super.c 23.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
/*
 *  linux/fs/super.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *
 *  super.c contains code to handle: - mount structures
 *                                   - super-block tables
 *                                   - filesystem drivers list
 *                                   - mount system call
 *                                   - umount system call
 *                                   - ustat system call
 *
 * GK 2/5/95  -  Changed to support mounting the root fs via NFS
 *
 *  Added kerneld support: Jacques Gelinas and Bjorn Ekwall
 *  Added change_root: Werner Almesberger & Hans Lermen, Feb '96
 *  Added options to /proc/mounts:
18
 *    Torbjörn Lindh (torbjorn.lindh@gopta.se), April 14, 1996.
L
Linus Torvalds 已提交
19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
 *  Added devfs support: Richard Gooch <rgooch@atnf.csiro.au>, 13-JAN-1998
 *  Heavily rewritten for 'one fs - one tree' dcache architecture. AV, Mar 2000
 */

#include <linux/module.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/smp_lock.h>
#include <linux/acct.h>
#include <linux/blkdev.h>
#include <linux/quotaops.h>
#include <linux/namei.h>
#include <linux/buffer_head.h>		/* for fsync_super() */
#include <linux/mount.h>
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/vfs.h>
#include <linux/writeback.h>		/* for the emergency remount stuff */
#include <linux/idr.h>
#include <linux/kobject.h>
I
Ingo Molnar 已提交
39
#include <linux/mutex.h>
40
#include <linux/file.h>
41
#include <linux/async.h>
L
Linus Torvalds 已提交
42
#include <asm/uaccess.h>
43
#include "internal.h"
L
Linus Torvalds 已提交
44 45 46 47 48 49 50


LIST_HEAD(super_blocks);
DEFINE_SPINLOCK(sb_lock);

/**
 *	alloc_super	-	create new superblock
51
 *	@type:	filesystem type superblock should belong to
L
Linus Torvalds 已提交
52 53 54 55
 *
 *	Allocates and initializes a new &struct super_block.  alloc_super()
 *	returns a pointer new superblock or %NULL if allocation had failed.
 */
I
Ingo Molnar 已提交
56
static struct super_block *alloc_super(struct file_system_type *type)
L
Linus Torvalds 已提交
57
{
58
	struct super_block *s = kzalloc(sizeof(struct super_block),  GFP_USER);
L
Linus Torvalds 已提交
59 60 61 62 63 64 65 66 67 68
	static struct super_operations default_op;

	if (s) {
		if (security_sb_alloc(s)) {
			kfree(s);
			s = NULL;
			goto out;
		}
		INIT_LIST_HEAD(&s->s_dirty);
		INIT_LIST_HEAD(&s->s_io);
69
		INIT_LIST_HEAD(&s->s_more_io);
L
Linus Torvalds 已提交
70 71 72 73
		INIT_LIST_HEAD(&s->s_files);
		INIT_LIST_HEAD(&s->s_instances);
		INIT_HLIST_HEAD(&s->s_anon);
		INIT_LIST_HEAD(&s->s_inodes);
74
		INIT_LIST_HEAD(&s->s_dentry_lru);
75
		INIT_LIST_HEAD(&s->s_async_list);
L
Linus Torvalds 已提交
76
		init_rwsem(&s->s_umount);
77
		mutex_init(&s->s_lock);
78
		lockdep_set_class(&s->s_umount, &type->s_umount_key);
I
Ingo Molnar 已提交
79 80 81 82 83 84
		/*
		 * The locking rules for s_lock are up to the
		 * filesystem. For example ext3fs has different
		 * lock ordering than usbfs:
		 */
		lockdep_set_class(&s->s_lock, &type->s_lock_key);
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
		/*
		 * sget() can have s_umount recursion.
		 *
		 * When it cannot find a suitable sb, it allocates a new
		 * one (this one), and tries again to find a suitable old
		 * one.
		 *
		 * In case that succeeds, it will acquire the s_umount
		 * lock of the old one. Since these are clearly distrinct
		 * locks, and this object isn't exposed yet, there's no
		 * risk of deadlocks.
		 *
		 * Annotate this by putting this lock in a different
		 * subclass.
		 */
		down_write_nested(&s->s_umount, SINGLE_DEPTH_NESTING);
L
Linus Torvalds 已提交
101 102
		s->s_count = S_BIAS;
		atomic_set(&s->s_active, 1);
103
		mutex_init(&s->s_vfs_rename_mutex);
I
Ingo Molnar 已提交
104 105
		mutex_init(&s->s_dquot.dqio_mutex);
		mutex_init(&s->s_dquot.dqonoff_mutex);
L
Linus Torvalds 已提交
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
		init_rwsem(&s->s_dquot.dqptr_sem);
		init_waitqueue_head(&s->s_wait_unfrozen);
		s->s_maxbytes = MAX_NON_LFS;
		s->dq_op = sb_dquot_ops;
		s->s_qcop = sb_quotactl_ops;
		s->s_op = &default_op;
		s->s_time_gran = 1000000000;
	}
out:
	return s;
}

/**
 *	destroy_super	-	frees a superblock
 *	@s: superblock to free
 *
 *	Frees a superblock.
 */
static inline void destroy_super(struct super_block *s)
{
	security_sb_free(s);
M
Miklos Szeredi 已提交
127
	kfree(s->s_subtype);
128
	kfree(s->s_options);
L
Linus Torvalds 已提交
129 130 131 132 133 134 135 136 137
	kfree(s);
}

/* Superblock refcounting  */

/*
 * Drop a superblock's refcount.  Returns non-zero if the superblock was
 * destroyed.  The caller must hold sb_lock.
 */
A
Adrian Bunk 已提交
138
static int __put_super(struct super_block *sb)
L
Linus Torvalds 已提交
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199
{
	int ret = 0;

	if (!--sb->s_count) {
		destroy_super(sb);
		ret = 1;
	}
	return ret;
}

/*
 * Drop a superblock's refcount.
 * Returns non-zero if the superblock is about to be destroyed and
 * at least is already removed from super_blocks list, so if we are
 * making a loop through super blocks then we need to restart.
 * The caller must hold sb_lock.
 */
int __put_super_and_need_restart(struct super_block *sb)
{
	/* check for race with generic_shutdown_super() */
	if (list_empty(&sb->s_list)) {
		/* super block is removed, need to restart... */
		__put_super(sb);
		return 1;
	}
	/* can't be the last, since s_list is still in use */
	sb->s_count--;
	BUG_ON(sb->s_count == 0);
	return 0;
}

/**
 *	put_super	-	drop a temporary reference to superblock
 *	@sb: superblock in question
 *
 *	Drops a temporary reference, frees superblock if there's no
 *	references left.
 */
static void put_super(struct super_block *sb)
{
	spin_lock(&sb_lock);
	__put_super(sb);
	spin_unlock(&sb_lock);
}


/**
 *	deactivate_super	-	drop an active reference to superblock
 *	@s: superblock to deactivate
 *
 *	Drops an active reference to superblock, acquiring a temprory one if
 *	there is no active references left.  In that case we lock superblock,
 *	tell fs driver to shut it down and drop the temporary reference we
 *	had just acquired.
 */
void deactivate_super(struct super_block *s)
{
	struct file_system_type *fs = s->s_type;
	if (atomic_dec_and_lock(&s->s_active, &sb_lock)) {
		s->s_count -= S_BIAS-1;
		spin_unlock(&sb_lock);
200
		DQUOT_OFF(s, 0);
L
Linus Torvalds 已提交
201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220
		down_write(&s->s_umount);
		fs->kill_sb(s);
		put_filesystem(fs);
		put_super(s);
	}
}

EXPORT_SYMBOL(deactivate_super);

/**
 *	grab_super - acquire an active reference
 *	@s: reference we are trying to make active
 *
 *	Tries to acquire an active reference.  grab_super() is used when we
 * 	had just found a superblock in super_blocks or fs_type->fs_supers
 *	and want to turn it into a full-blown active reference.  grab_super()
 *	is called with sb_lock held and drops it.  Returns 1 in case of
 *	success, 0 if we had failed (superblock contents was already dead or
 *	dying when grab_super() had been called).
 */
221
static int grab_super(struct super_block *s) __releases(sb_lock)
L
Linus Torvalds 已提交
222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241
{
	s->s_count++;
	spin_unlock(&sb_lock);
	down_write(&s->s_umount);
	if (s->s_root) {
		spin_lock(&sb_lock);
		if (s->s_count > S_BIAS) {
			atomic_inc(&s->s_active);
			s->s_count--;
			spin_unlock(&sb_lock);
			return 1;
		}
		spin_unlock(&sb_lock);
	}
	up_write(&s->s_umount);
	put_super(s);
	yield();
	return 0;
}

242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259
/*
 * Superblock locking.  We really ought to get rid of these two.
 */
void lock_super(struct super_block * sb)
{
	get_fs_excl();
	mutex_lock(&sb->s_lock);
}

void unlock_super(struct super_block * sb)
{
	put_fs_excl();
	mutex_unlock(&sb->s_lock);
}

EXPORT_SYMBOL(lock_super);
EXPORT_SYMBOL(unlock_super);

260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290
/*
 * Write out and wait upon all dirty data associated with this
 * superblock.  Filesystem data as well as the underlying block
 * device.  Takes the superblock lock.  Requires a second blkdev
 * flush by the caller to complete the operation.
 */
void __fsync_super(struct super_block *sb)
{
	sync_inodes_sb(sb, 0);
	DQUOT_SYNC(sb);
	lock_super(sb);
	if (sb->s_dirt && sb->s_op->write_super)
		sb->s_op->write_super(sb);
	unlock_super(sb);
	if (sb->s_op->sync_fs)
		sb->s_op->sync_fs(sb, 1);
	sync_blockdev(sb->s_bdev);
	sync_inodes_sb(sb, 1);
}

/*
 * Write out and wait upon all dirty data associated with this
 * superblock.  Filesystem data as well as the underlying block
 * device.  Takes the superblock lock.
 */
int fsync_super(struct super_block *sb)
{
	__fsync_super(sb);
	return sync_blockdev(sb->s_bdev);
}

L
Linus Torvalds 已提交
291 292 293 294 295 296 297 298 299
/**
 *	generic_shutdown_super	-	common helper for ->kill_sb()
 *	@sb: superblock to kill
 *
 *	generic_shutdown_super() does all fs-independent work on superblock
 *	shutdown.  Typical ->kill_sb() should pick all fs-specific objects
 *	that need destruction out of superblock, call generic_shutdown_super()
 *	and release aforementioned objects.  Note: dentries and inodes _are_
 *	taken care of and do not need specific handling.
300 301 302 303
 *
 *	Upon calling this function, the filesystem may no longer alter or
 *	rearrange the set of dentries belonging to this super_block, nor may it
 *	change the attachments of dentries to inodes.
L
Linus Torvalds 已提交
304 305 306
 */
void generic_shutdown_super(struct super_block *sb)
{
307
	const struct super_operations *sop = sb->s_op;
L
Linus Torvalds 已提交
308

309

310 311
	if (sb->s_root) {
		shrink_dcache_for_umount(sb);
L
Linus Torvalds 已提交
312 313 314
		fsync_super(sb);
		lock_super(sb);
		sb->s_flags &= ~MS_ACTIVE;
315 316 317 318

		/*
		 * wait for asynchronous fs operations to finish before going further
		 */
319
		async_synchronize_full_domain(&sb->s_async_list);
320

L
Linus Torvalds 已提交
321 322 323 324 325 326 327 328 329 330 331
		/* bad name - it should be evict_inodes() */
		invalidate_inodes(sb);
		lock_kernel();

		if (sop->write_super && sb->s_dirt)
			sop->write_super(sb);
		if (sop->put_super)
			sop->put_super(sb);

		/* Forget any remaining inodes */
		if (invalidate_inodes(sb)) {
332 333 334
			printk("VFS: Busy inodes after unmount of %s. "
			   "Self-destruct in 5 seconds.  Have a nice day...\n",
			   sb->s_id);
L
Linus Torvalds 已提交
335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362
		}

		unlock_kernel();
		unlock_super(sb);
	}
	spin_lock(&sb_lock);
	/* should be initialized for __put_super_and_need_restart() */
	list_del_init(&sb->s_list);
	list_del(&sb->s_instances);
	spin_unlock(&sb_lock);
	up_write(&sb->s_umount);
}

EXPORT_SYMBOL(generic_shutdown_super);

/**
 *	sget	-	find or create a superblock
 *	@type:	filesystem type superblock should belong to
 *	@test:	comparison callback
 *	@set:	setup callback
 *	@data:	argument to each of them
 */
struct super_block *sget(struct file_system_type *type,
			int (*test)(struct super_block *,void *),
			int (*set)(struct super_block *,void *),
			void *data)
{
	struct super_block *s = NULL;
363
	struct super_block *old;
L
Linus Torvalds 已提交
364 365 366 367
	int err;

retry:
	spin_lock(&sb_lock);
368 369 370 371 372 373
	if (test) {
		list_for_each_entry(old, &type->fs_supers, s_instances) {
			if (!test(old, data))
				continue;
			if (!grab_super(old))
				goto retry;
L
Li Zefan 已提交
374 375
			if (s) {
				up_write(&s->s_umount);
376
				destroy_super(s);
L
Li Zefan 已提交
377
			}
378 379
			return old;
		}
L
Linus Torvalds 已提交
380 381 382
	}
	if (!s) {
		spin_unlock(&sb_lock);
I
Ingo Molnar 已提交
383
		s = alloc_super(type);
L
Linus Torvalds 已提交
384 385 386 387 388 389 390 391
		if (!s)
			return ERR_PTR(-ENOMEM);
		goto retry;
	}
		
	err = set(s, data);
	if (err) {
		spin_unlock(&sb_lock);
L
Li Zefan 已提交
392
		up_write(&s->s_umount);
L
Linus Torvalds 已提交
393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430
		destroy_super(s);
		return ERR_PTR(err);
	}
	s->s_type = type;
	strlcpy(s->s_id, type->name, sizeof(s->s_id));
	list_add_tail(&s->s_list, &super_blocks);
	list_add(&s->s_instances, &type->fs_supers);
	spin_unlock(&sb_lock);
	get_filesystem(type);
	return s;
}

EXPORT_SYMBOL(sget);

void drop_super(struct super_block *sb)
{
	up_read(&sb->s_umount);
	put_super(sb);
}

EXPORT_SYMBOL(drop_super);

static inline void write_super(struct super_block *sb)
{
	lock_super(sb);
	if (sb->s_root && sb->s_dirt)
		if (sb->s_op->write_super)
			sb->s_op->write_super(sb);
	unlock_super(sb);
}

/*
 * Note: check the dirty flag before waiting, so we don't
 * hold up the sync while mounting a device. (The newly
 * mounted device won't need syncing.)
 */
void sync_supers(void)
{
431 432
	struct super_block *sb;

L
Linus Torvalds 已提交
433
	spin_lock(&sb_lock);
434 435
restart:
	list_for_each_entry(sb, &super_blocks, s_list) {
L
Linus Torvalds 已提交
436 437 438 439 440
		if (sb->s_dirt) {
			sb->s_count++;
			spin_unlock(&sb_lock);
			down_read(&sb->s_umount);
			write_super(sb);
441 442 443 444 445 446
			up_read(&sb->s_umount);
			spin_lock(&sb_lock);
			if (__put_super_and_need_restart(sb))
				goto restart;
		}
	}
L
Linus Torvalds 已提交
447 448 449 450
	spin_unlock(&sb_lock);
}

/*
451
 * Call the ->sync_fs super_op against all filesystems which are r/w and
L
Linus Torvalds 已提交
452 453 454 455 456 457 458
 * which implement it.
 *
 * This operation is careful to avoid the livelock which could easily happen
 * if two or more filesystems are being continuously dirtied.  s_need_sync_fs
 * is used only here.  We set it against all filesystems and then clear it as
 * we sync them.  So redirtied filesystems are skipped.
 *
459
 * But if process A is currently running sync_filesystems and then process B
L
Linus Torvalds 已提交
460 461 462 463 464 465 466 467 468
 * calls sync_filesystems as well, process B will set all the s_need_sync_fs
 * flags again, which will cause process A to resync everything.  Fix that with
 * a local mutex.
 *
 * (Fabian) Avoid sync_fs with clean fs & wait mode 0
 */
void sync_filesystems(int wait)
{
	struct super_block *sb;
I
Ingo Molnar 已提交
469
	static DEFINE_MUTEX(mutex);
L
Linus Torvalds 已提交
470

I
Ingo Molnar 已提交
471
	mutex_lock(&mutex);		/* Could be down_interruptible */
L
Linus Torvalds 已提交
472
	spin_lock(&sb_lock);
473
	list_for_each_entry(sb, &super_blocks, s_list) {
L
Linus Torvalds 已提交
474 475 476 477 478 479 480 481
		if (!sb->s_op->sync_fs)
			continue;
		if (sb->s_flags & MS_RDONLY)
			continue;
		sb->s_need_sync_fs = 1;
	}

restart:
482
	list_for_each_entry(sb, &super_blocks, s_list) {
L
Linus Torvalds 已提交
483 484 485 486 487 488 489 490
		if (!sb->s_need_sync_fs)
			continue;
		sb->s_need_sync_fs = 0;
		if (sb->s_flags & MS_RDONLY)
			continue;	/* hm.  Was remounted r/o meanwhile */
		sb->s_count++;
		spin_unlock(&sb_lock);
		down_read(&sb->s_umount);
491
		async_synchronize_full_domain(&sb->s_async_list);
L
Linus Torvalds 已提交
492 493
		if (sb->s_root && (wait || sb->s_dirt))
			sb->s_op->sync_fs(sb, wait);
494 495 496 497 498
		up_read(&sb->s_umount);
		/* restart only when sb is no longer on the list */
		spin_lock(&sb_lock);
		if (__put_super_and_need_restart(sb))
			goto restart;
L
Linus Torvalds 已提交
499 500
	}
	spin_unlock(&sb_lock);
I
Ingo Molnar 已提交
501
	mutex_unlock(&mutex);
L
Linus Torvalds 已提交
502 503 504 505 506 507 508 509 510 511 512 513
}

/**
 *	get_super - get the superblock of a device
 *	@bdev: device to get the superblock for
 *	
 *	Scans the superblock list and finds the superblock of the file system
 *	mounted on the device given. %NULL is returned if no match is found.
 */

struct super_block * get_super(struct block_device *bdev)
{
514 515
	struct super_block *sb;

L
Linus Torvalds 已提交
516 517
	if (!bdev)
		return NULL;
518

L
Linus Torvalds 已提交
519
	spin_lock(&sb_lock);
520 521 522 523
rescan:
	list_for_each_entry(sb, &super_blocks, s_list) {
		if (sb->s_bdev == bdev) {
			sb->s_count++;
L
Linus Torvalds 已提交
524
			spin_unlock(&sb_lock);
525 526 527 528 529 530 531 532
			down_read(&sb->s_umount);
			if (sb->s_root)
				return sb;
			up_read(&sb->s_umount);
			/* restart only when sb is no longer on the list */
			spin_lock(&sb_lock);
			if (__put_super_and_need_restart(sb))
				goto rescan;
L
Linus Torvalds 已提交
533 534 535 536 537 538 539 540 541 542
		}
	}
	spin_unlock(&sb_lock);
	return NULL;
}

EXPORT_SYMBOL(get_super);
 
struct super_block * user_get_super(dev_t dev)
{
543
	struct super_block *sb;
L
Linus Torvalds 已提交
544 545

	spin_lock(&sb_lock);
546 547 548 549
rescan:
	list_for_each_entry(sb, &super_blocks, s_list) {
		if (sb->s_dev ==  dev) {
			sb->s_count++;
L
Linus Torvalds 已提交
550
			spin_unlock(&sb_lock);
551 552 553 554 555 556 557 558
			down_read(&sb->s_umount);
			if (sb->s_root)
				return sb;
			up_read(&sb->s_umount);
			/* restart only when sb is no longer on the list */
			spin_lock(&sb_lock);
			if (__put_super_and_need_restart(sb))
				goto rescan;
L
Linus Torvalds 已提交
559 560 561 562 563 564
		}
	}
	spin_unlock(&sb_lock);
	return NULL;
}

565
SYSCALL_DEFINE2(ustat, unsigned, dev, struct ustat __user *, ubuf)
L
Linus Torvalds 已提交
566 567 568 569 570 571 572 573 574
{
        struct super_block *s;
        struct ustat tmp;
        struct kstatfs sbuf;
	int err = -EINVAL;

        s = user_get_super(new_decode_dev(dev));
        if (s == NULL)
                goto out;
575
	err = vfs_statfs(s->s_root, &sbuf);
L
Linus Torvalds 已提交
576 577 578 579 580 581 582 583 584 585 586 587 588 589
	drop_super(s);
	if (err)
		goto out;

        memset(&tmp,0,sizeof(struct ustat));
        tmp.f_tfree = sbuf.f_bfree;
        tmp.f_tinode = sbuf.f_ffree;

        err = copy_to_user(ubuf,&tmp,sizeof(struct ustat)) ? -EFAULT : 0;
out:
	return err;
}

/**
590
 *	mark_files_ro - mark all files read-only
L
Linus Torvalds 已提交
591 592
 *	@sb: superblock in question
 *
593 594
 *	All files are marked read-only.  We don't care about pending
 *	delete files so this should be used in 'force' mode only.
L
Linus Torvalds 已提交
595 596 597 598 599 600
 */

static void mark_files_ro(struct super_block *sb)
{
	struct file *f;

601
retry:
L
Linus Torvalds 已提交
602
	file_list_lock();
E
Eric Dumazet 已提交
603
	list_for_each_entry(f, &sb->s_files, f_u.fu_list) {
604 605 606 607 608 609 610 611
		struct vfsmount *mnt;
		if (!S_ISREG(f->f_path.dentry->d_inode->i_mode))
		       continue;
		if (!file_count(f))
			continue;
		if (!(f->f_mode & FMODE_WRITE))
			continue;
		f->f_mode &= ~FMODE_WRITE;
612 613 614
		if (file_check_writeable(f) != 0)
			continue;
		file_release_write(f);
615 616 617 618 619 620 621 622 623
		mnt = mntget(f->f_path.mnt);
		file_list_unlock();
		/*
		 * This can sleep, so we can't hold
		 * the file_list_lock() spinlock.
		 */
		mnt_drop_write(mnt);
		mntput(mnt);
		goto retry;
L
Linus Torvalds 已提交
624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639
	}
	file_list_unlock();
}

/**
 *	do_remount_sb - asks filesystem to change mount options.
 *	@sb:	superblock in question
 *	@flags:	numeric part of options
 *	@data:	the rest of options
 *      @force: whether or not to force the change
 *
 *	Alters the mount options of a mounted file system.
 */
int do_remount_sb(struct super_block *sb, int flags, void *data, int force)
{
	int retval;
640
	int remount_rw;
L
Linus Torvalds 已提交
641
	
642
#ifdef CONFIG_BLOCK
L
Linus Torvalds 已提交
643 644
	if (!(flags & MS_RDONLY) && bdev_read_only(sb->s_bdev))
		return -EACCES;
645
#endif
L
Linus Torvalds 已提交
646 647 648 649 650 651 652 653 654 655 656 657
	if (flags & MS_RDONLY)
		acct_auto_close(sb);
	shrink_dcache_sb(sb);
	fsync_super(sb);

	/* If we are remounting RDONLY and current sb is read/write,
	   make sure there are no rw files opened */
	if ((flags & MS_RDONLY) && !(sb->s_flags & MS_RDONLY)) {
		if (force)
			mark_files_ro(sb);
		else if (!fs_may_remount_ro(sb))
			return -EBUSY;
658 659 660
		retval = DQUOT_OFF(sb, 1);
		if (retval < 0 && retval != -ENOSYS)
			return -EBUSY;
L
Linus Torvalds 已提交
661
	}
662
	remount_rw = !(flags & MS_RDONLY) && (sb->s_flags & MS_RDONLY);
L
Linus Torvalds 已提交
663 664 665 666 667 668 669 670 671

	if (sb->s_op->remount_fs) {
		lock_super(sb);
		retval = sb->s_op->remount_fs(sb, &flags, data);
		unlock_super(sb);
		if (retval)
			return retval;
	}
	sb->s_flags = (sb->s_flags & ~MS_RMT_MASK) | (flags & MS_RMT_MASK);
672 673
	if (remount_rw)
		DQUOT_ON_REMOUNT(sb);
L
Linus Torvalds 已提交
674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712
	return 0;
}

static void do_emergency_remount(unsigned long foo)
{
	struct super_block *sb;

	spin_lock(&sb_lock);
	list_for_each_entry(sb, &super_blocks, s_list) {
		sb->s_count++;
		spin_unlock(&sb_lock);
		down_read(&sb->s_umount);
		if (sb->s_root && sb->s_bdev && !(sb->s_flags & MS_RDONLY)) {
			/*
			 * ->remount_fs needs lock_kernel().
			 *
			 * What lock protects sb->s_flags??
			 */
			lock_kernel();
			do_remount_sb(sb, MS_RDONLY, NULL, 1);
			unlock_kernel();
		}
		drop_super(sb);
		spin_lock(&sb_lock);
	}
	spin_unlock(&sb_lock);
	printk("Emergency Remount complete\n");
}

void emergency_remount(void)
{
	pdflush_operation(do_emergency_remount, 0);
}

/*
 * Unnamed block devices are dummy devices used by virtual
 * filesystems which don't use real block-devices.  -- jrs
 */

713
static DEFINE_IDA(unnamed_dev_ida);
L
Linus Torvalds 已提交
714 715 716 717 718 719 720 721
static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */

int set_anon_super(struct super_block *s, void *data)
{
	int dev;
	int error;

 retry:
722
	if (ida_pre_get(&unnamed_dev_ida, GFP_ATOMIC) == 0)
L
Linus Torvalds 已提交
723 724
		return -ENOMEM;
	spin_lock(&unnamed_dev_lock);
725
	error = ida_get_new(&unnamed_dev_ida, &dev);
L
Linus Torvalds 已提交
726 727 728 729 730 731 732 733 734
	spin_unlock(&unnamed_dev_lock);
	if (error == -EAGAIN)
		/* We raced and lost with another CPU. */
		goto retry;
	else if (error)
		return -EAGAIN;

	if ((dev & MAX_ID_MASK) == (1 << MINORBITS)) {
		spin_lock(&unnamed_dev_lock);
735
		ida_remove(&unnamed_dev_ida, dev);
L
Linus Torvalds 已提交
736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
		spin_unlock(&unnamed_dev_lock);
		return -EMFILE;
	}
	s->s_dev = MKDEV(0, dev & MINORMASK);
	return 0;
}

EXPORT_SYMBOL(set_anon_super);

void kill_anon_super(struct super_block *sb)
{
	int slot = MINOR(sb->s_dev);

	generic_shutdown_super(sb);
	spin_lock(&unnamed_dev_lock);
751
	ida_remove(&unnamed_dev_ida, slot);
L
Linus Torvalds 已提交
752 753 754 755 756 757 758 759 760 761 762 763 764 765
	spin_unlock(&unnamed_dev_lock);
}

EXPORT_SYMBOL(kill_anon_super);

void kill_litter_super(struct super_block *sb)
{
	if (sb->s_root)
		d_genocide(sb->s_root);
	kill_anon_super(sb);
}

EXPORT_SYMBOL(kill_litter_super);

766
#ifdef CONFIG_BLOCK
L
Linus Torvalds 已提交
767 768 769 770 771 772 773 774 775 776 777 778
static int set_bdev_super(struct super_block *s, void *data)
{
	s->s_bdev = data;
	s->s_dev = s->s_bdev->bd_dev;
	return 0;
}

static int test_bdev_super(struct super_block *s, void *data)
{
	return (void *)s->s_bdev == data;
}

779
int get_sb_bdev(struct file_system_type *fs_type,
L
Linus Torvalds 已提交
780
	int flags, const char *dev_name, void *data,
781 782
	int (*fill_super)(struct super_block *, void *, int),
	struct vfsmount *mnt)
L
Linus Torvalds 已提交
783 784 785
{
	struct block_device *bdev;
	struct super_block *s;
786
	fmode_t mode = FMODE_READ;
L
Linus Torvalds 已提交
787 788
	int error = 0;

789 790 791 792
	if (!(flags & MS_RDONLY))
		mode |= FMODE_WRITE;

	bdev = open_bdev_exclusive(dev_name, mode, fs_type);
L
Linus Torvalds 已提交
793
	if (IS_ERR(bdev))
794
		return PTR_ERR(bdev);
L
Linus Torvalds 已提交
795 796 797 798 799 800

	/*
	 * once the super is inserted into the list by sget, s_umount
	 * will protect the lockfs code from trying to start a snapshot
	 * while we are mounting
	 */
801
	down(&bdev->bd_mount_sem);
L
Linus Torvalds 已提交
802
	s = sget(fs_type, test_bdev_super, set_bdev_super, bdev);
803
	up(&bdev->bd_mount_sem);
L
Linus Torvalds 已提交
804
	if (IS_ERR(s))
805
		goto error_s;
L
Linus Torvalds 已提交
806 807 808 809 810

	if (s->s_root) {
		if ((flags ^ s->s_flags) & MS_RDONLY) {
			up_write(&s->s_umount);
			deactivate_super(s);
811 812
			error = -EBUSY;
			goto error_bdev;
L
Linus Torvalds 已提交
813
		}
814

815
		close_bdev_exclusive(bdev, mode);
L
Linus Torvalds 已提交
816 817 818 819
	} else {
		char b[BDEVNAME_SIZE];

		s->s_flags = flags;
820
		s->s_mode = mode;
L
Linus Torvalds 已提交
821
		strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
822
		sb_set_blocksize(s, block_size(bdev));
823
		error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
L
Linus Torvalds 已提交
824 825 826
		if (error) {
			up_write(&s->s_umount);
			deactivate_super(s);
827
			goto error;
828
		}
829 830

		s->s_flags |= MS_ACTIVE;
831
		bdev->bd_super = s;
L
Linus Torvalds 已提交
832 833
	}

834
	return simple_set_mnt(mnt, s);
L
Linus Torvalds 已提交
835

836 837 838
error_s:
	error = PTR_ERR(s);
error_bdev:
839
	close_bdev_exclusive(bdev, mode);
840 841
error:
	return error;
L
Linus Torvalds 已提交
842 843 844 845 846 847 848
}

EXPORT_SYMBOL(get_sb_bdev);

void kill_block_super(struct super_block *sb)
{
	struct block_device *bdev = sb->s_bdev;
849
	fmode_t mode = sb->s_mode;
L
Linus Torvalds 已提交
850

851
	bdev->bd_super = 0;
L
Linus Torvalds 已提交
852 853
	generic_shutdown_super(sb);
	sync_blockdev(bdev);
854
	close_bdev_exclusive(bdev, mode);
L
Linus Torvalds 已提交
855 856 857
}

EXPORT_SYMBOL(kill_block_super);
858
#endif
L
Linus Torvalds 已提交
859

860
int get_sb_nodev(struct file_system_type *fs_type,
L
Linus Torvalds 已提交
861
	int flags, void *data,
862 863
	int (*fill_super)(struct super_block *, void *, int),
	struct vfsmount *mnt)
L
Linus Torvalds 已提交
864 865 866 867 868
{
	int error;
	struct super_block *s = sget(fs_type, NULL, set_anon_super, NULL);

	if (IS_ERR(s))
869
		return PTR_ERR(s);
L
Linus Torvalds 已提交
870 871 872

	s->s_flags = flags;

873
	error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
L
Linus Torvalds 已提交
874 875 876
	if (error) {
		up_write(&s->s_umount);
		deactivate_super(s);
877
		return error;
L
Linus Torvalds 已提交
878 879
	}
	s->s_flags |= MS_ACTIVE;
880
	return simple_set_mnt(mnt, s);
L
Linus Torvalds 已提交
881 882 883 884 885 886 887 888 889
}

EXPORT_SYMBOL(get_sb_nodev);

static int compare_single(struct super_block *s, void *p)
{
	return 1;
}

890
int get_sb_single(struct file_system_type *fs_type,
L
Linus Torvalds 已提交
891
	int flags, void *data,
892 893
	int (*fill_super)(struct super_block *, void *, int),
	struct vfsmount *mnt)
L
Linus Torvalds 已提交
894 895 896 897 898 899
{
	struct super_block *s;
	int error;

	s = sget(fs_type, compare_single, set_anon_super, NULL);
	if (IS_ERR(s))
900
		return PTR_ERR(s);
L
Linus Torvalds 已提交
901 902
	if (!s->s_root) {
		s->s_flags = flags;
903
		error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
L
Linus Torvalds 已提交
904 905 906
		if (error) {
			up_write(&s->s_umount);
			deactivate_super(s);
907
			return error;
L
Linus Torvalds 已提交
908 909 910 911
		}
		s->s_flags |= MS_ACTIVE;
	}
	do_remount_sb(s, flags, data, 0);
912
	return simple_set_mnt(mnt, s);
L
Linus Torvalds 已提交
913 914 915 916 917
}

EXPORT_SYMBOL(get_sb_single);

struct vfsmount *
918
vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void *data)
L
Linus Torvalds 已提交
919 920 921
{
	struct vfsmount *mnt;
	char *secdata = NULL;
922
	int error;
L
Linus Torvalds 已提交
923 924 925 926

	if (!type)
		return ERR_PTR(-ENODEV);

927
	error = -ENOMEM;
L
Linus Torvalds 已提交
928 929 930 931
	mnt = alloc_vfsmnt(name);
	if (!mnt)
		goto out;

932
	if (data && !(type->fs_flags & FS_BINARY_MOUNTDATA)) {
L
Linus Torvalds 已提交
933
		secdata = alloc_secdata();
934
		if (!secdata)
L
Linus Torvalds 已提交
935 936
			goto out_mnt;

937
		error = security_sb_copy_data(data, secdata);
938
		if (error)
L
Linus Torvalds 已提交
939 940 941
			goto out_free_secdata;
	}

942 943
	error = type->get_sb(type, flags, name, data, mnt);
	if (error < 0)
L
Linus Torvalds 已提交
944
		goto out_free_secdata;
945
	BUG_ON(!mnt->mnt_sb);
946

947
 	error = security_sb_kern_mount(mnt->mnt_sb, flags, secdata);
L
Linus Torvalds 已提交
948 949
 	if (error)
 		goto out_sb;
950 951

	mnt->mnt_mountpoint = mnt->mnt_root;
L
Linus Torvalds 已提交
952
	mnt->mnt_parent = mnt;
953
	up_write(&mnt->mnt_sb->s_umount);
954
	free_secdata(secdata);
L
Linus Torvalds 已提交
955 956
	return mnt;
out_sb:
957 958 959
	dput(mnt->mnt_root);
	up_write(&mnt->mnt_sb->s_umount);
	deactivate_super(mnt->mnt_sb);
L
Linus Torvalds 已提交
960 961 962 963 964
out_free_secdata:
	free_secdata(secdata);
out_mnt:
	free_vfsmnt(mnt);
out:
965
	return ERR_PTR(error);
L
Linus Torvalds 已提交
966 967
}

968 969
EXPORT_SYMBOL_GPL(vfs_kern_mount);

M
Miklos Szeredi 已提交
970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992
static struct vfsmount *fs_set_subtype(struct vfsmount *mnt, const char *fstype)
{
	int err;
	const char *subtype = strchr(fstype, '.');
	if (subtype) {
		subtype++;
		err = -EINVAL;
		if (!subtype[0])
			goto err;
	} else
		subtype = "";

	mnt->mnt_sb->s_subtype = kstrdup(subtype, GFP_KERNEL);
	err = -ENOMEM;
	if (!mnt->mnt_sb->s_subtype)
		goto err;
	return mnt;

 err:
	mntput(mnt);
	return ERR_PTR(err);
}

993 994 995 996 997 998 999 1000
struct vfsmount *
do_kern_mount(const char *fstype, int flags, const char *name, void *data)
{
	struct file_system_type *type = get_fs_type(fstype);
	struct vfsmount *mnt;
	if (!type)
		return ERR_PTR(-ENODEV);
	mnt = vfs_kern_mount(type, flags, name, data);
M
Miklos Szeredi 已提交
1001 1002 1003
	if (!IS_ERR(mnt) && (type->fs_flags & FS_HAS_SUBTYPE) &&
	    !mnt->mnt_sb->s_subtype)
		mnt = fs_set_subtype(mnt, fstype);
1004 1005 1006
	put_filesystem(type);
	return mnt;
}
1007
EXPORT_SYMBOL_GPL(do_kern_mount);
1008

1009
struct vfsmount *kern_mount_data(struct file_system_type *type, void *data)
L
Linus Torvalds 已提交
1010
{
1011
	return vfs_kern_mount(type, MS_KERNMOUNT, type->name, data);
L
Linus Torvalds 已提交
1012 1013
}

1014
EXPORT_SYMBOL_GPL(kern_mount_data);