xattr.c 42.2 KB
Newer Older
1
/*
2
 * linux/fs/ext4/xattr.c
3 4 5 6
 *
 * Copyright (C) 2001-2003 Andreas Gruenbacher, <agruen@suse.de>
 *
 * Fix by Harrison Xing <harrison@mountainviewdata.com>.
7
 * Ext4 code with a lot of help from Eric Jarman <ejarman@acm.org>.
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
 * Extended attributes for symlinks and special files added per
 *  suggestion of Luka Renko <luka.renko@hermes.si>.
 * xattr consolidation Copyright (c) 2004 James Morris <jmorris@redhat.com>,
 *  Red Hat Inc.
 * ea-in-inode support by Alex Tomas <alex@clusterfs.com> aka bzzz
 *  and Andreas Gruenbacher <agruen@suse.de>.
 */

/*
 * Extended attributes are stored directly in inodes (on file systems with
 * inodes bigger than 128 bytes) and on additional disk blocks. The i_file_acl
 * field contains the block number if an inode uses an additional block. All
 * attributes must fit in the inode and one additional block. Blocks that
 * contain the identical set of attributes may be shared among several inodes.
 * Identical blocks are detected by keeping a cache of blocks that have
 * recently been accessed.
 *
 * The attributes in inodes and on blocks have a different header; the entries
 * are stored in the same format:
 *
 *   +------------------+
 *   | header           |
 *   | entry 1          | |
 *   | entry 2          | | growing downwards
 *   | entry 3          | v
 *   | four null bytes  |
 *   | . . .            |
 *   | value 1          | ^
 *   | value 3          | | growing upwards
 *   | value 2          | |
 *   +------------------+
 *
 * The header is followed by multiple entry descriptors. In disk blocks, the
 * entry descriptors are kept sorted. In inodes, they are unsorted. The
 * attribute values are aligned to the end of the block in no specific order.
 *
 * Locking strategy
 * ----------------
46
 * EXT4_I(inode)->i_file_acl is protected by EXT4_I(inode)->xattr_sem.
47 48 49 50 51 52 53 54 55 56 57 58
 * EA blocks are only changed if they are exclusive to an inode, so
 * holding xattr_sem also means that nothing but the EA block's reference
 * count can change. Multiple writers to the same block are synchronized
 * by the buffer lock.
 */

#include <linux/init.h>
#include <linux/fs.h>
#include <linux/slab.h>
#include <linux/mbcache.h>
#include <linux/quotaops.h>
#include <linux/rwsem.h>
59 60
#include "ext4_jbd2.h"
#include "ext4.h"
61 62 63
#include "xattr.h"
#include "acl.h"

64 65
#define BHDR(bh) ((struct ext4_xattr_header *)((bh)->b_data))
#define ENTRY(ptr) ((struct ext4_xattr_entry *)(ptr))
66 67 68
#define BFIRST(bh) ENTRY(BHDR(bh)+1)
#define IS_LAST_ENTRY(entry) (*(__u32 *)(entry) == 0)

69
#ifdef EXT4_XATTR_DEBUG
70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
# define ea_idebug(inode, f...) do { \
		printk(KERN_DEBUG "inode %s:%lu: ", \
			inode->i_sb->s_id, inode->i_ino); \
		printk(f); \
		printk("\n"); \
	} while (0)
# define ea_bdebug(bh, f...) do { \
		char b[BDEVNAME_SIZE]; \
		printk(KERN_DEBUG "block %s:%lu: ", \
			bdevname(bh->b_bdev, b), \
			(unsigned long) bh->b_blocknr); \
		printk(f); \
		printk("\n"); \
	} while (0)
#else
# define ea_idebug(f...)
# define ea_bdebug(f...)
#endif

89 90 91
static void ext4_xattr_cache_insert(struct buffer_head *);
static struct buffer_head *ext4_xattr_cache_find(struct inode *,
						 struct ext4_xattr_header *,
92
						 struct mb_cache_entry **);
93 94
static void ext4_xattr_rehash(struct ext4_xattr_header *,
			      struct ext4_xattr_entry *);
95
static int ext4_xattr_list(struct dentry *dentry, char *buffer,
96
			   size_t buffer_size);
97

98
static struct mb_cache *ext4_xattr_cache;
99

100 101
static struct xattr_handler *ext4_xattr_handler_map[] = {
	[EXT4_XATTR_INDEX_USER]		     = &ext4_xattr_user_handler,
T
Theodore Ts'o 已提交
102
#ifdef CONFIG_EXT4_FS_POSIX_ACL
103 104
	[EXT4_XATTR_INDEX_POSIX_ACL_ACCESS]  = &ext4_xattr_acl_access_handler,
	[EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT] = &ext4_xattr_acl_default_handler,
105
#endif
106
	[EXT4_XATTR_INDEX_TRUSTED]	     = &ext4_xattr_trusted_handler,
T
Theodore Ts'o 已提交
107
#ifdef CONFIG_EXT4_FS_SECURITY
108
	[EXT4_XATTR_INDEX_SECURITY]	     = &ext4_xattr_security_handler,
109 110 111
#endif
};

112 113 114
struct xattr_handler *ext4_xattr_handlers[] = {
	&ext4_xattr_user_handler,
	&ext4_xattr_trusted_handler,
T
Theodore Ts'o 已提交
115
#ifdef CONFIG_EXT4_FS_POSIX_ACL
116 117
	&ext4_xattr_acl_access_handler,
	&ext4_xattr_acl_default_handler,
118
#endif
T
Theodore Ts'o 已提交
119
#ifdef CONFIG_EXT4_FS_SECURITY
120
	&ext4_xattr_security_handler,
121 122 123 124 125
#endif
	NULL
};

static inline struct xattr_handler *
126
ext4_xattr_handler(int name_index)
127 128 129
{
	struct xattr_handler *handler = NULL;

130 131
	if (name_index > 0 && name_index < ARRAY_SIZE(ext4_xattr_handler_map))
		handler = ext4_xattr_handler_map[name_index];
132 133 134 135 136 137 138 139 140
	return handler;
}

/*
 * Inode operation listxattr()
 *
 * dentry->d_inode->i_mutex: don't care
 */
ssize_t
141
ext4_listxattr(struct dentry *dentry, char *buffer, size_t size)
142
{
143
	return ext4_xattr_list(dentry, buffer, size);
144 145 146
}

static int
147
ext4_xattr_check_names(struct ext4_xattr_entry *entry, void *end)
148 149
{
	while (!IS_LAST_ENTRY(entry)) {
150
		struct ext4_xattr_entry *next = EXT4_XATTR_NEXT(entry);
151 152 153 154 155 156 157 158
		if ((void *)next >= end)
			return -EIO;
		entry = next;
	}
	return 0;
}

static inline int
159
ext4_xattr_check_block(struct buffer_head *bh)
160 161 162
{
	int error;

163
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
164 165
	    BHDR(bh)->h_blocks != cpu_to_le32(1))
		return -EIO;
166
	error = ext4_xattr_check_names(BFIRST(bh), bh->b_data + bh->b_size);
167 168 169 170
	return error;
}

static inline int
171
ext4_xattr_check_entry(struct ext4_xattr_entry *entry, size_t size)
172 173 174 175 176 177 178 179 180 181
{
	size_t value_size = le32_to_cpu(entry->e_value_size);

	if (entry->e_value_block != 0 || value_size > size ||
	    le16_to_cpu(entry->e_value_offs) + value_size > size)
		return -EIO;
	return 0;
}

static int
182
ext4_xattr_find_entry(struct ext4_xattr_entry **pentry, int name_index,
183 184
		      const char *name, size_t size, int sorted)
{
185
	struct ext4_xattr_entry *entry;
186 187 188 189 190 191 192
	size_t name_len;
	int cmp = 1;

	if (name == NULL)
		return -EINVAL;
	name_len = strlen(name);
	entry = *pentry;
193
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
194 195 196 197 198 199 200 201 202
		cmp = name_index - entry->e_name_index;
		if (!cmp)
			cmp = name_len - entry->e_name_len;
		if (!cmp)
			cmp = memcmp(name, entry->e_name, name_len);
		if (cmp <= 0 && (sorted || cmp == 0))
			break;
	}
	*pentry = entry;
203
	if (!cmp && ext4_xattr_check_entry(entry, size))
204 205 206 207 208
			return -EIO;
	return cmp ? -ENODATA : 0;
}

static int
209
ext4_xattr_block_get(struct inode *inode, int name_index, const char *name,
210 211 212
		     void *buffer, size_t buffer_size)
{
	struct buffer_head *bh = NULL;
213
	struct ext4_xattr_entry *entry;
214 215 216 217 218 219 220
	size_t size;
	int error;

	ea_idebug(inode, "name=%d.%s, buffer=%p, buffer_size=%ld",
		  name_index, name, buffer, (long)buffer_size);

	error = -ENODATA;
221
	if (!EXT4_I(inode)->i_file_acl)
222
		goto cleanup;
223 224
	ea_idebug(inode, "reading block %u", EXT4_I(inode)->i_file_acl);
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
225 226 227 228
	if (!bh)
		goto cleanup;
	ea_bdebug(bh, "b_count=%d, refcount=%d",
		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
229
	if (ext4_xattr_check_block(bh)) {
230 231
bad_block:
		ext4_error(inode->i_sb,
232
			   "inode %lu: bad block %llu", inode->i_ino,
233
			   EXT4_I(inode)->i_file_acl);
234 235 236
		error = -EIO;
		goto cleanup;
	}
237
	ext4_xattr_cache_insert(bh);
238
	entry = BFIRST(bh);
239
	error = ext4_xattr_find_entry(&entry, name_index, name, bh->b_size, 1);
240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259
	if (error == -EIO)
		goto bad_block;
	if (error)
		goto cleanup;
	size = le32_to_cpu(entry->e_value_size);
	if (buffer) {
		error = -ERANGE;
		if (size > buffer_size)
			goto cleanup;
		memcpy(buffer, bh->b_data + le16_to_cpu(entry->e_value_offs),
		       size);
	}
	error = size;

cleanup:
	brelse(bh);
	return error;
}

static int
260
ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name,
261 262
		     void *buffer, size_t buffer_size)
{
263 264 265 266
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_entry *entry;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
267 268 269 270
	size_t size;
	void *end;
	int error;

271
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
272
		return -ENODATA;
273
	error = ext4_get_inode_loc(inode, &iloc);
274 275
	if (error)
		return error;
276
	raw_inode = ext4_raw_inode(&iloc);
277 278
	header = IHDR(inode, raw_inode);
	entry = IFIRST(header);
279 280
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(entry, end);
281 282
	if (error)
		goto cleanup;
283
	error = ext4_xattr_find_entry(&entry, name_index, name,
284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302
				      end - (void *)entry, 0);
	if (error)
		goto cleanup;
	size = le32_to_cpu(entry->e_value_size);
	if (buffer) {
		error = -ERANGE;
		if (size > buffer_size)
			goto cleanup;
		memcpy(buffer, (void *)IFIRST(header) +
		       le16_to_cpu(entry->e_value_offs), size);
	}
	error = size;

cleanup:
	brelse(iloc.bh);
	return error;
}

/*
303
 * ext4_xattr_get()
304 305 306 307 308 309 310 311 312
 *
 * Copy an extended attribute into the buffer
 * provided, or compute the buffer size required.
 * Buffer is NULL to compute the size of the buffer required.
 *
 * Returns a negative error number on failure, or the number of bytes
 * used / required on success.
 */
int
313
ext4_xattr_get(struct inode *inode, int name_index, const char *name,
314 315 316 317
	       void *buffer, size_t buffer_size)
{
	int error;

318 319
	down_read(&EXT4_I(inode)->xattr_sem);
	error = ext4_xattr_ibody_get(inode, name_index, name, buffer,
320 321
				     buffer_size);
	if (error == -ENODATA)
322
		error = ext4_xattr_block_get(inode, name_index, name, buffer,
323
					     buffer_size);
324
	up_read(&EXT4_I(inode)->xattr_sem);
325 326 327 328
	return error;
}

static int
329
ext4_xattr_list_entries(struct dentry *dentry, struct ext4_xattr_entry *entry,
330 331 332 333
			char *buffer, size_t buffer_size)
{
	size_t rest = buffer_size;

334
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
335
		struct xattr_handler *handler =
336
			ext4_xattr_handler(entry->e_name_index);
337 338

		if (handler) {
339
			size_t size = handler->list(dentry, buffer, rest,
340
						    entry->e_name,
341 342
						    entry->e_name_len,
						    handler->flags);
343 344 345 346 347 348 349 350 351 352 353 354
			if (buffer) {
				if (size > rest)
					return -ERANGE;
				buffer += size;
			}
			rest -= size;
		}
	}
	return buffer_size - rest;
}

static int
355
ext4_xattr_block_list(struct dentry *dentry, char *buffer, size_t buffer_size)
356
{
357
	struct inode *inode = dentry->d_inode;
358 359 360 361 362 363 364
	struct buffer_head *bh = NULL;
	int error;

	ea_idebug(inode, "buffer=%p, buffer_size=%ld",
		  buffer, (long)buffer_size);

	error = 0;
365
	if (!EXT4_I(inode)->i_file_acl)
366
		goto cleanup;
367 368
	ea_idebug(inode, "reading block %u", EXT4_I(inode)->i_file_acl);
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
369 370 371 372 373
	error = -EIO;
	if (!bh)
		goto cleanup;
	ea_bdebug(bh, "b_count=%d, refcount=%d",
		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
374
	if (ext4_xattr_check_block(bh)) {
375
		ext4_error(inode->i_sb,
376
			   "inode %lu: bad block %llu", inode->i_ino,
377
			   EXT4_I(inode)->i_file_acl);
378 379 380
		error = -EIO;
		goto cleanup;
	}
381
	ext4_xattr_cache_insert(bh);
382
	error = ext4_xattr_list_entries(dentry, BFIRST(bh), buffer, buffer_size);
383 384 385 386 387 388 389 390

cleanup:
	brelse(bh);

	return error;
}

static int
391
ext4_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size)
392
{
393
	struct inode *inode = dentry->d_inode;
394 395 396
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
397 398 399
	void *end;
	int error;

400
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
401
		return 0;
402
	error = ext4_get_inode_loc(inode, &iloc);
403 404
	if (error)
		return error;
405
	raw_inode = ext4_raw_inode(&iloc);
406
	header = IHDR(inode, raw_inode);
407 408
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(IFIRST(header), end);
409 410
	if (error)
		goto cleanup;
411
	error = ext4_xattr_list_entries(dentry, IFIRST(header),
412 413 414 415 416 417 418 419
					buffer, buffer_size);

cleanup:
	brelse(iloc.bh);
	return error;
}

/*
420
 * ext4_xattr_list()
421 422 423 424 425 426 427 428
 *
 * Copy a list of attribute names into the buffer
 * provided, or compute the buffer size required.
 * Buffer is NULL to compute the size of the buffer required.
 *
 * Returns a negative error number on failure, or the number of bytes
 * used / required on success.
 */
429
static int
430
ext4_xattr_list(struct dentry *dentry, char *buffer, size_t buffer_size)
431 432 433
{
	int i_error, b_error;

434 435
	down_read(&EXT4_I(dentry->d_inode)->xattr_sem);
	i_error = ext4_xattr_ibody_list(dentry, buffer, buffer_size);
436 437 438 439 440 441 442
	if (i_error < 0) {
		b_error = 0;
	} else {
		if (buffer) {
			buffer += i_error;
			buffer_size -= i_error;
		}
443
		b_error = ext4_xattr_block_list(dentry, buffer, buffer_size);
444 445 446
		if (b_error < 0)
			i_error = 0;
	}
447
	up_read(&EXT4_I(dentry->d_inode)->xattr_sem);
448 449 450 451
	return i_error + b_error;
}

/*
452
 * If the EXT4_FEATURE_COMPAT_EXT_ATTR feature of this file system is
453 454
 * not set, set it.
 */
455
static void ext4_xattr_update_super_block(handle_t *handle,
456 457
					  struct super_block *sb)
{
458
	if (EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR))
459 460
		return;

461
	if (ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh) == 0) {
462
		EXT4_SET_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR);
463
		sb->s_dirt = 1;
464
		ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
465 466 467 468 469 470 471 472
	}
}

/*
 * Release the xattr block BH: If the reference count is > 1, decrement
 * it; otherwise free the block.
 */
static void
473
ext4_xattr_release_block(handle_t *handle, struct inode *inode,
474 475 476
			 struct buffer_head *bh)
{
	struct mb_cache_entry *ce = NULL;
477
	int error = 0;
478

479
	ce = mb_cache_entry_get(ext4_xattr_cache, bh->b_bdev, bh->b_blocknr);
480 481 482 483 484
	error = ext4_journal_get_write_access(handle, bh);
	if (error)
		goto out;

	lock_buffer(bh);
485 486 487 488 489
	if (BHDR(bh)->h_refcount == cpu_to_le32(1)) {
		ea_bdebug(bh, "refcount now=0; freeing");
		if (ce)
			mb_cache_entry_free(ce);
		get_bh(bh);
490 491 492
		ext4_free_blocks(handle, inode, bh, 0, 1,
				 EXT4_FREE_BLOCKS_METADATA |
				 EXT4_FREE_BLOCKS_FORGET);
493
	} else {
M
Marcin Slusarz 已提交
494
		le32_add_cpu(&BHDR(bh)->h_refcount, -1);
495
		error = ext4_handle_dirty_metadata(handle, inode, bh);
496
		if (IS_SYNC(inode))
497
			ext4_handle_sync(handle);
498
		dquot_free_block(inode, 1);
499 500
		ea_bdebug(bh, "refcount now=%d; releasing",
			  le32_to_cpu(BHDR(bh)->h_refcount));
501 502 503
		if (ce)
			mb_cache_entry_release(ce);
	}
504 505 506 507
	unlock_buffer(bh);
out:
	ext4_std_error(inode->i_sb, error);
	return;
508 509
}

510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527
/*
 * Find the available free space for EAs. This also returns the total number of
 * bytes used by EA entries.
 */
static size_t ext4_xattr_free_space(struct ext4_xattr_entry *last,
				    size_t *min_offs, void *base, int *total)
{
	for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) {
		*total += EXT4_XATTR_LEN(last->e_name_len);
		if (!last->e_value_block && last->e_value_size) {
			size_t offs = le16_to_cpu(last->e_value_offs);
			if (offs < *min_offs)
				*min_offs = offs;
		}
	}
	return (*min_offs - ((void *)last - base) - sizeof(__u32));
}

528
struct ext4_xattr_info {
529 530 531 532 533 534
	int name_index;
	const char *name;
	const void *value;
	size_t value_len;
};

535 536
struct ext4_xattr_search {
	struct ext4_xattr_entry *first;
537 538
	void *base;
	void *end;
539
	struct ext4_xattr_entry *here;
540 541 542 543
	int not_found;
};

static int
544
ext4_xattr_set_entry(struct ext4_xattr_info *i, struct ext4_xattr_search *s)
545
{
546
	struct ext4_xattr_entry *last;
547 548 549 550
	size_t free, min_offs = s->end - s->base, name_len = strlen(i->name);

	/* Compute min_offs and last. */
	last = s->first;
551
	for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) {
552 553 554 555 556 557 558 559 560 561
		if (!last->e_value_block && last->e_value_size) {
			size_t offs = le16_to_cpu(last->e_value_offs);
			if (offs < min_offs)
				min_offs = offs;
		}
	}
	free = min_offs - ((void *)last - s->base) - sizeof(__u32);
	if (!s->not_found) {
		if (!s->here->e_value_block && s->here->e_value_size) {
			size_t size = le32_to_cpu(s->here->e_value_size);
562
			free += EXT4_XATTR_SIZE(size);
563
		}
564
		free += EXT4_XATTR_LEN(name_len);
565 566
	}
	if (i->value) {
567 568 569
		if (free < EXT4_XATTR_SIZE(i->value_len) ||
		    free < EXT4_XATTR_LEN(name_len) +
			   EXT4_XATTR_SIZE(i->value_len))
570 571 572 573 574
			return -ENOSPC;
	}

	if (i->value && s->not_found) {
		/* Insert the new name. */
575
		size_t size = EXT4_XATTR_LEN(name_len);
576 577 578 579 580 581 582 583 584 585 586
		size_t rest = (void *)last - (void *)s->here + sizeof(__u32);
		memmove((void *)s->here + size, s->here, rest);
		memset(s->here, 0, size);
		s->here->e_name_index = i->name_index;
		s->here->e_name_len = name_len;
		memcpy(s->here->e_name, i->name, name_len);
	} else {
		if (!s->here->e_value_block && s->here->e_value_size) {
			void *first_val = s->base + min_offs;
			size_t offs = le16_to_cpu(s->here->e_value_offs);
			void *val = s->base + offs;
587
			size_t size = EXT4_XATTR_SIZE(
588 589
				le32_to_cpu(s->here->e_value_size));

590
			if (i->value && size == EXT4_XATTR_SIZE(i->value_len)) {
591 592 593 594
				/* The old and the new value have the same
				   size. Just replace. */
				s->here->e_value_size =
					cpu_to_le32(i->value_len);
595 596
				memset(val + size - EXT4_XATTR_PAD, 0,
				       EXT4_XATTR_PAD); /* Clear pad bytes. */
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615
				memcpy(val, i->value, i->value_len);
				return 0;
			}

			/* Remove the old value. */
			memmove(first_val + size, first_val, val - first_val);
			memset(first_val, 0, size);
			s->here->e_value_size = 0;
			s->here->e_value_offs = 0;
			min_offs += size;

			/* Adjust all value offsets. */
			last = s->first;
			while (!IS_LAST_ENTRY(last)) {
				size_t o = le16_to_cpu(last->e_value_offs);
				if (!last->e_value_block &&
				    last->e_value_size && o < offs)
					last->e_value_offs =
						cpu_to_le16(o + size);
616
				last = EXT4_XATTR_NEXT(last);
617 618 619 620
			}
		}
		if (!i->value) {
			/* Remove the old name. */
621
			size_t size = EXT4_XATTR_LEN(name_len);
622 623 624 625 626 627 628 629 630 631 632
			last = ENTRY((void *)last - size);
			memmove(s->here, (void *)s->here + size,
				(void *)last - (void *)s->here + sizeof(__u32));
			memset(last, 0, size);
		}
	}

	if (i->value) {
		/* Insert the new value. */
		s->here->e_value_size = cpu_to_le32(i->value_len);
		if (i->value_len) {
633
			size_t size = EXT4_XATTR_SIZE(i->value_len);
634 635
			void *val = s->base + min_offs - size;
			s->here->e_value_offs = cpu_to_le16(min_offs - size);
636 637
			memset(val + size - EXT4_XATTR_PAD, 0,
			       EXT4_XATTR_PAD); /* Clear the pad bytes. */
638 639 640 641 642 643
			memcpy(val, i->value, i->value_len);
		}
	}
	return 0;
}

644 645
struct ext4_xattr_block_find {
	struct ext4_xattr_search s;
646 647 648 649
	struct buffer_head *bh;
};

static int
650 651
ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
		      struct ext4_xattr_block_find *bs)
652 653 654 655 656 657 658
{
	struct super_block *sb = inode->i_sb;
	int error;

	ea_idebug(inode, "name=%d.%s, value=%p, value_len=%ld",
		  i->name_index, i->name, i->value, (long)i->value_len);

659
	if (EXT4_I(inode)->i_file_acl) {
660
		/* The inode already has an extended attribute block. */
661
		bs->bh = sb_bread(sb, EXT4_I(inode)->i_file_acl);
662 663 664 665 666 667
		error = -EIO;
		if (!bs->bh)
			goto cleanup;
		ea_bdebug(bs->bh, "b_count=%d, refcount=%d",
			atomic_read(&(bs->bh->b_count)),
			le32_to_cpu(BHDR(bs->bh)->h_refcount));
668
		if (ext4_xattr_check_block(bs->bh)) {
669 670
			ext4_error(sb, "inode %lu: bad block %llu",
				   inode->i_ino, EXT4_I(inode)->i_file_acl);
671 672 673 674 675 676 677 678
			error = -EIO;
			goto cleanup;
		}
		/* Find the named attribute. */
		bs->s.base = BHDR(bs->bh);
		bs->s.first = BFIRST(bs->bh);
		bs->s.end = bs->bh->b_data + bs->bh->b_size;
		bs->s.here = bs->s.first;
679
		error = ext4_xattr_find_entry(&bs->s.here, i->name_index,
680 681 682 683 684 685 686 687 688 689 690 691
					      i->name, bs->bh->b_size, 1);
		if (error && error != -ENODATA)
			goto cleanup;
		bs->s.not_found = error;
	}
	error = 0;

cleanup:
	return error;
}

static int
692 693 694
ext4_xattr_block_set(handle_t *handle, struct inode *inode,
		     struct ext4_xattr_info *i,
		     struct ext4_xattr_block_find *bs)
695 696 697
{
	struct super_block *sb = inode->i_sb;
	struct buffer_head *new_bh = NULL;
698
	struct ext4_xattr_search *s = &bs->s;
699
	struct mb_cache_entry *ce = NULL;
700
	int error = 0;
701

702
#define header(x) ((struct ext4_xattr_header *)(x))
703 704 705 706

	if (i->value && i->value_len > sb->s_blocksize)
		return -ENOSPC;
	if (s->base) {
707
		ce = mb_cache_entry_get(ext4_xattr_cache, bs->bh->b_bdev,
708
					bs->bh->b_blocknr);
709 710 711 712 713
		error = ext4_journal_get_write_access(handle, bs->bh);
		if (error)
			goto cleanup;
		lock_buffer(bs->bh);

714 715 716 717 718 719
		if (header(s->base)->h_refcount == cpu_to_le32(1)) {
			if (ce) {
				mb_cache_entry_free(ce);
				ce = NULL;
			}
			ea_bdebug(bs->bh, "modifying in-place");
720
			error = ext4_xattr_set_entry(i, s);
721 722
			if (!error) {
				if (!IS_LAST_ENTRY(s->first))
723
					ext4_xattr_rehash(header(s->base),
724
							  s->here);
725
				ext4_xattr_cache_insert(bs->bh);
726 727 728 729 730
			}
			unlock_buffer(bs->bh);
			if (error == -EIO)
				goto bad_block;
			if (!error)
731 732 733
				error = ext4_handle_dirty_metadata(handle,
								   inode,
								   bs->bh);
734 735 736 737 738 739
			if (error)
				goto cleanup;
			goto inserted;
		} else {
			int offset = (char *)s->here - bs->bh->b_data;

740 741
			unlock_buffer(bs->bh);
			jbd2_journal_release_buffer(handle, bs->bh);
742 743 744 745 746
			if (ce) {
				mb_cache_entry_release(ce);
				ce = NULL;
			}
			ea_bdebug(bs->bh, "cloning");
747
			s->base = kmalloc(bs->bh->b_size, GFP_NOFS);
748 749 750 751 752 753 754 755 756 757 758
			error = -ENOMEM;
			if (s->base == NULL)
				goto cleanup;
			memcpy(s->base, BHDR(bs->bh), bs->bh->b_size);
			s->first = ENTRY(header(s->base)+1);
			header(s->base)->h_refcount = cpu_to_le32(1);
			s->here = ENTRY(s->base + offset);
			s->end = s->base + bs->bh->b_size;
		}
	} else {
		/* Allocate a buffer where we construct the new block. */
759
		s->base = kzalloc(sb->s_blocksize, GFP_NOFS);
760 761 762 763
		/* assert(header == s->base) */
		error = -ENOMEM;
		if (s->base == NULL)
			goto cleanup;
764
		header(s->base)->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
765 766 767 768 769 770 771
		header(s->base)->h_blocks = cpu_to_le32(1);
		header(s->base)->h_refcount = cpu_to_le32(1);
		s->first = ENTRY(header(s->base)+1);
		s->here = ENTRY(header(s->base)+1);
		s->end = s->base + sb->s_blocksize;
	}

772
	error = ext4_xattr_set_entry(i, s);
773 774 775 776 777
	if (error == -EIO)
		goto bad_block;
	if (error)
		goto cleanup;
	if (!IS_LAST_ENTRY(s->first))
778
		ext4_xattr_rehash(header(s->base), s->here);
779 780 781

inserted:
	if (!IS_LAST_ENTRY(s->first)) {
782
		new_bh = ext4_xattr_cache_find(inode, header(s->base), &ce);
783 784 785 786 787 788 789
		if (new_bh) {
			/* We found an identical block in the cache. */
			if (new_bh == bs->bh)
				ea_bdebug(new_bh, "keeping");
			else {
				/* The old block is released after updating
				   the inode. */
790 791
				error = dquot_alloc_block(inode, 1);
				if (error)
792
					goto cleanup;
793
				error = ext4_journal_get_write_access(handle,
794 795 796 797
								      new_bh);
				if (error)
					goto cleanup_dquot;
				lock_buffer(new_bh);
M
Marcin Slusarz 已提交
798
				le32_add_cpu(&BHDR(new_bh)->h_refcount, 1);
799 800 801
				ea_bdebug(new_bh, "reusing; refcount now=%d",
					le32_to_cpu(BHDR(new_bh)->h_refcount));
				unlock_buffer(new_bh);
802 803 804
				error = ext4_handle_dirty_metadata(handle,
								   inode,
								   new_bh);
805 806 807 808 809 810 811 812 813 814 815 816
				if (error)
					goto cleanup_dquot;
			}
			mb_cache_entry_release(ce);
			ce = NULL;
		} else if (bs->bh && s->base == bs->bh->b_data) {
			/* We were modifying this block in-place. */
			ea_bdebug(bs->bh, "keeping this block");
			new_bh = bs->bh;
			get_bh(new_bh);
		} else {
			/* We need to allocate a new block */
817 818 819
			ext4_fsblk_t goal, block;

			goal = ext4_group_first_block_no(sb,
820
						EXT4_I(inode)->i_block_group);
821 822 823 824 825 826

			/* non-extent files can't have physical blocks past 2^32 */
			if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
				goal = goal & EXT4_MAX_BLOCK_FILE_PHYS;

			block = ext4_new_meta_blocks(handle, inode,
T
Theodore Ts'o 已提交
827
						  goal, NULL, &error);
828 829
			if (error)
				goto cleanup;
830 831 832 833

			if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
				BUG_ON(block > EXT4_MAX_BLOCK_FILE_PHYS);

834 835 836 837 838
			ea_idebug(inode, "creating block %d", block);

			new_bh = sb_getblk(sb, block);
			if (!new_bh) {
getblk_failed:
839 840
				ext4_free_blocks(handle, inode, 0, block, 1,
						 EXT4_FREE_BLOCKS_METADATA);
841 842 843 844
				error = -EIO;
				goto cleanup;
			}
			lock_buffer(new_bh);
845
			error = ext4_journal_get_create_access(handle, new_bh);
846 847 848 849 850 851 852
			if (error) {
				unlock_buffer(new_bh);
				goto getblk_failed;
			}
			memcpy(new_bh->b_data, s->base, new_bh->b_size);
			set_buffer_uptodate(new_bh);
			unlock_buffer(new_bh);
853
			ext4_xattr_cache_insert(new_bh);
854 855
			error = ext4_handle_dirty_metadata(handle,
							   inode, new_bh);
856 857 858 859 860 861
			if (error)
				goto cleanup;
		}
	}

	/* Update the inode. */
862
	EXT4_I(inode)->i_file_acl = new_bh ? new_bh->b_blocknr : 0;
863 864 865

	/* Drop the previous xattr block. */
	if (bs->bh && bs->bh != new_bh)
866
		ext4_xattr_release_block(handle, inode, bs->bh);
867 868 869 870 871 872 873 874 875 876 877 878
	error = 0;

cleanup:
	if (ce)
		mb_cache_entry_release(ce);
	brelse(new_bh);
	if (!(bs->bh && s->base == bs->bh->b_data))
		kfree(s->base);

	return error;

cleanup_dquot:
879
	dquot_free_block(inode, 1);
880 881 882
	goto cleanup;

bad_block:
883 884
	ext4_error(inode->i_sb, "inode %lu: bad block %llu",
		   inode->i_ino, EXT4_I(inode)->i_file_acl);
885 886 887 888 889
	goto cleanup;

#undef header
}

890 891 892
struct ext4_xattr_ibody_find {
	struct ext4_xattr_search s;
	struct ext4_iloc iloc;
893 894 895
};

static int
896 897
ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i,
		      struct ext4_xattr_ibody_find *is)
898
{
899 900
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
901 902
	int error;

903
	if (EXT4_I(inode)->i_extra_isize == 0)
904
		return 0;
905
	raw_inode = ext4_raw_inode(&is->iloc);
906 907 908
	header = IHDR(inode, raw_inode);
	is->s.base = is->s.first = IFIRST(header);
	is->s.here = is->s.first;
909
	is->s.end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
910
	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
911
		error = ext4_xattr_check_names(IFIRST(header), is->s.end);
912 913 914
		if (error)
			return error;
		/* Find the named attribute. */
915
		error = ext4_xattr_find_entry(&is->s.here, i->name_index,
916 917 918 919 920 921 922 923 924 925
					      i->name, is->s.end -
					      (void *)is->s.base, 0);
		if (error && error != -ENODATA)
			return error;
		is->s.not_found = error;
	}
	return 0;
}

static int
926 927 928
ext4_xattr_ibody_set(handle_t *handle, struct inode *inode,
		     struct ext4_xattr_info *i,
		     struct ext4_xattr_ibody_find *is)
929
{
930 931
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_search *s = &is->s;
932 933
	int error;

934
	if (EXT4_I(inode)->i_extra_isize == 0)
935
		return -ENOSPC;
936
	error = ext4_xattr_set_entry(i, s);
937 938
	if (error)
		return error;
939
	header = IHDR(inode, ext4_raw_inode(&is->iloc));
940
	if (!IS_LAST_ENTRY(s->first)) {
941
		header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
942
		ext4_set_inode_state(inode, EXT4_STATE_XATTR);
943 944
	} else {
		header->h_magic = cpu_to_le32(0);
945
		ext4_clear_inode_state(inode, EXT4_STATE_XATTR);
946 947 948 949 950
	}
	return 0;
}

/*
951
 * ext4_xattr_set_handle()
952 953 954 955 956 957 958 959 960 961 962
 *
 * Create, replace or remove an extended attribute for this inode. Buffer
 * is NULL to remove an existing extended attribute, and non-NULL to
 * either replace an existing extended attribute, or create a new extended
 * attribute. The flags XATTR_REPLACE and XATTR_CREATE
 * specify that an extended attribute must exist and must not exist
 * previous to the call, respectively.
 *
 * Returns 0, or a negative error number on failure.
 */
int
963
ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
964 965 966
		      const char *name, const void *value, size_t value_len,
		      int flags)
{
967
	struct ext4_xattr_info i = {
968 969 970 971 972 973
		.name_index = name_index,
		.name = name,
		.value = value,
		.value_len = value_len,

	};
974
	struct ext4_xattr_ibody_find is = {
975 976
		.s = { .not_found = -ENODATA, },
	};
977
	struct ext4_xattr_block_find bs = {
978 979
		.s = { .not_found = -ENODATA, },
	};
K
Kalpak Shah 已提交
980
	unsigned long no_expand;
981 982 983 984 985 986
	int error;

	if (!name)
		return -EINVAL;
	if (strlen(name) > 255)
		return -ERANGE;
987
	down_write(&EXT4_I(inode)->xattr_sem);
988 989
	no_expand = ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND);
	ext4_set_inode_state(inode, EXT4_STATE_NO_EXPAND);
K
Kalpak Shah 已提交
990

991
	error = ext4_get_inode_loc(inode, &is.iloc);
992 993 994
	if (error)
		goto cleanup;

995 996 997 998
	error = ext4_journal_get_write_access(handle, is.iloc.bh);
	if (error)
		goto cleanup;

999
	if (ext4_test_inode_state(inode, EXT4_STATE_NEW)) {
1000 1001
		struct ext4_inode *raw_inode = ext4_raw_inode(&is.iloc);
		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
1002
		ext4_clear_inode_state(inode, EXT4_STATE_NEW);
1003 1004
	}

1005
	error = ext4_xattr_ibody_find(inode, &i, &is);
1006 1007 1008
	if (error)
		goto cleanup;
	if (is.s.not_found)
1009
		error = ext4_xattr_block_find(inode, &i, &bs);
1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025
	if (error)
		goto cleanup;
	if (is.s.not_found && bs.s.not_found) {
		error = -ENODATA;
		if (flags & XATTR_REPLACE)
			goto cleanup;
		error = 0;
		if (!value)
			goto cleanup;
	} else {
		error = -EEXIST;
		if (flags & XATTR_CREATE)
			goto cleanup;
	}
	if (!value) {
		if (!is.s.not_found)
1026
			error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1027
		else if (!bs.s.not_found)
1028
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1029
	} else {
1030
		error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1031 1032
		if (!error && !bs.s.not_found) {
			i.value = NULL;
1033
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1034
		} else if (error == -ENOSPC) {
1035 1036 1037 1038 1039
			if (EXT4_I(inode)->i_file_acl && !bs.s.base) {
				error = ext4_xattr_block_find(inode, &i, &bs);
				if (error)
					goto cleanup;
			}
1040
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1041 1042 1043 1044
			if (error)
				goto cleanup;
			if (!is.s.not_found) {
				i.value = NULL;
1045
				error = ext4_xattr_ibody_set(handle, inode, &i,
1046 1047 1048 1049 1050
							     &is);
			}
		}
	}
	if (!error) {
1051
		ext4_xattr_update_super_block(handle, inode->i_sb);
K
Kalpak Shah 已提交
1052
		inode->i_ctime = ext4_current_time(inode);
1053
		if (!value)
1054
			ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1055
		error = ext4_mark_iloc_dirty(handle, inode, &is.iloc);
1056
		/*
1057
		 * The bh is consumed by ext4_mark_iloc_dirty, even with
1058 1059 1060 1061
		 * error != 0.
		 */
		is.iloc.bh = NULL;
		if (IS_SYNC(inode))
1062
			ext4_handle_sync(handle);
1063 1064 1065 1066 1067
	}

cleanup:
	brelse(is.iloc.bh);
	brelse(bs.bh);
K
Kalpak Shah 已提交
1068
	if (no_expand == 0)
1069
		ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1070
	up_write(&EXT4_I(inode)->xattr_sem);
1071 1072 1073 1074
	return error;
}

/*
1075
 * ext4_xattr_set()
1076
 *
1077
 * Like ext4_xattr_set_handle, but start from an inode. This extended
1078 1079 1080 1081 1082
 * attribute modification is a filesystem transaction by itself.
 *
 * Returns 0, or a negative error number on failure.
 */
int
1083
ext4_xattr_set(struct inode *inode, int name_index, const char *name,
1084 1085 1086 1087 1088 1089
	       const void *value, size_t value_len, int flags)
{
	handle_t *handle;
	int error, retries = 0;

retry:
1090
	handle = ext4_journal_start(inode, EXT4_DATA_TRANS_BLOCKS(inode->i_sb));
1091 1092 1093 1094 1095
	if (IS_ERR(handle)) {
		error = PTR_ERR(handle);
	} else {
		int error2;

1096
		error = ext4_xattr_set_handle(handle, inode, name_index, name,
1097
					      value, value_len, flags);
1098
		error2 = ext4_journal_stop(handle);
1099
		if (error == -ENOSPC &&
1100
		    ext4_should_retry_alloc(inode->i_sb, &retries))
1101 1102 1103 1104 1105 1106 1107 1108
			goto retry;
		if (error == 0)
			error = error2;
	}

	return error;
}

1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150
/*
 * Shift the EA entries in the inode to create space for the increased
 * i_extra_isize.
 */
static void ext4_xattr_shift_entries(struct ext4_xattr_entry *entry,
				     int value_offs_shift, void *to,
				     void *from, size_t n, int blocksize)
{
	struct ext4_xattr_entry *last = entry;
	int new_offs;

	/* Adjust the value offsets of the entries */
	for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) {
		if (!last->e_value_block && last->e_value_size) {
			new_offs = le16_to_cpu(last->e_value_offs) +
							value_offs_shift;
			BUG_ON(new_offs + le32_to_cpu(last->e_value_size)
				 > blocksize);
			last->e_value_offs = cpu_to_le16(new_offs);
		}
	}
	/* Shift the entries by n bytes */
	memmove(to, from, n);
}

/*
 * Expand an inode by new_extra_isize bytes when EAs are present.
 * Returns 0 on success or negative error number on failure.
 */
int ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize,
			       struct ext4_inode *raw_inode, handle_t *handle)
{
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_entry *entry, *last, *first;
	struct buffer_head *bh = NULL;
	struct ext4_xattr_ibody_find *is = NULL;
	struct ext4_xattr_block_find *bs = NULL;
	char *buffer = NULL, *b_entry_name = NULL;
	size_t min_offs, free;
	int total_ino, total_blk;
	void *base, *start, *end;
	int extra_isize = 0, error = 0, tried_min_extra_isize = 0;
A
Aneesh Kumar K.V 已提交
1151
	int s_min_extra_isize = le16_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_min_extra_isize);
1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196

	down_write(&EXT4_I(inode)->xattr_sem);
retry:
	if (EXT4_I(inode)->i_extra_isize >= new_extra_isize) {
		up_write(&EXT4_I(inode)->xattr_sem);
		return 0;
	}

	header = IHDR(inode, raw_inode);
	entry = IFIRST(header);

	/*
	 * Check if enough free space is available in the inode to shift the
	 * entries ahead by new_extra_isize.
	 */

	base = start = entry;
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	min_offs = end - base;
	last = entry;
	total_ino = sizeof(struct ext4_xattr_ibody_header);

	free = ext4_xattr_free_space(last, &min_offs, base, &total_ino);
	if (free >= new_extra_isize) {
		entry = IFIRST(header);
		ext4_xattr_shift_entries(entry,	EXT4_I(inode)->i_extra_isize
				- new_extra_isize, (void *)raw_inode +
				EXT4_GOOD_OLD_INODE_SIZE + new_extra_isize,
				(void *)header, total_ino,
				inode->i_sb->s_blocksize);
		EXT4_I(inode)->i_extra_isize = new_extra_isize;
		error = 0;
		goto cleanup;
	}

	/*
	 * Enough free space isn't available in the inode, check if
	 * EA block can hold new_extra_isize bytes.
	 */
	if (EXT4_I(inode)->i_file_acl) {
		bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
		error = -EIO;
		if (!bh)
			goto cleanup;
		if (ext4_xattr_check_block(bh)) {
1197 1198
			ext4_error(inode->i_sb, "inode %lu: bad block %llu",
				   inode->i_ino, EXT4_I(inode)->i_file_acl);
1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302
			error = -EIO;
			goto cleanup;
		}
		base = BHDR(bh);
		first = BFIRST(bh);
		end = bh->b_data + bh->b_size;
		min_offs = end - base;
		free = ext4_xattr_free_space(first, &min_offs, base,
					     &total_blk);
		if (free < new_extra_isize) {
			if (!tried_min_extra_isize && s_min_extra_isize) {
				tried_min_extra_isize++;
				new_extra_isize = s_min_extra_isize;
				brelse(bh);
				goto retry;
			}
			error = -1;
			goto cleanup;
		}
	} else {
		free = inode->i_sb->s_blocksize;
	}

	while (new_extra_isize > 0) {
		size_t offs, size, entry_size;
		struct ext4_xattr_entry *small_entry = NULL;
		struct ext4_xattr_info i = {
			.value = NULL,
			.value_len = 0,
		};
		unsigned int total_size;  /* EA entry size + value size */
		unsigned int shift_bytes; /* No. of bytes to shift EAs by? */
		unsigned int min_total_size = ~0U;

		is = kzalloc(sizeof(struct ext4_xattr_ibody_find), GFP_NOFS);
		bs = kzalloc(sizeof(struct ext4_xattr_block_find), GFP_NOFS);
		if (!is || !bs) {
			error = -ENOMEM;
			goto cleanup;
		}

		is->s.not_found = -ENODATA;
		bs->s.not_found = -ENODATA;
		is->iloc.bh = NULL;
		bs->bh = NULL;

		last = IFIRST(header);
		/* Find the entry best suited to be pushed into EA block */
		entry = NULL;
		for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) {
			total_size =
			EXT4_XATTR_SIZE(le32_to_cpu(last->e_value_size)) +
					EXT4_XATTR_LEN(last->e_name_len);
			if (total_size <= free && total_size < min_total_size) {
				if (total_size < new_extra_isize) {
					small_entry = last;
				} else {
					entry = last;
					min_total_size = total_size;
				}
			}
		}

		if (entry == NULL) {
			if (small_entry) {
				entry = small_entry;
			} else {
				if (!tried_min_extra_isize &&
				    s_min_extra_isize) {
					tried_min_extra_isize++;
					new_extra_isize = s_min_extra_isize;
					goto retry;
				}
				error = -1;
				goto cleanup;
			}
		}
		offs = le16_to_cpu(entry->e_value_offs);
		size = le32_to_cpu(entry->e_value_size);
		entry_size = EXT4_XATTR_LEN(entry->e_name_len);
		i.name_index = entry->e_name_index,
		buffer = kmalloc(EXT4_XATTR_SIZE(size), GFP_NOFS);
		b_entry_name = kmalloc(entry->e_name_len + 1, GFP_NOFS);
		if (!buffer || !b_entry_name) {
			error = -ENOMEM;
			goto cleanup;
		}
		/* Save the entry name and the entry value */
		memcpy(buffer, (void *)IFIRST(header) + offs,
		       EXT4_XATTR_SIZE(size));
		memcpy(b_entry_name, entry->e_name, entry->e_name_len);
		b_entry_name[entry->e_name_len] = '\0';
		i.name = b_entry_name;

		error = ext4_get_inode_loc(inode, &is->iloc);
		if (error)
			goto cleanup;

		error = ext4_xattr_ibody_find(inode, &i, is);
		if (error)
			goto cleanup;

		/* Remove the chosen entry from the inode */
		error = ext4_xattr_ibody_set(handle, inode, &i, is);
1303 1304
		if (error)
			goto cleanup;
1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323

		entry = IFIRST(header);
		if (entry_size + EXT4_XATTR_SIZE(size) >= new_extra_isize)
			shift_bytes = new_extra_isize;
		else
			shift_bytes = entry_size + size;
		/* Adjust the offsets and shift the remaining entries ahead */
		ext4_xattr_shift_entries(entry, EXT4_I(inode)->i_extra_isize -
			shift_bytes, (void *)raw_inode +
			EXT4_GOOD_OLD_INODE_SIZE + extra_isize + shift_bytes,
			(void *)header, total_ino - entry_size,
			inode->i_sb->s_blocksize);

		extra_isize += shift_bytes;
		new_extra_isize -= shift_bytes;
		EXT4_I(inode)->i_extra_isize = extra_isize;

		i.name = b_entry_name;
		i.value = buffer;
A
Aneesh Kumar K.V 已提交
1324
		i.value_len = size;
1325 1326 1327 1328 1329 1330 1331 1332 1333 1334
		error = ext4_xattr_block_find(inode, &i, bs);
		if (error)
			goto cleanup;

		/* Add entry which was removed from the inode into the block */
		error = ext4_xattr_block_set(handle, inode, &i, bs);
		if (error)
			goto cleanup;
		kfree(b_entry_name);
		kfree(buffer);
1335 1336
		b_entry_name = NULL;
		buffer = NULL;
1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358
		brelse(is->iloc.bh);
		kfree(is);
		kfree(bs);
	}
	brelse(bh);
	up_write(&EXT4_I(inode)->xattr_sem);
	return 0;

cleanup:
	kfree(b_entry_name);
	kfree(buffer);
	if (is)
		brelse(is->iloc.bh);
	kfree(is);
	kfree(bs);
	brelse(bh);
	up_write(&EXT4_I(inode)->xattr_sem);
	return error;
}



1359
/*
1360
 * ext4_xattr_delete_inode()
1361 1362 1363 1364 1365 1366
 *
 * Free extended attribute resources associated with this inode. This
 * is called immediately before an inode is freed. We have exclusive
 * access to the inode.
 */
void
1367
ext4_xattr_delete_inode(handle_t *handle, struct inode *inode)
1368 1369 1370
{
	struct buffer_head *bh = NULL;

1371
	if (!EXT4_I(inode)->i_file_acl)
1372
		goto cleanup;
1373
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
1374
	if (!bh) {
1375 1376
		ext4_error(inode->i_sb, "inode %lu: block %llu read error",
			   inode->i_ino, EXT4_I(inode)->i_file_acl);
1377 1378
		goto cleanup;
	}
1379
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
1380
	    BHDR(bh)->h_blocks != cpu_to_le32(1)) {
1381 1382
		ext4_error(inode->i_sb, "inode %lu: bad block %llu",
			   inode->i_ino, EXT4_I(inode)->i_file_acl);
1383 1384
		goto cleanup;
	}
1385 1386
	ext4_xattr_release_block(handle, inode, bh);
	EXT4_I(inode)->i_file_acl = 0;
1387 1388 1389 1390 1391 1392

cleanup:
	brelse(bh);
}

/*
1393
 * ext4_xattr_put_super()
1394 1395 1396 1397
 *
 * This is called when a file system is unmounted.
 */
void
1398
ext4_xattr_put_super(struct super_block *sb)
1399 1400 1401 1402 1403
{
	mb_cache_shrink(sb->s_bdev);
}

/*
1404
 * ext4_xattr_cache_insert()
1405 1406 1407 1408 1409 1410 1411
 *
 * Create a new entry in the extended attribute cache, and insert
 * it unless such an entry is already in the cache.
 *
 * Returns 0, or a negative error number on failure.
 */
static void
1412
ext4_xattr_cache_insert(struct buffer_head *bh)
1413 1414 1415 1416 1417
{
	__u32 hash = le32_to_cpu(BHDR(bh)->h_hash);
	struct mb_cache_entry *ce;
	int error;

1418
	ce = mb_cache_entry_alloc(ext4_xattr_cache, GFP_NOFS);
1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436
	if (!ce) {
		ea_bdebug(bh, "out of memory");
		return;
	}
	error = mb_cache_entry_insert(ce, bh->b_bdev, bh->b_blocknr, &hash);
	if (error) {
		mb_cache_entry_free(ce);
		if (error == -EBUSY) {
			ea_bdebug(bh, "already in cache");
			error = 0;
		}
	} else {
		ea_bdebug(bh, "inserting [%x]", (int)hash);
		mb_cache_entry_release(ce);
	}
}

/*
1437
 * ext4_xattr_cmp()
1438 1439 1440 1441 1442 1443 1444
 *
 * Compare two extended attribute blocks for equality.
 *
 * Returns 0 if the blocks are equal, 1 if they differ, and
 * a negative error number on errors.
 */
static int
1445 1446
ext4_xattr_cmp(struct ext4_xattr_header *header1,
	       struct ext4_xattr_header *header2)
1447
{
1448
	struct ext4_xattr_entry *entry1, *entry2;
1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467

	entry1 = ENTRY(header1+1);
	entry2 = ENTRY(header2+1);
	while (!IS_LAST_ENTRY(entry1)) {
		if (IS_LAST_ENTRY(entry2))
			return 1;
		if (entry1->e_hash != entry2->e_hash ||
		    entry1->e_name_index != entry2->e_name_index ||
		    entry1->e_name_len != entry2->e_name_len ||
		    entry1->e_value_size != entry2->e_value_size ||
		    memcmp(entry1->e_name, entry2->e_name, entry1->e_name_len))
			return 1;
		if (entry1->e_value_block != 0 || entry2->e_value_block != 0)
			return -EIO;
		if (memcmp((char *)header1 + le16_to_cpu(entry1->e_value_offs),
			   (char *)header2 + le16_to_cpu(entry2->e_value_offs),
			   le32_to_cpu(entry1->e_value_size)))
			return 1;

1468 1469
		entry1 = EXT4_XATTR_NEXT(entry1);
		entry2 = EXT4_XATTR_NEXT(entry2);
1470 1471 1472 1473 1474 1475 1476
	}
	if (!IS_LAST_ENTRY(entry2))
		return 1;
	return 0;
}

/*
1477
 * ext4_xattr_cache_find()
1478 1479 1480 1481 1482 1483 1484
 *
 * Find an identical extended attribute block.
 *
 * Returns a pointer to the block found, or NULL if such a block was
 * not found or an error occurred.
 */
static struct buffer_head *
1485
ext4_xattr_cache_find(struct inode *inode, struct ext4_xattr_header *header,
1486 1487 1488 1489 1490 1491 1492 1493 1494
		      struct mb_cache_entry **pce)
{
	__u32 hash = le32_to_cpu(header->h_hash);
	struct mb_cache_entry *ce;

	if (!header->h_hash)
		return NULL;  /* never share */
	ea_idebug(inode, "looking for cached blocks [%x]", (int)hash);
again:
1495
	ce = mb_cache_entry_find_first(ext4_xattr_cache, 0,
1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506
				       inode->i_sb->s_bdev, hash);
	while (ce) {
		struct buffer_head *bh;

		if (IS_ERR(ce)) {
			if (PTR_ERR(ce) == -EAGAIN)
				goto again;
			break;
		}
		bh = sb_bread(inode->i_sb, ce->e_block);
		if (!bh) {
1507
			ext4_error(inode->i_sb,
1508 1509 1510
				"inode %lu: block %lu read error",
				inode->i_ino, (unsigned long) ce->e_block);
		} else if (le32_to_cpu(BHDR(bh)->h_refcount) >=
1511
				EXT4_XATTR_REFCOUNT_MAX) {
1512 1513 1514
			ea_idebug(inode, "block %lu refcount %d>=%d",
				  (unsigned long) ce->e_block,
				  le32_to_cpu(BHDR(bh)->h_refcount),
1515 1516
					  EXT4_XATTR_REFCOUNT_MAX);
		} else if (ext4_xattr_cmp(header, BHDR(bh)) == 0) {
1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529
			*pce = ce;
			return bh;
		}
		brelse(bh);
		ce = mb_cache_entry_find_next(ce, 0, inode->i_sb->s_bdev, hash);
	}
	return NULL;
}

#define NAME_HASH_SHIFT 5
#define VALUE_HASH_SHIFT 16

/*
1530
 * ext4_xattr_hash_entry()
1531 1532 1533
 *
 * Compute the hash of an extended attribute.
 */
1534 1535
static inline void ext4_xattr_hash_entry(struct ext4_xattr_header *header,
					 struct ext4_xattr_entry *entry)
1536 1537 1538 1539 1540
{
	__u32 hash = 0;
	char *name = entry->e_name;
	int n;

1541
	for (n = 0; n < entry->e_name_len; n++) {
1542 1543 1544 1545 1546 1547 1548 1549 1550
		hash = (hash << NAME_HASH_SHIFT) ^
		       (hash >> (8*sizeof(hash) - NAME_HASH_SHIFT)) ^
		       *name++;
	}

	if (entry->e_value_block == 0 && entry->e_value_size != 0) {
		__le32 *value = (__le32 *)((char *)header +
			le16_to_cpu(entry->e_value_offs));
		for (n = (le32_to_cpu(entry->e_value_size) +
1551
		     EXT4_XATTR_ROUND) >> EXT4_XATTR_PAD_BITS; n; n--) {
1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565
			hash = (hash << VALUE_HASH_SHIFT) ^
			       (hash >> (8*sizeof(hash) - VALUE_HASH_SHIFT)) ^
			       le32_to_cpu(*value++);
		}
	}
	entry->e_hash = cpu_to_le32(hash);
}

#undef NAME_HASH_SHIFT
#undef VALUE_HASH_SHIFT

#define BLOCK_HASH_SHIFT 16

/*
1566
 * ext4_xattr_rehash()
1567 1568 1569
 *
 * Re-compute the extended attribute hash value after an entry has changed.
 */
1570 1571
static void ext4_xattr_rehash(struct ext4_xattr_header *header,
			      struct ext4_xattr_entry *entry)
1572
{
1573
	struct ext4_xattr_entry *here;
1574 1575
	__u32 hash = 0;

1576
	ext4_xattr_hash_entry(header, entry);
1577 1578 1579 1580 1581 1582 1583 1584 1585 1586
	here = ENTRY(header+1);
	while (!IS_LAST_ENTRY(here)) {
		if (!here->e_hash) {
			/* Block is not shared if an entry's hash value == 0 */
			hash = 0;
			break;
		}
		hash = (hash << BLOCK_HASH_SHIFT) ^
		       (hash >> (8*sizeof(hash) - BLOCK_HASH_SHIFT)) ^
		       le32_to_cpu(here->e_hash);
1587
		here = EXT4_XATTR_NEXT(here);
1588 1589 1590 1591 1592 1593 1594
	}
	header->h_hash = cpu_to_le32(hash);
}

#undef BLOCK_HASH_SHIFT

int __init
1595
init_ext4_xattr(void)
1596
{
1597
	ext4_xattr_cache = mb_cache_create("ext4_xattr", NULL,
1598 1599
		sizeof(struct mb_cache_entry) +
		sizeof(((struct mb_cache_entry *) 0)->e_indexes[0]), 1, 6);
1600
	if (!ext4_xattr_cache)
1601 1602 1603 1604 1605
		return -ENOMEM;
	return 0;
}

void
1606
exit_ext4_xattr(void)
1607
{
1608 1609 1610
	if (ext4_xattr_cache)
		mb_cache_destroy(ext4_xattr_cache);
	ext4_xattr_cache = NULL;
1611
}