scrub.c 91.0 KB
Newer Older
A
Arne Jansen 已提交
1
/*
2
 * Copyright (C) 2011, 2012 STRATO.  All rights reserved.
A
Arne Jansen 已提交
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
 *
 * 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.
 */

#include <linux/blkdev.h>
20
#include <linux/ratelimit.h>
A
Arne Jansen 已提交
21 22 23 24
#include "ctree.h"
#include "volumes.h"
#include "disk-io.h"
#include "ordered-data.h"
25
#include "transaction.h"
26
#include "backref.h"
27
#include "extent_io.h"
28
#include "dev-replace.h"
29
#include "check-integrity.h"
30
#include "rcu-string.h"
D
David Woodhouse 已提交
31
#include "raid56.h"
A
Arne Jansen 已提交
32 33 34 35 36 37 38 39 40 41 42 43 44 45

/*
 * This is only the first step towards a full-features scrub. It reads all
 * extent and super block and verifies the checksums. In case a bad checksum
 * is found or the extent cannot be read, good data will be written back if
 * any can be found.
 *
 * Future enhancements:
 *  - In case an unrepairable extent is encountered, track which files are
 *    affected and report them
 *  - track and record media errors, throw out bad devices
 *  - add a mode to also read unallocated space
 */

46
struct scrub_block;
47
struct scrub_ctx;
A
Arne Jansen 已提交
48

49 50 51 52 53 54 55 56 57
/*
 * the following three values only influence the performance.
 * The last one configures the number of parallel and outstanding I/O
 * operations. The first two values configure an upper limit for the number
 * of (dynamically allocated) pages that are added to a bio.
 */
#define SCRUB_PAGES_PER_RD_BIO	32	/* 128k per bio */
#define SCRUB_PAGES_PER_WR_BIO	32	/* 128k per bio */
#define SCRUB_BIOS_PER_SCTX	64	/* 8MB per device in flight */
58 59 60 61 62 63

/*
 * the following value times PAGE_SIZE needs to be large enough to match the
 * largest node/leaf/sector size that shall be supported.
 * Values larger than BTRFS_STRIPE_LEN are not supported.
 */
64
#define SCRUB_MAX_PAGES_PER_BLOCK	16	/* 64k per node/leaf/sector */
A
Arne Jansen 已提交
65 66

struct scrub_page {
67 68
	struct scrub_block	*sblock;
	struct page		*page;
69
	struct btrfs_device	*dev;
A
Arne Jansen 已提交
70 71
	u64			flags;  /* extent flags */
	u64			generation;
72 73
	u64			logical;
	u64			physical;
74
	u64			physical_for_dev_replace;
75
	atomic_t		ref_count;
76 77 78 79 80
	struct {
		unsigned int	mirror_num:8;
		unsigned int	have_csum:1;
		unsigned int	io_error:1;
	};
A
Arne Jansen 已提交
81 82 83 84 85
	u8			csum[BTRFS_CSUM_SIZE];
};

struct scrub_bio {
	int			index;
86
	struct scrub_ctx	*sctx;
87
	struct btrfs_device	*dev;
A
Arne Jansen 已提交
88 89 90 91
	struct bio		*bio;
	int			err;
	u64			logical;
	u64			physical;
92 93 94 95 96
#if SCRUB_PAGES_PER_WR_BIO >= SCRUB_PAGES_PER_RD_BIO
	struct scrub_page	*pagev[SCRUB_PAGES_PER_WR_BIO];
#else
	struct scrub_page	*pagev[SCRUB_PAGES_PER_RD_BIO];
#endif
97
	int			page_count;
A
Arne Jansen 已提交
98
	int			next_free;
99
	struct btrfs_work	work;
A
Arne Jansen 已提交
100 101
};

102
struct scrub_block {
103
	struct scrub_page	*pagev[SCRUB_MAX_PAGES_PER_BLOCK];
104 105 106
	int			page_count;
	atomic_t		outstanding_pages;
	atomic_t		ref_count; /* free mem on transition to zero */
107
	struct scrub_ctx	*sctx;
108 109 110 111
	struct {
		unsigned int	header_error:1;
		unsigned int	checksum_error:1;
		unsigned int	no_io_error_seen:1;
112
		unsigned int	generation_error:1; /* also sets header_error */
113 114 115
	};
};

116 117 118 119 120 121 122 123
struct scrub_wr_ctx {
	struct scrub_bio *wr_curr_bio;
	struct btrfs_device *tgtdev;
	int pages_per_wr_bio; /* <= SCRUB_PAGES_PER_WR_BIO */
	atomic_t flush_all_writes;
	struct mutex wr_lock;
};

124
struct scrub_ctx {
125
	struct scrub_bio	*bios[SCRUB_BIOS_PER_SCTX];
126
	struct btrfs_root	*dev_root;
A
Arne Jansen 已提交
127 128
	int			first_free;
	int			curr;
129 130
	atomic_t		bios_in_flight;
	atomic_t		workers_pending;
A
Arne Jansen 已提交
131 132 133 134 135
	spinlock_t		list_lock;
	wait_queue_head_t	list_wait;
	u16			csum_size;
	struct list_head	csum_list;
	atomic_t		cancel_req;
A
Arne Jansen 已提交
136
	int			readonly;
137
	int			pages_per_rd_bio;
138 139 140
	u32			sectorsize;
	u32			nodesize;
	u32			leafsize;
141 142

	int			is_dev_replace;
143
	struct scrub_wr_ctx	wr_ctx;
144

A
Arne Jansen 已提交
145 146 147 148 149 150 151
	/*
	 * statistics
	 */
	struct btrfs_scrub_progress stat;
	spinlock_t		stat_lock;
};

152
struct scrub_fixup_nodatasum {
153
	struct scrub_ctx	*sctx;
154
	struct btrfs_device	*dev;
155 156
	u64			logical;
	struct btrfs_root	*root;
157
	struct btrfs_work	work;
158 159 160
	int			mirror_num;
};

161 162 163 164 165 166 167
struct scrub_nocow_inode {
	u64			inum;
	u64			offset;
	u64			root;
	struct list_head	list;
};

168 169 170 171 172 173
struct scrub_copy_nocow_ctx {
	struct scrub_ctx	*sctx;
	u64			logical;
	u64			len;
	int			mirror_num;
	u64			physical_for_dev_replace;
174
	struct list_head	inodes;
175
	struct btrfs_work	work;
176 177
};

178 179 180 181 182 183 184 185 186 187 188 189 190
struct scrub_warning {
	struct btrfs_path	*path;
	u64			extent_item_size;
	char			*scratch_buf;
	char			*msg_buf;
	const char		*errstr;
	sector_t		sector;
	u64			logical;
	struct btrfs_device	*dev;
	int			msg_bufsize;
	int			scratch_bufsize;
};

191

192 193 194 195
static void scrub_pending_bio_inc(struct scrub_ctx *sctx);
static void scrub_pending_bio_dec(struct scrub_ctx *sctx);
static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx);
static void scrub_pending_trans_workers_dec(struct scrub_ctx *sctx);
196
static int scrub_handle_errored_block(struct scrub_block *sblock_to_check);
197
static int scrub_setup_recheck_block(struct scrub_ctx *sctx,
198
				     struct btrfs_fs_info *fs_info,
199
				     struct scrub_block *original_sblock,
200
				     u64 length, u64 logical,
201
				     struct scrub_block *sblocks_for_recheck);
202 203 204 205
static void scrub_recheck_block(struct btrfs_fs_info *fs_info,
				struct scrub_block *sblock, int is_metadata,
				int have_csum, u8 *csum, u64 generation,
				u16 csum_size);
206 207 208 209 210 211 212 213 214 215 216
static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
					 struct scrub_block *sblock,
					 int is_metadata, int have_csum,
					 const u8 *csum, u64 generation,
					 u16 csum_size);
static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad,
					     struct scrub_block *sblock_good,
					     int force_write);
static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad,
					    struct scrub_block *sblock_good,
					    int page_num, int force_write);
217 218 219
static void scrub_write_block_to_dev_replace(struct scrub_block *sblock);
static int scrub_write_page_to_dev_replace(struct scrub_block *sblock,
					   int page_num);
220 221 222 223 224
static int scrub_checksum_data(struct scrub_block *sblock);
static int scrub_checksum_tree_block(struct scrub_block *sblock);
static int scrub_checksum_super(struct scrub_block *sblock);
static void scrub_block_get(struct scrub_block *sblock);
static void scrub_block_put(struct scrub_block *sblock);
225 226
static void scrub_page_get(struct scrub_page *spage);
static void scrub_page_put(struct scrub_page *spage);
227 228
static int scrub_add_page_to_rd_bio(struct scrub_ctx *sctx,
				    struct scrub_page *spage);
229
static int scrub_pages(struct scrub_ctx *sctx, u64 logical, u64 len,
230
		       u64 physical, struct btrfs_device *dev, u64 flags,
231 232
		       u64 gen, int mirror_num, u8 *csum, int force,
		       u64 physical_for_dev_replace);
S
Stefan Behrens 已提交
233
static void scrub_bio_end_io(struct bio *bio, int err);
234
static void scrub_bio_end_io_worker(struct btrfs_work *work);
235
static void scrub_block_complete(struct scrub_block *sblock);
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250
static void scrub_remap_extent(struct btrfs_fs_info *fs_info,
			       u64 extent_logical, u64 extent_len,
			       u64 *extent_physical,
			       struct btrfs_device **extent_dev,
			       int *extent_mirror_num);
static int scrub_setup_wr_ctx(struct scrub_ctx *sctx,
			      struct scrub_wr_ctx *wr_ctx,
			      struct btrfs_fs_info *fs_info,
			      struct btrfs_device *dev,
			      int is_dev_replace);
static void scrub_free_wr_ctx(struct scrub_wr_ctx *wr_ctx);
static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx,
				    struct scrub_page *spage);
static void scrub_wr_submit(struct scrub_ctx *sctx);
static void scrub_wr_bio_end_io(struct bio *bio, int err);
251
static void scrub_wr_bio_end_io_worker(struct btrfs_work *work);
252 253 254
static int write_page_nocow(struct scrub_ctx *sctx,
			    u64 physical_for_dev_replace, struct page *page);
static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root,
255
				      struct scrub_copy_nocow_ctx *ctx);
256 257
static int copy_nocow_pages(struct scrub_ctx *sctx, u64 logical, u64 len,
			    int mirror_num, u64 physical_for_dev_replace);
258
static void copy_nocow_pages_worker(struct btrfs_work *work);
259
static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info);
260
static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info);
S
Stefan Behrens 已提交
261 262


263 264 265 266 267 268 269 270 271 272 273
static void scrub_pending_bio_inc(struct scrub_ctx *sctx)
{
	atomic_inc(&sctx->bios_in_flight);
}

static void scrub_pending_bio_dec(struct scrub_ctx *sctx)
{
	atomic_dec(&sctx->bios_in_flight);
	wake_up(&sctx->list_wait);
}

274
static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info)
275 276 277 278 279 280 281 282 283
{
	while (atomic_read(&fs_info->scrub_pause_req)) {
		mutex_unlock(&fs_info->scrub_lock);
		wait_event(fs_info->scrub_pause_wait,
		   atomic_read(&fs_info->scrub_pause_req) == 0);
		mutex_lock(&fs_info->scrub_lock);
	}
}

284 285 286 287 288 289 290 291 292 293 294 295 296
static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info)
{
	atomic_inc(&fs_info->scrubs_paused);
	wake_up(&fs_info->scrub_pause_wait);

	mutex_lock(&fs_info->scrub_lock);
	__scrub_blocked_if_needed(fs_info);
	atomic_dec(&fs_info->scrubs_paused);
	mutex_unlock(&fs_info->scrub_lock);

	wake_up(&fs_info->scrub_pause_wait);
}

297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317
/*
 * used for workers that require transaction commits (i.e., for the
 * NOCOW case)
 */
static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx)
{
	struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info;

	/*
	 * increment scrubs_running to prevent cancel requests from
	 * completing as long as a worker is running. we must also
	 * increment scrubs_paused to prevent deadlocking on pause
	 * requests used for transactions commits (as the worker uses a
	 * transaction context). it is safe to regard the worker
	 * as paused for all matters practical. effectively, we only
	 * avoid cancellation requests from completing.
	 */
	mutex_lock(&fs_info->scrub_lock);
	atomic_inc(&fs_info->scrubs_running);
	atomic_inc(&fs_info->scrubs_paused);
	mutex_unlock(&fs_info->scrub_lock);
318 319 320 321 322 323 324 325 326 327

	/*
	 * check if @scrubs_running=@scrubs_paused condition
	 * inside wait_event() is not an atomic operation.
	 * which means we may inc/dec @scrub_running/paused
	 * at any time. Let's wake up @scrub_pause_wait as
	 * much as we can to let commit transaction blocked less.
	 */
	wake_up(&fs_info->scrub_pause_wait);

328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348
	atomic_inc(&sctx->workers_pending);
}

/* used for workers that require transaction commits */
static void scrub_pending_trans_workers_dec(struct scrub_ctx *sctx)
{
	struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info;

	/*
	 * see scrub_pending_trans_workers_inc() why we're pretending
	 * to be paused in the scrub counters
	 */
	mutex_lock(&fs_info->scrub_lock);
	atomic_dec(&fs_info->scrubs_running);
	atomic_dec(&fs_info->scrubs_paused);
	mutex_unlock(&fs_info->scrub_lock);
	atomic_dec(&sctx->workers_pending);
	wake_up(&fs_info->scrub_pause_wait);
	wake_up(&sctx->list_wait);
}

349
static void scrub_free_csums(struct scrub_ctx *sctx)
A
Arne Jansen 已提交
350
{
351
	while (!list_empty(&sctx->csum_list)) {
A
Arne Jansen 已提交
352
		struct btrfs_ordered_sum *sum;
353
		sum = list_first_entry(&sctx->csum_list,
A
Arne Jansen 已提交
354 355 356 357 358 359
				       struct btrfs_ordered_sum, list);
		list_del(&sum->list);
		kfree(sum);
	}
}

360
static noinline_for_stack void scrub_free_ctx(struct scrub_ctx *sctx)
A
Arne Jansen 已提交
361 362 363
{
	int i;

364
	if (!sctx)
A
Arne Jansen 已提交
365 366
		return;

367 368
	scrub_free_wr_ctx(&sctx->wr_ctx);

369
	/* this can happen when scrub is cancelled */
370 371
	if (sctx->curr != -1) {
		struct scrub_bio *sbio = sctx->bios[sctx->curr];
372 373

		for (i = 0; i < sbio->page_count; i++) {
374
			WARN_ON(!sbio->pagev[i]->page);
375 376 377 378 379
			scrub_block_put(sbio->pagev[i]->sblock);
		}
		bio_put(sbio->bio);
	}

380
	for (i = 0; i < SCRUB_BIOS_PER_SCTX; ++i) {
381
		struct scrub_bio *sbio = sctx->bios[i];
A
Arne Jansen 已提交
382 383 384 385 386 387

		if (!sbio)
			break;
		kfree(sbio);
	}

388 389
	scrub_free_csums(sctx);
	kfree(sctx);
A
Arne Jansen 已提交
390 391 392
}

static noinline_for_stack
393
struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace)
A
Arne Jansen 已提交
394
{
395
	struct scrub_ctx *sctx;
A
Arne Jansen 已提交
396 397
	int		i;
	struct btrfs_fs_info *fs_info = dev->dev_root->fs_info;
398 399
	int pages_per_rd_bio;
	int ret;
A
Arne Jansen 已提交
400

401 402 403 404 405 406 407 408 409 410 411 412
	/*
	 * the setting of pages_per_rd_bio is correct for scrub but might
	 * be wrong for the dev_replace code where we might read from
	 * different devices in the initial huge bios. However, that
	 * code is able to correctly handle the case when adding a page
	 * to a bio fails.
	 */
	if (dev->bdev)
		pages_per_rd_bio = min_t(int, SCRUB_PAGES_PER_RD_BIO,
					 bio_get_nr_vecs(dev->bdev));
	else
		pages_per_rd_bio = SCRUB_PAGES_PER_RD_BIO;
413 414
	sctx = kzalloc(sizeof(*sctx), GFP_NOFS);
	if (!sctx)
A
Arne Jansen 已提交
415
		goto nomem;
416
	sctx->is_dev_replace = is_dev_replace;
417
	sctx->pages_per_rd_bio = pages_per_rd_bio;
418
	sctx->curr = -1;
419
	sctx->dev_root = dev->dev_root;
420
	for (i = 0; i < SCRUB_BIOS_PER_SCTX; ++i) {
A
Arne Jansen 已提交
421 422 423 424 425
		struct scrub_bio *sbio;

		sbio = kzalloc(sizeof(*sbio), GFP_NOFS);
		if (!sbio)
			goto nomem;
426
		sctx->bios[i] = sbio;
A
Arne Jansen 已提交
427 428

		sbio->index = i;
429
		sbio->sctx = sctx;
430
		sbio->page_count = 0;
431 432
		btrfs_init_work(&sbio->work, scrub_bio_end_io_worker,
				NULL, NULL);
A
Arne Jansen 已提交
433

434
		if (i != SCRUB_BIOS_PER_SCTX - 1)
435
			sctx->bios[i]->next_free = i + 1;
436
		else
437 438 439 440 441 442
			sctx->bios[i]->next_free = -1;
	}
	sctx->first_free = 0;
	sctx->nodesize = dev->dev_root->nodesize;
	sctx->leafsize = dev->dev_root->leafsize;
	sctx->sectorsize = dev->dev_root->sectorsize;
443 444
	atomic_set(&sctx->bios_in_flight, 0);
	atomic_set(&sctx->workers_pending, 0);
445 446 447 448 449 450 451
	atomic_set(&sctx->cancel_req, 0);
	sctx->csum_size = btrfs_super_csum_size(fs_info->super_copy);
	INIT_LIST_HEAD(&sctx->csum_list);

	spin_lock_init(&sctx->list_lock);
	spin_lock_init(&sctx->stat_lock);
	init_waitqueue_head(&sctx->list_wait);
452 453 454 455 456 457 458

	ret = scrub_setup_wr_ctx(sctx, &sctx->wr_ctx, fs_info,
				 fs_info->dev_replace.tgtdev, is_dev_replace);
	if (ret) {
		scrub_free_ctx(sctx);
		return ERR_PTR(ret);
	}
459
	return sctx;
A
Arne Jansen 已提交
460 461

nomem:
462
	scrub_free_ctx(sctx);
A
Arne Jansen 已提交
463 464 465
	return ERR_PTR(-ENOMEM);
}

466 467
static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root,
				     void *warn_ctx)
468 469 470 471 472 473 474
{
	u64 isize;
	u32 nlink;
	int ret;
	int i;
	struct extent_buffer *eb;
	struct btrfs_inode_item *inode_item;
475
	struct scrub_warning *swarn = warn_ctx;
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
	struct btrfs_fs_info *fs_info = swarn->dev->dev_root->fs_info;
	struct inode_fs_paths *ipath = NULL;
	struct btrfs_root *local_root;
	struct btrfs_key root_key;

	root_key.objectid = root;
	root_key.type = BTRFS_ROOT_ITEM_KEY;
	root_key.offset = (u64)-1;
	local_root = btrfs_read_fs_root_no_name(fs_info, &root_key);
	if (IS_ERR(local_root)) {
		ret = PTR_ERR(local_root);
		goto err;
	}

	ret = inode_item_info(inum, 0, local_root, swarn->path);
	if (ret) {
		btrfs_release_path(swarn->path);
		goto err;
	}

	eb = swarn->path->nodes[0];
	inode_item = btrfs_item_ptr(eb, swarn->path->slots[0],
					struct btrfs_inode_item);
	isize = btrfs_inode_size(eb, inode_item);
	nlink = btrfs_inode_nlink(eb, inode_item);
	btrfs_release_path(swarn->path);

	ipath = init_ipath(4096, local_root, swarn->path);
504 505 506 507 508
	if (IS_ERR(ipath)) {
		ret = PTR_ERR(ipath);
		ipath = NULL;
		goto err;
	}
509 510 511 512 513 514 515 516 517 518
	ret = paths_from_inode(inum, ipath);

	if (ret < 0)
		goto err;

	/*
	 * we deliberately ignore the bit ipath might have been too small to
	 * hold all of the paths here
	 */
	for (i = 0; i < ipath->fspath->elem_cnt; ++i)
519
		printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev "
520 521
			"%s, sector %llu, root %llu, inode %llu, offset %llu, "
			"length %llu, links %u (path: %s)\n", swarn->errstr,
522
			swarn->logical, rcu_str_deref(swarn->dev->name),
523 524
			(unsigned long long)swarn->sector, root, inum, offset,
			min(isize - offset, (u64)PAGE_SIZE), nlink,
525
			(char *)(unsigned long)ipath->fspath->val[i]);
526 527 528 529 530

	free_ipath(ipath);
	return 0;

err:
531
	printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev "
532 533
		"%s, sector %llu, root %llu, inode %llu, offset %llu: path "
		"resolving failed with ret=%d\n", swarn->errstr,
534
		swarn->logical, rcu_str_deref(swarn->dev->name),
535 536 537 538 539 540
		(unsigned long long)swarn->sector, root, inum, offset, ret);

	free_ipath(ipath);
	return 0;
}

541
static void scrub_print_warning(const char *errstr, struct scrub_block *sblock)
542
{
543 544
	struct btrfs_device *dev;
	struct btrfs_fs_info *fs_info;
545 546 547 548 549
	struct btrfs_path *path;
	struct btrfs_key found_key;
	struct extent_buffer *eb;
	struct btrfs_extent_item *ei;
	struct scrub_warning swarn;
550 551 552
	unsigned long ptr = 0;
	u64 extent_item_pos;
	u64 flags = 0;
553
	u64 ref_root;
554
	u32 item_size;
555 556
	u8 ref_level;
	const int bufsize = 4096;
557
	int ret;
558

559
	WARN_ON(sblock->page_count < 1);
560
	dev = sblock->pagev[0]->dev;
561 562
	fs_info = sblock->sctx->dev_root->fs_info;

563 564 565 566
	path = btrfs_alloc_path();

	swarn.scratch_buf = kmalloc(bufsize, GFP_NOFS);
	swarn.msg_buf = kmalloc(bufsize, GFP_NOFS);
567 568
	swarn.sector = (sblock->pagev[0]->physical) >> 9;
	swarn.logical = sblock->pagev[0]->logical;
569
	swarn.errstr = errstr;
570
	swarn.dev = NULL;
571 572 573 574 575 576
	swarn.msg_bufsize = bufsize;
	swarn.scratch_bufsize = bufsize;

	if (!path || !swarn.scratch_buf || !swarn.msg_buf)
		goto out;

577 578
	ret = extent_from_logical(fs_info, swarn.logical, path, &found_key,
				  &flags);
579 580 581
	if (ret < 0)
		goto out;

J
Jan Schmidt 已提交
582
	extent_item_pos = swarn.logical - found_key.objectid;
583 584 585 586 587 588
	swarn.extent_item_size = found_key.offset;

	eb = path->nodes[0];
	ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item);
	item_size = btrfs_item_size_nr(eb, path->slots[0]);

589
	if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
590 591 592
		do {
			ret = tree_backref_for_extent(&ptr, eb, ei, item_size,
							&ref_root, &ref_level);
593
			printk_in_rcu(KERN_WARNING
594
				"BTRFS: %s at logical %llu on dev %s, "
595
				"sector %llu: metadata %s (level %d) in tree "
596 597
				"%llu\n", errstr, swarn.logical,
				rcu_str_deref(dev->name),
598 599 600 601 602
				(unsigned long long)swarn.sector,
				ref_level ? "node" : "leaf",
				ret < 0 ? -1 : ref_level,
				ret < 0 ? -1 : ref_root);
		} while (ret != 1);
603
		btrfs_release_path(path);
604
	} else {
605
		btrfs_release_path(path);
606
		swarn.path = path;
607
		swarn.dev = dev;
608 609
		iterate_extent_inodes(fs_info, found_key.objectid,
					extent_item_pos, 1,
610 611 612 613 614 615 616 617 618
					scrub_print_warning_inode, &swarn);
	}

out:
	btrfs_free_path(path);
	kfree(swarn.scratch_buf);
	kfree(swarn.msg_buf);
}

619
static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx)
620
{
621
	struct page *page = NULL;
622
	unsigned long index;
623
	struct scrub_fixup_nodatasum *fixup = fixup_ctx;
624
	int ret;
625
	int corrected = 0;
626
	struct btrfs_key key;
627
	struct inode *inode = NULL;
628
	struct btrfs_fs_info *fs_info;
629 630
	u64 end = offset + PAGE_SIZE - 1;
	struct btrfs_root *local_root;
631
	int srcu_index;
632 633 634 635

	key.objectid = root;
	key.type = BTRFS_ROOT_ITEM_KEY;
	key.offset = (u64)-1;
636 637 638 639 640 641 642

	fs_info = fixup->root->fs_info;
	srcu_index = srcu_read_lock(&fs_info->subvol_srcu);

	local_root = btrfs_read_fs_root_no_name(fs_info, &key);
	if (IS_ERR(local_root)) {
		srcu_read_unlock(&fs_info->subvol_srcu, srcu_index);
643
		return PTR_ERR(local_root);
644
	}
645 646 647 648

	key.type = BTRFS_INODE_ITEM_KEY;
	key.objectid = inum;
	key.offset = 0;
649 650
	inode = btrfs_iget(fs_info->sb, &key, local_root, NULL);
	srcu_read_unlock(&fs_info->subvol_srcu, srcu_index);
651 652 653 654 655 656
	if (IS_ERR(inode))
		return PTR_ERR(inode);

	index = offset >> PAGE_CACHE_SHIFT;

	page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
	if (!page) {
		ret = -ENOMEM;
		goto out;
	}

	if (PageUptodate(page)) {
		if (PageDirty(page)) {
			/*
			 * we need to write the data to the defect sector. the
			 * data that was in that sector is not in memory,
			 * because the page was modified. we must not write the
			 * modified page to that sector.
			 *
			 * TODO: what could be done here: wait for the delalloc
			 *       runner to write out that page (might involve
			 *       COW) and see whether the sector is still
			 *       referenced afterwards.
			 *
			 * For the meantime, we'll treat this error
			 * incorrectable, although there is a chance that a
			 * later scrub will find the bad sector again and that
			 * there's no dirty page in memory, then.
			 */
			ret = -EIO;
			goto out;
		}
683 684
		fs_info = BTRFS_I(inode)->root->fs_info;
		ret = repair_io_failure(fs_info, offset, PAGE_SIZE,
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
					fixup->logical, page,
					fixup->mirror_num);
		unlock_page(page);
		corrected = !ret;
	} else {
		/*
		 * we need to get good data first. the general readpage path
		 * will call repair_io_failure for us, we just have to make
		 * sure we read the bad mirror.
		 */
		ret = set_extent_bits(&BTRFS_I(inode)->io_tree, offset, end,
					EXTENT_DAMAGED, GFP_NOFS);
		if (ret) {
			/* set_extent_bits should give proper error */
			WARN_ON(ret > 0);
			if (ret > 0)
				ret = -EFAULT;
			goto out;
		}

		ret = extent_read_full_page(&BTRFS_I(inode)->io_tree, page,
						btrfs_get_extent,
						fixup->mirror_num);
		wait_on_page_locked(page);

		corrected = !test_range_bit(&BTRFS_I(inode)->io_tree, offset,
						end, EXTENT_DAMAGED, 0, NULL);
		if (!corrected)
			clear_extent_bits(&BTRFS_I(inode)->io_tree, offset, end,
						EXTENT_DAMAGED, GFP_NOFS);
	}

out:
	if (page)
		put_page(page);
	if (inode)
		iput(inode);
722 723 724 725 726 727 728 729 730 731 732 733 734 735 736

	if (ret < 0)
		return ret;

	if (ret == 0 && corrected) {
		/*
		 * we only need to call readpage for one of the inodes belonging
		 * to this extent. so make iterate_extent_inodes stop
		 */
		return 1;
	}

	return -EIO;
}

737
static void scrub_fixup_nodatasum(struct btrfs_work *work)
738 739 740
{
	int ret;
	struct scrub_fixup_nodatasum *fixup;
741
	struct scrub_ctx *sctx;
742 743 744 745 746
	struct btrfs_trans_handle *trans = NULL;
	struct btrfs_path *path;
	int uncorrectable = 0;

	fixup = container_of(work, struct scrub_fixup_nodatasum, work);
747
	sctx = fixup->sctx;
748 749 750

	path = btrfs_alloc_path();
	if (!path) {
751 752 753
		spin_lock(&sctx->stat_lock);
		++sctx->stat.malloc_errors;
		spin_unlock(&sctx->stat_lock);
754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781
		uncorrectable = 1;
		goto out;
	}

	trans = btrfs_join_transaction(fixup->root);
	if (IS_ERR(trans)) {
		uncorrectable = 1;
		goto out;
	}

	/*
	 * the idea is to trigger a regular read through the standard path. we
	 * read a page from the (failed) logical address by specifying the
	 * corresponding copynum of the failed sector. thus, that readpage is
	 * expected to fail.
	 * that is the point where on-the-fly error correction will kick in
	 * (once it's finished) and rewrite the failed sector if a good copy
	 * can be found.
	 */
	ret = iterate_inodes_from_logical(fixup->logical, fixup->root->fs_info,
						path, scrub_fixup_readpage,
						fixup);
	if (ret < 0) {
		uncorrectable = 1;
		goto out;
	}
	WARN_ON(ret != 1);

782 783 784
	spin_lock(&sctx->stat_lock);
	++sctx->stat.corrected_errors;
	spin_unlock(&sctx->stat_lock);
785 786 787 788 789

out:
	if (trans && !IS_ERR(trans))
		btrfs_end_transaction(trans, fixup->root);
	if (uncorrectable) {
790 791 792
		spin_lock(&sctx->stat_lock);
		++sctx->stat.uncorrectable_errors;
		spin_unlock(&sctx->stat_lock);
793 794 795
		btrfs_dev_replace_stats_inc(
			&sctx->dev_root->fs_info->dev_replace.
			num_uncorrectable_read_errors);
796 797
		printk_ratelimited_in_rcu(KERN_ERR "BTRFS: "
		    "unable to fixup (nodatasum) error at logical %llu on dev %s\n",
798
			fixup->logical, rcu_str_deref(fixup->dev->name));
799 800 801 802 803
	}

	btrfs_free_path(path);
	kfree(fixup);

804
	scrub_pending_trans_workers_dec(sctx);
805 806
}

A
Arne Jansen 已提交
807
/*
808 809 810 811 812 813
 * scrub_handle_errored_block gets called when either verification of the
 * pages failed or the bio failed to read, e.g. with EIO. In the latter
 * case, this function handles all pages in the bio, even though only one
 * may be bad.
 * The goal of this function is to repair the errored block by using the
 * contents of one of the mirrors.
A
Arne Jansen 已提交
814
 */
815
static int scrub_handle_errored_block(struct scrub_block *sblock_to_check)
A
Arne Jansen 已提交
816
{
817
	struct scrub_ctx *sctx = sblock_to_check->sctx;
818
	struct btrfs_device *dev;
819 820 821 822 823 824 825 826 827 828 829 830 831 832
	struct btrfs_fs_info *fs_info;
	u64 length;
	u64 logical;
	u64 generation;
	unsigned int failed_mirror_index;
	unsigned int is_metadata;
	unsigned int have_csum;
	u8 *csum;
	struct scrub_block *sblocks_for_recheck; /* holds one for each mirror */
	struct scrub_block *sblock_bad;
	int ret;
	int mirror_index;
	int page_num;
	int success;
833
	static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
834 835 836
				      DEFAULT_RATELIMIT_BURST);

	BUG_ON(sblock_to_check->page_count < 1);
837
	fs_info = sctx->dev_root->fs_info;
838 839 840 841 842 843 844 845 846 847 848
	if (sblock_to_check->pagev[0]->flags & BTRFS_EXTENT_FLAG_SUPER) {
		/*
		 * if we find an error in a super block, we just report it.
		 * They will get written with the next transaction commit
		 * anyway
		 */
		spin_lock(&sctx->stat_lock);
		++sctx->stat.super_errors;
		spin_unlock(&sctx->stat_lock);
		return 0;
	}
849
	length = sblock_to_check->page_count * PAGE_SIZE;
850 851 852 853 854
	logical = sblock_to_check->pagev[0]->logical;
	generation = sblock_to_check->pagev[0]->generation;
	BUG_ON(sblock_to_check->pagev[0]->mirror_num < 1);
	failed_mirror_index = sblock_to_check->pagev[0]->mirror_num - 1;
	is_metadata = !(sblock_to_check->pagev[0]->flags &
855
			BTRFS_EXTENT_FLAG_DATA);
856 857 858
	have_csum = sblock_to_check->pagev[0]->have_csum;
	csum = sblock_to_check->pagev[0]->csum;
	dev = sblock_to_check->pagev[0]->dev;
859

860 861 862 863 864
	if (sctx->is_dev_replace && !is_metadata && !have_csum) {
		sblocks_for_recheck = NULL;
		goto nodatasum_case;
	}

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
	/*
	 * read all mirrors one after the other. This includes to
	 * re-read the extent or metadata block that failed (that was
	 * the cause that this fixup code is called) another time,
	 * page by page this time in order to know which pages
	 * caused I/O errors and which ones are good (for all mirrors).
	 * It is the goal to handle the situation when more than one
	 * mirror contains I/O errors, but the errors do not
	 * overlap, i.e. the data can be repaired by selecting the
	 * pages from those mirrors without I/O error on the
	 * particular pages. One example (with blocks >= 2 * PAGE_SIZE)
	 * would be that mirror #1 has an I/O error on the first page,
	 * the second page is good, and mirror #2 has an I/O error on
	 * the second page, but the first page is good.
	 * Then the first page of the first mirror can be repaired by
	 * taking the first page of the second mirror, and the
	 * second page of the second mirror can be repaired by
	 * copying the contents of the 2nd page of the 1st mirror.
	 * One more note: if the pages of one mirror contain I/O
	 * errors, the checksum cannot be verified. In order to get
	 * the best data for repairing, the first attempt is to find
	 * a mirror without I/O errors and with a validated checksum.
	 * Only if this is not possible, the pages are picked from
	 * mirrors with I/O errors without considering the checksum.
	 * If the latter is the case, at the end, the checksum of the
	 * repaired area is verified in order to correctly maintain
	 * the statistics.
	 */

	sblocks_for_recheck = kzalloc(BTRFS_MAX_MIRRORS *
				     sizeof(*sblocks_for_recheck),
				     GFP_NOFS);
	if (!sblocks_for_recheck) {
898 899 900 901 902
		spin_lock(&sctx->stat_lock);
		sctx->stat.malloc_errors++;
		sctx->stat.read_errors++;
		sctx->stat.uncorrectable_errors++;
		spin_unlock(&sctx->stat_lock);
903
		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_READ_ERRS);
904
		goto out;
A
Arne Jansen 已提交
905 906
	}

907
	/* setup the context, map the logical blocks and alloc the pages */
908
	ret = scrub_setup_recheck_block(sctx, fs_info, sblock_to_check, length,
909 910
					logical, sblocks_for_recheck);
	if (ret) {
911 912 913 914
		spin_lock(&sctx->stat_lock);
		sctx->stat.read_errors++;
		sctx->stat.uncorrectable_errors++;
		spin_unlock(&sctx->stat_lock);
915
		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_READ_ERRS);
916 917 918 919
		goto out;
	}
	BUG_ON(failed_mirror_index >= BTRFS_MAX_MIRRORS);
	sblock_bad = sblocks_for_recheck + failed_mirror_index;
920

921
	/* build and submit the bios for the failed mirror, check checksums */
922 923
	scrub_recheck_block(fs_info, sblock_bad, is_metadata, have_csum,
			    csum, generation, sctx->csum_size);
A
Arne Jansen 已提交
924

925 926 927 928 929 930 931 932 933 934
	if (!sblock_bad->header_error && !sblock_bad->checksum_error &&
	    sblock_bad->no_io_error_seen) {
		/*
		 * the error disappeared after reading page by page, or
		 * the area was part of a huge bio and other parts of the
		 * bio caused I/O errors, or the block layer merged several
		 * read requests into one and the error is caused by a
		 * different bio (usually one of the two latter cases is
		 * the cause)
		 */
935 936 937
		spin_lock(&sctx->stat_lock);
		sctx->stat.unverified_errors++;
		spin_unlock(&sctx->stat_lock);
A
Arne Jansen 已提交
938

939 940
		if (sctx->is_dev_replace)
			scrub_write_block_to_dev_replace(sblock_bad);
941
		goto out;
A
Arne Jansen 已提交
942 943
	}

944
	if (!sblock_bad->no_io_error_seen) {
945 946 947
		spin_lock(&sctx->stat_lock);
		sctx->stat.read_errors++;
		spin_unlock(&sctx->stat_lock);
948 949
		if (__ratelimit(&_rs))
			scrub_print_warning("i/o error", sblock_to_check);
950
		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_READ_ERRS);
951
	} else if (sblock_bad->checksum_error) {
952 953 954
		spin_lock(&sctx->stat_lock);
		sctx->stat.csum_errors++;
		spin_unlock(&sctx->stat_lock);
955 956
		if (__ratelimit(&_rs))
			scrub_print_warning("checksum error", sblock_to_check);
957
		btrfs_dev_stat_inc_and_print(dev,
958
					     BTRFS_DEV_STAT_CORRUPTION_ERRS);
959
	} else if (sblock_bad->header_error) {
960 961 962
		spin_lock(&sctx->stat_lock);
		sctx->stat.verify_errors++;
		spin_unlock(&sctx->stat_lock);
963 964 965
		if (__ratelimit(&_rs))
			scrub_print_warning("checksum/header error",
					    sblock_to_check);
966
		if (sblock_bad->generation_error)
967
			btrfs_dev_stat_inc_and_print(dev,
968 969
				BTRFS_DEV_STAT_GENERATION_ERRS);
		else
970
			btrfs_dev_stat_inc_and_print(dev,
971
				BTRFS_DEV_STAT_CORRUPTION_ERRS);
972
	}
A
Arne Jansen 已提交
973

974 975 976 977
	if (sctx->readonly) {
		ASSERT(!sctx->is_dev_replace);
		goto out;
	}
A
Arne Jansen 已提交
978

979 980
	if (!is_metadata && !have_csum) {
		struct scrub_fixup_nodatasum *fixup_nodatasum;
A
Arne Jansen 已提交
981

982 983 984
nodatasum_case:
		WARN_ON(sctx->is_dev_replace);

985 986 987 988 989 990 991 992 993 994
		/*
		 * !is_metadata and !have_csum, this means that the data
		 * might not be COW'ed, that it might be modified
		 * concurrently. The general strategy to work on the
		 * commit root does not help in the case when COW is not
		 * used.
		 */
		fixup_nodatasum = kzalloc(sizeof(*fixup_nodatasum), GFP_NOFS);
		if (!fixup_nodatasum)
			goto did_not_correct_error;
995
		fixup_nodatasum->sctx = sctx;
996
		fixup_nodatasum->dev = dev;
997 998 999
		fixup_nodatasum->logical = logical;
		fixup_nodatasum->root = fs_info->extent_root;
		fixup_nodatasum->mirror_num = failed_mirror_index + 1;
1000
		scrub_pending_trans_workers_inc(sctx);
1001 1002 1003 1004
		btrfs_init_work(&fixup_nodatasum->work, scrub_fixup_nodatasum,
				NULL, NULL);
		btrfs_queue_work(fs_info->scrub_workers,
				 &fixup_nodatasum->work);
1005
		goto out;
A
Arne Jansen 已提交
1006 1007
	}

1008 1009
	/*
	 * now build and submit the bios for the other mirrors, check
1010 1011
	 * checksums.
	 * First try to pick the mirror which is completely without I/O
1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026
	 * errors and also does not have a checksum error.
	 * If one is found, and if a checksum is present, the full block
	 * that is known to contain an error is rewritten. Afterwards
	 * the block is known to be corrected.
	 * If a mirror is found which is completely correct, and no
	 * checksum is present, only those pages are rewritten that had
	 * an I/O error in the block to be repaired, since it cannot be
	 * determined, which copy of the other pages is better (and it
	 * could happen otherwise that a correct page would be
	 * overwritten by a bad one).
	 */
	for (mirror_index = 0;
	     mirror_index < BTRFS_MAX_MIRRORS &&
	     sblocks_for_recheck[mirror_index].page_count > 0;
	     mirror_index++) {
1027
		struct scrub_block *sblock_other;
1028

1029 1030 1031 1032 1033
		if (mirror_index == failed_mirror_index)
			continue;
		sblock_other = sblocks_for_recheck + mirror_index;

		/* build and submit the bios, check checksums */
1034 1035 1036 1037 1038
		scrub_recheck_block(fs_info, sblock_other, is_metadata,
				    have_csum, csum, generation,
				    sctx->csum_size);

		if (!sblock_other->header_error &&
1039 1040
		    !sblock_other->checksum_error &&
		    sblock_other->no_io_error_seen) {
1041 1042 1043 1044 1045 1046 1047 1048 1049
			if (sctx->is_dev_replace) {
				scrub_write_block_to_dev_replace(sblock_other);
			} else {
				int force_write = is_metadata || have_csum;

				ret = scrub_repair_block_from_good_copy(
						sblock_bad, sblock_other,
						force_write);
			}
1050 1051 1052 1053
			if (0 == ret)
				goto corrected_error;
		}
	}
A
Arne Jansen 已提交
1054 1055

	/*
1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113
	 * for dev_replace, pick good pages and write to the target device.
	 */
	if (sctx->is_dev_replace) {
		success = 1;
		for (page_num = 0; page_num < sblock_bad->page_count;
		     page_num++) {
			int sub_success;

			sub_success = 0;
			for (mirror_index = 0;
			     mirror_index < BTRFS_MAX_MIRRORS &&
			     sblocks_for_recheck[mirror_index].page_count > 0;
			     mirror_index++) {
				struct scrub_block *sblock_other =
					sblocks_for_recheck + mirror_index;
				struct scrub_page *page_other =
					sblock_other->pagev[page_num];

				if (!page_other->io_error) {
					ret = scrub_write_page_to_dev_replace(
							sblock_other, page_num);
					if (ret == 0) {
						/* succeeded for this page */
						sub_success = 1;
						break;
					} else {
						btrfs_dev_replace_stats_inc(
							&sctx->dev_root->
							fs_info->dev_replace.
							num_write_errors);
					}
				}
			}

			if (!sub_success) {
				/*
				 * did not find a mirror to fetch the page
				 * from. scrub_write_page_to_dev_replace()
				 * handles this case (page->io_error), by
				 * filling the block with zeros before
				 * submitting the write request
				 */
				success = 0;
				ret = scrub_write_page_to_dev_replace(
						sblock_bad, page_num);
				if (ret)
					btrfs_dev_replace_stats_inc(
						&sctx->dev_root->fs_info->
						dev_replace.num_write_errors);
			}
		}

		goto out;
	}

	/*
	 * for regular scrub, repair those pages that are errored.
	 * In case of I/O errors in the area that is supposed to be
1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134
	 * repaired, continue by picking good copies of those pages.
	 * Select the good pages from mirrors to rewrite bad pages from
	 * the area to fix. Afterwards verify the checksum of the block
	 * that is supposed to be repaired. This verification step is
	 * only done for the purpose of statistic counting and for the
	 * final scrub report, whether errors remain.
	 * A perfect algorithm could make use of the checksum and try
	 * all possible combinations of pages from the different mirrors
	 * until the checksum verification succeeds. For example, when
	 * the 2nd page of mirror #1 faces I/O errors, and the 2nd page
	 * of mirror #2 is readable but the final checksum test fails,
	 * then the 2nd page of mirror #3 could be tried, whether now
	 * the final checksum succeedes. But this would be a rare
	 * exception and is therefore not implemented. At least it is
	 * avoided that the good copy is overwritten.
	 * A more useful improvement would be to pick the sectors
	 * without I/O error based on sector sizes (512 bytes on legacy
	 * disks) instead of on PAGE_SIZE. Then maybe 512 byte of one
	 * mirror could be repaired by taking 512 byte of a different
	 * mirror, even if other 512 byte sectors in the same PAGE_SIZE
	 * area are unreadable.
A
Arne Jansen 已提交
1135 1136
	 */

1137 1138 1139 1140 1141 1142
	/* can only fix I/O errors from here on */
	if (sblock_bad->no_io_error_seen)
		goto did_not_correct_error;

	success = 1;
	for (page_num = 0; page_num < sblock_bad->page_count; page_num++) {
1143
		struct scrub_page *page_bad = sblock_bad->pagev[page_num];
1144 1145

		if (!page_bad->io_error)
A
Arne Jansen 已提交
1146
			continue;
1147 1148 1149 1150 1151 1152 1153

		for (mirror_index = 0;
		     mirror_index < BTRFS_MAX_MIRRORS &&
		     sblocks_for_recheck[mirror_index].page_count > 0;
		     mirror_index++) {
			struct scrub_block *sblock_other = sblocks_for_recheck +
							   mirror_index;
1154 1155
			struct scrub_page *page_other = sblock_other->pagev[
							page_num];
1156 1157 1158 1159 1160 1161 1162 1163 1164

			if (!page_other->io_error) {
				ret = scrub_repair_page_from_good_copy(
					sblock_bad, sblock_other, page_num, 0);
				if (0 == ret) {
					page_bad->io_error = 0;
					break; /* succeeded for this page */
				}
			}
I
Ilya Dryomov 已提交
1165
		}
A
Arne Jansen 已提交
1166

1167 1168 1169 1170
		if (page_bad->io_error) {
			/* did not find a mirror to copy the page from */
			success = 0;
		}
A
Arne Jansen 已提交
1171 1172
	}

1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183
	if (success) {
		if (is_metadata || have_csum) {
			/*
			 * need to verify the checksum now that all
			 * sectors on disk are repaired (the write
			 * request for data to be repaired is on its way).
			 * Just be lazy and use scrub_recheck_block()
			 * which re-reads the data before the checksum
			 * is verified, but most likely the data comes out
			 * of the page cache.
			 */
1184 1185 1186 1187
			scrub_recheck_block(fs_info, sblock_bad,
					    is_metadata, have_csum, csum,
					    generation, sctx->csum_size);
			if (!sblock_bad->header_error &&
1188 1189 1190 1191 1192 1193 1194
			    !sblock_bad->checksum_error &&
			    sblock_bad->no_io_error_seen)
				goto corrected_error;
			else
				goto did_not_correct_error;
		} else {
corrected_error:
1195 1196 1197
			spin_lock(&sctx->stat_lock);
			sctx->stat.corrected_errors++;
			spin_unlock(&sctx->stat_lock);
1198
			printk_ratelimited_in_rcu(KERN_ERR
1199
				"BTRFS: fixed up error at logical %llu on dev %s\n",
1200
				logical, rcu_str_deref(dev->name));
A
Arne Jansen 已提交
1201
		}
1202 1203
	} else {
did_not_correct_error:
1204 1205 1206
		spin_lock(&sctx->stat_lock);
		sctx->stat.uncorrectable_errors++;
		spin_unlock(&sctx->stat_lock);
1207
		printk_ratelimited_in_rcu(KERN_ERR
1208
			"BTRFS: unable to fixup (regular) error at logical %llu on dev %s\n",
1209
			logical, rcu_str_deref(dev->name));
I
Ilya Dryomov 已提交
1210
	}
A
Arne Jansen 已提交
1211

1212 1213 1214 1215 1216 1217 1218 1219
out:
	if (sblocks_for_recheck) {
		for (mirror_index = 0; mirror_index < BTRFS_MAX_MIRRORS;
		     mirror_index++) {
			struct scrub_block *sblock = sblocks_for_recheck +
						     mirror_index;
			int page_index;

1220 1221 1222 1223 1224
			for (page_index = 0; page_index < sblock->page_count;
			     page_index++) {
				sblock->pagev[page_index]->sblock = NULL;
				scrub_page_put(sblock->pagev[page_index]);
			}
1225 1226 1227
		}
		kfree(sblocks_for_recheck);
	}
A
Arne Jansen 已提交
1228

1229 1230
	return 0;
}
A
Arne Jansen 已提交
1231

1232
static int scrub_setup_recheck_block(struct scrub_ctx *sctx,
1233
				     struct btrfs_fs_info *fs_info,
1234
				     struct scrub_block *original_sblock,
1235 1236 1237 1238 1239 1240 1241 1242
				     u64 length, u64 logical,
				     struct scrub_block *sblocks_for_recheck)
{
	int page_index;
	int mirror_index;
	int ret;

	/*
1243
	 * note: the two members ref_count and outstanding_pages
1244 1245 1246 1247 1248 1249 1250 1251 1252
	 * are not used (and not set) in the blocks that are used for
	 * the recheck procedure
	 */

	page_index = 0;
	while (length > 0) {
		u64 sublen = min_t(u64, length, PAGE_SIZE);
		u64 mapped_length = sublen;
		struct btrfs_bio *bbio = NULL;
A
Arne Jansen 已提交
1253

1254 1255 1256 1257
		/*
		 * with a length of PAGE_SIZE, each returned stripe
		 * represents one mirror
		 */
1258 1259
		ret = btrfs_map_block(fs_info, REQ_GET_READ_MIRRORS, logical,
				      &mapped_length, &bbio, 0);
1260 1261 1262 1263
		if (ret || !bbio || mapped_length < sublen) {
			kfree(bbio);
			return -EIO;
		}
A
Arne Jansen 已提交
1264

1265
		BUG_ON(page_index >= SCRUB_PAGES_PER_RD_BIO);
1266 1267 1268 1269 1270 1271 1272 1273 1274
		for (mirror_index = 0; mirror_index < (int)bbio->num_stripes;
		     mirror_index++) {
			struct scrub_block *sblock;
			struct scrub_page *page;

			if (mirror_index >= BTRFS_MAX_MIRRORS)
				continue;

			sblock = sblocks_for_recheck + mirror_index;
1275 1276 1277 1278
			sblock->sctx = sctx;
			page = kzalloc(sizeof(*page), GFP_NOFS);
			if (!page) {
leave_nomem:
1279 1280 1281
				spin_lock(&sctx->stat_lock);
				sctx->stat.malloc_errors++;
				spin_unlock(&sctx->stat_lock);
1282
				kfree(bbio);
1283 1284
				return -ENOMEM;
			}
1285 1286 1287 1288
			scrub_page_get(page);
			sblock->pagev[page_index] = page;
			page->logical = logical;
			page->physical = bbio->stripes[mirror_index].physical;
1289 1290 1291 1292
			BUG_ON(page_index >= original_sblock->page_count);
			page->physical_for_dev_replace =
				original_sblock->pagev[page_index]->
				physical_for_dev_replace;
1293 1294 1295
			/* for missing devices, dev->bdev is NULL */
			page->dev = bbio->stripes[mirror_index].dev;
			page->mirror_num = mirror_index + 1;
1296
			sblock->page_count++;
1297 1298 1299
			page->page = alloc_page(GFP_NOFS);
			if (!page->page)
				goto leave_nomem;
1300 1301 1302 1303 1304 1305 1306 1307
		}
		kfree(bbio);
		length -= sublen;
		logical += sublen;
		page_index++;
	}

	return 0;
I
Ilya Dryomov 已提交
1308 1309
}

1310 1311 1312 1313 1314 1315 1316
/*
 * this function will check the on disk data for checksum errors, header
 * errors and read I/O errors. If any I/O errors happen, the exact pages
 * which are errored are marked as being bad. The goal is to enable scrub
 * to take those pages that are not errored from all the mirrors so that
 * the pages that are errored in the just handled mirror can be repaired.
 */
1317 1318 1319 1320
static void scrub_recheck_block(struct btrfs_fs_info *fs_info,
				struct scrub_block *sblock, int is_metadata,
				int have_csum, u8 *csum, u64 generation,
				u16 csum_size)
I
Ilya Dryomov 已提交
1321
{
1322
	int page_num;
I
Ilya Dryomov 已提交
1323

1324 1325 1326
	sblock->no_io_error_seen = 1;
	sblock->header_error = 0;
	sblock->checksum_error = 0;
I
Ilya Dryomov 已提交
1327

1328 1329
	for (page_num = 0; page_num < sblock->page_count; page_num++) {
		struct bio *bio;
1330
		struct scrub_page *page = sblock->pagev[page_num];
1331

1332
		if (page->dev->bdev == NULL) {
1333 1334 1335 1336 1337
			page->io_error = 1;
			sblock->no_io_error_seen = 0;
			continue;
		}

1338
		WARN_ON(!page->page);
1339
		bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
1340 1341 1342 1343 1344
		if (!bio) {
			page->io_error = 1;
			sblock->no_io_error_seen = 0;
			continue;
		}
1345
		bio->bi_bdev = page->dev->bdev;
1346 1347
		bio->bi_sector = page->physical >> 9;

1348
		bio_add_page(bio, page->page, PAGE_SIZE, 0);
1349
		if (btrfsic_submit_bio_wait(READ, bio))
1350
			sblock->no_io_error_seen = 0;
1351

1352 1353
		bio_put(bio);
	}
I
Ilya Dryomov 已提交
1354

1355 1356 1357 1358 1359
	if (sblock->no_io_error_seen)
		scrub_recheck_block_checksum(fs_info, sblock, is_metadata,
					     have_csum, csum, generation,
					     csum_size);

1360
	return;
A
Arne Jansen 已提交
1361 1362
}

1363 1364 1365 1366 1367
static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
					 struct scrub_block *sblock,
					 int is_metadata, int have_csum,
					 const u8 *csum, u64 generation,
					 u16 csum_size)
A
Arne Jansen 已提交
1368
{
1369 1370 1371 1372 1373
	int page_num;
	u8 calculated_csum[BTRFS_CSUM_SIZE];
	u32 crc = ~(u32)0;
	void *mapped_buffer;

1374
	WARN_ON(!sblock->pagev[0]->page);
1375 1376 1377
	if (is_metadata) {
		struct btrfs_header *h;

1378
		mapped_buffer = kmap_atomic(sblock->pagev[0]->page);
1379 1380
		h = (struct btrfs_header *)mapped_buffer;

1381
		if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h) ||
1382 1383
		    memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE) ||
		    memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid,
1384
			   BTRFS_UUID_SIZE)) {
1385
			sblock->header_error = 1;
1386
		} else if (generation != btrfs_stack_header_generation(h)) {
1387 1388 1389
			sblock->header_error = 1;
			sblock->generation_error = 1;
		}
1390 1391 1392 1393
		csum = h->csum;
	} else {
		if (!have_csum)
			return;
A
Arne Jansen 已提交
1394

1395
		mapped_buffer = kmap_atomic(sblock->pagev[0]->page);
1396
	}
A
Arne Jansen 已提交
1397

1398 1399
	for (page_num = 0;;) {
		if (page_num == 0 && is_metadata)
1400
			crc = btrfs_csum_data(
1401 1402 1403
				((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE,
				crc, PAGE_SIZE - BTRFS_CSUM_SIZE);
		else
1404
			crc = btrfs_csum_data(mapped_buffer, crc, PAGE_SIZE);
1405

1406
		kunmap_atomic(mapped_buffer);
1407 1408 1409
		page_num++;
		if (page_num >= sblock->page_count)
			break;
1410
		WARN_ON(!sblock->pagev[page_num]->page);
1411

1412
		mapped_buffer = kmap_atomic(sblock->pagev[page_num]->page);
1413 1414 1415 1416 1417
	}

	btrfs_csum_final(crc, calculated_csum);
	if (memcmp(calculated_csum, csum, csum_size))
		sblock->checksum_error = 1;
A
Arne Jansen 已提交
1418 1419
}

1420 1421 1422 1423 1424 1425
static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad,
					     struct scrub_block *sblock_good,
					     int force_write)
{
	int page_num;
	int ret = 0;
I
Ilya Dryomov 已提交
1426

1427 1428
	for (page_num = 0; page_num < sblock_bad->page_count; page_num++) {
		int ret_sub;
I
Ilya Dryomov 已提交
1429

1430 1431 1432 1433 1434 1435
		ret_sub = scrub_repair_page_from_good_copy(sblock_bad,
							   sblock_good,
							   page_num,
							   force_write);
		if (ret_sub)
			ret = ret_sub;
A
Arne Jansen 已提交
1436
	}
1437 1438 1439 1440 1441 1442 1443 1444

	return ret;
}

static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad,
					    struct scrub_block *sblock_good,
					    int page_num, int force_write)
{
1445 1446
	struct scrub_page *page_bad = sblock_bad->pagev[page_num];
	struct scrub_page *page_good = sblock_good->pagev[page_num];
1447

1448 1449
	BUG_ON(page_bad->page == NULL);
	BUG_ON(page_good->page == NULL);
1450 1451 1452 1453 1454
	if (force_write || sblock_bad->header_error ||
	    sblock_bad->checksum_error || page_bad->io_error) {
		struct bio *bio;
		int ret;

1455
		if (!page_bad->dev->bdev) {
1456 1457 1458
			printk_ratelimited(KERN_WARNING "BTRFS: "
				"scrub_repair_page_from_good_copy(bdev == NULL) "
				"is unexpected!\n");
1459 1460 1461
			return -EIO;
		}

1462
		bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
1463 1464
		if (!bio)
			return -EIO;
1465
		bio->bi_bdev = page_bad->dev->bdev;
1466 1467 1468 1469 1470 1471
		bio->bi_sector = page_bad->physical >> 9;

		ret = bio_add_page(bio, page_good->page, PAGE_SIZE, 0);
		if (PAGE_SIZE != ret) {
			bio_put(bio);
			return -EIO;
1472
		}
1473

1474
		if (btrfsic_submit_bio_wait(WRITE, bio)) {
1475 1476
			btrfs_dev_stat_inc_and_print(page_bad->dev,
				BTRFS_DEV_STAT_WRITE_ERRS);
1477 1478 1479
			btrfs_dev_replace_stats_inc(
				&sblock_bad->sctx->dev_root->fs_info->
				dev_replace.num_write_errors);
1480 1481 1482
			bio_put(bio);
			return -EIO;
		}
1483
		bio_put(bio);
A
Arne Jansen 已提交
1484 1485
	}

1486 1487 1488
	return 0;
}

1489 1490 1491 1492 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
static void scrub_write_block_to_dev_replace(struct scrub_block *sblock)
{
	int page_num;

	for (page_num = 0; page_num < sblock->page_count; page_num++) {
		int ret;

		ret = scrub_write_page_to_dev_replace(sblock, page_num);
		if (ret)
			btrfs_dev_replace_stats_inc(
				&sblock->sctx->dev_root->fs_info->dev_replace.
				num_write_errors);
	}
}

static int scrub_write_page_to_dev_replace(struct scrub_block *sblock,
					   int page_num)
{
	struct scrub_page *spage = sblock->pagev[page_num];

	BUG_ON(spage->page == NULL);
	if (spage->io_error) {
		void *mapped_buffer = kmap_atomic(spage->page);

		memset(mapped_buffer, 0, PAGE_CACHE_SIZE);
		flush_dcache_page(spage->page);
		kunmap_atomic(mapped_buffer);
	}
	return scrub_add_page_to_wr_bio(sblock->sctx, spage);
}

static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx,
				    struct scrub_page *spage)
{
	struct scrub_wr_ctx *wr_ctx = &sctx->wr_ctx;
	struct scrub_bio *sbio;
	int ret;

	mutex_lock(&wr_ctx->wr_lock);
again:
	if (!wr_ctx->wr_curr_bio) {
		wr_ctx->wr_curr_bio = kzalloc(sizeof(*wr_ctx->wr_curr_bio),
					      GFP_NOFS);
		if (!wr_ctx->wr_curr_bio) {
			mutex_unlock(&wr_ctx->wr_lock);
			return -ENOMEM;
		}
		wr_ctx->wr_curr_bio->sctx = sctx;
		wr_ctx->wr_curr_bio->page_count = 0;
	}
	sbio = wr_ctx->wr_curr_bio;
	if (sbio->page_count == 0) {
		struct bio *bio;

		sbio->physical = spage->physical_for_dev_replace;
		sbio->logical = spage->logical;
		sbio->dev = wr_ctx->tgtdev;
		bio = sbio->bio;
		if (!bio) {
1548
			bio = btrfs_io_bio_alloc(GFP_NOFS, wr_ctx->pages_per_wr_bio);
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 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617
			if (!bio) {
				mutex_unlock(&wr_ctx->wr_lock);
				return -ENOMEM;
			}
			sbio->bio = bio;
		}

		bio->bi_private = sbio;
		bio->bi_end_io = scrub_wr_bio_end_io;
		bio->bi_bdev = sbio->dev->bdev;
		bio->bi_sector = sbio->physical >> 9;
		sbio->err = 0;
	} else if (sbio->physical + sbio->page_count * PAGE_SIZE !=
		   spage->physical_for_dev_replace ||
		   sbio->logical + sbio->page_count * PAGE_SIZE !=
		   spage->logical) {
		scrub_wr_submit(sctx);
		goto again;
	}

	ret = bio_add_page(sbio->bio, spage->page, PAGE_SIZE, 0);
	if (ret != PAGE_SIZE) {
		if (sbio->page_count < 1) {
			bio_put(sbio->bio);
			sbio->bio = NULL;
			mutex_unlock(&wr_ctx->wr_lock);
			return -EIO;
		}
		scrub_wr_submit(sctx);
		goto again;
	}

	sbio->pagev[sbio->page_count] = spage;
	scrub_page_get(spage);
	sbio->page_count++;
	if (sbio->page_count == wr_ctx->pages_per_wr_bio)
		scrub_wr_submit(sctx);
	mutex_unlock(&wr_ctx->wr_lock);

	return 0;
}

static void scrub_wr_submit(struct scrub_ctx *sctx)
{
	struct scrub_wr_ctx *wr_ctx = &sctx->wr_ctx;
	struct scrub_bio *sbio;

	if (!wr_ctx->wr_curr_bio)
		return;

	sbio = wr_ctx->wr_curr_bio;
	wr_ctx->wr_curr_bio = NULL;
	WARN_ON(!sbio->bio->bi_bdev);
	scrub_pending_bio_inc(sctx);
	/* process all writes in a single worker thread. Then the block layer
	 * orders the requests before sending them to the driver which
	 * doubled the write performance on spinning disks when measured
	 * with Linux 3.5 */
	btrfsic_submit_bio(WRITE, sbio->bio);
}

static void scrub_wr_bio_end_io(struct bio *bio, int err)
{
	struct scrub_bio *sbio = bio->bi_private;
	struct btrfs_fs_info *fs_info = sbio->dev->dev_root->fs_info;

	sbio->err = err;
	sbio->bio = bio;

1618 1619
	btrfs_init_work(&sbio->work, scrub_wr_bio_end_io_worker, NULL, NULL);
	btrfs_queue_work(fs_info->scrub_wr_completion_workers, &sbio->work);
1620 1621
}

1622
static void scrub_wr_bio_end_io_worker(struct btrfs_work *work)
1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650
{
	struct scrub_bio *sbio = container_of(work, struct scrub_bio, work);
	struct scrub_ctx *sctx = sbio->sctx;
	int i;

	WARN_ON(sbio->page_count > SCRUB_PAGES_PER_WR_BIO);
	if (sbio->err) {
		struct btrfs_dev_replace *dev_replace =
			&sbio->sctx->dev_root->fs_info->dev_replace;

		for (i = 0; i < sbio->page_count; i++) {
			struct scrub_page *spage = sbio->pagev[i];

			spage->io_error = 1;
			btrfs_dev_replace_stats_inc(&dev_replace->
						    num_write_errors);
		}
	}

	for (i = 0; i < sbio->page_count; i++)
		scrub_page_put(sbio->pagev[i]);

	bio_put(sbio->bio);
	kfree(sbio);
	scrub_pending_bio_dec(sctx);
}

static int scrub_checksum(struct scrub_block *sblock)
1651 1652 1653 1654
{
	u64 flags;
	int ret;

1655 1656
	WARN_ON(sblock->page_count < 1);
	flags = sblock->pagev[0]->flags;
1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667
	ret = 0;
	if (flags & BTRFS_EXTENT_FLAG_DATA)
		ret = scrub_checksum_data(sblock);
	else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)
		ret = scrub_checksum_tree_block(sblock);
	else if (flags & BTRFS_EXTENT_FLAG_SUPER)
		(void)scrub_checksum_super(sblock);
	else
		WARN_ON(1);
	if (ret)
		scrub_handle_errored_block(sblock);
1668 1669

	return ret;
A
Arne Jansen 已提交
1670 1671
}

1672
static int scrub_checksum_data(struct scrub_block *sblock)
A
Arne Jansen 已提交
1673
{
1674
	struct scrub_ctx *sctx = sblock->sctx;
A
Arne Jansen 已提交
1675
	u8 csum[BTRFS_CSUM_SIZE];
1676 1677 1678
	u8 *on_disk_csum;
	struct page *page;
	void *buffer;
A
Arne Jansen 已提交
1679 1680
	u32 crc = ~(u32)0;
	int fail = 0;
1681 1682
	u64 len;
	int index;
A
Arne Jansen 已提交
1683

1684
	BUG_ON(sblock->page_count < 1);
1685
	if (!sblock->pagev[0]->have_csum)
A
Arne Jansen 已提交
1686 1687
		return 0;

1688 1689
	on_disk_csum = sblock->pagev[0]->csum;
	page = sblock->pagev[0]->page;
1690
	buffer = kmap_atomic(page);
1691

1692
	len = sctx->sectorsize;
1693 1694 1695 1696
	index = 0;
	for (;;) {
		u64 l = min_t(u64, len, PAGE_SIZE);

1697
		crc = btrfs_csum_data(buffer, crc, l);
1698
		kunmap_atomic(buffer);
1699 1700 1701 1702 1703
		len -= l;
		if (len == 0)
			break;
		index++;
		BUG_ON(index >= sblock->page_count);
1704 1705
		BUG_ON(!sblock->pagev[index]->page);
		page = sblock->pagev[index]->page;
1706
		buffer = kmap_atomic(page);
1707 1708
	}

A
Arne Jansen 已提交
1709
	btrfs_csum_final(crc, csum);
1710
	if (memcmp(csum, on_disk_csum, sctx->csum_size))
A
Arne Jansen 已提交
1711 1712 1713 1714 1715
		fail = 1;

	return fail;
}

1716
static int scrub_checksum_tree_block(struct scrub_block *sblock)
A
Arne Jansen 已提交
1717
{
1718
	struct scrub_ctx *sctx = sblock->sctx;
A
Arne Jansen 已提交
1719
	struct btrfs_header *h;
1720
	struct btrfs_root *root = sctx->dev_root;
A
Arne Jansen 已提交
1721
	struct btrfs_fs_info *fs_info = root->fs_info;
1722 1723 1724 1725 1726 1727
	u8 calculated_csum[BTRFS_CSUM_SIZE];
	u8 on_disk_csum[BTRFS_CSUM_SIZE];
	struct page *page;
	void *mapped_buffer;
	u64 mapped_size;
	void *p;
A
Arne Jansen 已提交
1728 1729 1730
	u32 crc = ~(u32)0;
	int fail = 0;
	int crc_fail = 0;
1731 1732 1733 1734
	u64 len;
	int index;

	BUG_ON(sblock->page_count < 1);
1735
	page = sblock->pagev[0]->page;
1736
	mapped_buffer = kmap_atomic(page);
1737
	h = (struct btrfs_header *)mapped_buffer;
1738
	memcpy(on_disk_csum, h->csum, sctx->csum_size);
A
Arne Jansen 已提交
1739 1740 1741 1742 1743 1744 1745

	/*
	 * we don't use the getter functions here, as we
	 * a) don't have an extent buffer and
	 * b) the page is already kmapped
	 */

1746
	if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h))
A
Arne Jansen 已提交
1747 1748
		++fail;

1749
	if (sblock->pagev[0]->generation != btrfs_stack_header_generation(h))
A
Arne Jansen 已提交
1750 1751 1752 1753 1754 1755 1756 1757 1758
		++fail;

	if (memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
		++fail;

	if (memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid,
		   BTRFS_UUID_SIZE))
		++fail;

1759
	WARN_ON(sctx->nodesize != sctx->leafsize);
1760
	len = sctx->nodesize - BTRFS_CSUM_SIZE;
1761 1762 1763 1764 1765 1766
	mapped_size = PAGE_SIZE - BTRFS_CSUM_SIZE;
	p = ((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE;
	index = 0;
	for (;;) {
		u64 l = min_t(u64, len, mapped_size);

1767
		crc = btrfs_csum_data(p, crc, l);
1768
		kunmap_atomic(mapped_buffer);
1769 1770 1771 1772 1773
		len -= l;
		if (len == 0)
			break;
		index++;
		BUG_ON(index >= sblock->page_count);
1774 1775
		BUG_ON(!sblock->pagev[index]->page);
		page = sblock->pagev[index]->page;
1776
		mapped_buffer = kmap_atomic(page);
1777 1778 1779 1780 1781
		mapped_size = PAGE_SIZE;
		p = mapped_buffer;
	}

	btrfs_csum_final(crc, calculated_csum);
1782
	if (memcmp(calculated_csum, on_disk_csum, sctx->csum_size))
A
Arne Jansen 已提交
1783 1784 1785 1786 1787
		++crc_fail;

	return fail || crc_fail;
}

1788
static int scrub_checksum_super(struct scrub_block *sblock)
A
Arne Jansen 已提交
1789 1790
{
	struct btrfs_super_block *s;
1791
	struct scrub_ctx *sctx = sblock->sctx;
1792
	struct btrfs_root *root = sctx->dev_root;
A
Arne Jansen 已提交
1793
	struct btrfs_fs_info *fs_info = root->fs_info;
1794 1795 1796 1797 1798 1799
	u8 calculated_csum[BTRFS_CSUM_SIZE];
	u8 on_disk_csum[BTRFS_CSUM_SIZE];
	struct page *page;
	void *mapped_buffer;
	u64 mapped_size;
	void *p;
A
Arne Jansen 已提交
1800
	u32 crc = ~(u32)0;
1801 1802
	int fail_gen = 0;
	int fail_cor = 0;
1803 1804
	u64 len;
	int index;
A
Arne Jansen 已提交
1805

1806
	BUG_ON(sblock->page_count < 1);
1807
	page = sblock->pagev[0]->page;
1808
	mapped_buffer = kmap_atomic(page);
1809
	s = (struct btrfs_super_block *)mapped_buffer;
1810
	memcpy(on_disk_csum, s->csum, sctx->csum_size);
A
Arne Jansen 已提交
1811

1812
	if (sblock->pagev[0]->logical != btrfs_super_bytenr(s))
1813
		++fail_cor;
A
Arne Jansen 已提交
1814

1815
	if (sblock->pagev[0]->generation != btrfs_super_generation(s))
1816
		++fail_gen;
A
Arne Jansen 已提交
1817 1818

	if (memcmp(s->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
1819
		++fail_cor;
A
Arne Jansen 已提交
1820

1821 1822 1823 1824 1825 1826 1827
	len = BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE;
	mapped_size = PAGE_SIZE - BTRFS_CSUM_SIZE;
	p = ((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE;
	index = 0;
	for (;;) {
		u64 l = min_t(u64, len, mapped_size);

1828
		crc = btrfs_csum_data(p, crc, l);
1829
		kunmap_atomic(mapped_buffer);
1830 1831 1832 1833 1834
		len -= l;
		if (len == 0)
			break;
		index++;
		BUG_ON(index >= sblock->page_count);
1835 1836
		BUG_ON(!sblock->pagev[index]->page);
		page = sblock->pagev[index]->page;
1837
		mapped_buffer = kmap_atomic(page);
1838 1839 1840 1841 1842
		mapped_size = PAGE_SIZE;
		p = mapped_buffer;
	}

	btrfs_csum_final(crc, calculated_csum);
1843
	if (memcmp(calculated_csum, on_disk_csum, sctx->csum_size))
1844
		++fail_cor;
A
Arne Jansen 已提交
1845

1846
	if (fail_cor + fail_gen) {
A
Arne Jansen 已提交
1847 1848 1849 1850 1851
		/*
		 * if we find an error in a super block, we just report it.
		 * They will get written with the next transaction commit
		 * anyway
		 */
1852 1853 1854
		spin_lock(&sctx->stat_lock);
		++sctx->stat.super_errors;
		spin_unlock(&sctx->stat_lock);
1855
		if (fail_cor)
1856
			btrfs_dev_stat_inc_and_print(sblock->pagev[0]->dev,
1857 1858
				BTRFS_DEV_STAT_CORRUPTION_ERRS);
		else
1859
			btrfs_dev_stat_inc_and_print(sblock->pagev[0]->dev,
1860
				BTRFS_DEV_STAT_GENERATION_ERRS);
A
Arne Jansen 已提交
1861 1862
	}

1863
	return fail_cor + fail_gen;
A
Arne Jansen 已提交
1864 1865
}

1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876
static void scrub_block_get(struct scrub_block *sblock)
{
	atomic_inc(&sblock->ref_count);
}

static void scrub_block_put(struct scrub_block *sblock)
{
	if (atomic_dec_and_test(&sblock->ref_count)) {
		int i;

		for (i = 0; i < sblock->page_count; i++)
1877
			scrub_page_put(sblock->pagev[i]);
1878 1879 1880 1881
		kfree(sblock);
	}
}

1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895
static void scrub_page_get(struct scrub_page *spage)
{
	atomic_inc(&spage->ref_count);
}

static void scrub_page_put(struct scrub_page *spage)
{
	if (atomic_dec_and_test(&spage->ref_count)) {
		if (spage->page)
			__free_page(spage->page);
		kfree(spage);
	}
}

1896
static void scrub_submit(struct scrub_ctx *sctx)
A
Arne Jansen 已提交
1897 1898 1899
{
	struct scrub_bio *sbio;

1900
	if (sctx->curr == -1)
S
Stefan Behrens 已提交
1901
		return;
A
Arne Jansen 已提交
1902

1903 1904
	sbio = sctx->bios[sctx->curr];
	sctx->curr = -1;
1905
	scrub_pending_bio_inc(sctx);
A
Arne Jansen 已提交
1906

1907 1908 1909 1910 1911 1912 1913 1914 1915
	if (!sbio->bio->bi_bdev) {
		/*
		 * this case should not happen. If btrfs_map_block() is
		 * wrong, it could happen for dev-replace operations on
		 * missing devices when no mirrors are available, but in
		 * this case it should already fail the mount.
		 * This case is handled correctly (but _very_ slowly).
		 */
		printk_ratelimited(KERN_WARNING
1916
			"BTRFS: scrub_submit(bio bdev == NULL) is unexpected!\n");
1917 1918 1919 1920
		bio_endio(sbio->bio, -EIO);
	} else {
		btrfsic_submit_bio(READ, sbio->bio);
	}
A
Arne Jansen 已提交
1921 1922
}

1923 1924
static int scrub_add_page_to_rd_bio(struct scrub_ctx *sctx,
				    struct scrub_page *spage)
A
Arne Jansen 已提交
1925
{
1926
	struct scrub_block *sblock = spage->sblock;
A
Arne Jansen 已提交
1927
	struct scrub_bio *sbio;
1928
	int ret;
A
Arne Jansen 已提交
1929 1930 1931 1932 1933

again:
	/*
	 * grab a fresh bio or wait for one to become available
	 */
1934 1935 1936 1937 1938 1939 1940 1941
	while (sctx->curr == -1) {
		spin_lock(&sctx->list_lock);
		sctx->curr = sctx->first_free;
		if (sctx->curr != -1) {
			sctx->first_free = sctx->bios[sctx->curr]->next_free;
			sctx->bios[sctx->curr]->next_free = -1;
			sctx->bios[sctx->curr]->page_count = 0;
			spin_unlock(&sctx->list_lock);
A
Arne Jansen 已提交
1942
		} else {
1943 1944
			spin_unlock(&sctx->list_lock);
			wait_event(sctx->list_wait, sctx->first_free != -1);
A
Arne Jansen 已提交
1945 1946
		}
	}
1947
	sbio = sctx->bios[sctx->curr];
1948
	if (sbio->page_count == 0) {
1949 1950
		struct bio *bio;

1951 1952
		sbio->physical = spage->physical;
		sbio->logical = spage->logical;
1953
		sbio->dev = spage->dev;
1954 1955
		bio = sbio->bio;
		if (!bio) {
1956
			bio = btrfs_io_bio_alloc(GFP_NOFS, sctx->pages_per_rd_bio);
1957 1958 1959 1960
			if (!bio)
				return -ENOMEM;
			sbio->bio = bio;
		}
1961 1962 1963

		bio->bi_private = sbio;
		bio->bi_end_io = scrub_bio_end_io;
1964 1965
		bio->bi_bdev = sbio->dev->bdev;
		bio->bi_sector = sbio->physical >> 9;
1966
		sbio->err = 0;
1967 1968 1969
	} else if (sbio->physical + sbio->page_count * PAGE_SIZE !=
		   spage->physical ||
		   sbio->logical + sbio->page_count * PAGE_SIZE !=
1970 1971
		   spage->logical ||
		   sbio->dev != spage->dev) {
1972
		scrub_submit(sctx);
A
Arne Jansen 已提交
1973 1974
		goto again;
	}
1975

1976 1977 1978 1979 1980 1981 1982 1983
	sbio->pagev[sbio->page_count] = spage;
	ret = bio_add_page(sbio->bio, spage->page, PAGE_SIZE, 0);
	if (ret != PAGE_SIZE) {
		if (sbio->page_count < 1) {
			bio_put(sbio->bio);
			sbio->bio = NULL;
			return -EIO;
		}
1984
		scrub_submit(sctx);
1985 1986 1987
		goto again;
	}

1988
	scrub_block_get(sblock); /* one for the page added to the bio */
1989 1990
	atomic_inc(&sblock->outstanding_pages);
	sbio->page_count++;
1991
	if (sbio->page_count == sctx->pages_per_rd_bio)
1992
		scrub_submit(sctx);
1993 1994 1995 1996

	return 0;
}

1997
static int scrub_pages(struct scrub_ctx *sctx, u64 logical, u64 len,
1998
		       u64 physical, struct btrfs_device *dev, u64 flags,
1999 2000
		       u64 gen, int mirror_num, u8 *csum, int force,
		       u64 physical_for_dev_replace)
2001 2002 2003 2004 2005 2006
{
	struct scrub_block *sblock;
	int index;

	sblock = kzalloc(sizeof(*sblock), GFP_NOFS);
	if (!sblock) {
2007 2008 2009
		spin_lock(&sctx->stat_lock);
		sctx->stat.malloc_errors++;
		spin_unlock(&sctx->stat_lock);
2010
		return -ENOMEM;
A
Arne Jansen 已提交
2011
	}
2012

2013 2014
	/* one ref inside this function, plus one for each page added to
	 * a bio later on */
2015
	atomic_set(&sblock->ref_count, 1);
2016
	sblock->sctx = sctx;
2017 2018 2019
	sblock->no_io_error_seen = 1;

	for (index = 0; len > 0; index++) {
2020
		struct scrub_page *spage;
2021 2022
		u64 l = min_t(u64, len, PAGE_SIZE);

2023 2024 2025
		spage = kzalloc(sizeof(*spage), GFP_NOFS);
		if (!spage) {
leave_nomem:
2026 2027 2028
			spin_lock(&sctx->stat_lock);
			sctx->stat.malloc_errors++;
			spin_unlock(&sctx->stat_lock);
2029
			scrub_block_put(sblock);
2030 2031
			return -ENOMEM;
		}
2032 2033 2034
		BUG_ON(index >= SCRUB_MAX_PAGES_PER_BLOCK);
		scrub_page_get(spage);
		sblock->pagev[index] = spage;
2035
		spage->sblock = sblock;
2036
		spage->dev = dev;
2037 2038 2039 2040
		spage->flags = flags;
		spage->generation = gen;
		spage->logical = logical;
		spage->physical = physical;
2041
		spage->physical_for_dev_replace = physical_for_dev_replace;
2042 2043 2044
		spage->mirror_num = mirror_num;
		if (csum) {
			spage->have_csum = 1;
2045
			memcpy(spage->csum, csum, sctx->csum_size);
2046 2047 2048 2049
		} else {
			spage->have_csum = 0;
		}
		sblock->page_count++;
2050 2051 2052
		spage->page = alloc_page(GFP_NOFS);
		if (!spage->page)
			goto leave_nomem;
2053 2054 2055
		len -= l;
		logical += l;
		physical += l;
2056
		physical_for_dev_replace += l;
2057 2058
	}

2059
	WARN_ON(sblock->page_count == 0);
2060
	for (index = 0; index < sblock->page_count; index++) {
2061
		struct scrub_page *spage = sblock->pagev[index];
2062 2063
		int ret;

2064
		ret = scrub_add_page_to_rd_bio(sctx, spage);
2065 2066
		if (ret) {
			scrub_block_put(sblock);
2067
			return ret;
2068
		}
2069
	}
A
Arne Jansen 已提交
2070

2071
	if (force)
2072
		scrub_submit(sctx);
A
Arne Jansen 已提交
2073

2074 2075
	/* last one frees, either here or in bio completion for last page */
	scrub_block_put(sblock);
A
Arne Jansen 已提交
2076 2077 2078
	return 0;
}

2079 2080 2081
static void scrub_bio_end_io(struct bio *bio, int err)
{
	struct scrub_bio *sbio = bio->bi_private;
2082
	struct btrfs_fs_info *fs_info = sbio->dev->dev_root->fs_info;
2083 2084 2085 2086

	sbio->err = err;
	sbio->bio = bio;

2087
	btrfs_queue_work(fs_info->scrub_workers, &sbio->work);
2088 2089
}

2090
static void scrub_bio_end_io_worker(struct btrfs_work *work)
2091 2092
{
	struct scrub_bio *sbio = container_of(work, struct scrub_bio, work);
2093
	struct scrub_ctx *sctx = sbio->sctx;
2094 2095
	int i;

2096
	BUG_ON(sbio->page_count > SCRUB_PAGES_PER_RD_BIO);
2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117
	if (sbio->err) {
		for (i = 0; i < sbio->page_count; i++) {
			struct scrub_page *spage = sbio->pagev[i];

			spage->io_error = 1;
			spage->sblock->no_io_error_seen = 0;
		}
	}

	/* now complete the scrub_block items that have all pages completed */
	for (i = 0; i < sbio->page_count; i++) {
		struct scrub_page *spage = sbio->pagev[i];
		struct scrub_block *sblock = spage->sblock;

		if (atomic_dec_and_test(&sblock->outstanding_pages))
			scrub_block_complete(sblock);
		scrub_block_put(sblock);
	}

	bio_put(sbio->bio);
	sbio->bio = NULL;
2118 2119 2120 2121
	spin_lock(&sctx->list_lock);
	sbio->next_free = sctx->first_free;
	sctx->first_free = sbio->index;
	spin_unlock(&sctx->list_lock);
2122 2123 2124 2125 2126 2127 2128 2129

	if (sctx->is_dev_replace &&
	    atomic_read(&sctx->wr_ctx.flush_all_writes)) {
		mutex_lock(&sctx->wr_ctx.wr_lock);
		scrub_wr_submit(sctx);
		mutex_unlock(&sctx->wr_ctx.wr_lock);
	}

2130
	scrub_pending_bio_dec(sctx);
2131 2132 2133 2134
}

static void scrub_block_complete(struct scrub_block *sblock)
{
2135
	if (!sblock->no_io_error_seen) {
2136
		scrub_handle_errored_block(sblock);
2137 2138 2139 2140 2141 2142 2143 2144 2145
	} else {
		/*
		 * if has checksum error, write via repair mechanism in
		 * dev replace case, otherwise write here in dev replace
		 * case.
		 */
		if (!scrub_checksum(sblock) && sblock->sctx->is_dev_replace)
			scrub_write_block_to_dev_replace(sblock);
	}
2146 2147
}

2148
static int scrub_find_csum(struct scrub_ctx *sctx, u64 logical, u64 len,
A
Arne Jansen 已提交
2149 2150 2151
			   u8 *csum)
{
	struct btrfs_ordered_sum *sum = NULL;
2152
	unsigned long index;
A
Arne Jansen 已提交
2153 2154
	unsigned long num_sectors;

2155 2156
	while (!list_empty(&sctx->csum_list)) {
		sum = list_first_entry(&sctx->csum_list,
A
Arne Jansen 已提交
2157 2158 2159 2160 2161 2162
				       struct btrfs_ordered_sum, list);
		if (sum->bytenr > logical)
			return 0;
		if (sum->bytenr + sum->len > logical)
			break;

2163
		++sctx->stat.csum_discards;
A
Arne Jansen 已提交
2164 2165 2166 2167 2168 2169 2170
		list_del(&sum->list);
		kfree(sum);
		sum = NULL;
	}
	if (!sum)
		return 0;

2171
	index = ((u32)(logical - sum->bytenr)) / sctx->sectorsize;
2172
	num_sectors = sum->len / sctx->sectorsize;
2173 2174
	memcpy(csum, sum->sums + index, sctx->csum_size);
	if (index == num_sectors - 1) {
A
Arne Jansen 已提交
2175 2176 2177
		list_del(&sum->list);
		kfree(sum);
	}
2178
	return 1;
A
Arne Jansen 已提交
2179 2180 2181
}

/* scrub extent tries to collect up to 64 kB for each bio */
2182
static int scrub_extent(struct scrub_ctx *sctx, u64 logical, u64 len,
2183
			u64 physical, struct btrfs_device *dev, u64 flags,
2184
			u64 gen, int mirror_num, u64 physical_for_dev_replace)
A
Arne Jansen 已提交
2185 2186 2187
{
	int ret;
	u8 csum[BTRFS_CSUM_SIZE];
2188 2189 2190
	u32 blocksize;

	if (flags & BTRFS_EXTENT_FLAG_DATA) {
2191 2192 2193 2194 2195
		blocksize = sctx->sectorsize;
		spin_lock(&sctx->stat_lock);
		sctx->stat.data_extents_scrubbed++;
		sctx->stat.data_bytes_scrubbed += len;
		spin_unlock(&sctx->stat_lock);
2196
	} else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
2197
		WARN_ON(sctx->nodesize != sctx->leafsize);
2198 2199 2200 2201 2202
		blocksize = sctx->nodesize;
		spin_lock(&sctx->stat_lock);
		sctx->stat.tree_extents_scrubbed++;
		sctx->stat.tree_bytes_scrubbed += len;
		spin_unlock(&sctx->stat_lock);
2203
	} else {
2204
		blocksize = sctx->sectorsize;
2205
		WARN_ON(1);
2206
	}
A
Arne Jansen 已提交
2207 2208

	while (len) {
2209
		u64 l = min_t(u64, len, blocksize);
A
Arne Jansen 已提交
2210 2211 2212 2213
		int have_csum = 0;

		if (flags & BTRFS_EXTENT_FLAG_DATA) {
			/* push csums to sbio */
2214
			have_csum = scrub_find_csum(sctx, logical, l, csum);
A
Arne Jansen 已提交
2215
			if (have_csum == 0)
2216
				++sctx->stat.no_csum;
2217 2218 2219 2220 2221 2222
			if (sctx->is_dev_replace && !have_csum) {
				ret = copy_nocow_pages(sctx, logical, l,
						       mirror_num,
						      physical_for_dev_replace);
				goto behind_scrub_pages;
			}
A
Arne Jansen 已提交
2223
		}
2224
		ret = scrub_pages(sctx, logical, l, physical, dev, flags, gen,
2225 2226 2227
				  mirror_num, have_csum ? csum : NULL, 0,
				  physical_for_dev_replace);
behind_scrub_pages:
A
Arne Jansen 已提交
2228 2229 2230 2231 2232
		if (ret)
			return ret;
		len -= l;
		logical += l;
		physical += l;
2233
		physical_for_dev_replace += l;
A
Arne Jansen 已提交
2234 2235 2236 2237
	}
	return 0;
}

2238
static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
2239 2240
					   struct map_lookup *map,
					   struct btrfs_device *scrub_dev,
2241 2242
					   int num, u64 base, u64 length,
					   int is_dev_replace)
A
Arne Jansen 已提交
2243 2244
{
	struct btrfs_path *path;
2245
	struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info;
A
Arne Jansen 已提交
2246 2247 2248
	struct btrfs_root *root = fs_info->extent_root;
	struct btrfs_root *csum_root = fs_info->csum_root;
	struct btrfs_extent_item *extent;
2249
	struct blk_plug plug;
A
Arne Jansen 已提交
2250 2251 2252 2253 2254 2255 2256 2257
	u64 flags;
	int ret;
	int slot;
	u64 nstripes;
	struct extent_buffer *l;
	struct btrfs_key key;
	u64 physical;
	u64 logical;
L
Liu Bo 已提交
2258
	u64 logic_end;
A
Arne Jansen 已提交
2259
	u64 generation;
2260
	int mirror_num;
A
Arne Jansen 已提交
2261 2262 2263 2264
	struct reada_control *reada1;
	struct reada_control *reada2;
	struct btrfs_key key_start;
	struct btrfs_key key_end;
A
Arne Jansen 已提交
2265 2266
	u64 increment = map->stripe_len;
	u64 offset;
2267 2268 2269 2270 2271
	u64 extent_logical;
	u64 extent_physical;
	u64 extent_len;
	struct btrfs_device *extent_dev;
	int extent_mirror_num;
L
Liu Bo 已提交
2272
	int stop_loop;
A
Arne Jansen 已提交
2273

D
David Woodhouse 已提交
2274 2275 2276 2277 2278 2279 2280
	if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
			 BTRFS_BLOCK_GROUP_RAID6)) {
		if (num >= nr_data_stripes(map)) {
			return 0;
		}
	}

A
Arne Jansen 已提交
2281 2282 2283 2284 2285 2286
	nstripes = length;
	offset = 0;
	do_div(nstripes, map->stripe_len);
	if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
		offset = map->stripe_len * num;
		increment = map->stripe_len * map->num_stripes;
2287
		mirror_num = 1;
A
Arne Jansen 已提交
2288 2289 2290 2291
	} else if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
		int factor = map->num_stripes / map->sub_stripes;
		offset = map->stripe_len * (num / map->sub_stripes);
		increment = map->stripe_len * factor;
2292
		mirror_num = num % map->sub_stripes + 1;
A
Arne Jansen 已提交
2293 2294
	} else if (map->type & BTRFS_BLOCK_GROUP_RAID1) {
		increment = map->stripe_len;
2295
		mirror_num = num % map->num_stripes + 1;
A
Arne Jansen 已提交
2296 2297
	} else if (map->type & BTRFS_BLOCK_GROUP_DUP) {
		increment = map->stripe_len;
2298
		mirror_num = num % map->num_stripes + 1;
A
Arne Jansen 已提交
2299 2300
	} else {
		increment = map->stripe_len;
2301
		mirror_num = 1;
A
Arne Jansen 已提交
2302 2303 2304 2305 2306 2307
	}

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

2308 2309 2310 2311 2312
	/*
	 * work on commit root. The related disk blocks are static as
	 * long as COW is applied. This means, it is save to rewrite
	 * them to repair disk errors without any race conditions
	 */
A
Arne Jansen 已提交
2313 2314 2315 2316
	path->search_commit_root = 1;
	path->skip_locking = 1;

	/*
A
Arne Jansen 已提交
2317 2318 2319
	 * trigger the readahead for extent tree csum tree and wait for
	 * completion. During readahead, the scrub is officially paused
	 * to not hold off transaction commits
A
Arne Jansen 已提交
2320 2321 2322
	 */
	logical = base + offset;

2323
	wait_event(sctx->list_wait,
2324
		   atomic_read(&sctx->bios_in_flight) == 0);
2325
	scrub_blocked_if_needed(fs_info);
A
Arne Jansen 已提交
2326 2327 2328 2329 2330 2331

	/* FIXME it might be better to start readahead at commit root */
	key_start.objectid = logical;
	key_start.type = BTRFS_EXTENT_ITEM_KEY;
	key_start.offset = (u64)0;
	key_end.objectid = base + offset + nstripes * increment;
2332 2333
	key_end.type = BTRFS_METADATA_ITEM_KEY;
	key_end.offset = (u64)-1;
A
Arne Jansen 已提交
2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348
	reada1 = btrfs_reada_add(root, &key_start, &key_end);

	key_start.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
	key_start.type = BTRFS_EXTENT_CSUM_KEY;
	key_start.offset = logical;
	key_end.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
	key_end.type = BTRFS_EXTENT_CSUM_KEY;
	key_end.offset = base + offset + nstripes * increment;
	reada2 = btrfs_reada_add(csum_root, &key_start, &key_end);

	if (!IS_ERR(reada1))
		btrfs_reada_wait(reada1);
	if (!IS_ERR(reada2))
		btrfs_reada_wait(reada2);

A
Arne Jansen 已提交
2349 2350 2351 2352 2353

	/*
	 * collect all data csums for the stripe to avoid seeking during
	 * the scrub. This might currently (crc32) end up to be about 1MB
	 */
2354
	blk_start_plug(&plug);
A
Arne Jansen 已提交
2355 2356 2357 2358

	/*
	 * now find all extents for each stripe and scrub them
	 */
A
Arne Jansen 已提交
2359 2360
	logical = base + offset;
	physical = map->stripes[num].physical;
L
Liu Bo 已提交
2361
	logic_end = logical + increment * nstripes;
A
Arne Jansen 已提交
2362
	ret = 0;
L
Liu Bo 已提交
2363
	while (logical < logic_end) {
A
Arne Jansen 已提交
2364 2365 2366 2367
		/*
		 * canceled?
		 */
		if (atomic_read(&fs_info->scrub_cancel_req) ||
2368
		    atomic_read(&sctx->cancel_req)) {
A
Arne Jansen 已提交
2369 2370 2371 2372 2373 2374 2375 2376
			ret = -ECANCELED;
			goto out;
		}
		/*
		 * check to see if we have to pause
		 */
		if (atomic_read(&fs_info->scrub_pause_req)) {
			/* push queued extents */
2377
			atomic_set(&sctx->wr_ctx.flush_all_writes, 1);
2378
			scrub_submit(sctx);
2379 2380 2381
			mutex_lock(&sctx->wr_ctx.wr_lock);
			scrub_wr_submit(sctx);
			mutex_unlock(&sctx->wr_ctx.wr_lock);
2382
			wait_event(sctx->list_wait,
2383
				   atomic_read(&sctx->bios_in_flight) == 0);
2384
			atomic_set(&sctx->wr_ctx.flush_all_writes, 0);
2385
			scrub_blocked_if_needed(fs_info);
A
Arne Jansen 已提交
2386 2387
		}

2388 2389 2390 2391
		if (btrfs_fs_incompat(fs_info, SKINNY_METADATA))
			key.type = BTRFS_METADATA_ITEM_KEY;
		else
			key.type = BTRFS_EXTENT_ITEM_KEY;
A
Arne Jansen 已提交
2392
		key.objectid = logical;
L
Liu Bo 已提交
2393
		key.offset = (u64)-1;
A
Arne Jansen 已提交
2394 2395 2396 2397

		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0)
			goto out;
2398

2399
		if (ret > 0) {
2400
			ret = btrfs_previous_extent_item(root, path, 0);
A
Arne Jansen 已提交
2401 2402
			if (ret < 0)
				goto out;
2403 2404 2405 2406 2407 2408 2409 2410 2411
			if (ret > 0) {
				/* there's no smaller item, so stick with the
				 * larger one */
				btrfs_release_path(path);
				ret = btrfs_search_slot(NULL, root, &key,
							path, 0, 0);
				if (ret < 0)
					goto out;
			}
A
Arne Jansen 已提交
2412 2413
		}

L
Liu Bo 已提交
2414
		stop_loop = 0;
A
Arne Jansen 已提交
2415
		while (1) {
2416 2417
			u64 bytes;

A
Arne Jansen 已提交
2418 2419 2420 2421 2422 2423 2424 2425 2426
			l = path->nodes[0];
			slot = path->slots[0];
			if (slot >= btrfs_header_nritems(l)) {
				ret = btrfs_next_leaf(root, path);
				if (ret == 0)
					continue;
				if (ret < 0)
					goto out;

L
Liu Bo 已提交
2427
				stop_loop = 1;
A
Arne Jansen 已提交
2428 2429 2430 2431
				break;
			}
			btrfs_item_key_to_cpu(l, &key, slot);

2432 2433 2434 2435 2436 2437
			if (key.type == BTRFS_METADATA_ITEM_KEY)
				bytes = root->leafsize;
			else
				bytes = key.offset;

			if (key.objectid + bytes <= logical)
A
Arne Jansen 已提交
2438 2439
				goto next;

L
Liu Bo 已提交
2440 2441 2442
			if (key.type != BTRFS_EXTENT_ITEM_KEY &&
			    key.type != BTRFS_METADATA_ITEM_KEY)
				goto next;
A
Arne Jansen 已提交
2443

L
Liu Bo 已提交
2444 2445 2446 2447 2448 2449
			if (key.objectid >= logical + map->stripe_len) {
				/* out of this device extent */
				if (key.objectid >= logic_end)
					stop_loop = 1;
				break;
			}
A
Arne Jansen 已提交
2450 2451 2452 2453 2454 2455 2456 2457

			extent = btrfs_item_ptr(l, slot,
						struct btrfs_extent_item);
			flags = btrfs_extent_flags(l, extent);
			generation = btrfs_extent_generation(l, extent);

			if (key.objectid < logical &&
			    (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)) {
2458 2459 2460
				btrfs_err(fs_info,
					   "scrub: tree block %llu spanning "
					   "stripes, ignored. logical=%llu",
2461
				       key.objectid, logical);
A
Arne Jansen 已提交
2462 2463 2464
				goto next;
			}

L
Liu Bo 已提交
2465 2466 2467 2468
again:
			extent_logical = key.objectid;
			extent_len = bytes;

A
Arne Jansen 已提交
2469 2470 2471
			/*
			 * trim extent to this stripe
			 */
L
Liu Bo 已提交
2472 2473 2474
			if (extent_logical < logical) {
				extent_len -= logical - extent_logical;
				extent_logical = logical;
A
Arne Jansen 已提交
2475
			}
L
Liu Bo 已提交
2476
			if (extent_logical + extent_len >
A
Arne Jansen 已提交
2477
			    logical + map->stripe_len) {
L
Liu Bo 已提交
2478 2479
				extent_len = logical + map->stripe_len -
					     extent_logical;
A
Arne Jansen 已提交
2480 2481
			}

L
Liu Bo 已提交
2482
			extent_physical = extent_logical - logical + physical;
2483 2484 2485 2486 2487 2488 2489
			extent_dev = scrub_dev;
			extent_mirror_num = mirror_num;
			if (is_dev_replace)
				scrub_remap_extent(fs_info, extent_logical,
						   extent_len, &extent_physical,
						   &extent_dev,
						   &extent_mirror_num);
L
Liu Bo 已提交
2490 2491 2492 2493 2494 2495 2496

			ret = btrfs_lookup_csums_range(csum_root, logical,
						logical + map->stripe_len - 1,
						&sctx->csum_list, 1);
			if (ret)
				goto out;

2497 2498 2499
			ret = scrub_extent(sctx, extent_logical, extent_len,
					   extent_physical, extent_dev, flags,
					   generation, extent_mirror_num,
2500
					   extent_logical - logical + physical);
A
Arne Jansen 已提交
2501 2502 2503
			if (ret)
				goto out;

2504
			scrub_free_csums(sctx);
L
Liu Bo 已提交
2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519
			if (extent_logical + extent_len <
			    key.objectid + bytes) {
				logical += increment;
				physical += map->stripe_len;

				if (logical < key.objectid + bytes) {
					cond_resched();
					goto again;
				}

				if (logical >= logic_end) {
					stop_loop = 1;
					break;
				}
			}
A
Arne Jansen 已提交
2520 2521 2522
next:
			path->slots[0]++;
		}
C
Chris Mason 已提交
2523
		btrfs_release_path(path);
A
Arne Jansen 已提交
2524 2525
		logical += increment;
		physical += map->stripe_len;
2526
		spin_lock(&sctx->stat_lock);
L
Liu Bo 已提交
2527 2528 2529 2530 2531
		if (stop_loop)
			sctx->stat.last_physical = map->stripes[num].physical +
						   length;
		else
			sctx->stat.last_physical = physical;
2532
		spin_unlock(&sctx->stat_lock);
L
Liu Bo 已提交
2533 2534
		if (stop_loop)
			break;
A
Arne Jansen 已提交
2535
	}
2536
out:
A
Arne Jansen 已提交
2537
	/* push queued extents */
2538
	scrub_submit(sctx);
2539 2540 2541
	mutex_lock(&sctx->wr_ctx.wr_lock);
	scrub_wr_submit(sctx);
	mutex_unlock(&sctx->wr_ctx.wr_lock);
A
Arne Jansen 已提交
2542

2543
	blk_finish_plug(&plug);
A
Arne Jansen 已提交
2544 2545 2546 2547
	btrfs_free_path(path);
	return ret < 0 ? ret : 0;
}

2548
static noinline_for_stack int scrub_chunk(struct scrub_ctx *sctx,
2549 2550 2551
					  struct btrfs_device *scrub_dev,
					  u64 chunk_tree, u64 chunk_objectid,
					  u64 chunk_offset, u64 length,
2552
					  u64 dev_offset, int is_dev_replace)
A
Arne Jansen 已提交
2553 2554
{
	struct btrfs_mapping_tree *map_tree =
2555
		&sctx->dev_root->fs_info->mapping_tree;
A
Arne Jansen 已提交
2556 2557 2558
	struct map_lookup *map;
	struct extent_map *em;
	int i;
2559
	int ret = 0;
A
Arne Jansen 已提交
2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575

	read_lock(&map_tree->map_tree.lock);
	em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1);
	read_unlock(&map_tree->map_tree.lock);

	if (!em)
		return -EINVAL;

	map = (struct map_lookup *)em->bdev;
	if (em->start != chunk_offset)
		goto out;

	if (em->len < length)
		goto out;

	for (i = 0; i < map->num_stripes; ++i) {
2576
		if (map->stripes[i].dev->bdev == scrub_dev->bdev &&
2577
		    map->stripes[i].physical == dev_offset) {
2578
			ret = scrub_stripe(sctx, map, scrub_dev, i,
2579 2580
					   chunk_offset, length,
					   is_dev_replace);
A
Arne Jansen 已提交
2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591
			if (ret)
				goto out;
		}
	}
out:
	free_extent_map(em);

	return ret;
}

static noinline_for_stack
2592
int scrub_enumerate_chunks(struct scrub_ctx *sctx,
2593 2594
			   struct btrfs_device *scrub_dev, u64 start, u64 end,
			   int is_dev_replace)
A
Arne Jansen 已提交
2595 2596 2597
{
	struct btrfs_dev_extent *dev_extent = NULL;
	struct btrfs_path *path;
2598
	struct btrfs_root *root = sctx->dev_root;
A
Arne Jansen 已提交
2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609
	struct btrfs_fs_info *fs_info = root->fs_info;
	u64 length;
	u64 chunk_tree;
	u64 chunk_objectid;
	u64 chunk_offset;
	int ret;
	int slot;
	struct extent_buffer *l;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_block_group_cache *cache;
2610
	struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace;
A
Arne Jansen 已提交
2611 2612 2613 2614 2615 2616 2617 2618 2619

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	path->reada = 2;
	path->search_commit_root = 1;
	path->skip_locking = 1;

2620
	key.objectid = scrub_dev->devid;
A
Arne Jansen 已提交
2621 2622 2623 2624 2625 2626
	key.offset = 0ull;
	key.type = BTRFS_DEV_EXTENT_KEY;

	while (1) {
		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0)
2627 2628 2629 2630 2631 2632 2633 2634 2635
			break;
		if (ret > 0) {
			if (path->slots[0] >=
			    btrfs_header_nritems(path->nodes[0])) {
				ret = btrfs_next_leaf(root, path);
				if (ret)
					break;
			}
		}
A
Arne Jansen 已提交
2636 2637 2638 2639 2640 2641

		l = path->nodes[0];
		slot = path->slots[0];

		btrfs_item_key_to_cpu(l, &found_key, slot);

2642
		if (found_key.objectid != scrub_dev->devid)
A
Arne Jansen 已提交
2643 2644
			break;

2645
		if (btrfs_key_type(&found_key) != BTRFS_DEV_EXTENT_KEY)
A
Arne Jansen 已提交
2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658
			break;

		if (found_key.offset >= end)
			break;

		if (found_key.offset < key.offset)
			break;

		dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
		length = btrfs_dev_extent_length(l, dev_extent);

		if (found_key.offset + length <= start) {
			key.offset = found_key.offset + length;
C
Chris Mason 已提交
2659
			btrfs_release_path(path);
A
Arne Jansen 已提交
2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673
			continue;
		}

		chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent);
		chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent);
		chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent);

		/*
		 * get a reference on the corresponding block group to prevent
		 * the chunk from going away while we scrub it
		 */
		cache = btrfs_lookup_block_group(fs_info, chunk_offset);
		if (!cache) {
			ret = -ENOENT;
2674
			break;
A
Arne Jansen 已提交
2675
		}
2676 2677 2678
		dev_replace->cursor_right = found_key.offset + length;
		dev_replace->cursor_left = found_key.offset;
		dev_replace->item_needs_writeback = 1;
2679
		ret = scrub_chunk(sctx, scrub_dev, chunk_tree, chunk_objectid,
2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700
				  chunk_offset, length, found_key.offset,
				  is_dev_replace);

		/*
		 * flush, submit all pending read and write bios, afterwards
		 * wait for them.
		 * Note that in the dev replace case, a read request causes
		 * write requests that are submitted in the read completion
		 * worker. Therefore in the current situation, it is required
		 * that all write requests are flushed, so that all read and
		 * write requests are really completed when bios_in_flight
		 * changes to 0.
		 */
		atomic_set(&sctx->wr_ctx.flush_all_writes, 1);
		scrub_submit(sctx);
		mutex_lock(&sctx->wr_ctx.wr_lock);
		scrub_wr_submit(sctx);
		mutex_unlock(&sctx->wr_ctx.wr_lock);

		wait_event(sctx->list_wait,
			   atomic_read(&sctx->bios_in_flight) == 0);
2701 2702 2703 2704 2705 2706 2707 2708
		atomic_inc(&fs_info->scrubs_paused);
		wake_up(&fs_info->scrub_pause_wait);

		/*
		 * must be called before we decrease @scrub_paused.
		 * make sure we don't block transaction commit while
		 * we are waiting pending workers finished.
		 */
2709 2710
		wait_event(sctx->list_wait,
			   atomic_read(&sctx->workers_pending) == 0);
2711 2712 2713 2714 2715 2716 2717
		atomic_set(&sctx->wr_ctx.flush_all_writes, 0);

		mutex_lock(&fs_info->scrub_lock);
		__scrub_blocked_if_needed(fs_info);
		atomic_dec(&fs_info->scrubs_paused);
		mutex_unlock(&fs_info->scrub_lock);
		wake_up(&fs_info->scrub_pause_wait);
2718

A
Arne Jansen 已提交
2719 2720 2721
		btrfs_put_block_group(cache);
		if (ret)
			break;
2722 2723
		if (is_dev_replace &&
		    atomic64_read(&dev_replace->num_write_errors) > 0) {
2724 2725 2726 2727 2728 2729 2730
			ret = -EIO;
			break;
		}
		if (sctx->stat.malloc_errors > 0) {
			ret = -ENOMEM;
			break;
		}
A
Arne Jansen 已提交
2731

2732 2733 2734
		dev_replace->cursor_left = dev_replace->cursor_right;
		dev_replace->item_needs_writeback = 1;

A
Arne Jansen 已提交
2735
		key.offset = found_key.offset + length;
C
Chris Mason 已提交
2736
		btrfs_release_path(path);
A
Arne Jansen 已提交
2737 2738 2739
	}

	btrfs_free_path(path);
2740 2741 2742 2743 2744 2745

	/*
	 * ret can still be 1 from search_slot or next_leaf,
	 * that's not an error
	 */
	return ret < 0 ? ret : 0;
A
Arne Jansen 已提交
2746 2747
}

2748 2749
static noinline_for_stack int scrub_supers(struct scrub_ctx *sctx,
					   struct btrfs_device *scrub_dev)
A
Arne Jansen 已提交
2750 2751 2752 2753 2754
{
	int	i;
	u64	bytenr;
	u64	gen;
	int	ret;
2755
	struct btrfs_root *root = sctx->dev_root;
A
Arne Jansen 已提交
2756

2757
	if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state))
2758 2759
		return -EIO;

A
Arne Jansen 已提交
2760 2761 2762 2763
	gen = root->fs_info->last_trans_committed;

	for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
		bytenr = btrfs_sb_offset(i);
2764
		if (bytenr + BTRFS_SUPER_INFO_SIZE > scrub_dev->total_bytes)
A
Arne Jansen 已提交
2765 2766
			break;

2767
		ret = scrub_pages(sctx, bytenr, BTRFS_SUPER_INFO_SIZE, bytenr,
2768
				  scrub_dev, BTRFS_EXTENT_FLAG_SUPER, gen, i,
2769
				  NULL, 1, bytenr);
A
Arne Jansen 已提交
2770 2771 2772
		if (ret)
			return ret;
	}
2773
	wait_event(sctx->list_wait, atomic_read(&sctx->bios_in_flight) == 0);
A
Arne Jansen 已提交
2774 2775 2776 2777 2778 2779 2780

	return 0;
}

/*
 * get a reference count on fs_info->scrub_workers. start worker if necessary
 */
2781 2782
static noinline_for_stack int scrub_workers_get(struct btrfs_fs_info *fs_info,
						int is_dev_replace)
A
Arne Jansen 已提交
2783
{
2784
	int ret = 0;
2785 2786
	int flags = WQ_FREEZABLE | WQ_UNBOUND;
	int max_active = fs_info->thread_pool_size;
A
Arne Jansen 已提交
2787

A
Arne Jansen 已提交
2788
	if (fs_info->scrub_workers_refcnt == 0) {
2789
		if (is_dev_replace)
2790 2791 2792
			fs_info->scrub_workers =
				btrfs_alloc_workqueue("btrfs-scrub", flags,
						      1, 4);
2793
		else
2794 2795 2796 2797 2798
			fs_info->scrub_workers =
				btrfs_alloc_workqueue("btrfs-scrub", flags,
						      max_active, 4);
		if (!fs_info->scrub_workers) {
			ret = -ENOMEM;
2799
			goto out;
2800 2801 2802 2803 2804 2805
		}
		fs_info->scrub_wr_completion_workers =
			btrfs_alloc_workqueue("btrfs-scrubwrc", flags,
					      max_active, 2);
		if (!fs_info->scrub_wr_completion_workers) {
			ret = -ENOMEM;
2806
			goto out;
2807 2808 2809 2810 2811
		}
		fs_info->scrub_nocow_workers =
			btrfs_alloc_workqueue("btrfs-scrubnc", flags, 1, 0);
		if (!fs_info->scrub_nocow_workers) {
			ret = -ENOMEM;
2812
			goto out;
2813
		}
A
Arne Jansen 已提交
2814
	}
A
Arne Jansen 已提交
2815
	++fs_info->scrub_workers_refcnt;
2816 2817
out:
	return ret;
A
Arne Jansen 已提交
2818 2819
}

2820
static noinline_for_stack void scrub_workers_put(struct btrfs_fs_info *fs_info)
A
Arne Jansen 已提交
2821
{
2822
	if (--fs_info->scrub_workers_refcnt == 0) {
2823 2824 2825
		btrfs_destroy_workqueue(fs_info->scrub_workers);
		btrfs_destroy_workqueue(fs_info->scrub_wr_completion_workers);
		btrfs_destroy_workqueue(fs_info->scrub_nocow_workers);
2826
	}
A
Arne Jansen 已提交
2827 2828 2829
	WARN_ON(fs_info->scrub_workers_refcnt < 0);
}

2830 2831
int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
		    u64 end, struct btrfs_scrub_progress *progress,
2832
		    int readonly, int is_dev_replace)
A
Arne Jansen 已提交
2833
{
2834
	struct scrub_ctx *sctx;
A
Arne Jansen 已提交
2835 2836 2837
	int ret;
	struct btrfs_device *dev;

2838
	if (btrfs_fs_closing(fs_info))
A
Arne Jansen 已提交
2839 2840 2841 2842 2843
		return -EINVAL;

	/*
	 * check some assumptions
	 */
2844
	if (fs_info->chunk_root->nodesize != fs_info->chunk_root->leafsize) {
2845 2846
		btrfs_err(fs_info,
			   "scrub: size assumption nodesize == leafsize (%d == %d) fails",
2847 2848
		       fs_info->chunk_root->nodesize,
		       fs_info->chunk_root->leafsize);
2849 2850 2851
		return -EINVAL;
	}

2852
	if (fs_info->chunk_root->nodesize > BTRFS_STRIPE_LEN) {
2853 2854 2855 2856 2857
		/*
		 * in this case scrub is unable to calculate the checksum
		 * the way scrub is implemented. Do not handle this
		 * situation at all because it won't ever happen.
		 */
2858 2859
		btrfs_err(fs_info,
			   "scrub: size assumption nodesize <= BTRFS_STRIPE_LEN (%d <= %d) fails",
2860
		       fs_info->chunk_root->nodesize, BTRFS_STRIPE_LEN);
2861 2862 2863
		return -EINVAL;
	}

2864
	if (fs_info->chunk_root->sectorsize != PAGE_SIZE) {
2865
		/* not supported for data w/o checksums */
2866 2867 2868
		btrfs_err(fs_info,
			   "scrub: size assumption sectorsize != PAGE_SIZE "
			   "(%d != %lu) fails",
2869
		       fs_info->chunk_root->sectorsize, PAGE_SIZE);
A
Arne Jansen 已提交
2870 2871 2872
		return -EINVAL;
	}

2873 2874 2875 2876 2877 2878 2879 2880
	if (fs_info->chunk_root->nodesize >
	    PAGE_SIZE * SCRUB_MAX_PAGES_PER_BLOCK ||
	    fs_info->chunk_root->sectorsize >
	    PAGE_SIZE * SCRUB_MAX_PAGES_PER_BLOCK) {
		/*
		 * would exhaust the array bounds of pagev member in
		 * struct scrub_block
		 */
2881 2882
		btrfs_err(fs_info, "scrub: size assumption nodesize and sectorsize "
			   "<= SCRUB_MAX_PAGES_PER_BLOCK (%d <= %d && %d <= %d) fails",
2883 2884 2885 2886 2887 2888 2889
		       fs_info->chunk_root->nodesize,
		       SCRUB_MAX_PAGES_PER_BLOCK,
		       fs_info->chunk_root->sectorsize,
		       SCRUB_MAX_PAGES_PER_BLOCK);
		return -EINVAL;
	}

A
Arne Jansen 已提交
2890

2891 2892
	mutex_lock(&fs_info->fs_devices->device_list_mutex);
	dev = btrfs_find_device(fs_info, devid, NULL, NULL);
2893
	if (!dev || (dev->missing && !is_dev_replace)) {
2894
		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
A
Arne Jansen 已提交
2895 2896 2897
		return -ENODEV;
	}

2898
	mutex_lock(&fs_info->scrub_lock);
2899
	if (!dev->in_fs_metadata || dev->is_tgtdev_for_dev_replace) {
A
Arne Jansen 已提交
2900
		mutex_unlock(&fs_info->scrub_lock);
2901 2902
		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
		return -EIO;
A
Arne Jansen 已提交
2903 2904
	}

2905 2906 2907 2908 2909
	btrfs_dev_replace_lock(&fs_info->dev_replace);
	if (dev->scrub_device ||
	    (!is_dev_replace &&
	     btrfs_dev_replace_is_ongoing(&fs_info->dev_replace))) {
		btrfs_dev_replace_unlock(&fs_info->dev_replace);
A
Arne Jansen 已提交
2910
		mutex_unlock(&fs_info->scrub_lock);
2911
		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
A
Arne Jansen 已提交
2912 2913
		return -EINPROGRESS;
	}
2914
	btrfs_dev_replace_unlock(&fs_info->dev_replace);
2915 2916 2917 2918 2919 2920 2921 2922

	ret = scrub_workers_get(fs_info, is_dev_replace);
	if (ret) {
		mutex_unlock(&fs_info->scrub_lock);
		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
		return ret;
	}

2923
	sctx = scrub_setup_ctx(dev, is_dev_replace);
2924
	if (IS_ERR(sctx)) {
A
Arne Jansen 已提交
2925
		mutex_unlock(&fs_info->scrub_lock);
2926 2927
		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
		scrub_workers_put(fs_info);
2928
		return PTR_ERR(sctx);
A
Arne Jansen 已提交
2929
	}
2930 2931
	sctx->readonly = readonly;
	dev->scrub_device = sctx;
2932
	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
A
Arne Jansen 已提交
2933

2934 2935 2936 2937
	/*
	 * checking @scrub_pause_req here, we can avoid
	 * race between committing transaction and scrubbing.
	 */
2938
	__scrub_blocked_if_needed(fs_info);
A
Arne Jansen 已提交
2939 2940 2941
	atomic_inc(&fs_info->scrubs_running);
	mutex_unlock(&fs_info->scrub_lock);

2942
	if (!is_dev_replace) {
2943 2944 2945 2946
		/*
		 * by holding device list mutex, we can
		 * kick off writing super in log tree sync.
		 */
2947
		mutex_lock(&fs_info->fs_devices->device_list_mutex);
2948
		ret = scrub_supers(sctx, dev);
2949
		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
2950
	}
A
Arne Jansen 已提交
2951 2952

	if (!ret)
2953 2954
		ret = scrub_enumerate_chunks(sctx, dev, start, end,
					     is_dev_replace);
A
Arne Jansen 已提交
2955

2956
	wait_event(sctx->list_wait, atomic_read(&sctx->bios_in_flight) == 0);
A
Arne Jansen 已提交
2957 2958 2959
	atomic_dec(&fs_info->scrubs_running);
	wake_up(&fs_info->scrub_pause_wait);

2960
	wait_event(sctx->list_wait, atomic_read(&sctx->workers_pending) == 0);
2961

A
Arne Jansen 已提交
2962
	if (progress)
2963
		memcpy(progress, &sctx->stat, sizeof(*progress));
A
Arne Jansen 已提交
2964 2965 2966

	mutex_lock(&fs_info->scrub_lock);
	dev->scrub_device = NULL;
2967
	scrub_workers_put(fs_info);
A
Arne Jansen 已提交
2968 2969
	mutex_unlock(&fs_info->scrub_lock);

2970
	scrub_free_ctx(sctx);
A
Arne Jansen 已提交
2971 2972 2973 2974

	return ret;
}

2975
void btrfs_scrub_pause(struct btrfs_root *root)
A
Arne Jansen 已提交
2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991
{
	struct btrfs_fs_info *fs_info = root->fs_info;

	mutex_lock(&fs_info->scrub_lock);
	atomic_inc(&fs_info->scrub_pause_req);
	while (atomic_read(&fs_info->scrubs_paused) !=
	       atomic_read(&fs_info->scrubs_running)) {
		mutex_unlock(&fs_info->scrub_lock);
		wait_event(fs_info->scrub_pause_wait,
			   atomic_read(&fs_info->scrubs_paused) ==
			   atomic_read(&fs_info->scrubs_running));
		mutex_lock(&fs_info->scrub_lock);
	}
	mutex_unlock(&fs_info->scrub_lock);
}

2992
void btrfs_scrub_continue(struct btrfs_root *root)
A
Arne Jansen 已提交
2993 2994 2995 2996 2997 2998 2999
{
	struct btrfs_fs_info *fs_info = root->fs_info;

	atomic_dec(&fs_info->scrub_pause_req);
	wake_up(&fs_info->scrub_pause_wait);
}

3000
int btrfs_scrub_cancel(struct btrfs_fs_info *fs_info)
A
Arne Jansen 已提交
3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020
{
	mutex_lock(&fs_info->scrub_lock);
	if (!atomic_read(&fs_info->scrubs_running)) {
		mutex_unlock(&fs_info->scrub_lock);
		return -ENOTCONN;
	}

	atomic_inc(&fs_info->scrub_cancel_req);
	while (atomic_read(&fs_info->scrubs_running)) {
		mutex_unlock(&fs_info->scrub_lock);
		wait_event(fs_info->scrub_pause_wait,
			   atomic_read(&fs_info->scrubs_running) == 0);
		mutex_lock(&fs_info->scrub_lock);
	}
	atomic_dec(&fs_info->scrub_cancel_req);
	mutex_unlock(&fs_info->scrub_lock);

	return 0;
}

3021 3022
int btrfs_scrub_cancel_dev(struct btrfs_fs_info *fs_info,
			   struct btrfs_device *dev)
3023
{
3024
	struct scrub_ctx *sctx;
A
Arne Jansen 已提交
3025 3026

	mutex_lock(&fs_info->scrub_lock);
3027 3028
	sctx = dev->scrub_device;
	if (!sctx) {
A
Arne Jansen 已提交
3029 3030 3031
		mutex_unlock(&fs_info->scrub_lock);
		return -ENOTCONN;
	}
3032
	atomic_inc(&sctx->cancel_req);
A
Arne Jansen 已提交
3033 3034 3035 3036 3037 3038 3039 3040 3041 3042
	while (dev->scrub_device) {
		mutex_unlock(&fs_info->scrub_lock);
		wait_event(fs_info->scrub_pause_wait,
			   dev->scrub_device == NULL);
		mutex_lock(&fs_info->scrub_lock);
	}
	mutex_unlock(&fs_info->scrub_lock);

	return 0;
}
S
Stefan Behrens 已提交
3043

A
Arne Jansen 已提交
3044 3045 3046 3047
int btrfs_scrub_progress(struct btrfs_root *root, u64 devid,
			 struct btrfs_scrub_progress *progress)
{
	struct btrfs_device *dev;
3048
	struct scrub_ctx *sctx = NULL;
A
Arne Jansen 已提交
3049 3050

	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
3051
	dev = btrfs_find_device(root->fs_info, devid, NULL, NULL);
A
Arne Jansen 已提交
3052
	if (dev)
3053 3054 3055
		sctx = dev->scrub_device;
	if (sctx)
		memcpy(progress, &sctx->stat, sizeof(*progress));
A
Arne Jansen 已提交
3056 3057
	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);

3058
	return dev ? (sctx ? 0 : -ENOTCONN) : -ENODEV;
A
Arne Jansen 已提交
3059
}
3060 3061 3062 3063 3064 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 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135

static void scrub_remap_extent(struct btrfs_fs_info *fs_info,
			       u64 extent_logical, u64 extent_len,
			       u64 *extent_physical,
			       struct btrfs_device **extent_dev,
			       int *extent_mirror_num)
{
	u64 mapped_length;
	struct btrfs_bio *bbio = NULL;
	int ret;

	mapped_length = extent_len;
	ret = btrfs_map_block(fs_info, READ, extent_logical,
			      &mapped_length, &bbio, 0);
	if (ret || !bbio || mapped_length < extent_len ||
	    !bbio->stripes[0].dev->bdev) {
		kfree(bbio);
		return;
	}

	*extent_physical = bbio->stripes[0].physical;
	*extent_mirror_num = bbio->mirror_num;
	*extent_dev = bbio->stripes[0].dev;
	kfree(bbio);
}

static int scrub_setup_wr_ctx(struct scrub_ctx *sctx,
			      struct scrub_wr_ctx *wr_ctx,
			      struct btrfs_fs_info *fs_info,
			      struct btrfs_device *dev,
			      int is_dev_replace)
{
	WARN_ON(wr_ctx->wr_curr_bio != NULL);

	mutex_init(&wr_ctx->wr_lock);
	wr_ctx->wr_curr_bio = NULL;
	if (!is_dev_replace)
		return 0;

	WARN_ON(!dev->bdev);
	wr_ctx->pages_per_wr_bio = min_t(int, SCRUB_PAGES_PER_WR_BIO,
					 bio_get_nr_vecs(dev->bdev));
	wr_ctx->tgtdev = dev;
	atomic_set(&wr_ctx->flush_all_writes, 0);
	return 0;
}

static void scrub_free_wr_ctx(struct scrub_wr_ctx *wr_ctx)
{
	mutex_lock(&wr_ctx->wr_lock);
	kfree(wr_ctx->wr_curr_bio);
	wr_ctx->wr_curr_bio = NULL;
	mutex_unlock(&wr_ctx->wr_lock);
}

static int copy_nocow_pages(struct scrub_ctx *sctx, u64 logical, u64 len,
			    int mirror_num, u64 physical_for_dev_replace)
{
	struct scrub_copy_nocow_ctx *nocow_ctx;
	struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info;

	nocow_ctx = kzalloc(sizeof(*nocow_ctx), GFP_NOFS);
	if (!nocow_ctx) {
		spin_lock(&sctx->stat_lock);
		sctx->stat.malloc_errors++;
		spin_unlock(&sctx->stat_lock);
		return -ENOMEM;
	}

	scrub_pending_trans_workers_inc(sctx);

	nocow_ctx->sctx = sctx;
	nocow_ctx->logical = logical;
	nocow_ctx->len = len;
	nocow_ctx->mirror_num = mirror_num;
	nocow_ctx->physical_for_dev_replace = physical_for_dev_replace;
3136
	btrfs_init_work(&nocow_ctx->work, copy_nocow_pages_worker, NULL, NULL);
3137
	INIT_LIST_HEAD(&nocow_ctx->inodes);
3138 3139
	btrfs_queue_work(fs_info->scrub_nocow_workers,
			 &nocow_ctx->work);
3140 3141 3142 3143

	return 0;
}

3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160
static int record_inode_for_nocow(u64 inum, u64 offset, u64 root, void *ctx)
{
	struct scrub_copy_nocow_ctx *nocow_ctx = ctx;
	struct scrub_nocow_inode *nocow_inode;

	nocow_inode = kzalloc(sizeof(*nocow_inode), GFP_NOFS);
	if (!nocow_inode)
		return -ENOMEM;
	nocow_inode->inum = inum;
	nocow_inode->offset = offset;
	nocow_inode->root = root;
	list_add_tail(&nocow_inode->list, &nocow_ctx->inodes);
	return 0;
}

#define COPY_COMPLETE 1

3161
static void copy_nocow_pages_worker(struct btrfs_work *work)
3162 3163 3164 3165 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 3192 3193 3194 3195
{
	struct scrub_copy_nocow_ctx *nocow_ctx =
		container_of(work, struct scrub_copy_nocow_ctx, work);
	struct scrub_ctx *sctx = nocow_ctx->sctx;
	u64 logical = nocow_ctx->logical;
	u64 len = nocow_ctx->len;
	int mirror_num = nocow_ctx->mirror_num;
	u64 physical_for_dev_replace = nocow_ctx->physical_for_dev_replace;
	int ret;
	struct btrfs_trans_handle *trans = NULL;
	struct btrfs_fs_info *fs_info;
	struct btrfs_path *path;
	struct btrfs_root *root;
	int not_written = 0;

	fs_info = sctx->dev_root->fs_info;
	root = fs_info->extent_root;

	path = btrfs_alloc_path();
	if (!path) {
		spin_lock(&sctx->stat_lock);
		sctx->stat.malloc_errors++;
		spin_unlock(&sctx->stat_lock);
		not_written = 1;
		goto out;
	}

	trans = btrfs_join_transaction(root);
	if (IS_ERR(trans)) {
		not_written = 1;
		goto out;
	}

	ret = iterate_inodes_from_logical(logical, fs_info, path,
3196
					  record_inode_for_nocow, nocow_ctx);
3197
	if (ret != 0 && ret != -ENOENT) {
3198 3199
		btrfs_warn(fs_info, "iterate_inodes_from_logical() failed: log %llu, "
			"phys %llu, len %llu, mir %u, ret %d",
3200 3201
			logical, physical_for_dev_replace, len, mirror_num,
			ret);
3202 3203 3204 3205
		not_written = 1;
		goto out;
	}

3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223
	btrfs_end_transaction(trans, root);
	trans = NULL;
	while (!list_empty(&nocow_ctx->inodes)) {
		struct scrub_nocow_inode *entry;
		entry = list_first_entry(&nocow_ctx->inodes,
					 struct scrub_nocow_inode,
					 list);
		list_del_init(&entry->list);
		ret = copy_nocow_pages_for_inode(entry->inum, entry->offset,
						 entry->root, nocow_ctx);
		kfree(entry);
		if (ret == COPY_COMPLETE) {
			ret = 0;
			break;
		} else if (ret) {
			break;
		}
	}
3224
out:
3225 3226 3227 3228 3229 3230 3231 3232
	while (!list_empty(&nocow_ctx->inodes)) {
		struct scrub_nocow_inode *entry;
		entry = list_first_entry(&nocow_ctx->inodes,
					 struct scrub_nocow_inode,
					 list);
		list_del_init(&entry->list);
		kfree(entry);
	}
3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244
	if (trans && !IS_ERR(trans))
		btrfs_end_transaction(trans, root);
	if (not_written)
		btrfs_dev_replace_stats_inc(&fs_info->dev_replace.
					    num_uncorrectable_read_errors);

	btrfs_free_path(path);
	kfree(nocow_ctx);

	scrub_pending_trans_workers_dec(sctx);
}

3245 3246
static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root,
				      struct scrub_copy_nocow_ctx *nocow_ctx)
3247
{
3248
	struct btrfs_fs_info *fs_info = nocow_ctx->sctx->dev_root->fs_info;
3249
	struct btrfs_key key;
3250 3251
	struct inode *inode;
	struct page *page;
3252
	struct btrfs_root *local_root;
3253 3254 3255 3256
	struct btrfs_ordered_extent *ordered;
	struct extent_map *em;
	struct extent_state *cached_state = NULL;
	struct extent_io_tree *io_tree;
3257
	u64 physical_for_dev_replace;
3258 3259
	u64 len = nocow_ctx->len;
	u64 lockstart = offset, lockend = offset + len - 1;
3260
	unsigned long index;
3261
	int srcu_index;
3262 3263
	int ret = 0;
	int err = 0;
3264 3265 3266 3267

	key.objectid = root;
	key.type = BTRFS_ROOT_ITEM_KEY;
	key.offset = (u64)-1;
3268 3269 3270

	srcu_index = srcu_read_lock(&fs_info->subvol_srcu);

3271
	local_root = btrfs_read_fs_root_no_name(fs_info, &key);
3272 3273
	if (IS_ERR(local_root)) {
		srcu_read_unlock(&fs_info->subvol_srcu, srcu_index);
3274
		return PTR_ERR(local_root);
3275
	}
3276 3277 3278 3279 3280

	key.type = BTRFS_INODE_ITEM_KEY;
	key.objectid = inum;
	key.offset = 0;
	inode = btrfs_iget(fs_info->sb, &key, local_root, NULL);
3281
	srcu_read_unlock(&fs_info->subvol_srcu, srcu_index);
3282 3283 3284
	if (IS_ERR(inode))
		return PTR_ERR(inode);

3285 3286 3287 3288
	/* Avoid truncate/dio/punch hole.. */
	mutex_lock(&inode->i_mutex);
	inode_dio_wait(inode);

3289
	physical_for_dev_replace = nocow_ctx->physical_for_dev_replace;
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
	io_tree = &BTRFS_I(inode)->io_tree;

	lock_extent_bits(io_tree, lockstart, lockend, 0, &cached_state);
	ordered = btrfs_lookup_ordered_range(inode, lockstart, len);
	if (ordered) {
		btrfs_put_ordered_extent(ordered);
		goto out_unlock;
	}

	em = btrfs_get_extent(inode, NULL, 0, lockstart, len, 0);
	if (IS_ERR(em)) {
		ret = PTR_ERR(em);
		goto out_unlock;
	}

	/*
	 * This extent does not actually cover the logical extent anymore,
	 * move on to the next inode.
	 */
	if (em->block_start > nocow_ctx->logical ||
	    em->block_start + em->block_len < nocow_ctx->logical + len) {
		free_extent_map(em);
		goto out_unlock;
	}
	free_extent_map(em);

3316 3317
	while (len >= PAGE_CACHE_SIZE) {
		index = offset >> PAGE_CACHE_SHIFT;
3318
again:
3319 3320
		page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
		if (!page) {
3321
			btrfs_err(fs_info, "find_or_create_page() failed");
3322
			ret = -ENOMEM;
3323
			goto out;
3324 3325 3326 3327 3328 3329 3330
		}

		if (PageUptodate(page)) {
			if (PageDirty(page))
				goto next_page;
		} else {
			ClearPageError(page);
3331 3332 3333
			err = extent_read_full_page_nolock(io_tree, page,
							   btrfs_get_extent,
							   nocow_ctx->mirror_num);
3334 3335
			if (err) {
				ret = err;
3336 3337
				goto next_page;
			}
3338

3339
			lock_page(page);
3340 3341 3342 3343 3344 3345 3346
			/*
			 * If the page has been remove from the page cache,
			 * the data on it is meaningless, because it may be
			 * old one, the new data may be written into the new
			 * page in the page cache.
			 */
			if (page->mapping != inode->i_mapping) {
3347
				unlock_page(page);
3348 3349 3350
				page_cache_release(page);
				goto again;
			}
3351 3352 3353 3354 3355
			if (!PageUptodate(page)) {
				ret = -EIO;
				goto next_page;
			}
		}
3356 3357 3358 3359
		err = write_page_nocow(nocow_ctx->sctx,
				       physical_for_dev_replace, page);
		if (err)
			ret = err;
3360
next_page:
3361 3362 3363 3364 3365 3366
		unlock_page(page);
		page_cache_release(page);

		if (ret)
			break;

3367 3368 3369 3370
		offset += PAGE_CACHE_SIZE;
		physical_for_dev_replace += PAGE_CACHE_SIZE;
		len -= PAGE_CACHE_SIZE;
	}
3371 3372 3373 3374
	ret = COPY_COMPLETE;
out_unlock:
	unlock_extent_cached(io_tree, lockstart, lockend, &cached_state,
			     GFP_NOFS);
3375
out:
3376
	mutex_unlock(&inode->i_mutex);
3377
	iput(inode);
3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392
	return ret;
}

static int write_page_nocow(struct scrub_ctx *sctx,
			    u64 physical_for_dev_replace, struct page *page)
{
	struct bio *bio;
	struct btrfs_device *dev;
	int ret;

	dev = sctx->wr_ctx.tgtdev;
	if (!dev)
		return -EIO;
	if (!dev->bdev) {
		printk_ratelimited(KERN_WARNING
3393
			"BTRFS: scrub write_page_nocow(bdev == NULL) is unexpected!\n");
3394 3395
		return -EIO;
	}
3396
	bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413
	if (!bio) {
		spin_lock(&sctx->stat_lock);
		sctx->stat.malloc_errors++;
		spin_unlock(&sctx->stat_lock);
		return -ENOMEM;
	}
	bio->bi_size = 0;
	bio->bi_sector = physical_for_dev_replace >> 9;
	bio->bi_bdev = dev->bdev;
	ret = bio_add_page(bio, page, PAGE_CACHE_SIZE, 0);
	if (ret != PAGE_CACHE_SIZE) {
leave_with_eio:
		bio_put(bio);
		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
		return -EIO;
	}

3414
	if (btrfsic_submit_bio_wait(WRITE_SYNC, bio))
3415 3416 3417 3418 3419
		goto leave_with_eio;

	bio_put(bio);
	return 0;
}