namespace.c 47.0 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 23 24 25 26
#include <linux/seq_file.h>
#include <linux/namespace.h>
#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_namespace == current->nsproxy->namespace;
L
Linus Torvalds 已提交
137 138
}

A
Al Viro 已提交
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
static void touch_namespace(struct namespace *ns)
{
	if (ns) {
		ns->event = ++event;
		wake_up_interruptible(&ns->poll);
	}
}

static void __touch_namespace(struct namespace *ns)
{
	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);
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202
}

/*
 * 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);
	struct namespace *n = parent->mnt_namespace;

	BUG_ON(parent == mnt);

	list_add_tail(&head, &mnt->mnt_list);
	list_for_each_entry(m, &head, mnt_list)
		m->mnt_namespace = n;
	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);
	touch_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
319 320 321 322 323 324 325 326

/* iterator */
static void *m_start(struct seq_file *m, loff_t *pos)
{
	struct namespace *n = m->private;
	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
328 329 330 331 332 333 334 335 336 337 338
	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)
{
	struct namespace *n = m->private;
	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" },
L
Linus Torvalds 已提交
371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
		{ 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
};

402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439
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
440 441 442 443 444 445 446 447 448 449
/**
 * 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
450 451 452
	int actual_refs = 0;
	int minimum_refs = 0;
	struct vfsmount *p;
L
Linus Torvalds 已提交
453 454

	spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
455
	for (p = mnt; p; p = next_mnt(p, mnt)) {
L
Linus Torvalds 已提交
456 457 458 459 460 461
		actual_refs += atomic_read(&p->mnt_count);
		minimum_refs += 2;
	}
	spin_unlock(&vfsmount_lock);

	if (actual_refs > minimum_refs)
462
		return 0;
L
Linus Torvalds 已提交
463

464
	return 1;
L
Linus Torvalds 已提交
465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483
}

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)
{
484
	int ret = 1;
R
Ram Pai 已提交
485 486
	spin_lock(&vfsmount_lock);
	if (propagate_mount_busy(mnt, 2))
487
		ret = 0;
R
Ram Pai 已提交
488 489
	spin_unlock(&vfsmount_lock);
	return ret;
L
Linus Torvalds 已提交
490 491 492 493
}

EXPORT_SYMBOL(may_umount);

494
void release_mounts(struct list_head *head)
R
Ram Pai 已提交
495 496
{
	struct vfsmount *mnt;
M
Miklos Szeredi 已提交
497
	while (!list_empty(head)) {
R
Ram Pai 已提交
498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
		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
516
void umount_tree(struct vfsmount *mnt, int propagate, struct list_head *kill)
L
Linus Torvalds 已提交
517 518 519
{
	struct vfsmount *p;

A
Akinobu Mita 已提交
520 521
	for (p = mnt; p; p = next_mnt(p, mnt))
		list_move(&p->mnt_hash, kill);
L
Linus Torvalds 已提交
522

R
Ram Pai 已提交
523 524 525
	if (propagate)
		propagate_umount(kill);

R
Ram Pai 已提交
526 527 528 529 530 531 532
	list_for_each_entry(p, kill, mnt_hash) {
		list_del_init(&p->mnt_expire);
		list_del_init(&p->mnt_list);
		__touch_namespace(p->mnt_namespace);
		p->mnt_namespace = NULL;
		list_del_init(&p->mnt_child);
		if (p->mnt_parent != p)
533
			p->mnt_mountpoint->d_mounted--;
R
Ram Pai 已提交
534
		change_mnt_propagation(p, MS_PRIVATE);
L
Linus Torvalds 已提交
535 536 537 538 539
	}
}

static int do_umount(struct vfsmount *mnt, int flags)
{
R
Ram Pai 已提交
540
	struct super_block *sb = mnt->mnt_sb;
L
Linus Torvalds 已提交
541
	int retval;
R
Ram Pai 已提交
542
	LIST_HEAD(umount_list);
L
Linus Torvalds 已提交
543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576

	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();
577 578
	if (sb->s_op->umount_begin)
		sb->s_op->umount_begin(mnt, flags);
L
Linus Torvalds 已提交
579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605
	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
606
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
607
	spin_lock(&vfsmount_lock);
A
Al Viro 已提交
608
	event++;
L
Linus Torvalds 已提交
609 610

	retval = -EBUSY;
R
Ram Pai 已提交
611
	if (flags & MNT_DETACH || !propagate_mount_busy(mnt, 2)) {
L
Linus Torvalds 已提交
612
		if (!list_empty(&mnt->mnt_list))
R
Ram Pai 已提交
613
			umount_tree(mnt, 1, &umount_list);
L
Linus Torvalds 已提交
614 615 616 617 618
		retval = 0;
	}
	spin_unlock(&vfsmount_lock);
	if (retval)
		security_sb_umount_busy(mnt);
R
Ram Pai 已提交
619
	up_write(&namespace_sem);
R
Ram Pai 已提交
620
	release_mounts(&umount_list);
L
Linus Torvalds 已提交
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659
	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
660
 *	The 2.0 compatible umount. No flags.
L
Linus Torvalds 已提交
661 662 663
 */
asmlinkage long sys_oldumount(char __user * name)
{
R
Ram Pai 已提交
664
	return sys_umount(name, 0);
L
Linus Torvalds 已提交
665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680
}

#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;
	}
681
	if (vfs_permission(nd, MAY_WRITE))
L
Linus Torvalds 已提交
682 683 684 685 686
		return -EPERM;
	return 0;
#endif
}

R
Ram Pai 已提交
687
static int lives_below_in_same_fs(struct dentry *d, struct dentry *dentry)
L
Linus Torvalds 已提交
688 689 690 691 692 693 694 695 696 697
{
	while (1) {
		if (d == dentry)
			return 1;
		if (d == NULL || d == d->d_parent)
			return 0;
		d = d->d_parent;
	}
}

698
struct vfsmount *copy_tree(struct vfsmount *mnt, struct dentry *dentry,
R
Ram Pai 已提交
699
					int flag)
L
Linus Torvalds 已提交
700 701 702 703
{
	struct vfsmount *res, *p, *q, *r, *s;
	struct nameidata nd;

R
Ram Pai 已提交
704 705 706
	if (!(flag & CL_COPY_ALL) && IS_MNT_UNBINDABLE(mnt))
		return NULL;

R
Ram Pai 已提交
707
	res = q = clone_mnt(mnt, dentry, flag);
L
Linus Torvalds 已提交
708 709 710 711 712
	if (!q)
		goto Enomem;
	q->mnt_mountpoint = mnt->mnt_mountpoint;

	p = mnt;
713
	list_for_each_entry(r, &mnt->mnt_mounts, mnt_child) {
L
Linus Torvalds 已提交
714 715 716 717
		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
718 719 720 721
			if (!(flag & CL_COPY_ALL) && IS_MNT_UNBINDABLE(s)) {
				s = skip_mnt_tree(s);
				continue;
			}
L
Linus Torvalds 已提交
722 723 724 725 726 727 728
			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
729
			q = clone_mnt(p, p->mnt_root, flag);
L
Linus Torvalds 已提交
730 731 732 733 734 735 736 737 738
			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
739
Enomem:
L
Linus Torvalds 已提交
740
	if (res) {
R
Ram Pai 已提交
741
		LIST_HEAD(umount_list);
L
Linus Torvalds 已提交
742
		spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
743
		umount_tree(res, 0, &umount_list);
L
Linus Torvalds 已提交
744
		spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
745
		release_mounts(&umount_list);
L
Linus Torvalds 已提交
746 747 748 749
	}
	return NULL;
}

750 751
/*
 *  @source_mnt : mount tree to be attached
R
Ram Pai 已提交
752 753 754 755
 *  @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)
756 757 758
 *
 *  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
759 760 761 762 763 764 765 766 767 768 769 770
 * ---------------------------------------------------------------------------
 * |         BIND MOUNT OPERATION                                            |
 * |**************************************************************************
 * | source-->| shared        |       private  |       slave    | unbindable |
 * | dest     |               |                |                |            |
 * |   |      |               |                |                |            |
 * |   v      |               |                |                |            |
 * |**************************************************************************
 * |  shared  | shared (++)   |     shared (+) |     shared(+++)|  invalid   |
 * |          |               |                |                |            |
 * |non-shared| shared (+)    |      private   |      slave (*) |  invalid   |
 * ***************************************************************************
771 772 773 774 775 776 777 778 779
 * 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
780 781 782 783 784 785 786
 * (+++) 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
787 788 789 790 791 792 793 794 795 796 797 798
 * ---------------------------------------------------------------------------
 * |         		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
799 800 801
 *
 * (+)  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
802
 * (+*)  the mount is moved to the destination.
R
Ram Pai 已提交
803 804 805 806
 * (+++)  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.
807 808 809 810 811 812 813
 *
 * 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
814
			struct nameidata *nd, struct nameidata *parent_nd)
815 816 817 818 819 820 821 822 823 824 825 826 827 828 829
{
	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
830 831 832
	if (parent_nd) {
		detach_mnt(source_mnt, parent_nd);
		attach_mnt(source_mnt, nd);
833
		touch_namespace(current->nsproxy->namespace);
R
Ram Pai 已提交
834 835 836 837
	} else {
		mnt_set_mountpoint(dest_mnt, dest_dentry, source_mnt);
		commit_tree(source_mnt);
	}
838 839 840 841 842 843 844 845 846

	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
847 848 849 850 851 852 853 854 855 856 857
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;
858
	mutex_lock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
859 860 861 862 863 864 865 866
	if (IS_DEADDIR(nd->dentry->d_inode))
		goto out_unlock;

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

	err = -ENOENT;
867
	if (IS_ROOT(nd->dentry) || !d_unhashed(nd->dentry))
R
Ram Pai 已提交
868
		err = attach_recursive_mnt(mnt, nd, NULL);
L
Linus Torvalds 已提交
869
out_unlock:
870
	mutex_unlock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
871 872 873 874 875
	if (!err)
		security_sb_post_addmount(mnt, nd);
	return err;
}

876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896
/*
 * 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
897 898 899
/*
 * do loopback mount.
 */
900
static int do_loopback(struct nameidata *nd, char *old_name, int recurse)
L
Linus Torvalds 已提交
901 902 903 904 905 906 907 908 909 910 911 912
{
	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
913
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
914
	err = -EINVAL;
R
Ram Pai 已提交
915 916 917
	if (IS_MNT_UNBINDABLE(old_nd.mnt))
 		goto out;

918 919
	if (!check_mnt(nd->mnt) || !check_mnt(old_nd.mnt))
		goto out;
L
Linus Torvalds 已提交
920

921 922
	err = -ENOMEM;
	if (recurse)
R
Ram Pai 已提交
923
		mnt = copy_tree(old_nd.mnt, old_nd.dentry, 0);
924
	else
R
Ram Pai 已提交
925
		mnt = clone_mnt(old_nd.mnt, old_nd.dentry, 0);
926 927 928 929 930 931

	if (!mnt)
		goto out;

	err = graft_tree(mnt, nd);
	if (err) {
R
Ram Pai 已提交
932
		LIST_HEAD(umount_list);
L
Linus Torvalds 已提交
933
		spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
934
		umount_tree(mnt, 0, &umount_list);
L
Linus Torvalds 已提交
935
		spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
936
		release_mounts(&umount_list);
937
	}
L
Linus Torvalds 已提交
938

939
out:
R
Ram Pai 已提交
940
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
941 942 943 944 945 946 947 948 949 950 951 952 953
	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
954
	struct super_block *sb = nd->mnt->mnt_sb;
L
Linus Torvalds 已提交
955 956 957 958 959 960 961 962 963 964 965 966 967

	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
968
		nd->mnt->mnt_flags = mnt_flags;
L
Linus Torvalds 已提交
969 970 971 972 973 974
	up_write(&sb->s_umount);
	if (!err)
		security_sb_post_remount(nd->mnt, flags, data);
	return err;
}

R
Ram Pai 已提交
975 976 977 978 979 980 981 982 983 984
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
985 986 987 988 989 990 991 992 993 994 995 996 997
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
998
	down_write(&namespace_sem);
R
Ram Pai 已提交
999
	while (d_mountpoint(nd->dentry) && follow_down(&nd->mnt, &nd->dentry))
L
Linus Torvalds 已提交
1000 1001 1002 1003 1004 1005
		;
	err = -EINVAL;
	if (!check_mnt(nd->mnt) || !check_mnt(old_nd.mnt))
		goto out;

	err = -ENOENT;
1006
	mutex_lock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
1007 1008 1009 1010
	if (IS_DEADDIR(nd->dentry->d_inode))
		goto out1;

	if (!IS_ROOT(nd->dentry) && d_unhashed(nd->dentry))
R
Ram Pai 已提交
1011
		goto out1;
L
Linus Torvalds 已提交
1012 1013 1014

	err = -EINVAL;
	if (old_nd.dentry != old_nd.mnt->mnt_root)
R
Ram Pai 已提交
1015
		goto out1;
L
Linus Torvalds 已提交
1016 1017

	if (old_nd.mnt == old_nd.mnt->mnt_parent)
R
Ram Pai 已提交
1018
		goto out1;
L
Linus Torvalds 已提交
1019 1020 1021

	if (S_ISDIR(nd->dentry->d_inode->i_mode) !=
	      S_ISDIR(old_nd.dentry->d_inode->i_mode))
R
Ram Pai 已提交
1022 1023 1024 1025 1026 1027
		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
1028 1029 1030 1031 1032 1033
	/*
	 * 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
1034
	err = -ELOOP;
R
Ram Pai 已提交
1035
	for (p = nd->mnt; p->mnt_parent != p; p = p->mnt_parent)
L
Linus Torvalds 已提交
1036
		if (p == old_nd.mnt)
R
Ram Pai 已提交
1037
			goto out1;
L
Linus Torvalds 已提交
1038

R
Ram Pai 已提交
1039 1040
	if ((err = attach_recursive_mnt(old_nd.mnt, nd, &parent_nd)))
		goto out1;
L
Linus Torvalds 已提交
1041

R
Ram Pai 已提交
1042
	spin_lock(&vfsmount_lock);
L
Linus Torvalds 已提交
1043 1044
	/* if the mount is moved, it should no longer be expire
	 * automatically */
1045
	list_del_init(&old_nd.mnt->mnt_expire);
L
Linus Torvalds 已提交
1046 1047
	spin_unlock(&vfsmount_lock);
out1:
1048
	mutex_unlock(&nd->dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
1049
out:
R
Ram Pai 已提交
1050
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1051 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
	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
1089
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
1090
	/* Something was mounted here while we slept */
R
Ram Pai 已提交
1091
	while (d_mountpoint(nd->dentry) && follow_down(&nd->mnt, &nd->dentry))
L
Linus Torvalds 已提交
1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
		;
	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;
1108 1109
	if ((err = graft_tree(newmnt, nd)))
		goto unlock;
L
Linus Torvalds 已提交
1110

1111
	if (fslist) {
L
Linus Torvalds 已提交
1112 1113
		/* add to the specified expiration list */
		spin_lock(&vfsmount_lock);
1114
		list_add_tail(&newmnt->mnt_expire, fslist);
L
Linus Torvalds 已提交
1115 1116
		spin_unlock(&vfsmount_lock);
	}
R
Ram Pai 已提交
1117
	up_write(&namespace_sem);
1118
	return 0;
L
Linus Torvalds 已提交
1119 1120

unlock:
R
Ram Pai 已提交
1121
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1122 1123 1124 1125 1126 1127
	mntput(newmnt);
	return err;
}

EXPORT_SYMBOL_GPL(do_add_mount);

R
Ram Pai 已提交
1128 1129
static void expire_mount(struct vfsmount *mnt, struct list_head *mounts,
				struct list_head *umounts)
1130 1131 1132
{
	spin_lock(&vfsmount_lock);

1133 1134 1135 1136 1137 1138 1139 1140 1141
	/*
	 * 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;
	}

1142 1143 1144 1145
	/*
	 * 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
1146
	if (!propagate_mount_busy(mnt, 2)) {
1147
		/* delete from the namespace */
A
Al Viro 已提交
1148
		touch_namespace(mnt->mnt_namespace);
1149
		list_del_init(&mnt->mnt_list);
1150
		mnt->mnt_namespace = NULL;
R
Ram Pai 已提交
1151
		umount_tree(mnt, 1, umounts);
1152 1153 1154 1155 1156 1157
		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
		 */
1158
		list_add_tail(&mnt->mnt_expire, mounts);
1159 1160 1161 1162
		spin_unlock(&vfsmount_lock);
	}
}

T
Trond Myklebust 已提交
1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
/*
 * 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)
{
	struct namespace *namespace;
	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 */
		namespace = mnt->mnt_namespace;
		if (!namespace || !namespace->root)
			continue;
		get_namespace(namespace);

		spin_unlock(&vfsmount_lock);
		down_write(&namespace_sem);
		expire_mount(mnt, mounts, &umounts);
		up_write(&namespace_sem);
		release_mounts(&umounts);
		mntput(mnt);
		put_namespace(namespace);
		spin_lock(&vfsmount_lock);
	}
}

L
Linus Torvalds 已提交
1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217
/*
 * 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())
	 */
1218
	list_for_each_entry_safe(mnt, next, mounts, mnt_expire) {
L
Linus Torvalds 已提交
1219 1220 1221 1222 1223
		if (!xchg(&mnt->mnt_expiry_mark, 1) ||
		    atomic_read(&mnt->mnt_count) != 1)
			continue;

		mntget(mnt);
1224
		list_move(&mnt->mnt_expire, &graveyard);
L
Linus Torvalds 已提交
1225 1226
	}

T
Trond Myklebust 已提交
1227
	expire_mount_list(&graveyard, mounts);
L
Linus Torvalds 已提交
1228

T
Trond Myklebust 已提交
1229 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
	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
1255
			continue;
T
Trond Myklebust 已提交
1256 1257 1258 1259 1260 1261 1262
		/*
		 * 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
1263

T
Trond Myklebust 已提交
1264 1265 1266 1267 1268
		if (!propagate_mount_busy(mnt, 1)) {
			mntget(mnt);
			list_move_tail(&mnt->mnt_expire, graveyard);
			found++;
		}
L
Linus Torvalds 已提交
1269
	}
T
Trond Myklebust 已提交
1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294
	/*
	 * 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
1295 1296 1297 1298

	spin_unlock(&vfsmount_lock);
}

T
Trond Myklebust 已提交
1299
EXPORT_SYMBOL_GPL(shrink_submounts);
L
Linus Torvalds 已提交
1300 1301 1302 1303 1304 1305 1306

/*
 * 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
1307 1308
static long exact_copy_from_user(void *to, const void __user * from,
				 unsigned long n)
L
Linus Torvalds 已提交
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328
{
	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
1329
int copy_mount_options(const void __user * data, unsigned long *where)
L
Linus Torvalds 已提交
1330 1331 1332 1333
{
	int i;
	unsigned long page;
	unsigned long size;
R
Ram Pai 已提交
1334

L
Linus Torvalds 已提交
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352
	*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
1353
		free_page(page);
L
Linus Torvalds 已提交
1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375
		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
1376
long do_mount(char *dev_name, char *dir_name, char *type_page,
L
Linus Torvalds 已提交
1377 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
		  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;
1404 1405 1406 1407 1408 1409 1410
	if (flags & MS_NOATIME)
		mnt_flags |= MNT_NOATIME;
	if (flags & MS_NODIRATIME)
		mnt_flags |= MNT_NODIRATIME;

	flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | MS_ACTIVE |
		   MS_NOATIME | MS_NODIRATIME);
L
Linus Torvalds 已提交
1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424

	/* ... 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)
1425
		retval = do_loopback(&nd, dev_name, flags & MS_REC);
R
Ram Pai 已提交
1426
	else if (flags & (MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE))
1427
		retval = do_change_type(&nd, flags);
L
Linus Torvalds 已提交
1428 1429 1430 1431 1432 1433 1434 1435 1436 1437
	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;
}

1438 1439 1440 1441 1442
/*
 * Allocate a new namespace structure and populate it with contents
 * copied from the namespace of the passed in task structure.
 */
struct namespace *dup_namespace(struct task_struct *tsk, struct fs_struct *fs)
L
Linus Torvalds 已提交
1443
{
1444
	struct namespace *namespace = tsk->nsproxy->namespace;
L
Linus Torvalds 已提交
1445 1446 1447 1448 1449 1450
	struct namespace *new_ns;
	struct vfsmount *rootmnt = NULL, *pwdmnt = NULL, *altrootmnt = NULL;
	struct vfsmount *p, *q;

	new_ns = kmalloc(sizeof(struct namespace), GFP_KERNEL);
	if (!new_ns)
1451
		return NULL;
L
Linus Torvalds 已提交
1452 1453 1454

	atomic_set(&new_ns->count, 1);
	INIT_LIST_HEAD(&new_ns->list);
A
Al Viro 已提交
1455 1456
	init_waitqueue_head(&new_ns->poll);
	new_ns->event = 0;
L
Linus Torvalds 已提交
1457

R
Ram Pai 已提交
1458
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
1459
	/* First pass: copy the tree topology */
R
Ram Pai 已提交
1460
	new_ns->root = copy_tree(namespace->root, namespace->root->mnt_root,
R
Ram Pai 已提交
1461
					CL_COPY_ALL | CL_EXPIRE);
L
Linus Torvalds 已提交
1462
	if (!new_ns->root) {
R
Ram Pai 已提交
1463
		up_write(&namespace_sem);
L
Linus Torvalds 已提交
1464
		kfree(new_ns);
1465
		return NULL;
L
Linus Torvalds 已提交
1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496
	}
	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.
	 */
	p = namespace->root;
	q = new_ns->root;
	while (p) {
		q->mnt_namespace = new_ns;
		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);
			}
		}
		p = next_mnt(p, namespace->root);
		q = next_mnt(q, new_ns->root);
	}
R
Ram Pai 已提交
1497
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1498 1499 1500 1501 1502 1503 1504 1505

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

1506 1507 1508 1509 1510
	return new_ns;
}

int copy_namespace(int flags, struct task_struct *tsk)
{
1511
	struct namespace *namespace = tsk->nsproxy->namespace;
1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533
	struct namespace *new_ns;
	int err = 0;

	if (!namespace)
		return 0;

	get_namespace(namespace);

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

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

	new_ns = dup_namespace(tsk, tsk->fs);
	if (!new_ns) {
		err = -ENOMEM;
		goto out;
	}

1534
	tsk->nsproxy->namespace = new_ns;
L
Linus Torvalds 已提交
1535 1536 1537

out:
	put_namespace(namespace);
1538
	return err;
L
Linus Torvalds 已提交
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550
}

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
1551
	retval = copy_mount_options(type, &type_page);
L
Linus Torvalds 已提交
1552 1553 1554 1555 1556 1557 1558 1559
	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
1560
	retval = copy_mount_options(dev_name, &dev_page);
L
Linus Torvalds 已提交
1561 1562 1563
	if (retval < 0)
		goto out2;

R
Ram Pai 已提交
1564
	retval = copy_mount_options(data, &data_page);
L
Linus Torvalds 已提交
1565 1566 1567 1568
	if (retval < 0)
		goto out3;

	lock_kernel();
R
Ram Pai 已提交
1569 1570
	retval = do_mount((char *)dev_page, dir_page, (char *)type_page,
			  flags, (void *)data_page);
L
Linus Torvalds 已提交
1571 1572 1573 1574 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
	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
1639 1640
			if (fs->root == old_nd->dentry
			    && fs->rootmnt == old_nd->mnt)
L
Linus Torvalds 已提交
1641
				set_fs_root(fs, new_nd->mnt, new_nd->dentry);
R
Ram Pai 已提交
1642 1643
			if (fs->pwd == old_nd->dentry
			    && fs->pwdmnt == old_nd->mnt)
L
Linus Torvalds 已提交
1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664
				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
1665 1666 1667 1668
 * 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
1669 1670 1671 1672 1673 1674 1675 1676
 * 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
1677 1678
asmlinkage long sys_pivot_root(const char __user * new_root,
			       const char __user * put_old)
L
Linus Torvalds 已提交
1679 1680 1681 1682 1683 1684 1685 1686 1687 1688
{
	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
1689 1690
	error = __user_walk(new_root, LOOKUP_FOLLOW | LOOKUP_DIRECTORY,
			    &new_nd);
L
Linus Torvalds 已提交
1691 1692 1693 1694 1695 1696
	if (error)
		goto out0;
	error = -EINVAL;
	if (!check_mnt(new_nd.mnt))
		goto out1;

R
Ram Pai 已提交
1697
	error = __user_walk(put_old, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &old_nd);
L
Linus Torvalds 已提交
1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710
	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
1711
	down_write(&namespace_sem);
1712
	mutex_lock(&old_nd.dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
1713
	error = -EINVAL;
R
Ram Pai 已提交
1714 1715 1716 1717
	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
1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732
	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 */
1733 1734
	if (user_nd.mnt->mnt_parent == user_nd.mnt)
		goto out2; /* not attached */
L
Linus Torvalds 已提交
1735 1736
	if (new_nd.mnt->mnt_root != new_nd.dentry)
		goto out2; /* not a mountpoint */
1737 1738
	if (new_nd.mnt->mnt_parent == new_nd.mnt)
		goto out2; /* not attached */
L
Linus Torvalds 已提交
1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756
	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 / */
1757
	touch_namespace(current->nsproxy->namespace);
L
Linus Torvalds 已提交
1758 1759 1760 1761 1762 1763 1764
	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:
1765
	mutex_unlock(&old_nd.dentry->d_inode->i_mutex);
R
Ram Pai 已提交
1766
	up_write(&namespace_sem);
L
Linus Torvalds 已提交
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791
	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;
	struct namespace *namespace;

	mnt = do_kern_mount("rootfs", 0, "rootfs", NULL);
	if (IS_ERR(mnt))
		panic("Can't create rootfs");
	namespace = kmalloc(sizeof(*namespace), GFP_KERNEL);
	if (!namespace)
		panic("Can't allocate initial namespace");
	atomic_set(&namespace->count, 1);
	INIT_LIST_HEAD(&namespace->list);
A
Al Viro 已提交
1792 1793
	init_waitqueue_head(&namespace->poll);
	namespace->event = 0;
L
Linus Torvalds 已提交
1794 1795 1796 1797
	list_add(&mnt->mnt_list, &namespace->list);
	namespace->root = mnt;
	mnt->mnt_namespace = namespace;

1798 1799
	init_task.nsproxy->namespace = namespace;
	get_namespace(namespace);
L
Linus Torvalds 已提交
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809

	set_fs_pwd(current->fs, namespace->root, namespace->root->mnt_root);
	set_fs_root(current->fs, namespace->root, namespace->root->mnt_root);
}

void __init mnt_init(unsigned long mempages)
{
	struct list_head *d;
	unsigned int nr_hash;
	int i;
1810
	int err;
L
Linus Torvalds 已提交
1811

R
Ram Pai 已提交
1812 1813
	init_rwsem(&namespace_sem);

L
Linus Torvalds 已提交
1814
	mnt_cache = kmem_cache_create("mnt_cache", sizeof(struct vfsmount),
R
Ram Pai 已提交
1815
			0, SLAB_HWCACHE_ALIGN | SLAB_PANIC, NULL, NULL);
L
Linus Torvalds 已提交
1816

R
Ram Pai 已提交
1817
	mount_hashtable = (struct list_head *)__get_free_page(GFP_ATOMIC);
L
Linus Torvalds 已提交
1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838

	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
1839
	hash_mask = nr_hash - 1;
L
Linus Torvalds 已提交
1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850

	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);
1851 1852 1853 1854 1855 1856 1857 1858
	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
1859 1860 1861 1862 1863 1864
	init_rootfs();
	init_mount_tree();
}

void __put_namespace(struct namespace *namespace)
{
1865
	struct vfsmount *root = namespace->root;
R
Ram Pai 已提交
1866
	LIST_HEAD(umount_list);
1867 1868
	namespace->root = NULL;
	spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
1869
	down_write(&namespace_sem);
L
Linus Torvalds 已提交
1870
	spin_lock(&vfsmount_lock);
R
Ram Pai 已提交
1871
	umount_tree(root, 0, &umount_list);
L
Linus Torvalds 已提交
1872
	spin_unlock(&vfsmount_lock);
R
Ram Pai 已提交
1873
	up_write(&namespace_sem);
R
Ram Pai 已提交
1874
	release_mounts(&umount_list);
L
Linus Torvalds 已提交
1875 1876
	kfree(namespace);
}