super.c 22.8 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
 *  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/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
38
#include <linux/mutex.h>
39
#include <linux/file.h>
L
Linus Torvalds 已提交
40
#include <asm/uaccess.h>
41
#include "internal.h"
L
Linus Torvalds 已提交
42 43 44 45 46 47 48


LIST_HEAD(super_blocks);
DEFINE_SPINLOCK(sb_lock);

/**
 *	alloc_super	-	create new superblock
49
 *	@type:	filesystem type superblock should belong to
L
Linus Torvalds 已提交
50 51 52 53
 *
 *	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
54
static struct super_block *alloc_super(struct file_system_type *type)
L
Linus Torvalds 已提交
55
{
56
	struct super_block *s = kzalloc(sizeof(struct super_block),  GFP_USER);
57
	static const struct super_operations default_op;
L
Linus Torvalds 已提交
58 59 60 61 62 63 64 65 66 67 68

	if (s) {
		if (security_sb_alloc(s)) {
			kfree(s);
			s = NULL;
			goto out;
		}
		INIT_LIST_HEAD(&s->s_files);
		INIT_LIST_HEAD(&s->s_instances);
		INIT_HLIST_HEAD(&s->s_anon);
		INIT_LIST_HEAD(&s->s_inodes);
69
		INIT_LIST_HEAD(&s->s_dentry_lru);
L
Linus Torvalds 已提交
70
		init_rwsem(&s->s_umount);
71
		mutex_init(&s->s_lock);
72
		lockdep_set_class(&s->s_umount, &type->s_umount_key);
I
Ingo Molnar 已提交
73 74 75 76 77 78
		/*
		 * 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);
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
		/*
		 * 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
95 96
		s->s_count = S_BIAS;
		atomic_set(&s->s_active, 1);
97
		mutex_init(&s->s_vfs_rename_mutex);
I
Ingo Molnar 已提交
98 99
		mutex_init(&s->s_dquot.dqio_mutex);
		mutex_init(&s->s_dquot.dqonoff_mutex);
L
Linus Torvalds 已提交
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
		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
121
	kfree(s->s_subtype);
122
	kfree(s->s_options);
L
Linus Torvalds 已提交
123 124 125 126 127 128 129 130 131
	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
132
static int __put_super(struct super_block *sb)
L
Linus Torvalds 已提交
133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
{
	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.
 */
171
void put_super(struct super_block *sb)
L
Linus Torvalds 已提交
172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193
{
	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);
194
		vfs_dq_off(s, 0);
L
Linus Torvalds 已提交
195 196 197 198 199 200 201 202 203
		down_write(&s->s_umount);
		fs->kill_sb(s);
		put_filesystem(fs);
		put_super(s);
	}
}

EXPORT_SYMBOL(deactivate_super);

A
Al Viro 已提交
204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231
/**
 *	deactivate_locked_super	-	drop an active reference to superblock
 *	@s: superblock to deactivate
 *
 *	Equivalent of up_write(&s->s_umount); deactivate_super(s);, except that
 *	it does not unlock it until it's all over.  As the result, it's safe to
 *	use to dispose of new superblock on ->get_sb() failure exits - nobody
 *	will see the sucker until it's all over.  Equivalent using up_write +
 *	deactivate_super is safe for that purpose only if superblock is either
 *	safe to use or has NULL ->s_root when we unlock.
 */
void deactivate_locked_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);
		vfs_dq_off(s, 0);
		fs->kill_sb(s);
		put_filesystem(fs);
		put_super(s);
	} else {
		up_write(&s->s_umount);
	}
}

EXPORT_SYMBOL(deactivate_locked_super);

L
Linus Torvalds 已提交
232 233 234 235 236 237 238 239 240 241 242
/**
 *	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).
 */
243
static int grab_super(struct super_block *s) __releases(sb_lock)
L
Linus Torvalds 已提交
244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263
{
	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;
}

264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
/*
 * 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);

L
Linus Torvalds 已提交
282 283 284 285 286 287 288 289 290
/**
 *	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.
291 292 293 294
 *
 *	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
295 296 297
 */
void generic_shutdown_super(struct super_block *sb)
{
298
	const struct super_operations *sop = sb->s_op;
L
Linus Torvalds 已提交
299

300

301 302
	if (sb->s_root) {
		shrink_dcache_for_umount(sb);
303
		sync_filesystem(sb);
304
		get_fs_excl();
L
Linus Torvalds 已提交
305
		sb->s_flags &= ~MS_ACTIVE;
306

L
Linus Torvalds 已提交
307 308 309 310 311 312 313 314
		/* bad name - it should be evict_inodes() */
		invalidate_inodes(sb);

		if (sop->put_super)
			sop->put_super(sb);

		/* Forget any remaining inodes */
		if (invalidate_inodes(sb)) {
315 316 317
			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
318
		}
319
		put_fs_excl();
L
Linus Torvalds 已提交
320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343
	}
	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;
344
	struct super_block *old;
L
Linus Torvalds 已提交
345 346 347 348
	int err;

retry:
	spin_lock(&sb_lock);
349 350 351 352 353 354
	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
355 356
			if (s) {
				up_write(&s->s_umount);
357
				destroy_super(s);
L
Li Zefan 已提交
358
			}
359 360
			return old;
		}
L
Linus Torvalds 已提交
361 362 363
	}
	if (!s) {
		spin_unlock(&sb_lock);
I
Ingo Molnar 已提交
364
		s = alloc_super(type);
L
Linus Torvalds 已提交
365 366 367 368 369 370 371 372
		if (!s)
			return ERR_PTR(-ENOMEM);
		goto retry;
	}
		
	err = set(s, data);
	if (err) {
		spin_unlock(&sb_lock);
L
Li Zefan 已提交
373
		up_write(&s->s_umount);
L
Linus Torvalds 已提交
374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395
		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);

C
Christoph Hellwig 已提交
396 397 398 399 400 401 402 403
/**
 * sync_supers - helper for periodic superblock writeback
 *
 * Call the write_super method if present on all dirty superblocks in
 * the system.  This is for the periodic writeback used by most older
 * filesystems.  For data integrity superblock writeback use
 * sync_filesystems() instead.
 *
L
Linus Torvalds 已提交
404 405 406 407 408 409
 * 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)
{
410 411
	struct super_block *sb;

L
Linus Torvalds 已提交
412
	spin_lock(&sb_lock);
413 414
restart:
	list_for_each_entry(sb, &super_blocks, s_list) {
C
Christoph Hellwig 已提交
415
		if (sb->s_op->write_super && sb->s_dirt) {
L
Linus Torvalds 已提交
416 417
			sb->s_count++;
			spin_unlock(&sb_lock);
C
Christoph Hellwig 已提交
418

L
Linus Torvalds 已提交
419
			down_read(&sb->s_umount);
C
Christoph Hellwig 已提交
420 421
			if (sb->s_root && sb->s_dirt)
				sb->s_op->write_super(sb);
422
			up_read(&sb->s_umount);
C
Christoph Hellwig 已提交
423

424 425 426 427 428
			spin_lock(&sb_lock);
			if (__put_super_and_need_restart(sb))
				goto restart;
		}
	}
L
Linus Torvalds 已提交
429 430 431 432 433 434 435 436 437 438 439 440 441
	spin_unlock(&sb_lock);
}

/**
 *	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)
{
442 443
	struct super_block *sb;

L
Linus Torvalds 已提交
444 445
	if (!bdev)
		return NULL;
446

L
Linus Torvalds 已提交
447
	spin_lock(&sb_lock);
448 449 450 451
rescan:
	list_for_each_entry(sb, &super_blocks, s_list) {
		if (sb->s_bdev == bdev) {
			sb->s_count++;
L
Linus Torvalds 已提交
452
			spin_unlock(&sb_lock);
453 454 455 456 457 458 459 460
			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
461 462 463 464 465 466 467 468 469 470
		}
	}
	spin_unlock(&sb_lock);
	return NULL;
}

EXPORT_SYMBOL(get_super);
 
struct super_block * user_get_super(dev_t dev)
{
471
	struct super_block *sb;
L
Linus Torvalds 已提交
472 473

	spin_lock(&sb_lock);
474 475 476 477
rescan:
	list_for_each_entry(sb, &super_blocks, s_list) {
		if (sb->s_dev ==  dev) {
			sb->s_count++;
L
Linus Torvalds 已提交
478
			spin_unlock(&sb_lock);
479 480 481 482 483 484 485 486
			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
487 488 489 490 491 492
		}
	}
	spin_unlock(&sb_lock);
	return NULL;
}

493
SYSCALL_DEFINE2(ustat, unsigned, dev, struct ustat __user *, ubuf)
L
Linus Torvalds 已提交
494 495 496 497 498 499 500 501 502
{
        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;
503
	err = vfs_statfs(s->s_root, &sbuf);
L
Linus Torvalds 已提交
504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528
	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;
}

/**
 *	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;
529
	int remount_rw;
L
Linus Torvalds 已提交
530
	
531
#ifdef CONFIG_BLOCK
L
Linus Torvalds 已提交
532 533
	if (!(flags & MS_RDONLY) && bdev_read_only(sb->s_bdev))
		return -EACCES;
534
#endif
L
Linus Torvalds 已提交
535 536 537
	if (flags & MS_RDONLY)
		acct_auto_close(sb);
	shrink_dcache_sb(sb);
538
	sync_filesystem(sb);
L
Linus Torvalds 已提交
539 540 541 542 543 544

	/* 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);
545
		else if (!fs_may_remount_ro(sb))
L
Linus Torvalds 已提交
546
			return -EBUSY;
547
		retval = vfs_dq_off(sb, 1);
548
		if (retval < 0 && retval != -ENOSYS)
549
			return -EBUSY;
L
Linus Torvalds 已提交
550
	}
551
	remount_rw = !(flags & MS_RDONLY) && (sb->s_flags & MS_RDONLY);
L
Linus Torvalds 已提交
552 553 554

	if (sb->s_op->remount_fs) {
		retval = sb->s_op->remount_fs(sb, &flags, data);
555
		if (retval)
L
Linus Torvalds 已提交
556 557 558
			return retval;
	}
	sb->s_flags = (sb->s_flags & ~MS_RMT_MASK) | (flags & MS_RMT_MASK);
559
	if (remount_rw)
560
		vfs_dq_quota_on_remount(sb);
L
Linus Torvalds 已提交
561 562 563
	return 0;
}

564
static void do_emergency_remount(struct work_struct *work)
L
Linus Torvalds 已提交
565 566 567 568 569 570 571
{
	struct super_block *sb;

	spin_lock(&sb_lock);
	list_for_each_entry(sb, &super_blocks, s_list) {
		sb->s_count++;
		spin_unlock(&sb_lock);
572
		down_write(&sb->s_umount);
L
Linus Torvalds 已提交
573 574 575 576 577 578 579 580
		if (sb->s_root && sb->s_bdev && !(sb->s_flags & MS_RDONLY)) {
			/*
			 * ->remount_fs needs lock_kernel().
			 *
			 * What lock protects sb->s_flags??
			 */
			do_remount_sb(sb, MS_RDONLY, NULL, 1);
		}
581 582
		up_write(&sb->s_umount);
		put_super(sb);
L
Linus Torvalds 已提交
583 584 585
		spin_lock(&sb_lock);
	}
	spin_unlock(&sb_lock);
586
	kfree(work);
L
Linus Torvalds 已提交
587 588 589 590 591
	printk("Emergency Remount complete\n");
}

void emergency_remount(void)
{
592 593 594 595 596 597 598
	struct work_struct *work;

	work = kmalloc(sizeof(*work), GFP_ATOMIC);
	if (work) {
		INIT_WORK(work, do_emergency_remount);
		schedule_work(work);
	}
L
Linus Torvalds 已提交
599 600 601 602 603 604 605
}

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

606
static DEFINE_IDA(unnamed_dev_ida);
L
Linus Torvalds 已提交
607
static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */
608
static int unnamed_dev_start = 0; /* don't bother trying below it */
L
Linus Torvalds 已提交
609 610 611 612 613 614 615

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

 retry:
616
	if (ida_pre_get(&unnamed_dev_ida, GFP_ATOMIC) == 0)
L
Linus Torvalds 已提交
617 618
		return -ENOMEM;
	spin_lock(&unnamed_dev_lock);
619
	error = ida_get_new_above(&unnamed_dev_ida, unnamed_dev_start, &dev);
620 621
	if (!error)
		unnamed_dev_start = dev + 1;
L
Linus Torvalds 已提交
622 623 624 625 626 627 628 629 630
	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);
631
		ida_remove(&unnamed_dev_ida, dev);
632 633
		if (unnamed_dev_start > dev)
			unnamed_dev_start = dev;
L
Linus Torvalds 已提交
634 635 636 637 638 639 640 641 642 643 644 645 646 647 648
		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);
649
	ida_remove(&unnamed_dev_ida, slot);
650 651
	if (slot < unnamed_dev_start)
		unnamed_dev_start = slot;
L
Linus Torvalds 已提交
652 653 654 655 656 657 658 659 660 661 662 663 664 665
	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);

666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691
static int ns_test_super(struct super_block *sb, void *data)
{
	return sb->s_fs_info == data;
}

static int ns_set_super(struct super_block *sb, void *data)
{
	sb->s_fs_info = data;
	return set_anon_super(sb, NULL);
}

int get_sb_ns(struct file_system_type *fs_type, int flags, void *data,
	int (*fill_super)(struct super_block *, void *, int),
	struct vfsmount *mnt)
{
	struct super_block *sb;

	sb = sget(fs_type, ns_test_super, ns_set_super, data);
	if (IS_ERR(sb))
		return PTR_ERR(sb);

	if (!sb->s_root) {
		int err;
		sb->s_flags = flags;
		err = fill_super(sb, data, flags & MS_SILENT ? 1 : 0);
		if (err) {
A
Al Viro 已提交
692
			deactivate_locked_super(sb);
693 694 695 696 697 698 699 700 701 702 703 704
			return err;
		}

		sb->s_flags |= MS_ACTIVE;
	}

	simple_set_mnt(mnt, sb);
	return 0;
}

EXPORT_SYMBOL(get_sb_ns);

705
#ifdef CONFIG_BLOCK
L
Linus Torvalds 已提交
706 707 708 709
static int set_bdev_super(struct super_block *s, void *data)
{
	s->s_bdev = data;
	s->s_dev = s->s_bdev->bd_dev;
J
Jens Axboe 已提交
710 711 712 713 714 715

	/*
	 * We set the bdi here to the queue backing, file systems can
	 * overwrite this in ->fill_super()
	 */
	s->s_bdi = &bdev_get_queue(s->s_bdev)->backing_dev_info;
L
Linus Torvalds 已提交
716 717 718 719 720 721 722 723
	return 0;
}

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

724
int get_sb_bdev(struct file_system_type *fs_type,
L
Linus Torvalds 已提交
725
	int flags, const char *dev_name, void *data,
726 727
	int (*fill_super)(struct super_block *, void *, int),
	struct vfsmount *mnt)
L
Linus Torvalds 已提交
728 729 730
{
	struct block_device *bdev;
	struct super_block *s;
731
	fmode_t mode = FMODE_READ;
L
Linus Torvalds 已提交
732 733
	int error = 0;

734 735 736 737
	if (!(flags & MS_RDONLY))
		mode |= FMODE_WRITE;

	bdev = open_bdev_exclusive(dev_name, mode, fs_type);
L
Linus Torvalds 已提交
738
	if (IS_ERR(bdev))
739
		return PTR_ERR(bdev);
L
Linus Torvalds 已提交
740 741 742 743 744 745

	/*
	 * 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
	 */
746 747 748 749 750 751
	mutex_lock(&bdev->bd_fsfreeze_mutex);
	if (bdev->bd_fsfreeze_count > 0) {
		mutex_unlock(&bdev->bd_fsfreeze_mutex);
		error = -EBUSY;
		goto error_bdev;
	}
L
Linus Torvalds 已提交
752
	s = sget(fs_type, test_bdev_super, set_bdev_super, bdev);
753
	mutex_unlock(&bdev->bd_fsfreeze_mutex);
L
Linus Torvalds 已提交
754
	if (IS_ERR(s))
755
		goto error_s;
L
Linus Torvalds 已提交
756 757 758

	if (s->s_root) {
		if ((flags ^ s->s_flags) & MS_RDONLY) {
A
Al Viro 已提交
759
			deactivate_locked_super(s);
760 761
			error = -EBUSY;
			goto error_bdev;
L
Linus Torvalds 已提交
762
		}
763

764
		close_bdev_exclusive(bdev, mode);
L
Linus Torvalds 已提交
765 766 767 768
	} else {
		char b[BDEVNAME_SIZE];

		s->s_flags = flags;
769
		s->s_mode = mode;
L
Linus Torvalds 已提交
770
		strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
771
		sb_set_blocksize(s, block_size(bdev));
772
		error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
L
Linus Torvalds 已提交
773
		if (error) {
A
Al Viro 已提交
774
			deactivate_locked_super(s);
775
			goto error;
776
		}
777 778

		s->s_flags |= MS_ACTIVE;
779
		bdev->bd_super = s;
L
Linus Torvalds 已提交
780 781
	}

782 783
	simple_set_mnt(mnt, s);
	return 0;
L
Linus Torvalds 已提交
784

785 786 787
error_s:
	error = PTR_ERR(s);
error_bdev:
788
	close_bdev_exclusive(bdev, mode);
789 790
error:
	return error;
L
Linus Torvalds 已提交
791 792 793 794 795 796 797
}

EXPORT_SYMBOL(get_sb_bdev);

void kill_block_super(struct super_block *sb)
{
	struct block_device *bdev = sb->s_bdev;
798
	fmode_t mode = sb->s_mode;
L
Linus Torvalds 已提交
799

800
	bdev->bd_super = NULL;
L
Linus Torvalds 已提交
801 802
	generic_shutdown_super(sb);
	sync_blockdev(bdev);
803
	close_bdev_exclusive(bdev, mode);
L
Linus Torvalds 已提交
804 805 806
}

EXPORT_SYMBOL(kill_block_super);
807
#endif
L
Linus Torvalds 已提交
808

809
int get_sb_nodev(struct file_system_type *fs_type,
L
Linus Torvalds 已提交
810
	int flags, void *data,
811 812
	int (*fill_super)(struct super_block *, void *, int),
	struct vfsmount *mnt)
L
Linus Torvalds 已提交
813 814 815 816 817
{
	int error;
	struct super_block *s = sget(fs_type, NULL, set_anon_super, NULL);

	if (IS_ERR(s))
818
		return PTR_ERR(s);
L
Linus Torvalds 已提交
819 820 821

	s->s_flags = flags;

822
	error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
L
Linus Torvalds 已提交
823
	if (error) {
A
Al Viro 已提交
824
		deactivate_locked_super(s);
825
		return error;
L
Linus Torvalds 已提交
826 827
	}
	s->s_flags |= MS_ACTIVE;
828 829
	simple_set_mnt(mnt, s);
	return 0;
L
Linus Torvalds 已提交
830 831 832 833 834 835 836 837 838
}

EXPORT_SYMBOL(get_sb_nodev);

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

839
int get_sb_single(struct file_system_type *fs_type,
L
Linus Torvalds 已提交
840
	int flags, void *data,
841 842
	int (*fill_super)(struct super_block *, void *, int),
	struct vfsmount *mnt)
L
Linus Torvalds 已提交
843 844 845 846 847 848
{
	struct super_block *s;
	int error;

	s = sget(fs_type, compare_single, set_anon_super, NULL);
	if (IS_ERR(s))
849
		return PTR_ERR(s);
L
Linus Torvalds 已提交
850 851
	if (!s->s_root) {
		s->s_flags = flags;
852
		error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
L
Linus Torvalds 已提交
853
		if (error) {
A
Al Viro 已提交
854
			deactivate_locked_super(s);
855
			return error;
L
Linus Torvalds 已提交
856 857 858 859
		}
		s->s_flags |= MS_ACTIVE;
	}
	do_remount_sb(s, flags, data, 0);
860 861
	simple_set_mnt(mnt, s);
	return 0;
L
Linus Torvalds 已提交
862 863 864 865 866
}

EXPORT_SYMBOL(get_sb_single);

struct vfsmount *
867
vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void *data)
L
Linus Torvalds 已提交
868 869 870
{
	struct vfsmount *mnt;
	char *secdata = NULL;
871
	int error;
L
Linus Torvalds 已提交
872 873 874 875

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

876
	error = -ENOMEM;
L
Linus Torvalds 已提交
877 878 879 880
	mnt = alloc_vfsmnt(name);
	if (!mnt)
		goto out;

881
	if (data && !(type->fs_flags & FS_BINARY_MOUNTDATA)) {
L
Linus Torvalds 已提交
882
		secdata = alloc_secdata();
883
		if (!secdata)
L
Linus Torvalds 已提交
884 885
			goto out_mnt;

886
		error = security_sb_copy_data(data, secdata);
887
		if (error)
L
Linus Torvalds 已提交
888 889 890
			goto out_free_secdata;
	}

891 892
	error = type->get_sb(type, flags, name, data, mnt);
	if (error < 0)
L
Linus Torvalds 已提交
893
		goto out_free_secdata;
894
	BUG_ON(!mnt->mnt_sb);
895

896
 	error = security_sb_kern_mount(mnt->mnt_sb, flags, secdata);
L
Linus Torvalds 已提交
897 898
 	if (error)
 		goto out_sb;
899

900 901 902 903 904 905 906 907 908 909
	/*
	 * filesystems should never set s_maxbytes larger than MAX_LFS_FILESIZE
	 * but s_maxbytes was an unsigned long long for many releases. Throw
	 * this warning for a little while to try and catch filesystems that
	 * violate this rule. This warning should be either removed or
	 * converted to a BUG() in 2.6.34.
	 */
	WARN((mnt->mnt_sb->s_maxbytes < 0), "%s set sb->s_maxbytes to "
		"negative value (%lld)\n", type->name, mnt->mnt_sb->s_maxbytes);

910
	mnt->mnt_mountpoint = mnt->mnt_root;
L
Linus Torvalds 已提交
911
	mnt->mnt_parent = mnt;
912
	up_write(&mnt->mnt_sb->s_umount);
913
	free_secdata(secdata);
L
Linus Torvalds 已提交
914 915
	return mnt;
out_sb:
916
	dput(mnt->mnt_root);
A
Al Viro 已提交
917
	deactivate_locked_super(mnt->mnt_sb);
L
Linus Torvalds 已提交
918 919 920 921 922
out_free_secdata:
	free_secdata(secdata);
out_mnt:
	free_vfsmnt(mnt);
out:
923
	return ERR_PTR(error);
L
Linus Torvalds 已提交
924 925
}

926 927
EXPORT_SYMBOL_GPL(vfs_kern_mount);

M
Miklos Szeredi 已提交
928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950
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);
}

951 952 953 954 955 956 957 958
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
959 960 961
	if (!IS_ERR(mnt) && (type->fs_flags & FS_HAS_SUBTYPE) &&
	    !mnt->mnt_sb->s_subtype)
		mnt = fs_set_subtype(mnt, fstype);
962 963 964
	put_filesystem(type);
	return mnt;
}
965
EXPORT_SYMBOL_GPL(do_kern_mount);
966

967
struct vfsmount *kern_mount_data(struct file_system_type *type, void *data)
L
Linus Torvalds 已提交
968
{
969
	return vfs_kern_mount(type, MS_KERNMOUNT, type->name, data);
L
Linus Torvalds 已提交
970 971
}

972
EXPORT_SYMBOL_GPL(kern_mount_data);