xattr.c 43.8 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
# 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
85 86
# define ea_idebug(inode, fmt, ...)	no_printk(fmt, ##__VA_ARGS__)
# define ea_bdebug(bh, fmt, ...)	no_printk(fmt, ##__VA_ARGS__)
87 88
#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

S
Stephen Hemminger 已提交
100
static const struct xattr_handler *ext4_xattr_handler_map[] = {
101
	[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
};

S
Stephen Hemminger 已提交
112
const struct xattr_handler *ext4_xattr_handlers[] = {
113 114
	&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
#endif
	NULL
};

125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
static __le32 ext4_xattr_block_csum(struct inode *inode,
				    sector_t block_nr,
				    struct ext4_xattr_header *hdr)
{
	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
	struct ext4_inode_info *ei = EXT4_I(inode);
	__u32 csum, old;

	old = hdr->h_checksum;
	hdr->h_checksum = 0;
	if (le32_to_cpu(hdr->h_refcount) != 1) {
		block_nr = cpu_to_le64(block_nr);
		csum = ext4_chksum(sbi, sbi->s_csum_seed, (__u8 *)&block_nr,
				   sizeof(block_nr));
	} else
		csum = ei->i_csum_seed;
	csum = ext4_chksum(sbi, csum, (__u8 *)hdr,
			   EXT4_BLOCK_SIZE(inode->i_sb));
	hdr->h_checksum = old;
	return cpu_to_le32(csum);
}

static int ext4_xattr_block_csum_verify(struct inode *inode,
					sector_t block_nr,
					struct ext4_xattr_header *hdr)
{
	if (EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb,
		EXT4_FEATURE_RO_COMPAT_METADATA_CSUM) &&
	    (hdr->h_checksum != ext4_xattr_block_csum(inode, block_nr, hdr)))
		return 0;
	return 1;
}

static void ext4_xattr_block_csum_set(struct inode *inode,
				      sector_t block_nr,
				      struct ext4_xattr_header *hdr)
{
	if (!EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb,
		EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
		return;

	hdr->h_checksum = ext4_xattr_block_csum(inode, block_nr, hdr);
}

static inline int ext4_handle_dirty_xattr_block(handle_t *handle,
						struct inode *inode,
						struct buffer_head *bh)
{
	ext4_xattr_block_csum_set(inode, bh->b_blocknr, BHDR(bh));
	return ext4_handle_dirty_metadata(handle, inode, bh);
}

S
Stephen Hemminger 已提交
177
static inline const struct xattr_handler *
178
ext4_xattr_handler(int name_index)
179
{
S
Stephen Hemminger 已提交
180
	const struct xattr_handler *handler = NULL;
181

182 183
	if (name_index > 0 && name_index < ARRAY_SIZE(ext4_xattr_handler_map))
		handler = ext4_xattr_handler_map[name_index];
184 185 186 187 188 189 190 191 192
	return handler;
}

/*
 * Inode operation listxattr()
 *
 * dentry->d_inode->i_mutex: don't care
 */
ssize_t
193
ext4_listxattr(struct dentry *dentry, char *buffer, size_t size)
194
{
195
	return ext4_xattr_list(dentry, buffer, size);
196 197 198
}

static int
199
ext4_xattr_check_names(struct ext4_xattr_entry *entry, void *end)
200 201
{
	while (!IS_LAST_ENTRY(entry)) {
202
		struct ext4_xattr_entry *next = EXT4_XATTR_NEXT(entry);
203 204 205 206 207 208 209 210
		if ((void *)next >= end)
			return -EIO;
		entry = next;
	}
	return 0;
}

static inline int
211
ext4_xattr_check_block(struct inode *inode, struct buffer_head *bh)
212
{
213 214 215 216 217
	int error;

	if (buffer_verified(bh))
		return 0;

218
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
219 220
	    BHDR(bh)->h_blocks != cpu_to_le32(1))
		return -EIO;
221 222 223 224 225 226
	if (!ext4_xattr_block_csum_verify(inode, bh->b_blocknr, BHDR(bh)))
		return -EIO;
	error = ext4_xattr_check_names(BFIRST(bh), bh->b_data + bh->b_size);
	if (!error)
		set_buffer_verified(bh);
	return error;
227 228 229
}

static inline int
230
ext4_xattr_check_entry(struct ext4_xattr_entry *entry, size_t size)
231 232 233 234 235 236 237 238 239 240
{
	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
241
ext4_xattr_find_entry(struct ext4_xattr_entry **pentry, int name_index,
242 243
		      const char *name, size_t size, int sorted)
{
244
	struct ext4_xattr_entry *entry;
245 246 247 248 249 250 251
	size_t name_len;
	int cmp = 1;

	if (name == NULL)
		return -EINVAL;
	name_len = strlen(name);
	entry = *pentry;
252
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
253 254 255 256 257 258 259 260 261
		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;
262
	if (!cmp && ext4_xattr_check_entry(entry, size))
263 264 265 266 267
			return -EIO;
	return cmp ? -ENODATA : 0;
}

static int
268
ext4_xattr_block_get(struct inode *inode, int name_index, const char *name,
269 270 271
		     void *buffer, size_t buffer_size)
{
	struct buffer_head *bh = NULL;
272
	struct ext4_xattr_entry *entry;
273 274 275 276 277 278 279
	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;
280
	if (!EXT4_I(inode)->i_file_acl)
281
		goto cleanup;
282 283
	ea_idebug(inode, "reading block %llu",
		  (unsigned long long)EXT4_I(inode)->i_file_acl);
284
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
285 286 287 288
	if (!bh)
		goto cleanup;
	ea_bdebug(bh, "b_count=%d, refcount=%d",
		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
289
	if (ext4_xattr_check_block(inode, bh)) {
290
bad_block:
291 292
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
293 294 295
		error = -EIO;
		goto cleanup;
	}
296
	ext4_xattr_cache_insert(bh);
297
	entry = BFIRST(bh);
298
	error = ext4_xattr_find_entry(&entry, name_index, name, bh->b_size, 1);
299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
	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
319
ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name,
320 321
		     void *buffer, size_t buffer_size)
{
322 323 324 325
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_entry *entry;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
326 327 328 329
	size_t size;
	void *end;
	int error;

330
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
331
		return -ENODATA;
332
	error = ext4_get_inode_loc(inode, &iloc);
333 334
	if (error)
		return error;
335
	raw_inode = ext4_raw_inode(&iloc);
336 337
	header = IHDR(inode, raw_inode);
	entry = IFIRST(header);
338 339
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(entry, end);
340 341
	if (error)
		goto cleanup;
342
	error = ext4_xattr_find_entry(&entry, name_index, name,
343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361
				      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;
}

/*
362
 * ext4_xattr_get()
363 364 365 366 367 368 369 370 371
 *
 * 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
372
ext4_xattr_get(struct inode *inode, int name_index, const char *name,
373 374 375 376
	       void *buffer, size_t buffer_size)
{
	int error;

377 378
	down_read(&EXT4_I(inode)->xattr_sem);
	error = ext4_xattr_ibody_get(inode, name_index, name, buffer,
379 380
				     buffer_size);
	if (error == -ENODATA)
381
		error = ext4_xattr_block_get(inode, name_index, name, buffer,
382
					     buffer_size);
383
	up_read(&EXT4_I(inode)->xattr_sem);
384 385 386 387
	return error;
}

static int
388
ext4_xattr_list_entries(struct dentry *dentry, struct ext4_xattr_entry *entry,
389 390 391 392
			char *buffer, size_t buffer_size)
{
	size_t rest = buffer_size;

393
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
S
Stephen Hemminger 已提交
394
		const struct xattr_handler *handler =
395
			ext4_xattr_handler(entry->e_name_index);
396 397

		if (handler) {
398
			size_t size = handler->list(dentry, buffer, rest,
399
						    entry->e_name,
400 401
						    entry->e_name_len,
						    handler->flags);
402 403 404 405 406 407 408 409 410 411 412 413
			if (buffer) {
				if (size > rest)
					return -ERANGE;
				buffer += size;
			}
			rest -= size;
		}
	}
	return buffer_size - rest;
}

static int
414
ext4_xattr_block_list(struct dentry *dentry, char *buffer, size_t buffer_size)
415
{
416
	struct inode *inode = dentry->d_inode;
417 418 419 420 421 422 423
	struct buffer_head *bh = NULL;
	int error;

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

	error = 0;
424
	if (!EXT4_I(inode)->i_file_acl)
425
		goto cleanup;
426 427
	ea_idebug(inode, "reading block %llu",
		  (unsigned long long)EXT4_I(inode)->i_file_acl);
428
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
429 430 431 432 433
	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));
434
	if (ext4_xattr_check_block(inode, bh)) {
435 436
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
437 438 439
		error = -EIO;
		goto cleanup;
	}
440
	ext4_xattr_cache_insert(bh);
441
	error = ext4_xattr_list_entries(dentry, BFIRST(bh), buffer, buffer_size);
442 443 444 445 446 447 448 449

cleanup:
	brelse(bh);

	return error;
}

static int
450
ext4_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size)
451
{
452
	struct inode *inode = dentry->d_inode;
453 454 455
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
456 457 458
	void *end;
	int error;

459
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
460
		return 0;
461
	error = ext4_get_inode_loc(inode, &iloc);
462 463
	if (error)
		return error;
464
	raw_inode = ext4_raw_inode(&iloc);
465
	header = IHDR(inode, raw_inode);
466 467
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(IFIRST(header), end);
468 469
	if (error)
		goto cleanup;
470
	error = ext4_xattr_list_entries(dentry, IFIRST(header),
471 472 473 474 475 476 477 478
					buffer, buffer_size);

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

/*
479
 * ext4_xattr_list()
480 481 482 483 484 485 486 487
 *
 * 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.
 */
488
static int
489
ext4_xattr_list(struct dentry *dentry, char *buffer, size_t buffer_size)
490
{
491
	int ret, ret2;
492

493
	down_read(&EXT4_I(dentry->d_inode)->xattr_sem);
494 495 496 497 498 499
	ret = ret2 = ext4_xattr_ibody_list(dentry, buffer, buffer_size);
	if (ret < 0)
		goto errout;
	if (buffer) {
		buffer += ret;
		buffer_size -= ret;
500
	}
501 502 503 504 505
	ret = ext4_xattr_block_list(dentry, buffer, buffer_size);
	if (ret < 0)
		goto errout;
	ret += ret2;
errout:
506
	up_read(&EXT4_I(dentry->d_inode)->xattr_sem);
507
	return ret;
508 509 510
}

/*
511
 * If the EXT4_FEATURE_COMPAT_EXT_ATTR feature of this file system is
512 513
 * not set, set it.
 */
514
static void ext4_xattr_update_super_block(handle_t *handle,
515 516
					  struct super_block *sb)
{
517
	if (EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR))
518 519
		return;

520
	if (ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh) == 0) {
521
		EXT4_SET_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR);
T
Theodore Ts'o 已提交
522
		ext4_handle_dirty_super(handle, sb);
523 524 525 526 527 528 529 530
	}
}

/*
 * Release the xattr block BH: If the reference count is > 1, decrement
 * it; otherwise free the block.
 */
static void
531
ext4_xattr_release_block(handle_t *handle, struct inode *inode,
532 533 534
			 struct buffer_head *bh)
{
	struct mb_cache_entry *ce = NULL;
535
	int error = 0;
536

537
	ce = mb_cache_entry_get(ext4_xattr_cache, bh->b_bdev, bh->b_blocknr);
538 539 540 541 542
	error = ext4_journal_get_write_access(handle, bh);
	if (error)
		goto out;

	lock_buffer(bh);
543 544 545 546 547
	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);
548 549 550
		ext4_free_blocks(handle, inode, bh, 0, 1,
				 EXT4_FREE_BLOCKS_METADATA |
				 EXT4_FREE_BLOCKS_FORGET);
551
		unlock_buffer(bh);
552
	} else {
M
Marcin Slusarz 已提交
553
		le32_add_cpu(&BHDR(bh)->h_refcount, -1);
554 555 556
		if (ce)
			mb_cache_entry_release(ce);
		unlock_buffer(bh);
557
		error = ext4_handle_dirty_xattr_block(handle, inode, bh);
558
		if (IS_SYNC(inode))
559
			ext4_handle_sync(handle);
560
		dquot_free_block(inode, 1);
561 562
		ea_bdebug(bh, "refcount now=%d; releasing",
			  le32_to_cpu(BHDR(bh)->h_refcount));
563
	}
564 565 566
out:
	ext4_std_error(inode->i_sb, error);
	return;
567 568
}

569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
/*
 * 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));
}

587
struct ext4_xattr_info {
588 589 590 591 592 593
	int name_index;
	const char *name;
	const void *value;
	size_t value_len;
};

594 595
struct ext4_xattr_search {
	struct ext4_xattr_entry *first;
596 597
	void *base;
	void *end;
598
	struct ext4_xattr_entry *here;
599 600 601 602
	int not_found;
};

static int
603
ext4_xattr_set_entry(struct ext4_xattr_info *i, struct ext4_xattr_search *s)
604
{
605
	struct ext4_xattr_entry *last;
606 607 608 609
	size_t free, min_offs = s->end - s->base, name_len = strlen(i->name);

	/* Compute min_offs and last. */
	last = s->first;
610
	for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) {
611 612 613 614 615 616 617 618 619 620
		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);
621
			free += EXT4_XATTR_SIZE(size);
622
		}
623
		free += EXT4_XATTR_LEN(name_len);
624 625
	}
	if (i->value) {
626 627 628
		if (free < EXT4_XATTR_SIZE(i->value_len) ||
		    free < EXT4_XATTR_LEN(name_len) +
			   EXT4_XATTR_SIZE(i->value_len))
629 630 631 632 633
			return -ENOSPC;
	}

	if (i->value && s->not_found) {
		/* Insert the new name. */
634
		size_t size = EXT4_XATTR_LEN(name_len);
635 636 637 638 639 640 641 642 643 644 645
		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;
646
			size_t size = EXT4_XATTR_SIZE(
647 648
				le32_to_cpu(s->here->e_value_size));

649
			if (i->value && size == EXT4_XATTR_SIZE(i->value_len)) {
650 651 652 653
				/* The old and the new value have the same
				   size. Just replace. */
				s->here->e_value_size =
					cpu_to_le32(i->value_len);
654 655
				memset(val + size - EXT4_XATTR_PAD, 0,
				       EXT4_XATTR_PAD); /* Clear pad bytes. */
656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674
				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);
675
				last = EXT4_XATTR_NEXT(last);
676 677 678 679
			}
		}
		if (!i->value) {
			/* Remove the old name. */
680
			size_t size = EXT4_XATTR_LEN(name_len);
681 682 683 684 685 686 687 688 689 690 691
			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) {
692
			size_t size = EXT4_XATTR_SIZE(i->value_len);
693 694
			void *val = s->base + min_offs - size;
			s->here->e_value_offs = cpu_to_le16(min_offs - size);
695 696
			memset(val + size - EXT4_XATTR_PAD, 0,
			       EXT4_XATTR_PAD); /* Clear the pad bytes. */
697 698 699 700 701 702
			memcpy(val, i->value, i->value_len);
		}
	}
	return 0;
}

703 704
struct ext4_xattr_block_find {
	struct ext4_xattr_search s;
705 706 707 708
	struct buffer_head *bh;
};

static int
709 710
ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
		      struct ext4_xattr_block_find *bs)
711 712 713 714 715 716 717
{
	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);

718
	if (EXT4_I(inode)->i_file_acl) {
719
		/* The inode already has an extended attribute block. */
720
		bs->bh = sb_bread(sb, EXT4_I(inode)->i_file_acl);
721 722 723 724 725 726
		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));
727
		if (ext4_xattr_check_block(inode, bs->bh)) {
728 729
			EXT4_ERROR_INODE(inode, "bad block %llu",
					 EXT4_I(inode)->i_file_acl);
730 731 732 733 734 735 736 737
			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;
738
		error = ext4_xattr_find_entry(&bs->s.here, i->name_index,
739 740 741 742 743 744 745 746 747 748 749 750
					      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
751 752 753
ext4_xattr_block_set(handle_t *handle, struct inode *inode,
		     struct ext4_xattr_info *i,
		     struct ext4_xattr_block_find *bs)
754 755 756
{
	struct super_block *sb = inode->i_sb;
	struct buffer_head *new_bh = NULL;
757
	struct ext4_xattr_search *s = &bs->s;
758
	struct mb_cache_entry *ce = NULL;
759
	int error = 0;
760

761
#define header(x) ((struct ext4_xattr_header *)(x))
762 763 764 765

	if (i->value && i->value_len > sb->s_blocksize)
		return -ENOSPC;
	if (s->base) {
766
		ce = mb_cache_entry_get(ext4_xattr_cache, bs->bh->b_bdev,
767
					bs->bh->b_blocknr);
768 769 770 771 772
		error = ext4_journal_get_write_access(handle, bs->bh);
		if (error)
			goto cleanup;
		lock_buffer(bs->bh);

773 774 775 776 777 778
		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");
779
			error = ext4_xattr_set_entry(i, s);
780 781
			if (!error) {
				if (!IS_LAST_ENTRY(s->first))
782
					ext4_xattr_rehash(header(s->base),
783
							  s->here);
784
				ext4_xattr_cache_insert(bs->bh);
785 786 787 788 789
			}
			unlock_buffer(bs->bh);
			if (error == -EIO)
				goto bad_block;
			if (!error)
790 791 792
				error = ext4_handle_dirty_xattr_block(handle,
								      inode,
								      bs->bh);
793 794 795 796 797 798
			if (error)
				goto cleanup;
			goto inserted;
		} else {
			int offset = (char *)s->here - bs->bh->b_data;

799
			unlock_buffer(bs->bh);
800
			ext4_handle_release_buffer(handle, bs->bh);
801 802 803 804 805
			if (ce) {
				mb_cache_entry_release(ce);
				ce = NULL;
			}
			ea_bdebug(bs->bh, "cloning");
806
			s->base = kmalloc(bs->bh->b_size, GFP_NOFS);
807 808 809 810 811 812 813 814 815 816 817
			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. */
818
		s->base = kzalloc(sb->s_blocksize, GFP_NOFS);
819 820 821 822
		/* assert(header == s->base) */
		error = -ENOMEM;
		if (s->base == NULL)
			goto cleanup;
823
		header(s->base)->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
824 825 826 827 828 829 830
		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;
	}

831
	error = ext4_xattr_set_entry(i, s);
832 833 834 835 836
	if (error == -EIO)
		goto bad_block;
	if (error)
		goto cleanup;
	if (!IS_LAST_ENTRY(s->first))
837
		ext4_xattr_rehash(header(s->base), s->here);
838 839 840

inserted:
	if (!IS_LAST_ENTRY(s->first)) {
841
		new_bh = ext4_xattr_cache_find(inode, header(s->base), &ce);
842 843 844 845 846 847 848
		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. */
849 850
				error = dquot_alloc_block(inode, 1);
				if (error)
851
					goto cleanup;
852
				error = ext4_journal_get_write_access(handle,
853 854 855 856
								      new_bh);
				if (error)
					goto cleanup_dquot;
				lock_buffer(new_bh);
M
Marcin Slusarz 已提交
857
				le32_add_cpu(&BHDR(new_bh)->h_refcount, 1);
858 859 860
				ea_bdebug(new_bh, "reusing; refcount now=%d",
					le32_to_cpu(BHDR(new_bh)->h_refcount));
				unlock_buffer(new_bh);
861 862 863
				error = ext4_handle_dirty_xattr_block(handle,
								      inode,
								      new_bh);
864 865 866 867 868 869 870 871 872 873 874 875
				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 */
876 877 878
			ext4_fsblk_t goal, block;

			goal = ext4_group_first_block_no(sb,
879
						EXT4_I(inode)->i_block_group);
880 881

			/* non-extent files can't have physical blocks past 2^32 */
882
			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
883 884
				goal = goal & EXT4_MAX_BLOCK_FILE_PHYS;

885 886 887 888 889
			/*
			 * take i_data_sem because we will test
			 * i_delalloc_reserved_flag in ext4_mb_new_blocks
			 */
			down_read((&EXT4_I(inode)->i_data_sem));
890 891
			block = ext4_new_meta_blocks(handle, inode, goal, 0,
						     NULL, &error);
892
			up_read((&EXT4_I(inode)->i_data_sem));
893 894
			if (error)
				goto cleanup;
895

896
			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
897 898
				BUG_ON(block > EXT4_MAX_BLOCK_FILE_PHYS);

899 900
			ea_idebug(inode, "creating block %llu",
				  (unsigned long long)block);
901 902 903 904

			new_bh = sb_getblk(sb, block);
			if (!new_bh) {
getblk_failed:
905
				ext4_free_blocks(handle, inode, NULL, block, 1,
906
						 EXT4_FREE_BLOCKS_METADATA);
907 908 909 910
				error = -EIO;
				goto cleanup;
			}
			lock_buffer(new_bh);
911
			error = ext4_journal_get_create_access(handle, new_bh);
912 913 914 915 916 917 918
			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);
919
			ext4_xattr_cache_insert(new_bh);
920 921
			error = ext4_handle_dirty_xattr_block(handle,
							      inode, new_bh);
922 923 924 925 926 927
			if (error)
				goto cleanup;
		}
	}

	/* Update the inode. */
928
	EXT4_I(inode)->i_file_acl = new_bh ? new_bh->b_blocknr : 0;
929 930 931

	/* Drop the previous xattr block. */
	if (bs->bh && bs->bh != new_bh)
932
		ext4_xattr_release_block(handle, inode, bs->bh);
933 934 935 936 937 938 939 940 941 942 943 944
	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:
945
	dquot_free_block(inode, 1);
946 947 948
	goto cleanup;

bad_block:
949 950
	EXT4_ERROR_INODE(inode, "bad block %llu",
			 EXT4_I(inode)->i_file_acl);
951 952 953 954 955
	goto cleanup;

#undef header
}

956 957 958
struct ext4_xattr_ibody_find {
	struct ext4_xattr_search s;
	struct ext4_iloc iloc;
959 960 961
};

static int
962 963
ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i,
		      struct ext4_xattr_ibody_find *is)
964
{
965 966
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
967 968
	int error;

969
	if (EXT4_I(inode)->i_extra_isize == 0)
970
		return 0;
971
	raw_inode = ext4_raw_inode(&is->iloc);
972 973 974
	header = IHDR(inode, raw_inode);
	is->s.base = is->s.first = IFIRST(header);
	is->s.here = is->s.first;
975
	is->s.end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
976
	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
977
		error = ext4_xattr_check_names(IFIRST(header), is->s.end);
978 979 980
		if (error)
			return error;
		/* Find the named attribute. */
981
		error = ext4_xattr_find_entry(&is->s.here, i->name_index,
982 983 984 985 986 987 988 989 990 991
					      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
992 993 994
ext4_xattr_ibody_set(handle_t *handle, struct inode *inode,
		     struct ext4_xattr_info *i,
		     struct ext4_xattr_ibody_find *is)
995
{
996 997
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_search *s = &is->s;
998 999
	int error;

1000
	if (EXT4_I(inode)->i_extra_isize == 0)
1001
		return -ENOSPC;
1002
	error = ext4_xattr_set_entry(i, s);
1003 1004
	if (error)
		return error;
1005
	header = IHDR(inode, ext4_raw_inode(&is->iloc));
1006
	if (!IS_LAST_ENTRY(s->first)) {
1007
		header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
1008
		ext4_set_inode_state(inode, EXT4_STATE_XATTR);
1009 1010
	} else {
		header->h_magic = cpu_to_le32(0);
1011
		ext4_clear_inode_state(inode, EXT4_STATE_XATTR);
1012 1013 1014 1015 1016
	}
	return 0;
}

/*
1017
 * ext4_xattr_set_handle()
1018
 *
1019
 * Create, replace or remove an extended attribute for this inode.  Value
1020 1021 1022 1023 1024 1025 1026 1027 1028
 * 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
1029
ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
1030 1031 1032
		      const char *name, const void *value, size_t value_len,
		      int flags)
{
1033
	struct ext4_xattr_info i = {
1034 1035 1036 1037 1038 1039
		.name_index = name_index,
		.name = name,
		.value = value,
		.value_len = value_len,

	};
1040
	struct ext4_xattr_ibody_find is = {
1041 1042
		.s = { .not_found = -ENODATA, },
	};
1043
	struct ext4_xattr_block_find bs = {
1044 1045
		.s = { .not_found = -ENODATA, },
	};
K
Kalpak Shah 已提交
1046
	unsigned long no_expand;
1047 1048 1049 1050 1051 1052
	int error;

	if (!name)
		return -EINVAL;
	if (strlen(name) > 255)
		return -ERANGE;
1053
	down_write(&EXT4_I(inode)->xattr_sem);
1054 1055
	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
1056

1057
	error = ext4_reserve_inode_write(handle, inode, &is.iloc);
1058 1059 1060
	if (error)
		goto cleanup;

1061
	if (ext4_test_inode_state(inode, EXT4_STATE_NEW)) {
1062 1063
		struct ext4_inode *raw_inode = ext4_raw_inode(&is.iloc);
		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
1064
		ext4_clear_inode_state(inode, EXT4_STATE_NEW);
1065 1066
	}

1067
	error = ext4_xattr_ibody_find(inode, &i, &is);
1068 1069 1070
	if (error)
		goto cleanup;
	if (is.s.not_found)
1071
		error = ext4_xattr_block_find(inode, &i, &bs);
1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087
	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)
1088
			error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1089
		else if (!bs.s.not_found)
1090
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1091
	} else {
1092
		error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1093 1094
		if (!error && !bs.s.not_found) {
			i.value = NULL;
1095
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1096
		} else if (error == -ENOSPC) {
1097 1098 1099 1100 1101
			if (EXT4_I(inode)->i_file_acl && !bs.s.base) {
				error = ext4_xattr_block_find(inode, &i, &bs);
				if (error)
					goto cleanup;
			}
1102
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1103 1104 1105 1106
			if (error)
				goto cleanup;
			if (!is.s.not_found) {
				i.value = NULL;
1107
				error = ext4_xattr_ibody_set(handle, inode, &i,
1108 1109 1110 1111 1112
							     &is);
			}
		}
	}
	if (!error) {
1113
		ext4_xattr_update_super_block(handle, inode->i_sb);
K
Kalpak Shah 已提交
1114
		inode->i_ctime = ext4_current_time(inode);
1115
		if (!value)
1116
			ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1117
		error = ext4_mark_iloc_dirty(handle, inode, &is.iloc);
1118
		/*
1119
		 * The bh is consumed by ext4_mark_iloc_dirty, even with
1120 1121 1122 1123
		 * error != 0.
		 */
		is.iloc.bh = NULL;
		if (IS_SYNC(inode))
1124
			ext4_handle_sync(handle);
1125 1126 1127 1128 1129
	}

cleanup:
	brelse(is.iloc.bh);
	brelse(bs.bh);
K
Kalpak Shah 已提交
1130
	if (no_expand == 0)
1131
		ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1132
	up_write(&EXT4_I(inode)->xattr_sem);
1133 1134 1135 1136
	return error;
}

/*
1137
 * ext4_xattr_set()
1138
 *
1139
 * Like ext4_xattr_set_handle, but start from an inode. This extended
1140 1141 1142 1143 1144
 * attribute modification is a filesystem transaction by itself.
 *
 * Returns 0, or a negative error number on failure.
 */
int
1145
ext4_xattr_set(struct inode *inode, int name_index, const char *name,
1146 1147 1148 1149 1150 1151
	       const void *value, size_t value_len, int flags)
{
	handle_t *handle;
	int error, retries = 0;

retry:
1152
	handle = ext4_journal_start(inode, EXT4_DATA_TRANS_BLOCKS(inode->i_sb));
1153 1154 1155 1156 1157
	if (IS_ERR(handle)) {
		error = PTR_ERR(handle);
	} else {
		int error2;

1158
		error = ext4_xattr_set_handle(handle, inode, name_index, name,
1159
					      value, value_len, flags);
1160
		error2 = ext4_journal_stop(handle);
1161
		if (error == -ENOSPC &&
1162
		    ext4_should_retry_alloc(inode->i_sb, &retries))
1163 1164 1165 1166 1167 1168 1169 1170
			goto retry;
		if (error == 0)
			error = error2;
	}

	return error;
}

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 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
/*
 * 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
1213
	int s_min_extra_isize = le16_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_min_extra_isize);
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

	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;
1258
		if (ext4_xattr_check_block(inode, bh)) {
1259 1260
			EXT4_ERROR_INODE(inode, "bad block %llu",
					 EXT4_I(inode)->i_file_acl);
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 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364
			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);
1365 1366
		if (error)
			goto cleanup;
1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385

		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
1386
		i.value_len = size;
1387 1388 1389 1390 1391 1392 1393 1394 1395 1396
		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);
1397 1398
		b_entry_name = NULL;
		buffer = NULL;
1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420
		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;
}



1421
/*
1422
 * ext4_xattr_delete_inode()
1423 1424 1425 1426 1427 1428
 *
 * 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
1429
ext4_xattr_delete_inode(handle_t *handle, struct inode *inode)
1430 1431 1432
{
	struct buffer_head *bh = NULL;

1433
	if (!EXT4_I(inode)->i_file_acl)
1434
		goto cleanup;
1435
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
1436
	if (!bh) {
1437 1438
		EXT4_ERROR_INODE(inode, "block %llu read error",
				 EXT4_I(inode)->i_file_acl);
1439 1440
		goto cleanup;
	}
1441
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
1442
	    BHDR(bh)->h_blocks != cpu_to_le32(1)) {
1443 1444
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
1445 1446
		goto cleanup;
	}
1447 1448
	ext4_xattr_release_block(handle, inode, bh);
	EXT4_I(inode)->i_file_acl = 0;
1449 1450 1451 1452 1453 1454

cleanup:
	brelse(bh);
}

/*
1455
 * ext4_xattr_put_super()
1456 1457 1458 1459
 *
 * This is called when a file system is unmounted.
 */
void
1460
ext4_xattr_put_super(struct super_block *sb)
1461 1462 1463 1464 1465
{
	mb_cache_shrink(sb->s_bdev);
}

/*
1466
 * ext4_xattr_cache_insert()
1467 1468 1469 1470 1471 1472 1473
 *
 * 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
1474
ext4_xattr_cache_insert(struct buffer_head *bh)
1475 1476 1477 1478 1479
{
	__u32 hash = le32_to_cpu(BHDR(bh)->h_hash);
	struct mb_cache_entry *ce;
	int error;

1480
	ce = mb_cache_entry_alloc(ext4_xattr_cache, GFP_NOFS);
1481 1482 1483 1484
	if (!ce) {
		ea_bdebug(bh, "out of memory");
		return;
	}
1485
	error = mb_cache_entry_insert(ce, bh->b_bdev, bh->b_blocknr, hash);
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
	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);
	}
}

/*
1499
 * ext4_xattr_cmp()
1500 1501 1502 1503 1504 1505 1506
 *
 * 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
1507 1508
ext4_xattr_cmp(struct ext4_xattr_header *header1,
	       struct ext4_xattr_header *header2)
1509
{
1510
	struct ext4_xattr_entry *entry1, *entry2;
1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529

	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;

1530 1531
		entry1 = EXT4_XATTR_NEXT(entry1);
		entry2 = EXT4_XATTR_NEXT(entry2);
1532 1533 1534 1535 1536 1537 1538
	}
	if (!IS_LAST_ENTRY(entry2))
		return 1;
	return 0;
}

/*
1539
 * ext4_xattr_cache_find()
1540 1541 1542 1543 1544 1545 1546
 *
 * 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 *
1547
ext4_xattr_cache_find(struct inode *inode, struct ext4_xattr_header *header,
1548 1549 1550 1551 1552 1553 1554 1555 1556
		      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:
1557 1558
	ce = mb_cache_entry_find_first(ext4_xattr_cache, inode->i_sb->s_bdev,
				       hash);
1559 1560 1561 1562 1563 1564 1565 1566 1567 1568
	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) {
1569 1570
			EXT4_ERROR_INODE(inode, "block %lu read error",
					 (unsigned long) ce->e_block);
1571
		} else if (le32_to_cpu(BHDR(bh)->h_refcount) >=
1572
				EXT4_XATTR_REFCOUNT_MAX) {
1573 1574 1575
			ea_idebug(inode, "block %lu refcount %d>=%d",
				  (unsigned long) ce->e_block,
				  le32_to_cpu(BHDR(bh)->h_refcount),
1576 1577
					  EXT4_XATTR_REFCOUNT_MAX);
		} else if (ext4_xattr_cmp(header, BHDR(bh)) == 0) {
1578 1579 1580 1581
			*pce = ce;
			return bh;
		}
		brelse(bh);
1582
		ce = mb_cache_entry_find_next(ce, inode->i_sb->s_bdev, hash);
1583 1584 1585 1586 1587 1588 1589 1590
	}
	return NULL;
}

#define NAME_HASH_SHIFT 5
#define VALUE_HASH_SHIFT 16

/*
1591
 * ext4_xattr_hash_entry()
1592 1593 1594
 *
 * Compute the hash of an extended attribute.
 */
1595 1596
static inline void ext4_xattr_hash_entry(struct ext4_xattr_header *header,
					 struct ext4_xattr_entry *entry)
1597 1598 1599 1600 1601
{
	__u32 hash = 0;
	char *name = entry->e_name;
	int n;

1602
	for (n = 0; n < entry->e_name_len; n++) {
1603 1604 1605 1606 1607 1608 1609 1610 1611
		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) +
1612
		     EXT4_XATTR_ROUND) >> EXT4_XATTR_PAD_BITS; n; n--) {
1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626
			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

/*
1627
 * ext4_xattr_rehash()
1628 1629 1630
 *
 * Re-compute the extended attribute hash value after an entry has changed.
 */
1631 1632
static void ext4_xattr_rehash(struct ext4_xattr_header *header,
			      struct ext4_xattr_entry *entry)
1633
{
1634
	struct ext4_xattr_entry *here;
1635 1636
	__u32 hash = 0;

1637
	ext4_xattr_hash_entry(header, entry);
1638 1639 1640 1641 1642 1643 1644 1645 1646 1647
	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);
1648
		here = EXT4_XATTR_NEXT(here);
1649 1650 1651 1652 1653 1654 1655
	}
	header->h_hash = cpu_to_le32(hash);
}

#undef BLOCK_HASH_SHIFT

int __init
1656
ext4_init_xattr(void)
1657
{
1658
	ext4_xattr_cache = mb_cache_create("ext4_xattr", 6);
1659
	if (!ext4_xattr_cache)
1660 1661 1662 1663 1664
		return -ENOMEM;
	return 0;
}

void
1665
ext4_exit_xattr(void)
1666
{
1667 1668 1669
	if (ext4_xattr_cache)
		mb_cache_destroy(ext4_xattr_cache);
	ext4_xattr_cache = NULL;
1670
}