ext4_jbd2.h 15.5 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0+
2
/*
3
 * ext4_jbd2.h
4 5 6 7 8 9 10 11
 *
 * Written by Stephen C. Tweedie <sct@redhat.com>, 1999
 *
 * Copyright 1998--1999 Red Hat corp --- All Rights Reserved
 *
 * Ext4-specific journaling extensions.
 */

12 13
#ifndef _EXT4_JBD2_H
#define _EXT4_JBD2_H
14 15

#include <linux/fs.h>
16
#include <linux/jbd2.h>
17
#include "ext4.h"
18 19 20 21 22 23 24 25

#define EXT4_JOURNAL(inode)	(EXT4_SB((inode)->i_sb)->s_journal)

/* Define the number of blocks we need to account to a transaction to
 * modify one block of data.
 *
 * We may have to touch one inode, one bitmap buffer, up to three
 * indirection blocks, the group and superblock summaries, and the data
A
Alex Tomas 已提交
26 27
 * block to complete the transaction.
 *
28
 * For extents-enabled fs we may have to allocate and modify up to
29 30 31
 * 5 levels of tree, data block (for each of these we need bitmap + group
 * summaries), root which is stored in the inode, sb
 */
32

A
Alex Tomas 已提交
33
#define EXT4_SINGLEDATA_TRANS_BLOCKS(sb)				\
34
	(ext4_has_feature_extents(sb) ? 20U : 8U)
35 36 37 38 39 40 41 42 43 44 45 46 47

/* Extended attribute operations touch at most two data buffers,
 * two bitmap buffers, and two group summaries, in addition to the inode
 * and the superblock, which are already accounted for. */

#define EXT4_XATTR_TRANS_BLOCKS		6U

/* Define the minimum size for a transaction which modifies data.  This
 * needs to take into account the fact that we may end up modifying two
 * quota files too (one for the group, one for the user quota).  The
 * superblock only gets updated once, of course, so don't bother
 * counting that again for the quota updates. */

A
Alex Tomas 已提交
48
#define EXT4_DATA_TRANS_BLOCKS(sb)	(EXT4_SINGLEDATA_TRANS_BLOCKS(sb) + \
49
					 EXT4_XATTR_TRANS_BLOCKS - 2 + \
D
Dmitry Monakhov 已提交
50
					 EXT4_MAXQUOTAS_TRANS_BLOCKS(sb))
51

52 53 54 55 56 57
/*
 * Define the number of metadata blocks we need to account to modify data.
 *
 * This include super block, inode block, quota blocks and xattr blocks
 */
#define EXT4_META_TRANS_BLOCKS(sb)	(EXT4_XATTR_TRANS_BLOCKS + \
D
Dmitry Monakhov 已提交
58
					EXT4_MAXQUOTAS_TRANS_BLOCKS(sb))
59

60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76
/* Define an arbitrary limit for the amount of data we will anticipate
 * writing to any given transaction.  For unbounded transactions such as
 * write(2) and truncate(2) we can write more than this, but we always
 * start off at the maximum transaction size and grow the transaction
 * optimistically as we go. */

#define EXT4_MAX_TRANS_DATA		64U

/* We break up a large truncate or write transaction once the handle's
 * buffer credits gets this low, we need either to extend the
 * transaction or to start a new one.  Reserve enough space here for
 * inode, bitmap, superblock, group and indirection updates for at least
 * one block, plus two quota updates.  Quota allocations are not
 * needed. */

#define EXT4_RESERVE_TRANS_BLOCKS	12U

A
Artem Blagodarenko 已提交
77 78 79 80 81 82 83 84
/*
 * Number of credits needed if we need to insert an entry into a
 * directory.  For each new index block, we need 4 blocks (old index
 * block, new index block, bitmap block, bg summary).  For normal
 * htree directories there are 2 levels; if the largedir feature
 * enabled it's 3 levels.
 */
#define EXT4_INDEX_EXTRA_TRANS_BLOCKS	12U
85 86 87

#ifdef CONFIG_QUOTA
/* Amount of blocks needed for quota update - we know that the structure was
88
 * allocated so we need to update only data block */
89
#define EXT4_QUOTA_TRANS_BLOCKS(sb) ((test_opt(sb, QUOTA) ||\
90
		ext4_has_feature_quota(sb)) ? 1 : 0)
91 92
/* Amount of blocks needed for quota insert/delete - we do some block writes
 * but inode, sb and group updates are done only once */
93
#define EXT4_QUOTA_INIT_BLOCKS(sb) ((test_opt(sb, QUOTA) ||\
94
		ext4_has_feature_quota(sb)) ?\
95 96 97 98
		(DQUOT_INIT_ALLOC*(EXT4_SINGLEDATA_TRANS_BLOCKS(sb)-3)\
		 +3+DQUOT_INIT_REWRITE) : 0)

#define EXT4_QUOTA_DEL_BLOCKS(sb) ((test_opt(sb, QUOTA) ||\
99
		ext4_has_feature_quota(sb)) ?\
100 101
		(DQUOT_DEL_ALLOC*(EXT4_SINGLEDATA_TRANS_BLOCKS(sb)-3)\
		 +3+DQUOT_DEL_REWRITE) : 0)
102 103 104 105 106
#else
#define EXT4_QUOTA_TRANS_BLOCKS(sb) 0
#define EXT4_QUOTA_INIT_BLOCKS(sb) 0
#define EXT4_QUOTA_DEL_BLOCKS(sb) 0
#endif
J
Jan Kara 已提交
107 108 109
#define EXT4_MAXQUOTAS_TRANS_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_TRANS_BLOCKS(sb))
#define EXT4_MAXQUOTAS_INIT_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_INIT_BLOCKS(sb))
#define EXT4_MAXQUOTAS_DEL_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_DEL_BLOCKS(sb))
110

111 112 113 114 115 116 117 118 119 120 121 122 123 124
/*
 * Ext4 handle operation types -- for logging purposes
 */
#define EXT4_HT_MISC             0
#define EXT4_HT_INODE            1
#define EXT4_HT_WRITE_PAGE       2
#define EXT4_HT_MAP_BLOCKS       3
#define EXT4_HT_DIR              4
#define EXT4_HT_TRUNCATE         5
#define EXT4_HT_QUOTA            6
#define EXT4_HT_RESIZE           7
#define EXT4_HT_MIGRATE          8
#define EXT4_HT_MOVE_EXTENTS     9
#define EXT4_HT_XATTR           10
125 126
#define EXT4_HT_EXT_CONVERT     11
#define EXT4_HT_MAX             12
127

B
Bobi Jam 已提交
128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
/**
 *   struct ext4_journal_cb_entry - Base structure for callback information.
 *
 *   This struct is a 'seed' structure for a using with your own callback
 *   structs. If you are using callbacks you must allocate one of these
 *   or another struct of your own definition which has this struct
 *   as it's first element and pass it to ext4_journal_callback_add().
 */
struct ext4_journal_cb_entry {
	/* list information for other callbacks attached to the same handle */
	struct list_head jce_list;

	/*  Function to call with this callback structure */
	void (*jce_func)(struct super_block *sb,
			 struct ext4_journal_cb_entry *jce, int error);

	/* user data goes here */
};

/**
 * ext4_journal_callback_add: add a function to call after transaction commit
 * @handle: active journal transaction handle to register callback on
 * @func: callback function to call after the transaction has committed:
 *        @sb: superblock of current filesystem for transaction
 *        @jce: returned journal callback data
 *        @rc: journal state at commit (0 = transaction committed properly)
 * @jce: journal callback data (internal and function private data struct)
 *
 * The registered function will be called in the context of the journal thread
 * after the transaction for which the handle was created has completed.
 *
 * No locks are held when the callback function is called, so it is safe to
 * call blocking functions from within the callback, but the callback should
 * not block or run for too long, or the filesystem will be blocked waiting for
 * the next transaction to commit. No journaling functions can be used, or
 * there is a risk of deadlock.
 *
 * There is no guaranteed calling order of multiple registered callbacks on
 * the same transaction.
 */
168 169 170 171 172 173 174
static inline void _ext4_journal_callback_add(handle_t *handle,
			struct ext4_journal_cb_entry *jce)
{
	/* Add the jce to transaction's private list */
	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
}

B
Bobi Jam 已提交
175 176 177 178 179 180 181 182 183 184 185 186
static inline void ext4_journal_callback_add(handle_t *handle,
			void (*func)(struct super_block *sb,
				     struct ext4_journal_cb_entry *jce,
				     int rc),
			struct ext4_journal_cb_entry *jce)
{
	struct ext4_sb_info *sbi =
			EXT4_SB(handle->h_transaction->t_journal->j_private);

	/* Add the jce to transaction's private list */
	jce->jce_func = func;
	spin_lock(&sbi->s_md_lock);
187
	_ext4_journal_callback_add(handle, jce);
B
Bobi Jam 已提交
188 189 190
	spin_unlock(&sbi->s_md_lock);
}

191

B
Bobi Jam 已提交
192 193 194 195
/**
 * ext4_journal_callback_del: delete a registered callback
 * @handle: active journal transaction handle on which callback was registered
 * @jce: registered journal callback entry to unregister
196
 * Return true if object was successfully removed
B
Bobi Jam 已提交
197
 */
198
static inline bool ext4_journal_callback_try_del(handle_t *handle,
B
Bobi Jam 已提交
199 200
					     struct ext4_journal_cb_entry *jce)
{
201
	bool deleted;
B
Bobi Jam 已提交
202 203 204 205
	struct ext4_sb_info *sbi =
			EXT4_SB(handle->h_transaction->t_journal->j_private);

	spin_lock(&sbi->s_md_lock);
206
	deleted = !list_empty(&jce->jce_list);
B
Bobi Jam 已提交
207 208
	list_del_init(&jce->jce_list);
	spin_unlock(&sbi->s_md_lock);
209
	return deleted;
B
Bobi Jam 已提交
210 211
}

212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
int
ext4_mark_iloc_dirty(handle_t *handle,
		     struct inode *inode,
		     struct ext4_iloc *iloc);

/*
 * On success, We end up with an outstanding reference count against
 * iloc->bh.  This _must_ be cleaned up later.
 */

int ext4_reserve_inode_write(handle_t *handle, struct inode *inode,
			struct ext4_iloc *iloc);

int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode);

227 228 229
int ext4_expand_extra_isize(struct inode *inode,
			    unsigned int new_extra_isize,
			    struct ext4_iloc *iloc);
230
/*
231
 * Wrapper functions with which ext4 calls into JBD.
232
 */
233 234
int __ext4_journal_get_write_access(const char *where, unsigned int line,
				    handle_t *handle, struct buffer_head *bh);
235

236 237 238
int __ext4_forget(const char *where, unsigned int line, handle_t *handle,
		  int is_metadata, struct inode *inode,
		  struct buffer_head *bh, ext4_fsblk_t blocknr);
239

240
int __ext4_journal_get_create_access(const char *where, unsigned int line,
241
				handle_t *handle, struct buffer_head *bh);
242

243 244 245
int __ext4_handle_dirty_metadata(const char *where, unsigned int line,
				 handle_t *handle, struct inode *inode,
				 struct buffer_head *bh);
246

247
int __ext4_handle_dirty_super(const char *where, unsigned int line,
248
			      handle_t *handle, struct super_block *sb);
T
Theodore Ts'o 已提交
249

250
#define ext4_journal_get_write_access(handle, bh) \
251
	__ext4_journal_get_write_access(__func__, __LINE__, (handle), (bh))
252
#define ext4_forget(handle, is_metadata, inode, bh, block_nr) \
253 254
	__ext4_forget(__func__, __LINE__, (handle), (is_metadata), (inode), \
		      (bh), (block_nr))
255
#define ext4_journal_get_create_access(handle, bh) \
256
	__ext4_journal_get_create_access(__func__, __LINE__, (handle), (bh))
257
#define ext4_handle_dirty_metadata(handle, inode, bh) \
258 259
	__ext4_handle_dirty_metadata(__func__, __LINE__, (handle), (inode), \
				     (bh))
T
Theodore Ts'o 已提交
260
#define ext4_handle_dirty_super(handle, sb) \
261
	__ext4_handle_dirty_super(__func__, __LINE__, (handle), (sb))
262

263
handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line,
264
				  int type, int blocks, int rsv_blocks);
265
int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle);
266

267
#define EXT4_NOJOURNAL_MAX_REF_COUNT ((unsigned long) 4096)
268

269 270
/* Note:  Do not use this for NULL handles.  This is only to determine if
 * a properly allocated handle is using a journal or not. */
271 272
static inline int ext4_handle_valid(handle_t *handle)
{
273
	if ((unsigned long)handle < EXT4_NOJOURNAL_MAX_REF_COUNT)
274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297
		return 0;
	return 1;
}

static inline void ext4_handle_sync(handle_t *handle)
{
	if (ext4_handle_valid(handle))
		handle->h_sync = 1;
}

static inline int ext4_handle_is_aborted(handle_t *handle)
{
	if (ext4_handle_valid(handle))
		return is_handle_aborted(handle);
	return 0;
}

static inline int ext4_handle_has_enough_credits(handle_t *handle, int needed)
{
	if (ext4_handle_valid(handle) && handle->h_buffer_credits < needed)
		return 0;
	return 1;
}

298
#define ext4_journal_start_sb(sb, type, nblocks)			\
299
	__ext4_journal_start_sb((sb), __LINE__, (type), (nblocks), 0)
300 301

#define ext4_journal_start(inode, type, nblocks)			\
302 303 304 305
	__ext4_journal_start((inode), __LINE__, (type), (nblocks), 0)

#define ext4_journal_start_with_reserve(inode, type, blocks, rsv_blocks) \
	__ext4_journal_start((inode), __LINE__, (type), (blocks), (rsv_blocks))
306 307 308

static inline handle_t *__ext4_journal_start(struct inode *inode,
					     unsigned int line, int type,
309
					     int blocks, int rsv_blocks)
310
{
311 312
	return __ext4_journal_start_sb(inode->i_sb, line, type, blocks,
				       rsv_blocks);
313 314 315
}

#define ext4_journal_stop(handle) \
316
	__ext4_journal_stop(__func__, __LINE__, (handle))
317

318
#define ext4_journal_start_reserved(handle, type) \
319 320 321 322 323 324 325 326 327 328 329
	__ext4_journal_start_reserved((handle), __LINE__, (type))

handle_t *__ext4_journal_start_reserved(handle_t *handle, unsigned int line,
					int type);

static inline void ext4_journal_free_reserved(handle_t *handle)
{
	if (ext4_handle_valid(handle))
		jbd2_journal_free_reserved(handle);
}

330 331 332 333 334 335 336
static inline handle_t *ext4_journal_current_handle(void)
{
	return journal_current_handle();
}

static inline int ext4_journal_extend(handle_t *handle, int nblocks)
{
337 338 339
	if (ext4_handle_valid(handle))
		return jbd2_journal_extend(handle, nblocks);
	return 0;
340 341 342 343
}

static inline int ext4_journal_restart(handle_t *handle, int nblocks)
{
344 345 346
	if (ext4_handle_valid(handle))
		return jbd2_journal_restart(handle, nblocks);
	return 0;
347 348 349 350
}

static inline int ext4_journal_blocks_per_page(struct inode *inode)
{
351 352 353
	if (EXT4_JOURNAL(inode) != NULL)
		return jbd2_journal_blocks_per_page(inode);
	return 0;
354 355 356 357
}

static inline int ext4_journal_force_commit(journal_t *journal)
{
358 359 360
	if (journal)
		return jbd2_journal_force_commit(journal);
	return 0;
361 362
}

363
static inline int ext4_jbd2_inode_add_write(handle_t *handle,
364
		struct inode *inode, loff_t start_byte, loff_t length)
365
{
366
	if (ext4_handle_valid(handle))
367 368
		return jbd2_journal_inode_ranged_write(handle,
				EXT4_I(inode)->jinode, start_byte, length);
369
	return 0;
370 371
}

372
static inline int ext4_jbd2_inode_add_wait(handle_t *handle,
373
		struct inode *inode, loff_t start_byte, loff_t length)
374 375
{
	if (ext4_handle_valid(handle))
376 377
		return jbd2_journal_inode_ranged_wait(handle,
				EXT4_I(inode)->jinode, start_byte, length);
378 379 380
	return 0;
}

381 382 383 384 385 386
static inline void ext4_update_inode_fsync_trans(handle_t *handle,
						 struct inode *inode,
						 int datasync)
{
	struct ext4_inode_info *ei = EXT4_I(inode);

387
	if (ext4_handle_valid(handle) && !is_handle_aborted(handle)) {
388 389 390 391 392 393
		ei->i_sync_tid = handle->h_transaction->t_tid;
		if (datasync)
			ei->i_datasync_tid = handle->h_transaction->t_tid;
	}
}

394 395 396
/* super.c */
int ext4_force_commit(struct super_block *sb);

397 398 399 400 401 402 403 404
/*
 * Ext4 inode journal modes
 */
#define EXT4_INODE_JOURNAL_DATA_MODE	0x01 /* journal data mode */
#define EXT4_INODE_ORDERED_DATA_MODE	0x02 /* ordered data mode */
#define EXT4_INODE_WRITEBACK_DATA_MODE	0x04 /* writeback data mode */

static inline int ext4_inode_journal_mode(struct inode *inode)
405
{
406
	if (EXT4_JOURNAL(inode) == NULL)
407 408 409
		return EXT4_INODE_WRITEBACK_DATA_MODE;	/* writeback */
	/* We do not support data journalling with delayed allocation */
	if (!S_ISREG(inode->i_mode) ||
410 411 412 413 414 415
	    test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA ||
	    (ext4_test_inode_flag(inode, EXT4_INODE_JOURNAL_DATA) &&
	    !test_opt(inode->i_sb, DELALLOC))) {
		/* We do not support data journalling for encrypted data */
		if (S_ISREG(inode->i_mode) && ext4_encrypted_inode(inode))
			return EXT4_INODE_ORDERED_DATA_MODE;  /* ordered */
416
		return EXT4_INODE_JOURNAL_DATA_MODE;	/* journal data */
417
	}
418 419 420 421
	if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA)
		return EXT4_INODE_ORDERED_DATA_MODE;	/* ordered */
	if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)
		return EXT4_INODE_WRITEBACK_DATA_MODE;	/* writeback */
422
	BUG();
423 424 425 426 427
}

static inline int ext4_should_journal_data(struct inode *inode)
{
	return ext4_inode_journal_mode(inode) & EXT4_INODE_JOURNAL_DATA_MODE;
428 429 430 431
}

static inline int ext4_should_order_data(struct inode *inode)
{
432
	return ext4_inode_journal_mode(inode) & EXT4_INODE_ORDERED_DATA_MODE;
433 434 435 436
}

static inline int ext4_should_writeback_data(struct inode *inode)
{
437
	return ext4_inode_journal_mode(inode) & EXT4_INODE_WRITEBACK_DATA_MODE;
438 439
}

440 441 442 443
/*
 * This function controls whether or not we should try to go down the
 * dioread_nolock code paths, which makes it safe to avoid taking
 * i_mutex for direct I/O reads.  This only works for extent-based
444 445 446 447
 * files, and it doesn't work if data journaling is enabled, since the
 * dioread_nolock code uses b_private to pass information back to the
 * I/O completion handler, and this conflicts with the jbd's use of
 * b_private.
448 449 450 451 452 453 454
 */
static inline int ext4_should_dioread_nolock(struct inode *inode)
{
	if (!test_opt(inode->i_sb, DIOREAD_NOLOCK))
		return 0;
	if (!S_ISREG(inode->i_mode))
		return 0;
455
	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
456 457 458
		return 0;
	if (ext4_should_journal_data(inode))
		return 0;
459 460 461
	/* temporary fix to prevent generic/422 test failures */
	if (!test_opt(inode->i_sb, DELALLOC))
		return 0;
462 463 464
	return 1;
}

465
#endif	/* _EXT4_JBD2_H */