super.c 25.3 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
#include "overlayfs.h"
M
Miklos Szeredi 已提交
21
#include "ovl_entry.h"
M
Miklos Szeredi 已提交
22 23 24 25 26 27 28 29

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


struct ovl_dir_cache;

30 31
#define OVL_MAX_STACK 500

32 33 34 35
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
36 37 38 39 40 41

static void ovl_dentry_release(struct dentry *dentry)
{
	struct ovl_entry *oe = dentry->d_fsdata;

	if (oe) {
42 43 44 45
		unsigned int i;

		for (i = 0; i < oe->numlower; i++)
			dput(oe->lowerstack[i].dentry);
M
Miklos Szeredi 已提交
46 47 48 49
		kfree_rcu(oe, rcu);
	}
}

50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
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;
}

66 67 68
static struct dentry *ovl_d_real(struct dentry *dentry,
				 const struct inode *inode,
				 unsigned int open_flags)
M
Miklos Szeredi 已提交
69 70
{
	struct dentry *real;
71
	int err;
M
Miklos Szeredi 已提交
72

73
	if (!d_is_reg(dentry)) {
M
Miklos Szeredi 已提交
74 75 76 77 78
		if (!inode || inode == d_inode(dentry))
			return dentry;
		goto bug;
	}

79 80 81 82
	if (d_is_negative(dentry))
		return dentry;

	if (open_flags) {
83
		err = ovl_open_maybe_copy_up(dentry, open_flags);
84 85 86 87
		if (err)
			return ERR_PTR(err);
	}

M
Miklos Szeredi 已提交
88
	real = ovl_dentry_upper(dentry);
89 90 91 92 93 94
	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
95
		return real;
96
	}
M
Miklos Szeredi 已提交
97 98 99 100 101

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

M
Miklos Szeredi 已提交
102 103 104
	/* Handle recursion */
	real = d_real(real, inode, open_flags);

M
Miklos Szeredi 已提交
105 106 107
	if (!inode || inode == d_inode(real))
		return real;
bug:
M
Miklos Szeredi 已提交
108
	WARN(1, "ovl_d_real(%pd4, %s:%lu): real dentry not found\n", dentry,
M
Miklos Szeredi 已提交
109 110 111 112
	     inode ? inode->i_sb->s_id : "NULL", inode ? inode->i_ino : 0);
	return dentry;
}

113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153
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
154 155
static const struct dentry_operations ovl_dentry_operations = {
	.d_release = ovl_dentry_release,
M
Miklos Szeredi 已提交
156
	.d_real = ovl_d_real,
M
Miklos Szeredi 已提交
157 158
};

159 160
static const struct dentry_operations ovl_reval_dentry_operations = {
	.d_release = ovl_dentry_release,
M
Miklos Szeredi 已提交
161
	.d_real = ovl_d_real,
162 163 164 165
	.d_revalidate = ovl_dentry_revalidate,
	.d_weak_revalidate = ovl_dentry_weak_revalidate,
};

166 167 168 169 170 171
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);

M
Miklos Szeredi 已提交
172
	oi->redirect = NULL;
173
	oi->__upperdentry = NULL;
174 175
	oi->lower = NULL;

176 177 178 179 180 181 182 183 184 185 186 187
	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)
{
188 189 190
	struct ovl_inode *oi = OVL_I(inode);

	dput(oi->__upperdentry);
M
Miklos Szeredi 已提交
191
	kfree(oi->redirect);
192

193 194 195
	call_rcu(&inode->i_rcu, ovl_i_callback);
}

M
Miklos Szeredi 已提交
196 197 198
static void ovl_put_super(struct super_block *sb)
{
	struct ovl_fs *ufs = sb->s_fs_info;
199
	unsigned i;
M
Miklos Szeredi 已提交
200 201 202

	dput(ufs->workdir);
	mntput(ufs->upper_mnt);
203 204
	for (i = 0; i < ufs->numlower; i++)
		mntput(ufs->lower_mnt[i]);
205
	kfree(ufs->lower_mnt);
M
Miklos Szeredi 已提交
206

E
Erez Zadok 已提交
207 208 209
	kfree(ufs->config.lowerdir);
	kfree(ufs->config.upperdir);
	kfree(ufs->config.workdir);
210
	put_cred(ufs->creator_cred);
M
Miklos Szeredi 已提交
211 212 213
	kfree(ufs);
}

214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232
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
233 234 235 236 237 238
/**
 * 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
239
 * filesystem pass the statfs to the upper filesystem (if it exists)
A
Andy Whitcroft 已提交
240 241 242 243 244 245 246 247
 */
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;

248
	ovl_path_real(root_dentry, &path);
A
Andy Whitcroft 已提交
249 250 251

	err = vfs_statfs(&path, buf);
	if (!err) {
M
Miklos Szeredi 已提交
252
		buf->f_namelen = ofs->namelen;
A
Andy Whitcroft 已提交
253 254 255 256 257 258
		buf->f_type = OVERLAYFS_SUPER_MAGIC;
	}

	return err;
}

E
Erez Zadok 已提交
259 260 261 262 263 264 265 266 267 268 269
/**
 * 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;

270
	seq_show_option(m, "lowerdir", ufs->config.lowerdir);
M
Miklos Szeredi 已提交
271
	if (ufs->config.upperdir) {
272 273
		seq_show_option(m, "upperdir", ufs->config.upperdir);
		seq_show_option(m, "workdir", ufs->config.workdir);
M
Miklos Szeredi 已提交
274
	}
M
Miklos Szeredi 已提交
275 276
	if (ufs->config.default_permissions)
		seq_puts(m, ",default_permissions");
277 278 279
	if (ufs->config.redirect_dir != ovl_redirect_dir_def)
		seq_printf(m, ",redirect_dir=%s",
			   ufs->config.redirect_dir ? "on" : "off");
E
Erez Zadok 已提交
280 281 282
	return 0;
}

283 284 285 286
static int ovl_remount(struct super_block *sb, int *flags, char *data)
{
	struct ovl_fs *ufs = sb->s_fs_info;

287
	if (!(*flags & MS_RDONLY) && (!ufs->upper_mnt || !ufs->workdir))
288 289 290 291 292
		return -EROFS;

	return 0;
}

M
Miklos Szeredi 已提交
293
static const struct super_operations ovl_super_operations = {
294 295 296
	.alloc_inode	= ovl_alloc_inode,
	.destroy_inode	= ovl_destroy_inode,
	.drop_inode	= generic_delete_inode,
M
Miklos Szeredi 已提交
297
	.put_super	= ovl_put_super,
298
	.sync_fs	= ovl_sync_fs,
A
Andy Whitcroft 已提交
299
	.statfs		= ovl_statfs,
E
Erez Zadok 已提交
300
	.show_options	= ovl_show_options,
301
	.remount_fs	= ovl_remount,
M
Miklos Szeredi 已提交
302 303 304 305 306 307
};

enum {
	OPT_LOWERDIR,
	OPT_UPPERDIR,
	OPT_WORKDIR,
M
Miklos Szeredi 已提交
308
	OPT_DEFAULT_PERMISSIONS,
M
Miklos Szeredi 已提交
309 310
	OPT_REDIRECT_DIR_ON,
	OPT_REDIRECT_DIR_OFF,
M
Miklos Szeredi 已提交
311 312 313 314 315 316 317
	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
318
	{OPT_DEFAULT_PERMISSIONS,	"default_permissions"},
M
Miklos Szeredi 已提交
319 320
	{OPT_REDIRECT_DIR_ON,		"redirect_dir=on"},
	{OPT_REDIRECT_DIR_OFF,		"redirect_dir=off"},
M
Miklos Szeredi 已提交
321 322 323
	{OPT_ERR,			NULL}
};

M
Miklos Szeredi 已提交
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346
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
347 348 349 350
static int ovl_parse_opt(char *opt, struct ovl_config *config)
{
	char *p;

M
Miklos Szeredi 已提交
351
	while ((p = ovl_next_opt(&opt)) != NULL) {
M
Miklos Szeredi 已提交
352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380
		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
381 382 383 384
		case OPT_DEFAULT_PERMISSIONS:
			config->default_permissions = true;
			break;

M
Miklos Szeredi 已提交
385 386 387 388 389 390 391 392
		case OPT_REDIRECT_DIR_ON:
			config->redirect_dir = true;
			break;

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

M
Miklos Szeredi 已提交
393
		default:
394
			pr_err("overlayfs: unrecognized mount option \"%s\" or missing value\n", p);
M
Miklos Szeredi 已提交
395 396 397
			return -EINVAL;
		}
	}
H
hujianyang 已提交
398 399 400 401 402 403 404 405 406

	/* 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
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423
	return 0;
}

#define OVL_WORKDIR_NAME "work"

static struct dentry *ovl_workdir_create(struct vfsmount *mnt,
					 struct dentry *dentry)
{
	struct inode *dir = dentry->d_inode;
	struct dentry *work;
	int err;
	bool retried = false;

	err = mnt_want_write(mnt);
	if (err)
		return ERR_PTR(err);

A
Al Viro 已提交
424
	inode_lock_nested(dir, I_MUTEX_PARENT);
M
Miklos Szeredi 已提交
425 426 427 428 429
retry:
	work = lookup_one_len(OVL_WORKDIR_NAME, dentry,
			      strlen(OVL_WORKDIR_NAME));

	if (!IS_ERR(work)) {
430 431
		struct iattr attr = {
			.ia_valid = ATTR_MODE,
A
Al Viro 已提交
432
			.ia_mode = S_IFDIR | 0,
433
		};
M
Miklos Szeredi 已提交
434 435 436 437 438 439 440

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

			retried = true;
M
Miklos Szeredi 已提交
441
			ovl_workdir_cleanup(dir, mnt, work, 0);
M
Miklos Szeredi 已提交
442 443 444 445
			dput(work);
			goto retry;
		}

A
Al Viro 已提交
446 447 448
		err = ovl_create_real(dir, work,
				      &(struct cattr){.mode = S_IFDIR | 0},
				      NULL, true);
M
Miklos Szeredi 已提交
449 450
		if (err)
			goto out_dput;
451

452 453 454 455 456 457 458 459 460 461 462 463 464
		/*
		 * 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.
		 */
465
		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_DEFAULT);
M
Miklos Szeredi 已提交
466
		if (err && err != -ENODATA && err != -EOPNOTSUPP)
467 468 469
			goto out_dput;

		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_ACCESS);
M
Miklos Szeredi 已提交
470
		if (err && err != -ENODATA && err != -EOPNOTSUPP)
471 472 473 474 475 476 477 478
			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;
M
Miklos Szeredi 已提交
479 480
	}
out_unlock:
A
Al Viro 已提交
481
	inode_unlock(dir);
M
Miklos Szeredi 已提交
482 483 484 485 486 487 488 489 490 491
	mnt_drop_write(mnt);

	return work;

out_dput:
	dput(work);
	work = ERR_PTR(err);
	goto out_unlock;
}

M
Miklos Szeredi 已提交
492 493 494 495 496 497 498 499 500 501 502 503 504
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
505 506
static int ovl_mount_dir_noesc(const char *name, struct path *path)
{
507
	int err = -EINVAL;
M
Miklos Szeredi 已提交
508

509 510 511 512
	if (!*name) {
		pr_err("overlayfs: empty lowerdir\n");
		goto out;
	}
M
Miklos Szeredi 已提交
513 514 515 516 517 518
	err = kern_path(name, LOOKUP_FOLLOW, path);
	if (err) {
		pr_err("overlayfs: failed to resolve '%s': %i\n", name, err);
		goto out;
	}
	err = -EINVAL;
519
	if (ovl_dentry_weird(path->dentry)) {
M
Miklos Szeredi 已提交
520 521 522
		pr_err("overlayfs: filesystem on '%s' not supported\n", name);
		goto out_put;
	}
M
Miklos Szeredi 已提交
523
	if (!d_is_dir(path->dentry)) {
M
Miklos Szeredi 已提交
524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542
		pr_err("overlayfs: '%s' not a directory\n", name);
		goto out_put;
	}
	return 0;

out_put:
	path_put(path);
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);
543 544 545 546 547 548 549 550

		if (!err)
			if (ovl_dentry_remote(path->dentry)) {
				pr_err("overlayfs: filesystem on '%s' not supported as upperdir\n",
				       tmp);
				path_put(path);
				err = -EINVAL;
			}
M
Miklos Szeredi 已提交
551 552 553 554 555
		kfree(tmp);
	}
	return err;
}

M
Miklos Szeredi 已提交
556 557
static int ovl_check_namelen(struct path *path, struct ovl_fs *ofs,
			     const char *name)
M
Miklos Szeredi 已提交
558 559
{
	struct kstatfs statfs;
M
Miklos Szeredi 已提交
560 561 562 563 564 565 566 567 568 569 570 571 572 573
	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
574

575
	err = ovl_mount_dir_noesc(name, path);
M
Miklos Szeredi 已提交
576 577 578
	if (err)
		goto out;

M
Miklos Szeredi 已提交
579 580
	err = ovl_check_namelen(path, ofs, name);
	if (err)
M
Miklos Szeredi 已提交
581
		goto out_put;
M
Miklos Szeredi 已提交
582

M
Miklos Szeredi 已提交
583 584
	*stack_depth = max(*stack_depth, path->mnt->mnt_sb->s_stack_depth);

585 586 587
	if (ovl_dentry_remote(path->dentry))
		*remote = true;

M
Miklos Szeredi 已提交
588 589 590 591 592 593 594 595
	return 0;

out_put:
	path_put(path);
out:
	return err;
}

M
Miklos Szeredi 已提交
596 597 598 599 600 601 602 603 604 605 606 607
/* 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;
}

608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627
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;
}

628 629 630 631 632 633 634 635
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)
{
	return ovl_xattr_get(dentry, handler->name, buffer, size);
}

636 637 638 639 640
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
641 642
{
	struct dentry *workdir = ovl_workdir(dentry);
643
	struct inode *realinode = ovl_inode_real(inode);
M
Miklos Szeredi 已提交
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667
	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);

668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
	/*
	 * 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;
	}

683 684
	err = ovl_xattr_set(dentry, handler->name, value, size, flags);
	if (!err)
685
		ovl_copyattr(ovl_inode_real(inode), inode);
686 687

	return err;
M
Miklos Szeredi 已提交
688 689 690 691 692 693

out_acl_release:
	posix_acl_release(acl);
	return err;
}

694 695 696 697
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
698
	return -EOPNOTSUPP;
699 700
}

M
Miklos Szeredi 已提交
701 702 703 704 705
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
706
	return -EOPNOTSUPP;
M
Miklos Szeredi 已提交
707 708
}

709 710 711 712 713 714 715
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)
{
	return ovl_xattr_get(dentry, name, buffer, size);
}

716 717 718 719 720 721 722 723
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)
{
	return ovl_xattr_set(dentry, name, value, size, flags);
}

724 725
static const struct xattr_handler __maybe_unused
ovl_posix_acl_access_xattr_handler = {
M
Miklos Szeredi 已提交
726 727
	.name = XATTR_NAME_POSIX_ACL_ACCESS,
	.flags = ACL_TYPE_ACCESS,
728
	.get = ovl_posix_acl_xattr_get,
M
Miklos Szeredi 已提交
729 730 731
	.set = ovl_posix_acl_xattr_set,
};

732 733
static const struct xattr_handler __maybe_unused
ovl_posix_acl_default_xattr_handler = {
M
Miklos Szeredi 已提交
734 735
	.name = XATTR_NAME_POSIX_ACL_DEFAULT,
	.flags = ACL_TYPE_DEFAULT,
736
	.get = ovl_posix_acl_xattr_get,
M
Miklos Szeredi 已提交
737 738 739 740 741
	.set = ovl_posix_acl_xattr_set,
};

static const struct xattr_handler ovl_own_xattr_handler = {
	.prefix	= OVL_XATTR_PREFIX,
742
	.get = ovl_own_xattr_get,
M
Miklos Szeredi 已提交
743 744 745 746 747
	.set = ovl_own_xattr_set,
};

static const struct xattr_handler ovl_other_xattr_handler = {
	.prefix	= "", /* catch all */
748
	.get = ovl_other_xattr_get,
M
Miklos Szeredi 已提交
749 750 751 752
	.set = ovl_other_xattr_set,
};

static const struct xattr_handler *ovl_xattr_handlers[] = {
753
#ifdef CONFIG_FS_POSIX_ACL
M
Miklos Szeredi 已提交
754 755
	&ovl_posix_acl_access_xattr_handler,
	&ovl_posix_acl_default_xattr_handler,
756
#endif
M
Miklos Szeredi 已提交
757 758 759 760 761
	&ovl_own_xattr_handler,
	&ovl_other_xattr_handler,
	NULL
};

M
Miklos Szeredi 已提交
762 763
static int ovl_fill_super(struct super_block *sb, void *data, int silent)
{
K
Kees Cook 已提交
764 765
	struct path upperpath = { };
	struct path workpath = { };
M
Miklos Szeredi 已提交
766 767 768
	struct dentry *root_dentry;
	struct ovl_entry *oe;
	struct ovl_fs *ufs;
769 770 771 772 773
	struct path *stack = NULL;
	char *lowertmp;
	char *lower;
	unsigned int numlower;
	unsigned int stacklen = 0;
774
	unsigned int i;
775
	bool remote = false;
776
	struct cred *cred;
M
Miklos Szeredi 已提交
777 778
	int err;

E
Erez Zadok 已提交
779 780 781
	err = -ENOMEM;
	ufs = kzalloc(sizeof(struct ovl_fs), GFP_KERNEL);
	if (!ufs)
M
Miklos Szeredi 已提交
782 783
		goto out;

784
	init_waitqueue_head(&ufs->copyup_wq);
785
	ufs->config.redirect_dir = ovl_redirect_dir_def;
E
Erez Zadok 已提交
786 787 788 789
	err = ovl_parse_opt((char *) data, &ufs->config);
	if (err)
		goto out_free_config;

M
Miklos Szeredi 已提交
790
	err = -EINVAL;
M
Miklos Szeredi 已提交
791
	if (!ufs->config.lowerdir) {
792 793
		if (!silent)
			pr_err("overlayfs: missing 'lowerdir'\n");
M
Miklos Szeredi 已提交
794 795 796
		goto out_free_config;
	}

M
Miklos Szeredi 已提交
797
	sb->s_stack_depth = 0;
798
	sb->s_maxbytes = MAX_LFS_FILESIZE;
M
Miklos Szeredi 已提交
799 800 801 802 803
	if (ufs->config.upperdir) {
		if (!ufs->config.workdir) {
			pr_err("overlayfs: missing 'workdir'\n");
			goto out_free_config;
		}
M
Miklos Szeredi 已提交
804

M
Miklos Szeredi 已提交
805 806 807
		err = ovl_mount_dir(ufs->config.upperdir, &upperpath);
		if (err)
			goto out_free_config;
M
Miklos Szeredi 已提交
808

H
hujianyang 已提交
809 810 811 812 813 814 815
		/* Upper fs should not be r/o */
		if (upperpath.mnt->mnt_sb->s_flags & MS_RDONLY) {
			pr_err("overlayfs: upper fs is r/o, try multi-lower layers mount\n");
			err = -EINVAL;
			goto out_put_upperpath;
		}

M
Miklos Szeredi 已提交
816 817 818 819
		err = ovl_check_namelen(&upperpath, ufs, ufs->config.upperdir);
		if (err)
			goto out_put_upperpath;

M
Miklos Szeredi 已提交
820 821 822 823
		err = ovl_mount_dir(ufs->config.workdir, &workpath);
		if (err)
			goto out_put_upperpath;

824
		err = -EINVAL;
M
Miklos Szeredi 已提交
825 826 827 828 829 830 831 832 833
		if (upperpath.mnt != workpath.mnt) {
			pr_err("overlayfs: workdir and upperdir must reside under the same mount\n");
			goto out_put_workpath;
		}
		if (!ovl_workdir_ok(workpath.dentry, upperpath.dentry)) {
			pr_err("overlayfs: workdir and upperdir must be separate subtrees\n");
			goto out_put_workpath;
		}
		sb->s_stack_depth = upperpath.mnt->mnt_sb->s_stack_depth;
A
Andy Whitcroft 已提交
834
	}
835 836 837
	err = -ENOMEM;
	lowertmp = kstrdup(ufs->config.lowerdir, GFP_KERNEL);
	if (!lowertmp)
M
Miklos Szeredi 已提交
838
		goto out_put_workpath;
839

840 841
	err = -EINVAL;
	stacklen = ovl_split_lowerdirs(lowertmp);
842
	if (stacklen > OVL_MAX_STACK) {
843
		pr_err("overlayfs: too many lower directories, limit is %d\n",
844
		       OVL_MAX_STACK);
845
		goto out_free_lowertmp;
846 847 848 849
	} else if (!ufs->config.upperdir && stacklen == 1) {
		pr_err("overlayfs: at least 2 lowerdir are needed while upperdir nonexistent\n");
		goto out_free_lowertmp;
	}
850

851
	err = -ENOMEM;
852 853 854 855
	stack = kcalloc(stacklen, sizeof(struct path), GFP_KERNEL);
	if (!stack)
		goto out_free_lowertmp;

856
	err = -EINVAL;
857 858
	lower = lowertmp;
	for (numlower = 0; numlower < stacklen; numlower++) {
M
Miklos Szeredi 已提交
859 860
		err = ovl_lower_dir(lower, &stack[numlower], ufs,
				    &sb->s_stack_depth, &remote);
861 862 863 864 865 866
		if (err)
			goto out_put_lowerpath;

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

867
	err = -EINVAL;
M
Miklos Szeredi 已提交
868
	sb->s_stack_depth++;
869 870
	if (sb->s_stack_depth > FILESYSTEM_MAX_STACK_DEPTH) {
		pr_err("overlayfs: maximum fs stacking depth exceeded\n");
871
		goto out_put_lowerpath;
872 873
	}

M
Miklos Szeredi 已提交
874 875 876 877 878 879 880
	if (ufs->config.upperdir) {
		ufs->upper_mnt = clone_private_mount(&upperpath);
		err = PTR_ERR(ufs->upper_mnt);
		if (IS_ERR(ufs->upper_mnt)) {
			pr_err("overlayfs: failed to clone upperpath\n");
			goto out_put_lowerpath;
		}
M
Miklos Szeredi 已提交
881 882 883 884
		/* Don't inherit atime flags */
		ufs->upper_mnt->mnt_flags &= ~(MNT_NOATIME | MNT_NODIRATIME | MNT_RELATIME);

		sb->s_time_gran = ufs->upper_mnt->mnt_sb->s_time_gran;
885

M
Miklos Szeredi 已提交
886 887 888
		ufs->workdir = ovl_workdir_create(ufs->upper_mnt, workpath.dentry);
		err = PTR_ERR(ufs->workdir);
		if (IS_ERR(ufs->workdir)) {
889 890 891 892
			pr_warn("overlayfs: failed to create directory %s/%s (errno: %i); mounting read-only\n",
				ufs->config.workdir, OVL_WORKDIR_NAME, -err);
			sb->s_flags |= MS_RDONLY;
			ufs->workdir = NULL;
M
Miklos Szeredi 已提交
893
		}
894 895 896 897

		/*
		 * Upper should support d_type, else whiteouts are visible.
		 * Given workdir and upper are on same fs, we can do
898 899
		 * iterate_dir() on workdir. This check requires successful
		 * creation of workdir in previous step.
900
		 */
901
		if (ufs->workdir) {
902 903
			struct dentry *temp;

904 905 906
			err = ovl_check_d_type_supported(&workpath);
			if (err < 0)
				goto out_put_workdir;
907

908 909 910 911 912 913 914
			/*
			 * 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");
915 916 917 918 919 920 921 922

			/* 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");
923 924 925 926 927 928 929 930 931 932 933 934 935

			/*
			 * 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);
			}
936
		}
M
Miklos Szeredi 已提交
937 938
	}

939
	err = -ENOMEM;
940
	ufs->lower_mnt = kcalloc(numlower, sizeof(struct vfsmount *), GFP_KERNEL);
941
	if (ufs->lower_mnt == NULL)
942
		goto out_put_workdir;
943 944
	for (i = 0; i < numlower; i++) {
		struct vfsmount *mnt = clone_private_mount(&stack[i]);
945

946
		err = PTR_ERR(mnt);
947 948 949 950 951 952 953 954
		if (IS_ERR(mnt)) {
			pr_err("overlayfs: failed to clone lowerpath\n");
			goto out_put_lower_mnt;
		}
		/*
		 * Make lower_mnt R/O.  That way fchmod/fchown on lower file
		 * will fail instead of modifying lower fs.
		 */
M
Miklos Szeredi 已提交
955
		mnt->mnt_flags |= MNT_READONLY | MNT_NOATIME;
956

957 958
		ufs->lower_mnt[ufs->numlower] = mnt;
		ufs->numlower++;
959 960 961 962 963 964

		/* 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;
965
	}
M
Miklos Szeredi 已提交
966

H
hujianyang 已提交
967 968
	/* If the upper fs is nonexistent, we mark overlayfs r/o too */
	if (!ufs->upper_mnt)
M
Miklos Szeredi 已提交
969
		sb->s_flags |= MS_RDONLY;
970 971
	else if (ufs->upper_mnt->mnt_sb != ufs->same_sb)
		ufs->same_sb = NULL;
M
Miklos Szeredi 已提交
972

973 974 975 976
	if (remote)
		sb->s_d_op = &ovl_reval_dentry_operations;
	else
		sb->s_d_op = &ovl_dentry_operations;
M
Miklos Szeredi 已提交
977

978 979
	ufs->creator_cred = cred = prepare_creds();
	if (!cred)
980 981
		goto out_put_lower_mnt;

982 983 984
	/* Never override disk quota limits or use reserved space */
	cap_lower(cred->cap_effective, CAP_SYS_RESOURCE);

M
Miklos Szeredi 已提交
985
	err = -ENOMEM;
986
	oe = ovl_alloc_entry(numlower);
987
	if (!oe)
988
		goto out_put_cred;
M
Miklos Szeredi 已提交
989

990 991 992 993 994 995
	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;

996
	root_dentry = d_make_root(ovl_new_inode(sb, S_IFDIR, 0));
M
Miklos Szeredi 已提交
997
	if (!root_dentry)
998
		goto out_free_oe;
M
Miklos Szeredi 已提交
999 1000

	mntput(upperpath.mnt);
1001 1002
	for (i = 0; i < numlower; i++)
		mntput(stack[i].mnt);
M
Miklos Szeredi 已提交
1003
	path_put(&workpath);
1004
	kfree(lowertmp);
M
Miklos Szeredi 已提交
1005

1006 1007 1008
	if (upperpath.dentry) {
		oe->impure = ovl_is_impuredir(upperpath.dentry);
	}
1009 1010 1011 1012
	for (i = 0; i < numlower; i++) {
		oe->lowerstack[i].dentry = stack[i].dentry;
		oe->lowerstack[i].mnt = ufs->lower_mnt[i];
	}
1013
	kfree(stack);
M
Miklos Szeredi 已提交
1014 1015 1016

	root_dentry->d_fsdata = oe;

1017 1018
	ovl_inode_init(d_inode(root_dentry), upperpath.dentry,
		       ovl_dentry_lower(root_dentry));
M
Miklos Szeredi 已提交
1019

M
Miklos Szeredi 已提交
1020 1021 1022 1023
	sb->s_root = root_dentry;

	return 0;

1024 1025
out_free_oe:
	kfree(oe);
1026 1027
out_put_cred:
	put_cred(ufs->creator_cred);
M
Miklos Szeredi 已提交
1028
out_put_lower_mnt:
1029 1030 1031
	for (i = 0; i < ufs->numlower; i++)
		mntput(ufs->lower_mnt[i]);
	kfree(ufs->lower_mnt);
1032 1033
out_put_workdir:
	dput(ufs->workdir);
M
Miklos Szeredi 已提交
1034 1035
	mntput(ufs->upper_mnt);
out_put_lowerpath:
1036 1037 1038 1039 1040
	for (i = 0; i < numlower; i++)
		path_put(&stack[i]);
	kfree(stack);
out_free_lowertmp:
	kfree(lowertmp);
1041 1042
out_put_workpath:
	path_put(&workpath);
M
Miklos Szeredi 已提交
1043 1044 1045
out_put_upperpath:
	path_put(&upperpath);
out_free_config:
E
Erez Zadok 已提交
1046 1047 1048 1049
	kfree(ufs->config.lowerdir);
	kfree(ufs->config.upperdir);
	kfree(ufs->config.workdir);
	kfree(ufs);
M
Miklos Szeredi 已提交
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
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,
1062
	.name		= "overlay",
M
Miklos Szeredi 已提交
1063 1064 1065
	.mount		= ovl_mount,
	.kill_sb	= kill_anon_super,
};
1066
MODULE_ALIAS_FS("overlay");
M
Miklos Szeredi 已提交
1067

1068 1069 1070 1071 1072 1073 1074
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
1075 1076
static int __init ovl_init(void)
{
1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091
	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
1092 1093 1094 1095 1096
}

static void __exit ovl_exit(void)
{
	unregister_filesystem(&ovl_fs_type);
1097 1098 1099 1100 1101 1102 1103 1104

	/*
	 * 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
1105 1106 1107 1108
}

module_init(ovl_init);
module_exit(ovl_exit);