super.c 107.0 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0
2
/*
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
 * fs/f2fs/super.c
 *
 * Copyright (c) 2012 Samsung Electronics Co., Ltd.
 *             http://www.samsung.com/
 */
#include <linux/module.h>
#include <linux/init.h>
#include <linux/fs.h>
#include <linux/statfs.h>
#include <linux/buffer_head.h>
#include <linux/backing-dev.h>
#include <linux/kthread.h>
#include <linux/parser.h>
#include <linux/mount.h>
#include <linux/seq_file.h>
18
#include <linux/proc_fs.h>
19 20
#include <linux/random.h>
#include <linux/exportfs.h>
21
#include <linux/blkdev.h>
22
#include <linux/quotaops.h>
23
#include <linux/f2fs_fs.h>
24
#include <linux/sysfs.h>
25
#include <linux/quota.h>
26
#include <linux/unicode.h>
27
#include <linux/part_stat.h>
28 29 30

#include "f2fs.h"
#include "node.h"
31
#include "segment.h"
32
#include "xattr.h"
33
#include "gc.h"
34
#include "trace.h"
35

36 37 38
#define CREATE_TRACE_POINTS
#include <trace/events/f2fs.h>

39 40
static struct kmem_cache *f2fs_inode_cachep;

41
#ifdef CONFIG_F2FS_FAULT_INJECTION
42

43
const char *f2fs_fault_name[FAULT_MAX] = {
44
	[FAULT_KMALLOC]		= "kmalloc",
45
	[FAULT_KVMALLOC]	= "kvmalloc",
46
	[FAULT_PAGE_ALLOC]	= "page alloc",
47
	[FAULT_PAGE_GET]	= "page get",
48
	[FAULT_ALLOC_BIO]	= "alloc bio",
49 50 51 52
	[FAULT_ALLOC_NID]	= "alloc nid",
	[FAULT_ORPHAN]		= "orphan",
	[FAULT_BLOCK]		= "no more block",
	[FAULT_DIR_DEPTH]	= "too big dir depth",
53
	[FAULT_EVICT_INODE]	= "evict_inode fail",
54
	[FAULT_TRUNCATE]	= "truncate fail",
55
	[FAULT_READ_IO]		= "read IO error",
56
	[FAULT_CHECKPOINT]	= "checkpoint error",
57
	[FAULT_DISCARD]		= "discard error",
58
	[FAULT_WRITE_IO]	= "write IO error",
59
};
60

61 62
void f2fs_build_fault_attr(struct f2fs_sb_info *sbi, unsigned int rate,
							unsigned int type)
63
{
64
	struct f2fs_fault_info *ffi = &F2FS_OPTION(sbi).fault_info;
65

66
	if (rate) {
67 68
		atomic_set(&ffi->inject_ops, 0);
		ffi->inject_rate = rate;
69
	}
70 71 72 73 74 75

	if (type)
		ffi->inject_type = type;

	if (!rate && !type)
		memset(ffi, 0, sizeof(struct f2fs_fault_info));
76
}
77 78
#endif

79 80 81 82 83 84 85
/* f2fs-wide shrinker description */
static struct shrinker f2fs_shrinker_info = {
	.scan_objects = f2fs_shrink_scan,
	.count_objects = f2fs_shrink_count,
	.seeks = DEFAULT_SEEKS,
};

86
enum {
87
	Opt_gc_background,
88
	Opt_disable_roll_forward,
89
	Opt_norecovery,
90
	Opt_discard,
91
	Opt_nodiscard,
92
	Opt_noheap,
93
	Opt_heap,
94
	Opt_user_xattr,
95
	Opt_nouser_xattr,
96
	Opt_acl,
97 98 99
	Opt_noacl,
	Opt_active_logs,
	Opt_disable_ext_identify,
100
	Opt_inline_xattr,
101
	Opt_noinline_xattr,
102
	Opt_inline_xattr_size,
103
	Opt_inline_data,
104
	Opt_inline_dentry,
105
	Opt_noinline_dentry,
106
	Opt_flush_merge,
107
	Opt_noflush_merge,
108
	Opt_nobarrier,
109
	Opt_fastboot,
110
	Opt_extent_cache,
111
	Opt_noextent_cache,
112
	Opt_noinline_data,
113
	Opt_data_flush,
114
	Opt_reserve_root,
115 116
	Opt_resgid,
	Opt_resuid,
117
	Opt_mode,
118
	Opt_io_size_bits,
119
	Opt_fault_injection,
120
	Opt_fault_type,
121 122
	Opt_lazytime,
	Opt_nolazytime,
123 124
	Opt_quota,
	Opt_noquota,
125 126
	Opt_usrquota,
	Opt_grpquota,
C
Chao Yu 已提交
127
	Opt_prjquota,
128 129 130 131 132 133 134 135 136
	Opt_usrjquota,
	Opt_grpjquota,
	Opt_prjjquota,
	Opt_offusrjquota,
	Opt_offgrpjquota,
	Opt_offprjjquota,
	Opt_jqfmt_vfsold,
	Opt_jqfmt_vfsv0,
	Opt_jqfmt_vfsv1,
137
	Opt_whint,
138
	Opt_alloc,
139
	Opt_fsync,
140
	Opt_test_dummy_encryption,
141
	Opt_inlinecrypt,
142 143 144 145
	Opt_checkpoint_disable,
	Opt_checkpoint_disable_cap,
	Opt_checkpoint_disable_cap_perc,
	Opt_checkpoint_enable,
146 147 148
	Opt_compress_algorithm,
	Opt_compress_log_size,
	Opt_compress_extension,
149
	Opt_compress_chksum,
150
	Opt_compress_mode,
151
	Opt_atgc,
152 153 154 155
	Opt_err,
};

static match_table_t f2fs_tokens = {
156
	{Opt_gc_background, "background_gc=%s"},
157
	{Opt_disable_roll_forward, "disable_roll_forward"},
158
	{Opt_norecovery, "norecovery"},
159
	{Opt_discard, "discard"},
160
	{Opt_nodiscard, "nodiscard"},
161
	{Opt_noheap, "no_heap"},
162
	{Opt_heap, "heap"},
163
	{Opt_user_xattr, "user_xattr"},
164
	{Opt_nouser_xattr, "nouser_xattr"},
165
	{Opt_acl, "acl"},
166 167 168
	{Opt_noacl, "noacl"},
	{Opt_active_logs, "active_logs=%u"},
	{Opt_disable_ext_identify, "disable_ext_identify"},
169
	{Opt_inline_xattr, "inline_xattr"},
170
	{Opt_noinline_xattr, "noinline_xattr"},
171
	{Opt_inline_xattr_size, "inline_xattr_size=%u"},
172
	{Opt_inline_data, "inline_data"},
173
	{Opt_inline_dentry, "inline_dentry"},
174
	{Opt_noinline_dentry, "noinline_dentry"},
175
	{Opt_flush_merge, "flush_merge"},
176
	{Opt_noflush_merge, "noflush_merge"},
177
	{Opt_nobarrier, "nobarrier"},
178
	{Opt_fastboot, "fastboot"},
179
	{Opt_extent_cache, "extent_cache"},
180
	{Opt_noextent_cache, "noextent_cache"},
181
	{Opt_noinline_data, "noinline_data"},
182
	{Opt_data_flush, "data_flush"},
183
	{Opt_reserve_root, "reserve_root=%u"},
184 185
	{Opt_resgid, "resgid=%u"},
	{Opt_resuid, "resuid=%u"},
186
	{Opt_mode, "mode=%s"},
187
	{Opt_io_size_bits, "io_bits=%u"},
188
	{Opt_fault_injection, "fault_injection=%u"},
189
	{Opt_fault_type, "fault_type=%u"},
190 191
	{Opt_lazytime, "lazytime"},
	{Opt_nolazytime, "nolazytime"},
192 193
	{Opt_quota, "quota"},
	{Opt_noquota, "noquota"},
194 195
	{Opt_usrquota, "usrquota"},
	{Opt_grpquota, "grpquota"},
C
Chao Yu 已提交
196
	{Opt_prjquota, "prjquota"},
197 198 199 200 201 202 203 204 205
	{Opt_usrjquota, "usrjquota=%s"},
	{Opt_grpjquota, "grpjquota=%s"},
	{Opt_prjjquota, "prjjquota=%s"},
	{Opt_offusrjquota, "usrjquota="},
	{Opt_offgrpjquota, "grpjquota="},
	{Opt_offprjjquota, "prjjquota="},
	{Opt_jqfmt_vfsold, "jqfmt=vfsold"},
	{Opt_jqfmt_vfsv0, "jqfmt=vfsv0"},
	{Opt_jqfmt_vfsv1, "jqfmt=vfsv1"},
206
	{Opt_whint, "whint_mode=%s"},
207
	{Opt_alloc, "alloc_mode=%s"},
208
	{Opt_fsync, "fsync_mode=%s"},
209
	{Opt_test_dummy_encryption, "test_dummy_encryption=%s"},
210
	{Opt_test_dummy_encryption, "test_dummy_encryption"},
211
	{Opt_inlinecrypt, "inlinecrypt"},
212 213 214 215
	{Opt_checkpoint_disable, "checkpoint=disable"},
	{Opt_checkpoint_disable_cap, "checkpoint=disable:%u"},
	{Opt_checkpoint_disable_cap_perc, "checkpoint=disable:%u%%"},
	{Opt_checkpoint_enable, "checkpoint=enable"},
216 217 218
	{Opt_compress_algorithm, "compress_algorithm=%s"},
	{Opt_compress_log_size, "compress_log_size=%u"},
	{Opt_compress_extension, "compress_extension=%s"},
219
	{Opt_compress_chksum, "compress_chksum"},
220
	{Opt_compress_mode, "compress_mode=%s"},
221
	{Opt_atgc, "atgc"},
222 223 224
	{Opt_err, NULL},
};

225
void f2fs_printk(struct f2fs_sb_info *sbi, const char *fmt, ...)
226 227 228
{
	struct va_format vaf;
	va_list args;
229
	int level;
230 231

	va_start(args, fmt);
232 233 234

	level = printk_get_level(fmt);
	vaf.fmt = printk_skip_level(fmt);
235
	vaf.va = &args;
236 237 238
	printk("%c%cF2FS-fs (%s): %pV\n",
	       KERN_SOH_ASCII, level, sbi->sb->s_id, &vaf);

239 240 241
	va_end(args);
}

242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271
#ifdef CONFIG_UNICODE
static const struct f2fs_sb_encodings {
	__u16 magic;
	char *name;
	char *version;
} f2fs_sb_encoding_map[] = {
	{F2FS_ENC_UTF8_12_1, "utf8", "12.1.0"},
};

static int f2fs_sb_read_encoding(const struct f2fs_super_block *sb,
				 const struct f2fs_sb_encodings **encoding,
				 __u16 *flags)
{
	__u16 magic = le16_to_cpu(sb->s_encoding);
	int i;

	for (i = 0; i < ARRAY_SIZE(f2fs_sb_encoding_map); i++)
		if (magic == f2fs_sb_encoding_map[i].magic)
			break;

	if (i >= ARRAY_SIZE(f2fs_sb_encoding_map))
		return -EINVAL;

	*encoding = &f2fs_sb_encoding_map[i];
	*flags = le16_to_cpu(sb->s_encoding_flags);

	return 0;
}
#endif

272 273
static inline void limit_reserve_root(struct f2fs_sb_info *sbi)
{
274 275
	block_t limit = min((sbi->user_block_count << 1) / 1000,
			sbi->user_block_count - sbi->reserved_blocks);
276 277

	/* limit is 0.2% */
278 279 280
	if (test_opt(sbi, RESERVE_ROOT) &&
			F2FS_OPTION(sbi).root_reserved_blocks > limit) {
		F2FS_OPTION(sbi).root_reserved_blocks = limit;
281 282
		f2fs_info(sbi, "Reduce reserved blocks for root = %u",
			  F2FS_OPTION(sbi).root_reserved_blocks);
283
	}
284
	if (!test_opt(sbi, RESERVE_ROOT) &&
285
		(!uid_eq(F2FS_OPTION(sbi).s_resuid,
286
				make_kuid(&init_user_ns, F2FS_DEF_RESUID)) ||
287
		!gid_eq(F2FS_OPTION(sbi).s_resgid,
288
				make_kgid(&init_user_ns, F2FS_DEF_RESGID))))
289 290 291 292 293
		f2fs_info(sbi, "Ignore s_resuid=%u, s_resgid=%u w/o reserve_root",
			  from_kuid_munged(&init_user_ns,
					   F2FS_OPTION(sbi).s_resuid),
			  from_kgid_munged(&init_user_ns,
					   F2FS_OPTION(sbi).s_resgid));
294 295
}

296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311
static inline void adjust_unusable_cap_perc(struct f2fs_sb_info *sbi)
{
	if (!F2FS_OPTION(sbi).unusable_cap_perc)
		return;

	if (F2FS_OPTION(sbi).unusable_cap_perc == 100)
		F2FS_OPTION(sbi).unusable_cap = sbi->user_block_count;
	else
		F2FS_OPTION(sbi).unusable_cap = (sbi->user_block_count / 100) *
					F2FS_OPTION(sbi).unusable_cap_perc;

	f2fs_info(sbi, "Adjust unusable cap for checkpoint=disable = %u / %u%%",
			F2FS_OPTION(sbi).unusable_cap,
			F2FS_OPTION(sbi).unusable_cap_perc);
}

312 313 314 315 316 317 318
static void init_once(void *foo)
{
	struct f2fs_inode_info *fi = (struct f2fs_inode_info *) foo;

	inode_init_once(&fi->vfs_inode);
}

319 320 321 322 323 324 325 326 327 328
#ifdef CONFIG_QUOTA
static const char * const quotatypes[] = INITQFNAMES;
#define QTYPE2NAME(t) (quotatypes[t])
static int f2fs_set_qf_name(struct super_block *sb, int qtype,
							substring_t *args)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	char *qname;
	int ret = -EINVAL;

329
	if (sb_any_quota_loaded(sb) && !F2FS_OPTION(sbi).s_qf_names[qtype]) {
330
		f2fs_err(sbi, "Cannot change journaled quota options when quota turned on");
331 332
		return -EINVAL;
	}
333
	if (f2fs_sb_has_quota_ino(sbi)) {
334
		f2fs_info(sbi, "QUOTA feature is enabled, so ignore qf_name");
335 336 337
		return 0;
	}

338 339
	qname = match_strdup(args);
	if (!qname) {
340
		f2fs_err(sbi, "Not enough memory for storing quotafile name");
341
		return -ENOMEM;
342
	}
343 344
	if (F2FS_OPTION(sbi).s_qf_names[qtype]) {
		if (strcmp(F2FS_OPTION(sbi).s_qf_names[qtype], qname) == 0)
345 346
			ret = 0;
		else
347
			f2fs_err(sbi, "%s quota file already specified",
348 349 350 351
				 QTYPE2NAME(qtype));
		goto errout;
	}
	if (strchr(qname, '/')) {
352
		f2fs_err(sbi, "quotafile must be on filesystem root");
353 354
		goto errout;
	}
355
	F2FS_OPTION(sbi).s_qf_names[qtype] = qname;
356 357 358
	set_opt(sbi, QUOTA);
	return 0;
errout:
359
	kfree(qname);
360 361 362 363 364 365 366
	return ret;
}

static int f2fs_clear_qf_name(struct super_block *sb, int qtype)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);

367
	if (sb_any_quota_loaded(sb) && F2FS_OPTION(sbi).s_qf_names[qtype]) {
368
		f2fs_err(sbi, "Cannot change journaled quota options when quota turned on");
369 370
		return -EINVAL;
	}
371
	kfree(F2FS_OPTION(sbi).s_qf_names[qtype]);
372
	F2FS_OPTION(sbi).s_qf_names[qtype] = NULL;
373 374 375 376 377 378 379 380 381 382
	return 0;
}

static int f2fs_check_quota_options(struct f2fs_sb_info *sbi)
{
	/*
	 * We do the test below only for project quotas. 'usrquota' and
	 * 'grpquota' mount options are allowed even without quota feature
	 * to support legacy quotas in quota files.
	 */
383
	if (test_opt(sbi, PRJQUOTA) && !f2fs_sb_has_project_quota(sbi)) {
384
		f2fs_err(sbi, "Project quota feature not enabled. Cannot enable project quota enforcement.");
385 386
		return -1;
	}
387 388 389 390 391
	if (F2FS_OPTION(sbi).s_qf_names[USRQUOTA] ||
			F2FS_OPTION(sbi).s_qf_names[GRPQUOTA] ||
			F2FS_OPTION(sbi).s_qf_names[PRJQUOTA]) {
		if (test_opt(sbi, USRQUOTA) &&
				F2FS_OPTION(sbi).s_qf_names[USRQUOTA])
392 393
			clear_opt(sbi, USRQUOTA);

394 395
		if (test_opt(sbi, GRPQUOTA) &&
				F2FS_OPTION(sbi).s_qf_names[GRPQUOTA])
396 397
			clear_opt(sbi, GRPQUOTA);

398 399
		if (test_opt(sbi, PRJQUOTA) &&
				F2FS_OPTION(sbi).s_qf_names[PRJQUOTA])
400 401 402 403
			clear_opt(sbi, PRJQUOTA);

		if (test_opt(sbi, GRPQUOTA) || test_opt(sbi, USRQUOTA) ||
				test_opt(sbi, PRJQUOTA)) {
404
			f2fs_err(sbi, "old and new quota format mixing");
405 406 407
			return -1;
		}

408
		if (!F2FS_OPTION(sbi).s_jquota_fmt) {
409
			f2fs_err(sbi, "journaled quota format not specified");
410 411 412
			return -1;
		}
	}
413

414
	if (f2fs_sb_has_quota_ino(sbi) && F2FS_OPTION(sbi).s_jquota_fmt) {
415
		f2fs_info(sbi, "QUOTA feature is enabled, so ignore jquota_fmt");
416
		F2FS_OPTION(sbi).s_jquota_fmt = 0;
417
	}
418 419 420 421
	return 0;
}
#endif

422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441
static int f2fs_set_test_dummy_encryption(struct super_block *sb,
					  const char *opt,
					  const substring_t *arg,
					  bool is_remount)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
#ifdef CONFIG_FS_ENCRYPTION
	int err;

	if (!f2fs_sb_has_encrypt(sbi)) {
		f2fs_err(sbi, "Encrypt feature is off");
		return -EINVAL;
	}

	/*
	 * This mount option is just for testing, and it's not worthwhile to
	 * implement the extra complexity (e.g. RCU protection) that would be
	 * needed to allow it to be set or changed during remount.  We do allow
	 * it to be specified during remount, but only if there is no change.
	 */
442
	if (is_remount && !F2FS_OPTION(sbi).dummy_enc_policy.policy) {
443 444 445 446
		f2fs_warn(sbi, "Can't set test_dummy_encryption on remount");
		return -EINVAL;
	}
	err = fscrypt_set_test_dummy_encryption(
447
		sb, arg->from, &F2FS_OPTION(sbi).dummy_enc_policy);
448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467
	if (err) {
		if (err == -EEXIST)
			f2fs_warn(sbi,
				  "Can't change test_dummy_encryption on remount");
		else if (err == -EINVAL)
			f2fs_warn(sbi, "Value of option \"%s\" is unrecognized",
				  opt);
		else
			f2fs_warn(sbi, "Error processing option \"%s\" [%d]",
				  opt, err);
		return -EINVAL;
	}
	f2fs_warn(sbi, "Test dummy encryption mode enabled");
#else
	f2fs_warn(sbi, "Test dummy encryption mount option ignored");
#endif
	return 0;
}

static int parse_options(struct super_block *sb, char *options, bool is_remount)
468 469 470
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	substring_t args[MAX_OPT_ARGS];
471
#ifdef CONFIG_F2FS_FS_COMPRESSION
472
	unsigned char (*ext)[F2FS_EXTENSION_LEN];
473 474
	int ext_cnt;
#endif
475
	char *p, *name;
476
	int arg = 0;
477 478
	kuid_t uid;
	kgid_t gid;
479
	int ret;
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500

	if (!options)
		return 0;

	while ((p = strsep(&options, ",")) != NULL) {
		int token;
		if (!*p)
			continue;
		/*
		 * Initialize args struct so we know whether arg was
		 * found; some options take optional arguments.
		 */
		args[0].to = args[0].from = NULL;
		token = match_token(p, f2fs_tokens, args);

		switch (token) {
		case Opt_gc_background:
			name = match_strdup(&args[0]);

			if (!name)
				return -ENOMEM;
501
			if (!strcmp(name, "on")) {
502
				F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_ON;
503
			} else if (!strcmp(name, "off")) {
504
				F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_OFF;
505
			} else if (!strcmp(name, "sync")) {
506
				F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_SYNC;
507
			} else {
508
				kfree(name);
509 510
				return -EINVAL;
			}
511
			kfree(name);
512 513 514 515
			break;
		case Opt_disable_roll_forward:
			set_opt(sbi, DISABLE_ROLL_FORWARD);
			break;
516 517
		case Opt_norecovery:
			/* this option mounts f2fs with ro */
518
			set_opt(sbi, NORECOVERY);
519 520 521
			if (!f2fs_readonly(sb))
				return -EINVAL;
			break;
522
		case Opt_discard:
523
			set_opt(sbi, DISCARD);
524
			break;
525
		case Opt_nodiscard:
526
			if (f2fs_sb_has_blkzoned(sbi)) {
527
				f2fs_warn(sbi, "discard is required for zoned block devices");
528 529
				return -EINVAL;
			}
530
			clear_opt(sbi, DISCARD);
531
			break;
532 533 534
		case Opt_noheap:
			set_opt(sbi, NOHEAP);
			break;
535 536 537
		case Opt_heap:
			clear_opt(sbi, NOHEAP);
			break;
538
#ifdef CONFIG_F2FS_FS_XATTR
539 540 541
		case Opt_user_xattr:
			set_opt(sbi, XATTR_USER);
			break;
542 543 544
		case Opt_nouser_xattr:
			clear_opt(sbi, XATTR_USER);
			break;
545 546 547
		case Opt_inline_xattr:
			set_opt(sbi, INLINE_XATTR);
			break;
548 549 550
		case Opt_noinline_xattr:
			clear_opt(sbi, INLINE_XATTR);
			break;
551 552 553 554
		case Opt_inline_xattr_size:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
			set_opt(sbi, INLINE_XATTR_SIZE);
555
			F2FS_OPTION(sbi).inline_xattr_size = arg;
556
			break;
557
#else
558
		case Opt_user_xattr:
559
			f2fs_info(sbi, "user_xattr options not supported");
560
			break;
561
		case Opt_nouser_xattr:
562
			f2fs_info(sbi, "nouser_xattr options not supported");
563
			break;
564
		case Opt_inline_xattr:
565
			f2fs_info(sbi, "inline_xattr options not supported");
566
			break;
567
		case Opt_noinline_xattr:
568
			f2fs_info(sbi, "noinline_xattr options not supported");
569
			break;
570 571
#endif
#ifdef CONFIG_F2FS_FS_POSIX_ACL
572 573 574
		case Opt_acl:
			set_opt(sbi, POSIX_ACL);
			break;
575 576 577 578
		case Opt_noacl:
			clear_opt(sbi, POSIX_ACL);
			break;
#else
579
		case Opt_acl:
580
			f2fs_info(sbi, "acl options not supported");
581
			break;
582
		case Opt_noacl:
583
			f2fs_info(sbi, "noacl options not supported");
584 585 586 587 588
			break;
#endif
		case Opt_active_logs:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
C
Chao Yu 已提交
589 590
			if (arg != 2 && arg != 4 &&
				arg != NR_CURSEG_PERSIST_TYPE)
591
				return -EINVAL;
592
			F2FS_OPTION(sbi).active_logs = arg;
593 594 595 596
			break;
		case Opt_disable_ext_identify:
			set_opt(sbi, DISABLE_EXT_IDENTIFY);
			break;
597 598 599
		case Opt_inline_data:
			set_opt(sbi, INLINE_DATA);
			break;
600 601 602
		case Opt_inline_dentry:
			set_opt(sbi, INLINE_DENTRY);
			break;
603 604 605
		case Opt_noinline_dentry:
			clear_opt(sbi, INLINE_DENTRY);
			break;
606 607 608
		case Opt_flush_merge:
			set_opt(sbi, FLUSH_MERGE);
			break;
609 610 611
		case Opt_noflush_merge:
			clear_opt(sbi, FLUSH_MERGE);
			break;
612 613 614
		case Opt_nobarrier:
			set_opt(sbi, NOBARRIER);
			break;
615 616 617
		case Opt_fastboot:
			set_opt(sbi, FASTBOOT);
			break;
618 619 620
		case Opt_extent_cache:
			set_opt(sbi, EXTENT_CACHE);
			break;
621 622 623
		case Opt_noextent_cache:
			clear_opt(sbi, EXTENT_CACHE);
			break;
624 625 626
		case Opt_noinline_data:
			clear_opt(sbi, INLINE_DATA);
			break;
627 628 629
		case Opt_data_flush:
			set_opt(sbi, DATA_FLUSH);
			break;
630 631 632 633
		case Opt_reserve_root:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
			if (test_opt(sbi, RESERVE_ROOT)) {
634 635
				f2fs_info(sbi, "Preserve previous reserve_root=%u",
					  F2FS_OPTION(sbi).root_reserved_blocks);
636
			} else {
637
				F2FS_OPTION(sbi).root_reserved_blocks = arg;
638 639 640
				set_opt(sbi, RESERVE_ROOT);
			}
			break;
641 642 643 644 645
		case Opt_resuid:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
			uid = make_kuid(current_user_ns(), arg);
			if (!uid_valid(uid)) {
646
				f2fs_err(sbi, "Invalid uid value %d", arg);
647 648
				return -EINVAL;
			}
649
			F2FS_OPTION(sbi).s_resuid = uid;
650 651 652 653 654 655
			break;
		case Opt_resgid:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
			gid = make_kgid(current_user_ns(), arg);
			if (!gid_valid(gid)) {
656
				f2fs_err(sbi, "Invalid gid value %d", arg);
657 658
				return -EINVAL;
			}
659
			F2FS_OPTION(sbi).s_resgid = gid;
660
			break;
661 662 663 664 665
		case Opt_mode:
			name = match_strdup(&args[0]);

			if (!name)
				return -ENOMEM;
666
			if (!strcmp(name, "adaptive")) {
667
				if (f2fs_sb_has_blkzoned(sbi)) {
668
					f2fs_warn(sbi, "adaptive mode is not allowed with zoned block device feature");
669
					kfree(name);
670 671
					return -EINVAL;
				}
672
				F2FS_OPTION(sbi).fs_mode = FS_MODE_ADAPTIVE;
673
			} else if (!strcmp(name, "lfs")) {
674
				F2FS_OPTION(sbi).fs_mode = FS_MODE_LFS;
675
			} else {
676
				kfree(name);
677 678
				return -EINVAL;
			}
679
			kfree(name);
680
			break;
681 682 683
		case Opt_io_size_bits:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
684
			if (arg <= 0 || arg > __ilog2_u32(BIO_MAX_PAGES)) {
685 686
				f2fs_warn(sbi, "Not support %d, larger than %d",
					  1 << arg, BIO_MAX_PAGES);
687 688
				return -EINVAL;
			}
689
			F2FS_OPTION(sbi).write_io_size_bits = arg;
690
			break;
691
#ifdef CONFIG_F2FS_FAULT_INJECTION
692 693 694
		case Opt_fault_injection:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
695 696 697
			f2fs_build_fault_attr(sbi, arg, F2FS_ALL_FAULT_TYPE);
			set_opt(sbi, FAULT_INJECTION);
			break;
698

699 700 701 702
		case Opt_fault_type:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
			f2fs_build_fault_attr(sbi, 0, arg);
703
			set_opt(sbi, FAULT_INJECTION);
704
			break;
705
#else
706
		case Opt_fault_injection:
707
			f2fs_info(sbi, "fault_injection options not supported");
708
			break;
709 710

		case Opt_fault_type:
711
			f2fs_info(sbi, "fault_type options not supported");
712 713
			break;
#endif
714
		case Opt_lazytime:
715
			sb->s_flags |= SB_LAZYTIME;
716 717
			break;
		case Opt_nolazytime:
718
			sb->s_flags &= ~SB_LAZYTIME;
719
			break;
720
#ifdef CONFIG_QUOTA
721
		case Opt_quota:
722 723 724 725 726 727
		case Opt_usrquota:
			set_opt(sbi, USRQUOTA);
			break;
		case Opt_grpquota:
			set_opt(sbi, GRPQUOTA);
			break;
C
Chao Yu 已提交
728 729 730
		case Opt_prjquota:
			set_opt(sbi, PRJQUOTA);
			break;
731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761
		case Opt_usrjquota:
			ret = f2fs_set_qf_name(sb, USRQUOTA, &args[0]);
			if (ret)
				return ret;
			break;
		case Opt_grpjquota:
			ret = f2fs_set_qf_name(sb, GRPQUOTA, &args[0]);
			if (ret)
				return ret;
			break;
		case Opt_prjjquota:
			ret = f2fs_set_qf_name(sb, PRJQUOTA, &args[0]);
			if (ret)
				return ret;
			break;
		case Opt_offusrjquota:
			ret = f2fs_clear_qf_name(sb, USRQUOTA);
			if (ret)
				return ret;
			break;
		case Opt_offgrpjquota:
			ret = f2fs_clear_qf_name(sb, GRPQUOTA);
			if (ret)
				return ret;
			break;
		case Opt_offprjjquota:
			ret = f2fs_clear_qf_name(sb, PRJQUOTA);
			if (ret)
				return ret;
			break;
		case Opt_jqfmt_vfsold:
762
			F2FS_OPTION(sbi).s_jquota_fmt = QFMT_VFS_OLD;
763 764
			break;
		case Opt_jqfmt_vfsv0:
765
			F2FS_OPTION(sbi).s_jquota_fmt = QFMT_VFS_V0;
766 767
			break;
		case Opt_jqfmt_vfsv1:
768
			F2FS_OPTION(sbi).s_jquota_fmt = QFMT_VFS_V1;
769 770 771 772 773 774 775
			break;
		case Opt_noquota:
			clear_opt(sbi, QUOTA);
			clear_opt(sbi, USRQUOTA);
			clear_opt(sbi, GRPQUOTA);
			clear_opt(sbi, PRJQUOTA);
			break;
776
#else
777
		case Opt_quota:
778 779
		case Opt_usrquota:
		case Opt_grpquota:
C
Chao Yu 已提交
780
		case Opt_prjquota:
781 782 783 784 785 786 787 788 789 790
		case Opt_usrjquota:
		case Opt_grpjquota:
		case Opt_prjjquota:
		case Opt_offusrjquota:
		case Opt_offgrpjquota:
		case Opt_offprjjquota:
		case Opt_jqfmt_vfsold:
		case Opt_jqfmt_vfsv0:
		case Opt_jqfmt_vfsv1:
		case Opt_noquota:
791
			f2fs_info(sbi, "quota operations not supported");
792 793
			break;
#endif
794 795 796 797
		case Opt_whint:
			name = match_strdup(&args[0]);
			if (!name)
				return -ENOMEM;
798
			if (!strcmp(name, "user-based")) {
799
				F2FS_OPTION(sbi).whint_mode = WHINT_MODE_USER;
800
			} else if (!strcmp(name, "off")) {
801
				F2FS_OPTION(sbi).whint_mode = WHINT_MODE_OFF;
802
			} else if (!strcmp(name, "fs-based")) {
803
				F2FS_OPTION(sbi).whint_mode = WHINT_MODE_FS;
804
			} else {
805
				kfree(name);
806 807
				return -EINVAL;
			}
808
			kfree(name);
809
			break;
810 811 812 813 814
		case Opt_alloc:
			name = match_strdup(&args[0]);
			if (!name)
				return -ENOMEM;

815
			if (!strcmp(name, "default")) {
816
				F2FS_OPTION(sbi).alloc_mode = ALLOC_MODE_DEFAULT;
817
			} else if (!strcmp(name, "reuse")) {
818
				F2FS_OPTION(sbi).alloc_mode = ALLOC_MODE_REUSE;
819
			} else {
820
				kfree(name);
821 822
				return -EINVAL;
			}
823
			kfree(name);
824
			break;
825 826 827 828
		case Opt_fsync:
			name = match_strdup(&args[0]);
			if (!name)
				return -ENOMEM;
829
			if (!strcmp(name, "posix")) {
830
				F2FS_OPTION(sbi).fsync_mode = FSYNC_MODE_POSIX;
831
			} else if (!strcmp(name, "strict")) {
832
				F2FS_OPTION(sbi).fsync_mode = FSYNC_MODE_STRICT;
833
			} else if (!strcmp(name, "nobarrier")) {
834 835
				F2FS_OPTION(sbi).fsync_mode =
							FSYNC_MODE_NOBARRIER;
836
			} else {
837
				kfree(name);
838 839
				return -EINVAL;
			}
840
			kfree(name);
841
			break;
842
		case Opt_test_dummy_encryption:
843 844 845 846
			ret = f2fs_set_test_dummy_encryption(sb, p, &args[0],
							     is_remount);
			if (ret)
				return ret;
847
			break;
848 849 850 851 852 853 854
		case Opt_inlinecrypt:
#ifdef CONFIG_FS_ENCRYPTION_INLINE_CRYPT
			sb->s_flags |= SB_INLINECRYPT;
#else
			f2fs_info(sbi, "inline encryption not supported");
#endif
			break;
855 856
		case Opt_checkpoint_disable_cap_perc:
			if (args->from && match_int(args, &arg))
857
				return -EINVAL;
858 859
			if (arg < 0 || arg > 100)
				return -EINVAL;
860
			F2FS_OPTION(sbi).unusable_cap_perc = arg;
861 862 863 864 865 866 867 868 869 870 871 872 873
			set_opt(sbi, DISABLE_CHECKPOINT);
			break;
		case Opt_checkpoint_disable_cap:
			if (args->from && match_int(args, &arg))
				return -EINVAL;
			F2FS_OPTION(sbi).unusable_cap = arg;
			set_opt(sbi, DISABLE_CHECKPOINT);
			break;
		case Opt_checkpoint_disable:
			set_opt(sbi, DISABLE_CHECKPOINT);
			break;
		case Opt_checkpoint_enable:
			clear_opt(sbi, DISABLE_CHECKPOINT);
874
			break;
875
#ifdef CONFIG_F2FS_FS_COMPRESSION
876 877
		case Opt_compress_algorithm:
			if (!f2fs_sb_has_compression(sbi)) {
878 879
				f2fs_info(sbi, "Image doesn't support compression");
				break;
880 881 882 883
			}
			name = match_strdup(&args[0]);
			if (!name)
				return -ENOMEM;
884
			if (!strcmp(name, "lzo")) {
885 886
				F2FS_OPTION(sbi).compress_algorithm =
								COMPRESS_LZO;
887
			} else if (!strcmp(name, "lz4")) {
888 889
				F2FS_OPTION(sbi).compress_algorithm =
								COMPRESS_LZ4;
890
			} else if (!strcmp(name, "zstd")) {
891 892
				F2FS_OPTION(sbi).compress_algorithm =
								COMPRESS_ZSTD;
893 894 895
			} else if (!strcmp(name, "lzo-rle")) {
				F2FS_OPTION(sbi).compress_algorithm =
								COMPRESS_LZORLE;
896 897 898 899 900 901 902 903
			} else {
				kfree(name);
				return -EINVAL;
			}
			kfree(name);
			break;
		case Opt_compress_log_size:
			if (!f2fs_sb_has_compression(sbi)) {
904 905
				f2fs_info(sbi, "Image doesn't support compression");
				break;
906 907 908 909 910 911 912 913 914 915 916 917 918
			}
			if (args->from && match_int(args, &arg))
				return -EINVAL;
			if (arg < MIN_COMPRESS_LOG_SIZE ||
				arg > MAX_COMPRESS_LOG_SIZE) {
				f2fs_err(sbi,
					"Compress cluster log size is out of range");
				return -EINVAL;
			}
			F2FS_OPTION(sbi).compress_log_size = arg;
			break;
		case Opt_compress_extension:
			if (!f2fs_sb_has_compression(sbi)) {
919 920
				f2fs_info(sbi, "Image doesn't support compression");
				break;
921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940
			}
			name = match_strdup(&args[0]);
			if (!name)
				return -ENOMEM;

			ext = F2FS_OPTION(sbi).extensions;
			ext_cnt = F2FS_OPTION(sbi).compress_ext_cnt;

			if (strlen(name) >= F2FS_EXTENSION_LEN ||
				ext_cnt >= COMPRESS_EXT_NUM) {
				f2fs_err(sbi,
					"invalid extension length/number");
				kfree(name);
				return -EINVAL;
			}

			strcpy(ext[ext_cnt], name);
			F2FS_OPTION(sbi).compress_ext_cnt++;
			kfree(name);
			break;
941 942 943
		case Opt_compress_chksum:
			F2FS_OPTION(sbi).compress_chksum = true;
			break;
944 945 946 947 948 949 950 951 952 953 954 955 956 957
		case Opt_compress_mode:
			name = match_strdup(&args[0]);
			if (!name)
				return -ENOMEM;
			if (!strcmp(name, "fs")) {
				F2FS_OPTION(sbi).compress_mode = COMPR_MODE_FS;
			} else if (!strcmp(name, "user")) {
				F2FS_OPTION(sbi).compress_mode = COMPR_MODE_USER;
			} else {
				kfree(name);
				return -EINVAL;
			}
			kfree(name);
			break;
958 959 960 961
#else
		case Opt_compress_algorithm:
		case Opt_compress_log_size:
		case Opt_compress_extension:
962
		case Opt_compress_chksum:
963
		case Opt_compress_mode:
964 965 966
			f2fs_info(sbi, "compression options not supported");
			break;
#endif
967 968 969
		case Opt_atgc:
			set_opt(sbi, ATGC);
			break;
970
		default:
971 972
			f2fs_err(sbi, "Unrecognized mount option \"%s\" or missing value",
				 p);
973 974 975
			return -EINVAL;
		}
	}
976 977 978
#ifdef CONFIG_QUOTA
	if (f2fs_check_quota_options(sbi))
		return -EINVAL;
979
#else
980
	if (f2fs_sb_has_quota_ino(sbi) && !f2fs_readonly(sbi->sb)) {
981
		f2fs_info(sbi, "Filesystem with quota feature cannot be mounted RDWR without CONFIG_QUOTA");
982 983
		return -EINVAL;
	}
984
	if (f2fs_sb_has_project_quota(sbi) && !f2fs_readonly(sbi->sb)) {
985
		f2fs_err(sbi, "Filesystem with project quota feature cannot be mounted RDWR without CONFIG_QUOTA");
986 987
		return -EINVAL;
	}
988
#endif
989 990 991 992 993 994 995
#ifndef CONFIG_UNICODE
	if (f2fs_sb_has_casefold(sbi)) {
		f2fs_err(sbi,
			"Filesystem with casefold feature cannot be mounted without CONFIG_UNICODE");
		return -EINVAL;
	}
#endif
996 997 998 999 1000 1001 1002 1003 1004 1005 1006
	/*
	 * The BLKZONED feature indicates that the drive was formatted with
	 * zone alignment optimization. This is optional for host-aware
	 * devices, but mandatory for host-managed zoned block devices.
	 */
#ifndef CONFIG_BLK_DEV_ZONED
	if (f2fs_sb_has_blkzoned(sbi)) {
		f2fs_err(sbi, "Zoned block device support is not enabled");
		return -EINVAL;
	}
#endif
1007

1008
	if (F2FS_IO_SIZE_BITS(sbi) && !f2fs_lfs_mode(sbi)) {
1009 1010
		f2fs_err(sbi, "Should set mode=lfs with %uKB-sized IO",
			 F2FS_IO_SIZE_KB(sbi));
1011 1012
		return -EINVAL;
	}
1013 1014

	if (test_opt(sbi, INLINE_XATTR_SIZE)) {
1015 1016
		int min_size, max_size;

1017 1018
		if (!f2fs_sb_has_extra_attr(sbi) ||
			!f2fs_sb_has_flexible_inline_xattr(sbi)) {
1019
			f2fs_err(sbi, "extra_attr or flexible_inline_xattr feature is off");
1020 1021
			return -EINVAL;
		}
1022
		if (!test_opt(sbi, INLINE_XATTR)) {
1023
			f2fs_err(sbi, "inline_xattr_size option should be set with inline_xattr option");
1024 1025
			return -EINVAL;
		}
1026 1027

		min_size = sizeof(struct f2fs_xattr_header) / sizeof(__le32);
1028
		max_size = MAX_INLINE_XATTR_SIZE;
1029 1030 1031

		if (F2FS_OPTION(sbi).inline_xattr_size < min_size ||
				F2FS_OPTION(sbi).inline_xattr_size > max_size) {
1032 1033
			f2fs_err(sbi, "inline xattr size is out of range: %d ~ %d",
				 min_size, max_size);
1034 1035 1036
			return -EINVAL;
		}
	}
1037

1038
	if (test_opt(sbi, DISABLE_CHECKPOINT) && f2fs_lfs_mode(sbi)) {
1039
		f2fs_err(sbi, "LFS not compatible with checkpoint=disable\n");
1040 1041 1042
		return -EINVAL;
	}

1043
	/* Not pass down write hints if the number of active logs is lesser
C
Chao Yu 已提交
1044
	 * than NR_CURSEG_PERSIST_TYPE.
1045
	 */
1046 1047
	if (F2FS_OPTION(sbi).active_logs != NR_CURSEG_TYPE)
		F2FS_OPTION(sbi).whint_mode = WHINT_MODE_OFF;
1048 1049 1050
	return 0;
}

1051 1052 1053 1054
static struct inode *f2fs_alloc_inode(struct super_block *sb)
{
	struct f2fs_inode_info *fi;

1055
	fi = kmem_cache_alloc(f2fs_inode_cachep, GFP_F2FS_ZERO);
1056 1057 1058 1059 1060
	if (!fi)
		return NULL;

	init_once((void *) fi);

1061
	/* Initialize f2fs-specific inode info */
1062
	atomic_set(&fi->dirty_pages, 0);
1063
	atomic_set(&fi->i_compr_blocks, 0);
1064
	init_rwsem(&fi->i_sem);
1065
	spin_lock_init(&fi->i_size_lock);
1066
	INIT_LIST_HEAD(&fi->dirty_list);
1067
	INIT_LIST_HEAD(&fi->gdirty_list);
1068
	INIT_LIST_HEAD(&fi->inmem_ilist);
J
Jaegeuk Kim 已提交
1069 1070
	INIT_LIST_HEAD(&fi->inmem_pages);
	mutex_init(&fi->inmem_lock);
1071 1072
	init_rwsem(&fi->i_gc_rwsem[READ]);
	init_rwsem(&fi->i_gc_rwsem[WRITE]);
1073
	init_rwsem(&fi->i_mmap_sem);
1074
	init_rwsem(&fi->i_xattr_sem);
1075

1076 1077
	/* Will be used by directory only */
	fi->i_dir_level = F2FS_SB(sb)->dir_level;
1078

1079 1080
	fi->ra_offset = -1;

1081 1082 1083
	return &fi->vfs_inode;
}

1084 1085
static int f2fs_drop_inode(struct inode *inode)
{
1086
	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
1087
	int ret;
1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100

	/*
	 * during filesystem shutdown, if checkpoint is disabled,
	 * drop useless meta/node dirty pages.
	 */
	if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED))) {
		if (inode->i_ino == F2FS_NODE_INO(sbi) ||
			inode->i_ino == F2FS_META_INO(sbi)) {
			trace_f2fs_drop_inode(inode, 1);
			return 1;
		}
	}

1101 1102 1103 1104 1105 1106 1107
	/*
	 * This is to avoid a deadlock condition like below.
	 * writeback_single_inode(inode)
	 *  - f2fs_write_data_page
	 *    - f2fs_gc -> iput -> evict
	 *       - inode_wait_for_writeback(inode)
	 */
1108
	if ((!inode_unhashed(inode) && inode->i_state & I_SYNC)) {
1109
		if (!inode->i_nlink && !is_bad_inode(inode)) {
1110 1111
			/* to avoid evict_inode call simultaneously */
			atomic_inc(&inode->i_count);
1112 1113 1114 1115
			spin_unlock(&inode->i_lock);

			/* some remained atomic pages should discarded */
			if (f2fs_is_atomic_file(inode))
1116
				f2fs_drop_inmem_pages(inode);
1117

1118 1119 1120
			/* should remain fi->extent_tree for writepage */
			f2fs_destroy_extent_node(inode);

1121
			sb_start_intwrite(inode->i_sb);
1122
			f2fs_i_size_write(inode, 0);
1123

1124 1125 1126 1127
			f2fs_submit_merged_write_cond(F2FS_I_SB(inode),
					inode, NULL, 0, DATA);
			truncate_inode_pages_final(inode->i_mapping);

1128
			if (F2FS_HAS_BLOCKS(inode))
1129
				f2fs_truncate(inode);
1130 1131 1132 1133

			sb_end_intwrite(inode->i_sb);

			spin_lock(&inode->i_lock);
1134
			atomic_dec(&inode->i_count);
1135
		}
1136
		trace_f2fs_drop_inode(inode, 0);
1137
		return 0;
1138
	}
1139
	ret = generic_drop_inode(inode);
1140 1141
	if (!ret)
		ret = fscrypt_drop_inode(inode);
1142 1143
	trace_f2fs_drop_inode(inode, ret);
	return ret;
1144 1145
}

1146
int f2fs_inode_dirtied(struct inode *inode, bool sync)
1147
{
1148
	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
1149
	int ret = 0;
1150 1151 1152

	spin_lock(&sbi->inode_lock[DIRTY_META]);
	if (is_inode_flag_set(inode, FI_DIRTY_INODE)) {
1153 1154 1155 1156
		ret = 1;
	} else {
		set_inode_flag(inode, FI_DIRTY_INODE);
		stat_inc_dirty_inode(sbi, DIRTY_META);
1157
	}
1158 1159
	if (sync && list_empty(&F2FS_I(inode)->gdirty_list)) {
		list_add_tail(&F2FS_I(inode)->gdirty_list,
1160
				&sbi->inode_list[DIRTY_META]);
1161 1162
		inc_page_count(sbi, F2FS_DIRTY_IMETA);
	}
1163
	spin_unlock(&sbi->inode_lock[DIRTY_META]);
1164
	return ret;
1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175
}

void f2fs_inode_synced(struct inode *inode)
{
	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);

	spin_lock(&sbi->inode_lock[DIRTY_META]);
	if (!is_inode_flag_set(inode, FI_DIRTY_INODE)) {
		spin_unlock(&sbi->inode_lock[DIRTY_META]);
		return;
	}
1176 1177 1178 1179
	if (!list_empty(&F2FS_I(inode)->gdirty_list)) {
		list_del_init(&F2FS_I(inode)->gdirty_list);
		dec_page_count(sbi, F2FS_DIRTY_IMETA);
	}
1180
	clear_inode_flag(inode, FI_DIRTY_INODE);
1181
	clear_inode_flag(inode, FI_AUTO_RECOVER);
1182
	stat_dec_dirty_inode(F2FS_I_SB(inode), DIRTY_META);
1183
	spin_unlock(&sbi->inode_lock[DIRTY_META]);
1184 1185
}

1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204
/*
 * f2fs_dirty_inode() is called from __mark_inode_dirty()
 *
 * We should call set_dirty_inode to write the dirty inode through write_inode.
 */
static void f2fs_dirty_inode(struct inode *inode, int flags)
{
	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);

	if (inode->i_ino == F2FS_NODE_INO(sbi) ||
			inode->i_ino == F2FS_META_INO(sbi))
		return;

	if (flags == I_DIRTY_TIME)
		return;

	if (is_inode_flag_set(inode, FI_AUTO_RECOVER))
		clear_inode_flag(inode, FI_AUTO_RECOVER);

1205
	f2fs_inode_dirtied(inode, false);
1206 1207
}

1208
static void f2fs_free_inode(struct inode *inode)
1209
{
1210
	fscrypt_free_inode(inode);
1211 1212 1213
	kmem_cache_free(f2fs_inode_cachep, F2FS_I(inode));
}

1214 1215
static void destroy_percpu_info(struct f2fs_sb_info *sbi)
{
1216
	percpu_counter_destroy(&sbi->alloc_valid_block_count);
1217
	percpu_counter_destroy(&sbi->total_valid_inode_count);
1218 1219
}

1220 1221 1222 1223 1224 1225 1226
static void destroy_device_list(struct f2fs_sb_info *sbi)
{
	int i;

	for (i = 0; i < sbi->s_ndevs; i++) {
		blkdev_put(FDEV(i).bdev, FMODE_EXCL);
#ifdef CONFIG_BLK_DEV_ZONED
1227
		kvfree(FDEV(i).blkz_seq);
1228
		kfree(FDEV(i).zone_capacity_blocks);
1229 1230
#endif
	}
1231
	kvfree(sbi->devs);
1232 1233
}

1234 1235 1236
static void f2fs_put_super(struct super_block *sb)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
C
Chao Yu 已提交
1237
	int i;
1238
	bool dropped;
1239

1240 1241 1242
	/* unregister procfs/sysfs entries in advance to avoid race case */
	f2fs_unregister_sysfs(sbi);

1243
	f2fs_quota_off_umount(sb);
1244

1245 1246 1247
	/* prevent remaining shrinker jobs */
	mutex_lock(&sbi->umount_mutex);

1248 1249 1250 1251 1252
	/*
	 * We don't need to do checkpoint when superblock is clean.
	 * But, the previous checkpoint was not done by umount, it needs to do
	 * clean checkpoint again.
	 */
1253 1254
	if ((is_sbi_flag_set(sbi, SBI_IS_DIRTY) ||
			!is_set_ckpt_flags(sbi, CP_UMOUNT_FLAG))) {
1255 1256 1257
		struct cp_control cpc = {
			.reason = CP_UMOUNT,
		};
1258
		f2fs_write_checkpoint(sbi, &cpc);
1259
	}
1260

1261
	/* be sure to wait for any on-going discard commands */
1262
	dropped = f2fs_issue_discard_timeout(sbi);
1263

1264 1265
	if ((f2fs_hw_support_discard(sbi) || f2fs_hw_should_discard(sbi)) &&
					!sbi->discard_blks && !dropped) {
1266 1267 1268
		struct cp_control cpc = {
			.reason = CP_UMOUNT | CP_TRIMMED,
		};
1269
		f2fs_write_checkpoint(sbi, &cpc);
1270 1271
	}

1272 1273 1274 1275
	/*
	 * normally superblock is clean, so we need to release this.
	 * In addition, EIO will skip do checkpoint, we need this as well.
	 */
1276
	f2fs_release_ino_entry(sbi, true);
1277

1278 1279 1280
	f2fs_leave_shrinker(sbi);
	mutex_unlock(&sbi->umount_mutex);

1281
	/* our cp_error case, we can wait for any writeback page */
1282
	f2fs_flush_merged_writes(sbi);
1283

1284
	f2fs_wait_on_all_pages(sbi, F2FS_WB_CP_DATA);
1285 1286 1287

	f2fs_bug_on(sbi, sbi->fsync_node_num);

1288
	iput(sbi->node_inode);
1289 1290
	sbi->node_inode = NULL;

1291
	iput(sbi->meta_inode);
1292
	sbi->meta_inode = NULL;
1293

1294 1295 1296 1297 1298 1299
	/*
	 * iput() can update stat information, if f2fs_write_checkpoint()
	 * above failed with error.
	 */
	f2fs_destroy_stats(sbi);

1300
	/* destroy f2fs internal modules */
1301 1302
	f2fs_destroy_node_manager(sbi);
	f2fs_destroy_segment_manager(sbi);
1303

1304 1305
	f2fs_destroy_post_read_wq(sbi);

1306
	kvfree(sbi->ckpt);
C
Chao Yu 已提交
1307

1308
	sb->s_fs_info = NULL;
1309 1310
	if (sbi->s_chksum_driver)
		crypto_free_shash(sbi->s_chksum_driver);
1311
	kfree(sbi->raw_super);
1312

1313
	destroy_device_list(sbi);
1314
	f2fs_destroy_page_array_cache(sbi);
1315
	f2fs_destroy_xattr_caches(sbi);
1316
	mempool_destroy(sbi->write_io_dummy);
1317 1318
#ifdef CONFIG_QUOTA
	for (i = 0; i < MAXQUOTAS; i++)
1319
		kfree(F2FS_OPTION(sbi).s_qf_names[i]);
1320
#endif
1321
	fscrypt_free_dummy_policy(&F2FS_OPTION(sbi).dummy_enc_policy);
1322
	destroy_percpu_info(sbi);
J
Jaegeuk Kim 已提交
1323
	for (i = 0; i < NR_PAGE_TYPE; i++)
1324
		kvfree(sbi->write_io[i]);
1325
#ifdef CONFIG_UNICODE
1326
	utf8_unload(sb->s_encoding);
1327
#endif
1328
	kfree(sbi);
1329 1330 1331 1332 1333
}

int f2fs_sync_fs(struct super_block *sb, int sync)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
1334
	int err = 0;
1335

1336 1337
	if (unlikely(f2fs_cp_error(sbi)))
		return 0;
1338 1339
	if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED)))
		return 0;
1340

1341 1342
	trace_f2fs_sync_fs(sb, sync);

1343 1344 1345
	if (unlikely(is_sbi_flag_set(sbi, SBI_POR_DOING)))
		return -EAGAIN;

1346
	if (sync) {
1347 1348
		struct cp_control cpc;

1349 1350
		cpc.reason = __get_cp_reason(sbi);

1351
		down_write(&sbi->gc_lock);
1352
		err = f2fs_write_checkpoint(sbi, &cpc);
1353
		up_write(&sbi->gc_lock);
1354
	}
1355
	f2fs_trace_ios(NULL, 1);
1356

1357
	return err;
1358 1359
}

1360 1361
static int f2fs_freeze(struct super_block *sb)
{
J
Jaegeuk Kim 已提交
1362
	if (f2fs_readonly(sb))
1363 1364
		return 0;

1365 1366 1367 1368 1369 1370 1371 1372
	/* IO error happened before */
	if (unlikely(f2fs_cp_error(F2FS_SB(sb))))
		return -EIO;

	/* must be clean, since sync_filesystem() was already called */
	if (is_sbi_flag_set(F2FS_SB(sb), SBI_IS_DIRTY))
		return -EINVAL;
	return 0;
1373 1374 1375 1376 1377 1378 1379
}

static int f2fs_unfreeze(struct super_block *sb)
{
	return 0;
}

1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392
#ifdef CONFIG_QUOTA
static int f2fs_statfs_project(struct super_block *sb,
				kprojid_t projid, struct kstatfs *buf)
{
	struct kqid qid;
	struct dquot *dquot;
	u64 limit;
	u64 curblock;

	qid = make_kqid_projid(projid);
	dquot = dqget(sb, qid);
	if (IS_ERR(dquot))
		return PTR_ERR(dquot);
1393
	spin_lock(&dquot->dq_dqb_lock);
1394

1395 1396
	limit = min_not_zero(dquot->dq_dqb.dqb_bsoftlimit,
					dquot->dq_dqb.dqb_bhardlimit);
1397 1398
	if (limit)
		limit >>= sb->s_blocksize_bits;
1399

1400
	if (limit && buf->f_blocks > limit) {
1401 1402
		curblock = (dquot->dq_dqb.dqb_curspace +
			    dquot->dq_dqb.dqb_rsvspace) >> sb->s_blocksize_bits;
1403 1404 1405 1406 1407 1408
		buf->f_blocks = limit;
		buf->f_bfree = buf->f_bavail =
			(buf->f_blocks > curblock) ?
			 (buf->f_blocks - curblock) : 0;
	}

1409 1410
	limit = min_not_zero(dquot->dq_dqb.dqb_isoftlimit,
					dquot->dq_dqb.dqb_ihardlimit);
1411

1412 1413 1414 1415 1416 1417 1418
	if (limit && buf->f_files > limit) {
		buf->f_files = limit;
		buf->f_ffree =
			(buf->f_files > dquot->dq_dqb.dqb_curinodes) ?
			 (buf->f_files - dquot->dq_dqb.dqb_curinodes) : 0;
	}

1419
	spin_unlock(&dquot->dq_dqb_lock);
1420 1421 1422 1423 1424
	dqput(dquot);
	return 0;
}
#endif

1425 1426 1427 1428 1429
static int f2fs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
	struct super_block *sb = dentry->d_sb;
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	u64 id = huge_encode_dev(sb->s_bdev->bd_dev);
1430
	block_t total_count, user_block_count, start_count;
1431
	u64 avail_node_count;
1432 1433 1434 1435 1436 1437 1438 1439

	total_count = le64_to_cpu(sbi->raw_super->block_count);
	user_block_count = sbi->user_block_count;
	start_count = le32_to_cpu(sbi->raw_super->segment0_blkaddr);
	buf->f_type = F2FS_SUPER_MAGIC;
	buf->f_bsize = sbi->blocksize;

	buf->f_blocks = total_count - start_count;
1440
	buf->f_bfree = user_block_count - valid_user_blocks(sbi) -
1441
						sbi->current_reserved_blocks;
1442 1443

	spin_lock(&sbi->stat_lock);
1444 1445 1446 1447
	if (unlikely(buf->f_bfree <= sbi->unusable_block_count))
		buf->f_bfree = 0;
	else
		buf->f_bfree -= sbi->unusable_block_count;
1448
	spin_unlock(&sbi->stat_lock);
1449

1450 1451 1452
	if (buf->f_bfree > F2FS_OPTION(sbi).root_reserved_blocks)
		buf->f_bavail = buf->f_bfree -
				F2FS_OPTION(sbi).root_reserved_blocks;
1453 1454
	else
		buf->f_bavail = 0;
1455

1456
	avail_node_count = sbi->total_node_count - F2FS_RESERVED_NODE_NUM;
1457 1458 1459 1460 1461 1462 1463 1464 1465

	if (avail_node_count > user_block_count) {
		buf->f_files = user_block_count;
		buf->f_ffree = buf->f_bavail;
	} else {
		buf->f_files = avail_node_count;
		buf->f_ffree = min(avail_node_count - valid_node_count(sbi),
					buf->f_bavail);
	}
1466

1467
	buf->f_namelen = F2FS_NAME_LEN;
1468
	buf->f_fsid    = u64_to_fsid(id);
1469

1470 1471 1472 1473 1474 1475
#ifdef CONFIG_QUOTA
	if (is_inode_flag_set(dentry->d_inode, FI_PROJ_INHERIT) &&
			sb_has_quota_limits_enabled(sb, PRJQUOTA)) {
		f2fs_statfs_project(sb, F2FS_I(dentry->d_inode)->i_projid, buf);
	}
#endif
1476 1477 1478
	return 0;
}

1479 1480 1481 1482 1483 1484
static inline void f2fs_show_quota_options(struct seq_file *seq,
					   struct super_block *sb)
{
#ifdef CONFIG_QUOTA
	struct f2fs_sb_info *sbi = F2FS_SB(sb);

1485
	if (F2FS_OPTION(sbi).s_jquota_fmt) {
1486 1487
		char *fmtname = "";

1488
		switch (F2FS_OPTION(sbi).s_jquota_fmt) {
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501
		case QFMT_VFS_OLD:
			fmtname = "vfsold";
			break;
		case QFMT_VFS_V0:
			fmtname = "vfsv0";
			break;
		case QFMT_VFS_V1:
			fmtname = "vfsv1";
			break;
		}
		seq_printf(seq, ",jqfmt=%s", fmtname);
	}

1502 1503 1504
	if (F2FS_OPTION(sbi).s_qf_names[USRQUOTA])
		seq_show_option(seq, "usrjquota",
			F2FS_OPTION(sbi).s_qf_names[USRQUOTA]);
1505

1506 1507 1508
	if (F2FS_OPTION(sbi).s_qf_names[GRPQUOTA])
		seq_show_option(seq, "grpjquota",
			F2FS_OPTION(sbi).s_qf_names[GRPQUOTA]);
1509

1510 1511 1512
	if (F2FS_OPTION(sbi).s_qf_names[PRJQUOTA])
		seq_show_option(seq, "prjjquota",
			F2FS_OPTION(sbi).s_qf_names[PRJQUOTA]);
1513 1514 1515
#endif
}

1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532
static inline void f2fs_show_compress_options(struct seq_file *seq,
							struct super_block *sb)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	char *algtype = "";
	int i;

	if (!f2fs_sb_has_compression(sbi))
		return;

	switch (F2FS_OPTION(sbi).compress_algorithm) {
	case COMPRESS_LZO:
		algtype = "lzo";
		break;
	case COMPRESS_LZ4:
		algtype = "lz4";
		break;
1533 1534 1535
	case COMPRESS_ZSTD:
		algtype = "zstd";
		break;
1536 1537 1538
	case COMPRESS_LZORLE:
		algtype = "lzo-rle";
		break;
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548
	}
	seq_printf(seq, ",compress_algorithm=%s", algtype);

	seq_printf(seq, ",compress_log_size=%u",
			F2FS_OPTION(sbi).compress_log_size);

	for (i = 0; i < F2FS_OPTION(sbi).compress_ext_cnt; i++) {
		seq_printf(seq, ",compress_extension=%s",
			F2FS_OPTION(sbi).extensions[i]);
	}
1549 1550 1551

	if (F2FS_OPTION(sbi).compress_chksum)
		seq_puts(seq, ",compress_chksum");
1552 1553 1554 1555 1556

	if (F2FS_OPTION(sbi).compress_mode == COMPR_MODE_FS)
		seq_printf(seq, ",compress_mode=%s", "fs");
	else if (F2FS_OPTION(sbi).compress_mode == COMPR_MODE_USER)
		seq_printf(seq, ",compress_mode=%s", "user");
1557 1558
}

1559 1560 1561 1562
static int f2fs_show_options(struct seq_file *seq, struct dentry *root)
{
	struct f2fs_sb_info *sbi = F2FS_SB(root->d_sb);

1563 1564 1565 1566 1567
	if (F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_SYNC)
		seq_printf(seq, ",background_gc=%s", "sync");
	else if (F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_ON)
		seq_printf(seq, ",background_gc=%s", "on");
	else if (F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_OFF)
1568
		seq_printf(seq, ",background_gc=%s", "off");
1569

1570 1571
	if (test_opt(sbi, DISABLE_ROLL_FORWARD))
		seq_puts(seq, ",disable_roll_forward");
1572 1573
	if (test_opt(sbi, NORECOVERY))
		seq_puts(seq, ",norecovery");
1574 1575
	if (test_opt(sbi, DISCARD))
		seq_puts(seq, ",discard");
1576 1577
	else
		seq_puts(seq, ",nodiscard");
1578
	if (test_opt(sbi, NOHEAP))
1579 1580 1581
		seq_puts(seq, ",no_heap");
	else
		seq_puts(seq, ",heap");
1582 1583 1584 1585 1586
#ifdef CONFIG_F2FS_FS_XATTR
	if (test_opt(sbi, XATTR_USER))
		seq_puts(seq, ",user_xattr");
	else
		seq_puts(seq, ",nouser_xattr");
1587 1588
	if (test_opt(sbi, INLINE_XATTR))
		seq_puts(seq, ",inline_xattr");
1589 1590
	else
		seq_puts(seq, ",noinline_xattr");
1591 1592
	if (test_opt(sbi, INLINE_XATTR_SIZE))
		seq_printf(seq, ",inline_xattr_size=%u",
1593
					F2FS_OPTION(sbi).inline_xattr_size);
1594 1595 1596 1597 1598 1599 1600 1601
#endif
#ifdef CONFIG_F2FS_FS_POSIX_ACL
	if (test_opt(sbi, POSIX_ACL))
		seq_puts(seq, ",acl");
	else
		seq_puts(seq, ",noacl");
#endif
	if (test_opt(sbi, DISABLE_EXT_IDENTIFY))
1602
		seq_puts(seq, ",disable_ext_identify");
1603 1604
	if (test_opt(sbi, INLINE_DATA))
		seq_puts(seq, ",inline_data");
1605 1606
	else
		seq_puts(seq, ",noinline_data");
1607 1608
	if (test_opt(sbi, INLINE_DENTRY))
		seq_puts(seq, ",inline_dentry");
1609 1610
	else
		seq_puts(seq, ",noinline_dentry");
1611
	if (!f2fs_readonly(sbi->sb) && test_opt(sbi, FLUSH_MERGE))
1612
		seq_puts(seq, ",flush_merge");
1613 1614
	if (test_opt(sbi, NOBARRIER))
		seq_puts(seq, ",nobarrier");
1615 1616
	if (test_opt(sbi, FASTBOOT))
		seq_puts(seq, ",fastboot");
1617 1618
	if (test_opt(sbi, EXTENT_CACHE))
		seq_puts(seq, ",extent_cache");
1619 1620
	else
		seq_puts(seq, ",noextent_cache");
1621 1622
	if (test_opt(sbi, DATA_FLUSH))
		seq_puts(seq, ",data_flush");
1623 1624

	seq_puts(seq, ",mode=");
1625
	if (F2FS_OPTION(sbi).fs_mode == FS_MODE_ADAPTIVE)
1626
		seq_puts(seq, "adaptive");
1627
	else if (F2FS_OPTION(sbi).fs_mode == FS_MODE_LFS)
1628
		seq_puts(seq, "lfs");
1629
	seq_printf(seq, ",active_logs=%u", F2FS_OPTION(sbi).active_logs);
1630
	if (test_opt(sbi, RESERVE_ROOT))
1631
		seq_printf(seq, ",reserve_root=%u,resuid=%u,resgid=%u",
1632 1633 1634 1635 1636
				F2FS_OPTION(sbi).root_reserved_blocks,
				from_kuid_munged(&init_user_ns,
					F2FS_OPTION(sbi).s_resuid),
				from_kgid_munged(&init_user_ns,
					F2FS_OPTION(sbi).s_resgid));
1637
	if (F2FS_IO_SIZE_BITS(sbi))
1638 1639
		seq_printf(seq, ",io_bits=%u",
				F2FS_OPTION(sbi).write_io_size_bits);
1640
#ifdef CONFIG_F2FS_FAULT_INJECTION
1641
	if (test_opt(sbi, FAULT_INJECTION)) {
1642
		seq_printf(seq, ",fault_injection=%u",
1643
				F2FS_OPTION(sbi).fault_info.inject_rate);
1644 1645 1646
		seq_printf(seq, ",fault_type=%u",
				F2FS_OPTION(sbi).fault_info.inject_type);
	}
1647
#endif
1648
#ifdef CONFIG_QUOTA
1649 1650
	if (test_opt(sbi, QUOTA))
		seq_puts(seq, ",quota");
1651 1652 1653 1654
	if (test_opt(sbi, USRQUOTA))
		seq_puts(seq, ",usrquota");
	if (test_opt(sbi, GRPQUOTA))
		seq_puts(seq, ",grpquota");
C
Chao Yu 已提交
1655 1656
	if (test_opt(sbi, PRJQUOTA))
		seq_puts(seq, ",prjquota");
1657
#endif
1658
	f2fs_show_quota_options(seq, sbi->sb);
1659
	if (F2FS_OPTION(sbi).whint_mode == WHINT_MODE_USER)
1660
		seq_printf(seq, ",whint_mode=%s", "user-based");
1661
	else if (F2FS_OPTION(sbi).whint_mode == WHINT_MODE_FS)
1662
		seq_printf(seq, ",whint_mode=%s", "fs-based");
1663 1664

	fscrypt_show_test_dummy_encryption(seq, ',', sbi->sb);
1665

1666 1667 1668
	if (sbi->sb->s_flags & SB_INLINECRYPT)
		seq_puts(seq, ",inlinecrypt");

1669
	if (F2FS_OPTION(sbi).alloc_mode == ALLOC_MODE_DEFAULT)
1670
		seq_printf(seq, ",alloc_mode=%s", "default");
1671
	else if (F2FS_OPTION(sbi).alloc_mode == ALLOC_MODE_REUSE)
1672
		seq_printf(seq, ",alloc_mode=%s", "reuse");
1673

1674
	if (test_opt(sbi, DISABLE_CHECKPOINT))
1675 1676
		seq_printf(seq, ",checkpoint=disable:%u",
				F2FS_OPTION(sbi).unusable_cap);
1677
	if (F2FS_OPTION(sbi).fsync_mode == FSYNC_MODE_POSIX)
1678
		seq_printf(seq, ",fsync_mode=%s", "posix");
1679
	else if (F2FS_OPTION(sbi).fsync_mode == FSYNC_MODE_STRICT)
1680
		seq_printf(seq, ",fsync_mode=%s", "strict");
1681 1682
	else if (F2FS_OPTION(sbi).fsync_mode == FSYNC_MODE_NOBARRIER)
		seq_printf(seq, ",fsync_mode=%s", "nobarrier");
1683

1684
#ifdef CONFIG_F2FS_FS_COMPRESSION
1685
	f2fs_show_compress_options(seq, sbi->sb);
1686
#endif
1687 1688 1689

	if (test_opt(sbi, ATGC))
		seq_puts(seq, ",atgc");
1690 1691 1692
	return 0;
}

1693 1694 1695
static void default_options(struct f2fs_sb_info *sbi)
{
	/* init some FS parameters */
C
Chao Yu 已提交
1696
	F2FS_OPTION(sbi).active_logs = NR_CURSEG_PERSIST_TYPE;
1697 1698 1699 1700
	F2FS_OPTION(sbi).inline_xattr_size = DEFAULT_INLINE_XATTR_ADDRS;
	F2FS_OPTION(sbi).whint_mode = WHINT_MODE_OFF;
	F2FS_OPTION(sbi).alloc_mode = ALLOC_MODE_DEFAULT;
	F2FS_OPTION(sbi).fsync_mode = FSYNC_MODE_POSIX;
1701 1702
	F2FS_OPTION(sbi).s_resuid = make_kuid(&init_user_ns, F2FS_DEF_RESUID);
	F2FS_OPTION(sbi).s_resgid = make_kgid(&init_user_ns, F2FS_DEF_RESGID);
1703
	F2FS_OPTION(sbi).compress_algorithm = COMPRESS_LZ4;
1704 1705
	F2FS_OPTION(sbi).compress_log_size = MIN_COMPRESS_LOG_SIZE;
	F2FS_OPTION(sbi).compress_ext_cnt = 0;
1706
	F2FS_OPTION(sbi).compress_mode = COMPR_MODE_FS;
1707
	F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_ON;
1708

1709 1710
	sbi->sb->s_flags &= ~SB_INLINECRYPT;

1711
	set_opt(sbi, INLINE_XATTR);
1712
	set_opt(sbi, INLINE_DATA);
1713
	set_opt(sbi, INLINE_DENTRY);
1714
	set_opt(sbi, EXTENT_CACHE);
1715
	set_opt(sbi, NOHEAP);
1716
	clear_opt(sbi, DISABLE_CHECKPOINT);
1717
	F2FS_OPTION(sbi).unusable_cap = 0;
1718
	sbi->sb->s_flags |= SB_LAZYTIME;
1719
	set_opt(sbi, FLUSH_MERGE);
1720
	set_opt(sbi, DISCARD);
1721
	if (f2fs_sb_has_blkzoned(sbi))
1722
		F2FS_OPTION(sbi).fs_mode = FS_MODE_LFS;
1723
	else
1724
		F2FS_OPTION(sbi).fs_mode = FS_MODE_ADAPTIVE;
1725 1726 1727 1728 1729 1730 1731

#ifdef CONFIG_F2FS_FS_XATTR
	set_opt(sbi, XATTR_USER);
#endif
#ifdef CONFIG_F2FS_FS_POSIX_ACL
	set_opt(sbi, POSIX_ACL);
#endif
1732

1733
	f2fs_build_fault_attr(sbi, 0, 0);
1734 1735
}

1736 1737 1738
#ifdef CONFIG_QUOTA
static int f2fs_enable_quotas(struct super_block *sb);
#endif
1739 1740 1741

static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi)
{
1742
	unsigned int s_flags = sbi->sb->s_flags;
1743
	struct cp_control cpc;
1744 1745
	int err = 0;
	int ret;
1746
	block_t unusable;
1747

1748
	if (s_flags & SB_RDONLY) {
1749
		f2fs_err(sbi, "checkpoint=disable on readonly fs");
1750 1751
		return -EINVAL;
	}
1752 1753 1754 1755 1756
	sbi->sb->s_flags |= SB_ACTIVE;

	f2fs_update_time(sbi, DISABLE_TIME);

	while (!f2fs_time_over(sbi, DISABLE_TIME)) {
1757
		down_write(&sbi->gc_lock);
1758
		err = f2fs_gc(sbi, true, false, NULL_SEGNO);
1759 1760
		if (err == -ENODATA) {
			err = 0;
1761
			break;
1762
		}
1763
		if (err && err != -EAGAIN)
1764
			break;
1765 1766
	}

1767 1768 1769 1770 1771
	ret = sync_filesystem(sbi->sb);
	if (ret || err) {
		err = ret ? ret: err;
		goto restore_flag;
	}
1772

1773 1774
	unusable = f2fs_get_unusable_blocks(sbi);
	if (f2fs_disable_cp_again(sbi, unusable)) {
1775 1776 1777
		err = -EAGAIN;
		goto restore_flag;
	}
1778

1779
	down_write(&sbi->gc_lock);
1780 1781
	cpc.reason = CP_PAUSE;
	set_sbi_flag(sbi, SBI_CP_DISABLED);
1782 1783 1784
	err = f2fs_write_checkpoint(sbi, &cpc);
	if (err)
		goto out_unlock;
1785

1786
	spin_lock(&sbi->stat_lock);
1787
	sbi->unusable_block_count = unusable;
1788 1789
	spin_unlock(&sbi->stat_lock);

1790
out_unlock:
1791
	up_write(&sbi->gc_lock);
1792
restore_flag:
1793
	sbi->sb->s_flags = s_flags;	/* Restore SB_RDONLY status */
1794
	return err;
1795 1796 1797 1798
}

static void f2fs_enable_checkpoint(struct f2fs_sb_info *sbi)
{
1799
	down_write(&sbi->gc_lock);
1800 1801 1802 1803
	f2fs_dirty_to_prefree(sbi);

	clear_sbi_flag(sbi, SBI_CP_DISABLED);
	set_sbi_flag(sbi, SBI_IS_DIRTY);
1804
	up_write(&sbi->gc_lock);
1805 1806 1807 1808

	f2fs_sync_fs(sbi->sb, 1);
}

1809 1810 1811 1812
static int f2fs_remount(struct super_block *sb, int *flags, char *data)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	struct f2fs_mount_info org_mount_opt;
1813
	unsigned long old_sb_flags;
1814
	int err;
1815 1816
	bool need_restart_gc = false;
	bool need_stop_gc = false;
1817
	bool no_extent_cache = !test_opt(sbi, EXTENT_CACHE);
1818
	bool disable_checkpoint = test_opt(sbi, DISABLE_CHECKPOINT);
1819
	bool no_io_align = !F2FS_IO_ALIGNED(sbi);
1820
	bool no_atgc = !test_opt(sbi, ATGC);
1821
	bool checkpoint_changed;
1822 1823 1824
#ifdef CONFIG_QUOTA
	int i, j;
#endif
1825 1826 1827 1828 1829 1830

	/*
	 * Save the old mount options in case we
	 * need to restore them.
	 */
	org_mount_opt = sbi->mount_opt;
1831
	old_sb_flags = sb->s_flags;
1832

1833
#ifdef CONFIG_QUOTA
1834
	org_mount_opt.s_jquota_fmt = F2FS_OPTION(sbi).s_jquota_fmt;
1835
	for (i = 0; i < MAXQUOTAS; i++) {
1836 1837 1838 1839 1840
		if (F2FS_OPTION(sbi).s_qf_names[i]) {
			org_mount_opt.s_qf_names[i] =
				kstrdup(F2FS_OPTION(sbi).s_qf_names[i],
				GFP_KERNEL);
			if (!org_mount_opt.s_qf_names[i]) {
1841
				for (j = 0; j < i; j++)
1842
					kfree(org_mount_opt.s_qf_names[j]);
1843 1844 1845
				return -ENOMEM;
			}
		} else {
1846
			org_mount_opt.s_qf_names[i] = NULL;
1847 1848 1849 1850
		}
	}
#endif

1851
	/* recover superblocks we couldn't write due to previous RO mount */
1852
	if (!(*flags & SB_RDONLY) && is_sbi_flag_set(sbi, SBI_NEED_SB_WRITE)) {
1853
		err = f2fs_commit_super(sbi, false);
1854 1855
		f2fs_info(sbi, "Try to recover all the superblocks, ret: %d",
			  err);
1856 1857 1858 1859
		if (!err)
			clear_sbi_flag(sbi, SBI_NEED_SB_WRITE);
	}

1860
	default_options(sbi);
1861

1862
	/* parse mount options */
1863
	err = parse_options(sb, data, true);
1864 1865
	if (err)
		goto restore_opts;
1866 1867
	checkpoint_changed =
			disable_checkpoint != test_opt(sbi, DISABLE_CHECKPOINT);
1868 1869 1870

	/*
	 * Previous and new state of filesystem is RO,
1871
	 * so skip checking GC and FLUSH_MERGE conditions.
1872
	 */
1873
	if (f2fs_readonly(sb) && (*flags & SB_RDONLY))
1874 1875
		goto skip;

1876
#ifdef CONFIG_QUOTA
1877
	if (!f2fs_readonly(sb) && (*flags & SB_RDONLY)) {
1878 1879 1880
		err = dquot_suspend(sb, -1);
		if (err < 0)
			goto restore_opts;
1881
	} else if (f2fs_readonly(sb) && !(*flags & SB_RDONLY)) {
1882
		/* dquot_resume needs RW */
1883
		sb->s_flags &= ~SB_RDONLY;
1884 1885
		if (sb_any_quota_suspended(sb)) {
			dquot_resume(sb, -1);
1886
		} else if (f2fs_sb_has_quota_ino(sbi)) {
1887 1888 1889 1890
			err = f2fs_enable_quotas(sb);
			if (err)
				goto restore_opts;
		}
1891
	}
1892
#endif
1893 1894 1895 1896 1897 1898 1899
	/* disallow enable atgc dynamically */
	if (no_atgc == !!test_opt(sbi, ATGC)) {
		err = -EINVAL;
		f2fs_warn(sbi, "switch atgc option is not allowed");
		goto restore_opts;
	}

1900 1901 1902
	/* disallow enable/disable extent_cache dynamically */
	if (no_extent_cache == !!test_opt(sbi, EXTENT_CACHE)) {
		err = -EINVAL;
1903
		f2fs_warn(sbi, "switch extent_cache option is not allowed");
1904 1905 1906
		goto restore_opts;
	}

1907 1908 1909 1910 1911 1912
	if (no_io_align == !!F2FS_IO_ALIGNED(sbi)) {
		err = -EINVAL;
		f2fs_warn(sbi, "switch io_bits option is not allowed");
		goto restore_opts;
	}

1913 1914
	if ((*flags & SB_RDONLY) && test_opt(sbi, DISABLE_CHECKPOINT)) {
		err = -EINVAL;
1915
		f2fs_warn(sbi, "disabling checkpoint not compatible with read-only");
1916 1917 1918
		goto restore_opts;
	}

1919 1920 1921 1922 1923
	/*
	 * We stop the GC thread if FS is mounted as RO
	 * or if background_gc = off is passed in mount
	 * option. Also sync the filesystem.
	 */
1924 1925
	if ((*flags & SB_RDONLY) ||
			F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_OFF) {
1926
		if (sbi->gc_thread) {
1927
			f2fs_stop_gc_thread(sbi);
1928
			need_restart_gc = true;
1929
		}
1930
	} else if (!sbi->gc_thread) {
1931
		err = f2fs_start_gc_thread(sbi);
1932 1933
		if (err)
			goto restore_opts;
1934 1935 1936
		need_stop_gc = true;
	}

1937 1938
	if (*flags & SB_RDONLY ||
		F2FS_OPTION(sbi).whint_mode != org_mount_opt.whint_mode) {
1939 1940 1941 1942 1943 1944 1945 1946
		sync_inodes_sb(sb);

		set_sbi_flag(sbi, SBI_IS_DIRTY);
		set_sbi_flag(sbi, SBI_IS_CLOSE);
		f2fs_sync_fs(sb, 1);
		clear_sbi_flag(sbi, SBI_IS_CLOSE);
	}

1947 1948 1949 1950 1951 1952 1953 1954 1955 1956
	if (checkpoint_changed) {
		if (test_opt(sbi, DISABLE_CHECKPOINT)) {
			err = f2fs_disable_checkpoint(sbi);
			if (err)
				goto restore_gc;
		} else {
			f2fs_enable_checkpoint(sbi);
		}
	}

1957 1958 1959 1960
	/*
	 * We stop issue flush thread if FS is mounted as RO
	 * or if flush_merge is not passed in mount option.
	 */
1961
	if ((*flags & SB_RDONLY) || !test_opt(sbi, FLUSH_MERGE)) {
1962
		clear_opt(sbi, FLUSH_MERGE);
1963
		f2fs_destroy_flush_cmd_control(sbi, false);
1964
	} else {
1965
		err = f2fs_create_flush_cmd_control(sbi);
1966
		if (err)
1967
			goto restore_gc;
1968 1969
	}
skip:
1970 1971 1972
#ifdef CONFIG_QUOTA
	/* Release old quota file names */
	for (i = 0; i < MAXQUOTAS; i++)
1973
		kfree(org_mount_opt.s_qf_names[i]);
1974
#endif
1975
	/* Update the POSIXACL Flag */
1976 1977
	sb->s_flags = (sb->s_flags & ~SB_POSIXACL) |
		(test_opt(sbi, POSIX_ACL) ? SB_POSIXACL : 0);
1978

1979
	limit_reserve_root(sbi);
1980
	adjust_unusable_cap_perc(sbi);
1981
	*flags = (*flags & ~SB_LAZYTIME) | (sb->s_flags & SB_LAZYTIME);
1982
	return 0;
1983 1984
restore_gc:
	if (need_restart_gc) {
1985
		if (f2fs_start_gc_thread(sbi))
1986
			f2fs_warn(sbi, "background gc thread has stopped");
1987
	} else if (need_stop_gc) {
1988
		f2fs_stop_gc_thread(sbi);
1989
	}
1990
restore_opts:
1991
#ifdef CONFIG_QUOTA
1992
	F2FS_OPTION(sbi).s_jquota_fmt = org_mount_opt.s_jquota_fmt;
1993
	for (i = 0; i < MAXQUOTAS; i++) {
1994
		kfree(F2FS_OPTION(sbi).s_qf_names[i]);
1995
		F2FS_OPTION(sbi).s_qf_names[i] = org_mount_opt.s_qf_names[i];
1996 1997
	}
#endif
1998
	sbi->mount_opt = org_mount_opt;
1999
	sb->s_flags = old_sb_flags;
2000 2001 2002
	return err;
}

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026
#ifdef CONFIG_QUOTA
/* Read data from quotafile */
static ssize_t f2fs_quota_read(struct super_block *sb, int type, char *data,
			       size_t len, loff_t off)
{
	struct inode *inode = sb_dqopt(sb)->files[type];
	struct address_space *mapping = inode->i_mapping;
	block_t blkidx = F2FS_BYTES_TO_BLK(off);
	int offset = off & (sb->s_blocksize - 1);
	int tocopy;
	size_t toread;
	loff_t i_size = i_size_read(inode);
	struct page *page;
	char *kaddr;

	if (off > i_size)
		return 0;

	if (off + len > i_size)
		len = i_size - off;
	toread = len;
	while (toread > 0) {
		tocopy = min_t(unsigned long, sb->s_blocksize - offset, toread);
repeat:
2027
		page = read_cache_page_gfp(mapping, blkidx, GFP_NOFS);
2028 2029
		if (IS_ERR(page)) {
			if (PTR_ERR(page) == -ENOMEM) {
2030 2031
				congestion_wait(BLK_RW_ASYNC,
						DEFAULT_IO_TIMEOUT);
2032 2033
				goto repeat;
			}
2034
			set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR);
2035
			return PTR_ERR(page);
2036
		}
2037 2038 2039 2040 2041 2042 2043 2044 2045

		lock_page(page);

		if (unlikely(page->mapping != mapping)) {
			f2fs_put_page(page, 1);
			goto repeat;
		}
		if (unlikely(!PageUptodate(page))) {
			f2fs_put_page(page, 1);
2046
			set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR);
2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072
			return -EIO;
		}

		kaddr = kmap_atomic(page);
		memcpy(data, kaddr + offset, tocopy);
		kunmap_atomic(kaddr);
		f2fs_put_page(page, 1);

		offset = 0;
		toread -= tocopy;
		data += tocopy;
		blkidx++;
	}
	return len;
}

/* Write to quotafile */
static ssize_t f2fs_quota_write(struct super_block *sb, int type,
				const char *data, size_t len, loff_t off)
{
	struct inode *inode = sb_dqopt(sb)->files[type];
	struct address_space *mapping = inode->i_mapping;
	const struct address_space_operations *a_ops = mapping->a_ops;
	int offset = off & (sb->s_blocksize - 1);
	size_t towrite = len;
	struct page *page;
2073
	void *fsdata = NULL;
2074 2075 2076 2077 2078 2079 2080
	char *kaddr;
	int err = 0;
	int tocopy;

	while (towrite > 0) {
		tocopy = min_t(unsigned long, sb->s_blocksize - offset,
								towrite);
2081
retry:
2082
		err = a_ops->write_begin(NULL, mapping, off, tocopy, 0,
2083
							&page, &fsdata);
2084 2085
		if (unlikely(err)) {
			if (err == -ENOMEM) {
2086 2087
				congestion_wait(BLK_RW_ASYNC,
						DEFAULT_IO_TIMEOUT);
2088 2089
				goto retry;
			}
2090
			set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR);
2091
			break;
2092
		}
2093 2094 2095 2096 2097 2098 2099

		kaddr = kmap_atomic(page);
		memcpy(kaddr + offset, data, tocopy);
		kunmap_atomic(kaddr);
		flush_dcache_page(page);

		a_ops->write_end(NULL, mapping, off, tocopy, tocopy,
2100
						page, fsdata);
2101 2102 2103 2104 2105 2106 2107 2108
		offset = 0;
		towrite -= tocopy;
		off += tocopy;
		data += tocopy;
		cond_resched();
	}

	if (len == towrite)
2109
		return err;
2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124
	inode->i_mtime = inode->i_ctime = current_time(inode);
	f2fs_mark_inode_dirty_sync(inode, false);
	return len - towrite;
}

static struct dquot **f2fs_get_dquots(struct inode *inode)
{
	return F2FS_I(inode)->i_dquot;
}

static qsize_t *f2fs_get_reserved_space(struct inode *inode)
{
	return &F2FS_I(inode)->i_reserved_quota;
}

2125 2126
static int f2fs_quota_on_mount(struct f2fs_sb_info *sbi, int type)
{
2127
	if (is_set_ckpt_flags(sbi, CP_QUOTA_NEED_FSCK_FLAG)) {
2128
		f2fs_err(sbi, "quota sysfile may be corrupted, skip loading it");
2129 2130 2131
		return 0;
	}

2132 2133
	return dquot_quota_on_mount(sbi->sb, F2FS_OPTION(sbi).s_qf_names[type],
					F2FS_OPTION(sbi).s_jquota_fmt, type);
2134 2135
}

2136
int f2fs_enable_quota_files(struct f2fs_sb_info *sbi, bool rdonly)
2137
{
2138 2139 2140
	int enabled = 0;
	int i, err;

2141
	if (f2fs_sb_has_quota_ino(sbi) && rdonly) {
2142 2143
		err = f2fs_enable_quotas(sbi->sb);
		if (err) {
2144
			f2fs_err(sbi, "Cannot turn on quota_ino: %d", err);
2145 2146 2147 2148
			return 0;
		}
		return 1;
	}
2149 2150

	for (i = 0; i < MAXQUOTAS; i++) {
2151
		if (F2FS_OPTION(sbi).s_qf_names[i]) {
2152 2153 2154 2155 2156
			err = f2fs_quota_on_mount(sbi, i);
			if (!err) {
				enabled = 1;
				continue;
			}
2157 2158
			f2fs_err(sbi, "Cannot turn on quotas: %d on %d",
				 err, i);
2159 2160
		}
	}
2161 2162 2163 2164 2165 2166 2167 2168 2169 2170
	return enabled;
}

static int f2fs_quota_enable(struct super_block *sb, int type, int format_id,
			     unsigned int flags)
{
	struct inode *qf_inode;
	unsigned long qf_inum;
	int err;

2171
	BUG_ON(!f2fs_sb_has_quota_ino(F2FS_SB(sb)));
2172 2173 2174 2175 2176 2177 2178

	qf_inum = f2fs_qf_ino(sb, type);
	if (!qf_inum)
		return -EPERM;

	qf_inode = f2fs_iget(sb, qf_inum);
	if (IS_ERR(qf_inode)) {
2179
		f2fs_err(F2FS_SB(sb), "Bad quota inode %u:%lu", type, qf_inum);
2180 2181 2182 2183 2184
		return PTR_ERR(qf_inode);
	}

	/* Don't account quota for quota files to avoid recursion */
	qf_inode->i_flags |= S_NOQUOTA;
2185
	err = dquot_load_quota_inode(qf_inode, type, format_id, flags);
2186 2187 2188 2189 2190 2191
	iput(qf_inode);
	return err;
}

static int f2fs_enable_quotas(struct super_block *sb)
{
2192
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
2193 2194 2195
	int type, err = 0;
	unsigned long qf_inum;
	bool quota_mopt[MAXQUOTAS] = {
2196 2197 2198
		test_opt(sbi, USRQUOTA),
		test_opt(sbi, GRPQUOTA),
		test_opt(sbi, PRJQUOTA),
2199 2200
	};

2201
	if (is_set_ckpt_flags(F2FS_SB(sb), CP_QUOTA_NEED_FSCK_FLAG)) {
2202
		f2fs_err(sbi, "quota file may be corrupted, skip loading it");
2203 2204 2205 2206 2207
		return 0;
	}

	sb_dqopt(sb)->flags |= DQUOT_QUOTA_SYS_FILE;

2208 2209 2210 2211 2212 2213 2214
	for (type = 0; type < MAXQUOTAS; type++) {
		qf_inum = f2fs_qf_ino(sb, type);
		if (qf_inum) {
			err = f2fs_quota_enable(sb, type, QFMT_VFS_V1,
				DQUOT_USAGE_ENABLED |
				(quota_mopt[type] ? DQUOT_LIMITS_ENABLED : 0));
			if (err) {
2215 2216
				f2fs_err(sbi, "Failed to enable quota tracking (type=%d, err=%d). Please run fsck to fix.",
					 type, err);
2217 2218
				for (type--; type >= 0; type--)
					dquot_quota_off(sb, type);
2219 2220
				set_sbi_flag(F2FS_SB(sb),
						SBI_QUOTA_NEED_REPAIR);
2221 2222
				return err;
			}
2223 2224
		}
	}
2225
	return 0;
2226 2227
}

2228
int f2fs_quota_sync(struct super_block *sb, int type)
2229
{
2230
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
2231 2232 2233 2234
	struct quota_info *dqopt = sb_dqopt(sb);
	int cnt;
	int ret;

2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246
	/*
	 * do_quotactl
	 *  f2fs_quota_sync
	 *  down_read(quota_sem)
	 *  dquot_writeback_dquots()
	 *  f2fs_dquot_commit
	 *                            block_operation
	 *                            down_read(quota_sem)
	 */
	f2fs_lock_op(sbi);

	down_read(&sbi->quota_sem);
2247 2248
	ret = dquot_writeback_dquots(sb, type);
	if (ret)
2249
		goto out;
2250 2251 2252 2253 2254 2255

	/*
	 * Now when everything is written we can discard the pagecache so
	 * that userspace sees the changes.
	 */
	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
2256 2257
		struct address_space *mapping;

2258 2259 2260 2261 2262
		if (type != -1 && cnt != type)
			continue;
		if (!sb_has_quota_active(sb, cnt))
			continue;

2263 2264 2265
		mapping = dqopt->files[cnt]->i_mapping;

		ret = filemap_fdatawrite(mapping);
2266
		if (ret)
2267 2268 2269 2270 2271 2272 2273 2274 2275
			goto out;

		/* if we are using journalled quota */
		if (is_journalled_quota(sbi))
			continue;

		ret = filemap_fdatawait(mapping);
		if (ret)
			set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR);
2276 2277 2278 2279 2280

		inode_lock(dqopt->files[cnt]);
		truncate_inode_pages(&dqopt->files[cnt]->i_data, 0);
		inode_unlock(dqopt->files[cnt]);
	}
2281 2282 2283
out:
	if (ret)
		set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR);
2284 2285
	up_read(&sbi->quota_sem);
	f2fs_unlock_op(sbi);
2286
	return ret;
2287 2288 2289 2290 2291 2292 2293 2294
}

static int f2fs_quota_on(struct super_block *sb, int type, int format_id,
							const struct path *path)
{
	struct inode *inode;
	int err;

2295 2296 2297 2298 2299 2300
	/* if quota sysfile exists, deny enabling quota with specific file */
	if (f2fs_sb_has_quota_ino(F2FS_SB(sb))) {
		f2fs_err(F2FS_SB(sb), "quota sysfile already exists");
		return -EBUSY;
	}

2301
	err = f2fs_quota_sync(sb, type);
2302 2303 2304 2305 2306 2307 2308 2309 2310 2311
	if (err)
		return err;

	err = dquot_quota_on(sb, type, format_id, path);
	if (err)
		return err;

	inode = d_inode(path->dentry);

	inode_lock(inode);
2312
	F2FS_I(inode)->i_flags |= F2FS_NOATIME_FL | F2FS_IMMUTABLE_FL;
2313
	f2fs_set_inode_flags(inode);
2314 2315 2316 2317 2318 2319
	inode_unlock(inode);
	f2fs_mark_inode_dirty_sync(inode, false);

	return 0;
}

2320
static int __f2fs_quota_off(struct super_block *sb, int type)
2321 2322 2323 2324 2325 2326 2327
{
	struct inode *inode = sb_dqopt(sb)->files[type];
	int err;

	if (!inode || !igrab(inode))
		return dquot_quota_off(sb, type);

2328 2329 2330
	err = f2fs_quota_sync(sb, type);
	if (err)
		goto out_put;
2331 2332

	err = dquot_quota_off(sb, type);
2333
	if (err || f2fs_sb_has_quota_ino(F2FS_SB(sb)))
2334 2335 2336
		goto out_put;

	inode_lock(inode);
2337
	F2FS_I(inode)->i_flags &= ~(F2FS_NOATIME_FL | F2FS_IMMUTABLE_FL);
2338
	f2fs_set_inode_flags(inode);
2339 2340 2341 2342 2343 2344 2345
	inode_unlock(inode);
	f2fs_mark_inode_dirty_sync(inode, false);
out_put:
	iput(inode);
	return err;
}

2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362
static int f2fs_quota_off(struct super_block *sb, int type)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	int err;

	err = __f2fs_quota_off(sb, type);

	/*
	 * quotactl can shutdown journalled quota, result in inconsistence
	 * between quota record and fs data by following updates, tag the
	 * flag to let fsck be aware of it.
	 */
	if (is_journalled_quota(sbi))
		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
	return err;
}

2363
void f2fs_quota_off_umount(struct super_block *sb)
2364 2365
{
	int type;
2366 2367 2368
	int err;

	for (type = 0; type < MAXQUOTAS; type++) {
2369
		err = __f2fs_quota_off(sb, type);
2370 2371
		if (err) {
			int ret = dquot_quota_off(sb, type);
2372

2373 2374
			f2fs_err(F2FS_SB(sb), "Fail to turn off disk quota (type: %d, err: %d, ret:%d), Please run fsck to fix it.",
				 type, err, ret);
2375
			set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR);
2376 2377
		}
	}
2378 2379 2380 2381 2382 2383
	/*
	 * In case of checkpoint=disable, we must flush quota blocks.
	 * This can cause NULL exception for node_inode in end_io, since
	 * put_super already dropped it.
	 */
	sync_filesystem(sb);
2384 2385
}

2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397
static void f2fs_truncate_quota_inode_pages(struct super_block *sb)
{
	struct quota_info *dqopt = sb_dqopt(sb);
	int type;

	for (type = 0; type < MAXQUOTAS; type++) {
		if (!dqopt->files[type])
			continue;
		f2fs_inode_synced(dqopt->files[type]);
	}
}

2398 2399
static int f2fs_dquot_commit(struct dquot *dquot)
{
2400
	struct f2fs_sb_info *sbi = F2FS_SB(dquot->dq_sb);
2401 2402
	int ret;

2403
	down_read_nested(&sbi->quota_sem, SINGLE_DEPTH_NESTING);
2404 2405
	ret = dquot_commit(dquot);
	if (ret < 0)
2406 2407
		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
	up_read(&sbi->quota_sem);
2408 2409 2410 2411 2412
	return ret;
}

static int f2fs_dquot_acquire(struct dquot *dquot)
{
2413
	struct f2fs_sb_info *sbi = F2FS_SB(dquot->dq_sb);
2414 2415
	int ret;

2416
	down_read(&sbi->quota_sem);
2417 2418
	ret = dquot_acquire(dquot);
	if (ret < 0)
2419 2420
		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
	up_read(&sbi->quota_sem);
2421 2422 2423 2424 2425
	return ret;
}

static int f2fs_dquot_release(struct dquot *dquot)
{
2426
	struct f2fs_sb_info *sbi = F2FS_SB(dquot->dq_sb);
2427
	int ret = dquot_release(dquot);
2428 2429

	if (ret < 0)
2430
		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
2431 2432 2433 2434 2435 2436 2437
	return ret;
}

static int f2fs_dquot_mark_dquot_dirty(struct dquot *dquot)
{
	struct super_block *sb = dquot->dq_sb;
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
2438
	int ret = dquot_mark_dquot_dirty(dquot);
2439 2440 2441 2442 2443 2444 2445 2446 2447 2448

	/* if we are using journalled quota */
	if (is_journalled_quota(sbi))
		set_sbi_flag(sbi, SBI_QUOTA_NEED_FLUSH);

	return ret;
}

static int f2fs_dquot_commit_info(struct super_block *sb, int type)
{
2449
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
2450
	int ret = dquot_commit_info(sb, type);
2451 2452

	if (ret < 0)
2453
		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
2454 2455
	return ret;
}
2456

2457
static int f2fs_get_projid(struct inode *inode, kprojid_t *projid)
C
Chao Yu 已提交
2458 2459 2460 2461 2462
{
	*projid = F2FS_I(inode)->i_projid;
	return 0;
}

2463 2464
static const struct dquot_operations f2fs_quota_operations = {
	.get_reserved_space = f2fs_get_reserved_space,
2465 2466 2467 2468 2469
	.write_dquot	= f2fs_dquot_commit,
	.acquire_dquot	= f2fs_dquot_acquire,
	.release_dquot	= f2fs_dquot_release,
	.mark_dirty	= f2fs_dquot_mark_dquot_dirty,
	.write_info	= f2fs_dquot_commit_info,
2470 2471
	.alloc_dquot	= dquot_alloc,
	.destroy_dquot	= dquot_destroy,
C
Chao Yu 已提交
2472
	.get_projid	= f2fs_get_projid,
2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486
	.get_next_id	= dquot_get_next_id,
};

static const struct quotactl_ops f2fs_quotactl_ops = {
	.quota_on	= f2fs_quota_on,
	.quota_off	= f2fs_quota_off,
	.quota_sync	= f2fs_quota_sync,
	.get_state	= dquot_get_state,
	.set_info	= dquot_set_dqinfo,
	.get_dqblk	= dquot_get_dqblk,
	.set_dqblk	= dquot_set_dqblk,
	.get_nextdqblk	= dquot_get_next_dqblk,
};
#else
2487 2488 2489 2490 2491
int f2fs_quota_sync(struct super_block *sb, int type)
{
	return 0;
}

2492
void f2fs_quota_off_umount(struct super_block *sb)
2493 2494 2495 2496
{
}
#endif

2497
static const struct super_operations f2fs_sops = {
2498
	.alloc_inode	= f2fs_alloc_inode,
2499
	.free_inode	= f2fs_free_inode,
2500
	.drop_inode	= f2fs_drop_inode,
2501
	.write_inode	= f2fs_write_inode,
2502
	.dirty_inode	= f2fs_dirty_inode,
2503
	.show_options	= f2fs_show_options,
2504 2505 2506 2507 2508
#ifdef CONFIG_QUOTA
	.quota_read	= f2fs_quota_read,
	.quota_write	= f2fs_quota_write,
	.get_dquots	= f2fs_get_dquots,
#endif
2509 2510 2511
	.evict_inode	= f2fs_evict_inode,
	.put_super	= f2fs_put_super,
	.sync_fs	= f2fs_sync_fs,
2512 2513
	.freeze_fs	= f2fs_freeze,
	.unfreeze_fs	= f2fs_unfreeze,
2514
	.statfs		= f2fs_statfs,
2515
	.remount_fs	= f2fs_remount,
2516 2517
};

2518
#ifdef CONFIG_FS_ENCRYPTION
2519 2520 2521 2522 2523 2524 2525 2526 2527 2528
static int f2fs_get_context(struct inode *inode, void *ctx, size_t len)
{
	return f2fs_getxattr(inode, F2FS_XATTR_INDEX_ENCRYPTION,
				F2FS_XATTR_NAME_ENCRYPTION_CONTEXT,
				ctx, len, NULL);
}

static int f2fs_set_context(struct inode *inode, const void *ctx, size_t len,
							void *fs_data)
{
2529 2530 2531 2532 2533 2534 2535 2536
	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);

	/*
	 * Encrypting the root directory is not allowed because fsck
	 * expects lost+found directory to exist and remain unencrypted
	 * if LOST_FOUND feature is enabled.
	 *
	 */
2537
	if (f2fs_sb_has_lost_found(sbi) &&
2538 2539 2540
			inode->i_ino == F2FS_ROOT_INO(sbi))
		return -EPERM;

2541 2542 2543 2544 2545
	return f2fs_setxattr(inode, F2FS_XATTR_INDEX_ENCRYPTION,
				F2FS_XATTR_NAME_ENCRYPTION_CONTEXT,
				ctx, len, fs_data, XATTR_CREATE);
}

2546
static const union fscrypt_policy *f2fs_get_dummy_policy(struct super_block *sb)
2547
{
2548
	return F2FS_OPTION(F2FS_SB(sb)).dummy_enc_policy.policy;
2549 2550
}

2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562
static bool f2fs_has_stable_inodes(struct super_block *sb)
{
	return true;
}

static void f2fs_get_ino_and_lblk_bits(struct super_block *sb,
				       int *ino_bits_ret, int *lblk_bits_ret)
{
	*ino_bits_ret = 8 * sizeof(nid_t);
	*lblk_bits_ret = 8 * sizeof(block_t);
}

2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581
static int f2fs_get_num_devices(struct super_block *sb)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);

	if (f2fs_is_multi_device(sbi))
		return sbi->s_ndevs;
	return 1;
}

static void f2fs_get_devices(struct super_block *sb,
			     struct request_queue **devs)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	int i;

	for (i = 0; i < sbi->s_ndevs; i++)
		devs[i] = bdev_get_queue(FDEV(i).bdev);
}

2582
static const struct fscrypt_operations f2fs_cryptops = {
2583 2584 2585
	.key_prefix		= "f2fs:",
	.get_context		= f2fs_get_context,
	.set_context		= f2fs_set_context,
2586
	.get_dummy_policy	= f2fs_get_dummy_policy,
2587 2588 2589 2590
	.empty_dir		= f2fs_empty_dir,
	.max_namelen		= F2FS_NAME_LEN,
	.has_stable_inodes	= f2fs_has_stable_inodes,
	.get_ino_and_lblk_bits	= f2fs_get_ino_and_lblk_bits,
2591 2592
	.get_num_devices	= f2fs_get_num_devices,
	.get_devices		= f2fs_get_devices,
2593 2594 2595
};
#endif

2596 2597 2598 2599 2600 2601
static struct inode *f2fs_nfs_get_inode(struct super_block *sb,
		u64 ino, u32 generation)
{
	struct f2fs_sb_info *sbi = F2FS_SB(sb);
	struct inode *inode;

2602
	if (f2fs_check_nid_range(sbi, ino))
2603
		return ERR_PTR(-ESTALE);
2604 2605 2606 2607 2608 2609 2610 2611 2612

	/*
	 * f2fs_iget isn't quite right if the inode is currently unallocated!
	 * However f2fs_iget currently does appropriate checks to handle stale
	 * inodes so everything is OK.
	 */
	inode = f2fs_iget(sb, ino);
	if (IS_ERR(inode))
		return ERR_CAST(inode);
2613
	if (unlikely(generation && inode->i_generation != generation)) {
2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640
		/* we didn't find the right inode.. */
		iput(inode);
		return ERR_PTR(-ESTALE);
	}
	return inode;
}

static struct dentry *f2fs_fh_to_dentry(struct super_block *sb, struct fid *fid,
		int fh_len, int fh_type)
{
	return generic_fh_to_dentry(sb, fid, fh_len, fh_type,
				    f2fs_nfs_get_inode);
}

static struct dentry *f2fs_fh_to_parent(struct super_block *sb, struct fid *fid,
		int fh_len, int fh_type)
{
	return generic_fh_to_parent(sb, fid, fh_len, fh_type,
				    f2fs_nfs_get_inode);
}

static const struct export_operations f2fs_export_ops = {
	.fh_to_dentry = f2fs_fh_to_dentry,
	.fh_to_parent = f2fs_fh_to_parent,
	.get_parent = f2fs_get_parent,
};

2641
static loff_t max_file_blocks(void)
2642
{
2643
	loff_t result = 0;
2644
	loff_t leaf_count = DEF_ADDRS_PER_BLOCK;
2645

2646 2647
	/*
	 * note: previously, result is equal to (DEF_ADDRS_PER_INODE -
2648
	 * DEFAULT_INLINE_XATTR_ADDRS), but now f2fs try to reserve more
2649 2650 2651 2652
	 * space in inode.i_addr, it will be more safe to reassign
	 * result as zero.
	 */

2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666
	/* two direct node blocks */
	result += (leaf_count * 2);

	/* two indirect node blocks */
	leaf_count *= NIDS_PER_BLOCK;
	result += (leaf_count * 2);

	/* one double indirect node block */
	leaf_count *= NIDS_PER_BLOCK;
	result += leaf_count;

	return result;
}

2667 2668 2669 2670 2671 2672 2673 2674 2675 2676
static int __f2fs_commit_super(struct buffer_head *bh,
			struct f2fs_super_block *super)
{
	lock_buffer(bh);
	if (super)
		memcpy(bh->b_data + F2FS_SUPER_OFFSET, super, sizeof(*super));
	set_buffer_dirty(bh);
	unlock_buffer(bh);

	/* it's rare case, we can do fua all the time */
2677
	return __sync_dirty_buffer(bh, REQ_SYNC | REQ_PREFLUSH | REQ_FUA);
2678 2679
}

2680
static inline bool sanity_check_area_boundary(struct f2fs_sb_info *sbi,
2681
					struct buffer_head *bh)
2682
{
2683 2684
	struct f2fs_super_block *raw_super = (struct f2fs_super_block *)
					(bh->b_data + F2FS_SUPER_OFFSET);
2685
	struct super_block *sb = sbi->sb;
2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698
	u32 segment0_blkaddr = le32_to_cpu(raw_super->segment0_blkaddr);
	u32 cp_blkaddr = le32_to_cpu(raw_super->cp_blkaddr);
	u32 sit_blkaddr = le32_to_cpu(raw_super->sit_blkaddr);
	u32 nat_blkaddr = le32_to_cpu(raw_super->nat_blkaddr);
	u32 ssa_blkaddr = le32_to_cpu(raw_super->ssa_blkaddr);
	u32 main_blkaddr = le32_to_cpu(raw_super->main_blkaddr);
	u32 segment_count_ckpt = le32_to_cpu(raw_super->segment_count_ckpt);
	u32 segment_count_sit = le32_to_cpu(raw_super->segment_count_sit);
	u32 segment_count_nat = le32_to_cpu(raw_super->segment_count_nat);
	u32 segment_count_ssa = le32_to_cpu(raw_super->segment_count_ssa);
	u32 segment_count_main = le32_to_cpu(raw_super->segment_count_main);
	u32 segment_count = le32_to_cpu(raw_super->segment_count);
	u32 log_blocks_per_seg = le32_to_cpu(raw_super->log_blocks_per_seg);
2699 2700 2701 2702
	u64 main_end_blkaddr = main_blkaddr +
				(segment_count_main << log_blocks_per_seg);
	u64 seg_end_blkaddr = segment0_blkaddr +
				(segment_count << log_blocks_per_seg);
2703 2704

	if (segment0_blkaddr != cp_blkaddr) {
2705 2706
		f2fs_info(sbi, "Mismatch start address, segment0(%u) cp_blkaddr(%u)",
			  segment0_blkaddr, cp_blkaddr);
2707 2708 2709 2710 2711
		return true;
	}

	if (cp_blkaddr + (segment_count_ckpt << log_blocks_per_seg) !=
							sit_blkaddr) {
2712 2713 2714
		f2fs_info(sbi, "Wrong CP boundary, start(%u) end(%u) blocks(%u)",
			  cp_blkaddr, sit_blkaddr,
			  segment_count_ckpt << log_blocks_per_seg);
2715 2716 2717 2718 2719
		return true;
	}

	if (sit_blkaddr + (segment_count_sit << log_blocks_per_seg) !=
							nat_blkaddr) {
2720 2721 2722
		f2fs_info(sbi, "Wrong SIT boundary, start(%u) end(%u) blocks(%u)",
			  sit_blkaddr, nat_blkaddr,
			  segment_count_sit << log_blocks_per_seg);
2723 2724 2725 2726 2727
		return true;
	}

	if (nat_blkaddr + (segment_count_nat << log_blocks_per_seg) !=
							ssa_blkaddr) {
2728 2729 2730
		f2fs_info(sbi, "Wrong NAT boundary, start(%u) end(%u) blocks(%u)",
			  nat_blkaddr, ssa_blkaddr,
			  segment_count_nat << log_blocks_per_seg);
2731 2732 2733 2734 2735
		return true;
	}

	if (ssa_blkaddr + (segment_count_ssa << log_blocks_per_seg) !=
							main_blkaddr) {
2736 2737 2738
		f2fs_info(sbi, "Wrong SSA boundary, start(%u) end(%u) blocks(%u)",
			  ssa_blkaddr, main_blkaddr,
			  segment_count_ssa << log_blocks_per_seg);
2739 2740 2741
		return true;
	}

2742
	if (main_end_blkaddr > seg_end_blkaddr) {
2743 2744
		f2fs_info(sbi, "Wrong MAIN_AREA boundary, start(%u) end(%llu) block(%u)",
			  main_blkaddr, seg_end_blkaddr,
2745
			  segment_count_main << log_blocks_per_seg);
2746
		return true;
2747 2748 2749 2750 2751 2752 2753 2754 2755
	} else if (main_end_blkaddr < seg_end_blkaddr) {
		int err = 0;
		char *res;

		/* fix in-memory information all the time */
		raw_super->segment_count = cpu_to_le32((main_end_blkaddr -
				segment0_blkaddr) >> log_blocks_per_seg);

		if (f2fs_readonly(sb) || bdev_read_only(sb->s_bdev)) {
2756
			set_sbi_flag(sbi, SBI_NEED_SB_WRITE);
2757 2758 2759 2760 2761
			res = "internally";
		} else {
			err = __f2fs_commit_super(bh, NULL);
			res = err ? "failed" : "done";
		}
2762 2763
		f2fs_info(sbi, "Fix alignment : %s, start(%u) end(%llu) block(%u)",
			  res, main_blkaddr, seg_end_blkaddr,
2764
			  segment_count_main << log_blocks_per_seg);
2765 2766
		if (err)
			return true;
2767 2768 2769 2770
	}
	return false;
}

2771
static int sanity_check_raw_super(struct f2fs_sb_info *sbi,
2772
				struct buffer_head *bh)
2773
{
2774
	block_t segment_count, segs_per_sec, secs_per_zone, segment_count_main;
2775
	block_t total_sections, blocks_per_seg;
2776 2777
	struct f2fs_super_block *raw_super = (struct f2fs_super_block *)
					(bh->b_data + F2FS_SUPER_OFFSET);
2778
	unsigned int blocksize;
2779 2780 2781
	size_t crc_offset = 0;
	__u32 crc = 0;

2782 2783 2784 2785 2786 2787
	if (le32_to_cpu(raw_super->magic) != F2FS_SUPER_MAGIC) {
		f2fs_info(sbi, "Magic Mismatch, valid(0x%x) - read(0x%x)",
			  F2FS_SUPER_MAGIC, le32_to_cpu(raw_super->magic));
		return -EINVAL;
	}

2788
	/* Check checksum_offset and crc in superblock */
2789
	if (__F2FS_HAS_FEATURE(raw_super, F2FS_FEATURE_SB_CHKSUM)) {
2790 2791 2792
		crc_offset = le32_to_cpu(raw_super->checksum_offset);
		if (crc_offset !=
			offsetof(struct f2fs_super_block, crc)) {
2793 2794
			f2fs_info(sbi, "Invalid SB checksum offset: %zu",
				  crc_offset);
2795
			return -EFSCORRUPTED;
2796 2797 2798
		}
		crc = le32_to_cpu(raw_super->crc);
		if (!f2fs_crc_valid(sbi, crc, raw_super, crc_offset)) {
2799
			f2fs_info(sbi, "Invalid SB checksum value: %u", crc);
2800
			return -EFSCORRUPTED;
2801 2802
		}
	}
2803

2804
	/* Currently, support only 4KB page cache size */
2805
	if (F2FS_BLKSIZE != PAGE_SIZE) {
2806 2807
		f2fs_info(sbi, "Invalid page_cache_size (%lu), supports only 4KB",
			  PAGE_SIZE);
2808
		return -EFSCORRUPTED;
2809 2810
	}

2811 2812
	/* Currently, support only 4KB block size */
	blocksize = 1 << le32_to_cpu(raw_super->log_blocksize);
2813
	if (blocksize != F2FS_BLKSIZE) {
2814 2815
		f2fs_info(sbi, "Invalid blocksize (%u), supports only 4KB",
			  blocksize);
2816
		return -EFSCORRUPTED;
2817
	}
2818

2819 2820
	/* check log blocks per segment */
	if (le32_to_cpu(raw_super->log_blocks_per_seg) != 9) {
2821 2822
		f2fs_info(sbi, "Invalid log blocks per segment (%u)",
			  le32_to_cpu(raw_super->log_blocks_per_seg));
2823
		return -EFSCORRUPTED;
2824 2825
	}

C
Chao Yu 已提交
2826 2827 2828 2829 2830
	/* Currently, support 512/1024/2048/4096 bytes sector size */
	if (le32_to_cpu(raw_super->log_sectorsize) >
				F2FS_MAX_LOG_SECTOR_SIZE ||
		le32_to_cpu(raw_super->log_sectorsize) <
				F2FS_MIN_LOG_SECTOR_SIZE) {
2831 2832
		f2fs_info(sbi, "Invalid log sectorsize (%u)",
			  le32_to_cpu(raw_super->log_sectorsize));
2833
		return -EFSCORRUPTED;
2834
	}
C
Chao Yu 已提交
2835 2836 2837
	if (le32_to_cpu(raw_super->log_sectors_per_block) +
		le32_to_cpu(raw_super->log_sectorsize) !=
			F2FS_MAX_LOG_SECTOR_SIZE) {
2838 2839 2840
		f2fs_info(sbi, "Invalid log sectors per block(%u) log sectorsize(%u)",
			  le32_to_cpu(raw_super->log_sectors_per_block),
			  le32_to_cpu(raw_super->log_sectorsize));
2841
		return -EFSCORRUPTED;
2842
	}
2843

2844
	segment_count = le32_to_cpu(raw_super->segment_count);
2845
	segment_count_main = le32_to_cpu(raw_super->segment_count_main);
2846 2847 2848 2849 2850 2851 2852 2853 2854
	segs_per_sec = le32_to_cpu(raw_super->segs_per_sec);
	secs_per_zone = le32_to_cpu(raw_super->secs_per_zone);
	total_sections = le32_to_cpu(raw_super->section_count);

	/* blocks_per_seg should be 512, given the above check */
	blocks_per_seg = 1 << le32_to_cpu(raw_super->log_blocks_per_seg);

	if (segment_count > F2FS_MAX_SEGMENT ||
				segment_count < F2FS_MIN_SEGMENTS) {
2855
		f2fs_info(sbi, "Invalid segment count (%u)", segment_count);
2856
		return -EFSCORRUPTED;
2857 2858
	}

2859
	if (total_sections > segment_count_main || total_sections < 1 ||
2860
			segs_per_sec > segment_count || !segs_per_sec) {
2861 2862
		f2fs_info(sbi, "Invalid segment/section count (%u, %u x %u)",
			  segment_count, total_sections, segs_per_sec);
2863
		return -EFSCORRUPTED;
2864 2865
	}

2866 2867 2868 2869 2870 2871
	if (segment_count_main != total_sections * segs_per_sec) {
		f2fs_info(sbi, "Invalid segment/section count (%u != %u * %u)",
			  segment_count_main, total_sections, segs_per_sec);
		return -EFSCORRUPTED;
	}

2872
	if ((segment_count / segs_per_sec) < total_sections) {
2873 2874
		f2fs_info(sbi, "Small segment_count (%u < %u * %u)",
			  segment_count, segs_per_sec, total_sections);
2875
		return -EFSCORRUPTED;
2876 2877
	}

2878
	if (segment_count > (le64_to_cpu(raw_super->block_count) >> 9)) {
2879 2880
		f2fs_info(sbi, "Wrong segment_count / block_count (%u > %llu)",
			  segment_count, le64_to_cpu(raw_super->block_count));
2881
		return -EFSCORRUPTED;
2882 2883
	}

2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896
	if (RDEV(0).path[0]) {
		block_t dev_seg_count = le32_to_cpu(RDEV(0).total_segments);
		int i = 1;

		while (i < MAX_DEVICES && RDEV(i).path[0]) {
			dev_seg_count += le32_to_cpu(RDEV(i).total_segments);
			i++;
		}
		if (segment_count != dev_seg_count) {
			f2fs_info(sbi, "Segment count (%u) mismatch with total segments from devices (%u)",
					segment_count, dev_seg_count);
			return -EFSCORRUPTED;
		}
2897 2898 2899 2900 2901 2902
	} else {
		if (__F2FS_HAS_FEATURE(raw_super, F2FS_FEATURE_BLKZONED) &&
					!bdev_is_zoned(sbi->sb->s_bdev)) {
			f2fs_info(sbi, "Zoned block device path is missing");
			return -EFSCORRUPTED;
		}
2903 2904
	}

2905
	if (secs_per_zone > total_sections || !secs_per_zone) {
2906 2907
		f2fs_info(sbi, "Wrong secs_per_zone / total_sections (%u, %u)",
			  secs_per_zone, total_sections);
2908
		return -EFSCORRUPTED;
2909 2910 2911 2912 2913
	}
	if (le32_to_cpu(raw_super->extension_count) > F2FS_MAX_EXTENSION ||
			raw_super->hot_ext_count > F2FS_MAX_EXTENSION ||
			(le32_to_cpu(raw_super->extension_count) +
			raw_super->hot_ext_count) > F2FS_MAX_EXTENSION) {
2914 2915 2916 2917
		f2fs_info(sbi, "Corrupted extension count (%u + %u > %u)",
			  le32_to_cpu(raw_super->extension_count),
			  raw_super->hot_ext_count,
			  F2FS_MAX_EXTENSION);
2918
		return -EFSCORRUPTED;
2919 2920 2921 2922
	}

	if (le32_to_cpu(raw_super->cp_payload) >
				(blocks_per_seg - F2FS_CP_PACKS)) {
2923 2924 2925
		f2fs_info(sbi, "Insane cp_payload (%u > %u)",
			  le32_to_cpu(raw_super->cp_payload),
			  blocks_per_seg - F2FS_CP_PACKS);
2926
		return -EFSCORRUPTED;
2927 2928
	}

2929 2930 2931 2932
	/* check reserved ino info */
	if (le32_to_cpu(raw_super->node_ino) != 1 ||
		le32_to_cpu(raw_super->meta_ino) != 2 ||
		le32_to_cpu(raw_super->root_ino) != 3) {
2933 2934 2935 2936
		f2fs_info(sbi, "Invalid Fs Meta Ino: node(%u) meta(%u) root(%u)",
			  le32_to_cpu(raw_super->node_ino),
			  le32_to_cpu(raw_super->meta_ino),
			  le32_to_cpu(raw_super->root_ino));
2937
		return -EFSCORRUPTED;
2938 2939 2940
	}

	/* check CP/SIT/NAT/SSA/MAIN_AREA area boundary */
2941
	if (sanity_check_area_boundary(sbi, bh))
2942
		return -EFSCORRUPTED;
2943

2944 2945 2946
	return 0;
}

2947
int f2fs_sanity_check_ckpt(struct f2fs_sb_info *sbi)
2948 2949
{
	unsigned int total, fsmeta;
2950 2951
	struct f2fs_super_block *raw_super = F2FS_RAW_SUPER(sbi);
	struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi);
J
Jaegeuk Kim 已提交
2952
	unsigned int ovp_segments, reserved_segments;
2953
	unsigned int main_segs, blocks_per_seg;
2954 2955 2956
	unsigned int sit_segs, nat_segs;
	unsigned int sit_bitmap_size, nat_bitmap_size;
	unsigned int log_blocks_per_seg;
2957
	unsigned int segment_count_main;
2958
	unsigned int cp_pack_start_sum, cp_payload;
2959 2960
	block_t user_block_count, valid_user_blocks;
	block_t avail_node_count, valid_node_count;
2961
	int i, j;
2962 2963 2964

	total = le32_to_cpu(raw_super->segment_count);
	fsmeta = le32_to_cpu(raw_super->segment_count_ckpt);
2965 2966 2967 2968
	sit_segs = le32_to_cpu(raw_super->segment_count_sit);
	fsmeta += sit_segs;
	nat_segs = le32_to_cpu(raw_super->segment_count_nat);
	fsmeta += nat_segs;
2969 2970 2971
	fsmeta += le32_to_cpu(ckpt->rsvd_segment_count);
	fsmeta += le32_to_cpu(raw_super->segment_count_ssa);

2972
	if (unlikely(fsmeta >= total))
2973
		return 1;
2974

J
Jaegeuk Kim 已提交
2975 2976 2977
	ovp_segments = le32_to_cpu(ckpt->overprov_segment_count);
	reserved_segments = le32_to_cpu(ckpt->rsvd_segment_count);

2978
	if (unlikely(fsmeta < F2FS_MIN_META_SEGMENTS ||
J
Jaegeuk Kim 已提交
2979
			ovp_segments == 0 || reserved_segments == 0)) {
2980
		f2fs_err(sbi, "Wrong layout: check mkfs.f2fs version");
J
Jaegeuk Kim 已提交
2981 2982 2983
		return 1;
	}

2984 2985 2986 2987 2988
	user_block_count = le64_to_cpu(ckpt->user_block_count);
	segment_count_main = le32_to_cpu(raw_super->segment_count_main);
	log_blocks_per_seg = le32_to_cpu(raw_super->log_blocks_per_seg);
	if (!user_block_count || user_block_count >=
			segment_count_main << log_blocks_per_seg) {
2989 2990
		f2fs_err(sbi, "Wrong user_block_count: %u",
			 user_block_count);
2991 2992 2993
		return 1;
	}

2994 2995
	valid_user_blocks = le64_to_cpu(ckpt->valid_block_count);
	if (valid_user_blocks > user_block_count) {
2996 2997
		f2fs_err(sbi, "Wrong valid_user_blocks: %u, user_block_count: %u",
			 valid_user_blocks, user_block_count);
2998 2999 3000 3001
		return 1;
	}

	valid_node_count = le32_to_cpu(ckpt->valid_node_count);
3002
	avail_node_count = sbi->total_node_count - F2FS_RESERVED_NODE_NUM;
3003
	if (valid_node_count > avail_node_count) {
3004 3005
		f2fs_err(sbi, "Wrong valid_node_count: %u, avail_node_count: %u",
			 valid_node_count, avail_node_count);
3006 3007 3008
		return 1;
	}

3009 3010 3011 3012 3013 3014 3015
	main_segs = le32_to_cpu(raw_super->segment_count_main);
	blocks_per_seg = sbi->blocks_per_seg;

	for (i = 0; i < NR_CURSEG_NODE_TYPE; i++) {
		if (le32_to_cpu(ckpt->cur_node_segno[i]) >= main_segs ||
			le16_to_cpu(ckpt->cur_node_blkoff[i]) >= blocks_per_seg)
			return 1;
3016 3017 3018
		for (j = i + 1; j < NR_CURSEG_NODE_TYPE; j++) {
			if (le32_to_cpu(ckpt->cur_node_segno[i]) ==
				le32_to_cpu(ckpt->cur_node_segno[j])) {
3019 3020 3021
				f2fs_err(sbi, "Node segment (%u, %u) has the same segno: %u",
					 i, j,
					 le32_to_cpu(ckpt->cur_node_segno[i]));
3022 3023 3024
				return 1;
			}
		}
3025 3026 3027 3028 3029
	}
	for (i = 0; i < NR_CURSEG_DATA_TYPE; i++) {
		if (le32_to_cpu(ckpt->cur_data_segno[i]) >= main_segs ||
			le16_to_cpu(ckpt->cur_data_blkoff[i]) >= blocks_per_seg)
			return 1;
3030 3031 3032
		for (j = i + 1; j < NR_CURSEG_DATA_TYPE; j++) {
			if (le32_to_cpu(ckpt->cur_data_segno[i]) ==
				le32_to_cpu(ckpt->cur_data_segno[j])) {
3033 3034 3035
				f2fs_err(sbi, "Data segment (%u, %u) has the same segno: %u",
					 i, j,
					 le32_to_cpu(ckpt->cur_data_segno[i]));
3036 3037 3038 3039 3040
				return 1;
			}
		}
	}
	for (i = 0; i < NR_CURSEG_NODE_TYPE; i++) {
3041
		for (j = 0; j < NR_CURSEG_DATA_TYPE; j++) {
3042 3043
			if (le32_to_cpu(ckpt->cur_node_segno[i]) ==
				le32_to_cpu(ckpt->cur_data_segno[j])) {
3044
				f2fs_err(sbi, "Node segment (%u) and Data segment (%u) has the same segno: %u",
3045 3046
					 i, j,
					 le32_to_cpu(ckpt->cur_node_segno[i]));
3047 3048 3049
				return 1;
			}
		}
3050 3051
	}

3052 3053 3054 3055 3056
	sit_bitmap_size = le32_to_cpu(ckpt->sit_ver_bitmap_bytesize);
	nat_bitmap_size = le32_to_cpu(ckpt->nat_ver_bitmap_bytesize);

	if (sit_bitmap_size != ((sit_segs / 2) << log_blocks_per_seg) / 8 ||
		nat_bitmap_size != ((nat_segs / 2) << log_blocks_per_seg) / 8) {
3057 3058
		f2fs_err(sbi, "Wrong bitmap size: sit: %u, nat:%u",
			 sit_bitmap_size, nat_bitmap_size);
3059 3060
		return 1;
	}
3061 3062 3063 3064 3065

	cp_pack_start_sum = __start_sum_addr(sbi);
	cp_payload = __cp_payload(sbi);
	if (cp_pack_start_sum < cp_payload + 1 ||
		cp_pack_start_sum > blocks_per_seg - 1 -
C
Chao Yu 已提交
3066
			NR_CURSEG_PERSIST_TYPE) {
3067 3068
		f2fs_err(sbi, "Wrong cp_pack_start_sum: %u",
			 cp_pack_start_sum);
3069 3070
		return 1;
	}
3071

3072 3073
	if (__is_set_ckpt_flags(ckpt, CP_LARGE_NAT_BITMAP_FLAG) &&
		le32_to_cpu(ckpt->checksum_offset) != CP_MIN_CHKSUM_OFFSET) {
3074 3075 3076
		f2fs_warn(sbi, "using deprecated layout of large_nat_bitmap, "
			  "please run fsck v1.13.0 or higher to repair, chksum_offset: %u, "
			  "fixed with patch: \"f2fs-tools: relocate chksum_offset for large_nat_bitmap feature\"",
3077
			  le32_to_cpu(ckpt->checksum_offset));
3078 3079 3080
		return 1;
	}

3081
	if (unlikely(f2fs_cp_error(sbi))) {
3082
		f2fs_err(sbi, "A bug case: need to run fsck");
3083 3084
		return 1;
	}
3085 3086 3087 3088 3089 3090
	return 0;
}

static void init_sb_info(struct f2fs_sb_info *sbi)
{
	struct f2fs_super_block *raw_super = sbi->raw_super;
3091
	int i;
3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104

	sbi->log_sectors_per_block =
		le32_to_cpu(raw_super->log_sectors_per_block);
	sbi->log_blocksize = le32_to_cpu(raw_super->log_blocksize);
	sbi->blocksize = 1 << sbi->log_blocksize;
	sbi->log_blocks_per_seg = le32_to_cpu(raw_super->log_blocks_per_seg);
	sbi->blocks_per_seg = 1 << sbi->log_blocks_per_seg;
	sbi->segs_per_sec = le32_to_cpu(raw_super->segs_per_sec);
	sbi->secs_per_zone = le32_to_cpu(raw_super->secs_per_zone);
	sbi->total_sections = le32_to_cpu(raw_super->section_count);
	sbi->total_node_count =
		(le32_to_cpu(raw_super->segment_count_nat) / 2)
			* sbi->blocks_per_seg * NAT_ENTRY_PER_BLOCK;
3105 3106 3107
	F2FS_ROOT_INO(sbi) = le32_to_cpu(raw_super->root_ino);
	F2FS_NODE_INO(sbi) = le32_to_cpu(raw_super->node_ino);
	F2FS_META_INO(sbi) = le32_to_cpu(raw_super->meta_ino);
3108
	sbi->cur_victim_sec = NULL_SECNO;
3109 3110
	sbi->next_victim_seg[BG_GC] = NULL_SEGNO;
	sbi->next_victim_seg[FG_GC] = NULL_SEGNO;
3111
	sbi->max_victim_search = DEF_MAX_VICTIM_SEARCH;
3112
	sbi->migration_granularity = sbi->segs_per_sec;
3113

3114
	sbi->dir_level = DEF_DIR_LEVEL;
3115
	sbi->interval_time[CP_TIME] = DEF_CP_INTERVAL;
3116
	sbi->interval_time[REQ_TIME] = DEF_IDLE_INTERVAL;
3117 3118
	sbi->interval_time[DISCARD_TIME] = DEF_IDLE_INTERVAL;
	sbi->interval_time[GC_TIME] = DEF_IDLE_INTERVAL;
3119
	sbi->interval_time[DISABLE_TIME] = DEF_DISABLE_INTERVAL;
3120 3121
	sbi->interval_time[UMOUNT_DISCARD_TIMEOUT] =
				DEF_UMOUNT_DISCARD_TIMEOUT;
3122
	clear_sbi_flag(sbi, SBI_NEED_FSCK);
3123

3124 3125 3126
	for (i = 0; i < NR_COUNT_TYPE; i++)
		atomic_set(&sbi->nr_pages[i], 0);

3127 3128
	for (i = 0; i < META; i++)
		atomic_set(&sbi->wb_sync_req[i], 0);
3129

3130 3131
	INIT_LIST_HEAD(&sbi->s_list);
	mutex_init(&sbi->umount_mutex);
3132
	init_rwsem(&sbi->io_order_lock);
3133
	spin_lock_init(&sbi->cp_lock);
3134 3135 3136

	sbi->dirty_device = 0;
	spin_lock_init(&sbi->dev_lock);
3137

3138
	init_rwsem(&sbi->sb_lock);
3139
	init_rwsem(&sbi->pin_sem);
3140 3141
}

3142 3143
static int init_percpu_info(struct f2fs_sb_info *sbi)
{
3144
	int err;
3145

3146 3147 3148 3149
	err = percpu_counter_init(&sbi->alloc_valid_block_count, 0, GFP_KERNEL);
	if (err)
		return err;

3150
	err = percpu_counter_init(&sbi->total_valid_inode_count, 0,
3151
								GFP_KERNEL);
3152 3153 3154 3155
	if (err)
		percpu_counter_destroy(&sbi->alloc_valid_block_count);

	return err;
3156 3157
}

3158
#ifdef CONFIG_BLK_DEV_ZONED
3159 3160 3161 3162 3163 3164

struct f2fs_report_zones_args {
	struct f2fs_dev_info *dev;
	bool zone_cap_mismatch;
};

3165
static int f2fs_report_zone_cb(struct blk_zone *zone, unsigned int idx,
3166
			      void *data)
3167
{
3168 3169 3170 3171 3172 3173 3174 3175 3176 3177
	struct f2fs_report_zones_args *rz_args = data;

	if (zone->type == BLK_ZONE_TYPE_CONVENTIONAL)
		return 0;

	set_bit(idx, rz_args->dev->blkz_seq);
	rz_args->dev->zone_capacity_blocks[idx] = zone->capacity >>
						F2FS_LOG_SECTORS_PER_BLOCK;
	if (zone->len != zone->capacity && !rz_args->zone_cap_mismatch)
		rz_args->zone_cap_mismatch = true;
3178 3179 3180 3181

	return 0;
}

3182
static int init_blkz_info(struct f2fs_sb_info *sbi, int devi)
3183
{
3184
	struct block_device *bdev = FDEV(devi).bdev;
3185
	sector_t nr_sectors = bdev->bd_part->nr_sects;
3186
	struct f2fs_report_zones_args rep_zone_arg;
3187
	int ret;
3188

3189
	if (!f2fs_sb_has_blkzoned(sbi))
3190 3191
		return 0;

3192
	if (sbi->blocks_per_blkz && sbi->blocks_per_blkz !=
3193
				SECTOR_TO_BLOCK(bdev_zone_sectors(bdev)))
3194
		return -EINVAL;
3195
	sbi->blocks_per_blkz = SECTOR_TO_BLOCK(bdev_zone_sectors(bdev));
3196 3197 3198
	if (sbi->log_blocks_per_blkz && sbi->log_blocks_per_blkz !=
				__ilog2_u32(sbi->blocks_per_blkz))
		return -EINVAL;
3199
	sbi->log_blocks_per_blkz = __ilog2_u32(sbi->blocks_per_blkz);
3200 3201
	FDEV(devi).nr_blkz = SECTOR_TO_BLOCK(nr_sectors) >>
					sbi->log_blocks_per_blkz;
3202
	if (nr_sectors & (bdev_zone_sectors(bdev) - 1))
3203
		FDEV(devi).nr_blkz++;
3204

3205
	FDEV(devi).blkz_seq = f2fs_kvzalloc(sbi,
3206 3207 3208 3209
					BITS_TO_LONGS(FDEV(devi).nr_blkz)
					* sizeof(unsigned long),
					GFP_KERNEL);
	if (!FDEV(devi).blkz_seq)
3210 3211
		return -ENOMEM;

3212 3213 3214 3215 3216 3217 3218 3219 3220 3221
	/* Get block zones type and zone-capacity */
	FDEV(devi).zone_capacity_blocks = f2fs_kzalloc(sbi,
					FDEV(devi).nr_blkz * sizeof(block_t),
					GFP_KERNEL);
	if (!FDEV(devi).zone_capacity_blocks)
		return -ENOMEM;

	rep_zone_arg.dev = &FDEV(devi);
	rep_zone_arg.zone_cap_mismatch = false;

3222
	ret = blkdev_report_zones(bdev, 0, BLK_ALL_ZONES, f2fs_report_zone_cb,
3223
				  &rep_zone_arg);
3224 3225
	if (ret < 0)
		return ret;
3226

3227 3228 3229 3230 3231
	if (!rep_zone_arg.zone_cap_mismatch) {
		kfree(FDEV(devi).zone_capacity_blocks);
		FDEV(devi).zone_capacity_blocks = NULL;
	}

3232
	return 0;
3233 3234 3235
}
#endif

3236 3237
/*
 * Read f2fs raw super block.
3238 3239 3240
 * Because we have two copies of super block, so read both of them
 * to get the first valid one. If any one of them is broken, we pass
 * them recovery flag back to the caller.
3241
 */
3242
static int read_raw_super_block(struct f2fs_sb_info *sbi,
3243
			struct f2fs_super_block **raw_super,
3244
			int *valid_super_block, int *recovery)
3245
{
3246
	struct super_block *sb = sbi->sb;
3247
	int block;
3248
	struct buffer_head *bh;
3249
	struct f2fs_super_block *super;
3250
	int err = 0;
3251

3252 3253 3254
	super = kzalloc(sizeof(struct f2fs_super_block), GFP_KERNEL);
	if (!super)
		return -ENOMEM;
3255 3256 3257 3258

	for (block = 0; block < 2; block++) {
		bh = sb_bread(sb, block);
		if (!bh) {
3259 3260
			f2fs_err(sbi, "Unable to read %dth superblock",
				 block + 1);
3261
			err = -EIO;
3262
			*recovery = 1;
3263 3264
			continue;
		}
3265

3266
		/* sanity checking of raw super */
3267 3268
		err = sanity_check_raw_super(sbi, bh);
		if (err) {
3269 3270
			f2fs_err(sbi, "Can't find valid F2FS filesystem in %dth superblock",
				 block + 1);
3271
			brelse(bh);
3272
			*recovery = 1;
3273 3274
			continue;
		}
3275

3276
		if (!*raw_super) {
3277 3278
			memcpy(super, bh->b_data + F2FS_SUPER_OFFSET,
							sizeof(*super));
3279 3280 3281 3282
			*valid_super_block = block;
			*raw_super = super;
		}
		brelse(bh);
3283 3284 3285
	}

	/* No valid superblock */
3286
	if (!*raw_super)
3287
		kfree(super);
3288 3289
	else
		err = 0;
3290

3291
	return err;
3292 3293
}

3294
int f2fs_commit_super(struct f2fs_sb_info *sbi, bool recover)
3295
{
3296
	struct buffer_head *bh;
3297
	__u32 crc = 0;
3298 3299
	int err;

3300 3301 3302
	if ((recover && f2fs_readonly(sbi->sb)) ||
				bdev_read_only(sbi->sb->s_bdev)) {
		set_sbi_flag(sbi, SBI_NEED_SB_WRITE);
3303
		return -EROFS;
3304
	}
3305

3306
	/* we should update superblock crc here */
3307
	if (!recover && f2fs_sb_has_sb_chksum(sbi)) {
3308 3309 3310 3311 3312
		crc = f2fs_crc32(sbi, F2FS_RAW_SUPER(sbi),
				offsetof(struct f2fs_super_block, crc));
		F2FS_RAW_SUPER(sbi)->crc = cpu_to_le32(crc);
	}

3313
	/* write back-up superblock first */
3314
	bh = sb_bread(sbi->sb, sbi->valid_super_block ? 0 : 1);
3315 3316
	if (!bh)
		return -EIO;
3317
	err = __f2fs_commit_super(bh, F2FS_RAW_SUPER(sbi));
3318
	brelse(bh);
3319 3320 3321

	/* if we are in recovery path, skip writing valid superblock */
	if (recover || err)
3322
		return err;
3323 3324

	/* write current valid superblock */
3325
	bh = sb_bread(sbi->sb, sbi->valid_super_block);
3326 3327 3328 3329 3330
	if (!bh)
		return -EIO;
	err = __f2fs_commit_super(bh, F2FS_RAW_SUPER(sbi));
	brelse(bh);
	return err;
3331 3332
}

3333 3334 3335
static int f2fs_scan_devices(struct f2fs_sb_info *sbi)
{
	struct f2fs_super_block *raw_super = F2FS_RAW_SUPER(sbi);
3336
	unsigned int max_devices = MAX_DEVICES;
3337 3338
	int i;

3339 3340 3341
	/* Initialize single device information */
	if (!RDEV(0).path[0]) {
		if (!bdev_is_zoned(sbi->sb->s_bdev))
3342
			return 0;
3343 3344
		max_devices = 1;
	}
3345

3346 3347 3348 3349
	/*
	 * Initialize multiple devices information, or single
	 * zoned block device information.
	 */
3350 3351 3352 3353
	sbi->devs = f2fs_kzalloc(sbi,
				 array_size(max_devices,
					    sizeof(struct f2fs_dev_info)),
				 GFP_KERNEL);
3354 3355
	if (!sbi->devs)
		return -ENOMEM;
3356

3357
	for (i = 0; i < max_devices; i++) {
3358

3359 3360 3361 3362 3363 3364 3365
		if (i > 0 && !RDEV(i).path[0])
			break;

		if (max_devices == 1) {
			/* Single zoned block device mount */
			FDEV(0).bdev =
				blkdev_get_by_dev(sbi->sb->s_bdev->bd_dev,
3366
					sbi->sb->s_mode, sbi->sb->s_type);
3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384
		} else {
			/* Multi-device mount */
			memcpy(FDEV(i).path, RDEV(i).path, MAX_PATH_LEN);
			FDEV(i).total_segments =
				le32_to_cpu(RDEV(i).total_segments);
			if (i == 0) {
				FDEV(i).start_blk = 0;
				FDEV(i).end_blk = FDEV(i).start_blk +
				    (FDEV(i).total_segments <<
				    sbi->log_blocks_per_seg) - 1 +
				    le32_to_cpu(raw_super->segment0_blkaddr);
			} else {
				FDEV(i).start_blk = FDEV(i - 1).end_blk + 1;
				FDEV(i).end_blk = FDEV(i).start_blk +
					(FDEV(i).total_segments <<
					sbi->log_blocks_per_seg) - 1;
			}
			FDEV(i).bdev = blkdev_get_by_path(FDEV(i).path,
3385
					sbi->sb->s_mode, sbi->sb->s_type);
3386
		}
3387 3388 3389 3390 3391 3392 3393 3394
		if (IS_ERR(FDEV(i).bdev))
			return PTR_ERR(FDEV(i).bdev);

		/* to release errored devices */
		sbi->s_ndevs = i + 1;

#ifdef CONFIG_BLK_DEV_ZONED
		if (bdev_zoned_model(FDEV(i).bdev) == BLK_ZONED_HM &&
3395
				!f2fs_sb_has_blkzoned(sbi)) {
3396
			f2fs_err(sbi, "Zoned block device feature not enabled\n");
3397 3398 3399 3400
			return -EINVAL;
		}
		if (bdev_zoned_model(FDEV(i).bdev) != BLK_ZONED_NONE) {
			if (init_blkz_info(sbi, i)) {
3401
				f2fs_err(sbi, "Failed to initialize F2FS blkzone information");
3402 3403
				return -EINVAL;
			}
3404 3405
			if (max_devices == 1)
				break;
3406 3407 3408 3409 3410 3411
			f2fs_info(sbi, "Mount Device [%2d]: %20s, %8u, %8x - %8x (zone: %s)",
				  i, FDEV(i).path,
				  FDEV(i).total_segments,
				  FDEV(i).start_blk, FDEV(i).end_blk,
				  bdev_zoned_model(FDEV(i).bdev) == BLK_ZONED_HA ?
				  "Host-aware" : "Host-managed");
3412 3413 3414
			continue;
		}
#endif
3415 3416 3417 3418 3419 3420 3421
		f2fs_info(sbi, "Mount Device [%2d]: %20s, %8u, %8x - %8x",
			  i, FDEV(i).path,
			  FDEV(i).total_segments,
			  FDEV(i).start_blk, FDEV(i).end_blk);
	}
	f2fs_info(sbi,
		  "IO Block Size: %8d KB", F2FS_IO_SIZE_KB(sbi));
3422 3423 3424
	return 0;
}

3425 3426 3427
static int f2fs_setup_casefold(struct f2fs_sb_info *sbi)
{
#ifdef CONFIG_UNICODE
3428
	if (f2fs_sb_has_casefold(sbi) && !sbi->sb->s_encoding) {
3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452
		const struct f2fs_sb_encodings *encoding_info;
		struct unicode_map *encoding;
		__u16 encoding_flags;

		if (f2fs_sb_read_encoding(sbi->raw_super, &encoding_info,
					  &encoding_flags)) {
			f2fs_err(sbi,
				 "Encoding requested by superblock is unknown");
			return -EINVAL;
		}

		encoding = utf8_load(encoding_info->version);
		if (IS_ERR(encoding)) {
			f2fs_err(sbi,
				 "can't mount with superblock charset: %s-%s "
				 "not supported by the kernel. flags: 0x%x.",
				 encoding_info->name, encoding_info->version,
				 encoding_flags);
			return PTR_ERR(encoding);
		}
		f2fs_info(sbi, "Using encoding defined by superblock: "
			 "%s-%s with flags 0x%hx", encoding_info->name,
			 encoding_info->version?:"\b", encoding_flags);

3453 3454
		sbi->sb->s_encoding = encoding;
		sbi->sb->s_encoding_flags = encoding_flags;
3455 3456 3457 3458 3459 3460 3461 3462 3463 3464
	}
#else
	if (f2fs_sb_has_casefold(sbi)) {
		f2fs_err(sbi, "Filesystem with casefold feature cannot be mounted without CONFIG_UNICODE");
		return -EINVAL;
	}
#endif
	return 0;
}

3465 3466 3467 3468 3469 3470
static void f2fs_tuning_parameters(struct f2fs_sb_info *sbi)
{
	struct f2fs_sm_info *sm_i = SM_I(sbi);

	/* adjust parameters according to the volume size */
	if (sm_i->main_segments <= SMALL_VOLUME_SEGMENTS) {
3471
		F2FS_OPTION(sbi).alloc_mode = ALLOC_MODE_REUSE;
3472 3473 3474
		sm_i->dcc_info->discard_granularity = 1;
		sm_i->ipu_policy = 1 << F2FS_IPU_FORCE;
	}
3475 3476

	sbi->readdir_ra = 1;
3477 3478
}

3479 3480 3481
static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
{
	struct f2fs_sb_info *sbi;
3482
	struct f2fs_super_block *raw_super;
3483
	struct inode *root;
3484
	int err;
3485
	bool skip_recovery = false, need_fsck = false;
3486
	char *options = NULL;
3487
	int recovery, i, valid_super_block;
3488
	struct curseg_info *seg_i;
3489
	int retry_cnt = 1;
3490

3491
try_onemore:
3492 3493
	err = -EINVAL;
	raw_super = NULL;
3494
	valid_super_block = -1;
3495 3496
	recovery = 0;

3497 3498 3499 3500 3501
	/* allocate memory for f2fs-specific super block info */
	sbi = kzalloc(sizeof(struct f2fs_sb_info), GFP_KERNEL);
	if (!sbi)
		return -ENOMEM;

3502 3503
	sbi->sb = sb;

3504 3505 3506
	/* Load the checksum driver */
	sbi->s_chksum_driver = crypto_alloc_shash("crc32", 0, 0);
	if (IS_ERR(sbi->s_chksum_driver)) {
3507
		f2fs_err(sbi, "Cannot load crc32 driver.");
3508 3509 3510 3511 3512
		err = PTR_ERR(sbi->s_chksum_driver);
		sbi->s_chksum_driver = NULL;
		goto free_sbi;
	}

3513
	/* set a block size */
3514
	if (unlikely(!sb_set_blocksize(sb, F2FS_BLKSIZE))) {
3515
		f2fs_err(sbi, "unable to set blocksize");
3516
		goto free_sbi;
3517
	}
3518

3519
	err = read_raw_super_block(sbi, &raw_super, &valid_super_block,
3520
								&recovery);
3521 3522 3523
	if (err)
		goto free_sbi;

3524
	sb->s_fs_info = sbi;
3525 3526
	sbi->raw_super = raw_super;

C
Chao Yu 已提交
3527
	/* precompute checksum seed for metadata */
3528
	if (f2fs_sb_has_inode_chksum(sbi))
C
Chao Yu 已提交
3529 3530 3531
		sbi->s_chksum_seed = f2fs_chksum(sbi, ~0, raw_super->uuid,
						sizeof(raw_super->uuid));

3532
	default_options(sbi);
3533
	/* parse mount options */
3534 3535 3536
	options = kstrdup((const char *)data, GFP_KERNEL);
	if (data && !options) {
		err = -ENOMEM;
3537
		goto free_sb_buf;
3538 3539
	}

3540
	err = parse_options(sb, options, false);
3541 3542
	if (err)
		goto free_options;
3543

3544 3545 3546
	sbi->max_file_blocks = max_file_blocks();
	sb->s_maxbytes = sbi->max_file_blocks <<
				le32_to_cpu(raw_super->log_blocksize);
3547 3548
	sb->s_max_links = F2FS_LINK_MAX;

3549 3550 3551 3552
	err = f2fs_setup_casefold(sbi);
	if (err)
		goto free_options;

3553 3554
#ifdef CONFIG_QUOTA
	sb->dq_op = &f2fs_quota_operations;
3555
	sb->s_qcop = &f2fs_quotactl_ops;
C
Chao Yu 已提交
3556
	sb->s_quota_types = QTYPE_MASK_USR | QTYPE_MASK_GRP | QTYPE_MASK_PRJ;
3557

3558
	if (f2fs_sb_has_quota_ino(sbi)) {
3559 3560 3561 3562 3563
		for (i = 0; i < MAXQUOTAS; i++) {
			if (f2fs_qf_ino(sbi->sb, i))
				sbi->nquota_files++;
		}
	}
3564 3565
#endif

3566
	sb->s_op = &f2fs_sops;
3567
#ifdef CONFIG_FS_ENCRYPTION
3568
	sb->s_cop = &f2fs_cryptops;
3569 3570 3571
#endif
#ifdef CONFIG_FS_VERITY
	sb->s_vop = &f2fs_verityops;
3572
#endif
3573 3574 3575 3576
	sb->s_xattr = f2fs_xattr_handlers;
	sb->s_export_op = &f2fs_export_ops;
	sb->s_magic = F2FS_SUPER_MAGIC;
	sb->s_time_gran = 1;
3577 3578
	sb->s_flags = (sb->s_flags & ~SB_POSIXACL) |
		(test_opt(sbi, POSIX_ACL) ? SB_POSIXACL : 0);
3579
	memcpy(&sb->s_uuid, raw_super->uuid, sizeof(raw_super->uuid));
3580
	sb->s_iflags |= SB_I_CGROUPWB;
3581 3582

	/* init f2fs-specific super block info */
3583
	sbi->valid_super_block = valid_super_block;
3584
	init_rwsem(&sbi->gc_lock);
3585
	mutex_init(&sbi->writepages);
3586
	init_rwsem(&sbi->cp_global_sem);
3587
	init_rwsem(&sbi->node_write);
3588
	init_rwsem(&sbi->node_change);
3589 3590 3591

	/* disallow all the data/node/meta page writes */
	set_sbi_flag(sbi, SBI_POR_DOING);
3592
	spin_lock_init(&sbi->stat_lock);
3593

C
Chao Yu 已提交
3594 3595 3596
	/* init iostat info */
	spin_lock_init(&sbi->iostat_lock);
	sbi->iostat_enable = false;
3597
	sbi->iostat_period_ms = DEFAULT_IOSTAT_PERIOD_MS;
C
Chao Yu 已提交
3598

3599
	for (i = 0; i < NR_PAGE_TYPE; i++) {
J
Jaegeuk Kim 已提交
3600 3601 3602
		int n = (i == META) ? 1: NR_TEMP_TYPE;
		int j;

3603 3604 3605 3606 3607
		sbi->write_io[i] =
			f2fs_kmalloc(sbi,
				     array_size(n,
						sizeof(struct f2fs_bio_info)),
				     GFP_KERNEL);
3608 3609
		if (!sbi->write_io[i]) {
			err = -ENOMEM;
3610
			goto free_bio_info;
3611
		}
J
Jaegeuk Kim 已提交
3612 3613 3614 3615 3616

		for (j = HOT; j < n; j++) {
			init_rwsem(&sbi->write_io[i][j].io_rwsem);
			sbi->write_io[i][j].sbi = sbi;
			sbi->write_io[i][j].bio = NULL;
3617 3618
			spin_lock_init(&sbi->write_io[i][j].io_lock);
			INIT_LIST_HEAD(&sbi->write_io[i][j].io_list);
C
Chao Yu 已提交
3619 3620
			INIT_LIST_HEAD(&sbi->write_io[i][j].bio_list);
			init_rwsem(&sbi->write_io[i][j].bio_list_lock);
J
Jaegeuk Kim 已提交
3621
		}
3622
	}
3623

3624
	init_rwsem(&sbi->cp_rwsem);
3625
	init_rwsem(&sbi->quota_sem);
3626
	init_waitqueue_head(&sbi->cp_wait);
3627 3628
	init_sb_info(sbi);

3629 3630
	err = init_percpu_info(sbi);
	if (err)
3631
		goto free_bio_info;
3632

3633
	if (F2FS_IO_ALIGNED(sbi)) {
3634
		sbi->write_io_dummy =
3635
			mempool_create_page_pool(2 * (F2FS_IO_SIZE(sbi) - 1), 0);
3636 3637
		if (!sbi->write_io_dummy) {
			err = -ENOMEM;
3638
			goto free_percpu;
3639
		}
3640 3641
	}

3642 3643 3644 3645
	/* init per sbi slab cache */
	err = f2fs_init_xattr_caches(sbi);
	if (err)
		goto free_io_dummy;
3646 3647 3648
	err = f2fs_init_page_array_cache(sbi);
	if (err)
		goto free_xattr_cache;
3649

3650 3651 3652
	/* get an inode for meta space */
	sbi->meta_inode = f2fs_iget(sb, F2FS_META_INO(sbi));
	if (IS_ERR(sbi->meta_inode)) {
3653
		f2fs_err(sbi, "Failed to read F2FS meta data inode");
3654
		err = PTR_ERR(sbi->meta_inode);
3655
		goto free_page_array_cache;
3656 3657
	}

3658
	err = f2fs_get_valid_checkpoint(sbi);
3659
	if (err) {
3660
		f2fs_err(sbi, "Failed to get valid F2FS checkpoint");
3661
		goto free_meta_inode;
3662
	}
3663

3664 3665
	if (__is_set_ckpt_flags(F2FS_CKPT(sbi), CP_QUOTA_NEED_FSCK_FLAG))
		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
3666 3667 3668 3669
	if (__is_set_ckpt_flags(F2FS_CKPT(sbi), CP_DISABLED_QUICK_FLAG)) {
		set_sbi_flag(sbi, SBI_CP_DISABLED_QUICK);
		sbi->interval_time[DISABLE_TIME] = DEF_DISABLE_QUICK_INTERVAL;
	}
3670

3671 3672 3673
	if (__is_set_ckpt_flags(F2FS_CKPT(sbi), CP_FSCK_FLAG))
		set_sbi_flag(sbi, SBI_NEED_FSCK);

3674 3675 3676
	/* Initialize device list */
	err = f2fs_scan_devices(sbi);
	if (err) {
3677
		f2fs_err(sbi, "Failed to find devices");
3678 3679 3680
		goto free_devices;
	}

3681 3682 3683 3684 3685 3686
	err = f2fs_init_post_read_wq(sbi);
	if (err) {
		f2fs_err(sbi, "Failed to initialize post read workqueue");
		goto free_devices;
	}

3687 3688
	sbi->total_valid_node_count =
				le32_to_cpu(sbi->ckpt->valid_node_count);
3689 3690
	percpu_counter_set(&sbi->total_valid_inode_count,
				le32_to_cpu(sbi->ckpt->valid_inode_count));
3691 3692 3693 3694
	sbi->user_block_count = le64_to_cpu(sbi->ckpt->user_block_count);
	sbi->total_valid_block_count =
				le64_to_cpu(sbi->ckpt->valid_block_count);
	sbi->last_valid_block_count = sbi->total_valid_block_count;
3695
	sbi->reserved_blocks = 0;
3696
	sbi->current_reserved_blocks = 0;
3697
	limit_reserve_root(sbi);
3698
	adjust_unusable_cap_perc(sbi);
3699

3700 3701 3702 3703
	for (i = 0; i < NR_INODE_TYPE; i++) {
		INIT_LIST_HEAD(&sbi->inode_list[i]);
		spin_lock_init(&sbi->inode_lock[i]);
	}
3704
	mutex_init(&sbi->flush_lock);
3705

3706
	f2fs_init_extent_cache_info(sbi);
3707

3708
	f2fs_init_ino_entry_info(sbi);
3709

3710 3711
	f2fs_init_fsync_node_info(sbi);

3712
	/* setup f2fs internal modules */
3713
	err = f2fs_build_segment_manager(sbi);
3714
	if (err) {
3715 3716
		f2fs_err(sbi, "Failed to initialize F2FS segment manager (%d)",
			 err);
3717
		goto free_sm;
3718
	}
3719
	err = f2fs_build_node_manager(sbi);
3720
	if (err) {
3721 3722
		f2fs_err(sbi, "Failed to initialize F2FS node manager (%d)",
			 err);
3723
		goto free_nm;
3724
	}
3725

3726
	/* For write statistics */
3727
	sbi->sectors_written_start = f2fs_get_sectors_written(sbi);
3728 3729 3730 3731 3732

	/* Read accumulated write IO statistics if exists */
	seg_i = CURSEG_I(sbi, CURSEG_HOT_NODE);
	if (__exist_node_summaries(sbi))
		sbi->kbytes_written =
3733
			le64_to_cpu(seg_i->journal->info.kbytes_written);
3734

3735
	f2fs_build_gc_manager(sbi);
3736

3737 3738 3739 3740
	err = f2fs_build_stats(sbi);
	if (err)
		goto free_nm;

3741 3742 3743
	/* get an inode for node space */
	sbi->node_inode = f2fs_iget(sb, F2FS_NODE_INO(sbi));
	if (IS_ERR(sbi->node_inode)) {
3744
		f2fs_err(sbi, "Failed to read node inode");
3745
		err = PTR_ERR(sbi->node_inode);
3746
		goto free_stats;
3747 3748 3749 3750 3751
	}

	/* read root inode and dentry */
	root = f2fs_iget(sb, F2FS_ROOT_INO(sbi));
	if (IS_ERR(root)) {
3752
		f2fs_err(sbi, "Failed to read root inode");
3753
		err = PTR_ERR(root);
3754
		goto free_node_inode;
3755
	}
3756 3757
	if (!S_ISDIR(root->i_mode) || !root->i_blocks ||
			!root->i_size || !root->i_nlink) {
3758
		iput(root);
3759
		err = -EINVAL;
3760
		goto free_node_inode;
3761
	}
3762 3763 3764 3765

	sb->s_root = d_make_root(root); /* allocate root dentry */
	if (!sb->s_root) {
		err = -ENOMEM;
3766
		goto free_node_inode;
3767 3768
	}

3769
	err = f2fs_register_sysfs(sbi);
3770
	if (err)
C
Chao Yu 已提交
3771
		goto free_root_inode;
3772

3773
#ifdef CONFIG_QUOTA
3774
	/* Enable quota usage during mount */
3775
	if (f2fs_sb_has_quota_ino(sbi) && !f2fs_readonly(sb)) {
3776
		err = f2fs_enable_quotas(sb);
3777
		if (err)
3778
			f2fs_err(sbi, "Cannot turn on quotas: error %d", err);
3779 3780
	}
#endif
3781
	/* if there are any orphan inodes, free them */
3782
	err = f2fs_recover_orphan_inodes(sbi);
3783
	if (err)
3784
		goto free_meta;
3785

3786
	if (unlikely(is_set_ckpt_flags(sbi, CP_DISABLED_FLAG)))
3787
		goto reset_checkpoint;
3788

3789
	/* recover fsynced data */
3790 3791
	if (!test_opt(sbi, DISABLE_ROLL_FORWARD) &&
			!test_opt(sbi, NORECOVERY)) {
3792 3793 3794 3795
		/*
		 * mount should be failed, when device has readonly mode, and
		 * previous checkpoint was not done by clean system shutdown.
		 */
3796 3797 3798
		if (f2fs_hw_is_readonly(sbi)) {
			if (!is_set_ckpt_flags(sbi, CP_UMOUNT_FLAG)) {
				err = -EROFS;
3799
				f2fs_err(sbi, "Need to recover fsync data, but write access unavailable");
3800 3801
				goto free_meta;
			}
3802
			f2fs_info(sbi, "write access unavailable, skipping recovery");
3803
			goto reset_checkpoint;
3804
		}
3805 3806 3807 3808

		if (need_fsck)
			set_sbi_flag(sbi, SBI_NEED_FSCK);

3809 3810
		if (skip_recovery)
			goto reset_checkpoint;
3811

3812
		err = f2fs_recover_fsync_data(sbi, false);
3813
		if (err < 0) {
3814 3815
			if (err != -ENOMEM)
				skip_recovery = true;
3816
			need_fsck = true;
3817 3818
			f2fs_err(sbi, "Cannot recover all fsync data errno=%d",
				 err);
3819
			goto free_meta;
3820
		}
3821
	} else {
3822
		err = f2fs_recover_fsync_data(sbi, true);
3823 3824 3825

		if (!f2fs_readonly(sb) && err > 0) {
			err = -EINVAL;
3826
			f2fs_err(sbi, "Need to recover fsync data");
3827
			goto free_meta;
3828
		}
3829
	}
3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840

	/*
	 * If the f2fs is not readonly and fsync data recovery succeeds,
	 * check zoned block devices' write pointer consistency.
	 */
	if (!err && !f2fs_readonly(sb) && f2fs_sb_has_blkzoned(sbi)) {
		err = f2fs_check_write_pointer(sbi);
		if (err)
			goto free_meta;
	}

3841
reset_checkpoint:
3842 3843
	f2fs_init_inmem_curseg(sbi);

3844
	/* f2fs_recover_fsync_data() cleared this already */
3845
	clear_sbi_flag(sbi, SBI_POR_DOING);
3846

3847 3848 3849
	if (test_opt(sbi, DISABLE_CHECKPOINT)) {
		err = f2fs_disable_checkpoint(sbi);
		if (err)
3850
			goto sync_free_meta;
3851 3852 3853 3854
	} else if (is_set_ckpt_flags(sbi, CP_DISABLED_FLAG)) {
		f2fs_enable_checkpoint(sbi);
	}

3855 3856 3857 3858
	/*
	 * If filesystem is not mounted as read-only then
	 * do start the gc_thread.
	 */
3859
	if (F2FS_OPTION(sbi).bggc_mode != BGGC_MODE_OFF && !f2fs_readonly(sb)) {
3860
		/* After POR, we can run background GC thread.*/
3861
		err = f2fs_start_gc_thread(sbi);
3862
		if (err)
3863
			goto sync_free_meta;
3864
	}
3865
	kvfree(options);
3866 3867

	/* recover broken superblock */
3868
	if (recovery) {
3869
		err = f2fs_commit_super(sbi, true);
3870 3871
		f2fs_info(sbi, "Try to recover %dth superblock, ret: %d",
			  sbi->valid_super_block ? 1 : 2, err);
3872 3873
	}

3874 3875
	f2fs_join_shrinker(sbi);

3876 3877
	f2fs_tuning_parameters(sbi);

3878 3879
	f2fs_notice(sbi, "Mounted with checkpoint version = %llx",
		    cur_cp_version(F2FS_CKPT(sbi)));
3880
	f2fs_update_time(sbi, CP_TIME);
3881
	f2fs_update_time(sbi, REQ_TIME);
3882
	clear_sbi_flag(sbi, SBI_CP_DISABLED_QUICK);
3883
	return 0;
3884

3885 3886 3887
sync_free_meta:
	/* safe to flush all the data */
	sync_filesystem(sbi->sb);
3888
	retry_cnt = 0;
3889

3890
free_meta:
3891
#ifdef CONFIG_QUOTA
3892
	f2fs_truncate_quota_inode_pages(sb);
3893
	if (f2fs_sb_has_quota_ino(sbi) && !f2fs_readonly(sb))
3894 3895
		f2fs_quota_off_umount(sbi->sb);
#endif
3896
	/*
3897
	 * Some dirty meta pages can be produced by f2fs_recover_orphan_inodes()
3898
	 * failed by EIO. Then, iput(node_inode) can trigger balance_fs_bg()
3899 3900
	 * followed by f2fs_write_checkpoint() through f2fs_write_node_pages(), which
	 * falls into an infinite loop in f2fs_sync_meta_pages().
3901 3902
	 */
	truncate_inode_pages_final(META_MAPPING(sbi));
3903 3904
	/* evict some inodes being cached by GC */
	evict_inodes(sb);
3905
	f2fs_unregister_sysfs(sbi);
3906 3907 3908 3909
free_root_inode:
	dput(sb->s_root);
	sb->s_root = NULL;
free_node_inode:
3910
	f2fs_release_ino_entry(sbi, true);
3911
	truncate_inode_pages_final(NODE_MAPPING(sbi));
3912
	iput(sbi->node_inode);
3913
	sbi->node_inode = NULL;
3914 3915
free_stats:
	f2fs_destroy_stats(sbi);
3916
free_nm:
3917
	f2fs_destroy_node_manager(sbi);
3918
free_sm:
3919
	f2fs_destroy_segment_manager(sbi);
3920
	f2fs_destroy_post_read_wq(sbi);
3921 3922
free_devices:
	destroy_device_list(sbi);
3923
	kvfree(sbi->ckpt);
3924 3925 3926
free_meta_inode:
	make_bad_inode(sbi->meta_inode);
	iput(sbi->meta_inode);
3927
	sbi->meta_inode = NULL;
3928 3929
free_page_array_cache:
	f2fs_destroy_page_array_cache(sbi);
3930 3931
free_xattr_cache:
	f2fs_destroy_xattr_caches(sbi);
3932 3933
free_io_dummy:
	mempool_destroy(sbi->write_io_dummy);
3934 3935 3936
free_percpu:
	destroy_percpu_info(sbi);
free_bio_info:
J
Jaegeuk Kim 已提交
3937
	for (i = 0; i < NR_PAGE_TYPE; i++)
3938
		kvfree(sbi->write_io[i]);
3939 3940

#ifdef CONFIG_UNICODE
3941
	utf8_unload(sb->s_encoding);
3942
	sb->s_encoding = NULL;
3943
#endif
3944
free_options:
3945 3946
#ifdef CONFIG_QUOTA
	for (i = 0; i < MAXQUOTAS; i++)
3947
		kfree(F2FS_OPTION(sbi).s_qf_names[i]);
3948
#endif
3949
	fscrypt_free_dummy_policy(&F2FS_OPTION(sbi).dummy_enc_policy);
3950
	kvfree(options);
3951
free_sb_buf:
3952
	kfree(raw_super);
3953
free_sbi:
3954 3955
	if (sbi->s_chksum_driver)
		crypto_free_shash(sbi->s_chksum_driver);
3956
	kfree(sbi);
3957 3958

	/* give only one another chance */
3959 3960
	if (retry_cnt > 0 && skip_recovery) {
		retry_cnt--;
3961 3962 3963
		shrink_dcache_sb(sb);
		goto try_onemore;
	}
3964 3965 3966 3967 3968 3969 3970 3971 3972
	return err;
}

static struct dentry *f2fs_mount(struct file_system_type *fs_type, int flags,
			const char *dev_name, void *data)
{
	return mount_bdev(fs_type, flags, dev_name, data, f2fs_fill_super);
}

3973 3974
static void kill_f2fs_super(struct super_block *sb)
{
3975
	if (sb->s_root) {
3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988
		struct f2fs_sb_info *sbi = F2FS_SB(sb);

		set_sbi_flag(sbi, SBI_IS_CLOSE);
		f2fs_stop_gc_thread(sbi);
		f2fs_stop_discard_thread(sbi);

		if (is_sbi_flag_set(sbi, SBI_IS_DIRTY) ||
				!is_set_ckpt_flags(sbi, CP_UMOUNT_FLAG)) {
			struct cp_control cpc = {
				.reason = CP_UMOUNT,
			};
			f2fs_write_checkpoint(sbi, &cpc);
		}
3989 3990 3991

		if (is_sbi_flag_set(sbi, SBI_IS_RECOVERED) && f2fs_readonly(sb))
			sb->s_flags &= ~SB_RDONLY;
3992
	}
3993 3994 3995
	kill_block_super(sb);
}

3996 3997 3998 3999
static struct file_system_type f2fs_fs_type = {
	.owner		= THIS_MODULE,
	.name		= "f2fs",
	.mount		= f2fs_mount,
4000
	.kill_sb	= kill_f2fs_super,
4001 4002
	.fs_flags	= FS_REQUIRES_DEV,
};
4003
MODULE_ALIAS_FS("f2fs");
4004

4005
static int __init init_inodecache(void)
4006
{
4007 4008 4009
	f2fs_inode_cachep = kmem_cache_create("f2fs_inode_cache",
			sizeof(struct f2fs_inode_info), 0,
			SLAB_RECLAIM_ACCOUNT|SLAB_ACCOUNT, NULL);
4010
	if (!f2fs_inode_cachep)
4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028
		return -ENOMEM;
	return 0;
}

static void destroy_inodecache(void)
{
	/*
	 * Make sure all delayed rcu free inodes are flushed before we
	 * destroy cache.
	 */
	rcu_barrier();
	kmem_cache_destroy(f2fs_inode_cachep);
}

static int __init init_f2fs_fs(void)
{
	int err;

4029 4030 4031 4032 4033 4034
	if (PAGE_SIZE != F2FS_BLKSIZE) {
		printk("F2FS not supported on PAGE_SIZE(%lu) != %d\n",
				PAGE_SIZE, F2FS_BLKSIZE);
		return -EINVAL;
	}

4035 4036
	f2fs_build_trace_ios();

4037 4038 4039
	err = init_inodecache();
	if (err)
		goto fail;
4040
	err = f2fs_create_node_manager_caches();
4041
	if (err)
4042
		goto free_inodecache;
4043
	err = f2fs_create_segment_manager_caches();
4044
	if (err)
4045
		goto free_node_manager_caches;
4046
	err = f2fs_create_checkpoint_caches();
4047
	if (err)
4048
		goto free_segment_manager_caches;
4049
	err = f2fs_create_extent_cache();
4050 4051
	if (err)
		goto free_checkpoint_caches;
4052
	err = f2fs_create_garbage_collection_cache();
C
Chao Yu 已提交
4053
	if (err)
4054
		goto free_extent_cache;
4055 4056 4057
	err = f2fs_init_sysfs();
	if (err)
		goto free_garbage_collection_cache;
4058
	err = register_shrinker(&f2fs_shrinker_info);
4059
	if (err)
C
Chao Yu 已提交
4060
		goto free_sysfs;
4061 4062 4063
	err = register_filesystem(&f2fs_fs_type);
	if (err)
		goto free_shrinker;
4064
	f2fs_create_root_stats();
4065 4066 4067
	err = f2fs_init_post_read_processing();
	if (err)
		goto free_root_stats;
C
Chao Yu 已提交
4068 4069 4070
	err = f2fs_init_bio_entry_cache();
	if (err)
		goto free_post_read;
4071 4072 4073
	err = f2fs_init_bioset();
	if (err)
		goto free_bio_enrty_cache;
4074 4075 4076
	err = f2fs_init_compress_mempool();
	if (err)
		goto free_bioset;
4077 4078 4079
	err = f2fs_init_compress_cache();
	if (err)
		goto free_compress_mempool;
4080
	return 0;
4081 4082
free_compress_mempool:
	f2fs_destroy_compress_mempool();
4083 4084
free_bioset:
	f2fs_destroy_bioset();
4085 4086
free_bio_enrty_cache:
	f2fs_destroy_bio_entry_cache();
C
Chao Yu 已提交
4087 4088
free_post_read:
	f2fs_destroy_post_read_processing();
4089 4090
free_root_stats:
	f2fs_destroy_root_stats();
4091
	unregister_filesystem(&f2fs_fs_type);
4092 4093
free_shrinker:
	unregister_shrinker(&f2fs_shrinker_info);
C
Chao Yu 已提交
4094
free_sysfs:
4095
	f2fs_exit_sysfs();
4096 4097
free_garbage_collection_cache:
	f2fs_destroy_garbage_collection_cache();
4098
free_extent_cache:
4099
	f2fs_destroy_extent_cache();
4100
free_checkpoint_caches:
4101
	f2fs_destroy_checkpoint_caches();
4102
free_segment_manager_caches:
4103
	f2fs_destroy_segment_manager_caches();
4104
free_node_manager_caches:
4105
	f2fs_destroy_node_manager_caches();
4106 4107
free_inodecache:
	destroy_inodecache();
4108 4109 4110 4111 4112 4113
fail:
	return err;
}

static void __exit exit_f2fs_fs(void)
{
4114
	f2fs_destroy_compress_cache();
4115
	f2fs_destroy_compress_mempool();
4116
	f2fs_destroy_bioset();
C
Chao Yu 已提交
4117
	f2fs_destroy_bio_entry_cache();
4118
	f2fs_destroy_post_read_processing();
4119
	f2fs_destroy_root_stats();
4120
	unregister_filesystem(&f2fs_fs_type);
4121
	unregister_shrinker(&f2fs_shrinker_info);
4122
	f2fs_exit_sysfs();
4123
	f2fs_destroy_garbage_collection_cache();
4124 4125 4126 4127
	f2fs_destroy_extent_cache();
	f2fs_destroy_checkpoint_caches();
	f2fs_destroy_segment_manager_caches();
	f2fs_destroy_node_manager_caches();
4128
	destroy_inodecache();
4129
	f2fs_destroy_trace_ios();
4130 4131 4132 4133 4134 4135 4136 4137
}

module_init(init_f2fs_fs)
module_exit(exit_f2fs_fs)

MODULE_AUTHOR("Samsung Electronics's Praesto Team");
MODULE_DESCRIPTION("Flash Friendly File System");
MODULE_LICENSE("GPL");
4138
反馈
建议
客服 返回
顶部