xattr.c 44.5 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
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);
125 126 127
	__u32 csum;
	__le32 save_csum;
	__le64 dsk_block_nr = cpu_to_le64(block_nr);
128

129
	save_csum = hdr->h_checksum;
130
	hdr->h_checksum = 0;
131 132
	csum = ext4_chksum(sbi, sbi->s_csum_seed, (__u8 *)&dsk_block_nr,
			   sizeof(dsk_block_nr));
133 134
	csum = ext4_chksum(sbi, csum, (__u8 *)hdr,
			   EXT4_BLOCK_SIZE(inode->i_sb));
135

136
	hdr->h_checksum = save_csum;
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
	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
170
static inline const struct xattr_handler *
171
ext4_xattr_handler(int name_index)
172
{
S
Stephen Hemminger 已提交
173
	const struct xattr_handler *handler = NULL;
174

175 176
	if (name_index > 0 && name_index < ARRAY_SIZE(ext4_xattr_handler_map))
		handler = ext4_xattr_handler_map[name_index];
177 178 179 180 181 182 183 184 185
	return handler;
}

/*
 * Inode operation listxattr()
 *
 * dentry->d_inode->i_mutex: don't care
 */
ssize_t
186
ext4_listxattr(struct dentry *dentry, char *buffer, size_t size)
187
{
188
	return ext4_xattr_list(dentry, buffer, size);
189 190 191
}

static int
192
ext4_xattr_check_names(struct ext4_xattr_entry *entry, void *end)
193 194
{
	while (!IS_LAST_ENTRY(entry)) {
195
		struct ext4_xattr_entry *next = EXT4_XATTR_NEXT(entry);
196 197 198 199 200 201 202 203
		if ((void *)next >= end)
			return -EIO;
		entry = next;
	}
	return 0;
}

static inline int
204
ext4_xattr_check_block(struct inode *inode, struct buffer_head *bh)
205
{
206 207 208 209 210
	int error;

	if (buffer_verified(bh))
		return 0;

211
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
212 213
	    BHDR(bh)->h_blocks != cpu_to_le32(1))
		return -EIO;
214 215 216 217 218 219
	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;
220 221 222
}

static inline int
223
ext4_xattr_check_entry(struct ext4_xattr_entry *entry, size_t size)
224 225 226 227 228 229 230 231 232 233
{
	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
234
ext4_xattr_find_entry(struct ext4_xattr_entry **pentry, int name_index,
235 236
		      const char *name, size_t size, int sorted)
{
237
	struct ext4_xattr_entry *entry;
238 239 240 241 242 243 244
	size_t name_len;
	int cmp = 1;

	if (name == NULL)
		return -EINVAL;
	name_len = strlen(name);
	entry = *pentry;
245
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
246 247 248 249 250 251 252 253 254
		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;
255
	if (!cmp && ext4_xattr_check_entry(entry, size))
256 257 258 259 260
			return -EIO;
	return cmp ? -ENODATA : 0;
}

static int
261
ext4_xattr_block_get(struct inode *inode, int name_index, const char *name,
262 263 264
		     void *buffer, size_t buffer_size)
{
	struct buffer_head *bh = NULL;
265
	struct ext4_xattr_entry *entry;
266 267 268 269 270 271 272
	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;
273
	if (!EXT4_I(inode)->i_file_acl)
274
		goto cleanup;
275 276
	ea_idebug(inode, "reading block %llu",
		  (unsigned long long)EXT4_I(inode)->i_file_acl);
277
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
278 279 280 281
	if (!bh)
		goto cleanup;
	ea_bdebug(bh, "b_count=%d, refcount=%d",
		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
282
	if (ext4_xattr_check_block(inode, bh)) {
283
bad_block:
284 285
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
286 287 288
		error = -EIO;
		goto cleanup;
	}
289
	ext4_xattr_cache_insert(bh);
290
	entry = BFIRST(bh);
291
	error = ext4_xattr_find_entry(&entry, name_index, name, bh->b_size, 1);
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
	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
311
int
312
ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name,
313 314
		     void *buffer, size_t buffer_size)
{
315 316 317 318
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_entry *entry;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
319 320 321 322
	size_t size;
	void *end;
	int error;

323
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
324
		return -ENODATA;
325
	error = ext4_get_inode_loc(inode, &iloc);
326 327
	if (error)
		return error;
328
	raw_inode = ext4_raw_inode(&iloc);
329 330
	header = IHDR(inode, raw_inode);
	entry = IFIRST(header);
331 332
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(entry, end);
333 334
	if (error)
		goto cleanup;
335
	error = ext4_xattr_find_entry(&entry, name_index, name,
336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354
				      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;
}

/*
355
 * ext4_xattr_get()
356 357 358 359 360 361 362 363 364
 *
 * 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
365
ext4_xattr_get(struct inode *inode, int name_index, const char *name,
366 367 368 369
	       void *buffer, size_t buffer_size)
{
	int error;

370 371
	down_read(&EXT4_I(inode)->xattr_sem);
	error = ext4_xattr_ibody_get(inode, name_index, name, buffer,
372 373
				     buffer_size);
	if (error == -ENODATA)
374
		error = ext4_xattr_block_get(inode, name_index, name, buffer,
375
					     buffer_size);
376
	up_read(&EXT4_I(inode)->xattr_sem);
377 378 379 380
	return error;
}

static int
381
ext4_xattr_list_entries(struct dentry *dentry, struct ext4_xattr_entry *entry,
382 383 384 385
			char *buffer, size_t buffer_size)
{
	size_t rest = buffer_size;

386
	for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) {
S
Stephen Hemminger 已提交
387
		const struct xattr_handler *handler =
388
			ext4_xattr_handler(entry->e_name_index);
389 390

		if (handler) {
391
			size_t size = handler->list(dentry, buffer, rest,
392
						    entry->e_name,
393 394
						    entry->e_name_len,
						    handler->flags);
395 396 397 398 399 400 401 402 403 404 405 406
			if (buffer) {
				if (size > rest)
					return -ERANGE;
				buffer += size;
			}
			rest -= size;
		}
	}
	return buffer_size - rest;
}

static int
407
ext4_xattr_block_list(struct dentry *dentry, char *buffer, size_t buffer_size)
408
{
409
	struct inode *inode = dentry->d_inode;
410 411 412 413 414 415 416
	struct buffer_head *bh = NULL;
	int error;

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

	error = 0;
417
	if (!EXT4_I(inode)->i_file_acl)
418
		goto cleanup;
419 420
	ea_idebug(inode, "reading block %llu",
		  (unsigned long long)EXT4_I(inode)->i_file_acl);
421
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
422 423 424 425 426
	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));
427
	if (ext4_xattr_check_block(inode, bh)) {
428 429
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
430 431 432
		error = -EIO;
		goto cleanup;
	}
433
	ext4_xattr_cache_insert(bh);
434
	error = ext4_xattr_list_entries(dentry, BFIRST(bh), buffer, buffer_size);
435 436 437 438 439 440 441 442

cleanup:
	brelse(bh);

	return error;
}

static int
443
ext4_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size)
444
{
445
	struct inode *inode = dentry->d_inode;
446 447 448
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
	struct ext4_iloc iloc;
449 450 451
	void *end;
	int error;

452
	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
453
		return 0;
454
	error = ext4_get_inode_loc(inode, &iloc);
455 456
	if (error)
		return error;
457
	raw_inode = ext4_raw_inode(&iloc);
458
	header = IHDR(inode, raw_inode);
459 460
	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
	error = ext4_xattr_check_names(IFIRST(header), end);
461 462
	if (error)
		goto cleanup;
463
	error = ext4_xattr_list_entries(dentry, IFIRST(header),
464 465 466 467 468 469 470 471
					buffer, buffer_size);

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

/*
472
 * ext4_xattr_list()
473 474 475 476 477 478 479 480
 *
 * 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.
 */
481
static int
482
ext4_xattr_list(struct dentry *dentry, char *buffer, size_t buffer_size)
483
{
484
	int ret, ret2;
485

486
	down_read(&EXT4_I(dentry->d_inode)->xattr_sem);
487 488 489 490 491 492
	ret = ret2 = ext4_xattr_ibody_list(dentry, buffer, buffer_size);
	if (ret < 0)
		goto errout;
	if (buffer) {
		buffer += ret;
		buffer_size -= ret;
493
	}
494 495 496 497 498
	ret = ext4_xattr_block_list(dentry, buffer, buffer_size);
	if (ret < 0)
		goto errout;
	ret += ret2;
errout:
499
	up_read(&EXT4_I(dentry->d_inode)->xattr_sem);
500
	return ret;
501 502 503
}

/*
504
 * If the EXT4_FEATURE_COMPAT_EXT_ATTR feature of this file system is
505 506
 * not set, set it.
 */
507
static void ext4_xattr_update_super_block(handle_t *handle,
508 509
					  struct super_block *sb)
{
510
	if (EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR))
511 512
		return;

513
	if (ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh) == 0) {
514
		EXT4_SET_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR);
T
Theodore Ts'o 已提交
515
		ext4_handle_dirty_super(handle, sb);
516 517 518 519 520 521 522 523
	}
}

/*
 * Release the xattr block BH: If the reference count is > 1, decrement
 * it; otherwise free the block.
 */
static void
524
ext4_xattr_release_block(handle_t *handle, struct inode *inode,
525 526 527
			 struct buffer_head *bh)
{
	struct mb_cache_entry *ce = NULL;
528
	int error = 0;
529

530
	ce = mb_cache_entry_get(ext4_xattr_cache, bh->b_bdev, bh->b_blocknr);
531 532 533 534 535
	error = ext4_journal_get_write_access(handle, bh);
	if (error)
		goto out;

	lock_buffer(bh);
536 537 538 539 540
	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);
541 542 543
		ext4_free_blocks(handle, inode, bh, 0, 1,
				 EXT4_FREE_BLOCKS_METADATA |
				 EXT4_FREE_BLOCKS_FORGET);
544
		unlock_buffer(bh);
545
	} else {
M
Marcin Slusarz 已提交
546
		le32_add_cpu(&BHDR(bh)->h_refcount, -1);
547 548 549
		if (ce)
			mb_cache_entry_release(ce);
		unlock_buffer(bh);
550
		error = ext4_handle_dirty_xattr_block(handle, inode, bh);
551
		if (IS_SYNC(inode))
552
			ext4_handle_sync(handle);
553
		dquot_free_block(inode, EXT4_C2B(EXT4_SB(inode->i_sb), 1));
554 555
		ea_bdebug(bh, "refcount now=%d; releasing",
			  le32_to_cpu(BHDR(bh)->h_refcount));
556
	}
557 558 559
out:
	ext4_std_error(inode->i_sb, error);
	return;
560 561
}

562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579
/*
 * 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));
}

580
static int
581
ext4_xattr_set_entry(struct ext4_xattr_info *i, struct ext4_xattr_search *s)
582
{
583
	struct ext4_xattr_entry *last;
584 585 586 587
	size_t free, min_offs = s->end - s->base, name_len = strlen(i->name);

	/* Compute min_offs and last. */
	last = s->first;
588
	for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) {
589 590 591 592 593 594 595 596 597 598
		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);
599
			free += EXT4_XATTR_SIZE(size);
600
		}
601
		free += EXT4_XATTR_LEN(name_len);
602 603
	}
	if (i->value) {
604 605 606
		if (free < EXT4_XATTR_SIZE(i->value_len) ||
		    free < EXT4_XATTR_LEN(name_len) +
			   EXT4_XATTR_SIZE(i->value_len))
607 608 609 610 611
			return -ENOSPC;
	}

	if (i->value && s->not_found) {
		/* Insert the new name. */
612
		size_t size = EXT4_XATTR_LEN(name_len);
613 614 615 616 617 618 619 620 621 622 623
		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;
624
			size_t size = EXT4_XATTR_SIZE(
625 626
				le32_to_cpu(s->here->e_value_size));

627
			if (i->value && size == EXT4_XATTR_SIZE(i->value_len)) {
628 629 630 631
				/* The old and the new value have the same
				   size. Just replace. */
				s->here->e_value_size =
					cpu_to_le32(i->value_len);
632 633 634 635 636 637 638 639
				if (i->value == EXT4_ZERO_XATTR_VALUE) {
					memset(val, 0, size);
				} else {
					/* Clear pad bytes first. */
					memset(val + size - EXT4_XATTR_PAD, 0,
					       EXT4_XATTR_PAD);
					memcpy(val, i->value, i->value_len);
				}
640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657
				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);
658
				last = EXT4_XATTR_NEXT(last);
659 660 661 662
			}
		}
		if (!i->value) {
			/* Remove the old name. */
663
			size_t size = EXT4_XATTR_LEN(name_len);
664 665 666 667 668 669 670 671 672 673 674
			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) {
675
			size_t size = EXT4_XATTR_SIZE(i->value_len);
676 677
			void *val = s->base + min_offs - size;
			s->here->e_value_offs = cpu_to_le16(min_offs - size);
678 679 680 681 682 683 684 685
			if (i->value == EXT4_ZERO_XATTR_VALUE) {
				memset(val, 0, size);
			} else {
				/* Clear the pad bytes first. */
				memset(val + size - EXT4_XATTR_PAD, 0,
				       EXT4_XATTR_PAD);
				memcpy(val, i->value, i->value_len);
			}
686 687 688 689 690
		}
	}
	return 0;
}

691 692
struct ext4_xattr_block_find {
	struct ext4_xattr_search s;
693 694 695 696
	struct buffer_head *bh;
};

static int
697 698
ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
		      struct ext4_xattr_block_find *bs)
699 700 701 702 703 704 705
{
	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);

706
	if (EXT4_I(inode)->i_file_acl) {
707
		/* The inode already has an extended attribute block. */
708
		bs->bh = sb_bread(sb, EXT4_I(inode)->i_file_acl);
709 710 711 712 713 714
		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));
715
		if (ext4_xattr_check_block(inode, bs->bh)) {
716 717
			EXT4_ERROR_INODE(inode, "bad block %llu",
					 EXT4_I(inode)->i_file_acl);
718 719 720 721 722 723 724 725
			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;
726
		error = ext4_xattr_find_entry(&bs->s.here, i->name_index,
727 728 729 730 731 732 733 734 735 736 737 738
					      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
739 740 741
ext4_xattr_block_set(handle_t *handle, struct inode *inode,
		     struct ext4_xattr_info *i,
		     struct ext4_xattr_block_find *bs)
742 743 744
{
	struct super_block *sb = inode->i_sb;
	struct buffer_head *new_bh = NULL;
745
	struct ext4_xattr_search *s = &bs->s;
746
	struct mb_cache_entry *ce = NULL;
747
	int error = 0;
748

749
#define header(x) ((struct ext4_xattr_header *)(x))
750 751 752 753

	if (i->value && i->value_len > sb->s_blocksize)
		return -ENOSPC;
	if (s->base) {
754
		ce = mb_cache_entry_get(ext4_xattr_cache, bs->bh->b_bdev,
755
					bs->bh->b_blocknr);
756 757 758 759 760
		error = ext4_journal_get_write_access(handle, bs->bh);
		if (error)
			goto cleanup;
		lock_buffer(bs->bh);

761 762 763 764 765 766
		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");
767
			error = ext4_xattr_set_entry(i, s);
768 769
			if (!error) {
				if (!IS_LAST_ENTRY(s->first))
770
					ext4_xattr_rehash(header(s->base),
771
							  s->here);
772
				ext4_xattr_cache_insert(bs->bh);
773 774 775 776 777
			}
			unlock_buffer(bs->bh);
			if (error == -EIO)
				goto bad_block;
			if (!error)
778 779 780
				error = ext4_handle_dirty_xattr_block(handle,
								      inode,
								      bs->bh);
781 782 783 784 785 786
			if (error)
				goto cleanup;
			goto inserted;
		} else {
			int offset = (char *)s->here - bs->bh->b_data;

787
			unlock_buffer(bs->bh);
788 789 790 791 792
			if (ce) {
				mb_cache_entry_release(ce);
				ce = NULL;
			}
			ea_bdebug(bs->bh, "cloning");
793
			s->base = kmalloc(bs->bh->b_size, GFP_NOFS);
794 795 796 797 798 799 800 801 802 803 804
			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. */
805
		s->base = kzalloc(sb->s_blocksize, GFP_NOFS);
806 807 808 809
		/* assert(header == s->base) */
		error = -ENOMEM;
		if (s->base == NULL)
			goto cleanup;
810
		header(s->base)->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
811 812 813 814 815 816 817
		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;
	}

818
	error = ext4_xattr_set_entry(i, s);
819 820 821 822 823
	if (error == -EIO)
		goto bad_block;
	if (error)
		goto cleanup;
	if (!IS_LAST_ENTRY(s->first))
824
		ext4_xattr_rehash(header(s->base), s->here);
825 826 827

inserted:
	if (!IS_LAST_ENTRY(s->first)) {
828
		new_bh = ext4_xattr_cache_find(inode, header(s->base), &ce);
829 830 831 832 833 834 835
		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. */
836 837
				error = dquot_alloc_block(inode,
						EXT4_C2B(EXT4_SB(sb), 1));
838
				if (error)
839
					goto cleanup;
840
				error = ext4_journal_get_write_access(handle,
841 842 843 844
								      new_bh);
				if (error)
					goto cleanup_dquot;
				lock_buffer(new_bh);
M
Marcin Slusarz 已提交
845
				le32_add_cpu(&BHDR(new_bh)->h_refcount, 1);
846 847 848
				ea_bdebug(new_bh, "reusing; refcount now=%d",
					le32_to_cpu(BHDR(new_bh)->h_refcount));
				unlock_buffer(new_bh);
849 850 851
				error = ext4_handle_dirty_xattr_block(handle,
								      inode,
								      new_bh);
852 853 854 855 856 857 858 859 860 861 862 863
				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 */
864 865 866
			ext4_fsblk_t goal, block;

			goal = ext4_group_first_block_no(sb,
867
						EXT4_I(inode)->i_block_group);
868 869

			/* non-extent files can't have physical blocks past 2^32 */
870
			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
871 872
				goal = goal & EXT4_MAX_BLOCK_FILE_PHYS;

873 874 875 876 877
			/*
			 * 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));
878 879
			block = ext4_new_meta_blocks(handle, inode, goal, 0,
						     NULL, &error);
880
			up_read((&EXT4_I(inode)->i_data_sem));
881 882
			if (error)
				goto cleanup;
883

884
			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
885 886
				BUG_ON(block > EXT4_MAX_BLOCK_FILE_PHYS);

887 888
			ea_idebug(inode, "creating block %llu",
				  (unsigned long long)block);
889 890

			new_bh = sb_getblk(sb, block);
891
			if (unlikely(!new_bh)) {
892
				error = -ENOMEM;
893
getblk_failed:
894
				ext4_free_blocks(handle, inode, NULL, block, 1,
895
						 EXT4_FREE_BLOCKS_METADATA);
896 897 898
				goto cleanup;
			}
			lock_buffer(new_bh);
899
			error = ext4_journal_get_create_access(handle, new_bh);
900 901
			if (error) {
				unlock_buffer(new_bh);
902
				error = -EIO;
903 904 905 906 907
				goto getblk_failed;
			}
			memcpy(new_bh->b_data, s->base, new_bh->b_size);
			set_buffer_uptodate(new_bh);
			unlock_buffer(new_bh);
908
			ext4_xattr_cache_insert(new_bh);
909 910
			error = ext4_handle_dirty_xattr_block(handle,
							      inode, new_bh);
911 912 913 914 915 916
			if (error)
				goto cleanup;
		}
	}

	/* Update the inode. */
917
	EXT4_I(inode)->i_file_acl = new_bh ? new_bh->b_blocknr : 0;
918 919 920

	/* Drop the previous xattr block. */
	if (bs->bh && bs->bh != new_bh)
921
		ext4_xattr_release_block(handle, inode, bs->bh);
922 923 924 925 926 927 928 929 930 931 932 933
	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:
934
	dquot_free_block(inode, EXT4_C2B(EXT4_SB(sb), 1));
935 936 937
	goto cleanup;

bad_block:
938 939
	EXT4_ERROR_INODE(inode, "bad block %llu",
			 EXT4_I(inode)->i_file_acl);
940 941 942 943 944
	goto cleanup;

#undef header
}

T
Tao Ma 已提交
945 946
int ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i,
			  struct ext4_xattr_ibody_find *is)
947
{
948 949
	struct ext4_xattr_ibody_header *header;
	struct ext4_inode *raw_inode;
950 951
	int error;

952
	if (EXT4_I(inode)->i_extra_isize == 0)
953
		return 0;
954
	raw_inode = ext4_raw_inode(&is->iloc);
955 956 957
	header = IHDR(inode, raw_inode);
	is->s.base = is->s.first = IFIRST(header);
	is->s.here = is->s.first;
958
	is->s.end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
959
	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
960
		error = ext4_xattr_check_names(IFIRST(header), is->s.end);
961 962 963
		if (error)
			return error;
		/* Find the named attribute. */
964
		error = ext4_xattr_find_entry(&is->s.here, i->name_index,
965 966 967 968 969 970 971 972 973
					      i->name, is->s.end -
					      (void *)is->s.base, 0);
		if (error && error != -ENODATA)
			return error;
		is->s.not_found = error;
	}
	return 0;
}

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 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014
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)
1015
{
1016 1017
	struct ext4_xattr_ibody_header *header;
	struct ext4_xattr_search *s = &is->s;
1018 1019
	int error;

1020
	if (EXT4_I(inode)->i_extra_isize == 0)
1021
		return -ENOSPC;
1022
	error = ext4_xattr_set_entry(i, s);
1023 1024
	if (error)
		return error;
1025
	header = IHDR(inode, ext4_raw_inode(&is->iloc));
1026
	if (!IS_LAST_ENTRY(s->first)) {
1027
		header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
1028
		ext4_set_inode_state(inode, EXT4_STATE_XATTR);
1029 1030
	} else {
		header->h_magic = cpu_to_le32(0);
1031
		ext4_clear_inode_state(inode, EXT4_STATE_XATTR);
1032 1033 1034 1035 1036
	}
	return 0;
}

/*
1037
 * ext4_xattr_set_handle()
1038
 *
1039
 * Create, replace or remove an extended attribute for this inode.  Value
1040 1041 1042 1043 1044 1045 1046 1047 1048
 * 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
1049
ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
1050 1051 1052
		      const char *name, const void *value, size_t value_len,
		      int flags)
{
1053
	struct ext4_xattr_info i = {
1054 1055 1056 1057 1058 1059
		.name_index = name_index,
		.name = name,
		.value = value,
		.value_len = value_len,

	};
1060
	struct ext4_xattr_ibody_find is = {
1061 1062
		.s = { .not_found = -ENODATA, },
	};
1063
	struct ext4_xattr_block_find bs = {
1064 1065
		.s = { .not_found = -ENODATA, },
	};
K
Kalpak Shah 已提交
1066
	unsigned long no_expand;
1067 1068 1069 1070 1071 1072
	int error;

	if (!name)
		return -EINVAL;
	if (strlen(name) > 255)
		return -ERANGE;
1073
	down_write(&EXT4_I(inode)->xattr_sem);
1074 1075
	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
1076

1077
	error = ext4_reserve_inode_write(handle, inode, &is.iloc);
1078 1079 1080
	if (error)
		goto cleanup;

1081
	if (ext4_test_inode_state(inode, EXT4_STATE_NEW)) {
1082 1083
		struct ext4_inode *raw_inode = ext4_raw_inode(&is.iloc);
		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
1084
		ext4_clear_inode_state(inode, EXT4_STATE_NEW);
1085 1086
	}

1087
	error = ext4_xattr_ibody_find(inode, &i, &is);
1088 1089 1090
	if (error)
		goto cleanup;
	if (is.s.not_found)
1091
		error = ext4_xattr_block_find(inode, &i, &bs);
1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
	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)
1108
			error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1109
		else if (!bs.s.not_found)
1110
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1111
	} else {
1112
		error = ext4_xattr_ibody_set(handle, inode, &i, &is);
1113 1114
		if (!error && !bs.s.not_found) {
			i.value = NULL;
1115
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1116
		} else if (error == -ENOSPC) {
1117 1118 1119 1120 1121
			if (EXT4_I(inode)->i_file_acl && !bs.s.base) {
				error = ext4_xattr_block_find(inode, &i, &bs);
				if (error)
					goto cleanup;
			}
1122
			error = ext4_xattr_block_set(handle, inode, &i, &bs);
1123 1124 1125 1126
			if (error)
				goto cleanup;
			if (!is.s.not_found) {
				i.value = NULL;
1127
				error = ext4_xattr_ibody_set(handle, inode, &i,
1128 1129 1130 1131 1132
							     &is);
			}
		}
	}
	if (!error) {
1133
		ext4_xattr_update_super_block(handle, inode->i_sb);
K
Kalpak Shah 已提交
1134
		inode->i_ctime = ext4_current_time(inode);
1135
		if (!value)
1136
			ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1137
		error = ext4_mark_iloc_dirty(handle, inode, &is.iloc);
1138
		/*
1139
		 * The bh is consumed by ext4_mark_iloc_dirty, even with
1140 1141 1142 1143
		 * error != 0.
		 */
		is.iloc.bh = NULL;
		if (IS_SYNC(inode))
1144
			ext4_handle_sync(handle);
1145 1146 1147 1148 1149
	}

cleanup:
	brelse(is.iloc.bh);
	brelse(bs.bh);
K
Kalpak Shah 已提交
1150
	if (no_expand == 0)
1151
		ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
1152
	up_write(&EXT4_I(inode)->xattr_sem);
1153 1154 1155 1156
	return error;
}

/*
1157
 * ext4_xattr_set()
1158
 *
1159
 * Like ext4_xattr_set_handle, but start from an inode. This extended
1160 1161 1162 1163 1164
 * attribute modification is a filesystem transaction by itself.
 *
 * Returns 0, or a negative error number on failure.
 */
int
1165
ext4_xattr_set(struct inode *inode, int name_index, const char *name,
1166 1167 1168 1169
	       const void *value, size_t value_len, int flags)
{
	handle_t *handle;
	int error, retries = 0;
1170
	int credits = ext4_jbd2_credits_xattr(inode);
1171 1172

retry:
1173
	handle = ext4_journal_start(inode, EXT4_HT_XATTR, credits);
1174 1175 1176 1177 1178
	if (IS_ERR(handle)) {
		error = PTR_ERR(handle);
	} else {
		int error2;

1179
		error = ext4_xattr_set_handle(handle, inode, name_index, name,
1180
					      value, value_len, flags);
1181
		error2 = ext4_journal_stop(handle);
1182
		if (error == -ENOSPC &&
1183
		    ext4_should_retry_alloc(inode->i_sb, &retries))
1184 1185 1186 1187 1188 1189 1190 1191
			goto retry;
		if (error == 0)
			error = error2;
	}

	return error;
}

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 1228 1229 1230 1231 1232 1233
/*
 * 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
1234
	int s_min_extra_isize = le16_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_min_extra_isize);
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278

	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;
1279
		if (ext4_xattr_check_block(inode, bh)) {
1280 1281
			EXT4_ERROR_INODE(inode, "bad block %llu",
					 EXT4_I(inode)->i_file_acl);
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
			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;
D
Dave Jones 已提交
1353 1354
					kfree(is); is = NULL;
					kfree(bs); bs = NULL;
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 1380 1381 1382 1383 1384 1385 1386 1387
					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);
1388 1389
		if (error)
			goto cleanup;
1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408

		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
1409
		i.value_len = size;
1410 1411 1412 1413 1414 1415 1416 1417 1418 1419
		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);
1420 1421
		b_entry_name = NULL;
		buffer = NULL;
1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443
		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;
}



1444
/*
1445
 * ext4_xattr_delete_inode()
1446 1447 1448 1449 1450 1451
 *
 * 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
1452
ext4_xattr_delete_inode(handle_t *handle, struct inode *inode)
1453 1454 1455
{
	struct buffer_head *bh = NULL;

1456
	if (!EXT4_I(inode)->i_file_acl)
1457
		goto cleanup;
1458
	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
1459
	if (!bh) {
1460 1461
		EXT4_ERROR_INODE(inode, "block %llu read error",
				 EXT4_I(inode)->i_file_acl);
1462 1463
		goto cleanup;
	}
1464
	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
1465
	    BHDR(bh)->h_blocks != cpu_to_le32(1)) {
1466 1467
		EXT4_ERROR_INODE(inode, "bad block %llu",
				 EXT4_I(inode)->i_file_acl);
1468 1469
		goto cleanup;
	}
1470 1471
	ext4_xattr_release_block(handle, inode, bh);
	EXT4_I(inode)->i_file_acl = 0;
1472 1473 1474 1475 1476 1477

cleanup:
	brelse(bh);
}

/*
1478
 * ext4_xattr_put_super()
1479 1480 1481 1482
 *
 * This is called when a file system is unmounted.
 */
void
1483
ext4_xattr_put_super(struct super_block *sb)
1484 1485 1486 1487 1488
{
	mb_cache_shrink(sb->s_bdev);
}

/*
1489
 * ext4_xattr_cache_insert()
1490 1491 1492 1493 1494 1495 1496
 *
 * 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
1497
ext4_xattr_cache_insert(struct buffer_head *bh)
1498 1499 1500 1501 1502
{
	__u32 hash = le32_to_cpu(BHDR(bh)->h_hash);
	struct mb_cache_entry *ce;
	int error;

1503
	ce = mb_cache_entry_alloc(ext4_xattr_cache, GFP_NOFS);
1504 1505 1506 1507
	if (!ce) {
		ea_bdebug(bh, "out of memory");
		return;
	}
1508
	error = mb_cache_entry_insert(ce, bh->b_bdev, bh->b_blocknr, hash);
1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521
	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);
	}
}

/*
1522
 * ext4_xattr_cmp()
1523 1524 1525 1526 1527 1528 1529
 *
 * 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
1530 1531
ext4_xattr_cmp(struct ext4_xattr_header *header1,
	       struct ext4_xattr_header *header2)
1532
{
1533
	struct ext4_xattr_entry *entry1, *entry2;
1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552

	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;

1553 1554
		entry1 = EXT4_XATTR_NEXT(entry1);
		entry2 = EXT4_XATTR_NEXT(entry2);
1555 1556 1557 1558 1559 1560 1561
	}
	if (!IS_LAST_ENTRY(entry2))
		return 1;
	return 0;
}

/*
1562
 * ext4_xattr_cache_find()
1563 1564 1565 1566 1567 1568 1569
 *
 * 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 *
1570
ext4_xattr_cache_find(struct inode *inode, struct ext4_xattr_header *header,
1571 1572 1573 1574 1575 1576 1577 1578 1579
		      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:
1580 1581
	ce = mb_cache_entry_find_first(ext4_xattr_cache, inode->i_sb->s_bdev,
				       hash);
1582 1583 1584 1585 1586 1587 1588 1589 1590 1591
	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) {
1592 1593
			EXT4_ERROR_INODE(inode, "block %lu read error",
					 (unsigned long) ce->e_block);
1594
		} else if (le32_to_cpu(BHDR(bh)->h_refcount) >=
1595
				EXT4_XATTR_REFCOUNT_MAX) {
1596 1597 1598
			ea_idebug(inode, "block %lu refcount %d>=%d",
				  (unsigned long) ce->e_block,
				  le32_to_cpu(BHDR(bh)->h_refcount),
1599 1600
					  EXT4_XATTR_REFCOUNT_MAX);
		} else if (ext4_xattr_cmp(header, BHDR(bh)) == 0) {
1601 1602 1603 1604
			*pce = ce;
			return bh;
		}
		brelse(bh);
1605
		ce = mb_cache_entry_find_next(ce, inode->i_sb->s_bdev, hash);
1606 1607 1608 1609 1610 1611 1612 1613
	}
	return NULL;
}

#define NAME_HASH_SHIFT 5
#define VALUE_HASH_SHIFT 16

/*
1614
 * ext4_xattr_hash_entry()
1615 1616 1617
 *
 * Compute the hash of an extended attribute.
 */
1618 1619
static inline void ext4_xattr_hash_entry(struct ext4_xattr_header *header,
					 struct ext4_xattr_entry *entry)
1620 1621 1622 1623 1624
{
	__u32 hash = 0;
	char *name = entry->e_name;
	int n;

1625
	for (n = 0; n < entry->e_name_len; n++) {
1626 1627 1628 1629 1630 1631 1632 1633 1634
		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) +
1635
		     EXT4_XATTR_ROUND) >> EXT4_XATTR_PAD_BITS; n; n--) {
1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649
			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

/*
1650
 * ext4_xattr_rehash()
1651 1652 1653
 *
 * Re-compute the extended attribute hash value after an entry has changed.
 */
1654 1655
static void ext4_xattr_rehash(struct ext4_xattr_header *header,
			      struct ext4_xattr_entry *entry)
1656
{
1657
	struct ext4_xattr_entry *here;
1658 1659
	__u32 hash = 0;

1660
	ext4_xattr_hash_entry(header, entry);
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670
	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);
1671
		here = EXT4_XATTR_NEXT(here);
1672 1673 1674 1675 1676 1677 1678
	}
	header->h_hash = cpu_to_le32(hash);
}

#undef BLOCK_HASH_SHIFT

int __init
1679
ext4_init_xattr(void)
1680
{
1681
	ext4_xattr_cache = mb_cache_create("ext4_xattr", 6);
1682
	if (!ext4_xattr_cache)
1683 1684 1685 1686 1687
		return -ENOMEM;
	return 0;
}

void
1688
ext4_exit_xattr(void)
1689
{
1690 1691 1692
	if (ext4_xattr_cache)
		mb_cache_destroy(ext4_xattr_cache);
	ext4_xattr_cache = NULL;
1693
}