namei.c 22.9 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
static int ovl_acceptable(void *ctx, struct dentry *dentry)
{
87 88 89 90 91 92 93 94 95 96 97 98 99
	/*
	 * A non-dir origin may be disconnected, which is fine, because
	 * we only need it for its unique inode number.
	 */
	if (!d_is_dir(dentry))
		return 1;

	/* Don't decode a deleted empty directory */
	if (d_unhashed(dentry))
		return 0;

	/* Check if directory belongs to the layer we are decoding from */
	return is_subdir(dentry, ((struct vfsmount *)ctx)->mnt_root);
100 101
}

102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
/*
 * 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;
}

129
static struct ovl_fh *ovl_get_fh(struct dentry *dentry, const char *name)
130
{
131
	int res, err;
132 133
	struct ovl_fh *fh = NULL;

134
	res = vfs_getxattr(dentry, name, NULL, 0);
135 136 137 138 139 140 141 142 143
	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;

144
	fh = kzalloc(res, GFP_KERNEL);
145 146 147
	if (!fh)
		return ERR_PTR(-ENOMEM);

148
	res = vfs_getxattr(dentry, name, fh, res);
149 150 151
	if (res < 0)
		goto fail;

152 153 154 155
	err = ovl_check_fh_len(fh, res);
	if (err < 0) {
		if (err == -ENODATA)
			goto out;
156
		goto invalid;
157
	}
158

159 160 161 162 163 164 165 166 167 168 169 170 171 172
	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;
}

173
static struct dentry *ovl_decode_fh(struct ovl_fh *fh, struct vfsmount *mnt)
174
{
175
	struct dentry *real;
176 177
	int bytes;

178 179 180 181
	/*
	 * Make sure that the stored uuid matches the uuid of the lower
	 * layer where file handle will be decoded.
	 */
182
	if (!uuid_equal(&fh->uuid, &mnt->mnt_sb->s_uuid))
183
		return NULL;
184

185
	bytes = (fh->len - offsetof(struct ovl_fh, fid));
186 187 188 189 190 191 192 193 194 195 196 197 198 199
	real = exportfs_decode_fh(mnt, (struct fid *)fh->fid,
				  bytes >> 2, (int)fh->type,
				  ovl_acceptable, mnt);
	if (IS_ERR(real)) {
		/*
		 * Treat stale file handle to lower file as "origin unknown".
		 * upper file handle could become stale when upper file is
		 * unlinked and this information is needed to handle stale
		 * index entries correctly.
		 */
		if (real == ERR_PTR(-ESTALE) &&
		    !(fh->flags & OVL_FH_FLAG_PATH_UPPER))
			real = NULL;
		return real;
200 201
	}

202 203
	if (ovl_dentry_weird(real)) {
		dput(real);
204 205
		return NULL;
	}
206

207
	return real;
208 209
}

210 211 212 213 214
static bool ovl_is_opaquedir(struct dentry *dentry)
{
	return ovl_check_dir_xattr(dentry, OVL_XATTR_OPAQUE);
}

215 216
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
217
			     size_t prelen, const char *post,
218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253
			     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
254 255 256
	err = ovl_check_redirect(this, d, prelen, post);
	if (err)
		goto out_err;
257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273
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)
{
274 275
	/* Counting down from the end, since the prefix can change */
	size_t rem = d->name.len - 1;
M
Miklos Szeredi 已提交
276 277 278
	struct dentry *dentry = NULL;
	int err;

279
	if (d->name.name[0] != '/')
M
Miklos Szeredi 已提交
280 281 282
		return ovl_lookup_single(base, d, d->name.name, d->name.len,
					 0, "", ret);

283 284
	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
285
		const char *next = strchrnul(s, '/');
286 287
		size_t thislen = next - s;
		bool end = !next[0];
M
Miklos Szeredi 已提交
288

289 290
		/* Verify we did not go off the rails */
		if (WARN_ON(s[-1] != '/'))
M
Miklos Szeredi 已提交
291 292
			return -EIO;

293 294
		err = ovl_lookup_single(base, d, s, thislen,
					d->name.len - rem, next, &base);
M
Miklos Szeredi 已提交
295 296 297 298
		dput(dentry);
		if (err)
			return err;
		dentry = base;
299 300 301 302 303 304 305
		if (end)
			break;

		rem -= thislen + 1;

		if (WARN_ON(rem >= d->name.len))
			return -EIO;
M
Miklos Szeredi 已提交
306 307 308
	}
	*ret = dentry;
	return 0;
309 310
}

311

312 313
static int ovl_check_origin_fh(struct ovl_fs *ofs, struct ovl_fh *fh,
			       struct dentry *upperdentry,
314
			       struct ovl_path **stackp)
315
{
316 317
	struct dentry *origin = NULL;
	int i;
318

319 320
	for (i = 0; i < ofs->numlower; i++) {
		origin = ovl_decode_fh(fh, ofs->lower_layers[i].mnt);
321 322 323 324 325
		if (origin)
			break;
	}

	if (!origin)
326 327 328 329 330 331 332
		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;
333

334
	if (!*stackp)
335
		*stackp = kmalloc(sizeof(struct ovl_path), GFP_KERNEL);
336 337 338 339
	if (!*stackp) {
		dput(origin);
		return -ENOMEM;
	}
340 341 342 343
	**stackp = (struct ovl_path){
		.dentry = origin,
		.layer = &ofs->lower_layers[i]
	};
344 345

	return 0;
346 347 348 349 350 351 352 353 354

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;
}

355
static int ovl_check_origin(struct ovl_fs *ofs, struct dentry *upperdentry,
356 357
			    struct ovl_path **stackp, unsigned int *ctrp)
{
358
	struct ovl_fh *fh = ovl_get_fh(upperdentry, OVL_XATTR_ORIGIN);
359 360 361 362 363
	int err;

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

364
	err = ovl_check_origin_fh(ofs, fh, upperdentry, stackp);
365 366 367 368 369 370 371 372 373 374 375 376 377
	kfree(fh);

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

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

	*ctrp = 1;
	return 0;
378 379
}

380
/*
381
 * Verify that @fh matches the file handle stored in xattr @name.
382 383
 * Return 0 on match, -ESTALE on mismatch, < 0 on error.
 */
384 385
static int ovl_verify_fh(struct dentry *dentry, const char *name,
			 const struct ovl_fh *fh)
386
{
387
	struct ovl_fh *ofh = ovl_get_fh(dentry, name);
388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403
	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;
}

/*
404
 * Verify that @real dentry matches the file handle stored in xattr @name.
405
 *
406 407
 * If @set is true and there is no stored file handle, encode @real and store
 * file handle in xattr @name.
408
 *
409
 * Return 0 on match, -ESTALE on mismatch, -ENODATA on no xattr, < 0 on error.
410
 */
411 412
int ovl_verify_set_fh(struct dentry *dentry, const char *name,
		      struct dentry *real, bool is_upper, bool set)
413 414 415 416 417
{
	struct inode *inode;
	struct ovl_fh *fh;
	int err;

418
	fh = ovl_encode_fh(real, is_upper);
419 420 421 422
	err = PTR_ERR(fh);
	if (IS_ERR(fh))
		goto fail;

423
	err = ovl_verify_fh(dentry, name, fh);
424
	if (set && err == -ENODATA)
425
		err = ovl_do_setxattr(dentry, name, fh, fh->len, 0);
426 427 428 429 430 431 432 433
	if (err)
		goto fail;

out:
	kfree(fh);
	return err;

fail:
434 435 436 437
	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);
438 439 440
	goto out;
}

441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469
/* Get upper dentry from index */
static struct dentry *ovl_index_upper(struct ovl_fs *ofs, struct dentry *index)
{
	struct ovl_fh *fh;
	struct dentry *upper;

	if (!d_is_dir(index))
		return dget(index);

	fh = ovl_get_fh(index, OVL_XATTR_UPPER);
	if (IS_ERR_OR_NULL(fh))
		return ERR_CAST(fh);

	upper = ovl_decode_fh(fh, ofs->upper_mnt);
	kfree(fh);

	if (IS_ERR_OR_NULL(upper))
		return upper ?: ERR_PTR(-ESTALE);

	if (!d_is_dir(upper)) {
		pr_warn_ratelimited("overlayfs: invalid index upper (%pd2, upper=%pd2).\n",
				    index, upper);
		dput(upper);
		return ERR_PTR(-EIO);
	}

	return upper;
}

470 471 472 473 474
/*
 * 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.
 */
475
int ovl_verify_index(struct ovl_fs *ofs, struct dentry *index)
476 477 478
{
	struct ovl_fh *fh = NULL;
	size_t len;
479 480
	struct ovl_path origin = { };
	struct ovl_path *stack = &origin;
481
	struct dentry *upper = NULL;
482 483 484 485 486
	int err;

	if (!d_inode(index))
		return 0;

487
	err = -EINVAL;
488 489 490 491 492
	if (index->d_name.len < sizeof(struct ovl_fh)*2)
		goto fail;

	err = -ENOMEM;
	len = index->d_name.len / 2;
493
	fh = kzalloc(len, GFP_KERNEL);
494 495 496 497
	if (!fh)
		goto fail;

	err = -EINVAL;
498 499 500 501 502
	if (hex2bin((u8 *)fh, index->d_name.name, len))
		goto fail;

	err = ovl_check_fh_len(fh, len);
	if (err)
503 504
		goto fail;

505 506 507 508 509 510 511 512
	/*
	 * 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;

513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531
	/*
	 * Verifying directory index entries are not stale is expensive, so
	 * only verify stale dir index if NFS export is enabled.
	 */
	if (d_is_dir(index) && !ofs->config.nfs_export)
		goto out;

	/*
	 * Directory index entries should have 'upper' xattr pointing to the
	 * real upper dir. Non-dir index entries are hardlinks to the upper
	 * real inode. For non-dir index, we can read the copy up origin xattr
	 * directly from the index dentry, but for dir index we first need to
	 * decode the upper directory.
	 */
	upper = ovl_index_upper(ofs, index);
	if (IS_ERR_OR_NULL(upper)) {
		err = PTR_ERR(upper);
		if (!err)
			err = -ESTALE;
532
		goto fail;
533
	}
534

535 536
	err = ovl_verify_fh(upper, OVL_XATTR_ORIGIN, fh);
	dput(upper);
537 538 539
	if (err)
		goto fail;

540 541 542 543 544 545 546 547 548
	/* Check if non-dir index is orphan and don't warn before cleaning it */
	if (!d_is_dir(index) && d_inode(index)->i_nlink == 1) {
		err = ovl_check_origin_fh(ofs, fh, index, &stack);
		if (err)
			goto fail;

		if (ovl_get_nlink(origin.dentry, index, 0) == 0)
			err = -ENOENT;
	}
549

550
out:
551
	dput(origin.dentry);
552 553 554 555
	kfree(fh);
	return err;

fail:
556 557
	pr_warn_ratelimited("overlayfs: failed to verify index (%pd2, ftype=%x, err=%i)\n",
			    index, d_inode(index)->i_mode & S_IFMT, err);
558 559 560
	goto out;
}

561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
/*
 * 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;
587
	n = kzalloc(fh->len * 2, GFP_KERNEL);
588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606
	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;
607
	bool is_dir = d_is_dir(origin);
608 609 610 611 612 613 614 615
	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)) {
616
		err = PTR_ERR(index);
617 618 619 620
		if (err == -ENOENT) {
			index = NULL;
			goto out;
		}
621 622 623 624 625 626 627
		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;
	}

628
	inode = d_inode(index);
629
	if (d_is_negative(index)) {
630
		goto out_dput;
631 632 633 634 635 636 637 638 639 640 641 642
	} 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);
643
		goto fail;
644 645 646 647 648 649
	} else if (is_dir) {
		if (!upper) {
			pr_warn_ratelimited("overlayfs: suspected uncovered redirected dir found (origin=%pd2, index=%pd2).\n",
					    origin, index);
			goto fail;
		}
650

651 652 653 654 655 656 657 658 659 660 661 662
		/* 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;
	}
663 664 665 666
out:
	kfree(name.name);
	return index;

667 668 669 670 671
out_dput:
	dput(index);
	index = NULL;
	goto out;

672 673 674 675 676 677
fail:
	dput(index);
	index = ERR_PTR(-EIO);
	goto out;
}

M
Miklos Szeredi 已提交
678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693
/*
 * 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);
694 695
	path->dentry = oe->lowerstack[idx - 1].dentry;
	path->mnt = oe->lowerstack[idx - 1].layer->mnt;
M
Miklos Szeredi 已提交
696 697 698 699

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

700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720
/* 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
721 722 723 724 725
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
726
	struct ovl_fs *ofs = dentry->d_sb->s_fs_info;
M
Miklos Szeredi 已提交
727
	struct ovl_entry *poe = dentry->d_parent->d_fsdata;
728
	struct ovl_entry *roe = dentry->d_sb->s_root->d_fsdata;
729
	struct ovl_path *stack = NULL;
M
Miklos Szeredi 已提交
730
	struct dentry *upperdir, *upperdentry = NULL;
731
	struct dentry *origin = NULL;
732
	struct dentry *index = NULL;
M
Miklos Szeredi 已提交
733 734 735
	unsigned int ctr = 0;
	struct inode *inode = NULL;
	bool upperopaque = false;
M
Miklos Szeredi 已提交
736
	char *upperredirect = NULL;
M
Miklos Szeredi 已提交
737 738 739
	struct dentry *this;
	unsigned int i;
	int err;
740 741 742 743 744 745
	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
746
		.redirect = NULL,
747
	};
M
Miklos Szeredi 已提交
748

M
Miklos Szeredi 已提交
749 750 751
	if (dentry->d_name.len > ofs->namelen)
		return ERR_PTR(-ENAMETOOLONG);

M
Miklos Szeredi 已提交
752
	old_cred = ovl_override_creds(dentry->d_sb);
753
	upperdir = ovl_dentry_upper(dentry->d_parent);
M
Miklos Szeredi 已提交
754
	if (upperdir) {
755 756
		err = ovl_lookup_layer(upperdir, &d, &upperdentry);
		if (err)
M
Miklos Szeredi 已提交
757 758
			goto out;

759 760 761 762
		if (upperdentry && unlikely(ovl_dentry_remote(upperdentry))) {
			dput(upperdentry);
			err = -EREMOTE;
			goto out;
M
Miklos Szeredi 已提交
763
		}
764 765
		if (upperdentry && !d.is_dir) {
			BUG_ON(!d.stop || d.redirect);
766 767 768 769 770 771 772 773 774 775
			/*
			 * 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.
			 */
776
			err = ovl_check_origin(ofs, upperdentry, &stack, &ctr);
777
			if (err)
778
				goto out_put_upper;
779
		}
M
Miklos Szeredi 已提交
780 781

		if (d.redirect) {
782
			err = -ENOMEM;
M
Miklos Szeredi 已提交
783 784 785 786
			upperredirect = kstrdup(d.redirect, GFP_KERNEL);
			if (!upperredirect)
				goto out_put_upper;
			if (d.redirect[0] == '/')
787
				poe = roe;
M
Miklos Szeredi 已提交
788
		}
789
		upperopaque = d.opaque;
M
Miklos Szeredi 已提交
790 791
	}

792
	if (!d.stop && poe->numlower) {
M
Miklos Szeredi 已提交
793
		err = -ENOMEM;
794
		stack = kcalloc(ofs->numlower, sizeof(struct ovl_path),
795
				GFP_KERNEL);
M
Miklos Szeredi 已提交
796 797 798 799
		if (!stack)
			goto out_put_upper;
	}

800
	for (i = 0; !d.stop && i < poe->numlower; i++) {
801
		struct ovl_path lower = poe->lowerstack[i];
M
Miklos Szeredi 已提交
802

803
		d.last = i == poe->numlower - 1;
804
		err = ovl_lookup_layer(lower.dentry, &d, &this);
805
		if (err)
M
Miklos Szeredi 已提交
806
			goto out_put;
M
Miklos Szeredi 已提交
807

M
Miklos Szeredi 已提交
808 809 810
		if (!this)
			continue;

811 812 813 814 815 816 817 818 819 820 821 822
		/*
		 * 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;
			}
		}

823 824
		/*
		 * When "verify_lower" feature is enabled, do not merge with a
825 826
		 * lower dir that does not match a stored origin xattr. In any
		 * case, only verified origin is used for index lookup.
827 828 829 830 831 832 833
		 */
		if (upperdentry && !ctr && ovl_verify_lower(dentry->d_sb)) {
			err = ovl_verify_origin(upperdentry, this, false);
			if (err) {
				dput(this);
				break;
			}
834 835 836

			/* Bless lower dir as verified origin */
			origin = this;
837 838
		}

M
Miklos Szeredi 已提交
839
		stack[ctr].dentry = this;
840
		stack[ctr].layer = lower.layer;
M
Miklos Szeredi 已提交
841
		ctr++;
M
Miklos Szeredi 已提交
842 843 844 845

		if (d.stop)
			break;

846 847 848 849 850 851 852 853 854 855 856 857
		/*
		 * 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) {
858 859
			pr_warn_ratelimited("overlayfs: refusing to follow redirect for (%pd2)\n",
					    dentry);
860 861 862
			goto out_put;
		}

863 864
		if (d.redirect && d.redirect[0] == '/' && poe != roe) {
			poe = roe;
M
Miklos Szeredi 已提交
865
			/* Find the current layer on the root dentry */
866
			i = lower.layer->idx - 1;
M
Miklos Szeredi 已提交
867
		}
M
Miklos Szeredi 已提交
868 869
	}

870 871 872 873 874 875 876 877
	/*
	 * 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;
878

879 880
	if (origin && ovl_indexdir(dentry->d_sb) &&
	    (!d.is_dir || ovl_index_all(dentry->d_sb))) {
881 882 883 884 885 886 887 888
		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
889 890 891 892 893
	oe = ovl_alloc_entry(ctr);
	err = -ENOMEM;
	if (!oe)
		goto out_put;

M
Miklos Szeredi 已提交
894
	oe->opaque = upperopaque;
895
	memcpy(oe->lowerstack, stack, sizeof(struct ovl_path) * ctr);
M
Miklos Szeredi 已提交
896
	dentry->d_fsdata = oe;
M
Miklos Szeredi 已提交
897

898 899 900
	if (upperdentry)
		ovl_dentry_set_upper_alias(dentry);
	else if (index)
901 902
		upperdentry = dget(index);

M
Miklos Szeredi 已提交
903
	if (upperdentry || ctr) {
904
		inode = ovl_get_inode(dentry, upperdentry, index);
905 906
		err = PTR_ERR(inode);
		if (IS_ERR(inode))
M
Miklos Szeredi 已提交
907
			goto out_free_oe;
M
Miklos Szeredi 已提交
908 909

		OVL_I(inode)->redirect = upperredirect;
910 911
		if (index)
			ovl_set_flag(OVL_INDEX, inode);
M
Miklos Szeredi 已提交
912 913 914
	}

	revert_creds(old_cred);
915
	dput(index);
M
Miklos Szeredi 已提交
916
	kfree(stack);
M
Miklos Szeredi 已提交
917
	kfree(d.redirect);
M
Miklos Szeredi 已提交
918 919 920 921 922
	d_add(dentry, inode);

	return NULL;

out_free_oe:
M
Miklos Szeredi 已提交
923
	dentry->d_fsdata = NULL;
M
Miklos Szeredi 已提交
924 925
	kfree(oe);
out_put:
926
	dput(index);
M
Miklos Szeredi 已提交
927 928 929 930 931
	for (i = 0; i < ctr; i++)
		dput(stack[i].dentry);
	kfree(stack);
out_put_upper:
	dput(upperdentry);
M
Miklos Szeredi 已提交
932
	kfree(upperredirect);
M
Miklos Szeredi 已提交
933
out:
M
Miklos Szeredi 已提交
934
	kfree(d.redirect);
M
Miklos Szeredi 已提交
935 936 937 938 939 940 941 942 943
	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;
944
	const struct cred *old_cred;
M
Miklos Szeredi 已提交
945 946 947 948 949 950 951 952 953 954 955 956
	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 */
957
	if (!ovl_dentry_upper(dentry))
M
Miklos Szeredi 已提交
958 959
		return true;

960
	old_cred = ovl_override_creds(dentry->d_sb);
M
Miklos Szeredi 已提交
961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989
	/* 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);
		}
	}
990
	revert_creds(old_cred);
M
Miklos Szeredi 已提交
991 992 993

	return positive;
}