namei.c 21.5 KB
Newer Older
M
Miklos Szeredi 已提交
1 2 3 4 5 6 7 8 9 10
/*
 * Copyright (C) 2011 Novell Inc.
 * Copyright (C) 2016 Red Hat, 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.
 */

#include <linux/fs.h>
11
#include <linux/cred.h>
M
Miklos Szeredi 已提交
12 13
#include <linux/namei.h>
#include <linux/xattr.h>
M
Miklos Szeredi 已提交
14
#include <linux/ratelimit.h>
15 16
#include <linux/mount.h>
#include <linux/exportfs.h>
M
Miklos Szeredi 已提交
17 18
#include "overlayfs.h"

19 20 21 22 23 24
struct ovl_lookup_data {
	struct qstr name;
	bool is_dir;
	bool opaque;
	bool stop;
	bool last;
M
Miklos Szeredi 已提交
25
	char *redirect;
26
};
M
Miklos Szeredi 已提交
27

M
Miklos Szeredi 已提交
28 29 30 31 32 33 34 35 36 37 38 39
static int ovl_check_redirect(struct dentry *dentry, struct ovl_lookup_data *d,
			      size_t prelen, const char *post)
{
	int res;
	char *s, *next, *buf = NULL;

	res = vfs_getxattr(dentry, OVL_XATTR_REDIRECT, NULL, 0);
	if (res < 0) {
		if (res == -ENODATA || res == -EOPNOTSUPP)
			return 0;
		goto fail;
	}
40
	buf = kzalloc(prelen + res + strlen(post) + 1, GFP_KERNEL);
M
Miklos Szeredi 已提交
41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
	if (!buf)
		return -ENOMEM;

	if (res == 0)
		goto invalid;

	res = vfs_getxattr(dentry, OVL_XATTR_REDIRECT, buf, res);
	if (res < 0)
		goto fail;
	if (res == 0)
		goto invalid;
	if (buf[0] == '/') {
		for (s = buf; *s++ == '/'; s = next) {
			next = strchrnul(s, '/');
			if (s == next)
				goto invalid;
		}
	} else {
		if (strchr(buf, '/') != NULL)
			goto invalid;

		memmove(buf + prelen, buf, res);
		memcpy(buf, d->name.name, prelen);
	}

	strcat(buf, post);
	kfree(d->redirect);
	d->redirect = buf;
	d->name.name = d->redirect;
	d->name.len = strlen(d->redirect);

	return 0;

err_free:
	kfree(buf);
	return 0;
fail:
	pr_warn_ratelimited("overlayfs: failed to get redirect (%i)\n", res);
	goto err_free;
invalid:
	pr_warn_ratelimited("overlayfs: invalid redirect (%s)\n", buf);
	goto err_free;
}

85 86 87 88 89
static int ovl_acceptable(void *ctx, struct dentry *dentry)
{
	return 1;
}

90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
/*
 * Check validity of an overlay file handle buffer.
 *
 * Return 0 for a valid file handle.
 * Return -ENODATA for "origin unknown".
 * Return <0 for an invalid file handle.
 */
static int ovl_check_fh_len(struct ovl_fh *fh, int fh_len)
{
	if (fh_len < sizeof(struct ovl_fh) || fh_len < fh->len)
		return -EINVAL;

	if (fh->magic != OVL_FH_MAGIC)
		return -EINVAL;

	/* Treat larger version and unknown flags as "origin unknown" */
	if (fh->version > OVL_FH_VERSION || fh->flags & ~OVL_FH_FLAG_ALL)
		return -ENODATA;

	/* Treat endianness mismatch as "origin unknown" */
	if (!(fh->flags & OVL_FH_FLAG_ANY_ENDIAN) &&
	    (fh->flags & OVL_FH_FLAG_BIG_ENDIAN) != OVL_FH_FLAG_CPU_ENDIAN)
		return -ENODATA;

	return 0;
}

117
static struct ovl_fh *ovl_get_fh(struct dentry *dentry, const char *name)
118
{
119
	int res, err;
120 121
	struct ovl_fh *fh = NULL;

122
	res = vfs_getxattr(dentry, name, NULL, 0);
123 124 125 126 127 128 129 130 131
	if (res < 0) {
		if (res == -ENODATA || res == -EOPNOTSUPP)
			return NULL;
		goto fail;
	}
	/* Zero size value means "copied up but origin unknown" */
	if (res == 0)
		return NULL;

132
	fh = kzalloc(res, GFP_KERNEL);
133 134 135
	if (!fh)
		return ERR_PTR(-ENOMEM);

136
	res = vfs_getxattr(dentry, name, fh, res);
137 138 139
	if (res < 0)
		goto fail;

140 141 142 143
	err = ovl_check_fh_len(fh, res);
	if (err < 0) {
		if (err == -ENODATA)
			goto out;
144
		goto invalid;
145
	}
146

147 148 149 150 151 152 153 154 155 156 157 158 159 160
	return fh;

out:
	kfree(fh);
	return NULL;

fail:
	pr_warn_ratelimited("overlayfs: failed to get origin (%i)\n", res);
	goto out;
invalid:
	pr_warn_ratelimited("overlayfs: invalid origin (%*phN)\n", res, fh);
	goto out;
}

161
static struct dentry *ovl_decode_fh(struct ovl_fh *fh, struct vfsmount *mnt)
162
{
163
	struct dentry *origin;
164 165
	int bytes;

166 167 168 169
	/*
	 * Make sure that the stored uuid matches the uuid of the lower
	 * layer where file handle will be decoded.
	 */
170
	if (!uuid_equal(&fh->uuid, &mnt->mnt_sb->s_uuid))
171
		return NULL;
172

173
	bytes = (fh->len - offsetof(struct ovl_fh, fid));
174 175 176 177 178 179 180
	origin = exportfs_decode_fh(mnt, (struct fid *)fh->fid,
				    bytes >> 2, (int)fh->type,
				    ovl_acceptable, NULL);
	if (IS_ERR(origin)) {
		/* Treat stale file handle as "origin unknown" */
		if (origin == ERR_PTR(-ESTALE))
			origin = NULL;
181
		return origin;
182 183
	}

184 185 186 187
	if (ovl_dentry_weird(origin)) {
		dput(origin);
		return NULL;
	}
188 189 190 191

	return origin;
}

192 193 194 195 196
static bool ovl_is_opaquedir(struct dentry *dentry)
{
	return ovl_check_dir_xattr(dentry, OVL_XATTR_OPAQUE);
}

197 198
static int ovl_lookup_single(struct dentry *base, struct ovl_lookup_data *d,
			     const char *name, unsigned int namelen,
M
Miklos Szeredi 已提交
199
			     size_t prelen, const char *post,
200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235
			     struct dentry **ret)
{
	struct dentry *this;
	int err;

	this = lookup_one_len_unlocked(name, base, namelen);
	if (IS_ERR(this)) {
		err = PTR_ERR(this);
		this = NULL;
		if (err == -ENOENT || err == -ENAMETOOLONG)
			goto out;
		goto out_err;
	}
	if (!this->d_inode)
		goto put_and_out;

	if (ovl_dentry_weird(this)) {
		/* Don't support traversing automounts and other weirdness */
		err = -EREMOTE;
		goto out_err;
	}
	if (ovl_is_whiteout(this)) {
		d->stop = d->opaque = true;
		goto put_and_out;
	}
	if (!d_can_lookup(this)) {
		d->stop = true;
		if (d->is_dir)
			goto put_and_out;
		goto out;
	}
	d->is_dir = true;
	if (!d->last && ovl_is_opaquedir(this)) {
		d->stop = d->opaque = true;
		goto out;
	}
M
Miklos Szeredi 已提交
236 237 238
	err = ovl_check_redirect(this, d, prelen, post);
	if (err)
		goto out_err;
239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255
out:
	*ret = this;
	return 0;

put_and_out:
	dput(this);
	this = NULL;
	goto out;

out_err:
	dput(this);
	return err;
}

static int ovl_lookup_layer(struct dentry *base, struct ovl_lookup_data *d,
			    struct dentry **ret)
{
256 257
	/* Counting down from the end, since the prefix can change */
	size_t rem = d->name.len - 1;
M
Miklos Szeredi 已提交
258 259 260
	struct dentry *dentry = NULL;
	int err;

261
	if (d->name.name[0] != '/')
M
Miklos Szeredi 已提交
262 263 264
		return ovl_lookup_single(base, d, d->name.name, d->name.len,
					 0, "", ret);

265 266
	while (!IS_ERR_OR_NULL(base) && d_can_lookup(base)) {
		const char *s = d->name.name + d->name.len - rem;
M
Miklos Szeredi 已提交
267
		const char *next = strchrnul(s, '/');
268 269
		size_t thislen = next - s;
		bool end = !next[0];
M
Miklos Szeredi 已提交
270

271 272
		/* Verify we did not go off the rails */
		if (WARN_ON(s[-1] != '/'))
M
Miklos Szeredi 已提交
273 274
			return -EIO;

275 276
		err = ovl_lookup_single(base, d, s, thislen,
					d->name.len - rem, next, &base);
M
Miklos Szeredi 已提交
277 278 279 280
		dput(dentry);
		if (err)
			return err;
		dentry = base;
281 282 283 284 285 286 287
		if (end)
			break;

		rem -= thislen + 1;

		if (WARN_ON(rem >= d->name.len))
			return -EIO;
M
Miklos Szeredi 已提交
288 289 290
	}
	*ret = dentry;
	return 0;
291 292
}

293

294 295
static int ovl_check_origin_fh(struct ovl_fs *ofs, struct ovl_fh *fh,
			       struct dentry *upperdentry,
296
			       struct ovl_path **stackp)
297
{
298 299
	struct dentry *origin = NULL;
	int i;
300

301 302
	for (i = 0; i < ofs->numlower; i++) {
		origin = ovl_decode_fh(fh, ofs->lower_layers[i].mnt);
303 304 305 306 307
		if (origin)
			break;
	}

	if (!origin)
308 309 310 311 312 313 314
		return -ESTALE;
	else if (IS_ERR(origin))
		return PTR_ERR(origin);

	if (!ovl_is_whiteout(upperdentry) &&
	    ((d_inode(origin)->i_mode ^ d_inode(upperdentry)->i_mode) & S_IFMT))
		goto invalid;
315

316
	if (!*stackp)
317
		*stackp = kmalloc(sizeof(struct ovl_path), GFP_KERNEL);
318 319 320 321
	if (!*stackp) {
		dput(origin);
		return -ENOMEM;
	}
322 323 324 325
	**stackp = (struct ovl_path){
		.dentry = origin,
		.layer = &ofs->lower_layers[i]
	};
326 327

	return 0;
328 329 330 331 332 333 334 335 336

invalid:
	pr_warn_ratelimited("overlayfs: invalid origin (%pd2, ftype=%x, origin ftype=%x).\n",
			    upperdentry, d_inode(upperdentry)->i_mode & S_IFMT,
			    d_inode(origin)->i_mode & S_IFMT);
	dput(origin);
	return -EIO;
}

337
static int ovl_check_origin(struct ovl_fs *ofs, struct dentry *upperdentry,
338 339
			    struct ovl_path **stackp, unsigned int *ctrp)
{
340
	struct ovl_fh *fh = ovl_get_fh(upperdentry, OVL_XATTR_ORIGIN);
341 342 343 344 345
	int err;

	if (IS_ERR_OR_NULL(fh))
		return PTR_ERR(fh);

346
	err = ovl_check_origin_fh(ofs, fh, upperdentry, stackp);
347 348 349 350 351 352 353 354 355 356 357 358 359
	kfree(fh);

	if (err) {
		if (err == -ESTALE)
			return 0;
		return err;
	}

	if (WARN_ON(*ctrp))
		return -EIO;

	*ctrp = 1;
	return 0;
360 361
}

362
/*
363
 * Verify that @fh matches the file handle stored in xattr @name.
364 365
 * Return 0 on match, -ESTALE on mismatch, < 0 on error.
 */
366 367
static int ovl_verify_fh(struct dentry *dentry, const char *name,
			 const struct ovl_fh *fh)
368
{
369
	struct ovl_fh *ofh = ovl_get_fh(dentry, name);
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385
	int err = 0;

	if (!ofh)
		return -ENODATA;

	if (IS_ERR(ofh))
		return PTR_ERR(ofh);

	if (fh->len != ofh->len || memcmp(fh, ofh, fh->len))
		err = -ESTALE;

	kfree(ofh);
	return err;
}

/*
386
 * Verify that @real dentry matches the file handle stored in xattr @name.
387
 *
388 389
 * If @set is true and there is no stored file handle, encode @real and store
 * file handle in xattr @name.
390
 *
391
 * Return 0 on match, -ESTALE on mismatch, -ENODATA on no xattr, < 0 on error.
392
 */
393 394
int ovl_verify_set_fh(struct dentry *dentry, const char *name,
		      struct dentry *real, bool is_upper, bool set)
395 396 397 398 399
{
	struct inode *inode;
	struct ovl_fh *fh;
	int err;

400
	fh = ovl_encode_fh(real, is_upper);
401 402 403 404
	err = PTR_ERR(fh);
	if (IS_ERR(fh))
		goto fail;

405
	err = ovl_verify_fh(dentry, name, fh);
406
	if (set && err == -ENODATA)
407
		err = ovl_do_setxattr(dentry, name, fh, fh->len, 0);
408 409 410 411 412 413 414 415
	if (err)
		goto fail;

out:
	kfree(fh);
	return err;

fail:
416 417 418 419
	inode = d_inode(real);
	pr_warn_ratelimited("overlayfs: failed to verify %s (%pd2, ino=%lu, err=%i)\n",
			    is_upper ? "upper" : "origin", real,
			    inode ? inode->i_ino : 0, err);
420 421 422
	goto out;
}

423 424 425 426 427
/*
 * Verify that an index entry name matches the origin file handle stored in
 * OVL_XATTR_ORIGIN and that origin file handle can be decoded to lower path.
 * Return 0 on match, -ESTALE on mismatch or stale origin, < 0 on error.
 */
428
int ovl_verify_index(struct ovl_fs *ofs, struct dentry *index)
429 430 431
{
	struct ovl_fh *fh = NULL;
	size_t len;
432 433
	struct ovl_path origin = { };
	struct ovl_path *stack = &origin;
434 435 436 437 438
	int err;

	if (!d_inode(index))
		return 0;

439 440 441
	/*
	 * Directory index entries are going to be used for looking up
	 * redirected upper dirs by lower dir fh when decoding an overlay
442 443 444 445
	 * file handle of a merge dir.  We don't know the verification rules
	 * for directory index entries, because they have not been implemented
	 * yet, so return EINVAL if those entries are found to abort the mount
	 * and to avoid corrupting an index that was created by a newer kernel.
446
	 */
447
	err = -EINVAL;
448
	if (d_is_dir(index))
449 450 451 452 453 454 455
		goto fail;

	if (index->d_name.len < sizeof(struct ovl_fh)*2)
		goto fail;

	err = -ENOMEM;
	len = index->d_name.len / 2;
456
	fh = kzalloc(len, GFP_KERNEL);
457 458 459 460
	if (!fh)
		goto fail;

	err = -EINVAL;
461 462 463 464 465
	if (hex2bin((u8 *)fh, index->d_name.name, len))
		goto fail;

	err = ovl_check_fh_len(fh, len);
	if (err)
466 467
		goto fail;

468 469 470 471 472 473 474 475
	/*
	 * Whiteout index entries are used as an indication that an exported
	 * overlay file handle should be treated as stale (i.e. after unlink
	 * of the overlay inode). These entries contain no origin xattr.
	 */
	if (ovl_is_whiteout(index))
		goto out;

476
	err = ovl_verify_fh(index, OVL_XATTR_ORIGIN, fh);
477 478 479
	if (err)
		goto fail;

480
	err = ovl_check_origin_fh(ofs, fh, index, &stack);
481 482 483
	if (err)
		goto fail;

484 485
	/* Check if index is orphan and don't warn before cleaning it */
	if (d_inode(index)->i_nlink == 1 &&
486
	    ovl_get_nlink(origin.dentry, index, 0) == 0)
487 488
		err = -ENOENT;

489 490 491 492 493 494
	dput(origin.dentry);
out:
	kfree(fh);
	return err;

fail:
495 496
	pr_warn_ratelimited("overlayfs: failed to verify index (%pd2, ftype=%x, err=%i)\n",
			    index, d_inode(index)->i_mode & S_IFMT, err);
497 498 499
	goto out;
}

500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525
/*
 * Lookup in indexdir for the index entry of a lower real inode or a copy up
 * origin inode. The index entry name is the hex representation of the lower
 * inode file handle.
 *
 * If the index dentry in negative, then either no lower aliases have been
 * copied up yet, or aliases have been copied up in older kernels and are
 * not indexed.
 *
 * If the index dentry for a copy up origin inode is positive, but points
 * to an inode different than the upper inode, then either the upper inode
 * has been copied up and not indexed or it was indexed, but since then
 * index dir was cleared. Either way, that index cannot be used to indentify
 * the overlay inode.
 */
int ovl_get_index_name(struct dentry *origin, struct qstr *name)
{
	int err;
	struct ovl_fh *fh;
	char *n, *s;

	fh = ovl_encode_fh(origin, false);
	if (IS_ERR(fh))
		return PTR_ERR(fh);

	err = -ENOMEM;
526
	n = kzalloc(fh->len * 2, GFP_KERNEL);
527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545
	if (n) {
		s  = bin2hex(n, fh, fh->len);
		*name = (struct qstr) QSTR_INIT(n, s - n);
		err = 0;
	}
	kfree(fh);

	return err;

}

static struct dentry *ovl_lookup_index(struct dentry *dentry,
				       struct dentry *upper,
				       struct dentry *origin)
{
	struct ovl_fs *ofs = dentry->d_sb->s_fs_info;
	struct dentry *index;
	struct inode *inode;
	struct qstr name;
546
	bool is_dir = d_is_dir(origin);
547 548 549 550 551 552 553 554
	int err;

	err = ovl_get_index_name(origin, &name);
	if (err)
		return ERR_PTR(err);

	index = lookup_one_len_unlocked(name.name, ofs->indexdir, name.len);
	if (IS_ERR(index)) {
555
		err = PTR_ERR(index);
556 557 558 559
		if (err == -ENOENT) {
			index = NULL;
			goto out;
		}
560 561 562 563 564 565 566
		pr_warn_ratelimited("overlayfs: failed inode index lookup (ino=%lu, key=%*s, err=%i);\n"
				    "overlayfs: mount with '-o index=off' to disable inodes index.\n",
				    d_inode(origin)->i_ino, name.len, name.name,
				    err);
		goto out;
	}

567
	inode = d_inode(index);
568
	if (d_is_negative(index)) {
569
		goto out_dput;
570 571 572 573 574 575 576 577 578 579 580 581
	} else if (ovl_dentry_weird(index) || ovl_is_whiteout(index) ||
		   ((inode->i_mode ^ d_inode(origin)->i_mode) & S_IFMT)) {
		/*
		 * Index should always be of the same file type as origin
		 * except for the case of a whiteout index. A whiteout
		 * index should only exist if all lower aliases have been
		 * unlinked, which means that finding a lower origin on lookup
		 * whose index is a whiteout should be treated as an error.
		 */
		pr_warn_ratelimited("overlayfs: bad index found (index=%pd2, ftype=%x, origin ftype=%x).\n",
				    index, d_inode(index)->i_mode & S_IFMT,
				    d_inode(origin)->i_mode & S_IFMT);
582
		goto fail;
583 584 585 586 587 588
	} else if (is_dir) {
		if (!upper) {
			pr_warn_ratelimited("overlayfs: suspected uncovered redirected dir found (origin=%pd2, index=%pd2).\n",
					    origin, index);
			goto fail;
		}
589

590 591 592 593 594 595 596 597 598 599 600 601
		/* Verify that dir index 'upper' xattr points to upper dir */
		err = ovl_verify_upper(index, upper, false);
		if (err) {
			if (err == -ESTALE) {
				pr_warn_ratelimited("overlayfs: suspected multiply redirected dir found (upper=%pd2, origin=%pd2, index=%pd2).\n",
						    upper, origin, index);
			}
			goto fail;
		}
	} else if (upper && d_inode(upper) != inode) {
		goto out_dput;
	}
602 603 604 605
out:
	kfree(name.name);
	return index;

606 607 608 609 610
out_dput:
	dput(index);
	index = NULL;
	goto out;

611 612 613 614 615 616
fail:
	dput(index);
	index = ERR_PTR(-EIO);
	goto out;
}

M
Miklos Szeredi 已提交
617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632
/*
 * Returns next layer in stack starting from top.
 * Returns -1 if this is the last layer.
 */
int ovl_path_next(int idx, struct dentry *dentry, struct path *path)
{
	struct ovl_entry *oe = dentry->d_fsdata;

	BUG_ON(idx < 0);
	if (idx == 0) {
		ovl_path_upper(dentry, path);
		if (path->dentry)
			return oe->numlower ? 1 : -1;
		idx++;
	}
	BUG_ON(idx > oe->numlower);
633 634
	path->dentry = oe->lowerstack[idx - 1].dentry;
	path->mnt = oe->lowerstack[idx - 1].layer->mnt;
M
Miklos Szeredi 已提交
635 636 637 638

	return (idx < oe->numlower) ? idx + 1 : -1;
}

639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659
/* Fix missing 'origin' xattr */
static int ovl_fix_origin(struct dentry *dentry, struct dentry *lower,
			  struct dentry *upper)
{
	int err;

	if (ovl_check_origin_xattr(upper))
		return 0;

	err = ovl_want_write(dentry);
	if (err)
		return err;

	err = ovl_set_origin(dentry, lower, upper);
	if (!err)
		err = ovl_set_impure(dentry->d_parent, upper->d_parent);

	ovl_drop_write(dentry);
	return err;
}

M
Miklos Szeredi 已提交
660 661 662 663 664
struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry,
			  unsigned int flags)
{
	struct ovl_entry *oe;
	const struct cred *old_cred;
M
Miklos Szeredi 已提交
665
	struct ovl_fs *ofs = dentry->d_sb->s_fs_info;
M
Miklos Szeredi 已提交
666
	struct ovl_entry *poe = dentry->d_parent->d_fsdata;
667
	struct ovl_entry *roe = dentry->d_sb->s_root->d_fsdata;
668
	struct ovl_path *stack = NULL;
M
Miklos Szeredi 已提交
669
	struct dentry *upperdir, *upperdentry = NULL;
670
	struct dentry *origin = NULL;
671
	struct dentry *index = NULL;
M
Miklos Szeredi 已提交
672 673 674
	unsigned int ctr = 0;
	struct inode *inode = NULL;
	bool upperopaque = false;
M
Miklos Szeredi 已提交
675
	char *upperredirect = NULL;
M
Miklos Szeredi 已提交
676 677 678
	struct dentry *this;
	unsigned int i;
	int err;
679 680 681 682 683 684
	struct ovl_lookup_data d = {
		.name = dentry->d_name,
		.is_dir = false,
		.opaque = false,
		.stop = false,
		.last = !poe->numlower,
M
Miklos Szeredi 已提交
685
		.redirect = NULL,
686
	};
M
Miklos Szeredi 已提交
687

M
Miklos Szeredi 已提交
688 689 690
	if (dentry->d_name.len > ofs->namelen)
		return ERR_PTR(-ENAMETOOLONG);

M
Miklos Szeredi 已提交
691
	old_cred = ovl_override_creds(dentry->d_sb);
692
	upperdir = ovl_dentry_upper(dentry->d_parent);
M
Miklos Szeredi 已提交
693
	if (upperdir) {
694 695
		err = ovl_lookup_layer(upperdir, &d, &upperdentry);
		if (err)
M
Miklos Szeredi 已提交
696 697
			goto out;

698 699 700 701
		if (upperdentry && unlikely(ovl_dentry_remote(upperdentry))) {
			dput(upperdentry);
			err = -EREMOTE;
			goto out;
M
Miklos Szeredi 已提交
702
		}
703 704
		if (upperdentry && !d.is_dir) {
			BUG_ON(!d.stop || d.redirect);
705 706 707 708 709 710 711 712 713 714
			/*
			 * Lookup copy up origin by decoding origin file handle.
			 * We may get a disconnected dentry, which is fine,
			 * because we only need to hold the origin inode in
			 * cache and use its inode number.  We may even get a
			 * connected dentry, that is not under any of the lower
			 * layers root.  That is also fine for using it's inode
			 * number - it's the same as if we held a reference
			 * to a dentry in lower layer that was moved under us.
			 */
715
			err = ovl_check_origin(ofs, upperdentry, &stack, &ctr);
716
			if (err)
717
				goto out_put_upper;
718
		}
M
Miklos Szeredi 已提交
719 720

		if (d.redirect) {
721
			err = -ENOMEM;
M
Miklos Szeredi 已提交
722 723 724 725
			upperredirect = kstrdup(d.redirect, GFP_KERNEL);
			if (!upperredirect)
				goto out_put_upper;
			if (d.redirect[0] == '/')
726
				poe = roe;
M
Miklos Szeredi 已提交
727
		}
728
		upperopaque = d.opaque;
M
Miklos Szeredi 已提交
729 730
	}

731
	if (!d.stop && poe->numlower) {
M
Miklos Szeredi 已提交
732
		err = -ENOMEM;
733
		stack = kcalloc(ofs->numlower, sizeof(struct ovl_path),
734
				GFP_KERNEL);
M
Miklos Szeredi 已提交
735 736 737 738
		if (!stack)
			goto out_put_upper;
	}

739
	for (i = 0; !d.stop && i < poe->numlower; i++) {
740
		struct ovl_path lower = poe->lowerstack[i];
M
Miklos Szeredi 已提交
741

742
		d.last = i == poe->numlower - 1;
743
		err = ovl_lookup_layer(lower.dentry, &d, &this);
744
		if (err)
M
Miklos Szeredi 已提交
745
			goto out_put;
M
Miklos Szeredi 已提交
746

M
Miklos Szeredi 已提交
747 748 749
		if (!this)
			continue;

750 751 752 753 754 755 756 757 758 759 760 761
		/*
		 * If no origin fh is stored in upper of a merge dir, store fh
		 * of lower dir and set upper parent "impure".
		 */
		if (upperdentry && !ctr && !ofs->noxattr) {
			err = ovl_fix_origin(dentry, this, upperdentry);
			if (err) {
				dput(this);
				goto out_put;
			}
		}

762 763
		/*
		 * When "verify_lower" feature is enabled, do not merge with a
764 765
		 * lower dir that does not match a stored origin xattr. In any
		 * case, only verified origin is used for index lookup.
766 767 768 769 770 771 772
		 */
		if (upperdentry && !ctr && ovl_verify_lower(dentry->d_sb)) {
			err = ovl_verify_origin(upperdentry, this, false);
			if (err) {
				dput(this);
				break;
			}
773 774 775

			/* Bless lower dir as verified origin */
			origin = this;
776 777
		}

M
Miklos Szeredi 已提交
778
		stack[ctr].dentry = this;
779
		stack[ctr].layer = lower.layer;
M
Miklos Szeredi 已提交
780
		ctr++;
M
Miklos Szeredi 已提交
781 782 783 784

		if (d.stop)
			break;

785 786 787 788 789 790 791 792 793 794 795 796
		/*
		 * Following redirects can have security consequences: it's like
		 * a symlink into the lower layer without the permission checks.
		 * This is only a problem if the upper layer is untrusted (e.g
		 * comes from an USB drive).  This can allow a non-readable file
		 * or directory to become readable.
		 *
		 * Only following redirects when redirects are enabled disables
		 * this attack vector when not necessary.
		 */
		err = -EPERM;
		if (d.redirect && !ofs->config.redirect_follow) {
797 798
			pr_warn_ratelimited("overlayfs: refusing to follow redirect for (%pd2)\n",
					    dentry);
799 800 801
			goto out_put;
		}

802 803
		if (d.redirect && d.redirect[0] == '/' && poe != roe) {
			poe = roe;
M
Miklos Szeredi 已提交
804
			/* Find the current layer on the root dentry */
805
			i = lower.layer->idx - 1;
M
Miklos Szeredi 已提交
806
		}
M
Miklos Szeredi 已提交
807 808
	}

809 810 811 812 813 814 815 816
	/*
	 * Lookup index by lower inode and verify it matches upper inode.
	 * We only trust dir index if we verified that lower dir matches
	 * origin, otherwise dir index entries may be inconsistent and we
	 * ignore them. Always lookup index of non-dir and non-upper.
	 */
	if (ctr && (!upperdentry || !d.is_dir))
		origin = stack[0].dentry;
817

818 819
	if (origin && ovl_indexdir(dentry->d_sb) &&
	    (!d.is_dir || ovl_index_all(dentry->d_sb))) {
820 821 822 823 824 825 826 827
		index = ovl_lookup_index(dentry, upperdentry, origin);
		if (IS_ERR(index)) {
			err = PTR_ERR(index);
			index = NULL;
			goto out_put;
		}
	}

M
Miklos Szeredi 已提交
828 829 830 831 832
	oe = ovl_alloc_entry(ctr);
	err = -ENOMEM;
	if (!oe)
		goto out_put;

M
Miklos Szeredi 已提交
833
	oe->opaque = upperopaque;
834
	memcpy(oe->lowerstack, stack, sizeof(struct ovl_path) * ctr);
M
Miklos Szeredi 已提交
835
	dentry->d_fsdata = oe;
M
Miklos Szeredi 已提交
836

837 838 839
	if (upperdentry)
		ovl_dentry_set_upper_alias(dentry);
	else if (index)
840 841
		upperdentry = dget(index);

M
Miklos Szeredi 已提交
842
	if (upperdentry || ctr) {
843
		inode = ovl_get_inode(dentry, upperdentry, index);
844 845
		err = PTR_ERR(inode);
		if (IS_ERR(inode))
M
Miklos Szeredi 已提交
846
			goto out_free_oe;
M
Miklos Szeredi 已提交
847 848

		OVL_I(inode)->redirect = upperredirect;
849 850
		if (index)
			ovl_set_flag(OVL_INDEX, inode);
M
Miklos Szeredi 已提交
851 852 853
	}

	revert_creds(old_cred);
854
	dput(index);
M
Miklos Szeredi 已提交
855
	kfree(stack);
M
Miklos Szeredi 已提交
856
	kfree(d.redirect);
M
Miklos Szeredi 已提交
857 858 859 860 861
	d_add(dentry, inode);

	return NULL;

out_free_oe:
M
Miklos Szeredi 已提交
862
	dentry->d_fsdata = NULL;
M
Miklos Szeredi 已提交
863 864
	kfree(oe);
out_put:
865
	dput(index);
M
Miklos Szeredi 已提交
866 867 868 869 870
	for (i = 0; i < ctr; i++)
		dput(stack[i].dentry);
	kfree(stack);
out_put_upper:
	dput(upperdentry);
M
Miklos Szeredi 已提交
871
	kfree(upperredirect);
M
Miklos Szeredi 已提交
872
out:
M
Miklos Szeredi 已提交
873
	kfree(d.redirect);
M
Miklos Szeredi 已提交
874 875 876 877 878 879 880 881 882
	revert_creds(old_cred);
	return ERR_PTR(err);
}

bool ovl_lower_positive(struct dentry *dentry)
{
	struct ovl_entry *oe = dentry->d_fsdata;
	struct ovl_entry *poe = dentry->d_parent->d_fsdata;
	const struct qstr *name = &dentry->d_name;
883
	const struct cred *old_cred;
M
Miklos Szeredi 已提交
884 885 886 887 888 889 890 891 892 893 894 895
	unsigned int i;
	bool positive = false;
	bool done = false;

	/*
	 * If dentry is negative, then lower is positive iff this is a
	 * whiteout.
	 */
	if (!dentry->d_inode)
		return oe->opaque;

	/* Negative upper -> positive lower */
896
	if (!ovl_dentry_upper(dentry))
M
Miklos Szeredi 已提交
897 898
		return true;

899
	old_cred = ovl_override_creds(dentry->d_sb);
M
Miklos Szeredi 已提交
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
	/* Positive upper -> have to look up lower to see whether it exists */
	for (i = 0; !done && !positive && i < poe->numlower; i++) {
		struct dentry *this;
		struct dentry *lowerdir = poe->lowerstack[i].dentry;

		this = lookup_one_len_unlocked(name->name, lowerdir,
					       name->len);
		if (IS_ERR(this)) {
			switch (PTR_ERR(this)) {
			case -ENOENT:
			case -ENAMETOOLONG:
				break;

			default:
				/*
				 * Assume something is there, we just couldn't
				 * access it.
				 */
				positive = true;
				break;
			}
		} else {
			if (this->d_inode) {
				positive = !ovl_is_whiteout(this);
				done = true;
			}
			dput(this);
		}
	}
929
	revert_creds(old_cred);
M
Miklos Szeredi 已提交
930 931 932

	return positive;
}