super.c 30.2 KB
Newer Older
M
Miklos Szeredi 已提交
1 2 3 4 5 6 7 8 9
/*
 *
 * Copyright (C) 2011 Novell Inc.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published by
 * the Free Software Foundation.
 */

10
#include <uapi/linux/magic.h>
M
Miklos Szeredi 已提交
11 12 13 14 15 16
#include <linux/fs.h>
#include <linux/namei.h>
#include <linux/xattr.h>
#include <linux/mount.h>
#include <linux/parser.h>
#include <linux/module.h>
A
Andy Whitcroft 已提交
17
#include <linux/statfs.h>
E
Erez Zadok 已提交
18
#include <linux/seq_file.h>
M
Miklos Szeredi 已提交
19
#include <linux/posix_acl_xattr.h>
M
Miklos Szeredi 已提交
20 21 22 23 24 25 26 27 28
#include "overlayfs.h"

MODULE_AUTHOR("Miklos Szeredi <miklos@szeredi.hu>");
MODULE_DESCRIPTION("Overlay filesystem");
MODULE_LICENSE("GPL");


struct ovl_dir_cache;

29 30
#define OVL_MAX_STACK 500

31 32 33 34
static bool ovl_redirect_dir_def = IS_ENABLED(CONFIG_OVERLAY_FS_REDIRECT_DIR);
module_param_named(redirect_dir, ovl_redirect_dir_def, bool, 0644);
MODULE_PARM_DESC(ovl_redirect_dir_def,
		 "Default to on or off for the redirect_dir feature");
M
Miklos Szeredi 已提交
35

36 37 38 39 40
static bool ovl_index_def = IS_ENABLED(CONFIG_OVERLAY_FS_INDEX);
module_param_named(index, ovl_index_def, bool, 0644);
MODULE_PARM_DESC(ovl_index_def,
		 "Default to on or off for the inodes index feature");

M
Miklos Szeredi 已提交
41 42 43 44 45
static void ovl_dentry_release(struct dentry *dentry)
{
	struct ovl_entry *oe = dentry->d_fsdata;

	if (oe) {
46 47 48 49
		unsigned int i;

		for (i = 0; i < oe->numlower; i++)
			dput(oe->lowerstack[i].dentry);
M
Miklos Szeredi 已提交
50 51 52 53
		kfree_rcu(oe, rcu);
	}
}

54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
static int ovl_check_append_only(struct inode *inode, int flag)
{
	/*
	 * This test was moot in vfs may_open() because overlay inode does
	 * not have the S_APPEND flag, so re-check on real upper inode
	 */
	if (IS_APPEND(inode)) {
		if  ((flag & O_ACCMODE) != O_RDONLY && !(flag & O_APPEND))
			return -EPERM;
		if (flag & O_TRUNC)
			return -EPERM;
	}

	return 0;
}

70 71
static struct dentry *ovl_d_real(struct dentry *dentry,
				 const struct inode *inode,
M
Miklos Szeredi 已提交
72
				 unsigned int open_flags, unsigned int flags)
M
Miklos Szeredi 已提交
73 74
{
	struct dentry *real;
75
	int err;
M
Miklos Szeredi 已提交
76

77 78 79
	if (flags & D_REAL_UPPER)
		return ovl_dentry_upper(dentry);

80
	if (!d_is_reg(dentry)) {
M
Miklos Szeredi 已提交
81 82 83 84 85
		if (!inode || inode == d_inode(dentry))
			return dentry;
		goto bug;
	}

86
	if (open_flags) {
87
		err = ovl_open_maybe_copy_up(dentry, open_flags);
88 89 90 91
		if (err)
			return ERR_PTR(err);
	}

M
Miklos Szeredi 已提交
92
	real = ovl_dentry_upper(dentry);
93 94 95 96 97 98
	if (real && (!inode || inode == d_inode(real))) {
		if (!inode) {
			err = ovl_check_append_only(d_inode(real), open_flags);
			if (err)
				return ERR_PTR(err);
		}
M
Miklos Szeredi 已提交
99
		return real;
100
	}
M
Miklos Szeredi 已提交
101 102 103 104 105

	real = ovl_dentry_lower(dentry);
	if (!real)
		goto bug;

M
Miklos Szeredi 已提交
106
	/* Handle recursion */
M
Miklos Szeredi 已提交
107
	real = d_real(real, inode, open_flags, 0);
M
Miklos Szeredi 已提交
108

M
Miklos Szeredi 已提交
109 110 111
	if (!inode || inode == d_inode(real))
		return real;
bug:
M
Miklos Szeredi 已提交
112
	WARN(1, "ovl_d_real(%pd4, %s:%lu): real dentry not found\n", dentry,
M
Miklos Szeredi 已提交
113 114 115 116
	     inode ? inode->i_sb->s_id : "NULL", inode ? inode->i_ino : 0);
	return dentry;
}

117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157
static int ovl_dentry_revalidate(struct dentry *dentry, unsigned int flags)
{
	struct ovl_entry *oe = dentry->d_fsdata;
	unsigned int i;
	int ret = 1;

	for (i = 0; i < oe->numlower; i++) {
		struct dentry *d = oe->lowerstack[i].dentry;

		if (d->d_flags & DCACHE_OP_REVALIDATE) {
			ret = d->d_op->d_revalidate(d, flags);
			if (ret < 0)
				return ret;
			if (!ret) {
				if (!(flags & LOOKUP_RCU))
					d_invalidate(d);
				return -ESTALE;
			}
		}
	}
	return 1;
}

static int ovl_dentry_weak_revalidate(struct dentry *dentry, unsigned int flags)
{
	struct ovl_entry *oe = dentry->d_fsdata;
	unsigned int i;
	int ret = 1;

	for (i = 0; i < oe->numlower; i++) {
		struct dentry *d = oe->lowerstack[i].dentry;

		if (d->d_flags & DCACHE_OP_WEAK_REVALIDATE) {
			ret = d->d_op->d_weak_revalidate(d, flags);
			if (ret <= 0)
				break;
		}
	}
	return ret;
}

M
Miklos Szeredi 已提交
158 159
static const struct dentry_operations ovl_dentry_operations = {
	.d_release = ovl_dentry_release,
M
Miklos Szeredi 已提交
160
	.d_real = ovl_d_real,
M
Miklos Szeredi 已提交
161 162
};

163 164
static const struct dentry_operations ovl_reval_dentry_operations = {
	.d_release = ovl_dentry_release,
M
Miklos Szeredi 已提交
165
	.d_real = ovl_d_real,
166 167 168 169
	.d_revalidate = ovl_dentry_revalidate,
	.d_weak_revalidate = ovl_dentry_weak_revalidate,
};

170 171 172 173 174 175
static struct kmem_cache *ovl_inode_cachep;

static struct inode *ovl_alloc_inode(struct super_block *sb)
{
	struct ovl_inode *oi = kmem_cache_alloc(ovl_inode_cachep, GFP_KERNEL);

176 177 178
	if (!oi)
		return NULL;

179
	oi->cache = NULL;
M
Miklos Szeredi 已提交
180
	oi->redirect = NULL;
181
	oi->version = 0;
M
Miklos Szeredi 已提交
182
	oi->flags = 0;
183
	oi->__upperdentry = NULL;
184
	oi->lower = NULL;
185
	mutex_init(&oi->lock);
186

187 188 189 190 191 192 193 194 195 196 197 198
	return &oi->vfs_inode;
}

static void ovl_i_callback(struct rcu_head *head)
{
	struct inode *inode = container_of(head, struct inode, i_rcu);

	kmem_cache_free(ovl_inode_cachep, OVL_I(inode));
}

static void ovl_destroy_inode(struct inode *inode)
{
199 200 201
	struct ovl_inode *oi = OVL_I(inode);

	dput(oi->__upperdentry);
M
Miklos Szeredi 已提交
202
	kfree(oi->redirect);
203
	ovl_dir_cache_free(inode);
204
	mutex_destroy(&oi->lock);
205

206 207 208
	call_rcu(&inode->i_rcu, ovl_i_callback);
}

209
static void ovl_free_fs(struct ovl_fs *ufs)
M
Miklos Szeredi 已提交
210
{
211
	unsigned i;
M
Miklos Szeredi 已提交
212

213
	dput(ufs->indexdir);
M
Miklos Szeredi 已提交
214
	dput(ufs->workdir);
215 216
	if (ufs->workdir_locked)
		ovl_inuse_unlock(ufs->workbasedir);
217
	dput(ufs->workbasedir);
218
	if (ufs->upperdir_locked)
219
		ovl_inuse_unlock(ufs->upper_mnt->mnt_root);
M
Miklos Szeredi 已提交
220
	mntput(ufs->upper_mnt);
221
	for (i = 0; i < ufs->numlower; i++) {
222
		mntput(ufs->lower_layers[i].mnt);
223 224
		free_anon_bdev(ufs->lower_layers[i].pseudo_dev);
	}
225
	kfree(ufs->lower_layers);
M
Miklos Szeredi 已提交
226

E
Erez Zadok 已提交
227 228 229
	kfree(ufs->config.lowerdir);
	kfree(ufs->config.upperdir);
	kfree(ufs->config.workdir);
230 231
	if (ufs->creator_cred)
		put_cred(ufs->creator_cred);
M
Miklos Szeredi 已提交
232 233 234
	kfree(ufs);
}

235 236 237 238 239 240 241
static void ovl_put_super(struct super_block *sb)
{
	struct ovl_fs *ofs = sb->s_fs_info;

	ovl_free_fs(ofs);
}

242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
static int ovl_sync_fs(struct super_block *sb, int wait)
{
	struct ovl_fs *ufs = sb->s_fs_info;
	struct super_block *upper_sb;
	int ret;

	if (!ufs->upper_mnt)
		return 0;
	upper_sb = ufs->upper_mnt->mnt_sb;
	if (!upper_sb->s_op->sync_fs)
		return 0;

	/* real inodes have already been synced by sync_filesystem(ovl_sb) */
	down_read(&upper_sb->s_umount);
	ret = upper_sb->s_op->sync_fs(upper_sb, wait);
	up_read(&upper_sb->s_umount);
	return ret;
}

A
Andy Whitcroft 已提交
261 262 263 264 265 266
/**
 * ovl_statfs
 * @sb: The overlayfs super block
 * @buf: The struct kstatfs to fill in with stats
 *
 * Get the filesystem statistics.  As writes always target the upper layer
267
 * filesystem pass the statfs to the upper filesystem (if it exists)
A
Andy Whitcroft 已提交
268 269 270 271 272 273 274 275
 */
static int ovl_statfs(struct dentry *dentry, struct kstatfs *buf)
{
	struct ovl_fs *ofs = dentry->d_sb->s_fs_info;
	struct dentry *root_dentry = dentry->d_sb->s_root;
	struct path path;
	int err;

276
	ovl_path_real(root_dentry, &path);
A
Andy Whitcroft 已提交
277 278 279

	err = vfs_statfs(&path, buf);
	if (!err) {
M
Miklos Szeredi 已提交
280
		buf->f_namelen = ofs->namelen;
A
Andy Whitcroft 已提交
281 282 283 284 285 286
		buf->f_type = OVERLAYFS_SUPER_MAGIC;
	}

	return err;
}

287 288 289 290 291 292
/* Will this overlay be forced to mount/remount ro? */
static bool ovl_force_readonly(struct ovl_fs *ufs)
{
	return (!ufs->upper_mnt || !ufs->workdir);
}

E
Erez Zadok 已提交
293 294 295 296 297 298 299 300 301 302 303
/**
 * ovl_show_options
 *
 * Prints the mount options for a given superblock.
 * Returns zero; does not fail.
 */
static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
{
	struct super_block *sb = dentry->d_sb;
	struct ovl_fs *ufs = sb->s_fs_info;

304
	seq_show_option(m, "lowerdir", ufs->config.lowerdir);
M
Miklos Szeredi 已提交
305
	if (ufs->config.upperdir) {
306 307
		seq_show_option(m, "upperdir", ufs->config.upperdir);
		seq_show_option(m, "workdir", ufs->config.workdir);
M
Miklos Szeredi 已提交
308
	}
M
Miklos Szeredi 已提交
309 310
	if (ufs->config.default_permissions)
		seq_puts(m, ",default_permissions");
311 312 313
	if (ufs->config.redirect_dir != ovl_redirect_dir_def)
		seq_printf(m, ",redirect_dir=%s",
			   ufs->config.redirect_dir ? "on" : "off");
314 315 316
	if (ufs->config.index != ovl_index_def)
		seq_printf(m, ",index=%s",
			   ufs->config.index ? "on" : "off");
E
Erez Zadok 已提交
317 318 319
	return 0;
}

320 321 322 323
static int ovl_remount(struct super_block *sb, int *flags, char *data)
{
	struct ovl_fs *ufs = sb->s_fs_info;

324
	if (!(*flags & MS_RDONLY) && ovl_force_readonly(ufs))
325 326 327 328 329
		return -EROFS;

	return 0;
}

M
Miklos Szeredi 已提交
330
static const struct super_operations ovl_super_operations = {
331 332 333
	.alloc_inode	= ovl_alloc_inode,
	.destroy_inode	= ovl_destroy_inode,
	.drop_inode	= generic_delete_inode,
M
Miklos Szeredi 已提交
334
	.put_super	= ovl_put_super,
335
	.sync_fs	= ovl_sync_fs,
A
Andy Whitcroft 已提交
336
	.statfs		= ovl_statfs,
E
Erez Zadok 已提交
337
	.show_options	= ovl_show_options,
338
	.remount_fs	= ovl_remount,
M
Miklos Szeredi 已提交
339 340 341 342 343 344
};

enum {
	OPT_LOWERDIR,
	OPT_UPPERDIR,
	OPT_WORKDIR,
M
Miklos Szeredi 已提交
345
	OPT_DEFAULT_PERMISSIONS,
M
Miklos Szeredi 已提交
346 347
	OPT_REDIRECT_DIR_ON,
	OPT_REDIRECT_DIR_OFF,
348 349
	OPT_INDEX_ON,
	OPT_INDEX_OFF,
M
Miklos Szeredi 已提交
350 351 352 353 354 355 356
	OPT_ERR,
};

static const match_table_t ovl_tokens = {
	{OPT_LOWERDIR,			"lowerdir=%s"},
	{OPT_UPPERDIR,			"upperdir=%s"},
	{OPT_WORKDIR,			"workdir=%s"},
M
Miklos Szeredi 已提交
357
	{OPT_DEFAULT_PERMISSIONS,	"default_permissions"},
M
Miklos Szeredi 已提交
358 359
	{OPT_REDIRECT_DIR_ON,		"redirect_dir=on"},
	{OPT_REDIRECT_DIR_OFF,		"redirect_dir=off"},
360 361
	{OPT_INDEX_ON,			"index=on"},
	{OPT_INDEX_OFF,			"index=off"},
M
Miklos Szeredi 已提交
362 363 364
	{OPT_ERR,			NULL}
};

M
Miklos Szeredi 已提交
365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
static char *ovl_next_opt(char **s)
{
	char *sbegin = *s;
	char *p;

	if (sbegin == NULL)
		return NULL;

	for (p = sbegin; *p; p++) {
		if (*p == '\\') {
			p++;
			if (!*p)
				break;
		} else if (*p == ',') {
			*p = '\0';
			*s = p + 1;
			return sbegin;
		}
	}
	*s = NULL;
	return sbegin;
}

M
Miklos Szeredi 已提交
388 389 390 391
static int ovl_parse_opt(char *opt, struct ovl_config *config)
{
	char *p;

M
Miklos Szeredi 已提交
392
	while ((p = ovl_next_opt(&opt)) != NULL) {
M
Miklos Szeredi 已提交
393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
		int token;
		substring_t args[MAX_OPT_ARGS];

		if (!*p)
			continue;

		token = match_token(p, ovl_tokens, args);
		switch (token) {
		case OPT_UPPERDIR:
			kfree(config->upperdir);
			config->upperdir = match_strdup(&args[0]);
			if (!config->upperdir)
				return -ENOMEM;
			break;

		case OPT_LOWERDIR:
			kfree(config->lowerdir);
			config->lowerdir = match_strdup(&args[0]);
			if (!config->lowerdir)
				return -ENOMEM;
			break;

		case OPT_WORKDIR:
			kfree(config->workdir);
			config->workdir = match_strdup(&args[0]);
			if (!config->workdir)
				return -ENOMEM;
			break;

M
Miklos Szeredi 已提交
422 423 424 425
		case OPT_DEFAULT_PERMISSIONS:
			config->default_permissions = true;
			break;

M
Miklos Szeredi 已提交
426 427 428 429 430 431 432 433
		case OPT_REDIRECT_DIR_ON:
			config->redirect_dir = true;
			break;

		case OPT_REDIRECT_DIR_OFF:
			config->redirect_dir = false;
			break;

434 435 436 437 438 439 440 441
		case OPT_INDEX_ON:
			config->index = true;
			break;

		case OPT_INDEX_OFF:
			config->index = false;
			break;

M
Miklos Szeredi 已提交
442
		default:
443
			pr_err("overlayfs: unrecognized mount option \"%s\" or missing value\n", p);
M
Miklos Szeredi 已提交
444 445 446
			return -EINVAL;
		}
	}
H
hujianyang 已提交
447 448 449 450 451 452 453 454 455

	/* Workdir is useless in non-upper mount */
	if (!config->upperdir && config->workdir) {
		pr_info("overlayfs: option \"workdir=%s\" is useless in a non-upper mount, ignore\n",
			config->workdir);
		kfree(config->workdir);
		config->workdir = NULL;
	}

M
Miklos Szeredi 已提交
456 457 458 459
	return 0;
}

#define OVL_WORKDIR_NAME "work"
460
#define OVL_INDEXDIR_NAME "index"
M
Miklos Szeredi 已提交
461

462
static struct dentry *ovl_workdir_create(struct ovl_fs *ufs,
463
					 const char *name, bool persist)
M
Miklos Szeredi 已提交
464
{
465
	struct inode *dir =  ufs->workbasedir->d_inode;
466
	struct vfsmount *mnt = ufs->upper_mnt;
M
Miklos Szeredi 已提交
467 468 469
	struct dentry *work;
	int err;
	bool retried = false;
470
	bool locked = false;
M
Miklos Szeredi 已提交
471 472 473

	err = mnt_want_write(mnt);
	if (err)
474
		goto out_err;
M
Miklos Szeredi 已提交
475

A
Al Viro 已提交
476
	inode_lock_nested(dir, I_MUTEX_PARENT);
477 478
	locked = true;

M
Miklos Szeredi 已提交
479
retry:
480
	work = lookup_one_len(name, ufs->workbasedir, strlen(name));
M
Miklos Szeredi 已提交
481 482

	if (!IS_ERR(work)) {
483 484
		struct iattr attr = {
			.ia_valid = ATTR_MODE,
A
Al Viro 已提交
485
			.ia_mode = S_IFDIR | 0,
486
		};
M
Miklos Szeredi 已提交
487 488 489 490 491 492

		if (work->d_inode) {
			err = -EEXIST;
			if (retried)
				goto out_dput;

493 494 495
			if (persist)
				goto out_unlock;

M
Miklos Szeredi 已提交
496
			retried = true;
M
Miklos Szeredi 已提交
497
			ovl_workdir_cleanup(dir, mnt, work, 0);
M
Miklos Szeredi 已提交
498 499 500 501
			dput(work);
			goto retry;
		}

A
Al Viro 已提交
502 503 504
		err = ovl_create_real(dir, work,
				      &(struct cattr){.mode = S_IFDIR | 0},
				      NULL, true);
M
Miklos Szeredi 已提交
505 506
		if (err)
			goto out_dput;
507

508 509 510 511 512 513 514 515 516 517 518 519 520
		/*
		 * Try to remove POSIX ACL xattrs from workdir.  We are good if:
		 *
		 * a) success (there was a POSIX ACL xattr and was removed)
		 * b) -ENODATA (there was no POSIX ACL xattr)
		 * c) -EOPNOTSUPP (POSIX ACL xattrs are not supported)
		 *
		 * There are various other error values that could effectively
		 * mean that the xattr doesn't exist (e.g. -ERANGE is returned
		 * if the xattr name is too long), but the set of filesystems
		 * allowed as upper are limited to "normal" ones, where checking
		 * for the above two errors is sufficient.
		 */
521
		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_DEFAULT);
M
Miklos Szeredi 已提交
522
		if (err && err != -ENODATA && err != -EOPNOTSUPP)
523 524 525
			goto out_dput;

		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_ACCESS);
M
Miklos Szeredi 已提交
526
		if (err && err != -ENODATA && err != -EOPNOTSUPP)
527 528 529 530 531 532 533 534
			goto out_dput;

		/* Clear any inherited mode bits */
		inode_lock(work->d_inode);
		err = notify_change(work, &attr, NULL);
		inode_unlock(work->d_inode);
		if (err)
			goto out_dput;
535 536 537
	} else {
		err = PTR_ERR(work);
		goto out_err;
M
Miklos Szeredi 已提交
538 539 540
	}
out_unlock:
	mnt_drop_write(mnt);
541 542
	if (locked)
		inode_unlock(dir);
M
Miklos Szeredi 已提交
543 544 545 546 547

	return work;

out_dput:
	dput(work);
548 549 550 551
out_err:
	pr_warn("overlayfs: failed to create directory %s/%s (errno: %i); mounting read-only\n",
		ufs->config.workdir, name, -err);
	work = NULL;
M
Miklos Szeredi 已提交
552 553 554
	goto out_unlock;
}

M
Miklos Szeredi 已提交
555 556 557 558 559 560 561 562 563 564 565 566 567
static void ovl_unescape(char *s)
{
	char *d = s;

	for (;; s++, d++) {
		if (*s == '\\')
			s++;
		*d = *s;
		if (!*s)
			break;
	}
}

M
Miklos Szeredi 已提交
568 569
static int ovl_mount_dir_noesc(const char *name, struct path *path)
{
570
	int err = -EINVAL;
M
Miklos Szeredi 已提交
571

572 573 574 575
	if (!*name) {
		pr_err("overlayfs: empty lowerdir\n");
		goto out;
	}
M
Miklos Szeredi 已提交
576 577 578 579 580 581
	err = kern_path(name, LOOKUP_FOLLOW, path);
	if (err) {
		pr_err("overlayfs: failed to resolve '%s': %i\n", name, err);
		goto out;
	}
	err = -EINVAL;
582
	if (ovl_dentry_weird(path->dentry)) {
M
Miklos Szeredi 已提交
583 584 585
		pr_err("overlayfs: filesystem on '%s' not supported\n", name);
		goto out_put;
	}
M
Miklos Szeredi 已提交
586
	if (!d_is_dir(path->dentry)) {
M
Miklos Szeredi 已提交
587 588 589 590 591 592
		pr_err("overlayfs: '%s' not a directory\n", name);
		goto out_put;
	}
	return 0;

out_put:
593
	path_put_init(path);
M
Miklos Szeredi 已提交
594 595 596 597 598 599 600 601 602 603 604 605
out:
	return err;
}

static int ovl_mount_dir(const char *name, struct path *path)
{
	int err = -ENOMEM;
	char *tmp = kstrdup(name, GFP_KERNEL);

	if (tmp) {
		ovl_unescape(tmp);
		err = ovl_mount_dir_noesc(tmp, path);
606 607 608 609 610

		if (!err)
			if (ovl_dentry_remote(path->dentry)) {
				pr_err("overlayfs: filesystem on '%s' not supported as upperdir\n",
				       tmp);
611
				path_put_init(path);
612 613
				err = -EINVAL;
			}
M
Miklos Szeredi 已提交
614 615 616 617 618
		kfree(tmp);
	}
	return err;
}

M
Miklos Szeredi 已提交
619 620
static int ovl_check_namelen(struct path *path, struct ovl_fs *ofs,
			     const char *name)
M
Miklos Szeredi 已提交
621 622
{
	struct kstatfs statfs;
M
Miklos Szeredi 已提交
623 624 625 626 627 628 629 630 631 632 633 634 635 636
	int err = vfs_statfs(path, &statfs);

	if (err)
		pr_err("overlayfs: statfs failed on '%s'\n", name);
	else
		ofs->namelen = max(ofs->namelen, statfs.f_namelen);

	return err;
}

static int ovl_lower_dir(const char *name, struct path *path,
			 struct ovl_fs *ofs, int *stack_depth, bool *remote)
{
	int err;
M
Miklos Szeredi 已提交
637

638
	err = ovl_mount_dir_noesc(name, path);
M
Miklos Szeredi 已提交
639 640 641
	if (err)
		goto out;

M
Miklos Szeredi 已提交
642 643
	err = ovl_check_namelen(path, ofs, name);
	if (err)
M
Miklos Szeredi 已提交
644
		goto out_put;
M
Miklos Szeredi 已提交
645

M
Miklos Szeredi 已提交
646 647
	*stack_depth = max(*stack_depth, path->mnt->mnt_sb->s_stack_depth);

648 649 650
	if (ovl_dentry_remote(path->dentry))
		*remote = true;

651 652 653 654 655 656 657 658 659
	/*
	 * The inodes index feature needs to encode and decode file
	 * handles, so it requires that all layers support them.
	 */
	if (ofs->config.index && !ovl_can_decode_fh(path->dentry->d_sb)) {
		ofs->config.index = false;
		pr_warn("overlayfs: fs on '%s' does not support file handles, falling back to index=off.\n", name);
	}

M
Miklos Szeredi 已提交
660 661 662
	return 0;

out_put:
663
	path_put_init(path);
M
Miklos Szeredi 已提交
664 665 666 667
out:
	return err;
}

M
Miklos Szeredi 已提交
668 669 670 671 672 673 674 675 676 677 678 679
/* Workdir should not be subdir of upperdir and vice versa */
static bool ovl_workdir_ok(struct dentry *workdir, struct dentry *upperdir)
{
	bool ok = false;

	if (workdir != upperdir) {
		ok = (lock_rename(workdir, upperdir) == NULL);
		unlock_rename(workdir, upperdir);
	}
	return ok;
}

680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699
static unsigned int ovl_split_lowerdirs(char *str)
{
	unsigned int ctr = 1;
	char *s, *d;

	for (s = d = str;; s++, d++) {
		if (*s == '\\') {
			s++;
		} else if (*s == ':') {
			*d = '\0';
			ctr++;
			continue;
		}
		*d = *s;
		if (!*s)
			break;
	}
	return ctr;
}

700 701 702 703 704
static int __maybe_unused
ovl_posix_acl_xattr_get(const struct xattr_handler *handler,
			struct dentry *dentry, struct inode *inode,
			const char *name, void *buffer, size_t size)
{
705
	return ovl_xattr_get(dentry, inode, handler->name, buffer, size);
706 707
}

708 709 710 711 712
static int __maybe_unused
ovl_posix_acl_xattr_set(const struct xattr_handler *handler,
			struct dentry *dentry, struct inode *inode,
			const char *name, const void *value,
			size_t size, int flags)
M
Miklos Szeredi 已提交
713 714
{
	struct dentry *workdir = ovl_workdir(dentry);
715
	struct inode *realinode = ovl_inode_real(inode);
M
Miklos Szeredi 已提交
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739
	struct posix_acl *acl = NULL;
	int err;

	/* Check that everything is OK before copy-up */
	if (value) {
		acl = posix_acl_from_xattr(&init_user_ns, value, size);
		if (IS_ERR(acl))
			return PTR_ERR(acl);
	}
	err = -EOPNOTSUPP;
	if (!IS_POSIXACL(d_inode(workdir)))
		goto out_acl_release;
	if (!realinode->i_op->set_acl)
		goto out_acl_release;
	if (handler->flags == ACL_TYPE_DEFAULT && !S_ISDIR(inode->i_mode)) {
		err = acl ? -EACCES : 0;
		goto out_acl_release;
	}
	err = -EPERM;
	if (!inode_owner_or_capable(inode))
		goto out_acl_release;

	posix_acl_release(acl);

740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
	/*
	 * Check if sgid bit needs to be cleared (actual setacl operation will
	 * be done with mounter's capabilities and so that won't do it for us).
	 */
	if (unlikely(inode->i_mode & S_ISGID) &&
	    handler->flags == ACL_TYPE_ACCESS &&
	    !in_group_p(inode->i_gid) &&
	    !capable_wrt_inode_uidgid(inode, CAP_FSETID)) {
		struct iattr iattr = { .ia_valid = ATTR_KILL_SGID };

		err = ovl_setattr(dentry, &iattr);
		if (err)
			return err;
	}

755
	err = ovl_xattr_set(dentry, inode, handler->name, value, size, flags);
756
	if (!err)
757
		ovl_copyattr(ovl_inode_real(inode), inode);
758 759

	return err;
M
Miklos Szeredi 已提交
760 761 762 763 764 765

out_acl_release:
	posix_acl_release(acl);
	return err;
}

766 767 768 769
static int ovl_own_xattr_get(const struct xattr_handler *handler,
			     struct dentry *dentry, struct inode *inode,
			     const char *name, void *buffer, size_t size)
{
A
Amir Goldstein 已提交
770
	return -EOPNOTSUPP;
771 772
}

M
Miklos Szeredi 已提交
773 774 775 776 777
static int ovl_own_xattr_set(const struct xattr_handler *handler,
			     struct dentry *dentry, struct inode *inode,
			     const char *name, const void *value,
			     size_t size, int flags)
{
A
Amir Goldstein 已提交
778
	return -EOPNOTSUPP;
M
Miklos Szeredi 已提交
779 780
}

781 782 783 784
static int ovl_other_xattr_get(const struct xattr_handler *handler,
			       struct dentry *dentry, struct inode *inode,
			       const char *name, void *buffer, size_t size)
{
785
	return ovl_xattr_get(dentry, inode, name, buffer, size);
786 787
}

788 789 790 791 792
static int ovl_other_xattr_set(const struct xattr_handler *handler,
			       struct dentry *dentry, struct inode *inode,
			       const char *name, const void *value,
			       size_t size, int flags)
{
793
	return ovl_xattr_set(dentry, inode, name, value, size, flags);
794 795
}

796 797
static const struct xattr_handler __maybe_unused
ovl_posix_acl_access_xattr_handler = {
M
Miklos Szeredi 已提交
798 799
	.name = XATTR_NAME_POSIX_ACL_ACCESS,
	.flags = ACL_TYPE_ACCESS,
800
	.get = ovl_posix_acl_xattr_get,
M
Miklos Szeredi 已提交
801 802 803
	.set = ovl_posix_acl_xattr_set,
};

804 805
static const struct xattr_handler __maybe_unused
ovl_posix_acl_default_xattr_handler = {
M
Miklos Szeredi 已提交
806 807
	.name = XATTR_NAME_POSIX_ACL_DEFAULT,
	.flags = ACL_TYPE_DEFAULT,
808
	.get = ovl_posix_acl_xattr_get,
M
Miklos Szeredi 已提交
809 810 811 812 813
	.set = ovl_posix_acl_xattr_set,
};

static const struct xattr_handler ovl_own_xattr_handler = {
	.prefix	= OVL_XATTR_PREFIX,
814
	.get = ovl_own_xattr_get,
M
Miklos Szeredi 已提交
815 816 817 818 819
	.set = ovl_own_xattr_set,
};

static const struct xattr_handler ovl_other_xattr_handler = {
	.prefix	= "", /* catch all */
820
	.get = ovl_other_xattr_get,
M
Miklos Szeredi 已提交
821 822 823 824
	.set = ovl_other_xattr_set,
};

static const struct xattr_handler *ovl_xattr_handlers[] = {
825
#ifdef CONFIG_FS_POSIX_ACL
M
Miklos Szeredi 已提交
826 827
	&ovl_posix_acl_access_xattr_handler,
	&ovl_posix_acl_default_xattr_handler,
828
#endif
M
Miklos Szeredi 已提交
829 830 831 832 833
	&ovl_own_xattr_handler,
	&ovl_other_xattr_handler,
	NULL
};

M
Miklos Szeredi 已提交
834
static int ovl_get_upper(struct ovl_fs *ufs, struct path *upperpath)
835
{
M
Miklos Szeredi 已提交
836
	struct vfsmount *upper_mnt;
837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862
	int err;

	err = ovl_mount_dir(ufs->config.upperdir, upperpath);
	if (err)
		goto out;

	/* Upper fs should not be r/o */
	if (sb_rdonly(upperpath->mnt->mnt_sb)) {
		pr_err("overlayfs: upper fs is r/o, try multi-lower layers mount\n");
		err = -EINVAL;
		goto out;
	}

	err = ovl_check_namelen(upperpath, ufs, ufs->config.upperdir);
	if (err)
		goto out;

	err = -EBUSY;
	if (ovl_inuse_trylock(upperpath->dentry)) {
		ufs->upperdir_locked = true;
	} else if (ufs->config.index) {
		pr_err("overlayfs: upperdir is in-use by another mount, mount with '-o index=off' to override exclusive upperdir protection.\n");
		goto out;
	} else {
		pr_warn("overlayfs: upperdir is in-use by another mount, accessing files from both mounts will result in undefined behavior.\n");
	}
M
Miklos Szeredi 已提交
863 864 865 866 867 868 869 870 871 872 873

	upper_mnt = clone_private_mount(upperpath);
	err = PTR_ERR(upper_mnt);
	if (IS_ERR(upper_mnt)) {
		pr_err("overlayfs: failed to clone upperpath\n");
		goto out;
	}

	/* Don't inherit atime flags */
	upper_mnt->mnt_flags &= ~(MNT_NOATIME | MNT_NODIRATIME | MNT_RELATIME);
	ufs->upper_mnt = upper_mnt;
874 875 876 877 878
	err = 0;
out:
	return err;
}

879
static int ovl_get_workdir(struct ovl_fs *ufs, struct path *workpath)
880 881 882 883
{
	struct dentry *temp;
	int err;

884
	ufs->workdir = ovl_workdir_create(ufs, OVL_WORKDIR_NAME, false);
885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933
	if (!ufs->workdir)
		return 0;

	/*
	 * Upper should support d_type, else whiteouts are visible.  Given
	 * workdir and upper are on same fs, we can do iterate_dir() on
	 * workdir. This check requires successful creation of workdir in
	 * previous step.
	 */
	err = ovl_check_d_type_supported(workpath);
	if (err < 0)
		return err;

	/*
	 * We allowed this configuration and don't want to break users over
	 * kernel upgrade. So warn instead of erroring out.
	 */
	if (!err)
		pr_warn("overlayfs: upper fs needs to support d_type.\n");

	/* Check if upper/work fs supports O_TMPFILE */
	temp = ovl_do_tmpfile(ufs->workdir, S_IFREG | 0);
	ufs->tmpfile = !IS_ERR(temp);
	if (ufs->tmpfile)
		dput(temp);
	else
		pr_warn("overlayfs: upper fs does not support tmpfile.\n");

	/*
	 * Check if upper/work fs supports trusted.overlay.* xattr
	 */
	err = ovl_do_setxattr(ufs->workdir, OVL_XATTR_OPAQUE, "0", 1, 0);
	if (err) {
		ufs->noxattr = true;
		pr_warn("overlayfs: upper fs does not support xattr.\n");
	} else {
		vfs_removexattr(ufs->workdir, OVL_XATTR_OPAQUE);
	}

	/* Check if upper/work fs supports file handles */
	if (ufs->config.index &&
	    !ovl_can_decode_fh(ufs->workdir->d_sb)) {
		ufs->config.index = false;
		pr_warn("overlayfs: upper fs does not support file handles, falling back to index=off.\n");
	}

	return 0;
}

934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968
static int ovl_get_workpath(struct ovl_fs *ufs, struct path *upperpath,
			    struct path *workpath)
{
	int err;

	err = ovl_mount_dir(ufs->config.workdir, workpath);
	if (err)
		goto out;

	err = -EINVAL;
	if (upperpath->mnt != workpath->mnt) {
		pr_err("overlayfs: workdir and upperdir must reside under the same mount\n");
		goto out;
	}
	if (!ovl_workdir_ok(workpath->dentry, upperpath->dentry)) {
		pr_err("overlayfs: workdir and upperdir must be separate subtrees\n");
		goto out;
	}

	err = -EBUSY;
	if (ovl_inuse_trylock(workpath->dentry)) {
		ufs->workdir_locked = true;
	} else if (ufs->config.index) {
		pr_err("overlayfs: workdir is in-use by another mount, mount with '-o index=off' to override exclusive workdir protection.\n");
		goto out;
	} else {
		pr_warn("overlayfs: workdir is in-use by another mount, accessing files from both mounts will result in undefined behavior.\n");
	}

	ufs->workbasedir = dget(workpath->dentry);
	err = 0;
out:
	return err;
}

969
static int ovl_get_indexdir(struct ovl_fs *ufs, struct ovl_entry *oe,
970
			    struct path *upperpath)
971 972 973 974 975 976 977 978 979 980 981 982 983
{
	int err;

	/* Verify lower root is upper root origin */
	err = ovl_verify_origin(upperpath->dentry,
				oe->lowerstack[0].layer->mnt,
				oe->lowerstack[0].dentry,
				false, true);
	if (err) {
		pr_err("overlayfs: failed to verify upper root origin\n");
		goto out;
	}

984
	ufs->indexdir = ovl_workdir_create(ufs, OVL_INDEXDIR_NAME, true);
985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005
	if (ufs->indexdir) {
		/* Verify upper root is index dir origin */
		err = ovl_verify_origin(ufs->indexdir, ufs->upper_mnt,
					upperpath->dentry, true, true);
		if (err)
			pr_err("overlayfs: failed to verify index dir origin\n");

		/* Cleanup bad/stale/orphan index entries */
		if (!err)
			err = ovl_indexdir_cleanup(ufs->indexdir,
						   ufs->upper_mnt,
						   oe->lowerstack,
						   oe->numlower);
	}
	if (err || !ufs->indexdir)
		pr_warn("overlayfs: try deleting index dir or mounting with '-o index=off' to disable inodes index.\n");

out:
	return err;
}

1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
static int ovl_get_lower_layers(struct ovl_fs *ufs, struct path *stack,
				unsigned int numlower)
{
	int err;
	unsigned int i;

	err = -ENOMEM;
	ufs->lower_layers = kcalloc(numlower, sizeof(struct ovl_layer),
				    GFP_KERNEL);
	if (ufs->lower_layers == NULL)
		goto out;
	for (i = 0; i < numlower; i++) {
		struct vfsmount *mnt;
		dev_t dev;

		err = get_anon_bdev(&dev);
		if (err) {
			pr_err("overlayfs: failed to get anonymous bdev for lowerpath\n");
			goto out;
		}

		mnt = clone_private_mount(&stack[i]);
		err = PTR_ERR(mnt);
		if (IS_ERR(mnt)) {
			pr_err("overlayfs: failed to clone lowerpath\n");
			free_anon_bdev(dev);
			goto out;
		}
		/*
		 * Make lower layers R/O.  That way fchmod/fchown on lower file
		 * will fail instead of modifying lower fs.
		 */
		mnt->mnt_flags |= MNT_READONLY | MNT_NOATIME;

		ufs->lower_layers[ufs->numlower].mnt = mnt;
		ufs->lower_layers[ufs->numlower].pseudo_dev = dev;
		ufs->numlower++;

		/* Check if all lower layers are on same sb */
		if (i == 0)
			ufs->same_sb = mnt->mnt_sb;
		else if (ufs->same_sb != mnt->mnt_sb)
			ufs->same_sb = NULL;
	}
	err = 0;
out:
	return err;
}

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 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123
static int ovl_get_lowerstack(struct super_block *sb, struct ovl_fs *ufs,
			      struct path **stackp, unsigned int *stacklenp)
{
	int err;
	char *lowertmp, *lower;
	struct path *stack;
	unsigned int stacklen, numlower, i;
	bool remote = false;

	err = -ENOMEM;
	lowertmp = kstrdup(ufs->config.lowerdir, GFP_KERNEL);
	if (!lowertmp)
		goto out;

	err = -EINVAL;
	stacklen = ovl_split_lowerdirs(lowertmp);
	if (stacklen > OVL_MAX_STACK) {
		pr_err("overlayfs: too many lower directories, limit is %d\n",
		       OVL_MAX_STACK);
		goto out;
	} else if (!ufs->config.upperdir && stacklen == 1) {
		pr_err("overlayfs: at least 2 lowerdir are needed while upperdir nonexistent\n");
		goto out;
	}

	err = -ENOMEM;
	stack = kcalloc(stacklen, sizeof(struct path), GFP_KERNEL);
	if (!stack)
		goto out;

	err = -EINVAL;
	lower = lowertmp;
	for (numlower = 0; numlower < stacklen; numlower++) {
		err = ovl_lower_dir(lower, &stack[numlower], ufs,
				    &sb->s_stack_depth, &remote);
		if (err)
			goto out_free_stack;

		lower = strchr(lower, '\0') + 1;
	}

	err = -EINVAL;
	sb->s_stack_depth++;
	if (sb->s_stack_depth > FILESYSTEM_MAX_STACK_DEPTH) {
		pr_err("overlayfs: maximum fs stacking depth exceeded\n");
		goto out_free_stack;
	}

	*stackp = stack;
	*stacklenp = numlower;

	if (remote)
		sb->s_d_op = &ovl_reval_dentry_operations;
	else
		sb->s_d_op = &ovl_dentry_operations;

	err = 0;

out:
	kfree(lowertmp);
	return err;

out_free_stack:
	for (i = 0; i < numlower; i++)
		path_put(&stack[i]);
	kfree(stack);
	goto out;
}

M
Miklos Szeredi 已提交
1124 1125
static int ovl_fill_super(struct super_block *sb, void *data, int silent)
{
K
Kees Cook 已提交
1126 1127
	struct path upperpath = { };
	struct path workpath = { };
M
Miklos Szeredi 已提交
1128
	struct dentry *root_dentry;
1129
	struct ovl_entry *oe = NULL;
M
Miklos Szeredi 已提交
1130
	struct ovl_fs *ufs;
1131
	struct path *stack = NULL;
1132
	unsigned int numlower = 0;
1133
	unsigned int i;
1134
	struct cred *cred;
M
Miklos Szeredi 已提交
1135 1136
	int err;

E
Erez Zadok 已提交
1137 1138 1139
	err = -ENOMEM;
	ufs = kzalloc(sizeof(struct ovl_fs), GFP_KERNEL);
	if (!ufs)
M
Miklos Szeredi 已提交
1140 1141
		goto out;

1142 1143 1144 1145
	ufs->creator_cred = cred = prepare_creds();
	if (!cred)
		goto out_err;

1146
	ufs->config.redirect_dir = ovl_redirect_dir_def;
1147
	ufs->config.index = ovl_index_def;
E
Erez Zadok 已提交
1148 1149
	err = ovl_parse_opt((char *) data, &ufs->config);
	if (err)
1150
		goto out_err;
E
Erez Zadok 已提交
1151

M
Miklos Szeredi 已提交
1152
	err = -EINVAL;
M
Miklos Szeredi 已提交
1153
	if (!ufs->config.lowerdir) {
1154 1155
		if (!silent)
			pr_err("overlayfs: missing 'lowerdir'\n");
1156
		goto out_err;
M
Miklos Szeredi 已提交
1157 1158
	}

M
Miklos Szeredi 已提交
1159
	sb->s_stack_depth = 0;
1160
	sb->s_maxbytes = MAX_LFS_FILESIZE;
M
Miklos Szeredi 已提交
1161 1162 1163
	if (ufs->config.upperdir) {
		if (!ufs->config.workdir) {
			pr_err("overlayfs: missing 'workdir'\n");
1164
			goto out_err;
M
Miklos Szeredi 已提交
1165
		}
M
Miklos Szeredi 已提交
1166

1167
		err = ovl_get_upper(ufs, &upperpath);
M
Miklos Szeredi 已提交
1168
		if (err)
1169
			goto out_err;
1170

1171
		err = ovl_get_workpath(ufs, &upperpath, &workpath);
1172
		if (err)
1173
			goto out_err;
M
Miklos Szeredi 已提交
1174

1175
		err = ovl_get_workdir(ufs, &workpath);
1176
		if (err)
1177
			goto out_err;
1178

1179 1180 1181
		if (!ufs->workdir)
			sb->s_flags |= MS_RDONLY;

1182 1183 1184
		sb->s_stack_depth = ufs->upper_mnt->mnt_sb->s_stack_depth;
		sb->s_time_gran = ufs->upper_mnt->mnt_sb->s_time_gran;

M
Miklos Szeredi 已提交
1185
	}
1186 1187 1188
	err = ovl_get_lowerstack(sb, ufs, &stack, &numlower);
	if (err)
		goto out_err;
M
Miklos Szeredi 已提交
1189

1190 1191
	err = ovl_get_lower_layers(ufs, stack, numlower);
	if (err)
1192
		goto out_err;
M
Miklos Szeredi 已提交
1193

H
hujianyang 已提交
1194 1195
	/* If the upper fs is nonexistent, we mark overlayfs r/o too */
	if (!ufs->upper_mnt)
M
Miklos Szeredi 已提交
1196
		sb->s_flags |= MS_RDONLY;
1197 1198
	else if (ufs->upper_mnt->mnt_sb != ufs->same_sb)
		ufs->same_sb = NULL;
M
Miklos Szeredi 已提交
1199

1200 1201 1202
	err = -ENOMEM;
	oe = ovl_alloc_entry(numlower);
	if (!oe)
1203
		goto out_err;
1204 1205 1206 1207 1208 1209

	for (i = 0; i < numlower; i++) {
		oe->lowerstack[i].dentry = stack[i].dentry;
		oe->lowerstack[i].layer = &(ufs->lower_layers[i]);
	}

1210
	if (!(ovl_force_readonly(ufs)) && ufs->config.index) {
1211
		err = ovl_get_indexdir(ufs, oe, &upperpath);
1212
		if (err)
1213
			goto out_err;
1214 1215 1216

		if (!ufs->indexdir)
			sb->s_flags |= MS_RDONLY;
1217 1218 1219 1220 1221 1222
	}

	/* Show index=off/on in /proc/mounts for any of the reasons above */
	if (!ufs->indexdir)
		ufs->config.index = false;

1223 1224 1225
	/* Never override disk quota limits or use reserved space */
	cap_lower(cred->cap_effective, CAP_SYS_RESOURCE);

1226 1227 1228 1229 1230 1231
	sb->s_magic = OVERLAYFS_SUPER_MAGIC;
	sb->s_op = &ovl_super_operations;
	sb->s_xattr = ovl_xattr_handlers;
	sb->s_fs_info = ufs;
	sb->s_flags |= MS_POSIXACL | MS_NOREMOTELOCK;

1232
	err = -ENOMEM;
1233
	root_dentry = d_make_root(ovl_new_inode(sb, S_IFDIR, 0));
M
Miklos Szeredi 已提交
1234
	if (!root_dentry)
1235
		goto out_err;
M
Miklos Szeredi 已提交
1236 1237

	mntput(upperpath.mnt);
1238 1239
	for (i = 0; i < numlower; i++)
		mntput(stack[i].mnt);
1240
	kfree(stack);
1241
	path_put(&workpath);
M
Miklos Szeredi 已提交
1242

1243
	if (upperpath.dentry) {
1244
		oe->has_upper = true;
M
Miklos Szeredi 已提交
1245 1246
		if (ovl_is_impuredir(upperpath.dentry))
			ovl_set_flag(OVL_IMPURE, d_inode(root_dentry));
1247
	}
M
Miklos Szeredi 已提交
1248 1249 1250

	root_dentry->d_fsdata = oe;

1251 1252
	/* Root is always merge -> can have whiteouts */
	ovl_set_flag(OVL_WHITEOUTS, d_inode(root_dentry));
1253 1254
	ovl_inode_init(d_inode(root_dentry), upperpath.dentry,
		       ovl_dentry_lower(root_dentry));
M
Miklos Szeredi 已提交
1255

M
Miklos Szeredi 已提交
1256 1257 1258 1259
	sb->s_root = root_dentry;

	return 0;

1260
out_err:
1261
	kfree(oe);
1262 1263 1264
	for (i = 0; i < numlower; i++)
		path_put(&stack[i]);
	kfree(stack);
1265
	path_put(&workpath);
M
Miklos Szeredi 已提交
1266
	path_put(&upperpath);
1267
	ovl_free_fs(ufs);
M
Miklos Szeredi 已提交
1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
out:
	return err;
}

static struct dentry *ovl_mount(struct file_system_type *fs_type, int flags,
				const char *dev_name, void *raw_data)
{
	return mount_nodev(fs_type, flags, raw_data, ovl_fill_super);
}

static struct file_system_type ovl_fs_type = {
	.owner		= THIS_MODULE,
1280
	.name		= "overlay",
M
Miklos Szeredi 已提交
1281 1282 1283
	.mount		= ovl_mount,
	.kill_sb	= kill_anon_super,
};
1284
MODULE_ALIAS_FS("overlay");
M
Miklos Szeredi 已提交
1285

1286 1287 1288 1289 1290 1291 1292
static void ovl_inode_init_once(void *foo)
{
	struct ovl_inode *oi = foo;

	inode_init_once(&oi->vfs_inode);
}

M
Miklos Szeredi 已提交
1293 1294
static int __init ovl_init(void)
{
1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
	int err;

	ovl_inode_cachep = kmem_cache_create("ovl_inode",
					     sizeof(struct ovl_inode), 0,
					     (SLAB_RECLAIM_ACCOUNT|
					      SLAB_MEM_SPREAD|SLAB_ACCOUNT),
					     ovl_inode_init_once);
	if (ovl_inode_cachep == NULL)
		return -ENOMEM;

	err = register_filesystem(&ovl_fs_type);
	if (err)
		kmem_cache_destroy(ovl_inode_cachep);

	return err;
M
Miklos Szeredi 已提交
1310 1311 1312 1313 1314
}

static void __exit ovl_exit(void)
{
	unregister_filesystem(&ovl_fs_type);
1315 1316 1317 1318 1319 1320 1321 1322

	/*
	 * Make sure all delayed rcu free inodes are flushed before we
	 * destroy cache.
	 */
	rcu_barrier();
	kmem_cache_destroy(ovl_inode_cachep);

M
Miklos Szeredi 已提交
1323 1324 1325 1326
}

module_init(ovl_init);
module_exit(ovl_exit);