super.c 111.2 KB
Newer Older
1
/*
2
 *  linux/fs/ext4/super.c
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
 *
 * Copyright (C) 1992, 1993, 1994, 1995
 * Remy Card (card@masi.ibp.fr)
 * Laboratoire MASI - Institut Blaise Pascal
 * Universite Pierre et Marie Curie (Paris VI)
 *
 *  from
 *
 *  linux/fs/minix/inode.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *
 *  Big-endian to little-endian byte-swapping/bitmaps by
 *        David S. Miller (davem@caip.rutgers.edu), 1995
 */

#include <linux/module.h>
#include <linux/string.h>
#include <linux/fs.h>
#include <linux/time.h>
23
#include <linux/jbd2.h>
24 25 26 27 28 29
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/blkdev.h>
#include <linux/parser.h>
#include <linux/smp_lock.h>
#include <linux/buffer_head.h>
30
#include <linux/exportfs.h>
31 32 33 34 35 36
#include <linux/vfs.h>
#include <linux/random.h>
#include <linux/mount.h>
#include <linux/namei.h>
#include <linux/quotaops.h>
#include <linux/seq_file.h>
37
#include <linux/proc_fs.h>
T
Theodore Ts'o 已提交
38
#include <linux/ctype.h>
39
#include <linux/marker.h>
V
Vignesh Babu 已提交
40
#include <linux/log2.h>
A
Andreas Dilger 已提交
41
#include <linux/crc16.h>
42 43
#include <asm/uaccess.h>

44 45
#include "ext4.h"
#include "ext4_jbd2.h"
46 47 48
#include "xattr.h"
#include "acl.h"
#include "namei.h"
A
Andreas Dilger 已提交
49
#include "group.h"
50

51
struct proc_dir_entry *ext4_proc_root;
T
Theodore Ts'o 已提交
52
static struct kset *ext4_kset;
53

54
static int ext4_load_journal(struct super_block *, struct ext4_super_block *,
55
			     unsigned long journal_devnum);
56
static int ext4_commit_super(struct super_block *sb,
57 58 59 60 61
			      struct ext4_super_block *es, int sync);
static void ext4_mark_recovery_complete(struct super_block *sb,
					struct ext4_super_block *es);
static void ext4_clear_journal_err(struct super_block *sb,
				   struct ext4_super_block *es);
62
static int ext4_sync_fs(struct super_block *sb, int wait);
63
static const char *ext4_decode_error(struct super_block *sb, int errno,
64
				     char nbuf[16]);
65 66
static int ext4_remount(struct super_block *sb, int *flags, char *data);
static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf);
67
static int ext4_unfreeze(struct super_block *sb);
68
static void ext4_write_super(struct super_block *sb);
69
static int ext4_freeze(struct super_block *sb);
70

L
Laurent Vivier 已提交
71

72 73
ext4_fsblk_t ext4_block_bitmap(struct super_block *sb,
			       struct ext4_group_desc *bg)
L
Laurent Vivier 已提交
74
{
75
	return le32_to_cpu(bg->bg_block_bitmap_lo) |
76
		(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
77
		(ext4_fsblk_t)le32_to_cpu(bg->bg_block_bitmap_hi) << 32 : 0);
L
Laurent Vivier 已提交
78 79
}

80 81
ext4_fsblk_t ext4_inode_bitmap(struct super_block *sb,
			       struct ext4_group_desc *bg)
L
Laurent Vivier 已提交
82
{
83
	return le32_to_cpu(bg->bg_inode_bitmap_lo) |
84
		(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
85
		(ext4_fsblk_t)le32_to_cpu(bg->bg_inode_bitmap_hi) << 32 : 0);
L
Laurent Vivier 已提交
86 87
}

88 89
ext4_fsblk_t ext4_inode_table(struct super_block *sb,
			      struct ext4_group_desc *bg)
L
Laurent Vivier 已提交
90
{
91
	return le32_to_cpu(bg->bg_inode_table_lo) |
92
		(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
93
		(ext4_fsblk_t)le32_to_cpu(bg->bg_inode_table_hi) << 32 : 0);
L
Laurent Vivier 已提交
94 95
}

96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
__u32 ext4_free_blks_count(struct super_block *sb,
			      struct ext4_group_desc *bg)
{
	return le16_to_cpu(bg->bg_free_blocks_count_lo) |
		(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
		(__u32)le16_to_cpu(bg->bg_free_blocks_count_hi) << 16 : 0);
}

__u32 ext4_free_inodes_count(struct super_block *sb,
			      struct ext4_group_desc *bg)
{
	return le16_to_cpu(bg->bg_free_inodes_count_lo) |
		(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
		(__u32)le16_to_cpu(bg->bg_free_inodes_count_hi) << 16 : 0);
}

__u32 ext4_used_dirs_count(struct super_block *sb,
			      struct ext4_group_desc *bg)
{
	return le16_to_cpu(bg->bg_used_dirs_count_lo) |
		(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
		(__u32)le16_to_cpu(bg->bg_used_dirs_count_hi) << 16 : 0);
}

__u32 ext4_itable_unused_count(struct super_block *sb,
			      struct ext4_group_desc *bg)
{
	return le16_to_cpu(bg->bg_itable_unused_lo) |
		(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
		(__u32)le16_to_cpu(bg->bg_itable_unused_hi) << 16 : 0);
}

128 129
void ext4_block_bitmap_set(struct super_block *sb,
			   struct ext4_group_desc *bg, ext4_fsblk_t blk)
L
Laurent Vivier 已提交
130
{
131
	bg->bg_block_bitmap_lo = cpu_to_le32((u32)blk);
132 133
	if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
		bg->bg_block_bitmap_hi = cpu_to_le32(blk >> 32);
L
Laurent Vivier 已提交
134 135
}

136 137
void ext4_inode_bitmap_set(struct super_block *sb,
			   struct ext4_group_desc *bg, ext4_fsblk_t blk)
L
Laurent Vivier 已提交
138
{
139
	bg->bg_inode_bitmap_lo  = cpu_to_le32((u32)blk);
140 141
	if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
		bg->bg_inode_bitmap_hi = cpu_to_le32(blk >> 32);
L
Laurent Vivier 已提交
142 143
}

144 145
void ext4_inode_table_set(struct super_block *sb,
			  struct ext4_group_desc *bg, ext4_fsblk_t blk)
L
Laurent Vivier 已提交
146
{
147
	bg->bg_inode_table_lo = cpu_to_le32((u32)blk);
148 149
	if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
		bg->bg_inode_table_hi = cpu_to_le32(blk >> 32);
L
Laurent Vivier 已提交
150 151
}

152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183
void ext4_free_blks_set(struct super_block *sb,
			  struct ext4_group_desc *bg, __u32 count)
{
	bg->bg_free_blocks_count_lo = cpu_to_le16((__u16)count);
	if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
		bg->bg_free_blocks_count_hi = cpu_to_le16(count >> 16);
}

void ext4_free_inodes_set(struct super_block *sb,
			  struct ext4_group_desc *bg, __u32 count)
{
	bg->bg_free_inodes_count_lo = cpu_to_le16((__u16)count);
	if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
		bg->bg_free_inodes_count_hi = cpu_to_le16(count >> 16);
}

void ext4_used_dirs_set(struct super_block *sb,
			  struct ext4_group_desc *bg, __u32 count)
{
	bg->bg_used_dirs_count_lo = cpu_to_le16((__u16)count);
	if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
		bg->bg_used_dirs_count_hi = cpu_to_le16(count >> 16);
}

void ext4_itable_unused_set(struct super_block *sb,
			  struct ext4_group_desc *bg, __u32 count)
{
	bg->bg_itable_unused_lo = cpu_to_le16((__u16)count);
	if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
		bg->bg_itable_unused_hi = cpu_to_le16(count >> 16);
}

184
/*
185
 * Wrappers for jbd2_journal_start/end.
186 187 188 189 190 191
 *
 * The only special thing we need to do here is to make sure that all
 * journal_end calls result in the superblock being marked dirty, so
 * that sync() will call the filesystem's write_super callback if
 * appropriate.
 */
192
handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks)
193 194 195 196 197 198 199 200 201
{
	journal_t *journal;

	if (sb->s_flags & MS_RDONLY)
		return ERR_PTR(-EROFS);

	/* Special case here: if the journal has aborted behind our
	 * backs (eg. EIO in the commit thread), then we still need to
	 * take the FS itself readonly cleanly. */
202
	journal = EXT4_SB(sb)->s_journal;
203 204 205 206 207 208 209
	if (journal) {
		if (is_journal_aborted(journal)) {
			ext4_abort(sb, __func__,
				   "Detected aborted journal");
			return ERR_PTR(-EROFS);
		}
		return jbd2_journal_start(journal, nblocks);
210
	}
211 212 213 214 215
	/*
	 * We're not journaling, return the appropriate indication.
	 */
	current->journal_info = EXT4_NOJOURNAL_HANDLE;
	return current->journal_info;
216 217 218 219
}

/*
 * The only special thing we need to do here is to make sure that all
220
 * jbd2_journal_stop calls result in the superblock being marked dirty, so
221 222 223
 * that sync() will call the filesystem's write_super callback if
 * appropriate.
 */
224
int __ext4_journal_stop(const char *where, handle_t *handle)
225 226 227 228 229
{
	struct super_block *sb;
	int err;
	int rc;

230 231 232 233 234 235 236 237
	if (!ext4_handle_valid(handle)) {
		/*
		 * Do this here since we don't call jbd2_journal_stop() in
		 * no-journal mode.
		 */
		current->journal_info = NULL;
		return 0;
	}
238 239
	sb = handle->h_transaction->t_journal->j_private;
	err = handle->h_err;
240
	rc = jbd2_journal_stop(handle);
241 242 243 244

	if (!err)
		err = rc;
	if (err)
245
		__ext4_std_error(sb, where, err);
246 247 248
	return err;
}

249
void ext4_journal_abort_handle(const char *caller, const char *err_fn,
250 251 252
		struct buffer_head *bh, handle_t *handle, int err)
{
	char nbuf[16];
253
	const char *errstr = ext4_decode_error(NULL, err, nbuf);
254

255 256
	BUG_ON(!ext4_handle_valid(handle));

257 258 259 260 261 262 263 264 265 266 267 268
	if (bh)
		BUFFER_TRACE(bh, "abort");

	if (!handle->h_err)
		handle->h_err = err;

	if (is_handle_aborted(handle))
		return;

	printk(KERN_ERR "%s: aborting transaction: %s in %s\n",
	       caller, errstr, err_fn);

269
	jbd2_journal_abort_handle(handle);
270 271 272 273 274 275
}

/* Deal with the reporting of failure conditions on a filesystem such as
 * inconsistencies detected or read IO failures.
 *
 * On ext2, we can store the error state of the filesystem in the
276
 * superblock.  That is not possible on ext4, because we may have other
277 278 279 280 281
 * write ordering constraints on the superblock which prevent us from
 * writing it out straight away; and given that the journal is about to
 * be aborted, we can't rely on the current, or future, transactions to
 * write out the superblock safely.
 *
282
 * We'll just use the jbd2_journal_abort() error code to record an error in
283 284 285 286
 * the journal instead.  On recovery, the journal will compain about
 * that error until we've noted it down and cleared it.
 */

287
static void ext4_handle_error(struct super_block *sb)
288
{
289
	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
290

291 292
	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
	es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
293 294 295 296

	if (sb->s_flags & MS_RDONLY)
		return;

297
	if (!test_opt(sb, ERRORS_CONT)) {
298
		journal_t *journal = EXT4_SB(sb)->s_journal;
299

300
		EXT4_SB(sb)->s_mount_opt |= EXT4_MOUNT_ABORT;
301
		if (journal)
302
			jbd2_journal_abort(journal, -EIO);
303
	}
304 305
	if (test_opt(sb, ERRORS_RO)) {
		printk(KERN_CRIT "Remounting filesystem read-only\n");
306 307
		sb->s_flags |= MS_RDONLY;
	}
308
	ext4_commit_super(sb, es, 1);
309
	if (test_opt(sb, ERRORS_PANIC))
310
		panic("EXT4-fs (device %s): panic forced after error\n",
311 312 313
			sb->s_id);
}

314 315
void ext4_error(struct super_block *sb, const char *function,
		const char *fmt, ...)
316 317 318 319
{
	va_list args;

	va_start(args, fmt);
320
	printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function);
321 322 323 324
	vprintk(fmt, args);
	printk("\n");
	va_end(args);

325
	ext4_handle_error(sb);
326 327
}

328
static const char *ext4_decode_error(struct super_block *sb, int errno,
329 330 331 332 333 334 335 336 337 338 339 340
				     char nbuf[16])
{
	char *errstr = NULL;

	switch (errno) {
	case -EIO:
		errstr = "IO failure";
		break;
	case -ENOMEM:
		errstr = "Out of memory";
		break;
	case -EROFS:
341
		if (!sb || EXT4_SB(sb)->s_journal->j_flags & JBD2_ABORT)
342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360
			errstr = "Journal has aborted";
		else
			errstr = "Readonly filesystem";
		break;
	default:
		/* If the caller passed in an extra buffer for unknown
		 * errors, textualise them now.  Else we just return
		 * NULL. */
		if (nbuf) {
			/* Check for truncated error codes... */
			if (snprintf(nbuf, 16, "error %d", -errno) >= 0)
				errstr = nbuf;
		}
		break;
	}

	return errstr;
}

361
/* __ext4_std_error decodes expected errors from journaling functions
362 363
 * automatically and invokes the appropriate error response.  */

364
void __ext4_std_error(struct super_block *sb, const char *function, int errno)
365 366 367 368 369 370 371 372 373 374 375
{
	char nbuf[16];
	const char *errstr;

	/* Special case: if the error is EROFS, and we're not already
	 * inside a transaction, then there's really no point in logging
	 * an error. */
	if (errno == -EROFS && journal_current_handle() == NULL &&
	    (sb->s_flags & MS_RDONLY))
		return;

376
	errstr = ext4_decode_error(sb, errno, nbuf);
377 378
	printk(KERN_CRIT "EXT4-fs error (device %s) in %s: %s\n",
	       sb->s_id, function, errstr);
379

380
	ext4_handle_error(sb);
381 382 383
}

/*
384
 * ext4_abort is a much stronger failure handler than ext4_error.  The
385 386 387 388 389 390 391 392
 * abort function may be used to deal with unrecoverable failures such
 * as journal IO errors or ENOMEM at a critical moment in log management.
 *
 * We unconditionally force the filesystem into an ABORT|READONLY state,
 * unless the error response on the fs has been set to panic in which
 * case we take the easy way out and panic immediately.
 */

393 394
void ext4_abort(struct super_block *sb, const char *function,
		const char *fmt, ...)
395 396 397
{
	va_list args;

398
	printk(KERN_CRIT "ext4_abort called.\n");
399 400

	va_start(args, fmt);
401
	printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function);
402 403 404 405 406
	vprintk(fmt, args);
	printk("\n");
	va_end(args);

	if (test_opt(sb, ERRORS_PANIC))
407
		panic("EXT4-fs panic from previous error\n");
408 409 410 411 412

	if (sb->s_flags & MS_RDONLY)
		return;

	printk(KERN_CRIT "Remounting filesystem read-only\n");
413
	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
414
	sb->s_flags |= MS_RDONLY;
415
	EXT4_SB(sb)->s_mount_opt |= EXT4_MOUNT_ABORT;
416 417
	if (EXT4_SB(sb)->s_journal)
		jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO);
418 419
}

420 421
void ext4_warning(struct super_block *sb, const char *function,
		  const char *fmt, ...)
422 423 424 425
{
	va_list args;

	va_start(args, fmt);
426
	printk(KERN_WARNING "EXT4-fs warning (device %s): %s: ",
427 428 429 430 431 432
	       sb->s_id, function);
	vprintk(fmt, args);
	printk("\n");
	va_end(args);
}

433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470
void ext4_grp_locked_error(struct super_block *sb, ext4_group_t grp,
				const char *function, const char *fmt, ...)
__releases(bitlock)
__acquires(bitlock)
{
	va_list args;
	struct ext4_super_block *es = EXT4_SB(sb)->s_es;

	va_start(args, fmt);
	printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function);
	vprintk(fmt, args);
	printk("\n");
	va_end(args);

	if (test_opt(sb, ERRORS_CONT)) {
		EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
		es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
		ext4_commit_super(sb, es, 0);
		return;
	}
	ext4_unlock_group(sb, grp);
	ext4_handle_error(sb);
	/*
	 * We only get here in the ERRORS_RO case; relocking the group
	 * may be dangerous, but nothing bad will happen since the
	 * filesystem will have already been marked read/only and the
	 * journal has been aborted.  We return 1 as a hint to callers
	 * who might what to use the return value from
	 * ext4_grp_locked_error() to distinguish beween the
	 * ERRORS_CONT and ERRORS_RO case, and perhaps return more
	 * aggressively from the ext4 function in question, with a
	 * more appropriate error code.
	 */
	ext4_lock_group(sb, grp);
	return;
}


471
void ext4_update_dynamic_rev(struct super_block *sb)
472
{
473
	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
474

475
	if (le32_to_cpu(es->s_rev_level) > EXT4_GOOD_OLD_REV)
476 477
		return;

478
	ext4_warning(sb, __func__,
479 480
		     "updating to rev %d because of new feature flag, "
		     "running e2fsck is recommended",
481
		     EXT4_DYNAMIC_REV);
482

483 484 485
	es->s_first_ino = cpu_to_le32(EXT4_GOOD_OLD_FIRST_INO);
	es->s_inode_size = cpu_to_le16(EXT4_GOOD_OLD_INODE_SIZE);
	es->s_rev_level = cpu_to_le32(EXT4_DYNAMIC_REV);
486 487 488 489 490 491 492 493 494 495 496 497 498
	/* leave es->s_feature_*compat flags alone */
	/* es->s_uuid will be set by e2fsck if empty */

	/*
	 * The rest of the superblock fields should be zero, and if not it
	 * means they are likely already in use, so leave them alone.  We
	 * can leave it up to e2fsck to clean up any inconsistencies there.
	 */
}

/*
 * Open the external journal device
 */
499
static struct block_device *ext4_blkdev_get(dev_t dev)
500 501 502 503 504 505 506 507 508 509
{
	struct block_device *bdev;
	char b[BDEVNAME_SIZE];

	bdev = open_by_devnum(dev, FMODE_READ|FMODE_WRITE);
	if (IS_ERR(bdev))
		goto fail;
	return bdev;

fail:
510
	printk(KERN_ERR "EXT4-fs: failed to open journal device %s: %ld\n",
511 512 513 514 515 516 517
			__bdevname(dev, b), PTR_ERR(bdev));
	return NULL;
}

/*
 * Release the journal device
 */
518
static int ext4_blkdev_put(struct block_device *bdev)
519 520
{
	bd_release(bdev);
A
Al Viro 已提交
521
	return blkdev_put(bdev, FMODE_READ|FMODE_WRITE);
522 523
}

524
static int ext4_blkdev_remove(struct ext4_sb_info *sbi)
525 526 527 528 529 530
{
	struct block_device *bdev;
	int ret = -ENODEV;

	bdev = sbi->journal_bdev;
	if (bdev) {
531
		ret = ext4_blkdev_put(bdev);
532 533 534 535 536 537 538
		sbi->journal_bdev = NULL;
	}
	return ret;
}

static inline struct inode *orphan_list_entry(struct list_head *l)
{
539
	return &list_entry(l, struct ext4_inode_info, i_orphan)->vfs_inode;
540 541
}

542
static void dump_orphan_list(struct super_block *sb, struct ext4_sb_info *sbi)
543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
{
	struct list_head *l;

	printk(KERN_ERR "sb orphan head is %d\n",
	       le32_to_cpu(sbi->s_es->s_last_orphan));

	printk(KERN_ERR "sb_info orphan list:\n");
	list_for_each(l, &sbi->s_orphan) {
		struct inode *inode = orphan_list_entry(l);
		printk(KERN_ERR "  "
		       "inode %s:%lu at %p: mode %o, nlink %d, next %d\n",
		       inode->i_sb->s_id, inode->i_ino, inode,
		       inode->i_mode, inode->i_nlink,
		       NEXT_ORPHAN(inode));
	}
}

560
static void ext4_put_super(struct super_block *sb)
561
{
562 563
	struct ext4_sb_info *sbi = EXT4_SB(sb);
	struct ext4_super_block *es = sbi->s_es;
564
	int i, err;
565

566
	ext4_mb_release(sb);
A
Alex Tomas 已提交
567
	ext4_ext_release(sb);
568
	ext4_xattr_put_super(sb);
569 570 571 572 573 574 575
	if (sbi->s_journal) {
		err = jbd2_journal_destroy(sbi->s_journal);
		sbi->s_journal = NULL;
		if (err < 0)
			ext4_abort(sb, __func__,
				   "Couldn't clean up the journal");
	}
576
	if (!(sb->s_flags & MS_RDONLY)) {
577
		EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
578
		es->s_state = cpu_to_le16(sbi->s_mount_state);
579
		ext4_commit_super(sb, es, 1);
580
	}
581
	if (sbi->s_proc) {
582
		remove_proc_entry(sb->s_id, ext4_proc_root);
583
	}
T
Theodore Ts'o 已提交
584
	kobject_del(&sbi->s_kobj);
585 586 587 588

	for (i = 0; i < sbi->s_gdb_count; i++)
		brelse(sbi->s_group_desc[i]);
	kfree(sbi->s_group_desc);
589
	kfree(sbi->s_flex_groups);
590 591 592
	percpu_counter_destroy(&sbi->s_freeblocks_counter);
	percpu_counter_destroy(&sbi->s_freeinodes_counter);
	percpu_counter_destroy(&sbi->s_dirs_counter);
593
	percpu_counter_destroy(&sbi->s_dirtyblocks_counter);
594 595 596 597 598 599 600 601 602 603 604 605 606 607
	brelse(sbi->s_sbh);
#ifdef CONFIG_QUOTA
	for (i = 0; i < MAXQUOTAS; i++)
		kfree(sbi->s_qf_names[i]);
#endif

	/* Debugging code just in case the in-memory inode orphan list
	 * isn't empty.  The on-disk one can be non-empty if we've
	 * detected an error and taken the fs readonly, but the
	 * in-memory list had better be clean by this point. */
	if (!list_empty(&sbi->s_orphan))
		dump_orphan_list(sb, sbi);
	J_ASSERT(list_empty(&sbi->s_orphan));

608
	invalidate_bdev(sb->s_bdev);
609 610 611 612 613 614 615
	if (sbi->journal_bdev && sbi->journal_bdev != sb->s_bdev) {
		/*
		 * Invalidate the journal device's buffers.  We don't want them
		 * floating about in memory - the physical journal device may
		 * hotswapped, and it breaks the `ro-after' testing code.
		 */
		sync_blockdev(sbi->journal_bdev);
616
		invalidate_bdev(sbi->journal_bdev);
617
		ext4_blkdev_remove(sbi);
618 619
	}
	sb->s_fs_info = NULL;
T
Theodore Ts'o 已提交
620 621 622 623 624 625 626 627 628 629
	/*
	 * Now that we are completely done shutting down the
	 * superblock, we need to actually destroy the kobject.
	 */
	unlock_kernel();
	unlock_super(sb);
	kobject_put(&sbi->s_kobj);
	wait_for_completion(&sbi->s_kobj_unregister);
	lock_super(sb);
	lock_kernel();
630
	kfree(sbi->s_blockgroup_lock);
631 632 633 634
	kfree(sbi);
	return;
}

635
static struct kmem_cache *ext4_inode_cachep;
636 637 638 639

/*
 * Called inside transaction, so use GFP_NOFS
 */
640
static struct inode *ext4_alloc_inode(struct super_block *sb)
641
{
642
	struct ext4_inode_info *ei;
643

C
Christoph Lameter 已提交
644
	ei = kmem_cache_alloc(ext4_inode_cachep, GFP_NOFS);
645 646
	if (!ei)
		return NULL;
T
Theodore Ts'o 已提交
647
#ifdef CONFIG_EXT4_FS_POSIX_ACL
648 649
	ei->i_acl = EXT4_ACL_NOT_CACHED;
	ei->i_default_acl = EXT4_ACL_NOT_CACHED;
650 651
#endif
	ei->vfs_inode.i_version = 1;
652
	ei->vfs_inode.i_data.writeback_index = 0;
A
Alex Tomas 已提交
653
	memset(&ei->i_cached_extent, 0, sizeof(struct ext4_ext_cache));
654 655
	INIT_LIST_HEAD(&ei->i_prealloc_list);
	spin_lock_init(&ei->i_prealloc_lock);
656 657 658 659 660
	/*
	 * Note:  We can be called before EXT4_SB(sb)->s_journal is set,
	 * therefore it can be null here.  Don't check it, just initialize
	 * jinode.
	 */
661
	jbd2_journal_init_jbd_inode(&ei->jinode, &ei->vfs_inode);
662 663 664 665 666
	ei->i_reserved_data_blocks = 0;
	ei->i_reserved_meta_blocks = 0;
	ei->i_allocated_meta_blocks = 0;
	ei->i_delalloc_reserved_flag = 0;
	spin_lock_init(&(ei->i_block_reservation_lock));
667 668 669
	return &ei->vfs_inode;
}

670
static void ext4_destroy_inode(struct inode *inode)
671
{
672 673 674 675 676 677 678 679
	if (!list_empty(&(EXT4_I(inode)->i_orphan))) {
		printk("EXT4 Inode %p: orphan list check failed!\n",
			EXT4_I(inode));
		print_hex_dump(KERN_INFO, "", DUMP_PREFIX_ADDRESS, 16, 4,
				EXT4_I(inode), sizeof(struct ext4_inode_info),
				true);
		dump_stack();
	}
680
	kmem_cache_free(ext4_inode_cachep, EXT4_I(inode));
681 682
}

683
static void init_once(void *foo)
684
{
685
	struct ext4_inode_info *ei = (struct ext4_inode_info *) foo;
686

C
Christoph Lameter 已提交
687
	INIT_LIST_HEAD(&ei->i_orphan);
T
Theodore Ts'o 已提交
688
#ifdef CONFIG_EXT4_FS_XATTR
C
Christoph Lameter 已提交
689
	init_rwsem(&ei->xattr_sem);
690
#endif
691
	init_rwsem(&ei->i_data_sem);
C
Christoph Lameter 已提交
692
	inode_init_once(&ei->vfs_inode);
693 694 695 696
}

static int init_inodecache(void)
{
697 698
	ext4_inode_cachep = kmem_cache_create("ext4_inode_cache",
					     sizeof(struct ext4_inode_info),
699 700
					     0, (SLAB_RECLAIM_ACCOUNT|
						SLAB_MEM_SPREAD),
701
					     init_once);
702
	if (ext4_inode_cachep == NULL)
703 704 705 706 707 708
		return -ENOMEM;
	return 0;
}

static void destroy_inodecache(void)
{
709
	kmem_cache_destroy(ext4_inode_cachep);
710 711
}

712
static void ext4_clear_inode(struct inode *inode)
713
{
T
Theodore Ts'o 已提交
714
#ifdef CONFIG_EXT4_FS_POSIX_ACL
715 716 717 718 719 720 721 722 723
	if (EXT4_I(inode)->i_acl &&
			EXT4_I(inode)->i_acl != EXT4_ACL_NOT_CACHED) {
		posix_acl_release(EXT4_I(inode)->i_acl);
		EXT4_I(inode)->i_acl = EXT4_ACL_NOT_CACHED;
	}
	if (EXT4_I(inode)->i_default_acl &&
			EXT4_I(inode)->i_default_acl != EXT4_ACL_NOT_CACHED) {
		posix_acl_release(EXT4_I(inode)->i_default_acl);
		EXT4_I(inode)->i_default_acl = EXT4_ACL_NOT_CACHED;
724 725
	}
#endif
726
	ext4_discard_preallocations(inode);
727 728
	if (EXT4_JOURNAL(inode))
		jbd2_journal_release_jbd_inode(EXT4_SB(inode->i_sb)->s_journal,
729
				       &EXT4_I(inode)->jinode);
730 731
}

732 733
static inline void ext4_show_quota_options(struct seq_file *seq,
					   struct super_block *sb)
734 735
{
#if defined(CONFIG_QUOTA)
736
	struct ext4_sb_info *sbi = EXT4_SB(sb);
737 738 739

	if (sbi->s_jquota_fmt)
		seq_printf(seq, ",jqfmt=%s",
740
		(sbi->s_jquota_fmt == QFMT_VFS_OLD) ? "vfsold" : "vfsv0");
741 742 743 744 745 746 747

	if (sbi->s_qf_names[USRQUOTA])
		seq_printf(seq, ",usrjquota=%s", sbi->s_qf_names[USRQUOTA]);

	if (sbi->s_qf_names[GRPQUOTA])
		seq_printf(seq, ",grpjquota=%s", sbi->s_qf_names[GRPQUOTA]);

748
	if (sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA)
749 750
		seq_puts(seq, ",usrquota");

751
	if (sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA)
752 753 754 755
		seq_puts(seq, ",grpquota");
#endif
}

M
Miklos Szeredi 已提交
756 757 758 759 760
/*
 * Show an option if
 *  - it's set to a non-default value OR
 *  - if the per-sb default is different from the global default
 */
761
static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
762
{
763 764
	int def_errors;
	unsigned long def_mount_opts;
765
	struct super_block *sb = vfs->mnt_sb;
M
Miklos Szeredi 已提交
766 767 768 769
	struct ext4_sb_info *sbi = EXT4_SB(sb);
	struct ext4_super_block *es = sbi->s_es;

	def_mount_opts = le32_to_cpu(es->s_default_mount_opts);
770
	def_errors     = le16_to_cpu(es->s_errors);
M
Miklos Szeredi 已提交
771 772 773 774 775

	if (sbi->s_sb_block != 1)
		seq_printf(seq, ",sb=%llu", sbi->s_sb_block);
	if (test_opt(sb, MINIX_DF))
		seq_puts(seq, ",minixdf");
776
	if (test_opt(sb, GRPID) && !(def_mount_opts & EXT4_DEFM_BSDGROUPS))
M
Miklos Szeredi 已提交
777 778 779 780 781 782 783 784 785 786 787
		seq_puts(seq, ",grpid");
	if (!test_opt(sb, GRPID) && (def_mount_opts & EXT4_DEFM_BSDGROUPS))
		seq_puts(seq, ",nogrpid");
	if (sbi->s_resuid != EXT4_DEF_RESUID ||
	    le16_to_cpu(es->s_def_resuid) != EXT4_DEF_RESUID) {
		seq_printf(seq, ",resuid=%u", sbi->s_resuid);
	}
	if (sbi->s_resgid != EXT4_DEF_RESGID ||
	    le16_to_cpu(es->s_def_resgid) != EXT4_DEF_RESGID) {
		seq_printf(seq, ",resgid=%u", sbi->s_resgid);
	}
788
	if (test_opt(sb, ERRORS_RO)) {
M
Miklos Szeredi 已提交
789
		if (def_errors == EXT4_ERRORS_PANIC ||
790 791
		    def_errors == EXT4_ERRORS_CONTINUE) {
			seq_puts(seq, ",errors=remount-ro");
M
Miklos Szeredi 已提交
792 793
		}
	}
794
	if (test_opt(sb, ERRORS_CONT) && def_errors != EXT4_ERRORS_CONTINUE)
795
		seq_puts(seq, ",errors=continue");
796
	if (test_opt(sb, ERRORS_PANIC) && def_errors != EXT4_ERRORS_PANIC)
M
Miklos Szeredi 已提交
797
		seq_puts(seq, ",errors=panic");
798
	if (test_opt(sb, NO_UID32) && !(def_mount_opts & EXT4_DEFM_UID16))
M
Miklos Szeredi 已提交
799
		seq_puts(seq, ",nouid32");
800
	if (test_opt(sb, DEBUG) && !(def_mount_opts & EXT4_DEFM_DEBUG))
M
Miklos Szeredi 已提交
801 802 803
		seq_puts(seq, ",debug");
	if (test_opt(sb, OLDALLOC))
		seq_puts(seq, ",oldalloc");
T
Theodore Ts'o 已提交
804
#ifdef CONFIG_EXT4_FS_XATTR
805 806
	if (test_opt(sb, XATTR_USER) &&
		!(def_mount_opts & EXT4_DEFM_XATTR_USER))
M
Miklos Szeredi 已提交
807 808 809 810 811 812
		seq_puts(seq, ",user_xattr");
	if (!test_opt(sb, XATTR_USER) &&
	    (def_mount_opts & EXT4_DEFM_XATTR_USER)) {
		seq_puts(seq, ",nouser_xattr");
	}
#endif
T
Theodore Ts'o 已提交
813
#ifdef CONFIG_EXT4_FS_POSIX_ACL
814
	if (test_opt(sb, POSIX_ACL) && !(def_mount_opts & EXT4_DEFM_ACL))
M
Miklos Szeredi 已提交
815 816 817 818 819 820
		seq_puts(seq, ",acl");
	if (!test_opt(sb, POSIX_ACL) && (def_mount_opts & EXT4_DEFM_ACL))
		seq_puts(seq, ",noacl");
#endif
	if (!test_opt(sb, RESERVATION))
		seq_puts(seq, ",noreservation");
821
	if (sbi->s_commit_interval != JBD2_DEFAULT_MAX_COMMIT_AGE*HZ) {
M
Miklos Szeredi 已提交
822 823 824
		seq_printf(seq, ",commit=%u",
			   (unsigned) (sbi->s_commit_interval / HZ));
	}
825 826 827 828 829 830 831 832 833
	if (sbi->s_min_batch_time != EXT4_DEF_MIN_BATCH_TIME) {
		seq_printf(seq, ",min_batch_time=%u",
			   (unsigned) sbi->s_min_batch_time);
	}
	if (sbi->s_max_batch_time != EXT4_DEF_MAX_BATCH_TIME) {
		seq_printf(seq, ",max_batch_time=%u",
			   (unsigned) sbi->s_min_batch_time);
	}

E
Eric Sandeen 已提交
834 835 836 837 838 839 840
	/*
	 * We're changing the default of barrier mount option, so
	 * let's always display its mount state so it's clear what its
	 * status is.
	 */
	seq_puts(seq, ",barrier=");
	seq_puts(seq, test_opt(sb, BARRIER) ? "1" : "0");
841 842
	if (test_opt(sb, JOURNAL_ASYNC_COMMIT))
		seq_puts(seq, ",journal_async_commit");
M
Miklos Szeredi 已提交
843 844
	if (test_opt(sb, NOBH))
		seq_puts(seq, ",nobh");
845 846
	if (test_opt(sb, I_VERSION))
		seq_puts(seq, ",i_version");
847 848 849
	if (!test_opt(sb, DELALLOC))
		seq_puts(seq, ",nodelalloc");

850

851 852
	if (sbi->s_stripe)
		seq_printf(seq, ",stripe=%lu", sbi->s_stripe);
853 854 855 856
	/*
	 * journal mode get enabled in different ways
	 * So just print the value even if we didn't specify it
	 */
857
	if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA)
858
		seq_puts(seq, ",data=journal");
859
	else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA)
860
		seq_puts(seq, ",data=ordered");
861
	else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)
862 863
		seq_puts(seq, ",data=writeback");

864 865 866 867
	if (sbi->s_inode_readahead_blks != EXT4_DEF_INODE_READAHEAD_BLKS)
		seq_printf(seq, ",inode_readahead_blks=%u",
			   sbi->s_inode_readahead_blks);

868 869 870
	if (test_opt(sb, DATA_ERR_ABORT))
		seq_puts(seq, ",data_err=abort");

871
	ext4_show_quota_options(seq, sb);
872 873 874 875
	return 0;
}


C
Christoph Hellwig 已提交
876 877
static struct inode *ext4_nfs_get_inode(struct super_block *sb,
		u64 ino, u32 generation)
878 879 880
{
	struct inode *inode;

881
	if (ino < EXT4_FIRST_INO(sb) && ino != EXT4_ROOT_INO)
882
		return ERR_PTR(-ESTALE);
883
	if (ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))
884 885 886 887
		return ERR_PTR(-ESTALE);

	/* iget isn't really right if the inode is currently unallocated!!
	 *
888
	 * ext4_read_inode will return a bad_inode if the inode had been
889 890 891 892 893
	 * deleted, so we should be safe.
	 *
	 * Currently we don't know the generation for parent directory, so
	 * a generation of 0 means "accept any"
	 */
894 895 896 897
	inode = ext4_iget(sb, ino);
	if (IS_ERR(inode))
		return ERR_CAST(inode);
	if (generation && inode->i_generation != generation) {
898 899 900
		iput(inode);
		return ERR_PTR(-ESTALE);
	}
C
Christoph Hellwig 已提交
901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916

	return inode;
}

static struct dentry *ext4_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,
				    ext4_nfs_get_inode);
}

static struct dentry *ext4_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,
				    ext4_nfs_get_inode);
917 918
}

919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937
/*
 * Try to release metadata pages (indirect blocks, directories) which are
 * mapped via the block device.  Since these pages could have journal heads
 * which would prevent try_to_free_buffers() from freeing them, we must use
 * jbd2 layer's try_to_free_buffers() function to release them.
 */
static int bdev_try_to_free_page(struct super_block *sb, struct page *page, gfp_t wait)
{
	journal_t *journal = EXT4_SB(sb)->s_journal;

	WARN_ON(PageChecked(page));
	if (!page_has_buffers(page))
		return 0;
	if (journal)
		return jbd2_journal_try_to_free_buffers(journal, page,
							wait & ~__GFP_WAIT);
	return try_to_free_buffers(page);
}

938
#ifdef CONFIG_QUOTA
939
#define QTYPE2NAME(t) ((t) == USRQUOTA ? "user" : "group")
940
#define QTYPE2MOPT(on, t) ((t) == USRQUOTA?((on)##USRJQUOTA):((on)##GRPJQUOTA))
941

942 943 944 945 946
static int ext4_write_dquot(struct dquot *dquot);
static int ext4_acquire_dquot(struct dquot *dquot);
static int ext4_release_dquot(struct dquot *dquot);
static int ext4_mark_dquot_dirty(struct dquot *dquot);
static int ext4_write_info(struct super_block *sb, int type);
947 948
static int ext4_quota_on(struct super_block *sb, int type, int format_id,
				char *path, int remount);
949 950
static int ext4_quota_on_mount(struct super_block *sb, int type);
static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data,
951
			       size_t len, loff_t off);
952
static ssize_t ext4_quota_write(struct super_block *sb, int type,
953 954
				const char *data, size_t len, loff_t off);

955
static struct dquot_operations ext4_quota_operations = {
956 957
	.initialize	= dquot_initialize,
	.drop		= dquot_drop,
958
	.alloc_space	= dquot_alloc_space,
959 960 961 962
	.reserve_space	= dquot_reserve_space,
	.claim_space	= dquot_claim_space,
	.release_rsv	= dquot_release_reserved_space,
	.get_reserved_space = ext4_get_reserved_space,
963 964 965 966
	.alloc_inode	= dquot_alloc_inode,
	.free_space	= dquot_free_space,
	.free_inode	= dquot_free_inode,
	.transfer	= dquot_transfer,
967 968 969 970
	.write_dquot	= ext4_write_dquot,
	.acquire_dquot	= ext4_acquire_dquot,
	.release_dquot	= ext4_release_dquot,
	.mark_dirty	= ext4_mark_dquot_dirty,
971 972 973
	.write_info	= ext4_write_info,
	.alloc_dquot	= dquot_alloc,
	.destroy_dquot	= dquot_destroy,
974 975
};

976 977
static struct quotactl_ops ext4_qctl_operations = {
	.quota_on	= ext4_quota_on,
978 979 980 981 982 983 984 985 986
	.quota_off	= vfs_quota_off,
	.quota_sync	= vfs_quota_sync,
	.get_info	= vfs_get_dqinfo,
	.set_info	= vfs_set_dqinfo,
	.get_dqblk	= vfs_get_dqblk,
	.set_dqblk	= vfs_set_dqblk
};
#endif

987
static const struct super_operations ext4_sops = {
988 989 990 991 992 993 994 995
	.alloc_inode	= ext4_alloc_inode,
	.destroy_inode	= ext4_destroy_inode,
	.write_inode	= ext4_write_inode,
	.dirty_inode	= ext4_dirty_inode,
	.delete_inode	= ext4_delete_inode,
	.put_super	= ext4_put_super,
	.write_super	= ext4_write_super,
	.sync_fs	= ext4_sync_fs,
996 997
	.freeze_fs	= ext4_freeze,
	.unfreeze_fs	= ext4_unfreeze,
998 999 1000 1001
	.statfs		= ext4_statfs,
	.remount_fs	= ext4_remount,
	.clear_inode	= ext4_clear_inode,
	.show_options	= ext4_show_options,
1002
#ifdef CONFIG_QUOTA
1003 1004
	.quota_read	= ext4_quota_read,
	.quota_write	= ext4_quota_write,
1005
#endif
1006
	.bdev_try_to_free_page = bdev_try_to_free_page,
1007 1008
};

1009
static const struct export_operations ext4_export_ops = {
C
Christoph Hellwig 已提交
1010 1011
	.fh_to_dentry = ext4_fh_to_dentry,
	.fh_to_parent = ext4_fh_to_parent,
1012
	.get_parent = ext4_get_parent,
1013 1014 1015 1016 1017
};

enum {
	Opt_bsd_df, Opt_minix_df, Opt_grpid, Opt_nogrpid,
	Opt_resgid, Opt_resuid, Opt_sb, Opt_err_cont, Opt_err_panic, Opt_err_ro,
1018
	Opt_nouid32, Opt_debug, Opt_oldalloc, Opt_orlov,
1019 1020
	Opt_user_xattr, Opt_nouser_xattr, Opt_acl, Opt_noacl,
	Opt_reservation, Opt_noreservation, Opt_noload, Opt_nobh, Opt_bh,
1021
	Opt_commit, Opt_min_batch_time, Opt_max_batch_time,
1022
	Opt_journal_update, Opt_journal_dev,
1023
	Opt_journal_checksum, Opt_journal_async_commit,
1024
	Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback,
1025
	Opt_data_err_abort, Opt_data_err_ignore,
1026 1027 1028
	Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota,
	Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota,
	Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota,
1029
	Opt_grpquota, Opt_i_version,
1030
	Opt_stripe, Opt_delalloc, Opt_nodelalloc,
1031
	Opt_inode_readahead_blks, Opt_journal_ioprio
1032 1033
};

1034
static const match_table_t tokens = {
1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060
	{Opt_bsd_df, "bsddf"},
	{Opt_minix_df, "minixdf"},
	{Opt_grpid, "grpid"},
	{Opt_grpid, "bsdgroups"},
	{Opt_nogrpid, "nogrpid"},
	{Opt_nogrpid, "sysvgroups"},
	{Opt_resgid, "resgid=%u"},
	{Opt_resuid, "resuid=%u"},
	{Opt_sb, "sb=%u"},
	{Opt_err_cont, "errors=continue"},
	{Opt_err_panic, "errors=panic"},
	{Opt_err_ro, "errors=remount-ro"},
	{Opt_nouid32, "nouid32"},
	{Opt_debug, "debug"},
	{Opt_oldalloc, "oldalloc"},
	{Opt_orlov, "orlov"},
	{Opt_user_xattr, "user_xattr"},
	{Opt_nouser_xattr, "nouser_xattr"},
	{Opt_acl, "acl"},
	{Opt_noacl, "noacl"},
	{Opt_reservation, "reservation"},
	{Opt_noreservation, "noreservation"},
	{Opt_noload, "noload"},
	{Opt_nobh, "nobh"},
	{Opt_bh, "bh"},
	{Opt_commit, "commit=%u"},
1061 1062
	{Opt_min_batch_time, "min_batch_time=%u"},
	{Opt_max_batch_time, "max_batch_time=%u"},
1063 1064
	{Opt_journal_update, "journal=update"},
	{Opt_journal_dev, "journal_dev=%u"},
1065 1066
	{Opt_journal_checksum, "journal_checksum"},
	{Opt_journal_async_commit, "journal_async_commit"},
1067 1068 1069 1070
	{Opt_abort, "abort"},
	{Opt_data_journal, "data=journal"},
	{Opt_data_ordered, "data=ordered"},
	{Opt_data_writeback, "data=writeback"},
1071 1072
	{Opt_data_err_abort, "data_err=abort"},
	{Opt_data_err_ignore, "data_err=ignore"},
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083
	{Opt_offusrjquota, "usrjquota="},
	{Opt_usrjquota, "usrjquota=%s"},
	{Opt_offgrpjquota, "grpjquota="},
	{Opt_grpjquota, "grpjquota=%s"},
	{Opt_jqfmt_vfsold, "jqfmt=vfsold"},
	{Opt_jqfmt_vfsv0, "jqfmt=vfsv0"},
	{Opt_grpquota, "grpquota"},
	{Opt_noquota, "noquota"},
	{Opt_quota, "quota"},
	{Opt_usrquota, "usrquota"},
	{Opt_barrier, "barrier=%u"},
1084
	{Opt_i_version, "i_version"},
1085
	{Opt_stripe, "stripe=%u"},
1086
	{Opt_resize, "resize"},
1087
	{Opt_delalloc, "delalloc"},
1088
	{Opt_nodelalloc, "nodelalloc"},
1089
	{Opt_inode_readahead_blks, "inode_readahead_blks=%u"},
1090
	{Opt_journal_ioprio, "journal_ioprio=%u"},
J
Josef Bacik 已提交
1091
	{Opt_err, NULL},
1092 1093
};

1094
static ext4_fsblk_t get_sb_block(void **data)
1095
{
1096
	ext4_fsblk_t	sb_block;
1097 1098 1099 1100 1101
	char		*options = (char *) *data;

	if (!options || strncmp(options, "sb=", 3) != 0)
		return 1;	/* Default location */
	options += 3;
1102
	/*todo: use simple_strtoll with >32bit ext4 */
1103 1104
	sb_block = simple_strtoul(options, &options, 0);
	if (*options && *options != ',') {
1105
		printk(KERN_ERR "EXT4-fs: Invalid sb specification: %s\n",
1106 1107 1108 1109 1110 1111 1112 1113 1114
		       (char *) *data);
		return 1;
	}
	if (*options == ',')
		options++;
	*data = (void *) options;
	return sb_block;
}

1115 1116
#define DEFAULT_JOURNAL_IOPRIO (IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, 3))

1117
static int parse_options(char *options, struct super_block *sb,
1118
			 unsigned long *journal_devnum,
1119
			 unsigned int *journal_ioprio,
1120
			 ext4_fsblk_t *n_blocks_count, int is_remount)
1121
{
1122
	struct ext4_sb_info *sbi = EXT4_SB(sb);
1123
	char *p;
1124 1125 1126 1127
	substring_t args[MAX_OPT_ARGS];
	int data_opt = 0;
	int option;
#ifdef CONFIG_QUOTA
1128
	int qtype, qfmt;
1129 1130 1131 1132 1133 1134
	char *qname;
#endif

	if (!options)
		return 1;

1135
	while ((p = strsep(&options, ",")) != NULL) {
1136 1137 1138 1139 1140 1141 1142
		int token;
		if (!*p)
			continue;

		token = match_token(p, tokens, args);
		switch (token) {
		case Opt_bsd_df:
1143
			clear_opt(sbi->s_mount_opt, MINIX_DF);
1144 1145
			break;
		case Opt_minix_df:
1146
			set_opt(sbi->s_mount_opt, MINIX_DF);
1147 1148
			break;
		case Opt_grpid:
1149
			set_opt(sbi->s_mount_opt, GRPID);
1150 1151
			break;
		case Opt_nogrpid:
1152
			clear_opt(sbi->s_mount_opt, GRPID);
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
			break;
		case Opt_resuid:
			if (match_int(&args[0], &option))
				return 0;
			sbi->s_resuid = option;
			break;
		case Opt_resgid:
			if (match_int(&args[0], &option))
				return 0;
			sbi->s_resgid = option;
			break;
		case Opt_sb:
			/* handled by get_sb_block() instead of here */
			/* *sb_block = match_int(&args[0]); */
			break;
		case Opt_err_panic:
1169 1170 1171
			clear_opt(sbi->s_mount_opt, ERRORS_CONT);
			clear_opt(sbi->s_mount_opt, ERRORS_RO);
			set_opt(sbi->s_mount_opt, ERRORS_PANIC);
1172 1173
			break;
		case Opt_err_ro:
1174 1175 1176
			clear_opt(sbi->s_mount_opt, ERRORS_CONT);
			clear_opt(sbi->s_mount_opt, ERRORS_PANIC);
			set_opt(sbi->s_mount_opt, ERRORS_RO);
1177 1178
			break;
		case Opt_err_cont:
1179 1180 1181
			clear_opt(sbi->s_mount_opt, ERRORS_RO);
			clear_opt(sbi->s_mount_opt, ERRORS_PANIC);
			set_opt(sbi->s_mount_opt, ERRORS_CONT);
1182 1183
			break;
		case Opt_nouid32:
1184
			set_opt(sbi->s_mount_opt, NO_UID32);
1185 1186
			break;
		case Opt_debug:
1187
			set_opt(sbi->s_mount_opt, DEBUG);
1188 1189
			break;
		case Opt_oldalloc:
1190
			set_opt(sbi->s_mount_opt, OLDALLOC);
1191 1192
			break;
		case Opt_orlov:
1193
			clear_opt(sbi->s_mount_opt, OLDALLOC);
1194
			break;
T
Theodore Ts'o 已提交
1195
#ifdef CONFIG_EXT4_FS_XATTR
1196
		case Opt_user_xattr:
1197
			set_opt(sbi->s_mount_opt, XATTR_USER);
1198 1199
			break;
		case Opt_nouser_xattr:
1200
			clear_opt(sbi->s_mount_opt, XATTR_USER);
1201 1202 1203 1204
			break;
#else
		case Opt_user_xattr:
		case Opt_nouser_xattr:
1205 1206
			printk(KERN_ERR "EXT4 (no)user_xattr options "
			       "not supported\n");
1207 1208
			break;
#endif
T
Theodore Ts'o 已提交
1209
#ifdef CONFIG_EXT4_FS_POSIX_ACL
1210 1211 1212 1213 1214 1215 1216 1217 1218
		case Opt_acl:
			set_opt(sbi->s_mount_opt, POSIX_ACL);
			break;
		case Opt_noacl:
			clear_opt(sbi->s_mount_opt, POSIX_ACL);
			break;
#else
		case Opt_acl:
		case Opt_noacl:
1219 1220
			printk(KERN_ERR "EXT4 (no)acl options "
			       "not supported\n");
1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235
			break;
#endif
		case Opt_reservation:
			set_opt(sbi->s_mount_opt, RESERVATION);
			break;
		case Opt_noreservation:
			clear_opt(sbi->s_mount_opt, RESERVATION);
			break;
		case Opt_journal_update:
			/* @@@ FIXME */
			/* Eventually we will want to be able to create
			   a journal file here.  For now, only allow the
			   user to specify an existing inode to be the
			   journal file. */
			if (is_remount) {
1236
				printk(KERN_ERR "EXT4-fs: cannot specify "
1237 1238 1239
				       "journal on remount\n");
				return 0;
			}
1240
			set_opt(sbi->s_mount_opt, UPDATE_JOURNAL);
1241 1242 1243
			break;
		case Opt_journal_dev:
			if (is_remount) {
1244
				printk(KERN_ERR "EXT4-fs: cannot specify "
1245 1246 1247 1248 1249 1250 1251
				       "journal on remount\n");
				return 0;
			}
			if (match_int(&args[0], &option))
				return 0;
			*journal_devnum = option;
			break;
1252 1253 1254 1255 1256 1257 1258
		case Opt_journal_checksum:
			set_opt(sbi->s_mount_opt, JOURNAL_CHECKSUM);
			break;
		case Opt_journal_async_commit:
			set_opt(sbi->s_mount_opt, JOURNAL_ASYNC_COMMIT);
			set_opt(sbi->s_mount_opt, JOURNAL_CHECKSUM);
			break;
1259
		case Opt_noload:
1260
			set_opt(sbi->s_mount_opt, NOLOAD);
1261 1262 1263 1264 1265 1266 1267
			break;
		case Opt_commit:
			if (match_int(&args[0], &option))
				return 0;
			if (option < 0)
				return 0;
			if (option == 0)
1268
				option = JBD2_DEFAULT_MAX_COMMIT_AGE;
1269 1270
			sbi->s_commit_interval = HZ * option;
			break;
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286
		case Opt_max_batch_time:
			if (match_int(&args[0], &option))
				return 0;
			if (option < 0)
				return 0;
			if (option == 0)
				option = EXT4_DEF_MAX_BATCH_TIME;
			sbi->s_max_batch_time = option;
			break;
		case Opt_min_batch_time:
			if (match_int(&args[0], &option))
				return 0;
			if (option < 0)
				return 0;
			sbi->s_min_batch_time = option;
			break;
1287
		case Opt_data_journal:
1288
			data_opt = EXT4_MOUNT_JOURNAL_DATA;
1289 1290
			goto datacheck;
		case Opt_data_ordered:
1291
			data_opt = EXT4_MOUNT_ORDERED_DATA;
1292 1293
			goto datacheck;
		case Opt_data_writeback:
1294
			data_opt = EXT4_MOUNT_WRITEBACK_DATA;
1295 1296
		datacheck:
			if (is_remount) {
1297
				if ((sbi->s_mount_opt & EXT4_MOUNT_DATA_FLAGS)
1298 1299
						!= data_opt) {
					printk(KERN_ERR
1300
						"EXT4-fs: cannot change data "
1301 1302 1303 1304
						"mode on remount\n");
					return 0;
				}
			} else {
1305
				sbi->s_mount_opt &= ~EXT4_MOUNT_DATA_FLAGS;
1306 1307 1308
				sbi->s_mount_opt |= data_opt;
			}
			break;
1309 1310 1311 1312 1313 1314
		case Opt_data_err_abort:
			set_opt(sbi->s_mount_opt, DATA_ERR_ABORT);
			break;
		case Opt_data_err_ignore:
			clear_opt(sbi->s_mount_opt, DATA_ERR_ABORT);
			break;
1315 1316 1317 1318 1319 1320 1321
#ifdef CONFIG_QUOTA
		case Opt_usrjquota:
			qtype = USRQUOTA;
			goto set_qf_name;
		case Opt_grpjquota:
			qtype = GRPQUOTA;
set_qf_name:
1322
			if (sb_any_quota_loaded(sb) &&
1323
			    !sbi->s_qf_names[qtype]) {
1324
				printk(KERN_ERR
1325 1326
				       "EXT4-fs: Cannot change journaled "
				       "quota options when quota turned on.\n");
1327 1328 1329 1330 1331
				return 0;
			}
			qname = match_strdup(&args[0]);
			if (!qname) {
				printk(KERN_ERR
1332
					"EXT4-fs: not enough memory for "
1333 1334 1335 1336 1337 1338
					"storing quotafile name.\n");
				return 0;
			}
			if (sbi->s_qf_names[qtype] &&
			    strcmp(sbi->s_qf_names[qtype], qname)) {
				printk(KERN_ERR
1339
					"EXT4-fs: %s quota file already "
1340 1341 1342 1343 1344 1345 1346
					"specified.\n", QTYPE2NAME(qtype));
				kfree(qname);
				return 0;
			}
			sbi->s_qf_names[qtype] = qname;
			if (strchr(sbi->s_qf_names[qtype], '/')) {
				printk(KERN_ERR
1347
					"EXT4-fs: quotafile must be on "
1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360
					"filesystem root.\n");
				kfree(sbi->s_qf_names[qtype]);
				sbi->s_qf_names[qtype] = NULL;
				return 0;
			}
			set_opt(sbi->s_mount_opt, QUOTA);
			break;
		case Opt_offusrjquota:
			qtype = USRQUOTA;
			goto clear_qf_name;
		case Opt_offgrpjquota:
			qtype = GRPQUOTA;
clear_qf_name:
1361
			if (sb_any_quota_loaded(sb) &&
1362
			    sbi->s_qf_names[qtype]) {
1363
				printk(KERN_ERR "EXT4-fs: Cannot change "
1364
					"journaled quota options when "
1365 1366 1367 1368 1369 1370 1371 1372 1373 1374
					"quota turned on.\n");
				return 0;
			}
			/*
			 * The space will be released later when all options
			 * are confirmed to be correct
			 */
			sbi->s_qf_names[qtype] = NULL;
			break;
		case Opt_jqfmt_vfsold:
1375 1376
			qfmt = QFMT_VFS_OLD;
			goto set_qf_format;
1377
		case Opt_jqfmt_vfsv0:
1378 1379
			qfmt = QFMT_VFS_V0;
set_qf_format:
1380
			if (sb_any_quota_loaded(sb) &&
1381 1382 1383 1384 1385 1386 1387
			    sbi->s_jquota_fmt != qfmt) {
				printk(KERN_ERR "EXT4-fs: Cannot change "
					"journaled quota options when "
					"quota turned on.\n");
				return 0;
			}
			sbi->s_jquota_fmt = qfmt;
1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398
			break;
		case Opt_quota:
		case Opt_usrquota:
			set_opt(sbi->s_mount_opt, QUOTA);
			set_opt(sbi->s_mount_opt, USRQUOTA);
			break;
		case Opt_grpquota:
			set_opt(sbi->s_mount_opt, QUOTA);
			set_opt(sbi->s_mount_opt, GRPQUOTA);
			break;
		case Opt_noquota:
1399
			if (sb_any_quota_loaded(sb)) {
1400
				printk(KERN_ERR "EXT4-fs: Cannot change quota "
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411
					"options when quota turned on.\n");
				return 0;
			}
			clear_opt(sbi->s_mount_opt, QUOTA);
			clear_opt(sbi->s_mount_opt, USRQUOTA);
			clear_opt(sbi->s_mount_opt, GRPQUOTA);
			break;
#else
		case Opt_quota:
		case Opt_usrquota:
		case Opt_grpquota:
1412 1413 1414
			printk(KERN_ERR
				"EXT4-fs: quota options not supported.\n");
			break;
1415 1416 1417 1418 1419 1420 1421
		case Opt_usrjquota:
		case Opt_grpjquota:
		case Opt_offusrjquota:
		case Opt_offgrpjquota:
		case Opt_jqfmt_vfsold:
		case Opt_jqfmt_vfsv0:
			printk(KERN_ERR
1422
				"EXT4-fs: journaled quota options not "
1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442
				"supported.\n");
			break;
		case Opt_noquota:
			break;
#endif
		case Opt_abort:
			set_opt(sbi->s_mount_opt, ABORT);
			break;
		case Opt_barrier:
			if (match_int(&args[0], &option))
				return 0;
			if (option)
				set_opt(sbi->s_mount_opt, BARRIER);
			else
				clear_opt(sbi->s_mount_opt, BARRIER);
			break;
		case Opt_ignore:
			break;
		case Opt_resize:
			if (!is_remount) {
1443
				printk("EXT4-fs: resize option only available "
1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456
					"for remount\n");
				return 0;
			}
			if (match_int(&args[0], &option) != 0)
				return 0;
			*n_blocks_count = option;
			break;
		case Opt_nobh:
			set_opt(sbi->s_mount_opt, NOBH);
			break;
		case Opt_bh:
			clear_opt(sbi->s_mount_opt, NOBH);
			break;
1457 1458 1459 1460
		case Opt_i_version:
			set_opt(sbi->s_mount_opt, I_VERSION);
			sb->s_flags |= MS_I_VERSION;
			break;
1461 1462 1463
		case Opt_nodelalloc:
			clear_opt(sbi->s_mount_opt, DELALLOC);
			break;
1464 1465 1466 1467 1468 1469 1470
		case Opt_stripe:
			if (match_int(&args[0], &option))
				return 0;
			if (option < 0)
				return 0;
			sbi->s_stripe = option;
			break;
1471 1472 1473
		case Opt_delalloc:
			set_opt(sbi->s_mount_opt, DELALLOC);
			break;
1474 1475 1476 1477 1478
		case Opt_inode_readahead_blks:
			if (match_int(&args[0], &option))
				return 0;
			if (option < 0 || option > (1 << 30))
				return 0;
T
Theodore Ts'o 已提交
1479 1480 1481 1482 1483
			if (option & (option - 1)) {
				printk(KERN_ERR "EXT4-fs: inode_readahead_blks"
				       " must be a power of 2\n");
				return 0;
			}
1484 1485
			sbi->s_inode_readahead_blks = option;
			break;
1486 1487 1488 1489 1490 1491 1492 1493
		case Opt_journal_ioprio:
			if (match_int(&args[0], &option))
				return 0;
			if (option < 0 || option > 7)
				break;
			*journal_ioprio = IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE,
							    option);
			break;
1494
		default:
1495 1496 1497
			printk(KERN_ERR
			       "EXT4-fs: Unrecognized mount option \"%s\" "
			       "or missing value\n", p);
1498 1499 1500 1501 1502
			return 0;
		}
	}
#ifdef CONFIG_QUOTA
	if (sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) {
1503
		if ((sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA) &&
1504 1505 1506
		     sbi->s_qf_names[USRQUOTA])
			clear_opt(sbi->s_mount_opt, USRQUOTA);

1507
		if ((sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA) &&
1508 1509 1510 1511
		     sbi->s_qf_names[GRPQUOTA])
			clear_opt(sbi->s_mount_opt, GRPQUOTA);

		if ((sbi->s_qf_names[USRQUOTA] &&
1512
				(sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA)) ||
1513
		    (sbi->s_qf_names[GRPQUOTA] &&
1514 1515
				(sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA))) {
			printk(KERN_ERR "EXT4-fs: old and new quota "
1516 1517 1518 1519 1520
					"format mixing.\n");
			return 0;
		}

		if (!sbi->s_jquota_fmt) {
1521
			printk(KERN_ERR "EXT4-fs: journaled quota format "
1522 1523 1524 1525 1526
					"not specified.\n");
			return 0;
		}
	} else {
		if (sbi->s_jquota_fmt) {
1527 1528
			printk(KERN_ERR "EXT4-fs: journaled quota format "
					"specified with no journaling "
1529 1530 1531 1532 1533 1534 1535 1536
					"enabled.\n");
			return 0;
		}
	}
#endif
	return 1;
}

1537
static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es,
1538 1539
			    int read_only)
{
1540
	struct ext4_sb_info *sbi = EXT4_SB(sb);
1541 1542
	int res = 0;

1543
	if (le32_to_cpu(es->s_rev_level) > EXT4_MAX_SUPP_REV) {
1544 1545
		printk(KERN_ERR "EXT4-fs warning: revision level too high, "
		       "forcing read-only mode\n");
1546 1547 1548 1549
		res = MS_RDONLY;
	}
	if (read_only)
		return res;
1550
	if (!(sbi->s_mount_state & EXT4_VALID_FS))
1551 1552
		printk(KERN_WARNING "EXT4-fs warning: mounting unchecked fs, "
		       "running e2fsck is recommended\n");
1553
	else if ((sbi->s_mount_state & EXT4_ERROR_FS))
1554 1555 1556
		printk(KERN_WARNING
		       "EXT4-fs warning: mounting fs with errors, "
		       "running e2fsck is recommended\n");
1557 1558 1559
	else if ((__s16) le16_to_cpu(es->s_max_mnt_count) >= 0 &&
		 le16_to_cpu(es->s_mnt_count) >=
		 (unsigned short) (__s16) le16_to_cpu(es->s_max_mnt_count))
1560 1561 1562
		printk(KERN_WARNING
		       "EXT4-fs warning: maximal mount count reached, "
		       "running e2fsck is recommended\n");
1563 1564 1565
	else if (le32_to_cpu(es->s_checkinterval) &&
		(le32_to_cpu(es->s_lastcheck) +
			le32_to_cpu(es->s_checkinterval) <= get_seconds()))
1566 1567 1568
		printk(KERN_WARNING
		       "EXT4-fs warning: checktime reached, "
		       "running e2fsck is recommended\n");
1569 1570
	if (!sbi->s_journal) 
		es->s_state &= cpu_to_le16(~EXT4_VALID_FS);
1571
	if (!(__s16) le16_to_cpu(es->s_max_mnt_count))
1572
		es->s_max_mnt_count = cpu_to_le16(EXT4_DFL_MAX_MNT_COUNT);
M
Marcin Slusarz 已提交
1573
	le16_add_cpu(&es->s_mnt_count, 1);
1574
	es->s_mtime = cpu_to_le32(get_seconds());
1575
	ext4_update_dynamic_rev(sb);
1576 1577
	if (sbi->s_journal)
		EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
1578

1579
	ext4_commit_super(sb, es, 1);
1580
	if (test_opt(sb, DEBUG))
1581
		printk(KERN_INFO "[EXT4 FS bs=%lu, gc=%u, "
1582 1583 1584
				"bpg=%lu, ipg=%lu, mo=%04lx]\n",
			sb->s_blocksize,
			sbi->s_groups_count,
1585 1586
			EXT4_BLOCKS_PER_GROUP(sb),
			EXT4_INODES_PER_GROUP(sb),
1587 1588
			sbi->s_mount_opt);

1589 1590 1591 1592 1593 1594 1595
	if (EXT4_SB(sb)->s_journal) {
		printk(KERN_INFO "EXT4 FS on %s, %s journal on %s\n",
		       sb->s_id, EXT4_SB(sb)->s_journal->j_inode ? "internal" :
		       "external", EXT4_SB(sb)->s_journal->j_devname);
	} else {
		printk(KERN_INFO "EXT4 FS on %s, no journal\n", sb->s_id);
	}
1596 1597 1598
	return res;
}

1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
static int ext4_fill_flex_info(struct super_block *sb)
{
	struct ext4_sb_info *sbi = EXT4_SB(sb);
	struct ext4_group_desc *gdp = NULL;
	struct buffer_head *bh;
	ext4_group_t flex_group_count;
	ext4_group_t flex_group;
	int groups_per_flex = 0;
	int i;

	if (!sbi->s_es->s_log_groups_per_flex) {
		sbi->s_log_groups_per_flex = 0;
		return 1;
	}

	sbi->s_log_groups_per_flex = sbi->s_es->s_log_groups_per_flex;
	groups_per_flex = 1 << sbi->s_log_groups_per_flex;

1617 1618
	/* We allocate both existing and potentially added groups */
	flex_group_count = ((sbi->s_groups_count + groups_per_flex - 1) +
1619 1620
			((le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) + 1) <<
			      EXT4_DESC_PER_BLOCK_BITS(sb))) / groups_per_flex;
1621
	sbi->s_flex_groups = kzalloc(flex_group_count *
1622 1623
				     sizeof(struct flex_groups), GFP_KERNEL);
	if (sbi->s_flex_groups == NULL) {
1624
		printk(KERN_ERR "EXT4-fs: not enough memory for "
1625
				"%u flex groups\n", flex_group_count);
1626 1627 1628 1629 1630 1631 1632
		goto failed;
	}

	for (i = 0; i < sbi->s_groups_count; i++) {
		gdp = ext4_get_group_desc(sb, i, &bh);

		flex_group = ext4_flex_group(sbi, i);
1633 1634 1635 1636
		atomic_set(&sbi->s_flex_groups[flex_group].free_inodes,
			   ext4_free_inodes_count(sb, gdp));
		atomic_set(&sbi->s_flex_groups[flex_group].free_blocks,
			   ext4_free_blks_count(sb, gdp));
1637 1638
		atomic_set(&sbi->s_flex_groups[flex_group].used_dirs,
			   ext4_used_dirs_count(sb, gdp));
1639 1640 1641 1642 1643 1644 1645
	}

	return 1;
failed:
	return 0;
}

A
Andreas Dilger 已提交
1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682
__le16 ext4_group_desc_csum(struct ext4_sb_info *sbi, __u32 block_group,
			    struct ext4_group_desc *gdp)
{
	__u16 crc = 0;

	if (sbi->s_es->s_feature_ro_compat &
	    cpu_to_le32(EXT4_FEATURE_RO_COMPAT_GDT_CSUM)) {
		int offset = offsetof(struct ext4_group_desc, bg_checksum);
		__le32 le_group = cpu_to_le32(block_group);

		crc = crc16(~0, sbi->s_es->s_uuid, sizeof(sbi->s_es->s_uuid));
		crc = crc16(crc, (__u8 *)&le_group, sizeof(le_group));
		crc = crc16(crc, (__u8 *)gdp, offset);
		offset += sizeof(gdp->bg_checksum); /* skip checksum */
		/* for checksum of struct ext4_group_desc do the rest...*/
		if ((sbi->s_es->s_feature_incompat &
		     cpu_to_le32(EXT4_FEATURE_INCOMPAT_64BIT)) &&
		    offset < le16_to_cpu(sbi->s_es->s_desc_size))
			crc = crc16(crc, (__u8 *)gdp + offset,
				    le16_to_cpu(sbi->s_es->s_desc_size) -
					offset);
	}

	return cpu_to_le16(crc);
}

int ext4_group_desc_csum_verify(struct ext4_sb_info *sbi, __u32 block_group,
				struct ext4_group_desc *gdp)
{
	if ((sbi->s_es->s_feature_ro_compat &
	     cpu_to_le32(EXT4_FEATURE_RO_COMPAT_GDT_CSUM)) &&
	    (gdp->bg_checksum != ext4_group_desc_csum(sbi, block_group, gdp)))
		return 0;

	return 1;
}

1683
/* Called at mount-time, super-block is locked */
1684
static int ext4_check_descriptors(struct super_block *sb)
1685
{
1686 1687 1688
	struct ext4_sb_info *sbi = EXT4_SB(sb);
	ext4_fsblk_t first_block = le32_to_cpu(sbi->s_es->s_first_data_block);
	ext4_fsblk_t last_block;
L
Laurent Vivier 已提交
1689 1690 1691
	ext4_fsblk_t block_bitmap;
	ext4_fsblk_t inode_bitmap;
	ext4_fsblk_t inode_table;
J
Jose R. Santos 已提交
1692
	int flexbg_flag = 0;
1693
	ext4_group_t i;
1694

J
Jose R. Santos 已提交
1695 1696 1697
	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG))
		flexbg_flag = 1;

1698
	ext4_debug("Checking group descriptors");
1699

1700 1701 1702
	for (i = 0; i < sbi->s_groups_count; i++) {
		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, i, NULL);

J
Jose R. Santos 已提交
1703
		if (i == sbi->s_groups_count - 1 || flexbg_flag)
L
Laurent Vivier 已提交
1704
			last_block = ext4_blocks_count(sbi->s_es) - 1;
1705 1706
		else
			last_block = first_block +
1707
				(EXT4_BLOCKS_PER_GROUP(sb) - 1);
1708

1709
		block_bitmap = ext4_block_bitmap(sb, gdp);
1710
		if (block_bitmap < first_block || block_bitmap > last_block) {
1711
			printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: "
1712
			       "Block bitmap for group %u not in group "
1713
			       "(block %llu)!\n", i, block_bitmap);
1714 1715
			return 0;
		}
1716
		inode_bitmap = ext4_inode_bitmap(sb, gdp);
1717
		if (inode_bitmap < first_block || inode_bitmap > last_block) {
1718
			printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: "
1719
			       "Inode bitmap for group %u not in group "
1720
			       "(block %llu)!\n", i, inode_bitmap);
1721 1722
			return 0;
		}
1723
		inode_table = ext4_inode_table(sb, gdp);
L
Laurent Vivier 已提交
1724
		if (inode_table < first_block ||
1725
		    inode_table + sbi->s_itb_per_group - 1 > last_block) {
1726
			printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: "
1727
			       "Inode table for group %u not in group "
1728
			       "(block %llu)!\n", i, inode_table);
1729 1730
			return 0;
		}
1731
		spin_lock(sb_bgl_lock(sbi, i));
A
Andreas Dilger 已提交
1732
		if (!ext4_group_desc_csum_verify(sbi, i, gdp)) {
1733
			printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: "
1734
			       "Checksum for group %u failed (%u!=%u)\n",
1735 1736
			       i, le16_to_cpu(ext4_group_desc_csum(sbi, i,
			       gdp)), le16_to_cpu(gdp->bg_checksum));
1737 1738
			if (!(sb->s_flags & MS_RDONLY)) {
				spin_unlock(sb_bgl_lock(sbi, i));
1739
				return 0;
1740
			}
A
Andreas Dilger 已提交
1741
		}
1742
		spin_unlock(sb_bgl_lock(sbi, i));
J
Jose R. Santos 已提交
1743 1744
		if (!flexbg_flag)
			first_block += EXT4_BLOCKS_PER_GROUP(sb);
1745 1746
	}

L
Laurent Vivier 已提交
1747
	ext4_free_blocks_count_set(sbi->s_es, ext4_count_free_blocks(sb));
1748
	sbi->s_es->s_free_inodes_count = cpu_to_le32(ext4_count_free_inodes(sb));
1749 1750 1751
	return 1;
}

1752
/* ext4_orphan_cleanup() walks a singly-linked list of inodes (starting at
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764
 * the superblock) which were deleted from all directories, but held open by
 * a process at the time of a crash.  We walk the list and try to delete these
 * inodes at recovery time (only with a read-write filesystem).
 *
 * In order to keep the orphan inode chain consistent during traversal (in
 * case of crash during recovery), we link each inode into the superblock
 * orphan list_head and handle it the same way as an inode deletion during
 * normal operation (which journals the operations for us).
 *
 * We only do an iget() and an iput() on each inode, which is very safe if we
 * accidentally point at an in-use or already deleted inode.  The worst that
 * can happen in this case is that we get a "bit already cleared" message from
1765
 * ext4_free_inode().  The only reason we would point at a wrong inode is if
1766 1767 1768
 * e2fsck was run on this filesystem, and it must have already done the orphan
 * inode cleanup for us, so we can safely abort without any further action.
 */
1769 1770
static void ext4_orphan_cleanup(struct super_block *sb,
				struct ext4_super_block *es)
1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781
{
	unsigned int s_flags = sb->s_flags;
	int nr_orphans = 0, nr_truncates = 0;
#ifdef CONFIG_QUOTA
	int i;
#endif
	if (!es->s_last_orphan) {
		jbd_debug(4, "no orphan inodes to clean up\n");
		return;
	}

1782 1783 1784 1785 1786 1787
	if (bdev_read_only(sb->s_bdev)) {
		printk(KERN_ERR "EXT4-fs: write access "
			"unavailable, skipping orphan cleanup.\n");
		return;
	}

1788
	if (EXT4_SB(sb)->s_mount_state & EXT4_ERROR_FS) {
1789 1790 1791 1792 1793 1794 1795 1796 1797
		if (es->s_last_orphan)
			jbd_debug(1, "Errors on filesystem, "
				  "clearing orphan list.\n");
		es->s_last_orphan = 0;
		jbd_debug(1, "Skipping orphan recovery on fs with errors.\n");
		return;
	}

	if (s_flags & MS_RDONLY) {
1798
		printk(KERN_INFO "EXT4-fs: %s: orphan cleanup on readonly fs\n",
1799 1800 1801 1802 1803 1804 1805 1806
		       sb->s_id);
		sb->s_flags &= ~MS_RDONLY;
	}
#ifdef CONFIG_QUOTA
	/* Needed for iput() to work correctly and not trash data */
	sb->s_flags |= MS_ACTIVE;
	/* Turn on quotas so that they are updated correctly */
	for (i = 0; i < MAXQUOTAS; i++) {
1807 1808
		if (EXT4_SB(sb)->s_qf_names[i]) {
			int ret = ext4_quota_on_mount(sb, i);
1809 1810
			if (ret < 0)
				printk(KERN_ERR
1811
					"EXT4-fs: Cannot turn on journaled "
1812 1813 1814 1815 1816 1817 1818 1819
					"quota: error %d\n", ret);
		}
	}
#endif

	while (es->s_last_orphan) {
		struct inode *inode;

1820 1821
		inode = ext4_orphan_get(sb, le32_to_cpu(es->s_last_orphan));
		if (IS_ERR(inode)) {
1822 1823 1824 1825
			es->s_last_orphan = 0;
			break;
		}

1826
		list_add(&EXT4_I(inode)->i_orphan, &EXT4_SB(sb)->s_orphan);
1827
		vfs_dq_init(inode);
1828 1829
		if (inode->i_nlink) {
			printk(KERN_DEBUG
1830
				"%s: truncating inode %lu to %lld bytes\n",
1831
				__func__, inode->i_ino, inode->i_size);
1832
			jbd_debug(2, "truncating inode %lu to %lld bytes\n",
1833
				  inode->i_ino, inode->i_size);
1834
			ext4_truncate(inode);
1835 1836 1837 1838
			nr_truncates++;
		} else {
			printk(KERN_DEBUG
				"%s: deleting unreferenced inode %lu\n",
1839
				__func__, inode->i_ino);
1840 1841 1842 1843 1844 1845 1846
			jbd_debug(2, "deleting unreferenced inode %lu\n",
				  inode->i_ino);
			nr_orphans++;
		}
		iput(inode);  /* The delete magic happens here! */
	}

1847
#define PLURAL(x) (x), ((x) == 1) ? "" : "s"
1848 1849

	if (nr_orphans)
1850
		printk(KERN_INFO "EXT4-fs: %s: %d orphan inode%s deleted\n",
1851 1852
		       sb->s_id, PLURAL(nr_orphans));
	if (nr_truncates)
1853
		printk(KERN_INFO "EXT4-fs: %s: %d truncate%s cleaned up\n",
1854 1855 1856 1857 1858
		       sb->s_id, PLURAL(nr_truncates));
#ifdef CONFIG_QUOTA
	/* Turn quotas off */
	for (i = 0; i < MAXQUOTAS; i++) {
		if (sb_dqopt(sb)->files[i])
1859
			vfs_quota_off(sb, i, 0);
1860 1861 1862 1863
	}
#endif
	sb->s_flags = s_flags; /* Restore MS_RDONLY status */
}
1864 1865 1866 1867 1868 1869 1870 1871 1872
/*
 * Maximal extent format file size.
 * Resulting logical blkno at s_maxbytes must fit in our on-disk
 * extent format containers, within a sector_t, and within i_blocks
 * in the vfs.  ext4 inode has 48 bits of i_block in fsblock units,
 * so that won't be a limiting factor.
 *
 * Note, this does *not* consider any metadata overhead for vfs i_blocks.
 */
1873
static loff_t ext4_max_size(int blkbits, int has_huge_files)
1874 1875 1876 1877 1878
{
	loff_t res;
	loff_t upper_limit = MAX_LFS_FILESIZE;

	/* small i_blocks in vfs inode? */
1879
	if (!has_huge_files || sizeof(blkcnt_t) < sizeof(u64)) {
1880
		/*
J
Jens Axboe 已提交
1881
		 * CONFIG_LBD is not enabled implies the inode
1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902
		 * i_block represent total blocks in 512 bytes
		 * 32 == size of vfs inode i_blocks * 8
		 */
		upper_limit = (1LL << 32) - 1;

		/* total blocks in file system block size */
		upper_limit >>= (blkbits - 9);
		upper_limit <<= blkbits;
	}

	/* 32-bit extent-start container, ee_block */
	res = 1LL << 32;
	res <<= blkbits;
	res -= 1;

	/* Sanity check against vm- & vfs- imposed limits */
	if (res > upper_limit)
		res = upper_limit;

	return res;
}
1903 1904

/*
1905
 * Maximal bitmap file size.  There is a direct, and {,double-,triple-}indirect
1906 1907
 * block limit, and also a limit of (2^48 - 1) 512-byte sectors in i_blocks.
 * We need to be 1 filesystem block less than the 2^48 sector limit.
1908
 */
1909
static loff_t ext4_max_bitmap_size(int bits, int has_huge_files)
1910
{
1911
	loff_t res = EXT4_NDIR_BLOCKS;
1912 1913 1914
	int meta_blocks;
	loff_t upper_limit;
	/* This is calculated to be the largest file size for a
1915
	 * dense, bitmapped file such that the total number of
1916
	 * sectors in the file, including data and all indirect blocks,
1917 1918 1919 1920 1921
	 * does not exceed 2^48 -1
	 * __u32 i_blocks_lo and _u16 i_blocks_high representing the
	 * total number of  512 bytes blocks of the file
	 */

1922
	if (!has_huge_files || sizeof(blkcnt_t) < sizeof(u64)) {
1923
		/*
J
Jens Axboe 已提交
1924
		 * !has_huge_files or CONFIG_LBD is not enabled
1925 1926
		 * implies the inode i_block represent total blocks in
		 * 512 bytes 32 == size of vfs inode i_blocks * 8
1927 1928 1929 1930 1931 1932 1933
		 */
		upper_limit = (1LL << 32) - 1;

		/* total blocks in file system block size */
		upper_limit >>= (bits - 9);

	} else {
A
Aneesh Kumar K.V 已提交
1934 1935 1936 1937 1938 1939
		/*
		 * We use 48 bit ext4_inode i_blocks
		 * With EXT4_HUGE_FILE_FL set the i_blocks
		 * represent total number of blocks in
		 * file system block size
		 */
1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952
		upper_limit = (1LL << 48) - 1;

	}

	/* indirect blocks */
	meta_blocks = 1;
	/* double indirect blocks */
	meta_blocks += 1 + (1LL << (bits-2));
	/* tripple indirect blocks */
	meta_blocks += 1 + (1LL << (bits-2)) + (1LL << (2*(bits-2)));

	upper_limit -= meta_blocks;
	upper_limit <<= bits;
1953 1954 1955 1956 1957 1958 1959

	res += 1LL << (bits-2);
	res += 1LL << (2*(bits-2));
	res += 1LL << (3*(bits-2));
	res <<= bits;
	if (res > upper_limit)
		res = upper_limit;
1960 1961 1962 1963

	if (res > MAX_LFS_FILESIZE)
		res = MAX_LFS_FILESIZE;

1964 1965 1966
	return res;
}

1967
static ext4_fsblk_t descriptor_loc(struct super_block *sb,
1968
				ext4_fsblk_t logical_sb_block, int nr)
1969
{
1970
	struct ext4_sb_info *sbi = EXT4_SB(sb);
1971
	ext4_group_t bg, first_meta_bg;
1972 1973 1974 1975
	int has_super = 0;

	first_meta_bg = le32_to_cpu(sbi->s_es->s_first_meta_bg);

1976
	if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_META_BG) ||
1977
	    nr < first_meta_bg)
1978
		return logical_sb_block + nr + 1;
1979
	bg = sbi->s_desc_per_block * nr;
1980
	if (ext4_bg_has_super(sb, bg))
1981
		has_super = 1;
1982
	return (has_super + ext4_group_first_block_no(sb, bg));
1983 1984
}

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
/**
 * ext4_get_stripe_size: Get the stripe size.
 * @sbi: In memory super block info
 *
 * If we have specified it via mount option, then
 * use the mount option value. If the value specified at mount time is
 * greater than the blocks per group use the super block value.
 * If the super block value is greater than blocks per group return 0.
 * Allocator needs it be less than blocks per group.
 *
 */
static unsigned long ext4_get_stripe_size(struct ext4_sb_info *sbi)
{
	unsigned long stride = le16_to_cpu(sbi->s_es->s_raid_stride);
	unsigned long stripe_width =
			le32_to_cpu(sbi->s_es->s_raid_stripe_width);

	if (sbi->s_stripe && sbi->s_stripe <= sbi->s_blocks_per_group)
		return sbi->s_stripe;

	if (stripe_width <= sbi->s_blocks_per_group)
		return stripe_width;

	if (stride <= sbi->s_blocks_per_group)
		return stride;

	return 0;
}
2013

T
Theodore Ts'o 已提交
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188
/* sysfs supprt */

struct ext4_attr {
	struct attribute attr;
	ssize_t (*show)(struct ext4_attr *, struct ext4_sb_info *, char *);
	ssize_t (*store)(struct ext4_attr *, struct ext4_sb_info *, 
			 const char *, size_t);
	int offset;
};

static int parse_strtoul(const char *buf,
		unsigned long max, unsigned long *value)
{
	char *endp;

	while (*buf && isspace(*buf))
		buf++;
	*value = simple_strtoul(buf, &endp, 0);
	while (*endp && isspace(*endp))
		endp++;
	if (*endp || *value > max)
		return -EINVAL;

	return 0;
}

static ssize_t delayed_allocation_blocks_show(struct ext4_attr *a,
					      struct ext4_sb_info *sbi,
					      char *buf)
{
	return snprintf(buf, PAGE_SIZE, "%llu\n",
			(s64) percpu_counter_sum(&sbi->s_dirtyblocks_counter));
}

static ssize_t session_write_kbytes_show(struct ext4_attr *a,
					 struct ext4_sb_info *sbi, char *buf)
{
	struct super_block *sb = sbi->s_buddy_cache->i_sb;

	return snprintf(buf, PAGE_SIZE, "%lu\n",
			(part_stat_read(sb->s_bdev->bd_part, sectors[1]) -
			 sbi->s_sectors_written_start) >> 1);
}

static ssize_t lifetime_write_kbytes_show(struct ext4_attr *a,
					  struct ext4_sb_info *sbi, char *buf)
{
	struct super_block *sb = sbi->s_buddy_cache->i_sb;

	return snprintf(buf, PAGE_SIZE, "%llu\n",
			sbi->s_kbytes_written + 
			((part_stat_read(sb->s_bdev->bd_part, sectors[1]) -
			  EXT4_SB(sb)->s_sectors_written_start) >> 1));
}

static ssize_t inode_readahead_blks_store(struct ext4_attr *a,
					  struct ext4_sb_info *sbi,
					  const char *buf, size_t count)
{
	unsigned long t;

	if (parse_strtoul(buf, 0x40000000, &t))
		return -EINVAL;

	/* inode_readahead_blks must be a power of 2 */
	if (t & (t-1))
		return -EINVAL;

	sbi->s_inode_readahead_blks = t;
	return count;
}

static ssize_t sbi_ui_show(struct ext4_attr *a,
				struct ext4_sb_info *sbi, char *buf)
{
	unsigned int *ui = (unsigned int *) (((char *) sbi) + a->offset);

	return snprintf(buf, PAGE_SIZE, "%u\n", *ui);
}

static ssize_t sbi_ui_store(struct ext4_attr *a,
			    struct ext4_sb_info *sbi,
			    const char *buf, size_t count)
{
	unsigned int *ui = (unsigned int *) (((char *) sbi) + a->offset);
	unsigned long t;

	if (parse_strtoul(buf, 0xffffffff, &t))
		return -EINVAL;
	*ui = t;
	return count;
}

#define EXT4_ATTR_OFFSET(_name,_mode,_show,_store,_elname) \
static struct ext4_attr ext4_attr_##_name = {			\
	.attr = {.name = __stringify(_name), .mode = _mode },	\
	.show	= _show,					\
	.store	= _store,					\
	.offset = offsetof(struct ext4_sb_info, _elname),	\
}
#define EXT4_ATTR(name, mode, show, store) \
static struct ext4_attr ext4_attr_##name = __ATTR(name, mode, show, store)

#define EXT4_RO_ATTR(name) EXT4_ATTR(name, 0444, name##_show, NULL)
#define EXT4_RW_ATTR(name) EXT4_ATTR(name, 0644, name##_show, name##_store)
#define EXT4_RW_ATTR_SBI_UI(name, elname)	\
	EXT4_ATTR_OFFSET(name, 0644, sbi_ui_show, sbi_ui_store, elname)
#define ATTR_LIST(name) &ext4_attr_##name.attr

EXT4_RO_ATTR(delayed_allocation_blocks);
EXT4_RO_ATTR(session_write_kbytes);
EXT4_RO_ATTR(lifetime_write_kbytes);
EXT4_ATTR_OFFSET(inode_readahead_blks, 0644, sbi_ui_show,
		 inode_readahead_blks_store, s_inode_readahead_blks);
EXT4_RW_ATTR_SBI_UI(mb_stats, s_mb_stats);
EXT4_RW_ATTR_SBI_UI(mb_max_to_scan, s_mb_max_to_scan);
EXT4_RW_ATTR_SBI_UI(mb_min_to_scan, s_mb_min_to_scan);
EXT4_RW_ATTR_SBI_UI(mb_order2_req, s_mb_order2_reqs);
EXT4_RW_ATTR_SBI_UI(mb_stream_req, s_mb_stream_request);
EXT4_RW_ATTR_SBI_UI(mb_group_prealloc, s_mb_group_prealloc);

static struct attribute *ext4_attrs[] = {
	ATTR_LIST(delayed_allocation_blocks),
	ATTR_LIST(session_write_kbytes),
	ATTR_LIST(lifetime_write_kbytes),
	ATTR_LIST(inode_readahead_blks),
	ATTR_LIST(mb_stats),
	ATTR_LIST(mb_max_to_scan),
	ATTR_LIST(mb_min_to_scan),
	ATTR_LIST(mb_order2_req),
	ATTR_LIST(mb_stream_req),
	ATTR_LIST(mb_group_prealloc),
	NULL,
};

static ssize_t ext4_attr_show(struct kobject *kobj,
			      struct attribute *attr, char *buf)
{
	struct ext4_sb_info *sbi = container_of(kobj, struct ext4_sb_info,
						s_kobj);
	struct ext4_attr *a = container_of(attr, struct ext4_attr, attr);

	return a->show ? a->show(a, sbi, buf) : 0;
}

static ssize_t ext4_attr_store(struct kobject *kobj,
			       struct attribute *attr,
			       const char *buf, size_t len)
{
	struct ext4_sb_info *sbi = container_of(kobj, struct ext4_sb_info,
						s_kobj);
	struct ext4_attr *a = container_of(attr, struct ext4_attr, attr);

	return a->store ? a->store(a, sbi, buf, len) : 0;
}

static void ext4_sb_release(struct kobject *kobj)
{
	struct ext4_sb_info *sbi = container_of(kobj, struct ext4_sb_info,
						s_kobj);
	complete(&sbi->s_kobj_unregister);
}


static struct sysfs_ops ext4_attr_ops = {
	.show	= ext4_attr_show,
	.store	= ext4_attr_store,
};

static struct kobj_type ext4_ktype = {
	.default_attrs	= ext4_attrs,
	.sysfs_ops	= &ext4_attr_ops,
	.release	= ext4_sb_release,
};

2189
static int ext4_fill_super(struct super_block *sb, void *data, int silent)
A
Aneesh Kumar K.V 已提交
2190 2191
				__releases(kernel_lock)
				__acquires(kernel_lock)
A
Aneesh Kumar K.V 已提交
2192

2193
{
2194
	struct buffer_head *bh;
2195 2196 2197 2198
	struct ext4_super_block *es = NULL;
	struct ext4_sb_info *sbi;
	ext4_fsblk_t block;
	ext4_fsblk_t sb_block = get_sb_block(&data);
2199
	ext4_fsblk_t logical_sb_block;
2200 2201 2202 2203
	unsigned long offset = 0;
	unsigned long journal_devnum = 0;
	unsigned long def_mount_opts;
	struct inode *root;
2204
	char *cp;
2205
	const char *descr;
2206
	int ret = -EINVAL;
2207
	int blocksize;
2208 2209
	unsigned int db_count;
	unsigned int i;
2210
	int needs_recovery, has_huge_files;
A
Aneesh Kumar K.V 已提交
2211
	int features;
L
Laurent Vivier 已提交
2212
	__u64 blocks_count;
2213
	int err;
2214
	unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO;
2215 2216 2217 2218

	sbi = kzalloc(sizeof(*sbi), GFP_KERNEL);
	if (!sbi)
		return -ENOMEM;
2219 2220 2221 2222 2223 2224 2225

	sbi->s_blockgroup_lock =
		kzalloc(sizeof(struct blockgroup_lock), GFP_KERNEL);
	if (!sbi->s_blockgroup_lock) {
		kfree(sbi);
		return -ENOMEM;
	}
2226 2227
	sb->s_fs_info = sbi;
	sbi->s_mount_opt = 0;
2228 2229
	sbi->s_resuid = EXT4_DEF_RESUID;
	sbi->s_resgid = EXT4_DEF_RESGID;
2230
	sbi->s_inode_readahead_blks = EXT4_DEF_INODE_READAHEAD_BLKS;
M
Miklos Szeredi 已提交
2231
	sbi->s_sb_block = sb_block;
T
Theodore Ts'o 已提交
2232 2233
	sbi->s_sectors_written_start = part_stat_read(sb->s_bdev->bd_part,
						      sectors[1]);
2234 2235 2236

	unlock_kernel();

2237 2238 2239 2240
	/* Cleanup superblock name */
	for (cp = sb->s_id; (cp = strchr(cp, '/'));)
		*cp = '!';

2241
	blocksize = sb_min_blocksize(sb, EXT4_MIN_BLOCK_SIZE);
2242
	if (!blocksize) {
2243
		printk(KERN_ERR "EXT4-fs: unable to set blocksize\n");
2244 2245 2246 2247
		goto out_fail;
	}

	/*
2248
	 * The ext4 superblock will not be buffer aligned for other than 1kB
2249 2250
	 * block sizes.  We need to calculate the offset from buffer start.
	 */
2251
	if (blocksize != EXT4_MIN_BLOCK_SIZE) {
2252 2253
		logical_sb_block = sb_block * EXT4_MIN_BLOCK_SIZE;
		offset = do_div(logical_sb_block, blocksize);
2254
	} else {
2255
		logical_sb_block = sb_block;
2256 2257
	}

2258
	if (!(bh = sb_bread(sb, logical_sb_block))) {
2259
		printk(KERN_ERR "EXT4-fs: unable to read superblock\n");
2260 2261 2262 2263
		goto out_fail;
	}
	/*
	 * Note: s_es must be initialized as soon as possible because
2264
	 *       some ext4 macro-instructions depend on its value
2265
	 */
2266
	es = (struct ext4_super_block *) (((char *)bh->b_data) + offset);
2267 2268
	sbi->s_es = es;
	sb->s_magic = le16_to_cpu(es->s_magic);
2269 2270
	if (sb->s_magic != EXT4_SUPER_MAGIC)
		goto cantfind_ext4;
T
Theodore Ts'o 已提交
2271
	sbi->s_kbytes_written = le64_to_cpu(es->s_kbytes_written);
2272 2273 2274

	/* Set defaults before we parse the mount options */
	def_mount_opts = le32_to_cpu(es->s_default_mount_opts);
2275
	if (def_mount_opts & EXT4_DEFM_DEBUG)
2276
		set_opt(sbi->s_mount_opt, DEBUG);
2277
	if (def_mount_opts & EXT4_DEFM_BSDGROUPS)
2278
		set_opt(sbi->s_mount_opt, GRPID);
2279
	if (def_mount_opts & EXT4_DEFM_UID16)
2280
		set_opt(sbi->s_mount_opt, NO_UID32);
T
Theodore Ts'o 已提交
2281
#ifdef CONFIG_EXT4_FS_XATTR
2282
	if (def_mount_opts & EXT4_DEFM_XATTR_USER)
2283
		set_opt(sbi->s_mount_opt, XATTR_USER);
2284
#endif
T
Theodore Ts'o 已提交
2285
#ifdef CONFIG_EXT4_FS_POSIX_ACL
2286
	if (def_mount_opts & EXT4_DEFM_ACL)
2287
		set_opt(sbi->s_mount_opt, POSIX_ACL);
2288
#endif
2289 2290 2291 2292 2293 2294 2295 2296
	if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_DATA)
		sbi->s_mount_opt |= EXT4_MOUNT_JOURNAL_DATA;
	else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_ORDERED)
		sbi->s_mount_opt |= EXT4_MOUNT_ORDERED_DATA;
	else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_WBACK)
		sbi->s_mount_opt |= EXT4_MOUNT_WRITEBACK_DATA;

	if (le16_to_cpu(sbi->s_es->s_errors) == EXT4_ERRORS_PANIC)
2297
		set_opt(sbi->s_mount_opt, ERRORS_PANIC);
2298
	else if (le16_to_cpu(sbi->s_es->s_errors) == EXT4_ERRORS_CONTINUE)
D
Dmitry Mishin 已提交
2299
		set_opt(sbi->s_mount_opt, ERRORS_CONT);
2300 2301
	else
		set_opt(sbi->s_mount_opt, ERRORS_RO);
2302 2303 2304

	sbi->s_resuid = le16_to_cpu(es->s_def_resuid);
	sbi->s_resgid = le16_to_cpu(es->s_def_resgid);
2305 2306 2307
	sbi->s_commit_interval = JBD2_DEFAULT_MAX_COMMIT_AGE * HZ;
	sbi->s_min_batch_time = EXT4_DEF_MIN_BATCH_TIME;
	sbi->s_max_batch_time = EXT4_DEF_MAX_BATCH_TIME;
2308 2309

	set_opt(sbi->s_mount_opt, RESERVATION);
E
Eric Sandeen 已提交
2310
	set_opt(sbi->s_mount_opt, BARRIER);
2311

2312 2313 2314 2315 2316 2317 2318
	/*
	 * enable delayed allocation by default
	 * Use -o nodelalloc to turn it off
	 */
	set_opt(sbi->s_mount_opt, DELALLOC);


2319 2320
	if (!parse_options((char *) data, sb, &journal_devnum,
			   &journal_ioprio, NULL, 0))
2321 2322 2323
		goto failed_mount;

	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
2324
		((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
2325

2326 2327 2328 2329
	if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV &&
	    (EXT4_HAS_COMPAT_FEATURE(sb, ~0U) ||
	     EXT4_HAS_RO_COMPAT_FEATURE(sb, ~0U) ||
	     EXT4_HAS_INCOMPAT_FEATURE(sb, ~0U)))
2330
		printk(KERN_WARNING
2331
		       "EXT4-fs warning: feature flags set on rev 0 fs, "
2332
		       "running e2fsck is recommended\n");
2333

2334 2335 2336 2337 2338
	/*
	 * Check feature flags regardless of the revision level, since we
	 * previously didn't change the revision level when setting the flags,
	 * so there is a chance incompat flags are set on a rev 0 filesystem.
	 */
2339
	features = EXT4_HAS_INCOMPAT_FEATURE(sb, ~EXT4_FEATURE_INCOMPAT_SUPP);
2340
	if (features) {
2341
		printk(KERN_ERR "EXT4-fs: %s: couldn't mount because of "
A
Aneesh Kumar K.V 已提交
2342 2343 2344
		       "unsupported optional features (%x).\n", sb->s_id,
			(le32_to_cpu(EXT4_SB(sb)->s_es->s_feature_incompat) &
			~EXT4_FEATURE_INCOMPAT_SUPP));
2345 2346
		goto failed_mount;
	}
2347
	features = EXT4_HAS_RO_COMPAT_FEATURE(sb, ~EXT4_FEATURE_RO_COMPAT_SUPP);
2348
	if (!(sb->s_flags & MS_RDONLY) && features) {
2349
		printk(KERN_ERR "EXT4-fs: %s: couldn't mount RDWR because of "
A
Aneesh Kumar K.V 已提交
2350 2351 2352
		       "unsupported optional features (%x).\n", sb->s_id,
			(le32_to_cpu(EXT4_SB(sb)->s_es->s_feature_ro_compat) &
			~EXT4_FEATURE_RO_COMPAT_SUPP));
2353 2354
		goto failed_mount;
	}
2355 2356 2357
	has_huge_files = EXT4_HAS_RO_COMPAT_FEATURE(sb,
				    EXT4_FEATURE_RO_COMPAT_HUGE_FILE);
	if (has_huge_files) {
2358 2359
		/*
		 * Large file size enabled file system can only be
J
Jens Axboe 已提交
2360
		 * mount if kernel is build with CONFIG_LBD
2361 2362 2363 2364 2365
		 */
		if (sizeof(root->i_blocks) < sizeof(u64) &&
				!(sb->s_flags & MS_RDONLY)) {
			printk(KERN_ERR "EXT4-fs: %s: Filesystem with huge "
					"files cannot be mounted read-write "
J
Jens Axboe 已提交
2366
					"without CONFIG_LBD.\n", sb->s_id);
2367 2368 2369
			goto failed_mount;
		}
	}
2370 2371
	blocksize = BLOCK_SIZE << le32_to_cpu(es->s_log_block_size);

2372 2373
	if (blocksize < EXT4_MIN_BLOCK_SIZE ||
	    blocksize > EXT4_MAX_BLOCK_SIZE) {
2374
		printk(KERN_ERR
2375
		       "EXT4-fs: Unsupported filesystem blocksize %d on %s.\n",
2376 2377 2378 2379 2380
		       blocksize, sb->s_id);
		goto failed_mount;
	}

	if (sb->s_blocksize != blocksize) {
2381 2382 2383 2384 2385

		/* Validate the filesystem blocksize */
		if (!sb_set_blocksize(sb, blocksize)) {
			printk(KERN_ERR "EXT4-fs: bad block size %d.\n",
					blocksize);
2386 2387 2388
			goto failed_mount;
		}

2389
		brelse(bh);
2390 2391 2392
		logical_sb_block = sb_block * EXT4_MIN_BLOCK_SIZE;
		offset = do_div(logical_sb_block, blocksize);
		bh = sb_bread(sb, logical_sb_block);
2393 2394
		if (!bh) {
			printk(KERN_ERR
2395
			       "EXT4-fs: Can't read superblock on 2nd try.\n");
2396 2397
			goto failed_mount;
		}
2398
		es = (struct ext4_super_block *)(((char *)bh->b_data) + offset);
2399
		sbi->s_es = es;
2400
		if (es->s_magic != cpu_to_le16(EXT4_SUPER_MAGIC)) {
2401 2402
			printk(KERN_ERR
			       "EXT4-fs: Magic mismatch, very weird !\n");
2403 2404 2405 2406
			goto failed_mount;
		}
	}

2407 2408 2409
	sbi->s_bitmap_maxbytes = ext4_max_bitmap_size(sb->s_blocksize_bits,
						      has_huge_files);
	sb->s_maxbytes = ext4_max_size(sb->s_blocksize_bits, has_huge_files);
2410

2411 2412 2413
	if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV) {
		sbi->s_inode_size = EXT4_GOOD_OLD_INODE_SIZE;
		sbi->s_first_ino = EXT4_GOOD_OLD_FIRST_INO;
2414 2415 2416
	} else {
		sbi->s_inode_size = le16_to_cpu(es->s_inode_size);
		sbi->s_first_ino = le32_to_cpu(es->s_first_ino);
2417
		if ((sbi->s_inode_size < EXT4_GOOD_OLD_INODE_SIZE) ||
V
Vignesh Babu 已提交
2418
		    (!is_power_of_2(sbi->s_inode_size)) ||
2419
		    (sbi->s_inode_size > blocksize)) {
2420 2421 2422
			printk(KERN_ERR
			       "EXT4-fs: unsupported inode size: %d\n",
			       sbi->s_inode_size);
2423 2424
			goto failed_mount;
		}
K
Kalpak Shah 已提交
2425 2426
		if (sbi->s_inode_size > EXT4_GOOD_OLD_INODE_SIZE)
			sb->s_time_gran = 1 << (EXT4_EPOCH_BITS - 2);
2427
	}
2428 2429
	sbi->s_desc_size = le16_to_cpu(es->s_desc_size);
	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_64BIT)) {
2430
		if (sbi->s_desc_size < EXT4_MIN_DESC_SIZE_64BIT ||
2431
		    sbi->s_desc_size > EXT4_MAX_DESC_SIZE ||
V
vignesh babu 已提交
2432
		    !is_power_of_2(sbi->s_desc_size)) {
2433
			printk(KERN_ERR
2434
			       "EXT4-fs: unsupported descriptor size %lu\n",
2435 2436 2437 2438 2439
			       sbi->s_desc_size);
			goto failed_mount;
		}
	} else
		sbi->s_desc_size = EXT4_MIN_DESC_SIZE;
2440 2441
	sbi->s_blocks_per_group = le32_to_cpu(es->s_blocks_per_group);
	sbi->s_inodes_per_group = le32_to_cpu(es->s_inodes_per_group);
2442
	if (EXT4_INODE_SIZE(sb) == 0 || EXT4_INODES_PER_GROUP(sb) == 0)
2443 2444
		goto cantfind_ext4;
	sbi->s_inodes_per_block = blocksize / EXT4_INODE_SIZE(sb);
2445
	if (sbi->s_inodes_per_block == 0)
2446
		goto cantfind_ext4;
2447 2448
	sbi->s_itb_per_group = sbi->s_inodes_per_group /
					sbi->s_inodes_per_block;
2449
	sbi->s_desc_per_block = blocksize / EXT4_DESC_SIZE(sb);
2450 2451
	sbi->s_sbh = bh;
	sbi->s_mount_state = le16_to_cpu(es->s_state);
2452 2453
	sbi->s_addr_per_block_bits = ilog2(EXT4_ADDR_PER_BLOCK(sb));
	sbi->s_desc_per_block_bits = ilog2(EXT4_DESC_PER_BLOCK(sb));
2454
	for (i = 0; i < 4; i++)
2455 2456
		sbi->s_hash_seed[i] = le32_to_cpu(es->s_hash_seed[i]);
	sbi->s_def_hash_version = es->s_def_hash_version;
2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468
	i = le32_to_cpu(es->s_flags);
	if (i & EXT2_FLAGS_UNSIGNED_HASH)
		sbi->s_hash_unsigned = 3;
	else if ((i & EXT2_FLAGS_SIGNED_HASH) == 0) {
#ifdef __CHAR_UNSIGNED__
		es->s_flags |= cpu_to_le32(EXT2_FLAGS_UNSIGNED_HASH);
		sbi->s_hash_unsigned = 3;
#else
		es->s_flags |= cpu_to_le32(EXT2_FLAGS_SIGNED_HASH);
#endif
		sb->s_dirt = 1;
	}
2469 2470

	if (sbi->s_blocks_per_group > blocksize * 8) {
2471 2472 2473
		printk(KERN_ERR
		       "EXT4-fs: #blocks per group too big: %lu\n",
		       sbi->s_blocks_per_group);
2474 2475 2476
		goto failed_mount;
	}
	if (sbi->s_inodes_per_group > blocksize * 8) {
2477 2478 2479
		printk(KERN_ERR
		       "EXT4-fs: #inodes per group too big: %lu\n",
		       sbi->s_inodes_per_group);
2480 2481 2482
		goto failed_mount;
	}

L
Laurent Vivier 已提交
2483
	if (ext4_blocks_count(es) >
2484
		    (sector_t)(~0ULL) >> (sb->s_blocksize_bits - 9)) {
2485
		printk(KERN_ERR "EXT4-fs: filesystem on %s:"
2486 2487
			" too large to mount safely\n", sb->s_id);
		if (sizeof(sector_t) < 8)
2488
			printk(KERN_WARNING "EXT4-fs: CONFIG_LBD not "
2489 2490 2491 2492
					"enabled\n");
		goto failed_mount;
	}

2493 2494
	if (EXT4_BLOCKS_PER_GROUP(sb) == 0)
		goto cantfind_ext4;
2495

2496 2497 2498 2499 2500 2501 2502 2503 2504
        /*
         * It makes no sense for the first data block to be beyond the end
         * of the filesystem.
         */
        if (le32_to_cpu(es->s_first_data_block) >= ext4_blocks_count(es)) {
                printk(KERN_WARNING "EXT4-fs: bad geometry: first data"
		       "block %u is beyond end of filesystem (%llu)\n",
		       le32_to_cpu(es->s_first_data_block),
		       ext4_blocks_count(es));
2505 2506
		goto failed_mount;
	}
L
Laurent Vivier 已提交
2507 2508 2509 2510
	blocks_count = (ext4_blocks_count(es) -
			le32_to_cpu(es->s_first_data_block) +
			EXT4_BLOCKS_PER_GROUP(sb) - 1);
	do_div(blocks_count, EXT4_BLOCKS_PER_GROUP(sb));
2511 2512 2513 2514 2515 2516 2517 2518 2519
	if (blocks_count > ((uint64_t)1<<32) - EXT4_DESC_PER_BLOCK(sb)) {
		printk(KERN_WARNING "EXT4-fs: groups count too large: %u "
		       "(block count %llu, first data block %u, "
		       "blocks per group %lu)\n", sbi->s_groups_count,
		       ext4_blocks_count(es),
		       le32_to_cpu(es->s_first_data_block),
		       EXT4_BLOCKS_PER_GROUP(sb));
		goto failed_mount;
	}
L
Laurent Vivier 已提交
2520
	sbi->s_groups_count = blocks_count;
2521 2522
	db_count = (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) - 1) /
		   EXT4_DESC_PER_BLOCK(sb);
2523
	sbi->s_group_desc = kmalloc(db_count * sizeof(struct buffer_head *),
2524 2525
				    GFP_KERNEL);
	if (sbi->s_group_desc == NULL) {
2526
		printk(KERN_ERR "EXT4-fs: not enough memory\n");
2527 2528 2529
		goto failed_mount;
	}

2530
#ifdef CONFIG_PROC_FS
2531 2532
	if (ext4_proc_root)
		sbi->s_proc = proc_mkdir(sb->s_id, ext4_proc_root);
2533
#endif
2534

2535
	bgl_lock_init(sbi->s_blockgroup_lock);
2536 2537

	for (i = 0; i < db_count; i++) {
2538
		block = descriptor_loc(sb, logical_sb_block, i);
2539 2540
		sbi->s_group_desc[i] = sb_bread(sb, block);
		if (!sbi->s_group_desc[i]) {
2541 2542
			printk(KERN_ERR "EXT4-fs: "
			       "can't read group descriptor %d\n", i);
2543 2544 2545 2546
			db_count = i;
			goto failed_mount2;
		}
	}
2547
	if (!ext4_check_descriptors(sb)) {
2548
		printk(KERN_ERR "EXT4-fs: group descriptors corrupted!\n");
2549 2550
		goto failed_mount2;
	}
2551 2552 2553 2554 2555 2556 2557 2558
	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG))
		if (!ext4_fill_flex_info(sb)) {
			printk(KERN_ERR
			       "EXT4-fs: unable to initialize "
			       "flex_bg meta info!\n");
			goto failed_mount2;
		}

2559 2560 2561 2562
	sbi->s_gdb_count = db_count;
	get_random_bytes(&sbi->s_next_generation, sizeof(u32));
	spin_lock_init(&sbi->s_next_gen_lock);

2563 2564 2565 2566 2567 2568 2569 2570 2571 2572
	err = percpu_counter_init(&sbi->s_freeblocks_counter,
			ext4_count_free_blocks(sb));
	if (!err) {
		err = percpu_counter_init(&sbi->s_freeinodes_counter,
				ext4_count_free_inodes(sb));
	}
	if (!err) {
		err = percpu_counter_init(&sbi->s_dirs_counter,
				ext4_count_dirs(sb));
	}
2573 2574 2575
	if (!err) {
		err = percpu_counter_init(&sbi->s_dirtyblocks_counter, 0);
	}
2576 2577 2578 2579
	if (err) {
		printk(KERN_ERR "EXT4-fs: insufficient memory\n");
		goto failed_mount3;
	}
2580

2581 2582
	sbi->s_stripe = ext4_get_stripe_size(sbi);

2583 2584 2585
	/*
	 * set up enough so that it can read an inode
	 */
2586 2587 2588
	sb->s_op = &ext4_sops;
	sb->s_export_op = &ext4_export_ops;
	sb->s_xattr = ext4_xattr_handlers;
2589
#ifdef CONFIG_QUOTA
2590 2591
	sb->s_qcop = &ext4_qctl_operations;
	sb->dq_op = &ext4_quota_operations;
2592 2593 2594 2595 2596 2597
#endif
	INIT_LIST_HEAD(&sbi->s_orphan); /* unlinked but open files */

	sb->s_root = NULL;

	needs_recovery = (es->s_last_orphan != 0 ||
2598 2599
			  EXT4_HAS_INCOMPAT_FEATURE(sb,
				    EXT4_FEATURE_INCOMPAT_RECOVER));
2600 2601 2602 2603 2604 2605

	/*
	 * The first inode we look at is the journal inode.  Don't try
	 * root first: it may be modified in the journal!
	 */
	if (!test_opt(sb, NOLOAD) &&
2606 2607
	    EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL)) {
		if (ext4_load_journal(sb, es, journal_devnum))
2608
			goto failed_mount3;
2609 2610 2611 2612
		if (!(sb->s_flags & MS_RDONLY) &&
		    EXT4_SB(sb)->s_journal->j_failed_commit) {
			printk(KERN_CRIT "EXT4-fs error (device %s): "
			       "ext4_fill_super: Journal transaction "
2613
			       "%u is corrupt\n", sb->s_id,
2614
			       EXT4_SB(sb)->s_journal->j_failed_commit);
2615 2616 2617
			if (test_opt(sb, ERRORS_RO)) {
				printk(KERN_CRIT
				       "Mounting filesystem read-only\n");
2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628
				sb->s_flags |= MS_RDONLY;
				EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
				es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
			}
			if (test_opt(sb, ERRORS_PANIC)) {
				EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
				es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
				ext4_commit_super(sb, es, 1);
				goto failed_mount4;
			}
		}
2629 2630 2631 2632 2633
	} else if (test_opt(sb, NOLOAD) && !(sb->s_flags & MS_RDONLY) &&
	      EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER)) {
		printk(KERN_ERR "EXT4-fs: required journal recovery "
		       "suppressed and not mounted read-only\n");
		goto failed_mount4;
2634
	} else {
2635 2636 2637 2638 2639
		clear_opt(sbi->s_mount_opt, DATA_FLAGS);
		set_opt(sbi->s_mount_opt, WRITEBACK_DATA);
		sbi->s_journal = NULL;
		needs_recovery = 0;
		goto no_journal;
2640 2641
	}

2642 2643 2644
	if (ext4_blocks_count(es) > 0xffffffffULL &&
	    !jbd2_journal_set_features(EXT4_SB(sb)->s_journal, 0, 0,
				       JBD2_FEATURE_INCOMPAT_64BIT)) {
2645
		printk(KERN_ERR "EXT4-fs: Failed to set 64-bit journal feature\n");
2646 2647 2648
		goto failed_mount4;
	}

2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663
	if (test_opt(sb, JOURNAL_ASYNC_COMMIT)) {
		jbd2_journal_set_features(sbi->s_journal,
				JBD2_FEATURE_COMPAT_CHECKSUM, 0,
				JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT);
	} else if (test_opt(sb, JOURNAL_CHECKSUM)) {
		jbd2_journal_set_features(sbi->s_journal,
				JBD2_FEATURE_COMPAT_CHECKSUM, 0, 0);
		jbd2_journal_clear_features(sbi->s_journal, 0, 0,
				JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT);
	} else {
		jbd2_journal_clear_features(sbi->s_journal,
				JBD2_FEATURE_COMPAT_CHECKSUM, 0,
				JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT);
	}

2664 2665 2666 2667 2668
	/* We have now updated the journal if required, so we can
	 * validate the data journaling mode. */
	switch (test_opt(sb, DATA_FLAGS)) {
	case 0:
		/* No mode set, assume a default based on the journal
A
Andrew Morton 已提交
2669 2670 2671
		 * capabilities: ORDERED_DATA if the journal can
		 * cope, else JOURNAL_DATA
		 */
2672 2673
		if (jbd2_journal_check_available_features
		    (sbi->s_journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE))
2674 2675 2676 2677 2678
			set_opt(sbi->s_mount_opt, ORDERED_DATA);
		else
			set_opt(sbi->s_mount_opt, JOURNAL_DATA);
		break;

2679 2680
	case EXT4_MOUNT_ORDERED_DATA:
	case EXT4_MOUNT_WRITEBACK_DATA:
2681 2682
		if (!jbd2_journal_check_available_features
		    (sbi->s_journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE)) {
2683
			printk(KERN_ERR "EXT4-fs: Journal does not support "
2684 2685 2686 2687 2688 2689
			       "requested data journaling mode\n");
			goto failed_mount4;
		}
	default:
		break;
	}
2690
	set_task_ioprio(sbi->s_journal->j_task, journal_ioprio);
2691

2692
no_journal:
2693 2694

	if (test_opt(sb, NOBH)) {
2695 2696
		if (!(test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)) {
			printk(KERN_WARNING "EXT4-fs: Ignoring nobh option - "
2697 2698 2699 2700 2701
				"its supported only with writeback mode\n");
			clear_opt(sbi->s_mount_opt, NOBH);
		}
	}
	/*
2702
	 * The jbd2_journal_load will have done any necessary log recovery,
2703 2704 2705
	 * so we can safely mount the rest of the filesystem now.
	 */

2706 2707
	root = ext4_iget(sb, EXT4_ROOT_INO);
	if (IS_ERR(root)) {
2708
		printk(KERN_ERR "EXT4-fs: get root inode failed\n");
2709
		ret = PTR_ERR(root);
2710 2711 2712
		goto failed_mount4;
	}
	if (!S_ISDIR(root->i_mode) || !root->i_blocks || !root->i_size) {
2713
		iput(root);
2714
		printk(KERN_ERR "EXT4-fs: corrupt root inode, run e2fsck\n");
2715 2716
		goto failed_mount4;
	}
2717 2718 2719 2720 2721 2722 2723
	sb->s_root = d_alloc_root(root);
	if (!sb->s_root) {
		printk(KERN_ERR "EXT4-fs: get root dentry failed\n");
		iput(root);
		ret = -ENOMEM;
		goto failed_mount4;
	}
2724

2725
	ext4_setup_super(sb, es, sb->s_flags & MS_RDONLY);
K
Kalpak Shah 已提交
2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751

	/* determine the minimum size of new large inodes, if present */
	if (sbi->s_inode_size > EXT4_GOOD_OLD_INODE_SIZE) {
		sbi->s_want_extra_isize = sizeof(struct ext4_inode) -
						     EXT4_GOOD_OLD_INODE_SIZE;
		if (EXT4_HAS_RO_COMPAT_FEATURE(sb,
				       EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE)) {
			if (sbi->s_want_extra_isize <
			    le16_to_cpu(es->s_want_extra_isize))
				sbi->s_want_extra_isize =
					le16_to_cpu(es->s_want_extra_isize);
			if (sbi->s_want_extra_isize <
			    le16_to_cpu(es->s_min_extra_isize))
				sbi->s_want_extra_isize =
					le16_to_cpu(es->s_min_extra_isize);
		}
	}
	/* Check if enough inode space is available */
	if (EXT4_GOOD_OLD_INODE_SIZE + sbi->s_want_extra_isize >
							sbi->s_inode_size) {
		sbi->s_want_extra_isize = sizeof(struct ext4_inode) -
						       EXT4_GOOD_OLD_INODE_SIZE;
		printk(KERN_INFO "EXT4-fs: required extra inode space not"
			"available.\n");
	}

2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766
	if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) {
		printk(KERN_WARNING "EXT4-fs: Ignoring delalloc option - "
				"requested data journaling mode\n");
		clear_opt(sbi->s_mount_opt, DELALLOC);
	} else if (test_opt(sb, DELALLOC))
		printk(KERN_INFO "EXT4-fs: delayed allocation enabled\n");

	ext4_ext_init(sb);
	err = ext4_mb_init(sb, needs_recovery);
	if (err) {
		printk(KERN_ERR "EXT4-fs: failed to initalize mballoc (%d)\n",
		       err);
		goto failed_mount4;
	}

T
Theodore Ts'o 已提交
2767 2768 2769 2770 2771 2772 2773 2774 2775 2776
	sbi->s_kobj.kset = ext4_kset;
	init_completion(&sbi->s_kobj_unregister);
	err = kobject_init_and_add(&sbi->s_kobj, &ext4_ktype, NULL,
				   "%s", sb->s_id);
	if (err) {
		ext4_mb_release(sb);
		ext4_ext_release(sb);
		goto failed_mount4;
	};

2777 2778 2779 2780 2781 2782 2783 2784
	/*
	 * akpm: core read_super() calls in here with the superblock locked.
	 * That deadlocks, because orphan cleanup needs to lock the superblock
	 * in numerous places.  Here we just pop the lock - it's relatively
	 * harmless, because we are now ready to accept write_super() requests,
	 * and aviro says that's the only reason for hanging onto the
	 * superblock lock.
	 */
2785 2786 2787
	EXT4_SB(sb)->s_mount_state |= EXT4_ORPHAN_FS;
	ext4_orphan_cleanup(sb, es);
	EXT4_SB(sb)->s_mount_state &= ~EXT4_ORPHAN_FS;
2788
	if (needs_recovery) {
2789
		printk(KERN_INFO "EXT4-fs: recovery complete.\n");
2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803
		ext4_mark_recovery_complete(sb, es);
	}
	if (EXT4_SB(sb)->s_journal) {
		if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA)
			descr = " journalled data mode";
		else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA)
			descr = " ordered data mode";
		else
			descr = " writeback data mode";
	} else
		descr = "out journal";

	printk(KERN_INFO "EXT4-fs: mounted filesystem %s with%s\n",
	       sb->s_id, descr);
2804 2805 2806 2807

	lock_kernel();
	return 0;

2808
cantfind_ext4:
2809
	if (!silent)
2810
		printk(KERN_ERR "VFS: Can't find ext4 filesystem on dev %s.\n",
2811 2812 2813 2814
		       sb->s_id);
	goto failed_mount;

failed_mount4:
2815 2816 2817 2818 2819
	printk(KERN_ERR "EXT4-fs (device %s): mount failed\n", sb->s_id);
	if (sbi->s_journal) {
		jbd2_journal_destroy(sbi->s_journal);
		sbi->s_journal = NULL;
	}
2820 2821 2822 2823
failed_mount3:
	percpu_counter_destroy(&sbi->s_freeblocks_counter);
	percpu_counter_destroy(&sbi->s_freeinodes_counter);
	percpu_counter_destroy(&sbi->s_dirs_counter);
2824
	percpu_counter_destroy(&sbi->s_dirtyblocks_counter);
2825 2826 2827 2828 2829
failed_mount2:
	for (i = 0; i < db_count; i++)
		brelse(sbi->s_group_desc[i]);
	kfree(sbi->s_group_desc);
failed_mount:
2830
	if (sbi->s_proc) {
2831
		remove_proc_entry(sb->s_id, ext4_proc_root);
2832
	}
2833 2834 2835 2836
#ifdef CONFIG_QUOTA
	for (i = 0; i < MAXQUOTAS; i++)
		kfree(sbi->s_qf_names[i]);
#endif
2837
	ext4_blkdev_remove(sbi);
2838 2839 2840 2841 2842
	brelse(bh);
out_fail:
	sb->s_fs_info = NULL;
	kfree(sbi);
	lock_kernel();
2843
	return ret;
2844 2845 2846 2847 2848 2849 2850
}

/*
 * Setup any per-fs journal parameters now.  We'll do this both on
 * initial mount, once the journal has been initialised but before we've
 * done any recovery; and again on any subsequent remount.
 */
2851
static void ext4_init_journal_params(struct super_block *sb, journal_t *journal)
2852
{
2853
	struct ext4_sb_info *sbi = EXT4_SB(sb);
2854

2855 2856 2857
	journal->j_commit_interval = sbi->s_commit_interval;
	journal->j_min_batch_time = sbi->s_min_batch_time;
	journal->j_max_batch_time = sbi->s_max_batch_time;
2858 2859 2860

	spin_lock(&journal->j_state_lock);
	if (test_opt(sb, BARRIER))
2861
		journal->j_flags |= JBD2_BARRIER;
2862
	else
2863
		journal->j_flags &= ~JBD2_BARRIER;
2864 2865 2866 2867
	if (test_opt(sb, DATA_ERR_ABORT))
		journal->j_flags |= JBD2_ABORT_ON_SYNCDATA_ERR;
	else
		journal->j_flags &= ~JBD2_ABORT_ON_SYNCDATA_ERR;
2868 2869 2870
	spin_unlock(&journal->j_state_lock);
}

2871
static journal_t *ext4_get_journal(struct super_block *sb,
2872 2873 2874 2875 2876
				   unsigned int journal_inum)
{
	struct inode *journal_inode;
	journal_t *journal;

2877 2878
	BUG_ON(!EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL));

2879 2880 2881 2882
	/* First, test for the existence of a valid inode on disk.  Bad
	 * things happen if we iget() an unused inode, as the subsequent
	 * iput() will try to delete it. */

2883 2884
	journal_inode = ext4_iget(sb, journal_inum);
	if (IS_ERR(journal_inode)) {
2885
		printk(KERN_ERR "EXT4-fs: no journal found.\n");
2886 2887 2888 2889 2890
		return NULL;
	}
	if (!journal_inode->i_nlink) {
		make_bad_inode(journal_inode);
		iput(journal_inode);
2891
		printk(KERN_ERR "EXT4-fs: journal inode is deleted.\n");
2892 2893 2894
		return NULL;
	}

2895
	jbd_debug(2, "Journal inode found at %p: %lld bytes\n",
2896
		  journal_inode, journal_inode->i_size);
2897
	if (!S_ISREG(journal_inode->i_mode)) {
2898
		printk(KERN_ERR "EXT4-fs: invalid journal inode.\n");
2899 2900 2901 2902
		iput(journal_inode);
		return NULL;
	}

2903
	journal = jbd2_journal_init_inode(journal_inode);
2904
	if (!journal) {
2905
		printk(KERN_ERR "EXT4-fs: Could not load journal inode\n");
2906 2907 2908 2909
		iput(journal_inode);
		return NULL;
	}
	journal->j_private = sb;
2910
	ext4_init_journal_params(sb, journal);
2911 2912 2913
	return journal;
}

2914
static journal_t *ext4_get_dev_journal(struct super_block *sb,
2915 2916
				       dev_t j_dev)
{
2917
	struct buffer_head *bh;
2918
	journal_t *journal;
2919 2920
	ext4_fsblk_t start;
	ext4_fsblk_t len;
2921
	int hblock, blocksize;
2922
	ext4_fsblk_t sb_block;
2923
	unsigned long offset;
2924
	struct ext4_super_block *es;
2925 2926
	struct block_device *bdev;

2927 2928
	BUG_ON(!EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL));

2929
	bdev = ext4_blkdev_get(j_dev);
2930 2931 2932 2933 2934
	if (bdev == NULL)
		return NULL;

	if (bd_claim(bdev, sb)) {
		printk(KERN_ERR
2935
			"EXT4-fs: failed to claim external journal device.\n");
A
Al Viro 已提交
2936
		blkdev_put(bdev, FMODE_READ|FMODE_WRITE);
2937 2938 2939 2940 2941 2942 2943
		return NULL;
	}

	blocksize = sb->s_blocksize;
	hblock = bdev_hardsect_size(bdev);
	if (blocksize < hblock) {
		printk(KERN_ERR
2944
			"EXT4-fs: blocksize too small for journal device.\n");
2945 2946 2947
		goto out_bdev;
	}

2948 2949
	sb_block = EXT4_MIN_BLOCK_SIZE / blocksize;
	offset = EXT4_MIN_BLOCK_SIZE % blocksize;
2950 2951
	set_blocksize(bdev, blocksize);
	if (!(bh = __bread(bdev, sb_block, blocksize))) {
2952
		printk(KERN_ERR "EXT4-fs: couldn't read superblock of "
2953 2954 2955 2956
		       "external journal\n");
		goto out_bdev;
	}

2957 2958
	es = (struct ext4_super_block *) (((char *)bh->b_data) + offset);
	if ((le16_to_cpu(es->s_magic) != EXT4_SUPER_MAGIC) ||
2959
	    !(le32_to_cpu(es->s_feature_incompat) &
2960 2961
	      EXT4_FEATURE_INCOMPAT_JOURNAL_DEV)) {
		printk(KERN_ERR "EXT4-fs: external journal has "
2962 2963 2964 2965 2966
					"bad superblock\n");
		brelse(bh);
		goto out_bdev;
	}

2967 2968
	if (memcmp(EXT4_SB(sb)->s_es->s_journal_uuid, es->s_uuid, 16)) {
		printk(KERN_ERR "EXT4-fs: journal UUID does not match\n");
2969 2970 2971 2972
		brelse(bh);
		goto out_bdev;
	}

L
Laurent Vivier 已提交
2973
	len = ext4_blocks_count(es);
2974 2975 2976
	start = sb_block + 1;
	brelse(bh);	/* we're done with the superblock */

2977
	journal = jbd2_journal_init_dev(bdev, sb->s_bdev,
2978 2979
					start, len, blocksize);
	if (!journal) {
2980
		printk(KERN_ERR "EXT4-fs: failed to create device journal\n");
2981 2982 2983 2984 2985 2986
		goto out_bdev;
	}
	journal->j_private = sb;
	ll_rw_block(READ, 1, &journal->j_sb_buffer);
	wait_on_buffer(journal->j_sb_buffer);
	if (!buffer_uptodate(journal->j_sb_buffer)) {
2987
		printk(KERN_ERR "EXT4-fs: I/O error on journal device\n");
2988 2989 2990
		goto out_journal;
	}
	if (be32_to_cpu(journal->j_superblock->s_nr_users) != 1) {
2991
		printk(KERN_ERR "EXT4-fs: External journal has more than one "
2992 2993 2994 2995
					"user (unsupported) - %d\n",
			be32_to_cpu(journal->j_superblock->s_nr_users));
		goto out_journal;
	}
2996 2997
	EXT4_SB(sb)->journal_bdev = bdev;
	ext4_init_journal_params(sb, journal);
2998 2999
	return journal;
out_journal:
3000
	jbd2_journal_destroy(journal);
3001
out_bdev:
3002
	ext4_blkdev_put(bdev);
3003 3004 3005
	return NULL;
}

3006 3007
static int ext4_load_journal(struct super_block *sb,
			     struct ext4_super_block *es,
3008 3009 3010 3011 3012 3013 3014 3015
			     unsigned long journal_devnum)
{
	journal_t *journal;
	unsigned int journal_inum = le32_to_cpu(es->s_journal_inum);
	dev_t journal_dev;
	int err = 0;
	int really_read_only;

3016 3017
	BUG_ON(!EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL));

3018 3019
	if (journal_devnum &&
	    journal_devnum != le32_to_cpu(es->s_journal_dev)) {
3020
		printk(KERN_INFO "EXT4-fs: external journal device major/minor "
3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033
			"numbers have changed\n");
		journal_dev = new_decode_dev(journal_devnum);
	} else
		journal_dev = new_decode_dev(le32_to_cpu(es->s_journal_dev));

	really_read_only = bdev_read_only(sb->s_bdev);

	/*
	 * Are we loading a blank journal or performing recovery after a
	 * crash?  For recovery, we need to check in advance whether we
	 * can get read-write access to the device.
	 */

3034
	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER)) {
3035
		if (sb->s_flags & MS_RDONLY) {
3036
			printk(KERN_INFO "EXT4-fs: INFO: recovery "
3037 3038
					"required on readonly filesystem.\n");
			if (really_read_only) {
3039
				printk(KERN_ERR "EXT4-fs: write access "
3040 3041 3042
					"unavailable, cannot proceed.\n");
				return -EROFS;
			}
3043 3044
			printk(KERN_INFO "EXT4-fs: write access will "
			       "be enabled during recovery.\n");
3045 3046 3047 3048
		}
	}

	if (journal_inum && journal_dev) {
3049
		printk(KERN_ERR "EXT4-fs: filesystem has both journal "
3050 3051 3052 3053 3054
		       "and inode journals!\n");
		return -EINVAL;
	}

	if (journal_inum) {
3055
		if (!(journal = ext4_get_journal(sb, journal_inum)))
3056 3057
			return -EINVAL;
	} else {
3058
		if (!(journal = ext4_get_dev_journal(sb, journal_dev)))
3059 3060 3061
			return -EINVAL;
	}

3062 3063 3064 3065 3066
	if (journal->j_flags & JBD2_BARRIER)
		printk(KERN_INFO "EXT4-fs: barriers enabled\n");
	else
		printk(KERN_INFO "EXT4-fs: barriers disabled\n");

3067
	if (!really_read_only && test_opt(sb, UPDATE_JOURNAL)) {
3068
		err = jbd2_journal_update_format(journal);
3069
		if (err)  {
3070
			printk(KERN_ERR "EXT4-fs: error updating journal.\n");
3071
			jbd2_journal_destroy(journal);
3072 3073 3074 3075
			return err;
		}
	}

3076
	if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER))
3077
		err = jbd2_journal_wipe(journal, !really_read_only);
3078
	if (!err)
3079
		err = jbd2_journal_load(journal);
3080 3081

	if (err) {
3082
		printk(KERN_ERR "EXT4-fs: error loading journal.\n");
3083
		jbd2_journal_destroy(journal);
3084 3085 3086
		return err;
	}

3087 3088
	EXT4_SB(sb)->s_journal = journal;
	ext4_clear_journal_err(sb, es);
3089 3090 3091 3092 3093 3094 3095

	if (journal_devnum &&
	    journal_devnum != le32_to_cpu(es->s_journal_dev)) {
		es->s_journal_dev = cpu_to_le32(journal_devnum);
		sb->s_dirt = 1;

		/* Make sure we flush the recovery flag to disk. */
3096
		ext4_commit_super(sb, es, 1);
3097 3098 3099 3100 3101
	}

	return 0;
}

3102
static int ext4_commit_super(struct super_block *sb,
3103
			      struct ext4_super_block *es, int sync)
3104
{
3105
	struct buffer_head *sbh = EXT4_SB(sb)->s_sbh;
3106
	int error = 0;
3107 3108

	if (!sbh)
3109
		return error;
3110 3111 3112 3113 3114 3115 3116 3117 3118
	if (buffer_write_io_error(sbh)) {
		/*
		 * Oh, dear.  A previous attempt to write the
		 * superblock failed.  This could happen because the
		 * USB device was yanked out.  Or it could happen to
		 * be a transient write error and maybe the block will
		 * be remapped.  Nothing we can do but to retry the
		 * write and hope for the best.
		 */
3119
		printk(KERN_ERR "EXT4-fs: previous I/O error to "
3120 3121 3122 3123
		       "superblock detected for %s.\n", sb->s_id);
		clear_buffer_write_io_error(sbh);
		set_buffer_uptodate(sbh);
	}
3124
	es->s_wtime = cpu_to_le32(get_seconds());
T
Theodore Ts'o 已提交
3125 3126 3127 3128
	es->s_kbytes_written =
		cpu_to_le64(EXT4_SB(sb)->s_kbytes_written + 
			    ((part_stat_read(sb->s_bdev->bd_part, sectors[1]) -
			      EXT4_SB(sb)->s_sectors_written_start) >> 1));
3129 3130 3131 3132 3133
	ext4_free_blocks_count_set(es, percpu_counter_sum_positive(
					&EXT4_SB(sb)->s_freeblocks_counter));
	es->s_free_inodes_count = cpu_to_le32(percpu_counter_sum_positive(
					&EXT4_SB(sb)->s_freeinodes_counter));

3134 3135
	BUFFER_TRACE(sbh, "marking dirty");
	mark_buffer_dirty(sbh);
3136
	if (sync) {
3137 3138 3139 3140 3141 3142
		error = sync_dirty_buffer(sbh);
		if (error)
			return error;

		error = buffer_write_io_error(sbh);
		if (error) {
3143
			printk(KERN_ERR "EXT4-fs: I/O error while writing "
3144 3145 3146 3147 3148
			       "superblock for %s.\n", sb->s_id);
			clear_buffer_write_io_error(sbh);
			set_buffer_uptodate(sbh);
		}
	}
3149
	return error;
3150 3151 3152 3153 3154 3155 3156 3157
}


/*
 * Have we just finished recovery?  If so, and if we are mounting (or
 * remounting) the filesystem readonly, then we will end up with a
 * consistent fs on disk.  Record that fact.
 */
3158 3159
static void ext4_mark_recovery_complete(struct super_block *sb,
					struct ext4_super_block *es)
3160
{
3161
	journal_t *journal = EXT4_SB(sb)->s_journal;
3162

3163 3164 3165 3166
	if (!EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL)) {
		BUG_ON(journal != NULL);
		return;
	}
3167
	jbd2_journal_lock_updates(journal);
3168 3169 3170
	if (jbd2_journal_flush(journal) < 0)
		goto out;

3171
	lock_super(sb);
3172
	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER) &&
3173
	    sb->s_flags & MS_RDONLY) {
3174
		EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
3175
		sb->s_dirt = 0;
3176
		ext4_commit_super(sb, es, 1);
3177
	}
3178
	unlock_super(sb);
3179 3180

out:
3181
	jbd2_journal_unlock_updates(journal);
3182 3183 3184 3185 3186 3187 3188
}

/*
 * If we are mounting (or read-write remounting) a filesystem whose journal
 * has recorded an error from a previous lifetime, move that error to the
 * main filesystem now.
 */
3189 3190
static void ext4_clear_journal_err(struct super_block *sb,
				   struct ext4_super_block *es)
3191 3192 3193 3194 3195
{
	journal_t *journal;
	int j_errno;
	const char *errstr;

3196 3197
	BUG_ON(!EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL));

3198
	journal = EXT4_SB(sb)->s_journal;
3199 3200 3201

	/*
	 * Now check for any error status which may have been recorded in the
3202
	 * journal by a prior ext4_error() or ext4_abort()
3203 3204
	 */

3205
	j_errno = jbd2_journal_errno(journal);
3206 3207 3208
	if (j_errno) {
		char nbuf[16];

3209
		errstr = ext4_decode_error(sb, j_errno, nbuf);
3210
		ext4_warning(sb, __func__, "Filesystem error recorded "
3211
			     "from previous mount: %s", errstr);
3212
		ext4_warning(sb, __func__, "Marking fs in need of "
3213 3214
			     "filesystem check.");

3215 3216
		EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
		es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
3217
		ext4_commit_super(sb, es, 1);
3218

3219
		jbd2_journal_clear_err(journal);
3220 3221 3222 3223 3224 3225 3226
	}
}

/*
 * Force the running and committing transactions to commit,
 * and wait on the commit.
 */
3227
int ext4_force_commit(struct super_block *sb)
3228 3229
{
	journal_t *journal;
3230
	int ret = 0;
3231 3232 3233 3234

	if (sb->s_flags & MS_RDONLY)
		return 0;

3235
	journal = EXT4_SB(sb)->s_journal;
3236 3237 3238 3239 3240
	if (journal) {
		sb->s_dirt = 0;
		ret = ext4_journal_force_commit(journal);
	}

3241 3242 3243 3244
	return ret;
}

/*
3245
 * Ext4 always journals updates to the superblock itself, so we don't
3246
 * have to propagate any other updates to the superblock on disk at this
3247 3248
 * point.  (We can probably nuke this function altogether, and remove
 * any mention to sb->s_dirt in all of fs/ext4; eventual cleanup...)
3249
 */
3250
static void ext4_write_super(struct super_block *sb)
3251
{
3252 3253 3254 3255 3256 3257 3258
	if (EXT4_SB(sb)->s_journal) {
		if (mutex_trylock(&sb->s_lock) != 0)
			BUG();
		sb->s_dirt = 0;
	} else {
		ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1);
	}
3259 3260
}

3261
static int ext4_sync_fs(struct super_block *sb, int wait)
3262
{
3263
	int ret = 0;
3264
	tid_t target;
3265

3266
	trace_mark(ext4_sync_fs, "dev %s wait %d", sb->s_id, wait);
3267
	sb->s_dirt = 0;
3268
	if (EXT4_SB(sb)->s_journal) {
3269 3270 3271 3272 3273 3274
		if (jbd2_journal_start_commit(EXT4_SB(sb)->s_journal,
					      &target)) {
			if (wait)
				jbd2_log_wait_commit(EXT4_SB(sb)->s_journal,
						     target);
		}
3275 3276 3277
	} else {
		ext4_commit_super(sb, EXT4_SB(sb)->s_es, wait);
	}
3278
	return ret;
3279 3280 3281 3282 3283 3284
}

/*
 * LVM calls this function before a (read-only) snapshot is created.  This
 * gives us a chance to flush the journal completely and mark the fs clean.
 */
3285
static int ext4_freeze(struct super_block *sb)
3286
{
3287 3288
	int error = 0;
	journal_t *journal;
3289 3290 3291
	sb->s_dirt = 0;

	if (!(sb->s_flags & MS_RDONLY)) {
3292
		journal = EXT4_SB(sb)->s_journal;
3293

3294 3295 3296
		if (journal) {
			/* Now we set up the journal barrier. */
			jbd2_journal_lock_updates(journal);
3297

3298 3299 3300 3301
			/*
			 * We don't want to clear needs_recovery flag when we
			 * failed to flush the journal.
			 */
3302 3303 3304
			error = jbd2_journal_flush(journal);
			if (error < 0)
				goto out;
3305
		}
3306 3307

		/* Journal blocked and flushed, clear needs_recovery flag. */
3308
		EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
3309 3310 3311
		error = ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1);
		if (error)
			goto out;
3312
	}
3313 3314 3315 3316
	return 0;
out:
	jbd2_journal_unlock_updates(journal);
	return error;
3317 3318 3319 3320 3321 3322
}

/*
 * Called by LVM after the snapshot is done.  We need to reset the RECOVER
 * flag here, even though the filesystem is not technically dirty yet.
 */
3323
static int ext4_unfreeze(struct super_block *sb)
3324
{
3325
	if (EXT4_SB(sb)->s_journal && !(sb->s_flags & MS_RDONLY)) {
3326 3327
		lock_super(sb);
		/* Reser the needs_recovery flag before the fs is unlocked. */
3328 3329
		EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
		ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1);
3330
		unlock_super(sb);
3331
		jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal);
3332
	}
3333
	return 0;
3334 3335
}

3336
static int ext4_remount(struct super_block *sb, int *flags, char *data)
3337
{
3338
	struct ext4_super_block *es;
3339 3340
	struct ext4_sb_info *sbi = EXT4_SB(sb);
	ext4_fsblk_t n_blocks_count = 0;
3341
	unsigned long old_sb_flags;
3342
	struct ext4_mount_options old_opts;
3343
	ext4_group_t g;
3344
	unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO;
3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355
	int err;
#ifdef CONFIG_QUOTA
	int i;
#endif

	/* Store the original options */
	old_sb_flags = sb->s_flags;
	old_opts.s_mount_opt = sbi->s_mount_opt;
	old_opts.s_resuid = sbi->s_resuid;
	old_opts.s_resgid = sbi->s_resgid;
	old_opts.s_commit_interval = sbi->s_commit_interval;
3356 3357
	old_opts.s_min_batch_time = sbi->s_min_batch_time;
	old_opts.s_max_batch_time = sbi->s_max_batch_time;
3358 3359 3360 3361 3362
#ifdef CONFIG_QUOTA
	old_opts.s_jquota_fmt = sbi->s_jquota_fmt;
	for (i = 0; i < MAXQUOTAS; i++)
		old_opts.s_qf_names[i] = sbi->s_qf_names[i];
#endif
3363 3364
	if (sbi->s_journal && sbi->s_journal->j_task->io_context)
		journal_ioprio = sbi->s_journal->j_task->io_context->ioprio;
3365 3366 3367 3368

	/*
	 * Allow the "check" option to be passed as a remount option.
	 */
3369 3370
	if (!parse_options(data, sb, NULL, &journal_ioprio,
			   &n_blocks_count, 1)) {
3371 3372 3373 3374
		err = -EINVAL;
		goto restore_opts;
	}

3375
	if (sbi->s_mount_opt & EXT4_MOUNT_ABORT)
3376
		ext4_abort(sb, __func__, "Abort forced by user");
3377 3378

	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
3379
		((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
3380 3381 3382

	es = sbi->s_es;

3383
	if (sbi->s_journal) {
3384
		ext4_init_journal_params(sb, sbi->s_journal);
3385 3386
		set_task_ioprio(sbi->s_journal->j_task, journal_ioprio);
	}
3387 3388

	if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY) ||
L
Laurent Vivier 已提交
3389
		n_blocks_count > ext4_blocks_count(es)) {
3390
		if (sbi->s_mount_opt & EXT4_MOUNT_ABORT) {
3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406
			err = -EROFS;
			goto restore_opts;
		}

		if (*flags & MS_RDONLY) {
			/*
			 * First of all, the unconditional stuff we have to do
			 * to disable replay of the journal when we next remount
			 */
			sb->s_flags |= MS_RDONLY;

			/*
			 * OK, test if we are remounting a valid rw partition
			 * readonly, and if so set the rdonly flag and then
			 * mark the partition as valid again.
			 */
3407 3408
			if (!(es->s_state & cpu_to_le16(EXT4_VALID_FS)) &&
			    (sbi->s_mount_state & EXT4_VALID_FS))
3409 3410
				es->s_state = cpu_to_le16(sbi->s_mount_state);

3411 3412 3413 3414
			/*
			 * We have to unlock super so that we can wait for
			 * transactions.
			 */
3415 3416 3417 3418 3419
			if (sbi->s_journal) {
				unlock_super(sb);
				ext4_mark_recovery_complete(sb, es);
				lock_super(sb);
			}
3420
		} else {
A
Aneesh Kumar K.V 已提交
3421
			int ret;
3422 3423 3424
			if ((ret = EXT4_HAS_RO_COMPAT_FEATURE(sb,
					~EXT4_FEATURE_RO_COMPAT_SUPP))) {
				printk(KERN_WARNING "EXT4-fs: %s: couldn't "
3425
				       "remount RDWR because of unsupported "
A
Aneesh Kumar K.V 已提交
3426 3427 3428
				       "optional features (%x).\n", sb->s_id,
				(le32_to_cpu(sbi->s_es->s_feature_ro_compat) &
					~EXT4_FEATURE_RO_COMPAT_SUPP));
3429 3430 3431
				err = -EROFS;
				goto restore_opts;
			}
3432

3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444
			/*
			 * Make sure the group descriptor checksums
			 * are sane.  If they aren't, refuse to
			 * remount r/w.
			 */
			for (g = 0; g < sbi->s_groups_count; g++) {
				struct ext4_group_desc *gdp =
					ext4_get_group_desc(sb, g, NULL);

				if (!ext4_group_desc_csum_verify(sbi, g, gdp)) {
					printk(KERN_ERR
	       "EXT4-fs: ext4_remount: "
3445
		"Checksum for group %u failed (%u!=%u)\n",
3446 3447 3448 3449 3450 3451 3452
		g, le16_to_cpu(ext4_group_desc_csum(sbi, g, gdp)),
					       le16_to_cpu(gdp->bg_checksum));
					err = -EINVAL;
					goto restore_opts;
				}
			}

3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467
			/*
			 * If we have an unprocessed orphan list hanging
			 * around from a previously readonly bdev mount,
			 * require a full umount/remount for now.
			 */
			if (es->s_last_orphan) {
				printk(KERN_WARNING "EXT4-fs: %s: couldn't "
				       "remount RDWR because of unprocessed "
				       "orphan inode list.  Please "
				       "umount/remount instead.\n",
				       sb->s_id);
				err = -EINVAL;
				goto restore_opts;
			}

3468 3469 3470 3471 3472 3473
			/*
			 * Mounting a RDONLY partition read-write, so reread
			 * and store the current valid flag.  (It may have
			 * been changed by e2fsck since we originally mounted
			 * the partition.)
			 */
3474 3475
			if (sbi->s_journal)
				ext4_clear_journal_err(sb, es);
3476
			sbi->s_mount_state = le16_to_cpu(es->s_state);
3477
			if ((err = ext4_group_extend(sb, es, n_blocks_count)))
3478
				goto restore_opts;
3479
			if (!ext4_setup_super(sb, es, 0))
3480 3481 3482
				sb->s_flags &= ~MS_RDONLY;
		}
	}
3483 3484 3485
	if (sbi->s_journal == NULL)
		ext4_commit_super(sb, es, 1);

3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499
#ifdef CONFIG_QUOTA
	/* Release old quota file names */
	for (i = 0; i < MAXQUOTAS; i++)
		if (old_opts.s_qf_names[i] &&
		    old_opts.s_qf_names[i] != sbi->s_qf_names[i])
			kfree(old_opts.s_qf_names[i]);
#endif
	return 0;
restore_opts:
	sb->s_flags = old_sb_flags;
	sbi->s_mount_opt = old_opts.s_mount_opt;
	sbi->s_resuid = old_opts.s_resuid;
	sbi->s_resgid = old_opts.s_resgid;
	sbi->s_commit_interval = old_opts.s_commit_interval;
3500 3501
	sbi->s_min_batch_time = old_opts.s_min_batch_time;
	sbi->s_max_batch_time = old_opts.s_max_batch_time;
3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513
#ifdef CONFIG_QUOTA
	sbi->s_jquota_fmt = old_opts.s_jquota_fmt;
	for (i = 0; i < MAXQUOTAS; i++) {
		if (sbi->s_qf_names[i] &&
		    old_opts.s_qf_names[i] != sbi->s_qf_names[i])
			kfree(sbi->s_qf_names[i]);
		sbi->s_qf_names[i] = old_opts.s_qf_names[i];
	}
#endif
	return err;
}

3514
static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf)
3515 3516
{
	struct super_block *sb = dentry->d_sb;
3517 3518
	struct ext4_sb_info *sbi = EXT4_SB(sb);
	struct ext4_super_block *es = sbi->s_es;
P
Pekka Enberg 已提交
3519
	u64 fsid;
3520

B
Badari Pulavarty 已提交
3521 3522
	if (test_opt(sb, MINIX_DF)) {
		sbi->s_overhead_last = 0;
3523
	} else if (sbi->s_blocks_last != ext4_blocks_count(es)) {
3524
		ext4_group_t ngroups = sbi->s_groups_count, i;
B
Badari Pulavarty 已提交
3525
		ext4_fsblk_t overhead = 0;
3526 3527 3528
		smp_rmb();

		/*
B
Badari Pulavarty 已提交
3529 3530 3531
		 * Compute the overhead (FS structures).  This is constant
		 * for a given filesystem unless the number of block groups
		 * changes so we cache the previous value until it does.
3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545
		 */

		/*
		 * All of the blocks before first_data_block are
		 * overhead
		 */
		overhead = le32_to_cpu(es->s_first_data_block);

		/*
		 * Add the overhead attributed to the superblock and
		 * block group descriptors.  If the sparse superblocks
		 * feature is turned on, then not all groups have this.
		 */
		for (i = 0; i < ngroups; i++) {
3546 3547
			overhead += ext4_bg_has_super(sb, i) +
				ext4_bg_num_gdb(sb, i);
3548 3549 3550 3551 3552 3553 3554
			cond_resched();
		}

		/*
		 * Every block group has an inode bitmap, a block
		 * bitmap, and an inode table.
		 */
B
Badari Pulavarty 已提交
3555 3556 3557
		overhead += ngroups * (2 + sbi->s_itb_per_group);
		sbi->s_overhead_last = overhead;
		smp_wmb();
3558
		sbi->s_blocks_last = ext4_blocks_count(es);
3559 3560
	}

3561
	buf->f_type = EXT4_SUPER_MAGIC;
3562
	buf->f_bsize = sb->s_blocksize;
B
Badari Pulavarty 已提交
3563
	buf->f_blocks = ext4_blocks_count(es) - sbi->s_overhead_last;
3564 3565
	buf->f_bfree = percpu_counter_sum_positive(&sbi->s_freeblocks_counter) -
		       percpu_counter_sum_positive(&sbi->s_dirtyblocks_counter);
3566
	ext4_free_blocks_count_set(es, buf->f_bfree);
L
Laurent Vivier 已提交
3567 3568
	buf->f_bavail = buf->f_bfree - ext4_r_blocks_count(es);
	if (buf->f_bfree < ext4_r_blocks_count(es))
3569 3570
		buf->f_bavail = 0;
	buf->f_files = le32_to_cpu(es->s_inodes_count);
P
Peter Zijlstra 已提交
3571
	buf->f_ffree = percpu_counter_sum_positive(&sbi->s_freeinodes_counter);
B
Badari Pulavarty 已提交
3572
	es->s_free_inodes_count = cpu_to_le32(buf->f_ffree);
3573
	buf->f_namelen = EXT4_NAME_LEN;
P
Pekka Enberg 已提交
3574 3575 3576 3577
	fsid = le64_to_cpup((void *)es->s_uuid) ^
	       le64_to_cpup((void *)es->s_uuid + sizeof(u64));
	buf->f_fsid.val[0] = fsid & 0xFFFFFFFFUL;
	buf->f_fsid.val[1] = (fsid >> 32) & 0xFFFFFFFFUL;
3578 3579 3580 3581 3582 3583
	return 0;
}

/* Helper function for writing quotas on sync - we need to start transaction before quota file
 * is locked for write. Otherwise the are possible deadlocks:
 * Process 1                         Process 2
3584
 * ext4_create()                     quota_sync()
3585 3586
 *   jbd2_journal_start()                  write_dquot()
 *   vfs_dq_init()                         down(dqio_mutex)
3587
 *     down(dqio_mutex)                    jbd2_journal_start()
3588 3589 3590 3591 3592 3593 3594 3595 3596 3597
 *
 */

#ifdef CONFIG_QUOTA

static inline struct inode *dquot_to_inode(struct dquot *dquot)
{
	return sb_dqopt(dquot->dq_sb)->files[dquot->dq_type];
}

3598
static int ext4_write_dquot(struct dquot *dquot)
3599 3600 3601 3602 3603 3604
{
	int ret, err;
	handle_t *handle;
	struct inode *inode;

	inode = dquot_to_inode(dquot);
3605 3606
	handle = ext4_journal_start(inode,
					EXT4_QUOTA_TRANS_BLOCKS(dquot->dq_sb));
3607 3608 3609
	if (IS_ERR(handle))
		return PTR_ERR(handle);
	ret = dquot_commit(dquot);
3610
	err = ext4_journal_stop(handle);
3611 3612 3613 3614 3615
	if (!ret)
		ret = err;
	return ret;
}

3616
static int ext4_acquire_dquot(struct dquot *dquot)
3617 3618 3619 3620
{
	int ret, err;
	handle_t *handle;

3621 3622
	handle = ext4_journal_start(dquot_to_inode(dquot),
					EXT4_QUOTA_INIT_BLOCKS(dquot->dq_sb));
3623 3624 3625
	if (IS_ERR(handle))
		return PTR_ERR(handle);
	ret = dquot_acquire(dquot);
3626
	err = ext4_journal_stop(handle);
3627 3628 3629 3630 3631
	if (!ret)
		ret = err;
	return ret;
}

3632
static int ext4_release_dquot(struct dquot *dquot)
3633 3634 3635 3636
{
	int ret, err;
	handle_t *handle;

3637 3638
	handle = ext4_journal_start(dquot_to_inode(dquot),
					EXT4_QUOTA_DEL_BLOCKS(dquot->dq_sb));
J
Jan Kara 已提交
3639 3640 3641
	if (IS_ERR(handle)) {
		/* Release dquot anyway to avoid endless cycle in dqput() */
		dquot_release(dquot);
3642
		return PTR_ERR(handle);
J
Jan Kara 已提交
3643
	}
3644
	ret = dquot_release(dquot);
3645
	err = ext4_journal_stop(handle);
3646 3647 3648 3649 3650
	if (!ret)
		ret = err;
	return ret;
}

3651
static int ext4_mark_dquot_dirty(struct dquot *dquot)
3652
{
3653
	/* Are we journaling quotas? */
3654 3655
	if (EXT4_SB(dquot->dq_sb)->s_qf_names[USRQUOTA] ||
	    EXT4_SB(dquot->dq_sb)->s_qf_names[GRPQUOTA]) {
3656
		dquot_mark_dquot_dirty(dquot);
3657
		return ext4_write_dquot(dquot);
3658 3659 3660 3661 3662
	} else {
		return dquot_mark_dquot_dirty(dquot);
	}
}

3663
static int ext4_write_info(struct super_block *sb, int type)
3664 3665 3666 3667 3668
{
	int ret, err;
	handle_t *handle;

	/* Data block + inode block */
3669
	handle = ext4_journal_start(sb->s_root->d_inode, 2);
3670 3671 3672
	if (IS_ERR(handle))
		return PTR_ERR(handle);
	ret = dquot_commit_info(sb, type);
3673
	err = ext4_journal_stop(handle);
3674 3675 3676 3677 3678 3679 3680 3681 3682
	if (!ret)
		ret = err;
	return ret;
}

/*
 * Turn on quotas during mount time - we need to find
 * the quota file and such...
 */
3683
static int ext4_quota_on_mount(struct super_block *sb, int type)
3684
{
3685 3686
	return vfs_quota_on_mount(sb, EXT4_SB(sb)->s_qf_names[type],
			EXT4_SB(sb)->s_jquota_fmt, type);
3687 3688 3689 3690 3691
}

/*
 * Standard function to be called on quota_on
 */
3692
static int ext4_quota_on(struct super_block *sb, int type, int format_id,
3693
			 char *name, int remount)
3694 3695
{
	int err;
3696
	struct path path;
3697 3698 3699

	if (!test_opt(sb, QUOTA))
		return -EINVAL;
3700
	/* When remounting, no checks are needed and in fact, name is NULL */
3701
	if (remount)
3702
		return vfs_quota_on(sb, type, format_id, name, remount);
3703

3704
	err = kern_path(name, LOOKUP_FOLLOW, &path);
3705 3706
	if (err)
		return err;
3707

3708
	/* Quotafile not on the same filesystem? */
3709 3710
	if (path.mnt->mnt_sb != sb) {
		path_put(&path);
3711 3712
		return -EXDEV;
	}
3713 3714
	/* Journaling quota? */
	if (EXT4_SB(sb)->s_qf_names[type]) {
3715
		/* Quotafile not in fs root? */
3716
		if (path.dentry->d_parent != sb->s_root)
3717 3718 3719
			printk(KERN_WARNING
				"EXT4-fs: Quota file not on filesystem root. "
				"Journaled quota will not work.\n");
3720
	}
3721 3722 3723 3724 3725

	/*
	 * When we journal data on quota file, we have to flush journal to see
	 * all updates to the file when we bypass pagecache...
	 */
3726 3727
	if (EXT4_SB(sb)->s_journal &&
	    ext4_should_journal_data(path.dentry->d_inode)) {
3728 3729 3730 3731 3732
		/*
		 * We don't need to lock updates but journal_flush() could
		 * otherwise be livelocked...
		 */
		jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal);
3733
		err = jbd2_journal_flush(EXT4_SB(sb)->s_journal);
3734
		jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal);
3735
		if (err) {
3736
			path_put(&path);
3737 3738
			return err;
		}
3739 3740
	}

3741 3742
	err = vfs_quota_on_path(sb, type, format_id, &path);
	path_put(&path);
3743
	return err;
3744 3745 3746 3747 3748 3749
}

/* Read data from quotafile - avoid pagecache and such because we cannot afford
 * acquiring the locks... As quota files are never truncated and quota code
 * itself serializes the operations (and noone else should touch the files)
 * we don't have to be afraid of races */
3750
static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data,
3751 3752 3753
			       size_t len, loff_t off)
{
	struct inode *inode = sb_dqopt(sb)->files[type];
A
Aneesh Kumar K.V 已提交
3754
	ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb);
3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769
	int err = 0;
	int offset = off & (sb->s_blocksize - 1);
	int tocopy;
	size_t toread;
	struct buffer_head *bh;
	loff_t i_size = i_size_read(inode);

	if (off > i_size)
		return 0;
	if (off+len > i_size)
		len = i_size-off;
	toread = len;
	while (toread > 0) {
		tocopy = sb->s_blocksize - offset < toread ?
				sb->s_blocksize - offset : toread;
3770
		bh = ext4_bread(NULL, inode, blk, 0, &err);
3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787
		if (err)
			return err;
		if (!bh)	/* A hole? */
			memset(data, 0, tocopy);
		else
			memcpy(data, bh->b_data+offset, tocopy);
		brelse(bh);
		offset = 0;
		toread -= tocopy;
		data += tocopy;
		blk++;
	}
	return len;
}

/* Write to quotafile (we know the transaction is already started and has
 * enough credits) */
3788
static ssize_t ext4_quota_write(struct super_block *sb, int type,
3789 3790 3791
				const char *data, size_t len, loff_t off)
{
	struct inode *inode = sb_dqopt(sb)->files[type];
A
Aneesh Kumar K.V 已提交
3792
	ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb);
3793 3794 3795
	int err = 0;
	int offset = off & (sb->s_blocksize - 1);
	int tocopy;
3796
	int journal_quota = EXT4_SB(sb)->s_qf_names[type] != NULL;
3797 3798 3799 3800
	size_t towrite = len;
	struct buffer_head *bh;
	handle_t *handle = journal_current_handle();

3801
	if (EXT4_SB(sb)->s_journal && !handle) {
3802
		printk(KERN_WARNING "EXT4-fs: Quota write (off=%llu, len=%llu)"
J
Jan Kara 已提交
3803 3804 3805 3806
			" cancelled because transaction is not started.\n",
			(unsigned long long)off, (unsigned long long)len);
		return -EIO;
	}
3807 3808 3809 3810
	mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA);
	while (towrite > 0) {
		tocopy = sb->s_blocksize - offset < towrite ?
				sb->s_blocksize - offset : towrite;
3811
		bh = ext4_bread(handle, inode, blk, 1, &err);
3812 3813 3814
		if (!bh)
			goto out;
		if (journal_quota) {
3815
			err = ext4_journal_get_write_access(handle, bh);
3816 3817 3818 3819 3820 3821 3822 3823 3824 3825
			if (err) {
				brelse(bh);
				goto out;
			}
		}
		lock_buffer(bh);
		memcpy(bh->b_data+offset, data, tocopy);
		flush_dcache_page(bh->b_page);
		unlock_buffer(bh);
		if (journal_quota)
3826
			err = ext4_handle_dirty_metadata(handle, NULL, bh);
3827 3828
		else {
			/* Always do at least ordered writes for quotas */
3829
			err = ext4_jbd2_file_inode(handle, inode);
3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840
			mark_buffer_dirty(bh);
		}
		brelse(bh);
		if (err)
			goto out;
		offset = 0;
		towrite -= tocopy;
		data += tocopy;
		blk++;
	}
out:
3841 3842
	if (len == towrite) {
		mutex_unlock(&inode->i_mutex);
3843
		return err;
3844
	}
3845 3846
	if (inode->i_size < off+len-towrite) {
		i_size_write(inode, off+len-towrite);
3847
		EXT4_I(inode)->i_disksize = inode->i_size;
3848 3849
	}
	inode->i_mtime = inode->i_ctime = CURRENT_TIME;
3850
	ext4_mark_inode_dirty(handle, inode);
3851 3852 3853 3854 3855 3856
	mutex_unlock(&inode->i_mutex);
	return len - towrite;
}

#endif

3857
static int ext4_get_sb(struct file_system_type *fs_type,
3858 3859
	int flags, const char *dev_name, void *data, struct vfsmount *mnt)
{
3860
	return get_sb_bdev(fs_type, flags, dev_name, data, ext4_fill_super, mnt);
3861 3862
}

T
Theodore Ts'o 已提交
3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881
static struct file_system_type ext4_fs_type = {
	.owner		= THIS_MODULE,
	.name		= "ext4",
	.get_sb		= ext4_get_sb,
	.kill_sb	= kill_block_super,
	.fs_flags	= FS_REQUIRES_DEV,
};

#ifdef CONFIG_EXT4DEV_COMPAT
static int ext4dev_get_sb(struct file_system_type *fs_type,
	int flags, const char *dev_name, void *data, struct vfsmount *mnt)
{
	printk(KERN_WARNING "EXT4-fs: Update your userspace programs "
	       "to mount using ext4\n");
	printk(KERN_WARNING "EXT4-fs: ext4dev backwards compatibility "
	       "will go away by 2.6.31\n");
	return get_sb_bdev(fs_type, flags, dev_name, data, ext4_fill_super, mnt);
}

3882
static struct file_system_type ext4dev_fs_type = {
3883
	.owner		= THIS_MODULE,
3884
	.name		= "ext4dev",
T
Theodore Ts'o 已提交
3885
	.get_sb		= ext4dev_get_sb,
3886 3887 3888
	.kill_sb	= kill_block_super,
	.fs_flags	= FS_REQUIRES_DEV,
};
T
Theodore Ts'o 已提交
3889 3890
MODULE_ALIAS("ext4dev");
#endif
3891

3892
static int __init init_ext4_fs(void)
3893
{
3894 3895
	int err;

T
Theodore Ts'o 已提交
3896 3897 3898
	ext4_kset = kset_create_and_add("ext4", NULL, fs_kobj);
	if (!ext4_kset)
		return -ENOMEM;
3899
	ext4_proc_root = proc_mkdir("fs/ext4", NULL);
3900
	err = init_ext4_mballoc();
3901 3902
	if (err)
		return err;
3903 3904 3905 3906

	err = init_ext4_xattr();
	if (err)
		goto out2;
3907 3908 3909
	err = init_inodecache();
	if (err)
		goto out1;
T
Theodore Ts'o 已提交
3910
	err = register_filesystem(&ext4_fs_type);
3911 3912
	if (err)
		goto out;
T
Theodore Ts'o 已提交
3913 3914 3915 3916 3917 3918 3919
#ifdef CONFIG_EXT4DEV_COMPAT
	err = register_filesystem(&ext4dev_fs_type);
	if (err) {
		unregister_filesystem(&ext4_fs_type);
		goto out;
	}
#endif
3920 3921 3922 3923
	return 0;
out:
	destroy_inodecache();
out1:
3924
	exit_ext4_xattr();
3925 3926
out2:
	exit_ext4_mballoc();
3927 3928 3929
	return err;
}

3930
static void __exit exit_ext4_fs(void)
3931
{
T
Theodore Ts'o 已提交
3932 3933
	unregister_filesystem(&ext4_fs_type);
#ifdef CONFIG_EXT4DEV_COMPAT
3934
	unregister_filesystem(&ext4dev_fs_type);
T
Theodore Ts'o 已提交
3935
#endif
3936
	destroy_inodecache();
3937
	exit_ext4_xattr();
3938
	exit_ext4_mballoc();
3939
	remove_proc_entry("fs/ext4", NULL);
T
Theodore Ts'o 已提交
3940
	kset_unregister(ext4_kset);
3941 3942 3943
}

MODULE_AUTHOR("Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others");
3944
MODULE_DESCRIPTION("Fourth Extended Filesystem");
3945
MODULE_LICENSE("GPL");
3946 3947
module_init(init_ext4_fs)
module_exit(exit_ext4_fs)