super.c 24.1 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
		unsigned int i;

M
Miklos Szeredi 已提交
44
		dput(oe->__upperdentry);
M
Miklos Szeredi 已提交
45
		kfree(oe->redirect);
46 47
		for (i = 0; i < oe->numlower; i++)
			dput(oe->lowerstack[i].dentry);
M
Miklos Szeredi 已提交
48 49 50 51
		kfree_rcu(oe, rcu);
	}
}

52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
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;
}

68 69 70
static struct dentry *ovl_d_real(struct dentry *dentry,
				 const struct inode *inode,
				 unsigned int open_flags)
M
Miklos Szeredi 已提交
71 72
{
	struct dentry *real;
73
	int err;
M
Miklos Szeredi 已提交
74

75
	if (!d_is_reg(dentry)) {
M
Miklos Szeredi 已提交
76 77 78 79 80
		if (!inode || inode == d_inode(dentry))
			return dentry;
		goto bug;
	}

81 82 83 84
	if (d_is_negative(dentry))
		return dentry;

	if (open_flags) {
85
		err = ovl_open_maybe_copy_up(dentry, open_flags);
86 87 88 89
		if (err)
			return ERR_PTR(err);
	}

M
Miklos Szeredi 已提交
90
	real = ovl_dentry_upper(dentry);
91 92 93 94 95 96
	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
97
		return real;
98
	}
M
Miklos Szeredi 已提交
99 100 101 102 103

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

M
Miklos Szeredi 已提交
104 105 106
	/* Handle recursion */
	real = d_real(real, inode, open_flags);

M
Miklos Szeredi 已提交
107 108 109
	if (!inode || inode == d_inode(real))
		return real;
bug:
M
Miklos Szeredi 已提交
110
	WARN(1, "ovl_d_real(%pd4, %s:%lu): real dentry not found\n", dentry,
M
Miklos Szeredi 已提交
111 112 113 114
	     inode ? inode->i_sb->s_id : "NULL", inode ? inode->i_ino : 0);
	return dentry;
}

115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
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
156 157
static const struct dentry_operations ovl_dentry_operations = {
	.d_release = ovl_dentry_release,
M
Miklos Szeredi 已提交
158
	.d_real = ovl_d_real,
M
Miklos Szeredi 已提交
159 160
};

161 162
static const struct dentry_operations ovl_reval_dentry_operations = {
	.d_release = ovl_dentry_release,
M
Miklos Szeredi 已提交
163
	.d_real = ovl_d_real,
164 165 166 167
	.d_revalidate = ovl_dentry_revalidate,
	.d_weak_revalidate = ovl_dentry_weak_revalidate,
};

M
Miklos Szeredi 已提交
168 169 170
static void ovl_put_super(struct super_block *sb)
{
	struct ovl_fs *ufs = sb->s_fs_info;
171
	unsigned i;
M
Miklos Szeredi 已提交
172 173 174

	dput(ufs->workdir);
	mntput(ufs->upper_mnt);
175 176
	for (i = 0; i < ufs->numlower; i++)
		mntput(ufs->lower_mnt[i]);
177
	kfree(ufs->lower_mnt);
M
Miklos Szeredi 已提交
178

E
Erez Zadok 已提交
179 180 181
	kfree(ufs->config.lowerdir);
	kfree(ufs->config.upperdir);
	kfree(ufs->config.workdir);
182
	put_cred(ufs->creator_cred);
M
Miklos Szeredi 已提交
183 184 185
	kfree(ufs);
}

186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204
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
205 206 207 208 209 210
/**
 * 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
211
 * filesystem pass the statfs to the upper filesystem (if it exists)
A
Andy Whitcroft 已提交
212 213 214 215 216 217 218 219
 */
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;

220
	ovl_path_real(root_dentry, &path);
A
Andy Whitcroft 已提交
221 222 223

	err = vfs_statfs(&path, buf);
	if (!err) {
M
Miklos Szeredi 已提交
224
		buf->f_namelen = ofs->namelen;
A
Andy Whitcroft 已提交
225 226 227 228 229 230
		buf->f_type = OVERLAYFS_SUPER_MAGIC;
	}

	return err;
}

E
Erez Zadok 已提交
231 232 233 234 235 236 237 238 239 240 241
/**
 * 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;

242
	seq_show_option(m, "lowerdir", ufs->config.lowerdir);
M
Miklos Szeredi 已提交
243
	if (ufs->config.upperdir) {
244 245
		seq_show_option(m, "upperdir", ufs->config.upperdir);
		seq_show_option(m, "workdir", ufs->config.workdir);
M
Miklos Szeredi 已提交
246
	}
M
Miklos Szeredi 已提交
247 248
	if (ufs->config.default_permissions)
		seq_puts(m, ",default_permissions");
249 250 251
	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
252 253 254
	return 0;
}

255 256 257 258
static int ovl_remount(struct super_block *sb, int *flags, char *data)
{
	struct ovl_fs *ufs = sb->s_fs_info;

259
	if (!(*flags & MS_RDONLY) && (!ufs->upper_mnt || !ufs->workdir))
260 261 262 263 264
		return -EROFS;

	return 0;
}

M
Miklos Szeredi 已提交
265 266
static const struct super_operations ovl_super_operations = {
	.put_super	= ovl_put_super,
267
	.sync_fs	= ovl_sync_fs,
A
Andy Whitcroft 已提交
268
	.statfs		= ovl_statfs,
E
Erez Zadok 已提交
269
	.show_options	= ovl_show_options,
270
	.remount_fs	= ovl_remount,
M
Miklos Szeredi 已提交
271
	.drop_inode	= generic_delete_inode,
M
Miklos Szeredi 已提交
272 273 274 275 276 277
};

enum {
	OPT_LOWERDIR,
	OPT_UPPERDIR,
	OPT_WORKDIR,
M
Miklos Szeredi 已提交
278
	OPT_DEFAULT_PERMISSIONS,
M
Miklos Szeredi 已提交
279 280
	OPT_REDIRECT_DIR_ON,
	OPT_REDIRECT_DIR_OFF,
M
Miklos Szeredi 已提交
281 282 283 284 285 286 287
	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
288
	{OPT_DEFAULT_PERMISSIONS,	"default_permissions"},
M
Miklos Szeredi 已提交
289 290
	{OPT_REDIRECT_DIR_ON,		"redirect_dir=on"},
	{OPT_REDIRECT_DIR_OFF,		"redirect_dir=off"},
M
Miklos Szeredi 已提交
291 292 293
	{OPT_ERR,			NULL}
};

M
Miklos Szeredi 已提交
294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316
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
317 318 319 320
static int ovl_parse_opt(char *opt, struct ovl_config *config)
{
	char *p;

M
Miklos Szeredi 已提交
321
	while ((p = ovl_next_opt(&opt)) != NULL) {
M
Miklos Szeredi 已提交
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350
		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
351 352 353 354
		case OPT_DEFAULT_PERMISSIONS:
			config->default_permissions = true;
			break;

M
Miklos Szeredi 已提交
355 356 357 358 359 360 361 362
		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
363
		default:
364
			pr_err("overlayfs: unrecognized mount option \"%s\" or missing value\n", p);
M
Miklos Szeredi 已提交
365 366 367
			return -EINVAL;
		}
	}
H
hujianyang 已提交
368 369 370 371 372 373 374 375 376

	/* 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
377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393
	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
394
	inode_lock_nested(dir, I_MUTEX_PARENT);
M
Miklos Szeredi 已提交
395 396 397 398 399
retry:
	work = lookup_one_len(OVL_WORKDIR_NAME, dentry,
			      strlen(OVL_WORKDIR_NAME));

	if (!IS_ERR(work)) {
400 401
		struct iattr attr = {
			.ia_valid = ATTR_MODE,
A
Al Viro 已提交
402
			.ia_mode = S_IFDIR | 0,
403
		};
M
Miklos Szeredi 已提交
404 405 406 407 408 409 410

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

			retried = true;
M
Miklos Szeredi 已提交
411
			ovl_workdir_cleanup(dir, mnt, work, 0);
M
Miklos Szeredi 已提交
412 413 414 415
			dput(work);
			goto retry;
		}

A
Al Viro 已提交
416 417 418
		err = ovl_create_real(dir, work,
				      &(struct cattr){.mode = S_IFDIR | 0},
				      NULL, true);
M
Miklos Szeredi 已提交
419 420
		if (err)
			goto out_dput;
421

422 423 424 425 426 427 428 429 430 431 432 433 434
		/*
		 * 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.
		 */
435
		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_DEFAULT);
M
Miklos Szeredi 已提交
436
		if (err && err != -ENODATA && err != -EOPNOTSUPP)
437 438 439
			goto out_dput;

		err = vfs_removexattr(work, XATTR_NAME_POSIX_ACL_ACCESS);
M
Miklos Szeredi 已提交
440
		if (err && err != -ENODATA && err != -EOPNOTSUPP)
441 442 443 444 445 446 447 448
			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
449 450
	}
out_unlock:
A
Al Viro 已提交
451
	inode_unlock(dir);
M
Miklos Szeredi 已提交
452 453 454 455 456 457 458 459 460 461
	mnt_drop_write(mnt);

	return work;

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

M
Miklos Szeredi 已提交
462 463 464 465 466 467 468 469 470 471 472 473 474
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
475 476
static int ovl_mount_dir_noesc(const char *name, struct path *path)
{
477
	int err = -EINVAL;
M
Miklos Szeredi 已提交
478

479 480 481 482
	if (!*name) {
		pr_err("overlayfs: empty lowerdir\n");
		goto out;
	}
M
Miklos Szeredi 已提交
483 484 485 486 487 488
	err = kern_path(name, LOOKUP_FOLLOW, path);
	if (err) {
		pr_err("overlayfs: failed to resolve '%s': %i\n", name, err);
		goto out;
	}
	err = -EINVAL;
489
	if (ovl_dentry_weird(path->dentry)) {
M
Miklos Szeredi 已提交
490 491 492
		pr_err("overlayfs: filesystem on '%s' not supported\n", name);
		goto out_put;
	}
M
Miklos Szeredi 已提交
493
	if (!d_is_dir(path->dentry)) {
M
Miklos Szeredi 已提交
494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512
		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);
513 514 515 516 517 518 519 520

		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
521 522 523 524 525
		kfree(tmp);
	}
	return err;
}

M
Miklos Szeredi 已提交
526 527
static int ovl_check_namelen(struct path *path, struct ovl_fs *ofs,
			     const char *name)
M
Miklos Szeredi 已提交
528 529
{
	struct kstatfs statfs;
M
Miklos Szeredi 已提交
530 531 532 533 534 535 536 537 538 539 540 541 542 543
	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
544

545
	err = ovl_mount_dir_noesc(name, path);
M
Miklos Szeredi 已提交
546 547 548
	if (err)
		goto out;

M
Miklos Szeredi 已提交
549 550
	err = ovl_check_namelen(path, ofs, name);
	if (err)
M
Miklos Szeredi 已提交
551
		goto out_put;
M
Miklos Szeredi 已提交
552

M
Miklos Szeredi 已提交
553 554
	*stack_depth = max(*stack_depth, path->mnt->mnt_sb->s_stack_depth);

555 556 557
	if (ovl_dentry_remote(path->dentry))
		*remote = true;

M
Miklos Szeredi 已提交
558 559 560 561 562 563 564 565
	return 0;

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

M
Miklos Szeredi 已提交
566 567 568 569 570 571 572 573 574 575 576 577
/* 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;
}

578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597
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;
}

598 599 600 601 602 603 604 605
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);
}

606 607 608 609 610
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
611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637
{
	struct dentry *workdir = ovl_workdir(dentry);
	struct inode *realinode = ovl_inode_real(inode, NULL);
	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);

638 639 640 641 642 643 644 645 646 647 648 649 650 651 652
	/*
	 * 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;
	}

653 654 655 656 657
	err = ovl_xattr_set(dentry, handler->name, value, size, flags);
	if (!err)
		ovl_copyattr(ovl_inode_real(inode, NULL), inode);

	return err;
M
Miklos Szeredi 已提交
658 659 660 661 662 663

out_acl_release:
	posix_acl_release(acl);
	return err;
}

664 665 666 667
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
668
	return -EOPNOTSUPP;
669 670
}

M
Miklos Szeredi 已提交
671 672 673 674 675
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
676
	return -EOPNOTSUPP;
M
Miklos Szeredi 已提交
677 678
}

679 680 681 682 683 684 685
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);
}

686 687 688 689 690 691 692 693
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);
}

694 695
static const struct xattr_handler __maybe_unused
ovl_posix_acl_access_xattr_handler = {
M
Miklos Szeredi 已提交
696 697
	.name = XATTR_NAME_POSIX_ACL_ACCESS,
	.flags = ACL_TYPE_ACCESS,
698
	.get = ovl_posix_acl_xattr_get,
M
Miklos Szeredi 已提交
699 700 701
	.set = ovl_posix_acl_xattr_set,
};

702 703
static const struct xattr_handler __maybe_unused
ovl_posix_acl_default_xattr_handler = {
M
Miklos Szeredi 已提交
704 705
	.name = XATTR_NAME_POSIX_ACL_DEFAULT,
	.flags = ACL_TYPE_DEFAULT,
706
	.get = ovl_posix_acl_xattr_get,
M
Miklos Szeredi 已提交
707 708 709 710 711
	.set = ovl_posix_acl_xattr_set,
};

static const struct xattr_handler ovl_own_xattr_handler = {
	.prefix	= OVL_XATTR_PREFIX,
712
	.get = ovl_own_xattr_get,
M
Miklos Szeredi 已提交
713 714 715 716 717
	.set = ovl_own_xattr_set,
};

static const struct xattr_handler ovl_other_xattr_handler = {
	.prefix	= "", /* catch all */
718
	.get = ovl_other_xattr_get,
M
Miklos Szeredi 已提交
719 720 721 722
	.set = ovl_other_xattr_set,
};

static const struct xattr_handler *ovl_xattr_handlers[] = {
723
#ifdef CONFIG_FS_POSIX_ACL
M
Miklos Szeredi 已提交
724 725
	&ovl_posix_acl_access_xattr_handler,
	&ovl_posix_acl_default_xattr_handler,
726
#endif
M
Miklos Szeredi 已提交
727 728 729 730 731
	&ovl_own_xattr_handler,
	&ovl_other_xattr_handler,
	NULL
};

M
Miklos Szeredi 已提交
732 733
static int ovl_fill_super(struct super_block *sb, void *data, int silent)
{
K
Kees Cook 已提交
734 735
	struct path upperpath = { };
	struct path workpath = { };
M
Miklos Szeredi 已提交
736
	struct dentry *root_dentry;
737
	struct inode *realinode;
M
Miklos Szeredi 已提交
738 739
	struct ovl_entry *oe;
	struct ovl_fs *ufs;
740 741 742 743 744
	struct path *stack = NULL;
	char *lowertmp;
	char *lower;
	unsigned int numlower;
	unsigned int stacklen = 0;
745
	unsigned int i;
746
	bool remote = false;
747
	struct cred *cred;
M
Miklos Szeredi 已提交
748 749
	int err;

E
Erez Zadok 已提交
750 751 752
	err = -ENOMEM;
	ufs = kzalloc(sizeof(struct ovl_fs), GFP_KERNEL);
	if (!ufs)
M
Miklos Szeredi 已提交
753 754
		goto out;

755
	init_waitqueue_head(&ufs->copyup_wq);
756
	ufs->config.redirect_dir = ovl_redirect_dir_def;
E
Erez Zadok 已提交
757 758 759 760
	err = ovl_parse_opt((char *) data, &ufs->config);
	if (err)
		goto out_free_config;

M
Miklos Szeredi 已提交
761
	err = -EINVAL;
M
Miklos Szeredi 已提交
762
	if (!ufs->config.lowerdir) {
763 764
		if (!silent)
			pr_err("overlayfs: missing 'lowerdir'\n");
M
Miklos Szeredi 已提交
765 766 767
		goto out_free_config;
	}

M
Miklos Szeredi 已提交
768
	sb->s_stack_depth = 0;
769
	sb->s_maxbytes = MAX_LFS_FILESIZE;
M
Miklos Szeredi 已提交
770 771 772 773 774
	if (ufs->config.upperdir) {
		if (!ufs->config.workdir) {
			pr_err("overlayfs: missing 'workdir'\n");
			goto out_free_config;
		}
M
Miklos Szeredi 已提交
775

M
Miklos Szeredi 已提交
776 777 778
		err = ovl_mount_dir(ufs->config.upperdir, &upperpath);
		if (err)
			goto out_free_config;
M
Miklos Szeredi 已提交
779

H
hujianyang 已提交
780 781 782 783 784 785 786
		/* 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
787 788 789 790
		err = ovl_check_namelen(&upperpath, ufs, ufs->config.upperdir);
		if (err)
			goto out_put_upperpath;

M
Miklos Szeredi 已提交
791 792 793 794
		err = ovl_mount_dir(ufs->config.workdir, &workpath);
		if (err)
			goto out_put_upperpath;

795
		err = -EINVAL;
M
Miklos Szeredi 已提交
796 797 798 799 800 801 802 803 804
		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
805
	}
806 807 808
	err = -ENOMEM;
	lowertmp = kstrdup(ufs->config.lowerdir, GFP_KERNEL);
	if (!lowertmp)
M
Miklos Szeredi 已提交
809
		goto out_put_workpath;
810

811 812
	err = -EINVAL;
	stacklen = ovl_split_lowerdirs(lowertmp);
813
	if (stacklen > OVL_MAX_STACK) {
814
		pr_err("overlayfs: too many lower directories, limit is %d\n",
815
		       OVL_MAX_STACK);
816
		goto out_free_lowertmp;
817 818 819 820
	} else if (!ufs->config.upperdir && stacklen == 1) {
		pr_err("overlayfs: at least 2 lowerdir are needed while upperdir nonexistent\n");
		goto out_free_lowertmp;
	}
821

822
	err = -ENOMEM;
823 824 825 826
	stack = kcalloc(stacklen, sizeof(struct path), GFP_KERNEL);
	if (!stack)
		goto out_free_lowertmp;

827
	err = -EINVAL;
828 829
	lower = lowertmp;
	for (numlower = 0; numlower < stacklen; numlower++) {
M
Miklos Szeredi 已提交
830 831
		err = ovl_lower_dir(lower, &stack[numlower], ufs,
				    &sb->s_stack_depth, &remote);
832 833 834 835 836 837
		if (err)
			goto out_put_lowerpath;

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

838
	err = -EINVAL;
M
Miklos Szeredi 已提交
839
	sb->s_stack_depth++;
840 841
	if (sb->s_stack_depth > FILESYSTEM_MAX_STACK_DEPTH) {
		pr_err("overlayfs: maximum fs stacking depth exceeded\n");
842
		goto out_put_lowerpath;
843 844
	}

M
Miklos Szeredi 已提交
845 846 847 848 849 850 851
	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
852 853 854 855
		/* 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;
856

M
Miklos Szeredi 已提交
857 858 859
		ufs->workdir = ovl_workdir_create(ufs->upper_mnt, workpath.dentry);
		err = PTR_ERR(ufs->workdir);
		if (IS_ERR(ufs->workdir)) {
860 861 862 863
			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
864
		}
865 866 867 868

		/*
		 * Upper should support d_type, else whiteouts are visible.
		 * Given workdir and upper are on same fs, we can do
869 870
		 * iterate_dir() on workdir. This check requires successful
		 * creation of workdir in previous step.
871
		 */
872
		if (ufs->workdir) {
873 874
			struct dentry *temp;

875 876 877
			err = ovl_check_d_type_supported(&workpath);
			if (err < 0)
				goto out_put_workdir;
878

879 880 881 882 883 884 885
			/*
			 * 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");
886 887 888 889 890 891 892 893

			/* 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");
894 895 896 897 898 899 900 901 902 903 904 905 906

			/*
			 * 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);
			}
907
		}
M
Miklos Szeredi 已提交
908 909
	}

910
	err = -ENOMEM;
911
	ufs->lower_mnt = kcalloc(numlower, sizeof(struct vfsmount *), GFP_KERNEL);
912
	if (ufs->lower_mnt == NULL)
913
		goto out_put_workdir;
914 915
	for (i = 0; i < numlower; i++) {
		struct vfsmount *mnt = clone_private_mount(&stack[i]);
916

917
		err = PTR_ERR(mnt);
918 919 920 921 922 923 924 925
		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
926
		mnt->mnt_flags |= MNT_READONLY | MNT_NOATIME;
927

928 929
		ufs->lower_mnt[ufs->numlower] = mnt;
		ufs->numlower++;
930 931 932 933 934 935

		/* 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;
936
	}
M
Miklos Szeredi 已提交
937

H
hujianyang 已提交
938 939
	/* If the upper fs is nonexistent, we mark overlayfs r/o too */
	if (!ufs->upper_mnt)
M
Miklos Szeredi 已提交
940
		sb->s_flags |= MS_RDONLY;
941 942
	else if (ufs->upper_mnt->mnt_sb != ufs->same_sb)
		ufs->same_sb = NULL;
M
Miklos Szeredi 已提交
943

944 945 946 947
	if (remote)
		sb->s_d_op = &ovl_reval_dentry_operations;
	else
		sb->s_d_op = &ovl_dentry_operations;
M
Miklos Szeredi 已提交
948

949 950
	ufs->creator_cred = cred = prepare_creds();
	if (!cred)
951 952
		goto out_put_lower_mnt;

953 954 955
	/* Never override disk quota limits or use reserved space */
	cap_lower(cred->cap_effective, CAP_SYS_RESOURCE);

M
Miklos Szeredi 已提交
956
	err = -ENOMEM;
957
	oe = ovl_alloc_entry(numlower);
958
	if (!oe)
959
		goto out_put_cred;
M
Miklos Szeredi 已提交
960

961 962 963 964 965 966
	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;

967
	root_dentry = d_make_root(ovl_new_inode(sb, S_IFDIR, 0));
M
Miklos Szeredi 已提交
968
	if (!root_dentry)
969
		goto out_free_oe;
M
Miklos Szeredi 已提交
970 971

	mntput(upperpath.mnt);
972 973
	for (i = 0; i < numlower; i++)
		mntput(stack[i].mnt);
M
Miklos Szeredi 已提交
974
	path_put(&workpath);
975
	kfree(lowertmp);
M
Miklos Szeredi 已提交
976 977

	oe->__upperdentry = upperpath.dentry;
978 979 980 981
	for (i = 0; i < numlower; i++) {
		oe->lowerstack[i].dentry = stack[i].dentry;
		oe->lowerstack[i].mnt = ufs->lower_mnt[i];
	}
982
	kfree(stack);
M
Miklos Szeredi 已提交
983 984 985

	root_dentry->d_fsdata = oe;

986 987 988
	realinode = d_inode(ovl_dentry_real(root_dentry));
	ovl_inode_init(d_inode(root_dentry), realinode, !!upperpath.dentry);
	ovl_copyattr(realinode, d_inode(root_dentry));
M
Miklos Szeredi 已提交
989

M
Miklos Szeredi 已提交
990 991 992 993
	sb->s_root = root_dentry;

	return 0;

994 995
out_free_oe:
	kfree(oe);
996 997
out_put_cred:
	put_cred(ufs->creator_cred);
M
Miklos Szeredi 已提交
998
out_put_lower_mnt:
999 1000 1001
	for (i = 0; i < ufs->numlower; i++)
		mntput(ufs->lower_mnt[i]);
	kfree(ufs->lower_mnt);
1002 1003
out_put_workdir:
	dput(ufs->workdir);
M
Miklos Szeredi 已提交
1004 1005
	mntput(ufs->upper_mnt);
out_put_lowerpath:
1006 1007 1008 1009 1010
	for (i = 0; i < numlower; i++)
		path_put(&stack[i]);
	kfree(stack);
out_free_lowertmp:
	kfree(lowertmp);
1011 1012
out_put_workpath:
	path_put(&workpath);
M
Miklos Szeredi 已提交
1013 1014 1015
out_put_upperpath:
	path_put(&upperpath);
out_free_config:
E
Erez Zadok 已提交
1016 1017 1018 1019
	kfree(ufs->config.lowerdir);
	kfree(ufs->config.upperdir);
	kfree(ufs->config.workdir);
	kfree(ufs);
M
Miklos Szeredi 已提交
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
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,
1032
	.name		= "overlay",
M
Miklos Szeredi 已提交
1033 1034 1035
	.mount		= ovl_mount,
	.kill_sb	= kill_anon_super,
};
1036
MODULE_ALIAS_FS("overlay");
M
Miklos Szeredi 已提交
1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049

static int __init ovl_init(void)
{
	return register_filesystem(&ovl_fs_type);
}

static void __exit ovl_exit(void)
{
	unregister_filesystem(&ovl_fs_type);
}

module_init(ovl_init);
module_exit(ovl_exit);