namespace.c 46.8 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
/*
 *  linux/fs/namespace.c
 *
 * (C) Copyright Al Viro 2000, 2001
 *	Released under GPL v2.
 *
 * Based on code from fs/super.c, copyright Linus Torvalds and others.
 * Heavily rewritten.
 */

#include <linux/syscalls.h>
#include <linux/slab.h>
#include <linux/sched.h>
#include <linux/smp_lock.h>
#include <linux/init.h>
16
#include <linux/kernel.h>
L
Linus Torvalds 已提交
17 18
#include <linux/quotaops.h>
#include <linux/acct.h>
19
#include <linux/capability.h>
L
Linus Torvalds 已提交
20
#include <linux/module.h>
21
#include <linux/sysfs.h>
L
Linus Torvalds 已提交
22
#include <linux/seq_file.h>
23
#include <linux/mnt_namespace.h>
L
Linus Torvalds 已提交
24 25 26
#include <linux/namei.h>
#include <linux/security.h>
#include <linux/mount.h>
27
#include <linux/ramfs.h>
L
Linus Torvalds 已提交
28 29
#include <asm/uaccess.h>
#include <asm/unistd.h>
30
#include "pnode.h"
L
Linus Torvalds 已提交
31 32

/* spinlock for vfsmount related operations, inplace of dcache_lock */
A
Al Viro 已提交
33 34 35
__cacheline_aligned_in_smp DEFINE_SPINLOCK(vfsmount_lock);

static int event;
L
Linus Torvalds 已提交
36

37
static struct list_head *mount_hashtable __read_mostly;
38
static int hash_mask __read_mostly, hash_bits __read_mostly;
39
static struct kmem_cache *mnt_cache __read_mostly;
R
Ram Pai 已提交
40
static struct rw_semaphore namespace_sem;
L
Linus Torvalds 已提交
41

M
Miklos Szeredi 已提交
42 43 44 45
/* /sys/fs */
decl_subsys(fs, NULL, NULL);
EXPORT_SYMBOL_GPL(fs_subsys);

L
Linus Torvalds 已提交
46 47
static inline unsigned long hash(struct vfsmount *mnt, struct dentry *dentry)
{
R
Ram Pai 已提交
48 49
	unsigned long tmp = ((unsigned long)mnt / L1_CACHE_BYTES);
	tmp += ((unsigned long)dentry / L1_CACHE_BYTES);
L
Linus Torvalds 已提交
50 51 52 53 54 55
	tmp = tmp + (tmp >> hash_bits);
	return tmp & hash_mask;
}

struct vfsmount *alloc_vfsmnt(const char *name)
{
R
Ram Pai 已提交
56
	struct vfsmount *mnt = kmem_cache_alloc(mnt_cache, GFP_KERNEL);
L
Linus Torvalds 已提交
57 58
	if (mnt) {
		memset(mnt, 0, sizeof(struct vfsmount));
R
Ram Pai 已提交
59
		atomic_set(&mnt->mnt_count, 1);
L
Linus Torvalds 已提交
60 61 62 63
		INIT_LIST_HEAD(&mnt->mnt_hash);
		INIT_LIST_HEAD(&mnt->mnt_child);
		INIT_LIST_HEAD(&mnt->mnt_mounts);
		INIT_LIST_HEAD(&mnt->mnt_list);
64
		INIT_LIST_HEAD(&mnt->mnt_expire);
R
Ram Pai 已提交
65
		INIT_LIST_HEAD(&mnt->mnt_share);
R
Ram Pai 已提交
66 67
		INIT_LIST_HEAD(&mnt->mnt_slave_list);
		INIT_LIST_HEAD(&mnt->mnt_slave);
L
Linus Torvalds 已提交
68
		if (name) {
R
Ram Pai 已提交
69
			int size = strlen(name) + 1;
L
Linus Torvalds 已提交
70 71 72 73 74 75 76 77 78 79
			char *newname = kmalloc(size, GFP_KERNEL);
			if (newname) {
				memcpy(newname, name, size);
				mnt->mnt_devname = newname;
			}
		}
	}
	return mnt;
}

80 81 82 83 84 85 86 87 88
int simple_set_mnt(struct vfsmount *mnt, struct super_block *sb)
{
	mnt->mnt_sb = sb;
	mnt->mnt_root = dget(sb->s_root);
	return 0;
}

EXPORT_SYMBOL(simple_set_mnt);

L
Linus Torvalds 已提交
89 90 91 92 93 94 95
void free_vfsmnt(struct vfsmount *mnt)
{
	kfree(mnt->mnt_devname);
	kmem_cache_free(mnt_cache, mnt);
}

/*
R
Ram Pai 已提交
96 97
 * find the first or last mount at @dentry on vfsmount @mnt depending on
 * @dir. If @dir is set return the first mount else return the last mount.
L
Linus Torvalds 已提交
98
 */
R
Ram Pai 已提交
99 100
struct vfsmount *__lookup_mnt(struct vfsmount *mnt, struct dentry *dentry,
			      int dir)
L
Linus Torvalds 已提交
101
{
R
Ram Pai 已提交
102 103
	struct list_head *head = mount_hashtable + hash(mnt, dentry);
	struct list_head *tmp = head;
L
Linus Torvalds 已提交
104 105 106
	struct vfsmount *p, *found = NULL;

	for (;;) {
R
Ram Pai 已提交
107
		tmp = dir ? tmp->next : tmp->prev;
L
Linus Torvalds 已提交
108 109 110 111 112
		p = NULL;
		if (tmp == head)
			break;
		p = list_entry(tmp, struct vfsmount, mnt_hash);
		if (p->mnt_parent == mnt && p->mnt_mountpoint == dentry) {
R
Ram Pai 已提交
113
			found = p;
L
Linus Torvalds 已提交
114 115 116 117 118 119
			break;
		}
	}
	return found;
}

R
Ram Pai 已提交
120 121 122 123 124 125 126 127 128 129 130 131 132 133
/*
 * lookup_mnt increments the ref count before returning
 * the vfsmount struct.
 */
struct vfsmount *lookup_mnt(struct vfsmount *mnt, struct dentry *dentry)
{
	struct vfsmount *child_mnt;
	spin_lock(&vfsmount_lock);
	if ((child_mnt = __lookup_mnt(mnt, dentry, 1)))
		mntget(child_mnt);
	spin_unlock(&vfsmount_lock);
	return child_mnt;
}

L
Linus Torvalds 已提交
134 135
static inline int check_mnt(struct vfsmount *mnt)
{
136
	return mnt->mnt_ns == current->nsproxy->mnt_ns;
L
Linus Torvalds 已提交
137 138
}

139
static void touch_mnt_namespace(struct mnt_namespace *ns)
A
Al Viro 已提交
140 141 142 143 144 145 146
{
	if (ns) {
		ns->event = ++event;
		wake_up_interruptible(&ns->poll);
	}
}

147
static void __touch_mnt_namespace(struct mnt_namespace *ns)
A
Al Viro 已提交
148 149 150 151 152 153 154
{
	if (ns && ns->event != event) {
		ns->event = event;
		wake_up_interruptible(&ns->poll);
	}
}

L
Linus Torvalds 已提交
155 156 157 158 159 160 161 162 163 164 165
static void detach_mnt(struct vfsmount *mnt, struct nameidata *old_nd)
{
	old_nd->dentry = mnt->mnt_mountpoint;
	old_nd->mnt = mnt->mnt_parent;
	mnt->mnt_parent = mnt;
	mnt->mnt_mountpoint = mnt->mnt_root;
	list_del_init(&mnt->mnt_child);
	list_del_init(&mnt->mnt_hash);
	old_nd->dentry->d_mounted--;
}

166 167 168 169 170 171 172 173
void mnt_set_mountpoint(struct vfsmount *mnt, struct dentry *dentry,
			struct vfsmount *child_mnt)
{
	child_mnt->mnt_parent = mntget(mnt);
	child_mnt->mnt_mountpoint = dget(dentry);
	dentry->d_mounted++;
}

L
Linus Torvalds 已提交
174 175
static void attach_mnt(struct vfsmount *mnt, struct nameidata *nd)
{
176 177 178
	mnt_set_mountpoint(nd->mnt, nd->dentry, mnt);
	list_add_tail(&mnt->mnt_hash, mount_hashtable +
			hash(nd->mnt, nd->dentry));
L
Linus Torvalds 已提交
179
	list_add_tail(&mnt->mnt_child, &nd->mnt->mnt_mounts);
180 181 182 183 184 185 186 187 188 189
}

/*
 * the caller must hold vfsmount_lock
 */
static void commit_tree(struct vfsmount *mnt)
{
	struct vfsmount *parent = mnt->mnt_parent;
	struct vfsmount *m;
	LIST_HEAD(head);
190
	struct mnt_namespace *n = parent->mnt_ns;
191 192 193 194 195

	BUG_ON(parent == mnt);

	list_add_tail(&head, &mnt->mnt_list);
	list_for_each_entry(m, &head, mnt_list)
196
		m->mnt_ns = n;
197 198 199 200 201
	list_splice(&head, n->list.prev);

	list_add_tail(&mnt->mnt_hash, mount_hashtable +
				hash(parent, mnt->mnt_mountpoint));
	list_add_tail(&mnt->mnt_child, &parent->mnt_mounts);
202
	touch_mnt_namespace(n);
L
Linus Torvalds 已提交
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220
}

static struct vfsmount *next_mnt(struct vfsmount *p, struct vfsmount *root)
{
	struct list_head *next = p->mnt_mounts.next;
	if (next == &p->mnt_mounts) {
		while (1) {
			if (p == root)
				return NULL;
			next = p->mnt_child.next;
			if (next != &p->mnt_parent->mnt_mounts)
				break;
			p = p->mnt_parent;
		}
	}
	return list_entry(next, struct vfsmount, mnt_child);
}

R
Ram Pai 已提交
221 222 223 224 225 226 227 228 229 230
static struct vfsmount *skip_mnt_tree(struct vfsmount *p)
{
	struct list_head *prev = p->mnt_mounts.prev;
	while (prev != &p->mnt_mounts) {
		p = list_entry(prev, struct vfsmount, mnt_child);
		prev = p->mnt_mounts.prev;
	}
	return p;
}

R
Ram Pai 已提交
231 232
static struct vfsmount *clone_mnt(struct vfsmount *old, struct dentry *root,
					int flag)
L
Linus Torvalds 已提交
233 234 235 236 237 238 239 240 241 242 243
{
	struct super_block *sb = old->mnt_sb;
	struct vfsmount *mnt = alloc_vfsmnt(old->mnt_devname);

	if (mnt) {
		mnt->mnt_flags = old->mnt_flags;
		atomic_inc(&sb->s_active);
		mnt->mnt_sb = sb;
		mnt->mnt_root = dget(root);
		mnt->mnt_mountpoint = mnt->mnt_root;
		mnt->mnt_parent = mnt;
244

R
Ram Pai 已提交
245 246 247 248 249 250 251 252 253 254 255
		if (flag & CL_SLAVE) {
			list_add(&mnt->mnt_slave, &old->mnt_slave_list);
			mnt->mnt_master = old;
			CLEAR_MNT_SHARED(mnt);
		} else {
			if ((flag & CL_PROPAGATION) || IS_MNT_SHARED(old))
				list_add(&mnt->mnt_share, &old->mnt_share);
			if (IS_MNT_SLAVE(old))
				list_add(&mnt->mnt_slave, &old->mnt_slave);
			mnt->mnt_master = old->mnt_master;
		}
256 257
		if (flag & CL_MAKE_SHARED)
			set_mnt_shared(mnt);
L
Linus Torvalds 已提交
258 259 260

		/* stick the duplicate mount on the same expiry list
		 * as the original if that was on one */
R
Ram Pai 已提交
261 262 263 264 265 266
		if (flag & CL_EXPIRE) {
			spin_lock(&vfsmount_lock);
			if (!list_empty(&old->mnt_expire))
				list_add(&mnt->mnt_expire, &old->mnt_expire);
			spin_unlock(&vfsmount_lock);
		}
L
Linus Torvalds 已提交
267 268 269 270
	}
	return mnt;
}

271
static inline void __mntput(struct vfsmount *mnt)
L
Linus Torvalds 已提交
272 273 274 275 276 277 278
{
	struct super_block *sb = mnt->mnt_sb;
	dput(mnt->mnt_root);
	free_vfsmnt(mnt);
	deactivate_super(sb);
}

279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
void mntput_no_expire(struct vfsmount *mnt)
{
repeat:
	if (atomic_dec_and_lock(&mnt->mnt_count, &vfsmount_lock)) {
		if (likely(!mnt->mnt_pinned)) {
			spin_unlock(&vfsmount_lock);
			__mntput(mnt);
			return;
		}
		atomic_add(mnt->mnt_pinned + 1, &mnt->mnt_count);
		mnt->mnt_pinned = 0;
		spin_unlock(&vfsmount_lock);
		acct_auto_close_mnt(mnt);
		security_sb_umount_close(mnt);
		goto repeat;
	}
}

EXPORT_SYMBOL(mntput_no_expire);

void mnt_pin(struct vfsmount *mnt)
{
	spin_lock(&vfsmount_lock);
	mnt->mnt_pinned++;
	spin_unlock(&vfsmount_lock);
}

EXPORT_SYMBOL(mnt_pin);

void mnt_unpin(struct vfsmount *mnt)
{
	spin_lock(&vfsmount_lock);
	if (mnt->mnt_pinned) {
		atomic_inc(&mnt->mnt_count);
		mnt->mnt_pinned--;
	}
	spin_unlock(&vfsmount_lock);
}

EXPORT_SYMBOL(mnt_unpin);
L
Linus Torvalds 已提交
319 320 321 322

/* iterator */
static void *m_start(struct seq_file *m, loff_t *pos)
{
323
	struct mnt_namespace *n = m->private;
L
Linus Torvalds 已提交
324 325 326
	struct list_head *p;
	loff_t l = *pos;

R
Ram Pai 已提交
327
	down_read(&namespace_sem);
L
Linus Torvalds 已提交
328 329 330 331 332 333 334 335
	list_for_each(p, &n->list)
		if (!l--)
			return list_entry(p, struct vfsmount, mnt_list);
	return NULL;
}

static void *m_next(struct seq_file *m, void *v, loff_t *pos)
{
336
	struct mnt_namespace *n = m->private;
L
Linus Torvalds 已提交
337 338
	struct list_head *p = ((struct vfsmount *)v)->mnt_list.next;
	(*pos)++;
R
Ram Pai 已提交
339
	return p == &n->list ? NULL : list_entry(p, struct vfsmount, mnt_list);
L
Linus Torvalds 已提交
340 341 342 343
}

static void m_stop(struct seq_file *m, void *v)
{
R
Ram Pai 已提交
344
	up_read(&namespace_sem);
L
Linus Torvalds 已提交
345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368
}

static inline void mangle(struct seq_file *m, const char *s)
{
	seq_escape(m, s, " \t\n\\");
}

static int show_vfsmnt(struct seq_file *m, void *v)
{
	struct vfsmount *mnt = v;
	int err = 0;
	static struct proc_fs_info {
		int flag;
		char *str;
	} fs_info[] = {
		{ MS_SYNCHRONOUS, ",sync" },
		{ MS_DIRSYNC, ",dirsync" },
		{ MS_MANDLOCK, ",mand" },
		{ 0, NULL }
	};
	static struct proc_fs_info mnt_info[] = {
		{ MNT_NOSUID, ",nosuid" },
		{ MNT_NODEV, ",nodev" },
		{ MNT_NOEXEC, ",noexec" },
369 370
		{ MNT_NOATIME, ",noatime" },
		{ MNT_NODIRATIME, ",nodiratime" },
V
Valerie Henson 已提交
371
		{ MNT_RELATIME, ",relatime" },
L
Linus Torvalds 已提交
372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402
		{ 0, NULL }
	};
	struct proc_fs_info *fs_infop;

	mangle(m, mnt->mnt_devname ? mnt->mnt_devname : "none");
	seq_putc(m, ' ');
	seq_path(m, mnt, mnt->mnt_root, " \t\n\\");
	seq_putc(m, ' ');
	mangle(m, mnt->mnt_sb->s_type->name);
	seq_puts(m, mnt->mnt_sb->s_flags & MS_RDONLY ? " ro" : " rw");
	for (fs_infop = fs_info; fs_infop->flag; fs_infop++) {
		if (mnt->mnt_sb->s_flags & fs_infop->flag)
			seq_puts(m, fs_infop->str);
	}
	for (fs_infop = mnt_info; fs_infop->flag; fs_infop++) {
		if (mnt->mnt_flags & fs_infop->flag)
			seq_puts(m, fs_infop->str);
	}
	if (mnt->mnt_sb->s_op->show_options)
		err = mnt->mnt_sb->s_op->show_options(m, mnt);
	seq_puts(m, " 0 0\n");
	return err;
}

struct seq_operations mounts_op = {
	.start	= m_start,
	.next	= m_next,
	.stop	= m_stop,
	.show	= show_vfsmnt
};

403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
static int show_vfsstat(struct seq_file *m, void *v)
{
	struct vfsmount *mnt = v;
	int err = 0;

	/* device */
	if (mnt->mnt_devname) {
		seq_puts(m, "device ");
		mangle(m, mnt->mnt_devname);
	} else
		seq_puts(m, "no device");

	/* mount point */
	seq_puts(m, " mounted on ");
	seq_path(m, mnt, mnt->mnt_root, " \t\n\\");
	seq_putc(m, ' ');

	/* file system type */
	seq_puts(m, "with fstype ");
	mangle(m, mnt->mnt_sb->s_type->name);

	/* optional statistics */
	if (mnt->mnt_sb->s_op->show_stats) {
		seq_putc(m, ' ');
		err = mnt->mnt_sb->s_op->show_stats(m, mnt);
	}

	seq_putc(m, '\n');
	return err;
}

struct seq_operations mountstats_op = {
	.start	= m_start,
	.next	= m_next,
	.stop	= m_stop,
	.show	= show_vfsstat,
};

L
Linus Torvalds 已提交
441 442 443 444 445 446 447 448 449 450
/**
 * may_umount_tree - check if a mount tree is busy
 * @mnt: root of mount tree
 *
 * This is called to check if a tree of mounts has any
 * open files, pwds, chroots or sub mounts that are
 * busy.
 */
int may_umount_tree(struct vfsmount *mnt)
{
R
Ram Pai 已提交
451 452 453
	int actual_refs = 0;
	int minimum_refs = 0;
	struct vfsmount *p;
L
Linus Torvalds 已提交
454 455

	spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
456
	for (p = mnt; p; p = next_mnt(p, mnt)) {
L
Linus Torvalds 已提交
457 458 459 460 461 462
		actual_refs += atomic_read(&p->mnt_count);
		minimum_refs += 2;
	}
	spin_unlock(&vfsmount_lock);

	if (actual_refs > minimum_refs)
463
		return 0;
L
Linus Torvalds 已提交
464

465
	return 1;
L
Linus Torvalds 已提交
466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484
}

EXPORT_SYMBOL(may_umount_tree);

/**
 * may_umount - check if a mount point is busy
 * @mnt: root of mount
 *
 * This is called to check if a mount point has any
 * open files, pwds, chroots or sub mounts. If the
 * mount has sub mounts this will return busy
 * regardless of whether the sub mounts are busy.
 *
 * Doesn't take quota and stuff into account. IOW, in some cases it will
 * give false negatives. The main reason why it's here is that we need
 * a non-destructive way to look for easily umountable filesystems.
 */
int may_umount(struct vfsmount *mnt)
{
485
	int ret = 1;
R
Ram Pai 已提交
486 487
	spin_lock(&vfsmount_lock);
	if (propagate_mount_busy(mnt, 2))
488
		ret = 0;
R
Ram Pai 已提交
489 490
	spin_unlock(&vfsmount_lock);
	return ret;
L
Linus Torvalds 已提交
491 492 493 494
}

EXPORT_SYMBOL(may_umount);

495
void release_mounts(struct list_head *head)
R
Ram Pai 已提交
496 497
{
	struct vfsmount *mnt;
M
Miklos Szeredi 已提交
498
	while (!list_empty(head)) {
R
Ram Pai 已提交
499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516
		mnt = list_entry(head->next, struct vfsmount, mnt_hash);
		list_del_init(&mnt->mnt_hash);
		if (mnt->mnt_parent != mnt) {
			struct dentry *dentry;
			struct vfsmount *m;
			spin_lock(&vfsmount_lock);
			dentry = mnt->mnt_mountpoint;
			m = mnt->mnt_parent;
			mnt->mnt_mountpoint = mnt->mnt_root;
			mnt->mnt_parent = mnt;
			spin_unlock(&vfsmount_lock);
			dput(dentry);
			mntput(m);
		}
		mntput(mnt);
	}
}

R
Ram Pai 已提交
517
void umount_tree(struct vfsmount *mnt, int propagate, struct list_head *kill)
L
Linus Torvalds 已提交
518 519 520
{
	struct vfsmount *p;

A
Akinobu Mita 已提交
521 522
	for (p = mnt; p; p = next_mnt(p, mnt))
		list_move(&p->mnt_hash, kill);
L
Linus Torvalds 已提交
523

R
Ram Pai 已提交
524 525 526
	if (propagate)
		propagate_umount(kill);

R
Ram Pai 已提交
527 528 529
	list_for_each_entry(p, kill, mnt_hash) {
		list_del_init(&p->mnt_expire);
		list_del_init(&p->mnt_list);
530 531
		__touch_mnt_namespace(p->mnt_ns);
		p->mnt_ns = NULL;
R
Ram Pai 已提交
532 533
		list_del_init(&p->mnt_child);
		if (p->mnt_parent != p)
534
			p->mnt_mountpoint->d_mounted--;
R
Ram Pai 已提交
535
		change_mnt_propagation(p, MS_PRIVATE);
L
Linus Torvalds 已提交
536 537 538 539 540
	}
}

static int do_umount(struct vfsmount *mnt, int flags)
{
R
Ram Pai 已提交
541
	struct super_block *sb = mnt->mnt_sb;
L
Linus Torvalds 已提交
542
	int retval;
R
Ram Pai 已提交
543
	LIST_HEAD(umount_list);
L
Linus Torvalds 已提交
544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577

	retval = security_sb_umount(mnt, flags);
	if (retval)
		return retval;

	/*
	 * Allow userspace to request a mountpoint be expired rather than
	 * unmounting unconditionally. Unmount only happens if:
	 *  (1) the mark is already set (the mark is cleared by mntput())
	 *  (2) the usage count == 1 [parent vfsmount] + 1 [sys_umount]
	 */
	if (flags & MNT_EXPIRE) {
		if (mnt == current->fs->rootmnt ||
		    flags & (MNT_FORCE | MNT_DETACH))
			return -EINVAL;

		if (atomic_read(&mnt->mnt_count) != 2)
			return -EBUSY;

		if (!xchg(&mnt->mnt_expiry_mark, 1))
			return -EAGAIN;
	}

	/*
	 * If we may have to abort operations to get out of this
	 * mount, and they will themselves hold resources we must
	 * allow the fs to do things. In the Unix tradition of
	 * 'Gee thats tricky lets do it in userspace' the umount_begin
	 * might fail to complete on the first run through as other tasks
	 * must return, and the like. Thats for the mount program to worry
	 * about for the moment.
	 */

	lock_kernel();
578 579
	if (sb->s_op->umount_begin)
		sb->s_op->umount_begin(mnt, flags);
L
Linus Torvalds 已提交
580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606
	unlock_kernel();

	/*
	 * No sense to grab the lock for this test, but test itself looks
	 * somewhat bogus. Suggestions for better replacement?
	 * Ho-hum... In principle, we might treat that as umount + switch
	 * to rootfs. GC would eventually take care of the old vfsmount.
	 * Actually it makes sense, especially if rootfs would contain a
	 * /reboot - static binary that would close all descriptors and
	 * call reboot(9). Then init(8) could umount root and exec /reboot.
	 */
	if (mnt == current->fs->rootmnt && !(flags & MNT_DETACH)) {
		/*
		 * Special case for "unmounting" root ...
		 * we just try to remount it readonly.
		 */
		down_write(&sb->s_umount);
		if (!(sb->s_flags & MS_RDONLY)) {
			lock_kernel();
			DQUOT_OFF(sb);
			retval = do_remount_sb(sb, MS_RDONLY, NULL, 0);
			unlock_kernel();
		}
		up_write(&sb->s_umount);
		return retval;
	}

R
Ram Pai 已提交
607
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
608
	spin_lock(&vfsmount_lock);
A
Al Viro 已提交
609
	event++;
L
Linus Torvalds 已提交
610 611

	retval = -EBUSY;
R
Ram Pai 已提交
612
	if (flags & MNT_DETACH || !propagate_mount_busy(mnt, 2)) {
L
Linus Torvalds 已提交
613
		if (!list_empty(&mnt->mnt_list))
R
Ram Pai 已提交
614
			umount_tree(mnt, 1, &umount_list);
L
Linus Torvalds 已提交
615 616 617 618 619
		retval = 0;
	}
	spin_unlock(&vfsmount_lock);
	if (retval)
		security_sb_umount_busy(mnt);
R
Ram Pai 已提交
620
	up_write(&namespace_sem);
R
Ram Pai 已提交
621
	release_mounts(&umount_list);
L
Linus Torvalds 已提交
622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
	return retval;
}

/*
 * Now umount can handle mount points as well as block devices.
 * This is important for filesystems which use unnamed block devices.
 *
 * We now support a flag for forced unmount like the other 'big iron'
 * unixes. Our API is identical to OSF/1 to avoid making a mess of AMD
 */

asmlinkage long sys_umount(char __user * name, int flags)
{
	struct nameidata nd;
	int retval;

	retval = __user_walk(name, LOOKUP_FOLLOW, &nd);
	if (retval)
		goto out;
	retval = -EINVAL;
	if (nd.dentry != nd.mnt->mnt_root)
		goto dput_and_out;
	if (!check_mnt(nd.mnt))
		goto dput_and_out;

	retval = -EPERM;
	if (!capable(CAP_SYS_ADMIN))
		goto dput_and_out;

	retval = do_umount(nd.mnt, flags);
dput_and_out:
	path_release_on_umount(&nd);
out:
	return retval;
}

#ifdef __ARCH_WANT_SYS_OLDUMOUNT

/*
R
Ram Pai 已提交
661
 *	The 2.0 compatible umount. No flags.
L
Linus Torvalds 已提交
662 663 664
 */
asmlinkage long sys_oldumount(char __user * name)
{
R
Ram Pai 已提交
665
	return sys_umount(name, 0);
L
Linus Torvalds 已提交
666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681
}

#endif

static int mount_is_safe(struct nameidata *nd)
{
	if (capable(CAP_SYS_ADMIN))
		return 0;
	return -EPERM;
#ifdef notyet
	if (S_ISLNK(nd->dentry->d_inode->i_mode))
		return -EPERM;
	if (nd->dentry->d_inode->i_mode & S_ISVTX) {
		if (current->uid != nd->dentry->d_inode->i_uid)
			return -EPERM;
	}
682
	if (vfs_permission(nd, MAY_WRITE))
L
Linus Torvalds 已提交
683 684 685 686 687
		return -EPERM;
	return 0;
#endif
}

R
Ram Pai 已提交
688
static int lives_below_in_same_fs(struct dentry *d, struct dentry *dentry)
L
Linus Torvalds 已提交
689 690 691 692 693 694 695 696 697 698
{
	while (1) {
		if (d == dentry)
			return 1;
		if (d == NULL || d == d->d_parent)
			return 0;
		d = d->d_parent;
	}
}

699
struct vfsmount *copy_tree(struct vfsmount *mnt, struct dentry *dentry,
R
Ram Pai 已提交
700
					int flag)
L
Linus Torvalds 已提交
701 702 703 704
{
	struct vfsmount *res, *p, *q, *r, *s;
	struct nameidata nd;

R
Ram Pai 已提交
705 706 707
	if (!(flag & CL_COPY_ALL) && IS_MNT_UNBINDABLE(mnt))
		return NULL;

R
Ram Pai 已提交
708
	res = q = clone_mnt(mnt, dentry, flag);
L
Linus Torvalds 已提交
709 710 711 712 713
	if (!q)
		goto Enomem;
	q->mnt_mountpoint = mnt->mnt_mountpoint;

	p = mnt;
714
	list_for_each_entry(r, &mnt->mnt_mounts, mnt_child) {
L
Linus Torvalds 已提交
715 716 717 718
		if (!lives_below_in_same_fs(r->mnt_mountpoint, dentry))
			continue;

		for (s = r; s; s = next_mnt(s, r)) {
R
Ram Pai 已提交
719 720 721 722
			if (!(flag & CL_COPY_ALL) && IS_MNT_UNBINDABLE(s)) {
				s = skip_mnt_tree(s);
				continue;
			}
L
Linus Torvalds 已提交
723 724 725 726 727 728 729
			while (p != s->mnt_parent) {
				p = p->mnt_parent;
				q = q->mnt_parent;
			}
			p = s;
			nd.mnt = q;
			nd.dentry = p->mnt_mountpoint;
R
Ram Pai 已提交
730
			q = clone_mnt(p, p->mnt_root, flag);
L
Linus Torvalds 已提交
731 732 733 734 735 736 737 738 739
			if (!q)
				goto Enomem;
			spin_lock(&vfsmount_lock);
			list_add_tail(&q->mnt_list, &res->mnt_list);
			attach_mnt(q, &nd);
			spin_unlock(&vfsmount_lock);
		}
	}
	return res;
R
Ram Pai 已提交
740
Enomem:
L
Linus Torvalds 已提交
741
	if (res) {
R
Ram Pai 已提交
742
		LIST_HEAD(umount_list);
L
Linus Torvalds 已提交
743
		spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
744
		umount_tree(res, 0, &umount_list);
L
Linus Torvalds 已提交
745
		spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
746
		release_mounts(&umount_list);
L
Linus Torvalds 已提交
747 748 749 750
	}
	return NULL;
}

751 752
/*
 *  @source_mnt : mount tree to be attached
R
Ram Pai 已提交
753 754 755 756
 *  @nd         : place the mount tree @source_mnt is attached
 *  @parent_nd  : if non-null, detach the source_mnt from its parent and
 *  		   store the parent mount and mountpoint dentry.
 *  		   (done when source_mnt is moved)
757 758 759
 *
 *  NOTE: in the table below explains the semantics when a source mount
 *  of a given type is attached to a destination mount of a given type.
R
Ram Pai 已提交
760 761 762 763 764 765 766 767 768 769 770 771
 * ---------------------------------------------------------------------------
 * |         BIND MOUNT OPERATION                                            |
 * |**************************************************************************
 * | source-->| shared        |       private  |       slave    | unbindable |
 * | dest     |               |                |                |            |
 * |   |      |               |                |                |            |
 * |   v      |               |                |                |            |
 * |**************************************************************************
 * |  shared  | shared (++)   |     shared (+) |     shared(+++)|  invalid   |
 * |          |               |                |                |            |
 * |non-shared| shared (+)    |      private   |      slave (*) |  invalid   |
 * ***************************************************************************
772 773 774 775 776 777 778 779 780
 * A bind operation clones the source mount and mounts the clone on the
 * destination mount.
 *
 * (++)  the cloned mount is propagated to all the mounts in the propagation
 * 	 tree of the destination mount and the cloned mount is added to
 * 	 the peer group of the source mount.
 * (+)   the cloned mount is created under the destination mount and is marked
 *       as shared. The cloned mount is added to the peer group of the source
 *       mount.
R
Ram Pai 已提交
781 782 783 784 785 786 787
 * (+++) the mount is propagated to all the mounts in the propagation tree
 *       of the destination mount and the cloned mount is made slave
 *       of the same master as that of the source mount. The cloned mount
 *       is marked as 'shared and slave'.
 * (*)   the cloned mount is made a slave of the same master as that of the
 * 	 source mount.
 *
R
Ram Pai 已提交
788 789 790 791 792 793 794 795 796 797 798 799
 * ---------------------------------------------------------------------------
 * |         		MOVE MOUNT OPERATION                                 |
 * |**************************************************************************
 * | source-->| shared        |       private  |       slave    | unbindable |
 * | dest     |               |                |                |            |
 * |   |      |               |                |                |            |
 * |   v      |               |                |                |            |
 * |**************************************************************************
 * |  shared  | shared (+)    |     shared (+) |    shared(+++) |  invalid   |
 * |          |               |                |                |            |
 * |non-shared| shared (+*)   |      private   |    slave (*)   | unbindable |
 * ***************************************************************************
R
Ram Pai 已提交
800 801 802
 *
 * (+)  the mount is moved to the destination. And is then propagated to
 * 	all the mounts in the propagation tree of the destination mount.
R
Ram Pai 已提交
803
 * (+*)  the mount is moved to the destination.
R
Ram Pai 已提交
804 805 806 807
 * (+++)  the mount is moved to the destination and is then propagated to
 * 	all the mounts belonging to the destination mount's propagation tree.
 * 	the mount is marked as 'shared and slave'.
 * (*)	the mount continues to be a slave at the new location.
808 809 810 811 812 813 814
 *
 * if the source mount is a tree, the operations explained above is
 * applied to each mount in the tree.
 * Must be called without spinlocks held, since this function can sleep
 * in allocations.
 */
static int attach_recursive_mnt(struct vfsmount *source_mnt,
R
Ram Pai 已提交
815
			struct nameidata *nd, struct nameidata *parent_nd)
816 817 818 819 820 821 822 823 824 825 826 827 828 829 830
{
	LIST_HEAD(tree_list);
	struct vfsmount *dest_mnt = nd->mnt;
	struct dentry *dest_dentry = nd->dentry;
	struct vfsmount *child, *p;

	if (propagate_mnt(dest_mnt, dest_dentry, source_mnt, &tree_list))
		return -EINVAL;

	if (IS_MNT_SHARED(dest_mnt)) {
		for (p = source_mnt; p; p = next_mnt(p, source_mnt))
			set_mnt_shared(p);
	}

	spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
831 832 833
	if (parent_nd) {
		detach_mnt(source_mnt, parent_nd);
		attach_mnt(source_mnt, nd);
834
		touch_mnt_namespace(current->nsproxy->mnt_ns);
R
Ram Pai 已提交
835 836 837 838
	} else {
		mnt_set_mountpoint(dest_mnt, dest_dentry, source_mnt);
		commit_tree(source_mnt);
	}
839 840 841 842 843 844 845 846 847

	list_for_each_entry_safe(child, p, &tree_list, mnt_hash) {
		list_del_init(&child->mnt_hash);
		commit_tree(child);
	}
	spin_unlock(&vfsmount_lock);
	return 0;
}

L
Linus Torvalds 已提交
848 849 850 851 852 853 854 855 856 857 858
static int graft_tree(struct vfsmount *mnt, struct nameidata *nd)
{
	int err;
	if (mnt->mnt_sb->s_flags & MS_NOUSER)
		return -EINVAL;

	if (S_ISDIR(nd->dentry->d_inode->i_mode) !=
	      S_ISDIR(mnt->mnt_root->d_inode->i_mode))
		return -ENOTDIR;

	err = -ENOENT;
859
	mutex_lock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
860 861 862 863 864 865 866 867
	if (IS_DEADDIR(nd->dentry->d_inode))
		goto out_unlock;

	err = security_sb_check_sb(mnt, nd);
	if (err)
		goto out_unlock;

	err = -ENOENT;
868
	if (IS_ROOT(nd->dentry) || !d_unhashed(nd->dentry))
R
Ram Pai 已提交
869
		err = attach_recursive_mnt(mnt, nd, NULL);
L
Linus Torvalds 已提交
870
out_unlock:
871
	mutex_unlock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
872 873 874 875 876
	if (!err)
		security_sb_post_addmount(mnt, nd);
	return err;
}

877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897
/*
 * recursively change the type of the mountpoint.
 */
static int do_change_type(struct nameidata *nd, int flag)
{
	struct vfsmount *m, *mnt = nd->mnt;
	int recurse = flag & MS_REC;
	int type = flag & ~MS_REC;

	if (nd->dentry != nd->mnt->mnt_root)
		return -EINVAL;

	down_write(&namespace_sem);
	spin_lock(&vfsmount_lock);
	for (m = mnt; m; m = (recurse ? next_mnt(m, mnt) : NULL))
		change_mnt_propagation(m, type);
	spin_unlock(&vfsmount_lock);
	up_write(&namespace_sem);
	return 0;
}

L
Linus Torvalds 已提交
898 899 900
/*
 * do loopback mount.
 */
901
static int do_loopback(struct nameidata *nd, char *old_name, int recurse)
L
Linus Torvalds 已提交
902 903 904 905 906 907 908 909 910 911 912 913
{
	struct nameidata old_nd;
	struct vfsmount *mnt = NULL;
	int err = mount_is_safe(nd);
	if (err)
		return err;
	if (!old_name || !*old_name)
		return -EINVAL;
	err = path_lookup(old_name, LOOKUP_FOLLOW, &old_nd);
	if (err)
		return err;

R
Ram Pai 已提交
914
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
915
	err = -EINVAL;
R
Ram Pai 已提交
916 917 918
	if (IS_MNT_UNBINDABLE(old_nd.mnt))
 		goto out;

919 920
	if (!check_mnt(nd->mnt) || !check_mnt(old_nd.mnt))
		goto out;
L
Linus Torvalds 已提交
921

922 923
	err = -ENOMEM;
	if (recurse)
R
Ram Pai 已提交
924
		mnt = copy_tree(old_nd.mnt, old_nd.dentry, 0);
925
	else
R
Ram Pai 已提交
926
		mnt = clone_mnt(old_nd.mnt, old_nd.dentry, 0);
927 928 929 930 931 932

	if (!mnt)
		goto out;

	err = graft_tree(mnt, nd);
	if (err) {
R
Ram Pai 已提交
933
		LIST_HEAD(umount_list);
L
Linus Torvalds 已提交
934
		spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
935
		umount_tree(mnt, 0, &umount_list);
L
Linus Torvalds 已提交
936
		spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
937
		release_mounts(&umount_list);
938
	}
L
Linus Torvalds 已提交
939

940
out:
R
Ram Pai 已提交
941
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
942 943 944 945 946 947 948 949 950 951 952 953 954
	path_release(&old_nd);
	return err;
}

/*
 * change filesystem flags. dir should be a physical root of filesystem.
 * If you've mounted a non-root directory somewhere and want to do remount
 * on it - tough luck.
 */
static int do_remount(struct nameidata *nd, int flags, int mnt_flags,
		      void *data)
{
	int err;
R
Ram Pai 已提交
955
	struct super_block *sb = nd->mnt->mnt_sb;
L
Linus Torvalds 已提交
956 957 958 959 960 961 962 963 964 965 966 967 968

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	if (!check_mnt(nd->mnt))
		return -EINVAL;

	if (nd->dentry != nd->mnt->mnt_root)
		return -EINVAL;

	down_write(&sb->s_umount);
	err = do_remount_sb(sb, flags, data, 0);
	if (!err)
R
Ram Pai 已提交
969
		nd->mnt->mnt_flags = mnt_flags;
L
Linus Torvalds 已提交
970 971 972 973 974 975
	up_write(&sb->s_umount);
	if (!err)
		security_sb_post_remount(nd->mnt, flags, data);
	return err;
}

R
Ram Pai 已提交
976 977 978 979 980 981 982 983 984 985
static inline int tree_contains_unbindable(struct vfsmount *mnt)
{
	struct vfsmount *p;
	for (p = mnt; p; p = next_mnt(p, mnt)) {
		if (IS_MNT_UNBINDABLE(p))
			return 1;
	}
	return 0;
}

L
Linus Torvalds 已提交
986 987 988 989 990 991 992 993 994 995 996 997 998
static int do_move_mount(struct nameidata *nd, char *old_name)
{
	struct nameidata old_nd, parent_nd;
	struct vfsmount *p;
	int err = 0;
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;
	if (!old_name || !*old_name)
		return -EINVAL;
	err = path_lookup(old_name, LOOKUP_FOLLOW, &old_nd);
	if (err)
		return err;

R
Ram Pai 已提交
999
	down_write(&namespace_sem);
R
Ram Pai 已提交
1000
	while (d_mountpoint(nd->dentry) && follow_down(&nd->mnt, &nd->dentry))
L
Linus Torvalds 已提交
1001 1002 1003 1004 1005 1006
		;
	err = -EINVAL;
	if (!check_mnt(nd->mnt) || !check_mnt(old_nd.mnt))
		goto out;

	err = -ENOENT;
1007
	mutex_lock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
1008 1009 1010 1011
	if (IS_DEADDIR(nd->dentry->d_inode))
		goto out1;

	if (!IS_ROOT(nd->dentry) && d_unhashed(nd->dentry))
R
Ram Pai 已提交
1012
		goto out1;
L
Linus Torvalds 已提交
1013 1014 1015

	err = -EINVAL;
	if (old_nd.dentry != old_nd.mnt->mnt_root)
R
Ram Pai 已提交
1016
		goto out1;
L
Linus Torvalds 已提交
1017 1018

	if (old_nd.mnt == old_nd.mnt->mnt_parent)
R
Ram Pai 已提交
1019
		goto out1;
L
Linus Torvalds 已提交
1020 1021 1022

	if (S_ISDIR(nd->dentry->d_inode->i_mode) !=
	      S_ISDIR(old_nd.dentry->d_inode->i_mode))
R
Ram Pai 已提交
1023 1024 1025 1026 1027 1028
		goto out1;
	/*
	 * Don't move a mount residing in a shared parent.
	 */
	if (old_nd.mnt->mnt_parent && IS_MNT_SHARED(old_nd.mnt->mnt_parent))
		goto out1;
R
Ram Pai 已提交
1029 1030 1031 1032 1033 1034
	/*
	 * Don't move a mount tree containing unbindable mounts to a destination
	 * mount which is shared.
	 */
	if (IS_MNT_SHARED(nd->mnt) && tree_contains_unbindable(old_nd.mnt))
		goto out1;
L
Linus Torvalds 已提交
1035
	err = -ELOOP;
R
Ram Pai 已提交
1036
	for (p = nd->mnt; p->mnt_parent != p; p = p->mnt_parent)
L
Linus Torvalds 已提交
1037
		if (p == old_nd.mnt)
R
Ram Pai 已提交
1038
			goto out1;
L
Linus Torvalds 已提交
1039

R
Ram Pai 已提交
1040 1041
	if ((err = attach_recursive_mnt(old_nd.mnt, nd, &parent_nd)))
		goto out1;
L
Linus Torvalds 已提交
1042

R
Ram Pai 已提交
1043
	spin_lock(&vfsmount_lock);
L
Linus Torvalds 已提交
1044 1045
	/* if the mount is moved, it should no longer be expire
	 * automatically */
1046
	list_del_init(&old_nd.mnt->mnt_expire);
L
Linus Torvalds 已提交
1047 1048
	spin_unlock(&vfsmount_lock);
out1:
1049
	mutex_unlock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
1050
out:
R
Ram Pai 已提交
1051
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089
	if (!err)
		path_release(&parent_nd);
	path_release(&old_nd);
	return err;
}

/*
 * create a new mount for userspace and request it to be added into the
 * namespace's tree
 */
static int do_new_mount(struct nameidata *nd, char *type, int flags,
			int mnt_flags, char *name, void *data)
{
	struct vfsmount *mnt;

	if (!type || !memchr(type, 0, PAGE_SIZE))
		return -EINVAL;

	/* we need capabilities... */
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	mnt = do_kern_mount(type, flags, name, data);
	if (IS_ERR(mnt))
		return PTR_ERR(mnt);

	return do_add_mount(mnt, nd, mnt_flags, NULL);
}

/*
 * add a mount into a namespace's mount tree
 * - provide the option of adding the new mount to an expiration list
 */
int do_add_mount(struct vfsmount *newmnt, struct nameidata *nd,
		 int mnt_flags, struct list_head *fslist)
{
	int err;

R
Ram Pai 已提交
1090
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
1091
	/* Something was mounted here while we slept */
R
Ram Pai 已提交
1092
	while (d_mountpoint(nd->dentry) && follow_down(&nd->mnt, &nd->dentry))
L
Linus Torvalds 已提交
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108
		;
	err = -EINVAL;
	if (!check_mnt(nd->mnt))
		goto unlock;

	/* Refuse the same filesystem on the same mount point */
	err = -EBUSY;
	if (nd->mnt->mnt_sb == newmnt->mnt_sb &&
	    nd->mnt->mnt_root == nd->dentry)
		goto unlock;

	err = -EINVAL;
	if (S_ISLNK(newmnt->mnt_root->d_inode->i_mode))
		goto unlock;

	newmnt->mnt_flags = mnt_flags;
1109 1110
	if ((err = graft_tree(newmnt, nd)))
		goto unlock;
L
Linus Torvalds 已提交
1111

1112
	if (fslist) {
L
Linus Torvalds 已提交
1113 1114
		/* add to the specified expiration list */
		spin_lock(&vfsmount_lock);
1115
		list_add_tail(&newmnt->mnt_expire, fslist);
L
Linus Torvalds 已提交
1116 1117
		spin_unlock(&vfsmount_lock);
	}
R
Ram Pai 已提交
1118
	up_write(&namespace_sem);
1119
	return 0;
L
Linus Torvalds 已提交
1120 1121

unlock:
R
Ram Pai 已提交
1122
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1123 1124 1125 1126 1127 1128
	mntput(newmnt);
	return err;
}

EXPORT_SYMBOL_GPL(do_add_mount);

R
Ram Pai 已提交
1129 1130
static void expire_mount(struct vfsmount *mnt, struct list_head *mounts,
				struct list_head *umounts)
1131 1132 1133
{
	spin_lock(&vfsmount_lock);

1134 1135 1136 1137 1138 1139 1140 1141 1142
	/*
	 * Check if mount is still attached, if not, let whoever holds it deal
	 * with the sucker
	 */
	if (mnt->mnt_parent == mnt) {
		spin_unlock(&vfsmount_lock);
		return;
	}

1143 1144 1145 1146
	/*
	 * Check that it is still dead: the count should now be 2 - as
	 * contributed by the vfsmount parent and the mntget above
	 */
R
Ram Pai 已提交
1147
	if (!propagate_mount_busy(mnt, 2)) {
1148
		/* delete from the namespace */
1149
		touch_mnt_namespace(mnt->mnt_ns);
1150
		list_del_init(&mnt->mnt_list);
1151
		mnt->mnt_ns = NULL;
R
Ram Pai 已提交
1152
		umount_tree(mnt, 1, umounts);
1153 1154 1155 1156 1157 1158
		spin_unlock(&vfsmount_lock);
	} else {
		/*
		 * Someone brought it back to life whilst we didn't have any
		 * locks held so return it to the expiration list
		 */
1159
		list_add_tail(&mnt->mnt_expire, mounts);
1160 1161 1162 1163
		spin_unlock(&vfsmount_lock);
	}
}

T
Trond Myklebust 已提交
1164 1165 1166 1167 1168 1169 1170 1171
/*
 * go through the vfsmounts we've just consigned to the graveyard to
 * - check that they're still dead
 * - delete the vfsmount from the appropriate namespace under lock
 * - dispose of the corpse
 */
static void expire_mount_list(struct list_head *graveyard, struct list_head *mounts)
{
1172
	struct mnt_namespace *ns;
T
Trond Myklebust 已提交
1173 1174 1175 1176 1177 1178 1179 1180 1181
	struct vfsmount *mnt;

	while (!list_empty(graveyard)) {
		LIST_HEAD(umounts);
		mnt = list_entry(graveyard->next, struct vfsmount, mnt_expire);
		list_del_init(&mnt->mnt_expire);

		/* don't do anything if the namespace is dead - all the
		 * vfsmounts from it are going away anyway */
1182 1183
		ns = mnt->mnt_ns;
		if (!ns || !ns->root)
T
Trond Myklebust 已提交
1184
			continue;
1185
		get_mnt_ns(ns);
T
Trond Myklebust 已提交
1186 1187 1188 1189 1190 1191 1192

		spin_unlock(&vfsmount_lock);
		down_write(&namespace_sem);
		expire_mount(mnt, mounts, &umounts);
		up_write(&namespace_sem);
		release_mounts(&umounts);
		mntput(mnt);
1193
		put_mnt_ns(ns);
T
Trond Myklebust 已提交
1194 1195 1196 1197
		spin_lock(&vfsmount_lock);
	}
}

L
Linus Torvalds 已提交
1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
/*
 * process a list of expirable mountpoints with the intent of discarding any
 * mountpoints that aren't in use and haven't been touched since last we came
 * here
 */
void mark_mounts_for_expiry(struct list_head *mounts)
{
	struct vfsmount *mnt, *next;
	LIST_HEAD(graveyard);

	if (list_empty(mounts))
		return;

	spin_lock(&vfsmount_lock);

	/* extract from the expiration list every vfsmount that matches the
	 * following criteria:
	 * - only referenced by its parent vfsmount
	 * - still marked for expiry (marked on the last call here; marks are
	 *   cleared by mntput())
	 */
1219
	list_for_each_entry_safe(mnt, next, mounts, mnt_expire) {
L
Linus Torvalds 已提交
1220 1221 1222 1223 1224
		if (!xchg(&mnt->mnt_expiry_mark, 1) ||
		    atomic_read(&mnt->mnt_count) != 1)
			continue;

		mntget(mnt);
1225
		list_move(&mnt->mnt_expire, &graveyard);
L
Linus Torvalds 已提交
1226 1227
	}

T
Trond Myklebust 已提交
1228
	expire_mount_list(&graveyard, mounts);
L
Linus Torvalds 已提交
1229

T
Trond Myklebust 已提交
1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255
	spin_unlock(&vfsmount_lock);
}

EXPORT_SYMBOL_GPL(mark_mounts_for_expiry);

/*
 * Ripoff of 'select_parent()'
 *
 * search the list of submounts for a given mountpoint, and move any
 * shrinkable submounts to the 'graveyard' list.
 */
static int select_submounts(struct vfsmount *parent, struct list_head *graveyard)
{
	struct vfsmount *this_parent = parent;
	struct list_head *next;
	int found = 0;

repeat:
	next = this_parent->mnt_mounts.next;
resume:
	while (next != &this_parent->mnt_mounts) {
		struct list_head *tmp = next;
		struct vfsmount *mnt = list_entry(tmp, struct vfsmount, mnt_child);

		next = tmp->next;
		if (!(mnt->mnt_flags & MNT_SHRINKABLE))
L
Linus Torvalds 已提交
1256
			continue;
T
Trond Myklebust 已提交
1257 1258 1259 1260 1261 1262 1263
		/*
		 * Descend a level if the d_mounts list is non-empty.
		 */
		if (!list_empty(&mnt->mnt_mounts)) {
			this_parent = mnt;
			goto repeat;
		}
L
Linus Torvalds 已提交
1264

T
Trond Myklebust 已提交
1265 1266 1267 1268 1269
		if (!propagate_mount_busy(mnt, 1)) {
			mntget(mnt);
			list_move_tail(&mnt->mnt_expire, graveyard);
			found++;
		}
L
Linus Torvalds 已提交
1270
	}
T
Trond Myklebust 已提交
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295
	/*
	 * All done at this level ... ascend and resume the search
	 */
	if (this_parent != parent) {
		next = this_parent->mnt_child.next;
		this_parent = this_parent->mnt_parent;
		goto resume;
	}
	return found;
}

/*
 * process a list of expirable mountpoints with the intent of discarding any
 * submounts of a specific parent mountpoint
 */
void shrink_submounts(struct vfsmount *mountpoint, struct list_head *mounts)
{
	LIST_HEAD(graveyard);
	int found;

	spin_lock(&vfsmount_lock);

	/* extract submounts of 'mountpoint' from the expiration list */
	while ((found = select_submounts(mountpoint, &graveyard)) != 0)
		expire_mount_list(&graveyard, mounts);
L
Linus Torvalds 已提交
1296 1297 1298 1299

	spin_unlock(&vfsmount_lock);
}

T
Trond Myklebust 已提交
1300
EXPORT_SYMBOL_GPL(shrink_submounts);
L
Linus Torvalds 已提交
1301 1302 1303 1304 1305 1306 1307

/*
 * Some copy_from_user() implementations do not return the exact number of
 * bytes remaining to copy on a fault.  But copy_mount_options() requires that.
 * Note that this function differs from copy_from_user() in that it will oops
 * on bad values of `to', rather than returning a short copy.
 */
R
Ram Pai 已提交
1308 1309
static long exact_copy_from_user(void *to, const void __user * from,
				 unsigned long n)
L
Linus Torvalds 已提交
1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329
{
	char *t = to;
	const char __user *f = from;
	char c;

	if (!access_ok(VERIFY_READ, from, n))
		return n;

	while (n) {
		if (__get_user(c, f)) {
			memset(t, 0, n);
			break;
		}
		*t++ = c;
		f++;
		n--;
	}
	return n;
}

R
Ram Pai 已提交
1330
int copy_mount_options(const void __user * data, unsigned long *where)
L
Linus Torvalds 已提交
1331 1332 1333 1334
{
	int i;
	unsigned long page;
	unsigned long size;
R
Ram Pai 已提交
1335

L
Linus Torvalds 已提交
1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353
	*where = 0;
	if (!data)
		return 0;

	if (!(page = __get_free_page(GFP_KERNEL)))
		return -ENOMEM;

	/* We only care that *some* data at the address the user
	 * gave us is valid.  Just in case, we'll zero
	 * the remainder of the page.
	 */
	/* copy_from_user cannot cross TASK_SIZE ! */
	size = TASK_SIZE - (unsigned long)data;
	if (size > PAGE_SIZE)
		size = PAGE_SIZE;

	i = size - exact_copy_from_user((void *)page, data, size);
	if (!i) {
R
Ram Pai 已提交
1354
		free_page(page);
L
Linus Torvalds 已提交
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
		return -EFAULT;
	}
	if (i != PAGE_SIZE)
		memset((char *)page + i, 0, PAGE_SIZE - i);
	*where = page;
	return 0;
}

/*
 * Flags is a 32-bit value that allows up to 31 non-fs dependent flags to
 * be given to the mount() call (ie: read-only, no-dev, no-suid etc).
 *
 * data is a (void *) that can point to any structure up to
 * PAGE_SIZE-1 bytes, which can contain arbitrary fs-dependent
 * information (or be NULL).
 *
 * Pre-0.97 versions of mount() didn't have a flags word.
 * When the flags word was introduced its top half was required
 * to have the magic value 0xC0ED, and this remained so until 2.4.0-test9.
 * Therefore, if this magic number is present, it carries no information
 * and must be discarded.
 */
R
Ram Pai 已提交
1377
long do_mount(char *dev_name, char *dir_name, char *type_page,
L
Linus Torvalds 已提交
1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404
		  unsigned long flags, void *data_page)
{
	struct nameidata nd;
	int retval = 0;
	int mnt_flags = 0;

	/* Discard magic */
	if ((flags & MS_MGC_MSK) == MS_MGC_VAL)
		flags &= ~MS_MGC_MSK;

	/* Basic sanity checks */

	if (!dir_name || !*dir_name || !memchr(dir_name, 0, PAGE_SIZE))
		return -EINVAL;
	if (dev_name && !memchr(dev_name, 0, PAGE_SIZE))
		return -EINVAL;

	if (data_page)
		((char *)data_page)[PAGE_SIZE - 1] = 0;

	/* Separate the per-mountpoint flags */
	if (flags & MS_NOSUID)
		mnt_flags |= MNT_NOSUID;
	if (flags & MS_NODEV)
		mnt_flags |= MNT_NODEV;
	if (flags & MS_NOEXEC)
		mnt_flags |= MNT_NOEXEC;
1405 1406 1407 1408
	if (flags & MS_NOATIME)
		mnt_flags |= MNT_NOATIME;
	if (flags & MS_NODIRATIME)
		mnt_flags |= MNT_NODIRATIME;
V
Valerie Henson 已提交
1409 1410
	if (flags & MS_RELATIME)
		mnt_flags |= MNT_RELATIME;
1411 1412

	flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | MS_ACTIVE |
V
Valerie Henson 已提交
1413
		   MS_NOATIME | MS_NODIRATIME | MS_RELATIME);
L
Linus Torvalds 已提交
1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427

	/* ... and get the mountpoint */
	retval = path_lookup(dir_name, LOOKUP_FOLLOW, &nd);
	if (retval)
		return retval;

	retval = security_sb_mount(dev_name, &nd, type_page, flags, data_page);
	if (retval)
		goto dput_out;

	if (flags & MS_REMOUNT)
		retval = do_remount(&nd, flags & ~MS_REMOUNT, mnt_flags,
				    data_page);
	else if (flags & MS_BIND)
1428
		retval = do_loopback(&nd, dev_name, flags & MS_REC);
R
Ram Pai 已提交
1429
	else if (flags & (MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE))
1430
		retval = do_change_type(&nd, flags);
L
Linus Torvalds 已提交
1431 1432 1433 1434 1435 1436 1437 1438 1439 1440
	else if (flags & MS_MOVE)
		retval = do_move_mount(&nd, dev_name);
	else
		retval = do_new_mount(&nd, type_page, flags, mnt_flags,
				      dev_name, data_page);
dput_out:
	path_release(&nd);
	return retval;
}

1441 1442 1443 1444
/*
 * Allocate a new namespace structure and populate it with contents
 * copied from the namespace of the passed in task structure.
 */
1445 1446
struct mnt_namespace *dup_mnt_ns(struct task_struct *tsk,
		struct fs_struct *fs)
L
Linus Torvalds 已提交
1447
{
1448 1449
	struct mnt_namespace *mnt_ns = tsk->nsproxy->mnt_ns;
	struct mnt_namespace *new_ns;
L
Linus Torvalds 已提交
1450 1451 1452
	struct vfsmount *rootmnt = NULL, *pwdmnt = NULL, *altrootmnt = NULL;
	struct vfsmount *p, *q;

1453
	new_ns = kmalloc(sizeof(struct mnt_namespace), GFP_KERNEL);
L
Linus Torvalds 已提交
1454
	if (!new_ns)
1455
		return NULL;
L
Linus Torvalds 已提交
1456 1457 1458

	atomic_set(&new_ns->count, 1);
	INIT_LIST_HEAD(&new_ns->list);
A
Al Viro 已提交
1459 1460
	init_waitqueue_head(&new_ns->poll);
	new_ns->event = 0;
L
Linus Torvalds 已提交
1461

R
Ram Pai 已提交
1462
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
1463
	/* First pass: copy the tree topology */
1464
	new_ns->root = copy_tree(mnt_ns->root, mnt_ns->root->mnt_root,
R
Ram Pai 已提交
1465
					CL_COPY_ALL | CL_EXPIRE);
L
Linus Torvalds 已提交
1466
	if (!new_ns->root) {
R
Ram Pai 已提交
1467
		up_write(&namespace_sem);
L
Linus Torvalds 已提交
1468
		kfree(new_ns);
1469
		return NULL;
L
Linus Torvalds 已提交
1470 1471 1472 1473 1474 1475 1476 1477 1478 1479
	}
	spin_lock(&vfsmount_lock);
	list_add_tail(&new_ns->list, &new_ns->root->mnt_list);
	spin_unlock(&vfsmount_lock);

	/*
	 * Second pass: switch the tsk->fs->* elements and mark new vfsmounts
	 * as belonging to new namespace.  We have already acquired a private
	 * fs_struct, so tsk->fs->lock is not needed.
	 */
1480
	p = mnt_ns->root;
L
Linus Torvalds 已提交
1481 1482
	q = new_ns->root;
	while (p) {
1483
		q->mnt_ns = new_ns;
L
Linus Torvalds 已提交
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497
		if (fs) {
			if (p == fs->rootmnt) {
				rootmnt = p;
				fs->rootmnt = mntget(q);
			}
			if (p == fs->pwdmnt) {
				pwdmnt = p;
				fs->pwdmnt = mntget(q);
			}
			if (p == fs->altrootmnt) {
				altrootmnt = p;
				fs->altrootmnt = mntget(q);
			}
		}
1498
		p = next_mnt(p, mnt_ns->root);
L
Linus Torvalds 已提交
1499 1500
		q = next_mnt(q, new_ns->root);
	}
R
Ram Pai 已提交
1501
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1502 1503 1504 1505 1506 1507 1508 1509

	if (rootmnt)
		mntput(rootmnt);
	if (pwdmnt)
		mntput(pwdmnt);
	if (altrootmnt)
		mntput(altrootmnt);

1510 1511 1512
	return new_ns;
}

1513
int copy_mnt_ns(int flags, struct task_struct *tsk)
1514
{
1515 1516
	struct mnt_namespace *ns = tsk->nsproxy->mnt_ns;
	struct mnt_namespace *new_ns;
1517 1518
	int err = 0;

1519
	if (!ns)
1520 1521
		return 0;

1522
	get_mnt_ns(ns);
1523 1524 1525 1526 1527 1528 1529 1530 1531

	if (!(flags & CLONE_NEWNS))
		return 0;

	if (!capable(CAP_SYS_ADMIN)) {
		err = -EPERM;
		goto out;
	}

1532
	new_ns = dup_mnt_ns(tsk, tsk->fs);
1533 1534 1535 1536 1537
	if (!new_ns) {
		err = -ENOMEM;
		goto out;
	}

1538
	tsk->nsproxy->mnt_ns = new_ns;
L
Linus Torvalds 已提交
1539 1540

out:
1541
	put_mnt_ns(ns);
1542
	return err;
L
Linus Torvalds 已提交
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554
}

asmlinkage long sys_mount(char __user * dev_name, char __user * dir_name,
			  char __user * type, unsigned long flags,
			  void __user * data)
{
	int retval;
	unsigned long data_page;
	unsigned long type_page;
	unsigned long dev_page;
	char *dir_page;

R
Ram Pai 已提交
1555
	retval = copy_mount_options(type, &type_page);
L
Linus Torvalds 已提交
1556 1557 1558 1559 1560 1561 1562 1563
	if (retval < 0)
		return retval;

	dir_page = getname(dir_name);
	retval = PTR_ERR(dir_page);
	if (IS_ERR(dir_page))
		goto out1;

R
Ram Pai 已提交
1564
	retval = copy_mount_options(dev_name, &dev_page);
L
Linus Torvalds 已提交
1565 1566 1567
	if (retval < 0)
		goto out2;

R
Ram Pai 已提交
1568
	retval = copy_mount_options(data, &data_page);
L
Linus Torvalds 已提交
1569 1570 1571 1572
	if (retval < 0)
		goto out3;

	lock_kernel();
R
Ram Pai 已提交
1573 1574
	retval = do_mount((char *)dev_page, dir_page, (char *)type_page,
			  flags, (void *)data_page);
L
Linus Torvalds 已提交
1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642
	unlock_kernel();
	free_page(data_page);

out3:
	free_page(dev_page);
out2:
	putname(dir_page);
out1:
	free_page(type_page);
	return retval;
}

/*
 * Replace the fs->{rootmnt,root} with {mnt,dentry}. Put the old values.
 * It can block. Requires the big lock held.
 */
void set_fs_root(struct fs_struct *fs, struct vfsmount *mnt,
		 struct dentry *dentry)
{
	struct dentry *old_root;
	struct vfsmount *old_rootmnt;
	write_lock(&fs->lock);
	old_root = fs->root;
	old_rootmnt = fs->rootmnt;
	fs->rootmnt = mntget(mnt);
	fs->root = dget(dentry);
	write_unlock(&fs->lock);
	if (old_root) {
		dput(old_root);
		mntput(old_rootmnt);
	}
}

/*
 * Replace the fs->{pwdmnt,pwd} with {mnt,dentry}. Put the old values.
 * It can block. Requires the big lock held.
 */
void set_fs_pwd(struct fs_struct *fs, struct vfsmount *mnt,
		struct dentry *dentry)
{
	struct dentry *old_pwd;
	struct vfsmount *old_pwdmnt;

	write_lock(&fs->lock);
	old_pwd = fs->pwd;
	old_pwdmnt = fs->pwdmnt;
	fs->pwdmnt = mntget(mnt);
	fs->pwd = dget(dentry);
	write_unlock(&fs->lock);

	if (old_pwd) {
		dput(old_pwd);
		mntput(old_pwdmnt);
	}
}

static void chroot_fs_refs(struct nameidata *old_nd, struct nameidata *new_nd)
{
	struct task_struct *g, *p;
	struct fs_struct *fs;

	read_lock(&tasklist_lock);
	do_each_thread(g, p) {
		task_lock(p);
		fs = p->fs;
		if (fs) {
			atomic_inc(&fs->count);
			task_unlock(p);
R
Ram Pai 已提交
1643 1644
			if (fs->root == old_nd->dentry
			    && fs->rootmnt == old_nd->mnt)
L
Linus Torvalds 已提交
1645
				set_fs_root(fs, new_nd->mnt, new_nd->dentry);
R
Ram Pai 已提交
1646 1647
			if (fs->pwd == old_nd->dentry
			    && fs->pwdmnt == old_nd->mnt)
L
Linus Torvalds 已提交
1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668
				set_fs_pwd(fs, new_nd->mnt, new_nd->dentry);
			put_fs_struct(fs);
		} else
			task_unlock(p);
	} while_each_thread(g, p);
	read_unlock(&tasklist_lock);
}

/*
 * pivot_root Semantics:
 * Moves the root file system of the current process to the directory put_old,
 * makes new_root as the new root file system of the current process, and sets
 * root/cwd of all processes which had them on the current root to new_root.
 *
 * Restrictions:
 * The new_root and put_old must be directories, and  must not be on the
 * same file  system as the current process root. The put_old  must  be
 * underneath new_root,  i.e. adding a non-zero number of /.. to the string
 * pointed to by put_old must yield the same directory as new_root. No other
 * file system may be mounted on put_old. After all, new_root is a mountpoint.
 *
N
Neil Brown 已提交
1669 1670 1671 1672
 * Also, the current root cannot be on the 'rootfs' (initial ramfs) filesystem.
 * See Documentation/filesystems/ramfs-rootfs-initramfs.txt for alternatives
 * in this situation.
 *
L
Linus Torvalds 已提交
1673 1674 1675 1676 1677 1678 1679 1680
 * Notes:
 *  - we don't move root/cwd if they are not at the root (reason: if something
 *    cared enough to change them, it's probably wrong to force them elsewhere)
 *  - it's okay to pick a root that isn't the root of a file system, e.g.
 *    /nfs/my_root where /nfs is the mount point. It must be a mountpoint,
 *    though, so you may need to say mount --bind /nfs/my_root /nfs/my_root
 *    first.
 */
R
Ram Pai 已提交
1681 1682
asmlinkage long sys_pivot_root(const char __user * new_root,
			       const char __user * put_old)
L
Linus Torvalds 已提交
1683 1684 1685 1686 1687 1688 1689 1690 1691 1692
{
	struct vfsmount *tmp;
	struct nameidata new_nd, old_nd, parent_nd, root_parent, user_nd;
	int error;

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	lock_kernel();

R
Ram Pai 已提交
1693 1694
	error = __user_walk(new_root, LOOKUP_FOLLOW | LOOKUP_DIRECTORY,
			    &new_nd);
L
Linus Torvalds 已提交
1695 1696 1697 1698 1699 1700
	if (error)
		goto out0;
	error = -EINVAL;
	if (!check_mnt(new_nd.mnt))
		goto out1;

R
Ram Pai 已提交
1701
	error = __user_walk(put_old, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &old_nd);
L
Linus Torvalds 已提交
1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714
	if (error)
		goto out1;

	error = security_sb_pivotroot(&old_nd, &new_nd);
	if (error) {
		path_release(&old_nd);
		goto out1;
	}

	read_lock(&current->fs->lock);
	user_nd.mnt = mntget(current->fs->rootmnt);
	user_nd.dentry = dget(current->fs->root);
	read_unlock(&current->fs->lock);
R
Ram Pai 已提交
1715
	down_write(&namespace_sem);
1716
	mutex_lock(&old_nd.dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
1717
	error = -EINVAL;
R
Ram Pai 已提交
1718 1719 1720 1721
	if (IS_MNT_SHARED(old_nd.mnt) ||
		IS_MNT_SHARED(new_nd.mnt->mnt_parent) ||
		IS_MNT_SHARED(user_nd.mnt->mnt_parent))
		goto out2;
L
Linus Torvalds 已提交
1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736
	if (!check_mnt(user_nd.mnt))
		goto out2;
	error = -ENOENT;
	if (IS_DEADDIR(new_nd.dentry->d_inode))
		goto out2;
	if (d_unhashed(new_nd.dentry) && !IS_ROOT(new_nd.dentry))
		goto out2;
	if (d_unhashed(old_nd.dentry) && !IS_ROOT(old_nd.dentry))
		goto out2;
	error = -EBUSY;
	if (new_nd.mnt == user_nd.mnt || old_nd.mnt == user_nd.mnt)
		goto out2; /* loop, on the same file system  */
	error = -EINVAL;
	if (user_nd.mnt->mnt_root != user_nd.dentry)
		goto out2; /* not a mountpoint */
1737 1738
	if (user_nd.mnt->mnt_parent == user_nd.mnt)
		goto out2; /* not attached */
L
Linus Torvalds 已提交
1739 1740
	if (new_nd.mnt->mnt_root != new_nd.dentry)
		goto out2; /* not a mountpoint */
1741 1742
	if (new_nd.mnt->mnt_parent == new_nd.mnt)
		goto out2; /* not attached */
L
Linus Torvalds 已提交
1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760
	tmp = old_nd.mnt; /* make sure we can reach put_old from new_root */
	spin_lock(&vfsmount_lock);
	if (tmp != new_nd.mnt) {
		for (;;) {
			if (tmp->mnt_parent == tmp)
				goto out3; /* already mounted on put_old */
			if (tmp->mnt_parent == new_nd.mnt)
				break;
			tmp = tmp->mnt_parent;
		}
		if (!is_subdir(tmp->mnt_mountpoint, new_nd.dentry))
			goto out3;
	} else if (!is_subdir(old_nd.dentry, new_nd.dentry))
		goto out3;
	detach_mnt(new_nd.mnt, &parent_nd);
	detach_mnt(user_nd.mnt, &root_parent);
	attach_mnt(user_nd.mnt, &old_nd);     /* mount old root on put_old */
	attach_mnt(new_nd.mnt, &root_parent); /* mount new_root on / */
1761
	touch_mnt_namespace(current->nsproxy->mnt_ns);
L
Linus Torvalds 已提交
1762 1763 1764 1765 1766 1767 1768
	spin_unlock(&vfsmount_lock);
	chroot_fs_refs(&user_nd, &new_nd);
	security_sb_post_pivotroot(&user_nd, &new_nd);
	error = 0;
	path_release(&root_parent);
	path_release(&parent_nd);
out2:
1769
	mutex_unlock(&old_nd.dentry->d_inode->i_mutex);
R
Ram Pai 已提交
1770
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785
	path_release(&user_nd);
	path_release(&old_nd);
out1:
	path_release(&new_nd);
out0:
	unlock_kernel();
	return error;
out3:
	spin_unlock(&vfsmount_lock);
	goto out2;
}

static void __init init_mount_tree(void)
{
	struct vfsmount *mnt;
1786
	struct mnt_namespace *ns;
L
Linus Torvalds 已提交
1787 1788 1789 1790

	mnt = do_kern_mount("rootfs", 0, "rootfs", NULL);
	if (IS_ERR(mnt))
		panic("Can't create rootfs");
1791 1792
	ns = kmalloc(sizeof(*ns), GFP_KERNEL);
	if (!ns)
L
Linus Torvalds 已提交
1793
		panic("Can't allocate initial namespace");
1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806
	atomic_set(&ns->count, 1);
	INIT_LIST_HEAD(&ns->list);
	init_waitqueue_head(&ns->poll);
	ns->event = 0;
	list_add(&mnt->mnt_list, &ns->list);
	ns->root = mnt;
	mnt->mnt_ns = ns;

	init_task.nsproxy->mnt_ns = ns;
	get_mnt_ns(ns);

	set_fs_pwd(current->fs, ns->root, ns->root->mnt_root);
	set_fs_root(current->fs, ns->root, ns->root->mnt_root);
L
Linus Torvalds 已提交
1807 1808 1809 1810 1811 1812 1813
}

void __init mnt_init(unsigned long mempages)
{
	struct list_head *d;
	unsigned int nr_hash;
	int i;
1814
	int err;
L
Linus Torvalds 已提交
1815

R
Ram Pai 已提交
1816 1817
	init_rwsem(&namespace_sem);

L
Linus Torvalds 已提交
1818
	mnt_cache = kmem_cache_create("mnt_cache", sizeof(struct vfsmount),
R
Ram Pai 已提交
1819
			0, SLAB_HWCACHE_ALIGN | SLAB_PANIC, NULL, NULL);
L
Linus Torvalds 已提交
1820

R
Ram Pai 已提交
1821
	mount_hashtable = (struct list_head *)__get_free_page(GFP_ATOMIC);
L
Linus Torvalds 已提交
1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842

	if (!mount_hashtable)
		panic("Failed to allocate mount hash table\n");

	/*
	 * Find the power-of-two list-heads that can fit into the allocation..
	 * We don't guarantee that "sizeof(struct list_head)" is necessarily
	 * a power-of-two.
	 */
	nr_hash = PAGE_SIZE / sizeof(struct list_head);
	hash_bits = 0;
	do {
		hash_bits++;
	} while ((nr_hash >> hash_bits) != 0);
	hash_bits--;

	/*
	 * Re-calculate the actual number of entries and the mask
	 * from the number of bits we can fit.
	 */
	nr_hash = 1UL << hash_bits;
R
Ram Pai 已提交
1843
	hash_mask = nr_hash - 1;
L
Linus Torvalds 已提交
1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854

	printk("Mount-cache hash table entries: %d\n", nr_hash);

	/* And initialize the newly allocated array */
	d = mount_hashtable;
	i = nr_hash;
	do {
		INIT_LIST_HEAD(d);
		d++;
		i--;
	} while (i);
1855 1856 1857 1858 1859 1860 1861 1862
	err = sysfs_init();
	if (err)
		printk(KERN_WARNING "%s: sysfs_init error: %d\n",
			__FUNCTION__, err);
	err = subsystem_register(&fs_subsys);
	if (err)
		printk(KERN_WARNING "%s: subsystem_register error: %d\n",
			__FUNCTION__, err);
L
Linus Torvalds 已提交
1863 1864 1865 1866
	init_rootfs();
	init_mount_tree();
}

1867
void __put_mnt_ns(struct mnt_namespace *ns)
L
Linus Torvalds 已提交
1868
{
1869
	struct vfsmount *root = ns->root;
R
Ram Pai 已提交
1870
	LIST_HEAD(umount_list);
1871
	ns->root = NULL;
1872
	spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
1873
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
1874
	spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
1875
	umount_tree(root, 0, &umount_list);
L
Linus Torvalds 已提交
1876
	spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
1877
	up_write(&namespace_sem);
R
Ram Pai 已提交
1878
	release_mounts(&umount_list);
1879
	kfree(ns);
L
Linus Torvalds 已提交
1880
}