readdir.c 22.7 KB
Newer Older
M
Miklos Szeredi 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
/*
 *
 * 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.
 */

#include <linux/fs.h>
#include <linux/slab.h>
#include <linux/namei.h>
#include <linux/file.h>
#include <linux/xattr.h>
#include <linux/rbtree.h>
#include <linux/security.h>
#include <linux/cred.h>
18
#include <linux/ratelimit.h>
M
Miklos Szeredi 已提交
19 20 21 22 23
#include "overlayfs.h"

struct ovl_cache_entry {
	unsigned int len;
	unsigned int type;
24
	u64 real_ino;
M
Miklos Szeredi 已提交
25 26 27
	u64 ino;
	struct list_head l_node;
	struct rb_node node;
28
	struct ovl_cache_entry *next_maybe_whiteout;
M
Miklos Szeredi 已提交
29
	bool is_whiteout;
30
	char name[];
M
Miklos Szeredi 已提交
31 32 33 34 35 36
};

struct ovl_dir_cache {
	long refcount;
	u64 version;
	struct list_head entries;
37
	struct rb_root root;
M
Miklos Szeredi 已提交
38 39 40 41
};

struct ovl_readdir_data {
	struct dir_context ctx;
42
	struct dentry *dentry;
43
	bool is_lowest;
44
	struct rb_root *root;
M
Miklos Szeredi 已提交
45
	struct list_head *list;
46
	struct list_head middle;
47
	struct ovl_cache_entry *first_maybe_whiteout;
M
Miklos Szeredi 已提交
48 49
	int count;
	int err;
50
	bool is_upper;
51
	bool d_type_supported;
M
Miklos Szeredi 已提交
52 53 54 55 56 57
};

struct ovl_dir_file {
	bool is_real;
	bool is_upper;
	struct ovl_dir_cache *cache;
58
	struct list_head *cursor;
M
Miklos Szeredi 已提交
59 60 61 62 63 64
	struct file *realfile;
	struct file *upperfile;
};

static struct ovl_cache_entry *ovl_cache_entry_from_node(struct rb_node *n)
{
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
	return rb_entry(n, struct ovl_cache_entry, node);
}

static bool ovl_cache_entry_find_link(const char *name, int len,
				      struct rb_node ***link,
				      struct rb_node **parent)
{
	bool found = false;
	struct rb_node **newp = *link;

	while (!found && *newp) {
		int cmp;
		struct ovl_cache_entry *tmp;

		*parent = *newp;
		tmp = ovl_cache_entry_from_node(*newp);
		cmp = strncmp(name, tmp->name, len);
		if (cmp > 0)
			newp = &tmp->node.rb_right;
		else if (cmp < 0 || len < tmp->len)
			newp = &tmp->node.rb_left;
		else
			found = true;
	}
	*link = newp;

	return found;
M
Miklos Szeredi 已提交
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
}

static struct ovl_cache_entry *ovl_cache_entry_find(struct rb_root *root,
						    const char *name, int len)
{
	struct rb_node *node = root->rb_node;
	int cmp;

	while (node) {
		struct ovl_cache_entry *p = ovl_cache_entry_from_node(node);

		cmp = strncmp(name, p->name, len);
		if (cmp > 0)
			node = p->node.rb_right;
		else if (cmp < 0 || len < p->len)
			node = p->node.rb_left;
		else
			return p;
	}

	return NULL;
}

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
static bool ovl_calc_d_ino(struct ovl_readdir_data *rdd,
			   struct ovl_cache_entry *p)
{
	/* Don't care if not doing ovl_iter() */
	if (!rdd->dentry)
		return false;

	/* Always recalc d_ino for parent */
	if (strcmp(p->name, "..") == 0)
		return true;

	/* If this is lower, then native d_ino will do */
	if (!rdd->is_upper)
		return false;

	/*
	 * Recalc d_ino for '.' and for all entries if dir is impure (contains
	 * copied up entries)
	 */
	if ((p->name[0] == '.' && p->len == 1) ||
	    ovl_test_flag(OVL_IMPURE, d_inode(rdd->dentry)))
		return true;

	return false;
}

141
static struct ovl_cache_entry *ovl_cache_entry_new(struct ovl_readdir_data *rdd,
142
						   const char *name, int len,
M
Miklos Szeredi 已提交
143 144 145
						   u64 ino, unsigned int d_type)
{
	struct ovl_cache_entry *p;
146
	size_t size = offsetof(struct ovl_cache_entry, name[len + 1]);
M
Miklos Szeredi 已提交
147

148
	p = kmalloc(size, GFP_KERNEL);
149 150 151 152 153 154 155
	if (!p)
		return NULL;

	memcpy(p->name, name, len);
	p->name[len] = '\0';
	p->len = len;
	p->type = d_type;
156
	p->real_ino = ino;
157
	p->ino = ino;
158 159 160
	/* Defer setting d_ino for upper entry to ovl_iterate() */
	if (ovl_calc_d_ino(rdd, p))
		p->ino = 0;
161 162 163
	p->is_whiteout = false;

	if (d_type == DT_CHR) {
164 165
		p->next_maybe_whiteout = rdd->first_maybe_whiteout;
		rdd->first_maybe_whiteout = p;
166
	}
M
Miklos Szeredi 已提交
167 168 169 170 171 172 173
	return p;
}

static int ovl_cache_entry_add_rb(struct ovl_readdir_data *rdd,
				  const char *name, int len, u64 ino,
				  unsigned int d_type)
{
174
	struct rb_node **newp = &rdd->root->rb_node;
M
Miklos Szeredi 已提交
175 176 177
	struct rb_node *parent = NULL;
	struct ovl_cache_entry *p;

178 179
	if (ovl_cache_entry_find_link(name, len, &newp, &parent))
		return 0;
M
Miklos Szeredi 已提交
180

181
	p = ovl_cache_entry_new(rdd, name, len, ino, d_type);
M
Miklos Szeredi 已提交
182 183
	if (p == NULL) {
		rdd->err = -ENOMEM;
M
Miklos Szeredi 已提交
184
		return -ENOMEM;
M
Miklos Szeredi 已提交
185
	}
M
Miklos Szeredi 已提交
186 187 188

	list_add_tail(&p->l_node, rdd->list);
	rb_link_node(&p->node, parent, newp);
189
	rb_insert_color(&p->node, rdd->root);
M
Miklos Szeredi 已提交
190 191 192 193

	return 0;
}

194 195 196
static int ovl_fill_lowest(struct ovl_readdir_data *rdd,
			   const char *name, int namelen,
			   loff_t offset, u64 ino, unsigned int d_type)
M
Miklos Szeredi 已提交
197 198 199
{
	struct ovl_cache_entry *p;

200
	p = ovl_cache_entry_find(rdd->root, name, namelen);
M
Miklos Szeredi 已提交
201
	if (p) {
202
		list_move_tail(&p->l_node, &rdd->middle);
M
Miklos Szeredi 已提交
203
	} else {
204
		p = ovl_cache_entry_new(rdd, name, namelen, ino, d_type);
M
Miklos Szeredi 已提交
205 206 207
		if (p == NULL)
			rdd->err = -ENOMEM;
		else
208
			list_add_tail(&p->l_node, &rdd->middle);
M
Miklos Szeredi 已提交
209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224
	}

	return rdd->err;
}

void ovl_cache_free(struct list_head *list)
{
	struct ovl_cache_entry *p;
	struct ovl_cache_entry *n;

	list_for_each_entry_safe(p, n, list, l_node)
		kfree(p);

	INIT_LIST_HEAD(list);
}

225 226 227 228 229 230 231 232 233 234
void ovl_dir_cache_free(struct inode *inode)
{
	struct ovl_dir_cache *cache = ovl_dir_cache(inode);

	if (cache) {
		ovl_cache_free(&cache->entries);
		kfree(cache);
	}
}

M
Miklos Szeredi 已提交
235 236 237 238 239 240 241
static void ovl_cache_put(struct ovl_dir_file *od, struct dentry *dentry)
{
	struct ovl_dir_cache *cache = od->cache;

	WARN_ON(cache->refcount <= 0);
	cache->refcount--;
	if (!cache->refcount) {
242 243
		if (ovl_dir_cache(d_inode(dentry)) == cache)
			ovl_set_dir_cache(d_inode(dentry), NULL);
M
Miklos Szeredi 已提交
244 245 246 247 248 249

		ovl_cache_free(&cache->entries);
		kfree(cache);
	}
}

250 251 252
static int ovl_fill_merge(struct dir_context *ctx, const char *name,
			  int namelen, loff_t offset, u64 ino,
			  unsigned int d_type)
M
Miklos Szeredi 已提交
253
{
254 255
	struct ovl_readdir_data *rdd =
		container_of(ctx, struct ovl_readdir_data, ctx);
M
Miklos Szeredi 已提交
256 257

	rdd->count++;
258
	if (!rdd->is_lowest)
M
Miklos Szeredi 已提交
259 260
		return ovl_cache_entry_add_rb(rdd, name, namelen, ino, d_type);
	else
261
		return ovl_fill_lowest(rdd, name, namelen, offset, ino, d_type);
M
Miklos Szeredi 已提交
262 263
}

264 265 266 267 268 269 270
static int ovl_check_whiteouts(struct dentry *dir, struct ovl_readdir_data *rdd)
{
	int err;
	struct ovl_cache_entry *p;
	struct dentry *dentry;
	const struct cred *old_cred;

271
	old_cred = ovl_override_creds(rdd->dentry->d_sb);
272

273
	err = down_write_killable(&dir->d_inode->i_rwsem);
274 275 276 277 278 279 280 281 282 283
	if (!err) {
		while (rdd->first_maybe_whiteout) {
			p = rdd->first_maybe_whiteout;
			rdd->first_maybe_whiteout = p->next_maybe_whiteout;
			dentry = lookup_one_len(p->name, dir, p->len);
			if (!IS_ERR(dentry)) {
				p->is_whiteout = ovl_is_whiteout(dentry);
				dput(dentry);
			}
		}
A
Al Viro 已提交
284
		inode_unlock(dir->d_inode);
285 286 287 288 289 290
	}
	revert_creds(old_cred);

	return err;
}

M
Miklos Szeredi 已提交
291 292 293 294 295 296 297 298 299 300
static inline int ovl_dir_read(struct path *realpath,
			       struct ovl_readdir_data *rdd)
{
	struct file *realfile;
	int err;

	realfile = ovl_path_open(realpath, O_RDONLY | O_DIRECTORY);
	if (IS_ERR(realfile))
		return PTR_ERR(realfile);

301
	rdd->first_maybe_whiteout = NULL;
M
Miklos Szeredi 已提交
302 303 304 305 306 307 308 309
	rdd->ctx.pos = 0;
	do {
		rdd->count = 0;
		rdd->err = 0;
		err = iterate_dir(realfile, &rdd->ctx);
		if (err >= 0)
			err = rdd->err;
	} while (!err && rdd->count);
310

M
Miklos Szeredi 已提交
311
	if (!err && rdd->first_maybe_whiteout && rdd->dentry)
312 313
		err = ovl_check_whiteouts(realpath->dentry, rdd);

M
Miklos Szeredi 已提交
314 315 316 317 318 319 320 321 322 323 324 325 326 327 328
	fput(realfile);

	return err;
}

static void ovl_dir_reset(struct file *file)
{
	struct ovl_dir_file *od = file->private_data;
	struct ovl_dir_cache *cache = od->cache;
	struct dentry *dentry = file->f_path.dentry;
	enum ovl_path_type type = ovl_path_type(dentry);

	if (cache && ovl_dentry_version_get(dentry) != cache->version) {
		ovl_cache_put(od, dentry);
		od->cache = NULL;
329
		od->cursor = NULL;
M
Miklos Szeredi 已提交
330
	}
M
Miklos Szeredi 已提交
331 332
	WARN_ON(!od->is_real && !OVL_TYPE_MERGE(type));
	if (od->is_real && OVL_TYPE_MERGE(type))
M
Miklos Szeredi 已提交
333 334 335
		od->is_real = false;
}

336 337
static int ovl_dir_read_merged(struct dentry *dentry, struct list_head *list,
	struct rb_root *root)
M
Miklos Szeredi 已提交
338 339
{
	int err;
M
Miklos Szeredi 已提交
340
	struct path realpath;
M
Miklos Szeredi 已提交
341 342
	struct ovl_readdir_data rdd = {
		.ctx.actor = ovl_fill_merge,
343
		.dentry = dentry,
M
Miklos Szeredi 已提交
344
		.list = list,
345
		.root = root,
346
		.is_lowest = false,
M
Miklos Szeredi 已提交
347
	};
M
Miklos Szeredi 已提交
348
	int idx, next;
M
Miklos Szeredi 已提交
349

M
Miklos Szeredi 已提交
350 351
	for (idx = 0; idx != -1; idx = next) {
		next = ovl_path_next(idx, dentry, &realpath);
352
		rdd.is_upper = ovl_dentry_upper(dentry) == realpath.dentry;
353

M
Miklos Szeredi 已提交
354 355
		if (next != -1) {
			err = ovl_dir_read(&realpath, &rdd);
M
Miklos Szeredi 已提交
356
			if (err)
M
Miklos Szeredi 已提交
357 358 359 360 361 362 363
				break;
		} else {
			/*
			 * Insert lowest layer entries before upper ones, this
			 * allows offsets to be reasonably constant
			 */
			list_add(&rdd.middle, rdd.list);
364
			rdd.is_lowest = true;
M
Miklos Szeredi 已提交
365 366
			err = ovl_dir_read(&realpath, &rdd);
			list_del(&rdd.middle);
M
Miklos Szeredi 已提交
367 368 369 370 371 372 373
		}
	}
	return err;
}

static void ovl_seek_cursor(struct ovl_dir_file *od, loff_t pos)
{
374
	struct list_head *p;
M
Miklos Szeredi 已提交
375 376
	loff_t off = 0;

377
	list_for_each(p, &od->cache->entries) {
M
Miklos Szeredi 已提交
378 379 380 381
		if (off >= pos)
			break;
		off++;
	}
382 383
	/* Cursor is safe since the cache is stable */
	od->cursor = p;
M
Miklos Szeredi 已提交
384 385 386 387 388 389 390
}

static struct ovl_dir_cache *ovl_cache_get(struct dentry *dentry)
{
	int res;
	struct ovl_dir_cache *cache;

391
	cache = ovl_dir_cache(d_inode(dentry));
M
Miklos Szeredi 已提交
392
	if (cache && ovl_dentry_version_get(dentry) == cache->version) {
393
		WARN_ON(!cache->refcount);
M
Miklos Szeredi 已提交
394 395 396
		cache->refcount++;
		return cache;
	}
397
	ovl_set_dir_cache(d_inode(dentry), NULL);
M
Miklos Szeredi 已提交
398 399 400 401 402 403 404

	cache = kzalloc(sizeof(struct ovl_dir_cache), GFP_KERNEL);
	if (!cache)
		return ERR_PTR(-ENOMEM);

	cache->refcount = 1;
	INIT_LIST_HEAD(&cache->entries);
405
	cache->root = RB_ROOT;
M
Miklos Szeredi 已提交
406

407
	res = ovl_dir_read_merged(dentry, &cache->entries, &cache->root);
M
Miklos Szeredi 已提交
408 409 410 411 412 413 414
	if (res) {
		ovl_cache_free(&cache->entries);
		kfree(cache);
		return ERR_PTR(res);
	}

	cache->version = ovl_dentry_version_get(dentry);
415
	ovl_set_dir_cache(d_inode(dentry), cache);
M
Miklos Szeredi 已提交
416 417 418 419

	return cache;
}

420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488
/*
 * Set d_ino for upper entries. Non-upper entries should always report
 * the uppermost real inode ino and should not call this function.
 *
 * When not all layer are on same fs, report real ino also for upper.
 *
 * When all layers are on the same fs, and upper has a reference to
 * copy up origin, call vfs_getattr() on the overlay entry to make
 * sure that d_ino will be consistent with st_ino from stat(2).
 */
static int ovl_cache_update_ino(struct path *path, struct ovl_cache_entry *p)

{
	struct dentry *dir = path->dentry;
	struct dentry *this = NULL;
	enum ovl_path_type type;
	u64 ino = p->real_ino;
	int err = 0;

	if (!ovl_same_sb(dir->d_sb))
		goto out;

	if (p->name[0] == '.') {
		if (p->len == 1) {
			this = dget(dir);
			goto get;
		}
		if (p->len == 2 && p->name[1] == '.') {
			/* we shall not be moved */
			this = dget(dir->d_parent);
			goto get;
		}
	}
	this = lookup_one_len(p->name, dir, p->len);
	if (IS_ERR_OR_NULL(this) || !this->d_inode) {
		if (IS_ERR(this)) {
			err = PTR_ERR(this);
			this = NULL;
			goto fail;
		}
		goto out;
	}

get:
	type = ovl_path_type(this);
	if (OVL_TYPE_ORIGIN(type)) {
		struct kstat stat;
		struct path statpath = *path;

		statpath.dentry = this;
		err = vfs_getattr(&statpath, &stat, STATX_INO, 0);
		if (err)
			goto fail;

		WARN_ON_ONCE(dir->d_sb->s_dev != stat.dev);
		ino = stat.ino;
	}

out:
	p->ino = ino;
	dput(this);
	return err;

fail:
	pr_warn_ratelimited("overlay: failed to look up (%s) for ino (%i)\n",
			    p->name, err);
	goto out;
}

489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651
static int ovl_fill_plain(struct dir_context *ctx, const char *name,
			  int namelen, loff_t offset, u64 ino,
			  unsigned int d_type)
{
	struct ovl_cache_entry *p;
	struct ovl_readdir_data *rdd =
		container_of(ctx, struct ovl_readdir_data, ctx);

	rdd->count++;
	p = ovl_cache_entry_new(rdd, name, namelen, ino, d_type);
	if (p == NULL) {
		rdd->err = -ENOMEM;
		return -ENOMEM;
	}
	list_add_tail(&p->l_node, rdd->list);

	return 0;
}

static int ovl_dir_read_impure(struct path *path,  struct list_head *list,
			       struct rb_root *root)
{
	int err;
	struct path realpath;
	struct ovl_cache_entry *p, *n;
	struct ovl_readdir_data rdd = {
		.ctx.actor = ovl_fill_plain,
		.list = list,
		.root = root,
	};

	INIT_LIST_HEAD(list);
	*root = RB_ROOT;
	ovl_path_upper(path->dentry, &realpath);

	err = ovl_dir_read(&realpath, &rdd);
	if (err)
		return err;

	list_for_each_entry_safe(p, n, list, l_node) {
		if (strcmp(p->name, ".") != 0 &&
		    strcmp(p->name, "..") != 0) {
			err = ovl_cache_update_ino(path, p);
			if (err)
				return err;
		}
		if (p->ino == p->real_ino) {
			list_del(&p->l_node);
			kfree(p);
		} else {
			struct rb_node **newp = &root->rb_node;
			struct rb_node *parent = NULL;

			if (WARN_ON(ovl_cache_entry_find_link(p->name, p->len,
							      &newp, &parent)))
				return -EIO;

			rb_link_node(&p->node, parent, newp);
			rb_insert_color(&p->node, root);
		}
	}
	return 0;
}

static struct ovl_dir_cache *ovl_cache_get_impure(struct path *path)
{
	int res;
	struct dentry *dentry = path->dentry;
	struct ovl_dir_cache *cache;

	cache = ovl_dir_cache(d_inode(dentry));
	if (cache && ovl_dentry_version_get(dentry) == cache->version)
		return cache;

	/* Impure cache is not refcounted, free it here */
	ovl_dir_cache_free(d_inode(dentry));
	ovl_set_dir_cache(d_inode(dentry), NULL);

	cache = kzalloc(sizeof(struct ovl_dir_cache), GFP_KERNEL);
	if (!cache)
		return ERR_PTR(-ENOMEM);

	res = ovl_dir_read_impure(path, &cache->entries, &cache->root);
	if (res) {
		ovl_cache_free(&cache->entries);
		kfree(cache);
		return ERR_PTR(res);
	}
	if (list_empty(&cache->entries)) {
		/* Good oportunity to get rid of an unnecessary "impure" flag */
		ovl_do_removexattr(ovl_dentry_upper(dentry), OVL_XATTR_IMPURE);
		ovl_clear_flag(OVL_IMPURE, d_inode(dentry));
		kfree(cache);
		return NULL;
	}

	cache->version = ovl_dentry_version_get(dentry);
	ovl_set_dir_cache(d_inode(dentry), cache);

	return cache;
}

struct ovl_readdir_translate {
	struct dir_context *orig_ctx;
	struct ovl_dir_cache *cache;
	struct dir_context ctx;
	u64 parent_ino;
};

static int ovl_fill_real(struct dir_context *ctx, const char *name,
			   int namelen, loff_t offset, u64 ino,
			   unsigned int d_type)
{
	struct ovl_readdir_translate *rdt =
		container_of(ctx, struct ovl_readdir_translate, ctx);
	struct dir_context *orig_ctx = rdt->orig_ctx;

	if (rdt->parent_ino && strcmp(name, "..") == 0)
		ino = rdt->parent_ino;
	else if (rdt->cache) {
		struct ovl_cache_entry *p;

		p = ovl_cache_entry_find(&rdt->cache->root, name, namelen);
		if (p)
			ino = p->ino;
	}

	return orig_ctx->actor(orig_ctx, name, namelen, offset, ino, d_type);
}

static int ovl_iterate_real(struct file *file, struct dir_context *ctx)
{
	int err;
	struct ovl_dir_file *od = file->private_data;
	struct dentry *dir = file->f_path.dentry;
	struct ovl_readdir_translate rdt = {
		.ctx.actor = ovl_fill_real,
		.orig_ctx = ctx,
	};

	if (OVL_TYPE_MERGE(ovl_path_type(dir->d_parent))) {
		struct kstat stat;
		struct path statpath = file->f_path;

		statpath.dentry = dir->d_parent;
		err = vfs_getattr(&statpath, &stat, STATX_INO, 0);
		if (err)
			return err;

		WARN_ON_ONCE(dir->d_sb->s_dev != stat.dev);
		rdt.parent_ino = stat.ino;
	}

	if (ovl_test_flag(OVL_IMPURE, d_inode(dir))) {
		rdt.cache = ovl_cache_get_impure(&file->f_path);
		if (IS_ERR(rdt.cache))
			return PTR_ERR(rdt.cache);
	}

	return iterate_dir(od->realfile, &rdt.ctx);
}


M
Miklos Szeredi 已提交
652 653 654 655
static int ovl_iterate(struct file *file, struct dir_context *ctx)
{
	struct ovl_dir_file *od = file->private_data;
	struct dentry *dentry = file->f_path.dentry;
656
	struct ovl_cache_entry *p;
657
	int err;
M
Miklos Szeredi 已提交
658 659 660 661

	if (!ctx->pos)
		ovl_dir_reset(file);

662 663 664 665 666 667 668 669 670 671 672
	if (od->is_real) {
		/*
		 * If parent is merge, then need to adjust d_ino for '..', if
		 * dir is impure then need to adjust d_ino for copied up
		 * entries.
		 */
		if (ovl_same_sb(dentry->d_sb) &&
		    (ovl_test_flag(OVL_IMPURE, d_inode(dentry)) ||
		     OVL_TYPE_MERGE(ovl_path_type(dentry->d_parent)))) {
			return ovl_iterate_real(file, ctx);
		}
M
Miklos Szeredi 已提交
673
		return iterate_dir(od->realfile, ctx);
674
	}
M
Miklos Szeredi 已提交
675 676 677 678 679 680 681 682 683 684 685 686

	if (!od->cache) {
		struct ovl_dir_cache *cache;

		cache = ovl_cache_get(dentry);
		if (IS_ERR(cache))
			return PTR_ERR(cache);

		od->cache = cache;
		ovl_seek_cursor(od, ctx->pos);
	}

687 688
	while (od->cursor != &od->cache->entries) {
		p = list_entry(od->cursor, struct ovl_cache_entry, l_node);
689 690 691 692 693 694
		if (!p->is_whiteout) {
			if (!p->ino) {
				err = ovl_cache_update_ino(&file->f_path, p);
				if (err)
					return err;
			}
695 696
			if (!dir_emit(ctx, p->name, p->len, p->ino, p->type))
				break;
697
		}
698 699
		od->cursor = p->l_node.next;
		ctx->pos++;
M
Miklos Szeredi 已提交
700 701 702 703 704 705 706 707 708
	}
	return 0;
}

static loff_t ovl_dir_llseek(struct file *file, loff_t offset, int origin)
{
	loff_t res;
	struct ovl_dir_file *od = file->private_data;

A
Al Viro 已提交
709
	inode_lock(file_inode(file));
M
Miklos Szeredi 已提交
710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738
	if (!file->f_pos)
		ovl_dir_reset(file);

	if (od->is_real) {
		res = vfs_llseek(od->realfile, offset, origin);
		file->f_pos = od->realfile->f_pos;
	} else {
		res = -EINVAL;

		switch (origin) {
		case SEEK_CUR:
			offset += file->f_pos;
			break;
		case SEEK_SET:
			break;
		default:
			goto out_unlock;
		}
		if (offset < 0)
			goto out_unlock;

		if (offset != file->f_pos) {
			file->f_pos = offset;
			if (od->cache)
				ovl_seek_cursor(od, offset);
		}
		res = offset;
	}
out_unlock:
A
Al Viro 已提交
739
	inode_unlock(file_inode(file));
M
Miklos Szeredi 已提交
740 741 742 743 744 745 746 747 748 749 750 751 752 753

	return res;
}

static int ovl_dir_fsync(struct file *file, loff_t start, loff_t end,
			 int datasync)
{
	struct ovl_dir_file *od = file->private_data;
	struct dentry *dentry = file->f_path.dentry;
	struct file *realfile = od->realfile;

	/*
	 * Need to check if we started out being a lower dir, but got copied up
	 */
M
Miklos Szeredi 已提交
754
	if (!od->is_upper && OVL_TYPE_UPPER(ovl_path_type(dentry))) {
M
Miklos Szeredi 已提交
755 756
		struct inode *inode = file_inode(file);

M
Miklos Szeredi 已提交
757
		realfile = lockless_dereference(od->upperfile);
M
Miklos Szeredi 已提交
758 759 760 761 762
		if (!realfile) {
			struct path upperpath;

			ovl_path_upper(dentry, &upperpath);
			realfile = ovl_path_open(&upperpath, O_RDONLY);
763

A
Al Viro 已提交
764
			inode_lock(inode);
765 766
			if (!od->upperfile) {
				if (IS_ERR(realfile)) {
A
Al Viro 已提交
767
					inode_unlock(inode);
768 769
					return PTR_ERR(realfile);
				}
770
				smp_store_release(&od->upperfile, realfile);
771 772 773 774 775
			} else {
				/* somebody has beaten us to it */
				if (!IS_ERR(realfile))
					fput(realfile);
				realfile = od->upperfile;
M
Miklos Szeredi 已提交
776
			}
A
Al Viro 已提交
777
			inode_unlock(inode);
M
Miklos Szeredi 已提交
778 779 780 781 782 783 784 785 786 787 788
		}
	}

	return vfs_fsync_range(realfile, start, end, datasync);
}

static int ovl_dir_release(struct inode *inode, struct file *file)
{
	struct ovl_dir_file *od = file->private_data;

	if (od->cache) {
A
Al Viro 已提交
789
		inode_lock(inode);
M
Miklos Szeredi 已提交
790
		ovl_cache_put(od, file->f_path.dentry);
A
Al Viro 已提交
791
		inode_unlock(inode);
M
Miklos Szeredi 已提交
792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818
	}
	fput(od->realfile);
	if (od->upperfile)
		fput(od->upperfile);
	kfree(od);

	return 0;
}

static int ovl_dir_open(struct inode *inode, struct file *file)
{
	struct path realpath;
	struct file *realfile;
	struct ovl_dir_file *od;
	enum ovl_path_type type;

	od = kzalloc(sizeof(struct ovl_dir_file), GFP_KERNEL);
	if (!od)
		return -ENOMEM;

	type = ovl_path_real(file->f_path.dentry, &realpath);
	realfile = ovl_path_open(&realpath, file->f_flags);
	if (IS_ERR(realfile)) {
		kfree(od);
		return PTR_ERR(realfile);
	}
	od->realfile = realfile;
M
Miklos Szeredi 已提交
819 820
	od->is_real = !OVL_TYPE_MERGE(type);
	od->is_upper = OVL_TYPE_UPPER(type);
M
Miklos Szeredi 已提交
821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838
	file->private_data = od;

	return 0;
}

const struct file_operations ovl_dir_operations = {
	.read		= generic_read_dir,
	.open		= ovl_dir_open,
	.iterate	= ovl_iterate,
	.llseek		= ovl_dir_llseek,
	.fsync		= ovl_dir_fsync,
	.release	= ovl_dir_release,
};

int ovl_check_empty_dir(struct dentry *dentry, struct list_head *list)
{
	int err;
	struct ovl_cache_entry *p;
839
	struct rb_root root = RB_ROOT;
M
Miklos Szeredi 已提交
840

841
	err = ovl_dir_read_merged(dentry, list, &root);
M
Miklos Szeredi 已提交
842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867
	if (err)
		return err;

	err = 0;

	list_for_each_entry(p, list, l_node) {
		if (p->is_whiteout)
			continue;

		if (p->name[0] == '.') {
			if (p->len == 1)
				continue;
			if (p->len == 2 && p->name[1] == '.')
				continue;
		}
		err = -ENOTEMPTY;
		break;
	}

	return err;
}

void ovl_cleanup_whiteouts(struct dentry *upper, struct list_head *list)
{
	struct ovl_cache_entry *p;

A
Al Viro 已提交
868
	inode_lock_nested(upper->d_inode, I_MUTEX_CHILD);
M
Miklos Szeredi 已提交
869 870 871 872 873 874 875 876 877 878 879 880 881
	list_for_each_entry(p, list, l_node) {
		struct dentry *dentry;

		if (!p->is_whiteout)
			continue;

		dentry = lookup_one_len(p->name, upper, p->len);
		if (IS_ERR(dentry)) {
			pr_err("overlayfs: lookup '%s/%.*s' failed (%i)\n",
			       upper->d_name.name, p->len, p->name,
			       (int) PTR_ERR(dentry));
			continue;
		}
882 883
		if (dentry->d_inode)
			ovl_cleanup(upper->d_inode, dentry);
M
Miklos Szeredi 已提交
884 885
		dput(dentry);
	}
A
Al Viro 已提交
886
	inode_unlock(upper->d_inode);
M
Miklos Szeredi 已提交
887
}
888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923

static int ovl_check_d_type(struct dir_context *ctx, const char *name,
			  int namelen, loff_t offset, u64 ino,
			  unsigned int d_type)
{
	struct ovl_readdir_data *rdd =
		container_of(ctx, struct ovl_readdir_data, ctx);

	/* Even if d_type is not supported, DT_DIR is returned for . and .. */
	if (!strncmp(name, ".", namelen) || !strncmp(name, "..", namelen))
		return 0;

	if (d_type != DT_UNKNOWN)
		rdd->d_type_supported = true;

	return 0;
}

/*
 * Returns 1 if d_type is supported, 0 not supported/unknown. Negative values
 * if error is encountered.
 */
int ovl_check_d_type_supported(struct path *realpath)
{
	int err;
	struct ovl_readdir_data rdd = {
		.ctx.actor = ovl_check_d_type,
		.d_type_supported = false,
	};

	err = ovl_dir_read(realpath, &rdd);
	if (err)
		return err;

	return rdd.d_type_supported;
}
M
Miklos Szeredi 已提交
924 925 926 927 928 929

static void ovl_workdir_cleanup_recurse(struct path *path, int level)
{
	int err;
	struct inode *dir = path->dentry->d_inode;
	LIST_HEAD(list);
930
	struct rb_root root = RB_ROOT;
M
Miklos Szeredi 已提交
931 932 933 934 935
	struct ovl_cache_entry *p;
	struct ovl_readdir_data rdd = {
		.ctx.actor = ovl_fill_merge,
		.dentry = NULL,
		.list = &list,
936
		.root = &root,
M
Miklos Szeredi 已提交
937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985
		.is_lowest = false,
	};

	err = ovl_dir_read(path, &rdd);
	if (err)
		goto out;

	inode_lock_nested(dir, I_MUTEX_PARENT);
	list_for_each_entry(p, &list, l_node) {
		struct dentry *dentry;

		if (p->name[0] == '.') {
			if (p->len == 1)
				continue;
			if (p->len == 2 && p->name[1] == '.')
				continue;
		}
		dentry = lookup_one_len(p->name, path->dentry, p->len);
		if (IS_ERR(dentry))
			continue;
		if (dentry->d_inode)
			ovl_workdir_cleanup(dir, path->mnt, dentry, level);
		dput(dentry);
	}
	inode_unlock(dir);
out:
	ovl_cache_free(&list);
}

void ovl_workdir_cleanup(struct inode *dir, struct vfsmount *mnt,
			 struct dentry *dentry, int level)
{
	int err;

	if (!d_is_dir(dentry) || level > 1) {
		ovl_cleanup(dir, dentry);
		return;
	}

	err = ovl_do_rmdir(dir, dentry);
	if (err) {
		struct path path = { .mnt = mnt, .dentry = dentry };

		inode_unlock(dir);
		ovl_workdir_cleanup_recurse(&path, level + 1);
		inode_lock_nested(dir, I_MUTEX_PARENT);
		ovl_cleanup(dir, dentry);
	}
}
986 987 988 989 990

int ovl_indexdir_cleanup(struct dentry *dentry, struct vfsmount *mnt,
			 struct path *lowerstack, unsigned int numlower)
{
	int err;
991
	struct dentry *index = NULL;
992 993 994
	struct inode *dir = dentry->d_inode;
	struct path path = { .mnt = mnt, .dentry = dentry };
	LIST_HEAD(list);
995
	struct rb_root root = RB_ROOT;
996 997 998 999 1000
	struct ovl_cache_entry *p;
	struct ovl_readdir_data rdd = {
		.ctx.actor = ovl_fill_merge,
		.dentry = NULL,
		.list = &list,
1001
		.root = &root,
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019
		.is_lowest = false,
	};

	err = ovl_dir_read(&path, &rdd);
	if (err)
		goto out;

	inode_lock_nested(dir, I_MUTEX_PARENT);
	list_for_each_entry(p, &list, l_node) {
		if (p->name[0] == '.') {
			if (p->len == 1)
				continue;
			if (p->len == 2 && p->name[1] == '.')
				continue;
		}
		index = lookup_one_len(p->name, dentry, p->len);
		if (IS_ERR(index)) {
			err = PTR_ERR(index);
1020
			index = NULL;
1021 1022
			break;
		}
1023
		err = ovl_verify_index(index, lowerstack, numlower);
1024 1025
		/* Cleanup stale and orphan index entries */
		if (err && (err == -ESTALE || err == -ENOENT))
1026
			err = ovl_cleanup(dir, index);
1027 1028 1029
		if (err)
			break;

1030
		dput(index);
1031
		index = NULL;
1032
	}
1033
	dput(index);
1034 1035 1036 1037 1038 1039 1040
	inode_unlock(dir);
out:
	ovl_cache_free(&list);
	if (err)
		pr_err("overlayfs: failed index dir cleanup (%i)\n", err);
	return err;
}