xattr.c 44.3 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
#ifdef EXT4_XATTR_DEBUG
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
# 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
80 81
# define ea_idebug(inode, fmt, ...)	no_printk(fmt, ##__VA_ARGS__)
# define ea_bdebug(bh, fmt, ...)	no_printk(fmt, ##__VA_ARGS__)
82 83
#endif

84 85 86
static void ext4_xattr_cache_insert(struct buffer_head *);
static struct buffer_head *ext4_xattr_cache_find(struct inode *,
						 struct ext4_xattr_header *,
87
						 struct mb_cache_entry **);
88 89
static void ext4_xattr_rehash(struct ext4_xattr_header *,
			      struct ext4_xattr_entry *);
90
static int ext4_xattr_list(struct dentry *dentry, char *buffer,
91
			   size_t buffer_size);
92

93
static struct mb_cache *ext4_xattr_cache;
94

S
Stephen Hemminger 已提交
95
static const struct xattr_handler *ext4_xattr_handler_map[] = {
96
	[EXT4_XATTR_INDEX_USER]		     = &ext4_xattr_user_handler,
T
Theodore Ts'o 已提交
97
#ifdef CONFIG_EXT4_FS_POSIX_ACL
98 99
	[EXT4_XATTR_INDEX_POSIX_ACL_ACCESS]  = &ext4_xattr_acl_access_handler,
	[EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT] = &ext4_xattr_acl_default_handler,
100
#endif
101
	[EXT4_XATTR_INDEX_TRUSTED]	     = &ext4_xattr_trusted_handler,
T
Theodore Ts'o 已提交
102
#ifdef CONFIG_EXT4_FS_SECURITY
103
	[EXT4_XATTR_INDEX_SECURITY]	     = &ext4_xattr_security_handler,
104 105 106
#endif
};

S
Stephen Hemminger 已提交
107
const struct xattr_handler *ext4_xattr_handlers[] = {
108 109
	&ext4_xattr_user_handler,
	&ext4_xattr_trusted_handler,
T
Theodore Ts'o 已提交
110
#ifdef CONFIG_EXT4_FS_POSIX_ACL
111 112
	&ext4_xattr_acl_access_handler,
	&ext4_xattr_acl_default_handler,
113
#endif
T
Theodore Ts'o 已提交
114
#ifdef CONFIG_EXT4_FS_SECURITY
115
	&ext4_xattr_security_handler,
116 117 118 119
#endif
	NULL
};

120 121 122 123 124 125 126 127 128
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);
	__u32 csum, old;

	old = hdr->h_checksum;
	hdr->h_checksum = 0;
129 130 131
	block_nr = cpu_to_le64(block_nr);
	csum = ext4_chksum(sbi, sbi->s_csum_seed, (__u8 *)&block_nr,
			   sizeof(block_nr));
132 133
	csum = ext4_chksum(sbi, csum, (__u8 *)hdr,
			   EXT4_BLOCK_SIZE(inode->i_sb));
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
	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
169
static inline const struct xattr_handler *
170
ext4_xattr_handler(int name_index)
171
{
S
Stephen Hemminger 已提交
172
	const struct xattr_handler *handler = NULL;
173

174 175
	if (name_index > 0 && name_index < ARRAY_SIZE(ext4_xattr_handler_map))
		handler = ext4_xattr_handler_map[name_index];
176 177 178 179 180 181 182 183 184
	return handler;
}

/*
 * Inode operation listxattr()
 *
 * dentry->d_inode->i_mutex: don't care
 */
ssize_t
185
ext4_listxattr(struct dentry *dentry, char *buffer, size_t size)
186
{
187
	return ext4_xattr_list(dentry, buffer, size);
188 189 190
}

static int
191
ext4_xattr_check_names(struct ext4_xattr_entry *entry, void *end)
192 193
{
	while (!IS_LAST_ENTRY(entry)) {
194
		struct ext4_xattr_entry *next = EXT4_XATTR_NEXT(entry);
195 196 197 198 199 200 201 202
		if ((void *)next >= end)
			return -EIO;
		entry = next;
	}
	return 0;
}

static inline int
203
ext4_xattr_check_block(struct inode *inode, struct buffer_head *bh)
204
{
205 206 207 208 209
	int error;

	if (buffer_verified(bh))
		return 0;

210
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
211 212
	    BHDR(bh)->h_blocks != cpu_to_le32(1))
		return -EIO;
213 214 215 216 217 218
	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;
219 220 221
}

static inline int
222
ext4_xattr_check_entry(struct ext4_xattr_entry *entry, size_t size)
223 224 225 226 227 228 229 230 231 232
{
	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
233
ext4_xattr_find_entry(struct ext4_xattr_entry **pentry, int name_index,
234 235
		      const char *name, size_t size, int sorted)
{
236
	struct ext4_xattr_entry *entry;
237 238 239 240 241 242 243
	size_t name_len;
	int cmp = 1;

	if (name == NULL)
		return -EINVAL;
	name_len = strlen(name);
	entry = *pentry;
244
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
245 246 247 248 249 250 251 252 253
		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;
254
	if (!cmp && ext4_xattr_check_entry(entry, size))
255 256 257 258 259
			return -EIO;
	return cmp ? -ENODATA : 0;
}

static int
260
ext4_xattr_block_get(struct inode *inode, int name_index, const char *name,
261 262 263
		     void *buffer, size_t buffer_size)
{
	struct buffer_head *bh = NULL;
264
	struct ext4_xattr_entry *entry;
265 266 267 268 269 270 271
	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;
272
	if (!EXT4_I(inode)->i_file_acl)
273
		goto cleanup;
274 275
	ea_idebug(inode, "reading block %llu",
		  (unsigned long long)EXT4_I(inode)->i_file_acl);
276
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
277 278 279 280
	if (!bh)
		goto cleanup;
	ea_bdebug(bh, "b_count=%d, refcount=%d",
		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
281
	if (ext4_xattr_check_block(inode, bh)) {
282
bad_block:
283 284
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
285 286 287
		error = -EIO;
		goto cleanup;
	}
288
	ext4_xattr_cache_insert(bh);
289
	entry = BFIRST(bh);
290
	error = ext4_xattr_find_entry(&entry, name_index, name, bh->b_size, 1);
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309
	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;
}

T
Tao Ma 已提交
310
int
311
ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name,
312 313
		     void *buffer, size_t buffer_size)
{
314 315 316 317
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_entry *entry;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
318 319 320 321
	size_t size;
	void *end;
	int error;

322
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
323
		return -ENODATA;
324
	error = ext4_get_inode_loc(inode, &iloc);
325 326
	if (error)
		return error;
327
	raw_inode = ext4_raw_inode(&iloc);
328 329
	header = IHDR(inode, raw_inode);
	entry = IFIRST(header);
330 331
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(entry, end);
332 333
	if (error)
		goto cleanup;
334
	error = ext4_xattr_find_entry(&entry, name_index, name,
335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353
				      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;
}

/*
354
 * ext4_xattr_get()
355 356 357 358 359 360 361 362 363
 *
 * 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
364
ext4_xattr_get(struct inode *inode, int name_index, const char *name,
365 366 367 368
	       void *buffer, size_t buffer_size)
{
	int error;

369 370
	down_read(&EXT4_I(inode)->xattr_sem);
	error = ext4_xattr_ibody_get(inode, name_index, name, buffer,
371 372
				     buffer_size);
	if (error == -ENODATA)
373
		error = ext4_xattr_block_get(inode, name_index, name, buffer,
374
					     buffer_size);
375
	up_read(&EXT4_I(inode)->xattr_sem);
376 377 378 379
	return error;
}

static int
380
ext4_xattr_list_entries(struct dentry *dentry, struct ext4_xattr_entry *entry,
381 382 383 384
			char *buffer, size_t buffer_size)
{
	size_t rest = buffer_size;

385
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
S
Stephen Hemminger 已提交
386
		const struct xattr_handler *handler =
387
			ext4_xattr_handler(entry->e_name_index);
388 389

		if (handler) {
390
			size_t size = handler->list(dentry, buffer, rest,
391
						    entry->e_name,
392 393
						    entry->e_name_len,
						    handler->flags);
394 395 396 397 398 399 400 401 402 403 404 405
			if (buffer) {
				if (size > rest)
					return -ERANGE;
				buffer += size;
			}
			rest -= size;
		}
	}
	return buffer_size - rest;
}

static int
406
ext4_xattr_block_list(struct dentry *dentry, char *buffer, size_t buffer_size)
407
{
408
	struct inode *inode = dentry->d_inode;
409 410 411 412 413 414 415
	struct buffer_head *bh = NULL;
	int error;

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

	error = 0;
416
	if (!EXT4_I(inode)->i_file_acl)
417
		goto cleanup;
418 419
	ea_idebug(inode, "reading block %llu",
		  (unsigned long long)EXT4_I(inode)->i_file_acl);
420
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
421 422 423 424 425
	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));
426
	if (ext4_xattr_check_block(inode, bh)) {
427 428
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
429 430 431
		error = -EIO;
		goto cleanup;
	}
432
	ext4_xattr_cache_insert(bh);
433
	error = ext4_xattr_list_entries(dentry, BFIRST(bh), buffer, buffer_size);
434 435 436 437 438 439 440 441

cleanup:
	brelse(bh);

	return error;
}

static int
442
ext4_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size)
443
{
444
	struct inode *inode = dentry->d_inode;
445 446 447
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
448 449 450
	void *end;
	int error;

451
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
452
		return 0;
453
	error = ext4_get_inode_loc(inode, &iloc);
454 455
	if (error)
		return error;
456
	raw_inode = ext4_raw_inode(&iloc);
457
	header = IHDR(inode, raw_inode);
458 459
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(IFIRST(header), end);
460 461
	if (error)
		goto cleanup;
462
	error = ext4_xattr_list_entries(dentry, IFIRST(header),
463 464 465 466 467 468 469 470
					buffer, buffer_size);

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

/*
471
 * ext4_xattr_list()
472 473 474 475 476 477 478 479
 *
 * 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.
 */
480
static int
481
ext4_xattr_list(struct dentry *dentry, char *buffer, size_t buffer_size)
482
{
483
	int ret, ret2;
484

485
	down_read(&EXT4_I(dentry->d_inode)->xattr_sem);
486 487 488 489 490 491
	ret = ret2 = ext4_xattr_ibody_list(dentry, buffer, buffer_size);
	if (ret < 0)
		goto errout;
	if (buffer) {
		buffer += ret;
		buffer_size -= ret;
492
	}
493 494 495 496 497
	ret = ext4_xattr_block_list(dentry, buffer, buffer_size);
	if (ret < 0)
		goto errout;
	ret += ret2;
errout:
498
	up_read(&EXT4_I(dentry->d_inode)->xattr_sem);
499
	return ret;
500 501 502
}

/*
503
 * If the EXT4_FEATURE_COMPAT_EXT_ATTR feature of this file system is
504 505
 * not set, set it.
 */
506
static void ext4_xattr_update_super_block(handle_t *handle,
507 508
					  struct super_block *sb)
{
509
	if (EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR))
510 511
		return;

512
	if (ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh) == 0) {
513
		EXT4_SET_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR);
T
Theodore Ts'o 已提交
514
		ext4_handle_dirty_super(handle, sb);
515 516 517 518 519 520 521 522
	}
}

/*
 * Release the xattr block BH: If the reference count is > 1, decrement
 * it; otherwise free the block.
 */
static void
523
ext4_xattr_release_block(handle_t *handle, struct inode *inode,
524 525 526
			 struct buffer_head *bh)
{
	struct mb_cache_entry *ce = NULL;
527
	int error = 0;
528

529
	ce = mb_cache_entry_get(ext4_xattr_cache, bh->b_bdev, bh->b_blocknr);
530 531 532 533 534
	error = ext4_journal_get_write_access(handle, bh);
	if (error)
		goto out;

	lock_buffer(bh);
535 536 537 538 539
	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);
540 541 542
		ext4_free_blocks(handle, inode, bh, 0, 1,
				 EXT4_FREE_BLOCKS_METADATA |
				 EXT4_FREE_BLOCKS_FORGET);
543
		unlock_buffer(bh);
544
	} else {
M
Marcin Slusarz 已提交
545
		le32_add_cpu(&BHDR(bh)->h_refcount, -1);
546 547 548
		if (ce)
			mb_cache_entry_release(ce);
		unlock_buffer(bh);
549
		error = ext4_handle_dirty_xattr_block(handle, inode, bh);
550
		if (IS_SYNC(inode))
551
			ext4_handle_sync(handle);
552
		dquot_free_block(inode, 1);
553 554
		ea_bdebug(bh, "refcount now=%d; releasing",
			  le32_to_cpu(BHDR(bh)->h_refcount));
555
	}
556 557 558
out:
	ext4_std_error(inode->i_sb, error);
	return;
559 560
}

561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578
/*
 * 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));
}

579
static int
580
ext4_xattr_set_entry(struct ext4_xattr_info *i, struct ext4_xattr_search *s)
581
{
582
	struct ext4_xattr_entry *last;
583 584 585 586
	size_t free, min_offs = s->end - s->base, name_len = strlen(i->name);

	/* Compute min_offs and last. */
	last = s->first;
587
	for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) {
588 589 590 591 592 593 594 595 596 597
		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);
598
			free += EXT4_XATTR_SIZE(size);
599
		}
600
		free += EXT4_XATTR_LEN(name_len);
601 602
	}
	if (i->value) {
603 604 605
		if (free < EXT4_XATTR_SIZE(i->value_len) ||
		    free < EXT4_XATTR_LEN(name_len) +
			   EXT4_XATTR_SIZE(i->value_len))
606 607 608 609 610
			return -ENOSPC;
	}

	if (i->value && s->not_found) {
		/* Insert the new name. */
611
		size_t size = EXT4_XATTR_LEN(name_len);
612 613 614 615 616 617 618 619 620 621 622
		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;
623
			size_t size = EXT4_XATTR_SIZE(
624 625
				le32_to_cpu(s->here->e_value_size));

626
			if (i->value && size == EXT4_XATTR_SIZE(i->value_len)) {
627 628 629 630
				/* The old and the new value have the same
				   size. Just replace. */
				s->here->e_value_size =
					cpu_to_le32(i->value_len);
631 632
				memset(val + size - EXT4_XATTR_PAD, 0,
				       EXT4_XATTR_PAD); /* Clear pad bytes. */
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651
				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);
652
				last = EXT4_XATTR_NEXT(last);
653 654 655 656
			}
		}
		if (!i->value) {
			/* Remove the old name. */
657
			size_t size = EXT4_XATTR_LEN(name_len);
658 659 660 661 662 663 664 665 666 667 668
			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) {
669
			size_t size = EXT4_XATTR_SIZE(i->value_len);
670 671
			void *val = s->base + min_offs - size;
			s->here->e_value_offs = cpu_to_le16(min_offs - size);
672 673
			memset(val + size - EXT4_XATTR_PAD, 0,
			       EXT4_XATTR_PAD); /* Clear the pad bytes. */
674 675 676 677 678 679
			memcpy(val, i->value, i->value_len);
		}
	}
	return 0;
}

680 681
struct ext4_xattr_block_find {
	struct ext4_xattr_search s;
682 683 684 685
	struct buffer_head *bh;
};

static int
686 687
ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
		      struct ext4_xattr_block_find *bs)
688 689 690 691 692 693 694
{
	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);

695
	if (EXT4_I(inode)->i_file_acl) {
696
		/* The inode already has an extended attribute block. */
697
		bs->bh = sb_bread(sb, EXT4_I(inode)->i_file_acl);
698 699 700 701 702 703
		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));
704
		if (ext4_xattr_check_block(inode, bs->bh)) {
705 706
			EXT4_ERROR_INODE(inode, "bad block %llu",
					 EXT4_I(inode)->i_file_acl);
707 708 709 710 711 712 713 714
			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;
715
		error = ext4_xattr_find_entry(&bs->s.here, i->name_index,
716 717 718 719 720 721 722 723 724 725 726 727
					      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
728 729 730
ext4_xattr_block_set(handle_t *handle, struct inode *inode,
		     struct ext4_xattr_info *i,
		     struct ext4_xattr_block_find *bs)
731 732 733
{
	struct super_block *sb = inode->i_sb;
	struct buffer_head *new_bh = NULL;
734
	struct ext4_xattr_search *s = &bs->s;
735
	struct mb_cache_entry *ce = NULL;
736
	int error = 0;
737

738
#define header(x) ((struct ext4_xattr_header *)(x))
739 740 741 742

	if (i->value && i->value_len > sb->s_blocksize)
		return -ENOSPC;
	if (s->base) {
743
		ce = mb_cache_entry_get(ext4_xattr_cache, bs->bh->b_bdev,
744
					bs->bh->b_blocknr);
745 746 747 748 749
		error = ext4_journal_get_write_access(handle, bs->bh);
		if (error)
			goto cleanup;
		lock_buffer(bs->bh);

750 751 752 753 754 755
		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");
756
			error = ext4_xattr_set_entry(i, s);
757 758
			if (!error) {
				if (!IS_LAST_ENTRY(s->first))
759
					ext4_xattr_rehash(header(s->base),
760
							  s->here);
761
				ext4_xattr_cache_insert(bs->bh);
762 763 764 765 766
			}
			unlock_buffer(bs->bh);
			if (error == -EIO)
				goto bad_block;
			if (!error)
767 768 769
				error = ext4_handle_dirty_xattr_block(handle,
								      inode,
								      bs->bh);
770 771 772 773 774 775
			if (error)
				goto cleanup;
			goto inserted;
		} else {
			int offset = (char *)s->here - bs->bh->b_data;

776
			unlock_buffer(bs->bh);
777 778 779 780 781
			if (ce) {
				mb_cache_entry_release(ce);
				ce = NULL;
			}
			ea_bdebug(bs->bh, "cloning");
782
			s->base = kmalloc(bs->bh->b_size, GFP_NOFS);
783 784 785 786 787 788 789 790 791 792 793
			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. */
794
		s->base = kzalloc(sb->s_blocksize, GFP_NOFS);
795 796 797 798
		/* assert(header == s->base) */
		error = -ENOMEM;
		if (s->base == NULL)
			goto cleanup;
799
		header(s->base)->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
800 801 802 803 804 805 806
		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;
	}

807
	error = ext4_xattr_set_entry(i, s);
808 809 810 811 812
	if (error == -EIO)
		goto bad_block;
	if (error)
		goto cleanup;
	if (!IS_LAST_ENTRY(s->first))
813
		ext4_xattr_rehash(header(s->base), s->here);
814 815 816

inserted:
	if (!IS_LAST_ENTRY(s->first)) {
817
		new_bh = ext4_xattr_cache_find(inode, header(s->base), &ce);
818 819 820 821 822 823 824
		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. */
825 826
				error = dquot_alloc_block(inode, 1);
				if (error)
827
					goto cleanup;
828
				error = ext4_journal_get_write_access(handle,
829 830 831 832
								      new_bh);
				if (error)
					goto cleanup_dquot;
				lock_buffer(new_bh);
M
Marcin Slusarz 已提交
833
				le32_add_cpu(&BHDR(new_bh)->h_refcount, 1);
834 835 836
				ea_bdebug(new_bh, "reusing; refcount now=%d",
					le32_to_cpu(BHDR(new_bh)->h_refcount));
				unlock_buffer(new_bh);
837 838 839
				error = ext4_handle_dirty_xattr_block(handle,
								      inode,
								      new_bh);
840 841 842 843 844 845 846 847 848 849 850 851
				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 */
852 853 854
			ext4_fsblk_t goal, block;

			goal = ext4_group_first_block_no(sb,
855
						EXT4_I(inode)->i_block_group);
856 857

			/* non-extent files can't have physical blocks past 2^32 */
858
			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
859 860
				goal = goal & EXT4_MAX_BLOCK_FILE_PHYS;

861 862 863 864 865
			/*
			 * 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));
866 867
			block = ext4_new_meta_blocks(handle, inode, goal, 0,
						     NULL, &error);
868
			up_read((&EXT4_I(inode)->i_data_sem));
869 870
			if (error)
				goto cleanup;
871

872
			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
873 874
				BUG_ON(block > EXT4_MAX_BLOCK_FILE_PHYS);

875 876
			ea_idebug(inode, "creating block %llu",
				  (unsigned long long)block);
877 878 879 880

			new_bh = sb_getblk(sb, block);
			if (!new_bh) {
getblk_failed:
881
				ext4_free_blocks(handle, inode, NULL, block, 1,
882
						 EXT4_FREE_BLOCKS_METADATA);
883 884 885 886
				error = -EIO;
				goto cleanup;
			}
			lock_buffer(new_bh);
887
			error = ext4_journal_get_create_access(handle, new_bh);
888 889 890 891 892 893 894
			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);
895
			ext4_xattr_cache_insert(new_bh);
896 897
			error = ext4_handle_dirty_xattr_block(handle,
							      inode, new_bh);
898 899 900 901 902 903
			if (error)
				goto cleanup;
		}
	}

	/* Update the inode. */
904
	EXT4_I(inode)->i_file_acl = new_bh ? new_bh->b_blocknr : 0;
905 906 907

	/* Drop the previous xattr block. */
	if (bs->bh && bs->bh != new_bh)
908
		ext4_xattr_release_block(handle, inode, bs->bh);
909 910 911 912 913 914 915 916 917 918 919 920
	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:
921
	dquot_free_block(inode, 1);
922 923 924
	goto cleanup;

bad_block:
925 926
	EXT4_ERROR_INODE(inode, "bad block %llu",
			 EXT4_I(inode)->i_file_acl);
927 928 929 930 931
	goto cleanup;

#undef header
}

T
Tao Ma 已提交
932 933
int ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i,
			  struct ext4_xattr_ibody_find *is)
934
{
935 936
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
937 938
	int error;

939
	if (EXT4_I(inode)->i_extra_isize == 0)
940
		return 0;
941
	raw_inode = ext4_raw_inode(&is->iloc);
942 943 944
	header = IHDR(inode, raw_inode);
	is->s.base = is->s.first = IFIRST(header);
	is->s.here = is->s.first;
945
	is->s.end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
946
	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
947
		error = ext4_xattr_check_names(IFIRST(header), is->s.end);
948 949 950
		if (error)
			return error;
		/* Find the named attribute. */
951
		error = ext4_xattr_find_entry(&is->s.here, i->name_index,
952 953 954 955 956 957 958 959 960
					      i->name, is->s.end -
					      (void *)is->s.base, 0);
		if (error && error != -ENODATA)
			return error;
		is->s.not_found = error;
	}
	return 0;
}

961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
int ext4_xattr_ibody_inline_set(handle_t *handle, struct inode *inode,
				struct ext4_xattr_info *i,
				struct ext4_xattr_ibody_find *is)
{
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_search *s = &is->s;
	int error;

	if (EXT4_I(inode)->i_extra_isize == 0)
		return -ENOSPC;
	error = ext4_xattr_set_entry(i, s);
	if (error) {
		if (error == -ENOSPC &&
		    ext4_has_inline_data(inode)) {
			error = ext4_try_to_evict_inline_data(handle, inode,
					EXT4_XATTR_LEN(strlen(i->name) +
					EXT4_XATTR_SIZE(i->value_len)));
			if (error)
				return error;
			error = ext4_xattr_ibody_find(inode, i, is);
			if (error)
				return error;
			error = ext4_xattr_set_entry(i, s);
		}
		if (error)
			return error;
	}
	header = IHDR(inode, ext4_raw_inode(&is->iloc));
	if (!IS_LAST_ENTRY(s->first)) {
		header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
		ext4_set_inode_state(inode, EXT4_STATE_XATTR);
	} else {
		header->h_magic = cpu_to_le32(0);
		ext4_clear_inode_state(inode, EXT4_STATE_XATTR);
	}
	return 0;
}

static int ext4_xattr_ibody_set(handle_t *handle, struct inode *inode,
				struct ext4_xattr_info *i,
				struct ext4_xattr_ibody_find *is)
1002
{
1003 1004
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_search *s = &is->s;
1005 1006
	int error;

1007
	if (EXT4_I(inode)->i_extra_isize == 0)
1008
		return -ENOSPC;
1009
	error = ext4_xattr_set_entry(i, s);
1010 1011
	if (error)
		return error;
1012
	header = IHDR(inode, ext4_raw_inode(&is->iloc));
1013
	if (!IS_LAST_ENTRY(s->first)) {
1014
		header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
1015
		ext4_set_inode_state(inode, EXT4_STATE_XATTR);
1016 1017
	} else {
		header->h_magic = cpu_to_le32(0);
1018
		ext4_clear_inode_state(inode, EXT4_STATE_XATTR);
1019 1020 1021 1022 1023
	}
	return 0;
}

/*
1024
 * ext4_xattr_set_handle()
1025
 *
1026
 * Create, replace or remove an extended attribute for this inode.  Value
1027 1028 1029 1030 1031 1032 1033 1034 1035
 * 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
1036
ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
1037 1038 1039
		      const char *name, const void *value, size_t value_len,
		      int flags)
{
1040
	struct ext4_xattr_info i = {
1041 1042 1043 1044 1045 1046
		.name_index = name_index,
		.name = name,
		.value = value,
		.value_len = value_len,

	};
1047
	struct ext4_xattr_ibody_find is = {
1048 1049
		.s = { .not_found = -ENODATA, },
	};
1050
	struct ext4_xattr_block_find bs = {
1051 1052
		.s = { .not_found = -ENODATA, },
	};
K
Kalpak Shah 已提交
1053
	unsigned long no_expand;
1054 1055 1056 1057 1058 1059
	int error;

	if (!name)
		return -EINVAL;
	if (strlen(name) > 255)
		return -ERANGE;
1060
	down_write(&EXT4_I(inode)->xattr_sem);
1061 1062
	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
1063

1064
	error = ext4_reserve_inode_write(handle, inode, &is.iloc);
1065 1066 1067
	if (error)
		goto cleanup;

1068
	if (ext4_test_inode_state(inode, EXT4_STATE_NEW)) {
1069 1070
		struct ext4_inode *raw_inode = ext4_raw_inode(&is.iloc);
		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
1071
		ext4_clear_inode_state(inode, EXT4_STATE_NEW);
1072 1073
	}

1074
	error = ext4_xattr_ibody_find(inode, &i, &is);
1075 1076 1077
	if (error)
		goto cleanup;
	if (is.s.not_found)
1078
		error = ext4_xattr_block_find(inode, &i, &bs);
1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094
	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)
1095
			error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1096
		else if (!bs.s.not_found)
1097
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1098
	} else {
1099
		error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1100 1101
		if (!error && !bs.s.not_found) {
			i.value = NULL;
1102
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1103
		} else if (error == -ENOSPC) {
1104 1105 1106 1107 1108
			if (EXT4_I(inode)->i_file_acl && !bs.s.base) {
				error = ext4_xattr_block_find(inode, &i, &bs);
				if (error)
					goto cleanup;
			}
1109
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1110 1111 1112 1113
			if (error)
				goto cleanup;
			if (!is.s.not_found) {
				i.value = NULL;
1114
				error = ext4_xattr_ibody_set(handle, inode, &i,
1115 1116 1117 1118 1119
							     &is);
			}
		}
	}
	if (!error) {
1120
		ext4_xattr_update_super_block(handle, inode->i_sb);
K
Kalpak Shah 已提交
1121
		inode->i_ctime = ext4_current_time(inode);
1122
		if (!value)
1123
			ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1124
		error = ext4_mark_iloc_dirty(handle, inode, &is.iloc);
1125
		/*
1126
		 * The bh is consumed by ext4_mark_iloc_dirty, even with
1127 1128 1129 1130
		 * error != 0.
		 */
		is.iloc.bh = NULL;
		if (IS_SYNC(inode))
1131
			ext4_handle_sync(handle);
1132 1133 1134 1135 1136
	}

cleanup:
	brelse(is.iloc.bh);
	brelse(bs.bh);
K
Kalpak Shah 已提交
1137
	if (no_expand == 0)
1138
		ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1139
	up_write(&EXT4_I(inode)->xattr_sem);
1140 1141 1142 1143
	return error;
}

/*
1144
 * ext4_xattr_set()
1145
 *
1146
 * Like ext4_xattr_set_handle, but start from an inode. This extended
1147 1148 1149 1150 1151
 * attribute modification is a filesystem transaction by itself.
 *
 * Returns 0, or a negative error number on failure.
 */
int
1152
ext4_xattr_set(struct inode *inode, int name_index, const char *name,
1153 1154 1155 1156
	       const void *value, size_t value_len, int flags)
{
	handle_t *handle;
	int error, retries = 0;
1157
	int credits = EXT4_DATA_TRANS_BLOCKS(inode->i_sb);
1158 1159

retry:
1160 1161 1162 1163 1164 1165 1166 1167
	/*
	 * In case of inline data, we may push out the data to a block,
	 * So reserve the journal space first.
	 */
	if (ext4_has_inline_data(inode))
		credits += ext4_writepage_trans_blocks(inode) + 1;

	handle = ext4_journal_start(inode, credits);
1168 1169 1170 1171 1172
	if (IS_ERR(handle)) {
		error = PTR_ERR(handle);
	} else {
		int error2;

1173
		error = ext4_xattr_set_handle(handle, inode, name_index, name,
1174
					      value, value_len, flags);
1175
		error2 = ext4_journal_stop(handle);
1176
		if (error == -ENOSPC &&
1177
		    ext4_should_retry_alloc(inode->i_sb, &retries))
1178 1179 1180 1181 1182 1183 1184 1185
			goto retry;
		if (error == 0)
			error = error2;
	}

	return error;
}

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 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227
/*
 * 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
1228
	int s_min_extra_isize = le16_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_min_extra_isize);
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

	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;
1273
		if (ext4_xattr_check_block(inode, bh)) {
1274 1275
			EXT4_ERROR_INODE(inode, "bad block %llu",
					 EXT4_I(inode)->i_file_acl);
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 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379
			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);
1380 1381
		if (error)
			goto cleanup;
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400

		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
1401
		i.value_len = size;
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411
		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);
1412 1413
		b_entry_name = NULL;
		buffer = NULL;
1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435
		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;
}



1436
/*
1437
 * ext4_xattr_delete_inode()
1438 1439 1440 1441 1442 1443
 *
 * 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
1444
ext4_xattr_delete_inode(handle_t *handle, struct inode *inode)
1445 1446 1447
{
	struct buffer_head *bh = NULL;

1448
	if (!EXT4_I(inode)->i_file_acl)
1449
		goto cleanup;
1450
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
1451
	if (!bh) {
1452 1453
		EXT4_ERROR_INODE(inode, "block %llu read error",
				 EXT4_I(inode)->i_file_acl);
1454 1455
		goto cleanup;
	}
1456
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
1457
	    BHDR(bh)->h_blocks != cpu_to_le32(1)) {
1458 1459
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
1460 1461
		goto cleanup;
	}
1462 1463
	ext4_xattr_release_block(handle, inode, bh);
	EXT4_I(inode)->i_file_acl = 0;
1464 1465 1466 1467 1468 1469

cleanup:
	brelse(bh);
}

/*
1470
 * ext4_xattr_put_super()
1471 1472 1473 1474
 *
 * This is called when a file system is unmounted.
 */
void
1475
ext4_xattr_put_super(struct super_block *sb)
1476 1477 1478 1479 1480
{
	mb_cache_shrink(sb->s_bdev);
}

/*
1481
 * ext4_xattr_cache_insert()
1482 1483 1484 1485 1486 1487 1488
 *
 * 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
1489
ext4_xattr_cache_insert(struct buffer_head *bh)
1490 1491 1492 1493 1494
{
	__u32 hash = le32_to_cpu(BHDR(bh)->h_hash);
	struct mb_cache_entry *ce;
	int error;

1495
	ce = mb_cache_entry_alloc(ext4_xattr_cache, GFP_NOFS);
1496 1497 1498 1499
	if (!ce) {
		ea_bdebug(bh, "out of memory");
		return;
	}
1500
	error = mb_cache_entry_insert(ce, bh->b_bdev, bh->b_blocknr, hash);
1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513
	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);
	}
}

/*
1514
 * ext4_xattr_cmp()
1515 1516 1517 1518 1519 1520 1521
 *
 * 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
1522 1523
ext4_xattr_cmp(struct ext4_xattr_header *header1,
	       struct ext4_xattr_header *header2)
1524
{
1525
	struct ext4_xattr_entry *entry1, *entry2;
1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544

	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;

1545 1546
		entry1 = EXT4_XATTR_NEXT(entry1);
		entry2 = EXT4_XATTR_NEXT(entry2);
1547 1548 1549 1550 1551 1552 1553
	}
	if (!IS_LAST_ENTRY(entry2))
		return 1;
	return 0;
}

/*
1554
 * ext4_xattr_cache_find()
1555 1556 1557 1558 1559 1560 1561
 *
 * 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 *
1562
ext4_xattr_cache_find(struct inode *inode, struct ext4_xattr_header *header,
1563 1564 1565 1566 1567 1568 1569 1570 1571
		      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:
1572 1573
	ce = mb_cache_entry_find_first(ext4_xattr_cache, inode->i_sb->s_bdev,
				       hash);
1574 1575 1576 1577 1578 1579 1580 1581 1582 1583
	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) {
1584 1585
			EXT4_ERROR_INODE(inode, "block %lu read error",
					 (unsigned long) ce->e_block);
1586
		} else if (le32_to_cpu(BHDR(bh)->h_refcount) >=
1587
				EXT4_XATTR_REFCOUNT_MAX) {
1588 1589 1590
			ea_idebug(inode, "block %lu refcount %d>=%d",
				  (unsigned long) ce->e_block,
				  le32_to_cpu(BHDR(bh)->h_refcount),
1591 1592
					  EXT4_XATTR_REFCOUNT_MAX);
		} else if (ext4_xattr_cmp(header, BHDR(bh)) == 0) {
1593 1594 1595 1596
			*pce = ce;
			return bh;
		}
		brelse(bh);
1597
		ce = mb_cache_entry_find_next(ce, inode->i_sb->s_bdev, hash);
1598 1599 1600 1601 1602 1603 1604 1605
	}
	return NULL;
}

#define NAME_HASH_SHIFT 5
#define VALUE_HASH_SHIFT 16

/*
1606
 * ext4_xattr_hash_entry()
1607 1608 1609
 *
 * Compute the hash of an extended attribute.
 */
1610 1611
static inline void ext4_xattr_hash_entry(struct ext4_xattr_header *header,
					 struct ext4_xattr_entry *entry)
1612 1613 1614 1615 1616
{
	__u32 hash = 0;
	char *name = entry->e_name;
	int n;

1617
	for (n = 0; n < entry->e_name_len; n++) {
1618 1619 1620 1621 1622 1623 1624 1625 1626
		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) +
1627
		     EXT4_XATTR_ROUND) >> EXT4_XATTR_PAD_BITS; n; n--) {
1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641
			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

/*
1642
 * ext4_xattr_rehash()
1643 1644 1645
 *
 * Re-compute the extended attribute hash value after an entry has changed.
 */
1646 1647
static void ext4_xattr_rehash(struct ext4_xattr_header *header,
			      struct ext4_xattr_entry *entry)
1648
{
1649
	struct ext4_xattr_entry *here;
1650 1651
	__u32 hash = 0;

1652
	ext4_xattr_hash_entry(header, entry);
1653 1654 1655 1656 1657 1658 1659 1660 1661 1662
	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);
1663
		here = EXT4_XATTR_NEXT(here);
1664 1665 1666 1667 1668 1669 1670
	}
	header->h_hash = cpu_to_le32(hash);
}

#undef BLOCK_HASH_SHIFT

int __init
1671
ext4_init_xattr(void)
1672
{
1673
	ext4_xattr_cache = mb_cache_create("ext4_xattr", 6);
1674
	if (!ext4_xattr_cache)
1675 1676 1677 1678 1679
		return -ENOMEM;
	return 0;
}

void
1680
ext4_exit_xattr(void)
1681
{
1682 1683 1684
	if (ext4_xattr_cache)
		mb_cache_destroy(ext4_xattr_cache);
	ext4_xattr_cache = NULL;
1685
}