *)
	    (bh->b_data + (dev_bytenr & 4095));

	if (btrfs_super_bytenr(super_tmp) != dev_bytenr ||
816
	    super_tmp->magic != cpu_to_le64(BTRFS_MAGIC) ||
817 818 819 820
	    memcmp(device->uuid, super_tmp->dev_item.uuid, BTRFS_UUID_SIZE) ||
	    btrfs_super_nodesize(super_tmp) != state->metablock_size ||
	    btrfs_super_leafsize(super_tmp) != state->metablock_size ||
	    btrfs_super_sectorsize(super_tmp) != state->datablock_size) {
821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846
		brelse(bh);
		return 0;
	}

	superblock_tmp =
	    btrfsic_block_hashtable_lookup(superblock_bdev,
					   dev_bytenr,
					   &state->block_hashtable);
	if (NULL == superblock_tmp) {
		superblock_tmp = btrfsic_block_alloc();
		if (NULL == superblock_tmp) {
			printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
			brelse(bh);
			return -1;
		}
		/* for superblock, only the dev_bytenr makes sense */
		superblock_tmp->dev_bytenr = dev_bytenr;
		superblock_tmp->dev_state = dev_state;
		superblock_tmp->logical_bytenr = dev_bytenr;
		superblock_tmp->generation = btrfs_super_generation(super_tmp);
		superblock_tmp->is_metadata = 1;
		superblock_tmp->is_superblock = 1;
		superblock_tmp->is_iodone = 1;
		superblock_tmp->never_written = 0;
		superblock_tmp->mirror_num = 1 + superblock_mirror_num;
		if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE)
847 848 849 850 851 852 853 854
			printk_in_rcu(KERN_INFO "New initial S-block (bdev %p, %s)"
				     " @%llu (%s/%llu/%d)\n",
				     superblock_bdev,
				     rcu_str_deref(device->name),
				     (unsigned long long)dev_bytenr,
				     dev_state->name,
				     (unsigned long long)dev_bytenr,
				     superblock_mirror_num);
855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904
		list_add(&superblock_tmp->all_blocks_node,
			 &state->all_blocks_list);
		btrfsic_block_hashtable_add(superblock_tmp,
					    &state->block_hashtable);
	}

	/* select the one with the highest generation field */
	if (btrfs_super_generation(super_tmp) >
	    state->max_superblock_generation ||
	    0 == state->max_superblock_generation) {
		memcpy(selected_super, super_tmp, sizeof(*selected_super));
		*selected_dev_state = dev_state;
		state->max_superblock_generation =
		    btrfs_super_generation(super_tmp);
		state->latest_superblock = superblock_tmp;
	}

	for (pass = 0; pass < 3; pass++) {
		u64 next_bytenr;
		int num_copies;
		int mirror_num;
		const char *additional_string = NULL;
		struct btrfs_disk_key tmp_disk_key;

		tmp_disk_key.type = BTRFS_ROOT_ITEM_KEY;
		tmp_disk_key.offset = 0;
		switch (pass) {
		case 0:
			tmp_disk_key.objectid =
			    cpu_to_le64(BTRFS_ROOT_TREE_OBJECTID);
			additional_string = "initial root ";
			next_bytenr = btrfs_super_root(super_tmp);
			break;
		case 1:
			tmp_disk_key.objectid =
			    cpu_to_le64(BTRFS_CHUNK_TREE_OBJECTID);
			additional_string = "initial chunk ";
			next_bytenr = btrfs_super_chunk_root(super_tmp);
			break;
		case 2:
			tmp_disk_key.objectid =
			    cpu_to_le64(BTRFS_TREE_LOG_OBJECTID);
			additional_string = "initial log ";
			next_bytenr = btrfs_super_log_root(super_tmp);
			if (0 == next_bytenr)
				continue;
			break;
		}

		num_copies =
905
		    btrfs_num_copies(state->root->fs_info,
906
				     next_bytenr, state->metablock_size);
907 908 909 910 911 912 913 914
		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
			       (unsigned long long)next_bytenr, num_copies);
		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
			struct btrfsic_block *next_block;
			struct btrfsic_block_data_ctx tmp_next_block_ctx;
			struct btrfsic_block_link *l;

915 916
			if (btrfsic_map_block(state, next_bytenr,
					      state->metablock_size,
917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984
					      &tmp_next_block_ctx,
					      mirror_num)) {
				printk(KERN_INFO "btrfsic: btrfsic_map_block("
				       "bytenr @%llu, mirror %d) failed!\n",
				       (unsigned long long)next_bytenr,
				       mirror_num);
				brelse(bh);
				return -1;
			}

			next_block = btrfsic_block_lookup_or_add(
					state, &tmp_next_block_ctx,
					additional_string, 1, 1, 0,
					mirror_num, NULL);
			if (NULL == next_block) {
				btrfsic_release_block_ctx(&tmp_next_block_ctx);
				brelse(bh);
				return -1;
			}

			next_block->disk_key = tmp_disk_key;
			next_block->generation = BTRFSIC_GENERATION_UNKNOWN;
			l = btrfsic_block_link_lookup_or_add(
					state, &tmp_next_block_ctx,
					next_block, superblock_tmp,
					BTRFSIC_GENERATION_UNKNOWN);
			btrfsic_release_block_ctx(&tmp_next_block_ctx);
			if (NULL == l) {
				brelse(bh);
				return -1;
			}
		}
	}
	if (state->print_mask & BTRFSIC_PRINT_MASK_INITIAL_ALL_TREES)
		btrfsic_dump_tree_sub(state, superblock_tmp, 0);

	brelse(bh);
	return 0;
}

static struct btrfsic_stack_frame *btrfsic_stack_frame_alloc(void)
{
	struct btrfsic_stack_frame *sf;

	sf = kzalloc(sizeof(*sf), GFP_NOFS);
	if (NULL == sf)
		printk(KERN_INFO "btrfsic: alloc memory failed!\n");
	else
		sf->magic = BTRFSIC_BLOCK_STACK_FRAME_MAGIC_NUMBER;
	return sf;
}

static void btrfsic_stack_frame_free(struct btrfsic_stack_frame *sf)
{
	BUG_ON(!(NULL == sf ||
		 BTRFSIC_BLOCK_STACK_FRAME_MAGIC_NUMBER == sf->magic));
	kfree(sf);
}

static int btrfsic_process_metablock(
		struct btrfsic_state *state,
		struct btrfsic_block *const first_block,
		struct btrfsic_block_data_ctx *const first_block_ctx,
		int first_limit_nesting, int force_iodone_flag)
{
	struct btrfsic_stack_frame initial_stack_frame = { 0 };
	struct btrfsic_stack_frame *sf;
	struct btrfsic_stack_frame *next_stack;
985 986
	struct btrfs_header *const first_hdr =
		(struct btrfs_header *)first_block_ctx->datav[0];
987

988
	BUG_ON(!first_hdr);
989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027
	sf = &initial_stack_frame;
	sf->error = 0;
	sf->i = -1;
	sf->limit_nesting = first_limit_nesting;
	sf->block = first_block;
	sf->block_ctx = first_block_ctx;
	sf->next_block = NULL;
	sf->hdr = first_hdr;
	sf->prev = NULL;

continue_with_new_stack_frame:
	sf->block->generation = le64_to_cpu(sf->hdr->generation);
	if (0 == sf->hdr->level) {
		struct btrfs_leaf *const leafhdr =
		    (struct btrfs_leaf *)sf->hdr;

		if (-1 == sf->i) {
			sf->nr = le32_to_cpu(leafhdr->header.nritems);

			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO
				       "leaf %llu items %d generation %llu"
				       " owner %llu\n",
				       (unsigned long long)
				       sf->block_ctx->start,
				       sf->nr,
				       (unsigned long long)
				       le64_to_cpu(leafhdr->header.generation),
				       (unsigned long long)
				       le64_to_cpu(leafhdr->header.owner));
		}

continue_with_current_leaf_stack_frame:
		if (0 == sf->num_copies || sf->mirror_num > sf->num_copies) {
			sf->i++;
			sf->num_copies = 0;
		}

		if (sf->i < sf->nr) {
1028 1029 1030 1031 1032
			struct btrfs_item disk_item;
			u32 disk_item_offset =
				(uintptr_t)(leafhdr->items + sf->i) -
				(uintptr_t)leafhdr;
			struct btrfs_disk_key *disk_key;
1033
			u8 type;
1034
			u32 item_offset;
1035
			u32 item_size;
1036

1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050
			if (disk_item_offset + sizeof(struct btrfs_item) >
			    sf->block_ctx->len) {
leaf_item_out_of_bounce_error:
				printk(KERN_INFO
				       "btrfsic: leaf item out of bounce at logical %llu, dev %s\n",
				       sf->block_ctx->start,
				       sf->block_ctx->dev->name);
				goto one_stack_frame_backwards;
			}
			btrfsic_read_from_block_data(sf->block_ctx,
						     &disk_item,
						     disk_item_offset,
						     sizeof(struct btrfs_item));
			item_offset = le32_to_cpu(disk_item.offset);
1051
			item_size = le32_to_cpu(disk_item.size);
1052
			disk_key = &disk_item.key;
1053 1054 1055
			type = disk_key->type;

			if (BTRFS_ROOT_ITEM_KEY == type) {
1056 1057 1058 1059 1060 1061
				struct btrfs_root_item root_item;
				u32 root_item_offset;
				u64 next_bytenr;

				root_item_offset = item_offset +
					offsetof(struct btrfs_leaf, items);
1062
				if (root_item_offset + item_size >
1063 1064 1065 1066 1067
				    sf->block_ctx->len)
					goto leaf_item_out_of_bounce_error;
				btrfsic_read_from_block_data(
					sf->block_ctx, &root_item,
					root_item_offset,
1068
					item_size);
1069
				next_bytenr = le64_to_cpu(root_item.bytenr);
1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083

				sf->error =
				    btrfsic_create_link_to_next_block(
						state,
						sf->block,
						sf->block_ctx,
						next_bytenr,
						sf->limit_nesting,
						&sf->next_block_ctx,
						&sf->next_block,
						force_iodone_flag,
						&sf->num_copies,
						&sf->mirror_num,
						disk_key,
1084
						le64_to_cpu(root_item.
1085 1086 1087 1088 1089 1090 1091
						generation));
				if (sf->error)
					goto one_stack_frame_backwards;

				if (NULL != sf->next_block) {
					struct btrfs_header *const next_hdr =
					    (struct btrfs_header *)
1092
					    sf->next_block_ctx.datav[0];
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153

					next_stack =
					    btrfsic_stack_frame_alloc();
					if (NULL == next_stack) {
						btrfsic_release_block_ctx(
								&sf->
								next_block_ctx);
						goto one_stack_frame_backwards;
					}

					next_stack->i = -1;
					next_stack->block = sf->next_block;
					next_stack->block_ctx =
					    &sf->next_block_ctx;
					next_stack->next_block = NULL;
					next_stack->hdr = next_hdr;
					next_stack->limit_nesting =
					    sf->limit_nesting - 1;
					next_stack->prev = sf;
					sf = next_stack;
					goto continue_with_new_stack_frame;
				}
			} else if (BTRFS_EXTENT_DATA_KEY == type &&
				   state->include_extent_data) {
				sf->error = btrfsic_handle_extent_data(
						state,
						sf->block,
						sf->block_ctx,
						item_offset,
						force_iodone_flag);
				if (sf->error)
					goto one_stack_frame_backwards;
			}

			goto continue_with_current_leaf_stack_frame;
		}
	} else {
		struct btrfs_node *const nodehdr = (struct btrfs_node *)sf->hdr;

		if (-1 == sf->i) {
			sf->nr = le32_to_cpu(nodehdr->header.nritems);

			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO "node %llu level %d items %d"
				       " generation %llu owner %llu\n",
				       (unsigned long long)
				       sf->block_ctx->start,
				       nodehdr->header.level, sf->nr,
				       (unsigned long long)
				       le64_to_cpu(nodehdr->header.generation),
				       (unsigned long long)
				       le64_to_cpu(nodehdr->header.owner));
		}

continue_with_current_node_stack_frame:
		if (0 == sf->num_copies || sf->mirror_num > sf->num_copies) {
			sf->i++;
			sf->num_copies = 0;
		}

		if (sf->i < sf->nr) {
1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171
			struct btrfs_key_ptr key_ptr;
			u32 key_ptr_offset;
			u64 next_bytenr;

			key_ptr_offset = (uintptr_t)(nodehdr->ptrs + sf->i) -
					  (uintptr_t)nodehdr;
			if (key_ptr_offset + sizeof(struct btrfs_key_ptr) >
			    sf->block_ctx->len) {
				printk(KERN_INFO
				       "btrfsic: node item out of bounce at logical %llu, dev %s\n",
				       sf->block_ctx->start,
				       sf->block_ctx->dev->name);
				goto one_stack_frame_backwards;
			}
			btrfsic_read_from_block_data(
				sf->block_ctx, &key_ptr, key_ptr_offset,
				sizeof(struct btrfs_key_ptr));
			next_bytenr = le64_to_cpu(key_ptr.blockptr);
1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183

			sf->error = btrfsic_create_link_to_next_block(
					state,
					sf->block,
					sf->block_ctx,
					next_bytenr,
					sf->limit_nesting,
					&sf->next_block_ctx,
					&sf->next_block,
					force_iodone_flag,
					&sf->num_copies,
					&sf->mirror_num,
1184 1185
					&key_ptr.key,
					le64_to_cpu(key_ptr.generation));
1186 1187 1188 1189 1190 1191
			if (sf->error)
				goto one_stack_frame_backwards;

			if (NULL != sf->next_block) {
				struct btrfs_header *const next_hdr =
				    (struct btrfs_header *)
1192
				    sf->next_block_ctx.datav[0];
1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237

				next_stack = btrfsic_stack_frame_alloc();
				if (NULL == next_stack)
					goto one_stack_frame_backwards;

				next_stack->i = -1;
				next_stack->block = sf->next_block;
				next_stack->block_ctx = &sf->next_block_ctx;
				next_stack->next_block = NULL;
				next_stack->hdr = next_hdr;
				next_stack->limit_nesting =
				    sf->limit_nesting - 1;
				next_stack->prev = sf;
				sf = next_stack;
				goto continue_with_new_stack_frame;
			}

			goto continue_with_current_node_stack_frame;
		}
	}

one_stack_frame_backwards:
	if (NULL != sf->prev) {
		struct btrfsic_stack_frame *const prev = sf->prev;

		/* the one for the initial block is freed in the caller */
		btrfsic_release_block_ctx(sf->block_ctx);

		if (sf->error) {
			prev->error = sf->error;
			btrfsic_stack_frame_free(sf);
			sf = prev;
			goto one_stack_frame_backwards;
		}

		btrfsic_stack_frame_free(sf);
		sf = prev;
		goto continue_with_new_stack_frame;
	} else {
		BUG_ON(&initial_stack_frame != sf);
	}

	return sf->error;
}

1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266
static void btrfsic_read_from_block_data(
	struct btrfsic_block_data_ctx *block_ctx,
	void *dstv, u32 offset, size_t len)
{
	size_t cur;
	size_t offset_in_page;
	char *kaddr;
	char *dst = (char *)dstv;
	size_t start_offset = block_ctx->start & ((u64)PAGE_CACHE_SIZE - 1);
	unsigned long i = (start_offset + offset) >> PAGE_CACHE_SHIFT;

	WARN_ON(offset + len > block_ctx->len);
	offset_in_page = (start_offset + offset) &
			 ((unsigned long)PAGE_CACHE_SIZE - 1);

	while (len > 0) {
		cur = min(len, ((size_t)PAGE_CACHE_SIZE - offset_in_page));
		BUG_ON(i >= (block_ctx->len + PAGE_CACHE_SIZE - 1) >>
			    PAGE_CACHE_SHIFT);
		kaddr = block_ctx->datav[i];
		memcpy(dst, kaddr + offset_in_page, cur);

		dst += cur;
		len -= cur;
		offset_in_page = 0;
		i++;
	}
}

1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288
static int btrfsic_create_link_to_next_block(
		struct btrfsic_state *state,
		struct btrfsic_block *block,
		struct btrfsic_block_data_ctx *block_ctx,
		u64 next_bytenr,
		int limit_nesting,
		struct btrfsic_block_data_ctx *next_block_ctx,
		struct btrfsic_block **next_blockp,
		int force_iodone_flag,
		int *num_copiesp, int *mirror_nump,
		struct btrfs_disk_key *disk_key,
		u64 parent_generation)
{
	struct btrfsic_block *next_block = NULL;
	int ret;
	struct btrfsic_block_link *l;
	int did_alloc_block_link;
	int block_was_created;

	*next_blockp = NULL;
	if (0 == *num_copiesp) {
		*num_copiesp =
1289
		    btrfs_num_copies(state->root->fs_info,
1290
				     next_bytenr, state->metablock_size);
1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304
		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
			       (unsigned long long)next_bytenr, *num_copiesp);
		*mirror_nump = 1;
	}

	if (*mirror_nump > *num_copiesp)
		return 0;

	if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
		printk(KERN_INFO
		       "btrfsic_create_link_to_next_block(mirror_num=%d)\n",
		       *mirror_nump);
	ret = btrfsic_map_block(state, next_bytenr,
1305
				state->metablock_size,
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399
				next_block_ctx, *mirror_nump);
	if (ret) {
		printk(KERN_INFO
		       "btrfsic: btrfsic_map_block(@%llu, mirror=%d) failed!\n",
		       (unsigned long long)next_bytenr, *mirror_nump);
		btrfsic_release_block_ctx(next_block_ctx);
		*next_blockp = NULL;
		return -1;
	}

	next_block = btrfsic_block_lookup_or_add(state,
						 next_block_ctx, "referenced ",
						 1, force_iodone_flag,
						 !force_iodone_flag,
						 *mirror_nump,
						 &block_was_created);
	if (NULL == next_block) {
		btrfsic_release_block_ctx(next_block_ctx);
		*next_blockp = NULL;
		return -1;
	}
	if (block_was_created) {
		l = NULL;
		next_block->generation = BTRFSIC_GENERATION_UNKNOWN;
	} else {
		if (next_block->logical_bytenr != next_bytenr &&
		    !(!next_block->is_metadata &&
		      0 == next_block->logical_bytenr)) {
			printk(KERN_INFO
			       "Referenced block @%llu (%s/%llu/%d)"
			       " found in hash table, %c,"
			       " bytenr mismatch (!= stored %llu).\n",
			       (unsigned long long)next_bytenr,
			       next_block_ctx->dev->name,
			       (unsigned long long)next_block_ctx->dev_bytenr,
			       *mirror_nump,
			       btrfsic_get_block_type(state, next_block),
			       (unsigned long long)next_block->logical_bytenr);
		} else if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "Referenced block @%llu (%s/%llu/%d)"
			       " found in hash table, %c.\n",
			       (unsigned long long)next_bytenr,
			       next_block_ctx->dev->name,
			       (unsigned long long)next_block_ctx->dev_bytenr,
			       *mirror_nump,
			       btrfsic_get_block_type(state, next_block));
		next_block->logical_bytenr = next_bytenr;

		next_block->mirror_num = *mirror_nump;
		l = btrfsic_block_link_hashtable_lookup(
				next_block_ctx->dev->bdev,
				next_block_ctx->dev_bytenr,
				block_ctx->dev->bdev,
				block_ctx->dev_bytenr,
				&state->block_link_hashtable);
	}

	next_block->disk_key = *disk_key;
	if (NULL == l) {
		l = btrfsic_block_link_alloc();
		if (NULL == l) {
			printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
			btrfsic_release_block_ctx(next_block_ctx);
			*next_blockp = NULL;
			return -1;
		}

		did_alloc_block_link = 1;
		l->block_ref_to = next_block;
		l->block_ref_from = block;
		l->ref_cnt = 1;
		l->parent_generation = parent_generation;

		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			btrfsic_print_add_link(state, l);

		list_add(&l->node_ref_to, &block->ref_to_list);
		list_add(&l->node_ref_from, &next_block->ref_from_list);

		btrfsic_block_link_hashtable_add(l,
						 &state->block_link_hashtable);
	} else {
		did_alloc_block_link = 0;
		if (0 == limit_nesting) {
			l->ref_cnt++;
			l->parent_generation = parent_generation;
			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				btrfsic_print_add_link(state, l);
		}
	}

	if (limit_nesting > 0 && did_alloc_block_link) {
		ret = btrfsic_read_block(state, next_block_ctx);
1400
		if (ret < (int)next_block_ctx->len) {
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424
			printk(KERN_INFO
			       "btrfsic: read block @logical %llu failed!\n",
			       (unsigned long long)next_bytenr);
			btrfsic_release_block_ctx(next_block_ctx);
			*next_blockp = NULL;
			return -1;
		}

		*next_blockp = next_block;
	} else {
		*next_blockp = NULL;
	}
	(*mirror_nump)++;

	return 0;
}

static int btrfsic_handle_extent_data(
		struct btrfsic_state *state,
		struct btrfsic_block *block,
		struct btrfsic_block_data_ctx *block_ctx,
		u32 item_offset, int force_iodone_flag)
{
	int ret;
1425 1426 1427 1428 1429
	struct btrfs_file_extent_item file_extent_item;
	u64 file_extent_item_offset;
	u64 next_bytenr;
	u64 num_bytes;
	u64 generation;
1430 1431
	struct btrfsic_block_link *l;

1432 1433
	file_extent_item_offset = offsetof(struct btrfs_leaf, items) +
				  item_offset;
1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455
	if (file_extent_item_offset +
	    offsetof(struct btrfs_file_extent_item, disk_num_bytes) >
	    block_ctx->len) {
		printk(KERN_INFO
		       "btrfsic: file item out of bounce at logical %llu, dev %s\n",
		       block_ctx->start, block_ctx->dev->name);
		return -1;
	}

	btrfsic_read_from_block_data(block_ctx, &file_extent_item,
		file_extent_item_offset,
		offsetof(struct btrfs_file_extent_item, disk_num_bytes));
	if (BTRFS_FILE_EXTENT_REG != file_extent_item.type ||
	    ((u64)0) == le64_to_cpu(file_extent_item.disk_bytenr)) {
		if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
			printk(KERN_INFO "extent_data: type %u, disk_bytenr = %llu\n",
			       file_extent_item.type,
			       (unsigned long long)
			       le64_to_cpu(file_extent_item.disk_bytenr));
		return 0;
	}

1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471
	if (file_extent_item_offset + sizeof(struct btrfs_file_extent_item) >
	    block_ctx->len) {
		printk(KERN_INFO
		       "btrfsic: file item out of bounce at logical %llu, dev %s\n",
		       block_ctx->start, block_ctx->dev->name);
		return -1;
	}
	btrfsic_read_from_block_data(block_ctx, &file_extent_item,
				     file_extent_item_offset,
				     sizeof(struct btrfs_file_extent_item));
	next_bytenr = le64_to_cpu(file_extent_item.disk_bytenr) +
		      le64_to_cpu(file_extent_item.offset);
	generation = le64_to_cpu(file_extent_item.generation);
	num_bytes = le64_to_cpu(file_extent_item.num_bytes);
	generation = le64_to_cpu(file_extent_item.generation);

1472 1473 1474
	if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
		printk(KERN_INFO "extent_data: type %u, disk_bytenr = %llu,"
		       " offset = %llu, num_bytes = %llu\n",
1475
		       file_extent_item.type,
1476
		       (unsigned long long)
1477 1478 1479
		       le64_to_cpu(file_extent_item.disk_bytenr),
		       (unsigned long long)le64_to_cpu(file_extent_item.offset),
		       (unsigned long long)num_bytes);
1480 1481 1482 1483 1484
	while (num_bytes > 0) {
		u32 chunk_len;
		int num_copies;
		int mirror_num;

1485 1486
		if (num_bytes > state->datablock_size)
			chunk_len = state->datablock_size;
1487 1488 1489 1490
		else
			chunk_len = num_bytes;

		num_copies =
1491
		    btrfs_num_copies(state->root->fs_info,
1492
				     next_bytenr, state->datablock_size);
1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583
		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
			       (unsigned long long)next_bytenr, num_copies);
		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
			struct btrfsic_block_data_ctx next_block_ctx;
			struct btrfsic_block *next_block;
			int block_was_created;

			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO "btrfsic_handle_extent_data("
				       "mirror_num=%d)\n", mirror_num);
			if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
				printk(KERN_INFO
				       "\tdisk_bytenr = %llu, num_bytes %u\n",
				       (unsigned long long)next_bytenr,
				       chunk_len);
			ret = btrfsic_map_block(state, next_bytenr,
						chunk_len, &next_block_ctx,
						mirror_num);
			if (ret) {
				printk(KERN_INFO
				       "btrfsic: btrfsic_map_block(@%llu,"
				       " mirror=%d) failed!\n",
				       (unsigned long long)next_bytenr,
				       mirror_num);
				return -1;
			}

			next_block = btrfsic_block_lookup_or_add(
					state,
					&next_block_ctx,
					"referenced ",
					0,
					force_iodone_flag,
					!force_iodone_flag,
					mirror_num,
					&block_was_created);
			if (NULL == next_block) {
				printk(KERN_INFO
				       "btrfsic: error, kmalloc failed!\n");
				btrfsic_release_block_ctx(&next_block_ctx);
				return -1;
			}
			if (!block_was_created) {
				if (next_block->logical_bytenr != next_bytenr &&
				    !(!next_block->is_metadata &&
				      0 == next_block->logical_bytenr)) {
					printk(KERN_INFO
					       "Referenced block"
					       " @%llu (%s/%llu/%d)"
					       " found in hash table, D,"
					       " bytenr mismatch"
					       " (!= stored %llu).\n",
					       (unsigned long long)next_bytenr,
					       next_block_ctx.dev->name,
					       (unsigned long long)
					       next_block_ctx.dev_bytenr,
					       mirror_num,
					       (unsigned long long)
					       next_block->logical_bytenr);
				}
				next_block->logical_bytenr = next_bytenr;
				next_block->mirror_num = mirror_num;
			}

			l = btrfsic_block_link_lookup_or_add(state,
							     &next_block_ctx,
							     next_block, block,
							     generation);
			btrfsic_release_block_ctx(&next_block_ctx);
			if (NULL == l)
				return -1;
		}

		next_bytenr += chunk_len;
		num_bytes -= chunk_len;
	}

	return 0;
}

static int btrfsic_map_block(struct btrfsic_state *state, u64 bytenr, u32 len,
			     struct btrfsic_block_data_ctx *block_ctx_out,
			     int mirror_num)
{
	int ret;
	u64 length;
	struct btrfs_bio *multi = NULL;
	struct btrfs_device *device;

	length = len;
1584
	ret = btrfs_map_block(state->root->fs_info, READ,
1585 1586
			      bytenr, &length, &multi, mirror_num);

1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598
	if (ret) {
		block_ctx_out->start = 0;
		block_ctx_out->dev_bytenr = 0;
		block_ctx_out->len = 0;
		block_ctx_out->dev = NULL;
		block_ctx_out->datav = NULL;
		block_ctx_out->pagev = NULL;
		block_ctx_out->mem_to_free = NULL;

		return ret;
	}

1599 1600 1601 1602 1603
	device = multi->stripes[0].dev;
	block_ctx_out->dev = btrfsic_dev_state_lookup(device->bdev);
	block_ctx_out->dev_bytenr = multi->stripes[0].physical;
	block_ctx_out->start = bytenr;
	block_ctx_out->len = len;
1604 1605 1606
	block_ctx_out->datav = NULL;
	block_ctx_out->pagev = NULL;
	block_ctx_out->mem_to_free = NULL;
1607

1608
	kfree(multi);
1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624
	if (NULL == block_ctx_out->dev) {
		ret = -ENXIO;
		printk(KERN_INFO "btrfsic: error, cannot lookup dev (#1)!\n");
	}

	return ret;
}

static int btrfsic_map_superblock(struct btrfsic_state *state, u64 bytenr,
				  u32 len, struct block_device *bdev,
				  struct btrfsic_block_data_ctx *block_ctx_out)
{
	block_ctx_out->dev = btrfsic_dev_state_lookup(bdev);
	block_ctx_out->dev_bytenr = bytenr;
	block_ctx_out->start = bytenr;
	block_ctx_out->len = len;
1625 1626 1627
	block_ctx_out->datav = NULL;
	block_ctx_out->pagev = NULL;
	block_ctx_out->mem_to_free = NULL;
1628 1629 1630 1631 1632 1633 1634 1635 1636 1637
	if (NULL != block_ctx_out->dev) {
		return 0;
	} else {
		printk(KERN_INFO "btrfsic: error, cannot lookup dev (#2)!\n");
		return -ENXIO;
	}
}

static void btrfsic_release_block_ctx(struct btrfsic_block_data_ctx *block_ctx)
{
1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660
	if (block_ctx->mem_to_free) {
		unsigned int num_pages;

		BUG_ON(!block_ctx->datav);
		BUG_ON(!block_ctx->pagev);
		num_pages = (block_ctx->len + (u64)PAGE_CACHE_SIZE - 1) >>
			    PAGE_CACHE_SHIFT;
		while (num_pages > 0) {
			num_pages--;
			if (block_ctx->datav[num_pages]) {
				kunmap(block_ctx->pagev[num_pages]);
				block_ctx->datav[num_pages] = NULL;
			}
			if (block_ctx->pagev[num_pages]) {
				__free_page(block_ctx->pagev[num_pages]);
				block_ctx->pagev[num_pages] = NULL;
			}
		}

		kfree(block_ctx->mem_to_free);
		block_ctx->mem_to_free = NULL;
		block_ctx->pagev = NULL;
		block_ctx->datav = NULL;
1661 1662 1663 1664 1665 1666
	}
}

static int btrfsic_read_block(struct btrfsic_state *state,
			      struct btrfsic_block_data_ctx *block_ctx)
{
1667 1668 1669 1670 1671 1672 1673 1674 1675
	unsigned int num_pages;
	unsigned int i;
	u64 dev_bytenr;
	int ret;

	BUG_ON(block_ctx->datav);
	BUG_ON(block_ctx->pagev);
	BUG_ON(block_ctx->mem_to_free);
	if (block_ctx->dev_bytenr & ((u64)PAGE_CACHE_SIZE - 1)) {
1676 1677 1678 1679 1680
		printk(KERN_INFO
		       "btrfsic: read_block() with unaligned bytenr %llu\n",
		       (unsigned long long)block_ctx->dev_bytenr);
		return -1;
	}
1681 1682 1683 1684 1685 1686 1687

	num_pages = (block_ctx->len + (u64)PAGE_CACHE_SIZE - 1) >>
		    PAGE_CACHE_SHIFT;
	block_ctx->mem_to_free = kzalloc((sizeof(*block_ctx->datav) +
					  sizeof(*block_ctx->pagev)) *
					 num_pages, GFP_NOFS);
	if (!block_ctx->mem_to_free)
1688
		return -1;
1689 1690 1691 1692 1693 1694
	block_ctx->datav = block_ctx->mem_to_free;
	block_ctx->pagev = (struct page **)(block_ctx->datav + num_pages);
	for (i = 0; i < num_pages; i++) {
		block_ctx->pagev[i] = alloc_page(GFP_NOFS);
		if (!block_ctx->pagev[i])
			return -1;
1695 1696
	}

1697 1698 1699 1700 1701 1702
	dev_bytenr = block_ctx->dev_bytenr;
	for (i = 0; i < num_pages;) {
		struct bio *bio;
		unsigned int j;
		DECLARE_COMPLETION_ONSTACK(complete);

1703
		bio = btrfs_io_bio_alloc(GFP_NOFS, num_pages - i);
1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749
		if (!bio) {
			printk(KERN_INFO
			       "btrfsic: bio_alloc() for %u pages failed!\n",
			       num_pages - i);
			return -1;
		}
		bio->bi_bdev = block_ctx->dev->bdev;
		bio->bi_sector = dev_bytenr >> 9;
		bio->bi_end_io = btrfsic_complete_bio_end_io;
		bio->bi_private = &complete;

		for (j = i; j < num_pages; j++) {
			ret = bio_add_page(bio, block_ctx->pagev[j],
					   PAGE_CACHE_SIZE, 0);
			if (PAGE_CACHE_SIZE != ret)
				break;
		}
		if (j == i) {
			printk(KERN_INFO
			       "btrfsic: error, failed to add a single page!\n");
			return -1;
		}
		submit_bio(READ, bio);

		/* this will also unplug the queue */
		wait_for_completion(&complete);

		if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) {
			printk(KERN_INFO
			       "btrfsic: read error at logical %llu dev %s!\n",
			       block_ctx->start, block_ctx->dev->name);
			bio_put(bio);
			return -1;
		}
		bio_put(bio);
		dev_bytenr += (j - i) * PAGE_CACHE_SIZE;
		i = j;
	}
	for (i = 0; i < num_pages; i++) {
		block_ctx->datav[i] = kmap(block_ctx->pagev[i]);
		if (!block_ctx->datav[i]) {
			printk(KERN_INFO "btrfsic: kmap() failed (dev %s)!\n",
			       block_ctx->dev->name);
			return -1;
		}
	}
1750 1751 1752 1753

	return block_ctx->len;
}

1754 1755 1756 1757 1758
static void btrfsic_complete_bio_end_io(struct bio *bio, int err)
{
	complete((struct completion *)bio->bi_private);
}

1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 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
static void btrfsic_dump_database(struct btrfsic_state *state)
{
	struct list_head *elem_all;

	BUG_ON(NULL == state);

	printk(KERN_INFO "all_blocks_list:\n");
	list_for_each(elem_all, &state->all_blocks_list) {
		const struct btrfsic_block *const b_all =
		    list_entry(elem_all, struct btrfsic_block,
			       all_blocks_node);
		struct list_head *elem_ref_to;
		struct list_head *elem_ref_from;

		printk(KERN_INFO "%c-block @%llu (%s/%llu/%d)\n",
		       btrfsic_get_block_type(state, b_all),
		       (unsigned long long)b_all->logical_bytenr,
		       b_all->dev_state->name,
		       (unsigned long long)b_all->dev_bytenr,
		       b_all->mirror_num);

		list_for_each(elem_ref_to, &b_all->ref_to_list) {
			const struct btrfsic_block_link *const l =
			    list_entry(elem_ref_to,
				       struct btrfsic_block_link,
				       node_ref_to);

			printk(KERN_INFO " %c @%llu (%s/%llu/%d)"
			       " refers %u* to"
			       " %c @%llu (%s/%llu/%d)\n",
			       btrfsic_get_block_type(state, b_all),
			       (unsigned long long)b_all->logical_bytenr,
			       b_all->dev_state->name,
			       (unsigned long long)b_all->dev_bytenr,
			       b_all->mirror_num,
			       l->ref_cnt,
			       btrfsic_get_block_type(state, l->block_ref_to),
			       (unsigned long long)
			       l->block_ref_to->logical_bytenr,
			       l->block_ref_to->dev_state->name,
			       (unsigned long long)l->block_ref_to->dev_bytenr,
			       l->block_ref_to->mirror_num);
		}

		list_for_each(elem_ref_from, &b_all->ref_from_list) {
			const struct btrfsic_block_link *const l =
			    list_entry(elem_ref_from,
				       struct btrfsic_block_link,
				       node_ref_from);

			printk(KERN_INFO " %c @%llu (%s/%llu/%d)"
			       " is ref %u* from"
			       " %c @%llu (%s/%llu/%d)\n",
			       btrfsic_get_block_type(state, b_all),
			       (unsigned long long)b_all->logical_bytenr,
			       b_all->dev_state->name,
			       (unsigned long long)b_all->dev_bytenr,
			       b_all->mirror_num,
			       l->ref_cnt,
			       btrfsic_get_block_type(state, l->block_ref_from),
			       (unsigned long long)
			       l->block_ref_from->logical_bytenr,
			       l->block_ref_from->dev_state->name,
			       (unsigned long long)
			       l->block_ref_from->dev_bytenr,
			       l->block_ref_from->mirror_num);
		}

		printk(KERN_INFO "\n");
	}
}

/*
 * Test whether the disk block contains a tree block (leaf or node)
 * (note that this test fails for the super block)
 */
static int btrfsic_test_for_metadata(struct btrfsic_state *state,
1836
				     char **datav, unsigned int num_pages)
1837 1838 1839 1840
{
	struct btrfs_header *h;
	u8 csum[BTRFS_CSUM_SIZE];
	u32 crc = ~(u32)0;
1841
	unsigned int i;
1842

1843 1844 1845 1846
	if (num_pages * PAGE_CACHE_SIZE < state->metablock_size)
		return 1; /* not metadata */
	num_pages = state->metablock_size >> PAGE_CACHE_SHIFT;
	h = (struct btrfs_header *)datav[0];
1847 1848

	if (memcmp(h->fsid, state->root->fs_info->fsid, BTRFS_UUID_SIZE))
1849
		return 1;
1850

1851 1852 1853 1854 1855 1856 1857
	for (i = 0; i < num_pages; i++) {
		u8 *data = i ? datav[i] : (datav[i] + BTRFS_CSUM_SIZE);
		size_t sublen = i ? PAGE_CACHE_SIZE :
				    (PAGE_CACHE_SIZE - BTRFS_CSUM_SIZE);

		crc = crc32c(crc, data, sublen);
	}
1858 1859
	btrfs_csum_final(crc, csum);
	if (memcmp(csum, h->csum, state->csum_size))
1860
		return 1;
1861

1862
	return 0; /* is metadata */
1863 1864 1865
}

static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
1866 1867 1868
					  u64 dev_bytenr, char **mapped_datav,
					  unsigned int num_pages,
					  struct bio *bio, int *bio_is_patched,
1869 1870 1871 1872 1873 1874 1875 1876 1877
					  struct buffer_head *bh,
					  int submit_bio_bh_rw)
{
	int is_metadata;
	struct btrfsic_block *block;
	struct btrfsic_block_data_ctx block_ctx;
	int ret;
	struct btrfsic_state *state = dev_state->state;
	struct block_device *bdev = dev_state->bdev;
1878
	unsigned int processed_len;
1879 1880 1881 1882

	if (NULL != bio_is_patched)
		*bio_is_patched = 0;

1883 1884 1885 1886 1887 1888 1889 1890
again:
	if (num_pages == 0)
		return;

	processed_len = 0;
	is_metadata = (0 == btrfsic_test_for_metadata(state, mapped_datav,
						      num_pages));

1891 1892 1893
	block = btrfsic_block_hashtable_lookup(bdev, dev_bytenr,
					       &state->block_hashtable);
	if (NULL != block) {
1894
		u64 bytenr = 0;
1895 1896 1897 1898 1899
		struct list_head *elem_ref_to;
		struct list_head *tmp_ref_to;

		if (block->is_superblock) {
			bytenr = le64_to_cpu(((struct btrfs_super_block *)
1900 1901 1902 1903 1904 1905 1906
					      mapped_datav[0])->bytenr);
			if (num_pages * PAGE_CACHE_SIZE <
			    BTRFS_SUPER_INFO_SIZE) {
				printk(KERN_INFO
				       "btrfsic: cannot work with too short bios!\n");
				return;
			}
1907
			is_metadata = 1;
1908 1909
			BUG_ON(BTRFS_SUPER_INFO_SIZE & (PAGE_CACHE_SIZE - 1));
			processed_len = BTRFS_SUPER_INFO_SIZE;
1910 1911 1912 1913 1914 1915 1916 1917 1918
			if (state->print_mask &
			    BTRFSIC_PRINT_MASK_TREE_BEFORE_SB_WRITE) {
				printk(KERN_INFO
				       "[before new superblock is written]:\n");
				btrfsic_dump_tree_sub(state, block, 0);
			}
		}
		if (is_metadata) {
			if (!block->is_superblock) {
1919 1920 1921 1922 1923 1924 1925
				if (num_pages * PAGE_CACHE_SIZE <
				    state->metablock_size) {
					printk(KERN_INFO
					       "btrfsic: cannot work with too short bios!\n");
					return;
				}
				processed_len = state->metablock_size;
1926
				bytenr = le64_to_cpu(((struct btrfs_header *)
1927
						      mapped_datav[0])->bytenr);
1928 1929
				btrfsic_cmp_log_and_dev_bytenr(state, bytenr,
							       dev_state,
1930
							       dev_bytenr);
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
			}
			if (block->logical_bytenr != bytenr) {
				printk(KERN_INFO
				       "Written block @%llu (%s/%llu/%d)"
				       " found in hash table, %c,"
				       " bytenr mismatch"
				       " (!= stored %llu).\n",
				       (unsigned long long)bytenr,
				       dev_state->name,
				       (unsigned long long)dev_bytenr,
				       block->mirror_num,
				       btrfsic_get_block_type(state, block),
				       (unsigned long long)
				       block->logical_bytenr);
				block->logical_bytenr = bytenr;
			} else if (state->print_mask &
				   BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO
				       "Written block @%llu (%s/%llu/%d)"
				       " found in hash table, %c.\n",
				       (unsigned long long)bytenr,
				       dev_state->name,
				       (unsigned long long)dev_bytenr,
				       block->mirror_num,
				       btrfsic_get_block_type(state, block));
		} else {
1957 1958 1959 1960 1961 1962 1963
			if (num_pages * PAGE_CACHE_SIZE <
			    state->datablock_size) {
				printk(KERN_INFO
				       "btrfsic: cannot work with too short bios!\n");
				return;
			}
			processed_len = state->datablock_size;
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
			bytenr = block->logical_bytenr;
			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO
				       "Written block @%llu (%s/%llu/%d)"
				       " found in hash table, %c.\n",
				       (unsigned long long)bytenr,
				       dev_state->name,
				       (unsigned long long)dev_bytenr,
				       block->mirror_num,
				       btrfsic_get_block_type(state, block));
		}

		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "ref_to_list: %cE, ref_from_list: %cE\n",
			       list_empty(&block->ref_to_list) ? ' ' : '!',
			       list_empty(&block->ref_from_list) ? ' ' : '!');
		if (btrfsic_is_block_ref_by_superblock(state, block, 0)) {
			printk(KERN_INFO "btrfs: attempt to overwrite %c-block"
			       " @%llu (%s/%llu/%d), old(gen=%llu,"
			       " objectid=%llu, type=%d, offset=%llu),"
			       " new(gen=%llu),"
			       " which is referenced by most recent superblock"
			       " (superblockgen=%llu)!\n",
			       btrfsic_get_block_type(state, block),
			       (unsigned long long)bytenr,
			       dev_state->name,
			       (unsigned long long)dev_bytenr,
			       block->mirror_num,
			       (unsigned long long)block->generation,
			       (unsigned long long)
			       le64_to_cpu(block->disk_key.objectid),
			       block->disk_key.type,
			       (unsigned long long)
			       le64_to_cpu(block->disk_key.offset),
			       (unsigned long long)
			       le64_to_cpu(((struct btrfs_header *)
2001
					    mapped_datav[0])->generation),
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
			       (unsigned long long)
			       state->max_superblock_generation);
			btrfsic_dump_tree(state);
		}

		if (!block->is_iodone && !block->never_written) {
			printk(KERN_INFO "btrfs: attempt to overwrite %c-block"
			       " @%llu (%s/%llu/%d), oldgen=%llu, newgen=%llu,"
			       " which is not yet iodone!\n",
			       btrfsic_get_block_type(state, block),
			       (unsigned long long)bytenr,
			       dev_state->name,
			       (unsigned long long)dev_bytenr,
			       block->mirror_num,
			       (unsigned long long)block->generation,
			       (unsigned long long)
			       le64_to_cpu(((struct btrfs_header *)
2019
					    mapped_datav[0])->generation));
2020 2021
			/* it would not be safe to go on */
			btrfsic_dump_tree(state);
2022
			goto continue_loop;
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
		}

		/*
		 * Clear all references of this block. Do not free
		 * the block itself even if is not referenced anymore
		 * because it still carries valueable information
		 * like whether it was ever written and IO completed.
		 */
		list_for_each_safe(elem_ref_to, tmp_ref_to,
				   &block->ref_to_list) {
			struct btrfsic_block_link *const l =
			    list_entry(elem_ref_to,
				       struct btrfsic_block_link,
				       node_ref_to);

			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				btrfsic_print_rem_link(state, l);
			l->ref_cnt--;
			if (0 == l->ref_cnt) {
				list_del(&l->node_ref_to);
				list_del(&l->node_ref_from);
				btrfsic_block_link_hashtable_remove(l);
				btrfsic_block_link_free(l);
			}
		}

		if (block->is_superblock)
2050 2051
			ret = btrfsic_map_superblock(state, bytenr,
						     processed_len,
2052 2053
						     bdev, &block_ctx);
		else
2054
			ret = btrfsic_map_block(state, bytenr, processed_len,
2055 2056 2057 2058 2059
						&block_ctx, 0);
		if (ret) {
			printk(KERN_INFO
			       "btrfsic: btrfsic_map_block(root @%llu)"
			       " failed!\n", (unsigned long long)bytenr);
2060
			goto continue_loop;
2061
		}
2062
		block_ctx.datav = mapped_datav;
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 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117
		/* the following is required in case of writes to mirrors,
		 * use the same that was used for the lookup */
		block_ctx.dev = dev_state;
		block_ctx.dev_bytenr = dev_bytenr;

		if (is_metadata || state->include_extent_data) {
			block->never_written = 0;
			block->iodone_w_error = 0;
			if (NULL != bio) {
				block->is_iodone = 0;
				BUG_ON(NULL == bio_is_patched);
				if (!*bio_is_patched) {
					block->orig_bio_bh_private =
					    bio->bi_private;
					block->orig_bio_bh_end_io.bio =
					    bio->bi_end_io;
					block->next_in_same_bio = NULL;
					bio->bi_private = block;
					bio->bi_end_io = btrfsic_bio_end_io;
					*bio_is_patched = 1;
				} else {
					struct btrfsic_block *chained_block =
					    (struct btrfsic_block *)
					    bio->bi_private;

					BUG_ON(NULL == chained_block);
					block->orig_bio_bh_private =
					    chained_block->orig_bio_bh_private;
					block->orig_bio_bh_end_io.bio =
					    chained_block->orig_bio_bh_end_io.
					    bio;
					block->next_in_same_bio = chained_block;
					bio->bi_private = block;
				}
			} else if (NULL != bh) {
				block->is_iodone = 0;
				block->orig_bio_bh_private = bh->b_private;
				block->orig_bio_bh_end_io.bh = bh->b_end_io;
				block->next_in_same_bio = NULL;
				bh->b_private = block;
				bh->b_end_io = btrfsic_bh_end_io;
			} else {
				block->is_iodone = 1;
				block->orig_bio_bh_private = NULL;
				block->orig_bio_bh_end_io.bio = NULL;
				block->next_in_same_bio = NULL;
			}
		}

		block->flush_gen = dev_state->last_flush_gen + 1;
		block->submit_bio_bh_rw = submit_bio_bh_rw;
		if (is_metadata) {
			block->logical_bytenr = bytenr;
			block->is_metadata = 1;
			if (block->is_superblock) {
2118 2119
				BUG_ON(PAGE_CACHE_SIZE !=
				       BTRFS_SUPER_INFO_SIZE);
2120 2121 2122 2123
				ret = btrfsic_process_written_superblock(
						state,
						block,
						(struct btrfs_super_block *)
2124
						mapped_datav[0]);
2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166
				if (state->print_mask &
				    BTRFSIC_PRINT_MASK_TREE_AFTER_SB_WRITE) {
					printk(KERN_INFO
					"[after new superblock is written]:\n");
					btrfsic_dump_tree_sub(state, block, 0);
				}
			} else {
				block->mirror_num = 0;	/* unknown */
				ret = btrfsic_process_metablock(
						state,
						block,
						&block_ctx,
						0, 0);
			}
			if (ret)
				printk(KERN_INFO
				       "btrfsic: btrfsic_process_metablock"
				       "(root @%llu) failed!\n",
				       (unsigned long long)dev_bytenr);
		} else {
			block->is_metadata = 0;
			block->mirror_num = 0;	/* unknown */
			block->generation = BTRFSIC_GENERATION_UNKNOWN;
			if (!state->include_extent_data
			    && list_empty(&block->ref_from_list)) {
				/*
				 * disk block is overwritten with extent
				 * data (not meta data) and we are configured
				 * to not include extent data: take the
				 * chance and free the block's memory
				 */
				btrfsic_block_hashtable_remove(block);
				list_del(&block->all_blocks_node);
				btrfsic_block_free(block);
			}
		}
		btrfsic_release_block_ctx(&block_ctx);
	} else {
		/* block has not been found in hash table */
		u64 bytenr;

		if (!is_metadata) {
2167
			processed_len = state->datablock_size;
2168 2169 2170 2171 2172
			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO "Written block (%s/%llu/?)"
				       " !found in hash table, D.\n",
				       dev_state->name,
				       (unsigned long long)dev_bytenr);
2173 2174 2175 2176
			if (!state->include_extent_data) {
				/* ignore that written D block */
				goto continue_loop;
			}
2177 2178 2179 2180 2181

			/* this is getting ugly for the
			 * include_extent_data case... */
			bytenr = 0;	/* unknown */
			block_ctx.start = bytenr;
2182 2183 2184
			block_ctx.len = processed_len;
			block_ctx.mem_to_free = NULL;
			block_ctx.pagev = NULL;
2185
		} else {
2186
			processed_len = state->metablock_size;
2187
			bytenr = le64_to_cpu(((struct btrfs_header *)
2188
					      mapped_datav[0])->bytenr);
2189
			btrfsic_cmp_log_and_dev_bytenr(state, bytenr, dev_state,
2190
						       dev_bytenr);
2191 2192 2193 2194 2195 2196 2197 2198
			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO
				       "Written block @%llu (%s/%llu/?)"
				       " !found in hash table, M.\n",
				       (unsigned long long)bytenr,
				       dev_state->name,
				       (unsigned long long)dev_bytenr);

2199 2200
			ret = btrfsic_map_block(state, bytenr, processed_len,
						&block_ctx, 0);
2201 2202 2203 2204 2205
			if (ret) {
				printk(KERN_INFO
				       "btrfsic: btrfsic_map_block(root @%llu)"
				       " failed!\n",
				       (unsigned long long)dev_bytenr);
2206
				goto continue_loop;
2207 2208
			}
		}
2209
		block_ctx.datav = mapped_datav;
2210 2211 2212 2213 2214 2215 2216 2217 2218
		/* the following is required in case of writes to mirrors,
		 * use the same that was used for the lookup */
		block_ctx.dev = dev_state;
		block_ctx.dev_bytenr = dev_bytenr;

		block = btrfsic_block_alloc();
		if (NULL == block) {
			printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
			btrfsic_release_block_ctx(&block_ctx);
2219
			goto continue_loop;
2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278
		}
		block->dev_state = dev_state;
		block->dev_bytenr = dev_bytenr;
		block->logical_bytenr = bytenr;
		block->is_metadata = is_metadata;
		block->never_written = 0;
		block->iodone_w_error = 0;
		block->mirror_num = 0;	/* unknown */
		block->flush_gen = dev_state->last_flush_gen + 1;
		block->submit_bio_bh_rw = submit_bio_bh_rw;
		if (NULL != bio) {
			block->is_iodone = 0;
			BUG_ON(NULL == bio_is_patched);
			if (!*bio_is_patched) {
				block->orig_bio_bh_private = bio->bi_private;
				block->orig_bio_bh_end_io.bio = bio->bi_end_io;
				block->next_in_same_bio = NULL;
				bio->bi_private = block;
				bio->bi_end_io = btrfsic_bio_end_io;
				*bio_is_patched = 1;
			} else {
				struct btrfsic_block *chained_block =
				    (struct btrfsic_block *)
				    bio->bi_private;

				BUG_ON(NULL == chained_block);
				block->orig_bio_bh_private =
				    chained_block->orig_bio_bh_private;
				block->orig_bio_bh_end_io.bio =
				    chained_block->orig_bio_bh_end_io.bio;
				block->next_in_same_bio = chained_block;
				bio->bi_private = block;
			}
		} else if (NULL != bh) {
			block->is_iodone = 0;
			block->orig_bio_bh_private = bh->b_private;
			block->orig_bio_bh_end_io.bh = bh->b_end_io;
			block->next_in_same_bio = NULL;
			bh->b_private = block;
			bh->b_end_io = btrfsic_bh_end_io;
		} else {
			block->is_iodone = 1;
			block->orig_bio_bh_private = NULL;
			block->orig_bio_bh_end_io.bio = NULL;
			block->next_in_same_bio = NULL;
		}
		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "New written %c-block @%llu (%s/%llu/%d)\n",
			       is_metadata ? 'M' : 'D',
			       (unsigned long long)block->logical_bytenr,
			       block->dev_state->name,
			       (unsigned long long)block->dev_bytenr,
			       block->mirror_num);
		list_add(&block->all_blocks_node, &state->all_blocks_list);
		btrfsic_block_hashtable_add(block, &state->block_hashtable);

		if (is_metadata) {
			ret = btrfsic_process_metablock(state, block,
2279
							&block_ctx, 0, 0);
2280 2281 2282 2283 2284 2285 2286 2287
			if (ret)
				printk(KERN_INFO
				       "btrfsic: process_metablock(root @%llu)"
				       " failed!\n",
				       (unsigned long long)dev_bytenr);
		}
		btrfsic_release_block_ctx(&block_ctx);
	}
2288 2289 2290 2291 2292 2293 2294

continue_loop:
	BUG_ON(!processed_len);
	dev_bytenr += processed_len;
	mapped_datav += processed_len >> PAGE_CACHE_SHIFT;
	num_pages -= processed_len >> PAGE_CACHE_SHIFT;
	goto again;
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 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 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475
}

static void btrfsic_bio_end_io(struct bio *bp, int bio_error_status)
{
	struct btrfsic_block *block = (struct btrfsic_block *)bp->bi_private;
	int iodone_w_error;

	/* mutex is not held! This is not save if IO is not yet completed
	 * on umount */
	iodone_w_error = 0;
	if (bio_error_status)
		iodone_w_error = 1;

	BUG_ON(NULL == block);
	bp->bi_private = block->orig_bio_bh_private;
	bp->bi_end_io = block->orig_bio_bh_end_io.bio;

	do {
		struct btrfsic_block *next_block;
		struct btrfsic_dev_state *const dev_state = block->dev_state;

		if ((dev_state->state->print_mask &
		     BTRFSIC_PRINT_MASK_END_IO_BIO_BH))
			printk(KERN_INFO
			       "bio_end_io(err=%d) for %c @%llu (%s/%llu/%d)\n",
			       bio_error_status,
			       btrfsic_get_block_type(dev_state->state, block),
			       (unsigned long long)block->logical_bytenr,
			       dev_state->name,
			       (unsigned long long)block->dev_bytenr,
			       block->mirror_num);
		next_block = block->next_in_same_bio;
		block->iodone_w_error = iodone_w_error;
		if (block->submit_bio_bh_rw & REQ_FLUSH) {
			dev_state->last_flush_gen++;
			if ((dev_state->state->print_mask &
			     BTRFSIC_PRINT_MASK_END_IO_BIO_BH))
				printk(KERN_INFO
				       "bio_end_io() new %s flush_gen=%llu\n",
				       dev_state->name,
				       (unsigned long long)
				       dev_state->last_flush_gen);
		}
		if (block->submit_bio_bh_rw & REQ_FUA)
			block->flush_gen = 0; /* FUA completed means block is
					       * on disk */
		block->is_iodone = 1; /* for FLUSH, this releases the block */
		block = next_block;
	} while (NULL != block);

	bp->bi_end_io(bp, bio_error_status);
}

static void btrfsic_bh_end_io(struct buffer_head *bh, int uptodate)
{
	struct btrfsic_block *block = (struct btrfsic_block *)bh->b_private;
	int iodone_w_error = !uptodate;
	struct btrfsic_dev_state *dev_state;

	BUG_ON(NULL == block);
	dev_state = block->dev_state;
	if ((dev_state->state->print_mask & BTRFSIC_PRINT_MASK_END_IO_BIO_BH))
		printk(KERN_INFO
		       "bh_end_io(error=%d) for %c @%llu (%s/%llu/%d)\n",
		       iodone_w_error,
		       btrfsic_get_block_type(dev_state->state, block),
		       (unsigned long long)block->logical_bytenr,
		       block->dev_state->name,
		       (unsigned long long)block->dev_bytenr,
		       block->mirror_num);

	block->iodone_w_error = iodone_w_error;
	if (block->submit_bio_bh_rw & REQ_FLUSH) {
		dev_state->last_flush_gen++;
		if ((dev_state->state->print_mask &
		     BTRFSIC_PRINT_MASK_END_IO_BIO_BH))
			printk(KERN_INFO
			       "bh_end_io() new %s flush_gen=%llu\n",
			       dev_state->name,
			       (unsigned long long)dev_state->last_flush_gen);
	}
	if (block->submit_bio_bh_rw & REQ_FUA)
		block->flush_gen = 0; /* FUA completed means block is on disk */

	bh->b_private = block->orig_bio_bh_private;
	bh->b_end_io = block->orig_bio_bh_end_io.bh;
	block->is_iodone = 1; /* for FLUSH, this releases the block */
	bh->b_end_io(bh, uptodate);
}

static int btrfsic_process_written_superblock(
		struct btrfsic_state *state,
		struct btrfsic_block *const superblock,
		struct btrfs_super_block *const super_hdr)
{
	int pass;

	superblock->generation = btrfs_super_generation(super_hdr);
	if (!(superblock->generation > state->max_superblock_generation ||
	      0 == state->max_superblock_generation)) {
		if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE)
			printk(KERN_INFO
			       "btrfsic: superblock @%llu (%s/%llu/%d)"
			       " with old gen %llu <= %llu\n",
			       (unsigned long long)superblock->logical_bytenr,
			       superblock->dev_state->name,
			       (unsigned long long)superblock->dev_bytenr,
			       superblock->mirror_num,
			       (unsigned long long)
			       btrfs_super_generation(super_hdr),
			       (unsigned long long)
			       state->max_superblock_generation);
	} else {
		if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE)
			printk(KERN_INFO
			       "btrfsic: got new superblock @%llu (%s/%llu/%d)"
			       " with new gen %llu > %llu\n",
			       (unsigned long long)superblock->logical_bytenr,
			       superblock->dev_state->name,
			       (unsigned long long)superblock->dev_bytenr,
			       superblock->mirror_num,
			       (unsigned long long)
			       btrfs_super_generation(super_hdr),
			       (unsigned long long)
			       state->max_superblock_generation);

		state->max_superblock_generation =
		    btrfs_super_generation(super_hdr);
		state->latest_superblock = superblock;
	}

	for (pass = 0; pass < 3; pass++) {
		int ret;
		u64 next_bytenr;
		struct btrfsic_block *next_block;
		struct btrfsic_block_data_ctx tmp_next_block_ctx;
		struct btrfsic_block_link *l;
		int num_copies;
		int mirror_num;
		const char *additional_string = NULL;
		struct btrfs_disk_key tmp_disk_key;

		tmp_disk_key.type = BTRFS_ROOT_ITEM_KEY;
		tmp_disk_key.offset = 0;

		switch (pass) {
		case 0:
			tmp_disk_key.objectid =
			    cpu_to_le64(BTRFS_ROOT_TREE_OBJECTID);
			additional_string = "root ";
			next_bytenr = btrfs_super_root(super_hdr);
			if (state->print_mask &
			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
				printk(KERN_INFO "root@%llu\n",
				       (unsigned long long)next_bytenr);
			break;
		case 1:
			tmp_disk_key.objectid =
			    cpu_to_le64(BTRFS_CHUNK_TREE_OBJECTID);
			additional_string = "chunk ";
			next_bytenr = btrfs_super_chunk_root(super_hdr);
			if (state->print_mask &
			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
				printk(KERN_INFO "chunk@%llu\n",
				       (unsigned long long)next_bytenr);
			break;
		case 2:
			tmp_disk_key.objectid =
			    cpu_to_le64(BTRFS_TREE_LOG_OBJECTID);
			additional_string = "log ";
			next_bytenr = btrfs_super_log_root(super_hdr);
			if (0 == next_bytenr)
				continue;
			if (state->print_mask &
			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
				printk(KERN_INFO "log@%llu\n",
				       (unsigned long long)next_bytenr);
			break;
		}

		num_copies =
2476
		    btrfs_num_copies(state->root->fs_info,
2477
				     next_bytenr, BTRFS_SUPER_INFO_SIZE);
2478 2479 2480 2481 2482 2483 2484 2485 2486 2487
		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
			       (unsigned long long)next_bytenr, num_copies);
		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
			int was_created;

			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				printk(KERN_INFO
				       "btrfsic_process_written_superblock("
				       "mirror_num=%d)\n", mirror_num);
2488 2489
			ret = btrfsic_map_block(state, next_bytenr,
						BTRFS_SUPER_INFO_SIZE,
2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614
						&tmp_next_block_ctx,
						mirror_num);
			if (ret) {
				printk(KERN_INFO
				       "btrfsic: btrfsic_map_block(@%llu,"
				       " mirror=%d) failed!\n",
				       (unsigned long long)next_bytenr,
				       mirror_num);
				return -1;
			}

			next_block = btrfsic_block_lookup_or_add(
					state,
					&tmp_next_block_ctx,
					additional_string,
					1, 0, 1,
					mirror_num,
					&was_created);
			if (NULL == next_block) {
				printk(KERN_INFO
				       "btrfsic: error, kmalloc failed!\n");
				btrfsic_release_block_ctx(&tmp_next_block_ctx);
				return -1;
			}

			next_block->disk_key = tmp_disk_key;
			if (was_created)
				next_block->generation =
				    BTRFSIC_GENERATION_UNKNOWN;
			l = btrfsic_block_link_lookup_or_add(
					state,
					&tmp_next_block_ctx,
					next_block,
					superblock,
					BTRFSIC_GENERATION_UNKNOWN);
			btrfsic_release_block_ctx(&tmp_next_block_ctx);
			if (NULL == l)
				return -1;
		}
	}

	if (-1 == btrfsic_check_all_ref_blocks(state, superblock, 0)) {
		WARN_ON(1);
		btrfsic_dump_tree(state);
	}

	return 0;
}

static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state,
					struct btrfsic_block *const block,
					int recursion_level)
{
	struct list_head *elem_ref_to;
	int ret = 0;

	if (recursion_level >= 3 + BTRFS_MAX_LEVEL) {
		/*
		 * Note that this situation can happen and does not
		 * indicate an error in regular cases. It happens
		 * when disk blocks are freed and later reused.
		 * The check-integrity module is not aware of any
		 * block free operations, it just recognizes block
		 * write operations. Therefore it keeps the linkage
		 * information for a block until a block is
		 * rewritten. This can temporarily cause incorrect
		 * and even circular linkage informations. This
		 * causes no harm unless such blocks are referenced
		 * by the most recent super block.
		 */
		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "btrfsic: abort cyclic linkage (case 1).\n");

		return ret;
	}

	/*
	 * This algorithm is recursive because the amount of used stack
	 * space is very small and the max recursion depth is limited.
	 */
	list_for_each(elem_ref_to, &block->ref_to_list) {
		const struct btrfsic_block_link *const l =
		    list_entry(elem_ref_to, struct btrfsic_block_link,
			       node_ref_to);

		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "rl=%d, %c @%llu (%s/%llu/%d)"
			       " %u* refers to %c @%llu (%s/%llu/%d)\n",
			       recursion_level,
			       btrfsic_get_block_type(state, block),
			       (unsigned long long)block->logical_bytenr,
			       block->dev_state->name,
			       (unsigned long long)block->dev_bytenr,
			       block->mirror_num,
			       l->ref_cnt,
			       btrfsic_get_block_type(state, l->block_ref_to),
			       (unsigned long long)
			       l->block_ref_to->logical_bytenr,
			       l->block_ref_to->dev_state->name,
			       (unsigned long long)l->block_ref_to->dev_bytenr,
			       l->block_ref_to->mirror_num);
		if (l->block_ref_to->never_written) {
			printk(KERN_INFO "btrfs: attempt to write superblock"
			       " which references block %c @%llu (%s/%llu/%d)"
			       " which is never written!\n",
			       btrfsic_get_block_type(state, l->block_ref_to),
			       (unsigned long long)
			       l->block_ref_to->logical_bytenr,
			       l->block_ref_to->dev_state->name,
			       (unsigned long long)l->block_ref_to->dev_bytenr,
			       l->block_ref_to->mirror_num);
			ret = -1;
		} else if (!l->block_ref_to->is_iodone) {
			printk(KERN_INFO "btrfs: attempt to write superblock"
			       " which references block %c @%llu (%s/%llu/%d)"
			       " which is not yet iodone!\n",
			       btrfsic_get_block_type(state, l->block_ref_to),
			       (unsigned long long)
			       l->block_ref_to->logical_bytenr,
			       l->block_ref_to->dev_state->name,
			       (unsigned long long)l->block_ref_to->dev_bytenr,
			       l->block_ref_to->mirror_num);
			ret = -1;
2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625
		} else if (l->block_ref_to->iodone_w_error) {
			printk(KERN_INFO "btrfs: attempt to write superblock"
			       " which references block %c @%llu (%s/%llu/%d)"
			       " which has write error!\n",
			       btrfsic_get_block_type(state, l->block_ref_to),
			       (unsigned long long)
			       l->block_ref_to->logical_bytenr,
			       l->block_ref_to->dev_state->name,
			       (unsigned long long)l->block_ref_to->dev_bytenr,
			       l->block_ref_to->mirror_num);
			ret = -1;
2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964
		} else if (l->parent_generation !=
			   l->block_ref_to->generation &&
			   BTRFSIC_GENERATION_UNKNOWN !=
			   l->parent_generation &&
			   BTRFSIC_GENERATION_UNKNOWN !=
			   l->block_ref_to->generation) {
			printk(KERN_INFO "btrfs: attempt to write superblock"
			       " which references block %c @%llu (%s/%llu/%d)"
			       " with generation %llu !="
			       " parent generation %llu!\n",
			       btrfsic_get_block_type(state, l->block_ref_to),
			       (unsigned long long)
			       l->block_ref_to->logical_bytenr,
			       l->block_ref_to->dev_state->name,
			       (unsigned long long)l->block_ref_to->dev_bytenr,
			       l->block_ref_to->mirror_num,
			       (unsigned long long)l->block_ref_to->generation,
			       (unsigned long long)l->parent_generation);
			ret = -1;
		} else if (l->block_ref_to->flush_gen >
			   l->block_ref_to->dev_state->last_flush_gen) {
			printk(KERN_INFO "btrfs: attempt to write superblock"
			       " which references block %c @%llu (%s/%llu/%d)"
			       " which is not flushed out of disk's write cache"
			       " (block flush_gen=%llu,"
			       " dev->flush_gen=%llu)!\n",
			       btrfsic_get_block_type(state, l->block_ref_to),
			       (unsigned long long)
			       l->block_ref_to->logical_bytenr,
			       l->block_ref_to->dev_state->name,
			       (unsigned long long)l->block_ref_to->dev_bytenr,
			       l->block_ref_to->mirror_num,
			       (unsigned long long)block->flush_gen,
			       (unsigned long long)
			       l->block_ref_to->dev_state->last_flush_gen);
			ret = -1;
		} else if (-1 == btrfsic_check_all_ref_blocks(state,
							      l->block_ref_to,
							      recursion_level +
							      1)) {
			ret = -1;
		}
	}

	return ret;
}

static int btrfsic_is_block_ref_by_superblock(
		const struct btrfsic_state *state,
		const struct btrfsic_block *block,
		int recursion_level)
{
	struct list_head *elem_ref_from;

	if (recursion_level >= 3 + BTRFS_MAX_LEVEL) {
		/* refer to comment at "abort cyclic linkage (case 1)" */
		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "btrfsic: abort cyclic linkage (case 2).\n");

		return 0;
	}

	/*
	 * This algorithm is recursive because the amount of used stack space
	 * is very small and the max recursion depth is limited.
	 */
	list_for_each(elem_ref_from, &block->ref_from_list) {
		const struct btrfsic_block_link *const l =
		    list_entry(elem_ref_from, struct btrfsic_block_link,
			       node_ref_from);

		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "rl=%d, %c @%llu (%s/%llu/%d)"
			       " is ref %u* from %c @%llu (%s/%llu/%d)\n",
			       recursion_level,
			       btrfsic_get_block_type(state, block),
			       (unsigned long long)block->logical_bytenr,
			       block->dev_state->name,
			       (unsigned long long)block->dev_bytenr,
			       block->mirror_num,
			       l->ref_cnt,
			       btrfsic_get_block_type(state, l->block_ref_from),
			       (unsigned long long)
			       l->block_ref_from->logical_bytenr,
			       l->block_ref_from->dev_state->name,
			       (unsigned long long)
			       l->block_ref_from->dev_bytenr,
			       l->block_ref_from->mirror_num);
		if (l->block_ref_from->is_superblock &&
		    state->latest_superblock->dev_bytenr ==
		    l->block_ref_from->dev_bytenr &&
		    state->latest_superblock->dev_state->bdev ==
		    l->block_ref_from->dev_state->bdev)
			return 1;
		else if (btrfsic_is_block_ref_by_superblock(state,
							    l->block_ref_from,
							    recursion_level +
							    1))
			return 1;
	}

	return 0;
}

static void btrfsic_print_add_link(const struct btrfsic_state *state,
				   const struct btrfsic_block_link *l)
{
	printk(KERN_INFO
	       "Add %u* link from %c @%llu (%s/%llu/%d)"
	       " to %c @%llu (%s/%llu/%d).\n",
	       l->ref_cnt,
	       btrfsic_get_block_type(state, l->block_ref_from),
	       (unsigned long long)l->block_ref_from->logical_bytenr,
	       l->block_ref_from->dev_state->name,
	       (unsigned long long)l->block_ref_from->dev_bytenr,
	       l->block_ref_from->mirror_num,
	       btrfsic_get_block_type(state, l->block_ref_to),
	       (unsigned long long)l->block_ref_to->logical_bytenr,
	       l->block_ref_to->dev_state->name,
	       (unsigned long long)l->block_ref_to->dev_bytenr,
	       l->block_ref_to->mirror_num);
}

static void btrfsic_print_rem_link(const struct btrfsic_state *state,
				   const struct btrfsic_block_link *l)
{
	printk(KERN_INFO
	       "Rem %u* link from %c @%llu (%s/%llu/%d)"
	       " to %c @%llu (%s/%llu/%d).\n",
	       l->ref_cnt,
	       btrfsic_get_block_type(state, l->block_ref_from),
	       (unsigned long long)l->block_ref_from->logical_bytenr,
	       l->block_ref_from->dev_state->name,
	       (unsigned long long)l->block_ref_from->dev_bytenr,
	       l->block_ref_from->mirror_num,
	       btrfsic_get_block_type(state, l->block_ref_to),
	       (unsigned long long)l->block_ref_to->logical_bytenr,
	       l->block_ref_to->dev_state->name,
	       (unsigned long long)l->block_ref_to->dev_bytenr,
	       l->block_ref_to->mirror_num);
}

static char btrfsic_get_block_type(const struct btrfsic_state *state,
				   const struct btrfsic_block *block)
{
	if (block->is_superblock &&
	    state->latest_superblock->dev_bytenr == block->dev_bytenr &&
	    state->latest_superblock->dev_state->bdev == block->dev_state->bdev)
		return 'S';
	else if (block->is_superblock)
		return 's';
	else if (block->is_metadata)
		return 'M';
	else
		return 'D';
}

static void btrfsic_dump_tree(const struct btrfsic_state *state)
{
	btrfsic_dump_tree_sub(state, state->latest_superblock, 0);
}

static void btrfsic_dump_tree_sub(const struct btrfsic_state *state,
				  const struct btrfsic_block *block,
				  int indent_level)
{
	struct list_head *elem_ref_to;
	int indent_add;
	static char buf[80];
	int cursor_position;

	/*
	 * Should better fill an on-stack buffer with a complete line and
	 * dump it at once when it is time to print a newline character.
	 */

	/*
	 * This algorithm is recursive because the amount of used stack space
	 * is very small and the max recursion depth is limited.
	 */
	indent_add = sprintf(buf, "%c-%llu(%s/%llu/%d)",
			     btrfsic_get_block_type(state, block),
			     (unsigned long long)block->logical_bytenr,
			     block->dev_state->name,
			     (unsigned long long)block->dev_bytenr,
			     block->mirror_num);
	if (indent_level + indent_add > BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL) {
		printk("[...]\n");
		return;
	}
	printk(buf);
	indent_level += indent_add;
	if (list_empty(&block->ref_to_list)) {
		printk("\n");
		return;
	}
	if (block->mirror_num > 1 &&
	    !(state->print_mask & BTRFSIC_PRINT_MASK_TREE_WITH_ALL_MIRRORS)) {
		printk(" [...]\n");
		return;
	}

	cursor_position = indent_level;
	list_for_each(elem_ref_to, &block->ref_to_list) {
		const struct btrfsic_block_link *const l =
		    list_entry(elem_ref_to, struct btrfsic_block_link,
			       node_ref_to);

		while (cursor_position < indent_level) {
			printk(" ");
			cursor_position++;
		}
		if (l->ref_cnt > 1)
			indent_add = sprintf(buf, " %d*--> ", l->ref_cnt);
		else
			indent_add = sprintf(buf, " --> ");
		if (indent_level + indent_add >
		    BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL) {
			printk("[...]\n");
			cursor_position = 0;
			continue;
		}

		printk(buf);

		btrfsic_dump_tree_sub(state, l->block_ref_to,
				      indent_level + indent_add);
		cursor_position = 0;
	}
}

static struct btrfsic_block_link *btrfsic_block_link_lookup_or_add(
		struct btrfsic_state *state,
		struct btrfsic_block_data_ctx *next_block_ctx,
		struct btrfsic_block *next_block,
		struct btrfsic_block *from_block,
		u64 parent_generation)
{
	struct btrfsic_block_link *l;

	l = btrfsic_block_link_hashtable_lookup(next_block_ctx->dev->bdev,
						next_block_ctx->dev_bytenr,
						from_block->dev_state->bdev,
						from_block->dev_bytenr,
						&state->block_link_hashtable);
	if (NULL == l) {
		l = btrfsic_block_link_alloc();
		if (NULL == l) {
			printk(KERN_INFO
			       "btrfsic: error, kmalloc" " failed!\n");
			return NULL;
		}

		l->block_ref_to = next_block;
		l->block_ref_from = from_block;
		l->ref_cnt = 1;
		l->parent_generation = parent_generation;

		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			btrfsic_print_add_link(state, l);

		list_add(&l->node_ref_to, &from_block->ref_to_list);
		list_add(&l->node_ref_from, &next_block->ref_from_list);

		btrfsic_block_link_hashtable_add(l,
						 &state->block_link_hashtable);
	} else {
		l->ref_cnt++;
		l->parent_generation = parent_generation;
		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			btrfsic_print_add_link(state, l);
	}

	return l;
}

static struct btrfsic_block *btrfsic_block_lookup_or_add(
		struct btrfsic_state *state,
		struct btrfsic_block_data_ctx *block_ctx,
		const char *additional_string,
		int is_metadata,
		int is_iodone,
		int never_written,
		int mirror_num,
		int *was_created)
{
	struct btrfsic_block *block;

	block = btrfsic_block_hashtable_lookup(block_ctx->dev->bdev,
					       block_ctx->dev_bytenr,
					       &state->block_hashtable);
	if (NULL == block) {
		struct btrfsic_dev_state *dev_state;

		block = btrfsic_block_alloc();
		if (NULL == block) {
			printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
			return NULL;
		}
		dev_state = btrfsic_dev_state_lookup(block_ctx->dev->bdev);
		if (NULL == dev_state) {
			printk(KERN_INFO
			       "btrfsic: error, lookup dev_state failed!\n");
			btrfsic_block_free(block);
			return NULL;
		}
		block->dev_state = dev_state;
		block->dev_bytenr = block_ctx->dev_bytenr;
		block->logical_bytenr = block_ctx->start;
		block->is_metadata = is_metadata;
		block->is_iodone = is_iodone;
		block->never_written = never_written;
		block->mirror_num = mirror_num;
		if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
			printk(KERN_INFO
			       "New %s%c-block @%llu (%s/%llu/%d)\n",
			       additional_string,
			       btrfsic_get_block_type(state, block),
			       (unsigned long long)block->logical_bytenr,
			       dev_state->name,
			       (unsigned long long)block->dev_bytenr,
			       mirror_num);
		list_add(&block->all_blocks_node, &state->all_blocks_list);
		btrfsic_block_hashtable_add(block, &state->block_hashtable);
		if (NULL != was_created)
			*was_created = 1;
	} else {
		if (NULL != was_created)
			*was_created = 0;
	}

	return block;
}

static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
					   u64 bytenr,
					   struct btrfsic_dev_state *dev_state,
2965
					   u64 dev_bytenr)
2966 2967 2968 2969 2970 2971 2972
{
	int num_copies;
	int mirror_num;
	int ret;
	struct btrfsic_block_data_ctx block_ctx;
	int match = 0;

2973
	num_copies = btrfs_num_copies(state->root->fs_info,
2974
				      bytenr, state->metablock_size);
2975 2976

	for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
2977
		ret = btrfsic_map_block(state, bytenr, state->metablock_size,
2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002
					&block_ctx, mirror_num);
		if (ret) {
			printk(KERN_INFO "btrfsic:"
			       " btrfsic_map_block(logical @%llu,"
			       " mirror %d) failed!\n",
			       (unsigned long long)bytenr, mirror_num);
			continue;
		}

		if (dev_state->bdev == block_ctx.dev->bdev &&
		    dev_bytenr == block_ctx.dev_bytenr) {
			match++;
			btrfsic_release_block_ctx(&block_ctx);
			break;
		}
		btrfsic_release_block_ctx(&block_ctx);
	}

	if (!match) {
		printk(KERN_INFO "btrfs: attempt to write M-block which contains logical bytenr that doesn't map to dev+physical bytenr of submit_bio,"
		       " buffer->log_bytenr=%llu, submit_bio(bdev=%s,"
		       " phys_bytenr=%llu)!\n",
		       (unsigned long long)bytenr, dev_state->name,
		       (unsigned long long)dev_bytenr);
		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
3003 3004
			ret = btrfsic_map_block(state, bytenr,
						state->metablock_size,
3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052
						&block_ctx, mirror_num);
			if (ret)
				continue;

			printk(KERN_INFO "Read logical bytenr @%llu maps to"
			       " (%s/%llu/%d)\n",
			       (unsigned long long)bytenr,
			       block_ctx.dev->name,
			       (unsigned long long)block_ctx.dev_bytenr,
			       mirror_num);
		}
		WARN_ON(1);
	}
}

static struct btrfsic_dev_state *btrfsic_dev_state_lookup(
		struct block_device *bdev)
{
	struct btrfsic_dev_state *ds;

	ds = btrfsic_dev_state_hashtable_lookup(bdev,
						&btrfsic_dev_state_hashtable);
	return ds;
}

int btrfsic_submit_bh(int rw, struct buffer_head *bh)
{
	struct btrfsic_dev_state *dev_state;

	if (!btrfsic_is_initialized)
		return submit_bh(rw, bh);

	mutex_lock(&btrfsic_mutex);
	/* since btrfsic_submit_bh() might also be called before
	 * btrfsic_mount(), this might return NULL */
	dev_state = btrfsic_dev_state_lookup(bh->b_bdev);

	/* Only called to write the superblock (incl. FLUSH/FUA) */
	if (NULL != dev_state &&
	    (rw & WRITE) && bh->b_size > 0) {
		u64 dev_bytenr;

		dev_bytenr = 4096 * bh->b_blocknr;
		if (dev_state->state->print_mask &
		    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
			printk(KERN_INFO
			       "submit_bh(rw=0x%x, blocknr=%lu (bytenr %llu),"
			       " size=%lu, data=%p, bdev=%p)\n",
3053 3054 3055 3056
			       rw, (unsigned long)bh->b_blocknr,
			       (unsigned long long)dev_bytenr,
			       (unsigned long)bh->b_size, bh->b_data,
			       bh->b_bdev);
3057
		btrfsic_process_written_block(dev_state, dev_bytenr,
3058
					      &bh->b_data, 1, NULL,
3059 3060 3061 3062 3063
					      NULL, bh, rw);
	} else if (NULL != dev_state && (rw & REQ_FLUSH)) {
		if (dev_state->state->print_mask &
		    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
			printk(KERN_INFO
3064
			       "submit_bh(rw=0x%x FLUSH, bdev=%p)\n",
3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112
			       rw, bh->b_bdev);
		if (!dev_state->dummy_block_for_bio_bh_flush.is_iodone) {
			if ((dev_state->state->print_mask &
			     (BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH |
			      BTRFSIC_PRINT_MASK_VERBOSE)))
				printk(KERN_INFO
				       "btrfsic_submit_bh(%s) with FLUSH"
				       " but dummy block already in use"
				       " (ignored)!\n",
				       dev_state->name);
		} else {
			struct btrfsic_block *const block =
				&dev_state->dummy_block_for_bio_bh_flush;

			block->is_iodone = 0;
			block->never_written = 0;
			block->iodone_w_error = 0;
			block->flush_gen = dev_state->last_flush_gen + 1;
			block->submit_bio_bh_rw = rw;
			block->orig_bio_bh_private = bh->b_private;
			block->orig_bio_bh_end_io.bh = bh->b_end_io;
			block->next_in_same_bio = NULL;
			bh->b_private = block;
			bh->b_end_io = btrfsic_bh_end_io;
		}
	}
	mutex_unlock(&btrfsic_mutex);
	return submit_bh(rw, bh);
}

void btrfsic_submit_bio(int rw, struct bio *bio)
{
	struct btrfsic_dev_state *dev_state;

	if (!btrfsic_is_initialized) {
		submit_bio(rw, bio);
		return;
	}

	mutex_lock(&btrfsic_mutex);
	/* since btrfsic_submit_bio() is also called before
	 * btrfsic_mount(), this might return NULL */
	dev_state = btrfsic_dev_state_lookup(bio->bi_bdev);
	if (NULL != dev_state &&
	    (rw & WRITE) && NULL != bio->bi_io_vec) {
		unsigned int i;
		u64 dev_bytenr;
		int bio_is_patched;
3113
		char **mapped_datav;
3114 3115 3116 3117 3118 3119 3120 3121

		dev_bytenr = 512 * bio->bi_sector;
		bio_is_patched = 0;
		if (dev_state->state->print_mask &
		    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
			printk(KERN_INFO
			       "submit_bio(rw=0x%x, bi_vcnt=%u,"
			       " bi_sector=%lu (bytenr %llu), bi_bdev=%p)\n",
3122
			       rw, bio->bi_vcnt, (unsigned long)bio->bi_sector,
3123 3124 3125
			       (unsigned long long)dev_bytenr,
			       bio->bi_bdev);

3126 3127 3128 3129
		mapped_datav = kmalloc(sizeof(*mapped_datav) * bio->bi_vcnt,
				       GFP_NOFS);
		if (!mapped_datav)
			goto leave;
3130
		for (i = 0; i < bio->bi_vcnt; i++) {
3131 3132 3133 3134 3135 3136 3137 3138 3139 3140
			BUG_ON(bio->bi_io_vec[i].bv_len != PAGE_CACHE_SIZE);
			mapped_datav[i] = kmap(bio->bi_io_vec[i].bv_page);
			if (!mapped_datav[i]) {
				while (i > 0) {
					i--;
					kunmap(bio->bi_io_vec[i].bv_page);
				}
				kfree(mapped_datav);
				goto leave;
			}
3141 3142 3143 3144 3145 3146
			if ((BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH |
			     BTRFSIC_PRINT_MASK_VERBOSE) ==
			    (dev_state->state->print_mask &
			     (BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH |
			      BTRFSIC_PRINT_MASK_VERBOSE)))
				printk(KERN_INFO
3147
				       "#%u: page=%p, len=%u, offset=%u\n",
3148 3149 3150
				       i, bio->bi_io_vec[i].bv_page,
				       bio->bi_io_vec[i].bv_len,
				       bio->bi_io_vec[i].bv_offset);
3151 3152 3153 3154 3155 3156 3157
		}
		btrfsic_process_written_block(dev_state, dev_bytenr,
					      mapped_datav, bio->bi_vcnt,
					      bio, &bio_is_patched,
					      NULL, rw);
		while (i > 0) {
			i--;
3158 3159
			kunmap(bio->bi_io_vec[i].bv_page);
		}
3160
		kfree(mapped_datav);
3161 3162 3163 3164
	} else if (NULL != dev_state && (rw & REQ_FLUSH)) {
		if (dev_state->state->print_mask &
		    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
			printk(KERN_INFO
3165
			       "submit_bio(rw=0x%x FLUSH, bdev=%p)\n",
3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191
			       rw, bio->bi_bdev);
		if (!dev_state->dummy_block_for_bio_bh_flush.is_iodone) {
			if ((dev_state->state->print_mask &
			     (BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH |
			      BTRFSIC_PRINT_MASK_VERBOSE)))
				printk(KERN_INFO
				       "btrfsic_submit_bio(%s) with FLUSH"
				       " but dummy block already in use"
				       " (ignored)!\n",
				       dev_state->name);
		} else {
			struct btrfsic_block *const block =
				&dev_state->dummy_block_for_bio_bh_flush;

			block->is_iodone = 0;
			block->never_written = 0;
			block->iodone_w_error = 0;
			block->flush_gen = dev_state->last_flush_gen + 1;
			block->submit_bio_bh_rw = rw;
			block->orig_bio_bh_private = bio->bi_private;
			block->orig_bio_bh_end_io.bio = bio->bi_end_io;
			block->next_in_same_bio = NULL;
			bio->bi_private = block;
			bio->bi_end_io = btrfsic_bio_end_io;
		}
	}
3192
leave:
3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206
	mutex_unlock(&btrfsic_mutex);

	submit_bio(rw, bio);
}

int btrfsic_mount(struct btrfs_root *root,
		  struct btrfs_fs_devices *fs_devices,
		  int including_extent_data, u32 print_mask)
{
	int ret;
	struct btrfsic_state *state;
	struct list_head *dev_head = &fs_devices->devices;
	struct btrfs_device *device;

3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230
	if (root->nodesize != root->leafsize) {
		printk(KERN_INFO
		       "btrfsic: cannot handle nodesize %d != leafsize %d!\n",
		       root->nodesize, root->leafsize);
		return -1;
	}
	if (root->nodesize & ((u64)PAGE_CACHE_SIZE - 1)) {
		printk(KERN_INFO
		       "btrfsic: cannot handle nodesize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
		       root->nodesize, (unsigned long)PAGE_CACHE_SIZE);
		return -1;
	}
	if (root->leafsize & ((u64)PAGE_CACHE_SIZE - 1)) {
		printk(KERN_INFO
		       "btrfsic: cannot handle leafsize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
		       root->leafsize, (unsigned long)PAGE_CACHE_SIZE);
		return -1;
	}
	if (root->sectorsize & ((u64)PAGE_CACHE_SIZE - 1)) {
		printk(KERN_INFO
		       "btrfsic: cannot handle sectorsize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
		       root->sectorsize, (unsigned long)PAGE_CACHE_SIZE);
		return -1;
	}
3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246
	state = kzalloc(sizeof(*state), GFP_NOFS);
	if (NULL == state) {
		printk(KERN_INFO "btrfs check-integrity: kmalloc() failed!\n");
		return -1;
	}

	if (!btrfsic_is_initialized) {
		mutex_init(&btrfsic_mutex);
		btrfsic_dev_state_hashtable_init(&btrfsic_dev_state_hashtable);
		btrfsic_is_initialized = 1;
	}
	mutex_lock(&btrfsic_mutex);
	state->root = root;
	state->print_mask = print_mask;
	state->include_extent_data = including_extent_data;
	state->csum_size = 0;
3247 3248
	state->metablock_size = root->nodesize;
	state->datablock_size = root->sectorsize;
3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 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 3362 3363 3364
	INIT_LIST_HEAD(&state->all_blocks_list);
	btrfsic_block_hashtable_init(&state->block_hashtable);
	btrfsic_block_link_hashtable_init(&state->block_link_hashtable);
	state->max_superblock_generation = 0;
	state->latest_superblock = NULL;

	list_for_each_entry(device, dev_head, dev_list) {
		struct btrfsic_dev_state *ds;
		char *p;

		if (!device->bdev || !device->name)
			continue;

		ds = btrfsic_dev_state_alloc();
		if (NULL == ds) {
			printk(KERN_INFO
			       "btrfs check-integrity: kmalloc() failed!\n");
			mutex_unlock(&btrfsic_mutex);
			return -1;
		}
		ds->bdev = device->bdev;
		ds->state = state;
		bdevname(ds->bdev, ds->name);
		ds->name[BDEVNAME_SIZE - 1] = '\0';
		for (p = ds->name; *p != '\0'; p++);
		while (p > ds->name && *p != '/')
			p--;
		if (*p == '/')
			p++;
		strlcpy(ds->name, p, sizeof(ds->name));
		btrfsic_dev_state_hashtable_add(ds,
						&btrfsic_dev_state_hashtable);
	}

	ret = btrfsic_process_superblock(state, fs_devices);
	if (0 != ret) {
		mutex_unlock(&btrfsic_mutex);
		btrfsic_unmount(root, fs_devices);
		return ret;
	}

	if (state->print_mask & BTRFSIC_PRINT_MASK_INITIAL_DATABASE)
		btrfsic_dump_database(state);
	if (state->print_mask & BTRFSIC_PRINT_MASK_INITIAL_TREE)
		btrfsic_dump_tree(state);

	mutex_unlock(&btrfsic_mutex);
	return 0;
}

void btrfsic_unmount(struct btrfs_root *root,
		     struct btrfs_fs_devices *fs_devices)
{
	struct list_head *elem_all;
	struct list_head *tmp_all;
	struct btrfsic_state *state;
	struct list_head *dev_head = &fs_devices->devices;
	struct btrfs_device *device;

	if (!btrfsic_is_initialized)
		return;

	mutex_lock(&btrfsic_mutex);

	state = NULL;
	list_for_each_entry(device, dev_head, dev_list) {
		struct btrfsic_dev_state *ds;

		if (!device->bdev || !device->name)
			continue;

		ds = btrfsic_dev_state_hashtable_lookup(
				device->bdev,
				&btrfsic_dev_state_hashtable);
		if (NULL != ds) {
			state = ds->state;
			btrfsic_dev_state_hashtable_remove(ds);
			btrfsic_dev_state_free(ds);
		}
	}

	if (NULL == state) {
		printk(KERN_INFO
		       "btrfsic: error, cannot find state information"
		       " on umount!\n");
		mutex_unlock(&btrfsic_mutex);
		return;
	}

	/*
	 * Don't care about keeping the lists' state up to date,
	 * just free all memory that was allocated dynamically.
	 * Free the blocks and the block_links.
	 */
	list_for_each_safe(elem_all, tmp_all, &state->all_blocks_list) {
		struct btrfsic_block *const b_all =
		    list_entry(elem_all, struct btrfsic_block,
			       all_blocks_node);
		struct list_head *elem_ref_to;
		struct list_head *tmp_ref_to;

		list_for_each_safe(elem_ref_to, tmp_ref_to,
				   &b_all->ref_to_list) {
			struct btrfsic_block_link *const l =
			    list_entry(elem_ref_to,
				       struct btrfsic_block_link,
				       node_ref_to);

			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
				btrfsic_print_rem_link(state, l);

			l->ref_cnt--;
			if (0 == l->ref_cnt)
				btrfsic_block_link_free(l);
		}

3365
		if (b_all->is_iodone || b_all->never_written)
3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381
			btrfsic_block_free(b_all);
		else
			printk(KERN_INFO "btrfs: attempt to free %c-block"
			       " @%llu (%s/%llu/%d) on umount which is"
			       " not yet iodone!\n",
			       btrfsic_get_block_type(state, b_all),
			       (unsigned long long)b_all->logical_bytenr,
			       b_all->dev_state->name,
			       (unsigned long long)b_all->dev_bytenr,
			       b_all->mirror_num);
	}

	mutex_unlock(&btrfsic_mutex);

	kfree(state);
}